Bug#658670: linux-image-3.1.0-1-amd64: 3w-9xxx abnormally slows the system down
Package: linux-2.6
Version: 3.1.8-2
Severity: important
Tags: d-i upstream
Dear Sirs,
first of all thank You for all the efforts making Debian GNU/Linux what it
is today. It has definitely come a long way since my first encounter with
rex. At first it seemed that I have found a bug in GNU Parted in the beta
Debian Installer (debian-testing-amd-xfce+lxde-CD-1.iso dated 2012-01-30
05:54) as it cheerfully greeted me that I have found a bug in it. The
important part of the message was: “Unable to determine the size and length
of /dev/sd[a-b][0-6]” (the 3ware controller has a RAID-1 and a RAID-5 (256
KiB stripe size) unit and I created maximum 6 partitions, I created GPT
disk labels). Generally the installer felt rather slow especially after the
above mentioned error message(s). After loosing several days with the wheezy
installer I had a hunch that I should try squeeze's one for the
partitioning. As the partitioning went without a hiccup (resulting in
aligned partition thanks to using MiB as unit in parted) I've returned to
the beta installer and the difference was imminent while waiting looking on
the progression bar on the disk and partition detecting sections. Therefore
I decided having wheezy later and returned to the squeeze installer which
felt (the usual GNU/Linux) lightning fast (although I must confess that only
ssh server task was selected).
After a reboot I've made the dist-upgrade of the just installed base system,
and struck the recognition that it was the Linux kernel, not the beta Debian
Installer that caused me so much pain. Nevertheless I've already started the
dist-upgrade on my notebook, and there was no way to return. However to my
surprise by the time the machine with the 3ware controller finished booting
so did the notebook with the wheezy system. This was surprising because the
system on the notebook can't be accused being a fresh base system.
The vastly different experience with the notebook shed the light on the
specific reason of the system with the 3ware controller being sluggishly slow.
I've searched the Internet and found an e-mail sent by Russell L. CARTER to
the debian-user list (2011-12-02 08:51 UTC-7, subject: 3w-9xxx performance
bug in kernel 3.1.0-amd64 –
http://list.debian.org/debian-user/2011/12/msg00153.html) describing the exact
same problem. However it seems that for whatever reasons he didn't reported
the bug.
With returning to the kernel in linux-image-2.6.32-5-amd64 also did the speed.
This also confirms that the speed degradation with 3.1.0-1-amd64 kernel has
nothing to do with creating optimised XFS file system (mkfs.xfs -d sunit=512,
swidth=1024 -i size=512 -l sunit=512 -s size=512 -L given-label /dev/sdb1) on
top of an aligned (using the “–dataalignment 1024s” option in pvcreate) LVM
volume.
I think that the time required for booting speaks to itself: while with the
older kernel it takes 25 seconds from Grub 2 to have a login shell with the
new one it takes 29 minutes (I exaggerated: actually it takes 15 second less
than that).
I thank You in anticipation for Your efforts in solving this deal breaker
problem for 3ware 9000 series controller users or if it is not possible, than
when the time comes please add it to the errata of the wheezy release.
Yours truly,
Csaba GAAL
-- Package-specific info:
** Version:
Linux version 3.1.0-1-amd64 (Debian 3.1.8-2) (ben@decadent.org.uk) (gcc version 4.6.2 (Debian 4.6.2-11) ) #1 SMP Tue Jan 10 05:01:58 UTC 2012
** Kernel log:
[ 46.184463] [drm] radeon: irq initialized.
[ 46.187278] [drm] Detected VRAM RAM=128M, BAR=128M
[ 46.187281] [drm] RAM width 16bits DDR
[ 46.187361] [TTM] Zone kernel: Available graphics memory: 2030744 kiB.
[ 46.187362] [TTM] Initializing pool allocator.
[ 46.187396] [drm] radeon: 32M of VRAM memory ready
[ 46.187397] [drm] radeon: 512M of GTT memory ready.
[ 46.187417] [drm] GART: num cpu pages 131072, num gpu pages 131072
[ 46.208648] radeon 0000:01:05.0: WB enabled
[ 46.208693] [drm] Loading R100 Microcode
[ 47.176437] [drm] radeon: ring at 0x00000000C0001000
[ 47.176461] [drm] ring test succeeded in 2 usecs
[ 47.176562] [drm] radeon: ib pool ready.
[ 47.176624] [drm] ib test succeeded in 0 usecs
[ 47.176768] [drm] No valid Ext TMDS info found in BIOS
[ 47.176797] [drm] Radeon Display Connectors
[ 47.176799] [drm] Connector 0:
[ 47.176800] [drm] VGA
[ 47.176802] [drm] DDC: 0x60 0x60 0x60 0x60 0x60 0x60 0x60 0x60
[ 47.176803] [drm] Encoders:
[ 47.176804] [drm] CRT1: INTERNAL_DAC1
[ 47.176805] [drm] Connector 1:
[ 47.176806] [drm] DVI-D
[ 47.176807] [drm] HPD2
[ 47.176809] [drm] DDC: 0x64 0x64 0x64 0x64 0x64 0x64 0x64 0x64
[ 47.176810] [drm] Encoders:
[ 47.176812] [drm] DFP2: INTERNAL_DVO1
[ 47.229121] [drm] Radeon display connector VGA-1: Found valid EDID
[ 47.238890] [drm] Radeon display connector DVI-D-1: No monitor connected or invalid EDID
[ 47.307238] [drm] fb mappable at 0xE0040000
[ 47.307240] [drm] vram apper at 0xE0000000
[ 47.307241] [drm] size 1310720
[ 47.307242] [drm] fb depth is 8
[ 47.307243] [drm] pitch is 1280
[ 47.307308] fbcon: radeondrmfb (fb0) is primary device
[ 47.472277] Console: switching to colour frame buffer device 160x64
[ 47.484656] fb0: radeondrmfb frame buffer device
[ 47.484658] drm: registered panic notifier
[ 47.484662] [drm] Initialized radeon 2.11.0 20080528 for 0000:01:05.0 on minor 0
[ 104.076310] Intel AES-NI instructions are not detected.
[ 104.576172] padlock_aes: VIA PadLock not detected.
[ 105.176178] padlock_sha: VIA PadLock Hash Engine not detected.
[ 134.477458] loop: module loaded
[ 135.576289] w83627hf: w83627hf: Found W83627THF chip at 0x290
[ 213.680234] Adding 2228220k swap on /dev/mapper/sda3_crypt. Priority:-1 extents:1 across:2228220k
[ 238.284028] XFS (sda2): Mounting Filesystem
[ 445.388026] XFS (sda2): Ending clean mount
[ 445.588028] XFS (dm-1): Mounting Filesystem
[ 448.188025] XFS (dm-1): Ending clean mount
[ 448.388027] XFS (dm-3): Mounting Filesystem
[ 450.788076] XFS (dm-3): Ending clean mount
[ 450.988026] XFS (dm-4): Mounting Filesystem
[ 453.388025] XFS (dm-4): Ending clean mount
[ 453.588025] XFS (dm-2): Mounting Filesystem
[ 660.896026] XFS (dm-2): Ending clean mount
[ 661.096033] XFS (dm-5): Mounting Filesystem
[ 663.696028] XFS (dm-5): Ending clean mount
[ 663.896027] XFS (dm-8): Mounting Filesystem
[ 871.296039] XFS (dm-8): Ending clean mount
[ 871.496026] XFS (dm-7): Mounting Filesystem
[ 886.596024] XFS (dm-7): Ending clean mount
[ 886.896029] XFS (dm-9): Mounting Filesystem
[ 1094.404025] XFS (dm-9): Ending clean mount
[ 1094.704044] XFS (dm-6): Mounting Filesystem
[ 1302.116025] XFS (dm-6): Ending clean mount
[ 1302.316026] XFS (dm-10): Mounting Filesystem
[ 1360.216027] XFS (dm-10): Ending clean mount
[ 1360.416028] XFS (dm-11): Mounting Filesystem
[ 1364.416025] XFS (dm-11): Ending clean mount
