[PATCH] D145171: [FSAFDO] Improve FS discriminator encoding

[Rong Xu via llvm-commits]

On Thu, Mar 9, 2023 at 10:21 AM Hongtao Yu via Phabricator <
reviews at reviews.llvm.org> wrote:

> hoy added a comment.
>
> In D145171#4182014 <https://reviews.llvm.org/D145171#4182014>, @xur wrote:
>
> > The total improvement really depends on programs. On average, I would
> say 1.5% to 2.0%.  For some programs, like clang itself, FSAFDO improve ~3%
> over AFDO.
> > There is also a small change on the create_llvm_prof tool side. We will
> update the tool soon.
>
> The numbers are exciting! I'll let you know our numbers once ready.
>
> BTW, I'll started the integration with CSSPGO base off here. Pseudo probes
> have a potential to allow more bits for the FS discriminator and additional
> FS passes. Implementation-wise, the potential comes from placing the FS
> discriminator field as a separate int64 operand of the pseudo probe
> intrinsic, instead of using the existing dbg metadata. I would like your
> thoughts about whether it's worth the diversion. Thanks.
>
Interesting. Having more bits for different rounds definitely helps -- we
can encode more information to detect mismatch or do a partial match.
Looking forward to seeing your CSSPGO work in this regard.

In Google, we are also considering extending the discriminator in dbg info.
The DWARF standards specified DW_LNE_set_discriminator to take an ULEB128
integer. So in theory we can use arbitrarily large integer types. Extending
the discriminator to 64 bits in LLVM is doable.

>
>
> CHANGES SINCE LAST ACTION
>   https://reviews.llvm.org/D145171/new/
>
> https://reviews.llvm.org/D145171
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20230309/8c5a2ddf/attachment.html>