Hi, Dave.
This is resended e-mail for your advice.
I'm a newbie interested in linux kernel and qxl module.
Please check this patch and give me advice for me.
Also I'll be glad if there is any task that you bothered.
Thanks.
Sidong.
Replace deprecated function drm_modeset_lock/unlock_all with
helper function DRM_MODESET_LOCK_ALL_BEGIN/END.
Signed-off-by: Sidong Yang <[email protected]>
---
drivers/gpu/drm/qxl/qxl_display.c | 21 +++++++++++----------
1 file changed, 11 insertions(+), 10 deletions(-)
diff --git a/drivers/gpu/drm/qxl/qxl_display.c b/drivers/gpu/drm/qxl/qxl_display.c
index 1082cd5d2fd4..07e164cee868 100644
--- a/drivers/gpu/drm/qxl/qxl_display.c
+++ b/drivers/gpu/drm/qxl/qxl_display.c
@@ -162,7 +162,8 @@ static void qxl_update_offset_props(struct qxl_device *qdev)
void qxl_display_read_client_monitors_config(struct qxl_device *qdev)
{
struct drm_device *dev = &qdev->ddev;
- int status, retries;
+ struct drm_modeset_acquire_ctx ctx;
+ int status, retries, ret;
for (retries = 0; retries < 10; retries++) {
status = qxl_display_copy_rom_client_monitors_config(qdev);
@@ -183,9 +184,9 @@ void qxl_display_read_client_monitors_config(struct qxl_device *qdev)
return;
}
- drm_modeset_lock_all(dev);
+ DRM_MODESET_LOCK_ALL_BEGIN(dev, ctx, DRM_MODESET_ACQUIRE_INTERRUPTIBLE, ret);
qxl_update_offset_props(qdev);
- drm_modeset_unlock_all(dev);
+ DRM_MODESET_LOCK_ALL_END(ctx, ret);
if (!drm_helper_hpd_irq_event(dev)) {
/* notify that the monitor configuration changed, to
adjust at the arbitrary resolution */
@@ -403,18 +404,17 @@ static int qxl_framebuffer_surface_dirty(struct drm_framebuffer *fb,
struct qxl_device *qdev = to_qxl(fb->dev);
struct drm_clip_rect norect;
struct qxl_bo *qobj;
+ struct drm_modeset_acquire_ctx ctx;
bool is_primary;
- int inc = 1;
+ int inc = 1, ret;
- drm_modeset_lock_all(fb->dev);
+ DRM_MODESET_LOCK_ALL_BEGIN(fb->dev, ctx, DRM_MODESET_ACQUIRE_INTERRUPTIBLE, ret);
qobj = gem_to_qxl_bo(fb->obj[0]);
/* if we aren't primary surface ignore this */
is_primary = qobj->shadow ? qobj->shadow->is_primary : qobj->is_primary;
- if (!is_primary) {
- drm_modeset_unlock_all(fb->dev);
- return 0;
- }
+ if (!is_primary)
+ goto out_lock_end;
if (!num_clips) {
num_clips = 1;
@@ -430,7 +430,8 @@ static int qxl_framebuffer_surface_dirty(struct drm_framebuffer *fb,
qxl_draw_dirty_fb(qdev, fb, qobj, flags, color,
clips, num_clips, inc, 0);
- drm_modeset_unlock_all(fb->dev);
+out_lock_end:
+ DRM_MODESET_LOCK_ALL_END(ctx, ret);
return 0;
}
--
2.17.1
On Sun, May 24, 2020 at 07:42:25AM +1000, David Airlie wrote:
> On Sun, May 24, 2020 at 2:02 AM Sidong Yang <[email protected]> wrote:
> >
> > Hi, Dave.
> >
> > I'm a newbie kernel developer interested in qxl driver. And I want to participate in
> > contributing for QXL module.
> > I wrote some simple patch for refactoring task found in todos in gpu documentation.
> > I want to know it's okay to contribute and write some patch for qxl module.
> > If this patch is wrong, please give me some advice for me.
> > Or if you have some simple task for me, I'll be glad to do it.
> > Thanks.
>
> Hi Sidong,
>
> The best way to start is probably to email dri-devel list rather than
> just me, there are a few more people there who can help with
> onboarding and accepting patches.
>
> For QXL, Gerd Hoffmann ([email protected]) is also worth cc'ing as he
> is mostly maintaining it at the moment.
