Hi Alek, On Wed, 21 May 2008 10:21:05 +0800, Alek Du wrote: > On Tue, 20 May 2008 17:31:40 +0800 > "Jean Delvare" <[EMAIL PROTECTED]> wrote: > > > ETIMEOUT is indeed not defined, but ETIMEDOUT (what I wrote) is. > Ok, my bad. > > > > Well, the datasheet clearly says otherwise. How did you test bit 31? > > Your code uses pci_read_config_word() at the moment so you can't read > > bit 31, you'd need to use pci_read_config_dword() instead. If Bit 31 > Yes, my bad again. fixed now. > > Again, here is the updated patch, thanks for your kind review: > > This patch adds Intel SCH chipsets (AF82US15W, AF82US15L, AF82UL11L) i2c bus > support.
OK, this is the right one this time. I've added your patch to my i2c tree as: http://khali.linux-fr.org/devel/linux-2.6/jdelvare-i2c/i2c-add-intel-sch-i2c-smbus-support.patch It will go in linux-next and is scheduled for merge in kernel 2.6.27-rc1. I've also updated sensors-detect and http://www.lm-sensors.org/wiki/Devices . Any further patch to this driver should be an incremental one based on what I have in my tree. Things that could be improved include: * Allocate the device data dynamically to get rid of the assumption that only one SMBus interface is present. * Rework the timeout mechanism so that it doesn't depend on HZ. * Documentation. Most SMBus host controller drivers have a documentation file under Documentation/i2c/busses, it would be nice to have one for this new i2c-isch driver. Thanks, -- Jean Delvare _______________________________________________ i2c mailing list [email protected] http://lists.lm-sensors.org/mailman/listinfo/i2c
