Bug#814427: linux-image-4.3.0-1-amd64: memory leak in rtl_usb
Package: src:linux
Version: 4.3.3-7
Severity: critical
Tags: upstream
Justification: breaks the whole system
The bug triggers the oom-killer after memory gets exhausted.
Rebooting was the only option left.
Feb 10 17:09:01 debian kernel: [91971.006981] expr invoked oom-killer: gfp_mask=0x20858, order=0, oom_score_adj=0
Feb 10 17:13:16 debian kernel: [92224.364209] ssh invoked oom-killer: gfp_mask=0x201da, order=0, oom_score_adj=0
Feb 10 17:17:55 debian kernel: [92494.838897] chromium invoked oom-killer: gfp_mask=0x200da, order=0, oom_score_adj=300
Feb 10 17:18:10 debian kernel: [92512.491930] chromium invoked oom-killer: gfp_mask=0x201da, order=0, oom_score_adj=300
Feb 10 17:18:11 debian kernel: [92512.591229] chromium invoked oom-killer: gfp_mask=0x200da, order=0, oom_score_adj=200
Feb 10 17:18:11 debian kernel: [92512.627263] chromium invoked oom-killer: gfp_mask=0x200da, order=0, oom_score_adj=0
Feb 10 17:18:11 debian kernel: [92512.695305] chromium invoked oom-killer: gfp_mask=0x201da, order=0, oom_score_adj=300
Feb 10 17:18:11 debian kernel: [92518.620065] DOM Worker invoked oom-killer: gfp_mask=0x201da, order=0, oom_score_adj=0
Feb 10 17:19:49 debian kernel: [92616.049141] zsh invoked oom-killer: gfp_mask=0x200da, order=0, oom_score_adj=0
Feb 10 17:19:56 debian kernel: [92623.960756] ssh invoked oom-killer: gfp_mask=0x201da, order=0, oom_score_adj=0
Feb 10 17:19:56 debian kernel: [92624.064644] ssh invoked oom-killer: gfp_mask=0x201da, order=0, oom_score_adj=0
Feb 10 17:19:57 debian kernel: [92624.244884] fping invoked oom-killer: gfp_mask=0x200da, order=0, oom_score_adj=0
Feb 10 17:19:57 debian kernel: [92624.448912] expr invoked oom-killer: gfp_mask=0x200da, order=0, oom_score_adj=0
Feb 10 17:19:57 debian kernel: [92625.881507] chromium invoked oom-killer: gfp_mask=0x200da, order=0, oom_score_adj=300
Feb 10 17:20:07 debian kernel: [92635.538249] chromium invoked oom-killer: gfp_mask=0x200da, order=0, oom_score_adj=300
Feb 10 17:20:08 debian kernel: [92636.414606] Xorg invoked oom-killer: gfp_mask=0x200da, order=0, oom_score_adj=0
Feb 10 17:21:08 debian kernel: [92694.084031] sshd invoked oom-killer: gfp_mask=0x201da, order=0, oom_score_adj=0
Feb 10 17:22:28 debian kernel: [92744.218266] cron invoked oom-killer: gfp_mask=0x20858, order=0, oom_score_adj=0
Feb 10 17:22:28 debian kernel: [92773.248057] smtpd invoked oom-killer: gfp_mask=0x200da, order=0, oom_score_adj=0
Feb 10 17:22:28 debian kernel: [92777.365410] ssh invoked oom-killer: gfp_mask=0x201da, order=0, oom_score_adj=0
Feb 10 17:22:28 debian kernel: [92777.470042] hddtemp invoked oom-killer: gfp_mask=0x200da, order=0, oom_score_adj=0
Feb 10 17:22:28 debian kernel: [92777.514018] cron invoked oom-killer: gfp_mask=0x2040d0, order=0, oom_score_adj=0
Feb 10 17:22:30 debian kernel: [92781.628055] top invoked oom-killer: gfp_mask=0x200da, order=0, oom_score_adj=0
Feb 10 17:26:30 debian kernel: [93018.201374] gconfd-2 invoked oom-killer: gfp_mask=0x200da, order=0, oom_score_adj=0
I found this on the LKML mailing list:
https://lkml.org/lkml/2015/12/6/113
From Peter Wu <>
Subject [PATCH] rtlwifi: fix gigantic memleak in rtl_usb
Date Sun, 6 Dec 2015 18:57:57 +0100
which seems to be related.
