From: "Rafael J. Wysocki" <rafael@kernel.org>
To: "Marcin Ślusarz" <marcin.slusarz@gmail.com>
Cc: "moderated list:SOUND - SOC LAYER / DYNAMIC AUDIO POWER
MANAGEM..." <alsa-devel@alsa-project.org>,
"Rafael J. Wysocki" <rafael@kernel.org>,
"Rafael J. Wysocki" <rjw@rjwysocki.net>,
Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>,
ACPI Devel Maling List <linux-acpi@vger.kernel.org>,
Vinod Koul <vkoul@kernel.org>,
Bard Liao <yung-chuan.liao@linux.intel.com>,
Len Brown <lenb@kernel.org>
Subject: Re: Crash in acpi_ns_validate_handle triggered by soundwire on Linux 5.10
Date: Thu, 28 Jan 2021 14:31:59 +0100 [thread overview]
Message-ID: <CAJZ5v0gRjO-0Kt8BHXwv2EbBuJhxcx=afQWpKkiUJBdVc8=dEg@mail.gmail.com> (raw)
In-Reply-To: <CA+GA0_sXVNHr1048otvwGwORt17ET0EbvTq-kGMOQoh1YOWYRw@mail.gmail.com>
On Thu, Jan 28, 2021 at 2:25 PM Marcin Ślusarz <marcin.slusarz@gmail.com> wrote:
>
> śr., 27 sty 2021 o 23:02 Pierre-Louis Bossart
> <pierre-louis.bossart@linux.intel.com> napisał(a):
> > On 1/27/21 1:18 PM, Marcin Ślusarz wrote:
> > > śr., 27 sty 2021 o 18:28 Pierre-Louis Bossart
> > > <pierre-louis.bossart@linux.intel.com> napisał(a):
> > >>> Weird, I can't reproduce this problem with my self-compiled kernel :/
> > >>> I don't even see soundwire modules loaded in. Manually loading them of course
> > >>> doesn't do much.
> > >>>
> > >>> Previously I could boot into the "faulty" kernel by using "recovery mode", but
> > >>> I can't do that anymore - it crashes too.
> > >>>
> > >>> Maybe there's some kind of race and this bug depends on some specific
> > >>> ordering of events?
> > >>
> > >> missing Kconfig?
> > >> You need CONFIG_SOUNDWIRE and CONFIG_SND_SOC_SOF_INTEL_SOUNDWIRE
> > >> selected to enter this sdw_intel_acpi_scan() routine.
> > >
> > > It was a PEBKAC, but a slightly different one. I won't bore you with
> > > (embarrassing) details ;).
> > >
> > > I reproduced the problem, tested both your and Rafael's patches
> > > and the kernel still crashes, with the same stack trace.
> > > (Yes, I'm sure I booted the right kernel :)
> > >
> > > Why "recovery mode" stopped working (or worked previously) is still a mystery.
> > >
> >
> > Thanks Marcin for the information. If you have a consistent failure
> > that's better to some extent.
> >
> > Maybe a bit of explanation of what this routine tries to do:
> > when SoundWire is enabled in a system, we need to have the following
> > pattern in the DSDT:
> >
> > Scope (_SB.PCI0)
> > {
> > Device (HDAS)
> > {
> > Name (_ADR, 0x001F0003) // _ADR: Address
> > }
> >
> >
> > Scope (HDAS)
> > {
> > Device (SNDW)
> > {
> > Name (_ADR, 0x40000000) // _ADR: Address
> >
> > The only thing the code does is to walk through the children and check
> > if the valid _ADR 0x40000000 is found.
> >
> > You don't have SoundWire in your device so there should not be any
> > children found. I don't see anything in the DSDT that looks like
> > _SB.PCI0.HDAS.<something>, so in theory we should not even enter the
> > callback.
> >
> > The error happens in acpi_bus_get_device(), after we read the adr but
> > before we check it, so wondering if we shouldn't revert the checks. Can
> > you try the diff below? I am not sure why there is a crash and we should
> > root-cause this issue, just trying to triangulate what is happening.
