v1:
remove the selected DRM_KMS_CMA_HELPER in kconfig
drm-sprd-fix-potential-NULL-dereference
Kevin Tang (2):
drm/sprd: remove the selected DRM_KMS_CMA_HELPER in kconfig
drm/sprd: fix potential NULL dereference
drivers/gpu/drm/sprd/Kconfig | 1 -
drivers/gpu/drm/sprd/sprd_dpu.c | 3 +++
drivers/gpu/drm/sprd/sprd_drm.c | 8 ++------
drivers/gpu/drm/sprd/sprd_dsi.c | 3 +++
4 files changed, 8 insertions(+), 7 deletions(-)
--
2.29.0
On linux-next, commit 43531edd53f0 ("drm/sprd: add Unisoc's drm kms master") adds the config DRM_SPRD,
which selects DRM_KMS_CMA_HELPER.
However, commit 09717af7d13d ("drm: Remove CONFIG_DRM_KMS_CMA_HELPER option") just removed the
DRM_KMS_CMA_HELPER. So, the select DRM_KMS_CMA_HELPER refers to a non-existing kconfig symbol.
Cc: Orson Zhai <[email protected]>
Cc: Chunyan Zhang <[email protected]>
Signed-off-by: Kevin Tang <[email protected]>
---
drivers/gpu/drm/sprd/Kconfig | 1 -
1 file changed, 1 deletion(-)
diff --git a/drivers/gpu/drm/sprd/Kconfig b/drivers/gpu/drm/sprd/Kconfig
index 3edeaeca0..9a9c7ebfc 100644
--- a/drivers/gpu/drm/sprd/Kconfig
+++ b/drivers/gpu/drm/sprd/Kconfig
@@ -3,7 +3,6 @@ config DRM_SPRD
depends on ARCH_SPRD || COMPILE_TEST
depends on DRM && OF
select DRM_GEM_CMA_HELPER
- select DRM_KMS_CMA_HELPER
select DRM_KMS_HELPER
select DRM_MIPI_DSI
select VIDEOMODE_HELPERS
--
2.29.0
platform_get_resource() may fail and return NULL, so check it's value
before using it.
'drm' could be null in sprd_drm_shutdown, and drm_warn maybe dereference
it, remove this warning log.
Cc: Orson Zhai <[email protected]>
Cc: Chunyan Zhang <[email protected]>
Signed-off-by: Kevin Tang <[email protected]>
---
drivers/gpu/drm/sprd/sprd_dpu.c | 3 +++
drivers/gpu/drm/sprd/sprd_drm.c | 8 ++------
drivers/gpu/drm/sprd/sprd_dsi.c | 3 +++
3 files changed, 8 insertions(+), 6 deletions(-)
diff --git a/drivers/gpu/drm/sprd/sprd_dpu.c b/drivers/gpu/drm/sprd/sprd_dpu.c
index 06a3414ee..69683b7ba 100644
--- a/drivers/gpu/drm/sprd/sprd_dpu.c
+++ b/drivers/gpu/drm/sprd/sprd_dpu.c
@@ -790,6 +790,9 @@ static int sprd_dpu_context_init(struct sprd_dpu *dpu,
int ret;
res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+ if (!res)
+ return -EINVAL;
+
ctx->base = devm_ioremap(dev, res->start, resource_size(res));
if (!ctx->base) {
dev_err(dev, "failed to map dpu registers\n");
diff --git a/drivers/gpu/drm/sprd/sprd_drm.c b/drivers/gpu/drm/sprd/sprd_drm.c
index a077e2d4d..54030839e 100644
--- a/drivers/gpu/drm/sprd/sprd_drm.c
+++ b/drivers/gpu/drm/sprd/sprd_drm.c
@@ -154,12 +154,8 @@ static void sprd_drm_shutdown(struct platform_device *pdev)
{
struct drm_device *drm = platform_get_drvdata(pdev);
- if (!drm) {
- drm_warn(drm, "drm device is not available, no shutdown\n");
- return;
- }
-
- drm_atomic_helper_shutdown(drm);
+ if (drm)
+ drm_atomic_helper_shutdown(drm);
}
static const struct of_device_id drm_match_table[] = {
diff --git a/drivers/gpu/drm/sprd/sprd_dsi.c b/drivers/gpu/drm/sprd/sprd_dsi.c
index 911b3cddc..955c5995a 100644
--- a/drivers/gpu/drm/sprd/sprd_dsi.c
+++ b/drivers/gpu/drm/sprd/sprd_dsi.c
@@ -907,6 +907,9 @@ static int sprd_dsi_context_init(struct sprd_dsi *dsi,
struct resource *res;
res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+ if (!res)
+ return -EINVAL;
+
ctx->base = devm_ioremap(dev, res->start, resource_size(res));
if (!ctx->base) {
drm_err(dsi->drm, "failed to map dsi host registers\n");
--
2.29.0
Hello Kevin,
On 12/24/21 15:12, Kevin Tang wrote:
> On linux-next, commit 43531edd53f0 ("drm/sprd: add Unisoc's drm kms master") adds the config DRM_SPRD,
> which selects DRM_KMS_CMA_HELPER.
