Re: Fix type of malloc call in trans-expr.cc

2024-11-15 Thread Jakub Jelinek
On Fri, Nov 15, 2024 at 07:45:32AM +, Paul Richard Thomas wrote:
> Hi Jakub,

Honza's catch.

> Good catch! Does it fix any specific PR?

Dunno.  I think even without a PR it can be backported after a while.

Jakub



Re: Fix type of malloc call in trans-expr.cc

2024-11-15 Thread Jan Hubicka
> Hi Jakub,
> 
> Good catch! Does it fix any specific PR?
> 
> If you don't have the time, I would be happy to apply the correction to
> 13-branch through to mainline.

I caught it with my WIP patch to improve tree-ssa-dce.  I am not aware
it can produce wrong code.  It will likely lead to missed optimization.

I was wondering. We are very loose on type checking of calls because K&R
C is evil.  Perhaps we could have flag to enable more strict type
checking for gimple produced by other languages.  We would need to give
up with LTO, but that is fine.

I think bacporting the patch is fine.  It is quite obvious fix.

Honza