aboutsummaryrefslogtreecommitdiffstats
path: root/drivers/gpu/drm/i915/display/intel_frontbuffer.c
Commit message (Collapse)AuthorAgeFilesLines
* drm/i915/frontbuffer: convert intel_frontbuffer.[ch] to struct intel_displayJani Nikula2025-04-171-32/+28
| | | | | | | | | Going forward, struct intel_display is the main display device data pointer. Convert intel_frontbuffer.[ch] to struct intel_display. Reviewed-by: Chaitanya Kumar Borah <[email protected]> Link: https://lore.kernel.org/r/ef0860583b7d6ad141959f84c25657e0c102d6d2.1744222449.git.jani.nikula@intel.com Signed-off-by: Jani Nikula <[email protected]>
* drm/i915: relocate intel_plane_ggtt_offset() to intel_atomic_plane.cJani Nikula2025-03-031-0/+1
| | | | | | | | | | | | With the primary goal of removing #include "i915_vma.h" from intel_display_types.h, move intel_plane_ggtt_offset() to a proper function in intel_atomic_plane.c. This reveals tons of implicit dependencies all over the place that we pulled in via i915_vma.h. Fix the fallout. Reviewed-by: Rodrigo Vivi <[email protected]> Signed-off-by: Jani Nikula <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/70ac6d19518f355abf37ac8c4b0f1d18878be28c.1740412806.git.jani.nikula@intel.com
* drm/i915/fbc: convert intel_fbc.[ch] to struct intel_displayJani Nikula2025-02-271-3/+2
| | | | | | | | | | | Going forward, struct intel_display is the main display device data pointer. Convert as much as possible of intel_fbc.[ch] to struct intel_display. In a few places, change the variable declaration order to prefer having display first. Reviewed-by: Suraj Kandpal <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/4dfd7d62e99b75d26563bdf29f541ed2ffbe548e.1740502116.git.jani.nikula@intel.com Signed-off-by: Jani Nikula <[email protected]>
* drm/i915/drrs: convert intel_drrs.[ch] to struct intel_displayJani Nikula2025-02-271-2/+2
| | | | | | | | | | Going forward, struct intel_display is the main display device data pointer. Convert as much as possible of intel_drrs.[ch] to struct intel_display. Reviewed-by: Suraj Kandpal <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/02a5189e349f91e58d3907fdf8d0d3c4c5384f4d.1740502116.git.jani.nikula@intel.com Signed-off-by: Jani Nikula <[email protected]>
* drm/i915/tdf: convert intel_tdf.[ch] to struct intel_displayJani Nikula2025-02-271-1/+1
| | | | | | | | | Going forward, struct intel_display is the main display device data pointer. Convert the intel_tdf.[ch] glue to struct intel_display. Reviewed-by: Suraj Kandpal <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/26d976f23295713f9a7cda20e32b7ef5aad3dd9e.1740502116.git.jani.nikula@intel.com Signed-off-by: Jani Nikula <[email protected]>
* drm/i915/display: fix typos in i915/display filesNitin Gote2025-01-231-1/+1
| | | | | | | | | | | | | | | | | Fix all typos in files under drm/i915/display reported by codespell tool. v2: - Include british and american spelling, as those are not typos. - Fix commenting style. <Jani> v3: Fix "In case" wrongly capitalized and also fix comment style. <Krzysztof Niemiec> Signed-off-by: Nitin Gote <[email protected]> Reviewed-by: Krzysztof Niemiec <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected] Signed-off-by: Rodrigo Vivi <[email protected]>
* drm/i915/frontbuffer: convert frontbuffer->obj to struct drm_gem_objectJani Nikula2024-09-191-25/+31
| | | | | | | | | | | | | Prefer the driver agnostic struct drm_gem_object over i915 specific struct drm_i915_gem_object. Do some opportunistic struct intel_display conversions while at it, because it's more convenient to deal with. Reviewed-by: Maarten Lankhorst <[email protected]> Acked-by: Rodrigo Vivi <[email protected]> Signed-off-by: Jani Nikula <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/51bdb3c9b798e28bd70c259fc3874d80bc9b7443.1726589119.git.jani.nikula@intel.com
* drm/i915/frontbuffer: convert intel_frontbuffer_get() to struct drm_gem_objectJani Nikula2024-09-191-6/+6
| | | | | | | | | | Prefer the driver agnostic struct drm_gem_object over i915 specific struct drm_i915_gem_object. Reviewed-by: Maarten Lankhorst <[email protected]> Acked-by: Rodrigo Vivi <[email protected]> Signed-off-by: Jani Nikula <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/358cfcb5eb666732cd7ae21e4f63d07837960ec2.1726589119.git.jani.nikula@intel.com
* drm/i915/display: add intel_bo_get/set_frontbuffer() and use themJani Nikula2024-09-191-3/+4
