Hello,
New driver for SiS IDE chipsets with support for ATA16 to ATA100
chip generations.
I added support for ATAT16 and ATA33 chipsets to my previous patches and
rewrote large parts of the original driver.
The diff is 33345 bytes long (longer than the whole sis5513.c file in fact)
so I put it on a web server (thanks go to my employer for the web hosting):
http://inet6.dyn.dhs.org/sponsoring/sis5513/sis.patch.20020123_1
On the following page you'll find more information, links to current
sis5513.c code and previous patches:
http://inet6.dyn.dhs.org/sponsoring/sis5513/index.html
As always, success and failure reports are welcomed.
Comments on coding style are too.
LB.
Well, success :-)
Stock 2.4.17 doesn't hang any more :-)
One question though, hdparm -i/I says my drive is udma2 (UDMA/33 isnt it?)
capable. Why don't I get it on boot? (or is my chipset UDMA/33 capable?).
hdparm -X66 /dev/hda hangs the machine completely. Same with 65 and 64.
The debug messages that appear when I do so say:
sis5513_tune_chipset start, changed registers: none
sis5513_tune_chipset ,drive 0, speed 66
sis5513_tune_chipset end, changed registers: none
Attached is the boot time debug info.
Thanks,
-K.
On Fri, Jan 25, 2002 at 08:02:43PM +0200, Liakakis Kostas wrote:
>
> Well, success :-)
> Stock 2.4.17 doesn't hang any more :-)
>
Cool.
> One question though, hdparm -i/I says my drive is udma2 (UDMA/33 isnt it?)
> capable. Why don't I get it on boot? (or is my chipset UDMA/33 capable?).
Don't think so. IIRC (don't have the specs right now at hand) SiS5513 isn't UDMA capable.
> hdparm -X66 /dev/hda hangs the machine completely. Same with 65 and 64.
If I'm correct with UDMA above, this is expected behavior...
I don't know the IDE framework and hdparm enough yet to know if I can trap hdparm
actions at the chip driver level... it may speak (and I believe it to do so) to the drives directly.
> The debug messages that appear when I do so say:
>
> sis5513_tune_chipset start, changed registers: none
> sis5513_tune_chipset ,drive 0, speed 66
> sis5513_tune_chipset end, changed registers: none
>
> Attached is the boot time debug info.
Many thanks. My database of BIOS init behaviour will grow, this is *good*.
I'm mostly away (Palm with GSM is kind of... slow) from Net access this week-end and I'll be away too from tuesday
on next week. So don't be surprised if I don't answer bug reports or
enhancement requests in the following days, be patient and don't hesitate
to repost your message if you don't have any answer February the 6.
LB.