public/tomsfastmath
3
0
Fork 0
forked from ibphoenix/tomsfastmath
Code Releases Activity

ed1a1e90e9 was not only a x86-64 problem

Browse Source
This commit is contained in:
Steffen Jaeckel 2014-06-21 13:24:03 +02:00
parent 60cf64919f
commit 7882b0370d
3 changed files with 39 additions and 39 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
src/mont/fp_montgomery_reduce.c
View File

@ -30,7 +30,7 @@ asm( \
"movl %%edx,%1 \n\t" \
:"=g"(_c[LO]), "=r"(cy) \
:"0"(_c[LO]), "1"(cy), "g"(mu), "g"(*tmpm++) \
: "%eax", "%edx", "%cc")
: "%eax", "%edx", "cc")
#define PROPCARRY \
asm( \
@ -39,7 +39,7 @@ asm( \
"movzbl %%al,%1 \n\t" \
:"=g"(_c[LO]), "=r"(cy) \
:"0"(_c[LO]), "1"(cy) \
: "%eax", "%cc")
: "%eax", "cc")
/******************************************************************/
#elif defined(TFM_X86_64)
@ -280,7 +280,7 @@ asm( \
"movzbl %%al,%1 \n\t" \
:"=g"(_c[LO]), "=r"(cy) \
:"0"(_c[LO]), "1"(cy) \
: "%eax", "%cc")
: "%eax", "cc")
/******************************************************************/
#elif defined(TFM_ARM)
@ -300,7 +300,7 @@ asm( \
" MOVCC %0,#0 \n\t" \
" UMLAL r0,%0,%3,%4 \n\t" \
" STR r0,%1 \n\t" \
:"=r"(cy),"=m"(_c[0]):"0"(cy),"r"(mu),"r"(*tmpm++),"1"(_c[0]):"r0","%cc");
:"=r"(cy),"=m"(_c[0]):"0"(cy),"r"(mu),"r"(*tmpm++),"1"(_c[0]):"r0","cc");
#define PROPCARRY \
asm( \
@ -309,7 +309,7 @@ asm( \
" STR r0,%1 \n\t" \
" MOVCS %0,#1 \n\t" \
" MOVCC %0,#0 \n\t" \
:"=r"(cy),"=m"(_c[0]):"0"(cy),"1"(_c[0]):"r0","%cc");
:"=r"(cy),"=m"(_c[0]):"0"(cy),"1"(_c[0]):"r0","cc");
/******************************************************************/
#elif defined(TFM_PPC32)
@ -331,7 +331,7 @@ asm( \
" addc 16,16,18 \n\t" \
" addze %0,17 \n\t" \
" stw 16,%1 \n\t" \
:"=r"(cy),"=g"(_c[0]):"0"(cy),"r"(mu),"r"(tmpm[0]),"1"(_c[0]):"16", "17", "18","%cc"); ++tmpm;
:"=r"(cy),"=g"(_c[0]):"0"(cy),"r"(mu),"r"(tmpm[0]),"1"(_c[0]):"16", "17", "18","cc"); ++tmpm;
#define PROPCARRY \
asm( \
@ -340,7 +340,7 @@ asm( \
" stw 16,%1 \n\t" \
" xor %0,%0,%0 \n\t" \
" addze %0,%0 \n\t" \
:"=r"(cy),"=g"(_c[0]):"0"(cy),"1"(_c[0]):"16","%cc");
:"=r"(cy),"=g"(_c[0]):"0"(cy),"1"(_c[0]):"16","cc");
/******************************************************************/
#elif defined(TFM_PPC64)
@ -362,7 +362,7 @@ asm( \
" addc r16,r16,r18 \n\t" \
" addze %0,r17 \n\t" \
" sdx r16,0,%1 \n\t" \
:"=r"(cy),"=m"(_c[0]):"0"(cy),"r"(mu),"r"(tmpm[0]),"1"(_c[0]):"r16", "r17", "r18","%cc"); ++tmpm;
:"=r"(cy),"=m"(_c[0]):"0"(cy),"r"(mu),"r"(tmpm[0]),"1"(_c[0]):"r16", "r17", "r18","cc"); ++tmpm;