| | | | | | | | | Add the struct drm_gem_object based interfaces. Reviewed-by: Maarten Lankhorst <[email protected]> Acked-by: Rodrigo Vivi <[email protected]> Signed-off-by: Jani Nikula <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/3120ec430656f04701077dda39cce5f1ed415eee.1726589119.git.jani.nikula@intel.com
* drm/i915/display: convert intel_display_trace.h to struct intel_displayJani Nikula2024-09-051-2/+2
| | | | | | | | | | Going forward, struct intel_display shall replace struct drm_i915_private as the main display device data pointer type. Convert intel_display_trace.h to struct intel_display. Reviewed-by: Ville Syrjälä <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected] Signed-off-by: Jani Nikula <[email protected]>
* drm/i915/fb: hide the guts of intel_fb_obj()Jani Nikula2024-09-051-0/+1
| | | | | | | | | Use a proper function in intel_fb.[ch] for intel_fb_obj() to be able to drop the gem/i915_gem_object_types.h from intel_display_types.h. Reviewed-by: Luca Coelho <[email protected]> Signed-off-by: Jani Nikula <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/c5ff0d355911903809ba366403192243c05d3427.1724689818.git.jani.nikula@intel.com
* drm/i915/psr: convert intel_psr.[ch] to struct intel_displayJani Nikula2024-09-031-2/+5
| | | | | | | | | | | | Going forward, struct intel_display shall replace struct drm_i915_private as the main display device data pointer type. Convert intel_psr.[ch] to struct intel_display. Some stragglers are left behind where needed. Reviewed-by: Ville Syrjälä <[email protected]> Signed-off-by: Jani Nikula <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/4399b98b07019a8063adbec1043ff7eabb7c1080.1725012870.git.jani.nikula@intel.com
* drm/i915/display: perform transient flushMatthew Auld2024-05-031-0/+2
| | | | | | | | | | | | | | Perform manual transient cache flush prior to flip and at the end of frontbuffer_flush. This is needed to ensure display engine doesn't see garbage if the surface is L3:XD dirty. Testcase: igt@xe-pat@display-vs-wb-transient Signed-off-by: Matthew Auld <[email protected]> Signed-off-by: Balasubramani Vivekanandan <[email protected]> Acked-by: Nirmoy Das <[email protected]> Reviewed-by: Matt Roper <[email protected]> Signed-off-by: Radhakrishna Sripada <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
* drm/i915/display: Move releasing gem object away from fb trackingJouni Högander2023-10-231-2/+0
| | | | | | | | | | As a preparation for Xe we want to remove all i915_gem_object details away from frontbuffer tacking code. Due to this move releasing gem object reference to i915_gem_object_set_frontbuffer. Signed-off-by: Jouni Högander <[email protected]> Reviewed-by: Juha-Pekka Heikkila <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
* drm/i915/display: Include i915_active header in frontbuffer tracking codeJouni Högander2023-10-021-0/+1
| | | | | | | | | | | Intel frontbuffer tracking code is using i915_active functions -> include i915_active.h directly instead of relying getting indirectly included. Cc: Jani Nikula <[email protected]> Signed-off-by: Jouni Högander <[email protected]> Reviewed-by: Jani Nikula <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
* drm/i915: Warn on if set frontbuffer return value is not NULL on releaseJouni Högander2023-09-291-2/+3
| | | | | | | | | | | | | | | | i915_gem_object_set_frontbuffer returns set frontbuffer pointer. When we are releasing frontbuffer we are clearing the pointer from the object. Warn on if return value is not null. v3: Check i915_gem_object_set_frontbuffer return value separately v2: Instead of ignoring do drm_WARN_ON Cc: Rodrigo Vivi <[email protected]> Cc: Jani Nikula <[email protected]> Signed-off-by: Jouni Högander <[email protected]> Reviewed-by: Jani Nikula <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
* drm/i915: Add new frontbuffer tracking interface to queue flushJouni Högander2023-09-041-0/+28
| | | | | | | | | | | | | | | | | We want to wait dma fences in dirtyfb ioctl. As we don't want to make dirtyfb ioctl as blocking call we need to use dma_fence_add_callback. Callback used for dma_fence_add_callback is called from atomic context. Due to this we need to add a new frontbuffer tracking interface to queue flush. v3: - Check schedule work success rather than work being pending - Init flush work when frontbuffer struct is initialized v2: Check if flush work is already pending Signed-off-by: Jouni Högander <[email protected]> Reviewed-by: Ville Syrjälä <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
* drm/i915: add minimal i915_gem_object_frontbuffer.hJani Nikula2023-08-311-0/+1
