With architected flat scratch it becomes readonly. We must always
reserve SGPR pair for it even if we do not use scratch at all since
an attempt to write to SGPRs mapped to FLAT_SCRATCH results in
memory violation.
This is not needed since GFX10 with architected flat scratch though
since special SGPRs are not carving space from normal SGPRs.
Differential Revision: https://reviews.llvm.org/D110376
As suggested on D108522, if we're sign extending a v4i16 source before multiplying as a v4i32, then we can replace that with a zero extension and rely on the implicit sign-extension of PMADDWD.
This is motivated by the examples and discussion in:
https://llvm.org/PR51245
...and related bugs.
By using vector compares and vector logic, we can convert 2 'set'
instructions into 1 'movd' or 'movmsk' and generally improve
throughput/reduce instructions.
Unfortunately, we don't have a complete vector compare ISA before
AVX, so I left SSE-only out of this patch. Ie, we'd need extra logic
ops to simulate the missing predicates for SSE 'cmpp*', so it's not
as clearly a win.
Differential Revision: https://reviews.llvm.org/D110342
Add the tail policy argument to LLVM IR intrinsics. There are two policies for tail elements. Tail agnostic means users do not care about the values in the tail elements and tail undisturbed means the values in the tail elements need to be kept after the operation. In order to let users control the tail policy, we add an additional argument at the end of the argument list.
For unmasked operations, we have no maskedoff and the tail policy is always tail agnostic. If users want to keep tail elements under unmasked operations, they could use all one mask in the masked operations to do it. So, we only add the additional argument for masked operations for most cases. There are exceptions listed below.
In this patch, we do not handle the following cases to reduce the complexity of the patch. There could be two separate patches for them.
* Use dest argument to control tail policy
vmerge.vvm/vmerge.vxm/vmerge.vim (add _t builtins with additional dest argument)
vfmerge.vfm (add _t builtins with additional dest argument)
vmv.v.v (add _t builtins with additional dest argument)
vmv.v.x (add _t builtins with additional dest argument)
vmv.v.i (add _t builtins with additional dest argument)
vfmv.v.f (add _t builtins with additional dest argument)
vadc.vvm/vadc.vxm/vadc.vim (add _t builtins with additional dest argument)
vsbc.vvm/vsbc.vxm (add _t builtins with additional dest argument)
* Always has tail argument for masked/unmasked intrinsics
Vector Single-Width Integer Multiply-Add Instructions (add _t and _mt builtins)
Vector Widening Integer Multiply-Add Instructions (add _t and _mt builtins)
Vector Single-Width Floating-Point Fused Multiply-Add Instructions (add _t and _mt builtins)
Vector Widening Floating-Point Fused Multiply-Add Instructions (add _t and _mt builtins)
Vector Reduction Operations (add _t and _mt builtins)
Vector Slideup Instructions (add _t and _mt builtins)
Vector Slidedown Instructions (add _t and _mt builtins)
Discussion: https://github.com/riscv/rvv-intrinsic-doc/pull/101
Differential Revision: https://reviews.llvm.org/D105092
Testing on a SLM box suggests these can run on either port, but the throughput is 4cy on either (inc MMX versions). Confirmed with Intel AoM / Agner / InstLatX64.
The type legalizer has by default no method of doing this bitcast other than
storing and reloading the value from stack.
This patch implements a custom lowering of this operation using extractions
of subregs (z13 and earlier using FP128 register pairs), or of vector
elements (with 'vector enhancements 1' using VR128 FP registers).
Review: Ulrich Weigand
Differential Revision: https://reviews.llvm.org/D110346
It was assuming that GPR extends could only have destination sizes of 32 or 64
bits, but for AArch64 we allow < 32 bits even without matching size physregs.
We don't allow an initializer for LDS variables
and there is an early abort during instruction
selection. This patch legalizes them by ignoring
the init values. During assembly emission, proper
error reporting already exists for such instances.
Reviewed By: arsenm
Differential Revision: https://reviews.llvm.org/D109901
On targets that do not support AGPR to AGPR copying directly, try to find the
defining accvgpr_write and propagate its source vgpr register to the copies
before register allocation so the source vgpr register does not get clobbered.
The postrapseudos pass also attempt to propagate the defining accvgpr_write but
if the register to propagate is clobbered, it will give up and create new
temporary vgpr registers instead.
Reviewed By: rampitec
Differential Revision: https://reviews.llvm.org/D108830
Turn (and (shl x, c2), c1) -> (slli (srli x, c3-c2), c3) if c1 is a
shifted mask with no leading zeros and c3 trailing zeros where c3
is greater than c2.
We should use IMAGE_REL_I386_SECREL in the i386 section of this file.
IMAGE_REL_I386_SECREL and IMAGE_REL_AMD64_SECREL have the same
numeric value 0xB, so this doesn't change behavior.
