search for: xmmintrin

...ncoding 8000 Hz audio using narrowband mode (stereo)<br> Segmentation fault (core dumped)<br> <br> Stack&nbsp; trace:<br> Loaded symbols for /lib/ld-linux.so.2<br> #0&nbsp; fir_mem2_10 (x=0x80552a8, _num=0x80566b4, y=0x80558ac, N=160, ord=10, _mem=0x80568cc) at xmmintrin.h:790<br> 790&nbsp;&nbsp;&nbsp;&nbsp; {<br> (gdb) bt<br> #0&nbsp; fir_mem2_10 (x=0x80552a8, _num=0x80566b4, y=0x80558ac, N=160, ord=10, _mem=0x80568cc) at xmmintrin.h:790<br> #1&nbsp; 0xb75dc509 in fir_mem2 (x=0x80552a8, _num=0x80566b4, y=0x80558ac, N...

....html , specifically under "New Targets and Target Specific Improvements") to enable SSE instructions within the compiler's output (for appropriate architectures). Compiling speex on GCC 3.3.2 without the -msse option yields the error message shown at bottom. The referenced file (xmmintrin.h) contains a check to ensure that SSE is enabled (via -msse) before use: #ifndef __SSE__ # error "SSE instruction set not enabled" #else This check seems to have been added to the GCC 3.3.x series. Once the option is added to CFLAGS, libspeex (and the rest of the speex package) comp...

...# speexenc -n loup.wav loup.spx > Encoding 8000 Hz audio using narrowband mode (stereo) > Segmentation fault (core dumped) > > Stack trace: > Loaded symbols for /lib/ld-linux.so.2 > #0 fir_mem2_10 (x=0x80552a8, _num=0x80566b4, y=0x80558ac, N=160, > ord=10, _mem=0x80568cc) at xmmintrin.h:790 > 790 { > (gdb) bt > #0 fir_mem2_10 (x=0x80552a8, _num=0x80566b4, y=0x80558ac, N=160, > ord=10, _mem=0x80568cc) at xmmintrin.h:790 > #1 0xb75dc509 in fir_mem2 (x=0x80552a8, _num=0x80566b4, y=0x80558ac, > N=160, ord=0, _mem=0x1) at filters_sse.h:330 > #2 0xb75d2ca7...

...-speexenc --rate 8000 --16bit -n /dev/zero /tmp/foo Encoding 8000 Hz audio using narrowband mode (mono) Program received signal SIGSEGV, Segmentation fault. fir_mem2_10 (x=0x8054a88, _num=0x8055e94, y=0x805508c, N=160, ord=10, _mem=0x80560ac) at /usr/lib/gcc-lib/i386-redhat-linux/3.2/include/xmmintrin.h:774 774 { (gdb) (gdb) bt #0 fir_mem2_10 (x=0x8054a88, _num=0x8055e94, y=0x805508c, N=160, ord=10, _mem=0x80560ac) at /usr/lib/gcc-lib/i386-redhat-linux/3.2/include/xmmintrin.h:774 #1 0x40020add in fir_mem2 (x=0x8054a88, _num=0x8055e94, y=0x805508c, N=160, ord=0, _mem=0x1) at filters_s...

...t; #define float2int(x) lrint(x) -#elif (defined(_MSC_VER) && _MSC_VER >= 1400) && (defined (WIN64) || defined (_WIN64)) +#elif (defined(_MSC_VER) && _MSC_VER >= 1400) && (defined (WIN64) || defined (_WIN64)) && !defined(_M_ARM) #include <xmmintrin.h> __inline long int float2int(float value) { return _mm_cvtss_si32(_mm_load_ss(&value)); } -#elif (defined(_MSC_VER) && _MSC_VER >= 1400) && (defined (WIN32) || defined (_WIN32)) +#elif (defined(_MSC_VER) && _MSC_VER &...

...-#elif (defined(_MSC_VER) && _MSC_VER >= 1400) && (defined (WIN64) || >> defined (_WIN64)) >> +#elif (defined(_MSC_VER) && _MSC_VER >= 1400) && (defined (WIN64) || >> defined (_WIN64)) && !defined(_M_ARM) >> #include <xmmintrin.h> >> >> __inline long int float2int(float value) >> { >> return _mm_cvtss_si32(_mm_load_ss(&value)); >> } >> -#elif (defined(_MSC_VER) && _MSC_VER >= 1400) && (defined (WIN32) || >> defined...

