aboutsummaryrefslogtreecommitdiff
path: root/drivers/gpu
AgeCommit message (Collapse)AuthorFilesLines
2021-10-15drm/msm: unlock on error in get_sched_entity()Dan Carpenter1-0/+1
Add a missing unlock on the error path if drm_sched_entity_init() fails. Fixes: 68002469e571 ("drm/msm: One sched entity per process per priority") Signed-off-by: Dan Carpenter <[email protected]> Link: https://lore.kernel.org/r/20211011124005.GE15188@kili Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm: Remove redundant 'flush_workqueue()' callsChristophe JAILLET6-10/+2
'destroy_workqueue()' already drains the queue before destroying it, so there is no need to flush it explicitly. Remove the redundant 'flush_workqueue()' calls. This was generated with coccinelle: @@ expression E; @@ - flush_workqueue(E); destroy_workqueue(E); Signed-off-by: Christophe JAILLET <[email protected]> Acked-by: Jyri Sarha <[email protected]> Reviewed-by: Rob Clark <[email protected]> Link: https://lore.kernel.org/r/75e8ba40076ad707d47e3a3670e6b23c1b8b11bc.1633874223.git.christophe.jaillet@wanadoo.fr Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm/dp: Simplify the dp_debug debugfs show functionBjorn Andersson1-138/+25
The "dp_debug" show function allocates a buffer and piecemeal appends line by line, checking for buffer overflows etc. Migrate the function to seq_file, to remove all the extra book keeping and simplify the function. Reviewed-by: Stephen Boyd <[email protected]> Signed-off-by: Bjorn Andersson <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm/dp: Use the connector passed to dp_debug_get()Bjorn Andersson3-90/+47
The debugfs code is provided an array of a single drm_connector. Then to access the connector, the list of all connectors of the DRM device is traversed and all non-DisplayPort connectors are skipped, to find the one and only DisplayPort connector. But as we move to support multiple DisplayPort controllers this will now find multiple connectors and has no way to distinguish them. Pass the single connector to dp_debug_get() and use this in the debugfs functions instead, both to simplify the code and the support the multiple instances. Reviewed-by: Abhinav Kumar <[email protected]> Reviewed-by: Stephen Boyd <[email protected]> Signed-off-by: Bjorn Andersson <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm/a6xx: correct cx_debugbus_read argumentsDmitry Baryshkov1-1/+1
First argument of cx_debugbus_read() should be 'void __iomem *' rather than 'void * __iomem' to make sparse happy. Signed-off-by: Dmitry Baryshkov <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm/dsi: use bulk clk APIDmitry Baryshkov1-47/+12
Use clk_bulk_* API instead of hand-coding them. Note, this drops support for legacy clk naming (e.g. "iface_clk" instead of just "iface"), however all in-kernel device trees were converted long long ago. The warning is present there since 2017. Signed-off-by: Dmitry Baryshkov <[email protected]> Reviewed-by: Bjorn Andersson <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm/devfreq: Add 1ms delay before clamping freqRob Clark2-8/+37
Add a short delay before clamping to idle frequency on active->idle transition. It takes ~0.5ms to increase the freq again on the next idle->active transition, so this helps avoid extra freq transitions on workloads that bounce between CPU and GPU. Signed-off-by: Rob Clark <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm: Add hrtimer + kthread_work helperRob Clark4-17/+58
Before open-coding this a 2nd time, add a helper. Signed-off-by: Rob Clark <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm/dp: Allow sub-regions to be specified in DTBjorn Andersson1-11/+38
Not all platforms has P0 at an offset of 0x1000 from the base address, so add support for specifying each sub-region in DT. The code falls back to the predefined offsets in the case that only a single reg is specified, in order to support existing DT. Reviewed-by: Stephen Boyd <[email protected]> Reviewed-by: Abhinav Kumar <[email protected]> Signed-off-by: Bjorn Andersson <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm/dp: Store each subblock in the io regionBjorn Andersson3-48/+53
Not all platforms has DP_P0 at offset 0x1000 from the beginning of the DP block. So split the dss_io_data memory region into a set of sub-regions, to make it possible in the next patch to specify each of the sub-regions individually. Reviewed-by: Stephen Boyd <[email protected]> Signed-off-by: Bjorn Andersson <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm/dp: Refactor ioremap wrapperBjorn Andersson2-25/+12
