- 28 Feb, 2022 3 commits
-
-
Noralf Trønnes authored
Remove trailing whitespace from a comment. Acked-by:
Sam Ravnborg <sam@ravnborg.org> Signed-off-by:
Noralf Trønnes <noralf@tronnes.org> Signed-off-by:
Maxime Ripard <maxime@cerno.tech> Link: https://patchwork.freedesktop.org/patch/msgid/20220227124713.39766-3-noralf@tronnes.org
-
Noralf Trønnes authored
Add binding for MIPI DBI compatible SPI panels. v6: - Fix indentation (Rob) v5: - Add sainsmart18 to compatible items (Rob) - Expand write-only description (Sam) v4: - There should only be two compatible (Maxime) - s/panel-dbi-spi/panel-mipi-dbi-spi/in compatible v3: - Move properties to Device Tree (Maxime) - Use contains for compatible (Maxime) - Add backlight property to example - Flesh out description v2: - Fix path for panel-common.yaml - Use unevaluatedProperties - Drop properties which are in the allOf section - Drop model property (Rob) Acked-by:
Rob Herring <robh@kernel.org> Signed-off-by:
-
Daniel Vetter authored
Some things changed, and add two useful links. v2: Also include a link to the QR encoding work. Plus review from Javier. v3: Fix typo Guilherme spotted. Reviewed-by:
Guilherme G. Piccoli <gpiccoli@igalia.com> Acked-by:
Pekka Paalanen <pekka.paalanen@collabora.com> Reviewed-by:
Javier Martinez Canillas <javierm@redhat.com> Cc: Javier Martinez Canillas <javierm@redhat.com> Cc: Pekka Paalanen <ppaalanen@gmail.com> Cc: gpiccoli@igalia.com Signed-off-by:
Daniel Vetter <daniel.vetter@intel.com> Link: https://patchwork.freedesktop.org/patch/msgid/20220224132425.3463791-1-daniel.vetter@ffwll.ch
-
- 25 Feb, 2022 16 commits
-
-
Maxime Ripard authored
The omap KMS driver will call drm_plane_create_color_properties() with a default encoding and range values of BT601 and Full Range, respectively. Since the initial value wasn't carried over in the state, the driver had to set it again in omap_plane_reset(). However, the helpers have been adjusted to set it properly at reset, so this is not needed anymore. Reviewed-by:
Tomi Valkeinen <tomba@kernel.org> Signed-off-by:
-
Dave Stevenson authored
The drm_plane_create_color_properties() function asks for an initial value for the color encoding and range, and will set the associated plane state variable with that value if a state is present. However, that function is usually called at a time where there's no state yet. Since the drm_plane_state reset helper doesn't take care of reading that value when it's called, it means that in most cases the initial value will be 0 (so DRM_COLOR_YCBCR_BT601 and DRM_COLOR_YCBCR_LIMITED_RANGE, respectively), or the drivers will have to work around it. Let's add some code in __drm_atomic_helper_plane_state_reset() to get the initial encoding and range value if the property has been attached in order to fix this. Reviewed-by:
Harry Wentland <harry.wentland@amd.com> Signed-off-by:
Dave Stevenson <dave.stevenson@raspberrypi.com> Signed-off-by:
-
Maxime Ripard authored
The sun4i KMS driver will call drm_plane_create_zpos_property() with an init value depending on the plane type. Since the initial value wasn't carried over in the state, the driver had to set it again in sun4i_backend_layer_reset(). However, the helpers have been adjusted to set it properly at reset, so this is not needed anymore. Cc: linux-arm-kernel@lists.infradead.org Cc: linux-sunxi@lists.linux.dev Cc: Chen-Yu Tsai <wens@csie.org> Reviewed-by:
Jernej Skrabec <jernej.skrabec@gmail.com> Signed-off-by:
-
Maxime Ripard authored
The sti KMS driver will call drm_plane_create_zpos_property() with an init value depending on the plane type. Since the initial value wasn't carried over in the state, the driver had to set it again in sti_plane_reset(). However, the helpers have been adjusted to set it properly at reset, so this is not needed anymore. Reviewed-by:
Alain Volmat <alain.volmat@foss.st.com> Reviewed-by:
Philippe Cornu <philippe.cornu@foss.st.com> Signed-off-by:
-
Maxime Ripard authored
The rcar-du KMS driver will call drm_plane_create_zpos_property() with an init value depending on the plane type. Since the initial value wasn't carried over in the state, the driver had to set it again in rcar_du_plane_reset() and rcar_du_vsp_plane_reset(). However, the helpers have been adjusted to set it properly at reset, so this is not needed anymore. Cc: linux-renesas-soc@vger.kernel.org Reviewed-by:
Kieran Bingham <kieran.bingham+renesas@ideasonboard.com> Reviewed-by:
Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by:
-
Maxime Ripard authored
The omap KMS driver will call drm_plane_create_zpos_property() with an init value of the plane index and the plane type. Since the initial value wasn't carried over in the state, the driver had to set it again in omap_plane_reset(). However, the helpers have been adjusted to set it properly at reset, so this is not needed anymore. Reviewed-by:
-
Maxime Ripard authored
The nouveau KMS driver will call drm_plane_create_zpos_property() with an init value depending on the plane purpose. Since the initial value wasn't carried over in the state, the driver had to set it again in nv50_wndw_reset(). However, the helpers have been adjusted to set it properly at reset, so this is not needed anymore. Cc: nouveau@lists.freedesktop.org Cc: Ben Skeggs <bskeggs@redhat.com> Cc: Karol Herbst <kherbst@redhat.com> Cc: Lyude Paul <lyude@redhat.com> Signed-off-by:
Karol Herbst <kherbst@redhat.com> Reviewed-by:
Lyude Paul <lyude@redhat.com> Link: https://patchwork.freedesktop.org/patch/msgid/20220221095918.18763-14-maxime@cerno.tech
-
Maxime Ripard authored
The mdp KMS driver will call drm_plane_create_zpos_property() with an init value depending on the plane purpose. Since the initial value wasn't carried over in the state, the driver had to set it again in mdp5_plane_reset(). However, the helpers have been adjusted to set it properly at reset, so this is not needed anymore. Cc: freedreno@lists.freedesktop.org Cc: linux-arm-msm@vger.kernel.org Cc: Abhinav Kumar <quic_abhinavk@quicinc.com> Cc: Rob Clark <robdclark@gmail.com> Cc: Sean Paul <sean@poorly.run> Reviewed-by:
Dmitry Baryshkov <dmitry.baryshkov@linaro.org> Signed-off-by:
-
Dave Stevenson authored
The drm_plane_create_zpos_property() function asks for an initial value, and will set the associated plane state variable with that value if a state is present. However, that function is usually called at a time where there's no state yet. Since the drm_plane_state reset helper doesn't take care of reading that value when it's called, it means that in most cases the initial value will be 0, or the drivers will have to work around it. Let's add some code in __drm_atomic_helper_plane_state_reset() to get the initial zpos value if the property has been attached in order to fix this. Reviewed-by:
-
Dave Stevenson authored
Some functions to create properties (drm_plane_create_zpos_property or drm_plane_create_color_properties for example) will ask for a range of acceptable value and an initial one. This initial value is then stored in the values array for that property. Let's provide an helper to access this property. Acked-by:
Daniel Vetter <daniel.vetter@ffwll.ch> Reviewed-by:
-
Maxime Ripard authored
While the omap_plane_init() function calls drm_plane_create_zpos_property() with an initial value of 0, omap_plane_reset() will force it to another value depending on the plane type. Fix the discrepancy by setting the initial zpos value to the same value in the drm_plane_create_zpos_property() call. Reviewed-by:
-
Jagan Teki authored
