This is restricted to single use loads, which if we fold to sextloads we can
find more optimal addressing modes on AArch64.
This also fixes an overload the MachineFunction::getMachineMemOperand() method
which was incorrectly using the MF alignment instead of the MMO alignment.
Differential Revision: https://reviews.llvm.org/D85966
By detecting this sign extend pattern early, we can uncover opportunities for
more optimizations.
Differential Revision: https://reviews.llvm.org/D85965
We weren't looking through the parameters on calls at all.
E.g., say you had
```
declare i32 @zext(i32 zeroext %x)
...
%y = call i32 @zext(i32 %something)
...
```
At the point of the call, we wouldn't know that the %something should have the
zeroext attribute.
This sets flags in about the same way as
TargetLoweringBase::ArgListEntry::setAttributes.
Differential Revision: https://reviews.llvm.org/D86125
Theory was that we should never reach a non-type unit (eg: type in an
anonymous namespace) when we're already in the invalid "encountered an
address-use, so stop emitting types for now, until we throw out the
whole type tree to restart emitting in non-type unit" state. But that's
not the case (prior commit cleaned up one reason this wasn't exposed
sooner - but also makes it easier to test/demonstrate this issue)
This reads more like what you'd expect the DWARF to look like (from the
lexical order of C++ - template parameters come before members, etc),
and also happens to make it easier to tickle (& thus test) a bug related
to type units and Split DWARF I'm about to fix.
Some of the lower implementations were relying on this, however the
type was not set depending on which form .lower* helper form you were
using. For instance, if you used an unconditonal lower(), the type was
never set. Most of the lower actions do not benefit from a type
parameter, and just expand in terms of the original operation's types.
However, some lowerings could benefit from an additional type hint to
combine a promotion and an expansion. An example of this is for
add/sub sat. The DAG integer legalization tries to use smarter
expansions directly when promoting the integer type, and doesn't
always produce the same instruction with a wider type.
Treat this as an optional hint argument, that only means something for
specific lower actions. It may be useful to generalize this mechanism
to pass a full list of type indexes and desired types, but I haven't
run into a case like that yet.
The "isa" checks were less constrained because they allow
target constants, but the later matching code would bail
out on those anyway, so this should be slightly more
efficient.
(Forgot to land this a couple of weeks back.)
In a recent series of changes, I've introduced support for using the respective operand bundle kinds on the statepoint. At the moment, code supports either/or, but there's no need to keep the old support around. For the moment, I am simply changing the specification and verifier to require zero length argument sets in the intrinsic.
The intrinsic itself is experimental. Given that, there's no forward serialization needed. The in tree uses and generation have already been updated to use the new operand bundle based forms, the only folks broken by the change will be those with frontends generating statepoints directly and the updates should be easy.
Why not go ahead and just remove the arguments entirely? Well, I plan to. But while working on this I've found that almost all of the arguments to the statepoint can be expressed via operand bundles or attributes. Given that, I'm planning a radical simplification of the arguments and figured I'd do one update not several small ones.
Differential Revision: https://reviews.llvm.org/D80892
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
These should really match either G_BUILD_VECTOR or
G_BUILD_VECTOR_TRUNC, but there doesn't seem to be an existing
mechanism for matching alternative opcodes. There is GIM_SwitchOpcode,
but it seems to assume it's oly only used for matcher optimization.
I could also omit any opcode check and rely on the matcher directly
checking the opcode, but the table optimizer currently assumes there
has to be an opcode check.
Also doesn't try to handle undef elements like the DAG version.
Extend FixupStatepointCallerSaved pass with ability to spill
statepoint GC pointer arguments (optionally allowing them on CSRs).
Special handling is required for invoke statepoints, because at MI
level single landing pad may be shared by multiple statepoints, so
we must ensure we spill landing pad's live-ins into the same stack
slots.
Full statepoint refactoring change set is available at D81603.
Reviewed By: skatkov
Differential Revision: https://reviews.llvm.org/D81647
This patch changes SplitVecOp_EXTRACT_VECTOR_ELT to work correctly
for scalable vectors and also fixes an a bug in DAGCombiner where
the scalable property is dropped in visitTRUNCATE when attempting
to fold an extract + a truncate.
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D85754
In DAGTypeLegalizer::GenWidenVectorStores the algorithm assumes it only
ever deals with fixed width types, hence the offsets for each individual
store never take 'vscale' into account. I've changed the main loop in
that function to use TypeSize instead of unsigned for tracking the
remaining store amount and offset increment. In addition, I've changed
the loop to use the new IncrementPointer helper function for updating
the addresses in each iteration, since this handles scalable vector
types.
Whilst fixing this function I also fixed a minor issue in
IncrementPointer whereby we were not adding the no-unsigned-wrap flag
for the add instruction in the same way as the fixed width case does.
Also, I've added a report_fatal_error in GenWidenVectorTruncStores,
since this code currently uses a sequence of element-by-element scalar
stores.
I've added new tests in
CodeGen/AArch64/sve-intrinsics-stores.ll
CodeGen/AArch64/sve-st1-addressing-mode-reg-imm.ll
for the changes in GenWidenVectorStores.
