From: Thomas Zimmermann <tzimmermann@suse.de>
To: Igor Matheus Andrade Torrente <igormtorrente@gmail.com>,
rodrigosiqueiramelo@gmail.com, melissa.srw@gmail.com
Cc: hamohammed.sa@gmail.com, daniel@ffwll.ch, airlied@linux.ie,
contact@emersion.fr, leandro.ribeiro@collabora.com,
dri-devel@lists.freedesktop.org, linux-kernel@vger.kernel.org,
lkcamp@lists.libreplanetbr.org
Subject: Re: [PATCH 4/6] drm: vkms: Add fb information to `vkms_writeback_job`
Date: Mon, 18 Oct 2021 12:10:34 +0200 [thread overview]
Message-ID: <32c0eb54-46af-4cbd-34eb-c49e3d2505bc@suse.de> (raw)
In-Reply-To: <20211005201637.58563-5-igormtorrente@gmail.com>
[-- Attachment #1.1: Type: text/plain, Size: 4236 bytes --]
Hi
Am 05.10.21 um 22:16 schrieb Igor Matheus Andrade Torrente:
> This commit is the groundwork to introduce new formats to the planes and
> writeback buffer. As part of it, a new buffer metadata field is added to
> `vkms_writeback_job`, this metadata is represented by the `vkms_composer`
> struct.
>
> This will allow us, in the future, to have different compositing and wb
> format types.
>
> Signed-off-by: Igor Matheus Andrade Torrente <igormtorrente@gmail.com>
> ---
> drivers/gpu/drm/vkms/vkms_drv.h | 10 +++++-----
> drivers/gpu/drm/vkms/vkms_writeback.c | 16 +++++++++++++---
> 2 files changed, 18 insertions(+), 8 deletions(-)
>
> diff --git a/drivers/gpu/drm/vkms/vkms_drv.h b/drivers/gpu/drm/vkms/vkms_drv.h
> index 64e62993b06f..d62f8ebd454b 100644
> --- a/drivers/gpu/drm/vkms/vkms_drv.h
> +++ b/drivers/gpu/drm/vkms/vkms_drv.h
> @@ -20,11 +20,6 @@
> #define XRES_MAX 8192
> #define YRES_MAX 8192
>
> -struct vkms_writeback_job {
> - struct dma_buf_map map[DRM_FORMAT_MAX_PLANES];
> - struct dma_buf_map data[DRM_FORMAT_MAX_PLANES];
> -};
> -
> struct vkms_composer {
> struct drm_framebuffer fb;
> struct drm_rect src, dst;
> @@ -34,6 +29,11 @@ struct vkms_composer {
> unsigned int cpp;
> };
>
> +struct vkms_writeback_job {
> + struct dma_buf_map data[DRM_FORMAT_MAX_PLANES];
> + struct vkms_composer composer;
> +};
> +
> /**
> * vkms_plane_state - Driver specific plane state
> * @base: base plane state
> diff --git a/drivers/gpu/drm/vkms/vkms_writeback.c b/drivers/gpu/drm/vkms/vkms_writeback.c
> index 8694227f555f..5a3e12f105dc 100644
> --- a/drivers/gpu/drm/vkms/vkms_writeback.c
> +++ b/drivers/gpu/drm/vkms/vkms_writeback.c
> @@ -75,7 +75,7 @@ static int vkms_wb_prepare_job(struct drm_writeback_connector *wb_connector,
> if (!vkmsjob)
> return -ENOMEM;
>
> - ret = drm_gem_fb_vmap(job->fb, vkmsjob->map, vkmsjob->data);
> + ret = drm_gem_fb_vmap(job->fb, vkmsjob->composer.map, vkmsjob->data);
> if (ret) {
> DRM_ERROR("vmap failed: %d\n", ret);
> goto err_kfree;
> @@ -99,7 +99,7 @@ static void vkms_wb_cleanup_job(struct drm_writeback_connector *connector,
> if (!job->fb)
> return;
>
> - drm_gem_fb_vunmap(job->fb, vkmsjob->map);
> + drm_gem_fb_vunmap(job->fb, vkmsjob->composer.map);
>
> vkmsdev = drm_device_to_vkms_device(job->fb->dev);
> vkms_set_composer(&vkmsdev->output, false);
> @@ -116,14 +116,24 @@ static void vkms_wb_atomic_commit(struct drm_connector *conn,
> struct drm_writeback_connector *wb_conn = &output->wb_connector;
> struct drm_connector_state *conn_state = wb_conn->base.state;
> struct vkms_crtc_state *crtc_state = output->composer_state;
> + struct drm_framebuffer *fb = connector_state->writeback_job->fb;
> + struct vkms_writeback_job *active_wb;
> + struct vkms_composer *wb_composer;
>
> if (!conn_state)
> return;
>
> vkms_set_composer(&vkmsdev->output, true);
>
> + active_wb = conn_state->writeback_job->priv;
> + wb_composer = &active_wb->composer;
> +
> spin_lock_irq(&output->composer_lock);
> - crtc_state->active_writeback = conn_state->writeback_job->priv;
> + crtc_state->active_writeback = active_wb;
> + memcpy(&wb_composer->fb, fb, sizeof(struct drm_framebuffer));
> + wb_composer->offset = fb->offsets[0];
> + wb_composer->pitch = fb->pitches[0];
> + wb_composer->cpp = fb->format->cpp[0];
I don't know the exact cleanup requirements of these structures, but
copying fb into composer looks like it breaks reference counting of the
framebuffer's gem objects at least.
