https://bugzilla.kernel.org/show_bug.cgi?id=205657
--- Comment #11 from Zhang Rui (rui.zh...@intel.com) --- (In reply to Pask from comment #9) > - "so this seems like a kernel regression?" > > Somehow or, at least I tried to understand where this problem came from so I > tried to compile an earlier kernel lts 3.16.78, but I had to change to > 3.18.99 because some software had memfd functionality requirements. > So the problem does not exist in 3.18.99. > > - "following is part of the debug trace in your boot dmesg, let's make clear > if the trace log is related with the shutdown issue or not first." > > How do you proceed? boot into any good/bad kernel, check if the boot trace log exists or not, shutdown and check if the shutdown problem exists or not. > > > - "So can you please confirm if the debug trace always happens in bad > kernel, and does not happen in good kernel? Say, back to 3.26.78, which > shutdown works, does the log still exists? and does the shutdown > problem/trace log still happen in the latest upstream kernel?" > > I confirm that the problem only occurs when the kernel lts 4.19.xx and even > earlier as well as the kernel 5.x. > > This problem does not occur with 3.x kernels. So the boot trace log is probably related with the shutdown issue, right? Then please rebuild your 4.x and 5.x kernel with CONFIG_SSB cleared, and check a) if the boot trace log still exists b) if the shutdown issue still exists > > The debug message: > > INFO: task shutdown:1 blocked for more than 120 seconds. > Tainted: G W 4.19.81-1.0-lts #1 > "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. > Call Trace: > __schedule+0x264/0x8e0 > schedule+0x2e/0x80 > schedule_timeout+0x2a5/0x420 > ? acpi_ut_delete_internal_object_list+0x2a/0x2e > wait_for_common+0x9e/0x130 > ? wake_up_q+0x60/0x60 > wait_for_completion+0x17/0x20 > flush_workqueue+0xf9/0x3d0 > acpi_os_wait_events_complete+0x2b/0x30 > acpi_power_off_prepare+Ox1c/0x20 > kernel_power_off+040/0x70 > sys_reboot+0x104/0x200 > ? do_iter_write+0xc2/0x190 > ? vfs_wrtiev+0x7d/0xe0 > do_fast_syscall_32+0x87/0x1f0 > entry_SYSENTER_32+0x6b/0xbe > EIP: 0xb7ee2d19 > Code: Bad RIP value. > EAX: ffffffda EBX: fee1dead ECX: 28121969 EDX: 4321fedc > ESI: 00000000 EDI: bfee8c1c EBP: bfee8d18 ESP: bfee8b78 > DS: 007b ES: 007b FS: 0000 GS: 0033 SS: 007b EFLAGS: 00000292 > > this message only occurs during shutdown. It repeats itself 2 3 or even 4 > times before the laptop turns off. This problem does not arise if I request > a restart only the shutdown. I noticed something else, it's the power > button. In the power supply parameters I selected "on demand" for the power > button, what do you mean by "power supply parameters"? > but when I press it the time for the dialog window to appear is also > very long. While under kernel 3.18.99 this is done normally. > I reconfirm that this problem does not occur with the kernel 3.18.99 please attach the output of "cat /proc/interrupts" and "grep . /sys/firmware/acpi/interrupts/*" both before and after pressing the power button, in both good and bad kernels. -- You are receiving this mail because: You are watching the assignee of the bug. _______________________________________________ acpi-bugzilla mailing list acpi-bugzilla@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/acpi-bugzilla
