diff options
author | Anton Samokhvalov <pg83@yandex.ru> | 2022-02-10 16:45:15 +0300 |
---|---|---|
committer | Daniil Cherednik <dcherednik@yandex-team.ru> | 2022-02-10 16:45:15 +0300 |
commit | 72cb13b4aff9bc9cf22e49251bc8fd143f82538f (patch) | |
tree | da2c34829458c7d4e74bdfbdf85dff449e9e7fb8 /contrib/libs/cxxsupp/builtins/i386/lshrdi3.S | |
parent | 778e51ba091dc39e7b7fcab2b9cf4dbedfb6f2b5 (diff) | |
download | ydb-72cb13b4aff9bc9cf22e49251bc8fd143f82538f.tar.gz |
Restoring authorship annotation for Anton Samokhvalov <pg83@yandex.ru>. Commit 1 of 2.
Diffstat (limited to 'contrib/libs/cxxsupp/builtins/i386/lshrdi3.S')
-rw-r--r-- | contrib/libs/cxxsupp/builtins/i386/lshrdi3.S | 118 |
1 files changed, 59 insertions, 59 deletions
diff --git a/contrib/libs/cxxsupp/builtins/i386/lshrdi3.S b/contrib/libs/cxxsupp/builtins/i386/lshrdi3.S index b80f11a380..74673c03de 100644 --- a/contrib/libs/cxxsupp/builtins/i386/lshrdi3.S +++ b/contrib/libs/cxxsupp/builtins/i386/lshrdi3.S @@ -1,59 +1,59 @@ -// This file is dual licensed under the MIT and the University of Illinois Open -// Source Licenses. See LICENSE.TXT for details. - -#include "../assembly.h" - -// di_int __lshrdi3(di_int input, int count); - -// This routine has some extra memory traffic, loading the 64-bit input via two -// 32-bit loads, then immediately storing it back to the stack via a single 64-bit -// store. This is to avoid a write-small, read-large stall. -// However, if callers of this routine can be safely assumed to store the argument -// via a 64-bt store, this is unnecessary memory traffic, and should be avoided. -// It can be turned off by defining the TRUST_CALLERS_USE_64_BIT_STORES macro. - -#ifdef __i386__ -#ifdef __SSE2__ - -.text -.balign 4 -DEFINE_COMPILERRT_FUNCTION(__lshrdi3) - movd 12(%esp), %xmm2 // Load count -#ifndef TRUST_CALLERS_USE_64_BIT_STORES - movd 4(%esp), %xmm0 - movd 8(%esp), %xmm1 - punpckldq %xmm1, %xmm0 // Load input -#else - movq 4(%esp), %xmm0 // Load input -#endif - psrlq %xmm2, %xmm0 // shift input by count - movd %xmm0, %eax - psrlq $32, %xmm0 - movd %xmm0, %edx - ret -END_COMPILERRT_FUNCTION(__lshrdi3) - -#else // Use GPRs instead of SSE2 instructions, if they aren't available. - -.text -.balign 4 -DEFINE_COMPILERRT_FUNCTION(__lshrdi3) - movl 12(%esp), %ecx // Load count - movl 8(%esp), %edx // Load high - movl 4(%esp), %eax // Load low - - testl $0x20, %ecx // If count >= 32 - jnz 1f // goto 1 - - shrdl %cl, %edx, %eax // right shift low by count - shrl %cl, %edx // right shift high by count - ret - -1: movl %edx, %eax // Move high to low - xorl %edx, %edx // clear high - shrl %cl, %eax // shift low by count - 32 - ret -END_COMPILERRT_FUNCTION(__lshrdi3) - -#endif // __SSE2__ -#endif // __i386__ +// This file is dual licensed under the MIT and the University of Illinois Open +// Source Licenses. See LICENSE.TXT for details. + +#include "../assembly.h" + +// di_int __lshrdi3(di_int input, int count); + +// This routine has some extra memory traffic, loading the 64-bit input via two +// 32-bit loads, then immediately storing it back to the stack via a single 64-bit +// store. This is to avoid a write-small, read-large stall. +// However, if callers of this routine can be safely assumed to store the argument +// via a 64-bt store, this is unnecessary memory traffic, and should be avoided. +// It can be turned off by defining the TRUST_CALLERS_USE_64_BIT_STORES macro. + +#ifdef __i386__ +#ifdef __SSE2__ + +.text +.balign 4 +DEFINE_COMPILERRT_FUNCTION(__lshrdi3) + movd 12(%esp), %xmm2 // Load count +#ifndef TRUST_CALLERS_USE_64_BIT_STORES + movd 4(%esp), %xmm0 + movd 8(%esp), %xmm1 + punpckldq %xmm1, %xmm0 // Load input +#else + movq 4(%esp), %xmm0 // Load input +#endif + psrlq %xmm2, %xmm0 // shift input by count + movd %xmm0, %eax + psrlq $32, %xmm0 + movd %xmm0, %edx + ret +END_COMPILERRT_FUNCTION(__lshrdi3) + +#else // Use GPRs instead of SSE2 instructions, if they aren't available. + +.text +.balign 4 +DEFINE_COMPILERRT_FUNCTION(__lshrdi3) + movl 12(%esp), %ecx // Load count + movl 8(%esp), %edx // Load high + movl 4(%esp), %eax // Load low + + testl $0x20, %ecx // If count >= 32 + jnz 1f // goto 1 + + shrdl %cl, %edx, %eax // right shift low by count + shrl %cl, %edx // right shift high by count + ret + +1: movl %edx, %eax // Move high to low + xorl %edx, %edx // clear high + shrl %cl, %eax // shift low by count - 32 + ret +END_COMPILERRT_FUNCTION(__lshrdi3) + +#endif // __SSE2__ +#endif // __i386__ |