Re: [PATCH 3/3] cfgcleanup: small performance improvement in some cases

2025-05-16 Thread Richard Biener
On Tue, May 13, 2025 at 12:47 AM Andrew Pinski wrote: > > With some functions, there might be the case where every stack variable is a > live at the end of a > basic block. If that is the case then it is known that all stack variables > will conflict with each > other so there is no reason to go

[PATCH 3/3] cfgcleanup: small performance improvement in some cases

2025-05-12 Thread Andrew Pinski
With some functions, there might be the case where every stack variable is a live at the end of a basic block. If that is the case then it is known that all stack variables will conflict with each other so there is no reason to go through figuring out what variables conflict with each other. Sin