The device is:
Bus 003 Device 007: ID 2001:3308 D-Link Corp. DWA-121 802.11n Wireless N 150 Pico Adapter [Realtek RTL8188CUS]
The interface was not set to 'monitor' type, but 'managed'.
I ifdowned the wlan0 iface and:
# lsmod | egrep rtl
rtl8192cu 65536 0
rtl_usb 20480 1 rtl8192cu
rtl8192c_common 49152 1 rtl8192cu
rtlwifi 77824 3 rtl_usb,rtl8192c_common,rtl8192cu
mac80211 626688 3 rtl_usb,rtlwifi,rtl8192cu
cfg80211 540672 2 mac80211,rtlwifi
usbcore 233472 13 uas,uhci_hcd,rtl_usb,usb_storage,usbserial,ohci_hcd,ehci_hcd,ehci_pci,usbhid,ftdi_sio,rtl8192cu,xhci_hcd,xhci_pci
# modprobe -r -v rtl8192cu
rmmod rtl8192cu
rmmod rtl8192c_common
rmmod rtl_usb
rmmod rtlwifi
rmmod mac80211
rmmod cfg80211
rmmod rfkill
vmstat shows memory consumtion stabilized (at least it looks like
that). The leak was around 300 k/s.
Now it looks like the system is slowly recovering. Is no longer
just allocating memory, but deallocating too.
Simplified system information follows:
-- Package-specific info:
** Version:
Linux version 4.3.0-1-amd64 (debian-kernel@lists.debian.org) (gcc version 5.3.1 20160114 (Debian 5.3.1-6) ) #1 SMP Debian 4.3.3-7 (2016-01-19)
** Command line:
BOOT_IMAGE=/vmlinuz-4.3.0-1-amd64 root=/dev/mapper/vg0-root ro cgroup_enable=memory swapaccount vga=795
** Model information
sys_vendor:
product_name:
product_version:
chassis_vendor:
chassis_version:
bios_vendor: Intel Corp.
bios_version: BAP6710H.86A.0067.2011.0526.1448
board_vendor: Intel Corporation
board_name: DP67BA
board_version: AAG10219-300
-- System Information:
Debian Release: stretch/sid
APT prefers testing
APT policy: (500, 'testing'), (99, 'unstable'), (59, 'experimental')
Architecture: amd64 (x86_64)
Foreign Architectures: i386
Kernel: Linux 4.3.0-1-amd64 (SMP w/8 CPU cores)
Locale: LANG=, LC_CTYPE= (charmap=ANSI_X3.4-1968)
Shell: /bin/sh linked to /bin/bash
Init: sysvinit (via /sbin/init)
Versions of packages linux-image-4.3.0-1-amd64 depends on:
ii debconf [debconf-2.0] 1.5.58
ii initramfs-tools [linux-initramfs-tool] 0.120
ii kmod 22-1
ii linux-base 4.0
Versions of packages linux-image-4.3.0-1-amd64 recommends:
ii firmware-linux-free 3.4
ii irqbalance 1.1.0-2
Versions of packages linux-image-4.3.0-1-amd64 suggests:
pn debian-kernel-handbook <none>
ii extlinux 3:6.03+dfsg-11
ii grub-pc 2.02~beta2-36
pn linux-doc-4.3 <none>
Versions of packages linux-image-4.3.0-1-amd64 is related to:
ii firmware-realtek 20160110-1
-- debconf information excluded
Cheers,
--
Cristian
Reply to: