search for: preinline

...k = anX.getK(); or an alternative mechanism. The GCC compiler has the options '-finstrument-functions-exclude-file-list' and '-finstrument-functions-exclude-function-list' for this purpose, but these are not available in CLang/LLVM. I will experiment with the '-mllvm -disable-preinline' option, thanks for telling me about this too. All the best, MartinO -----Original Message----- From: vsk at apple.com [mailto:vsk at apple.com] Sent: 13 December 2016 23:12 To: Martin J. O'Riordan Cc: LLVM Developers Subject: Re: [llvm-dev] Reining in profile instrumentation >...

...ck {1, 1, 1}. That said, I don't know what kinds of issues this would cause in practice. I'd really like to hear about how the performance of your optimized application changes when you turn pre-inlining on during the instrumentation step. You can experiment with this with -mllvm -disable-preinline. > My question is whether there are any mechanisms in LLVM to control what functions get instrumented; for instance ‘#pragma’s that can be added to the code, especially headers, that can be used to disable the instrumentation of large groups of functions. Or an option to remove the instrumen...

...n even suffers slightly. Which is suprising because for C++ code using Clang's version of IR-level instrumentation & PGO brings signifcant gains (up to 5-10% from what I've seen in benchmarks for Firefox). One thing we noticed early on is that disabling the pre-inlining pass (`-disable-preinline`) seems to consistently improve the situation for Rust code. Doing that we sometimes see performance wins of almost 1% over not using PGO. This again is very different to C++ where disabling this pass causes dramatic performance loses for the Firefox benchmarks. And 1% performance improvement is st...

...> beginning of PassManagerBuilder::populateModulePassManager(), namely > >>> PreProfileInlinerPass and PGOLateInstrumentationPass. > >>> > >>> > >>> 4.1 Pre-inline pass: > >>> > >>> It is controlled by back-end option "-preinline" and > >>> "-disable-preinline". If the user specifies any llvm option of > >>> "-fprofile-late-instr-{generate|use}, option "-mllvm -preinline" will > be > >>> automatically inserted in the driver.. To disable the pre-inliner whe...

...is requires a slight profile format change. In our prototype implementation, two new passes are added in the beginning of PassManagerBuilder::populateModulePassManager(), namely PreProfileInlinerPass and PGOLateInstrumentationPass. 4.1 Pre-inline pass: It is controlled by back-end option "-preinline" and "-disable-preinline". If the user specifies any llvm option of "-fprofile-late-instr-{generate|use}, option "-mllvm -preinline" will be automatically inserted in the driver.. To disable the pre-inliner when late instrumentation is enabled, use option "-mllvm...

...e implementation, two new passes are added in the >> beginning of PassManagerBuilder::populateModulePassManager(), namely >> PreProfileInlinerPass and PGOLateInstrumentationPass. >> >> >> 4.1 Pre-inline pass: >> >> It is controlled by back-end option "-preinline" and >> "-disable-preinline". If the user specifies any llvm option of >> "-fprofile-late-instr-{generate|use}, option "-mllvm -preinline" will be >> automatically inserted in the driver.. To disable the pre-inliner when late >> instrumentation...

...or C++ code >> using Clang's version of IR-level instrumentation & PGO brings >> signifcant gains (up to 5-10% from what I've seen in >> benchmarks for Firefox). >> >> One thing we noticed early on is that disabling the >> pre-inlining pass (`-disable-preinline`) seems to consistently >> improve the situation for Rust code. Doing that we sometimes >> see performance wins of almost 1% over not using PGO. This >> again is very different to C++ where disabling this pass >> causes dramatic performance loses for the Firefox benchmarks....

...n of IR-level instrumentation & PGO brings > >>> signifcant gains (up to 5-10% from what I've seen in > >>> benchmarks for Firefox). > >>> > >>> One thing we noticed early on is that disabling the > >>> pre-inlining pass (`-disable-preinline`) seems to consistently > >>> improve the situation for Rust code. Doing that we sometimes > >>> see performance wins of almost 1% over not using PGO. This > >>> again is very different to C++ where disabling this pass > >>> causes dramatic performanc...

...n of IR-level instrumentation & PGO brings > >>> signifcant gains (up to 5-10% from what I've seen in > >>> benchmarks for Firefox). > >>> > >>> One thing we noticed early on is that disabling the > >>> pre-inlining pass (`-disable-preinline`) seems to consistently > >>> improve the situation for Rust code. Doing that we sometimes > >>> see performance wins of almost 1% over not using PGO. This > >>> again is very different to C++ where disabling this pass > >>> causes dramatic performanc...

When either '-pg' or '-finstrument-functions' is used, the compiler inserts the appropriate profiling hooks. This happens prior to inlining, so the hooks remain in place. Normally this is fine, but with C++ and the heavy use of inline functions and templates, there can be a vast number of trivial functions that are normally optimised away; but with the instrumentation hooks

...s > >> >>> signifcant gains (up to 5-10% from what I've seen in > >> >>> benchmarks for Firefox). > >> >>> > >> >>> One thing we noticed early on is that disabling the > >> >>> pre-inlining pass (`-disable-preinline`) seems to consistently > >> >>> improve the situation for Rust code. Doing that we sometimes > >> >>> see performance wins of almost 1% over not using PGO. This > >> >>> again is very different to C++ where disabling this pass > >> &gt...

...ant gains (up to 5-10% from what I've seen in >> > >> >>> benchmarks for Firefox). >> > >> >>> >> > >> >>> One thing we noticed early on is that disabling the >> > >> >>> pre-inlining pass (`-disable-preinline`) seems to consistently >> > >> >>> improve the situation for Rust code. Doing that we sometimes >> > >> >>> see performance wins of almost 1% over not using PGO. This >> > >> >>> again is very different to C++ where disabling...