The RISC-V Privileged Specification 1.11 defines `mcountinhibit`, which
has the same numeric CSR value as `mucounteren` from 1.09.1. This patch
enables the use of the old `mucounteren` name.
Patch by Yuichi Sugiyama.
Reviewed By: lenary, jrtc27, pzheng
Differential Revision: https://reviews.llvm.org/D85067
This fixes the "Unable to insert indirect branch" fatal error sometimes
seen when generating position-independent code.
Patch by msizanoen1
Reviewed By: jrtc27
Differential Revision: https://reviews.llvm.org/D84833
Perform lowerShuffleWithVPMOV as part of the v16i8/v8i16 shuffle lowering stages, which are the only types that are currently supported.
We need to expand support for lowering shuffles as truncations to fix the remaining regressions in D66004
Support f128 using VE instructions. Update regression tests.
I've noticed there is no load or store i128 test, so I add them too.
Reviewed By: simoll
Differential Revision: https://reviews.llvm.org/D86035
We can now enable this for AVX1 targets can now assist with canonicalizeShuffleMaskWithHorizOp cleanup.
There's still a few missed opportunities for merging subvector insert/extracts into shuffles, but they shouldn't cause any regressions now.
Instead of just attempting to fold shuffle(HOP,HOP) for a specific target shuffle, make this part of combineX86ShufflesRecursively so we can perform this on the combined shuffle chain, which is particularly useful for recognising more cases of where we're performing multiple HOPs that can be merged and pre-AVX where we don't have good blend/unary target shuffle support.
Split the isRepeatedTargetShuffleMask into a wrapper variant that takes a MVT describing the mask width, and an internal version that just needs the raw mask element bit size.
This will be necessary for an upcoming change where the horizontal ops element width might not match the shuffle mask element width.
This cleans up copies that the legalizer or other combines leave around. They
can occasionally end up escaping as moves.
Differential Revision: https://reviews.llvm.org/D85964
This was always set to 0. Use a default value of 0 in this context to
satisfy the instruction definition patterns. We can't unconditionally
use SLC with a default value of 0 due to limitations in TableGen's
handling of defaulted operands when followed by non-default operands.
The VGPR component is a 32-bit offset, not 64-bits.
I'm not sure what the correct syntax is for this. This maintains the
vaddr position and leaves saddr in the end "off" position. This is
particularly terrible for stores, since the operand order is now <vgpr
offset>, <data>, <sgpr base>, splitting the pointer operands. I
suppose this is a logical consequence from the mistake of not putting
the data operand first. I'm not sure what sp3 does.
This was only used for matching the saddr addressing mode of global
instructions, but this was not implemented correctly. The instruction
definitions aren't even correct, and are defined as using a 64-bit
VGPR component. Eliminate this pass to enable correcting the
instruction definitions. A new matching implementation can work in
GlobalISel or relying on DAG divergence information for the base
address.
It did not process hazard for ds_permute because it does not
load or store even though it is DS.
Differential Revision: https://reviews.llvm.org/D86003
This patch implements initial backend support for a -mtune CPU controlled by a "tune-cpu" function attribute. If the attribute is not present X86 will use the resolved CPU from target-cpu attribute or command line.
This patch adds MC layer support a tune CPU. Each CPU now has two sets of features stored in their GenSubtargetInfo.inc tables . These features lists are passed separately to the Processor and ProcessorModel classes in tablegen. The tune list defaults to an empty list to avoid changes to non-X86. This annoyingly increases the size of static tables on all target as we now store 24 more bytes per CPU. I haven't quantified the overall impact, but I can if we're concerned.
One new test is added to X86 to show a few tuning features with mismatched tune-cpu and target-cpu/target-feature attributes to demonstrate independent control. Another new test is added to demonstrate that the scheduler model follows the tune CPU.
I have not added a -mtune to llc/opt or MC layer command line yet. With no attributes we'll just use the -mcpu for both. MC layer tools will always follow the normal CPU for tuning.
Differential Revision: https://reviews.llvm.org/D85165
A unique module id, which is a part of sinit and sterm function names, is
necessary to be unique. However, `getUniqueModuleId` will fail if there is
no strong external symbol within a module. We turn to use Pid and timestamp
when this happens.
Differential Revision: https://reviews.llvm.org/D85527
This is beginning to look like a canonicalization stage that could be performed as part of shuffle combining
Another step towards PR41813
Recommit of rG9bd97d036398 with fixed offset adjustments
Unfortunately this ends up not working as expected on targets with
16-bit operations due to AMDGPUCodeGenPrepare's promotion of uniform
16-bit ops to i32.
The vector case annoyingly requires switching the checked opcode,
since constants for vectors aren't directly handled.
I also need to think more carefully about whether this is valid for i1.
Remove I8/I16 register classes which are prepared to implement previously
to implement VE ABI. However, it is possible to implement VE ABI correctly
without them. Therefore, removing them now.
Reviewed By: simoll
Differential Revision: https://reviews.llvm.org/D85905
PAL recently got support for multiple ELF sections and relocations,
therefore we can now use .rodata sections instead of forcing constants
into .text.
Differential Revision: https://reviews.llvm.org/D85895
The code wasn't taking into account that the two operands
passed to ptest could be identical and was trying to erase
them twice.
Differential Revision: https://reviews.llvm.org/D85892
If we need a scratch register for the spill don't use the same scratch
register that is being used for the MBUF offset.
Reviewed By: arsenm
Differential Revision: https://reviews.llvm.org/D85772
Allow inlining only when the Callee has a subset of the Caller's
features. In principle, we should be able to inline regardless of any
features because WebAssembly supports features at module granularity,
not function granularity, but without this restriction it would be
possible for a module to "forget" about features if all the functions
that used them were inlined.
Requested in PR46812.
Differential Revision: https://reviews.llvm.org/D85494
These operations take Qda and Rn register operands, which are
commutative so long as the instruction is not predicated.
Differential Revision: https://reviews.llvm.org/D85813
SIPreEmitPeephole does not process all terminators, which means
it can fail to handle SI_RETURN_TO_EPILOG if immediately preceeded
by a branch to the early exit block.
Reviewed By: rampitec
Differential Revision: https://reviews.llvm.org/D85872
From the code after the 'break', they are processing 64bit scalar and
vector bitcast. So I think the break-condition should be (cond1 || cond2)
This means we only execute following code if (64bit and dest-is-vector).
Also remove a previous fix which is not needed with this new fix.
(introduced in: 1349a04ef5)
Reviewed By: arsenm
Differential Revision: https://reviews.llvm.org/D85804
This patch implements the builtins for the vector shifts (shl, srl, sra), and
adds the appropriate test cases for these builtins. The builtins utilize the
vector shift instructions introduced within ISA 3.1.
Differential Revision: https://reviews.llvm.org/D83338
Testing is performed when targeting 128, 256 and 512-bit wide vectors.
For 128-bit vectors, the original behavior of using NEON instructions is
preserved.
Differential Revision: https://reviews.llvm.org/D85479
Similar to the Two op + select patterns that were added recently, this
adds some patterns for select + fma to turn them into predicated
operations.
Differential Revision: https://reviews.llvm.org/D85824
Pull out element equivalence code from isShuffleEquivalent/isTargetShuffleEquivalent, I've also removed many of the index modulos where possible.
First step toward simply adding some additional equivalence tests.
We need to produce a setcc instruction which has an 8-bit result.
This gets rid of a bunch of cases that were using the s1->s8/s16/s32/s64
handling in selectZExt.
I'm not very familiar with GlobalISel yet so I'm not yet sure
the best way to do things. I'd especially like feedback on the
best way to handle the currently split 32-bit and 64-bit mode
handling.
Differential Revision: https://reviews.llvm.org/D85814
Widen the scope of memory operations that are allowed to be tail predicated
to include gathers and scatters, such that loops that are auto-vectorized
with the option -enable-arm-maskedgatscat (and actually end up containing
an MVE gather or scatter) can be tail predicated.
