Skip to content

Commit

Permalink
Remove commented-out inline asm, use clang/gcc intrinsics
Browse files Browse the repository at this point in the history
  • Loading branch information
@hikari-no-yume
hikari-no-yume committed Jan 6, 2015
1 parent 5dc8960 commit bdcd9bb
Show file tree
Hide file tree
Showing 2 changed files with 59 additions and 264 deletions.
103 changes: 25 additions & 78 deletions Zend/zend_multiply.h
View file
Original file line number Diff line number Diff line change
Expand Up @@ -24,92 +24,39 @@

#include "zend_bigint.h"


/* assembly commented-out as it uses the old float overflow behaviour
* however, now longs overflow to bigints, so we can't use it */

#if 0
#if defined(__i386__) && defined(__GNUC__)

#define ZEND_SIGNED_MULTIPLY_LONG(a, b, lval, dval, usedval) do { \
zend_long __tmpvar; \
__asm__ ("imul %3,%0\n" \
"adc $0,%1" \
: "=r"(__tmpvar),"=r"(usedval) \
: "0"(a), "r"(b), "1"(0)); \
if (usedval) (dval) = (double) (a) * (double) (b); \
else (lval) = __tmpvar; \
} while (0)

#elif defined(__x86_64__) && defined(__GNUC__)

#define ZEND_SIGNED_MULTIPLY_LONG(a, b, lval, dval, usedval) do { \
zend_long __tmpvar; \
__asm__ ("imul %3,%0\n" \
"adc $0,%1" \
: "=r"(__tmpvar),"=r"(usedval) \
: "0"(a), "r"(b), "1"(0)); \
if (usedval) (dval) = (double) (a) * (double) (b); \
else (lval) = __tmpvar; \
} while (0)

#elif defined(__arm__) && defined(__GNUC__)

#define ZEND_SIGNED_MULTIPLY_LONG(a, b, lval, dval, usedval) do { \
zend_long __tmpvar; \
__asm__("smull %0, %1, %2, %3\n" \
"sub %1, %1, %0, asr #31" \
: "=r"(__tmpvar), "=r"(usedval) \
: "r"(a), "r"(b)); \
if (usedval) (dval) = (double) (a) * (double) (b); \
else (lval) = __tmpvar; \
} while (0)

#elif defined(__aarch64__) && defined(__GNUC__)

#define ZEND_SIGNED_MULTIPLY_LONG(a, b, lval, dval, usedval) do { \
zend_long __tmpvar; \
__asm__("mul %0, %2, %3\n" \
"smulh %1, %2, %3\n" \
"sub %1, %1, %0, asr #63\n" \
: "=X"(__tmpvar), "=X"(usedval) \
: "X"(a), "X"(b)); \
if (usedval) (dval) = (double) (a) * (double) (b); \
else (lval) = __tmpvar; \
} while (0)
#endif
/* So __has_builtin won't cause an error if it's unavailable */
#ifndef __has_builtin
# define __has_builtin(x) 0
#endif

#if 0
#elif defined(ZEND_WIN32)
#if __has_builtin(__builtin_smul_overflow) && SIZEOF_LONG == SIZEOF_ZEND_LONG

#define ZEND_SIGNED_MULTIPLY_LONG(a, b, lval, dval, usedval) do { \
zend_long __lres = (a) * (b); \
long double __dres = (long double)(a) * (long double)(b); \
long double __delta = (long double) __lres - __dres; \
if ( ((usedval) = (( __dres + __delta ) != __dres))) { \
(dval) = __dres; \
#define ZEND_SIGNED_MULTIPLY_LONG(a, b, lval, big, usedval) do { \
long __lres = (a) * (b); \
if (__builtin_smull_overflow((a), (b), &__lres)) { \
zend_bigint *__out = zend_bigint_init_alloc(); \
zend_bigint_long_multiply_long(__out, a, b); \
(big) = __out; \
(usedval) = 1; \
} else { \
(lval) = __lres; \
(usedval) = 0; \
} \
} while (0)

#elif defined(__powerpc64__) && defined(__GNUC__)

