linux-kernel - Re: [PATCH] kbuild: Demote 'sign-compare' warning to W=2

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: Windows password security audit tool. GUI, reports in PDF.

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

Date:	Tue, 12 Jan 2016 15:05:20 +0100
From:	Arnd Bergmann <arnd@...db.de>
To:	Lee Jones <lee.jones@...aro.org>
Cc:	linux-arm-kernel@...ts.infradead.org, linux-kernel@...r.kernel.org,
	kernel@...inux.com, maxime.coquelin@...com, mmarek@...e.com,
	linux-kbuild@...r.kernel.org,
	Behan Webster <behanw@...verseincode.com>,
	Jan-Simon Möller <dl9pf@....de>
Subject: Re: [PATCH] kbuild: Demote 'sign-compare' warning to W=2

On Tuesday 12 January 2016 13:31:59 Lee Jones wrote:
> Ideally, a kernel compile with W=1 enabled should complete cleanly;
> however, when we run one currently we are presented with ~25k warnings.
> 'sign-compare' accounts for ~22k of those ~25k.
> 
> In this patch we're demoting 'sign-compare' warnings to W=2, with a view
> to fixing the remaining 3k W=1 warnings required for a clean build.
> 
> Cc: Arnd Bergmann <arnd@...db.de>
> Signed-off-by: Lee Jones <lee.jones@...aro.org>
> 

As per our discussion, I'd add that this was inadvertedly introduced
by Behan when he moved the clang specific warnings into an ifdef block
and did not notice that -Wsign-compare was interpreted by both gcc
and clang.

Earlier, it was introduced in just the same way by Jan-Simon as part
of 3d3d6b847420 ("kbuild: LLVMLinux: Adapt warnings for compilation with clang").

The new patch looks good to me, please add

Acked-by: Arnd Bergmann <arnd@...db.de>
Fixes: 26ea6bb1fef0 ("kbuild, LLVMLinux: Supress warnings unless W=1-3")

I've started looking into building the kernel with clang, and will likely
follow up with further unrelated changes to the list, as clang-3.9
seems to generate a couple of extra warnings we want to suppress.

	Arnd