[Bug target/104188] gcc omitting AVX-512 broadcast instruction

2022-01-22 Thread jakub at gcc dot gnu.org via Gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104188 Jakub Jelinek changed: What|Removed |Added Status|WAITING |NEW --- Comment #5 from Jakub Jelinek

[Bug target/104188] gcc omitting AVX-512 broadcast instruction

2022-01-22 Thread kvr000 at gmail dot com via Gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104188 --- Comment #4 from Zbynek Vyskovsky --- Sure, the code: #include #ifndef NO_VECTORIZE #ifdef __x86_64__ #include #include #endif #ifdef __aarch64__ #include #endif #endif typedef union Mat44 { float m[4][4]; #ifndef NO_VECTORIZE

[Bug target/104188] gcc omitting AVX-512 broadcast instruction

2022-01-22 Thread jakub at gcc dot gnu.org via Gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104188 Jakub Jelinek changed: What|Removed |Added CC||jakub at gcc dot gnu.org --- Comment #3

[Bug target/104188] gcc omitting AVX-512 broadcast instruction

2022-01-22 Thread kvr000 at gmail dot com via Gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104188 --- Comment #2 from Zbynek Vyskovsky --- > {1to16} says to broadcast from first element to all 16. The vbroadcastf32x4 is supposed to copy first four elmenents to 4-7, 8-11 and 12-15 . > Why do you think this is wrong code? It doesn't work.

[Bug target/104188] gcc omitting AVX-512 broadcast instruction

2022-01-22 Thread pinskia at gcc dot gnu.org via Gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104188 Andrew Pinski changed: What|Removed |Added Status|UNCONFIRMED |WAITING Last reconfirmed|