| | | | | | | | | | | | | | | | | | | Split out frontbuffer related declarations and static inlines from gem/i915_gem_object.h into new gem/i915_gem_object_frontbuffer.h. The main goal is to reduce header interdependencies. With gem/i915_gem_object.h including display/intel_frontbuffer.h, modification of the latter causes a whopping 300+ objects to be rebuilt, while many of the source files actually needing it aren't explicitly including it at all. After the change, only 21 objects depend on display/intel_frontbuffer.h, directly or indirectly. Cc: Jouni Högander <[email protected]> Reviewed-by: Jouni Högander <[email protected]> Signed-off-by: Jani Nikula <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
* drm/i915: Add function to clear scanout flag for vmasJouni Högander2023-08-011-7/+1
| | | | | | | | | | | | | Currently frontbuffer tracking code is directly iterating over object vmas and clearing scanout flags for them. Add function to clear scanout flag for vmas and use it from frontbuffer tracking code. v2: describe function parameter. Signed-off-by: Jouni Högander <[email protected]> Reviewed-by: Jani Nikula <[email protected]> Reviewed-by: Nirmoy Das <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
* drm/i915: Add getter/setter for i915_gem_object->frontbufferJouni Högander2023-08-011-13/+7
| | | | | | | | | | | | | | Add getter/setter for i915_gem_object->frontbuffer and use it instead of directly touching i915_gem_object->frontbuffer frontbuffer pointer. v3: - Fix intel_frontbuffer_get return value - s/front_ret/cur/ v2: Move getter/setter into i915_gem_object.h Signed-off-by: Jouni Högander <[email protected]> Reviewed-by: Nirmoy Das <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
* drm/i915: Add macros to get i915 device from i915_gem_objectJouni Högander2023-08-011-9/+9
| | | | | | | | | | | | | We want to stop touching directly i915_gem_object struct members in intel_frontbuffer code. As a part of this we add helper macro to get i915 device from i915_gem_object. v2: operate on and return pointer in defined macros Signed-off-by: Jouni Högander <[email protected]> Reviewed-by: Jani Nikula <[email protected]> Reviewed-by: Nirmoy Das <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
* drm/i915: Pass i915 to frontbuffer tracepointsVille Syrjälä2022-11-171-2/+2
| | | | | | | | | Pass the device to the frontbuffer tracpoints. Will be used later to include the device name in the tracpoints. Signed-off-by: Ville Syrjälä <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected] Acked-by: Jani Nikula <[email protected]>
* drm/i915: move fb_tracking under display sub-structJani Nikula2022-08-311-27/+27
| | | | | | | | | | | | | | | Move display frontbuffer tracking related members under drm_i915_private display sub-struct. Rename struct i915_frontbuffer_tracking to intel_frontbuffer_tracking while at it. FIXME: fb_tracking.lock mutex init should be moved away from i915_gem_init_early(). Signed-off-by: Jani Nikula <[email protected]> Reviewed-by: Lucas De Marchi <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/a5444d0a373afca46da9a2f6e4db442af21b429b.1661779055.git.jani.nikula@intel.com
* drm/i915: move INTEL_FRONTBUFFER_* macros to intel_frontbuffer.hJani Nikula2022-08-291-0/+2
| | | | | | | | | | | | The macros clearly don't belong in i915_drv.h. Move to intel_frontbuffer.h. Also split the BUILD_BUG_ON()s to intel_frontbuffer_track() to avoid depending on some other macros in the header. Signed-off-by: Jani Nikula <[email protected]> Reviewed-by: Lucas De Marchi <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/c899fd1ef2220564a876cd88c03e53c4c7b0168b.1661346845.git.jani.nikula@intel.com
* drm/i915/trace: split out display trace to a separate fileJani Nikula2021-12-091-4/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | Add display/intel_display_trace.[ch] for defining display tracepoints. The main goal is to reduce cross-includes between gem and display. It would be possible split up tracing even further, but that would lead to more boilerplate. We end up having to include intel_crtc.h in a few places because it was pulled in implicitly via intel_de.h -> i915_trace.h -> intel_crtc.h, and that's no longer the case. There should be no changes to tracepoints. v3: - Rebase v2: - Define TRACE_INCLUDE_PATH relative to define_trace.h (Chris) - Remove useless comments (Ville) Cc: Ville Syrjälä <[email protected]> Signed-off-by: Jani Nikula <[email protected]> Reviewed-by: Ville Syrjälä <[email protected]> Reviewed-by: Lucas De Marchi <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/7862ad764fbd0748d903c76bc632d3d277874e5b.1638961423.git.jani.nikula@intel.com
* drm/i915/display: Renaming DRRS functions to intel_drrs_*()José Roberto de Souza2021-08-301-2/+2
| | | | | | | | | | | | | | | | | | | We had a mix of intel_edp_drrs_*(), intel_dp_drrs_*() and intel_dp_set_drrs_state(), so properly renaming all functions to keep the same pattern. While at it, also dropping intel_dp_set_drrs_state from the documentation as it is a static function. v3: - dropping documentation style comment in static function (intel_drrs_set_state) Reviewed-by: Rodrigo Vivi <[email protected]> Cc: Jani Nikula <[email protected]> Cc: Rodrigo Vivi <[email protected]> Signed-off-by: José Roberto de Souza <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
* drm/i915/display: Move DRRS code its own fileJosé Roberto de Souza2021-08-301-0/+1
| | | | | | | | | | | intel_dp.c is a 5k lines monster, so moving DRRS out of it to reduce some lines from it. Reviewed-by: Rodrigo Vivi <[email protected]> Cc: Jani Nikula <[email protected]> Cc: Rodrigo Vivi <[email protected]> Signed-off-by: José Roberto de Souza <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
* Merge tag 'drm-intel-gt-next-2021-05-28' of ↵Dave Airlie2021-06-021-2/+2
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://anongit.freedesktop.org/drm/drm-intel into drm-next UAPI Changes: - Add reworked uAPI for DG1 behind CONFIG_BROKEN (Matt A, Abdiel) Driver Changes: - Fix for Gitlab issues #3293 and #3450: Avoid kernel crash on older L-shape memory machines - Add Wa_14010733141 (VDBox SFC reset) for Gen11+ (Aditya) - Fix crash in auto_retire active retire callback due to misalignment (Stephane) - Fix overlay active retire callback alignment (Tvrtko) - Eliminate need to align active retire callbacks (Matt A, Ville, Daniel) - Program FF_MODE2 tuning value for all Gen12 platforms (Caz) - Add Wa_14011060649 for TGL,RKL,DG1 and ADLS (Swathi) - Create stolen memory region from local memory on DG1 (CQ) - Place PD in LMEM on dGFX (Matt A) - Use WC when default state object is allocated in LMEM (Venkata) - Determine the coherent map type based on object location (Venkata) - Use lmem physical addresses for fb_mmap() on discrete (Mohammed) - Bypass aperture on fbdev when LMEM is available (Anusha) - Return error value when displayable BO not in LMEM for dGFX (Mohammed) - Do release kernel context if breadcrumb measure fails (Janusz) - Hide modparams for compiled-out features (Tvrtko) - Apply Wa_22010271021 for all Gen11 platforms (Caz) - Fix unlikely ref count race in arming the watchdog timer (Tvrtko) - Check actual RC6 enable status in PMU (Tvrtko) - Fix a double free in gen8_preallocate_top_level_pdp (Lv) - Use trylock in shrinker for GGTT on BSW VT-d and BXT (Maarten) - Remove erroneous i915_is_ggtt check for I915_GEM_OBJECT_UNBIND_VM_TRYLOCK (Maarten) - Convert uAPI headers to real kerneldoc (Matt A) - Clean up kerneldoc warnings headers (Matt A, Maarten) - Fail driver if LMEM training failed (Matt R) - Avoid div-by-zero on Gen2 (Ville) - Read C0DRB3/C1DRB3 as 16 bits again and add _BW suffix (Ville) - Remove reference to struct drm_device.pdev (Thomas) - Increase separation between GuC and execlists code (Chris, Matt B) - Use might_alloc() (Bernard) - Split DGFX_FEATURES from GEN12_FEATURES (Lucas) - Deduplicate Wa_22010271021 programming on (Jose) - Drop duplicate WaDisable4x2SubspanOptimization:hsw (Tvrtko) - Selftest improvements (Chris, Hsin-Yi, Tvrtko) - Shuffle around init_memory_region for stolen (Matt) - Typo fixes (wengjianfeng) [airlied: fix conflict with fixes in i915_active.c] Signed-off-by: Dave Airlie <[email protected]> From: Joonas Lahtinen <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/YLCbBR22BsQ/[email protected]
| * drm/i915: drop the __i915_active_call pointer packingMatthew Auld2021-05-051-2/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We use some of the lower bits of the retire function pointer for potential flags, which is quite thorny, since the caller needs to remember to give the function the correct alignment with __i915_active_call, otherwise we might incorrectly unpack the pointer and jump to some garbage address later. Instead of all this let's just pass the flags along as a separate parameter. Suggested-by: Ville Syrjälä <[email protected]> Suggested-by: Daniel Vetter <[email protected]> References: ca419f407b43 ("drm/i915: Fix crash in auto_retire") References: d8e44e4dd221 ("drm/i915/overlay: Fix active retire callback alignment") References: fd5f262db118 ("drm/i915/selftests: Fix active retire callback alignment") Signed-off-by: Matthew Auld <[email protected]> Reviewed-by: Matthew Brost <[email protected]> Acked-by: Daniel Vetter <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
