This version contains changes for removing unnecessary typecasting only. Backing out remaining changes i.e. loop-unrolling. Though loop-unrolling makes sense from more space/less time perspective, code generated by GCC 4.8.2 with "gcc -O3 -mavx -s" and "gcc -O3 -m64 -s" for loop of 2, 4 and 8 iterations are same and "memcpy perf" from "make test" reveals similar results for "with and without loop". Will investigate this later. Ravi Kerur (1): Clean up rte_memcpy.h file .../common/include/arch/x86/rte_memcpy.h | 340 +++++++++++---------- 1 file changed, 175 insertions(+), 165 deletions(-) -- 1.9.1
Remove unnecessary type casting in functions. Tested on Ubuntu (14.04 x86_64) with "make test". "make test" results match the results with baseline. "Memcpy perf" results match the results with baseline. Signed-off-by: Ravi Kerur <rkerur@gmail.com> --- .../common/include/arch/x86/rte_memcpy.h | 340 +++++++++++---------- 1 file changed, 175 insertions(+), 165 deletions(-) diff --git a/lib/librte_eal/common/include/arch/x86/rte_memcpy.h b/lib/librte_eal/common/include/arch/x86/rte_memcpy.h index 6a57426..839d4ec 100644 --- a/lib/librte_eal/common/include/arch/x86/rte_memcpy.h +++ b/lib/librte_eal/common/include/arch/x86/rte_memcpy.h @@ -106,8 +106,8 @@ rte_mov32(uint8_t *dst, const uint8_t *src) static inline void rte_mov64(uint8_t *dst, const uint8_t *src) { - rte_mov32((uint8_t *)dst + 0 * 32, (const uint8_t *)src + 0 * 32); - rte_mov32((uint8_t *)dst + 1 * 32, (const uint8_t *)src + 1 * 32); + rte_mov32(dst + 0 * 32, src + 0 * 32); + rte_mov32(dst + 1 * 32, src + 1 * 32); } /** @@ -117,10 +117,10 @@ rte_mov64(uint8_t *dst, const uint8_t *src) static inline void rte_mov128(uint8_t *dst, const uint8_t *src) { - rte_mov32((uint8_t *)dst + 0 * 32, (const uint8_t *)src + 0 * 32); - rte_mov32((uint8_t *)dst + 1 * 32, (const uint8_t *)src + 1 * 32); - rte_mov32((uint8_t *)dst + 2 * 32, (const uint8_t *)src + 2 * 32); - rte_mov32((uint8_t *)dst + 3 * 32, (const uint8_t *)src + 3 * 32); + rte_mov32(dst + 0 * 32, src + 0 * 32); + rte_mov32(dst + 1 * 32, src + 1 * 32); + rte_mov32(dst + 2 * 32, src + 2 * 32); + rte_mov32(dst + 3 * 32, src + 3 * 32); } /** @@ -130,14 +130,14 @@ rte_mov128(uint8_t *dst, const uint8_t *src) static inline void rte_mov256(uint8_t *dst, const uint8_t *src) { - rte_mov32((uint8_t *)dst + 0 * 32, (const uint8_t *)src + 0 * 32); - rte_mov32((uint8_t *)dst + 1 * 32, (const uint8_t *)src + 1 * 32); - rte_mov32((uint8_t *)dst + 2 * 32, (const uint8_t *)src + 2 * 32); - rte_mov32((uint8_t *)dst + 3 * 32, (const uint8_t *)src + 3 * 32); - rte_mov32((uint8_t *)dst + 4 * 32, (const uint8_t *)src + 4 * 32); - rte_mov32((uint8_t *)dst + 5 * 32, (const uint8_t *)src + 5 * 32); - rte_mov32((uint8_t *)dst + 6 * 32, (const uint8_t *)src + 6 * 32); - rte_mov32((uint8_t *)dst + 7 * 32, (const uint8_t *)src + 7 * 32); + rte_mov32(dst + 0 * 32, src + 0 * 32); + rte_mov32(dst + 1 * 32, src + 1 * 32); + rte_mov32(dst + 2 * 32, src + 2 * 32); + rte_mov32(dst + 3 * 32, src + 3 * 32); + rte_mov32(dst + 4 * 32, src + 4 * 32); + rte_mov32(dst + 5 * 32, src + 5 * 32); + rte_mov32(dst + 6 * 32, src + 6 * 32); + rte_mov32(dst + 7 * 32, src + 7 * 32); } /** @@ -150,13 +150,16 @@ rte_mov64blocks(uint8_t *dst, const uint8_t *src, size_t n) __m256i ymm0, ymm1; while (n >= 64) { - ymm0 = _mm256_loadu_si256((const __m256i *)((const uint8_t *)src + 0 * 32)); + + ymm0 = _mm256_loadu_si256((const __m256i *)(src + 0 * 32)); + ymm1 = _mm256_loadu_si256((const __m256i *)(src + 1 * 32)); + + _mm256_storeu_si256((__m256i *)(dst + 0 * 32), ymm0); + _mm256_storeu_si256((__m256i *)(dst + 1 * 32), ymm1); + n -= 64; - ymm1 = _mm256_loadu_si256((const __m256i *)((const uint8_t *)src + 1 * 32)); - src = (const uint8_t *)src + 64; - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 0 * 32), ymm0); - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 1 * 32), ymm1); - dst = (uint8_t *)dst + 64; + src = src + 64; + dst = dst + 64; } } @@ -170,34 +173,39 @@ rte_mov256blocks(uint8_t *dst, const uint8_t *src, size_t n) __m256i ymm0, ymm1, ymm2, ymm3, ymm4, ymm5, ymm6, ymm7; while (n >= 256) { - ymm0 = _mm256_loadu_si256((const __m256i *)((const uint8_t *)src + 0 * 32)); + + ymm0 = _mm256_loadu_si256((const __m256i *)(src + 0 * 32)); + ymm1 = _mm256_loadu_si256((const __m256i *)(src + 1 * 32)); + ymm2 = _mm256_loadu_si256((const __m256i *)(src + 2 * 32)); + ymm3 = _mm256_loadu_si256((const __m256i *)(src + 3 * 32)); + ymm4 = _mm256_loadu_si256((const __m256i *)(src + 4 * 32)); + ymm5 = _mm256_loadu_si256((const __m256i *)(src + 5 * 32)); + ymm6 = _mm256_loadu_si256((const __m256i *)(src + 6 * 32)); + ymm7 = _mm256_loadu_si256((const __m256i *)(src + 7 * 32)); + + _mm256_storeu_si256((__m256i *)(dst + 0 * 32), ymm0); + _mm256_storeu_si256((__m256i *)(dst + 1 * 32), ymm1); + _mm256_storeu_si256((__m256i *)(dst + 2 * 32), ymm2); + _mm256_storeu_si256((__m256i *)(dst + 3 * 32), ymm3); + _mm256_storeu_si256((__m256i *)(dst + 4 * 32), ymm4); + _mm256_storeu_si256((__m256i *)(dst + 5 * 32), ymm5); + _mm256_storeu_si256((__m256i *)(dst + 6 * 32), ymm6); + _mm256_storeu_si256((__m256i *)(dst + 7 * 32), ymm7); + n -= 256; - ymm1 = _mm256_loadu_si256((const __m256i *)((const uint8_t *)src + 1 * 32)); - ymm2 = _mm256_loadu_si256((const __m256i *)((const uint8_t *)src + 2 * 32)); - ymm3 = _mm256_loadu_si256((const __m256i *)((const uint8_t *)src + 3 * 32)); - ymm4 = _mm256_loadu_si256((const __m256i *)((const uint8_t *)src + 4 * 32)); - ymm5 = _mm256_loadu_si256((const __m256i *)((const uint8_t *)src + 5 * 32)); - ymm6 = _mm256_loadu_si256((const __m256i *)((const uint8_t *)src + 6 * 32)); - ymm7 = _mm256_loadu_si256((const __m256i *)((const uint8_t *)src + 7 * 32)); - src = (const uint8_t *)src + 256; - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 0 * 32), ymm0); - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 1 * 32), ymm1); - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 2 * 32), ymm2); - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 3 * 32), ymm3); - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 4 * 32), ymm4); - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 5 * 32), ymm5); - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 6 * 32), ymm6); - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 7 * 32), ymm7); - dst = (uint8_t *)dst + 256; + src = src + 256; + dst = dst + 256; } } static inline void * -rte_memcpy(void *dst, const void *src, size_t n) +rte_memcpy(void *_dst, const void *_src, size_t n) { - uintptr_t dstu = (uintptr_t)dst; - uintptr_t srcu = (uintptr_t)src; - void *ret = dst; + const uint8_t *src = (const uint8_t *)_src; + uint8_t *dst = (uint8_t *)_dst; + uintptr_t dstu = (uintptr_t)_dst; + uintptr_t srcu = (uintptr_t)_src; + void *ret = _dst; size_t dstofss; size_t bits; @@ -230,43 +238,43 @@ rte_memcpy(void *dst, const void *src, size_t n) * Fast way when copy size doesn't exceed 512 bytes */ if (n <= 32) { - rte_mov16((uint8_t *)dst, (const uint8_t *)src); - rte_mov16((uint8_t *)dst - 16 + n, (const uint8_t *)src - 16 + n); + rte_mov16(dst, src); + rte_mov16(dst - 16 + n, src - 16 + n); return