[Gossamer]

Hopefully this won't get buried in the barrage of Mac Pro threads, but here goes:
I've got a B&W G3 I'm getting set up for a friend to use this coming school year. It was my server last year, but I procured a Cube at work that I use for that now, so this one's going out on loan.
The problem is that it's starting to act quirky. Suddenly, firewire has stopped working. I wiped the stock Apple 13GB drive and installed a fresh copy of 10.4.6 straight from the DVD. Upon restart, I set up accounts and such, but Software update will either unexpectedly quit, or fail when installing the 10.4.7 update. I downloaded the stand along 10.4.7 installer straight from Apple, but that fails to verify, and if verification is skipped, it will say that there was an error when installing. Apps randomly quit, such as Safari (when I try to open it, it will bounce a couple times, then stop) or the Console, I can't repair disk permissions with Disk Utility, because halfway through it tells me it can no longer communicate with Disk Management Tool and I must quit and try again (scratch that, I just successfully repaired permissions). I've zapped the PRAM numerous times to no avail. I haven't tried the motherboard reset switch, but that's coming right up.
It's a G3/450 Rev 2, with 512MB RAM (4x128MB), OS 10.4.6, 2x13GB HDs (one on its own bus, one with the CD-ROM drive).
Should I start pulling memory and see if that's it? Is there a hardware test I can run? I'm hurtin!

[Gossamer]

Okay, I just realized that it's report me as having 10.4.7 installed even though the installation never successfully completed. Here's the log from the Safari crash:
**********

Host Name: Xenon
Date/Time: 2006-08-10 17:30:39.696 -0500
OS Version: 10.4.7 (Build 8J135)
Report Version: 4

Command: Safari
Path: /Applications/Safari.app/Contents/MacOS/Safari
Parent: WindowServer [62]

Version: 2.0.4 (419.3)

PID: 205
Thread: Unknown

Link (dyld) error:

bad local relocation length
**********

Host Name: Xenon
Date/Time: 2006-08-10 17:30:43.379 -0500
OS Version: 10.4.7 (Build 8J135)
Report Version: 4

Command: Safari
Path: /Applications/Safari.app/Contents/MacOS/Safari
Parent: WindowServer [62]

Version: 2.0.4 (419.3)

PID: 207
Thread: Unknown

Link (dyld) error:

bad local relocation length

[dimmer]

Do you have the Apple Hardware Test CD? It sounds like a HD failure overall, sad to say.

[seanc]

Ah one of your B&W G3's again Ram would be a likely suspect or else the CPU but i'm sure you already know that.

Edit: Whats the SMART status on the HDD and is it making any weird noises?

[Gossamer]

Update: I successfully ran the 10.4.7 update from my iBook with the G3's HD in a FW enclosure. I tried running Software Update but it quit unexpectedly in the middle of an update.
I tried installing the Java 1.3.1 and 1.4.2 update by itself, but it told me that the update is corrupted. Both drives passed a SMART test, and I'm running Rember right now to test the RAM. Thanks for the quick replies, I'll try to keep you guys updated.

[Gossamer]

Okay, Rember decided it would be fun to quit unexpectedly...let's try again

[seanc]

Or you could run it 1 128mb dimm at a time to see if the machine is stable and then hopefully you'll hit the bad one somewhere along the line.

[seanc]

And BTW, wouldn't it be likely that Rember quits when it hits the bad piece of Ram (if there is any)?

[Gossamer]

Okay, here's what Rember had to say: (I used '...' to cut down on space)


Memtest version 4.13 (32-bit)
Copyright (C) 2004 Charles Cazabon
Copyright (C) 2004, 2005 Tony Scaminaci (Macintosh port)
Licensed under the GNU General Public License version 2 only

MacOS X (Darwin) running in multiuser mode
POSIX version 200112
Pagesize is 4096
Requested memory: 350MB (367333376 bytes)
Available memory: 350MB (367333376 bytes)
Allocated memory: 350MB (367333376 bytes) at local address 0x02008000
Attempting memory lock... locked successfully
Creating test buffers...
Buffer A: 175MB (183666688 bytes) at local address 0x02008000
Buffer B: 175MB (183666688 bytes) at local address 0x0cf30800

Running 1 test sequence...

Test sequence 1 of 1:
Stuck Address : setting 1 of 16testing 1 of 16setting 2 of 16testing 2 of 16setting 3 of 16testing 3 of 16setting 4 of 16testing 4 of 16setting 5 of 16testing 5 of 16setting 6 of 16testing 6 of 16setting 7 of 16testing 7 of 16setting 8 of 16testing 8 of 16setting 9 of 16testing 9 of 16setting 10 of 16testing 10 of 16setting 11 of 16testing 11 of 16setting 12 of 16testing 12 of 16setting 13 of 16testing 13 of 16setting 14 of 16testing 14 of 16setting 15 of 16testing 15 of 16setting 16 of 16testing 16 of 16 ok
Random Value : \|/-\|/-\|/-\|/-\|/-\|/-\|/-\|/-\|/-\|/-\|/-\|/-\|/-\|/-\|/-\|/-\|/-\|/-\|/-\ ok
Compare XOR :  ok
Compare SUB :  ok
Compare MUL :  ok
Compare DIV :  ok
Compare OR :  ok
Compare AND :  ok
Sequential Increment:  ok

Solid Bits : setting 1 of 64testing
...