Differential Revision: https://reviews.llvm.org/D85138
When TTI was updated to use an explicit cost, TCK_CodeSize was used
although the default implicit cost would have been the hand-wavey
cost of size and latency. So, revert back to this behaviour. This is
not expected to have (much) impact on targets since most (all?) of
them return the same value for SizeAndLatency and CodeSize.
When optimising for size, the logic has been changed to query
CodeSize costs instead of SizeAndLatency.
This patch also adds a testing option in the unroller so that
OptSize thresholds can be specified.
Differential Revision: https://reviews.llvm.org/D85723
This pick ups the work on the overflow checks for get.active.lane.mask,
which ensure that it is safe to insert the VCTP intrinisc that enables
tail-predication. For a 2d auto-correlation kernel and its inner loop j:
M = Size - i;
for (j = 0; j < M; j++)
Sum += Input[j] * Input[j+i];
For this inner loop, the SCEV backedge taken count (BTC) expression is:
(-1 + (sext i16 %Size to i32)),+,-1}<nw><%for.body>
and LoopUtil cannotBeMaxInLoop couldn't calculate a bound on this, thus "BTC
cannot be max" could not be determined. So overflow behaviour had to be assumed
in the loop tripcount expression that uses the BTC. As a result
tail-predication had to be forced (with an option) for this case.
This change solves that by using ScalarEvolution's helper
getConstantMaxBackedgeTakenCount which is able to determine the range of BTC,
thus can determine it is safe, so that we no longer need to force tail-predication
as reflected in the changed test cases.
Differential Revision: https://reviews.llvm.org/D85737
In this patch I have fixed two issues:
1. Our SVE tuple get/set intrinsics were using the wrong constant type
for the index passed to EXTRACT_SUBVECTOR. I have fixed this by using the
function SelectionDAG::getVectorIdxConstant to create the value. Also, I
have updated the documentation for EXTRACT_SUBVECTOR describing what type
the constant index should be and we now enforce this when creating the
node.
2. The AArch64 backend was missing the appropriate patterns for
extracting certain subvectors (nxv4f16 and nxv2f32) from legal SVE types.
I have added them as part of this patch.
The only way that I could find to test the new patterns was to use the
SVE tuple get intrinsics, although I realise it looks a bit unusual.
Tests added here:
test/CodeGen/AArch64/sve-extract-subvector.ll
Differential Revision: https://reviews.llvm.org/D85516
VE has only 64 bits AND/OR/XOR instructions. We pretended that VE has 32 bits
instructions also, but doing it increase the number of generated instructions.
Therefore, we decide to promote 32 bits operations and use only 64 bits
instructions in back end. We also avoid pretending that VE has 32 bits LEA
instruction. Update regression tests also.
Reviewed By: simoll
Differential Revision: https://reviews.llvm.org/D85726
SUBREG_TO_REG is supposed to be used when we know the producing
instruction already zeroed the bits we're extending. But that's
not the case here. So INSERT_SUBREG with an IMPLICIT_DEF is the
correct thing to use.
Rather than just saying that some feature is missing, report the exact
features to make the error message more useful and actionable.
Differential Revision: https://reviews.llvm.org/D85795
The officially specified abbreviation for WebAssembly is Wasm and the
spec explicitly calls out WASM as being an incorrect spelling. This
patch fixes a few comments and error messages to use the
spec-compliant abbreviation.
Differential Revision: https://reviews.llvm.org/D85764
SUMMARY:
1. in the patch , remove setting storageclass in function .getXCOFFSection and construct function of class MCSectionXCOFF
there are
XCOFF::StorageMappingClass MappingClass;
XCOFF::SymbolType Type;
XCOFF::StorageClass StorageClass;
in the MCSectionXCOFF class,
these attribute only used in the XCOFFObjectWriter, (asm path do not need the StorageClass)
we need get the value of StorageClass, Type,MappingClass before we invoke the getXCOFFSection every time.
actually , we can get the StorageClass of the MCSectionXCOFF from it's delegated symbol.
2. we also change the oprand of branch instruction from symbol name to qualify symbol name.
for example change
bl .foo
extern .foo
to
bl .foo[PR]
extern .foo[PR]
3. and if there is reference indirect call a function bar.
we also add
extern .bar[PR]
Reviewers: Jason liu, Xiangling Liao
Differential Revision: https://reviews.llvm.org/D84765
This implements
```
(logic_op (op x...), (op y...)) -> (op (logic_op x, y))
```
when `op` is an extend, a shift, or an and.
This is similar to `DAGCombiner::hoistLogicOpWithSameOpcodeHands`
(with a bunch of missing cases, e.g. G_TRUNC, G_BITCAST, etc.)
This is implemented so it works both pre and post-legalization.
This also adds a general way to add a series of instructions in a combine.
(`applyBuildInstructionSteps`).
Differential Revision: https://reviews.llvm.org/D85050
This mirrors the support for the equivalent extracts. This also
creates a huge mess that would be greatly improved if we had any bit
operation combines.
ISD::ATOMIC_STORE arbitrarily has the operands in the opposite order
from regular ISD::STORE, which always introduced an annoying
duplication of patterns to handle both cases. Since in GlobalISel
there's just the one G_STORE, we need to swap the operands to
correctly emit the type check for the pointer operand.
Some work started in 20aafa3156 to
migrate SelectionDAG to use ISD::STORE for atomics, but that work
seems to have stalled. Since this is the pretty much the last
operation which matters which isn't supported for AMDGPU, use this
compatibility hack to unblock declaring it functionally complete.
Not sure what's going on with the pending_phis AArch64 test. It seems
it didn't always use atomics, and I'm not sure what it was originally
testing matters anymore.
Changes the Offset arguments to both functions from int64_t to TypeSize
& updates all uses of the functions to create the offset using TypeSize::Fixed()
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D85220
Change bitreverse/bswap/ctlz/ctpop/cttz regression tests to support i128
and signext/zeroext i32 types. This patch also change the way to support
i32 types using 64 bits VE instructions.
Reviewed By: simoll
Differential Revision: https://reviews.llvm.org/D85712
These are useful instructions when lowering fixed length vector
extends, so I've broken this patch out as kind of NFC like work.
Differential Revision: https://reviews.llvm.org/D85546
By factoring out the end of tryVPTERNLOG, we can use the same code
to directly match X86ISD::VPTERNLOG. This allows us to remove
around 3-4K worth of X86GenDAGISel.inc.
When we use mask compare intrinsics under strict FP option, the masked
elements shouldn't raise any exception. So, we cann't replace the
intrinsic with a full compare + "and" operation.
Reviewed By: craig.topper
Differential Revision: https://reviews.llvm.org/D85385
Summary:
Use TE SMC instead of TC SMC in large code model mode,
so that large code model TOC entries could get placed after all
the small code model TOC entries, which reduces the chance of TOC overflow.
Reviewed By: Xiangling_L
Differential Revision: https://reviews.llvm.org/D85455
In cases where MachineOutliner candidates either are:
* noreturn
* have calls with no available LR or free regs
* Don't use SP
we can end up hitting stack fixup code for the caller and the callee for
a FrameID of MachineOutlinerDefault. This triggers the assert:
`assert(OF.FrameConstructionID != MachineOutlinerDefault &&
"Can only fix up stack references once");`
in AArch64InstrInfo.cpp. This assert exists for now because a lot of the
fixup code is not tested to handle fixing up more than once and needs
some better checks and enhancements to avoid potentially generating
illegal code.
I've filed a Bugzilla report to track this until these cases are handled
by the AArch64 MachineOutliner: https://bugs.llvm.org/show_bug.cgi?id=46767
This diff detects cases that will cause these multiple stack fixups and
prune the Candidates from `RepeatedSequenceLocs`.
Differential Revision: https://reviews.llvm.org/D83923
This can fold the immediate into the physical destination, but this
should not look for further users of the register. Fixes regression
introduced by 766cb615a3.
X86 is the only user of this interface in tree. Previously the
X86 pass would loop over operands looking for one undef operand for
the pass to fix. But there could theoretically be multiple operands
to fix. So it makes more sense for the pass to do the looping and
ask the target if an operand needs to be fixed.
If a shuffle is referring to both the lower and upper half lanes of an unary horizontal op, then canonicalize the mask to only refer to the lower half.
Summary:
AIX assembler does not generate correct relocation when .rename
appear between tc entry label and .tc directive.
So only emit .rename after .tc/.comm or other linkage is emitted.
Reviewed By: daltenty, hubert.reinterpretcast
Differential Revision: https://reviews.llvm.org/D85317
On the frontend side, this patch recovers AIX static init implementation to
use the linkage type and function names Clang chooses for sinit related function.
On the backend side, this patch sets correct linkage and function names on aliases
created for sinit/sterm functions.
Differential Revision: https://reviews.llvm.org/D84534
Add a hidden option to the compiler to control a the PC Relative GOT indirect
linker optimization.
If this option is set to false the compiler will no loger produce the
relocations required by the linker to perform the optimization.
Reviewed By: nemanjai, NeHuang, #powerpc
Differential Revision: https://reviews.llvm.org/D85377
IT blocks with more than one instruction were performance deprecated in Armv8
but that doesn't mean we should follow that advise when optimising for size.
Differential Revision: https://reviews.llvm.org/D85638
Check that we're shuffling hadd/pack ops first before altering shuffle masks.
First step towards adding extra functionality, plus it avoids costly shuffle mask manipulation if not necessary.
This patch introduces two intrinsics: llvm.ppc.setflm and
llvm.ppc.readflm. They read from or write to FPSCR register
(floating-point status & control) which contains rounding mode and
exception status.
To ensure correctness of program, we need to prevent FP operations from
being moved across these intrinsics (mffs/mtfsf instruction), so here I
set them as scheduling boundaries. We can relax such restriction if
FPSCR is modeled well in the future.
Reviewed By: steven.zhang
Differential Revision: https://reviews.llvm.org/D84914
Fix 64-bit copy to SCC by restricting the pattern resulting
in such a copy to subtargets supporting 64-bit scalar compare,
and mapping the copy to S_CMP_LG_U64.
Before introducing the S_CSELECT pattern with explicit SCC
(0045786f14), there was no need
for handling 64-bit copy to SCC ($scc = COPY sreg_64).
The proposed handling to read only the low bits was however
based on a false premise that it is only one bit that matters,
while in fact the copy source might be a vector of booleans and
all bits need to be considered.
The practical problem of mapping the 64-bit copy to SCC is that
the natural instruction to use (S_CMP_LG_U64) is not available
on old hardware. Fix it by restricting the problematic pattern
to subtargets supporting the instruction (hasScalarCompareEq64).
Differential Revision: https://reviews.llvm.org/D85207
This adds patterns for v16i16's vecreduce, using all the existing code
to go via an i32 VADDV/VMLAV and truncating the result.
Differential Revision: https://reviews.llvm.org/D85452
This formats some of the MVE patterns, and adds a missing
Predicates = [HasMVEInt] to some VRHADD patterns I noticed
as going through. Although I don't believe NEON would ever
use the patterns (as it would use ADDL and VSHRN instead)
they should ideally be predicated on having MVE instructions.
Fixes PR47040, in which an assertion was improperly triggered during
FastISel's address computation. The issue was that an `Address` set to
be relative to the FrameIndex with offset zero was incorrectly
considered to have an unset base. When the left hand side of an add
set the Address to be 0 off the FrameIndex, the right side would not
detect that the Address base had already been set and could try to set
the Address to be relative to a register instead, triggering an
assertion.
This patch fixes the issue by explicitly tracking whether an `Address`
has been set rather than interpreting an offset of zero to mean the
`Address` has not been set.
Differential Revision: https://reviews.llvm.org/D85581
This was blocking isTypeLegal call so that we could do a particular
transform on illegal types before type legalization. But the we
create a target specific node using that type. We shouldn't do
that if the type isn't legal. So I think we should just always
make sure the type is legal.
I suspect that in order to get the condition VT to not be a vector
of i1 we already completed type legalization anyway so this probably
doesn't matter much in practice.
This is just a thin wrapper around computeRegisterLivness which
we can just call directly. The only real difference is that
isSafeToClobberEFLAGS returns a bool and computeRegisterLivness
returns an enum. So we need to check for the specific enum value
that isSafeToClobberEFLAGS was hiding.
I've also adjusted which sites pass an explicit value for
Neighborhood since the default for computeRegisterLivness is 10.
I messed up the bug numbers in the commit message before
Previously this function searched 4 instructions forwards or
backwards to determine if it was ok to clobber eflags.
This is called in 3 places: rematerialization, turning 2 operand
leas into adds or splitting 3 ops leas into an lea and add on some
CPU targets.
This patch increases the search limit to 10 instructions for
rematerialization and 2 operand lea to add. I've left the old
treshold for 3 ops lea spliting as that increases code size.
Fixes PR47024 and PR46315.
Previously this function searched 4 instructions forwards or
backwards to determine if it was ok to clobber eflags.
This is called in 3 places: rematerialization, turning 2 operand
leas into adds or splitting 3 ops leas into an lea and add on some
CPU targets.
This patch increases the search limit to 10 instructions for
rematerialization and 2 operand lea to add. I've left the old
treshold for 3 ops lea spliting as that increases code size.
Fixes PR47024 and PR43014
Previously the transform was doing these two canonicalizations
(x > y) ? x : y -> (x >= y) ? x : y
(x < y) ? x : y -> (x <= y) ? x : y
But those don't seem to be useful generally. And they actively
pessimize the cases in PR47049.
This patch limits it to
(x > 0) ? x : 0 -> (x >= 0) ? x : 0
(x < -1) ? x : -1 -> (x <= -1) ? x : -1
These are the cases mentioned in the comments as the motivation
for the canonicalization. These allow the CMOV to use the S
flag from the compare thus improving opportunities to use a TEST
or the flags from an arithmetic instruction.
In D85499, I attempted to fix this same issue by canonicalizing
andnp for i1 vectors, but since there was some opposition to such
a change, this commit just fixes the bug by using two different
forms depending on which kind of vector type is in use. We can
then always decide to switch the canonical forms later.
Description of the original bug:
We have a DAG combine that tries to fold (vselect cond, 0000..., X) -> (andnp cond, x).
However, it does so by attempting to create an i64 vector with the number
of elements obtained by truncating division by 64 from the bitwidth. This is
bad for mask vectors like v8i1, since that division is just zero. Besides,
we don't want i64 vectors anyway. For i1 vectors, switch the pattern
to (andnp (not cond), x), which is the canonical form for `kandn`
on mask registers.
Fixes https://github.com/JuliaLang/julia/issues/36955.
Differential Revision: https://reviews.llvm.org/D85553
Regions are sometimes skipped which should be rescheduled without memory op
clustering. RegionIdx is not incremented when iterating over regions that
are flagged to be skipped, causing the index to be incorrect.
Thanks to Vang Thao for discovering this bug!
Reviewed By: rampitec
Differential Revision: https://reviews.llvm.org/D85498
This patch adds the instruction definitions and assembly/disassembly tests for
the following set of instructions:
Vector Extract [byte | half | word | doubleword | quad] with mask
Vector Expand [byte | half | word | doubleword | quad] with mask
Move to VSR [byte | byte immediate | half | word | doubleword | quad] with mask
Vector Count Mask Bits [byte | half | word | doubleword]
Differential Revision: https://reviews.llvm.org/D83724
Introduce a fatal error if any thread local storage code is compiled
using pc relative memory operations as well as a hidden override
option `-enable-ppc-pcrel-tls` so that this support can be incrementally
added if possible.
