Re: [PATCH-RFC] Promise TX4 implement hw-bug workaround

[Alexander Sabourenkov <screwdriver@lxnt.info>]

Alexander Sabourenkov wrote:
> Alan Cox wrote:
>>> I can't think of a way to avoid second pass over scatterlist without
>>> duplicating code (ata_qc_prep() and ata_fill_sg() from libata-core.c).
>> This appears to be incomplete:
>>
> 
> [...]
> 
>> What guarantees you have enough PRD entries to do this without changing
>> the limit in the structures ?
>>
>> Otherwise looks good
> 
> PRD entries count is 256
> include/linux/ata.h:
> 	ATA_MAX_PRD		= 256,
> 	ATA_PRD_TBL_SZ          = (ATA_MAX_PRD * ATA_PRD_SZ),
> 
> drivers/ata/libata-core.c:
>  ap->prd = dmam_alloc_coherent(dev, ATA_PRD_TBL_SZ, &ap->prd_dma,
> 
> sata_promise Scsi_Host declares support for half of that:
> 
> include/linux/libata.h:
> LIBATA_MAX_PRD		= ATA_MAX_PRD / 2,
> 
> drivers/ata/sata_promise.c
>     .sg_tablesize           = LIBATA_MAX_PRD,
> 
> 
> PS: Vendor code has this limit at 32.
> 

That's an interesting question of itself. I don't know what limits PRD
count, but if it's hardware, then the driver should somehow make sure
that it gets no more than hw can handle minus one for this errata.

Right now driver declares that any hardware it supports can handle 128
PRD entries. If this is not true for any possibly existing specimen,
we're welcoming trouble.

-- 

./lxnt