December 2024 - Linux-stable-mirror

[PATCH 5.10 1/1] Bluetooth: L2CAP: Fix uaf in l2cap_connect

by d.privalov

From: Luiz Augusto von Dentz <luiz.von.dentz(a)intel.com> commit 333b4fd11e89b29c84c269123f871883a30be586 upstream. [Syzbot reported] BUG: KASAN: slab-use-after-free in l2cap_connect.constprop.0+0x10d8/0x1270 net/bluetooth/l2cap_core.c:3949 Read of size 8 at addr ffff8880241e9800 by task kworker/u9:0/54 CPU: 0 UID: 0 PID: 54 Comm: kworker/u9:0 Not tainted 6.11.0-rc6-syzkaller-00268-g788220eee30d #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 08/06/2024 Workqueue: hci2 hci_rx_work Call Trace: <TASK> __dump_stack lib/dump_stack.c:93 [inline] dump_stack_lvl+0x116/0x1f0 lib/dump_stack.c:119 print_address_description mm/kasan/report.c:377 [inline] print_report+0xc3/0x620 mm/kasan/report.c:488 kasan_report+0xd9/0x110 mm/kasan/report.c:601 l2cap_connect.constprop.0+0x10d8/0x1270 net/bluetooth/l2cap_core.c:3949 l2cap_connect_req net/bluetooth/l2cap_core.c:4080 [inline] l2cap_bredr_sig_cmd net/bluetooth/l2cap_core.c:4772 [inline] l2cap_sig_channel net/bluetooth/l2cap_core.c:5543 [inline] l2cap_recv_frame+0xf0b/0x8eb0 net/bluetooth/l2cap_core.c:6825 l2cap_recv_acldata+0x9b4/0xb70 net/bluetooth/l2cap_core.c:7514 hci_acldata_packet net/bluetooth/hci_core.c:3791 [inline] hci_rx_work+0xaab/0x1610 net/bluetooth/hci_core.c:4028 process_one_work+0x9c5/0x1b40 kernel/workqueue.c:3231 process_scheduled_works kernel/workqueue.c:3312 [inline] worker_thread+0x6c8/0xed0 kernel/workqueue.c:3389 kthread+0x2c1/0x3a0 kernel/kthread.c:389 ret_from_fork+0x45/0x80 arch/x86/kernel/process.c:147 ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:244 ... Freed by task 5245: kasan_save_stack+0x33/0x60 mm/kasan/common.c:47 kasan_save_track+0x14/0x30 mm/kasan/common.c:68 kasan_save_free_info+0x3b/0x60 mm/kasan/generic.c:579 poison_slab_object+0xf7/0x160 mm/kasan/common.c:240 __kasan_slab_free+0x32/0x50 mm/kasan/common.c:256 kasan_slab_free include/linux/kasan.h:184 [inline] slab_free_hook mm/slub.c:2256 [inline] slab_free mm/slub.c:4477 [inline] kfree+0x12a/0x3b0 mm/slub.c:4598 l2cap_conn_free net/bluetooth/l2cap_core.c:1810 [inline] kref_put include/linux/kref.h:65 [inline] l2cap_conn_put net/bluetooth/l2cap_core.c:1822 [inline] l2cap_conn_del+0x59d/0x730 net/bluetooth/l2cap_core.c:1802 l2cap_connect_cfm+0x9e6/0xf80 net/bluetooth/l2cap_core.c:7241 hci_connect_cfm include/net/bluetooth/hci_core.h:1960 [inline] hci_conn_failed+0x1c3/0x370 net/bluetooth/hci_conn.c:1265 hci_abort_conn_sync+0x75a/0xb50 net/bluetooth/hci_sync.c:5583 abort_conn_sync+0x197/0x360 net/bluetooth/hci_conn.c:2917 hci_cmd_sync_work+0x1a4/0x410 net/bluetooth/hci_sync.c:328 process_one_work+0x9c5/0x1b40 kernel/workqueue.c:3231 process_scheduled_works kernel/workqueue.c:3312 [inline] worker_thread+0x6c8/0xed0 kernel/workqueue.c:3389 kthread+0x2c1/0x3a0 kernel/kthread.c:389 ret_from_fork+0x45/0x80 arch/x86/kernel/process.c:147 ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:244 Reported-by: syzbot+c12e2f941af1feb5632c(a)syzkaller.appspotmail.com Tested-by: syzbot+c12e2f941af1feb5632c(a)syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=c12e2f941af1feb5632c Fixes: 7b064edae38d ("Bluetooth: Fix authentication if acl data comes before remote feature evt") Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz(a)intel.com> Signed-off-by: Dmitriy Privalov <d.privalov(a)omp.ru> --- net/bluetooth/hci_core.c | 2 ++ net/bluetooth/hci_event.c | 2 +- net/bluetooth/l2cap_core.c | 9 --------- 3 files changed, 3 insertions(+), 10 deletions(-) diff --git a/net/bluetooth/hci_core.c b/net/bluetooth/hci_core.c index 9787a4c55113..c4c86407b920 100644 --- a/net/bluetooth/hci_core.c +++ b/net/bluetooth/hci_core.c @@ -4769,6 +4769,8 @@ static void hci_acldata_packet(struct hci_dev *hdev, struct sk_buff *skb) hci_dev_lock(hdev); conn = hci_conn_hash_lookup_handle(hdev, handle); + if (conn && hci_dev_test_flag(hdev, HCI_MGMT)) + mgmt_device_connected(hdev, conn, 0, NULL, 0); hci_dev_unlock(hdev); if (conn) { diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c index 58c029958759..634b12b19b32 100644 --- a/net/bluetooth/hci_event.c +++ b/net/bluetooth/hci_event.c @@ -3245,7 +3245,7 @@ static void hci_remote_features_evt(struct hci_dev *hdev, goto unlock; } - if (!ev->status && !test_bit(HCI_CONN_MGMT_CONNECTED, &conn->flags)) { + if (!ev->status) { struct hci_cp_remote_name_req cp; memset(&cp, 0, sizeof(cp)); bacpy(&cp.bdaddr, &conn->dst); diff --git a/net/bluetooth/l2cap_core.c b/net/bluetooth/l2cap_core.c index 23fc03f7bf31..cad0e535ff81 100644 --- a/net/bluetooth/l2cap_core.c +++ b/net/bluetooth/l2cap_core.c @@ -4272,18 +4272,9 @@ static struct l2cap_chan *l2cap_connect(struct l2cap_conn *conn, static int l2cap_connect_req(struct l2cap_conn *conn, struct l2cap_cmd_hdr *cmd, u16 cmd_len, u8 *data) { - struct hci_dev *hdev = conn->hcon->hdev; - struct hci_conn *hcon = conn->hcon; - if (cmd_len < sizeof(struct l2cap_conn_req)) return -EPROTO; - hci_dev_lock(hdev); - if (hci_dev_test_flag(hdev, HCI_MGMT) && - !test_and_set_bit(HCI_CONN_MGMT_CONNECTED, &hcon->flags)) - mgmt_device_connected(hdev, hcon, 0, NULL, 0); - hci_dev_unlock(hdev); - l2cap_connect(conn, cmd, data, L2CAP_CONN_RSP, 0); return 0; } -- 2.34.1

