From ae4814a3aab54ce548950161937176b7f9ec6f77 Mon Sep 17 00:00:00 2001 From: Pei Xiao Date: Mon, 7 Apr 2025 09:53:28 +0800 Subject: [PATCH 01/16] iommu: remove unneeded semicolon cocci warnings: drivers/iommu/dma-iommu.c:1788:2-3: Unneeded semicolon so remove unneeded semicolon to fix cocci warnings. Signed-off-by: Pei Xiao Reviewed-by: Jason Gunthorpe Link: https://lore.kernel.org/r/tencent_73EEE47E6ECCF538229C9B9E6A0272DA2B05@qq.com Signed-off-by: Joerg Roedel --- drivers/iommu/dma-iommu.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/iommu/dma-iommu.c b/drivers/iommu/dma-iommu.c index cb7e29dcac15..a775e4dbe06f 100644 --- a/drivers/iommu/dma-iommu.c +++ b/drivers/iommu/dma-iommu.c @@ -1754,7 +1754,7 @@ static size_t cookie_msi_granule(const struct iommu_domain *domain) return PAGE_SIZE; default: BUG(); - }; + } } static struct list_head *cookie_msi_pages(const struct iommu_domain *domain) @@ -1766,7 +1766,7 @@ static struct list_head *cookie_msi_pages(const struct iommu_domain *domain) return &domain->msi_cookie->msi_page_list; default: BUG(); - }; + } } static struct iommu_dma_msi_page *iommu_dma_get_msi_page(struct device *dev, -- 2.50.1 From 767e22001dfce64cc03b7def1562338591ab6031 Mon Sep 17 00:00:00 2001 From: Nicolin Chen Date: Mon, 7 Apr 2025 13:19:08 -0700 Subject: [PATCH 02/16] iommu/tegra241-cmdqv: Fix warnings due to dmam_free_coherent() Two WARNINGs are observed when SMMU driver rolls back upon failure: arm-smmu-v3.9.auto: Failed to register iommu arm-smmu-v3.9.auto: probe with driver arm-smmu-v3 failed with error -22 ------------[ cut here ]------------ WARNING: CPU: 5 PID: 1 at kernel/dma/mapping.c:74 dmam_free_coherent+0xc0/0xd8 Call trace: dmam_free_coherent+0xc0/0xd8 (P) tegra241_vintf_free_lvcmdq+0x74/0x188 tegra241_cmdqv_remove_vintf+0x60/0x148 tegra241_cmdqv_remove+0x48/0xc8 arm_smmu_impl_remove+0x28/0x60 devm_action_release+0x1c/0x40 ------------[ cut here ]------------ 128 pages are still in use! WARNING: CPU: 16 PID: 1 at mm/page_alloc.c:6902 free_contig_range+0x18c/0x1c8 Call trace: free_contig_range+0x18c/0x1c8 (P) cma_release+0x154/0x2f0 dma_free_contiguous+0x38/0xa0 dma_direct_free+0x10c/0x248 dma_free_attrs+0x100/0x290 dmam_free_coherent+0x78/0xd8 tegra241_vintf_free_lvcmdq+0x74/0x160 tegra241_cmdqv_remove+0x98/0x198 arm_smmu_impl_remove+0x28/0x60 devm_action_release+0x1c/0x40 This is because the LVCMDQ queue memory are managed by devres, while that dmam_free_coherent() is called in the context of devm_action_release(). Jason pointed out that "arm_smmu_impl_probe() has mis-ordered the devres callbacks if ops->device_remove() is going to be manually freeing things that probe allocated": https://lore.kernel.org/linux-iommu/20250407174408.GB1722458@nvidia.com/ In fact, tegra241_cmdqv_init_structures() only allocates memory resources which means any failure that it generates would be similar to -ENOMEM, so there is no point in having that "falling back to standard SMMU" routine, as the standard SMMU would likely fail to allocate memory too. Remove the unwind part in tegra241_cmdqv_init_structures(), and return a proper error code to ask SMMU driver to call tegra241_cmdqv_remove() via impl_ops->device_remove(). Then, drop tegra241_vintf_free_lvcmdq() since devres will take care of that. Fixes: 483e0bd8883a ("iommu/tegra241-cmdqv: Do not allocate vcmdq until dma_set_mask_and_coherent") Cc: stable@vger.kernel.org Suggested-by: Jason Gunthorpe Signed-off-by: Nicolin Chen Reviewed-by: Jason Gunthorpe Link: https://lore.kernel.org/r/20250407201908.172225-1-nicolinc@nvidia.com Signed-off-by: Joerg Roedel --- .../iommu/arm/arm-smmu-v3/tegra241-cmdqv.c | 32 +++---------------- 1 file changed, 5 insertions(+), 27 deletions(-) diff --git a/drivers/iommu/arm/arm-smmu-v3/tegra241-cmdqv.c b/drivers/iommu/arm/arm-smmu-v3/tegra241-cmdqv.c index d525ab43a4ae..dd7d030d2e89 100644 --- a/drivers/iommu/arm/arm-smmu-v3/tegra241-cmdqv.c +++ b/drivers/iommu/arm/arm-smmu-v3/tegra241-cmdqv.c @@ -487,17 +487,6 @@ static int tegra241_cmdqv_hw_reset(struct arm_smmu_device *smmu) /* VCMDQ Resource Helpers */ -static void tegra241_vcmdq_free_smmu_cmdq(struct tegra241_vcmdq *vcmdq) -{ - struct arm_smmu_queue *q = &vcmdq->cmdq.q; - size_t nents = 1 << q->llq.max_n_shift; - size_t qsz = nents << CMDQ_ENT_SZ_SHIFT; - - if (!q->base) - return; - dmam_free_coherent(vcmdq->cmdqv->smmu.dev, qsz, q->base, q->base_dma); -} - static int tegra241_vcmdq_alloc_smmu_cmdq(struct tegra241_vcmdq *vcmdq) { struct arm_smmu_device *smmu = &vcmdq->cmdqv->smmu; @@ -560,7 +549,8 @@ static void tegra241_vintf_free_lvcmdq(struct tegra241_vintf *vintf, u16 lidx) struct tegra241_vcmdq *vcmdq = vintf->lvcmdqs[lidx]; char header[64]; - tegra241_vcmdq_free_smmu_cmdq(vcmdq); + /* Note that the lvcmdq queue memory space is managed by devres */ + tegra241_vintf_deinit_lvcmdq(vintf, lidx); dev_dbg(vintf->cmdqv->dev, @@ -768,13 +758,13 @@ static int tegra241_cmdqv_init_structures(struct arm_smmu_device *smmu) vintf = kzalloc(sizeof(*vintf), GFP_KERNEL); if (!vintf) - goto out_fallback; + return -ENOMEM; /* Init VINTF0 for in-kernel use */ ret = tegra241_cmdqv_init_vintf(cmdqv, 0, vintf); if (ret) { dev_err(cmdqv->dev, "failed to init vintf0: %d\n", ret); - goto free_vintf; + return ret; } /* Preallocate logical VCMDQs to VINTF0 */ @@ -783,24 +773,12 @@ static int tegra241_cmdqv_init_structures(struct arm_smmu_device *smmu) vcmdq = tegra241_vintf_alloc_lvcmdq(vintf, lidx); if (IS_ERR(vcmdq)) - goto free_lvcmdq; + return PTR_ERR(vcmdq); } /* Now, we are ready to run all the impl ops */ smmu->impl_ops = &tegra241_cmdqv_impl_ops; return 0; - -free_lvcmdq: - for (lidx--; lidx >= 0; lidx--) - tegra241_vintf_free_lvcmdq(vintf, lidx); - tegra241_cmdqv_deinit_vintf(cmdqv, vintf->idx); -free_vintf: - kfree(vintf); -out_fallback: - dev_info(smmu->impl_dev, "Falling back to standard SMMU CMDQ\n"); - smmu->options &= ~ARM_SMMU_OPT_TEGRA241_CMDQV; - tegra241_cmdqv_remove(smmu); - return 0; } #ifdef CONFIG_IOMMU_DEBUGFS -- 2.50.1 From 3b607b75a345b1d808031bf1bb1038e4dac8d521 Mon Sep 17 00:00:00 2001 From: Thorsten Blum Date: Thu, 10 Apr 2025 17:47:23 +0200 Subject: [PATCH 03/16] null_blk: Use strscpy() instead of strscpy_pad() in null_add_dev() blk_mq_alloc_disk() already zero-initializes the destination buffer, making strscpy() sufficient for safely copying the disk's name. The additional NUL-padding performed by strscpy_pad() is unnecessary. If the destination buffer has a fixed length, strscpy() automatically determines its size using sizeof() when the argument is omitted. This makes the explicit size argument unnecessary. The source string is also NUL-terminated and meets the __must_be_cstr() requirement of strscpy(). No functional changes intended. Signed-off-by: Thorsten Blum Reviewed-by: Zhu Yanjun Reviewed-by: Damien Le Moal Link: https://lore.kernel.org/r/20250410154727.883207-1-thorsten.blum@linux.dev Signed-off-by: Jens Axboe --- drivers/block/null_blk/main.