Rollup merge of #126617 - sayantn:veorq, r=workingjubilee

Expand `avx512_target_feature` to include VEX variants

Added 5 new target features for x86:

 - `AVX-IFMA`
 - `AVX-NE-CONVERT`
 - `AVX-VNNI`
 - `AVX-VNNI_INT8`
 - `AVX-VNNI_INT16`

Both LLVM and GCC already have support for these.

See also the [stdarch PR](rust-lang/stdarch#1586)

Author: matthiaskrgr