- 04 Oct, 2019 23 commits
-
-
Chris Wilson authored
A few callers need to serialise the destruction of their drm_mm_node and ensure it is removed from the drm_mm before freeing. However, to be completely sure that any access from another thread is complete before we free the struct, we require the RELEASE semantics of clear_bit_unlock(). This allows the conditional locking such as Thread A Thread B mutex_lock(mm_lock); if (drm_mm_node_allocated(node)) { drm_mm_node_remove(node); mutex_lock(mm_lock); mutex_unlock(mm_lock); if (drm_mm_node_allocated(node)) drm_mm_node_remove(node); mutex_unlock(mm_lock); } kfree(node); to serialise correctly without any lingering accesses from A to the freed node. Allocation / insertion of the node is assumed never to race with removal or eviction scanning. Signed-off-by:Chris Wilson <chris@chris-wilson.co.uk> Reviewed-by:
Tvrtko Ursulin <tvrtko.ursulin@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20191003210100.22250-5-chris@chris-wilson.co.uk
-
Chris Wilson authored
A straightforward conversion of assignment and checking of the boolean state flags (allocated, scanned) into non-atomic bitops. The caller remains responsible for all locking around the drm_mm and its nodes. Signed-off-by:
-
Chris Wilson authored
In preparation for rearranging the booleans into a flags field, ensure all the current users are using the inline helpers and not directly accessing the members. Signed-off-by:
-
Chris Wilson authored
Make dma_fence_enable_sw_signaling() behave like its dma_fence_add_callback() and dma_fence_default_wait() counterparts and perform the test to enable signaling under the fence->lock, along with the action to do so. This ensure that should an implementation be trying to flush the cb_list (by signaling) on retirement before freeing the fence, it can do so in a race-free manner. See also 0fc89b68 ("dma-fence: Simply wrap dma_fence_signal_locked with dma_fence_signal"). v2: Refactor all 3 enable_signaling paths to use a common function. v3: Don't argue, just keep the tracepoint in the existing spot. Signed-off-by:
-
Tomi Valkeinen authored
If use_mclk is false, mclk_mode is written to a register without initialization. This doesn't cause any ill effects as the written value is not used when use_mclk is false. To fix this, write use_mclk only when use_mclk is true. Signed-off-by:
Tomi Valkeinen <tomi.valkeinen@ti.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190930103840.18970-8-tomi.valkeinen@ti.comReviewed-by:
Laurent Pinchart <laurent.pinchart@ideasonboard.com>
-
Tomi Valkeinen authored
Currently the HDMI driver uses always limited range RGB output. This patch improves the behavior by using limited range only if the output is identified as a HDMI display, and VIC > 1. Signed-off-by:
-
Jyri Sarha authored
The core.c just for registering the drivers is kind of useless. Let's get rid of it and register the dss drivers in dss.c. Signed-off-by:
Jyri Sarha <jsarha@ti.com> Signed-off-by:
-
Tomi Valkeinen authored
OMAP2 and OMAP3/AM4 have limitations with the scaler: - OMAP2 can only scale XRGB8888 - OMAP3/AM4 can only scale XRGB8888, RGB565, YUYV and UYVY The driver doesn't check these limitations, which leads to sync-lost floods. This patch adds a check for the pixel formats when scaling. Signed-off-by:
-
Alejandro Hernandez authored
A "HDMI I2C Master Error" is sometimes reported with the current DDC SCL timings. The current settings for a 10us SCL period (100 KHz) causes the error with some displays. This patch increases the SCL signal period from 10us to 10.2us, with the new settings the error is not observed Signed-off-by:
Alejandro Hernandez <ajhernandez@ti.com> Signed-off-by:
-
Tomi Valkeinen authored
Avoid unnecessary copy in mgr_fld_read/write by taking a pointer to the reg_resc and using that. Signed-off-by:
-
Tomi Valkeinen authored
