BIOS HAS CORRUPT [hw] PMU RESOURCES....

[CSG-IT]

This error comes up, booting to a fresh install of

Debian/CentOS/Fedora

on a DL 380 G5 with latest BIOS and FW.

RESULT: Server doesn't boot, hangs after grub, blank screen, no more details.

I urgently need a fix or workaround for this problem.

Any suggestions?

Thank you,

Oliver

Moved from ProLiant Servers - Netservers to ProLiant Servers (ML,DL,SL)

[TVVJ]

1. You may try clearing the server’s NVRAM. To do so, you may perform the following steps:

i. Power down the server and remove the access panel.

ii. Please refer to item 9 of the System board components diagram, on page 84 of 104 of the server’s maintenance and service guide to identify the system maintenance switch. The guide is locatable at the following link:

https://support.hpe.com/hpesc/public/docDisplay?docId=emr_na-c00710359

iii. Locate individual switch S6, and set it from the default position of ‘Off’ to ‘On’. Replace the access panel and power it on. Wait till the message indicates that the server has been reset. Then power off the server, remove the power cable from the wall jack, and remove the access panel, revert switch S6 to position back to ‘Off’ and then power it on.

After clearing NVRAM, please re-enter the serial number in the server's RBSU. You may perform the following steps:

i. During the server startup sequence, press the F9 key to access RBSU.

ii. Select the System Options menu.

iii. Select Serial Number. The following warning is displayed:  

                                WARNING! WARNING! WARNING! The serial number is loaded into the system during the manufacturing process and should NOT be modified. This option should only be used by qualified service personnel. This value should always match the serial number sticker located on the chassis.  

iv. Press the Enter key to clear the warning.

v. Enter the serial number and press the Enter key.

vi. Select Product ID.

vii. Enter the product ID and press the Enter key.

viii. Press the Esc key to close the menu.

ix. Press the Esc key to exit RBSU.

x. Press the F10 key to confirm exiting RBSU. The server will automatically reboot.

2. If the issue is not resolved, then activate redundant ROM. To do so, locate the system maintenance switch as listed above and set switches S1, S5, S6 to “On” position from the default “OFF” position and power it on. Once the message appears, then proceed to power off the server and reverse the switch settings.

3. If the issue is persists, then re-flash the server’s BIOS. To do so, you may perform the following steps:

                i. Download the ROMPaq from the following link and follow instructions seen under tab ‘Installation instructions’ to create a bootable USB ROMPaq flashkey:

                x86:

https://support.hpe.com/hpsc/swd/public/detail?swItemId=MTX_57720d956df94dfcbaa0e28256&swEnvOid=4168

                x64:

https://support.hpe.com/hpsc/swd/public/detail?swItemId=MTX_57720d956df94dfcbaa0e28256&swEnvOid=4168

https://techlibrary.hpe.com/docs/iss/proliant-gen10-uefi/s_enbable_intel_perf_montoring.html

                ii. Once the bootable USB ROMPaq flash key  is created, connect it to the server and power cycle it so that the server boots from  the USB key. If there is a message that the BIOS is already updated and option to re-flash it, then re-flash it again.

4. If the issue is still not resolved, then check if there are any POST error messages and update it.

Regards,

Vijayasarathy

 Moderator Note: [Updated with new links]

[CSG-IT]

Hi Vijayasarathy,

I am sorry but none of your given options works for me. Any other suggestions?
Here is (maybe) a more specific error message: 
After this "BIOS has corrupt..", the 2nd line reads
"ERST: Failed to get Error-Log Adress range..."

Thank you!

[Jimmy Vance]

"the BIOS has corrupted hw-PMU resources" gets displayed as the OS is trying to use a performance counter the BIOS is using.  You can either ignore the message, or turn off performance monitoring in the BIOS as shown in this advisory:

 http://h20564.www2.hpe.com/hpsc/doc/public/display?docId=emr_na-c03265132

Currently ProLiant BIOS does not support the ACPI ERST feature and can be safely ignored

[CSG-IT]

Hi Jimmy,

thanks for your reply.

Ignoring this message is not an option, because the server doesn't boot.

The link above is for G7 servers, I do own G5s only.

(Pressing CTRL+A in RBSU does nothing  ;) )

Regards

[valor88]

Hi, could I also confirm that regarding the ERST Failed to get Error Log message also applies to G7 proliant servers?

According to the this doc,

https://support.hpe.com/hpesc/public/docDisplay?docId=mmr_sf-EN_US000020344

it affects only g8 servers?

[Moderator Note: Removed the broken link. Please refer to https://support.hpe.com/]

[Jimmy Vance]

---

@CSG-IT wrote:

Ignoring this message is not an option, because the server doesn't boot.

The link above is for G7 servers, I do own G5s only.

(Pressing CTRL+A in RBSU does nothing  ;) )

Both of these messages are warnings and will not stop a system from booting. I had to go back and check, G5 doesn't have the CTRL+A feature.

[Jimmy Vance]

One thing you might try is edit the boot line and add nomodeset

[Jimmy Vance]

---

valor88 wrote:

Hi, could I also confirm that regarding the ERST Failed to get Error Log message also applies to G7 proliant servers?

According to the this doc,

https://support.hpe.com/hpesc/public/docDisplay?docId=c03731635

it affects only g8 servers?

