"Reserved Range Exhausted", what does it mean??
 

OK, I normally don't post things until I have exhausted
all possibilities and looked everywhere to solve the problem, but I give up!

I got a new dual 1.25 GHz PowerPC, with lots of RAM (1.5 G)
and a dual video card (10.2.6). Nice machine except for
lots and lots of messages like this that show up in the
Console log:

After a few hours and a few hundred of these, the system starts
getting really slow, lots of spinning balls, doesn't wake up
from sleep and a few hard crashes. I looked around and
several folks have been having this problem so here are
the hypothesized causes and what I have tried (none
of the worked for me):

1. Corrupted fonts: several people reported this to be
the cause either in your local ~/Library/Fonts folder or
in the /System/Library/Fonts or /Library/Fonts. Using font
repair diagnostics (like FontDoctor) takes care of the problem;
other people just move half the fonts to their desktop,
see if the problem persists, then mover the half, etc..until
you narrow the culprit by elimination. I did both of these
with no luck..Fonts don't seem to be the problem..

2. Corrupted Font Cache: Utilities like Jaguar Cache
Cleaner remove cache font files that might be corrupted.
I used it and nope...no avail..

3. Corrupted pref plist files. I found that the console messages
would appear right after opening any of the Microsoft office
apps when no other apps where opened. I removed all the
Microsoft pref files and although the messages did not
appear after opening MS apps right from the bat, they did
appear later when other apps where opened (see below).
Additionally, the problem persists when other users are
logged in (so it doesn't seem to be a user specific issue).

The only consistent pattern I have found is that the console
messages start popping after a few apps are opened
(usually 8-10) after login. It doesn't matter which apps
are or on what order they are opened; the only thing
that matters is the number of apps opened. I checked
memory (extensive Hardware test) to see if that could be
the culprit; no issues found there. I also re-installed the
system to make sure nothing was wrong there..nada.

So I give up! The only other thing I can think off is corrupted
cache memory, but haven't heard that ever mentioned in
any forums..Anyway does anybody have any ideas??

And BTW what do these messages mean exactly? What does
the WindowServer do?

Thanks,

ateles

Re: "Reserved Range Exhausted", what does it mean??
 

I gave up too. After trying most of the same procedures you did, and having much the same substantive experience of this error (tho no hard crashes that I would attribute to this problem), I decided that nobody knew the answer. So, in the hope that there is someone out there who can shed some light on this, I've added to this thread.

I wonder if it's got something to do with dual video.. Seems zed has(d?) a similar problem with a similar set up:
http://forums.macosxhints.com/showth...ange+exhausted

Does the Geforce4MX that I use count as dual video? I know it has two outputs but I've only ever used one of them.

Yes, I have a ATY card. The problem persists with either one or two monitors connected to it, so I wonder if its the card memory (I have 64 MB). But again, the exyensive hardware tests found nothing..

ateles

Interesting.. I get this error when I run Microsloth Word.

Jun 4 09:15:33 null WindowServer[206]: Reserved range exhausted. (0xbbc94000 to 0xbc16c000 goes out of bounds)

May I assume that you both have Microslow Office X as well? (Hopefully zed will read this and reply as well.)

Hi,

Thanks for the ping...

Yes I do have Microsoft office X..

My crashes seem to have dissipated for now...

I'm on 10.2.6 with dual screen...

---Zed :cool:

Upon reflection, I think most video cards sold now-a-days support multiple video, so that's probably not an issue.
Microflacid making a buggy piece of software.. hell no that's not possible! :D

As far as I know the WindowServer is the Aqua window server.. what makes the OS X GUI look like it does.

I've got similar messages appearing in my console log. I've had lots of problems with Codetek Virtual Desktop today, where it stops working normally. When I try restarting the program, the entire system hangs. The only thing I can do is hold down the power button and force it to shut down. I suspect that the culprit may be the Quicktime 6.3 update, since everything was working normally before that.