Reviewed By: #powerpc, nemanjai
Differential Revision: https://reviews.llvm.org/D85448
Change to expand MULHU/MULHS/UMUL_LOHI/SMUL_LOHI for i32 and i64 since
those instructions are not available on Aurora SX VE. Some of them
are used in expansion of i128 multiply, so need to modify them to
support i128. Then, update basic arithmetic regression tests of
i128 and signed/unsigned i32 typed integer values.
Reviewed By: simoll
Differential Revision: https://reviews.llvm.org/D85490
Fixed an incorrect pattern in lib/Target/AArch64/AArch64SVEInstrInfo.td
for storing out <vscale x 2 x f32> unpacked scalable vectors. Added
a couple of tests to
test/CodeGen/AArch64/sve-st1-addressing-mode-reg-imm.ll
Differential Revision: https://reviews.llvm.org/D85441
This patch implements the function prototypes vec_extractl and vec_extracth in altivec.h to utilize the vector extract double element instructions introduced in Power10.
Differential Revision: https://reviews.llvm.org/D84622
Change to not generate truncate instructions if all use of a truncate
operation don't care about higher bits. For example, an i32 add
instruction doesn't care about higher 32 bits in 64 bit registers.
Updates regression tests also.
Reviewed By: simoll
Differential Revision: https://reviews.llvm.org/D85418
Buildbot reported a build failure when building shared
library libLLVMBPFCodeGen.so with unknown reference to
"createCFGSimplificationPass".
Commit 87cba43402 ("BPF: add a SimplifyCFG IR pass during
generic Scalar/IPO optimization") added an IR pass SimplifyCFG
by BPF target. The commit called function
createCFGSimplificationPass() defined in "Scalar" library.
Add this library in Target/BPF/LLVMBuild.txt so
shared library build can succeed.
The following bpf linux kernel selftest failed with latest
llvm:
$ ./test_progs -n 7/10
...
The sequence of 8193 jumps is too complex.
verification time 126272 usec
stack depth 320
processed 114799 insns (limit 1000000)
...
libbpf: failed to load object 'pyperf600_nounroll.o'
test_bpf_verif_scale:FAIL:110
#7/10 pyperf600_nounroll.o:FAIL
#7 bpf_verif_scale:FAIL
After some investigation, I found the following llvm patch
https://reviews.llvm.org/D84108
is responsible. The patch disabled hoisting common instructions
in SimplifyCFG by default. Later on, the code changes and a
SimplifyCFG phase with hoisting on cannot do the work any more.
A test is provided to demonstrate the problem.
The IR before simplifyCFG looks like:
for.cond:
%i.0 = phi i32 [ 0, %entry ], [ %inc, %for.inc ]
%cmp = icmp ult i32 %i.0, 6
br i1 %cmp, label %for.body, label %for.cond.cleanup
for.cond.cleanup:
%2 = load i8*, i8** %frame_ptr, align 8, !tbaa !2
%cmp2 = icmp eq i8* %2, null
%conv = zext i1 %cmp2 to i32
call void @llvm.lifetime.end.p0i8(i64 8, i8* nonnull %1) #3
call void @llvm.lifetime.end.p0i8(i64 8, i8* nonnull %0) #3
ret i32 %conv
for.body:
%3 = load i8*, i8** %frame_ptr, align 8, !tbaa !2
%tobool.not = icmp eq i8* %3, null
br i1 %tobool.not, label %for.inc, label %land.lhs.true
The first two insns of `for.cond.cleanup` and `for.body`, load and
icmp, can be hoisted to `for.cond` block. With Patch D84108, the
optimization is delayed. But unfortunately, later on loop rotation
added addition phi nodes to `for.body` and hoisting cannot
be done any more.
Note such a hoisting is beneficial to bpf programs as
bpf verifier does path sensitive analysis and verification.
The hoisting preverts reloading from stack which will assume
conservative value and increase exploited insns. In this case,
it caused verifier failure.
To fix this problem, I added an IR pass from bpf target
to performance additional simplifycfg with hoisting common inst
enabled.
Differential Revision: https://reviews.llvm.org/D85434
Add cases of fused fmul+fadd/fsub with f16 and f64 operands to cost model.
Also added operations with contract attribute.
Fixed line endings in test.
Reviewed By: rampitec
Differential Revision: https://reviews.llvm.org/D84995
Use the same basic strategy as LegalizeVectorTypes. Try to index into
smaller pieces if there's a constant index, and otherwise fall back to
a stack temporary.
If we were to have an operation with an s16 def that needs to be
executed in a waterfall loop, not having s16 legal would place an
avoidable burden on RegBankSelect to widen it.
This was trying to constrain a physical register. By the verifier's
understanding, it's impossible to have a 1-bit copy to vcc/vcc_lo so
don't try to handle physregs.
This allows us to remove extra patterns from AArch64SVEInstrInfo.td
because we can reuse those required for fixed length vectors.
Differential Revision: https://reviews.llvm.org/D85328
NOTE: Also uses SVE code generation for NEON size vectors, instead
of expanding i64 based vector multiplications.
Differential Revision: https://reviews.llvm.org/D85327
These might occur in seemingly generic assembly. Previously when
targeting COFF, they were silently ignored, which certainly won't
give the right result. Instead clearly error out, to make it clear
that the assembly needs to be adjusted for this target.
Also change a preexisting report_fatal_error into a proper error
message, pointing out the offending source instruction. This isn't
strictly an internal error, as it can be triggered by user input.
Differential Revision: https://reviews.llvm.org/D85242
We need to have special handling of i128 div/rem on Windows due
to a weird calling convention needed for the libcall. There was
also some code that made it look like we do the same for sdivrem/udiv,
but the code didn't account for multiple return values of those
functions so couldn't possibly work. I think this code never
triggers because we don't have libcall names defined for those
functions by default so DAGCombine never creates DIVREM nodes.
The functionality is used when calling imageAtomicExhange() on float
type imageBuffer in Graphics shaders.
Reviewed By: arsenm
Differential Revision: https://reviews.llvm.org/D85187
For example a v4f16 argument is scalarized to 4 i32 values. So
the values are spread out instead of being packed tightly like
in the original vector.
Fixes PR47000.
We've had issues in the past where isHorizontalBinOp calls would affect later combines as the LHS/RHS references had been commuted but still failed to match.
As with other targets, set the throughput cost of control-flow
instructions to free so that we don't miss out of vectorization
opportunities.
Differential Revision: https://reviews.llvm.org/D85283
Since there are no ill effects when performing these operations
with undefined elements, they are lowered to the already supported
unpredicated scalable vector equivalents.
Differential Revision: https://reviews.llvm.org/D85117
This fixes an issue triggered by the following code, where emitEpilogue
got confused when trying to restore the SVE registers after the call,
whereas the call to bar() is implemented as a TCReturn:
int non_sve();
int sve(svint32_t x) { return non_sve(); }
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D84869
This patch simplified IR generation for __builtin_btf_type_id().
For __builtin_btf_type_id(obj, flag), previously IR builtin
looks like
if (obj is a lvalue)
llvm.bpf.btf.type.id(obj.ptr, 1, flag) !type
else
llvm.bpf.btf.type.id(obj, 0, flag) !type
The purpose of the 2nd argument is to differentiate
__builtin_btf_type_id(obj, flag) where obj is a lvalue
vs.
__builtin_btf_type_id(obj.ptr, flag)
Note that obj or obj.ptr is never used by the backend
and the `obj` argument is only used to derive the type.
This code sequence is subject to potential llvm CSE when
- obj is the same .e.g., nullptr
- flag is the same
- metadata type is different, e.g., typedef of struct "s"
and strust "s".
In the above, we don't want CSE since their metadata is different.
