Inline spiller can decide to move a spill as early as possible in the basic block.
It will skip phis and label, but we also need to make sure it skips instructions
in the basic block prologue which restore exec mask.
Added isPositionLike callback in TargetInstrInfo to detect instructions which
shall be skipped in addition to common phis, labels etc.
Differential Revision: https://reviews.llvm.org/D27997
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@292554 91177308-0d34-0410-b5e6-96231b3b80d8
For -(x + y) -> (-x) + (-y), if x == -y, this would
change the result from -0.0 to 0.0. Since the fma/fmad
combine is an extension of this problem it also
applies there.
fmul should be fine, and I don't think any of the unary
operators or conversions should be a problem either.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@292473 91177308-0d34-0410-b5e6-96231b3b80d8
Limit register coalescer by not allowing it to artificially increase
size of registers beyond dword. Such super-registers are in fact
register sequences and not distinct HW registers.
With more super-regs we would need to allocate adjacent registers
and constraint regalloc more than needed. Moreover, our super
registers are overlapping. For instance we have VGPR0_VGPR1_VGPR2,
VGPR1_VGPR2_VGPR3, VGPR2_VGPR3_VGPR4 etc, which complicates registers
allocation even more, resulting in excessive spilling.
Differential Revision: https://reviews.llvm.org/D28782
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@292413 91177308-0d34-0410-b5e6-96231b3b80d8
This is the unsafe conversion pattern, but not guarded by
an unsafe math check. It is also already done in LegalizeDAG.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@292173 91177308-0d34-0410-b5e6-96231b3b80d8
With some minor manual fixes for using function_ref instead of
std::function. No functional change intended.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@291904 91177308-0d34-0410-b5e6-96231b3b80d8
This produces worse code when i16 is legal, mostly
due to combines getting confused by conversions inserted
for uniform 16-bit operations.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@291717 91177308-0d34-0410-b5e6-96231b3b80d8
This was shrinking the instruction even though the carry output
register was a virtual register, not known VCC.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@291716 91177308-0d34-0410-b5e6-96231b3b80d8
updated instructions:
pmulld, pmullw, pmulhw, mulsd, mulps, mulpd, divss, divps, divsd, divpd, addpd and subpd.
special optimization case which replaces pmulld with pmullw\pmulhw\pshuf seq.
In case if the real operands bitwidth <= 16.
Differential Revision: https://reviews.llvm.org/D28104
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@291657 91177308-0d34-0410-b5e6-96231b3b80d8
For i16 zeroext arguments when i16 was a legal type, the
known bits information from the truncate was lost. Insert
a zeroext so the known bits optimizations work with the 32-bit
loads.
Fixes code quality regressions vs. SI in min.ll test.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@291461 91177308-0d34-0410-b5e6-96231b3b80d8
Among other stuff, this allows to use predefined .option.machine_version_major
/minor/stepping symbols in the directive.
Relevant test expanded at once (also file renamed for clarity).
Differential Revision: https://reviews.llvm.org/D28140
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@290710 91177308-0d34-0410-b5e6-96231b3b80d8
The feature allows for conditional assembly, filling the entries
of .amd_kernel_code_t etc.
Symbols are defined with value 0 at the beginning of each kernel scope.
After each register usage, the respective symbol is set to:
value = max( value, ( register index + 1 ) )
Thus, at the end of scope the value represents a count of used registers.
Kernel scopes begin at .amdgpu_hsa_kernel directive, end at the
next .amdgpu_hsa_kernel (or EOF, whichever comes first). There is also
dummy scope that lies from the beginning of source file til the
first .amdgpu_hsa_kernel.
Test added.
Differential Revision: https://reviews.llvm.org/D27859
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@290608 91177308-0d34-0410-b5e6-96231b3b80d8
Canonicalize a select with a constant to the false side. This
enables more instruction shrinking opportunities since an
inline immediate can be used for the false side of v_cndmask_b32_e32.
This seems to usually be better but causes some code size regressions
in some tests.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@290372 91177308-0d34-0410-b5e6-96231b3b80d8
