We've noticed that for large builds executing thin-link can take on the
order of 10s of minutes. We are only using a single thread to write the
sharded indices and import files for each input bitcode file. While we
need to ensure the index file produced lists modules in a deterministic
order, that doesn't prevent us from executing the rest of the work in
parallel.
In this change we use a thread pool to execute as much of the backend's
work as possible in parallel. In local testing on a machine with 80
cores, this change makes a thin-link for ~100,000 input files run in ~2
minutes. Without this change it takes upwards of 10 minutes.
---------
Co-authored-by: Nuri Amari <nuriamari@fb.com>
The primary motivation is to remove `EntryCount` from `FunctionSummary`.
This frees 8 bytes out of `sizeof(FunctionSummary)` (136 bytes as of
64498c5483).
While I'm at it, this PR clean up {SummaryBasedOptimizations,
SyntheticCountsPropagation} since they were not used and there are no
plans to further invest on them.
With this patch, bitcode writer writes a placeholder 0 at the byte
offset of `EntryCount` and bitcode reader can parse the function entry
count at the correct byte offset. Added a TODO to stop writing
`EntryCount` and bump bitcode version
The background is as follows. I'm planning to reduce the memory
footprint of ThinLTO indexing by changing ImportMapTy, the data
structure used for an import list. Once this patch lands, I'm
planning to change the type slightly. The new type alias allows us to
update the type without touching many places.
This patch introduces type alias ModuleToSummariesForIndexTy.
I'm planning to change the type slightly to allow heterogeneous lookup
(that is, std::map<K, V, std::less<>>) in a subsequent patch. The
problem is that changing the type affects many places. Using a type
alias reduces the impact.
Without this patch, computeLTOCacheKey computes SHA1, creates its
hexadecimal representation with toHex, which returns std::string, and
then copies it to an output parameter of type SmallString.
This patch removes the redirection and teaches computeLTOCacheKey to
directly return std::string computed by toHex. With the move
semantics, no buffer copy should be involved.
While I am at it, this patch adds a Twine to concatenate two strings.
\#92331 tried to make `ObjCARCContractPass` by default, but it caused a
regression on O0 builds and was reverted.
This patch trys to bring that back by:
1. reverts the
[revert](1579e9ca9c).
2. `createObjCARCContractPass` only on optimized builds.
Tests are updated to refelect the changes. Specifically, all `O0` tests
should not include `ObjCARCContractPass`
Signed-off-by: Peter Rong <PeterRong@meta.com>
https://github.com/llvm/llvm-project/pull/87600 was reverted in order to
revert
6262763341.
Now https://github.com/llvm/llvm-project/pull/95482 is fix forward for
6262763341.
This patch is a reland for
https://github.com/llvm/llvm-project/pull/87600
**Changes on top of original patch**
In `llvm/include/llvm/IR/ModuleSummaryIndex.h`, make the type of
`GVSummaryPtrSet` an `unordered_set` which is more memory efficient when
the number of elements is smaller than 128 [1]
**Original commit message**
For distributed ThinLTO, the LTO indexing step generates combined
summary for each module, and postlink pipeline reads the combined
summary which stores the information for link-time optimization.
This patch populates the 'import type' of a summary in bitcode, and
updates bitcode reader to parse the bit correctly.
[1]
393eff4e02/llvm/lib/Support/SmallPtrSet.cpp (L43)
This reverts commit 8cc8e5d6c6ac9bfc888f3449f7e424678deae8c2.
This reverts commit dae55c89835347a353619f506ee5c8f8a2c136a7.
Causes major compile-time regressions for unoptimized builds.
Prior to this patch, when using -fthinlto-index= the ObjCARCContractPass isn't run prior to CodeGen, and instruction selection fails on IR containing arc intrinsics. This patch is motivated by that usecase.
The pass was previously added in various places codegen is performed. This patch adds the pass to the default codegen pipepline, makes sure it bails immediately if no arc intrinsics are found, and removes the adhoc scheduling of the pass.
Co-authored-by: Nuri Amari <nuriamari@fb.com>
For distributed ThinLTO, the LTO indexing step generates combined
summary for each module, and postlink pipeline reads the combined
summary which stores the information for link-time optimization.
This patch populates the 'import type' of a summary in bitcode, and
updates bitcode reader to parse the bit correctly.
The base class llvm::ThreadPoolInterface will be renamed
llvm::ThreadPool in a subsequent commit.
This is a breaking change: clients who use to create a ThreadPool must
now create a DefaultThreadPool instead.
When computing the module maps, `ThinLTOCodeGenerator` asserts if it
sees duplicating module names. This PR adds a debug print, so that the
list of modules already added can be printed. With this information, one
can identify which modules are causing the duplication.
