Re: ohci1394 unhandled interrupts bug in 2.6.14-rc2

[Stefan Richter <stefanr@s5r6.in-berlin.de>]

rob wrote:
> On Mon, 17 Oct 2005 02:42:19 -0700, Andrew Morton wrote
>>Stefan Richter <stefanr@s5r6.in-berlin.de> wrote:
>>>Earlier forms of the patch do DMI matching:
>>>http://marc.theaimsgroup.com/?l=linux1394-devel&m=110790513206094
>>>http://www.janerob.com/rob/ts5100/tosh-1394.patch
>>>[short-circuited by if (1) at the second URL]
>>
>>Rob, can you finish that patch off and send it?
> 
> Sorry, I was advised that this should be correctly handled as a pci-quirk
> (Jody McIntyre <scjody@modernduck.com>),

Since Jesse found that we really need to read & write back the 
PCI_CACHE_LINE_SIZE, I gather the workaround has to stay in ohci1394 
(and should be triggered by dmi_check_system), doesn't it?

> and subsequently my laptop
> motherboard died so I have no way of taking it further.

We should be able to finish it.

> The responses I got
> indicated that the code works as is for the followiung laptops
> 
>>System Vendor: TOSHIBA
>>Product Name: S5100-501
>>Version: PS510E-00NV7-EN
> 
> System Vendor: TOSHIBA
> Product Name: S5200-801
> Version: PS520E-31P1D-GR
> 
> Manufacturer: TOSHIBA
> Product Name: Satellite 5200
> Version: PS520C-31P0EP
> 
> Manufacturer: TOSHIBA
> Product Name: Satellite 5205
> Version: PS522U-XK00YV
> 
> Manufacturer: TOSHIBA
> Product Name: S5100-603
> Version: PS511E-05328-GR
> 
> toshiba satellite 5005-S504
> 
> Toshiba Satellite 5105-s607

Thanks a lot for the survey. Do they all _need_ the patch, or do some of 
them need it and the others are just not harmed by the patch?

Does anybody know a DMI_PRODUCT_NAME of a Libretto L1? Something like 
PAL1060TNMM or PAL1060TNCM?
-- 
Stefan Richter
-=====-=-=-= =-=- =---=
http://arcgraph.de/sr/