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/block/null_blk/main.c b/drivers/block/null_blk/main.c index 3bb9cee0a9b5..aa163ae9b2aa 100644 --- a/drivers/block/null_blk/main.c +++ b/drivers/block/null_blk/main.c @@ -2031,7 +2031,7 @@ static int null_add_dev(struct nullb_device *dev) nullb->disk->minors = 1; nullb->disk->fops = &null_ops; nullb->disk->private_data = nullb; - strscpy_pad(nullb->disk->disk_name, nullb->disk_name, DISK_NAME_LEN); + strscpy(nullb->disk->disk_name, nullb->disk_name); if (nullb->dev->zoned) { rv = null_register_zoned_dev(nullb); -- 2.50.1 From a650d38915c194b87616a0747a339b20958d17db Mon Sep 17 00:00:00 2001 From: Kumar Kartikeya Dwivedi Date: Fri, 11 Apr 2025 03:17:59 -0700 Subject: [PATCH 04/16] bpf: Convert ringbuf map to rqspinlock MIME-Version: 1.0 Content-Type: text/plain; charset=utf8 Content-Transfer-Encoding: 8bit Convert the raw spinlock used by BPF ringbuf to rqspinlock. Currently, we have an open syzbot report of a potential deadlock. In addition, the ringbuf can fail to reserve spuriously under contention from NMI context. It is potentially attractive to enable unconstrained usage (incl. NMIs) while ensuring no deadlocks manifest at runtime, perform the conversion to rqspinlock to achieve this. This change was benchmarked for BPF ringbuf's multi-producer contention case on an Intel Sapphire Rapids server, with hyperthreading disabled and performance governor turned on. 5 warm up runs were done for each case before obtaining the results. Before (raw_spinlock_t): Ringbuf, multi-producer contention ================================== rb-libbpf nr_prod 1 11.440 ± 0.019M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 2 2.706 ± 0.010M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 3 3.130 ± 0.004M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 4 2.472 ± 0.003M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 8 2.352 ± 0.001M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 12 2.813 ± 0.001M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 16 1.988 ± 0.001M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 20 2.245 ± 0.001M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 24 2.148 ± 0.001M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 28 2.190 ± 0.001M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 32 2.490 ± 0.001M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 36 2.180 ± 0.001M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 40 2.201 ± 0.001M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 44 2.226 ± 0.001M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 48 2.164 ± 0.001M/s (drops 0.000 ± 0.000M/s) rb-libbpf nr_prod 52 1.874 ± 0.001M/s (drops 0.000 ± 0.000M/s) After (rqspinlock_t): Ringbuf, multi-producer contention ================================== rb-libbpf nr_prod 1 11.078 ± 0.019M/s (drops 0.000 ± 0.000M/s) (-3.16%) rb-libbpf nr_prod 2 2.801 ± 0.014M/s (drops 0.000 ± 0.000M/s) (3.51%) rb-libbpf nr_prod 3 3.454 ± 0.005M/s (drops 0.000 ± 0.000M/s) (10.35%) rb-libbpf nr_prod 4 2.567 ± 0.002M/s (drops 0.000 ± 0.000M/s) (3.84%) rb-libbpf nr_prod 8 2.468 ± 0.001M/s (drops 0.000 ± 0.000M/s) (4.93%) rb-libbpf nr_prod 12 2.510 ± 0.001M/s (drops 0.000 ± 0.000M/s) (-10.77%) rb-libbpf nr_prod 16 2.075 ± 0.001M/s (drops 0.000 ± 0.000M/s) (4.38%) rb-libbpf nr_prod 20 2.640 ± 0.001M/s (drops 0.000 ± 0.000M/s) (17.59%) rb-libbpf nr_prod 24 2.092 ± 0.001M/s (drops 0.000 ± 0.000M/s) (-2.61%) rb-libbpf nr_prod 28 2.426 ± 0.005M/s (drops 0.000 ± 0.000M/s) (10.78%) rb-libbpf nr_prod 32 2.331 ± 0.004M/s (drops 0.000 ± 0.000M/s) (-6.39%) rb-libbpf nr_prod 36 2.306 ± 0.003M/s (drops 0.000 ± 0.000M/s) (5.78%) rb-libbpf nr_prod 40 2.178 ± 0.002M/s (drops 0.000 ± 0.000M/s) (-1.04%) rb-libbpf nr_prod 44 2.293 ± 0.001M/s (drops 0.000 ± 0.000M/s) (3.01%) rb-libbpf nr_prod 48 2.022 ± 0.001M/s (drops 0.000 ± 0.000M/s) (-6.56%) rb-libbpf nr_prod 52 1.809 ± 0.001M/s (drops 0.000 ± 0.000M/s) (-3.47%) There's a fair amount of noise in the benchmark, with numbers on reruns going up and down by 10%, so all changes are in the range of this disturbance, and we see no major regressions. Reported-by: syzbot+850aaf14624dc0c6d366@syzkaller.appspotmail.com Closes: https://lore.kernel.org/all/0000000000004aa700061379547e@google.com Signed-off-by: Kumar Kartikeya Dwivedi Link: https://lore.kernel.org/r/20250411101759.4061366-1-memxor@gmail.com Signed-off-by: Alexei Starovoitov --- kernel/bpf/ringbuf.c | 17 +++++++---------- 1 file changed, 7 insertions(+), 10 deletions(-) diff --git a/kernel/bpf/ringbuf.c b/kernel/bpf/ringbuf.c index 1499d8caa9a3..719d73299397 100644 --- a/kernel/bpf/ringbuf.c +++ b/kernel/bpf/ringbuf.c @@ -11,6 +11,7 @@ #include #include #include +#include #define RINGBUF_CREATE_FLAG_MASK (BPF_F_NUMA_NODE) @@ -29,7 +30,7 @@ struct bpf_ringbuf { u64 mask; struct page **pages; int nr_pages; - raw_spinlock_t spinlock ____cacheline_aligned_in_smp; + rqspinlock_t spinlock ____cacheline_aligned_in_smp; /* For user-space producer ring buffers, an atomic_t busy bit is used * to synchronize access to the ring buffers in the kernel, rather than * the spinlock that is used for kernel-producer ring buffers. This is @@ -173,7 +174,7 @@ static struct bpf_ringbuf *bpf_ringbuf_alloc(size_t data_sz, int numa_node) if (!rb) return NULL; - raw_spin_lock_init(&rb->spinlock); + raw_res_spin_lock_init(&rb->spinlock); atomic_set(&rb->busy, 0); init_waitqueue_head(&rb->waitq); init_irq_work(&rb->work, bpf_ringbuf_notify); @@ -416,12 +417,8 @@ static void *__bpf_ringbuf_reserve(struct bpf_ringbuf *rb, u64 size) cons_pos = smp_load_acquire(&rb->consumer_pos); - if (in_nmi()) { - if (!raw_spin_trylock_irqsave(&rb->spinlock, flags)) - return NULL; - } else { - raw_spin_lock_irqsave(&rb->spinlock, flags); - } + if (raw_res_spin_lock_irqsave(&rb->spinlock, flags)) + return NULL; pend_pos = rb->pending_pos; prod_pos = rb->producer_pos; @@ -446,7 +443,7 @@ static void *__bpf_ringbuf_reserve(struct bpf_ringbuf *rb, u64 size) */ if (new_prod_pos - cons_pos > rb->mask || new_prod_pos - pend_pos > rb->mask) { - raw_spin_unlock_irqrestore(&rb->spinlock, flags); + raw_res_spin_unlock_irqrestore(&rb->spinlock, flags); return NULL; } @@ -458,7 +455,7 @@ static void *__bpf_ringbuf_reserve(struct bpf_ringbuf *rb, u64 size) /* pairs with consumer's smp_load_acquire() */ smp_store_release(&rb->producer_pos, new_prod_pos); - raw_spin_unlock_irqrestore(&rb->spinlock, flags); + raw_res_spin_unlock_irqrestore(&rb->spinlock, flags); return (void *)hdr + BPF_RINGBUF_HDR_SZ; } -- 2.50.1 From c8ba3f8aff672a5ea36e895f0f8f657271d855d7 Mon Sep 17 00:00:00 2001 From: Zhangfei Gao Date: Mon, 17 Mar 2025 01:13:52 +0000 Subject: [PATCH 05/16] PCI: Run quirk_huawei_pcie_sva() before arm_smmu_probe_device() quirk_huawei_pcie_sva() sets properties needed by arm_smmu_probe_device(), but bcb81ac6ae3c ("iommu: Get DT/ACPI parsing into the proper probe path") changed the iommu_probe_device() flow so arm_smmu_probe_device() is now invoked before the quirk, leading to failures like this: reg-dummy reg-dummy: late IOMMU probe at driver bind, something fishy here! WARNING: CPU: 0 PID: 1 at drivers/iommu/iommu.c:449 __iommu_probe_device+0x140/0x570 RIP: 0010:__iommu_probe_device+0x140/0x570 