From: "Christian König" <ckoenig.leichtzumerken@gmail.com>
To: Andrey Grodzovsky <Andrey.Grodzovsky@amd.com>,
christian.koenig@amd.com, Daniel Vetter <daniel@ffwll.ch>
Cc: "Alex Deucher" <alexdeucher@gmail.com>,
"Michel Dänzer" <michel@daenzer.net>,
"amd-gfx list" <amd-gfx@lists.freedesktop.org>,
"Pekka Paalanen" <ppaalanen@gmail.com>,
dri-devel <dri-devel@lists.freedesktop.org>
Subject: Re: [PATCH v2 8/8] drm/amdgpu: Prevent any job recoveries after device is unplugged.
Date: Thu, 19 Nov 2020 08:55:33 +0100 [thread overview]
Message-ID: <f65c5d45-8cb2-fde1-1785-756088aa95d5@gmail.com> (raw)
In-Reply-To: <e462f296-75d0-316a-e30f-c985e7aac88d@amd.com>
Am 18.11.20 um 17:20 schrieb Andrey Grodzovsky:
>
> On 11/18/20 7:01 AM, Christian König wrote:
>> Am 18.11.20 um 08:39 schrieb Daniel Vetter:
>>> On Tue, Nov 17, 2020 at 9:07 PM Andrey Grodzovsky
>>> <Andrey.Grodzovsky@amd.com> wrote:
>>>>
>>>> On 11/17/20 2:49 PM, Daniel Vetter wrote:
>>>>> On Tue, Nov 17, 2020 at 02:18:49PM -0500, Andrey Grodzovsky wrote:
>>>>>> On 11/17/20 1:52 PM, Daniel Vetter wrote:
>>>>>>> On Tue, Nov 17, 2020 at 01:38:14PM -0500, Andrey Grodzovsky wrote:
>>>>>>>> On 6/22/20 5:53 AM, Daniel Vetter wrote:
>>>>>>>>> On Sun, Jun 21, 2020 at 02:03:08AM -0400, Andrey Grodzovsky
>>>>>>>>> wrote:
>>>>>>>>>> No point to try recovery if device is gone, just messes up
>>>>>>>>>> things.
>>>>>>>>>>
>>>>>>>>>> Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
>>>>>>>>>> ---
>>>>>>>>>> drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 16
>>>>>>>>>> ++++++++++++++++
>>>>>>>>>> drivers/gpu/drm/amd/amdgpu/amdgpu_job.c | 8 ++++++++
>>>>>>>>>> 2 files changed, 24 insertions(+)
>>>>>>>>>>
>>>>>>>>>> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
>>>>>>>>>> b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
>>>>>>>>>> index 6932d75..5d6d3d9 100644
>>>>>>>>>> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
>>>>>>>>>> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
>>>>>>>>>> @@ -1129,12 +1129,28 @@ static int amdgpu_pci_probe(struct
>>>>>>>>>> pci_dev *pdev,
>>>>>>>>>> return ret;
>>>>>>>>>> }
>>>>>>>>>> +static void amdgpu_cancel_all_tdr(struct amdgpu_device *adev)
>>>>>>>>>> +{
>>>>>>>>>> + int i;
>>>>>>>>>> +
>>>>>>>>>> + for (i = 0; i < AMDGPU_MAX_RINGS; ++i) {
>>>>>>>>>> + struct amdgpu_ring *ring = adev->rings[i];
>>>>>>>>>> +
>>>>>>>>>> + if (!ring || !ring->sched.thread)
>>>>>>>>>> + continue;
>>>>>>>>>> +
>>>>>>>>>> + cancel_delayed_work_sync(&ring->sched.work_tdr);
>>>>>>>>>> + }
>>>>>>>>>> +}
>>>>>>>>> I think this is a function that's supposed to be in
>>>>>>>>> drm/scheduler, not
>>>>>>>>> here. Might also just be your cleanup code being ordered
>>>>>>>>> wrongly, or your
>>>>>>>>> split in one of the earlier patches not done quite right.
>>>>>>>>> -Daniel
>>>>>>>> This function iterates across all the schedulers per amdgpu
>>>>>>>> device and accesses
>>>>>>>> amdgpu specific structures , drm/scheduler deals with single
>>>>>>>> scheduler at most
>>>>>>>> so looks to me like this is the right place for this function
>>>>>>> I guess we could keep track of all schedulers somewhere in a
>>>>>>> list in
>>>>>>> struct drm_device and wrap this up. That was kinda the idea.
>>>>>>>
>>>>>>> Minimally I think a tiny wrapper with docs for the
>>>>>>> cancel_delayed_work_sync(&sched->work_tdr); which explains what
>>>>>>> you must
>>>>>>> observe to make sure there's no race.
>>>>>> Will do
>>>>>>
>>>>>>
>>>>>>> I'm not exactly sure there's no
>>>>>>> guarantee here we won't get a new tdr work launched right
>>>>>>> afterwards at
>>>>>>> least, so this looks a bit like a hack.
