This reverts commit 6ea5bf436a.
6ea5bf436a made use of new c++17 rules regarding
order of evaluation (specifically: in function calls the expression naming the
function should be sequenced before the evalution of any operands) to simplify
some continuation-passing calls. Unfortunately this appears to break at least
one MSVC bot: https://lab.llvm.org/buildbot/#/builders/123/builds/12149 .
Includes an update to the comments to note that the workaround is now based on
MSVC limitations, not on LLVM adopting c++17.
std::iterator has been deprecated in C++17 and some standard library implementations such as MS STL or libc++ emit deperecation messages when using the class.
Since LLVM has now switched to C++17 these will emit warnings on these implementations, or worse, errors in build configurations using -Werror.
This patch fixes these issues by replacing them with LLVMs own llvm::iterator_facade_base which offers a superset of functionality of std::iterator.
Differential Revision: https://reviews.llvm.org/D131320
This function heap-allocates a ThreadSafeModule (the current C bindings assume
that TSMs are always heap-allocated), but was failing to free it.
Should fix http://llvm.org/PR56953.
If we're adding a constant that can't use addi we try a few tricks,
one of which is using li+sh3add. We should not do this if lui+add
would work. For example adding 8192. Using sh3add prevents folding
a sext.w to form addw, thus increasing instruction count.
This fixes a bug reported privately by @craig.topper. Here's an example which illustrates the problem:
vsetivli a1, a0, e32, m1, ta, mu # both DefInfo and PrevInfo
vsetivli a2, a1, e32, m4, ta, mu
With the unsound result being:
vsetivli a1, a0, e32, m1, ta, mu
vsetivli a2, a0, e32, m4, ta, mu
Consider the case where this is running on a machine with VLEN=512,. For this case, the VLMAXs are 16 and 64 respectively.
Consider for a0 = 33. The correct result is: a1 = 16, and a2 = 16
After the unsound optimization: a1 = 16 and a2 = 33
This particular example used VLMAXs which differed by more than a power of two. With a difference of only one power of two, there's another form of this bug which involves the AVL < 2 x VLMAX special case, but that ones more complicated to construct as many examples turn out accidentally sound.
This patch takes the approach of simply removing the unsound optimization, but there are multiple sound sub-cases of it. I plan to return to at least a couple of them, but figured it was cleaner to remove the unsound optimization (for ease of backporting), and then review the new optimizations on their own.
Differential Revision: https://reviews.llvm.org/D131264
Create function segments and emit unwind info of them.
A segment must be less than 1MB and no prolog or epilog is splitted between two
segments.
This patch should generate correct, though not optimal, unwind info for large
functions. Currently it only generate pacted info (.pdata) only for functions
that are less than 1MB (single-segment functions). This is NFC from before this
patch.
The next step is to enable (.pdata) only unwind info for the first segment or
segments that have neither prolog or epilog in a multi-segment function.
Another future work item is to further split segments that require more than 255
code words or have more than 65535 epilogs.
Reference:
https://docs.microsoft.com/en-us/cpp/build/arm64-exception-handling#function-fragments
Differential Revision: https://reviews.llvm.org/D130049
NOTE: i8 vector splats are ignored because the immediate range of
DUP already has full coverage.
Differential Revision: https://reviews.llvm.org/D131078
We get a couple of improvements from recognizing swapped
operand patterns that were not handled by the replicated
code.
This should also enable simplifying larger patterns as
seen in issue #56653 and issue #56654, but that requires
enhancements to isImpliedCondition() itself.
There are no AMDGPUSampleVariant versions for _G16, it is treated more like a
modifier for derivatives (_D) (also for intrinsics where it is overloaded type
instead of part of instrinsic name) so we ended up making more variants for
these instruction then we actually needed.
32-bit derivatives need 6 dwords at most, while 16-bit need 4 at most. Using
same AMDGPUSampleVariant for both, we ended up creating 2 extra variants per
instruction than were necessary.
In total this deletes 260 unused tablegen records.
Differential Revision: https://reviews.llvm.org/D131252
Given a poison constant as input, the dyn_cast to a ConstantInt would
fail so we would fall through to the generic code that attempts to fold
each element of the input vectors. The inputs to these intrinsics are
not vectors though, leading to a compile time crash. Instead bail out
properly for poison values by returning nullptr. This doesn't try to
define what poison means for these intrinsics.
Fixes#56945
This fixes 69 llvm tests that failed when EXPENSIVE_CHECKS was enabled.
llvm/test/Transforms/IROutliner/outlining-commutative-operands-opposite-order.ll
is one example.
When we have EXPENSIVE_CHECKS, _GLIBCXX_DEBUG is defined. This means
that libstdc++ will call the compare function to check if it is
implemented correctly (that !(a < a) is true).
This happens even if there is only one item and here, we expect
to see one return void or multiple return constant integer.
Don't sort if we have 1 item, but do assert that it is the 1
ret void we expect. In the comparator, assert that neither
Value is a nullptr in case one ended up in a the list somehow.
Reviewed By: AndrewLitteken
Differential Revision: https://reviews.llvm.org/D130230
According to the description of the LoongArch abi documentation,
(https://loongson.github.io/LoongArch-Documentation/LoongArch-ELF-ABI-EN.html)
the calling convention of LoongArch is almost the same as the RISCV's
(except for the vector part), so we borrow the implementation of RISCV.
This patch only guarantees the correctness of lp64d, because only the
part of lp64d is described in detail in the documentation.
Differential Revision: https://reviews.llvm.org/D130249
Closing https://github.com/llvm/llvm-project/issues/56919
It is meaningless to preserve the lifetime markers for the spilled
allocas in the coroutine frames and it would block some optimizations
too.
This is a simple addition to emitConditionalComparison, to match CCMP
with immediates using getIConstantVRegValWithLookThrough, letting it
select the CCMPri variants of the instructions.
Differential Revision: https://reviews.llvm.org/D131073
`BMI` new instruction `tzcnt` has better performance than `bsf` on new
processors. Its encoding has a mandatory prefix '0xf3' compared to
`bsf`. If we force emit `rep` prefix for `bsf`, we will gain better
performance when the same code run on new processors.
GCC has already done this way: https://c.godbolt.org/z/6xere6fs1Fixes#34191
Reviewed By: craig.topper, skan
Differential Revision: https://reviews.llvm.org/D130956
A const reference is preferred over a non-null const pointer.
`Type *` is kept as is to match the other overload.
Reviewed By: davidxl
Differential Revision: https://reviews.llvm.org/D131197
When folding (sra (add (shl X, 32), C1), 32 - C) -> (shl (sext_inreg (add X, C1), i32), C)
it's possible that the add is used by multiple sras. We should
allow the combine if all the SRAs will eventually be updated.
After transforming all of the sras, the shls will share a single
(sext_inreg (add X, C1), i32).
This pattern occurs if an sra with 32 is used as index in multiple
GEPs with different scales. The shl from the GEPs will be combined
with the sra before we get a chance to match the sra pattern.
COFF has a verifier check that private global variables don't have a comdat of the same name.
Reviewed By: rnk
Differential Revision: https://reviews.llvm.org/D131043
1) Overloaded (instruction-based) method is a wrapper around the current (opcode-based) method.
2) This patch also changes a few callsites (VectorCombine.cpp,
SLPVectorizer.cpp, CodeGenPrepare.cpp) to call the overloaded method.
3) This is a split of D128302.
Differential Revision: https://reviews.llvm.org/D131114
In contrast to AAPotentialValues, the constant values version can
contain implicit `undef` in the set. We had an assertion that could
misfire before. Handle it properly now.
When folding (sra (add (shl X, 32), C1), 32 - C) -> (shl (sext_inreg (add X, C1), C)
ignore the use count on the (shl X, 32).
The sext_inreg after the transform is free. So we're only making
2 new instructions, the add and the shl. So we only need to be
concerned with replacing the original sra+add. The original shl
can have other uses. This helps if there are multiple different
constants being added to the same shl.
This connects the Symbolizer to the markup filter and enables the first
working end-to-end flow using the filter.
Reviewed By: peter.smith
Differential Revision: https://reviews.llvm.org/D130187
As discussed in [0], this diff adds the `skipprofile` attribute to
prevent the function from being profiled while allowing profiled
functions to be inlined into it. The `noprofile` attribute remains
unchanged.
The `noprofile` attribute is used for functions where it is
dangerous to add instrumentation to while the `skipprofile` attribute is
used to reduce code size or performance overhead.
