| Message ID | 1471429061-23545-2-git-send-email-daniel.vetter@ffwll.ch (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
Hi, On 17 August 2016 at 18:17, Daniel Vetter <daniel.vetter@ffwll.ch> wrote: > I just broke the build :( > > Note that the cleanup function is a bit confused: ade_data was never > set as drvdata, and calling drm_crtc_cleanup directly is a bug - this Yes, this is a bug. Thanks for pointing out this. > is called indirectly through drm_mode_config_cleanup, which calls into > crtc->destroy, which already has the call to drm_crtc_cleanup. Which > means we can just nuke it. > > Aside: I have no idea why this driver depends upon ARM64. It doesn't > build warning-free on 32bit, but otherwise it's perfectly fine. Because this driver is written for ARM64 SoCs. Thanks, -xinliang > > Cc: Xinliang Liu <z.liuxinliang@hisilicon.com> > Cc: Xinwei Kong <kong.kongxinwei@hisilicon.com> > Cc: Chen Feng <puck.chen@hisilicon.com> > Cc: Sean Paul <seanpaul@chromium.org> > Fixes: d25bcfb8c2e1 ("drm/hisilicon: Don't set drm_device->platformdev") > Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> > --- > drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c | 9 ++------- > drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c | 4 ++-- > 2 files changed, 4 insertions(+), 9 deletions(-) > > diff --git a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c > index 91188f33b1d9..c64c82cb7e71 100644 > --- a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c > +++ b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c > @@ -991,7 +991,7 @@ static int ade_dts_parse(struct platform_device *pdev, struct ade_hw_ctx *ctx) > > static int ade_drm_init(struct platform_device *pdev) > { > - struct drm_device *drm_dev = platform_get_drvdata(dev); > + struct drm_device *dev = platform_get_drvdata(pdev); > struct ade_data *ade; > struct ade_hw_ctx *ctx; > struct ade_crtc *acrtc; > @@ -1052,14 +1052,9 @@ static int ade_drm_init(struct platform_device *pdev) > > static void ade_drm_cleanup(struct platform_device *pdev) > { > - struct drm_device *drm_dev = platform_get_drvdata(dev); > - struct ade_data *ade = platform_get_drvdata(pdev); > - struct drm_crtc *crtc = &ade->acrtc.base; > - > - drm_crtc_cleanup(crtc); > } > > const struct kirin_dc_ops ade_dc_ops = { > .init = ade_drm_init, > .cleanup = ade_drm_cleanup > -; > +}; > diff --git a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c > index 6b0f9f6c16e1..b9b8c25da5e3 100644 > --- a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c > +++ b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c > @@ -103,7 +103,7 @@ static int kirin_drm_kms_init(struct drm_device *dev) > kirin_drm_mode_config_init(dev); > > /* display controller init */ > - ret = dc_ops->init(to_platform_device(dev)); > + ret = dc_ops->init(to_platform_device(dev->dev)); > if (ret) > goto err_mode_config_cleanup; > > @@ -137,7 +137,7 @@ static int kirin_drm_kms_init(struct drm_device *dev) > err_unbind_all: > component_unbind_all(dev->dev, dev); > err_dc_cleanup: > - dc_ops->cleanup(dev); > + dc_ops->cleanup(to_platform_device(dev->dev)); > err_mode_config_cleanup: > drm_mode_config_cleanup(dev); > devm_kfree(dev->dev, priv); > -- > 2.8.1 > > _______________________________________________ > dri-devel mailing list > dri-devel@lists.freedesktop.org > https://lists.freedesktop.org/mailman/listinfo/dri-devel
On Wed, Aug 17, 2016 at 07:02:01PM +0800, Xinliang Liu wrote: > Hi, > > On 17 August 2016 at 18:17, Daniel Vetter <daniel.vetter@ffwll.ch> wrote: > > I just broke the build :( > > > > Note that the cleanup function is a bit confused: ade_data was never > > set as drvdata, and calling drm_crtc_cleanup directly is a bug - this > > Yes, this is a bug. Thanks for pointing out this. > > > is called indirectly through drm_mode_config_cleanup, which calls into > > crtc->destroy, which already has the call to drm_crtc_cleanup. Which > > means we can just nuke it. > > > > Aside: I have no idea why this driver