feedc0de/wolfssl
1
0
Fork 0
forked from wolfSSL/wolfssl
Code Pull Requests Activity

change inline assembly to more generic condition code in clobber list, clang likes it better this way

Browse Source
This commit is contained in:
toddouska
2013-03-24 11:09:58 -07:00
parent 75bb1a8fa7
commit 80e3c85737
1 changed files with 47 additions and 47 deletions
Download Patch File Download Diff File Expand all files Collapse all files

94
ctaocrypt/src/asm.c
View File

@@ -51,7 +51,7 @@ __asm__( \
"movl %%edx,%1 \n\t" \ "movl %%edx,%1 \n\t" \
:"=g"(_c[LO]), "=r"(cy) \ :"=g"(_c[LO]), "=r"(cy) \
:"0"(_c[LO]), "1"(cy), "g"(mu), "g"(*tmpm++) \ :"0"(_c[LO]), "1"(cy), "g"(mu), "g"(*tmpm++) \
: "%eax", "%edx", "%cc") : "%eax", "%edx", "cc")
#define PROPCARRY \ #define PROPCARRY \
__asm__( \ __asm__( \
@@ -60,7 +60,7 @@ __asm__( \
"movzbl %%al,%1 \n\t" \ "movzbl %%al,%1 \n\t" \
:"=g"(_c[LO]), "=r"(cy) \ :"=g"(_c[LO]), "=r"(cy) \
:"0"(_c[LO]), "1"(cy) \ :"0"(_c[LO]), "1"(cy) \
: "%eax", "%cc") : "%eax", "cc")
/******************************************************************/ /******************************************************************/
#elif defined(TFM_X86_64) #elif defined(TFM_X86_64)
@@ -83,7 +83,7 @@ __asm__( \
"movq %%rdx,%1 \n\t" \ "movq %%rdx,%1 \n\t" \
:"=g"(_c[LO]), "=r"(cy) \ :"=g"(_c[LO]), "=r"(cy) \
:"0"(_c[LO]), "1"(cy), "r"(mu), "r"(*tmpm++) \ :"0"(_c[LO]), "1"(cy), "r"(mu), "r"(*tmpm++) \
: "%rax", "%rdx", "%cc") : "%rax", "%rdx", "cc")
#define INNERMUL8 \ #define INNERMUL8 \
__asm__( \ __asm__( \
@@ -176,7 +176,7 @@ __asm__( \
\ \
:"=r"(_c), "=r"(cy) \ :"=r"(_c), "=r"(cy) \
: "0"(_c), "1"(cy), "g"(mu), "r"(tmpm)\ : "0"(_c), "1"(cy), "g"(mu), "r"(tmpm)\
: "%rax", "%rdx", "%r10", "%r11", "%cc") : "%rax", "%rdx", "%r10", "%r11", "cc")
#define PROPCARRY \ #define PROPCARRY \
@@ -186,7 +186,7 @@ __asm__( \
"movzbq %%al,%1 \n\t" \ "movzbq %%al,%1 \n\t" \
:"=g"(_c[LO]), "=r"(cy) \ :"=g"(_c[LO]), "=r"(cy) \
:"0"(_c[LO]), "1"(cy) \ :"0"(_c[LO]), "1"(cy) \
: "%rax", "%cc") : "%rax", "cc")
/******************************************************************/ /******************************************************************/
#elif defined(TFM_SSE2) #elif defined(TFM_SSE2)
@@ -304,7 +304,7 @@ __asm__( \
"movzbl %%al,%1 \n\t" \ "movzbl %%al,%1 \n\t" \
:"=g"(_c[LO]), "=r"(cy) \ :"=g"(_c[LO]), "=r"(cy) \
:"0"(_c[LO]), "1"(cy) \ :"0"(_c[LO]), "1"(cy) \
: "%eax", "%cc") : "%eax", "cc")
/******************************************************************/ /******************************************************************/
#elif defined(TFM_ARM) #elif defined(TFM_ARM)
@@ -324,7 +324,7 @@ __asm__( \
" MOVCC %0,#0 \n\t" \ " MOVCC %0,#0 \n\t" \
" UMLAL r0,%0,%3,%4 \n\t" \ " UMLAL r0,%0,%3,%4 \n\t" \
" STR r0,%1 \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 \ #define PROPCARRY \
