[Kernel-packages] [Bug 1793430] Re: Page leaking in cachefiles_read_backing_file while vmscan is active
This bug was fixed in the package linux - 4.4.0-138.164 --- linux (4.4.0-138.164) xenial; urgency=medium * linux: 4.4.0-138.164 -proposed tracker (LP: #1795582) * Linux 4.4.155 stable release build is broken on ppc64 (LP: #1795662) - powerpc/fadump: Return error when fadump registration fails * Kernel hang on drive pull caused by regression introduced by commit 287922eb0b18 (LP: #1791790) - block: Fix a race between blk_cleanup_queue() and timeout handling * qeth: use vzalloc for QUERY OAT buffer (LP: #1793086) - s390/qeth: use vzalloc for QUERY OAT buffer * Page leaking in cachefiles_read_backing_file while vmscan is active (LP: #1793430) - SAUCE: cachefiles: Page leaking in cachefiles_read_backing_file while vmscan is active * Bugfix for handling of shadow doorbell buffer (LP: #1788222) - nvme-pci: add a memory barrier to nvme_dbbuf_update_and_check_event * Xenial update to 4.4.155 stable release (LP: #1792419) - net: 6lowpan: fix reserved space for single frames - net: mac802154: tx: expand tailroom if necessary - 9p/net: Fix zero-copy path in the 9p virtio transport - net: lan78xx: Fix misplaced tasklet_schedule() call - spi: davinci: fix a NULL pointer dereference - drm/i915/userptr: reject zero user_size - powerpc/fadump: handle crash memory ranges array index overflow - powerpc/pseries: Fix endianness while restoring of r3 in MCE handler. - fs/9p/xattr.c: catch the error of p9_client_clunk when setting xattr failed - 9p/virtio: fix off-by-one error in sg list bounds check - net/9p/client.c: version pointer uninitialized - net/9p/trans_fd.c: fix race-condition by flushing workqueue before the kfree() - dm cache metadata: save in-core policy_hint_size to on-disk superblock - iio: ad9523: Fix displayed phase - iio: ad9523: Fix return value for ad952x_store() - vmw_balloon: fix inflation of 64-bit GFNs - vmw_balloon: do not use 2MB without batching - vmw_balloon: VMCI_DOORBELL_SET does not check status - vmw_balloon: fix VMCI use when balloon built into kernel - tracing: Do not call start/stop() functions when tracing_on does not change - tracing/blktrace: Fix to allow setting same value - kthread, tracing: Don't expose half-written comm when creating kthreads - uprobes: Use synchronize_rcu() not synchronize_sched() - 9p: fix multiple NULL-pointer-dereferences - PM / sleep: wakeup: Fix build error caused by missing SRCU support - pnfs/blocklayout: off by one in bl_map_stripe() - ARM: tegra: Fix Tegra30 Cardhu PCA954x reset - mm/tlb: Remove tlb_remove_table() non-concurrent condition - iommu/vt-d: Add definitions for PFSID - iommu/vt-d: Fix dev iotlb pfsid use - osf_getdomainname(): use copy_to_user() - sys: don't hold uts_sem while accessing userspace memory - userns: move user access out of the mutex - ubifs: Fix memory leak in lprobs self-check - Revert "UBIFS: Fix potential integer overflow in allocation" - ubifs: Check data node size before truncate - ubifs: Fix synced_i_size calculation for xattr inodes - pwm: tiehrpwm: Fix disabling of output of PWMs - fb: fix lost console when the user unplugs a USB adapter - udlfb: set optimal write delay - getxattr: use correct xattr length - bcache: release dc->writeback_lock properly in bch_writeback_thread() - perf auxtrace: Fix queue resize - fs/quota: Fix spectre gadget in do_quotactl - x86/io: add interface to reserve io memtype for a resource range. (v1.1) - drm/drivers: add support for using the arch wc mapping API. - Linux 4.4.155 * Xenial update to 4.4.154 stable release (LP: #1792392) - sched/sysctl: Check user input value of sysctl_sched_time_avg - Cipso: cipso_v4_optptr enter infinite loop - vti6: fix PMTU caching and reporting on xmit - xfrm: fix missing dst_release() after policy blocking lbcast and multicast - xfrm: free skb if nlsk pointer is NULL - mac80211: add stations tied to AP_VLANs during hw reconfig - nl80211: Add a missing break in parse_station_flags - drm/bridge: adv7511: Reset registers on hotplug - scsi: libiscsi: fix possible NULL pointer dereference in case of TMF - drm/imx: imx-ldb: disable LDB on driver bind - drm/imx: imx-ldb: check if channel is enabled before printing warning - usb: gadget: r8a66597: Fix two possible sleep-in-atomic-context bugs in init_controller() - usb: gadget: r8a66597: Fix a possible sleep-in-atomic-context bugs in r8a66597_queue() - usb/phy: fix PPC64 build errors in phy-fsl-usb.c - tools: usb: ffs-test: Fix build on big endian systems - usb: gadget: f_uac2: fix endianness of 'struct cntrl_*_lay3' - tools/power turbostat: fix -S on UP systems - net: caif: Add a missing rcu_read_unlock() in caif_flow_cb - qed: Fix possible race for the link state value.
