Bug#456203: dumb reportbug

2007-12-18 Thread maximilian attems
On Thu, 13 Dec 2007, Nagy Elemer Karoly wrote:

>
> Dear Debian,
>
> So, reportbug forgot to inlcude the majority of the report:
>
> USB pendrive fails when syncing file systems (works OK with big sequential 
> copies) with "reset high speed USB device" and "device descriptor read/64, 
> error -110" and "sd 0:0:0:0: scsi: Device offlined - not ready after error 
> recovery" and "scsi 0:0:0:0: rejecting I/O to dead device"
> Laptop is HP NX 6125, Pendrive is A-DATA PD7:
>  00:13.0 0c03: 1002:4374 (prog-if 10 [OHCI])
>  00:13.1 0c03: 1002:4375 (prog-if 10 [OHCI])
>  00:13.2 0c03: 1002:4373 (prog-if 20 [EHCI])
>  Bus 003 Device 013: ID 090c:1000 Feiya Technology Corp. Memory Bar
>  sd 4:0:0:0: [sda] 4046848 512-byte hardware sectors (2072 MB)
>  sd 4:0:0:0: [sda] Assuming drive cache: write through
>
> A workaround is to issue:
>  echo "64" >/sys/block/sda/device/max_sectors
> but that slows down copying to 2.6.18 level (which had no such issue).
>

can you please try 2.6.23 linux images from unstable they install just
fine in testing.

thanks for feedback.

-- 
maks



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]



Bug#456203: dumb reportbug

2007-12-13 Thread Nagy Elemer Karoly


Dear Debian,

So, reportbug forgot to inlcude the majority of the report:

USB pendrive fails when syncing file systems (works OK with big sequential 
copies) with "reset high speed USB device" and "device descriptor read/64, 
error -110" and "sd 0:0:0:0: scsi: Device offlined - not ready after error 
recovery" and "scsi 0:0:0:0: rejecting I/O to dead device"

Laptop is HP NX 6125, Pendrive is A-DATA PD7:
 00:13.0 0c03: 1002:4374 (prog-if 10 [OHCI])
 00:13.1 0c03: 1002:4375 (prog-if 10 [OHCI])
 00:13.2 0c03: 1002:4373 (prog-if 20 [EHCI])
 Bus 003 Device 013: ID 090c:1000 Feiya Technology Corp. Memory Bar
 sd 4:0:0:0: [sda] 4046848 512-byte hardware sectors (2072 MB)
 sd 4:0:0:0: [sda] Assuming drive cache: write through

A workaround is to issue:
 echo "64" >/sys/block/sda/device/max_sectors
but that slows down copying to 2.6.18 level (which had no such issue).

Without the workaround, I borked my FAT (fat_get_cluster: invalid 
cluster chain (i_pos 0), FAT: Filesystem panic (dev sda1)) and ended up 
having files filled with zeroes. Severe data loss, could be a problem when 
kernel gets into Etch.


Seems to affect only high-speed USB HUB (only EHCI?) +Pendrive.

Google is full of "reset high speed USB device" with many different USB 
stuff - from external HDDs to network cards.


Elemer




--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]