This patch does the following simplification:
```
sel1 = select cond1, X, Y
sel2 = select cond2, sel1, Z
-->
sel2 = select cond2, X, Z if cond2 implies cond1
sel2 = select cond2, Y, Z if cond2 implies !cond1
```
Alive2: https://alive2.llvm.org/ce/z/9A_arU
It cannot be done in CVP/SCCP since we should guarantee that `cond2` is
not an undef.
If the phi node found by matchSimpleRecurrence is not from the current
loop, then isAlmostDeadIV panics. With this patch we bail out early.
Signed-off-by: Patrick O'Neill <patrick@rivosinc.com>
---------
Signed-off-by: Patrick O'Neill <patrick@rivosinc.com>
This helps cover some missing cases in both and hopefully serves as
creating an easier framework for extending general condition based
analysis.
Closes#83161
If a gep has only one phi as one of its operands and the remaining
indexes are constant, we can unfold `gep ptr, (phi idx1, idx2)` to `phi
((gep ptr, idx1), (gep ptr, idx2))`.
Take care not to unfold recursive phis.
Followup to #80983.
This was initially was #83087. Initial PR did not handle allocas in
entry block that weren't at the beginning of the function, causing GEPs
to be inserted after the first chunk of allocas but potentially before
an alloca not at the beginning. Insert GEPs at the end of the entry
block instead since constants/arguments/static allocas can all be used
there.
Currently we model subvector inserts and extracts as shuffles,
potentially going as far as scalarizing. If the types are legal then
they can just be simple zip/unzip operations, or possible even no-ops.
Change the cost to a relatively small one to ensure that simple loops
featuring such operations between fixed and scalable vector types that
are effectively the same at a given sve width can be unrolled and
further optimized.
DSE uses BatchAA, which caches queries using pairs of MemoryLocations.
At the moment, DSE may remove instructions that are used as pointers in
cached MemoryLocations. If a new instruction used by a new MemoryLoation
and this instruction gets allocated at the same address as a previosuly
cached and then removed instruction, we may access an incorrect entry in
the cache.
To avoid this delay removing all instructions except MemoryDefs until
the end of DSE. This should avoid removing any values used in BatchAA's
cache.
Test case by @vporpo from
https://github.com/llvm/llvm-project/pull/83181.
(Test not precommitted because the results are non-determinstic - memset
only sometimes gets removed)
PR: https://github.com/llvm/llvm-project/pull/83411
Masked gather is very expensive operation and sometimes better to
represent it as a serie of consecutive/strided loads + insertsubvectors
sequences. Patch adds some basic estimation and if loads+insertsubvector
is cheaper, decides to represent it in this way rather than masked
gather.
Reviewers: RKSimon
Reviewed By: RKSimon
Pull Request: https://github.com/llvm/llvm-project/pull/83481
If the graph incudes only single "gather" node with only
extractelements/undefs, which used only in insertelement-based
buildvector sequences, it still might be profitable to vectorize it.
Need to rely on the cost model, not throw this graph away immediately.
Reviewers: RKSimon
Reviewed By: RKSimon
Pull Request: https://github.com/llvm/llvm-project/pull/83581
We can support these by changing the sext promotion to -zext(-C) and
replacing a sgt check with ugt. Reframing the logic in terms of how the
unsigned range are affected. More comments in the patch.
The new cases check isLegalAddImmediate to avoid some
regressions in lit tests.
When do the analysis for the (potential) masked gather node, we check
that not greater than half of the pointer operands are loop invariants
or potentially vectorizable.
Need to check actually, that we have a loop at first
and do better check for the potentially vectorizable
pointers.
Reviewers: RKSimon
Reviewed By: RKSimon
Pull Request: https://github.com/llvm/llvm-project/pull/83472
If the buildvector node contains extract, which later should be combined
with some other nodes by shuffling, need to estimate the cost of this
shuffle before building the mask after shuffle.