[0] https://discourse.llvm.org/t/why-does-the-noprofile-attribute-restrict-inlining/64108
Reviewed By: phosek
Differential Revision: https://reviews.llvm.org/D130807
This allows the construct to be shared between different backends. However, it
still remains illegal to use TypedPointerType in LLVM IR--the type is intended
to remain an auxiliary type, not a real LLVM type. So no support is provided for
LLVM-C, nor bitcode, nor LLVM assembly (besides the bare minimum needed to make
Type->dump() work properly).
Reviewed By: beanz, nikic, aeubanks
Differential Revision: https://reviews.llvm.org/D130592
This fixes warnings like these:
../lib/ExecutionEngine/Orc/MemoryMapper.cpp:364:9: warning: ignoring return value of function declared with 'warn_unused_result' attribute [-Wunused-result]
joinErrors(std::move(Err),
^~~~~~~~~~ ~~~~~~~~~~~~~~~
Differential Revision: https://reviews.llvm.org/D131056
In RVV, we use vwredsum.vs and vwredsumu.vs for vecreduce.add(ext(Ty A)) if the result type's width is twice of the input vector's SEW-width. In this situation, the cost of extended add reduction should be same as single-width add reduction. So as the vector float widenning reduction.
Differential Revision: https://reviews.llvm.org/D129994
The isOnlyUserOf prevented the fold if the chain result had any
users. What we really care about is the the data result from the
AND is only used by the TEST, and the flags results from the ANDs
aren't used at all. It's ok if the chain has users, we just need
to replace those users with the chain from the TESTrm.
Reviewed By: LuoYuanke
Differential Revision: https://reviews.llvm.org/D131117
BoundsChecking uses ObjectSizeOffsetEvaluator to keep track of the
underlying size/offset of pointers in allocations. However,
ObjectSizeOffsetVisitor (something ObjectSizeOffsetEvaluator
uses to check for constant sizes/offsets)
doesn't quite treat sizes and offsets the same way as
BoundsChecking. BoundsChecking wants to know the size of the
underlying allocation and the current pointer's offset within
it, but ObjectSizeOffsetVisitor only cares about the size
from the pointer to the end of the underlying allocation.
This only comes up when merging two size/offset pairs. Add a new mode to
ObjectSizeOffsetVisitor which cares about the underlying size/offset
rather than the size from the current pointer to the end of the
allocation.
Fixes a false positive with -fsanitize=bounds.
Reviewed By: vitalybuka, asbirlea
Differential Revision: https://reviews.llvm.org/D131001
This is the 2nd patch of the two-patch series (D130188, D130189) that
fix PR56275 (https://github.com/llvm/llvm-project/issues/56275) which
is a missed opportunity for loop interchange.
As follow-up on the dependence analysis (DA) patch D130188, this patch
normalizes DA results in loop interchange, such that negative dependence
vectors queried by loop interchange are reversed to be non-negative.
Now all tests in PR56275 can get interchanged. Those tests are added
in lit test as `pr56275.ll`.
Reviewed By: kawashima-fj, bmahjour, Meinersbur, #loopoptwg
Differential Revision: https://reviews.llvm.org/D130189
This patch is the first of the two-patch series (D130188, D130179) that
resolve PR56275 (https://github.com/llvm/llvm-project/issues/56275)
which is a missed opportunity, where a perfrectly valid case for loop
interchange failed interchange legality.
If the distance/direction vector produced by dependence analysis (DA) is
negative, it needs to be normalized (reversed). This patch provides helper
functions `isDirectionNegative()` and `normalize()` in DA that does the
normalization, and clients can query DA to do normalization if needed.
A pass option `<normalized-results>` is added to DependenceAnalysisPrinterPass,
and we leverage it to update DA test cases to make sure of test coverage. The
test cases added in `Banerjee.ll` shows that negative vectors are normalized
with `print<da><normalized-results>`.
Reviewed By: bmahjour, Meinersbur, #loopoptwg
Differential Revision: https://reviews.llvm.org/D130188
During LTO a local promoted to a global gets a unique suffix based on
a hash of the module IR. This means that changes in the local's module
can affect the contents in another module that imported it (because the name
of the imported promoted local is changed, but that doesn't reflect a
real change in the importing module). So any tool that's
validating changes to the importing module will see a superficial change.
Instead of using the module hash, we can use the "source_filename" if it
exists to generate a unique identifier that doesn't change due to LTO
shenanigans.
Differential Revision: https://reviews.llvm.org/D128863
This just shuffles implementations and declarations around. Now the
logger and the TF C API-based model evaluator are separate.
Differential Revision: https://reviews.llvm.org/D131116
D129980 converts (seteq (i64 (and X, 0xffffffff)), C1) into
(seteq (i64 (sext_inreg X, i32)), C1). If bit 31 of X is 0, it
will be turned back into an 'and' by SimplifyDemandedBits which
can cause an infinite loop.
To prevent this, check if bit 31 is 0 with computeKnownBits before
doing the transformation.
Fixes PR56905.
Reviewed By: reames
Differential Revision: https://reviews.llvm.org/D131113
If we're going to emit a rep prefix before bsf as proposed in
D130956, it makes sense to promote i16 operations to i32 to avoid
the false depedency of tzcntw.
Reviewed By: skan, pengfei
Differential Revision: https://reviews.llvm.org/D130995
The testcase was delta-reduced from an LTO build with sanitizer
coverage and the MIR tail duplication pass caused a machine basic
block to become unreachable in MIR. This caused the MBB to be invisible
to the reverse post-order traversal used to initialize the MBB <->
RPONumber lookup tables.
rdar://97226240
Differential Revision: https://reviews.llvm.org/D130999
The function `handleDebugValue` has custom logic to handle certain kinds
constants, namely integers, floats and null pointers. However, it does
not handle constant pointers created from IntToPtr ConstantExpressions.
This patch addresses the issue by replacing the Constant with its
integer operand.
A similar bug was addressed for GlobalISel in D130642.
Reviewed By: aprantl, #debug-info
Differential Revision: https://reviews.llvm.org/D130908
This is mostly a stylistic change to make the uniform memop widening cost
code fit more naturally with the sourounding code. Its not strictly
speaking NFC as I added in the store with invariant value case, and we
could in theory have a target where a gather/scatter is cheaper than a
single load/store... but it's probably NFC in practice. Note that the
scatter/gather result can still be overriden later if the result is
uniform-by-parts.
This patch ensures consistency in the construction of FP_ROUND nodes
such that they always use ISD::TargetConstant instead of ISD::Constant.
This additionally fixes a bug in the AArch64 SVE backend where patterns
were matching against TargetConstant nodes and sometimes failing when
passed a Constant node.
Reviewed By: paulwalker-arm
Differential Revision: https://reviews.llvm.org/D130370
An unnecessary sext.w is generated when masking the result of the
riscv_masked_cmpxchg_i64 intrinsic. Implementing handling of the
intrinsic in ComputeNumSignBitsForTargetNode allows it to be removed.
Although this isn't a particularly important optimisation, removing the
sext.w simplifies implementation of an additional cmpxchg-related
optimisation in D130192.
Although I can't produce a test with different codegen for the other
atomics intrinsics, these are added as well for completeness.
Differential Revision: https://reviews.llvm.org/D130191
Enable SGPRs for the following operands of these opcodes:
- src operands of VOP3 variant.
- src2 operand of DPP variants.
Differential Revision: https://reviews.llvm.org/D130989
`BMI` new instruction `tzcnt` has better performance than `bsf` on new
processors. Its encoding has a mandatory prefix '0xf3' compared to
`bsf`. If we force emit `rep` prefix for `bsf`, we will gain better
performance when the same code run on new processors.
GCC has already done this way: https://c.godbolt.org/z/6xere6fs1Fixes#34191
Reviewed By: skan
Differential Revision: https://reviews.llvm.org/D130956
Unfortunately, this overflow is extremely hard to reproduce reliably (in fact, I was unable to do so). The issue is that:
- getOperandsToCreate sometimes skips creating an SCEV for the LHS
- then, createSCEV is called for the BinaryOp
- ... which calls getNoWrapFlagsFromUB
- ... which under certain circumstances calls isSCEVExprNeverPoison
- ... which under certain circumstances requires the SCEVs of all operands
For certain deep dependency trees, this causes a stack overflow.
Reviewed By: bkramer, fhahn
Differential Revision: https://reviews.llvm.org/D129745
Mark ModRefInfo as a bitmask enum, which allows using normal
& and | operators on it. This supersedes various functions like
unionModRef() and intersectModRef(). I think this makes the code
cleaner than going through helper functions...
Differential Revision: https://reviews.llvm.org/D130870
Sometimes SCEV cannot infer nuw/nsw from something as simple as
```
len in [0, MAX_INT]
...
iv = phi(0, iv.next)
guard(iv <s len)
guard(iv <u len)
iv.next = iv + 1
```
just because flag strenthening only relies on definition and does not use local facts.
