Re: [bisected] Re: todays git: WARNING: at drivers/ata/libata-sff.c:1017 ata_sff_hsm_move+0x45e/0x750()

[Sergei Shtylyov <sshtylyov@ru.mvista.com>]

Alan Cox wrote:

>>* Ingo Molnar <mingo@elte.hu> wrote:

>>>build its random kernel and reboot into it (800 MHz P3), so i started a 
>>>full automated bisection instead driven from a faster box.

>>and the winner is:
>>| 871af1210f13966ab911ed2166e4ab2ce775b99d is first bad commit
>>| commit 871af1210f13966ab911ed2166e4ab2ce775b99d
>>| Author: Alan Cox <alan@redhat.com>
>>| Date:   Mon Jan 5 14:16:39 2009 +0000
>>|
>>|     libata: Add 32bit PIO support
>>
>>And the revert below does the trick.

> Unfortunately it only hides the bug - that code doesn't change the sg
> lists being passed, it just catches out anyone who pasts a non divisble
> by four malformed sg list.

    All the S/G counts printed out were divisible by 4 (36 for INQUIRY and 96 
for REQUSET SENSE). It's the *actual* byte count for the REQUEST SENSE that's 
no divisible. The SCSI/ATAPI devices are free to sent less data than requested 
on non block transfer commands.

MBR, Sergei