Currently every time we encounter an indirect call of a known function,
we try to evaluate the inline cost of that function. In case of a
recursion, that evaluation never stops.
The solution I propose is to evaluate only the indirect call of the
function, while any further indirect calls (of a known function) will be
treated just as direct function calls, which, actually, never tries to
evaluate the call.
Fixes PR35469.
Differential Revision: https://reviews.llvm.org/D69349
Currently every time we encounter an indirect call of a known function,
we try to evaluate the inline cost of that function. In case of a
recursion, that evaluation never stops.
The solution presented is to evaluate only the indirect call of the
function, while any further indirect calls (of a known function) will be
treated just as direct function calls, which, actually, never tries to
evaluate the call.
Fixes PR35469.
Differential Revision: https://reviews.llvm.org/D69349
This reverts commit 004ed2b0d1.
Original commit hash 6d03890384
Summary:
This adds a clang option to disable inline line tables. When it is used,
the inliner uses the call site as the location of the inlined function instead of
marking it as an inline location with the function location.
https://reviews.llvm.org/D67723
Summary:
This adds a clang option to disable inline line tables. When it is used,
the inliner uses the call site as the location of the inlined function instead of
marking it as an inline location with the function location.
See https://bugs.llvm.org/show_bug.cgi?id=42344
Reviewers: rnk
Subscribers: hiraditya, cfe-commits, llvm-commits
Tags: #clang, #llvm
Differential Revision: https://reviews.llvm.org/D67723
Summary:
Debug info affects output from "opt -inline", InlineFunction could
not handle the llvm.dbg.value when it exist between alloca
instructions.
Problem was that the first alloca in a sequence of allocas was
handled differently from the subsequence alloca instructions. Now
all static alloca instructions are treated the same (being removed
if the have no uses). So it does not matter if there are dbg
instructions (or any other instructions) in between.
Fix the issue: https://bugs.llvm.org/show_bug.cgi?id=43291k
Patch by: yechunliang (Chris Ye)
Reviewers: bjope, jmorse, vsk, probinson, jdoerfert, mtrofin, aprantl, fhahn
Reviewed By: bjope
Subscribers: uabelho, ormris, aprantl, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D68633
Summary:
The CallAnalyzer::visitSwitchInst has an early exit when the estimated
lower bound of the switch cost will put the overall cost of the inline
above the threshold. However, this code is not correctly estimating the
lower bound for switches that can be transformed into bit tests, leading
to unnecessary lost inlines, and also differing behavior with
optimization remarks enabled.
First, the early exit is controlled by whether ComputeFullInlineCost is
enabled or not, and that in turn is disabled by default but enabled when
enabling -pass-remarks=missed. This by itself wouldn't lead to a
problem, except that as described below, the lower bound can be above
the real lower bound, so we can sometimes get different inline decisions
with inline remarks enabled, which is problematic.
The early exit was added in along with a new switch cost model in D31085.
The reason why this early exit was added is due to a concern one reviewer
raised about compile time for large switches:
https://reviews.llvm.org/D31085?id=94559#inline-276200
However, the code just below there calls
getEstimatedNumberOfCaseClusters, which in turn immediately calls
BasicTTIImpl getEstimatedNumberOfCaseClusters, which in the worst case
does a linear scan of the cases to get the high and low values. The
bit test handling in particular is guarded by whether the number of
cases fits into the max bit width. There is no suggestion that anyone
measured a compile time issue, it appears to be theoretical.
The problem is that the reviewer's comment about the lower bound
calculation is incorrect, specifically in the case of a switch that can
be lowered to a bit test. This isn't followed up on the comment
thread, but the author does add a FIXME to that effect above the early
exit added when they subsequently revised the patch.
As a result, we were incorrectly early exiting and not inlining
functions with switch statements that would be lowered to bit tests in
cases where we were nearing the threshold. Combined with the fact that
this early exit was skipped with opt remarks enabled, this caused
different inlining decisions to be made when -pass-remarks=missed is
enabled to debug the missing inline.
