[openssl.org #783] ASN1_INTEGER_get/ASN1_INTEGER_set patch

2003-12-02 Thread Frédéric Giudicelli via RT
After looking at ASN1_INTEGER_get and ASN1_INTEGER_set I noticed there were minor optimizations possible. There were all value assignements that had no reasons to be there. The patch is based on openssl-0.9.7c. Frédéric Giudicelli http://www.newpki.org

Re: bn_lib.c:bn_expand2()

2003-12-02 Thread Richard Levitte - VMS Whacker
In message [EMAIL PROTECTED] on Mon, 1 Dec 2003 17:27:46 -0500, Geoff Thorpe [EMAIL PROTECTED] said: geoff Hi Andy (and anyone else that's interested), geoff geoff As part of the general hackathon/audit we're doing in geoff crypto/bn/ I once again came across the curious zeroing code in geoff

Re: bn_lib.c:bn_expand2()

2003-12-02 Thread Geoff Thorpe
On December 2, 2003 05:19 am, Richard Levitte - VMS Whacker wrote: geoff I understand the desire to cater for CPU pipelining with the geoff 8-wise loop unrolling, but is this a better solution than just geoff using memset() and letting the compiler take care of the same geoff sort of thing?

Re: bn_lib.c:bn_expand2()

2003-12-02 Thread Richard Levitte - VMS Whacker
In message [EMAIL PROTECTED] on Tue, 2 Dec 2003 08:52:04 -0500, Geoff Thorpe [EMAIL PROTECTED] said: geoff If we switch to use the mem* variant in bn_expand_internal(), it needs geoff to be corrected. The current form is unbelievingly buggy (I didn't geoff even look at it last I played

Re: bn_lib.c:bn_expand2()

2003-12-02 Thread Geoff Thorpe
On December 2, 2003 09:35 am, Richard Levitte - VMS Whacker wrote: geoff afterwards by the memcpy(), and likewise the use of geoff sizeof(b-d[0]) seems strange given the logic in the previous geoff line uses sizeof(BN_ULONG) (which is better). But unbelievably geoff buggy? How so? A missing

Re: bn_lib.c:bn_expand2()

2003-12-02 Thread Peter Waltenberg
Comment: valgrind on x86 Linux picks up those off-by-one errors quite well if you want an extra screening test. For some reason there was a significant reduction in noise from valgrind in the transition from OpenSSL-0.9.7b-OpenSSL-0.9.7c. Peter Geoff Thorpe [EMAIL PROTECTED] Sent by:

[openssl.org #515] 0.9.7a

2003-12-02 Thread Lutz Jaenicke via RT
[EMAIL PROTECTED] - Thu Mar 20 12:01:22 2003]: I'd like to ask you to please consider fetching ftp://ftp.openssl.org/snapshot/openssl-0.9.7-stable-SNAP- 20030319.tar.gz and test it, to determine if we need to do more fixing *before* release of 0.9.7b (if possible). Otherwise, you

Re: [openssl.org #781] [PATCH] NetWare Support for OpenSSL 0.9.7

2003-12-02 Thread Verdon Walker via RT
Addressing the points in order: ERR_free_state_table() is not meant to do the same as ERR_remove_state(). It is indeed meant to clean up the entire table and is used for process cleanup, not thread cleanup. NetWare will clean up process resources, but does complain when you don't clean up after

Library cleanup functionality

2003-12-02 Thread Verdon Walker
Although OpenSSL has an initialization routine (SSL_library_init), it does not have a corresponding cleanup routine (SSL_library_cleanup). It does have a few cleanup routines (e.g. EVP_cleanup, OBJ_cleanup, etc.) but no overall cleanup function. This has been an issue for us when running OpenSSL

Re: Library cleanup functionality

2003-12-02 Thread Peter Waltenberg
Potentially useful, but ... A problem we frequently hit is multiple instances of libraries in a single process - often due to applications dragging in shared library dependencies they aren't explicitly aware of. For that reason, this is only mostly safe if you can reference count somehow,