Linaro-mm-sig April 2026

linaro-mm-sig@lists.linaro.org

91 participants
214 discussions

[PATCH v5 0/2] dma-buf: heaps: system: add an option to allocate explicitly shared/decrypted memory

by Jiri Pirko

From: Jiri Pirko <jiri(a)nvidia.com> Confidential computing (CoCo) VMs/guests, such as AMD SEV and Intel TDX, run with private/encrypted memory which creates a challenge for devices that do not support DMA to it (no TDISP support). For kernel-only DMA operations, swiotlb bounce buffering provides a transparent solution by copying data through shared memory. However, the only way to get this memory into userspace is via the DMA API's dma_alloc_pages()/dma_mmap_pages() type interfaces which limits the use of the memory to a single DMA device, and is incompatible with pin_user_pages(). These limitations are particularly problematic for the RDMA subsystem which makes heavy use of pin_user_pages() and expects flexible memory usage between many different DMA devices. This patch series enables userspace to explicitly request shared (decrypted) memory allocations from new dma-buf system_cc_shared heap. Userspace can mmap this memory and pass the dma-buf fd to other existing importers such as RDMA or DRM devices to access the memory. The DMA API is improved to allow the dma heap exporter to DMA map the shared memory to each importing device. Based on dma-mapping-for-next e7442a68cd1ee797b585f045d348781e9c0dde0d Jiri Pirko (2): dma-mapping: introduce DMA_ATTR_CC_SHARED for shared memory dma-buf: heaps: system: add system_cc_shared heap for explicitly shared memory drivers/dma-buf/heaps/system_heap.c | 103 ++++++++++++++++++++++++++-- include/linux/dma-mapping.h | 10 +++ include/trace/events/dma.h | 3 +- kernel/dma/direct.h | 14 +++- kernel/dma/mapping.c | 13 +++- 5 files changed, 132 insertions(+), 11 deletions(-) -- 2.51.1