...program, ie did you use clang, llvm-gcc, or dragonegg, and which options did you use. From your message, I can't tell if you built at O0 or O3. In this case, no, LLVM does not have any auto-vectorization optimizations. However, LLVM does have good support for vector intrinsics, so if you use xmmintrin.h you should be able to get good performance. Reid

...64-bit compiler: --- float_cast.h Sat Oct 17 19:51:32 2009 +++ new\float_cast.h Fri Oct 15 12:09:38 2010 @@ -72,8 +72,14 @@ #include <math.h> #define float2int(x) lrint(x) -#elif (defined (WIN32) || defined (_WIN32)) +#elif (defined (WIN64) || defined (_WIN64)) + #include <xmmintrin.h> + __inline long int float2int(float value) + { + return _mm_cvtss_si32(_mm_load_ss(&value)); + } +#elif (defined (WIN32) || defined (_WIN32)) #include <math.h> /* Win32 doesn't seem to have these functions. BTW, I have also compiled CELT for 64...

...asm. Regards. -- O.S. diff --git a/celt/float_cast.h b/celt/float_cast.h index ed5a39b..b9b8484 100644 --- a/celt/float_cast.h +++ b/celt/float_cast.h @@ -61,7 +61,14 @@ ** the config.h file. */ -#if (HAVE_LRINTF) +#if (defined(__GNUC__) && defined(_WIN64)) + #include <xmmintrin.h> + static __inline long int float2int(float value) + { + return _mm_cvtss_si32(_mm_load_ss(&value)); + } + +#elif (HAVE_LRINTF) /* These defines enable functionality introduced with the 1999 ISO C ** standard. They must be defined before the...

...r "New > Targets and Target Specific Improvements") to enable SSE instructions > within the compiler's output (for appropriate architectures). > > Compiling speex on GCC 3.3.2 without the -msse option yields the error > message shown at bottom. The referenced file (xmmintrin.h) contains a > check to ensure that SSE is enabled (via -msse) before use: > > #ifndef __SSE__ > # error "SSE instruction set not enabled" > #else > > This check seems to have been added to the GCC 3.3.x series. > > Once the option is added to CFLAGS, libsp...

...sy. You have a couple options here: one is to manipulate the source to let you see the _mm_ names, and the other is to catch the _mm_ names before the inliner runs. Manipulating the source isn't actually very hard, although it's a non-trivial amount of work; basically, you create your own xmmintrin.h that doesn't have inline implementations, and mess with the include paths so the compiler picks your version rather than the builtin version. That way, once you transform to IL, the _mm_ calls will stay as _mm_ calls. If you're using the standard headers, the _mm_ function are defined a...

---------- Forwarded message ---------- From: Serg Anohovsky <serg.anohovsky at gmail.com> Date: 2011/5/22 Subject: Re: [LLVMdev] No SSE instructions To: Chris Lattner <clattner at apple.com> 2011/5/22 Chris Lattner <clattner at apple.com> > > On May 22, 2011, at 10:47 AM, Justin Holewinski wrote: > > On Sun, May 22, 2011 at 1:07 PM, Serg Anohovsky

...-DVERSION=\"1.1.4\" -D_USE_SSE=1 -DEPIC_48K=1 -I. -I. -s -O3 -march=i686 -MT nb_celp.lo -MD -MP -MF .deps/nb_celp.Tpo -c nb_celp.c -fPIC -DPIC -o .libs/nb_celp.o In file included from vq.h:43, from nb_celp.c:41: /usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.3.2/include/xmmintrin.h:34:3: #error "SSE instruction set not enabled" In file included from nb_celp.c:41: vq.h:44: error: parse error before '*' token vq.h:46: error: parse error before '*' token gmake[1]: *** [nb_celp.lo] Error 1 gmake[1]: Leaving directory `/tmp/speex-1.1.4/libspeex' gm...

...AY_HAVE_NEON_INTR" = x"1" && test x"$OPUS_ARM_PRESUME_NEON_INTR" != x"1"], @@ -521,10 +522,13 @@ AS_IF([test x"$enable_intrinsics" = x"yes"],[ [OPUS_X86_MAY_HAVE_SSE], [OPUS_X86_PRESUME_SSE], [[#include <xmmintrin.h> + #include <time.h> ]], [[ - static __m128 mtest; - mtest = _mm_setzero_ps(); + __m128 mtest; + mtest = _mm_set1_ps((float)time(NULL)); + mtest = _mm_mul_ps(mtest, mtest); + return _mm_...

