On Thu, 5 Nov 2026 00:21:18 GMT, Vladimir Ivanov <[email protected]> wrote:
>> Jatin Bhateja has updated the pull request incrementally with one additional >> commit since the last revision: >> >> Review comments resolution > > src/hotspot/cpu/x86/c2_MacroAssembler_x86.cpp line 1434: > >> 1432: andl(rtmp, 31); >> 1433: // The entry is the kind's first matrix. rtmp selects the count. >> 1434: lea(rbase, >> ExternalAddress(StubRoutines::x86::vector_byte_shift_gfni_matrix(kind))); > > It looks way too complicated on StubGenerator side. You could just represent > the offset as constant displacement: > > lea(rbase, > ExternalAddress(StubRoutines::x86::vector_byte_shift_gfni_matrix()); > > const int mats_per_kind = 32; > Address matrix(rbase, rtmp, Address::times_8, kind * mats_per_kind * > BytesPerLong); Thanks @iwanowww , addressed. Let me know if there are other comments. ------------- PR Review Comment: https://git.openjdk.org/jdk/pull/33113#discussion_r4476334967