[Kernel-packages] [Bug 1793430] Re: Page leaking in cachefiles_read_backing_file while vmscan is active
This bug was fixed in the package linux - 4.15.0-38.41 --- linux (4.15.0-38.41) bionic; urgency=medium * linux: 4.15.0-38.41 -proposed tracker (LP: #1797061) * Silent data corruption in Linux kernel 4.15 (LP: #1796542) - block: add a lower-level bio_add_page interface - block: bio_iov_iter_get_pages: fix size of last iovec - blkdev: __blkdev_direct_IO_simple: fix leak in error case - block: bio_iov_iter_get_pages: pin more pages for multi-segment IOs linux (4.15.0-37.40) bionic; urgency=medium * linux: 4.15.0-37.40 -proposed tracker (LP: #1795564) * hns3: enable ethtool rx-vlan-filter on supported hw (LP: #1793394) - net: hns3: Add vlan filter setting by ethtool command -K * hns3: Modifying channel parameters will reset ring parameters back to defaults (LP: #1793404) - net: hns3: Fix desc num set to default when setting channel * hisi_sas: Add SATA FIX check for v3 hw (LP: #1794151) - scsi: hisi_sas: Add SATA FIS check for v3 hw * Fix potential corruption using SAS controller on HiSilicon arm64 boards (LP: #1794156) - scsi: hisi_sas: add memory barrier in task delivery function * hisi_sas: Reduce unnecessary spin lock contention (LP: #1794165) - scsi: hisi_sas: Tidy hisi_sas_task_prep() * Add functional level reset support for the SAS controller on HiSilicon D06 systems (LP: #1794166) - scsi: hisi_sas: tidy host controller reset function a bit - scsi: hisi_sas: relocate some common code for v3 hw - scsi: hisi_sas: Implement handlers of PCIe FLR for v3 hw * HiSilicon SAS controller doesn't recover from PHY STP link timeout (LP: #1794172) - scsi: hisi_sas: tidy channel interrupt handler for v3 hw - scsi: hisi_sas: Fix the failure of recovering PHY from STP link timeout * getxattr: always handle namespaced attributes (LP: #1789746) - getxattr: use correct xattr length * Fix unusable NVIDIA GPU after S3 (LP: #1793338) - PCI: Reprogram bridge prefetch registers on resume * Fails to boot under Xen PV: BUG: unable to handle kernel paging request at edc21fd9 (LP: #1789118) - x86/EISA: Don't probe EISA bus for Xen PV guests * qeth: use vzalloc for QUERY OAT buffer (LP: #1793086) - s390/qeth: use vzalloc for QUERY OAT buffer * SRU: Enable middle button of touchpad on ThinkPad P72 (LP: #1793463) - Input: elantech - enable middle button of touchpad on ThinkPad P72 * Dell new AIO requires a new uart backlight driver (LP: #1727235) - SAUCE: platform/x86: dell-uart-backlight: new backlight driver for DELL AIO - updateconfigs for Dell UART backlight driver * [Ubuntu] s390/crypto: Fix return code checking in cbc_paes_crypt. (LP: #1794294) - s390/crypto: Fix return code checking in cbc_paes_crypt() * hns3: Retrieve RoCE MSI-X config from firmware (LP: #1793221) - net: hns3: Fix MSIX allocation issue for VF - net: hns3: Refine the MSIX allocation for PF * net: hns: Avoid hang when link is changed while handling packets (LP: #1792209) - net: hns: add the code for cleaning pkt in chip - net: hns: add netif_carrier_off before change speed and duplex * Page leaking in cachefiles_read_backing_file while vmscan is active (LP: #1793430) - SAUCE: cachefiles: Page leaking in cachefiles_read_backing_file while vmscan is active * some nvidia p1000 graphic cards hang during the boot (LP: #1791569) - drm/nouveau/gr/gf100-: virtualise tpc_mask + apply fixes from traces * Error reported when creating ZFS pool with "-t" option, despite successful pool creation (LP: #1769937) - SAUCE: (noup) Update zfs to 0.7.5-1ubuntu16.4 * Fix I2C touchpanels' interrupt storms after system suspend (LP: #1792309) - HID: i2c-hid: Fix flooded incomplete report after S3 on Rayd touchscreen - HID: i2c-hid: Don't reset device upon system resume * ipmmu is always registered (LP: #1783746) - iommu/ipmmu-vmsa: Don't register as BUS IOMMU if machine doesn't have IPMMU- VMSA * Bionic update: upstream stable patchset 2018-09-27 (LP: #1794889) - clocksource/drivers/imx-tpm: Correct some registers operation flow - Input: synaptics-rmi4 - fix an unchecked out of memory error path - KVM: X86: fix incorrect reference of trace_kvm_pi_irte_update - x86: Add check for APIC access address for vmentry of L2 guests - MIPS: io: Prevent compiler reordering writeX() - nfp: ignore signals when communicating with management FW - perf report: Fix switching to another perf.data file - fsnotify: fix ignore mask logic in send_to_group() - MIPS: io: Add barrier after register read in readX() - s390/smsgiucv: disable SMSG on module unload - isofs: fix potential memory leak in mount option parsing - MIPS: dts: Boston: Fix PCI bus dtc warnings: - spi: sh-msiof: Fix bit field overflow writes to TSCR/RSCR - doc: Add vendor prefix for Kieback & Peter GmbH - dt-bindings:
[Kernel-packages] [Bug 1793430] Re: Page leaking in cachefiles_read_backing_file while vmscan is active
This bug was fixed in the package linux - 4.18.0-9.10
---
linux (4.18.0-9.10) cosmic; urgency=medium
* linux: 4.18.0-9.10 -proposed tracker (LP: #1796346)
* Cosmic update: v4.18.12 upstream stable release (LP: #1796139)
- crypto: skcipher - Fix -Wstringop-truncation warnings
- iio: adc: ina2xx: avoid kthread_stop() with stale task_struct
- tsl2550: fix lux1_input error in low light
- misc: ibmvmc: Use GFP_ATOMIC under spin lock
- vmci: type promotion bug in qp_host_get_user_memory()
- siox: don't create a thread without starting it
- x86/numa_emulation: Fix emulated-to-physical node mapping
- staging: rts5208: fix missing error check on call to rtsx_write_register
- power: supply: axp288_charger: Fix initial constant_charge_current value
- misc: sram: enable clock before registering regions
- serial: sh-sci: Stop RX FIFO timer during port shutdown
- uwb: hwa-rc: fix memory leak at probe
- power: vexpress: fix corruption in notifier registration
- iommu/amd: make sure TLB to be flushed before IOVA freed
- Bluetooth: Add a new Realtek 8723DE ID 0bda:b009
- USB: serial: kobil_sct: fix modem-status error handling
- 6lowpan: iphc: reset mac_header after decompress to fix panic
- iommu/msm: Don't call iommu_device_{,un}link from atomic context
- s390/mm: correct allocate_pgste proc_handler callback
- power: remove possible deadlock when unregistering power_supply
- drm/amd/display/dc/dce: Fix multiple potential integer overflows
- drm/amd/display: fix use of uninitialized memory
- md-cluster: clear another node's suspend_area after the copy is finished
- cxgb4: Fix the condition to check if the card is T5
- RDMA/bnxt_re: Fix a couple off by one bugs
- RDMA/i40w: Hold read semaphore while looking after VMA
- RDMA/bnxt_re: Fix a bunch of off by one bugs in qplib_fp.c
- IB/core: type promotion bug in rdma_rw_init_one_mr()
- media: exynos4-is: Prevent NULL pointer dereference in
__isp_video_try_fmt()
- IB/mlx4: Test port number before querying type.
- powerpc/kdump: Handle crashkernel memory reservation failure
- media: fsl-viu: fix error handling in viu_of_probe()
- vhost_net: Avoid tx vring kicks during busyloop
- media: staging/imx: fill vb2_v4l2_buffer field entry
- IB/mlx5: Fix GRE flow specification
- include/rdma/opa_addr.h: Fix an endianness issue
- x86/tsc: Add missing header to tsc_msr.c
- ARM: hwmod: RTC: Don't assume lock/unlock will be called with irq enabled
- x86/entry/64: Add two more instruction suffixes
- ARM: dts: ls1021a: Add missing cooling device properties for CPUs
- scsi: target/iscsi: Make iscsit_ta_authentication() respect the output
buffer size
- thermal: i.MX: Allow thermal probe to fail gracefully in case of bad
calibration.
