Browse Source

Refs #113. Fixed the typo BOBCATE -> BOBCAT

tags/v0.2.0^2
Zhang Xianyi 13 years ago
parent
commit
d3b67d0bd8
28 changed files with 83 additions and 83 deletions
  1. +1
    -1
      TargetList.txt
  2. +2
    -2
      cpuid.h
  3. +6
    -6
      cpuid_x86.c
  4. +2
    -2
      driver/others/parameter.c
  5. +6
    -6
      getarch.c
  6. +1
    -1
      kernel/setparam-ref.c
  7. +0
    -0
      kernel/x86/KERNEL.BOBCAT
  8. +5
    -5
      kernel/x86/trsm_kernel_LN_2x4_sse2.S
  9. +11
    -11
      kernel/x86/trsm_kernel_LN_4x4_sse.S
  10. +5
    -5
      kernel/x86/trsm_kernel_LT_2x4_sse2.S
  11. +11
    -11
      kernel/x86/trsm_kernel_LT_4x4_sse.S
  12. +5
    -5
      kernel/x86/trsm_kernel_RT_2x4_sse2.S
  13. +11
    -11
      kernel/x86/trsm_kernel_RT_4x4_sse.S
  14. +2
    -2
      kernel/x86/ztrsm_kernel_LN_2x2_sse.S
  15. +2
    -2
      kernel/x86/ztrsm_kernel_LT_2x2_sse.S
  16. +2
    -2
      kernel/x86/ztrsm_kernel_RT_2x2_sse.S
  17. +0
    -0
      kernel/x86_64/KERNEL.BOBCAT
  18. +1
    -1
      kernel/x86_64/symv_L_sse.S
  19. +1
    -1
      kernel/x86_64/symv_L_sse2.S
  20. +1
    -1
      kernel/x86_64/symv_U_sse.S
  21. +1
    -1
      kernel/x86_64/symv_U_sse2.S
  22. +1
    -1
      kernel/x86_64/zsymv_L_sse.S
  23. +1
    -1
      kernel/x86_64/zsymv_L_sse2.S
  24. +1
    -1
      kernel/x86_64/zsymv_U_sse.S
  25. +1
    -1
      kernel/x86_64/zsymv_U_sse2.S
  26. +1
    -1
      l1param.h
  27. +1
    -1
      l2param.h
  28. +1
    -1
      param.h

+ 1
- 1
TargetList.txt View File

@@ -28,7 +28,7 @@ OPTERON_SSE3
BARCELONA
SHANGHAI
ISTANBUL
BOBCATE
BOBCAT

c)VIA CPU:
SSE_GENERIC


+ 2
- 2
cpuid.h View File

@@ -104,7 +104,7 @@
#define CORE_ATOM 18
#define CORE_NANO 19
#define CORE_SANDYBRIDGE 20
#define CORE_BOBCATE 21
#define CORE_BOBCAT 21

#define HAVE_SSE (1 << 0)
#define HAVE_SSE2 (1 << 1)
@@ -192,5 +192,5 @@ typedef struct {
#define CPUTYPE_VIAC3 42
#define CPUTYPE_NANO 43
#define CPUTYPE_SANDYBRIDGE 44
#define CPUTYPE_BOBCATE 45
#define CPUTYPE_BOBCAT 45
#endif

+ 6
- 6
cpuid_x86.c View File

@@ -1029,7 +1029,7 @@ int get_cpuname(void){
case 10:
return CPUTYPE_BARCELONA;
case 5:
return CPUTYPE_BOBCATE;
return CPUTYPE_BOBCAT;
}
break;
}
@@ -1150,7 +1150,7 @@ static char *cpuname[] = {
"VIAC3",
"NANO",
"SANDYBRIDGE",
"BOBCATE",
"BOBCAT",
};

static char *lowercpuname[] = {
@@ -1198,7 +1198,7 @@ static char *lowercpuname[] = {
"nsgeode",
"nano",
"sandybridge",
"bobcate",
"bobcat",
};

static char *corename[] = {
@@ -1223,7 +1223,7 @@ static char *corename[] = {
"ATOM",
"NANO",
"SANDYBRIDGE",
"BOBCATE",
"BOBCAT",
};

static char *corename_lower[] = {
@@ -1248,7 +1248,7 @@ static char *corename_lower[] = {
"atom",
"nano",
"sandybridge",
"bobcate",
"bobcat",
};


