cvs commit: src/sys/dev/ata ata-all.c ata-all.h ata-card.c ata-lowlevel.c ata-pci.c ata-pci.h ata-queue.c

[Søren Schmidt]

Roman Kurakin wrote:
> Søren Schmidt wrote:
> 
>>> Includes fixes to avoid checking the ATA_FLAG_54_58 bit, fixing LBA
>>> detection:
>>> Submitted by:    rk
>>
>>
>> not true, rk submitted the problem not the code.
> 
> 
> Probably you are talking about me "rik" :-)
> I realy doesn't submit any code. I just tell my suggestions that 
> checking this bit
> is bad idea according to ata spec, and according last ata specs it 
> shouldn't be
> used at all since it obsoleted. So I didn't submit any fixes about it. 
> Only idea.
> So this may be "Asked by" or "Pointed by" :-)

Right, it may break ATA on old devices though, but I guess none of these 
drives are in use any more.
The real problem here is that I cannot just use the latest ATA specs (it 
would make my life *alot* simpler though :) ), I need to make things 
work that are compatible to at least back to ATA rev3 so just relying on 
things be obsoleted in the latest spec revision doesn't cut it..

> Code I submitted was to make ata driver more verbose about CHS mode.
> They still are not there and should be revised to make them possible to
> apply to current code.

I'll think about it, but since CHS mode is almost no longer used anyways
and the info can pretty easily be calculated from the LBA's printed..

> By the way, may be it would be a nice idea to warn users that they have
> ATA_FLAG_54_58 bit off, and device will work in LBA mode not
> in CHS as was before?

Users wont notice at all, from users perspective the disk is just a 
bunch of sectors, the interface method used to talk to the disk is 
nowhere to be seen...

-- 
-Søren