- 21 Dec, 2023 40 commits
-
-
Lucas De Marchi authored
Based on commit 8d8d062b ("drm/i915/mtl: Fix MTL stolen memory GGTT mapping"). For stolen on MTL and beyond, the address in the PTE is the offset from DSM base. While at it, update the comments explaining each part of the calculation. Reviewed-by:
Matt Roper <matthew.d.roper@intel.com> Link: https://lore.kernel.org/r/20230726160708.3967790-9-lucas.demarchi@intel.comSigned-off-by:
Lucas De Marchi <lucas.demarchi@intel.com> Signed-off-by:
Rodrigo Vivi <rodrigo.vivi@intel.com>
-
Lucas De Marchi authored
Integrated graphics 1270 and beyond should set the PTE_LM bit in the PTE when it's stolen memory. Add a new function, xe_bo_is_stolen_devmem(), and use it when encoding the PTE. In some places in the spec the PTE bit is called "Local Memory", abbreviated as LM, and in others it's called "Device Memory" (DM). Since we moved away from "Local Memory" and preferred the "vram" terminology, also rename the macros as DM to follow the name of the new function. Reviewed-by:
-
Lucas De Marchi authored
The output arg is_vram in xe_bo_addr() is unused by several callers. It's also not what the function is mainly doing. Remove the argument and let the interested callers to call xe_bo_is_vram(). Reviewed-by:
-
Lucas De Marchi authored
All the callers pass a NULL vma, so the buffer is always the BO. Remove the argument and the side effects of dealing with it. Reviewed-by:
Matthew Brost <matthew.brost@intel.com> Link: https://lore.kernel.org/r/20230726160708.3967790-5-lucas.demarchi@intel.comSigned-off-by:
-
Daniele Ceraolo Spurio authored
in commit 81593af6 ("drm/xe: Convert xe_mmio_wait32 to us so we can stop using wait_for_us.") the mcr semaphore register read was accidentally switched from waiting for the register to go to 1 to waiting for the register to go to 0, so we need to flip it back. Signed-off-by:
Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Matthew Brost <matthew.brost@intel.com> Reviewed-by:
-
Lucas De Marchi authored
Commit 37430402 ("drm/xe: NULL binding implementation") introduced the NULL binding implementation, but left a case in which the out value is_vram is not set and the caller will use whatever was on stack. Eventually the is_vram out could be removed, but this should at least fix the current bug. Fixes: 37430402 ("drm/xe: NULL binding implementation") Reviewed-by:
-
Matthew Auld authored
Bind engines need to use the migration vm, however we don't have any rpm for such a vm, otherwise the kernel would prevent rpm suspend-resume. There are two issues here, first is the actual engine create which needs to touch the lrc, but since that is in VRAM we trigger loads of missing mem_access asserts. The second issue is when destroying the actual engine, which requires GuC CT to deregister the context. v2 (Rodrigo): - Just use ENGINE_FLAG_VM as the indicator that we need to hold an rpm ref. This also handles the case in xe_vm_create() where we create default bind engines. Closes: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/499 Closes: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/504 Cc: Rodrigo Vivi <rodrigo.vivi@intel.com> Cc: Matthew Brost <matthew.brost@intel.com> Signed-off-by:Matthew Auld <matthew.auld@intel.com> Reviewed-by:
-
Matthew Brost authored
If no operations are generated for VM binds the out-syncs must still be signaled. Signed-off-by:
-
Matthew Brost authored
Do not queue the rebind worker directly, rather use the helper xe_vm_queue_rebind_worker. This ensures we use the correct work queue. Signed-off-by:
-
Rodrigo Vivi authored
The module parameter should reflect the name of the optional, experimental and unsafe option, rather than the default one. Signed-off-by:
José Roberto de Souza <jose.souza@intel.com>
-
Rodrigo Vivi authored
This config is the only real one. If execlist remains in the code it will forever be experimental and we shouldn't maintain an uapi like that for that experimental piece of code that should never be used by real users. Signed-off-by:
-
Matthew Auld authored