@@ -1358,7 +1358,7 @@ int get_coretype(void){
if (family <= 0xe) return CORE_ATHLON;
if (family == 0xf){
if ((exfamily == 0) || (exfamily == 2)) return CORE_OPTERON;
else if (exfamily == 5) return CORE_BOBCATE;
else if (exfamily == 5) return CORE_BOBCAT;
else return CORE_BARCELONA;
}
}


+ 2
- 2
driver/others/parameter.c View File

@@ -163,7 +163,7 @@ int get_L2_size(void){

int eax, ebx, ecx, edx;

#if defined(ATHLON) || defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE) || \
#if defined(ATHLON) || defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT) || \
defined(CORE_PRESCOTT) || defined(CORE_CORE2) || defined(PENRYN) || defined(DUNNINGTON) || \
defined(CORE_NEHALEM) || defined(CORE_SANDYBRIDGE) || defined(ATOM) || defined(GENERIC)

@@ -446,7 +446,7 @@ void blas_set_parameter(void){
#endif
#endif

#if defined(CORE_BARCELONA) || defined(CORE_BOBCATE)
#if defined(CORE_BARCELONA) || defined(CORE_BOBCAT)
size >>= 8;

sgemm_p = 232 * size;


+ 6
- 6
getarch.c View File

@@ -102,7 +102,7 @@ USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
/* #define FORCE_BARCELONA */
/* #define FORCE_SHANGHAI */
/* #define FORCE_ISTANBUL */
/* #define FORCE_BOBCATE */
/* #define FORCE_BOBCAT */
/* #define FORCE_SSE_GENERIC */
/* #define FORCE_VIAC3 */
/* #define FORCE_NANO */
@@ -364,19 +364,19 @@ USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#define CORENAME "BARCELONA"
#endif

#if defined(FORCE_BOBCATE)
#if defined(FORCE_BOBCAT)
#define FORCE
#define FORCE_INTEL
#define ARCHITECTURE "X86"
#define SUBARCHITECTURE "BOBCATE"
#define ARCHCONFIG "-DBOBCATE " \
#define SUBARCHITECTURE "BOBCAT"
#define ARCHCONFIG "-DBOBCAT " \
"-DL1_DATA_SIZE=32768 -DL1_DATA_LINESIZE=64 " \
"-DL2_SIZE=524288 -DL2_LINESIZE=64 " \
"-DDTB_DEFAULT_ENTRIES=40 -DDTB_SIZE=4096 " \
"-DHAVE_MMX -DHAVE_SSE -DHAVE_SSE2 -DHAVE_SSE3 -DHAVE_SSSE3 " \
"-DHAVE_SSE4A -DHAVE_MISALIGNSSE -DHAVE_CFLUSH -DHAVE_CMOV"
#define LIBNAME "bobcate"
#define CORENAME "BOBCATE"
#define LIBNAME "bobcat"
#define CORENAME "BOBCAT"
#endif

#ifdef FORCE_SSE_GENERIC


+ 1
- 1
kernel/setparam-ref.c View File

@@ -794,7 +794,7 @@ static void init_parameter(void) {
#endif
#endif

#ifdef BOBCATE
#ifdef BOBCAT

#ifdef DEBUG
fprintf(stderr, "Bobcate\n");


kernel/x86/KERNEL.BOBCATE → kernel/x86/KERNEL.BOBCAT View File


+ 5
- 5
kernel/x86/trsm_kernel_LN_2x4_sse2.S View File

@@ -69,7 +69,7 @@
#define STACK_ALIGN 4096
#define STACK_OFFSET 1024

#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
#define PREFETCH prefetch
#define PREFETCHSIZE (8 * 10 + 4)
#endif
@@ -439,7 +439,7 @@
.L22:
mulsd %xmm0, %xmm2
addsd %xmm2, %xmm4
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
PREFETCH (PREFETCHSIZE + 0) * SIZE(AA)
#endif
movlpd 2 * SIZE(BB), %xmm2
@@ -488,7 +488,7 @@
movlpd 40 * SIZE(BB), %xmm3
addsd %xmm0, %xmm7
movlpd 8 * SIZE(AA), %xmm0
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
PREFETCH (PREFETCHSIZE + 8) * SIZE(AA)
#endif
mulsd %xmm1, %xmm2
@@ -1697,7 +1697,7 @@