1 hour, 58 minutes

[PATCH] drm/amdgpu: fix recursive ww_mutex acquire in amdgpu_devcoredump_format

by Mikhail Gavrilov

When dumping IB contents from a hung job, amdgpu_devcoredump_format() acquires the VM root PD's reservation lock via amdgpu_vm_lock_by_pasid() and then, for each IB referenced by the job, calls amdgpu_bo_reserve() on the BO that backs the IB. Both reservations are taken on reservation_ww_class_mutex objects but neither uses a ww_acquire_ctx, which trips lockdep: WARNING: possible recursive locking detected -------------------------------------------- kworker/u128:0 is trying to acquire lock: ffff88838b16e1f0 (reservation_ww_class_mutex){+.+.}-{4:4}, at: amdgpu_devcoredump_format+0x1594/0x23f0 [amdgpu] but task is already holding lock: ffff8882f82681f0 (reservation_ww_class_mutex){+.+.}-{4:4}, at: amdgpu_devcoredump_format+0x1594/0x23f0 [amdgpu] Possible unsafe locking scenario: CPU0 ---- lock(reservation_ww_class_mutex); lock(reservation_ww_class_mutex); *** DEADLOCK *** May be due to missing lock nesting notation Workqueue: events_unbound amdgpu_devcoredump_deferred_work [amdgpu] Call Trace: __ww_mutex_lock.constprop.0 ww_mutex_lock amdgpu_bo_reserve amdgpu_devcoredump_format+0x1594 [amdgpu] amdgpu_devcoredump_deferred_work+0xea [amdgpu] process_one_work worker_thread kthread The two reservations are on different BOs in the captured trace, so the splat is a lockdep-correctness warning, not an observed deadlock. It becomes a real self-deadlock whenever the IB BO shares its dma_resv with the root PD (the always-valid case, see amdgpu_vm_is_bo_always_valid()): amdgpu_bo_reserve(abo) re-acquires the same ww_mutex without a ticket and blocks forever. With amdgpu.gpu_recovery=0 the timeout handler refires every ~2 s and each invocation produces this splat, drowning the kernel ring buffer. Fix it by collecting the per-IB BO references under the root PD's reservation, then releasing the root before reserving each IB BO individually. The walk over the VM mapping tree must remain under the root lock (mappings can be torn down without it), but the actual content copies do not need to nest inside it. Each per-IB reservation is now an independent top-level acquire, eliminating the nested ww_mutex. The collect/release logic is factored out into two small helpers (amdgpu_devcoredump_collect_ib_refs / amdgpu_devcoredump_release_ib_refs) to keep the main function's indentation reasonable. This also fixes a BO refcount leak in the original code: when amdgpu_bo_reserve() failed, control jumped to free_ib_content without running amdgpu_bo_unref(). In the new structure the per-IB BO refs are released unconditionally in the cleanup helper. Reproducer (~150 LoC libdrm_amdgpu): submit a single GFX IB containing PACKET3_INDIRECT_BUFFER chained at GPU VA 0 and wait for the fence. The TDR fires within ~10 s and the deferred coredump worker produces the splat above on every invocation. Fixes: 7b15fc2d1f1a ("drm/amdgpu: dump job ibs in the devcoredump") Cc: stable(a)vger.kernel.org # 7.1 Signed-off-by: Mikhail Gavrilov <mikhail.v.gavrilov(a)gmail.com> --- .../gpu/drm/amd/amdgpu/amdgpu_dev_coredump.c | 147 +++++++++++++----- 1 file changed, 110 insertions(+), 37 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dev_coredump.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_dev_coredump.c index d386bc775d03..f6bb968de756 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dev_coredump.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dev_coredump.c @@ -207,6 +207,72 @@ static void amdgpu_devcoredump_fw_info(struct amdgpu_device *adev, } } +struct amdgpu_devcoredump_ib_ref { + struct amdgpu_bo *bo; + u64 offset; +}; + +/* + * Walk the VM's mapping tree under the root PD's reservation to obtain the BO + * that backs each IB and pin it with a refcount. The root PD reservation is + * dropped before this function returns; the caller can then reserve each IB + * BO individually without nesting ww_mutex acquires on + * reservation_ww_class_mutex. + * + * Returns an array of num_ibs entries (each ib_refs[i].bo may be NULL if its + * mapping was not found), or NULL on allocation failure / VM lookup failure. + * The caller must release the BO refs and free the array. + */ +static struct amdgpu_devcoredump_ib_ref * +amdgpu_devcoredump_collect_ib_refs(struct amdgpu_device *adev, + struct amdgpu_coredump_info *coredump) +{ + struct amdgpu_devcoredump_ib_ref *ib_refs; + struct amdgpu_bo_va_mapping *mapping; + struct amdgpu_bo *root; + struct amdgpu_vm *vm; + u64 va_start; + + ib_refs = kcalloc(coredump->num_ibs, sizeof(*ib_refs), GFP_KERNEL); + if (!ib_refs) + return NULL; + + vm = amdgpu_vm_lock_by_pasid(adev, &root, coredump->pasid); + if (!vm) { + kfree(ib_refs); + return NULL; + } + + for (int i = 0; i < coredump->num_ibs; i++) { + va_start = coredump->ibs[i].gpu_addr & AMDGPU_GMC_HOLE_MASK; + mapping = amdgpu_vm_bo_lookup_mapping(vm, va_start / AMDGPU_GPU_PAGE_SIZE); + if (!mapping) + continue; + + ib_refs[i].bo = amdgpu_bo_ref(mapping->bo_va->base.bo); + ib_refs[i].offset = va_start - + mapping->start * AMDGPU_GPU_PAGE_SIZE; + } + + amdgpu_bo_unreserve(root); + amdgpu_bo_unref(&root); + + return ib_refs; +} + +static void +amdgpu_devcoredump_release_ib_refs(struct amdgpu_devcoredump_ib_ref *ib_refs, + int num_ibs) +{ + if (!ib_refs) + return; + + for (int i = 0; i < num_ibs; i++) + if (ib_refs[i].bo) + amdgpu_bo_unref(&ib_refs[i].bo); + kfree(ib_refs); +} + static ssize_t amdgpu_devcoredump_format(char *buffer, size_t count, struct amdgpu_coredump_info *coredump) { @@ -214,13 +280,11 @@ amdgpu_devcoredump_format(char *buffer, size_t count, struct amdgpu_coredump_inf struct drm_printer p; struct drm_print_iterator iter; struct amdgpu_vm_fault_info *fault_info; - struct amdgpu_bo_va_mapping *mapping; struct amdgpu_ip_block *ip_block; struct amdgpu_res_cursor cursor; - struct amdgpu_bo *abo, *root; - uint64_t va_start, offset; + struct amdgpu_bo *abo; + uint64_t offset; struct amdgpu_ring *ring; - struct amdgpu_vm *vm; u32 *ib_content; uint8_t *kptr; int ver, i, j, r; @@ -343,43 +407,52 @@ amdgpu_devcoredump_format(char *buffer, size_t count, struct amdgpu_coredump_inf drm_printf(&p, "VRAM is lost due to GPU reset!\n"); if (coredump->num_ibs) { - /* Don't try to lookup the VM or map the BOs when calculating the - * size required to store the devcoredump. + struct amdgpu_devcoredump_ib_ref *ib_refs = NULL; + + /* + * Snapshot per-IB BO references under the root PD's reservation, + * then release the root before reserving each IB BO individually + * to copy its contents. + * + * Reserving an IB BO while the root PD is still reserved would + * be a nested ww_mutex acquire on reservation_ww_class_mutex + * without a ww_acquire_ctx, which trips lockdep's recursive- + * locking check and self-deadlocks for IB BOs that share their + * dma_resv with the root PD (always-valid BOs). + * + * Skip lookup/reservation entirely on the sizing pass: it does + * not write IB content, and the size estimate doesn't depend on + * whether the BOs are reachable. */ - if (sizing_pass) - vm = NULL; - else - vm = amdgpu_vm_lock_by_pasid(adev, &root, coredump->pasid); + if (!sizing_pass) + ib_refs = amdgpu_devcoredump_collect_ib_refs(adev, coredump); - for (int i = 0; i < coredump->num_ibs && (sizing_pass || vm); i++) { + for (int i = 0; i < coredump->num_ibs; i++) { ib_content = kvmalloc_array(coredump->ibs[i].ib_size_dw, 4, GFP_KERNEL); if (!ib_content) continue; - /* vm=NULL can only happen when 'sizing_pass' is true. Skip to the - * drm_printf() calls (ib_content doesn't need to be initialized - * as its content won't be written anywhere). - */ - if (!vm) + if (sizing_pass) goto output_ib_content; - va_start = coredump->ibs[i].gpu_addr & AMDGPU_GMC_HOLE_MASK; - mapping = amdgpu_vm_bo_lookup_mapping(vm, va_start / AMDGPU_GPU_PAGE_SIZE); - if (!mapping) - goto free_ib_content; + if (!ib_refs || !ib_refs[i].bo) + goto output_ib_content; + + abo = ib_refs[i].bo; + offset = ib_refs[i].offset; - offset = va_start - (mapping->start * AMDGPU_GPU_PAGE_SIZE); - abo = amdgpu_bo_ref(mapping->bo_va->base.bo); r = amdgpu_bo_reserve(abo, false); if (r) - goto free_ib_content; + goto output_ib_content; if (abo->flags & AMDGPU_GEM_CREATE_NO_CPU_ACCESS) { off = 0; - if (abo->tbo.resource->mem_type != TTM_PL_VRAM) - goto unreserve_abo; + if (abo->tbo.resource->mem_type != TTM_PL_VRAM) { + amdgpu_bo_unreserve(abo); + goto output_ib_content; + } amdgpu_res_first(abo->tbo.resource, offset, coredump->ibs[i].ib_size_dw * 4, @@ -395,8 +468,10 @@ amdgpu_devcoredump_format(char *buffer, size_t count, struct amdgpu_coredump_inf r = ttm_bo_kmap(&abo->tbo, 0, PFN_UP(abo->tbo.base.size), &abo->kmap); - if (r) - goto unreserve_abo; + if (r) { + amdgpu_bo_unreserve(abo); + goto output_ib_content; + } kptr = amdgpu_bo_kptr(abo); kptr += offset; @@ -406,21 +481,19 @@ amdgpu_devcoredump_format(char *buffer, size_t count, struct amdgpu_coredump_inf amdgpu_bo_kunmap(abo); } + amdgpu_bo_unreserve(abo); + output_ib_content: drm_printf(&p, "\nIB #%d 0x%llx %d dw\n", i, coredump->ibs[i].gpu_addr, coredump->ibs[i].ib_size_dw); - for (int j = 0; j < coredump->ibs[i].ib_size_dw; j++) - drm_printf(&p, "0x%08x\n", ib_content[j]); -unreserve_abo: - if (vm) - amdgpu_bo_unreserve(abo); -free_ib_content: + if (!sizing_pass && ib_refs && ib_refs[i].bo) { + for (int j = 0; j < coredump->ibs[i].ib_size_dw; j++) + drm_printf(&p, "0x%08x\n", ib_content[j]); + } kvfree(ib_content); } - if (vm) { - amdgpu_bo_unreserve(root); - amdgpu_bo_unref(&root); - } + + amdgpu_devcoredump_release_ib_refs(ib_refs, coredump->num_ibs); } return count - iter.remain; -- 2.54.0

6 hours, 50 minutes

[PATCH v5 0/4] dma-buf: heaps: Turn heaps into modules

by Maxime Ripard

Hi, The recent introduction of heaps in the optee driver [1] made possible the creation of heaps as modules. It's generally a good idea if possible, including for the already existing system and CMA heaps. The system one is pretty trivial, the CMA is now easy too with the reworks we got in 7.1-r1. Let me know what you think, Maxime 1: https://lore.kernel.org/dri-devel/20250911135007.1275833-4-jens.wiklander@l… Signed-off-by: Maxime Ripard <mripard(a)kernel.org> --- Changes in v5: - Rebase on 7.1-rc1 - Add a patch to enable the heaps in arm64 defconfig - Link to v4: https://lore.kernel.org/r/20260331-dma-buf-heaps-as-modules-v4-0-e18fda5044… Changes in v4: - Fix compilation failure - Rework to take into account OF_RESERVED_MEM - Fix regression making the default CMA area disappear if not created through the DT - Added some documentation and comments - Link to v3: https://lore.kernel.org/r/20260303-dma-buf-heaps-as-modules-v3-0-24344812c7… Changes in v3: - Squashed cma_get_name and cma_alloc/release patches - Fixed typo in Export dev_get_cma_area commit title - Fixed compilation failure with DMA_CMA but not OF_RESERVED_MEM - Link to v2: https://lore.kernel.org/r/20260227-dma-buf-heaps-as-modules-v2-0-454aee7e06… Changes in v2: - Collect tags - Don't export dma_contiguous_default_area anymore, but export dev_get_cma_area instead - Mentioned that heap modules can't be removed - Link to v1: https://lore.kernel.org/r/20260225-dma-buf-heaps-as-modules-v1-0-2109225a09… --- Maxime Ripard (4): dma-buf: heaps: Export mem_accounting parameter dma-buf: heaps: cma: Turn the heap into a module dma-buf: heaps: system: Turn the heap into a module arm64: defconfig: Enable dma-buf heaps arch/arm64/configs/defconfig | 3 +++ drivers/dma-buf/dma-heap.c | 1 + drivers/dma-buf/heaps/Kconfig | 4 ++-- drivers/dma-buf/heaps/cma_heap.c | 3 +++ drivers/dma-buf/heaps/system_heap.c | 5 +++++ 5 files changed, 14 insertions(+), 2 deletions(-) --- base-commit: 5e9b7d093f3f77cb0af4409559e3d139babfb443 change-id: 20260225-dma-buf-heaps-as-modules-1034b3ec9f2a Best regards, -- Maxime Ripard <mripard(a)kernel.org>