This allows vram_size > io_size, instead of just clamping the vram size to the BAR size, now that the driver supports it. Signed-off-by:
Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Reviewed-by:
Gwan-gyeong Mun <gwan-gyeong.mun@intel.com> Signed-off-by:
-
Matthew Auld authored
Mostly the same as i915. We add a new hint for userspace to force an object into the mappable part of vram. We also need to tell userspace how large the mappable part is. In Vulkan for example, there will be two vram heaps for small-bar systems. And here the size of each heap needs to be known. Likewise the used/avail tracking needs to account for the mappable part. We also limit the available tracking going forward, such that we limit to privileged users only, since these values are system wide and are technically considered an info leak. v2 (Maarten): - s/NEEDS_CPU_ACCESS/NEEDS_VISIBLE_VRAM/ in the uapi. We also no longer require smem as an extra placement. This is more flexible, and lets us use this for clear-color surfaces, since we need CPU access there but we don't want to attach smem, since that effectively disables CCS from kernel pov. - Reject clear-color CCS buffers where NEEDS_VISIBLE_VRAM is not set, instead of migrating it behind the scenes. v3 (José): - Split the changes that limit the accounting for perfmon_capable() into a separate patch. - Use XE_BO_CREATE_VRAM_MASK. v4 (Gwan-gyeong Mun): - Add some kernel-doc for the query bits. v5: - One small kernel-doc correction. The cpu_visible_size and corresponding used tracking are always zero for non XE_MEM_REGION_CLASS_VRAM. v6: - Without perfmon_capable() it likely makes more sense to report as zero, instead of reporting as used == total size. This should give similar behaviour as i915 which rather tracks free instead of used. - Only enforce NEEDS_VISIBLE_VRAM on rc_ccs_cc_plane surfaces when the device is actually small-bar. Testcase: igt/tests/xe_query Testcase: igt/tests/xe_mmap@small-bar Signed-off-by: -
Matthew Auld authored
Add the new flag XE_BO_NEEDS_CPU_ACCESS, to force allocating in the mappable part of vram. If no flag is specified we do a topdown allocation, to limit the chances of stealing the precious mappable part, if we don't need it. If this is a full-bar system, then this all gets nooped. For kernel users, it looks like xe_bo_create_pin_map() is the central place which users should call if they want CPU access to the object, so add the flag there. We still need to plumb this through for userspace allocations. Also it looks like page-tables are using pin_map(), which is less than ideal. If we can already use the GPU to do page-table management, then maybe we should just force that for small-bar. Signed-off-by:
-
Matt Roper authored
Platforms like MTL only have a single tile, but multiple GTs. Ensure XE_ENGINE_CREATE accepts engine creation on gt1 on such platforms. Reviewed-by:
-
Matt Roper authored
On MTL and beyond, the GPU performs non-coherent accesses to the PPGTT page tables. These page tables should be mapped as CPU:WC. Removes CAT errors triggered by xe_exec_basic@once-basic on MTL: xe 0000:00:02.0: [drm:__xe_pt_bind_vma [xe]] Preparing bind, with range [1a0000...1a0fff) engine 0000000000000000. xe 0000:00:02.0: [drm:xe_vm_dbg_print_entries [xe]] 1 entries to update xe 0000:00:02.0: [drm:xe_vm_dbg_print_entries [xe]] 0: Update level 3 at (0 + 1) [0...8000000000) f:0 xe 0000:00:02.0: [drm] Engine memory cat error: guc_id=2 xe 0000:00:02.0: [drm] Engine memory cat error: guc_id=2 xe 0000:00:02.0: [drm] Timedout job: seqno=4294967169, guc_id=2, flags=0x4 v2: - Rename to XE_BO_PAGETABLE to make it more clear that this BO is the pagetable itself, rather than just being bound in the PPGTT. (Lucas) Cc: Lucas De Marchi <lucas.demarchi@intel.com> Reviewed-by:
Nirmoy Das <nirmoy.das@intel.com> Link: https://lore.kernel.org/r/20230725003433.1992137-3-matthew.d.roper@intel.comSigned-off-by:
-
Matthew Auld authored
The main motivation is with d3cold which will make the suspend and resume callbacks even more scary, but is useful regardless. We already have the needed annotation on the acquire side with xe_device_mem_access_get(), and by adding the annotation on the release side we should have a lot more confidence that our locking hierarchy is correct. v2: - Move the annotation into both callbacks for better symmetry. Also don't hold over the entire mem_access_get(); we only need to lockep to understand what is being held upon entering mem_access_get(), and how that matches up with locks in the callbacks. Signed-off-by: -
Matthew Brost authored
We must use migrate engine for page fault binds in order to avoid a deadlock as the migrate engine has a reserved BCS instance which cannot be stuck on a fault. To use the migrate engine the engine argument to xe_migrate_update_pgtables must be NULL, this was incorrectly wired up so vm->eng[tile_id] was always being used. Fix this. Reviewed-by:
-
Matthew Brost authored
Only alloc userptr part of xe_vma for userptrs, this will save on space in the common BO case. Reviewed-by:
-
Matthew Brost authored
The callback kicks the worker thus mutually exclusive execution, combining saves a bit of space in xe_vma. Reviewed-by:
-
Matthew Brost authored
This will save us a few bytes in the xe_vma structure. v2: Use hweight8 rather than hweight_long (Rodrigo) Reviewed-by:
-
Matthew Brost authored
This list isn't used again, list_del is the proper call. Reviewed-by:
-
Matthew Brost authored
Combine the userptr, rebind, and destroy links into a union as the lists these links belong to are mutually exclusive. v2: Adjust which lists are combined (Thomas H) v3: Add kernel doc why this is safe (Thomas H), remove related change of list_del_init -> list_del (Rodrigo) Reviewed-by:
-
Matthew Brost authored
If we dont change page sizes we can avoid doing rebinds rather just do a partial unbind. The algorithm to determine its page size is greedy as we assume all pages in the removed VMA are the largest page used in the VMA. v2: Don't exceed 100 lines v3: struct xe_vma_op_unmap remove in different patch, remove XXX comment Reviewed-by:
-
Matthew Brost authored
xe_vma_op_unmap isn't used, remove it. Reviewed-by:
-
Matthew Brost authored
We currently have a race between bind engines which can result in corrupted page tables leading to faults. A simple example: bind A 0x0000-0x1000, engine A, has unsatisfied in-fence bind B 0x1000-0x2000, engine B, no in-fences exec A uses 0x1000-0x2000 Bind B will pass bind A and exec A will fault. This occurs as bind A programs the root of the page table in a bind job which is held up by an in-fence. Bind B in this case just programs a leaf entry of the structure. To fix use range-fence utility to track cross bind engine conflicts. In the above example bind A would insert an dependency into the range-fence tree with a key of 0x0-0x7fffffffff, bind B would find that dependency and its bind job would scheduled behind the unsatisfied in-fence and bind A's job. Reviewed-by: Maarten Lankhorst<maarten.lankhorst@linux.intel.com> Co-developed-by:
Thomas Hellström <thomas.hellstrom@linux.intel.com> Signed-off-by:
-
Thomas Hellström authored
Add generic utility to track range conflicts signaled by a dma-fence. Tracking implemented via an interval tree. An example use case being tracking conflicts for pending (un)binds from multiple bind engines. By being generic ths idea would this could moved to the DRM level and used in multiple drivers for similar problems. v2: Make interval tree functions static (CI) v3: Remove non-static cleanup function (CI) Reviewed-by:
-
Francois Dugast authored
Make explicit in the log that execlist submission is used to prevent from silently using it over GuC submission. Signed-off-by:
Francois Dugast <francois.dugast@intel.com> Reviewed-by:
-
Francois Dugast authored
Fix 6 errors and 20 warnings reported by checkpatch.pl. Signed-off-by:
-
Francois Dugast authored
Those look like leftover debug and are not even being used. If they were real debug/info, they should be using the drm helpers. Signed-off-by:
-
Francois Dugast authored
Those messages are unnecessary because a generic message is already produced in case of allocation failure. Besides, this also removes a misuse of the XE_IOCTL_DBG macro. Signed-off-by:
-
Lucas De Marchi authored
Use FIELD_PREP()/FIELD_GET() to encode the tile id into flags. Besides protecting for eventual overflow it also makes it easier to see a new flag can't be added as BIT(7). Reviewed-by:
-
Lucas De Marchi authored
Rename XE_VM_FLAGS_64K to XE_VM_FLAG_64K to follow the other names and s/GT/TILE/ that got missed in commit 08dea767 ("drm/xe: Move migration from GT to tile"). Reviewed-by:
-
Matthew Auld authored
It looks like bulk_move is set during object construction, but is only removed on object close, however in various places we might not yet have an actual fd to close, like on the error paths for the gem_create ioctl, and also one internal user for the evict_test_run_gt() selftest. Try to handle those cases by manually resetting the bulk_move. This should prevent triggering: WARNING: CPU: 7 PID: 8252 at drivers/gpu/drm/ttm/ttm_bo.c:327 ttm_bo_release+0x25e/0x2a0 [ttm] v2 (Nirmoy): - It should be safe to just unconditionally call __xe_bo_unset_bulk_move() in most places. Signed-off-by: -
Matthew Auld authored
Test seems to be failing badly after calling xe_bo_restore_kernel(). Taking a snapshot of the CTB and copying back a potentially old version seems risky, depending on what might have been inflight. Also it seems snapshotting the ADS object and copying back results in serious breakage. Normally when calling xe_bo_restore_kernel() we always fully restart the GT, which re-intializes such things. We could potentially skip saving and restoring such objects in xe_bo_evict_all() however seems quite fragile not to also restart the GT. Try to do that here by triggering a GT reset. Signed-off-by:
-
Matthew Auld authored
The GPU job will keep the device awake, however assumption here is that caller of xe_migrate_clear() is also holding mem_access.ref otherwise we hit the asserts in xe_sa_bo_flush_write() prior to the job construction. Signed-off-by:
-
Matthew Auld authored
We are calling fairly low level things like xe_bo_restore_kernel() which expect caller to be holding mem_access.ref. Since we are doing stuff like evict_all we likely don't want to race with rpm suspend, since that potentially wants to do the same thing, so just wrap the whole test. Signed-off-by:
-
Matthew Auld authored
The atomics here might hide potential issues, also rpm core is not holding any lock when calling our rpm resume callback, so add a dummy lock with the idea that xe_pm_runtime_resume() is eventually going to be called when we are holding it. This only needs to happen once and then lockdep can validate all callers and their locks. v2: (Thomas Hellström) - Prefer static lockdep_map instead of full blown mutex. Signed-off-by:
-
Matthew Auld authored
Lockdep gives the following splat: [ 594.158863] ffff888140da53f0 (&vm->userptr.notifier_lock){++++}-{3:3}, at: vma_userptr_invalidate+0xeb/0x330 [xe] [ 594.158921] but task is already holding lock: [ 594.158926] ffffffff82761940 (mmu_notifier_invalidate_range_start){+.+.}-{0:0}, at: unmap_vmas+0x0/0x1c0 [ 594.158941] which lock already depends on the new lock. [ 594.158947] the existing dependency chain (in reverse order) is: [ 594.158953] -> #5 (mmu_notifier_invalidate_range_start){+.+.}-{0:0}: [ 594.158961] fs_reclaim_acquire+0x68/0xd0 [ 594.158969] __kmem_cache_alloc_node+0x2c/0x1b0 [ 594.158975] kmalloc_node_trace+0x1d/0xb0 [ 594.158983] alloc_worker+0x18/0x50 [ 594.158989] init_rescuer.part.0+0x13/0xa0 [ 594.158995] workqueue_init+0xdf/0x210 [ 594.159001] kernel_init_freeable+0x5c/0x2f0 [ 594.159009] kernel_init+0x11/0x1a0 [ 594.159017] ret_from_fork+0x29/0x50 [ 594.159023] -> #4 (fs_reclaim){+.+.}-{0:0}: [ 594.159031] fs_reclaim_acquire+0xa0/0xd0 [ 594.159037] __kmem_cache_alloc_node+0x2c/0x1b0 [ 594.159042] kmalloc_trace+0x20/0xb0 [ 594.159048] acpi_device_add+0x25a/0x3f0 [ 594.159056] acpi_add_single_object+0x387/0x750 [ 594.159063] acpi_bus_check_add+0x108/0x280 [ 594.159069] acpi_bus_scan+0x34/0xf0 [ 594.159075] acpi_scan_init+0xed/0x2b0 [ 594.159082] acpi_init+0x21e/0x520 [ 594.159087] do_one_initcall+0x53/0x260 [ 594.159092] kernel_init_freeable+0x18a/0x2f0 [ 594.159099] kernel_init+0x11/0x1a0 [ 594.159105] ret_from_fork+0x29/0x50 [ 594.159110] -> #3 (acpi_device_lock){+.+.}-{3:3}: [ 594.159117] __mutex_lock+0x95/0xd10 [ 594.159122] acpi_enable_wakeup_device_power+0x30/0x120 [ 594.159130] __acpi_device_wakeup_enable+0x34/0x110 [ 594.159138] acpi_pm_set_device_wakeup+0x55/0x140 [ 594.159143] __pci_enable_wake+0x56/0xb0 [ 594.159150] pci_finish_runtime_suspend+0x35/0x80 [ 594.159157] pci_pm_runtime_suspend+0xb5/0x1a0 [ 594.159162] __rpm_callback+0x3c/0x110 [ 594.159170] rpm_callback+0x58/0x70 [ 594.159176] rpm_suspend+0x15c/0x6f0 [ 594.159182] pm_runtime_work+0x9b/0xb0 [ 594.159188] process_one_work+0x263/0x520 [ 594.159195] worker_thread+0x4d/0x3b0 [ 594.159200] kthread+0xeb/0x120 [ 594.159206] ret_from_fork+0x29/0x50 [ 594.159211] -> #2 (acpi_wakeup_lock){+.+.}-{3:3}: [ 594.159218] __mutex_lock+0x95/0xd10 [ 594.159223] acpi_pm_set_device_wakeup+0x7a/0x140 [ 594.159228] __pci_enable_wake+0x77/0xb0 [ 594.159234] pci_pm_runtime_resume+0x70/0xd0 [ 594.159240] __rpm_callback+0x3c/0x110 [ 594.159246] rpm_callback+0x58/0x70 [ 594.159252] rpm_resume+0x50d/0x7a0 [ 594.159258] rpm_resume+0x267/0x7a0 [ 594.159264] __pm_runtime_resume+0x45/0x90 [ 594.159270] xe_pm_runtime_resume_and_get+0x12/0x50 [xe] [ 594.159314] xe_device_mem_access_get+0x97/0xc0 [xe] [ 594.159346] hw_engines+0x65/0xf0 [xe] [ 594.159380] seq_read_iter+0x10d/0x4b0 [ 594.159385] seq_read+0x9e/0xd0 [ 594.159390] full_proxy_read+0x4e/0x80 [ 594.159396] vfs_read+0xb6/0x310 [ 594.159401] ksys_read+0x60/0xe0 [ 594.159406] do_syscall_64+0x38/0x90 [ 594.159413] entry_SYSCALL_64_after_hwframe+0x72/0xdc [ 594.159419] -> #1 (&xe->mem_access.lock){+.+.}-{3:3}: [ 594.159427] xe_device_mem_access_get+0x43/0xc0 [xe] [ 594.159457] xe_gt_tlb_invalidation_vma+0x53/0x190 [xe] [ 594.159490] invalidation_fence_init+0x1d2/0x2c0 [xe] [ 594.159529] __xe_pt_unbind_vma+0x151/0x4e0 [xe] [ 594.159564] vm_bind_ioctl+0x48a/0xae0 [xe] [ 594.159602] async_op_work_func+0x20c/0x530 [xe] [ 594.159634] process_one_work+0x263/0x520 [ 594.159640] worker_thread+0x4d/0x3b0 [ 594.159646] kthread+0xeb/0x120 [ 594.159650] ret_from_fork+0x29/0x50 [ 594.159655] -> #0 (&vm->userptr.notifier_lock){++++}-{3:3}: [ 594.159663] __lock_acquire+0x16fa/0x2850 [ 594.159670] lock_acquire+0xd2/0x2e0 [ 594.159676] down_write+0x36/0xd0 [ 594.159681] vma_userptr_invalidate+0xeb/0x330 [xe] [ 594.159714] __mmu_notifier_invalidate_range_start+0x239/0x2a0 [ 594.159722] unmap_vmas+0x1ac/0x1c0 [ 594.159727] unmap_region+0xb5/0x120 [ 594.159732] do_vmi_align_munmap+0x2be/0x430 [ 594.159739] do_vmi_munmap+0xea/0x120 [ 594.159744] __vm_munmap+0x9c/0x160 [ 594.159750] __x64_sys_munmap+0x12/0x20 [ 594.159756] do_syscall_64+0x38/0x90 [ 594.159761] entry_SYSCALL_64_after_hwframe+0x72/0xdc [ 594.159768] other info that might help us debug this: [ 594.159773] Chain exists of: &vm->userptr.notifier_lock --> fs_reclaim --> mmu_notifier_invalidate_range_start [ 594.159785] Possible unsafe locking scenario: [ 594.159790] CPU0 CPU1 [ 594.159794] ---- ---- [ 594.159797] lock(mmu_notifier_invalidate_range_start); [ 594.159802] lock(fs_reclaim); [ 594.159808] lock(mmu_notifier_invalidate_range_start); [ 594.159814] lock(&vm->userptr.notifier_lock); [ 594.159819] The VM should be holding a mem_access.ref so this looks like it should be a false positive and we can just drop the explicit mem_access in xe_gt_tlb_invalidation(). The GGTT invalidation path also takes care to hold mem_access.ref so should be fine there also, and we already assert that we hold access.ref for the GuC communication underneath. Signed-off-by: -
Matthew Auld authored
Increase the sensitivity of the ggtt->lock by priming it against FS_RECLAIM, such that allocating memory while holding will result in lockdep splats. Signed-off-by:
-
