Looks like no one’s replied in a while. To start the conversation again, simply ask a new question.

User profile for user: batondor
batondor Author
User level: Level 1
0 points

Mac OS X 10.6.3 system "freezing" problems: my experiences and conclusions

My last post was lost in the pile and so here it is in a new "topic" for consideration:

http://discussions.apple.com/thread.jspa?threadID=2403126&tstart=0&messageID=114 57265#11457265

The bottom line is that my iMac has been working without a problem since I replaced the ATIRadeonX2000 drivers from 10.6.3 with those from 10.6.2 as explained in my last post. This iMac has been running since Thursday night without problems, it has been rebooted manually once or twice to insure that it restarted without problems, and it has been running with both an external video monitor and has been accessed via VNC without incident. The temperature monitor indicates nothing out of the ordinary and it is operating with a single external USB drive and a UPS monitor via USB...

Just to be clear, I have not added any third party software yet other than the temperature monitor. I have not yet begun using the system for email or work other than via webmail interfaces, but I have used it for at least ten hours for video streaming... again, without incident.

Yesterday, I submitted a detailed report via the Apple support feedback mechanism that I will not copy verbatim here, but the gist of my conclusions are here for your consideration:

1) There clearly is an issue with accelerated graphics on this machine that seem resolved with a simple downgrade of the device drivers for this machine (and don't forget that it worked just fine, if slowly and with restricted functionality, when started with Safe Boot).

2) Other users seem to see different if similar "system freezing" issues on hardware that contain different graphics devices, and these issues sometimes involve "frozen" mice, etc... but that they almost always seem to be connected with display issues.

3) While some users report problems during installation and/or boot procedures, I am wondering whether the 10.6.3 boot doesn't evoke the latest generation graphics drivers and systems libraries during that process... because I also saw sporadic cases where the system would not even boot normally with the 10.6.3 drivers (though I should add that I had no difficulty whatsoever with the erase/install...).

My conclusion, if I may be so bold, is that there is a problem with either code common to some or all of the 10.6.3 graphics drivers, whether at the device level, in Core Graphics, or in the OpenGL libraries (or something related to any of these).

I decided to submit this post even though I offered it to Apple via feedback precisely because someone else might be able to benefit from it. I do not want to enter into an argument about how "unique" each of these problems might be and hope other contributors will respect this by not speculating about motives or imagining other scenarios. I am not able to check for responses very frequently but will attempt to respond to queries about my experience and how it apparently has been resolved (esp. those that pertain to similar hardware), but I obviously cannot be responsible for any similar attempts to temporarily correct this situation...

... and frankly, I hope that someone at Apple support does monitor these forums and that the problem will be explicitly correct sometime soon.

Good luck to all...

iMac/Intel 20 inch 2.0 Ghz (version 7,1), Mac OS X (10.6.3), 2 gb RAM and ATI Radeon HD2400

Posted on May 3, 2010 6:01 AM

Reply
74 replies
User profile for user: batondor
batondor Author
User level: Level 1
0 points

May 5, 2010 6:22 AM in response to R C-R

Frankly, R C-R, I'd suggest that you explain your dismissiveness in the face of my experience:

1) my iMac apparently does not have a hardware problem because it was completely tested by Apple...

2) my Mac OS X 10.6.3 v1.1 install has nothing in it that could be considered "third-party" software...

3) the problems do not appear at all during Safe Boot w/ the generic ATI 2400 driver running in isolation (the X2000 kext is not loaded...).

The only reason that I am guessing that there is something beyond the ATI Radeon X2000 driver at issue is because other users seem to have similar problems though they might kick in differently and in different usage scenarios... but it is hardly reasonable to dismiss it as "unique" to my machine if it can be reproduced so simply on a stock machine.

I'm not certain that Allan is correct in presuming that Apple support does not passively observe the user forums as a source of information, but I would agree that they would open up a can of worms by acknowledging it (not to mention participating in them). I would also suggest that you have contradicted yourself by dismissing my use of the feedback mechanism to offer my ideas on the nature of the problem both because that was what was requested by the Apple Expert with whom I spoke on the telephone as well as by you in a different thread.

In fact, I don't really understand the objective of your voluminous contributions to this forum because while sometimes constructive, as often as not you seem to think that Apple can do no wrong...

