Strange Constant Freezes w/ pixel glitch in 10.8.2 - Mac Pro Early 2008

Question

Level 1

8 points

Strange Constant Freezes w/ pixel glitch in 10.8.2 - Mac Pro Early 2008

Running an early 2008 Mac Pro (3,1) with 16GB RAM.

Installed 10.8.2 onto a clean, new hard drive. After a few weeks, I keep getting freezes on my machine. Not crashes (meaning I have to manually reset), but the screen freezes. It even gives me a few seconds of strange warning because I get some pixel glitches on my machine right before it happens (see below).

Newest piece of hardware on the machine is some RAM. I tested the RAM on a memtest and came back all clean (even pulled the new RAM and still got the freeze). I was also using an additional older internal hard drive for storage. Pulled it out and still got freezes. Freezes have also become increasingly frequent.

I'm pretty new to reading error logs from Console, but they all seem pretty inconsistent to me (but again, I'm a novice).

Here are some lines from my last crash logs

12/12/12 10:34:47.446 AM com.apple.launchd.peruser.502[359]: (com.seagate.notificationexec.plist) Throttling respawn: Will start in 10 seconds

12/12/12 10:34:57.537 AM com.apple.launchd.peruser.502[359]: (com.seagate.notificationexec.plist) Throttling respawn: Will start in 10 seconds

12/12/12 10:35:04.380 AM com.avast.proxy[68]: Error reading receive buffer: read(): Connection reset by peer

12/12/12 10:35:04.776 AM com.avast.proxy[68]: Error reading receive buffer: read(): Connection reset by peer

12/12/12 10:35:07.631 AM com.apple.launchd.peruser.502[359]: (com.seagate.notificationexec.plist) Throttling respawn: Will start in 10 seconds

12/12/12 10:35:17.722 AM com.apple.launchd.peruser.502[359]: (com.seagate.notificationexec.plist) Throttling respawn: Will start in 10 seconds

12/12/12 10:35:27.817 AM com.apple.launchd.peruser.502[359]: (com.seagate.notificationexec.plist) Throttling respawn: Will start in 10 seconds

12/12/12 10:35:37.905 AM com.apple.launchd.peruser.502[359]: (com.seagate.notificationexec.plist) Throttling respawn: Will start in 10 seconds

12/12/12 10:35:40.256 AM Google Chrome Helper[2386]: Failed to look up domain bootstrapper during domain extension: 0x44c

12/12/12 10:35:40.257 AM Google Chrome Helper[2386]: Failed to look up domain bootstrapper during domain extension: 0x44c

12/12/12 10:35:40.258 AM Google Chrome Helper[2386]: Failed to look up domain bootstrapper during domain extension: 0x44c

12/12/12 10:35:40.264 AM Google Chrome Helper[2386]: Failed to look up domain bootstrapper during domain extension: 0x44c

12/12/12 10:35:48.010 AM com.apple.launchd.peruser.502[359]: (com.seagate.notificationexec.plist) Throttling respawn: Will start in 10 seconds

12/12/12 10:35:58.100 AM com.apple.launchd.peruser.502[359]: (com.seagate.notificationexec.plist) Throttling respawn: Will start in 10 seconds

12/12/12 10:35:59.757 AM Google Chrome Helper EH[2431]: Unsure about the internals of CFAllocator but going to patch them anyway. If there is a crash inside of CFAllocatorAllocate, please report it at http://crbug.com/117476 . If there is a crash and it is NOT inside of CFAllocatorAllocate, it is NOT RELATED. DO NOT REPORT IT THERE but rather FILE A NEW BUG.

12/12/12 10:36:00.864 AM Google Chrome Helper EH[2431]: objc[2431]: Class WebCoreTextFieldCell is implemented in both /Applications/Google Chrome.app/Contents/Versions/21.0.1180.90/Google Chrome Framework.framework/Google Chrome Framework and /System/Library/Frameworks/WebKit.framework/Versions/A/Frameworks/WebCore.frame work/Versions/A/WebCore. One of the two will be used. Which one is undefined.

12/12/12 10:36:09.481 AM com.apple.launchd.peruser.502[359]: (com.seagate.notificationexec.plist) Throttling respawn: Will start in 9 seconds

12/12/12 10:36:18.577 AM com.apple.launchd.peruser.502[359]: (com.seagate.notificationexec.plist) Throttling respawn: Will start in 10 seconds

12/12/12 10:36:29.044 AM com.apple.launchd.peruser.502[359]: (com.seagate.notificationexec.plist) Throttling respawn: Will start in 10 seconds

****(USER NOTE: The freeze occured around here at 10:36:32)****

12/12/12 10:36:44.000 AM kernel[0]: ** Device in slot: SLOT-1 **

12/12/12 10:36:44.000 AM kernel[0]: ** GPU Debug Info Start **

12/12/12 10:36:44.000 AM kernel[0]: ** Device in slot: SLOT-1 **

12/12/12 10:36:44.000 AM kernel[0]: 0x00009588

12/12/12 10:36:44.000 AM kernel[0]: 0x000000d8

12/12/12 10:36:44.000 AM kernel[0]: ** GPU Debug Info Start **

12/12/12 10:36:44.000 AM kernel[0]: 0x00000001

12/12/12 10:36:44.000 AM kernel[0]: 0x00009588

12/12/12 10:36:44.000 AM kernel[0]: 0x00000018

12/12/12 10:36:44.000 AM kernel[0]: 0x000000d8

This is insanely frustrating. Any help would be amazing. Love what you guys do here. Thanks.

