Received-SPF: pass (google.com: domain of linux-kernel+bounces-109089-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.80.249 as permitted sender) client-ip=147.75.80.249;
Message-ID: <db4c8e74-5c79-49be-9781-a5d8669eccc1@leemhuis.info>
Date: Wed, 20 Mar 2024 16:25:48 +0100
Precedence: bulk
MIME-Version: 1.0
User-Agent: Mozilla Thunderbird
Subject: Re: [PATCH 1/1] drm/qxl: fixes qxl_fence_wait
Content-Language: en-US, de-DE
To: Alex Constantino <dreaming.about.electric.sheep@gmail.com>
Cc: 1054514@bugs.debian.org, airlied@redhat.com, carnil@debian.org,
 daniel@ffwll.ch, dri-devel@lists.freedesktop.org, kraxel@redhat.com,
 linux-kernel@vger.kernel.org, regressions@lists.linux.dev,
 spice-devel@lists.freedesktop.org, timo.lindfors@iki.fi,
 tzimmermann@suse.de, virtualization@lists.linux-foundation.org,
 Maarten Lankhorst <maarten.lankhorst@linux.intel.com>,
 Maxime Ripard <mripard@kernel.org>
References: <fb0fda6a-3750-4e1b-893f-97a3e402b9af@leemhuis.info>
 <20240308010851.17104-1-dreaming.about.electric.sheep@gmail.com>
 <20240308010851.17104-2-dreaming.about.electric.sheep@gmail.com>
From: "Linux regression tracking (Thorsten Leemhuis)"
 <regressions@leemhuis.info>
Reply-To: Linux regressions mailing list <regressions@lists.linux.dev>
In-Reply-To: <20240308010851.17104-2-dreaming.about.electric.sheep@gmail.com>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit

On 08.03.24 02:08, Alex Constantino wrote:
> Fix OOM scenario by doing multiple notifications to the OOM handler through
> a busy wait logic.
> Changes from commit 5a838e5d5825 ("drm/qxl: simplify qxl_fence_wait") would
> result in a '[TTM] Buffer eviction failed' exception whenever it reached a
> timeout.
> 
> Fixes: 5a838e5d5825 ("drm/qxl: simplify qxl_fence_wait")
> Link: https://lore.kernel.org/regressions/fb0fda6a-3750-4e1b-893f-97a3e402b9af@leemhuis.info
> Reported-by: Timo Lindfors <timo.lindfors@iki.fi>
> Closes: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1054514
> Signed-off-by: Alex Constantino <dreaming.about.electric.sheep@gmail.com>
> ---
>  drivers/gpu/drm/qxl/qxl_release.c | 20 ++++++++++++++------
>  1 file changed, 14 insertions(+), 6 deletions(-)

Hey Dave and Gerd as well as Thomas, Maarten and Maxime (the latter two
I just added to the CC), it seems to me this regression fix did not
maybe any progress since it was posted. Did I miss something, is it just
"we are busy with the merge window", or is there some other a reason?
Just wondering, I just saw someone on a Fedora IRC channel complaining
about the regression, that's why I'm asking. Would be really good to
finally get this resolved...

Ciao, Thorsten (wearing his 'the Linux kernel's regression tracker' hat)
--
Everything you wanna know about Linux kernel regression tracking:
https://linux-regtracking.leemhuis.info/about/#tldr
If I did something stupid, please tell me, as explained on that page.

#regzbot poke

> diff --git a/drivers/gpu/drm/qxl/qxl_release.c b/drivers/gpu/drm/qxl/qxl_release.c
> index 368d26da0d6a..51c22e7f9647 100644
> --- a/drivers/gpu/drm/qxl/qxl_release.c
> +++ b/drivers/gpu/drm/qxl/qxl_release.c
> @@ -20,8 +20,6 @@
>   * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
>   */
>  
> -#include <linux/delay.h>
> -
>  #include <trace/events/dma_fence.h>
>  
>  #include "qxl_drv.h"
> @@ -59,14 +57,24 @@ static long qxl_fence_wait(struct dma_fence *fence, bool intr,
>  {
>  	struct qxl_device *qdev;
>  	unsigned long cur, end = jiffies + timeout;
> +	signed long iterations = 1;
> +	signed long timeout_fraction = timeout;
>  
>  	qdev = container_of(fence->lock, struct qxl_device, release_lock);
>  
> -	if (!wait_event_timeout(qdev->release_event,
> +	// using HZ as a factor since it is used in ttm_bo_wait_ctx too
> +	if (timeout_fraction > HZ) {
> +		iterations = timeout_fraction / HZ;
> +		timeout_fraction = HZ;
> +	}
> +	for (int i = 0; i < iterations; i++) {
> +		if (wait_event_timeout(
> +				qdev->release_event,
>  				(dma_fence_is_signaled(fence) ||
> -				 (qxl_io_notify_oom(qdev), 0)),
> -				timeout))
> -		return 0;
> +					(qxl_io_notify_oom(qdev), 0)),
> +				timeout_fraction))
> +			break;
> +	}
>  
>  	cur = jiffies;
>  	if (time_after(cur, end))