From f6e6e78bebe005149309d71cd4cc54ced7ca378e Mon Sep 17 00:00:00 2001 From: ston1th Date: Tue, 25 Jun 2019 20:57:03 +0200 Subject: [PATCH] initial commit --- .gitignore | 1 + LICENSE | 24 + Makefile | 39 + go.mod | 8 + go.sum | 11 + main.go | 84 + pkg/crypto/crypto.go | 71 + pkg/crypto/crypto_test.go | 38 + .../ston1th/godrop/v2/.gitignore | 3 + .../git.giftfish.de/ston1th/godrop/v2/LICENSE | 24 + .../ston1th/godrop/v2/README.md | 5 + .../git.giftfish.de/ston1th/godrop/v2/go.mod | 3 + .../git.giftfish.de/ston1th/godrop/v2/go.sum | 2 + .../ston1th/godrop/v2/godrop.go | 136 + .../ston1th/godrop/v2/lookup.go | 34 + .../ston1th/godrop/v2/pledge.go | 14 + .../ston1th/godrop/v2/pledge_openbsd.go | 43 + .../ston1th/godrop/v2/unveil.go | 11 + .../ston1th/godrop/v2/unveil_openbsd.go | 32 + vendor/golang.org/x/crypto/AUTHORS | 3 + vendor/golang.org/x/crypto/CONTRIBUTORS | 3 + vendor/golang.org/x/crypto/LICENSE | 27 + vendor/golang.org/x/crypto/PATENTS | 22 + vendor/golang.org/x/crypto/argon2/argon2.go | 285 + vendor/golang.org/x/crypto/argon2/blake2b.go | 53 + .../x/crypto/argon2/blamka_amd64.go | 60 + .../golang.org/x/crypto/argon2/blamka_amd64.s | 243 + .../x/crypto/argon2/blamka_generic.go | 163 + .../golang.org/x/crypto/argon2/blamka_ref.go | 15 + vendor/golang.org/x/crypto/blake2b/blake2b.go | 289 + .../x/crypto/blake2b/blake2bAVX2_amd64.go | 37 + .../x/crypto/blake2b/blake2bAVX2_amd64.s | 750 ++ .../x/crypto/blake2b/blake2b_amd64.go | 24 + .../x/crypto/blake2b/blake2b_amd64.s | 281 + .../x/crypto/blake2b/blake2b_generic.go | 182 + .../x/crypto/blake2b/blake2b_ref.go | 11 + vendor/golang.org/x/crypto/blake2b/blake2x.go | 177 + .../golang.org/x/crypto/blake2b/register.go | 32 + .../chacha20poly1305/chacha20poly1305.go | 101 + .../chacha20poly1305_amd64.go | 86 + .../chacha20poly1305/chacha20poly1305_amd64.s | 2695 +++++++ .../chacha20poly1305_generic.go | 81 + .../chacha20poly1305_noasm.go | 15 + .../chacha20poly1305/xchacha20poly1305.go | 104 + .../x/crypto/internal/chacha20/asm_arm64.s | 308 + .../x/crypto/internal/chacha20/asm_ppc64le.s | 668 ++ .../crypto/internal/chacha20/chacha_arm64.go | 31 + .../internal/chacha20/chacha_generic.go | 264 + .../crypto/internal/chacha20/chacha_noasm.go | 16 + .../internal/chacha20/chacha_ppc64le.go | 51 + .../crypto/internal/chacha20/chacha_s390x.go | 29 + .../x/crypto/internal/chacha20/chacha_s390x.s | 260 + .../x/crypto/internal/chacha20/xor.go | 43 + .../x/crypto/internal/subtle/aliasing.go | 32 + .../internal/subtle/aliasing_appengine.go | 35 + .../golang.org/x/crypto/poly1305/mac_noasm.go | 11 + .../golang.org/x/crypto/poly1305/poly1305.go | 83 + .../golang.org/x/crypto/poly1305/sum_amd64.go | 68 + .../golang.org/x/crypto/poly1305/sum_amd64.s | 148 + .../golang.org/x/crypto/poly1305/sum_arm.go | 22 + vendor/golang.org/x/crypto/poly1305/sum_arm.s | 427 + .../x/crypto/poly1305/sum_generic.go | 172 + .../golang.org/x/crypto/poly1305/sum_noasm.go | 16 + .../x/crypto/poly1305/sum_ppc64le.go | 68 + .../x/crypto/poly1305/sum_ppc64le.s | 247 + .../golang.org/x/crypto/poly1305/sum_s390x.go | 42 + .../golang.org/x/crypto/poly1305/sum_s390x.s | 378 + .../x/crypto/poly1305/sum_vmsl_s390x.s | 909 +++ .../x/crypto/ssh/terminal/terminal.go | 966 +++ .../golang.org/x/crypto/ssh/terminal/util.go | 114 + .../x/crypto/ssh/terminal/util_aix.go | 12 + .../x/crypto/ssh/terminal/util_bsd.go | 12 + .../x/crypto/ssh/terminal/util_linux.go | 10 + .../x/crypto/ssh/terminal/util_plan9.go | 58 + .../x/crypto/ssh/terminal/util_solaris.go | 124 + .../x/crypto/ssh/terminal/util_windows.go | 105 + vendor/golang.org/x/sys/AUTHORS | 3 + vendor/golang.org/x/sys/CONTRIBUTORS | 3 + vendor/golang.org/x/sys/LICENSE | 27 + vendor/golang.org/x/sys/PATENTS | 22 + vendor/golang.org/x/sys/cpu/byteorder.go | 30 + vendor/golang.org/x/sys/cpu/cpu.go | 126 + vendor/golang.org/x/sys/cpu/cpu_aix_ppc64.go | 30 + vendor/golang.org/x/sys/cpu/cpu_arm.go | 9 + vendor/golang.org/x/sys/cpu/cpu_gc_s390x.go | 21 + vendor/golang.org/x/sys/cpu/cpu_gc_x86.go | 16 + vendor/golang.org/x/sys/cpu/cpu_gccgo.c | 43 + vendor/golang.org/x/sys/cpu/cpu_gccgo.go | 26 + .../golang.org/x/sys/cpu/cpu_gccgo_s390x.go | 22 + vendor/golang.org/x/sys/cpu/cpu_linux.go | 59 + .../golang.org/x/sys/cpu/cpu_linux_arm64.go | 67 + .../golang.org/x/sys/cpu/cpu_linux_ppc64x.go | 33 + .../golang.org/x/sys/cpu/cpu_linux_s390x.go | 161 + vendor/golang.org/x/sys/cpu/cpu_mips64x.go | 11 + vendor/golang.org/x/sys/cpu/cpu_mipsx.go | 11 + .../golang.org/x/sys/cpu/cpu_other_arm64.go | 11 + vendor/golang.org/x/sys/cpu/cpu_s390x.s | 57 + vendor/golang.org/x/sys/cpu/cpu_wasm.go | 15 + vendor/golang.org/x/sys/cpu/cpu_x86.go | 59 + vendor/golang.org/x/sys/cpu/cpu_x86.s | 27 + vendor/golang.org/x/sys/unix/.gitignore | 2 + vendor/golang.org/x/sys/unix/README.md | 173 + .../golang.org/x/sys/unix/affinity_linux.go | 124 + vendor/golang.org/x/sys/unix/aliases.go | 14 + vendor/golang.org/x/sys/unix/asm_aix_ppc64.s | 17 + vendor/golang.org/x/sys/unix/asm_darwin_386.s | 29 + .../golang.org/x/sys/unix/asm_darwin_amd64.s | 29 + vendor/golang.org/x/sys/unix/asm_darwin_arm.s | 30 + .../golang.org/x/sys/unix/asm_darwin_arm64.s | 30 + .../x/sys/unix/asm_dragonfly_amd64.s | 29 + .../golang.org/x/sys/unix/asm_freebsd_386.s | 29 + .../golang.org/x/sys/unix/asm_freebsd_amd64.s | 29 + .../golang.org/x/sys/unix/asm_freebsd_arm.s | 29 + .../golang.org/x/sys/unix/asm_freebsd_arm64.s | 29 + vendor/golang.org/x/sys/unix/asm_linux_386.s | 65 + .../golang.org/x/sys/unix/asm_linux_amd64.s | 57 + vendor/golang.org/x/sys/unix/asm_linux_arm.s | 56 + .../golang.org/x/sys/unix/asm_linux_arm64.s | 52 + .../golang.org/x/sys/unix/asm_linux_mips64x.s | 56 + .../golang.org/x/sys/unix/asm_linux_mipsx.s | 54 + .../golang.org/x/sys/unix/asm_linux_ppc64x.s | 44 + .../golang.org/x/sys/unix/asm_linux_s390x.s | 56 + vendor/golang.org/x/sys/unix/asm_netbsd_386.s | 29 + .../golang.org/x/sys/unix/asm_netbsd_amd64.s | 29 + vendor/golang.org/x/sys/unix/asm_netbsd_arm.s | 29 + .../golang.org/x/sys/unix/asm_netbsd_arm64.s | 29 + .../golang.org/x/sys/unix/asm_openbsd_386.s | 29 + .../golang.org/x/sys/unix/asm_openbsd_amd64.s | 29 + .../golang.org/x/sys/unix/asm_openbsd_arm.s | 29 + .../golang.org/x/sys/unix/asm_solaris_amd64.s | 17 + .../golang.org/x/sys/unix/bluetooth_linux.go | 35 + vendor/golang.org/x/sys/unix/cap_freebsd.go | 195 + vendor/golang.org/x/sys/unix/constants.go | 13 + vendor/golang.org/x/sys/unix/dev_aix_ppc.go | 27 + vendor/golang.org/x/sys/unix/dev_aix_ppc64.go | 29 + vendor/golang.org/x/sys/unix/dev_darwin.go | 24 + vendor/golang.org/x/sys/unix/dev_dragonfly.go | 30 + vendor/golang.org/x/sys/unix/dev_freebsd.go | 30 + vendor/golang.org/x/sys/unix/dev_linux.go | 42 + vendor/golang.org/x/sys/unix/dev_netbsd.go | 29 + vendor/golang.org/x/sys/unix/dev_openbsd.go | 29 + vendor/golang.org/x/sys/unix/dirent.go | 17 + vendor/golang.org/x/sys/unix/endian_big.go | 9 + vendor/golang.org/x/sys/unix/endian_little.go | 9 + vendor/golang.org/x/sys/unix/env_unix.go | 31 + .../x/sys/unix/errors_freebsd_386.go | 227 + .../x/sys/unix/errors_freebsd_amd64.go | 227 + .../x/sys/unix/errors_freebsd_arm.go | 226 + vendor/golang.org/x/sys/unix/fcntl.go | 32 + vendor/golang.org/x/sys/unix/fcntl_darwin.go | 18 + .../x/sys/unix/fcntl_linux_32bit.go | 13 + vendor/golang.org/x/sys/unix/gccgo.go | 62 + vendor/golang.org/x/sys/unix/gccgo_c.c | 39 + .../x/sys/unix/gccgo_linux_amd64.go | 20 + vendor/golang.org/x/sys/unix/ioctl.go | 30 + vendor/golang.org/x/sys/unix/mkall.sh | 212 + vendor/golang.org/x/sys/unix/mkasm_darwin.go | 61 + vendor/golang.org/x/sys/unix/mkerrors.sh | 659 ++ vendor/golang.org/x/sys/unix/mkpost.go | 106 + vendor/golang.org/x/sys/unix/mksyscall.go | 407 + .../x/sys/unix/mksyscall_aix_ppc.go | 404 + .../x/sys/unix/mksyscall_aix_ppc64.go | 602 ++ .../x/sys/unix/mksyscall_solaris.go | 335 + .../golang.org/x/sys/unix/mksysctl_openbsd.pl | 265 + vendor/golang.org/x/sys/unix/mksysnum.go | 190 + .../golang.org/x/sys/unix/openbsd_pledge.go | 166 + .../golang.org/x/sys/unix/openbsd_unveil.go | 44 + vendor/golang.org/x/sys/unix/pagesize_unix.go | 15 + vendor/golang.org/x/sys/unix/race.go | 30 + vendor/golang.org/x/sys/unix/race0.go | 25 + .../golang.org/x/sys/unix/sockcmsg_linux.go | 36 + vendor/golang.org/x/sys/unix/sockcmsg_unix.go | 117 + vendor/golang.org/x/sys/unix/str.go | 26 + vendor/golang.org/x/sys/unix/syscall.go | 54 + vendor/golang.org/x/sys/unix/syscall_aix.go | 549 ++ .../golang.org/x/sys/unix/syscall_aix_ppc.go | 34 + .../x/sys/unix/syscall_aix_ppc64.go | 34 + vendor/golang.org/x/sys/unix/syscall_bsd.go | 624 ++ .../golang.org/x/sys/unix/syscall_darwin.go | 706 ++ .../x/sys/unix/syscall_darwin_386.go | 63 + .../x/sys/unix/syscall_darwin_amd64.go | 63 + .../x/sys/unix/syscall_darwin_arm.go | 64 + .../x/sys/unix/syscall_darwin_arm64.go | 66 + .../x/sys/unix/syscall_darwin_libSystem.go | 31 + .../x/sys/unix/syscall_dragonfly.go | 539 ++ .../x/sys/unix/syscall_dragonfly_amd64.go | 52 + .../golang.org/x/sys/unix/syscall_freebsd.go | 824 ++ .../x/sys/unix/syscall_freebsd_386.go | 52 + .../x/sys/unix/syscall_freebsd_amd64.go | 52 + .../x/sys/unix/syscall_freebsd_arm.go | 52 + .../x/sys/unix/syscall_freebsd_arm64.go | 52 + vendor/golang.org/x/sys/unix/syscall_linux.go | 1771 +++++ .../x/sys/unix/syscall_linux_386.go | 386 + .../x/sys/unix/syscall_linux_amd64.go | 190 + .../x/sys/unix/syscall_linux_amd64_gc.go | 13 + .../x/sys/unix/syscall_linux_arm.go | 274 + .../x/sys/unix/syscall_linux_arm64.go | 223 + .../golang.org/x/sys/unix/syscall_linux_gc.go | 14 + .../x/sys/unix/syscall_linux_gc_386.go | 16 + .../x/sys/unix/syscall_linux_gccgo_386.go | 30 + .../x/sys/unix/syscall_linux_gccgo_arm.go | 20 + .../x/sys/unix/syscall_linux_mips64x.go | 222 + .../x/sys/unix/syscall_linux_mipsx.go | 234 + .../x/sys/unix/syscall_linux_ppc64x.go | 152 + .../x/sys/unix/syscall_linux_riscv64.go | 226 + .../x/sys/unix/syscall_linux_s390x.go | 338 + .../x/sys/unix/syscall_linux_sparc64.go | 147 + .../golang.org/x/sys/unix/syscall_netbsd.go | 622 ++ .../x/sys/unix/syscall_netbsd_386.go | 33 + .../x/sys/unix/syscall_netbsd_amd64.go | 33 + .../x/sys/unix/syscall_netbsd_arm.go | 33 + .../x/sys/unix/syscall_netbsd_arm64.go | 33 + .../golang.org/x/sys/unix/syscall_openbsd.go | 416 + .../x/sys/unix/syscall_openbsd_386.go | 37 + .../x/sys/unix/syscall_openbsd_amd64.go | 37 + .../x/sys/unix/syscall_openbsd_arm.go | 37 + .../golang.org/x/sys/unix/syscall_solaris.go | 737 ++ .../x/sys/unix/syscall_solaris_amd64.go | 23 + vendor/golang.org/x/sys/unix/syscall_unix.go | 431 ++ .../golang.org/x/sys/unix/syscall_unix_gc.go | 15 + .../x/sys/unix/syscall_unix_gc_ppc64x.go | 24 + vendor/golang.org/x/sys/unix/timestruct.go | 82 + vendor/golang.org/x/sys/unix/types_aix.go | 236 + vendor/golang.org/x/sys/unix/types_darwin.go | 283 + .../golang.org/x/sys/unix/types_dragonfly.go | 263 + vendor/golang.org/x/sys/unix/types_freebsd.go | 356 + vendor/golang.org/x/sys/unix/types_netbsd.go | 289 + vendor/golang.org/x/sys/unix/types_openbsd.go | 282 + vendor/golang.org/x/sys/unix/types_solaris.go | 266 + vendor/golang.org/x/sys/unix/xattr_bsd.go | 240 + .../golang.org/x/sys/unix/zerrors_aix_ppc.go | 1372 ++++ .../x/sys/unix/zerrors_aix_ppc64.go | 1373 ++++ .../x/sys/unix/zerrors_darwin_386.go | 1783 +++++ .../x/sys/unix/zerrors_darwin_amd64.go | 1783 +++++ .../x/sys/unix/zerrors_darwin_arm.go | 1783 +++++ .../x/sys/unix/zerrors_darwin_arm64.go | 1783 +++++ .../x/sys/unix/zerrors_dragonfly_amd64.go | 1650 ++++ .../x/sys/unix/zerrors_freebsd_386.go | 1793 +++++ .../x/sys/unix/zerrors_freebsd_amd64.go | 1794 +++++ .../x/sys/unix/zerrors_freebsd_arm.go | 1802 +++++ .../x/sys/unix/zerrors_freebsd_arm64.go | 1794 +++++ .../x/sys/unix/zerrors_linux_386.go | 2835 +++++++ .../x/sys/unix/zerrors_linux_amd64.go | 2835 +++++++ .../x/sys/unix/zerrors_linux_arm.go | 2841 +++++++ .../x/sys/unix/zerrors_linux_arm64.go | 2826 +++++++ .../x/sys/unix/zerrors_linux_mips.go | 2842 +++++++ .../x/sys/unix/zerrors_linux_mips64.go | 2842 +++++++ .../x/sys/unix/zerrors_linux_mips64le.go | 2842 +++++++ .../x/sys/unix/zerrors_linux_mipsle.go | 2842 +++++++ .../x/sys/unix/zerrors_linux_ppc64.go | 2897 +++++++ .../x/sys/unix/zerrors_linux_ppc64le.go | 2897 +++++++ .../x/sys/unix/zerrors_linux_riscv64.go | 2822 +++++++ .../x/sys/unix/zerrors_linux_s390x.go | 2895 +++++++ .../x/sys/unix/zerrors_linux_sparc64.go | 2891 +++++++ .../x/sys/unix/zerrors_netbsd_386.go | 1772 +++++ .../x/sys/unix/zerrors_netbsd_amd64.go | 1762 +++++ .../x/sys/unix/zerrors_netbsd_arm.go | 1751 +++++ .../x/sys/unix/zerrors_netbsd_arm64.go | 1762 +++++ .../x/sys/unix/zerrors_openbsd_386.go | 1654 ++++ .../x/sys/unix/zerrors_openbsd_amd64.go | 1765 +++++ .../x/sys/unix/zerrors_openbsd_arm.go | 1656 ++++ .../x/sys/unix/zerrors_solaris_amd64.go | 1532 ++++ .../golang.org/x/sys/unix/zptrace386_linux.go | 80 + .../golang.org/x/sys/unix/zptracearm_linux.go | 41 + .../x/sys/unix/zptracemips_linux.go | 50 + .../x/sys/unix/zptracemipsle_linux.go | 50 + .../golang.org/x/sys/unix/zsyscall_aix_ppc.go | 1450 ++++ .../x/sys/unix/zsyscall_aix_ppc64.go | 1416 ++++ .../x/sys/unix/zsyscall_aix_ppc64_gc.go | 1172 +++ .../x/sys/unix/zsyscall_aix_ppc64_gccgo.go | 1051 +++ .../x/sys/unix/zsyscall_darwin_386.1_11.go | 1810 +++++ .../x/sys/unix/zsyscall_darwin_386.go | 2505 ++++++ .../x/sys/unix/zsyscall_darwin_386.s | 284 + .../x/sys/unix/zsyscall_darwin_amd64.1_11.go | 1810 +++++ .../x/sys/unix/zsyscall_darwin_amd64.go | 2520 ++++++ .../x/sys/unix/zsyscall_darwin_amd64.s | 286 + .../x/sys/unix/zsyscall_darwin_arm.1_11.go | 1793 +++++ .../x/sys/unix/zsyscall_darwin_arm.go | 2483 ++++++ .../x/sys/unix/zsyscall_darwin_arm.s | 282 + .../x/sys/unix/zsyscall_darwin_arm64.1_11.go | 1793 +++++ .../x/sys/unix/zsyscall_darwin_arm64.go | 2483 ++++++ .../x/sys/unix/zsyscall_darwin_arm64.s | 282 + .../x/sys/unix/zsyscall_dragonfly_amd64.go | 1659 ++++ .../x/sys/unix/zsyscall_freebsd_386.go | 2015 +++++ .../x/sys/unix/zsyscall_freebsd_amd64.go | 2015 +++++ .../x/sys/unix/zsyscall_freebsd_arm.go | 2015 +++++ .../x/sys/unix/zsyscall_freebsd_arm64.go | 2015 +++++ .../x/sys/unix/zsyscall_linux_386.go | 2220 ++++++ .../x/sys/unix/zsyscall_linux_amd64.go | 2387 ++++++ .../x/sys/unix/zsyscall_linux_arm.go | 2342 ++++++ .../x/sys/unix/zsyscall_linux_arm64.go | 2244 ++++++ .../x/sys/unix/zsyscall_linux_mips.go | 2400 ++++++ .../x/sys/unix/zsyscall_linux_mips64.go | 2371 ++++++ .../x/sys/unix/zsyscall_linux_mips64le.go | 2371 ++++++ .../x/sys/unix/zsyscall_linux_mipsle.go | 2400 ++++++ .../x/sys/unix/zsyscall_linux_ppc64.go | 2449 ++++++ .../x/sys/unix/zsyscall_linux_ppc64le.go | 2449 ++++++ .../x/sys/unix/zsyscall_linux_riscv64.go | 2224 ++++++ .../x/sys/unix/zsyscall_linux_s390x.go | 2219 ++++++ .../x/sys/unix/zsyscall_linux_sparc64.go | 2382 ++++++ .../x/sys/unix/zsyscall_netbsd_386.go | 1826 +++++ .../x/sys/unix/zsyscall_netbsd_amd64.go | 1826 +++++ .../x/sys/unix/zsyscall_netbsd_arm.go | 1826 +++++ .../x/sys/unix/zsyscall_netbsd_arm64.go | 1826 +++++ .../x/sys/unix/zsyscall_openbsd_386.go | 1692 ++++ .../x/sys/unix/zsyscall_openbsd_amd64.go | 1692 ++++ .../x/sys/unix/zsyscall_openbsd_arm.go | 1692 ++++ .../x/sys/unix/zsyscall_solaris_amd64.go | 1953 +++++ .../x/sys/unix/zsysctl_openbsd_386.go | 270 + .../x/sys/unix/zsysctl_openbsd_amd64.go | 270 + .../x/sys/unix/zsysctl_openbsd_arm.go | 270 + .../x/sys/unix/zsysnum_darwin_386.go | 436 ++ .../x/sys/unix/zsysnum_darwin_amd64.go | 438 ++ .../x/sys/unix/zsysnum_darwin_arm.go | 436 ++ .../x/sys/unix/zsysnum_darwin_arm64.go | 436 ++ .../x/sys/unix/zsysnum_dragonfly_amd64.go | 315 + .../x/sys/unix/zsysnum_freebsd_386.go | 403 + .../x/sys/unix/zsysnum_freebsd_amd64.go | 403 + .../x/sys/unix/zsysnum_freebsd_arm.go | 403 + .../x/sys/unix/zsysnum_freebsd_arm64.go | 395 + .../x/sys/unix/zsysnum_linux_386.go | 392 + .../x/sys/unix/zsysnum_linux_amd64.go | 344 + .../x/sys/unix/zsysnum_linux_arm.go | 364 + .../x/sys/unix/zsysnum_linux_arm64.go | 289 + .../x/sys/unix/zsysnum_linux_mips.go | 377 + .../x/sys/unix/zsysnum_linux_mips64.go | 337 + .../x/sys/unix/zsysnum_linux_mips64le.go | 337 + .../x/sys/unix/zsysnum_linux_mipsle.go | 377 + .../x/sys/unix/zsysnum_linux_ppc64.go | 375 + .../x/sys/unix/zsysnum_linux_ppc64le.go | 375 + .../x/sys/unix/zsysnum_linux_riscv64.go | 288 + .../x/sys/unix/zsysnum_linux_s390x.go | 337 + .../x/sys/unix/zsysnum_linux_sparc64.go | 351 + .../x/sys/unix/zsysnum_netbsd_386.go | 274 + .../x/sys/unix/zsysnum_netbsd_amd64.go | 274 + .../x/sys/unix/zsysnum_netbsd_arm.go | 274 + .../x/sys/unix/zsysnum_netbsd_arm64.go | 274 + .../x/sys/unix/zsysnum_openbsd_386.go | 218 + .../x/sys/unix/zsysnum_openbsd_amd64.go | 218 + .../x/sys/unix/zsysnum_openbsd_arm.go | 218 + .../golang.org/x/sys/unix/ztypes_aix_ppc.go | 345 + .../golang.org/x/sys/unix/ztypes_aix_ppc64.go | 354 + .../x/sys/unix/ztypes_darwin_386.go | 499 ++ .../x/sys/unix/ztypes_darwin_amd64.go | 509 ++ .../x/sys/unix/ztypes_darwin_arm.go | 500 ++ .../x/sys/unix/ztypes_darwin_arm64.go | 509 ++ .../x/sys/unix/ztypes_dragonfly_amd64.go | 469 ++ .../x/sys/unix/ztypes_freebsd_386.go | 603 ++ .../x/sys/unix/ztypes_freebsd_amd64.go | 602 ++ .../x/sys/unix/ztypes_freebsd_arm.go | 602 ++ .../x/sys/unix/ztypes_freebsd_arm64.go | 602 ++ .../golang.org/x/sys/unix/ztypes_linux_386.go | 2135 +++++ .../x/sys/unix/ztypes_linux_amd64.go | 2148 ++++++ .../golang.org/x/sys/unix/ztypes_linux_arm.go | 2126 +++++ .../x/sys/unix/ztypes_linux_arm64.go | 2127 +++++ .../x/sys/unix/ztypes_linux_mips.go | 2132 +++++ .../x/sys/unix/ztypes_linux_mips64.go | 2129 +++++ .../x/sys/unix/ztypes_linux_mips64le.go | 2129 +++++ .../x/sys/unix/ztypes_linux_mipsle.go | 2132 +++++ .../x/sys/unix/ztypes_linux_ppc64.go | 2137 +++++ .../x/sys/unix/ztypes_linux_ppc64le.go | 2137 +++++ .../x/sys/unix/ztypes_linux_riscv64.go | 2154 ++++++ .../x/sys/unix/ztypes_linux_s390x.go | 2151 ++++++ .../x/sys/unix/ztypes_linux_sparc64.go | 2132 +++++ .../x/sys/unix/ztypes_netbsd_386.go | 465 ++ .../x/sys/unix/ztypes_netbsd_amd64.go | 472 ++ .../x/sys/unix/ztypes_netbsd_arm.go | 470 ++ .../x/sys/unix/ztypes_netbsd_arm64.go | 472 ++ .../x/sys/unix/ztypes_openbsd_386.go | 570 ++ .../x/sys/unix/ztypes_openbsd_amd64.go | 570 ++ .../x/sys/unix/ztypes_openbsd_arm.go | 571 ++ .../x/sys/unix/ztypes_solaris_amd64.go | 442 ++ vendor/golang.org/x/sys/windows/aliases.go | 13 + .../x/sys/windows/asm_windows_386.s | 13 + .../x/sys/windows/asm_windows_amd64.s | 13 + .../x/sys/windows/asm_windows_arm.s | 11 + .../golang.org/x/sys/windows/dll_windows.go | 378 + .../golang.org/x/sys/windows/env_windows.go | 29 + vendor/golang.org/x/sys/windows/eventlog.go | 20 + .../golang.org/x/sys/windows/exec_windows.go | 97 + .../x/sys/windows/memory_windows.go | 26 + vendor/golang.org/x/sys/windows/mkerrors.bash | 64 + vendor/golang.org/x/sys/windows/mkerrors.go | 7 + vendor/golang.org/x/sys/windows/mksyscall.go | 7 + vendor/golang.org/x/sys/windows/race.go | 30 + vendor/golang.org/x/sys/windows/race0.go | 25 + .../x/sys/windows/security_windows.go | 649 ++ vendor/golang.org/x/sys/windows/service.go | 181 + vendor/golang.org/x/sys/windows/str.go | 22 + vendor/golang.org/x/sys/windows/syscall.go | 74 + .../x/sys/windows/syscall_windows.go | 1219 +++ .../golang.org/x/sys/windows/types_windows.go | 1441 ++++ .../x/sys/windows/types_windows_386.go | 22 + .../x/sys/windows/types_windows_amd64.go | 22 + .../x/sys/windows/types_windows_arm.go | 22 + .../x/sys/windows/zerrors_windows.go | 6853 +++++++++++++++++ .../x/sys/windows/zsyscall_windows.go | 2766 +++++++ vendor/modules.txt | 14 + 398 files changed, 242114 insertions(+) create mode 100644 .gitignore create mode 100644 LICENSE create mode 100644 Makefile create mode 100644 go.mod create mode 100644 go.sum create mode 100644 main.go create mode 100644 pkg/crypto/crypto.go create mode 100644 pkg/crypto/crypto_test.go create mode 100644 vendor/git.giftfish.de/ston1th/godrop/v2/.gitignore create mode 100644 vendor/git.giftfish.de/ston1th/godrop/v2/LICENSE create mode 100644 vendor/git.giftfish.de/ston1th/godrop/v2/README.md create mode 100644 vendor/git.giftfish.de/ston1th/godrop/v2/go.mod create mode 100644 vendor/git.giftfish.de/ston1th/godrop/v2/go.sum create mode 100644 vendor/git.giftfish.de/ston1th/godrop/v2/godrop.go create mode 100644 vendor/git.giftfish.de/ston1th/godrop/v2/lookup.go create mode 100644 vendor/git.giftfish.de/ston1th/godrop/v2/pledge.go create mode 100644 vendor/git.giftfish.de/ston1th/godrop/v2/pledge_openbsd.go create mode 100644 vendor/git.giftfish.de/ston1th/godrop/v2/unveil.go create mode 100644 vendor/git.giftfish.de/ston1th/godrop/v2/unveil_openbsd.go create mode 100644 vendor/golang.org/x/crypto/AUTHORS create mode 100644 vendor/golang.org/x/crypto/CONTRIBUTORS create mode 100644 vendor/golang.org/x/crypto/LICENSE create mode 100644 vendor/golang.org/x/crypto/PATENTS create mode 100644 vendor/golang.org/x/crypto/argon2/argon2.go create mode 100644 vendor/golang.org/x/crypto/argon2/blake2b.go create mode 100644 vendor/golang.org/x/crypto/argon2/blamka_amd64.go create mode 100644 vendor/golang.org/x/crypto/argon2/blamka_amd64.s create mode 100644 vendor/golang.org/x/crypto/argon2/blamka_generic.go create mode 100644 vendor/golang.org/x/crypto/argon2/blamka_ref.go create mode 100644 vendor/golang.org/x/crypto/blake2b/blake2b.go create mode 100644 vendor/golang.org/x/crypto/blake2b/blake2bAVX2_amd64.go create mode 100644 vendor/golang.org/x/crypto/blake2b/blake2bAVX2_amd64.s create mode 100644 vendor/golang.org/x/crypto/blake2b/blake2b_amd64.go create mode 100644 vendor/golang.org/x/crypto/blake2b/blake2b_amd64.s create mode 100644 vendor/golang.org/x/crypto/blake2b/blake2b_generic.go create mode 100644 vendor/golang.org/x/crypto/blake2b/blake2b_ref.go create mode 100644 vendor/golang.org/x/crypto/blake2b/blake2x.go create mode 100644 vendor/golang.org/x/crypto/blake2b/register.go create mode 100644 vendor/golang.org/x/crypto/chacha20poly1305/chacha20poly1305.go create mode 100644 vendor/golang.org/x/crypto/chacha20poly1305/chacha20poly1305_amd64.go create mode 100644 vendor/golang.org/x/crypto/chacha20poly1305/chacha20poly1305_amd64.s create mode 100644 vendor/golang.org/x/crypto/chacha20poly1305/chacha20poly1305_generic.go create mode 100644 vendor/golang.org/x/crypto/chacha20poly1305/chacha20poly1305_noasm.go create mode 100644 vendor/golang.org/x/crypto/chacha20poly1305/xchacha20poly1305.go create mode 100644 vendor/golang.org/x/crypto/internal/chacha20/asm_arm64.s create mode 100644 vendor/golang.org/x/crypto/internal/chacha20/asm_ppc64le.s create mode 100644 vendor/golang.org/x/crypto/internal/chacha20/chacha_arm64.go create mode 100644 vendor/golang.org/x/crypto/internal/chacha20/chacha_generic.go create mode 100644 vendor/golang.org/x/crypto/internal/chacha20/chacha_noasm.go create mode 100644 vendor/golang.org/x/crypto/internal/chacha20/chacha_ppc64le.go create mode 100644 vendor/golang.org/x/crypto/internal/chacha20/chacha_s390x.go create mode 100644 vendor/golang.org/x/crypto/internal/chacha20/chacha_s390x.s create mode 100644 vendor/golang.org/x/crypto/internal/chacha20/xor.go create mode 100644 vendor/golang.org/x/crypto/internal/subtle/aliasing.go create mode 100644 vendor/golang.org/x/crypto/internal/subtle/aliasing_appengine.go create mode 100644 vendor/golang.org/x/crypto/poly1305/mac_noasm.go create mode 100644 vendor/golang.org/x/crypto/poly1305/poly1305.go create mode 100644 vendor/golang.org/x/crypto/poly1305/sum_amd64.go create mode 100644 vendor/golang.org/x/crypto/poly1305/sum_amd64.s create mode 100644 vendor/golang.org/x/crypto/poly1305/sum_arm.go create mode 100644 vendor/golang.org/x/crypto/poly1305/sum_arm.s create mode 100644 vendor/golang.org/x/crypto/poly1305/sum_generic.go create mode 100644 vendor/golang.org/x/crypto/poly1305/sum_noasm.go create mode 100644 vendor/golang.org/x/crypto/poly1305/sum_ppc64le.go create mode 100644 vendor/golang.org/x/crypto/poly1305/sum_ppc64le.s create mode 100644 vendor/golang.org/x/crypto/poly1305/sum_s390x.go create mode 100644 vendor/golang.org/x/crypto/poly1305/sum_s390x.s create mode 100644 vendor/golang.org/x/crypto/poly1305/sum_vmsl_s390x.s create mode 100644 vendor/golang.org/x/crypto/ssh/terminal/terminal.go create mode 100644 vendor/golang.org/x/crypto/ssh/terminal/util.go create mode 100644 vendor/golang.org/x/crypto/ssh/terminal/util_aix.go create mode 100644 vendor/golang.org/x/crypto/ssh/terminal/util_bsd.go create mode 100644 vendor/golang.org/x/crypto/ssh/terminal/util_linux.go create mode 100644 vendor/golang.org/x/crypto/ssh/terminal/util_plan9.go create mode 100644 vendor/golang.org/x/crypto/ssh/terminal/util_solaris.go create mode 100644 vendor/golang.org/x/crypto/ssh/terminal/util_windows.go create mode 100644 vendor/golang.org/x/sys/AUTHORS create mode 100644 vendor/golang.org/x/sys/CONTRIBUTORS create mode 100644 vendor/golang.org/x/sys/LICENSE create mode 100644 vendor/golang.org/x/sys/PATENTS create mode 100644 vendor/golang.org/x/sys/cpu/byteorder.go create mode 100644 vendor/golang.org/x/sys/cpu/cpu.go create mode 100644 vendor/golang.org/x/sys/cpu/cpu_aix_ppc64.go create mode 100644 vendor/golang.org/x/sys/cpu/cpu_arm.go create mode 100644 vendor/golang.org/x/sys/cpu/cpu_gc_s390x.go create mode 100644 vendor/golang.org/x/sys/cpu/cpu_gc_x86.go create mode 100644 vendor/golang.org/x/sys/cpu/cpu_gccgo.c create mode 100644 vendor/golang.org/x/sys/cpu/cpu_gccgo.go create mode 100644 vendor/golang.org/x/sys/cpu/cpu_gccgo_s390x.go create mode 100644 vendor/golang.org/x/sys/cpu/cpu_linux.go create mode 100644 vendor/golang.org/x/sys/cpu/cpu_linux_arm64.go create mode 100644 vendor/golang.org/x/sys/cpu/cpu_linux_ppc64x.go create mode 100644 vendor/golang.org/x/sys/cpu/cpu_linux_s390x.go create mode 100644 vendor/golang.org/x/sys/cpu/cpu_mips64x.go create mode 100644 vendor/golang.org/x/sys/cpu/cpu_mipsx.go create mode 100644 vendor/golang.org/x/sys/cpu/cpu_other_arm64.go create mode 100644 vendor/golang.org/x/sys/cpu/cpu_s390x.s create mode 100644 vendor/golang.org/x/sys/cpu/cpu_wasm.go create mode 100644 vendor/golang.org/x/sys/cpu/cpu_x86.go create mode 100644 vendor/golang.org/x/sys/cpu/cpu_x86.s create mode 100644 vendor/golang.org/x/sys/unix/.gitignore create mode 100644 vendor/golang.org/x/sys/unix/README.md create mode 100644 vendor/golang.org/x/sys/unix/affinity_linux.go create mode 100644 vendor/golang.org/x/sys/unix/aliases.go create mode 100644 vendor/golang.org/x/sys/unix/asm_aix_ppc64.s create mode 100644 vendor/golang.org/x/sys/unix/asm_darwin_386.s create mode 100644 vendor/golang.org/x/sys/unix/asm_darwin_amd64.s create mode 100644 vendor/golang.org/x/sys/unix/asm_darwin_arm.s create mode 100644 vendor/golang.org/x/sys/unix/asm_darwin_arm64.s create mode 100644 vendor/golang.org/x/sys/unix/asm_dragonfly_amd64.s create mode 100644 vendor/golang.org/x/sys/unix/asm_freebsd_386.s create mode 100644 vendor/golang.org/x/sys/unix/asm_freebsd_amd64.s create mode 100644 vendor/golang.org/x/sys/unix/asm_freebsd_arm.s create mode 100644 vendor/golang.org/x/sys/unix/asm_freebsd_arm64.s create mode 100644 vendor/golang.org/x/sys/unix/asm_linux_386.s create mode 100644 vendor/golang.org/x/sys/unix/asm_linux_amd64.s create mode 100644 vendor/golang.org/x/sys/unix/asm_linux_arm.s create mode 100644 vendor/golang.org/x/sys/unix/asm_linux_arm64.s create mode 100644 vendor/golang.org/x/sys/unix/asm_linux_mips64x.s create mode 100644 vendor/golang.org/x/sys/unix/asm_linux_mipsx.s create mode 100644 vendor/golang.org/x/sys/unix/asm_linux_ppc64x.s create mode 100644 vendor/golang.org/x/sys/unix/asm_linux_s390x.s create mode 100644 vendor/golang.org/x/sys/unix/asm_netbsd_386.s create mode 100644 vendor/golang.org/x/sys/unix/asm_netbsd_amd64.s create mode 100644 vendor/golang.org/x/sys/unix/asm_netbsd_arm.s create mode 100644 vendor/golang.org/x/sys/unix/asm_netbsd_arm64.s create mode 100644 vendor/golang.org/x/sys/unix/asm_openbsd_386.s create mode 100644 vendor/golang.org/x/sys/unix/asm_openbsd_amd64.s create mode 100644 vendor/golang.org/x/sys/unix/asm_openbsd_arm.s create mode 100644 vendor/golang.org/x/sys/unix/asm_solaris_amd64.s create mode 100644 vendor/golang.org/x/sys/unix/bluetooth_linux.go create mode 100644 vendor/golang.org/x/sys/unix/cap_freebsd.go create mode 100644 vendor/golang.org/x/sys/unix/constants.go create mode 100644 vendor/golang.org/x/sys/unix/dev_aix_ppc.go create mode 100644 vendor/golang.org/x/sys/unix/dev_aix_ppc64.go create mode 100644 vendor/golang.org/x/sys/unix/dev_darwin.go create mode 100644 vendor/golang.org/x/sys/unix/dev_dragonfly.go create mode 100644 vendor/golang.org/x/sys/unix/dev_freebsd.go create mode 100644 vendor/golang.org/x/sys/unix/dev_linux.go create mode 100644 vendor/golang.org/x/sys/unix/dev_netbsd.go create mode 100644 vendor/golang.org/x/sys/unix/dev_openbsd.go create mode 100644 vendor/golang.org/x/sys/unix/dirent.go create mode 100644 vendor/golang.org/x/sys/unix/endian_big.go create mode 100644 vendor/golang.org/x/sys/unix/endian_little.go create mode 100644 vendor/golang.org/x/sys/unix/env_unix.go create mode 100644 vendor/golang.org/x/sys/unix/errors_freebsd_386.go create mode 100644 vendor/golang.org/x/sys/unix/errors_freebsd_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/errors_freebsd_arm.go create mode 100644 vendor/golang.org/x/sys/unix/fcntl.go create mode 100644 vendor/golang.org/x/sys/unix/fcntl_darwin.go create mode 100644 vendor/golang.org/x/sys/unix/fcntl_linux_32bit.go create mode 100644 vendor/golang.org/x/sys/unix/gccgo.go create mode 100644 vendor/golang.org/x/sys/unix/gccgo_c.c create mode 100644 vendor/golang.org/x/sys/unix/gccgo_linux_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/ioctl.go create mode 100644 vendor/golang.org/x/sys/unix/mkall.sh create mode 100644 vendor/golang.org/x/sys/unix/mkasm_darwin.go create mode 100644 vendor/golang.org/x/sys/unix/mkerrors.sh create mode 100644 vendor/golang.org/x/sys/unix/mkpost.go create mode 100644 vendor/golang.org/x/sys/unix/mksyscall.go create mode 100644 vendor/golang.org/x/sys/unix/mksyscall_aix_ppc.go create mode 100644 vendor/golang.org/x/sys/unix/mksyscall_aix_ppc64.go create mode 100644 vendor/golang.org/x/sys/unix/mksyscall_solaris.go create mode 100644 vendor/golang.org/x/sys/unix/mksysctl_openbsd.pl create mode 100644 vendor/golang.org/x/sys/unix/mksysnum.go create mode 100644 vendor/golang.org/x/sys/unix/openbsd_pledge.go create mode 100644 vendor/golang.org/x/sys/unix/openbsd_unveil.go create mode 100644 vendor/golang.org/x/sys/unix/pagesize_unix.go create mode 100644 vendor/golang.org/x/sys/unix/race.go create mode 100644 vendor/golang.org/x/sys/unix/race0.go create mode 100644 vendor/golang.org/x/sys/unix/sockcmsg_linux.go create mode 100644 vendor/golang.org/x/sys/unix/sockcmsg_unix.go create mode 100644 vendor/golang.org/x/sys/unix/str.go create mode 100644 vendor/golang.org/x/sys/unix/syscall.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_aix.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_aix_ppc.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_aix_ppc64.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_bsd.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_darwin.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_darwin_386.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_darwin_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_darwin_arm.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_darwin_arm64.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_darwin_libSystem.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_dragonfly.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_dragonfly_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_freebsd.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_freebsd_386.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_freebsd_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_freebsd_arm.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_freebsd_arm64.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_linux.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_linux_386.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_linux_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_linux_amd64_gc.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_linux_arm.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_linux_arm64.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_linux_gc.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_linux_gc_386.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_linux_gccgo_386.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_linux_gccgo_arm.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_linux_mips64x.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_linux_mipsx.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_linux_ppc64x.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_linux_riscv64.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_linux_s390x.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_linux_sparc64.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_netbsd.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_netbsd_386.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_netbsd_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_netbsd_arm.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_netbsd_arm64.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_openbsd.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_openbsd_386.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_openbsd_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_openbsd_arm.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_solaris.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_solaris_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_unix.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_unix_gc.go create mode 100644 vendor/golang.org/x/sys/unix/syscall_unix_gc_ppc64x.go create mode 100644 vendor/golang.org/x/sys/unix/timestruct.go create mode 100644 vendor/golang.org/x/sys/unix/types_aix.go create mode 100644 vendor/golang.org/x/sys/unix/types_darwin.go create mode 100644 vendor/golang.org/x/sys/unix/types_dragonfly.go create mode 100644 vendor/golang.org/x/sys/unix/types_freebsd.go create mode 100644 vendor/golang.org/x/sys/unix/types_netbsd.go create mode 100644 vendor/golang.org/x/sys/unix/types_openbsd.go create mode 100644 vendor/golang.org/x/sys/unix/types_solaris.go create mode 100644 vendor/golang.org/x/sys/unix/xattr_bsd.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_aix_ppc.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_aix_ppc64.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_darwin_386.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_darwin_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_darwin_arm.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_darwin_arm64.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_dragonfly_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_freebsd_386.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_freebsd_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_freebsd_arm.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_freebsd_arm64.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_linux_386.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_linux_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_linux_arm.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_linux_arm64.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_linux_mips.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_linux_mips64.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_linux_mips64le.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_linux_mipsle.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_linux_ppc64.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_linux_ppc64le.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_linux_riscv64.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_linux_s390x.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_linux_sparc64.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_netbsd_386.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_netbsd_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_netbsd_arm.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_netbsd_arm64.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_openbsd_386.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_openbsd_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_openbsd_arm.go create mode 100644 vendor/golang.org/x/sys/unix/zerrors_solaris_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/zptrace386_linux.go create mode 100644 vendor/golang.org/x/sys/unix/zptracearm_linux.go create mode 100644 vendor/golang.org/x/sys/unix/zptracemips_linux.go create mode 100644 vendor/golang.org/x/sys/unix/zptracemipsle_linux.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_aix_ppc.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_aix_ppc64.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_aix_ppc64_gc.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_aix_ppc64_gccgo.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_darwin_386.1_11.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_darwin_386.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_darwin_386.s create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_darwin_amd64.1_11.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_darwin_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_darwin_amd64.s create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_darwin_arm.1_11.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_darwin_arm.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_darwin_arm.s create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_darwin_arm64.1_11.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_darwin_arm64.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_darwin_arm64.s create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_dragonfly_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_freebsd_386.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_freebsd_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_freebsd_arm.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_freebsd_arm64.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_linux_386.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_linux_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_linux_arm.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_linux_arm64.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_linux_mips.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_linux_mips64.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_linux_mips64le.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_linux_mipsle.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_linux_ppc64.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_linux_ppc64le.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_linux_riscv64.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_linux_s390x.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_linux_sparc64.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_netbsd_386.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_netbsd_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_netbsd_arm.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_netbsd_arm64.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_openbsd_386.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_openbsd_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_openbsd_arm.go create mode 100644 vendor/golang.org/x/sys/unix/zsyscall_solaris_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/zsysctl_openbsd_386.go create mode 100644 vendor/golang.org/x/sys/unix/zsysctl_openbsd_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/zsysctl_openbsd_arm.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_darwin_386.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_darwin_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_darwin_arm.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_darwin_arm64.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_dragonfly_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_freebsd_386.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_freebsd_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_freebsd_arm.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_freebsd_arm64.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_linux_386.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_linux_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_linux_arm.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_linux_arm64.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_linux_mips.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_linux_mips64.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_linux_mips64le.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_linux_mipsle.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_linux_ppc64.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_linux_ppc64le.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_linux_riscv64.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_linux_s390x.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_linux_sparc64.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_netbsd_386.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_netbsd_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_netbsd_arm.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_netbsd_arm64.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_openbsd_386.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_openbsd_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/zsysnum_openbsd_arm.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_aix_ppc.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_aix_ppc64.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_darwin_386.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_darwin_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_darwin_arm.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_darwin_arm64.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_dragonfly_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_freebsd_386.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_freebsd_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_freebsd_arm.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_freebsd_arm64.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_linux_386.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_linux_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_linux_arm.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_linux_arm64.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_linux_mips.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_linux_mips64.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_linux_mips64le.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_linux_mipsle.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_linux_ppc64.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_linux_ppc64le.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_linux_riscv64.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_linux_s390x.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_linux_sparc64.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_netbsd_386.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_netbsd_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_netbsd_arm.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_netbsd_arm64.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_openbsd_386.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_openbsd_amd64.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_openbsd_arm.go create mode 100644 vendor/golang.org/x/sys/unix/ztypes_solaris_amd64.go create mode 100644 vendor/golang.org/x/sys/windows/aliases.go create mode 100644 vendor/golang.org/x/sys/windows/asm_windows_386.s create mode 100644 vendor/golang.org/x/sys/windows/asm_windows_amd64.s create mode 100644 vendor/golang.org/x/sys/windows/asm_windows_arm.s create mode 100644 vendor/golang.org/x/sys/windows/dll_windows.go create mode 100644 vendor/golang.org/x/sys/windows/env_windows.go create mode 100644 vendor/golang.org/x/sys/windows/eventlog.go create mode 100644 vendor/golang.org/x/sys/windows/exec_windows.go create mode 100644 vendor/golang.org/x/sys/windows/memory_windows.go create mode 100644 vendor/golang.org/x/sys/windows/mkerrors.bash create mode 100644 vendor/golang.org/x/sys/windows/mkerrors.go create mode 100644 vendor/golang.org/x/sys/windows/mksyscall.go create mode 100644 vendor/golang.org/x/sys/windows/race.go create mode 100644 vendor/golang.org/x/sys/windows/race0.go create mode 100644 vendor/golang.org/x/sys/windows/security_windows.go create mode 100644 vendor/golang.org/x/sys/windows/service.go create mode 100644 vendor/golang.org/x/sys/windows/str.go create mode 100644 vendor/golang.org/x/sys/windows/syscall.go create mode 100644 vendor/golang.org/x/sys/windows/syscall_windows.go create mode 100644 vendor/golang.org/x/sys/windows/types_windows.go create mode 100644 vendor/golang.org/x/sys/windows/types_windows_386.go create mode 100644 vendor/golang.org/x/sys/windows/types_windows_amd64.go create mode 100644 vendor/golang.org/x/sys/windows/types_windows_arm.go create mode 100644 vendor/golang.org/x/sys/windows/zerrors_windows.go create mode 100644 vendor/golang.org/x/sys/windows/zsyscall_windows.go create mode 100644 vendor/modules.txt diff --git a/.gitignore b/.gitignore new file mode 100644 index 0000000..1e9c059 --- /dev/null +++ b/.gitignore @@ -0,0 +1 @@ +gocrypt diff --git a/LICENSE b/LICENSE new file mode 100644 index 0000000..45785b1 --- /dev/null +++ b/LICENSE @@ -0,0 +1,24 @@ +Copyright (C) 2019 Marius Schellenberger +All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are met: + * Redistributions of source code must retain the above copyright + notice, this list of conditions and the following disclaimer. + * Redistributions in binary form must reproduce the above copyright + notice, this list of conditions and the following disclaimer in the + documentation and/or other materials provided with the distribution. + * The names of the authors and/or contributors may not be used to + endorse or promote products derived from this software without + specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND +ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED +WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE +DISCLAIMED. IN NO EVENT SHALL ston1th BE LIABLE FOR ANY +DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES +(INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; +LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND +ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS +SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. diff --git a/Makefile b/Makefile new file mode 100644 index 0000000..3e1778a --- /dev/null +++ b/Makefile @@ -0,0 +1,39 @@ +CC=go +BUILD=build -v +GCFLAGS=-gcflags '-e' +LDFLAGS=-ldflags '-s -w' +PROGRAM=gocrypt +ENV=CGO_ENABLED=0 GO111MODULE=on + +all: $(PROGRAM) + +release: clean gofmt test + $(ENV) $(CC) $(BUILD) $(GCFLAGS) $(LDFLAGS) + +release-vendor: clean gofmt test + $(ENV) $(CC) $(BUILD) -mod=vendor $(GCFLAGS) $(LDFLAGS) + +vendor: clean gofmt + $(ENV) $(CC) $(BUILD) -mod=vendor $(GCFLAGS) $(LDFLAGS) + +$(PROGRAM): clean gofmt + $(ENV) $(CC) $(BUILD) $(GCFLAGS) $(LDFLAGS) + +clean: + $(CC) clean -x + +gofmt: + gofmt -w pkg/* main.go + +test: +ifeq ($(shell go env GOARCH), $(shell go env GOHOSTARCH)) +ifeq ($(shell go env GOOS), $(shell go env GOHOSTOS)) + $(CC) test ./... +else + $(info skipping tests of other platforms) +endif +else + $(info skipping tests of other platforms) +endif + +.PHONY: release release-vendor vendor build clean gofmt test diff --git a/go.mod b/go.mod new file mode 100644 index 0000000..6e471c0 --- /dev/null +++ b/go.mod @@ -0,0 +1,8 @@ +module git.giftfish.de/ston1th/gocrypt + +go 1.12 + +require ( + git.giftfish.de/ston1th/godrop/v2 v2.1.0 + golang.org/x/crypto v0.0.0-20190605123033-f99c8df09eb5 +) diff --git a/go.sum b/go.sum new file mode 100644 index 0000000..02a0277 --- /dev/null +++ b/go.sum @@ -0,0 +1,11 @@ +git.giftfish.de/ston1th/godrop/v2 v2.1.0 h1:vR4aWVeMGJ+tHQH0p44esZrSl5dYJzuCmZqhKWP/vwI= +git.giftfish.de/ston1th/godrop/v2 v2.1.0/go.mod h1:z8KA3HP/vB04wddLXZfYVKURYDxBGau9hlHtV61x9mM= +golang.org/x/crypto v0.0.0-20190308221718-c2843e01d9a2/go.mod h1:djNgcEr1/C05ACkg1iLfiJU5Ep61QUkGW8qpdssI0+w= +golang.org/x/crypto v0.0.0-20190605123033-f99c8df09eb5 h1:58fnuSXlxZmFdJyvtTFVmVhcMLU6v5fEb/ok4wyqtNU= +golang.org/x/crypto v0.0.0-20190605123033-f99c8df09eb5/go.mod h1:yigFU9vqHzYiE8UmvKecakEJjdnWj3jj499lnFckfCI= +golang.org/x/net v0.0.0-20190404232315-eb5bcb51f2a3/go.mod h1:t9HGtf8HONx5eT2rtn7q6eTqICYqUVnKs3thJo3Qplg= +golang.org/x/sys v0.0.0-20181026144532-2772b66316d2/go.mod h1:STP8DvDyc/dI5b8T5hshtkjS+E42TnysNCUPdjciGhY= +golang.org/x/sys v0.0.0-20190215142949-d0b11bdaac8a/go.mod h1:STP8DvDyc/dI5b8T5hshtkjS+E42TnysNCUPdjciGhY= +golang.org/x/sys v0.0.0-20190412213103-97732733099d h1:+R4KGOnez64A81RvjARKc4UT5/tI9ujCIVX+P5KiHuI= +golang.org/x/sys v0.0.0-20190412213103-97732733099d/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs= +golang.org/x/text v0.3.0/go.mod h1:NqM8EUOU14njkJ3fqMW+pc6Ldnwhi/IjpwHt7yyuwOQ= diff --git a/main.go b/main.go new file mode 100644 index 0000000..c399f82 --- /dev/null +++ b/main.go @@ -0,0 +1,84 @@ +// Copyright (C) 2019 Marius Schellenberger + +package main + +import ( + "flag" + "fmt" + "git.giftfish.de/ston1th/gocrypt/pkg/crypto" + "git.giftfish.de/ston1th/godrop/v2" + "io" + "io/ioutil" + "os" +) + +var ( + encMode bool + decMode bool + input string + output string +) + +func fatal(err error) { + if err != nil { + fatals(err.Error()) + } +} + +func fatals(s string) { + stderr("error: " + s + "\n") + os.Exit(1) +} + +func stderr(s string) { + fmt.Fprintf(os.Stderr, s) +} + +func init() { + fatal(godrop.PledgePromises("stdio rpath wpath cpath tty unveil")) + fatal(godrop.Unveil("/", "")) + flag.BoolVar(&encMode, "enc", false, "encryption mode - password is read from stdin") + flag.BoolVar(&decMode, "dec", false, "decryption mode - password is read from stdin") + flag.StringVar(&input, "f", "", "input file") + flag.StringVar(&output, "o", "-", "output file (default stdout)") + flag.Parse() +} + +func main() { + var ( + in []byte + out io.WriteCloser + err error + ) + if input == "" { + fatals("input file is empty") + } else { + fatal(godrop.Unveil(input, "r")) + in, err = ioutil.ReadFile(input) + fatal(err) + } + if output == "" { + fatals("output file is empty") + } else if output == "-" { + out = os.Stdout + } else { + fatal(godrop.Unveil(output, "wc")) + out, err = os.OpenFile(output, os.O_WRONLY|os.O_CREATE, 0640) + fatal(err) + } + fatal(godrop.PledgePromises("stdio tty")) + stderr("Enter password: ") + pass, err := crypto.Password() + stderr("\n") + fatal(err) + fatal(godrop.PledgePromises("stdio")) + switch { + case encMode == true: + err = crypto.Encrypt(pass, in, out) + fatal(err) + case decMode == true: + err = crypto.Decrypt(pass, in, out) + fatal(err) + } + fatal(out.Close()) +} diff --git a/pkg/crypto/crypto.go b/pkg/crypto/crypto.go new file mode 100644 index 0000000..47078ff --- /dev/null +++ b/pkg/crypto/crypto.go @@ -0,0 +1,71 @@ +// Copyright (C) 2019 Marius Schellenberger + +package crypto + +import ( + "crypto/rand" + "errors" + "golang.org/x/crypto/argon2" + "golang.org/x/crypto/chacha20poly1305" + "golang.org/x/crypto/ssh/terminal" + "io" + "os" +) + +const ( + SaltSize = 16 + PrefixSize = SaltSize + chacha20poly1305.NonceSizeX +) + +func randomBytes(l int) (b []byte) { + b = make([]byte, l) + _, err := rand.Read(b) + if err != nil { + panic(err) + } + return +} + +func kdf(password, salt []byte) []byte { + return argon2.IDKey(password, salt, 1, 64*1024, 4, chacha20poly1305.KeySize) +} + +func Password() ([]byte, error) { + return terminal.ReadPassword(int(os.Stdin.Fd())) +} + +func Encrypt(pass, src []byte, dst io.Writer) (err error) { + salt := randomBytes(SaltSize) + nonce := randomBytes(chacha20poly1305.NonceSizeX) + aead, err := chacha20poly1305.NewX(kdf(pass, salt)) + if err != nil { + return + } + _, err = dst.Write(salt) + if err != nil { + return + } + _, err = dst.Write(nonce) + if err != nil { + return + } + _, err = dst.Write(aead.Seal(nil, nonce, src, nil)) + return +} + +func Decrypt(pass, src []byte, dst io.Writer) (err error) { + if len(src) <= PrefixSize { + return errors.New("input file corrupted") + } + nonce := src[SaltSize:PrefixSize] + aead, err := chacha20poly1305.NewX(kdf(pass, src[:SaltSize])) + if err != nil { + return + } + dec, err := aead.Open(nil, nonce, src[PrefixSize:], nil) + if err != nil { + return + } + _, err = dst.Write(dec) + return +} diff --git a/pkg/crypto/crypto_test.go b/pkg/crypto/crypto_test.go new file mode 100644 index 0000000..4b32a64 --- /dev/null +++ b/pkg/crypto/crypto_test.go @@ -0,0 +1,38 @@ +// Copyright (C) 2019 Marius Schellenberger + +package crypto + +import ( + "bytes" + "encoding/base64" + "testing" +) + +const ( + text = "Lorem ipsum dolor sit amet, consetetur sadipscing elitr, sed diam nonumy eirmod tempor invidunt ut labore et dolore magna aliquyam erat, sed diam voluptua." + cipher = "kkU4cAJeNmbFGaZFdubxFzUPfo/YNu3y02vKOD97mh48hJshRsBTbAkMo6TCq2yre89zcr54ZY6olAohOqSZAux5cpowHJQ1Hj7YDq+xAjWybERhApYTvhc283DW48Fc5usXmxoWbcqQ1eBl+QRC/RHBUO2qJ0p/MBu6sI2+nCu8z4ZLRqkh6JoMcLekYrKp6Tn34gaGgWi0ElSkD/uJzQmt3dqGHu5Hufj9MzzZlezzIHI3qzKFUaOKwFtLyoPBO2V6liMYCn2yxGWpsJPJQjvbcg==" + pass = "SecurePassword123" +) + +func TestEncrypt(t *testing.T) { + buf := new(bytes.Buffer) + err := Encrypt([]byte(pass), []byte(text), buf) + if err != nil { + t.Error(err) + } +} + +func TestDecrypt(t *testing.T) { + buf := new(bytes.Buffer) + src, err := base64.StdEncoding.DecodeString(cipher) + if err != nil { + t.Error(err) + } + err = Decrypt([]byte(pass), src, buf) + if err != nil { + t.Error(err) + } + if buf.String() != text { + t.Error("decrypted text does not match") + } +} diff --git a/vendor/git.giftfish.de/ston1th/godrop/v2/.gitignore b/vendor/git.giftfish.de/ston1th/godrop/v2/.gitignore new file mode 100644 index 0000000..af033d5 --- /dev/null +++ b/vendor/git.giftfish.de/ston1th/godrop/v2/.gitignore @@ -0,0 +1,3 @@ +port80 +port80fg +port80and443 diff --git a/vendor/git.giftfish.de/ston1th/godrop/v2/LICENSE b/vendor/git.giftfish.de/ston1th/godrop/v2/LICENSE new file mode 100644 index 0000000..2bb060c --- /dev/null +++ b/vendor/git.giftfish.de/ston1th/godrop/v2/LICENSE @@ -0,0 +1,24 @@ +Copyright (C) 2017 Marius Schellenberger +All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are met: + * Redistributions of source code must retain the above copyright + notice, this list of conditions and the following disclaimer. + * Redistributions in binary form must reproduce the above copyright + notice, this list of conditions and the following disclaimer in the + documentation and/or other materials provided with the distribution. + * The names of the authors and/or contributors may not be used to + endorse or promote products derived from this software without + specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND +ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED +WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE +DISCLAIMED. IN NO EVENT SHALL ston1th BE LIABLE FOR ANY +DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES +(INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; +LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND +ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS +SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. diff --git a/vendor/git.giftfish.de/ston1th/godrop/v2/README.md b/vendor/git.giftfish.de/ston1th/godrop/v2/README.md new file mode 100644 index 0000000..bcd5479 --- /dev/null +++ b/vendor/git.giftfish.de/ston1th/godrop/v2/README.md @@ -0,0 +1,5 @@ +# godrop - drop privileges + +Godrop is a simple library to drop privileges on Linux and OpenBSD. + +See the examples directory on how to use the `Drop` and `MultiDrop` functions. diff --git a/vendor/git.giftfish.de/ston1th/godrop/v2/go.mod b/vendor/git.giftfish.de/ston1th/godrop/v2/go.mod new file mode 100644 index 0000000..a41c9af --- /dev/null +++ b/vendor/git.giftfish.de/ston1th/godrop/v2/go.mod @@ -0,0 +1,3 @@ +module git.giftfish.de/ston1th/godrop/v2 + +require golang.org/x/sys v0.0.0-20181026144532-2772b66316d2 // indirect diff --git a/vendor/git.giftfish.de/ston1th/godrop/v2/go.sum b/vendor/git.giftfish.de/ston1th/godrop/v2/go.sum new file mode 100644 index 0000000..36d0abd --- /dev/null +++ b/vendor/git.giftfish.de/ston1th/godrop/v2/go.sum @@ -0,0 +1,2 @@ +golang.org/x/sys v0.0.0-20181026144532-2772b66316d2 h1:W7CqTdBJ1CmxLKe7LptKDnBYV6PHrVLiGnoyBjaG/JQ= +golang.org/x/sys v0.0.0-20181026144532-2772b66316d2/go.mod h1:STP8DvDyc/dI5b8T5hshtkjS+E42TnysNCUPdjciGhY= diff --git a/vendor/git.giftfish.de/ston1th/godrop/v2/godrop.go b/vendor/git.giftfish.de/ston1th/godrop/v2/godrop.go new file mode 100644 index 0000000..55f39ec --- /dev/null +++ b/vendor/git.giftfish.de/ston1th/godrop/v2/godrop.go @@ -0,0 +1,136 @@ +// Copyright (C) 2018 Marius Schellenberger + +// +build go1.11 + +// Package godrop provides a simple library to drop privileges on Linux and OpenBSD. +package godrop + +import ( + "errors" + "fmt" + "net" + "os" + "os/exec" + "os/signal" + "syscall" +) + +// Config represents the drop config +type Config struct { + // User is the user to drop privileges to. + User string + // Group is the group to drop privileges to. + Group string + // Chroot is the directory to chroot into. Leave this emptry for no chroot. + // When compiling without cgo, make sure the chroot directory contains the /etc/passwd and /etc/group files. + Chroot string + // Set to true, to run the process in the foreground. + Foreground bool +} + +// Drop will spawn a new process and hand over the listening socket file descriptor +func Drop(c Config, f func() (net.Listener, error)) error { + return MultiDrop(c, func() ([]net.Listener, error) { + l, err := f() + return []net.Listener{l}, err + }) +} + +// MultiDrop will spawn a new process and hand over the all listening sockets +func MultiDrop(c Config, f func() ([]net.Listener, error)) error { + uid, err := userID(c.User) + if err != nil { + return errors.New("godrop: " + err.Error()) + } + if uid == 0 { + return fmt.Errorf("godrop: you can't drop privileges to uid 0 (%s)", c.User) + } + gid, err := groupID(c.Group) + if err != nil { + return errors.New("godrop: " + err.Error()) + } + switch os.Getuid() { + case 0: + cmd := exec.Command(os.Args[0], os.Args[1:]...) + ln, err := f() + if err != nil { + return errors.New("godrop: " + err.Error()) + } + for i, v := range ln { + var f *os.File + switch l := v.