Remove the early exit for the above reasons.
I also copied over an existing AArch64 inlining test to X86, and
adjusted the threshold so that the bit test inline only occurs with the
fix in this patch.
Reviewers: davidxl
Subscribers: eraman, kristof.beyls, haicheng, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D67716
llvm-svn: 372440
Summary:
This tests inlining size thresholds, but relies on the output of running
the full O2 pipeline, making it brittle against changes in unrelated
passes.
Only run the inlining pass and set thresholds on the test RUN line
instead.
Found while investigating D60318.
Reviewers: RKSimon, qcolombet
Subscribers: llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D67349
llvm-svn: 371397
Summary:
In D62801, new function attribute `willreturn` was introduced. In short, a function with `willreturn` is guaranteed to come back to the call site(more precise definition is in LangRef).
In this patch, willreturn is annotated for LLVM intrinsics.
Reviewers: jdoerfert
Reviewed By: jdoerfert
Subscribers: jvesely, nhaehnle, sstefan1, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D64904
llvm-svn: 367184
Summary:
Since the target has no significant advantage of vectorization,
vector instructions bous threshold bonus should be optional.
amdgpu-inline-arg-alloca-cost parameter default value and the target
InliningThresholdMultiplier value tuned then respectively.
Reviewers: arsenm, rampitec
Subscribers: kzhuravl, jvesely, wdng, nhaehnle, yaxunl, dstuttard, tpr, t-tye, eraman, hiraditya, haicheng, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D64642
llvm-svn: 366348
Add "memtag" sanitizer that detects and mitigates stack memory issues
using armv8.5 Memory Tagging Extension.
It is similar in principle to HWASan, which is a software implementation
of the same idea, but there are enough differencies to warrant a new
sanitizer type IMHO. It is also expected to have very different
performance properties.
The new sanitizer does not have a runtime library (it may grow one
later, along with a "debugging" mode). Similar to SafeStack and
StackProtector, the instrumentation pass (in a follow up change) will be
inserted in all cases, but will only affect functions marked with the
new sanitize_memtag attribute.
Reviewers: pcc, hctim, vitalybuka, ostannard
Subscribers: srhines, mehdi_amini, javed.absar, kristof.beyls, hiraditya, cryptoad, steven_wu, dexonsmith, cfe-commits, llvm-commits
Tags: #clang, #llvm
Differential Revision: https://reviews.llvm.org/D64169
llvm-svn: 366123
This reverts r364422 (git commit 1a3dc76186)
The inlining cost calculation is incorrect, leading to stack overflow due to large stack frames from heavy inlining.
llvm-svn: 365000
Summary:
Doing better separation of Cost and Threshold.
Cost counts the abstract complexity of live instructions, while Threshold is an upper bound of complexity that inlining is comfortable to pay.
There are two parts:
- huge 15K last-call-to-static bonus is no longer subtracted from Cost
but rather is now added to Threshold.
That makes much more sense, as the cost of inlining (Cost) is not changed by the fact
that internal function is called once. It only changes the likelyhood of this inlining
being profitable (Threshold).
- bonus for calls proved-to-be-inlinable into callee is no longer subtracted from Cost
but added to Threshold instead.
While calculations are somewhat different, overall InlineResult should stay the same since Cost >= Threshold compares the same.
Reviewers: eraman, greened, chandlerc, yrouban, apilipenko
Reviewed By: apilipenko
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D60740
llvm-svn: 364422
This adds support for unary fneg based on the implementation of BinaryOperator without the soft float FP cost.
Previously we would just delegate to visitUnaryInstruction. I think the only real change is that we will pass the FastMath flags to SimplifyFNeg now.
