[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-Id: <CF8E694D-D8ED-417F-967F-E83E34E6CBB7@linux.dev>
Date: Mon, 27 Oct 2025 22:33:01 +0100
From: Thorsten Blum <thorsten.blum@...ux.dev>
To: Josh Poimboeuf <jpoimboe@...nel.org>
Cc: Peter Zijlstra <peterz@...radead.org>,
Miroslav Benes <mbenes@...e.cz>,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH] objtool: Re-add resume_play_dead() to noreturn list
On 27. Oct 2025, at 21:37, Josh Poimboeuf wrote:
> Hm? There's no "resume_play_dead" in the diff for 6245ce4ab670.
>
> That was actually done intentionally with a different commit:
>
> 34245659debd ("objtool: Remove superfluous global_noreturns entries")
Sorry, I must have missed this. Please drop this patch then.
It's a bit unfortunate that this is neither documented in
objtool/noreturns.h, where it rather misleadingly says "all functions":
/*
* This is a (sorted!) list of all known __noreturn functions in the kernel.
* ...
*/
nor in objtool/Documentation/objtool.txt:
"A noreturn function must be marked __noreturn in both its declaration
and its definition, and must have a NORETURN() annotation in
tools/objtool/noreturns.h."
Only the __dead_end_function() function comment in objtool/check.c and
the implementation itself make this clear.
Thanks,
Thorsten
Powered by blists - more mailing lists