(type) { + case *net.TCPListener: + f, err = l.File() + l.Close() + case *net.UnixListener: + f, err = l.File() + l.Close() + default: + return fmt.Errorf("godrop: index %d listener is not type of either *net.TCPListener or *net.UnixListener", i) + } + if err != nil { + return fmt.Errorf("godrop: index %d %s", i, err) + } + cmd.ExtraFiles = append(cmd.ExtraFiles, f) + } + + cmd.SysProcAttr = &syscall.SysProcAttr{ + Chroot: c.Chroot, + Credential: &syscall.Credential{ + Uid: uint32(uid), + Gid: uint32(gid), + }, + Setsid: true, + } + + if c.Foreground { + cmd.Stdin = os.Stdin + cmd.Stdout = os.Stdout + cmd.Stderr = os.Stderr + } + + if err := cmd.Start(); err != nil { + return errors.New("godrop: " + err.Error()) + } + + if c.Foreground { + go func() { + term := make(chan os.Signal) + signal.Notify(term, syscall.SIGHUP, syscall.SIGINT, syscall.SIGTERM, syscall.SIGQUIT) + sigs := make(chan os.Signal) + signal.Notify(sigs, syscall.SIGSTOP, syscall.SIGCONT, syscall.SIGUSR1, syscall.SIGUSR2) + for { + select { + case sig := <-term: + cmd.Process.Signal(sig) + return + case sig := <-sigs: + cmd.Process.Signal(sig) + } + } + }() + _ = cmd.Wait() + os.Exit(int(cmd.ProcessState.Sys().(syscall.WaitStatus))) + } + cmd.Process.Release() + os.Exit(0) + case uid: + return nil + } + return errors.New("godrop: dropping priviledges failed") +} + +// GetListener returns the listener socket of file descriptor 3 +func GetListener() (net.Listener, error) { + return GetListenerFd(3) +} + +// GetListenerFd returns the listener socket of the given file descriptor +func GetListenerFd(fd int) (net.Listener, error) { + if fd < 3 { + return nil, errors.New("godrop: fd is less than 3") + } + f := os.NewFile(uintptr(fd), "") + defer f.Close() + return net.FileListener(f) +} diff --git a/vendor/git.giftfish.de/ston1th/godrop/v2/lookup.go b/vendor/git.giftfish.de/ston1th/godrop/v2/lookup.go new file mode 100644 index 0000000..2d0e1a9 --- /dev/null +++ b/vendor/git.giftfish.de/ston1th/godrop/v2/lookup.go @@ -0,0 +1,34 @@ +// Copyright (C) 2018 Marius Schellenberger + +// +build go1.11 + +package godrop + +import ( + "os/user" + "strconv" +) + +func atoi(a string) (int, error) { + i, err := strconv.Atoi(a) + if err != nil { + return -1, err + } + return i, nil +} + +func userID(username string) (int, error) { + u, err := user.Lookup(username) + if err != nil { + return -1, err + } + return atoi(u.Uid) +} + +func groupID(name string) (int, error) { + g, err := user.LookupGroup(name) + if err != nil { + return -1, err + } + return atoi(g.Gid) +} diff --git a/vendor/git.giftfish.de/ston1th/godrop/v2/pledge.go b/vendor/git.giftfish.de/ston1th/godrop/v2/pledge.go new file mode 100644 index 0000000..d0e361e --- /dev/null +++ b/vendor/git.giftfish.de/ston1th/godrop/v2/pledge.go @@ -0,0 +1,14 @@ +// Copyright (C) 2018 Marius Schellenberger + +// +build !openbsd + +package godrop + +// Pledge is currently only supported on OpenBSD. +func Pledge(promises, execpromises string) error { return nil } + +// PledgePromises is currently only supported on OpenBSD. +func PledgePromises(promises string) error { return nil } + +// PledgeExecPromises is currently only supported on OpenBSD. +func PledgeExecpromises(execpromises string) error { return nil } diff --git a/vendor/git.giftfish.de/ston1th/godrop/v2/pledge_openbsd.go b/vendor/git.giftfish.de/ston1th/godrop/v2/pledge_openbsd.go new file mode 100644 index 0000000..7369b03 --- /dev/null +++ b/vendor/git.giftfish.de/ston1th/godrop/v2/pledge_openbsd.go @@ -0,0 +1,43 @@ +// Copyright (C) 2018 Marius Schellenberger + +// +build openbsd + +package godrop + +import ( + "errors" + "golang.org/x/sys/unix" +) + +// Pledge is a wrapper for x/sys/unix Pledge. +// +// See https://go.googlesource.com/sys/+/master/unix/openbsd_pledge.go for usage. +func Pledge(promises, execpromises string) (err error) { + err = unix.Pledge(promises, execpromises) + if err != nil { + err = errors.New("pledge: " + err.Error()) + } + return +} + +// PledgePromises is a wrapper for x/sys/unix PledgePromises. +// +// See https://go.googlesource.com/sys/+/master/unix/openbsd_pledge.go for usage. +func PledgePromises(promises string) (err error) { + err = unix.PledgePromises(promises) + if err != nil { + err = errors.New("pledge: " + err.Error()) + } + return +} + +// PledgeExecpromises is a wrapper for x/sys/unix PledgeExecpromises. +// +// See https://go.googlesource.com/sys/+/master/unix/openbsd_pledge.go for usage. +func PledgeExecpromises(execpromises string) (err error) { + err = unix.PledgeExecpromises(execpromises) + if err != nil { + err = errors.New("pledge: " + err.Error()) + } + return +} diff --git a/vendor/git.giftfish.de/ston1th/godrop/v2/unveil.go b/vendor/git.giftfish.de/ston1th/godrop/v2/unveil.go new file mode 100644 index 0000000..06c8c90 --- /dev/null +++ b/vendor/git.giftfish.de/ston1th/godrop/v2/unveil.go @@ -0,0 +1,11 @@ +// Copyright (C) 2018 Marius Schellenberger + +// +build !openbsd + +package godrop + +// Unveil is currently only supported on OpenBSD. +func Unveil(path, flags string) error { return nil } + +// UnveilBlock is currently only supported on OpenBSD. +func UnveilBlock() error { return nil } diff --git a/vendor/git.giftfish.de/ston1th/godrop/v2/unveil_openbsd.go b/vendor/git.giftfish.de/ston1th/godrop/v2/unveil_openbsd.go new file mode 100644 index 0000000..f32d630 --- /dev/null +++ b/vendor/git.giftfish.de/ston1th/godrop/v2/unveil_openbsd.go @@ -0,0 +1,32 @@ +// Copyright (C) 2018 Marius Schellenberger + +// +build openbsd + +package godrop + +import ( + "errors" + "golang.org/x/sys/unix" +) + +// Unveil is a wrapper for x/sys/unix Unveil. +// +// See https://go.googlesource.com/sys/+/master/unix/openbsd_unveil.go for usage. +func Unveil(path, flags string) (err error) { + err = unix.Unveil(path, flags) + if err != nil { + err = errors.New("unveil: " + err.Error()) + } + return +} + +// UnveilBlock is a wrapper for x/sys/unix UnveilBlock. +// +// See https://go.googlesource.com/sys/+/master/unix/openbsd_unveil.go for usage. +func UnveilBlock() (err error) { + err = unix.UnveilBlock() + if err != nil { + err = errors.New("unveil: " + err.Error()) + } + return +} diff --git a/vendor/golang.org/x/crypto/AUTHORS b/vendor/golang.org/x/crypto/AUTHORS new file mode 100644 index 0000000..2b00ddb --- /dev/null +++ b/vendor/golang.org/x/crypto/AUTHORS @@ -0,0 +1,3 @@ +# This source code refers to The Go Authors for copyright purposes. +# The master list of authors is in the main Go distribution, +# visible at https://tip.golang.org/AUTHORS. diff --git a/vendor/golang.org/x/crypto/CONTRIBUTORS b/vendor/golang.org/x/crypto/CONTRIBUTORS new file mode 100644 index 0000000..1fbd3e9 --- /dev/null +++ b/vendor/golang.org/x/crypto/CONTRIBUTORS @@ -0,0 +1,3 @@ +# This source code was written by the Go contributors. +# The master list of contributors is in the main Go distribution, +# visible at https://tip.golang.org/CONTRIBUTORS. diff --git a/vendor/golang.org/x/crypto/LICENSE b/vendor/golang.org/x/crypto/LICENSE new file mode 100644 index 0000000..6a66aea --- /dev/null +++ b/vendor/golang.org/x/crypto/LICENSE @@ -0,0 +1,27 @@ +Copyright (c) 2009 The Go Authors. All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are +met: + + * Redistributions of source code must retain the above copyright +notice, this list of conditions and the following disclaimer. + * Redistributions in binary form must reproduce the above +copyright notice, this list of conditions and the following disclaimer +in the documentation and/or other materials provided with the +distribution. + * Neither the name of Google Inc. nor the names of its +contributors may be used to endorse or promote products derived from +this software without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT +LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, +DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY +THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. diff --git a/vendor/golang.org/x/crypto/PATENTS b/vendor/golang.org/x/crypto/PATENTS new file mode 100644 index 0000000..7330990 --- /dev/null +++ b/vendor/golang.org/x/crypto/PATENTS @@ -0,0 +1,22 @@ +Additional IP Rights Grant (Patents) + +"This implementation" means the copyrightable works distributed by +Google as part of the Go project. + +Google hereby grants to You a perpetual, worldwide, non-exclusive, +no-charge, royalty-free, irrevocable (except as stated in this section) +patent license to make, have made, use, offer to sell, sell, import, +transfer and otherwise run, modify and propagate the contents of this +implementation of Go, where such license applies only to those patent +claims, both currently owned or controlled by Google and acquired in +the future, licensable by Google that are necessarily infringed by this +implementation of Go. This grant does not include claims that would be +infringed only as a consequence of further modification of this +implementation. If you or your agent or exclusive licensee institute or +order or agree to the institution of patent litigation against any +entity (including a cross-claim or counterclaim in a lawsuit) alleging +that this implementation of Go or any code incorporated within this +implementation of Go constitutes direct or contributory patent +infringement, or inducement of patent infringement, then any patent +rights granted to you under this License for this implementation of Go +shall terminate as of the date such litigation is filed. diff --git a/vendor/golang.org/x/crypto/argon2/argon2.go b/vendor/golang.org/x/crypto/argon2/argon2.go new file mode 100644 index 0000000..b423fea --- /dev/null +++ b/vendor/golang.org/x/crypto/argon2/argon2.go @@ -0,0 +1,285 @@ +// Copyright 2017 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// Package argon2 implements the key derivation function Argon2. +// Argon2 was selected as the winner of the Password Hashing Competition and can +// be used to derive cryptographic keys from passwords. +// +// For a detailed specification of Argon2 see [1]. +// +// If you aren't sure which function you need, use Argon2id (IDKey) and +// the parameter recommendations for your scenario. +// +// +// Argon2i +// +// Argon2i (implemented by Key) is the side-channel resistant version of Argon2. +// It uses data-independent memory access, which is preferred for password +// hashing and password-based key derivation. Argon2i requires more passes over +// memory than Argon2id to protect from trade-off attacks. The recommended +// parameters (taken from [2]) for non-interactive operations are time=3 and to +// use the maximum available memory. +// +// +// Argon2id +// +// Argon2id (implemented by IDKey) is a hybrid version of Argon2 combining +// Argon2i and Argon2d. It uses data-independent memory access for the first +// half of the first iteration over the memory and data-dependent memory access +// for the rest. Argon2id is side-channel resistant and provides better brute- +// force cost savings due to time-memory tradeoffs than Argon2i. The recommended +// parameters for non-interactive operations (taken from [2]) are time=1 and to +// use the maximum available memory. +// +// [1] https://github.com/P-H-C/phc-winner-argon2/blob/master/argon2-specs.pdf +// [2] https://tools.ietf.org/html/draft-irtf-cfrg-argon2-03#section-9.3 +package argon2 + +import ( + "encoding/binary" + "sync" + + "golang.org/x/crypto/blake2b" +) + +// The Argon2 version implemented by this package. +const Version = 0x13 + +const ( + argon2d = iota + argon2i + argon2id +) + +// Key derives a key from the password, salt, and cost parameters using Argon2i +// returning a byte slice of length keyLen that can be used as cryptographic +// key. The CPU cost and parallelism degree must be greater than zero. +// +// For example, you can get a derived key for e.g. AES-256 (which needs a +// 32-byte key) by doing: +// +// key := argon2.Key([]byte("some password"), salt, 3, 32*1024, 4, 32) +// +// The draft RFC recommends[2] time=3, and memory=32*1024 is a sensible number. +// If using that amount of memory (32 MB) is not possible in some contexts then +// the time parameter can be increased to compensate. +// +// The time parameter specifies the number of passes over the memory and the +// memory parameter specifies the size of the memory in KiB. For example +// memory=32*1024 sets the memory cost to ~32 MB. The number of threads can be +// adjusted to the number of available CPUs. The cost parameters should be +// increased as memory latency and CPU parallelism increases. Remember to get a +// good random salt. +func Key(password, salt []byte, time, memory uint32, threads uint8, keyLen uint32) []byte { + return deriveKey(argon2i, password, salt, nil, nil, time, memory, threads, keyLen) +} + +// IDKey derives a key from the password, salt, and cost parameters using +// Argon2id returning a byte slice of length keyLen that can be used as +// cryptographic key. The CPU cost and parallelism degree must be greater than +// zero. +// +// For example, you can get a derived key for e.g. AES-256 (which needs a +// 32-byte key) by doing: +// +// key := argon2.IDKey([]byte("some password"), salt, 1, 64*1024, 4, 32) +// +// The draft RFC recommends[2] time=1, and memory=64*1024 is a sensible number. +// If using that amount of memory (64 MB) is not possible in some contexts then +// the time parameter can be increased to compensate. +// +// The time parameter specifies the number of passes over the memory and the +// memory parameter specifies the size of the memory in KiB. For example +// memory=64*1024 sets the memory cost to ~64 MB. The number of threads can be +// adjusted to the numbers of available CPUs. The cost parameters should be +// increased as memory latency and CPU parallelism increases. Remember to get a +// good random salt. +func IDKey(password, salt []byte, time, memory uint32, threads uint8, keyLen uint32) []byte { + return deriveKey(argon2id, password, salt, nil, nil, time, memory, threads, keyLen) +} + +func deriveKey(mode int, password, salt, secret, data []byte, time, memory uint32, threads uint8, keyLen uint32) []byte { + if time < 1 { + panic("argon2: number of rounds too small") + } + if threads < 1 { + panic("argon2: parallelism degree too low") + } + h0 := initHash(password, salt, secret, data, time, memory, uint32(threads), keyLen, mode) + + memory = memory / (syncPoints * uint32(threads)) * (syncPoints * uint32(threads)) + if memory < 2*syncPoints*uint32(threads) { + memory = 2 * syncPoints * uint32(threads) + } + B := initBlocks(&h0, memory, uint32(threads)) + processBlocks(B, time, memory, uint32(threads), mode) + return extractKey(B, memory, uint32(threads), keyLen) +} + +const ( + blockLength = 128 + syncPoints = 4 +) + +type block [blockLength]uint64 + +func initHash(password, salt, key, data []byte, time, memory, threads, keyLen uint32, mode int) [blake2b.Size + 8]byte { + var ( + h0 [blake2b.Size + 8]byte + params [24]byte + tmp [4]byte + ) + + b2, _ := blake2b.New512(nil) + binary.LittleEndian.PutUint32(params[0:4], threads) + binary.LittleEndian.PutUint32(params[4:8], keyLen) + binary.LittleEndian.PutUint32(params[8:12], memory) + binary.LittleEndian.PutUint32(params[12:16], time) + binary.LittleEndian.PutUint32(params[16:20], uint32(Version)) + binary.LittleEndian.PutUint32(params[20:24], uint32(mode)) + b2.Write(params[:]) + binary.LittleEndian.PutUint32(tmp[:], uint32(len(password))) + b2.Write(tmp[:]) + b2.Write(password) + binary.LittleEndian.PutUint32(tmp[:], uint32(len(salt))) + b2.Write(tmp[:]) + b2.Write(salt) + binary.LittleEndian.PutUint32(tmp[:], uint32(len(key))) + b2.Write(tmp[:]) + b2.Write(key) + binary.LittleEndian.PutUint32(tmp[:], uint32(len(data))) + b2.Write(tmp[:]) + b2.Write(data) + b2.Sum(h0[:0]) + return h0 +} + +func initBlocks(h0 *[blake2b.Size + 8]byte, memory, threads uint32) []block { + var block0 [1024]byte + B := make([]block, memory) + for lane := uint32(0); lane < threads; lane++ { + j := lane * (memory / threads) + binary.LittleEndian.PutUint32(h0[blake2b.Size+4:], lane) + + binary.LittleEndian.PutUint32(h0[blake2b.Size:], 0) + blake2bHash(block0[:], h0[:]) + for i := range B[j+0] { + B[j+0][i] = binary.LittleEndian.Uint64(block0[i*8:]) + } + + binary.LittleEndian.PutUint32(h0[blake2b.Size:], 1) + blake2bHash(block0[:], h0[:]) + for i := range B[j+1] { + B[j+1][i] = binary.LittleEndian.Uint64(block0[i*8:]) + } + } + return B +} + +func processBlocks(B []block, time, memory, threads uint32, mode int) { + lanes := memory / threads + segments := lanes / syncPoints + + processSegment := func(n, slice, lane uint32, wg *sync.WaitGroup) { + var addresses, in, zero block + if mode == argon2i || (mode == argon2id && n == 0 && slice < syncPoints/2) { + in[0] = uint64(n) + in[1] = uint64(lane) + in[2] = uint64(slice) + in[3] = uint64(memory) + in[4] = uint64(time) + in[5] = uint64(mode) + } + + index := uint32(0) + if n == 0 && slice == 0 { + index = 2 // we have already generated the first two blocks + if mode == argon2i || mode == argon2id { + in[6]++ + processBlock(&addresses, &in, &zero) + processBlock(&addresses, &addresses, &zero) + } + } + + offset := lane*lanes + slice*segments + index + var random uint64 + for index < segments { + prev := offset - 1 + if index == 0 && slice == 0 { + prev += lanes // last block in lane + } + if mode == argon2i || (mode == argon2id && n == 0 && slice < syncPoints/2) { + if index%blockLength == 0 { + in[6]++ + processBlock(&addresses, &in, &zero) + processBlock(&addresses, &addresses, &zero) + } + random = addresses[index%blockLength] + } else { + random = B[prev][0] + } + newOffset := indexAlpha(random, lanes, segments, threads, n, slice, lane, index) + processBlockXOR(&B[offset], &B[prev], &B[newOffset]) + index, offset = index+1, offset+1 + } + wg.Done() + } + + for n := uint32(0); n < time; n++ { + for slice := uint32(0); slice < syncPoints; slice++ { + var wg sync.WaitGroup + for lane := uint32(0); lane < threads; lane++ { + wg.Add(1) + go processSegment(n, slice, lane, &wg) + } + wg.Wait() + } + } + +} + +func extractKey(B []block, memory, threads, keyLen uint32) []byte { + lanes := memory / threads + for lane := uint32(0); lane < threads-1; lane++ { + for i, v := range B[(lane*lanes)+lanes-1] { + B[memory-1][i] ^= v + } + } + + var block [1024]byte + for i, v := range B[memory-1] { + binary.LittleEndian.PutUint64(block[i*8:], v) + } + key := make([]byte, keyLen) + blake2bHash(key, block[:]) + return key +} + +func indexAlpha(rand uint64, lanes, segments, threads, n, slice, lane, index uint32) uint32 { + refLane := uint32(rand>>32) % threads + if n == 0 && slice == 0 { + refLane = lane + } + m, s := 3*segments, ((slice+1)%syncPoints)*segments + if lane == refLane { + m += index + } + if n == 0 { + m, s = slice*segments, 0 + if slice == 0 || lane == refLane { + m += index + } + } + if index == 0 || lane == refLane { + m-- + } + return phi(rand, uint64(m), uint64(s), refLane, lanes) +} + +func phi(rand, m, s uint64, lane, lanes uint32) uint32 { + p := rand & 0xFFFFFFFF + p = (p * p) >> 32 + p = (p * m) >> 32 + return lane*lanes + uint32((s+m-(p+1))%uint64(lanes)) +} diff --git a/vendor/golang.org/x/crypto/argon2/blake2b.go b/vendor/golang.org/x/crypto/argon2/blake2b.go new file mode 100644 index 0000000..10f4694 --- /dev/null +++ b/vendor/golang.org/x/crypto/argon2/blake2b.go @@ -0,0 +1,53 @@ +// Copyright 2017 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package argon2 + +import ( + "encoding/binary" + "hash" + + "golang.org/x/crypto/blake2b" +) + +// blake2bHash computes an arbitrary long hash value of in +// and writes the hash to out. +func blake2bHash(out []byte, in []byte) { + var b2 hash.Hash + if n := len(out); n < blake2b.Size { + b2, _ = blake2b.New(n, nil) + } else { + b2, _ = blake2b.New512(nil) + } + + var buffer [blake2b.Size]byte + binary.LittleEndian.PutUint32(buffer[:4], uint32(len(out))) + b2.Write(buffer[:4]) + b2.Write(in) + + if len(out) <= blake2b.Size { + b2.Sum(out[:0]) + return + } + + outLen := len(out) + b2.Sum(buffer[:0]) + b2.Reset() + copy(out, buffer[:32]) + out = out[32:] + for len(out) > blake2b.Size { + b2.Write(buffer[:]) + b2.Sum(buffer[:0]) + copy(out, buffer[:32]) + out = out[32:] + b2.Reset() + } + + if outLen%blake2b.Size > 0 { // outLen > 64 + r := ((outLen + 31) / 32) - 2 // ⌈τ /32⌉-2 + b2, _ = blake2b.New(outLen-32*r, nil) + } + b2.Write(buffer[:]) + b2.Sum(out[:0]) +} diff --git a/vendor/golang.org/x/crypto/argon2/blamka_amd64.go b/vendor/golang.org/x/crypto/argon2/blamka_amd64.go new file mode 100644 index 0000000..2fc1ec0 --- /dev/null +++ b/vendor/golang.org/x/crypto/argon2/blamka_amd64.go @@ -0,0 +1,60 @@ +// Copyright 2017 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build amd64,!gccgo,!appengine + +package argon2 + +import "golang.org/x/sys/cpu" + +func init() { + useSSE4 = cpu.X86.HasSSE41 +} + +//go:noescape +func mixBlocksSSE2(out, a, b, c *block) + +//go:noescape +func xorBlocksSSE2(out, a, b, c *block) + +//go:noescape +func blamkaSSE4(b *block) + +func processBlockSSE(out, in1, in2 *block, xor bool) { + var t block + mixBlocksSSE2(&t, in1, in2, &t) + if useSSE4 { + blamkaSSE4(&t) + } else { + for i := 0; i < blockLength; i += 16 { + blamkaGeneric( + &t[i+0], &t[i+1], &t[i+2], &t[i+3], + &t[i+4], &t[i+5], &t[i+6], &t[i+7], + &t[i+8], &t[i+9], &t[i+10], &t[i+11], + &t[i+12], &t[i+13], &t[i+14], &t[i+15], + ) + } + for i := 0; i < blockLength/8; i += 2 { + blamkaGeneric( + &t[i], &t[i+1], &t[16+i], &t[16+i+1], + &t[32+i], &t[32+i+1], &t[48+i], &t[48+i+1], + &t[64+i], &t[64+i+1], &t[80+i], &t[80+i+1], + &t[96+i], &t[96+i+1], &t[112+i], &t[112+i+1], + ) + } + } + if xor { + xorBlocksSSE2(out, in1, in2, &t) + } else { + mixBlocksSSE2(out, in1, in2, &t) + } +} + +func processBlock(out, in1, in2 *block) { + processBlockSSE(out, in1, in2, false) +} + +func processBlockXOR(out, in1, in2 *block) { + processBlockSSE(out, in1, in2, true) +} diff --git a/vendor/golang.org/x/crypto/argon2/blamka_amd64.s b/vendor/golang.org/x/crypto/argon2/blamka_amd64.s new file mode 100644 index 0000000..74a6e73 --- /dev/null +++ b/vendor/golang.org/x/crypto/argon2/blamka_amd64.s @@ -0,0 +1,243 @@ +// Copyright 2017 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build amd64,!gccgo,!appengine + +#include "textflag.h" + +DATA ·c40<>+0x00(SB)/8, $0x0201000706050403 +DATA ·c40<>+0x08(SB)/8, $0x0a09080f0e0d0c0b +GLOBL ·c40<>(SB), (NOPTR+RODATA), $16 + +DATA ·c48<>+0x00(SB)/8, $0x0100070605040302 +DATA ·c48<>+0x08(SB)/8, $0x09080f0e0d0c0b0a +GLOBL ·c48<>(SB), (NOPTR+RODATA), $16 + +#define SHUFFLE(v2, v3, v4, v5, v6, v7, t1, t2) \ + MOVO v4, t1; \ + MOVO v5, v4; \ + MOVO t1, v5; \ + MOVO v6, t1; \ + PUNPCKLQDQ v6, t2; \ + PUNPCKHQDQ v7, v6; \ + PUNPCKHQDQ t2, v6; \ + PUNPCKLQDQ v7, t2; \ + MOVO t1, v7; \ + MOVO v2, t1; \ + PUNPCKHQDQ t2, v7; \ + PUNPCKLQDQ v3, t2; \ + PUNPCKHQDQ t2, v2; \ + PUNPCKLQDQ t1, t2; \ + PUNPCKHQDQ t2, v3 + +#define SHUFFLE_INV(v2, v3, v4, v5, v6, v7, t1, t2) \ + MOVO v4, t1; \ + MOVO v5, v4; \ + MOVO t1, v5; \ + MOVO v2, t1; \ + PUNPCKLQDQ v2, t2; \ + PUNPCKHQDQ v3, v2; \ + PUNPCKHQDQ t2, v2; \ + PUNPCKLQDQ v3, t2; \ + MOVO t1, v3; \ + MOVO v6, t1; \ + PUNPCKHQDQ t2, v3; \ + PUNPCKLQDQ v7, t2; \ + PUNPCKHQDQ t2, v6; \ + PUNPCKLQDQ t1, t2; \ + PUNPCKHQDQ t2, v7 + +#define HALF_ROUND(v0, v1, v2, v3, v4, v5, v6, v7, t0, c40, c48) \ + MOVO v0, t0; \ + PMULULQ v2, t0; \ + PADDQ v2, v0; \ + PADDQ t0, v0; \ + PADDQ t0, v0; \ + PXOR v0, v6; \ + PSHUFD $0xB1, v6, v6; \ + MOVO v4, t0; \ + PMULULQ v6, t0; \ + PADDQ v6, v4; \ + PADDQ t0, v4; \ + PADDQ t0, v4; \ + PXOR v4, v2; \ + PSHUFB c40, v2; \ + MOVO v0, t0; \ + PMULULQ v2, t0; \ + PADDQ v2, v0; \ + PADDQ t0, v0; \ + PADDQ t0, v0; \ + PXOR v0, v6; \ + PSHUFB c48, v6; \ + MOVO v4, t0; \ + PMULULQ v6, t0; \ + PADDQ v6, v4; \ + PADDQ t0, v4; \ + PADDQ t0, v4; \ + PXOR v4, v2; \ + MOVO v2, t0; \ + PADDQ v2, t0; \ + PSRLQ $63, v2; \ + PXOR t0, v2; \ + MOVO v1, t0; \ + PMULULQ v3, t0; \ + PADDQ v3, v1; \ + PADDQ t0, v1; \ + PADDQ t0, v1; \ + PXOR v1, v7; \ + PSHUFD $0xB1, v7, v7; \ + MOVO v5, t0; \ + PMULULQ v7, t0; \ + PADDQ v7, v5; \ + PADDQ t0, v5; \ + PADDQ t0, v5; \ + PXOR v5, v3; \ + PSHUFB c40, v3; \ + MOVO v1, t0; \ + PMULULQ v3, t0; \ + PADDQ v3, v1; \ + PADDQ t0, v1; \ + PADDQ t0, v1; \ + PXOR v1, v7; \ + PSHUFB c48, v7; \ + MOVO v5, t0; \ + PMULULQ v7, t0; \ + PADDQ v7, v5; \ + PADDQ t0, v5; \ + PADDQ t0, v5; \ + PXOR v5, v3; \ + MOVO v3, t0; \ + PADDQ v3, t0; \ + PSRLQ $63, v3; \ + PXOR t0, v3 + +#define LOAD_MSG_0(block, off) \ + MOVOU 8*(off+0)(block), X0; \ + MOVOU 8*(off+2)(block), X1; \ + MOVOU 8*(off+4)(block), X2; \ + MOVOU 8*(off+6)(block), X3; \ + MOVOU 8*(off+8)(block), X4; \ + MOVOU 8*(off+10)(block), X5; \ + MOVOU 8*(off+12)(block), X6; \ + MOVOU 8*(off+14)(block), X7 + +#define STORE_MSG_0(block, off) \ + MOVOU X0, 8*(off+0)(block); \ + MOVOU X1, 8*(off+2)(block); \ + MOVOU X2, 8*(off+4)(block); \ + MOVOU X3, 8*(off+6)(block); \ + MOVOU X4, 8*(off+8)(block); \ + MOVOU X5, 8*(off+10)(block); \ + MOVOU X6, 8*(off+12)(block); \ + MOVOU X7, 8*(off+14)(block) + +#define LOAD_MSG_1(block, off) \ + MOVOU 8*off+0*8(block), X0; \ + MOVOU 8*off+16*8(block), X1; \ + MOVOU 8*off+32*8(block), X2; \ + MOVOU 8*off+48*8(block), X3; \ + MOVOU 8*off+64*8(block), X4; \ + MOVOU 8*off+80*8(block), X5; \ + MOVOU 8*off+96*8(block), X6; \ + MOVOU 8*off+112*8(block), X7 + +#define STORE_MSG_1(block, off) \ + MOVOU X0, 8*off+0*8(block); \ + MOVOU X1, 8*off+16*8(block); \ + MOVOU X2, 8*off+32*8(block); \ + MOVOU X3, 8*off+48*8(block); \ + MOVOU X4, 8*off+64*8(block); \ + MOVOU X5, 8*off+80*8(block); \ + MOVOU X6, 8*off+96*8(block); \ + MOVOU X7, 8*off+112*8(block) + +#define BLAMKA_ROUND_0(block, off, t0, t1, c40, c48) \ + LOAD_MSG_0(block, off); \ + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, t0, c40, c48); \ + SHUFFLE(X2, X3, X4, X5, X6, X7, t0, t1); \ + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, t0, c40, c48); \ + SHUFFLE_INV(X2, X3, X4, X5, X6, X7, t0, t1); \ + STORE_MSG_0(block, off) + +#define BLAMKA_ROUND_1(block, off, t0, t1, c40, c48) \ + LOAD_MSG_1(block, off); \ + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, t0, c40, c48); \ + SHUFFLE(X2, X3, X4, X5, X6, X7, t0, t1); \ + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, t0, c40, c48); \ + SHUFFLE_INV(X2, X3, X4, X5, X6, X7, t0, t1); \ + STORE_MSG_1(block, off) + +// func blamkaSSE4(b *block) +TEXT ·blamkaSSE4(SB), 4, $0-8 + MOVQ b+0(FP), AX + + MOVOU ·c40<>(SB), X10 + MOVOU ·c48<>(SB), X11 + + BLAMKA_ROUND_0(AX, 0, X8, X9, X10, X11) + BLAMKA_ROUND_0(AX, 16, X8, X9, X10, X11) + BLAMKA_ROUND_0(AX, 32, X8, X9, X10, X11) + BLAMKA_ROUND_0(AX, 48, X8, X9, X10, X11) + BLAMKA_ROUND_0(AX, 64, X8, X9, X10, X11) + BLAMKA_ROUND_0(AX, 80, X8, X9, X10, X11) + BLAMKA_ROUND_0(AX, 96, X8, X9, X10, X11) + BLAMKA_ROUND_0(AX, 112, X8, X9, X10, X11) + + BLAMKA_ROUND_1(AX, 0, X8, X9, X10, X11) + BLAMKA_ROUND_1(AX, 2, X8, X9, X10, X11) + BLAMKA_ROUND_1(AX, 4, X8, X9, X10, X11) + BLAMKA_ROUND_1(AX, 6, X8, X9, X10, X11) + BLAMKA_ROUND_1(AX, 8, X8, X9, X10, X11) + BLAMKA_ROUND_1(AX, 10, X8, X9, X10, X11) + BLAMKA_ROUND_1(AX, 12, X8, X9, X10, X11) + BLAMKA_ROUND_1(AX, 14, X8, X9, X10, X11) + RET + +// func mixBlocksSSE2(out, a, b, c *block) +TEXT ·mixBlocksSSE2(SB), 4, $0-32 + MOVQ out+0(FP), DX + MOVQ a+8(FP), AX + MOVQ b+16(FP), BX + MOVQ a+24(FP), CX + MOVQ $128, BP + +loop: + MOVOU 0(AX), X0 + MOVOU 0(BX), X1 + MOVOU 0(CX), X2 + PXOR X1, X0 + PXOR X2, X0 + MOVOU X0, 0(DX) + ADDQ $16, AX + ADDQ $16, BX + ADDQ $16, CX + ADDQ $16, DX + SUBQ $2, BP + JA loop + RET + +// func xorBlocksSSE2(out, a, b, c *block) +TEXT ·xorBlocksSSE2(SB), 4, $0-32 + MOVQ out+0(FP), DX + MOVQ a+8(FP), AX + MOVQ b+16(FP), BX + MOVQ a+24(FP), CX + MOVQ $128, BP + +loop: + MOVOU 0(AX), X0 + MOVOU 0(BX), X1 + MOVOU 0(CX), X2 + MOVOU 0(DX), X3 + PXOR X1, X0 + PXOR X2, X0 + PXOR X3, X0 + MOVOU X0, 0(DX) + ADDQ $16, AX + ADDQ $16, BX + ADDQ $16, CX + ADDQ $16, DX + SUBQ $2, BP + JA loop + RET diff --git a/vendor/golang.org/x/crypto/argon2/blamka_generic.go b/vendor/golang.org/x/crypto/argon2/blamka_generic.go new file mode 100644 index 0000000..a481b22 --- /dev/null +++ b/vendor/golang.org/x/crypto/argon2/blamka_generic.go @@ -0,0 +1,163 @@ +// Copyright 2017 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package argon2 + +var useSSE4 bool + +func processBlockGeneric(out, in1, in2 *block, xor bool) { + var t block + for i := range t { + t[i] = in1[i] ^ in2[i] + } + for i := 0; i < blockLength; i += 16 { + blamkaGeneric( + &t[i+0], &t[i+1], &t[i+2], &t[i+3], + &t[i+4], &t[i+5], &t[i+6], &t[i+7], + &t[i+8], &t[i+9], &t[i+10], &t[i+11], + &t[i+12], &t[i+13], &t[i+14], &t[i+15], + ) + } + for i := 0; i < blockLength/8; i += 2 { + blamkaGeneric( + &t[i], &t[i+1], &t[16+i], &t[16+i+1], + &t[32+i], &t[32+i+1], &t[48+i], &t[48+i+1], + &t[64+i], &t[64+i+1], &t[80+i], &t[80+i+1], + &t[96+i], &t[96+i+1], &t[112+i], &t[112+i+1], + ) + } + if xor { + for i := range t { + out[i] ^= in1[i] ^ in2[i] ^ t[i] + } + } else { + for i := range t { + out[i] = in1[i] ^ in2[i] ^ t[i] + } + } +} + +func blamkaGeneric(t00, t01, t02, t03, t04, t05, t06, t07, t08, t09, t10, t11, t12, t13, t14, t15 *uint64) { + v00, v01, v02, v03 := *t00, *t01, *t02, *t03 + v04, v05, v06, v07 := *t04, *t05, *t06, *t07 + v08, v09, v10, v11 := *t08, *t09, *t10, *t11 + v12, v13, v14, v15 := *t12, *t13, *t14, *t15 + + v00 += v04 + 2*uint64(uint32(v00))*uint64(uint32(v04)) + v12 ^= v00 + v12 = v12>>32 | v12<<32 + v08 += v12 + 2*uint64(uint32(v08))*uint64(uint32(v12)) + v04 ^= v08 + v04 = v04>>24 | v04<<40 + + v00 += v04 + 2*uint64(uint32(v00))*uint64(uint32(v04)) + v12 ^= v00 + v12 = v12>>16 | v12<<48 + v08 += v12 + 2*uint64(uint32(v08))*uint64(uint32(v12)) + v04 ^= v08 + v04 = v04>>63 | v04<<1 + + v01 += v05 + 2*uint64(uint32(v01))*uint64(uint32(v05)) + v13 ^= v01 + v13 = v13>>32 | v13<<32 + v09 += v13 + 2*uint64(uint32(v09))*uint64(uint32(v13)) + v05 ^= v09 + v05 = v05>>24 | v05<<40 + + v01 += v05 + 2*uint64(uint32(v01))*uint64(uint32(v05)) + v13 ^= v01 + v13 = v13>>16 | v13<<48 + v09 += v13 + 2*uint64(uint32(v09))*uint64(uint32(v13)) + v05 ^= v09 + v05 = v05>>63 | v05<<1 + + v02 += v06 + 2*uint64(uint32(v02))*uint64(uint32(v06)) + v14 ^= v02 + v14 = v14>>32 | v14<<32 + v10 += v14 + 2*uint64(uint32(v10))*uint64(uint32(v14)) + v06 ^= v10 + v06 = v06>>24 | v06<<40 + + v02 += v06 + 2*uint64(uint32(v02))*uint64(uint32(v06)) + v14 ^= v02 + v14 = v14>>16 | v14<<48 + v10 += v14 + 2*uint64(uint32(v10))*uint64(uint32(v14)) + v06 ^= v10 + v06 = v06>>63 | v06<<1 + + v03 += v07 + 2*uint64(uint32(v03))*uint64(uint32(v07)) + v15 ^= v03 + v15 = v15>>32 | v15<<32 + v11 += v15 + 2*uint64(uint32(v11))*uint64(uint32(v15)) + v07 ^= v11 + v07 = v07>>24 | v07<<40 + + v03 += v07 + 2*uint64(uint32(v03))*uint64(uint32(v07)) + v15 ^= v03 + v15 = v15>>16 | v15<<48 + v11 += v15 + 2*uint64(uint32(v11))*uint64(uint32(v15)) + v07 ^= v11 + v07 = v07>>63 | v07<<1 + + v00 += v05 + 2*uint64(uint32(v00))*uint64(uint32(v05)) + v15 ^= v00 + v15 = v15>>32 | v15<<32 + v10 += v15 + 2*uint64(uint32(v10))*uint64(uint32(v15)) + v05 ^= v10 + v05 = v05>>24 | v05<<40 + + v00 += v05 + 2*uint64(uint32(v00))*uint64(uint32(v05)) + v15 ^= v00 + v15 = v15>>16 | v15<<48 + v10 += v15 + 2*uint64(uint32(v10))*uint64(uint32(v15)) + v05 ^= v10 + v05 = v05>>63 | v05<<1 + + v01 += v06 + 2*uint64(uint32(v01))*uint64(uint32(v06)) + v12 ^= v01 + v12 = v12>>32 | v12<<32 + v11 += v12 + 2*uint64(uint32(v11))*uint64(uint32(v12)) + v06 ^= v11 + v06 = v06>>24 | v06<<40 + + v01 += v06 + 2*uint64(uint32(v01))*uint64(uint32(v06)) + v12 ^= v01 + v12 = v12>>16 | v12<<48 + v11 += v12 + 2*uint64(uint32(v11))*uint64(uint32(v12)) + v06 ^= v11 + v06 = v06>>63 | v06<<1 + + v02 += v07 + 2*uint64(uint32(v02))*uint64(uint32(v07)) + v13 ^= v02 + v13 = v13>>32 | v13<<32 + v08 += v13 + 2*uint64(uint32(v08))*uint64(uint32(v13)) + v07 ^= v08 + v07 = v07>>24 | v07<<40 + + v02 += v07 + 2*uint64(uint32(v02))*uint64(uint32(v07)) + v13 ^= v02 + v13 = v13>>16 | v13<<48 + v08 += v13 + 2*uint64(uint32(v08))*uint64(uint32(v13)) + v07 ^= v08 + v07 = v07>>63 | v07<<1 + + v03 += v04 + 2*uint64(uint32(v03))*uint64(uint32(v04)) + v14 ^= v03 + v14 = v14>>32 | v14<<32 + v09 += v14 + 2*uint64(uint32(v09))*uint64(uint32(v14)) + v04 ^= v09 + v04 = v04>>24 | v04<<40 + + v03 += v04 + 2*uint64(uint32(v03))*uint64(uint32(v04)) + v14 ^= v03 + v14 = v14>>16 | v14<<48 + v09 += v14 + 2*uint64(uint32(v09))*uint64(uint32(v14)) + v04 ^= v09 + v04 = v04>>63 | v04<<1 + + *t00, *t01, *t02, *t03 = v00, v01, v02, v03 + *t04, *t05, *t06, *t07 = v04, v05, v06, v07 + *t08, *t09, *t10, *t11 = v08, v09, v10, v11 + *t12, *t13, *t14, *t15 = v12, v13, v14, v15 +} diff --git a/vendor/golang.org/x/crypto/argon2/blamka_ref.go b/vendor/golang.org/x/crypto/argon2/blamka_ref.go new file mode 100644 index 0000000..baf7b55 --- /dev/null +++ b/vendor/golang.org/x/crypto/argon2/blamka_ref.go @@ -0,0 +1,15 @@ +// Copyright 2017 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build !amd64 appengine gccgo + +package argon2 + +func processBlock(out, in1, in2 *block) { + processBlockGeneric(out, in1, in2, false) +} + +func processBlockXOR(out, in1, in2 *block) { + processBlockGeneric(out, in1, in2, true) +} diff --git a/vendor/golang.org/x/crypto/blake2b/blake2b.go b/vendor/golang.org/x/crypto/blake2b/blake2b.go new file mode 100644 index 0000000..c160e1a --- /dev/null +++ b/vendor/golang.org/x/crypto/blake2b/blake2b.go @@ -0,0 +1,289 @@ +// Copyright 2016 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// Package blake2b implements the BLAKE2b hash algorithm defined by RFC 7693 +// and the extendable output function (XOF) BLAKE2Xb. +// +// For a detailed specification of BLAKE2b see https://blake2.net/blake2.pdf +// and for BLAKE2Xb see https://blake2.net/blake2x.pdf +// +// If you aren't sure which function you need, use BLAKE2b (Sum512 or New512). +// If you need a secret-key MAC (message authentication code), use the New512 +// function with a non-nil key. +// +// BLAKE2X is a construction to compute hash values larger than 64 bytes. It +// can produce hash values between 0 and 4 GiB. +package blake2b + +import ( + "encoding/binary" + "errors" + "hash" +) + +const ( + // The blocksize of BLAKE2b in bytes. + BlockSize = 128 + // The hash size of BLAKE2b-512 in bytes. + Size = 64 + // The hash size of BLAKE2b-384 in bytes. + Size384 = 48 + // The hash size of BLAKE2b-256 in bytes. + Size256 = 32 +) + +var ( + useAVX2 bool + useAVX bool + useSSE4 bool +) + +var ( + errKeySize = errors.New("blake2b: invalid key size") + errHashSize = errors.New("blake2b: invalid hash size") +) + +var iv = [8]uint64{ + 0x6a09e667f3bcc908, 0xbb67ae8584caa73b, 0x3c6ef372fe94f82b, 0xa54ff53a5f1d36f1, + 0x510e527fade682d1, 0x9b05688c2b3e6c1f, 0x1f83d9abfb41bd6b, 0x5be0cd19137e2179, +} + +// Sum512 returns the BLAKE2b-512 checksum of the data. +func Sum512(data []byte) [Size]byte { + var sum [Size]byte + checkSum(&sum, Size, data) + return sum +} + +// Sum384 returns the BLAKE2b-384 checksum of the data. +func Sum384(data []byte) [Size384]byte { + var sum [Size]byte + var sum384 [Size384]byte + checkSum(&sum, Size384, data) + copy(sum384[:], sum[:Size384]) + return sum384 +} + +// Sum256 returns the BLAKE2b-256 checksum of the data. +func Sum256(data []byte) [Size256]byte { + var sum [Size]byte + var sum256 [Size256]byte + checkSum(&sum, Size256, data) + copy(sum256[:], sum[:Size256]) + return sum256 +} + +// New512 returns a new hash.Hash computing the BLAKE2b-512 checksum. A non-nil +// key turns the hash into a MAC. The key must be between zero and 64 bytes long. +func New512(key []byte) (hash.Hash, error) { return newDigest(Size, key) } + +// New384 returns a new hash.Hash computing the BLAKE2b-384 checksum. A non-nil +// key turns the hash into a MAC. The key must be between zero and 64 bytes long. +func New384(key []byte) (hash.Hash, error) { return newDigest(Size384, key) } + +// New256 returns a new hash.Hash computing the BLAKE2b-256 checksum. A non-nil +// key turns the hash into a MAC. The key must be between zero and 64 bytes long. +func New256(key []byte) (hash.Hash, error) { return newDigest(Size256, key) } + +// New returns a new hash.Hash computing the BLAKE2b checksum with a custom length. +// A non-nil key turns the hash into a MAC. The key must be between zero and 64 bytes long. +// The hash size can be a value between 1 and 64 but it is highly recommended to use +// values equal or greater than: +// - 32 if BLAKE2b is used as a hash function (The key is zero bytes long). +// - 16 if BLAKE2b is used as a MAC function (The key is at least 16 bytes long). +// When the key is nil, the returned hash.Hash implements BinaryMarshaler +// and BinaryUnmarshaler for state (de)serialization as documented by hash.Hash. +func New(size int, key []byte) (hash.Hash, error) { return newDigest(size, key) } + +func newDigest(hashSize int, key []byte) (*digest, error) { + if hashSize < 1 || hashSize > Size { + return nil, errHashSize + } + if len(key) > Size { + return nil, errKeySize + } + d := &digest{ + size: hashSize, + keyLen: len(key), + } + copy(d.key[:], key) + d.Reset() + return d, nil +} + +func checkSum(sum *[Size]byte, hashSize int, data []byte) { + h := iv + h[0] ^= uint64(hashSize) | (1 << 16) | (1 << 24) + var c [2]uint64 + + if length := len(data); length > BlockSize { + n := length &^ (BlockSize - 1) + if length == n { + n -= BlockSize + } + hashBlocks(&h, &c, 0, data[:n]) + data = data[n:] + } + + var block [BlockSize]byte + offset := copy(block[:], data) + remaining := uint64(BlockSize - offset) + if c[0] < remaining { + c[1]-- + } + c[0] -= remaining + + hashBlocks(&h, &c, 0xFFFFFFFFFFFFFFFF, block[:]) + + for i, v := range h[:(hashSize+7)/8] { + binary.LittleEndian.PutUint64(sum[8*i:], v) + } +} + +type digest struct { + h [8]uint64 + c [2]uint64 + size int + block [BlockSize]byte + offset int + + key [BlockSize]byte + keyLen int +} + +const ( + magic = "b2b" + marshaledSize = len(magic) + 8*8 + 2*8 + 1 + BlockSize + 1 +) + +func (d *digest) MarshalBinary() ([]byte, error) { + if d.keyLen != 0 { + return nil, errors.New("crypto/blake2b: cannot marshal MACs") + } + b := make([]byte, 0, marshaledSize) + b = append(b, magic...) + for i := 0; i < 8; i++ { + b = appendUint64(b, d.h[i]) + } + b = appendUint64(b, d.c[0]) + b = appendUint64(b, d.c[1]) + // Maximum value for size is 64 + b = append(b, byte(d.size)) + b = append(b, d.block[:]...) + b = append(b, byte(d.offset)) + return b, nil +} + +func (d *digest) UnmarshalBinary(b []byte) error { + if len(b) < len(magic) || string(b[:len(magic)]) != magic { + return errors.New("crypto/blake2b: invalid hash state identifier") + } + if len(b) != marshaledSize { + return errors.New("crypto/blake2b: invalid hash state size") + } + b = b[len(magic):] + for i := 0; i < 8; i++ { + b, d.h[i] = consumeUint64(b) + } + b, d.c[0] = consumeUint64(b) + b, d.c[1] = consumeUint64(b) + d.size = int(b[0]) + b = b[1:] + copy(d.block[:], b[:BlockSize]) + b = b[BlockSize:] + d.offset = int(b[0]) + return nil +} + +func (d *digest) BlockSize() int { return BlockSize } + +func (d *digest) Size() int { return d.size } + +func (d *digest) Reset() { + d.h = iv + d.h[0] ^= uint64(d.size) | (uint64(d.keyLen) << 8) | (1 << 16) | (1 << 24) + d.offset, d.c[0], d.c[1] = 0, 0, 0 + if d.keyLen > 0 { + d.block = d.key + d.offset = BlockSize + } +} + +func (d *digest) Write(p []byte) (n int, err error) { + n = len(p) + + if d.offset > 0 { + remaining := BlockSize - d.offset + if n <= remaining { + d.offset += copy(d.block[d.offset:], p) + return + } + copy(d.block[d.offset:], p[:remaining]) + hashBlocks(&d.h, &d.c, 0, d.block[:]) + d.offset = 0 + p = p[remaining:] + } + + if length := len(p); length > BlockSize { + nn := length &^ (BlockSize - 1) + if length == nn { + nn -= BlockSize + } + hashBlocks(&d.h, &d.c, 0, p[:nn]) + p = p[nn:] + } + + if len(p) > 0 { + d.offset += copy(d.block[:], p) + } + + return +} + +func (d *digest) Sum(sum []byte) []byte { + var hash [Size]byte + d.finalize(&hash) + return append(sum, hash[:d.size]...) +} + +func (d *digest) finalize(hash *[Size]byte) { + var block [BlockSize]byte + copy(block[:], d.block[:d.offset]) + remaining := uint64(BlockSize - d.offset) + + c := d.c + if c[0] < remaining { + c[1]-- + } + c[0] -= remaining + + h := d.h + hashBlocks(&h, &c, 0xFFFFFFFFFFFFFFFF, block[:]) + + for i, v := range h { + binary.LittleEndian.PutUint64(hash[8*i:], v) + } +} + +func appendUint64(b []byte, x uint64) []byte { + var a [8]byte + binary.BigEndian.PutUint64(a[:], x) + return append(b, a[:]...) +} + +func appendUint32(b []byte, x uint32) []byte { + var a [4]byte + binary.BigEndian.PutUint32(a[:], x) + return append(b, a[:]...) +} + +func consumeUint64(b []byte) ([]byte, uint64) { + x := binary.BigEndian.Uint64(b) + return b[8:], x +} + +func consumeUint32(b []byte) ([]byte, uint32) { + x := binary.BigEndian.Uint32(b) + return b[4:], x +} diff --git a/vendor/golang.org/x/crypto/blake2b/blake2bAVX2_amd64.go b/vendor/golang.org/x/crypto/blake2b/blake2bAVX2_amd64.go new file mode 100644 index 0000000..4d31dd0 --- /dev/null +++ b/vendor/golang.org/x/crypto/blake2b/blake2bAVX2_amd64.go @@ -0,0 +1,37 @@ +// Copyright 2016 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build go1.7,amd64,!gccgo,!appengine + +package blake2b + +import "golang.org/x/sys/cpu" + +func init() { + useAVX2 = cpu.X86.HasAVX2 + useAVX = cpu.X86.HasAVX + useSSE4 = cpu.X86.HasSSE41 +} + +//go:noescape +func hashBlocksAVX2(h *[8]uint64, c *[2]uint64, flag uint64, blocks []byte) + +//go:noescape +func hashBlocksAVX(h *[8]uint64, c *[2]uint64, flag uint64, blocks []byte) + +//go:noescape +func hashBlocksSSE4(h *[8]uint64, c *[2]uint64, flag uint64, blocks []byte) + +func hashBlocks(h *[8]uint64, c *[2]uint64, flag uint64, blocks []byte) { + switch { + case useAVX2: + hashBlocksAVX2(h, c, flag, blocks) + case useAVX: + hashBlocksAVX(h, c, flag, blocks) + case useSSE4: + hashBlocksSSE4(h, c, flag, blocks) + default: + hashBlocksGeneric(h, c, flag, blocks) + } +} diff --git a/vendor/golang.org/x/crypto/blake2b/blake2bAVX2_amd64.s b/vendor/golang.org/x/crypto/blake2b/blake2bAVX2_amd64.s new file mode 100644 index 0000000..5593b1b --- /dev/null +++ b/vendor/golang.org/x/crypto/blake2b/blake2bAVX2_amd64.s @@ -0,0 +1,750 @@ +// Copyright 2016 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build go1.7,amd64,!gccgo,!appengine + +#include "textflag.h" + +DATA ·AVX2_iv0<>+0x00(SB)/8, $0x6a09e667f3bcc908 +DATA ·AVX2_iv0<>+0x08(SB)/8, $0xbb67ae8584caa73b +DATA ·AVX2_iv0<>+0x10(SB)/8, $0x3c6ef372fe94f82b +DATA ·AVX2_iv0<>+0x18(SB)/8, $0xa54ff53a5f1d36f1 +GLOBL ·AVX2_iv0<>(SB), (NOPTR+RODATA), $32 + +DATA ·AVX2_iv1<>+0x00(SB)/8, $0x510e527fade682d1 +DATA ·AVX2_iv1<>+0x08(SB)/8, $0x9b05688c2b3e6c1f +DATA ·AVX2_iv1<>+0x10(SB)/8, $0x1f83d9abfb41bd6b +DATA ·AVX2_iv1<>+0x18(SB)/8, $0x5be0cd19137e2179 +GLOBL ·AVX2_iv1<>(SB), (NOPTR+RODATA), $32 + +DATA ·AVX2_c40<>+0x00(SB)/8, $0x0201000706050403 +DATA ·AVX2_c40<>+0x08(SB)/8, $0x0a09080f0e0d0c0b +DATA ·AVX2_c40<>+0x10(SB)/8, $0x0201000706050403 +DATA ·AVX2_c40<>+0x18(SB)/8, $0x0a09080f0e0d0c0b +GLOBL ·AVX2_c40<>(SB), (NOPTR+RODATA), $32 + +DATA ·AVX2_c48<>+0x00(SB)/8, $0x0100070605040302 +DATA ·AVX2_c48<>+0x08(SB)/8, $0x09080f0e0d0c0b0a +DATA ·AVX2_c48<>+0x10(SB)/8, $0x0100070605040302 +DATA ·AVX2_c48<>+0x18(SB)/8, $0x09080f0e0d0c0b0a +GLOBL ·AVX2_c48<>(SB), (NOPTR+RODATA), $32 + +DATA ·AVX_iv0<>+0x00(SB)/8, $0x6a09e667f3bcc908 +DATA ·AVX_iv0<>+0x08(SB)/8, $0xbb67ae8584caa73b +GLOBL ·AVX_iv0<>(SB), (NOPTR+RODATA), $16 + +DATA ·AVX_iv1<>+0x00(SB)/8, $0x3c6ef372fe94f82b +DATA ·AVX_iv1<>+0x08(SB)/8, $0xa54ff53a5f1d36f1 +GLOBL ·AVX_iv1<>(SB), (NOPTR+RODATA), $16 + +DATA ·AVX_iv2<>+0x00(SB)/8, $0x510e527fade682d1 +DATA ·AVX_iv2<>+0x08(SB)/8, $0x9b05688c2b3e6c1f +GLOBL ·AVX_iv2<>(SB), (NOPTR+RODATA), $16 + +DATA ·AVX_iv3<>+0x00(SB)/8, $0x1f83d9abfb41bd6b +DATA ·AVX_iv3<>+0x08(SB)/8, $0x5be0cd19137e2179 +GLOBL ·AVX_iv3<>(SB), (NOPTR+RODATA), $16 + +DATA ·AVX_c40<>+0x00(SB)/8, $0x0201000706050403 +DATA ·AVX_c40<>+0x08(SB)/8, $0x0a09080f0e0d0c0b +GLOBL ·AVX_c40<>(SB), (NOPTR+RODATA), $16 + +DATA ·AVX_c48<>+0x00(SB)/8, $0x0100070605040302 +DATA ·AVX_c48<>+0x08(SB)/8, $0x09080f0e0d0c0b0a +GLOBL ·AVX_c48<>(SB), (NOPTR+RODATA), $16 + +#define VPERMQ_0x39_Y1_Y1 BYTE $0xc4; BYTE $0xe3; BYTE $0xfd; BYTE $0x00; BYTE $0xc9; BYTE $0x39 +#define VPERMQ_0x93_Y1_Y1 BYTE $0xc4; BYTE $0xe3; BYTE $0xfd; BYTE $0x00; BYTE $0xc9; BYTE $0x93 +#define VPERMQ_0x4E_Y2_Y2 BYTE $0xc4; BYTE $0xe3; BYTE $0xfd; BYTE $0x00; BYTE $0xd2; BYTE $0x4e +#define VPERMQ_0x93_Y3_Y3 BYTE $0xc4; BYTE $0xe3; BYTE $0xfd; BYTE $0x00; BYTE $0xdb; BYTE $0x93 +#define VPERMQ_0x39_Y3_Y3 BYTE $0xc4; BYTE $0xe3; BYTE $0xfd; BYTE $0x00; BYTE $0xdb; BYTE $0x39 + +#define ROUND_AVX2(m0, m1, m2, m3, t, c40, c48) \ + VPADDQ m0, Y0, Y0; \ + VPADDQ Y1, Y0, Y0; \ + VPXOR Y0, Y3, Y3; \ + VPSHUFD $-79, Y3, Y3; \ + VPADDQ Y3, Y2, Y2; \ + VPXOR Y2, Y1, Y1; \ + VPSHUFB c40, Y1, Y1; \ + VPADDQ m1, Y0, Y0; \ + VPADDQ Y1, Y0, Y0; \ + VPXOR Y0, Y3, Y3; \ + VPSHUFB c48, Y3, Y3; \ + VPADDQ Y3, Y2, Y2; \ + VPXOR Y2, Y1, Y1; \ + VPADDQ Y1, Y1, t; \ + VPSRLQ $63, Y1, Y1; \ + VPXOR t, Y1, Y1; \ + VPERMQ_0x39_Y1_Y1; \ + VPERMQ_0x4E_Y2_Y2; \ + VPERMQ_0x93_Y3_Y3; \ + VPADDQ m2, Y0, Y0; \ + VPADDQ Y1, Y0, Y0; \ + VPXOR Y0, Y3, Y3; \ + VPSHUFD $-79, Y3, Y3; \ + VPADDQ Y3, Y2, Y2; \ + VPXOR Y2, Y1, Y1; \ + VPSHUFB c40, Y1, Y1; \ + VPADDQ m3, Y0, Y0; \ + VPADDQ Y1, Y0, Y0; \ + VPXOR Y0, Y3, Y3; \ + VPSHUFB c48, Y3, Y3; \ + VPADDQ Y3, Y2, Y2; \ + VPXOR Y2, Y1, Y1; \ + VPADDQ Y1, Y1, t; \ + VPSRLQ $63, Y1, Y1; \ + VPXOR t, Y1, Y1; \ + VPERMQ_0x39_Y3_Y3; \ + VPERMQ_0x4E_Y2_Y2; \ + VPERMQ_0x93_Y1_Y1 + +#define VMOVQ_SI_X11_0 BYTE $0xC5; BYTE $0x7A; BYTE $0x7E; BYTE $0x1E +#define VMOVQ_SI_X12_0 BYTE $0xC5; BYTE $0x7A; BYTE $0x7E; BYTE $0x26 +#define VMOVQ_SI_X13_0 BYTE $0xC5; BYTE $0x7A; BYTE $0x7E; BYTE $0x2E +#define VMOVQ_SI_X14_0 BYTE $0xC5; BYTE $0x7A; BYTE $0x7E; BYTE $0x36 +#define VMOVQ_SI_X15_0 BYTE $0xC5; BYTE $0x7A; BYTE $0x7E; BYTE $0x3E + +#define VMOVQ_SI_X11(n) BYTE $0xC5; BYTE $0x7A; BYTE $0x7E; BYTE $0x5E; BYTE $n +#define VMOVQ_SI_X12(n) BYTE $0xC5; BYTE $0x7A; BYTE $0x7E; BYTE $0x66; BYTE $n +#define VMOVQ_SI_X13(n) BYTE $0xC5; BYTE $0x7A; BYTE $0x7E; BYTE $0x6E; BYTE $n +#define VMOVQ_SI_X14(n) BYTE $0xC5; BYTE $0x7A; BYTE $0x7E; BYTE $0x76; BYTE $n +#define VMOVQ_SI_X15(n) BYTE $0xC5; BYTE $0x7A; BYTE $0x7E; BYTE $0x7E; BYTE $n + +#define VPINSRQ_1_SI_X11_0 BYTE $0xC4; BYTE $0x63; BYTE $0xA1; BYTE $0x22; BYTE $0x1E; BYTE $0x01 +#define VPINSRQ_1_SI_X12_0 BYTE $0xC4; BYTE $0x63; BYTE $0x99; BYTE $0x22; BYTE $0x26; BYTE $0x01 +#define VPINSRQ_1_SI_X13_0 BYTE $0xC4; BYTE $0x63; BYTE $0x91; BYTE $0x22; BYTE $0x2E; BYTE $0x01 +#define VPINSRQ_1_SI_X14_0 BYTE $0xC4; BYTE $0x63; BYTE $0x89; BYTE $0x22; BYTE $0x36; BYTE $0x01 +#define VPINSRQ_1_SI_X15_0 BYTE $0xC4; BYTE $0x63; BYTE $0x81; BYTE $0x22; BYTE $0x3E; BYTE $0x01 + +#define VPINSRQ_1_SI_X11(n) BYTE $0xC4; BYTE $0x63; BYTE $0xA1; BYTE $0x22; BYTE $0x5E; BYTE $n; BYTE $0x01 +#define VPINSRQ_1_SI_X12(n) BYTE $0xC4; BYTE $0x63; BYTE $0x99; BYTE $0x22; BYTE $0x66; BYTE $n; BYTE $0x01 +#define VPINSRQ_1_SI_X13(n) BYTE $0xC4; BYTE $0x63; BYTE $0x91; BYTE $0x22; BYTE $0x6E; BYTE $n; BYTE $0x01 +#define VPINSRQ_1_SI_X14(n) BYTE $0xC4; BYTE $0x63; BYTE $0x89; BYTE $0x22; BYTE $0x76; BYTE $n; BYTE $0x01 +#define VPINSRQ_1_SI_X15(n) BYTE $0xC4; BYTE $0x63; BYTE $0x81; BYTE $0x22; BYTE $0x7E; BYTE $n; BYTE $0x01 + +#define VMOVQ_R8_X15 BYTE $0xC4; BYTE $0x41; BYTE $0xF9; BYTE $0x6E; BYTE $0xF8 +#define VPINSRQ_1_R9_X15 BYTE $0xC4; BYTE $0x43; BYTE $0x81; BYTE $0x22; BYTE $0xF9; BYTE $0x01 + +// load msg: Y12 = (i0, i1, i2, i3) +// i0, i1, i2, i3 must not be 0 +#define LOAD_MSG_AVX2_Y12(i0, i1, i2, i3) \ + VMOVQ_SI_X12(i0*8); \ + VMOVQ_SI_X11(i2*8); \ + VPINSRQ_1_SI_X12(i1*8); \ + VPINSRQ_1_SI_X11(i3*8); \ + VINSERTI128 $1, X11, Y12, Y12 + +// load msg: Y13 = (i0, i1, i2, i3) +// i0, i1, i2, i3 must not be 0 +#define LOAD_MSG_AVX2_Y13(i0, i1, i2, i3) \ + VMOVQ_SI_X13(i0*8); \ + VMOVQ_SI_X11(i2*8); \ + VPINSRQ_1_SI_X13(i1*8); \ + VPINSRQ_1_SI_X11(i3*8); \ + VINSERTI128 $1, X11, Y13, Y13 + +// load msg: Y14 = (i0, i1, i2, i3) +// i0, i1, i2, i3 must not be 0 +#define LOAD_MSG_AVX2_Y14(i0, i1, i2, i3) \ + VMOVQ_SI_X14(i0*8); \ + VMOVQ_SI_X11(i2*8); \ + VPINSRQ_1_SI_X14(i1*8); \ + VPINSRQ_1_SI_X11(i3*8); \ + VINSERTI128 $1, X11, Y14, Y14 + +// load msg: Y15 = (i0, i1, i2, i3) +// i0, i1, i2, i3 must not be 0 +#define LOAD_MSG_AVX2_Y15(i0, i1, i2, i3) \ + VMOVQ_SI_X15(i0*8); \ + VMOVQ_SI_X11(i2*8); \ + VPINSRQ_1_SI_X15(i1*8); \ + VPINSRQ_1_SI_X11(i3*8); \ + VINSERTI128 $1, X11, Y15, Y15 + +#define LOAD_MSG_AVX2_0_2_4_6_1_3_5_7_8_10_12_14_9_11_13_15() \ + VMOVQ_SI_X12_0; \ + VMOVQ_SI_X11(4*8); \ + VPINSRQ_1_SI_X12(2*8); \ + VPINSRQ_1_SI_X11(6*8); \ + VINSERTI128 $1, X11, Y12, Y12; \ + LOAD_MSG_AVX2_Y13(1, 3, 5, 7); \ + LOAD_MSG_AVX2_Y14(8, 10, 12, 14); \ + LOAD_MSG_AVX2_Y15(9, 11, 13, 15) + +#define LOAD_MSG_AVX2_14_4_9_13_10_8_15_6_1_0_11_5_12_2_7_3() \ + LOAD_MSG_AVX2_Y12(14, 4, 9, 13); \ + LOAD_MSG_AVX2_Y13(10, 8, 15, 6); \ + VMOVQ_SI_X11(11*8); \ + VPSHUFD $0x4E, 0*8(SI), X14; \ + VPINSRQ_1_SI_X11(5*8); \ + VINSERTI128 $1, X11, Y14, Y14; \ + LOAD_MSG_AVX2_Y15(12, 2, 7, 3) + +#define LOAD_MSG_AVX2_11_12_5_15_8_0_2_13_10_3_7_9_14_6_1_4() \ + VMOVQ_SI_X11(5*8); \ + VMOVDQU 11*8(SI), X12; \ + VPINSRQ_1_SI_X11(15*8); \ + VINSERTI128 $1, X11, Y12, Y12; \ + VMOVQ_SI_X13(8*8); \ + VMOVQ_SI_X11(2*8); \ + VPINSRQ_1_SI_X13_0; \ + VPINSRQ_1_SI_X11(13*8); \ + VINSERTI128 $1, X11, Y13, Y13; \ + LOAD_MSG_AVX2_Y14(10, 3, 7, 9); \ + LOAD_MSG_AVX2_Y15(14, 6, 1, 4) + +#define LOAD_MSG_AVX2_7_3_13_11_9_1_12_14_2_5_4_15_6_10_0_8() \ + LOAD_MSG_AVX2_Y12(7, 3, 13, 11); \ + LOAD_MSG_AVX2_Y13(9, 1, 12, 14); \ + LOAD_MSG_AVX2_Y14(2, 5, 4, 15); \ + VMOVQ_SI_X15(6*8); \ + VMOVQ_SI_X11_0; \ + VPINSRQ_1_SI_X15(10*8); \ + VPINSRQ_1_SI_X11(8*8); \ + VINSERTI128 $1, X11, Y15, Y15 + +#define LOAD_MSG_AVX2_9_5_2_10_0_7_4_15_14_11_6_3_1_12_8_13() \ + LOAD_MSG_AVX2_Y12(9, 5, 2, 10); \ + VMOVQ_SI_X13_0; \ + VMOVQ_SI_X11(4*8); \ + VPINSRQ_1_SI_X13(7*8); \ + VPINSRQ_1_SI_X11(15*8); \ + VINSERTI128 $1, X11, Y13, Y13; \ + LOAD_MSG_AVX2_Y14(14, 11, 6, 3); \ + LOAD_MSG_AVX2_Y15(1, 12, 8, 13) + +#define LOAD_MSG_AVX2_2_6_0_8_12_10_11_3_4_7_15_1_13_5_14_9() \ + VMOVQ_SI_X12(2*8); \ + VMOVQ_SI_X11_0; \ + VPINSRQ_1_SI_X12(6*8); \ + VPINSRQ_1_SI_X11(8*8); \ + VINSERTI128 $1, X11, Y12, Y12; \ + LOAD_MSG_AVX2_Y13(12, 10, 11, 3); \ + LOAD_MSG_AVX2_Y14(4, 7, 15, 1); \ + LOAD_MSG_AVX2_Y15(13, 5, 14, 9) + +#define LOAD_MSG_AVX2_12_1_14_4_5_15_13_10_0_6_9_8_7_3_2_11() \ + LOAD_MSG_AVX2_Y12(12, 1, 14, 4); \ + LOAD_MSG_AVX2_Y13(5, 15, 13, 10); \ + VMOVQ_SI_X14_0; \ + VPSHUFD $0x4E, 8*8(SI), X11; \ + VPINSRQ_1_SI_X14(6*8); \ + VINSERTI128 $1, X11, Y14, Y14; \ + LOAD_MSG_AVX2_Y15(7, 3, 2, 11) + +#define LOAD_MSG_AVX2_13_7_12_3_11_14_1_9_5_15_8_2_0_4_6_10() \ + LOAD_MSG_AVX2_Y12(13, 7, 12, 3); \ + LOAD_MSG_AVX2_Y13(11, 14, 1, 9); \ + LOAD_MSG_AVX2_Y14(5, 15, 8, 2); \ + VMOVQ_SI_X15_0; \ + VMOVQ_SI_X11(6*8); \ + VPINSRQ_1_SI_X15(4*8); \ + VPINSRQ_1_SI_X11(10*8); \ + VINSERTI128 $1, X11, Y15, Y15 + +#define LOAD_MSG_AVX2_6_14_11_0_15_9_3_8_12_13_1_10_2_7_4_5() \ + VMOVQ_SI_X12(6*8); \ + VMOVQ_SI_X11(11*8); \ + VPINSRQ_1_SI_X12(14*8); \ + VPINSRQ_1_SI_X11_0; \ + VINSERTI128 $1, X11, Y12, Y12; \ + LOAD_MSG_AVX2_Y13(15, 9, 3, 8); \ + VMOVQ_SI_X11(1*8); \ + VMOVDQU 12*8(SI), X14; \ + VPINSRQ_1_SI_X11(10*8); \ + VINSERTI128 $1, X11, Y14, Y14; \ + VMOVQ_SI_X15(2*8); \ + VMOVDQU 4*8(SI), X11; \ + VPINSRQ_1_SI_X15(7*8); \ + VINSERTI128 $1, X11, Y15, Y15 + +#define LOAD_MSG_AVX2_10_8_7_1_2_4_6_5_15_9_3_13_11_14_12_0() \ + LOAD_MSG_AVX2_Y12(10, 8, 7, 1); \ + VMOVQ_SI_X13(2*8); \ + VPSHUFD $0x4E, 5*8(SI), X11; \ + VPINSRQ_1_SI_X13(4*8); \ + VINSERTI128 $1, X11, Y13, Y13; \ + LOAD_MSG_AVX2_Y14(15, 9, 3, 13); \ + VMOVQ_SI_X15(11*8); \ + VMOVQ_SI_X11(12*8); \ + VPINSRQ_1_SI_X15(14*8); \ + VPINSRQ_1_SI_X11_0; \ + VINSERTI128 $1, X11, Y15, Y15 + +// func hashBlocksAVX2(h *[8]uint64, c *[2]uint64, flag uint64, blocks []byte) +TEXT ·hashBlocksAVX2(SB), 4, $320-48 // frame