ret; } if (n <= 64) { - rte_mov32((uint8_t *)dst, (const uint8_t *)src); - rte_mov32((uint8_t *)dst - 32 + n, (const uint8_t *)src - 32 + n); + rte_mov32(dst, src); + rte_mov32(dst - 32 + n, src - 32 + n); return ret; } if (n <= 512) { if (n >= 256) { n -= 256; - rte_mov256((uint8_t *)dst, (const uint8_t *)src); - src = (const uint8_t *)src + 256; - dst = (uint8_t *)dst + 256; + rte_mov256(dst, src); + src = src + 256; + dst = dst + 256; } if (n >= 128) { n -= 128; - rte_mov128((uint8_t *)dst, (const uint8_t *)src); - src = (const uint8_t *)src + 128; - dst = (uint8_t *)dst + 128; + rte_mov128(dst, src); + src = src + 128; + dst = dst + 128; } if (n >= 64) { n -= 64; - rte_mov64((uint8_t *)dst, (const uint8_t *)src); - src = (const uint8_t *)src + 64; - dst = (uint8_t *)dst + 64; + rte_mov64(dst, src); + src = src + 64; + dst = dst + 64; } COPY_BLOCK_64_BACK31: if (n > 32) { - rte_mov32((uint8_t *)dst, (const uint8_t *)src); - rte_mov32((uint8_t *)dst - 32 + n, (const uint8_t *)src - 32 + n); + rte_mov32(dst, src); + rte_mov32(dst - 32 + n, src - 32 + n); return ret; } - if (n > 0) { - rte_mov32((uint8_t *)dst - 32 + n, (const uint8_t *)src - 32 + n); - } + if (n > 0) + rte_mov32(dst - 32 + n, src - 32 + n); + return ret; } @@ -275,21 +283,21 @@ COPY_BLOCK_64_BACK31: */ dstofss = 32 - ((uintptr_t)dst & 0x1F); n -= dstofss; - rte_mov32((uint8_t *)dst, (const uint8_t *)src); - src = (const uint8_t *)src + dstofss; - dst = (uint8_t *)dst + dstofss; + rte_mov32(dst, src); + src = src + dstofss; + dst = dst + dstofss; /** * Copy 256-byte blocks. * Use copy block function for better instruction order control, * which is important when load is unaligned. */ - rte_mov256blocks((uint8_t *)dst, (const uint8_t *)src, n); + rte_mov256blocks(dst, src, n); bits = n; n = n & 255; bits -= n; - src = (const uint8_t *)src + bits; - dst = (uint8_t *)dst + bits; + src = src + bits; + dst = dst + bits; /** * Copy 64-byte blocks. @@ -297,12 +305,12 @@ COPY_BLOCK_64_BACK31: * which is important when load is unaligned. */ if (n >= 64) { - rte_mov64blocks((uint8_t *)dst, (const uint8_t *)src, n); + rte_mov64blocks(dst, src, n); bits = n; n = n & 63; bits -= n; - src = (const uint8_t *)src + bits; - dst = (uint8_t *)dst + bits; + src = src + bits; + dst = dst + bits; } /** @@ -337,8 +345,8 @@ rte_mov16(uint8_t *dst, const uint8_t *src) static inline void rte_mov32(uint8_t *dst, const uint8_t *src) { - rte_mov16((uint8_t *)dst + 0 * 16, (const uint8_t *)src + 0 * 16); - rte_mov16((uint8_t *)dst + 1 * 16, (const uint8_t *)src + 1 * 16); + rte_mov16(dst + 0 * 16, src + 0 * 16); + rte_mov16(dst + 1 * 16, src + 1 * 16); } /** @@ -348,10 +356,10 @@ rte_mov32(uint8_t *dst, const uint8_t *src) static inline void rte_mov64(uint8_t *dst, const uint8_t *src) { - rte_mov16((uint8_t *)dst + 0 * 16, (const uint8_t *)src + 0 * 16); - rte_mov16((uint8_t *)dst + 1 * 16, (const uint8_t *)src + 1 * 16); - rte_mov16((uint8_t *)dst + 2 * 16, (const uint8_t *)src + 2 * 16); - rte_mov16((uint8_t *)dst + 3 * 16, (const uint8_t *)src + 3 * 16); + rte_mov16(dst + 0 * 16, src + 0 * 16); + rte_mov16(dst + 1 * 16, src + 1 * 16); + rte_mov16(dst + 2 * 16, src + 2 * 16); + rte_mov16(dst + 3 * 16, src + 3 * 16); } /** @@ -361,14 +369,14 @@ rte_mov64(uint8_t *dst, const uint8_t *src) static inline void rte_mov128(uint8_t *dst, const uint8_t *src) { - rte_mov16((uint8_t *)dst + 0 * 16, (const uint8_t *)src + 0 * 16); - rte_mov16((uint8_t *)dst + 1 * 16, (const uint8_t *)src + 1 * 16); - rte_mov16((uint8_t *)dst + 2 * 16, (const uint8_t *)src + 2 * 16); - rte_mov16((uint8_t *)dst + 3 * 16, (const uint8_t *)src + 3 * 16); - rte_mov16((uint8_t *)dst + 4 * 16, (const uint8_t *)src + 4 * 16); - rte_mov16((uint8_t *)dst + 5 * 16, (const uint8_t *)src + 5 * 16); - rte_mov16((uint8_t *)dst + 6 * 16, (const uint8_t *)src + 6 * 16); - rte_mov16((uint8_t *)dst + 7 * 16, (const uint8_t *)src + 7 * 16); + rte_mov16(dst + 0 * 16, src + 0 * 16); + rte_mov16(dst + 1 * 16, src + 1 * 16); + rte_mov16(dst + 2 * 16, src + 2 * 16); + rte_mov16(dst + 3 * 16, src + 3 * 16); + rte_mov16(dst + 4 * 16, src + 4 * 16); + rte_mov16(dst + 5 * 16, src + 5 * 16); + rte_mov16(dst + 6 * 16, src + 6 * 16); + rte_mov16(dst + 7 * 16, src + 7 * 16); } /** @@ -378,22 +386,22 @@ rte_mov128(uint8_t *dst, const uint8_t *src) static inline void rte_mov256(uint8_t *dst, const uint8_t *src) { - rte_mov16((uint8_t *)dst + 0 * 16, (const uint8_t *)src + 0 * 16); - rte_mov16((uint8_t *)dst + 1 * 16, (const uint8_t *)src + 1 * 16); - rte_mov16((uint8_t *)dst + 2 * 16, (const uint8_t *)src + 2 * 16); - rte_mov16((uint8_t *)dst + 3 * 16, (const uint8_t *)src + 3 * 16); - rte_mov16((uint8_t *)dst + 4 * 16, (const uint8_t *)src + 4 * 16); - rte_mov16((uint8_t *)dst + 5 * 16, (const uint8_t *)src + 5 * 16); - rte_mov16((uint8_t *)dst + 6 * 16, (const uint8_t *)src + 6 * 16); - rte_mov16((uint8_t *)dst + 7 * 16, (const uint8_t *)src + 7 * 16); - rte_mov16((uint8_t *)dst + 8 * 16, (const uint8_t *)src + 8 * 16); - rte_mov16((uint8_t *)dst + 9 * 16, (const uint8_t *)src + 9 * 16); - rte_mov16((uint8_t *)dst + 10 * 16, (const uint8_t *)src + 10 * 16); - rte_mov16((uint8_t *)dst + 11 * 16, (const uint8_t *)src + 11 * 16); - rte_mov16((uint8_t *)dst + 12 * 16, (const uint8_t *)src + 12 * 16); - rte_mov16((uint8_t *)dst + 13 * 16, (const uint8_t *)src + 13 * 16); - rte_mov16((uint8_t *)dst + 14 * 16, (const uint8_t *)src + 14 * 16); - rte_mov16((uint8_t *)dst + 15 * 16, (const uint8_t *)src + 15 * 16); + rte_mov16(dst + 0 * 16, src + 0 * 16); + rte_mov16(dst + 1 * 16, src + 1 * 16); + rte_mov16(dst + 2 * 16, src + 2 * 16); + rte_mov16(dst + 3 * 16, src + 3 * 16); + rte_mov16(dst + 4 * 16, src + 4 * 16); + rte_mov16(dst + 5 * 16, src + 5 * 16); + rte_mov16(dst + 6 * 16, src + 6 * 16); + rte_mov16(dst + 7 * 16, src + 7 * 16); + rte_mov16(dst + 8 * 16, src + 8 * 16); + rte_mov16(dst + 9 * 16, src + 9 * 16); + rte_mov16(dst + 10 * 16, src + 10 * 16); + rte_mov16(dst + 11 * 16, src + 11 * 16); + rte_mov16(dst + 12 * 16, src + 12 * 16); + rte_mov16(dst + 13 * 16, src + 13 * 16); + rte_mov16(dst + 14 * 16, src + 14 * 16); + rte_mov16(dst + 15 * 16, src + 15 * 16); } /** @@ -411,48 +419,48 @@ rte_mov256(uint8_t *dst, const uint8_t *src) ({ \ int tmp; \ while (len >= 128 + 16 - offset) { \ - xmm0 = _mm_loadu_si128((const __m128i *)((const uint8_t *)src - offset + 0 * 16)); \ - len -= 128; \ - xmm1 = _mm_loadu_si128((const __m128i *)((const uint8_t *)src - offset + 1 * 16)); \ - xmm2 = _mm_loadu_si128((const __m128i *)((const uint8_t *)src - offset + 2 * 16)); \ - xmm3 = _mm_loadu_si128((const __m128i *)((const uint8_t *)src - offset + 3 * 16)); \ - xmm4 = _mm_loadu_si128((const __m128i *)((const uint8_t *)src - offset + 4 * 16)); \ - xmm5 = _mm_loadu_si128((const __m128i *)((const uint8_t *)src - offset + 5 * 16)); \ - xmm6 = _mm_loadu_si128((const __m128i *)((const uint8_t *)src - offset + 6 * 16)); \ - xmm7 = _mm_loadu_si128((const __m128i *)((const uint8_t *)src - offset + 7 * 16)); \ - xmm8 = _mm_loadu_si128((const __m128i *)((const uint8_t *)src - offset + 8 * 16)); \ + xmm0 = _mm_loadu_si128((const __m128i *)(src - offset + 0 * 16)); \ + len -= 128; \ + xmm1 = _mm_loadu_si128((const __m128i *)(src - offset + 1 * 16)); \ + xmm2 = _mm_loadu_si128((const __m128i *)(src - offset + 2 * 16)); \ + xmm3 = _mm_loadu_si128((const __m128i *)(src - offset + 3 * 16)); \ + xmm4 = _mm_loadu_si128((const __m128i *)(src - offset + 4 * 16)); \ + xmm5 = _mm_loadu_si128((const __m128i *)(src - offset + 5 * 16)); \ + xmm6 = _mm_loadu_si128((const __m128i *)(src - offset + 6 * 16)); \ + xmm7 = _mm_loadu_si128((const __m128i *)(src - offset + 7 * 16)); \ + xmm8 = _mm_loadu_si128((const __m128i *)(src - offset + 8 * 16)); \ src = (const uint8_t *)src + 128; \ - _mm_storeu_si128((__m128i *)((uint8_t *)dst + 0 * 16), _mm_alignr_epi8(xmm1, xmm0, offset)); \ - _mm_storeu_si128((__m128i *)((uint8_t *)dst + 1 * 16), _mm_alignr_epi8(xmm2, xmm1, offset)); \ - _mm_storeu_si128((__m128i *)((uint8_t *)dst + 2 * 16), _mm_alignr_epi8(xmm3, xmm2, offset)); \ - _mm_storeu_si128((__m128i *)((uint8_t *)dst + 3 * 16), _mm_alignr_epi8(xmm4, xmm3, offset)); \ - _mm_storeu_si128((__m128i *)((uint8_t *)dst + 4 * 16), _mm_alignr_epi8(xmm5, xmm4, offset)); \ - _mm_storeu_si128((__m128i *)((uint8_t *)dst + 5 * 16), _mm_alignr_epi8(xmm6, xmm5, offset)); \ - _mm_storeu_si128((__m128i *)((uint8_t *)dst + 6 * 16), _mm_alignr_epi8(xmm7, xmm6, offset)); \ - _mm_storeu_si128((__m128i *)((uint8_t *)dst + 7 * 16), _mm_alignr_epi8(xmm8, xmm7, offset)); \ + _mm_storeu_si128((__m128i *)(dst + 0 * 16), _mm_alignr_epi8(xmm1, xmm0, offset)); \ + _mm_storeu_si128((__m128i *)(dst + 1 * 16), _mm_alignr_epi8(xmm2, xmm1, offset)); \ + _mm_storeu_si128((__m128i *)(dst + 2 * 16), _mm_alignr_epi8(xmm3, xmm2, offset)); \ + _mm_storeu_si128((__m128i *)(dst + 3 * 16), _mm_alignr_epi8(xmm4, xmm3, offset)); \ + _mm_storeu_si128((__m128i *)(dst + 4 * 16), _mm_alignr_epi8(xmm5, xmm4, offset)); \ + _mm_storeu_si128((__m128i *)(dst + 5 * 16), _mm_alignr_epi8(xmm6, xmm5, offset)); \ + _mm_storeu_si128((__m128i *)(dst + 6 * 16), _mm_alignr_epi8(xmm7, xmm6, offset)); \ + _mm_storeu_si128((__m128i *)(dst + 7 * 16), _mm_alignr_epi8(xmm8, xmm7, offset)); \ dst = (uint8_t *)dst + 128; \ } \ tmp = len; \ len = ((len - 16 + offset) & 127) + 16 - offset; \ tmp -= len; \ - src = (const uint8_t *)src + tmp; \ - dst = (uint8_t *)dst + tmp; \ + src = src + tmp; \ + dst = dst + tmp; \ if (len >= 32 + 16 - offset) { \ while (len >= 32 + 16 - offset) { \ - xmm0 = _mm_loadu_si128((const __m128i *)((const uint8_t *)src - offset + 0 * 16)); \ len -= 32; \ - xmm1 = _mm_loadu_si128((const __m128i *)((const uint8_t *)src - offset + 1 * 16)); \ - xmm2 = _mm_loadu_si128((const __m128i *)((const uint8_t *)src - offset + 2 * 16)); \ - src = (const uint8_t *)src + 32; \ - _mm_storeu_si128((__m128i *)((uint8_t *)dst + 0 * 16), _mm_alignr_epi8(xmm1, xmm0, offset)); \ - _mm_storeu_si128((__m128i *)((uint8_t *)dst + 1 * 16), _mm_alignr_epi8(xmm2, xmm1, offset)); \ - dst = (uint8_t *)dst + 32; \ + xmm0 = _mm_loadu_si128((const __m128i *)(src - offset + 0 * 16)); \ + xmm1 = _mm_loadu_si128((const __m128i *)(src - offset + 1 * 16)); \ + xmm2 = _mm_loadu_si128((const __m128i *)(src - offset + 2 * 16)); \ + src = src + 32; \ + _mm_storeu_si128((__m128i *)(dst + 0 * 16), _mm_alignr_epi8(xmm1, xmm0, offset)); \ + _mm_storeu_si128((__m128i *)(dst + 1 * 16), _mm_alignr_epi8(xmm2, xmm1, offset)); \ + dst = dst + 32; \ } \ tmp = len; \ len = ((len - 16 + offset) & 31) + 16 - offset; \ tmp -= len; \ - src = (const uint8_t *)src + tmp; \ - dst = (uint8_t *)dst + tmp; \ + src = src + tmp; \ + dst = dst + tmp; \ } \ }) @@ -491,12 +499,14 @@ rte_mov256(uint8_t *dst, const uint8_t *src) }) static inline void * -rte_memcpy(void *dst, const void *src, size_t n) +rte_memcpy(void *_dst, const void *_src, size_t n) { __m128i xmm0, xmm1, xmm2, xmm3, xmm4, xmm5, xmm6, xmm7, xmm8; - uintptr_t dstu = (uintptr_t)dst; - uintptr_t srcu = (uintptr_t)src; - void *ret = dst; + const uint8_t *src = (const uint8_t *)_src; + uint8_t *dst = (uint8_t *)_dst; + uintptr_t dstu = (uintptr_t)_dst; + uintptr_t srcu = (uintptr_t)_src; + void *ret = _dst; size_t dstofss; size_t srcofs; @@ -529,61 +539,61 @@ rte_memcpy(void *dst, const void *src, size_t n) * Fast way when copy size doesn't exceed 512 bytes */ if (n <= 32) { - rte_mov16((uint8_t *)dst, (const uint8_t *)src); - rte_mov16((uint8_t *)dst - 16 + n, (const uint8_t *)src - 16 + n); + rte_mov16(dst, src); + rte_mov16(dst - 16 + n, src - 16 + n); return ret; } if (n <= 48) { - rte_mov32((uint8_t *)dst, (const uint8_t *)src); - rte_mov16((uint8_t *)dst - 16 + n, (const uint8_t *)src - 16 + n); + rte_mov32(dst, src); + rte_mov16(dst - 16 + n, src - 16 + n); return ret; } if (n <= 64) { - rte_mov32((uint8_t *)dst, (const uint8_t *)src); - rte_mov16((uint8_t *)dst + 32, (const uint8_t *)src + 32); - rte_mov16((uint8_t *)dst - 16 + n, (const uint8_t *)src - 16 + n); + rte_mov32(dst, src); + rte_mov16(dst + 32, src + 32); + rte_mov16(dst - 16 + n, src - 16 + n); return ret; } - if (n <= 128) { + if (n <= 128) goto COPY_BLOCK_128_BACK15; - } + if (n <= 512) { if (n >= 256) { n -= 256; - rte_mov128((uint8_t *)dst, (const uint8_t *)src); - rte_mov128((uint8_t *)dst + 128, (const uint8_t *)src + 128); - src = (const uint8_t *)src + 256; - dst = (uint8_t *)dst + 256; + rte_mov128(dst, src); + rte_mov128(dst + 128, src + 128); + src = src + 256; + dst = dst + 256; } COPY_BLOCK_255_BACK15: if (n >= 128) { n -= 128; - rte_mov128((uint8_t *)dst, (const uint8_t *)src); - src = (const uint8_t *)src + 128; - dst = (uint8_t *)dst + 128; + rte_mov128(dst, src); + src = src + 128; + dst = dst + 128; } COPY_BLOCK_128_BACK15: if (n >= 64) { n -= 64; - rte_mov64((uint8_t *)dst, (const uint8_t *)src); - src = (const uint8_t *)src + 64; - dst = (uint8_t *)dst + 64; + rte_mov64(dst, src); + src = src + 64; + dst = dst + 64; } COPY_BLOCK_64_BACK15: if (n >= 32) { n -= 32; - rte_mov32((uint8_t *)dst, (const uint8_t *)src); - src = (const uint8_t *)src + 32; - dst = (uint8_t *)dst + 32; + rte_mov32(dst, src); + src = src + 32; + dst = dst + 32; } if (n > 16) { - rte_mov16((uint8_t *)dst, (const uint8_t *)src); - rte_mov16((uint8_t *)dst - 16 + n, (const uint8_t *)src - 16 + n); + rte_mov16(dst, src); + rte_mov16(dst - 16 + n, src - 16 + n); return ret; } - if (n > 0) { - rte_mov16((uint8_t *)dst - 16 + n, (const uint8_t *)src - 16 + n); - } + if (n > 0) + rte_mov16(dst - 16 + n, src - 16 + n); + return ret; } @@ -595,9 +605,9 @@ COPY_BLOCK_64_BACK15: */ dstofss = 16 - ((uintptr_t)dst & 0x0F) + 16; n -= dstofss; - rte_mov32((uint8_t *)dst, (const uint8_t *)src); - src = (const uint8_t *)src + dstofss; - dst = (uint8_t *)dst + dstofss; + rte_mov32(dst, src); + src = src + dstofss; + dst = dst + dstofss; srcofs = ((uintptr_t)src & 0x0F); /** @@ -608,9 +618,9 @@ COPY_BLOCK_64_BACK15: * Copy 256-byte blocks */ for (; n >= 256; n -= 256) { - rte_mov256((uint8_t *)dst, (const uint8_t *)src); - dst = (uint8_t *)dst + 256; - src = (const uint8_t *)src + 256; + rte_mov256(dst, src); + dst = dst + 256; + src = src + 256; } /** -- 1.9.1
Any inputs here? No functionality change just cleanup. I have run "make
test" and "memcpy_perf_autotest". I have not noticed any changes in numbers.