9 months

2
3
0 0

[PATCH] drm/amdgpu: fix backport of commit 73dae652dcac

by Alex Deucher

Commit 73dae652dcac ("drm/amdgpu: rework resume handling for display (v2)") missed a small code change when it was backported resulting in an automatic backlight control breakage. Fix the backport. Note that this patch is not in Linus' tree as it is not required there; the bug was introduced in the backport. Fixes: 99a02eab8251 ("drm/amdgpu: rework resume handling for display (v2)") Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/3853 Signed-off-by: Alex Deucher <alexander.deucher(a)amd.com> Cc: stable(a)vger.kernel.org # 6.11.x --- drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 51904906545e..45e28726e148 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -3721,8 +3721,12 @@ static int amdgpu_device_ip_resume_phase3(struct amdgpu_device *adev) continue; if (adev->ip_blocks[i].version->type == AMD_IP_BLOCK_TYPE_DCE) { r = adev->ip_blocks[i].version->funcs->resume(adev); - if (r) + if (r) { + DRM_ERROR("resume of IP block <%s> failed %d\n", + adev->ip_blocks[i].version->funcs->name, r); return r; + } + adev->ip_blocks[i].status.hw = true; } } -- 2.47.1

9 months

5
6
0 0

[PATCH] Revert "MIPS: csrc-r4k: Select HAVE_UNSTABLE_SCHED_CLOCK if SMP && 64BIT"

by Xi Ruoyao

This reverts commit 426fa8e4fe7bb914b5977cbce453a9926bf5b2e6. The commit has caused two issues on Loongson 3A4000: 1. The timestamp in dmesg become erratic, like: [3.736957] amdgpu 0000:04:00.0: ... ... [3.748895] [drm] Initialized amdgpu ... ... [18446744073.381141] amdgpu 0000:04:00:0: ... ... [1.613326] igb 0000:03:00.0 enp3s0: ... ... 2. More seriously, some workloads (for example, the test stdlib/test-cxa_atexit2 in the Glibc test suite) triggers an RCU stall and hang the system with a high probably (4 hangs out of 5 tests). Revert this commit to use jiffie on Loongson MIPS systems and fix these issues for now. The root cause may need more investigation. Cc: stable(a)vger.kernel.org # 6.11+ Cc: Jiaxun Yang <jiaxun.yang(a)flygoat.com> Cc: Icenowy Zheng <icenowy(a)aosc.io> Signed-off-by: Xi Ruoyao <xry111(a)xry111.site> --- arch/mips/Kconfig | 1 - 1 file changed, 1 deletion(-) diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig index 467b10f4361a..5078ebf071ec 100644 --- a/arch/mips/Kconfig +++ b/arch/mips/Kconfig @@ -1084,7 +1084,6 @@ config CSRC_IOASIC config CSRC_R4K select CLOCKSOURCE_WATCHDOG if CPU_FREQ - select HAVE_UNSTABLE_SCHED_CLOCK if SMP && 64BIT bool config CSRC_SB1250 -- 2.47.1