size = 288 + 32 byte alignment + MOVQ h+0(FP), AX + MOVQ c+8(FP), BX + MOVQ flag+16(FP), CX + MOVQ blocks_base+24(FP), SI + MOVQ blocks_len+32(FP), DI + + MOVQ SP, DX + MOVQ SP, R9 + ADDQ $31, R9 + ANDQ $~31, R9 + MOVQ R9, SP + + MOVQ CX, 16(SP) + XORQ CX, CX + MOVQ CX, 24(SP) + + VMOVDQU ·AVX2_c40<>(SB), Y4 + VMOVDQU ·AVX2_c48<>(SB), Y5 + + VMOVDQU 0(AX), Y8 + VMOVDQU 32(AX), Y9 + VMOVDQU ·AVX2_iv0<>(SB), Y6 + VMOVDQU ·AVX2_iv1<>(SB), Y7 + + MOVQ 0(BX), R8 + MOVQ 8(BX), R9 + MOVQ R9, 8(SP) + +loop: + ADDQ $128, R8 + MOVQ R8, 0(SP) + CMPQ R8, $128 + JGE noinc + INCQ R9 + MOVQ R9, 8(SP) + +noinc: + VMOVDQA Y8, Y0 + VMOVDQA Y9, Y1 + VMOVDQA Y6, Y2 + VPXOR 0(SP), Y7, Y3 + + LOAD_MSG_AVX2_0_2_4_6_1_3_5_7_8_10_12_14_9_11_13_15() + VMOVDQA Y12, 32(SP) + VMOVDQA Y13, 64(SP) + VMOVDQA Y14, 96(SP) + VMOVDQA Y15, 128(SP) + ROUND_AVX2(Y12, Y13, Y14, Y15, Y10, Y4, Y5) + LOAD_MSG_AVX2_14_4_9_13_10_8_15_6_1_0_11_5_12_2_7_3() + VMOVDQA Y12, 160(SP) + VMOVDQA Y13, 192(SP) + VMOVDQA Y14, 224(SP) + VMOVDQA Y15, 256(SP) + + ROUND_AVX2(Y12, Y13, Y14, Y15, Y10, Y4, Y5) + LOAD_MSG_AVX2_11_12_5_15_8_0_2_13_10_3_7_9_14_6_1_4() + ROUND_AVX2(Y12, Y13, Y14, Y15, Y10, Y4, Y5) + LOAD_MSG_AVX2_7_3_13_11_9_1_12_14_2_5_4_15_6_10_0_8() + ROUND_AVX2(Y12, Y13, Y14, Y15, Y10, Y4, Y5) + LOAD_MSG_AVX2_9_5_2_10_0_7_4_15_14_11_6_3_1_12_8_13() + ROUND_AVX2(Y12, Y13, Y14, Y15, Y10, Y4, Y5) + LOAD_MSG_AVX2_2_6_0_8_12_10_11_3_4_7_15_1_13_5_14_9() + ROUND_AVX2(Y12, Y13, Y14, Y15, Y10, Y4, Y5) + LOAD_MSG_AVX2_12_1_14_4_5_15_13_10_0_6_9_8_7_3_2_11() + ROUND_AVX2(Y12, Y13, Y14, Y15, Y10, Y4, Y5) + LOAD_MSG_AVX2_13_7_12_3_11_14_1_9_5_15_8_2_0_4_6_10() + ROUND_AVX2(Y12, Y13, Y14, Y15, Y10, Y4, Y5) + LOAD_MSG_AVX2_6_14_11_0_15_9_3_8_12_13_1_10_2_7_4_5() + ROUND_AVX2(Y12, Y13, Y14, Y15, Y10, Y4, Y5) + LOAD_MSG_AVX2_10_8_7_1_2_4_6_5_15_9_3_13_11_14_12_0() + ROUND_AVX2(Y12, Y13, Y14, Y15, Y10, Y4, Y5) + + ROUND_AVX2(32(SP), 64(SP), 96(SP), 128(SP), Y10, Y4, Y5) + ROUND_AVX2(160(SP), 192(SP), 224(SP), 256(SP), Y10, Y4, Y5) + + VPXOR Y0, Y8, Y8 + VPXOR Y1, Y9, Y9 + VPXOR Y2, Y8, Y8 + VPXOR Y3, Y9, Y9 + + LEAQ 128(SI), SI + SUBQ $128, DI + JNE loop + + MOVQ R8, 0(BX) + MOVQ R9, 8(BX) + + VMOVDQU Y8, 0(AX) + VMOVDQU Y9, 32(AX) + VZEROUPPER + + MOVQ DX, SP + RET + +#define VPUNPCKLQDQ_X2_X2_X15 BYTE $0xC5; BYTE $0x69; BYTE $0x6C; BYTE $0xFA +#define VPUNPCKLQDQ_X3_X3_X15 BYTE $0xC5; BYTE $0x61; BYTE $0x6C; BYTE $0xFB +#define VPUNPCKLQDQ_X7_X7_X15 BYTE $0xC5; BYTE $0x41; BYTE $0x6C; BYTE $0xFF +#define VPUNPCKLQDQ_X13_X13_X15 BYTE $0xC4; BYTE $0x41; BYTE $0x11; BYTE $0x6C; BYTE $0xFD +#define VPUNPCKLQDQ_X14_X14_X15 BYTE $0xC4; BYTE $0x41; BYTE $0x09; BYTE $0x6C; BYTE $0xFE + +#define VPUNPCKHQDQ_X15_X2_X2 BYTE $0xC4; BYTE $0xC1; BYTE $0x69; BYTE $0x6D; BYTE $0xD7 +#define VPUNPCKHQDQ_X15_X3_X3 BYTE $0xC4; BYTE $0xC1; BYTE $0x61; BYTE $0x6D; BYTE $0xDF +#define VPUNPCKHQDQ_X15_X6_X6 BYTE $0xC4; BYTE $0xC1; BYTE $0x49; BYTE $0x6D; BYTE $0xF7 +#define VPUNPCKHQDQ_X15_X7_X7 BYTE $0xC4; BYTE $0xC1; BYTE $0x41; BYTE $0x6D; BYTE $0xFF +#define VPUNPCKHQDQ_X15_X3_X2 BYTE $0xC4; BYTE $0xC1; BYTE $0x61; BYTE $0x6D; BYTE $0xD7 +#define VPUNPCKHQDQ_X15_X7_X6 BYTE $0xC4; BYTE $0xC1; BYTE $0x41; BYTE $0x6D; BYTE $0xF7 +#define VPUNPCKHQDQ_X15_X13_X3 BYTE $0xC4; BYTE $0xC1; BYTE $0x11; BYTE $0x6D; BYTE $0xDF +#define VPUNPCKHQDQ_X15_X13_X7 BYTE $0xC4; BYTE $0xC1; BYTE $0x11; BYTE $0x6D; BYTE $0xFF + +#define SHUFFLE_AVX() \ + VMOVDQA X6, X13; \ + VMOVDQA X2, X14; \ + VMOVDQA X4, X6; \ + VPUNPCKLQDQ_X13_X13_X15; \ + VMOVDQA X5, X4; \ + VMOVDQA X6, X5; \ + VPUNPCKHQDQ_X15_X7_X6; \ + VPUNPCKLQDQ_X7_X7_X15; \ + VPUNPCKHQDQ_X15_X13_X7; \ + VPUNPCKLQDQ_X3_X3_X15; \ + VPUNPCKHQDQ_X15_X2_X2; \ + VPUNPCKLQDQ_X14_X14_X15; \ + VPUNPCKHQDQ_X15_X3_X3; \ + +#define SHUFFLE_AVX_INV() \ + VMOVDQA X2, X13; \ + VMOVDQA X4, X14; \ + VPUNPCKLQDQ_X2_X2_X15; \ + VMOVDQA X5, X4; \ + VPUNPCKHQDQ_X15_X3_X2; \ + VMOVDQA X14, X5; \ + VPUNPCKLQDQ_X3_X3_X15; \ + VMOVDQA X6, X14; \ + VPUNPCKHQDQ_X15_X13_X3; \ + VPUNPCKLQDQ_X7_X7_X15; \ + VPUNPCKHQDQ_X15_X6_X6; \ + VPUNPCKLQDQ_X14_X14_X15; \ + VPUNPCKHQDQ_X15_X7_X7; \ + +#define HALF_ROUND_AVX(v0, v1, v2, v3, v4, v5, v6, v7, m0, m1, m2, m3, t0, c40, c48) \ + VPADDQ m0, v0, v0; \ + VPADDQ v2, v0, v0; \ + VPADDQ m1, v1, v1; \ + VPADDQ v3, v1, v1; \ + VPXOR v0, v6, v6; \ + VPXOR v1, v7, v7; \ + VPSHUFD $-79, v6, v6; \ + VPSHUFD $-79, v7, v7; \ + VPADDQ v6, v4, v4; \ + VPADDQ v7, v5, v5; \ + VPXOR v4, v2, v2; \ + VPXOR v5, v3, v3; \ + VPSHUFB c40, v2, v2; \ + VPSHUFB c40, v3, v3; \ + VPADDQ m2, v0, v0; \ + VPADDQ v2, v0, v0; \ + VPADDQ m3, v1, v1; \ + VPADDQ v3, v1, v1; \ + VPXOR v0, v6, v6; \ + VPXOR v1, v7, v7; \ + VPSHUFB c48, v6, v6; \ + VPSHUFB c48, v7, v7; \ + VPADDQ v6, v4, v4; \ + VPADDQ v7, v5, v5; \ + VPXOR v4, v2, v2; \ + VPXOR v5, v3, v3; \ + VPADDQ v2, v2, t0; \ + VPSRLQ $63, v2, v2; \ + VPXOR t0, v2, v2; \ + VPADDQ v3, v3, t0; \ + VPSRLQ $63, v3, v3; \ + VPXOR t0, v3, v3 + +// load msg: X12 = (i0, i1), X13 = (i2, i3), X14 = (i4, i5), X15 = (i6, i7) +// i0, i1, i2, i3, i4, i5, i6, i7 must not be 0 +#define LOAD_MSG_AVX(i0, i1, i2, i3, i4, i5, i6, i7) \ + VMOVQ_SI_X12(i0*8); \ + VMOVQ_SI_X13(i2*8); \ + VMOVQ_SI_X14(i4*8); \ + VMOVQ_SI_X15(i6*8); \ + VPINSRQ_1_SI_X12(i1*8); \ + VPINSRQ_1_SI_X13(i3*8); \ + VPINSRQ_1_SI_X14(i5*8); \ + VPINSRQ_1_SI_X15(i7*8) + +// load msg: X12 = (0, 2), X13 = (4, 6), X14 = (1, 3), X15 = (5, 7) +#define LOAD_MSG_AVX_0_2_4_6_1_3_5_7() \ + VMOVQ_SI_X12_0; \ + VMOVQ_SI_X13(4*8); \ + VMOVQ_SI_X14(1*8); \ + VMOVQ_SI_X15(5*8); \ + VPINSRQ_1_SI_X12(2*8); \ + VPINSRQ_1_SI_X13(6*8); \ + VPINSRQ_1_SI_X14(3*8); \ + VPINSRQ_1_SI_X15(7*8) + +// load msg: X12 = (1, 0), X13 = (11, 5), X14 = (12, 2), X15 = (7, 3) +#define LOAD_MSG_AVX_1_0_11_5_12_2_7_3() \ + VPSHUFD $0x4E, 0*8(SI), X12; \ + VMOVQ_SI_X13(11*8); \ + VMOVQ_SI_X14(12*8); \ + VMOVQ_SI_X15(7*8); \ + VPINSRQ_1_SI_X13(5*8); \ + VPINSRQ_1_SI_X14(2*8); \ + VPINSRQ_1_SI_X15(3*8) + +// load msg: X12 = (11, 12), X13 = (5, 15), X14 = (8, 0), X15 = (2, 13) +#define LOAD_MSG_AVX_11_12_5_15_8_0_2_13() \ + VMOVDQU 11*8(SI), X12; \ + VMOVQ_SI_X13(5*8); \ + VMOVQ_SI_X14(8*8); \ + VMOVQ_SI_X15(2*8); \ + VPINSRQ_1_SI_X13(15*8); \ + VPINSRQ_1_SI_X14_0; \ + VPINSRQ_1_SI_X15(13*8) + +// load msg: X12 = (2, 5), X13 = (4, 15), X14 = (6, 10), X15 = (0, 8) +#define LOAD_MSG_AVX_2_5_4_15_6_10_0_8() \ + VMOVQ_SI_X12(2*8); \ + VMOVQ_SI_X13(4*8); \ + VMOVQ_SI_X14(6*8); \ + VMOVQ_SI_X15_0; \ + VPINSRQ_1_SI_X12(5*8); \ + VPINSRQ_1_SI_X13(15*8); \ + VPINSRQ_1_SI_X14(10*8); \ + VPINSRQ_1_SI_X15(8*8) + +// load msg: X12 = (9, 5), X13 = (2, 10), X14 = (0, 7), X15 = (4, 15) +#define LOAD_MSG_AVX_9_5_2_10_0_7_4_15() \ + VMOVQ_SI_X12(9*8); \ + VMOVQ_SI_X13(2*8); \ + VMOVQ_SI_X14_0; \ + VMOVQ_SI_X15(4*8); \ + VPINSRQ_1_SI_X12(5*8); \ + VPINSRQ_1_SI_X13(10*8); \ + VPINSRQ_1_SI_X14(7*8); \ + VPINSRQ_1_SI_X15(15*8) + +// load msg: X12 = (2, 6), X13 = (0, 8), X14 = (12, 10), X15 = (11, 3) +#define LOAD_MSG_AVX_2_6_0_8_12_10_11_3() \ + VMOVQ_SI_X12(2*8); \ + VMOVQ_SI_X13_0; \ + VMOVQ_SI_X14(12*8); \ + VMOVQ_SI_X15(11*8); \ + VPINSRQ_1_SI_X12(6*8); \ + VPINSRQ_1_SI_X13(8*8); \ + VPINSRQ_1_SI_X14(10*8); \ + VPINSRQ_1_SI_X15(3*8) + +// load msg: X12 = (0, 6), X13 = (9, 8), X14 = (7, 3), X15 = (2, 11) +#define LOAD_MSG_AVX_0_6_9_8_7_3_2_11() \ + MOVQ 0*8(SI), X12; \ + VPSHUFD $0x4E, 8*8(SI), X13; \ + MOVQ 7*8(SI), X14; \ + MOVQ 2*8(SI), X15; \ + VPINSRQ_1_SI_X12(6*8); \ + VPINSRQ_1_SI_X14(3*8); \ + VPINSRQ_1_SI_X15(11*8) + +// load msg: X12 = (6, 14), X13 = (11, 0), X14 = (15, 9), X15 = (3, 8) +#define LOAD_MSG_AVX_6_14_11_0_15_9_3_8() \ + MOVQ 6*8(SI), X12; \ + MOVQ 11*8(SI), X13; \ + MOVQ 15*8(SI), X14; \ + MOVQ 3*8(SI), X15; \ + VPINSRQ_1_SI_X12(14*8); \ + VPINSRQ_1_SI_X13_0; \ + VPINSRQ_1_SI_X14(9*8); \ + VPINSRQ_1_SI_X15(8*8) + +// load msg: X12 = (5, 15), X13 = (8, 2), X14 = (0, 4), X15 = (6, 10) +#define LOAD_MSG_AVX_5_15_8_2_0_4_6_10() \ + MOVQ 5*8(SI), X12; \ + MOVQ 8*8(SI), X13; \ + MOVQ 0*8(SI), X14; \ + MOVQ 6*8(SI), X15; \ + VPINSRQ_1_SI_X12(15*8); \ + VPINSRQ_1_SI_X13(2*8); \ + VPINSRQ_1_SI_X14(4*8); \ + VPINSRQ_1_SI_X15(10*8) + +// load msg: X12 = (12, 13), X13 = (1, 10), X14 = (2, 7), X15 = (4, 5) +#define LOAD_MSG_AVX_12_13_1_10_2_7_4_5() \ + VMOVDQU 12*8(SI), X12; \ + MOVQ 1*8(SI), X13; \ + MOVQ 2*8(SI), X14; \ + VPINSRQ_1_SI_X13(10*8); \ + VPINSRQ_1_SI_X14(7*8); \ + VMOVDQU 4*8(SI), X15 + +// load msg: X12 = (15, 9), X13 = (3, 13), X14 = (11, 14), X15 = (12, 0) +#define LOAD_MSG_AVX_15_9_3_13_11_14_12_0() \ + MOVQ 15*8(SI), X12; \ + MOVQ 3*8(SI), X13; \ + MOVQ 11*8(SI), X14; \ + MOVQ 12*8(SI), X15; \ + VPINSRQ_1_SI_X12(9*8); \ + VPINSRQ_1_SI_X13(13*8); \ + VPINSRQ_1_SI_X14(14*8); \ + VPINSRQ_1_SI_X15_0 + +// func hashBlocksAVX(h *[8]uint64, c *[2]uint64, flag uint64, blocks []byte) +TEXT ·hashBlocksAVX(SB), 4, $288-48 // frame size = 272 + 16 byte alignment + MOVQ h+0(FP), AX + MOVQ c+8(FP), BX + MOVQ flag+16(FP), CX + MOVQ blocks_base+24(FP), SI + MOVQ blocks_len+32(FP), DI + + MOVQ SP, BP + MOVQ SP, R9 + ADDQ $15, R9 + ANDQ $~15, R9 + MOVQ R9, SP + + VMOVDQU ·AVX_c40<>(SB), X0 + VMOVDQU ·AVX_c48<>(SB), X1 + VMOVDQA X0, X8 + VMOVDQA X1, X9 + + VMOVDQU ·AVX_iv3<>(SB), X0 + VMOVDQA X0, 0(SP) + XORQ CX, 0(SP) // 0(SP) = ·AVX_iv3 ^ (CX || 0) + + VMOVDQU 0(AX), X10 + VMOVDQU 16(AX), X11 + VMOVDQU 32(AX), X2 + VMOVDQU 48(AX), X3 + + MOVQ 0(BX), R8 + MOVQ 8(BX), R9 + +loop: + ADDQ $128, R8 + CMPQ R8, $128 + JGE noinc + INCQ R9 + +noinc: + VMOVQ_R8_X15 + VPINSRQ_1_R9_X15 + + VMOVDQA X10, X0 + VMOVDQA X11, X1 + VMOVDQU ·AVX_iv0<>(SB), X4 + VMOVDQU ·AVX_iv1<>(SB), X5 + VMOVDQU ·AVX_iv2<>(SB), X6 + + VPXOR X15, X6, X6 + VMOVDQA 0(SP), X7 + + LOAD_MSG_AVX_0_2_4_6_1_3_5_7() + VMOVDQA X12, 16(SP) + VMOVDQA X13, 32(SP) + VMOVDQA X14, 48(SP) + VMOVDQA X15, 64(SP) + HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9) + SHUFFLE_AVX() + LOAD_MSG_AVX(8, 10, 12, 14, 9, 11, 13, 15) + VMOVDQA X12, 80(SP) + VMOVDQA X13, 96(SP) + VMOVDQA X14, 112(SP) + VMOVDQA X15, 128(SP) + HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9) + SHUFFLE_AVX_INV() + + LOAD_MSG_AVX(14, 4, 9, 13, 10, 8, 15, 6) + VMOVDQA X12, 144(SP) + VMOVDQA X13, 160(SP) + VMOVDQA X14, 176(SP) + VMOVDQA X15, 192(SP) + HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9) + SHUFFLE_AVX() + LOAD_MSG_AVX_1_0_11_5_12_2_7_3() + VMOVDQA X12, 208(SP) + VMOVDQA X13, 224(SP) + VMOVDQA X14, 240(SP) + VMOVDQA X15, 256(SP) + HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9) + SHUFFLE_AVX_INV() + + LOAD_MSG_AVX_11_12_5_15_8_0_2_13() + HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9) + SHUFFLE_AVX() + LOAD_MSG_AVX(10, 3, 7, 9, 14, 6, 1, 4) + HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9) + SHUFFLE_AVX_INV() + + LOAD_MSG_AVX(7, 3, 13, 11, 9, 1, 12, 14) + HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9) + SHUFFLE_AVX() + LOAD_MSG_AVX_2_5_4_15_6_10_0_8() + HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9) + SHUFFLE_AVX_INV() + + LOAD_MSG_AVX_9_5_2_10_0_7_4_15() + HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9) + SHUFFLE_AVX() + LOAD_MSG_AVX(14, 11, 6, 3, 1, 12, 8, 13) + HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9) + SHUFFLE_AVX_INV() + + LOAD_MSG_AVX_2_6_0_8_12_10_11_3() + HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9) + SHUFFLE_AVX() + LOAD_MSG_AVX(4, 7, 15, 1, 13, 5, 14, 9) + HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9) + SHUFFLE_AVX_INV() + + LOAD_MSG_AVX(12, 1, 14, 4, 5, 15, 13, 10) + HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9) + SHUFFLE_AVX() + LOAD_MSG_AVX_0_6_9_8_7_3_2_11() + HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9) + SHUFFLE_AVX_INV() + + LOAD_MSG_AVX(13, 7, 12, 3, 11, 14, 1, 9) + HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9) + SHUFFLE_AVX() + LOAD_MSG_AVX_5_15_8_2_0_4_6_10() + HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9) + SHUFFLE_AVX_INV() + + LOAD_MSG_AVX_6_14_11_0_15_9_3_8() + HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9) + SHUFFLE_AVX() + LOAD_MSG_AVX_12_13_1_10_2_7_4_5() + HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9) + SHUFFLE_AVX_INV() + + LOAD_MSG_AVX(10, 8, 7, 1, 2, 4, 6, 5) + HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9) + SHUFFLE_AVX() + LOAD_MSG_AVX_15_9_3_13_11_14_12_0() + HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, X12, X13, X14, X15, X15, X8, X9) + SHUFFLE_AVX_INV() + + HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, 16(SP), 32(SP), 48(SP), 64(SP), X15, X8, X9) + SHUFFLE_AVX() + HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, 80(SP), 96(SP), 112(SP), 128(SP), X15, X8, X9) + SHUFFLE_AVX_INV() + + HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, 144(SP), 160(SP), 176(SP), 192(SP), X15, X8, X9) + SHUFFLE_AVX() + HALF_ROUND_AVX(X0, X1, X2, X3, X4, X5, X6, X7, 208(SP), 224(SP), 240(SP), 256(SP), X15, X8, X9) + SHUFFLE_AVX_INV() + + VMOVDQU 32(AX), X14 + VMOVDQU 48(AX), X15 + VPXOR X0, X10, X10 + VPXOR X1, X11, X11 + VPXOR X2, X14, X14 + VPXOR X3, X15, X15 + VPXOR X4, X10, X10 + VPXOR X5, X11, X11 + VPXOR X6, X14, X2 + VPXOR X7, X15, X3 + VMOVDQU X2, 32(AX) + VMOVDQU X3, 48(AX) + + LEAQ 128(SI), SI + SUBQ $128, DI + JNE loop + + VMOVDQU X10, 0(AX) + VMOVDQU X11, 16(AX) + + MOVQ R8, 0(BX) + MOVQ R9, 8(BX) + VZEROUPPER + + MOVQ BP, SP + RET diff --git a/vendor/golang.org/x/crypto/blake2b/blake2b_amd64.go b/vendor/golang.org/x/crypto/blake2b/blake2b_amd64.go new file mode 100644 index 0000000..30e2fcd --- /dev/null +++ b/vendor/golang.org/x/crypto/blake2b/blake2b_amd64.go @@ -0,0 +1,24 @@ +// Copyright 2016 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build !go1.7,amd64,!gccgo,!appengine + +package blake2b + +import "golang.org/x/sys/cpu" + +func init() { + useSSE4 = cpu.X86.HasSSE41 +} + +//go:noescape +func hashBlocksSSE4(h *[8]uint64, c *[2]uint64, flag uint64, blocks []byte) + +func hashBlocks(h *[8]uint64, c *[2]uint64, flag uint64, blocks []byte) { + if useSSE4 { + hashBlocksSSE4(h, c, flag, blocks) + } else { + hashBlocksGeneric(h, c, flag, blocks) + } +} diff --git a/vendor/golang.org/x/crypto/blake2b/blake2b_amd64.s b/vendor/golang.org/x/crypto/blake2b/blake2b_amd64.s new file mode 100644 index 0000000..578e947 --- /dev/null +++ b/vendor/golang.org/x/crypto/blake2b/blake2b_amd64.s @@ -0,0 +1,281 @@ +// Copyright 2016 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build amd64,!gccgo,!appengine + +#include "textflag.h" + +DATA ·iv0<>+0x00(SB)/8, $0x6a09e667f3bcc908 +DATA ·iv0<>+0x08(SB)/8, $0xbb67ae8584caa73b +GLOBL ·iv0<>(SB), (NOPTR+RODATA), $16 + +DATA ·iv1<>+0x00(SB)/8, $0x3c6ef372fe94f82b +DATA ·iv1<>+0x08(SB)/8, $0xa54ff53a5f1d36f1 +GLOBL ·iv1<>(SB), (NOPTR+RODATA), $16 + +DATA ·iv2<>+0x00(SB)/8, $0x510e527fade682d1 +DATA ·iv2<>+0x08(SB)/8, $0x9b05688c2b3e6c1f +GLOBL ·iv2<>(SB), (NOPTR+RODATA), $16 + +DATA ·iv3<>+0x00(SB)/8, $0x1f83d9abfb41bd6b +DATA ·iv3<>+0x08(SB)/8, $0x5be0cd19137e2179 +GLOBL ·iv3<>(SB), (NOPTR+RODATA), $16 + +DATA ·c40<>+0x00(SB)/8, $0x0201000706050403 +DATA ·c40<>+0x08(SB)/8, $0x0a09080f0e0d0c0b +GLOBL ·c40<>(SB), (NOPTR+RODATA), $16 + +DATA ·c48<>+0x00(SB)/8, $0x0100070605040302 +DATA ·c48<>+0x08(SB)/8, $0x09080f0e0d0c0b0a +GLOBL ·c48<>(SB), (NOPTR+RODATA), $16 + +#define SHUFFLE(v2, v3, v4, v5, v6, v7, t1, t2) \ + MOVO v4, t1; \ + MOVO v5, v4; \ + MOVO t1, v5; \ + MOVO v6, t1; \ + PUNPCKLQDQ v6, t2; \ + PUNPCKHQDQ v7, v6; \ + PUNPCKHQDQ t2, v6; \ + PUNPCKLQDQ v7, t2; \ + MOVO t1, v7; \ + MOVO v2, t1; \ + PUNPCKHQDQ t2, v7; \ + PUNPCKLQDQ v3, t2; \ + PUNPCKHQDQ t2, v2; \ + PUNPCKLQDQ t1, t2; \ + PUNPCKHQDQ t2, v3 + +#define SHUFFLE_INV(v2, v3, v4, v5, v6, v7, t1, t2) \ + MOVO v4, t1; \ + MOVO v5, v4; \ + MOVO t1, v5; \ + MOVO v2, t1; \ + PUNPCKLQDQ v2, t2; \ + PUNPCKHQDQ v3, v2; \ + PUNPCKHQDQ t2, v2; \ + PUNPCKLQDQ v3, t2; \ + MOVO t1, v3; \ + MOVO v6, t1; \ + PUNPCKHQDQ t2, v3; \ + PUNPCKLQDQ v7, t2; \ + PUNPCKHQDQ t2, v6; \ + PUNPCKLQDQ t1, t2; \ + PUNPCKHQDQ t2, v7 + +#define HALF_ROUND(v0, v1, v2, v3, v4, v5, v6, v7, m0, m1, m2, m3, t0, c40, c48) \ + PADDQ m0, v0; \ + PADDQ m1, v1; \ + PADDQ v2, v0; \ + PADDQ v3, v1; \ + PXOR v0, v6; \ + PXOR v1, v7; \ + PSHUFD $0xB1, v6, v6; \ + PSHUFD $0xB1, v7, v7; \ + PADDQ v6, v4; \ + PADDQ v7, v5; \ + PXOR v4, v2; \ + PXOR v5, v3; \ + PSHUFB c40, v2; \ + PSHUFB c40, v3; \ + PADDQ m2, v0; \ + PADDQ m3, v1; \ + PADDQ v2, v0; \ + PADDQ v3, v1; \ + PXOR v0, v6; \ + PXOR v1, v7; \ + PSHUFB c48, v6; \ + PSHUFB c48, v7; \ + PADDQ v6, v4; \ + PADDQ v7, v5; \ + PXOR v4, v2; \ + PXOR v5, v3; \ + MOVOU v2, t0; \ + PADDQ v2, t0; \ + PSRLQ $63, v2; \ + PXOR t0, v2; \ + MOVOU v3, t0; \ + PADDQ v3, t0; \ + PSRLQ $63, v3; \ + PXOR t0, v3 + +#define LOAD_MSG(m0, m1, m2, m3, src, i0, i1, i2, i3, i4, i5, i6, i7) \ + MOVQ i0*8(src), m0; \ + PINSRQ $1, i1*8(src), m0; \ + MOVQ i2*8(src), m1; \ + PINSRQ $1, i3*8(src), m1; \ + MOVQ i4*8(src), m2; \ + PINSRQ $1, i5*8(src), m2; \ + MOVQ i6*8(src), m3; \ + PINSRQ $1, i7*8(src), m3 + +// func hashBlocksSSE4(h *[8]uint64, c *[2]uint64, flag uint64, blocks []byte) +TEXT ·hashBlocksSSE4(SB), 4, $288-48 // frame size = 272 + 16 byte alignment + MOVQ h+0(FP), AX + MOVQ c+8(FP), BX + MOVQ flag+16(FP), CX + MOVQ blocks_base+24(FP), SI + MOVQ blocks_len+32(FP), DI + + MOVQ SP, BP + MOVQ SP, R9 + ADDQ $15, R9 + ANDQ $~15, R9 + MOVQ R9, SP + + MOVOU ·iv3<>(SB), X0 + MOVO X0, 0(SP) + XORQ CX, 0(SP) // 0(SP) = ·iv3 ^ (CX || 0) + + MOVOU ·c40<>(SB), X13 + MOVOU ·c48<>(SB), X14 + + MOVOU 0(AX), X12 + MOVOU 16(AX), X15 + + MOVQ 0(BX), R8 + MOVQ 8(BX), R9 + +loop: + ADDQ $128, R8 + CMPQ R8, $128 + JGE noinc + INCQ R9 + +noinc: + MOVQ R8, X8 + PINSRQ $1, R9, X8 + + MOVO X12, X0 + MOVO X15, X1 + MOVOU 32(AX), X2 + MOVOU 48(AX), X3 + MOVOU ·iv0<>(SB), X4 + MOVOU ·iv1<>(SB), X5 + MOVOU ·iv2<>(SB), X6 + + PXOR X8, X6 + MOVO 0(SP), X7 + + LOAD_MSG(X8, X9, X10, X11, SI, 0, 2, 4, 6, 1, 3, 5, 7) + MOVO X8, 16(SP) + MOVO X9, 32(SP) + MOVO X10, 48(SP) + MOVO X11, 64(SP) + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14) + SHUFFLE(X2, X3, X4, X5, X6, X7, X8, X9) + LOAD_MSG(X8, X9, X10, X11, SI, 8, 10, 12, 14, 9, 11, 13, 15) + MOVO X8, 80(SP) + MOVO X9, 96(SP) + MOVO X10, 112(SP) + MOVO X11, 128(SP) + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14) + SHUFFLE_INV(X2, X3, X4, X5, X6, X7, X8, X9) + + LOAD_MSG(X8, X9, X10, X11, SI, 14, 4, 9, 13, 10, 8, 15, 6) + MOVO X8, 144(SP) + MOVO X9, 160(SP) + MOVO X10, 176(SP) + MOVO X11, 192(SP) + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14) + SHUFFLE(X2, X3, X4, X5, X6, X7, X8, X9) + LOAD_MSG(X8, X9, X10, X11, SI, 1, 0, 11, 5, 12, 2, 7, 3) + MOVO X8, 208(SP) + MOVO X9, 224(SP) + MOVO X10, 240(SP) + MOVO X11, 256(SP) + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14) + SHUFFLE_INV(X2, X3, X4, X5, X6, X7, X8, X9) + + LOAD_MSG(X8, X9, X10, X11, SI, 11, 12, 5, 15, 8, 0, 2, 13) + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14) + SHUFFLE(X2, X3, X4, X5, X6, X7, X8, X9) + LOAD_MSG(X8, X9, X10, X11, SI, 10, 3, 7, 9, 14, 6, 1, 4) + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14) + SHUFFLE_INV(X2, X3, X4, X5, X6, X7, X8, X9) + + LOAD_MSG(X8, X9, X10, X11, SI, 7, 3, 13, 11, 9, 1, 12, 14) + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14) + SHUFFLE(X2, X3, X4, X5, X6, X7, X8, X9) + LOAD_MSG(X8, X9, X10, X11, SI, 2, 5, 4, 15, 6, 10, 0, 8) + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14) + SHUFFLE_INV(X2, X3, X4, X5, X6, X7, X8, X9) + + LOAD_MSG(X8, X9, X10, X11, SI, 9, 5, 2, 10, 0, 7, 4, 15) + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14) + SHUFFLE(X2, X3, X4, X5, X6, X7, X8, X9) + LOAD_MSG(X8, X9, X10, X11, SI, 14, 11, 6, 3, 1, 12, 8, 13) + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14) + SHUFFLE_INV(X2, X3, X4, X5, X6, X7, X8, X9) + + LOAD_MSG(X8, X9, X10, X11, SI, 2, 6, 0, 8, 12, 10, 11, 3) + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14) + SHUFFLE(X2, X3, X4, X5, X6, X7, X8, X9) + LOAD_MSG(X8, X9, X10, X11, SI, 4, 7, 15, 1, 13, 5, 14, 9) + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14) + SHUFFLE_INV(X2, X3, X4, X5, X6, X7, X8, X9) + + LOAD_MSG(X8, X9, X10, X11, SI, 12, 1, 14, 4, 5, 15, 13, 10) + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14) + SHUFFLE(X2, X3, X4, X5, X6, X7, X8, X9) + LOAD_MSG(X8, X9, X10, X11, SI, 0, 6, 9, 8, 7, 3, 2, 11) + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14) + SHUFFLE_INV(X2, X3, X4, X5, X6, X7, X8, X9) + + LOAD_MSG(X8, X9, X10, X11, SI, 13, 7, 12, 3, 11, 14, 1, 9) + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14) + SHUFFLE(X2, X3, X4, X5, X6, X7, X8, X9) + LOAD_MSG(X8, X9, X10, X11, SI, 5, 15, 8, 2, 0, 4, 6, 10) + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14) + SHUFFLE_INV(X2, X3, X4, X5, X6, X7, X8, X9) + + LOAD_MSG(X8, X9, X10, X11, SI, 6, 14, 11, 0, 15, 9, 3, 8) + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14) + SHUFFLE(X2, X3, X4, X5, X6, X7, X8, X9) + LOAD_MSG(X8, X9, X10, X11, SI, 12, 13, 1, 10, 2, 7, 4, 5) + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14) + SHUFFLE_INV(X2, X3, X4, X5, X6, X7, X8, X9) + + LOAD_MSG(X8, X9, X10, X11, SI, 10, 8, 7, 1, 2, 4, 6, 5) + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14) + SHUFFLE(X2, X3, X4, X5, X6, X7, X8, X9) + LOAD_MSG(X8, X9, X10, X11, SI, 15, 9, 3, 13, 11, 14, 12, 0) + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, X8, X9, X10, X11, X11, X13, X14) + SHUFFLE_INV(X2, X3, X4, X5, X6, X7, X8, X9) + + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, 16(SP), 32(SP), 48(SP), 64(SP), X11, X13, X14) + SHUFFLE(X2, X3, X4, X5, X6, X7, X8, X9) + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, 80(SP), 96(SP), 112(SP), 128(SP), X11, X13, X14) + SHUFFLE_INV(X2, X3, X4, X5, X6, X7, X8, X9) + + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, 144(SP), 160(SP), 176(SP), 192(SP), X11, X13, X14) + SHUFFLE(X2, X3, X4, X5, X6, X7, X8, X9) + HALF_ROUND(X0, X1, X2, X3, X4, X5, X6, X7, 208(SP), 224(SP), 240(SP), 256(SP), X11, X13, X14) + SHUFFLE_INV(X2, X3, X4, X5, X6, X7, X8, X9) + + MOVOU 32(AX), X10 + MOVOU 48(AX), X11 + PXOR X0, X12 + PXOR X1, X15 + PXOR X2, X10 + PXOR X3, X11 + PXOR X4, X12 + PXOR X5, X15 + PXOR X6, X10 + PXOR X7, X11 + MOVOU X10, 32(AX) + MOVOU X11, 48(AX) + + LEAQ 128(SI), SI + SUBQ $128, DI + JNE loop + + MOVOU X12, 0(AX) + MOVOU X15, 16(AX) + + MOVQ R8, 0(BX) + MOVQ R9, 8(BX) + + MOVQ BP, SP + RET diff --git a/vendor/golang.org/x/crypto/blake2b/blake2b_generic.go b/vendor/golang.org/x/crypto/blake2b/blake2b_generic.go new file mode 100644 index 0000000..3168a8a --- /dev/null +++ b/vendor/golang.org/x/crypto/blake2b/blake2b_generic.go @@ -0,0 +1,182 @@ +// Copyright 2016 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package blake2b + +import ( + "encoding/binary" + "math/bits" +) + +// the precomputed values for BLAKE2b +// there are 12 16-byte arrays - one for each round +// the entries are calculated from the sigma constants. +var precomputed = [12][16]byte{ + {0, 2, 4, 6, 1, 3, 5, 7, 8, 10, 12, 14, 9, 11, 13, 15}, + {14, 4, 9, 13, 10, 8, 15, 6, 1, 0, 11, 5, 12, 2, 7, 3}, + {11, 12, 5, 15, 8, 0, 2, 13, 10, 3, 7, 9, 14, 6, 1, 4}, + {7, 3, 13, 11, 9, 1, 12, 14, 2, 5, 4, 15, 6, 10, 0, 8}, + {9, 5, 2, 10, 0, 7, 4, 15, 14, 11, 6, 3, 1, 12, 8, 13}, + {2, 6, 0, 8, 12, 10, 11, 3, 4, 7, 15, 1, 13, 5, 14, 9}, + {12, 1, 14, 4, 5, 15, 13, 10, 0, 6, 9, 8, 7, 3, 2, 11}, + {13, 7, 12, 3, 11, 14, 1, 9, 5, 15, 8, 2, 0, 4, 6, 10}, + {6, 14, 11, 0, 15, 9, 3, 8, 12, 13, 1, 10, 2, 7, 4, 5}, + {10, 8, 7, 1, 2, 4, 6, 5, 15, 9, 3, 13, 11, 14, 12, 0}, + {0, 2, 4, 6, 1, 3, 5, 7, 8, 10, 12, 14, 9, 11, 13, 15}, // equal to the first + {14, 4, 9, 13, 10, 8, 15, 6, 1, 0, 11, 5, 12, 2, 7, 3}, // equal to the second +} + +func hashBlocksGeneric(h *[8]uint64, c *[2]uint64, flag uint64, blocks []byte) { + var m [16]uint64 + c0, c1 := c[0], c[1] + + for i := 0; i < len(blocks); { + c0 += BlockSize + if c0 < BlockSize { + c1++ + } + + v0, v1, v2, v3, v4, v5, v6, v7 := h[0], h[1], h[2], h[3], h[4], h[5], h[6], h[7] + v8, v9, v10, v11, v12, v13, v14, v15 := iv[0], iv[1], iv[2], iv[3], iv[4], iv[5], iv[6], iv[7] + v12 ^= c0 + v13 ^= c1 + v14 ^= flag + + for j := range m { + m[j] = binary.LittleEndian.Uint64(blocks[i:]) + i += 8 + } + + for j := range precomputed { + s := &(precomputed[j]) + + v0 += m[s[0]] + v0 += v4 + v12 ^= v0 + v12 = bits.RotateLeft64(v12, -32) + v8 += v12 + v4 ^= v8 + v4 = bits.RotateLeft64(v4, -24) + v1 += m[s[1]] + v1 += v5 + v13 ^= v1 + v13 = bits.RotateLeft64(v13, -32) + v9 += v13 + v5 ^= v9 + v5 = bits.RotateLeft64(v5, -24) + v2 += m[s[2]] + v2 += v6 + v14 ^= v2 + v14 = bits.RotateLeft64(v14, -32) + v10 += v14 + v6 ^= v10 + v6 = bits.RotateLeft64(v6, -24) + v3 += m[s[3]] + v3 += v7 + v15 ^= v3 + v15 = bits.RotateLeft64(v15, -32) + v11 += v15 + v7 ^= v11 + v7 = bits.RotateLeft64(v7, -24) + + v0 += m[s[4]] + v0 += v4 + v12 ^= v0 + v12 = bits.RotateLeft64(v12, -16) + v8 += v12 + v4 ^= v8 + v4 = bits.RotateLeft64(v4, -63) + v1 += m[s[5]] + v1 += v5 + v13 ^= v1 + v13 = bits.RotateLeft64(v13, -16) + v9 += v13 + v5 ^= v9 + v5 = bits.RotateLeft64(v5, -63) + v2 += m[s[6]] + v2 += v6 + v14 ^= v2 + v14 = bits.RotateLeft64(v14, -16) + v10 += v14 + v6 ^= v10 + v6 = bits.RotateLeft64(v6, -63) + v3 += m[s[7]] + v3 += v7 + v15 ^= v3 + v15 = bits.RotateLeft64(v15, -16) + v11 += v15 + v7 ^= v11 + v7 = bits.RotateLeft64(v7, -63) + + v0 += m[s[8]] + v0 += v5 + v15 ^= v0 + v15 = bits.RotateLeft64(v15, -32) + v10 += v15 + v5 ^= v10 + v5 = bits.RotateLeft64(v5, -24) + v1 += m[s[9]] + v1 += v6 + v12 ^= v1 + v12 = bits.RotateLeft64(v12, -32) + v11 += v12 + v6 ^= v11 + v6 = bits.RotateLeft64(v6, -24) + v2 += m[s[10]] + v2 += v7 + v13 ^= v2 + v13 = bits.RotateLeft64(v13, -32) + v8 += v13 + v7 ^= v8 + v7 = bits.RotateLeft64(v7, -24) + v3 += m[s[11]] + v3 += v4 + v14 ^= v3 + v14 = bits.RotateLeft64(v14, -32) + v9 += v14 + v4 ^= v9 + v4 = bits.RotateLeft64(v4, -24) + + v0 += m[s[12]] + v0 += v5 + v15 ^= v0 + v15 = bits.RotateLeft64(v15, -16) + v10 += v15 + v5 ^= v10 + v5 = bits.RotateLeft64(v5, -63) + v1 += m[s[13]] + v1 += v6 + v12 ^= v1 + v12 = bits.RotateLeft64(v12, -16) + v11 += v12 + v6 ^= v11 + v6 = bits.RotateLeft64(v6, -63) + v2 += m[s[14]] + v2 += v7 + v13 ^= v2 + v13 = bits.RotateLeft64(v13, -16) + v8 += v13 + v7 ^= v8 + v7 = bits.RotateLeft64(v7, -63) + v3 += m[s[15]] + v3 += v4 + v14 ^= v3 + v14 = bits.RotateLeft64(v14, -16) + v9 += v14 + v4 ^= v9 + v4 = bits.RotateLeft64(v4, -63) + + } + + h[0] ^= v0 ^ v8 + h[1] ^= v1 ^ v9 + h[2] ^= v2 ^ v10 + h[3] ^= v3 ^ v11 + h[4] ^= v4 ^ v12 + h[5] ^= v5 ^ v13 + h[6] ^= v6 ^ v14 + h[7] ^= v7 ^ v15 + } + c[0], c[1] = c0, c1 +} diff --git a/vendor/golang.org/x/crypto/blake2b/blake2b_ref.go b/vendor/golang.org/x/crypto/blake2b/blake2b_ref.go new file mode 100644 index 0000000..da156a1 --- /dev/null +++ b/vendor/golang.org/x/crypto/blake2b/blake2b_ref.go @@ -0,0 +1,11 @@ +// Copyright 2016 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build !amd64 appengine gccgo + +package blake2b + +func hashBlocks(h *[8]uint64, c *[2]uint64, flag uint64, blocks []byte) { + hashBlocksGeneric(h, c, flag, blocks) +} diff --git a/vendor/golang.org/x/crypto/blake2b/blake2x.go b/vendor/golang.org/x/crypto/blake2b/blake2x.go new file mode 100644 index 0000000..52c414d --- /dev/null +++ b/vendor/golang.org/x/crypto/blake2b/blake2x.go @@ -0,0 +1,177 @@ +// Copyright 2017 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package blake2b + +import ( + "encoding/binary" + "errors" + "io" +) + +// XOF defines the interface to hash functions that +// support arbitrary-length output. +type XOF interface { + // Write absorbs more data into the hash's state. It panics if called + // after Read. + io.Writer + + // Read reads more output from the hash. It returns io.EOF if the limit + // has been reached. + io.Reader + + // Clone returns a copy of the XOF in its current state. + Clone() XOF + + // Reset resets the XOF to its initial state. + Reset() +} + +// OutputLengthUnknown can be used as the size argument to NewXOF to indicate +// the length of the output is not known in advance. +const OutputLengthUnknown = 0 + +// magicUnknownOutputLength is a magic value for the output size that indicates +// an unknown number of output bytes. +const magicUnknownOutputLength = (1 << 32) - 1 + +// maxOutputLength is the absolute maximum number of bytes to produce when the +// number of output bytes is unknown. +const maxOutputLength = (1 << 32) * 64 + +// NewXOF creates a new variable-output-length hash. The hash either produce a +// known number of bytes (1 <= size < 2**32-1), or an unknown number of bytes +// (size == OutputLengthUnknown). In the latter case, an absolute limit of +// 256GiB applies. +// +// A non-nil key turns the hash into a MAC. The key must between +// zero and 32 bytes long. +func NewXOF(size uint32, key []byte) (XOF, error) { + if len(key) > Size { + return nil, errKeySize + } + if size == magicUnknownOutputLength { + // 2^32-1 indicates an unknown number of bytes and thus isn't a + // valid length. + return nil, errors.New("blake2b: XOF length too large") + } + if size == OutputLengthUnknown { + size = magicUnknownOutputLength + } + x := &xof{ + d: digest{ + size: Size, + keyLen: len(key), + }, + length: size, + } + copy(x.d.key[:], key) + x.Reset() + return x, nil +} + +type xof struct { + d digest + length uint32 + remaining uint64 + cfg, root, block [Size]byte + offset int + nodeOffset uint32 + readMode bool +} + +func (x *xof) Write(p []byte) (n int, err error) { + if x.readMode { + panic("blake2b: write to XOF after read") + } + return x.d.Write(p) +} + +func (x *xof) Clone() XOF { + clone := *x + return &clone +} + +func (x *xof) Reset() { + x.cfg[0] = byte(Size) + binary.LittleEndian.PutUint32(x.cfg[4:], uint32(Size)) // leaf length + binary.LittleEndian.PutUint32(x.cfg[12:], x.length) // XOF length + x.cfg[17] = byte(Size) // inner hash size + + x.d.Reset() + x.d.h[1] ^= uint64(x.length) << 32 + + x.remaining = uint64(x.length) + if x.remaining == magicUnknownOutputLength { + x.remaining = maxOutputLength + } + x.offset, x.nodeOffset = 0, 0 + x.readMode = false +} + +func (x *xof) Read(p []byte) (n int, err error) { + if !x.readMode { + x.d.finalize(&x.root) + x.readMode = true + } + + if x.remaining == 0 { + return 0, io.EOF + } + + n = len(p) + if uint64(n) > x.remaining { + n = int(x.remaining) + p = p[:n] + } + + if x.offset > 0 { + blockRemaining := Size - x.offset + if n < blockRemaining { + x.offset += copy(p, x.block[x.offset:]) + x.remaining -= uint64(n) + return + } + copy(p, x.block[x.offset:]) + p = p[blockRemaining:] + x.offset = 0 + x.remaining -= uint64(blockRemaining) + } + + for len(p) >= Size { + binary.LittleEndian.PutUint32(x.cfg[8:], x.nodeOffset) + x.nodeOffset++ + + x.d.initConfig(&x.cfg) + x.d.Write(x.root[:]) + x.d.finalize(&x.block) + + copy(p, x.block[:]) + p = p[Size:] + x.remaining -= uint64(Size) + } + + if todo := len(p); todo > 0 { + if x.remaining < uint64(Size) { + x.cfg[0] = byte(x.remaining) + } + binary.LittleEndian.PutUint32(x.cfg[8:], x.nodeOffset) + x.nodeOffset++ + + x.d.initConfig(&x.cfg) + x.d.Write(x.root[:]) + x.d.finalize(&x.block) + + x.offset = copy(p, x.block[:todo]) + x.remaining -= uint64(todo) + } + return +} + +func (d *digest) initConfig(cfg *[Size]byte) { + d.offset, d.c[0], d.c[1] = 0, 0, 0 + for i := range d.h { + d.h[i] = iv[i] ^ binary.LittleEndian.Uint64(cfg[i*8:]) + } +} diff --git a/vendor/golang.org/x/crypto/blake2b/register.go b/vendor/golang.org/x/crypto/blake2b/register.go new file mode 100644 index 0000000..efd689a --- /dev/null +++ b/vendor/golang.org/x/crypto/blake2b/register.go @@ -0,0 +1,32 @@ +// Copyright 2017 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build go1.9 + +package blake2b + +import ( + "crypto" + "hash" +) + +func init() { + newHash256 := func() hash.Hash { + h, _ := New256(nil) + return h + } + newHash384 := func() hash.Hash { + h, _ := New384(nil) + return h + } + + newHash512 := func() hash.Hash { + h, _ := New512(nil) + return h + } + + crypto.RegisterHash(crypto.BLAKE2b_256, newHash256) + crypto.RegisterHash(crypto.BLAKE2b_384, newHash384) + crypto.RegisterHash(crypto.BLAKE2b_512, newHash512) +} diff --git a/vendor/golang.org/x/crypto/chacha20poly1305/chacha20poly1305.go b/vendor/golang.org/x/crypto/chacha20poly1305/chacha20poly1305.go new file mode 100644 index 0000000..bbb86ef --- /dev/null +++ b/vendor/golang.org/x/crypto/chacha20poly1305/chacha20poly1305.go @@ -0,0 +1,101 @@ +// Copyright 2016 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// Package chacha20poly1305 implements the ChaCha20-Poly1305 AEAD as specified in RFC 7539, +// and its extended nonce variant XChaCha20-Poly1305. +package chacha20poly1305 // import "golang.org/x/crypto/chacha20poly1305" + +import ( + "crypto/cipher" + "encoding/binary" + "errors" +) + +const ( + // KeySize is the size of the key used by this AEAD, in bytes. + KeySize = 32 + + // NonceSize is the size of the nonce used with the standard variant of this + // AEAD, in bytes. + // + // Note that this is too short to be safely generated at random if the same + // key is reused more than 2³² times. + NonceSize = 12 + + // NonceSizeX is the size of the nonce used with the XChaCha20-Poly1305 + // variant of this AEAD, in bytes. + NonceSizeX = 24 +) + +type chacha20poly1305 struct { + key [8]uint32 +} + +// New returns a ChaCha20-Poly1305 AEAD that uses the given 256-bit key. +func New(key []byte) (cipher.AEAD, error) { + if len(key) != KeySize { + return nil, errors.New("chacha20poly1305: bad key length") + } + ret := new(chacha20poly1305) + ret.key[0] = binary.LittleEndian.Uint32(key[0:4]) + ret.key[1] = binary.LittleEndian.Uint32(key[4:8]) + ret.key[2] = binary.LittleEndian.Uint32(key[8:12]) + ret.key[3] = binary.LittleEndian.Uint32(key[12:16]) + ret.key[4] = binary.LittleEndian.Uint32(key[16:20]) + ret.key[5] = binary.LittleEndian.Uint32(key[20:24]) + ret.key[6] = binary.LittleEndian.Uint32(key[24:28]) + ret.key[7] = binary.LittleEndian.Uint32(key[28:32]) + return ret, nil +} + +func (c *chacha20poly1305) NonceSize() int { + return NonceSize +} + +func (c *chacha20poly1305) Overhead() int { + return 16 +} + +func (c *chacha20poly1305) Seal(dst, nonce, plaintext, additionalData []byte) []byte { + if len(nonce) != NonceSize { + panic("chacha20poly1305: bad nonce length passed to Seal") + } + + if uint64(len(plaintext)) > (1<<38)-64 { + panic("chacha20poly1305: plaintext too large") + } + + return c.seal(dst, nonce, plaintext, additionalData) +} + +var errOpen = errors.New("chacha20poly1305: message authentication failed") + +func (c *chacha20poly1305) Open(dst, nonce, ciphertext, additionalData []byte) ([]byte, error) { + if len(nonce) != NonceSize { + panic("chacha20poly1305: bad nonce length passed to Open") + } + if len(ciphertext) < 16 { + return nil, errOpen + } + if uint64(len(ciphertext)) > (1<<38)-48 { + panic("chacha20poly1305: ciphertext too large") + } + + return c.open(dst, nonce, ciphertext, additionalData) +} + +// sliceForAppend takes a slice and a requested number of bytes. It returns a +// slice with the contents of the given slice followed by that many bytes and a +// second slice that aliases into it and contains only the extra bytes. If the +// original slice has sufficient capacity then no allocation is performed. +func sliceForAppend(in []byte, n int) (head, tail []byte) { + if total := len(in) + n; cap(in) >= total { + head = in[:total] + } else { + head = make([]byte, total) + copy(head, in) + } + tail = head[len(in):] + return +} diff --git a/vendor/golang.org/x/crypto/chacha20poly1305/chacha20poly1305_amd64.go b/vendor/golang.org/x/crypto/chacha20poly1305/chacha20poly1305_amd64.go new file mode 100644 index 0000000..2aa4fd8 --- /dev/null +++ b/vendor/golang.org/x/crypto/chacha20poly1305/chacha20poly1305_amd64.go @@ -0,0 +1,86 @@ +// Copyright 2016 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build go1.7,amd64,!gccgo,!appengine + +package chacha20poly1305 + +import ( + "encoding/binary" + + "golang.org/x/crypto/internal/subtle" + "golang.org/x/sys/cpu" +) + +//go:noescape +func chacha20Poly1305Open(dst []byte, key []uint32, src, ad []byte) bool + +//go:noescape +func chacha20Poly1305Seal(dst []byte, key []uint32, src, ad []byte) + +var ( + useAVX2 = cpu.X86.HasAVX2 && cpu.X86.HasBMI2 +) + +// setupState writes a ChaCha20 input matrix to state. See +// https://tools.ietf.org/html/rfc7539#section-2.3. +func setupState(state *[16]uint32, key *[8]uint32, nonce []byte) { + state[0] = 0x61707865 + state[1] = 0x3320646e + state[2] = 0x79622d32 + state[3] = 0x6b206574 + + state[4] = key[0] + state[5] = key[1] + state[6] = key[2] + state[7] = key[3] + state[8] = key[4] + state[9] = key[5] + state[10] = key[6] + state[11] = key[7] + + state[12] = 0 + state[13] = binary.LittleEndian.Uint32(nonce[:4]) + state[14] = binary.LittleEndian.Uint32(nonce[4:8]) + state[15] = binary.LittleEndian.Uint32(nonce[8:12]) +} + +func (c *chacha20poly1305) seal(dst, nonce, plaintext, additionalData []byte) []byte { + if !cpu.X86.HasSSSE3 { + return c.sealGeneric(dst, nonce, plaintext, additionalData) + } + + var state [16]uint32 + setupState(&state, &c.key, nonce) + + ret, out := sliceForAppend(dst, len(plaintext)+16) + if subtle.InexactOverlap(out, plaintext) { + panic("chacha20poly1305: invalid buffer overlap") + } + chacha20Poly1305Seal(out[:], state[:], plaintext, additionalData) + return ret +} + +func (c *chacha20poly1305) open(dst, nonce, ciphertext, additionalData []byte) ([]byte, error) { + if !cpu.X86.HasSSSE3 { + return c.openGeneric(dst, nonce, ciphertext, additionalData) + } + + var state [16]uint32 + setupState(&state, &c.key, nonce) + + ciphertext = ciphertext[:len(ciphertext)-16] + ret, out := sliceForAppend(dst, len(ciphertext)) + if subtle.InexactOverlap(out, ciphertext) { + panic("chacha20poly1305: invalid buffer overlap") + } + if !chacha20Poly1305Open(out, state[:], ciphertext, additionalData) { + for i := range out { + out[i] = 0 + } + return nil, errOpen + } + + return ret, nil +} diff --git a/vendor/golang.org/x/crypto/chacha20poly1305/chacha20poly1305_amd64.s b/vendor/golang.org/x/crypto/chacha20poly1305/chacha20poly1305_amd64.s new file mode 100644 index 0000000..9dd5d7a --- /dev/null +++ b/vendor/golang.org/x/crypto/chacha20poly1305/chacha20poly1305_amd64.s @@ -0,0 +1,2695 @@ +// Copyright 2016 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// This file was originally from https://golang.org/cl/24717 by Vlad Krasnov of CloudFlare. + +// +build go1.7,amd64,!gccgo,!appengine + +#include "textflag.h" +// General register allocation +#define oup DI +#define inp SI +#define inl BX +#define adp CX // free to reuse, after we hash the additional data +#define keyp R8 // free to reuse, when we copy the key to stack +#define itr2 R9 // general iterator +#define itr1 CX // general iterator +#define acc0 R10 +#define acc1 R11 +#define acc2 R12 +#define t0 R13 +#define t1 R14 +#define t2 R15 +#define t3 R8 +// Register and stack allocation for the SSE code +#define rStore (0*16)(BP) +#define sStore (1*16)(BP) +#define state1Store (2*16)(BP) +#define state2Store (3*16)(BP) +#define tmpStore (4*16)(BP) +#define ctr0Store (5*16)(BP) +#define ctr1Store (6*16)(BP) +#define ctr2Store (7*16)(BP) +#define ctr3Store (8*16)(BP) +#define A0 X0 +#define A1 X1 +#define A2 X2 +#define B0 X3 +#define B1 X4 +#define B2 X5 +#define C0 X6 +#define C1 X7 +#define C2 X8 +#define D0 X9 +#define D1 X10 +#define D2 X11 +#define T0 X12 +#define T1 X13 +#define T2 X14 +#define T3 X15 +#define A3 T0 +#define B3 T1 +#define C3 T2 +#define D3 T3 +// Register and stack allocation for the AVX2 code +#define rsStoreAVX2 (0*32)(BP) +#define state1StoreAVX2 (1*32)(BP) +#define state2StoreAVX2 (2*32)(BP) +#define ctr0StoreAVX2 (3*32)(BP) +#define ctr1StoreAVX2 (4*32)(BP) +#define ctr2StoreAVX2 (5*32)(BP) +#define ctr3StoreAVX2 (6*32)(BP) +#define tmpStoreAVX2 (7*32)(BP) // 256 bytes on stack +#define AA0 Y0 +#define AA1 Y5 +#define AA2 Y6 +#define AA3 Y7 +#define BB0 Y14 +#define BB1 Y9 +#define BB2 Y10 +#define BB3 Y11 +#define CC0 Y12 +#define CC1 Y13 +#define CC2 Y8 +#define CC3 Y15 +#define DD0 Y4 +#define DD1 Y1 +#define DD2 Y2 +#define DD3 Y3 +#define TT0 DD3 +#define TT1 AA3 +#define TT2 BB3 +#define TT3 CC3 +// ChaCha20 constants +DATA ·chacha20Constants<>+0x00(SB)/4, $0x61707865 +DATA ·chacha20Constants<>+0x04(SB)/4, $0x3320646e +DATA ·chacha20Constants<>+0x08(SB)/4, $0x79622d32 +DATA ·chacha20Constants<>+0x0c(SB)/4, $0x6b206574 +DATA ·chacha20Constants<>+0x10(SB)/4, $0x61707865 +DATA ·chacha20Constants<>+0x14(SB)/4, $0x3320646e +DATA ·chacha20Constants<>+0x18(SB)/4, $0x79622d32 +DATA ·chacha20Constants<>+0x1c(SB)/4, $0x6b206574 +// <<< 16 with PSHUFB +DATA ·rol16<>+0x00(SB)/8, $0x0504070601000302 +DATA ·rol16<>+0x08(SB)/8, $0x0D0C0F0E09080B0A +DATA ·rol16<>+0x10(SB)/8, $0x0504070601000302 +DATA ·rol16<>+0x18(SB)/8, $0x0D0C0F0E09080B0A +// <<< 8 with PSHUFB +DATA ·rol8<>+0x00(SB)/8, $0x0605040702010003 +DATA ·rol8<>+0x08(SB)/8, $0x0E0D0C0F0A09080B +DATA ·rol8<>+0x10(SB)/8, $0x0605040702010003 +DATA ·rol8<>+0x18(SB)/8, $0x0E0D0C0F0A09080B + +DATA ·avx2InitMask<>+0x00(SB)/8, $0x0 +DATA ·avx2InitMask<>+0x08(SB)/8, $0x0 +DATA ·avx2InitMask<>+0x10(SB)/8, $0x1 +DATA ·avx2InitMask<>+0x18(SB)/8, $0x0 + +DATA ·avx2IncMask<>+0x00(SB)/8, $0x2 +DATA ·avx2IncMask<>+0x08(SB)/8, $0x0 +DATA ·avx2IncMask<>+0x10(SB)/8, $0x2 +DATA ·avx2IncMask<>+0x18(SB)/8, $0x0 +// Poly1305 key clamp +DATA ·polyClampMask<>+0x00(SB)/8, $0x0FFFFFFC0FFFFFFF +DATA ·polyClampMask<>+0x08(SB)/8, $0x0FFFFFFC0FFFFFFC +DATA ·polyClampMask<>+0x10(SB)/8, $0xFFFFFFFFFFFFFFFF +DATA ·polyClampMask<>+0x18(SB)/8, $0xFFFFFFFFFFFFFFFF + +DATA ·sseIncMask<>+0x00(SB)/8, $0x1 +DATA ·sseIncMask<>+0x08(SB)/8, $0x0 +// To load/store the last < 16 bytes in a buffer +DATA ·andMask<>+0x00(SB)/8, $0x00000000000000ff +DATA ·andMask<>+0x08(SB)/8, $0x0000000000000000 +DATA ·andMask<>+0x10(SB)/8, $0x000000000000ffff +DATA ·andMask<>+0x18(SB)/8, $0x0000000000000000 +DATA ·andMask<>+0x20(SB)/8, $0x0000000000ffffff +DATA ·andMask<>+0x28(SB)/8, $0x0000000000000000 +DATA ·andMask<>+0x30(SB)/8, $0x00000000ffffffff +DATA ·andMask<>+0x38(SB)/8, $0x0000000000000000 +DATA ·andMask<>+0x40(SB)/8, $0x000000ffffffffff +DATA ·andMask<>+0x48(SB)/8, $0x0000000000000000 +DATA ·andMask<>+0x50(SB)/8, $0x0000ffffffffffff +DATA ·andMask<>+0x58(SB)/8, $0x0000000000000000 +DATA ·andMask<>+0x60(SB)/8, $0x00ffffffffffffff +DATA ·andMask<>+0x68(SB)/8, $0x0000000000000000 +DATA ·andMask<>+0x70(SB)/8, $0xffffffffffffffff +DATA ·andMask<>+0x78(SB)/8, $0x0000000000000000 +DATA ·andMask<>+0x80(SB)/8, $0xffffffffffffffff +DATA ·andMask<>+0x88(SB)/8, $0x00000000000000ff +DATA ·andMask<>+0x90(SB)/8, $0xffffffffffffffff +DATA ·andMask<>+0x98(SB)/8, $0x000000000000ffff +DATA ·andMask<>+0xa0(SB)/8, $0xffffffffffffffff +DATA ·andMask<>+0xa8(SB)/8, $0x0000000000ffffff +DATA ·andMask<>+0xb0(SB)/8, $0xffffffffffffffff +DATA ·andMask<>+0xb8(SB)/8, $0x00000000ffffffff +DATA ·andMask<>+0xc0(SB)/8, $0xffffffffffffffff +DATA ·andMask<>+0xc8(SB)/8, $0x000000ffffffffff +DATA ·andMask<>+0xd0(SB)/8, $0xffffffffffffffff +DATA ·andMask<>+0xd8(SB)/8, $0x0000ffffffffffff +DATA ·andMask<>+0xe0(SB)/8, $0xffffffffffffffff +DATA ·andMask<>+0xe8(SB)/8, $0x00ffffffffffffff + +GLOBL ·chacha20Constants<>(SB), (NOPTR+RODATA), $32 +GLOBL ·rol16<>(SB), (NOPTR+RODATA), $32 +GLOBL ·rol8<>(SB), (NOPTR+RODATA), $32 +GLOBL ·sseIncMask<>(SB), (NOPTR+RODATA), $16 +GLOBL ·avx2IncMask<>(SB), (NOPTR+RODATA), $32 +GLOBL ·avx2InitMask<>(SB), (NOPTR+RODATA), $32 +GLOBL ·polyClampMask<>(SB), (NOPTR+RODATA), $32 +GLOBL ·andMask<>(SB), (NOPTR+RODATA), $240 +// No PALIGNR in Go ASM yet (but VPALIGNR is present). +#define shiftB0Left BYTE $0x66; BYTE $0x0f; BYTE $0x3a; BYTE $0x0f; BYTE $0xdb; BYTE $0x04 // PALIGNR $4, X3, X3 +#define shiftB1Left BYTE $0x66; BYTE $0x0f; BYTE $0x3a; BYTE $0x0f; BYTE $0xe4; BYTE $0x04 // PALIGNR $4, X4, X4 +#define shiftB2Left BYTE $0x66; BYTE $0x0f; BYTE $0x3a; BYTE $0x0f; BYTE $0xed; BYTE $0x04 // PALIGNR $4, X5, X5 +#define shiftB3Left BYTE $0x66; BYTE $0x45; BYTE $0x0f; BYTE $0x3a; BYTE $0x0f; BYTE $0xed; BYTE $0x04 // PALIGNR $4, X13, X13 +#define shiftC0Left BYTE $0x66; BYTE $0x0f; BYTE $0x3a; BYTE $0x0f; BYTE $0xf6; BYTE $0x08 // PALIGNR $8, X6, X6 +#define shiftC1Left BYTE $0x66; BYTE $0x0f; BYTE $0x3a; BYTE $0x0f; BYTE $0xff; BYTE $0x08 // PALIGNR $8, X7, X7 +#define shiftC2Left BYTE $0x66; BYTE $0x45; BYTE $0x0f; BYTE $0x3a; BYTE $0x0f; BYTE $0xc0; BYTE $0x08 // PALIGNR $8, X8, X8 +#define shiftC3Left BYTE $0x66; BYTE $0x45; BYTE $0x0f; BYTE $0x3a; BYTE $0x0f; BYTE $0xf6; BYTE $0x08 // PALIGNR $8, X14, X14 +#define shiftD0Left BYTE $0x66; BYTE $0x45; BYTE $0x0f; BYTE $0x3a; BYTE $0x0f; BYTE $0xc9; BYTE $0x0c // PALIGNR $12, X9, X9 +#define shiftD1Left BYTE $0x66; BYTE $0x45; BYTE $0x0f; BYTE $0x3a; BYTE $0x0f; BYTE $0xd2; BYTE $0x0c // PALIGNR $12, X10, X10 +#define shiftD2Left BYTE $0x66; BYTE $0x45; BYTE $0x0f; BYTE $0x3a; BYTE $0x0f; BYTE $0xdb; BYTE $0x0c // PALIGNR $12, X11, X11 +#define shiftD3Left BYTE $0x66; BYTE $0x45; BYTE $0x0f; BYTE $0x3a; BYTE $0x0f; BYTE $0xff; BYTE $0x0c // PALIGNR $12, X15, X15 +#define shiftB0Right BYTE $0x66; BYTE $0x0f; BYTE $0x3a; BYTE $0x0f; BYTE $0xdb; BYTE $0x0c // PALIGNR $12, X3, X3 +#define shiftB1Right BYTE $0x66; BYTE $0x0f; BYTE $0x3a; BYTE $0x0f; BYTE $0xe4; BYTE $0x0c // PALIGNR $12, X4, X4 +#define shiftB2Right BYTE $0x66; BYTE $0x0f; BYTE $0x3a; BYTE $0x0f; BYTE $0xed; BYTE $0x0c // PALIGNR $12, X5, X5 +#define shiftB3Right BYTE $0x66; BYTE $0x45; BYTE $0x0f; BYTE $0x3a; BYTE $0x0f; BYTE $0xed; BYTE $0x0c // PALIGNR $12, X13, X13 +#define shiftC0Right shiftC0Left +#define shiftC1Right shiftC1Left +#define shiftC2Right shiftC2Left +#define shiftC3Right shiftC3Left +#define shiftD0Right BYTE $0x66; BYTE $0x45; BYTE $0x0f; BYTE $0x3a; BYTE $0x0f; BYTE $0xc9; BYTE $0x04 // PALIGNR $4, X9, X9 +#define shiftD1Right BYTE $0x66; BYTE $0x45; BYTE $0x0f; BYTE $0x3a; BYTE $0x0f; BYTE $0xd2; BYTE $0x04 // PALIGNR $4, X10, X10 +#define shiftD2Right BYTE $0x66; BYTE $0x45; BYTE $0x0f; BYTE $0x3a; BYTE $0x0f; BYTE $0xdb; BYTE $0x04 // PALIGNR $4, X11, X11 +#define shiftD3Right BYTE $0x66; BYTE $0x45; BYTE $0x0f; BYTE $0x3a; BYTE $0x0f; BYTE $0xff; BYTE $0x04 // PALIGNR $4, X15, X15 +// Some macros +#define chachaQR(A, B, C, D, T) \ + PADDD B, A; PXOR A, D; PSHUFB ·rol16<>(SB), D \ + PADDD D, C; PXOR C, B; MOVO B, T; PSLLL $12, T; PSRLL $20, B; PXOR T, B \ + PADDD B, A; PXOR A, D; PSHUFB ·rol8<>(SB), D \ + PADDD D, C; PXOR C, B; MOVO B, T; PSLLL $7, T; PSRLL $25, B; PXOR T, B + +#define chachaQR_AVX2(A, B, C, D, T) \ + VPADDD B, A, A; VPXOR A, D, D; VPSHUFB ·rol16<>(SB), D, D \ + VPADDD D, C, C; VPXOR C, B, B; VPSLLD $12, B, T; VPSRLD $20, B, B; VPXOR T, B, B \ + VPADDD B, A, A; VPXOR A, D, D; VPSHUFB ·rol8<>(SB), D, D \ + VPADDD D, C, C; VPXOR C, B, B; VPSLLD $7, B, T; VPSRLD $25, B, B; VPXOR T, B, B + +#define polyAdd(S) ADDQ S, acc0; ADCQ 8+S, acc1; ADCQ $1, acc2 +#define polyMulStage1 MOVQ (0*8)(BP), AX; MOVQ AX, t2; MULQ acc0; MOVQ AX, t0; MOVQ DX, t1; MOVQ (0*8)(BP), AX; MULQ acc1; IMULQ acc2, t2; ADDQ AX, t1; ADCQ DX, t2 +#define polyMulStage2 MOVQ (1*8)(BP), AX; MOVQ AX, t3; MULQ acc0; ADDQ AX, t1; ADCQ $0, DX; MOVQ DX, acc0; MOVQ (1*8)(BP), AX; MULQ acc1; ADDQ AX, t2; ADCQ $0, DX +#define polyMulStage3 IMULQ acc2, t3; ADDQ acc0, t2; ADCQ DX, t3 +#define polyMulReduceStage MOVQ t0, acc0; MOVQ t1, acc1; MOVQ t2, acc2; ANDQ $3, acc2; MOVQ t2, t0; ANDQ $-4, t0; MOVQ t3, t1; SHRQ $2, t3, t2; SHRQ $2, t3; ADDQ t0, acc0; ADCQ t1, acc1; ADCQ $0, acc2; ADDQ t2, acc0; ADCQ t3, acc1; ADCQ $0, acc2 + +#define polyMulStage1_AVX2 MOVQ (0*8)(BP), DX; MOVQ DX, t2; MULXQ acc0, t0, t1; IMULQ acc2, t2; MULXQ acc1, AX, DX; ADDQ AX, t1; ADCQ DX, t2 +#define polyMulStage2_AVX2 MOVQ (1*8)(BP), DX; MULXQ acc0, acc0, AX; ADDQ acc0, t1; MULXQ acc1, acc1, t3; ADCQ acc1, t2; ADCQ $0, t3 +#define polyMulStage3_AVX2 IMULQ acc2, DX; ADDQ AX, t2; ADCQ DX, t3 + +#define polyMul polyMulStage1; polyMulStage2; polyMulStage3; polyMulReduceStage +#define polyMulAVX2 polyMulStage1_AVX2; polyMulStage2_AVX2; polyMulStage3_AVX2; polyMulReduceStage +// ---------------------------------------------------------------------------- +TEXT polyHashADInternal<>(SB), NOSPLIT, $0 + // adp points to beginning of additional data + // itr2 holds ad length + XORQ acc0, acc0 + XORQ acc1, acc1 + XORQ acc2, acc2 + CMPQ itr2, $13 + JNE hashADLoop + +openFastTLSAD: + // Special treatment for the TLS case of 13 bytes + MOVQ (adp), acc0 + MOVQ 5(adp), acc1 + SHRQ $24, acc1 + MOVQ $1, acc2 + polyMul + RET + +hashADLoop: + // Hash in 16 byte chunks + CMPQ itr2, $16 + JB hashADTail + polyAdd(0(adp)) + LEAQ (1*16)(adp), adp + SUBQ $16, itr2 + polyMul + JMP hashADLoop + +hashADTail: + CMPQ itr2, $0 + JE hashADDone + + // Hash last < 16 byte tail + XORQ t0, t0 + XORQ t1, t1 + XORQ t2, t2 + ADDQ itr2, adp + +hashADTailLoop: + SHLQ $8, t0, t1 + SHLQ $8, t0 + MOVB -1(adp), t2 + XORQ t2, t0 + DECQ adp + DECQ itr2 + JNE hashADTailLoop + +hashADTailFinish: + ADDQ t0, acc0; ADCQ t1, acc1; ADCQ $1, acc2 + polyMul + + // Finished AD +hashADDone: + RET + +// ---------------------------------------------------------------------------- +// func chacha20Poly1305Open(dst, key, src, ad []byte) bool +TEXT ·chacha20Poly1305Open(SB), 0, $288-97 + // For aligned stack access + MOVQ SP, BP + ADDQ $32, BP + ANDQ $-32, BP + MOVQ dst+0(FP), oup + MOVQ key+24(FP), keyp + MOVQ src+48(FP), inp + MOVQ src_len+56(FP), inl + MOVQ ad+72(FP), adp + + // Check for AVX2 support + CMPB ·useAVX2(SB), $1 + JE chacha20Poly1305Open_AVX2 + + // Special optimization, for very short buffers + CMPQ inl, $128 + JBE openSSE128 // About 16% faster + + // For long buffers, prepare the poly key first + MOVOU ·chacha20Constants<>(SB), A0 + MOVOU (1*16)(keyp), B0 + MOVOU (2*16)(keyp), C0 + MOVOU (3*16)(keyp), D0 + MOVO D0, T1 + + // Store state on stack for future use + MOVO B0, state1Store + MOVO C0, state2Store + MOVO D0, ctr3Store + MOVQ $10, itr2 + +openSSEPreparePolyKey: + chachaQR(A0, B0, C0, D0, T0) + shiftB0Left; shiftC0Left; shiftD0Left + chachaQR(A0, B0, C0, D0, T0) + shiftB0Right; shiftC0Right; shiftD0Right + DECQ itr2 + JNE openSSEPreparePolyKey + + // A0|B0 hold the Poly1305 32-byte key, C0,D0 can be discarded + PADDL ·chacha20Constants<>(SB), A0; PADDL state1Store, B0 + + // Clamp and store the key + PAND ·polyClampMask<>(SB), A0 + MOVO A0, rStore; MOVO B0, sStore + + // Hash AAD + MOVQ ad_len+80(FP), itr2 + CALL polyHashADInternal<>(SB) + +openSSEMainLoop: + CMPQ inl, $256 + JB openSSEMainLoopDone + + // Load state, increment counter blocks + MOVO ·chacha20Constants<>(SB), A0; MOVO state1Store, B0; MOVO state2Store, C0; MOVO ctr3Store, D0; PADDL ·sseIncMask<>(SB), D0 + MOVO A0, A1; MOVO B0, B1; MOVO C0, C1; MOVO D0, D1; PADDL ·sseIncMask<>(SB), D1 + MOVO A1, A2; MOVO B1, B2; MOVO C1, C2; MOVO D1, D2; PADDL ·sseIncMask<>(SB), D2 + MOVO A2, A3; MOVO B2, B3; MOVO C2, C3; MOVO D2, D3; PADDL ·sseIncMask<>(SB), D3 + + // Store counters + MOVO D0, ctr0Store; MOVO D1, ctr1Store; MOVO D2, ctr2Store; MOVO D3, ctr3Store + + // There are 10 ChaCha20 iterations of 2QR each, so for 6 iterations we hash 2 blocks, and for the remaining 4 only 1 block - for a total of 16 + MOVQ $4, itr1 + MOVQ inp, itr2 + +openSSEInternalLoop: + MOVO C3, tmpStore + chachaQR(A0, B0, C0, D0, C3); chachaQR(A1, B1, C1, D1, C3); chachaQR(A2, B2, C2, D2, C3) + MOVO tmpStore, C3 + MOVO C1, tmpStore + chachaQR(A3, B3, C3, D3, C1) + MOVO tmpStore, C1 + polyAdd(0(itr2)) + shiftB0Left; shiftB1Left; shiftB2Left; shiftB3Left + shiftC0Left; shiftC1Left; shiftC2Left; shiftC3Left + shiftD0Left; shiftD1Left; shiftD2Left; shiftD3Left + polyMulStage1 + polyMulStage2 + LEAQ (2*8)(itr2), itr2 + MOVO C3, tmpStore + chachaQR(A0, B0, C0, D0, C3); chachaQR(A1, B1, C1, D1, C3); chachaQR(A2, B2, C2, D2, C3) + MOVO tmpStore, C3 + MOVO C1, tmpStore + polyMulStage3 + chachaQR(A3, B3, C3, D3, C1) + MOVO tmpStore, C1 + polyMulReduceStage + shiftB0Right; shiftB1Right; shiftB2Right; shiftB3Right + shiftC0Right; shiftC1Right; shiftC2Right; shiftC3Right + shiftD0Right; shiftD1Right; shiftD2Right; shiftD3Right + DECQ itr1 + JGE openSSEInternalLoop + + polyAdd(0(itr2)) + polyMul + LEAQ (2*8)(itr2), itr2 + + CMPQ itr1, $-6 + JG openSSEInternalLoop + + // Add in the state + PADDD ·chacha20Constants<>(SB), A0; PADDD ·chacha20Constants<>(SB), A1; PADDD ·chacha20Constants<>(SB), A2; PADDD ·chacha20Constants<>(SB), A3 + PADDD state1Store, B0; PADDD state1Store, B1; PADDD state1Store, B2; PADDD state1Store, B3 + PADDD state2Store, C0; PADDD state2Store, C1; PADDD state2Store, C2; PADDD state2Store, C3 + PADDD ctr0Store, D0; PADDD ctr1Store, D1; PADDD ctr2Store, D2; PADDD ctr3Store, D3 + + // Load - xor - store + MOVO D3, tmpStore + MOVOU (0*16)(inp), D3; PXOR D3, A0; MOVOU A0, (0*16)(oup) + MOVOU (1*16)(inp), D3; PXOR D3, B0; MOVOU B0, (1*16)(oup) + MOVOU (2*16)(inp), D3; PXOR D3, C0; MOVOU C0, (2*16)(oup) + MOVOU (3*16)(inp), D3; PXOR D3, D0; MOVOU D0, (3*16)(oup) + MOVOU (4*16)(inp), D0; PXOR D0, A1; MOVOU A1, (4*16)(oup) + MOVOU (5*16)(inp), D0; PXOR D0, B1; MOVOU B1, (5*16)(oup) + MOVOU (6*16)(inp), D0; PXOR D0, C1; MOVOU C1, (6*16)(oup) + MOVOU (7*16)(inp), D0; PXOR D0, D1; MOVOU D1, (7*16)(oup) + MOVOU (8*16)(inp), D0; PXOR D0, A2; MOVOU A2, (8*16)(oup) + MOVOU (9*16)(inp), D0; PXOR D0, B2; MOVOU B2, (9*16)(oup) + MOVOU (10*16)(inp), D0; PXOR D0, C2; MOVOU C2, (10*16)(oup) + MOVOU (11*16)(inp), D0; PXOR D0, D2; MOVOU D2, (11*16)(oup) + MOVOU (12*16)(inp), D0; PXOR D0, A3; MOVOU A3, (12*16)(oup) + MOVOU (13*16)(inp), D0; PXOR D0, B3; MOVOU B3, (13*16)(oup) + MOVOU (14*16)(inp), D0; PXOR D0, C3; MOVOU C3, (14*16)(oup) + MOVOU (15*16)(inp), D0; PXOR tmpStore, D0; MOVOU D0, (15*16)(oup) + LEAQ 256(inp), inp + LEAQ 256(oup), oup + SUBQ $256, inl + JMP openSSEMainLoop + +openSSEMainLoopDone: + // Handle the various tail sizes efficiently + TESTQ inl, inl + JE openSSEFinalize + CMPQ inl, $64 + JBE openSSETail64 + CMPQ inl, $128 + JBE openSSETail128 + CMPQ inl, $192 + JBE openSSETail192 + JMP openSSETail256 + +openSSEFinalize: + // Hash in the PT, AAD lengths + ADDQ ad_len+80(FP), acc0; ADCQ src_len+56(FP), acc1; ADCQ $1, acc2 + polyMul + + // Final reduce + MOVQ acc0, t0 + MOVQ acc1, t1 + MOVQ acc2, t2 + SUBQ $-5, acc0 + SBBQ $-1, acc1 + SBBQ $3, acc2 + CMOVQCS t0, acc0 + CMOVQCS t1, acc1 + CMOVQCS t2, acc2 + + // Add in the "s" part of the key + ADDQ 0+sStore, acc0 + ADCQ 8+sStore, acc1 + + // Finally, constant time compare to the tag at the end of the message + XORQ AX, AX + MOVQ $1, DX + XORQ (0*8)(inp), acc0 + XORQ (1*8)(inp), acc1 + ORQ acc1, acc0 + CMOVQEQ DX, AX + + // Return true iff tags are equal + MOVB AX, ret+96(FP) + RET + +// ---------------------------------------------------------------------------- +// Special optimization for buffers smaller than 129 bytes +openSSE128: + // For up to 128 bytes of ciphertext and 64 bytes for the poly key, we require to process three blocks + MOVOU ·chacha20Constants<>(SB), A0; MOVOU (1*16)(keyp), B0; MOVOU (2*16)(keyp), C0; MOVOU (3*16)(keyp), D0 + MOVO A0, A1; MOVO B0, B1; MOVO C0, C1; MOVO D0, D1; PADDL ·sseIncMask<>(SB), D1 + MOVO A1, A2; MOVO B1, B2; MOVO C1, C2; MOVO D1, D2; PADDL ·sseIncMask<>(SB), D2 + MOVO B0, T1; MOVO C0, T2; MOVO D1, T3 + MOVQ $10, itr2 + +openSSE128InnerCipherLoop: + chachaQR(A0, B0, C0, D0, T0); chachaQR(A1, B1, C1, D1, T0); chachaQR(A2, B2, C2, D2, T0) + shiftB0Left; shiftB1Left; shiftB2Left + shiftC0Left; shiftC1Left; shiftC2Left + shiftD0Left; shiftD1Left; shiftD2Left + chachaQR(A0, B0, C0, D0, T0); chachaQR(A1, B1, C1, D1, T0); chachaQR(A2, B2, C2, D2, T0) + shiftB0Right; shiftB1Right; shiftB2Right + shiftC0Right; shiftC1Right; shiftC2Right + shiftD0Right; shiftD1Right; shiftD2Right + DECQ itr2 + JNE openSSE128InnerCipherLoop + + // A0|B0 hold the Poly1305 32-byte key, C0,D0 can be discarded + PADDL ·chacha20Constants<>(SB), A0; PADDL ·chacha20Constants<>(SB), A1; PADDL ·chacha20Constants<>(SB), A2 + PADDL T1, B0; PADDL T1, B1; PADDL T1, B2 + PADDL T2, C1; PADDL T2, C2 + PADDL T3, D1; PADDL ·sseIncMask<>(SB), T3; PADDL T3, D2 + + // Clamp and store the key + PAND ·polyClampMask<>(SB), A0 + MOVOU A0, rStore; MOVOU B0, sStore + + // Hash + MOVQ ad_len+80(FP), itr2 + CALL polyHashADInternal<>(SB) + +openSSE128Open: + CMPQ inl, $16 + JB openSSETail16 + SUBQ $16, inl + + // Load for hashing + polyAdd(0(inp)) + + // Load for decryption + MOVOU (inp), T0; PXOR T0, A1; MOVOU A1, (oup) + LEAQ (1*16)(inp), inp + LEAQ (1*16)(oup), oup + polyMul + + // Shift the stream "left" + MOVO B1, A1 + MOVO C1, B1 + MOVO D1, C1 + MOVO A2, D1 + MOVO B2, A2 + MOVO C2, B2 + MOVO D2, C2 + JMP openSSE128Open + +openSSETail16: + TESTQ inl, inl + JE openSSEFinalize + + // We can safely load the CT from the end, because it is padded with the MAC + MOVQ inl, itr2 + SHLQ $4, itr2 + LEAQ ·andMask<>(SB), t0 + MOVOU (inp), T0 + ADDQ inl, inp + PAND -16(t0)(itr2*1), T0 + MOVO T0, 0+tmpStore + MOVQ T0, t0 + MOVQ 8+tmpStore, t1 + PXOR A1, T0 + + // We can only store one byte at a time, since plaintext can be shorter than 16 bytes +openSSETail16Store: + MOVQ T0, t3 + MOVB t3, (oup) + PSRLDQ $1, T0 + INCQ oup + DECQ inl + JNE openSSETail16Store + ADDQ t0, acc0; ADCQ t1, acc1; ADCQ $1, acc2 + polyMul + JMP openSSEFinalize + +// ---------------------------------------------------------------------------- +// Special optimization for the last 64 bytes of ciphertext +openSSETail64: + // Need to decrypt up to 64 bytes - prepare single block + MOVO ·chacha20Constants<>(SB), A0; MOVO state1Store, B0; MOVO state2Store, C0; MOVO ctr3Store, D0; PADDL ·sseIncMask<>(SB), D0; MOVO D0, ctr0Store + XORQ itr2, itr2 + MOVQ inl, itr1 + CMPQ itr1, $16 + JB openSSETail64LoopB + +openSSETail64LoopA: + // Perform ChaCha rounds, while hashing the remaining input + polyAdd(0(inp)(itr2*1)) + polyMul + SUBQ $16, itr1 + +openSSETail64LoopB: + ADDQ $16, itr2 + chachaQR(A0, B0, C0, D0, T0) + shiftB0Left; shiftC0Left; shiftD0Left + chachaQR(A0, B0, C0, D0, T0) + shiftB0Right; shiftC0Right; shiftD0Right + + CMPQ itr1, $16 + JAE openSSETail64LoopA + + CMPQ itr2, $160 + JNE openSSETail64LoopB + + PADDL ·chacha20Constants<>(SB), A0; PADDL state1Store, B0; PADDL state2Store, C0; PADDL ctr0Store, D0 + +openSSETail64DecLoop: + CMPQ inl, $16 + JB openSSETail64DecLoopDone + SUBQ $16, inl + MOVOU (inp), T0 + PXOR T0, A0 + MOVOU A0, (oup) + LEAQ 16(inp), inp + LEAQ 16(oup), oup + MOVO B0, A0 + MOVO C0, B0 + MOVO D0, C0 + JMP openSSETail64DecLoop + +openSSETail64DecLoopDone: + MOVO A0, A1 + JMP openSSETail16 + +// ---------------------------------------------------------------------------- +// Special optimization for the last 128 bytes of ciphertext +openSSETail128: + // Need to decrypt up to 128 bytes - prepare two blocks + MOVO ·chacha20Constants<>(SB), A1; MOVO state1Store, B1; MOVO state2Store, C1; MOVO ctr3Store, D1; PADDL ·sseIncMask<>(SB), D1; MOVO D1, ctr0Store + MOVO A1, A0; MOVO B1, B0; MOVO C1, C0; MOVO D1, D0; PADDL ·sseIncMask<>(SB), D0; MOVO D0, ctr1Store + XORQ itr2, itr2 + MOVQ inl, itr1 + ANDQ $-16, itr1 + +openSSETail128LoopA: + // Perform ChaCha rounds, while hashing the remaining input + polyAdd(0(inp)(itr2*1)) + polyMul + +openSSETail128LoopB: + ADDQ $16, itr2 + chachaQR(A0, B0, C0, D0, T0); chachaQR(A1, B1, C1, D1, T0) + shiftB0Left; shiftC0Left; shiftD0Left + shiftB1Left; shiftC1Left; shiftD1Left + chachaQR(A0, B0, C0, D0, T0); chachaQR(A1, B1, C1, D1, T0) + shiftB0Right; shiftC0Right; shiftD0Right + shiftB1Right; shiftC1Right; shiftD1Right + + CMPQ itr2, itr1 + JB openSSETail128LoopA + + CMPQ itr2, $160 + JNE openSSETail128LoopB + + PADDL ·chacha20Constants<>(SB), A0; PADDL ·chacha20Constants<>(SB), A1 + PADDL state1Store, B0; PADDL state1Store, B1 + PADDL state2Store, C0; PADDL state2Store, C1 + PADDL ctr1Store, D0; PADDL ctr0Store, D1 + + MOVOU (0*16)(inp), T0; MOVOU (1*16)(inp), T1; MOVOU (2*16)(inp), T2; MOVOU (3*16)(inp), T3 + PXOR T0, A1; PXOR T1, B1; PXOR T2, C1; PXOR T3, D1 + MOVOU A1, (0*16)(oup); MOVOU B1, (1*16)(oup); MOVOU C1, (2*16)(oup); MOVOU D1, (3*16)(oup) + + SUBQ $64, inl + LEAQ 64(inp), inp + LEAQ 64(oup), oup + JMP openSSETail64DecLoop + +// ---------------------------------------------------------------------------- +// Special optimization for the last 192 bytes of ciphertext +openSSETail192: + // Need to decrypt up to 192 bytes - prepare three blocks + MOVO ·chacha20Constants<>(SB), A2; MOVO state1Store, B2; MOVO state2Store, C2; MOVO ctr3Store, D2; PADDL ·sseIncMask<>(SB), D2; MOVO D2, ctr0Store + MOVO A2, A1; MOVO B2, B1; MOVO C2, C1; MOVO D2, D1; PADDL ·sseIncMask<>(SB), D1; MOVO D1, ctr1Store + MOVO A1, A0; MOVO B1, B0; MOVO C1, C0; MOVO D1, D0; PADDL ·sseIncMask<>(SB), D0; MOVO D0, ctr2Store + + MOVQ inl, itr1 + MOVQ $160, itr2 + CMPQ itr1, $160 + CMOVQGT itr2, itr1 + ANDQ $-16, itr1 + XORQ itr2, itr2 + +openSSLTail192LoopA: + // Perform ChaCha rounds, while hashing the remaining input + polyAdd(0(inp)(itr2*1)) + polyMul + +openSSLTail192LoopB: + ADDQ $16, itr2 + chachaQR(A0, B0, C0, D0, T0); chachaQR(A1, B1, C1, D1, T0); chachaQR(A2, B2, C2, D2, T0) + shiftB0Left; shiftC0Left; shiftD0Left + shiftB1Left; shiftC1Left; shiftD1Left + shiftB2Left; shiftC2Left; shiftD2Left + + chachaQR(A0, B0, C0, D0, T0); chachaQR(A1, B1, C1, D1, T0); chachaQR(A2, B2, C2, D2, T0) + shiftB0Right; shiftC0Right; shiftD0Right + shiftB1Right; shiftC1Right; shiftD1Right + shiftB2Right; shiftC2Right; shiftD2Right + + CMPQ itr2, itr1 + JB openSSLTail192LoopA + + CMPQ itr2, $160 + JNE openSSLTail192LoopB + + CMPQ inl, $176 + JB openSSLTail192Store + + polyAdd(160(inp)) + polyMul + + CMPQ inl, $192 + JB openSSLTail192Store + + polyAdd(176(inp)) + polyMul + +openSSLTail192Store: + PADDL ·chacha20Constants<>(SB), A0; PADDL ·chacha20Constants<>(SB), A1; PADDL ·chacha20Constants<>(SB), A2 + PADDL state1Store, B0; PADDL state1Store, B1; PADDL state1Store, B2 + PADDL state2Store, C0; PADDL state2Store, C1; PADDL state2Store, C2 + PADDL ctr2Store, D0; PADDL ctr1Store, D1; PADDL ctr0Store, D2 + + MOVOU (0*16)(inp), T0; MOVOU (1*16)(inp), T1; MOVOU (2*16)(inp), T2; MOVOU (3*16)(inp), T3 + PXOR T0, A2; PXOR T1, B2; PXOR T2, C2; PXOR T3, D2 + MOVOU A2, (0*16)(oup); MOVOU B2, (1*16)(oup); MOVOU C2, (2*16)(oup); MOVOU D2, (3*16)(oup) + + MOVOU (4*16)(inp), T0; MOVOU (5*16)(inp), T1; MOVOU (6*16)(inp), T2; MOVOU (7*16)(inp), T3 + PXOR T0, A1; PXOR T1, B1; PXOR T2, C1; PXOR T3, D1 + MOVOU A1, (4*16)(oup); MOVOU B1, (5*16)(oup); MOVOU C1, (6*16)(oup); MOVOU D1, (7*16)(oup) + + SUBQ $128, inl + LEAQ 128(inp), inp + LEAQ 128(oup), oup + JMP openSSETail64DecLoop + +// ---------------------------------------------------------------------------- +// Special optimization for the last 256 bytes of ciphertext +openSSETail256: + // Need to decrypt up to 256 bytes - prepare four blocks + MOVO ·chacha20Constants<>(SB), A0; MOVO state1Store, B0; MOVO state2Store, C0; MOVO ctr3Store, D0; PADDL ·sseIncMask<>(SB), D0 + MOVO A0, A1; MOVO B0, B1; MOVO C0, C1; MOVO D0, D1; PADDL ·sseIncMask<>(SB), D1 + MOVO A1, A2; MOVO B1, B2; MOVO C1, C2; MOVO D1, D2; PADDL ·sseIncMask<>(SB), D2 + MOVO A2, A3; MOVO B2, B3; MOVO C2, C3; MOVO D2, D3; PADDL ·sseIncMask<>(SB), D3 + + // Store counters + MOVO D0, ctr0Store; MOVO D1, ctr1Store; MOVO D2, ctr2Store; MOVO D3, ctr3Store + XORQ itr2, itr2 + +openSSETail256Loop: + // This loop inteleaves 8 ChaCha quarter rounds with 1 poly multiplication + polyAdd(0(inp)(itr2*1)) + MOVO C3, tmpStore + chachaQR(A0, B0, C0, D0, C3); chachaQR(A1, B1, C1, D1, C3); chachaQR(A2, B2, C2, D2, C3) + MOVO tmpStore, C3 + MOVO C1, tmpStore + chachaQR(A3, B3, C3, D3, C1) + MOVO tmpStore, C1 + shiftB0Left; shiftB1Left; shiftB2Left; shiftB3Left + shiftC0Left; shiftC1Left; shiftC2Left; shiftC3Left + shiftD0Left; shiftD1Left; shiftD2Left; shiftD3Left + polyMulStage1 + polyMulStage2 + MOVO C3, tmpStore + chachaQR(A0, B0, C0, D0, C3); chachaQR(A1, B1, C1, D1, C3); chachaQR(A2, B2, C2, D2, C3) + MOVO tmpStore, C3 + MOVO C1, tmpStore + chachaQR(A3, B3, C3, D3, C1) + MOVO tmpStore, C1 + polyMulStage3 + polyMulReduceStage + shiftB0Right; shiftB1Right; shiftB2Right; shiftB3Right + shiftC0Right; shiftC1Right; shiftC2Right; shiftC3Right + shiftD0Right; shiftD1Right; shiftD2Right; shiftD3Right + ADDQ $2*8, itr2 + CMPQ itr2, $160 + JB openSSETail256Loop + MOVQ inl, itr1 + ANDQ $-16, itr1 + +openSSETail256HashLoop: + polyAdd(0(inp)(itr2*1)) + polyMul + ADDQ $2*8, itr2 + CMPQ itr2, itr1 + JB openSSETail256HashLoop + + // Add in the state + PADDD ·chacha20Constants<>(SB), A0; PADDD ·chacha20Constants<>(SB), A1; PADDD ·chacha20Constants<>(SB), A2; PADDD ·chacha20Constants<>(SB), A3 + PADDD state1Store, B0; PADDD state1Store, B1; PADDD state1Store, B2; PADDD state1Store, B3 + PADDD state2Store, C0; PADDD state2Store, C1; PADDD state2Store, C2; PADDD state2Store, C3 + PADDD ctr0Store, D0; PADDD ctr1Store, D1; PADDD ctr2Store, D2; PADDD ctr3Store, D3 + MOVO D3, tmpStore + + // Load - xor - store + MOVOU (0*16)(inp), D3; PXOR D3, A0 + MOVOU (1*16)(inp), D3; PXOR D3, B0 + MOVOU (2*16)(inp), D3; PXOR D3, C0 + MOVOU (3*16)(inp), D3; PXOR D3, D0 + MOVOU A0, (0*16)(oup) + MOVOU B0, (1*16)(oup) + MOVOU C0, (2*16)(oup) + MOVOU D0, (3*16)(oup) + MOVOU (4*16)(inp), A0; MOVOU (5*16)(inp), B0; MOVOU (6*16)(inp), C0; MOVOU (7*16)(inp), D0 + PXOR A0, A1; PXOR B0, B1; PXOR C0, C1; PXOR D0, D1 + MOVOU A1, (4*16)(oup); MOVOU B1, (5*16)(oup); MOVOU C1, (6*16)(oup); MOVOU D1, (7*16)(oup) + MOVOU (8*16)(inp), A0; MOVOU (9*16)(inp), B0; MOVOU (10*16)(inp), C0; MOVOU (11*16)(inp), D0 + PXOR A0, A2; PXOR B0, B2; PXOR C0, C2; PXOR D0, D2 + MOVOU A2, (8*16)(oup); MOVOU B2, (9*16)(oup); MOVOU C2, (10*16)(oup); MOVOU D2, (11*16)(oup) + LEAQ 192(inp), inp + LEAQ 192(oup), oup + SUBQ $192, inl + MOVO A3, A0 + MOVO B3, B0 + MOVO C3, C0 + MOVO tmpStore, D0 + + JMP openSSETail64DecLoop + +// ---------------------------------------------------------------------------- +// ------------------------- AVX2 Code ---------------------------------------- +chacha20Poly1305Open_AVX2: + VZEROUPPER + VMOVDQU ·chacha20Constants<>(SB), AA0 + BYTE $0xc4; BYTE $0x42; BYTE $0x7d; BYTE $0x5a; BYTE $0x70; BYTE $0x10 // broadcasti128 16(r8), ymm14 + BYTE $0xc4; BYTE $0x42; BYTE $0x7d; BYTE $0x5a; BYTE $0x60; BYTE $0x20 // broadcasti128 32(r8), ymm12 + BYTE $0xc4; BYTE $0xc2; BYTE $0x7d; BYTE $0x5a; BYTE $0x60; BYTE $0x30 // broadcasti128 48(r8), ymm4 + VPADDD ·avx2InitMask<>(SB), DD0, DD0 + + // Special optimization, for very short buffers + CMPQ inl, $192 + JBE openAVX2192 + CMPQ inl, $320 + JBE openAVX2320 + + // For the general key prepare the key first - as a byproduct we have 64 bytes of cipher stream + VMOVDQA BB0, state1StoreAVX2 + VMOVDQA CC0, state2StoreAVX2 + VMOVDQA DD0, ctr3StoreAVX2 + MOVQ $10, itr2 + +openAVX2PreparePolyKey: + chachaQR_AVX2(AA0, BB0, CC0, DD0, TT0) + VPALIGNR $4, BB0, BB0, BB0; VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $12, DD0, DD0, DD0 + chachaQR_AVX2(AA0, BB0, CC0, DD0, TT0) + VPALIGNR $12, BB0, BB0, BB0; VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $4, DD0, DD0, DD0 + DECQ itr2 + JNE openAVX2PreparePolyKey + + VPADDD ·chacha20Constants<>(SB), AA0, AA0 + VPADDD state1StoreAVX2, BB0, BB0 + VPADDD state2StoreAVX2, CC0, CC0 + VPADDD ctr3StoreAVX2, DD0, DD0 + + VPERM2I128 $0x02, AA0, BB0, TT0 + + // Clamp and store poly key + VPAND ·polyClampMask<>(SB), TT0, TT0 + VMOVDQA TT0, rsStoreAVX2 + + // Stream for the first 64 bytes + VPERM2I128 $0x13, AA0, BB0, AA0 + VPERM2I128 $0x13, CC0, DD0, BB0 + + // Hash AD + first 64 bytes + MOVQ ad_len+80(FP), itr2 + CALL polyHashADInternal<>(SB) + XORQ itr1, itr1 + +openAVX2InitialHash64: + polyAdd(0(inp)(itr1*1)) + polyMulAVX2 + ADDQ $16, itr1 + CMPQ itr1, $64 + JNE openAVX2InitialHash64 + + // Decrypt the first 64 bytes + VPXOR (0*32)(inp), AA0, AA0 + VPXOR (1*32)(inp), BB0, BB0 + VMOVDQU AA0, (0*32)(oup) + VMOVDQU BB0, (1*32)(oup) + LEAQ (2*32)(inp), inp + LEAQ (2*32)(oup), oup + SUBQ $64, inl + +openAVX2MainLoop: + CMPQ inl, $512 + JB openAVX2MainLoopDone + + // Load state, increment counter blocks, store the incremented counters + VMOVDQU ·chacha20Constants<>(SB), AA0; VMOVDQA AA0, AA1; VMOVDQA AA0, AA2; VMOVDQA AA0, AA3 + VMOVDQA state1StoreAVX2, BB0; VMOVDQA BB0, BB1; VMOVDQA BB0, BB2; VMOVDQA BB0, BB3 + VMOVDQA state2StoreAVX2, CC0; VMOVDQA CC0, CC1; VMOVDQA CC0, CC2; VMOVDQA CC0, CC3 + VMOVDQA ctr3StoreAVX2, DD0; VPADDD ·avx2IncMask<>(SB), DD0, DD0; VPADDD ·avx2IncMask<>(SB), DD0, DD1; VPADDD ·avx2IncMask<>(SB), DD1, DD2; VPADDD ·avx2IncMask<>(SB), DD2, DD3 + VMOVDQA DD0, ctr0StoreAVX2; VMOVDQA DD1, ctr1StoreAVX2; VMOVDQA DD2, ctr2StoreAVX2; VMOVDQA DD3, ctr3StoreAVX2 + XORQ itr1, itr1 + +openAVX2InternalLoop: + // Lets just say this spaghetti loop interleaves 2 quarter rounds with 3 poly multiplications + // Effectively per 512 bytes of stream we hash 480 bytes of ciphertext + polyAdd(0*8(inp)(itr1*1)) + VPADDD BB0, AA0, AA0; VPADDD BB1, AA1, AA1; VPADDD BB2, AA2, AA2; VPADDD BB3, AA3, AA3 + polyMulStage1_AVX2 + VPXOR AA0, DD0, DD0; VPXOR AA1, DD1, DD1; VPXOR AA2, DD2, DD2; VPXOR AA3, DD3, DD3 + VPSHUFB ·rol16<>(SB), DD0, DD0; VPSHUFB ·rol16<>(SB), DD1, DD1; VPSHUFB ·rol16<>(SB), DD2, DD2; VPSHUFB ·rol16<>(SB), DD3, DD3 + polyMulStage2_AVX2 + VPADDD DD0, CC0, CC0; VPADDD DD1, CC1, CC1; VPADDD DD2, CC2, CC2; VPADDD DD3, CC3, CC3 + VPXOR CC0, BB0, BB0; VPXOR CC1, BB1, BB1; VPXOR CC2, BB2, BB2; VPXOR CC3, BB3, BB3 + polyMulStage3_AVX2 + VMOVDQA CC3, tmpStoreAVX2 + VPSLLD $12, BB0, CC3; VPSRLD $20, BB0, BB0; VPXOR CC3, BB0, BB0 + VPSLLD $12, BB1, CC3; VPSRLD $20, BB1, BB1; VPXOR CC3, BB1, BB1 + VPSLLD $12, BB2, CC3; VPSRLD $20, BB2, BB2; VPXOR CC3, BB2, BB2 + VPSLLD $12, BB3, CC3; VPSRLD $20, BB3, BB3; VPXOR CC3, BB3, BB3 + VMOVDQA tmpStoreAVX2, CC3 + polyMulReduceStage + VPADDD BB0, AA0, AA0; VPADDD BB1, AA1, AA1; VPADDD BB2, AA2, AA2; VPADDD BB3, AA3, AA3 + VPXOR AA0, DD0, DD0; VPXOR AA1, DD1, DD1; VPXOR AA2, DD2, DD2; VPXOR AA3, DD3, DD3 + VPSHUFB ·rol8<>(SB), DD0, DD0; VPSHUFB ·rol8<>(SB), DD1, DD1; VPSHUFB ·rol8<>(SB), DD2, DD2; VPSHUFB ·rol8<>(SB), DD3, DD3 + polyAdd(2*8(inp)(itr1*1)) + VPADDD DD0, CC0, CC0; VPADDD DD1, CC1, CC1; VPADDD DD2, CC2, CC2; VPADDD DD3, CC3, CC3 + polyMulStage1_AVX2 + VPXOR CC0, BB0, BB0; VPXOR CC1, BB1, BB1; VPXOR CC2, BB2, BB2; VPXOR CC3, BB3, BB3 + VMOVDQA CC3, tmpStoreAVX2 + VPSLLD $7, BB0, CC3; VPSRLD $25, BB0, BB0; VPXOR CC3, BB0, BB0 + VPSLLD $7, BB1, CC3; VPSRLD $25, BB1, BB1; VPXOR CC3, BB1, BB1 + VPSLLD $7, BB2, CC3; VPSRLD $25, BB2, BB2; VPXOR CC3, BB2, BB2 + VPSLLD $7, BB3, CC3; VPSRLD $25, BB3, BB3; VPXOR CC3, BB3, BB3 + VMOVDQA tmpStoreAVX2, CC3 + polyMulStage2_AVX2 + VPALIGNR $4, BB0, BB0, BB0; VPALIGNR $4, BB1, BB1, BB1; VPALIGNR $4, BB2, BB2, BB2; VPALIGNR $4, BB3, BB3, BB3 + VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $8, CC1, CC1, CC1; VPALIGNR $8, CC2, CC2, CC2; VPALIGNR $8, CC3, CC3, CC3 + VPALIGNR $12, DD0, DD0, DD0; VPALIGNR $12, DD1, DD1, DD1; VPALIGNR $12, DD2, DD2, DD2; VPALIGNR $12, DD3, DD3, DD3 + VPADDD BB0, AA0, AA0; VPADDD BB1, AA1, AA1; VPADDD BB2, AA2, AA2; VPADDD BB3, AA3, AA3 + polyMulStage3_AVX2 + VPXOR AA0, DD0, DD0; VPXOR AA1, DD1, DD1; VPXOR AA2, DD2, DD2; VPXOR AA3, DD3, DD3 + VPSHUFB ·rol16<>(SB), DD0, DD0; VPSHUFB ·rol16<>(SB), DD1, DD1; VPSHUFB ·rol16<>(SB), DD2, DD2; VPSHUFB ·rol16<>(SB), DD3, DD3 + polyMulReduceStage + VPADDD DD0, CC0, CC0; VPADDD DD1, CC1, CC1; VPADDD DD2, CC2, CC2; VPADDD DD3, CC3, CC3 + VPXOR CC0, BB0, BB0; VPXOR CC1, BB1, BB1; VPXOR CC2, BB2, BB2; VPXOR CC3, BB3, BB3 + polyAdd(4*8(inp)(itr1*1)) + LEAQ (6*8)(itr1), itr1 + VMOVDQA CC3, tmpStoreAVX2 + VPSLLD $12, BB0, CC3; VPSRLD $20, BB0, BB0; VPXOR CC3, BB0, BB0 + VPSLLD $12, BB1, CC3; VPSRLD $20, BB1, BB1; VPXOR CC3, BB1, BB1 + VPSLLD $12, BB2, CC3; VPSRLD $20, BB2, BB2; VPXOR CC3, BB2, BB2 + VPSLLD $12, BB3, CC3; VPSRLD $20, BB3, BB3; VPXOR CC3, BB3, BB3 + VMOVDQA tmpStoreAVX2, CC3 + polyMulStage1_AVX2 + VPADDD BB0, AA0, AA0; VPADDD BB1, AA1, AA1; VPADDD BB2, AA2, AA2; VPADDD BB3, AA3, AA3 + VPXOR AA0, DD0, DD0; VPXOR AA1, DD1, DD1; VPXOR AA2, DD2, DD2; VPXOR AA3, DD3, DD3 + polyMulStage2_AVX2 + VPSHUFB ·rol8<>(SB), DD0, DD0; VPSHUFB ·rol8<>(SB), DD1, DD1; VPSHUFB ·rol8<>(SB), DD2, DD2; VPSHUFB ·rol8<>(SB), DD3, DD3 + VPADDD DD0, CC0, CC0; VPADDD DD1, CC1, CC1; VPADDD DD2, CC2, CC2; VPADDD DD3, CC3, CC3 + polyMulStage3_AVX2 + VPXOR CC0, BB0, BB0; VPXOR CC1, BB1, BB1; VPXOR CC2, BB2, BB2; VPXOR CC3, BB3, BB3 + VMOVDQA CC3, tmpStoreAVX2 + VPSLLD $7, BB0, CC3; VPSRLD $25, BB0, BB0; VPXOR CC3, BB0, BB0 + VPSLLD $7, BB1, CC3; VPSRLD $25, BB1, BB1; VPXOR CC3, BB1, BB1 + VPSLLD $7, BB2, CC3; VPSRLD $25, BB2, BB2; VPXOR CC3, BB2, BB2 + VPSLLD $7, BB3, CC3; VPSRLD $25, BB3, BB3; VPXOR CC3, BB3, BB3 + VMOVDQA tmpStoreAVX2, CC3 + polyMulReduceStage + VPALIGNR $12, BB0, BB0, BB0; VPALIGNR $12, BB1, BB1, BB1; VPALIGNR $12, BB2, BB2, BB2; VPALIGNR $12, BB3, BB3, BB3 + VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $8, CC1, CC1, CC1; VPALIGNR $8, CC2, CC2, CC2; VPALIGNR $8, CC3, CC3, CC3 + VPALIGNR $4, DD0, DD0, DD0; VPALIGNR $4, DD1, DD1, DD1; VPALIGNR $4, DD2, DD2, DD2; VPALIGNR $4, DD3, DD3, DD3 + CMPQ itr1, $480 + JNE openAVX2InternalLoop + + VPADDD ·chacha20Constants<>(SB), AA0, AA0; VPADDD ·chacha20Constants<>(SB), AA1, AA1; VPADDD ·chacha20Constants<>(SB), AA2, AA2; VPADDD ·chacha20Constants<>(SB), AA3, AA3 + VPADDD state1StoreAVX2, BB0, BB0; VPADDD state1StoreAVX2, BB1, BB1; VPADDD state1StoreAVX2, BB2, BB2; VPADDD state1StoreAVX2, BB3, BB3 + VPADDD state2StoreAVX2, CC0, CC0; VPADDD state2StoreAVX2, CC1, CC1; VPADDD state2StoreAVX2, CC2, CC2; VPADDD state2StoreAVX2, CC3, CC3 + VPADDD ctr0StoreAVX2, DD0, DD0; VPADDD ctr1StoreAVX2, DD1, DD1; VPADDD ctr2StoreAVX2, DD2, DD2; VPADDD ctr3StoreAVX2, DD3, DD3 + VMOVDQA CC3, tmpStoreAVX2 + + // We only hashed 480 of the 512 bytes available - hash the remaining 32 here + polyAdd(480(inp)) + polyMulAVX2 + VPERM2I128 $0x02, AA0, BB0, CC3; VPERM2I128 $0x13, AA0, BB0, BB0; VPERM2I128 $0x02, CC0, DD0, AA0; VPERM2I128 $0x13, CC0, DD0, CC0 + VPXOR (0*32)(inp), CC3, CC3; VPXOR (1*32)(inp), AA0, AA0; VPXOR (2*32)(inp), BB0, BB0; VPXOR (3*32)(inp), CC0, CC0 + VMOVDQU CC3, (0*32)(oup); VMOVDQU AA0, (1*32)(oup); VMOVDQU BB0, (2*32)(oup); VMOVDQU CC0, (3*32)(oup) + VPERM2I128 $0x02, AA1, BB1, AA0; VPERM2I128 $0x02, CC1, DD1, BB0; VPERM2I128 $0x13, AA1, BB1, CC0; VPERM2I128 $0x13, CC1, DD1, DD0 + VPXOR (4*32)(inp), AA0, AA0; VPXOR (5*32)(inp), BB0, BB0; VPXOR (6*32)(inp), CC0, CC0; VPXOR (7*32)(inp), DD0, DD0 + VMOVDQU AA0, (4*32)(oup); VMOVDQU BB0, (5*32)(oup); VMOVDQU CC0, (6*32)(oup); VMOVDQU DD0, (7*32)(oup) + + // and here + polyAdd(496(inp)) + polyMulAVX2 + VPERM2I128 $0x02, AA2, BB2, AA0; VPERM2I128 $0x02, CC2, DD2, BB0; VPERM2I128 $0x13, AA2, BB2, CC0; VPERM2I128 $0x13, CC2, DD2, DD0 + VPXOR (8*32)(inp), AA0, AA0; VPXOR (9*32)(inp), BB0, BB0; VPXOR (10*32)(inp), CC0, CC0; VPXOR (11*32)(inp), DD0, DD0 + VMOVDQU AA0, (8*32)(oup); VMOVDQU BB0, (9*32)(oup); VMOVDQU CC0, (10*32)(oup); VMOVDQU DD0, (11*32)(oup) + VPERM2I128 $0x02, AA3, BB3, AA0; VPERM2I128 $0x02, tmpStoreAVX2, DD3, BB0; VPERM2I128 $0x13, AA3, BB3, CC0; VPERM2I128 $0x13, tmpStoreAVX2, DD3, DD0 + VPXOR (12*32)(inp), AA0, AA0; VPXOR (13*32)(inp), BB0, BB0; VPXOR (14*32)(inp), CC0, CC0; VPXOR (15*32)(inp), DD0, DD0 + VMOVDQU AA0, (12*32)(oup); VMOVDQU BB0, (13*32)(oup); VMOVDQU CC0, (14*32)(oup); VMOVDQU DD0, (15*32)(oup) + LEAQ (32*16)(inp), inp + LEAQ (32*16)(oup), oup + SUBQ $(32*16), inl + JMP openAVX2MainLoop + +openAVX2MainLoopDone: + // Handle the various tail sizes efficiently + TESTQ inl, inl + JE openSSEFinalize + CMPQ inl, $128 + JBE openAVX2Tail128 + CMPQ inl, $256 + JBE openAVX2Tail256 + CMPQ inl, $384 + JBE openAVX2Tail384 + JMP openAVX2Tail512 + +// ---------------------------------------------------------------------------- +// Special optimization for buffers smaller than 193 bytes +openAVX2192: + // For up to 192 bytes of ciphertext and 64 bytes for the poly key, we process four blocks + VMOVDQA AA0, AA1 + VMOVDQA BB0, BB1 + VMOVDQA CC0, CC1 + VPADDD ·avx2IncMask<>(SB), DD0, DD1 + VMOVDQA AA0, AA2 + VMOVDQA BB0, BB2 + VMOVDQA CC0, CC2 + VMOVDQA DD0, DD2 + VMOVDQA DD1, TT3 + MOVQ $10, itr2 + +openAVX2192InnerCipherLoop: + chachaQR_AVX2(AA0, BB0, CC0, DD0, TT0); chachaQR_AVX2(AA1, BB1, CC1, DD1, TT0) + VPALIGNR $4, BB0, BB0, BB0; VPALIGNR $4, BB1, BB1, BB1 + VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $8, CC1, CC1, CC1 + VPALIGNR $12, DD0, DD0, DD0; VPALIGNR $12, DD1, DD1, DD1 + chachaQR_AVX2(AA0, BB0, CC0, DD0, TT0); chachaQR_AVX2(AA1, BB1, CC1, DD1, TT0) + VPALIGNR $12, BB0, BB0, BB0; VPALIGNR $12, BB1, BB1, BB1 + VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $8, CC1, CC1, CC1 + VPALIGNR $4, DD0, DD0, DD0; VPALIGNR $4, DD1, DD1, DD1 + DECQ itr2 + JNE openAVX2192InnerCipherLoop + VPADDD AA2, AA0, AA0; VPADDD AA2, AA1, AA1 + VPADDD BB2, BB0, BB0; VPADDD BB2, BB1, BB1 + VPADDD CC2, CC0, CC0; VPADDD CC2, CC1, CC1 + VPADDD DD2, DD0, DD0; VPADDD TT3, DD1, DD1 + VPERM2I128 $0x02, AA0, BB0, TT0 + + // Clamp and store poly key + VPAND ·polyClampMask<>(SB), TT0, TT0 + VMOVDQA TT0, rsStoreAVX2 + + // Stream for up to 192 bytes + VPERM2I128 $0x13, AA0, BB0, AA0 + VPERM2I128 $0x13, CC0, DD0, BB0 + VPERM2I128 $0x02, AA1, BB1, CC0 + VPERM2I128 $0x02, CC1, DD1, DD0 + VPERM2I128 $0x13, AA1, BB1, AA1 + VPERM2I128 $0x13, CC1, DD1, BB1 + +openAVX2ShortOpen: + // Hash + MOVQ ad_len+80(FP), itr2 + CALL polyHashADInternal<>(SB) + +openAVX2ShortOpenLoop: + CMPQ inl, $32 + JB openAVX2ShortTail32 + SUBQ $32, inl + + // Load for hashing + polyAdd(0*8(inp)) + polyMulAVX2 + polyAdd(2*8(inp)) + polyMulAVX2 + + // Load for decryption + VPXOR (inp), AA0, AA0 + VMOVDQU AA0, (oup) + LEAQ (1*32)(inp), inp + LEAQ (1*32)(oup), oup + + // Shift stream left + VMOVDQA BB0, AA0 + VMOVDQA CC0, BB0 + VMOVDQA DD0, CC0 + VMOVDQA AA1, DD0 + VMOVDQA BB1, AA1 + VMOVDQA CC1, BB1 + VMOVDQA DD1, CC1 + VMOVDQA AA2, DD1 + VMOVDQA BB2, AA2 + JMP openAVX2ShortOpenLoop + +openAVX2ShortTail32: + CMPQ inl, $16 + VMOVDQA A0, A1 + JB openAVX2ShortDone + + SUBQ $16, inl + + // Load for hashing + polyAdd(0*8(inp)) + polyMulAVX2 + + // Load for decryption + VPXOR (inp), A0, T0 + VMOVDQU T0, (oup) + LEAQ (1*16)(inp), inp + LEAQ (1*16)(oup), oup + VPERM2I128 $0x11, AA0, AA0, AA0 + VMOVDQA A0, A1 + +openAVX2ShortDone: + VZEROUPPER + JMP openSSETail16 + +// ---------------------------------------------------------------------------- +// Special optimization for buffers smaller than 321 bytes +openAVX2320: + // For up to 320 bytes of ciphertext and 64 bytes for the poly key, we process six blocks + VMOVDQA AA0, AA1; VMOVDQA BB0, BB1; VMOVDQA CC0, CC1; VPADDD ·avx2IncMask<>(SB), DD0, DD1 + VMOVDQA AA0, AA2; VMOVDQA BB0, BB2; VMOVDQA CC0, CC2; VPADDD ·avx2IncMask<>(SB), DD1, DD2 + VMOVDQA BB0, TT1; VMOVDQA CC0, TT2; VMOVDQA DD0, TT3 + MOVQ $10, itr2 + +openAVX2320InnerCipherLoop: + chachaQR_AVX2(AA0, BB0, CC0, DD0, TT0); chachaQR_AVX2(AA1, BB1, CC1, DD1, TT0); chachaQR_AVX2(AA2, BB2, CC2, DD2, TT0) + VPALIGNR $4, BB0, BB0, BB0; VPALIGNR $4, BB1, BB1, BB1; VPALIGNR $4, BB2, BB2, BB2 + VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $8, CC1, CC1, CC1; VPALIGNR $8, CC2, CC2, CC2 + VPALIGNR $12, DD0, DD0, DD0; VPALIGNR $12, DD1, DD1, DD1; VPALIGNR $12, DD2, DD2, DD2 + chachaQR_AVX2(AA0, BB0, CC0, DD0, TT0); chachaQR_AVX2(AA1, BB1, CC1, DD1, TT0); chachaQR_AVX2(AA2, BB2, CC2, DD2, TT0) + VPALIGNR $12, BB0, BB0, BB0; VPALIGNR $12, BB1, BB1, BB1; VPALIGNR $12, BB2, BB2, BB2 + VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $8, CC1, CC1, CC1; VPALIGNR $8, CC2, CC2, CC2 + VPALIGNR $4, DD0, DD0, DD0; VPALIGNR $4, DD1, DD1, DD1; VPALIGNR $4, DD2, DD2, DD2 + DECQ itr2 + JNE openAVX2320InnerCipherLoop + + VMOVDQA ·chacha20Constants<>(SB), TT0 + VPADDD TT0, AA0, AA0; VPADDD TT0, AA1, AA1; VPADDD TT0, AA2, AA2 + VPADDD TT1, BB0, BB0; VPADDD TT1, BB1, BB1; VPADDD TT1, BB2, BB2 + VPADDD TT2, CC0, CC0; VPADDD TT2, CC1, CC1; VPADDD TT2, CC2, CC2 + VMOVDQA ·avx2IncMask<>(SB), TT0 + VPADDD TT3, DD0, DD0; VPADDD TT0, TT3, TT3 + VPADDD TT3, DD1, DD1; VPADDD TT0, TT3, TT3 + VPADDD TT3, DD2, DD2 + + // Clamp and store poly key + VPERM2I128 $0x02, AA0, BB0, TT0 + VPAND ·polyClampMask<>(SB), TT0, TT0 + VMOVDQA TT0, rsStoreAVX2 + + // Stream for up to 320 bytes + VPERM2I128 $0x13, AA0, BB0, AA0 + VPERM2I128 $0x13, CC0, DD0, BB0 + VPERM2I128 $0x02, AA1, BB1, CC0 + VPERM2I128 $0x02, CC1, DD1, DD0 + VPERM2I128 $0x13, AA1, BB1, AA1 + VPERM2I128 $0x13, CC1, DD1, BB1 + VPERM2I128 $0x02, AA2, BB2, CC1 + VPERM2I128 $0x02, CC2, DD2, DD1 + VPERM2I128 $0x13, AA2, BB2, AA2 + VPERM2I128 $0x13, CC2, DD2, BB2 + JMP openAVX2ShortOpen + +// ---------------------------------------------------------------------------- +// Special optimization for the last 128 bytes of ciphertext +openAVX2Tail128: + // Need to decrypt up to 128 bytes - prepare two blocks + VMOVDQA ·chacha20Constants<>(SB), AA1 + VMOVDQA state1StoreAVX2, BB1 + VMOVDQA state2StoreAVX2, CC1 + VMOVDQA ctr3StoreAVX2, DD1 + VPADDD ·avx2IncMask<>(SB), DD1, DD1 + VMOVDQA DD1, DD0 + + XORQ itr2, itr2 + MOVQ inl, itr1 + ANDQ $-16, itr1 + TESTQ itr1, itr1 + JE openAVX2Tail128LoopB + +openAVX2Tail128LoopA: + // Perform ChaCha rounds, while hashing the remaining input + polyAdd(0(inp)(itr2*1)) + polyMulAVX2 + +openAVX2Tail128LoopB: + ADDQ $16, itr2 + chachaQR_AVX2(AA1, BB1, CC1, DD1, TT0) + VPALIGNR $4, BB1, BB1, BB1 + VPALIGNR $8, CC1, CC1, CC1 + VPALIGNR $12, DD1, DD1, DD1 + chachaQR_AVX2(AA1, BB1, CC1, DD1, TT0) + VPALIGNR $12, BB1, BB1, BB1 + VPALIGNR $8, CC1, CC1, CC1 + VPALIGNR $4, DD1, DD1, DD1 + CMPQ itr2, itr1 + JB openAVX2Tail128LoopA + CMPQ itr2, $160 + JNE openAVX2Tail128LoopB + + VPADDD ·chacha20Constants<>(SB), AA1, AA1 + VPADDD state1StoreAVX2, BB1, BB1 + VPADDD state2StoreAVX2, CC1, CC1 + VPADDD DD0, DD1, DD1 + VPERM2I128 $0x02, AA1, BB1, AA0; VPERM2I128 $0x02, CC1, DD1, BB0; VPERM2I128 $0x13, AA1, BB1, CC0; VPERM2I128 $0x13, CC1, DD1, DD0 + +openAVX2TailLoop: + CMPQ inl, $32 + JB openAVX2Tail + SUBQ $32, inl + + // Load for decryption + VPXOR (inp), AA0, AA0 + VMOVDQU AA0, (oup) + LEAQ (1*32)(inp), inp + LEAQ (1*32)(oup), oup + VMOVDQA BB0, AA0 + VMOVDQA CC0, BB0 + VMOVDQA DD0, CC0 + JMP openAVX2TailLoop + +openAVX2Tail: + CMPQ inl, $16 + VMOVDQA A0, A1 + JB openAVX2TailDone + SUBQ $16, inl + + // Load for decryption + VPXOR (inp), A0, T0 + VMOVDQU T0, (oup) + LEAQ (1*16)(inp), inp + LEAQ (1*16)(oup), oup + VPERM2I128 $0x11, AA0, AA0, AA0 + VMOVDQA A0, A1 + +openAVX2TailDone: + VZEROUPPER + JMP openSSETail16 + +// ---------------------------------------------------------------------------- +// Special optimization for the last 256 bytes of ciphertext +openAVX2Tail256: + // Need to decrypt up to 256 bytes - prepare four blocks + VMOVDQA ·chacha20Constants<>(SB), AA0; VMOVDQA AA0, AA1 + VMOVDQA state1StoreAVX2, BB0; VMOVDQA BB0, BB1 + VMOVDQA state2StoreAVX2, CC0; VMOVDQA CC0, CC1 + VMOVDQA ctr3StoreAVX2, DD0 + VPADDD ·avx2IncMask<>(SB), DD0, DD0 + VPADDD ·avx2IncMask<>(SB), DD0, DD1 + VMOVDQA DD0, TT1 + VMOVDQA DD1, TT2 + + // Compute the number of iterations that will hash data + MOVQ inl, tmpStoreAVX2 + MOVQ inl, itr1 + SUBQ $128, itr1 + SHRQ $4, itr1 + MOVQ $10, itr2 + CMPQ itr1, $10 + CMOVQGT itr2, itr1 + MOVQ inp, inl + XORQ itr2, itr2 + +openAVX2Tail256LoopA: + polyAdd(0(inl)) + polyMulAVX2 + LEAQ 16(inl), inl + + // Perform ChaCha rounds, while hashing the remaining input +openAVX2Tail256LoopB: + chachaQR_AVX2(AA0, BB0, CC0, DD0, TT0); chachaQR_AVX2(AA1, BB1, CC1, DD1, TT0) + VPALIGNR $4, BB0, BB0, BB0; VPALIGNR $4, BB1, BB1, BB1 + VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $8, CC1, CC1, CC1 + VPALIGNR $12, DD0, DD0, DD0; VPALIGNR $12, DD1, DD1, DD1 + INCQ itr2 + chachaQR_AVX2(AA0, BB0, CC0, DD0, TT0); chachaQR_AVX2(AA1, BB1, CC1, DD1, TT0) + VPALIGNR $12, BB0, BB0, BB0; VPALIGNR $12, BB1, BB1, BB1 + VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $8, CC1, CC1, CC1 + VPALIGNR $4, DD0, DD0, DD0; VPALIGNR $4, DD1, DD1, DD1 + CMPQ itr2, itr1 + JB openAVX2Tail256LoopA + + CMPQ itr2, $10 + JNE openAVX2Tail256LoopB + + MOVQ inl, itr2 + SUBQ inp, inl + MOVQ inl, itr1 + MOVQ tmpStoreAVX2, inl + + // Hash the remainder of data (if any) +openAVX2Tail256Hash: + ADDQ $16, itr1 + CMPQ itr1, inl + JGT openAVX2Tail256HashEnd + polyAdd (0(itr2)) + polyMulAVX2 + LEAQ 16(itr2), itr2 + JMP openAVX2Tail256Hash + +// Store 128 bytes safely, then go to store loop +openAVX2Tail256HashEnd: + VPADDD ·chacha20Constants<>(SB), AA0, AA0; VPADDD ·chacha20Constants<>(SB), AA1, AA1 + VPADDD state1StoreAVX2, BB0, BB0; VPADDD state1StoreAVX2, BB1, BB1 + VPADDD state2StoreAVX2, CC0, CC0; VPADDD state2StoreAVX2, CC1, CC1 + VPADDD TT1, DD0, DD0; VPADDD TT2, DD1, DD1 + VPERM2I128 $0x02, AA0, BB0, AA2; VPERM2I128 $0x02, CC0, DD0, BB2; VPERM2I128 $0x13, AA0, BB0, CC2; VPERM2I128 $0x13, CC0, DD0, DD2 + VPERM2I128 $0x02, AA1, BB1, AA0; VPERM2I128 $0x02, CC1, DD1, BB0; VPERM2I128 $0x13, AA1, BB1, CC0; VPERM2I128 $0x13, CC1, DD1, DD0 + + VPXOR (0*32)(inp), AA2, AA2; VPXOR (1*32)(inp), BB2, BB2; VPXOR (2*32)(inp), CC2, CC2; VPXOR (3*32)(inp), DD2, DD2 + VMOVDQU AA2, (0*32)(oup); VMOVDQU BB2, (1*32)(oup); VMOVDQU CC2, (2*32)(oup); VMOVDQU DD2, (3*32)(oup) + LEAQ (4*32)(inp), inp + LEAQ (4*32)(oup), oup + SUBQ $4*32, inl + + JMP openAVX2TailLoop + +// ---------------------------------------------------------------------------- +// Special optimization for the last 384 bytes of ciphertext +openAVX2Tail384: + // Need to decrypt up to 384 bytes - prepare six blocks + VMOVDQA ·chacha20Constants<>(SB), AA0; VMOVDQA AA0, AA1; VMOVDQA AA0, AA2 + VMOVDQA state1StoreAVX2, BB0; VMOVDQA BB0, BB1; VMOVDQA BB0, BB2 + VMOVDQA state2StoreAVX2, CC0; VMOVDQA CC0, CC1; VMOVDQA CC0, CC2 + VMOVDQA ctr3StoreAVX2, DD0 + VPADDD ·avx2IncMask<>(SB), DD0, DD0 + VPADDD ·avx2IncMask<>(SB), DD0, DD1 + VPADDD ·avx2IncMask<>(SB), DD1, DD2 + VMOVDQA DD0, ctr0StoreAVX2 + VMOVDQA DD1, ctr1StoreAVX2 + VMOVDQA DD2, ctr2StoreAVX2 + + // Compute the number of iterations that will hash two blocks of data + MOVQ inl, tmpStoreAVX2 + MOVQ inl, itr1 + SUBQ $256, itr1 + SHRQ $4, itr1 + ADDQ $6, itr1 + MOVQ $10, itr2 + CMPQ itr1, $10 + CMOVQGT itr2, itr1 + MOVQ inp, inl + XORQ itr2, itr2 + + // Perform ChaCha rounds, while hashing the remaining input +openAVX2Tail384LoopB: + polyAdd(0(inl)) + polyMulAVX2 + LEAQ 16(inl), inl + +openAVX2Tail384LoopA: + chachaQR_AVX2(AA0, BB0, CC0, DD0, TT0); chachaQR_AVX2(AA1, BB1, CC1, DD1, TT0); chachaQR_AVX2(AA2, BB2, CC2, DD2, TT0) + VPALIGNR $4, BB0, BB0, BB0; VPALIGNR $4, BB1, BB1, BB1; VPALIGNR $4, BB2, BB2, BB2 + VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $8, CC1, CC1, CC1; VPALIGNR $8, CC2, CC2, CC2 + VPALIGNR $12, DD0, DD0, DD0; VPALIGNR $12, DD1, DD1, DD1; VPALIGNR $12, DD2, DD2, DD2 + polyAdd(0(inl)) + polyMulAVX2 + LEAQ 16(inl), inl + INCQ itr2 + chachaQR_AVX2(AA0, BB0, CC0, DD0, TT0); chachaQR_AVX2(AA1, BB1, CC1, DD1, TT0); chachaQR_AVX2(AA2, BB2, CC2, DD2, TT0) + VPALIGNR $12, BB0, BB0, BB0; VPALIGNR $12, BB1, BB1, BB1; VPALIGNR $12, BB2, BB2, BB2 + VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $8, CC1, CC1, CC1; VPALIGNR $8, CC2, CC2, CC2 + VPALIGNR $4, DD0, DD0, DD0; VPALIGNR $4, DD1, DD1, DD1; VPALIGNR $4, DD2, DD2, DD2 + + CMPQ itr2, itr1 + JB openAVX2Tail384LoopB + + CMPQ itr2, $10 + JNE openAVX2Tail384LoopA + + MOVQ inl, itr2 + SUBQ inp, inl + MOVQ inl, itr1 + MOVQ tmpStoreAVX2, inl + +openAVX2Tail384Hash: + ADDQ $16, itr1 + CMPQ itr1, inl + JGT openAVX2Tail384HashEnd + polyAdd(0(itr2)) + polyMulAVX2 + LEAQ 16(itr2), itr2 + JMP openAVX2Tail384Hash + +// Store 256 bytes safely, then go to store loop +openAVX2Tail384HashEnd: + VPADDD ·chacha20Constants<>(SB), AA0, AA0; VPADDD ·chacha20Constants<>(SB), AA1, AA1; VPADDD ·chacha20Constants<>(SB), AA2, AA2 + VPADDD state1StoreAVX2, BB0, BB0; VPADDD state1StoreAVX2, BB1, BB1; VPADDD state1StoreAVX2, BB2, BB2 + VPADDD state2StoreAVX2, CC0, CC0; VPADDD state2StoreAVX2, CC1, CC1; VPADDD state2StoreAVX2, CC2, CC2 + VPADDD ctr0StoreAVX2, DD0, DD0; VPADDD ctr1StoreAVX2, DD1, DD1; VPADDD ctr2StoreAVX2, DD2, DD2 + VPERM2I128 $0x02, AA0, BB0, TT0; VPERM2I128 $0x02, CC0, DD0, TT1; VPERM2I128 $0x13, AA0, BB0, TT2; VPERM2I128 $0x13, CC0, DD0, TT3 + VPXOR (0*32)(inp), TT0, TT0; VPXOR (1*32)(inp), TT1, TT1; VPXOR (2*32)(inp), TT2, TT2; VPXOR (3*32)(inp), TT3, TT3 + VMOVDQU TT0, (0*32)(oup); VMOVDQU TT1, (1*32)(oup); VMOVDQU TT2, (2*32)(oup); VMOVDQU TT3, (3*32)(oup) + VPERM2I128 $0x02, AA1, BB1, TT0; VPERM2I128 $0x02, CC1, DD1, TT1; VPERM2I128 $0x13, AA1, BB1, TT2; VPERM2I128 $0x13, CC1, DD1, TT3 + VPXOR (4*32)(inp), TT0, TT0; VPXOR (5*32)(inp), TT1, TT1; VPXOR (6*32)(inp), TT2, TT2; VPXOR (7*32)(inp), TT3, TT3 + VMOVDQU TT0, (4*32)(oup); VMOVDQU TT1, (5*32)(oup); VMOVDQU TT2, (6*32)(oup); VMOVDQU TT3, (7*32)(oup) + VPERM2I128 $0x02, AA2, BB2, AA0; VPERM2I128 $0x02, CC2, DD2, BB0; VPERM2I128 $0x13, AA2, BB2, CC0; VPERM2I128 $0x13, CC2, DD2, DD0 + LEAQ (8*32)(inp), inp + LEAQ (8*32)(oup), oup + SUBQ $8*32, inl + JMP openAVX2TailLoop + +// ---------------------------------------------------------------------------- +// Special optimization for the last 512 bytes of ciphertext +openAVX2Tail512: + VMOVDQU ·chacha20Constants<>(SB), AA0; VMOVDQA AA0, AA1; VMOVDQA AA0, AA2; VMOVDQA AA0, AA3 + VMOVDQA state1StoreAVX2, BB0; VMOVDQA BB0, BB1; VMOVDQA BB0, BB2; VMOVDQA BB0, BB3 + VMOVDQA state2StoreAVX2, CC0; VMOVDQA CC0, CC1; VMOVDQA CC0, CC2; VMOVDQA CC0, CC3 + VMOVDQA ctr3StoreAVX2, DD0; VPADDD ·avx2IncMask<>(SB), DD0, DD0; VPADDD ·avx2IncMask<>(SB), DD0, DD1; VPADDD ·avx2IncMask<>(SB), DD1, DD2; VPADDD ·avx2IncMask<>(SB), DD2, DD3 + VMOVDQA DD0, ctr0StoreAVX2; VMOVDQA DD1, ctr1StoreAVX2; VMOVDQA DD2, ctr2StoreAVX2; VMOVDQA DD3, ctr3StoreAVX2 + XORQ itr1, itr1 + MOVQ inp, itr2 + +openAVX2Tail512LoopB: + polyAdd(0(itr2)) + polyMulAVX2 + LEAQ (2*8)(itr2), itr2 + +openAVX2Tail512LoopA: + VPADDD BB0, AA0, AA0; VPADDD BB1, AA1, AA1; VPADDD BB2, AA2, AA2; VPADDD BB3, AA3, AA3 + VPXOR AA0, DD0, DD0; VPXOR AA1, DD1, DD1; VPXOR AA2, DD2, DD2; VPXOR AA3, DD3, DD3 + VPSHUFB ·rol16<>(SB), DD0, DD0; VPSHUFB ·rol16<>(SB), DD1, DD1; VPSHUFB ·rol16<>(SB), DD2, DD2; VPSHUFB ·rol16<>(SB), DD3, DD3 + VPADDD DD0, CC0, CC0; VPADDD DD1, CC1, CC1; VPADDD DD2, CC2, CC2; VPADDD DD3, CC3, CC3 + VPXOR CC0, BB0, BB0; VPXOR CC1, BB1, BB1; VPXOR CC2, BB2, BB2; VPXOR CC3, BB3, BB3 + VMOVDQA CC3, tmpStoreAVX2 + VPSLLD $12, BB0, CC3; VPSRLD $20, BB0, BB0; VPXOR CC3, BB0, BB0 + VPSLLD $12, BB1, CC3; VPSRLD $20, BB1, BB1; VPXOR CC3, BB1, BB1 + VPSLLD $12, BB2, CC3; VPSRLD $20, BB2, BB2; VPXOR CC3, BB2, BB2 + VPSLLD $12, BB3, CC3; VPSRLD $20, BB3, BB3; VPXOR CC3, BB3, BB3 + VMOVDQA tmpStoreAVX2, CC3 + polyAdd(0*8(itr2)) + polyMulAVX2 + VPADDD BB0, AA0, AA0; VPADDD BB1, AA1, AA1; VPADDD BB2, AA2, AA2; VPADDD BB3, AA3, AA3 + VPXOR AA0, DD0, DD0; VPXOR AA1, DD1, DD1; VPXOR AA2, DD2, DD2; VPXOR AA3, DD3, DD3 + VPSHUFB ·rol8<>(SB), DD0, DD0; VPSHUFB ·rol8<>(SB), DD1, DD1; VPSHUFB ·rol8<>(SB), DD2, DD2; VPSHUFB ·rol8<>(SB), DD3, DD3 + VPADDD DD0, CC0, CC0; VPADDD DD1, CC1, CC1; VPADDD DD2, CC2, CC2; VPADDD DD3, CC3, CC3 + VPXOR CC0, BB0, BB0; VPXOR CC1, BB1, BB1; VPXOR CC2, BB2, BB2; VPXOR CC3, BB3, BB3 + VMOVDQA CC3, tmpStoreAVX2 + VPSLLD $7, BB0, CC3; VPSRLD $25, BB0, BB0; VPXOR CC3, BB0, BB0 + VPSLLD $7, BB1, CC3; VPSRLD $25, BB1, BB1; VPXOR CC3, BB1, BB1 + VPSLLD $7, BB2, CC3; VPSRLD $25, BB2, BB2; VPXOR CC3, BB2, BB2 + VPSLLD $7, BB3, CC3; VPSRLD $25, BB3, BB3; VPXOR CC3, BB3, BB3 + VMOVDQA tmpStoreAVX2, CC3 + VPALIGNR $4, BB0, BB0, BB0; VPALIGNR $4, BB1, BB1, BB1; VPALIGNR $4, BB2, BB2, BB2; VPALIGNR $4, BB3, BB3, BB3 + VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $8, CC1, CC1, CC1; VPALIGNR $8, CC2, CC2, CC2; VPALIGNR $8, CC3, CC3, CC3 + VPALIGNR $12, DD0, DD0, DD0; VPALIGNR $12, DD1, DD1, DD1; VPALIGNR $12, DD2, DD2, DD2; VPALIGNR $12, DD3, DD3, DD3 + VPADDD BB0, AA0, AA0; VPADDD BB1, AA1, AA1; VPADDD BB2, AA2, AA2; VPADDD BB3, AA3, AA3 + VPXOR AA0, DD0, DD0; VPXOR AA1, DD1, DD1; VPXOR AA2, DD2, DD2; VPXOR AA3, DD3, DD3 + VPSHUFB ·rol16<>(SB), DD0, DD0; VPSHUFB ·rol16<>(SB), DD1, DD1; VPSHUFB ·rol16<>(SB), DD2, DD2; VPSHUFB ·rol16<>(SB), DD3, DD3 + VPADDD DD0, CC0, CC0; VPADDD DD1, CC1, CC1; VPADDD DD2, CC2, CC2; VPADDD DD3, CC3, CC3 + VPXOR CC0, BB0, BB0; VPXOR CC1, BB1, BB1; VPXOR CC2, BB2, BB2; VPXOR CC3, BB3, BB3 + polyAdd(2*8(itr2)) + polyMulAVX2 + LEAQ (4*8)(itr2), itr2 + VMOVDQA CC3, tmpStoreAVX2 + VPSLLD $12, BB0, CC3; VPSRLD $20, BB0, BB0; VPXOR CC3, BB0, BB0 + VPSLLD $12, BB1, CC3; VPSRLD $20, BB1, BB1; VPXOR CC3, BB1, BB1 + VPSLLD $12, BB2, CC3; VPSRLD $20, BB2, BB2; VPXOR CC3, BB2, BB2 + VPSLLD $12, BB3, CC3; VPSRLD $20, BB3, BB3; VPXOR CC3, BB3, BB3 + VMOVDQA tmpStoreAVX2, CC3 + VPADDD BB0, AA0, AA0; VPADDD BB1, AA1, AA1; VPADDD BB2, AA2, AA2; VPADDD BB3, AA3, AA3 + VPXOR AA0, DD0, DD0; VPXOR AA1, DD1, DD1; VPXOR AA2, DD2, DD2; VPXOR AA3, DD3, DD3 + VPSHUFB ·rol8<>(SB), DD0, DD0; VPSHUFB ·rol8<>(SB), DD1, DD1; VPSHUFB ·rol8<>(SB), DD2, DD2; VPSHUFB ·rol8<>(SB), DD3, DD3 + VPADDD DD0, CC0, CC0; VPADDD DD1, CC1, CC1; VPADDD DD2, CC2, CC2; VPADDD DD3, CC3, CC3 + VPXOR CC0, BB0, BB0; VPXOR CC1, BB1, BB1; VPXOR CC2, BB2, BB2; VPXOR CC3, BB3, BB3 + VMOVDQA CC3, tmpStoreAVX2 + VPSLLD $7, BB0, CC3; VPSRLD $25, BB0, BB0; VPXOR CC3, BB0, BB0 + VPSLLD $7, BB1, CC3; VPSRLD $25, BB1, BB1; VPXOR CC3, BB1, BB1 + VPSLLD $7, BB2, CC3; VPSRLD $25, BB2, BB2; VPXOR CC3, BB2, BB2 + VPSLLD $7, BB3, CC3; VPSRLD $25, BB3, BB3; VPXOR CC3, BB3, BB3 + VMOVDQA tmpStoreAVX2, CC3 + VPALIGNR $12, BB0, BB0, BB0; VPALIGNR $12, BB1, BB1, BB1; VPALIGNR $12, BB2, BB2, BB2; VPALIGNR $12, BB3, BB3, BB3 + VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $8, CC1, CC1, CC1; VPALIGNR $8, CC2, CC2, CC2; VPALIGNR $8, CC3, CC3, CC3 + VPALIGNR $4, DD0, DD0, DD0; VPALIGNR $4, DD1, DD1, DD1; VPALIGNR $4, DD2, DD2, DD2; VPALIGNR $4, DD3, DD3, DD3 + INCQ itr1 + CMPQ itr1, $4 + JLT openAVX2Tail512LoopB + + CMPQ itr1, $10 + JNE openAVX2Tail512LoopA + + MOVQ inl, itr1 + SUBQ $384, itr1 + ANDQ $-16, itr1 + +openAVX2Tail512HashLoop: + TESTQ itr1, itr1 + JE openAVX2Tail512HashEnd + polyAdd(0(itr2)) + polyMulAVX2 + LEAQ 16(itr2), itr2 + SUBQ $16, itr1 + JMP openAVX2Tail512HashLoop + +openAVX2Tail512HashEnd: + VPADDD ·chacha20Constants<>(SB), AA0, AA0; VPADDD ·chacha20Constants<>(SB), AA1, AA1; VPADDD ·chacha20Constants<>(SB), AA2, AA2; VPADDD ·chacha20Constants<>(SB), AA3, AA3 + VPADDD state1StoreAVX2, BB0, BB0; VPADDD state1StoreAVX2, BB1, BB1; VPADDD state1StoreAVX2, BB2, BB2; VPADDD state1StoreAVX2, BB3, BB3 + VPADDD state2StoreAVX2, CC0, CC0; VPADDD state2StoreAVX2, CC1, CC1; VPADDD state2StoreAVX2, CC2, CC2; VPADDD state2StoreAVX2, CC3, CC3 + VPADDD ctr0StoreAVX2, DD0, DD0; VPADDD ctr1StoreAVX2, DD1, DD1; VPADDD ctr2StoreAVX2, DD2, DD2; VPADDD ctr3StoreAVX2, DD3, DD3 + VMOVDQA CC3, tmpStoreAVX2 + VPERM2I128 $0x02, AA0, BB0, CC3; VPERM2I128 $0x13, AA0, BB0, BB0; VPERM2I128 $0x02, CC0, DD0, AA0; VPERM2I128 $0x13, CC0, DD0, CC0 + VPXOR (0*32)(inp), CC3, CC3; VPXOR (1*32)(inp), AA0, AA0; VPXOR (2*32)(inp), BB0, BB0; VPXOR (3*32)(inp), CC0, CC0 + VMOVDQU CC3, (0*32)(oup); VMOVDQU AA0, (1*32)(oup); VMOVDQU BB0, (2*32)(oup); VMOVDQU CC0, (3*32)(oup) + VPERM2I128 $0x02, AA1, BB1, AA0; VPERM2I128 $0x02, CC1, DD1, BB0; VPERM2I128 $0x13, AA1, BB1, CC0; VPERM2I128 $0x13, CC1, DD1, DD0 + VPXOR (4*32)(inp), AA0, AA0; VPXOR (5*32)(inp), BB0, BB0; VPXOR (6*32)(inp), CC0, CC0; VPXOR (7*32)(inp), DD0, DD0 + VMOVDQU AA0, (4*32)(oup); VMOVDQU BB0, (5*32)(oup); VMOVDQU CC0, (6*32)(oup); VMOVDQU DD0, (7*32)(oup) + VPERM2I128 $0x02, AA2, BB2, AA0; VPERM2I128 $0x02, CC2, DD2, BB0; VPERM2I128 $0x13, AA2, BB2, CC0; VPERM2I128 $0x13, CC2, DD2, DD0 + VPXOR (8*32)(inp), AA0, AA0; VPXOR (9*32)(inp), BB0, BB0; VPXOR (10*32)(inp), CC0, CC0; VPXOR (11*32)(inp), DD0, DD0 + VMOVDQU AA0, (8*32)(oup); VMOVDQU BB0, (9*32)(oup); VMOVDQU CC0, (10*32)(oup); VMOVDQU DD0, (11*32)(oup) + VPERM2I128 $0x02, AA3, BB3, AA0; VPERM2I128 $0x02, tmpStoreAVX2, DD3, BB0; VPERM2I128 $0x13, AA3, BB3, CC0; VPERM2I128 $0x13, tmpStoreAVX2, DD3, DD0 + + LEAQ (12*32)(inp), inp + LEAQ (12*32)(oup), oup + SUBQ $12*32, inl + + JMP openAVX2TailLoop + +// ---------------------------------------------------------------------------- +// ---------------------------------------------------------------------------- +// func chacha20Poly1305Seal(dst, key, src, ad []byte) +TEXT ·chacha20Poly1305Seal(SB), 0, $288-96 + // For aligned stack access + MOVQ SP, BP + ADDQ $32, BP + ANDQ $-32, BP + MOVQ dst+0(FP), oup + MOVQ key+24(FP), keyp + MOVQ src+48(FP), inp + MOVQ src_len+56(FP), inl + MOVQ ad+72(FP), adp + + CMPB ·useAVX2(SB), $1 + JE chacha20Poly1305Seal_AVX2 + + // Special optimization, for very short buffers + CMPQ inl, $128 + JBE sealSSE128 // About 15% faster + + // In the seal case - prepare the poly key + 3 blocks of stream in the first iteration + MOVOU ·chacha20Constants<>(SB), A0 + MOVOU (1*16)(keyp), B0 + MOVOU (2*16)(keyp), C0 + MOVOU (3*16)(keyp), D0 + + // Store state on stack for future use + MOVO B0, state1Store + MOVO C0, state2Store + + // Load state, increment counter blocks + MOVO A0, A1; MOVO B0, B1; MOVO C0, C1; MOVO D0, D1; PADDL ·sseIncMask<>(SB), D1 + MOVO A1, A2; MOVO B1, B2; MOVO C1, C2; MOVO D1, D2; PADDL ·sseIncMask<>(SB), D2 + MOVO A2, A3; MOVO B2, B3; MOVO C2, C3; MOVO D2, D3; PADDL ·sseIncMask<>(SB), D3 + + // Store counters + MOVO D0, ctr0Store; MOVO D1, ctr1Store; MOVO D2, ctr2Store; MOVO D3, ctr3Store + MOVQ $10, itr2 + +sealSSEIntroLoop: + MOVO C3, tmpStore + chachaQR(A0, B0, C0, D0, C3); chachaQR(A1, B1, C1, D1, C3); chachaQR(A2, B2, C2, D2, C3) + MOVO tmpStore, C3 + MOVO C1, tmpStore + chachaQR(A3, B3, C3, D3, C1) + MOVO tmpStore, C1 + shiftB0Left; shiftB1Left; shiftB2Left; shiftB3Left + shiftC0Left; shiftC1Left; shiftC2Left; shiftC3Left + shiftD0Left; shiftD1Left; shiftD2Left; shiftD3Left + + MOVO C3, tmpStore + chachaQR(A0, B0, C0, D0, C3); chachaQR(A1, B1, C1, D1, C3); chachaQR(A2, B2, C2, D2, C3) + MOVO tmpStore, C3 + MOVO C1, tmpStore + chachaQR(A3, B3, C3, D3, C1) + MOVO tmpStore, C1 + shiftB0Right; shiftB1Right; shiftB2Right; shiftB3Right + shiftC0Right; shiftC1Right; shiftC2Right; shiftC3Right + shiftD0Right; shiftD1Right; shiftD2Right; shiftD3Right + DECQ itr2 + JNE sealSSEIntroLoop + + // Add in the state + PADDD ·chacha20Constants<>(SB), A0; PADDD ·chacha20Constants<>(SB), A1; PADDD ·chacha20Constants<>(SB), A2; PADDD ·chacha20Constants<>(SB), A3 + PADDD state1Store, B0; PADDD state1Store, B1; PADDD state1Store, B2; PADDD state1Store, B3 + PADDD state2Store, C1; PADDD state2Store, C2; PADDD state2Store, C3 + PADDD ctr1Store, D1; PADDD ctr2Store, D2; PADDD ctr3Store, D3 + + // Clamp and store the key + PAND ·polyClampMask<>(SB), A0 + MOVO A0, rStore + MOVO B0, sStore + + // Hash AAD + MOVQ ad_len+80(FP), itr2 + CALL polyHashADInternal<>(SB) + + MOVOU (0*16)(inp), A0; MOVOU (1*16)(inp), B0; MOVOU (2*16)(inp), C0; MOVOU (3*16)(inp), D0 + PXOR A0, A1; PXOR B0, B1; PXOR C0, C1; PXOR D0, D1 + MOVOU A1, (0*16)(oup); MOVOU B1, (1*16)(oup); MOVOU C1, (2*16)(oup); MOVOU D1, (3*16)(oup) + MOVOU (4*16)(inp), A0; MOVOU (5*16)(inp), B0; MOVOU (6*16)(inp), C0; MOVOU (7*16)(inp), D0 + PXOR A0, A2; PXOR B0, B2; PXOR C0, C2; PXOR D0, D2 + MOVOU A2, (4*16)(oup); MOVOU B2, (5*16)(oup); MOVOU C2, (6*16)(oup); MOVOU D2, (7*16)(oup) + + MOVQ $128, itr1 + SUBQ $128, inl + LEAQ 128(inp), inp + + MOVO A3, A1; MOVO B3, B1; MOVO C3, C1; MOVO D3, D1 + + CMPQ inl, $64 + JBE sealSSE128SealHash + + MOVOU (0*16)(inp), A0; MOVOU (1*16)(inp), B0; MOVOU (2*16)(inp), C0; MOVOU (3*16)(inp), D0 + PXOR A0, A3; PXOR B0, B3; PXOR C0, C3; PXOR D0, D3 + MOVOU A3, (8*16)(oup); MOVOU B3, (9*16)(oup); MOVOU C3, (10*16)(oup); MOVOU D3, (11*16)(oup) + + ADDQ $64, itr1 + SUBQ $64, inl + LEAQ 64(inp), inp + + MOVQ $2, itr1 + MOVQ $8, itr2 + + CMPQ inl, $64 + JBE sealSSETail64 + CMPQ inl, $128 + JBE sealSSETail128 + CMPQ inl, $192 + JBE sealSSETail192 + +sealSSEMainLoop: + // Load state, increment counter blocks + MOVO ·chacha20Constants<>(SB), A0; MOVO state1Store, B0; MOVO state2Store, C0; MOVO ctr3Store, D0; PADDL ·sseIncMask<>(SB), D0 + MOVO A0, A1; MOVO B0, B1; MOVO C0, C1; MOVO D0, D1; PADDL ·sseIncMask<>(SB), D1 + MOVO A1, A2; MOVO B1, B2; MOVO C1, C2; MOVO D1, D2; PADDL ·sseIncMask<>(SB), D2 + MOVO A2, A3; MOVO B2, B3; MOVO C2, C3; MOVO D2, D3; PADDL ·sseIncMask<>(SB), D3 + + // Store counters + MOVO D0, ctr0Store; MOVO D1, ctr1Store; MOVO D2, ctr2Store; MOVO D3, ctr3Store + +sealSSEInnerLoop: + MOVO C3, tmpStore + chachaQR(A0, B0, C0, D0, C3); chachaQR(A1, B1, C1, D1, C3); chachaQR(A2, B2, C2, D2, C3) + MOVO tmpStore, C3 + MOVO C1, tmpStore + chachaQR(A3, B3, C3, D3, C1) + MOVO tmpStore, C1 + polyAdd(0(oup)) + shiftB0Left; shiftB1Left; shiftB2Left; shiftB3Left + shiftC0Left; shiftC1Left; shiftC2Left; shiftC3Left + shiftD0Left; shiftD1Left; shiftD2Left; shiftD3Left + polyMulStage1 + polyMulStage2 + LEAQ (2*8)(oup), oup + MOVO C3, tmpStore + chachaQR(A0, B0, C0, D0, C3); chachaQR(A1, B1, C1, D1, C3); chachaQR(A2, B2, C2, D2, C3) + MOVO tmpStore, C3 + MOVO C1, tmpStore + polyMulStage3 + chachaQR(A3, B3, C3, D3, C1) + MOVO tmpStore, C1 + polyMulReduceStage + shiftB0Right; shiftB1Right; shiftB2Right; shiftB3Right + shiftC0Right; shiftC1Right; shiftC2Right; shiftC3Right + shiftD0Right; shiftD1Right; shiftD2Right; shiftD3Right + DECQ itr2 + JGE sealSSEInnerLoop + polyAdd(0(oup)) + polyMul + LEAQ (2*8)(oup), oup + DECQ itr1 + JG sealSSEInnerLoop + + // Add in the state + PADDD ·chacha20Constants<>(SB), A0; PADDD ·chacha20Constants<>(SB), A1; PADDD ·chacha20Constants<>(SB), A2; PADDD ·chacha20Constants<>(SB), A3 + PADDD state1Store, B0; PADDD state1Store, B1; PADDD state1Store, B2; PADDD state1Store, B3 + PADDD state2Store, C0; PADDD state2Store, C1; PADDD state2Store, C2; PADDD state2Store, C3 + PADDD ctr0Store, D0; PADDD ctr1Store, D1; PADDD ctr2Store, D2; PADDD ctr3Store, D3 + MOVO D3, tmpStore + + // Load - xor - store + MOVOU (0*16)(inp), D3; PXOR D3, A0 + MOVOU (1*16)(inp), D3; PXOR D3, B0 + MOVOU (2*16)(inp), D3; PXOR D3, C0 + MOVOU (3*16)(inp), D3; PXOR D3, D0 + MOVOU A0, (0*16)(oup) + MOVOU B0, (1*16)(oup) + MOVOU C0, (2*16)(oup) + MOVOU D0, (3*16)(oup) + MOVO tmpStore, D3 + + MOVOU (4*16)(inp), A0; MOVOU (5*16)(inp), B0; MOVOU (6*16)(inp), C0; MOVOU (7*16)(inp), D0 + PXOR A0, A1; PXOR B0, B1; PXOR C0, C1; PXOR D0, D1 + MOVOU A1, (4*16)(oup); MOVOU B1, (5*16)(oup); MOVOU C1, (6*16)(oup); MOVOU D1, (7*16)(oup) + MOVOU (8*16)(inp), A0; MOVOU (9*16)(inp), B0; MOVOU (10*16)(inp), C0; MOVOU (11*16)(inp), D0 + PXOR A0, A2; PXOR B0, B2; PXOR C0, C2; PXOR D0, D2 + MOVOU A2, (8*16)(oup); MOVOU B2, (9*16)(oup); MOVOU C2, (10*16)(oup); MOVOU D2, (11*16)(oup) + ADDQ $192, inp + MOVQ $192, itr1 + SUBQ $192, inl + MOVO A3, A1 + MOVO B3, B1 + MOVO C3, C1 + MOVO D3, D1 + CMPQ inl, $64 + JBE sealSSE128SealHash + MOVOU (0*16)(inp), A0; MOVOU (1*16)(inp), B0; MOVOU (2*16)(inp), C0; MOVOU (3*16)(inp), D0 + PXOR A0, A3; PXOR B0, B3; PXOR C0, C3; PXOR D0, D3 + MOVOU A3, (12*16)(oup); MOVOU B3, (13*16)(oup); MOVOU C3, (14*16)(oup); MOVOU D3, (15*16)(oup) + LEAQ 64(inp), inp + SUBQ $64, inl + MOVQ $6, itr1 + MOVQ $4, itr2 + CMPQ inl, $192 + JG sealSSEMainLoop + + MOVQ inl, itr1 + TESTQ inl, inl + JE sealSSE128SealHash + MOVQ $6, itr1 + CMPQ inl, $64 + JBE sealSSETail64 + CMPQ inl, $128 + JBE sealSSETail128 + JMP sealSSETail192 + +// ---------------------------------------------------------------------------- +// Special optimization for the last 64 bytes of plaintext +sealSSETail64: + // Need to encrypt up to 64 bytes - prepare single block, hash 192 or 256 bytes + MOVO ·chacha20Constants<>(SB), A1 + MOVO state1Store, B1 + MOVO state2Store, C1 + MOVO ctr3Store, D1 + PADDL ·sseIncMask<>(SB), D1 + MOVO D1, ctr0Store + +sealSSETail64LoopA: + // Perform ChaCha rounds, while hashing the previously encrypted ciphertext + polyAdd(0(oup)) + polyMul + LEAQ 16(oup), oup + +sealSSETail64LoopB: + chachaQR(A1, B1, C1, D1, T1) + shiftB1Left; shiftC1Left; shiftD1Left + chachaQR(A1, B1, C1, D1, T1) + shiftB1Right; shiftC1Right; shiftD1Right + polyAdd(0(oup)) + polyMul + LEAQ 16(oup), oup + + DECQ itr1 + JG sealSSETail64LoopA + + DECQ itr2 + JGE sealSSETail64LoopB + PADDL ·chacha20Constants<>(SB), A1 + PADDL state1Store, B1 + PADDL state2Store, C1 + PADDL ctr0Store, D1 + + JMP sealSSE128Seal + +// ---------------------------------------------------------------------------- +// Special optimization for the last 128 bytes of plaintext +sealSSETail128: + // Need to encrypt up to 128 bytes - prepare two blocks, hash 192 or 256 bytes + MOVO ·chacha20Constants<>(SB), A0; MOVO state1Store, B0; MOVO state2Store, C0; MOVO ctr3Store, D0; PADDL ·sseIncMask<>(SB), D0; MOVO D0, ctr0Store + MOVO A0, A1; MOVO B0, B1; MOVO C0, C1; MOVO D0, D1; PADDL ·sseIncMask<>(SB), D1; MOVO D1, ctr1Store + +sealSSETail128LoopA: + // Perform ChaCha rounds, while hashing the previously encrypted ciphertext + polyAdd(0(oup)) + polyMul + LEAQ 16(oup), oup + +sealSSETail128LoopB: + chachaQR(A0, B0, C0, D0, T0); chachaQR(A1, B1, C1, D1, T0) + shiftB0Left; shiftC0Left; shiftD0Left + shiftB1Left; shiftC1Left; shiftD1Left + polyAdd(0(oup)) + polyMul + LEAQ 16(oup), oup + chachaQR(A0, B0, C0, D0, T0); chachaQR(A1, B1, C1, D1, T0) + shiftB0Right; shiftC0Right; shiftD0Right + shiftB1Right; shiftC1Right; shiftD1Right + + DECQ itr1 + JG sealSSETail128LoopA + + DECQ itr2 + JGE sealSSETail128LoopB + + PADDL ·chacha20Constants<>(SB), A0; PADDL ·chacha20Constants<>(SB), A1 + PADDL state1Store, B0; PADDL state1Store, B1 + PADDL state2Store, C0; PADDL state2Store, C1 + PADDL ctr0Store, D0; PADDL ctr1Store, D1 + + MOVOU (0*16)(inp), T0; MOVOU (1*16)(inp), T1; MOVOU (2*16)(inp), T2; MOVOU (3*16)(inp), T3 + PXOR T0, A0; PXOR T1, B0; PXOR T2, C0; PXOR T3, D0 + MOVOU A0, (0*16)(oup); MOVOU B0, (1*16)(oup); MOVOU C0, (2*16)(oup); MOVOU D0, (3*16)(oup) + + MOVQ $64, itr1 + LEAQ 64(inp), inp + SUBQ $64, inl + + JMP sealSSE128SealHash + +// ---------------------------------------------------------------------------- +// Special optimization for the last 192 bytes of plaintext +sealSSETail192: + // Need to encrypt up to 192 bytes - prepare three blocks, hash 192 or 256 bytes + MOVO ·chacha20Constants<>(SB), A0; MOVO state1Store, B0; MOVO state2Store, C0; MOVO ctr3Store, D0; PADDL ·sseIncMask<>(SB), D0; MOVO D0, ctr0Store + MOVO A0, A1; MOVO B0, B1; MOVO C0, C1; MOVO D0, D1; PADDL ·sseIncMask<>(SB), D1; MOVO D1, ctr1Store + MOVO A1, A2; MOVO B1, B2; MOVO C1, C2; MOVO D1, D2; PADDL ·sseIncMask<>(SB), D2; MOVO D2, ctr2Store + +sealSSETail192LoopA: + // Perform ChaCha rounds, while hashing the previously encrypted ciphertext + polyAdd(0(oup)) + polyMul + LEAQ 16(oup), oup + +sealSSETail192LoopB: + chachaQR(A0, B0, C0, D0, T0); chachaQR(A1, B1, C1, D1, T0); chachaQR(A2, B2, C2, D2, T0) + shiftB0Left; shiftC0Left; shiftD0Left + shiftB1Left; shiftC1Left; shiftD1Left + shiftB2Left; shiftC2Left; shiftD2Left + + polyAdd(0(oup)) + polyMul + LEAQ 16(oup), oup + + chachaQR(A0, B0, C0, D0, T0); chachaQR(A1, B1, C1, D1, T0); chachaQR(A2, B2, C2, D2, T0) + shiftB0Right; shiftC0Right; shiftD0Right + shiftB1Right; shiftC1Right; shiftD1Right + shiftB2Right; shiftC2Right; shiftD2Right + + DECQ itr1 + JG sealSSETail192LoopA + + DECQ itr2 + JGE sealSSETail192LoopB + + PADDL ·chacha20Constants<>(SB), A0; PADDL ·chacha20Constants<>(SB), A1; PADDL ·chacha20Constants<>(SB), A2 + PADDL state1Store, B0; PADDL state1Store, B1; PADDL state1Store, B2 + PADDL state2Store, C0; PADDL state2Store, C1; PADDL state2Store, C2 + PADDL ctr0Store, D0; PADDL ctr1Store, D1; PADDL ctr2Store, D2 + + MOVOU (0*16)(inp), T0; MOVOU (1*16)(inp), T1; MOVOU (2*16)(inp), T2; MOVOU (3*16)(inp), T3 + PXOR T0, A0; PXOR T1, B0; PXOR T2, C0; PXOR T3, D0 + MOVOU A0, (0*16)(oup); MOVOU B0, (1*16)(oup); MOVOU C0, (2*16)(oup); MOVOU D0, (3*16)(oup) + MOVOU (4*16)(inp), T0; MOVOU (5*16)(inp), T1; MOVOU (6*16)(inp), T2; MOVOU (7*16)(inp), T3 + PXOR T0, A1; PXOR T1, B1; PXOR T2, C1; PXOR T3, D1 + MOVOU A1, (4*16)(oup); MOVOU B1, (5*16)(oup); MOVOU C1, (6*16)(oup); MOVOU D1, (7*16)(oup) + + MOVO A2, A1 + MOVO B2, B1 + MOVO C2, C1 + MOVO D2, D1 + MOVQ $128, itr1 + LEAQ 128(inp), inp + SUBQ $128, inl + + JMP sealSSE128SealHash + +// ---------------------------------------------------------------------------- +// Special seal optimization for buffers smaller than 129 bytes +sealSSE128: + // For up to 128 bytes of ciphertext and 64 bytes for the poly key, we require to process three blocks + MOVOU ·chacha20Constants<>(SB), A0; MOVOU (1*16)(keyp), B0; MOVOU (2*16)(keyp), C0; MOVOU (3*16)(keyp), D0 + MOVO A0, A1; MOVO B0, B1; MOVO C0, C1; MOVO D0, D1; PADDL ·sseIncMask<>(SB), D1 + MOVO A1, A2; MOVO B1, B2; MOVO C1, C2; MOVO D1, D2; PADDL ·sseIncMask<>(SB), D2 + MOVO B0, T1; MOVO C0, T2; MOVO D1, T3 + MOVQ $10, itr2 + +sealSSE128InnerCipherLoop: + chachaQR(A0, B0, C0, D0, T0); chachaQR(A1, B1, C1, D1, T0); chachaQR(A2, B2, C2, D2, T0) + shiftB0Left; shiftB1Left; shiftB2Left + shiftC0Left; shiftC1Left; shiftC2Left + shiftD0Left; shiftD1Left; shiftD2Left + chachaQR(A0, B0, C0, D0, T0); chachaQR(A1, B1, C1, D1, T0); chachaQR(A2, B2, C2, D2, T0) + shiftB0Right; shiftB1Right; shiftB2Right + shiftC0Right; shiftC1Right; shiftC2Right + shiftD0Right; shiftD1Right; shiftD2Right + DECQ itr2 + JNE sealSSE128InnerCipherLoop + + // A0|B0 hold the Poly1305 32-byte key, C0,D0 can be discarded + PADDL ·chacha20Constants<>(SB), A0; PADDL ·chacha20Constants<>(SB), A1; PADDL ·chacha20Constants<>(SB), A2 + PADDL T1, B0; PADDL T1, B1; PADDL T1, B2 + PADDL T2, C1; PADDL T2, C2 + PADDL T3, D1; PADDL ·sseIncMask<>(SB), T3; PADDL T3, D2 + PAND ·polyClampMask<>(SB), A0 + MOVOU A0, rStore + MOVOU B0, sStore + + // Hash + MOVQ ad_len+80(FP), itr2 + CALL polyHashADInternal<>(SB) + XORQ itr1, itr1 + +sealSSE128SealHash: + // itr1 holds the number of bytes encrypted but not yet hashed + CMPQ itr1, $16 + JB sealSSE128Seal + polyAdd(0(oup)) + polyMul + + SUBQ $16, itr1 + ADDQ $16, oup + + JMP sealSSE128SealHash + +sealSSE128Seal: + CMPQ inl, $16 + JB sealSSETail + SUBQ $16, inl + + // Load for decryption + MOVOU (inp), T0 + PXOR T0, A1 + MOVOU A1, (oup) + LEAQ (1*16)(inp), inp + LEAQ (1*16)(oup), oup + + // Extract for hashing + MOVQ A1, t0 + PSRLDQ $8, A1 + MOVQ A1, t1 + ADDQ t0, acc0; ADCQ t1, acc1; ADCQ $1, acc2 + polyMul + + // Shift the stream "left" + MOVO B1, A1 + MOVO C1, B1 + MOVO D1, C1 + MOVO A2, D1 + MOVO B2, A2 + MOVO C2, B2 + MOVO D2, C2 + JMP sealSSE128Seal + +sealSSETail: + TESTQ inl, inl + JE sealSSEFinalize + + // We can only load the PT one byte at a time to avoid read after end of buffer + MOVQ inl, itr2 + SHLQ $4, itr2 + LEAQ ·andMask<>(SB), t0 + MOVQ inl, itr1 + LEAQ -1(inp)(inl*1), inp + XORQ t2, t2 + XORQ t3, t3 + XORQ AX, AX + +sealSSETailLoadLoop: + SHLQ $8, t2, t3 + SHLQ $8, t2 + MOVB (inp), AX + XORQ AX, t2 + LEAQ -1(inp), inp + DECQ itr1 + JNE sealSSETailLoadLoop + MOVQ t2, 0+tmpStore + MOVQ t3, 8+tmpStore + PXOR 0+tmpStore, A1 + MOVOU A1, (oup) + MOVOU -16(t0)(itr2*1), T0 + PAND T0, A1 + MOVQ A1, t0 + PSRLDQ $8, A1 + MOVQ A1, t1 + ADDQ t0, acc0; ADCQ t1, acc1; ADCQ $1, acc2 + polyMul + + ADDQ inl, oup + +sealSSEFinalize: + // Hash in the buffer lengths + ADDQ ad_len+80(FP), acc0 + ADCQ src_len+56(FP), acc1 + ADCQ $1, acc2 + polyMul + + // Final reduce + MOVQ acc0, t0 + MOVQ acc1, t1 + MOVQ acc2, t2 + SUBQ $-5, acc0 + SBBQ $-1, acc1 + SBBQ $3, acc2 + CMOVQCS t0, acc0 + CMOVQCS t1, acc1 + CMOVQCS t2, acc2 + + // Add in the "s" part of the key + ADDQ 0+sStore, acc0 + ADCQ 8+sStore, acc1 + + // Finally store the tag at the end of the message + MOVQ acc0, (0*8)(oup) + MOVQ acc1, (1*8)(oup) + RET + +// ---------------------------------------------------------------------------- +// ------------------------- AVX2 Code ---------------------------------------- +chacha20Poly1305Seal_AVX2: + VZEROUPPER + VMOVDQU ·chacha20Constants<>(SB), AA0 + BYTE $0xc4; BYTE $0x42; BYTE $0x7d; BYTE $0x5a; BYTE $0x70; BYTE $0x10 // broadcasti128 16(r8), ymm14 + BYTE $0xc4; BYTE $0x42; BYTE $0x7d; BYTE $0x5a; BYTE $0x60; BYTE $0x20 // broadcasti128 32(r8), ymm12 + BYTE $0xc4; BYTE $0xc2; BYTE $0x7d; BYTE $0x5a; BYTE $0x60; BYTE $0x30 // broadcasti128 48(r8), ymm4 + VPADDD ·avx2InitMask<>(SB), DD0, DD0 + + // Special optimizations, for very short buffers + CMPQ inl, $192 + JBE seal192AVX2 // 33% faster + CMPQ inl, $320 + JBE seal320AVX2 // 17% faster + + // For the general key prepare the key first - as a byproduct we have 64 bytes of cipher stream + VMOVDQA AA0, AA1; VMOVDQA AA0, AA2; VMOVDQA AA0, AA3 + VMOVDQA BB0, BB1; VMOVDQA BB0, BB2; VMOVDQA BB0, BB3; VMOVDQA BB0, state1StoreAVX2 + VMOVDQA CC0, CC1; VMOVDQA CC0, CC2; VMOVDQA CC0, CC3; VMOVDQA CC0, state2StoreAVX2 + VPADDD ·avx2IncMask<>(SB), DD0, DD1; VMOVDQA DD0, ctr0StoreAVX2 + VPADDD ·avx2IncMask<>(SB), DD1, DD2; VMOVDQA DD1, ctr1StoreAVX2 + VPADDD ·avx2IncMask<>(SB), DD2, DD3; VMOVDQA DD2, ctr2StoreAVX2 + VMOVDQA DD3, ctr3StoreAVX2 + MOVQ $10, itr2 + +sealAVX2IntroLoop: + VMOVDQA CC3, tmpStoreAVX2 + chachaQR_AVX2(AA0, BB0, CC0, DD0, CC3); chachaQR_AVX2(AA1, BB1, CC1, DD1, CC3); chachaQR_AVX2(AA2, BB2, CC2, DD2, CC3) + VMOVDQA tmpStoreAVX2, CC3 + VMOVDQA CC1, tmpStoreAVX2 + chachaQR_AVX2(AA3, BB3, CC3, DD3, CC1) + VMOVDQA tmpStoreAVX2, CC1 + + VPALIGNR $4, BB0, BB0, BB0; VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $12, DD0, DD0, DD0 + VPALIGNR $4, BB1, BB1, BB1; VPALIGNR $8, CC1, CC1, CC1; VPALIGNR $12, DD1, DD1, DD1 + VPALIGNR $4, BB2, BB2, BB2; VPALIGNR $8, CC2, CC2, CC2; VPALIGNR $12, DD2, DD2, DD2 + VPALIGNR $4, BB3, BB3, BB3; VPALIGNR $8, CC3, CC3, CC3; VPALIGNR $12, DD3, DD3, DD3 + + VMOVDQA CC3, tmpStoreAVX2 + chachaQR_AVX2(AA0, BB0, CC0, DD0, CC3); chachaQR_AVX2(AA1, BB1, CC1, DD1, CC3); chachaQR_AVX2(AA2, BB2, CC2, DD2, CC3) + VMOVDQA tmpStoreAVX2, CC3 + VMOVDQA CC1, tmpStoreAVX2 + chachaQR_AVX2(AA3, BB3, CC3, DD3, CC1) + VMOVDQA tmpStoreAVX2, CC1 + + VPALIGNR $12, BB0, BB0, BB0; VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $4, DD0, DD0, DD0 + VPALIGNR $12, BB1, BB1, BB1; VPALIGNR $8, CC1, CC1, CC1; VPALIGNR $4, DD1, DD1, DD1 + VPALIGNR $12, BB2, BB2, BB2; VPALIGNR $8, CC2, CC2, CC2; VPALIGNR $4, DD2, DD2, DD2 + VPALIGNR $12, BB3, BB3, BB3; VPALIGNR $8, CC3, CC3, CC3; VPALIGNR $4, DD3, DD3, DD3 + DECQ itr2 + JNE sealAVX2IntroLoop + + VPADDD ·chacha20Constants<>(SB), AA0, AA0; VPADDD ·chacha20Constants<>(SB), AA1, AA1; VPADDD ·chacha20Constants<>(SB), AA2, AA2; VPADDD ·chacha20Constants<>(SB), AA3, AA3 + VPADDD state1StoreAVX2, BB0, BB0; VPADDD state1StoreAVX2, BB1, BB1; VPADDD state1StoreAVX2, BB2, BB2; VPADDD state1StoreAVX2, BB3, BB3 + VPADDD state2StoreAVX2, CC0, CC0; VPADDD state2StoreAVX2, CC1, CC1; VPADDD state2StoreAVX2, CC2, CC2; VPADDD state2StoreAVX2, CC3, CC3 + VPADDD ctr0StoreAVX2, DD0, DD0; VPADDD ctr1StoreAVX2, DD1, DD1; VPADDD ctr2StoreAVX2, DD2, DD2; VPADDD ctr3StoreAVX2, DD3, DD3 + + VPERM2I128 $0x13, CC0, DD0, CC0 // Stream bytes 96 - 127 + VPERM2I128 $0x02, AA0, BB0, DD0 // The Poly1305 key + VPERM2I128 $0x13, AA0, BB0, AA0 // Stream bytes 64 - 95 + + // Clamp and store poly key + VPAND ·polyClampMask<>(SB), DD0, DD0 + VMOVDQA DD0, rsStoreAVX2 + + // Hash AD + MOVQ ad_len+80(FP), itr2 + CALL polyHashADInternal<>(SB) + + // Can store at least 320 bytes + VPXOR (0*32)(inp), AA0, AA0 + VPXOR (1*32)(inp), CC0, CC0 + VMOVDQU AA0, (0*32)(oup) + VMOVDQU CC0, (1*32)(oup) + + VPERM2I128 $0x02, AA1, BB1, AA0; VPERM2I128 $0x02, CC1, DD1, BB0; VPERM2I128 $0x13, AA1, BB1, CC0; VPERM2I128 $0x13, CC1, DD1, DD0 + VPXOR (2*32)(inp), AA0, AA0; VPXOR (3*32)(inp), BB0, BB0; VPXOR (4*32)(inp), CC0, CC0; VPXOR (5*32)(inp), DD0, DD0 + VMOVDQU AA0, (2*32)(oup); VMOVDQU BB0, (3*32)(oup); VMOVDQU CC0, (4*32)(oup); VMOVDQU DD0, (5*32)(oup) + VPERM2I128 $0x02, AA2, BB2, AA0; VPERM2I128 $0x02, CC2, DD2, BB0; VPERM2I128 $0x13, AA2, BB2, CC0; VPERM2I128 $0x13, CC2, DD2, DD0 + VPXOR (6*32)(inp), AA0, AA0; VPXOR (7*32)(inp), BB0, BB0; VPXOR (8*32)(inp), CC0, CC0; VPXOR (9*32)(inp), DD0, DD0 + VMOVDQU AA0, (6*32)(oup); VMOVDQU BB0, (7*32)(oup); VMOVDQU CC0, (8*32)(oup); VMOVDQU DD0, (9*32)(oup) + + MOVQ $320, itr1 + SUBQ $320, inl + LEAQ 320(inp), inp + + VPERM2I128 $0x02, AA3, BB3, AA0; VPERM2I128 $0x02, CC3, DD3, BB0; VPERM2I128 $0x13, AA3, BB3, CC0; VPERM2I128 $0x13, CC3, DD3, DD0 + CMPQ inl, $128 + JBE sealAVX2SealHash + + VPXOR (0*32)(inp), AA0, AA0; VPXOR (1*32)(inp), BB0, BB0; VPXOR (2*32)(inp), CC0, CC0; VPXOR (3*32)(inp), DD0, DD0 + VMOVDQU AA0, (10*32)(oup); VMOVDQU BB0, (11*32)(oup); VMOVDQU CC0, (12*32)(oup); VMOVDQU DD0, (13*32)(oup) + SUBQ $128, inl + LEAQ 128(inp), inp + + MOVQ $8, itr1 + MOVQ $2, itr2 + + CMPQ inl, $128 + JBE sealAVX2Tail128 + CMPQ inl, $256 + JBE sealAVX2Tail256 + CMPQ inl, $384 + JBE sealAVX2Tail384 + CMPQ inl, $512 + JBE sealAVX2Tail512 + + // We have 448 bytes to hash, but main loop hashes 512 bytes at a time - perform some rounds, before the main loop + VMOVDQA ·chacha20Constants<>(SB), AA0; VMOVDQA AA0, AA1; VMOVDQA AA0, AA2; VMOVDQA AA0, AA3 + VMOVDQA state1StoreAVX2, BB0; VMOVDQA BB0, BB1; VMOVDQA BB0, BB2; VMOVDQA BB0, BB3 + VMOVDQA state2StoreAVX2, CC0; VMOVDQA CC0, CC1; VMOVDQA CC0, CC2; VMOVDQA CC0, CC3 + VMOVDQA ctr3StoreAVX2, DD0 + VPADDD ·avx2IncMask<>(SB), DD0, DD0; VPADDD ·avx2IncMask<>(SB), DD0, DD1; VPADDD ·avx2IncMask<>(SB), DD1, DD2; VPADDD ·avx2IncMask<>(SB), DD2, DD3 + VMOVDQA DD0, ctr0StoreAVX2; VMOVDQA DD1, ctr1StoreAVX2; VMOVDQA DD2, ctr2StoreAVX2; VMOVDQA DD3, ctr3StoreAVX2 + + VMOVDQA CC3, tmpStoreAVX2 + chachaQR_AVX2(AA0, BB0, CC0, DD0, CC3); chachaQR_AVX2(AA1, BB1, CC1, DD1, CC3); chachaQR_AVX2(AA2, BB2, CC2, DD2, CC3) + VMOVDQA tmpStoreAVX2, CC3 + VMOVDQA CC1, tmpStoreAVX2 + chachaQR_AVX2(AA3, BB3, CC3, DD3, CC1) + VMOVDQA tmpStoreAVX2, CC1 + + VPALIGNR $4, BB0, BB0, BB0; VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $12, DD0, DD0, DD0 + VPALIGNR $4, BB1, BB1, BB1; VPALIGNR $8, CC1, CC1, CC1; VPALIGNR $12, DD1, DD1, DD1 + VPALIGNR $4, BB2, BB2, BB2; VPALIGNR $8, CC2, CC2, CC2; VPALIGNR $12, DD2, DD2, DD2 + VPALIGNR $4, BB3, BB3, BB3; VPALIGNR $8, CC3, CC3, CC3; VPALIGNR $12, DD3, DD3, DD3 + + VMOVDQA CC3, tmpStoreAVX2 + chachaQR_AVX2(AA0, BB0, CC0, DD0, CC3); chachaQR_AVX2(AA1, BB1, CC1, DD1, CC3); chachaQR_AVX2(AA2, BB2, CC2, DD2, CC3) + VMOVDQA tmpStoreAVX2, CC3 + VMOVDQA CC1, tmpStoreAVX2 + chachaQR_AVX2(AA3, BB3, CC3, DD3, CC1) + VMOVDQA tmpStoreAVX2, CC1 + + VPALIGNR $12, BB0, BB0, BB0; VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $4, DD0, DD0, DD0 + VPALIGNR $12, BB1, BB1, BB1; VPALIGNR $8, CC1, CC1, CC1; VPALIGNR $4, DD1, DD1, DD1 + VPALIGNR $12, BB2, BB2, BB2; VPALIGNR $8, CC2, CC2, CC2; VPALIGNR $4, DD2, DD2, DD2 + VPALIGNR $12, BB3, BB3, BB3; VPALIGNR $8, CC3, CC3, CC3; VPALIGNR $4, DD3, DD3, DD3 + VPADDD BB0, AA0, AA0; VPADDD BB1, AA1, AA1; VPADDD BB2, AA2, AA2; VPADDD BB3, AA3, AA3 + VPXOR AA0, DD0, DD0; VPXOR AA1, DD1, DD1; VPXOR AA2, DD2, DD2; VPXOR AA3, DD3, DD3 + VPSHUFB ·rol16<>(SB), DD0, DD0; VPSHUFB ·rol16<>(SB), DD1, DD1; VPSHUFB ·rol16<>(SB), DD2, DD2; VPSHUFB ·rol16<>(SB), DD3, DD3 + VPADDD DD0, CC0, CC0; VPADDD DD1, CC1, CC1; VPADDD DD2, CC2, CC2; VPADDD DD3, CC3, CC3 + VPXOR CC0, BB0, BB0; VPXOR CC1, BB1, BB1; VPXOR CC2, BB2, BB2; VPXOR CC3, BB3, BB3 + VMOVDQA CC3, tmpStoreAVX2 + VPSLLD $12, BB0, CC3; VPSRLD $20, BB0, BB0; VPXOR CC3, BB0, BB0 + VPSLLD $12, BB1, CC3; VPSRLD $20, BB1, BB1; VPXOR CC3, BB1, BB1 + VPSLLD $12, BB2, CC3; VPSRLD $20, BB2, BB2; VPXOR CC3, BB2, BB2 + VPSLLD $12, BB3, CC3; VPSRLD $20, BB3, BB3; VPXOR CC3, BB3, BB3 + VMOVDQA tmpStoreAVX2, CC3 + + SUBQ $16, oup // Adjust the pointer + MOVQ $9, itr1 + JMP sealAVX2InternalLoopStart + +sealAVX2MainLoop: + // Load state, increment counter blocks, store the incremented counters + VMOVDQU ·chacha20Constants<>(SB), AA0; VMOVDQA AA0, AA1; VMOVDQA AA0, AA2; VMOVDQA AA0, AA3 + VMOVDQA state1StoreAVX2, BB0; VMOVDQA BB0, BB1; VMOVDQA BB0, BB2; VMOVDQA BB0, BB3 + VMOVDQA state2StoreAVX2, CC0; VMOVDQA CC0, CC1; VMOVDQA CC0, CC2; VMOVDQA CC0, CC3 + VMOVDQA ctr3StoreAVX2, DD0; VPADDD ·avx2IncMask<>(SB), DD0, DD0; VPADDD ·avx2IncMask<>(SB), DD0, DD1; VPADDD ·avx2IncMask<>(SB), DD1, DD2; VPADDD ·avx2IncMask<>(SB), DD2, DD3 + VMOVDQA DD0, ctr0StoreAVX2; VMOVDQA DD1, ctr1StoreAVX2; VMOVDQA DD2, ctr2StoreAVX2; VMOVDQA DD3, ctr3StoreAVX2 + MOVQ $10, itr1 + +sealAVX2InternalLoop: + polyAdd(0*8(oup)) + VPADDD BB0, AA0, AA0; VPADDD BB1, AA1, AA1; VPADDD BB2, AA2, AA2; VPADDD BB3, AA3, AA3 + polyMulStage1_AVX2 + VPXOR AA0, DD0, DD0; VPXOR AA1, DD1, DD1; VPXOR AA2, DD2, DD2; VPXOR AA3, DD3, DD3 + VPSHUFB ·rol16<>(SB), DD0, DD0; VPSHUFB ·rol16<>(SB), DD1, DD1; VPSHUFB ·rol16<>(SB), DD2, DD2; VPSHUFB ·rol16<>(SB), DD3, DD3 + polyMulStage2_AVX2 + VPADDD DD0, CC0, CC0; VPADDD DD1, CC1, CC1; VPADDD DD2, CC2, CC2; VPADDD DD3, CC3, CC3 + VPXOR CC0, BB0, BB0; VPXOR CC1, BB1, BB1; VPXOR CC2, BB2, BB2; VPXOR CC3, BB3, BB3 + polyMulStage3_AVX2 + VMOVDQA CC3, tmpStoreAVX2 + VPSLLD $12, BB0, CC3; VPSRLD $20, BB0, BB0; VPXOR CC3, BB0, BB0 + VPSLLD $12, BB1, CC3; VPSRLD $20, BB1, BB1; VPXOR CC3, BB1, BB1 + VPSLLD $12, BB2, CC3; VPSRLD $20, BB2, BB2; VPXOR CC3, BB2, BB2 + VPSLLD $12, BB3, CC3; VPSRLD $20, BB3, BB3; VPXOR CC3, BB3, BB3 + VMOVDQA tmpStoreAVX2, CC3 + polyMulReduceStage + +sealAVX2InternalLoopStart: + VPADDD BB0, AA0, AA0; VPADDD BB1, AA1, AA1; VPADDD BB2, AA2, AA2; VPADDD BB3, AA3, AA3 + VPXOR AA0, DD0, DD0; VPXOR AA1, DD1, DD1; VPXOR AA2, DD2, DD2; VPXOR AA3, DD3, DD3 + VPSHUFB ·rol8<>(SB), DD0, DD0; VPSHUFB ·rol8<>(SB), DD1, DD1; VPSHUFB ·rol8<>(SB), DD2, DD2; VPSHUFB ·rol8<>(SB), DD3, DD3 + polyAdd(2*8(oup)) + VPADDD DD0, CC0, CC0; VPADDD DD1, CC1, CC1; VPADDD DD2, CC2, CC2; VPADDD DD3, CC3, CC3 + polyMulStage1_AVX2 + VPXOR CC0, BB0, BB0; VPXOR CC1, BB1, BB1; VPXOR CC2, BB2, BB2; VPXOR CC3, BB3, BB3 + VMOVDQA CC3, tmpStoreAVX2 + VPSLLD $7, BB0, CC3; VPSRLD $25, BB0, BB0; VPXOR CC3, BB0, BB0 + VPSLLD $7, BB1, CC3; VPSRLD $25, BB1, BB1; VPXOR CC3, BB1, BB1 + VPSLLD $7, BB2, CC3; VPSRLD $25, BB2, BB2; VPXOR CC3, BB2, BB2 + VPSLLD $7, BB3, CC3; VPSRLD $25, BB3, BB3; VPXOR CC3, BB3, BB3 + VMOVDQA tmpStoreAVX2, CC3 + polyMulStage2_AVX2 + VPALIGNR $4, BB0, BB0, BB0; VPALIGNR $4, BB1, BB1, BB1; VPALIGNR $4, BB2, BB2, BB2; VPALIGNR $4, BB3, BB3, BB3 + VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $8, CC1, CC1, CC1; VPALIGNR $8, CC2, CC2, CC2; VPALIGNR $8, CC3, CC3, CC3 + VPALIGNR $12, DD0, DD0, DD0; VPALIGNR $12, DD1, DD1, DD1; VPALIGNR $12, DD2, DD2, DD2; VPALIGNR $12, DD3, DD3, DD3 + VPADDD BB0, AA0, AA0; VPADDD BB1, AA1, AA1; VPADDD BB2, AA2, AA2; VPADDD BB3, AA3, AA3 + polyMulStage3_AVX2 + VPXOR AA0, DD0, DD0; VPXOR AA1, DD1, DD1; VPXOR AA2, DD2, DD2; VPXOR AA3, DD3, DD3 + VPSHUFB ·rol16<>(SB), DD0, DD0; VPSHUFB ·rol16<>(SB), DD1, DD1; VPSHUFB ·rol16<>(SB), DD2, DD2; VPSHUFB ·rol16<>(SB), DD3, DD3 + polyMulReduceStage + VPADDD DD0, CC0, CC0; VPADDD DD1, CC1, CC1; VPADDD DD2, CC2, CC2; VPADDD DD3, CC3, CC3 + VPXOR CC0, BB0, BB0; VPXOR CC1, BB1, BB1; VPXOR CC2, BB2, BB2; VPXOR CC3, BB3, BB3 + polyAdd(4*8(oup)) + LEAQ (6*8)(oup), oup + VMOVDQA CC3, tmpStoreAVX2 + VPSLLD $12, BB0, CC3; VPSRLD $20, BB0, BB0; VPXOR CC3, BB0, BB0 + VPSLLD $12, BB1, CC3; VPSRLD $20, BB1, BB1; VPXOR CC3, BB1, BB1 + VPSLLD $12, BB2, CC3; VPSRLD $20, BB2, BB2; VPXOR CC3, BB2, BB2 + VPSLLD $12, BB3, CC3; VPSRLD $20, BB3, BB3; VPXOR CC3, BB3, BB3 + VMOVDQA tmpStoreAVX2, CC3 + polyMulStage1_AVX2 + VPADDD BB0, AA0, AA0; VPADDD BB1, AA1, AA1; VPADDD BB2, AA2, AA2; VPADDD BB3, AA3, AA3 + VPXOR AA0, DD0, DD0; VPXOR AA1, DD1, DD1; VPXOR AA2, DD2, DD2; VPXOR AA3, DD3, DD3 + polyMulStage2_AVX2 + VPSHUFB ·rol8<>(SB), DD0, DD0; VPSHUFB ·rol8<>(SB), DD1, DD1; VPSHUFB ·rol8<>(SB), DD2, DD2; VPSHUFB ·rol8<>(SB), DD3, DD3 + VPADDD DD0, CC0, CC0; VPADDD DD1, CC1, CC1; VPADDD DD2, CC2, CC2; VPADDD DD3, CC3, CC3 + polyMulStage3_AVX2 + VPXOR CC0, BB0, BB0; VPXOR CC1, BB1, BB1; VPXOR CC2, BB2, BB2; VPXOR CC3, BB3, BB3 + VMOVDQA CC3, tmpStoreAVX2 + VPSLLD $7, BB0, CC3; VPSRLD $25, BB0, BB0; VPXOR CC3, BB0, BB0 + VPSLLD $7, BB1, CC3; VPSRLD $25, BB1, BB1; VPXOR CC3, BB1, BB1 + VPSLLD $7, BB2, CC3; VPSRLD $25, BB2, BB2; VPXOR CC3, BB2, BB2 + VPSLLD $7, BB3, CC3; VPSRLD $25, BB3, BB3; VPXOR CC3, BB3, BB3 + VMOVDQA tmpStoreAVX2, CC3 + polyMulReduceStage + VPALIGNR $12, BB0, BB0, BB0; VPALIGNR $12, BB1, BB1, BB1; VPALIGNR $12, BB2, BB2, BB2; VPALIGNR $12, BB3, BB3, BB3 + VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $8, CC1, CC1, CC1; VPALIGNR $8, CC2, CC2, CC2; VPALIGNR $8, CC3, CC3, CC3 + VPALIGNR $4, DD0, DD0, DD0; VPALIGNR $4, DD1, DD1, DD1; VPALIGNR $4, DD2, DD2, DD2; VPALIGNR $4, DD3, DD3, DD3 + DECQ itr1 + JNE sealAVX2InternalLoop + + VPADDD ·chacha20Constants<>(SB), AA0, AA0; VPADDD ·chacha20Constants<>(SB), AA1, AA1; VPADDD ·chacha20Constants<>(SB), AA2, AA2; VPADDD ·chacha20Constants<>(SB), AA3, AA3 + VPADDD state1StoreAVX2, BB0, BB0; VPADDD state1StoreAVX2, BB1, BB1; VPADDD state1StoreAVX2, BB2, BB2; VPADDD state1StoreAVX2, BB3, BB3 + VPADDD state2StoreAVX2, CC0, CC0; VPADDD state2StoreAVX2, CC1, CC1; VPADDD state2StoreAVX2, CC2, CC2; VPADDD state2StoreAVX2, CC3, CC3 + VPADDD ctr0StoreAVX2, DD0, DD0; VPADDD ctr1StoreAVX2, DD1, DD1; VPADDD ctr2StoreAVX2, DD2, DD2; VPADDD ctr3StoreAVX2, DD3, DD3 + VMOVDQA CC3, tmpStoreAVX2 + + // We only hashed 480 of the 512 bytes available - hash the remaining 32 here + polyAdd(0*8(oup)) + polyMulAVX2 + LEAQ (4*8)(oup), oup + VPERM2I128 $0x02, AA0, BB0, CC3; VPERM2I128 $0x13, AA0, BB0, BB0; VPERM2I128 $0x02, CC0, DD0, AA0; VPERM2I128 $0x13, CC0, DD0, CC0 + VPXOR (0*32)(inp), CC3, CC3; VPXOR (1*32)(inp), AA0, AA0; VPXOR (2*32)(inp), BB0, BB0; VPXOR (3*32)(inp), CC0, CC0 + VMOVDQU CC3, (0*32)(oup); VMOVDQU AA0, (1*32)(oup); VMOVDQU BB0, (2*32)(oup); VMOVDQU CC0, (3*32)(oup) + VPERM2I128 $0x02, AA1, BB1, AA0; VPERM2I128 $0x02, CC1, DD1, BB0; VPERM2I128 $0x13, AA1, BB1, CC0; VPERM2I128 $0x13, CC1, DD1, DD0 + VPXOR (4*32)(inp), AA0, AA0; VPXOR (5*32)(inp), BB0, BB0; VPXOR (6*32)(inp), CC0, CC0; VPXOR (7*32)(inp), DD0, DD0 + VMOVDQU AA0, (4*32)(oup); VMOVDQU BB0, (5*32)(oup); VMOVDQU CC0, (6*32)(oup); VMOVDQU DD0, (7*32)(oup) + + // and here + polyAdd(-2*8(oup)) + polyMulAVX2 + VPERM2I128 $0x02, AA2, BB2, AA0; VPERM2I128 $0x02, CC2, DD2, BB0; VPERM2I128 $0x13, AA2, BB2, CC0; VPERM2I128 $0x13, CC2, DD2, DD0 + VPXOR (8*32)(inp), AA0, AA0; VPXOR (9*32)(inp), BB0, BB0; VPXOR (10*32)(inp), CC0, CC0; VPXOR (11*32)(inp), DD0, DD0 + VMOVDQU AA0, (8*32)(oup); VMOVDQU BB0, (9*32)(oup); VMOVDQU CC0, (10*32)(oup); VMOVDQU DD0, (11*32)(oup) + VPERM2I128 $0x02, AA3, BB3, AA0; VPERM2I128 $0x02, tmpStoreAVX2, DD3, BB0; VPERM2I128 $0x13, AA3, BB3, CC0; VPERM2I128 $0x13, tmpStoreAVX2, DD3, DD0 + VPXOR (12*32)(inp), AA0, AA0; VPXOR (13*32)(inp), BB0, BB0; VPXOR (14*32)(inp), CC0, CC0; VPXOR (15*32)(inp), DD0, DD0 + VMOVDQU AA0, (12*32)(oup); VMOVDQU BB0, (13*32)(oup); VMOVDQU CC0, (14*32)(oup); VMOVDQU DD0, (15*32)(oup) + LEAQ (32*16)(inp), inp + SUBQ $(32*16), inl + CMPQ inl, $512 + JG sealAVX2MainLoop + + // Tail can only hash 480 bytes + polyAdd(0*8(oup)) + polyMulAVX2 + polyAdd(2*8(oup)) + polyMulAVX2 + LEAQ 32(oup), oup + + MOVQ $10, itr1 + MOVQ $0, itr2 + CMPQ inl, $128 + JBE sealAVX2Tail128 + CMPQ inl, $256 + JBE sealAVX2Tail256 + CMPQ inl, $384 + JBE sealAVX2Tail384 + JMP sealAVX2Tail512 + +// ---------------------------------------------------------------------------- +// Special optimization for buffers smaller than 193 bytes +seal192AVX2: + // For up to 192 bytes of ciphertext and 64 bytes for the poly key, we process four blocks + VMOVDQA AA0, AA1 + VMOVDQA BB0, BB1 + VMOVDQA CC0, CC1 + VPADDD ·avx2IncMask<>(SB), DD0, DD1 + VMOVDQA AA0, AA2 + VMOVDQA BB0, BB2 + VMOVDQA CC0, CC2 + VMOVDQA DD0, DD2 + VMOVDQA DD1, TT3 + MOVQ $10, itr2 + +sealAVX2192InnerCipherLoop: + chachaQR_AVX2(AA0, BB0, CC0, DD0, TT0); chachaQR_AVX2(AA1, BB1, CC1, DD1, TT0) + VPALIGNR $4, BB0, BB0, BB0; VPALIGNR $4, BB1, BB1, BB1 + VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $8, CC1, CC1, CC1 + VPALIGNR $12, DD0, DD0, DD0; VPALIGNR $12, DD1, DD1, DD1 + chachaQR_AVX2(AA0, BB0, CC0, DD0, TT0); chachaQR_AVX2(AA1, BB1, CC1, DD1, TT0) + VPALIGNR $12, BB0, BB0, BB0; VPALIGNR $12, BB1, BB1, BB1 + VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $8, CC1, CC1, CC1 + VPALIGNR $4, DD0, DD0, DD0; VPALIGNR $4, DD1, DD1, DD1 + DECQ itr2 + JNE sealAVX2192InnerCipherLoop + VPADDD AA2, AA0, AA0; VPADDD AA2, AA1, AA1 + VPADDD BB2, BB0, BB0; VPADDD BB2, BB1, BB1 + VPADDD CC2, CC0, CC0; VPADDD CC2, CC1, CC1 + VPADDD DD2, DD0, DD0; VPADDD TT3, DD1, DD1 + VPERM2I128 $0x02, AA0, BB0, TT0 + + // Clamp and store poly key + VPAND ·polyClampMask<>(SB), TT0, TT0 + VMOVDQA TT0, rsStoreAVX2 + + // Stream for up to 192 bytes + VPERM2I128 $0x13, AA0, BB0, AA0 + VPERM2I128 $0x13, CC0, DD0, BB0 + VPERM2I128 $0x02, AA1, BB1, CC0 + VPERM2I128 $0x02, CC1, DD1, DD0 + VPERM2I128 $0x13, AA1, BB1, AA1 + VPERM2I128 $0x13, CC1, DD1, BB1 + +sealAVX2ShortSeal: + // Hash aad + MOVQ ad_len+80(FP), itr2 + CALL polyHashADInternal<>(SB) + XORQ itr1, itr1 + +sealAVX2SealHash: + // itr1 holds the number of bytes encrypted but not yet hashed + CMPQ itr1, $16 + JB sealAVX2ShortSealLoop + polyAdd(0(oup)) + polyMul + SUBQ $16, itr1 + ADDQ $16, oup + JMP sealAVX2SealHash + +sealAVX2ShortSealLoop: + CMPQ inl, $32 + JB sealAVX2ShortTail32 + SUBQ $32, inl + + // Load for encryption + VPXOR (inp), AA0, AA0 + VMOVDQU AA0, (oup) + LEAQ (1*32)(inp), inp + + // Now can hash + polyAdd(0*8(oup)) + polyMulAVX2 + polyAdd(2*8(oup)) + polyMulAVX2 + LEAQ (1*32)(oup), oup + + // Shift stream left + VMOVDQA BB0, AA0 + VMOVDQA CC0, BB0 + VMOVDQA DD0, CC0 + VMOVDQA AA1, DD0 + VMOVDQA BB1, AA1 + VMOVDQA CC1, BB1 + VMOVDQA DD1, CC1 + VMOVDQA AA2, DD1 + VMOVDQA BB2, AA2 + JMP sealAVX2ShortSealLoop + +sealAVX2ShortTail32: + CMPQ inl, $16 + VMOVDQA A0, A1 + JB sealAVX2ShortDone + + SUBQ $16, inl + + // Load for encryption + VPXOR (inp), A0, T0 + VMOVDQU T0, (oup) + LEAQ (1*16)(inp), inp + + // Hash + polyAdd(0*8(oup)) + polyMulAVX2 + LEAQ (1*16)(oup), oup + VPERM2I128 $0x11, AA0, AA0, AA0 + VMOVDQA A0, A1 + +sealAVX2ShortDone: + VZEROUPPER + JMP sealSSETail + +// ---------------------------------------------------------------------------- +// Special optimization for buffers smaller than 321 bytes +seal320AVX2: + // For up to 320 bytes of ciphertext and 64 bytes for the poly key, we process six blocks + VMOVDQA AA0, AA1; VMOVDQA BB0, BB1; VMOVDQA CC0, CC1; VPADDD ·avx2IncMask<>(SB), DD0, DD1 + VMOVDQA AA0, AA2; VMOVDQA BB0, BB2; VMOVDQA CC0, CC2; VPADDD ·avx2IncMask<>(SB), DD1, DD2 + VMOVDQA BB0, TT1; VMOVDQA CC0, TT2; VMOVDQA DD0, TT3 + MOVQ $10, itr2 + +sealAVX2320InnerCipherLoop: + chachaQR_AVX2(AA0, BB0, CC0, DD0, TT0); chachaQR_AVX2(AA1, BB1, CC1, DD1, TT0); chachaQR_AVX2(AA2, BB2, CC2, DD2, TT0) + VPALIGNR $4, BB0, BB0, BB0; VPALIGNR $4, BB1, BB1, BB1; VPALIGNR $4, BB2, BB2, BB2 + VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $8, CC1, CC1, CC1; VPALIGNR $8, CC2, CC2, CC2 + VPALIGNR $12, DD0, DD0, DD0; VPALIGNR $12, DD1, DD1, DD1; VPALIGNR $12, DD2, DD2, DD2 + chachaQR_AVX2(AA0, BB0, CC0, DD0, TT0); chachaQR_AVX2(AA1, BB1, CC1, DD1, TT0); chachaQR_AVX2(AA2, BB2, CC2, DD2, TT0) + VPALIGNR $12, BB0, BB0, BB0; VPALIGNR $12, BB1, BB1, BB1; VPALIGNR $12, BB2, BB2, BB2 + VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $8, CC1, CC1, CC1; VPALIGNR $8, CC2, CC2, CC2 + VPALIGNR $4, DD0, DD0, DD0; VPALIGNR $4, DD1, DD1, DD1; VPALIGNR $4, DD2, DD2, DD2 + DECQ itr2 + JNE sealAVX2320InnerCipherLoop + + VMOVDQA ·chacha20Constants<>(SB), TT0 + VPADDD TT0, AA0, AA0; VPADDD TT0, AA1, AA1; VPADDD TT0, AA2, AA2 + VPADDD TT1, BB0, BB0; VPADDD TT1, BB1, BB1; VPADDD TT1, BB2, BB2 + VPADDD TT2, CC0, CC0; VPADDD TT2, CC1, CC1; VPADDD TT2, CC2, CC2 + VMOVDQA ·avx2IncMask<>(SB), TT0 + VPADDD TT3, DD0, DD0; VPADDD TT0, TT3, TT3 + VPADDD TT3, DD1, DD1; VPADDD TT0, TT3, TT3 + VPADDD TT3, DD2, DD2 + + // Clamp and store poly key + VPERM2I128 $0x02, AA0, BB0, TT0 + VPAND ·polyClampMask<>(SB), TT0, TT0 + VMOVDQA TT0, rsStoreAVX2 + + // Stream for up to 320 bytes + VPERM2I128 $0x13, AA0, BB0, AA0 + VPERM2I128 $0x13, CC0, DD0, BB0 + VPERM2I128 $0x02, AA1, BB1, CC0 + VPERM2I128 $0x02, CC1, DD1, DD0 + VPERM2I128 $0x13, AA1, BB1, AA1 + VPERM2I128 $0x13, CC1, DD1, BB1 + VPERM2I128 $0x02, AA2, BB2, CC1 + VPERM2I128 $0x02, CC2, DD2, DD1 + VPERM2I128 $0x13, AA2, BB2, AA2 + VPERM2I128 $0x13, CC2, DD2, BB2 + JMP sealAVX2ShortSeal + +// ---------------------------------------------------------------------------- +// Special optimization for the last 128 bytes of ciphertext +sealAVX2Tail128: + // Need to decrypt up to 128 bytes - prepare two blocks + // If we got here after the main loop - there are 512 encrypted bytes waiting to be hashed + // If we got here before the main loop - there are 448 encrpyred bytes waiting to be hashed + VMOVDQA ·chacha20Constants<>(SB), AA0 + VMOVDQA state1StoreAVX2, BB0 + VMOVDQA state2StoreAVX2, CC0 + VMOVDQA ctr3StoreAVX2, DD0 + VPADDD ·avx2IncMask<>(SB), DD0, DD0 + VMOVDQA DD0, DD1 + +sealAVX2Tail128LoopA: + polyAdd(0(oup)) + polyMul + LEAQ 16(oup), oup + +sealAVX2Tail128LoopB: + chachaQR_AVX2(AA0, BB0, CC0, DD0, TT0) + polyAdd(0(oup)) + polyMul + VPALIGNR $4, BB0, BB0, BB0 + VPALIGNR $8, CC0, CC0, CC0 + VPALIGNR $12, DD0, DD0, DD0 + chachaQR_AVX2(AA0, BB0, CC0, DD0, TT0) + polyAdd(16(oup)) + polyMul + LEAQ 32(oup), oup + VPALIGNR $12, BB0, BB0, BB0 + VPALIGNR $8, CC0, CC0, CC0 + VPALIGNR $4, DD0, DD0, DD0 + DECQ itr1 + JG sealAVX2Tail128LoopA + DECQ itr2 + JGE sealAVX2Tail128LoopB + + VPADDD ·chacha20Constants<>(SB), AA0, AA1 + VPADDD state1StoreAVX2, BB0, BB1 + VPADDD state2StoreAVX2, CC0, CC1 + VPADDD DD1, DD0, DD1 + + VPERM2I128 $0x02, AA1, BB1, AA0 + VPERM2I128 $0x02, CC1, DD1, BB0 + VPERM2I128 $0x13, AA1, BB1, CC0 + VPERM2I128 $0x13, CC1, DD1, DD0 + JMP sealAVX2ShortSealLoop + +// ---------------------------------------------------------------------------- +// Special optimization for the last 256 bytes of ciphertext +sealAVX2Tail256: + // Need to decrypt up to 256 bytes - prepare two blocks + // If we got here after the main loop - there are 512 encrypted bytes waiting to be hashed + // If we got here before the main loop - there are 448 encrpyred bytes waiting to be hashed + VMOVDQA ·chacha20Constants<>(SB), AA0; VMOVDQA ·chacha20Constants<>(SB), AA1 + VMOVDQA state1StoreAVX2, BB0; VMOVDQA state1StoreAVX2, BB1 + VMOVDQA state2StoreAVX2, CC0; VMOVDQA state2StoreAVX2, CC1 + VMOVDQA ctr3StoreAVX2, DD0 + VPADDD ·avx2IncMask<>(SB), DD0, DD0 + VPADDD ·avx2IncMask<>(SB), DD0, DD1 + VMOVDQA DD0, TT1 + VMOVDQA DD1, TT2 + +sealAVX2Tail256LoopA: + polyAdd(0(oup)) + polyMul + LEAQ 16(oup), oup + +sealAVX2Tail256LoopB: + chachaQR_AVX2(AA0, BB0, CC0, DD0, TT0); chachaQR_AVX2(AA1, BB1, CC1, DD1, TT0) + polyAdd(0(oup)) + polyMul + VPALIGNR $4, BB0, BB0, BB0; VPALIGNR $4, BB1, BB1, BB1 + VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $8, CC1, CC1, CC1 + VPALIGNR $12, DD0, DD0, DD0; VPALIGNR $12, DD1, DD1, DD1 + chachaQR_AVX2(AA0, BB0, CC0, DD0, TT0); chachaQR_AVX2(AA1, BB1, CC1, DD1, TT0) + polyAdd(16(oup)) + polyMul + LEAQ 32(oup), oup + VPALIGNR $12, BB0, BB0, BB0; VPALIGNR $12, BB1, BB1, BB1 + VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $8, CC1, CC1, CC1 + VPALIGNR $4, DD0, DD0, DD0; VPALIGNR $4, DD1, DD1, DD1 + DECQ itr1 + JG sealAVX2Tail256LoopA + DECQ itr2 + JGE sealAVX2Tail256LoopB + + VPADDD ·chacha20Constants<>(SB), AA0, AA0; VPADDD ·chacha20Constants<>(SB), AA1, AA1 + VPADDD state1StoreAVX2, BB0, BB0; VPADDD state1StoreAVX2, BB1, BB1 + VPADDD state2StoreAVX2, CC0, CC0; VPADDD state2StoreAVX2, CC1, CC1 + VPADDD TT1, DD0, DD0; VPADDD TT2, DD1, DD1 + VPERM2I128 $0x02, AA0, BB0, TT0 + VPERM2I128 $0x02, CC0, DD0, TT1 + VPERM2I128 $0x13, AA0, BB0, TT2 + VPERM2I128 $0x13, CC0, DD0, TT3 + VPXOR (0*32)(inp), TT0, TT0; VPXOR (1*32)(inp), TT1, TT1; VPXOR (2*32)(inp), TT2, TT2; VPXOR (3*32)(inp), TT3, TT3 + VMOVDQU TT0, (0*32)(oup); VMOVDQU TT1, (1*32)(oup); VMOVDQU TT2, (2*32)(oup); VMOVDQU TT3, (3*32)(oup) + MOVQ $128, itr1 + LEAQ 128(inp), inp + SUBQ $128, inl + VPERM2I128 $0x02, AA1, BB1, AA0 + VPERM2I128 $0x02, CC1, DD1, BB0 + VPERM2I128 $0x13, AA1, BB1, CC0 + VPERM2I128 $0x13, CC1, DD1, DD0 + + JMP sealAVX2SealHash + +// ---------------------------------------------------------------------------- +// Special optimization for the last 384 bytes of ciphertext +sealAVX2Tail384: + // Need to decrypt up to 384 bytes - prepare two blocks + // If we got here after the main loop - there are 512 encrypted bytes waiting to be hashed + // If we got here before the main loop - there are 448 encrpyred bytes waiting to be hashed + VMOVDQA ·chacha20Constants<>(SB), AA0; VMOVDQA AA0, AA1; VMOVDQA AA0, AA2 + VMOVDQA state1StoreAVX2, BB0; VMOVDQA BB0, BB1; VMOVDQA BB0, BB2 + VMOVDQA state2StoreAVX2, CC0; VMOVDQA CC0, CC1; VMOVDQA CC0, CC2 + VMOVDQA ctr3StoreAVX2, DD0 + VPADDD ·avx2IncMask<>(SB), DD0, DD0; VPADDD ·avx2IncMask<>(SB), DD0, DD1; VPADDD ·avx2IncMask<>(SB), DD1, DD2 + VMOVDQA DD0, TT1; VMOVDQA DD1, TT2; VMOVDQA DD2, TT3 + +sealAVX2Tail384LoopA: + polyAdd(0(oup)) + polyMul + LEAQ 16(oup), oup + +sealAVX2Tail384LoopB: + chachaQR_AVX2(AA0, BB0, CC0, DD0, TT0); chachaQR_AVX2(AA1, BB1, CC1, DD1, TT0); chachaQR_AVX2(AA2, BB2, CC2, DD2, TT0) + polyAdd(0(oup)) + polyMul + VPALIGNR $4, BB0, BB0, BB0; VPALIGNR $4, BB1, BB1, BB1; VPALIGNR $4, BB2, BB2, BB2 + VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $8, CC1, CC1, CC1; VPALIGNR $8, CC2, CC2, CC2 + VPALIGNR $12, DD0, DD0, DD0; VPALIGNR $12, DD1, DD1, DD1; VPALIGNR $12, DD2, DD2, DD2 + chachaQR_AVX2(AA0, BB0, CC0, DD0, TT0); chachaQR_AVX2(AA1, BB1, CC1, DD1, TT0); chachaQR_AVX2(AA2, BB2, CC2, DD2, TT0) + polyAdd(16(oup)) + polyMul + LEAQ 32(oup), oup + VPALIGNR $12, BB0, BB0, BB0; VPALIGNR $12, BB1, BB1, BB1; VPALIGNR $12, BB2, BB2, BB2 + VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $8, CC1, CC1, CC1; VPALIGNR $8, CC2, CC2, CC2 + VPALIGNR $4, DD0, DD0, DD0; VPALIGNR $4, DD1, DD1, DD1; VPALIGNR $4, DD2, DD2, DD2 + DECQ itr1 + JG sealAVX2Tail384LoopA + DECQ itr2 + JGE sealAVX2Tail384LoopB + + VPADDD ·chacha20Constants<>(SB), AA0, AA0; VPADDD ·chacha20Constants<>(SB), AA1, AA1; VPADDD ·chacha20Constants<>(SB), AA2, AA2 + VPADDD state1StoreAVX2, BB0, BB0; VPADDD state1StoreAVX2, BB1, BB1; VPADDD state1StoreAVX2, BB2, BB2 + VPADDD state2StoreAVX2, CC0, CC0; VPADDD state2StoreAVX2, CC1, CC1; VPADDD state2StoreAVX2, CC2, CC2 + VPADDD TT1, DD0, DD0; VPADDD TT2, DD1, DD1; VPADDD TT3, DD2, DD2 + VPERM2I128 $0x02, AA0, BB0, TT0 + VPERM2I128 $0x02, CC0, DD0, TT1 + VPERM2I128 $0x13, AA0, BB0, TT2 + VPERM2I128 $0x13, CC0, DD0, TT3 + VPXOR (0*32)(inp), TT0, TT0; VPXOR (1*32)(inp), TT1, TT1; VPXOR (2*32)(inp), TT2, TT2; VPXOR (3*32)(inp), TT3, TT3 + VMOVDQU TT0, (0*32)(oup); VMOVDQU TT1, (1*32)(oup); VMOVDQU TT2, (2*32)(oup); VMOVDQU TT3, (3*32)(oup) + VPERM2I128 $0x02, AA1, BB1, TT0 + VPERM2I128 $0x02, CC1, DD1, TT1 + VPERM2I128 $0x13, AA1, BB1, TT2 + VPERM2I128 $0x13, CC1, DD1, TT3 + VPXOR (4*32)(inp), TT0, TT0; VPXOR (5*32)(inp), TT1, TT1; VPXOR (6*32)(inp), TT2, TT2; VPXOR (7*32)(inp), TT3, TT3 + VMOVDQU TT0, (4*32)(oup); VMOVDQU TT1, (5*32)(oup); VMOVDQU TT2, (6*32)(oup); VMOVDQU TT3, (7*32)(oup) + MOVQ $256, itr1 + LEAQ 256(inp), inp + SUBQ $256, inl + VPERM2I128 $0x02, AA2, BB2, AA0 + VPERM2I128 $0x02, CC2, DD2, BB0 + VPERM2I128 $0x13, AA2, BB2, CC0 + VPERM2I128 $0x13, CC2, DD2, DD0 + + JMP sealAVX2SealHash + +// ---------------------------------------------------------------------------- +// Special optimization for the last 512 bytes of ciphertext +sealAVX2Tail512: + // Need to decrypt up to 512 bytes - prepare two blocks + // If we got here after the main loop - there are 512 encrypted bytes waiting to be hashed + // If we got here before the main loop - there are 448 encrpyred bytes waiting to be hashed + VMOVDQA ·chacha20Constants<>(SB), AA0; VMOVDQA AA0, AA1; VMOVDQA AA0, AA2; VMOVDQA AA0, AA3 + VMOVDQA state1StoreAVX2, BB0; VMOVDQA BB0, BB1; VMOVDQA BB0, BB2; VMOVDQA BB0, BB3 + VMOVDQA state2StoreAVX2, CC0; VMOVDQA CC0, CC1; VMOVDQA CC0, CC2; VMOVDQA CC0, CC3 + VMOVDQA ctr3StoreAVX2, DD0 + VPADDD ·avx2IncMask<>(SB), DD0, DD0; VPADDD ·avx2IncMask<>(SB), DD0, DD1; VPADDD ·avx2IncMask<>(SB), DD1, DD2; VPADDD ·avx2IncMask<>(SB), DD2, DD3 + VMOVDQA DD0, ctr0StoreAVX2; VMOVDQA DD1, ctr1StoreAVX2; VMOVDQA DD2, ctr2StoreAVX2; VMOVDQA DD3, ctr3StoreAVX2 + +sealAVX2Tail512LoopA: + polyAdd(0(oup)) + polyMul + LEAQ 16(oup), oup + +sealAVX2Tail512LoopB: + VPADDD BB0, AA0, AA0; VPADDD BB1, AA1, AA1; VPADDD BB2, AA2, AA2; VPADDD BB3, AA3, AA3 + VPXOR AA0, DD0, DD0; VPXOR AA1, DD1, DD1; VPXOR AA2, DD2, DD2; VPXOR AA3, DD3, DD3 + VPSHUFB ·rol16<>(SB), DD0, DD0; VPSHUFB ·rol16<>(SB), DD1, DD1; VPSHUFB ·rol16<>(SB), DD2, DD2; VPSHUFB ·rol16<>(SB), DD3, DD3 + VPADDD DD0, CC0, CC0; VPADDD DD1, CC1, CC1; VPADDD DD2, CC2, CC2; VPADDD DD3, CC3, CC3 + VPXOR CC0, BB0, BB0; VPXOR CC1, BB1, BB1; VPXOR CC2, BB2, BB2; VPXOR CC3, BB3, BB3 + VMOVDQA CC3, tmpStoreAVX2 + VPSLLD $12, BB0, CC3; VPSRLD $20, BB0, BB0; VPXOR CC3, BB0, BB0 + VPSLLD $12, BB1, CC3; VPSRLD $20, BB1, BB1; VPXOR CC3, BB1, BB1 + VPSLLD $12, BB2, CC3; VPSRLD $20, BB2, BB2; VPXOR CC3, BB2, BB2 + VPSLLD $12, BB3, CC3; VPSRLD $20, BB3, BB3; VPXOR CC3, BB3, BB3 + VMOVDQA tmpStoreAVX2, CC3 + polyAdd(0*8(oup)) + polyMulAVX2 + VPADDD BB0, AA0, AA0; VPADDD BB1, AA1, AA1; VPADDD BB2, AA2, AA2; VPADDD BB3, AA3, AA3 + VPXOR AA0, DD0, DD0; VPXOR AA1, DD1, DD1; VPXOR AA2, DD2, DD2; VPXOR AA3, DD3, DD3 + VPSHUFB ·rol8<>(SB), DD0, DD0; VPSHUFB ·rol8<>(SB), DD1, DD1; VPSHUFB ·rol8<>(SB), DD2, DD2; VPSHUFB ·rol8<>(SB), DD3, DD3 + VPADDD DD0, CC0, CC0; VPADDD DD1, CC1, CC1; VPADDD DD2, CC2, CC2; VPADDD DD3, CC3, CC3 + VPXOR CC0, BB0, BB0; VPXOR CC1, BB1, BB1; VPXOR CC2, BB2, BB2; VPXOR CC3, BB3, BB3 + VMOVDQA CC3, tmpStoreAVX2 + VPSLLD $7, BB0, CC3; VPSRLD $25, BB0, BB0; VPXOR CC3, BB0, BB0 + VPSLLD $7, BB1, CC3; VPSRLD $25, BB1, BB1; VPXOR CC3, BB1, BB1 + VPSLLD $7, BB2, CC3; VPSRLD $25, BB2, BB2; VPXOR CC3, BB2, BB2 + VPSLLD $7, BB3, CC3; VPSRLD $25, BB3, BB3; VPXOR CC3, BB3, BB3 + VMOVDQA tmpStoreAVX2, CC3 + VPALIGNR $4, BB0, BB0, BB0; VPALIGNR $4, BB1, BB1, BB1; VPALIGNR $4, BB2, BB2, BB2; VPALIGNR $4, BB3, BB3, BB3 + VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $8, CC1, CC1, CC1; VPALIGNR $8, CC2, CC2, CC2; VPALIGNR $8, CC3, CC3, CC3 + VPALIGNR $12, DD0, DD0, DD0; VPALIGNR $12, DD1, DD1, DD1; VPALIGNR $12, DD2, DD2, DD2; VPALIGNR $12, DD3, DD3, DD3 + VPADDD BB0, AA0, AA0; VPADDD BB1, AA1, AA1; VPADDD BB2, AA2, AA2; VPADDD BB3, AA3, AA3 + VPXOR AA0, DD0, DD0; VPXOR AA1, DD1, DD1; VPXOR AA2, DD2, DD2; VPXOR AA3, DD3, DD3 + VPSHUFB ·rol16<>(SB), DD0, DD0; VPSHUFB ·rol16<>(SB), DD1, DD1; VPSHUFB ·rol16<>(SB), DD2, DD2; VPSHUFB ·rol16<>(SB), DD3, DD3 + VPADDD DD0, CC0, CC0; VPADDD DD1, CC1, CC1; VPADDD DD2, CC2, CC2; VPADDD DD3, CC3, CC3 + VPXOR CC0, BB0, BB0; VPXOR CC1, BB1, BB1; VPXOR CC2, BB2, BB2; VPXOR CC3, BB3, BB3 + polyAdd(2*8(oup)) + polyMulAVX2 + LEAQ (4*8)(oup), oup + VMOVDQA CC3, tmpStoreAVX2 + VPSLLD $12, BB0, CC3; VPSRLD $20, BB0, BB0; VPXOR CC3, BB0, BB0 + VPSLLD $12, BB1, CC3; VPSRLD $20, BB1, BB1; VPXOR CC3, BB1, BB1 + VPSLLD $12, BB2, CC3; VPSRLD $20, BB2, BB2; VPXOR CC3, BB2, BB2 + VPSLLD $12, BB3, CC3; VPSRLD $20, BB3, BB3; VPXOR CC3, BB3, BB3 + VMOVDQA tmpStoreAVX2, CC3 + VPADDD BB0, AA0, AA0; VPADDD BB1, AA1, AA1; VPADDD BB2, AA2, AA2; VPADDD BB3, AA3, AA3 + VPXOR AA0, DD0, DD0; VPXOR AA1, DD1, DD1; VPXOR AA2, DD2, DD2; VPXOR AA3, DD3, DD3 + VPSHUFB ·rol8<>(SB), DD0, DD0; VPSHUFB ·rol8<>(SB), DD1, DD1; VPSHUFB ·rol8<>(SB), DD2, DD2; VPSHUFB ·rol8<>(SB), DD3, DD3 + VPADDD DD0, CC0, CC0; VPADDD DD1, CC1, CC1; VPADDD DD2, CC2, CC2; VPADDD DD3, CC3, CC3 + VPXOR CC0, BB0, BB0; VPXOR CC1, BB1, BB1; VPXOR CC2, BB2, BB2; VPXOR CC3, BB3, BB3 + VMOVDQA CC3, tmpStoreAVX2 + VPSLLD $7, BB0, CC3; VPSRLD $25, BB0, BB0; VPXOR CC3, BB0, BB0 + VPSLLD $7, BB1, CC3; VPSRLD $25, BB1, BB1; VPXOR CC3, BB1, BB1 + VPSLLD $7, BB2, CC3; VPSRLD $25, BB2, BB2; VPXOR CC3, BB2, BB2 + VPSLLD $7, BB3, CC3; VPSRLD $25, BB3, BB3; VPXOR CC3, BB3, BB3 + VMOVDQA tmpStoreAVX2, CC3 + VPALIGNR $12, BB0, BB0, BB0; VPALIGNR $12, BB1, BB1, BB1; VPALIGNR $12, BB2, BB2, BB2; VPALIGNR $12, BB3, BB3, BB3 + VPALIGNR $8, CC0, CC0, CC0; VPALIGNR $8, CC1, CC1, CC1; VPALIGNR $8, CC2, CC2, CC2; VPALIGNR $8, CC3, CC3, CC3 + VPALIGNR $4, DD0, DD0, DD0; VPALIGNR $4, DD1, DD1, DD1; VPALIGNR $4, DD2, DD2, DD2; VPALIGNR $4, DD3, DD3, DD3 + + DECQ itr1 + JG sealAVX2Tail512LoopA + DECQ itr2 + JGE sealAVX2Tail512LoopB + + VPADDD ·chacha20Constants<>(SB), AA0, AA0; VPADDD ·chacha20Constants<>(SB), AA1, AA1; VPADDD ·chacha20Constants<>(SB), AA2, AA2; VPADDD ·chacha20Constants<>(SB), AA3, AA3 + VPADDD state1StoreAVX2, BB0, BB0; VPADDD state1StoreAVX2, BB1, BB1; VPADDD state1StoreAVX2, BB2, BB2; VPADDD state1StoreAVX2, BB3, BB3 + VPADDD state2StoreAVX2, CC0, CC0; VPADDD state2StoreAVX2, CC1, CC1; VPADDD state2StoreAVX2, CC2, CC2; VPADDD state2StoreAVX2, CC3, CC3 + VPADDD ctr0StoreAVX2, DD0, DD0; VPADDD ctr1StoreAVX2, DD1, DD1; VPADDD ctr2StoreAVX2, DD2, DD2; VPADDD ctr3StoreAVX2, DD3, DD3 + VMOVDQA CC3, tmpStoreAVX2 + VPERM2I128 $0x02, AA0, BB0, CC3 + VPXOR (0*32)(inp), CC3, CC3 + VMOVDQU CC3, (0*32)(oup) + VPERM2I128 $0x02, CC0, DD0, CC3 + VPXOR (1*32)(inp), CC3, CC3 + VMOVDQU CC3, (1*32)(oup) + VPERM2I128 $0x13, AA0, BB0, CC3 + VPXOR (2*32)(inp), CC3, CC3 + VMOVDQU CC3, (2*32)(oup) + VPERM2I128 $0x13, CC0, DD0, CC3 + VPXOR (3*32)(inp), CC3, CC3 + VMOVDQU CC3, (3*32)(oup) + + VPERM2I128 $0x02, AA1, BB1, AA0 + VPERM2I128 $0x02, CC1, DD1, BB0 + VPERM2I128 $0x13, AA1, BB1, CC0 + VPERM2I128 $0x13, CC1, DD1, DD0 + VPXOR (4*32)(inp), AA0, AA0; VPXOR (5*32)(inp), BB0, BB0; VPXOR (6*32)(inp), CC0, CC0; VPXOR (7*32)(inp), DD0, DD0 + VMOVDQU AA0, (4*32)(oup); VMOVDQU BB0, (5*32)(oup); VMOVDQU CC0, (6*32)(oup); VMOVDQU DD0, (7*32)(oup) + + VPERM2I128 $0x02, AA2, BB2, AA0 + VPERM2I128 $0x02, CC2, DD2, BB0 + VPERM2I128 $0x13, AA2, BB2, CC0 + VPERM2I128 $0x13, CC2, DD2, DD0 + VPXOR (8*32)(inp), AA0, AA0; VPXOR (9*32)(inp), BB0, BB0; VPXOR (10*32)(inp), CC0, CC0; VPXOR (11*32)(inp), DD0, DD0 + VMOVDQU AA0, (8*32)(oup); VMOVDQU BB0, (9*32)(oup); VMOVDQU CC0, (10*32)(oup); VMOVDQU DD0, (11*32)(oup) + + MOVQ $384, itr1 + LEAQ 384(inp), inp + SUBQ $384, inl + VPERM2I128 $0x02, AA3, BB3, AA0 + VPERM2I128 $0x02, tmpStoreAVX2, DD3, BB0 + VPERM2I128 $0x13, AA3, BB3, CC0 + VPERM2I128 $0x13, tmpStoreAVX2, DD3, DD0 + + JMP sealAVX2SealHash diff --git a/vendor/golang.org/x/crypto/chacha20poly1305/chacha20poly1305_generic.go b/vendor/golang.org/x/crypto/chacha20poly1305/chacha20poly1305_generic.go new file mode 100644 index 0000000..c279712 --- /dev/null +++ b/vendor/golang.org/x/crypto/chacha20poly1305/chacha20poly1305_generic.go @@ -0,0 +1,81 @@ +// Copyright 2016 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package chacha20poly1305 + +import ( + "encoding/binary" + + "golang.org/x/crypto/internal/chacha20" + "golang.org/x/crypto/internal/subtle" + "golang.org/x/crypto/poly1305" +) + +func roundTo16(n int) int { + return 16 * ((n + 15) / 16) +} + +func (c *chacha20poly1305) sealGeneric(dst, nonce, plaintext, additionalData []byte) []byte { + ret, out := sliceForAppend(dst, len(plaintext)+poly1305.TagSize) + if subtle.InexactOverlap(out, plaintext) { + panic("chacha20poly1305: invalid buffer overlap") + } + + var polyKey [32]byte + s := chacha20.New(c.key, [3]uint32{ + binary.LittleEndian.Uint32(nonce[0:4]), + binary.LittleEndian.Uint32(nonce[4:8]), + binary.LittleEndian.Uint32(nonce[8:12]), + }) + s.XORKeyStream(polyKey[:], polyKey[:]) + s.Advance() // skip the next 32 bytes + s.XORKeyStream(out, plaintext) + + polyInput := make([]byte, roundTo16(len(additionalData))+roundTo16(len(plaintext))+8+8) + copy(polyInput, additionalData) + copy(polyInput[roundTo16(len(additionalData)):], out[:len(plaintext)]) + binary.LittleEndian.PutUint64(polyInput[len(polyInput)-16:], uint64(len(additionalData))) + binary.LittleEndian.PutUint64(polyInput[len(polyInput)-8:], uint64(len(plaintext))) + + var tag [poly1305.TagSize]byte + poly1305.Sum(&tag, polyInput, &polyKey) + copy(out[len(plaintext):], tag[:]) + + return ret +} + +func (c *chacha20poly1305) openGeneric(dst, nonce, ciphertext, additionalData []byte) ([]byte, error) { + var tag [poly1305.TagSize]byte + copy(tag[:], ciphertext[len(ciphertext)-16:]) + ciphertext = ciphertext[:len(ciphertext)-16] + + var polyKey [32]byte + s := chacha20.New(c.key, [3]uint32{ + binary.LittleEndian.Uint32(nonce[0:4]), + binary.LittleEndian.Uint32(nonce[4:8]), + binary.LittleEndian.Uint32(nonce[8:12]), + }) + s.XORKeyStream(polyKey[:], polyKey[:]) + s.Advance() // skip the next 32 bytes + + polyInput := make([]byte, roundTo16(len(additionalData))+roundTo16(len(ciphertext))+8+8) + copy(polyInput, additionalData) + copy(polyInput[roundTo16(len(additionalData)):], ciphertext) + binary.LittleEndian.PutUint64(polyInput[len(polyInput)-16:], uint64(len(additionalData))) + binary.LittleEndian.PutUint64(polyInput[len(polyInput)-8:], uint64(len(ciphertext))) + + ret, out := sliceForAppend(dst, len(ciphertext)) + if subtle.InexactOverlap(out, ciphertext) { + panic("chacha20poly1305: invalid buffer overlap") + } + if !poly1305.Verify(&tag, polyInput, &polyKey) { + for i := range out { + out[i] = 0 + } + return nil, errOpen + } + + s.XORKeyStream(out, ciphertext) + return ret, nil +} diff --git a/vendor/golang.org/x/crypto/chacha20poly1305/chacha20poly1305_noasm.go b/vendor/golang.org/x/crypto/chacha20poly1305/chacha20poly1305_noasm.go new file mode 100644 index 0000000..4c2eb70 --- /dev/null +++ b/vendor/golang.org/x/crypto/chacha20poly1305/chacha20poly1305_noasm.go @@ -0,0 +1,15 @@ +// Copyright 2016 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build !amd64 !go1.7 gccgo appengine + +package chacha20poly1305 + +func (c *chacha20poly1305) seal(dst, nonce, plaintext, additionalData []byte) []byte { + return c.sealGeneric(dst, nonce, plaintext, additionalData) +} + +func (c *chacha20poly1305) open(dst, nonce, ciphertext, additionalData []byte) ([]byte, error) { + return c.openGeneric(dst, nonce, ciphertext, additionalData) +} diff --git a/vendor/golang.org/x/crypto/chacha20poly1305/xchacha20poly1305.go b/vendor/golang.org/x/crypto/chacha20poly1305/xchacha20poly1305.go new file mode 100644 index 0000000..a02fa57 --- /dev/null +++ b/vendor/golang.org/x/crypto/chacha20poly1305/xchacha20poly1305.go @@ -0,0 +1,104 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package chacha20poly1305 + +import ( + "crypto/cipher" + "encoding/binary" + "errors" + + "golang.org/x/crypto/internal/chacha20" +) + +type xchacha20poly1305 struct { + key [8]uint32 +} + +// NewX returns a XChaCha20-Poly1305 AEAD that uses the given 256-bit key. +// +// XChaCha20-Poly1305 is a ChaCha20-Poly1305 variant that takes a longer nonce, +// suitable to be generated randomly without risk of collisions. It should be +// preferred when nonce uniqueness cannot be trivially ensured, or whenever +// nonces are randomly generated. +func NewX(key []byte) (cipher.AEAD, error) { + if len(key) != KeySize { + return nil, errors.New("chacha20poly1305: bad key length") + } + ret := new(xchacha20poly1305) + ret.key[0] = binary.LittleEndian.Uint32(key[0:4]) + ret.key[1] = binary.LittleEndian.Uint32(key[4:8]) + ret.key[2] = binary.LittleEndian.Uint32(key[8:12]) + ret.key[3] = binary.LittleEndian.Uint32(key[12:16]) + ret.key[4] = binary.LittleEndian.Uint32(key[16:20]) + ret.key[5] = binary.LittleEndian.Uint32(key[20:24]) + ret.key[6] = binary.LittleEndian.Uint32(key[24:28]) + ret.key[7] = binary.LittleEndian.Uint32(key[28:32]) + return ret, nil +} + +func (*xchacha20poly1305) NonceSize() int { + return NonceSizeX +} + +func (*xchacha20poly1305) Overhead() int { + return 16 +} + +func (x *xchacha20poly1305) Seal(dst, nonce, plaintext, additionalData []byte) []byte { + if len(nonce) != NonceSizeX { + panic("chacha20poly1305: bad nonce length passed to Seal") + } + + // XChaCha20-Poly1305 technically supports a 64-bit counter, so there is no + // size limit. However, since we reuse the ChaCha20-Poly1305 implementation, + // the second half of the counter is not available. This is unlikely to be + // an issue because the cipher.AEAD API requires the entire message to be in + // memory, and the counter overflows at 256 GB. + if uint64(len(plaintext)) > (1<<38)-64 { + panic("chacha20poly1305: plaintext too large") + } + + hNonce := [4]uint32{ + binary.LittleEndian.Uint32(nonce[0:4]), + binary.LittleEndian.Uint32(nonce[4:8]), + binary.LittleEndian.Uint32(nonce[8:12]), + binary.LittleEndian.Uint32(nonce[12:16]), + } + c := &chacha20poly1305{ + key: chacha20.HChaCha20(&x.key, &hNonce), + } + // The first 4 bytes of the final nonce are unused counter space. + cNonce := make([]byte, NonceSize) + copy(cNonce[4:12], nonce[16:24]) + + return c.seal(dst, cNonce[:], plaintext, additionalData) +} + +func (x *xchacha20poly1305) Open(dst, nonce, ciphertext, additionalData []byte) ([]byte, error) { + if len(nonce) != NonceSizeX { + panic("chacha20poly1305: bad nonce length passed to Open") + } + if len(ciphertext) < 16 { + return nil, errOpen + } + if uint64(len(ciphertext)) > (1<<38)-48 { + panic("chacha20poly1305: ciphertext too large") + } + + hNonce := [4]uint32{ + binary.LittleEndian.Uint32(nonce[0:4]), + binary.LittleEndian.Uint32(nonce[4:8]), + binary.LittleEndian.Uint32(nonce[8:12]), + binary.LittleEndian.Uint32(nonce[12:16]), + } + c := &chacha20poly1305{ + key: chacha20.HChaCha20(&x.key, &hNonce), + } + // The first 4 bytes of the final nonce are unused counter space. + cNonce := make([]byte, NonceSize) + copy(cNonce[4:12], nonce[16:24]) + + return c.open(dst, cNonce[:], ciphertext, additionalData) +} diff --git a/vendor/golang.org/x/crypto/internal/chacha20/asm_arm64.s b/vendor/golang.org/x/crypto/internal/chacha20/asm_arm64.s new file mode 100644 index 0000000..b3a16ef --- /dev/null +++ b/vendor/golang.org/x/crypto/internal/chacha20/asm_arm64.s @@ -0,0 +1,308 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build go1.11 +// +build !gccgo,!appengine + +#include "textflag.h" + +#define NUM_ROUNDS 10 + +// func xorKeyStreamVX(dst, src []byte, key *[8]uint32, nonce *[3]uint32, counter *uint32) +TEXT ·xorKeyStreamVX(SB), NOSPLIT, $0 + MOVD dst+0(FP), R1 + MOVD src+24(FP), R2 + MOVD src_len+32(FP), R3 + MOVD key+48(FP), R4 + MOVD nonce+56(FP), R6 + MOVD counter+64(FP), R7 + + MOVD $·constants(SB), R10 + MOVD $·incRotMatrix(SB), R11 + + MOVW (R7), R20 + + AND $~255, R3, R13 + ADD R2, R13, R12 // R12 for block end + AND $255, R3, R13 +loop: + MOVD $NUM_ROUNDS, R21 + VLD1 (R11), [V30.S4, V31.S4] + + // load contants + // VLD4R (R10), [V0.S4, V1.S4, V2.S4, V3.S4] + WORD $0x4D60E940 + + // load keys + // VLD4R 16(R4), [V4.S4, V5.S4, V6.S4, V7.S4] + WORD $0x4DFFE884 + // VLD4R 16(R4), [V8.S4, V9.S4, V10.S4, V11.S4] + WORD $0x4DFFE888 + SUB $32, R4 + + // load counter + nonce + // VLD1R (R7), [V12.S4] + WORD $0x4D40C8EC + + // VLD3R (R6), [V13.S4, V14.S4, V15.S4] + WORD $0x4D40E8CD + + // update counter + VADD V30.S4, V12.S4, V12.S4 + +chacha: + // V0..V3 += V4..V7 + // V12..V15 <<<= ((V12..V15 XOR V0..V3), 16) + VADD V0.S4, V4.S4, V0.S4 + VADD V1.S4, V5.S4, V1.S4 + VADD V2.S4, V6.S4, V2.S4 + VADD V3.S4, V7.S4, V3.S4 + VEOR V12.B16, V0.B16, V12.B16 + VEOR V13.B16, V1.B16, V13.B16 + VEOR V14.B16, V2.B16, V14.B16 + VEOR V15.B16, V3.B16, V15.B16 + VREV32 V12.H8, V12.H8 + VREV32 V13.H8, V13.H8 + VREV32 V14.H8, V14.H8 + VREV32 V15.H8, V15.H8 + // V8..V11 += V12..V15 + // V4..V7 <<<= ((V4..V7 XOR V8..V11), 12) + VADD V8.S4, V12.S4, V8.S4 + VADD V9.S4, V13.S4, V9.S4 + VADD V10.S4, V14.S4, V10.S4 + VADD V11.S4, V15.S4, V11.S4 + VEOR V8.B16, V4.B16, V16.B16 + VEOR V9.B16, V5.B16, V17.B16 + VEOR V10.B16, V6.B16, V18.B16 + VEOR V11.B16, V7.B16, V19.B16 + VSHL $12, V16.S4, V4.S4 + VSHL $12, V17.S4, V5.S4 + VSHL $12, V18.S4, V6.S4 + VSHL $12, V19.S4, V7.S4 + VSRI $20, V16.S4, V4.S4 + VSRI $20, V17.S4, V5.S4 + VSRI $20, V18.S4, V6.S4 + VSRI $20, V19.S4, V7.S4 + + // V0..V3 += V4..V7 + // V12..V15 <<<= ((V12..V15 XOR V0..V3), 8) + VADD V0.S4, V4.S4, V0.S4 + VADD V1.S4, V5.S4, V1.S4 + VADD V2.S4, V6.S4, V2.S4 + VADD V3.S4, V7.S4, V3.S4 + VEOR V12.B16, V0.B16, V12.B16 + VEOR V13.B16, V1.B16, V13.B16 + VEOR V14.B16, V2.B16, V14.B16 + VEOR V15.B16, V3.B16, V15.B16 + VTBL V31.B16, [V12.B16], V12.B16 + VTBL V31.B16, [V13.B16], V13.B16 + VTBL V31.B16, [V14.B16], V14.B16 + VTBL V31.B16, [V15.B16], V15.B16 + + // V8..V11 += V12..V15 + // V4..V7 <<<= ((V4..V7 XOR V8..V11), 7) + VADD V12.S4, V8.S4, V8.S4 + VADD V13.S4, V9.S4, V9.S4 + VADD V14.S4, V10.S4, V10.S4 + VADD V15.S4, V11.S4, V11.S4 + VEOR V8.B16, V4.B16, V16.B16 + VEOR V9.B16, V5.B16, V17.B16 + VEOR V10.B16, V6.B16, V18.B16 + VEOR V11.B16, V7.B16, V19.B16 + VSHL $7, V16.S4, V4.S4 + VSHL $7, V17.S4, V5.S4 + VSHL $7, V18.S4, V6.S4 + VSHL $7, V19.S4, V7.S4 + VSRI $25, V16.S4, V4.S4 + VSRI $25, V17.S4, V5.S4 + VSRI $25, V18.S4, V6.S4 + VSRI $25, V19.S4, V7.S4 + + // V0..V3 += V5..V7, V4 + // V15,V12-V14 <<<= ((V15,V12-V14 XOR V0..V3), 16) + VADD V0.S4, V5.S4, V0.S4 + VADD V1.S4, V6.S4, V1.S4 + VADD V2.S4, V7.S4, V2.S4 + VADD V3.S4, V4.S4, V3.S4 + VEOR V15.B16, V0.B16, V15.B16 + VEOR V12.B16, V1.B16, V12.B16 + VEOR V13.B16, V2.B16, V13.B16 + VEOR V14.B16, V3.B16, V14.B16 + VREV32 V12.H8, V12.H8 + VREV32 V13.H8, V13.H8 + VREV32 V14.H8, V14.H8 + VREV32 V15.H8, V15.H8 + + // V10 += V15; V5 <<<= ((V10 XOR V5), 12) + // ... + VADD V15.S4, V10.S4, V10.S4 + VADD V12.S4, V11.S4, V11.S4 + VADD V13.S4, V8.S4, V8.S4 + VADD V14.S4, V9.S4, V9.S4 + VEOR V10.B16, V5.B16, V16.B16 + VEOR V11.B16, V6.B16, V17.B16 + VEOR V8.B16, V7.B16, V18.B16 + VEOR V9.B16, V4.B16, V19.B16 + VSHL $12, V16.S4, V5.S4 + VSHL $12, V17.S4, V6.S4 + VSHL $12, V18.S4, V7.S4 + VSHL $12, V19.S4, V4.S4 + VSRI $20, V16.S4, V5.S4 + VSRI $20, V17.S4, V6.S4 + VSRI $20, V18.S4, V7.S4 + VSRI $20, V19.S4, V4.S4 + + // V0 += V5; V15 <<<= ((V0 XOR V15), 8) + // ... + VADD V5.S4, V0.S4, V0.S4 + VADD V6.S4, V1.S4, V1.S4 + VADD V7.S4, V2.S4, V2.S4 + VADD V4.S4, V3.S4, V3.S4 + VEOR V0.B16, V15.B16, V15.B16 + VEOR V1.B16, V12.B16, V12.B16 + VEOR V2.B16, V13.B16, V13.B16 + VEOR V3.B16, V14.B16, V14.B16 + VTBL V31.B16, [V12.B16], V12.B16 + VTBL V31.B16, [V13.B16], V13.B16 + VTBL V31.B16, [V14.B16], V14.B16 + VTBL V31.B16, [V15.B16], V15.B16 + + // V10 += V15; V5 <<<= ((V10 XOR V5), 7) + // ... + VADD V15.S4, V10.S4, V10.S4 + VADD V12.S4, V11.S4, V11.S4 + VADD V13.S4, V8.S4, V8.S4 + VADD V14.S4, V9.S4, V9.S4 + VEOR V10.B16, V5.B16, V16.B16 + VEOR V11.B16, V6.B16, V17.B16 + VEOR V8.B16, V7.B16, V18.B16 + VEOR V9.B16, V4.B16, V19.B16 + VSHL $7, V16.S4, V5.S4 + VSHL $7, V17.S4, V6.S4 + VSHL $7, V18.S4, V7.S4 + VSHL $7, V19.S4, V4.S4 + VSRI $25, V16.S4, V5.S4 + VSRI $25, V17.S4, V6.S4 + VSRI $25, V18.S4, V7.S4 + VSRI $25, V19.S4, V4.S4 + + SUB $1, R21 + CBNZ R21, chacha + + // VLD4R (R10), [V16.S4, V17.S4, V18.S4, V19.S4] + WORD $0x4D60E950 + + // VLD4R 16(R4), [V20.S4, V21.S4, V22.S4, V23.S4] + WORD $0x4DFFE894 + VADD V30.S4, V12.S4, V12.S4 + VADD V16.S4, V0.S4, V0.S4 + VADD V17.S4, V1.S4, V1.S4 + VADD V18.S4, V2.S4, V2.S4 + VADD V19.S4, V3.S4, V3.S4 + // VLD4R 16(R4), [V24.S4, V25.S4, V26.S4, V27.S4] + WORD $0x4DFFE898 + // restore R4 + SUB $32, R4 + + // load counter + nonce + // VLD1R (R7), [V28.S4] + WORD $0x4D40C8FC + // VLD3R (R6), [V29.S4, V30.S4, V31.S4] + WORD $0x4D40E8DD + + VADD V20.S4, V4.S4, V4.S4 + VADD V21.S4, V5.S4, V5.S4 + VADD V22.S4, V6.S4, V6.S4 + VADD V23.S4, V7.S4, V7.S4 + VADD V24.S4, V8.S4, V8.S4 + VADD V25.S4, V9.S4, V9.S4 + VADD V26.S4, V10.S4, V10.S4 + VADD V27.S4, V11.S4, V11.S4 + VADD V28.S4, V12.S4, V12.S4 + VADD V29.S4, V13.S4, V13.S4 + VADD V30.S4, V14.S4, V14.S4 + VADD V31.S4, V15.S4, V15.S4 + + VZIP1 V1.S4, V0.S4, V16.S4 + VZIP2 V1.S4, V0.S4, V17.S4 + VZIP1 V3.S4, V2.S4, V18.S4 + VZIP2 V3.S4, V2.S4, V19.S4 + VZIP1 V5.S4, V4.S4, V20.S4 + VZIP2 V5.S4, V4.S4, V21.S4 + VZIP1 V7.S4, V6.S4, V22.S4 + VZIP2 V7.S4, V6.S4, V23.S4 + VZIP1 V9.S4, V8.S4, V24.S4 + VZIP2 V9.S4, V8.S4, V25.S4 + VZIP1 V11.S4, V10.S4, V26.S4 + VZIP2 V11.S4, V10.S4, V27.S4 + VZIP1 V13.S4, V12.S4, V28.S4 + VZIP2 V13.S4, V12.S4, V29.S4 + VZIP1 V15.S4, V14.S4, V30.S4 + VZIP2 V15.S4, V14.S4, V31.S4 + VZIP1 V18.D2, V16.D2, V0.D2 + VZIP2 V18.D2, V16.D2, V4.D2 + VZIP1 V19.D2, V17.D2, V8.D2 + VZIP2 V19.D2, V17.D2, V12.D2 + VLD1.P 64(R2), [V16.B16, V17.B16, V18.B16, V19.B16] + + VZIP1 V22.D2, V20.D2, V1.D2 + VZIP2 V22.D2, V20.D2, V5.D2 + VZIP1 V23.D2, V21.D2, V9.D2 + VZIP2 V23.D2, V21.D2, V13.D2 + VLD1.P 64(R2), [V20.B16, V21.B16, V22.B16, V23.B16] + VZIP1 V26.D2, V24.D2, V2.D2 + VZIP2 V26.D2, V24.D2, V6.D2 + VZIP1 V27.D2, V25.D2, V10.D2 + VZIP2 V27.D2, V25.D2, V14.D2 + VLD1.P 64(R2), [V24.B16, V25.B16, V26.B16, V27.B16] + VZIP1 V30.D2, V28.D2, V3.D2 + VZIP2 V30.D2, V28.D2, V7.D2 + VZIP1 V31.D2, V29.D2, V11.D2 + VZIP2 V31.D2, V29.D2, V15.D2 + VLD1.P 64(R2), [V28.B16, V29.B16, V30.B16, V31.B16] + VEOR V0.B16, V16.B16, V16.B16 + VEOR V1.B16, V17.B16, V17.B16 + VEOR V2.B16, V18.B16, V18.B16 + VEOR V3.B16, V19.B16, V19.B16 + VST1.P [V16.B16, V17.B16, V18.B16, V19.B16], 64(R1) + VEOR V4.B16, V20.B16, V20.B16 + VEOR V5.B16, V21.B16, V21.B16 + VEOR V6.B16, V22.B16, V22.B16 + VEOR V7.B16, V23.B16, V23.B16 + VST1.P [V20.B16, V21.B16, V22.B16, V23.B16], 64(R1) + VEOR V8.B16, V24.B16, V24.B16 + VEOR V9.B16, V25.B16, V25.B16 + VEOR V10.B16, V26.B16, V26.B16 + VEOR V11.B16, V27.B16, V27.B16 + VST1.P [V24.B16, V25.B16, V26.B16, V27.B16], 64(R1) + VEOR V12.B16, V28.B16, V28.B16 + VEOR V13.B16, V29.B16, V29.B16 + VEOR V14.B16, V30.B16, V30.B16 + VEOR V15.B16, V31.B16, V31.B16 + VST1.P [V28.B16, V29.B16, V30.B16, V31.B16], 64(R1) + + ADD $4, R20 + MOVW R20, (R7) // update counter + + CMP R2, R12 + BGT loop + + RET + + +DATA ·constants+0x00(SB)/4, $0x61707865 +DATA ·constants+0x04(SB)/4, $0x3320646e +DATA ·constants+0x08(SB)/4, $0x79622d32 +DATA ·constants+0x0c(SB)/4, $0x6b206574 +GLOBL ·constants(SB), NOPTR|RODATA, $32 + +DATA ·incRotMatrix+0x00(SB)/4, $0x00000000 +DATA ·incRotMatrix+0x04(SB)/4, $0x00000001 +DATA ·incRotMatrix+0x08(SB)/4, $0x00000002 +DATA ·incRotMatrix+0x0c(SB)/4, $0x00000003 +DATA ·incRotMatrix+0x10(SB)/4, $0x02010003 +DATA ·incRotMatrix+0x14(SB)/4, $0x06050407 +DATA ·incRotMatrix+0x18(SB)/4, $0x0A09080B +DATA ·incRotMatrix+0x1c(SB)/4, $0x0E0D0C0F +GLOBL ·incRotMatrix(SB), NOPTR|RODATA, $32 diff --git a/vendor/golang.org/x/crypto/internal/chacha20/asm_ppc64le.s b/vendor/golang.org/x/crypto/internal/chacha20/asm_ppc64le.s new file mode 100644 index 0000000..8815cb0 --- /dev/null +++ b/vendor/golang.org/x/crypto/internal/chacha20/asm_ppc64le.s @@ -0,0 +1,668 @@ +// Copyright 2019 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// Based on CRYPTOGAMS code with the following comment: +// # ==================================================================== +// # Written by Andy Polyakov for the OpenSSL +// # project. The module is, however, dual licensed under OpenSSL and +// # CRYPTOGAMS licenses depending on where you obtain it. For further +// # details see http://www.openssl.org/~appro/cryptogams/. +// # ==================================================================== + +// Original code can be found at the link below: +// https://github.com/dot-asm/cryptogams/commit/a60f5b50ed908e91e5c39ca79126a4a876d5d8ff + +// There are some differences between CRYPTOGAMS code and this one. The round +// loop for "_int" isn't the same as the original. Some adjustments were +// necessary because there are less vector registers available. For example, some +// X variables (r12, r13, r14, and r15) share the same register used by the +// counter. The original code uses ctr to name the counter. Here we use CNT +// because golang uses CTR as the counter register name. + +// +build ppc64le,!gccgo,!appengine + +#include "textflag.h" + +#define OUT R3 +#define INP R4 +#define LEN R5 +#define KEY R6 +#define CNT R7 + +#define TEMP R8 + +#define X0 R11 +#define X1 R12 +#define X2 R14 +#define X3 R15 +#define X4 R16 +#define X5 R17 +#define X6 R18 +#define X7 R19 +#define X8 R20 +#define X9 R21 +#define X10 R22 +#define X11 R23 +#define X12 R24 +#define X13 R25 +#define X14 R26 +#define X15 R27 + +#define CON0 X0 +#define CON1 X1 +#define CON2 X2 +#define CON3 X3 + +#define KEY0 X4 +#define KEY1 X5 +#define KEY2 X6 +#define KEY3 X7 +#define KEY4 X8 +#define KEY5 X9 +#define KEY6 X10 +#define KEY7 X11 + +#define CNT0 X12 +#define CNT1 X13 +#define CNT2 X14 +#define CNT3 X15 + +#define TMP0 R9 +#define TMP1 R10 +#define TMP2 R28 +#define TMP3 R29 + +#define CONSTS R8 + +#define A0 V0 +#define B0 V1 +#define C0 V2 +#define D0 V3 +#define A1 V4 +#define B1 V5 +#define C1 V6 +#define D1 V7 +#define A2 V8 +#define B2 V9 +#define C2 V10 +#define D2 V11 +#define T0 V12 +#define T1 V13 +#define T2 V14 + +#define K0 V15 +#define K1 V16 +#define K2 V17 +#define K3 V18 +#define K4 V19 +#define K5 V20 + +#define FOUR V21 +#define SIXTEEN V22 +#define TWENTY4 V23 +#define TWENTY V24 +#define TWELVE V25 +#define TWENTY5 V26 +#define SEVEN V27 + +#define INPPERM V28 +#define OUTPERM V29 +#define OUTMASK V30 + +#define DD0 V31 +#define DD1 SEVEN +#define DD2 T0 +#define DD3 T1 +#define DD4 T2 + +DATA ·consts+0x00(SB)/8, $0x3320646e61707865 +DATA ·consts+0x08(SB)/8, $0x6b20657479622d32 +DATA ·consts+0x10(SB)/8, $0x0000000000000001 +DATA ·consts+0x18(SB)/8, $0x0000000000000000 +DATA ·consts+0x20(SB)/8, $0x0000000000000004 +DATA ·consts+0x28(SB)/8, $0x0000000000000000 +DATA ·consts+0x30(SB)/8, $0x0a0b08090e0f0c0d +DATA ·consts+0x38(SB)/8, $0x0203000106070405 +DATA ·consts+0x40(SB)/8, $0x090a0b080d0e0f0c +DATA ·consts+0x48(SB)/8, $0x0102030005060704 +GLOBL ·consts(SB), RODATA, $80 + +//func chaCha20_ctr32_vmx(out, inp *byte, len int, key *[32]byte, counter *[16]byte) +TEXT ·chaCha20_ctr32_vmx(SB),NOSPLIT|NOFRAME,$0 + // Load the arguments inside the registers + MOVD out+0(FP), OUT + MOVD inp+8(FP), INP + MOVD len+16(FP), LEN + MOVD key+24(FP), KEY + MOVD cnt+32(FP), CNT + + MOVD $·consts(SB), CONSTS // point to consts addr + + MOVD $16, X0 + MOVD $32, X1 + MOVD $48, X2 + MOVD $64, X3 + MOVD $31, X4 + MOVD $15, X5 + + // Load key + LVX (KEY)(R0), K1 + LVSR (KEY)(R0), T0 + LVX (KEY)(X0), K2 + LVX (KEY)(X4), DD0 + + // Load counter + LVX (CNT)(R0), K3 + LVSR (CNT)(R0), T1 + LVX (CNT)(X5), DD1 + + // Load constants + LVX (CONSTS)(R0), K0 + LVX (CONSTS)(X0), K5 + LVX (CONSTS)(X1), FOUR + LVX (CONSTS)(X2), SIXTEEN + LVX (CONSTS)(X3), TWENTY4 + + // Align key and counter + VPERM K2, K1, T0, K1 + VPERM DD0, K2, T0, K2 + VPERM DD1, K3, T1, K3 + + // Load counter to GPR + MOVWZ 0(CNT), CNT0 + MOVWZ 4(CNT), CNT1 + MOVWZ 8(CNT), CNT2 + MOVWZ 12(CNT), CNT3 + + // Adjust vectors for the initial state + VADDUWM K3, K5, K3 + VADDUWM K3, K5, K4 + VADDUWM K4, K5, K5 + + // Synthesized constants + VSPLTISW $-12, TWENTY + VSPLTISW $12, TWELVE + VSPLTISW $-7, TWENTY5 + + VXOR T0, T0, T0 + VSPLTISW $-1, OUTMASK + LVSR (INP)(R0), INPPERM + LVSL (OUT)(R0), OUTPERM + VPERM