Hi Jean Thank you for clarification. But is it possible that when I/O ports is controlled by motherboard, the kernels can still "read" the information from the I/O ports rather than read and write data to I/O ports? I think in this case, at least, I can monitor my fan speed without really controlling the ports, and I think it's still safe. Thanks. Best Wishes, Dong-Bang Tsai  èæé Ph.D. Student in Applied Physics at Stanford University åääååæçççæååç ----------------------------------- Web : http://www.dbtsai.com Phone : +1-650-383-8392 (USA)       +886-910-008-392 (Taiwan) On Thu, Apr 7, 2011 at 12:29 AM, Jean Delvare <khali@xxxxxxxxxxxx> wrote: > On Wed, 6 Apr 2011 13:47:45 -0700, D.B. Tsai wrote: >> Hi >> >> Sure, I can take the risk of the disabling acpi resource conflict >> check, and I'm willing to test your code. Please let send me the code >> when you finish it. By the way, I'm using Ubuntu 11.04 with kernel >> 2.6.38-8-server, and wondering to know that when will the mainline >> kernel fix this conflict issue? > > Never, as this isn't a bug per se. Your ACPI BIOS wants to control the > I/O ports, the kernels wants them too, and that's about it. If your > BIOS requests the I/O ports but doesn't use them, then it's a bug in > your BIOS. > > Some vendors (Asus) request the I/O ports for a reason, and they > implement a proprietary interface on top of the hardware monitoring > chip. In this case it is possible to write an ACPI driver as a > replacement for the native driver. Unfortunately, most other vendors > which request the I/O ports don't implement anything like this. > > -- > Jean Delvare > _______________________________________________ lm-sensors mailing list lm-sensors@xxxxxxxxxxxxxx http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