... to which I will retort that I am actually quite pleased by my experience with both Apple hardware and as a Mac OS X user these past five years after over thirty years as a UNIX developer and system integrator (among many other operating systems). The very existence of this forum is an example of their connection with users... and the fact that these support and collaboration mechanisms can always be improved does not undermine their intrinsic value.

In other words, please show a bit of respect...

Message was edited by: batondor
Reply
User profile for user: batondor
batondor Author
User level: Level 1
0 points

May 5, 2010 12:28 PM in response to R C-R

There are three good answers to your question:

1) Anyone experiencing the same problems with 10.6.3 who might wish to attempt the fix themselves...

2) Anyone who needed convincing that there was an intrinsic system-related problem with the 10.6.3 update on some otherwise viable hardware configurations...

3) Anyone who was trying to understand the origins of the problem (though I accept this is particularly for Apple support people who might not, in fact, peruse these forums... but how about professional endusers who might be experiencing similar issues with this update...)

I'm perfectly willing to admit, R C-R, that I would drop the matter if I had received any feedback whatsoever from the Apple Store geniuses, the Apple Expert with whom I spoke, or a support entity directly... and while I am not surprised I have not received such feedback, I still do not look forward to the next "update" that might be recommended for other reasons...

Your dismissiveness notwithstanding, I think it's reasonable to bring temporary closure to the matter from my perspective, which is all I was attempting to do.
Reply
User profile for user: alemao_locao
alemao_locao
User level: Level 1
0 points

May 5, 2010 12:52 PM in response to batondor

Hi.

I'm also having the same problem. After 10.6.3, when i'm changing through spaces (via cmd +1/2/3/4), the space indicator stays on the desktop, even releasing the keys. Them my mouse works, but i cannot click. My system do not freeze, if you wait about 15 seconds, it back to work normally. If you are playing itunes in the background, you can still listen them when it 'freezes'. I'm getting into this problem maybe 5-10 times a day and i don't know what's causing it.
Reply
User profile for user: R C-R
R C-R
User level: Level 6
17,782 points

May 5, 2010 1:24 PM in response to batondor

How would anyone else know they have the same problem(s)? If they thought they did, how would what you have posted help them fix that themselves? Apple warns users not to mix together parts of different OS versions. It isn't a solution. It is a workaround that masks the real issue & often causes other ones, especially with full functionality & future updates.

At best, it is the method of last resort, after everything else has been tried. If you convince users that there is an intrinsic problem with the update, they won't try everything else; in fact, there is a good chance they won't try anything, not even applying the update in the first place.

Your experiences are not representative of anyone's but your own. There is no reason to believe they are typical.
Reply
User profile for user: batondor
batondor Author
User level: Level 1
0 points

May 5, 2010 3:50 PM in response to R C-R

RC-R...

If I thought you were commenting in good faith I would not be so skeptical of your arguments... which are tortured, at best... so I will ask you a simple question:

Are you suggesting that my experience in being able to reproduce the "freezing" problem with 10.6.3 v1.1 on a stock iMac that had been checked out at the Apple Store and that has a fresh erase/install is somehow not a generic issue of some sort?

For what it's worth, I agree with your observation that in most cases these problems arise when people add third-party software, especially that which effects the kernel or system-level functionality... and I also agree that many users refuse to believe this is the case (and yes, I've had clients who swore their graphics applications were faulty when, in fact, their display was fried or even just unplugged...).

I also agree, to some extent, that steps like the replacement of the 10.6.3 ATI X2000 driver with the one from 10.6.2 is both a workaround and is not for the fainthearted... but I know enough about the driver integration model on Mac OS X to have attempted it without too much worry (and I am fortunate enough to know an ex-Apple graphics systems developer from the original NextStep era to make the test with an additional vote of confidence).

This is what makes this particular situation unusual... but not worthy of being swept under the rug or underplayed, in my opinion.

Message was edited by: batondor
Reply
User profile for user: R C-R
R C-R
User level: Level 6
17,782 points

May 5, 2010 9:55 PM in response to batondor

batondor wrote:
Are you suggesting that my experience in being able to reproduce the "freezing" problem with 10.6.3 v1.1 on a stock iMac that had been checked out at the Apple Store and that has a fresh erase/install is somehow not a generic issue of some sort?