In order to deal with multiple memory ranges in the following commit change the ioremap wrapper to not poke directly into the dss_io_data struct. While at it, devm_ioremap_resource() already prints useful error messages on failure, so omit the unnecessary prints from the caller. Signed-off-by: Bjorn Andersson <[email protected]> Reviewed-by: Stephen Boyd <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm/dp: Use devres for ioremap()Bjorn Andersson1-25/+4
The non-devres version of ioremap is used, which requires manual cleanup. But the code paths leading here is mixed with other devres users, so rely on this for ioremap as well to simplify the code. Reviewed-by: Abhinav Kumar <[email protected]> Reviewed-by: Stephen Boyd <[email protected]> Signed-off-by: Bjorn Andersson <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm: potential error pointer dereference in init()Dan Carpenter1-0/+4
The msm_iommu_new() returns error pointers on failure so check for that to avoid an Oops. Fixes: ccac7ce373c1 ("drm/msm: Refactor address space initialization") Signed-off-by: Dan Carpenter <[email protected]> Reviewed-by: Abhinav Kumar <[email protected]> Reviewed-by: Dmitry Baryshkov <[email protected]> Link: https://lore.kernel.org/r/20211004103806.GD25015@kili Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm: Fix potential Oops in a6xx_gmu_rpmh_init()Dan Carpenter1-3/+3
There are two problems here: 1) The "seqptr" is used uninitalized when we free it at the end. 2) The a6xx_gmu_get_mmio() function returns error pointers. It never returns true. Fixes: 64245fc55172 ("drm/msm/a6xx: use AOP-initialized PDC for a650") Fixes: f8fc924e088e ("drm/msm/a6xx: Fix PDC register overlap") Signed-off-by: Dan Carpenter <[email protected]> Reviewed-by: Dmitry Baryshkov <[email protected]> Link: https://lore.kernel.org/r/20211004134530.GB11689@kili Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm/dsi: do not enable irq handler before powering up the hostDmitry Baryshkov3-17/+49
The DSI host might be left in some state by the bootloader. If this state generates an IRQ, it might hang the system by holding the interrupt line before the driver sets up the DSI host to the known state. Move the request_irq into msm_dsi_host_init and pass IRQF_NO_AUTOEN to it. Call enable/disable_irq after msm_dsi_host_power_on/_off() functions, so that we can be sure that the interrupt is delivered when the host is in the known state. It is not possible to defer the interrupt enablement to a later point, because drm_panel_prepare might need to communicate with the panel over the DSI link and that requires working interrupt. Fixes: a689554ba6ed ("drm/msm: Initial add DSI connector support") Signed-off-by: Dmitry Baryshkov <[email protected]> Reviewed-by: Abhinav Kumar <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm/mdp5: Add configuration for MDP v1.16Vladimir Lypak1-0/+89
MDP version v1.16 is almost identical to v1.15 with most significant difference being presence of second DSI interface. MDP v1.16 is found on SoCs such as MSM8x53, SDM450, SDM632 (All with Adreno 506). Signed-off-by: Vladimir Lypak <[email protected]> Signed-off-by: Sireesh Kodali <[email protected]> Reviewed-by: Dmitry Baryshkov <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm/dsi: Add phy configuration for MSM8953Vladimir Lypak3-0/+24
Add phy configuration for 14nm dsi phy found on MSM8953 SoC. Only difference from existing configurations are io_start addresses. Signed-off-by: Vladimir Lypak <[email protected]> Reviewed-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Sireesh Kodali <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm: Use IS_ERR() instead of IS_ERR_OR_NULL()Wang Qing1-2/+2
dev_pm_opp_find_freq_exact never return null but ERR_PTR(),so we should use IS_ERR() to clarify and avoid compilation warnings. Signed-off-by: Wang Qing <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm: msm: adreno: use DEFINE_DEBUGFS_ATTRIBUTE with debugfs_create_file_unsafe()Deepak R Varma1-3/+3
Using DEFINE_DEBUGFS_ATTRIBUTE macro with debugfs_create_file_unsafe() function in place of the debugfs_create_file() function will make the file operation struct "reset" aware of the file's lifetime. Additional details here: https://lists.archive.carbon60.com/linux/kernel/2369498 Issue reported by Coccinelle script: scripts/coccinelle/api/debugfs/debugfs_simple_attr.cocci Signed-off-by: Deepak R Varma <[email protected]> Link: https://lore.kernel.org/r/9ca2c2e4cbd9ebb282b90f742305fd9b481aacc2.1603998014.git.mh12gx2825@gmail.com Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm: delete conversion from bool value to bool returnBernard1-1/+1
Delete the conversion from bool value to bool function return. Signed-off-by: Bernard Zhao <[email protected]> Link: https://lore.kernel.org/r/AI*[email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm/mdp5: Remove redundant null check before ↵Xu Wang2-20/+10