On Mon, Apr 20, 2015 at 1:33 PM, Ravi Kerur <rkerur@gmail.com> wrote:
> Remove unnecessary type casting in functions.
>
> Tested on Ubuntu (14.04 x86_64) with "make test".
> "make test" results match the results with baseline.
> "Memcpy perf" results match the results with baseline.
>
> Signed-off-by: Ravi Kerur <rkerur@gmail.com>
> ---
> .../common/include/arch/x86/rte_memcpy.h | 340
> +++++++++++----------
> 1 file changed, 175 insertions(+), 165 deletions(-)
>
> diff --git a/lib/librte_eal/common/include/arch/x86/rte_memcpy.h
> b/lib/librte_eal/common/include/arch/x86/rte_memcpy.h
> index 6a57426..839d4ec 100644
> --- a/lib/librte_eal/common/include/arch/x86/rte_memcpy.h
> +++ b/lib/librte_eal/common/include/arch/x86/rte_memcpy.h
> @@ -106,8 +106,8 @@ rte_mov32(uint8_t *dst, const uint8_t *src)
> static inline void
> rte_mov64(uint8_t *dst, const uint8_t *src)
> {
> - rte_mov32((uint8_t *)dst + 0 * 32, (const uint8_t *)src + 0 * 32);
> - rte_mov32((uint8_t *)dst + 1 * 32, (const uint8_t *)src + 1 * 32);
> + rte_mov32(dst + 0 * 32, src + 0 * 32);
> + rte_mov32(dst + 1 * 32, src + 1 * 32);
> }
>
> /**
> @@ -117,10 +117,10 @@ rte_mov64(uint8_t *dst, const uint8_t *src)
> static inline void
> rte_mov128(uint8_t *dst, const uint8_t *src)
> {
> - rte_mov32((uint8_t *)dst + 0 * 32, (const uint8_t *)src + 0 * 32);
> - rte_mov32((uint8_t *)dst + 1 * 32, (const uint8_t *)src + 1 * 32);
> - rte_mov32((uint8_t *)dst + 2 * 32, (const uint8_t *)src + 2 * 32);
> - rte_mov32((uint8_t *)dst + 3 * 32, (const uint8_t *)src + 3 * 32);
> + rte_mov32(dst + 0 * 32, src + 0 * 32);
> + rte_mov32(dst + 1 * 32, src + 1 * 32);
> + rte_mov32(dst + 2 * 32, src + 2 * 32);
> + rte_mov32(dst + 3 * 32, src + 3 * 32);
> }
>
> /**
> @@ -130,14 +130,14 @@ rte_mov128(uint8_t *dst, const uint8_t *src)
> static inline void
> rte_mov256(uint8_t *dst, const uint8_t *src)
> {
> - rte_mov32((uint8_t *)dst + 0 * 32, (const uint8_t *)src + 0 * 32);
> - rte_mov32((uint8_t *)dst + 1 * 32, (const uint8_t *)src + 1 * 32);
> - rte_mov32((uint8_t *)dst + 2 * 32, (const uint8_t *)src + 2 * 32);
> - rte_mov32((uint8_t *)dst + 3 * 32, (const uint8_t *)src + 3 * 32);
> - rte_mov32((uint8_t *)dst + 4 * 32, (const uint8_t *)src + 4 * 32);
> - rte_mov32((uint8_t *)dst + 5 * 32, (const uint8_t *)src + 5 * 32);
> - rte_mov32((uint8_t *)dst + 6 * 32, (const uint8_t *)src + 6 * 32);
> - rte_mov32((uint8_t *)dst + 7 * 32, (const uint8_t *)src + 7 * 32);
> + rte_mov32(dst + 0 * 32, src + 0 * 32);
> + rte_mov32(dst + 1 * 32, src + 1 * 32);
> + rte_mov32(dst + 2 * 32, src + 2 * 32);
> + rte_mov32(dst + 3 * 32, src + 3 * 32);
> + rte_mov32(dst + 4 * 32, src + 4 * 32);
> + rte_mov32(dst + 5 * 32, src + 5 * 32);
> + rte_mov32(dst + 6 * 32, src + 6 * 32);
> + rte_mov32(dst + 7 * 32, src + 7 * 32);
> }
>
> /**
> @@ -150,13 +150,16 @@ rte_mov64blocks(uint8_t *dst, const uint8_t *src,
> size_t n)
> __m256i ymm0, ymm1;
>
> while (n >= 64) {
> - ymm0 = _mm256_loadu_si256((const __m256i *)((const uint8_t
> *)src + 0 * 32));
> +
> + ymm0 = _mm256_loadu_si256((const __m256i *)(src + 0 * 32));
> + ymm1 = _mm256_loadu_si256((const __m256i *)(src + 1 * 32));
> +
> + _mm256_storeu_si256((__m256i *)(dst + 0 * 32), ymm0);
> + _mm256_storeu_si256((__m256i *)(dst + 1 * 32), ymm1);
> +
> n -= 64;
> - ymm1 = _mm256_loadu_si256((const __m256i *)((const uint8_t
> *)src + 1 * 32));
> - src = (const uint8_t *)src + 64;
> - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 0 * 32),
> ymm0);
> - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 1 * 32),
> ymm1);
> - dst = (uint8_t *)dst + 64;
> + src = src + 64;
> + dst = dst + 64;
> }
> }
>
> @@ -170,34 +173,39 @@ rte_mov256blocks(uint8_t *dst, const uint8_t *src,
> size_t n)
> __m256i ymm0, ymm1, ymm2, ymm3, ymm4, ymm5, ymm6, ymm7;
>
> while (n >= 256) {
> - ymm0 = _mm256_loadu_si256((const __m256i *)((const uint8_t
> *)src + 0 * 32));
> +
> + ymm0 = _mm256_loadu_si256((const __m256i *)(src + 0 * 32));
> + ymm1 = _mm256_loadu_si256((const __m256i *)(src + 1 * 32));
> + ymm2 = _mm256_loadu_si256((const __m256i *)(src + 2 * 32));
> + ymm3 = _mm256_loadu_si256((const __m256i *)(src + 3 * 32));
> + ymm4 = _mm256_loadu_si256((const __m256i *)(src + 4 * 32));
> + ymm5 = _mm256_loadu_si256((const __m256i *)(src + 5 * 32));
> + ymm6 = _mm256_loadu_si256((const __m256i *)(src + 6 * 32));
> + ymm7 = _mm256_loadu_si256((const __m256i *)(src + 7 * 32));
> +
> + _mm256_storeu_si256((__m256i *)(dst + 0 * 32), ymm0);
> + _mm256_storeu_si256((__m256i *)(dst + 1 * 32), ymm1);
> + _mm256_storeu_si256((__m256i *)(dst + 2 * 32), ymm2);
> + _mm256_storeu_si256((__m256i *)(dst + 3 * 32), ymm3);
> + _mm256_storeu_si256((__m256i *)(dst + 4 * 32), ymm4);
> + _mm256_storeu_si256((__m256i *)(dst + 5 * 32), ymm5);
> + _mm256_storeu_si256((__m256i *)(dst + 6 * 32), ymm6);
> + _mm256_storeu_si256((__m256i *)(dst + 7 * 32), ymm7);
> +
> n -= 256;
> - ymm1 = _mm256_loadu_si256((const __m256i *)((const uint8_t
> *)src + 1 * 32));
> - ymm2 = _mm256_loadu_si256((const __m256i *)((const uint8_t
> *)src + 2 * 32));
> - ymm3 = _mm256_loadu_si256((const __m256i *)((const uint8_t
> *)src + 3 * 32));
> - ymm4 = _mm256_loadu_si256((const __m256i *)((const uint8_t
> *)src + 4 * 32));
> - ymm5 = _mm256_loadu_si256((const __m256i *)((const uint8_t
> *)src + 5 * 32));
> - ymm6 = _mm256_loadu_si256((const __m256i *)((const uint8_t
> *)src + 6 * 32));
> - ymm7 = _mm256_loadu_si256((const __m256i *)((const uint8_t
> *)src + 7 * 32));
> - src = (const uint8_t *)src + 256;
> - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 0 * 32),
> ymm0);
> - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 1 * 32),
