We do 48+ hour stress testing each week on various machines with the latest upstream kernel. Our two Dell Precision 5510 machines have this issue approximately 5 times out of every 2000 tests (0.4% fail rate). It only occurs in S3 suspend/resume. Here is the dmesg log snippet where the error occurs: [95583.861796] PM: resume of devices complete after 1022.883 msecs [95583.864827] acpi LNXPOWER:1b: Turning OFF [95583.877239] ACPI BIOS Error (bug): Could not resolve [\_SB.PCI0.LPCB.HEC.ECAV], AE_NOT_FOUND (20181213/psargs-330) [95583.877270] [95583.877273] Local0: 00000000677be56b <Obj> Integer 0000000000000001 [95583.877286] Local1: 000000002e7bf245 <Obj> Integer 0000000000000001 [95583.877294] Local2: 000000006f034898 <Obj> Integer 0000000000000001 [95583.877303] Local3: 00000000fdde280a <Obj> Integer 0000000000000001 [95583.877311] Local4: 000000001c00f842 <Obj> Integer 0000000000000000 [95583.877319] Local5: 00000000f8575737 <Obj> Integer 0000000000000000 [95583.877327] Local6: 000000004a7ca0c2 <Obj> Integer 0000000000000000 [95583.877337] No Arguments are initialized for method [FNCL] [95583.877342] ACPI Error: Method parse/execution failed \_TZ.FNCL, AE_NOT_FOUND (20181213/psparse-531) [95583.877363] ACPI Error: Method parse/execution failed \_TZ.FN04._OFF, AE_NOT_FOUND (20181213/psparse-531) [95583.877485] acpi LNXPOWER:1a: Turning OFF [95583.889232] ACPI BIOS Error (bug): Could not resolve [\_SB.PCI0.LPCB.HEC.ECAV], AE_NOT_FOUND (20181213/psargs-330) [95583.889263] [95583.889266] Local0: 00000000f8575737 <Obj> Integer 0000000000000001 [95583.889279] Local1: 0000000093d19097 <Obj> Integer 0000000000000001 [95583.889287] Local2: 000000006f034898 <Obj> Integer 0000000000000001 [95583.889295] Local3: 00000000fdde280a <Obj> Integer 0000000000000000 [95583.889303] Local4: 00000000677be56b <Obj> Integer 0000000000000000 [95583.889311] Local5: 000000004bc152cc <Obj> Integer 0000000000000000 [95583.889319] Local6: 000000004a7ca0c2 <Obj> Integer 0000000000000000 [95583.889330] No Arguments are initialized for method [FNCL] [95583.889335] ACPI Error: Method parse/execution failed \_TZ.FNCL, AE_NOT_FOUND (20181213/psparse-531) [95583.889356] ACPI Error: Method parse/execution failed \_TZ.FN03._OFF, AE_NOT_FOUND (20181213/psparse-531) [95583.889476] acpi LNXPOWER:19: Turning OFF [95583.901232] ACPI BIOS Error (bug): Could not resolve [\_SB.PCI0.LPCB.HEC.ECAV], AE_NOT_FOUND (20181213/psargs-330) [95583.901263] [95583.901265] Local0: 000000004bc152cc <Obj> Integer 0000000000000001 [95583.901278] Local1: 00000000ad5243df <Obj> Integer 0000000000000001 [95583.901287] Local2: 000000006f034898 <Obj> Integer 0000000000000000 [95583.901295] Local3: 00000000fdde280a <Obj> Integer 0000000000000000 [95583.901303] Local4: 00000000f8575737 <Obj> Integer 0000000000000000 [95583.901311] Local5: 000000004c0e10cd <Obj> Integer 0000000000000000 [95583.901319] Local6: 000000004a7ca0c2 <Obj> Integer 0000000000000000 [95583.901329] No Arguments are initialized for method [FNCL] [95583.901335] ACPI Error: Method parse/execution failed \_TZ.FNCL, AE_NOT_FOUND (20181213/psparse-531) [95583.901356] ACPI Error: Method parse/execution failed \_TZ.FN02._OFF, AE_NOT_FOUND (20181213/psparse-531) [95583.901476] acpi LNXPOWER:18: Turning OFF [95583.913281] ACPI BIOS Error (bug): Could not resolve [\_SB.PCI0.LPCB.HEC.ECAV], AE_NOT_FOUND (20181213/psargs-330) [95583.913312] [95583.913315] Local0: 000000004c0e10cd <Obj> Integer 0000000000000001 [95583.913327] Local1: 000000002e7bf245 <Obj> Integer 0000000000000000 [95583.913335] Local2: 000000006f034898 <Obj> Integer 0000000000000000 [95583.913344] Local3: 00000000fdde280a <Obj> Integer 0000000000000000 [95583.913352] Local4: 000000004bc152cc <Obj> Integer 0000000000000000 [95583.913360] Local5: 000000005b1d7087 <Obj> Integer 0000000000000000 [95583.913368] Local6: 000000004a7ca0c2 <Obj> Integer 0000000000000000 [95583.913378] No Arguments are initialized for method [FNCL] [95583.913383] ACPI Error: Method parse/execution failed \_TZ.FNCL, AE_NOT_FOUND (20181213/psparse-531) [95583.913404] ACPI Error: Method parse/execution failed \_TZ.FN01._OFF, AE_NOT_FOUND (20181213/psparse-531) [95583.913525] acpi LNXPOWER:17: Turning OFF [95583.925258] ACPI BIOS Error (bug): Could not resolve [\_SB.PCI0.LPCB.HEC.ECAV], AE_NOT_FOUND (20181213/psargs-330) [95583.925290] [95583.925293] Local0: 000000005b1d7087 <Obj> Integer 0000000000000000 [95583.925305] Local1: 0000000093d19097 <Obj> Integer 0000000000000000 [95583.925313] Local2: 000000006f034898 <Obj> Integer 0000000000000000 [95583.925322] Local3: 00000000fdde280a <Obj> Integer 0000000000000000 [95583.925330] Local4: 000000004c0e10cd <Obj> Integer 0000000000000000 [95583.925338] Local5: 00000000282fe88a <Obj> Integer 0000000000000000 [95583.925346] Local6: 000000004a7ca0c2 <Obj> Integer 0000000000000000 [95583.925356] No Arguments are initialized for method [FNCL] [95583.925361] ACPI Error: Method parse/execution failed \_TZ.FNCL, AE_NOT_FOUND (20181213/psparse-531) [95583.925382] ACPI Error: Method parse/execution failed \_TZ.FN00._OFF, AE_NOT_FOUND (20181213/psparse-531) [95583.927984] PM: Finishing wakeup.
Created attachment 281155 [details] otcpl-dell-p5510-xeon-1_mem.html
Created attachment 281157 [details] otcpl-dell-p5510-xeon-2_mem.html
Created attachment 281219 [details] dell-precision-5510-xeon-1-acpidump.out
Created attachment 281221 [details] dell-precision-5510-xeon-2-acpidump.out
P.S. the dmesg logs are embedded in the html timelines attached. Open them in a browser and click the "dmesg" button in the upper right hand corner.
Created attachment 281237 [details] otcpl-dell-p5510-xeon-1_bootdmesg.log
Created attachment 281239 [details] otcpl-dell-p5510-xeon-2_bootdmesg.log
Like 202583 please try to go into BIOS setup and turn off C-states and repeat this test run. See if it happens again.
I tried disabling cstate switching from the kernel command line and via sysfs. These are the changes I tried with a rerun of the stress test: 1) processor.max_cstate=0 2) intel_idle.max_cstate=0 3) processor.max_cstate=0 intel_idle.max_cstate=0 4) disable all but state0 via /sys/devices/system/cpu/cpuN/cpuidle/stateN/disable No change.
This is a BIOS bug. $ grep ECAV *.dsl ssdt6.dsl: External (_SB_.PCI0.LPCB.H_EC.ECAV, IntObj) ssdt6.dsl: If ((\_SB.PCI0.LPCB.H_EC.ECAV && ETMD)) ssdt6.dsl: If (\_SB.PCI0.LPCB.H_EC.ECAV) ssdt6.dsl: If (\_SB.PCI0.LPCB.H_EC.ECAV) ssdt8.dsl: External (_SB_.PCI0.LPCB.H_EC.ECAV, UnknownObj) ssdt8.dsl: If ((\_SB.PCI0.LPCB.H_EC.ECAV == One)) ssdt8.dsl: If ((\_SB.PCI0.LPCB.H_EC.ECAV == One)) ssdt8.dsl: If ((\_SB.PCI0.LPCB.H_EC.ECAV == One)) ssdt8.dsl: If ((\_SB.PCI0.LPCB.H_EC.ECAV == One)) ssdt8.dsl: If ((\_SB.PCI0.LPCB.H_EC.ECAV == One)) ssdt8.dsl: If ((\_SB.PCI0.LPCB.H_EC.ECAV == One)) ssdt8.dsl: If ((\_SB.PCI0.LPCB.H_EC.ECAV == One)) ssdt8.dsl: If ((\_SB.PCI0.LPCB.H_EC.ECAV == One)) ssdt8.dsl: If ((\_SB.PCI0.LPCB.H_EC.ECAV == One)) ssdt8.dsl: If ((\_SB.PCI0.LPCB.H_EC.ECAV == One)) ssdt8.dsl: If ((\_SB.PCI0.LPCB.H_EC.ECAV == One)) ssdt8.dsl: If ((\_SB.PCI0.LPCB.H_EC.ECAV == One)) There is no definition of ECAV in any AML tables, thus we will get the ACPI warnings every time when \_TZ.FNCL is invoked (ACPI Fan is turned on). The reason why this issue is seen 5 times out of every 2000 tests is because we turn on the ACPI Fan device only if the temperature is high after resume. And apparently, the system does not overheat after resume most of the times. I think we can close this bug report as it is a BIOS issue. Mario, can you please raise this to BIOS team to get this fixed?
Created attachment 289125 [details] issue.def