|
|
Is it normal for Window Manager to eat 101% CPU...?
|
|
|
|
Senior User
Join Date: Jun 2001
Location: Massachusetts, USA
Status:
Offline
|
|
Well, is it?
I'm on a duallie (1.25 GHz, even), and Window Manager is still managing to chomp 101% of my CPU... I'm not noticing a slowdown (since I still have a processor to spare), but it's a bit worrying... admittedly, I'm running quite a few programs (Mail, Proteus, OmniWeb, and Audion), but it seems odd that the Window Manager should care either way. Anyone have any clue why this might be happening, and how to fix it?
|
|
|
|
|
|
|
|
|
Professional Poster
Join Date: Apr 2001
Location: Asheville, NC
Status:
Offline
|
|
drawing graphics on the screen takes processor time. The only way we can answer whether 101% is abnormal is if we know what you are doing--because there would be situations in which that would be normal.
|
ACSA 10.4/10.3, ACTC 10.3, ACHDS 10.3
|
|
|
|
|
|
|
|
Senior User
Join Date: Jun 2001
Location: Massachusetts, USA
Status:
Offline
|
|
Sorry, should have specified that other than the things the aforementioned programs generally do (play MP3s, render some standard windows, etc.), it was idle. I guess the answer is "no", because now after a restart it's taking up only about 10% on average. Very strange.
|
|
|
|
|
|
|
|
|
Mac Elite
Join Date: Mar 2001
Location: East Texas (omg)
Status:
Offline
|
|
Originally posted by Phoenix1701:
Sorry, should have specified that other than the things the aforementioned programs generally do (play MP3s, render some standard windows, etc.), it was idle. I guess the answer is "no", because now after a restart it's taking up only about 10% on average. Very strange.
it was busy sending some "statistical info" to 17.254.0.60 ( http://littlebuddy.apple.com )
if it happens again, run 'netstat' in the terminal and you'll see that connection.
usually only sends about 20k of data and spikes the WM process briefly, but i've seen it hang at handshaking like that too. easiest thing to do is just reboot.
|
|
|
|
|
|
|
|
|
Addicted to MacNN
Join Date: Oct 2001
Location: Yokohama, Japan
Status:
Offline
|
|
Really!? This is the first I've heard of that. Can you give us any more information? Like what exactly is sent, do other processes send similar info, etc.?
|
|
|
|
|
|
|
|
|
Senior User
Join Date: Jun 2001
Location: Massachusetts, USA
Status:
Offline
|
|
"Statistical info"? Smells like Microsoft! I'm tempted to fire up a bit more than just netstat... what port's this going out on?
|
|
|
|
|
|
|
|
|
Professional Poster
Join Date: Nov 2000
Location: Norway (I eat whales)
Status:
Offline
|
|
In that case, i wonder what those 20 kb of statistic information really are..?
|
Sniffer gone old-school sig
|
|
|
|
|
|
|
|
Professional Poster
Join Date: Nov 2000
Location: Norway (I eat whales)
Status:
Offline
|
|
..
Perhaps they track down how many percent that just upgraded to jag, so they know excactly when it is about time to move over to os 10.3..
(W/full cd upgrade ofcourse.)
|
Sniffer gone old-school sig
|
|
|
|
|
|
|
|
Mac Elite
Join Date: Mar 2001
Location: East Texas (omg)
Status:
Offline
|
|
"Statistical info"? Smells like Microsoft! I'm tempted to fire up a bit more than just netstat... what port's this going out on?
It actually goes out on port 160, but the built in firewall is setup to allow it even if you specifially block it.
I haven't tried using a 3rd part firewall but I imagine that'd work.
I've portsniffed it... most of it seemed harmless (uptime, proc load avgs, etc), but some of it looked like garbarge (possibly encrypted? I dunno).
Also, if you look at the iptables.. haha... guys, i can't keep this up... i was totally joking.
|
|
|
|
|
|
|
|
|
Professional Poster
Join Date: Nov 2000
Location: Norway (I eat whales)
Status:
Offline
|
|
Originally posted by chris_h:
It actually goes out on port 160, but the built in firewall is setup to allow it even if you specifially block it.
I haven't tried using a 3rd part firewall but I imagine that'd work.
I've portsniffed it... most of it seemed harmless (uptime, proc load avgs, etc), but some of it looked like garbarge (possibly encrypted? I dunno).
Also, if you look at the iptables.. haha... guys, i can't keep this up... i was totally joking.
LOL! That's totally funny!
|
Sniffer gone old-school sig
|
|
|
|
|
|
|
|
Addicted to MacNN
Join Date: Oct 2001
Location: Yokohama, Japan
Status:
Offline
|
|
Originally posted by chris_h:
Also, if you look at the iptables.. haha... guys, i can't keep this up... i was totally joking.
Wow, you got me hook, line, and sinker! Good one.
|
|
|
|
|
|
|
|
|
Senior User
Join Date: Jun 2001
Location: Massachusetts, USA
Status:
Offline
|
|
Damn, you need to start writing fake technotes or something... you had me too!
