Hi Jean: * Jean Delvare <khali at linux-fr.org> [2008-03-09 13:34:28 +0100]: > Only request I/O ports 0x295-0x296 instead of the full I/O address > range. This solves a conflict with PNP resources on a few motherboards. > > Also request the I/O ports in two parts (4 low ports, 4 high ports) > during device detection, otherwise the PNP resource makes the request > (and thus the detection) fail. > > This fixes lm-sensors ticket #2306: > http://www.lm-sensors.org/ticket/2306 > > Signed-off-by: Jean Delvare <khali at linux-fr.org> > --- > Mark, this fixes a regression so this should go to Linus before 2.6.25 > final. This should also be backported to 2.6.24-stable (I'll take care.) > > Note: I do agree that my fix is not universal and it is possible that > some boards out there have a different PNP setup not covered by this > fix. However it's hard to predict what the various BIOS do, and I also > do not expect that many boards old enough to use a W83781D to declare > it as a PNP device. So, my current position is to wait for further > reports before updating the code again if needed. > > drivers/hwmon/w83781d.c | 21 ++++++++++++++++----- > 1 file changed, 16 insertions(+), 5 deletions(-) There is a special corner of hell reserved for authors of crappy BIOS. I've applied this to my release tree, and I'll send it to Linus later today (as a straight patch instead of pull request, if necessary). Thanks & regards, -- Mark M. Hoffman mhoffman at lightlink.com