clk_prepare_enable/clk_disable_unprepare Because clk_prepare_enable() and clk_disable_unprepare() already checked NULL clock parameter, so the additional checks are unnecessary, just remove them. Signed-off-by: Xu Wang <[email protected]> Reviewed-by: Dmitry Baryshkov <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm: dsi: Remove redundant null check before ↵Xu Wang1-9/+6
clk_prepare_enable/clk_disable_unprepare Because clk_prepare_enable() and clk_disable_unprepare() already checked NULL clock parameter, so the additional checks are unnecessary, just remove them. Signed-off-by: Xu Wang <[email protected]> Reviewed-by: Dmitry Baryshkov <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm/mdp4: Remove redundant null check before ↵Xu Wang1-12/+6
clk_prepare_enable/clk_disable_unprepare Because clk_prepare_enable() and clk_disable_unprepare() already checked NULL clock parameter, so the additional checks are unnecessary, just remove them. Signed-off-by: Xu Wang <[email protected]> Reviewed-by: Dmitry Baryshkov <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm: fix warning "using plain integer as NULL pointer"Bernard Zhao1-2/+2
Fix sparse warning: drivers/gpu/drm/msm/disp/dpu1/dpu_plane.c:1195:41: warning: Using plain integer as NULL pointer drivers/gpu/drm/msm/disp/dpu1/dpu_plane.c:1348:32: warning: Using plain integer as NULL pointer Signed-off-by: Bernard Zhao <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm: prevent NULL dereference in msm_gpu_crashstate_capture()Tim Gardner1-1/+1
Coverity complains of a possible NULL dereference: CID 120718 (#1 of 1): Dereference null return value (NULL_RETURNS) 23. dereference: Dereferencing a pointer that might be NULL state->bos when calling msm_gpu_crashstate_get_bo. [show details] 301 msm_gpu_crashstate_get_bo(state, submit->bos[i].obj, 302 submit->bos[i].iova, submit->bos[i].flags); Fix this by employing the same state->bos NULL check as is used in the next for loop. Cc: Rob Clark <[email protected]> Cc: Sean Paul <[email protected]> Cc: David Airlie <[email protected]> Cc: Daniel Vetter <[email protected]> Cc: [email protected] Cc: [email protected] Cc: [email protected] Cc: [email protected] Signed-off-by: Tim Gardner <[email protected]> Reviewed-by: Dmitry Baryshkov <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm/dp: Remove redundant initialization of variable bppColin Ian King1-1/+1
The variable bpp is being initialized with a value that is never read, it is being updated later on in both paths of an if statement. The assignment is redundant and can be removed. Addresses-Coverity: ("Unused value") Signed-off-by: Colin Ian King <[email protected]> Reviewed-by: Dmitry Baryshkov <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm/dpu: Remove some nonsenseRob Clark2-14/+0
These aren't used. And if we add use for them later, we should probably do something a bit more structured than string parsing. Signed-off-by: Rob Clark <[email protected]> Reviewed-by: Dmitry Baryshkov <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm/dsi: Support NO_CONNECTOR bridgesRob Clark2-13/+39
For now, since we have a mix of bridges which support this flag, which which do *not* support this flag, or work both ways, try it once with NO_CONNECTOR and then fall back to the old way if that doesn't work. Eventually we can drop the fallback path. v2: Add missing drm_connector_attach_encoder() so display actually comes up when the bridge properly handles the NO_CONNECTOR flag Signed-off-by: Rob Clark <[email protected]> Reviewed-by: Laurent Pinchart <[email protected]> Reviewed-by: Dmitry Baryshkov <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm: msm: hdmi: Constify static structsRikard Falkeborn1-2/+2
The only usage of hdmi_8996_pll_ops is to assign its address to the ops field in the clk_init_data struct, and the only usage of pll_init is to assign its address to the init field in the clk_hw struct, both which are pointers to const. Make them const to allow the compiler to put them in read-only memory. Signed-off-by: Rikard Falkeborn <[email protected]> Reviewed-by: Dmitry Baryshkov <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm/dsi: Use division result from div_u64_rem in 7nm and 14nm PLLMarijn Suijten2-6/+2
div_u64_rem provides the result of the division and additionally the remainder; don't use this function to solely calculate the remainder while calculating the division again with div_u64. A similar improvement was applied earlier to the 10nm pll in 5c191fef4ce2 ("drm/msm/dsi_pll_10nm: Fix dividing the same numbers twice"). Signed-off-by: Marijn Suijten <[email protected]> Reviewed-By: AngeloGioacchino Del Regno <[email protected]> Reviewed-by: Abhinav Kumar <[email protected]> Reviewed-by: Stephen Boyd <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm: Remove initialization of static variableszhaoxiao1-2/+2
