5.0.4 Performance Issue (Similar to D3 Beta)

Mac Technical Support
Prev 1 12 13 14 44 Next
So unless I missed a posting or thread, its been about 24 hours since Blizz has responded to this issue. This is not cool. It takes less than a minute to go on make a thread and say "We are looking into the issue, but have not found out what that is." or even "Fck you Mac Client consumers, deal with and buy yourself a PC."

>.<

Hopefully it would be the first, but let us know something. If I did miss something I apologize and I am sorry, but if there is in fact nothing posted than I'm a little more than frustrated.
Growltiger:

ps uM results:

USER PID %CPU %MEM VSZ RSS TT STAT STARTED TIME COMMAND PRI STIME UTIME
75657 33.0 2.8 5118920 203120 ?? S 10:36PM 0:42.14 /Volume 46T 0:01.11 0:06.67
75657 0.0 2.8 5118920 203120 S 10:36PM 0:42.14 48T 0:00.00 0:00.00
75657 0.0 2.8 5118920 203120 S 10:36PM 0:42.14 46T 0:00.00 0:00.00
75657 60.0 2.8 5118920 203120 R 10:36PM 0:42.14 46T 0:02.39 0:08.75
75657 0.0 2.8 5118920 203120 S 10:36PM 0:42.14 46T 0:00.00 0:00.00
75657 0.0 2.8 5118920 203120 S 10:36PM 0:42.14 46T 0:00.00 0:00.00
75657 0.0 2.8 5118920 203120 S 10:36PM 0:42.14 46T 0:00.00 0:00.00
75657 99.1 2.8 5118920 203120 R 10:36PM 0:42.14 46T 0:00.05 0:22.53
75657 0.0 2.8 5118920 203120 S 10:36PM 0:42.14 62T 0:00.02 0:00.15
75657 0.0 2.8 5118920 203120 S 10:36PM 0:42.14 46T 0:00.00 0:00.00
75657 0.0 2.8 5118920 203120 S 10:36PM 0:42.14 46T 0:00.00 0:00.00
75657 0.0 2.8 5118920 203120 S 10:36PM 0:42.14 46T 0:00.00 0:00.00
75657 0.0 2.8 5118920 203120 S 10:36PM 0:42.14 46T 0:00.04 0:00.15
75657 0.0 2.8 5118920 203120 S 10:36PM 0:42.14 46T 0:00.00 0:00.00
75657 0.0 2.8 5118920 203120 S 10:36PM 0:42.14 46T 0:00.00 0:00.00
75657 0.6 2.8 5118920 203120 S 10:36PM 0:42.14 46T 0:00.08 0:00.04
75657 0.6 2.8 5118920 203120 S 10:36PM 0:42.14 97R 0:00.06 0:00.07
75657 0.0 2.8 5118920 203120 S 10:36PM 0:42.14 62R 0:00.01 0:00.01
75657 0.0 2.8 5118920 203120 S 10:36PM 0:42.14 46T 0:00.00 0:00.00
75657 0.0 2.8 5118920 203120 S 10:36PM 0:42.14 46T 0:00.00 0:00.00
75657 0.0 2.8 5118920 203120 S 10:36PM 0:42.14 46T 0:00.00 0:00.00


info threads results:

(gdb) info threads
21 0x00007fff85701a2a in __workq_kernreturn ()
20 0x00007fff8572b932 in select$DARWIN_EXTSN ()
19 0x00007fff85722a6a in __semwait_signal ()
18 0x00007fff856e7e82 in mach_wait_until ()
17 0x00007fff856e7dda in semaphore_timedwait_signal_trap ()
16 0x00007fff85722a6a in __semwait_signal ()
15 "BlizzardCore Thread" 0x00007fff85722a6a in __semwait_signal ()
14 0x00007fff85722a6a in __semwait_signal ()
13 0x00007fff85722a6a in __semwait_signal ()
12 "BlizzardCore Thread" 0x00007fff8579ab72 in sem_wait ()
11 "BlizzardCore Thread" 0x00007fff8579ab72 in sem_wait ()
10 "com.apple.CFSocket.private" 0x00007fff8572b932 in select$DARWIN_EXTSN ()
9 0x00007fff856e7d7a in mach_msg_trap ()
8 0x00007fff82bbef5f in _class_hasCxxStructorsNoSuper ()
7 "Fatal Exception Handler" 0x00007fff856e7d7a in mach_msg_trap ()
6 "Async Signal Handler" 0x00007fff85722a6a in __semwait_signal ()
5 "Exception Handler" 0x00007fff856e7d7a in mach_msg_trap ()
4 "GLWorker Thread" 0x00007fff856e7d7a in mach_msg_trap ()
3 0x00007fff856e7d7a in mach_msg_trap ()
2 "com.apple.libdispatch-manager" 0x00007fff85700c0a in kevent ()
* 1 "com.apple.main-thread" 0x000000010083cd5a in AssertAndCrash ()


thread where results:

Thread 8 (process 75657):
Thread 8 has current state "WAITING"
Mach port #0x6703 (gdb port #0x1b03)
frame 0: 0x00007fff82bbef5f in _class_hasCxxStructorsNoSuper ()
pthread ID: 0x115f81000
system-wide unique thread id: 0x66f0f
total user time: 765184008
total system time: 204765000
scaled cpu usage percentage: 0
scheduling policy in effect: 0x1
run state: 0x3 (WAITING)
flags: 0x1 (SWAPPED)
number of seconds that thread has slept: 0
current priority: 46
max priority: 63
suspend count: 0.
08/29/2012 10:40 PMPosted by Oddtech
So unless I missed a posting or thread, its been about 24 hours since Blizz has responded to this issue. This is not cool.