23 hours, 39 minutes

[PATCH v3 00/10] Add dmabuf read/write via io_uring

by Pavel Begunkov

The patch set allows to register a dmabuf to an io_uring instance for a specified file and use it with io_uring read / write requests. The infrastructure is not tied to io_uring and there could be more users in the future. A similar idea was attempted some years ago by Keith [1], from where I borrowed a good number of changes, and later was brough up by Tushar and Vishal from Intel. It's an opt-in feature for files, and they need to implement a new file operation to use it. Only NVMe block devices are supported in this series. The user API is built on top of io_uring's "registered buffers", where a dmabuf is registered in a special way, but after it can be used as any other "registered buffer" with IORING_OP_{READ,WRITE}_FIXED requests. It's created via a new file operation and the resulted map is then passed through the I/O stack in a new iterator type. There is some additional infrastructure to bind it all, which also counts requests using a dmabuf map and managing lifetimes, which is used to implement map invalidation. It was tested for GPU <-> NVMe transfers. Also, as it maintains a long-term dma mapping, it helps with the IOMMU cost. The numbers below are for udmabuf reads previously run by Anuj for different IOMMU modes: - STRICT: before = 570 KIOPS, after = 5.01 MIOPS - LAZY: before = 1.93 MIOPS, after = 5.01 MIOPS - PASSTHROUGH: before = 5.01 MIOPS, after = 5.01 MIOPS There are some liburing tests that can serve as an example: git: https://github.com/isilence/liburing.git rw-dmabuf-tests-v3 url: https://github.com/isilence/liburing/tree/rw-dmabuf-tests-v3 [1] https://lore.kernel.org/io-uring/20220805162444.3985535-1-kbusch@fb.com/ v3: - Rework io_uring registration - Move token/map infrastructure code out of blk-mq - Simplify callbacks: remove a separate blk-mq table, which was mostly just forwarding calls (to nvme). - Don't skip dma sync depending on request direction - Fix a couple of hangs - Rename s/dma/dmabuf/ - Other small changes v2: - Don't pass raw dma addresses, wrap it into a driver specific object - Split into two objects: token and map - Implement move_notify Pavel Begunkov (10): file: add callback for creating long-term dmabuf maps iov_iter: add iterator type for dmabuf maps block: move bvec init into __bio_clone block: introduce dma map backed bio type lib: add dmabuf token infrastructure block: forward create_dmabuf_token to drivers nvme-pci: implement dma_token backed requests io_uring/rsrc: introduce buf registration structure io_uring/rsrc: extend buffer update io_uring/rsrc: add dmabuf backed registered buffers block/bio.c | 28 +++- block/blk-merge.c | 14 ++ block/blk.h | 3 +- block/fops.c | 16 ++ drivers/nvme/host/pci.c | 282 ++++++++++++++++++++++++++++++++ include/linux/bio.h | 19 ++- include/linux/blk-mq.h | 9 + include/linux/blk_types.h | 8 +- include/linux/fs.h | 2 + include/linux/io_dmabuf_token.h | 92 +++++++++++ include/linux/io_uring_types.h | 5 + include/linux/uio.h | 11 ++ include/uapi/linux/io_uring.h | 31 +++- io_uring/io_uring.c | 3 +- io_uring/rsrc.c | 266 +++++++++++++++++++++++++----- io_uring/rsrc.h | 30 +++- io_uring/rw.c | 4 +- lib/Kconfig | 4 + lib/Makefile | 2 + lib/io_dmabuf_token.c | 272 ++++++++++++++++++++++++++++++ lib/iov_iter.c | 29 +++- 21 files changed, 1071 insertions(+), 59 deletions(-) create mode 100644 include/linux/io_dmabuf_token.h create mode 100644 lib/io_dmabuf_token.c -- 2.53.0

