With https://reviews.llvm.org/D136627, now we have the metrics for profile staleness based on profile statistics, monitoring the profile staleness in real-time can help user quickly identify performance issues. For a production scenario, the build is usually incremental and if we want the real-time metrics, we should store/cache all the old object's metrics somewhere and pull them in a post-build time. To make it more convenient, this patch add an option to persist them into the object binary, the metrics can be reported right away by decoding the binary rather than polling the previous stdout/stderrs from a cache system.
For implementation, it writes the statistics first into a new metadata section(llvm.stats) then encode into a special ELF `.llvm_stats` section. The section data is formatted as a list of key/value pair so that future statistics can be easily extended. This is also under a new switch(`-persist-profile-staleness`)
In terms of size overhead, the metrics are computed at module level, so the size overhead should be small, measured on one of our internal service, it costs less than < 1MB for a 10GB+ binary.
Reviewed By: wenlei
Differential Revision: https://reviews.llvm.org/D136698
This switches everything to use the memory attribute proposed in
https://discourse.llvm.org/t/rfc-unify-memory-effect-attributes/65579.
The old argmemonly, inaccessiblememonly and inaccessiblemem_or_argmemonly
attributes are dropped. The readnone, readonly and writeonly attributes
are restricted to parameters only.
The old attributes are auto-upgraded both in bitcode and IR.
The bitcode upgrade is a policy requirement that has to be retained
indefinitely. The IR upgrade is mainly there so it's not necessary
to update all tests using memory attributes in this patch, which
is already large enough. We could drop that part after migrating
tests, or retain it longer term, to make it easier to import IR
from older LLVM versions.
High-level Function/CallBase APIs like doesNotAccessMemory() or
setDoesNotAccessMemory() are mapped transparently to the memory
attribute. Code that directly manipulates attributes (e.g. via
AttributeList) on the other hand needs to switch to working with
the memory attribute instead.
Differential Revision: https://reviews.llvm.org/D135780
Currently pseudo probe encoding for a function is like:
- For the first probe, a relocation from it to its physical position in the code body
- For subsequent probes, an incremental offset from the current probe to the previous probe
The relocation could potentially cause relocation overflow during link time. I'm now replacing it with an offset from the first probe to the function start address.
A source function could be lowered into multiple binary functions due to outlining (e.g, coro-split). Since those binary function have independent link-time layout, to really avoid relocations from .pseudo_probe sections to .text sections, the offset to replace with should really be the offset from the probe's enclosing binary function, rather than from the entry of the source function. This requires some changes to previous section-based emission scheme which now switches to be function-based. The assembly form of pseudo probe directive is also changed correspondingly, i.e, reflecting the binary function name.
Most of the source functions end up with only one binary function. For those don't, a sentinel probe is emitted for each of the binary functions with a different name from the source. The sentinel probe indicates the binary function name to differentiate subsequent probes from the ones from a different binary function. For examples, given source function
```
Foo() {
…
Probe 1
…
Probe 2
}
```
If it is transformed into two binary functions:
```
Foo:
…
Foo.outlined:
…
```
The encoding for the two binary functions will be separate:
```
GUID of Foo
Probe 1
GUID of Foo
Sentinel probe of Foo.outlined
Probe 2
```
Then probe1 will be decoded against binary `Foo`'s address, and Probe 2 will be decoded against `Foo.outlined`. The sentinel probe of `Foo.outlined` makes sure there's not accidental relocation from `Foo.outlined`'s probes to `Foo`'s entry address.
On the BOLT side, to be minimal intrusive, the pseudo probe re-encoding sticks with the old encoding format. This is fine since unlike linker, Bolt processes the pseudo probe section as a whole and it is free from relocation overflow issues.
The change is downwards compatible as long as there's no mixed use of the old encoding and the new encoding.
Reviewed By: wenlei, maksfb
Differential Revision: https://reviews.llvm.org/D135912
Differential Revision: https://reviews.llvm.org/D135914
Differential Revision: https://reviews.llvm.org/D136394
When a profile is stale and profile mismatch could happen, the mismatched samples are discarded, so we'd like to compute the mismatch metrics to quantify how stale the profile is, which will suggest user to refresh the profile if the number is high.
