- 20 Jul, 2023 3 commits
-
-
Simon Ser authored
Introduce a new DRM_IOCTL_SYNCOBJ_EVENTFD IOCTL which signals an eventfd from a syncobj. This is useful for Wayland compositors to handle wait-before-submit. Wayland clients can send a timeline point to the compositor before the point has materialized yet, then compositors can wait for the point to materialize via this new IOCTL. The existing DRM_IOCTL_SYNCOBJ_TIMELINE_WAIT IOCTL is not suitable because it blocks. Compositors want to integrate the wait with their poll(2)-based event loop. Requirements for new uAPI: - User-space patch: https://gitlab.freedesktop.org/wlroots/wlroots/-/merge_requests/4262 - IGT: https://lists.freedesktop.org/archives/igt-dev/2023-July/057893.html v2: - Wait for fence when flags is zero - Improve documentation (Pekka) - Rename IOCTL (Christian) - Fix typo in drm_syncobj_add_eventfd() (Christian) v3: - Link user-space + IGT patches - Add reference from overview docs v4: fix IOCTL number conflict with GETFB2 (Nicholas Choi, Vitaly Prosyak) Signed-off-by:
Simon Ser <contact@emersion.fr> Reviewed-by:
Christian König <christian.koenig@amd.com> Acked-by:
Pekka Paalanen <pekka.paalanen@collabora.com> Cc: Jason Ekstrand <jason@jlekstrand.net> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl> Cc: Daniel Stone <daniel@fooishbar.org> Cc: James Jones <jajones@nvidia.com> Cc: Austin Shafer <ashafer@nvidia.com> Cc: Vitaly Prosyak <vprosyak@amd.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230714111257.11940-1-contact@emersion.fr
-
Danilo Krummrich authored
This commit adds a function to dump a DRM GPU VA space and a macro for drivers to register the struct drm_info_list 'gpuvas' entry. Most likely, most drivers might maintain one DRM GPU VA space per struct drm_file, but there might also be drivers not having a fixed relation between DRM GPU VA spaces and a DRM core infrastructure, hence we need the indirection via the driver iterating it's maintained DRM GPU VA spaces. Reviewed-by:
Boris Brezillon <boris.brezillon@collabora.com> Signed-off-by:
Danilo Krummrich <dakr@redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230720001443.2380-3-dakr@redhat.com
-
Danilo Krummrich authored
Add infrastructure to keep track of GPU virtual address (VA) mappings with a decicated VA space manager implementation. New UAPIs, motivated by Vulkan sparse memory bindings graphics drivers start implementing, allow userspace applications to request multiple and arbitrary GPU VA mappings of buffer objects. The DRM GPU VA manager is intended to serve the following purposes in this context. 1) Provide infrastructure to track GPU VA allocations and mappings, using an interval tree (RB-tree). 2) Generically connect GPU VA mappings to their backing buffers, in particular DRM GEM objects. 3) Provide a common implementation to perform more complex mapping operations on the GPU VA space. In particular splitting and merging of GPU VA mappings, e.g. for intersecting mapping requests or partial unmap requests. Acked-by:
Thomas Hellström <thomas.hellstrom@linux.intel.com> Acked-by:
Matthew Brost <matthew.brost@intel.com> Reviewed-by:
Donald Robson <donald.robson@imgtec.com> Suggested-by:
Dave Airlie <airlied@redhat.com> Signed-off-by:
-
- 17 Jul, 2023 4 commits
-
-
Chen-Yu Tsai authored
The DRM DP code has macros for the DP HDCP capabilities. Use them in the anx7625 driver instead of raw numbers. Fixes: cd1637c7 ("drm/bridge: anx7625: add HDCP support") Suggested-by:
Nícolas F. R. A. Prado <nfraprado@collabora.com> Signed-off-by:
Chen-Yu Tsai <wenst@chromium.org> Reviewed-by:
Robert Foss <rfoss@kernel.org> Signed-off-by:
-
Chen-Yu Tsai authored