Differential Revision: https://reviews.llvm.org/D84937
In narrowExtractedVectorLoad there is an optimisation that tries to
combine extract_subvector with a narrowing vector load. At the moment
this produces warnings due to the incorrect calls to
getVectorNumElements() for scalable vector types. I've got this
working for scalable vectors too when the extract subvector index
is a multiple of the minimum number of elements. I have added a
new variant of the function:
MachineFunction::getMachineMemOperand
that copies an existing MachineMemOperand, but replaces the pointer
info with a null version since we cannot currently represent scaled
offsets.
I've added a new test for this particular case in:
CodeGen/AArch64/sve-extract-subvector.ll
Differential Revision: https://reviews.llvm.org/D83950
This is mostly a straight port from SelectionDAG. We re-use the actual bit-test
analysis part from SwitchLoweringUtils, which was factored out earlier to
support jump-tables.
Differential Revision: https://reviews.llvm.org/D85233
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
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
Allow the GNU .debug_macro extension to be emitted for DWARF versions
earlier than 5. The extension is basically what became DWARF 5's format,
except that a DW_AT_GNU_macros attribute is emitted, and some entries
like the strx entries are missing. In this patch I emit GNU's indirect
entries, which are the same as DWARF 5's strp entries.
This patch adds the extension behind a hidden LLVM flag,
-use-gnu-debug-macro. I would later want to enable it by default when
tuning for GDB and targeting DWARF versions earlier than 5.
The size of a Clang 8.0 binary built with RelWithDebInfo and the flags
"-gdwarf-4 -fdebug-macro" reduces from 1533 MB to 1349 MB with
.debug_macro (compared to 1296 MB without -fdebug-macro).
Reviewed By: SouraVX, dblaikie
Differential Revision: https://reviews.llvm.org/D82975
Broken out from a review comment on D82975. This is an NFC expect for
that the Macinfo macro string is now emitted using a single emitBytes()
invocation, so it can be done using a single string directive.
Reviewed By: dblaikie
Differential Revision: https://reviews.llvm.org/D83557
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.
When the result type of insertelement needs to be split,
SplitVecRes_INSERT_VECTOR_ELT will try to store the vector to a
stack temporary, store the element at the location of the stack
temporary plus the index, and reload the Lo/Hi parts.
This patch does the following to ensure this works for scalable vectors:
- Sets the StackID with getStackIDForScalableVectors() in CreateStackTemporary
- Adds an IsScalable flag to getMemBasePlusOffset() and scales the
offset by VScale when this is true
- Ensures the immediate is clamped correctly by clampDynamicVectorIndex
so that we don't try to use an out of range index
Reviewed By: david-arm
Differential Revision: https://reviews.llvm.org/D84874
Move the Dwarf version checks that determine if the .debug_macro section
should be emitted, into a DwarfDebug member. This is a preparatory
refactoring for allowing the GNU .debug_macro extension, which is a
precursor to the DWARF 5 format, to be emitted by LLVM for earlier DWARF
versions.
Reviewed By: dblaikie
Differential Revision: https://reviews.llvm.org/D82971
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
We skip debug instructions in RDA so we cannot attempt to look them
up in our instruction map without causing a crash. But some of the
methods select the last instruction in the block and this
instruction may be a debug instruction... So, use getLastNonDebugInstr
instead of calling back on a MachineBasicBlock.
MachineBasicBlock iterators have also been updated to use
instructionsWithoutDebug so we can avoid the manual checks for debug
instructions.
Differential Revision: https://reviews.llvm.org/D85658
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
SplitKit forms invalid COPY subreg bundles without a leading
BUNDLE instruction. That manifests itself in post-RA scheduler
counting instruction and asserting on "Instruction count mismatch".
The bundle shall be undone by VirtRegRewriter::expandCopyBundle(),
but it does not because VirtRegRewriter::handleIdentityCopy() can
turn COPY bundle into a KILL bundle.
Process KILLs as well.
Differential Revision: https://reviews.llvm.org/D85484
This patch adds the missing information to the LF_BUILDINFO record, which allows for rebuilding a .CPP without any external dependency but the .OBJ itself (other than the compiler).
Some external tools that we are using (Recode, Live++) are extracting the information to reproduce a build without any knowledge of the build system. The LF_BUILDINFO stores a full path to the compiler, the PWD (CWD at program startup), a relative or absolute path to the TU, and the full CC1 command line. The command line needs to be freestanding (not depend on any environment variables). In the same way, MSVC doesn't store the provided command-line, but an expanded version (somehow their equivalent of CC1) which is also freestanding.
For more information see PR36198 and D43002.
Differential Revision: https://reviews.llvm.org/D80833
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.
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
As noticed on D66004, scalarization of an expandload with a constant mask as a chain of irregular loads+inserts makes it tricky to optimize before lowering, resulting in difficulties in merging loads etc.
This patch instead scalarizes the expansion to a build_vector(load0, load1, undef, load2,....) style pattern and then performs a blend shuffle with the pass through vector. This allows us to more easily make use of all the build_vector combines, merging of consecutive loads etc.