.L42:
mulpd %xmm0, %xmm2
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif
mulpd 2 * SIZE(BB), %xmm0
@@ -1727,7 +1727,7 @@
addpd %xmm0, %xmm7
movapd 16 * SIZE(AA), %xmm0

#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 8) * SIZE(AA)
#endif
mulpd %xmm1, %xmm2


+ 11
- 11
kernel/x86/trsm_kernel_LN_4x4_sse.S View File

@@ -64,7 +64,7 @@
#define BORIG 60(%esp)
#define BUFFER 128(%esp)

#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
#define PREFETCH prefetch
#define PREFETCHW prefetchw
#define PREFETCHSIZE (16 * 10 + 8)
@@ -437,7 +437,7 @@
.L32:
mulss %xmm0, %xmm2
addss %xmm2, %xmm4
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif
movss 4 * SIZE(BB), %xmm2
@@ -833,7 +833,7 @@
.L22:
mulps %xmm0, %xmm2
addps %xmm2, %xmm4
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif
movaps 4 * SIZE(BB), %xmm2
@@ -1848,7 +1848,7 @@

.L72:
mulss %xmm0, %xmm2
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif
mulss 4 * SIZE(BB), %xmm0
@@ -2109,7 +2109,7 @@
ALIGN_4

.L62:
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif

@@ -2429,7 +2429,7 @@

.L52:
mulps %xmm0, %xmm2
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif
mulps 4 * SIZE(BB), %xmm0
@@ -2459,7 +2459,7 @@
addps %xmm0, %xmm5
movaps 32 * SIZE(AA), %xmm0

#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 16) * SIZE(AA)
#endif
mulps %xmm1, %xmm2
@@ -2952,7 +2952,7 @@

.L112:
mulss %xmm0, %xmm2
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif
movss 1 * SIZE(AA), %xmm0
@@ -3148,7 +3148,7 @@

.L102:
mulps %xmm0, %xmm2
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif
movsd 2 * SIZE(AA), %xmm0
@@ -3389,7 +3389,7 @@

.L92:
mulps %xmm0, %xmm2
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif
movaps 4 * SIZE(AA), %xmm0
@@ -3404,7 +3404,7 @@
mulps 12 * SIZE(BB), %xmm0
addps %xmm0, %xmm7
movaps 32 * SIZE(AA), %xmm0
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 16) * SIZE(AA)
#endif
mulps %xmm1, %xmm3


+ 5
- 5
kernel/x86/trsm_kernel_LT_2x4_sse2.S View File

@@ -69,7 +69,7 @@
#define STACK_ALIGN 4096
#define STACK_OFFSET 1024

#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
#define PREFETCH prefetch
#define PREFETCHSIZE (8 * 10 + 4)
#endif
@@ -910,7 +910,7 @@
.L22:
mulsd %xmm0, %xmm2
addsd %xmm2, %xmm4
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
PREFETCH (PREFETCHSIZE + 0) * SIZE(AA)
#endif
movlpd 2 * SIZE(BB), %xmm2
@@ -959,7 +959,7 @@
movlpd 40 * SIZE(BB), %xmm3
addsd %xmm0, %xmm7
movlpd 8 * SIZE(AA), %xmm0
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
PREFETCH (PREFETCHSIZE + 8) * SIZE(AA)
#endif
mulsd %xmm1, %xmm2
@@ -1439,7 +1439,7 @@

.L42:
mulpd %xmm0, %xmm2
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif
mulpd 2 * SIZE(BB), %xmm0
@@ -1469,7 +1469,7 @@
addpd %xmm0, %xmm7
movapd 16 * SIZE(AA), %xmm0

#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 8) * SIZE(AA)
#endif
mulpd %xmm1, %xmm2


