diff options
| author |   deraadt <deraadt@openbsd.org> | 2017-12-11 05:48:01 +0000 |
|---|---|---|
| committer | deraadt <deraadt@openbsd.org> | 2017-12-11 05:48:01 +0000 |
| commit | 3f624d5f689b2964adb044dbb5971a0310c8869d (patch) | |
| tree | 54f5f3ab8d249de3ffc2608e6fc0375d50c09ee5 /lib/libcrypto/camellia | |
| parent | In uvm Chuck decided backing store would not be allocated proactively (diff) | |
| download | wireguard-openbsd-3f624d5f689b2964adb044dbb5971a0310c8869d.tar.xz wireguard-openbsd-3f624d5f689b2964adb044dbb5971a0310c8869d.zip | |
http://repzret.org/p/repzret/
My read of this: Long time ago (Think Conan, not dinasaurs) during the race
to make speedier processors, a cpu vendor built a pipeline with a bad stall,
and proposed a tremendously hasky workaround. A wizard adopted this into his
perl scroll, and failed to reflect later when no compiler adopted the practice.
This relic remains at the tail end of some functions in OpenSSL as
".byte 0xf3,0xc3". Banish it straight to hell.
ok mlarkin, others also stared blankly
Diffstat (limited to 'lib/libcrypto/camellia')
| -rw-r--r-- | lib/libcrypto/camellia/asm/cmll-x86_64.pl | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/lib/libcrypto/camellia/asm/cmll-x86_64.pl b/lib/libcrypto/camellia/asm/cmll-x86_64.pl index a171c654b2d..df6bf11a288 100644 --- a/lib/libcrypto/camellia/asm/cmll-x86_64.pl +++ b/lib/libcrypto/camellia/asm/cmll-x86_64.pl @@ -218,7 +218,7 @@ $code.=<<___; mov $t2,@S[2] mov $t3,@S[3] - .byte 0xf3,0xc3 # rep ret + retq .size _x86_64_Camellia_encrypt,.-_x86_64_Camellia_encrypt # V1.x API @@ -329,7 +329,7 @@ $code.=<<___; mov $t0,@S[2] mov $t1,@S[3] - .byte 0xf3,0xc3 # rep ret + retq .size _x86_64_Camellia_decrypt,.-_x86_64_Camellia_decrypt ___ |