9 months

3
2
0 0

[PATCH RESEND] [SCSI] raid class: Fix error handling in raid_component_add

by Ma Ke

The reference count of the device incremented in device_initialize() is not decremented when device_add() fails. Add a put_device() call before returning from the function to decrement reference count for cleanup. Or it could cause memory leak. As comment of device_add() says, if device_add() succeeds, you should call device_del() when you want to get rid of it. If device_add() has not succeeded, use only put_device() to drop the reference count. Found by code review. Cc: stable(a)vger.kernel.org Fixes: ed542bed126c ("[SCSI] raid class: handle component-add errors") Signed-off-by: Ma Ke <make_ruc2021(a)163.com> --- drivers/scsi/raid_class.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/scsi/raid_class.c b/drivers/scsi/raid_class.c index 898a0bdf8df6..2cb2949a78c6 100644 --- a/drivers/scsi/raid_class.c +++ b/drivers/scsi/raid_class.c @@ -251,6 +251,7 @@ int raid_component_add(struct raid_template *r,struct device *raid_dev, list_del(&rc->node); rd->component_count--; put_device(component_dev); + put_device(&rc->dev); kfree(rc); return err; } -- 2.25.1

9 months, 1 week

2
1
0 0

[PATCH] ufs: qcom: fix crypto key eviction

by Eric Biggers

From: Eric Biggers <ebiggers(a)google.com> Commit 56541c7c4468 ("scsi: ufs: ufs-qcom: Switch to the new ICE API") introduced an incorrect check of the algorithm ID into the key eviction path, and thus qcom_ice_evict_key() is no longer ever called. Fix it. Fixes: 56541c7c4468 ("scsi: ufs: ufs-qcom: Switch to the new ICE API") Cc: stable(a)vger.kernel.org Cc: Abel Vesa <abel.vesa(a)linaro.org> Signed-off-by: Eric Biggers <ebiggers(a)google.com> --- drivers/ufs/host/ufs-qcom.c | 18 ++++++++---------- 1 file changed, 8 insertions(+), 10 deletions(-) diff --git a/drivers/ufs/host/ufs-qcom.c b/drivers/ufs/host/ufs-qcom.c index 68040b2ab5f82..e33ae71245dd0 100644 --- a/drivers/ufs/host/ufs-qcom.c +++ b/drivers/ufs/host/ufs-qcom.c @@ -153,27 +153,25 @@ static int ufs_qcom_ice_program_key(struct ufs_hba *hba, const union ufs_crypto_cfg_entry *cfg, int slot) { struct ufs_qcom_host *host = ufshcd_get_variant(hba); union ufs_crypto_cap_entry cap; - bool config_enable = - cfg->config_enable & UFS_CRYPTO_CONFIGURATION_ENABLE; + + if (!(cfg->config_enable & UFS_CRYPTO_CONFIGURATION_ENABLE)) + return qcom_ice_evict_key(host->ice, slot); /* Only AES-256-XTS has been tested so far. */ cap = hba->crypto_cap_array[cfg->crypto_cap_idx]; if (cap.algorithm_id != UFS_CRYPTO_ALG_AES_XTS || cap.key_size != UFS_CRYPTO_KEY_SIZE_256) return -EOPNOTSUPP; - if (config_enable) - return qcom_ice_program_key(host->ice, - QCOM_ICE_CRYPTO_ALG_AES_XTS, - QCOM_ICE_CRYPTO_KEY_SIZE_256, - cfg->crypto_key, - cfg->data_unit_size, slot); - else - return qcom_ice_evict_key(host->ice, slot); + return qcom_ice_program_key(host->ice, + QCOM_ICE_CRYPTO_ALG_AES_XTS, + QCOM_ICE_CRYPTO_KEY_SIZE_256, + cfg->crypto_key, + cfg->data_unit_size, slot); } #else #define ufs_qcom_ice_program_key NULL base-commit: 7cb1b466315004af98f6ba6c2546bb713ca3c237 -- 2.47.1

