From: Sam Ravnborg <sam@ravnborg.org>
To: Thomas Zimmermann <tzimmermann@suse.de>
Cc: maxime.ripard@bootlin.com, dri-devel@lists.freedesktop.org,
kraxel@redhat.com, airlied@redhat.com, sean@poorly.run
Subject: Re: [PATCH v2 5/9] drm/ast: Pin framebuffer BO during dirty update
Date: Tue, 11 Jun 2019 22:29:53 +0200 [thread overview]
Message-ID: <20190611202953.GA18315@ravnborg.org> (raw)
In-Reply-To: <20190611130344.18988-6-tzimmermann@suse.de>
Hi Thomas.
On Tue, Jun 11, 2019 at 03:03:40PM +0200, Thomas Zimmermann wrote:
> Another explicit lock operation of a GEM VRAM BO is located in AST's
> framebuffer update code. Instead of locking the BO, we pin it to wherever
> it is.
>
> v2:
> * update with pin flag of 0
>
> Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
> ---
> drivers/gpu/drm/ast/ast_fb.c | 33 ++++++++++++++++-----------------
> 1 file changed, 16 insertions(+), 17 deletions(-)
>
> diff --git a/drivers/gpu/drm/ast/ast_fb.c b/drivers/gpu/drm/ast/ast_fb.c
> index 05f45222b702..b404b51c2c55 100644
> --- a/drivers/gpu/drm/ast/ast_fb.c
> +++ b/drivers/gpu/drm/ast/ast_fb.c
> @@ -48,32 +48,31 @@ static void ast_dirty_update(struct ast_fbdev *afbdev,
> int x, int y, int width, int height)
> {
> int i;
> - struct drm_gem_object *obj;
> struct drm_gem_vram_object *gbo;
> int src_offset, dst_offset;
> int bpp = afbdev->afb.base.format->cpp[0];
> - int ret = -EBUSY;
> + int ret;
> u8 *dst;
> bool unmap = false;
> bool store_for_later = false;
> int x2, y2;
> unsigned long flags;
>
> - obj = afbdev->afb.obj;
> - gbo = drm_gem_vram_of_gem(obj);
> -
> - /* Try to lock the BO. If we fail with -EBUSY then
> - * the BO is being moved and we should store up the
> - * damage until later.
> - */
> - if (drm_can_sleep())
> - ret = drm_gem_vram_lock(gbo, true);
> - if (ret) {
> - if (ret != -EBUSY)
> - return;
> -
> + gbo = drm_gem_vram_of_gem(afbdev->afb.obj);
> +
> + if (drm_can_sleep()) {
While touching this code, anyway we could get rid of drm_can_sleep()?
I only ask because Daniel V. said that we should not use it.
This was some months ago during some ehader file clean-up, so nothing
in particular related to the ast driver.
Sam
> + /* We pin the BO so it won't be moved during the
> + * update. The actual location, video RAM or system
> + * memory, is not important.
> + */
> + ret = drm_gem_vram_pin(gbo, 0);
> + if (ret) {
> + if (ret != -EBUSY)
> + return;
> + store_for_later = true;
> + }
> + } else
> store_for_later = true;
> - }
>
> x2 = x + width - 1;
> y2 = y + height - 1;
> @@ -126,7 +125,7 @@ static void ast_dirty_update(struct ast_fbdev *afbdev,
> drm_gem_vram_kunmap(gbo);
>
> out:
> - drm_gem_vram_unlock(gbo);
> + drm_gem_vram_unpin(gbo);
> }
>
> static void ast_fillrect(struct fb_info *info,
> --
> 2.21.0
_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel
next prev parent reply other threads:[~2019-06-11 20:29 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-06-11 13:03 [PATCH v2 0/9] Remove explicit locking and kmap arguments from GEM VRAM interface Thomas Zimmermann
2019-06-11 13:03 ` [PATCH v2 1/9] drm/gem-vram: Support pinning buffers to current location Thomas Zimmermann
2019-06-12 8:13 ` Gerd Hoffmann
2019-06-12 8:29 ` Thomas Zimmermann
2019-06-12 12:30 ` Gerd Hoffmann
2019-06-11 13:03 ` [PATCH v2 2/9] drm/ast: Unpin cursor BO during cleanup Thomas Zimmermann
2019-06-11 13:03 ` [PATCH v2 3/9] drm/ast: Remove obsolete or unused cursor state Thomas Zimmermann
2019-06-11 13:03 ` [PATCH v2 4/9] drm/ast: Pin and map cursor source BO during update Thomas Zimmermann
2019-06-12 8:15 ` Gerd Hoffmann
2019-06-11 13:03 ` [PATCH v2 5/9] drm/ast: Pin framebuffer BO during dirty update Thomas Zimmermann
2019-06-11 20:29 ` Sam Ravnborg [this message]
2019-06-12 7:35 ` Thomas Zimmermann
2019-06-12 7:48 ` Sam Ravnborg
2019-06-11 13:03 ` [PATCH v2 6/9] drm/mgag200: " Thomas Zimmermann
2019-06-11 13:03 ` [PATCH v2 7/9] drm/mgag200: Rewrite cursor handling Thomas Zimmermann
2019-06-12 8:17 ` Gerd Hoffmann
2019-06-11 13:03 ` [PATCH v2 8/9] drm: Remove lock interfaces from GEM VRAM helpers Thomas Zimmermann
2019-06-11 13:03 ` [PATCH v2 9/9] drm: Remove functions with kmap-object argument " Thomas Zimmermann
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=20190611202953.GA18315@ravnborg.org \
--to=sam@ravnborg.org \
--cc=airlied@redhat.com \
--cc=dri-devel@lists.freedesktop.org \
--cc=kraxel@redhat.com \
--cc=maxime.ripard@bootlin.com \
--cc=sean@poorly.run \
--cc=tzimmermann@suse.de \
/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