This patch change IR builtin to
llvm.bpf.btf.type.id(seq_num, flag) !type
and seq_num is always increasing. This will prevent potential
llvm CSE.
Also report an error if the type name is empty for
remote relocation since remote relocation needs non-empty
type name to do relocation against vmlinux.
Differential Revision: https://reviews.llvm.org/D85174
This is the last remaining use of ConstantProp, migrate it to InstSimplify in the goal of removing ConstantProp.
Add -hexagon-instsimplify option to enable skipping of instsimplify in
tests that can't handle the extra optimization.
Differential Revision: https://reviews.llvm.org/D85047
Get the argument register and ensure there's a copy to the virtual
register. AMDGPU and AArch64 have similarish code to get the livein
value, and I also want to use this in multiple places.
This is a bit more aggressive about setting the register class than
the original function, but that's probably OK.
I think we're missing a few verifier checks for function live ins. I
noticed AArch64's calling convention code is not actually adding
liveins to functions, only the entry block (which apparently might not
matter that much?). There should probably be a verifier check that
entry block live ins are also live into the function. We also might
need a verifier check that the copy to the livein virtual register is
in the entry block.
The SVE instruction set only supports sdiv/udiv for 32-bit and 64-bit
integers. If we see an 8-bit or 16-bit divide, widen the operands to 32
bits, and narrow the result.
Differential Revision: https://reviews.llvm.org/D85170
Four new CO-RE relocations are introduced:
- TYPE_EXISTENCE: whether a typedef/record/enum type exists
- TYPE_SIZE: the size of a typedef/record/enum type
- ENUM_VALUE_EXISTENCE: whether an enum value of an enum type exists
- ENUM_VALUE: the enum value of an enum type
These additional relocations will make CO-RE bpf programs
more adaptive for potential kernel internal data structure
changes.
Differential Revision: https://reviews.llvm.org/D83878
Introduced by fd6584a220
Following similar use of casts in AsmParser.cpp, for instance - ideally
this type would use unsigned chars as they're more representative of raw
data and don't get confused around implementation defined choices of
char's signedness, but this is what it is & the signed/unsigned
conversions are (so far as I understand) safe/bit preserving in this
usage and what's intended, given the API design here.
The swap removal pass looks to remove swaps when a loaded value is swapped, some
number of lane-insensitive operations are performed and then the value is
swapped again and stored.
However, in a situation where we load the value, swap it and then store it
without swapping again, the pass erroneously removes the single swap. The
reason is that both checks in the same equivalence class:
- load feeds a swap
- swap feeds a store
pass. However, there is no check that the two swaps are actually a single swap.
This patch just fixes that.
Differential revision: https://reviews.llvm.org/D84785
The custom lowering saves an instruction over the generic expansion, by
taking advantage of the fact that PowerPC shift instructions are well
defined in the shift-by-bitwidth case.
Differential Revision: https://reviews.llvm.org/D83948
Now that rG47cea9e82dda941e lets us aggressively decode multi-use shuffles for the OR(SHUFFLE(),SHUFFLE()) case we don't need the computeKnownBits variant any more.
Permit lane-crossing post shuffles on AVX1 targets as long as every element comes from the same source lane, which for v8f32/v4f64 cases can be efficiently lowered with the LowerShuffleAsLanePermuteAnd* style methods.
This patch adds a CFI entry for each SVE callee saved register
that needs unwind info at an offset from the CFA. The offset is
a DWARF expression because the offset is partly scalable.
The CFI entries only cover a subset of the SVE callee-saves and
only encodes the lower 64-bits, thus implementing the lowest
common denominator ABI. Existing unwinders may support VG but
only restore the lower 64-bits.
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D84044
The CFA is calculated as (SP/FP + offset), but when there are
SVE objects on the stack the SP offset is partly scalable and
should instead be expressed as the DWARF expression:
SP + offset + scalable_offset * VG
where VG is the Vector Granule register, containing the
number of 64bits 'granules' in a scalable vector.
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D84043
This is the final bit of work to relax the register allocation
requirements when code generating normal LLVM IR, which rarely
care about the result of inactive lanes. By using _PRED nodes
we can make better use of SVE's reversed instructions.
Also removes a redundant parameter from the min/max tests.
Differential Revision: https://reviews.llvm.org/D85142
Added patterns so that both SSAT and USAT instructions are generated with shifts. Added corresponding regression tests.
Differential Review: https://reviews.llvm.org/D85120
[X86][SSE] Shuffle combine blends to OR(X,Y) if the relevant elements are known zero (REAPPLIED)
This allows us to remove the (depth violating) code in getFauxShuffleMask where we were combining the OR(SHUFFLE,SHUFFLE) shuffle inputs as well, and not just the OR().
This is a minor step toward being able to shuffle combine from/to SELECT/BLENDV as a faux shuffle.
Reapplied with fixed signed/unsigned comparisons.
Currently, instruction level fast math flags are not considered when
generating patterns for the machine combiner.
This currently leads to some missed opportunities to generate FMAs in
combination with `#pragma clang fp contract (fast)`.
For example, when building the example below with -O3 for AArch64, no
FMADD is generated. If built with -O2 and the DAGCombiner is used
instead of the MachineCombiner for FMAs, an FMADD is generated.
With this patch, the same code is generated in both cases.
float madd_contract(float a, float b, float c) {
#pragma clang fp contract (fast)
return (a * b) + c;
}
Reviewed By: dmgreen
Differential Revision: https://reviews.llvm.org/D84930
For FP_TO_INT and INT_TO_FP lowering, we have direct-move and
non-direct-move methods. But they share some conversion logic, so we can
reduce redundant code by introducing new methods.
Reviewed By: steven.zhang
Differential Revision: https://reviews.llvm.org/D81818
Test function mask_cmp_128 failed during ISEL
LLVM ERROR: Cannot select: t37: v8i1 = X86ISD::KSHIFTL t48, TargetConstant:i8<4>
due to v8i1 only available under AVX512DQ.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D84922
When scavenging consider the sub-register of the source operand
to determine the bank of a candidate register (not just sub0).
Without this it is possible to introduce an infinite loop,
e.g. $sgpr15_sgpr16_sgpr17 can be assigned for a conflict between
$sgpr0 and SGPR_96:sub1.
Reviewed By: rampitec
Differential Revision: https://reviews.llvm.org/D84910
Adds the function createMCInst() to MCContext that creates a MCInst using
a typed bump alloctor.
MCInst contains a SmallVector<MCOperand, 8>. The SmallVector is POD only
for <= 8 operands. The default untyped bump pointer allocator of MCContext
does not delete the MCInst, so if the SmallVector grows, it's a leak.
This fixes https://bugs.llvm.org/show_bug.cgi?id=46900.
GlobalISel is the default ISel for aarch64 at -O0. Prior to D78465, GlobalISel
didn't have support for dealing with address-of-global lowerings, so it fell
back to SelectionDAGISel.
HWASan Globals require special handling, as they contain the pointer tag in the
top 16-bits, and are thus outside the code model. We need to generate a `movk`
in the instruction sequence with a G3 relocation to ensure the bits are
relocated properly. This is implemented in SelectionDAGISel, this patch does
the same for GlobalISel.
GlobalISel and SelectionDAGISel differ in their lowering sequence, so there are
differences in the final instruction sequence, explained in
`tagged-globals.ll`. Both of these implementations are correct, but GlobalISel
is slightly larger code size / slightly slower (by a couple of arithmetic
instructions). I don't see this as a problem for now as GlobalISel is only on
by default at `-O0`.
Reviewed By: aemerson, arsenm
Differential Revision: https://reviews.llvm.org/D82615
VPSEL has slightly different semantics under tail predication (it can
end up selecting from Qn, Qm and Qd). We do not model that at the moment
so they block tail predicated loops from being formed.
