Public bug reported:
I am running Karmic, freshly installed on an SSD. An upgrade yesterday
brought me from kernel 2.6.31-11.36 to 2.6.31-11.38. Since then the
system feels very slow when reading from disk (even slower than when
running from a hard disk instead of the SSD). "hdparm -t" confirmed
that something is wrong: I got a disk read speed of approximately 20 MB/s
when the day before it was 230 MB/s.
A little experimentation showed that the effect depends on CPU load.
This is a triple core machine and I normally run it with all three
cores loaded with Boinc, niced. If I stop Boinc the disk performance
is nearly back to normal.
A test with un-niced CPU load shows even worse behaviour:
With three CPU-burning processes running, "hdparm -t" shows that the
SSD delivers 10 MB/s and a hard disk (that does 83 MB/s normally)
10 MB/s, too.
This effect does not occur at all with the older kernels.
While I don't have 2.6.31-11.36 anymore I verified this using
2.6.31-10.35: The disk read performance is good independent of
CPU load.
I will attach the dmesg's of a boot with kernel 2.6.31-10.35 and
one with 2.6.31-11.38 (note that the latter boot, as far as dmesg
captures it, took half a second longer; might be related).
ProblemType: Bug
Architecture: amd64
ArecordDevices:
**** List of CAPTURE Hardware Devices ****
card 0: Solo1 [ESS ES1938 (Solo-1)], device 0: es-1938-1946 [ESS Solo-1]
Subdevices: 1/1
Subdevice #0: subdevice #0
AudioDevicesInUse:
USER PID ACCESS COMMAND
/dev/snd/controlC0: lutz 1922 F.... pulseaudio
CRDA: Error: [Errno 2] No such file or directory
Card0.Amixer.info:
Card hw:0 'Solo1'/'ESS ES1938 (Solo-1) rev 0, irq 20'
Mixer name : 'ESS Solo-1'
Components : ''
Controls : 29
Simple ctrls : 22
Card1.Amixer.info:
Card hw:1 'HDMI'/'HDA ATI HDMI at 0xfeaec000 irq 19'
Mixer name : 'ATI R6xx HDMI'
Components : 'HDA:1002aa01,00aa0100,00100000'
Controls : 4
Simple ctrls : 1
Card1.Amixer.values:
Simple mixer control 'IEC958',0
Capabilities: pswitch pswitch-joined
Playback channels: Mono
Mono: Playback [on]
Date: Sun Oct 4 15:18:10 2009
DistroRelease: Ubuntu 9.10
HibernationDevice: RESUME=UUID=5d494d1d-7e76-4d14-a092-b911976f2f28
IwConfig:
lo no wireless extensions.
eth0 no wireless extensions.
ppp0 no wireless extensions.
MachineType: System manufacturer System Product Name
Package: linux-image-2.6.31-11-generic 2.6.31-11.38
ProcCmdLine: BOOT_IMAGE=/vmlinuz-2.6.31-11-generic
root=UUID=da22260c-8cdb-4890-bae8-6ef40a4502ee ro quiet splash
ProcEnviron:
PATH=(custom, user)
LANG=de_DE.UTF-8
SHELL=/bin/bash
ProcVersionSignature: Ubuntu 2.6.31-11.38-generic
RelatedPackageVersions:
linux-backports-modules-2.6.31-11-generic N/A
linux-firmware 1.20
RfKill:
SourcePackage: linux
Uname: Linux 2.6.31-11-generic x86_64
XsessionErrors:
(gnome-settings-daemon:1967): GLib-CRITICAL **: g_propagate_error: assertion
`src != NULL' failed
(gnome-settings-daemon:1967): GLib-CRITICAL **: g_propagate_error: assertion
`src != NULL' failed
(nautilus:2068): Eel-CRITICAL **: eel_preferences_get_boolean: assertion
`preferences_is_initialized ()' failed
(polkit-gnome-authentication-agent-1:2087): GLib-CRITICAL **:
g_once_init_leave: assertion `initialization_value != 0' failed
dmi.bios.date: 11/13/2008
dmi.bios.vendor: American Megatrends Inc.
dmi.bios.version: 1102
dmi.board.asset.tag: To Be Filled By O.E.M.
dmi.board.name: M3A
dmi.board.vendor: ASUSTeK Computer INC.
dmi.board.version: Rev 1.xx
dmi.chassis.asset.tag: Asset-1234567890
dmi.chassis.type: 3
dmi.chassis.vendor: Chassis Manufacture
dmi.chassis.version: Chassis Version
dmi.modalias:
dmi:bvnAmericanMegatrendsInc.:bvr1102:bd11/13/2008:svnSystemmanufacturer:pnSystemProductName:pvrSystemVersion:rvnASUSTeKComputerINC.:rnM3A:rvrRev1.xx:cvnChassisManufacture:ct3:cvrChassisVersion:
dmi.product.name: System Product Name
dmi.product.version: System Version
dmi.sys.vendor: System manufacturer
** Affects: linux (Ubuntu)
Importance: Undecided
Status: New
** Tags: amd64 apport-bug
--
Severe disk performance regression in 2.6.31-11.38
https://bugs.launchpad.net/bugs/442232
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
--
ubuntu-bugs mailing list
ubuntu-bugs@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