Two sets of metrics are introduced here:
- (Num_of_mismatched_funchash/Total_profiled_funchash), (Samples_of_mismached_func_hash / Samples_of_profiled_function) : Here it leverages the FunctionSamples's checksums attribute which is a feature of pseudo probe. When the source code CFG changes, the function checksums will be different, later sample loader will discard the whole functions' samples, this metrics can show the percentage of samples are discarded due to this.
- (Num_of_mismatched_callsite/Total_profiled_callsite), (Samples_of_mismached_callsite / Samples_of_profiled_callsite) : This shows how many mismatching for the callsite location as callsite location mismatch will affect the inlining which is highly correlated with the performance. It goes through all the callsite location in the IR and profile, use the call target name to match, report the num of samples in the profile that doesn't match a IR callsite.
This is implemented in a new class(SampleProfileMatcher) and under a switch("--report-profile-staleness"), we plan to extend it with a fuzzy profile matching feature in the future.
Reviewed By: hoy, wenlei, davidxl
Differential Revision: https://reviews.llvm.org/D136627
MisExpect was occasionally crashing under SampleProfiling, due to a division by zero.
We worked around that in D124302 by changing the assert to an early return.
This patch is intended to add a test case for the crashing scenario and
re-enable MisExpect for SampleProfiling.
Reviewed By: tejohnson
Differential Revision: https://reviews.llvm.org/D124481
Probe-based profile leads to a better performance when combined with profi and ext-tsp block layout. I'm turning them on by default.
Reviewed By: wenlei
Differential Revision: https://reviews.llvm.org/D122442
When --disable-sample-loader-inlining is true, skip inline transformation, but merge profiles of inlined instances to outlining versions.
Differential Revision: https://reviews.llvm.org/D121862
The priority-based inliner currenlty uses block count combined with callee entry count to drive callsite inlining. This doesn't work well with LTO where postlink inlining is driven by prelink-annotated block count which could be based on the merge of all context profiles. I'm fixing it by using callee profile entry count only which should be context-sensitive.
I'm seeing 0.2% perf improvment for one of our internal large benchmarks with probe-based non-CS profile.
Reviewed By: wenlei
Differential Revision: https://reviews.llvm.org/D120784
Trimming whitespace or carriage returns from symbols allows this code
to work on Windows and makes it match other places symbol lists are
handled.
Reviewed By: MaskRay
Differential Revision: https://reviews.llvm.org/D117570
This is an extension of **profi** post-processing step that rebalances counts
in CFGs that have basic blocks w/o probes (aka "unknown" blocks). Specifically,
the new version finds many more "unknown" subgraphs and marks more "unknown"
basic blocks as hot (which prevents unwanted optimization passes).
I see up to 0.5% perf on some (large) binaries, e.g., clang-10 and gcc-8.
The algorithm is still linear and yields no build time overhead.
CSSPGO currently employs a flat profile format for context-sensitive profiles. Such a flat profile allows for precisely manipulating contexts that is either inlined or not inlined. This is a benefit over the nested profile format used by non-CS AutoFDO. A downside of this is the longer build time due to parsing the indexing the full CS contexts.
For a CS flat profile, though only the context profiles relevant to a module are loaded when that module is compiled, the cost to figure out what profiles are relevant is noticeably high when there're many contexts, since the sample reader will need to scan all context strings anyway. On the contrary, a nested function profile has its related inline subcontexts isolated from other unrelated contexts. Therefore when compiling a set of functions, unrelated contexts will never need to be scanned.
In this change we are exploring using nested profile format for CSSPGO. This is expected to work based on an assumption that with a preinliner-computed profile all contexts are precomputed and expected to be inlined by the compiler. Contexts not expected to be inlined will be cut off and returned to corresponding base profiles (for top-level outlined functions). This naturally forms a nested profile where all nested contexts are expected to be inlined. The compiler will less likely optimize on derived contexts that are not precomputed.
A CS-nested profile will look exactly the same with regular nested profile except that each nested profile can come with an attributes. With pseudo probes, a nested profile shown as below can also have a CFG checksum.
```
main:1968679:12
2: 24
3: 28 _Z5funcAi:18
3.1: 28 _Z5funcBi:30
3: _Z5funcAi:1467398
0: 10
1: 10 _Z8funcLeafi:11
3: 24
1: _Z8funcLeafi:1467299
0: 6
1: 6
3: 287884
4: 287864 _Z3fibi:315608
15: 23
!CFGChecksum: 138828622701
!Attributes: 2
!CFGChecksum: 281479271677951
!Attributes: 2
```
Specific work included in this change:
- A recursive profile converter to convert CS flat profile to nested profile.