depends upon ARM64. It doesn't > > build warning-free on 32bit, but otherwise it's perfectly fine. > > Because this driver is written for ARM64 SoCs. It makes compile-testing harder, the same reasons why depending upon specific arm platforms is really annoying (but lukily that problem seems to have stopped being one for the drm subsystem at least). Depending on CONFIG_OF and other stuff that's really needed is perfectly fine, but imo depending upon specific platforms and stuff really isn't. -Daniel > > Thanks, > -xinliang > > > > > Cc: Xinliang Liu <z.liuxinliang@hisilicon.com> > > Cc: Xinwei Kong <kong.kongxinwei@hisilicon.com> > > Cc: Chen Feng <puck.chen@hisilicon.com> > > Cc: Sean Paul <seanpaul@chromium.org> > > Fixes: d25bcfb8c2e1 ("drm/hisilicon: Don't set drm_device->platformdev") > > Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> > > --- > > drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c | 9 ++------- > > drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c | 4 ++-- > > 2 files changed, 4 insertions(+), 9 deletions(-) > > > > diff --git a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c > > index 91188f33b1d9..c64c82cb7e71 100644 > > --- a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c > > +++ b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c > > @@ -991,7 +991,7 @@ static int ade_dts_parse(struct platform_device *pdev, struct ade_hw_ctx *ctx) > > > > static int ade_drm_init(struct platform_device *pdev) > > { > > - struct drm_device *drm_dev = platform_get_drvdata(dev); > > + struct drm_device *dev = platform_get_drvdata(pdev); > > struct ade_data *ade; > > struct ade_hw_ctx *ctx; > > struct ade_crtc *acrtc; > > @@ -1052,14 +1052,9 @@ static int ade_drm_init(struct platform_device *pdev) > > > > static void ade_drm_cleanup(struct platform_device *pdev) > > { > > - struct drm_device *drm_dev = platform_get_drvdata(dev); > > - struct ade_data *ade = platform_get_drvdata(pdev); > > - struct drm_crtc *crtc = &ade->acrtc.base; > > - > > - drm_crtc_cleanup(crtc); > > } > > > > const struct kirin_dc_ops ade_dc_ops = { > > .init = ade_drm_init, > > .cleanup = ade_drm_cleanup > > -; > > +}; > > diff --git a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c > > index 6b0f9f6c16e1..b9b8c25da5e3 100644 > > --- a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c > > +++ b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c > > @@ -103,7 +103,7 @@ static int kirin_drm_kms_init(struct drm_device *dev) > > kirin_drm_mode_config_init(dev); > > > > /* display controller init */ > > - ret = dc_ops->init(to_platform_device(dev)); > > + ret = dc_ops->init(to_platform_device(dev->dev)); > > if (ret) > > goto err_mode_config_cleanup; > > > > @@ -137,7 +137,7 @@ static int kirin_drm_kms_init(struct drm_device *dev) > > err_unbind_all: > > component_unbind_all(dev->dev, dev); > > err_dc_cleanup: > > - dc_ops->cleanup(dev); > > + dc_ops->cleanup(to_platform_device(dev->dev)); > > err_mode_config_cleanup: > > drm_mode_config_cleanup(dev); > > devm_kfree(dev->dev, priv); > > -- > > 2.8.1 > > > > _______________________________________________ > > dri-devel mailing list > > dri-devel@lists.freedesktop.org > > https://lists.freedesktop.org/mailman/listinfo/dri-devel
On 17 August 2016 at 19:11, Daniel Vetter <daniel@ffwll.ch> wrote: > On Wed, Aug 17, 2016 at 07:02:01PM +0800, Xinliang Liu wrote: >> Hi, >> >> On 17 August 2016 at 18:17, Daniel Vetter <daniel.vetter@ffwll.ch> wrote: >> > I just broke the build :( >> > >> > Note that the cleanup function is a bit confused: ade_data was never >> > set as drvdata, and calling drm_crtc_cleanup directly is a bug - this >> >> Yes, this is a bug. Thanks for pointing out this. >> >> > is called indirectly through drm_mode_config_cleanup, which calls into >> > crtc->destroy, which already has the call to drm_crtc_cleanup. Which >> > means we can just nuke it. >> > >> > Aside: I have no idea why this driver depends upon ARM64. It doesn't >> > build warning-free