> >
> > diff --git a/drivers/soundwire/intel_init.c b/drivers/soundwire/intel_init.c
> > index cabdadb09a1b..6bc87a682fb3 100644
> > --- a/drivers/soundwire/intel_init.c
> > +++ b/drivers/soundwire/intel_init.c
> > @@ -369,13 +369,6 @@ static acpi_status sdw_intel_acpi_cb(acpi_handle
> > handle, u32 level,
> > if (ACPI_FAILURE(status))
> > return AE_OK; /* keep going */
> >
> > - if (acpi_bus_get_device(handle, &adev)) {
> > - pr_err("%s: Couldn't find ACPI handle\n", __func__);
> > - return AE_NOT_FOUND;
> > - }
> > -
> > - info->handle = handle;
> > -
> > /*
> > * On some Intel platforms, multiple children of the HDAS
> > * device can be found, but only one of them is the SoundWire
> > @@ -386,6 +379,13 @@ static acpi_status sdw_intel_acpi_cb(acpi_handle
> > handle, u32 level,
> > if (FIELD_GET(GENMASK(31, 28), adr) != SDW_LINK_TYPE)
> > return AE_OK; /* keep going */
> >
> > + if (acpi_bus_get_device(handle, &adev)) {
> > + pr_err("%s: Couldn't find ACPI handle\n", __func__);
> > + return AE_NOT_FOUND;
> > + }
> > +
> > + info->handle = handle;
> > +
> > /* device found, stop namespace walk */
> > return AE_CTRL_TERMINATE;
> > }
>
> still the same crash
The modification doesn't fundamentally change the conditions, but
since the flow gets to the acpi_bus_get_device() evaluation, adr
appears to make sense (which is kind of strange, because it shouldn't
in the "no SoundWire" case).
next prev parent reply other threads:[~2021-01-28 13:33 UTC|newest]
Thread overview: 33+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-01-20 19:56 Crash in acpi_ns_validate_handle triggered by soundwire on Linux 5.10 Marcin Ślusarz
2021-01-20 20:34 ` Rafael J. Wysocki
2021-01-20 22:28 ` Pierre-Louis Bossart
2021-01-21 17:47 ` Marcin Ślusarz
2021-01-27 16:36 ` Marcin Ślusarz
2021-01-27 17:28 ` Pierre-Louis Bossart
2021-01-27 19:18 ` Marcin Ślusarz
2021-01-27 21:52 ` Pierre-Louis Bossart
2021-01-27 22:02 ` Pierre-Louis Bossart
2021-01-28 13:25 ` Marcin Ślusarz
2021-01-28 13:31 ` Rafael J. Wysocki [this message]
2021-01-28 12:13 ` Rafael J. Wysocki
2021-01-28 12:39 ` Rafael J. Wysocki
2021-01-28 13:45 ` Marcin Ślusarz
2021-01-28 14:32 ` Marcin Ślusarz
2021-01-29 18:59 ` Marcin Ślusarz
2021-01-29 20:03 ` Marcin Ślusarz
2021-02-01 11:42 ` Rafael J. Wysocki
2021-02-01 12:16 ` Marcin Ślusarz
2021-02-04 12:11 ` Marcin Ślusarz
2021-02-04 12:48 ` Marcin Ślusarz
2021-02-05 15:40 ` [PATCH] soundwire: intel: fix possible crash when no device is detected (was Re: Crash in acpi_ns_validate_handle triggered by soundwire on Linux 5.10) Marcin Ślusarz
2021-02-05 16:16 ` Pierre-Louis Bossart
2021-02-08 12:01 ` [PATCH 1/2] soundwire: intel: fix possible crash when no device is detected Marcin Ślusarz
2021-02-08 12:01 ` [PATCH 2/2] ACPICA: update documentation of acpi_walk_namespace Marcin Ślusarz
2021-02-08 12:43 ` Rafael J. Wysocki
2021-02-08 12:37 ` [PATCH 1/2] soundwire: intel: fix possible crash when no device is detected Rafael J. Wysocki
2021-02-10 23:15 ` Pierre-Louis Bossart
2021-02-11 5:20 ` Vinod Koul
2021-01-28 13:29 ` Crash in acpi_ns_validate_handle triggered by soundwire on Linux 5.10 Marcin Ślusarz
2021-01-28 13:36 ` Rafael J. Wysocki
2021-01-25 18:38 ` Salvatore Bonaccorso
2021-01-25 19:26 ` Pierre-Louis Bossart
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to='CAJZ5v0gRjO-0Kt8BHXwv2EbBuJhxcx=afQWpKkiUJBdVc8=dEg@mail.gmail.com' \
--to=rafael@kernel.org \
--cc=alsa-devel@alsa-project.org \
--cc=lenb@kernel.org \
--cc=linux-acpi@vger.kernel.org \
--cc=marcin.slusarz@gmail.com \
--cc=pierre-louis.bossart@linux.intel.com \
--cc=rjw@rjwysocki.net \
--cc=vkoul@kernel.org \
--cc=yung-chuan.liao@linux.intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).