Devices can also be child nodes when we also control that device through the upstream device (ie, MIPI-DCS for a MIPI-DSI device). drm_of_find_panel_or_bridge can lookup panel or bridge for a given device has port and endpoint and it fails to lookup if the device has a child nodes. This patch add support to lookup for a child node of the given parent that isn't either port or ports. Example OF graph representation of DSI host, which has port but not has ports and has child panel node. dsi { compatible = "allwinner,sun6i-a31-mipi-dsi"; #address-cells = <1>; #size-cells = <0>; port { dsi_in_tcon0: endpoint { remote-endpoint = <tcon0_out_dsi>; }; panel@0 { reg = <0>; }; }; Example OF graph representation of DSI host, which has ports but not has port and has child panel node. dsi { compatible = "samsung,exynos5433-mipi-dsi"; #address-cells = <1>; #size-cells = <0>; ports { #address-cells = <1>; #size-cells = <0>; port@0 { reg = <0>; dsi_to_mic: endpoint { remote-endpoint = <&mic_to_dsi>; }; }; }; panel@0 { reg = <0>; }; }; Example OF graph representation of DSI host, which has neither a port nor a ports but has child panel node. dsi0 { compatible = "ste,mcde-dsi"; #address-cells = <1>; #size-cells = <0>; panel@0 { reg = <0>; }; }; Reviewed-by:Linus Walleij <linus.walleij@linaro.org> Signed-off-by:
Jagan Teki <jagan@amarulasolutions.com> Signed-off-by:
-
Raphael Gallais-Pou authored
This patch adds the CRC hashing feature supported by some recent hardware versions of the LTDC. This is useful for test suite such as IGT-GPU-tools [1] where a CRTC output frame can be compared to a test reference frame thanks to their respective CRC hash. [1] https://cgit.freedesktop.org/drm/igt-gpu-toolsSigned-off-by:
Raphael Gallais-Pou <raphael.gallais-pou@foss.st.com> Acked-by:
Yannick Fertre <yannick.fertre@foss.st.com> Signed-off-by:
-
Nathan Chancellor authored
Clang warns: drivers/gpu/drm/stm/ltdc.c:625:2: warning: variable 'val' is used uninitialized whenever switch default is taken [-Wsometimes-uninitialized] default: ^~~~~~~ drivers/gpu/drm/stm/ltdc.c:635:2: note: uninitialized use occurs here val |= LxPCR_YCEN; ^~~ drivers/gpu/drm/stm/ltdc.c:600:9: note: initialize the variable 'val' to silence this warning u32 val; ^ = 0 1 warning generated. Use a return instead of break in the default case to fix the warning. Add an error message so that this return is not silent, which could hide issues in the future. Fixes: 484e72d3 ("drm/stm: ltdc: add support of ycbcr pixel formats") Link: https://github.com/ClangBuiltLinux/linux/issues/1575Acked-by:Nick Desaulniers <ndesaulniers@google.com> Reviewed-by:
Nathan Chancellor <nathan@kernel.org> Signed-off-by:
-
Kees Cook authored
The link_status array was not large enough to read the Adjust Request Post Cursor2 register, so remove the common helper function to avoid an OOB read, found with a -Warray-bounds build: drivers/gpu/drm/drm_dp_helper.c: In function 'drm_dp_get_adjust_request_post_cursor': drivers/gpu/drm/drm_dp_helper.c:59:27: error: array subscript 10 is outside array bounds of 'const u8[6]' {aka 'const unsigned char[6]'} [-Werror=array-bounds] 59 | return link_status[r - DP_LANE0_1_STATUS]; | ~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~ drivers/gpu/drm/drm_dp_helper.c:147:51: note: while referencing 'link_status' 147 | u8 drm_dp_get_adjust_request_post_cursor(const u8 link_status[DP_LINK_STATUS_SIZE], | ~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Replace the only user of the helper with an open-coded fetch and decode, similar to drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c. Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Cc: Maxime Ripard <mripard@kernel.org> Cc: Thomas Zimmermann <tzimmermann@suse.de> Cc: David Airlie <airlied@linux.ie> Cc: dri-devel@lists.freedesktop.org Fixes: 79465e0f ("drm/dp: Add helper to get post-cursor adjustments") Signed-off-by:Kees Cook <keescook@chromium.org> Reviewed-by:
Gustavo A. R. Silva <gustavoars@kernel.org> Reviewed-by:
Jani Nikula <jani.nikula@intel.com> Link: https://lore.kernel.org/r/20220105173507.2420910-1-keescook@chromium.orgSigned-off-by:
Thierry Reding <treding@nvidia.com>
-
Kees Cook authored
The pcon_dsc_dpcd array holds 13 registers (0x92 through 0x9E). Fix the math to calculate the max size. Found from a -Warray-bounds build: drivers/gpu/drm/drm_dp_helper.c: In function 'drm_dp_pcon_dsc_bpp_incr': drivers/gpu/drm/drm_dp_helper.c:3130:28: error: array subscript 12 is outside array bounds of 'const u8[12]' {aka 'const unsigned char[12]'} [-Werror=array-bounds] 3130 | buf = pcon_dsc_dpcd[DP_PCON_DSC_BPP_INCR - DP_PCON_DSC_ENCODER]; | ~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ drivers/gpu/drm/drm_dp_helper.c:3126:39: note: while referencing 'pcon_dsc_dpcd' 3126 | int drm_dp_pcon_dsc_bpp_incr(const u8 pcon_dsc_dpcd[DP_PCON_DSC_ENCODER_CAP_SIZE]) | ~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Cc: Maxime Ripard <mripard@kernel.org> Cc: Thomas Zimmermann <tzimmermann@suse.de> Cc: David Airlie <airlied@linux.ie> Cc: dri-devel@lists.freedesktop.org Fixes: e2e16da3 ("drm/dp_helper: Add support for Configuring DSC for HDMI2.1 Pcon") Cc: stable@vger.kernel.org Reviewed-by:
-
- 24 Feb, 2022 2 commits
-
-
Cai Huoqing authored
The nouveau driver depends on include/linux/list.h instead of nvif/list.h, so remove the obstacle-nvif/list.h. Signed-off-by:
Cai Huoqing <cai.huoqing@linux.dev> Signed-off-by:
-
Javier Martinez Canillas authored
There is now a drm_fb_xrgb8888_to_mono_reversed() helper function to do format conversion from XRGB8888 to reversed monochrome. Use that helper and remove the open coded version in the repaper driver. Signed-off-by:
Thomas Zimmermann <tzimmermann@suse.de> Tested-by:
-
- 23 Feb, 2022 9 commits
-
-
Hans de Goede authored
Some devices use e.g. a portrait panel in a standard laptop casing made for landscape panels. efifb calls drm_get_panel_orientation_quirk() and sets fb_info.fbcon_rotate_hint to make fbcon rotate the console so that it shows up-right instead of on its side. When switching to simpledrm the fbcon renders on its side. Call the drm_connector_set_panel_orientation_with_quirk() helper to add a "panel orientation" property on devices listed in the quirk table, to make the fbcon (and aware userspace apps) rotate the image to display properly. Cc: Javier Martinez Canillas <javierm@redhat.com> Signed-off-by:
Hans de Goede <hdegoede@redhat.com> Reviewed-by:
-
Arunpravin authored
create a pot-sized mm, then allocate one of each possible order within. This should leave the mm with exactly one page left. Free the largest block, then whittle down again. Eventually we will have a fully 50% fragmented mm. v2(Matthew Auld): - removed unnecessary test succeeded print - replace list_del()/list_add_tail() with list_move_tail() Signed-off-by:
Arunpravin <Arunpravin.PaneerSelvam@amd.com> Reviewed-by:
Matthew Auld <matthew.auld@intel.com> Acked-by:
Christian König <christian.koenig@amd.com> Link: https://patchwork.freedesktop.org/patch/msgid/20220222174845.2175-7-Arunpravin.PaneerSelvam@amd.comSigned-off-by:
-
Arunpravin authored
- add a test to ascertain that the critical functionalities of the program is working fine - add a timeout helper function v2: - removed unnecessary test succeeded print - replace list_del()/list_add_tail() with list_move_tail() Signed-off-by:
-
Arunpravin authored
create a pot-sized mm, then allocate one of each possible order within. This should leave the mm with exactly one page left. v2: - removed unnecessary test succeeded print - replace list_del()/list_add_tail() with list_move_tail() Signed-off-by:
-
Arunpravin authored
create a mm with one block of each order available, and try to allocate them all. v2(Matthew Auld): - removed unnecessary test succeeded print - replace list_del()/list_add_tail() with list_move_tail() Signed-off-by:
-
Arunpravin authored
- add a test to check the range allocation - export get_buddy() function in drm_buddy.c - export drm_prandom_u32_max_state() in lib/drm_random.c - include helper functions - include prime number header file v2: - add drm_get_buddy() function description (Matthew Auld) - removed unnecessary test succeeded print Signed-off-by:
-
Arunpravin authored
add a test to check the maximum allocation limit v2(Matthew Auld): - added err = -EINVAL in block NULL check - removed unnecessary test succeeded print Signed-off-by:
-
Arunpravin authored
- move i915 buddy selftests into drm selftests folder - add Makefile and Kconfig support - add sanitycheck testcase Prerequisites - These series of selftests patches are created on top of drm buddy series - Enable kselftests for DRM as a module in .config Signed-off-by:
-
Jiawei Gu authored
Add device pointer so scheduler's printing can use DRM_DEV_ERROR() instead, which makes life easier under multiple GPU scenario. v2: amend all calls of drm_sched_init() v3: fill dev pointer for all drm_sched_init() calls Signed-off-by:
Jiawei Gu <Jiawei.Gu@amd.com> Reviewed-by:
Andrey Grodzovsky <andrey.grodzovsky@amd.com> Reviewed-by:
-
- 22 Feb, 2022 3 commits
-
-
Arunpravin authored
On contiguous allocation, we round up the size to the *next* power of 2, implement a function to free the unused pages after the newly allocate block. v2(Matthew Auld): - replace function name 'drm_buddy_free_unused_pages' with drm_buddy_block_trim - replace input argument name 'actual_size' with 'new_size' - add more validation checks for input arguments - add overlaps check to avoid needless searching and splitting - merged the below patch to see the feature in action - add free unused pages support to i915 driver - lock drm_buddy_block_trim() function as it calls mark_free/mark_split are all globally visible v3(Matthew Auld): - remove trim method error handling as we address the failure case at drm_buddy_block_trim() function v4: - in case of trim, at __alloc_range() split_block failure path marks the block as free and removes it from the original list, potentially also freeing it, to overcome this problem, we turn the drm_buddy_block_trim() input node into a temporary node to prevent recursively freeing itself, but still retain the un-splitting/freeing of the other nodes(Matthew Auld) - modify the drm_buddy_block_trim() function return type v5(Matthew Auld): - revert drm_buddy_block_trim() function return type changes in v4 - modify drm_buddy_block_trim() passing argument n_pages to original_size as n_pages has already been rounded up to the next power-of-two and passing n_pages results noop v6: - fix warnings reported by kernel test robot <lkp@intel.com> v7: - modify drm_buddy_block_trim() function doc description - at drm_buddy_block_trim() handle non-allocated block as a serious programmer error - fix a typo Signed-off-by: -
Arunpravin authored
Implemented a function which walk through the order list, compares the offset and returns the maximum offset block, this method is unpredictable in obtaining the high range address blocks which depends on allocation and deallocation. for instance, if driver requests address at a low specific range, allocator traverses from the root block and splits the larger blocks until it reaches the specific block and in the process of splitting, lower orders in the freelist are occupied with low range address blocks and for the subsequent TOPDOWN memory request we may return the low range blocks.To overcome this issue, we may go with the below approach. The other approach, sorting each order list entries in ascending order and compares the last entry of each order list in the freelist and return the max block. This creates sorting overhead on every drm_buddy_free() request and split up of larger blocks for a single page request. v2: - Fix alignment issues(Matthew Auld) - Remove unnecessary list_empty check(Matthew Auld) - merged the below patch to see the feature in action - add top-down alloc support to i915 driver Signed-off-by: -
