* Alexander Malysh <amalysh at web.de> [2003-08-01 23:21:59 +0200]: > Hi Jean, Mark, > > i do not know what is to do here :( > i2c-sis630 is not a pci driver and should not claim any pci devices. > i2c-sis630 use only io ports and pci device look up used only in order to > determine if it a right sis chip et all. I want to revert pci changes and > commit it, if nobody has any objections. On one hand it should be a PCI driver. After all it does read (and potentially write) in the PCI config space of the bridge device - even if that's only at device init. But then again, if there are other functions hidden there that some other driver might use, we're back to the same problem. > Mark, what do you think about it? Why should be such detect method not > accepted in 2.4.X ? Well maybe it's not forbidden, but discouraged. Again, I'm not sure what is the right thing. If we don't hear fresh advice in a few days I'll ask on LKML. > On Friday 01 August 2003 23:01, Jean Delvare wrote: > > Mark, Alexander, > > > > Is the i2c-sis630 driver considered fixed or not? Shall I close the > > tickets 1350 and 1351? > > > > (My reading of the thread makes me think there still is a discussion > > about what should be done, but I don't know what I am supposed to do > > with the tickets.) The code in CVS should work just fine. I would say the tickets can be closed because (unless we're very sloppy) that bug won't happen again. What we discuss above is related but more subtle. Regards, -- Mark M. Hoffman mhoffman at lightlink.com
