Linaro-mm-sig March 2026

linaro-mm-sig@lists.linaro.org

133 participants
1108 discussions

[PATCH 0/5] Replace the dmabuf custom test framework with kunit

by Jason Gunthorpe

Using kunit to write tests for new work on dmabuf is coming up: https://lore.kernel.org/all/26-v1-b5cab63049c0+191af-dmabuf_map_type_jgg@nv… Replace the custom test framework with kunit to avoid maintaining two concurrent test frameworks. The conversion minimizes code changes and uses simple pattern-oriented reworks to reduce the chance of breaking any tests. Aside from adding the kunit_test_suite() boilerplate, the conversion follows a number of patterns: Test failures without cleanup. For example: if (!ptr) return -ENOMEM; Becomes: KUNIT_ASSERT_NOT_NULL(test, ptr); In kunit ASSERT longjumps out of the test. Check for error, fail and cleanup: if (err) { pr_err("msg\n"); goto cleanup; } Becomes: if (err) { KUNIT_FAIL(test, "msg"); goto cleanup; } Preserve the existing failure messages and cleanup code. Cases where the test returns err but prints no message: if (err) goto cleanup; Becomes: if (err) { KUNIT_FAIL(test, "msg"); goto cleanup; } Use KUNIT_FAIL to retain the 'cleanup on err' behavior. Overall, the conversion is straightforward. The result can be run with kunit.py: $ tools/testing/kunit/kunit.py run --build_dir build_kunit_x86_64 --arch x86_64 --kunitconfig ./drivers/dma-buf/.kunitconfig [20:37:23] Configuring KUnit Kernel ... [20:37:23] Building KUnit Kernel ... Populating config with: $ make ARCH=x86_64 O=build_kunit_x86_64 olddefconfig Building with: $ make all compile_commands.json scripts_gdb ARCH=x86_64 O=build_kunit_x86_64 --jobs=20 [20:37:29] Starting KUnit Kernel (1/1)... [20:37:29] ============================================================ Running tests with: $ qemu-system-x86_64 -nodefaults -m 1024 -kernel build_kunit_x86_64/arch/x86/boot/bzImage -append 'kunit.enable=1 console=ttyS0 kunit_shutdown=reboot' -no-reboot -nographic -accel kvm -accel hvf -accel tcg -serial stdio -bios qboot.rom [20:37:30] ================ dma-buf-resv (5 subtests) ================= [20:37:30] [PASSED] test_sanitycheck [20:37:30] ===================== test_signaling ====================== [20:37:30] [PASSED] kernel [20:37:30] [PASSED] write [20:37:30] [PASSED] read [20:37:30] [PASSED] bookkeep [20:37:30] ================= [PASSED] test_signaling ================== ... [20:37:35] Testing complete. Ran 50 tests: passed: 49, skipped: 1 [20:37:35] Elapsed time: 12.635s total, 0.001s configuring, 6.551s building, 6.017s running One test that requires two CPUs is skipped since the default VM has a single CPU and cannot run the test. All other usual ways to run kunit work as well, and all tests are placed in a module to provide more options for how they are run. AI was used to do the large scale semantic search and replaces described above, then everything was hand checked. AI also deduced the issue with test_race_signal_callback() in a couple of seconds from the kunit crash (!!), again was hand checked though I am not so familiar with this test to be fully certain this is the best answer. Jason Gunthorpe (5): dma-buf: Change st-dma-resv.c to use kunit dma-buf: Change st-dma-fence.c to use kunit dma-buf: Change st-dma-fence-unwrap.c to use kunit dma-buf: Change st-dma-fence-chain.c to use kunit dma-buf: Remove the old selftest drivers/dma-buf/.kunitconfig | 2 + drivers/dma-buf/Kconfig | 11 +- drivers/dma-buf/Makefile | 5 +- drivers/dma-buf/selftest.c | 167 --------------- drivers/dma-buf/selftest.h | 30 --- drivers/dma-buf/selftests.h | 16 -- drivers/dma-buf/st-dma-fence-chain.c | 217 +++++++++---------- drivers/dma-buf/st-dma-fence-unwrap.c | 290 +++++++++++--------------- drivers/dma-buf/st-dma-fence.c | 200 ++++++++---------- drivers/dma-buf/st-dma-resv.c | 145 +++++++------ drivers/gpu/drm/i915/Kconfig.debug | 2 +- 11 files changed, 394 insertions(+), 691 deletions(-) create mode 100644 drivers/dma-buf/.kunitconfig delete mode 100644 drivers/dma-buf/selftest.c delete mode 100644 drivers/dma-buf/selftest.h delete mode 100644 drivers/dma-buf/selftests.h base-commit: 41dae5ac5e157b0bb260f381eb3df2f4a4610205 -- 2.43.0