__asm__( \ __asm__( \
@@ -333,7 +333,7 @@ __asm__( \
" STR r0,%1 \n\t" \ " STR r0,%1 \n\t" \
" MOVCS %0,#1 \n\t" \ " MOVCS %0,#1 \n\t" \
" MOVCC %0,#0 \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) #elif defined(TFM_PPC32)
@@ -354,7 +354,7 @@ __asm__( \
" addc 16,16,18 \n\t" \ " addc 16,16,18 \n\t" \
" addze %0,17 \n\t" \ " addze %0,17 \n\t" \
" stw 16,%1 \n\t" \ " stw 16,%1 \n\t" \
:"=r"(cy),"=m"(_c[0]):"0"(cy),"r"(mu),"r"(tmpm[0]),"1"(_c[0]):"16", "17", "18","%cc"); ++tmpm; :"=r"(cy),"=m"(_c[0]):"0"(cy),"r"(mu),"r"(tmpm[0]),"1"(_c[0]):"16", "17", "18","cc"); ++tmpm;
#define PROPCARRY \ #define PROPCARRY \
__asm__( \ __asm__( \
@@ -363,7 +363,7 @@ __asm__( \
" stw 16,%1 \n\t" \ " stw 16,%1 \n\t" \
" xor %0,%0,%0 \n\t" \ " xor %0,%0,%0 \n\t" \
" addze %0,%0 \n\t" \ " addze %0,%0 \n\t" \
:"=r"(cy),"=m"(_c[0]):"0"(cy),"1"(_c[0]):"16","%cc"); :"=r"(cy),"=m"(_c[0]):"0"(cy),"1"(_c[0]):"16","cc");
#elif defined(TFM_PPC64) #elif defined(TFM_PPC64)
@@ -384,7 +384,7 @@ __asm__( \
" addc 16,16,18 \n\t" \ " addc 16,16,18 \n\t" \
" addze %0,17 \n\t" \ " addze %0,17 \n\t" \
" sdx 16,0,%1 \n\t" \ " sdx 16,0,%1 \n\t" \
:"=r"(cy),"=m"(_c[0]):"0"(cy),"r"(mu),"r"(tmpm[0]),"1"(_c[0]):"16", "17", "18","%cc"); ++tmpm; :"=r"(cy),"=m"(_c[0]):"0"(cy),"r"(mu),"r"(tmpm[0]),"1"(_c[0]):"16", "17", "18","cc"); ++tmpm;
#define PROPCARRY \ #define PROPCARRY \
__asm__( \ __asm__( \
@@ -393,7 +393,7 @@ __asm__( \
" sdx 16,0,%1 \n\t" \ " sdx 16,0,%1 \n\t" \
" xor %0,%0,%0 \n\t" \ " xor %0,%0,%0 \n\t" \
" addze %0,%0 \n\t" \ " addze %0,%0 \n\t" \
:"=r"(cy),"=m"(_c[0]):"0"(cy),"1"(_c[0]):"16","%cc"); :"=r"(cy),"=m"(_c[0]):"0"(cy),"1"(_c[0]):"16","cc");
/******************************************************************/ /******************************************************************/
@@ -424,7 +424,7 @@ __asm__( \
" st.w %1,r2 \n\t" \ " st.w %1,r2 \n\t" \
" eor %0,%0 \n\t" \ " eor %0,%0 \n\t" \
" acr %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");
#else #else
@@ -482,7 +482,7 @@ __asm__( \
"addl %%eax,%0 \n\t" \ "addl %%eax,%0 \n\t" \
"adcl %%edx,%1 \n\t" \ "adcl %%edx,%1 \n\t" \
"adcl $0,%2 \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) \ #define SQRADD2(i, j) \
__asm__( \ __asm__( \
@@ -494,7 +494,7 @@ __asm__( \
"addl %%eax,%0 \n\t" \ "addl %%eax,%0 \n\t" \
"adcl %%edx,%1 \n\t" \ "adcl %%edx,%1 \n\t" \
"adcl $0,%2 \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) \ #define SQRADDSC(i, j) \
__asm__( \ __asm__( \
@@ -503,7 +503,7 @@ __asm__( \
"movl %%eax,%0 \n\t" \ "movl %%eax,%0 \n\t" \
"movl %%edx,%1 \n\t" \ "movl %%edx,%1 \n\t" \
"xorl %2,%2 \n\t" \ "xorl %2,%2 \n\t" \
:"=r"(sc0), "=r"(sc1), "=r"(sc2): "g"(i), "g"(j) :"%eax","%edx","%cc"); :"=r"(sc0), "=r"(sc1), "=r"(sc2): "g"(i), "g"(j) :"%eax","%edx","cc");
/* TAO removed sc0,1,2 as input to remove warning so %6,%7 become %3,%4 */ /* TAO removed sc0,1,2 as input to remove warning so %6,%7 become %3,%4 */
