2002-06-17 13:31:58

by Lionel Bouton

[permalink] [raw]
Subject: Re: 2.5.20 hardlock w/ hdparm

Zwane Mwaikambo wrote:

>Hi Lionel, Martin,
>2.5.20, hdparm + IDE deadlocks on my testbox
>
>

I don't follow 2.5 dev (yet). I merely follow Andre's work on 2.4 and
code a new chipset capabilities detection code in order to support newer
chipsets.
Is the v0.13 driver driver already forward ported to 2.5 by somebody ?
If there's a need (some 2.5 developpers needing a more uptodate driver
and uncomfortable with forward porting IDE chipset drivers), I'll do it...

>kernel:Linux version 2.5.20+prempt ([email protected]) (gcc version
>2.96 20000731 (Red Hat Linux 7.3 2.96-110)) #24 SMP Wed Jun 5 21:48:07 SAST 2002
>
>ata subsys:
>ATA/ATAPI device driver v7.0.0
>ATA: PCI bus speed 33.3MHz
>ATA: Silicon Integrated Systems [SiS] 5513 [IDE], PCI slot 00:00.1
>PCI: No IRQ known for interrupt pin A of device 00:00.1. Please try using pci=biosirq.
>ATA: chipset rev.: 208
>ATA: non-legacy mode: IRQ probe delayed
>SiS620
>

Unless the SiS620 is not compatable with the 630 IDE support spec or
affected by some bugs I corrected for PIO mode timings (unlikely as they
were unnoticed for a quiet long time) since sis5513.c v0.11 this should
not be an IDE chipset problem.

>[...] btw Martin you seem to like pain so get ready for when i whip out the old
>Quantum mavericks, 486 (SiS) and Opti621 card ;)
>
>

Hum, 486 SiS chipsets might bring pain to me also...
I've received several bugreports for old SiS IDE chipset (ie pre ATA66)
that I couldn't solve without disabling the SiS driver or passing
"ide=nodma". I've triple-checked the specs and couldn't see the problem.

>Thanks,
> Zwane Mwaikambo
>
>
LB


2002-06-17 15:32:55

by Zwane Mwaikambo

[permalink] [raw]
Subject: Re: 2.5.20 hardlock w/ hdparm

On Mon, 17 Jun 2002, Lionel Bouton wrote:

> Unless the SiS620 is not compatable with the 630 IDE support spec or
> affected by some bugs I corrected for PIO mode timings (unlikely as they
> were unnoticed for a quiet long time) since sis5513.c v0.11 this should
> not be an IDE chipset problem.

Nope definitely not, i've had a chance to dig a little deeper.

> Hum, 486 SiS chipsets might bring pain to me also...
> I've received several bugreports for old SiS IDE chipset (ie pre ATA66)
> that I couldn't solve without disabling the SiS driver or passing
> "ide=nodma". I've triple-checked the specs and couldn't see the problem.

That one is just for fun ;) ditto for the 386 w/ IDE box i just got =)

Cheers,
Zwane
--
http://function.linuxpower.ca