46 of 64setting 47 of 64testing 47 of 64FAILURE: 0xffffffff != 0xff00ffff at offset 0x02bc9d28.
FAILURE: 0xffffffff != 0xff00ffff at offset 0x02bc9d2a.
FAILURE: 0xffffffff != 0xff00ffff at offset 0x02bc9d2c.
FAILURE: 0xffffffff != 0xff00ffff at offset 0x02bc9d2e.

Block Sequential : setting 1 of 256testing 1 of 256setting 2 of 256testing 2 of 256setting 3 of 256testing 3 of 256setting 4 of 256testing 4 of 256setting 5 of 256testing 5 of 256setting 6 of 256testing 6 of 256setting 7 of 256testing 7 of 256setting 8 of 256testing 8 of 256FAILURE: 0x07070707 != 0x07060707 at offset 0x015d4ff0.
FAILURE: 0x07070707 != 0x07060707 at offset 0x015d4ff2.
FAILURE: 0x07070707 != 0x07060707 at offset 0x015d4ff4.
FAILURE: 0x07070707 != 0x07060707 at offset 0x015d4ff6.

Checkerboard : setting 1 of 64testing 1 of 64setting 2 of 64testing 2 of 64setting 3 of 64testing 3 of 64setting 4 of 64testing 4 of 64setting 5 of 64testing 5 of 64setting 6 of 64testing 6 of 64setting

...

4setting 61 of 64testing 61 of 64setting 62 of 64testing 62 of 64setting 63 of 64testing 63 of 64setting 64 of 64testing 64 of 64 ok

Bit Spread : setting 1 of 64testing 1 of 64setting 2 of 64testing 2 of 64setting 3 of 64testing 3 of 64setting 4 of 64testing 4 of 64setting 5 of 64testing 5 of 64setting 6 of 64testing

...

64setting 60 of 64testing 60 of 64setting 61 of 64testing 61 of 64setting 62 of 64testing 62 of 64setting 63 of 64testing 63 of 64setting 64 of 64testing 64 of 64 ok

Bit Flip : setting 1 of 256testing 1 of 256setting

...
49 of 256testing 49 of 256FAILURE: 0xffffffbf != 0xff00ffbf at offset 0x011e3d90.
FAILURE: 0xffffffbf != 0xff00ffbf at offset 0x011e3d92.
FAILURE: 0xffffffbf != 0xff00ffbf at offset 0x011e3d94.
FAILURE: 0xffffffbf != 0xff00ffbf at offset 0x011e3d96.
FAILURE: 0xffffffbf != 0xff00ffbf at offset 0x01f93978.
FAILURE: 0xffffffbf != 0xff00ffbf at offset 0x01f9397a.
FAILURE: 0xffffffbf != 0xff00ffbf at offset 0x01f9397c.
FAILURE: 0xffffffbf != 0xff00ffbf at offset 0x01f9397e.
FAILURE: 0xffffffbf != 0xff00ffbf at offset 0x01f93988.
FAILURE: 0xffffffbf != 0xff00ffbf at offset 0x01f9398a.
FAILURE: 0xffffffbf != 0xff00ffbf at offset 0x01f9398c.
FAILURE: 0xffffffbf != 0xff00ffbf at offset 0x01f9398e.


Walking Ones : setting 1 of 64testing 1 of 64setting

...

64testing 42 of 64setting 43 of 64testing 43 of 64setting 44 of 64testing 44 of 64setting 45 of 64testing 45 of 64setting 46 of 64testing 46 of 64FAILURE: 0xfffbffff != 0xfff7ffff at offset 0x015b0c60.
FAILURE: 0xfffbffff != 0xfff7ffff at offset 0x015b0c62.
FAILURE: 0xfffbffff != 0xfff7ffff at offset 0x015b0c64.
FAILURE: 0xfffbffff != 0xfff7ffff at offset 0x015b0c66.

Walking Zeroes : setting 1 of 6

...

testing 64 of 64 ok

*** Memory Test Failed *** Please check transcript for details.



So.....what the heck do I do. Can I tell from this info which chip is bad?

[seanc]

Well i certainly can't tell. One thing though, Rember says you have 350mb Ram, whereas you have 4 x 128mb dimms. That could imply that one is faulty and that it's essentially detecting 384mb. I guess you could pull out one of the dimms and see what happens, in theory if it's stable, you've pulled the bad dimm.

[Gossamer]

Originally Posted by seanc

Well i certainly can't tell. One thing though, Rember says you have 350mb Ram, whereas you have 4 x 128mb dimms. That could imply that one is faulty and that it's essentially detecting 384mb. I guess you could pull out one of the dimms and see what happens, in theory if it's stable, you've pulled the bad dimm.

Yeah, that's strange. OS X reports it has having 512MB. There is an option in Rember that lets you choose how much RAM to test, and I'm pretty sure I hit 'all,' but I guess it's possible I didn't. Once I get home from work tonight I'll start playing with it some more. Thanks for the replies!

[Gossamer]

Okay, so I pulled the topmost chip and ran Rember, and it passed. So then I put the top chip in the third chip's place, and it passed again. I added the third chip in the top slot and it failed. So it looks like I have a bad memory slot. I'm running one last test to make sure it's okay, and if so, I should be good to go. Thanks guys!