>
> Dave.
>
Thanks so much for advice Dave.
I'll add cc for qxl maintainer and dri-devel in next patch.
Sidong.
>
> >
> > Sincerely,
> > Sidong.
> >
> > Replace deprecated function drm_modeset_lock/unlock_all with
> > helper function DRM_MODESET_LOCK_ALL_BEGIN/END.
> >
> > Signed-off-by: Sidong Yang <[email protected]>
> > ---
> > drivers/gpu/drm/qxl/qxl_display.c | 21 +++++++++++----------
> > 1 file changed, 11 insertions(+), 10 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/qxl/qxl_display.c b/drivers/gpu/drm/qxl/qxl_display.c
> > index 1082cd5d2fd4..07e164cee868 100644
> > --- a/drivers/gpu/drm/qxl/qxl_display.c
> > +++ b/drivers/gpu/drm/qxl/qxl_display.c
> > @@ -162,7 +162,8 @@ static void qxl_update_offset_props(struct qxl_device *qdev)
> > void qxl_display_read_client_monitors_config(struct qxl_device *qdev)
> > {
> > struct drm_device *dev = &qdev->ddev;
> > - int status, retries;
> > + struct drm_modeset_acquire_ctx ctx;
> > + int status, retries, ret;
> >
> > for (retries = 0; retries < 10; retries++) {
> > status = qxl_display_copy_rom_client_monitors_config(qdev);
> > @@ -183,9 +184,9 @@ void qxl_display_read_client_monitors_config(struct qxl_device *qdev)
> > return;
> > }
> >
> > - drm_modeset_lock_all(dev);
> > + DRM_MODESET_LOCK_ALL_BEGIN(dev, ctx, DRM_MODESET_ACQUIRE_INTERRUPTIBLE, ret);
> > qxl_update_offset_props(qdev);
> > - drm_modeset_unlock_all(dev);
> > + DRM_MODESET_LOCK_ALL_END(ctx, ret);
> > if (!drm_helper_hpd_irq_event(dev)) {
> > /* notify that the monitor configuration changed, to
> > adjust at the arbitrary resolution */
> > @@ -403,18 +404,17 @@ static int qxl_framebuffer_surface_dirty(struct drm_framebuffer *fb,
> > struct qxl_device *qdev = to_qxl(fb->dev);
> > struct drm_clip_rect norect;
> > struct qxl_bo *qobj;
> > + struct drm_modeset_acquire_ctx ctx;
> > bool is_primary;
> > - int inc = 1;
> > + int inc = 1, ret;
> >
> > - drm_modeset_lock_all(fb->dev);
> > + DRM_MODESET_LOCK_ALL_BEGIN(fb->dev, ctx, DRM_MODESET_ACQUIRE_INTERRUPTIBLE, ret);
> >
> > qobj = gem_to_qxl_bo(fb->obj[0]);
> > /* if we aren't primary surface ignore this */
> > is_primary = qobj->shadow ? qobj->shadow->is_primary : qobj->is_primary;
> > - if (!is_primary) {
> > - drm_modeset_unlock_all(fb->dev);
> > - return 0;
> > - }
> > + if (!is_primary)
> > + goto out_lock_end;
> >
> > if (!num_clips) {
> > num_clips = 1;
> > @@ -430,7 +430,8 @@ static int qxl_framebuffer_surface_dirty(struct drm_framebuffer *fb,
> > qxl_draw_dirty_fb(qdev, fb, qobj, flags, color,
> > clips, num_clips, inc, 0);
> >
> > - drm_modeset_unlock_all(fb->dev);
> > +out_lock_end:
> > + DRM_MODESET_LOCK_ALL_END(ctx, ret);
> >
> > return 0;
> > }
> > --
> > 2.17.1
> >
>
On Sun, May 24, 2020 at 11:26:23AM +0900, Sidong Yang wrote:
> Hi, Dave.
>
> This is resended e-mail for your advice.
>
> I'm a newbie interested in linux kernel and qxl module.
> Please check this patch and give me advice for me.
> Also I'll be glad if there is any task that you bothered.
> Thanks.
>
> Sidong.
>
> Replace deprecated function drm_modeset_lock/unlock_all with
> helper function DRM_MODESET_LOCK_ALL_BEGIN/END.
>
> Signed-off-by: Sidong Yang <[email protected]>
Queued for drm-misc-next.
thanks,
Gerd