+ 11
- 11
kernel/x86/trsm_kernel_LT_4x4_sse.S View File

@@ -64,7 +64,7 @@
#define BORIG 60(%esp)
#define BUFFER 128(%esp)

#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
#define PREFETCH prefetch
#define PREFETCHW prefetchw
#define PREFETCHSIZE (16 * 10 + 8)
@@ -872,7 +872,7 @@
.L22:
mulps %xmm0, %xmm2
addps %xmm2, %xmm4
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif
movaps 4 * SIZE(BB), %xmm2
@@ -1316,7 +1316,7 @@
.L32:
mulss %xmm0, %xmm2
addss %xmm2, %xmm4
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif
movss 4 * SIZE(BB), %xmm2
@@ -1855,7 +1855,7 @@

.L52:
mulps %xmm0, %xmm2
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif
mulps 4 * SIZE(BB), %xmm0
@@ -1885,7 +1885,7 @@
addps %xmm0, %xmm5
movaps 32 * SIZE(AA), %xmm0

#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 16) * SIZE(AA)
#endif
mulps %xmm1, %xmm2
@@ -2249,7 +2249,7 @@
ALIGN_4

.L62:
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif

@@ -2562,7 +2562,7 @@

.L72:
mulss %xmm0, %xmm2
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif
mulss 4 * SIZE(BB), %xmm0
@@ -2957,7 +2957,7 @@

.L92:
mulps %xmm0, %xmm2
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif
movaps 4 * SIZE(AA), %xmm0
@@ -2972,7 +2972,7 @@
mulps 12 * SIZE(BB), %xmm0
addps %xmm0, %xmm7
movaps 32 * SIZE(AA), %xmm0
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 16) * SIZE(AA)
#endif
mulps %xmm1, %xmm3
@@ -3280,7 +3280,7 @@

.L102:
mulps %xmm0, %xmm2
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif
movsd 2 * SIZE(AA), %xmm0
@@ -3515,7 +3515,7 @@

.L112:
mulss %xmm0, %xmm2
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif
movss 1 * SIZE(AA), %xmm0


+ 5
- 5
kernel/x86/trsm_kernel_RT_2x4_sse2.S View File

@@ -69,7 +69,7 @@
#define STACK_ALIGN 4096
#define STACK_OFFSET 1024

#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
#define PREFETCH prefetch
#define PREFETCHSIZE (8 * 10 + 4)
#endif
@@ -1036,7 +1036,7 @@

.L42:
mulpd %xmm0, %xmm2
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif
mulpd 2 * SIZE(BB), %xmm0
@@ -1066,7 +1066,7 @@
addpd %xmm0, %xmm7
movapd 16 * SIZE(AA), %xmm0

#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 8) * SIZE(AA)
#endif
mulpd %xmm1, %xmm2
@@ -2224,7 +2224,7 @@
.L22:
mulsd %xmm0, %xmm2
addsd %xmm2, %xmm4
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
PREFETCH (PREFETCHSIZE + 0) * SIZE(AA)
#endif
movlpd 2 * SIZE(BB), %xmm2
@@ -2273,7 +2273,7 @@
movlpd 40 * SIZE(BB), %xmm3
addsd %xmm0, %xmm7
movlpd 8 * SIZE(AA), %xmm0
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
PREFETCH (PREFETCHSIZE + 8) * SIZE(AA)
#endif
mulsd %xmm1, %xmm2


+ 11
- 11
kernel/x86/trsm_kernel_RT_4x4_sse.S View File

@@ -64,7 +64,7 @@
#define BORIG 60(%esp)
#define BUFFER 128(%esp)

#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
#define PREFETCH prefetch
#define PREFETCHW prefetchw
#define PREFETCHSIZE (16 * 10 + 8)
@@ -439,7 +439,7 @@

.L92:
mulps %xmm0, %xmm2
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif
movaps 4 * SIZE(AA), %xmm0
@@ -454,7 +454,7 @@
mulps 12 * SIZE(BB), %xmm0
addps %xmm0, %xmm7
movaps 32 * SIZE(AA), %xmm0
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 16) * SIZE(AA)
#endif
mulps %xmm1, %xmm3
@@ -758,7 +758,7 @@

