linux-kernel - Re: [GIT PULL] objtool fixes and updates

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <Z-piwUbLx-OitJrv@gmail.com>
Date: Mon, 31 Mar 2025 11:39:13 +0200
From: Ingo Molnar <mingo@...nel.org>
To: Josh Poimboeuf <jpoimboe@...nel.org>
Cc: Linus Torvalds <torvalds@...ux-foundation.org>,
	linux-kernel@...r.kernel.org
Subject: Re: [GIT PULL] objtool fixes and updates


* Josh Poimboeuf <jpoimboe@...nel.org> wrote:

> On Mon, Mar 31, 2025 at 01:13:55AM +0200, Borislav Petkov wrote:
> > On Sun, Mar 30, 2025 at 03:19:40PM -0700, Linus Torvalds wrote:
> > > On Sat, 29 Mar 2025 at 08:33, Borislav Petkov <bp@...en8.de> wrote:
> > > >
> > > > Btw, test bot complains:
> > > >
> > > > https://lore.kernel.org/r/202503292202.Sge7ZEUc-lkp@intel.com
> > > 
> > > That's not a very helpful error message
> > 
> > I found this:
> > 
> > https://lore.kernel.org/r/202503280703.OARM8SrY-lkp@intel.com
> > 
> > which looks like the original report.
> > 
> > Looks unsolved yet...
> 
> The "new" warning is just the "skipping duplicate warning", which was
> already merged with commit 0a7fb6f07e3a ("objtool: Increase per-function
> WARN_FUNC() rate limit").  So none of the warnings are specific to this
> pull request.
> 
> Tiezhu, can you please look at this warning?
>   
>    arch/loongarch/kernel/traps.o: warning: objtool: show_stack+0xe0: stack state mismatch: reg1[22]=-1+0 reg2[22]=-2-160
>    arch/loongarch/kernel/traps.o: warning: objtool: show_stack+0xe0: stack state mismatch: reg1[23]=-1+0 reg2[23]=-2-152

I'm wondering why LKP didn't pick up the primary warning, only the 
final warning that skips duplicate warnings, which is kinda totally 
useless in isolation:

    `-- arch-loongarch-kernel-traps.o:warning:objtool:show_stack:skipping-duplicate-warning(s)

Maybe we should just do what is below - output a single warning, a bit 
like lockdep.

Or delay the reporting to until the entire pass is finished, and 
output something like:

  warning: objtool: show_stack+0xe0: stack state mismatch: reg1[22]=-1+0 reg2[22]=-2-160 [2+ more warnings]

Note: we shouldn't output the real warning count - just that it's 
multiple warnings that might or might not be resolved by fixing the 
primary warning.

Or something like that.

Thanks,

	Ingo

============>
 tools/objtool/include/objtool/warn.h | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/tools/objtool/include/objtool/warn.h b/tools/objtool/include/objtool/warn.h
index b29ac144e4f5..69dd29a3c267 100644
--- a/tools/objtool/include/objtool/warn.h
+++ b/tools/objtool/include/objtool/warn.h
@@ -57,7 +57,7 @@ static inline char *offstr(struct section *sec, unsigned long offset)
 	free(_str);					\
 })
 
-#define WARN_LIMIT 2
+#define WARN_LIMIT 1
 
 #define WARN_INSN(insn, format, ...)					\
 ({									\
@@ -69,8 +69,6 @@ static inline char *offstr(struct section *sec, unsigned long offset)
 		if (_insn->sym)						\
 			_insn->sym->warnings++;				\
 	} else if (_insn->sym && _insn->sym->warnings == WARN_LIMIT) {	\
-		WARN_FUNC("skipping duplicate warning(s)",		\
-			  _insn->sec, _insn->offset);			\
 		_insn->sym->warnings++;					\
 	}								\
 })