Date: Thu, 21 Jun 2012 08:05:02 -0500 From: "jfoug" <jfoug@....net> To: <john-dev@...ts.openwall.com> Subject: RE: 1.7.9-jumbo-6 >From: Solar Designer [mailto:solar@...nwall.com] > >I think we're almost ready to release the current magnum-jumbo as 1.7.9- >jumbo-6. If you're aware of anything that needs to happen before we >make this release, please let me know (post to this thread). EPI and RACF need work. >Also, now is the time to proceed with testing on lots of platforms. >For example, I am unsure if my recent changes to ALGORITHM_NAME for sse- >intrinsics work as intended on 32-bit builds (-any, -mmx, -sse2) - can >anyone test that, please? Win32-cygwin-x86-any. Works. Algo's display correctly RACF and EPI fail self tests Win32-cygwin-x86-mmx. Works. (but mmx is almost never a real target these days anyway) Dynamic formats: [64/64 SSE2 intrinsics 64x2]. But I thought MMX used the .s files. Others show this: Benchmarking: Kerberos AFS DES [48/64 4K MMX]. Benchmarking: HTTP Digest access authentication MD5 [64/64 MMX 2x]. RACF and EPI fail self tests Win32-cygwin-x86-sse2 (I always use sse2i) Algo's display correctly RACF and EPI fail self tests This one does intrinsics, but compiles the .c file (slower with cygwin's gcc) Win32-cygwin-x86-sse2i (Uses ICC's .S build of the intrinsic file, much faster) Algo's display correctly RACF and EPI fail self tests Jim
Powered by blists - more mailing lists
Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.