Intel-GFX Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: "Christian König" <christian.koenig@amd.com>
To: Daniel Vetter <daniel.vetter@ffwll.ch>,
	Intel Graphics Development <intel-gfx@lists.freedesktop.org>,
	DRI Development <dri-devel@lists.freedesktop.org>
Cc: Daniel Vetter <daniel.vetter@intel.com>
Subject: Re: [PATCH 1/3] drm/modeset: Prime modeset lock vs dma_resv
Date: Wed, 20 Nov 2019 09:34:03 +0100	[thread overview]
Message-ID: <9ce03b86-3bec-3a33-b5d0-4acbe9e861ec@amd.com> (raw)
In-Reply-To: <20191119210844.16947-2-daniel.vetter@ffwll.ch>

Am 19.11.19 um 22:08 schrieb Daniel Vetter:
> It's kinda really hard to get this wrong on a driver with both display
> and dma_resv locking. But who ever knows, so better to make sure that
> really all drivers nest these the same way.
>
> For actual lock semantics the acquire context nesting doesn't matter.
> But to teach lockdep what's going on with ww_mutex the acquire ctx is
> a fake lockdep lock, hence from a lockdep pov it does matter. That's
> why I figured better to include it.
>
> Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
> Cc: Chris Wilson <chris@chris-wilson.co.uk>
> Cc: Christian König <christian.koenig@amd.com>
> Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>

Why not add another __init function like we did for dma_resv? That 
looked rather clean to me.

Either why feel free to add an Acked-by: Christian König 
<christian.koenig@amd.com> to the patch.

Regards,
Christian.

> ---
>   drivers/gpu/drm/drm_mode_config.c | 28 ++++++++++++++++++++++++++++
>   1 file changed, 28 insertions(+)
>
> diff --git a/drivers/gpu/drm/drm_mode_config.c b/drivers/gpu/drm/drm_mode_config.c
> index 3b570a404933..08e6eff6a179 100644
> --- a/drivers/gpu/drm/drm_mode_config.c
> +++ b/drivers/gpu/drm/drm_mode_config.c
> @@ -27,6 +27,7 @@
>   #include <drm/drm_file.h>
>   #include <drm/drm_mode_config.h>
>   #include <drm/drm_print.h>
> +#include <linux/dma-resv.h>
>   
>   #include "drm_crtc_internal.h"
>   #include "drm_internal.h"
> @@ -415,6 +416,33 @@ void drm_mode_config_init(struct drm_device *dev)
>   	dev->mode_config.num_crtc = 0;
>   	dev->mode_config.num_encoder = 0;
>   	dev->mode_config.num_total_plane = 0;
> +
> +	if (IS_ENABLED(CONFIG_LOCKDEP)) {
> +		struct drm_modeset_acquire_ctx modeset_ctx;
> +		struct ww_acquire_ctx resv_ctx;
> +		struct dma_resv resv;
> +		int ret;
> +
> +		dma_resv_init(&resv);
> +
> +		drm_modeset_acquire_init(&modeset_ctx, 0);
> +		ret = drm_modeset_lock(&dev->mode_config.connection_mutex,
> +				       &modeset_ctx);
> +		if (ret == -EDEADLK)
> +			ret = drm_modeset_backoff(&modeset_ctx);
> +
> +		ww_acquire_init(&resv_ctx, &reservation_ww_class);
> +		ret = dma_resv_lock(&resv, &resv_ctx);
> +		if (ret == -EDEADLK)
> +			dma_resv_lock_slow(&resv, &resv_ctx);
> +
> +		dma_resv_unlock(&resv);
> +		ww_acquire_fini(&resv_ctx);
> +
> +		drm_modeset_drop_locks(&modeset_ctx);
> +		drm_modeset_acquire_fini(&modeset_ctx);
> +		dma_resv_fini(&resv);
> +	}
>   }
>   EXPORT_SYMBOL(drm_mode_config_init);
>   