I'm suggesting that a generic issue would of necessity affect some entire class of iMacs. Apple sold 3 million Macs last quarter alone. iMacs are the first or second best selling line. No matter how I slice it, I come up with at least several hundred thousand recent vintage iMacs in use, probably many more.

I do not doubt you can reproduce the freezing problem on yours, & that there are others that can do the same. However, the the elephant in the room is the relative paucity of reports compared to the size of the user base. If this was truly a widespread issue reports would be flooding in, it would receive a lot of attention on Mac oriented web sites, & so on.

Moreover, these same contentions come up after every update, as do the downgrade kernel extension workarounds, & all the rest of it, including whatever you might be tempted to say to explain the lack of more reports or coverage. It is such a regular occurrence, & the contentions so similar, that we could save time & effort by just publishing a numbered list of them so users could refer to them by number instead of repeating what are often almost word for word the same arguments used against previous updates.

It's not that I'm not acting in good faith. It is that I (& many others) have been through this so many times before, looking in good faith for some credible evidence that any of these problems really are generic but finding none. There are most certainly real problems that defy simple explanation, just not enough of them to conclude that they are common or typical of some large class of the user base.

If we can agree on that much then I'm done.
Reply
User profile for user: sysvr4
sysvr4
User level: Level 1
0 points

May 22, 2010 8:43 AM in response to R C-R

I'm suggesting that a generic issue would of necessity affect some entire class of iMacs. Apple sold 3 million Macs last quarter alone. iMacs are the first or second best selling line. No matter how I slice it, I come up with at least several hundred thousand recent vintage iMacs in use, probably many more.


Then I'm suggesting that you're not trying hard enough to envision a scenario which would not affect every system in the same manner. Just to name one, albeit a big one, you're aware there are often hardware revisions within subsystems of any given machine, yes?

I do not doubt you can reproduce the freezing problem on yours, & that there are others that can do the same. However, the the elephant in the room is the relative paucity of reports compared to the size of the user base. If this was truly a widespread issue reports would be flooding in, it would receive a lot of attention on Mac oriented web sites, & so on.


Your tired argument of "it would affect everyone", which has no basis in fact, and the "paucity of reports" aside, I'd love to hear your analysis of how the OP's system, hardware checked by Apple and no third-party software, is in ANY way indicative of a problem stemming from outside of Apple's development.

Somehow I think the conclusions you've so obviously reached would differ significantly if one of your several Macs were experiencing this problem...
Reply
User profile for user: R C-R
R C-R
User level: Level 6
17,782 points

May 22, 2010 11:39 AM in response to sysvr4

sysvr4 wrote:
Then I'm suggesting that you're not trying hard enough to envision a scenario which would not affect every system in the same manner. Just to name one, albeit a big one, you're aware there are often hardware revisions within subsystems of any given machine, yes?


Sure, but it is that "albeit a big one" qualifier that makes it hard to accept any scenario that blames all the freezing problems on the update. It isn't as if there are dozens of revisions among the tens of thousands of Macs of any particular model in use, but more importantly even if there were, if one or a few revisions were not compatible with 10.6.3, why blame the update instead of the revision?

... I'd love to hear your analysis of how the OP's system, hardware checked by Apple and no third-party software, is in ANY way indicative of a problem stemming from outside of Apple's development.


Perhaps this system has developed a hardware defect that is hard to diagnose. Maybe several components are near the limits of the tolerances for 'in spec' parts, individually still within spec but together adding up to a marginal system. Throughout this discussion some have assumed that a single overnight Apple Store checkout of this system would conclusively have found any hardware issue it had, but that is by no means a safe assumption:

In the first place, batondor said the Apple Store did not check it out with an external monitor connected, that this was the way he used it, that the problem returned about 24 hours after he got it home & attached it, & that it now continues to occur without the external monitor attached. This is not inconsistent with a deteriorating part or subsystem, probably but not necessarily in the graphics subsystem. The system was never returned to the Apple Store for further testing, so it is hard to say what if anything would be discovered now.

Moreover, an "Apple Expert" (quoted here because there is no such job description at Apple with that exact title) told him that "the upgrade is not stable yet on some configurations" & he concluded from this oddly vague statement that his must be one of those configurations.

Don't misunderstand me. I'm not saying this is the cause of the problem with his system. But I am saying that if it were my system, I would talk with the store manager at the Apple Store where the testing was done, repeat what Applecare said, ask for clarification, & depending on what the manager said, demand that the system be checked out more thoroughly.

