The associated function can be a nullptr if it is an indirect call.
This causes a crash in `CheckCallee` which always assumes the callee
is a valid pointer.
Fix#66904.
The builtin_expect(), and C++20's likely, unlikely attributes assign branch_weights to annotated branches.
This patch adds the the ability to query branch !prof metadata and improve static analysis based on that.
Fixes: https://github.com/llvm/llvm-project/issues/64998
Reviewers: tejohnson, efriedma
Differential Revision: https://reviews.llvm.org/D159336
Continuing the patch series to get rid of debug intrinsics [0], instruction
insertion needs to be done with iterators rather than instruction pointers,
so that we can communicate information in the iterator class. This patch
adds an iterator-taking insertBefore method and converts various call sites
to take iterators. These are all sites where such debug-info needs to be
preserved so that a stage2 clang can be built identically; it's likely that
many more will need to be changed in the future.
At this stage, this is just changing the spelling of a few operations,
which will eventually become signifiant once the debug-info bearing
iterator is used.
[0] https://discourse.llvm.org/t/rfc-instruction-api-changes-needed-to-eliminate-debug-intrinsics-from-ir/68939
Differential Revision: https://reviews.llvm.org/D152537
The Attributor has gained support for indirect calls but it is opt-in.
This patch makes AAKernelInfoCallSite able to handle multiple potential
callees.
Fix crash on RAUW due to locals and globals having different address
spaces. This is the intent of the original code, but it assumes the
alloca address space is 0. This patch fixes the code to check that the
global's address space matches `DL.getAllocaAddrSpace()` instead.
Fixes#65155
`AAAddressSpace` currently only works for `LoadInst` and `StoreInst`
currently. For `StoreInst`, the corresponding use can be the pointer
operand, or value operand, or both. When it is used as value operand, it
can prevent `AMDGPUPromoteAlloca` from optimization in certain cases.
This patch changes the manifest method such that only pointer operand
will be rewritten.
Through the new `Attributor::checkForAllCallees` we can look through
indirect calls and visit all potential callees if they are known. Most
AAs will do that implicitly now via `AACalleeToCallSite`, thus, most AAs
are able to deal with missing callees for call site IR positions.
Differential Revision: https://reviews.llvm.org/D112290
Update `ForceFunctionAttrs` pass to optionally take its input from a csv file, for example, function-level optimization attributes. A subsequent patch will enable the pass pipeline to be aware of these attributes, and this pass will be used to test that is the case. Eventually, the annotations would be driven by an agent, e.g. a machine learning-based policy.
This patch is a part of GSoC 2023, more details can be found [[ https://summerofcode.withgoogle.com/programs/2023/projects/T8rB84Sr | here ]]
Reviewed By: mtrofin, aeubanks
Differential Revision: https://reviews.llvm.org/D155617
As outlined in my proposal of how to get rid of debug intrinsics, this
patch adds a moveBefore method that signals the caller /intends/ the order
of moved instructions is to stay the same. This semantic difference has an
effect on debug-info, as it signals whether debug-info needs to move with
instructions or not.
The patch just replaces a few calls to moveBefore with calls to
moveBeforePreserving -- and the latter just calls the former, so it's all
NFC right now. A future patch will add an implementation of
moveBeforePreserving that takes action to correctly preserve debug-info,
but that's tightly coupled with our non-instruction debug-info
representation that's still being reviewed.
[0] https://discourse.llvm.org/t/rfc-instruction-api-changes-needed-to-eliminate-debug-intrinsics-from-ir/68939
Differential Revision: https://reviews.llvm.org/D156369
This patch replaces any_cast with llvm::any_cast. This in turn allows us
to gracefully switch to std::any in future by forwarding llvm::Any and
llvm::any_cast to:
using Any = std::any;
template <class T> T *any_cast(Any *Value) {
return std::any_cast<T>(Value);
}
respectively.
Without this patch, it's ambiguous whether any_cast refers to
std::any_cast or llvm::any_cast.
As an added bonus, this patch makes it easier to mechanically replace
llvm::any_cast with std::any_cast without affecting other occurrences of
any_cast (e.g. in libcxx).
Added a class to hold such common state. The goal is to both reduce the argument list of other utilities used by `computeImportForModule` (which will be brought as members in a subsequent patch), and to make it easy to extend such state later.
Many AAs translated callee information to the call site explicitly but
they now all use the helper we already had for callee return to call
site return propagation. In a follow up the helper is going to be
extended to handle multiple callees.
A direct call to a function casted to a different type is still not
really an address taken event. We allow the user to opt out of these
now.
Reviewed By: arsenm
Differential Revision: https://reviews.llvm.org/D159149
Also removed them from the header. They are there for test-only. This
simplifies further refactoring (as well as code comprehension)
Differential Revision: https://reviews.llvm.org/D159308
Address feedback in https://reviews.llvm.org/D158817. Since `extractProbe` can be used for both calliste and BB probe, we can leverage this to unify the callsite handling code.
Reviewed By: hoy, wenlei
Differential Revision: https://reviews.llvm.org/D159169
Allow specialization of functions with "dynamic" denormal modes to a
known IEEE or DAZ mode based on callers. This should make it possible
to implement a is-denormal-flushing-enabled test using
llvm.canonicalize and have it be free after LTO.
https://reviews.llvm.org/D156129
/data/home/jiefu/llvm-project/llvm/lib/Transforms/IPO/SampleProfile.cpp:2189:8: error: variable 'IsFuncHashMismatch' set but not used [-Werror,-Wunused-but-set-variable]
bool IsFuncHashMismatch = false;
^
1 error generated.