#define PROPCARRY \
asm( \
@ -371,7 +371,7 @@ asm( \
" sdx r16,0,%1 \n\t" \
" xor %0,%0,%0 \n\t" \
" addze %0,%0 \n\t" \
:"=r"(cy),"=m"(_c[0]):"0"(cy),"1"(_c[0]):"r16","%cc");
:"=r"(cy),"=m"(_c[0]):"0"(cy),"1"(_c[0]):"r16","cc");
/******************************************************************/
#elif defined(TFM_AVR32)
@ -401,7 +401,7 @@ asm( \
" st.w %1,r2 \n\t" \
" eor %0,%0 \n\t" \
" acr %0 \n\t" \
:"=r"(cy),"=r"(&_c[0]):"0"(cy),"1"(&_c[0]):"r2","%cc");
:"=r"(cy),"=r"(&_c[0]):"0"(cy),"1"(&_c[0]):"r2","cc");
/******************************************************************/
#elif defined(TFM_MIPS)

6
src/mul/fp_mul_comba.c
View File

@ -53,7 +53,7 @@ asm( \
"addl %%eax,%0 \n\t" \
"adcl %%edx,%1 \n\t" \
"adcl $0,%2 \n\t" \
:"=r"(c0), "=r"(c1), "=r"(c2): "0"(c0), "1"(c1), "2"(c2), "m"(i), "m"(j) :"%eax","%edx","%cc");
:"=r"(c0), "=r"(c1), "=r"(c2): "0"(c0), "1"(c1), "2"(c2), "m"(i), "m"(j) :"%eax","%edx","cc");
#elif defined(TFM_X86_64)
/* x86-64 optimized */
@ -128,7 +128,7 @@ asm( \
"movd %%mm0,%%eax \n\t" \
"adcl %%eax,%1 \n\t" \
"adcl $0,%2 \n\t" \
:"=r"(c0), "=r"(c1), "=r"(c2): "0"(c0), "1"(c1), "2"(c2), "m"(i), "m"(j) :"%eax","%cc");
:"=r"(c0), "=r"(c1), "=r"(c2): "0"(c0), "1"(c1), "2"(c2), "m"(i), "m"(j) :"%eax","cc");
#elif defined(TFM_ARM)
/* ARM code */
@ -155,7 +155,7 @@ asm( \
" ADDS %0,%0,r0 \n\t" \
" ADCS %1,%1,r1 \n\t" \
" ADC %2,%2,#0 \n\t" \
:"=r"(c0), "=r"(c1), "=r"(c2) : "0"(c0), "1"(c1), "2"(c2), "r"(i), "r"(j) : "r0", "r1", "%cc");
:"=r"(c0), "=r"(c1), "=r"(c2) : "0"(c0), "1"(c1), "2"(c2), "r"(i), "r"(j) : "r0", "r1", "cc");
#elif defined(TFM_PPC32)
/* For 32-bit PPC */

52
src/sqr/fp_sqr_comba.c
View File

@ -41,7 +41,7 @@ asm( \
"addl %%eax,%0 \n\t" \
"adcl %%edx,%1 \n\t" \
"adcl $0,%2 \n\t" \
:"=r"(c0), "=r"(c1), "=r"(c2): "0"(c0), "1"(c1), "2"(c2), "m"(i) :"%eax","%edx","%cc");
:"=r"(c0), "=r"(c1), "=r"(c2): "0"(c0), "1"(c1), "2"(c2), "m"(i) :"%eax","%edx","cc");
#define SQRADD2(i, j) \
asm( \
@ -53,7 +53,7 @@ asm( \
"addl %%eax,%0 \n\t" \
"adcl %%edx,%1 \n\t" \
"adcl $0,%2 \n\t" \
:"=r"(c0), "=r"(c1), "=r"(c2): "0"(c0), "1"(c1), "2"(c2), "m"(i), "m"(j) :"%eax","%edx","%cc");
:"=r"(c0), "=r"(c1), "=r"(c2): "0"(c0), "1"(c1), "2"(c2), "m"(i), "m"(j) :"%eax","%edx","cc");
#define SQRADDSC(i, j) \
asm( \
@ -62,7 +62,7 @@ asm( \
"movl %%eax,%0 \n\t" \
"movl %%edx,%1 \n\t" \
"xorl %2,%2 \n\t" \
