summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2014-07-10Remove bogus preprocessor statements trying to pick the largest integerMiod Vallat
type for BF_LONG, MD[45]_LONG and SHA_LONG. First, the preprocessor symbols they check for a 64-bit system is __ILP64__ which no sane system provides; second, on the platforms which have assembler code to speed things up, the assembler code assumes a 32-bit type will be used.
2014-07-10crank major for decompression featureTed Unangst
2014-07-10decompress libssl. ok beck jsingTed Unangst
2014-07-10KNF comments, reflowing and moving out of the middle of argument lists inPhilip Guenther
places ok jsing@
2014-07-10remove unused ecc_pkey_size.Brent Cook
ok jsing@ miod@
2014-07-10BN_free, BN_clear_free, BN_CTX_free, BN_BLINDING_free and BN_MONT_CTX_freeJoel Sing
all have implicit NULL checks, so we do not need them here. ok miod@
2014-07-10Use a while loop instead of an ifdowhile loop.Joel Sing
ok miod@ tedu@
2014-07-09include <limits.h> for LONG_MIN/LONG_MAX.Brent Cook
Also remove <sys/filio.h> added from previous commit. This was the wrong way to get FIONBIO. ok jsing@
2014-07-09ASN1_STRING_free can handle NULL, so callers don't need to check. ok miodTed Unangst
2014-07-09reset host, port, path to null after freeing so the caller doesn'tTed Unangst
accidentally free them again. actually a bug in the caller and (hey hey) apps/ocsp.c has exactly that bug, but it's easier/safer to fix here.
2014-07-09stale prototypeTed Unangst
2014-07-09Make comments readable.Joel Sing
2014-07-09More KNF.Joel Sing
2014-07-09KNF.Joel Sing
2014-07-09ocsp_check_ids says "If algoritm mismatch let caller deal with it" beforeTed Unangst
returning 2. The one and only caller doesn't check for that, so... Stop returning it. ok miod
2014-07-09Update to match the current state of crypto/rand, and remove MLINKS forMiod Vallat
RAND_event and RAND_screen.
2014-07-09Remove mention that the PRNG needs to be seeded before invoking someMiod Vallat
functions.
2014-07-09Unifdef -UPKCS_TESTVECT - we don't want the random data used in OAEP paddingMiod Vallat
to get overwritten by a known value, ever.
2014-07-09{malloc,reallocarray} + memset(,0,) -> callocMiod Vallat
ok tedu@
2014-07-09Replace memset(a, 0, ...); free(a); by explicit_bzero(a, ...); free(a); inMiod Vallat
fear a smartass-optimizing compiler decides memset is useless immediately before free(). ok jsing@ deraadt@ tedu@
2014-07-09Remove leading underscore from _BYTE_ORDER and _{LITTLE,BIG}_ENDIAN, to beMiod Vallat
more friendly to systems where the underscore flavours may be defined as empty. Found the hard way be bcook@; joint brainstrom with bcook beck and guenther
2014-07-09miod@ fixed the "bug" so remove reference to it in the man pod^Wpage.Joel Sing
2014-07-09add <sys/filio.h> header for FIONBIOBrent Cook
ok beck@
2014-07-09Clean up and simplify SSL_CIPHER_description by always using asprintf. IfJoel Sing
a buffer was supplied then we copy the result into it. Also make the failure case return values match the documentation. Joint work with beck@
2014-07-09Simplify error path of DH_check_pub_key()Miod Vallat
2014-07-09KNFMiod Vallat
2014-07-09remove unused variables getentropy for OS XBrent Cook
ok beck@
2014-07-09tedu the SSL export cipher handling - since we do not have enabled exportJoel Sing
ciphers we no longer need the flags or code to support it. ok beck@ miod@
2014-07-09cast ASN1_STRING (unsigned char *) to match strlcat's argument (char *)Brent Cook
ok beck@
2014-07-09remove unused, private version strings except SSL_version_strBrent Cook
Also remove unused des_ver.h, which exports some of these strings, but is not installed. ok miod@ tedu@
2014-07-09Make use of this wonderful modern C construct known as a `switch', instead ofMiod Vallat
8-line if() tests.
2014-07-09KNFMiod Vallat
2014-07-09Kill more FIPS tentacles by removing the private_AES_set_{enc,dec}rypt_key()Miod Vallat
internal interfaces, and promoting them to being the public AES_set_{enc,dec}rypt_key() interfaces. In non-FIPS mode, these public interfaces were directly calling the private ones. ok guenther@ jsing@
2014-07-09Be more strict in RSA_padding_check_X931(), and thus avoid a possibleMiod Vallat
memcpy() with a negative size. ok tedu@
2014-07-09In the old days (not in this century), SSLeay 0.4.5 would create X.509 RSAMiod Vallat
signatures using the wrong oid for the signature type. The signature verification code has thus been modified to allow these signatures to be accepted, with a printf to stderr to notify the user something was fishy. Remove this chunk; these signatures will no longer get accepted. ok deraadt@ guenther@ jsing@ tedu@
2014-07-09Remove RSA_memory_lock(). This undocumented function sort-of serializes yourMiod Vallat
RSA components to memory and clears them, but there is no unserializing function, so its usefulness is close to zero. A grep through the ports tree sources show that it is only present in ports embedding their own openssl copy, and never used otherwise. ok jsing@
2014-07-09remove unused dynamic_cmd_defns_empty structure.Brent Cook
ok miod@ deraadt@ guenther@
2014-07-09RSA_NULL used to be a compile option allowing the RSA interfaces to beMiod Vallat
compiled-in, with nonfunctional code, to be able to cope with the RSA patent. However, we don't use this option, and the RSA patent has expired more than 10 years ago, so just drop this piece.
2014-07-09KNFMiod Vallat
2014-07-08Mark the weakened 40-bit export ciphers as invalid - no one in their rightJoel Sing
mind should be using them. ok deraadt@ miod@
2014-07-08Remove SSL_FIPS.Joel Sing
ok deraadt@ miod@
2014-07-08Nuke SSL_NOT_EXP since it does nothing.Joel Sing
ok deraadt@ miod@
2014-07-08replace malloc(strlen())/strlcpy with strdupBrent Cook
ok beck@ jsing@
2014-07-08Define SMALL_REGISTER_BANK on arm and vax, for it generates faster code forMiod Vallat
these systems (vax being 30% faster!). (surprisingly, the prime candidate for SMALL_REGISTER_BANK, SuperH, runs actually slower in that case)
2014-07-08Remove undocumented _des_crypt() interface and its companion header file,Miod Vallat
which had never been installed, so it's unlikely something ever used this in the last 15~20 years. ok deraadt@ jsing@ beck@
2014-07-08pedantic avoidance of division by zero, likely not actuallyBob Beck
possible to hit for real. ok miod@ jsing@
2014-07-08add missing stdint. include for uint32_t, etc.Brent Cook
ok beck@ jsing@
2014-07-08We do not care about broken Borland C pre-processors.Joel Sing
ok beck@ deraadt@
2014-07-08Pretty much everything has SO_REUSEADDR - nuke the #ifdefs.Joel Sing
ok beck@
2014-07-08We have EAI_FAMILY - remove the #ifdefs.Joel Sing
ok beck@