9 months, 1 week

2
1
0 0

[PATCH v2] arm64: tegra: Fix Tegra234 PCIe interrupt-map

by Brad Griffis

For interrupt-map entries, the DTS specification requires that #address-cells is defined for both the child node and the interrupt parent. For the PCIe interrupt-map entries, the parent node ("gic") has not specified #address-cells. The existing layout of the PCIe interrupt-map entries indicates that it assumes that #address-cells is zero for this node. Explicitly set #address-cells to zero for "gic" so that it complies with the device tree specification. NVIDIA EDK2 works around this issue by assuming #address-cells is zero in this scenario, but that workaround is being removed and so this update is needed or else NVIDIA EDK2 cannot successfully parse the device tree and the board cannot boot. Fixes: ec142c44b026 ("arm64: tegra: Add P2U and PCIe controller nodes to Tegra234 DT") Signed-off-by: Brad Griffis <bgriffis(a)nvidia.com> Cc: stable(a)vger.kernel.org --- v2 * Add "Fixes" and CC stable * Give further context about why the patch is needed arch/arm64/boot/dts/nvidia/tegra234.dtsi | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/arm64/boot/dts/nvidia/tegra234.dtsi b/arch/arm64/boot/dts/nvidia/tegra234.dtsi index 984c85eab41a..e1c07c99e9bd 100644 --- a/arch/arm64/boot/dts/nvidia/tegra234.dtsi +++ b/arch/arm64/boot/dts/nvidia/tegra234.dtsi @@ -4010,6 +4010,7 @@ ccplex@e000000 { gic: interrupt-controller@f400000 { compatible = "arm,gic-v3"; + #address-cells = <0>; reg = <0x0 0x0f400000 0x0 0x010000>, /* GICD */ <0x0 0x0f440000 0x0 0x200000>; /* GICR */ interrupt-parent = <&gic>; -- 2.34.1

9 months, 1 week

2
1
0 0

[PATCH v2 0/2] rust: lockdep: Fix soundness issue affecting LockClassKeys

by Mitchell Levy

This series is aimed at fixing a soundness issue with how dynamically allocated LockClassKeys are handled. Currently, LockClassKeys can be used without being Pin'd, which can break lockdep since it relies on address stability. Similarly, these keys are not automatically (de)registered with lockdep. At the suggestion of Alice Ryhl, this series includes a patch for -stable kernels that disables dynamically allocated keys. This prevents backported patches from using the unsound implementation. Currently, this series requires that all dynamically allocated LockClassKeys have a lifetime of 'static (i.e., they must be leaked after allocation). This is because Lock does not currently keep a reference to the LockClassKey, instead passing it to C via FFI. This causes a problem because the rust compiler would allow creating a 'static Lock with a 'a LockClassKey (with 'a < 'static) while C would expect the LockClassKey to live as long as the lock. This problem represents an avenue for future work. --- Changes from RFC: - Split into two commits so that dynamically allocated LockClassKeys are removed from stable kernels. (Thanks Alice Ryhl) - Extract calls to C lockdep functions into helpers so things build properly when LOCKDEP=n. (Thanks Benno Lossin) - Remove extraneous `get_ref()` calls. (Thanks Benno Lossin) - Provide better documentation for `new_dynamic()`. (Thanks Benno Lossin) - Ran rustfmt to fix formatting and some extraneous changes. (Thanks Alice Ryhl and Benno Lossin) - Link to RFC: https://lore.kernel.org/r/20240905-rust-lockdep-v1-1-d2c9c21aa8b2@gmail.com --- Changes in v2: - Dropped formatting change that's already fixed upstream (Thanks Dirk Behme). - Moved safety comment to the right point in the patch series (Thanks Dirk Behme and Boqun Feng). - Added an example of dynamic LockClassKey usage (Thanks Boqun Feng). - Link to v1: https://lore.kernel.org/r/20241004-rust-lockdep-v1-0-e9a5c45721fc@gmail.com --- Mitchell Levy (2): rust: lockdep: Remove support for dynamically allocated LockClassKeys rust: lockdep: Use Pin for all LockClassKey usages rust/helpers/helpers.c | 1 + rust/helpers/sync.c | 13 +++++++++ rust/kernel/sync.rs | 63 ++++++++++++++++++++++++++++++++++------- rust/kernel/sync/condvar.rs | 5 ++-- rust/kernel/sync/lock.rs | 9 ++---- rust/kernel/sync/lock/global.rs | 5 ++-- rust/kernel/sync/poll.rs | 2 +- rust/kernel/workqueue.rs | 3 +- 8 files changed, 78 insertions(+), 23 deletions(-) --- base-commit: 0c5928deada15a8d075516e6e0d9ee19011bb000 change-id: 20240905-rust-lockdep-d3e30521c8ba Best regards, -- Mitchell Levy <levymitchell0(a)gmail.com>

9 months, 1 week

3
4
0 0

[PATCH v2] samples/landlock: Fix possible NULL dereference in parse_path()

by Gax-c

From: Zichen Xie <zichenxie0106(a)gmail.com> malloc() may return NULL, leading to NULL dereference. Add a NULL check. Fixes: ba84b0bf5a16 ("samples/landlock: Add a sandbox manager example") Signed-off-by: Zichen Xie <zichenxie0106(a)gmail.com> Cc: stable(a)vger.kernel.org --- v2: Modify logic & Add Fixes tag. --- samples/landlock/sandboxer.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/samples/landlock/sandboxer.c b/samples/landlock/sandboxer.c index 57565dfd74a2..ef2a34173d84 100644 --- a/samples/landlock/sandboxer.c +++ b/samples/landlock/sandboxer.c @@ -91,6 +91,9 @@ static int parse_path(char *env_path, const char ***const path_list) } } *path_list = malloc(num_paths * sizeof(**path_list)); + if (*path_list == NULL) + return -1; + for (i = 0; i < num_paths; i++) (*path_list)[i] = strsep(&env_path, ENV_DELIMITER); @@ -127,6 +130,11 @@ static int populate_ruleset_fs(const char *const env_var, const int ruleset_fd, env_path_name = strdup(env_path_name); unsetenv(env_var); num_paths = parse_path(env_path_name, &path_list); + if (num_paths == -1) { + fprintf(stderr, "Failed to allocate memory\n"); + ret = 1; + goto out_free_name; + } if (num_paths == 1 && path_list[0][0] == '\0') { /* * Allows to not use all possible restrictions (e.g. use -- 2.34.1

