Most recent kernel where this bug did not occur: Never worked be it with 2.4 or 2.6 Distribution: Fedora or Mandrake (several versions) Hardware Environment: Abit NF7 motherboard (first release) Nforce2 chipset Software Environment: Problem Description: When using the poweroff command all seems going normally, I hear the heads parking, screens go black and then... nothing happens. The box continues powered. But it is no longer possible to restart the box, be it by using the Soft poweron button, by magic keys or through reset. I have to use the switch on the alim to physically poweroff and then switch it on poweron. If I hit the soft poweroff I have the same symptoms (black screen, no poweroff but impoosible to restart) except that it is im
please verify that you're running the latest BIOS. please re-test with linux-2.6.13.1 or later
I have upgraded to the lastest BIOS and to 2.6.14rc1. Settings in BIOS were: poweroff button shifts to S1, action immediate (alternative shift to S3, delay 4 seconds, BTW when booting the kernel lists S0,S1,S4,S5 as supported but not S3). Results: no changes when using the poweroff command, ie system closes cleanly, heads park but poweroff does not take place and system cannot be awakened be it through the power button or a reset. Need to use the switch need in the alim and physycally power off/on. However, now if I push the power button Linux gets the signal and an orderly shutdown takes place but end result is the same that after a power off command (heads parked, machine powered on and not restartable)
Do you still have the problem with 2.6.14? Could you please attach output of dmesg and acpidump utilities?
Created attachment 6752 [details] Output of acpicdump
Created attachment 6753 [details] Output of dmesg
Created attachment 6754 [details] Output of acpitool I also include the output of acpitool becauseit contains a few interesting bits: AC_Adapter not available (does that mean that this alim cannot be powered off by software?). I also notice that while in my wife's Pentium 4 there is a VBTN and a KBD events who can awake the machine while in my box they are absent. Does that mean that this box cannot be awakened by the Soft Power button. Finally, even if I set up all wakeup events to enabled in the BIOS in Linux I ever get up to disabled. This doesn't happen in the P4.
Do you still have this problem in 2.6.22?
Don't know. I ended changing motherboard. Same chipset, different manufacturer. Works like a charm.