FWTS fails method and klog tests on SeaMicro SM15000-OP

[Samantha Jian-Pielak]

FWTS method and klog tests reported Critical and High failure against the SM150000-OP, AMD opteron config blade server. I will attach the fwts test logs and post the test summary below. Please let me know these should gate a certificate for this system.

Method:
Test Failure Summary
================================================================================

Critical failures: 21
 method: Detected error 'Not exist' when evaluating '\_SB_.PCI0.HPET._CRS'.
 method: Detected error 'Not exist' when evaluating '\_SB_.PCI0.LPC0.RTC_._CRS'.
 method: Detected error 'Not exist' when evaluating '\_SB_.PCI0.LPC0.TMR_._CRS'.
 method: Detected error 'Not exist' when evaluating '\_SB_.PCI0.HPET._STA'.
 method: Detected error 'Not exist' when evaluating '\_SB_.PCI0.SATA.PRID.P_D0._STA'.
 method: Detected error 'Not exist' when evaluating '\_SB_.PCI0.SATA.PRID.P_D1._STA'.
 method: Detected error 'Not exist' when evaluating '\_SB_.PCI0.SATA.SECD.S_D0._STA'.
 method: Detected error 'Not exist' when evaluating '\_SB_.PCI0.SATA.SECD.S_D1._STA'.
 method: Detected error 'Not exist' when evaluating '\_SB_.PCI0.SATA._INI'.
 method: Detected error 'Not exist' when evaluating '\_SB_.PCI0.SATA.PRID._PS0'.
 method: Detected error 'Not exist' when evaluating '\_SB_.PCI0.SATA.PRID.P_D0._PS0'.
 method: Detected error 'Not exist' when evaluating '\_SB_.PCI0.SATA.PRID.P_D1._PS0'.
 method: Detected error 'Not exist' when evaluating '\_SB_.PCI0.SATA.SECD._PS0'.
 method: Detected error 'Not exist' when evaluating '\_SB_.PCI0.SATA.SECD.S_D0._PS0'.
 method: Detected error 'Not exist' when evaluating '\_SB_.PCI0.SATA.SECD.S_D1._PS0'.
 method: Detected error 'Not exist' when evaluating '\_SB_.PCI0.USB0._PSW'.
 method: Detected error 'Not exist' when evaluating '\_SB_.PCI0.USB1._PSW'.
 method: Detected error 'Not exist' when evaluating '\_SB_.PCI0.USB2._PSW'.
 method: Detected error 'Not exist' when evaluating '\_SB_.PCI0.USB3._PSW'.
 method: Detected error 'Not exist' when evaluating '\_SB_.PCI0.USB4._PSW'.
 method: Detected error 'Not exist' when evaluating '\_WAK'.

High failures: NONE

Medium failures: 1
 method: Method \_WAK did not return ACPI_TYPE_PACKAGE.

Low failures: NONE

Other failures: NONE

Klog:

Test Failure Summary
================================================================================

Critical failures: NONE

High failures: 3
 klog: HIGH Kernel message: [ 4.652608] ACPI Exception: AE_NOT_FOUND, While evaluating Sleep State [\_S3_] (20131115/hwxface-580)
 klog: HIGH Kernel message: [ 4.763657] ACPI Exception: AE_NOT_FOUND, While evaluating Sleep State [\_S4_] (20131115/hwxface-580)
 klog: HIGH Kernel message: [ 15.108267] [Firmware Bug]: cpu 4, IBS interrupt offset 0 not available (MSRC001103A=0x0000000000000100)

