* [PATCH v2] drm/gma500: Create the primary plane in the driver
@ 2026-09-03 7:11 oushixiong1025
2026-09-03 7:30 ` sashiko-bot
0 siblings, 1 reply; 2+ messages in thread
From: oushixiong1025 @ 2026-09-03 7:11 UTC (permalink / raw)
To: Patrik Jakobsson
Cc: Maarten Lankhorst, Maxime Ripard, Thomas Zimmermann, David Airlie,
Simona Vetter, dri-devel, linux-kernel, Shixiong Ou
From: Shixiong Ou <oushixiong@kylinos.cn>
drm_crtc_init() creates the primary plane from a fixed format list
that includes ARGB8888. The display engine programs the primary plane
with DISPPLANE_32BPP_NO_ALPHA, so it does not support per-pixel alpha
and must not advertise alpha formats. Since commit 860e748bddcc
("drm: ensure blend mode supported if pixel format with alpha exposed"),
drm_mode_config_validate() warns about this at probe time.
Replace drm_crtc_init() with a driver-owned primary plane that
advertises only XRGB8888. The plane is allocated by
drmm_universal_plane_alloc() and the "pixel blend mode" property is
not needed because no format with alpha is exposed.
Signed-off-by: Shixiong Ou <oushixiong@kylinos.cn>
---
v1->v2:
- Open-code DRM_PLANE_NON_ATOMIC_FUNCS without .destroy, which
drmm_universal_plane_alloc() rejects
- Use drmm_universal_plane_alloc() for automatic cleanup
- Move the formats comment above the array
drivers/gpu/drm/gma500/psb_intel_display.c | 34 +++++++++++++++++++++-
1 file changed, 33 insertions(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/gma500/psb_intel_display.c b/drivers/gpu/drm/gma500/psb_intel_display.c
index 0df75a4a7739..f79484b9d593 100644
--- a/drivers/gpu/drm/gma500/psb_intel_display.c
+++ b/drivers/gpu/drm/gma500/psb_intel_display.c
@@ -9,8 +9,11 @@
#include <linux/delay.h>
#include <linux/i2c.h>
+#include <drm/drm_fourcc.h>
#include <drm/drm_modeset_helper.h>
#include <drm/drm_modeset_helper_vtables.h>
+#include <drm/drm_plane.h>
+#include <drm/drm_plane_helper.h>
#include <drm/drm_print.h>
#include "framebuffer.h"
@@ -471,11 +474,25 @@ static void psb_intel_cursor_init(struct drm_device *dev,
REG_WRITE(base[gma_crtc->pipe], 0);
}
+/*
+ * The display engine programs the primary plane with
+ * DISPPLANE_32BPP_NO_ALPHA, so no formats with alpha.
+ */
+static const uint32_t gma_primary_formats[] = {
+ DRM_FORMAT_XRGB8888,
+};
+
+static const struct drm_plane_funcs gma_primary_plane_funcs = {
+ .update_plane = drm_plane_helper_update_primary,
+ .disable_plane = drm_plane_helper_disable_primary,
+};
+
void psb_intel_crtc_init(struct drm_device *dev, int pipe,
struct psb_intel_mode_device *mode_dev)
{
struct drm_psb_private *dev_priv = to_drm_psb_private(dev);
struct gma_crtc *gma_crtc;
+ struct drm_plane *primary;
int i;
/* We allocate a extra array of drm_connector pointers
@@ -494,7 +511,17 @@ void psb_intel_crtc_init(struct drm_device *dev, int pipe,
return;
}
- drm_crtc_init(dev, &gma_crtc->base, &gma_crtc_funcs);
+ primary = drmm_universal_plane_alloc(dev, struct drm_plane, dev, 0,
+ &gma_primary_plane_funcs,
+ gma_primary_formats,
+ ARRAY_SIZE(gma_primary_formats),
+ NULL, DRM_PLANE_TYPE_PRIMARY, NULL);
+ if (IS_ERR(primary))
+ goto err_free_crtc;
+
+ if (drm_crtc_init_with_planes(dev, &gma_crtc->base, primary, NULL,
+ &gma_crtc_funcs, NULL))
+ goto err_free_crtc;
/* Set the CRTC clock functions from chip specific data */
gma_crtc->clock_funcs = dev_priv->ops->clock_funcs;
@@ -524,6 +551,11 @@ void psb_intel_crtc_init(struct drm_device *dev, int pipe,
/* Set to true so that the pipe is forced off on initial config. */
gma_crtc->active = true;
+ return;
+
+err_free_crtc:
+ kfree(gma_crtc->crtc_state);
+ kfree(gma_crtc);
}
struct drm_crtc *psb_intel_get_crtc_from_pipe(struct drm_device *dev, int pipe)
--
2.25.1
No virus found
Checked by Hillstone Network AntiVirus
^ permalink raw reply related [flat|nested] 2+ messages in thread
* Re: [PATCH v2] drm/gma500: Create the primary plane in the driver
2026-09-03 7:11 [PATCH v2] drm/gma500: Create the primary plane in the driver oushixiong1025
@ 2026-09-03 7:30 ` sashiko-bot
0 siblings, 0 replies; 2+ messages in thread
From: sashiko-bot @ 2026-09-03 7:30 UTC (permalink / raw)
To: oushixiong1025; +Cc: dri-devel
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] Freeing `gma_crtc` on `drm_crtc_init_with_planes()` failure causes a Use-After-Free because the CRTC remains linked in `dev->mode_config.crtc_list`.