The SR-IOV enumeration ordering changes like this: pci_iov_add_virtfn pci_device_add pci_fixup_device(pci_fixup_header) <-- device_add bus_notify iommu_bus_notifier + iommu_probe_device + arm_smmu_probe_device pci_bus_add_device pci_fixup_device(pci_fixup_final) <-- device_attach driver_probe_device really_probe pci_dma_configure acpi_dma_configure_id - iommu_probe_device - arm_smmu_probe_device The non-SR-IOV case is similar in that pci_device_add() is called from pci_scan_single_device() in the generic enumeration path and pci_bus_add_device() is called later, after all host bridges have been enumerated. Declare quirk_huawei_pcie_sva() as a header fixup to ensure that it happens before arm_smmu_probe_device(). Fixes: bcb81ac6ae3c ("iommu: Get DT/ACPI parsing into the proper probe path") Reported-by: Chaitanya Kumar Borah Closes: https://lore.kernel.org/all/SJ1PR11MB61295DE21A1184AEE0786E25B9D22@SJ1PR11MB6129.namprd11.prod.outlook.com/ Signed-off-by: Zhangfei Gao [bhelgaas: commit log, add failure info and reporter] Signed-off-by: Bjorn Helgaas Link: https://patch.msgid.link/20250317011352.5806-1-zhangfei.gao@linaro.org --- drivers/pci/quirks.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c index 8d610c17e0f2..94daca15a096 100644 --- a/drivers/pci/quirks.c +++ b/drivers/pci/quirks.c @@ -1990,12 +1990,12 @@ static void quirk_huawei_pcie_sva(struct pci_dev *pdev) device_create_managed_software_node(&pdev->dev, properties, NULL)) pci_warn(pdev, "could not add stall property"); } -DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_HUAWEI, 0xa250, quirk_huawei_pcie_sva); -DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_HUAWEI, 0xa251, quirk_huawei_pcie_sva); -DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_HUAWEI, 0xa255, quirk_huawei_pcie_sva); -DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_HUAWEI, 0xa256, quirk_huawei_pcie_sva); -DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_HUAWEI, 0xa258, quirk_huawei_pcie_sva); -DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_HUAWEI, 0xa259, quirk_huawei_pcie_sva); +DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_HUAWEI, 0xa250, quirk_huawei_pcie_sva); +DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_HUAWEI, 0xa251, quirk_huawei_pcie_sva); +DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_HUAWEI, 0xa255, quirk_huawei_pcie_sva); +DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_HUAWEI, 0xa256, quirk_huawei_pcie_sva); +DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_HUAWEI, 0xa258, quirk_huawei_pcie_sva); +DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_HUAWEI, 0xa259, quirk_huawei_pcie_sva); /* * It's possible for the MSI to get corrupted if SHPC and ACPI are used -- 2.50.1 From 04a80a34c22f4db245f553d8696d1318d1c00ece Mon Sep 17 00:00:00 2001 From: Andy Chiu Date: Wed, 9 Apr 2025 00:02:57 +0800 Subject: [PATCH 06/16] ftrace: Properly merge notrace hashes The global notrace hash should be jointly decided by the intersection of each subops's notrace hash, but not the filter hash. Cc: stable@vger.kernel.org Link: https://lore.kernel.org/20250408160258.48563-1-andybnac@gmail.com Fixes: 5fccc7552ccb ("ftrace: Add subops logic to allow one ops to manage many") Signed-off-by: Andy Chiu [ fixed removing of freeing of filter_hash ] Signed-off-by: Steven Rostedt (Google) --- kernel/trace/ftrace.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 1a48aedb5255..8939eeebb02e 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -3526,16 +3526,16 @@ int ftrace_startup_subops(struct ftrace_ops *ops, struct ftrace_ops *subops, int ftrace_hash_empty(subops->func_hash->notrace_hash)) { notrace_hash = EMPTY_HASH; } else { - size_bits = max(ops->func_hash->filter_hash->size_bits, - subops->func_hash->filter_hash->size_bits); + size_bits = max(ops->func_hash->notrace_hash->size_bits, + subops->func_hash->notrace_hash->size_bits); notrace_hash = alloc_ftrace_hash(size_bits); if (!notrace_hash) { free_ftrace_hash(filter_hash); return -ENOMEM; } - ret = intersect_hash(¬race_hash, ops->func_hash->filter_hash, - subops->func_hash->filter_hash); + ret = intersect_hash(¬race_hash, ops->func_hash->notrace_hash, + subops->func_hash->notrace_hash); if (ret < 0) { free_ftrace_hash(filter_hash); free_ftrace_hash(notrace_hash); -- 2.50.1 From 0ae6b8ce200da00a78f33c055fdc4fe3225d22ec Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Wed, 9 Apr 2025 11:15:50 -0400 Subject: [PATCH 07/16] ftrace: Fix accounting of subop hashes The function graph infrastructure uses ftrace to hook to functions. It has a single ftrace_ops to manage all the users of function graph. Each individual user (tracing, bpf, fprobes, etc) has its own ftrace_ops to track the functions it will have its callback called from. These ftrace_ops are "subops" to the main ftrace_ops of the function graph infrastructure. Each ftrace_ops has a filter_hash and a notrace_hash that is defined as: Only trace functions that are in the filter_hash but not in the notrace_hash. If the filter_hash is empty, it means to trace all functions. If the notrace_hash is empty, it means do not disable any function. The function graph main ftrace_ops needs to be a superset containing all the functions to be traced by all the subops it has. The algorithm to perform this merge was incorrect. When the first subops was added to the main ops, it simply made the main ops a copy of the subops (same filter_hash and notrace_hash). When a second ops was added, it joined the new subops filter_hash with the main ops filter_hash as a union of the two sets. The intersect between the new subops notrace_hash and the main ops notrace_hash was created as the new notrace_hash of the main ops. The issue here is that it would then start tracing functions than no subops were tracing. For example if you had two subops that had: subops 1: filter_hash = '*sched*' # trace all functions with "sched" in it notrace_hash = '*time*' # except do not trace functions with "time" subops 2: filter_hash = '*lock*' # trace all functions with "lock" in it notrace_hash = '*clock*' # except do not trace functions with "clock" The intersect of '*time*' functions with '*clock*' functions could be the empty set. That means the main ops will be tracing all functions with '*time*' and all "*clock*" in it! Instead, modify the algorithm to be a bit simpler and correct. First, when adding a new subops, even if it's the first one, do not add the notrace_hash if the filter_hash is not empty. Instead, just add the functions that are in the filter_hash of the subops but not in the notrace_hash of the subops into the main ops filter_hash. There's no reason to add anything to the main ops notrace_hash. The notrace_hash of the main ops should only be non empty iff all subops filter_hashes are empty (meaning to trace all functions) and all subops notrace_hashes include the same functions. That is, the main ops notrace_hash is empty if any subops filter_hash is non empty. The main ops notrace_hash only has content in it if all subops filter_hashes are empty, and the content