From: Paul Cercueil <paul@crapouillou.net>
To: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>,
Maxime Ripard <mripard@kernel.org>,
Thomas Zimmermann <tzimmermann@suse.de>,
David Airlie <airlied@linux.ie>, Daniel Vetter <daniel@ffwll.ch>
Cc: Christoph Hellwig <hch@infradead.org>,
list@opendingux.net, dri-devel@lists.freedesktop.org,
linux-kernel@vger.kernel.org, linux-mips@vger.kernel.org,
Paul Cercueil <paul@crapouillou.net>
Subject: [PATCH v5 2/3] drm: Add and export function drm_fb_cma_sync_non_coherent
Date: Sun, 23 May 2021 18:04:14 +0100 [thread overview]
Message-ID: <20210523170415.90410-3-paul@crapouillou.net> (raw)
In-Reply-To: <20210523170415.90410-1-paul@crapouillou.net>
This function can be used by drivers that use damage clips and have
CMA GEM objects backed by non-coherent memory. Calling this function
in a plane's .atomic_update ensures that all the data in the backing
memory have been written to RAM.
v3: - Only sync data if using GEM objects backed by non-coherent memory.
- Use a drm_device pointer instead of device pointer in prototype
v5: - Rename to drm_fb_cma_sync_non_coherent
- Invert loops for better cache locality
- Only sync BOs that have the non-coherent flag
- Move to drm_fb_cma_helper.c to avoid circular dependency
Signed-off-by: Paul Cercueil <paul@crapouillou.net>
---
drivers/gpu/drm/drm_fb_cma_helper.c | 46 +++++++++++++++++++++++++++++
include/drm/drm_fb_cma_helper.h | 4 +++
2 files changed, 50 insertions(+)
diff --git a/drivers/gpu/drm/drm_fb_cma_helper.c b/drivers/gpu/drm/drm_fb_cma_helper.c
index cb2349ad338d..69c57273b184 100644
--- a/drivers/gpu/drm/drm_fb_cma_helper.c
+++ b/drivers/gpu/drm/drm_fb_cma_helper.c
@@ -9,12 +9,14 @@
* Copyright (C) 2012 Red Hat
*/
+#include <drm/drm_damage_helper.h>
#include <drm/drm_fb_cma_helper.h>
#include <drm/drm_fourcc.h>
#include <drm/drm_framebuffer.h>
#include <drm/drm_gem_cma_helper.h>
#include <drm/drm_gem_framebuffer_helper.h>
#include <drm/drm_plane.h>
+#include <linux/dma-mapping.h>
#include <linux/module.h>
/**
@@ -97,3 +99,47 @@ dma_addr_t drm_fb_cma_get_gem_addr(struct drm_framebuffer *fb,
return paddr;
}
EXPORT_SYMBOL_GPL(drm_fb_cma_get_gem_addr);
+
+/**
+ * drm_fb_cma_sync_non_coherent - Sync GEM object to non-coherent backing
+ * memory
+ * @drm: DRM device
+ * @old_state: Old plane state
+ * @state: New plane state
+ *
+ * This function can be used by drivers that use damage clips and have
+ * CMA GEM objects backed by non-coherent memory. Calling this function
+ * in a plane's .atomic_update ensures that all the data in the backing
+ * memory have been written to RAM.
+ */
+void drm_fb_cma_sync_non_coherent(struct drm_device *drm,
+ struct drm_plane_state *old_state,
+ struct drm_plane_state *state)
+{
+ const struct drm_format_info *finfo = state->fb->format;
+ struct drm_atomic_helper_damage_iter iter;
+ const struct drm_gem_cma_object *cma_obj;
+ unsigned int offset, i;
+ struct drm_rect clip;
+ dma_addr_t daddr;
+ size_t nb_bytes;
+
+ for (i = 0; i < finfo->num_planes; i++) {
+ cma_obj = drm_fb_cma_get_gem_obj(state->fb, i);
+ if (!cma_obj->map_noncoherent)
+ continue;
+
+ daddr = drm_fb_cma_get_gem_addr(state->fb, state, i);
+ drm_atomic_helper_damage_iter_init(&iter, old_state, state);
+
+ drm_atomic_for_each_plane_damage(&iter, &clip) {
+ /* Ignore x1/x2 values, invalidate complete lines */
+ offset = clip.y1 * state->fb->pitches[i];
+
+ nb_bytes = (clip.y2 - clip.y1) * state->fb->pitches[i];
+ dma_sync_single_for_device(drm->dev, daddr + offset,
+ nb_bytes, DMA_TO_DEVICE);
+ }
+ }
+}
+EXPORT_SYMBOL_GPL(drm_fb_cma_sync_non_coherent);
diff --git a/include/drm/drm_fb_cma_helper.h b/include/drm/drm_fb_cma_helper.h
index 795aea1d0a25..3a177632b17e 100644
--- a/include/drm/drm_fb_cma_helper.h
+++ b/include/drm/drm_fb_cma_helper.h
@@ -14,5 +14,9 @@ dma_addr_t drm_fb_cma_get_gem_addr(struct drm_framebuffer *fb,
struct drm_plane_state *state,
unsigned int plane);
+void drm_fb_cma_sync_non_coherent(struct drm_device *drm,
+ struct drm_plane_state *old_state,
+ struct drm_plane_state *state);
+
#endif
--
2.30.2
next prev parent reply other threads:[~2021-05-23 17:04 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-05-23 17:04 [PATCH v5 0/3] Add option to mmap GEM buffers cached Paul Cercueil
2021-05-23 17:04 ` [PATCH v5 1/3] drm: Add support for GEM buffers backed by non-coherent memory Paul Cercueil
2021-05-27 10:40 ` Tomi Valkeinen
2021-05-27 12:43 ` Paul Cercueil
2021-05-27 23:03 ` [PATCH] drm: Fix for GEM buffers with write-combine memory Paul Cercueil
2021-05-28 5:59 ` Tomi Valkeinen
2021-05-28 10:10 ` Paul Cercueil
2021-05-23 17:04 ` Paul Cercueil [this message]
2021-05-23 18:40 ` [PATCH v5 2/3] drm: Add and export function drm_fb_cma_sync_non_coherent kernel test robot
2021-05-23 17:04 ` [PATCH v5 3/3] drm/ingenic: Add option to alloc cached GEM buffers Paul Cercueil
2021-05-23 19:05 ` [PATCH v5 0/3] Add option to mmap GEM buffers cached Thomas Zimmermann
2021-05-23 19:19 ` Paul Cercueil
2021-05-25 10:03 ` 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=20210523170415.90410-3-paul@crapouillou.net \
--to=paul@crapouillou.net \
--cc=airlied@linux.ie \
--cc=daniel@ffwll.ch \
--cc=dri-devel@lists.freedesktop.org \
--cc=hch@infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mips@vger.kernel.org \
--cc=list@opendingux.net \
--cc=maarten.lankhorst@linux.intel.com \
--cc=mripard@kernel.org \
--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;
as well as URLs for NNTP newsgroup(s).