:"=r"(sc0), "=r"(sc1), "=r"(sc2): "0"(sc0), "1"(sc1), "2"(sc2), "g"(i), "g"(j) :"%eax","%edx","%cc");
:"=r"(sc0), "=r"(sc1), "=r"(sc2): "0"(sc0), "1"(sc1), "2"(sc2), "g"(i), "g"(j) :"%eax","%edx","cc");
#define SQRADDAC(i, j) \
asm( \
@ -71,7 +71,7 @@ asm( \
"addl %%eax,%0 \n\t" \
"adcl %%edx,%1 \n\t" \
"adcl $0,%2 \n\t" \
:"=r"(sc0), "=r"(sc1), "=r"(sc2): "0"(sc0), "1"(sc1), "2"(sc2), "g"(i), "g"(j) :"%eax","%edx","%cc");
:"=r"(sc0), "=r"(sc1), "=r"(sc2): "0"(sc0), "1"(sc1), "2"(sc2), "g"(i), "g"(j) :"%eax","%edx","cc");
#define SQRADDDB \
asm( \
@ -81,7 +81,7 @@ asm( \
"addl %6,%0 \n\t" \
"adcl %7,%1 \n\t" \
"adcl %8,%2 \n\t" \
:"=r"(c0), "=r"(c1), "=r"(c2) : "0"(c0), "1"(c1), "2"(c2), "r"(sc0), "r"(sc1), "r"(sc2) : "%cc");
:"=r"(c0), "=r"(c1), "=r"(c2) : "0"(c0), "1"(c1), "2"(c2), "r"(sc0), "r"(sc1), "r"(sc2) : "cc");
#elif defined(TFM_X86_64)
/* x86-64 optimized */
@ -181,7 +181,7 @@ asm( \
"movd %%mm0,%%eax \n\t" \
"adcl %%eax,%1 \n\t" \
"adcl $0,%2 \n\t" \
:"=r"(c0), "=r"(c1), "=r"(c2): "0"(c0), "1"(c1), "2"(c2), "m"(i) :"%eax","%cc");
:"=r"(c0), "=r"(c1), "=r"(c2): "0"(c0), "1"(c1), "2"(c2), "m"(i) :"%eax","cc");
#define SQRADD2(i, j) \
asm( \
@ -197,7 +197,7 @@ asm( \
"addl %%eax,%0 \n\t" \
"adcl %%edx,%1 \n\t" \
"adcl $0,%2 \n\t" \
:"=r"(c0), "=r"(c1), "=r"(c2): "0"(c0), "1"(c1), "2"(c2), "m"(i), "m"(j) :"%eax","%edx","%cc");
:"=r"(c0), "=r"(c1), "=r"(c2): "0"(c0), "1"(c1), "2"(c2), "m"(i), "m"(j) :"%eax","%edx","cc");
#define SQRADDSC(i, j) \
asm( \
@ -221,7 +221,7 @@ asm( \
"addl %%eax,%0 \n\t" \
"adcl %%edx,%1 \n\t" \
"adcl $0,%2 \n\t" \
:"=r"(sc0), "=r"(sc1), "=r"(sc2): "0"(sc0), "1"(sc1), "2"(sc2), "m"(i), "m"(j) :"%eax","%edx","%cc");
:"=r"(sc0), "=r"(sc1), "=r"(sc2): "0"(sc0), "1"(sc1), "2"(sc2), "m"(i), "m"(j) :"%eax","%edx","cc");
#define SQRADDDB \
asm( \
@ -231,7 +231,7 @@ asm( \
"addl %6,%0 \n\t" \
"adcl %7,%1 \n\t" \
"adcl %8,%2 \n\t" \
:"=r"(c0), "=r"(c1), "=r"(c2) : "0"(c0), "1"(c1), "2"(c2), "r"(sc0), "r"(sc1), "r"(sc2) : "%cc");
:"=r"(c0), "=r"(c1), "=r"(c2) : "0"(c0), "1"(c1), "2"(c2), "r"(sc0), "r"(sc1), "r"(sc2) : "cc");
#elif defined(TFM_ARM)
@ -260,7 +260,7 @@ asm( \
" ADDS %0,%0,r0 \n\t" \
" ADCS %1,%1,r1 \n\t" \
" ADC %2,%2,#0 \n\t" \
:"=r"(c0), "=r"(c1), "=r"(c2) : "0"(c0), "1"(c1), "2"(c2), "r"(i) : "r0", "r1", "%cc");
:"=r"(c0), "=r"(c1), "=r"(c2) : "0"(c0), "1"(c1), "2"(c2), "r"(i) : "r0", "r1", "cc");
/* for squaring some of the terms are doubled... */
#define SQRADD2(i, j) \
@ -272,13 +272,13 @@ asm( \
