When true16 is enabled, isel start to emit sgpr_lo16 register when a
trunc/sext i16/i32 is generated, or a salu32 is used by vgpr16 or vice
versa. And this causes a problem as sgpr_lo16 is not fully supported in
the pipeline.
True16 mode works fine in -O3 mode since folding pass remove sgpr_lo16
from the pipeline. However it hit a problem in -O0 mode as folding pass
is skipped.
This patch did:
1. stop emitting sgpr_lo16 from isel
2. update codegen pattern to split uniformed/divergent pattern for
i16/i32 conversion
3. update fix-sgpr-copy pass to address legalization requirement in
true16 mode, update fix-sgpr-copies-f16-true16.mir
test to include all possible combinations
This patch is tested with cts and downstream repo with -O0 testing
When targeting GFX950, disassembly of v_permlane16_swap_b32 and
v_permlane32_swap_b32 instructions produces errors when they use certain
vdst operand values, e.g., v_permlane16_swap_b32 v218, v219. This patch
fixes this problem.
i16 can be selected into sgpr32 or vgpr16 in isel lowering in true16
mode. And thus, it creates cases that we copy from vgpr16 to sgpr32 in
ext selection and this seems inevitable without sgpr16 support.
legalize the src/dst reg when we decide to lower this special copy to a
readfirstlane in fix-sgpr-copy pass and add a lit test
We haven't implemented 16 bit SGPRs. Currently allow 32-bit SGPRs to be
folded into True16 bit instructions taking 16 bit values. Also use
sgpr_32 when Imm is copied to spgr_lo16 so it could be further folded.
This improves generated code quality.
M0 can only be written to by the SALU, so `v_readfirstlane_b32 m0` is
effectively useless. Represent this by restricting the dest RC of that
instruction to `SReg_32_XM0` which excludes M0.
There is a lot of test changes due to the register class changing, but
most changes are trivial. In some cases, an extra register and
`s_mov_b32` is needed.
Fixes SWDEV-513269
This was a bit annoying because these introduce a new special case
encoding usage. op_sel is repurposed as a subset of dpp controls,
and is eligible for VOP3->VOP1 shrinking. For some reason fi also
uses an enum value, so we need to convert the raw boolean to 1 instead
of -1.
The 2 registers are swapped, so this has 2 defs. Ideally the builtin
would return a pair, but that's difficult so return a vector instead.
This would make a hypothetical builtin that supports v2f16 directly
uglier.
There are a lot of messes in the special case
predicate handling. Currently broad let blocks
override specific predicates with more general
cases. For instructions with SDWA, the HasSDWA
predicate was overriding the SubtargetPredicate
for the instruction.
This fixes enough to properly disallow new instructions
that support SDWA on older targets.