9 months, 1 week

2
1
0 0

[PATCH v2 5/9] btrfs: do proper folio cleanup when run_delalloc_nocow() failed

by Qu Wenruo

[BUG] With CONFIG_DEBUG_VM set, test case generic/476 has some chance to crash with the following VM_BUG_ON_FOLIO(): BTRFS error (device dm-3): cow_file_range failed, start 1146880 end 1253375 len 106496 ret -28 BTRFS error (device dm-3): run_delalloc_nocow failed, start 1146880 end 1253375 len 106496 ret -28 page: refcount:4 mapcount:0 mapping:00000000592787cc index:0x12 pfn:0x10664 aops:btrfs_aops [btrfs] ino:101 dentry name(?):"f1774" flags: 0x2fffff80004028(uptodate|lru|private|node=0|zone=2|lastcpupid=0xfffff) page dumped because: VM_BUG_ON_FOLIO(!folio_test_locked(folio)) ------------[ cut here ]------------ kernel BUG at mm/page-writeback.c:2992! Internal error: Oops - BUG: 00000000f2000800 [#1] SMP CPU: 2 UID: 0 PID: 3943513 Comm: kworker/u24:15 Tainted: G OE 6.12.0-rc7-custom+ #87 Tainted: [O]=OOT_MODULE, [E]=UNSIGNED_MODULE Hardware name: QEMU KVM Virtual Machine, BIOS unknown 2/2/2022 Workqueue: events_unbound btrfs_async_reclaim_data_space [btrfs] pc : folio_clear_dirty_for_io+0x128/0x258 lr : folio_clear_dirty_for_io+0x128/0x258 Call trace: folio_clear_dirty_for_io+0x128/0x258 btrfs_folio_clamp_clear_dirty+0x80/0xd0 [btrfs] __process_folios_contig+0x154/0x268 [btrfs] extent_clear_unlock_delalloc+0x5c/0x80 [btrfs] run_delalloc_nocow+0x5f8/0x760 [btrfs] btrfs_run_delalloc_range+0xa8/0x220 [btrfs] writepage_delalloc+0x230/0x4c8 [btrfs] extent_writepage+0xb8/0x358 [btrfs] extent_write_cache_pages+0x21c/0x4e8 [btrfs] btrfs_writepages+0x94/0x150 [btrfs] do_writepages+0x74/0x190 filemap_fdatawrite_wbc+0x88/0xc8 start_delalloc_inodes+0x178/0x3a8 [btrfs] btrfs_start_delalloc_roots+0x174/0x280 [btrfs] shrink_delalloc+0x114/0x280 [btrfs] flush_space+0x250/0x2f8 [btrfs] btrfs_async_reclaim_data_space+0x180/0x228 [btrfs] process_one_work+0x164/0x408 worker_thread+0x25c/0x388 kthread+0x100/0x118 ret_from_fork+0x10/0x20 Code: 910a8021 a90363f7 a9046bf9 94012379 (d4210000) ---[ end trace 0000000000000000 ]--- [CAUSE] The first two lines of extra debug messages show the problem is caused by the error handling of run_delalloc_nocow(). E.g. we have the following dirtied range (4K blocksize 4K page size): 0 16K 32K |//////////////////////////////////////| | Pre-allocated | And the range [0, 16K) has a preallocated extent. - Enter run_delalloc_nocow() for range [0, 16K) Which found range [0, 16K) is preallocated, can do the proper NOCOW write. - Enter fallback_to_fow() for range [16K, 32K) Since the range [16K, 32K) is not backed by preallocated extent, we have to go COW. - cow_file_range() failed for range [16K, 32K) So cow_file_range() will do the clean up by clearing folio dirty, unlock the folios. Now the folios in range [16K, 32K) is unlocked. - Enter extent_clear_unlock_delalloc() from run_delalloc_nocow() Which is called with PAGE_START_WRITEBACK to start page writeback. But folios can only be marked writeback when it's properly locked, thus this triggered the VM_BUG_ON_FOLIO(). Furthermore there is another hidden but common bug that run_delalloc_nocow() is not clearing the folio dirty flags in its error handling path. This is the common bug shared between run_delalloc_nocow() and cow_file_range(). [FIX] - Clear folio dirty for range [@start, @cur_offset) Introduce a helper, cleanup_dirty_folios(), which will find and lock the folio in the range, clear the dirty flag and start/end the writeback, with the extra handling for the @locked_folio. - Introduce a helper to record the last failed COW range end This is to trace which range we should skip, to avoid double unlocking. - Skip the failed COW range for the error handling Cc: stable(a)vger.kernel.org Signed-off-by: Qu Wenruo <wqu(a)suse.com> --- fs/btrfs/inode.c | 93 ++++++++++++++++++++++++++++++++++++++++++++---- 1 file changed, 86 insertions(+), 7 deletions(-) diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 19c88b7d0363..bae8aceb3eae 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -1961,6 +1961,48 @@ static int can_nocow_file_extent(struct btrfs_path *path, return ret < 0 ? ret : can_nocow; } +static void cleanup_dirty_folios(struct btrfs_inode *inode, + struct