Commit d49cd155 ("OMAPDSS: DISPC: lock access to DISPC_CONTROL & DISPC_CONFIG") added locking to mgr_fld_write(). This was needed in omapfb times due to lack of good locking, especially in the case of both V4L2 and fbdev layers using the DSS driver. This is not needed for omapdrm, so we can remove the locking. Signed-off-by:
-
Thomas Zimmermann authored
By putting cursor BOs at the high end of the video memory, we can avoid memory fragmentation. Starting at the low end, contiguous video memory is available for framebuffers. The patch also simplifies the buffer swapping and aligns it with the ast driver. If there are more drivers with similar requirements, the code could be moved into a shared place. Signed-off-by:
Thomas Zimmermann <tzimmermann@suse.de> Acked-by:
Gerd Hoffmann <kraxel@redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190927091301.10574-8-tzimmermann@suse.de
-
Thomas Zimmermann authored
The double-buffered cursor image is currently stored in video memory by creating two BOs and pinning them to VRAM. The exact location is chosen by VRAM helpers. The pinned cursor BOs can conflict with framebuffer BOs and prevent the primary plane from displaying its framebuffer. As a first step to solving this problem, we reserve dedicated space at the high end of the video memory for the cursor images. As the amount of video memory now differs from the amount of available framebuffer memory, size tests are adapted accordingly. Signed-off-by:
-
Thomas Zimmermann authored
Selecting the correct BO for the new cursor image is not relevant outside of mgag200_show_cursor(). Let the function do the work. Signed-off-by:
-
Thomas Zimmermann authored
Separating the management of buffer objects from updating the hardware cursor buffer gives the code more structure. While doing this, we can further split the image-update code into code for writing the buffer, setting the base scan-out address, and enabling the cursor. The first two operations are in dedicated functions update() and set_base(). Signed-off-by:
-
Thomas Zimmermann authored
Adding mgag200_move_cursor() makes the cursor code more consistent and will become handy when we move to universal cursor planes. Signed-off-by:
-
Thomas Zimmermann authored
Moving the cursor initialization and cleanup into separate functions makes the overall code slightly more readable. Signed-off-by:
-
Thomas Zimmermann authored
Although the driver source code is fairly inconsistent wrt naming, the prefix should be mgag200. Rename cursor functions accordingly. Signed-off-by:
-
Thomas Zimmermann authored
By putting cursor BOs at the high end of the video memory, we can avoid memory fragmentation. Starting at the low end, contiguous video memory is available for framebuffers. The patch also simplifies the buffer swapping by splitting struct ast_private.cursor_cache BO into two separate boffer objects. Cursor images alternate between these buffers instead of offsets within cursor_cache. v3: * fixes space-before-tab error near AST_HWC_SIGNATURE_CHECKSUM Signed-off-by:
-
Thomas Zimmermann authored
Selecting the correct offset for the new cursor image is not relevant outside of ast_show_cursor(). Let the function do the work. Signed-off-by:
-
Thomas Zimmermann authored
A call to ast's show-cursor function now receives the cursor image and updates the buffer. The change splits off image update and base-address update into separate functions. v3: * move ast_{show,hide}_cursor() in a previous patch Signed-off-by: -
Thomas Zimmermann authored
This patch only moves around code for easier review of later patches. No functional cahnges are made. Signed-off-by:
-
Thomas Zimmermann authored
Separating the cursor's move() function from the show() function in preparation of further rework of the cursor update code. 'Showing' the cursor from within the move() function is required to update the cursor position. Signed-off-by:
-
- 03 Oct, 2019 5 commits
-
-
Maxime Ripard authored