> ymm1);
> - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 2 * 32),
> ymm2);
> - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 3 * 32),
> ymm3);
> - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 4 * 32),
> ymm4);
> - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 5 * 32),
> ymm5);
> - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 6 * 32),
> ymm6);
> - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 7 * 32),
> ymm7);
> - dst = (uint8_t *)dst + 256;
> + src = src + 256;
> + dst = dst + 256;
> }
> }
>
> static inline void *
> -rte_memcpy(void *dst, const void *src, size_t n)
> +rte_memcpy(void *_dst, const void *_src, size_t n)
> {
> - uintptr_t dstu = (uintptr_t)dst;
> - uintptr_t srcu = (uintptr_t)src;
> - void *ret = dst;
> + const uint8_t *src = (const uint8_t *)_src;
> + uint8_t *dst = (uint8_t *)_dst;
> + uintptr_t dstu = (uintptr_t)_dst;
> + uintptr_t srcu = (uintptr_t)_src;
> + void *ret = _dst;
> size_t dstofss;
> size_t bits;
>
> @@ -230,43 +238,43 @@ rte_memcpy(void *dst, const void *src, size_t n)
> * Fast way when copy size doesn't exceed 512 bytes
> */
> if (n <= 32) {
> - rte_mov16((uint8_t *)dst, (const uint8_t *)src);
> - rte_mov16((uint8_t *)dst - 16 + n, (const uint8_t *)src -
> 16 + n);
> + rte_mov16(dst, src);
> + rte_mov16(dst - 16 + n, src - 16 + n);
> return ret;
> }
> if (n <= 64) {
> - rte_mov32((uint8_t *)dst, (const uint8_t *)src);
> - rte_mov32((uint8_t *)dst - 32 + n, (const uint8_t *)src -
> 32 + n);
> + rte_mov32(dst, src);
> + rte_mov32(dst - 32 + n, src - 32 + n);
> return ret;
> }
> if (n <= 512) {
> if (n >= 256) {
> n -= 256;
> - rte_mov256((uint8_t *)dst, (const uint8_t *)src);
> - src = (const uint8_t *)src + 256;
> - dst = (uint8_t *)dst + 256;
> + rte_mov256(dst, src);
> + src = src + 256;
> + dst = dst + 256;
> }
> if (n >= 128) {
> n -= 128;
> - rte_mov128((uint8_t *)dst, (const uint8_t *)src);
> - src = (const uint8_t *)src + 128;
> - dst = (uint8_t *)dst + 128;
> + rte_mov128(dst, src);
> + src = src + 128;
> + dst = dst + 128;
> }
> if (n >= 64) {
> n -= 64;
> - rte_mov64((uint8_t *)dst, (const uint8_t *)src);
> - src = (const uint8_t *)src + 64;
> - dst = (uint8_t *)dst + 64;
> + rte_mov64(dst, src);
> + src = src + 64;
> + dst = dst + 64;
> }
> COPY_BLOCK_64_BACK31:
> if (n > 32) {
> - rte_mov32((uint8_t *)dst, (const uint8_t *)src);
> - rte_mov32((uint8_t *)dst - 32 + n, (const uint8_t
> *)src - 32 + n);
> + rte_mov32(dst, src);
> + rte_mov32(dst - 32 + n, src - 32 + n);
> return ret;
> }
> - if (n > 0) {
> - rte_mov32((uint8_t *)dst - 32 + n, (const uint8_t
> *)src - 32 + n);
> - }
> + if (n > 0)
> + rte_mov32(dst - 32 + n, src - 32 + n);
> +
> return ret;
> }
>
> @@ -275,21 +283,21 @@ COPY_BLOCK_64_BACK31:
> */
> dstofss = 32 - ((uintptr_t)dst & 0x1F);
> n -= dstofss;
> - rte_mov32((uint8_t *)dst, (const uint8_t *)src);
> - src = (const uint8_t *)src + dstofss;
> - dst = (uint8_t *)dst + dstofss;
> + rte_mov32(dst, src);
> + src = src + dstofss;
> + dst = dst + dstofss;
>
> /**
> * Copy 256-byte blocks.
> * Use copy block function for better instruction order control,
> * which is important when load is unaligned.
> */
> - rte_mov256blocks((uint8_t *)dst, (const uint8_t *)src, n);
> + rte_mov256blocks(dst, src, n);
> bits = n;
> n = n & 255;
> bits -= n;
> - src = (const uint8_t *)src + bits;
> - dst = (uint8_t *)dst + bits;
> + src = src + bits;
> + dst = dst + bits;
>
> /**
> * Copy 64-byte blocks.
> @@ -297,12 +305,12 @@ COPY_BLOCK_64_BACK31:
> * which is important when load is unaligned.
> */
> if (n >= 64) {
> - rte_mov64blocks((uint8_t *)dst, (const uint8_t *)src, n);
> + rte_mov64blocks(dst, src, n);
> bits = n;
> n = n & 63;
> bits -= n;
> - src = (const uint8_t *)src + bits;
> - dst = (uint8_t *)dst + bits;
> + src = src + bits;
> + dst = dst + bits;
> }
>
> /**
> @@ -337,8 +345,8 @@ rte_mov16(uint8_t *dst, const uint8_t *src)
> static inline void
> rte_mov32(uint8_t *dst, const uint8_t *src)
> {
> - rte_mov16((uint8_t *)dst + 0 * 16, (const uint8_t *)src + 0 * 16);
> - rte_mov16((uint8_t *)dst + 1 * 16, (const uint8_t *)src + 1 * 16);
> + rte_mov16(dst + 0 * 16, src + 0 * 16);
> + rte_mov16(dst + 1 * 16, src + 1 * 16);
> }
>
> /**
> @@ -348,10 +356,10 @@ rte_mov32(uint8_t *dst, const uint8_t *src)
> static inline void
> rte_mov64(uint8_t *dst, const uint8_t *src)
> {
> - rte_mov16((uint8_t *)dst + 0 * 16, (const uint8_t *)src + 0 * 16);
> - rte_mov16((uint8_t *)dst + 1 * 16, (const uint8_t *)src + 1 * 16);
> - rte_mov16((uint8_t *)dst + 2 * 16, (const uint8_t *)src + 2 * 16);
> - rte_mov16((uint8_t *)dst + 3 * 16, (const uint8_t *)src + 3 * 16);
> + rte_mov16(dst + 0 * 16, src + 0 * 16);
> + rte_mov16(dst + 1 * 16, src + 1 * 16);
> + rte_mov16(dst + 2 * 16, src + 2 * 16);
> + rte_mov16(dst + 3 * 16, src + 3 * 16);
> }
>
> /**
> @@ -361,14 +369,14 @@ rte_mov64(uint8_t *dst, const uint8_t *src)
> static inline void
> rte_mov128(uint8_t *dst, const uint8_t *src)
> {
> - rte_mov16((uint8_t *)dst + 0 * 16, (const uint8_t *)src + 0 * 16);
> - rte_mov16((uint8_t *)dst + 1 * 16, (const uint8_t *)src + 1 * 16);
> - rte_mov16((uint8_t *)dst + 2 * 16, (const uint8_t *)src + 2 * 16);
> - rte_mov16((uint8_t *)dst + 3 * 16, (const uint8_t *)src + 3 * 16);
> - rte_mov16((uint8_t *)dst + 4 * 16, (const uint8_t *)src + 4 * 16);
> - rte_mov16((uint8_t *)dst + 5 * 16, (const uint8_t *)src + 5 * 16);
> - rte_mov16((uint8_t *)dst + 6 * 16, (const uint8_t *)src + 6 * 16);
> - rte_mov16((uint8_t *)dst + 7 * 16, (const uint8_t *)src + 7 * 16);
> + rte_mov16(dst + 0 * 16, src + 0 * 16);
> + rte_mov16(dst + 1 * 16, src + 1 * 16);
> + rte_mov16(dst + 2 * 16, src + 2 * 16);