are only functions that intersect all the subops notrace_hashes. If any subops notrace_hash is empty, then so is the main ops notrace_hash. Cc: Masami Hiramatsu Cc: Mark Rutland Cc: Mathieu Desnoyers Cc: Andrew Morton Cc: Shuah Khan Cc: Andy Chiu Link: https://lore.kernel.org/20250409152720.216356767@goodmis.org Signed-off-by: Steven Rostedt (Google) --- kernel/trace/ftrace.c | 314 ++++++++++++++++++++++++------------------ 1 file changed, 177 insertions(+), 137 deletions(-) diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 8939eeebb02e..a8a02868b435 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -3255,6 +3255,31 @@ static int append_hash(struct ftrace_hash **hash, struct ftrace_hash *new_hash, return 0; } +/* + * Remove functions from @hash that are in @notrace_hash + */ +static void remove_hash(struct ftrace_hash *hash, struct ftrace_hash *notrace_hash) +{ + struct ftrace_func_entry *entry; + struct hlist_node *tmp; + int size; + int i; + + /* If the notrace hash is empty, there's nothing to do */ + if (ftrace_hash_empty(notrace_hash)) + return; + + size = 1 << hash->size_bits; + for (i = 0; i < size; i++) { + hlist_for_each_entry_safe(entry, tmp, &hash->buckets[i], hlist) { + if (!__ftrace_lookup_ip(notrace_hash, entry->ip)) + continue; + remove_hash_entry(hash, entry); + kfree(entry); + } + } +} + /* * Add to @hash only those that are in both @new_hash1 and @new_hash2 * @@ -3295,67 +3320,6 @@ static int intersect_hash(struct ftrace_hash **hash, struct ftrace_hash *new_has return 0; } -/* Return a new hash that has a union of all @ops->filter_hash entries */ -static struct ftrace_hash *append_hashes(struct ftrace_ops *ops) -{ - struct ftrace_hash *new_hash = NULL; - struct ftrace_ops *subops; - int size_bits; - int ret; - - if (ops->func_hash->filter_hash) - size_bits = ops->func_hash->filter_hash->size_bits; - else - size_bits = FTRACE_HASH_DEFAULT_BITS; - - list_for_each_entry(subops, &ops->subop_list, list) { - ret = append_hash(&new_hash, subops->func_hash->filter_hash, size_bits); - if (ret < 0) { - free_ftrace_hash(new_hash); - return NULL; - } - /* Nothing more to do if new_hash is empty */ - if (ftrace_hash_empty(new_hash)) - break; - } - /* Can't return NULL as that means this failed */ - return new_hash ? : EMPTY_HASH; -} - -/* Make @ops trace evenything except what all its subops do not trace */ -static struct ftrace_hash *intersect_hashes(struct ftrace_ops *ops) -{ - struct ftrace_hash *new_hash = NULL; - struct ftrace_ops *subops; - int size_bits; - int ret; - - list_for_each_entry(subops, &ops->subop_list, list) { - struct ftrace_hash *next_hash; - - if (!new_hash) { - size_bits = subops->func_hash->notrace_hash->size_bits; - new_hash = alloc_and_copy_ftrace_hash(size_bits, ops->func_hash->notrace_hash); - if (!new_hash) - return NULL; - continue; - } - size_bits = new_hash->size_bits; - next_hash = new_hash; - new_hash = alloc_ftrace_hash(size_bits); - ret = intersect_hash(&new_hash, next_hash, subops->func_hash->notrace_hash); - free_ftrace_hash(next_hash); - if (ret < 0) { - free_ftrace_hash(new_hash); - return NULL; - } - /* Nothing more to do if new_hash is empty */ - if (ftrace_hash_empty(new_hash)) - break; - } - return new_hash; -} - static bool ops_equal(struct ftrace_hash *A, struct ftrace_hash *B) { struct ftrace_func_entry *entry; @@ -3427,6 +3391,93 @@ static int ftrace_update_ops(struct ftrace_ops *ops, struct ftrace_hash *filter_ return 0; } +static int add_first_hash(struct ftrace_hash **filter_hash, struct ftrace_hash **notrace_hash, + struct ftrace_ops_hash *func_hash) +{ + /* If the filter hash is not empty, simply remove the nohash from it */ + if (!ftrace_hash_empty(func_hash->filter_hash)) { + *filter_hash = copy_hash(func_hash->filter_hash); + if (!*filter_hash) + return -ENOMEM; + remove_hash(*filter_hash, func_hash->notrace_hash); + *notrace_hash = EMPTY_HASH; + + } else { + *notrace_hash = copy_hash(func_hash->notrace_hash); + if (!*notrace_hash) + return -ENOMEM; + *filter_hash = EMPTY_HASH; + } + return 0; +} + +static int add_next_hash(struct ftrace_hash **filter_hash, struct ftrace_hash **notrace_hash, + struct ftrace_ops_hash *ops_hash, struct ftrace_ops_hash *subops_hash) +{ + int size_bits; + int ret; + + /* If the subops trace all functions so must the main ops */ + if (ftrace_hash_empty(ops_hash->filter_hash) || + ftrace_hash_empty(subops_hash->filter_hash)) { + *filter_hash = EMPTY_HASH; + } else { + /* + * The main ops filter hash is not empty, so its + * notrace_hash had better be, as the notrace hash + * is only used for empty main filter hashes. + */ + WARN_ON_ONCE(!ftrace_hash_empty(ops_hash->notrace_hash)); + + size_bits = max(ops_hash->filter_hash->size_bits, + subops_hash->filter_hash->size_bits); + + /* Copy the subops hash */ + *filter_hash = alloc_and_copy_ftrace_hash(size_bits, subops_hash->filter_hash); + if (!filter_hash) + return -ENOMEM; + /* Remove any notrace functions from the copy */ + remove_hash(*filter_hash, subops_hash->notrace_hash); + + ret = append_hash(filter_hash, ops_hash->filter_hash, + size_bits); + if (ret < 0) { + free_ftrace_hash(*filter_hash); + return ret; + } + } + + /* + * Only process notrace hashes if the main filter hash is empty + * (tracing all functions), otherwise the filter hash will just + * remove the notrace hash functions, and the notrace hash is + * not needed. + */ + if (ftrace_hash_empty(*filter_hash)) { + /* + * Intersect the notrace functions. That is, if two + * subops are not tracing a set of functions, the + * main ops will only not trace the functions that are + * in both subops, but has to trace the functions that + * are only notrace in one of the subops, for the other + * subops to be able to trace them. + */ + size_bits = max(ops_hash->notrace_hash->size_bits, + subops_hash->notrace_hash->size_bits); + *notrace_hash = alloc_ftrace_hash(size_bits); + if (!*notrace_hash) + return -ENOMEM; + + ret = intersect_hash(notrace_hash, ops_hash->notrace_hash, + subops_hash->notrace_hash); + if (ret < 0) { + free_ftrace_hash(*notrace_hash); + return ret; + } + } + return 0; +} + /** * ftrace_startup_subops - enable tracing for subops of an ops * @ops: Manager ops (used to pick all the functions of its subops) @@ -3443,7 +3494,6 @@ int ftrace_startup_subops(struct ftrace_ops *ops, struct ftrace_ops *subops, int struct ftrace_hash *notrace_hash; struct ftrace_hash *save_filter_hash; struct ftrace_hash *save_notrace_hash; - int size_bits; int ret; if (unlikely(ftrace_disabled)) @@ -3467,14 +3517,14 @@ int ftrace_startup_subops(struct ftrace_ops *ops, struct ftrace_ops *subops, int /* For the first subops to ops just enable it normally */ if (list_empty(&ops->subop_list)) { - /* Just use the subops hashes */ - filter_hash = copy_hash(subops->func_hash->filter_hash); - notrace_hash = copy_hash(subops->func_hash->notrace_hash); - if (!filter_hash || !notrace_hash) { - free_ftrace_hash(filter_hash); - free_ftrace_hash(notrace_hash); - return -ENOMEM; - } + + /* The ops was empty, should have empty hashes */ + WARN_ON_ONCE(!ftrace_hash_empty(ops->func_hash->filter_hash)); + WARN_ON_ONCE(!ftrace_hash_empty(ops->func_hash->notrace_hash)); + + ret = add_first_hash(&filter_hash, ¬race_hash, subops->func_hash); + if (ret < 0) + return ret; save_filter_hash = ops->func_hash->filter_hash; save_notrace_hash = ops->func_hash->notrace_hash; @@ -3500,48 +3550,16 @@ int ftrace_startup_subops(struct ftrace_ops *ops, struct ftrace_ops *subops, int /* * Here there's already something attached. Here are the rules: - * o If either filter_hash is empty then the final stays empty - * o Otherwise, the final is a superset of both hashes - * o If either notrace_hash is empty then the final stays empty - * o Otherwise, the final is an intersection between the hashes + * If the new subops and main ops filter hashes are not empty: + * o Make a copy of the subops filter hash + * o Remove all functions in the nohash from it. + * o Add in the main hash filter functions + * o Remove any of these functions from the main notrace hash */ - if (ftrace_hash_empty(ops->func_hash->filter_hash) || - ftrace_hash_empty(subops->func_hash->filter_hash)) { - filter_hash = EMPTY_HASH; - } else { - size_bits = max(ops->func_hash->filter_hash->size_bits, - subops->func_hash->filter_hash->size_bits); - filter_hash = alloc_and_copy_ftrace_hash(size_bits, ops->func_hash->filter_hash); - if (!filter_hash) - return -ENOMEM; - ret = append_hash(&filter_hash, subops->func_hash->filter_hash, - size_bits); - if (ret < 0) { - free_ftrace_hash(filter_hash); - return ret; - } - } - - if (ftrace_hash_empty(ops->func_hash->notrace_hash) || - ftrace_hash_empty(subops->func_hash->notrace_hash)) { - notrace_hash = EMPTY_HASH; - } else { - size_bits = max(ops->func_hash->notrace_hash->size_bits, - subops->func_hash->notrace_hash->size_bits); - notrace_hash = alloc_ftrace_hash(size_bits); - if (!notrace_hash) { - free_ftrace_hash(filter_hash); - return -ENOMEM; - } - ret = intersect_hash(¬race_hash, ops->func_hash->notrace_hash, - subops->func_hash->notrace_hash); - if (ret < 0) { - free_ftrace_hash(filter_hash); - free_ftrace_hash(notrace_hash); - return ret; - } - } + ret = add_next_hash(&filter_hash, ¬race_hash, ops->func_hash, subops->func_hash); + if (ret < 0) + return ret; list_add(&subops->list, &ops->subop_list); @@ -3557,6 +3575,42 @@ int ftrace_startup_subops(struct ftrace_ops *ops, struct ftrace_ops *subops, int return ret; } +static int rebuild_hashes(struct ftrace_hash **filter_hash, struct ftrace_hash **notrace_hash, + struct ftrace_ops *ops) +{ + struct ftrace_ops_hash temp_hash; + struct ftrace_ops *subops; + bool first = true; + int ret; + + temp_hash.filter_hash = EMPTY_HASH; + temp_hash.notrace_hash = EMPTY_HASH; + + list_for_each_entry(subops, &ops->subop_list, list) { + *filter_hash = EMPTY_HASH; + *notrace_hash = EMPTY_HASH; + + if (first) { + ret = add_first_hash(filter_hash, notrace_hash, subops->func_hash); + if (ret < 0) + return ret; + first = false; + } else { + ret = add_next_hash(filter_hash, notrace_hash, + &temp_hash, subops->func_hash); + if (ret < 0) { + free_ftrace_hash(temp_hash.filter_hash); + free_ftrace_hash(temp_hash.notrace_hash); + return ret; + } + } + + temp_hash.filter_hash = *filter_hash; + temp_hash.notrace_hash = *notrace_hash; + } + return 0; +} + /** * ftrace_shutdown_subops - Remove a subops from a manager ops * @ops: A manager ops to remove @subops from @@ -3605,14 +3659,9 @@ int ftrace_shutdown_subops(struct ftrace_ops *ops, struct ftrace_ops *subops, in } /* Rebuild the hashes without subops */ - filter_hash = append_hashes(ops); - notrace_hash = intersect_hashes(ops); - if (!filter_hash || !notrace_hash) { - free_ftrace_hash(filter_hash); - free_ftrace_hash(notrace_hash); - list_add(&subops->list, &ops->subop_list); - return -ENOMEM; - } + ret = rebuild_hashes(&filter_hash, ¬race_hash, ops); + if (ret < 0) + return ret; ret = ftrace_update_ops(ops, filter_hash, notrace_hash); if (ret < 0) { @@ -3628,11 +3677,11 @@ int ftrace_shutdown_subops(struct ftrace_ops *ops, struct ftrace_ops *subops, in static int ftrace_hash_move_and_update_subops(struct ftrace_ops *subops, struct ftrace_hash **orig_subhash, - struct ftrace_hash *hash, - int enable) + struct ftrace_hash *hash) { struct ftrace_ops *ops = subops->managed; - struct ftrace_hash **orig_hash; + struct ftrace_hash *notrace_hash; + struct ftrace_hash *filter_hash; struct ftrace_hash *save_hash; struct ftrace_hash *new_hash; int ret; @@ -3649,24 +3698,15 @@ static int ftrace_hash_move_and_update_subops(struct ftrace_ops *subops, return -ENOMEM; } - /* Create a new_hash to hold the ops new functions */ - if (enable) { - orig_hash = &ops->func_hash->filter_hash; - new_hash = append_hashes(ops); - } else { - orig_hash = &ops->func_hash->notrace_hash; - new_hash = intersect_hashes(ops); - } - - /* Move the hash over to the new hash */ - ret = __ftrace_hash_move_and_update_ops(ops, orig_hash, new_hash, enable); - - free_ftrace_hash(new_hash); + ret = rebuild_hashes(&filter_hash, ¬race_hash, ops); + if (!ret) + ret = ftrace_update_ops(ops, filter_hash, notrace_hash); if (ret) { /* Put back the original hash */ - free_ftrace_hash_rcu(*orig_subhash); + new_hash = *orig_subhash; *orig_subhash = save_hash; + free_ftrace_hash_rcu(new_hash); } else { free_ftrace_hash_rcu(save_hash); } @@ -4890,7 +4930,7 @@ static int ftrace_hash_move_and_update_ops(struct ftrace_ops *ops, int enable) { if (ops->flags & FTRACE_OPS_FL_SUBOP) - return ftrace_hash_move_and_update_subops(ops, orig_hash, hash, enable); + return ftrace_hash_move_and_update_subops(ops, orig_hash, hash); /* * If this ops is not enabled, it could be sharing its filters @@ -4909,7 +4949,7 @@ static int ftrace_hash_move_and_update_ops(struct ftrace_ops *ops, list_for_each_entry(subops, &op->subop_list, list) { if ((subops->flags & FTRACE_OPS_FL_ENABLED) && subops->func_hash == ops->func_hash) { - return ftrace_hash_move_and_update_subops(subops, orig_hash, hash, enable); + return ftrace_hash_move_and_update_subops(subops, orig_hash, hash); } } } while_for_each_ftrace_op(op); -- 2.50.1 From a1fc89d409d8fd927622c238b7c7d719e9ecab3d Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Wed, 9 Apr 2025 11:15:51 -0400 Subject: [PATCH 08/16] tracing/selftest: Add test to better test subops filtering of function graph A bug was discovered that showed the accounting of the subops of the ftrace_ops filtering was incorrect. Add a new test to better test the filtering. This test creates two instances, where it will add various filters to both the set_ftrace_filter and the set_ftrace_notrace files and enable function_graph. Then it looks into the enabled_functions file to make sure that the filters are behaving correctly. Cc: Masami Hiramatsu Cc: Mark Rutland Cc: Mathieu Desnoyers Cc: Andrew Morton Cc: Shuah Khan Cc: Andy Chiu Link: https://lore.kernel.org/20250409152720.380778379@goodmis.org Signed-off-by: Steven Rostedt (Google) --- .../test.d/ftrace/fgraph-multi-filter.tc | 177 ++++++++++++++++++ 1 file changed, 177 insertions(+) create mode 100644 tools/testing/selftests/ftrace/test.d/ftrace/fgraph-multi-filter.tc diff --git a/tools/testing/selftests/ftrace/test.d/ftrace/fgraph-multi-filter.tc b/tools/testing/selftests/ftrace/test.d/ftrace/fgraph-multi-filter.tc new file mode 100644 index 000000000000..b6d6a312ead5 --- /dev/null +++ b/tools/testing/selftests/ftrace/test.d/ftrace/fgraph-multi-filter.tc @@ -0,0 +1,177 @@ +#!/bin/sh +# SPDX-License-Identifier: GPL-2.0 +# description: ftrace - function graph filters +# requires: set_ftrace_filter function_graph:tracer + +# Make sure that function graph filtering works + +INSTANCE1="instances/test1_$$" +INSTANCE2="instances/test2_$$" + +WD=`pwd` + +do_reset() { + cd $WD + if [ -d $INSTANCE1 ]; then + echo nop > $INSTANCE1/current_tracer + rmdir $INSTANCE1 + fi + if [ -d $INSTANCE2 ]; then + echo nop > $INSTANCE2/current_tracer + rmdir $INSTANCE2 + fi +} + +mkdir $INSTANCE1 +if ! grep -q function_graph $INSTANCE1/available_tracers; then + echo "function_graph not allowed with instances" + rmdir $INSTANCE1 + exit_unsupported +fi + +mkdir $INSTANCE2 + +fail() { # msg + do_reset + echo $1 + exit_fail +} + +disable_tracing +clear_trace + +function_count() { + search=$1 + vsearch=$2 + + if [ -z "$search" ]; then + cat enabled_functions | wc -l + elif [ -z "$vsearch" ]; then + grep $search enabled_functions | wc -l + else + grep $search enabled_functions | grep $vsearch| wc -l + fi +} + +set_fgraph() { + instance=$1 + filter="$2" + notrace="$3" + + echo "$filter" > $instance/set_ftrace_filter + echo "$notrace" > $instance/set_ftrace_notrace + echo function_graph > $instance/current_tracer +} + +check_functions() { + orig_cnt=$1 + test=$2 + + cnt=`function_count $test` + if [ $cnt -gt $orig_cnt ]; then + fail + fi +} + +check_cnt() { + orig_cnt=$1 + search=$2 + vsearch=$3 + + cnt=`function_count $search $vsearch` + if [ $cnt -gt $orig_cnt ]; then + fail + fi +} + +reset_graph() { + instance=$1 + echo nop > $instance/current_tracer +} + +# get any functions that were enabled before the test +total_cnt=`function_count` +sched_cnt=`function_count sched` +lock_cnt=`function_count lock` +time_cnt=`function_count time` +clock_cnt=`function_count clock` +locks_clock_cnt=`function_count locks clock` +clock_locks_cnt=`function_count clock locks` + +# Trace functions with "sched" but not "time" +set_fgraph $INSTANCE1 '*sched*' '*time*' + +# Make sure "time" isn't listed +check_functions $time_cnt 'time' +instance1_cnt=`function_count` + +# Trace functions with "lock" but not "clock" +set_fgraph $INSTANCE2 '*lock*' '*clock*' +instance1_2_cnt=`function_count` + +# Turn off the first instance +reset_graph $INSTANCE1 + +# The second instance doesn't trace "clock" functions +check_functions $clock_cnt 'clock' +instance2_cnt=`function_count` + +# Start from a clean slate +reset_graph $INSTANCE2 +check_functions $total_cnt + +# Trace functions with "lock" but not "clock" +set_fgraph $INSTANCE2 '*lock*' '*clock*' + +# This should match the last time instance 2 was by itself +cnt=`function_count` +if [ $instance2_cnt -ne $cnt ]; then + fail +fi + +# And it should not be tracing "clock" functions +check_functions $clock_cnt 'clock' + +# Trace functions with "sched" but not "time" +set_fgraph $INSTANCE1 '*sched*' '*time*' + +# This should match the last time both instances were enabled +cnt=`function_count` +if [ $instance1_2_cnt -ne $cnt ]; then + fail +fi + +# Turn off the second instance +reset_graph $INSTANCE2 + +# This should match the last time instance 1 was by itself +cnt=`function_count` +if [ $instance1_cnt -ne $cnt ]; then + fail +fi + +# And it should not be tracing "time" functions +check_functions $time_cnt 'time' + +# Start from a clean slate +reset_graph $INSTANCE1 +check_functions $total_cnt + +# Enable all functions but those that have "locks" +set_fgraph $INSTANCE1 '' '*locks*' + +# Enable all functions but those that have "clock" +set_fgraph $INSTANCE2 '' '*clock*' + +# If a function has "locks" it should not have "clock" +check_cnt $locks_clock_cnt locks clock + +# If a function has "clock" it should not have "locks" +check_cnt $clock_locks_cnt clock locks + +reset_graph $INSTANCE1 +reset_graph $INSTANCE2 + +do_reset + +exit 0 -- 2.50.1 From 485acd207d7daf8cf941a5f0fd0c09bc6d049402 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Fri, 11 Apr 2025 13:30:15 -0400 Subject: [PATCH 09/16] ftrace: Do not have print_graph_retval() add a newline The retval and retaddr options for function_graph tracer will add a comment at the end of a function for both leaf and non leaf functions that looks like: __wake_up_common(); /* ret=0x1 */ } /* pick_next_task_fair ret=0x0 */ The function print_graph_retval() adds a newline after the "*/". But if that's not called, the caller function needs to make sure there's a newline added. This is confusing and when the function parameters code was added, it added a newline even when calling print_graph_retval() as the fact that the print_graph_retval() function prints a newline isn't obvious. This caused an extra newline to be printed and that made it fail the selftests when the retval option was set, as the selftests were not expecting blank lines being injected into the trace. Instead of having print_graph_retval() print a newline, just have the caller always print the newline regardless if it calls print_graph_retval() or not. This not only fixes this bug, but it also simplifies the code. Cc: Masami Hiramatsu Cc: Mark Rutland Cc: Mathieu Desnoyers Link: https://lore.kernel.org/20250411133015.015ca393@gandalf.local.home Reported-by: Mark Brown Tested-by: Mark Brown Closes: https://lore.kernel.org/all/ccc40f2b-4b9e-4abd-8daf-d22fce2a86f0@sirena.org.uk/ Fixes: ff5c9c576e754 ("ftrace: Add support for function argument to graph tracer") Signed-off-by: Steven Rostedt (Google) --- kernel/trace/trace_functions_graph.c | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c index 2f077d4158e5..0c357a89c58e 100644 --- a/kernel/trace/trace_functions_graph.c +++ b/kernel/trace/trace_functions_graph.c @@ -880,8 +880,6 @@ static void print_graph_retval(struct trace_seq *s, struct ftrace_graph_ent_entr if (print_retval || print_retaddr) trace_seq_puts(s, " /*"); - else - trace_seq_putc(s, '\n'); } else { print_retaddr = false; trace_seq_printf(s, "} /* %ps", func); @@ -899,7 +897,7 @@ static void print_graph_retval(struct trace_seq *s, struct ftrace_graph_ent_entr } if (!entry || print_retval || print_retaddr) - trace_seq_puts(s, " */\n"); + trace_seq_puts(s, " */"); } #else @@ -975,7 +973,7 @@ print_graph_entry_leaf(struct trace_iterator *iter, } else trace_seq_puts(s, "();"); } - trace_seq_printf(s, "\n"); + trace_seq_putc(s, '\n'); print_graph_irq(iter, graph_ret->func, TRACE_GRAPH_RET, cpu, iter->ent->pid, flags); @@ -1313,10 +1311,11 @@ print_graph_return(struct ftrace_graph_ret_entry *retentry, struct trace_seq *s, * that if the funcgraph-tail option is enabled. */ if (func_match && !