- scsi: klist: Make it safe to use klists in atomic context
- scsi: ibmvscsi: Improve strings handling
- scsi: target: Avoid that EXTENDED COPY commands trigger lock inversion
- usb: wusbcore: security: cast sizeof to int for comparison
- ath10k: sdio: use same endpoint id for all packets in a bundle
- ath10k: sdio: set skb len for all rx packets
- powerpc/powernv/ioda2: Reduce upper limit for DMA window size
- platform/x86: asus-wireless: Fix uninitialized symbol usage
- ACPI / button: increment wakeup count only when notified
- s390/sysinfo: add missing #ifdef CONFIG_PROC_FS
- alarmtimer: Prevent overflow for relative nanosleep
- s390/dasd: correct numa_node in dasd_alloc_queue
- s390/scm_blk: correct numa_node in scm_blk_dev_setup
- s390/extmem: fix gcc 8 stringop-overflow warning
- mtd: rawnand: atmel: add module param to avoid using dma
- iio: accel: adxl345: convert address field usage in iio_chan_spec
- posix-timers: Make forward callback return s64
- posix-timers: Sanitize overrun handling
- ALSA: snd-aoa: add of_node_put() in error path
- selftests: forwarding: Tweak tc filters for mirror-to-gretap tests
- ath10k: use locked skb_dequeue for rx completions
- media: s3c-camif: ignore -ENOIOCTLCMD from v4l2_subdev_call for s_power
- media: soc_camera: ov772x: correct setting of banding filter
- media: omap3isp: zero-initialize the isp cam_xclk{a,b} initial data
- media: ov772x: add checks for register read errors
- staging: android: ashmem: Fix mmap size validation
- media: ov772x: allow i2c controllers without I2C_FUNC_PROTOCOL_MANGLING
- staging: mt7621-eth: Fix memory leak in mtk_add_mac() error path
- drivers/tty: add error handling for pcmcia_loop_config
- arm64: dts: renesas: salvator-common: Fix adv7482 decimal unit addresses
- serial: pxa: Fix an error handling path in 'serial_pxa_probe()'
- staging: mt7621-dts: Fix remaining pcie warnings
- media: tm6000: add error handling for dvb_register_adapter
- ASoC: qdsp6: qdafe:
[Kernel-packages] [Bug 1793430] Re: Page leaking in cachefiles_read_backing_file while vmscan is active
** Tags removed: verification-needed-bionic ** Tags added: verification-done-bionic -- You received this bug notification because you are a member of Kernel Packages, which is subscribed to linux in Ubuntu. https://bugs.launchpad.net/bugs/1793430 Title: Page leaking in cachefiles_read_backing_file while vmscan is active Status in linux package in Ubuntu: Fix Committed Status in linux source package in Xenial: Fix Committed Status in linux source package in Bionic: Fix Committed Status in linux source package in Cosmic: Fix Committed Bug description: SRU Justification - [Description] In a heavily loaded system where the system pagecache is nearing memory limits and fscache is enabled, pages can be leaked by fscache while trying read pages from cachefiles backend. This can happen because two applications can be reading same page from a single mount, two threads can be trying to read the backing page at same time. This results in one of the thread finding that a page for the backing file or netfs file is already in the radix tree. During the error handling cachefiles does not cleanup the reference on backing page, leading to page leak. [Fix] The fix is straightforward, to decrement the reference when error is encounterd. [Testing] A user has tested the fix using following method for 12+ hrs. 1) mkdir -p /mnt/nfs ; mount -o vers=3,fsc :/export /mnt/nfs 2) create 1 files of 2.8MB in a NFS mount. 3) start a thread to simulate heavy VM presssure (while true ; do echo 3 > /proc/sys/vm/drop_caches ; sleep 1 ; done)& 4) start multiple parallel reader for data set at same time find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & .. .. find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & 5) finally check using cat /proc/fs/fscache/stats | grep -i pages ; free -h , cat /proc/meminfo and page-types -r -b lru to ensure all pages are freed. [Regression Potential] Limited to cachefiles. To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1793430/+subscriptions -- Mailing list: https://launchpad.net/~kernel-packages Post to : kernel-packages@lists.launchpad.net Unsubscribe : https://launchpad.net/~kernel-packages More help : https://help.launchpad.net/ListHelp
[Kernel-packages] [Bug 1793430] Re: Page leaking in cachefiles_read_backing_file while vmscan is active
test log attached ** Attachment added: "test log for fscache page leak." https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1793430/+attachment/5199707/+files/fscache_page_leak_test.log -- You received this bug notification because you are a member of Kernel Packages, which is subscribed to linux in Ubuntu. https://bugs.launchpad.net/bugs/1793430 Title: Page leaking in cachefiles_read_backing_file while vmscan is active Status in linux package in Ubuntu: Fix Committed Status in linux source package in Xenial: Fix Committed Status in linux source package in Bionic: Fix Committed Status in linux source package in Cosmic: Fix Committed Bug description: SRU Justification - [Description] In a heavily loaded system where the system pagecache is nearing memory limits and fscache is enabled, pages can be leaked by fscache while trying read pages from cachefiles backend. This can happen because two applications can be reading same page from a single mount, two threads can be trying to read the backing page at same time. This results in one of the thread finding that a page for the backing file or netfs file is already in the radix tree. During the error handling cachefiles does not cleanup the reference on backing page, leading to page leak. [Fix] The fix is straightforward, to decrement the reference when error is encounterd. [Testing] A user has tested the fix using following method for 12+ hrs. 1) mkdir -p /mnt/nfs ; mount -o vers=3,fsc :/export /mnt/nfs 2) create 1 files of 2.8MB in a NFS mount. 3) start a thread to simulate heavy VM presssure (while true ; do echo 3 > /proc/sys/vm/drop_caches ; sleep 1 ; done)& 4) start multiple parallel reader for data set at same time find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & .. .. find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & 5) finally check using cat /proc/fs/fscache/stats | grep -i pages ; free -h , cat /proc/meminfo and page-types -r -b lru to ensure all pages are freed. [Regression Potential] Limited to cachefiles. To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1793430/+subscriptions -- Mailing list: https://launchpad.net/~kernel-packages Post to : kernel-packages@lists.launchpad.net Unsubscribe : https://launchpad.net/~kernel-packages More help : https://help.launchpad.net/ListHelp
