Re: Kernel 2.6.18 and above

[Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>]

On Friday 20 April 2007, Benjamin Herrenschmidt wrote:
> I'm CC'ing linux-ide... looks like the disk which you say is 20GB is
> misdetected by the IDE layer as being 2GB.
> 
> Ok, so we have:
> 
> > >> Probing IDE interface ide0...
> > >> hda: WDC WD800JB-00FMA0, ATA DISK drive
> > >> hdb: IC35L020AVVN07-0, ATA DISK drive
> > >> hda: Enabling Ultra DMA 4
> > >> hdb: Enabling Ultra DMA 4
> > >> ide0 at 0xf139c000-0xf139c007,0xf139c160 on irq 19
> > >> ide1: Found Apple KeyLargo ATA-3 controller, bus ID 0, irq 20
> > >> Probing IDE interface ide1...
> > >> hdc: PIONEER DVD-RW DVR-111D, ATAPI CD/DVD-ROM drive
> > >> hdd: IOMEGA ZIP 250 ATAPI, ATAPI FLOPPY drive
> > >> hdc: Enabling MultiWord DMA 2
> > >> ide1 at 0xf139e000-0xf139e007,0xf139e160 on irq 20
> > >> ide2: Found Apple KeyLargo ATA-3 controller, bus ID 1, irq 21
> > >> Probing IDE interface ide2...
> > >> Probing IDE interface ide2...
> > >> hda: max request size: 128KiB
> > >> hda: 156301488 sectors (80026 MB) w/8192KiB Cache, CHS=65535/16/63, UDMA(66)
> > >> hda: cache flushes supported
> > >>  hda: [mac] hda1 hda2 hda3 hda4 hda5 hda6 hda7 hda8 hda9 hda10 hda11 hda12 hda13
> > >> hdb: max request size: 128KiB
> > >> hdb: 4128768 sectors (2113 MB) w/1863KiB Cache, CHS=4096/16/63, UDMA(66)
> 
> Here the drive claims to be 2GB, not 20GB (or is detected as such).
> 
> > /dev/hdb
> >         #                    type name                length   base    ( 
> > size )  system
> > /dev/hdb1     Apple_partition_map Apple                   63 @ 1       ( 
> > 31.5k)  Partition map
> > /dev/hdb2          Apple_Driver43 Macintosh               56 @ 64      ( 
> > 28.0k)  Driver 4.3
> > /dev/hdb3          Apple_Driver43 Macintosh               56 @ 120     ( 
> > 28.0k)  Driver 4.3
> > /dev/hdb4        Apple_Driver_ATA Macintosh               56 @ 176     ( 
> > 28.0k)  Unknown
> > /dev/hdb5        Apple_Driver_ATA Macintosh               56 @ 232     ( 
> > 28.0k)  Unknown
> > /dev/hdb6          Apple_FWDriver Macintosh              512 @ 288     
> > (256.0k)  Unknown
> > /dev/hdb7      Apple_Driver_IOKit Macintosh              512 @ 800     
> > (256.0k)  Unknown
> > /dev/hdb8           Apple_Patches Patch Partition        512 @ 1312    
> > (256.0k)  Unknown
> > /dev/hdb9              Apple_Free                     262144 @ 1824    
> > (128.0M)  Free space
> > /dev/hdb10              Apple_HFS Apple_HFS_Untitled_2 39924976 @ 
> > 263968  ( 19.0G)  HFS
> > /dev/hdb11             Apple_Free                         16 @ 40188944 
> > (  8.0k)  Free space
> > 
> > Block size=512, Number of Blocks=40188960
> > DeviceType=0x0, DeviceId=0x0
> > Drivers-
> > 1: @ 64 for 23, type=0x1
> > 2: @ 120 for 36, type=0xffff
> > 3: @ 176 for 21, type=0x701
> > 4: @ 232 for 34, type=0xf8ff
> 
> And here you can see that the partition map tries to use about 20GB...
> so there is a discrepancy between what the partition map says and what
> the IDE driver says. I suspect old kernel just ignores the later, while
> newer kernels enforce the limit which causes your problem.
> 
> Bart, I'm a bit at lost as to why the device size would be misdetected,
> do you have an idea ? Might be worth getting a dump of the ID block...
> Mike, can you send us a dump of the binary file /proc/ide/hdb/identify ?

Since this is a 2.6.17->2.6.18 regression narrowing the problem down to
the specific changeset (or at least -git or even -rc) seems like a most
promising way in discovering the source of the issue.

Bart