Differential Revision: https://reviews.llvm.org/D85416
This also fixes the condition in the assertion in
DwarfCompileUnit::getLabelBegin() because it checked something unrelated
to the returned value.
Differential Revision: https://reviews.llvm.org/D85437
These aren't the canonical forms we'd get from InstCombine, but
we do have X86 tests for them. Recognizing them is pretty cheap.
While there make use of APInt:isSignedMinValue/isSignedMaxValue
instead of creating a new APInt to compare with. Also use
SelectionDAG::getAllOnesConstant helper to hide the all ones
APInt creation.
Follow-up to D82716 / rGea71ba11ab11
We do not have the fabs removal fold in IR yet for the case
where the sqrt operand is repeated, so that's another potential
improvement.
As mentioned on D85463, we should be using SimplifyMultipleUseDemandedBits (which is the default fallback).
The minor regression in illegal-bitfield-loadstore.ll will be addressed properly by D77804.
This removes members of the DIEUnit class which were used only in unit
tests. Note also that child classes shadowed some of these methods,
namely, getDwarfVersion() was overridden in DwartfUnit and getLength()
was overridden in DwarfCompileUnit.
Differential Revision: https://reviews.llvm.org/D85436
If it is load cluster, we don't need to create the dependency edges(SUb->reg) from SUb to SUa
as they both depend on the base register "reg"
+-------+
+----> reg |
| +---+---+
| ^
| |
| |
| |
| +---+---+
| | SUa | Load 0(reg)
| +---+---+
| ^
| |
| |
| +---+---+
+----+ SUb | Load 4(reg)
+-------+
But if it is store cluster, we need to create it as follow shows to avoid the instruction store
depend on scheduled in-between SUb and SUa.
+-------+
+----> reg |
| +---+---+
| ^
| | Missing +-------+
| | +-------------------->+ y |
| | | +---+---+
| +---+-+-+ ^
| | SUa | Store x 0(reg) |
| +---+---+ |
| ^ |
| | +------------------------+
| | |
| +---+--++
+----+ SUb | Store y 4(reg)
+-------+
Reviewed By: evandro, arsenm, rampitec, foad, fhahn
Differential Revision: https://reviews.llvm.org/D72031
One of the callers only wants the condition, but the vselect can
be simplified by getNode making it hard or impossible to retrieve
the condition.
Instead, return the condition and make the other 2 callers
responsible for creating the vselect node using the condition.
Rename the function to WidenVSELECTMask accordingly.
Differential Revision: https://reviews.llvm.org/D85468
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 a function is in a unique section, putting all jump tables in
.rodata will prevent functions that have a jump table to get
garbage collect by the linker.
Therefore, we need to put jump table into a unique section as well.
Reviewed By: Xiangling_L
Differential Revision: https://reviews.llvm.org/D84761
Add given input and mark it as tied.
Doesn't create additional copy compared to
matching input constraint to virtual register.
Differential Revision: https://reviews.llvm.org/D85122
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
This patch changes the functionality of AsmPrinter to name the basic block end labels as LBB_END${i}_${j}, with ${i} being the identifier for the function and ${j} being the identifier for the basic block. The new naming scheme is consistent with how basic block labels are named (.LBB${i}_{j}), and how function end symbol are named (.Lfunc_end${i}) and helps to write stronger tests for the upcoming patch for BB-Info section (as proposed in https://lists.llvm.org/pipermail/llvm-dev/2020-July/143512.html). The end label is used with basicblock-labels (BB-Info section in future) and basicblock-sections to compute the size of basic blocks and basic block sections, respectively. For BB sections, the section containing the entry basic block will not have a BB end label since it already gets the function end-label.
This label is cached for every basic block (CachedEndMCSymbol) like the label for the basic block (CachedMCSymbol).
Differential Revision: https://reviews.llvm.org/D83885
Implement proper folding of statepoint meta operands (deopt and GC)
when statepoint uses tied registers.
For deopt operands it is just about properly preserving tiedness
in new instruction.
For tied GC operands folding is a little bit more tricky.
We can fold tied GC operands only from InlineSpiller, because it knows
how to properly reload tied def after it was turned into memory operand.
Other users (e.g. peephole) cannot properly fold such operands as they
do not know how (or when) to reload them from memory.
We do this by un-tieing operand we want to fold in InlineSpiller
and allowing to fold only untied operands in foldPatchpoint.
We currently don't do anything to fold any_extend vector loads as no target has such an instruction.
Instead I've added support for folding to a zextload, SimplifyDemandedBits does a good job of adjusting the zext(truncate(()) stages as required later on.
We still need the custom scalar extload handling instead of using the tryToFoldExtOfLoad helper as it has different legality tests - we can probably tweak that to reduce most of the code duplication.
Fixes the regression I mentioned in rG99a971cadff7
Differential Revision: https://reviews.llvm.org/D85129
Currently, we only test the `--stackmap` option here:
https://github.com/llvm/llvm-project/blob/master/llvm/test/Object/stackmap-dump.test
it uses a precompiled MachO binary currently and I've found no tests for this option for ELF.