2 days, 6 hours

[PATCH 0/9] vfio/pci: Add mmap() for DMABUFs

by Matt Evans

Hi all, This series is based on previous RFCs/discussions: Tech topic: https://lore.kernel.org/linux-iommu/20250918214425.2677057-1-amastro@fb.com/ RFCv1: https://lore.kernel.org/all/20260226202211.929005-1-mattev@meta.com/ RFCv2: https://lore.kernel.org/kvm/20260312184613.3710705-1-mattev@meta.com/ The background/rationale is covered in more detail in the RFC cover letters. The TL;DR is: The goal is to enable userspace driver designs that use VFIO to export DMABUFs representing subsets of PCI device BARs, and "vend" those buffers from a primary process to other subordinate processes by fd. These processes then mmap() the buffers and their access to the device is isolated to the exported ranges. This is an improvement on sharing the VFIO device fd to subordinate processes, which would allow unfettered access . This is achieved by enabling mmap() of vfio-pci DMABUFs. Second, a new ioctl()-based revocation mechanism is added to allow the primary process to forcibly revoke access to previously-shared BAR spans, even if the subordinate processes haven't cleanly exited. (The related topic of safe delegation of iommufd control to the subordinate processes is not addressed here, and is follow-up work.) As well as isolation and revocation, another advantage to accessing a BAR through a VMA backed by a DMABUF is that it's straightforward to create the buffer with access attributes, such as write-combining. Notes on patches ================ Feedback from the RFCs requested that, instead of creating DMABUF-specific vm_ops and .fault paths, to go the whole way and migrate the existing VFIO PCI BAR mmap() to be backed by a DMABUF too, resulting in a common vm_ops and fault handler for mmap()s of both the VFIO device and explicitly-exported DMABUFs. This has been done for vfio-pci, but not sub-drivers (nvgrace-gpu's special-case mappings are unchanged). vfio/pci: Fix vfio_pci_dma_buf_cleanup() double-put A bug fix to a related are, whose context is a depdency for later patches. vfio/pci: Add a helper to look up PFNs for DMABUFs vfio/pci: Add a helper to create a DMABUF for a BAR-map VMA The first is for a DMABUF VMA fault handler to determine arbitrary-sized PFNs from ranges in DMABUF. Secondly, refactor DMABUF export for use by the existing export feature and a new helper that creates a DMABUF corresponding to a VFIO BAR mmap() request. vfio/pci: Convert BAR mmap() to use a DMABUF The vfio-pci core mmap() creates a DMABUF with the helper, and the vm_ops fault handler uses the other helper to resolve the fault. Because this depends on DMABUF structs/code, CONFIG_VFIO_PCI_CORE needs to depend on CONFIG_DMA_SHARED_BUFFER. The CONFIG_VFIO_PCI_DMABUF still conditionally enables the export support code. NOTE: The user mmap()s a device fd, but the resulting VMA's vm_file becomes that of the DMABUF which takes ownership of the device and puts it on release. This maintains the existing behaviour of a VMA keeping the VFIO device open. BAR zapping then happens via the existing vfio_pci_dma_buf_move() path, which now needs to unmap PTEs in the DMABUF's address_space. vfio/pci: Provide a user-facing name for BAR mappings There was a request for decent debug naming in /proc/<pid>/maps etc. comparable to the existing VFIO names: since the VMAs are DMABUFs, they have a "dmabuf:" prefix and can't be 100% identical to before. This is a user-visible change, but this patch at least now gives us extra info on the BDF & BAR being mapped. vfio/pci: Clean up BAR zap and revocation In general (see NOTE!) the vfio_pci_zap_bars() is now obsolete, since it unmaps PTEs in the VFIO device address_space which is now unused. This consolidates all calls (e.g. around reset) with the neighbouring vfio_pci_dma_buf_move()s into new functions, to revoke-zap/unrevoke. NOTE: the nvgrace-gpu driver continues to use its own private vm_ops, fault handler, etc. for its special memregions, and these DO still add PTEs to the VFIO device address_space. So, a temporary flag, vdev->bar_needs_zap, maintains the old behaviour for this use. At least this patch's consolidation makes it easy to remove the remaining zap when this need goes away. A FIXME is added: if nvgrace-gpu is converted to DMABUFs, remove the flag and final zap. vfio/pci: Support mmap() of a VFIO DMABUF Adds mmap() for a DMABUF fd exported from vfio-pci. It was a goal to keep the VFIO device fd lifetime behaviour unchanged with respect to the DMABUFs. An application can close all device fds, and this will revoke/clean up all DMABUFs; no mappings or other access can be performed now. When enabling mmap() of the DMABUFs, this means access through the VMA is also revoked. This complicates the fault handler because whilst the DMABUF exists, it has no guarantee that the corresponding VFIO device is still alive. Adds synchronisation ensuring the vdev is available before vdev->memory_lock is touched. (I decided against the alternative of preventing cleanup by holding the VFIO device open if any DMABUFs exist, because it's both a change of behaviour and less clean overall.) I've added a chonky comment in place, happy to clarify more if you have ideas. vfio/pci: Permanently revoke a DMABUF on request By weight, this is mostly a rename of revoked to an enum, status. There are now 3 states for a buffer, usable and revoked temporary/permanent. A new VFIO device ioctl is added, VFIO_DEVICE_PCI_DMABUF_REVOKE, which passes a DMABUF (exported from that device) and permanently revokes it. Thus a userspace driver can guarantee any downstream consumers of a shared fd are prevented from accessing a BAR range, and that range can be reused. The code doing revocation in vfio_pci_dma_buf_move() is moved, unchanged, to a common function for use by _move() and the new ioctl path. Q: I can't think of a good reason to temporarily revoke/unrevoke buffers from userspace, so didn't add a 'flags' field to the ioctl struct. Easy to add if people think it's worthwhile for future use. vfio/pci: Add mmap() attributes to DMABUF feature Reserves bits [31:28] in vfio_device_feature_dma_buf to allow a (CPU) mapping attribute to be specified for an exported set of ranges. The default is the current UC, and a new flag can specify CPU access as WC. Q: I've taken 4 bits; the intention is for this field to be a scalar not a bitmap (i.e. mutually-exclusive access properties). Perhaps 4 is a bit too many? Testing ======= (The [RFC ONLY] userspace test program, for QEMU edu-plus, has been dropped, but can be found in the GitHub branch below.) This code has been tested in mapping DMABUFs of single/multiple ranges, aliasing mmap()s, aliasing ranges across DMABUFs, vm_pgoff > 0, revocation, shutdown/cleanup scenarios, and hugepage mappings seem to work correctly. I've lightly tested WC mappings also (by observing resulting PTEs as having the correct attributes...). No regressions observed on the VFIO selftests, or on our internal vfio-pci applications. End === This is based on -next (next-20260414 but will merge earlier), as it depends on Leon's series "vfio: Wait for dma-buf invalidation to complete": https://lore.kernel.org/linux-iommu/20260205-nocturnal-poetic-chamois-f566a… These commits are on GitHub, along with "[RFC ONLY] selftests: vfio: Add standalone vfio_dmabuf_mmap_test": https://github.com/metamev/linux/compare/next-20260414...metamev:linux:dev/… Thanks for reading, Matt ================================================================================ Change log: v1: - Cleanup of the common DMABUF-aware VMA vm_ops fault handler and export code. - Fixed a lot of races, particularly faults racing with DMABUF cleanup (if the VFIO device fds close, for example). - Added nicer human-readable names for VFIO mmap() VMAs RFCv2: Respin based on the feedback/suggestions: https://lore.kernel.org/kvm/20260312184613.3710705-1-mattev@meta.com/ - Transform the existing VFIO BAR mmap path to also use DMABUFs behind the scenes, and then simply share that code for explicitly-mapped DMABUFs. Jason wanted to go that direction to enable iommufd VFIO type 1 emulation to pick up a DMABUF for an IO mapping. - Revoke buffers using a VFIO device fd ioctl RFCv1: https://lore.kernel.org/all/20260226202211.929005-1-mattev@meta.com/ Matt Evans (9): vfio/pci: Fix vfio_pci_dma_buf_cleanup() double-put vfio/pci: Add a helper to look up PFNs for DMABUFs vfio/pci: Add a helper to create a DMABUF for a BAR-map VMA vfio/pci: Convert BAR mmap() to use a DMABUF vfio/pci: Provide a user-facing name for BAR mappings vfio/pci: Clean up BAR zap and revocation vfio/pci: Support mmap() of a VFIO DMABUF vfio/pci: Permanently revoke a DMABUF on request vfio/pci: Add mmap() attributes to DMABUF feature drivers/vfio/pci/Kconfig | 3 +- drivers/vfio/pci/Makefile | 3 +- drivers/vfio/pci/nvgrace-gpu/main.c | 5 + drivers/vfio/pci/vfio_pci_config.c | 30 +- drivers/vfio/pci/vfio_pci_core.c | 224 ++++++++++--- drivers/vfio/pci/vfio_pci_dmabuf.c | 500 +++++++++++++++++++++++----- drivers/vfio/pci/vfio_pci_priv.h | 49 ++- include/linux/vfio_pci_core.h | 1 + include/uapi/linux/vfio.h | 42 ++- 9 files changed, 690 insertions(+), 167 deletions(-) -- 2.47.3

3 days

[PATCH v2] dma-fence: Clarify external lock use case in dma_fence_init() docs

by Maíra Canal

The kerneldoc comment on dma_fence_init() and dma_fence_init64() describe the legacy reason to pass an external lock as a need to prevent multiple fences "from signaling out of order". However, this wording is a bit misleading: a shared spinlock does not (and cannot) prevent the signaler from signaling out of order. Signaling order is the driver's responsibility regardless of whether the lock is shared or per-fence. What a shared lock actually provides is serialization of signaling and observation across fences in a given context, so that observers never see a later fence signaled while an earlier one is not. Reword both comments to describe this more accurately. Signed-off-by: Maíra Canal <mcanal(a)igalia.com> --- v1 -> v2: https://lore.kernel.org/dri-devel/20260411185756.1887119-4-mcanal@igalia.co… - Be more explicit about not allowing new users to use an external lock. - De-duplicate the explanation in dma_fence_init64() by pointing to the dma_fence_init() documentation. --- drivers/dma-buf/dma-fence.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/drivers/dma-buf/dma-fence.c b/drivers/dma-buf/dma-fence.c index 1c1eaecaf1b0..63b349ba9a34 100644 --- a/drivers/dma-buf/dma-fence.c +++ b/drivers/dma-buf/dma-fence.c @@ -1102,9 +1102,11 @@ __dma_fence_init(struct dma_fence *fence, const struct dma_fence_ops *ops, * context and seqno are used for easy comparison between fences, allowing * to check which fence is later by simply using dma_fence_later(). * - * It is strongly discouraged to provide an external lock because this couples - * lock and fence life time. This is only allowed for legacy use cases when - * multiple fences need to be prevented from signaling out of order. + * External locks are a relic from legacy use cases that needed a shared lock + * to serialize signaling and observation of fences within a context, so that + * observers never see a later fence signaled while an earlier one isn't. New + * users MUST NOT use external locks, as they force the issuer to outlive all + * fences that reference the lock. */ void dma_fence_init(struct dma_fence *fence, const struct dma_fence_ops *ops, @@ -1129,9 +1131,8 @@ EXPORT_SYMBOL(dma_fence_init); * Context and seqno are used for easy comparison between fences, allowing * to check which fence is later by simply using dma_fence_later(). * - * It is strongly discouraged to provide an external lock because this couples - * lock and fence life time. This is only allowed for legacy use cases when - * multiple fences need to be prevented from signaling out of order. + * New users MUST NOT use external locks. Check the documentation in + * dma_fence_init() to understand the motives behind the legacy use cases. */ void dma_fence_init64(struct dma_fence *fence, const struct dma_fence_ops *ops, -- 2.53.0

4 days

[PATCH 00/13] Add DMA support for LINFlexD UART driver

by Larisa Grigore

This patchset enhances the LINFlexD UART driver and its device tree bindings to support DMA transfers, configurable clock inputs, dynamic baudrate changes, and termios features. It also includes a series of fixes and improvements to ensure reliable operation across various modes and configurations. The changes added can be summarized as follows: 1. Fixes with respect to FIFO handling, locking, interrupt related registers and INITM mode transition. 2. Removal of the earlycon workaround, as proper FIFO handling and INITM transitions now ensure stable behavior. 3. Support for configurable stop bits and dynamic baudrate changes based on clock inputs and termios settings. 4. Optional DMA support for RX and TX paths, preventing character loss during high-throughput operations like copy-paste. Cyclic DMA is used for RX to avoid gaps between transactions. Larisa Grigore (8): serial: linflexuart: Clean SLEEP bit in LINCR1 after suspend serial: linflexuart: Check FIFO full before writing serial: linflexuart: Correctly clear UARTSR in buffer mode serial: linflexuart: Update RXEN/TXEN outside INITM mode serial: linflexuart: Ensure FIFO is empty when entering INITM serial: linflexuart: Revert earlycon workaround serial: linflexuart: Add support for configurable stop bits serial: linflexuart: Add DMA support Radu Pirea (5): serial: linflexuart: Fix locking in set_termios dt-bindings: serial: fsl-linflexuart: add clock input properties dt-bindings: serial: fsl-linflexuart: add dma properties serial: linflexuart: Add support for changing baudrate serial: linflexuart: Avoid stopping DMA during receive operations .../bindings/serial/fsl,s32-linflexuart.yaml | 31 + drivers/tty/serial/fsl_linflexuart.c | 972 +++++++++++++++--- 2 files changed, 846 insertions(+), 157 deletions(-) -- 2.47.0

4 days, 1 hour

[PATCH v2 00/10] dma-bug: heaps: Add Tegra VPR support

by Thierry Reding

From: Thierry Reding <treding(a)nvidia.com> Hi, This series adds support for the video protection region (VPR) used on Tegra SoC devices. It's a special region of memory that is protected from accesses by the CPU and used to store DRM protected content (both decrypted stream data as well as decoded video frames). Patches 1 and 2 add DT binding documentation for the VPR and add the VPR to the list of memory-region items for display and host1x. Patch 3 adds bitmap_allocate(), which is like bitmap_allocate_region() but works on sizes that are not a power of two. Patch 4 introduces new APIs needed by the Tegra VPR implementation that allow CMA areas to be dynamically created at runtime rather than using the fixed, system-wide list. This is used in this driver specifically because it can use an arbitrary number of these areas (though they are currently limited to 4). Patch 5 adds some infrastructure for DMA heap implementations to provide information through debugfs. The Tegra VPR implementation is added in patch 6. See its commit message for more details about the specifics of this implementation. Finally, patches 7-10 add the VPR placeholder node on Tegra234 and hook it up to the host1x and GPU nodes so that they can make use of this region. Changes in v2: - Tegra VPR implementation is now more optimized to reduce the number of (very slow) resize operations, and allows cross-chunk allocations - dynamic CMA areas are now trackd separately from static ones, but the global number of CMA pages accounts for all areas Thierry Thierry Reding (10): dt-bindings: reserved-memory: Document Tegra VPR dt-bindings: display: tegra: Document memory regions bitmap: Add bitmap_allocate() function mm/cma: Allow dynamically creating CMA areas dma-buf: heaps: Add debugfs support dma-buf: heaps: Add support for Tegra VPR arm64: tegra: Add VPR placeholder node on Tegra234 arm64: tegra: Add GPU node on Tegra234 arm64: tegra: Hook up VPR to host1x arm64: tegra: Hook up VPR to the GPU .../display/tegra/nvidia,tegra186-dc.yaml | 10 + .../display/tegra/nvidia,tegra20-dc.yaml | 10 +- .../display/tegra/nvidia,tegra20-host1x.yaml | 7 + .../nvidia,tegra-video-protection-region.yaml | 55 + arch/arm/mm/dma-mapping.c | 2 +- arch/arm64/boot/dts/nvidia/tegra234.dtsi | 60 + arch/s390/mm/init.c | 2 +- drivers/dma-buf/dma-heap.c | 56 + drivers/dma-buf/heaps/Kconfig | 7 + drivers/dma-buf/heaps/Makefile | 1 + drivers/dma-buf/heaps/cma_heap.c | 2 +- drivers/dma-buf/heaps/tegra-vpr.c | 1265 +++++++++++++++++ include/linux/bitmap.h | 25 +- include/linux/cma.h | 7 +- include/linux/dma-heap.h | 2 + include/trace/events/tegra_vpr.h | 57 + mm/cma.c | 187 ++- mm/cma.h | 5 +- 18 files changed, 1713 insertions(+), 47 deletions(-) create mode 100644 Documentation/devicetree/bindings/reserved-memory/nvidia,tegra-video-protection-region.yaml create mode 100644 drivers/dma-buf/heaps/tegra-vpr.c create mode 100644 include/trace/events/tegra_vpr.h -- 2.52.0

2 weeks, 2 days

[PATCH v18 00/42] DEPT(DEPendency Tracker)

by Byungchul Park

I'm happy to see that DEPT reported real problems in practice: https://lore.kernel.org/lkml/6383cde5-cf4b-facf-6e07-1378a485657d@I-love.SA… https://lore.kernel.org/lkml/1674268856-31807-1-git-send-email-byungchul.pa… https://lore.kernel.org/all/b6e00e77-4a8c-4e05-ab79-266bf05fcc2d@igalia.com/ I’ve added documentation describing DEPT — this should help you understand what DEPT is and how it works. You can use DEPT simply by enabling CONFIG_DEPT and checking dmesg at runtime. --- Hi Linus and folks, I’ve been developing a tool to detect deadlock possibilities by tracking waits/events — rather than lock acquisition order — to cover all the synchronization mechanisms. To summarize the design rationale, starting from the problem statement, through analysis, to the solution: CURRENT STATUS -------------- Lockdep tracks lock acquisition order to identify deadlock conditions. Additionally, it tracks IRQ state changes — via {en,dis}able — to detect cases where locks are acquired unintentionally during interrupt handling. PROBLEM ------- Waits and their associated events that are never reachable can eventually lead to deadlocks. However, since Lockdep focuses solely on lock acquisition order, it has inherent limitations when handling waits and events. Moreover, by tracking only lock acquisition order, Lockdep cannot properly handle read locks or cross-event scenarios — such as wait_for_completion() and complete() — making it increasingly inadequate as a general-purpose deadlock detection tool. SOLUTION -------- Once again, waits and their associated events that are never reachable can eventually lead to deadlocks. The new solution, DEPT, focuses directly on waits and events. DEPT monitors waits and events, and reports them when any become unreachable. DEPT provides: * Correct handling of read locks. * Support for general waits and events. * Continuous operation, even after multiple reports. * Simple, intuitive annotation APIs. There are still false positives, and some are already being worked on for suppression. Especially splitting the folio class into several appropriate classes e.g. block device mapping class and regular file mapping class, is currently under active development by me and Yeoreum Yun. Anyway, these efforts will need to continue for a while, as we’ve seen with lockdep over two decades. DEPT is tagged as EXPERIMENTAL in Kconfig — meaning it’s not yet suitable for use as an automation tool. However, for those who are interested in using DEPT to analyze complex synchronization patterns and extract dependency insights, DEPT would be a great tool for the purpose. Thanks for your support and contributions to: Harry Yoo <harry.yoo(a)oracle.com> Gwan-gyeong Mun <gwan-gyeong.mun(a)intel.com> Yunseong Kim <ysk(a)kzalloc.com> Yeoreum Yun <yeoreum.yun(a)arm.com> FAQ --- Q. Is this the first attempt to solve this problem? A. No. The cross-release feature (commit b09be676e0ff2) attempted to address it — as a Lockdep extension. It was merged, but quickly reverted, because: While it uncovered valuable hidden issues, it also introduced false positives. Since these false positives mask further real problems with Lockdep — and developers strongly dislike them — the feature was rolled back. Q. Why wasn’t DEPT built as a Lockdep extension? A. Lockdep is the result of years of work by kernel developers — and is now very stable. But I chose to build DEPT separately, because: While reusing BFS(Breadth First Search) and Lockdep’s hashing is beneficial, the rest of the system must be rebuilt from scratch to align with DEPT’s wait-event model — since Lockdep was originally designed for tracking lock acquisition orders, not wait-event dependencies. Q. Do you plan to replace Lockdep entirely? A. Not at all — Lockdep still plays a vital role in validating correct lock usage. While its dependency-checking logic should eventually be superseded by DEPT, the rest of its functionality should stay. Q. Should we replace the dependency check immediately? A. Absolutely not. Lockdep’s stability is the result of years of hard work by kernel developers. Lockdep and DEPT should run side by side until DEPT matures. Q. Stronger detection often leads to more false positives — which was a major pain point when cross-release was added. Is DEPT designed to handle this? A. Yes. DEPT’s simple, generalized design enables flexible reporting — so while false positives still need fixing, they’re far less disruptive than they were under the Lockdep extension, cross-release. Q. Why not fix all false positives out-of-tree before merging? A. Since the affected subsystems span the entire kernel, like Lockdep, which has relied on annotations to avoid false positives over the last two decades, DEPT too will require the annotation efforts. Performing annotation work within the mainline will help us add annotations more appropriately and will also make DEPT a useful tool for a wider range of users more quickly. CONFIG_DEPT is marked EXPERIMENTAL, so it’s opt-in. Some users are already interested in using DEPT to analyze complex synchronization patterns and extract dependency insights. Byungchul --- Changes from v17: 1. Rebase on the mainline as of 2025 Dec 5. 2. Convert the documents' format from txt to rst. (feedbacked by Jonathan Corbet and Bagas Sanjaya) 3. Move the documents from 'Documentation/dependency' to 'Documentation/dev-tools'. (feedbakced by Jonathan Corbet) 4. Improve the documentation. (feedbacked by NeilBrown) 5. Use a common function, enter_from_user_mode(), instead of arch specific code, to notice context switch from user mode. (feedbacked by Dave Hansen, Mark Rutland, and Mark Brown) 6. Resolve the header dependency issue by using dept's internal header, instead of relocating 'struct llist_{head,node}' to another header. (feedbacked by Greg KH) 7. Improve page(or folio) usage type APIs. 8. Add rust helper for wait_for_completion(). (feedbacked by Guangbo Cui, Boqun Feng, and Danilo Krummrich) 9. Refine some commit messages. Changes from v16: 1. Rebase on v6.17. 2. Fix a false positive from rcu (by Yunseong Kim) 3. Introduce APIs to set page's usage, dept_set_page_usage() and dept_reset_page_usage() to avoid false positives. 4. Consider lock_page() as a potential wait unconditionally. 5. Consider folio_lock_killable() as a potential wait unconditionally. 6. Add support for tracking PG_writeback waits and events. 7. Fix two build errors due to the additional debug information added by dept. (by Yunseong Kim) Changes from v15: 1. Fix typo and improve comments and commit messages (feedbacked by ALOK TIWARI, Waiman Long, and kernel test robot). 2. Do not stop dept on detection of cicular dependency of recover event, allowing to keep reporting. 3. Add SK hynix to copyright. 4. Consider folio_lock() as a potential wait unconditionally. 5. Fix Kconfig dependency bug (feedbacked by kernel test rebot). 6. Do not suppress reports that involve classes even that have already involved in other reports, allowing to keep reporting. Changes from v14: 1. Rebase on the current latest, v6.15-rc6. 2. Refactor dept code. 3. With multi event sites for a single wait, even if an event forms a circular dependency, the event can be recovered by other event(or wake up) paths. Even though informing the circular dependency is worthy but it should be suppressed once informing it, if it doesn't lead an actual deadlock. So introduce APIs to annotate the relationship between event site and recover site, that are, event_site() and dept_recover_event(). 4. wait_for_completion() worked with dept map embedded in struct completion. However, it generates a few false positves since all the waits using the instance of struct completion, share the map and key. To avoid the false positves, make it not to share the map and key but each wait_for_completion() caller have its own key by default. Of course, external maps also can be used if needed. 5. Fix a bug about hardirq on/off tracing. 6. Implement basic unit test for dept. 7. Add more supports for dma fence synchronization. 8. Add emergency stop of dept e.g. on panic(). 9. Fix false positives by mmu_notifier_invalidate_*(). 10. Fix recursive call bug by DEPT_WARN_*() and DEPT_STOP(). 11. Fix trivial bugs in DEPT_WARN_*() and DEPT_STOP(). 12. Fix a bug that a spin lock, dept_pool_spin, is used in both contexts of irq disabled and enabled without irq disabled. 13. Suppress reports with classes, any of that already have been reported, even though they have different chains but being barely meaningful. 14. Print stacktrace of the wait that an event is now waking up, not only stacktrace of the event. 15. Make dept aware of lockdep_cmp_fn() that is used to avoid false positives in lockdep so that dept can also avoid them. 16. Do do_event() only if there are no ecxts have been delimited. 17. Fix a bug that was not synchronized for stage_m in struct dept_task, using a spin lock, dept_task()->stage_lock. 18. Fix a bug that dept didn't handle the case that multiple ttwus for a single waiter can be called at the same time e.i. a race issue. 19. Distinguish each kernel context from others, not only by system call but also by user oriented fault so that dept can work with more accuracy information about kernel context. That helps to avoid a few false positives. 20. Limit dept's working to x86_64 and arm64. Changes from v13: 1. Rebase on the current latest version, v6.9-rc7. 2. Add 'dept' documentation describing dept APIs. Changes from v12: 1. Refine the whole document for dept. 2. Add 'Interpret dept report' section in the document, using a deadlock report obtained in practice. Hope this version of document helps guys understand dept better. https://lore.kernel.org/lkml/6383cde5-cf4b-facf-6e07-1378a485657d@I-love.SA… https://lore.kernel.org/lkml/1674268856-31807-1-git-send-email-byungchul.pa… Changes from v11: 1. Add 'dept' documentation describing the concept of dept. 2. Rewrite the commit messages of the following commits for using weaker lockdep annotation, for better description. fs/jbd2: Use a weaker annotation in journal handling cpu/hotplug: Use a weaker annotation in AP thread (feedbacked by Thomas Gleixner) Changes from v10: 1. Fix noinstr warning when building kernel source. 2. dept has been reporting some false positives due to the folio lock's unfairness. Reflect it and make dept work based on dept annotaions instead of just wait and wake up primitives. 3. Remove the support for PG_writeback while working on 2. I will add the support later if needed. 4. dept didn't print stacktrace for [S] if the participant of a deadlock is not lock mechanism but general wait and event. However, it made hard to interpret the report in that case. So add support to print stacktrace of the requestor who asked the event context to run - usually a waiter of the event does it just before going to wait state. 