linux-kernel - Re: [PATCH v2 4/4] dynamic_debug: add jump label support

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>] [day] [month] [year] [list]

Message-ID: <4481273.afaatbzpKs@wuerfel>
Date:	Wed, 13 Jul 2016 18:03:24 +0200
From:	Arnd Bergmann <arnd@...db.de>
To:	Jason Baron <jbaron@...mai.com>,
	kernel-build-reports@...ts.linaro.org
Cc:	linux-kernel@...r.kernel.org, linux-media@...r.kernel.org
Subject: Re: [PATCH v2 4/4] dynamic_debug: add jump label support

On Wednesday, July 13, 2016 11:41:50 AM CEST Jason Baron wrote:
> 
> Hi Arnd,
> 
> Ok, this is back in linux-next now (with hopefully a fix for arm). I
> was never able to quite reproduce the arm failure you saw. So if
> you get the chance to test this it would be great.
> 

I've had a day's worth of randconfig tests without running into the problem
so far.

However, I did get one new compiler warning that I have just bisected
down to 21413cd0e4ed ("dynamic_debug: add jump label support"):

/git/arm-soc/drivers/media/dvb-frontends/cxd2841er.c: In function 'cxd2841er_tune_tc':
/git/arm-soc/drivers/media/dvb-frontends/cxd2841er.c:3253:40: error: 'carrier_offset' may be used uninitialized in this function [-Werror=maybe-uninitialized]
    if (ret)
                                        ^
     return ret;
     ~~~~~~~~~                           
/git/arm-soc/drivers/media/dvb-frontends/cxd2841er.c:3209:11: note: 'carrier_offset' was declared here
  int ret, carrier_offset;
           ^~~~~~~~~~~~~~

It's clearly a false positive warning, the code is correct, but if this is
the only one that the dynamic_debug jump labels introduce, we may as well
just work around it in the driver.

I think this is caused by the "unlikely" annotation in dynamic_dev_dbg(),
which confuses the compiler trying to figure out whether the variable
is initialized or not.

	Arnd