#define ZEND_SIGNED_MULTIPLY_LONG(a, b, lval, dval, usedval) do { \
long __tmpvar; \
__asm__("li 14, 0\n\t" \
"mtxer 14\n\t" \
"mulldo. %0, %2,%3\n\t" \
"xor %1, %1, %1\n\t" \
"bns+ 0f\n\t" \
"li %1, 1\n\t" \
"0:\n" \
: "=r"(__tmpvar),"=r"(usedval) \
: "r"(a), "r"(b) \
: "r14", "cc"); \
if (usedval) (dval) = (double) (a) * (double) (b); \
else (lval) = __tmpvar; \
#elif __has_builtin(__builtin_saddll_overflow) && SIZEOF_LONG_LONG == SIZEOF_ZEND_LONG

#define ZEND_SIGNED_MULTIPLY_LONG(a, b, lval, big, usedval) do { \
long long __llres = (a) * (b); \
if (__builtin_smulll_overflow((a), (b), &__llres)) { \
zend_bigint *__out = zend_bigint_init_alloc(); \
zend_bigint_long_multiply_long(__out, a, b); \
(big) = __out; \
(usedval) = 1; \
} else { \
(lval) = __llres; \
(usedval) = 0; \
} \
} while (0)

#elif SIZEOF_ZEND_LONG == 4
Expand Down
220 changes: 34 additions & 186 deletions Zend/zend_operators.h
View file
Original file line number Diff line number Diff line change
Expand Up @@ -42,6 +42,10 @@
#include "ext/bcmath/libbcmath/src/bcmath.h"
#endif

/* So that using __has_builtin where unavailable won't cause a compile error */
#ifndef __has_builtin
# define __has_builtin(x) 0
#endif
#define LONG_SIGN_MASK (((zend_long)1) << (8*sizeof(zend_long)-1))

BEGIN_EXTERN_C()
Expand Down Expand Up @@ -472,104 +476,30 @@ static zend_always_inline int fast_add_function(zval *result, zval *op1, zval *o
{
if (EXPECTED(Z_TYPE_P(op1) == IS_LONG)) {
if (EXPECTED(Z_TYPE_P(op2) == IS_LONG)) {
/* assembly commented-out as it uses the old float overflow behaviour
* however, now longs overflow to bigints, so we can't use it */
#if 0
#if defined(__GNUC__) && defined(__i386__)
__asm__(
"movl (%1), %%eax\n\t"
"addl (%2), %%eax\n\t"
"jo 0f\n\t"
"movl %%eax, (%0)\n\t"
"movl %3, %c5(%0)\n\t"
"jmp 1f\n"
"0:\n\t"
"fildl (%1)\n\t"
"fildl (%2)\n\t"
"faddp %%st, %%st(1)\n\t"
"movl %4, %c5(%0)\n\t"
"fstpl (%0)\n"
"1:"
:
: "r"(&result->value),
"r"(&op1->value),
"r"(&op2->value),
"n"(IS_LONG),
"n"(IS_DOUBLE),
"n"(ZVAL_OFFSETOF_TYPE)
: "eax","cc");
#elif defined(__GNUC__) && defined(__x86_64__)
__asm__(
"movq (%1), %%rax\n\t"
"addq (%2), %%rax\n\t"
"jo 0f\n\t"
"movq %%rax, (%0)\n\t"
"movl %3, %c5(%0)\n\t"
"jmp 1f\n"
"0:\n\t"
"fildq (%1)\n\t"
"fildq (%2)\n\t"
"faddp %%st, %%st(1)\n\t"
"movl %4, %c5(%0)\n\t"
"fstpl (%0)\n"
"1:"
:
: "r"(&result->value),
"r"(&op1->value),
"r"(&op2->value),
"n"(IS_LONG),
"n"(IS_DOUBLE),
"n"(ZVAL_OFFSETOF_TYPE)
: "rax","cc");
#elif defined(__GNUC__) && defined(__powerpc64__)
__asm__(
"ld 14, 0(%1)\n\t"
"ld 15, 0(%2)\n\t"
"li 16, 0 \n\t"
"mtxer 16\n\t"
"addo. 14, 14, 15\n\t"
"bso- 0f\n\t"
"std 14, 0(%0)\n\t"
"li 14, %3\n\t"
"stw 14, %c5(%0)\n\t"
"b 1f\n"
"0:\n\t"
"lfd 0, 0(%1)\n\t"
"lfd 1, 0(%2)\n\t"
"fcfid 0, 0\n\t"
"fcfid 1, 1\n\t"
"fadd 0, 0, 1\n\t"
"li 14, %4\n\t"
"stw 14, %c5(%0)\n\t"
"stfd 0, 0(%0)\n"
"1:"
:
: "r"(&result->value),
"r"(&op1->value),
"r"(&op2->value),
"n"(IS_LONG),
"n"(IS_DOUBLE),
"n"(ZVAL_OFFSETOF_TYPE)
: "r14","r15","r16","fr0","fr1","cc");
#if __has_builtin(__builtin_saddl_overflow) && SIZEOF_LONG == SIZEOF_ZEND_LONG
long lresult;
if (!__builtin_saddl_overflow(Z_LVAL_P(op1), Z_LVAL_P(op2), &lresult)) {
ZVAL_LONG(result, lresult);
#elif __has_builtin(__builtin_saddll_overflow) && SIZEOF_LONG_LONG == SIZEOF_ZEND_LONG
long long llresult;
if (!__builtin_saddll_overflow(Z_LVAL_P(op1), Z_LVAL_P(op2), &llresult)) {
ZVAL_LONG(result, llresult);
#else
if (!(UNEXPECTED((Z_LVAL_P(op1) & LONG_SIGN_MASK) == (Z_LVAL_P(op2) & LONG_SIGN_MASK)
&& (Z_LVAL_P(op1) & LONG_SIGN_MASK) != ((Z_LVAL_P(op1) + Z_LVAL_P(op2)) & LONG_SIGN_MASK)))) {