1 month, 2 weeks

[PATCH] mm/vmalloc: map contiguous pages in batches for vmap() whenever possible

by Barry Song

From: Barry Song <v-songbaohua(a)oppo.com> In many cases, the pages passed to vmap() may include high-order pages allocated with __GFP_COMP flags. For example, the systemheap often allocates pages in descending order: order 8, then 4, then 0. Currently, vmap() iterates over every page individually—even pages inside a high-order block are handled one by one. This patch detects high-order pages and maps them as a single contiguous block whenever possible. An alternative would be to implement a new API, vmap_sg(), but that change seems to be large in scope. When vmapping a 128MB dma-buf using the systemheap, this patch makes system_heap_do_vmap() roughly 17× faster. W/ patch: [ 10.404769] system_heap_do_vmap took 2494000 ns [ 12.525921] system_heap_do_vmap took 2467008 ns [ 14.517348] system_heap_do_vmap took 2471008 ns [ 16.593406] system_heap_do_vmap took 2444000 ns [ 19.501341] system_heap_do_vmap took 2489008 ns W/o patch: [ 7.413756] system_heap_do_vmap took 42626000 ns [ 9.425610] system_heap_do_vmap took 42500992 ns [ 11.810898] system_heap_do_vmap took 42215008 ns [ 14.336790] system_heap_do_vmap took 42134992 ns [ 16.373890] system_heap_do_vmap took 42750000 ns Cc: David Hildenbrand <david(a)kernel.org> Cc: Uladzislau Rezki <urezki(a)gmail.com> Cc: Sumit Semwal <sumit.semwal(a)linaro.org> Cc: John Stultz <jstultz(a)google.com> Cc: Maxime Ripard <mripard(a)kernel.org> Tested-by: Tangquan Zheng <zhengtangquan(a)oppo.com> Signed-off-by: Barry Song <v-songbaohua(a)oppo.com> --- * diff with rfc: Many code refinements based on David's suggestions, thanks! Refine comment and changelog according to Uladzislau, thanks! rfc link: https://lore.kernel.org/linux-mm/20251122090343.81243-1-21cnbao@gmail.com/ mm/vmalloc.c | 45 +++++++++++++++++++++++++++++++++++++++------ 1 file changed, 39 insertions(+), 6 deletions(-) diff --git a/mm/vmalloc.c b/mm/vmalloc.c index 41dd01e8430c..8d577767a9e5 100644 --- a/mm/vmalloc.c +++ b/mm/vmalloc.c @@ -642,6 +642,29 @@ static int vmap_small_pages_range_noflush(unsigned long addr, unsigned long end, return err; } +static inline int get_vmap_batch_order(struct page **pages, + unsigned int stride, unsigned int max_steps, unsigned int idx) +{ + int nr_pages = 1; + + /* + * Currently, batching is only supported in vmap_pages_range + * when page_shift == PAGE_SHIFT. + */ + if (stride != 1) + return 0; + + nr_pages = compound_nr(pages[idx]); + if (nr_pages == 1) + return 0; + if (max_steps < nr_pages) + return 0; + + if (num_pages_contiguous(&pages[idx], nr_pages) == nr_pages) + return compound_order(pages[idx]); + return 0; +} + /* * vmap_pages_range_noflush is similar to vmap_pages_range, but does not * flush caches. @@ -655,23 +678,33 @@ int __vmap_pages_range_noflush(unsigned long addr, unsigned long end, pgprot_t prot, struct page **pages, unsigned int page_shift) { unsigned int i, nr = (end - addr) >> PAGE_SHIFT; + unsigned int stride; WARN_ON(page_shift < PAGE_SHIFT); + /* + * For vmap(), users may allocate pages from high orders down to + * order 0, while always using PAGE_SHIFT as the page_shift. + * We first check whether the initial page is a compound page. If so, + * there may be an opportunity to batch multiple pages together. + */ if (!IS_ENABLED(CONFIG_HAVE_ARCH_HUGE_VMALLOC) || - page_shift == PAGE_SHIFT) + (page_shift == PAGE_SHIFT && !PageCompound(pages[0]))) return vmap_small_pages_range_noflush(addr, end, prot, pages); - for (i = 0; i < nr; i += 1U << (page_shift - PAGE_SHIFT)) { - int err; + stride = 1U << (page_shift - PAGE_SHIFT); + for (i = 0; i < nr; ) { + int err, order; - err = vmap_range_noflush(addr, addr + (1UL << page_shift), + order = get_vmap_batch_order(pages, stride, nr - i, i); + err = vmap_range_noflush(addr, addr + (1UL << (page_shift + order)), page_to_phys(pages[i]), prot, - page_shift); + page_shift + order); if (err) return err; - addr += 1UL << page_shift; + addr += 1UL << (page_shift + order); + i += 1U << (order + page_shift - PAGE_SHIFT); } return 0; -- 2.39.3 (Apple Git-146)

1 month, 2 weeks

Understanding Blockchain Forensics:

by joelwest689＠gmail.com

By combining cross-chain tracing, rapid response, and data-driven investigation, Cipher Rescue Chain stands as the global benchmark for crypto recovery. Every traced transaction, every reconstructed path, and every recovered asset reinforces the same conclusion: with the right forensic expertise, recovery is not only possible—it is highly achievable.