.L102:
mulps %xmm0, %xmm2
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif
movsd 2 * SIZE(AA), %xmm0
@@ -993,7 +993,7 @@

.L112:
mulss %xmm0, %xmm2
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif
movss 1 * SIZE(AA), %xmm0
@@ -1324,7 +1324,7 @@

.L52:
mulps %xmm0, %xmm2
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif
mulps 4 * SIZE(BB), %xmm0
@@ -1354,7 +1354,7 @@
addps %xmm0, %xmm5
movaps 32 * SIZE(AA), %xmm0

#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 16) * SIZE(AA)
#endif
mulps %xmm1, %xmm2
@@ -1718,7 +1718,7 @@
ALIGN_4

.L62:
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif

@@ -2031,7 +2031,7 @@

.L72:
mulss %xmm0, %xmm2
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif
mulss 4 * SIZE(BB), %xmm0
@@ -2859,7 +2859,7 @@
.L22:
mulps %xmm0, %xmm2
addps %xmm2, %xmm4
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif
movaps 4 * SIZE(BB), %xmm2
@@ -3303,7 +3303,7 @@
.L32:
mulss %xmm0, %xmm2
addss %xmm2, %xmm4
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht0 (PREFETCHSIZE + 0) * SIZE(AA)
#endif
movss 4 * SIZE(BB), %xmm2


+ 2
- 2
kernel/x86/ztrsm_kernel_LN_2x2_sse.S View File

@@ -75,7 +75,7 @@
#define STACK_ALIGN 4096
#define STACK_OFFSET 1024

#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
#define PREFETCHSIZE (16 * 10 + 8)
#define WPREFETCHSIZE 112
#define PREFETCH prefetch
@@ -533,7 +533,7 @@
addps %xmm0, %xmm7
movsd 16 * SIZE(AA), %xmm0
mulps %xmm1, %xmm2
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht1 (PREFETCHSIZE + 16) * SIZE(AA)
#endif
addps %xmm2, %xmm4


+ 2
- 2
kernel/x86/ztrsm_kernel_LT_2x2_sse.S View File

@@ -75,7 +75,7 @@
#define STACK_ALIGN 4096
#define STACK_OFFSET 1024

#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
#define PREFETCHSIZE (16 * 10 + 8)
#define WPREFETCHSIZE 112
#define PREFETCH prefetch
@@ -994,7 +994,7 @@
addps %xmm0, %xmm7
movsd 16 * SIZE(AA), %xmm0
mulps %xmm1, %xmm2
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht1 (PREFETCHSIZE + 16) * SIZE(AA)
#endif
addps %xmm2, %xmm4


+ 2
- 2
kernel/x86/ztrsm_kernel_RT_2x2_sse.S View File

@@ -75,7 +75,7 @@
#define STACK_ALIGN 4096
#define STACK_OFFSET 1024

#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
#define PREFETCHSIZE (16 * 10 + 8)
#define WPREFETCHSIZE 112
#define PREFETCH prefetch
@@ -1820,7 +1820,7 @@
addps %xmm0, %xmm7
movsd 16 * SIZE(AA), %xmm0
mulps %xmm1, %xmm2
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCATE)
#if defined(OPTERON) || defined(BARCELONA) || defined(BOBCAT)
prefetcht1 (PREFETCHSIZE + 16) * SIZE(AA)
#endif
addps %xmm2, %xmm4


kernel/x86_64/KERNEL.BOBCATE → kernel/x86_64/KERNEL.BOBCAT View File


+ 1
- 1
kernel/x86_64/symv_L_sse.S View File

@@ -76,7 +76,7 @@
#define movsd movlps
#endif

#if defined(BARCELONA) || defined(SHANGHAI) || defined(BOBCATE)
#if defined(BARCELONA) || defined(SHANGHAI) || defined(BOBCAT)
#define PREFETCH prefetch
#define PREFETCHW prefetchw
#define PREFETCHSIZE (16 * 16)