---

The ERST message was found when testing Gen 8 systems. Since the Gen 8 system ROM doesn't support this feature, I think I can safely say the G7 system ROM doesn't either.  I dont recall when the ERST option was added to the kernel, but I don't think it was there when the G7 was certified.

As I've stated, the message is informative, it doesn't effect system performance

[Moderator edit: Updated the valid link.]

[valor88]

I currently have another same server which is running on linux 6.2. However it doesn't encounter the BIOS error. The current server which encounters the error is on 6.3. 

Any idea why this is the case?

[zou15]

i am also ubuntu 14.04 lts but it giving me the error of "firmaware biu: the BIOS has corrupted hw-PMU resources (MSR 38d is 330)". What should i do? I am using HP DL120 servers

any help please?

[Jimmy Vance]

The answer is in the 4th post of this thread date 03/27//2013

[Hpoley]

Unfortunately, the 4th posting does NOT contain the solution to this issue. The original question was regarding a DL380 G5. I am having the same issue. This appears to be universal with newer Linux releases. If anyone has a fix, I'd love to hear it.

[Jimmy Vance]

---

@Hpoley wrote:
Unfortunately, the 4th posting does NOT contain the solution to this issue. The original question was regarding a DL380 G5. I am having the same issue. This appears to be universal with newer Linux releases. If anyone has a fix, I'd love to hear it.

---

Both the Customer Advisory listed in the 4th post and the Red Hat bugzilla that is referenced from within the advisory list the resolution

Either ignore the message as it has no impact on the server or OS performance or turn off the performance counters in the BIOS.

[Maddas69]

I am having the same problem after firmware update.

The message does not prevent the server from booting, but the local console connection goes "black" when the Linux part of the boot sequence starts. It still boots and can be managed through ILO, but the local console remains unusable.

[Jimmy Vance]

---

@Maddas69 wrote:

I am having the same problem after firmware update.

The message does not prevent the server from booting, but the local console connection goes "black" when the Linux part of the boot sequence starts. It still boots and can be managed through ILO, but the local console remains unusable.

---

Not sure what is casuing your issue, but it is not 'BIOS HAS CORRUPT [hw] PMU RESOURCES....'  This error is Linux wanting to use a performance counter that the ProLiant BIOS is already using.

What model server are you working with and what ROM firmware did you install?  Try adding 'nomodeset' to your boot options

[Haman_Oleh]

---

@Jimmy Vance wrote:

---

@Maddas69 wrote:

I am having the same problem after firmware update.

The message does not prevent the server from booting, but the local console connection goes "black" when the Linux part of the boot sequence starts. It still boots and can be managed through ILO, but the local console remains unusable.

---

Not sure what is casuing your issue, but it is not 'BIOS HAS CORRUPT [hw] PMU RESOURCES....'  This error is Linux wanting to use a performance counter that the ProLiant BIOS is already using.

What model server are you working with and what ROM firmware did you install?  Try adding 'nomodeset' to your boot options

---

I have same problem  whith installing ProxMox 4.3 to HP ProLiant D360 G7 , I am alredy Disabled

Processor Power Monitoring 

but that is not solution ( I steel coud not use Remoute Console in ILO3, it show black screen. That is video of booting http://screencast.com/t/cYvBo6TeVrE

[Torsten.]

Regarding

The BIOS has Corrupted hw-PMU Resources (MSR 38d is 330)

https://support.hpe.com/hpesc/public/docDisplay?docId=emr_na-c03265132

Did you google

"Error: Driver 'pcspkr' is already registered, aborting"?

There are some possible solutions.

[Haman_Oleh]

No i didnt, but i will try, I thought that the first message is the main trouble )

[Haman_Oleh]

Thank you, after doing /etc/modprobe.d/blacklist: blacklist pcspkr that the error is gone

But now i have black screen in remoute console iLO3 but all ok on monitor screen or i got this error http://screencast.com/t/lxnCdsb5gE

Somebody know why? Sorry if this is offtop.

[Jimmy Vance]

Wonder why the radeon video driver is loading, the embeded video should be using the mga driver.  Have you added a video card?

[coconutdog]

I am getting this issue with a ProLiant DL320 G5 with a fresh install of Ubuntu 17.04, released today. It seems the Linux world moves slowly. This should have been sorted by now.

[Jimmy Vance]

---

@coconutdog wrote:

I am getting this issue with a ProLiant DL320 G5 with a fresh install of Ubuntu 17.04, released today. It seems the Linux world moves slowly. This should have been sorted by now.

---

Not at all, the hardware is using a performance counter the OS wants to own. Linus stuck a message in the kernel saying the BIOS is corrupt. If you search back through the kernel mailing lists far enough you can find the original thread.

Switch to OS power control and the message will no longer display

[sajinma]

Is it applicable for DL380 Gen9 servers as well?

I got below message during reboot after OS install

[Firmware Bug]: the BIOS has corrupted hw-PMU resources (MSR 38d is 330)

Thanks,

Sajin

[Jimmy Vance]

---

@sajinma wrote:

Is it applicable for DL380 Gen9 servers as well?

I got below message during reboot after OS install

[Firmware Bug]: the BIOS has corrupted hw-PMU resources (MSR 38d is 330)

Thanks,

Sajin

---

Yes,  this is a harmless informational message . You can disable all of the processor power management and telemetry reporting in RBSU to avoid the messages