Mac Pro, OS X Mountain Lion (10.8.2)

Posted on Dec 12, 2012 8:08 AM

Reply

Answer 1

The hatter

Level 9

61,031 points

Dec 12, 2012 9:51 AM in response to radracer

Stick with 10.6.8. Problem solved. Leave ML and Lion for when you must so you need to setup dual or triple boot. Which is piece of cake.

Reply

Answer 2

radracer Author

Level 1

8 points

Dec 12, 2012 10:11 AM in response to The hatter

I'd defnitely considered going to 10.6.8. My only fear being Apple quickly phasing that out as it seems to be the minimum requirement for everything these days.

Interested to hear any feedback on stability issues of Lion.

Reply

Answer 3

The hatter

Level 9

61,031 points

Dec 12, 2012 10:31 AM in response to radracer

There are any number of people with issues beyond 10.6.8 with 3,1's many of which are power management.

2008 was the first to introduce EFI64 for one and the early units DID have a freeze/restart when waking from sleep issue at a minimum.

You wonder if they tested Lion + on these.

Reply

Answer 4

radracer Author

Level 1

8 points

Dec 12, 2012 11:40 AM in response to The hatter

I have been starting to hear more about that issue with the 3,1's. I imagine you're right about Apple probably not even testing Lion+ on these guys.

Makes me wonder if it might be time to sell and get any value I can out of this old machine.

Reply

Answer 5

Linc Davis

Level 10

209,739 points

Dec 12, 2012 12:03 PM in response to radracer

The graphics processor is overheating or has failed.

Reply

Answer 6

radracer Author

Level 1

8 points

Dec 12, 2012 12:08 PM in response to Linc Davis

Wow, thats interesting. Is there a line of code that lets you know that, or you can tell from the pixel glitches?

Also wondering if a hardware fix will help this or if ML just hates my machine.

Reply

Answer 7

Linc Davis

Level 10

209,739 points

Dec 12, 2012 12:23 PM in response to radracer

GPU Debug Info Start

Reply

Answer 8

radracer Author

Level 1

8 points

Dec 12, 2012 12:32 PM in response to Linc Davis

hmmm...

is that hardware failure i should be worried about or just ML not playing nice with my machine?

Reply

Answer 9

Linc Davis

Level 10

209,739 points

Dec 12, 2012 12:40 PM in response to radracer

It is a hardware fault, but not necessarily reproducible in another version of OS X due to differences in the graphics drivers and/or thermal management routines.

Reply

Answer 10

radracer Author

Level 1

8 points

Dec 17, 2012 6:37 PM in response to Linc Davis

I seem to have temporarily fixed it.

First, I decided to pop open my machine and clean out the dust (there was a LOT of accumulation, especially around my video card). Then I found this program called SMS Fan Control (http://www.macupdate.com/app/mac/23049/smcfancontrol) which is a really great program. I turned up my fans a little bit faster than the default setting to keep my machine cooler and haven't had another crash yet (fingers crossed).

However, I noticed that while booting in regular mode is fine, when I try and boot in safe mode, the progress bar freezes every time.

Reply

Answer 11

Linc Davis

Level 10

209,739 points

Dec 17, 2012 6:51 PM in response to radracer

Make a "Genius" appointment at an Apple Store to have the machine tested.

Back up all data on the internal drive(s) before you hand over your computer to anyone. If privacy is a concern, erase the data partition(s) with the option to write zeros* (do this only if you have at least two complete, independent backups, and you know how to restore to an empty drive from any of them.) Don’t erase the recovery partition, if present.

*An SSD doesn't need to be zeroed.

Reply

Answer 12

radracer Author

Level 1

8 points

Dec 19, 2012 8:04 AM in response to Linc Davis

Thanks.

My fan solution stopped working, so I think I'm finally going to break down and go to the said Genius bar and see if they'll diagnose it.

Starting to wonder if putting another (what I'm guessing will be) 500 or so dollars into this 5 year old machine is even worth it. Would be interested to know how many of you cling tight to older machines or if 5 years is usually time to give up the ghost.

Reply

Answer 13

Dec 19, 2012 9:46 AM in response to radracer

It's not a "number of years" issue -- it's a Computer Architecture issue.

No one is advised by the cognoscenti here to make major investments in Mac Pro 2008 or previous, except to upgrade a dead graphics card to keep it running, and maybe add modest amounts of overly-expensive memory.

Reply

Answer 14

radracer Author

Level 1

8 points

Dec 29, 2012 9:23 AM in response to Linc Davis

Just a follow up. Took my computer to the said Genius Bar and they had no definitive diagnosis but said graphics card seemed likely. Their solution was to have them install another card (the same brand and model that was apparently defective in the first place). Naturally, I declined.

Instead, I bought an Nvidia Geforce GT 120 512MB, installed it 2 days ago and things are running smoothly (only 140 bucks used).

Thanks for the help Linc. I think I'm gonna definitely try and get some more life out of this computer now.

Reply

Answer 15

Dec 29, 2012 10:45 AM in response to radracer

Readers need to use caution with the Apple-firmware GT120 -- there may be two versions out there, and the ones shipped in earliest Macs (pre-2008) may not properly support later models (2008 and later).

Buyer beware.

Reply