@@ -514,7 +514,7 @@ __asm__( \
"addl %%eax,%0 \n\t" \ "addl %%eax,%0 \n\t" \
"adcl %%edx,%1 \n\t" \ "adcl %%edx,%1 \n\t" \
"adcl $0,%2 \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 \ #define SQRADDDB \
__asm__( \ __asm__( \
@@ -524,7 +524,7 @@ __asm__( \
"addl %6,%0 \n\t" \ "addl %6,%0 \n\t" \
"adcl %7,%1 \n\t" \ "adcl %7,%1 \n\t" \
"adcl %8,%2 \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) #elif defined(TFM_X86_64)
/* x86-64 optimized */ /* x86-64 optimized */
@@ -552,7 +552,7 @@ __asm__( \
"addq %%rax,%0 \n\t" \ "addq %%rax,%0 \n\t" \
"adcq %%rdx,%1 \n\t" \ "adcq %%rdx,%1 \n\t" \
"adcq $0,%2 \n\t" \ "adcq $0,%2 \n\t" \
:"=r"(c0), "=r"(c1), "=r"(c2): "0"(c0), "1"(c1), "2"(c2), "g"(i) :"%rax","%rdx","%cc"); :"=r"(c0), "=r"(c1), "=r"(c2): "0"(c0), "1"(c1), "2"(c2), "g"(i) :"%rax","%rdx","cc");
#define SQRADD2(i, j) \ #define SQRADD2(i, j) \
__asm__( \ __asm__( \
@@ -564,7 +564,7 @@ __asm__( \
"addq %%rax,%0 \n\t" \ "addq %%rax,%0 \n\t" \
"adcq %%rdx,%1 \n\t" \ "adcq %%rdx,%1 \n\t" \
"adcq $0,%2 \n\t" \ "adcq $0,%2 \n\t" \
:"=r"(c0), "=r"(c1), "=r"(c2): "0"(c0), "1"(c1), "2"(c2), "g"(i), "g"(j) :"%rax","%rdx","%cc"); :"=r"(c0), "=r"(c1), "=r"(c2): "0"(c0), "1"(c1), "2"(c2), "g"(i), "g"(j) :"%rax","%rdx","cc");
#define SQRADDSC(i, j) \ #define SQRADDSC(i, j) \
__asm__( \ __asm__( \
@@ -573,7 +573,7 @@ __asm__( \
"movq %%rax,%0 \n\t" \ "movq %%rax,%0 \n\t" \
"movq %%rdx,%1 \n\t" \ "movq %%rdx,%1 \n\t" \
"xorq %2,%2 \n\t" \ "xorq %2,%2 \n\t" \
:"=r"(sc0), "=r"(sc1), "=r"(sc2): "g"(i), "g"(j) :"%rax","%rdx","%cc"); :"=r"(sc0), "=r"(sc1), "=r"(sc2): "g"(i), "g"(j) :"%rax","%rdx","cc");
/* TAO removed sc0,1,2 as input to remove warning so %6,%7 become %3,%4 */ /* TAO removed sc0,1,2 as input to remove warning so %6,%7 become %3,%4 */
@@ -584,7 +584,7 @@ __asm__( \
"addq %%rax,%0 \n\t" \ "addq %%rax,%0 \n\t" \
"adcq %%rdx,%1 \n\t" \ "adcq %%rdx,%1 \n\t" \
"adcq $0,%2 \n\t" \ "adcq $0,%2 \n\t" \
:"=r"(sc0), "=r"(sc1), "=r"(sc2): "0"(sc0), "1"(sc1), "2"(sc2), "g"(i), "g"(j) :"%rax","%rdx","%cc"); :"=r"(sc0), "=r"(sc1), "=r"(sc2): "0"(sc0), "1"(sc1), "2"(sc2), "g"(i), "g"(j) :"%rax","%rdx","cc");
#define SQRADDDB \ #define SQRADDDB \
__asm__( \ __asm__( \
@@ -594,7 +594,7 @@ __asm__( \
"addq %6,%0 \n\t" \ "addq %6,%0 \n\t" \
"adcq %7,%1 \n\t" \ "adcq %7,%1 \n\t" \
"adcq %8,%2 \n\t" \ "adcq %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_SSE2) #elif defined(TFM_SSE2)
@@ -626,7 +626,7 @@ __asm__( \
"movd %%mm0,%%eax \n\t" \ "movd %%mm0,%%eax \n\t" \
"adcl %%eax,%1 \n\t" \ "adcl %%eax,%1 \n\t" \
"adcl $0,%2 \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) \ #define SQRADD2(i, j) \
__asm__( \ __asm__( \
@@ -642,7 +642,7 @@ __asm__( \
"addl %%eax,%0 \n\t" \ "addl %%eax,%0 \n\t" \
"adcl %%edx,%1 \n\t" \ "adcl %%edx,%1 \n\t" \