Reviewers: RKSimon
Reviewed By: RKSimon
Pull Request: https://github.com/llvm/llvm-project/pull/83442
Many profitable optimizations cannot be performed at -Oz, due to
unrotated loops. While this is worse for size (minimally), many of the
optimizations significantly reduce code size, such as memcpy
optimizations and other patterns found by loop idiom recognition.
Related discussion can be found in issue #50308.
This patch adds an experimental, backend-only flag to allow loop header
duplication, regardless of the optimization level. Downstream consumers
can experiment with this flag, and if it is profitable, we can adjust
the compiler's defaults accordingly, and expose any useful frontend
flags to opt into the new behavior.
Fixes: aadd7650447b
The above commit landed with an incorrect test expect, missing a `metadata`
prefix. This patch adds the expected prefix to the test.
Fixes the error described here:
a93a4ec7dd (commitcomment-138965199)
The function `findDbgIntrinsics` is used to return a list of debug
intrinsics and DPValues that use a given value, with the intent that no
duplicates are returned in either list. For DPValues, we've guarded
against DPValues that use a value multiple times as part of a DIArgList,
but we have not guarded against DPValues that use a value multiple times
as separate operands (currently only possible for `dbg_assign`s,
something I missed in my implementation of that type!). This patch adds
a guard, and also updates a test to cover this case.
Follow the 2019 rules and order -0 as less than +0 and +0 as greater
than -0. As currently defined this isn't required for the intrinsics,
but is a better QoI.
This will avoid the workaround in libc added by #83158
Originally, when `EnableImportMetadata` enabled, `SourceFileName` will
be recorded as `thinlto_src_module`. Now `SourceFileName` will be
recorded as `thinlto_src_file` and `ModuleIdentifier` will be recorded
as `thinlto_src_module`.
This reverts commit 2eb63982e88b9ed8336158d35884b1a1d04a0f78.
This caused verifier error
```
Instruction does not dominate all uses!
```
for some projects using Halide.
The verifier error happens inside `Halide::Internal::CodeGen_LLVM::optimize_module`
and looks like a genuine SROA issue.
If a gep has only one phi as one of its operands and the remaining
indexes are constant, we can unfold `gep ptr, (phi idx1, idx2)` to `phi
((gep ptr, idx1), (gep ptr, idx2))`.
Take care not to unfold recursive phis.
Followup to #80983.
Recent set of changes (PR #67725) in loop interleaving algorithm caused removal of the loop trip count threshold for allowing interleaving. Therefore configuration option interleave-small-loop-scalar-reduction is no longer needed.
Some optimizations are apply after UF and VF have been chosen. This
patch adds an extra print of the final VPlan just before
codegen/execution.
In the future, there will be additional transforms that are applied
later (interleaving for example).
PR: https://github.com/llvm/llvm-project/pull/82269
New change on top of [reviewed
patch](https://github.com/llvm/llvm-project/pull/81691) are [in commits
after this
one](d0757f46b3).
Previous commits are restored from the remote branch with timestamps.
1. Fix build breakage for non-ELF platforms, by defining the missing
functions {`__llvm_profile_begin_vtables`, `__llvm_profile_end_vtables`,
`__llvm_profile_begin_vtabnames `, `__llvm_profile_end_vtabnames`}
everywhere.
* Tested on mac laptop (for darwins) and Windows. Specifically,
functions in `InstrProfilingPlatformWindows.c` returns `NULL` to make it
more explicit that type prof isn't supported; see comments for the
reason.
* For the rest (AIX, other), mostly follow existing examples (like this
[one](f95b2f1acf))
2. Rename `__llvm_prf_vtabnames` -> `__llvm_prf_vns` for shorter section
name, and make returned pointers
[const](a825d2a4ec (diff-4de780ce726d76b7abc9d3353aef95013e7b21e7bda01be8940cc6574fb0b5ffR120-R121))
**Original Description**
* Raw profile format
- Header: records the byte size of compressed vtable names, and the
number of profiled vtable entries (call it `VTableProfData`). Header
also records padded bytes of each section.
- Payload: adds a section for compressed vtable names, and a section to
store `VTableProfData`. Both sections are padded so the size is a
multiple of 8.
