USB file transfers extremely slow and eventually cause kernel freezes

Bug #183839 reported by Gustavo Carneiro
70
This bug affects 9 people
Affects Status Importance Assigned to Milestone
linux (Ubuntu)
Fix Released
Undecided
Unassigned
Nominated for Karmic by Berte

Bug Description

Binary package hint: linux-source-2.6.24

Every time I transfer a large file from my USB driver, the transfer starts extremely slow, then the kernel eventually freezes. This happened to me with all 2.6.24-[24] kernels so far, but the old gutsy 2.6.22-14 kernel that I luckily still have lying around works flawlessly.

Revision history for this message
Gustavo Carneiro (gjc) wrote :

kernel log of the last boot before the crash

Revision history for this message
Gustavo Carneiro (gjc) wrote :

OK, so kernel 2.6.22-14 was not exactly flawless either:

[ 1611.873718] usb 5-7: reset high speed USB device using ehci_hcd and address 3
[ 1612.133324] usb 5-7: reset high speed USB device using ehci_hcd and address 3
[ 1612.266244] sd 2:0:0:0: [sdb] Result: hostbyte=DID_ERROR driverbyte=DRIVER_OK,SUGGEST_OK
[ 1612.266253] end_request: I/O error, dev sdb, sector 222381591
[ 1613.942588] usb 5-7: reset high speed USB device using ehci_hcd and address 3
[ 1614.206179] usb 5-7: reset high speed USB device using ehci_hcd and address 3

Some errors like this appear, and nautilus gives me some errors, but copying resumes and progresses after I press retry. At least it does not freeze on me :P

Revision history for this message
Leann Ogasawara (leannogasawara) wrote :

Hi Gustavo,

Also per the kernel team's bug policy, can you please attach the additional information. Please be sure to attach each file as a separate attachment.

* cat /proc/version_signature > version.log
* sudo lspci -vvnn > lspci-vvnn.log

For more information regarding the kernel team bug policy, please refer to https://wiki.ubuntu.com/KernelTeamBugPolicies . Thanks again and we appreciate your help and feedback.

Changed in linux:
status: New → Incomplete
Revision history for this message
Gustavo Carneiro (gjc) wrote :
Revision history for this message
Gustavo Carneiro (gjc) wrote :
Changed in linux:
status: Incomplete → New
Matthew Woerly (nattgew)
Changed in linux:
status: New → Incomplete
Revision history for this message
Gustavo Carneiro (gjc) wrote :

This bug appears to be fixed in the new 2.6.24-5 kernel.

Changed in linux:
status: Incomplete → Fix Released
Revision history for this message
Daniel Swarbrick (pressureman) wrote :

I'm seeing similar results with Hardy and kernel 2.6.24-12. Copy an ISO file for example to my USB memory stick takes around 15-18 minutes, and Nautilus's file operations dialog reports transfer speeds of around 700 KB/sec. Copying the same file under Windows finishes in around 4-5 minutes.

I don't think this is Nautilus or GVFS related, since copying from is also extremely slow.

Revision history for this message
Daniel Swarbrick (pressureman) wrote :

I just tried booting in rescue mode, opened a shell as root and copied the same ISO file to the USB stick with Midnight Commander, and it copied MUCH faster - averaging 5.6 MB/sec.

I then tried booting normally, but at GDM screen switched to a text vty, and repeated the copy with MC again. Still good, about 5.5 MB/sec.

I then tried logging into Gnome, opening a terminal and copying the same file with MC again... it gets to about 25-30% complete, then just gets slower, and slower, and slower, eventually bottoming out at around 650 KB/sec.

This looks like it's something related to whatever is running during a Gnome session. Maybe GVFS-related, but I doubt it, since I specifically mounted the USB stick manually as root from a bash prompt, and used MC to copy (which wouldn't be using GVFS).

Revision history for this message
Lito (lito-eordes) wrote :
Download full text (5.2 KiB)

I have the same problem with a clean Ubuntu Hardy 8.04 installation.

I can copy some big files but others I can't. The original file isn't corrupted. The transfer speed is very very slow. I attach my log files. In windows works fine.

The pendrive is a 8Gb USB 2.0 Flash Disk in a Linux 2.6.24-12-generic #1 SMP Wed Mar 12 23:01:54 UTC 2008 i686 GNU/Linux

The continuous error log:

[12894.060439] usb 2-2: new high speed USB device using ehci_hcd and address 22
[12894.193325] usb 2-2: configuration #1 chosen from 1 choice
[12894.207708] scsi9 : SCSI emulation for USB Mass Storage devices
[12894.212951] usb-storage: device found at 22
[12894.212960] usb-storage: waiting for device to settle before scanning
[12899.210550] usb-storage: device scan complete
[12899.211781] scsi 9:0:0:0: Direct-Access USB 2.0 Flash Disk 5.00 PQ: 0 ANSI: 2
[12899.223506] sd 9:0:0:0: [sdf] 16384000 512-byte hardware sectors (8389 MB)
[12899.225375] sd 9:0:0:0: [sdf] Write Protect is off
[12899.225382] sd 9:0:0:0: [sdf] Mode Sense: 0b 00 00 08
[12899.225387] sd 9:0:0:0: [sdf] Assuming drive cache: write through
[12899.229746] sd 9:0:0:0: [sdf] 16384000 512-byte hardware sectors (8389 MB)
[12899.231373] sd 9:0:0:0: [sdf] Write Protect is off
[12899.231379] sd 9:0:0:0: [sdf] Mode Sense: 0b 00 00 08
[12899.231383] sd 9:0:0:0: [sdf] Assuming drive cache: write through
[12899.231392] sdf: sdf1
[12899.233688] sd 9:0:0:0: [sdf] Attached SCSI removable disk
[12899.233755] sd 9:0:0:0: Attached scsi generic sg5 type 0
[12998.488627] usb 2-2: reset high speed USB device using ehci_hcd and address 22
[13016.752977] usb 2-2: device descriptor read/64, error -110
[ 7239.431809] usb 2-2: device descriptor read/64, error -110
[ 7239.647676] usb 2-2: reset high speed USB device using ehci_hcd and address 22
[13053.141648] usb 2-2: device descriptor read/64, error -110
[ 6528.282910] usb 2-2: device descriptor read/64, error -110
[ 6528.496265] usb 2-2: reset high speed USB device using ehci_hcd and address 22
[13082.931139] usb 2-2: device not accepting address 22, error -110
[13083.043129] usb 2-2: reset high speed USB device using ehci_hcd and address 22
[13093.447038] usb 2-2: device not accepting address 22, error -110
[13093.447475] usb 2-2: USB disconnect, address 22
[13093.447897] sd 9:0:0:0: Device offlined - not ready after error recovery
[13093.448442] sd 9:0:0:0: [sdf] Result: hostbyte=DID_NO_CONNECT driverbyte=DRIVER_OK,SUGGEST_OK
[13093.448450] end_request: I/O error, dev sdf, sector 8217847
[13093.450457] printk: 1517 messages suppressed.
[13093.450463] Buffer I/O error on device sdf1, logical block 8305224
[13093.450469] lost page write due to I/O error on sdf1
[13093.450475] Buffer I/O error on device sdf1, logical block 8305225
[13093.450479] lost page write due to I/O error on sdf1
[13093.450484] Buffer I/O error on device sdf1, logical block 8305226
[13093.450488] lost page write due to I/O error on sdf1
[13093.450492] Buffer I/O error on device sdf1, logical block 8305227
[13093.450496] lost page write due to I/O error on sdf1
[13093.450501] Buffer I/O error on device sdf1, logical block 8305228
[13093.450505] lost page write due to I/O error on sdf1
[1...

Read more...

Revision history for this message
Lito (lito-eordes) wrote :
Revision history for this message
Lito (lito-eordes) wrote :
Revision history for this message
Lito (lito-eordes) wrote :
Revision history for this message
Lito (lito-eordes) wrote :
Revision history for this message
Lito (lito-eordes) wrote :
Revision history for this message
Lito (lito-eordes) wrote :
Revision history for this message
Lito (lito-eordes) wrote :
Revision history for this message
Lito (lito-eordes) wrote :
Revision history for this message
Lito (lito-eordes) wrote :
Revision history for this message
Lito (lito-eordes) wrote :

The same problem in all USB ports (8).

Revision history for this message
Cricket (shahmir-noorani) wrote :

Ditto on the slow copying from HDD to USB. In nautilus, after the file has reached 100% (i.e. 0 seconds left) it still sits there for about 20-30 seconds before it'll actually finish writing the data on the USB drive.

Revision history for this message
Bogdan Gribincea (bogdan-gribincea) wrote :

Same problem under Jaunty, kernel 2.6.28-8.26 using MC, Dolphin or whatever. Starts fast then starts crawling.

Revision history for this message
Andy Davie (etatdavie) wrote :

Same problem in Karmic. The file transfer starts off well, but slows to a crawl...

dmesg | grep usb
[ 124.371817] usb 1-9: reset high speed USB device using ehci_hcd and address 4
[ 130.340032] usb 1-9: reset high speed USB device using ehci_hcd and address 4
[ 218.350034] usb 1-9: reset high speed USB device using ehci_hcd and address 4
[ 368.340032] usb 1-9: reset high speed USB device using ehci_hcd and address 4
[ 368.650033] usb 1-9: reset high speed USB device using ehci_hcd and address 4
[ 388.340026] usb 1-9: reset high speed USB device using ehci_hcd and address 4
[ 430.341309] usb 1-9: reset high speed USB device using ehci_hcd and address 4
[ 462.340038] usb 1-9: reset high speed USB device using ehci_hcd and address 4
[ 572.350050] usb 1-9: reset high speed USB device using ehci_hcd and address 4
[ 676.340024] usb 1-9: reset high speed USB device using ehci_hcd and address 4
[ 738.340024] usb 1-9: reset high speed USB device using ehci_hcd and address 4
[ 808.340025] usb 1-9: reset high speed USB device using ehci_hcd and address 4
[ 863.340021] usb 1-9: reset high speed USB device using ehci_hcd and address 4
[ 881.350058] usb 1-9: reset high speed USB device using ehci_hcd and address 4
[ 884.350056] usb 1-9: reset high speed USB device using ehci_hcd and address 4
[ 1014.340028] usb 1-9: reset high speed USB device using ehci_hcd and address 4
[ 1040.410035] usb 1-9: reset high speed USB device using ehci_hcd and address 4
[ 1044.350036] usb 1-9: reset high speed USB device using ehci_hcd and address 4
[ 1057.340028] usb 1-9: reset high speed USB device using ehci_hcd and address 4
[ 1092.340027] usb 1-9: reset high speed USB device using ehci_hcd and address 4

Revision history for this message
eltopo (eltopo) wrote :

Hi all,
I'm also affected by this bug (or a very similar one).
on the french ubuntu forum there are several people with this issue (http://forum.ubuntu-fr.org/viewtopic.php?id=287690&p=31), they suggested me to come on launchpad and make a post, so here I am.
I'm not very familiar with comand lines, but I have time these days. what can I try to isolate the problem?
thanks in advance for the help.
greetings
eltopo

Revision history for this message
eltopo (eltopo) wrote :

I'm deeply sorry I don't find how to edit my last post.
see attachement for lsusb ; lspci -vvnn ; dmesg ; lshw

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.