on 32bit, but otherwise it's perfectly fine. >> >> Because this driver is written for ARM64 SoCs. > > It makes compile-testing harder, the same reasons why depending upon What compile-testing we will have? I just try to understand the problem. Thanks, -xinliang > specific arm platforms is really annoying (but lukily that problem seems > to have stopped being one for the drm subsystem at least). Depending on > CONFIG_OF and other stuff that's really needed is perfectly fine, but imo > depending upon specific platforms and stuff really isn't. > -Daniel > >> >> Thanks, >> -xinliang >> >> > >> > Cc: Xinliang Liu <z.liuxinliang@hisilicon.com> >> > Cc: Xinwei Kong <kong.kongxinwei@hisilicon.com> >> > Cc: Chen Feng <puck.chen@hisilicon.com> >> > Cc: Sean Paul <seanpaul@chromium.org> >> > Fixes: d25bcfb8c2e1 ("drm/hisilicon: Don't set drm_device->platformdev") >> > Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> >> > --- >> > drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c | 9 ++------- >> > drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c | 4 ++-- >> > 2 files changed, 4 insertions(+), 9 deletions(-) >> > >> > diff --git a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c >> > index 91188f33b1d9..c64c82cb7e71 100644 >> > --- a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c >> > +++ b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c >> > @@ -991,7 +991,7 @@ static int ade_dts_parse(struct platform_device *pdev, struct ade_hw_ctx *ctx) >> > >> > static int ade_drm_init(struct platform_device *pdev) >> > { >> > - struct drm_device *drm_dev = platform_get_drvdata(dev); >> > + struct drm_device *dev = platform_get_drvdata(pdev); >> > struct ade_data *ade; >> > struct ade_hw_ctx *ctx; >> > struct ade_crtc *acrtc; >> > @@ -1052,14 +1052,9 @@ static int ade_drm_init(struct platform_device *pdev) >> > >> > static void ade_drm_cleanup(struct platform_device *pdev) >> > { >> > - struct drm_device *drm_dev = platform_get_drvdata(dev); >> > - struct ade_data *ade = platform_get_drvdata(pdev); >> > - struct drm_crtc *crtc = &ade->acrtc.base; >> > - >> > - drm_crtc_cleanup(crtc); >> > } >> > >> > const struct kirin_dc_ops ade_dc_ops = { >> > .init = ade_drm_init, >> > .cleanup = ade_drm_cleanup >> > -; >> > +}; >> > diff --git a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c >> > index 6b0f9f6c16e1..b9b8c25da5e3 100644 >> > --- a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c >> > +++ b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c >> > @@ -103,7 +103,7 @@ static int kirin_drm_kms_init(struct drm_device *dev) >> > kirin_drm_mode_config_init(dev); >> > >> > /* display controller init */ >> > - ret = dc_ops->init(to_platform_device(dev)); >> > + ret = dc_ops->init(to_platform_device(dev->dev)); >> > if (ret) >> > goto err_mode_config_cleanup; >> > >> > @@ -137,7 +137,7 @@ static int kirin_drm_kms_init(struct drm_device *dev) >> > err_unbind_all: >> > component_unbind_all(dev->dev, dev); >> > err_dc_cleanup: >> > - dc_ops->cleanup(dev); >> > + dc_ops->cleanup(to_platform_device(dev->dev)); >> > err_mode_config_cleanup: >> > drm_mode_config_cleanup(dev); >> > devm_kfree(dev->dev, priv); >> > -- >> > 2.8.1 >> > >> > _______________________________________________ >> > dri-devel mailing list >> > dri-devel@lists.freedesktop.org >> > https://lists.freedesktop.org/mailman/listinfo/dri-devel > > -- > Daniel Vetter > Software Engineer, Intel Corporation > http://blog.ffwll.ch