5. Give up tracking raw_local_irq_{disable,enable}() since it totally messed up dept's irq tracking. So make it work in the same way as lockdep does. I will consider it once any false positives by those are observed again. 6. Change the manual rwsem_acquire_read(->j_trans_commit_map) annotation in fs/jbd2/transaction.c to the try version so that it works as much as it exactly needs. 7. Remove unnecessary 'inline' keyword in dept.c and add '__maybe_unused' to a needed place. Changes from v9: 1. Fix a bug. SDT tracking didn't work well because of my big mistake that I should've used waiter's map to indentify its class but it had been working with waker's one. FYI, PG_locked and PG_writeback weren't affected. They still worked well. (reported by YoungJun) Changes from v8: 1. Fix build error by adding EXPORT_SYMBOL(PG_locked_map) and EXPORT_SYMBOL(PG_writeback_map) for kernel module build - appologize for that. (reported by kernel test robot) 2. Fix build error by removing header file's circular dependency that was caused by "atomic.h", "kernel.h" and "irqflags.h", which I introduced - appolgize for that. (reported by kernel test robot) Changes from v7: 1. Fix a bug that cannot track rwlock dependency properly, introduced in v7. (reported by Boqun and lockdep selftest) 2. Track wait/event of PG_{locked,writeback} more aggressively assuming that when a bit of PG_{locked,writeback} is cleared there might be waits on the bit. (reported by Linus, Hillf and syzbot) 3. Fix and clean bad style code e.i. unnecessarily introduced a randome pattern and so on. (pointed out by Linux) 4. Clean code for applying dept to wait_for_completion(). Changes from v6: 1. Tie to task scheduler code to track sleep and try_to_wake_up() assuming sleeps cause waits, try_to_wake_up()s would be the events that those are waiting for, of course with proper dept annotations, sdt_might_sleep_weak(), sdt_might_sleep_strong() and so on. For these cases, class is classified at sleep entrance rather than the synchronization initialization code. Which would extremely reduce false alarms. 2. Remove the dept associated instance in each page struct for tracking dependencies by PG_locked and PG_writeback thanks to the 1. work above. 3. Introduce CONFIG_dept_AGGRESIVE_TIMEOUT_WAIT to suppress reports that waits with timeout set are involved, for those who don't like verbose reporting. 4. Add a mechanism to refill the internal memory pools on running out so that dept could keep working as long as free memory is available in the system. 5. Re-enable tracking hashed-waitqueue wait. That's going to no longer generate false positives because class is classified at sleep entrance rather than the waitqueue initailization. 6. Refactor to make it easier to port onto each new version of the kernel. 7. Apply dept to dma fence. 8. Do trivial optimizaitions. Changes from v5: 1. Use just pr_warn_once() rather than WARN_ONCE() on the lack of internal resources because WARN_*() printing stacktrace is too much for informing the lack. (feedback from Ted, Hyeonggon) 2. Fix trivial bugs like missing initializing a struct before using it. 3. Assign a different class per task when handling onstack variables for waitqueue or the like. Which makes dept distinguish between onstack variables of different tasks so as to prevent false positives. (reported by Hyeonggon) 4. Make dept aware of even raw_local_irq_*() to prevent false positives. (reported by Hyeonggon) 5. Don't consider dependencies between the events that might be triggered within __schedule() and the waits that requires __schedule(), real ones. (reported by Hyeonggon) 6. Unstage the staged wait that has prepare_to_wait_event()'ed *and* yet to get to __schedule(), if we encounter __schedule() in-between for another sleep, which is possible if e.g. a mutex_lock() exists in 'condition' of ___wait_event(). 7. Turn on CONFIG_PROVE_LOCKING when CONFIG_DEPT is on, to rely on the hardirq and softirq entrance tracing to make dept more portable for now. Changes from v4: 1. Fix some bugs that produce false alarms. 2. Distinguish each syscall context from another *for arm64*. 3. Make it not warn it but just print it in case dept ring buffer gets exhausted. (feedback from Hyeonggon) 4. Explicitely describe "EXPERIMENTAL" and "dept might produce false positive reports" in Kconfig. (feedback from Ted) Changes from v3: 1. dept shouldn't create dependencies between different depths of a class that were indicated by *_lock_nested(). dept normally doesn't but it does once another lock class comes in. So fixed it. (feedback from Hyeonggon) 2. dept considered a wait as a real wait once getting to __schedule() even if it has been set to TASK_RUNNING by wake up sources in advance. Fixed it so that dept doesn't consider the case as a real wait. (feedback from Jan Kara) 3. Stop tracking dependencies with a map once the event associated with the map has been handled. dept will start to work with the map again, on the next sleep. Changes from v2: 1. Disable dept on bit_wait_table[] in sched/wait_bit.c reporting a lot of false positives, which is my fault. Wait/event for bit_wait_table[] should've been tagged in a higher layer for better work, which is a future work. (feedback from Jan Kara) 2. Disable dept on crypto_larval's completion to prevent a false positive. Changes from v1: 1. Fix coding style and typo. (feedback from Steven) 2. Distinguish each work context from another in workqueue. 3. Skip checking lock acquisition with nest_lock, which is about correct lock usage that should be checked by lockdep. Changes from RFC(v0): 1. Prevent adding a wait tag at prepare_to_wait() but __schedule(). (feedback from Linus and Matthew) 2. Use try version at lockdep_acquire_cpus_lock() annotation. 3. Distinguish each syscall context from another. Byungchul Park (41): dept: implement DEPT(DEPendency Tracker) dept: add single event dependency tracker APIs dept: add lock dependency tracker APIs dept: tie to lockdep and IRQ tracing dept: add proc knobs to show stats and dependency graph dept: distinguish each kernel context from another dept: distinguish each work from another dept: add a mechanism to refill the internal memory pools on running out dept: record the latest one out of consecutive waits of the same class dept: apply sdt_might_sleep_{start,end}() to wait_for_completion()/complete() dept: apply sdt_might_sleep_{start,end}() to swait dept: apply sdt_might_sleep_{start,end}() to waitqueue wait dept: apply sdt_might_sleep_{start,end}() to hashed-waitqueue wait dept: apply sdt_might_sleep_{start,end}() to dma fence dept: track timeout waits separately with a new Kconfig dept: apply timeout consideration to wait_for_completion()/complete() dept: apply timeout consideration to swait dept: apply timeout consideration to waitqueue wait dept: apply timeout consideration to hashed-waitqueue wait dept: apply timeout consideration to dma fence wait dept: make dept able to work with an external wgen dept: track PG_locked with dept dept: print staged wait's stacktrace on report locking/lockdep: prevent various lockdep assertions when lockdep_off()'ed dept: add documents for dept cpu/hotplug: use a weaker annotation in AP thread dept: assign dept map to mmu notifier invalidation synchronization dept: assign unique dept_key to each distinct dma fence caller dept: make dept aware of lockdep_set_lock_cmp_fn() annotation dept: make dept stop from working on debug_locks_off() dept: assign unique dept_key to each distinct wait_for_completion() caller completion, dept: introduce init_completion_dmap() API dept: introduce a new type of dependency tracking between multi event sites dept: add module support for struct dept_event_site and dept_event_site_dep dept: introduce event_site() to disable event tracking if it's recoverable dept: implement a basic unit test for dept dept: call dept_hardirqs_off() in local_irq_*() regardless of irq state dept: introduce APIs to set page usage and use subclasses_evt for the usage dept: track PG_writeback with dept SUNRPC: relocate struct rcu_head to the first field of struct rpc_xprt mm: percpu: increase PERCPU_DYNAMIC_SIZE_SHIFT on DEPT and large PAGE_SIZE Yunseong Kim (1): rcu/update: fix same dept key collision between various types of RCU Documentation/dev-tools/dept.rst | 778 ++++++ Documentation/dev-tools/dept_api.rst | 125 + drivers/dma-buf/dma-fence.c | 23 +- include/asm-generic/vmlinux.lds.h | 13 +- include/linux/completion.h | 124 +- include/linux/dept.h | 402 +++ include/linux/dept_ldt.h | 78 + include/linux/dept_sdt.h | 68 + include/linux/dept_unit_test.h | 67 + include/linux/dma-fence.h | 74 +- include/linux/hardirq.h | 3 + include/linux/irq-entry-common.h | 4 + include/linux/irqflags.h | 21 +- include/linux/local_lock_internal.h | 1 + include/linux/lockdep.h | 105 +- include/linux/lockdep_types.h | 3 + include/linux/mm_types.h | 4 + include/linux/mmu_notifier.h | 26 + include/linux/module.h | 5 + include/linux/mutex.h | 1 + include/linux/page-flags.h | 217 +- include/linux/pagemap.h | 37 +- include/linux/percpu-rwsem.h | 2 +- include/linux/percpu.h | 4 + include/linux/rcupdate_wait.h | 13 +- include/linux/rtmutex.h | 1 + include/linux/rwlock_types.h | 1 + include/linux/rwsem.h | 1 + include/linux/sched.h | 118 + include/linux/seqlock.h | 2 +- include/linux/spinlock_types_raw.h | 3 + include/linux/srcu.h | 2 +- include/linux/sunrpc/xprt.h | 9 +- include/linux/swait.h | 3 + include/linux/wait.h | 3 + include/linux/wait_bit.h | 3 + init/init_task.c | 2 + init/main.c | 2 + kernel/Makefile | 1 + kernel/cpu.c | 2 +- kernel/dependency/Makefile | 5 + kernel/dependency/dept.c | 3499 ++++++++++++++++++++++++++ kernel/dependency/dept_hash.h | 10 + kernel/dependency/dept_internal.h | 314 +++ kernel/dependency/dept_object.h | 13 + kernel/dependency/dept_proc.c | 94 + kernel/dependency/dept_unit_test.c | 173 ++ kernel/exit.c | 1 + kernel/fork.c | 2 + kernel/locking/lockdep.c | 33 + kernel/module/main.c | 19 + kernel/rcu/rcu.h | 1 + kernel/rcu/update.c | 5 +- kernel/sched/completion.c | 62 +- kernel/sched/core.c | 9 + kernel/workqueue.c | 3 + lib/Kconfig.debug | 48 + lib/debug_locks.c | 2 + lib/locking-selftest.c | 2 + mm/filemap.c | 38 + mm/mm_init.c | 3 + mm/mmu_notifier.c | 31 +- rust/helpers/completion.c | 5 + 63 files changed, 6602 insertions(+), 121 deletions(-) create mode 100644 Documentation/dev-tools/dept.rst create mode 100644 Documentation/dev-tools/dept_api.rst create mode 100644 include/linux/dept.h create mode 100644 include/linux/dept_ldt.h create mode 100644 include/linux/dept_sdt.h create mode 100644 include/linux/dept_unit_test.h create mode 100644 kernel/dependency/Makefile create mode 100644 kernel/dependency/dept.c create mode 100644 kernel/dependency/dept_hash.h create mode 100644 kernel/dependency/dept_internal.h create mode 100644 kernel/dependency/dept_object.h create mode 100644 kernel/dependency/dept_proc.c create mode 100644 kernel/dependency/dept_unit_test.c base-commit: 43dfc13ca972988e620a6edb72956981b75ab6b0 -- 2.17.1