"adcl $0,%2 \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) \ #define SQRADDSC(i, j) \
__asm__( \ __asm__( \
@@ -668,7 +668,7 @@ __asm__( \
"addl %%eax,%0 \n\t" \ "addl %%eax,%0 \n\t" \
"adcl %%edx,%1 \n\t" \ "adcl %%edx,%1 \n\t" \
"adcl $0,%2 \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 \ #define SQRADDDB \
__asm__( \ __asm__( \
@@ -678,7 +678,7 @@ __asm__( \
"addl %6,%0 \n\t" \ "addl %6,%0 \n\t" \
"adcl %7,%1 \n\t" \ "adcl %7,%1 \n\t" \
"adcl %8,%2 \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) #elif defined(TFM_ARM)
@@ -707,7 +707,7 @@ __asm__( \
" ADDS %0,%0,r0 \n\t" \ " ADDS %0,%0,r0 \n\t" \
" ADCS %1,%1,r1 \n\t" \ " ADCS %1,%1,r1 \n\t" \
" ADC %2,%2,#0 \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... */ /* for squaring some of the terms are doubled... */
#define SQRADD2(i, j) \ #define SQRADD2(i, j) \
@@ -719,13 +719,13 @@ __asm__( \
" ADDS %0,%0,r0 \n\t" \ " ADDS %0,%0,r0 \n\t" \
" ADCS %1,%1,r1 \n\t" \ " ADCS %1,%1,r1 \n\t" \
" ADC %2,%2,#0 \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) \ #define SQRADDSC(i, j) \
__asm__( \ __asm__( \
" UMULL %0,%1,%6,%7 \n\t" \ " UMULL %0,%1,%6,%7 \n\t" \
" SUB %2,%2,%2 \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) \ #define SQRADDAC(i, j) \
__asm__( \ __asm__( \
@@ -733,7 +733,7 @@ __asm__( \
" ADDS %0,%0,r0 \n\t" \ " ADDS %0,%0,r0 \n\t" \
" ADCS %1,%1,r1 \n\t" \ " ADCS %1,%1,r1 \n\t" \
" ADC %2,%2,#0 \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 \ #define SQRADDDB \
__asm__( \ __asm__( \
@@ -743,7 +743,7 @@ __asm__( \
" ADDS %0,%0,%3 \n\t" \ " ADDS %0,%0,%3 \n\t" \
" ADCS %1,%1,%4 \n\t" \ " ADCS %1,%1,%4 \n\t" \
" ADC %2,%2,%5 \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) #elif defined(TFM_PPC32)
@@ -773,7 +773,7 @@ __asm__( \
" mulhwu 16,%6,%6 \n\t" \ " mulhwu 16,%6,%6 \n\t" \
" adde %1,%1,16 \n\t" \ " adde %1,%1,16 \n\t" \
" addze %2,%2 \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... */ /* for squaring some of the terms are doubled... */
#define SQRADD2(i, j) \ #define SQRADD2(i, j) \
@@ -786,14 +786,14 @@ __asm__( \
" addc %0,%0,16 \n\t" \ " addc %0,%0,16 \n\t" \
" adde %1,%1,17 \n\t" \ " adde %1,%1,17 \n\t" \
" addze %2,%2 \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) \ #define SQRADDSC(i, j) \
__asm__( \ __asm__( \
" mullw %0,%6,%7 \n\t" \ " mullw %0,%6,%7 \n\t" \
" mulhwu %1,%6,%7 \n\t" \ " mulhwu %1,%6,%7 \n\t" \
" xor %2,%2,%2 \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) \ #define SQRADDAC(i, j) \
__asm__( \ __asm__( \
@@ -802,7 +802,7 @@ __asm__( \
" mulhwu 16,%6,%7 \n\t" \ " mulhwu 16,%6,%7 \n\t" \
" adde %1,%1,16 \n\t" \ " adde %1,%1,16 \n\t" \
" addze %2,%2 \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 \ #define SQRADDDB \
__asm__( \ __asm__( \
@@ -812,7 +812,7 @@ __asm__( \
" addc %0,%0,%3 \n\t" \ " addc %0,%0,%3 \n\t" \