The DRM DP code has macros for the DP power sequencing commands. Use them in the anx7625 driver instead of raw numbers. Fixes: 548b512e ("drm/bridge: anx7625: send DPCD command to downstream") Fixes: 27f26359 ("drm/bridge: anx7625: Set downstream sink into normal status") Signed-off-by:
-
Simon Ser authored
This makes it clearer that the values cannot be changed because they are ABI. Signed-off-by:
James Zhu <James.Zhu@amd.com> Reviewed-by:
Thomas Zimmermann <tzimmermann@suse.de> Cc: Christian König <christian.koenig@amd.com> Cc: Marek Olšák <marek.olsak@amd.com> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Link: https://patchwork.freedesktop.org/patch/msgid/20230714104557.518457-2-contact@emersion.fr
-
Simon Ser authored
This makes it easier to figure out what the "type" variable can be set to when reading the implementation of these functions. Signed-off-by:
-
- 14 Jul, 2023 2 commits
-
-
Paul Cercueil authored
Register a backlight device to be able to switch between all the gamma levels. v2: Remove .get_brightness() callback, use bl_get_data() and backlight_get_brightness() Signed-off-by:Paul Cercueil <paul@crapouillou.net> Reviewed-by:
Sam Ravnborg <sam@ravnborg.org> Link: https://patchwork.freedesktop.org/patch/msgid/20230708084027.18352-3-paul@crapouillou.net
-
Paul Cercueil authored
I have no idea what the prior magic values mean, and I have no idea what my replacement (extracted from [1]) magic values mean. What I do know, is that these new values result in a much better picture, where the blacks are really black (as you would expect on an AMOLED display) instead of grey-ish. [1]: https://github.com/dorimanx/Dorimanx-SG2-I9100-Kernel/blob/master-jelly-bean/arch/arm/mach-exynos/u1-panel.h v2: Remove spurious new line Signed-off-by:
Neil Armstrong <neil.armstrong@linaro.org> Reviewed-by:
-
- 13 Jul, 2023 1 commit
-
-
Javier Martinez Canillas authored
The commit e254b584 ("drm/ssd130x: Remove hardcoded bits-per-pixel in ssd130x_buf_alloc()") used a pixel format info rather than a hardcoded bpp to calculate the size of the buffer allocated to store the native pixels. But it wrongly used the DRM_FORMAT_C1 fourcc pixel format. That is for color-indexed frame buffer formats, while the ssd103x controllers don't support different single-channel colors nor a Color Lookup Table (CLUT). So the correct pixel format to use in this case is DRM_FORMAT_R1 instead. Since both formats use a eight pixels/byte, there is no functional change in practice by this patch. Still, the correct pixel format should be used. Suggested-by:
Geert Uytterhoeven <geert@linux-m68k.org> Signed-off-by:
Javier Martinez Canillas <javierm@redhat.com> Reviewed-by:
-
- 12 Jul, 2023 7 commits
-
-
Sui Jingfeng authored
Because smatch warnings: drivers/gpu/drm/loongson/lsdc_plane.c:199 lsdc_cursor_plane_atomic_async_check() warn: variable dereferenced before check 'state' (see line 180) vim +/state +199 drivers/gpu/drm/loongson/lsdc_plane.c 174 static int lsdc_cursor_plane_atomic_async_check(struct drm_plane *plane, 175 struct drm_atomic_state *state) 176 { 177 struct drm_plane_state *new_state; 178 struct drm_crtc_state *crtc_state; 179 180 new_state = drm_atomic_get_new_plane_state(state, plane); ^^^^^ state is dereferenced inside this function 181 182 if (!plane->state || !plane->state->fb) { 183 drm_dbg(plane->dev, "%s: state is NULL\n", plane->name); 184 return -EINVAL; 185 } 186 187 if (new_state->crtc_w != new_state->crtc_h) { 188 drm_dbg(plane->dev, "unsupported cursor size: %ux%u\n", 189 new_state->crtc_w, new_state->crtc_h); 190 return -EINVAL; 191 } 192 193 if (new_state->crtc_w != 64 && new_state->crtc_w != 32) { 194 drm_dbg(plane->dev, "unsupported cursor size: %ux%u\n", 195 new_state->crtc_w, new_state->crtc_h); 196 return -EINVAL; 197 } 198 199 if (state) { ^^^^^ Checked too late! Reported-by:Dan Carpenter <dan.carpenter@linaro.org> Closes: https://lore.kernel.org/r/202307100423.rV7D05Uq-lkp@intel.com/Signed-off-by:
Sui Jingfeng <suijingfeng@loongson.cn> Acked-by:
Sui Jingfeng <sui.jingfeng@linux.dev> Link: https://patchwork.freedesktop.org/patch/msgid/20230710102411.257970-1-suijingfeng@loongson.cn
-
Christian König authored
Use the new component here as well and remove the old handling. v2: drop dupplicate handling v3: fix memory leak pointed out by Tatsuyuki Signed-off-by:
Alex Deucher <alexander.deucher@amd.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230711133122.3710-7-christian.koenig@amd.com
-
Christian König authored
Start using the new component here as well. Signed-off-by:
-
Christian König authored
Start using the new component here as well. v2: ignore duplicates to allow per VM BO mappings Signed-off-by:
-
Christian König authored
Avoids quite a bit of logic and kmalloc overhead. v2: fix multiple problems pointed out by Felix v3: two more nit picks from Felix fixed Signed-off-by:
Felix Kuehling <Felix.Kuehling@amd.com> Acked-by:
-
Christian König authored
Exercise at least all driver facing functions of this new component. v2: add array test as well v3: some kunit cleanups v4: more tests and cleanups Signed-off-by:
-
Christian König authored
This adds the infrastructure for an execution context for GEM buffers which is similar to the existing TTMs execbuf util and intended to replace it in the long term. The basic functionality is that we abstracts the necessary loop to lock many different GEM buffers with automated deadlock and duplicate handling. v2: drop xarray and use dynamic resized array instead, the locking overhead is unnecessary and measurable. v3: drop duplicate tracking, radeon is really the only one needing that. v4: fixes issues pointed out by Danilo, some typos in comments and a helper for lock arrays of GEM objects. v5: some suggestions by Boris Brezillon, especially just use one retry macro, drop loop in prepare_array, use flags instead of bool v6: minor changes suggested by Thomas, Boris and Danilo v7: minor typos pointed out by checkpatch.pl fixed Signed-off-by:
-
- 11 Jul, 2023 1 commit
-
-
Thomas Zimmermann authored
Include <linux/screen_info.h> to get the global screen_info state. Fixes the following errors: >> drivers/video/fbdev/hyperv_fb.c:1033:10: error: use of undeclared identifier 'screen_info' 1033 | base = screen_info.lfb_base; | ^ drivers/video/fbdev/hyperv_fb.c:1034:10: error: use of undeclared identifier 'screen_info' 1034 | size = screen_info.lfb_size; | ^ >> drivers/video/fbdev/hyperv_fb.c:1080:3: error: must use 'struct' tag to refer to type 'screen_info' 1080 | screen_info.lfb_size = 0; | ^ | struct >> drivers/video/fbdev/hyperv_fb.c:1080:14: error: expected identifier or '(' 1080 | screen_info.lfb_size = 0; | ^ drivers/video/fbdev/hyperv_fb.c:1081:3: error: must use 'struct' tag to refer to type 'screen_info' 1081 | screen_info.lfb_base = 0; | ^ | struct drivers/video/fbdev/hyperv_fb.c:1081:14: error: expected identifier or '(' 1081 | screen_info.lfb_base = 0; | ^ drivers/video/fbdev/hyperv_fb.c:1082:3: error: must use 'struct' tag to refer to type 'screen_info' 1082 | screen_info.orig_video_isVGA = 0; | ^ | struct drivers/video/fbdev/hyperv_fb.c:1082:14: error: expected identifier or '(' 1082 | screen_info.orig_video_isVGA = 0; | ^ 8 errors generated. Reported-by:kernel test robot <lkp@intel.com> Closes: https://lore.kernel.org/oe-kbuild-all/202307101042.rqehuauj-lkp@intel.com/ Fixes: 8b0d1354 ("efi: Do not include <linux/screen_info.h> from EFI header") Signed-off-by:
-
- 10 Jul, 2023 13 commits
-
-
Javier Martinez Canillas authored
Otherwise if CONFIG_DRM is disabled, menuconfig will show an empty menu. Signed-off-by:
Liviu Dudau <liviu.dudau@arm.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230703230534.997525-4-javierm@redhat.com
-
Rajneesh Bhardwaj authored
Delay release TTM BOs when the kernel default setting is init_on_free. This offloads the overhead of clearing the system memory to the work item and potentially a different CPU. This could be very beneficial when the application does a lot of malloc/free style allocations of system memory. Signed-off-by:
Rajneesh Bhardwaj <rajneesh.bhardwaj@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com>. Link: https://patchwork.freedesktop.org/patch/msgid/20230708011355.853-1-rajneesh.bhardwaj@amd.comSigned-off-by:
-
Nikhil Devshatwar authored
When removing the tidss driver, there is a warning reported by kernel about an unhandled interrupt for mhdp driver. [ 43.238895] irq 31: nobody cared (try booting with the "irqpoll" option) ... [snipped backtrace] [ 43.330735] handlers: [ 43.333020] [<000000005367c4f9>] irq_default_primary_handler threaded [<000000007e02b601>] cdns_mhdp_irq_handler [cdns_mhdp8546] [ 43.344607] Disabling IRQ #31 This happens because as part of cdns_mhdp_bridge_hpd_disable, driver tries to disable the interrupts. While disabling the SW_EVENT interrupts, it accidentally enables the MBOX interrupts, which are not handled by the driver. Fix this with a read-modify-write to update only required bits. Use the enable / disable function as required in other places. Signed-off-by:
Nikhil Devshatwar <nikhil.nd@ti.com> Reviewed-by:
Swapnil Jakhade <sjakhade@cadence.com> Reviewed-by:
Tomi Valkeinen <tomi.valkeinen@ti.com> Signed-off-by:
Aradhya Bhatia <a-bhatia1@ti.com> Signed-off-by:
-
Aradhya Bhatia authored
With the new encoder/bridge chain model, the display controller driver is required to create a drm_connector entity instead of asking the bridge to do so during drm_bridge_attach. Moreover, the controller driver should create a drm_bridge entity to negotiate bus formats and a 'simple' drm_encoder entity to expose it to userspace. Update the encoder/bridge initialization sequence in tidss as per the new model. Signed-off-by:
Tomi Valkeinen <tomi.valkeinen@ideasonboard.com> Signed-off-by:
-
Aradhya Bhatia authored
input_bus_flags are specified in drm_bridge_timings (legacy) as well as drm_bridge_state->input_bus_cfg.flags The flags from the timings will be deprecated. Bridges are supposed to validate and set the bridge state flags from atomic_check. Implement atomic_check hook for the same. Signed-off-by:
-
Aradhya Bhatia authored
With new connector model, sii902x will not create the connector, when DRM_BRIDGE_ATTACH_NO_CONNECTOR is set and SoC driver will rely on format negotiation to setup the encoder format. Support format negotiations hooks in the drm_bridge_funcs. Use helper functions for state management. Input format is selected to MEDIA_BUS_FMT_RGB888_1X24 as default, as is the case with older model. Signed-off-by:
-
Nikhil Devshatwar authored
input_bus_flags are specified in drm_bridge_timings (legacy) as well as drm_bridge_state->input_bus_cfg.flags The flags from the timings will be deprecated. Bridges are supposed to validate and set the bridge state flags from atomic_check. Signed-off-by:
-
Nikhil Devshatwar authored
With new connector model, mhdp bridge will not create the connector and SoC driver will rely on format negotiation to setup the encoder format. Support minimal format negotiations hooks in the drm_bridge_funcs. Complete format negotiation can be added based on EDID data. This patch adds the minimal required support to avoid failure after moving to new connector model. Signed-off-by:
-
Nikhil Devshatwar authored
input_bus_flags are specified in drm_bridge_timings (legacy) as well as drm_bridge_state->input_bus_cfg.flags The flags from the timings will be deprecated. Bridges are supposed to validate and set the bridge state flags from atomic_check. Implement atomic_check hook for the same. Signed-off-by:
-
Nikhil Devshatwar authored
With new connector model, tfp410 will not create the connector and SoC driver will rely on format negotiation to setup the encoder format. Support format negotiations hooks in the drm_bridge_funcs. Use helper functions for state management. Input format is the one selected by the bridge from DT properties. Signed-off-by:
-
Christophe JAILLET authored
The devm_clk_get_enabled() helper: - calls devm_clk_get() - calls clk_prepare_enable() and registers what is needed in order to call clk_disable_unprepare() when needed, as a managed resource. This simplifies the code and avoids the need of a dedicated function used with devm_add_action_or_reset(). Signed-off-by:Christophe JAILLET <christophe.jaillet@wanadoo.fr> Reviewed-by:
Andrzej Hajda <andrzej.hajda@intel.com> Signed-off-by:
-
Simon Ser authored
This adds more information to the hotplug uevent and lets user-space know that it's about a particular connector only. Signed-off-by:
Lyude Paul <lyude@redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20230620181547.272476-1-contact@emersion.fr
-
Simon Ser authored
The hardware needs a FB which is packed. Add checks to make sure this is the case. While at it, add debug logs for the existing checks. This allows user-space to more easily figure out why a configuration is rejected. v2: - Use drm_format_info instead of hardcoding bytes-per-pixel (Ilia) - Remove unnecessary size check (Ilia) v3: - Add missing newlines in debug messages (Lyude) - Use NV_ATOMIC (Lyude) - Add missing debug log for invalid format (Ilia) v4: add plane name in debug messages (Ilia) Signed-off-by:
Ilia Mirkin <imirkin@alum.mit.edu> Cc: Ben Skeggs <bskeggs@redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20210205224140.28174-1-contact@emersion.fr
-
- 09 Jul, 2023 2 commits
-
-
Gurchetan Singh authored
We don't want to create a fence for every command submission. It's only necessary when userspace provides a waitable token for submission. This could be: 1) bo_handles, to be used with VIRTGPU_WAIT 2) out_fence_fd, to be used with dma_fence apis 3) a ring_idx provided with VIRTGPU_CONTEXT_PARAM_POLL_RINGS_MASK + DRM event API 4) syncobjs in the future The use case for just submitting a command to the host, and expecting no response. For example, gfxstream has GFXSTREAM_CONTEXT_PING that just wakes up the host side worker threads. There's also CROSS_DOMAIN_CMD_SEND which just sends data to the Wayland server. This prevents the need to signal the automatically created virtio_gpu_fence. In addition, VIRTGPU_EXECBUF_RING_IDX is checked when creating a DRM event object. VIRTGPU_CONTEXT_PARAM_POLL_RINGS_MASK is already defined in terms of per-context rings. It was theoretically possible to create a DRM event on the global timeline (ring_idx == 0), if the context enabled DRM event polling. However, that wouldn't work and userspace (Sommelier). Explicitly disallow it for clarity. Signed-off-by:
Gurchetan Singh <gurchetansingh@chromium.org> Reviewed-by:
Dmitry Osipenko <dmitry.osipenko@collabora.com> Tested-by:
-
Sui Jingfeng authored