* Indexed profile format
- Header: records the byte offset of compressed vtable names.
- Payload: adds a section to store compressed vtable names. This section
is used by `llvm-profdata` to show the list of vtables profiled for an
instrumented site.
[The originally reviewed
patch](https://github.com/llvm/llvm-project/pull/66825) will have
profile reader/write change and llvm-profdata change.
- To ensure this PR has all the necessary profile format change along
with profile version bump, created a copy of the originally reviewed
patch in https://github.com/llvm/llvm-project/pull/80761. The copy
doesn't have profile format change, but it has the set of tests which
covers type profile generation, profile read and profile merge. Tests
pass there.
rfc in
https://discourse.llvm.org/t/rfc-dynamic-type-profiling-and-optimizations-in-llvm/74600
---------
Co-authored-by: modiking <modiking213@gmail.com>
Fix a bug in the handling of cases where a callsite's stack ids
partially overlap with the pruned context during matching of
calls to the graph contructed from the profiled contexts. This fix makes
the code match the comments.
This reapplication changes debug intrinsic declaration removal to only take
place when printing final IR, so that the processing format of the Module
does not affect the output.
This reverts commit d128448efdd4e2bf3c9bc9a5b43ae642aa78026f.
This is mostly NFC but some output does change due to consistently
inserting into poison rather than undef and using i64 as the index
type for inserts.
This is mostly NFC but some output does change due to consistently
inserting into poison rather than undef and using i64 as the index
type for inserts.
This is mostly NFC but some output does change due to consistently
inserting into poison rather than undef and using i64 as the index
type for inserts.
This removes a CHECK-NOT: vector.body line from the test which seems to
imply the test does not get vectorized, but it does now.
This line was left over from when the test was pre-committed, remove it.
This test triple was removed earlier to fix build errors. To preserve the original test intention the triple is re-added with an explicit target requirement.
At the moment, some VPInstructions create only a single scalar value,
but use VPTransformatState's 'vector' storage for this value. Those
values are effectively uniform-per-VF (or in some cases
uniform-across-VF-and-UF). Using the vector/per-part storage doesn't
interact well with other recipes, that more accurately using (Part,
Lane) to look up scalar values and prevents VPInstructions creating
scalars from interacting with other recipes working with scalars.
This PR tries to unify handling of scalars by using (Part, 0) for scalar
values where only the first lane is demanded. This allows using
VPInstructions with other recipes like VPScalarCastRecipe and is also
needed when using VPInstructions in more cases otuside the vector loop
region to generate scalars.
Depends on https://github.com/llvm/llvm-project/pull/80269
Reverted due to failures on buildbots, where a new cl flag was placed
in the wrong file, resulting in link errors.
https://lab.llvm.org/buildbot/#/builders/198/builds/8548
This reverts commit 0b398256b3f72204ad1f7c625efe4990204e898a.
This patch adds support for printing the proposed non-instruction debug
info ("RemoveDIs") out to textual IR. This patch does not add any
bitcode support, parsing support, or documentation.
Printing of the new format is controlled by a flag added in this patch,
`--write-experimental-debuginfo`, which defaults to false. The new
format will be printed *iff* this flag is true, so whether we use the IR
format is completely independent of whether we use non-instruction debug
info during LLVM passes (which is controlled by the
`--try-experimental-debuginfo-iterators` flag).
Even with the flag disabled, some existing tests need to be updated, as this
patch causes debug intrinsic declarations to be changed in a round trip,
such that they always appear at the end of a module and have no attributes
(this has no functional change on the module).
The design of this new IR format was proposed previously on
Discourse, and any further discussion about the design can still be
contributed there:
https://discourse.llvm.org/t/rfc-debuginfo-proposed-changes-to-the-textual-ir-representation-for-debug-values/73491
llvm.dbg.labels are deleted in SpeculativelyExecuteBB so DPLabels should
be too.
Modify existing test to check this (NB I couldn't find a dedicated
debug-info test that checks this behaviour).