" ADDS %0,%0,r0 \n\t" \
" ADCS %1,%1,r1 \n\t" \
" ADC %2,%2,#0 \n\t" \
:"=r"(c0), "=r"(c1), "=r"(c2) : "0"(c0), "1"(c1), "2"(c2), "r"(i), "r"(j) : "r0", "r1", "%cc");
:"=r"(c0), "=r"(c1), "=r"(c2) : "0"(c0), "1"(c1), "2"(c2), "r"(i), "r"(j) : "r0", "r1", "cc");
#define SQRADDSC(i, j) \
asm( \
" UMULL %0,%1,%6,%7 \n\t" \
" SUB %2,%2,%2 \n\t" \
:"=r"(sc0), "=r"(sc1), "=r"(sc2) : "0"(sc0), "1"(sc1), "2"(sc2), "r"(i), "r"(j) : "%cc");
:"=r"(sc0), "=r"(sc1), "=r"(sc2) : "0"(sc0), "1"(sc1), "2"(sc2), "r"(i), "r"(j) : "cc");
#define SQRADDAC(i, j) \
asm( \
@ -286,7 +286,7 @@ asm( \
" ADDS %0,%0,r0 \n\t" \
" ADCS %1,%1,r1 \n\t" \
" ADC %2,%2,#0 \n\t" \
:"=r"(sc0), "=r"(sc1), "=r"(sc2) : "0"(sc0), "1"(sc1), "2"(sc2), "r"(i), "r"(j) : "r0", "r1", "%cc");
:"=r"(sc0), "=r"(sc1), "=r"(sc2) : "0"(sc0), "1"(sc1), "2"(sc2), "r"(i), "r"(j) : "r0", "r1", "cc");
#define SQRADDDB \
asm( \
@ -296,7 +296,7 @@ asm( \
" ADDS %0,%0,%3 \n\t" \
" ADCS %1,%1,%4 \n\t" \
" ADC %2,%2,%5 \n\t" \
:"=r"(c0), "=r"(c1), "=r"(c2) : "r"(sc0), "r"(sc1), "r"(sc2), "0"(c0), "1"(c1), "2"(c2) : "%cc");
:"=r"(c0), "=r"(c1), "=r"(c2) : "r"(sc0), "r"(sc1), "r"(sc2), "0"(c0), "1"(c1), "2"(c2) : "cc");
#elif defined(TFM_PPC32)
@ -326,7 +326,7 @@ asm( \
" mulhwu 16,%6,%6 \n\t" \
" adde %1,%1,16 \n\t" \
" addze %2,%2 \n\t" \
:"=r"(c0), "=r"(c1), "=r"(c2):"0"(c0), "1"(c1), "2"(c2), "r"(i):"16","%cc");
:"=r"(c0), "=r"(c1), "=r"(c2):"0"(c0), "1"(c1), "2"(c2), "r"(i):"16","cc");
/* for squaring some of the terms are doubled... */
#define SQRADD2(i, j) \
@ -339,14 +339,14 @@ asm( \
" addc %0,%0,16 \n\t" \
" adde %1,%1,17 \n\t" \
" addze %2,%2 \n\t" \
:"=r"(c0), "=r"(c1), "=r"(c2):"0"(c0), "1"(c1), "2"(c2), "r"(i), "r"(j):"16", "17","%cc");
:"=r"(c0), "=r"(c1), "=r"(c2):"0"(c0), "1"(c1), "2"(c2), "r"(i), "r"(j):"16", "17","cc");
#define SQRADDSC(i, j) \
asm( \
" mullw %0,%6,%7 \n\t" \
" mulhwu %1,%6,%7 \n\t" \
" xor %2,%2,%2 \n\t" \
:"=r"(sc0), "=r"(sc1), "=r"(sc2):"0"(sc0), "1"(sc1), "2"(sc2), "r"(i),"r"(j) : "%cc");
:"=r"(sc0), "=r"(sc1), "=r"(sc2):"0"(sc0), "1"(sc1), "2"(sc2), "r"(i),"r"(j) : "cc");
#define SQRADDAC(i, j) \
asm( \
@ -355,7 +355,7 @@ asm( \
" mulhwu 16,%6,%7 \n\t" \
" adde %1,%1,16 \n\t" \
" addze %2,%2 \n\t" \
:"=r"(sc0), "=r"(sc1), "=r"(sc2):"0"(sc0), "1"(sc1), "2"(sc2), "r"(i), "r"(j):"16", "%cc");
:"=r"(sc0), "=r"(sc1), "=r"(sc2):"0"(sc0), "1"(sc1), "2"(sc2), "r"(i), "r"(j):"16", "cc");
#define SQRADDDB \
asm( \
@ -365,7 +365,7 @@ asm( \
" addc %0,%0,%3 \n\t" \