The implementation also has issues. For example, it might assert on a wrong version
of the .llvm-stackmaps section. Or it might crash on an empty or truncated section.
This patch introduces a new tools/llvm-readobj/ELF test file as well as implements a few
basic checks to catch simple crashes/issues
It also eliminates `unwrapOrError` calls in `printStackMap()`.
Differential revision: https://reviews.llvm.org/D85208
The sorting is needed, because reaching defs are (logically) ordered,
but are not collected in that order. This change will break up the
single call to std::sort into a series of smaller sorts, each of which
should use a cheaper comparison function than the original.
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.
This corresponds with the SelectionDAGISel change in D84056.
Also, rename some poorly named tests in CodeGen/X86/fast-isel-fneg.ll with NFC.
Differential Revision: https://reviews.llvm.org/D85149
This one is pretty easy and shrinks the list of unhandled
intrinsics. I'm not sure how relevant the insert point is. Using the
insert position of EntryBuilder will place this after
constants. SelectionDAG seems to end up emitting these after argument
copies and before anything else, but I don't think it really
matters. This also ends up emitting these in the opposite order from
SelectionDAG, but I don't think that matters either.
This also needs a fix to stop the later passes dropping this as a dead
instruction. DeadMachineInstructionElim's version of isDead special
cases LOCAL_ESCAPE for some reason, and I'm not sure why it's excluded
from MachineInstr::isLabel (or why isDead doesn't check it).
I also noticed DeadMachineInstructionElim never considers inline asm
as dead, but GlobalISel will drop asm with no constraints.
This patch stops unconditionally transforming FSUB(-0, X) into an FNEG(X) while building the MIR.
This corresponds with the SelectionDAGISel change in D84056.
Differential Revision: https://reviews.llvm.org/D85139
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
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
Part of https://bugs.llvm.org/show_bug.cgi?id=41734
LTO can drop externally available definitions. Such AssociatedSymbol is
not associated with a symbol. ELFWriter::writeSection() will assert.
Allow a SHF_LINK_ORDER section to have sh_link=0.
We need to give sh_link a syntax, a literal zero in the linked-to symbol
position, e.g. `.section name,"ao",@progbits,0`
Reviewed By: pcc
Differential Revision: https://reviews.llvm.org/D72899
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
Use pad with undef and unmerge with unused results. This is annoyingly
similar to several other places in LegalizerHelper, but they're all
slightly different.
The patch restricts DIEDelta::SizeOf() to accept only DWARF forms that
are actually used in the LLVM codebase. This should make the use of the
class more explicit and help to avoid issues similar to fixed in D83958
and D84094.
Differential Revision: https://reviews.llvm.org/D84095
DIELabel can emit only 32- or 64-bit values, while it was created in
some places with DW_FORM_udata, which implies emitting uleb128.
Nevertheless, these places also expected to emit U32 or U64, but just
used a misleading DWARF form. The patch updates those places to use more
appropriate DWARF forms and restricts DIELabel::SizeOf() to accept only
forms that are actually used in the LLVM codebase.
Differential Revision: https://reviews.llvm.org/D84094
DIELocList is used with a limited number of DWARF forms, see the only
place where it is instantiated, DwarfCompileUnit::addLocationList().
The patch marks the unexpected execution path in DIELocList::SizeOf()
as unreachable, to reduce ambiguity.
Differential Revision: https://reviews.llvm.org/D84092
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.
Try to be more consistent with the SDLoc param in the TargetLowering methods.
This also exposes an issue where we were passing a SDNode as a SDLoc, relying on the implicit SDLoc(SDNode) constructor.
D68049 created options for basic block sections: -fbasic-block-sections=,
-funique-basic-block-section-names. Rename options in llc and lld (--lto-)
to be consistent. Specifically,
+ Rename basicblock-sections to basic-block-sections
+ Rename unique-bb-section-names to unique-basic-block-section-names
Differential Revision: https://reviews.llvm.org/D84462
https://reviews.llvm.org/D84909
Patch adds two new GICombinerRules, one for G_INTTOPTR and one for
G_PTRTOINT. The G_INTTOPTR elides ptr2int(int2ptr(x)) to a copy of x, if
the cast is within the same address space. The G_PTRTOINT elides
int2ptr(ptr2int(x)) to a copy of x. Patch additionally adds new combiner
tests for the AArch64 target to test these new combiner rules.
Patch by mkitzan
Just the obvious implementation that rewrites the result type. Also fix
warning from EXTRACT_SUBVECTOR legalization that triggers on the test.
Differential Revision: https://reviews.llvm.org/D84706
This fixes an assertion failure that was being triggered in
SelectionDAG::getZeroExtendInReg(), where it was trying to extend the <2xi32>
to i64 (which should have been <2xi64>).
Fixes: rdar://66016901
Differential Revision: https://reviews.llvm.org/D84884
In cases where the alignment of the datatype is smaller than
expected by the instruction, the address is aligned. The aligned
address is used for the load, but wasn't used for the store
conditional, which resulted in a run-time alignment exception.
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
Summary:
In the phi-node-elimination pass, we set the killed flag incorrectly.
When we eliminate the PHI node, we replace the PHI with a copy for the
incoming value.
