Browse Source

Ok, there was a need for OPENSSL_cleanse() instead of bzero() to prevent

supposedly smart compilers from optimizing memory cleanups away. Understood.
Ok, in case of an hypothetically super smart compiler, OPENSSL_cleanse() had
to be convoluted enough for the compiler not to recognize that this was
actually bzero() in disguise. Understood.
But then why there had been optimized assembler versions of OPENSSL_cleanse()
is beyond me. Did someone not trust the C obfuscation?
OPENBSD_5_6
miod 10 years ago
parent
commit
7c11123d14
1 changed files with 2 additions and 2 deletions
  1. +2
    -2
      src/lib/libcrypto/crypto/Makefile

+ 2
- 2
src/lib/libcrypto/crypto/Makefile View File

@ -1,4 +1,4 @@
# $OpenBSD: Makefile,v 1.19 2014/04/17 18:11:46 miod Exp $
# $OpenBSD: Makefile,v 1.20 2014/04/17 18:16:45 miod Exp $
LIB= crypto LIB= crypto
@ -35,6 +35,7 @@ CFLAGS+= -I${LCRYPTO_SRC}/modes -I${LCRYPTO_SRC}/asn1 -I${LCRYPTO_SRC}/evp
# crypto/ # crypto/
SRCS+= cryptlib.c malloc-wrapper.c mem_dbg.c cversion.c ex_data.c cpt_err.c SRCS+= cryptlib.c malloc-wrapper.c mem_dbg.c cversion.c ex_data.c cpt_err.c
SRCS+= uid.c o_time.c o_str.c o_init.c SRCS+= uid.c o_time.c o_str.c o_init.c
SRCS+= mem_clr.c
# aes/ # aes/
SRCS+= aes_misc.c aes_ecb.c aes_cfb.c aes_ofb.c SRCS+= aes_misc.c aes_ecb.c aes_cfb.c aes_ofb.c
@ -432,7 +433,6 @@ obj_dat.h: obj_mac.h
.if empty (SSLASM) .if empty (SSLASM)
CFLAGS+=-DOPENSSL_NO_ASM CFLAGS+=-DOPENSSL_NO_ASM
SRCS+= mem_clr.c
SRCS+= aes_core.c aes_cbc.c SRCS+= aes_core.c aes_cbc.c
SRCS+= bf_enc.c SRCS+= bf_enc.c
SRCS+= bn_asm.c SRCS+= bn_asm.c


Loading…
Cancel
Save