This patch adds support for the simplest case: inference of flags of `add(x, constant)`
if we can contextually prove that `x <= max_int - constant`.
In case if it has negative CT impact, we can add an option to switch it off. I woudln't
expect that though.
Differential Revision: https://reviews.llvm.org/D129643
Reviewed By: apilipenko
In case that opaque pointers not enabled, there may be some constexpr
bitcast uses for thread local variables and the design of llvm allow
people to sink constant arbitrarily. This breaks the assumption of
IRBuilder::CreateThreadLocalAddress. This patch tries to handle the
case.
In this patch we replace common code patterns with the use of utility
functions for dealing with profiling metadata. There should be no change
in functionality, as the existing checks should be preserved in all
cases.
Reviewed By: bogner, davidxl
Differential Revision: https://reviews.llvm.org/D128860
When compiling for multiple targets the scheduler that is selected via the
-misched option is applied globally. This patch adds a target CL option instead.
Reviewed By: rampitec
Differential Revision: https://reviews.llvm.org/D131022
This used to print from the ADDI where the operand number was
correct. It recently changed to print from the LUI or AUIPC which
needs to use operand 1 instead of 2.
This shows up as a crash with -debug.
Prefer using these accessors to access the special sub-commands
corresponding to the top-level (no subcommand) and all sub-commands.
This is a preparatory step towards removing the use of ManagedStatic:
with a subsequent change, these global instances will be moved to
be regular function-scope statics.
It is split up to give downstream projects a (albeit short) window in
which they can switch to using the accessors in a forward-compatible
way.
Differential Revision: https://reviews.llvm.org/D129118
Creates a new scheduling strategy that attempts to maximize ILP for a single
wave.
Reviewed By: rampitec
Differential Revision: https://reviews.llvm.org/D130869
When emit-obj from clang directly, DirectX backend will hit assert caused by not initialize passes for AsmPrinter.
The fix will initialize the passes by calling createPassConfig.
Also ignore global variable which not has section in DXILAsmPrinter::emitGlobalVariable to avoid hit llvm_unreachable in DXILTargetObjectFile::SelectSectionForGlobal.
Reviewed By: beanz
Differential Revision: https://reviews.llvm.org/D130856
Fixes code in OrderedChangedData<T>::report which assumes that a string will only appear once in Before/After.
Reviewed By: jamieschmeiser
Differential Revision: https://reviews.llvm.org/D130587
The LegalizerHelper misses the code to lower G_MUL to a library call,
which this change adds.
Reviewed By: arsenm
Differential Revision: https://reviews.llvm.org/D130987
rGcf97e0ec42b8 makes $x18 to be treated as callee-saved in functions with
Windows calling convention on non-Windows OSes.
Here we mark $x18 as callee-saved for functions with Windows calling
convention on Darwin, as well as on other non-Windows platforms, in
order to prevent some miscompilations (like miscompilation of
win64cc-darwin-backup-x18.ll).
Since getCalleeSavedRegs doesn't return x18 in list of callee-saved
registers, assignCalleeSavedSpillSlots and determineCalleeSaves
consider different sets of registers as callee-saved. It causes an
error:
```
Assertion failed: ((!HasCalleeSavedStackSize || getCalleeSavedStackSize() == Size) && "Invalid size calculated for callee saves"), function getCalleeSavedStackSize, file
AArch64MachineFunctionInfo.h, line 292.
```
Differential Revision: https://reviews.llvm.org/D130676
The ZERO register should be exposed as a constant physical register through the interface TargetRegisterInfo::isConstantPhysReg.
Differential Revision: https://reviews.llvm.org/D130932
I think these pseudos will exist when the post-RA scheduler runs
so they should have sched classes.
Reviewed By: monkchiang
Differential Revision: https://reviews.llvm.org/D130945
In the 2e29b0138c we introduce a specific solving algorithm
that analyzes the VGPR to SGPR copies use chains and either lowers
the copy to v_readfirstlane_b32 or converts the whole chain to VALU forms.
Same time we still have the code that blindly converts to VALU REG_SEQUENCE and PHIs
in case they produce SGPR but have VGPRs input operands. In case the REG_SEQUENCE and PHIs
are in the VGPR to SGPR copy use chain, and this chain was considered long enough to convert
copy to v_readfistlane_b32, further lowering them to VALU leads to several kinds of issues.
At first, we have v_readfistlane_b32 which is completely useless because most parts of its use chain
were moved to VALU forms. Second, we may encounter subtle bugs related to the EXEC-dependent CF
because of the weird mixing of SALU and VALU instructions.
This change removes the code that moves REG_SEQUENCE and PHIs to VALU. Instead, we use the fact
that both REG_SEQUENCE and PHIs have copy semantics. That is, if they define SGPR but have VGPR inputs,
we insert VGPR to SGPR copies to make them pure SGPR. Then, the new copies are processed by the common
VGPR to SGPR lowering algorithm.
This is Part 2 in the series of commits aiming at the massive refactoring of the SIFixSGPRCopies pass.
Reviewed By: rampitec
Differential Revision: https://reviews.llvm.org/D130367
This pass seems to have very little effect because all it does is hoist
some instructions, but it is followed later in the codegen pipeline by
the IR CodeSinking pass which does the opposite.
Differential Revision: https://reviews.llvm.org/D130258
This improves a corner case where v_fmac can be converted to v_fma on
GFX10+ even if it has a literal operand.
Differential Revision: https://reviews.llvm.org/D130992
This extends the handling of uniform memory operations to handle the case where a store is storing a loop invariant value. Unlike the general case of a store to an invariant address where we must use the last active lane, in this case we can use any lane since all lanes must produce the same result.
For context, the basic structure of the existing code and how the change fits in:
* First, we select a widening strategy. (The result is irrelevant for this patch.)
* Then we determine if a computation is uniform within all lanes of VF. (Note this is the uniform-per-part definition, not LAI's uniform across all unrolled iterations definition.)
* If it is, we overrule the widening strategy, and unconditionally scalarize.
* VPReplicationRecipe - which is what actually does the scalarization - knows how to handle unform-per-part values including for scalable vectors. However, we do need to know that the expression is safe to execute without predication - e.g. the uniform mem op was unconditional in the original loop. (This part was split off and already landed.)
An obvious question is why not simply implement the generic case? The answer is that I'm going to, but doing so without a canonicalization towards uniform causes regressions due to bad interaction with scalarization/uniformity of values feeding the uniform mem-op. This patch is needed to avoid those regressions.
Differential Revision: https://reviews.llvm.org/D130364
The problem Alexander reported on D127982 was caused by an optimization
for AVX512-FP16 instruction. We must limit it to the feature enabled only.
During the investigation, I found we didn't expand for fp_round/fp_extend
without F16C. This may result runtime crash, so change them too.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D130817
Add a new IRBuilderBase::CreateIntrinsic which takes the return type and
argument values for the intrinsic call but does not take an explicit
list of types to mangle. Instead the builder works this out from the
intrinsic declaration and the types of the supplied arguments.
This means that the mangling is hidden from the client, which in turn
means that intrinsic definitions can change which arguments are mangled
without requiring any changes to the client code.
Differential Revision: https://reviews.llvm.org/D130776
This folds a v4i32 Mul(And(Srl(X, 15), 0x10001), 0xffff) into a v8i16
CMLTz instruction. The Srl and And extract the top bit (whether the
input is negative) and the Mul sets all values in the i16 half to all
1/0 depending on if that top bit was set. This is equivalent to a v8i16
CMLTz instruction. The same applies to other sizes with equivalent
constants.
Differential Revision: https://reviews.llvm.org/D130874
matchRotateSub is given shift amounts that will already have stripped any/zero-extend nodes from - so make sure those values are wide enough to take a mask.
If we have interleave groups in the loop we want to vectorise then
we should fall back on normal vectorisation with a scalar epilogue. In
such cases when tail-folding is enabled we'll almost certainly go on to
create vplans with very high costs for all vector VFs and fall back on
VF=1 anyway. This is likely to be worse than if we'd just used an
unpredicated vector loop in the first place.
Once the vectoriser has proper support for analysing all the costs
for each combination of VF and vectorisation style, then we should
be able to remove this.
Added an extra test here:
Transforms/LoopVectorize/AArch64/sve-tail-folding-option.ll
Differential Revision: https://reviews.llvm.org/D128342
* TargetFrameLowering has a TransientStackAlignment field that "returns
the number of bytes to which the stack pointer must be aligned at all
times, even between calls.
