diff --git a/libc/arch-x86/string/ssse3-strcmp-latest.S b/libc/arch-x86/string/ssse3-strcmp-latest.S index 69c6425be..673ba570d 100644 --- a/libc/arch-x86/string/ssse3-strcmp-latest.S +++ b/libc/arch-x86/string/ssse3-strcmp-latest.S @@ -102,7 +102,7 @@ name: \ /* calculate left number to compare */ \ mov $16, %esi; \ sub %ecx, %esi; \ - cmp %esi, %ebp; \ + cmpl %esi, %ebp; \ jbe L(more8byteseq); \ sub %esi, %ebp #endif @@ -116,7 +116,7 @@ ENTRY (ssse3_strcmp_latest) movl STR2(%esp), %eax #ifdef USE_AS_STRNCMP movl CNT(%esp), %ebp - cmp $16, %ebp + cmpl $16, %ebp jb L(less16bytes_sncmp) jmp L(more16bytes) #endif @@ -172,18 +172,18 @@ ENTRY (ssse3_strcmp_latest) add $8, %edx add $8, %eax #ifdef USE_AS_STRNCMP - cmp $8, %ebp + cmpl $8, %ebp lea -8(%ebp), %ebp je L(eq) L(more16bytes): #endif movl %edx, %ecx and $0xfff, %ecx - cmp $0xff0, %ecx + cmpl $0xff0, %ecx ja L(crosspage) mov %eax, %ecx and $0xfff, %ecx - cmp $0xff0, %ecx + cmpl $0xff0, %ecx ja L(crosspage) pxor %xmm0, %xmm0 movlpd (%eax), %xmm1 @@ -197,7 +197,7 @@ L(more16bytes): sub $0xffff, %ecx jnz L(less16bytes) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(eq) #endif @@ -220,7 +220,7 @@ L(crosspage): xor %ecx, %eax xor %edi, %edx xor %ebx, %ebx - cmp %edi, %ecx + cmpl %edi, %ecx je L(ashr_0) ja L(bigger) or $0x20, %ebx @@ -229,37 +229,37 @@ L(crosspage): L(bigger): lea 15(%edi), %edi sub %ecx, %edi - cmp $8, %edi + cmpl $8, %edi jle L(ashr_less_8) - cmp $14, %edi + cmpl $14, %edi je L(ashr_15) - cmp $13, %edi + cmpl $13, %edi je L(ashr_14) - cmp $12, %edi + cmpl $12, %edi je L(ashr_13) - cmp $11, %edi + cmpl $11, %edi je L(ashr_12) - cmp $10, %edi + cmpl $10, %edi je L(ashr_11) - cmp $9, %edi + cmpl $9, %edi je L(ashr_10) L(ashr_less_8): je L(ashr_9) - cmp $7, %edi + cmpl $7, %edi je L(ashr_8) - cmp $6, %edi + cmpl $6, %edi je L(ashr_7) - cmp $5, %edi + cmpl $5, %edi je L(ashr_6) - cmp $4, %edi + cmpl $4, %edi je L(ashr_5) - cmp $3, %edi + cmpl $3, %edi je L(ashr_4) - cmp $2, %edi + cmpl $2, %edi je L(ashr_3) - cmp $1, %edi + cmpl $1, %edi je L(ashr_2) - cmp $0, %edi + cmpl $0, %edi je L(ashr_1) /* @@ -297,7 +297,7 @@ L(loop_ashr_0): sub $0xffff, %esi jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -355,7 +355,7 @@ L(gobble_ashr_1): sub $0xffff, %esi jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -380,7 +380,7 @@ L(gobble_ashr_1): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -396,7 +396,7 @@ L(nibble_ashr_1): jnz L(ashr_1_exittail) #ifdef USE_AS_STRNCMP - cmp $15, %ebp + cmpl $15, %ebp jbe L(ashr_1_exittail) #endif pxor %xmm0, %xmm0 @@ -462,7 +462,7 @@ L(gobble_ashr_2): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -486,7 +486,7 @@ L(gobble_ashr_2): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -502,7 +502,7 @@ L(nibble_ashr_2): jnz L(ashr_2_exittail) #ifdef USE_AS_STRNCMP - cmp $14, %ebp + cmpl $14, %ebp jbe L(ashr_2_exittail) #endif @@ -569,7 +569,7 @@ L(gobble_ashr_3): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -593,7 +593,7 @@ L(gobble_ashr_3): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -609,7 +609,7 @@ L(nibble_ashr_3): jnz L(ashr_3_exittail) #ifdef USE_AS_STRNCMP - cmp $13, %ebp + cmpl $13, %ebp jbe L(ashr_3_exittail) #endif pxor %xmm0, %xmm0 @@ -675,7 +675,7 @@ L(gobble_ashr_4): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -700,7 +700,7 @@ L(gobble_ashr_4): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -717,7 +717,7 @@ L(nibble_ashr_4): jnz L(ashr_4_exittail) #ifdef USE_AS_STRNCMP - cmp $12, %ebp + cmpl $12, %ebp jbe L(ashr_4_exittail) #endif @@ -784,7 +784,7 @@ L(gobble_ashr_5): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -808,7 +808,7 @@ L(gobble_ashr_5): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -824,7 +824,7 @@ L(nibble_ashr_5): jnz L(ashr_5_exittail) #ifdef USE_AS_STRNCMP - cmp $11, %ebp + cmpl $11, %ebp jbe L(ashr_5_exittail) #endif pxor %xmm0, %xmm0 @@ -891,7 +891,7 @@ L(gobble_ashr_6): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -915,7 +915,7 @@ L(gobble_ashr_6): sub $0xffff, %esi jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -932,7 +932,7 @@ L(nibble_ashr_6): jnz L(ashr_6_exittail) #ifdef USE_AS_STRNCMP - cmp $10, %ebp + cmpl $10, %ebp jbe L(ashr_6_exittail) #endif pxor %xmm0, %xmm0 @@ -999,7 +999,7 @@ L(gobble_ashr_7): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -1024,7 +1024,7 @@ L(gobble_ashr_7): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -1041,7 +1041,7 @@ L(nibble_ashr_7): jnz L(ashr_7_exittail) #ifdef USE_AS_STRNCMP - cmp $9, %ebp + cmpl $9, %ebp jbe L(ashr_7_exittail) #endif pxor %xmm0, %xmm0 @@ -1108,7 +1108,7 @@ L(gobble_ashr_8): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -1132,7 +1132,7 @@ L(gobble_ashr_8): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -1148,7 +1148,7 @@ L(nibble_ashr_8): jnz L(ashr_8_exittail) #ifdef USE_AS_STRNCMP - cmp $8, %ebp + cmpl $8, %ebp jbe L(ashr_8_exittail) #endif pxor %xmm0, %xmm0 @@ -1215,7 +1215,7 @@ L(gobble_ashr_9): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -1239,7 +1239,7 @@ L(gobble_ashr_9): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -1255,7 +1255,7 @@ L(nibble_ashr_9): jnz L(ashr_9_exittail) #ifdef USE_AS_STRNCMP - cmp $7, %ebp + cmpl $7, %ebp jbe L(ashr_9_exittail) #endif pxor %xmm0, %xmm0 @@ -1321,7 +1321,7 @@ L(gobble_ashr_10): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -1345,7 +1345,7 @@ L(gobble_ashr_10): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -1361,7 +1361,7 @@ L(nibble_ashr_10): jnz L(ashr_10_exittail) #ifdef USE_AS_STRNCMP - cmp $6, %ebp + cmpl $6, %ebp jbe L(ashr_10_exittail) #endif pxor %xmm0, %xmm0 @@ -1427,7 +1427,7 @@ L(gobble_ashr_11): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -1451,7 +1451,7 @@ L(gobble_ashr_11): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -1467,7 +1467,7 @@ L(nibble_ashr_11): jnz L(ashr_11_exittail) #ifdef USE_AS_STRNCMP - cmp $5, %ebp + cmpl $5, %ebp jbe L(ashr_11_exittail) #endif pxor %xmm0, %xmm0 @@ -1533,7 +1533,7 @@ L(gobble_ashr_12): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -1558,7 +1558,7 @@ L(gobble_ashr_12): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -1574,7 +1574,7 @@ L(nibble_ashr_12): jnz L(ashr_12_exittail) #ifdef USE_AS_STRNCMP - cmp $4, %ebp + cmpl $4, %ebp jbe L(ashr_12_exittail) #endif pxor %xmm0, %xmm0 @@ -1640,7 +1640,7 @@ L(gobble_ashr_13): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -1664,7 +1664,7 @@ L(gobble_ashr_13): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -1680,7 +1680,7 @@ L(nibble_ashr_13): jnz L(ashr_13_exittail) #ifdef USE_AS_STRNCMP - cmp $3, %ebp + cmpl $3, %ebp jbe L(ashr_13_exittail) #endif pxor %xmm0, %xmm0 @@ -1746,7 +1746,7 @@ L(gobble_ashr_14): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -1770,7 +1770,7 @@ L(gobble_ashr_14): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -1786,7 +1786,7 @@ L(nibble_ashr_14): jnz L(ashr_14_exittail) #ifdef USE_AS_STRNCMP - cmp $2, %ebp + cmpl $2, %ebp jbe L(ashr_14_exittail) #endif pxor %xmm0, %xmm0 @@ -1853,7 +1853,7 @@ L(gobble_ashr_15): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -1877,7 +1877,7 @@ L(gobble_ashr_15): jnz L(exit) #ifdef USE_AS_STRNCMP - cmp $16, %ebp + cmpl $16, %ebp lea -16(%ebp), %ebp jbe L(more8byteseq) #endif @@ -1893,7 +1893,7 @@ L(nibble_ashr_15): jnz L(ashr_15_exittail) #ifdef USE_AS_STRNCMP - cmp $1, %ebp + cmpl $1, %ebp jbe L(ashr_15_exittail) #endif pxor %xmm0, %xmm0 @@ -1955,12 +1955,12 @@ L(less16bytes): test $0x40, %cl jnz L(Byte6) #ifdef USE_AS_STRNCMP - cmp $7, %ebp + cmpl $7, %ebp jbe L(eq) #endif - movzx 7(%eax), %ecx - movzx 7(%edx), %eax + movzbl 7(%eax), %ecx + movzbl 7(%edx), %eax sub %ecx, %eax RETURN @@ -1968,11 +1968,11 @@ L(less16bytes): .p2align 4 L(Byte0): #ifdef USE_AS_STRNCMP - cmp $0, %ebp + cmpl $0, %ebp jbe L(eq) #endif - movzx (%eax), %ecx - movzx (%edx), %eax + movzbl (%eax), %ecx + movzbl (%edx), %eax sub %ecx, %eax RETURN @@ -1980,11 +1980,11 @@ L(Byte0): .p2align 4 L(Byte1): #ifdef USE_AS_STRNCMP - cmp $1, %ebp + cmpl $1, %ebp jbe L(eq) #endif - movzx 1(%eax), %ecx - movzx 1(%edx), %eax + movzbl 1(%eax), %ecx + movzbl 1(%edx), %eax sub %ecx, %eax RETURN @@ -1992,11 +1992,11 @@ L(Byte1): .p2align 4 L(Byte2): #ifdef USE_AS_STRNCMP - cmp $2, %ebp + cmpl $2, %ebp jbe L(eq) #endif - movzx 2(%eax), %ecx - movzx 2(%edx), %eax + movzbl 2(%eax), %ecx + movzbl 2(%edx), %eax sub %ecx, %eax RETURN @@ -2004,11 +2004,11 @@ L(Byte2): .p2align 4 L(Byte3): #ifdef USE_AS_STRNCMP - cmp $3, %ebp + cmpl $3, %ebp jbe L(eq) #endif - movzx 3(%eax), %ecx - movzx 3(%edx), %eax + movzbl 3(%eax), %ecx + movzbl 3(%edx), %eax sub %ecx, %eax RETURN @@ -2016,11 +2016,11 @@ L(Byte3): .p2align 4 L(Byte4): #ifdef USE_AS_STRNCMP - cmp $4, %ebp + cmpl $4, %ebp