Wed, 13 May 2009 02:58:57 +0000 |
diego |
whitespace cosmetics: Remove all trailing whitespace.
|
Wed, 08 Apr 2009 20:21:21 +0000 |
ramiro |
Rename RUNTIME_CPUDETECT to CONFIG_RUNTIME_CPUDETECT and always define it.
|
Thu, 02 Apr 2009 09:20:48 +0000 |
diego |
Remove unnecessary malloc.h #includes and related #ifdeffery.
|
Fri, 27 Mar 2009 18:48:47 +0000 |
reimar |
Sync cpuid detection code with libavcodec: assume it is always available on x86_64
|
Thu, 12 Mar 2009 10:50:21 +0000 |
diego |
SSE3 support patch by Zhou Zongyi, zhouzongyi pset.suntec net
|
Mon, 23 Feb 2009 03:35:07 +0000 |
zuxy |
Get rid of the outdated and unmaintained CPU codename table.
|
Tue, 17 Feb 2009 11:16:19 +0000 |
diego |
Convert HAVE_MALLOC_H into a 0/1 definition, fixes the warning:
|
Mon, 09 Feb 2009 00:29:26 +0000 |
diego |
Print information about detected CPU in verbose mode only.
|
Fri, 30 Jan 2009 23:29:39 +0000 |
diego |
Use OS preprocessor checks with '#if defined()' consistently.
|
Mon, 26 Jan 2009 14:36:31 +0000 |
diego |
Replace another bunch of '#if HAVE_FOO' preprocessor checks by 'if (HAVE_FOO)'.
|
Mon, 26 Jan 2009 09:29:53 +0000 |
diego |
HAVE_3DNOW --> HAVE_AMD3DNOW
|
Sun, 25 Jan 2009 21:03:02 +0000 |
diego |
HAVE_3DNOWEX --> HAVE_3DNOWEXT
|
Fri, 16 Jan 2009 12:23:52 +0000 |
gpoirier |
fix wrong #ifdef/#ifndef -> #if conversion in r28323
|
Fri, 16 Jan 2009 09:44:36 +0000 |
reimar |
Fix a wrongly converted !defined(ARCH_X86_64)
|
Fri, 16 Jan 2009 08:43:05 +0000 |
reimar |
Fix first handful of #if vs. #ifdef for ARCH_, HAVE_SSE etc.
|
Fri, 16 Jan 2009 01:11:16 +0000 |
diego |
one more ARCH_ARMV4L --> ARCH_ARM, patch by Guillaume Lecerf, foxcore gmail com
|
Sat, 27 Dec 2008 12:00:11 +0000 |
diego |
Set and use only ARCH_PPC, not also ARCH_POWERPC.
|
Sun, 30 Nov 2008 09:17:21 +0000 |
zuxy |
Correct detection of SSSE3 and SSE4a feature bits.
|
Tue, 25 Nov 2008 18:36:50 +0000 |
cehoyos |
Replace pushf/popf by explicit pushfl/popfl (32 bit) or pushfq/popfq
|
Wed, 19 Nov 2008 16:41:59 +0000 |
gpoirier |
Add detection of x86 CPU features SSSE3 and SSE4a.
|
Thu, 16 Oct 2008 20:17:56 +0000 |
diego |
Replace all occurrences of '__volatile__' and '__volatile' by plain 'volatile'.
|
Thu, 16 Oct 2008 18:59:27 +0000 |
diego |
Change all occurrences of asm and __asm to __asm__, same as was done for FFmpeg.
|
Mon, 13 Oct 2008 16:23:55 +0000 |
diego |
Replace preprocessor check for WIN32 with checks for __MINGW32__ and __CYGWIN__.
|
Thu, 18 Sep 2008 15:30:23 +0000 |
reimar |
Simplify cpudetect OS-support detection code, e.g. using one mp_msg to print either yes or no instead of two.
|
Fri, 29 Aug 2008 17:44:51 +0000 |
diego |
Remove pointless '#if 1 [...] #endif' around has_cpuid() function.
|
Fri, 16 May 2008 09:42:28 +0000 |
diego |
cosmetics: Remove useless parentheses from return statements.
|
Sun, 24 Feb 2008 18:21:41 +0000 |
diego |
Wrap HAVE_XXX macros with RUNTIME_CPUDETECT, because when RUNTIME_CPUDETECT is
|
Sun, 24 Feb 2008 18:09:28 +0000 |
diego |
Enable SSE detection on OS/2.