Accumulating the staleness metrics from per-link is less accurate than doing it from post-link time(assuming we use the offline profile mismatch as baseline), the reason is that there are some duplicated reports for the same functions, for example, one template function could be included in multiple TUs, but in post thin link time, only one function are kept(linkonce_odr) and others are marked as available-externally function. Hence, this change skips reporting the metrics for imported functions(available-externally).
I saw the post-link number is now very close to the offline number(dump the mismatched functions and count the metrics offline based on the entire profile), sightly smaller than offline number due to some missing inlined functions.
Reviewed By: hoy, wenlei
Differential Revision: https://reviews.llvm.org/D156725
Follow-up diff for https://reviews.llvm.org/D158891. Compute the checksum mismatch based on the original nested profile. Additionally, use a recursive way to compute the children mismatched samples in the nested tree even the top-level func checksum is matched.
Reviewed By: hoy, wenlei
Differential Revision: https://reviews.llvm.org/D158900
- Always use flattened profile to find the profile anchors. Since profile under different contexts may have different inlined callsites, to get more profile anchors, we use a merged profile from all the contexts(the flattened profile) to find callsite anchors.
- Compute the staleness metrics based on the original nested profile, as currently once a callsite is mismatched, all its children profile are dropped.(TODO: in future, we can improve to reuse the children valid profile)
Reviewed By: hoy, wenlei
Differential Revision: https://reviews.llvm.org/D158891
As in per-link time, callsites could be optimized out by inlining, we don't have those original call targets in the IR in LTO time. Additionally, the inlined code doesn't actually belong to the original function, the IR locations or pseudo probe parsed from it are incorrect and could mislead the matching later.
This change adds the support to extract the original IR location info from the inlined code, specifically, it make sure to skip all the inlined code that doesn't belong the original function, but before that, it processes the inline frames of the debug info to extract the base frame and recover its callsite and callee target(name).
Measured on some stale profile instances, all showed some perf improvements.
Reviewed By: hoy, wenlei
Differential Revision: https://reviews.llvm.org/D156722
- rename `IRLocation` --> `IRAnchors`, `ProfileLocation` --> `ProfileAnchors`
- reorganize runOnFunction, fact out the finding IR anchors code into `findIRAnchors`
- introduce a new function `findProfileAnchors` to populate the profile related anchors, the result is saved into `ProfileAnchors`, it's later used for both mismatch report and matching, this can avoid to parse the `getBodySamples` and `getCallsiteSamples` for multiple times.
- move the `MatchedCallsiteLocs` stuffs from `findIRAnchors` to `countProfileMismatches` so that all the staleness metrics report are computed in one function.
- move all matching related into `runStaleProfileMatching`, and move all mismatching report into `countProfileMismatches`
Reviewed By: wenlei
Differential Revision: https://reviews.llvm.org/D158817
GlobalValues are often interesting, especially if they have local
linkage. We now track all uses of those and refine potential callees
with it. Effectively, if an internal function cannot reach an indirect
call site, it cannot be a potential callee, even if it has its address
taken.
The Attributor user can now set the closed world flag
(`AttributorConfig.IsClosedWorldModule` or
`-attributor-assume-closed-world`) in order to specialize call edges
based only on available callees. That means, we assume all functions are
known and hence all potential callees must be declared/defined in the
module. We will use this for GPUs and LTO cases, but for now the user
has to set it via a flag.
The code has changed significantly over time making the description
outdated. In this patch I am re-writing the description with an
emphasis to the cost model, where most of the changes have happened.
Differential Revision: https://reviews.llvm.org/D158723
If a potential callee has more arguments than the call site, the values
passed will be `poison`. If the potential callee would exhibit UB for
such `undef` argument, e.g., they are marked `noundef`, we can rule the
potential callee out.
The user can now limit the number of indirect calls specialized for a
given call site with `-attributor-max-specializations-per-call-base=N`
or the AttributorConfig callback. We further attach the `!callee`
metadata if all remaining callees are known.
This patch makes the -force-attribute and -force-remove-attribute flags
in the forceattrs pass additionally work over all the functions in a
module rather than on specific functions. I often
find myself dealing with bitcode from projects with arbitrary build
systems that have applied the optnone attribute for various reasons and
having the ability to do this in opt would be quite convenient. It's
possible to remove the attribute by disassembling to textual IR, running
sed, and then running opt over the result, but this option makes things
just a bit easier.
Differential Revision: https://reviews.llvm.org/D158815
Reviewed-By: nikic
Before, we checked the parallel region only once, and ignored updates in
the KernelInfo for the parallel region that happened later. This caused
us to think nested parallel sections are not present even if they are,
among other things.
The old code did not account for new queries during an update, which
caused us to leave stack RQIs in the map. We are now explicit about
temporary vs non-temporary RQIs.
Fixes: https://github.com/llvm/llvm-project/issues/64959
When we remove barriers, we might need to remove llvm.assume
assumptions as well. However, doing this early, thus in the module pass,
will cause us to miss out on information we might need. There are few
situations we can eliminate barriers across functions, for now we simply
disable elimination of barriers that require assumptions to be removed
during the early module pass.