> + rte_mov16(dst + 3 * 16, src + 3 * 16);
> + rte_mov16(dst + 4 * 16, src + 4 * 16);
> + rte_mov16(dst + 5 * 16, src + 5 * 16);
> + rte_mov16(dst + 6 * 16, src + 6 * 16);
> + rte_mov16(dst + 7 * 16, src + 7 * 16);
> }
>
> /**
> @@ -378,22 +386,22 @@ rte_mov128(uint8_t *dst, const uint8_t *src)
> static inline void
> rte_mov256(uint8_t *dst, const uint8_t *src)
> {
> - rte_mov16((uint8_t *)dst + 0 * 16, (const uint8_t *)src + 0 * 16);
> - rte_mov16((uint8_t *)dst + 1 * 16, (const uint8_t *)src + 1 * 16);
> - rte_mov16((uint8_t *)dst + 2 * 16, (const uint8_t *)src + 2 * 16);
> - rte_mov16((uint8_t *)dst + 3 * 16, (const uint8_t *)src + 3 * 16);
> - rte_mov16((uint8_t *)dst + 4 * 16, (const uint8_t *)src + 4 * 16);
> - rte_mov16((uint8_t *)dst + 5 * 16, (const uint8_t *)src + 5 * 16);
> - rte_mov16((uint8_t *)dst + 6 * 16, (const uint8_t *)src + 6 * 16);
> - rte_mov16((uint8_t *)dst + 7 * 16, (const uint8_t *)src + 7 * 16);
> - rte_mov16((uint8_t *)dst + 8 * 16, (const uint8_t *)src + 8 * 16);
> - rte_mov16((uint8_t *)dst + 9 * 16, (const uint8_t *)src + 9 * 16);
> - rte_mov16((uint8_t *)dst + 10 * 16, (const uint8_t *)src + 10 *
> 16);
> - rte_mov16((uint8_t *)dst + 11 * 16, (const uint8_t *)src + 11 *
> 16);
> - rte_mov16((uint8_t *)dst + 12 * 16, (const uint8_t *)src + 12 *
> 16);
> - rte_mov16((uint8_t *)dst + 13 * 16, (const uint8_t *)src + 13 *
> 16);
> - rte_mov16((uint8_t *)dst + 14 * 16, (const uint8_t *)src + 14 *
> 16);
> - rte_mov16((uint8_t *)dst + 15 * 16, (const uint8_t *)src + 15 *
> 16);
> + rte_mov16(dst + 0 * 16, src + 0 * 16);
> + rte_mov16(dst + 1 * 16, src + 1 * 16);
> + rte_mov16(dst + 2 * 16, src + 2 * 16);
> + rte_mov16(dst + 3 * 16, src + 3 * 16);
> + rte_mov16(dst + 4 * 16, src + 4 * 16);
> + rte_mov16(dst + 5 * 16, src + 5 * 16);
> + rte_mov16(dst + 6 * 16, src + 6 * 16);
> + rte_mov16(dst + 7 * 16, src + 7 * 16);
> + rte_mov16(dst + 8 * 16, src + 8 * 16);
> + rte_mov16(dst + 9 * 16, src + 9 * 16);
> + rte_mov16(dst + 10 * 16, src + 10 * 16);
> + rte_mov16(dst + 11 * 16, src + 11 * 16);
> + rte_mov16(dst + 12 * 16, src + 12 * 16);
> + rte_mov16(dst + 13 * 16, src + 13 * 16);
> + rte_mov16(dst + 14 * 16, src + 14 * 16);
> + rte_mov16(dst + 15 * 16, src + 15 * 16);
> }
>
> /**
> @@ -411,48 +419,48 @@ rte_mov256(uint8_t *dst, const uint8_t *src)
> ({
> \
> int tmp;
> \
> while (len >= 128 + 16 - offset) {
> \
> - xmm0 = _mm_loadu_si128((const __m128i *)((const uint8_t *)src -
> offset + 0 * 16)); \
> - len -= 128;
> \
> - xmm1 = _mm_loadu_si128((const __m128i *)((const uint8_t *)src -
> offset + 1 * 16)); \
> - xmm2 = _mm_loadu_si128((const __m128i *)((const uint8_t *)src -
> offset + 2 * 16)); \
> - xmm3 = _mm_loadu_si128((const __m128i *)((const uint8_t *)src -
> offset + 3 * 16)); \
> - xmm4 = _mm_loadu_si128((const __m128i *)((const uint8_t *)src -
> offset + 4 * 16)); \
> - xmm5 = _mm_loadu_si128((const __m128i *)((const uint8_t *)src -
> offset + 5 * 16)); \
> - xmm6 = _mm_loadu_si128((const __m128i *)((const uint8_t *)src -
> offset + 6 * 16)); \
> - xmm7 = _mm_loadu_si128((const __m128i *)((const uint8_t *)src -
> offset + 7 * 16)); \
> - xmm8 = _mm_loadu_si128((const __m128i *)((const uint8_t *)src -
> offset + 8 * 16)); \
> + xmm0 = _mm_loadu_si128((const __m128i *)(src - offset + 0 *
> 16)); \
> + len -= 128;
> \
> + xmm1 = _mm_loadu_si128((const __m128i *)(src - offset + 1 *
> 16)); \
> + xmm2 = _mm_loadu_si128((const __m128i *)(src - offset + 2 *
> 16)); \
> + xmm3 = _mm_loadu_si128((const __m128i *)(src - offset + 3 *
> 16)); \
> + xmm4 = _mm_loadu_si128((const __m128i *)(src - offset + 4 *
> 16)); \
> + xmm5 = _mm_loadu_si128((const __m128i *)(src - offset + 5 *
> 16)); \
> + xmm6 = _mm_loadu_si128((const __m128i *)(src - offset + 6 *
> 16)); \
> + xmm7 = _mm_loadu_si128((const __m128i *)(src - offset + 7 *
> 16)); \
> + xmm8 = _mm_loadu_si128((const __m128i *)(src - offset + 8 *
> 16)); \
> src = (const uint8_t *)src + 128;
> \
> - _mm_storeu_si128((__m128i *)((uint8_t *)dst + 0 * 16),
> _mm_alignr_epi8(xmm1, xmm0, offset)); \
> - _mm_storeu_si128((__m128i *)((uint8_t *)dst + 1 * 16),
> _mm_alignr_epi8(xmm2, xmm1, offset)); \
> - _mm_storeu_si128((__m128i *)((uint8_t *)dst + 2 * 16),
> _mm_alignr_epi8(xmm3, xmm2, offset)); \
> - _mm_storeu_si128((__m128i *)((uint8_t *)dst + 3 * 16),
> _mm_alignr_epi8(xmm4, xmm3, offset)); \
> - _mm_storeu_si128((__m128i *)((uint8_t *)dst + 4 * 16),
> _mm_alignr_epi8(xmm5, xmm4, offset)); \
> - _mm_storeu_si128((__m128i *)((uint8_t *)dst + 5 * 16),
> _mm_alignr_epi8(xmm6, xmm5, offset)); \
> - _mm_storeu_si128((__m128i *)((uint8_t *)dst + 6 * 16),
> _mm_alignr_epi8(xmm7, xmm6, offset)); \
> - _mm_storeu_si128((__m128i *)((uint8_t *)dst + 7 * 16),
> _mm_alignr_epi8(xmm8, xmm7, offset)); \
> + _mm_storeu_si128((__m128i *)(dst + 0 * 16), _mm_alignr_epi8(xmm1,
> xmm0, offset)); \
> + _mm_storeu_si128((__m128i *)(dst + 1 * 16), _mm_alignr_epi8(xmm2,
> xmm1, offset)); \
> + _mm_storeu_si128((__m128i *)(dst + 2 * 16), _mm_alignr_epi8(xmm3,
> xmm2, offset)); \
> + _mm_storeu_si128((__m128i *)(dst + 3 * 16), _mm_alignr_epi8(xmm4,
> xmm3, offset)); \
> + _mm_storeu_si128((__m128i *)(dst + 4 * 16), _mm_alignr_epi8(xmm5,
> xmm4, offset)); \
> + _mm_storeu_si128((__m128i *)(dst + 5 * 16), _mm_alignr_epi8(xmm6,
> xmm5, offset)); \
> + _mm_storeu_si128((__m128i *)(dst + 6 * 16), _mm_alignr_epi8(xmm7,
> xmm6, offset)); \
> + _mm_storeu_si128((__m128i *)(dst + 7 * 16), _mm_alignr_epi8(xmm8,
> xmm7, offset)); \
> dst = (uint8_t *)dst + 128;
> \
> }
> \
> tmp = len;
> \
> len = ((len - 16 + offset) & 127) + 16 - offset;
> \
> tmp -= len;
> \
> - src = (const uint8_t *)src + tmp;