Differential Revision: https://reviews.llvm.org/D62699
llvm-svn: 362732
Summary: Fneg can be implemented with an xor rather than a function call so we don't need to add the function call overhead. This was pointed out in D62699
Reviewers: efriedma, cameron.mcinally
Reviewed By: efriedma
Subscribers: javed.absar, eraman, hiraditya, haicheng, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D62747
llvm-svn: 362304
AMDGPU uses multiplier 9 for the inline cost. It is taken into account
everywhere except for inline hint threshold. As a result we are penalizing
functions with the inline hint making them less probable to be inlined
than those without the hint. Defaults are 225 for a normal function and
325 for a function with an inline hint. Currently we have effective
threshold 225 * 9 = 2025 for normal functions and just 325 for those with
the hint. That is fixed by this patch.
Differential Revision: https://reviews.llvm.org/D62707
llvm-svn: 362239
When we switch to opaque pointer types we will need some way to describe
how many bytes a 'byval' parameter should occupy on the stack. This adds
a (for now) optional extra type parameter.
If present, the type must match the pointee type of the argument.
The original commit did not remap byval types when linking modules, which broke
LTO. This version fixes that.
Note to front-end maintainers: if this causes test failures, it's probably
because the "byval" attribute is printed after attributes without any parameter
after this change.
llvm-svn: 362128
When we switch to opaque pointer types we will need some way to describe
how many bytes a 'byval' parameter should occupy on the stack. This adds
a (for now) optional extra type parameter.
If present, the type must match the pointee type of the argument.
Note to front-end maintainers: if this causes test failures, it's probably
because the "byval" attribute is printed after attributes without any parameter
after this change.
llvm-svn: 362012
Those two subtarget features were awkward because their semantics are
reversed: each one indicates the _lack_ of support for something in
the architecture, rather than the presence. As a consequence, you
don't get the behavior you want if you combine two sets of feature
bits.
Each SubtargetFeature for an FP architecture version now comes in four
versions, one for each combination of those options. So you can still
say (for example) '+vfp2' in a feature string and it will mean what
it's always meant, but there's a new string '+vfp2d16sp' meaning the
version without those extra options.
A lot of this change is just mechanically replacing positive checks
for the old features with negative checks for the new ones. But one
more interesting change is that I've rearranged getFPUFeatures() so
that the main FPU feature is appended to the output list *before*
rather than after the features derived from the Restriction field, so
that -fp64 and -d32 can override defaults added by the main feature.
Reviewers: dmgreen, samparker, SjoerdMeijer
Subscribers: srhines, javed.absar, eraman, kristof.beyls, hiraditya, zzheng, Petar.Avramovic, cfe-commits, llvm-commits
Tags: #clang, #llvm
Differential Revision: https://reviews.llvm.org/D60691
llvm-svn: 361845
This was skipping GetUnderlyingObject for nonprivate addresses, but an
alloca could also be found through an addrspacecast if it's flat.
llvm-svn: 361649
Summary:
It was supposed that Ref LazyCallGraph::Edge's were being inserted by
inlining, but that doesn't seem to be the case. Instead, it seems that
there was no test for a blockaddress Constant in an instruction that
referenced the function that contained the instruction. Ex:
```
define void @f() {
%1 = alloca i8*, align 8
2:
store i8* blockaddress(@f, %2), i8** %1, align 8
ret void
}
```
When iterating blockaddresses, do not add the function they refer to
back to the worklist if the blockaddress is referring to the contained
function (as opposed to an external function).
Because blockaddress has sligtly different semantics than GNU C's
address of labels, there are 3 cases that can occur with blockaddress,
where only 1 can happen in GNU C due to C's scoping rules:
* blockaddress is within the function it refers to (possible in GNU C).
* blockaddress is within a different function than the one it refers to
(not possible in GNU C).
* blockaddress is used in to declare a global (not possible in GNU C).
The second case is tested in:
```
$ ./llvm/build/unittests/Analysis/AnalysisTests \
--gtest_filter=LazyCallGraphTest.HandleBlockAddress
```
This patch adjusts the iteration of blockaddresses in
LazyCallGraph::visitReferences to not revisit the blockaddresses
function in the first case.