Address the following checkpatch errors: ERROR: do not initialise statics to false FILE: :drivers/gpu/drm/msm/msm_drv.c:21: -static bool reglog = false; FILE: :drivers/gpu/drm/msm/msm_drv.c:31: -bool dumpstate = false; Signed-off-by: zhaoxiao <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm: remove unneeded variableChi Minghao1-3/+1
Fix the following coccicheck REVIEW: ./drivers/gpu/drm/msm/edp/edp_ctrl.c:1245:5-8 Unneeded variable Reported-by: Zeal Robot <[email protected]> Signed-off-by: Chi Minghao <[email protected]> Reviewed-by: Lyude Paul <[email protected]> Reviewed-by: Dmitry Baryshkov <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm: Change dpu_crtc_get_vblank_counter to use vsync count.Mark Yacoub3-19/+9
[why] vsync_cnt atomic counter increments for every hw vsync. On the other hand, frame count is a register that increments when the frame gets actually pushed out. We cannnot read this register whenever the timing engine is off, but vblank counter should still return a valid number. This behavior also matches the downstream driver. [How] Read the encoder vsync count instead of the dpu_encoder_phys frame count. Suggested-by: Abhinav Kumar <[email protected]> CC: Rob Clark <[email protected]> Signed-off-by: Mark Yacoub <[email protected]> Reviewed-by: Abhinav Kumar <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm/dpu: merge struct dpu_irq into struct dpu_hw_intrDmitry Baryshkov3-41/+28
As dpu_core_irq was merged into dpu_hw_intr, merge data structures too, removing the need for a separate data structure. Signed-off-by: Dmitry Baryshkov <[email protected]> Reviewed-by: Abhinav Kumar <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm/dpu: don't clear IRQ register twiceDmitry Baryshkov1-17/+0
We already clear the IRQ status register before processing IRQs, so do not clear the register again. Especially do not clear the IRQ status _after_ processing the IRQ as this way we can loose the event. Signed-off-by: Dmitry Baryshkov <[email protected]> Reviewed-by: Abhinav Kumar <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/msm/dpu: squash dpu_core_irq into dpu_hw_interruptsDmitry Baryshkov4-399/+214
With dpu_core_irq being the wrapper around dpu_hw_interrupts, there is little sense in having them separate. Squash them together to remove another layer of abstraction (hw_intr ops). Signed-off-by: Dmitry Baryshkov <[email protected]> Reviewed-by: Abhinav Kumar <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Signed-off-by: Rob Clark <[email protected]>
2021-10-15drm/i915: Clean up PXP Kconfig info.Rodrigo Vivi1-5/+5
During the review I focused on stop the using of the "+" to reference the newer platforms, but I forgot that we are in a process of making things more clear and differentiate graphics and display versions. So, let me to clean up this a bit. Also, we don't need any version mentioned in the config menu entry, only in the help. Cc: Alan Previn <[email protected]> Cc: Daniele Ceraolo Spurio <[email protected]> Signed-off-by: Rodrigo Vivi <[email protected]> Reviewed-by: Daniele Ceraolo Spurio <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
2021-10-15drm/i915: Enable multi-bb execbufMatthew Brost1-3/+0
Enable multi-bb execbuf by enabling the set_parallel extension. Signed-off-by: Matthew Brost <[email protected]> Reviewed-by: John Harrison <[email protected]> Signed-off-by: John Harrison <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
2021-10-15drm/i915: Update I915_GEM_BUSY IOCTL to understand composite fencesMatthew Brost3-13/+55
Parallel submission create composite fences (dma_fence_array) for excl / shared slots in objects. The I915_GEM_BUSY IOCTL checks these slots to determine the busyness of the object. Prior to patch it only check if the fence in the slot was a i915_request. Update the check to understand composite fences and correctly report the busyness. v2: (Tvrtko) - Remove duplicate BUILD_BUG_ON Reviewed-by: Daniele Ceraolo Spurio <[email protected]> Signed-off-by: Matthew Brost <[email protected]> Signed-off-by: John Harrison <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
2021-10-15drm/i915: Make request conflict tracking understand parallel submitsMatthew Brost1-14/+29
