Re: [PATCH] libata: always use polling IDENTIFY

[Jeff Garzik <jgarzik@pobox.com>]

Alan wrote:
> On Sun, 3 Dec 2006 19:30:32 +0900
> Tejun Heo <htejun@gmail.com> wrote:
> 
>> libata switched to IRQ-driven IDENTIFY when IRQ-driven PIO was
>> introduced.  This has caused a lot of problems including device
>> misdetection and phantom device.
>>
>> ATA_FLAG_DETECT_POLLING was added recently to selectively use polling
>> IDENTIFY on problemetic drivers but many controllers and devices are
>> affected by this problem and trying to adding ATA_FLAG_DETECT_POLLING
>> for each such case is diffcult and not very rewarding.
>>
>> This patch makes libata always use polling IDENTIFY.  This is
>> consistent with libata's original behavior and drivers/ide's behavior.
>>
>> Signed-off-by: Tejun Heo <htejun@gmail.com>
> 
> Acked-by: Alan Cox <alan@redhat.com>
> 
> One possibility however would be to invert the flag we have now so we can
> try turning -on- IRQ polling on some devices in future and maybe figure
> it out that way around ?

Agreed, though it is low priority.

You'll note that Tejun's change only applies to the [large] set of 
drivers using libata-sff, rather than all drivers.  Modern, unbridged 
SATA controllers (ahci, sata_sil24) continue to do their own interrupts 
based on received packets (FIS's) through the use of the higher level 
libata hooks.

	Jeff