linux-kernel - Re: kernel panic with v5.18-rc1 on OpenPandora (only)

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]

Date:   Tue, 3 May 2022 09:28:44 +0200
From:   Ard Biesheuvel <ardb@...nel.org>
To:     Arnd Bergmann <arnd@...db.de>
Cc:     "H. Nikolaus Schaller" <hns@...delico.com>,
        Tony Lindgren <tony@...mide.com>,
        Discussions about the Letux Kernel 
        <letux-kernel@...nphoenux.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        Linux-OMAP <linux-omap@...r.kernel.org>,
        Linux ARM <linux-arm-kernel@...ts.infradead.org>
Subject: Re: kernel panic with v5.18-rc1 on OpenPandora (only)

On Sat, 30 Apr 2022 at 20:48, Arnd Bergmann <arnd@...db.de> wrote:
>
> On Sat, Apr 30, 2022 at 7:18 PM H. Nikolaus Schaller <hns@...delico.com> wrote:
> > > Am 30.04.2022 um 17:36 schrieb Arnd Bergmann <arnd@...db.de>:
> > >
> > >
> > > I suppose this could be anywhere then. The backtrace seems to point
> > > to re-enabling interupts in do_work_pending, so something probably
> > > accessed DMA memory asynchronously.
> >
> > Yes. I now (or still) sometimes see the same omap l3 irq issue when plugging in/out the USB/OTG
> > cable. Not with a kernel panic, but in the same driver omap_l3_smx.c.
> > This happens even if the wl1251 driver is removed.
>
> Is this also a regression, or did it happen before the vmap-stack
> change? If this only
> appeared now, then this points to another bug somewhere that you
> should find using
> CONFIG_DMA_API_DEBUG.
>
> I think what is going on here is that your platform is able to detect
> the broken DMA
> because of the l3 interrupt handler telling the kernel about it, when
> on other platforms
> we would see either silent data corruption or a DMA that never reaches
> its target.
>

I wonder if we could narrow this down by adding the possibility to use
IRQ stacks in the linear map, while using vmap'ed task stacks.