On Thu, Aug 25, 2016 at 9:48 PM, Xinliang Liu <xinliang.liu@linaro.org> wrote: > On 17 August 2016 at 19:11, Daniel Vetter <daniel@ffwll.ch> wrote: >> On Wed, Aug 17, 2016 at 07:02:01PM +0800, Xinliang Liu wrote: >>> Hi, >>> >>> On 17 August 2016 at 18:17, Daniel Vetter <daniel.vetter@ffwll.ch> wrote: >>> > I just broke the build :( >>> > >>> > Note that the cleanup function is a bit confused: ade_data was never >>> > set as drvdata, and calling drm_crtc_cleanup directly is a bug - this >>> >>> Yes, this is a bug. Thanks for pointing out this. >>> >>> > is called indirectly through drm_mode_config_cleanup, which calls into >>> > crtc->destroy, which already has the call to drm_crtc_cleanup. Which >>> > means we can just nuke it. >>> > >>> > Aside: I have no idea why this driver depends upon ARM64. It doesn't >>> > build warning-free on 32bit, but otherwise it's perfectly fine. >>> >>> Because this driver is written for ARM64 SoCs. >> >> It makes compile-testing harder, the same reasons why depending upon > > What compile-testing we will have? I just try to understand the problem. it is useful for others who don't have the hw to be able to compile test core changes (hence kconfig COMPILE_TEST option) to at least catch the sorts of problems that compilers can catch.. BR, -R > Thanks, > -xinliang > >> specific arm platforms is really annoying (but lukily that problem seems >> to have stopped being one for the drm subsystem at least). Depending on >> CONFIG_OF and other stuff that's really needed is perfectly fine, but imo >> depending upon specific platforms and stuff really isn't. >> -Daniel >> >>> >>> Thanks, >>> -xinliang >>> >>> > >>> > Cc: Xinliang Liu <z.liuxinliang@hisilicon.com> >>> > Cc: Xinwei Kong <kong.kongxinwei@hisilicon.com> >>> > Cc: Chen Feng <puck.chen@hisilicon.com> >>> > Cc: Sean Paul <seanpaul@chromium.org> >>> > Fixes: d25bcfb8c2e1 ("drm/hisilicon: Don't set drm_device->platformdev") >>> > Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> >>> > --- >>> > drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c | 9 ++------- >>> > drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c | 4 ++-- >>> > 2 files changed, 4 insertions(+), 9 deletions(-) >>> > >>> > diff --git a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c >>> > index 91188f33b1d9..c64c82cb7e71 100644 >>> > --- a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c >>> > +++ b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c >>> > @@ -991,7 +991,7 @@ static int ade_dts_parse(struct platform_device *pdev, struct ade_hw_ctx *ctx) >>> > >>> > static int ade_drm_init(struct platform_device *pdev) >>> > { >>> > - struct drm_device *drm_dev = platform_get_drvdata(dev); >>> > + struct drm_device *dev = platform_get_drvdata(pdev); >>> > struct ade_data *ade; >>> > struct ade_hw_ctx *ctx; >>> > struct ade_crtc *acrtc; >>> > @@ -1052,14 +1052,9 @@ static int ade_drm_init(struct platform_device *pdev) >>> > >>> > static void ade_drm_cleanup(struct platform_device *pdev) >>> > { >>> > - struct drm_device *drm_dev = platform_get_drvdata(dev); >>> > - struct ade_data *ade = platform_get_drvdata(pdev); >>> > - struct drm_crtc *crtc = &ade->acrtc.base; >>> > - >>> > - drm_crtc_cleanup(crtc); >>> > } >>> > >>> > const struct kirin_dc_ops ade_dc_ops = { >>> > .init = ade_drm_init, >>> > .cleanup = ade_drm_cleanup >>> > -; >>> > +}; >>> > diff --git a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c >>> > index 6b0f9f6c16e1..b9b8c25da5e3 100644 >>> > --- a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c >>> > +++ b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c >>> > @@ -103,7 +103,7 @@ static int kirin_drm_kms_init(struct drm_device *dev) >>> > kirin_drm_mode_config_init(dev); >>> > >>> > /* display controller init */ >>> > - ret = dc_ops->init(to_platform_device(dev)); >>> > + ret = dc_ops->init(to_platform_device(dev->dev)); >>> > if (ret) >>> > goto err_mode_config_cleanup; >>> > >>> > @@ -137,7 +137,7 @@ static int kirin_drm_kms_init(struct drm_device *dev) >>> > err_unbind_all: >>> > component_unbind_all(dev->dev, dev); >>> > err_dc_cleanup: >>> > - dc_ops->cleanup(dev); >>> > + dc_ops->cleanup(to_platform_device(dev->dev)); >>> > err_mode_config_cleanup: >>> > drm_mode_config_cleanup(dev); >>> > devm_kfree(dev->dev, priv); >>> > -- >>> > 2.8.1 >>> > >>> > _______________________________________________ >>> > dri-devel mailing list >>> > dri-devel@lists.freedesktop.org >>> > https://lists.freedesktop.org/mailman/listinfo/dri-devel >> >> -- >> Daniel Vetter >> Software Engineer, Intel Corporation >> http://blog.ffwll.ch > _______________________________________________ > dri-devel mailing list > dri-devel@lists.freedesktop.org > https://lists.freedesktop.org/mailman/listinfo/dri-devel