>>>>>> Note that for any TDR work happening post amdgpu_cancel_all_tdr
>>>>>> amdgpu_job_timedout->drm_dev_is_unplugged
>>>>>> will return true and so it will return early. To make it water
>>>>>> proof tight
>>>>>> against race
>>>>>> i can switch from drm_dev_is_unplugged to drm_dev_enter/exit
>>>>> Hm that's confusing. You do a work_cancel_sync, so that at least
>>>>> looks
>>>>> like "tdr work must not run after this point"
>>>>>
>>>>> If you only rely on drm_dev_enter/exit check with the tdr work, then
>>>>> there's no need to cancel anything.
>>>>
>>>> Agree, synchronize_srcu from drm_dev_unplug should play the role
>>>> of 'flushing' any earlier (in progress) tdr work which is
>>>> using drm_dev_enter/exit pair. Any later arising tdr will terminate
>>>> early when
>>>> drm_dev_enter
>>>> returns false.
>>> Nope, anything you put into the work itself cannot close this race.
>>> It's the schedule_work that matters here. Or I'm missing something ...
>>> I thought that the tdr work you're cancelling here is launched by
>>> drm/scheduler code, not by the amd callback?
>
>
> My bad, you are right, I am supposed to put drm_dev_enter.exit pair
> into drm_sched_job_timedout
>
>
>>
>> Yes that is correct. Canceling the work item is not the right
>> approach at all, nor is adding dev_enter/exit pair in the recovery
>> handler.
>
>
> Without adding the dev_enter/exit guarding pair in the recovery
> handler you are ending up with GPU reset starting while
> the device is already unplugged, this leads to multiple errors and
> general mess.
>
>
>>
>> What we need to do here is to stop the scheduler thread and then wait
>> for any timeout handling to have finished.
>>
>> Otherwise it can scheduler a new timeout just after we have canceled
>> this one.
>>
>> Regards,
>> Christian.
>
>
> Schedulers are stopped from amdgpu_driver_unload_kms which indeed
> happens after drm_dev_unplug
> so yes, there is still a chance for new work being scheduler and
> timeout armed after but, once i fix the code
> to place drm_dev_enter/exit pair into drm_sched_job_timeout I don't
> see why that not a good solution ?
Yeah that should work as well, but then you also don't need to cancel
the work item from the driver.
> Any tdr work started after drm_dev_unplug finished will simply abort
> on entry to drm_sched_job_timedout
> because drm_dev_enter will be false and the function will return
> without rearming the timeout timer and
> so will have no impact.
>
> The only issue i see here now is of possible use after free if some
> late tdr work will try to execute after
> drm device already gone, for this we probably should add
> cancel_delayed_work_sync(sched.work_tdr)
> to drm_sched_fini after sched->thread is stopped there.
Good point, that is indeed missing as far as I can see.
Christian.
>
> Andrey
_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
next prev parent reply other threads:[~2020-11-19 8:04 UTC|newest]
Thread overview: 97+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-06-21 6:03 [PATCH v2 0/8] RFC Support hot device unplug in amdgpu Andrey Grodzovsky
2020-06-21 6:03 ` [PATCH v2 1/8] drm: Add dummy page per device or GEM object Andrey Grodzovsky
2020-06-22 9:35 ` Daniel Vetter
2020-06-22 14:21 ` Pekka Paalanen
2020-06-22 14:24 ` Daniel Vetter
2020-06-22 14:28 ` Pekka Paalanen
2020-11-09 20:34 ` Andrey Grodzovsky
2020-11-15 6:39 ` Andrey Grodzovsky
2020-06-22 13:18 ` Christian König
2020-06-22 14:23 ` Daniel Vetter
2020-06-22 14:32 ` Andrey Grodzovsky
2020-06-22 17:45 ` Christian König
2020-06-22 17:50 ` Daniel Vetter
2020-11-09 20:53 ` Andrey Grodzovsky
2020-11-13 20:52 ` Andrey Grodzovsky
2020-11-14 8:41 ` Christian König
2020-11-14 9:51 ` Daniel Vetter