folio *locked_folio, + u64 start, u64 end, int error) +{ + struct btrfs_fs_info *fs_info = inode->root->fs_info; + struct address_space *mapping = inode->vfs_inode.i_mapping; + pgoff_t start_index = start >> PAGE_SHIFT; + pgoff_t end_index = end >> PAGE_SHIFT; + u32 len; + + ASSERT(end + 1 - start < U32_MAX); + ASSERT(IS_ALIGNED(start, fs_info->sectorsize) && + IS_ALIGNED(end + 1, fs_info->sectorsize)); + len = end + 1 - start; + + /* + * Handle the locked folio first. + * btrfs_folio_clamp_*() helpers can handle range out of the folio case. + */ + btrfs_folio_clamp_clear_dirty(fs_info, locked_folio, start, len); + btrfs_folio_clamp_set_writeback(fs_info, locked_folio, start, len); + btrfs_folio_clamp_clear_writeback(fs_info, locked_folio, start, len); + + for (pgoff_t index = start_index; index <= end_index; index++) { + struct folio *folio; + + /* Already handled at the beginning. */ + if (index == locked_folio->index) + continue; + folio = __filemap_get_folio(mapping, index, FGP_LOCK, GFP_NOFS); + /* Cache already dropped, no need to do any cleanup. */ + if (IS_ERR(folio)) + continue; + btrfs_folio_clamp_clear_dirty(fs_info, folio, start, len); + btrfs_folio_clamp_set_writeback(fs_info, folio, start, len); + btrfs_folio_clamp_clear_writeback(fs_info, folio, start, len); + folio_unlock(folio); + folio_put(folio); + } + mapping_set_error(mapping, error); +} + /* * when nowcow writeback call back. This checks for snapshots or COW copies * of the extents that exist in the file, and COWs the file as required. @@ -1976,6 +2018,11 @@ static noinline int run_delalloc_nocow(struct btrfs_inode *inode, struct btrfs_root *root = inode->root; struct btrfs_path *path; u64 cow_start = (u64)-1; + /* + * If not 0, represents the inclusive end of the last fallback_to_cow() + * range. Only for error handling. + */ + u64 cow_end = 0; u64 cur_offset = start; int ret; bool check_prev = true; @@ -2136,6 +2183,7 @@ static noinline int run_delalloc_nocow(struct btrfs_inode *inode, found_key.offset - 1); cow_start = (u64)-1; if (ret) { + cow_end = found_key.offset - 1; btrfs_dec_nocow_writers(nocow_bg); goto error; } @@ -2209,11 +2257,12 @@ static noinline int run_delalloc_nocow(struct btrfs_inode *inode, cow_start = cur_offset; if (cow_start != (u64)-1) { - cur_offset = end; ret = fallback_to_cow(inode, locked_folio, cow_start, end); cow_start = (u64)-1; - if (ret) + if (ret) { + cow_end = end; goto error; + } } btrfs_free_path(path); @@ -2221,12 +2270,42 @@ static noinline int run_delalloc_nocow(struct btrfs_inode *inode, error: /* - * If an error happened while a COW region is outstanding, cur_offset - * needs to be reset to cow_start to ensure the COW region is unlocked - * as well. + * There are several error cases: + * + * 1) Failed without falling back to COW + * start cur_start end + * |/////////////| | + * + * For range [start, cur_start) the folios are already unlocked (except + * @locked_folio), EXTENT_DELALLOC already removed. + * Only need to clear the dirty flag as they will never be submitted. + * Ordered extent and extent maps are handled by + * btrfs_mark_ordered_io_finished() inside run_delalloc_range(). + * + * 2) Failed with error from fallback_to_cow() + * start cur_start cow_end end + * |/////////////|-----------| | + * + * For range [start, cur_start) it's the same as case 1). + * But for range [cur_start, cow_end), the folios have dirty flag + * cleared and unlocked, EXTENT_DEALLLOC cleared. + * There may or may not be any ordered extents/extent maps allocated. + * + * We should not call extent_clear_unlock_delalloc() on range [cur_start, + * cow_end), as the folios are already unlocked. + * + * So clear the folio dirty flags for [start, cur_offset) first. */ - if (cow_start != (u64)-1) - cur_offset = cow_start; + if (cur_offset > start) + cleanup_dirty_folios(inode, locked_folio, start, cur_offset - 1, ret); + + /* + * If an error happened while a COW region is outstanding, cur_offset + * needs to be reset to @cow_end + 1 to skip the COW range, as + * cow_file_range() will do the proper cleanup at error. + */ + if (cow_end) + cur_offset = cow_end + 1; /* * We need to lock the extent here because we're clearing DELALLOC and -- 2.47.1