Before this patch, we will set incoming value as killed(PHICopy). And
we will remove the killed flag from last using incoming value(OldKill).
This is correct, only if the new PHICopy is after the OldKill.
Reviewed By: bjope
Differential Revision: https://reviews.llvm.org/D80886
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.
This change is mechanical, it just removes the restriction and updates tests. The key building blocks were submitted in 31342eb and 8fe2abc.
Note that this (and preceeding changes) entirely subsumes D83965. I did includes a couple of it's tests.
From the codegen changes, an interesting observation: this doesn't actual reduce spilling, it just let's the register allocator do it's job. That results in a slightly different overall result which has both pros and cons over the eager spill lowering. (i.e. We'll have some perf tuning to do once this is stable.)
Change the way we track how a particular pointer was relocated at a statepoint in selection dag. Previously, we used an optional<location> for the spill lowering, and a block local Register for the newly introduced vreg lowering. Combine all three lowerings (norelocate, spill, and vreg) into a single helper class, and keep a single copy of the information.
This is submitted separately as it really does make the code more readible on it's own, but the indirect motivation is to move vreg tracking from StatepointLowering to FunctionLoweringInfo. This is the last piece needed to support cross block relocations with vregs; that will follow in a separate (non-NFC) patch.
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
This builds on 3da1a96 on the path towards supporting invokes and cross block relocations. The actual change attempts to be NFC, but does fail in one corner-case explained below.
The change itself is fairly mechanical. Rather than remember SDValues - which are inherently block local - immediately produce a virtual register copy and remember that.
Once this lands, we'll update the FunctionLoweringInfo::StatepointSpillMap map to allow register based lowerings, delete VirtRegs from StatepointLowering, and drop the restriction against cross block relocations. I deliberately separate the semantic part into it's own change for easy of understanding and fault isolation.
The corner-case which isn't quite NFC is that the old implementation implicitly CSEd gc.relocates of the same SDValue regardless of type. The new implementation still only relocates once, but it produces distinct vregs for the bitcast and it's source, whereas SelectionDAG's generic CSE was able to remove the bitcast in the old implementation. Note that the final assembly doesn't change (at least in the test), as our MI level optimizations catch the duplication.
I assert that this is an uninteresting corner-case. It's functionally correct, and if we find a case where this influences performance, we should really be canonicalizing types to i8* at the IR level.
Differential Revision: https://reviews.llvm.org/D84692
Summary:
When doing MachineVerifier for LiveVariables, the MachineVerifier pass
will calculate the LiveVariables, and compares the result with the
result livevars pass gave. If they are different, verifyLiveVariables()
will give error.
But when we calculate the LiveVariables in MachineVerifier, we don't
consider the PHI node, while livevars considers.
This patch is to fix above bug.
Reviewed By: bjope
Differential Revision: https://reviews.llvm.org/D80274
In MachineCopyPropagation::BackwardPropagatableCopy(),
a check is added for multiple destination registers.
The copy propagation is avoided if the copied destination register
is the same register as another destination on the same instruction.
A new test is added. This used to fail on ARM like this:
error: unpredictable instruction, RdHi and RdLo must be different
umull r9, r9, lr, r0
Reviewed By: lkail
Differential Revision: https://reviews.llvm.org/D82638
I have added tests to:
CodeGen/AArch64/sve-intrinsics-int-arith.ll
for doing simple integer add operations on tuple types. Since these
tests introduced new warnings due to incorrect use of
getVectorNumElements() I have also fixed up these warnings in the
same patch. These fixes are:
1. In narrowExtractedVectorBinOp I have changed the code to bail out
early for scalable vector types, since we've not yet hit a case that
proves the optimisations are profitable for scalable vectors.
2. In DAGTypeLegalizer::WidenVecRes_CONCAT_VECTORS I have replaced
calls to getVectorNumElements with getVectorMinNumElements in cases
that work with scalable vectors. For the other cases I have added
asserts that the vector is not scalable because we should not be
using shuffle vectors and build vectors in such cases.
Differential revision: https://reviews.llvm.org/D84016
In DAGTypeLegalizer::SplitVecOp_EXTRACT_SUBVECTOR I have replaced
calls to getVectorNumElements with getVectorMinNumElements, since
this code path works for both fixed and scalable vector types. For
scalable vectors the index will be multiplied by VSCALE.
Fixes warnings in this test:
sve-sext-zext.ll
Differential revision: https://reviews.llvm.org/D83198
The refactoring encapsulates frequency calculation in MachineBlockFrequencyInfo,
and renames the API to clarify its motivation. It should clarify
frequencies may not be reset 'freely' by users of the analysis, as the
API serves as a partial update to avoid a full analysis recomputation.
Differential Revision: https://reviews.llvm.org/D84427
I think these were added as a workaround for SelectionDAG lacking half
legalization support in the past. I think they should probably be
removed from the IR, but clang does still have a target control to
emit these instead of the native half fpext/fptrunc.
The move constructor of MachineModuleInfo currently does not copy the
MachineFunctions map. This commit fixes this issue.
Patch by Sridhar Gopinath. Thanks!
