On Wed, 2012-02-01 at 11:19 -0500, Jean Delvare wrote: > On Wed, 1 Feb 2012 08:06:31 -0800, Guenter Roeck wrote: > > Another question: Several drivers read vid only once. That doesn't seem > > to be correct; I thought the whole point of having vid is to have > > dynamic voltage to the CPU. Should this be changed as well ? > > Congratulations Guenter, you just volunteered to convert all LPC-based > hwmon drivers to the MFD framework! :D > > Seriously, there can be two reasons why VID is read only once: > * Old driver, predating frequency-scalable CPUs, and having never been > updated. Not sure if there are such drivers remaining. > * LPC-based device with VID read from configuration space. Without > proper locking this is dangerous, so we only do it once at driver > load time. The only proper fix is to migrate the driver(s) to MFD > first. > I think I'll rather leave that alone for now ;). Guenter _______________________________________________ lm-sensors mailing list lm-sensors@xxxxxxxxxxxxxx http://lists.lm-sensors.org/mailman/listinfo/lm-sensors