* | drm/i915: Add frontbuffer tracking tracepointsVille Syrjälä2021-04-281-0/+5
|/ | | | | | | | | Add some tracpoints for frontbuffer tracking so we can try to figure out what's going on. Signed-off-by: Ville Syrjälä <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected] Acked-by: Jani Nikula <[email protected]>
* Merge drm/drm-next into drm-intel-nextJani Nikula2021-03-111-1/+3
|\ | | | | | | | | | | Sync up with upstream. Signed-off-by: Jani Nikula <[email protected]>
| * drm/i915/gem: Protect used framebuffers from casual evictionChris Wilson2021-01-201-1/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In the shrinker, we protect framebuffers from light reclaim as we typically expect framebuffers to be reused in the near future (and with low latency requirements). We can apply the same logic to the GGTT eviction and defer framebuffers to the second pass only used if the caller is desperate enough to wait for space to become available. In most cases, the caller will use a smaller partial vma instead of trying to force the object into the GGTT if doing so will cause other users to be evicted. Signed-off-by: Chris Wilson <[email protected]> Reviewed-by: Matthew Auld <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
* | drm/i915: Warn when releasing a frontbuffer while in useVille Syrjälä2021-02-091-0/+2
|/ | | | | | | | | Let's scream if we are about to release a frontbuffer which is still in use. Signed-off-by: Ville Syrjälä <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected] Reviewed-by: Chris Wilson <[email protected]>
* drm/i915: Add a couple of missing i915_active_fini()Chris Wilson2020-08-171-0/+2
| | | | | | | | | | | | We use i915_active_fini() as a debug check on the i915_active state before freeing. If we forget to call it, we may end up angering the debugobjects contained within. Signed-off-by: Chris Wilson <[email protected]> Reviewed-by: Tvrtko Ursulin <[email protected]> Reviewed-by: Thomas Hellström <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected] Signed-off-by: Rodrigo Vivi <[email protected]>
* drm/i915/display/frontbuffer: Prefer drm_WARN_ON over WARN_ONPankaj Bharadiya2020-04-211-2/+4
| | | | | | | | | | | struct drm_device specific drm_WARN* macros include device information in the backtrace, so we know what device the warnings originate from. Prefer drm_WARN_ON over WARN_ON. Signed-off-by: Pankaj Bharadiya <[email protected]> Signed-off-by: Jani Nikula <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
* drm/i915: Hold reference to intel_frontbuffer as we track activityChris Wilson2019-12-181-10/+6
| | | | | | | | | | | | | | | Since obj->frontbuffer is no longer protected by the struct_mutex, as we are processing the execbuf, it may be removed. Mark the intel_frontbuffer as rcu protected, and so acquire a reference to the struct as we track activity upon it. Closes: https://gitlab.freedesktop.org/drm/intel/issues/827 Fixes: 8e7cb1799b4f ("drm/i915: Extract intel_frontbuffer active tracking") Signed-off-by: Chris Wilson <[email protected]> Cc: Matthew Auld <[email protected]> Cc: <[email protected]> # v5.4+ Reviewed-by: Joonas Lahtinen <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
* drm/i915: Coordinate i915_active with its own mutexChris Wilson2019-10-041-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | Forgo the struct_mutex serialisation for i915_active, and interpose its own mutex handling for active/retire. This is a multi-layered sleight-of-hand. First, we had to ensure that no active/retire callbacks accidentally inverted the mutex ordering rules, nor assumed that they were themselves serialised by struct_mutex. More challenging though, is the rule over updating elements of the active rbtree. Instead of the whole i915_active now being serialised by struct_mutex, allocations/rotations of the tree are serialised by the i915_active.mutex and individual nodes are serialised by the caller using the i915_timeline.mutex (we need to use nested spinlocks to interact with the dma_fence callback lists). The pain point here is that instead of a single mutex around execbuf, we now have to take a mutex for active tracker (one for each vma, context, etc) and a couple of spinlocks for each fence update. The improvement in fine grained locking allowing for multiple concurrent clients (eventually!) should be worth it in typical loads. v2: Add some comments that barely elucidate anything :( Signed-off-by: Chris Wilson <[email protected]> Reviewed-by: Tvrtko Ursulin <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