--
commit 2d329a448a74770c14b7d09aef862579fa99672e
Author: Shixiong Ou <oushixiong@kylinos.cn>
drm/gma500: Create the primary plane in the driver
drm_crtc_init() creates the primary plane from a fixed format list
that includes ARGB8888. The display engine programs the primary plane
with DISPPLANE_32BPP_NO_ALPHA, so it does not support per-pixel alpha
and must not advertise alpha formats. This commit replaces drm_crtc_init()
with a driver-owned primary plane that advertises only XRGB8888.
> diff --git a/drivers/gpu/drm/gma500/psb_intel_display.c b/drivers/gpu/drm/gma500/psb_intel_display.c
> index 0df75a4a77396..f79484b9d593e 100644
> --- a/drivers/gpu/drm/gma500/psb_intel_display.c
> +++ b/drivers/gpu/drm/gma500/psb_intel_display.c
[ ... ]
> @@ -494,7 +511,17 @@ void psb_intel_crtc_init(struct drm_device *dev, int pipe,
> return;
> }
>
> - drm_crtc_init(dev, &gma_crtc->base, &gma_crtc_funcs);
> + primary = drmm_universal_plane_alloc(dev, struct drm_plane, dev, 0,
> + &gma_primary_plane_funcs,
> + gma_primary_formats,
> + ARRAY_SIZE(gma_primary_formats),
> + NULL, DRM_PLANE_TYPE_PRIMARY, NULL);
> + if (IS_ERR(primary))
> + goto err_free_crtc;
> +
> + if (drm_crtc_init_with_planes(dev, &gma_crtc->base, primary, NULL,
> + &gma_crtc_funcs, NULL))
> + goto err_free_crtc;
[Severity: High]
Could this error handling path leave a freed pointer linked in the DRM core?
Looking at __drm_crtc_init_with_planes(), the CRTC is added to the list early
in the function:
list_add_tail(&crtc->head, &config->crtc_list);
crtc->index = config->num_crtc++;
If a subsequent initialization step fails (such as drm_crtc_crc_init()),
__drm_crtc_init_with_planes() returns an error without removing the CRTC from
the list.
When jumping to err_free_crtc here, gma_crtc is freed via kfree() below. Might
this create a use-after-free condition when dev->mode_config.crtc_list is later
traversed?
[ ... ]
> @@ -524,6 +551,11 @@ void psb_intel_crtc_init(struct drm_device *dev, int pipe,
>
> /* Set to true so that the pipe is forced off on initial config. */
> gma_crtc->active = true;
> + return;
> +
> +err_free_crtc:
> + kfree(gma_crtc->crtc_state);
> + kfree(gma_crtc);
> }
>
> struct drm_crtc *psb_intel_get_crtc_from_pipe(struct drm_device *dev, int pipe)
--
Sashiko AI review · https://sashiko.dev/#/patchset/20260903071149.423579-1-oushixiong1025@163.com?part=1
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2026-09-03 7:30 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-09-03 7:11 [PATCH v2] drm/gma500: Create the primary plane in the driver oushixiong1025
2026-09-03 7:30 ` sashiko-bot
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox