Commit Graph

5 Commits

Author SHA1 Message Date
Mirko Brkusanin a6a72dfdf2 [AMDGPU][GlobalISel] Avoid selecting S_PACK with constants
If constants are hidden behind G_ANYEXT we can treat them same way as G_SEXT.
For that purpose we extend getConstantVRegValWithLookThrough with option
to handle G_ANYEXT same way as G_SEXT.

Differential Revision: https://reviews.llvm.org/D92219
2021-01-20 11:54:53 +01:00
Matt Arsenault 5819159995 AMDGPU/GlobalISel: Pack constant G_BUILD_VECTOR_TRUNCs when selecting 2020-07-26 09:55:34 -04:00
Matt Arsenault 6a479220b5 AMDGPU/GlobalISel: Commit test changes I forgot to squash
These should have been in ac7abe0ba9
2020-02-21 11:43:39 -05:00
Matt Arsenault ac7abe0ba9 AMDGPU/GlobalISel: Manually select G_BUILD_VECTOR_TRUNC
We have patterns for s_pack* selection, but they assume the inputs are
a build_vector with 16-bit inputs, not a truncating build
vector. Since there's still outstanding work for how to handle
mismatched result and source element vector operations, and since I'm
trying a different packed vector strategy than SelectionDAG, just
manually select this for now.
2020-02-21 10:34:11 -05:00
Matt Arsenault dfa9420f09 AMDGPU/GlobalISel: Don't use legal v2s16 G_BUILD_VECTOR
If we have s_pack_* instructions, legalize this to
G_BUILD_VECTOR_TRUNC from s32 elements. This is closer to how how the
s_pack_* instructions really behave.

If we don't have s_pack_ instructions, expand this by creating a merge
to s32 and bitcasting. This expands to the expected bit operations. I
think this eventually should go in a new bitcast legalize action type
in LegalizerHelper.

We already directly emit the shift operations in RegBankSelect for the
vector case. This could possibly be cleaned up, but I also may want to
defer doing this expansion to selection anyway. I'll see about that
when I try to actually match VOP3P instructions.

This breaks the selection of the build_vector since tablegen doesn't
know how to match G_BUILD_VECTOR_TRUNC yet, so just xfail it for now.
2020-02-05 11:52:18 -05:00