I get "reserve range exhausted " errors daily; they seem to appear in clusters. No functional problems, just lots of error messages.

I've recently noticed that when I start getting lots of these, I've generally been using a Carbon app. Do Carbon apps interact differently with the Window Server than Cocoa apps? Don't know if there's a correlation, just trying find a trend in these errors.

I only get them when using Word.

i don't get them using Micro$loth Word, or any of the office suet.

Merv, do you get them at all? If not, microshaft products may still be a factor.

hmmm.. what version of Office? Default install or add-ons (Equation Ed, EndNote)?

Stranger.. I don't get it on my office, machine, only my home machine.

no, i don't get "Reserved range exhausted" messages in the logs, at all.

full office install, v10.1.1 is the latest i have for word and entourage, i believe the result of the "Office X 10.1.2 Update" of 31 oct 2002

Microsoft products definitely had something to do with it, because when I deleted all MS Office preferences the frequency of messages decreased. It might have to do w/Carbon apps, but I don't get the messages in my Ti Laptop were I also use MS Office on a regular basis...
However, I have not tried checking for messages when there using only cocoa apps...I'll try that and let peple know.

ateles

Ditto that on my dual-display setup, stock 64MB GeForce 4 MX card.

Forgot to mention I'm not using *any* MS apps on my system, yet still get the errors. Have yet to pin down any particualr apps as causing the errors though.

I also recieve many of these messages in the log from time to time. I would attribute this error to not locking on to the drawing area as it should be done when drawing in quart, just an idea. In reality there is no way for us to fix it since this is a bug in the actual program code, and most probable than not the code is not accessible to us. I wouldn't doubt that the Finder itself was causing this problem.
Ateles, the windowserver is the main drawing engine for the MacOSX desktop environment. Applications which want to draw to the screen send notifications through network-like ports to the windowserver program. This in turn displays whatever it is you are asking it to draw. The windowserver is what in UNIX is called a window manager.
AtomicTuesday

Speaking about the windowserver, does anyone know how it is that a system can actually draw to screen? My guess is that firmware level calls have to be made directly to the videocard which in turn sends the information to the monitor. Is this statement right at all? Another thing, when I opened the firmware device for the video card in a G4 iMac and displayed the words (does anyone know what I am talking about here) available for this device, a very interesting word had to do with characters. I suppose that this is used for displaying characters to screen. I was thinking, the kernel doesn't probably draw the actual words to screen but rather calls a firmware call to display a character. Does anyone know if this is how it happens?
AtomicTuesday

Ok. Interesting qquestion, but lets keep focused here...thanks..
So I have tried a number of other things, but none have worked:

1. Reset PRAM
2. Work with Cocoa apps only
3. Re-nstall the video card driver - ATI Raedon 9000 Pro..although the driver for this card is not supposed to be used if you have later than 10.2.2 (I have 10.2.6). I wonder...
4. Re-order the memory sticks (512,512,256).

I am beggining to think what some people have suggested; this might be a OS glitch..

ateles

Personally I think this is a non-issue. I always remember seeing some kind of complaint of the windowserver in the logs. Now I'm getting:

WindowServer[175]: orderWindow: Invalid other window
WindowServer[175]: CGXOrderWinsAndPanels: error ordering window

And in the past I have seen:

WindowServer[175]: CGXDisableUpdate: Updates disabled
by connection 0x15e77 for over 1.000000 seconds

If you're getting any crashes or strange behaviour I'm not sure it's the windowserver, seeing that others have it in their logs without il effects.

v

vonleigh,

These messages are OK and I get them at my PB all the time. But an excesive number of "Reserved range exhausted" messages, is a signal something is not right...

I think more and more that it has to do with the video card, since most people that have reported the problem have the same ATI card as I do..

Maybe ATI has not updated its drivers for OS 10.2.6??

FWIW, I get this message on an nVidia, and not on an ATI.

Quartz Extreme may be a little too extreme.

I don't have access to my logs right now, but it would be interesting to see if it's related.

I think vonliegh is right, the window buffer might be hitting a brick wall, does redraw look painfully slow and ragged when you get a plethora of these messages? My tuppence.

For what its worth, Codetek issued an update yesterday for Virtual Desktop. I haven't had it running for long, but thus far there haven't been any problems and OS X certainly didn't crash.

:)

I can produce the Reserved range exhausted console message by clicking and rolling over the menus in Camino.

Opening a new window in the Finder can also cause the messages.

I get these messages also in nearly every program. Every second or fourth click in the menu bar produces this error in console.

UPDATE:
I did a little testing on this. I had my display set to 1280x960. I switched it to 1152x960 and the messages disappeared! Tested for 5 minutes and decided to switch back. Tested for 5 minutes. No messages at all. Then I did a restart. For 10 minutes everything was fine but then I got 16 messages in 10 seconds. The next 20 minutes I got only 1 mesage.
So this seems to be related somehow.

I get these errors all the time as well. I thought it was because I modified my shared memory settings to run oracle. I guess not. I don't have an ms products on my machine so that can't be the reason. I have a G4 with dual 1.25GHZ processors and 1.2GB of RAM. Here are the results of these messages on my machine.

[username@osx /]$ uptime
7:20PM up 2 days, 19:29, 3 users, load averages: 0.00, 0.02, 0.00
[username@osx /]$ grep "Reserved range exhausted" /var/tmp/console.log | wc -l
771

As you can see, this is happening A LOT. I can't say a lot of apps crash or I have any problems. But this problem is a bit disconcerting....

J

oleaf: I also notice a decrease in the number of messages after changing the resolution, but only in the beggining. After a few hours of working on the machine, they were everywhere especially after launching new apps or saving documents...

JavaOS X: Interestingly the messages only appear in the newest Powermacs (specially dual). I repeated your commands in My Powerbook G4 (15'') and here are the results:

Not a single one in my PB! Strange ahh? But what I have noticed in my Power Mac is that crashes have vanished, (maybe because of all the cache cleaning, renewal of preference files, etc..) although the number of "Reserved Range.." messages is quite high..Anyway..

ateles

That's interesting, so that begs the question, is anyone here having this problem with a single processor machine?

J

Yes, I am.

... and have done for some time.

Same here. Sometimes I get lots, sometimes I get very few.

Note, about 5 minutes later, 3x as many. This was reproducable only by opening 2 MS Office apps together.

After a frustrating couple of days where the 'reserved range exhausted' issue finally brought my machine to its knees (crashes, apps not opening ... and so on) I decided to do a clean install of the OS. On the assumption that 10.2.6 might be implicated I'm currently at 10.2.5 and haven't updated QT yet. However, one of the early things I reinstalled was Little Snitch and immediately started getting the 'reserved range exhausted' messages in my system log. So, of those of you who are experiencing this problem are there any of you who don't have Little Snitch installed?

Having said that, my machine hasn't been up long since I removed Little Snitch so I may well be wrong about this.

No little snitch here :(

I spoke to soon. I removed Little Snitch and still have the damn things. Which, given that I've just gone through (or, more accurately am part way through) a complete, clean install, is a bit of a bugger.

However, it means we can probably stop trying to fix this as I suspect that it was a system bug introduced in 10.2.5 (or maybe earlier) :confused: :( :mad:

The only thing I turned up in my search for this was

But there isn't source available for this, that I know of. So it's dead end unless someone feels like reverse engineering it.

J

Yep, I agree, though one final thing that might be worth reporting is that since I reinstalled my entire system I'm still getting a slew of these errors:

The above only three hours after a reboot.