[Kernel-packages] [Bug 1793430] Re: Page leaking in cachefiles_read_backing_file while vmscan is active
Yes we were able to test with 4.15.0-37-generic kernel aswell. Your sosreport has been generated and saved in: /tmp/sosreport-kmodukuri.00195310-20181010154537.tar.xz The checksum is: 8c057643e9995694678915ce550e422c Please send this file to your support representative. ** Attachment added: "sos report after testing on 4.15.0-37-generic" https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1793430/+attachment/5199706/+files/sosreport-kmodukuri.00195310-20181010154537.tar.xz -- You received this bug notification because you are a member of Kernel Packages, which is subscribed to linux in Ubuntu. https://bugs.launchpad.net/bugs/1793430 Title: Page leaking in cachefiles_read_backing_file while vmscan is active Status in linux package in Ubuntu: Fix Committed Status in linux source package in Xenial: Fix Committed Status in linux source package in Bionic: Fix Committed Status in linux source package in Cosmic: Fix Committed Bug description: SRU Justification - [Description] In a heavily loaded system where the system pagecache is nearing memory limits and fscache is enabled, pages can be leaked by fscache while trying read pages from cachefiles backend. This can happen because two applications can be reading same page from a single mount, two threads can be trying to read the backing page at same time. This results in one of the thread finding that a page for the backing file or netfs file is already in the radix tree. During the error handling cachefiles does not cleanup the reference on backing page, leading to page leak. [Fix] The fix is straightforward, to decrement the reference when error is encounterd. [Testing] A user has tested the fix using following method for 12+ hrs. 1) mkdir -p /mnt/nfs ; mount -o vers=3,fsc :/export /mnt/nfs 2) create 1 files of 2.8MB in a NFS mount. 3) start a thread to simulate heavy VM presssure (while true ; do echo 3 > /proc/sys/vm/drop_caches ; sleep 1 ; done)& 4) start multiple parallel reader for data set at same time find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & .. .. find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & 5) finally check using cat /proc/fs/fscache/stats | grep -i pages ; free -h , cat /proc/meminfo and page-types -r -b lru to ensure all pages are freed. [Regression Potential] Limited to cachefiles. To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1793430/+subscriptions -- Mailing list: https://launchpad.net/~kernel-packages Post to : kernel-packages@lists.launchpad.net Unsubscribe : https://launchpad.net/~kernel-packages More help : https://help.launchpad.net/ListHelp
[Kernel-packages] [Bug 1793430] Re: Page leaking in cachefiles_read_backing_file while vmscan is active
All tests passed from Nvidia testing for fscache. Your sosreport has been generated and saved in: /tmp/sosreport-tid1870983.00195310-20181009172638.tar.xz The checksum is: 2f7b671685cf8116920efb63e6397fe2 ** Attachment added: "sosreport-tid1870983.00195310-20181009172638.tar.xz" https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1793430/+attachment/5199600/+files/sosreport-tid1870983.00195310-20181009172638.tar.xz -- You received this bug notification because you are a member of Kernel Packages, which is subscribed to linux in Ubuntu. https://bugs.launchpad.net/bugs/1793430 Title: Page leaking in cachefiles_read_backing_file while vmscan is active Status in linux package in Ubuntu: Fix Committed Status in linux source package in Xenial: Fix Committed Status in linux source package in Bionic: Fix Committed Status in linux source package in Cosmic: Fix Committed Bug description: SRU Justification - [Description] In a heavily loaded system where the system pagecache is nearing memory limits and fscache is enabled, pages can be leaked by fscache while trying read pages from cachefiles backend. This can happen because two applications can be reading same page from a single mount, two threads can be trying to read the backing page at same time. This results in one of the thread finding that a page for the backing file or netfs file is already in the radix tree. During the error handling cachefiles does not cleanup the reference on backing page, leading to page leak. [Fix] The fix is straightforward, to decrement the reference when error is encounterd. [Testing] A user has tested the fix using following method for 12+ hrs. 1) mkdir -p /mnt/nfs ; mount -o vers=3,fsc :/export /mnt/nfs 2) create 1 files of 2.8MB in a NFS mount. 3) start a thread to simulate heavy VM presssure (while true ; do echo 3 > /proc/sys/vm/drop_caches ; sleep 1 ; done)& 4) start multiple parallel reader for data set at same time find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & .. .. find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & 5) finally check using cat /proc/fs/fscache/stats | grep -i pages ; free -h , cat /proc/meminfo and page-types -r -b lru to ensure all pages are freed. [Regression Potential] Limited to cachefiles. To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1793430/+subscriptions -- Mailing list: https://launchpad.net/~kernel-packages Post to : kernel-packages@lists.launchpad.net Unsubscribe : https://launchpad.net/~kernel-packages More help : https://help.launchpad.net/ListHelp