Turn (and (shr x, c2), c1) -> (slli (srli x, c2+c3), c3) if c1 is a
shifted mask with c2 leading zeros and c3 trailing zeros.
When the leading zeros is C2+32 we can use SRLIW in place of SRLI.
This function can be adapted to solve bugs like PR51245,
but it could require differentiating the combiner timing
between the existing and new transforms.
The pass amdgpu-propagate-attributes ("Early/Late propagate attributes
from kernels to functions") is currently run also for shaders, where
it does nothing. Modify the check so the pass only processes functions
for kernels.
Differential Revision: https://reviews.llvm.org/D109961
Only the most recent cpus support really 1cy 64-bit multiplies, and the X64 cost table represents a realistic worst case. The 1cy value was also discouraging vectorization when most vXi64 PMULDQ expansions aren't actually slower than scalarization.
Noticed while investigating PR51436.
This patch adds codegen support for lowering the vector-predicated
reduction intrinsics to RVV instructions. The process is similar to that
of the other reduction intrinsics, save for the fact that every VP
reduction has a start value. We reuse the existing custom "VL" nodes,
adding extra patterns where required to handle non-true masks.
To support these nodes, the `RISCVISD::VECREDUCE_*_VL` nodes have been
given an explicit "merge" operand. This is to faciliate the VP
reductions, where we must be careful to ensure that even if no operation
is performed (when VL=0) we still produce the start value. The RVV
reductions don't update the destination register under these conditions,
so we tie the splatted start value to the output register.
Reviewed By: craig.topper
Differential Revision: https://reviews.llvm.org/D107657
This simplifies the API and addresses a FIXME in
TwoAddressInstructionPass::convertInstTo3Addr.
Differential Revision: https://reviews.llvm.org/D110229
This patch is to support transform something like
_mm512_add_ph(acc, _mm512_fmadd_pch(a, b, _mm512_setzero_ph()))
to _mm512_fmadd_pch(a, b, acc).
Differential Revision: https://reviews.llvm.org/D109953
We can use riscv_vse intrinsic instead of riscv_vse_mask. The code here
is based on similar code for handling masked.scatter and vp.scatter.
Reviewed By: frasercrmck
Differential Revision: https://reviews.llvm.org/D110206
Currenlty PseudoProbeInserter is a pass conditioned on a target switch. It works well with a single clang invocation. It doesn't work so well when the backend is called separately (i.e, through the linker or llc), where user has always to pass -pseudo-probe-for-profiling explictly. I'm making the pass a default pass that requires no command line arg to trigger, but will be actually run depending on whether the CU comes with `llvm.pseudo_probe_desc` metadata.
Reviewed By: wenlei
Differential Revision: https://reviews.llvm.org/D110209
Based off a discussion on D110100, we should be avoiding default CostKinds whenever possible.
This initial patch removes them from the 'inner' target implementation callbacks - these should only be used by the main TTI calls, so this should guarantee that we don't cause changes in CostKind by missing it in an inner call. This exposed a few missing arguments in getGEPCost and reduction cost calls that I've cleaned up.
Differential Revision: https://reviews.llvm.org/D110242
Like normal atomicrmw operations, at -O0 the simple register-allocator can
insert spills into the LL/SC loop if it's expanded and visible when regalloc
runs. This can cause the operation to never succeed by repeatedly clearing the
monitor. Instead expand to a cmpxchg, which has a pseudo-instruction for -O0.
This allows VMOVL in tail predicated loops so long as the the vector
size the VMOVL is extending into is less than or equal to the size of
the VCTP in the tail predicated loop. These cases represent a
sign-extend-inreg (or zero-extend-inreg), which needn't block tail
predication as in https://godbolt.org/z/hdTsEbx8Y.
For this a vecsize has been added to the TSFlag bits of MVE
instructions, which stores the size of the elements that the MVE
instruction operates on. In the case of multiple size (such as a
MVE_VMOVLs8bh that extends from i8 to i16, the largest size was be
chosen). The sizes are encoded as 00 = i8, 01 = i16, 10 = i32 and 11 =
i64, which often (but not always) comes from the instruction encoding
directly. A unit test was added, and although only a subset of the
vecsizes are currently used, the rest should be useful for other cases.
Differential Revision: https://reviews.llvm.org/D109706
Use of output modifiers forces VOP3 encoding for a VOP2 mac/fmac
instruction, so we might as well convert it to the more flexible VOP3-
only mad/fma form.
With this change, the only way we should emit VOP3-encoded mac/fmac is
if regalloc chooses registers that require the VOP3 encoding, e.g. sgprs
for both src0 and src1. In all other cases the mac/fmac should either be
converted to mad/fma or shrunk to VOP2 encoding.
