- 13 Sep, 2018 5 commits
-
-
Ville Syrjälä authored
Disable atomic on a per-device basis instead of for all devices. Made possible by the new device.driver_features thing. Cc: Alex Deucher <alexander.deucher@amd.com> Cc: "Christian König" <christian.koenig@amd.com> Cc: "David (ChunMing) Zhou" <David1.Zhou@amd.com> Cc: Harry Wentland <harry.wentland@amd.com> Cc: Michel Dänzer <michel@daenzer.net> Suggested-by:
Michel Dänzer <michel@daenzer.net> Signed-off-by:
Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20180913163147.27900-2-ville.syrjala@linux.intel.comReviewed-by:
Michel Dänzer <michel.daenzer@amd.com> Reviewed-by:
Alex Deucher <alexander.deucher@amd.com>
-
Ville Syrjälä authored
Currently we're clearing DRIVER_ATOMIC in driver.driver_features for older platforms. This will not work correctly should we ever have a system with and old and new GPU in it. While that is not possible currently let's make the code more correct and use the per-device driver_features instead. Cc: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by:
Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by:
Daniel Vetter <daniel.vetter@ffwll.ch>
-
Ville Syrjälä authored
We wish to control certain driver_features flags on a per-device basis while still sharing a single drm_driver instance across all the devices. To that end introduce device.driver_features. By default it will be set to ~0 to not impose any limits beyond driver.driver_features. Drivers can then clear specific flags in the per-device bitmask to limit the capabilities of the device. An alternative approach would be to copy the driver_features from the driver into the device in drm_dev_init(), however that would require verifying that no driver is currently changing driver.driver_features after drm_dev_init(). Hence the ~0 apporach was easier. Ideally we'd also make drm_driver const but there is plenty of code left that wants to mutate it (eg. various vfunc assignments). We'll need to fix all that up before we can make it const. And while at it fix up the type of the feature flag passed to drm_core_check_feature(). v2: Streamline the && vs. & (Chris) s/int/u32/ in drm_core_check_feature() args Cc: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: -
Peter Rosin authored
Since commit 4a878c03 ("drm: bridge: Detach bridge from encoder at encoder cleanup time"), it is generally no longer correct to detach bridges from encoders manually. Document that. Signed-off-by:
Peter Rosin <peda@axentia.se> Signed-off-by:
Andrzej Hajda <a.hajda@samsung.com> Link: https://patchwork.freedesktop.org/patch/msgid/20180806061910.29914-3-peda@axentia.se
-
Marc Zyngier authored
The Analogix DP bridge driver is pretty verbose, and outputs things like [ 619.414067] rockchip-dp ff970000.edp: Link Training Clock Recovery success [ 619.429233] rockchip-dp ff970000.edp: Link Training success! each time the display gets unblanked. While it is good to know that the device is behaving correctly, users already know that because they can see some video output. Let's keep these messages for cases where we need to actually debug the driver (we have dynamic debug to enable them at runtime if need be), and let's keep the kernel quiet otherwise. Signed-off-by:
Marc Zyngier <marc.zyngier@arm.com> Reviewed-by:
Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by:
-
- 12 Sep, 2018 14 commits
-
-
Maxime Ripard authored
Having DRM_SUN4I built-in but DRM_SUN8I_MIXER as a loadable module results in a link error, as we try to access a symbol from the sun8i_tcon_top.ko module: ERROR: "sun8i_tcon_top_de_config" [drivers/gpu/drm/sun4i/sun4i-tcon.ko] undefined! ERROR: "sun8i_tcon_top_set_hdmi_src" [drivers/gpu/drm/sun4i/sun4i-tcon.ko] undefined! ERROR: "sun8i_tcon_top_of_table" [drivers/gpu/drm/sun4i/sun4i-tcon.ko] undefined! This solves the problem by adding a silent symbol for the tcon_top module, building it as a separate module in exactly the cases that we need it, but in a way that it is reachable by the other modules. Fixes: cf77d79b ("drm/sun4i: tcon: Add another way for matching mixers with tcon") Fixes: 0305189a ("drm/sun4i: tcon: Add support for R40 TCON") Reviewed-by:
Jon Hunter <jonathanh@nvidia.com> Tested-by:
Matt Hart <matthew.hart@linaro.org> Signed-off-by:
Maxime Ripard <maxime.ripard@bootlin.com> Link: https://patchwork.freedesktop.org/patch/msgid/20180911113325.11024-1-maxime.ripard@bootlin.com
-
