On Tue, Apr 26, 2005 at 02:45:06AM -0500, Dmitry Torokhov wrote: > Hi, > > On Monday 25 April 2005 00:36, Yani Ioannou wrote: > > Hi all, > > > > I'm sending this driver sysfs callback patch to the lm-sensors list to > > get more of the sensor driver maintainer's comments on the patch > > (specifically the type to be associated with each sysfs entry) before > > submitting to LKML. Jean and Greg have already given me invaluable > > feedback and are supportive of the general idea of the patch (the > > problem of which is epitomized by the kludge the 2.6 bmcsensors driver > > is at the moment - see previous discussions in the list archive). > > > > Included is a patch against drivers/base/core.c and > > include/linux/driver.h that adds a void * to the device attribute > > struct and passes it back to the two sysfs callbacks show/store. > > If void is added directly to the attribute structure that means that same > attribute can not be shared between several instances of the same device > -> unexpected. So far all attributes could be statically created. One reason I will not accept such a patch, without seeing how it will be used :) thanks, greg k-h