Medium failures: 9
 klog: MEDIUM Kernel message: [ 14.942070] [Firmware Bug]: cpu 4, try to use APIC500 (LVT offset 0) for vector 0x10400, but the register is already in use for vector 0xf9 on another cpu
 klog: MEDIUM Kernel message: [ 16.369633] pcieport 0000:00:02.0: device [1002:5a16] has invalid IRQ; check vendor BIOS
 klog: MEDIUM Kernel message: [ 16.467098] pcieport 0000:00:03.0: device [1002:5a17] has invalid IRQ; check vendor BIOS
 klog: MEDIUM Kernel message: [ 16.564513] pcieport 0000:00:04.0: device [1002:5a18] has invalid IRQ; check vendor BIOS
 klog: MEDIUM Kernel message: [ 16.661913] pcieport 0000:00:05.0: device [1002:5a19] has invalid IRQ; check vendor BIOS
 klog: MEDIUM Kernel message: [ 16.759312] pcieport 0000:00:06.0: device [1002:5a1a] has invalid IRQ; check vendor BIOS
 klog: MEDIUM Kernel message: [ 16.856714] pcieport 0000:00:07.0: device [1002:5a1b] has invalid IRQ; check vendor BIOS
 klog: MEDIUM Kernel message: [ 16.954187] pcieport 0000:00:09.0: device [1002:5a1c] has invalid IRQ; check vendor BIOS
 klog: MEDIUM Kernel message: [ 16.954368] pcieport 0000:00:0a.0: device [1002:5a1d] has invalid IRQ; check vendor BIOS

Low failures: NONE

Other failures: NONE

[Alex Hung (alexhung)]

I noticed it is done by FWTS V14.02.00. Newer version of FWTS fixes some of the ACPI method errors. Please run again and update the result.

[Rod Smith (rodsmith)]

Here are results run today. These include both the results from running canonical-certification-server and of running fwts directly.

[Alex Hung (alexhung)]

@Roderick,

the results.log shows something like

method: ACPI DSDT Method Semantic tests.
--------------------------------------------------------------------------------
Failed to initialise tables.
Cannot initialise ACPI.
Aborted test, initialisation failed.
================================================================================
0 passed, 0 failed, 0 warning, 155 aborted, 0 skipped, 0 info only.
================================================================================

mcfg: MCFG PCI Express* memory mapped config space test.
--------------------------------------------------------------------------------
Must be run as root or sudo to be able to read system information.
Aborted test, insufficient privilege.
================================================================================
0 passed, 0 failed, 0 warning, 2 aborted, 0 skipped, 0 info only.
================================================================================

Did you run as root, ex. sudo fwts?

[Rod Smith (rodsmith)]

I may have forgotten sudo; here's another run when I definitely did not forget sudo.

[Alex Hung (alexhung)]

#5 shows the method errors are gone, and they are fixed in latest fwts.

[Alex Hung (alexhung)]

The below two log errors are mistriggered by linux kernel. They have no harm (a patch sent to upstream but no response).

 klog: HIGH Kernel message: [ 4.652608] ACPI Exception: AE_NOT_FOUND, While evaluating Sleep State [\_S3_] (20131115/hwxface-580)
 klog: HIGH Kernel message: [ 4.763657] ACPI Exception: AE_NOT_FOUND, While evaluating Sleep State [\_S4_] (20131115/hwxface-580)

[Alex Hung (alexhung)]

Unfortunately AMD's spec to us to check the below message. I will try to ask around whether this will impact

klog: HIGH Kernel message: [ 15.108267] [Firmware Bug]: cpu 4, IBS interrupt offset 0 not available MSRC001103A=0x0000000000000100)

However, http://developer.amd.com/wordpress/media/2012/10/AMD_IBS_paper_EN.pdf presents IBS and it is used to "collects a wide range of performance information in a single program run, making it easier to conduct performance testing." IBS does not sound like a function normally used but for performance measuring.

I do not think the error message is harmful.

[Samantha Jian-Pielak (samantha-jian)]

@Alex,

Could you please check the attached fwts test results on another SM15000 configuration?

I am running the same version of fwts as Ron did in #5, and the method test reports the following critical failures:
Critical failures: 5
 method: Detected error 'Division by zero' when evaluating '\_SB_.PCI0.LPCB.H_EC.BAT1._BTP'.
 method: Detected error 'Type' when evaluating '\_SB_.PCI0.LPCB.H_EC.BAT0._PCL'.
 method: Detected error 'Type' when evaluating '\_SB_.PCI0.LPCB.H_EC.BAT1._PCL'.
 method: Detected error 'Type' when evaluating '\_SB_.PCI0.LPCB.H_EC.BAT2._PCL'.
 method: Detected error 'Type' when evaluating '\_SB_.ADP1._PCL'.

[Samantha Jian-Pielak (samantha-jian)]

Please ignore, I opened a new bug for this is different failures. Bug 1311777.