* As explained in the [RISC-V calling
convention](https://github.com/riscv-non-isa/riscv-elf-psabi-doc/blob/master/riscv-cc.adoc),
the stack pointer must remain fully aligned throughout execution for
compliant code. This is important for embedded targets that might avoid
realigning the stack pointer for interrupt service routines. Systems
running full OSes may always realign the stack anyway.
* TransientStackAlignment is used in estimateStackSize in
MachineFrameInfo and in PEI::calculateFrameObjectOffsets.
* estimateStackSize is only used in the RISC-V backend for scavenging
slots. It may be possible to craft a function where the difference
is observable, but it wouldn't be a meaningful test.
* calculateFrameObjectOffsets makes use of TransientStackAlignment,
but then sets the stack alignment to the max of that alignment and
MaxAlign, which is unconditionally set to 16 in
RISCVFrameLowering::processFunctionBeforeFrameFinalized
* I've changed this logic to only set MaxAlign if there are RVV frame
objects. There should be no functional change here for either RVV
targets (MaxAlign is set as before) or non-RVV targets
(TransientStackAlign is now 16 anyway).
Differential Revision: https://reviews.llvm.org/D130068
Now the API getExtendedAddReductionCost is used to determine the cost of extended Add reduction with optional Mul. For Arm, it could cover the cases. But for other target, for example: RISCV, they support other kinds of extended recution, such as FAdd.
This patch does the following changes:
1, Split getExtendedAddReductionCost into 2 new API: getExtendedReductionCost which handles the extended reduction with addtional input of Opcode; getMulAccReductionCost which handle the MLA cases the getExtendedAddReductionCost.
2, Refactor getReductionPatternCost, add some contraint condition to make sure the getMulAccReductionCost should only handle the reuction of Add + Mul.
Differential Revision: https://reviews.llvm.org/D130868
Reflect in the pointer's offset the length of the leading part
of the consumed string preceding the first converted digit.
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D130912
It's possible we have:
lui a0, %hi(sym)
addi a0, %lo(sym)
addi a0, <offset1>
lw a0, <offset2>(a0)
We want to arrive at
lui a0, %hi(sym+offset1+offset2)
lw a0, %lo(sym+offset1+offset2)
We currently fail to do this because we only consider loads/stores
if we didn't find any arithmetic.
This patch splits arithmetic folding and load/store folding into
two separate phases. The load/store folding can no longer assume
the offset in hi/lo is 0 so we must combine the offsets. I've applied
the same simm32 limit that we applied in the arithmetic folding.
Reviewed By: luismarques
Differential Revision: https://reviews.llvm.org/D130931
The patch replaces SPIRVBaseInfo.* previously created using macros by
the tablegen approach. There are many small changes in other files due to
differences in namespaces. Also, functions in SPIRVUtils are moved to
the llvm namespace.
Differential Revision: https://reviews.llvm.org/D130518
Co-authored-by: Aleksandr Bezzubikov <zuban32s@gmail.com>
Co-authored-by: Michal Paszkowski <michal.paszkowski@outlook.com>
Co-authored-by: Andrey Tretyakov <andrey1.tretyakov@intel.com>
Co-authored-by: Konrad Trifunovic <konrad.trifunovic@intel.com>
2xi64 is the legalized type for wide reductions (like 16xi64) and setting the
cost to 2 makes `load-reduce` and `load-zext-reduce` patterns profitable.
The few performance measurments that I did on an aarch64 machine confirm that
these patterns are actually faster when vectorized.
Differential Revision: https://reviews.llvm.org/D130740
Follow-up to D130434.
Move doSystemDiff to PrintPasses.cpp and call it in MachineFunctionPass.cpp.
Reviewed By: aeubanks
Differential Revision: https://reviews.llvm.org/D130833
For VALU write and memory (VM, L/DS, FLAT) instructions, SQ would insert
wait-states to avoid data hazard. However when there is a DGEMM instruction
in-between them, SQ incorrectly disables the wait-states thus the data hazard
needs to be handled with this workaround.
Reviewed By: rampitec
Differential Revision: https://reviews.llvm.org/D130677
At least based on the lit tests, the coalescer sometimes fails to
propagate the copy from X0 into the branch instruction. This patch
does it manually during isel. The majority of the changes are from
the select patterns.
Some of the changes are just register allocation changes. Only
the Select change affects the whether a b*z instruction is generated
in the tests. I changed the branch pattern for consistency.
Reviewed By: asb
Differential Revision: https://reviews.llvm.org/D130809
The only iterator we're holding points to HiLUI and we never
delete that so I think it is safe to delete everything else
immediately.
I want to split detectAndFoldOffset into two phases. First, combine
LUI+ADDI with any ADD/ADDI/SHXADD that comes after it. This may
open opportunities to fold the ADDI from the LUI+ADDI into a
load/store address. So the load/store folding should run as a
second phase even if the ADD/ADDI/SHXADD made changes.
In order to do this we need to eagerly delete instructions in the
first phase so that we don't have dead users of the LUI+ADDI
when we start the second phase.
Patches to split the phases will come later.
Reviewed By: asb, luismarques
Differential Revision: https://reviews.llvm.org/D130119
Extend hazard recognizer of ReadM0MovRelInterpHazard with
DS_READ_ADDTID and DS_WRITE_ADDTID, as they also
require a manually inserted S_NOP after SALU writing m0.
Reviewed By: rampitec
Differential Revision: https://reviews.llvm.org/D130783
Fix all instances of:
*** Bad machine code: Kill missing from LiveVariables ***
in the X86 CodeGen tests with D129213 applied, which adds verification
of LiveIntervals after the TwoAddressInstruction pass runs.
Differential Revision: https://reviews.llvm.org/D129634
According to the ABI for the Arm Architecture, the value for the
Tag_also_compatible_with eabi attribute is represented by an NTBS entry.
This string value, in turn, is composed of a pair of tag+value encoded
in one of two formats:
- ULEB128: tag, ULEB128: value, 0.
- ULEB128: tag, NBTS: data.
(See [[ 60a8eb8c55/addenda32/addenda32.rst (3373secondary-compatibility-tag) | section 3.3.7.3 on the Addenda to, and Errata in, the ABI for the Arm Architecture ]].)
Currently the Arm assembly parser and streamer ignore the encoding of
the attribute's NTBS value, which can result in incorrect attributes
being emitted in both assembly and object file outputs.
This patch fixes these issues by properly handing the value's encoding.
An update to llvm-readobj to properly handle the attribute's value will be
covered by a separate patch.
Patch by Victor Campos and Lucas Prates.
Reviewed By: vhscampos
Differential Revision: https://reviews.llvm.org/D129500
Scope of changes:
1) Added new function to generate loop versioning
2) Added support for if clause to applySimd function
2) Added tests which confirm that lowering is successful
If ifCond is specified, then collapsed loop is duplicated and if branch
is added. Duplicated loop is executed if simd ifCond is evaluated to false.
Reviewed By: Meinersbur
Differential Revision: https://reviews.llvm.org/D129368
Signed-off-by: Dominik Adamski <dominik.adamski@amd.com>
Eliminate an AND by redefining an anyext|sext|zext.
(and (extract_subvector (anyext|sext|zext v) _) iN_mask)
=> (extract_subvector (zeroext_iN v))
Differential Revision: https://reviews.llvm.org/D130782
Builds upon D123264, adding support for merging the low part of the LLA
address into the load/store instruction offsets.
Differential Revision: https://reviews.llvm.org/D123265
Salvage debug info of instruction that is about to be deleted as dead in
Combiner pass. Currently supported instructions are COPY and G_TRUNC.
It allows to salvage debug info of some dead arguments of functions, by putting
DWARF expression corresponding to the instruction being deleted into related
DBG_VALUE instruction.
Here is an example of missing variables location https://godbolt.org/z/K48osb9dK.
We see that arguments x, y of function foo are not available in debugger, and
corresponding DBG_VALUE instructions have undefined register operand instead of
variables locaton after Aarch64PreLegalizerCombiner pass. The reason is that
registers where variables are located are removed as dead (with instruction
G_TRUNC). We can use salvageDebugInfo analogue for gMIR to preserve debug
locations of dead variables.
Statistics of llvm object files built with vs without this commit on -O2
optimization level (CMAKE_BUILD_TYPE=RelWithDebInfo, -fglobal-isel) on Aarch64 (macOS):
Number of variables with 100% of parent scope covered by DW_AT_location has been increased by 7,9%.
Number of variables with 0% coverage of parent scope has been decreased by 1,2%.
Number of variables processed by location statistics has been increased by 2,9%.
Average PC ranges coverage has been increased by 1,8 percentage points.
Coverage can be improved by supporting more instructions, or by calling
salvageDebugInfo for instructions that are deleted during Combiner rules exection.