" adde %1,%1,%4 \n\t" \ " adde %1,%1,%4 \n\t" \
" adde %2,%2,%5 \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) #elif defined(TFM_PPC64)
/* PPC64 */ /* PPC64 */
@@ -841,7 +841,7 @@ __asm__( \
" mulhdu 16,%6,%6 \n\t" \ " mulhdu 16,%6,%6 \n\t" \
" adde %1,%1,16 \n\t" \ " adde %1,%1,16 \n\t" \
" addze %2,%2 \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... */ /* for squaring some of the terms are doubled... */
#define SQRADD2(i, j) \ #define SQRADD2(i, j) \
@@ -854,14 +854,14 @@ __asm__( \
" addc %0,%0,16 \n\t" \ " addc %0,%0,16 \n\t" \
" adde %1,%1,17 \n\t" \ " adde %1,%1,17 \n\t" \
" addze %2,%2 \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) \ #define SQRADDSC(i, j) \
__asm__( \ __asm__( \
" mulld %0,%6,%7 \n\t" \ " mulld %0,%6,%7 \n\t" \
" mulhdu %1,%6,%7 \n\t" \ " mulhdu %1,%6,%7 \n\t" \
" xor %2,%2,%2 \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) \ #define SQRADDAC(i, j) \
__asm__( \ __asm__( \
@@ -870,7 +870,7 @@ __asm__( \
" mulhdu 16,%6,%7 \n\t" \ " mulhdu 16,%6,%7 \n\t" \
" adde %1,%1,16 \n\t" \ " adde %1,%1,16 \n\t" \
" addze %2,%2 \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 \ #define SQRADDDB \
__asm__( \ __asm__( \
@@ -880,7 +880,7 @@ __asm__( \
" addc %0,%0,%3 \n\t" \ " addc %0,%0,%3 \n\t" \
" adde %1,%1,%4 \n\t" \ " adde %1,%1,%4 \n\t" \
" adde %2,%2,%5 \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) #elif defined(TFM_AVR32)
@@ -948,7 +948,7 @@ __asm__( \
" add %0,%0,%3 \n\t" \ " add %0,%0,%3 \n\t" \
" adc %1,%1,%4 \n\t" \ " adc %1,%1,%4 \n\t" \
" adc %2,%2,%5 \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");
#else #else
@@ -1069,7 +1069,7 @@ __asm__( \
"addl %%eax,%0 \n\t" \ "addl %%eax,%0 \n\t" \
"adcl %%edx,%1 \n\t" \ "adcl %%edx,%1 \n\t" \
"adcl $0,%2 \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) #elif defined(TFM_X86_64)
/* x86-64 optimized */ /* x86-64 optimized */
@@ -1104,7 +1104,7 @@ __asm__ ( \
"addq %%rax,%0 \n\t" \ "addq %%rax,%0 \n\t" \
"adcq %%rdx,%1 \n\t" \ "adcq %%rdx,%1 \n\t" \
"adcq $0,%2 \n\t" \ "adcq $0,%2 \n\t" \
:"=r"(c0), "=r"(c1), "=r"(c2): "0"(c0), "1"(c1), "2"(c2), "g"(i), "g"(j) :"%rax","%rdx","%cc"); :"=r"(c0), "=r"(c1), "=r"(c2): "0"(c0), "1"(c1), "2"(c2), "g"(i), "g"(j) :"%rax","%rdx","cc");
#elif defined(TFM_SSE2) #elif defined(TFM_SSE2)
/* use SSE2 optimizations */ /* use SSE2 optimizations */
@@ -1144,7 +1144,7 @@ __asm__( \
"movd %%mm0,%%eax \n\t" \ "movd %%mm0,%%eax \n\t" \
"adcl %%eax,%1 \n\t" \ "adcl %%eax,%1 \n\t" \
"adcl $0,%2 \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) #elif defined(TFM_ARM)
/* ARM code */ /* ARM code */
@@ -1171,7 +1171,7 @@ __asm__( \
" ADDS %0,%0,r0 \n\t" \ " ADDS %0,%0,r0 \n\t" \
" ADCS %1,%1,r1 \n\t" \ " ADCS %1,%1,r1 \n\t" \
" ADC %2,%2,#0 \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) #elif defined(TFM_PPC32)
/* For 32-bit PPC */ /* For 32-bit PPC */