Peter Wu authored
Lots of code can be removed by relying on fb-helper: - "struct drm_framebuffer" moves to fb_helper.fb. - "struct drm_gem_object" moves to fb_helper.obj[0]. - "struct qxl_device" can be inferred as drm_fb_helper is embedded. - qxl_user_framebuffer_create -> drm_gem_fb_create. - qxl_user_framebuffer_destroy -> drm_gem_fb_destroy. - qxl_fbdev_destroy -> drm_fb_helper_fbdev_teardown + vfree(shadow). Remove unused code: - qxl_fbdev_qobj_is_fb, qxl_fbdev_set_suspend. - Unused fields of qxl_fbdev: delayed_ops, delayed_ops_lock, size. Misc notes: - The dirty callback is preserved as it is necessary to trigger update commands in the hw (the screen stays black otherwise). - No idea when .create_handle in drm_framebuffer_funcs is used, but use the same drm_gem_fb_create_handle to match drm_gem_fb_funcs. - I don't know why qxl_fb_find_or_create_single used to check for an existing framebuffer and removed that check to match other drivers. - Use of drm_fb_helper_fbdev_teardown also requires "info->fbdefio" to be dynamically allocated. Replace the existing defio config by drm_fb_helper_defio_init to accomodate this. Testing results: startx with fbdev, modesetting and qxl all seems to work. Tested also with CONFIG_DRM_FBDEV_EMULATION=n, fbdev obviously fails but others are fine. QEMU -spice and QEMU -spice with vdagent and multiple (resized) displays (via remote-viewer) also works. unbind vtconsole and rmmod has *not* regressed (i.e. it still trips on a use-after-free in qxl_check_idle via qxl_ttm_fini). Ideally setup/teardown is replaced by drm_fbdev_generic_setup as that would result in further code reduction, improve error handling (like not leaking shadow memory), but unfortunately QXL has no implementation for qxl_gem_prime_vmap. Signed-off-by:
Peter Wu <peter@lekensteyn.nl> Link: http://patchwork.freedesktop.org/patch/msgid/20180910132156.23201-1-peter@lekensteyn.nlSigned-off-by:
Gerd Hoffmann <kraxel@redhat.com>
-
Gerd Hoffmann authored
Reported-by:
-
Gerd Hoffmann authored
Reported-by:
-
Gerd Hoffmann authored
Reported-by:
-
Gerd Hoffmann authored
Reported-by:
-
Gerd Hoffmann authored
Reported-by:
-
Gerd Hoffmann authored
Create variable for the list length limit. Serves as documentation, also allows to make it a module parameter if needed. Also add a total size limit. Signed-off-by:
-
Gerd Hoffmann authored
udmabuf builds without it, but if userspace can not create memfd handles in the first place it is rather pointless to include it, except for test builds. Signed-off-by:
-
Gerd Hoffmann authored
Reported-by:
-
Gerd Hoffmann authored
Reported-by:
-
Gerd Hoffmann authored
Reported-by:
-
Gerd Hoffmann authored
Reported-by:
-
Gerd Hoffmann authored
Reported-by:
-
- 11 Sep, 2018 10 commits
-
-
Haneen Mohammed authored
Add an initial kerneldoc entry for vkms with a todo list. Signed-off-by:
Haneen Mohammed <hamohammed.sa@gmail.com> [danvet: Keep the todo.rst entry to point at the vkms docs instead.] Signed-off-by:
-
Haneen Mohammed authored
Cursor support is not complete yet. Add module option 'enable_cursor' to enable/disable cursor support which is used for testing currently. Signed-off-by:
-
Haneen Mohammed authored
This patch compute CRC for output frame with cursor and primary plane. Blend cursor with primary plane and compute CRC on the resulted frame. This currently passes cursor-size-change, and cursor-64x64-[onscreen, offscreen, sliding, random, dpms, rapid-movement] from igt kms_cursor_crc tests. Signed-off-by:
-
Haneen Mohammed authored
Add cursor plane support and update vkms_plane_atomic_check to enable positioning cursor plane. Signed-off-by:
-
Daniel Vetter authored
For buffer sharing, use dma-buf instead. We can't set smem_start to 0 unconditionally since that's used by the fbdev mmap default implementation. And we have plenty of userspace which would like to keep that working. This might break legit userspace - if it does we need to look at a case-by-cases basis how to handle that. Worst case I expect overrides for only specific drivers, since anything remotely modern should be using dma-buf/prime now (which is about 7 years old now for DRM drivers). This issue was uncovered because Noralf's rework to implement a generic fb_probe also implements it's own fb_mmap callback. Which means smem_start didn't have to be set anymore, which blew up some blob in userspace rather badly. Acked-by:
Sean Paul <seanpaul@chromium.org> Cc: Gustavo Padovan <gustavo@padovan.org> Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Cc: Sean Paul <sean@poorly.run> Cc: David Airlie <airlied@linux.ie> Cc: John Stultz <john.stultz@linaro.org> Cc: Noralf Trønnes <noralf@tronnes.org> Signed-off-by:
Daniel Vetter <daniel.vetter@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20180822085405.10787-4-daniel.vetter@ffwll.ch
-
Daniel Vetter authored
DRM drivers really, really, really don't want random userspace to share buffer behind it's back, bypassing the dma-buf buffer sharing machanism. For that reason we've ruthlessly rejected any IOCTL exposing the physical address of any graphics buffer. Unfortunately fbdev comes with that built-in. We could just set smem_start to 0, but that means we'd have to hand-roll our own fb_mmap implementation. For good reasons many drivers do that, but smem_start/length is still super convenient. Hence instead just stop the leak in the ioctl, to keep fb mmap working as-is. A second patch will set this flag for all drm drivers. Acked-by:
Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com> Cc: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com> Cc: Kees Cook <keescook@chromium.org> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: linux-fbdev@vger.kernel.org Signed-off-by:
-
Daniel Vetter authored
It was only used by the panic support in fbcon, which is now gone. Remove this now dead code too. Acked-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Kees Cook <keescook@chromium.org> Cc: Joe Perches <joe@perches.com> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: Meng Xu <mengxu.gatech@gmail.com> Cc: Nicolas Pitre <nicolas.pitre@linaro.org> Cc: Thomas Meyer <thomas@m3y3r.de> Cc: Mike Frysinger <vapier@chromium.org> Cc: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com> Cc: Hans de Goede <hdegoede@redhat.com> Cc: Thierry Reding <treding@nvidia.com> Cc: David Lechner <david@lechnology.com> Cc: Philippe Ombredanne <pombredanne@nexb.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Kate Stewart <kstewart@linuxfoundation.org> Signed-off-by:
-
Daniel Vetter authored
This was only added for the drm's fbdev emulation support, so that it would try harder to show the Oops. Unfortunately this never really worked reliably, and in practice ended up pushing the real Oops off the screen due to plentyfull locking, sleep-while-atomic and other issues. So we removed all that support from the fbdev emulation a while back. Aside: We've also removed the kgdb support, for similar reasons. Since it's such a small patch I figured I don't split this up into the usual 3-phase removal. Acked-by:
-
Alexandru Gheorghe authored
DRM_MODE_REFLECT_X and DRM_MODE_REFLECT_Y meaning seems a bit unclear to me, so try to clarify that with a bit of ascii graphics. Changes since v1: - Move the ascii graphics in the kerneldoc where all plane properties are already documented and make sure it's properly rendered, suggestested by Daniel Vetter. Signed-off-by:Alexandru Gheorghe <alexandru-cosmin.gheorghe@arm.com> Reviewed-by:
-
YueHaibing authored
Fixes gcc '-Wunused-but-set-variable' warning: drivers/gpu/drm/drm_plane.c: In function 'drm_mode_getplane_res': drivers/gpu/drm/drm_plane.c:475:26: warning: variable 'config' set but not used [-Wunused-but-set-variable] Signed-off-by:
YueHaibing <yuehaibing@huawei.com> Signed-off-by:
-
- 10 Sep, 2018 7 commits
-
-
Marc Zyngier authored
Leaving the DRM driver enabled on reboot or kexec has the annoying effect of leaving the display generating transactions whilst the IOMMU has been shut down. In turn, the IOMMU driver (which shares its interrupt line with the VOP) starts warning either on shutdown or when entering the secondary kernel in the kexec case (nothing is expected on that front). A cheap way of ensuring that things are nicely shut down is to register a shutdown callback in the platform driver. Signed-off-by:
Vicente Bergas <vicencb@gmail.com> Signed-off-by:
Heiko Stuebner <heiko@sntech.de> Link: https://patchwork.freedesktop.org/patch/msgid/20180805124807.18169-1-marc.zyngier@arm.com Cc: stable@vger.kernel.org
-
Peter Wu authored
Clarify the relation between drm_fb_helper_fbdev_setup/teardown. Clarify requirements for the new generic fbdev emulation API and log some more details in case the driver does something wrong. Fix related typos. Cc: Noralf Trønnes <noralf@tronnes.org> Signed-off-by:
-
Peter Wu authored
The drm_get_pci_dev API is deprecated, replace it by drm_dev_register. Signed-off-by:
-
Peter Wu authored
Currently unloading bochs_drm (after unbinding the vtconsole) results in a warning about a leaked connector: [drm:drm_mode_config_cleanup] *ERROR* connector Virtual-3 leaked! While investigating a potential fix I noticed that a lot of open-coded functionality is already implemented elsewhere, so start converting it: bochs_fbdev_init -> drm_fb_helper_fbdev_setup: trivial (similar impl). bochs_fbdev_fini -> drm_fb_helper_fbdev_teardown: requires unembedding "struct drm_framebuffer" from "struct bochs_framebuffer". Unembedding drm_framebuffer is made easy using drm_gem_fbdev_fb_create which can replace bochs_fbdev_destroy and custom routines in bochs_mm.c. For this to work, the GEM object is moved into "drm_framebuffer". After that, "bochs_framebuffer" is no longer needed and therefore removed. Remove the unused "size" and "initialized" fields from fb, the latter is not necessary as drm_fb_helper_fbdev_teardown can be called even if bochsfb_create fails. This theory was tested by returning early and late (just before drm_gem_fbdev_fb_create). Both scenarios fail gracefully although the latter seems to leak the object from bochsfb_create_object (not a regression). Guess on the reason for the encoder leak: drm_framebuffer_cleanup was previously used, but did not destroy much. drm_fb_helper_fbdev_teardown is now used and calls drm_framebuffer_remove which does a bit more work. Tested with 'echo 0 > /sys/class/vtconsole/vtcon1/bind; rmmod bochs_drm' and also with Xorg + fbdev (startx -> xterm). The latter triggered a warning in ttm_bo_vm_open that existed before, see https://lkml.kernel.org/r/1464000533-13140-4-git-send-email-mstaudt@suse.deAcked-by: -
Peter Wu authored
The "initialized" member is going away. suspend/resume still works (even if bochsfb_create is forced to fail). Signed-off-by:
-
Gerd Hoffmann authored
Drivers must set the quirk_addfb_prefer_host_byte_order quirk to make the drm_mode_addfb() compat code work correctly on bigendian machines. If they don't they interpret pixel_format values incorrectly for bug compatibility, which in turn implies the ADDFB2 ioctl does not work correctly then. So block it to make userspace fallback to ADDFB. Signed-off-by:
-
YueHaibing authored
Fixes gcc '-Wunused-but-set-variable' warning: drivers/gpu/drm/virtio/virtgpu_display.c: In function 'virtio_gpu_framebuffer_init': drivers/gpu/drm/virtio/virtgpu_display.c:78:28: warning: variable 'bo' set but not used [-Wunused-but-set-variable] struct virtio_gpu_object *bo; ^ Signed-off-by:
-
- 09 Sep, 2018 4 commits
-
-
Daniel Vetter authored
This leaves all the commit/check and state handling in drm_atomic.c, while pulling all the uapi glue and the huge ioctl itself into a seprate file. This seems to almost perfectly split the rather big drm_atomic.c file into 2 equal sizes. Also adjust the kerneldoc and type a very terse overview text. v2: Rebase. v3: Fix tiny typo. v4: - Fixup armada, newly converted atomic driver hooray! - Fixup msm/dpu1, newly added too. Signed-off-by:
Rodrigo Vivi <rodrigo.vivi@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20180905135711.28370-7-daniel.vetter@ffwll.ch
-
Daniel Vetter authored
- drmP.h is now fully split up. - vkms is happening (and will gain its own todo and docs under a new vkms.rst file real soon) - legacy cruft is completely hidden now, drm_vblank.c is split out from drm_irq.c now. I've decided to drop the task to split out drm_legacy.ko, partially because Dave already rejected a patch to hide the old dri1 drivers better. Current state feels good enough to me. - best_encoder atomic cleanup is done (it's now the default, not even exported anymore) - bunch of smaller things v2: - Explain why the drm_legacy.ko task is dropped (Emil). - typos (Sam). v3: Fix typo (Ilia) Reviewed-by:
Emil Velikov <emil.velikov@collabora.com> Reviewed-by:
-
Daniel Vetter authored
Remove the kerneldoc and EXPORT_SYMBOL which aren't used and really shouldn't ever be used by drivers directly. Unfortunately this means we need to move the set_writeback_fb function around to avoid a forward decl. Acked-by:
-
Daniel Vetter authored
This is starting to become easy! v2: Compiles now, with drm/drm_util.h extracted. v3: Fix up commit message (Sam Ravnborg) Reviewed-by:
-
