On Mon, 11 Nov 2024, Alfie Richards wrote:
> I see this code is very unclear in hindsight. The logic of this code
> relies on that FMV functions are only allowed at file scope.
> This should have a DECL_FILE_SCOPE_P check to avoid some of the ridiculous
> cases you mentioned.
If you have an actua
Adding missing CC's
Forwarded Message
Subject: Re: [PATCH v3] C: Support Function multiversionsing in the C
front end
Date: Mon, 11 Nov 2024 12:29:57 +
From: Alfie Richards
To: Joseph Myers
Hi Joseph,
Thank you for the detailed feedback. I am quite j
I should also add: the ACLE specification for the details of how function
multiversioning is supposed to work in terms of interactions of
declarations for different versions in the same or different scopes and
what happens regarding forming composite types seems rather vague. So
maybe it would
On Mon, 4 Nov 2024, alfie.richa...@arm.com wrote:
> /* Subroutine of duplicate_decls. Compare NEWDECL to OLDDECL.
> Returns true if the caller should proceed to merge the two, false
> if OLDDECL should simply be discarded. As a side effect, issues
> @@ -3365,11 +3382,53 @@ pushdecl (tre
Hi,
Joseph pointed out a pretty major flaw that this previously
did not merge fmv declarations and definitons or check properly
for duplicate definitons.
I rewrote the fmv logic and moved it to pushdecl resulting in a
patch that I much prefer.
The logic in pushdecl now detects the FMV case and a
