https://gcc.gnu.org/g:b95f529df46e33c7ee7058b7cc51473e7a0d2f4c
commit r16-7565-gb95f529df46e33c7ee7058b7cc51473e7a0d2f4c Author: Jakub Jelinek <[email protected]> Date: Wed Feb 18 10:28:34 2026 +0100 i386: Fix up avx_vpermilp_parallel [PR124138] r16-3677 extended avx_vpermilp_parallel so that it handles also V8DImode (among others) and broke this testcase. For V8DFmode the bug existed there already since r0-127405 which extended avx_vpermilp_parallel to handle V8DFmode. For V16SImode/V16SFmode I think the code is correct, in that case it can for V4SImode/V4SFmode choose any permutation it wants and the V8S{I,F}mode and V16S{I,F}mode cases before FALLTHRU verify the upper lanes match those lower ones. For V[48]D[IF]mode the function uses different checking, where it checks each pair of ipar separately: case E_V8DFmode: case E_V8DImode: /* In the 512-bit DFmode case, we can only move elements within a 128-bit lane. First fill the second part of the mask, then fallthru. */ for (i = 4; i < 6; ++i) { if (ipar[i] < 4 || ipar[i] >= 6) return 0; mask |= (ipar[i] - 4) << i; } for (i = 6; i < 8; ++i) { if (ipar[i] < 6) return 0; mask |= (ipar[i] - 6) << i; } /* FALLTHRU */ case E_V4DFmode: case E_V4DImode: /* In the 256-bit DFmode case, we can only move elements within a 128-bit lane. */ for (i = 0; i < 2; ++i) { if (ipar[i] >= 2) return 0; mask |= ipar[i] << i; } for (i = 2; i < 4; ++i) { if (ipar[i] < 2) return 0; mask |= (ipar[i] - 2) << i; } break; so that it directly computes corresponding mask bit. Earlier in the function it verified ipar[i] wasn't above nelts (nor negative), so just checking if (ipar[i] >= 2) is ok, it can't be negative, and just checking if (ipar[i] < 6) is also fine, because it can never be 8 or above (because nelts is 8 in that case). Though, the if (ipar[i] < 2) case check used to be correct only when nelts could be only 4, when it can be 8 too, it will as in the following testcase happily accept values in the [4, 7] range which shouldn't be accepted (it can only handle [2, 3]). Fixed thusly. 2026-02-17 Jakub Jelinek <[email protected]> PR target/124138 * config/i386/i386.cc (avx_vpermilp_parallel): Verify ipar[2] and ipar[3] aren't larger than 3. * gcc.dg/pr124138.c: New test. * gcc.target/i386/avx512f-pr124138-1.c: New test. * gcc.target/i386/avx512f-pr124138-2.c: New test. Diff: --- gcc/config/i386/i386.cc | 2 +- gcc/testsuite/gcc.dg/pr124138.c | 20 +++++++++++++++++ gcc/testsuite/gcc.target/i386/avx512f-pr124138-1.c | 24 +++++++++++++++++++++ gcc/testsuite/gcc.target/i386/avx512f-pr124138-2.c | 25 ++++++++++++++++++++++ 4 files changed, 70 insertions(+), 1 deletion(-) diff --git a/gcc/config/i386/i386.cc b/gcc/config/i386/i386.cc index 42a73d65fb1d..407d652fa998 100644 --- a/gcc/config/i386/i386.cc +++ b/gcc/config/i386/i386.cc @@ -20701,7 +20701,7 @@ avx_vpermilp_parallel (rtx par, machine_mode mode) } for (i = 2; i < 4; ++i) { - if (ipar[i] < 2) + if (ipar[i] < 2 || ipar[i] >= 4) return 0; mask |= (ipar[i] - 2) << i; } diff --git a/gcc/testsuite/gcc.dg/pr124138.c b/gcc/testsuite/gcc.dg/pr124138.c new file mode 100644 index 000000000000..9dc78dca6c6f --- /dev/null +++ b/gcc/testsuite/gcc.dg/pr124138.c @@ -0,0 +1,20 @@ +/* PR target/124138 */ +/* { dg-do run } */ +/* { dg-options "-Wno-psabi" } */ + +typedef __attribute__((__vector_size__ (8 * sizeof (unsigned long long)))) unsigned long long V; + +[[gnu::noipa]] V +foo (V x) +{ + return __builtin_shuffle (x, (V) { 0, 1, 2, 5, 4, 5, 6, 7 }); +} + +int +main () +{ + V x = foo ((V) { 1, 2, 3, 4, 5, 6, 7, 8 }); + if (x[0] != 1 || x[1] != 2 || x[2] != 3 || x[3] != 6 + || x[4] != 5 || x[5] != 6 || x[6] != 7 || x[7] != 8) + __builtin_abort (); +} diff --git a/gcc/testsuite/gcc.target/i386/avx512f-pr124138-1.c b/gcc/testsuite/gcc.target/i386/avx512f-pr124138-1.c new file mode 100644 index 000000000000..79efd2bb221d --- /dev/null +++ b/gcc/testsuite/gcc.target/i386/avx512f-pr124138-1.c @@ -0,0 +1,24 @@ +/* PR target/124138 */ +/* { dg-do run } */ +/* { dg-options "-mavx512f" } */ +/* { dg-require-effective-target avx512f } */ + +#define AVX512F +#include "avx512f-helper.h" + +typedef __attribute__((__vector_size__ (64))) unsigned long long V; + +[[gnu::noipa]] V +foo (V x) +{ + return __builtin_shuffle (x, (V) { 0, 1, 2, 5, 4, 5, 6, 7 }); +} + +void +TEST () +{ + V x = foo ((V) { 1, 2, 3, 4, 5, 6, 7, 8 }); + if (x[0] != 1 || x[1] != 2 || x[2] != 3 || x[3] != 6 + || x[4] != 5 || x[5] != 6 || x[6] != 7 || x[7] != 8) + __builtin_abort (); +} diff --git a/gcc/testsuite/gcc.target/i386/avx512f-pr124138-2.c b/gcc/testsuite/gcc.target/i386/avx512f-pr124138-2.c new file mode 100644 index 000000000000..893dbd6c4776 --- /dev/null +++ b/gcc/testsuite/gcc.target/i386/avx512f-pr124138-2.c @@ -0,0 +1,25 @@ +/* PR target/124138 */ +/* { dg-do run } */ +/* { dg-options "-mavx512f" } */ +/* { dg-require-effective-target avx512f } */ + +#define AVX512F +#include "avx512f-helper.h" + +typedef __attribute__((__vector_size__ (64))) unsigned long long V; +typedef __attribute__((__vector_size__ (64))) double W; + +[[gnu::noipa]] W +foo (W x) +{ + return __builtin_shuffle (x, (V) { 0, 1, 2, 5, 4, 5, 6, 7 }); +} + +void +TEST () +{ + W x = foo ((W) { 1.5, 2.5, 3.5, 4.5, 5.5, 6.5, 7.5, 8.5 }); + if (x[0] != 1.5 || x[1] != 2.5 || x[2] != 3.5 || x[3] != 6.5 + || x[4] != 5.5 || x[5] != 6.5 || x[6] != 7.5 || x[7] != 8.5) + __builtin_abort (); +}
