Jaunty config changes: Build in CDROM_PKTCDVD and CHR_DEV_SG

Tim Gardner tcanonical at tpi.com
Tue Jan 6 14:56:01 UTC 2009


Andy Whitcroft wrote:
> On Fri, Dec 26, 2008 at 07:01:53AM -0700, Tim Gardner wrote:
>> Andy Whitcroft wrote:
>>> On Mon, Dec 22, 2008 at 10:40:32AM -0700, Tim Gardner wrote:
>>>> Isn't CHR_DEV_SG just for eject and other non-blockdev controls?
>>> I queried it less for what it does specifically than when it seems to
>>> appear in my logs.  Note the scsi generic sg0 is attached well before sd
>>> is attached to the device:
>>>
>>>     [    4.462868] ata2.00: configured for UDMA/100
>>>     [    4.784113] ata5: SATA link down (SStatus 0 SControl 300)
>>>     [    5.104153] ata6: SATA link down (SStatus 0 SControl 300)
>>>     [    5.104279] scsi 1:0:0:0: Direct-Access     ATA      WDC WD2500BEVT-7 11.0 PQ: 0 ANSI: 5
>>>     [    5.104379] scsi 1:0:0:0: Attached scsi generic sg0 type 0
>>>     [    5.105477] scsi 2:0:0:0: CD-ROM            Optiarc  DVD+-RW AD-7640S HD14 PQ: 0 ANSI: 5
>>>     [    5.105548] scsi 2:0:0:0: Attached scsi generic sg1 type 5
>>> [...]
>>>     [    5.717378] sd 1:0:0:0: [sda] 488397168 512-byte hardware sectors (250059 MB)
>>>     [    5.717392] sd 1:0:0:0: [sda] Write Protect is off
>>>     [    5.717394] sd 1:0:0:0: [sda] Mode Sense: 00 3a 00 00
>>>     [    5.717416] sd 1:0:0:0: [sda] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
>>>
>>> Now this may occur because we have sg in the initrd and just because it
>>> is loaded.  Perhaps if we boot without it then things continue without
>>> forcing a load.
>>>
>>> -apw
>>>
>> Perhaps I'm being dense, but I don't see the problem. What is occurring?
> 
> I was saying that when my sata drives are seen an sg is automatically
> attached to them even before we load up the partition table etc.  Now
> that may occur just because sg is already loaded.  The question was is
> this sg attach going to occur regardless and trigger sg to be loaded.
> Making sg a boot required module in effect.
> 
> -apw
> 

Well, sg is enabled by CONFIG_CHR_DEV_SG=y, so isn't it kind of a moot
point? Its already built in.

-- 
Tim Gardner tim.gardner at canonical.com




More information about the kernel-team mailing list