> \
> - dst = (uint8_t *)dst + tmp;
> \
> + src = src + tmp;
> \
> + dst = dst + tmp;
> \
> if (len >= 32 + 16 - offset) {
> \
> while (len >= 32 + 16 - offset) {
> \
> - xmm0 = _mm_loadu_si128((const __m128i *)((const uint8_t *)src
> - offset + 0 * 16)); \
> len -= 32;
> \
> - xmm1 = _mm_loadu_si128((const __m128i *)((const uint8_t *)src
> - offset + 1 * 16)); \
> - xmm2 = _mm_loadu_si128((const __m128i *)((const uint8_t *)src
> - offset + 2 * 16)); \
> - src = (const uint8_t *)src + 32;
> \
> - _mm_storeu_si128((__m128i *)((uint8_t *)dst + 0 * 16),
> _mm_alignr_epi8(xmm1, xmm0, offset)); \
> - _mm_storeu_si128((__m128i *)((uint8_t *)dst + 1 * 16),
> _mm_alignr_epi8(xmm2, xmm1, offset)); \
> - dst = (uint8_t *)dst + 32;
> \
> + xmm0 = _mm_loadu_si128((const __m128i *)(src - offset + 0 *
> 16)); \
> + xmm1 = _mm_loadu_si128((const __m128i *)(src - offset + 1 *
> 16)); \
> + xmm2 = _mm_loadu_si128((const __m128i *)(src - offset + 2 *
> 16)); \
> + src = src + 32;
> \
> + _mm_storeu_si128((__m128i *)(dst + 0 * 16),
> _mm_alignr_epi8(xmm1, xmm0, offset)); \
> + _mm_storeu_si128((__m128i *)(dst + 1 * 16),
> _mm_alignr_epi8(xmm2, xmm1, offset)); \
> + dst = dst + 32;
> \
> }
> \
> tmp = len;
> \
> len = ((len - 16 + offset) & 31) + 16 - offset;
> \
> tmp -= len;
> \
> - src = (const uint8_t *)src + tmp;
> \
> - dst = (uint8_t *)dst + tmp;
> \
> + src = src + tmp;
> \
> + dst = dst + tmp;
> \
> }
> \
> })
>
> @@ -491,12 +499,14 @@ rte_mov256(uint8_t *dst, const uint8_t *src)
> })
>
> static inline void *
> -rte_memcpy(void *dst, const void *src, size_t n)
> +rte_memcpy(void *_dst, const void *_src, size_t n)
> {
> __m128i xmm0, xmm1, xmm2, xmm3, xmm4, xmm5, xmm6, xmm7, xmm8;
> - uintptr_t dstu = (uintptr_t)dst;
> - uintptr_t srcu = (uintptr_t)src;
> - void *ret = dst;
> + const uint8_t *src = (const uint8_t *)_src;
> + uint8_t *dst = (uint8_t *)_dst;
> + uintptr_t dstu = (uintptr_t)_dst;
> + uintptr_t srcu = (uintptr_t)_src;
> + void *ret = _dst;
> size_t dstofss;
> size_t srcofs;
>
> @@ -529,61 +539,61 @@ rte_memcpy(void *dst, const void *src, size_t n)
> * Fast way when copy size doesn't exceed 512 bytes
> */
> if (n <= 32) {
> - rte_mov16((uint8_t *)dst, (const uint8_t *)src);
> - rte_mov16((uint8_t *)dst - 16 + n, (const uint8_t *)src -
> 16 + n);
> + rte_mov16(dst, src);
> + rte_mov16(dst - 16 + n, src - 16 + n);
> return ret;
> }
> if (n <= 48) {
> - rte_mov32((uint8_t *)dst, (const uint8_t *)src);
> - rte_mov16((uint8_t *)dst - 16 + n, (const uint8_t *)src -
> 16 + n);
> + rte_mov32(dst, src);
> + rte_mov16(dst - 16 + n, src - 16 + n);
> return ret;
> }
> if (n <= 64) {
> - rte_mov32((uint8_t *)dst, (const uint8_t *)src);
> - rte_mov16((uint8_t *)dst + 32, (const uint8_t *)src + 32);
> - rte_mov16((uint8_t *)dst - 16 + n, (const uint8_t *)src -
> 16 + n);
> + rte_mov32(dst, src);
> + rte_mov16(dst + 32, src + 32);
> + rte_mov16(dst - 16 + n, src - 16 + n);
> return ret;
> }
> - if (n <= 128) {
> + if (n <= 128)
> goto COPY_BLOCK_128_BACK15;
> - }
> +
> if (n <= 512) {
> if (n >= 256) {
> n -= 256;
> - rte_mov128((uint8_t *)dst, (const uint8_t *)src);
> - rte_mov128((uint8_t *)dst + 128, (const uint8_t
> *)src + 128);
> - src = (const uint8_t *)src + 256;
> - dst = (uint8_t *)dst + 256;
> + rte_mov128(dst, src);
> + rte_mov128(dst + 128, src + 128);
> + src = src + 256;
> + dst = dst + 256;
> }
> COPY_BLOCK_255_BACK15:
> if (n >= 128) {
> n -= 128;
> - rte_mov128((uint8_t *)dst, (const uint8_t *)src);
> - src = (const uint8_t *)src + 128;
> - dst = (uint8_t *)dst + 128;
> + rte_mov128(dst, src);
> + src = src + 128;
> + dst = dst + 128;
> }
> COPY_BLOCK_128_BACK15:
> if (n >= 64) {
> n -= 64;
> - rte_mov64((uint8_t *)dst, (const uint8_t *)src);
> - src = (const uint8_t *)src + 64;
> - dst = (uint8_t *)dst + 64;
> + rte_mov64(dst, src);
> + src = src + 64;
> + dst = dst + 64;
> }
> COPY_BLOCK_64_BACK15:
> if (n >= 32) {
> n -= 32;
> - rte_mov32((uint8_t *)dst, (const uint8_t *)src);
> - src = (const uint8_t *)src + 32;
> - dst = (uint8_t *)dst + 32;
> + rte_mov32(dst, src);
> + src = src + 32;
> + dst = dst + 32;
> }
> if (n > 16) {
> - rte_mov16((uint8_t *)dst, (const uint8_t *)src);
> - rte_mov16((uint8_t *)dst - 16 + n, (const uint8_t
> *)src - 16 + n);
> + rte_mov16(dst, src);
> + rte_mov16(dst - 16 + n, src - 16 + n);
> return ret;
> }
> - if (n > 0) {
> - rte_mov16((uint8_t *)dst - 16 + n, (const uint8_t
> *)src - 16 + n);
> - }
> + if (n > 0)
> + rte_mov16(dst - 16 + n, src - 16 + n);
> +
> return ret;
> }
>
> @@ -595,9 +605,9 @@ COPY_BLOCK_64_BACK15:
> */
> dstofss = 16 - ((uintptr_t)dst & 0x0F) + 16;
> n -= dstofss;
> - rte_mov32((uint8_t *)dst, (const uint8_t *)src);
> - src = (const uint8_t *)src + dstofss;
> - dst = (uint8_t *)dst + dstofss;
> + rte_mov32(dst, src);
> + src = src + dstofss;
> + dst = dst + dstofss;
> srcofs = ((uintptr_t)src & 0x0F);
>
> /**
> @@ -608,9 +618,9 @@ COPY_BLOCK_64_BACK15:
> * Copy 256-byte blocks
> */
> for (; n >= 256; n -= 256) {
> - rte_mov256((uint8_t *)dst, (const uint8_t *)src);
> - dst = (uint8_t *)dst + 256;
> - src = (const uint8_t *)src + 256;
> + rte_mov256(dst, src);
> + dst = dst + 256;
> + src = src + 256;
> }
>
> /**
> --
> 1.9.1
>
>
On Mon, 20 Apr 2015 13:33:29 -0700
Ravi Kerur <rkerur@gmail.com> wrote:
> Remove unnecessary type casting in functions.
>
> Tested on Ubuntu (14.04 x86_64) with "make test".
> "make test" results match the results with baseline.
> "Memcpy perf" results match the results with baseline.
>
> Signed-off-by: Ravi Kerur <rkerur@gmail.com>
Getting rid of casts looks good.
My guess is no one reviewed it because no one is using rte_memcpy much..