This atomic_commit helper would better acquire the correct references
and hand them over to the compositor thread. Why not have a ref-counted
composer or get a reference on the framebuffer instead?
Best regards
Thomas
> crtc_state->wb_pending = true;
> spin_unlock_irq(&output->composer_lock);
> drm_writeback_queue_job(wb_conn, connector_state);
>
--
Thomas Zimmermann
Graphics Driver Developer
SUSE Software Solutions Germany GmbH
Maxfeldstr. 5, 90409 Nürnberg, Germany
(HRB 36809, AG Nürnberg)
Geschäftsführer: Felix Imendörffer
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 840 bytes --]
next prev parent reply other threads:[~2021-10-18 10:10 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-10-05 20:16 [PATCH 0/6] Refactor the vkms to accept new formats Igor Matheus Andrade Torrente
2021-10-05 20:16 ` [PATCH 1/6] drm: vkms: Replace the deprecated drm_mode_config_init Igor Matheus Andrade Torrente
2021-10-18 10:02 ` Thomas Zimmermann
2021-10-18 18:21 ` Igor Matheus Andrade Torrente
2021-10-05 20:16 ` [PATCH 2/6] drm: vkms: Alloc the compose frame using vzalloc Igor Matheus Andrade Torrente
2021-10-05 20:16 ` [PATCH 3/6] drm: vkms: Replace hardcoded value of `vkms_composer.map` to DRM_FORMAT_MAX_PLANES Igor Matheus Andrade Torrente
2021-10-18 10:04 ` Thomas Zimmermann
2021-10-05 20:16 ` [PATCH 4/6] drm: vkms: Add fb information to `vkms_writeback_job` Igor Matheus Andrade Torrente
2021-10-18 10:10 ` Thomas Zimmermann [this message]
2021-10-05 20:16 ` [PATCH 5/6] drm: vkms: Prepare `vkms_wb_encoder_atomic_check` to accept multiple formats Igor Matheus Andrade Torrente
2021-10-18 10:14 ` Thomas Zimmermann
2021-10-18 17:41 ` Igor Matheus Andrade Torrente
2021-10-18 18:06 ` Thomas Zimmermann
2021-10-18 19:32 ` Igor Matheus Andrade Torrente
2021-10-19 7:17 ` Thomas Zimmermann
2021-10-19 19:12 ` Igor Matheus Andrade Torrente
2021-10-05 20:16 ` [PATCH 6/6] drm: vkms: Refactor the plane composer to accept new formats Igor Matheus Andrade Torrente
2021-10-05 22:20 ` kernel test robot
2021-10-05 23:18 ` kernel test robot
2021-10-05 23:36 ` kernel test robot
2021-10-18 8:30 ` Pekka Paalanen
2021-10-18 19:26 ` Igor Matheus Andrade Torrente
2021-10-19 8:05 ` Pekka Paalanen
2021-10-19 21:10 ` Igor Matheus Andrade Torrente
2021-10-20 8:25 ` Pekka Paalanen
2021-10-18 7:53 ` [PATCH 0/6] Refactor the vkms " Pekka Paalanen
2021-10-18 18:05 ` Igor Matheus Andrade Torrente
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=32c0eb54-46af-4cbd-34eb-c49e3d2505bc@suse.de \
--to=tzimmermann@suse.de \
--cc=airlied@linux.ie \
--cc=contact@emersion.fr \
--cc=daniel@ffwll.ch \
--cc=dri-devel@lists.freedesktop.org \
--cc=hamohammed.sa@gmail.com \
--cc=igormtorrente@gmail.com \
--cc=leandro.ribeiro@collabora.com \
--cc=linux-kernel@vger.kernel.org \
--cc=lkcamp@lists.libreplanetbr.org \
--cc=melissa.srw@gmail.com \
--cc=rodrigosiqueiramelo@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
all inboxes | Powered by JetHome®