This adds a function specialization pass to LLVM. Constant parameters
like function pointers and constant globals are propagated to the callee by
specializing the function.
This is a first version with a number of limitations:
- The pass is off by default, so needs to be enabled on the command line,
- It does not handle specialization of recursive functions,
- It does not yet handle constants and constant ranges,
- Only 1 argument per function is specialised,
- The cost-model could be further looked into, and perhaps related,
- We are not yet caching analysis results.
This is based on earlier work by Matthew Simpson (D36432) and Vinay Madhusudan.
More recently this was also discussed on the list, see:
https://lists.llvm.org/pipermail/llvm-dev/2021-March/149380.html.
The motivation for this work is that function specialisation often comes up as
a reason for performance differences of generated code between LLVM and GCC,
which has this enabled by default from optimisation level -O3 and up. And while
this certainly helps a few cpu benchmark cases, this also triggers in real
world codes and is thus a generally useful transformation to have in LLVM.
Function specialisation has great potential to increase compile-times and
code-size. The summary from some investigations with this patch is:
- Compile-time increases for short compile jobs is high relatively, but the
increase in absolute numbers still low.
- For longer compile-jobs, the extra compile time is around 1%, and very much
in line with GCC.
- It is difficult to blame one thing for compile-time increases: it looks like
everywhere a little bit more time is spent processing more functions and
instructions.
- But the function specialisation pass itself is not very expensive; it doesn't
show up very high in the profile of the optimisation passes.
The goal of this work is to reach parity with GCC which means that eventually
we would like to get this enabled by default. But first we would like to address
some of the limitations before that.
Differential Revision: https://reviews.llvm.org/D93838
Relaxing superclass constraint for VSX register classes helps reducing
32-byte spills and copies when register pressure is high.
In test case affected, some of them introduces more copies due to new
allocation order. However, this patch should not be the root cause, and
we may be able to fix it in other places of register allocation.
Reviewed By: nemanjai
Differential Revision: https://reviews.llvm.org/D104006
When using FP to access stack objects, the scalable stack objects will
be put at the lower end of the frame. It looks like
```
|-------------------| <-- FP
| callee-saved regs |
|-------------------|
| scalar local vars |
|-------------------|
| RVV local vars |
|-------------------| <-- SP
```
If there are scalar arguments that need to pass through memory and there
are vector objects on the stack using FP to access. The outgoing scalar
arguments will overwrite the vector objects. It looks like
```
|-------------------| <-- FP
| callee-saved regs |
|-------------------|
| scalar local vars |
|-------------------| |-------------------|
| RVV local vars | | outgoing args | <- outgoing arguments
|-------------------| <-- SP |-------------------| overwrite from here.
```
In this patch, we reserve the stack for the outgoing arguments before
function calls if using FP to access and there are scalable vector frame
objects. It looks like
```
|-------------------| <-- FP
| callee-saved regs |
|-------------------|
| scalar local vars |
|-------------------|
| RVV local vars |
|-------------------|
| outgoing args |
|-------------------| <-- SP
```
Differential Revision: https://reviews.llvm.org/D103622
This fixes the concern in single element store scalarization that the
alignment of new store may be larger than it should be. It calculates
the largest alignment if index is constant, and a safe one if not.
Reviewed By: lebedev.ri, spatel
Differential Revision: https://reviews.llvm.org/D103419
This helps us select W instructions in more cases. Most of the
affected tests have had the sign_extend_inreg or AND folded into
sextload/zextload.
Differential Revision: https://reviews.llvm.org/D104079
When the extend is from 8 or 16 bits, the addressing modes don't support those
extensions, but we weren't checking that and therefore always generated the 32->64b
extension mode. Fun.
Differential Revision: https://reviews.llvm.org/D104070
When reducing vector builds to shuffles it possible that
the DAG combiner may try to extract invalid subvectors.
This happens as the existing code assumes vectors will be power
of 2 sizes, which is already untrue, but becomes more noticable
with v6 and v7 types.
Specifically the existing code assumes that half PowerOf2Ceil of
a given vector index will fit twice into a given vector.
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D103880
The loads end up becoming sextload/zextload which prevent our
isel patterns from finding the sign_extend_inreg or AND instruction
we need.
The easiest way to fix this is to use computeKnownBits or
ComputeNumSignBits in our isel matching to catch this.
Value::SubclassID cannot be directly compared to Instruction enums, such as
Instruction::{Call,Invoke,CallBr}. We have to first subtract InstructionVal
from the SubclassID to get the OpCode, similar to Instruction::getOpCode().