[Kernel-packages] [Bug 1793430] Re: Page leaking in cachefiles_read_backing_file while vmscan is active
** Tags removed: verification-needed-xenial ** Tags added: verification-done-xenial -- You received this bug notification because you are a member of Kernel Packages, which is subscribed to linux in Ubuntu. https://bugs.launchpad.net/bugs/1793430 Title: Page leaking in cachefiles_read_backing_file while vmscan is active Status in linux package in Ubuntu: Fix Committed Status in linux source package in Xenial: Fix Committed Status in linux source package in Bionic: Fix Committed Status in linux source package in Cosmic: Fix Committed Bug description: SRU Justification - [Description] In a heavily loaded system where the system pagecache is nearing memory limits and fscache is enabled, pages can be leaked by fscache while trying read pages from cachefiles backend. This can happen because two applications can be reading same page from a single mount, two threads can be trying to read the backing page at same time. This results in one of the thread finding that a page for the backing file or netfs file is already in the radix tree. During the error handling cachefiles does not cleanup the reference on backing page, leading to page leak. [Fix] The fix is straightforward, to decrement the reference when error is encounterd. [Testing] A user has tested the fix using following method for 12+ hrs. 1) mkdir -p /mnt/nfs ; mount -o vers=3,fsc :/export /mnt/nfs 2) create 1 files of 2.8MB in a NFS mount. 3) start a thread to simulate heavy VM presssure (while true ; do echo 3 > /proc/sys/vm/drop_caches ; sleep 1 ; done)& 4) start multiple parallel reader for data set at same time find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & .. .. find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & 5) finally check using cat /proc/fs/fscache/stats | grep -i pages ; free -h , cat /proc/meminfo and page-types -r -b lru to ensure all pages are freed. [Regression Potential] Limited to cachefiles. To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1793430/+subscriptions -- Mailing list: https://launchpad.net/~kernel-packages Post to : kernel-packages@lists.launchpad.net Unsubscribe : https://launchpad.net/~kernel-packages More help : https://help.launchpad.net/ListHelp
[Kernel-packages] [Bug 1793430] Re: Page leaking in cachefiles_read_backing_file while vmscan is active
we are trying to get hardware resource to test this before 10/9. will keep you posted. -- You received this bug notification because you are a member of Kernel Packages, which is subscribed to linux in Ubuntu. https://bugs.launchpad.net/bugs/1793430 Title: Page leaking in cachefiles_read_backing_file while vmscan is active Status in linux package in Ubuntu: Fix Committed Status in linux source package in Xenial: Fix Committed Status in linux source package in Bionic: Fix Committed Status in linux source package in Cosmic: Fix Committed Bug description: SRU Justification - [Description] In a heavily loaded system where the system pagecache is nearing memory limits and fscache is enabled, pages can be leaked by fscache while trying read pages from cachefiles backend. This can happen because two applications can be reading same page from a single mount, two threads can be trying to read the backing page at same time. This results in one of the thread finding that a page for the backing file or netfs file is already in the radix tree. During the error handling cachefiles does not cleanup the reference on backing page, leading to page leak. [Fix] The fix is straightforward, to decrement the reference when error is encounterd. [Testing] A user has tested the fix using following method for 12+ hrs. 1) mkdir -p /mnt/nfs ; mount -o vers=3,fsc :/export /mnt/nfs 2) create 1 files of 2.8MB in a NFS mount. 3) start a thread to simulate heavy VM presssure (while true ; do echo 3 > /proc/sys/vm/drop_caches ; sleep 1 ; done)& 4) start multiple parallel reader for data set at same time find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & .. .. find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & 5) finally check using cat /proc/fs/fscache/stats | grep -i pages ; free -h , cat /proc/meminfo and page-types -r -b lru to ensure all pages are freed. [Regression Potential] Limited to cachefiles. To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1793430/+subscriptions -- Mailing list: https://launchpad.net/~kernel-packages Post to : kernel-packages@lists.launchpad.net Unsubscribe : https://launchpad.net/~kernel-packages More help : https://help.launchpad.net/ListHelp
[Kernel-packages] [Bug 1793430] Re: Page leaking in cachefiles_read_backing_file while vmscan is active
This bug is awaiting verification that the kernel in -proposed solves the problem. Please test the kernel and update this bug with the results. If the problem is solved, change the tag 'verification-needed- xenial' to 'verification-done-xenial'. If the problem still exists, change the tag 'verification-needed-xenial' to 'verification-failed- xenial'. If verification is not done by 5 working days from today, this fix will be dropped from the source code, and this bug will be closed. See https://wiki.ubuntu.com/Testing/EnableProposed for documentation how to enable and use -proposed. Thank you! ** Tags added: verification-needed-xenial -- You received this bug notification because you are a member of Kernel Packages, which is subscribed to linux in Ubuntu. https://bugs.launchpad.net/bugs/1793430 Title: Page leaking in cachefiles_read_backing_file while vmscan is active Status in linux package in Ubuntu: Fix Committed Status in linux source package