Anyone seen this error before?

[dirkjello]

Hi everyone, first time caller here.

I have a Mac Pro 2,1 that's been having some random reboot issues, and I'm trying to track down the cause in the console logs. While I haven't found anything conclusive yet, I have seen this odd error popping up in the logs repeatedly (usually more than 10x in the same second):

5/31/13 5:36:10 PM	/System/Library/CoreServices/SystemUIServer.app/Contents/MacOS/SystemUIServer[212]	FIXME: IOUnserialize has detected a string that is not valid UTF-8, "_�
<".

5/31/13 5:36:10 PM	/System/Library/CoreServices/SystemUIServer.app/Contents/MacOS/SystemUIServer[212]	FIXME: IOUnserialize has detected a string that is not valid UTF-8, "(��L HS-FLASH C".

I've done some Google searching, and haven't turned up much, unfortunately. Anyone have any ideas as to what might be causing it, and might this be a symptom of the random restarts?

The salient system details:
OS: Snow Leopard 10.6.8
Storage (internal):
- 240GB Samsung 840 SSD (with TRIM enabler installed)
- 2x 3TB Toshiba HDD (RAID1)
- 4TB Hitachi HDD
Storage (external):
- 3TB external (FW800)
- 500GB external (FW800)
- 1TB external (FW800)
- 500GB external (eSATA)
- Newer VoyagerQ (eSATA)
- Newer MAXPower 6G PCIe controller (2-port) - Slot 4
RAM: 8GB
Video: Radeon 3870
Numerous USB 2.0 accessories connected, and a couple of FW400 devices.

Thanks in advance for any help you might be able to provide!
Jeff

 

[ScottishCaptain]

"HS-FLASH" should be a clue.

This either has to do with your internal Samsung SSD, or some external USB device. I hate to be Captain Obvious here, but I would assume the device (whatever it is) has some kind of flash storage inside it.

How is your system restarting? If it is panicking and leaving kernel panics, then you should check those out to figure out what is going sideways. If the system is simply rebooting out of nowhere, then that suggests a hardware problem instead (loose cables, defective PSU/logic board/CPU/etc).

-SC

 

[dirkjello]

"HS-FLASH" should be a clue.

This either has to do with your internal Samsung SSD, or some external USB device. I hate to be Captain Obvious here, but I would assume the device (whatever it is) has some kind of flash storage inside it.

How is your system restarting? If it is panicking and leaving kernel panics, then you should check those out to figure out what is going sideways. If the system is simply rebooting out of nowhere, then that suggests a hardware problem instead (loose cables, defective PSU/logic board/CPU/etc).

-SC

Thanks, Captain. It's rebooting out of nowhere, at random; no discernible pattern that I can tell. No panics in the log; just notes of "Improper shutdown detected."

I'll check and see if I've knocked any cables loose in my various installs.

Thanks,
Jeff

 

[dirkjello]

Might a faulty RAM module be causing these? I'm getting quite a few parity errors from one module.

 

[ScottishCaptain]

Parity errors are not normal.

You should probably try reseating all your RAM, but swap the suspect module with another. If the errors follow the RAM stick, then that stick is defective and should be replaced. If the error stays with the slot (and another stick suddenly has issues), then the logic board might be defective.

I'm kind of thinking that your RAM got bumped or jostled and there is a problem with the mechanical connection. Usually when a machine spontaneously reboots due to bad RAM (assuming the kernel doesn't panic first), a Machine Check Exception (MCE) is generated and logged during next startup.

If that's not happening, then whatever is happening is low level enough that the machine simply doesn't have time to even recognize there's an issue or log it- it just reboots. Something like an intermittent mechanical connection on a cable or RAM stick would do that. A failing RAM module probably wouldn't, unless it was failing in such a way that it is doing something really spectacular electrically.

-SC

 

[dirkjello]

Parity errors are not normal.

You should probably try reseating all your RAM, but swap the suspect module with another. If the errors follow the RAM stick, then that stick is defective and should be replaced. If the error stays with the slot (and another stick suddenly has issues), then the logic board might be defective.

I'm kind of thinking that your RAM got bumped or jostled and there is a problem with the mechanical connection. Usually when a machine spontaneously reboots due to bad RAM (assuming the kernel doesn't panic first), a Machine Check Exception (MCE) is generated and logged during next startup.

If that's not happening, then whatever is happening is low level enough that the machine simply doesn't have time to even recognize there's an issue or log it- it just reboots. Something like an intermittent mechanical connection on a cable or RAM stick would do that. A failing RAM module probably wouldn't, unless it was failing in such a way that it is doing something really spectacular electrically.

Thanks, Captain. I removed, blew off, and re-seated all of the RAM this morning; I'll see if I continue to get parity errors (none so far). If so, I'll move them around and see if I can pin down the cause from there.

I also re-seated the SSD in its Icy Dock, and made sure it was firmly installed.

I do think that I've figured out the earlier HS-FLASH C issue though; the card reader in my monitor (Dell) went south a while back, and I still had the USB cable plugged in because the hub still worked. Removing the USB cable has apparently eliminated the IOUnserialize errors, as I would always see them in the logs on restart, and I'm not seeing them now (fingers crossed it stays that way).

In combing through the logs, I did see a PowerManager kernel panic issue that I didn't notice before. (It had been suggested to me elsewhere that the machine was being rebooted in the night due to power flickers triggering an issue with the UPS it's on, which is probably due for a replacement soon).

There's just all kinds of things wonky here!

Thanks again!

 

[dirkjello]

So far, so good.

In addition to removing/cleaning/re-seating the RAM and the SSD, I also nuked the com.apple.powermanager prefs, disconnected the USB connection to the UPS, unplugged the USB cable to the monitor (which fixed the "HS-FLASH" messages), reset the NVRAM, and ran memtest for 5 cycles (10.5 hours, didn't find anything wrong).

Since doing all of those steps (some of which I admit might've been overkill), I haven't seen the issues come back (touch wood).

Thanks again for the assistance, Captain.

Best,
Jeff