jbe L(eq) #endif - movzx 4(%eax), %ecx - movzx 4(%edx), %eax + movzbl 4(%eax), %ecx + movzbl 4(%edx), %eax sub %ecx, %eax RETURN @@ -2028,11 +2028,11 @@ L(Byte4): .p2align 4 L(Byte5): #ifdef USE_AS_STRNCMP - cmp $5, %ebp + cmpl $5, %ebp jbe L(eq) #endif - movzx 5(%eax), %ecx - movzx 5(%edx), %eax + movzbl 5(%eax), %ecx + movzbl 5(%edx), %eax sub %ecx, %eax RETURN @@ -2040,11 +2040,11 @@ L(Byte5): .p2align 4 L(Byte6): #ifdef USE_AS_STRNCMP - cmp $6, %ebp + cmpl $6, %ebp jbe L(eq) #endif - movzx 6(%eax), %ecx - movzx 6(%edx), %eax + movzbl 6(%eax), %ecx + movzbl 6(%edx), %eax sub %ecx, %eax RETURN @@ -2054,7 +2054,7 @@ L(2next_8_bytes): add $8, %eax add $8, %edx #ifdef USE_AS_STRNCMP - cmp $8, %ebp + cmpl $8, %ebp lea -8(%ebp), %ebp jbe L(eq) #endif @@ -2081,11 +2081,11 @@ L(2next_8_bytes): jnz L(Byte6) #ifdef USE_AS_STRNCMP - cmp $7, %ebp + cmpl $7, %ebp jbe L(eq) #endif - movzx 7(%eax), %ecx - movzx 7(%edx), %eax + movzbl 7(%eax), %ecx + movzbl 7(%edx), %eax sub %ecx, %eax RETURN @@ -2129,7 +2129,7 @@ L(less16bytes_sncmp): test %cl, %cl je L(eq) - cmp $1, %ebp + cmpl $1, %ebp je L(eq) movzbl 1(%eax), %ecx @@ -2138,7 +2138,7 @@ L(less16bytes_sncmp): test %cl, %cl je L(eq) - cmp $2, %ebp + cmpl $2, %ebp je L(eq) movzbl 2(%eax), %ecx @@ -2147,7 +2147,7 @@ L(less16bytes_sncmp): test %cl, %cl je L(eq) - cmp $3, %ebp + cmpl $3, %ebp je L(eq) movzbl 3(%eax), %ecx @@ -2156,7 +2156,7 @@ L(less16bytes_sncmp): test %cl, %cl je L(eq) - cmp $4, %ebp + cmpl $4, %ebp je L(eq) movzbl 4(%eax), %ecx @@ -2165,7 +2165,7 @@ L(less16bytes_sncmp): test %cl, %cl je L(eq) - cmp $5, %ebp + cmpl $5, %ebp je L(eq) movzbl 5(%eax), %ecx @@ -2174,7 +2174,7 @@ L(less16bytes_sncmp): test %cl, %cl je L(eq) - cmp $6, %ebp + cmpl $6, %ebp je L(eq) movzbl 6(%eax), %ecx @@ -2183,7 +2183,7 @@ L(less16bytes_sncmp): test %cl, %cl je L(eq) - cmp $7, %ebp + cmpl $7, %ebp je L(eq) movzbl 7(%eax), %ecx @@ -2193,7 +2193,7 @@ L(less16bytes_sncmp): je L(eq) - cmp $8, %ebp + cmpl $8, %ebp je L(eq) movzbl 8(%eax), %ecx @@ -2202,7 +2202,7 @@ L(less16bytes_sncmp): test %cl, %cl je L(eq) - cmp $9, %ebp + cmpl $9, %ebp je L(eq) movzbl 9(%eax), %ecx @@ -2211,7 +2211,7 @@ L(less16bytes_sncmp): test %cl, %cl je L(eq) - cmp $10, %ebp + cmpl $10, %ebp je L(eq) movzbl 10(%eax), %ecx @@ -2220,7 +2220,7 @@ L(less16bytes_sncmp): test %cl, %cl je L(eq) - cmp $11, %ebp + cmpl $11, %ebp je L(eq) movzbl 11(%eax), %ecx @@ -2230,7 +2230,7 @@ L(less16bytes_sncmp): je L(eq) - cmp $12, %ebp + cmpl $12, %ebp je L(eq) movzbl 12(%eax), %ecx @@ -2239,7 +2239,7 @@ L(less16bytes_sncmp): test %cl, %cl je L(eq) - cmp $13, %ebp + cmpl $13, %ebp je L(eq) movzbl 13(%eax), %ecx @@ -2248,7 +2248,7 @@ L(less16bytes_sncmp): test %cl, %cl je L(eq) - cmp $14, %ebp + cmpl $14, %ebp je L(eq) movzbl 14(%eax), %ecx @@ -2257,7 +2257,7 @@ L(less16bytes_sncmp): test %cl, %cl je L(eq) - cmp $15, %ebp + cmpl $15, %ebp je L(eq) movzbl 15(%eax), %ecx