Differential Revision: https://reviews.llvm.org/D110156
Add an overload to pass the flat workgroup range in separately. This
will allow the attributor to use the assumed value for
amdgpu-flat-workgroup-sizes when inferring amdgpu-waves-per-eu.
This is a follow-up of D105872. Now we are able to prepare for update
form with non-const increment.
Reviewed By: jsji
Differential Revision: https://reviews.llvm.org/D106032
And always print it.
This makes some LLVM diagnostics match up better with Clang's diagnostics.
Updated some AMDGPU uses of DiagnosticInfoResourceLimit and now we print
better diagnostics for those.
Reviewed By: dblaikie
Differential Revision: https://reviews.llvm.org/D110204
This change adds the ASan intrinsic to the list whihc are setting hasCopyImplyingStackAdjustment.
Reviewed By: eugenis
Differential Revision: https://reviews.llvm.org/D110012
This time with the right bug number.
When we rewrite the setcc we replace set old setcc output register
with the new CondReg. But since CondReg can be shared by other
replacements, we don't know if the kill flags for the old register
are valid for CondReg. So be conservative and remove them.
The test case has a SETCCr and a SETCCm on the same condition so
they end up sharing the same CondReg. The SETCCr had one use with
a kill flag. This kill flag isn't valid after the replacement because
CondReg needs a live range extending to the later SETCCm replacment.
Fixes PR51903.
When we rewrite the setcc we replace set old setcc output register
with the new CondReg. But since CondReg can be shared by other
replacements, we don't know if the kill flags for the old register
are valid for CondReg. So be conservative and remove them.
The test case has a SETCCr and a SETCCm on the same condition so
they end up sharing the same CondReg. The SETCCr had one use with
a kill flag. This kill flag isn't valid after the replacement because
CondReg needs a live range extending to the later SETCCm replacment.
Fixes PR51908.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D110046
Normally, given that the DA results are kept consistent over the selection DAG, uniform comparisons get selected to S_CMP_* but divergent to V_CMP_*. Sometimes, for the sake of efficiency, SSA subgraphs may be converted to VALU to avoid repeatedly copying data back and forth. Hence we have to be able to sustain the correctness passing the i1 from VALU to SALU context and vice versa.
VALU operations only process the active lanes of the VGPR and ignore inactive ones.
Active lanes correspond to 1 bit in the EXEC mask register.
SALU represents i1 as just one bit but VALU as 64bits: 0/1 and 0/(0xffffffffffffffff & EXEC) respectively.
SALU uses one-bit conditional flag SCC but VALU - VCC that is a pair of 32-bit SGPRs
To expose SCC to the VALU context we need to convert the one-bit boolean value to the appropriate 64bit.
To return back to the SALU context we need to do the opposite.
To correctly convert 64bit VALU boolean to either 0 or 1 we need to filter out the bits corresponding to the inactive lanes.
Reviewed By: piotr
Differential Revision: https://reviews.llvm.org/D109900
When adding alias.scope and noalias metadata to a memcpy function,
the alias.scope and noalias metadata from the operands are merged.
The rule for merging alias.scope is to take the intersection of
the domains and the union of the scopes within those domains.
The rule for merging noalias is to take the intersection.
The bug is that AMDGPULowerModuleLDS was using concatenation for
both alias.scope and noalias. For example, when f1 and f2 are added
to the LDS structure and there is a memcpy(f2, f1, sizeof(f1)).
Then, concatenation creates noalias metadata for the memcpy that
includes both {f1, f2}. That means that the memcpy is assumed
not to alias a prior load of f2, which enables the optimizer to
remove a load of f2 that occurs after mempcy.
The function MDNode::getmostGenericAliasScope defines the semantics
for alias.scope. There is a function, combineMetadata in Local.cpp,
that uses intersect for noalias.
Differential Revision: https://reviews.llvm.org/D110049
This patch adds a prefixed load pattern involving v2f32 fpext v2f64, where we
are dealing with a value with an offset that fits into a 34-bit signed immediate.
A reduced test case is also added to patch that tests the pattern, in which the
pattern is tested in the big endian CHECKs of the newly added test.
Differential Revision: https://reviews.llvm.org/D109887
This requires a minor change to CodeGenPrepare to ensure that
shouldSinkOperands will be called for And.
Reviewed By: frasercrmck
Differential Revision: https://reviews.llvm.org/D110106
Optimize (add (mul x, c0), c1) -> (ADDI (MUL (ADDI, c1/c0), c0), c1%c0),
if c1/c0 and c1%c0 are simm12, while c1 is not.
Optimize (add (mul x, c0), c1) -> (MUL (ADDI, c1/c0), c0),
if c1%c0 is zero, and c1/c0 is simm12 while c1 is not.
Reviewed By: craig.topper
Differential Revision: https://reviews.llvm.org/D108607
Stanislav Mekhanoshin