Unfortunately, that means I have no answer to my question... and it happened again tonight; everything started getting jerky and even moving windows was slow (and this on a dual 1.25 GHz!!)... no pageouts or pageins, just Window Manager happily gobbling 100+%. I quit all apps and idled, and it didn't go down even a percentage point. So then I killed it (yes, I knew this would log me out, but I just wanted to be really sure it was dead), and it hasn't happened again since. Weird, and annoying.
|
|
|
|
|
|
|
|
|
Posting Junkie
Join Date: Mar 2001
Location: Salamanca, EspaƱa
Status:
Offline
|
|
Originally posted by chris_h:
It actually goes out on port 160, but the built in firewall is setup to allow it even if you specifially block it.
I haven't tried using a 3rd part firewall but I imagine that'd work.
I've portsniffed it... most of it seemed harmless (uptime, proc load avgs, etc), but some of it looked like garbarge (possibly encrypted? I dunno).
Also, if you look at the iptables.. haha... guys, i can't keep this up... i was totally joking.
Ya, ok chris. But what is it?
tcp4 0 0 raf-a139.raflina.52211 littlebuddy.appl.http ESTABLISHED
I get this line in Terminal netstat as you said? Am I misunderstanding?
|
I could take Sean Connery in a fight... I could definitely take him.
|
|
|
|
|
|
|
|
Mac Elite
Join Date: Sep 1999
Status:
Offline
|
|
Originally posted by Phoenix1701:
Anyone have any clue why this might be happening, and how to fix it?
The aqua window manager has several bugs which will casue it to peg the CPU even on single cpu machines. For example, opening any menu (or sub-menu) that is taller than your screen is high will peg the cpu until you close the menu. This happens in any application with tall menus whether it is carbon or cocao. As in your example, it does not seem to effect the other processes (much) since it isn't really doing any 'real work' and only consumes the 'idle' cpu cycles.
asxless in iLand
|
|
|
|
|
|
|
|
|
Mac Elite
Join Date: Mar 2001
Location: East Texas (omg)
Status:
Offline
|
|
Originally posted by voodoo:
Am I misunderstanding?
You Fool!
Delete your post, you've said too much.
|
|
|
|
|
|
|
|
|
Professional Poster
Join Date: Nov 2000
Location: Norway (I eat whales)
Status:
Offline
|
|
Originally posted by chris_h:
You Fool!
Delete your post, you've said too much.
I don't get it..
|
Sniffer gone old-school sig
|
|
|
|
|
|
|
|
Professional Poster
Join Date: Jul 2001
Location: Dis
Status:
Offline
|
|
Originally posted by Phoenix1701:
Damn, you need to start writing fake technotes or something... you had me too!
Unfortunately, that means I have no answer to my question... and it happened again tonight; everything started getting jerky and even moving windows was slow (and this on a dual 1.25 GHz!!)... no pageouts or pageins, just Window Manager happily gobbling 100+%. I quit all apps and idled, and it didn't go down even a percentage point. So then I killed it (yes, I knew this would log me out, but I just wanted to be really sure it was dead), and it hasn't happened again since. Weird, and annoying.
Perhaps this is the same problem I'm having with system freezes. I only have a 300MHz processor, and if the WM (what should be a very high priority process) ran amok, it could cause precisely the type of freeze I see on a single proc machine...
If you get a crash log from killing the window manager, send it in to Apple please, because I don't have any crash logs to give them from my episodes.
BG
|
|
|
|
|
|
|
|
|
Senior User
Join Date: Jun 2001
Location: Massachusetts, USA
Status:
Offline
|
|
Hmm, good thought there. I do, in fact, have such logs, and here's what they say...
Exception: EXC_BAD_ACCESS (0x0001)
Codes: KERN_PROTECTION_FAILURE (0x0002) at 0x00000000
Thread 0:
#0 0x992af4e8 in mach_msg_overwrite_trap
#1 0x992b4844 in mach_msg_overwrite
#2 0x936f82b8 in CGXRunOneServerPass
#3 0x936f87e0 in CGXRunOneServicesPass
#4 0x9370959c in CGXServerLoop
#5 0x93860fb8 in CGXServer
#6 0x00002dac in main
#7 0x00002b48 in _start
#8 0x000029c8 in start
Thread 1:
#0 0x99322328 in mach_msg_trap
#1 0x992b4670 in mach_msg
#2 0x9371c590 in eventThread
#3 0x992cf428 in _pthread_body
Thread 2 Crashed:
#0 0x00000000 in 0x0
#1 0x0023e5b8 in 0x23e5b8
#2 0x0023e014 in 0x23e014
...
Not terribly helpful, as far as I'm concerned, but then I don't know PPC Assembly at all well... all the zeros make me nervous though.
What is the proper avenue for reporting this to Apple? I don't remember the URL for the Radar page, if that's it.
[edit: edited to include the other threads, which might be useful.]
(
Last edited by Phoenix1701; Nov 7, 2002 at 12:20 AM.
)
|
|
|
|
|
|
|
|
|
Senior User
Join Date: Jun 2001
Location: Massachusetts, USA
Status:
Offline
|
|
I think I've figured out what it was...
ShadowDisk.
Given the choice between my disk information on my desktop or having my CPU back, um... I think I'll take my CPU, thanks.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Forum Rules
|
|
|
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts
|
HTML code is Off
|
|
|
|
|
|
|
|
|
|
|
|