linux-kernel - Re: [PATCH 1/3] platform/x86: intel_scu_ipc: Check status after timeouts in busy

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: <CAE-0n52wAqsmm4cs6JX2W2G10VxjLzocXVmF9c_GC+52Fi4djQ@mail.gmail.com>
Date:   Tue, 5 Sep 2023 17:24:29 -0500
From:   Stephen Boyd <swboyd@...omium.org>
To:     Andy Shevchenko <andriy.shevchenko@...ux.intel.com>
Cc:     Mika Westerberg <mika.westerberg@...ux.intel.com>,
        Hans de Goede <hdegoede@...hat.com>,
        Mark Gross <markgross@...nel.org>,
        linux-kernel@...r.kernel.org, patches@...ts.linux.dev,
        platform-driver-x86@...r.kernel.org,
        Kuppuswamy Sathyanarayanan 
        <sathyanarayanan.kuppuswamy@...ux.intel.com>,
        Prashant Malani <pmalani@...omium.org>
Subject: Re: [PATCH 1/3] platform/x86: intel_scu_ipc: Check status after
 timeouts in busy_loop()

Quoting Andy Shevchenko (2023-08-31 06:53:14)
> On Wed, Aug 30, 2023 at 06:14:01PM -0700, Stephen Boyd wrote:
> > It's possible for the polling loop in busy_loop() to get scheduled away
> > for a long time.
> >
> >   status = ipc_read_status(scu);
> >   <long time scheduled away>
> >   if (!(status & IPC_STATUS_BUSY))
> >
> > If this happens, then the status bit could change and this function
> > would never test it again after checking the jiffies against the timeout
> > limit. Polling code should check the condition one more time after the
> > timeout in case this happens.
> >
> > The read_poll_timeout() helper implements this logic, and is shorter, so
> > simply use that helper here.
>
> I don't remember by heart, but on some older Intel hardware this might have
> been called during early stages where ktime() is not functional yet.
>
> Is this still a case here?

I have no idea if that happens in early stages. What about
suspend/resume though? I suppose timekeeping could be suspended in that
case, so we can't really check anything with ktime.

I can rework this patch to simply recheck the busy bit so that we don't
have to figure out if the code is called early or from suspend paths.