1 month, 2 weeks

[PATCH] dma-buf: set SB_I_NOEXEC and SB_I_NODEV on dmabuf filesystem

by Chia-Lin Kao (AceLan)

The VFS now warns if an inode flagged with S_ANON_INODE is located on a filesystem that does not have SB_I_NOEXEC set. dmabuf inodes are created using alloc_anon_inode(), which sets S_ANON_INODE. This triggers a warning in path_noexec() when a dmabuf is mmapped, for example by GStreamer's v4l2src element. [ 60.061328] WARNING: CPU: 2 PID: 2803 at fs/exec.c:125 path_noexec+0xa0/0xd0 ... [ 60.061637] do_mmap+0x2b5/0x680 The warning was introduced by commit 1e7ab6f67824 ("anon_inode: rework assertions") which added enforcement that anonymous inodes must be on filesystems with SB_I_NOEXEC set. Fix this by setting SB_I_NOEXEC and SB_I_NODEV on the dmabuf filesystem context, following the same pattern as commit ce7419b6cf23d ("anon_inode: raise SB_I_NODEV and SB_I_NOEXEC") and commit 98f99394a104c ("secretmem: use SB_I_NOEXEC"). Signed-off-by: Chia-Lin Kao (AceLan) <acelan.kao(a)canonical.com> --- drivers/dma-buf/dma-buf.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/dma-buf/dma-buf.c b/drivers/dma-buf/dma-buf.c index a4d8f2ff94e46..dea79aaab10ce 100644 --- a/drivers/dma-buf/dma-buf.c +++ b/drivers/dma-buf/dma-buf.c @@ -221,6 +221,8 @@ static int dma_buf_fs_init_context(struct fs_context *fc) if (!ctx) return -ENOMEM; ctx->dops = &dma_buf_dentry_ops; + fc->s_iflags |= SB_I_NOEXEC; + fc->s_iflags |= SB_I_NODEV; return 0; } -- 2.51.0

1 month, 2 weeks

[PATCH] dma-fence: Move signalling tracepoint to before ops detach

by Tvrtko Ursulin

Move the signalling tracepoint to before fence->ops are reset otherwise tracepoint will dereference a null pointer. Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin(a)igalia.com> Fixes: 541c8f2468b9 ("dma-buf: detach fence ops on signal v3") Cc: Christian König <christian.koenig(a)amd.com> Cc: Philipp Stanner <phasta(a)kernel.org> Cc: Boris Brezillon <boris.brezillon(a)collabora.com> Cc: linux-media(a)vger.kernel.org Cc: linaro-mm-sig(a)lists.linaro.org --- drivers/dma-buf/dma-fence.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/dma-buf/dma-fence.c b/drivers/dma-buf/dma-fence.c index 1826ba73094c..1c1eaecaf1b0 100644 --- a/drivers/dma-buf/dma-fence.c +++ b/drivers/dma-buf/dma-fence.c @@ -363,6 +363,8 @@ void dma_fence_signal_timestamp_locked(struct dma_fence *fence, &fence->flags))) return; + trace_dma_fence_signaled(fence); + /* * When neither a release nor a wait operation is specified set the ops * pointer to NULL to allow the fence structure to become independent @@ -377,7 +379,6 @@ void dma_fence_signal_timestamp_locked(struct dma_fence *fence, fence->timestamp = timestamp; set_bit(DMA_FENCE_FLAG_TIMESTAMP_BIT, &fence->flags); - trace_dma_fence_signaled(fence); list_for_each_entry_safe(cur, tmp, &cb_list, node) { INIT_LIST_HEAD(&cur->node); -- 2.52.0

1 month, 2 weeks

[PATCH v4 0/8] 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 one is a bit more involved, especially since we have a call from kernel/dma/contiguous.c to the CMA heap code. This was solved by turning the logic around and making the CMA heap call into the contiguous DMA code. 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 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 (8): dma: contiguous: Turn heap registration logic around dma: contiguous: Make dev_get_cma_area() a proper function dma: contiguous: Make dma_contiguous_default_area static dma: contiguous: Export dev_get_cma_area() mm: cma: Export cma_alloc(), cma_release() and cma_get_name() 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 drivers/dma-buf/dma-heap.c | 1 + drivers/dma-buf/heaps/Kconfig | 4 +-- drivers/dma-buf/heaps/cma_heap.c | 22 +++---------- drivers/dma-buf/heaps/system_heap.c | 5 +++ include/linux/dma-buf/heaps/cma.h | 16 --------- include/linux/dma-map-ops.h | 14 ++++---- kernel/dma/contiguous.c | 66 +++++++++++++++++++++++++++++++++---- mm/cma.c | 3 ++ 8 files changed, 82 insertions(+), 49 deletions(-) --- base-commit: c081b71f11732ad2c443f170ab19c3ebe8a1a422 change-id: 20260225-dma-buf-heaps-as-modules-1034b3ec9f2a Best regards, -- Maxime Ripard <mripard(a)kernel.org>

