eeaf0bdadc
Fix non-intel arch builds (cpuid.h header)
2023-02-09 20:54:46 +01:00
Ayan Shafqat
bb31125c94
Fix CPUID to detect AVX512F
...
CPUID is part of x86_64 ISA to query CPU features. In order to determine
AVX512F ISA extension, EAX and ECX needs to be set to 7 and 0
respectively before invoking `cpuid` instruction. This commit also
removes inline assembly for __cpuid in favor of using compiler provided
intrinsic functions. Both GCC and clang provides __cpuid like function
via __cpuid_count intrinsic.
This commit also creates a portable wrapper over compiler intrinsic
functions, __cpuid and __cpuidex. `cpuid' provides base level ISA query
and `cpuidex` provides extra extension information like AVX512F. These
wrappers lean towards MSVC like API.
References:
CPUID Docs: https://en.wikipedia.org/wiki/CPUID
GCC's ``docs" on __cpuid_count:
https://gcc.gnu.org/git/?p=gcc.git;a=blob;f=gcc/config/i386/cpuid.h
Clang's docs on __cpuid_count:
https://clang.llvm.org/doxygen/cpuid_8h.html
MSVC's docs on __cpuid and __cpuidex:
https://learn.microsoft.com/en-us/cpp/intrinsics/cpuid-cpuidex
2023-02-08 23:20:20 +01:00
Ayan Shafqat
bf8fced073
Add CPU detection
2023-02-04 23:07:17 +01:00
John Emmas
ef7a5e1b00
For 64-bit compilation, MSVC no longer supports inline assembly
2021-01-07 09:22:43 +00:00
Ayan Shafqat
407882d23d
Add support for Intel/AMD's FMA extension
...
By supporting FMA extension, the number of instruction needed
for multiply accumulate to mix channels are reduced. Since,
this extension has been around since middle of 2012, more
computers have this instruction set available.
2021-01-01 21:05:16 +01:00
34159e4594
Fix ARM runtime NEON detection
2020-08-24 22:03:28 +02:00
a0a4477c84
Fix aarch64 builds
2020-08-24 21:26:38 +02:00
Ayan Shafqat
1f878636c8
Adding NEON detection during runtime
2020-08-24 02:06:35 +02:00
0301c47f6b
Update core library GPL boilerplate and (C) from git log
2019-08-03 15:53:17 +02:00
2f91bdfa53
NO-OP: <tab> after <space> fixes in libs
2019-04-13 19:19:29 +02:00
ed09a68f24
Fix for conflicting definition of _xgetbv with mingw-w64 >= 5
...
Simplify the #ifdef logic so it is easier to follow and add exceptions in the
future if necessary.
2016-11-10 15:23:43 +10:00
cf52d6e4b4
enough with umpteen "i18n.h" files. Consolidate on pbd/i18n.h
2016-07-14 14:45:23 -04:00
Florian Weimer
5a41b86028
fix __cpuid() on x86
...
The previous version used memory operands that gcc (probably dependent
on optimization flags and/or version) could address relative to the
stack pointer, but pushing %ebx onto the stack changed it. Here, the
address of the regs array is put into %esi and the individual members
are written into directly.
2016-03-08 14:54:41 +01:00
3dc09d42ab
allow to override FPU detection
2016-03-04 18:05:47 +01:00
22b07e0233
NOOP, remove trailing tabs/whitespace.
2015-10-05 16:17:49 +02:00
4dc63966f0
globally remove all trailing whitespace from ardour code base.
...
Paul Davis was responsible for introducing almost all of this.
2015-10-04 14:51:05 -04:00
eb998b3264
plug some memory leaks in libs
2015-09-12 20:05:25 +02:00
1d2938b162
Add '_xgetbv()' for MSVC-9 and earlier
2015-08-16 13:14:54 +01:00
f77a409b0c
tweaks for AVX detection
2015-08-14 08:53:06 -04:00
5a304e9e33
fix PPC builds
2015-08-12 19:16:01 +02:00
c86b82aa8a
comment tweak to show that the __cpuid() intrinsic belongs to MSVC and mingw
2015-08-11 23:45:52 -04:00
94b6e7ffe4
fix mingw compilation
2015-08-12 05:25:29 +02:00
b6f9bbeb2f
can't use xgetbv instruction with apple gcc under OS X Lion
2015-08-11 23:06:47 -04:00
41ccfee7a4
clean up FPU code with some ideas from Chromium and the web
2015-08-11 23:00:13 -04:00
bb399c8787
fix stupid logic error in testing separate bits required for AVX support
2015-08-11 20:18:38 -04:00
bb5c969ac0
improved solution for xgetbv() on windows and linux with all compilers
2015-08-11 17:45:41 -04:00
53bc2ba2c1
#ifdef out AVX detection on APPLE
2015-08-11 17:30:33 -04:00
c0ad97c179
add required test of OS support for AVX register use
2015-08-11 17:23:11 -04:00
4079855507
add test for CPU/FPU AVX capabilities
2015-08-11 16:36:46 -04:00
a3209b9b91
Correctly dereference 'fxbuf' when building with MSVC
2015-04-30 18:17:24 +01:00
fe3aafec1c
When testing for x86 / x64 architecture, don't forget the symbols that MSVC can understand
2015-04-30 18:17:23 +01:00
6cb742c289
windows 32bit asm/fpu support.
