[5.10,052/132] btrfs: clear extent buffer uptodate when we fail to write it

From: Josef Bacik <josef@toxicpanda.com>

From: Josef Bacik <josef@toxicpanda.com>

commit c2e39305299f0118298c2201f6d6cc7d3485f29e upstream.

I got dmesg errors on generic/281 on our overnight fstests.  Looking at
the history this happens occasionally, with errors like this

  WARNING: CPU: 0 PID: 673217 at fs/btrfs/extent_io.c:6848 assert_eb_page_uptodate+0x3f/0x50
  CPU: 0 PID: 673217 Comm: kworker/u4:13 Tainted: G        W         5.16.0-rc2+ #469
  Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.13.0-2.fc32 04/01/2014
  Workqueue: btrfs-cache btrfs_work_helper
  RIP: 0010:assert_eb_page_uptodate+0x3f/0x50
  RSP: 0018:ffffae598230bc60 EFLAGS: 00010246
  RAX: 0017ffffc0002112 RBX: ffffebaec4100900 RCX: 0000000000001000
  RDX: ffffebaec45733c7 RSI: ffffebaec4100900 RDI: ffff9fd98919f340
  RBP: 0000000000000d56 R08: ffff9fd98e300000 R09: 0000000000000000
  R10: 0001207370a91c50 R11: 0000000000000000 R12: 00000000000007b0
  R13: ffff9fd98919f340 R14: 0000000001500000 R15: 0000000001cb0000
  FS:  0000000000000000(0000) GS:ffff9fd9fbc00000(0000) knlGS:0000000000000000
  CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
  CR2: 00007f549fcf8940 CR3: 0000000114908004 CR4: 0000000000370ef0
  Call Trace:

   extent_buffer_test_bit+0x3f/0x70
   free_space_test_bit+0xa6/0xc0
   load_free_space_tree+0x1d6/0x430
   caching_thread+0x454/0x630
   ? rcu_read_lock_sched_held+0x12/0x60
   ? rcu_read_lock_sched_held+0x12/0x60
   ? rcu_read_lock_sched_held+0x12/0x60
   ? lock_release+0x1f0/0x2d0
   btrfs_work_helper+0xf2/0x3e0
   ? lock_release+0x1f0/0x2d0
   ? finish_task_switch.isra.0+0xf9/0x3a0
   process_one_work+0x270/0x5a0
   worker_thread+0x55/0x3c0
   ? process_one_work+0x5a0/0x5a0
   kthread+0x174/0x1a0
   ? set_kthread_struct+0x40/0x40
   ret_from_fork+0x1f/0x30

This happens because we're trying to read from a extent buffer page that
is !PageUptodate.  This happens because we will clear the page uptodate
when we have an IO error, but we don't clear the extent buffer uptodate.
If we do a read later and find this extent buffer we'll think its valid
and not return an error, and then trip over this warning.

Fix this by also clearing uptodate on the extent buffer when this
happens, so that we get an error when we do a btrfs_search_slot() and
find this block later.

CC: stable@vger.kernel.org # 5.4+
Signed-off-by: Josef Bacik <josef@toxicpanda.com>
Reviewed-by: David Sterba <dsterba@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 fs/btrfs/extent_io.c |    6 ++++++
 1 file changed, 6 insertions(+)