2 months

[PATCH RFC 00/18] accel/qda: Introduce Qualcomm DSP Accelerator driver

by Ekansh Gupta

This patch series introduces the Qualcomm DSP Accelerator (QDA) driver, a modern DRM-based accelerator implementation for Qualcomm Hexagon DSPs. The driver provides a standardized interface for offloading computational tasks to DSPs found on Qualcomm SoCs, supporting all DSP domains (ADSP, CDSP, SDSP, GDSP). The QDA driver is designed as an alternative for the FastRPC driver in drivers/misc/, offering improved resource management, better integration with standard kernel subsystems, and alignment with the Linux kernel's Compute Accelerators framework. User-space staging branch ============ https://github.com/qualcomm/fastrpc/tree/accel/staging Key Features ============ * Standard DRM accelerator interface via /dev/accel/accelN * GEM-based buffer management with DMA-BUF import/export support * IOMMU-based memory isolation using per-process context banks * FastRPC protocol implementation for DSP communication * RPMsg transport layer for reliable message passing * Support for all DSP domains (ADSP, CDSP, SDSP, GDSP) * Comprehensive IOCTL interface for DSP operations High-Level Architecture Differences with Existing FastRPC Driver ================================================================= The QDA driver represents a significant architectural departure from the existing FastRPC driver (drivers/misc/fastrpc.c), addressing several key limitations while maintaining protocol compatibility: 1. DRM Accelerator Framework Integration - FastRPC: Custom character device (/dev/fastrpc-*) - QDA: Standard DRM accel device (/dev/accel/accelN) - Benefit: Leverages established DRM infrastructure for device management. 2. Memory Management - FastRPC: Custom memory allocator with ION/DMA-BUF integration - QDA: Native GEM objects with full PRIME support - Benefit: Seamless buffer sharing using standard DRM mechanisms 3. IOMMU Context Bank Management - FastRPC: Direct IOMMU domain manipulation, limited isolation - QDA: Custom compute bus (qda_cb_bus_type) with proper device model - Benefit: Each CB device is a proper struct device with IOMMU group support, enabling better isolation and resource tracking. - https://lore.kernel.org/all/245d602f-3037-4ae3-9af9-d98f37258aae@oss.qualco… 4. Memory Manager Architecture - FastRPC: Monolithic allocator - QDA: Pluggable memory manager with backend abstraction - Benefit: Currently uses DMA-coherent backend, easily extensible for future memory types (e.g., carveout, CMA) 5. Transport Layer - FastRPC: Direct RPMsg integration in core driver - QDA: Abstracted transport layer (qda_rpmsg.c) - Benefit: Clean separation of concerns, easier to add alternative transports if needed 8. Code Organization - FastRPC: ~3000 lines in single file - QDA: Modular design across multiple files (~4600 lines total) * qda_drv.c: Core driver and DRM integration * qda_gem.c: GEM object management * qda_memory_manager.c: Memory and IOMMU management * qda_fastrpc.c: FastRPC protocol implementation * qda_rpmsg.c: Transport layer * qda_cb.c: Context bank device management - Benefit: Better maintainability, clearer separation of concerns 9. UAPI Design - FastRPC: Custom IOCTL interface - QDA: DRM-style IOCTLs with proper versioning support - Benefit: Follows DRM conventions, easier userspace integration 10. Documentation - FastRPC: Minimal in-tree documentation - QDA: Comprehensive documentation in Documentation/accel/qda/ - Benefit: Better developer experience, clearer API contracts 11. Buffer Reference Mechanism - FastRPC: Uses buffer file descriptors (FDs) for all book-keeping in both kernel and DSP - QDA: Uses GEM handles for kernel-side management, providing better integration with DRM subsystem - Benefit: Leverages DRM GEM infrastructure for reference counting, lifetime management, and integration with other DRM components Key Technical Improvements =========================== * Proper device model: CB devices are real struct device instances on a custom bus, enabling proper IOMMU group management and power management integration * Reference-counted IOMMU devices: Multiple file descriptors from the same process share a single IOMMU device, reducing overhead * GEM-based buffer lifecycle: Automatic cleanup via DRM GEM reference counting, eliminating many resource leak scenarios * Modular memory backends: The memory manager supports pluggable backends, currently implementing DMA-coherent allocations with SID-prefixed addresses for DSP firmware * Context-based invocation tracking: XArray-based context management with proper synchronization and cleanup Patch Series Organization ========================== Patches 1-2: Driver skeleton and documentation Patches 3-6: RPMsg transport and IOMMU/CB infrastructure Patches 7-9: DRM device registration and basic IOCTL Patches 10-12: GEM buffer management and PRIME support Patches 13-17: FastRPC protocol implementation (attach, invoke, create, map/unmap) Patch 18: MAINTAINERS entry Open Items =========== The following items are identified as open items: 1. Privilege Level Management - Currently, daemon