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

WARNING: multiple messages have this Message-ID (diff)
From: "Christian König" <christian.koenig@amd.com>
To: Daniel Vetter <daniel.vetter@ffwll.ch>,
	Intel Graphics Development <intel-gfx@lists.freedesktop.org>,
	DRI Development <dri-devel@lists.freedesktop.org>
Cc: Daniel Vetter <daniel.vetter@intel.com>
Subject: Re: [Intel-gfx] [PATCH 1/3] drm/modeset: Prime modeset lock vs dma_resv
Date: Wed, 20 Nov 2019 09:34:03 +0100	[thread overview]
Message-ID: <9ce03b86-3bec-3a33-b5d0-4acbe9e861ec@amd.com> (raw)
Message-ID: <20191120083403.HS5Ij7d2hmhONdUiR0JZG1XuVNnrr4Ni0921PLQ428M@z> (raw)
In-Reply-To: <20191119210844.16947-2-daniel.vetter@ffwll.ch>

Am 19.11.19 um 22:08 schrieb Daniel Vetter:
> It's kinda really hard to get this wrong on a driver with both display
> and dma_resv locking. But who ever knows, so better to make sure that
> really all drivers nest these the same way.
>
> For actual lock semantics the acquire context nesting doesn't matter.
> But to teach lockdep what's going on with ww_mutex the acquire ctx is
> a fake lockdep lock, hence from a lockdep pov it does matter. That's
> why I figured better to include it.
>
> Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
> Cc: Chris Wilson <chris@chris-wilson.co.uk>
> Cc: Christian König <christian.koenig@amd.com>
> Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>

Why not add another __init function like we did for dma_resv? That 
looked rather clean to me.

Either why feel free to add an Acked-by: Christian König 
<christian.koenig@amd.com> to the patch.

Regards,
Christian.

> ---
>   drivers/gpu/drm/drm_mode_config.c | 28 ++++++++++++++++++++++++++++
>   1 file changed, 28 insertions(+)
>
> diff --git a/drivers/gpu/drm/drm_mode_config.c b/drivers/gpu/drm/drm_mode_config.c
> index 3b570a404933..08e6eff6a179 100644
> --- a/drivers/gpu/drm/drm_mode_config.c
> +++ b/drivers/gpu/drm/drm_mode_config.c
> @@ -27,6 +27,7 @@
>   #include <drm/drm_file.h>
>   #include <drm/drm_mode_config.h>
>   #include <drm/drm_print.h>
> +#include <linux/dma-resv.h>
>   
>   #include "drm_crtc_internal.h"
>   #include "drm_internal.h"
> @@ -415,6 +416,33 @@ void drm_mode_config_init(struct drm_device *dev)
>   	dev->mode_config.num_crtc = 0;
>   	dev->mode_config.num_encoder = 0;
>   	dev->mode_config.num_total_plane = 0;
> +
> +	if (IS_ENABLED(CONFIG_LOCKDEP)) {
> +		struct drm_modeset_acquire_ctx modeset_ctx;
> +		struct ww_acquire_ctx resv_ctx;
> +		struct dma_resv resv;
> +		int ret;
> +
> +		dma_resv_init(&resv);
> +
> +		drm_modeset_acquire_init(&modeset_ctx, 0);
> +		ret = drm_modeset_lock(&dev->mode_config.connection_mutex,
> +				       &modeset_ctx);
> +		if (ret == -EDEADLK)
> +			ret = drm_modeset_backoff(&modeset_ctx);
> +
> +		ww_acquire_init(&resv_ctx, &reservation_ww_class);
> +		ret = dma_resv_lock(&resv, &resv_ctx);
> +		if (ret == -EDEADLK)
> +			dma_resv_lock_slow(&resv, &resv_ctx);
> +
> +		dma_resv_unlock(&resv);
> +		ww_acquire_fini(&resv_ctx);
> +
> +		drm_modeset_drop_locks(&modeset_ctx);
> +		drm_modeset_acquire_fini(&modeset_ctx);
> +		dma_resv_fini(&resv);
> +	}
>   }
>   EXPORT_SYMBOL(drm_mode_config_init);
>   