Fixes the following build errors on arm64: drivers/video/fbdev/hyperv_fb.c: In function 'hvfb_getmem': >> drivers/video/fbdev/hyperv_fb.c:1033:24: error: 'screen_info' undeclared (first use in this function) 1033 | base = screen_info.lfb_base; | ^~~~~~~~~~~ drivers/video/fbdev/hyperv_fb.c:1033:24: note: each undeclared identifier is reported only once for each function it appears in >> drivers/gpu/drm/hyperv/hyperv_drm_drv.c:75:54: error: 'screen_info' undeclared (first use in this function) 75 | drm_aperture_remove_conflicting_framebuffers(screen_info.lfb_base, | ^~~~~~~~~~~ drivers/gpu/drm/hyperv/hyperv_drm_drv.c:75:54: note: each undeclared identifier is reported only once for each function it appears in Reported-by:
-
- 08 Jul, 2023 7 commits
-
-
Thomas Zimmermann authored
Use the existing generator macros to create deferred-I/O helpers for xen-fbfront and set them in the fb_ops structure. Functions for damage handling on memory ranges and areas are provided by the driver. Xen-fbfront's implementation of fb_write writes to system memory, so the generated code can use the respective helper internally. Signed-off-by:
Helge Deller <deller@gmx.de> Link: https://patchwork.freedesktop.org/patch/msgid/20230706151432.20674-11-tzimmermann@suse.de
-
Thomas Zimmermann authored
The Kconfig token FB_SYS_HELPERS_DEFERRED selects everything that is required for deferred I/O on system-memory framebuffers. Select it from XEN_FBDEV_FRONTEND in favor of the existing identical selection. Signed-off-by:
-
Thomas Zimmermann authored
Use the existing generator macros to create deferred-I/O helpers for ssd1307fb and set them in the fb_ops structure. Functions for damage handling on memory ranges and areas are provided by the driver. Ssd1307fb's implementation of fb_write writes to system memory, so the generated code can use the respective helper internally. This also fixes a long-standing bug where fb_write returned an errno code instead of the number of written bytes. See the commit message of commit 921b7383 ("fbdev: Return number of bytes read or written") for more details. Signed-off-by:
-
Thomas Zimmermann authored
The Kconfig token FB_SYS_HELPERS_DEFERRED selects everything that is required for deferred I/O on system-memory framebuffers. Select it from FB_SSD1307 in favor of the existing identical selection. Signed-off-by:
-
Thomas Zimmermann authored
Use the existing generator macros to create deferred-I/O helpers for metronomefb and set them in the fb_ops structure. Functions for damage handling on memory ranges and areas are provided by the driver. Metronomefb's implementation of fb_write writes to system memory, so the generated code can use the respective helper internally. This also fixes a long-standing bug where fb_write returned an errno code instead of the number of written bytes. See the commit message of commit 921b7383 ("fbdev: Return number of bytes read or written") for more details. Signed-off-by:
-
Thomas Zimmermann authored
The Kconfig token FB_SYS_HELPERS_DEFERRED selects everything that is required for deferred I/O on system-memory framebuffers. Select it from FB_METRONOME in favor of the existing identical selection. Signed-off-by:
-
Thomas Zimmermann authored
Use the existing generator macros to create deferred-I/O helpers for hecubafb and set them in the fb_ops structure. Functions for damage handling on memory ranges and areas are provided by the driver. Hecubafb's implementation of fb_write writes to system memory, so the generated code can use the respective helper internally. This also fixes a long-standing bug where fb_write returned an errno code instead of the number of written bytes. See the commit message of commit 921b7383 ("fbdev: Return number of bytes read or written") for more details. Signed-off-by:
-