>
According to "The canonical patch format" section in [0], the body of the
explanation has to be line wrapped at 75 columns. But your sentences are
much longer than that.
[0]: https://www.kernel.org/doc/html/latest/process/submitting-patches.html#the-canonical-patch-format
> However, commit 09717af7d13d ("drm: Remove CONFIG_DRM_KMS_CMA_HELPER option") just removed the
> DRM_KMS_CMA_HELPER. So, the select DRM_KMS_CMA_HELPER refers to a non-existing kconfig symbol.
>
> Cc: Orson Zhai <[email protected]>
> Cc: Chunyan Zhang <[email protected]>
> Signed-off-by: Kevin Tang <[email protected]>
> ---
Other than that, the patch looks good to me.
Reviewed-by: Javier Martinez Canillas <[email protected]>
Best regards,
--
Javier Martinez Canillas
Linux Engineering
Red Hat
On 12/24/21 15:12, Kevin Tang wrote:
> platform_get_resource() may fail and return NULL, so check it's value
> before using it.
>
> 'drm' could be null in sprd_drm_shutdown, and drm_warn maybe dereference
> it, remove this warning log.
>
I would split this second change in a separate patch and just keep this
one about checking the platform_get_resource() return value.
If you do that, feel free to add:
Reviewed-by: Javier Martinez Canillas <[email protected]>
Best regards,
--
Javier Martinez Canillas
Linux Engineering
Red Hat
Hi
Am 24.12.21 um 15:12 schrieb Kevin Tang:
> On linux-next, commit 43531edd53f0 ("drm/sprd: add Unisoc's drm kms master") adds the config DRM_SPRD,
> which selects DRM_KMS_CMA_HELPER.
>
> However, commit 09717af7d13d ("drm: Remove CONFIG_DRM_KMS_CMA_HELPER option") just removed the
> DRM_KMS_CMA_HELPER. So, the select DRM_KMS_CMA_HELPER refers to a non-existing kconfig symbol.
With the long lines fixed, you can add
Acked-by: Thomas Zimmermann <[email protected]>
Selecting DRM_GEM_CMA_HELPER and DRM_KMS_HELPER at the same time acts
like DRM_KMS_CMA_HELPER did before.
Best regards
Thomas
>
> Cc: Orson Zhai <[email protected]>
> Cc: Chunyan Zhang <[email protected]>
> Signed-off-by: Kevin Tang <[email protected]>
> ---
> drivers/gpu/drm/sprd/Kconfig | 1 -
> 1 file changed, 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/sprd/Kconfig b/drivers/gpu/drm/sprd/Kconfig
> index 3edeaeca0..9a9c7ebfc 100644
> --- a/drivers/gpu/drm/sprd/Kconfig
> +++ b/drivers/gpu/drm/sprd/Kconfig
> @@ -3,7 +3,6 @@ config DRM_SPRD
> depends on ARCH_SPRD || COMPILE_TEST
> depends on DRM && OF
> select DRM_GEM_CMA_HELPER
> - select DRM_KMS_CMA_HELPER
> select DRM_KMS_HELPER
> select DRM_MIPI_DSI
> select VIDEOMODE_HELPERS
--
Thomas Zimmermann
Graphics Driver Developer
SUSE Software Solutions Germany GmbH
Maxfeldstr. 5, 90409 Nürnberg, Germany
(HRB 36809, AG Nürnberg)
Geschäftsführer: Ivo Totev
On Fri, Dec 24, 2021 at 3:12 PM Kevin Tang <[email protected]> wrote:
>
> On linux-next, commit 43531edd53f0 ("drm/sprd: add Unisoc's drm kms master") adds the config DRM_SPRD,
> which selects DRM_KMS_CMA_HELPER.
>
That this commit is _currently_ on linux-next is just a matter of the
current state. The commit message that goes into the project's history
should probably not state "on linux-next";
this information is probably outdated or of no interest to any further
future reader at the time of reading.
So, just drop "On linux-next". The commit 43531edd53f0 ("drm/sprd: add
Unisoc's drm kms master") will exist until the end of time.
> However, commit 09717af7d13d ("drm: Remove CONFIG_DRM_KMS_CMA_HELPER option") just removed the
> DRM_KMS_CMA_HELPER. So, the select DRM_KMS_CMA_HELPER refers to a non-existing kconfig symbol.
>
I would be happy about acknowledging my work of reporting with a
Reported-by tag, although I accidently send the report only to you
without cc-ing the mailing lists.
