Betalogue

Last night, once again, I experienced a typical problem in Mac OS X. I wanted to put my G4 MDD INTO deep sleep (the Power Supply kit has helped, but it’s still a rather noisy machine). I disconnected from the Internet (modem connection in my AirPort Base Station), waited for the modem to reset, and pushed the power button on my Apple flat-panel display, which usually does the trick.

This time, it shut down the monitors (which went black), but the computer itself stayed on, with the power button in full glow (instead of the pulsating that it does while it’s sleeping). Worse, I could no longer wake the machine from this pseudo-sleep. Clicking on the mouse or hitting the keyboard wouldn’t do anything.

This is not the first time it’s happened. I went to the G4 itself and pressed on the power button for half a second. This worked before. (It would wake up the machine, for some reason.) It didn’t this time. I started fiddling with my FireWire and USB devices (which usually do not prevent the machine from going to sleep). I unplugged a FireWire external hard drive and plugged it back in. I unplugged my PowerMate and plugged it back in.

Then all of a sudden the machine came back on, and everything was fine — except that it complained that I had disconnected a FireWire device without ejecting it first, of course.

I am not sure what prevented it from going INTO deep sleep, and I am not sure what caused it to wake up from this pseudo-sleep. It probably has to do with FireWire or USB devices. But which ones? I don’t think I’ll ever know…

It still frustrates me that, after all this time, Apple has not come up with a reliable architecture for system sleep that is not susceptible to interference from such external devices — especially since the interference only occurs from time to time.

Unpredictability is one of the worst sins in computer hardware/software design.

[UPDATE] I wanted to submit a bug report to Apple about this, and had a closer look at the Console Log around that time, and I found this:

May 7 23:09:33 DualG4 WindowServer[624]: Reserved range exhausted. 
(0xbbffe000 to 0xbc01d000 goes out of bounds)

2003-05-07 23:09:58.536 iCal Helper[687] Launch of helper planned at : 
05/08/03 00:09:58
May 7 23:10:29 DualG4 WindowServer[624]: Reserved range exhausted. 
(0xbbffe000 to 0xbc052000 goes out of bounds)

May 7 23:10:33 DualG4 last message repeated 3 times

May 7 23:10:46 DualG4 WindowServer[624]: Reserved range exhausted. 
(0xbbffe000 to 0xbc01a000 goes out of bounds)
…

## Component Manager: attempting to find symbols in a component alias 
of type (regR/carP/x!bt)
May 7 23:16:53 DualG4 lookupd[377]: DNSAgent: dns_send_query_server - 
send failed for 142.177.1.2
…

May 7 23:16:53 DualG4 AppleFileServer: AFPServer::AbnormalATListenerShutDown 
- clear listener

PowerMate device removed.
PowerMate device added.
May 7 23:18:43 DualG4 AppleFileServer: AFPServer::AbnormalATListenerShutDown 
- clear listener

May 7 23:18:43 DualG4 WindowServer[624]: CGXDisableUpdate: Updates 
disabled by connection 0xec03 for over 1.000000 seconds

May 7 23:18:44 DualG4 WindowServer[624]: Reserved range exhausted. 
(0xbbffe000 to 0xbc05a000 goes out of bounds)

May 7 23:18:55 DualG4 /sbin/autodiskmount[367]: disk3s9 
hfs yes yes Transport [not mounted] 

May 7 23:18:55 DualG4 /sbin/autodiskmount[367]: disk3s10 
hfs yes yes Tools [not mounted] 

May 7 23:18:56 DualG4 /sbin/autodiskmount[367]: disk3s9 
hfs yes yes Transport /Volumes/Transport 

May 7 23:18:56 DualG4 /sbin/autodiskmount[367]: disk3s10 
hfs yes yes Tools /Volumes/Tools 

May 7 23:19:01 DualG4 WindowServer[624]: Reserved range exhausted. 
(0xbbffe000 to 0xbc03d000 goes out of bounds)

May 7 23:19:34 DualG4 AppleFileServer: AFPServer::AbnormalATListenerShutDown 
- clear listener

May 7 23:19:36 DualG4 lookupd[738]: DNSAgent: dns_send_query_server 
- send failed for 142.177.1.3
…

May 7 23:19:47 DualG4 lookupd[738]: DNSAgent: dns_send_query_server 
- send failed for 142.177.1.3

2003-05-07 23:19:47.631 iChatAgent[664] WARNING: SubNetSession: 
Couldn't determine my IP address — can't open session

May 8 08:43:04 DualG4 WindowServer[624]: Reserved range exhausted. 
(0xbbffe000 to 0xbc010000 goes out of bounds)

Interesting! (I have removed the many duplicates and inserted “…” instead.) I can clearly see when I disconnected and reconnected the PowerMate and the FireWire drive (two partitions: “Transport” and “Tools”) and there is something about “AbnormalATListenerShutDown“.

But more interestingly during all that time there are all kinds of “dns_send_query_server” lines! Yet I had disconnected from the Internet at that time. So obviously it’s trying to connect even though there’s no established connection. Maybe that’s what’s preventing it from going to sleep. And it looks like iChat might have something to do with it!

Well, I’ve sent all this, including the Console Log, to Apple. We’ll see what they have to say (if anything).

This entry was posted by Pierre Igot on Thursday, May 8th, 2003 at 10:20 pm and is filed under Macintosh. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.