Hi, On 26 August 2016 at 10:28, Rob Clark <robdclark@gmail.com> wrote: > On Thu, Aug 25, 2016 at 9:48 PM, Xinliang Liu <xinliang.liu@linaro.org> wrote: >> On 17 August 2016 at 19:11, Daniel Vetter <daniel@ffwll.ch> wrote: >>> On Wed, Aug 17, 2016 at 07:02:01PM +0800, Xinliang Liu wrote: >>>> Hi, >>>> >>>> On 17 August 2016 at 18:17, Daniel Vetter <daniel.vetter@ffwll.ch> wrote: >>>> > I just broke the build :( >>>> > >>>> > Note that the cleanup function is a bit confused: ade_data was never >>>> > set as drvdata, and calling drm_crtc_cleanup directly is a bug - this >>>> >>>> Yes, this is a bug. Thanks for pointing out this. >>>> >>>> > is called indirectly through drm_mode_config_cleanup, which calls into >>>> > crtc->destroy, which already has the call to drm_crtc_cleanup. Which >>>> > means we can just nuke it. >>>> > >>>> > Aside: I have no idea why this driver depends upon ARM64. It doesn't >>>> > build warning-free on 32bit, but otherwise it's perfectly fine. >>>> >>>> Because this driver is written for ARM64 SoCs. >>> >>> It makes compile-testing harder, the same reasons why depending upon >> >> What compile-testing we will have? I just try to understand the problem. > > it is useful for others who don't have the hw to be able to compile > test core changes (hence kconfig COMPILE_TEST option) to at least > catch the sorts of problems that compilers can catch.. Understand! Thanks Rob. -xinliang > > BR, > -R > > >> Thanks, >> -xinliang >> >>> specific arm platforms is really annoying (but lukily that problem seems >>> to have stopped being one for the drm subsystem at least). Depending on >>> CONFIG_OF and other stuff that's really needed is perfectly fine, but imo >>> depending upon specific platforms and stuff really isn't. >>> -Daniel >>> >>>> >>>> Thanks, >>>> -xinliang >>>> >>>> > >>>> > Cc: Xinliang Liu <z.liuxinliang@hisilicon.com> >>>> > Cc: Xinwei Kong <kong.kongxinwei@hisilicon.com> >>>> > Cc: Chen Feng <puck.chen@hisilicon.com> >>>> > Cc: Sean Paul <seanpaul@chromium.org> >>>> > Fixes: d25bcfb8c2e1 ("drm/hisilicon: Don't set drm_device->platformdev") >>>> > Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> >>>> > --- >>>> > drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c | 9 ++------- >>>> > drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c | 4 ++-- >>>> > 2 files changed, 4 insertions(+), 9 deletions(-) >>>> > >>>> > diff --git a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c >>>> > index 91188f33b1d9..c64c82cb7e71 100644 >>>> > --- a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c >>>> > +++ b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c >>>> > @@ -991,7 +991,7 @@ static int ade_dts_parse(struct platform_device *pdev, struct ade_hw_ctx *ctx) >>>> > >>>> > static int ade_drm_init(struct platform_device *pdev) >>>> > { >>>> > - struct drm_device *drm_dev = platform_get_drvdata(dev); >>>> > + struct drm_device *dev = platform_get_drvdata(pdev); >>>> > struct ade_data *ade; >>>> > struct ade_hw_ctx *ctx; >>>> > struct ade_crtc *acrtc; >>>> > @@ -1052,14 +1052,9 @@ static int ade_drm_init(struct platform_device *pdev) >>>> > >>>> > static void ade_drm_cleanup(struct platform_device *pdev) >>>> > { >>>> > - struct drm_device *drm_dev = platform_get_drvdata(dev); >>>> > - struct ade_data *ade = platform_get_drvdata(pdev); >>>> > - struct drm_crtc *crtc = &ade->acrtc.base; >>>> > - >>>> > - drm_crtc_cleanup(crtc); >>>> > } >>>> > >>>> > const struct kirin_dc_ops ade_dc_ops = { >>>> > .init = ade_drm_init, >>>> > .cleanup = ade_drm_cleanup >>>> > -; >>>> > +}; >>>> > diff --git a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c >>>> > index 6b0f9f6c16e1..b9b8c25da5e3 100644 >>>> > --- a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c >>>> > +++ b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c >>>> > @@ -103,7 +103,7 @@ static int kirin_drm_kms_init(struct drm_device *dev) >>>> > kirin_drm_mode_config_init(dev); >>>> > >>>> > /* display controller init */ >>>> > - ret = dc_ops->init(to_platform_device(dev)); >>>> > + ret = dc_ops->init(to_platform_device(dev->dev)); >>>> > if (ret) >>>> > goto err_mode_config_cleanup; >>>> > >>>> > @@ -137,7 +137,7 @@ static int kirin_drm_kms_init(struct drm_device *dev) >>>> > err_unbind_all: >>>> > component_unbind_all(dev->dev, dev); >>>> > err_dc_cleanup: >>>> > - dc_ops->cleanup(dev); >>>> > + dc_ops->cleanup(to_platform_device(dev->dev)); >>>> > err_mode_config_cleanup: >>>> > drm_mode_config_cleanup(dev); >>>> > devm_kfree(dev->dev, priv); >>>> > -- >>>> > 2.8.1 >>>> > >>>> > _______________________________________________ >>>> > dri-devel mailing list >>>> > dri-devel@lists.freedesktop.org >>>> > https://lists.freedesktop.org/mailman/listinfo/dri-devel >>> >>> -- >>> Daniel Vetter >>> Software Engineer, Intel Corporation >>> http://blog.ffwll.ch >> _______________________________________________ >> dri-devel mailing list >> dri-devel@lists.freedesktop.org >> https://lists.freedesktop.org/mailman/listinfo/dri-devel
diff --git a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c index 91188f33b1d9..c64c82cb7e71 100644 --- a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c +++ b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c @@ -991,7 +991,7 @@ static int ade_dts_parse(struct platform_device *pdev, struct ade_hw_ctx *ctx) static int ade_drm_init(struct platform_device *pdev) { - struct drm_device *drm_dev = platform_get_drvdata(dev); + struct drm_device *dev = platform_get_drvdata(pdev); struct ade_data *ade; struct ade_hw_ctx *ctx; struct ade_crtc *acrtc; @@ -1052,14 +1052,9 @@ static int ade_drm_init(struct platform_device *pdev) static void ade_drm_cleanup(struct platform_device *pdev) { - struct drm_device *drm_dev = platform_get_drvdata(dev); - struct ade_data *ade = platform_get_drvdata(pdev); - struct drm_crtc *crtc = &ade->acrtc.base; - - drm_crtc_cleanup(crtc); } const struct kirin_dc_ops ade_dc_ops = { .init = ade_drm_init, .cleanup = ade_drm_cleanup -; +}; diff --git a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c index 6b0f9f6c16e1..b9b8c25da5e3 100644 --- a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c +++ b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c @@ -103,7 +103,7 @@ static int kirin_drm_kms_init(struct drm_device *dev) kirin_drm_mode_config_init(dev); /* display controller init */ - ret = dc_ops->init(to_platform_device(dev)); + ret = dc_ops->init(to_platform_device(dev->dev)); if (ret) goto err_mode_config_cleanup; @@ -137,7 +137,7 @@ static int kirin_drm_kms_init(struct drm_device *dev) err_unbind_all: component_unbind_all(dev->dev, dev); err_dc_cleanup: - dc_ops->cleanup(dev); + dc_ops->cleanup(to_platform_device(dev->dev)); err_mode_config_cleanup: drm_mode_config_cleanup(dev); devm_kfree(dev->dev, priv);
I just broke the build :( Note that the cleanup function is a bit confused: ade_data was never set as drvdata, and calling drm_crtc_cleanup directly is a bug - this is called indirectly through drm_mode_config_cleanup, which calls into crtc->destroy, which already has the call to drm_crtc_cleanup. Which means we can just nuke it. Aside: I have no idea why this driver depends upon ARM64. It doesn't build warning-free on 32bit, but otherwise it's perfectly fine. Cc: Xinliang Liu <z.liuxinliang@hisilicon.com> Cc: Xinwei Kong <kong.kongxinwei@hisilicon.com> Cc: Chen Feng <puck.chen@hisilicon.com> Cc: Sean Paul <seanpaul@chromium.org> Fixes: d25bcfb8c2e1 ("drm/hisilicon: Don't set drm_device->platformdev") Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> --- drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c | 9 ++------- drivers/gpu/drm/hisilicon/kirin/kirin_drm_drv.c | 4 ++-- 2 files changed, 4 insertions(+), 9 deletions(-)