[5.7,256/376] btrfs: free alien device after device add

Message ID	20200619141722.444467186@linuxfoundation.org
State	New
Headers	show Return-Path: <SRS0=CyoZ=AA=vger.kernel.org=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, Anand Jain <anand.jain@oracle.com>, David Sterba <dsterba@suse.com> Subject: [PATCH 5.7 256/376] btrfs: free alien device after device add Date: Fri, 19 Jun 2020 16:32:54 +0200 Message-Id: <20200619141722.444467186@linuxfoundation.org> In-Reply-To: <20200619141710.350494719@linuxfoundation.org> References: <20200619141710.350494719@linuxfoundation.org> User-Agent: quilt/0.66 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Sender: stable-owner@vger.kernel.org Precedence: bulk
Series	None \| expand [5.7,003/376] ath10k: Fix the race condition in firmware dump work queue [5.7,006/376] media: imx: imx7-mipi-csis: Cleanup and fix subdev pad format handling [5.7,007/376] crypto: ccp -- dont "select" CONFIG_DMADEVICES [5.7,010/376] media: vicodec: Fix error codes in probe function [5.7,013/376] objtool: Ignore empty alternatives [5.7,014/376] drm/amd/display: Force watermark value propagation [5.7,015/376] drm/amd/display: fix virtual signal dsc setup [5.7,016/376] spi: spi-mem: Fix Dual/Quad modes on Octal-capable devices [5.7,018/376] drm/bridge: panel: Return always an error pointer in drm_panel_bridge_add() [5.7,019/376] net: ethernet: ti: fix return value check in k3_cppi_desc_pool_create_name() [5.7,020/376] arm64/kernel: Fix range on invalidating dcache for boot page tables [5.7,022/376] libbpf: Fix memory leak and possible double-free in hashmap__clear [5.7,026/376] net: atlantic: make hw_get_regs optional [5.7,027/376] net: ena: fix error returning in ena_com_get_hash_function() [5.7,030/376] arm64: cacheflush: Fix KGDB trap detection [5.7,032/376] libperf evlist: Fix a refcount leak [5.7,033/376] rtw88: fix an issue about leak system resources [5.7,034/376] spi: dw: Zero DMA Tx and Rx configurations on stack [5.7,035/376] soc: fsl: dpio: properly compute the consumer index [5.7,038/376] arm64: insn: Fix two bugs in encoding 32-bit logical immediates [5.7,039/376] tools/power/x86/intel-speed-select: Fix CLX-N package information output [5.7,040/376] mt76: mt7615: fix aid configuration in mt7615_mcu_wtbl_generic_tlv [5.7,041/376] block: reset mapping if failed to update hardware queue count [5.7,043/376] lockdown: Allow unprivileged users to see lockdown status [5.7,044/376] ixgbe: Fix XDP redirect on archs with PAGE_SIZE above 4K [5.7,045/376] platform/x86: dell-laptop: dont register micmute LED if there is no token [5.7,046/376] cpuidle: psci: Fixup execution order when entering a domain idle state [5.7,048/376] io_uring: cleanup io_poll_remove_one() logic [5.7,049/376] media: i2c: imx219: Fix a bug in imx219_enum_frame_size [5.7,051/376] kgdb: Disable WARN_CONSOLE_UNLOCKED for all kgdb [5.7,052/376] kgdb: Prevent infinite recursive entries to the debugger [5.7,053/376] pmu/smmuv3: Clear IRQ affinity hint on device removal [5.7,054/376] ath11k: Fix some resource leaks in error path in ath11k_thermal_register() [5.7,058/376] spi: dw: Enable interrupts in accordance with DMA xfer mode [5.7,059/376] clocksource/drivers/timer-versatile: Clear OF_POPULATED flag [5.7,061/376] clocksource: dw_apb_timer_of: Fix missing clockevent timers [5.7,065/376] ARM: 8978/1: mm: make act_mm() respect THREAD_SIZE [5.7,066/376] batman-adv: Revert "disable ethtool link speed detection when auto negotiation off" [5.7,069/376] ice: Fix for memory leaks and modify ICE_FREE_CQ_BUFS [5.7,072/376] mmc: meson-mx-sdio: trigger a soft reset after a timeout or CRC error [5.7,074/376] Bluetooth: hci_qca: Fix suspend/resume functionality failure [5.7,075/376] spi: dw: Fix Rx-only DMA transfers [5.7,077/376] x86/kvm/hyper-v: Explicitly align hcall param for kvm_hyperv_exit [5.7,080/376] x86: fix vmap arguments in map_irq_stack [5.7,082/376] ubsan: entirely disable alignment checks under UBSAN_TRAP [5.7,083/376] ath11k: fix error message to correctly report the command that failed [5.7,085/376] ath11k: Avoid mgmt tx count underflow [5.7,087/376] drm/amd/display: Revert to old formula in set_vtg_params [5.7,088/376] media: staging/intel-ipu3: Implement lock for stream on/off operations [5.7,090/376] spi: Respect DataBitLength field of SpiSerialBusV2() ACPI resource [5.7,091/376] brcmfmac: fix wrong location to get firmware feature [5.7,092/376] regulator: qcom-rpmh: Fix typos in pm8150 and pm8150l [5.7,093/376] tools api fs: Make xxx__mountpoint() more scalable [5.7,095/376] net: mscc: ocelot: deal with problematic MAC_ETYPE VCAP IS2 rules [5.7,099/376] audit: fix a net reference leak in audit_send_reply() [5.7,100/376] media: dvb: return -EREMOTEIO on i2c transfer failure. [5.7,103/376] media: platform: fcp: Set appropriate DMA parameters [5.7,106/376] Bluetooth: btbcm: Add 2 missing models to subver tables [5.7,107/376] audit: fix a net reference leak in audit_list_rules_send() [5.7,108/376] drm/amd/display: Correct updating logic of dcn21s pipe VM flags [5.7,110/376] Drivers: hv: vmbus: Always handle the VMBus messages on CPU0 [5.7,111/376] efi/libstub/random: Align allocate size to EFI_ALLOC_ALIGN [5.7,113/376] bcache: remove a duplicate ->make_request_fn assignment [5.7,117/376] ath11k: use GFP_ATOMIC under spin lock [5.7,118/376] Bluetooth: Adding driver and quirk defs for multi-role LE [5.7,121/376] selftests/bpf: Ensure test flavors use correct skeletons [5.7,123/376] selftests/bpf: Fix memory leak in extract_build_id() [5.7,126/376] selftests/bpf: Fix bpf_link leak in ns_current_pid_tgid selftest [5.7,127/376] selftests/bpf: Add runqslower binary to .gitignore [5.7,129/376] ARM: 8969/1: decompressor: simplify libfdt builds [5.7,130/376] drm/bridge: fix stack usage warning on old gcc [5.7,132/376] net: bcmgenet: Fix WoL with password after deep sleep [5.7,135/376] exit: Move preemption fixup up, move blocking operations down [5.7,136/376] sched/core: Fix illegal RCU from offline CPUs [5.7,138/376] arm64: kexec_file: print appropriate variable [5.7,140/376] iocost_monitor: drop string wrap around numbers when outputting json [5.7,141/376] octeontx2-pf: Fix error return code in otx2_probe() [5.7,143/376] net: lpc-enet: fix error return code in lpc_mii_init() [5.7,144/376] selinux: fix error return code in policydb_read() [5.7,146/376] cpufreq: qcom: fix wrong compatible binding [5.7,147/376] ath10k: fix possible memory leak in ath10k_bmi_lz_data_large() [5.7,148/376] ath11k: fix error return code in ath11k_dp_alloc() [5.7,149/376] media: sun8i: Fix an error handling path in deinterlace_runtime_resume() [5.7,153/376] ath11k: fix kernel panic by freeing the msdu received with invalid length [5.7,154/376] libertas_tf: avoid a null dereference in pointer priv [5.7,155/376] xfs: clean up the error handling in xfs_swap_extents [5.7,157/376] Crypto/chcr: fix for ccm(aes) failed test [5.7,158/376] dsa: sja1105: dynamically allocate stats structure [5.7,159/376] MIPS: Truncate link address into 32bit for 32bit kernel [5.7,161/376] kgdb: Fix spurious true from in_dbg_master() [5.7,163/376] xfs: fix duplicate verification from xfs_qm_dqflush() [5.7,165/376] platform/x86: intel-vbtn: Split keymap into buttons and switches parts [5.7,168/376] iwlwifi: avoid debug max amsdu config overwriting itself [5.7,169/376] nvme: refine the Qemu Identify CNS quirk [5.7,170/376] nvme-fc: avoid gcc-10 zero-length-bounds warning [5.7,172/376] nvme-tcp: use bh_lock in data_ready [5.7,175/376] wcn36xx: Fix error handling path in wcn36xx_probe() [5.7,184/376] mt76: avoid rx reorder buffer overflow [5.7,186/376] brcmfmac: fix WPA/WPA2-PSK 4-way handshake offload and SAE offload failures [5.7,188/376] raid5: remove gfp flags from scribble_alloc() [5.7,189/376] iocost: dont let vrate run wild while theres no saturation signal [5.7,191/376] crypto: blake2b - Fix clang optimization for ARMv7-M [5.7,193/376] net/mlx5e: IPoIB, Drop multicast packets that this interface sent [5.7,195/376] net: ipa: do not clear interrupt in gsi_channel_start() [5.7,196/376] rtlwifi: Fix a double free in _rtl_usb_tx_urb_setup() [5.7,197/376] mwifiex: Fix memory corruption in dump_station [5.7,199/376] x86/boot: Correct relocation destination on old linkers [5.7,201/376] sched: Defend cfs and rt bandwidth quota against overflow [5.7,202/376] mips: MAAR: Use more precise address mask [5.7,204/376] ice: Fix resource leak on early exit from function [5.7,206/376] crypto: stm32/crc32 - fix ext4 chksum BUG_ON() [5.7,208/376] crypto: stm32/crc32 - fix multi-instance [5.7,209/376] drm/amd/powerpay: Disable gfxoff when setting manual mode on picasso and raven [5.7,210/376] drm/amdgpu: Sync with VM root BO when switching VM to CPU update mode [5.7,211/376] selftests/bpf: CONFIG_IPV6_SEG6_BPF required for test_seg6_loop.o [5.7,212/376] selftests/bpf: CONFIG_LIRC required for test_lirc_mode2.sh [5.7,214/376] x86/mm: Stop printing BRK addresses [5.7,215/376] MIPS: Fix exception handler memcpy() [5.7,218/376] btrfs: improve global reserve stealing logic [5.7,221/376] net: ethernet: fec: move GPR register offset and bit into DT [5.7,222/376] nvme-pci: make sure write/poll_queues less or equal then cpu count [5.7,223/376] nvmet: fix memory leak when removing namespaces and controllers concurrently [5.7,226/376] MIPS: Fix IRQ tracing when call handle_fpe() and handle_msa_fpe() [5.7,228/376] mmc: mmci: Switch to mmc_regulator_set_vqmmc() [5.7,230/376] staging: greybus: sdio: Respect the cmd->busy_timeout from the mmc core [5.7,231/376] mmc: owl-mmc: Respect the cmd->busy_timeout from the mmc core [5.7,232/376] mmc: via-sdmmc: Respect the cmd->busy_timeout from the mmc core [5.7,234/376] ice: fix potential double free in probe unrolling [5.7,235/376] ixgbe: fix signed-integer-overflow warning [5.7,237/376] mmc: sdhci-esdhc-imx: fix the mask for tuning start point [5.7,238/376] spi: dw: Return any value retrieved from the dma_transfer callback [5.7,241/376] io_uring: fix overflowed reqs cancellation [5.7,245/376] platform/x86: intel-vbtn: Only blacklist SW_TABLET_MODE on the 9 / "Laptop" chasis-... [5.7,247/376] libbpf: Fix perf_buffer__free() API for sparse allocs [5.7,249/376] bpf: Refactor sockmap redirect code so its easy to reuse [5.7,251/376] selftests/bpf, flow_dissector: Close TAP device FD after the test [5.7,254/376] kasan: stop tests being eliminated as dead code with FORTIFY_SOURCE [5.7,255/376] string.h: fix incompatibility between FORTIFY_SOURCE and KASAN [5.7,256/376] btrfs: free alien device after device add [5.7,258/376] btrfs: fix a race between scrub and block group removal/allocation [5.7,261/376] btrfs: reloc: fix reloc root leak and NULL pointer dereference [5.7,264/376] btrfs: fix wrong file range cleanup after an error filling dealloc range [5.7,265/376] btrfs: fix space_info bytes_may_use underflow after nocow buffered write [5.7,266/376] btrfs: fix space_info bytes_may_use underflow during space cache writeout [5.7,268/376] mm: thp: make the THP mapcount atomic against __split_huge_pmd_locked() [5.7,271/376] mm/pagealloc.c: call touch_nmi_watchdog() on max order boundaries in deferred init [5.7,273/376] ima: Fix ima digest hash table key calculation [5.7,275/376] ima: Evaluate error in init_ima() [5.7,276/376] ima: Directly assign the ima_default_policy pointer to ima_rules [5.7,277/376] ima: Call ima_calc_boot_aggregate() in ima_eventdigest_init() [5.7,279/376] evm: Fix possible memory leak in evm_calc_hmac_or_hash() [5.7,283/376] ext4: fix race between ext4_sync_parent() and rename() [5.7,284/376] PCI: Avoid Pericom USB controller OHCI/EHCI PME# defect [5.7,286/376] PCI: Avoid FLR for AMD Starship USB 3.0 [5.7,288/376] serial: 8250_pci: Move Pericom IDs to pci_ids.h [5.7,289/376] x86/amd_nb: Add AMD family 17h model 60h PCI IDs [5.7,291/376] ima: Set again build_ima_appraise variable [5.7,292/376] PCI: Program MPS for RCiEP devices [5.7,297/376] carl9170: remove P2P_GO support [5.7,298/376] clocksource: Remove obsolete ifdef [5.7,300/376] media: cedrus: Program output format during each run [5.7,302/376] Bluetooth: hci_bcm: respect IRQ polarity from DT [5.7,304/376] b43legacy: Fix case where channel status is corrupted [5.7,307/376] media: ov5640: fix use of destroyed mutex [5.7,308/376] usb: musb: mediatek: add reset FADDR to zero in reset interrupt handle [5.7,309/376] clk: mediatek: assign the initial value to clk_init_data of mtk_mux [5.7,312/376] EDAC/amd64: Add AMD family 17h model 60h PCI IDs [5.7,314/376] iommu/vt-d: Allocate domain info for real DMA sub-devices [5.7,315/376] power: vexpress: add suppress_bind_attrs to true [5.7,316/376] power: supply: core: fix HWMON temperature labels [5.7,317/376] power: supply: core: fix memory leak in HWMON error path [5.7,319/376] pinctrl: samsung: Save/restore eint_mask over suspend for EINT_TYPE GPIOs [5.7,321/376] sparc32: fix register window handling in genregs32_[gs]et() [5.7,322/376] sparc64: fix misuses of access_process_vm() in genregs32_[sg]et() [5.7,324/376] dm crypt: avoid truncating the logical block size [5.7,326/376] memory: samsung: exynos5422-dmc: Fix tFAW timings alignment [5.7,328/376] powerpc/fadump: consider reserved ranges while reserving memory [5.7,330/376] kernel/cpu_pm: Fix uninitted local in cpu_pm [5.7,331/376] ARM: tegra: Correct PL310 Auxiliary Control Register initialization [5.7,334/376] ARM: dts: exynos: Fix GPIO polarity for thr GalaxyS3 CM36651 sensors bus [5.7,335/376] ARM: dts: at91: sama5d2_ptc_ek: fix vbus pin [5.7,336/376] ARM: dts: s5pv210: Set keep-power-in-suspend for SDHCI1 on Aries [5.7,338/376] powerpc/32s: Fix another build failure with CONFIG_PPC_KUAP_DEBUG [5.7,340/376] powerpc/kasan: Fix shadow pages allocation failure [5.7,344/376] kbuild: force to build vmlinux if CONFIG_MODVERSION=y [5.7,345/376] virtio-balloon: Disable free page reporting if page poison reporting is not enabled [5.7,346/376] sunrpc: svcauth_gss_register_pseudoflavor must reject duplicate registrations. [5.7,348/376] block: nr_sects_write(): Disable preemption on seqcount write [5.7,350/376] pwm: lpss: Fix get_state runtime-pm reference handling [5.7,353/376] mtd: rawnand: onfi: Fix redundancy detection check [5.7,354/376] mtd: rawnand: brcmnand: fix hamming oob layout [5.7,356/376] mtd: rawnand: sharpsl: Fix the probe error path [5.7,358/376] mtd: rawnand: xway: Fix the probe error path [5.7,360/376] mtd: rawnand: socrates: Fix the probe error path [5.7,363/376] mtd: rawnand: plat_nand: Fix the probe error path [5.7,365/376] mtd: rawnand: mtk: Fix the probe error path [5.7,367/376] w1: omap-hdq: cleanup to add missing newline for some dev_dbg [5.7,370/376] f2fs: dont leak filename in f2fs_try_convert_inline_dir() [5.7,373/376] perf probe: Fix to check blacklist address correctly [5.7,374/376] perf probe: Check address correctness by map instead of _etext [5.7,376/376] perf symbols: Fix kernel maps for kcore and eBPF