2020-11-14 9:57 ` Daniel Vetter
2020-11-16 9:42 ` Michel Dänzer
2020-11-15 6:34 ` Andrey Grodzovsky
2020-11-16 9:48 ` Christian König
2020-11-16 19:00 ` Andrey Grodzovsky
2020-11-16 20:36 ` Christian König
2020-11-16 20:42 ` Andrey Grodzovsky
2020-11-19 10:01 ` Christian König
2020-06-21 6:03 ` [PATCH v2 2/8] drm/ttm: Remap all page faults to per process dummy page Andrey Grodzovsky
2020-06-22 9:41 ` Daniel Vetter
2020-06-24 3:31 ` Andrey Grodzovsky
2020-06-24 7:19 ` Daniel Vetter
2020-11-10 17:41 ` Andrey Grodzovsky
2020-06-22 19:30 ` Christian König
2020-06-21 6:03 ` [PATCH v2 3/8] drm/ttm: Add unampping of the entire device address space Andrey Grodzovsky
2020-06-22 9:45 ` Daniel Vetter
2020-06-23 5:00 ` Andrey Grodzovsky
2020-06-23 10:25 ` Daniel Vetter
2020-06-23 12:55 ` Christian König
2020-06-22 19:37 ` Christian König
2020-06-22 19:47 ` Alex Deucher
2020-06-21 6:03 ` [PATCH v2 4/8] drm/amdgpu: Split amdgpu_device_fini into early and late Andrey Grodzovsky
2020-06-22 9:48 ` Daniel Vetter
2020-11-12 4:19 ` Andrey Grodzovsky
2020-11-12 9:29 ` Daniel Vetter
2020-06-21 6:03 ` [PATCH v2 5/8] drm/amdgpu: Refactor sysfs removal Andrey Grodzovsky
2020-06-22 9:51 ` Daniel Vetter
2020-06-22 11:21 ` Greg KH
2020-06-22 16:07 ` Andrey Grodzovsky
2020-06-22 16:45 ` Greg KH
2020-06-23 4:51 ` Andrey Grodzovsky
2020-06-23 6:05 ` Greg KH
2020-06-24 3:04 ` Andrey Grodzovsky
2020-06-24 6:11 ` Greg KH
2020-06-25 1:52 ` Andrey Grodzovsky
2020-11-10 17:54 ` Andrey Grodzovsky
2020-11-10 17:59 ` Greg KH
2020-11-11 15:13 ` Andrey Grodzovsky
2020-11-11 15:34 ` Greg KH
2020-11-11 15:45 ` Andrey Grodzovsky
2020-11-11 16:06 ` Greg KH
2020-11-11 16:34 ` Andrey Grodzovsky
2020-12-02 15:48 ` Andrey Grodzovsky
2020-12-02 17:34 ` Greg KH
2020-12-02 18:02 ` Andrey Grodzovsky
2020-12-02 18:20 ` Greg KH
2020-12-02 18:40 ` Andrey Grodzovsky
2020-06-22 13:19 ` Christian König
2020-06-21 6:03 ` [PATCH v2 6/8] drm/amdgpu: Unmap entire device address space on device remove Andrey Grodzovsky
2020-06-22 9:56 ` Daniel Vetter
2020-06-22 19:38 ` Christian König
2020-06-22 19:48 ` Alex Deucher
2020-06-23 10:22 ` Daniel Vetter
2020-06-23 13:16 ` Christian König
2020-06-24 3:12 ` Andrey Grodzovsky
2020-06-21 6:03 ` [PATCH v2 7/8] drm/amdgpu: Fix sdma code crash post device unplug Andrey Grodzovsky
2020-06-22 9:55 ` Daniel Vetter
2020-06-22 19:40 ` Christian König
2020-06-23 5:11 ` Andrey Grodzovsky
2020-06-23 7:14 ` Christian König
2020-06-21 6:03 ` [PATCH v2 8/8] drm/amdgpu: Prevent any job recoveries after device is unplugged Andrey Grodzovsky
2020-06-22 9:53 ` Daniel Vetter
2020-11-17 18:38 ` Andrey Grodzovsky
2020-11-17 18:52 ` Daniel Vetter
2020-11-17 19:18 ` Andrey Grodzovsky
2020-11-17 19:49 ` Daniel Vetter
2020-11-17 20:07 ` Andrey Grodzovsky
2020-11-18 7:39 ` Daniel Vetter
2020-11-18 12:01 ` Christian König
2020-11-18 15:43 ` Luben Tuikov
2020-11-18 16:20 ` Andrey Grodzovsky
2020-11-19 7:55 ` Christian König [this message]
2020-11-19 15:02 ` Andrey Grodzovsky
2020-11-19 15:29 ` Daniel Vetter
2020-11-19 21:24 ` Andrey Grodzovsky
2020-11-18 0:46 ` Luben Tuikov
2020-06-22 9:46 ` [PATCH v2 0/8] RFC Support hot device unplug in amdgpu Daniel Vetter
2020-06-23 5:14 ` Andrey Grodzovsky
2020-06-23 9:04 ` Michel Dänzer
2020-06-24 3:21 ` Andrey Grodzovsky
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=f65c5d45-8cb2-fde1-1785-756088aa95d5@gmail.com \
--to=ckoenig.leichtzumerken@gmail.com \
--cc=Andrey.Grodzovsky@amd.com \
--cc=alexdeucher@gmail.com \
--cc=amd-gfx@lists.freedesktop.org \
--cc=christian.koenig@amd.com \
--cc=daniel@ffwll.ch \
--cc=dri-devel@lists.freedesktop.org \
--cc=michel@daenzer.net \
--cc=ppaalanen@gmail.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).