in Xenial: Fix Committed Status in linux source package in Bionic: Fix Committed Status in linux source package in Cosmic: Fix Committed Bug description: SRU Justification - [Description] In a heavily loaded system where the system pagecache is nearing memory limits and fscache is enabled, pages can be leaked by fscache while trying read pages from cachefiles backend. This can happen because two applications can be reading same page from a single mount, two threads can be trying to read the backing page at same time. This results in one of the thread finding that a page for the backing file or netfs file is already in the radix tree. During the error handling cachefiles does not cleanup the reference on backing page, leading to page leak. [Fix] The fix is straightforward, to decrement the reference when error is encounterd. [Testing] A user has tested the fix using following method for 12+ hrs. 1) mkdir -p /mnt/nfs ; mount -o vers=3,fsc :/export /mnt/nfs 2) create 1 files of 2.8MB in a NFS mount. 3) start a thread to simulate heavy VM presssure (while true ; do echo 3 > /proc/sys/vm/drop_caches ; sleep 1 ; done)& 4) start multiple parallel reader for data set at same time find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & .. .. find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & 5) finally check using cat /proc/fs/fscache/stats | grep -i pages ; free -h , cat /proc/meminfo and page-types -r -b lru to ensure all pages are freed. [Regression Potential] Limited to cachefiles. To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1793430/+subscriptions -- Mailing list: https://launchpad.net/~kernel-packages Post to : kernel-packages@lists.launchpad.net Unsubscribe : https://launchpad.net/~kernel-packages More help : https://help.launchpad.net/ListHelp
[Kernel-packages] [Bug 1793430] Re: Page leaking in cachefiles_read_backing_file while vmscan is active
This bug is awaiting verification that the kernel in -proposed solves the problem. Please test the kernel and update this bug with the results. If the problem is solved, change the tag 'verification-needed- bionic' to 'verification-done-bionic'. If the problem still exists, change the tag 'verification-needed-bionic' to 'verification-failed- bionic'. If verification is not done by 5 working days from today, this fix will be dropped from the source code, and this bug will be closed. See https://wiki.ubuntu.com/Testing/EnableProposed for documentation how to enable and use -proposed. Thank you! ** Tags added: verification-needed-bionic -- You received this bug notification because you are a member of Kernel Packages, which is subscribed to linux in Ubuntu. https://bugs.launchpad.net/bugs/1793430 Title: Page leaking in cachefiles_read_backing_file while vmscan is active Status in linux package in Ubuntu: Fix Committed Status in linux source package in Xenial: Fix Committed Status in linux source package in Bionic: Fix Committed Status in linux source package in Cosmic: Fix Committed Bug description: SRU Justification - [Description] In a heavily loaded system where the system pagecache is nearing memory limits and fscache is enabled, pages can be leaked by fscache while trying read pages from cachefiles backend. This can happen because two applications can be reading same page from a single mount, two threads can be trying to read the backing page at same time. This results in one of the thread finding that a page for the backing file or netfs file is already in the radix tree. During the error handling cachefiles does not cleanup the reference on backing page, leading to page leak. [Fix] The fix is straightforward, to decrement the reference when error is encounterd. [Testing] A user has tested the fix using following method for 12+ hrs. 1) mkdir -p /mnt/nfs ; mount -o vers=3,fsc :/export /mnt/nfs 2) create 1 files of 2.8MB in a NFS mount. 3) start a thread to simulate heavy VM presssure (while true ; do echo 3 > /proc/sys/vm/drop_caches ; sleep 1 ; done)& 4) start multiple parallel reader for data set at same time find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & .. .. find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & 5) finally check using cat /proc/fs/fscache/stats | grep -i pages ; free -h , cat /proc/meminfo and page-types -r -b lru to ensure all pages are freed. [Regression Potential] Limited to cachefiles. To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1793430/+subscriptions -- Mailing list: https://launchpad.net/~kernel-packages Post to : kernel-packages@lists.launchpad.net Unsubscribe : https://launchpad.net/~kernel-packages More help : https://help.launchpad.net/ListHelp
[Kernel-packages] [Bug 1793430] Re: Page leaking in cachefiles_read_backing_file while vmscan is active
** Changed in: linux (Ubuntu Xenial) Status: New => Fix Committed -- You received this bug notification because you are a member of Kernel Packages, which is subscribed to linux in Ubuntu. https://bugs.launchpad.net/bugs/1793430 Title: Page leaking in cachefiles_read_backing_file while vmscan is active Status in linux package in Ubuntu: Fix Committed Status in linux source package in Xenial: Fix Committed Status in linux source package in Bionic: Fix Committed Status in linux source package in Cosmic: Fix Committed Bug description: SRU Justification - [Description] In a heavily loaded system where the system pagecache is nearing memory limits and fscache is enabled, pages can be leaked by fscache while trying read pages from cachefiles backend. This can happen because two applications can be reading same page from a single mount, two threads can be trying to read the backing page at same time. This results in one of the thread finding that a page for the backing file or netfs file is already in the radix tree. During the error handling cachefiles does not cleanup the reference on backing page, leading to page leak. [Fix] The fix is straightforward, to decrement the reference when error is encounterd. [Testing] A user has tested the fix using following method for 12+ hrs. 1) mkdir -p /mnt/nfs ; mount -o vers=3,fsc :/export /mnt/nfs 2) create 1 files of 2.8MB in a NFS mount. 