Arunpravin authored
- Make drm_buddy_alloc a single function to handle range allocation and non-range allocation demands - Implemented a new function alloc_range() which allocates the requested power-of-two block comply with range limitations - Moved order computation and memory alignment logic from i915 driver to drm buddy v2: merged below changes to keep the build unbroken - drm_buddy_alloc_range() becomes obsolete and may be removed - enable ttm range allocation (fpfn / lpfn) support in i915 driver - apply enhanced drm_buddy_alloc() function to i915 driver v3(Matthew Auld): - Fix alignment issues and remove unnecessary list_empty check - add more validation checks for input arguments - make alloc_range() block allocations as bottom-up - optimize order computation logic - replace uint64_t with u64, which is preferred in the kernel v4(Matthew Auld): - keep drm_buddy_alloc_range() function implementation for generic actual range allocations - keep alloc_range() implementation for end bias allocations v5(Matthew Auld): - modify drm_buddy_alloc() passing argument place->lpfn to lpfn as place->lpfn will currently always be zero for i915 v6(Matthew Auld): - fixup potential uaf - If we are unlucky and can't allocate enough memory when splitting blocks, where we temporarily end up with the given block and its buddy on the respective free list, then we need to ensure we delete both blocks, and no just the buddy, before potentially freeing them - fix warnings reported by kernel test robot <lkp@intel.com> v7(Matthew Auld): - revert fixup potential uaf - keep __alloc_range() add node to the list logic same as drm_buddy_alloc_blocks() by having a temporary list variable - at drm_buddy_alloc_blocks() keep i915 range_overflows macro and add a new check for end variable v8: - fix warnings reported by kernel test robot <lkp@intel.com> v9(Matthew Auld): - remove DRM_BUDDY_RANGE_ALLOCATION flag - remove unnecessary function description v10: - keep DRM_BUDDY_RANGE_ALLOCATION flag as removing the flag and replacing with (end < size) logic fails amdgpu driver load Signed-off-by:
-
- 21 Feb, 2022 1 commit
-
-
Ville Syrjälä authored
struct drm_display_mode embeds a list head, so overwriting the full struct with another one will corrupt the list (if the destination mode is on a list). Use drm_mode_copy() instead which explicitly preserves the list head of the destination mode. Even if we know the destination mode is not on any list using drm_mode_copy() seems decent as it sets a good example. Bad examples of not using it might eventually get copied into code where preserving the list head actually matters. Obviously one case not covered here is when the mode itself is embedded in a larger structure and the whole structure is copied. But if we are careful when copying into modes embedded in structures I think we can be a little more reassured that bogus list heads haven't been propagated in. @is_mode_copy@ @@ drm_mode_copy(...) { ... } @depends on !is_mode_copy@ struct drm_display_mode *mode; expression E, S; @@ ( - *mode = E + drm_mode_copy(mode, &E) | - memcpy(mode, E, S) + drm_mode_copy(mode, E) ) @depends on !is_mode_copy@ struct drm_display_mode mode; expression E; @@ ( - mode = E + drm_mode_copy(&mode, &E) | - memcpy(&mode, E, S) + drm_mode_copy(&mode, E) ) @@ struct drm_display_mode *mode; @@ - &*mode + mode Cc: Emma Anholt <emma@anholt.net> Cc: Maxime Ripard <mripard@kernel.org> Signed-off-by:Ville Syrjälä <ville.syrjala@linux.intel.com> Signed-off-by:
-
- 20 Feb, 2022 3 commits
-
-
Linus Walleij authored