Reviewed By: aprantl
Differential Revision: https://reviews.llvm.org/D129909
SimplifyCFG does some common code hoisting, which is limited to hoisting a
sequence of identical instruction in identical order and stops at the first
non-identical instruction.
This patch allows hoisting instruction pairs over same-length sequences of
non-matching instructions. The linear asymptotic complexity of the algorithm
stays the same, there's an extra parameter `simplifycfg-hoist-common-skip-limit`
serving to limit compilation time and/or the size of the hoisted live ranges.
The patch improves SPECv6/525.x264_r by about 10%.
Reviewed By: dmgreen
Differential Revision: https://reviews.llvm.org/D129370
getModRefInfo() queries currently track whether the result is a
MustAlias on a best-effort basis. The only user of this functionality
is the optimized memory access type in MemorySSA -- which in turn
has no users. Given that this functionality has not found a user
since it was introduced five years ago (in D38862), I think we
should drop it again.
The context is that I'm working to separate FunctionModRefBehavior
to track mod/ref for different location kinds (like argmem or
inaccessiblemem) separately, and the fact that ModRefInfo also has
an unrelated Must flag makes this quite awkward, especially as this
means that NoModRef is not a zero value. If we want to retain the
functionality, I would probably split getModRefInfo() results into
a part that just contains the ModRef information, and a separate
part containing a (best-effort) AliasResult.
Differential Revision: https://reviews.llvm.org/D130713
This belongs to a series of patches which try to solve the thread
identification problem in coroutines. See
https://discourse.llvm.org/t/address-thread-identification-problems-with-coroutine/62015
for a full background.
The problem consists of two concrete problems: TLS variable and readnone
functions. This patch tries to convert the TLS problem to readnone
problem by converting the access of TLS variable to an intrinsic which
is marked as readnone.
The readnone problem would be addressed in following patches.
Reviewed By: nikic, jyknight, nhaehnle, ychen
Differential Revision: https://reviews.llvm.org/D125291
Expand load address pseudo-instructions earlier (pre-ra) to allow follow-up
patches to fold the addi of PseudoLLA instructions into the immediate
operand of load/store instructions.
Differential Revision: https://reviews.llvm.org/D123264
issue #56775
I rearranged the Thumb2 codegen test to avoid simplifying the chain
of rounding instructions. I'm assuming the intent of the test is
to verify lowering of each of those intrinsics.
Only PACKSS/PACKUS faux shuffles make use of the demanded elts at the moment, but this at least improves the handling of a couple of truncation patterns.
Handles COMDAT symbol with an offset and refactor the code to only generated symbol if the second symbol was encountered. This happens very infrequently but happens in recursive_mutex implementation of MSVC STL library.
Reviewed By: lhames
Differential Revision: https://reviews.llvm.org/D130454
Implements remaining IMAGE_REL_AMD64_REL32_*. We only need IMAGE_REL_AMD64_REL32_4 for now but doing all remaining ones for completeness. (clang only uses IMAGE_REL_AMD64_REL32_1 and IMAGE_REL_AMD64_REL32)
Reviewed By: lhames
Differential Revision: https://reviews.llvm.org/D130452
Relax zero-fill edge assertions to only consider relocation edges. Keep-alive edges to zero-fill blocks can cause this assertion which is too strict.
Reviewed By: lhames
Differential Revision: https://reviews.llvm.org/D130450
Implements include/alternatename linker directive. Alternatename is used by static msvc runtime library. Alias symbol is technically incorrect (we have to search for external definition) but we don't have a way to represent this in jitlink/orc yet, this is solved in the following up patch.
Inlcude linker directive is used in ucrt to forcelly lookup the static initializer symbols so that they will be emitted. It's implemented as extenral symbols with live flag on that cause the lookup of these symbols.
Reviewed By: lhames
Differential Revision: https://reviews.llvm.org/D130276
This adds a merge operand to all of the binary _VL nodes. Including
integer and widening. They all share multiclasses in tablegen
so doing them all at once was easiest.
I plan to use FADD_VL in an upcoming patch. The rest are just for
consistency to keep tablegen working.
This does reduce the isel table size by about 25k so that's nice.
Reviewed By: reames
Differential Revision: https://reviews.llvm.org/D130816
Noticed by inspection and I can't seem to make a test case, but SSE arithmetic bit shifts clamp to the max shift amount (i.e. create a sign splat) - combineVectorShiftImm already does something similar.
This patch fixes the error llvm/lib/CodeGen/MachineScheduler.cpp(755): error C2065: 'MISchedCutoff': undeclared identifier in case of NDEBUG and LLVM_ENABLE_ABI_BREAKING_CHECKS.
Note MISchedCutoff is declared under #ifndef NDEBUG.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D130425
This is a follow-up to 2ebfda2417
(replace "if" with "else if" since the cases nuw/nsw
were meant to be handled separately).
Test plan:
1/ ninja check-llvm check-clang check-lld
2/ Bootstrapped LLVM/Clang pass tests
The isa<Constant> check could misfire on an instruction with 2 constant
operands. This bug was introduced with bb789381fc (D36988).
See issue #56810 for a C source example that exposed the bug.
For constants in the range [-2047, 2048] we use addi. If the constant
is -2048 we can use xori. If we don't match this explicitly, we'll
emit an LI for the -2048 followed by an XOR.
It is not necessary to wait for all outstanding memory operations before
barriers on hardware that can back off of the barrier in the event of an
exception when traps are enabled. Add a new subtarget feature which
tracks which HW has this ability.
Reviewed By: #amdgpu, rampitec
Differential Revision: https://reviews.llvm.org/D130722
https://alive2.llvm.org/ce/z/3jYbEH
We should choose one of these forms, and the option that uses
the narrow type allows the motivating example from issue #56294
to reduce. In the best case (no 'not' needed and 'trunc' remains),
this does remove an instruction.
Note that there is what looks like a regression because there
is an existing canonicalization that turns trunc into and+icmp.
That is a long-standing transform, and I'm not sure what effect
reversing it would have.
If the LHS op has a single use then using the more general AND op is likely to allow commutation, load folding, generic folds etc.
Updated version - original version rG057db2002bb3 didn't correctly account for multiple uses of the mask that might be folding "OR(AND(X,C),AND(Y,~C)) -> OR(AND(X,C),ANDNP(C,Y))" in canonicalizeBitSelect
This review is extracted from D96035.
DWARF Debuginfo classes have two representations for DIEs: DWARFDebugInfoEntry
(short) and DWARFDie(extended). Depending on the task, it might be more convenient
to use DWARFDebugInfoEntry or/and DWARFDie. DWARFUnit class already has methods
working with DWARFDie and DWARFDebugInfoEntry. This patch adds more
methods working with DWARFDebugInfoEntry to have paired functionality.
Reviewed By: aprantl
Differential Revision: https://reviews.llvm.org/D126059
Since 814a0abcce, this would break if we
had a function in the module that becomes dead in any codegen IR
pass. The function wasn't deleted since it was initially used in dead
code, but is detached from the call graph and doesn't appear in the PO
traversal. Do a second walk over the module to populate the resources
of any functions which weren't already processed.
If the subregister uses were dead, this would leave the main range
segment pointing to a deleted instruction.
Not sure if this should try to avoid shrinking if we know we don't
have dead components.
Add a method for the various cases where we need to concatenate 2 KnownBits together (BUILD_PAIR and SHIFT_PARTS in particular) - uses the existing APInt::concat 'HiBits.concat(LoBits)' convention
Differential Revision: https://reviews.llvm.org/D130557
Ensure non-terminators don't follow terminators.
This patch fixes the `sdiv-udiv-srem-urem.ll` test failure with
expensive check.
Differential Revision: https://reviews.llvm.org/D130247
A build vector of two extracted elements is equivalent to an extract
subvector where the inner vector is any-extended to the
extract_vector_elt VT, because extract_vector_elt has the effect of an
any-extend.
(build_vector (extract_elt_i16_to_i32 vec Idx+0) (extract_elt_i16_to_i32 vec Idx+1))
=> (extract_subvector (anyext_i16_to_i32 vec) Idx)
Depends on D130697
Differential Revision: https://reviews.llvm.org/D130698
At the moment, proveNoSignedWrapViaInduction may be called for the
same AddRec a large number of times via getSignExtendExpr. This can have
a severe compile-time impact for very loop-heavy code.
If proveNoSignedWrapViaInduction failed to prove NSW the first time,
it is unlikely to succeed on subsequent tries and the cost doesn't seem
to be justified.
This is the signed version of 8daa338297 / D130648.