This just converts them into a predicated VMOV instead (via a VORR),
allowing tail predication to happen whilst still modelling the original
behaviour of the input.
Differential Revision: https://reviews.llvm.org/D85110
Specified in https://github.com/WebAssembly/simd/pull/237, these
instructions load the first vector lane from memory and zero the other
lanes. Since these instructions are not officially part of the SIMD
proposal, they are only available on an opt-in basis via LLVM
intrinsics and clang builtin functions. If these instructions are
merged to the proposal, this implementation will change so that the
instructions will be generated from normal IR. At that point the
intrinsics and builtin functions would be removed.
This PR also changes the opcodes for the experimental f32x4.qfm{a,s}
instructions because their opcodes conflicted with those of the
v128.load{32,64}_zero instructions. The new opcodes were chosen to
match those used in V8.
Differential Revision: https://reviews.llvm.org/D84820
Fixes test-suite compile failure caused by 8dfb5d7.
While I'm in the area, add some more test coverage to related
operations, to make sure we aren't missing any other patterns.
Instructions should not be scheduled across ENDBR instructions, as this would result in the ENDBR being displaced, breaking the parity needed for the Indirect Branch Tracking feature of CET.
Currently, the X86IndirectBranchTracking pass is later than the instruction scheduling in the pipeline, what causes the bug to be unnoticeable and very hard (if not unfeasible) to be triggered while compiling C files with the standard LLVM setup. Yet, for correctness and to prevent issues in future changes, the compiler should prevent the such scheduling.
Differential Revision: https://reviews.llvm.org/D84862
This allows us to remove the (depth violating) code in getFauxShuffleMask where we were combining the OR(SHUFFLE,SHUFFLE) shuffle inputs as well, and not just the OR().
This is a minor step toward being able to shuffle combine from/to SELECT/BLENDV as a faux shuffle.
This adds an isel pattern and special XOR8rr_NOREX instruction
to enable the use of h-registers for __builtin_parity. This avoids
a copy and a shift instruction. The NOREX instruction is in case
register allocation doesn't use the matching l-register for some
reason. If a R8-R15 register gets picked instead, we won't be
able to encode the instruction since an h-register can't be used
with a REX prefix.
Fixes PR46954
This patch stops unconditionally transforming FSUB(-0,X) into an FNEG(X) while building the DAG. There is also one small change to handle the new FSUB(-0,X) similarly to FNEG(X) in the AMDGPU backend.
Differential Revision: https://reviews.llvm.org/D84056
There were various hacks used to try to avoid making s1 SGPR vs. s1
VCC ambiguous after constraining the register before we had a strategy
to deal with this. This also attempted to handle undef operands, which
are now illegal gMIR.
We already do this on AVX (+ for ZERO_EXTEND_VECTOR_INREG), but this enables it for all SSE targets - we attempted something similar back at rL357057 but hit issues with the ZERO_EXTEND_VECTOR_INREG handling (PR41249).
I'm still looking at the vector-mul.ll regression - which is due to 32-bit targets performing the load as a f64, resulting in the shuffle combiner thinking it has to create a shuffle in the float domain.
Fixes a regression caused by D82439, in which IT blocks were no longer being
generated when -Oz is present. This was due to the CPSR register being marked as
dead, while this case was not accounted for.
Differential Revision: https://reviews.llvm.org/D83667
This is a refactor patch to prepare for adding the support for strict-fsetcc
in PowerPC backend. We want to move their definition into a uniform way so that,
we could add the strict node easier.
Reviewed By: shchenz
Differential Revision: https://reviews.llvm.org/D81712
If the upper bits of the __builtin_parity idiom are known to be
0 we were previously emitting an xor with 0 to get the parity flag.
But we can use cmp/test instead which may expose opportunities for
load folding or combining an AND.
For AMDGPU, vectors with elements < 32 bits should be indexed in
32-bit elements and the desired bits extracted from there. For
elements > 64-bits, these should be reduce to 64/32 elements to enable
the normal dynamic indexing paths.
In the dynamic index cases, this produces shorter code most of the
time. This does immediately regress the constant index cases, but this
should be fixed once we have the most basic of shift combines.
The element size > 64 case is pretty much ported from the exisiting
DAG implementation for extract element promote. The increasing element
size case is new.
These prefixes should override the default behavior and force a larger immediate size. I don't believe gas issues any warning if you use {disp8} when a 32-bit displacement is already required. And this patch doesn't either.
This completes the {disp8} and {disp32} support from PR46650.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D84793
Noticed while investigating combining from concatenated shuffle vectors, we weren't checking that PSHUFLW/PSHUFHW was legal - we were depending on lowering splitting to subvectors.
This adds sign/zero extending scalar loads/stores to the MVE
instructions added in D77813, allowing us to create up more post-inc
instructions. These are comparatively simple, compared to LDR/STR (which
may be better turned into an LDRD/LDM), but still require some additions
over MVE instructions. Because there are i12 and i8 variants of the
offset loads/stores dealing with different signs, we may need to convert
an i12 address to a i8 negative instruction. t2LDRBi12 can also be
shrunk to a tLDRi under the right conditions, so we need to be careful
with codesize too.
Differential Revision: https://reviews.llvm.org/D78625
Now we try to load and broadcast together for operand 1. Followed
by load and broadcast for operand 1. Previously we tried load
operand 1, load operand 1, broadcast operand 0, broadcast operand 1.
Now we have a single helper that tries load and broadcast for
one operand that we can just call twice.
SPE doesn't have a fsel instruction, so don't try to lower to it.
This fixes a "Cannot select: tN: f64 = PPCISD::FSEL tX, tY, tZ" error.
Reviewed By: #powerpc, lkail
Differential Revision: https://reviews.llvm.org/D77773
The patterns were incorrect copies from the FPU code, and are
unnecessary, since there's no extended load for SPE. Just let LLVM
itself do the work by marking it expand.
Reviewed By: #powerpc, lkail
Differential Revision: https://reviews.llvm.org/D78670
Change to expand all arguments and return values to i64 to follow ABI.
Update regression tests also.
Reviewed By: simoll
Differential Revision: https://reviews.llvm.org/D84581
Refer to LangRef http://llvm.org/docs/LangRef.html#llvm-masked-load-intrinsics
'llvm.masked.load/store.*’ intrinsics are overloaded intrinsic, which allow the
load/store data to be a vector of any integer, floating-point or pointer data type.
Therefore, allow pointer data type when checking 'isLegalMaskedLoadStore()'.
Reviewed By: paulwalker-arm
Differential Revision: https://reviews.llvm.org/D85045
Rather than hardcoding immediate values for 12 different combinations
in a nested pair of switches, we can perform the matched logic
operation on 3 magic constants to calculate the immediate.
Special thanks to this tweet https://twitter.com/rygorous/status/1187034321992871936
for making me realize I could do this.
Summary: This patch separates the Loop Peeling Utilities from Loop Unrolling.
The reason for this change is that Loop Peeling is no longer only being used by
loop unrolling; Patch D82927 introduces loop peeling with fusion, such that
loops can be modified to have to same trip count, making them legal to be
peeled.
Reviewed By: Meinersbur
Differential Revision: https://reviews.llvm.org/D83056
This patch implements the instruction definition and MC tests for the vector
string isolate instructions.
Differential Revision: https://reviews.llvm.org/D84197
Scheduler will try to retrieve the offset and base addr to determine if two
loads/stores are disjoint memory access. PowerPC failed to handle this for
frame index which will bring extra memory dependency for loads/stores.
Reviewed By: jji
Differential Revision: https://reviews.llvm.org/D84308
Similar to what was recently done to ParseATTOperand. Make
ParseIntelOperand directly responsible for adding to the operand
vector instead of returning the operand. Return a bool for error.
Remove ErrorOperand since it is no longer used.
findAllocaForValue uses AllocaForValue to cache resolved values.
The function is used only to resolve arguments of lifetime
intrinsic which usually are not fare for allocas. So result reuse
is likely unnoticeable.
In followup patches I'd like to replace the function with
GetUnderlyingObjects.
Depends on D84616.
Differential Revision: https://reviews.llvm.org/D84617
Fix for the issue raised in https://github.com/rust-lang/rust/issues/74632.
The current heuristic for inserting LFENCEs uses a quadratic-time algorithm. This can apparently cause substantial compilation slowdowns for building Rust projects, where functions > 5000 LoC are apparently common.
The updated heuristic in this patch implements a linear-time algorithm. On a set of benchmarks, the slowdown factor for the generated code was comparable (2.55x geo mean for the quadratic-time heuristic, vs. 2.58x for the linear-time heuristic). Both heuristics offer the same security properties, namely, mitigating LVI.
This patch also includes some formatting fixes.
Differential Revision: https://reviews.llvm.org/D84471
After the recent change to the tuning settings for pentium4 to improve our default 32-bit behavior, I've decided to see about implementing -mtune support. This way we could have a default architecture CPU of "pentium4" or "x86-64" and a default tuning cpu of "generic". And we could change our "pentium4" tuning settings back to what they were before.
As a step to supporting this, this patch separates all of the features lists for the CPUs into 2 lists. I'm using the Proc class and a new ProcModel class to concat the 2 lists before passing to the target independent ProcessorModel. Future work to truly support mtune would change ProcessorModel to take 2 lists separately. I've diffed the X86GenSubtargetInfo.inc file before and after this patch to ensure that the final feature list for the CPUs isn't changed.
Differential Revision: https://reviews.llvm.org/D84879
Avoid recursively calling copyPhysReg for AGPR handling. This was
dropping the necessary super register implicit defs to avoid liveness
verifier errors.
LLVM selection dag assumes "switch" indices are pointer sized, which causes problems for our 32-bit br_table. The new function ensures 32-bit operands don't get unnecessarily extended, and 64-bit operands get truncated.
Note that the changes to the existing test test exactly that: the addition of -NEXT in 2 places ensures no extension is inserted (which the test previously ignored) and that the wrap is present (previously omitted in wasm64 mode).
Differential Revision: https://reviews.llvm.org/D84705
We are using undef on the indirect move source subreg and then
using implicit super-reg. This creates a problem in RA when
Greedy decides to split the register. It reassigns the implicit
super-reg but does not bother to change undef source because
it is really does not matter. The fix is to stop lying to RA and
drop undef flag.
This has also hit a problem in SIFoldOperands as it can fold
immediate into an indirect move since there is no undef flag
anymore. That results in multiple test failures, so added the
check for this case.
Differential Revision: https://reviews.llvm.org/D84899
Get rid of all fixmes and base heuristic on `num-clustered-dwords`. The main intuition behind this is as
follows. The existing heuristic roughly summarizes as below:
* Assume, all the mem ops instructions participating in the clustering process, loads/stores same num bytes
* If num bytes loaded by each mem op is 4 bytes, then cluster at max 5 mem ops, that is at max 20 bytes
* If num bytes loaded by each mem op is 8 bytes, then cluster at max 3 mem ops, that is at max 24 bytes
* If num bytes loaded by each mem op is 16 bytes, then cluster at max 2 mem ops, that is at max 32 bytes
So, we need to make sure that the new heuristic do not completey deviate away from the above one, and it
properly handles both the sub-word loads and the wide loads.
Reviewed By: arsenm, rampitec
Differential Revision: https://reviews.llvm.org/D84354
We parse .arch so that some `.arch i386; .code32` code can assemble. It seems
that X86AsmParser does not do a good job tracking what features are needed to
assemble instructions. GNU as's x86 port supports a very wide range of .arch
operands. Ignore the operand for now.
Reviewed By: craig.topper
Differential Revision: https://reviews.llvm.org/D84900
The operand to these instructions is both input and output.
These are not yet emitted by the compiler and the assembler already
works fine, so can't test in this patch. But D75044 will use XPACI
and provide test coverage for this patch as well.
Differential Revision: https://reviews.llvm.org/D84298
Summary:
This patch implements -ffunction-sections on AIX.
This patch focuses on assembly generation.
Follow-on patch needs to handle:
1. -ffunction-sections implication for jump table.
2. Object file generation path and associated testing.
Differential Revision: https://reviews.llvm.org/D83875
As long as we can extract the lowest 128-bit subvector from the pre-truncated source vector, then we don't care what size it is.
The next stage will be to support non-zero extraction indices, as long as its still coming from the lowest 128-bit subvector.
When building code at -O0 We weren't falling back to DAG ISel correctly
when encountering alloca instructions with scalable vector types. This
is because the alloca has no operands that are scalable. I've fixed this by
adding a check in AArch64ISelLowering::fallBackToDAGISel for alloca
instructions with scalable types.
Differential Revision: https://reviews.llvm.org/D84746
Continue the change made to ParseATTOperand to take the vector by
reference. Let ParseMemOperand add its memory operand to the
vector and just return true/false to indicate error.
A '*' after the segment is equivalent to a '*' before the segment register. To make the AsmMatcher table work we need to place the '*' token into the operand vector before the full memory operand. To accomplish this I've modified some portions of operand parsing to expose the operand vector to ParseATTOperand so that the token can be pushed to the vector after parsing the segment register and before creating the memory operand using that segment register.
Fixes PR46879
Reviewed By: MaskRay
Differential Revision: https://reviews.llvm.org/D84895
Summary:
Some instructions have set the wrong [RM] flag, this patch is to fix it.
Instructions x(v|s)r(d|s)pi[zmp]? and fri[npzm] use fixed rounding
directions without referencing current rounding mode.
Also, the SETRNDi, SETRND, BCLRn, MTFSFI, MTFSB0, MTFSB1, MTFSFb,
MTFSFI, MTFSFI_rec, MTFSF, MTFSF_rec should also fix the RM flag.
Reviewed By: jsji
Differential Revision: https://reviews.llvm.org/D81360
I still think it's highly questionable that we have two intrinsics
with identical behavior and only vary by the name of the libcall used
if it happens to be lowered that way, but try to reduce the feature
delta between SDAG and GlobalISel for recently added intrinsics. I'm
not sure which opcode should be considered the canonical one, but
lower roundeven back to round.
Instead of never accepting v8f32/v4f64 FHADD/FHSUB if the input shuffle masks cross lanes, perform the matching and determine if the post shuffle mask simplifies to a 'whole lane shuffle' mask - in which case we are guaranteed to cheaply perform this as a VPERM2F128 shuffle.
MFMA instructions shall not be scheduled back to back
to avoid MAI SIMD stall. Tell post-RA schedule we would
prefer some other instruction instead.
Differential Revision: https://reviews.llvm.org/D84883
Adds frontend and backend options to enable and disable the
PowerPC paired vector memory operations added in ISA 3.1.
Instructions using these options will be added in subsequent patches.
Differential Revision: https://reviews.llvm.org/D83722
In future, we'd like to use the perfect-shuffle mechanism to deal with these
shuffle permutations. For now, this improves performance by avoiding the
super-expensive const-pool load + tbl instruction.
Differential Revision: https://reviews.llvm.org/D84866
Port the wide immediate case from AArch64DAGToDAGISel::SelectAddrModeXRO.
If we have a wide immediate which can't be represented in an add, we can end up
with code like this:
```
mov x0, imm
add x1, base, x0
ldr x2, [x1, 0]
```
If we use the [base, xN] addressing mode instead, we can produce this:
```
mov x0, imm
ldr x2, [base, x0]
```
This saves 0.4% code size on 7zip at -O3, and gives a geomean code size
improvement of 0.1% on CTMark.
