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

[Christian Borntraeger <borntraeger@de.ibm.com>]

Am Freitag 09 Januar 2009 schrieb Alan Cox:
> On Fri, 9 Jan 2009 13:34:55 +0100
> Christian Borntraeger <borntraeger@de.ibm.com> wrote:
> 
> > Hello,
> > 
> > with todays git I lots of these warnings:
> > 
> > [    1.246807] ata1.00: ATA-7: ST910021AS, 4.06, max UDMA/100
> > [    1.253066] ata1.00: 195371568 sectors, multi 16: LBA48
> > [    1.262561] ata2.00: ATAPI: HL-DT-ST DVDRAM GSA-U10N, 1.05, max UDMA/33
> > [    1.273159] ata1.00: configured for UDMA/100
> > [    1.275279] ata2.00: configured for UDMA/33
> > [    1.285546] scsi 0:0:0:0: Direct-Access     ATA      ST910021AS       4.06 PQ: 0 ANSI: 5
> > [    1.292049] sd 0:0:0:0: Attached scsi generic sg0 type 0
> > [    1.298427] sd 0:0:0:0: [sda] 195371568 512-byte hardware sectors: (100 GB/93.1 GiB)
> > [    1.299631] scsi 1:0:0:0: CD-ROM            HL-DT-ST DVDRAM GSA-U10N  1.05 PQ: 0 ANSI: 5
> > [    1.311059] sd 0:0:0:0: [sda] Write Protect is off
> > [    1.317409] sd 0:0:0:0: [sda] Mode Sense: 00 3a 00 00
> > [    1.317426] sd 0:0:0:0: [sda] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
> > [    1.318378] ------------[ cut here ]------------
> > [    1.318380] WARNING: at drivers/ata/libata-sff.c:1017 ata_sff_hsm_move+0x45e/0x750()
> > [    1.318381] Hardware name: 64575GG
> 
> Can you see if this goes away if you back out all Arjan's async stuff.
> Then if not can you back out my 32bit PIO patch and see if that is the
> trigger.

I did some printk debugging. 
__atapi_pio_bytes is called with bytes=18. That means with 32bit transfers
there is a 2 byte trailing transfer, which triggers this WARN_ON.

Christian