This can drastically improve compile-time in some excessive cases and
also has a slightly positive compile-time impact on CTMark:
NewPM-O3: -0.06%
NewPM-ReleaseThinLTO: -0.04%
NewPM-ReleaseLTO-g: -0.04%
https://llvm-compile-time-tracker.com/compare.php?from=8daa338297d533db4d1ae8d3770613eb25c29688&to=aed126a196e7a5a9803543d9b4d6bdb233d0009c&stat=instructions
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D130694
Implements include/alternatename linker directive. Alternatename is used by static msvc runtime library. Alias symbol is technically incorrect (we have to search for external definition) but we don't have a way to represent this in jitlink/orc yet, this is solved in the following up patch.
Inlcude linker directive is used in ucrt to forcelly lookup the static initializer symbols so that they will be emitted. It's implemented as extenral symbols with live flag on that cause the lookup of these symbols.
Reviewed By: lhames
Differential Revision: https://reviews.llvm.org/D130276
Implements dllimport stubs using GOT table manager. Benefit of using GOT table manager is that we can just reuse jitlink-check architecture.
Reviewed By: lhames
Differential Revision: https://reviews.llvm.org/D130175
Handles COFF import files of static archive. Changes static library genrator to build up object file map keyed by symbol name that excludes the symbols from dllimported symbols so that static generator will not be responsible for them. It exposes the list of dynamic libraries that need to be imported. Client should properly load the libraries in this list beforehand. Object file map is also an improvment from the past in terms of performance. Archive.findSym does a slow O(n) linear serach of symbol list to find the symbol. (we call findSym O(n) times, thus full time complexity is O(n^2); we were the only user of findSym function in fact)
There is a room for improvements in how to load the libraries in the list. We currently just hand the responsibility over to the clinet. A better way would be let ORC read this list and hand them over to JITLink side that would also help validation (e.g. not trying to generate stub for non dllimported targets) Nevertheless, we will have to exclude the symbols from COFF import object file list and need a way to access this list, which this patch offers.
Reviewed By: lhames
Differential Revision: https://reviews.llvm.org/D129952
Summary:
Flat scratch load of D16 type by default has tied vdst_in operand (with vdst). This should be taken
care of at the time of "removeOperand" in eliminateFrameIndex. Otherwise we will hit an assert saying
"Cannot move tied operands". This patch unties vdst_in before the move, and retie it with vdst afterwards.
Reviewers:
arsenm, foad
Differential Revision: https://reviews.llvm.org/D130537
Currently, the LLVM IR -> MIR translator fails to translate dbg.values
whose first argument is a null pointer. However, in other portions of
the code, such pointers are always lowered to the constant zero, for
example see IRTranslator::Translate(Constant, Register).
This patch addresses the limitation by following the same approach of
lowering null pointers to zero.
A prior test was checking that null pointers were always lowered to
$noreg; this test is changed to check for zero, and the previous
behavior is now checked by introducing a dbg.value whose first argument
is the address of a global variable.
Differential Revision: https://reviews.llvm.org/D130721
The getOperand method already returns a Constant when it is called on
a ConstantExpression, as such the cast is not needed. To prevent a type
mismatch between the different return statements of the lambda, the
lambda return type is explicitly provided.
Differential Revision: https://reviews.llvm.org/D130719
When register pressure tracking is disabled, the scheduler attempts to load
pressures at SReg_32 and VGPR_32. This causes an index out of bounds error.
This patch fixes this issue by disabling the initialization of RPTracker
when not needed. NFC
Reviewed By: rampitec, kerbowa, arsenm
Differential Revision: https://reviews.llvm.org/D129322
This still only includes the dwo name if it's in the DW_AT_dwo_name
attribute in the split unit - though it could be improved/modified to
use the dwo name from the command line (if linking raw dwo files) or
retrieved from the DW_AT_dwo_name in the executable (when using -e).
It's useful in any case because you might have a large command line with
many files and knowing exactly which dwo files are relevant will
simplify debugging, but especially with '-e' when you didn't pass the
dwo files explicitly in nthe first place it would be quite non-obvious
where the duplicate units are coming from.
Current DWARFLinker implementation does not support some debug sections
(mainly DWARF v5 sections). This patch adds diagnostic for such sections.
The warning would be displayed for critical(such that could not be removed)
sections and the source file would be skipped. Other unsupported sections
would be removed and warning message should be displayed. The zero exit
status would be returned for both cases.
Reviewed By: JDevlieghere
Differential Revision: https://reviews.llvm.org/D123623
This builtin allows the creation of custom scheduling pipelines on a per-region
basis. Like the sched_barrier builtin this is intended to be used either for
testing, in situations where the default scheduler heuristics cannot be
improved, or in critical kernels where users are trying to get performance that
is close to handwritten assembly. Obviously using these builtins will require
extra work from the kernel writer to maintain the desired behavior.
The builtin can be used to create groups of instructions called "scheduling
groups" where ordering between the groups is enforced by the scheduler.
__builtin_amdgcn_sched_group_barrier takes three parameters. The first parameter
is a mask that determines the types of instructions that you would like to
synchronize around and add to a scheduling group. These instructions will be
selected from the bottom up starting from the sched_group_barrier's location
during instruction scheduling. The second parameter is the number of matching
instructions that will be associated with this sched_group_barrier. The third
parameter is an identifier which is used to describe what other
sched_group_barriers should be synchronized with. Note that multiple
sched_group_barriers must be added in order for them to be useful since they
only synchronize with other sched_group_barriers. Only "scheduling groups" with
a matching third parameter will have any enforced ordering between them.
As an example, the code below tries to create a pipeline of 1 VMEM_READ
instruction followed by 1 VALU instruction followed by 5 MFMA instructions...
// 1 VMEM_READ
__builtin_amdgcn_sched_group_barrier(32, 1, 0)
// 1 VALU
__builtin_amdgcn_sched_group_barrier(2, 1, 0)
// 5 MFMA
__builtin_amdgcn_sched_group_barrier(8, 5, 0)
// 1 VMEM_READ
__builtin_amdgcn_sched_group_barrier(32, 1, 0)
// 3 VALU
__builtin_amdgcn_sched_group_barrier(2, 3, 0)
// 2 VMEM_WRITE
__builtin_amdgcn_sched_group_barrier(64, 2, 0)
Reviewed By: jrbyrnes
Differential Revision: https://reviews.llvm.org/D128158
This avoids a vmerge at the end and avoids spurious fflags updates.
This isn't used for constrained intrinsic so we technically don't have
to worry about fflags, but it doesn't cost much to support it.
To support I've extend our FCOPYSIGN_VL node to support a passthru
operand. Similar to what was done for VRGATHER*_VL nodes.
I plan to do a similar update for trunc, floor, and ceil.
Reviewed By: reames, frasercrmck
Differential Revision: https://reviews.llvm.org/D130659
GetDemandedBits is mainly a wrapper around SimplifyMultipleUseDemandedBits now, and is only used by DAGCombiner::visitSTORE so I've moved all remaining functionality there.
visitSTORE was making use of this to 'simplify' constants for a trunc-store. Just removing this code left to a mixture of regressions and gains - it came down to whether a target preferred a sign or zero extended constant for materialization/truncation. I've just moved the code over for now, but a next step would be to move this to targetShrinkDemandedConstant, but some targets that override the method expect a basic binop, and might react badly to a store node.....
We already had the reasoning about uniform mem op loads; if the address is accessed at least once, we know the instruction doesn't need predicated to ensure fault safety. For stores, we do need to ensure that the values visible in memory are the same with and without predication. The easiest sub-case to check for is that all the values being stored are the same. Since we know that at least one lane is active, this tells us that the value must be visible.
Warning on confusing terminology: "uniform" vs "uniform mem op" mean two different things here, and this patch is specific to the later. It would *not* be legal to make this same change for merely "uniform" operations.
Differential Revision: https://reviews.llvm.org/D130637
This patch introduces the inline cost priority into the
module inliner, which uses the same computation as
InlineCost.
Reviewed By: kazu
Differential Revision: https://reviews.llvm.org/D130012
I'm actually trying to get rid of GetDemandedBits - but while dismantling it I noticed that we were altering opaque constants. Fixing that causes a FP_TO_INT_SAT regression that should be addressed separately - I'll raise a bug.
This patch introduces the inline cost priority into the
module inliner, which uses the same computation as
InlineCost.
Reviewed By: kazu
Differential Revision: https://reviews.llvm.org/D130012
This patch allows SimplifyDemandedBits to call SimplifyMultipleUseDemandedBits in cases where the ISD::SRL source operand has other uses, enabling us to peek through the shifted value if we don't demand all the bits/elts.
This is another step towards removing SelectionDAG::GetDemandedBits and just using TargetLowering::SimplifyMultipleUseDemandedBits.
There a few cases where we end up with extra register moves which I think we can accept in exchange for the increased ILP.
Differential Revision: https://reviews.llvm.org/D77804
It simplifies the code overall and removes the need for manual bookkeeping.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D130447
It simplifies the code overall and removes the need for manual bookkeeping.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D130444
In the 2e29b0138c we introduce a specific solving algorithm
that analyzes the VGPR to SGPR copies use chains and either lowers
the copy to v_readfirstlane_b32 or converts the whole chain to VALU forms.
Same time we still have the code that blindly converts to VALU REG_SEQUENCE and PHIs
in case they produce SGPR but have VGPRs input operands. In case the REG_SEQUENCE and PHIs
are in the VGPR to SGPR copy use chain, and this chain was considered long enough to convert
copy to v_readfistlane_b32, further lowering them to VALU leads to several kinds of issues.
At first, we have v_readfistlane_b32 which is completely useless because most parts of its use chain
were moved to VALU forms. Second, we may encounter subtle bugs related to the EXEC-dependent CF
because of the weird mixing of SALU and VALU instructions.
This change removes the code that moves REG_SEQUENCE and PHIs to VALU. Instead, we use the fact
that both REG_SEQUENCE and PHIs have copy semantics. That is, if they define SGPR but have VGPR inputs,
we insert VGPR to SGPR copies to make them pure SGPR. Then, the new copies are processed by the common
VGPR to SGPR lowering algorithm.
This is Part 2 in the series of commits aiming at the massive refactoring of the SIFixSGPRCopies pass.
Reviewed By: rampitec
Differential Revision: https://reviews.llvm.org/D130367
Fix "JIT session error: Symbols not found: [ DW.ref.__gxx_personality_v0 ] error" which happens when trying to use exceptions on ppc linux. To do this, it expands AutoClaimSymbols option in RTDyldObjectLinkingLayer to also claim weak symbols before they are tried to be resovled. In ppc linux, DW.ref symbols is emitted as weak hidden symbols in the later stage of MC pipeline. This means when using IRLayer (i.e. LLJIT), IRLayer will not claim responsibility for such symbols and RuntimeDyld will skip defining this symbol even though it couldn't resolve corresponding external symbol.
Reviewed By: sgraenitz
Differential Revision: https://reviews.llvm.org/D129175
This works with any logic + extend:
https://alive2.llvm.org/ce/z/vzsqQD
The motivating case is from issue #56294, but that's still not optimal
(it should simplify completely).
At the moment, proveNoUnsignedWrapViaInduction may be called for the
same AddRec a large number of times via getZeroExtendExpr. This can have
a severe compile-time impact for very loop-heavy code. One one
particular workload, LSR takes ~51s without this patch, almost
exlusively in proveNoUnsignedWrapViaInduction. With this patch, the time
in LSR drops to ~0.4s.
If proveNoUnsignedWrapViaInduction failed to prove NUW the first time,
it is unlikely to succeed on subsequent tries and the cost doesn't seem
to be justified.
Besides drastically improving compile-time in some excessive cases, this
also has a slightly positive compile-time impact on CTMark:
NewPM-O3: -0.07%
NewPM-ReleaseThinLTO: -0.08%
NewPM-ReleaseLTO-g: -0.06
https://llvm-compile-time-tracker.com/compare.php?from=b435da027d7774c24cdb8c88d09f6b771e07fb14&to=f2729e33e8284b502f6c35a43345272252f35d12&stat=instructions
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D130648
DebugLocEntry assumes that it either contains 1 item that has no fragment
or many items that all have fragments (see the assert in addValues).
When EXPENSIVE_CHECKS is enabled, _GLIBCXX_DEBUG is defined. On a few machines
I've checked, this causes std::sort to call the comparator even
if there is only 1 item to sort. Perhaps to check that it is implemented
properly ordering wise, I didn't find out exactly why.
operator< for a DbgValueLoc will crash if this happens because the
optional Fragment is empty.
Compiler/linker/optimisation level seems to make this happen
or not. So I've seen this happen on x86 Ubuntu but the buildbot
for release EXPENSIVE_CHECKS did not have this issue.
Add an explicit check whether we have 1 item.
Reviewed By: aprantl
Differential Revision: https://reviews.llvm.org/D130156
As test in PR56672 shows, LAA produces different results which lead to either
positive or negative vectorization decisions depending on the order of blocks
in loop. The exact reason of this is not clear to me, however this makes investigation
of related bugs extremely complex.
Current order of blocks in the loop is arbitrary. It may change, for example, if loop
info analysis is dropped and recomputed. Seems that it interferes with LAA's logic.
This patch chooses fixed traversal order of blocks in loops, making it RPOT.
Note: this is *not* a fix for bug with incorrect analysis result. It just makes
the answer more robust to make the investigation easier.
Differential Revision: https://reviews.llvm.org/D130482
Reviewed By: aeubanks, fhahn
We will insert a new operand which is identical to the Dest for complex
FMUL with a mask. https://godbolt.org/z/eTEdnYv3q
Complex FMA and FMUL with maskz don't have this problem.
Reviewed By: LuoYuanke, skan
Differential Revision: https://reviews.llvm.org/D130638
By not clustering loads and adjusting heuristics to more aggressively reduce
register pressure we may be able to increase occupancy for the function if it
was dropped in a first pass scheduling.
Similarly, try to reduce spilling if register usage exceeds lower bound
occupancy.
Reviewed By: rampitec
Differential Revision: https://reviews.llvm.org/D130329
Clear all kill flags on source register when folding a COPY.
This is necessary because the kills may now be out of order with the uses.
Reviewed By: foad
Differential Revision: https://reviews.llvm.org/D130622
InstCombine and DAGCombine prefer to keep shl before binops.
This patch teaches isel to convert to (shl (and/or/xor X, C1 >> C2), C2)
if (C1 >> C2) is a simm12. The idea was taken from X86's isel code.
There's a special case implemented for a sext_inreg between the
shift and the binop.
Differential Revision: https://reviews.llvm.org/D130610
tryLastChanceRecoloring iterates over the set of LiveInterval pointers
and used that to seed the recoloring stack, which was
nondeterministic. Fixes a future test failing about 20% of the time.
This just takes the order the interfering vreg was encountered. Not
sure if we should try to order this more intelligently.
In this patch we replace common code patterns with the use of utility
functions for dealing with profiling metadata. There should be no change
in functionality, as the existing checks should be preserved in all
cases.
Reviewed By: bogner, davidxl
Differential Revision: https://reviews.llvm.org/D128860
Currently, there is significant code duplication for dealing with
MD_prof metadata throughout the compiler. These utility functions can
improve code reuse and simplify boilerplate code when dealing with
profiling metadata, such as branch weights. The inent is to provide a
uniform set of APIs that allow common tasks, such as identifying
specific types of MD_prof metadata and extracting branch weights.
Future patches can build on this initial implementation and clean up the
different implementations across the compiler.
Reviewed By: bogner
Differential Revision: https://reviews.llvm.org/D128858
Currently, the IR to MIR translator can only handle two kinds of constant
inputs to dbg.values intrinsics: constant integers and constant floats. In
particular, it cannot handle pointers created from IntToPtr ConstantExpression
objects.
This patch addresses the limitation above by replacing the IntToPtr with
its input integer prior to converting the dbg.value input.
Patch by Felipe Piovezan!
Differential Revision: https://reviews.llvm.org/D130642
This change enables vectorization (using scalable vectorization only, fixed vectors are not yet enabled) for RISCV when vector instructions are available for the target configuration.
At this point, the resulting configuration should be both stable (e.g. no crashes), and profitable (i.e. few cases where scalar loops beat vector ones), but is not going to be particularly well tuned (i.e. we emit the best possible vector loop). The goal of this change is to align testing across organizations and ensure the default configuration matches what downstreams are using as closely as possible.
This exposes a large amount of code which hasn't otherwise been on by default, and thus may not have been fully exercised. Given that, having issues fall out is not unexpected. If you find issues, please make sure to include as much information as you can when reverting this change.
Differential Revision: https://reviews.llvm.org/D129013
The instruction is used to modify wave priority with the intent
to affect VALU execution and currently we can reschedule VALU
around it since that VALU does not have side effects.
Differential Revision: https://reviews.llvm.org/D130654
Teach libDebugInfo (llvm-dwarfdump) and lldb about DWARF tags and
attributes for pointer authentication. These values have been emitted by
Apple clang for several releases. Although upstream LLVM doesn't emit
these values yet, we hope to upstream that part sometime soon.
Differential revision: https://reviews.llvm.org/D130215
Without this, the intrinsic will be expanded to an integer; thereby an
explicit copy (from GPR to SIMD register) will be codegen'd. This matches the
general convention of using "v1" types to represent scalar integer operations in
vector registers.
