Random Black Screens

[Mac... nificent]

It appears that this "might" be a Intel hardware issue with the HD4000 (not a big surprise to most of us). Still need to collect more data to confirm. If this is correct Apple will most likely be replacing the motherboard with a new one that has the new chips, but as I said - more data still needs to be collected before this will look to be the case.

[motrek]

Originally Posted by Mac... nificent

It appears that this "might" be a Intel hardware issue with the HD4000 (not a big surprise to most of us). Still need to collect more data to confirm. If this is correct Apple will most likely be replacing the motherboard with a new one that has the new chips, but as I said - more data still needs to be collected before this will look to be the case.

Why do you think it might be a hardware issue that would require motherboard replacement?

Here's some information from people who actually work at Intel:

http://communities.intel.com/message/171244

There's some indication, according to an actual Intel employee, that it has to do with a memory vs. BIOS incompatibility which Apple should be able to fix easily with a software update.

Please post your evidence that you think it's a hardware issue. If you don't have any then I don't see the point of trying to scare everybody.

[OldSchoolMacGuy]

It won't be a hardware replacement issue. It's going to be a firmware update and that'll be the end of it.

[Mac... nificent]

Originally Posted by motrek

Why do you think it might be a hardware issue that would require motherboard replacement?

First, if you read my post I said several times that this "might" be the issue, not that it is the issue. Why? Because someone on the Apple site had the issue, sent their mini back to Apple, and Apple replaced their motherboard and the problem has not come back Just trying to determine if the motherboard had new chips on it or if anything else was updated, that's all.

----------

Originally Posted by motrek

Here's some information from people who actually work at Intel:

There's a better link from Intel which says that they (Intel) are aware of the MM issue and they have sent Apple new drivers for this. I think Apple just needs more time.

[jollino]

Originally Posted by Mac... nificent

First, if you read my post I said several times that this "might" be the issue, not that it is the issue. Why? Because someone on the Apple site had the issue, sent their mini back to Apple, and Apple replaced their motherboard and the problem has not come back Just trying to determine if the motherboard had new chips on it or if anything else was updated, that's all.

It may be that the new motherboard simply had a (tentative) fix already applied, and it happens to be fixed indeed... besides, a firmware update may work around hardware problems. Actually even the OS itself can do so: the Linux kernel has long contained a workaround for the infamous F00F bug of early Pentium CPUs; at boot time the bug is tested, and if it is present the workaround code is loaded into memory. So we have this:
- Easiest solution: a kernel extension in 10.8.3 avoids the bug
- Easy solution: firmware upgrade for EFI to fix the bug or install a permanent workaround
- Hard solution: recall all the Minis sold so far for replacement under warranty

Since the issue has gone public, Apple has all the interest in having this fixed and is certainly aiming at doing so at the most effective cost, ie. issuing an OS or firmware upgrade.

Also, thanks Macnificent for having stopped writing in a larger font.

[Mac... nificent]

Originally Posted by jollino

Easiest solution: a kernel extension in 10.8.3 avoids the bug

10.8.3 was just seeded and it DOES address graphics, so fingers crossed

[Carl%R]

The reply from Intel on their forum has been edited now. I hope it doesn't mean bad news.

[Rob.G]

I just had another black screen crash. In fact, I got home from work again and got a snowy secondary screen minutes later. But moments ago, about a half hour after being home, I was playing a came of Civilization V and the screen went black. I heard a "ticking" noise coming from the speakers for 5-10 seconds, and then the machine automatically rebooted.

What's interesting is that I ssh'd into the thing with my MacBook Air and did tail -f on /var/log/system.log just in case this might happen. Here's what I saw:

et up page flip mode on display 0x042490e4 device: 0x10e8a8320 isBackBuffered:
1 numComp: 3 numDisp: 3
Nov 29 16:22:46 Grue kernel[0]: CODE SIGNING: cs_invalid_page(0x1000): p=1827[Go
ogleSoftwareUp] clearing CS_VALID
Nov 29 16:22:55 Grue.local KernelEventAgent[84]: tid 00000000 received event(s)
VQ_LOWDISK, VQ_VERYLOWDISK (516)
Nov 29 16:22:55 Grue kernel[0]: HFS: Vol: Google Chrome 23.0.1271.91-23.0.1271.9
5 Update Very Low Disk: freeblks: 0, dangerlimit: 10
Nov 29 16:22:56 Grue.local defaults[1866]:
The domain/default pair of (/Applications/Google Chrome.app/Contents/Info, KSBrandID) does not exist
Nov 29 16:22:56 Grue kernel[0]: CODE SIGNING: cs_invalid_page(0x1000): p=1857[ksadmin] clearing CS_VALID
Nov 29 16:22:56 Grue kernel[0]: CODE SIGNING: cs_invalid_page(0x1000): p=1860[ksadmin] clearing CS_VALID
Nov 29 16:22:56 Grue kernel[0]: CODE SIGNING: cs_invalid_page(0x1000): p=1862[ksadmin] clearing CS_VALID
Nov 29 16:23:07 Grue.local defaults[4352]:
The domain/default pair of (/Applications/Google Chrome.app/Contents/Info, KSChannelID) does not exist
Nov 29 16:23:07 Grue kernel[0]: CODE SIGNING: cs_invalid_page(0x1000): p=4358[ksadmin] clearing CS_VALID
Nov 29 16:30:04 localhost bootlog[0]: BOOT_TIME 1354235404 0
Nov 29 16:30:05 localhost kernel[0]: PMAP: PCID enabled
Nov 29 16:30:05 localhost kernel[0]: PMAP: Supervisor Mode Execute Protection enabled


When I got the black screen, I looked at the MBA and the last log entry was the one at 16:23:07... the ones starting at 16:30 are the reboot processes.

Am I seeing memory issues causing this? Not sure if this is related to the "regular" black screen problem or if this was game-related.

Rob

[Mac... nificent]

Originally Posted by Rob.G

Am I seeing memory issues causing this?

Intel did mention something about memory regarding this, but not sure if it applied to Apples version or PC's. As you said it might also just be game related. Can you get it to do it again?

[motrek]

Originally Posted by Mac... nificent

Intel did mention something about memory regarding this, but not sure if it applied to Apples version or PC's. As you said it might also just be game related. Can you get it to do it again?

There's no such thing as an Apple or PC version of a Core i5/i7 or RAM.

[Rob.G]

Originally Posted by Mac... nificent

Intel did mention something about memory regarding this, but not sure if it applied to Apples version or PC's. As you said it might also just be game related. Can you get it to do it again?

I'll try, but it lost a lot of my game progress, so I'm a little burned out on CivV at the moment.

Rob

[Westyfield2]

Originally Posted by Rob.G

I just had another black screen crash. In fact, I got home from work again and got a snowy secondary screen minutes later. But moments ago, about a half hour after being home, I was playing a came of Civilization V and the screen went black. I heard a "ticking" noise coming from the speakers for 5-10 seconds, and then the machine automatically rebooted.

What's interesting is that I ssh'd into the thing with my MacBook Air and did tail -f on /var/log/system.log just in case this might happen. Here's what I saw:

et up page flip mode on display 0x042490e4 device: 0x10e8a8320 isBackBuffered:
1 numComp: 3 numDisp: 3
Nov 29 16:22:46 Grue kernel[0]: CODE SIGNING: cs_invalid_page(0x1000): p=1827[Go
ogleSoftwareUp] clearing CS_VALID
Nov 29 16:22:55 Grue.local KernelEventAgent[84]: tid 00000000 received event(s)
VQ_LOWDISK, VQ_VERYLOWDISK (516)
Nov 29 16:22:55 Grue kernel[0]: HFS: Vol: Google Chrome 23.0.1271.91-23.0.1271.9
5 Update Very Low Disk: freeblks: 0, dangerlimit: 10
Nov 29 16:22:56 Grue.local defaults[1866]:
The domain/default pair of (/Applications/Google Chrome.app/Contents/Info, KSBrandID) does not exist
Nov 29 16:22:56 Grue kernel[0]: CODE SIGNING: cs_invalid_page(0x1000): p=1857[ksadmin] clearing CS_VALID
Nov 29 16:22:56 Grue kernel[0]: CODE SIGNING: cs_invalid_page(0x1000): p=1860[ksadmin] clearing CS_VALID
Nov 29 16:22:56 Grue kernel[0]: CODE SIGNING: cs_invalid_page(0x1000): p=1862[ksadmin] clearing CS_VALID
Nov 29 16:23:07 Grue.local defaults[4352]:
The domain/default pair of (/Applications/Google Chrome.app/Contents/Info, KSChannelID) does not exist
Nov 29 16:23:07 Grue kernel[0]: CODE SIGNING: cs_invalid_page(0x1000): p=4358[ksadmin] clearing CS_VALID
Nov 29 16:30:04 localhost bootlog[0]: BOOT_TIME 1354235404 0
Nov 29 16:30:05 localhost kernel[0]: PMAP: PCID enabled
Nov 29 16:30:05 localhost kernel[0]: PMAP: Supervisor Mode Execute Protection enabled


When I got the black screen, I looked at the MBA and the last log entry was the one at 16:23:07... the ones starting at 16:30 are the reboot processes.