processes and user processes have the same access level as both use the same accel device node. This needs to be addressed as daemons attach to privileged DSP PDs and require higher privilege levels for system-level operations - Seeking guidance on the best approach: separate device nodes, capability-based checks, or DRM master/authentication mechanisms 2. UAPI Compatibility Layer - Add UAPI compat layer to facilitate migration of client applications from existing FastRPC UAPI to the new QDA accel driver UAPI, ensuring smooth transition for existing userspace code - Seeking guidance on implementation approach: in-kernel translation layer, userspace wrapper library, or hybrid solution 3. Documentation Improvements - Add detailed IOCTL usage examples - Document DSP firmware interface requirements - Create migration guide from existing FastRPC 4. Per-Domain Memory Allocation - Develop new userspace API to support memory allocation on a per domain basis, enabling domain-specific memory management and optimization 5. Audio and Sensors PD Support - The current patch series does not handle Audio PD and Sensors PD functionalities. These specialized protection domains require additional support for real-time constraints and power management Interface Compatibility ======================== The QDA driver maintains compatibility with existing FastRPC infrastructure: * Device Tree Bindings: The driver uses the same device tree bindings as the existing FastRPC driver, ensuring no changes are required to device tree sources. The "qcom,fastrpc" compatible string and child node structure remain unchanged. * Userspace Interface: While the driver provides a new DRM-based UAPI, the underlying FastRPC protocol and DSP firmware interface remain compatible. This ensures that DSP firmware and libraries continue to work without modification. * Migration Path: The modular design allows for gradual migration, where both drivers can coexist during the transition period. Applications can be migrated incrementally to the new UAPI with the help of the planned compatibility layer. References ========== Previous discussions on this migration: - https://lkml.org/lkml/2024/6/24/479 - https://lkml.org/lkml/2024/6/21/1252 Testing ======= The driver has been tested on Qualcomm platforms with: - Basic FastRPC attach/release operations - DSP process creation and initialization - Memory mapping/unmapping operations - Dynamic invocation with various buffer types - GEM buffer allocation and mmap - PRIME buffer import from other subsystems Signed-off-by: Ekansh Gupta <ekansh.gupta(a)oss.qualcomm.com> --- Ekansh Gupta (18): accel/qda: Add Qualcomm QDA DSP accelerator driver docs accel/qda: Add Qualcomm DSP accelerator driver skeleton accel/qda: Add RPMsg transport for Qualcomm DSP accelerator accel/qda: Add built-in compute CB bus for QDA and integrate with IOMMU accel/qda: Create compute CB devices on QDA compute bus accel/qda: Add memory manager for CB devices accel/qda: Add DRM accel device registration for QDA driver accel/qda: Add per-file DRM context and open/close handling accel/qda: Add QUERY IOCTL and basic QDA UAPI header accel/qda: Add DMA-backed GEM objects and memory manager integration accel/qda: Add GEM_CREATE and GEM_MMAP_OFFSET IOCTLs accel/qda: Add PRIME dma-buf import support accel/qda: Add initial FastRPC attach and release support accel/qda: Add FastRPC dynamic invocation support accel/qda: Add FastRPC DSP process creation support accel/qda: Add FastRPC-based DSP memory mapping support accel/qda: Add FastRPC-based DSP memory unmapping support MAINTAINERS: Add MAINTAINERS entry for QDA driver Documentation/accel/index.rst | 1 + Documentation/accel/qda/index.rst | 14 + Documentation/accel/qda/qda.rst | 129 ++++ MAINTAINERS | 9 + arch/arm64/configs/defconfig | 2 + drivers/accel/Kconfig | 1 + drivers/accel/Makefile | 2 + drivers/accel/qda/Kconfig | 35 ++ drivers/accel/qda/Makefile | 19 + drivers/accel/qda/qda_cb.c | 182 ++++++ drivers/accel/qda/qda_cb.h | 26 + drivers/accel/qda/qda_compute_bus.c | 23 + drivers/accel/qda/qda_drv.c | 375 ++++++++++++ drivers/accel/qda/qda_drv.h | 171 ++++++ drivers/accel/qda/qda_fastrpc.c | 1002 ++++++++++++++++++++++++++++++++ drivers/accel/qda/qda_fastrpc.h | 433 ++++++++++++++ drivers/accel/qda/qda_gem.c | 211 +++++++ drivers/accel/qda/qda_gem.h | 103 ++++ drivers/accel/qda/qda_ioctl.c | 271 +++++++++ drivers/accel/qda/qda_ioctl.h | 118 ++++ drivers/accel/qda/qda_memory_dma.c | 91 +++ drivers/accel/qda/qda_memory_dma.h | 46 ++ drivers/accel/qda/qda_memory_manager.c | 382 ++++++++++++ drivers/accel/qda/qda_memory_manager.h | 148 +++++ drivers/accel/qda/qda_prime.c | 194 +++++++ drivers/accel/qda/qda_prime.h | 43 ++ drivers/accel/qda/qda_rpmsg.c | 327 +++++++++++ drivers/accel/qda/qda_rpmsg.h | 57 ++ drivers/iommu/iommu.c | 4 + include/linux/qda_compute_bus.h | 22 + include/uapi/drm/qda_accel.h | 224 +++++++ 31 files changed, 4665 insertions(+) --- base-commit: d4906ae14a5f136ceb671bb14cedbf13fa560da6 change-id: 20260223-qda-firstpost-4ab05249e2cc Best regards, -- Ekansh Gupta <ekansh.gupta(a)oss.qualcomm.com>