" adde %1,%1,%4 \n\t" \
" adde %2,%2,%5 \n\t" \
:"=r"(c0), "=r"(c1), "=r"(c2) : "r"(sc0), "r"(sc1), "r"(sc2), "0"(c0), "1"(c1), "2"(c2) : "%cc");
:"=r"(c0), "=r"(c1), "=r"(c2) : "r"(sc0), "r"(sc1), "r"(sc2), "0"(c0), "1"(c1), "2"(c2) : "cc");
#elif defined(TFM_PPC64)
/* PPC64 */
@ -394,7 +394,7 @@ asm( \
" mulhdu r16,%6,%6 \n\t" \
" adde %1,%1,r16 \n\t" \
" addze %2,%2 \n\t" \
:"=r"(c0), "=r"(c1), "=r"(c2):"0"(c0), "1"(c1), "2"(c2), "r"(i):"r16","%cc");
:"=r"(c0), "=r"(c1), "=r"(c2):"0"(c0), "1"(c1), "2"(c2), "r"(i):"r16","cc");
/* for squaring some of the terms are doubled... */
#define SQRADD2(i, j) \
@ -407,14 +407,14 @@ asm( \
" addc %0,%0,r16 \n\t" \
" adde %1,%1,r17 \n\t" \
" addze %2,%2 \n\t" \
:"=r"(c0), "=r"(c1), "=r"(c2):"0"(c0), "1"(c1), "2"(c2), "r"(i), "r"(j):"r16", "r17","%cc");
:"=r"(c0), "=r"(c1), "=r"(c2):"0"(c0), "1"(c1), "2"(c2), "r"(i), "r"(j):"r16", "r17","cc");
#define SQRADDSC(i, j) \
asm( \
" mulld %0,%6,%7 \n\t" \
" mulhdu %1,%6,%7 \n\t" \
" xor %2,%2,%2 \n\t" \
:"=r"(sc0), "=r"(sc1), "=r"(sc2):"0"(sc0), "1"(sc1), "2"(sc2), "r"(i),"r"(j) : "%cc");
:"=r"(sc0), "=r"(sc1), "=r"(sc2):"0"(sc0), "1"(sc1), "2"(sc2), "r"(i),"r"(j) : "cc");
#define SQRADDAC(i, j) \
asm( \
@ -423,7 +423,7 @@ asm( \
" mulhdu r16,%6,%7 \n\t" \
" adde %1,%1,r16 \n\t" \
" addze %2,%2 \n\t" \
:"=r"(sc0), "=r"(sc1), "=r"(sc2):"0"(sc0), "1"(sc1), "2"(sc2), "r"(i), "r"(j):"r16", "%cc");
:"=r"(sc0), "=r"(sc1), "=r"(sc2):"0"(sc0), "1"(sc1), "2"(sc2), "r"(i), "r"(j):"r16", "cc");
#define SQRADDDB \
asm( \
@ -433,7 +433,7 @@ asm( \
" addc %0,%0,%3 \n\t" \
" adde %1,%1,%4 \n\t" \
" adde %2,%2,%5 \n\t" \
:"=r"(c0), "=r"(c1), "=r"(c2) : "r"(sc0), "r"(sc1), "r"(sc2), "0"(c0), "1"(c1), "2"(c2) : "%cc");
:"=r"(c0), "=r"(c1), "=r"(c2) : "r"(sc0), "r"(sc1), "r"(sc2), "0"(c0), "1"(c1), "2"(c2) : "cc");
#elif defined(TFM_AVR32)
@ -501,7 +501,7 @@ asm( \
" add %0,%0,%3 \n\t" \
" adc %1,%1,%4 \n\t" \
" adc %2,%2,%5 \n\t" \
:"=r"(c0), "=r"(c1), "=r"(c2) : "r"(sc0), "r"(sc1), "r"(sc2), "0"(c0), "1"(c1), "2"(c2) : "%cc");
:"=r"(c0), "=r"(c1), "=r"(c2) : "r"(sc0), "r"(sc1), "r"(sc2), "0"(c0), "1"(c1), "2"(c2) : "cc");
#elif defined(TFM_MIPS)
@ -571,7 +571,7 @@ asm( \
" mflo %0 \n\t" \
" mfhi %1 \n\t" \
" xor %2,%2,%2 \n\t" \
:"=r"(sc0), "=r"(sc1), "=r"(sc2):"0"(sc0), "1"(sc1), "2"(sc2), "r"(i),"r"(j) : "%cc");
:"=r"(sc0), "=r"(sc1), "=r"(sc2):"0"(sc0), "1"(sc1), "2"(sc2), "r"(i),"r"(j) : "cc");
#define SQRADDAC(i, j) \
asm( \