...
This still leaves cache_aligned_malloc() in
libs/pbd/malign.cc pending _aligned_free (windows crashes when using
free() on memory allocated with _aligned_malloc()).
So far however there seems to be no issue with default malloc
for audio+midi buffers on win32…
2015-04-24 02:43:52 +02:00
e015c57663
<intrin.h> is for windows only
2015-04-21 12:31:02 -04:00
e3156c2cc4
add required header for windows compile of FPU code
2015-04-21 12:25:21 -04:00
9241f58188
fix build of FPU code on OS X by reverting to use of _LP64 to identify 32/64 bit situation
2015-04-21 12:13:00 -04:00
a5d7e8446b
merge all mingw and msvc specific code for FPU information into libs/pbd/fpu.cc and remove msvc-specific version
2015-04-21 11:39:31 -04:00
Greg Zharun
8af992c449
[Summary] Added SSE sound processing functions support for Windows. Version 1.
...
Conflicts:
wscript
2015-04-21 10:49:00 -04:00
658bb3ccd4
finished merge of cairocanvas with windows and windows+cc branches
2014-01-10 17:11:10 -05:00
3020b224fa
Merge windows+cc branch into cairocanvas branch. Not finished, need to now merge windows branch to get changes from there
2014-01-10 16:07:57 -05:00
b7835cd7b1
workaround compilation issues with assembler in fpu code and mingw
2013-12-05 14:45:17 -05:00
0cca272431
try to fix various warnings from gcc when optimization flags are enabled
2013-11-04 21:32:41 -05:00
7b480eaa27
Windows (compiler specific) includes for libpbd
2013-07-16 18:00:49 +01:00
58a027b7a2
add copyright comments
...
git-svn-id: svn://localhost/ardour2/branches/3.0@13857 d708f5d6-7413-0410-9779-e7cbd77b26cf
2013-01-16 18:15:38 +00:00
8647471379
Fix posix_memalign checks.
...
git-svn-id: svn://localhost/ardour2/branches/3.0@12316 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-05-17 00:54:43 +00:00
Carl Hetherington
312c4e22a7
Align the address of the pointer to the fxsave block to a 16-byte boundary (as well as the pointer itself), which the internets seem to suggest is required.
...
git-svn-id: svn://localhost/ardour2/branches/3.0@12313 d708f5d6-7413-0410-9779-e7cbd77b26cf
2012-05-16 22:43:23 +00:00
a556546291
explain MXCSR shenanigans in libs/pbd/fpu.cc
...
git-svn-id: svn://localhost/ardour2/branches/3.0@11079 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-12-26 22:32:21 +00:00
639bb7dc5d
use compiler-generated _LP64 macro to detect x86_64 rather than USE_X86_64_ASM
...
git-svn-id: svn://localhost/ardour2/branches/3.0@10871 d708f5d6-7413-0410-9779-e7cbd77b26cf
2011-12-02 20:38:31 +00:00
639542b410
build fixes for OS X
...
git-svn-id: svn://localhost/ardour2/branches/3.0@8006 d708f5d6-7413-0410-9779-e7cbd77b26cf
2010-11-10 23:13:29 +00:00
c6d5124269
update comments on register clobbering
...
git-svn-id: svn://localhost/ardour2/branches/3.0@6423 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-31 18:19:46 +00:00
569226e64e
un-fix register clobber list for gcc on x86
...
git-svn-id: svn://localhost/ardour2/branches/3.0@6422 d708f5d6-7413-0410-9779-e7cbd77b26cf
2009-12-31 18:16:03 +00:00
