Re: [PATCH] [x86] reenable dword MOVE_MAX for better memmove inlining

[Alexandre Oliva <oliva@adacore.com>]

On May 25, 2023, Richard Biener <richard.guenther@gmail.com> wrote:

> I mean we could do what RTL expansion would do later and do
> by-pieces, thus emit multiple loads/stores but not n loads and then
> n stores but interleaved.

That wouldn't help e.g. gcc.dg/memcpy-6.c's fold_move_8, because
MOVE_MAX and MOVE_MAX_PIECES currently limits inline expansion to 4
bytes on x86 without SSE, both in gimple and RTL, and interleaved loads
and stores wouldn't help with memmove.  We can't fix that by changing
code that uses MOVE_MAX and/or MOVE_MAX_PIECES, when these limits are
set too low.

I'm also concerned that doing more such expansion in gimple folding
would be reversed in later gimple passes.  That's good in that it would
enable efficient rtl movmem/cpymem instruction selection, but it's not
clear to me that there would generally be benefits to such early
open-coding in gimple.

-- 
Alexandre Oliva, happy hacker                https://FSFLA.org/blogs/lxo/
   Free Software Activist                       GNU Toolchain Engineer
Disinformation flourishes because many people care deeply about injustice
but very few check the facts.  Ask me about <https://stallmansupport.org>