2 weeks, 2 days

[PATCH] dma-buf: system_heap: Optimize sg_table-to-pages conversion in vmap

by Barry Song (Xiaomi)

From: Xueyuan Chen <Xueyuan.chen21(a)gmail.com> Replace the heavy for_each_sgtable_page() iterator in system_heap_do_vmap() with a more efficient nested loop approach. Instead of iterating page by page, we now iterate through the scatterlist entries via for_each_sgtable_sg(). Because pages within a single sg entry are physically contiguous, we can populate the page array with a in an inner loop using simple pointer math. This save a lot of time. The WARN_ON check is also pulled out of the loop to save branch instructions. Performance results mapping a 2GB buffer on Radxa O6: - Before: ~1440000 ns - After: ~232000 ns (~84% reduction in iteration time, or ~6.2x faster) Cc: Sumit Semwal <sumit.semwal(a)linaro.org> Cc: Benjamin Gaignard <benjamin.gaignard(a)collabora.com> Cc: Brian Starkey <Brian.Starkey(a)arm.com> Cc: John Stultz <jstultz(a)google.com> Cc: T.J. Mercier <tjmercier(a)google.com> Cc: Christian König <christian.koenig(a)amd.com> Signed-off-by: Xueyuan Chen <Xueyuan.chen21(a)gmail.com> Signed-off-by: Barry Song (Xiaomi) <baohua(a)kernel.org> --- drivers/dma-buf/heaps/system_heap.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/drivers/dma-buf/heaps/system_heap.c b/drivers/dma-buf/heaps/system_heap.c index b3650d8fd651..769f01f0cc96 100644 --- a/drivers/dma-buf/heaps/system_heap.c +++ b/drivers/dma-buf/heaps/system_heap.c @@ -224,16 +224,21 @@ static void *system_heap_do_vmap(struct system_heap_buffer *buffer) int npages = PAGE_ALIGN(buffer->len) / PAGE_SIZE; struct page **pages = vmalloc(sizeof(struct page *) * npages); struct page **tmp = pages; - struct sg_page_iter piter; void *vaddr; + u32 i, j, count; + struct page *base_page; + struct scatterlist *sg; if (!pages) return ERR_PTR(-ENOMEM); - for_each_sgtable_page(table, &piter, 0) { - WARN_ON(tmp - pages >= npages); - *tmp++ = sg_page_iter_page(&piter); + for_each_sgtable_sg(table, sg, i) { + base_page = sg_page(sg); + count = sg->length >> PAGE_SHIFT; + for (j = 0; j < count; j++) + *tmp++ = base_page + j; } + WARN_ON(tmp - pages != npages); vaddr = vmap(pages, npages, VM_MAP, PAGE_KERNEL); vfree(pages); -- 2.39.3 (Apple Git-146)

2 weeks, 3 days

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

Linaro-mm-sig April 2026