Reviewed By: nickdesaulniers
Differential Revision: https://reviews.llvm.org/D104043
First we refactor the code which does no wrapping add sequences
match: we need to allow different operand orders for
the key add instructions involved in the match.
Then we use the refactored code trying 4 variants of matching operands.
Originally the code relied on the fact that the matching operands
of the two last add instructions of memory index calculations
had the same LHS argument. But which operand is the same
in the two instructions is actually not essential, so now we allow
that to be any of LHS or RHS of each of the two instructions.
This increases the chances of vectorization to happen.
Reviewed By: volkan
Differential Revision: https://reviews.llvm.org/D103912
-Wframe-larger-than= is an interesting warning; we can't know the frame
size until PrologueEpilogueInsertion (PEI); very late in the compilation
pipeline.
-Wframe-larger-than= was propagated through CC1 as an -mllvm flag, then
was a cl::opt in LLVM's PEI pass; this meant it was dropped during LTO
and needed to be re-specified via -plugin-opt.
Instead, make it part of the IR proper as a module level attribute,
similar to D103048. Introduce -fwarn-stack-size CC1 option.
Reviewed By: rsmith, qcolombet
Differential Revision: https://reviews.llvm.org/D103928
SROA sometimes preserves MD_mem_parallel_loop_access and MD_access_group metadata on loads/stores, and sometimes fails to do so. This change adds copying of the MD after other CreateAlignedLoad/CreateAlignedStores. Also fix a case where the metadata was being copied from a load, rather than the store.
Added a LIT test to catch one case.
Patch by Mark Mendell
Differential Revision: https://reviews.llvm.org/D103254
This adds legalization for scalar G_CTTZ and G_CTTZ_ZERO_UNDEF. Vector support
requires handling vector G_BITREVERSE, which I haven't gotten around to yet.
For G_CTTZ_ZERO_UNDEF, we just lower it to G_CTTZ.
For G_CTTZ, we match SelectionDAG's lowering to a G_BITREVERSE + G_CTLZ.
e.g. https://godbolt.org/z/nPEseYh1s
(With this patch, we have slightly worse codegen than SDAG for types smaller
than s32; it seems like we're missing a combine.)
Also, this adds in a function to build G_BITREVERSE to MachineIRBuilder.
Differential Revision: https://reviews.llvm.org/D104065
As noted in https://bugs.llvm.org/show_bug.cgi?id=46666, the current behavior of assuming if-conversion safety if a loop is annotated parallel (`!llvm.loop.parallel_accesses`), is not expectable, the documentation for this behavior was since removed from the LangRef again, and can lead to invalid reads.
This was observed in POCL (https://github.com/pocl/pocl/issues/757) and would require similar workarounds in current work at hipSYCL.
The question remains why this was initially added and what the implications of removing this optimization would be.
Do we need an alternative mechanism to propagate the information about legality of if-conversion?
Or is the idea that conditional loads in `#pragma clang loop vectorize(assume_safety)` can be executed unmasked without additional checks flawed in general?
I think this implication is not part of what a user of that pragma (and corresponding metadata) would expect and thus dangerous.
Only two additional tests failed, which are adapted in this patch. Depending on the further direction force-ifcvt.ll should be removed or further adapted.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D103907
Essentially, the cover function simply combines the loop level check and the function level scope into one call. This simplifies several callers and is (subjectively) less error prone.
This addresses a performance regression reported against 3c6e4191. That change (correctly) limited a transform based on assumed finiteness to mustprogress loops, but the previous change (38540d7) which introduced the mustprogress check utility only handled function attributes, not the loop metadata form.
It turns out that clang uses the function attribute form for C++, and the loop metadata form for C. As a result, 3c6e4191 ended up being a large regression in practice for C code as loops weren't being considered mustprogress despite the language semantics.
Implementation of the unroll directive introduced in OpenMP 5.1. Follows the approach from D76342 for the tile directive (i.e. AST-based, not using the OpenMPIRBuilder). Tries to use `llvm.loop.unroll.*` metadata where possible, but has to fall back to an AST representation of the outer loop if the partially unrolled generated loop is associated with another directive (because it needs to compute the number of iterations).
Reviewed By: ABataev
Differential Revision: https://reviews.llvm.org/D99459
Surprisingly, not all instructions are always simplified after unrolling
and before MVE gather/scatter lowering. Notably dead gather operations
can be left around which cause the gather/scatter lowering pass to crash
if there are multiple gathers, some of which are dead.