Differential Revision: https://reviews.llvm.org/D84274
Speed up the method RequiresStackProtector by checking the intrinsic
value of the call. The original code calls getName() that returns an
allocating std::string on each check. This change removes about 96072
std::string instances when compiling sqlite3.c; The function was
discovered with a Facebook-internal performance tool.
Differential Revision: https://reviews.llvm.org/D84620
I have introduced a new TargetFrameLowering query function:
isStackIdSafeForLocalArea
that queries whether or not it is safe for objects of a given stack
id to be bundled into the local area. The default behaviour is to
always bundle regardless of the stack id, however for AArch64 this is
overriden so that it's only safe for fixed-size stack objects.
There is future work here to extend this algorithm for multiple local
areas so that SVE stack objects can be bundled together and accessed
from their own virtual base-pointer.
Differential Revision: https://reviews.llvm.org/D83859
Store Addr and Store Addr+8 are clusterable pair. They have memory(ctrl) dependency on different loads.
Current implementation will put these two stores into different group and miss to cluster them.
Reviewed By: evandro
Differential Revision: https://reviews.llvm.org/D84139
Summary:
In parallelizeChainedStores, a TokenFactor was created with the size greater than 3000.
We found that DAGCombiner::visitTokenFactor will consume a huge amount of time on
such nodes. Since the number of operands already exceeds TokenFactorInlineLimit, we propose
to give up simplification with the consideration of compile time.
Reviewers:
@spatel, @arsenm
Differential Revision:
https://reviews.llvm.org/D84204
(Disabled under flag for the moment)
This is part of a larger project wherein we are finally integrating lowering of gc live operands with the register allocator. Today, we force spill all operands in SelectionDAG. The code to do so is distinctly non-optimal. The approach this patch is working towards is to instead lower the relocations directly into the MI form, and let the register allocator pick which ones get spilled and which stack slots they get spilled to. In terms of performance, the later part is actually more important as it avoids redundant shuffling of values between stack slots.
This particular change adds ISEL support to produce the variadic def STATEPOINT form required by the above. In particular, the first N are lowered to variadic tied def/use pairs. So new statepoint looks like this:
reloc1,reloc2,... = STATEPOINT ..., base1, derived1<tied-def0>, base2, derived2<tied-def1>, ...
N is limited by the maximal number of tied registers machine instruction can have (15 at the moment).
The current patch is restricted to handling relocations within a single basic block. Cross block relocations (e.g. invokes) are handled via the legacy mechanism. This restriction will be relaxed in future patches.
Patch By: dantrushin
Differential Revision: https://reviews.llvm.org/D81648
Common up some existing MBB name printing logic into a single place.
Note that basic block dumping now prints the same set of attributes as
the MIRPrinter.
Change-Id: I8f022bbd922e831bc96d63143d7472c03282530b
Differential Revision: https://reviews.llvm.org/D83253
Emit DWARF 5 call-site symbols even though DWARF 4 is set,
only in the case of LLDB tuning.
This patch addresses PR46643.
Differential Revision: https://reviews.llvm.org/D83463
PassManager.h is one of the top headers in the ClangBuildAnalyzer frontend worst offenders list.
This exposes a large number of implicit dependencies on various forward declarations/includes in other headers that need addressing.
SONY debugger does not prefer debug entry values feature, so
the plan is to avoid production of the entry values
by default when the tuning is SCE debugger.
The feature still can be enabled with the -debug-entry-values
option for the testing/development purposes.
This patch addresses PR46643.
Differential Revision: https://reviews.llvm.org/D83462
In the included test case the align 16 allowed the v23f32 load to handled as load v16f32, load v4f32, and load v4f32(one element not used). These loads all need to be concatenated together into a final vector. In this case we tried to concatenate the two v4f32 loads to match the type of the v16f32 load so we could do a second concat_vectors, but those loads alone only add up to v8f32. So we need to two v4f32 undefs to pad it.
It appears we've tried to hack around a similar issue in this code before by adding undef padding to loads in one of the earlier loops in this function. Originally in r147964 by padding all loads narrower than previous loads to the same size. Later modifed to only the last load in r293088. This patch removes that earlier code and just handles it on demand where we know we need it.
Fixes PR46820
Differential Revision: https://reviews.llvm.org/D84463
Widen or narrow a type to a type with the same scalar size as
another. This can be used to force G_PTR_ADD/G_PTRMASK's scalar
operand to match the bitwidth of the pointer type. Use this to
disallow narrower types for G_PTRMASK.
This adds the llvm.abs(), llvm.umin(), llvm.umax(), llvm.smin(),
and llvm.smax() intrinsics specified in D81829. For SelectionDAG,
the ISD opcodes and all the legalization and lowering already exist,
so this just wires them up to the intrinsic in the SDAG builder and
adds rudimentary tests. For GlobalISel only the min/max intrinsics
are wired up, as llvm.abs() will require the addition of a G_ABS op,
and corresponding legalization support.
Differential Revision: https://reviews.llvm.org/D84125
Clarify the relation between a block's BlockFrequency and the
getEntryFreq() API, and added an API for the relatively common case of
finding a block's frequency relative to the entrypoint.