But, I didn't reinstall any of the Unsanity haxies (or their Application Enhancer) and my system, despite all the reserved range errors, seems way more stable. So, for anyone still suffering with this problem it might be worth uninstalling any Unsanity products to see whether it makes any difference.

nVidia cards are part of the problem
 

To wind up this issue: I spoke to someone I know who works for the Apple Seed Programme and asked him if he knew anything about this problem and if it could cause any serious problems. This is what he had to say:

I just checked our bug database on this, and, in general, it could end up crashing applications, but usually wouldn't (programatically, it could happen by sending AppleEvents with references it shouldn't - that could happen with drag & drop/copying things into the pasteboard, etc). Reading through the engineering notes- they believe that NVidia graphics cards are 'burning up most of shared address space'. There are things developers can do to reduce the occurrences of this happening due to their applications. The good thing is that Panther changes some underlying core graphics routines which seems to alleviate this particular error (among others).

When I asked him if I could quote his message here and if he knew why this seemed to be seemed to be such a problem with recent versions of 10.2 he replied:

Sure- I'm not sure if it will actually help people to know that it's (probably) an nVidia specific problem, but certainly, you can give them hope that Apple has looked into it, and Panther should improve on this. Looking at the bugs, it looks like there are reports on it since 10.2, and not just recently- perhaps your activities have changed somewhat recently so you see it more? or maybe some specific applications are causing this to happen more frequently.

So, that’s that I guess, other than I might try and do a bit more troubleshooting to see exactly which apps throw up this error more than others (which I suspect may be a futile exercise). So it’s either wait for Panther in the hope it fixes this issue, or fork out for an equivalent card (e.g. the Radeon 9000). Panther will probably be the cheaper option so I think I’ll probably try and wait for that. Having said that it’s worth emphasising a part of the comment above: “the good thing is that Panther changes some underlying core graphics routines which seems to alleviate this particular error”. Note that the word used is “alleviate”, not “eradicate” (or it’s equivalent).

On a more general note, I think that Apple have had problems with the nVidia cards (and from the above it seems as though they still have). When I first got my machine I had problems with the 4MX and weird glitches on my monitor – this was later resolved by Apple as a problem with the 10.1 drivers for the card. There were also problems with the 10.1 driver for the 4Ti. I would have included a link to the Apple board discussions but they both seem to have been deleted.

<edit>
duh ... I've just read back through this thread and realised that at least some of the people experiencing this problem are using ATI cards. Hmmm ... I'll get back to the guy I know at Apple.
</edit>

I am using an ATI mach64 3DU card, I still get these errors on my log. I'm not sure if this is really a hardware specific issue as more of a software issue. Maybe it is a bug common to all their graphics drivers for both ATI and NVidia.

Anyone know if Apple distributes the code for their graphics drivers?

Well my system has frozen twice today...

sometimes it goes weeks without a freeze... and then only a few hours!!

:( :( :(

--Zed :cool:

I guess the obvious question would be what are you doing that's different? But I suspect that the answer is nothing - at least my experience of trying to troubleshoot these errors is that there seems to be no rhyme or reason to them or what impact they have on the system. I'm still waiting on my contact at Apple to get back to me as to whether there's any more information on this.

<edit/afterthought>

In an effort to monitor these errors I've written a Konfabulator widget that puts the total number of errors on the desktop. If anyone would like a copy, drop me a line.

Your right... nothing different...

the last few freezes all I've had running is:

Mulberry - mail client
Terminal
Fire - IM message client
iChatAV
Safari
iCal
AddressBook
Stickies
ICA client
ITV - TV card viewer..

nothing too exciting...

Cheers,
--Zed :cool:

What Unsanity stuff do you have loaded?

nothing...

I've stayed away from anything too doggy...

I think it might be memory related... I never shutdown my system and I never even logout..

I do monitor it using top but so far not seen anything that looks odd :-(

---Zed :cool:

Oh well, so much for that hunch. My own experience is that the errors do increase over time and do seem to occur more readily when you get down to relatively small amounts of free memory. Having said which, the number of errors and the amount of free memory are both time related and there's no real reason to suppose that one can cause the other.

I've just had a thought...

My G4 2x1Ghz Quicksilver which does freeze has been upgraded from 10.1..

Were as my G4 powerbook 12" was clean installed with 10.2 and does not freeze

Anyone else have the same, ie the freezing system is upgraded? I'm wondering if it could be an artifact from this upgrade route..

Cheers,
--Zed :cool:

My PowerMac which has a clean install of Jaguar, shows the errors all the time...Looking at some messages back, I think it is going to be graphics-card related (although I have an ATI Raedon, not NVIDIA, as suggested to be the culprit)...Well, lets wait for Panther..

ateles

Well, I think I finally found the problem with my crashes, and as some of you suggested it had nothing to do with the "Reserved Range Exhausted" message. I still get that a lot of them, but since I removed one of my 256 MB dimms the crashes have dissapear. General conclusion: The RRE messages do not seem to be of major concern and as some have pointed out there might be some problems with certain graphic cards in certain systems that might contribute to it, but they don't neccesarily bring the system down once every day...bad memory does!

thanks to everybody who contributed to the discussion.

ateles

I get them ALL THE TIME on my 2x1.25Ghz G4.

No pattern I've been able to find, and I'm not sure they really hurt anything.

Mike

I get them on a 2x142 with two monitors and a single processor laptop, but the former more frequently. They show up in 10.2.4 and 10.2.6 with clean installs and no 3rd party hardware.

I have been on to Apple support about it (****-hot, expensive extended support at that), but no luck yet. Getting a different support rep each time who begins by asking me to do quite simple things that a child of four months would already have done doesn't help. I would posit that with the impending release of 10.3, curing such problems in 10.2.x is low priority.

I would suspect that if the console is actually reporting them, the situation isn't as dangerous as if memory writes were stepping out of bounds and not being detected. I don't like it though.

They do seem to be related to memory reservations for new windows. Simple test: open Textedit and make 150 or so new windows. See how many Reserved Range Exhausted messages you get. Now, close all those windows. Not a one.

Me neither, I've currently been up for just under 12 hours and have 79 of them. Before I rebooted (with an uptime of five days - damn that kp) I had approaching 1500. Anything that generates that many errors, while obviously not immediately fatal, certainly can't be helping things tick along smoothly.

Well here is a definitive from Ars Technica:

http://arstechnica.infopop.net/OpenT...1&m=9670962145

So it seems that this is a warning rather than an error, as I suggested previously, and it's not a worry.

Still irritating though.

Thanks for that. And yes, it is irritating - I use Desktop Console to tail my system log and reading hundreds upon hundreds of these entries is really rather tedious. All we need now is for someone to patch the windowserver to stop it generating these errors ;)

Yes, it's very annoying, I grep it out:

Let's hope this thread dies now too : )

v

It might be better to use:
... that way you don't dump all the WindowServer messages. Not that any of them seem all that interesting ...

Well, the windowserver is whiny:

"orderWindow: Invalid other window" and also "error ordering window". After I while I just decided to ignore the whole process. It's never really given me an interesting or useful error message anyway.

Plus no point in setting such a large n if I keep the window open always, a screenful is enough.

v

I guess you're probably right - perhaps I'm just living in hope that one day it will say something interesting ;)

I was hoping that this might be fixed in 10.2.8, but no - 102 'reserved range exhausted' errors with an uptime of 2 hours and 42 mins. Oh well :(

It will take a major overhaul of the WindowServer to actually fix the error, but it could be glossed over by reducing its log level or hushing its output a bit.

I still don't have my copy of Panther, but I was just curious to know if somebody has seen the infamous "Reserved Range Exhausted " messages in the console? I hoped Apple fixed this!

ateles

Mine vanished completely. :D
I did update from 10.2, no clean install.