Message ID

20200619141722.444467186@linuxfoundation.org

State

New

Headers

From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	stable@vger.kernel.org, Anand Jain <anand.jain@oracle.com>,
	David Sterba <dsterba@suse.com>
Subject: [PATCH 5.7 256/376] btrfs: free alien device after device add
Date: Fri, 19 Jun 2020 16:32:54 +0200
Message-Id: <20200619141722.444467186@linuxfoundation.org>
In-Reply-To: <20200619141710.350494719@linuxfoundation.org>
References: <20200619141710.350494719@linuxfoundation.org>
User-Agent: quilt/0.66
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Sender: stable-owner@vger.kernel.org
Precedence: bulk

Series

None | expand

Commit Message

gregkh@linuxfoundation.org June 19, 2020, 2:32 p.m. UTC

From: Anand Jain <anand.jain@oracle.com>

commit 7f551d969037cc128eca60688d9c5a300d84e665 upstream.

When an old device has new fsid through 'btrfs device add -f <dev>' our
fs_devices list has an alien device in one of the fs_devices lists.

By having an alien device in fs_devices, we have two issues so far

1. missing device does not not show as missing in the userland

2. degraded mount will fail

Both issues are caused by the fact that there's an alien device in the
fs_devices list. (Alien means that it does not belong to the filesystem,
identified by fsid, or does not contain btrfs filesystem at all, eg. due
to overwrite).

A device can be scanned/added through the control device ioctls
SCAN_DEV, DEVICES_READY or by ADD_DEV.

And device coming through the control device is checked against the all
other devices in the lists, but this was not the case for ADD_DEV.

This patch fixes both issues above by removing the alien device.

CC: stable@vger.kernel.org # 5.4+
Signed-off-by: Anand Jain <anand.jain@oracle.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/volumes.c |   12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)

--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -2663,8 +2663,18 @@  int btrfs_init_new_device(struct btrfs_f
 		ret = btrfs_commit_transaction(trans);
 	}
 
-	/* Update ctime/mtime for libblkid */
+	/*
+	 * Now that we have written a new super block to this device, check all
+	 * other fs_devices list if device_path alienates any other scanned
+	 * device.
+	 * We can ignore the return value as it typically returns -EINVAL and
+	 * only succeeds if the device was an alien.
+	 */
+	btrfs_forget_devices(device_path);
+
+	/* Update ctime/mtime for blkid or udev */
 	update_dev_time(device_path);
+
 	return ret;
 
 error_sysfs:

[5.7,256/376] btrfs: free alien device after device add

Commit Message

Patch