3) start a thread to simulate heavy VM presssure (while true ; do echo 3 > /proc/sys/vm/drop_caches ; sleep 1 ; done)& 4) start multiple parallel reader for data set at same time find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & .. .. find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & 5) finally check using cat /proc/fs/fscache/stats | grep -i pages ; free -h , cat /proc/meminfo and page-types -r -b lru to ensure all pages are freed. [Regression Potential] Limited to cachefiles. To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1793430/+subscriptions -- Mailing list: https://launchpad.net/~kernel-packages Post to : kernel-packages@lists.launchpad.net Unsubscribe : https://launchpad.net/~kernel-packages More help : https://help.launchpad.net/ListHelp
[Kernel-packages] [Bug 1793430] Re: Page leaking in cachefiles_read_backing_file while vmscan is active
** Changed in: linux (Ubuntu Bionic) Status: New => Fix Committed -- You received this bug notification because you are a member of Kernel Packages, which is subscribed to linux in Ubuntu. https://bugs.launchpad.net/bugs/1793430 Title: Page leaking in cachefiles_read_backing_file while vmscan is active Status in linux package in Ubuntu: Fix Committed Status in linux source package in Xenial: New Status in linux source package in Bionic: Fix Committed Status in linux source package in Cosmic: Fix Committed Bug description: SRU Justification - [Description] In a heavily loaded system where the system pagecache is nearing memory limits and fscache is enabled, pages can be leaked by fscache while trying read pages from cachefiles backend. This can happen because two applications can be reading same page from a single mount, two threads can be trying to read the backing page at same time. This results in one of the thread finding that a page for the backing file or netfs file is already in the radix tree. During the error handling cachefiles does not cleanup the reference on backing page, leading to page leak. [Fix] The fix is straightforward, to decrement the reference when error is encounterd. [Testing] A user has tested the fix using following method for 12+ hrs. 1) mkdir -p /mnt/nfs ; mount -o vers=3,fsc :/export /mnt/nfs 2) create 1 files of 2.8MB in a NFS mount. 3) start a thread to simulate heavy VM presssure (while true ; do echo 3 > /proc/sys/vm/drop_caches ; sleep 1 ; done)& 4) start multiple parallel reader for data set at same time find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & .. .. find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & 5) finally check using cat /proc/fs/fscache/stats | grep -i pages ; free -h , cat /proc/meminfo and page-types -r -b lru to ensure all pages are freed. [Regression Potential] Limited to cachefiles. To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1793430/+subscriptions -- Mailing list: https://launchpad.net/~kernel-packages Post to : kernel-packages@lists.launchpad.net Unsubscribe : https://launchpad.net/~kernel-packages More help : https://help.launchpad.net/ListHelp
[Kernel-packages] [Bug 1793430] Re: Page leaking in cachefiles_read_backing_file while vmscan is active
** Also affects: linux (Ubuntu Bionic) Importance: Undecided Status: New ** Also affects: linux (Ubuntu Cosmic) Importance: Undecided Assignee: Daniel Axtens (daxtens) Status: Confirmed ** Also affects: linux (Ubuntu Xenial) Importance: Undecided Status: New ** Changed in: linux (Ubuntu Cosmic) Status: Confirmed => Fix Committed -- You received this bug notification because you are a member of Kernel Packages, which is subscribed to linux in Ubuntu. https://bugs.launchpad.net/bugs/1793430 Title: Page leaking in cachefiles_read_backing_file while vmscan is active Status in linux package in Ubuntu: Fix Committed Status in linux source package in Xenial: New Status in linux source package in Bionic: New Status in linux source package in Cosmic: Fix Committed Bug description: SRU Justification - [Description] In a heavily loaded system where the system pagecache is nearing memory limits and fscache is enabled, pages can be leaked by fscache while trying read pages from cachefiles backend. This can happen because two applications can be reading same page from a single mount, two threads can be trying to read the backing page at same time. This results in one of the thread finding that a page for the backing file or netfs file is already in the radix tree. During the error handling cachefiles does not cleanup the reference on backing page, leading to page leak. [Fix] The fix is straightforward, to decrement the reference when error is encounterd. [Testing] A user has tested the fix using following method for 12+ hrs. 1) mkdir -p /mnt/nfs ; mount -o vers=3,fsc :/export /mnt/nfs 2) create 1 files of 2.8MB in a NFS mount. 3) start a thread to simulate heavy VM presssure (while true ; do echo 3 > /proc/sys/vm/drop_caches ; sleep 1 ; done)& 4) start multiple parallel reader for data set at same time find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & .. .. find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & find /mnt/nfs -type f | xargs -P 80 cat > /dev/null & 5) finally check using cat /proc/fs/fscache/stats | grep -i pages ; free -h , cat /proc/meminfo and page-types -r -b lru to ensure all pages are freed. [Regression Potential] Limited to cachefiles. To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1793430/+subscriptions -- Mailing list: https://launchpad.net/~kernel-packages Post to : kernel-packages@lists.launchpad.net Unsubscribe : https://launchpad.net/~kernel-packages More help : https://help.launchpad.net/ListHelp
