Properly fix AVX-512 CRC calculation bug

The problem that led to the workaround in f83f14881c7 was not in fact
a compiler bug, but a failure to zero the upper bits of the vector
register containing the initial scalar CRC value. Fix that and revert
the workaround.

Diagnosed-by: Nathan Bossart <[email protected]>
Diagnosed-by: Raghuveer Devulapalli <[email protected]>
Tested-by: Andy Fan <[email protected]>
Tested-by: Soumyadeep Chakraborty <[email protected]>
Reviewed-by: Nathan Bossart <[email protected]>
Reviewed-by: Raghuveer Devulapalli <[email protected]>
Discussion: https://postgr.es/m/PH8PR11MB82866B07AA6758D12F699C00FB70A@PH8PR11MB8286.namprd11.prod.outlook.com

diff --git a/src/port/pg_crc32c_sse42.c b/src/port/pg_crc32c_sse42.c
index 9af3474a6ca95eab8f33432308d51ee18103c503..1a7172553551afdd07af956b5218b8d7d2de908c 100644 (file)
--- a/src/port/pg_crc32c_sse42.c
+++ b/src/port/pg_crc32c_sse42.c
@@ -123,7 +123,7 @@ pg_comp_crc32c_avx512(pg_crc32c crc, const void *data, size_t len)
        __m512i     k;
 
        k = _mm512_broadcast_i32x4(_mm_setr_epi32(0x740eef02, 0, 0x9e4addf8, 0));
-       x0 = _mm512_xor_si512(_mm512_castsi128_si512(_mm_cvtsi32_si128(crc0)), x0);
+       x0 = _mm512_xor_si512(_mm512_zextsi128_si512(_mm_cvtsi32_si128(crc0)), x0);
        buf += 64;
 
        /* Main loop. */

diff --git a/src/port/pg_crc32c_sse42_choose.c b/src/port/pg_crc32c_sse42_choose.c
index 802e47788c10c34f907880a9fe646ad6767b0199..74d2421ba2be99cd73883388469e3fad8e5e557b 100644 (file)
--- a/src/port/pg_crc32c_sse42_choose.c
+++ b/src/port/pg_crc32c_sse42_choose.c
@@ -95,9 +95,7 @@ pg_comp_crc32c_choose(pg_crc32c crc, const void *data, size_t len)
            __cpuidex(exx, 7, 0);
 #endif
 
-#if defined(__clang__) && !defined(__OPTIMIZE__)
-           /* Some versions of clang are broken at -O0 */
-#elif defined(USE_AVX512_CRC32C_WITH_RUNTIME_CHECK)
+#ifdef USE_AVX512_CRC32C_WITH_RUNTIME_CHECK
            if (exx[2] & (1 << 10) &&   /* VPCLMULQDQ */
                exx[1] & (1 << 31)) /* AVX512-VL */
                pg_comp_crc32c = pg_comp_crc32c_avx512;