(flags & TRACE_GRAPH_PRINT_TAIL)) - trace_seq_puts(s, "}\n"); + trace_seq_puts(s, "}"); else - trace_seq_printf(s, "} /* %ps */\n", (void *)func); + trace_seq_printf(s, "} /* %ps */", (void *)func); } + trace_seq_putc(s, '\n'); /* Overrun */ if (flags & TRACE_GRAPH_PRINT_OVERRUN) -- 2.50.1 From 8d7861ac507d23024c7d74b6cb59a9cca248bcb7 Mon Sep 17 00:00:00 2001 From: Nam Cao Date: Fri, 11 Apr 2025 09:37:17 +0200 Subject: [PATCH 10/16] rv: Fix out-of-bound memory access in rv_is_container_monitor() When rv_is_container_monitor() is called on the last monitor in rv_monitors_list, KASAN yells: BUG: KASAN: global-out-of-bounds in rv_is_container_monitor+0x101/0x110 Read of size 8 at addr ffffffff97c7c798 by task setup/221 The buggy address belongs to the variable: rv_monitors_list+0x18/0x40 This is due to list_next_entry() is called on the last entry in the list. It wraps around to the first list_head, and the first list_head is not embedded in struct rv_monitor_def. Fix it by checking if the monitor is last in the list. Cc: stable@vger.kernel.org Cc: Gabriele Monaco Fixes: cb85c660fcd4 ("rv: Add option for nested monitors and include sched") Link: https://lore.kernel.org/e85b5eeb7228bfc23b8d7d4ab5411472c54ae91b.1744355018.git.namcao@linutronix.de Signed-off-by: Nam Cao Signed-off-by: Steven Rostedt (Google) --- kernel/trace/rv/rv.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/kernel/trace/rv/rv.c b/kernel/trace/rv/rv.c index 968c5c3b0246..e4077500a91d 100644 --- a/kernel/trace/rv/rv.c +++ b/kernel/trace/rv/rv.c @@ -225,7 +225,12 @@ bool rv_is_nested_monitor(struct rv_monitor_def *mdef) */ bool rv_is_container_monitor(struct rv_monitor_def *mdef) { - struct rv_monitor_def *next = list_next_entry(mdef, list); + struct rv_monitor_def *next; + + if (list_is_last(&mdef->list, &rv_monitors_list)) + return false; + + next = list_next_entry(mdef, list); return next->parent == mdef->monitor || !mdef->monitor->enable; } -- 2.50.1 From ce7e8a65aa1b7e8a6833403b314fa8f2cf133119 Mon Sep 17 00:00:00 2001 From: Tom Vierjahn Date: Mon, 24 Mar 2025 23:09:30 +0100 Subject: [PATCH 11/16] Documentation: ext4: Add fields to ext4_super_block documentation Documentation and implementation of the ext4 super block have slightly diverged: Padding has been removed in order to make room for new fields that are still missing in the documentation. Add the new fields s_encryption_level, s_first_error_errorcode, s_last_error_errorcode to the documentation of the ext4 super block. Fixes: f542fbe8d5e8 ("ext4 crypto: reserve codepoints used by the ext4 encryption feature") Fixes: 878520ac45f9 ("ext4: save the error code which triggered an ext4_error() in the superblock") Signed-off-by: Tom Vierjahn Reviewed-by: Ojaswin Mujoo Link: https://patch.msgid.link/20250324221004.5268-1-tom.vierjahn@acm.org Signed-off-by: Theodore Ts'o --- Documentation/filesystems/ext4/super.rst | 20 ++++++++++++++------ 1 file changed, 14 insertions(+), 6 deletions(-) diff --git a/Documentation/filesystems/ext4/super.rst b/Documentation/filesystems/ext4/super.rst index a1eb4a11a1d0..1b240661bfa3 100644 --- a/Documentation/filesystems/ext4/super.rst +++ b/Documentation/filesystems/ext4/super.rst @@ -328,9 +328,13 @@ The ext4 superblock is laid out as follows in - s_checksum_type - Metadata checksum algorithm type. The only valid value is 1 (crc32c). * - 0x176 - - __le16 - - s_reserved_pad - - + - \_\_u8 + - s\_encryption\_level + - Versioning level for encryption. + * - 0x177 + - \_\_u8 + - s\_reserved\_pad + - Padding to next 32bits. * - 0x178 - __le64 - s_kbytes_written @@ -466,9 +470,13 @@ The ext4 superblock is laid out as follows in - s_last_error_time_hi - Upper 8 bits of the s_last_error_time field. * - 0x27A - - __u8 - - s_pad[2] - - Zero padding. + - \_\_u8 + - s\_first\_error\_errcode + - + * - 0x27B + - \_\_u8 + - s\_last\_error\_errcode + - * - 0x27C - __le16 - s_encoding -- 2.50.1 From 7e50bbb134aba1df0854f171b596b3a42d35605a Mon Sep 17 00:00:00 2001 From: "Gustavo A. R. Silva" Date: Wed, 26 Mar 2025 16:55:51 -0600 Subject: [PATCH 12/16] ext4: avoid -Wflex-array-member-not-at-end warning -Wflex-array-member-not-at-end was introduced in GCC-14, and we are getting ready to enable it, globally. Use the `DEFINE_RAW_FLEX()` helper for an on-stack definition of a flexible structure where the size of the flexible-array member is known at compile-time, and refactor the rest of the code, accordingly. So, with these changes, fix the following warning: fs/ext4/mballoc.c:3041:40: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end] Signed-off-by: Gustavo A. R. Silva Reviewed-by: Kees Cook Link: https://patch.msgid.link/Z-SF97N3AxcIMlSi@kspp Signed-off-by: Theodore Ts'o --- fs/ext4/mballoc.c | 18 ++++++++---------- 1 file changed, 8 insertions(+), 10 deletions(-) diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 0d523e9fb3d5..f88424c28194 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -3037,10 +3037,8 @@ static int ext4_mb_seq_groups_show(struct seq_file *seq, void *v) unsigned char blocksize_bits = min_t(unsigned char, sb->s_blocksize_bits, EXT4_MAX_BLOCK_LOG_SIZE); - struct sg { - struct ext4_group_info info; - ext4_grpblk_t counters[EXT4_MAX_BLOCK_LOG_SIZE + 2]; - } sg; + DEFINE_RAW_FLEX(struct ext4_group_info, sg, bb_counters, + EXT4_MAX_BLOCK_LOG_SIZE + 2); group--; if (group == 0) @@ -3048,7 +3046,7 @@ static int ext4_mb_seq_groups_show(struct seq_file *seq, void *v) " 2^0 2^1 2^2 2^3 2^4 2^5 2^6 " " 2^7 2^8 2^9 2^10 2^11 2^12 2^13 ]\n"); - i = (blocksize_bits + 2) * sizeof(sg.info.bb_counters[0]) + + i = (blocksize_bits + 2) * sizeof(sg->bb_counters[0]) + sizeof(struct ext4_group_info); grinfo = ext4_get_group_info(sb, group); @@ -3068,14 +3066,14 @@ static int ext4_mb_seq_groups_show(struct seq_file *seq, void *v) * We care only about free space counters in the group info and * these are safe to access even after the buddy has been unloaded */ - memcpy(&sg, grinfo, i); - seq_printf(seq, "#%-5u: %-5u %-5u %-5u [", group, sg.info.bb_free, - sg.info.bb_fragments, sg.info.bb_first_free); + memcpy(sg, grinfo, i); + seq_printf(seq, "#%-5u: %-5u %-5u %-5u [", group, sg->bb_free, + sg->bb_fragments, sg->bb_first_free); for (i = 0; i <= 13; i++) seq_printf(seq, " %-5u", i <= blocksize_bits + 1 ? - sg.info.bb_counters[i] : 0); + sg->bb_counters[i] : 0); seq_puts(seq, " ]"); - if (EXT4_MB_GRP_BBITMAP_CORRUPT(&sg.info)) + if (EXT4_MB_GRP_BBITMAP_CORRUPT(sg)) seq_puts(seq, " Block bitmap corrupted!"); seq_putc(seq, '\n'); return 0; -- 2.50.1 From ccad447a3d331a239477c281533bacb585b54a98 Mon Sep 17 00:00:00 2001 From: Ojaswin Mujoo Date: Fri, 28 Mar 2025 11:54:52 +0530 Subject: [PATCH 13/16] ext4: make block validity check resistent to sb bh corruption Block validity checks need to be skipped in case they are called for journal blocks since they are part of system's protected zone. Currently, this is done by checking inode->ino against sbi->s_es->s_journal_inum, which is a direct read from the ext4 sb buffer head. If someone modifies this underneath us then the s_journal_inum field might get corrupted. To prevent against this, change the check to directly compare the inode with journal->j_inode. **Slight change in behavior**: During journal init path, check_block_validity etc might be called for journal inode when sbi->s_journal is not set yet. In this case