2 months

[PATCH v3 0/6] dma-buf: heaps: add coherent reserved-memory heap

by Albert Esteve

This patch series adds a new dma-buf heap driver that exposes coherent, non‑reusable reserved-memory regions as named heaps, so userspace can explicitly allocate buffers from those device‑specific pools. Motivation: we want cgroup accounting for all userspace‑visible buffer allocations (DRM, v4l2, dma‑buf heaps, etc.). That’s hard to do when drivers call dma_alloc_attrs() directly because the accounting controller (memcg vs dmem) is ambiguous. The long‑term plan is to steer those paths toward dma‑buf heaps, where each heap can unambiguously charge a single controller. To reach that goal, we need a heap backend for each dma_alloc_attrs() memory type. CMA and system heaps already exist; coherent reserved‑memory was the missing piece, since many SoCs define dedicated, device‑local coherent pools in DT under /reserved-memory using "shared-dma-pool" with non‑reusable regions (i.e., not CMA) that are carved out exclusively for coherent DMA and are currently only usable by in‑kernel drivers. Because these regions are device‑dependent, each heap instance binds a heap device to its reserved‑mem region via a newly introduced helper function -namely, of_reserved_mem_device_init_with_mem()- so coherent allocations use the correct dev->dma_mem. Charging to cgroups for these buffers is intentionally left out to keep review focused on the new heap; I plan to follow up based on Eric’s [1] and Maxime’s [2] work on dmem charging from userspace. This series also makes the new heap driver modular, in line with the CMA heap change in [3]. [1] https://lore.kernel.org/all/20260218-dmabuf-heap-cma-dmem-v2-0-b249886fb7b2… [2] https://lore.kernel.org/all/20250310-dmem-cgroups-v1-0-2984c1bc9312@kernel.… [3] https://lore.kernel.org/all/20260303-dma-buf-heaps-as-modules-v3-0-24344812… Signed-off-by: Albert Esteve <aesteve(a)redhat.com> --- Changes in v3: - Reorganized changesets among patches to ensure bisectability - Removed unused dma_heap_coherent_register() leftover - Removed fallback when setting mask in coherent heap dev, since dma_set_mask() already truncates to supported masks - Moved struct rmem_assigned_device (rd) logic to of_reserved_mem_device_init_with_mem() to allow listing the device - Link to v2: https://lore.kernel.org/r/20260303-b4-dmabuf-heap-coherent-rmem-v2-0-65a465… Changes in v2: - Removed dmem charging parts - Moved coherent heap registering logic to coherent.c - Made heap device a member of struct dma_heap - Split dma_heap_add logic into create/register, to be able to access the stored heap device before registered. - Avoid platform device in favour of heap device - Added a wrapper to rmem device_init() op - Switched from late_initcall() to module_init() - Made the coherent heap driver modular - Link to v1: https://lore.kernel.org/r/20260224-b4-dmabuf-heap-coherent-rmem-v1-1-dffef4… --- Albert Esteve (5): dma-buf: dma-heap: split dma_heap_add of_reserved_mem: add a helper for rmem device_init op dma: coherent: store reserved memory coherent regions dma-buf: heaps: Add Coherent heap to dmabuf heaps dma-buf: heaps: coherent: Turn heap into a module John Stultz (1): dma-buf: dma-heap: Keep track of the heap device struct drivers/dma-buf/dma-heap.c | 138 +++++++++-- drivers/dma-buf/heaps/Kconfig | 9 + drivers/dma-buf/heaps/Makefile | 1 + drivers/dma-buf/heaps/coherent_heap.c | 417 ++++++++++++++++++++++++++++++++++ drivers/of/of_reserved_mem.c | 68 ++++-- include/linux/dma-heap.h | 5 + include/linux/dma-map-ops.h | 7 + include/linux/of_reserved_mem.h | 8 + kernel/dma/coherent.c | 34 +++ 9 files changed, 640 insertions(+), 47 deletions(-) --- base-commit: 6de23f81a5e08be8fbf5e8d7e9febc72a5b5f27f change-id: 20260223-b4-dmabuf-heap-coherent-rmem-91fd3926afe9 Best regards, -- Albert Esteve <aesteve(a)redhat.com>

