On Mon, 7 Apr 2008, Marcelo Tosatti wrote: > I think the logic to wakeup tasks in HLT is racy. Nothing prevents > a timer event from being lost if it fires in between guest exit and > vcpu_block(). > > Please try the patch below. >
Hi Marcelo, the problem persists even with patch. [ 1905.899112] [<ffffffff8020a55b>] ? default_idle+0x3b/0x70 [ 1905.899112] [<ffffffff8020a520>] ? default_idle+0x0/0x70 [ 1905.899112] [<ffffffff8020a60e>] ? cpu_idle+0x7e/0xe0 [ 1905.899112] [<ffffffff80211630>] ? pda_init+0x30/0xb0 As I already said, if I enable ACPI, system does not boot at all (even with init=/bin/bash) and ends with mentioned trace immediately. If I disable ACPI, it hangs on "Starting udev", with init=/bin/bash it at least boots and I can use shell. At least for some time. Funny thing is, that If I then use poweroff, the system shuts down, ends with "System halted." message, but everything still seems to be working, shell, etc... ------------------------------------------------------------------------- This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Register now and save $200. Hurry, offer ends at 11:59 p.m., Monday, April 7! Use priority code J8TLD2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone _______________________________________________ kvm-devel mailing list kvm-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/kvm-devel