Acked-by: Stephen Hemminger <stephen@networkplumber.org>
> Remove unnecessary type casting in functions. > > Tested on Ubuntu (14.04 x86_64) with "make test". > "make test" results match the results with baseline. > "Memcpy perf" results match the results with baseline. > > Signed-off-by: Ravi Kerur <rkerur@gmail.com> > Acked-by: Stephen Hemminger <stephen@networkplumber.org> > > --- > .../common/include/arch/x86/rte_memcpy.h | 340 +++++++++++---------- > 1 file changed, 175 insertions(+), 165 deletions(-) > > diff --git a/lib/librte_eal/common/include/arch/x86/rte_memcpy.h b/lib/librte_eal/common/include/arch/x86/rte_memcpy.h > index 6a57426..839d4ec 100644 > --- a/lib/librte_eal/common/include/arch/x86/rte_memcpy.h > +++ b/lib/librte_eal/common/include/arch/x86/rte_memcpy.h [...] > /** > @@ -150,13 +150,16 @@ rte_mov64blocks(uint8_t *dst, const uint8_t *src, size_t n) > __m256i ymm0, ymm1; > > while (n >= 64) { > - ymm0 = _mm256_loadu_si256((const __m256i *)((const uint8_t *)src + 0 * 32)); > + > + ymm0 = _mm256_loadu_si256((const __m256i *)(src + 0 * 32)); > + ymm1 = _mm256_loadu_si256((const __m256i *)(src + 1 * 32)); > + > + _mm256_storeu_si256((__m256i *)(dst + 0 * 32), ymm0); > + _mm256_storeu_si256((__m256i *)(dst + 1 * 32), ymm1); > + Any particular reason to change the order of the statements here? :) Overall this patch looks good. > n -= 64; > - ymm1 = _mm256_loadu_si256((const __m256i *)((const uint8_t *)src + 1 * 32)); > - src = (const uint8_t *)src + 64; > - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 0 * 32), ymm0); > - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 1 * 32), ymm1); > - dst = (uint8_t *)dst + 64; > + src = src + 64; > + dst = dst + 64; > } > } >
On Wed, Jan 27, 2016 at 8:18 PM, Zhihong Wang <zhihong.wang@intel.com> wrote: > > Remove unnecessary type casting in functions. > > > > Tested on Ubuntu (14.04 x86_64) with "make test". > > "make test" results match the results with baseline. > > "Memcpy perf" results match the results with baseline. > > > > Signed-off-by: Ravi Kerur <rkerur@gmail.com> > > Acked-by: Stephen Hemminger <stephen@networkplumber.org> > > > > --- > > .../common/include/arch/x86/rte_memcpy.h | 340 > +++++++++++---------- > > 1 file changed, 175 insertions(+), 165 deletions(-) > > > > diff --git a/lib/librte_eal/common/include/arch/x86/rte_memcpy.h > b/lib/librte_eal/common/include/arch/x86/rte_memcpy.h > > index 6a57426..839d4ec 100644 > > --- a/lib/librte_eal/common/include/arch/x86/rte_memcpy.h > > +++ b/lib/librte_eal/common/include/arch/x86/rte_memcpy.h > > [...] > > > /** > > @@ -150,13 +150,16 @@ rte_mov64blocks(uint8_t *dst, const uint8_t *src, > size_t n) > > __m256i ymm0, ymm1; > > > > while (n >= 64) { > > - ymm0 = _mm256_loadu_si256((const __m256i *)((const uint8_t > *)src + 0 * 32)); > > + > > + ymm0 = _mm256_loadu_si256((const __m256i *)(src + 0 * 32)); > > + ymm1 = _mm256_loadu_si256((const __m256i *)(src + 1 * 32)); > > + > > + _mm256_storeu_si256((__m256i *)(dst + 0 * 32), ymm0); > > + _mm256_storeu_si256((__m256i *)(dst + 1 * 32), ymm1); > > + > > Any particular reason to change the order of the statements here? :) > Overall this patch looks good. > Sorry for the late response. Let me double check and get back to you, it's been a while since I did the changes. > > n -= 64; > > - ymm1 = _mm256_loadu_si256((const __m256i *)((const uint8_t > *)src + 1 * 32)); > > - src = (const uint8_t *)src + 64; > > - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 0 * 32), > ymm0); > > - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 1 * 32), > ymm1); > > - dst = (uint8_t *)dst + 64; > > + src = src + 64; > > + dst = dst + 64; > > } > > } > > > >
On Wed, Jan 27, 2016 at 8:18 PM, Zhihong Wang <zhihong.wang@intel.com> wrote: > > Remove unnecessary type casting in functions. > > > > Tested on Ubuntu (14.04 x86_64) with "make test". > > "make test" results match the results with baseline. > > "Memcpy perf" results match the results with baseline. > > > > Signed-off-by: Ravi Kerur <rkerur@gmail.com> > > Acked-by: Stephen Hemminger <stephen@networkplumber.org> > > > > --- > > .../common/include/arch/x86/rte_memcpy.h | 340 > +++++++++++---------- > > 1 file changed, 175 insertions(+), 165 deletions(-) > > > > diff --git a/lib/librte_eal/common/include/arch/x86/rte_memcpy.h > b/lib/librte_eal/common/include/arch/x86/rte_memcpy.h > > index 6a57426..839d4ec 100644 > > --- a/lib/librte_eal/common/include/arch/x86/rte_memcpy.h > > +++ b/lib/librte_eal/common/include/arch/x86/rte_memcpy.h > > [...] > > > /** > > @@ -150,13 +150,16 @@ rte_mov64blocks(uint8_t *dst, const uint8_t *src, > size_t n) > > __m256i ymm0, ymm1; > > > > while (n >= 64) { > > - ymm0 = _mm256_loadu_si256((const __m256i *)((const uint8_t > *)src + 0 * 32)); > > + > > + ymm0 = _mm256_loadu_si256((const __m256i *)(src + 0 * 32)); > > + ymm1 = _mm256_loadu_si256((const __m256i *)(src + 1 * 32)); > > + > > + _mm256_storeu_si256((__m256i *)(dst + 0 * 32), ymm0); > > + _mm256_storeu_si256((__m256i *)(dst + 1 * 32), ymm1); > > + > > Any particular reason to change the order of the statements here? :) > Overall this patch looks good. > I checked the code changes, initial code had moving addresses (src and dst) and decrement counter scattered between store and load instructions. I changed it to loads, followed by stores and handle address/counters increment/decrement without changing functionality. > > > n -= 64; > > - ymm1 = _mm256_loadu_si256((const __m256i *)((const uint8_t > *)src + 1 * 32)); > > - src = (const uint8_t *)src + 64; > > - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 0 * 32), > ymm0); > > - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 1 * 32), > ymm1); > > - dst = (uint8_t *)dst + 64; > > + src = src + 64; > > + dst = dst + 64; > > } > > } > > > >
> -----Original Message----- > From: Ravi Kerur [mailto:rkerur@gmail.com] > Sent: Saturday, February 27, 2016 10:06 PM > To: Wang, Zhihong <zhihong.wang@intel.com> > Cc: dev@dpdk.org > Subject: Re: [dpdk-dev,v2] Clean up rte_memcpy.h file > > > > On Wed, Jan 27, 2016 at 8:18 PM, Zhihong Wang <zhihong.wang@intel.com> > wrote: > > Remove unnecessary type casting in functions. > > > > Tested on Ubuntu (14.04 x86_64) with "make test". > > "make test" results match the results with baseline. > > "Memcpy perf" results match the results with baseline. > > > > Signed-off-by: Ravi Kerur <rkerur@gmail.com> > > Acked-by: Stephen Hemminger <stephen@networkplumber.org> > > > > --- > > .../common/include/arch/x86/rte_memcpy.h | 340 +++++++++++--- > ------- > > 1 file changed, 175 insertions(+), 165 deletions(-) > > > > diff --git a/lib/librte_eal/common/include/arch/x86/rte_memcpy.h > b/lib/librte_eal/common/include/arch/x86/rte_memcpy.h > > index 6a57426..839d4ec 100644 > > --- a/lib/librte_eal/common/include/arch/x86/rte_memcpy.h > > +++ b/lib/librte_eal/common/include/arch/x86/rte_memcpy.h > > [...] > > > /** > > @@ -150,13 +150,16 @@ rte_mov64blocks(uint8_t *dst, const uint8_t *src, > size_t n) > > __m256i ymm0, ymm1; > > > > while (n >= 64) { > > - ymm0 = _mm256_loadu_si256((const __m256i *)((const uint8_t > *)src + 0 * 32)); > > + > > + ymm0 = _mm256_loadu_si256((const __m256i *)(src + 0 * 32)); > > + ymm1 = _mm256_loadu_si256((const __m256i *)(src + 1 * 32)); > > + > > + _mm256_storeu_si256((__m256i *)(dst + 0 * 32), ymm0); > > + _mm256_storeu_si256((__m256i *)(dst + 1 * 32), ymm1); > > + > > Any particular reason to change the order of the statements here? :) > Overall this patch looks good. > > I checked the code changes, initial code had moving addresses (src and dst) and > decrement counter scattered between store and load instructions. I changed it to > loads, followed by stores and handle address/counters increment/decrement > without changing functionality. > It's definitely okay to do this. Actually changing it or not won't affect the final output at all since gcc will optimize it while generating code. It's C code we're writing after all. But personally I prefer to keep the original order just as a comment that what's needed in the future should be calculated ASAP, and different kinds (CPU port) of instructions should be mixed together. :) Could you please rebase this patch since there has been some changes already? > > n -= 64; > > - ymm1 = _mm256_loadu_si256((const __m256i *)((const uint8_t > *)src + 1 * 32)); > > - src = (const uint8_t *)src + 64; > > - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 0 * 32), > ymm0); > > - _mm256_storeu_si256((__m256i *)((uint8_t *)dst + 1 * 32), > ymm1); > > - dst = (uint8_t *)dst + 64; > > + src = src + 64; > > + dst = dst + 64; > > } > > } > >