If an object in the excl or shared slot is a composite fence from a parallel submit and the current request in the conflict tracking is from the same parallel context there is no need to enforce ordering as the ordering is already implicit. Make the request conflict tracking understand this by comparing a parallel submit's parent context and skipping conflict insertion if the values match. v2: (John Harrison) - Reword commit message Signed-off-by: Matthew Brost <[email protected]> Reviewed-by: John Harrison <[email protected]> Signed-off-by: John Harrison <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
2021-10-15drm/i915/guc: Handle errors in multi-lrc requestsMatthew Brost1-3/+66
If an error occurs in the front end when multi-lrc requests are getting generated we need to skip these in the backend but we still need to emit the breadcrumbs seqno. An issues arises because with multi-lrc breadcrumbs there is a handshake between the parent and children to make forward progress. If all the requests are not present this handshake doesn't work. To work around this, if multi-lrc request has an error we skip the handshake but still emit the breadcrumbs seqno. v2: (John Harrison) - Add comment explaining the skipping of the handshake logic - Fix typos in the commit message v3: (John Harrison) - Fix up some comments about the math to NOP the ring Signed-off-by: Matthew Brost <[email protected]> Reviewed-by: John Harrison <[email protected]> Signed-off-by: John Harrison <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
2021-10-15drm/i915: Multi-BB execbufMatthew Brost7-251/+595
Allow multiple batch buffers to be submitted in a single execbuf IOCTL after a context has been configured with the 'set_parallel' extension. The number batches is implicit based on the contexts configuration. This is implemented with a series of loops. First a loop is used to find all the batches, a loop to pin all the HW contexts, a loop to create all the requests, a loop to submit (emit BB start, etc...) all the requests, a loop to tie the requests to the VMAs they touch, and finally a loop to commit the requests to the backend. A composite fence is also created for the generated requests to return to the user and to stick in dma resv slots. No behavior from the existing IOCTL should be changed aside from when throttling because the ring for a context is full. In this situation, i915 will now wait while holding the object locks. This change was done because the code is much simpler to wait while holding the locks and we believe there isn't a huge benefit of dropping these locks. If this proves false we can restructure the code to drop the locks during the wait. IGT: https://patchwork.freedesktop.org/patch/447008/?series=93071&rev=1 media UMD: https://github.com/intel/media-driver/pull/1252 v2: (Matthew Brost) - Return proper error value if i915_request_create fails v3: (John Harrison) - Add comment explaining create / add order loops + locking - Update commit message explaining different in IOCTL behavior - Line wrap some comments - eb_add_request returns void - Return -EINVAL rather triggering BUG_ON if cmd parser used (Checkpatch) - Check eb->batch_len[*current_batch] v4: (CI) - Set batch len if passed if via execbuf args - Call __i915_request_skip after __i915_request_commit (Kernel test robot) - Initialize rq to NULL in eb_pin_timeline v5: (John Harrison) - Fix typo in comments near bb order loops Signed-off-by: Matthew Brost <[email protected]> Reviewed-by: John Harrison <[email protected]> Signed-off-by: John Harrison <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
2021-10-15drm/i915/guc: Implement no mid batch preemption for multi-lrcMatthew Brost4-13/+326
For some users of multi-lrc, e.g. split frame, it isn't safe to preempt mid BB. To safely enable preemption at the BB boundary, a handshake between parent and child is needed, syncing the set of BBs at the beginning and end of each batch. This is implemented via custom emit_bb_start & emit_fini_breadcrumb functions and enabled by default if a context is configured by set parallel extension. Lastly, this patch updates the process descriptor to the correct size as the memory used in the handshake is directly after the process descriptor. v2: (John Harrison) - Fix a few comments wording - Add struture for parent page layout v3: (John Harrison) - A structure for sync semaphore - Use offsetof to calc address - Update commit message v4: (John Harrison) - Fix typos in comment explaining memory map of scratch page Signed-off-by: Matthew Brost <[email protected]> Reviewed-by: John Harrison <[email protected]> Signed-off-by: John Harrison <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
2021-10-15drm/i915/guc: Add basic GuC multi-lrc selftestMatthew Brost3-0/+181
Add very basic (single submission) multi-lrc selftest. Signed-off-by: Matthew Brost <[email protected]> Reviewed-by: John Harrison <[email protected]> Signed-off-by: John Harrison <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
