Meta Continues Working On BOLT'ing The Linux Kernel For Greater Performance

[phoronix]

Phoronix: Meta Continues Working On BOLT'ing The Linux Kernel For Greater Performance

Merged to the LLVM compiler stack two years ago was the BOLT tool for optimizing the layout of generated binaries for offering even greater performance than the likes of Profile Guided Optimizations (PGO) alone. BOLT had been in development for years by Facebook/Meta engineers and has continued to be improved upon for enhancing the code layout of binaries to yield enhanced performance. Recently there's been renewed work on using BOLT to optimize Linux kernel images...

https://www.phoronix.com/news/LLVM-BOLT-Linux-Kernel-2024

[V1tol]

Researchers: Kernel LTO\PGO\WTF speedup
Users: waiting for x86-64-v3 OOB support in distros

[ptr1337]

On CachyOS we already tried with a custom branch of llvm from maksfb to bolt the kernel.
Sadly it failed, at the profile conversion, but we are in contact with the developers and hopefully they will implement soon that it will work.
Im really looking forward to it to have a alternative optimization for the kernel.

Mitigations seem to currently maybe problematic:

perf2bolt -p perf.data -o perf.fdate /usr/lib/modules/6.6.7-1-bolt/build/vmlinux
PERF2BOLT: Starting data aggregation job for perf.data
PERF2BOLT: spawning perf job to read branch events
PERF2BOLT: spawning perf job to read mem events
PERF2BOLT: spawning perf job to read process events
PERF2BOLT: spawning perf job to read task events
BOLT-INFO: Target architecture: x86_64
BOLT-INFO: BOLT version: 2a9654db14584dae6b6a5ccc6c47ac18ce89174c
BOLT-INFO: first alloc address is 0x0
BOLT-INFO: creating new program header table at address 0xffffffff84400000, offset 0xffffffff84400000
BOLT-INFO: static input executable detected
BOLT-INFO: disabling -align-macro-fusion in non-relocation mode
BOLT-INFO: enabling lite mode
BOLT-ERROR: symbol seen in the middle of the function retbleed_untrain_ret. Skipping.
BOLT-ERROR: symbol seen in the middle of the function srso_untrain_ret. Skipping.
BOLT-INFO: pre-processing profile using perf data aggregator
BOLT-WARNING: build-id will not be checked because we could not read one from input binary
PERF2BOLT: waiting for perf task events collection to finish...
PERF2BOLT: parsing perf-script task events output
PERF2BOLT: input binary is associated with 0 PID(s)
PERF2BOLT: waiting for perf events collection to finish...
PERF2BOLT: parse branch events...
PERF2BOLT: read 11895504 samples and 190265770 LBR entries
PERF2BOLT: 0 samples (0.0%) were ignored
PERF2BOLT: traces mismatching disassembled function contents: 0 (0.0%)
PERF2BOLT: out of range traces involving unknown regions: 178370422 (100.0%)
PERF2BOLT: waiting for perf mem events collection to finish...
BOLT-ERROR: while running linux-kernel-rewriter in pre-CFG state: out of bounds while reading .altinstructions​

[ms178]

I hope that PGO will also be a re-newed focus for the Kernel, especially since PGO would further the benefits of LTO and BOLT.