We haven't done any backmerge for a while due to the merge window, and it starts to become an issue for komeda. Let's bring 5.4-rc1 in. Signed-off-by:Maxime Ripard <mripard@kernel.org>
-
Icenowy Zheng authored
This reverts commit da676c6a. The original commit adds a start parameter to the calculation of the start delay according to some old BSP versions from Allwinner. However, there're two ways to add this delay -- add it in DSI controller or add it in the TCON. Add it in both controllers won't work. The code before this commit is picked from new versions of BSP kernel, which has a comment for the 1 that says "put start_delay to tcon". By checking the sun4i_tcon0_mode_set_cpu() in sun4i_tcon driver, it has already added this delay, so we shouldn't repeat to add the delay in DSI controller, otherwise the timing won't match. Signed-off-by:
Icenowy Zheng <icenowy@aosc.io> Reviewed-by:
Jagan Teki <jagan@amarulasolutions.com> Signed-off-by:
-
Jagan Teki authored
The LCD timing definitions between Linux DRM vs Allwinner are different, below diagram shows this clear differences. Active Front Sync Back Region Porch Porch <-----------------------><----------------><--------------><--------------> //////////////////////| ////////////////////// | ////////////////////// |.................. ................ ________________ <----- [hv]display -----> <------------- [hv]sync_start ------------> <--------------------- [hv]sync_end ----------------------> <-------------------------------- [hv]total ------------------------------> <----- lcd_[xy] --------> <- lcd_[hv]spw -> <---------- lcd_[hv]bp ---------> <-------------------------------- lcd_[hv]t ------------------------------> The DSI driver misinterpreted the hbp term from the BSP code to refer only to the backporch, when in fact it was backporch + sync. Thus the driver incorrectly used the horizontal front porch plus sync in its calculation of the DRQ set bit value, when it should not have included the sync timing. Including additional sync timings leads to flip_done timed out as: WARNING: CPU: 0 PID: 31 at drivers/gpu/drm/drm_atomic_helper.c:1429 drm_atomic_helper_wait_for_vblanks.part.1+0x298/0x2a0 [CRTC:46:crtc-0] vblank wait timed out Modules linked in: CPU: 0 PID: 31 Comm: kworker/0:1 Not tainted 5.1.0-next-20190514-00026-g01f0c75b902d-dirty #13 Hardware name: Allwinner sun8i Family Workqueue: events deferred_probe_work_func [<c010ed54>] (unwind_backtrace) from [<c010b76c>] (show_stack+0x10/0x14) [<c010b76c>] (show_stack) from [<c0688c70>] (dump_stack+0x84/0x98) [<c0688c70>] (dump_stack) from [<c011d9e4>] (__warn+0xfc/0x114) [<c011d9e4>] (__warn) from [<c011da40>] (warn_slowpath_fmt+0x44/0x68) [<c011da40>] (warn_slowpath_fmt) from [<c040cd50>] (drm_atomic_helper_wait_for_vblanks.part.1+0x298/0x2a0) [<c040cd50>] (drm_atomic_helper_wait_for_vblanks.part.1) from [<c040e694>] (drm_atomic_helper_commit_tail_rpm+0x5c/0x6c) [<c040e694>] (drm_atomic_helper_commit_tail_rpm) from [<c040e4dc>] (commit_tail+0x40/0x6c) [<c040e4dc>] (commit_tail) from [<c040e5cc>] (drm_atomic_helper_commit+0xbc/0x128) [<c040e5cc>] (drm_atomic_helper_commit) from [<c0411b64>] (restore_fbdev_mode_atomic+0x1cc/0x1dc) [<c0411b64>] (restore_fbdev_mode_atomic) from [<c04156f8>] (drm_fb_helper_restore_fbdev_mode_unlocked+0x54/0xa0) [<c04156f8>] (drm_fb_helper_restore_fbdev_mode_unlocked) from [<c0415774>] (drm_fb_helper_set_par+0x30/0x54) [<c0415774>] (drm_fb_helper_set_par) from [<c03ad450>] (fbcon_init+0x560/0x5ac) [<c03ad450>] (fbcon_init) from [<c03eb8a0>] (visual_init+0xbc/0x104) [<c03eb8a0>] (visual_init) from [<c03ed1b8>] (do_bind_con_driver+0x1b0/0x390) [<c03ed1b8>] (do_bind_con_driver) from [<c03ed780>] (do_take_over_console+0x13c/0x1c4) [<c03ed780>] (do_take_over_console) from [<c03ad800>] (do_fbcon_takeover+0x74/0xcc) [<c03ad800>] (do_fbcon_takeover) from [<c013c9c8>] (notifier_call_chain+0x44/0x84) [<c013c9c8>] (notifier_call_chain) from [<c013cd20>] (__blocking_notifier_call_chain+0x48/0x60) [<c013cd20>] (__blocking_notifier_call_chain) from [<c013cd50>] (blocking_notifier_call_chain+0x18/0x20) [<c013cd50>] (blocking_notifier_call_chain) from [<c03a6e44>] (register_framebuffer+0x1e0/0x2f8) [<c03a6e44>] (register_framebuffer) from [<c04153c0>] (__drm_fb_helper_initial_config_and_unlock+0x2fc/0x50c) [<c04153c0>] (__drm_fb_helper_initial_config_and_unlock) from [<c04158c8>] (drm_fbdev_client_hotplug+0xe8/0x1b8) [<c04158c8>] (drm_fbdev_client_hotplug) from [<c0415a20>] (drm_fbdev_generic_setup+0x88/0x118) [<c0415a20>] (drm_fbdev_generic_setup) from [<c043f060>] (sun4i_drv_bind+0x128/0x160) [<c043f060>] (sun4i_drv_bind) from [<c044b598>] (try_to_bring_up_master+0x164/0x1a0) [<c044b598>] (try_to_bring_up_master) from [<c044b668>] (__component_add+0x94/0x140) [<c044b668>] (__component_add) from [<c0445e1c>] (sun6i_dsi_probe+0x144/0x234) [<c0445e1c>] (sun6i_dsi_probe) from [<c0452ef4>] (platform_drv_probe+0x48/0x9c) [<c0452ef4>] (platform_drv_probe) from [<c04512cc>] (really_probe+0x1dc/0x2c8) [<c04512cc>] (really_probe) from [<c0451518>] (driver_probe_device+0x60/0x160) [<c0451518>] (driver_probe_device) from [<c044f7a4>] (bus_for_each_drv+0x74/0xb8) [<c044f7a4>] (bus_for_each_drv) from [<c045107c>] (__device_attach+0xd0/0x13c) [<c045107c>] (__device_attach) from [<c0450474>] (bus_probe_device+0x84/0x8c) [<c0450474>] (bus_probe_device) from [<c0450900>] (deferred_probe_work_func+0x64/0x90) [<c0450900>] (deferred_probe_work_func) from [<c0135970>] (process_one_work+0x204/0x420) [<c0135970>] (process_one_work) from [<c013690c>] (worker_thread+0x274/0x5a0) [<c013690c>] (worker_thread) from [<c013b3d8>] (kthread+0x11c/0x14c) [<c013b3d8>] (kthread) from [<c01010e8>] (ret_from_fork+0x14/0x2c) Exception stack(0xde539fb0 to 0xde539ff8) 9fa0: 00000000 00000000 00000000 00000000 9fc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 9fe0: 00000000 00000000 00000000 00000000 00000013 00000000 ---[ end trace b57eb1e5c64c6b8b ]--- random: fast init done [drm:drm_atomic_helper_wait_for_dependencies] *ERROR* [CRTC:46:crtc-0] flip_done timed out [drm:drm_atomic_helper_wait_for_dependencies] *ERROR* [CONNECTOR:48:DSI-1] flip_done timed out [drm:drm_atomic_helper_wait_for_dependencies] *ERROR* [PLANE:30:plane-0] flip_done timed out With the terms(as described in above diagram) fixed, the panel displays correctly without any timeouts. Tested-by:Merlijn Wajer <merlijn@wizzup.org> Signed-off-by:
-
Jagan Teki authored
Allwinner MIPI DSI controllers are supplied with SoC DSI power rails via VCC-DSI pin. Add support for this supply pin by adding voltage regulator handling code to MIPI DSI driver. Tested-by:
Chen-Yu Tsai <wens@csie.org> Signed-off-by:
-
Jagan Teki authored
Allwinner MIPI DSI controllers are supplied with SoC DSI power rails via VCC-DSI pin. Some board still work without supplying this but give more faith on datasheet and hardware schematics and document this supply property in required property list. Reviewed-by:
Rob Herring <robh@kernel.org> Tested-by:
-
- 02 Oct, 2019 5 commits
-
-
Jani Nikula authored
Allow better abstraction of the drm_debug global variable in the future. No functional changes. v2: Move unlikely() to drm_debug_enabled() Cc: Rob Clark <robdclark@gmail.com> Cc: Sean Paul <sean@poorly.run> Cc: linux-arm-msm@vger.kernel.org Cc: freedreno@lists.freedesktop.org Reviewed-by:
Rob Clark <robdclark@gmail.com> Signed-off-by:
Jani Nikula <jani.nikula@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/c7142cdebb5f6fed527272b333cd6c43c0aa68ec.1569329774.git.jani.nikula@intel.com
-
Jani Nikula authored
Allow better abstraction of the drm_debug global variable in the future. No functional changes. Cc: Francisco Jerez <currojerez@riseup.net> Reviewed-by:
Francisco Jerez <currojerez@riseup.net> Signed-off-by:
-
Jani Nikula authored
Allow better abstraction of the drm_debug global variable in the future. No functional changes. Cc: Lucas Stach <l.stach@pengutronix.de> Cc: Russell King <linux+etnaviv@armlinux.org.uk> Cc: Christian Gmeiner <christian.gmeiner@gmail.com> Cc: etnaviv@lists.freedesktop.org Acked-by:
Lucas Stach <l.stach@pengutronix.de> Signed-off-by:
-
Jani Nikula authored
Add helper to check if a drm debug category is enabled. Convert drm core to use it. No functional changes. v2: Move unlikely() to drm_debug_enabled() (Eric) v3: Keep unlikely() when combined with other conditions (Eric) Cc: Eric Engestrom <eric@engestrom.ch> Acked-by:
Alex Deucher <alexander.deucher@amd.com> Reviewed-by:
Eric Engestrom <eric@engestrom.ch> Signed-off-by:
-
Jani Nikula authored
Move drm_debug variable declaration and definition to where they are relevant and needed. No functional changes. Acked-by:
-
- 01 Oct, 2019 5 commits
-
-
Ville Syrjälä authored
Add a helper to initialize a rectangle from x/y/w/h information. Signed-off-by:
Ville Syrjälä <ville.syrjala@linux.intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20190930134214.24702-2-ville.syrjala@linux.intel.comReviewed-by:
-
Ville Syrjälä authored
Add a helper to translate a rectangle to an absolute position. Signed-off-by:
-
Mihail Atanassov authored
When initially turning a crtc on, drm_reset_vblank_timestamp will set the vblank timestamp to 0 for any driver that doesn't provide a ->get_vblank_timestamp() hook. Unfortunately, the FLIP_COMPLETE event depends on that timestamp, and the only way to regenerate a valid one is to have vblank interrupts enabled and have a valid in-ISR call to drm_crtc_handle_vblank. Additionally, if the user doesn't request vblanks but _does_ request FLIP_COMPLETE events, we still don't have a good timestamp: it'll be the same stamp as the last vblank one. Work around the issue by always enabling vblanks when the CRTC is on. Reducing the amount of time that PL0 has to be unmasked would be nice to fix at a later time. Changes since v1 [https://patchwork.freedesktop.org/patch/331727/]: - moved drm_crtc_vblank_put call to the ->atomic_disable() hook Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Liviu Dudau <Liviu.Dudau@arm.com> Signed-off-by:
Mihail Atanassov <mihail.atanassov@arm.com> Reviewed-by:
James Qian Wang (Arm Technology China) <james.qian.wang@arm.com> Signed-off-by:
Ayan kumar halder <ayan.halder@arm.com> Link: https://patchwork.freedesktop.org/patch/msgid/20191001142121.13939-1-mihail.atanassov@arm.com
-
Mihail Atanassov authored
No change in behaviour; IRQ_RETVAL is about twice as popular as manually writing out the ternary. Signed-off-by:
-
Ulf Magnusson authored
Commit 554b3529 ("thermal/drivers/core: Remove the module Kconfig's option") changed the type of THERMAL from tristate to bool, so THERMAL || !THERMAL is now always y. Remove the redundant dependency. Discovered through Kconfiglib detecting a dependency loop. The C tools simplify the expression to y before running dependency loop detection, and so don't see it. Changing the type of THERMAL back to tristate makes the C tools detect the same loop. Not sure if running dep. loop detection after simplification can be called a bug. Fixing this nit unbreaks Kconfiglib on the kernel at least. Signed-off-by:
Ulf Magnusson <ulfalizer@gmail.com> Signed-off-by:
Noralf Trønnes <noralf@tronnes.org> Link: https://patchwork.freedesktop.org/patch/msgid/20190927174218.GA32085@huvuddator
-
- 30 Sep, 2019 2 commits
-
-
Ville Syrjälä authored
Get rid of the drm_fixp_from_fraction() usage and just do the straightforward calculation directly. Cc: Lyude Paul <lyude@redhat.com> Cc: Harry Wentland <harry.wentland@amd.com> Cc: Alex Deucher <alexander.deucher@amd.com> Signed-off-by:
Lyude Paul <lyude@redhat.com>
-
Ville Syrjälä authored
Make drm_dp_get_vc_payload() tolerate arbitrary DP_LINK_BW_* values, just like drm_dp_bw_code_to_link_rate() does since commit 57a1b089 ("drm: Make the bw/link rate calculations more forgiving"). Cc: Lyude Paul <lyude@redhat.com> Cc: Sean Paul <seanpaul@chromium.org> Signed-off-by:
-