Discussion about this approach: https://discourse.llvm.org/t/rfc-safer-whole-program-class-hierarchy-analysis/65144/18
When enabling WPD in an environment where native binaries are present, types we want to optimize can be derived from inside these native files and devirtualizing them can lead to correctness issues. RTTI can be used as a way to determine all such types in native files and exclude them from WPD providing a safe checked way to enable WPD.
The approach is:
1. In the linker, identify if RTTI is available for all native types. If not, under `--lto-validate-all-vtables-have-type-infos` `--lto-whole-program-visibility` is automatically disabled. This is done by examining all .symtab symbols in object files and .dynsym symbols in DSOs for vtable (_ZTV) and typeinfo (_ZTI) symbols and ensuring there's always a match for every vtable symbol.
2. During thinlink, if `--lto-validate-all-vtables-have-type-infos` is set and RTTI is available for all native types, identify all typename (_ZTS) symbols via their corresponding typeinfo (_ZTI) symbols that are used natively or outside of our summary and exclude them from WPD.
Testing:
ninja check-all
large Meta service that uses boost, glog and libstdc++.so runs successfully with WPD via --lto-whole-program-visibility. Previously, native types in boost caused incorrect devirtualization that led to crashes.
Reviewed By: MaskRay, tejohnson
Differential Revision: https://reviews.llvm.org/D155659
This will make it easy for callers to see issues with and fix up calls
to createTargetMachine after a future change to the params of
TargetMachine.
This matches other nearby enums.
For downstream users, this should be a fairly straightforward
replacement,
e.g. s/CodeGenOpt::Aggressive/CodeGenOptLevel::Aggressive
or s/CGFT_/CodeGenFileType::
The module paths string table mapped to both an id sequentially assigned
during LTO linking, and the module hash. The former is leftover from
before the module hash was added for caching and subsequently replaced
use of the module id when renaming promoted symbols (to avoid affects
due to link order changes). The sequentially assigned module id was not
removed, however, as it was still a convenience when serializing to/from
bitcode and assembly.
This patch removes the module id from this table, since it isn't
strictly needed and can lead to confusion on when it is appropriate to
use (e.g. see fix in D156525). It also takes a (likely not significant)
amount of overhead. Where an integer module id is needed (e.g. bitcode
writing), one is assigned on the fly.
There are a couple of test changes since the paths are now sorted
alphanumerically when assigning ids on the fly during assembly writing,
in order to ensure deterministic behavior.
Differential Revision: https://reviews.llvm.org/D156730
The import/export maps, and the ModuleToDefinedGVSummaries map, are all
indexed by module paths, which are StringRef obtained from the module
summary index, which already has a data structure than owns these
strings (the ModulePathStringTable). Because these other maps are also
StringMap, which makes a copy of the string key, we were keeping
multiple extra copies of the module paths, leading to memory overhead.
Change these to DenseMap keyed by StringRef, and document that the
strings are owned by the index.
The only exception is the llvm-link tool which synthesizes an import list
from command line options, and I have added a string cache to maintain
ownership there.
I measured around 5% memory reduction in the thin link of a large
binary.
Differential Revision: https://reviews.llvm.org/D156580
SubtargetFeature.h is currently part of MC while it doesn't depend on
anything in MC. Since some LLVM components might have the need to work
with target features without necessarily needing MC, it might be
worthwhile to move SubtargetFeature.h to a different location. This will
reduce the dependencies of said components.
Note that I choose TargetParser as the destination because that's where
Triple lives and SubtargetFeatures feels related to that.
This issues came up during a JITLink review (D149522). JITLink would
like to avoid a dependency on MC while still needing to store target
features.
Reviewed By: MaskRay, arsenm
Differential Revision: https://reviews.llvm.org/D150549
I noticed that we are converting llvm.public.type.test to regular
llvm.type.test too early, and thus not updating those in imported
functions. This would result in losing out on WPD opportunities. Move
the update to after function importing, and improve test to cover this
case.
Differential Revision: https://reviews.llvm.org/D150326
This logic was added in https://reviews.llvm.org/D95943 specifically to
handle an issue for non-prevailing global variables. It turns out that
it adds a new issue for prevailing glboal variables, since those could
be replaced by an available_externally definition and hence incorrectly
omitted from the output object file. Limit the import to non-prevailing
global variables to fix this, as suggested by @tejohnson.
The bulk of the diff is mechanical changes to thread isPrevailing
through to where it's needed and ensure it's available before the
relevant calls; the actual logic change itself is straightforward.
Fixes https://github.com/llvm/llvm-project/issues/61677
Reviewed By: tejohnson
Differential Revision: https://reviews.llvm.org/D146876
This patch mechanically replaces None with std::nullopt where the
compiler would warn if None were deprecated. The intent is to reduce
the amount of manual work required in migrating from Optional to
std::optional.