Added / moved some comments to header.
Differential Revision: https://reviews.llvm.org/D84357
Add support in LegalizerHelper for lowering G_SADDSAT etc. either
using add/subtract-with-overflow or using max/min instructions.
Enable this lowering for AMDGPU so it can be tested. The legalization
rules are still approximate and skips out on using the clamp bit to
treat these as legal, which has never been used before. This also
doesn't yet try to deal with expanding SALU cases.
Summary: We do this already for output operands, but missed it for (non-tied) input operands.
Reviewers: arsenm, Petar.Avramovic
Reviewed By: arsenm
Subscribers: jvesely, wdng, nhaehnle, rovka, hiraditya, llvm-commits, kerbowa
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D83763
On systems where size() doesn't return unsigned long, this leads to an
overloading mismatch. Convert the constant to whatever type is used for
Q.size() on the system.
Currently popFromQueueImpl iterates over all candidates to find the best
one. While the candidate queue is small, this is not a problem. But it
becomes a problem once the queue gets larger. For example, the snippet
below takes 330s to compile with llc -O0, but completes in 3s with this
patch.
define void @test(i4000000* %ptr) {
entry:
store i4000000 0, i4000000* %ptr, align 4
ret void
}
This patch limits the number of candidates to check to 1000. This limit
ensures that it never triggers for test-suite/SPEC2000/SPEC2006 on X86
and AArch64 with -O3, while still drastically limiting the compile-time
in case of very large queues.
It would be even better to use a binary heap to manage to queue
(D83335), but some heuristics change the score of a node in the queue
after another node has been scheduled. I plan to address this for
backends that use the MachineScheduler in the future, but that requires
a more careful evaluation. In the meantime, the limit should help users
impacted by this issue.
The patch includes a slightly smaller version of the motivating example
as test case, to guard against the issue.
Reviewers: efriedma, paquette, niravd
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D84328
Test case `test/CodeGen/WebAssembly/stackified-debug.ll`
was failing due to malformed DwarfExpression.
This failure has been seen in lot of bots, for instance in:
http://lab.llvm.org:8011/builders/lld-x86_64-ubuntu-fast/builds/18794
: 'RUN: at line 1'
/home/buildbot/as-builder-4/lld-x86_64-ubuntu-fast/build/bin/llc
/home/buildbot/as-builder-4/lld-x86_64-ubuntu-fast/build/bin/FileCheck /home/buildbot/as-builder-4/lld-x86_64-ubuntu-fast/llvm-project/llvm/test/CodeGen/WebAssembly/stackified-debug.ll
home/buildbot/as-builder-4/lld-x86_64-ubuntu-fast/llvm-project/llvm/test/CodeGen/WebAssembly/stackified-debug.ll:26:10: error: CHECK: expected string not found in input
CHECK: .int16 4 # Loc expr size
^
<stdin>:34:2: note: scanning from here
.int16 3 # Loc expr size
Differential Revision: https://reviews.llvm.org/D83560
This patch was reverted in 9d2da6759b due to assertion failure seen
in `test/DebugInfo/Sparc/subreg.ll`. Assertion failure was happening
due to malformed/unhandeled DwarfExpression.
Differential Revision: https://reviews.llvm.org/D83560
Summary:
llvm is missing support for DW_OP_implicit_value operation.
DW_OP_implicit_value op is indispensable for cases such as
optimized out long double variables.
For intro refer: DWARFv5 Spec Pg: 40 2.6.1.1.4 Implicit Location Descriptions
Consider the following example:
```
int main() {
long double ld = 3.14;
printf("dummy\n");
ld *= ld;
return 0;
}
```
when compiled with tunk `clang` as
`clang test.c -g -O1` produces following location description
of variable `ld`:
```
DW_AT_location (0x00000000:
[0x0000000000201691, 0x000000000020169b): DW_OP_constu 0xc8f5c28f5c28f800, DW_OP_stack_value, DW_OP_piece 0x8, DW_OP_constu 0x4000, DW_OP_stack_value, DW_OP_bit_piece 0x10 0x40, DW_OP_stack_value)
DW_AT_name ("ld")
```
Here one may notice that this representation is incorrect(DWARF4
stack could only hold integers(and only up to the size of address)).
Here the variable size itself is `128` bit.
GDB and LLDB confirms this:
```
(gdb) p ld
$1 = <invalid float value>
(lldb) frame variable ld
(long double) ld = <extracting data from value failed>
```
GCC represents/uses DW_OP_implicit_value in these sort of situations.
Based on the discussion with Jakub Jelinek regarding GCC's motivation
for using this, I concluded that DW_OP_implicit_value is most appropriate
in this case.
Link: https://gcc.gnu.org/pipermail/gcc/2020-July/233057.html
GDB seems happy after this patch:(LLDB doesn't have support
for DW_OP_implicit_value)
```
(gdb) p ld
p ld
$1 = 3.14000000000000012434
```
Reviewed By: aprantl
Differential Revision: https://reviews.llvm.org/D83560
Summary: These should have half float as the element type
Reviewers: cameron.mcinally, efriedma, sdesmalen, paulwalker-arm
Reviewed By: paulwalker-arm
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D84211
This is an alternative proposal to D81476 (and D82084) - the details were sufficiently confusing to me it seemed easier to write some code and see how it looks.