2021-10-15drm/i915/guc: Connect UAPI to GuC multi-lrc interfaceMatthew Brost8-31/+374
Introduce 'set parallel submit' extension to connect UAPI to GuC multi-lrc interface. Kernel doc in new uAPI should explain it all. IGT: https://patchwork.freedesktop.org/patch/447008/?series=93071&rev=1 media UMD: https://github.com/intel/media-driver/pull/1252 v2: (Daniel Vetter) - Add IGT link and placeholder for media UMD link v3: (Kernel test robot) - Fix warning in unpin engines call (John Harrison) - Reword a bunch of the kernel doc v4: (John Harrison) - Add comment why perma-pin is done after setting gem context - Update some comments / docs for proto contexts v5: (John Harrison) - Rework perma-pin comment - Add BUG_IN if context is pinned when setting gem context Cc: Tvrtko Ursulin <[email protected]> Signed-off-by: Matthew Brost <[email protected]> Reviewed-by: John Harrison <[email protected]> Signed-off-by: John Harrison <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
2021-10-15drm/i915/guc: Update debugfs for GuC multi-lrcMatthew Brost1-14/+38
Display the workqueue status in debugfs for GuC contexts that are in parent-child relationship. v2: (John Harrison) - Output number children in debugfs Signed-off-by: Matthew Brost <[email protected]> Reviewed-by: John Harrison <[email protected]> Signed-off-by: John Harrison <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
2021-10-15drm/i915/guc: Implement multi-lrc resetMatthew Brost2-21/+58
Update context and full GPU reset to work with multi-lrc. The idea is parent context tracks all the active requests inflight for itself and its children. The parent context owns the reset replaying / canceling requests as needed. v2: (John Harrison) - Simply loop in find active request - Add comments to find ative request / reset loop v3: (John Harrison) - s/its'/its/g - Fix comment when searching for active request - Reorder if state in __guc_reset_context v4: (Kernel test robot) - Delete unused is_multi_lrc function Reviewed-by: John Harrison <[email protected]> Signed-off-by: Matthew Brost <[email protected]> Signed-off-by: John Harrison <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
2021-10-15drm/i915/guc: Insert submit fences between requests in parent-child relationshipMatthew Brost4-28/+108
The GuC must receive requests in the order submitted for contexts in a parent-child relationship to function correctly. To ensure this, insert a submit fence between the current request and last request submitted for requests / contexts in a parent child relationship. This is conceptually similar to a single timeline. Signed-off-by: Matthew Brost <[email protected]> Cc: John Harrison <[email protected]> Reviewed-by: John Harrison <[email protected]> Signed-off-by: John Harrison <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
2021-10-15drm/i915/guc: Implement multi-lrc submissionMatthew Brost6-75/+350
Implement multi-lrc submission via a single workqueue entry and single H2G. The workqueue entry contains an updated tail value for each request, of all the contexts in the multi-lrc submission, and updates these values simultaneously. As such, the tasklet and bypass path have been updated to coalesce requests into a single submission. v2: (John Harrison) - s/wqe/wqi - Use FIELD_PREP macros - Add GEM_BUG_ONs ensures length fits within field - Add comment / white space to intel_guc_write_barrier (Kernel test robot) - Make need_tasklet a static function v3: (Docs) - A comment for submission_stall_reason v4: (Kernel test robot) - Initialize return value in bypass tasklt submit function (John Harrison) - Add comment near work queue defs - Add BUILD_BUG_ON to ensure WQ_SIZE is a power of 2 - Update write_barrier comment to talk about work queue v5: (John Harrison) - Fix typo in work queue comment Reviewed-by: John Harrison <[email protected]> Signed-off-by: Matthew Brost <[email protected]> Signed-off-by: John Harrison <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
2021-10-15drm/i915/guc: Implement parallel context pin / unpin functionsMatthew Brost1-0/+70
Parallel contexts are perma-pinned by the upper layers which makes the backend implementation rather simple. The parent pins the guc_id and children increment the parent's pin count on pin to ensure all the contexts are unpinned before we disable scheduling with the GuC / or deregister the context. v2: (Daniel Vetter) - Perma-pin parallel contexts Signed-off-by: Matthew Brost <[email protected]> Reviewed-by: John Harrison <[email protected]> Signed-off-by: John Harrison <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
Powered by cgit, Gruvboxed by Blaster4385.