Blizzard's lack of response to threads is usually a good sign. It means that they're past the info gathering phase of development, and they're hard at work on it. A patch or fix is probably well on it's way.
Oh heck, I missed seeing the post where you said you got it working, lol, nvm!

But if anyone else is having a problem getting it to work, maybe that will help you figure it out, or help someone else help you figure it out.

I don't really want to get any further into analyzing the problem thread, that's Blizzard's job, they have the source code. It should be quite easy to figure out what it's doing wrong and fix it. Only problem really seems to be convincing Blizzard it IS a problem.
I'd say heating up a Mac Pro's anodized aluminum chassis to the point you can't touch it without being burned is a good indication of a problem. :D
I can confirm that the steps Growltiger provided worked perfectly for me. My macbook pro no longer overheats needlessly as soon as I log in.
I concur, my macbook does not like to melt.
Okay, so, for those of us who are almost entirely code ignorant but still overheating.

Can someone fill in the basic first nubby steps for this workaround? Because Terminal, frankly, scares me, and I'm afraid I'm going to brick my imac by meddling with it without knowing what I'm doing.

Please assume I have never opened Terminal in my life before explaining... because I haven't.
Turns out I'm experiencing the same issue, to the letter -- though for reasons that escape me, my overheating problems are no where near as severe as some of what I've read in this thread.

Thanks Growltiger for the refresher course in gdb. And, on that note, IMO: if you find Terminal and CLI tools in general scary, why on earth would you attempt it? Just sit tight and wait for the Blizzard developers to deliver a fix!
xCode is only available for 10.7+ and I'm on 10.6.8 atm, so I believe I'm out of luck for now. :/
I don't think there's a way to package this so it's simple. To use the workaround, you have to install Xcode from the App Store, then install the command-line tools from within Xcode, then know how to find the process id for WoW, start GDB, possibly list off the threads to tell which one is the "bad" one, then suspend it.

My experience was mixed. It worked a couple times (the game crashes fairly regularly now anyway, so I had to apply the fix many times), but then ceased working; I'm guessing the thread order is not static and you will have to list the threads and figure out which one is bad each time.

I then spent the next half hour blazing through Karazhan to try and reclaim my Voidheart Crown from Prince Malchezzar before discovering the Chess event right before him is bugged so as to be completely impossible, even for a full raid (Medivh cheats every round and enrages in under a minute). Found a list of known issues that includes this and about a mile long of others - I'm in awe of how many things they broke, but released it anyway. There is no QA here whatsoever.
If you have the installation disk for 10.6, it had the development tools on it, plus most of the stuff in there is GNU licensed, so should be freely available out there, just do a search.

If you're not comfortable with Terminal and using the command line, I'd suggest just being patient and waiting. I'm not even suggesting people who do have gdb, know how to use the command line, etc, use this as a fix, just as a way to verify that it solves the problem.
Confirmed that Growltiger's fix worked for me. Thus far, I haven't found the thread order jumping around.
Yeah, probably best.

I wish we could get some kind of confirm that this is being seen to, I really do. I count myself fortunate that I'm only (!!!) running at about 175º with the bugged thread going haywire, so I /can/ play. But I really don't want to do it for long stretches. I suppose it's good that I have company coming tomorrow, so maybe by the time she heads home again, this will have been seen to.
08/29/2012 11:19 PMPosted by Niernen
xCode is only available for 10.7+ and I'm on 10.6.8 atm, so I believe I'm out of luck for now. :/


You can access the older XCode downloads from the Developer Center on Apple's site (developer.apple.com). You'll need to be registered as a developer, but the general license is free. Only the pre-release software requires a paid account such as the one I have.

175°F is borderline destructive to many components. Usually the highest rated upper limit is 185°F, with a few components having 200°F ratings. So yeah, I'd wait for a fix for this.
I seriously hope that blizz is actually working on this and this will get fixed as soon as possible because i honestly cant play for more then a few minutes without my computer melting and thats really disappointing. Not everyone, including me, is able to do what Growltiger has done in order to fix this problem to be able to play the game so for the people like me which i know there is a lot of us out there we really need this to be fixed so we can go back to playing.
Thanks Lickme :)

And a huge thanks to Growltiger, again, for finding a verifiable solution to this problem.

Since we don't know the repercussions of using the method to play the game...um...someone I know just used it to play the game for 2 hours, and had no problems whatsoever. Terminal was running the whole time because *my friend* decided not to mess with it anymore, and then just shut terminal down after exiting the WoW app when finished playing.

Pretty awesome "fix", or, "verifiable solution", or whatever. You deserve a job with Blizz, Tiger, thanks a million.
Hello, I am chiming in to share my own experience with this problem. Here are my specs, for a mid-2009 MacBook Pro running OS 10.6.8:

Model Name: MacBook Pro
Model Identifier: MacBookPro5,4
Processor Name: Intel Core 2 Duo
Processor Speed: 2.53 GHz
Number Of Processors: 1
Total Number Of Cores: 2
L2 Cache: 3 MB
Memory: 4 GB
Bus Speed: 1.07 GHz
Boot ROM Version: MBP53.00AC.B03
SMC Version (system): 1.49f2
Serial Number (system): W89280R37XJ
Hardware UUID: 1DF111F5-DE04-5398-A025-6B8A3C1CDB79
Sudden Motion Sensor:
State: Enabled

Chipset Model: NVIDIA GeForce 9400M
Type: GPU
Bus: PCI
VRAM (Total): 256 MB
Vendor: NVIDIA (0x10de)
Device ID: 0x0863
Revision ID: 0x00b1
ROM Revision: 3448

I do not know very much about computers, but I downloaded SMC Fan Control like people recommended, and this is what I found:

• not running WoW: 120-130 °F -- 2000 rpm
• launcher: about the same
• log-in screen: immediate jump to 150-180+ °F -- rpm starts to climb -- 105-110% CPU
• Orgrimmar, standing around or hitting target dummies: 180-210+ °F -- 5500+ rpm -- 140-170+% CPU

All of my graphics settings are on low with foreground FPS minimized. My laptop has always run a bit hot, but not burning hot like it has been doing since the patch.
Ok, i noticed that some of the things this thread is doing sets some interlocks, which means if you suspend it while it holds a resource it can hang up the whole process.

The safest way to do this correctly is (after entering the gdb command):

thread 7
where

Now find the last line with GetBattlenetAllocator in it (sometimes there will be two). Take the number of the line above it, call that f#

type:

frame f#
finish

If you do "where" again, GetBattlenetAllocator should be shown as #0 (the last line generated by finish also shows that).

Now you can safely

thread suspend 7
quit

gdb) thread 7
[Switching to thread 7 (process 3450)]
0x932b5711 in CFBasicHashFindBucket ()
(gdb) where
#0 0x932b5711 in CFBasicHashFindBucket ()
#1 0x932dfdcd in CFSetGetValue ()
#2 0x932eaee3 in __CFRunLoopFindMode ()
#3 0x93327f0c in _CFRunLoopFinished ()
#4 0x9865aa32 in -[NSRunLoop(NSRunLoop) runMode:beforeDate:] ()
#5 0x00088599 in GetBattlenetAllocator ()
#6 0x98655318 in -[NSThread main] ()
#7 0x9865529b in __NSThread__main__ ()
#8 0x9350d557 in _pthread_start ()
#9 0x934f7cee in thread_start ()
(gdb) frame 4
#4 0x9865aa32 in -[NSRunLoop(NSRunLoop) runMode:beforeDate:] ()
(gdb) finish
Run till exit from #4 0x9865aa32 in -[NSRunLoop(NSRunLoop) runMode:beforeDate:] ()
[Switching to process 3450 thread 0x6f03]
0x00088599 in GetBattlenetAllocator ()
(gdb) where
#0 0x00088599 in GetBattlenetAllocator ()
#1 0x98655318 in -[NSThread main] ()
#2 0x9865529b in __NSThread__main__ ()
#3 0x9350d557 in _pthread_start ()
#4 0x934f7cee in thread_start ()
(gdb) thread suspend 7
(gdb) quit
The program is running. Quit anyway (and detach it)? (y or n) y
Detaching from process 3450.
Running 7 fps at the best of times.

Hardware Overview:

Model Name: MacBook Pro
Model Identifier: MacBookPro5,4
Processor Name: Intel Core 2 Duo
Processor Speed: 2.53 GHz
Number of Processors: 1
Total Number of Cores: 2
L2 Cache: 3 MB
Memory: 4 GB
Bus Speed: 1.07 GHz
Boot ROM Version: MBP53.00AC.B03
SMC Version (system): 1.49f2
Hardware UUID: FFA083B5-8D01-50CD-AFCB-7BFE34E3860E
Sudden Motion Sensor:
State: Disabled

NVIDIA GeForce 9400M:

Chipset Model: NVIDIA GeForce 9400M
Type: GPU
Bus: PCI
VRAM (Total): 256 MB
Vendor: NVIDIA (0x10de)
Device ID: 0x0863
Revision ID: 0x00b1
ROM Revision: 3448
Displays:
Color LCD:
Display Type: LCD
Resolution: 1440 x 900
Pixel Depth: 32-Bit Color (ARGB8888)
Main Display: Yes
Mirror: Off
Online: Yes
Built-In: Yes

System Software Overview:

System Version: OS X 10.8.1 (12B19)
Kernel Version: Darwin 12.1.0
Boot Volume: Macintosh HD
Boot Mode: Normal
Secure Virtual Memory: Enabled
Time since boot: 1 day6:23

Join the Conversation

Return to Forum