Reviewers: SouraVX
Differential Revision: https://reviews.llvm.org/D84278
This was structured in a way that implied every split argument is in
memory, or in registers. It is possible to pass an original argument
partially in registers, and partially in memory. Transpose the logic
here to only consider a single piece at a time. Every individual
CCValAssign should be treated independently, and any merge to original
value needs to be handled later.
This is in preparation for merging some preprocessing hacks in the
AMDGPU calling convention lowering into the generic code.
I'm also not sure what the correct behavior for memlocs where the
promoted size is larger than the original value. I've opted to clamp
the memory access size to not exceed the value register to avoid the
explicit trunc/extend/vector widen/vector extract instruction. This
happens for AMDGPU for i8 arguments that end up stack passed, which
are promoted to i16 (I think this is a preexisting DAG bug though, and
they should not really be promoted when in memory).
Summary:
AIX assembly's .set directive is not usable for aliasing purpose.
We need to use extra-label-at-defintion strategy to generate symbol
aliasing on AIX.
Reviewed By: DiggerLin, Xiangling_L
Differential Revision: https://reviews.llvm.org/D83252
Summary:
This patch reduces file size in debug builds by dropping variable locations a
debugger user will not see.
After building the debug entity history map we loop through it. For each
variable we look at each entry. If the entry opens a location range which does
not intersect any of the variable's scope's ranges then we mark it for removal.
After visiting the entries for each variable we also mark any clobbering
entries which will no longer be referenced for removal, and then finally erase
the marked entries. This all requires the ability to query the order of
instructions, so before this runs we number them.
Tests:
Added llvm/test/DebugInfo/X86/trim-var-locs.mir
Modified llvm/test/DebugInfo/COFF/register-variables.ll
Branch folding merges the tails of if.then and if.else into if.else. Each
blocks' debug-locations point to different scopes so when they're merged we
can't use either. Because of this the variable 'c' ends up with a location
range which doesn't cover any instructions in its scope; with the patch
applied the location range is dropped and its flag changes to IsOptimizedOut.
Modified llvm/test/DebugInfo/X86/live-debug-variables.ll
Modified llvm/test/DebugInfo/ARM/PR26163.ll
In both tests an out of scope location is now removed. The remaining location
covers the entire scope of the variable allowing us to emit it as a single
location.
Reviewed By: aprantl
Differential Revision: https://reviews.llvm.org/D82129
There are a few questionable things about this intrinsic and existing
DAG implementation. For some reason the intrinsic hardcodes the second
operand to be scalar-only i32, and SelectionDAG builder makes a
legalization decision based on whether the operand is constant.
The AMDGPU handling of f16 vectors is terrible still since it gets
scalarized even when the vector operation is legal.
The code is is essentially duplicated between the non-strict and
strict case. Apparently no other expansions are currently trying to do
this. This is mostly because I found the behavior of
getStrictFPOperationAction to be confusing. In the ARM case, it would
expand strict_fsub even though it shouldn't due to the later check. At
that point, the logic required to check for legality was more complex
than just duplicating the 2 instruction expansion.
Current tail duplication in machine block placement pass uses block frequency
information in cost model. But frequency number has only relative meaning
compared to other basic blocks in the same function. A large frequency number
doesn't mean it is hot and a small frequency number doesn't mean it is cold.
To overcome this problem, this patch uses profile count in cost model if it's
available. So we can tail duplicate real hot basic blocks.
Differential Revision: https://reviews.llvm.org/D83265
Try to make the behavior more consistent with getGCDType, and bias
towards returning something closer to the source type whenever there's
an ambiguity.
Try harder to find a canonical unmerge type when trying to cover the
desired target type. Handle finding a compatible unmerge type for two
vectors with different element types. This will return the largest
multiple of the source vector element that will evenly divide the
target vector type.
Also make the handling mixing scalars and vectors, and prefer the
source element type as the unmerge target type.
This isn't a natively supported operation, so convert it to a
mask+compare.
In addition to the operation itself, fix up some surrounding stuff to
make the testcase work: we need concat_vectors on i1 vectors, we need
legalization of i1 vector truncates, and we need to fix up all the
relevant uses of getVectorNumElements().
Differential Revision: https://reviews.llvm.org/D83811
Its effect could be achieved by
`-stop-after`,`-print-after`,`-print-after-all`. But a few tests need to
print MIR after ISel which could not be done with
`-print-after`/`-stop-after` since isel pass does not have commandline name.
That's the reason `--print-machineinstrs` is downgraded to
`--print-after-isel` in this patch. `--print-after-isel` could be
removed after we switch to new pass manager since isel pass would have a
commandline text name to use `print-after` or equivalent switches.
The motivation of this patch is to reduce tests dependency on
would-be-deprecated feature.
Reviewed By: arsenm, dsanders
Differential Revision: https://reviews.llvm.org/D83275
Amara Emerson