...; _MSC_VER >= 1400) && (defined (WIN64) || > >>> defined (_WIN64)) > >>> +#elif (defined(_MSC_VER) && _MSC_VER >= 1400) && (defined (WIN64) || > >>> defined (_WIN64)) && !defined(_M_ARM) > >>> #include <xmmintrin.h> > >>> > >>> __inline long int float2int(float value) > >>> { > >>> return _mm_cvtss_si32(_mm_load_ss(&value)); > >>> } > >>> -#elif (defined(_MSC_VER) && _MSC_VER >=...

Hii, Thanks for the response, yes I couldn't find any way to extract the names through any of the passes. Where could I potentially insert a hack so that any function call to intrinsic functions or library functions can be retrieved? Could you gimme any ideas for the start? -Nipun On Fri, Jan 30, 2009 at 10:39 PM, Eli Friedman <eli.friedman at gmail.com>wrote: > On Fri, Jan 30,

...eex 1.1.5 release, and the current CVS (which self-IDs as 1.1.4), and the result is the same. I suspect some funk in the use of the SSE intrinsics macros. Backtrace: #0 0x40024594 in filter_mem2_10 (x=0x805f31c, _num=0x8061fb8, _den=0x8061fe4, y=0x806071c, N=160, ord=10, _mem=0x8062150) at xmmintrin.h:790 #1 0x400248b4 in filter_mem2 (x=0x805f31c, _num=0x8061fb8, _den=0x8061fe4, y=0x806071c, N=1, ord=0, _mem=0x8061fe4) at filters_sse.h:135 #2 0x40019d1e in nb_encode (state=0x805ebd0, vin=0x80582b4, bits=0xbfffe840) at nb_celp.c:314 #3 0x40025e33 in speex_encode_native (state=0x8061fe4,...

...alpha.cpp # LLVM LOCAL end ;; ... arm*-*-*) cpu_type=arm extra_headers="mmintrin.h" ;; ... i[34567]86-*-*) cpu_type=i386 # LLVM LOCAL begin out_cxx_file=i386/llvm-i386.cpp # LLVM LOCAL end # APPLE LOCAL begin 5612787 mainline sse4 extra_headers="mmintrin.h mm3dnow.h xmmintrin.h emmintrin.h pmmintrin.h tmmintrin.h ammintrin.h smmintrin.h nmmintrin.h" (out_cxx_file variable is empty for ARM target) I wonder if llvm-gcc 4.2 front-end support bitcode conversion for ARM target. Thank you. Best regards, Jin-Gu Kang _____________________________...

...xtern void (*const CELT_FIR_IMPL[OPUS_ARCHMASK + 1])( #endif #endif + +#endif diff --git a/celt/x86/pitch_sse.c b/celt/x86/pitch_sse.c index e3bc6d7..20e7312 100644 --- a/celt/x86/pitch_sse.c +++ b/celt/x86/pitch_sse.c @@ -29,223 +29,157 @@ #include "config.h" #endif -#include <xmmintrin.h> -#include <emmintrin.h> - #include "macros.h" #include "celt_lpc.h" #include "stack_alloc.h" #include "mathops.h" #include "pitch.h" -#if defined(OPUS_X86_MAY_HAVE_SSE4_1) -#include <smmintrin.h> -#include "x86cpu.h&quo...

...xtern void (*const CELT_FIR_IMPL[OPUS_ARCHMASK + 1])( #endif #endif + +#endif diff --git a/celt/x86/pitch_sse.c b/celt/x86/pitch_sse.c index e3bc6d7..20e7312 100644 --- a/celt/x86/pitch_sse.c +++ b/celt/x86/pitch_sse.c @@ -29,223 +29,157 @@ #include "config.h" #endif -#include <xmmintrin.h> -#include <emmintrin.h> - #include "macros.h" #include "celt_lpc.h" #include "stack_alloc.h" #include "mathops.h" #include "pitch.h" -#if defined(OPUS_X86_MAY_HAVE_SSE4_1) -#include <smmintrin.h> -#include "x86cpu.h&quo...