The similar approach is observed in D56616, and the pattern likely applies on
other intrinsic that accepts integer scalars (e.g.,
int_aarch64_neon_sqdmulls_scalar)
Differential Revision: https://reviews.llvm.org/D130548
This adds similar heuristics to G_GLOBAL_VALUE, querying the cost of
materializing a specific constant in code size. Doing so prevents us from
sinking constants which require multiple instructions to generate into
use blocks.
Code size savings on CTMark -Os:
Program size.__text
before after diff
ClamAV/clamscan 381940.00 382052.00 0.0%
lencod/lencod 428408.00 428428.00 0.0%
SPASS/SPASS 411868.00 411876.00 0.0%
kimwitu++/kc 449944.00 449944.00 0.0%
Bullet/bullet 463588.00 463556.00 -0.0%
sqlite3/sqlite3 284696.00 284668.00 -0.0%
consumer-typeset/consumer-typeset 414492.00 414424.00 -0.0%
7zip/7zip-benchmark 595244.00 594972.00 -0.0%
mafft/pairlocalalign 247512.00 247368.00 -0.1%
tramp3d-v4/tramp3d-v4 372884.00 372044.00 -0.2%
Geomean difference -0.0%
Differential Revision: https://reviews.llvm.org/D130554
I am playing with the LoopDataPrefetch pass and found out that it
bails to work with a pointer in a non-zero address space. This
patch adds the target callback to check if an address space is to
be considered for prefetching. Default implementation still only
allows address space 0, so this is NFCI.
This does not currently affect any known targets, but seems to be
generally useful for the future.
Differential Revision: https://reviews.llvm.org/D129795
I don't have any evidence these particular uses are actually causing any
issues, but we should avoid accidentally truncating immediate values
depending on the host.
We can use slli.uw by C followed by sh1add. Similar can be done
for multiples of 5 and 9. We need to make sure that C is less than
32 to stay in bounds of the 5-bit immediate for slli.uw.
We have existing patterns for (mul X, 3<<C) that use sh1add
followed by slli. That order doesn't allow the and to be folded.
Reviewed By: reames
Differential Revision: https://reviews.llvm.org/D130146
A mul by a negated power of 2 is a slli followed by neg. This doesn't
require any constant materialization and may be lower latency than mul.
The neg may also be foldable into other arithmetic.
Reviewed By: reames
Differential Revision: https://reviews.llvm.org/D130047
We can't guarantee the long always 64 bits like WINDOWS or LLP64 data
model (rare but we should consider).
So use int64_t from inttypes.h and safe in this case.
Fixes https://github.com/llvm/llvm-project/issues/55911 .
It errors out in the Bazel CI:
AMDGPULowerModuleLDSPass.cpp:384:12: error: chosen constructor is
explicit in copy-initialization
return {SGV, std::move(Map)};
Reviewed By: rupprecht
Differential Revision: https://reviews.llvm.org/D130623
The behaviour of this patch is not great, but it has some side-effects
that are required for OpenMPOpt to work. The problem is that when we use
`-mlink-builtin-bitcode` we only import used symbols from the runtime.
Then OpenMPOpt will insert calls to symbols that were not previously
included. This patch removed this implicit behaviour as these functions
were kept alive by the `noinline` simply because it kept calls to them
in the module. This caused regression in some tests that relied on some
OpenMPOpt passes without using LTO. Reverting for the LLVM15 release but
will try to fix it more correctly on main.
This reverts commit d61d72dae6.
Fixes#56752
This patch changes legacy LTO to set data-sections by default. The user can
explicitly unset data-sections. The reason for this patch is to match the
behaviour of lld and gold plugin. Both lld and gold plugin have data-sections on
by default.
This patch also fixes the forwarding of the clang options -fno-data-sections and
-fno-function-sections to libLTO. Now, when -fno-data/function-sections are
specified in clang, -data/function-sections=0 will be passed to libLTO to
explicitly unset data/function-sections.
Reviewed By: w2yehia, MaskRay
Differential Revision: https://reviews.llvm.org/D129401
This patch changes legacy LTO to set data-sections by default. The user can
explicitly unset data-sections. The reason for this patch is to match the
behaviour of lld and gold plugin. Both lld and gold plugin have data-sections on
by default.
This patch also fixes the forwarding of the clang options -fno-data-sections and
-fno-function-sections to libLTO. Now, when -fno-data/function-sections are
specified in clang, -data/function-sections=0 will be passed to libLTO to
explicitly unset data/function-sections.
Reviewed By: w2yehia, MaskRay
Differential Revision: https://reviews.llvm.org/D129401
Instructions between two adjacent loops will be hoisted above the first
loop, or sunk below the second to facilitate loop fusion. Hoisting will
be attempted for an instruction that dominates the first loop.
Otherwise, sinking this instructions will be attempted.
Instructions with side effects will not be considered for sinking or
hoisting. Hoisting/sinking of any instructions between loops will only
be performed if all the instructions can be moved. As well,
sinking/hoisting is considered for each instruction in isolation,
without taking into account sinking/hoisting decisions for other
instructions in the preheader.
Differential Revision: https://reviews.llvm.org/D118076
While working on D118450 <https://reviews.llvm.org/D118450>, I noticed that
`sys::getHostCPUName` lacks SPARC support.
This patch implements it. The code is taken from/inspired by GCC's
`gcc/config/sparc/driver-sparc.cc`. There's one caveat: since LLVM, unlike
GCC, doesn't support the SPARC-M7, -S7, and -M8 CPUs, I map all those to
the latest supported one (UltraSparc T4/`niagara4`).
Tested on `sparcv9-sun-solaris2.11` and `sparc64-unknown-linux-gnu` by
running `savcov --version` on
- Netra SPARC S7-2 (SPARC-S7, Solaris 11.4)
- SPARC T5-2 (SPARC T5, Solaris 11.4)
- SPARC Enterprise T5220 (UltraSPARC T2, Solaris 11.3)
- SPARC T5 (UltraSPARC T5, Debian sid)
- SPARC T3 (UltraSPARC T3, Debian sid)
- SPARC Enterprise T5220 (Debian sid)
Differential Revision: https://reviews.llvm.org/D130272
SimplifyDemandedBits currently early-outs for multi-use values beyond the root node (just returning the knownbits), which is missing a number of optimizations as there are plenty of cases where we can still simplify when initially demanding all elements/bits.
@lenary has confirmed that the test cases in aea-erratum-fix.ll need refactoring and the current increase codegen is not a major concern.
Differential Revision: https://reviews.llvm.org/D129765
Normally, generic processor does not have any SubtargetFeature. And it
can just generate most basic instructions which have no Predicates to
guard.
But it needs to enbale predicate for the btsti16 instruction as one of the most basic instructions.
Or the generic processor can't finish codegen process. So Add FeatureBTST16 SubtargetFeature to generic ProcessorModel.
Given a patch like D129506, using instructions not valid for the current
feature set becomes an error. This updates the Arm hint-space
instructions for pac/bti to require thumbv7m as opposed to 8.1-m.main, to
make them valid when compiling for thumbv7m with -mbranch-protection.
Differential Revision: https://reviews.llvm.org/D129692
This adds a +atomic-32 target feature, which instructs LLVM to assume
that lock-free 32-bit atomics are available for this target, even
if they usually wouldn't be.
If only atomic loads/stores are used, then this won't emit libcalls.
If atomic CAS is used, then the user is responsible for providing
any necessary __sync implementations (e.g. by masking interrupts
for single-core privileged use cases).
See https://reviews.llvm.org/D120026#3674333 for context on this
change. The tl;dr is that the thumbv6m target in Rust has
historically made atomic load/store only available, which is
incompatible with the change from D120026, which switched these to
use libatomic.
Differential Revision: https://reviews.llvm.org/D130480
Added alloca optimization which was missed during the implemenation of D112098.
Reviewed By: vitalybuka
Differential Revision: https://reviews.llvm.org/D130503
This patch fixes the following error with MSVC 16.9.2 in case of NDEBUG and LLVM_ENABLE_DUMP:
llvm/lib/CodeGen/CodeGenPrepare.cpp(2581): error C2872: 'ExtAddrMode': ambiguous symbol
llvm/include/llvm/CodeGen/TargetInstrInfo.h(86): note: could be 'llvm::ExtAddrMode'
llvm/lib/CodeGen/CodeGenPrepare.cpp(2447): note: or '`anonymous-namespace'::ExtAddrMode'
llvm/lib/CodeGen/CodeGenPrepare.cpp(2581): error C2039: 'print': is not a member of 'llvm::ExtAddrMode'
Reviewed By: aaron.ballman
Differential Revision: https://reviews.llvm.org/D130426
Sanjay Patel