This is part of an effort to migrate from llvm::Optional to
std::optional:
https://discourse.llvm.org/t/deprecating-llvm-optional-x-hasvalue-getvalue-getvalueor/63716
A specific case for ThinLTO cache pruning is that the current build is huge, and the cache wasn't big enough to hold the intermediate object files of that build. So in doing that build, a file would be cached, and later in that same build it would be evicted. This was significantly decreasing the effectiveness of the cache. By giving this warning, the user could identify the required cache size/files and improve ThinLTO link speed.
Differential Revision: https://reviews.llvm.org/D135590
Turning on opaque pointers has uncovered an issue with WPD where we currently pattern match away `assume(type.test)` in WPD so that a later LTT doesn't resolve the type test to undef and introduce an `assume(false)`. The pattern matching can fail in cases where we transform two `assume(type.test)`s into `assume(phi(type.test.1, type.test.2))`.
Currently we create `assume(type.test)` for all virtual calls that might be devirtualized. This is to support `-Wl,--lto-whole-program-visibility`.
To prevent this, all virtual calls that may not be in the same LTO module instead use a new `llvm.public.type.test` intrinsic in place of the `llvm.type.test`. Then when we know if `-Wl,--lto-whole-program-visibility` is passed or not, we can either replace all `llvm.public.type.test` with `llvm.type.test`, or replace all `llvm.public.type.test` with `true`. This prevents WPD from trying to pattern match away `assume(type.test)` for public virtual calls when failing the pattern matching will result in miscompiles.
Reviewed By: tejohnson
Differential Revision: https://reviews.llvm.org/D128955
Most of `MemoryBuffer` interfaces expose a `RequiresNullTerminator` parameter that's being used to:
* determine how to open a file (`mmap` vs `open`),
* assert newly initialized buffer indeed has an implicit null terminator.
This patch adds the paramater to the `SmallVectorMemoryBuffer` constructors, meaning:
* null terminator can now be added to `SmallVector`s that didn't have one before,
* `SmallVectors` that had a null terminator before keep it even after the move.
In line with existing code, the new parameter is defaulted to `true`. This patch makes sure all calls to the `SmallVectorMemoryBuffer` constructor set it to `false` to preserve the current semantics.
Reviewed By: dexonsmith
Differential Revision: https://reviews.llvm.org/D115331
We almost always want to use the default AA pipeline. It's very easy for
users of PassBuilder to forget to customize the AAManager to use the
default AA pipeline (for example, the NewPM C API forgets to do this).
If somebody wants a custom AA pipeline, similar to what is being done
now with the default AA pipeline registration, they can
FAM.registerPass([&] { return std::move(MyAA); });
before calling
PB.registerFunctionAnalyses(FAM);
For example, LTOBackend.cpp and NewPMDriver.cpp do this.
Reviewed By: asbirlea
Differential Revision: https://reviews.llvm.org/D113210
This moves the registry higher in the LLVM library dependency stack.
Every client of the target registry needs to link against MC anyway to
actually use the target, so we might as well move this out of Support.
This allows us to ensure that Support doesn't have includes from MC/*.
Differential Revision: https://reviews.llvm.org/D111454
As described on D111049, we're trying to remove the <string> dependency from error handling and replace uses of report_fatal_error(const std::string&) with the Twine() variant which can be forward declared.
Thinlink provides an opportunity to propagate function attributes across modules, enabling additional propagation opportunities.
This change propagates (currently default off, turn on with `disable-thinlto-funcattrs=1`) noRecurse and noUnwind based off of function summaries of the prevailing functions in bottom-up call-graph order. Testing on clang self-build:
1. There's a 35-40% increase in noUnwind functions due to the additional propagation opportunities.
2. Throughput is measured at 10-15% increase in thinlink time which itself is 1.5% of E2E link time.
Implementation-wise this adds the following summary function attributes:
1. noUnwind: function is noUnwind
2. mayThrow: function contains a non-call instruction that `Instruction::mayThrow` returns true on (e.g. windows SEH instructions)
3. hasUnknownCall: function contains calls that don't make it into the summary call-graph thus should not be propagated from (e.g. indirect for now, could add no-opt functions as well)
Testing:
Clang self-build passes and 2nd stage build passes check-all
ninja check-all with newly added tests passing
Reviewed By: tejohnson
Differential Revision: https://reviews.llvm.org/D36850
Pulled out the OptimizationLevel class from PassBuilder in order to be able to access it from within the PassManager and avoid include conflicts.
Reviewed By: mtrofin
Differential Revision: https://reviews.llvm.org/D107025