- Extend function checksum and attribute metadata to be stored in nested way for text profile and extbinary profile.
- Unifiy sample loader inliner path for CS and preinlined nested profile.
- Changes in the sample loader to support probe-based nested profile.
I've seen promising results regarding build time. A nested profile can result in a 20% shorter build time than a CS flat profile while keep an on-par performance. This is with -duplicate-contexts-into-base=1.
Test Plan:
Reviewed By: wenlei
Differential Revision: https://reviews.llvm.org/D115205
This is a continuation of D109860 and D109903.
An important challenge for profile inference is caused by the fact that the
sample profile is collected on a fully optimized binary, while the block and
edge frequencies are consumed on an early stage of the compilation that operates
with a non-optimized IR. As a result, some of the basic blocks may not have
associated sample counts, and it is up to the algorithm to deduce missing
frequencies. The problem is illustrated in the figure where three basic
blocks are not present in the optimized binary and hence, receive no samples
during profiling.
We found that it is beneficial to treat all such blocks equally. Otherwise the
compiler may decide that some blocks are “cold” and apply undesirable
optimizations (e.g., hot-cold splitting) regressing the performance. Therefore,
we want to distribute the counts evenly along the blocks with missing samples.
This is achieved by a post-processing step that identifies "dangling" subgraphs
consisting of basic blocks with no sampled counts; once the subgraphs are
found, we rebalance the flow so as every branch probability is 50:50 within the
subgraphs.
Our experiments indicate up to 1% performance win using the optimization on
some binaries and a significant improvement in the quality of profile counts
(when compared to ground-truth instrumentation-based counts)
{F19093045}
Reviewed By: hoy
Differential Revision: https://reviews.llvm.org/D109980
This is a continuation of D109860.
Traditional flow-based algorithms cannot guarantee that the resulting edge
frequencies correspond to a *connected* flow in the control-flow graph. For
example, for an instance in the attached figure, a flow-based (or any other)
inference algorithm may produce an output in which the hot loop is disconnected
from the entry block (refer to the rightmost graph in the figure). Furthermore,
creating a connected minimum-cost maximum flow is a computationally NP-hard
problem. Hence, we apply a post-processing adjustments to the computed flow
by connecting all isolated flow components ("islands").
This feature helps to keep all blocks with sample counts connected and results
in significant performance wins for some binaries.
{F19077343}
Reviewed By: hoy
Differential Revision: https://reviews.llvm.org/D109903
The benefits of sampling-based PGO crucially depends on the quality of profile
data. This diff implements a flow-based algorithm, called profi, that helps to
overcome the inaccuracies in a profile after it is collected.
Profi is an extended and significantly re-engineered classic MCMF (min-cost
max-flow) approach suggested by Levin, Newman, and Haber [2008, Complementing
missing and inaccurate profiling using a minimum cost circulation algorithm]. It
models profile inference as an optimization problem on a control-flow graph with
the objectives and constraints capturing the desired properties of profile data.
Three important challenges that are being solved by profi:
- "fixing" errors in profiles caused by sampling;
- converting basic block counts to edge frequencies (branch probabilities);
- dealing with "dangling" blocks having no samples in the profile.
The main implementation (and required docs) are in SampleProfileInference.cpp.
The worst-time complexity is quadratic in the number of blocks in a function,
O(|V|^2). However a careful engineering and extensive evaluation shows that
the running time is (slightly) super-linear. In particular, instances with
1000 blocks are solved within 0.1 second.
The algorithm has been extensively tested internally on prod workloads,
significantly improving the quality of generated profile data and providing
speedups in the range from 0% to 5%. For "smaller" benchmarks (SPEC06/17), it
generally improves the performance (with a few outliers) but extra work in
the compiler might be needed to re-tune existing optimization passes relying on
profile counts.
UPD Dec 1st 2021:
- synced the declaration and definition of the option `SampleProfileUseProfi ` to use type `cl::opt<bool`;
- added `inline` for `SampleProfileInference<BT>::findUnlikelyJumps` and `SampleProfileInference<BT>::isExit` to avoid linking problems on windows.