+ 1
- 1
kernel/x86_64/symv_L_sse2.S View File

@@ -76,7 +76,7 @@
#define movsd movlpd
#endif

#if defined(BARCELONA) || defined(SHANGHAI) || defined(BOBCATE)
#if defined(BARCELONA) || defined(SHANGHAI) || defined(BOBCAT)
#define PREFETCH prefetch
#define PREFETCHW prefetchw
#define PREFETCHSIZE (16 * 16)


+ 1
- 1
kernel/x86_64/symv_U_sse.S View File

@@ -76,7 +76,7 @@
#define movsd movlps
#endif

#if defined(BARCELONA) || defined(SHANGHAI) || defined(BOBCATE)
#if defined(BARCELONA) || defined(SHANGHAI) || defined(BOBCAT)
#define PREFETCH prefetch
#define PREFETCHW prefetchw
#define PREFETCHSIZE (16 * 16)


+ 1
- 1
kernel/x86_64/symv_U_sse2.S View File

@@ -76,7 +76,7 @@
#define movsd movlpd
#endif

#if defined(BARCELONA) || defined(SHANGHAI) || defined(BOBCATE)
#if defined(BARCELONA) || defined(SHANGHAI) || defined(BOBCAT)
#define PREFETCH prefetch
#define PREFETCHW prefetchw
#define PREFETCHSIZE (16 * 16)


+ 1
- 1
kernel/x86_64/zsymv_L_sse.S View File

@@ -160,7 +160,7 @@
#define a3 %xmm14
#define xt1 %xmm15

#if (defined(HAVE_SSE3) && !defined(CORE_OPTERON)) || defined(BARCELONA) || defined(SHANGHAI) || defined(BOBCATE)
#if (defined(HAVE_SSE3) && !defined(CORE_OPTERON)) || defined(BARCELONA) || defined(SHANGHAI) || defined(BOBCAT)
#define MOVDDUP(a, b, c) movddup a(b), c
#define MOVDDUP2(a, b, c) movddup a##b, c
#else


+ 1
- 1
kernel/x86_64/zsymv_L_sse2.S View File

@@ -76,7 +76,7 @@
#define movsd movlpd
#endif

#if defined(BARCELONA) || defined(SHANGHAI) || defined(BOBCATE)
#if defined(BARCELONA) || defined(SHANGHAI) || defined(BOBCAT)
#define PREFETCH prefetch
#define PREFETCHW prefetchw
#define PREFETCHSIZE (16 * 16)


+ 1
- 1
kernel/x86_64/zsymv_U_sse.S View File

@@ -76,7 +76,7 @@
#define movsd movlpd
#endif

#if defined(BARCELONA) || defined(SHANGHAI) || defined(BOBCATE)
#if defined(BARCELONA) || defined(SHANGHAI) || defined(BOBCAT)
#define PREFETCH prefetch
#define PREFETCHW prefetchw
#define PREFETCHSIZE (16 * 16)


+ 1
- 1
kernel/x86_64/zsymv_U_sse2.S View File

@@ -76,7 +76,7 @@
#define movsd movlpd
#endif

#if defined(BARCELONA) || defined(SHANGHAI) || defined(BOBCATE)
#if defined(BARCELONA) || defined(SHANGHAI) || defined(BOBCAT)
#define PREFETCH prefetch
#define PREFETCHW prefetchw
#define PREFETCHSIZE (16 * 16)


+ 1
- 1
l1param.h View File

@@ -67,7 +67,7 @@
#define ALIGNED_ACCESS
#endif

#ifdef BOBCATE
#ifdef BOBCAT
#define PREFETCH prefetch
#define PREFETCHW prefetchw
#define PREFETCHSIZE (128 * 5)


+ 1
- 1
l2param.h View File

@@ -85,7 +85,7 @@
#define movsd movlps
#endif

#if defined(BARCELONA) || defined(SHANGHAI) || defined(BOBCATE)
#if defined(BARCELONA) || defined(SHANGHAI) || defined(BOBCAT)
#define ALIGNED_ACCESS
#define MOVUPS_A movaps
#define MOVUPS_XL movaps


+ 1
- 1
param.h View File

@@ -222,7 +222,7 @@ USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

#endif

#if defined(BOBCATE)
#if defined(BOBCAT)

#define SNUMOPT 8
#define DNUMOPT 4


Loading…
Cancel
Save