* drm/i915: Push the i915_active.retire into a workerChris Wilson2019-10-041-1/+3
| | | | | | | | | | | As we need to use a mutex to serialise i915_active activation (because we want to allow the callback to sleep), we need to push the i915_active.retire into a worker callback in case we get need to retire from an atomic context. Signed-off-by: Chris Wilson <[email protected]> Reviewed-by: Matthew Auld <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
* drm/i915: Replace obj->pin_global with obj->frontbufferChris Wilson2019-09-031-3/+10
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | obj->pin_global was originally used as a means to keep the shrinker off the active scanout, but we use the vma->pin_count itself for that and the obj->frontbuffer to delay shrinking active framebuffers. The other role that obj->pin_global gained was for spotting display objects inside GEM and working harder to keep those coherent; for which we can again simply inspect obj->frontbuffer directly. Coming up next, we will want to manipulate the pin_global counter outside of the principle locks, so would need to make pin_global atomic. However, since obj->frontbuffer is already managed atomically, it makes sense to use that the primary key for display objects instead of having pin_global. Ville pointed out the principle difference is that obj->frontbuffer is set for as long as an intel_framebuffer is attached to an object, but obj->pin_global was only raised for as long as the object was active. In practice, this means that we consider the object as being on the scanout for longer than is strictly required, causing us to be more proactive in flushing -- though it should be true that we would have flushed eventually when the back became the front, except that on the flip path that flush is async but when hit from another ioctl it will be synchronous. v2: i915_gem_object_is_framebuffer() Signed-off-by: Chris Wilson <[email protected]> Cc: Mika Kuoppala <[email protected]> Cc: Ville Syrjälä <[email protected]> Reviewed-by: Ville Syrjälä <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
* drm/i915: Extract intel_frontbuffer active trackingChris Wilson2019-08-161-73/+182
| | | | | | | | | | | | | Move the active tracking for the frontbuffer operations out of the i915_gem_object and into its own first class (refcounted) object. In the process of detangling, we switch from low level request tracking to the easier i915_active -- with the plan that this avoids any potential atomic callbacks as the frontbuffer tracking wishes to sleep as it flushes. Signed-off-by: Chris Wilson <[email protected]> Reviewed-by: Matthew Auld <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
* drm/i915: rename intel_drv.h to display/intel_display_types.hJani Nikula2019-08-071-1/+1
| | | | | | | | | | | | | | | Everything about the file is about display, and mostly about types related to display. Move under display/ as intel_display_types.h to reflect the facts. There's still plenty to clean up, but start off with moving the file where it logically belongs and naming according to contents. v2: fix the include guard name in the renamed file Reviewed-by: Chris Wilson <[email protected]> Signed-off-by: Jani Nikula <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
* drm/i915: move modesetting core code under display/Jani Nikula2019-06-171-0/+199
Now that we have a new subdirectory for display code, continue by moving modesetting core code. display/intel_frontbuffer.h sticks out like a sore thumb, otherwise this is, again, a surprisingly clean operation. v2: - don't move intel_sideband.[ch] (Ville) - use tabs for Makefile file lists and sort them Cc: Chris Wilson <[email protected]> Cc: Joonas Lahtinen <[email protected]> Cc: Rodrigo Vivi <[email protected]> Cc: Ville Syrjälä <[email protected]> Reviewed-by: Chris Wilson <[email protected]> Acked-by: Rodrigo Vivi <[email protected]> Acked-by: Maarten Lankhorst <[email protected]> Acked-by: Ville Syrjälä <[email protected]> Signed-off-by: Jani Nikula <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
generated by cgit (git 2.51.2) at 2025-12-01 04:41:35 +0000