2 months

[PATCH] udmabuf: fix DMA direction mismatch in release_udmabuf()

by Mikhail Gavrilov

begin_cpu_udmabuf() maps the sg_table with the caller-provided direction (e.g., DMA_TO_DEVICE for a write-only sync), and caches it in ubuf->sg for reuse. However, release_udmabuf() always unmaps this sg_table with a hardcoded DMA_BIDIRECTIONAL, regardless of the direction that was originally used for the mapping. With CONFIG_DMA_API_DEBUG=y this produces: DMA-API: misc udmabuf: device driver frees DMA memory with different direction [device address=0x000000044a123000] [size=4096 bytes] [mapped with DMA_TO_DEVICE] [unmapped with DMA_BIDIRECTIONAL] The issue was found during video playback when GStreamer performed a write-only DMA_BUF_IOCTL_SYNC on a udmabuf. It can be reproduced with CONFIG_DMA_API_DEBUG=y by creating a udmabuf from a memfd, performing a write-only sync (DMA_BUF_SYNC_WRITE without DMA_BUF_SYNC_READ), and closing the file descriptor. Fix this by storing the DMA direction used when the sg_table is first created in begin_cpu_udmabuf(), and passing that same direction to put_sg_table() in release_udmabuf(). Fixes: 284562e1f348 ("udmabuf: implement begin_cpu_access/end_cpu_access hooks") Cc: stable(a)vger.kernel.org Signed-off-by: Mikhail Gavrilov <mikhail.v.gavrilov(a)gmail.com> --- drivers/dma-buf/udmabuf.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/dma-buf/udmabuf.c b/drivers/dma-buf/udmabuf.c index 94b8ecb892bb..d0836febefdd 100644 --- a/drivers/dma-buf/udmabuf.c +++ b/drivers/dma-buf/udmabuf.c @@ -40,6 +40,7 @@ struct udmabuf { struct folio **pinned_folios; struct sg_table *sg; + enum dma_data_direction sg_dir; struct miscdevice *device; pgoff_t *offsets; }; @@ -235,7 +236,7 @@ static void release_udmabuf(struct dma_buf *buf) struct device *dev = ubuf->device->this_device; if (ubuf->sg) - put_sg_table(dev, ubuf->sg, DMA_BIDIRECTIONAL); + put_sg_table(dev, ubuf->sg, ubuf->sg_dir); deinit_udmabuf(ubuf); kfree(ubuf); @@ -253,6 +254,8 @@ static int begin_cpu_udmabuf(struct dma_buf *buf, if (IS_ERR(ubuf->sg)) { ret = PTR_ERR(ubuf->sg); ubuf->sg = NULL; + } else { + ubuf->sg_dir = direction; } } else { dma_sync_sgtable_for_cpu(dev, ubuf->sg, direction); -- 2.53.0

2 months

[RFC v2 PATCH 00/10] vfio/pci: Add mmap() for DMABUFs

by Matt Evans

Hi all, There were various suggestions in the September 2025 thread "[TECH TOPIC] vfio, iommufd: Enabling user space drivers to vend more granular access to client processes" [0], and LPC discussions, around improving the situation for multi-process userspace driver designs. This RFC series implements some of these ideas. (Thanks for feedback on v1! Revised series, with changes noted inline.) Background: Multi-process USDs ============================== The userspace driver scenario discussed in that thread involves a primary process driving a PCIe function through VFIO/iommufd, which manages the function-wide ownership/lifecycle. The function is designed to provide multiple distinct programming interfaces (for example, several independent MMIO register frames in one function), and the primary process delegates control of these interfaces to multiple independent client processes (which do the actual work). This scenario clearly relies on a HW design that provides appropriate isolation between the programming interfaces. The two key needs are: 1. Mechanisms to safely delegate a subset of the device MMIO resources to a client process without over-sharing wider access (or influence over whole-device activities, such as reset). 2. Mechanisms to allow a client process to do its own iommufd management w.r.t. its address space, in a way that's isolated from DMA relating to other clients. mmap() of VFIO DMABUFs ====================== This RFC addresses #1 in "vfio/pci: Support mmap() of a VFIO DMABUF", implementing the proposals in [0] to add mmap() support to the existing VFIO DMABUF exporter. This enables a userspace driver to define DMABUF ranges corresponding to sub-ranges of a BAR, and grant a given client (via a shared fd) the capability to access (only) those sub-ranges. The VFIO device fds would be kept private to the primary process. All the client can do with that fd is map (or iomap via iommufd) that specific subset of resources, and the impact of bugs/malice is contained. (We'll follow up on #2 separately, as a related-but-distinct problem. PASIDs are one way to achieve per-client isolation of DMA; another could be sharing of a single IOVA space via 'constrained' iommufds.) New in v2: To achieve this, the existing VFIO BAR mmap() path is converted to use DMABUFs behind the scenes, in "vfio/pci: Convert BAR mmap() to use a DMABUF" plus new helper functions, as Jason/Christian suggested in the v1 discussion [3]. This means: - Both regular and new DMABUF BAR mappings share the same vm_ops, i.e. mmap()ing DMABUFs is a smaller change on top of the existing mmap(). - The zapping of mappings occurs via vfio_pci_dma_buf_move(), and the vfio_pci_zap_bars() originally paired with the _move()s can go away. Each DMABUF has a unique address_space. - It's a step towards future iommufd VFIO Type1 emulation implementing P2P, since iommufd can now get a DMABUF from a VA that it's mapping for IO; the VMAs' vm_file is that of the backing DMABUF. Revocation/reclaim ================== Mapping a BAR subset is useful, but the lifetime of access granted to a client needs to be managed well. For example, a protocol between the primary process and the client can indicate when the client is done, and when it's safe to reuse the resources elsewhere, but cleanup can't practically be cooperative. For robustness, we enable the driver to make the resources guaranteed-inaccessible when it chooses, so that it can re-assign them to other uses in future. "vfio/pci: Permanently revoke a DMABUF on request" adds a new VFIO device fd ioctl, VFIO_DEVICE_PCI_DMABUF_REVOKE. This takes a DMABUF fd parameter previously exported (from that device!) and permanently revokes the DMABUF. This notifies/detaches importers, zaps PTEs for any mappings, and guarantees no future attachment/import/map/access is possible by any means. A primary driver process would use this operation when the client's tenure ends to reclaim "loaned-out" MMIO interfaces, at which point the interfaces could be safely re-used. New in v2: ioctl() on VFIO driver fd, rather than DMABUF fd. A DMABUF is revoked using code common to vfio_pci_dma_buf_move(), selectively zapping mappings (after waiting for completion on the dma_buf_invalidate_mappings() request). BAR mapping access attributes ============================= Inspired by Alex [Mastro] and Jason's comments in [0] and Mahmoud's work in [1] with the goal of controlling CPU access attributes for VFIO BAR mappings (e.g. WC), we can decorate DMABUFs with access attributes that are then used by a mapping's PTEs. I've proposed reserving a field in struct vfio_device_feature_dma_buf's flags to specify an attribute for its ranges. Although that keeps the (UAPI) struct unchanged, it means all ranges in a DMABUF share the same attribute. I feel a single attribute-to-mmap() relation is logical/reasonable. An application can also create multiple DMABUFs to describe any BAR layout and mix of attributes. Tests ===== (Still sharing the [RFC ONLY] userspace test/demo program for context, not for merge.) It illustrates & tests various map/revoke cases, but doesn't use the existing VFIO selftests and relies on a (tweaked) QEMU EDU function. I'm (still) working on integrating the scenarios into the existing VFIO selftests. 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...). Fin === v2 is based on next-20260310 (to build on Leon's recent series "vfio: Wait for dma-buf invalidation to complete" [2]). Please share your thoughts! I'd like to de-RFC if we feel this approach is now fair. Many thanks, Matt References: [0]: https://lore.kernel.org/linux-iommu/20250918214425.2677057-1-amastro@fb.com/ [1]: https://lore.kernel.org/all/20250804104012.87915-1-mngyadam@amazon.de/ [2]: https://lore.kernel.org/linux-iommu/20260205-nocturnal-poetic-chamois-f566a… [3]: https://lore.kernel.org/all/20260226202211.929005-1-mattev@meta.com/ -------------------------------------------------------------------------------- Changelog: v2: Respin based on the feedback/suggestions: - Transform the existing VFIO BAR mmap path to also use DMABUFs behind the scenes, and then simply share that code for explicitly-mapped DMABUFs. - Refactors the export itself out of vfio_pci_core_feature_dma_buf, and shared by a new vfio_pci_core_mmap_prep_dmabuf helper used by the regular VFIO mmap to create a DMABUF. - Revoke buffers using a VFIO device fd ioctl v1: https://lore.kernel.org/all/20260226202211.929005-1-mattev@meta.com/ Matt Evans (10): vfio/pci: Set up VFIO barmap before creating a DMABUF vfio/pci: Clean up DMABUFs before disabling function vfio/pci: Add 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: Remove vfio_pci_zap_bars() vfio/pci: Support mmap() of a VFIO DMABUF vfio/pci: Permanently revoke a DMABUF on request vfio/pci: Add mmap() attributes to DMABUF feature [RFC ONLY] selftests: vfio: Add standalone vfio_dmabuf_mmap_test drivers/vfio/pci/Kconfig | 3 +- drivers/vfio/pci/Makefile | 3 +- drivers/vfio/pci/vfio_pci_config.c | 18 +- drivers/vfio/pci/vfio_pci_core.c | 123 +-- drivers/vfio/pci/vfio_pci_dmabuf.c | 425 +++++++-- drivers/vfio/pci/vfio_pci_priv.h | 46 +- include/uapi/linux/vfio.h | 42 +- tools/testing/selftests/vfio/Makefile | 1 + .../vfio/standalone/vfio_dmabuf_mmap_test.c | 837 ++++++++++++++++++ 9 files changed, 1339 insertions(+), 159 deletions(-) create mode 100644 tools/testing/selftests/vfio/standalone/vfio_dmabuf_mmap_test.c -- 2.47.3

2 months

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

Linaro-mm-sig March 2026