9 months, 1 week

3
2
0 0

[PATCH v2 4/9] btrfs: do proper folio cleanup when cow_file_range() failed

by Qu Wenruo

[BUG] When testing with COW fixup marked as BUG_ON() (this is involved with the new pin_user_pages*() change, which should not result new out-of-band dirty pages), I hit a crash triggered by the BUG_ON() from hitting COW fixup path. This BUG_ON() happens just after a failed btrfs_run_delalloc_range(): BTRFS error (device dm-2): failed to run delalloc range, root 348 ino 405 folio 65536 submit_bitmap 6-15 start 90112 len 106496: -28 ------------[ cut here ]------------ kernel BUG at fs/btrfs/extent_io.c:1444! Internal error: Oops - BUG: 00000000f2000800 [#1] SMP CPU: 0 UID: 0 PID: 434621 Comm: kworker/u24:8 Tainted: G OE 6.12.0-rc7-custom+ #86 Hardware name: QEMU KVM Virtual Machine, BIOS unknown 2/2/2022 Workqueue: events_unbound btrfs_async_reclaim_data_space [btrfs] pc : extent_writepage_io+0x2d4/0x308 [btrfs] lr : extent_writepage_io+0x2d4/0x308 [btrfs] Call trace: extent_writepage_io+0x2d4/0x308 [btrfs] extent_writepage+0x218/0x330 [btrfs] extent_write_cache_pages+0x1d4/0x4b0 [btrfs] btrfs_writepages+0x94/0x150 [btrfs] do_writepages+0x74/0x190 filemap_fdatawrite_wbc+0x88/0xc8 start_delalloc_inodes+0x180/0x3b0 [btrfs] btrfs_start_delalloc_roots+0x174/0x280 [btrfs] shrink_delalloc+0x114/0x280 [btrfs] flush_space+0x250/0x2f8 [btrfs] btrfs_async_reclaim_data_space+0x180/0x228 [btrfs] process_one_work+0x164/0x408 worker_thread+0x25c/0x388 kthread+0x100/0x118 ret_from_fork+0x10/0x20 Code: aa1403e1 9402f3ef aa1403e0 9402f36f (d4210000) ---[ end trace 0000000000000000 ]--- [CAUSE] That failure is mostly from cow_file_range(), where we can hit -ENOSPC. Although the -ENOSPC is already a bug related to our space reservation code, let's just focus on the error handling. For example, we have the following dirty range [0, 64K) of an inode, with 4K sector size and 4K page size: 0 16K 32K 48K 64K |///////////////////////////////////////| |#######################################| Where |///| means page are still dirty, and |###| means the extent io tree has EXTENT_DELALLOC flag. - Enter extent_writepage() for page 0 - Enter btrfs_run_delalloc_range() for range [0, 64K) - Enter cow_file_range() for range [0, 64K) - Function btrfs_reserve_extent() only reserved one 16K extent So we created extent map and ordered extent for range [0, 16K) 0 16K 32K 48K 64K |////////|//////////////////////////////| |<- OE ->|##############################| And range [0, 16K) has its delalloc flag cleared. But since we haven't yet submit any bio, involved 4 pages are still dirty. - Function btrfs_reserve_extent() return with -ENOSPC Now we have to run error cleanup, which will clear all EXTENT_DELALLOC* flags and clear the dirty flags for the remaining ranges: 0 16K 32K 48K 64K |////////| | | | | Note that range [0, 16K) still has their pages dirty. - Some time later, writeback are triggered again for the range [0, 16K) since the page range still have dirty flags. - btrfs_run_delalloc_range() will do nothing because there is no EXTENT_DELALLOC flag. - extent_writepage_io() find page 0 has no ordered flag Which falls into the COW fixup path, triggering the BUG_ON(). Unfortunately this error handling bug dates back to the introduction of btrfs. Thankfully with the abuse of cow fixup, at least it won't crash the kernel. [FIX] Instead of immediately unlock the extent and folios, we keep the extent and folios locked until either erroring out or the whole delalloc range finished. When the whole delalloc range finished without error, we just unlock the whole range with PAGE_SET_ORDERED (and PAGE_UNLOCK for !keep_locked cases), with EXTENT_DELALLOC and EXTENT_LOCKED cleared. And those involved folios will be properly submitted, with their dirty flags cleared during submission. For the error path, it will be a little more complex: - The range with ordered extent allocated (range (1)) We only clear the EXTENT_DELALLOC and EXTENT_LOCKED, as the remaining flags are cleaned up by btrfs_mark_ordered_io_finished()->btrfs_finish_one_ordered(). For folios we finish the IO (clear dirty, start writeback and immediately finish the writeback) and unlock the folios. - The range with reserved extent but no ordered extent (range(2)) - The range we never touched (range(3)) For both range (2) and range(3) the behavior is not changed. Now even if cow_file_range() failed halfway with some successfully reserved extents/ordered extents, we will keep all folios