OUTMASK, T0, OUTPERM, OUTMASK + +loop_outer_vmx: + // Load constant + MOVD $0x61707865, CON0 + MOVD $0x3320646e, CON1 + MOVD $0x79622d32, CON2 + MOVD $0x6b206574, CON3 + + VOR K0, K0, A0 + VOR K0, K0, A1 + VOR K0, K0, A2 + VOR K1, K1, B0 + + MOVD $10, TEMP + + // Load key to GPR + MOVWZ 0(KEY), X4 + MOVWZ 4(KEY), X5 + MOVWZ 8(KEY), X6 + MOVWZ 12(KEY), X7 + VOR K1, K1, B1 + VOR K1, K1, B2 + MOVWZ 16(KEY), X8 + MOVWZ 0(CNT), X12 + MOVWZ 20(KEY), X9 + MOVWZ 4(CNT), X13 + VOR K2, K2, C0 + VOR K2, K2, C1 + MOVWZ 24(KEY), X10 + MOVWZ 8(CNT), X14 + VOR K2, K2, C2 + VOR K3, K3, D0 + MOVWZ 28(KEY), X11 + MOVWZ 12(CNT), X15 + VOR K4, K4, D1 + VOR K5, K5, D2 + + MOVD X4, TMP0 + MOVD X5, TMP1 + MOVD X6, TMP2 + MOVD X7, TMP3 + VSPLTISW $7, SEVEN + + MOVD TEMP, CTR + +loop_vmx: + // CRYPTOGAMS uses a macro to create a loop using perl. This isn't possible + // using assembly macros. Therefore, the macro expansion result was used + // in order to maintain the algorithm efficiency. + // This loop generates three keystream blocks using VMX instructions and, + // in parallel, one keystream block using scalar instructions. + ADD X4, X0, X0 + ADD X5, X1, X1 + VADDUWM A0, B0, A0 + VADDUWM A1, B1, A1 + ADD X6, X2, X2 + ADD X7, X3, X3 + VADDUWM A2, B2, A2 + VXOR D0, A0, D0 + XOR X0, X12, X12 + XOR X1, X13, X13 + VXOR D1, A1, D1 + VXOR D2, A2, D2 + XOR X2, X14, X14 + XOR X3, X15, X15 + VPERM D0, D0, SIXTEEN, D0 + VPERM D1, D1, SIXTEEN, D1 + ROTLW $16, X12, X12 + ROTLW $16, X13, X13 + VPERM D2, D2, SIXTEEN, D2 + VADDUWM C0, D0, C0 + ROTLW $16, X14, X14 + ROTLW $16, X15, X15 + VADDUWM C1, D1, C1 + VADDUWM C2, D2, C2 + ADD X12, X8, X8 + ADD X13, X9, X9 + VXOR B0, C0, T0 + VXOR B1, C1, T1 + ADD X14, X10, X10 + ADD X15, X11, X11 + VXOR B2, C2, T2 + VRLW T0, TWELVE, B0 + XOR X8, X4, X4 + XOR X9, X5, X5 + VRLW T1, TWELVE, B1 + VRLW T2, TWELVE, B2 + XOR X10, X6, X6 + XOR X11, X7, X7 + VADDUWM A0, B0, A0 + VADDUWM A1, B1, A1 + ROTLW $12, X4, X4 + ROTLW $12, X5, X5 + VADDUWM A2, B2, A2 + VXOR D0, A0, D0 + ROTLW $12, X6, X6 + ROTLW $12, X7, X7 + VXOR D1, A1, D1 + VXOR D2, A2, D2 + ADD X4, X0, X0 + ADD X5, X1, X1 + VPERM D0, D0, TWENTY4, D0 + VPERM D1, D1, TWENTY4, D1 + ADD X6, X2, X2 + ADD X7, X3, X3 + VPERM D2, D2, TWENTY4, D2 + VADDUWM C0, D0, C0 + XOR X0, X12, X12 + XOR X1, X13, X13 + VADDUWM C1, D1, C1 + VADDUWM C2, D2, C2 + XOR X2, X14, X14 + XOR X3, X15, X15 + VXOR B0, C0, T0 + VXOR B1, C1, T1 + ROTLW $8, X12, X12 + ROTLW $8, X13, X13 + VXOR B2, C2, T2 + VRLW T0, SEVEN, B0 + ROTLW $8, X14, X14 + ROTLW $8, X15, X15 + VRLW T1, SEVEN, B1 + VRLW T2, SEVEN, B2 + ADD X12, X8, X8 + ADD X13, X9, X9 + VSLDOI $8, C0, C0, C0 + VSLDOI $8, C1, C1, C1 + ADD X14, X10, X10 + ADD X15, X11, X11 + VSLDOI $8, C2, C2, C2 + VSLDOI $12, B0, B0, B0 + XOR X8, X4, X4 + XOR X9, X5, X5 + VSLDOI $12, B1, B1, B1 + VSLDOI $12, B2, B2, B2 + XOR X10, X6, X6 + XOR X11, X7, X7 + VSLDOI $4, D0, D0, D0 + VSLDOI $4, D1, D1, D1 + ROTLW $7, X4, X4 + ROTLW $7, X5, X5 + VSLDOI $4, D2, D2, D2 + VADDUWM A0, B0, A0 + ROTLW $7, X6, X6 + ROTLW $7, X7, X7 + VADDUWM A1, B1, A1 + VADDUWM A2, B2, A2 + ADD X5, X0, X0 + ADD X6, X1, X1 + VXOR D0, A0, D0 + VXOR D1, A1, D1 + ADD X7, X2, X2 + ADD X4, X3, X3 + VXOR D2, A2, D2 + VPERM D0, D0, SIXTEEN, D0 + XOR X0, X15, X15 + XOR X1, X12, X12 + VPERM D1, D1, SIXTEEN, D1 + VPERM D2, D2, SIXTEEN, D2 + XOR X2, X13, X13 + XOR X3, X14, X14 + VADDUWM C0, D0, C0 + VADDUWM C1, D1, C1 + ROTLW $16, X15, X15 + ROTLW $16, X12, X12 + VADDUWM C2, D2, C2 + VXOR B0, C0, T0 + ROTLW $16, X13, X13 + ROTLW $16, X14, X14 + VXOR B1, C1, T1 + VXOR B2, C2, T2 + ADD X15, X10, X10 + ADD X12, X11, X11 + VRLW T0, TWELVE, B0 + VRLW T1, TWELVE, B1 + ADD X13, X8, X8 + ADD X14, X9, X9 + VRLW T2, TWELVE, B2 + VADDUWM A0, B0, A0 + XOR X10, X5, X5 + XOR X11, X6, X6 + VADDUWM A1, B1, A1 + VADDUWM A2, B2, A2 + XOR X8, X7, X7 + XOR X9, X4, X4 + VXOR D0, A0, D0 + VXOR D1, A1, D1 + ROTLW $12, X5, X5 + ROTLW $12, X6, X6 + VXOR D2, A2, D2 + VPERM D0, D0, TWENTY4, D0 + ROTLW $12, X7, X7 + ROTLW $12, X4, X4 + VPERM D1, D1, TWENTY4, D1 + VPERM D2, D2, TWENTY4, D2 + ADD X5, X0, X0 + ADD X6, X1, X1 + VADDUWM C0, D0, C0 + VADDUWM C1, D1, C1 + ADD X7, X2, X2 + ADD X4, X3, X3 + VADDUWM C2, D2, C2 + VXOR B0, C0, T0 + XOR X0, X15, X15 + XOR X1, X12, X12 + VXOR B1, C1, T1 + VXOR B2, C2, T2 + XOR X2, X13, X13 + XOR X3, X14, X14 + VRLW T0, SEVEN, B0 + VRLW T1, SEVEN, B1 + ROTLW $8, X15, X15 + ROTLW $8, X12, X12 + VRLW T2, SEVEN, B2 + VSLDOI $8, C0, C0, C0 + ROTLW $8, X13, X13 + ROTLW $8, X14, X14 + VSLDOI $8, C1, C1, C1 + VSLDOI $8, C2, C2, C2 + ADD X15, X10, X10 + ADD X12, X11, X11 + VSLDOI $4, B0, B0, B0 + VSLDOI $4, B1, B1, B1 + ADD X13, X8, X8 + ADD X14, X9, X9 + VSLDOI $4, B2, B2, B2 + VSLDOI $12, D0, D0, D0 + XOR X10, X5, X5 + XOR X11, X6, X6 + VSLDOI $12, D1, D1, D1 + VSLDOI $12, D2, D2, D2 + XOR X8, X7, X7 + XOR X9, X4, X4 + ROTLW $7, X5, X5 + ROTLW $7, X6, X6 + ROTLW $7, X7, X7 + ROTLW $7, X4, X4 + BC 0x10, 0, loop_vmx + + SUB $256, LEN, LEN + + // Accumulate key block + ADD $0x61707865, X0, X0 + ADD $0x3320646e, X1, X1 + ADD $0x79622d32, X2, X2 + ADD $0x6b206574, X3, X3 + ADD TMP0, X4, X4 + ADD TMP1, X5, X5 + ADD TMP2, X6, X6 + ADD TMP3, X7, X7 + MOVWZ 16(KEY), TMP0 + MOVWZ 20(KEY), TMP1 + MOVWZ 24(KEY), TMP2 + MOVWZ 28(KEY), TMP3 + ADD TMP0, X8, X8 + ADD TMP1, X9, X9 + ADD TMP2, X10, X10 + ADD TMP3, X11, X11 + + MOVWZ 12(CNT), TMP0 + MOVWZ 8(CNT), TMP1 + MOVWZ 4(CNT), TMP2 + MOVWZ 0(CNT), TEMP + ADD TMP0, X15, X15 + ADD TMP1, X14, X14 + ADD TMP2, X13, X13 + ADD TEMP, X12, X12 + + // Accumulate key block + VADDUWM A0, K0, A0 + VADDUWM A1, K0, A1 + VADDUWM A2, K0, A2 + VADDUWM B0, K1, B0 + VADDUWM B1, K1, B1 + VADDUWM B2, K1, B2 + VADDUWM C0, K2, C0 + VADDUWM C1, K2, C1 + VADDUWM C2, K2, C2 + VADDUWM D0, K3, D0 + VADDUWM D1, K4, D1 + VADDUWM D2, K5, D2 + + // Increment counter + ADD $4, TEMP, TEMP + MOVW TEMP, 0(CNT) + + VADDUWM K3, FOUR, K3 + VADDUWM K4, FOUR, K4 + VADDUWM K5, FOUR, K5 + + // XOR the input slice (INP) with the keystream, which is stored in GPRs (X0-X3). + + // Load input (aligned or not) + MOVWZ 0(INP), TMP0 + MOVWZ 4(INP), TMP1 + MOVWZ 8(INP), TMP2 + MOVWZ 12(INP), TMP3 + + // XOR with input + XOR TMP0, X0, X0 + XOR TMP1, X1, X1 + XOR TMP2, X2, X2 + XOR TMP3, X3, X3 + MOVWZ 16(INP), TMP0 + MOVWZ 20(INP), TMP1 + MOVWZ 24(INP), TMP2 + MOVWZ 28(INP), TMP3 + XOR TMP0, X4, X4 + XOR TMP1, X5, X5 + XOR TMP2, X6, X6 + XOR TMP3, X7, X7 + MOVWZ 32(INP), TMP0 + MOVWZ 36(INP), TMP1 + MOVWZ 40(INP), TMP2 + MOVWZ 44(INP), TMP3 + XOR TMP0, X8, X8 + XOR TMP1, X9, X9 + XOR TMP2, X10, X10 + XOR TMP3, X11, X11 + MOVWZ 48(INP), TMP0 + MOVWZ 52(INP), TMP1 + MOVWZ 56(INP), TMP2 + MOVWZ 60(INP), TMP3 + XOR TMP0, X12, X12 + XOR TMP1, X13, X13 + XOR TMP2, X14, X14 + XOR TMP3, X15, X15 + + // Store output (aligned or not) + MOVW X0, 0(OUT) + MOVW X1, 4(OUT) + MOVW X2, 8(OUT) + MOVW X3, 12(OUT) + + ADD $64, INP, INP // INP points to the end of the slice for the alignment code below + + MOVW X4, 16(OUT) + MOVD $16, TMP0 + MOVW X5, 20(OUT) + MOVD $32, TMP1 + MOVW X6, 24(OUT) + MOVD $48, TMP2 + MOVW X7, 28(OUT) + MOVD $64, TMP3 + MOVW X8, 32(OUT) + MOVW X9, 36(OUT) + MOVW X10, 40(OUT) + MOVW X11, 44(OUT) + MOVW X12, 48(OUT) + MOVW X13, 52(OUT) + MOVW X14, 56(OUT) + MOVW X15, 60(OUT) + ADD $64, OUT, OUT + + // Load input + LVX (INP)(R0), DD0 + LVX (INP)(TMP0), DD1 + LVX (INP)(TMP1), DD2 + LVX (INP)(TMP2), DD3 + LVX (INP)(TMP3), DD4 + ADD $64, INP, INP + + VPERM DD1, DD0, INPPERM, DD0 // Align input + VPERM DD2, DD1, INPPERM, DD1 + VPERM DD3, DD2, INPPERM, DD2 + VPERM DD4, DD3, INPPERM, DD3 + VXOR A0, DD0, A0 // XOR with input + VXOR B0, DD1, B0 + LVX (INP)(TMP0), DD1 // Keep loading input + VXOR C0, DD2, C0 + LVX (INP)(TMP1), DD2 + VXOR D0, DD3, D0 + LVX (INP)(TMP2), DD3 + LVX (INP)(TMP3), DD0 + ADD $64, INP, INP + MOVD $63, TMP3 // 63 is not a typo + VPERM A0, A0, OUTPERM, A0 + VPERM B0, B0, OUTPERM, B0 + VPERM C0, C0, OUTPERM, C0 + VPERM D0, D0, OUTPERM, D0 + + VPERM DD1, DD4, INPPERM, DD4 // Align input + VPERM DD2, DD1, INPPERM, DD1 + VPERM DD3, DD2, INPPERM, DD2 + VPERM DD0, DD3, INPPERM, DD3 + VXOR A1, DD4, A1 + VXOR B1, DD1, B1 + LVX (INP)(TMP0), DD1 // Keep loading + VXOR C1, DD2, C1 + LVX (INP)(TMP1), DD2 + VXOR D1, DD3, D1 + LVX (INP)(TMP2), DD3 + + // Note that the LVX address is always rounded down to the nearest 16-byte + // boundary, and that it always points to at most 15 bytes beyond the end of + // the slice, so we cannot cross a page boundary. + LVX (INP)(TMP3), DD4 // Redundant in aligned case. + ADD $64, INP, INP + VPERM A1, A1, OUTPERM, A1 // Pre-misalign output + VPERM B1, B1, OUTPERM, B1 + VPERM C1, C1, OUTPERM, C1 + VPERM D1, D1, OUTPERM, D1 + + VPERM DD1, DD0, INPPERM, DD0 // Align Input + VPERM DD2, DD1, INPPERM, DD1 + VPERM DD3, DD2, INPPERM, DD2 + VPERM DD4, DD3, INPPERM, DD3 + VXOR A2, DD0, A2 + VXOR B2, DD1, B2 + VXOR C2, DD2, C2 + VXOR D2, DD3, D2 + VPERM A2, A2, OUTPERM, A2 + VPERM B2, B2, OUTPERM, B2 + VPERM C2, C2, OUTPERM, C2 + VPERM D2, D2, OUTPERM, D2 + + ANDCC $15, OUT, X1 // Is out aligned? + MOVD OUT, X0 + + VSEL A0, B0, OUTMASK, DD0 // Collect pre-misaligned output + VSEL B0, C0, OUTMASK, DD1 + VSEL C0, D0, OUTMASK, DD2 + VSEL D0, A1, OUTMASK, DD3 + VSEL A1, B1, OUTMASK, B0 + VSEL B1, C1, OUTMASK, C0 + VSEL C1, D1, OUTMASK, D0 + VSEL D1, A2, OUTMASK, A1 + VSEL A2, B2, OUTMASK, B1 + VSEL B2, C2, OUTMASK, C1 + VSEL C2, D2, OUTMASK, D1 + + STVX DD0, (OUT+TMP0) + STVX DD1, (OUT+TMP1) + STVX DD2, (OUT+TMP2) + ADD $64, OUT, OUT + STVX DD3, (OUT+R0) + STVX B0, (OUT+TMP0) + STVX C0, (OUT+TMP1) + STVX D0, (OUT+TMP2) + ADD $64, OUT, OUT + STVX A1, (OUT+R0) + STVX B1, (OUT+TMP0) + STVX C1, (OUT+TMP1) + STVX D1, (OUT+TMP2) + ADD $64, OUT, OUT + + BEQ aligned_vmx + + SUB X1, OUT, X2 // in misaligned case edges + MOVD $0, X3 // are written byte-by-byte + +unaligned_tail_vmx: + STVEBX D2, (X2+X3) + ADD $1, X3, X3 + CMPW X3, X1 + BNE unaligned_tail_vmx + SUB X1, X0, X2 + +unaligned_head_vmx: + STVEBX A0, (X2+X1) + CMPW X1, $15 + ADD $1, X1, X1 + BNE unaligned_head_vmx + + CMPU LEN, $255 // done with 256-byte block yet? + BGT loop_outer_vmx + + JMP done_vmx + +aligned_vmx: + STVX A0, (X0+R0) + CMPU LEN, $255 // done with 256-byte block yet? + BGT loop_outer_vmx + +done_vmx: + RET diff --git a/vendor/golang.org/x/crypto/internal/chacha20/chacha_arm64.go b/vendor/golang.org/x/crypto/internal/chacha20/chacha_arm64.go new file mode 100644 index 0000000..ad74e23 --- /dev/null +++ b/vendor/golang.org/x/crypto/internal/chacha20/chacha_arm64.go @@ -0,0 +1,31 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build go1.11 +// +build !gccgo + +package chacha20 + +const ( + haveAsm = true + bufSize = 256 +) + +//go:noescape +func xorKeyStreamVX(dst, src []byte, key *[8]uint32, nonce *[3]uint32, counter *uint32) + +func (c *Cipher) xorKeyStreamAsm(dst, src []byte) { + + if len(src) >= bufSize { + xorKeyStreamVX(dst, src, &c.key, &c.nonce, &c.counter) + } + + if len(src)%bufSize != 0 { + i := len(src) - len(src)%bufSize + c.buf = [bufSize]byte{} + copy(c.buf[:], src[i:]) + xorKeyStreamVX(c.buf[:], c.buf[:], &c.key, &c.nonce, &c.counter) + c.len = bufSize - copy(dst[i:], c.buf[:len(src)%bufSize]) + } +} diff --git a/vendor/golang.org/x/crypto/internal/chacha20/chacha_generic.go b/vendor/golang.org/x/crypto/internal/chacha20/chacha_generic.go new file mode 100644 index 0000000..6570847 --- /dev/null +++ b/vendor/golang.org/x/crypto/internal/chacha20/chacha_generic.go @@ -0,0 +1,264 @@ +// Copyright 2016 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// Package ChaCha20 implements the core ChaCha20 function as specified +// in https://tools.ietf.org/html/rfc7539#section-2.3. +package chacha20 + +import ( + "crypto/cipher" + "encoding/binary" + + "golang.org/x/crypto/internal/subtle" +) + +// assert that *Cipher implements cipher.Stream +var _ cipher.Stream = (*Cipher)(nil) + +// Cipher is a stateful instance of ChaCha20 using a particular key +// and nonce. A *Cipher implements the cipher.Stream interface. +type Cipher struct { + key [8]uint32 + counter uint32 // incremented after each block + nonce [3]uint32 + buf [bufSize]byte // buffer for unused keystream bytes + len int // number of unused keystream bytes at end of buf +} + +// New creates a new ChaCha20 stream cipher with the given key and nonce. +// The initial counter value is set to 0. +func New(key [8]uint32, nonce [3]uint32) *Cipher { + return &Cipher{key: key, nonce: nonce} +} + +// ChaCha20 constants spelling "expand 32-byte k" +const ( + j0 uint32 = 0x61707865 + j1 uint32 = 0x3320646e + j2 uint32 = 0x79622d32 + j3 uint32 = 0x6b206574 +) + +func quarterRound(a, b, c, d uint32) (uint32, uint32, uint32, uint32) { + a += b + d ^= a + d = (d << 16) | (d >> 16) + c += d + b ^= c + b = (b << 12) | (b >> 20) + a += b + d ^= a + d = (d << 8) | (d >> 24) + c += d + b ^= c + b = (b << 7) | (b >> 25) + return a, b, c, d +} + +// XORKeyStream XORs each byte in the given slice with a byte from the +// cipher's key stream. Dst and src must overlap entirely or not at all. +// +// If len(dst) < len(src), XORKeyStream will panic. It is acceptable +// to pass a dst bigger than src, and in that case, XORKeyStream will +// only update dst[:len(src)] and will not touch the rest of dst. +// +// Multiple calls to XORKeyStream behave as if the concatenation of +// the src buffers was passed in a single run. That is, Cipher +// maintains state and does not reset at each XORKeyStream call. +func (s *Cipher) XORKeyStream(dst, src []byte) { + if len(dst) < len(src) { + panic("chacha20: output smaller than input") + } + if subtle.InexactOverlap(dst[:len(src)], src) { + panic("chacha20: invalid buffer overlap") + } + + // xor src with buffered keystream first + if s.len != 0 { + buf := s.buf[len(s.buf)-s.len:] + if len(src) < len(buf) { + buf = buf[:len(src)] + } + td, ts := dst[:len(buf)], src[:len(buf)] // BCE hint + for i, b := range buf { + td[i] = ts[i] ^ b + } + s.len -= len(buf) + if s.len != 0 { + return + } + s.buf = [len(s.buf)]byte{} // zero the empty buffer + src = src[len(buf):] + dst = dst[len(buf):] + } + + if len(src) == 0 { + return + } + if haveAsm { + if uint64(len(src))+uint64(s.counter)*64 > (1<<38)-64 { + panic("chacha20: counter overflow") + } + s.xorKeyStreamAsm(dst, src) + return + } + + // set up a 64-byte buffer to pad out the final block if needed + // (hoisted out of the main loop to avoid spills) + rem := len(src) % 64 // length of final block + fin := len(src) - rem // index of final block + if rem > 0 { + copy(s.buf[len(s.buf)-64:], src[fin:]) + } + + // pre-calculate most of the first round + s1, s5, s9, s13 := quarterRound(j1, s.key[1], s.key[5], s.nonce[0]) + s2, s6, s10, s14 := quarterRound(j2, s.key[2], s.key[6], s.nonce[1]) + s3, s7, s11, s15 := quarterRound(j3, s.key[3], s.key[7], s.nonce[2]) + + n := len(src) + src, dst = src[:n:n], dst[:n:n] // BCE hint + for i := 0; i < n; i += 64 { + // calculate the remainder of the first round + s0, s4, s8, s12 := quarterRound(j0, s.key[0], s.key[4], s.counter) + + // execute the second round + x0, x5, x10, x15 := quarterRound(s0, s5, s10, s15) + x1, x6, x11, x12 := quarterRound(s1, s6, s11, s12) + x2, x7, x8, x13 := quarterRound(s2, s7, s8, s13) + x3, x4, x9, x14 := quarterRound(s3, s4, s9, s14) + + // execute the remaining 18 rounds + for i := 0; i < 9; i++ { + x0, x4, x8, x12 = quarterRound(x0, x4, x8, x12) + x1, x5, x9, x13 = quarterRound(x1, x5, x9, x13) + x2, x6, x10, x14 = quarterRound(x2, x6, x10, x14) + x3, x7, x11, x15 = quarterRound(x3, x7, x11, x15) + + x0, x5, x10, x15 = quarterRound(x0, x5, x10, x15) + x1, x6, x11, x12 = quarterRound(x1, x6, x11, x12) + x2, x7, x8, x13 = quarterRound(x2, x7, x8, x13) + x3, x4, x9, x14 = quarterRound(x3, x4, x9, x14) + } + + x0 += j0 + x1 += j1 + x2 += j2 + x3 += j3 + + x4 += s.key[0] + x5 += s.key[1] + x6 += s.key[2] + x7 += s.key[3] + x8 += s.key[4] + x9 += s.key[5] + x10 += s.key[6] + x11 += s.key[7] + + x12 += s.counter + x13 += s.nonce[0] + x14 += s.nonce[1] + x15 += s.nonce[2] + + // increment the counter + s.counter += 1 + if s.counter == 0 { + panic("chacha20: counter overflow") + } + + // pad to 64 bytes if needed + in, out := src[i:], dst[i:] + if i == fin { + // src[fin:] has already been copied into s.buf before + // the main loop + in, out = s.buf[len(s.buf)-64:], s.buf[len(s.buf)-64:] + } + in, out = in[:64], out[:64] // BCE hint + + // XOR the key stream with the source and write out the result + xor(out[0:], in[0:], x0) + xor(out[4:], in[4:], x1) + xor(out[8:], in[8:], x2) + xor(out[12:], in[12:], x3) + xor(out[16:], in[16:], x4) + xor(out[20:], in[20:], x5) + xor(out[24:], in[24:], x6) + xor(out[28:], in[28:], x7) + xor(out[32:], in[32:], x8) + xor(out[36:], in[36:], x9) + xor(out[40:], in[40:], x10) + xor(out[44:], in[44:], x11) + xor(out[48:], in[48:], x12) + xor(out[52:], in[52:], x13) + xor(out[56:], in[56:], x14) + xor(out[60:], in[60:], x15) + } + // copy any trailing bytes out of the buffer and into dst + if rem != 0 { + s.len = 64 - rem + copy(dst[fin:], s.buf[len(s.buf)-64:]) + } +} + +// Advance discards bytes in the key stream until the next 64 byte block +// boundary is reached and updates the counter accordingly. If the key +// stream is already at a block boundary no bytes will be discarded and +// the counter will be unchanged. +func (s *Cipher) Advance() { + s.len -= s.len % 64 + if s.len == 0 { + s.buf = [len(s.buf)]byte{} + } +} + +// XORKeyStream crypts bytes from in to out using the given key and counters. +// In and out must overlap entirely or not at all. Counter contains the raw +// ChaCha20 counter bytes (i.e. block counter followed by nonce). +func XORKeyStream(out, in []byte, counter *[16]byte, key *[32]byte) { + s := Cipher{ + key: [8]uint32{ + binary.LittleEndian.Uint32(key[0:4]), + binary.LittleEndian.Uint32(key[4:8]), + binary.LittleEndian.Uint32(key[8:12]), + binary.LittleEndian.Uint32(key[12:16]), + binary.LittleEndian.Uint32(key[16:20]), + binary.LittleEndian.Uint32(key[20:24]), + binary.LittleEndian.Uint32(key[24:28]), + binary.LittleEndian.Uint32(key[28:32]), + }, + nonce: [3]uint32{ + binary.LittleEndian.Uint32(counter[4:8]), + binary.LittleEndian.Uint32(counter[8:12]), + binary.LittleEndian.Uint32(counter[12:16]), + }, + counter: binary.LittleEndian.Uint32(counter[0:4]), + } + s.XORKeyStream(out, in) +} + +// HChaCha20 uses the ChaCha20 core to generate a derived key from a key and a +// nonce. It should only be used as part of the XChaCha20 construction. +func HChaCha20(key *[8]uint32, nonce *[4]uint32) [8]uint32 { + x0, x1, x2, x3 := j0, j1, j2, j3 + x4, x5, x6, x7 := key[0], key[1], key[2], key[3] + x8, x9, x10, x11 := key[4], key[5], key[6], key[7] + x12, x13, x14, x15 := nonce[0], nonce[1], nonce[2], nonce[3] + + for i := 0; i < 10; i++ { + x0, x4, x8, x12 = quarterRound(x0, x4, x8, x12) + x1, x5, x9, x13 = quarterRound(x1, x5, x9, x13) + x2, x6, x10, x14 = quarterRound(x2, x6, x10, x14) + x3, x7, x11, x15 = quarterRound(x3, x7, x11, x15) + + x0, x5, x10, x15 = quarterRound(x0, x5, x10, x15) + x1, x6, x11, x12 = quarterRound(x1, x6, x11, x12) + x2, x7, x8, x13 = quarterRound(x2, x7, x8, x13) + x3, x4, x9, x14 = quarterRound(x3, x4, x9, x14) + } + + var out [8]uint32 + out[0], out[1], out[2], out[3] = x0, x1, x2, x3 + out[4], out[5], out[6], out[7] = x12, x13, x14, x15 + return out +} diff --git a/vendor/golang.org/x/crypto/internal/chacha20/chacha_noasm.go b/vendor/golang.org/x/crypto/internal/chacha20/chacha_noasm.go new file mode 100644 index 0000000..bf8beba --- /dev/null +++ b/vendor/golang.org/x/crypto/internal/chacha20/chacha_noasm.go @@ -0,0 +1,16 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build !ppc64le,!arm64,!s390x arm64,!go1.11 gccgo appengine + +package chacha20 + +const ( + bufSize = 64 + haveAsm = false +) + +func (*Cipher) xorKeyStreamAsm(dst, src []byte) { + panic("not implemented") +} diff --git a/vendor/golang.org/x/crypto/internal/chacha20/chacha_ppc64le.go b/vendor/golang.org/x/crypto/internal/chacha20/chacha_ppc64le.go new file mode 100644 index 0000000..9bd6df6 --- /dev/null +++ b/vendor/golang.org/x/crypto/internal/chacha20/chacha_ppc64le.go @@ -0,0 +1,51 @@ +// Copyright 2019 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build ppc64le,!gccgo,!appengine + +package chacha20 + +import "encoding/binary" + +var haveAsm = true + +const bufSize = 256 + +//go:noescape +func chaCha20_ctr32_vmx(out, inp *byte, len int, key *[8]uint32, counter *uint32) + +func (c *Cipher) xorKeyStreamAsm(dst, src []byte) { + if len(src) >= bufSize { + chaCha20_ctr32_vmx(&dst[0], &src[0], len(src)-len(src)%bufSize, &c.key, &c.counter) + } + if len(src)%bufSize != 0 { + chaCha20_ctr32_vmx(&c.buf[0], &c.buf[0], bufSize, &c.key, &c.counter) + start := len(src) - len(src)%bufSize + ts, td, tb := src[start:], dst[start:], c.buf[:] + // Unroll loop to XOR 32 bytes per iteration. + for i := 0; i < len(ts)-32; i += 32 { + td, tb = td[:len(ts)], tb[:len(ts)] // bounds check elimination + s0 := binary.LittleEndian.Uint64(ts[0:8]) + s1 := binary.LittleEndian.Uint64(ts[8:16]) + s2 := binary.LittleEndian.Uint64(ts[16:24]) + s3 := binary.LittleEndian.Uint64(ts[24:32]) + b0 := binary.LittleEndian.Uint64(tb[0:8]) + b1 := binary.LittleEndian.Uint64(tb[8:16]) + b2 := binary.LittleEndian.Uint64(tb[16:24]) + b3 := binary.LittleEndian.Uint64(tb[24:32]) + binary.LittleEndian.PutUint64(td[0:8], s0^b0) + binary.LittleEndian.PutUint64(td[8:16], s1^b1) + binary.LittleEndian.PutUint64(td[16:24], s2^b2) + binary.LittleEndian.PutUint64(td[24:32], s3^b3) + ts, td, tb = ts[32:], td[32:], tb[32:] + } + td, tb = td[:len(ts)], tb[:len(ts)] // bounds check elimination + for i, v := range ts { + td[i] = tb[i] ^ v + } + c.len = bufSize - (len(src) % bufSize) + + } + +} diff --git a/vendor/golang.org/x/crypto/internal/chacha20/chacha_s390x.go b/vendor/golang.org/x/crypto/internal/chacha20/chacha_s390x.go new file mode 100644 index 0000000..aad645b --- /dev/null +++ b/vendor/golang.org/x/crypto/internal/chacha20/chacha_s390x.go @@ -0,0 +1,29 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build s390x,!gccgo,!appengine + +package chacha20 + +import ( + "golang.org/x/sys/cpu" +) + +var haveAsm = cpu.S390X.HasVX + +const bufSize = 256 + +// xorKeyStreamVX is an assembly implementation of XORKeyStream. It must only +// be called when the vector facility is available. +// Implementation in asm_s390x.s. +//go:noescape +func xorKeyStreamVX(dst, src []byte, key *[8]uint32, nonce *[3]uint32, counter *uint32, buf *[256]byte, len *int) + +func (c *Cipher) xorKeyStreamAsm(dst, src []byte) { + xorKeyStreamVX(dst, src, &c.key, &c.nonce, &c.counter, &c.buf, &c.len) +} + +// EXRL targets, DO NOT CALL! +func mvcSrcToBuf() +func mvcBufToDst() diff --git a/vendor/golang.org/x/crypto/internal/chacha20/chacha_s390x.s b/vendor/golang.org/x/crypto/internal/chacha20/chacha_s390x.s new file mode 100644 index 0000000..57df404 --- /dev/null +++ b/vendor/golang.org/x/crypto/internal/chacha20/chacha_s390x.s @@ -0,0 +1,260 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build s390x,!gccgo,!appengine + +#include "go_asm.h" +#include "textflag.h" + +// This is an implementation of the ChaCha20 encryption algorithm as +// specified in RFC 7539. It uses vector instructions to compute +// 4 keystream blocks in parallel (256 bytes) which are then XORed +// with the bytes in the input slice. + +GLOBL ·constants<>(SB), RODATA|NOPTR, $32 +// BSWAP: swap bytes in each 4-byte element +DATA ·constants<>+0x00(SB)/4, $0x03020100 +DATA ·constants<>+0x04(SB)/4, $0x07060504 +DATA ·constants<>+0x08(SB)/4, $0x0b0a0908 +DATA ·constants<>+0x0c(SB)/4, $0x0f0e0d0c +// J0: [j0, j1, j2, j3] +DATA ·constants<>+0x10(SB)/4, $0x61707865 +DATA ·constants<>+0x14(SB)/4, $0x3320646e +DATA ·constants<>+0x18(SB)/4, $0x79622d32 +DATA ·constants<>+0x1c(SB)/4, $0x6b206574 + +// EXRL targets: +TEXT ·mvcSrcToBuf(SB), NOFRAME|NOSPLIT, $0 + MVC $1, (R1), (R8) + RET + +TEXT ·mvcBufToDst(SB), NOFRAME|NOSPLIT, $0 + MVC $1, (R8), (R9) + RET + +#define BSWAP V5 +#define J0 V6 +#define KEY0 V7 +#define KEY1 V8 +#define NONCE V9 +#define CTR V10 +#define M0 V11 +#define M1 V12 +#define M2 V13 +#define M3 V14 +#define INC V15 +#define X0 V16 +#define X1 V17 +#define X2 V18 +#define X3 V19 +#define X4 V20 +#define X5 V21 +#define X6 V22 +#define X7 V23 +#define X8 V24 +#define X9 V25 +#define X10 V26 +#define X11 V27 +#define X12 V28 +#define X13 V29 +#define X14 V30 +#define X15 V31 + +#define NUM_ROUNDS 20 + +#define ROUND4(a0, a1, a2, a3, b0, b1, b2, b3, c0, c1, c2, c3, d0, d1, d2, d3) \ + VAF a1, a0, a0 \ + VAF b1, b0, b0 \ + VAF c1, c0, c0 \ + VAF d1, d0, d0 \ + VX a0, a2, a2 \ + VX b0, b2, b2 \ + VX c0, c2, c2 \ + VX d0, d2, d2 \ + VERLLF $16, a2, a2 \ + VERLLF $16, b2, b2 \ + VERLLF $16, c2, c2 \ + VERLLF $16, d2, d2 \ + VAF a2, a3, a3 \ + VAF b2, b3, b3 \ + VAF c2, c3, c3 \ + VAF d2, d3, d3 \ + VX a3, a1, a1 \ + VX b3, b1, b1 \ + VX c3, c1, c1 \ + VX d3, d1, d1 \ + VERLLF $12, a1, a1 \ + VERLLF $12, b1, b1 \ + VERLLF $12, c1, c1 \ + VERLLF $12, d1, d1 \ + VAF a1, a0, a0 \ + VAF b1, b0, b0 \ + VAF c1, c0, c0 \ + VAF d1, d0, d0 \ + VX a0, a2, a2 \ + VX b0, b2, b2 \ + VX c0, c2, c2 \ + VX d0, d2, d2 \ + VERLLF $8, a2, a2 \ + VERLLF $8, b2, b2 \ + VERLLF $8, c2, c2 \ + VERLLF $8, d2, d2 \ + VAF a2, a3, a3 \ + VAF b2, b3, b3 \ + VAF c2, c3, c3 \ + VAF d2, d3, d3 \ + VX a3, a1, a1 \ + VX b3, b1, b1 \ + VX c3, c1, c1 \ + VX d3, d1, d1 \ + VERLLF $7, a1, a1 \ + VERLLF $7, b1, b1 \ + VERLLF $7, c1, c1 \ + VERLLF $7, d1, d1 + +#define PERMUTE(mask, v0, v1, v2, v3) \ + VPERM v0, v0, mask, v0 \ + VPERM v1, v1, mask, v1 \ + VPERM v2, v2, mask, v2 \ + VPERM v3, v3, mask, v3 + +#define ADDV(x, v0, v1, v2, v3) \ + VAF x, v0, v0 \ + VAF x, v1, v1 \ + VAF x, v2, v2 \ + VAF x, v3, v3 + +#define XORV(off, dst, src, v0, v1, v2, v3) \ + VLM off(src), M0, M3 \ + PERMUTE(BSWAP, v0, v1, v2, v3) \ + VX v0, M0, M0 \ + VX v1, M1, M1 \ + VX v2, M2, M2 \ + VX v3, M3, M3 \ + VSTM M0, M3, off(dst) + +#define SHUFFLE(a, b, c, d, t, u, v, w) \ + VMRHF a, c, t \ // t = {a[0], c[0], a[1], c[1]} + VMRHF b, d, u \ // u = {b[0], d[0], b[1], d[1]} + VMRLF a, c, v \ // v = {a[2], c[2], a[3], c[3]} + VMRLF b, d, w \ // w = {b[2], d[2], b[3], d[3]} + VMRHF t, u, a \ // a = {a[0], b[0], c[0], d[0]} + VMRLF t, u, b \ // b = {a[1], b[1], c[1], d[1]} + VMRHF v, w, c \ // c = {a[2], b[2], c[2], d[2]} + VMRLF v, w, d // d = {a[3], b[3], c[3], d[3]} + +// func xorKeyStreamVX(dst, src []byte, key *[8]uint32, nonce *[3]uint32, counter *uint32, buf *[256]byte, len *int) +TEXT ·xorKeyStreamVX(SB), NOSPLIT, $0 + MOVD $·constants<>(SB), R1 + MOVD dst+0(FP), R2 // R2=&dst[0] + LMG src+24(FP), R3, R4 // R3=&src[0] R4=len(src) + MOVD key+48(FP), R5 // R5=key + MOVD nonce+56(FP), R6 // R6=nonce + MOVD counter+64(FP), R7 // R7=counter + MOVD buf+72(FP), R8 // R8=buf + MOVD len+80(FP), R9 // R9=len + + // load BSWAP and J0 + VLM (R1), BSWAP, J0 + + // set up tail buffer + ADD $-1, R4, R12 + MOVBZ R12, R12 + CMPUBEQ R12, $255, aligned + MOVD R4, R1 + AND $~255, R1 + MOVD $(R3)(R1*1), R1 + EXRL $·mvcSrcToBuf(SB), R12 + MOVD $255, R0 + SUB R12, R0 + MOVD R0, (R9) // update len + +aligned: + // setup + MOVD $95, R0 + VLM (R5), KEY0, KEY1 + VLL R0, (R6), NONCE + VZERO M0 + VLEIB $7, $32, M0 + VSRLB M0, NONCE, NONCE + + // initialize counter values + VLREPF (R7), CTR + VZERO INC + VLEIF $1, $1, INC + VLEIF $2, $2, INC + VLEIF $3, $3, INC + VAF INC, CTR, CTR + VREPIF $4, INC + +chacha: + VREPF $0, J0, X0 + VREPF $1, J0, X1 + VREPF $2, J0, X2 + VREPF $3, J0, X3 + VREPF $0, KEY0, X4 + VREPF $1, KEY0, X5 + VREPF $2, KEY0, X6 + VREPF $3, KEY0, X7 + VREPF $0, KEY1, X8 + VREPF $1, KEY1, X9 + VREPF $2, KEY1, X10 + VREPF $3, KEY1, X11 + VLR CTR, X12 + VREPF $1, NONCE, X13 + VREPF $2, NONCE, X14 + VREPF $3, NONCE, X15 + + MOVD $(NUM_ROUNDS/2), R1 + +loop: + ROUND4(X0, X4, X12, X8, X1, X5, X13, X9, X2, X6, X14, X10, X3, X7, X15, X11) + ROUND4(X0, X5, X15, X10, X1, X6, X12, X11, X2, X7, X13, X8, X3, X4, X14, X9) + + ADD $-1, R1 + BNE loop + + // decrement length + ADD $-256, R4 + BLT tail + +continue: + // rearrange vectors + SHUFFLE(X0, X1, X2, X3, M0, M1, M2, M3) + ADDV(J0, X0, X1, X2, X3) + SHUFFLE(X4, X5, X6, X7, M0, M1, M2, M3) + ADDV(KEY0, X4, X5, X6, X7) + SHUFFLE(X8, X9, X10, X11, M0, M1, M2, M3) + ADDV(KEY1, X8, X9, X10, X11) + VAF CTR, X12, X12 + SHUFFLE(X12, X13, X14, X15, M0, M1, M2, M3) + ADDV(NONCE, X12, X13, X14, X15) + + // increment counters + VAF INC, CTR, CTR + + // xor keystream with plaintext + XORV(0*64, R2, R3, X0, X4, X8, X12) + XORV(1*64, R2, R3, X1, X5, X9, X13) + XORV(2*64, R2, R3, X2, X6, X10, X14) + XORV(3*64, R2, R3, X3, X7, X11, X15) + + // increment pointers + MOVD $256(R2), R2 + MOVD $256(R3), R3 + + CMPBNE R4, $0, chacha + CMPUBEQ R12, $255, return + EXRL $·mvcBufToDst(SB), R12 // len was updated during setup + +return: + VSTEF $0, CTR, (R7) + RET + +tail: + MOVD R2, R9 + MOVD R8, R2 + MOVD R8, R3 + MOVD $0, R4 + JMP continue diff --git a/vendor/golang.org/x/crypto/internal/chacha20/xor.go b/vendor/golang.org/x/crypto/internal/chacha20/xor.go new file mode 100644 index 0000000..9c5ba0b --- /dev/null +++ b/vendor/golang.org/x/crypto/internal/chacha20/xor.go @@ -0,0 +1,43 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found src the LICENSE file. + +package chacha20 + +import ( + "runtime" +) + +// Platforms that have fast unaligned 32-bit little endian accesses. +const unaligned = runtime.GOARCH == "386" || + runtime.GOARCH == "amd64" || + runtime.GOARCH == "arm64" || + runtime.GOARCH == "ppc64le" || + runtime.GOARCH == "s390x" + +// xor reads a little endian uint32 from src, XORs it with u and +// places the result in little endian byte order in dst. +func xor(dst, src []byte, u uint32) { + _, _ = src[3], dst[3] // eliminate bounds checks + if unaligned { + // The compiler should optimize this code into + // 32-bit unaligned little endian loads and stores. + // TODO: delete once the compiler does a reliably + // good job with the generic code below. + // See issue #25111 for more details. + v := uint32(src[0]) + v |= uint32(src[1]) << 8 + v |= uint32(src[2]) << 16 + v |= uint32(src[3]) << 24 + v ^= u + dst[0] = byte(v) + dst[1] = byte(v >> 8) + dst[2] = byte(v >> 16) + dst[3] = byte(v >> 24) + } else { + dst[0] = src[0] ^ byte(u) + dst[1] = src[1] ^ byte(u>>8) + dst[2] = src[2] ^ byte(u>>16) + dst[3] = src[3] ^ byte(u>>24) + } +} diff --git a/vendor/golang.org/x/crypto/internal/subtle/aliasing.go b/vendor/golang.org/x/crypto/internal/subtle/aliasing.go new file mode 100644 index 0000000..f38797b --- /dev/null +++ b/vendor/golang.org/x/crypto/internal/subtle/aliasing.go @@ -0,0 +1,32 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build !appengine + +// Package subtle implements functions that are often useful in cryptographic +// code but require careful thought to use correctly. +package subtle // import "golang.org/x/crypto/internal/subtle" + +import "unsafe" + +// AnyOverlap reports whether x and y share memory at any (not necessarily +// corresponding) index. The memory beyond the slice length is ignored. +func AnyOverlap(x, y []byte) bool { + return len(x) > 0 && len(y) > 0 && + uintptr(unsafe.Pointer(&x[0])) <= uintptr(unsafe.Pointer(&y[len(y)-1])) && + uintptr(unsafe.Pointer(&y[0])) <= uintptr(unsafe.Pointer(&x[len(x)-1])) +} + +// InexactOverlap reports whether x and y share memory at any non-corresponding +// index. The memory beyond the slice length is ignored. Note that x and y can +// have different lengths and still not have any inexact overlap. +// +// InexactOverlap can be used to implement the requirements of the crypto/cipher +// AEAD, Block, BlockMode and Stream interfaces. +func InexactOverlap(x, y []byte) bool { + if len(x) == 0 || len(y) == 0 || &x[0] == &y[0] { + return false + } + return AnyOverlap(x, y) +} diff --git a/vendor/golang.org/x/crypto/internal/subtle/aliasing_appengine.go b/vendor/golang.org/x/crypto/internal/subtle/aliasing_appengine.go new file mode 100644 index 0000000..0cc4a8a --- /dev/null +++ b/vendor/golang.org/x/crypto/internal/subtle/aliasing_appengine.go @@ -0,0 +1,35 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build appengine + +// Package subtle implements functions that are often useful in cryptographic +// code but require careful thought to use correctly. +package subtle // import "golang.org/x/crypto/internal/subtle" + +// This is the Google App Engine standard variant based on reflect +// because the unsafe package and cgo are disallowed. + +import "reflect" + +// AnyOverlap reports whether x and y share memory at any (not necessarily +// corresponding) index. The memory beyond the slice length is ignored. +func AnyOverlap(x, y []byte) bool { + return len(x) > 0 && len(y) > 0 && + reflect.ValueOf(&x[0]).Pointer() <= reflect.ValueOf(&y[len(y)-1]).Pointer() && + reflect.ValueOf(&y[0]).Pointer() <= reflect.ValueOf(&x[len(x)-1]).Pointer() +} + +// InexactOverlap reports whether x and y share memory at any non-corresponding +// index. The memory beyond the slice length is ignored. Note that x and y can +// have different lengths and still not have any inexact overlap. +// +// InexactOverlap can be used to implement the requirements of the crypto/cipher +// AEAD, Block, BlockMode and Stream interfaces. +func InexactOverlap(x, y []byte) bool { + if len(x) == 0 || len(y) == 0 || &x[0] == &y[0] { + return false + } + return AnyOverlap(x, y) +} diff --git a/vendor/golang.org/x/crypto/poly1305/mac_noasm.go b/vendor/golang.org/x/crypto/poly1305/mac_noasm.go new file mode 100644 index 0000000..a8dd589 --- /dev/null +++ b/vendor/golang.org/x/crypto/poly1305/mac_noasm.go @@ -0,0 +1,11 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build !amd64,!ppc64le gccgo appengine + +package poly1305 + +type mac struct{ macGeneric } + +func newMAC(key *[32]byte) mac { return mac{newMACGeneric(key)} } diff --git a/vendor/golang.org/x/crypto/poly1305/poly1305.go b/vendor/golang.org/x/crypto/poly1305/poly1305.go new file mode 100644 index 0000000..d076a56 --- /dev/null +++ b/vendor/golang.org/x/crypto/poly1305/poly1305.go @@ -0,0 +1,83 @@ +// Copyright 2012 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// Package poly1305 implements Poly1305 one-time message authentication code as +// specified in https://cr.yp.to/mac/poly1305-20050329.pdf. +// +// Poly1305 is a fast, one-time authentication function. It is infeasible for an +// attacker to generate an authenticator for a message without the key. However, a +// key must only be used for a single message. Authenticating two different +// messages with the same key allows an attacker to forge authenticators for other +// messages with the same key. +// +// Poly1305 was originally coupled with AES in order to make Poly1305-AES. AES was +// used with a fixed key in order to generate one-time keys from an nonce. +// However, in this package AES isn't used and the one-time key is specified +// directly. +package poly1305 // import "golang.org/x/crypto/poly1305" + +import "crypto/subtle" + +// TagSize is the size, in bytes, of a poly1305 authenticator. +const TagSize = 16 + +// Verify returns true if mac is a valid authenticator for m with the given +// key. +func Verify(mac *[16]byte, m []byte, key *[32]byte) bool { + var tmp [16]byte + Sum(&tmp, m, key) + return subtle.ConstantTimeCompare(tmp[:], mac[:]) == 1 +} + +// New returns a new MAC computing an authentication +// tag of all data written to it with the given key. +// This allows writing the message progressively instead +// of passing it as a single slice. Common users should use +// the Sum function instead. +// +// The key must be unique for each message, as authenticating +// two different messages with the same key allows an attacker +// to forge messages at will. +func New(key *[32]byte) *MAC { + return &MAC{ + mac: newMAC(key), + finalized: false, + } +} + +// MAC is an io.Writer computing an authentication tag +// of the data written to it. +// +// MAC cannot be used like common hash.Hash implementations, +// because using a poly1305 key twice breaks its security. +// Therefore writing data to a running MAC after calling +// Sum causes it to panic. +type MAC struct { + mac // platform-dependent implementation + + finalized bool +} + +// Size returns the number of bytes Sum will return. +func (h *MAC) Size() int { return TagSize } + +// Write adds more data to the running message authentication code. +// It never returns an error. +// +// It must not be called after the first call of Sum. +func (h *MAC) Write(p []byte) (n int, err error) { + if h.finalized { + panic("poly1305: write to MAC after Sum") + } + return h.mac.Write(p) +} + +// Sum computes the authenticator of all data written to the +// message authentication code. +func (h *MAC) Sum(b []byte) []byte { + var mac [TagSize]byte + h.mac.Sum(&mac) + h.finalized = true + return append(b, mac[:]...) +} diff --git a/vendor/golang.org/x/crypto/poly1305/sum_amd64.go b/vendor/golang.org/x/crypto/poly1305/sum_amd64.go new file mode 100644 index 0000000..2dbf42a --- /dev/null +++ b/vendor/golang.org/x/crypto/poly1305/sum_amd64.go @@ -0,0 +1,68 @@ +// Copyright 2012 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build amd64,!gccgo,!appengine + +package poly1305 + +//go:noescape +func initialize(state *[7]uint64, key *[32]byte) + +//go:noescape +func update(state *[7]uint64, msg []byte) + +//go:noescape +func finalize(tag *[TagSize]byte, state *[7]uint64) + +// Sum generates an authenticator for m using a one-time key and puts the +// 16-byte result into out. Authenticating two different messages with the same +// key allows an attacker to forge messages at will. +func Sum(out *[16]byte, m []byte, key *[32]byte) { + h := newMAC(key) + h.Write(m) + h.Sum(out) +} + +func newMAC(key *[32]byte) (h mac) { + initialize(&h.state, key) + return +} + +type mac struct { + state [7]uint64 // := uint64{ h0, h1, h2, r0, r1, pad0, pad1 } + + buffer [TagSize]byte + offset int +} + +func (h *mac) Write(p []byte) (n int, err error) { + n = len(p) + if h.offset > 0 { + remaining := TagSize - h.offset + if n < remaining { + h.offset += copy(h.buffer[h.offset:], p) + return n, nil + } + copy(h.buffer[h.offset:], p[:remaining]) + p = p[remaining:] + h.offset = 0 + update(&h.state, h.buffer[:]) + } + if nn := len(p) - (len(p) % TagSize); nn > 0 { + update(&h.state, p[:nn]) + p = p[nn:] + } + if len(p) > 0 { + h.offset += copy(h.buffer[h.offset:], p) + } + return n, nil +} + +func (h *mac) Sum(out *[16]byte) { + state := h.state + if h.offset > 0 { + update(&state, h.buffer[:h.offset]) + } + finalize(out, &state) +} diff --git a/vendor/golang.org/x/crypto/poly1305/sum_amd64.s b/vendor/golang.org/x/crypto/poly1305/sum_amd64.s new file mode 100644 index 0000000..7d600f1 --- /dev/null +++ b/vendor/golang.org/x/crypto/poly1305/sum_amd64.s @@ -0,0 +1,148 @@ +// Copyright 2012 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build amd64,!gccgo,!appengine + +#include "textflag.h" + +#define POLY1305_ADD(msg, h0, h1, h2) \ + ADDQ 0(msg), h0; \ + ADCQ 8(msg), h1; \ + ADCQ $1, h2; \ + LEAQ 16(msg), msg + +#define POLY1305_MUL(h0, h1, h2, r0, r1, t0, t1, t2, t3) \ + MOVQ r0, AX; \ + MULQ h0; \ + MOVQ AX, t0; \ + MOVQ DX, t1; \ + MOVQ r0, AX; \ + MULQ h1; \ + ADDQ AX, t1; \ + ADCQ $0, DX; \ + MOVQ r0, t2; \ + IMULQ h2, t2; \ + ADDQ DX, t2; \ + \ + MOVQ r1, AX; \ + MULQ h0; \ + ADDQ AX, t1; \ + ADCQ $0, DX; \ + MOVQ DX, h0; \ + MOVQ r1, t3; \ + IMULQ h2, t3; \ + MOVQ r1, AX; \ + MULQ h1; \ + ADDQ AX, t2; \ + ADCQ DX, t3; \ + ADDQ h0, t2; \ + ADCQ $0, t3; \ + \ + MOVQ t0, h0; \ + MOVQ t1, h1; \ + MOVQ t2, h2; \ + ANDQ $3, h2; \ + MOVQ t2, t0; \ + ANDQ $0xFFFFFFFFFFFFFFFC, t0; \ + ADDQ t0, h0; \ + ADCQ t3, h1; \ + ADCQ $0, h2; \ + SHRQ $2, t3, t2; \ + SHRQ $2, t3; \ + ADDQ t2, h0; \ + ADCQ t3, h1; \ + ADCQ $0, h2 + +DATA ·poly1305Mask<>+0x00(SB)/8, $0x0FFFFFFC0FFFFFFF +DATA ·poly1305Mask<>+0x08(SB)/8, $0x0FFFFFFC0FFFFFFC +GLOBL ·poly1305Mask<>(SB), RODATA, $16 + +// func update(state *[7]uint64, msg []byte) +TEXT ·update(SB), $0-32 + MOVQ state+0(FP), DI + MOVQ msg_base+8(FP), SI + MOVQ msg_len+16(FP), R15 + + MOVQ 0(DI), R8 // h0 + MOVQ 8(DI), R9 // h1 + MOVQ 16(DI), R10 // h2 + MOVQ 24(DI), R11 // r0 + MOVQ 32(DI), R12 // r1 + + CMPQ R15, $16 + JB bytes_between_0_and_15 + +loop: + POLY1305_ADD(SI, R8, R9, R10) + +multiply: + POLY1305_MUL(R8, R9, R10, R11, R12, BX, CX, R13, R14) + SUBQ $16, R15 + CMPQ R15, $16 + JAE loop + +bytes_between_0_and_15: + TESTQ R15, R15 + JZ done + MOVQ $1, BX + XORQ CX, CX + XORQ R13, R13 + ADDQ R15, SI + +flush_buffer: + SHLQ $8, BX, CX + SHLQ $8, BX + MOVB -1(SI), R13 + XORQ R13, BX + DECQ SI + DECQ R15 + JNZ flush_buffer + + ADDQ BX, R8 + ADCQ CX, R9 + ADCQ $0, R10 + MOVQ $16, R15 + JMP multiply + +done: + MOVQ R8, 0(DI) + MOVQ R9, 8(DI) + MOVQ R10, 16(DI) + RET + +// func initialize(state *[7]uint64, key *[32]byte) +TEXT ·initialize(SB), $0-16 + MOVQ state+0(FP), DI + MOVQ key+8(FP), SI + + // state[0...7] is initialized with zero + MOVOU 0(SI), X0 + MOVOU 16(SI), X1 + MOVOU ·poly1305Mask<>(SB), X2 + PAND X2, X0 + MOVOU X0, 24(DI) + MOVOU X1, 40(DI) + RET + +// func finalize(tag *[TagSize]byte, state *[7]uint64) +TEXT ·finalize(SB), $0-16 + MOVQ tag+0(FP), DI + MOVQ state+8(FP), SI + + MOVQ 0(SI), AX + MOVQ 8(SI), BX + MOVQ 16(SI), CX + MOVQ AX, R8 + MOVQ BX, R9 + SUBQ $0xFFFFFFFFFFFFFFFB, AX + SBBQ $0xFFFFFFFFFFFFFFFF, BX + SBBQ $3, CX + CMOVQCS R8, AX + CMOVQCS R9, BX + ADDQ 40(SI), AX + ADCQ 48(SI), BX + + MOVQ AX, 0(DI) + MOVQ BX, 8(DI) + RET diff --git a/vendor/golang.org/x/crypto/poly1305/sum_arm.go b/vendor/golang.org/x/crypto/poly1305/sum_arm.go new file mode 100644 index 0000000..5dc321c --- /dev/null +++ b/vendor/golang.org/x/crypto/poly1305/sum_arm.go @@ -0,0 +1,22 @@ +// Copyright 2015 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build arm,!gccgo,!appengine,!nacl + +package poly1305 + +// This function is implemented in sum_arm.s +//go:noescape +func poly1305_auth_armv6(out *[16]byte, m *byte, mlen uint32, key *[32]byte) + +// Sum generates an authenticator for m using a one-time key and puts the +// 16-byte result into out. Authenticating two different messages with the same +// key allows an attacker to forge messages at will. +func Sum(out *[16]byte, m []byte, key *[32]byte) { + var mPtr *byte + if len(m) > 0 { + mPtr = &m[0] + } + poly1305_auth_armv6(out, mPtr, uint32(len(m)), key) +} diff --git a/vendor/golang.org/x/crypto/poly1305/sum_arm.s b/vendor/golang.org/x/crypto/poly1305/sum_arm.s new file mode 100644 index 0000000..f70b4ac --- /dev/null +++ b/vendor/golang.org/x/crypto/poly1305/sum_arm.s @@ -0,0 +1,427 @@ +// Copyright 2015 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build arm,!gccgo,!appengine,!nacl + +#include "textflag.h" + +// This code was translated into a form compatible with 5a from the public +// domain source by Andrew Moon: github.com/floodyberry/poly1305-opt/blob/master/app/extensions/poly1305. + +DATA ·poly1305_init_constants_armv6<>+0x00(SB)/4, $0x3ffffff +DATA ·poly1305_init_constants_armv6<>+0x04(SB)/4, $0x3ffff03 +DATA ·poly1305_init_constants_armv6<>+0x08(SB)/4, $0x3ffc0ff +DATA ·poly1305_init_constants_armv6<>+0x0c(SB)/4, $0x3f03fff +DATA ·poly1305_init_constants_armv6<>+0x10(SB)/4, $0x00fffff +GLOBL ·poly1305_init_constants_armv6<>(SB), 8, $20 + +// Warning: the linker may use R11 to synthesize certain instructions. Please +// take care and verify that no synthetic instructions use it. + +TEXT poly1305_init_ext_armv6<>(SB), NOSPLIT, $0 + // Needs 16 bytes of stack and 64 bytes of space pointed to by R0. (It + // might look like it's only 60 bytes of space but the final four bytes + // will be written by another function.) We need to skip over four + // bytes of stack because that's saving the value of 'g'. + ADD $4, R13, R8 + MOVM.IB [R4-R7], (R8) + MOVM.IA.W (R1), [R2-R5] + MOVW $·poly1305_init_constants_armv6<>(SB), R7 + MOVW R2, R8 + MOVW R2>>26, R9 + MOVW R3>>20, g + MOVW R4>>14, R11 + MOVW R5>>8, R12 + ORR R3<<6, R9, R9 + ORR R4<<12, g, g + ORR R5<<18, R11, R11 + MOVM.IA (R7), [R2-R6] + AND R8, R2, R2 + AND R9, R3, R3 + AND g, R4, R4 + AND R11, R5, R5 + AND R12, R6, R6 + MOVM.IA.W [R2-R6], (R0) + EOR R2, R2, R2 + EOR R3, R3, R3 + EOR R4, R4, R4 + EOR R5, R5, R5 + EOR R6, R6, R6 + MOVM.IA.W [R2-R6], (R0) + MOVM.IA.W (R1), [R2-R5] + MOVM.IA [R2-R6], (R0) + ADD $20, R13, R0 + MOVM.DA (R0), [R4-R7] + RET + +#define MOVW_UNALIGNED(Rsrc, Rdst, Rtmp, offset) \ + MOVBU (offset+0)(Rsrc), Rtmp; \ + MOVBU Rtmp, (offset+0)(Rdst); \ + MOVBU (offset+1)(Rsrc), Rtmp; \ + MOVBU Rtmp, (offset+1)(Rdst); \ + MOVBU (offset+2)(Rsrc), Rtmp; \ + MOVBU Rtmp, (offset+2)(Rdst); \ + MOVBU (offset+3)(Rsrc), Rtmp; \ + MOVBU Rtmp, (offset+3)(Rdst) + +TEXT poly1305_blocks_armv6<>(SB), NOSPLIT, $0 + // Needs 24 bytes of stack for saved registers and then 88 bytes of + // scratch space after that. We assume that 24 bytes at (R13) have + // already been used: four bytes for the link register saved in the + // prelude of poly1305_auth_armv6, four bytes for saving the value of g + // in that function and 16 bytes of scratch space used around + // poly1305_finish_ext_armv6_skip1. + ADD $24, R13, R12 + MOVM.IB [R4-R8, R14], (R12) + MOVW R0, 88(R13) + MOVW R1, 92(R13) + MOVW R2, 96(R13) + MOVW R1, R14 + MOVW R2, R12 + MOVW 56(R0), R8 + WORD $0xe1180008 // TST R8, R8 not working see issue 5921 + EOR R6, R6, R6 + MOVW.EQ $(1<<24), R6 + MOVW R6, 84(R13) + ADD $116, R13, g + MOVM.IA (R0), [R0-R9] + MOVM.IA [R0-R4], (g) + CMP $16, R12 + BLO poly1305_blocks_armv6_done + +poly1305_blocks_armv6_mainloop: + WORD $0xe31e0003 // TST R14, #3 not working see issue 5921 + BEQ poly1305_blocks_armv6_mainloop_aligned + ADD $100, R13, g + MOVW_UNALIGNED(R14, g, R0, 0) + MOVW_UNALIGNED(R14, g, R0, 4) + MOVW_UNALIGNED(R14, g, R0, 8) + MOVW_UNALIGNED(R14, g, R0, 12) + MOVM.IA (g), [R0-R3] + ADD $16, R14 + B poly1305_blocks_armv6_mainloop_loaded + +poly1305_blocks_armv6_mainloop_aligned: + MOVM.IA.W (R14), [R0-R3] + +poly1305_blocks_armv6_mainloop_loaded: + MOVW R0>>26, g + MOVW R1>>20, R11 + MOVW R2>>14, R12 + MOVW R14, 92(R13) + MOVW R3>>8, R4 + ORR R1<<6, g, g + ORR R2<<12, R11, R11 + ORR R3<<18, R12, R12 + BIC $0xfc000000, R0, R0 + BIC $0xfc000000, g, g + MOVW 84(R13), R3 + BIC $0xfc000000, R11, R11 + BIC $0xfc000000, R12, R12 + ADD R0, R5, R5 + ADD g, R6, R6 + ORR R3, R4, R4 + ADD R11, R7, R7 + ADD $116, R13, R14 + ADD R12, R8, R8 + ADD R4, R9, R9 + MOVM.IA (R14), [R0-R4] + MULLU R4, R5, (R11, g) + MULLU R3, R5, (R14, R12) + MULALU R3, R6, (R11, g) + MULALU R2, R6, (R14, R12) + MULALU R2, R7, (R11, g) + MULALU R1, R7, (R14, R12) + ADD R4<<2, R4, R4 + ADD R3<<2, R3, R3 + MULALU R1, R8, (R11, g) + MULALU R0, R8, (R14, R12) + MULALU R0, R9, (R11, g) + MULALU R4, R9, (R14, R12) + MOVW g, 76(R13) + MOVW R11, 80(R13) + MOVW R12, 68(R13) + MOVW R14, 72(R13) + MULLU R2, R5, (R11, g) + MULLU R1, R5, (R14, R12) + MULALU R1, R6, (R11, g) + MULALU R0, R6, (R14, R12) + MULALU R0, R7, (R11, g) + MULALU R4, R7, (R14, R12) + ADD R2<<2, R2, R2 + ADD R1<<2, R1, R1 + MULALU R4, R8, (R11, g) + MULALU R3, R8, (R14, R12) + MULALU R3, R9, (R11, g) + MULALU R2, R9, (R14, R12) + MOVW g, 60(R13) + MOVW R11, 64(R13) + MOVW R12, 52(R13) + MOVW R14, 56(R13) + MULLU R0, R5, (R11, g) + MULALU R4, R6, (R11, g) + MULALU R3, R7, (R11, g) + MULALU R2, R8, (R11, g) + MULALU R1, R9, (R11, g) + ADD $52, R13, R0 + MOVM.IA (R0), [R0-R7] + MOVW g>>26, R12 + MOVW R4>>26, R14 + ORR R11<<6, R12, R12 + ORR R5<<6, R14, R14 + BIC $0xfc000000, g, g + BIC $0xfc000000, R4, R4 + ADD.S R12, R0, R0 + ADC $0, R1, R1 + ADD.S R14, R6, R6 + ADC $0, R7, R7 + MOVW R0>>26, R12 + MOVW R6>>26, R14 + ORR R1<<6, R12, R12 + ORR R7<<6, R14, R14 + BIC $0xfc000000, R0, R0 + BIC $0xfc000000, R6, R6 + ADD R14<<2, R14, R14 + ADD.S R12, R2, R2 + ADC $0, R3, R3 + ADD R14, g, g + MOVW R2>>26, R12 + MOVW g>>26, R14 + ORR R3<<6, R12, R12 + BIC $0xfc000000, g, R5 + BIC $0xfc000000, R2, R7 + ADD R12, R4, R4 + ADD R14, R0, R0 + MOVW R4>>26, R12 + BIC $0xfc000000, R4, R8 + ADD R12, R6, R9 + MOVW 96(R13), R12 + MOVW 92(R13), R14 + MOVW R0, R6 + CMP $32, R12 + SUB $16, R12, R12 + MOVW R12, 96(R13) + BHS poly1305_blocks_armv6_mainloop + +poly1305_blocks_armv6_done: + MOVW 88(R13), R12 + MOVW R5, 20(R12) + MOVW R6, 24(R12) + MOVW R7, 28(R12) + MOVW R8, 32(R12) + MOVW R9, 36(R12) + ADD $48, R13, R0 + MOVM.DA (R0), [R4-R8, R14] + RET + +#define MOVHUP_UNALIGNED(Rsrc, Rdst, Rtmp) \ + MOVBU.P 1(Rsrc), Rtmp; \ + MOVBU.P Rtmp, 1(Rdst); \ + MOVBU.P 1(Rsrc), Rtmp; \ + MOVBU.P Rtmp, 1(Rdst) + +#define MOVWP_UNALIGNED(Rsrc, Rdst, Rtmp) \ + MOVHUP_UNALIGNED(Rsrc, Rdst, Rtmp); \ + MOVHUP_UNALIGNED(Rsrc, Rdst, Rtmp) + +// func poly1305_auth_armv6(out *[16]byte, m *byte, mlen uint32, key *[32]key) +TEXT ·poly1305_auth_armv6(SB), $196-16 + // The value 196, just above, is the sum of 64 (the size of the context + // structure) and 132 (the amount of stack needed). + // + // At this point, the stack pointer (R13) has been moved down. It + // points to the saved link register and there's 196 bytes of free + // space above it. + // + // The stack for this function looks like: + // + // +--------------------- + // | + // | 64 bytes of context structure + // | + // +--------------------- + // | + // | 112 bytes for poly1305_blocks_armv6 + // | + // +--------------------- + // | 16 bytes of final block, constructed at + // | poly1305_finish_ext_armv6_skip8 + // +--------------------- + // | four bytes of saved 'g' + // +--------------------- + // | lr, saved by prelude <- R13 points here + // +--------------------- + MOVW g, 4(R13) + + MOVW out+0(FP), R4 + MOVW m+4(FP), R5 + MOVW mlen+8(FP), R6 + MOVW key+12(FP), R7 + + ADD $136, R13, R0 // 136 = 4 + 4 + 16 + 112 + MOVW R7, R1 + + // poly1305_init_ext_armv6 will write to the stack from R13+4, but + // that's ok because none of the other values have been written yet. + BL poly1305_init_ext_armv6<>(SB) + BIC.S $15, R6, R2 + BEQ poly1305_auth_armv6_noblocks + ADD $136, R13, R0 + MOVW R5, R1 + ADD R2, R5, R5 + SUB R2, R6, R6 + BL poly1305_blocks_armv6<>(SB) + +poly1305_auth_armv6_noblocks: + ADD $136, R13, R0 + MOVW R5, R1 + MOVW R6, R2 + MOVW R4, R3 + + MOVW R0, R5 + MOVW R1, R6 + MOVW R2, R7 + MOVW R3, R8 + AND.S R2, R2, R2 + BEQ poly1305_finish_ext_armv6_noremaining + EOR R0, R0 + ADD $8, R13, R9 // 8 = offset to 16 byte scratch space + MOVW R0, (R9) + MOVW R0, 4(R9) + MOVW R0, 8(R9) + MOVW R0, 12(R9) + WORD $0xe3110003 // TST R1, #3 not working see issue 5921 + BEQ poly1305_finish_ext_armv6_aligned + WORD $0xe3120008 // TST R2, #8 not working see issue 5921 + BEQ poly1305_finish_ext_armv6_skip8 + MOVWP_UNALIGNED(R1, R9, g) + MOVWP_UNALIGNED(R1, R9, g) + +poly1305_finish_ext_armv6_skip8: + WORD $0xe3120004 // TST $4, R2 not working see issue 5921 + BEQ poly1305_finish_ext_armv6_skip4 + MOVWP_UNALIGNED(R1, R9, g) + +poly1305_finish_ext_armv6_skip4: + WORD $0xe3120002 // TST $2, R2 not working see issue 5921 + BEQ poly1305_finish_ext_armv6_skip2 + MOVHUP_UNALIGNED(R1, R9, g) + B poly1305_finish_ext_armv6_skip2 + +poly1305_finish_ext_armv6_aligned: + WORD $0xe3120008 // TST R2, #8 not working see issue 5921 + BEQ poly1305_finish_ext_armv6_skip8_aligned + MOVM.IA.W (R1), [g-R11] + MOVM.IA.W [g-R11], (R9) + +poly1305_finish_ext_armv6_skip8_aligned: + WORD $0xe3120004 // TST $4, R2 not working see issue 5921 + BEQ poly1305_finish_ext_armv6_skip4_aligned + MOVW.P 4(R1), g + MOVW.P g, 4(R9) + +poly1305_finish_ext_armv6_skip4_aligned: + WORD $0xe3120002 // TST $2, R2 not working see issue 5921 + BEQ poly1305_finish_ext_armv6_skip2 + MOVHU.P 2(R1), g + MOVH.P g, 2(R9) + +poly1305_finish_ext_armv6_skip2: + WORD $0xe3120001 // TST $1, R2 not working see issue 5921 + BEQ poly1305_finish_ext_armv6_skip1 + MOVBU.P 1(R1), g + MOVBU.P g, 1(R9) + +poly1305_finish_ext_armv6_skip1: + MOVW $1, R11 + MOVBU R11, 0(R9) + MOVW R11, 56(R5) + MOVW R5, R0 + ADD $8, R13, R1 + MOVW $16, R2 + BL poly1305_blocks_armv6<>(SB) + +poly1305_finish_ext_armv6_noremaining: + MOVW 20(R5), R0 + MOVW 24(R5), R1 + MOVW 28(R5), R2 + MOVW 32(R5), R3 + MOVW 36(R5), R4 + MOVW R4>>26, R12 + BIC $0xfc000000, R4, R4 + ADD R12<<2, R12, R12 + ADD R12, R0, R0 + MOVW R0>>26, R12 + BIC $0xfc000000, R0, R0 + ADD R12, R1, R1 + MOVW R1>>26, R12 + BIC $0xfc000000, R1, R1 + ADD R12, R2, R2 + MOVW R2>>26, R12 + BIC $0xfc000000, R2, R2 + ADD R12, R3, R3 + MOVW R3>>26, R12 + BIC $0xfc000000, R3, R3 + ADD R12, R4, R4 + ADD $5, R0, R6 + MOVW R6>>26, R12 + BIC $0xfc000000, R6, R6 + ADD R12, R1, R7 + MOVW R7>>26, R12 + BIC $0xfc000000, R7, R7 + ADD R12, R2, g + MOVW g>>26, R12 + BIC $0xfc000000, g, g + ADD R12, R3, R11 + MOVW $-(1<<26), R12 + ADD R11>>26, R12, R12 + BIC $0xfc000000, R11, R11 + ADD R12, R4, R9 + MOVW R9>>31, R12 + SUB $1, R12 + AND R12, R6, R6 + AND R12, R7, R7 + AND R12, g, g + AND R12, R11, R11 + AND R12, R9, R9 + MVN R12, R12 + AND R12, R0, R0 + AND R12, R1, R1 + AND R12, R2, R2 + AND R12, R3, R3 + AND R12, R4, R4 + ORR R6, R0, R0 + ORR R7, R1, R1 + ORR g, R2, R2 + ORR R11, R3, R3 + ORR R9, R4, R4 + ORR R1<<26, R0, R0 + MOVW R1>>6, R1 + ORR R2<<20, R1, R1 + MOVW R2>>12, R2 + ORR R3<<14, R2, R2 + MOVW R3>>18, R3 + ORR R4<<8, R3, R3 + MOVW 40(R5), R6 + MOVW 44(R5), R7 + MOVW 48(R5), g + MOVW 52(R5), R11 + ADD.S R6, R0, R0 + ADC.S R7, R1, R1 + ADC.S g, R2, R2 + ADC.S R11, R3, R3 + MOVM.IA [R0-R3], (R8) + MOVW R5, R12 + EOR R0, R0, R0 + EOR R1, R1, R1 + EOR R2, R2, R2 + EOR R3, R3, R3 + EOR R4, R4, R4 + EOR R5, R5, R5 + EOR R6, R6, R6 + EOR R7, R7, R7 + MOVM.IA.W [R0-R7], (R12) + MOVM.IA [R0-R7], (R12) + MOVW 4(R13), g + RET diff --git a/vendor/golang.org/x/crypto/poly1305/sum_generic.go b/vendor/golang.org/x/crypto/poly1305/sum_generic.go new file mode 100644 index 0000000..bab76ef --- /dev/null +++ b/vendor/golang.org/x/crypto/poly1305/sum_generic.go @@ -0,0 +1,172 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package poly1305 + +import "encoding/binary" + +const ( + msgBlock = uint32(1 << 24) + finalBlock = uint32(0) +) + +// sumGeneric generates an authenticator for msg using a one-time key and +// puts the 16-byte result into out. This is the generic implementation of +// Sum and should be called if no assembly implementation is available. +func sumGeneric(out *[TagSize]byte, msg []byte, key *[32]byte) { + h := newMACGeneric(key) + h.Write(msg) + h.Sum(out) +} + +func newMACGeneric(key *[32]byte) (h macGeneric) { + h.r[0] = binary.LittleEndian.Uint32(key[0:]) & 0x3ffffff + h.r[1] = (binary.LittleEndian.Uint32(key[3:]) >> 2) & 0x3ffff03 + h.r[2] = (binary.LittleEndian.Uint32(key[6:]) >> 4) & 0x3ffc0ff + h.r[3] = (binary.LittleEndian.Uint32(key[9:]) >> 6) & 0x3f03fff + h.r[4] = (binary.LittleEndian.Uint32(key[12:]) >> 8) & 0x00fffff + + h.s[0] = binary.LittleEndian.Uint32(key[16:]) + h.s[1] = binary.LittleEndian.Uint32(key[20:]) + h.s[2] = binary.LittleEndian.Uint32(key[24:]) + h.s[3] = binary.LittleEndian.Uint32(key[28:]) + return +} + +type macGeneric struct { + h, r [5]uint32 + s [4]uint32 + + buffer [TagSize]byte + offset int +} + +func (h *macGeneric) Write(p []byte) (n int, err error) { + n = len(p) + if h.offset > 0 { + remaining := TagSize - h.offset + if n < remaining { + h.offset += copy(h.buffer[h.offset:], p) + return n, nil + } + copy(h.buffer[h.offset:], p[:remaining]) + p = p[remaining:] + h.offset = 0 + updateGeneric(h.buffer[:], msgBlock, &(h.h), &(h.r)) + } + if nn := len(p) - (len(p) % TagSize); nn > 0 { + updateGeneric(p, msgBlock, &(h.h), &(h.r)) + p = p[nn:] + } + if len(p) > 0 { + h.offset += copy(h.buffer[h.offset:], p) + } + return n, nil +} + +func (h *macGeneric) Sum(out *[16]byte) { + H, R := h.h, h.r + if h.offset > 0 { + var buffer [TagSize]byte + copy(buffer[:], h.buffer[:h.offset]) + buffer[h.offset] = 1 // invariant: h.offset < TagSize + updateGeneric(buffer[:], finalBlock, &H, &R) + } + finalizeGeneric(out, &H, &(h.s)) +} + +func updateGeneric(msg []byte, flag uint32, h, r *[5]uint32) { + h0, h1, h2, h3, h4 := h[0], h[1], h[2], h[3], h[4] + r0, r1, r2, r3, r4 := uint64(r[0]), uint64(r[1]), uint64(r[2]), uint64(r[3]), uint64(r[4]) + R1, R2, R3, R4 := r1*5, r2*5, r3*5, r4*5 + + for len(msg) >= TagSize { + // h += msg + h0 += binary.LittleEndian.Uint32(msg[0:]) & 0x3ffffff + h1 += (binary.LittleEndian.Uint32(msg[3:]) >> 2) & 0x3ffffff + h2 += (binary.LittleEndian.Uint32(msg[6:]) >> 4) & 0x3ffffff + h3 += (binary.LittleEndian.Uint32(msg[9:]) >> 6) & 0x3ffffff + h4 += (binary.LittleEndian.Uint32(msg[12:]) >> 8) | flag + + // h *= r + d0 := (uint64(h0) * r0) + (uint64(h1) * R4) + (uint64(h2) * R3) + (uint64(h3) * R2) + (uint64(h4) * R1) + d1 := (d0 >> 26) + (uint64(h0) * r1) + (uint64(h1) * r0) + (uint64(h2) * R4) + (uint64(h3) * R3) + (uint64(h4) * R2) + d2 := (d1 >> 26) + (uint64(h0) * r2) + (uint64(h1) * r1) + (uint64(h2) * r0) + (uint64(h3) * R4) + (uint64(h4) * R3) + d3 := (d2 >> 26) + (uint64(h0) * r3) + (uint64(h1) * r2) + (uint64(h2) * r1) + (uint64(h3) * r0) + (uint64(h4) * R4) + d4 := (d3 >> 26) + (uint64(h0) * r4) + (uint64(h1) * r3) + (uint64(h2) * r2) + (uint64(h3) * r1) + (uint64(h4) * r0) + + // h %= p + h0 = uint32(d0) & 0x3ffffff + h1 = uint32(d1) & 0x3ffffff + h2 = uint32(d2) & 0x3ffffff + h3 = uint32(d3) & 0x3ffffff + h4 = uint32(d4) & 0x3ffffff + + h0 += uint32(d4>>26) * 5 + h1 += h0 >> 26 + h0 = h0 & 0x3ffffff + + msg = msg[TagSize:] + } + + h[0], h[1], h[2], h[3], h[4] = h0, h1, h2, h3, h4 +} + +func finalizeGeneric(out *[TagSize]byte, h *[5]uint32, s *[4]uint32) { + h0, h1, h2, h3, h4 := h[0], h[1], h[2], h[3], h[4] + + // h %= p reduction + h2 += h1 >> 26 + h1 &= 0x3ffffff + h3 += h2 >> 26 + h2 &= 0x3ffffff + h4 += h3 >> 26 + h3 &= 0x3ffffff + h0 += 5 * (h4 >> 26) + h4 &= 0x3ffffff + h1 += h0 >> 26 + h0 &= 0x3ffffff + + // h - p + t0 := h0 + 5 + t1 := h1 + (t0 >> 26) + t2 := h2 + (t1 >> 26) + t3 := h3 + (t2 >> 26) + t4 := h4 + (t3 >> 26) - (1 << 26) + t0 &= 0x3ffffff + t1 &= 0x3ffffff + t2 &= 0x3ffffff + t3 &= 0x3ffffff + + // select h if h < p else h - p + t_mask := (t4 >> 31) - 1 + h_mask := ^t_mask + h0 = (h0 & h_mask) | (t0 & t_mask) + h1 = (h1 & h_mask) | (t1 & t_mask) + h2 = (h2 & h_mask) | (t2 & t_mask) + h3 = (h3 & h_mask) | (t3 & t_mask) + h4 = (h4 & h_mask) | (t4 & t_mask) + + // h %= 2^128 + h0 |= h1 << 26 + h1 = ((h1 >> 6) | (h2 << 20)) + h2 = ((h2 >> 12) | (h3 << 14)) + h3 = ((h3 >> 18) | (h4 << 8)) + + // s: the s part of the key + // tag = (h + s) % (2^128) + t := uint64(h0) + uint64(s[0]) + h0 = uint32(t) + t = uint64(h1) + uint64(s[1]) + (t >> 32) + h1 = uint32(t) + t = uint64(h2) + uint64(s[2]) + (t >> 32) + h2 = uint32(t) + t = uint64(h3) + uint64(s[3]) + (t >> 32) + h3 = uint32(t) + + binary.LittleEndian.PutUint32(out[0:], h0) + binary.LittleEndian.PutUint32(out[4:], h1) + binary.LittleEndian.PutUint32(out[8:], h2) + binary.LittleEndian.PutUint32(out[12:], h3) +} diff --git a/vendor/golang.org/x/crypto/poly1305/sum_noasm.go b/vendor/golang.org/x/crypto/poly1305/sum_noasm.go new file mode 100644 index 0000000..8a9c207 --- /dev/null +++ b/vendor/golang.org/x/crypto/poly1305/sum_noasm.go @@ -0,0 +1,16 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build s390x,!go1.11 !arm,!amd64,!s390x,!ppc64le gccgo appengine nacl + +package poly1305 + +// Sum generates an authenticator for msg using a one-time key and puts the +// 16-byte result into out. Authenticating two different messages with the same +// key allows an attacker to forge messages at will. +func Sum(out *[TagSize]byte, msg []byte, key *[32]byte) { + h := newMAC(key) + h.Write(msg) + h.Sum(out) +} diff --git a/vendor/golang.org/x/crypto/poly1305/sum_ppc64le.go b/vendor/golang.org/x/crypto/poly1305/sum_ppc64le.go new file mode 100644 index 0000000..2402b63 --- /dev/null +++ b/vendor/golang.org/x/crypto/poly1305/sum_ppc64le.go @@ -0,0 +1,68 @@ +// Copyright 2019 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build ppc64le,!gccgo,!appengine + +package poly1305 + +//go:noescape +func initialize(state *[7]uint64, key *[32]byte) + +//go:noescape +func update(state *[7]uint64, msg []byte) + +//go:noescape +func finalize(tag *[TagSize]byte, state *[7]uint64) + +// Sum generates an authenticator for m using a one-time key and puts the +// 16-byte result into out. Authenticating two different messages with the same +// key allows an attacker to forge messages at will. +func Sum(out *[16]byte, m []byte, key *[32]byte) { + h := newMAC(key) + h.Write(m) + h.Sum(out) +} + +func newMAC(key *[32]byte) (h mac) { + initialize(&h.state, key) + return +} + +type mac struct { + state [7]uint64 // := uint64{ h0, h1, h2, r0, r1, pad0, pad1 } + + buffer [TagSize]byte + offset int +} + +func (h *mac) Write(p []byte) (n int, err error) { + n = len(p) + if h.offset > 0 { + remaining := TagSize - h.offset + if n < remaining { + h.offset += copy(h.buffer[h.offset:], p) + return n, nil + } + copy(h.buffer[h.offset:], p[:remaining]) + p = p[remaining:] + h.offset = 0 + update(&h.state, h.buffer[:]) + } + if nn := len(p) - (len(p) % TagSize); nn > 0 { + update(&h.state, p[:nn]) + p = p[nn:] + } + if len(p) > 0 { + h.offset += copy(h.buffer[h.offset:], p) + } + return n, nil +} + +func (h *mac) Sum(out *[16]byte) { + state := h.state + if h.offset > 0 { + update(&state, h.buffer[:h.offset]) + } + finalize(out, &state) +} diff --git a/vendor/golang.org/x/crypto/poly1305/sum_ppc64le.s b/vendor/golang.org/x/crypto/poly1305/sum_ppc64le.s new file mode 100644 index 0000000..55c7167 --- /dev/null +++ b/vendor/golang.org/x/crypto/poly1305/sum_ppc64le.s @@ -0,0 +1,247 @@ +// Copyright 2019 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build ppc64le,!gccgo,!appengine + +#include "textflag.h" + +// This was ported from the amd64 implementation. + +#define POLY1305_ADD(msg, h0, h1, h2, t0, t1, t2) \ + MOVD (msg), t0; \ + MOVD 8(msg), t1; \ + MOVD $1, t2; \ + ADDC t0, h0, h0; \ + ADDE t1, h1, h1; \ + ADDE t2, h2; \ + ADD $16, msg + +#define POLY1305_MUL(h0, h1, h2, r0, r1, t0, t1, t2, t3, t4, t5) \ + MULLD r0, h0, t0; \ + MULLD r0, h1, t4; \ + MULHDU r0, h0, t1; \ + MULHDU r0, h1, t5; \ + ADDC t4, t1, t1; \ + MULLD r0, h2, t2; \ + ADDZE t5; \ + MULHDU r1, h0, t4; \ + MULLD r1, h0, h0; \ + ADD t5, t2, t2; \ + ADDC h0, t1, t1; \ + MULLD h2, r1, t3; \ + ADDZE t4, h0; \ + MULHDU r1, h1, t5; \ + MULLD r1, h1, t4; \ + ADDC t4, t2, t2; \ + ADDE t5, t3, t3; \ + ADDC h0, t2, t2; \ + MOVD $-4, t4; \ + MOVD t0, h0; \ + MOVD t1, h1; \ + ADDZE t3; \ + ANDCC $3, t2, h2; \ + AND t2, t4, t0; \ + ADDC t0, h0, h0; \ + ADDE t3, h1, h1; \ + SLD $62, t3, t4; \ + SRD $2, t2; \ + ADDZE h2; \ + OR t4, t2, t2; \ + SRD $2, t3; \ + ADDC t2, h0, h0; \ + ADDE t3, h1, h1; \ + ADDZE h2 + +DATA ·poly1305Mask<>+0x00(SB)/8, $0x0FFFFFFC0FFFFFFF +DATA ·poly1305Mask<>+0x08(SB)/8, $0x0FFFFFFC0FFFFFFC +GLOBL ·poly1305Mask<>(SB), RODATA, $16 + +// func update(state *[7]uint64, msg []byte) + +TEXT ·update(SB), $0-32 + MOVD state+0(FP), R3 + MOVD msg_base+8(FP), R4 + MOVD msg_len+16(FP), R5 + + MOVD 0(R3), R8 // h0 + MOVD 8(R3), R9 // h1 + MOVD 16(R3), R10 // h2 + MOVD 24(R3), R11 // r0 + MOVD 32(R3), R12 // r1 + + CMP R5, $16 + BLT bytes_between_0_and_15 + +loop: + POLY1305_ADD(R4, R8, R9, R10, R20, R21, R22) + +multiply: + POLY1305_MUL(R8, R9, R10, R11, R12, R16, R17, R18, R14, R20, R21) + ADD $-16, R5 + CMP R5, $16 + BGE loop + +bytes_between_0_and_15: + CMP $0, R5 + BEQ done + MOVD $0, R16 // h0 + MOVD $0, R17 // h1 + +flush_buffer: + CMP R5, $8 + BLE just1 + + MOVD $8, R21 + SUB R21, R5, R21 + + // Greater than 8 -- load the rightmost remaining bytes in msg + // and put into R17 (h1) + MOVD (R4)(R21), R17 + MOVD $16, R22 + + // Find the offset to those bytes + SUB R5, R22, R22 + SLD $3, R22 + + // Shift to get only the bytes in msg + SRD R22, R17, R17 + + // Put 1 at high end + MOVD $1, R23 + SLD $3, R21 + SLD R21, R23, R23 + OR R23, R17, R17 + + // Remainder is 8 + MOVD $8, R5 + +just1: + CMP R5, $8 + BLT less8 + + // Exactly 8 + MOVD (R4), R16 + + CMP $0, R17 + + // Check if we've already set R17; if not + // set 1 to indicate end of msg. + BNE carry + MOVD $1, R17 + BR carry + +less8: + MOVD $0, R16 // h0 + MOVD $0, R22 // shift count + CMP R5, $4 + BLT less4 + MOVWZ (R4), R16 + ADD $4, R4 + ADD $-4, R5 + MOVD $32, R22 + +less4: + CMP R5, $2 + BLT less2 + MOVHZ (R4), R21 + SLD R22, R21, R21 + OR R16, R21, R16 + ADD $16, R22 + ADD $-2, R5 + ADD $2, R4 + +less2: + CMP $0, R5 + BEQ insert1 + MOVBZ (R4), R21 + SLD R22, R21, R21 + OR R16, R21, R16 + ADD $8, R22 + +insert1: + // Insert 1 at end of msg + MOVD $1, R21 + SLD R22, R21, R21 + OR R16, R21, R16 + +carry: + // Add new values to h0, h1, h2 + ADDC R16, R8 + ADDE R17, R9 + ADDE $0, R10 + MOVD $16, R5 + ADD R5, R4 + BR multiply + +done: + // Save h0, h1, h2 in state + MOVD R8, 0(R3) + MOVD R9, 8(R3) + MOVD R10, 16(R3) + RET + +// func initialize(state *[7]uint64, key *[32]byte) +TEXT ·initialize(SB), $0-16 + MOVD state+0(FP), R3 + MOVD key+8(FP), R4 + + // state[0...7] is initialized with zero + // Load key + MOVD 0(R4), R5 + MOVD 8(R4), R6 + MOVD 16(R4), R7 + MOVD 24(R4), R8 + + // Address of key mask + MOVD $·poly1305Mask<>(SB), R9 + + // Save original key in state + MOVD R7, 40(R3) + MOVD R8, 48(R3) + + // Get mask + MOVD (R9), R7 + MOVD 8(R9), R8 + + // And with key + AND R5, R7, R5 + AND R6, R8, R6 + + // Save masked key in state + MOVD R5, 24(R3) + MOVD R6, 32(R3) + RET + +// func finalize(tag *[TagSize]byte, state *[7]uint64) +TEXT ·finalize(SB), $0-16 + MOVD tag+0(FP), R3 + MOVD state+8(FP), R4 + + // Get h0, h1, h2 from state + MOVD 0(R4), R5 + MOVD 8(R4), R6 + MOVD 16(R4), R7 + + // Save h0, h1 + MOVD R5, R8 + MOVD R6, R9 + MOVD $3, R20 + MOVD $-1, R21 + SUBC $-5, R5 + SUBE R21, R6 + SUBE R20, R7 + MOVD $0, R21 + SUBZE R21 + + // Check for carry + CMP $0, R21 + ISEL $2, R5, R8, R5 + ISEL $2, R6, R9, R6 + MOVD 40(R4), R8 + MOVD 48(R4), R9 + ADDC R8, R5 + ADDE R9, R6 + MOVD R5, 0(R3) + MOVD R6, 8(R3) + RET diff --git a/vendor/golang.org/x/crypto/poly1305/sum_s390x.go b/vendor/golang.org/x/crypto/poly1305/sum_s390x.go new file mode 100644 index 0000000..ec99e07 --- /dev/null +++ b/vendor/golang.org/x/crypto/poly1305/sum_s390x.go @@ -0,0 +1,42 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build s390x,go1.11,!gccgo,!appengine + +package poly1305 + +import ( + "golang.org/x/sys/cpu" +) + +// poly1305vx is an assembly implementation of Poly1305 that uses vector +// instructions. It must only be called if the vector facility (vx) is +// available. +//go:noescape +func poly1305vx(out *[16]byte, m *byte, mlen uint64, key *[32]byte) + +// poly1305vmsl is an assembly implementation of Poly1305 that uses vector +// instructions, including VMSL. It must only be called if the vector facility (vx) is +// available and if VMSL is supported. +//go:noescape +func poly1305vmsl(out *[16]byte, m *byte, mlen uint64, key *[32]byte) + +// Sum generates an authenticator for m using a one-time key and puts the +// 16-byte result into out. Authenticating two different messages with the same +// key allows an attacker to forge messages at will. +func Sum(out *[16]byte, m []byte, key *[32]byte) { + if cpu.S390X.HasVX { + var mPtr *byte + if len(m) > 0 { + mPtr = &m[0] + } + if cpu.S390X.HasVXE && len(m) > 256 { + poly1305vmsl(out, mPtr, uint64(len(m)), key) + } else { + poly1305vx(out, mPtr, uint64(len(m)), key) + } + } else { + sumGeneric(out, m, key) + } +} diff --git a/vendor/golang.org/x/crypto/poly1305/sum_s390x.s b/vendor/golang.org/x/crypto/poly1305/sum_s390x.s new file mode 100644 index 0000000..ca5a309 --- /dev/null +++ b/vendor/golang.org/x/crypto/poly1305/sum_s390x.s @@ -0,0 +1,378 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build s390x,go1.11,!gccgo,!appengine + +#include "textflag.h" + +// Implementation of Poly1305 using the vector facility (vx). + +// constants +#define MOD26 V0 +#define EX0 V1 +#define EX1 V2 +#define EX2 V3 + +// temporaries +#define T_0 V4 +#define T_1 V5 +#define T_2 V6 +#define T_3 V7 +#define T_4 V8 + +// key (r) +#define R_0 V9 +#define R_1 V10 +#define R_2 V11 +#define R_3 V12 +#define R_4 V13 +#define R5_1 V14 +#define R5_2 V15 +#define R5_3 V16 +#define R5_4 V17 +#define RSAVE_0 R5 +#define RSAVE_1 R6 +#define RSAVE_2 R7 +#define RSAVE_3 R8 +#define RSAVE_4 R9 +#define R5SAVE_1 V28 +#define R5SAVE_2 V29 +#define R5SAVE_3 V30 +#define R5SAVE_4 V31 + +// message block +#define F_0 V18 +#define F_1 V19 +#define F_2 V20 +#define F_3 V21 +#define F_4 V22 + +// accumulator +#define H_0 V23 +#define H_1 V24 +#define H_2 V25 +#define H_3 V26 +#define H_4 V27 + +GLOBL ·keyMask<>(SB), RODATA, $16 +DATA ·keyMask<>+0(SB)/8, $0xffffff0ffcffff0f +DATA ·keyMask<>+8(SB)/8, $0xfcffff0ffcffff0f + +GLOBL ·bswapMask<>(SB), RODATA, $16 +DATA ·bswapMask<>+0(SB)/8, $0x0f0e0d0c0b0a0908 +DATA ·bswapMask<>+8(SB)/8, $0x0706050403020100 + +GLOBL ·constants<>(SB), RODATA, $64 +// MOD26 +DATA ·constants<>+0(SB)/8, $0x3ffffff +DATA ·constants<>+8(SB)/8, $0x3ffffff +// EX0 +DATA ·constants<>+16(SB)/8, $0x0006050403020100 +DATA ·constants<>+24(SB)/8, $0x1016151413121110 +// EX1 +DATA ·constants<>+32(SB)/8, $0x060c0b0a09080706 +DATA ·constants<>+40(SB)/8, $0x161c1b1a19181716 +// EX2 +DATA ·constants<>+48(SB)/8, $0x0d0d0d0d0d0f0e0d +DATA ·constants<>+56(SB)/8, $0x1d1d1d1d1d1f1e1d + +// h = (f*g) % (2**130-5) [partial reduction] +#define MULTIPLY(f0, f1, f2, f3, f4, g0, g1, g2, g3, g4, g51, g52, g53, g54, h0, h1, h2, h3, h4) \ + VMLOF f0, g0, h0 \ + VMLOF f0, g1, h1 \ + VMLOF f0, g2, h2 \ + VMLOF f0, g3, h3 \ + VMLOF f0, g4, h4 \ + VMLOF f1, g54, T_0 \ + VMLOF f1, g0, T_1 \ + VMLOF f1, g1, T_2 \ + VMLOF f1, g2, T_3 \ + VMLOF f1, g3, T_4 \ + VMALOF f2, g53, h0, h0 \ + VMALOF f2, g54, h1, h1 \ + VMALOF f2, g0, h2, h2 \ + VMALOF f2, g1, h3, h3 \ + VMALOF f2, g2, h4, h4 \ + VMALOF f3, g52, T_0, T_0 \ + VMALOF f3, g53, T_1, T_1 \ + VMALOF f3, g54, T_2, T_2 \ + VMALOF f3, g0, T_3, T_3 \ + VMALOF f3, g1, T_4, T_4 \ + VMALOF f4, g51, h0, h0 \ + VMALOF f4, g52, h1, h1 \ + VMALOF f4, g53, h2, h2 \ + VMALOF f4, g54, h3, h3 \ + VMALOF f4, g0, h4, h4 \ + VAG T_0, h0, h0 \ + VAG T_1, h1, h1 \ + VAG T_2, h2, h2 \ + VAG T_3, h3, h3 \ + VAG T_4, h4, h4 + +// carry h0->h1 h3->h4, h1->h2 h4->h0, h0->h1 h2->h3, h3->h4 +#define REDUCE(h0, h1, h2, h3, h4) \ + VESRLG $26, h0, T_0 \ + VESRLG $26, h3, T_1 \ + VN MOD26, h0, h0 \ + VN MOD26, h3, h3 \ + VAG T_0, h1, h1 \ + VAG T_1, h4, h4 \ + VESRLG $26, h1, T_2 \ + VESRLG $26, h4, T_3 \ + VN MOD26, h1, h1 \ + VN MOD26, h4, h4 \ + VESLG $2, T_3, T_4 \ + VAG T_3, T_4, T_4 \ + VAG T_2, h2, h2 \ + VAG T_4, h0, h0 \ + VESRLG $26, h2, T_0 \ + VESRLG $26, h0, T_1 \ + VN MOD26, h2, h2 \ + VN MOD26, h0, h0 \ + VAG T_0, h3, h3 \ + VAG T_1, h1, h1 \ + VESRLG $26, h3, T_2 \ + VN MOD26, h3, h3 \ + VAG T_2, h4, h4 + +// expand in0 into d[0] and in1 into d[1] +#define EXPAND(in0, in1, d0, d1, d2, d3, d4) \ + VGBM $0x0707, d1 \ // d1=tmp + VPERM in0, in1, EX2, d4 \ + VPERM in0, in1, EX0, d0 \ + VPERM in0, in1, EX1, d2 \ + VN d1, d4, d4 \ + VESRLG $26, d0, d1 \ + VESRLG $30, d2, d3 \ + VESRLG $4, d2, d2 \ + VN MOD26, d0, d0 \ + VN MOD26, d1, d1 \ + VN MOD26, d2, d2 \ + VN MOD26, d3, d3 + +// pack h4:h0 into h1:h0 (no carry) +#define PACK(h0, h1, h2, h3, h4) \ + VESLG $26, h1, h1 \ + VESLG $26, h3, h3 \ + VO h0, h1, h0 \ + VO h2, h3, h2 \ + VESLG $4, h2, h2 \ + VLEIB $7, $48, h1 \ + VSLB h1, h2, h2 \ + VO h0, h2, h0 \ + VLEIB $7, $104, h1 \ + VSLB h1, h4, h3 \ + VO h3, h0, h0 \ + VLEIB $7, $24, h1 \ + VSRLB h1, h4, h1 + +// if h > 2**130-5 then h -= 2**130-5 +#define MOD(h0, h1, t0, t1, t2) \ + VZERO t0 \ + VLEIG $1, $5, t0 \ + VACCQ h0, t0, t1 \ + VAQ h0, t0, t0 \ + VONE t2 \ + VLEIG $1, $-4, t2 \ + VAQ t2, t1, t1 \ + VACCQ h1, t1, t1 \ + VONE t2 \ + VAQ t2, t1, t1 \ + VN h0, t1, t2 \ + VNC t0, t1, t1 \ + VO t1, t2, h0 + +// func poly1305vx(out *[16]byte, m *byte, mlen uint64, key *[32]key) +TEXT ·poly1305vx(SB), $0-32 + // This code processes up to 2 blocks (32 bytes) per iteration + // using the algorithm described in: + // NEON crypto, Daniel J. Bernstein & Peter Schwabe + // https://cryptojedi.org/papers/neoncrypto-20120320.pdf + LMG out+0(FP), R1, R4 // R1=out, R2=m, R3=mlen, R4=key + + // load MOD26, EX0, EX1 and EX2 + MOVD $·constants<>(SB), R5 + VLM (R5), MOD26, EX2 + + // setup r + VL (R4), T_0 + MOVD $·keyMask<>(SB), R6 + VL (R6), T_1 + VN T_0, T_1, T_0 + EXPAND(T_0, T_0, R_0, R_1, R_2, R_3, R_4) + + // setup r*5 + VLEIG $0, $5, T_0 + VLEIG $1, $5, T_0 + + // store r (for final block) + VMLOF T_0, R_1, R5SAVE_1 + VMLOF T_0, R_2, R5SAVE_2 + VMLOF T_0, R_3, R5SAVE_3 + VMLOF T_0, R_4, R5SAVE_4 + VLGVG $0, R_0, RSAVE_0 + VLGVG $0, R_1, RSAVE_1 + VLGVG $0, R_2, RSAVE_2 + VLGVG $0, R_3, RSAVE_3 + VLGVG $0, R_4, RSAVE_4 + + // skip r**2 calculation + CMPBLE R3, $16, skip + + // calculate r**2 + MULTIPLY(R_0, R_1, R_2, R_3, R_4, R_0, R_1, R_2, R_3, R_4, R5SAVE_1, R5SAVE_2, R5SAVE_3, R5SAVE_4, H_0, H_1, H_2, H_3, H_4) + REDUCE(H_0, H_1, H_2, H_3, H_4) + VLEIG $0, $5, T_0 + VLEIG $1, $5, T_0 + VMLOF T_0, H_1, R5_1 + VMLOF T_0, H_2, R5_2 + VMLOF T_0, H_3, R5_3 + VMLOF T_0, H_4, R5_4 + VLR H_0, R_0 + VLR H_1, R_1 + VLR H_2, R_2 + VLR H_3, R_3 + VLR H_4, R_4 + + // initialize h + VZERO H_0 + VZERO H_1 + VZERO H_2 + VZERO H_3 + VZERO H_4 + +loop: + CMPBLE R3, $32, b2 + VLM (R2), T_0, T_1 + SUB $32, R3 + MOVD $32(R2), R2 + EXPAND(T_0, T_1, F_0, F_1, F_2, F_3, F_4) + VLEIB $4, $1, F_4 + VLEIB $12, $1, F_4 + +multiply: + VAG H_0, F_0, F_0 + VAG H_1, F_1, F_1 + VAG H_2, F_2, F_2 + VAG H_3, F_3, F_3 + VAG H_4, F_4, F_4 + MULTIPLY(F_0, F_1, F_2, F_3, F_4, R_0, R_1, R_2, R_3, R_4, R5_1, R5_2, R5_3, R5_4, H_0, H_1, H_2, H_3, H_4) + REDUCE(H_0, H_1, H_2, H_3, H_4) + CMPBNE R3, $0, loop + +finish: + // sum vectors + VZERO T_0 + VSUMQG H_0, T_0, H_0 + VSUMQG H_1, T_0, H_1 + VSUMQG H_2, T_0, H_2 + VSUMQG H_3, T_0, H_3 + VSUMQG H_4, T_0, H_4 + + // h may be >= 2*(2**130-5) so we need to reduce it again + REDUCE(H_0, H_1, H_2, H_3, H_4) + + // carry h1->h4 + VESRLG $26, H_1, T_1 + VN MOD26, H_1, H_1 + VAQ T_1, H_2, H_2 + VESRLG $26, H_2, T_2 + VN MOD26, H_2, H_2 + VAQ T_2, H_3, H_3 + VESRLG $26, H_3, T_3 + VN MOD26, H_3, H_3 + VAQ T_3, H_4, H_4 + + // h is now < 2*(2**130-5) + // pack h into h1 (hi) and h0 (lo) + PACK(H_0, H_1, H_2, H_3, H_4) + + // if h > 2**130-5 then h -= 2**130-5 + MOD(H_0, H_1, T_0, T_1, T_2) + + // h += s + MOVD $·bswapMask<>(SB), R5 + VL (R5), T_1 + VL 16(R4), T_0 + VPERM T_0, T_0, T_1, T_0 // reverse bytes (to big) + VAQ T_0, H_0, H_0 + VPERM H_0, H_0, T_1, H_0 // reverse bytes (to little) + VST H_0, (R1) + + RET + +b2: + CMPBLE R3, $16, b1 + + // 2 blocks remaining + SUB $17, R3 + VL (R2), T_0 + VLL R3, 16(R2), T_1 + ADD $1, R3 + MOVBZ $1, R0 + CMPBEQ R3, $16, 2(PC) + VLVGB R3, R0, T_1 + EXPAND(T_0, T_1, F_0, F_1, F_2, F_3, F_4) + CMPBNE R3, $16, 2(PC) + VLEIB $12, $1, F_4 + VLEIB $4, $1, F_4 + + // setup [r²,r] + VLVGG $1, RSAVE_0, R_0 + VLVGG $1, RSAVE_1, R_1 + VLVGG $1, RSAVE_2, R_2 + VLVGG $1, RSAVE_3, R_3 + VLVGG $1, RSAVE_4, R_4 + VPDI $0, R5_1, R5SAVE_1, R5_1 + VPDI $0, R5_2, R5SAVE_2, R5_2 + VPDI $0, R5_3, R5SAVE_3, R5_3 + VPDI $0, R5_4, R5SAVE_4, R5_4 + + MOVD $0, R3 + BR multiply + +skip: + VZERO H_0 + VZERO H_1 + VZERO H_2 + VZERO H_3 + VZERO H_4 + + CMPBEQ R3, $0, finish + +b1: + // 1 block remaining + SUB $1, R3 + VLL R3, (R2), T_0 + ADD $1, R3 + MOVBZ $1, R0 + CMPBEQ R3, $16, 2(PC) + VLVGB R3, R0, T_0 + VZERO T_1 + EXPAND(T_0, T_1, F_0, F_1, F_2, F_3, F_4) + CMPBNE R3, $16, 2(PC) + VLEIB $4, $1, F_4 + VLEIG $1, $1, R_0 + VZERO R_1 + VZERO R_2 + VZERO R_3 + VZERO R_4 + VZERO R5_1 + VZERO R5_2 + VZERO R5_3 + VZERO R5_4 + + // setup [r, 1] + VLVGG $0, RSAVE_0, R_0 + VLVGG $0, RSAVE_1, R_1 + VLVGG $0, RSAVE_2, R_2 + VLVGG $0, RSAVE_3, R_3 + VLVGG $0, RSAVE_4, R_4 + VPDI $0, R5SAVE_1, R5_1, R5_1 + VPDI $0, R5SAVE_2, R5_2, R5_2 + VPDI $0, R5SAVE_3, R5_3, R5_3 + VPDI $0, R5SAVE_4, R5_4, R5_4 + + MOVD $0, R3 + BR multiply diff --git a/vendor/golang.org/x/crypto/poly1305/sum_vmsl_s390x.s b/vendor/golang.org/x/crypto/poly1305/sum_vmsl_s390x.s new file mode 100644 index 0000000..e60bbc1 --- /dev/null +++ b/vendor/golang.org/x/crypto/poly1305/sum_vmsl_s390x.s @@ -0,0 +1,909 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build s390x,go1.11,!gccgo,!appengine + +#include "textflag.h" + +// Implementation of Poly1305 using the vector facility (vx) and the VMSL instruction. + +// constants +#define EX0 V1 +#define EX1 V2 +#define EX2 V3 + +// temporaries +#define T_0 V4 +#define T_1 V5 +#define T_2 V6 +#define T_3 V7 +#define T_4 V8 +#define T_5 V9 +#define T_6 V10 +#define T_7 V11 +#define T_8 V12 +#define T_9 V13 +#define T_10 V14 + +// r**2 & r**4 +#define R_0 V15 +#define R_1 V16 +#define R_2 V17 +#define R5_1 V18 +#define R5_2 V19 +// key (r) +#define RSAVE_0 R7 +#define RSAVE_1 R8 +#define RSAVE_2 R9 +#define R5SAVE_1 R10 +#define R5SAVE_2 R11 + +// message block +#define M0 V20 +#define M1 V21 +#define M2 V22 +#define M3 V23 +#define M4 V24 +#define M5 V25 + +// accumulator +#define H0_0 V26 +#define H1_0 V27 +#define H2_0 V28 +#define H0_1 V29 +#define H1_1 V30 +#define H2_1 V31 + +GLOBL ·keyMask<>(SB), RODATA, $16 +DATA ·keyMask<>+0(SB)/8, $0xffffff0ffcffff0f +DATA ·keyMask<>+8(SB)/8, $0xfcffff0ffcffff0f + +GLOBL ·bswapMask<>(SB), RODATA, $16 +DATA ·bswapMask<>+0(SB)/8, $0x0f0e0d0c0b0a0908 +DATA ·bswapMask<>+8(SB)/8, $0x0706050403020100 + +GLOBL ·constants<>(SB), RODATA, $48 +// EX0 +DATA ·constants<>+0(SB)/8, $0x18191a1b1c1d1e1f +DATA ·constants<>+8(SB)/8, $0x0000050403020100 +// EX1 +DATA ·constants<>+16(SB)/8, $0x18191a1b1c1d1e1f +DATA ·constants<>+24(SB)/8, $0x00000a0908070605 +// EX2 +DATA ·constants<>+32(SB)/8, $0x18191a1b1c1d1e1f +DATA ·constants<>+40(SB)/8, $0x0000000f0e0d0c0b + +GLOBL ·c<>(SB), RODATA, $48 +// EX0 +DATA ·c<>+0(SB)/8, $0x0000050403020100 +DATA ·c<>+8(SB)/8, $0x0000151413121110 +// EX1 +DATA ·c<>+16(SB)/8, $0x00000a0908070605 +DATA ·c<>+24(SB)/8, $0x00001a1918171615 +// EX2 +DATA ·c<>+32(SB)/8, $0x0000000f0e0d0c0b +DATA ·c<>+40(SB)/8, $0x0000001f1e1d1c1b + +GLOBL ·reduce<>(SB), RODATA, $32 +// 44 bit +DATA ·reduce<>+0(SB)/8, $0x0 +DATA ·reduce<>+8(SB)/8, $0xfffffffffff +// 42 bit +DATA ·reduce<>+16(SB)/8, $0x0 +DATA ·reduce<>+24(SB)/8, $0x3ffffffffff + +// h = (f*g) % (2**130-5) [partial reduction] +// uses T_0...T_9 temporary registers +// input: m02_0, m02_1, m02_2, m13_0, m13_1, m13_2, r_0, r_1, r_2, r5_1, r5_2, m4_0, m4_1, m4_2, m5_0, m5_1, m5_2 +// temp: t0, t1, t2, t3, t4, t5, t6, t7, t8, t9 +// output: m02_0, m02_1, m02_2, m13_0, m13_1, m13_2 +#define MULTIPLY(m02_0, m02_1, m02_2, m13_0, m13_1, m13_2, r_0, r_1, r_2, r5_1, r5_2, m4_0, m4_1, m4_2, m5_0, m5_1, m5_2, t0, t1, t2, t3, t4, t5, t6, t7, t8, t9) \ + \ // Eliminate the dependency for the last 2 VMSLs + VMSLG m02_0, r_2, m4_2, m4_2 \ + VMSLG m13_0, r_2, m5_2, m5_2 \ // 8 VMSLs pipelined + VMSLG m02_0, r_0, m4_0, m4_0 \ + VMSLG m02_1, r5_2, V0, T_0 \ + VMSLG m02_0, r_1, m4_1, m4_1 \ + VMSLG m02_1, r_0, V0, T_1 \ + VMSLG m02_1, r_1, V0, T_2 \ + VMSLG m02_2, r5_1, V0, T_3 \ + VMSLG m02_2, r5_2, V0, T_4 \ + VMSLG m13_0, r_0, m5_0, m5_0 \ + VMSLG m13_1, r5_2, V0, T_5 \ + VMSLG m13_0, r_1, m5_1, m5_1 \ + VMSLG m13_1, r_0, V0, T_6 \ + VMSLG m13_1, r_1, V0, T_7 \ + VMSLG m13_2, r5_1, V0, T_8 \ + VMSLG m13_2, r5_2, V0, T_9 \ + VMSLG m02_2, r_0, m4_2, m4_2 \ + VMSLG m13_2, r_0, m5_2, m5_2 \ + VAQ m4_0, T_0, m02_0 \ + VAQ m4_1, T_1, m02_1 \ + VAQ m5_0, T_5, m13_0 \ + VAQ m5_1, T_6, m13_1 \ + VAQ m02_0, T_3, m02_0 \ + VAQ m02_1, T_4, m02_1 \ + VAQ m13_0, T_8, m13_0 \ + VAQ m13_1, T_9, m13_1 \ + VAQ m4_2, T_2, m02_2 \ + VAQ m5_2, T_7, m13_2 \ + +// SQUARE uses three limbs of r and r_2*5 to output square of r +// uses T_1, T_5 and T_7 temporary registers +// input: r_0, r_1, r_2, r5_2 +// temp: TEMP0, TEMP1, TEMP2 +// output: p0, p1, p2 +#define SQUARE(r_0, r_1, r_2, r5_2, p0, p1, p2, TEMP0, TEMP1, TEMP2) \ + VMSLG r_0, r_0, p0, p0 \ + VMSLG r_1, r5_2, V0, TEMP0 \ + VMSLG r_2, r5_2, p1, p1 \ + VMSLG r_0, r_1, V0, TEMP1 \ + VMSLG r_1, r_1, p2, p2 \ + VMSLG r_0, r_2, V0, TEMP2 \ + VAQ TEMP0, p0, p0 \ + VAQ TEMP1, p1, p1 \ + VAQ TEMP2, p2, p2 \ + VAQ TEMP0, p0, p0 \ + VAQ TEMP1, p1, p1 \ + VAQ TEMP2, p2, p2 \ + +// carry h0->h1->h2->h0 || h3->h4->h5->h3 +// uses T_2, T_4, T_5, T_7, T_8, T_9 +// t6, t7, t8, t9, t10, t11 +// input: h0, h1, h2, h3, h4, h5 +// temp: t0, t1, t2, t3, t4, t5, t6, t7, t8, t9, t10, t11 +// output: h0, h1, h2, h3, h4, h5 +#define REDUCE(h0, h1, h2, h3, h4, h5, t0, t1, t2, t3, t4, t5, t6, t7, t8, t9, t10, t11) \ + VLM (R12), t6, t7 \ // 44 and 42 bit clear mask + VLEIB $7, $0x28, t10 \ // 5 byte shift mask + VREPIB $4, t8 \ // 4 bit shift mask + VREPIB $2, t11 \ // 2 bit shift mask + VSRLB t10, h0, t0 \ // h0 byte shift + VSRLB t10, h1, t1 \ // h1 byte shift + VSRLB t10, h2, t2 \ // h2 byte shift + VSRLB t10, h3, t3 \ // h3 byte shift + VSRLB t10, h4, t4 \ // h4 byte shift + VSRLB t10, h5, t5 \ // h5 byte shift + VSRL t8, t0, t0 \ // h0 bit shift + VSRL t8, t1, t1 \ // h2 bit shift + VSRL t11, t2, t2 \ // h2 bit shift + VSRL t8, t3, t3 \ // h3 bit shift + VSRL t8, t4, t4 \ // h4 bit shift + VESLG $2, t2, t9 \ // h2 carry x5 + VSRL t11, t5, t5 \ // h5 bit shift + VN t6, h0, h0 \ // h0 clear carry + VAQ t2, t9, t2 \ // h2 carry x5 + VESLG $2, t5, t9 \ // h5 carry x5 + VN t6, h1, h1 \ // h1 clear carry + VN t7, h2, h2 \ // h2 clear carry + VAQ t5, t9, t5 \ // h5 carry x5 + VN t6, h3, h3 \ // h3 clear carry + VN t6, h4, h4 \ // h4 clear carry + VN t7, h5, h5 \ // h5 clear carry + VAQ t0, h1, h1 \ // h0->h1 + VAQ t3, h4, h4 \ // h3->h4 + VAQ t1, h2, h2 \ // h1->h2 + VAQ t4, h5, h5 \ // h4->h5 + VAQ t2, h0, h0 \ // h2->h0 + VAQ t5, h3, h3 \ // h5->h3 + VREPG $1, t6, t6 \ // 44 and 42 bit masks across both halves + VREPG $1, t7, t7 \ + VSLDB $8, h0, h0, h0 \ // set up [h0/1/2, h3/4/5] + VSLDB $8, h1, h1, h1 \ + VSLDB $8, h2, h2, h2 \ + VO h0, h3, h3 \ + VO h1, h4, h4 \ + VO h2, h5, h5 \ + VESRLG $44, h3, t0 \ // 44 bit shift right + VESRLG $44, h4, t1 \ + VESRLG $42, h5, t2 \ + VN t6, h3, h3 \ // clear carry bits + VN t6, h4, h4 \ + VN t7, h5, h5 \ + VESLG $2, t2, t9 \ // multiply carry by 5 + VAQ t9, t2, t2 \ + VAQ t0, h4, h4 \ + VAQ t1, h5, h5 \ + VAQ t2, h3, h3 \ + +// carry h0->h1->h2->h0 +// input: h0, h1, h2 +// temp: t0, t1, t2, t3, t4, t5, t6, t7, t8 +// output: h0, h1, h2 +#define REDUCE2(h0, h1, h2, t0, t1, t2, t3, t4, t5, t6, t7, t8) \ + VLEIB $7, $0x28, t3 \ // 5 byte shift mask + VREPIB $4, t4 \ // 4 bit shift mask + VREPIB $2, t7 \ // 2 bit shift mask + VGBM $0x003F, t5 \ // mask to clear carry bits + VSRLB t3, h0, t0 \ + VSRLB t3, h1, t1 \ + VSRLB t3, h2, t2 \ + VESRLG $4, t5, t5 \ // 44 bit clear mask + VSRL t4, t0, t0 \ + VSRL t4, t1, t1 \ + VSRL t7, t2, t2 \ + VESRLG $2, t5, t6 \ // 42 bit clear mask + VESLG $2, t2, t8 \ + VAQ t8, t2, t2 \ + VN t5, h0, h0 \ + VN t5, h1, h1 \ + VN t6, h2, h2 \ + VAQ t0, h1, h1 \ + VAQ t1, h2, h2 \ + VAQ t2, h0, h0 \ + VSRLB t3, h0, t0 \ + VSRLB t3, h1, t1 \ + VSRLB t3, h2, t2 \ + VSRL t4, t0, t0 \ + VSRL t4, t1, t1 \ + VSRL t7, t2, t2 \ + VN t5, h0, h0 \ + VN t5, h1, h1 \ + VESLG $2, t2, t8 \ + VN t6, h2, h2 \ + VAQ t0, h1, h1 \ + VAQ t8, t2, t2 \ + VAQ t1, h2, h2 \ + VAQ t2, h0, h0 \ + +// expands two message blocks into the lower halfs of the d registers +// moves the contents of the d registers into upper halfs +// input: in1, in2, d0, d1, d2, d3, d4, d5 +// temp: TEMP0, TEMP1, TEMP2, TEMP3 +// output: d0, d1, d2, d3, d4, d5 +#define EXPACC(in1, in2, d0, d1, d2, d3, d4, d5, TEMP0, TEMP1, TEMP2, TEMP3) \ + VGBM $0xff3f, TEMP0 \ + VGBM $0xff1f, TEMP1 \ + VESLG $4, d1, TEMP2 \ + VESLG $4, d4, TEMP3 \ + VESRLG $4, TEMP0, TEMP0 \ + VPERM in1, d0, EX0, d0 \ + VPERM in2, d3, EX0, d3 \ + VPERM in1, d2, EX2, d2 \ + VPERM in2, d5, EX2, d5 \ + VPERM in1, TEMP2, EX1, d1 \ + VPERM in2, TEMP3, EX1, d4 \ + VN TEMP0, d0, d0 \ + VN TEMP0, d3, d3 \ + VESRLG $4, d1, d1 \ + VESRLG $4, d4, d4 \ + VN TEMP1, d2, d2 \ + VN TEMP1, d5, d5 \ + VN TEMP0, d1, d1 \ + VN TEMP0, d4, d4 \ + +// expands one message block into the lower halfs of the d registers +// moves the contents of the d registers into upper halfs +// input: in, d0, d1, d2 +// temp: TEMP0, TEMP1, TEMP2 +// output: d0, d1, d2 +#define EXPACC2(in, d0, d1, d2, TEMP0, TEMP1, TEMP2) \ + VGBM $0xff3f, TEMP0 \ + VESLG $4, d1, TEMP2 \ + VGBM $0xff1f, TEMP1 \ + VPERM in, d0, EX0, d0 \ + VESRLG $4, TEMP0, TEMP0 \ + VPERM in, d2, EX2, d2 \ + VPERM in, TEMP2, EX1, d1 \ + VN TEMP0, d0, d0 \ + VN TEMP1, d2, d2 \ + VESRLG $4, d1, d1 \ + VN TEMP0, d1, d1 \ + +// pack h2:h0 into h1:h0 (no carry) +// input: h0, h1, h2 +// output: h0, h1, h2 +#define PACK(h0, h1, h2) \ + VMRLG h1, h2, h2 \ // copy h1 to upper half h2 + VESLG $44, h1, h1 \ // shift limb 1 44 bits, leaving 20 + VO h0, h1, h0 \ // combine h0 with 20 bits from limb 1 + VESRLG $20, h2, h1 \ // put top 24 bits of limb 1 into h1 + VLEIG $1, $0, h1 \ // clear h2 stuff from lower half of h1 + VO h0, h1, h0 \ // h0 now has 88 bits (limb 0 and 1) + VLEIG $0, $0, h2 \ // clear upper half of h2 + VESRLG $40, h2, h1 \ // h1 now has upper two bits of result + VLEIB $7, $88, h1 \ // for byte shift (11 bytes) + VSLB h1, h2, h2 \ // shift h2 11 bytes to the left + VO h0, h2, h0 \ // combine h0 with 20 bits from limb 1 + VLEIG $0, $0, h1 \ // clear upper half of h1 + +// if h > 2**130-5 then h -= 2**130-5 +// input: h0, h1 +// temp: t0, t1, t2 +// output: h0 +#define MOD(h0, h1, t0, t1, t2) \ + VZERO t0 \ + VLEIG $1, $5, t0 \ + VACCQ h0, t0, t1 \ + VAQ h0, t0, t0 \ + VONE t2 \ + VLEIG $1, $-4, t2 \ + VAQ t2, t1, t1 \ + VACCQ h1, t1, t1 \ + VONE t2 \ + VAQ t2, t1, t1 \ + VN h0, t1, t2 \ + VNC t0, t1, t1 \ + VO t1, t2, h0 \ + +// func poly1305vmsl(out *[16]byte, m *byte, mlen uint64, key *[32]key) +TEXT ·poly1305vmsl(SB), $0-32 + // This code processes 6 + up to 4 blocks (32 bytes) per iteration + // using the algorithm described in: + // NEON crypto, Daniel J. Bernstein & Peter Schwabe + // https://cryptojedi.org/papers/neoncrypto-20120320.pdf + // And as moddified for VMSL as described in + // Accelerating Poly1305 Cryptographic Message Authentication on the z14 + // O'Farrell et al, CASCON 2017, p48-55 + // https://ibm.ent.box.com/s/jf9gedj0e9d2vjctfyh186shaztavnht + + LMG out+0(FP), R1, R4 // R1=out, R2=m, R3=mlen, R4=key + VZERO V0 // c + + // load EX0, EX1 and EX2 + MOVD $·constants<>(SB), R5 + VLM (R5), EX0, EX2 // c + + // setup r + VL (R4), T_0 + MOVD $·keyMask<>(SB), R6 + VL (R6), T_1 + VN T_0, T_1, T_0 + VZERO T_2 // limbs for r + VZERO T_3 + VZERO T_4 + EXPACC2(T_0, T_2, T_3, T_4, T_1, T_5, T_7) + + // T_2, T_3, T_4: [0, r] + + // setup r*20 + VLEIG $0, $0, T_0 + VLEIG $1, $20, T_0 // T_0: [0, 20] + VZERO T_5 + VZERO T_6 + VMSLG T_0, T_3, T_5, T_5 + VMSLG T_0, T_4, T_6, T_6 + + // store r for final block in GR + VLGVG $1, T_2, RSAVE_0 // c + VLGVG $1, T_3, RSAVE_1 // c + VLGVG $1, T_4, RSAVE_2 // c + VLGVG $1, T_5, R5SAVE_1 // c + VLGVG $1, T_6, R5SAVE_2 // c + + // initialize h + VZERO H0_0 + VZERO H1_0 + VZERO H2_0 + VZERO H0_1 + VZERO H1_1 + VZERO H2_1 + + // initialize pointer for reduce constants + MOVD $·reduce<>(SB), R12 + + // calculate r**2 and 20*(r**2) + VZERO R_0 + VZERO R_1 + VZERO R_2 + SQUARE(T_2, T_3, T_4, T_6, R_0, R_1, R_2, T_1, T_5, T_7) + REDUCE2(R_0, R_1, R_2, M0, M1, M2, M3, M4, R5_1, R5_2, M5, T_1) + VZERO R5_1 + VZERO R5_2 + VMSLG T_0, R_1, R5_1, R5_1 + VMSLG T_0, R_2, R5_2, R5_2 + + // skip r**4 calculation if 3 blocks or less + CMPBLE R3, $48, b4 + + // calculate r**4 and 20*(r**4) + VZERO T_8 + VZERO T_9 + VZERO T_10 + SQUARE(R_0, R_1, R_2, R5_2, T_8, T_9, T_10, T_1, T_5, T_7) + REDUCE2(T_8, T_9, T_10, M0, M1, M2, M3, M4, T_2, T_3, M5, T_1) + VZERO T_2 + VZERO T_3 + VMSLG T_0, T_9, T_2, T_2 + VMSLG T_0, T_10, T_3, T_3 + + // put r**2 to the right and r**4 to the left of R_0, R_1, R_2 + VSLDB $8, T_8, T_8, T_8 + VSLDB $8, T_9, T_9, T_9 + VSLDB $8, T_10, T_10, T_10 + VSLDB $8, T_2, T_2, T_2 + VSLDB $8, T_3, T_3, T_3 + + VO T_8, R_0, R_0 + VO T_9, R_1, R_1 + VO T_10, R_2, R_2 + VO T_2, R5_1, R5_1 + VO T_3, R5_2, R5_2 + + CMPBLE R3, $80, load // less than or equal to 5 blocks in message + + // 6(or 5+1) blocks + SUB $81, R3 + VLM (R2), M0, M4 + VLL R3, 80(R2), M5 + ADD $1, R3 + MOVBZ $1, R0 + CMPBGE R3, $16, 2(PC) + VLVGB R3, R0, M5 + MOVD $96(R2), R2 + EXPACC(M0, M1, H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, T_0, T_1, T_2, T_3) + EXPACC(M2, M3, H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, T_0, T_1, T_2, T_3) + VLEIB $2, $1, H2_0 + VLEIB $2, $1, H2_1 + VLEIB $10, $1, H2_0 + VLEIB $10, $1, H2_1 + + VZERO M0 + VZERO M1 + VZERO M2 + VZERO M3 + VZERO T_4 + VZERO T_10 + EXPACC(M4, M5, M0, M1, M2, M3, T_4, T_10, T_0, T_1, T_2, T_3) + VLR T_4, M4 + VLEIB $10, $1, M2 + CMPBLT R3, $16, 2(PC) + VLEIB $10, $1, T_10 + MULTIPLY(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, R_0, R_1, R_2, R5_1, R5_2, M0, M1, M2, M3, M4, T_10, T_0, T_1, T_2, T_3, T_4, T_5, T_6, T_7, T_8, T_9) + REDUCE(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, T_10, M0, M1, M2, M3, M4, T_4, T_5, T_2, T_7, T_8, T_9) + VMRHG V0, H0_1, H0_0 + VMRHG V0, H1_1, H1_0 + VMRHG V0, H2_1, H2_0 + VMRLG V0, H0_1, H0_1 + VMRLG V0, H1_1, H1_1 + VMRLG V0, H2_1, H2_1 + + SUB $16, R3 + CMPBLE R3, $0, square + +load: + // load EX0, EX1 and EX2 + MOVD $·c<>(SB), R5 + VLM (R5), EX0, EX2 + +loop: + CMPBLE R3, $64, add // b4 // last 4 or less blocks left + + // next 4 full blocks + VLM (R2), M2, M5 + SUB $64, R3 + MOVD $64(R2), R2 + REDUCE(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, T_10, M0, M1, T_0, T_1, T_3, T_4, T_5, T_2, T_7, T_8, T_9) + + // expacc in-lined to create [m2, m3] limbs + VGBM $0x3f3f, T_0 // 44 bit clear mask + VGBM $0x1f1f, T_1 // 40 bit clear mask + VPERM M2, M3, EX0, T_3 + VESRLG $4, T_0, T_0 // 44 bit clear mask ready + VPERM M2, M3, EX1, T_4 + VPERM M2, M3, EX2, T_5 + VN T_0, T_3, T_3 + VESRLG $4, T_4, T_4 + VN T_1, T_5, T_5 + VN T_0, T_4, T_4 + VMRHG H0_1, T_3, H0_0 + VMRHG H1_1, T_4, H1_0 + VMRHG H2_1, T_5, H2_0 + VMRLG H0_1, T_3, H0_1 + VMRLG H1_1, T_4, H1_1 + VMRLG H2_1, T_5, H2_1 + VLEIB $10, $1, H2_0 + VLEIB $10, $1, H2_1 + VPERM M4, M5, EX0, T_3 + VPERM M4, M5, EX1, T_4 + VPERM M4, M5, EX2, T_5 + VN T_0, T_3, T_3 + VESRLG $4, T_4, T_4 + VN T_1, T_5, T_5 + VN T_0, T_4, T_4 + VMRHG V0, T_3, M0 + VMRHG V0, T_4, M1 + VMRHG V0, T_5, M2 + VMRLG V0, T_3, M3 + VMRLG V0, T_4, M4 + VMRLG V0, T_5, M5 + VLEIB $10, $1, M2 + VLEIB $10, $1, M5 + + MULTIPLY(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, R_0, R_1, R_2, R5_1, R5_2, M0, M1, M2, M3, M4, M5, T_0, T_1, T_2, T_3, T_4, T_5, T_6, T_7, T_8, T_9) + CMPBNE R3, $0, loop + REDUCE(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, T_10, M0, M1, M3, M4, M5, T_4, T_5, T_2, T_7, T_8, T_9) + VMRHG V0, H0_1, H0_0 + VMRHG V0, H1_1, H1_0 + VMRHG V0, H2_1, H2_0 + VMRLG V0, H0_1, H0_1 + VMRLG V0, H1_1, H1_1 + VMRLG V0, H2_1, H2_1 + + // load EX0, EX1, EX2 + MOVD $·constants<>(SB), R5 + VLM (R5), EX0, EX2 + + // sum vectors + VAQ H0_0, H0_1, H0_0 + VAQ H1_0, H1_1, H1_0 + VAQ H2_0, H2_1, H2_0 + + // h may be >= 2*(2**130-5) so we need to reduce it again + // M0...M4 are used as temps here + REDUCE2(H0_0, H1_0, H2_0, M0, M1, M2, M3, M4, T_9, T_10, H0_1, M5) + +next: // carry h1->h2 + VLEIB $7, $0x28, T_1 + VREPIB $4, T_2 + VGBM $0x003F, T_3 + VESRLG $4, T_3 + + // byte shift + VSRLB T_1, H1_0, T_4 + + // bit shift + VSRL T_2, T_4, T_4 + + // clear h1 carry bits + VN T_3, H1_0, H1_0 + + // add carry + VAQ T_4, H2_0, H2_0 + + // h is now < 2*(2**130-5) + // pack h into h1 (hi) and h0 (lo) + PACK(H0_0, H1_0, H2_0) + + // if h > 2**130-5 then h -= 2**130-5 + MOD(H0_0, H1_0, T_0, T_1, T_2) + + // h += s + MOVD $·bswapMask<>(SB), R5 + VL (R5), T_1 + VL 16(R4), T_0 + VPERM T_0, T_0, T_1, T_0 // reverse bytes (to big) + VAQ T_0, H0_0, H0_0 + VPERM H0_0, H0_0, T_1, H0_0 // reverse bytes (to little) + VST H0_0, (R1) + RET + +add: + // load EX0, EX1, EX2 + MOVD $·constants<>(SB), R5 + VLM (R5), EX0, EX2 + + REDUCE(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, T_10, M0, M1, M3, M4, M5, T_4, T_5, T_2, T_7, T_8, T_9) + VMRHG V0, H0_1, H0_0 + VMRHG V0, H1_1, H1_0 + VMRHG V0, H2_1, H2_0 + VMRLG V0, H0_1, H0_1 + VMRLG V0, H1_1, H1_1 + VMRLG V0, H2_1, H2_1 + CMPBLE R3, $64, b4 + +b4: + CMPBLE R3, $48, b3 // 3 blocks or less + + // 4(3+1) blocks remaining + SUB $49, R3 + VLM (R2), M0, M2 + VLL R3, 48(R2), M3 + ADD $1, R3 + MOVBZ $1, R0 + CMPBEQ R3, $16, 2(PC) + VLVGB R3, R0, M3 + MOVD $64(R2), R2 + EXPACC(M0, M1, H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, T_0, T_1, T_2, T_3) + VLEIB $10, $1, H2_0 + VLEIB $10, $1, H2_1 + VZERO M0 + VZERO M1 + VZERO M4 + VZERO M5 + VZERO T_4 + VZERO T_10 + EXPACC(M2, M3, M0, M1, M4, M5, T_4, T_10, T_0, T_1, T_2, T_3) + VLR T_4, M2 + VLEIB $10, $1, M4 + CMPBNE R3, $16, 2(PC) + VLEIB $10, $1, T_10 + MULTIPLY(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, R_0, R_1, R_2, R5_1, R5_2, M0, M1, M4, M5, M2, T_10, T_0, T_1, T_2, T_3, T_4, T_5, T_6, T_7, T_8, T_9) + REDUCE(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, T_10, M0, M1, M3, M4, M5, T_4, T_5, T_2, T_7, T_8, T_9) + VMRHG V0, H0_1, H0_0 + VMRHG V0, H1_1, H1_0 + VMRHG V0, H2_1, H2_0 + VMRLG V0, H0_1, H0_1 + VMRLG V0, H1_1, H1_1 + VMRLG V0, H2_1, H2_1 + SUB $16, R3 + CMPBLE R3, $0, square // this condition must always hold true! + +b3: + CMPBLE R3, $32, b2 + + // 3 blocks remaining + + // setup [r²,r] + VSLDB $8, R_0, R_0, R_0 + VSLDB $8, R_1, R_1, R_1 + VSLDB $8, R_2, R_2, R_2 + VSLDB $8, R5_1, R5_1, R5_1 + VSLDB $8, R5_2, R5_2, R5_2 + + VLVGG $1, RSAVE_0, R_0 + VLVGG $1, RSAVE_1, R_1 + VLVGG $1, RSAVE_2, R_2 + VLVGG $1, R5SAVE_1, R5_1 + VLVGG $1, R5SAVE_2, R5_2 + + // setup [h0, h1] + VSLDB $8, H0_0, H0_0, H0_0 + VSLDB $8, H1_0, H1_0, H1_0 + VSLDB $8, H2_0, H2_0, H2_0 + VO H0_1, H0_0, H0_0 + VO H1_1, H1_0, H1_0 + VO H2_1, H2_0, H2_0 + VZERO H0_1 + VZERO H1_1 + VZERO H2_1 + + VZERO M0 + VZERO M1 + VZERO M2 + VZERO M3 + VZERO M4 + VZERO M5 + + // H*[r**2, r] + MULTIPLY(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, R_0, R_1, R_2, R5_1, R5_2, M0, M1, M2, M3, M4, M5, T_0, T_1, T_2, T_3, T_4, T_5, T_6, T_7, T_8, T_9) + REDUCE2(H0_0, H1_0, H2_0, M0, M1, M2, M3, M4, H0_1, H1_1, T_10, M5) + + SUB $33, R3 + VLM (R2), M0, M1 + VLL R3, 32(R2), M2 + ADD $1, R3 + MOVBZ $1, R0 + CMPBEQ R3, $16, 2(PC) + VLVGB R3, R0, M2 + + // H += m0 + VZERO T_1 + VZERO T_2 + VZERO T_3 + EXPACC2(M0, T_1, T_2, T_3, T_4, T_5, T_6) + VLEIB $10, $1, T_3 + VAG H0_0, T_1, H0_0 + VAG H1_0, T_2, H1_0 + VAG H2_0, T_3, H2_0 + + VZERO M0 + VZERO M3 + VZERO M4 + VZERO M5 + VZERO T_10 + + // (H+m0)*r + MULTIPLY(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, R_0, R_1, R_2, R5_1, R5_2, M0, M3, M4, M5, V0, T_10, T_0, T_1, T_2, T_3, T_4, T_5, T_6, T_7, T_8, T_9) + REDUCE2(H0_0, H1_0, H2_0, M0, M3, M4, M5, T_10, H0_1, H1_1, H2_1, T_9) + + // H += m1 + VZERO V0 + VZERO T_1 + VZERO T_2 + VZERO T_3 + EXPACC2(M1, T_1, T_2, T_3, T_4, T_5, T_6) + VLEIB $10, $1, T_3 + VAQ H0_0, T_1, H0_0 + VAQ H1_0, T_2, H1_0 + VAQ H2_0, T_3, H2_0 + REDUCE2(H0_0, H1_0, H2_0, M0, M3, M4, M5, T_9, H0_1, H1_1, H2_1, T_10) + + // [H, m2] * [r**2, r] + EXPACC2(M2, H0_0, H1_0, H2_0, T_1, T_2, T_3) + CMPBNE R3, $16, 2(PC) + VLEIB $10, $1, H2_0 + VZERO M0 + VZERO M1 + VZERO M2 + VZERO M3 + VZERO M4 + VZERO M5 + MULTIPLY(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, R_0, R_1, R_2, R5_1, R5_2, M0, M1, M2, M3, M4, M5, T_0, T_1, T_2, T_3, T_4, T_5, T_6, T_7, T_8, T_9) + REDUCE2(H0_0, H1_0, H2_0, M0, M1, M2, M3, M4, H0_1, H1_1, M5, T_10) + SUB $16, R3 + CMPBLE R3, $0, next // this condition must always hold true! + +b2: + CMPBLE R3, $16, b1 + + // 2 blocks remaining + + // setup [r²,r] + VSLDB $8, R_0, R_0, R_0 + VSLDB $8, R_1, R_1, R_1 + VSLDB $8, R_2, R_2, R_2 + VSLDB $8, R5_1, R5_1, R5_1 + VSLDB $8, R5_2, R5_2, R5_2 + + VLVGG $1, RSAVE_0, R_0 + VLVGG $1, RSAVE_1, R_1 + VLVGG $1, RSAVE_2, R_2 + VLVGG $1, R5SAVE_1, R5_1 + VLVGG $1, R5SAVE_2, R5_2 + + // setup [h0, h1] + VSLDB $8, H0_0, H0_0, H0_0 + VSLDB $8, H1_0, H1_0, H1_0 + VSLDB $8, H2_0, H2_0, H2_0 + VO H0_1, H0_0, H0_0 + VO H1_1, H1_0, H1_0 + VO H2_1, H2_0, H2_0 + VZERO H0_1 + VZERO H1_1 + VZERO H2_1 + + VZERO M0 + VZERO M1 + VZERO M2 + VZERO M3 + VZERO M4 + VZERO M5 + + // H*[r**2, r] + MULTIPLY(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, R_0, R_1, R_2, R5_1, R5_2, M0, M1, M2, M3, M4, M5, T_0, T_1, T_2, T_3, T_4, T_5, T_6, T_7, T_8, T_9) + REDUCE(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, T_10, M0, M1, M2, M3, M4, T_4, T_5, T_2, T_7, T_8, T_9) + VMRHG V0, H0_1, H0_0 + VMRHG V0, H1_1, H1_0 + VMRHG V0, H2_1, H2_0 + VMRLG V0, H0_1, H0_1 + VMRLG V0, H1_1, H1_1 + VMRLG V0, H2_1, H2_1 + + // move h to the left and 0s at the right + VSLDB $8, H0_0, H0_0, H0_0 + VSLDB $8, H1_0, H1_0, H1_0 + VSLDB $8, H2_0, H2_0, H2_0 + + // get message blocks and append 1 to start + SUB $17, R3 + VL (R2), M0 + VLL R3, 16(R2), M1 + ADD $1, R3 + MOVBZ $1, R0 + CMPBEQ R3, $16, 2(PC) + VLVGB R3, R0, M1 + VZERO T_6 + VZERO T_7 + VZERO T_8 + EXPACC2(M0, T_6, T_7, T_8, T_1, T_2, T_3) + EXPACC2(M1, T_6, T_7, T_8, T_1, T_2, T_3) + VLEIB $2, $1, T_8 + CMPBNE R3, $16, 2(PC) + VLEIB $10, $1, T_8 + + // add [m0, m1] to h + VAG H0_0, T_6, H0_0 + VAG H1_0, T_7, H1_0 + VAG H2_0, T_8, H2_0 + + VZERO M2 + VZERO M3 + VZERO M4 + VZERO M5 + VZERO T_10 + VZERO M0 + + // at this point R_0 .. R5_2 look like [r**2, r] + MULTIPLY(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, R_0, R_1, R_2, R5_1, R5_2, M2, M3, M4, M5, T_10, M0, T_0, T_1, T_2, T_3, T_4, T_5, T_6, T_7, T_8, T_9) + REDUCE2(H0_0, H1_0, H2_0, M2, M3, M4, M5, T_9, H0_1, H1_1, H2_1, T_10) + SUB $16, R3, R3 + CMPBLE R3, $0, next + +b1: + CMPBLE R3, $0, next + + // 1 block remaining + + // setup [r²,r] + VSLDB $8, R_0, R_0, R_0 + VSLDB $8, R_1, R_1, R_1 + VSLDB $8, R_2, R_2, R_2 + VSLDB $8, R5_1, R5_1, R5_1 + VSLDB $8, R5_2, R5_2, R5_2 + + VLVGG $1, RSAVE_0, R_0 + VLVGG $1, RSAVE_1, R_1 + VLVGG $1, RSAVE_2, R_2 + VLVGG $1, R5SAVE_1, R5_1 + VLVGG $1, R5SAVE_2, R5_2 + + // setup [h0, h1] + VSLDB $8, H0_0, H0_0, H0_0 + VSLDB $8, H1_0, H1_0, H1_0 + VSLDB $8, H2_0, H2_0, H2_0 + VO H0_1, H0_0, H0_0 + VO H1_1, H1_0, H1_0 + VO H2_1, H2_0, H2_0 + VZERO H0_1 + VZERO H1_1 + VZERO H2_1 + + VZERO M0 + VZERO M1 + VZERO M2 + VZERO M3 + VZERO M4 + VZERO M5 + + // H*[r**2, r] + MULTIPLY(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, R_0, R_1, R_2, R5_1, R5_2, M0, M1, M2, M3, M4, M5, T_0, T_1, T_2, T_3, T_4, T_5, T_6, T_7, T_8, T_9) + REDUCE2(H0_0, H1_0, H2_0, M0, M1, M2, M3, M4, T_9, T_10, H0_1, M5) + + // set up [0, m0] limbs + SUB $1, R3 + VLL R3, (R2), M0 + ADD $1, R3 + MOVBZ $1, R0 + CMPBEQ R3, $16, 2(PC) + VLVGB R3, R0, M0 + VZERO T_1 + VZERO T_2 + VZERO T_3 + EXPACC2(M0, T_1, T_2, T_3, T_4, T_5, T_6)// limbs: [0, m] + CMPBNE R3, $16, 2(PC) + VLEIB $10, $1, T_3 + + // h+m0 + VAQ H0_0, T_1, H0_0 + VAQ H1_0, T_2, H1_0 + VAQ H2_0, T_3, H2_0 + + VZERO M0 + VZERO M1 + VZERO M2 + VZERO M3 + VZERO M4 + VZERO M5 + MULTIPLY(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, R_0, R_1, R_2, R5_1, R5_2, M0, M1, M2, M3, M4, M5, T_0, T_1, T_2, T_3, T_4, T_5, T_6, T_7, T_8, T_9) + REDUCE2(H0_0, H1_0, H2_0, M0, M1, M2, M3, M4, T_9, T_10, H0_1, M5) + + BR next + +square: + // setup [r²,r] + VSLDB $8, R_0, R_0, R_0 + VSLDB $8, R_1, R_1, R_1 + VSLDB $8, R_2, R_2, R_2 + VSLDB $8, R5_1, R5_1, R5_1 + VSLDB $8, R5_2, R5_2, R5_2 + + VLVGG $1, RSAVE_0, R_0 + VLVGG $1, RSAVE_1, R_1 + VLVGG $1, RSAVE_2, R_2 + VLVGG $1, R5SAVE_1, R5_1 + VLVGG $1, R5SAVE_2, R5_2 + + // setup [h0, h1] + VSLDB $8, H0_0, H0_0, H0_0 + VSLDB $8, H1_0, H1_0, H1_0 + VSLDB $8, H2_0, H2_0, H2_0 + VO H0_1, H0_0, H0_0 + VO H1_1, H1_0, H1_0 + VO H2_1, H2_0, H2_0 + VZERO H0_1 + VZERO H1_1 + VZERO H2_1 + + VZERO M0 + VZERO M1 + VZERO M2 + VZERO M3 + VZERO M4 + VZERO M5 + + // (h0*r**2) + (h1*r) + MULTIPLY(H0_0, H1_0, H2_0, H0_1, H1_1, H2_1, R_0, R_1, R_2, R5_1, R5_2, M0, M1, M2, M3, M4, M5, T_0, T_1, T_2, T_3, T_4, T_5, T_6, T_7, T_8, T_9) + REDUCE2(H0_0, H1_0, H2_0, M0, M1, M2, M3, M4, T_9, T_10, H0_1, M5) + BR next diff --git a/vendor/golang.org/x/crypto/ssh/terminal/terminal.go b/vendor/golang.org/x/crypto/ssh/terminal/terminal.go new file mode 100644 index 0000000..2f04ee5 --- /dev/null +++ b/vendor/golang.org/x/crypto/ssh/terminal/terminal.go @@ -0,0 +1,966 @@ +// Copyright 2011 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package terminal + +import ( + "bytes" + "io" + "strconv" + "sync" + "unicode/utf8" +) + +// EscapeCodes contains escape sequences that can be written to the terminal in +// order to achieve different styles of text. +type EscapeCodes struct { + // Foreground colors + Black, Red, Green, Yellow, Blue, Magenta, Cyan, White []byte + + // Reset all attributes + Reset []byte +} + +var vt100EscapeCodes = EscapeCodes{ + Black: []byte{keyEscape, '[', '3', '0', 'm'}, + Red: []byte{keyEscape, '[', '3', '1', 'm'}, + Green: []byte{keyEscape, '[', '3', '2', 'm'}, + Yellow: []byte{keyEscape, '[', '3', '3', 'm'}, + Blue: []byte{keyEscape, '[', '3', '4', 'm'}, + Magenta: []byte{keyEscape, '[', '3', '5', 'm'}, + Cyan: []byte{keyEscape, '[', '3', '6', 'm'}, + White: []byte{keyEscape, '[', '3', '7', 'm'}, + + Reset: []byte{keyEscape, '[', '0', 'm'}, +} + +// Terminal contains the state for running a VT100 terminal that is capable of +// reading lines of input. +type Terminal struct { + // AutoCompleteCallback, if non-null, is called for each keypress with + // the full input line and the current position of the cursor (in + // bytes, as an index into |line|). If it returns ok=false, the key + // press is processed normally. Otherwise it returns a replacement line + // and the new cursor position. + AutoCompleteCallback func(line string, pos int, key rune) (newLine string, newPos int, ok bool) + + // Escape contains a pointer to the escape codes for this terminal. + // It's always a valid pointer, although the escape codes themselves + // may be empty if the terminal doesn't support them. + Escape *EscapeCodes + + // lock protects the terminal and the state in this object from + // concurrent processing of a key press and a Write() call. + lock sync.Mutex + + c io.ReadWriter + prompt []rune + + // line is the current line being entered. + line []rune + // pos is the logical position of the cursor in line + pos int + // echo is true if local echo is enabled + echo bool + // pasteActive is true iff there is a bracketed paste operation in + // progress. + pasteActive bool + + // cursorX contains the current X value of the cursor where the left + // edge is 0. cursorY contains the row number where the first row of + // the current line is 0. + cursorX, cursorY int + // maxLine is the greatest value of cursorY so far. + maxLine int + + termWidth, termHeight int + + // outBuf contains the terminal data to be sent. + outBuf []byte + // remainder contains the remainder of any partial key sequences after + // a read. It aliases into inBuf. + remainder []byte + inBuf [256]byte + + // history contains previously entered commands so that they can be + // accessed with the up and down keys. + history stRingBuffer + // historyIndex stores the currently accessed history entry, where zero + // means the immediately previous entry. + historyIndex int + // When navigating up and down the history it's possible to return to + // the incomplete, initial line. That value is stored in + // historyPending. + historyPending string +} + +// NewTerminal runs a VT100 terminal on the given ReadWriter. If the ReadWriter is +// a local terminal, that terminal must first have been put into raw mode. +// prompt is a string that is written at the start of each input line (i.e. +// "> "). +func NewTerminal(c io.ReadWriter, prompt string) *Terminal { + return &Terminal{ + Escape: &vt100EscapeCodes, + c: c, + prompt: []rune(prompt), + termWidth: 80, + termHeight: 24, + echo: true, + historyIndex: -1, + } +} + +const ( + keyCtrlD = 4 + keyCtrlU = 21 + keyEnter = '\r' + keyEscape = 27 + keyBackspace = 127 + keyUnknown = 0xd800 /* UTF-16 surrogate area */ + iota + keyUp + keyDown + keyLeft + keyRight + keyAltLeft + keyAltRight + keyHome + keyEnd + keyDeleteWord + keyDeleteLine + keyClearScreen + keyPasteStart + keyPasteEnd +) + +var ( + crlf = []byte{'\r', '\n'} + pasteStart = []byte{keyEscape, '[', '2', '0', '0', '~'} + pasteEnd = []byte{keyEscape, '[', '2', '0', '1', '~'} +) + +// bytesToKey tries to parse a key sequence from b. If successful, it returns +// the key and the remainder of the input. Otherwise it returns utf8.RuneError. +func bytesToKey(b []byte, pasteActive bool) (rune, []byte) { + if len(b) == 0 { + return utf8.RuneError, nil + } + + if !pasteActive { + switch b[0] { + case 1: // ^A + return keyHome, b[1:] + case 5: // ^E + return keyEnd, b[1:] + case 8: // ^H + return keyBackspace, b[1:] + case 11: // ^K + return keyDeleteLine, b[1:] + case 12: // ^L + return keyClearScreen, b[1:] + case 23: // ^W + return keyDeleteWord, b[1:] + case 14: // ^N + return keyDown, b[1:] + case 16: // ^P + return keyUp, b[1:] + } + } + + if b[0] != keyEscape { + if !utf8.FullRune(b) { + return utf8.RuneError, b + } + r, l := utf8.DecodeRune(b) + return r, b[l:] + } + + if !pasteActive && len(b) >= 3 && b[0] == keyEscape && b[1] == '[' { + switch b[2] { + case 'A': + return keyUp, b[3:] + case 'B': + return keyDown, b[3:] + case 'C': + return keyRight, b[3:] + case 'D': + return keyLeft, b[3:] + case 'H': + return keyHome, b[3:] + case 'F': + return keyEnd, b[3:] + } + } + + if !pasteActive && len(b) >= 6 && b[0] == keyEscape && b[1] == '[' && b[2] == '1' && b[3] == ';' && b[4] == '3' { + switch b[5] { + case 'C': + return keyAltRight, b[6:] + case 'D': + return keyAltLeft, b[6:] + } + } + + if !pasteActive && len(b) >= 6 && bytes.Equal(b[:6], pasteStart) { + return keyPasteStart, b[6:] + } + + if pasteActive && len(b) >= 6 && bytes.Equal(b[:6], pasteEnd) { + return keyPasteEnd, b[6:] + } + + // If we get here then we have a key that we don't recognise, or a + // partial sequence. It's not clear how one should find the end of a + // sequence without knowing them all, but it seems that [a-zA-Z~] only + // appears at the end of a sequence. + for i, c := range b[0:] { + if c >= 'a' && c <= 'z' || c >= 'A' && c <= 'Z' || c == '~' { + return keyUnknown, b[i+1:] + } + } + + return utf8.RuneError, b +} + +// queue appends data to the end of t.outBuf +func (t *Terminal) queue(data []rune) { + t.outBuf = append(t.outBuf, []byte(string(data))...) +} + +var eraseUnderCursor = []rune{' ', keyEscape, '[', 'D'} +var space = []rune{' '} + +func isPrintable(key rune) bool { + isInSurrogateArea := key >= 0xd800 && key <= 0xdbff + return key >= 32 && !isInSurrogateArea +} + +// moveCursorToPos appends data to t.outBuf which will move the cursor to the +// given, logical position in the text. +func (t *Terminal) moveCursorToPos(pos int) { + if !t.echo { + return + } + + x := visualLength(t.prompt) + pos + y := x / t.termWidth + x = x % t.termWidth + + up := 0 + if y < t.cursorY { + up = t.cursorY - y + } + + down := 0 + if y > t.cursorY { + down = y - t.cursorY + } + + left := 0 + if x < t.cursorX { + left = t.cursorX - x + } + + right := 0 + if x > t.cursorX { + right = x - t.cursorX + } + + t.cursorX = x + t.cursorY = y + t.move(up, down, left, right) +} + +func (t *Terminal) move(up, down, left, right int) { + m := []rune{} + + // 1 unit up can be expressed as ^[[A or ^[A + // 5 units up can be expressed as ^[[5A + + if up == 1 { + m = append(m, keyEscape, '[', 'A') + } else if up > 1 { + m = append(m, keyEscape, '[') + m = append(m, []rune(strconv.Itoa(up))...) + m = append(m, 'A') + } + + if down == 1 { + m = append(m, keyEscape, '[', 'B') + } else if down > 1 { + m = append(m, keyEscape, '[') + m = append(m, []rune(strconv.Itoa(down))...) + m = append(m, 'B') + } + + if right == 1 { + m = append(m, keyEscape, '[', 'C') + } else if right > 1 { + m = append(m, keyEscape, '[') + m = append(m, []rune(strconv.Itoa(right))...) + m = append(m, 'C') + } + + if left == 1 { + m = append(m, keyEscape, '[', 'D') + } else if left > 1 { + m = append(m, keyEscape, '[') + m = append(m, []rune(strconv.Itoa(left))...) + m = append(m, 'D') + } + + t.queue(m) +} + +func (t *Terminal) clearLineToRight() { + op := []rune{keyEscape, '[', 'K'} + t.queue(op) +} + +const maxLineLength = 4096 + +func (t *Terminal) setLine(newLine []rune, newPos int) { + if t.echo { + t.moveCursorToPos(0) + t.writeLine(newLine) + for i := len(newLine); i < len(t.line); i++ { + t.writeLine(space) + } + t.moveCursorToPos(newPos) + } + t.line = newLine + t.pos = newPos +} + +func (t *Terminal) advanceCursor(places int) { + t.cursorX += places + t.cursorY += t.cursorX / t.termWidth + if t.cursorY > t.maxLine { + t.maxLine = t.cursorY + } + t.cursorX = t.cursorX % t.termWidth + + if places > 0 && t.cursorX == 0 { + // Normally terminals will advance the current position + // when writing a character. But that doesn't happen + // for the last character in a line. However, when + // writing a character (except a new line) that causes + // a line wrap, the position will be advanced two + // places. + // + // So, if we are stopping at the end of a line, we + // need to write a newline so that our cursor can be + // advanced to the next line. + t.outBuf = append(t.outBuf, '\r', '\n') + } +} + +func (t *Terminal) eraseNPreviousChars(n int) { + if n == 0 { + return + } + + if t.pos < n { + n = t.pos + } + t.pos -= n + t.moveCursorToPos(t.pos) + + copy(t.line[t.pos:], t.line[n+t.pos:]) + t.line = t.line[:len(t.line)-n] + if t.echo { + t.writeLine(t.line[t.pos:]) + for i := 0; i < n; i++ { + t.queue(space) + } + t.advanceCursor(n) + t.moveCursorToPos(t.pos) + } +} + +// countToLeftWord returns then number of characters from the cursor to the +// start of the previous word. +func (t *Terminal) countToLeftWord() int { + if t.pos == 0 { + return 0 + } + + pos := t.pos - 1 + for pos > 0 { + if t.line[pos] != ' ' { + break + } + pos-- + } + for pos > 0 { + if t.line[pos] == ' ' { + pos++ + break + } + pos-- + } + + return t.pos - pos +} + +// countToRightWord returns then number of characters from the cursor to the +// start of the next word. +func (t *Terminal) countToRightWord() int { + pos := t.pos + for pos < len(t.line) { + if t.line[pos] == ' ' { + break + } + pos++ + } + for pos < len(t.line) { + if t.line[pos] != ' ' { + break + } + pos++ + } + return pos - t.pos +} + +// visualLength returns the number of visible glyphs in s. +func visualLength(runes []rune) int { + inEscapeSeq := false + length := 0 + + for _, r := range runes { + switch { + case inEscapeSeq: + if (r >= 'a' && r <= 'z') || (r >= 'A' && r <= 'Z') { + inEscapeSeq = false + } + case r == '\x1b': + inEscapeSeq = true + default: + length++ + } + } + + return length +} + +// handleKey processes the given key and, optionally, returns a line of text +// that the user has entered. +func (t *Terminal) handleKey(key rune) (line string, ok bool) { + if t.pasteActive && key != keyEnter { + t.addKeyToLine(key) + return + } + + switch key { + case keyBackspace: + if t.pos == 0 { + return + } + t.eraseNPreviousChars(1) + case keyAltLeft: + // move left by a word. + t.pos -= t.countToLeftWord() + t.moveCursorToPos(t.pos) + case keyAltRight: + // move right by a word. + t.pos += t.countToRightWord() + t.moveCursorToPos(t.pos) + case keyLeft: + if t.pos == 0 { + return + } + t.pos-- + t.moveCursorToPos(t.pos) + case keyRight: + if t.pos == len(t.line) { + return + } + t.pos++ + t.moveCursorToPos(t.pos) + case