[ 1364.716029] XFS (dm-12): Mounting Filesystem
[ 1572.216025] XFS (dm-12): Ending clean mount
[ 1572.516034] XFS (dm-13): Mounting Filesystem
[ 1575.116025] XFS (dm-13): Ending clean mount
[ 1575.316026] XFS (dm-14): Mounting Filesystem
[ 1588.116025] XFS (dm-14): Ending clean mount
[ 1588.316026] XFS (dm-15): Mounting Filesystem
[ 1591.516027] XFS (dm-15): Ending clean mount
[ 1591.716028] XFS (dm-16): Mounting Filesystem
[ 1631.316025] XFS (dm-16): Ending clean mount
[ 1631.516040] XFS (dm-17): Mounting Filesystem
[ 1633.916025] XFS (dm-17): Ending clean mount
[ 1641.470850] tg3 0000:05:00.0: irq 46 for MSI/MSI-X
[ 1641.516590] ADDRCONF(NETDEV_UP): eth1: link is not ready
[ 1643.067435] tg3 0000:05:00.0: eth1: Link is up at 100 Mbps, full duplex
[ 1643.067438] tg3 0000:05:00.0: eth1: Flow control is on for TX and on for RX
[ 1643.067921] ADDRCONF(NETDEV_CHANGE): eth1: link becomes ready
[ 1647.222786] RPC: Registered named UNIX socket transport module.
[ 1647.222789] RPC: Registered udp transport module.
[ 1647.222791] RPC: Registered tcp transport module.
[ 1647.222792] RPC: Registered tcp NFSv4.1 backchannel transport module.
[ 1648.020623] FS-Cache: Loaded
[ 1648.622861] FS-Cache: Netfs 'nfs' registered for caching
[ 1649.021209] Installing knfsd (copyright (C) 1996 okir@monad.swb.de).
[ 1653.408004] eth1: no IPv6 routers present
[ 1679.424317] powernow-k8: Found 1 AMD Processor model unknown (4 cpu cores) (version 2.20.00)
[ 1679.424358] powernow-k8: 0 : pstate 0 (2800 MHz)
[ 1679.424360] powernow-k8: 1 : pstate 1 (2100 MHz)
[ 1679.424361] powernow-k8: 2 : pstate 2 (1600 MHz)
[ 1679.424363] powernow-k8: 3 : pstate 3 (800 MHz)
[ 7335.428019] 3w-9xxx: scsi0: AEN: INFO (0x04:0x000C): Initialize started:unit=0.
** Model information
sys_vendor: ASUSTek Computer INC.
product_name: M2N-LR
product_version: Rev 1.xx
chassis_vendor: ASUSTek Computer INC.
chassis_version: Rev 1.xx
bios_vendor: American Megatrends Inc.
bios_version: 0515
board_vendor: ASUSTeK Computer INC.
board_name: M2N-LR
board_version: Rev 1.xx
Kernel: Linux 3.1.0-1-amd64 (SMP w/4 CPU cores)
Locale: LANG=hu_HU.UTF-8, LC_CTYPE=hu_HU.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/bash
Versions of packages linux-image-3.1.0-1-amd64 depends on:
ii debconf [debconf-2.0] 1.5.41
ii initramfs-tools [linux-initramfs-tool] 0.99
ii linux-base 3.4
ii module-init-tools 3.16-1
Versions of packages linux-image-3.1.0-1-amd64 recommends:
ii firmware-linux-free 3
Versions of packages linux-image-3.1.0-1-amd64 is related to:
ii firmware-bnx2 <none>
ii firmware-bnx2x <none>
ii firmware-ipw2x00 <none>
ii firmware-ivtv <none>
ii firmware-iwlwifi <none>
ii firmware-linux 0.35
ii firmware-linux-nonfree 0.35
ii firmware-qlogic <none>
ii firmware-ralink <none>
ii xen-hypervisor-4.1-amd64 [xen-hypervisor] 4.1.2-2
--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 20120205012439.14292.14144.reportbug@cirmos.presse matic.hu">http://lists.debian.org/20120205012439.14292.14144.reportbug@cirmos.presse matic.hu