Differential Revision: https://reviews.llvm.org/D84784
I never completed the work on the patches referenced by
f8bf7d7f42, but this was intended to
avoid folding immediate writes into m0 which the coalescer doesn't
understand very well. Relax this to allow simple SGPR immediates to
fold directly into VGPR copies. This pattern shows up routinely in
current GlobalISel code since nothing is smart enough to emit VGPR
constants yet.
When it was first created, CFGSort only made sure BBs in each
`MachineLoop` are sorted together. After we added exception support,
CFGSort now also sorts BBs in each `WebAssemblyException`, which
represents a `catch` block, together, and
`Region` class was introduced to be a thin wrapper for both
`MachineLoop` and `WebAssemblyException`.
But how we compute those loops and exceptions is different.
`MachineLoopInfo` is constructed using the standard loop computation
algorithm in LLVM; the definition of loop is "a set of BBs that are
dominated by a loop header and have a path back to the loop header". So
even if some BBs are semantically contained by a loop in the original
program, or in other words dominated by a loop header, if they don't
have a path back to the loop header, they are not considered a part of
the loop. For example, if a BB is dominated by a loop header but
contains `call abort()` or `rethrow`, it wouldn't have a path back to
the header, so it is not included in the loop.
But `WebAssemblyException` is wasm-specific data structure, and its
algorithm is simple: a `WebAssemblyException` consists of an EH pad and
all BBs dominated by the EH pad. So this scenario is possible: (This is
also the situation in the newly added test in cfg-stackify-eh.ll)
```
Loop L: header, A, ehpad, latch
Exception E: ehpad, latch, B
```
(B contains `abort()`, so it does not have a path back to the loop
header, so it is not included in L.)
And it is sorted in this order:
```
header
A
ehpad
latch
B
```
And when CFGStackify places `end_loop` or `end_try` markers, it
previously used `WebAssembly::getBottom()`, which returns the latest BB
in the sorted order, and placed the marker there. So in this case the
marker placements will be like this:
```
loop
header
try
A
catch
ehpad
latch
end_loop <-- misplaced!
B
end_try
```
in which nesting between the loop and the exception is not correct.
`end_loop` marker has to be placed after `B`, and also after `end_try`.
Maybe the fundamental way to solve this problem is to come up with our
own algorithm for computing loop region too, in which we include all BBs
dominated by a loop header in a loop. But this takes a lot more effort.
The only thing we need to fix is actually, `getBottom()`. If we make it
return the right BB, which means in case of a loop, the latest BB of the
loop itself and all exceptions contained in there, we are good.
This renames `Region` and `RegionInfo` to `SortRegion` and
`SortRegionInfo` and extracts them into their own file. And add
`getBottom` to `SortRegionInfo` class, from which it can access
`WebAssemblyExceptionInfo`, so that it can compute a correct bottom
block for loops.
Reviewed By: dschuff
Differential Revision: https://reviews.llvm.org/D84724
I've been looking at missed vectorizations in one codebase.
One particular thing that stands out is that some of the loops
reach vectorizer in a rather mangled form, with weird PHI's,
and some of the loops aren't even in a rotated form.
After taking a more detailed look, that happened because
the loop's headers were too big by then. It is evident that
SimplifyCFG's common code hoisting transform is at fault there,
because the pattern it handles is precisely the unrotated
loop basic block structure.
Surprizingly, `SimplifyCFGOpt::HoistThenElseCodeToIf()` is enabled
by default, and is always run, unlike it's friend, common code sinking
transform, `SinkCommonCodeFromPredecessors()`, which is not enabled
by default and is only run once very late in the pipeline.
I'm proposing to harmonize this, and disable common code hoisting
until //late// in pipeline. Definition of //late// may vary,
here currently i've picked the same one as for code sinking,
but i suppose we could enable it as soon as right after
loop rotation happens.
Experimentation shows that this does indeed unsurprizingly help,
more loops got rotated, although other issues remain elsewhere.
Now, this undoubtedly seriously shakes phase ordering.
This will undoubtedly be a mixed bag in terms of both compile- and
run- time performance, codesize. Since we no longer aggressively
hoist+deduplicate common code, we don't pay the price of said hoisting
(which wasn't big). That may allow more loops to be rotated,
so we pay that price. That, in turn, that may enable all the transforms
that require canonical (rotated) loop form, including but not limited to
vectorization, so we pay that too. And in general, no deduplication means
more [duplicate] instructions going through the optimizations. But there's still
late hoisting, some of them will be caught late.
As per benchmarks i've run {F12360204}, this is mostly within the noise,
there are some small improvements, some small regressions.
One big regression i saw i fixed in rG8d487668d09fb0e4e54f36207f07c1480ffabbfd, but i'm sure
this will expose many more pre-existing missed optimizations, as usual :S
llvm-compile-time-tracker.com thoughts on this:
http://llvm-compile-time-tracker.com/compare.php?from=e40315d2b4ed1e38962a8f33ff151693ed4ada63&to=c8289c0ecbf235da9fb0e3bc052e3c0d6bff5cf9&stat=instructions
* this does regress compile-time by +0.5% geomean (unsurprizingly)
* size impact varies; for ThinLTO it's actually an improvement
The largest fallout appears to be in GVN's load partial redundancy
elimination, it spends *much* more time in
`MemoryDependenceResults::getNonLocalPointerDependency()`.
Non-local `MemoryDependenceResults` is widely-known to be, uh, costly.
There does not appear to be a proper solution to this issue,
other than silencing the compile-time performance regression
by tuning cut-off thresholds in `MemoryDependenceResults`,
at the cost of potentially regressing run-time performance.
D84609 attempts to move in that direction, but the path is unclear
and is going to take some time.
If we look at stats before/after diffs, some excerpts:
* RawSpeed (the target) {F12360200}
* -14 (-73.68%) loops not rotated due to the header size (yay)
* -272 (-0.67%) `"Number of live out of a loop variables"` - good for vectorizer
* -3937 (-64.19%) common instructions hoisted
* +561 (+0.06%) x86 asm instructions
* -2 basic blocks
* +2418 (+0.11%) IR instructions
* vanilla test-suite + RawSpeed + darktable {F12360201}
* -36396 (-65.29%) common instructions hoisted
* +1676 (+0.02%) x86 asm instructions
* +662 (+0.06%) basic blocks
* +4395 (+0.04%) IR instructions
It is likely to be sub-optimal for when optimizing for code size,
so one might want to change tune pipeline by enabling sinking/hoisting
when optimizing for size.
Reviewed By: mkazantsev
Differential Revision: https://reviews.llvm.org/D84108
Summary:
PPC only supports the instruction selection for v16i8, v8i16, v4i32,
v2i64, v4f32 and v2f64 for ISD::SETCC, don't support the v1i128, so
v1i128 for ISD::SETCC will crash.
This patch is to set v1i128 to expand to avoid crash.
Reviewed By: steven.zhang
Differential Revision: https://reviews.llvm.org/D84238
expressions.
Also "fix" the longstanding bug where the computed size depends on the
order of the visitation. We could try to predict the allocation order
used by legalization, but it would never be 100% perfect. Until we
start fixing the addresses somehow (or have a more reliable allocation
scheme later), just try to compute the size based on the worst case
padding.
This patch uses the feature added in D79162 to fix the cost of a
sext/zext of a masked load, or a trunc for a masked store.
Previously, those were considered cheap or even free, but it's
not the case as we cannot split the load in the same way we would for
normal loads.
This updates the costs to better reflect reality, and adds a test for it
in test/Analysis/CostModel/ARM/cast.ll.
It also adds a vectorizer test that showcases the improvement: in some
cases, the vectorizer will now choose a smaller VF when
tail-predication is enabled, which results in better codegen. (Because
if it were to use a higher VF in those cases, the code we see above
would be generated, and the vmovs would block tail-predication later in
the process, resulting in very poor codegen overall)
Original Patch by Pierre van Houtryve
Differential Revision: https://reviews.llvm.org/D79163
Matt Arsenault