clean, so there will be no future writeback triggered on them. Cc: stable(a)vger.kernel.org Signed-off-by: Qu Wenruo <wqu(a)suse.com> --- fs/btrfs/inode.c | 65 ++++++++++++++++++++++++------------------------ 1 file changed, 32 insertions(+), 33 deletions(-) diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 5ba8d044757b..19c88b7d0363 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -1364,6 +1364,17 @@ static noinline int cow_file_range(struct btrfs_inode *inode, alloc_hint = btrfs_get_extent_allocation_hint(inode, start, num_bytes); + /* + * We're not doing compressed IO, don't unlock the first page + * (which the caller expects to stay locked), don't clear any + * dirty bits and don't set any writeback bits + * + * Do set the Ordered (Private2) bit so we know this page was + * properly setup for writepage. + */ + page_ops = (keep_locked ? 0 : PAGE_UNLOCK); + page_ops |= PAGE_SET_ORDERED; + /* * Relocation relies on the relocated extents to have exactly the same * size as the original extents. Normally writeback for relocation data @@ -1423,6 +1434,10 @@ static noinline int cow_file_range(struct btrfs_inode *inode, file_extent.offset = 0; file_extent.compression = BTRFS_COMPRESS_NONE; + /* + * Locked range will be released either during error clean up or + * after the whole range is finished. + */ lock_extent(&inode->io_tree, start, start + cur_alloc_size - 1, &cached); @@ -1468,21 +1483,6 @@ static noinline int cow_file_range(struct btrfs_inode *inode, btrfs_dec_block_group_reservations(fs_info, ins.objectid); - /* - * We're not doing compressed IO, don't unlock the first page - * (which the caller expects to stay locked), don't clear any - * dirty bits and don't set any writeback bits - * - * Do set the Ordered flag so we know this page was - * properly setup for writepage. - */ - page_ops = (keep_locked ? 0 : PAGE_UNLOCK); - page_ops |= PAGE_SET_ORDERED; - - extent_clear_unlock_delalloc(inode, start, start + cur_alloc_size - 1, - locked_folio, &cached, - EXTENT_LOCKED | EXTENT_DELALLOC, - page_ops); if (num_bytes < cur_alloc_size) num_bytes = 0; else @@ -1499,6 +1499,9 @@ static noinline int cow_file_range(struct btrfs_inode *inode, if (ret) goto out_unlock; } + extent_clear_unlock_delalloc(inode, orig_start, end, locked_folio, &cached, + EXTENT_LOCKED | EXTENT_DELALLOC, + page_ops); done: if (done_offset) *done_offset = end; @@ -1519,35 +1522,31 @@ static noinline int cow_file_range(struct btrfs_inode *inode, * We process each region below. */ - clear_bits = EXTENT_LOCKED | EXTENT_DELALLOC | EXTENT_DELALLOC_NEW | - EXTENT_DEFRAG | EXTENT_CLEAR_META_RESV; - page_ops = PAGE_UNLOCK | PAGE_START_WRITEBACK | PAGE_END_WRITEBACK; - /* * For the range (1). We have already instantiated the ordered extents * for this region. They are cleaned up by * btrfs_cleanup_ordered_extents() in e.g, - * btrfs_run_delalloc_range(). EXTENT_LOCKED | EXTENT_DELALLOC are - * already cleared in the above loop. And, EXTENT_DELALLOC_NEW | - * EXTENT_DEFRAG | EXTENT_CLEAR_META_RESV are handled by the cleanup - * function. + * btrfs_run_delalloc_range(). + * EXTENT_DELALLOC_NEW | EXTENT_DEFRAG | EXTENT_CLEAR_META_RESV + * are also handled by the cleanup function. * - * However, in case of @keep_locked, we still need to unlock the pages - * (except @locked_folio) to ensure all the pages are unlocked. + * So here we only clear EXTENT_LOCKED and EXTENT_DELALLOC flag, + * and finish the writeback of the involved folios, which will be + * never submitted. */ - if (keep_locked && orig_start < start) { + if (orig_start < start) { + clear_bits = EXTENT_LOCKED | EXTENT_DELALLOC; + page_ops = PAGE_UNLOCK | PAGE_START_WRITEBACK | PAGE_END_WRITEBACK; + if (!locked_folio) mapping_set_error(inode->vfs_inode.i_mapping, ret); extent_clear_unlock_delalloc(inode, orig_start, start - 1, - locked_folio, NULL, 0, page_ops); + locked_folio, NULL, clear_bits, page_ops); } - /* - * At this point we're unlocked, we want to make sure we're only - * clearing these flags under the extent lock, so lock the rest of the - * range and clear everything up. - */ - lock_extent(&inode->io_tree, start, end, NULL); + clear_bits = EXTENT_LOCKED | EXTENT_DELALLOC | EXTENT_DELALLOC_NEW | + EXTENT_DEFRAG | EXTENT_CLEAR_META_RESV; + page_ops = PAGE_UNLOCK | PAGE_START_WRITEBACK | PAGE_END_WRITEBACK; /* * For the range (2). If we reserved an extent for our delalloc range -- 2.47.1

9 months, 1 week

2
2
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror December 2024