|
Sun, 24 Feb 2008 18:04:34 +0000 |
diego |
cosmetics: Break one unreadable long line.
|
Sun, 24 Feb 2008 12:05:20 +0000 |
diego |
Add #include <sys/sysctl.h> for Mac OS X, fixes the warning
|
Tue, 11 Dec 2007 22:37:36 +0000 |
diego |
support for xtensa CPU architecture
|
Tue, 11 Dec 2007 22:31:57 +0000 |
diego |
Slightly simplify preprocessor conditionals.
|
Tue, 11 Dec 2007 22:29:22 +0000 |
diego |
Ahem, fix breakage of last commit: The AltiVec detection code has three
|
Tue, 11 Dec 2007 20:30:30 +0000 |
diego |
Remove redundant and obfuscating preprocessor conditional.
|
Tue, 11 Dec 2007 20:27:13 +0000 |
diego |
Replace SYS_DARWIN conditional by the more correct __APPLE__.
|
Mon, 01 Oct 2007 22:48:39 +0000 |
diego |
Fix warning:
|
Sat, 29 Sep 2007 19:48:45 +0000 |
diego |
Enable SSE on MinGW, many builds out there seem to use it without ill effect.
|
Fri, 14 Sep 2007 14:19:22 +0000 |
reimar |
sigill_handler_sse is not needed and can not compile on 64 bit systems
|
Fri, 14 Sep 2007 13:49:28 +0000 |
reimar |
Do not check for X86_FXSR_MAGIC define, it is missing in newer
|
Fri, 14 Sep 2007 13:43:06 +0000 |
reimar |
Remove unused sigfpe handler
|
Fri, 14 Sep 2007 13:25:20 +0000 |
reimar |
Remove test for SSE exception support that has been commented out since ages.
|
Wed, 10 Jan 2007 14:56:23 +0000 |
diego |
GNU/kFreeBSD support, closes Bugzilla #704.
|
Wed, 01 Nov 2006 18:41:25 +0000 |
reimar |
ARCH_X86 simplifications
|
Thu, 05 Oct 2006 00:12:04 +0000 |
nplourde |
Detect sse/2 on intel mac, Valtteri Vuorikoski(vuori@sci.fi)
|
Sat, 01 Jul 2006 03:56:13 +0000 |
reynaldo |
rm unnecesary void* casts - part 1
|
Sun, 25 Jun 2006 17:37:58 +0000 |
gpoirier |
Typo: use ptmpstr instead of tmpstr to strip leading spaces + add
|
Wed, 17 May 2006 21:10:11 +0000 |
gpoirier |
Retrieve CPU built-in namestring, and if it exists, print it during cpu detection; t it doesn't exist, fallback to the cpu table. Patch by Zuxy Meng
|
Tue, 28 Feb 2006 23:24:38 +0000 |
pacman |
add Amiga-style AltiVec detection, patch from andrea at amigasoft dot net
|
Thu, 09 Feb 2006 14:08:03 +0000 |
rathann |
Patch by Stefan Huehner / stefan % huehner ! org \
|
Mon, 07 Nov 2005 16:16:14 +0000 |
reimar |
Do not show cache-line size message, I've never seen a case where it was useful
|
Tue, 04 Oct 2005 21:06:07 +0000 |
gpoirier |
according to Intel/AMD official documentations, CPU family should be displayed as
|
Wed, 25 May 2005 08:48:32 +0000 |
diego |
DragonFly BSD support
|
Wed, 12 Jan 2005 21:56:25 +0000 |
reimar |
ensure null-termination after snprintf
|
Mon, 10 Jan 2005 18:33:51 +0000 |
aurel |
assume OS support SSE on x86_64
|
Thu, 21 Oct 2004 11:55:20 +0000 |
aurel |
adapting existing mmx/mmx2/sse/3dnow optimizations so they work on x86_64
|
Wed, 13 Oct 2004 09:37:32 +0000 |
diego |
Remove CPU speed detection since
|
Fri, 18 Jun 2004 20:56:06 +0000 |
faust3 |
disable buggy sse on mingw
|
Tue, 06 Apr 2004 17:08:02 +0000 |
alex |
SSE support under OpenBSD, patch by Bjorn Sandell
|
Thu, 01 Apr 2004 10:19:57 +0000 |
faust3 |
reenable SSE support for mingw
|
Mon, 16 Feb 2004 23:44:05 +0000 |
gabucino |
architecture type reporting for non-x86 CPUs (try 2, tested on x86 and x86-64)
|