Re: [PATCH v7 1/7] Add AutoFDO support for Clang build

[Peter Jung <ptr1337@cachyos.org>]

On 02.11.24 20:46, Peter Jung wrote:
> 
> 
> On 02.11.24 18:51, Rong Xu wrote:
>> Add the build support for using Clang's AutoFDO. Building the kernel
>> with AutoFDO does not reduce the optimization level from the
>> compiler. AutoFDO uses hardware sampling to gather information about
>> the frequency of execution of different code paths within a binary.
>> This information is then used to guide the compiler's optimization
>> decisions, resulting in a more efficient binary. Experiments
>> showed that the kernel can improve up to 10% in latency.
>>
>> The support requires a Clang compiler after LLVM 17. This submission
>> is limited to x86 platforms that support PMU features like LBR on
>> Intel machines and AMD Zen3 BRS. Support for SPE on ARM 1,
>>   and BRBE on ARM 1 is part of planned future work.
>>
>> Here is an example workflow for AutoFDO kernel:
>>
>> 1) Build the kernel on the host machine with LLVM enabled, for example,
>>         $ make menuconfig LLVM=1
>>      Turn on AutoFDO build config:
>>        CONFIG_AUTOFDO_CLANG=y
>>      With a configuration that has LLVM enabled, use the following
>>      command:
>>         scripts/config -e AUTOFDO_CLANG
>>      After getting the config, build with
>>        $ make LLVM=1
>>
>> 2) Install the kernel on the test machine.
>>
>> 3) Run the load tests. The '-c' option in perf specifies the sample
>>     event period. We suggest     using a suitable prime number,
>>     like 500009, for this purpose.
>>     For Intel platforms:
>>        $ perf record -e BR_INST_RETIRED.NEAR_TAKEN:k -a -N -b -c 
>> <count> \
>>          -o <perf_file> -- <loadtest>
>>     For AMD platforms:
>>        The supported system are: Zen3 with BRS, or Zen4 with amd_lbr_v2
>>       For Zen3:
>>        $ cat proc/cpuinfo | grep " brs"
>>        For Zen4:
>>        $ cat proc/cpuinfo | grep amd_lbr_v2
>>        $ perf record --pfm-events RETIRED_TAKEN_BRANCH_INSTRUCTIONS:k 
>> -a \
>>          -N -b -c <count> -o <perf_file> -- <loadtest>
>>
>> 4) (Optional) Download the raw perf file to the host machine.
>>
>> 5) To generate an AutoFDO profile, two offline tools are available:
>>     create_llvm_prof and llvm_profgen. The create_llvm_prof tool is part
>>     of the AutoFDO project and can be found on GitHub
>>     (https://github.com/google/autofdo), version v0.30.1 or later. The
>>     llvm_profgen tool is included in the LLVM compiler itself. It's
>>     important to note that the version of llvm_profgen doesn't need to
>>     match the version of Clang. It needs to be the LLVM 19 release or
>>     later, or from the LLVM trunk.
>>        $ llvm-profgen --kernel --binary=<vmlinux> -- 
>> perfdata=<perf_file> \
>>          -o <profile_file>
>>     or
>>        $ create_llvm_prof --binary=<vmlinux> --profile=<perf_file> \
>>          --format=extbinary --out=<profile_file>
>>
>>     Note that multiple AutoFDO profile files can be merged into one via:
>>        $ llvm-profdata merge -o <profile_file>  <profile_1> ... 
>> <profile_n>
>>
>> 6) Rebuild the kernel using the AutoFDO profile file with the same config
>>     as step 1, (Note CONFIG_AUTOFDO_CLANG needs to be enabled):
>>        $ make LLVM=1 CLANG_AUTOFDO_PROFILE=<profile_file>
>>
>> Co-developed-by: Han Shen<shenhan@google.com>
>> Signed-off-by: Han Shen<shenhan@google.com>
>> Signed-off-by: Rong Xu<xur@google.com>
>> Suggested-by: Sriraman Tallam<tmsriram@google.com>
>> Suggested-by: Krzysztof Pszeniczny<kpszeniczny@google.com>
>> Suggested-by: Nick Desaulniers<ndesaulniers@google.com>
>> Suggested-by: Stephane Eranian<eranian@google.com>
>> Tested-by: Yonghong Song<yonghong.song@linux.dev>
>> Tested-by: Yabin Cui<yabinc@google.com>
>> Tested-by: Nathan Chancellor<nathan@kernel.org>
>> Reviewed-by: Kees Cook<kees@kernel.org>
> 
> Tested-by: Peter Jung <ptr1337@cachyos.org>
> 

The compilations and testing with the "make pacman-pkg" function from 
the kernel worked fine.

One problem I do face:
When I apply a AutoFDO profile together with the PKGBUILD [1] from 
archlinux im running into issues at "module_install" at the packaging.

See following log:
```
make[2]: *** [scripts/Makefile.modinst:125: 
/tmp/makepkg/linux-cachyos-rc-autofdo/pkg/linux-cachyos-rc-autofdo/usr/lib/modules/6.12.0-rc5-5-cachyos-rc-autofdo/kernel/arch/x86/kvm/kvm.ko] 
Error 1
make[2]: *** Deleting file 
'/tmp/makepkg/linux-cachyos-rc-autofdo/pkg/linux-cachyos-rc-autofdo/usr/lib/modules/6.12.0-rc5-5-cachyos-rc-autofdo/kernel/arch/x86/kvm/kvm.ko'
   INSTALL 
/tmp/makepkg/linux-cachyos-rc-autofdo/pkg/linux-cachyos-rc-autofdo/usr/lib/modules/6.12.0-rc5-5-cachyos-rc-autofdo/kernel/crypto/cryptd.ko
make[2]: *** Waiting for unfinished jobs....
```


This can be fixed with removed "INSTALL_MOD_STRIP=1" to the passed 
parameters of module_install.

This explicitly only happens, if a profile is passed - otherwise the 
packaging works without problems.

Regards,

Peter Jung