Am I seeing memory issues causing this? Not sure if this is related to the "regular" black screen problem or if this was game-related.

Rob

That just looks like Chrome being rubbish with memory management to me and not to do with the HDMI bug .

----------

Originally Posted by Mac... nificent

First, if you read my post I said several times that this "might" be the issue, not that it is the issue. Why? Because someone on the Apple site had the issue, sent their mini back to Apple, and Apple replaced their motherboard and the problem has not come back Just trying to determine if the motherboard had new chips on it or if anything else was updated, that's all. [COLOR="#808080"]

Yeah I saw that... but then over on Apple Discussions everyone was obsessed with the amount of VRAM allocated and didn't realise it's based on your DRAM so I'm not sure whether to believe it...

[simon.hibbs]

Until yesterday I had occasional (once every few days) black screen issues for a second or two, then it would come back and be fine.

Yesterday it black screened twice for extended periods. Once for about an hour and again for 20 minutes.

I called Applecare and went through a system power off and reset process, but it didn't fix the issue so they arranged an appointment for me at an Apple Store to have the firmware updated. The rep said they are working on a software fix, but they already have a firmware fix they can do in-store. The appointment is for Sunday morning, so fingers crossed.

Simon Hibbs

[jollino]

Originally Posted by simon.hibbs

The rep said they are working on a software fix, but they already have a firmware fix they can do in-store. The appointment is for Sunday morning, so fingers crossed.

This is great news, please keep us posted!

[Westyfield2]

Originally Posted by simon.hibbs

I called Applecare and went through a system power off and reset process, but it didn't fix the issue so they arranged an appointment for me at an Apple Store to have the firmware updated. The rep said they are working on a software fix, but they already have a firmware fix they can do in-store. The appointment is for Sunday morning, so fingers crossed.

Now that is interesting!

[simon.hibbs]

Thinking about it, it makes sense that if Intel have provided the firmware fix to Apple but that Apple need a few days to wrap it in a downloadable installer and do some thorough testing.

It's actually pretty impressive that you can just pop over to a store and have an engineer do a while-you-wait manual firmware install for you for free though. Well, I paid for Applecare, but I wouldn't be at all surprised if you could get this done without it.

Simon Hibbs

[Westyfield2]

Originally Posted by simon.hibbs

Thinking about it, it makes sense that if Intel have provided the firmware fix to Apple but that Apple need a few days to wrap it in a downloadable installer and do some thorough testing.

It's actually pretty impressive that you can just pop over to a store and have an engineer do a while-you-wait manual firmware install for you for free though. Well, I paid for Applecare, but I wouldn't be at all surprised if you could get this done without it.

Simon Hibbs

I'd be careful that you don't start celebrating prematurely yet...

[adonis3k]

Originally Posted by Mac... nificent

10.8.3 was just seeded and it DOES address graphics, so fingers crossed

x2

[OldSchoolMacGuy]

The 10.8.3 graphics fixes are related to all Macs, not the mini specifically. We'll see our fix come in the form of a dedicated firmware or graphics update.

[motrek]

Originally Posted by OldSchoolMacGuy

The 10.8.3 graphics fixes are related to all Macs, not the mini specifically. We'll see our fix come in the form of a dedicated firmware or graphics update.

... Which could be easily bundled within a general OS update.

[OldSchoolMacGuy]

Originally Posted by motrek

... Which could be easily bundled within a general OS update.

Something that Apple doesn't generally do. OS updates lack hardware model specific fixes in most all cases.

[Mac... nificent]

Originally Posted by Westyfield2

over on Apple Discussions everyone was obsessed with the amount of VRAM allocated and didn't realise it's based on your DRAM

I don't think it is. There have been exceptions to that rule.

Anyway, Intel has sent the new drivers to Apple. Apple is testing them now.

[Jig3n]

Just had my 1st hdmi monitor glitch a moment ago.
I've returned it twice and this 3rd time around I didn't get any glitches for 13 days.
The difference this time was I used 10.8.1 the whole time.
Yesterday I updated to 10.8.2 and 24 hours later the monitor blanked out and went all snowy.

I hope that firmware update works out.

[iMarvin]

Updated mine through the osx server to 10.8.2 and i don't have no more black screens.

[Mac... nificent]

Originally Posted by Jig3n

The difference this time was I used 10.8.1 the whole time. Yesterday I updated to 10.8.2 and 24 hours later the monitor blanked out and went all snowy.

So it worked fine with 10.8.1?

Why not just roll back to it via your backup? You do practice regular backups right?