Message ID	20211213092940.911900222@linuxfoundation.org
State	New
Headers	show Return-Path: <stable-owner@kernel.org> From: Greg Kroah-Hartman <gregkh@linuxfoundation.org> To: linux-kernel@vger.kernel.org Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>, stable@vger.kernel.org, Josef Bacik <josef@toxicpanda.com>, David Sterba <dsterba@suse.com> Subject: [PATCH 5.10 052/132] btrfs: clear extent buffer uptodate when we fail to write it Date: Mon, 13 Dec 2021 10:29:53 +0100 Message-Id: <20211213092940.911900222@linuxfoundation.org> In-Reply-To: <20211213092939.074326017@linuxfoundation.org> References: <20211213092939.074326017@linuxfoundation.org> User-Agent: quilt/0.66 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Precedence: bulk
Series	None \| expand [5.10,002/132] gcc-plugins: simplify GCC plugin-dev capability test [5.10,008/132] HID: add USB_HID dependancy to hid-chicony [5.10,009/132] HID: add USB_HID dependancy on some USB HID drivers [5.10,010/132] HID: bigbenff: prevent null pointer dereference [5.10,012/132] HID: check for valid USB device for many HID drivers [5.10,015/132] IB/hfi1: Fix early init panic [5.10,019/132] can: sja1000: fix use after free in ems_pcmcia_add_card() [5.10,022/132] drm/amdkfd: separate kfd_iommu_resume from kfd_resume [5.10,027/132] nfc: fix potential NULL pointer deref in nfc_genl_dump_ses_done [5.10,028/132] selftests: netfilter: add a vrf+conntrack testcase [5.10,029/132] vrf: dont run conntrack on vrf with !dflt qdisc [5.10,033/132] bonding: make tx_rebalance_counter an atomic [5.10,035/132] seg6: fix the iif in the IPv6 socket control block [5.10,036/132] udp: using datalen to cap max gso segments [5.10,037/132] netfilter: conntrack: annotate data-races around ct->timeout [5.10,039/132] iavf: Fix reporting when setting descriptor count [5.10,041/132] devlink: fix netns refcount leak in devlink_nl_cmd_reload() [5.10,043/132] KVM: x86: Wait for IPIs to be delivered when handling Hyper-V TLB flush hypercall [5.10,044/132] mm: bdi: initialize bdi_min_ratio when bdi is unregistered [5.10,047/132] ALSA: hda/realtek: Fix quirk for TongFang PHxTxX1 [5.10,048/132] ALSA: pcm: oss: Fix negative period/buffer sizes [5.10,049/132] ALSA: pcm: oss: Limit the period size to 16MB [5.10,052/132] btrfs: clear extent buffer uptodate when we fail to write it [5.10,053/132] btrfs: replace the BUG_ON in btrfs_del_root_ref with proper error handling [5.10,055/132] nfsd: fix use-after-free due to delegation race [5.10,056/132] nfsd: Fix nsfd startup race (again) [5.10,059/132] clk: qcom: regmap-mux: fix parent clock lookup [5.10,061/132] can: pch_can: pch_can_rx_normal: fix use after free [5.10,063/132] libata: add horkage for ASMedia 1092 [5.10,064/132] wait: add wake_up_pollfree() [5.10,066/132] signalfd: use wake_up_pollfree() [5.10,067/132] aio: keep poll requests on waitqueue until completed [5.10,069/132] net: mvpp2: fix XDP rx queues registering [5.10,074/132] qede: validate non LSO skb length [5.10,075/132] PM: runtime: Fix pm_runtime_active() kerneldoc comment [5.10,076/132] ASoC: rt5682: Fix crash due to out of scope stack vars [5.10,080/132] ASoC: codecs: wcd934x: return correct value from mixer put [5.10,082/132] RDMA/hns: Do not destroy QP resources in the hw resetting phase [5.10,083/132] clk: imx: use module_platform_driver [5.10,086/132] mtd: rawnand: fsmc: Take instruction delay into account [5.10,089/132] Revert "PCI: aardvark: Fix support for PCI_ROM_ADDRESS1 on emulated bridge" [5.10,090/132] perf tools: Fix SMT detection fast read path [5.10,092/132] dt-bindings: net: Reintroduce PHY no lane swap binding [5.10,095/132] net: altera: set a couple error code in probe() [5.10,097/132] net, neigh: clear whole pneigh_entry at alloc time [5.10,098/132] net/qla3xxx: fix an error code in ql_adapter_up() [5.10,101/132] USB: gadget: zero allocate endpoint 0 buffers [5.10,102/132] usb: core: config: fix validation of wMaxPacketValue entries [5.10,103/132] xhci: Remove CONFIG_USB_DEFAULT_PERSIST to prevent xHCI from runtime suspending [5.10,104/132] usb: core: config: using bit mask instead of individual bits [5.10,108/132] iio: trigger: stm32-timer: fix MODULE_ALIAS [5.10,109/132] iio: stk3310: Dont return error code in interrupt handler [5.10,111/132] iio: ltr501: Dont return error code in trigger handler [5.10,113/132] iio: itg3200: Call iio_trigger_notify_done() on error [5.10,117/132] iio: adc: stm32: fix a current leak by resetting pcsel before disabling vdda [5.10,118/132] iio: adc: axp20x_adc: fix charging current reporting on AXP22x [5.10,119/132] iio: ad7768-1: Call iio_trigger_notify_done() on error [5.10,120/132] iio: accel: kxcjk-1013: Fix possible memory leak in probe and remove [5.10,121/132] csky: fix typo of fpu config macro [5.10,122/132] irqchip/aspeed-scu: Replace update_bits with write_bits. [5.10,125/132] irqchip/irq-gic-v3-its.c: Force synchronisation when issuing INVALL [5.10,126/132] irqchip: nvic: Fix offset for Interrupt Priority Offsets [5.10,128/132] bpf: Add selftests to cover packet access corner cases [5.10,129/132] kbuild: simplify GCC_PLUGINS enablement in dummy-tools/gcc [5.10,132/132] Documentation/Kbuild: Remove references to gcc-plugin.sh

[5.10,052/132] btrfs: clear extent buffer uptodate when we fail to write it

Commit Message

Patch