This patch ensures they are simplified before we modify anything, which
can change some of the existing tests, including making them no-longer
test what they originally tested. This uses a combination of disabling
the gather/scatter lowering pass and adjusting the test to keep them as
before.
Differential Revision: https://reviews.llvm.org/D103150
If a macro is defined on the command line and then overridden in the source code, this is likely to be an error in the user's build system. We should warn on this.
Reviewed By: thakis
Differential Revision: https://reviews.llvm.org/D104008
We fall back on G_CTTZ_ZERO_UNDEF a lot when building clang for arm64 with
gisel.
Handling this will require that we can handle G_BITREVERSE.
This patch marks G_BITREVERSE instructions with natively supported types as
legal. We get selection on these types for free via the importer.
Differential Revision: https://reviews.llvm.org/D103999
There is no need to schedule insertelement instructions. The compiler
did not schedule them before it started support their vectorization and
it should not do it after. We pre-schedule them manually when finding
a build vector sequence.
Disabling scheduling of insertelement instructions improves compile
time and vectorization of the very large basic blocks by saving
scheduling budget for other instructions.
Differential Revision: https://reviews.llvm.org/D104026
On CMake versions greater that >= 3.16 on AIX, shared libraries are
created as archives (which is the normal form for the platform). However
plugins libraries which are passed directly to a executable, like
libLTO to the linker, are usual build as plain `.so`, so this patch
restores this behaviour for libLTO on AIX (and adjust the name if need be
to account for the fact that llvm_add_library likes to force an empty
name prefix on modules), so we end up with the expected libLTO.so
Reviewed By: w2yehia
Differential Revision: https://reviews.llvm.org/D103824
```
llvm-project/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp:8024:19: warning: loop variable 'VF' of type 'const llvm::ElementCount' creates a copy from type 'const llvm::ElementCount' [-Wrange-loop-analysis]
for (const auto VF : VFCandidates) {
^
llvm-project/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp:8024:8: note: use reference type 'const llvm::ElementCount &' to prevent copying
for (const auto VF : VFCandidates) {
^~~~~~~~~~~~~~~
&
1 warning generated.
```
Differential Revision: https://reviews.llvm.org/D103970
I noticed that I did not update the command guide when introducing the
--rsp-quoting option. This change fixes this.
Differential Revision: https://reviews.llvm.org/D103915
Debug info is currently preventing VPT block creation, leading to
different codegen. This patch attempts to skip any debug instructions
during vpt block creation, making sure they do not interfere.
Differential Revision: https://reviews.llvm.org/D103610
This patch allows folding stepvector + extract to the lane when the lane is
lower than the minimum size of the scalable vector. This fold is possible
because lane X of a stepvector is also X!
For instance, extracting element 3 of a <vscale x 4 x i64>stepvector is 3.
Differential Revision: https://reviews.llvm.org/D103153
MASM specifies that all variable definitions are redefinable, except for EQU definitions to expressions. (TEXTEQU is unspecified, but appears to be fully redefinable as well.)
Also, in practice, ML.EXE allows redefinitions where the value doesn't change.
Make variable redefinition possible for text macros, suppressing expansion if written as the first argument to an EQU or TEXTEQU directive.
Reviewed By: thakis
Differential Revision: https://reviews.llvm.org/D103993
This patch allows that scalable vector can fold extractelement and constant splat
only when the lane index is lower than the minimum number of elements of the vector.
Differential Revision: https://reviews.llvm.org/D103180
This is a roll forward of D102679.
This patch simplifies the implementation of Sequence and makes it compatible with llvm::reverse.
It exposes the reverse iterators through rbegin/rend which prevents a dangling reference in std::reverse_iterator::operator++().
Note: Compared to D102679, this patch introduces a `asSmallVector()` member function and fixes compilation issue with GCC 5.
Differential Revision: https://reviews.llvm.org/D103948
Summary: Some structs like FileHeader32/SectionHeader32
defined in llvm/include/llvm/BinaryFormat/XCOFF.h seem
unnecessary, because we only need their size. So this
patch removes them and defines size constants directly.
Reviewed By: jhenderson
Differential Revision: https://reviews.llvm.org/D103901
This change implements new DAG notes GLOBAL_GET/GLOBAL_SET, and
lowering methods for load and stores of reference types from IR
globals. Once the lowering creates the new nodes, tablegen pattern
matches those and converts them to Wasm global.get/set.
Reviewed By: tlively
Differential Revision: https://reviews.llvm.org/D95425
