linux-kernel - Re: rc9 + orinoco WPA patchset: BUG: scheduling while atomic loading firmware with PCMCIA adapter

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

Message-Id: <200810092013.55476.arvidjaar@newmail.ru>
Date:	Thu, 9 Oct 2008 20:13:54 +0400
From:	Andrey Borzenkov <arvidjaar@...mail.ru>
To:	Dan Williams <dcbw@...hat.com>
Cc:	Matthew Wilcox <matthew@....cx>,
	orinoco-devel@...ts.sourceforge.net,
	linux-wireless@...r.kernel.org, linux-pcmcia@...ts.infradead.org,
	Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: rc9 + orinoco WPA patchset: BUG: scheduling while atomic loading firmware with PCMCIA adapter

On Thursday 09 October 2008, Dan Williams wrote:
> 
> On Thu, 2008-10-09 at 19:59 +0400, Andrey Borzenkov wrote:
> > On Thursday 09 October 2008, Matthew Wilcox wrote:
> > > 
> > > On Thu, Oct 09, 2008 at 05:22:55PM +0400, Andrey Borzenkov wrote:
> > > > After loading orinoco and "inserting" adapter I get either BUG with
> > > > endless loop (attached) or kernel panic on NULL pointer dereference
> > > > (was not able capture). Adapter works fine with wlags49 driver.
> > > 
> > > It looks like you've fallen off the bottom of the kernel stack.  Do you
> > > have 4k stacks enabled in your config?
> > > 
> > 
> > You are right. Using 8K stacks load and runs fine. Hmm ... not nice
> > from it :p
> 
> == driver bug; the driver should not require large stacks and this
> should get fixed.
> 
I'll check for obvious places, still ...

/* Download either STA or AP firmware into the card. */
static int
orinoco_dl_firmware(struct orinoco_private *priv,
                    const struct fw_info *fw,
                    int ap)
{
        /* Plug Data Area (PDA) */
        __le16 pda[512] = { 0 };

allocating 1K on stack is not nice; but what has eaten 3K more?

Is there any instrumentation to analyze stack usage?

Download attachment "signature.asc " of type "application/pgp-signature" (198 bytes)