@@ -125,7 +125,7 @@ Here's a glance at the effect that a PGOed LLVM has on *rustc*'s performance:
125125
126126[![Performance improvements gained from apply PGO to LLVM][rustc-perf-pgo-llvm-thumb]][rustc-perf-pgo-llvm]
127127
128- [rustc-perf-pgo-llvm-thumb ]: /images/inside-rust/2020-10-30 -exploring-pgo-for-the-rust-compiler/rustc-perf-pgo-llvm-thumb.png
128+ [rustc-perf-pgo-llvm-thumb ]: /images/inside-rust/2020-11-11 -exploring-pgo-for-the-rust-compiler/rustc-perf-pgo-llvm-thumb.png
129129[rustc-perf-pgo-llvm ]: https://perf.rust-lang.org/compare.html?start=pgo-2020-10-30-none&end=pgo-2020-10-30-llvm&stat=instructions%3Au
130130
131131The results are not quite as spectacular as the anecdotal 20% improvement from Clang's documentation;
@@ -134,7 +134,7 @@ Diving more into details shows the expected profile:
134134
135135![Performance improvements gained from apply PGO to LLVM (details)][rustc-perf-pgo-llvm-expanded]
136136
137- [rustc-perf-pgo-llvm-expanded ]: /images/inside-rust/2020-10-30 -exploring-pgo-for-the-rust-compiler/rustc-perf-pgo-llvm-expanded.png
137+ [rustc-perf-pgo-llvm-expanded ]: /images/inside-rust/2020-11-11 -exploring-pgo-for-the-rust-compiler/rustc-perf-pgo-llvm-expanded.png
138138
139139Workloads that spend most of their time in LLVM (e.g. optimized builds) show the most improvement, while workloads that don't invoke LLVM at all (e.g. check builds) also don't profit from a faster LLVM.
140140Let's take a look at how we can take things further by applying PGO to the other half of the compiler.
@@ -213,7 +213,7 @@ As expected the results are similar to when PGO was applied to LLVM: a reduction
213213
214214[ ![ Performance improvements gained from applying PGO to (only) the Rust part of the compiler] [ rustc-perf-pgo-rust-thumb ]] [ rustc-perf-pgo-rust ]
215215
216- [ rustc-perf-pgo-rust-thumb ] : /images/inside-rust/2020-10-30 -exploring-pgo-for-the-rust-compiler/rustc-perf-pgo-rust-thumb.png
216+ [ rustc-perf-pgo-rust-thumb ] : /images/inside-rust/2020-11-11 -exploring-pgo-for-the-rust-compiler/rustc-perf-pgo-rust-thumb.png
217217[ rustc-perf-pgo-rust ] : https://perf.rust-lang.org/compare.html?start=pgo-2020-10-30-none&end=pgo-2020-10-30-rust&stat=instructions%3Au
218218
219219Because different workloads execute different amounts of Rust code (vs C++/LLVM code), the total reduction can be a lot less for LLVM-heavy cases.
@@ -222,7 +222,7 @@ On the other hand, a *check* build or an *incr-unchanged* build spends almost no
222222
223223![ Performance improvements gained from applying PGO to (only) the Rust part of the compiler (details)] [ rustc-perf-pgo-rust-expanded ]
224224
225- [ rustc-perf-pgo-rust-expanded ] : /images/inside-rust/2020-10-30 -exploring-pgo-for-the-rust-compiler/rustc-perf-pgo-rust-expanded.png
225+ [ rustc-perf-pgo-rust-expanded ] : /images/inside-rust/2020-11-11 -exploring-pgo-for-the-rust-compiler/rustc-perf-pgo-rust-expanded.png
226226
227227
228228### Can we apply PGO to Rust and LLVM at the same time?
@@ -261,7 +261,7 @@ After roughly eight hours to complete both the PGO and the non-PGO versions of t
261261
262262[ ![ Wall time improvements gained from applying PGO to the entire compiler] [ rustc-perf-pgo-both-walltime-thumb ]] [ rustc-perf-pgo-both-walltime ]
263263
264- [ rustc-perf-pgo-both-walltime-thumb ] : /images/inside-rust/2020-10-30 -exploring-pgo-for-the-rust-compiler/rustc-perf-pgo-both-walltime-thumb.png
264+ [ rustc-perf-pgo-both-walltime-thumb ] : /images/inside-rust/2020-11-11 -exploring-pgo-for-the-rust-compiler/rustc-perf-pgo-both-walltime-thumb.png
265265[ rustc-perf-pgo-both-walltime ] : https://perf.rust-lang.org/compare.html?start=pgo-2020-10-30-none-20&end=pgo-2020-10-30-both-20&stat=wall-time
266266
267267As you can see we get a 10-16% reduction of build times almost across the board for real world test cases.
0 commit comments