Please add:
Reported-by: Lukas Bulwahn <[email protected]>
That said you may also add a Reviewed-by tag now:
Reviewed-by: Lukas Bulwahn <[email protected]>
Lukas
> Cc: Orson Zhai <[email protected]>
> Cc: Chunyan Zhang <[email protected]>
> Signed-off-by: Kevin Tang <[email protected]>
> ---
> drivers/gpu/drm/sprd/Kconfig | 1 -
> 1 file changed, 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/sprd/Kconfig b/drivers/gpu/drm/sprd/Kconfig
> index 3edeaeca0..9a9c7ebfc 100644
> --- a/drivers/gpu/drm/sprd/Kconfig
> +++ b/drivers/gpu/drm/sprd/Kconfig
> @@ -3,7 +3,6 @@ config DRM_SPRD
> depends on ARCH_SPRD || COMPILE_TEST
> depends on DRM && OF
> select DRM_GEM_CMA_HELPER
> - select DRM_KMS_CMA_HELPER
> select DRM_KMS_HELPER
> select DRM_MIPI_DSI
> select VIDEOMODE_HELPERS
> --
> 2.29.0
>
Hi,
on the changes for platform_get_resource(), you can
Acked-by: Thomas Zimmermann <[email protected]>
but see my comments below.
Am 24.12.21 um 15:12 schrieb Kevin Tang:
> platform_get_resource() may fail and return NULL, so check it's value
> before using it.
>
> 'drm' could be null in sprd_drm_shutdown, and drm_warn maybe dereference
> it, remove this warning log.
>
> Cc: Orson Zhai <[email protected]>
> Cc: Chunyan Zhang <[email protected]>
> Signed-off-by: Kevin Tang <[email protected]>
> ---
> drivers/gpu/drm/sprd/sprd_dpu.c | 3 +++
> drivers/gpu/drm/sprd/sprd_drm.c | 8 ++------
> drivers/gpu/drm/sprd/sprd_dsi.c | 3 +++
> 3 files changed, 8 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/gpu/drm/sprd/sprd_dpu.c b/drivers/gpu/drm/sprd/sprd_dpu.c
> index 06a3414ee..69683b7ba 100644
> --- a/drivers/gpu/drm/sprd/sprd_dpu.c
> +++ b/drivers/gpu/drm/sprd/sprd_dpu.c
> @@ -790,6 +790,9 @@ static int sprd_dpu_context_init(struct sprd_dpu *dpu,
> int ret;
>
> res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
> + if (!res)
> + return -EINVAL;
> +
You can add an error message if this fails.
> ctx->base = devm_ioremap(dev, res->start, resource_size(res));
> if (!ctx->base) {
> dev_err(dev, "failed to map dpu registers\n");
> diff --git a/drivers/gpu/drm/sprd/sprd_drm.c b/drivers/gpu/drm/sprd/sprd_drm.c
> index a077e2d4d..54030839e 100644
> --- a/drivers/gpu/drm/sprd/sprd_drm.c
> +++ b/drivers/gpu/drm/sprd/sprd_drm.c
> @@ -154,12 +154,8 @@ static void sprd_drm_shutdown(struct platform_device *pdev)
> {
> struct drm_device *drm = platform_get_drvdata(pdev);
>
> - if (!drm) {
> - drm_warn(drm, "drm device is not available, no shutdown\n");
> - return;
> - }
> -
> - drm_atomic_helper_shutdown(drm);
> + if (drm)
> + drm_atomic_helper_shutdown(drm);
This change should be in a separate patch. Instead of removing the
warning, you should rather use dev_err() or dev_warn() from [1]. Not
being able to shut down here looks like a serious driver bug that the
user should know about.
> }
>
> static const struct of_device_id drm_match_table[] = {
> diff --git a/drivers/gpu/drm/sprd/sprd_dsi.c b/drivers/gpu/drm/sprd/sprd_dsi.c
> index 911b3cddc..955c5995a 100644
> --- a/drivers/gpu/drm/sprd/sprd_dsi.c
> +++ b/drivers/gpu/drm/sprd/sprd_dsi.c
> @@ -907,6 +907,9 @@ static int sprd_dsi_context_init(struct sprd_dsi *dsi,
> struct resource *res;
>
> res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
> + if (!res)
> + return -EINVAL;
> +
Again, an error message seems appropriate.
Best regards
Thomas
[1]
https://elixir.bootlin.com/linux/latest/source/include/linux/dev_printk.h#L145
> ctx->base = devm_ioremap(dev, res->start, resource_size(res));
> if (!ctx->base) {
> drm_err(dsi->drm, "failed to map dsi host registers\n");
--
Thomas Zimmermann
Graphics Driver Developer
SUSE Software Solutions Germany GmbH
Maxfeldstr. 5, 90409 Nürnberg, Germany
(HRB 36809, AG Nürnberg)
Geschäftsführer: Ivo Totev