The Linux kernel contains code that's not semantically valid at -O0;
specifically code passed to asm goto. It requires that asm goto be
inline-able. This patch conservatively does not attempt to handle the
more general case of inlining blockaddresses that have non-callbr users
(pr/39560).
https://bugs.llvm.org/show_bug.cgi?id=39560https://bugs.llvm.org/show_bug.cgi?id=40722https://github.com/ClangBuiltLinux/linux/issues/6https://reviews.llvm.org/rL212077
Reviewers: jyknight, eli.friedman, chandlerc
Reviewed By: chandlerc
Subscribers: george.burgess.iv, nathanchance, mgorny, craig.topper, mengxu.gatech, void, mehdi_amini, E5ten, chandlerc, efriedma, eraman, hiraditya, haicheng, pirama, llvm-commits, srhines
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D58260
llvm-svn: 361173
This reverts commit 95805bc425.
I've squashed the test fix into this commit.
[DebugInfo] Update loop metadata for inlined loops
Currently, when a loop is cloned while inlining function (A) into function (B)
the loop metadata is copied and then not modified at all. The loop metadata can
encode the loop's start and end DILocations. Therefore, the new inlined loop in
function (B) may have loop metadata which shows start and end locations residing
in function (A).
This patch ensures loop metadata is updated while inlining so that the start and
end DILocations are given the "inlinedAt" operand. I've also added a regression
test for this.
This fix is required for D60831 because that patch uses loop metadata to
determine the DILocation for the branches of new loop preheaders.
Reviewers: aprantl, dblaikie, anemet
Reviewed By: aprantl
Subscribers: eraman, hiraditya, llvm-commits
Tags: #debug-info, #llvm
Differential Revision: https://reviews.llvm.org/D61933
llvm-svn: 361149
Summary:
Currently, when a loop is cloned while inlining function (A) into function (B) the loop metadata is copied and then not modified at all. The loop metadata can encode the loop's start and end DILocations. Therefore, the new inlined loop in function (B) may have loop metadata which shows start and end locations residing in function (A).
This patch ensures loop metadata is updated while inlining so that the start and end DILocations are given the "inlinedAt" operand. I've also added a regression test for this.
This fix is required for D60831 because that patch uses loop metadata to determine the DILocation for the branches of new loop preheaders.
Reviewers: aprantl, dblaikie, anemet
Reviewed By: aprantl
Subscribers: eraman, hiraditya, llvm-commits
Tags: #debug-info, #llvm
Differential Revision: https://reviews.llvm.org/D61933
llvm-svn: 361132
This is the conservatively correct default. It is always safe to
assume xnack is enabled, but not the converse.
Introduce a feature to blacklist targets where xnack can never be
meaningfully enabled. I'm not sure the targets this is applied to is
100% correct.
llvm-svn: 360903
As it's causing some bot failures (and per request from kbarton).
This reverts commit r358543/ab70da07286e618016e78247e4a24fcb84077fda.
llvm-svn: 358546
The test should really be checking for the property directly in the
code object headers, but there are problems with this. I don't see
this directly represented in the text form, and for the binary
emission this is depending on a function level subtarget feature to
emit a global flag.
llvm-svn: 357558
Since this can be set with s_setreg*, it should not be a subtarget
property. Set a default based on the calling convention, and Introduce
a new amdgpu-dx10-clamp attribute to override this if desired.
Also introduce a new amdgpu-ieee attribute to match.
The values need to match to allow inlining. I think it is OK for the
caller's dx10-clamp attribute to override the callee, but there
doesn't appear to be the infrastructure to do this currently without
definining the attribute in the generic Attributes.td.
Eventually the calling convention lowering will need to insert a mode
switch somewhere for these.
llvm-svn: 357302
The issue here is that we actually allow CGSCC passes to mutate IR (and
therefore invalidate analyses) outside of the current SCC. At a minimum,
we need to support mutating parent and ancestor SCCs to support the
ArgumentPromotion pass which rewrites all calls to a function.
However, the analysis invalidation infrastructure is heavily based
around not needing to invalidate the same IR-unit at multiple levels.
With Loop passes for example, they don't invalidate other Loops. So we
need to customize how we handle CGSCC invalidation. Doing this without
gratuitously re-running analyses is even harder. I've avoided most of
these by using an out-of-band preserved set to accumulate the cross-SCC
invalidation, but it still isn't perfect in the case of re-visiting the
same SCC repeatedly *but* it coming off the worklist. Unclear how
important this use case really is, but I wanted to call it out.
Another wrinkle is that in order for this to successfully propagate to
function analyses, we have to make sure we have a proxy from the SCC to
the Function level. That requires pre-creating the necessary proxy.
The motivating test case now works cleanly and is added for
ArgumentPromotion.
Thanks for the review from Philip and Wei!
Differential Revision: https://reviews.llvm.org/D59869
llvm-svn: 357137
Tuning flags don't have any effect on the available instructions so aren't a good reason to prevent inlining.
There are also some ISA flags that don't have any intrinsics our ABI requirements that we can exclude. I've put only the most basic ones like cmpxchg16b and lahfsahf. These are interesting because they aren't present in all 64-bit CPUs, but we have codegen workarounds when they aren't present.
Loosening these checks can help with scenarios where a caller has a more specific CPU than a callee. The default tuning flags on our generic 'x86-64' CPU can currently make it inline compatible with other CPUs. I've also added an example test for 'nocona' and 'prescott' where 'nocona' is just a 64-bit capable version of 'prescott' but in 32-bit mode they should be completely compatible.
I've based the implementation here of the similar code in AMDGPU.
Differential Revision: https://reviews.llvm.org/D58371
llvm-svn: 354355
This was inhibiting inlining of library functions when clang was
invoking the inliner directly. This is covering a bit of a mess with
subtarget feature handling, and this shouldn't be a subtarget
feature. The behavior is different depending on whether you are using
a -mattr flag in clang, or llc, opt.
llvm-svn: 353899
InlineCost's isInlineViable() is changed to return InlineResult
instead of bool. This provides messages for failure reasons and
allows to get more specific messages for cases where callsites
are not viable for inlining.
Reviewed By: xbolva00, anemet
Differential Revision: https://reviews.llvm.org/D57089
llvm-svn: 352849
The current llvm.mem.parallel_loop_access metadata has a problem in that
it uses LoopIDs. LoopID unfortunately is not loop identifier. It is
neither unique (there's even a regression test assigning the some LoopID
to multiple loops; can otherwise happen if passes such as LoopVersioning
make copies of entire loops) nor persistent (every time a property is
removed/added from a LoopID's MDNode, it will also receive a new LoopID;
this happens e.g. when calling Loop::setLoopAlreadyUnrolled()).
Since most loop transformation passes change the loop attributes (even
if it just to mark that a loop should not be processed again as
llvm.loop.isvectorized does, for the versioned and unversioned loop),
the parallel access information is lost for any subsequent pass.
This patch unlinks LoopIDs and parallel accesses.
llvm.mem.parallel_loop_access metadata on instruction is replaced by
llvm.access.group metadata. llvm.access.group points to a distinct
MDNode with no operands (avoiding the problem to ever need to add/remove
operands), called "access group". Alternatively, it can point to a list
of access groups. The LoopID then has an attribute
llvm.loop.parallel_accesses with all the access groups that are parallel
(no dependencies carries by this loop).
This intentionally avoid any kind of "ID". Loops that are clones/have
their attributes modifies retain the llvm.loop.parallel_accesses
attribute. Access instructions that a cloned point to the same access
group. It is not necessary for each access to have it's own "ID" MDNode,
but those memory access instructions with the same behavior can be
grouped together.
The behavior of llvm.mem.parallel_loop_access is not changed by this
patch, but should be considered deprecated.
Differential Revision: https://reviews.llvm.org/D52116
llvm-svn: 349725
ProfileSampleAccurate is used to indicate the profile has exact match to the
code to be optimized.
Previously ProfileSampleAccurate is handled in ProfileSummaryInfo::isColdCallSite
and ProfileSummaryInfo::isColdBlock. A better solution is to initialize function
entry count to 0 when ProfileSampleAccurate is true, so we don't have to handle
ProfileSampleAccurate in multiple places.
Differential Revision: https://reviews.llvm.org/D55660
llvm-svn: 349088
Lack of an attribute means that the function hasn't been checked for what vector width it requires. So if the caller or the callee doesn't have the attribute we should make sure the combined function after inlining does not have the attribute.
If the caller already doesn't have the attribute we can just avoid adding it. Otherwise if the callee doesn't have the attribute just remove the caller's attribute.
llvm-svn: 347841
We currently seem to underestimate the size of functions with loops in them,
both in terms of absolute code size and in the difficulties of dealing with
such code. (Calls, for example, can be tail merged to further reduce
codesize). At -Oz, we can then increase code size by inlining small loops
multiple times.
This attempts to penalise functions with loops at -Oz by adding a CallPenalty
for each top level loop in the function. It uses LI (and hence DT) to calculate
the number of loops. As we are dealing with minsize, the inline threshold is
small and functions at this point should be relatively small, making the
construction of these cheap.
Differential Revision: https://reviews.llvm.org/D52716
llvm-svn: 346134
in the same round of SCC update.
In https://reviews.llvm.org/rL309784, inline history is added to prevent
infinite inlining across multiple run of inliner and SCC update, but the
history will only be kept when new SCC is actually generated during SCC update.
We found a case that SCC can be split and then merge into itself in the same
round of SCC update, so the same SCC will be pop out from UR.CWorklist and
then added back immediately, without any new SCC generated, that is why the
existing patch cannot catch the infinite inline case.
What the patch does is even if no new SCC is generated, if only the current
SCC appears in UR.CWorklist again, then keep the inline history.
Differential Revision: https://reviews.llvm.org/D52915
llvm-svn: 345103
If you have the string /usr/bin, prior to this patch it would not
be quoted by our YAML serializer. But a string like C:\src would
be, due to the presence of a backslash. This makes the quoting
rules of basically every single file path different depending on
the path syntax (posix vs. Windows).
While technically not required by the YAML specification to quote
forward slashes, when the behavior of paths is inconsistent it
makes it difficult to portably write FileCheck lines that will
work with either kind of path.
Differential Revision: https://reviews.llvm.org/D53169
llvm-svn: 344359
This reverts commit b86c16ad8c97dadc1f529da72a5bb74e9eaed344.
This is being reverted because I forgot to write a useful
commit message, so I'm going to resubmit it with an actual
commit message.
llvm-svn: 344358
In r339636 the alias analysis rules were changed with regards to tail calls
and byval arguments. Previously, tail calls were assumed not to alias
allocas from the current frame. This has been updated, to not assume this
for arguments with the byval attribute.
This patch aligns TailCallElim with the new rule. Tail marking can now be
more aggressive and mark more calls as tails, e.g.:
define void @test() {
%f = alloca %struct.foo
call void @bar(%struct.foo* byval %f)
ret void
}
define void @test2(%struct.foo* byval %f) {
call void @bar(%struct.foo* byval %f)
ret void
}
define void @test3(%struct.foo* byval %f) {
%agg.tmp = alloca %struct.foo
%0 = bitcast %struct.foo* %agg.tmp to i8*
%1 = bitcast %struct.foo* %f to i8*
call void @llvm.memcpy.p0i8.p0i8.i64(i8* %0, i8* %1, i64 40, i1 false)
call void @bar(%struct.foo* byval %agg.tmp)
ret void
}
The problematic case where a byval parameter is captured by a call is still
handled correctly, and will not be marked as a tail (see PR7272).
llvm-svn: 343986
Summary:
rL323619 marks functions that are calling va_end as not viable for
inlining. This patch reverses that since this va_end doesn't need
access to the vriadic arguments list that are saved on the stack, only
va_start does.
Reviewers: efriedma, fhahn
Reviewed By: fhahn
Subscribers: eraman, haicheng, llvm-commits
Differential Revision: https://reviews.llvm.org/D52067
llvm-svn: 342675
Ehud Katz