llvm dev - Apr 2021 - Using TBAA type descriptor's name as hints for optimizations

Hello all,

There was a discussion (https://reviews.llvm.org/D100717) about whether
using TBAA type descriptor metadata's type name as hints for optimizations
is okay.

The patch is using the type name to determine whether converting memcpy to
load/store of integer is beneficial or not.
If the memcpy was copying a struct containing a pointer member, this
becomes the source of a lot of inttoptr casts because the later store
forwarding needs them.
The introduced inttoptr casts are later removed by cast elimination
(`inttoptr(ptrtoint p) -> p`), but there are issues (e.g.
http://llvm.org/pr34548) which is deeply related with its validity.

A suggested clarification to LangRef is made at
https://reviews.llvm.org/D101185 as well -
feel free to leave comments here and there.

Thanks,
Juneyoung
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20210424/6f355ab0/attachment.html>

Hi,

I think it's clear that using TBAA or anything else for *hinting* 
optimisation is fine.  The discussion in the review was rooted in two 
problems:

  - In some cases the transform is not valid.  `memcpy` must be a 
type-oblivious copy.  On some platforms (e.g. CHERI, including Arm's 
Morello, and with some language VMs), turning these into a typed 
load-store is not valid.  We lose tracking of pointers in these 
environments.  It is *always* valid to delete metadata and an 
optimisation may not become unsound if metadata is elided.

  - The transform was introducing things that made later analyses worse 
and so was not improving the optimisation pipeline overall.

Of these, the first is more significant.  If a transform is valid in the 
absence of metadata and generates better code in the presence of 
metadata, this is fine (and follows from the description of metadata in 
the LangRef today).  If an optimisation becomes unsound when metadata is 
omitted, this is categorically incorrect.

I don't object to the clarification in the LangRef, but all of this 
follows from the general rules about metadata:

  - Removing all metadata from a module should not cause miscompilations.

  - Adding metadata may alter code generation in any way that does not 
violate the extra semantics defined by the metadata.

David

On 23/04/2021 19:03, Juneyoung Lee via llvm-dev wrote:
> Hello all,
> 
> There was a discussion (https://reviews.llvm.org/D100717 
> <https://reviews.llvm.org/D100717>) about whether using TBAA type 
> descriptor metadata's type name as hints for optimizations is okay.
> 
> The patch is using the type name to determine whether converting memcpy 
> to load/store of integer is beneficial or not.
> If the memcpy was copying a struct containing a pointer member, this 
> becomes the source of a lot of inttoptr casts because the later store 
> forwarding needs them.
> The introduced inttoptr casts are later removed by cast elimination 
> (`inttoptr(ptrtoint p) -> p`), but there are issues (e.g. 
> http://llvm.org/pr34548 <http://llvm.org/pr34548>) which is deeply 
> related with its validity.
> 
> A suggested clarification to LangRef is made at 
> https://reviews.llvm.org/D101185 <https://reviews.llvm.org/D101185>
as
> well -
> feel free to leave comments here and there.
> 
> Thanks,
> Juneyoung
> 
> 
> _______________________________________________
> LLVM Developers mailing list
> llvm-dev at lists.llvm.org
> https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev
>