Basically, if I knew that other similar systems were running 10.6.3 without problems, I would want Apple to explain to me exactly why mine could not.
Reply
User profile for user: batondor
batondor Author
User level: Level 1
0 points

May 23, 2010 6:58 PM in response to R C-R

R C-R...

You really won't let up, will you?

(1) There is such a thing as an Apple Expert as indicated here (and yes, this is where I called for advice after the Apple Store failed to find a solution to the problems on my iMac):

http://www.apple.com/support/expert/

(2) The problem in question always occurred on my iMac both with and without an external monitor adapter connected, and you are mistaken in your concrete presumption that there are other tests that could have been performed at the Apple Store because they assured me that their test harness in the "back room" did a stress test on all components... and I was not instructed to return it by the "Expert" at a higher level with whom I spoke by telephone because I was told that if they wanted the machine to test they would make that request directly (and even pay for the shipping, etc.).

(3) While I have taken every step imaginable suggested by Apple, including the offer to return to the Apple Store for more tests, I am simply being a bit patient to see if their tests based on the bug report that I submitted and that they acknowledged leads to a further request on their part (and I called again at the end of last week to see if anything was pending). I fully accept the proposition that my system might have a hardware fault that was heretofore unknown and undetectable or, as I suspect is more likely, it is an unexpected performance-related limitation in the graphics hardware pipeline that is not consistent across all machines and that was triggered by the 10.6.3 upgrade of the OpenGL and/or Core Graphics libraries... and I would be prepared to bet from experience that the latter explanation is at work, that it affects many models other than my own, and that Apple will find a way to mitigate the effect in a generic fashion that will appear with a subsequent upgrade.

For what it's worth, I am perfectly ready to accept that this problem is unusual and that it may not effect all machines of a version that is now running 10.6.3... but I really disapprove of your lurking in this thread with belated critiques that are patently untrue. Yes, I could go back to the Apple Store and berate the technicians there who are doing their jobs... and could go from one Apple Store to another hoping to find an exceptional capable "Genius" who can do what the others were unable to do... but I don't have that kind of time. I'd rather be a bit more patient and trust that they take the problem seriously... as I was told.

Message was edited by: batondor

Message was edited by: batondor
Reply
User profile for user: batondor
batondor Author
User level: Level 1
0 points

May 23, 2010 7:33 PM in response to R C-R

I forgot to add that nobody ever told me that an upgrade was actually available as per your statement...

I am not going to go over all these posts to find something that I am reasonably certain that I never said, so if you can point to the one in question where you found this statement, please offer a link...
Reply
User profile for user: R C-R
R C-R
User level: Level 6
17,782 points

May 23, 2010 10:15 PM in response to batondor

1. Do you understand that my last reply was to sysvr4's specific question about how your problem could possibly be in some way not indicative of a problem with the 10.6.3 update?

2. "Apple Expert" is not a job title. You talk with Apple Customer Relations Representatives, whose primary responsibility is to resolve customer issues, which may include escalating issues in Apple's words "upward and outward" to engineering, product design, or other departments. These reps may not even be Apple experts in the strict sense of the term. (You can get an idea of the job skills Apple is looking for for these positions by browsing the job listings at http://www.apple.com/jobs/us/corporate.html#applecare, for instance this recent listing for a Portuguese Speaking Customer Rep.) Accordingly, do not assume they are at a "higher level" than an Apple Store employee.

3. Don't assume that an Apple Store "test harness" can find any & every hardware problem. Obviously, it did not find whatever caused your Mac to misbehave soon after you got it back & connected your external monitor, as you described in the 'Spoke too soon' topic. Whether hardware or software was the cause, once a Mac appears to be working correctly, no responsible tech will continue prolonged stress tests because a) they understand that customers want their working Mac back as soon as possible & b) prolonged stress tests (as you might expect) shorten product life.

4. You said in one paragraph that you weren't instructed to return the Mac for more testing but in another that there was an offer to return it to the Apple Store for more tests. Since your issue now seems repeatable & immediately obvious with or without an external monitor attached, it would be a good idea to take Apple up on that offer.
Reply

Mac OS X 10.6.3 system "freezing" problems: my experiences and conclusions

Welcome to Apple Support Community
A forum where Apple customers help each other with their products. Get started with your Apple ID.
Learn more Sign up