Add some code and config to also support the ACX424AKM used in some Sony (Ericsson) Mobile phones. Signed-off-by:
-
Linus Walleij authored
These IDs were found in the wild in a Sony Xperia vendor tree. Signed-off-by:
-
Linus Walleij authored
A code drop from Sony Mobile reveals that the ACX424 panels are built around the Novatek NT35560 panel controllers so just bite the bullet and rename the driver and all basic symbols so that we can modify this driver to cover any other panels also using the Novatek NT35560 display controller. Signed-off-by:
-
- 16 Feb, 2022 3 commits
-
-
Sankeerth Billakanti authored
Add support for the 14" sharp,lq140m1jw46 eDP panel. Signed-off-by:
Sankeerth Billakanti <quic_sbillaka@quicinc.com> Reviewed-by:
Douglas Anderson <dianders@chromium.org> Signed-off-by:
-
Sankeerth Billakanti authored
Add support for sharp LQ140M1JW46 display panel. It is a 14" eDP panel with 1920x1080 display resolution. Signed-off-by:
Stephen Boyd <swboyd@chromium.org> Reviewed-by:
-
Thomas Zimmermann authored
Fbdev's deferred I/O sorts all dirty pages by default, which incurs a significant overhead. Make the sorting step optional and update the few drivers that require it. Use a FIFO list by default. Most fbdev drivers with deferred I/O build a bounding rectangle around the dirty pages or simply flush the whole screen. The only two affected DRM drivers, generic fbdev and vmwgfx, both use a bounding rectangle. In those cases, the exact order of the pages doesn't matter. The other drivers look at the page index or handle pages one-by-one. The patch sets the sort_pagelist flag for those, even though some of them would probably work correctly without sorting. Driver maintainers should update their driver accordingly. Sorting pages by memory offset for deferred I/O performs an implicit bubble-sort step on the list of dirty pages. The algorithm goes through the list of dirty pages and inserts each new page according to its index field. Even worse, list traversal always starts at the first entry. As video memory is most likely updated scanline by scanline, the algorithm traverses through the complete list for each updated page. For example, with 1024x768x32bpp each page covers exactly one scanline. Writing a single screen update from top to bottom requires updating 768 pages. With an average list length of 384 entries, a screen update creates (768 * 384 =) 294912 compare operation. Fix this by making the sorting step opt-in and update the few drivers that require it. All other drivers work with unsorted page lists. Pages are appended to the list. Therefore, in the common case of writing the framebuffer top to bottom, pages are still sorted by offset, which may have a positive effect on performance. Playing a video [1] in mplayer's benchmark mode shows the difference (i7-4790, FullHD, simpledrm, kernel with debugging). mplayer -benchmark -nosound -vo fbdev ./big_buck_bunny_720p_stereo.ogg With sorted page lists: BENCHMARKs: VC: 32.960s VO: 73.068s A: 0.000s Sys: 2.413s = 108.441s BENCHMARK%: VC: 30.3947% VO: 67.3802% A: 0.0000% Sys: 2.2251% = 100.0000% With unsorted page lists: BENCHMARKs: VC: 31.005s VO: 42.889s A: 0.000s Sys: 2.256s = 76.150s BENCHMARK%: VC: 40.7156% VO: 56.3219% A: 0.0000% Sys: 2.9625% = 100.0000% VC shows the overhead of video decoding, VO shows the overhead of the video output. Using unsorted page lists reduces the benchmark's run time by ~32s/~25%. v2: * Make sorted pagelists the special case (Sam) * Comment on drivers' use of pagelist (Sam) * Warn about the overhead in comment Signed-off-by:
Andy Shevchenko <andriy.shevchenko@linux.intel.com> Link: https://download.blender.org/peach/bigbuckbunny_movies/big_buck_bunny_720p_stereo.ogg # [1] Link: https://patchwork.freedesktop.org/patch/msgid/20220211094640.21632-3-tzimmermann@suse.de
-