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

  parent reply	other threads:[~2019-11-20  8:34 UTC|newest]

Thread overview: 34+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-11-19 21:08 [PATCH 0/3] more dma-buf lockdep priming Daniel Vetter
2019-11-19 21:08 ` [Intel-gfx] " Daniel Vetter
2019-11-19 21:08 ` [PATCH 1/3] drm/modeset: Prime modeset lock vs dma_resv Daniel Vetter
2019-11-19 21:08   ` [Intel-gfx] " Daniel Vetter
2019-11-20  8:34   ` Christian König [this message]
2019-11-20  8:34     ` Christian König
2019-11-20 10:55     ` Daniel Vetter
2019-11-20 10:55       ` [Intel-gfx] " Daniel Vetter
2019-11-19 21:08 ` [PATCH 2/3] dma-resv: Also prime acquire ctx for lockdep Daniel Vetter
2019-11-19 21:08   ` [Intel-gfx] " Daniel Vetter
2019-11-20 11:30   ` Christian König
2019-11-20 11:30     ` [Intel-gfx] " Christian König
2019-11-20 13:56     ` Maarten Lankhorst
2019-11-20 13:56       ` [Intel-gfx] " Maarten Lankhorst
2019-11-19 21:08 ` [PATCH 3/3] drm/msm: Don't init ww_mutec acquire ctx before needed Daniel Vetter
2019-11-19 21:08   ` [Intel-gfx] " Daniel Vetter
     [not found]   ` <20191119210844.16947-4-daniel.vetter-/w4YWyX8dFk@public.gmane.org>
2019-11-20  2:07     ` Rob Clark
2019-11-20  2:07       ` [Intel-gfx] " Rob Clark
     [not found]       ` <CAF6AEGuPtwFjM12MRBf_X7GUqJBHt+Xb8vKrT+dddic093Ki6g-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
2019-11-20 10:22         ` Daniel Vetter
2019-11-20 10:22           ` [Intel-gfx] " Daniel Vetter
2019-11-20 10:56   ` [PATCH] " Daniel Vetter
2019-11-20 10:56     ` [Intel-gfx] " Daniel Vetter
     [not found]     ` <20191120105607.3023-1-daniel.vetter-/w4YWyX8dFk@public.gmane.org>
2019-11-21  0:24       ` Rob Clark
2019-11-21  0:24         ` [Intel-gfx] " Rob Clark
2019-11-21 10:03         ` Daniel Vetter
2019-11-21 10:03           ` [Intel-gfx] " Daniel Vetter
2019-11-19 22:31 ` ✗ Fi.CI.CHECKPATCH: warning for more dma-buf lockdep priming Patchwork
2019-11-19 22:31   ` [Intel-gfx] " Patchwork
2019-11-19 22:53 ` ✓ Fi.CI.BAT: success " Patchwork
2019-11-19 22:53   ` [Intel-gfx] " Patchwork
2019-11-20 13:55 ` ✓ Fi.CI.IGT: " Patchwork
2019-11-20 13:55   ` [Intel-gfx] " Patchwork
2019-11-20 15:31 ` ✗ Fi.CI.BUILD: failure for more dma-buf lockdep priming (rev2) Patchwork
2019-11-20 15:31   ` [Intel-gfx] " Patchwork

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=9ce03b86-3bec-3a33-b5d0-4acbe9e861ec@amd.com \
    --to=christian.koenig@amd.com \
    --cc=daniel.vetter@ffwll.ch \
    --cc=daniel.vetter@intel.com \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=intel-gfx@lists.freedesktop.org \
    /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