Reviewed By: wenlei, hoy
Differential Revision: https://reviews.llvm.org/D109860
For nodes that are in a cycle of a profiled call graph, the current order the underlying scc_iter computes purely depends on how those nodes are reached from outside the SCC and inside the SCC, based on the Tarjan algorithm. This does not honor profile edge hotness, thus does not gurantee hot callsites to be inlined prior to cold callsites. To mitigate that, I'm adding an extra sorter on top of scc_iter to sort scc functions in the order of callsite hotness, instead of changing the internal of scc_iter.
Sorting on callsite hotness can be optimally based on detecting cycles on a directed call graph, i.e, to remove the coldest edge until a cycle is broken. However, detecting cycles isn't cheap. I'm using an MST-based approach which is faster and appear to deliver some performance wins.
Reviewed By: wenlei
Differential Revision: https://reviews.llvm.org/D114204
The benefits of sampling-based PGO crucially depends on the quality of profile
data. This diff implements a flow-based algorithm, called profi, that helps to
overcome the inaccuracies in a profile after it is collected.
Profi is an extended and significantly re-engineered classic MCMF (min-cost
max-flow) approach suggested by Levin, Newman, and Haber [2008, Complementing
missing and inaccurate profiling using a minimum cost circulation algorithm]. It
models profile inference as an optimization problem on a control-flow graph with
the objectives and constraints capturing the desired properties of profile data.
Three important challenges that are being solved by profi:
- "fixing" errors in profiles caused by sampling;
- converting basic block counts to edge frequencies (branch probabilities);
- dealing with "dangling" blocks having no samples in the profile.
The main implementation (and required docs) are in SampleProfileInference.cpp.
The worst-time complexity is quadratic in the number of blocks in a function,
O(|V|^2). However a careful engineering and extensive evaluation shows that
the running time is (slightly) super-linear. In particular, instances with
1000 blocks are solved within 0.1 second.
The algorithm has been extensively tested internally on prod workloads,
significantly improving the quality of generated profile data and providing
speedups in the range from 0% to 5%. For "smaller" benchmarks (SPEC06/17), it
generally improves the performance (with a few outliers) but extra work in
the compiler might be needed to re-tune existing optimization passes relying on
profile counts.
Reviewed By: wenlei, hoy
Differential Revision: https://reviews.llvm.org/D109860
The benefits of sampling-based PGO crucially depends on the quality of profile
data. This diff implements a flow-based algorithm, called profi, that helps to
overcome the inaccuracies in a profile after it is collected.
Profi is an extended and significantly re-engineered classic MCMF (min-cost
max-flow) approach suggested by Levin, Newman, and Haber [2008, Complementing
missing and inaccurate profiling using a minimum cost circulation algorithm]. It
models profile inference as an optimization problem on a control-flow graph with
the objectives and constraints capturing the desired properties of profile data.
Three important challenges that are being solved by profi:
- "fixing" errors in profiles caused by sampling;
- converting basic block counts to edge frequencies (branch probabilities);
- dealing with "dangling" blocks having no samples in the profile.
The main implementation (and required docs) are in SampleProfileInference.cpp.
The worst-time complexity is quadratic in the number of blocks in a function,
O(|V|^2). However a careful engineering and extensive evaluation shows that
the running time is (slightly) super-linear. In particular, instances with
1000 blocks are solved within 0.1 second.
The algorithm has been extensively tested internally on prod workloads,
significantly improving the quality of generated profile data and providing
speedups in the range from 0% to 5%. For "smaller" benchmarks (SPEC06/17), it
generally improves the performance (with a few outliers) but extra work in
the compiler might be needed to re-tune existing optimization passes relying on
profile counts.
Reviewed By: wenlei, hoy
Differential Revision: https://reviews.llvm.org/D109860
Be more consistent in the naming convention for the various RET instructions to specify in terms of bitwidth.
Helps prevent future scheduler model mismatches like those that were only addressed in D44687.
Differential Revision: https://reviews.llvm.org/D113302
When linkage name isn't available in dwarf (ususally the case of C code), looking up callee samples should be based on the dwarf name instead of using an empty string.
Also fixing a test issue where using empty string to look up callee samples accidentally returns the correct samples because it is treated as indirect call.
