Re: IDE Patch SIS ATA100

[Lionel Bouton]

Jussi Laako wrote:

> Andre Hedrick wrote:
> 
>>Thanks for the feedback, but lkml needs it or it will not be adopted.
>>I know the driver is stable and effectively perfect in operations.
>>So I do not understand the total ignore I receive about it.
>>
> 
> Just to avoid overperfectness... ;))
> 
> Has anyone succeeded in fixing the sis5513 driver to work with ATA100 chips?
> I get heavy disk corruption with SiS730S chipset mobo (ASUS A7S-VM).
> 
> 
> Best regards,
> 
> 	- Jussi Laako
> 
> 

I'm on it. Had some intermittent successes and thought to have a correct 
patch until today (one 2002 week with ATA100)...
But I had a bugreport today and same dma problems on my machine after a 
BIOS flashing and a new drive in the system.
Currently in heavy debugging.

PB: chip seems to init itself correctly on extended periods of time on 
my config! Makes testing rather difficult :-(

For example, with the *exact* same code this evening I had:
- a system freeze just after /sbin/init load,
- a crash after ext3 fs errors before init,
- a somewhat working boot (some weird library errors caused "ip" to not 
work),
- a fully functionnal system (no error reported, some file copies, 
reboot with ide=nodma, e2fsck -f -> no error).

If you have a system you can test on without fear of breaking things 
I'll send you patches as soon as they'll work again on my config.
Better if you want to proof-read the code :
http://gyver.homeip.net/sis5513/index.html

LB.