ZVAL_LONG(result, Z_LVAL_P(op1) + Z_LVAL_P(op2));
#endif
#endif
/*
* 'result' may alias with op1 or op2, so we need to
* ensure that 'result' is not updated until after we
* have read the values of op1 and op2.
*/

if (UNEXPECTED((Z_LVAL_P(op1) & LONG_SIGN_MASK) == (Z_LVAL_P(op2) & LONG_SIGN_MASK)
&& (Z_LVAL_P(op1) & LONG_SIGN_MASK) != ((Z_LVAL_P(op1) + Z_LVAL_P(op2)) & LONG_SIGN_MASK))) {
} else {
/*
* 'result' may alias with op1 or op2, so we need to
* ensure that 'result' is not updated until after we
* have read the values of op1 and op2.
*/
zend_bigint *out = zend_bigint_init_alloc();
zend_bigint_long_add_long(out, Z_LVAL_P(op1), Z_LVAL_P(op2));
ZVAL_BIGINT(result, out);
} else {
ZVAL_LONG(result, Z_LVAL_P(op1) + Z_LVAL_P(op2));
}
#if 0
#endif
return SUCCESS;
} else if (EXPECTED(Z_TYPE_P(op2) == IS_DOUBLE)) {
ZVAL_DOUBLE(result, ((double)Z_LVAL_P(op1)) + Z_DVAL_P(op2));
Expand All @@ -591,106 +521,24 @@ static zend_always_inline int fast_sub_function(zval *result, zval *op1, zval *o
{
if (EXPECTED(Z_TYPE_P(op1) == IS_LONG)) {
if (EXPECTED(Z_TYPE_P(op2) == IS_LONG)) {
/* assembly commented-out as it uses the old float overflow behaviour
* however, now longs overflow to bigints, so we can't use it */
#if 0
#if defined(__GNUC__) && defined(__i386__)
__asm__(
"movl (%1), %%eax\n\t"
"subl (%2), %%eax\n\t"
"jo 0f\n\t"
"movl %%eax, (%0)\n\t"
"movl %3, %c5(%0)\n\t"
"jmp 1f\n"
"0:\n\t"
"fildl (%2)\n\t"
"fildl (%1)\n\t"
#if defined(__clang__) && (__clang_major__ < 2 || (__clang_major__ == 2 && __clang_minor__ < 10))
"fsubp %%st(1), %%st\n\t" /* LLVM bug #9164 */
#if __has_builtin(__builtin_ssubl_overflow) && SIZEOF_LONG == SIZEOF_ZEND_LONG
long lresult;
if (!__builtin_ssubl_overflow(Z_LVAL_P(op1), Z_LVAL_P(op2), &lresult)) {
ZVAL_LONG(result, lresult);
#elif __has_builtin(__builtin_ssubll_overflow) && SIZEOF_LONG_LONG == SIZEOF_ZEND_LONG
long long llresult;
if (!__builtin_ssubll_overflow(Z_LVAL_P(op1), Z_LVAL_P(op2), &llresult)) {
ZVAL_LONG(result, llresult);
#else
"fsubp %%st, %%st(1)\n\t"
if (!(UNEXPECTED((Z_LVAL_P(op1) & LONG_SIGN_MASK) != (Z_LVAL_P(op2) & LONG_SIGN_MASK)
&& (Z_LVAL_P(op1) & LONG_SIGN_MASK) != (Z_LVAL_P(result) & LONG_SIGN_MASK)))) {
ZVAL_LONG(result, Z_LVAL_P(op1) - Z_LVAL_P(op2));
#endif
"movl %4, %c5(%0)\n\t"
"fstpl (%0)\n"
"1:"
:
: "r"(&result->value),
"r"(&op1->value),
"r"(&op2->value),
"n"(IS_LONG),
"n"(IS_DOUBLE),
"n"(ZVAL_OFFSETOF_TYPE)
: "eax","cc");
#elif defined(__GNUC__) && defined(__x86_64__)
__asm__(
"movq (%1), %%rax\n\t"
"subq (%2), %%rax\n\t"
"jo 0f\n\t"
"movq %%rax, (%0)\n\t"
"movl %3, %c5(%0)\n\t"
"jmp 1f\n"
"0:\n\t"
"fildq (%2)\n\t"
"fildq (%1)\n\t"
#if defined(__clang__) && (__clang_major__ < 2 || (__clang_major__ == 2 && __clang_minor__ < 10))
"fsubp %%st(1), %%st\n\t" /* LLVM bug #9164 */
#else
"fsubp %%st, %%st(1)\n\t"
#endif
"movl %4, %c5(%0)\n\t"
"fstpl (%0)\n"
"1:"
:
: "r"(&result->value),
"r"(&op1->value),
"r"(&op2->value),
"n"(IS_LONG),
"n"(IS_DOUBLE),
"n"(ZVAL_OFFSETOF_TYPE)
: "rax","cc");
#elif defined(__GNUC__) && defined(__powerpc64__)
__asm__(
"ld 14, 0(%1)\n\t"
"ld 15, 0(%2)\n\t"
"li 16, 0\n\t"
"mtxer 16\n\t"
"subo. 14, 14, 15\n\t"
"bso- 0f\n\t"
"std 14, 0(%0)\n\t"
"li 14, %3\n\t"
"stw 14, %c5(%0)\n\t"
"b 1f\n"
"0:\n\t"
"lfd 0, 0(%1)\n\t"
"lfd 1, 0(%2)\n\t"
"fcfid 0, 0\n\t"
"fcfid 1, 1\n\t"
"fsub 0, 0, 1\n\t"
"li 14, %4\n\t"
"stw 14, %c5(%0)\n\t"
"stfd 0, 0(%0)\n"
"1:"
:
: "r"(&result->value),
"r"(&op1->value),
"r"(&op2->value),
"n"(IS_LONG),
"n"(IS_DOUBLE),
"n"(ZVAL_OFFSETOF_TYPE)
: "r14","r15","r16","fr0","fr1","cc");
#else
#endif
#endif
ZVAL_LONG(result, Z_LVAL_P(op1) - Z_LVAL_P(op2));

if (UNEXPECTED((Z_LVAL_P(op1) & LONG_SIGN_MASK) != (Z_LVAL_P(op2) & LONG_SIGN_MASK)
&& (Z_LVAL_P(op1) & LONG_SIGN_MASK) != (Z_LVAL_P(result) & LONG_SIGN_MASK))) {
} else {
zend_bigint *out = zend_bigint_init_alloc();
zend_bigint_long_subtract_long(out, Z_LVAL_P(op1), Z_LVAL_P(op2));
ZVAL_BIGINT(result, out);
}
#if 0
#endif
return SUCCESS;
} else if (EXPECTED(Z_TYPE_P(op2) == IS_DOUBLE)) {
ZVAL_DOUBLE(result, ((double)Z_LVAL_P(op1)) - Z_DVAL_P(op2));
Expand Down

0 comments on commit bdcd9bb

Please sign in to comment.