keyHome: + if t.pos == 0 { + return + } + t.pos = 0 + t.moveCursorToPos(t.pos) + case keyEnd: + if t.pos == len(t.line) { + return + } + t.pos = len(t.line) + t.moveCursorToPos(t.pos) + case keyUp: + entry, ok := t.history.NthPreviousEntry(t.historyIndex + 1) + if !ok { + return "", false + } + if t.historyIndex == -1 { + t.historyPending = string(t.line) + } + t.historyIndex++ + runes := []rune(entry) + t.setLine(runes, len(runes)) + case keyDown: + switch t.historyIndex { + case -1: + return + case 0: + runes := []rune(t.historyPending) + t.setLine(runes, len(runes)) + t.historyIndex-- + default: + entry, ok := t.history.NthPreviousEntry(t.historyIndex - 1) + if ok { + t.historyIndex-- + runes := []rune(entry) + t.setLine(runes, len(runes)) + } + } + case keyEnter: + t.moveCursorToPos(len(t.line)) + t.queue([]rune("\r\n")) + line = string(t.line) + ok = true + t.line = t.line[:0] + t.pos = 0 + t.cursorX = 0 + t.cursorY = 0 + t.maxLine = 0 + case keyDeleteWord: + // Delete zero or more spaces and then one or more characters. + t.eraseNPreviousChars(t.countToLeftWord()) + case keyDeleteLine: + // Delete everything from the current cursor position to the + // end of line. + for i := t.pos; i < len(t.line); i++ { + t.queue(space) + t.advanceCursor(1) + } + t.line = t.line[:t.pos] + t.moveCursorToPos(t.pos) + case keyCtrlD: + // Erase the character under the current position. + // The EOF case when the line is empty is handled in + // readLine(). + if t.pos < len(t.line) { + t.pos++ + t.eraseNPreviousChars(1) + } + case keyCtrlU: + t.eraseNPreviousChars(t.pos) + case keyClearScreen: + // Erases the screen and moves the cursor to the home position. + t.queue([]rune("\x1b[2J\x1b[H")) + t.queue(t.prompt) + t.cursorX, t.cursorY = 0, 0 + t.advanceCursor(visualLength(t.prompt)) + t.setLine(t.line, t.pos) + default: + if t.AutoCompleteCallback != nil { + prefix := string(t.line[:t.pos]) + suffix := string(t.line[t.pos:]) + + t.lock.Unlock() + newLine, newPos, completeOk := t.AutoCompleteCallback(prefix+suffix, len(prefix), key) + t.lock.Lock() + + if completeOk { + t.setLine([]rune(newLine), utf8.RuneCount([]byte(newLine)[:newPos])) + return + } + } + if !isPrintable(key) { + return + } + if len(t.line) == maxLineLength { + return + } + t.addKeyToLine(key) + } + return +} + +// addKeyToLine inserts the given key at the current position in the current +// line. +func (t *Terminal) addKeyToLine(key rune) { + if len(t.line) == cap(t.line) { + newLine := make([]rune, len(t.line), 2*(1+len(t.line))) + copy(newLine, t.line) + t.line = newLine + } + t.line = t.line[:len(t.line)+1] + copy(t.line[t.pos+1:], t.line[t.pos:]) + t.line[t.pos] = key + if t.echo { + t.writeLine(t.line[t.pos:]) + } + t.pos++ + t.moveCursorToPos(t.pos) +} + +func (t *Terminal) writeLine(line []rune) { + for len(line) != 0 { + remainingOnLine := t.termWidth - t.cursorX + todo := len(line) + if todo > remainingOnLine { + todo = remainingOnLine + } + t.queue(line[:todo]) + t.advanceCursor(visualLength(line[:todo])) + line = line[todo:] + } +} + +// writeWithCRLF writes buf to w but replaces all occurrences of \n with \r\n. +func writeWithCRLF(w io.Writer, buf []byte) (n int, err error) { + for len(buf) > 0 { + i := bytes.IndexByte(buf, '\n') + todo := len(buf) + if i >= 0 { + todo = i + } + + var nn int + nn, err = w.Write(buf[:todo]) + n += nn + if err != nil { + return n, err + } + buf = buf[todo:] + + if i >= 0 { + if _, err = w.Write(crlf); err != nil { + return n, err + } + n++ + buf = buf[1:] + } + } + + return n, nil +} + +func (t *Terminal) Write(buf []byte) (n int, err error) { + t.lock.Lock() + defer t.lock.Unlock() + + if t.cursorX == 0 && t.cursorY == 0 { + // This is the easy case: there's nothing on the screen that we + // have to move out of the way. + return writeWithCRLF(t.c, buf) + } + + // We have a prompt and possibly user input on the screen. We + // have to clear it first. + t.move(0 /* up */, 0 /* down */, t.cursorX /* left */, 0 /* right */) + t.cursorX = 0 + t.clearLineToRight() + + for t.cursorY > 0 { + t.move(1 /* up */, 0, 0, 0) + t.cursorY-- + t.clearLineToRight() + } + + if _, err = t.c.Write(t.outBuf); err != nil { + return + } + t.outBuf = t.outBuf[:0] + + if n, err = writeWithCRLF(t.c, buf); err != nil { + return + } + + t.writeLine(t.prompt) + if t.echo { + t.writeLine(t.line) + } + + t.moveCursorToPos(t.pos) + + if _, err = t.c.Write(t.outBuf); err != nil { + return + } + t.outBuf = t.outBuf[:0] + return +} + +// ReadPassword temporarily changes the prompt and reads a password, without +// echo, from the terminal. +func (t *Terminal) ReadPassword(prompt string) (line string, err error) { + t.lock.Lock() + defer t.lock.Unlock() + + oldPrompt := t.prompt + t.prompt = []rune(prompt) + t.echo = false + + line, err = t.readLine() + + t.prompt = oldPrompt + t.echo = true + + return +} + +// ReadLine returns a line of input from the terminal. +func (t *Terminal) ReadLine() (line string, err error) { + t.lock.Lock() + defer t.lock.Unlock() + + return t.readLine() +} + +func (t *Terminal) readLine() (line string, err error) { + // t.lock must be held at this point + + if t.cursorX == 0 && t.cursorY == 0 { + t.writeLine(t.prompt) + t.c.Write(t.outBuf) + t.outBuf = t.outBuf[:0] + } + + lineIsPasted := t.pasteActive + + for { + rest := t.remainder + lineOk := false + for !lineOk { + var key rune + key, rest = bytesToKey(rest, t.pasteActive) + if key == utf8.RuneError { + break + } + if !t.pasteActive { + if key == keyCtrlD { + if len(t.line) == 0 { + return "", io.EOF + } + } + if key == keyPasteStart { + t.pasteActive = true + if len(t.line) == 0 { + lineIsPasted = true + } + continue + } + } else if key == keyPasteEnd { + t.pasteActive = false + continue + } + if !t.pasteActive { + lineIsPasted = false + } + line, lineOk = t.handleKey(key) + } + if len(rest) > 0 { + n := copy(t.inBuf[:], rest) + t.remainder = t.inBuf[:n] + } else { + t.remainder = nil + } + t.c.Write(t.outBuf) + t.outBuf = t.outBuf[:0] + if lineOk { + if t.echo { + t.historyIndex = -1 + t.history.Add(line) + } + if lineIsPasted { + err = ErrPasteIndicator + } + return + } + + // t.remainder is a slice at the beginning of t.inBuf + // containing a partial key sequence + readBuf := t.inBuf[len(t.remainder):] + var n int + + t.lock.Unlock() + n, err = t.c.Read(readBuf) + t.lock.Lock() + + if err != nil { + return + } + + t.remainder = t.inBuf[:n+len(t.remainder)] + } +} + +// SetPrompt sets the prompt to be used when reading subsequent lines. +func (t *Terminal) SetPrompt(prompt string) { + t.lock.Lock() + defer t.lock.Unlock() + + t.prompt = []rune(prompt) +} + +func (t *Terminal) clearAndRepaintLinePlusNPrevious(numPrevLines int) { + // Move cursor to column zero at the start of the line. + t.move(t.cursorY, 0, t.cursorX, 0) + t.cursorX, t.cursorY = 0, 0 + t.clearLineToRight() + for t.cursorY < numPrevLines { + // Move down a line + t.move(0, 1, 0, 0) + t.cursorY++ + t.clearLineToRight() + } + // Move back to beginning. + t.move(t.cursorY, 0, 0, 0) + t.cursorX, t.cursorY = 0, 0 + + t.queue(t.prompt) + t.advanceCursor(visualLength(t.prompt)) + t.writeLine(t.line) + t.moveCursorToPos(t.pos) +} + +func (t *Terminal) SetSize(width, height int) error { + t.lock.Lock() + defer t.lock.Unlock() + + if width == 0 { + width = 1 + } + + oldWidth := t.termWidth + t.termWidth, t.termHeight = width, height + + switch { + case width == oldWidth: + // If the width didn't change then nothing else needs to be + // done. + return nil + case len(t.line) == 0 && t.cursorX == 0 && t.cursorY == 0: + // If there is nothing on current line and no prompt printed, + // just do nothing + return nil + case width < oldWidth: + // Some terminals (e.g. xterm) will truncate lines that were + // too long when shinking. Others, (e.g. gnome-terminal) will + // attempt to wrap them. For the former, repainting t.maxLine + // works great, but that behaviour goes badly wrong in the case + // of the latter because they have doubled every full line. + + // We assume that we are working on a terminal that wraps lines + // and adjust the cursor position based on every previous line + // wrapping and turning into two. This causes the prompt on + // xterms to move upwards, which isn't great, but it avoids a + // huge mess with gnome-terminal. + if t.cursorX >= t.termWidth { + t.cursorX = t.termWidth - 1 + } + t.cursorY *= 2 + t.clearAndRepaintLinePlusNPrevious(t.maxLine * 2) + case width > oldWidth: + // If the terminal expands then our position calculations will + // be wrong in the future because we think the cursor is + // |t.pos| chars into the string, but there will be a gap at + // the end of any wrapped line. + // + // But the position will actually be correct until we move, so + // we can move back to the beginning and repaint everything. + t.clearAndRepaintLinePlusNPrevious(t.maxLine) + } + + _, err := t.c.Write(t.outBuf) + t.outBuf = t.outBuf[:0] + return err +} + +type pasteIndicatorError struct{} + +func (pasteIndicatorError) Error() string { + return "terminal: ErrPasteIndicator not correctly handled" +} + +// ErrPasteIndicator may be returned from ReadLine as the error, in addition +// to valid line data. It indicates that bracketed paste mode is enabled and +// that the returned line consists only of pasted data. Programs may wish to +// interpret pasted data more literally than typed data. +var ErrPasteIndicator = pasteIndicatorError{} + +// SetBracketedPasteMode requests that the terminal bracket paste operations +// with markers. Not all terminals support this but, if it is supported, then +// enabling this mode will stop any autocomplete callback from running due to +// pastes. Additionally, any lines that are completely pasted will be returned +// from ReadLine with the error set to ErrPasteIndicator. +func (t *Terminal) SetBracketedPasteMode(on bool) { + if on { + io.WriteString(t.c, "\x1b[?2004h") + } else { + io.WriteString(t.c, "\x1b[?2004l") + } +} + +// stRingBuffer is a ring buffer of strings. +type stRingBuffer struct { + // entries contains max elements. + entries []string + max int + // head contains the index of the element most recently added to the ring. + head int + // size contains the number of elements in the ring. + size int +} + +func (s *stRingBuffer) Add(a string) { + if s.entries == nil { + const defaultNumEntries = 100 + s.entries = make([]string, defaultNumEntries) + s.max = defaultNumEntries + } + + s.head = (s.head + 1) % s.max + s.entries[s.head] = a + if s.size < s.max { + s.size++ + } +} + +// NthPreviousEntry returns the value passed to the nth previous call to Add. +// If n is zero then the immediately prior value is returned, if one, then the +// next most recent, and so on. If such an element doesn't exist then ok is +// false. +func (s *stRingBuffer) NthPreviousEntry(n int) (value string, ok bool) { + if n >= s.size { + return "", false + } + index := s.head - n + if index < 0 { + index += s.max + } + return s.entries[index], true +} + +// readPasswordLine reads from reader until it finds \n or io.EOF. +// The slice returned does not include the \n. +// readPasswordLine also ignores any \r it finds. +func readPasswordLine(reader io.Reader) ([]byte, error) { + var buf [1]byte + var ret []byte + + for { + n, err := reader.Read(buf[:]) + if n > 0 { + switch buf[0] { + case '\n': + return ret, nil + case '\r': + // remove \r from passwords on Windows + default: + ret = append(ret, buf[0]) + } + continue + } + if err != nil { + if err == io.EOF && len(ret) > 0 { + return ret, nil + } + return ret, err + } + } +} diff --git a/vendor/golang.org/x/crypto/ssh/terminal/util.go b/vendor/golang.org/x/crypto/ssh/terminal/util.go new file mode 100644 index 0000000..3911040 --- /dev/null +++ b/vendor/golang.org/x/crypto/ssh/terminal/util.go @@ -0,0 +1,114 @@ +// Copyright 2011 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build aix darwin dragonfly freebsd linux,!appengine netbsd openbsd + +// Package terminal provides support functions for dealing with terminals, as +// commonly found on UNIX systems. +// +// Putting a terminal into raw mode is the most common requirement: +// +// oldState, err := terminal.MakeRaw(0) +// if err != nil { +// panic(err) +// } +// defer terminal.Restore(0, oldState) +package terminal // import "golang.org/x/crypto/ssh/terminal" + +import ( + "golang.org/x/sys/unix" +) + +// State contains the state of a terminal. +type State struct { + termios unix.Termios +} + +// IsTerminal returns whether the given file descriptor is a terminal. +func IsTerminal(fd int) bool { + _, err := unix.IoctlGetTermios(fd, ioctlReadTermios) + return err == nil +} + +// MakeRaw put the terminal connected to the given file descriptor into raw +// mode and returns the previous state of the terminal so that it can be +// restored. +func MakeRaw(fd int) (*State, error) { + termios, err := unix.IoctlGetTermios(fd, ioctlReadTermios) + if err != nil { + return nil, err + } + + oldState := State{termios: *termios} + + // This attempts to replicate the behaviour documented for cfmakeraw in + // the termios(3) manpage. + termios.Iflag &^= unix.IGNBRK | unix.BRKINT | unix.PARMRK | unix.ISTRIP | unix.INLCR | unix.IGNCR | unix.ICRNL | unix.IXON + termios.Oflag &^= unix.OPOST + termios.Lflag &^= unix.ECHO | unix.ECHONL | unix.ICANON | unix.ISIG | unix.IEXTEN + termios.Cflag &^= unix.CSIZE | unix.PARENB + termios.Cflag |= unix.CS8 + termios.Cc[unix.VMIN] = 1 + termios.Cc[unix.VTIME] = 0 + if err := unix.IoctlSetTermios(fd, ioctlWriteTermios, termios); err != nil { + return nil, err + } + + return &oldState, nil +} + +// GetState returns the current state of a terminal which may be useful to +// restore the terminal after a signal. +func GetState(fd int) (*State, error) { + termios, err := unix.IoctlGetTermios(fd, ioctlReadTermios) + if err != nil { + return nil, err + } + + return &State{termios: *termios}, nil +} + +// Restore restores the terminal connected to the given file descriptor to a +// previous state. +func Restore(fd int, state *State) error { + return unix.IoctlSetTermios(fd, ioctlWriteTermios, &state.termios) +} + +// GetSize returns the dimensions of the given terminal. +func GetSize(fd int) (width, height int, err error) { + ws, err := unix.IoctlGetWinsize(fd, unix.TIOCGWINSZ) + if err != nil { + return -1, -1, err + } + return int(ws.Col), int(ws.Row), nil +} + +// passwordReader is an io.Reader that reads from a specific file descriptor. +type passwordReader int + +func (r passwordReader) Read(buf []byte) (int, error) { + return unix.Read(int(r), buf) +} + +// ReadPassword reads a line of input from a terminal without local echo. This +// is commonly used for inputting passwords and other sensitive data. The slice +// returned does not include the \n. +func ReadPassword(fd int) ([]byte, error) { + termios, err := unix.IoctlGetTermios(fd, ioctlReadTermios) + if err != nil { + return nil, err + } + + newState := *termios + newState.Lflag &^= unix.ECHO + newState.Lflag |= unix.ICANON | unix.ISIG + newState.Iflag |= unix.ICRNL + if err := unix.IoctlSetTermios(fd, ioctlWriteTermios, &newState); err != nil { + return nil, err + } + + defer unix.IoctlSetTermios(fd, ioctlWriteTermios, termios) + + return readPasswordLine(passwordReader(fd)) +} diff --git a/vendor/golang.org/x/crypto/ssh/terminal/util_aix.go b/vendor/golang.org/x/crypto/ssh/terminal/util_aix.go new file mode 100644 index 0000000..dfcd627 --- /dev/null +++ b/vendor/golang.org/x/crypto/ssh/terminal/util_aix.go @@ -0,0 +1,12 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build aix + +package terminal + +import "golang.org/x/sys/unix" + +const ioctlReadTermios = unix.TCGETS +const ioctlWriteTermios = unix.TCSETS diff --git a/vendor/golang.org/x/crypto/ssh/terminal/util_bsd.go b/vendor/golang.org/x/crypto/ssh/terminal/util_bsd.go new file mode 100644 index 0000000..cb23a59 --- /dev/null +++ b/vendor/golang.org/x/crypto/ssh/terminal/util_bsd.go @@ -0,0 +1,12 @@ +// Copyright 2013 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build darwin dragonfly freebsd netbsd openbsd + +package terminal + +import "golang.org/x/sys/unix" + +const ioctlReadTermios = unix.TIOCGETA +const ioctlWriteTermios = unix.TIOCSETA diff --git a/vendor/golang.org/x/crypto/ssh/terminal/util_linux.go b/vendor/golang.org/x/crypto/ssh/terminal/util_linux.go new file mode 100644 index 0000000..5fadfe8 --- /dev/null +++ b/vendor/golang.org/x/crypto/ssh/terminal/util_linux.go @@ -0,0 +1,10 @@ +// Copyright 2013 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package terminal + +import "golang.org/x/sys/unix" + +const ioctlReadTermios = unix.TCGETS +const ioctlWriteTermios = unix.TCSETS diff --git a/vendor/golang.org/x/crypto/ssh/terminal/util_plan9.go b/vendor/golang.org/x/crypto/ssh/terminal/util_plan9.go new file mode 100644 index 0000000..9317ac7 --- /dev/null +++ b/vendor/golang.org/x/crypto/ssh/terminal/util_plan9.go @@ -0,0 +1,58 @@ +// Copyright 2016 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// Package terminal provides support functions for dealing with terminals, as +// commonly found on UNIX systems. +// +// Putting a terminal into raw mode is the most common requirement: +// +// oldState, err := terminal.MakeRaw(0) +// if err != nil { +// panic(err) +// } +// defer terminal.Restore(0, oldState) +package terminal + +import ( + "fmt" + "runtime" +) + +type State struct{} + +// IsTerminal returns whether the given file descriptor is a terminal. +func IsTerminal(fd int) bool { + return false +} + +// MakeRaw put the terminal connected to the given file descriptor into raw +// mode and returns the previous state of the terminal so that it can be +// restored. +func MakeRaw(fd int) (*State, error) { + return nil, fmt.Errorf("terminal: MakeRaw not implemented on %s/%s", runtime.GOOS, runtime.GOARCH) +} + +// GetState returns the current state of a terminal which may be useful to +// restore the terminal after a signal. +func GetState(fd int) (*State, error) { + return nil, fmt.Errorf("terminal: GetState not implemented on %s/%s", runtime.GOOS, runtime.GOARCH) +} + +// Restore restores the terminal connected to the given file descriptor to a +// previous state. +func Restore(fd int, state *State) error { + return fmt.Errorf("terminal: Restore not implemented on %s/%s", runtime.GOOS, runtime.GOARCH) +} + +// GetSize returns the dimensions of the given terminal. +func GetSize(fd int) (width, height int, err error) { + return 0, 0, fmt.Errorf("terminal: GetSize not implemented on %s/%s", runtime.GOOS, runtime.GOARCH) +} + +// ReadPassword reads a line of input from a terminal without local echo. This +// is commonly used for inputting passwords and other sensitive data. The slice +// returned does not include the \n. +func ReadPassword(fd int) ([]byte, error) { + return nil, fmt.Errorf("terminal: ReadPassword not implemented on %s/%s", runtime.GOOS, runtime.GOARCH) +} diff --git a/vendor/golang.org/x/crypto/ssh/terminal/util_solaris.go b/vendor/golang.org/x/crypto/ssh/terminal/util_solaris.go new file mode 100644 index 0000000..3d5f06a --- /dev/null +++ b/vendor/golang.org/x/crypto/ssh/terminal/util_solaris.go @@ -0,0 +1,124 @@ +// Copyright 2015 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build solaris + +package terminal // import "golang.org/x/crypto/ssh/terminal" + +import ( + "golang.org/x/sys/unix" + "io" + "syscall" +) + +// State contains the state of a terminal. +type State struct { + termios unix.Termios +} + +// IsTerminal returns whether the given file descriptor is a terminal. +func IsTerminal(fd int) bool { + _, err := unix.IoctlGetTermio(fd, unix.TCGETA) + return err == nil +} + +// ReadPassword reads a line of input from a terminal without local echo. This +// is commonly used for inputting passwords and other sensitive data. The slice +// returned does not include the \n. +func ReadPassword(fd int) ([]byte, error) { + // see also: http://src.illumos.org/source/xref/illumos-gate/usr/src/lib/libast/common/uwin/getpass.c + val, err := unix.IoctlGetTermios(fd, unix.TCGETS) + if err != nil { + return nil, err + } + oldState := *val + + newState := oldState + newState.Lflag &^= syscall.ECHO + newState.Lflag |= syscall.ICANON | syscall.ISIG + newState.Iflag |= syscall.ICRNL + err = unix.IoctlSetTermios(fd, unix.TCSETS, &newState) + if err != nil { + return nil, err + } + + defer unix.IoctlSetTermios(fd, unix.TCSETS, &oldState) + + var buf [16]byte + var ret []byte + for { + n, err := syscall.Read(fd, buf[:]) + if err != nil { + return nil, err + } + if n == 0 { + if len(ret) == 0 { + return nil, io.EOF + } + break + } + if buf[n-1] == '\n' { + n-- + } + ret = append(ret, buf[:n]...) + if n < len(buf) { + break + } + } + + return ret, nil +} + +// MakeRaw puts the terminal connected to the given file descriptor into raw +// mode and returns the previous state of the terminal so that it can be +// restored. +// see http://cr.illumos.org/~webrev/andy_js/1060/ +func MakeRaw(fd int) (*State, error) { + termios, err := unix.IoctlGetTermios(fd, unix.TCGETS) + if err != nil { + return nil, err + } + + oldState := State{termios: *termios} + + termios.Iflag &^= unix.IGNBRK | unix.BRKINT | unix.PARMRK | unix.ISTRIP | unix.INLCR | unix.IGNCR | unix.ICRNL | unix.IXON + termios.Oflag &^= unix.OPOST + termios.Lflag &^= unix.ECHO | unix.ECHONL | unix.ICANON | unix.ISIG | unix.IEXTEN + termios.Cflag &^= unix.CSIZE | unix.PARENB + termios.Cflag |= unix.CS8 + termios.Cc[unix.VMIN] = 1 + termios.Cc[unix.VTIME] = 0 + + if err := unix.IoctlSetTermios(fd, unix.TCSETS, termios); err != nil { + return nil, err + } + + return &oldState, nil +} + +// Restore restores the terminal connected to the given file descriptor to a +// previous state. +func Restore(fd int, oldState *State) error { + return unix.IoctlSetTermios(fd, unix.TCSETS, &oldState.termios) +} + +// GetState returns the current state of a terminal which may be useful to +// restore the terminal after a signal. +func GetState(fd int) (*State, error) { + termios, err := unix.IoctlGetTermios(fd, unix.TCGETS) + if err != nil { + return nil, err + } + + return &State{termios: *termios}, nil +} + +// GetSize returns the dimensions of the given terminal. +func GetSize(fd int) (width, height int, err error) { + ws, err := unix.IoctlGetWinsize(fd, unix.TIOCGWINSZ) + if err != nil { + return 0, 0, err + } + return int(ws.Col), int(ws.Row), nil +} diff --git a/vendor/golang.org/x/crypto/ssh/terminal/util_windows.go b/vendor/golang.org/x/crypto/ssh/terminal/util_windows.go new file mode 100644 index 0000000..5cfdf8f --- /dev/null +++ b/vendor/golang.org/x/crypto/ssh/terminal/util_windows.go @@ -0,0 +1,105 @@ +// Copyright 2011 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build windows + +// Package terminal provides support functions for dealing with terminals, as +// commonly found on UNIX systems. +// +// Putting a terminal into raw mode is the most common requirement: +// +// oldState, err := terminal.MakeRaw(0) +// if err != nil { +// panic(err) +// } +// defer terminal.Restore(0, oldState) +package terminal + +import ( + "os" + + "golang.org/x/sys/windows" +) + +type State struct { + mode uint32 +} + +// IsTerminal returns whether the given file descriptor is a terminal. +func IsTerminal(fd int) bool { + var st uint32 + err := windows.GetConsoleMode(windows.Handle(fd), &st) + return err == nil +} + +// MakeRaw put the terminal connected to the given file descriptor into raw +// mode and returns the previous state of the terminal so that it can be +// restored. +func MakeRaw(fd int) (*State, error) { + var st uint32 + if err := windows.GetConsoleMode(windows.Handle(fd), &st); err != nil { + return nil, err + } + raw := st &^ (windows.ENABLE_ECHO_INPUT | windows.ENABLE_PROCESSED_INPUT | windows.ENABLE_LINE_INPUT | windows.ENABLE_PROCESSED_OUTPUT) + if err := windows.SetConsoleMode(windows.Handle(fd), raw); err != nil { + return nil, err + } + return &State{st}, nil +} + +// GetState returns the current state of a terminal which may be useful to +// restore the terminal after a signal. +func GetState(fd int) (*State, error) { + var st uint32 + if err := windows.GetConsoleMode(windows.Handle(fd), &st); err != nil { + return nil, err + } + return &State{st}, nil +} + +// Restore restores the terminal connected to the given file descriptor to a +// previous state. +func Restore(fd int, state *State) error { + return windows.SetConsoleMode(windows.Handle(fd), state.mode) +} + +// GetSize returns the visible dimensions of the given terminal. +// +// These dimensions don't include any scrollback buffer height. +func GetSize(fd int) (width, height int, err error) { + var info windows.ConsoleScreenBufferInfo + if err := windows.GetConsoleScreenBufferInfo(windows.Handle(fd), &info); err != nil { + return 0, 0, err + } + return int(info.Window.Right - info.Window.Left + 1), int(info.Window.Bottom - info.Window.Top + 1), nil +} + +// ReadPassword reads a line of input from a terminal without local echo. This +// is commonly used for inputting passwords and other sensitive data. The slice +// returned does not include the \n. +func ReadPassword(fd int) ([]byte, error) { + var st uint32 + if err := windows.GetConsoleMode(windows.Handle(fd), &st); err != nil { + return nil, err + } + old := st + + st &^= (windows.ENABLE_ECHO_INPUT) + st |= (windows.ENABLE_PROCESSED_INPUT | windows.ENABLE_LINE_INPUT | windows.ENABLE_PROCESSED_OUTPUT) + if err := windows.SetConsoleMode(windows.Handle(fd), st); err != nil { + return nil, err + } + + defer windows.SetConsoleMode(windows.Handle(fd), old) + + var h windows.Handle + p, _ := windows.GetCurrentProcess() + if err := windows.DuplicateHandle(p, windows.Handle(fd), p, &h, 0, false, windows.DUPLICATE_SAME_ACCESS); err != nil { + return nil, err + } + + f := os.NewFile(uintptr(h), "stdin") + defer f.Close() + return readPasswordLine(f) +} diff --git a/vendor/golang.org/x/sys/AUTHORS b/vendor/golang.org/x/sys/AUTHORS new file mode 100644 index 0000000..15167cd --- /dev/null +++ b/vendor/golang.org/x/sys/AUTHORS @@ -0,0 +1,3 @@ +# This source code refers to The Go Authors for copyright purposes. +# The master list of authors is in the main Go distribution, +# visible at http://tip.golang.org/AUTHORS. diff --git a/vendor/golang.org/x/sys/CONTRIBUTORS b/vendor/golang.org/x/sys/CONTRIBUTORS new file mode 100644 index 0000000..1c4577e --- /dev/null +++ b/vendor/golang.org/x/sys/CONTRIBUTORS @@ -0,0 +1,3 @@ +# This source code was written by the Go contributors. +# The master list of contributors is in the main Go distribution, +# visible at http://tip.golang.org/CONTRIBUTORS. diff --git a/vendor/golang.org/x/sys/LICENSE b/vendor/golang.org/x/sys/LICENSE new file mode 100644 index 0000000..6a66aea --- /dev/null +++ b/vendor/golang.org/x/sys/LICENSE @@ -0,0 +1,27 @@ +Copyright (c) 2009 The Go Authors. All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are +met: + + * Redistributions of source code must retain the above copyright +notice, this list of conditions and the following disclaimer. + * Redistributions in binary form must reproduce the above +copyright notice, this list of conditions and the following disclaimer +in the documentation and/or other materials provided with the +distribution. + * Neither the name of Google Inc. nor the names of its +contributors may be used to endorse or promote products derived from +this software without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT +LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, +DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY +THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. diff --git a/vendor/golang.org/x/sys/PATENTS b/vendor/golang.org/x/sys/PATENTS new file mode 100644 index 0000000..7330990 --- /dev/null +++ b/vendor/golang.org/x/sys/PATENTS @@ -0,0 +1,22 @@ +Additional IP Rights Grant (Patents) + +"This implementation" means the copyrightable works distributed by +Google as part of the Go project. + +Google hereby grants to You a perpetual, worldwide, non-exclusive, +no-charge, royalty-free, irrevocable (except as stated in this section) +patent license to make, have made, use, offer to sell, sell, import, +transfer and otherwise run, modify and propagate the contents of this +implementation of Go, where such license applies only to those patent +claims, both currently owned or controlled by Google and acquired in +the future, licensable by Google that are necessarily infringed by this +implementation of Go. This grant does not include claims that would be +infringed only as a consequence of further modification of this +implementation. If you or your agent or exclusive licensee institute or +order or agree to the institution of patent litigation against any +entity (including a cross-claim or counterclaim in a lawsuit) alleging +that this implementation of Go or any code incorporated within this +implementation of Go constitutes direct or contributory patent +infringement, or inducement of patent infringement, then any patent +rights granted to you under this License for this implementation of Go +shall terminate as of the date such litigation is filed. diff --git a/vendor/golang.org/x/sys/cpu/byteorder.go b/vendor/golang.org/x/sys/cpu/byteorder.go new file mode 100644 index 0000000..da6b9e4 --- /dev/null +++ b/vendor/golang.org/x/sys/cpu/byteorder.go @@ -0,0 +1,30 @@ +// Copyright 2019 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package cpu + +import ( + "encoding/binary" + "runtime" +) + +// hostByteOrder returns binary.LittleEndian on little-endian machines and +// binary.BigEndian on big-endian machines. +func hostByteOrder() binary.ByteOrder { + switch runtime.GOARCH { + case "386", "amd64", "amd64p32", + "arm", "arm64", + "mipsle", "mips64le", "mips64p32le", + "ppc64le", + "riscv", "riscv64": + return binary.LittleEndian + case "armbe", "arm64be", + "mips", "mips64", "mips64p32", + "ppc", "ppc64", + "s390", "s390x", + "sparc", "sparc64": + return binary.BigEndian + } + panic("unknown architecture") +} diff --git a/vendor/golang.org/x/sys/cpu/cpu.go b/vendor/golang.org/x/sys/cpu/cpu.go new file mode 100644 index 0000000..679e78c --- /dev/null +++ b/vendor/golang.org/x/sys/cpu/cpu.go @@ -0,0 +1,126 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// Package cpu implements processor feature detection for +// various CPU architectures. +package cpu + +// Initialized reports whether the CPU features were initialized. +// +// For some GOOS/GOARCH combinations initialization of the CPU features depends +// on reading an operating specific file, e.g. /proc/self/auxv on linux/arm +// Initialized will report false if reading the file fails. +var Initialized bool + +// CacheLinePad is used to pad structs to avoid false sharing. +type CacheLinePad struct{ _ [cacheLineSize]byte } + +// X86 contains the supported CPU features of the +// current X86/AMD64 platform. If the current platform +// is not X86/AMD64 then all feature flags are false. +// +// X86 is padded to avoid false sharing. Further the HasAVX +// and HasAVX2 are only set if the OS supports XMM and YMM +// registers in addition to the CPUID feature bit being set. +var X86 struct { + _ CacheLinePad + HasAES bool // AES hardware implementation (AES NI) + HasADX bool // Multi-precision add-carry instruction extensions + HasAVX bool // Advanced vector extension + HasAVX2 bool // Advanced vector extension 2 + HasBMI1 bool // Bit manipulation instruction set 1 + HasBMI2 bool // Bit manipulation instruction set 2 + HasERMS bool // Enhanced REP for MOVSB and STOSB + HasFMA bool // Fused-multiply-add instructions + HasOSXSAVE bool // OS supports XSAVE/XRESTOR for saving/restoring XMM registers. + HasPCLMULQDQ bool // PCLMULQDQ instruction - most often used for AES-GCM + HasPOPCNT bool // Hamming weight instruction POPCNT. + HasRDRAND bool // RDRAND instruction (on-chip random number generator) + HasRDSEED bool // RDSEED instruction (on-chip random number generator) + HasSSE2 bool // Streaming SIMD extension 2 (always available on amd64) + HasSSE3 bool // Streaming SIMD extension 3 + HasSSSE3 bool // Supplemental streaming SIMD extension 3 + HasSSE41 bool // Streaming SIMD extension 4 and 4.1 + HasSSE42 bool // Streaming SIMD extension 4 and 4.2 + _ CacheLinePad +} + +// ARM64 contains the supported CPU features of the +// current ARMv8(aarch64) platform. If the current platform +// is not arm64 then all feature flags are false. +var ARM64 struct { + _ CacheLinePad + HasFP bool // Floating-point instruction set (always available) + HasASIMD bool // Advanced SIMD (always available) + HasEVTSTRM bool // Event stream support + HasAES bool // AES hardware implementation + HasPMULL bool // Polynomial multiplication instruction set + HasSHA1 bool // SHA1 hardware implementation + HasSHA2 bool // SHA2 hardware implementation + HasCRC32 bool // CRC32 hardware implementation + HasATOMICS bool // Atomic memory operation instruction set + HasFPHP bool // Half precision floating-point instruction set + HasASIMDHP bool // Advanced SIMD half precision instruction set + HasCPUID bool // CPUID identification scheme registers + HasASIMDRDM bool // Rounding double multiply add/subtract instruction set + HasJSCVT bool // Javascript conversion from floating-point to integer + HasFCMA bool // Floating-point multiplication and addition of complex numbers + HasLRCPC bool // Release Consistent processor consistent support + HasDCPOP bool // Persistent memory support + HasSHA3 bool // SHA3 hardware implementation + HasSM3 bool // SM3 hardware implementation + HasSM4 bool // SM4 hardware implementation + HasASIMDDP bool // Advanced SIMD double precision instruction set + HasSHA512 bool // SHA512 hardware implementation + HasSVE bool // Scalable Vector Extensions + HasASIMDFHM bool // Advanced SIMD multiplication FP16 to FP32 + _ CacheLinePad +} + +// PPC64 contains the supported CPU features of the current ppc64/ppc64le platforms. +// If the current platform is not ppc64/ppc64le then all feature flags are false. +// +// For ppc64/ppc64le, it is safe to check only for ISA level starting on ISA v3.00, +// since there are no optional categories. There are some exceptions that also +// require kernel support to work (DARN, SCV), so there are feature bits for +// those as well. The minimum processor requirement is POWER8 (ISA 2.07). +// The struct is padded to avoid false sharing. +var PPC64 struct { + _ CacheLinePad + HasDARN bool // Hardware random number generator (requires kernel enablement) + HasSCV bool // Syscall vectored (requires kernel enablement) + IsPOWER8 bool // ISA v2.07 (POWER8) + IsPOWER9 bool // ISA v3.00 (POWER9) + _ CacheLinePad +} + +// S390X contains the supported CPU features of the current IBM Z +// (s390x) platform. If the current platform is not IBM Z then all +// feature flags are false. +// +// S390X is padded to avoid false sharing. Further HasVX is only set +// if the OS supports vector registers in addition to the STFLE +// feature bit being set. +var S390X struct { + _ CacheLinePad + HasZARCH bool // z/Architecture mode is active [mandatory] + HasSTFLE bool // store facility list extended + HasLDISP bool // long (20-bit) displacements + HasEIMM bool // 32-bit immediates + HasDFP bool // decimal floating point + HasETF3EH bool // ETF-3 enhanced + HasMSA bool // message security assist (CPACF) + HasAES bool // KM-AES{128,192,256} functions + HasAESCBC bool // KMC-AES{128,192,256} functions + HasAESCTR bool // KMCTR-AES{128,192,256} functions + HasAESGCM bool // KMA-GCM-AES{128,192,256} functions + HasGHASH bool // KIMD-GHASH function + HasSHA1 bool // K{I,L}MD-SHA-1 functions + HasSHA256 bool // K{I,L}MD-SHA-256 functions + HasSHA512 bool // K{I,L}MD-SHA-512 functions + HasSHA3 bool // K{I,L}MD-SHA3-{224,256,384,512} and K{I,L}MD-SHAKE-{128,256} functions + HasVX bool // vector facility + HasVXE bool // vector-enhancements facility 1 + _ CacheLinePad +} diff --git a/vendor/golang.org/x/sys/cpu/cpu_aix_ppc64.go b/vendor/golang.org/x/sys/cpu/cpu_aix_ppc64.go new file mode 100644 index 0000000..d8c26a0 --- /dev/null +++ b/vendor/golang.org/x/sys/cpu/cpu_aix_ppc64.go @@ -0,0 +1,30 @@ +// Copyright 2019 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build aix,ppc64 + +package cpu + +import "golang.org/x/sys/unix" + +const cacheLineSize = 128 + +const ( + // getsystemcfg constants + _SC_IMPL = 2 + _IMPL_POWER8 = 0x10000 + _IMPL_POWER9 = 0x20000 +) + +func init() { + impl := unix.Getsystemcfg(_SC_IMPL) + if impl&_IMPL_POWER8 != 0 { + PPC64.IsPOWER8 = true + } + if impl&_IMPL_POWER9 != 0 { + PPC64.IsPOWER9 = true + } + + Initialized = true +} diff --git a/vendor/golang.org/x/sys/cpu/cpu_arm.go b/vendor/golang.org/x/sys/cpu/cpu_arm.go new file mode 100644 index 0000000..7f2348b --- /dev/null +++ b/vendor/golang.org/x/sys/cpu/cpu_arm.go @@ -0,0 +1,9 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package cpu + +const cacheLineSize = 32 + +func doinit() {} diff --git a/vendor/golang.org/x/sys/cpu/cpu_gc_s390x.go b/vendor/golang.org/x/sys/cpu/cpu_gc_s390x.go new file mode 100644 index 0000000..568bcd0 --- /dev/null +++ b/vendor/golang.org/x/sys/cpu/cpu_gc_s390x.go @@ -0,0 +1,21 @@ +// Copyright 2019 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build !gccgo + +package cpu + +// haveAsmFunctions reports whether the other functions in this file can +// be safely called. +func haveAsmFunctions() bool { return true } + +// The following feature detection functions are defined in cpu_s390x.s. +// They are likely to be expensive to call so the results should be cached. +func stfle() facilityList +func kmQuery() queryResult +func kmcQuery() queryResult +func kmctrQuery() queryResult +func kmaQuery() queryResult +func kimdQuery() queryResult +func klmdQuery() queryResult diff --git a/vendor/golang.org/x/sys/cpu/cpu_gc_x86.go b/vendor/golang.org/x/sys/cpu/cpu_gc_x86.go new file mode 100644 index 0000000..f7cb469 --- /dev/null +++ b/vendor/golang.org/x/sys/cpu/cpu_gc_x86.go @@ -0,0 +1,16 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build 386 amd64 amd64p32 +// +build !gccgo + +package cpu + +// cpuid is implemented in cpu_x86.s for gc compiler +// and in cpu_gccgo.c for gccgo. +func cpuid(eaxArg, ecxArg uint32) (eax, ebx, ecx, edx uint32) + +// xgetbv with ecx = 0 is implemented in cpu_x86.s for gc compiler +// and in cpu_gccgo.c for gccgo. +func xgetbv() (eax, edx uint32) diff --git a/vendor/golang.org/x/sys/cpu/cpu_gccgo.c b/vendor/golang.org/x/sys/cpu/cpu_gccgo.c new file mode 100644 index 0000000..e363c7d --- /dev/null +++ b/vendor/golang.org/x/sys/cpu/cpu_gccgo.c @@ -0,0 +1,43 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build 386 amd64 amd64p32 +// +build gccgo + +#include +#include + +// Need to wrap __get_cpuid_count because it's declared as static. +int +gccgoGetCpuidCount(uint32_t leaf, uint32_t subleaf, + uint32_t *eax, uint32_t *ebx, + uint32_t *ecx, uint32_t *edx) +{ + return __get_cpuid_count(leaf, subleaf, eax, ebx, ecx, edx); +} + +// xgetbv reads the contents of an XCR (Extended Control Register) +// specified in the ECX register into registers EDX:EAX. +// Currently, the only supported value for XCR is 0. +// +// TODO: Replace with a better alternative: +// +// #include +// +// #pragma GCC target("xsave") +// +// void gccgoXgetbv(uint32_t *eax, uint32_t *edx) { +// unsigned long long x = _xgetbv(0); +// *eax = x & 0xffffffff; +// *edx = (x >> 32) & 0xffffffff; +// } +// +// Note that _xgetbv is defined starting with GCC 8. +void +gccgoXgetbv(uint32_t *eax, uint32_t *edx) +{ + __asm(" xorl %%ecx, %%ecx\n" + " xgetbv" + : "=a"(*eax), "=d"(*edx)); +} diff --git a/vendor/golang.org/x/sys/cpu/cpu_gccgo.go b/vendor/golang.org/x/sys/cpu/cpu_gccgo.go new file mode 100644 index 0000000..ba49b91 --- /dev/null +++ b/vendor/golang.org/x/sys/cpu/cpu_gccgo.go @@ -0,0 +1,26 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build 386 amd64 amd64p32 +// +build gccgo + +package cpu + +//extern gccgoGetCpuidCount +func gccgoGetCpuidCount(eaxArg, ecxArg uint32, eax, ebx, ecx, edx *uint32) + +func cpuid(eaxArg, ecxArg uint32) (eax, ebx, ecx, edx uint32) { + var a, b, c, d uint32 + gccgoGetCpuidCount(eaxArg, ecxArg, &a, &b, &c, &d) + return a, b, c, d +} + +//extern gccgoXgetbv +func gccgoXgetbv(eax, edx *uint32) + +func xgetbv() (eax, edx uint32) { + var a, d uint32 + gccgoXgetbv(&a, &d) + return a, d +} diff --git a/vendor/golang.org/x/sys/cpu/cpu_gccgo_s390x.go b/vendor/golang.org/x/sys/cpu/cpu_gccgo_s390x.go new file mode 100644 index 0000000..aa986f7 --- /dev/null +++ b/vendor/golang.org/x/sys/cpu/cpu_gccgo_s390x.go @@ -0,0 +1,22 @@ +// Copyright 2019 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build gccgo + +package cpu + +// haveAsmFunctions reports whether the other functions in this file can +// be safely called. +func haveAsmFunctions() bool { return false } + +// TODO(mundaym): the following feature detection functions are currently +// stubs. See https://golang.org/cl/162887 for how to fix this. +// They are likely to be expensive to call so the results should be cached. +func stfle() facilityList { panic("not implemented for gccgo") } +func kmQuery() queryResult { panic("not implemented for gccgo") } +func kmcQuery() queryResult { panic("not implemented for gccgo") } +func kmctrQuery() queryResult { panic("not implemented for gccgo") } +func kmaQuery() queryResult { panic("not implemented for gccgo") } +func kimdQuery() queryResult { panic("not implemented for gccgo") } +func klmdQuery() queryResult { panic("not implemented for gccgo") } diff --git a/vendor/golang.org/x/sys/cpu/cpu_linux.go b/vendor/golang.org/x/sys/cpu/cpu_linux.go new file mode 100644 index 0000000..76b5f50 --- /dev/null +++ b/vendor/golang.org/x/sys/cpu/cpu_linux.go @@ -0,0 +1,59 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +//+build !amd64,!amd64p32,!386 + +package cpu + +import ( + "io/ioutil" +) + +const ( + _AT_HWCAP = 16 + _AT_HWCAP2 = 26 + + procAuxv = "/proc/self/auxv" + + uintSize = int(32 << (^uint(0) >> 63)) +) + +// For those platforms don't have a 'cpuid' equivalent we use HWCAP/HWCAP2 +// These are initialized in cpu_$GOARCH.go +// and should not be changed after they are initialized. +var hwCap uint +var hwCap2 uint + +func init() { + buf, err := ioutil.ReadFile(procAuxv) + if err != nil { + // e.g. on android /proc/self/auxv is not accessible, so silently + // ignore the error and leave Initialized = false + return + } + + bo := hostByteOrder() + for len(buf) >= 2*(uintSize/8) { + var tag, val uint + switch uintSize { + case 32: + tag = uint(bo.Uint32(buf[0:])) + val = uint(bo.Uint32(buf[4:])) + buf = buf[8:] + case 64: + tag = uint(bo.Uint64(buf[0:])) + val = uint(bo.Uint64(buf[8:])) + buf = buf[16:] + } + switch tag { + case _AT_HWCAP: + hwCap = val + case _AT_HWCAP2: + hwCap2 = val + } + } + doinit() + + Initialized = true +} diff --git a/vendor/golang.org/x/sys/cpu/cpu_linux_arm64.go b/vendor/golang.org/x/sys/cpu/cpu_linux_arm64.go new file mode 100644 index 0000000..fa7fb1b --- /dev/null +++ b/vendor/golang.org/x/sys/cpu/cpu_linux_arm64.go @@ -0,0 +1,67 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package cpu + +const cacheLineSize = 64 + +// HWCAP/HWCAP2 bits. These are exposed by Linux. +const ( + hwcap_FP = 1 << 0 + hwcap_ASIMD = 1 << 1 + hwcap_EVTSTRM = 1 << 2 + hwcap_AES = 1 << 3 + hwcap_PMULL = 1 << 4 + hwcap_SHA1 = 1 << 5 + hwcap_SHA2 = 1 << 6 + hwcap_CRC32 = 1 << 7 + hwcap_ATOMICS = 1 << 8 + hwcap_FPHP = 1 << 9 + hwcap_ASIMDHP = 1 << 10 + hwcap_CPUID = 1 << 11 + hwcap_ASIMDRDM = 1 << 12 + hwcap_JSCVT = 1 << 13 + hwcap_FCMA = 1 << 14 + hwcap_LRCPC = 1 << 15 + hwcap_DCPOP = 1 << 16 + hwcap_SHA3 = 1 << 17 + hwcap_SM3 = 1 << 18 + hwcap_SM4 = 1 << 19 + hwcap_ASIMDDP = 1 << 20 + hwcap_SHA512 = 1 << 21 + hwcap_SVE = 1 << 22 + hwcap_ASIMDFHM = 1 << 23 +) + +func doinit() { + // HWCAP feature bits + ARM64.HasFP = isSet(hwCap, hwcap_FP) + ARM64.HasASIMD = isSet(hwCap, hwcap_ASIMD) + ARM64.HasEVTSTRM = isSet(hwCap, hwcap_EVTSTRM) + ARM64.HasAES = isSet(hwCap, hwcap_AES) + ARM64.HasPMULL = isSet(hwCap, hwcap_PMULL) + ARM64.HasSHA1 = isSet(hwCap, hwcap_SHA1) + ARM64.HasSHA2 = isSet(hwCap, hwcap_SHA2) + ARM64.HasCRC32 = isSet(hwCap, hwcap_CRC32) + ARM64.HasATOMICS = isSet(hwCap, hwcap_ATOMICS) + ARM64.HasFPHP = isSet(hwCap, hwcap_FPHP) + ARM64.HasASIMDHP = isSet(hwCap, hwcap_ASIMDHP) + ARM64.HasCPUID = isSet(hwCap, hwcap_CPUID) + ARM64.HasASIMDRDM = isSet(hwCap, hwcap_ASIMDRDM) + ARM64.HasJSCVT = isSet(hwCap, hwcap_JSCVT) + ARM64.HasFCMA = isSet(hwCap, hwcap_FCMA) + ARM64.HasLRCPC = isSet(hwCap, hwcap_LRCPC) + ARM64.HasDCPOP = isSet(hwCap, hwcap_DCPOP) + ARM64.HasSHA3 = isSet(hwCap, hwcap_SHA3) + ARM64.HasSM3 = isSet(hwCap, hwcap_SM3) + ARM64.HasSM4 = isSet(hwCap, hwcap_SM4) + ARM64.HasASIMDDP = isSet(hwCap, hwcap_ASIMDDP) + ARM64.HasSHA512 = isSet(hwCap, hwcap_SHA512) + ARM64.HasSVE = isSet(hwCap, hwcap_SVE) + ARM64.HasASIMDFHM = isSet(hwCap, hwcap_ASIMDFHM) +} + +func isSet(hwc uint, value uint) bool { + return hwc&value != 0 +} diff --git a/vendor/golang.org/x/sys/cpu/cpu_linux_ppc64x.go b/vendor/golang.org/x/sys/cpu/cpu_linux_ppc64x.go new file mode 100644 index 0000000..6c8d975 --- /dev/null +++ b/vendor/golang.org/x/sys/cpu/cpu_linux_ppc64x.go @@ -0,0 +1,33 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build linux +// +build ppc64 ppc64le + +package cpu + +const cacheLineSize = 128 + +// HWCAP/HWCAP2 bits. These are exposed by the kernel. +const ( + // ISA Level + _PPC_FEATURE2_ARCH_2_07 = 0x80000000 + _PPC_FEATURE2_ARCH_3_00 = 0x00800000 + + // CPU features + _PPC_FEATURE2_DARN = 0x00200000 + _PPC_FEATURE2_SCV = 0x00100000 +) + +func doinit() { + // HWCAP2 feature bits + PPC64.IsPOWER8 = isSet(hwCap2, _PPC_FEATURE2_ARCH_2_07) + PPC64.IsPOWER9 = isSet(hwCap2, _PPC_FEATURE2_ARCH_3_00) + PPC64.HasDARN = isSet(hwCap2, _PPC_FEATURE2_DARN) + PPC64.HasSCV = isSet(hwCap2, _PPC_FEATURE2_SCV) +} + +func isSet(hwc uint, value uint) bool { + return hwc&value != 0 +} diff --git a/vendor/golang.org/x/sys/cpu/cpu_linux_s390x.go b/vendor/golang.org/x/sys/cpu/cpu_linux_s390x.go new file mode 100644 index 0000000..d579eae --- /dev/null +++ b/vendor/golang.org/x/sys/cpu/cpu_linux_s390x.go @@ -0,0 +1,161 @@ +// Copyright 2019 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +package cpu + +const cacheLineSize = 256 + +const ( + // bit mask values from /usr/include/bits/hwcap.h + hwcap_ZARCH = 2 + hwcap_STFLE = 4 + hwcap_MSA = 8 + hwcap_LDISP = 16 + hwcap_EIMM = 32 + hwcap_DFP = 64 + hwcap_ETF3EH = 256 + hwcap_VX = 2048 + hwcap_VXE = 8192 +) + +// bitIsSet reports whether the bit at index is set. The bit index +// is in big endian order, so bit index 0 is the leftmost bit. +func bitIsSet(bits []uint64, index uint) bool { + return bits[index/64]&((1<<63)>>(index%64)) != 0 +} + +// function is the code for the named cryptographic function. +type function uint8 + +const ( + // KM{,A,C,CTR} function codes + aes128 function = 18 // AES-128 + aes192 function = 19 // AES-192 + aes256 function = 20 // AES-256 + + // K{I,L}MD function codes + sha1 function = 1 // SHA-1 + sha256 function = 2 // SHA-256 + sha512 function = 3 // SHA-512 + sha3_224 function = 32 // SHA3-224 + sha3_256 function = 33 // SHA3-256 + sha3_384 function = 34 // SHA3-384 + sha3_512 function = 35 // SHA3-512 + shake128 function = 36 // SHAKE-128 + shake256 function = 37 // SHAKE-256 + + // KLMD function codes + ghash function = 65 // GHASH +) + +// queryResult contains the result of a Query function +// call. Bits are numbered in big endian order so the +// leftmost bit (the MSB) is at index 0. +type queryResult struct { + bits [2]uint64 +} + +// Has reports whether the given functions are present. +func (q *queryResult) Has(fns ...function) bool { + if len(fns) == 0 { + panic("no function codes provided") + } + for _, f := range fns { + if !bitIsSet(q.bits[:], uint(f)) { + return false + } + } + return true +} + +// facility is a bit index for the named facility. +type facility uint8 + +const ( + // cryptography facilities + msa4 facility = 77 // message-security-assist extension 4 + msa8 facility = 146 // message-security-assist extension 8 +) + +// facilityList contains the result of an STFLE call. +// Bits are numbered in big endian order so the +// leftmost bit (the MSB) is at index 0. +type facilityList struct { + bits [4]uint64 +} + +// Has reports whether the given facilities are present. +func (s *facilityList) Has(fs ...facility) bool { + if len(fs) == 0 { + panic("no facility bits provided") + } + for _, f := range fs { + if !bitIsSet(s.bits[:], uint(f)) { + return false + } + } + return true +} + +func doinit() { + // test HWCAP bit vector + has := func(featureMask uint) bool { + return hwCap&featureMask == featureMask + } + + // mandatory + S390X.HasZARCH = has(hwcap_ZARCH) + + // optional + S390X.HasSTFLE = has(hwcap_STFLE) + S390X.HasLDISP = has(hwcap_LDISP) + S390X.HasEIMM = has(hwcap_EIMM) + S390X.HasETF3EH = has(hwcap_ETF3EH) + S390X.HasDFP = has(hwcap_DFP) + S390X.HasMSA = has(hwcap_MSA) + S390X.HasVX = has(hwcap_VX) + if S390X.HasVX { + S390X.HasVXE = has(hwcap_VXE) + } + + // We need implementations of stfle, km and so on + // to detect cryptographic features. + if !haveAsmFunctions() { + return + } + + // optional cryptographic functions + if S390X.HasMSA { + aes := []function{aes128, aes192, aes256} + + // cipher message + km, kmc := kmQuery(), kmcQuery() + S390X.HasAES = km.Has(aes...) + S390X.HasAESCBC = kmc.Has(aes...) + if S390X.HasSTFLE { + facilities := stfle() + if facilities.Has(msa4) { + kmctr := kmctrQuery() + S390X.HasAESCTR = kmctr.Has(aes...) + } + if facilities.Has(msa8) { + kma := kmaQuery() + S390X.HasAESGCM = kma.Has(aes...) + } + } + + // compute message digest + kimd := kimdQuery() // intermediate (no padding) + klmd := klmdQuery() // last (padding) + S390X.HasSHA1 = kimd.Has(sha1) && klmd.Has(sha1) + S390X.HasSHA256 = kimd.Has(sha256) && klmd.Has(sha256) + S390X.HasSHA512 = kimd.Has(sha512) && klmd.Has(sha512) + S390X.HasGHASH = kimd.Has(ghash) // KLMD-GHASH does not exist + sha3 := []function{ + sha3_224, sha3_256, sha3_384, sha3_512, + shake128, shake256, + } + S390X.HasSHA3 = kimd.Has(sha3...) && klmd.Has(sha3...) + } +} diff --git a/vendor/golang.org/x/sys/cpu/cpu_mips64x.go b/vendor/golang.org/x/sys/cpu/cpu_mips64x.go new file mode 100644 index 0000000..f55e0c8 --- /dev/null +++ b/vendor/golang.org/x/sys/cpu/cpu_mips64x.go @@ -0,0 +1,11 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build mips64 mips64le + +package cpu + +const cacheLineSize = 32 + +func doinit() {} diff --git a/vendor/golang.org/x/sys/cpu/cpu_mipsx.go b/vendor/golang.org/x/sys/cpu/cpu_mipsx.go new file mode 100644 index 0000000..cda87b1 --- /dev/null +++ b/vendor/golang.org/x/sys/cpu/cpu_mipsx.go @@ -0,0 +1,11 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build mips mipsle + +package cpu + +const cacheLineSize = 32 + +func doinit() {} diff --git a/vendor/golang.org/x/sys/cpu/cpu_other_arm64.go b/vendor/golang.org/x/sys/cpu/cpu_other_arm64.go new file mode 100644 index 0000000..dd1e76d --- /dev/null +++ b/vendor/golang.org/x/sys/cpu/cpu_other_arm64.go @@ -0,0 +1,11 @@ +// Copyright 2019 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build !linux,arm64 + +package cpu + +const cacheLineSize = 64 + +func doinit() {} diff --git a/vendor/golang.org/x/sys/cpu/cpu_s390x.s b/vendor/golang.org/x/sys/cpu/cpu_s390x.s new file mode 100644 index 0000000..e5037d9 --- /dev/null +++ b/vendor/golang.org/x/sys/cpu/cpu_s390x.s @@ -0,0 +1,57 @@ +// Copyright 2019 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build !gccgo + +#include "textflag.h" + +// func stfle() facilityList +TEXT ·stfle(SB), NOSPLIT|NOFRAME, $0-32 + MOVD $ret+0(FP), R1 + MOVD $3, R0 // last doubleword index to store + XC $32, (R1), (R1) // clear 4 doublewords (32 bytes) + WORD $0xb2b01000 // store facility list extended (STFLE) + RET + +// func kmQuery() queryResult +TEXT ·kmQuery(SB), NOSPLIT|NOFRAME, $0-16 + MOVD $0, R0 // set function code to 0 (KM-Query) + MOVD $ret+0(FP), R1 // address of 16-byte return value + WORD $0xB92E0024 // cipher message (KM) + RET + +// func kmcQuery() queryResult +TEXT ·kmcQuery(SB), NOSPLIT|NOFRAME, $0-16 + MOVD $0, R0 // set function code to 0 (KMC-Query) + MOVD $ret+0(FP), R1 // address of 16-byte return value + WORD $0xB92F0024 // cipher message with chaining (KMC) + RET + +// func kmctrQuery() queryResult +TEXT ·kmctrQuery(SB), NOSPLIT|NOFRAME, $0-16 + MOVD $0, R0 // set function code to 0 (KMCTR-Query) + MOVD $ret+0(FP), R1 // address of 16-byte return value + WORD $0xB92D4024 // cipher message with counter (KMCTR) + RET + +// func kmaQuery() queryResult +TEXT ·kmaQuery(SB), NOSPLIT|NOFRAME, $0-16 + MOVD $0, R0 // set function code to 0 (KMA-Query) + MOVD $ret+0(FP), R1 // address of 16-byte return value + WORD $0xb9296024 // cipher message with authentication (KMA) + RET + +// func kimdQuery() queryResult +TEXT ·kimdQuery(SB), NOSPLIT|NOFRAME, $0-16 + MOVD $0, R0 // set function code to 0 (KIMD-Query) + MOVD $ret+0(FP), R1 // address of 16-byte return value + WORD $0xB93E0024 // compute intermediate message digest (KIMD) + RET + +// func klmdQuery() queryResult +TEXT ·klmdQuery(SB), NOSPLIT|NOFRAME, $0-16 + MOVD $0, R0 // set function code to 0 (KLMD-Query) + MOVD $ret+0(FP), R1 // address of 16-byte return value + WORD $0xB93F0024 // compute last message digest (KLMD) + RET diff --git a/vendor/golang.org/x/sys/cpu/cpu_wasm.go b/vendor/golang.org/x/sys/cpu/cpu_wasm.go new file mode 100644 index 0000000..bd9bbda --- /dev/null +++ b/vendor/golang.org/x/sys/cpu/cpu_wasm.go @@ -0,0 +1,15 @@ +// Copyright 2019 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build wasm + +package cpu + +// We're compiling the cpu package for an unknown (software-abstracted) CPU. +// Make CacheLinePad an empty struct and hope that the usual struct alignment +// rules are good enough. + +const cacheLineSize = 0 + +func doinit() {} diff --git a/vendor/golang.org/x/sys/cpu/cpu_x86.go b/vendor/golang.org/x/sys/cpu/cpu_x86.go new file mode 100644 index 0000000..d70d317 --- /dev/null +++ b/vendor/golang.org/x/sys/cpu/cpu_x86.go @@ -0,0 +1,59 @@ +// Copyright 2018 The Go Authors. All rights reserved. +// Use of this source code is governed by a BSD-style +// license that can be found in the LICENSE file. + +// +build 386 amd64 amd64p32 + +package cpu + +const cacheLineSize = 64 + +func init() { + Initialized = true + + maxID, _, _, _ := cpuid(0, 0) + + if maxID < 1 { + return + } + + _, _, ecx1, edx1 := cpuid(1, 0) + X86.HasSSE2 = isSet(26, edx1) + + X86.HasSSE3 = isSet(0, ecx1) + X86.HasPCLMULQDQ = isSet(1, ecx1) + X86.HasSSSE3 = isSet(9, ecx1) + X86.HasFMA = isSet(12, ecx1) + X86.HasSSE41 = isSet(19, ecx1) + X86.HasSSE42 = isSet(20, ecx1) + X86.HasPOPCNT = isSet(23, ecx1) + X86.HasAES = isSet(25, ecx1) + X86.HasOSXSAVE = isSet(27, ecx1) + X86.HasRDRAND = isSet(30, ecx1) + + osSupportsAVX := false + // For XGETBV, OSXSAVE bit is required and sufficient. + if X86.HasOSXSAVE { + eax, _ := xgetbv() + // Check if XMM and YMM registers have OS support. + osSupportsAVX = isSet(1, eax) && isSet(2, eax) + } + + X86.HasAVX = isSet(28, ecx1) && osSupportsAVX + + if maxID < 7 { + return + } + + _, ebx7, _, _ := cpuid(7, 0) + X86.HasBMI1 = isSet(3, ebx7) + X86.HasAVX2 = isSet(5, ebx7) && osSupportsAVX + X86.HasBMI2 = isSet(8, ebx7) + X86.HasERMS = isSet(9, ebx7) + X86.HasRDSEED = isSet(18, ebx7) + X86.HasADX = isSet(19, ebx7) +} + +func isSet(bitpos uint, value uint32) bool { + return value&(1<