we now proceed with ext4_inode_block_valid() instead of returning early. Since systems zones have not been set yet, it is okay to proceed so we can perform basic checks on the blocks. Suggested-by: Baokun Li Reviewed-by: Baokun Li Reviewed-by: Jan Kara Reviewed-by: Zhang Yi Signed-off-by: Ojaswin Mujoo Link: https://patch.msgid.link/0c06bc9ebfcd6ccfed84a36e79147bf45ff5adc1.1743142920.git.ojaswin@linux.ibm.com Signed-off-by: Theodore Ts'o --- fs/ext4/block_validity.c | 5 ++--- fs/ext4/inode.c | 7 ++++--- 2 files changed, 6 insertions(+), 6 deletions(-) diff --git a/fs/ext4/block_validity.c b/fs/ext4/block_validity.c index 87ee3a17bd29..e8c5525afc67 100644 --- a/fs/ext4/block_validity.c +++ b/fs/ext4/block_validity.c @@ -351,10 +351,9 @@ int ext4_check_blockref(const char *function, unsigned int line, { __le32 *bref = p; unsigned int blk; + journal_t *journal = EXT4_SB(inode->i_sb)->s_journal; - if (ext4_has_feature_journal(inode->i_sb) && - (inode->i_ino == - le32_to_cpu(EXT4_SB(inode->i_sb)->s_es->s_journal_inum))) + if (journal && inode == journal->j_inode) return 0; while (bref < p+max) { diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 5c57a34cd82c..f386de8c12f6 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -384,10 +384,11 @@ static int __check_block_validity(struct inode *inode, const char *func, unsigned int line, struct ext4_map_blocks *map) { - if (ext4_has_feature_journal(inode->i_sb) && - (inode->i_ino == - le32_to_cpu(EXT4_SB(inode->i_sb)->s_es->s_journal_inum))) + journal_t *journal = EXT4_SB(inode->i_sb)->s_journal; + + if (journal && inode == journal->j_inode) return 0; + if (!ext4_inode_block_valid(inode, map->m_pblk, map->m_len)) { ext4_error_inode(inode, func, line, map->m_pblk, "lblock %lu mapped to illegal pblock %llu " -- 2.50.1 From 94824ac9a8aaf2fb3c54b4bdde842db80ffa555d Mon Sep 17 00:00:00 2001 From: Artem Sadovnikov Date: Fri, 4 Apr 2025 08:28:05 +0000 Subject: [PATCH 14/16] ext4: fix off-by-one error in do_split Syzkaller detected a use-after-free issue in ext4_insert_dentry that was caused by out-of-bounds access due to incorrect splitting in do_split. BUG: KASAN: use-after-free in ext4_insert_dentry+0x36a/0x6d0 fs/ext4/namei.c:2109 Write of size 251 at addr ffff888074572f14 by task syz-executor335/5847 CPU: 0 UID: 0 PID: 5847 Comm: syz-executor335 Not tainted 6.12.0-rc6-syzkaller-00318-ga9cda7c0ffed #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 10/30/2024 Call Trace: __dump_stack lib/dump_stack.c:94 [inline] dump_stack_lvl+0x241/0x360 lib/dump_stack.c:120 print_address_description mm/kasan/report.c:377 [inline] print_report+0x169/0x550 mm/kasan/report.c:488 kasan_report+0x143/0x180 mm/kasan/report.c:601 kasan_check_range+0x282/0x290 mm/kasan/generic.c:189 __asan_memcpy+0x40/0x70 mm/kasan/shadow.c:106 ext4_insert_dentry+0x36a/0x6d0 fs/ext4/namei.c:2109 add_dirent_to_buf+0x3d9/0x750 fs/ext4/namei.c:2154 make_indexed_dir+0xf98/0x1600 fs/ext4/namei.c:2351 ext4_add_entry+0x222a/0x25d0 fs/ext4/namei.c:2455 ext4_add_nondir+0x8d/0x290 fs/ext4/namei.c:2796 ext4_symlink+0x920/0xb50 fs/ext4/namei.c:3431 vfs_symlink+0x137/0x2e0 fs/namei.c:4615 do_symlinkat+0x222/0x3a0 fs/namei.c:4641 __do_sys_symlink fs/namei.c:4662 [inline] __se_sys_symlink fs/namei.c:4660 [inline] __x64_sys_symlink+0x7a/0x90 fs/namei.c:4660 do_syscall_x64 arch/x86/entry/common.c:52 [inline] do_syscall_64+0xf3/0x230 arch/x86/entry/common.c:83 entry_SYSCALL_64_after_hwframe+0x77/0x7f The following loop is located right above 'if' statement. for (i = count-1; i >= 0; i--) { /* is more than half of this entry in 2nd half of the block? */ if (size + map[i].size/2 > blocksize/2) break; size += map[i].size; move++; } 'i' in this case could go down to -1, in which case sum of active entries wouldn't exceed half the block size, but previous behaviour would also do split in half if sum would exceed at the very last block, which in case of having too many long name files in a single block could lead to out-of-bounds access and following use-after-free. Found by Linux Verification Center (linuxtesting.org) with Syzkaller. Cc: stable@vger.kernel.org Fixes: 5872331b3d91 ("ext4: fix potential negative array index in do_split()") Signed-off-by: Artem Sadovnikov Reviewed-by: Jan Kara Link: https://patch.msgid.link/20250404082804.2567-3-a.sadovnikov@ispras.ru Signed-off-by: Theodore Ts'o --- fs/ext4/namei.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index 43f96ef4aa01..dda1791e9e1a 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -1971,7 +1971,7 @@ static struct ext4_dir_entry_2 *do_split(handle_t *handle, struct inode *dir, * split it in half by count; each resulting block will have at least * half the space free. */ - if (i > 0) + if (i >= 0) split = count - move; else split = count/2; -- 2.50.1 From 8ffd015db85fea3e15a77027fda6c02ced4d2444 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Sun, 13 Apr 2025 11:54:49 -0700 Subject: [PATCH 15/16] Linux 6.15-rc2 --- Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Makefile b/Makefile index f42418556507..c91afd55099e 100644 --- a/Makefile +++ b/Makefile @@ -2,7 +2,7 @@ VERSION = 6 PATCHLEVEL = 15 SUBLEVEL = 0 -EXTRAVERSION = -rc1 +EXTRAVERSION = -rc2 NAME = Baby Opossum Posse # *DOCUMENTATION* -- 2.50.1 From 447fab30955cf7dba7dd563f42b67c02284860c8 Mon Sep 17 00:00:00 2001 From: =?utf8?q?Christian=20K=C3=B6nig?= Date: Fri, 28 Mar 2025 18:58:17 +0100 Subject: [PATCH 16/16] drm/amdgpu: use a dummy owner for sysfs triggered cleaner shaders v4 MIME-Version: 1.0 Content-Type: text/plain; charset=utf8 Content-Transfer-Encoding: 8bit Otherwise triggering sysfs multiple times without other submissions in between only runs the shader once. v2: add some comment v3: re-add missing cast v4: squash in semicolon fix Signed-off-by: Christian König Reviewed-by: Srinivasan Shanmugam Signed-off-by: Alex Deucher (cherry picked from commit 8b2ae7d492675e8af8902f103364bef59382b935) --- drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c | 14 +++++++++++--- 1 file changed, 11 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c index 72af5e5a894a..cf2df7790077 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c @@ -1438,9 +1438,11 @@ static int amdgpu_gfx_run_cleaner_shader_job(struct amdgpu_ring *ring) struct amdgpu_device *adev = ring->adev; struct drm_gpu_scheduler *sched = &ring->sched; struct drm_sched_entity entity; + static atomic_t counter; struct dma_fence *f; struct amdgpu_job *job; struct amdgpu_ib *ib; + void *owner; int i, r; /* Initialize the scheduler entity */ @@ -1451,9 +1453,15 @@ static int amdgpu_gfx_run_cleaner_shader_job(struct amdgpu_ring *ring) goto err; } - r = amdgpu_job_alloc_with_ib(ring->adev, &entity, NULL, - 64, 0, - &job); + /* + * Use some unique dummy value as the owner to make sure we execute + * the cleaner shader on each submission. The value just need to change + * for each submission and is otherwise meaningless. + */ + owner = (void *)(unsigned long)atomic_inc_return(&counter); + + r = amdgpu_job_alloc_with_ib(ring->adev, &entity, owner, + 64, 0, &job); if (r) goto err; -- 2.50.1