Reviewed By: wenlei
Differential Revision: https://reviews.llvm.org/D112948
Adds the following switches:
1. --sample-profile-inline-replay-fallback/--cgscc-inline-replay-fallback: controls what the replay advisor does for inline sites that are not present in the replay. Options are:
1. Original: defers to original advisor
2. AlwaysInline: inline all sites not in replay
3. NeverInline: inline no sites not in replay
2. --sample-profile-inline-replay-format/--cgscc-inline-replay-format: controls what format should be generated to match against the replay remarks. Options are:
1. Line
2. LineColumn
3. LineDiscriminator
4. LineColumnDiscriminator
Adds support for negative inlining decisions. These are denoted by "will not be inlined into" as compared to the positive "inlined into" in the remarks.
All of these together with the previous `--sample-profile-inline-replay-scope/--cgscc-inline-replay-scope` allow tweaking in how to apply replay. In my testing, I'm using:
1. --sample-profile-inline-replay-scope/--cgscc-inline-replay-scope = Function to only replay on a function
2. --sample-profile-inline-replay-fallback/--cgscc-inline-replay-fallback = NeverInline since I'm feeding in only positive remarks to the replay system
3. --sample-profile-inline-replay-format/--cgscc-inline-replay-format = Line since I'm generating the remarks from DWARF information from GCC which can conflict quite heavily in column number compared to Clang
An alternative configuration could be to do Function, AlwaysInline, Line fallback with negative remarks which closer matches the final call-sites. Note that this can lead to unbounded inlining if a negative remark doesn't match/exist for one reason or another.
Updated various tests to cover the new switches and negative remarks
Testing:
ninja check-all
Reviewed By: wenlei, mtrofin
Differential Revision: https://reviews.llvm.org/D112040
Replay in sample profiling needs to be asked on candidates that may not have counts or below the threshold. If replay is in effect for a function make sure these are captured and also imported during thinLTO.
Testing:
ninja check-all
Reviewed By: wenlei
Differential Revision: https://reviews.llvm.org/D112033
Operating systems such as AIX allow open and read on directories, passing in a direcotry as the replay file triggers `Invalid remark format` instead of `Could not open remarks file: Is a directory`. This patch substitutes the directory with a non-existent filename. The current filecheck should still work as-is.
Reviewed By: modimo, Whitney
Differential Revision: https://reviews.llvm.org/D112745
The goal is to allow grafting an inline tree from Clang or GCC into a new compilation without affecting other functions. For GCC, we're doing this by extracting the inline tree from dwarf information and generating the equivalent remarks.
This allows easier side-by-side asm analysis and a trial way to see if a particular inlining setup provides benefits by itself.
Testing:
ninja check-all
Reviewed By: wenlei, mtrofin
Differential Revision: https://reviews.llvm.org/D110658
Fix a bug when getInlineCost incorrectly returns a
cost/threshold pair instead of an explicit never inline.
Reviewed By: mtrofin
Differential Revision: https://reviews.llvm.org/D111687
This patch continues unblocking optimizations that are blocked by pseudo probe instrumentation.
Not exactly like DbgIntrinsics, PseudoProbe intrinsic has other attributes (such as mayread, maywrite, mayhaveSideEffect) that can block optimizations. The issues fixed are:
- Flipped default param of getFirstNonPHIOrDbg API to skip pseudo probes
- Unblocked CSE by avoiding pseudo probe from clobbering memory SSA
- Unblocked induction variable simpliciation
- Allow empty loop deletion by treating probe intrinsic isDroppable
- Some refactoring.
Reviewed By: wenlei
Differential Revision: https://reviews.llvm.org/D110847
Currenlty PseudoProbeInserter is a pass conditioned on a target switch. It works well with a single clang invocation. It doesn't work so well when the backend is called separately (i.e, through the linker or llc), where user has always to pass -pseudo-probe-for-profiling explictly. I'm making the pass a default pass that requires no command line arg to trigger, but will be actually run depending on whether the CU comes with `llvm.pseudo_probe_desc` metadata.
Reviewed By: wenlei
Differential Revision: https://reviews.llvm.org/D110209
A couple tweaks to
1. allow more thinlto importing by excluding probe intrinsics from IR size in module summary
2. Allow general default attributes (nofree nosync nounwind) for pseudo probe intrinsic. Without those attributes, pseudo probes will be basically treated as unknown calls which will in turn block their containing functions from annotated with those attributes.
Reviewed By: wenlei
Differential Revision: https://reviews.llvm.org/D109976
Bjorn Pettersson