58100 crashes with fglrx driver and radeon driver

I've done some more testing to try and pin down why 58100 is crashing for me, at first I thought it may be the fglrx driver as it's not the most stable at the best of times, so I reverted to the open source radeon driver in xorg 7.1.

Still the same crash which makes me think it probably isn't the fglrx driver:

2007-02-20T11:44:03Z INFO: _queueDataRequest: Starting transfer for 0d368e35-2451-83b3-7386-20a7db9fbb20
2007-02-20T11:44:03Z INFO: processTransferInfo: Receiving 1467ee0e-08f1-b074-4b5d-6356c2f51b6b, size 165661 bytes
*********************************WARN_ONCE*********************************
File r300_vertexprog.c function t_dst_index line 184
Unknown output 3
***************************************************************************
2007-02-20T11:44:06Z WARNING: LLVolumeMessage::constrainVolumeParams() - forced to constrain incoming volume params.
2007-02-20T11:44:06Z WARNING: LLVolumeMessage::constrainVolumeParams() - forced to constrain incoming volume params.
2007-02-20T11:44:07Z WARNING: LLVolumeMessage::constrainVolumeParams() - forced to constrain incoming volume params.
2007-02-20T11:44:07Z WARNING: processUpdateMessage: Bogus volume parameters in object f12921c1-1069-6b29-2ef9-5523fec37ddc
2007-02-20T11:44:07Z WARNING: processUpdateMessage: { 256512, 257280, 0 }
2007-02-20T11:44:09Z WARNING: signal_handlers: *** Caught signal 11
2007-02-20T11:44:09Z INFO: remove_marker_file()
2007-02-20T11:44:09Z INFO: do_elfio_glibc_backtrace: Opening stack trace file /home/mike/.secondlife/logs/stack_trace.log
2007-02-20T11:44:11Z INFO: do_elfio_glibc_backtrace: Finished generating stack trace.

Stack trace:

0: ELF(do_elfio_glibc_backtrace()+0x107) [0x95f2137]
1: ELF(viewer_crash_callback()+0x1a4) [0x9604ff4]
2: ELF(signal_handlers(int)+0x112) [0x95ff8c2]
3: [0xb7fa2420]
4: /lib/tls/i686/cmov/libc.so.6 [0xb6567a07]
5: /lib/tls/i686/cmov/libc.so.6(malloc+0x7f) [0xb656992f]
6: /home/mike/tmp/SecondLife_i686_1_13_3_58100_FIRSTLOOK/lib/libstdc++.so.6(_Znwj+0x26) [0xb670de26]
7: /home/mike/tmp/SecondLife_i686_1_13_3_58100_FIRSTLOOK/lib/libkdu_v42R.so(_ZN14kdu_codestream6createEP21kdu_compressed_source+0x16f) [0xb3eecc0f]
8: /home/mike/tmp/SecondLife_i686_1_13_3_58100_FIRSTLOOK/bin/libllkdu.so(_ZN13LLImageJ2CKDU15setupCodeStreamER10LLImageJ2CiNS_15ECodeStreamModeE+0xa0) [0xb3f65c3a]
9: /home/mike/tmp/SecondLife_i686_1_13_3_58100_FIRSTLOOK/bin/libllkdu.so(_ZN13LLImageJ2CKDU10initDecodeER10LLImageJ2CR10LLImageRawfNS_15ECodeStreamModeEii+0x97) [0xb3f667f3]
10: /home/mike/tmp/SecondLife_i686_1_13_3_58100_FIRSTLOOK/bin/libllkdu.so(_ZN13LLImageJ2CKDU10decodeImplER10LLImageJ2CR10LLImageRawfii+0x71) [0xb3f66d6f]
11: ELF(LLImageJ2C::decode(LLImageRaw*, float, int, int)+0x110) [0x96e2430]
12: ELF(LLImageJ2C::decode(LLImageRaw*, float)+0x1b) [0x96e1b4b]
13: ELF(LLImageWorker::doWork(int)+0x73) [0x96e81b3]
14: ELF(LLWorkerThread::WorkRequest:rocessRequest()+0x22) [0x99de952]
15: ELF(LLQueuedThread:rocessNextRequest()+0x210) [0x99b9f90]
16: ELF(LLQueuedThread::run()+0x35) [0x99ba135]
17: ELF(LLThread:taticRun(apr_thread_t*, void*)+0x1 [0x99d6308]
18: /home/mike/tmp/SecondLife_i686_1_13_3_58100_FIRSTLOOK/lib/libapr-1.so.0
[0xb6b096e6]
19: /lib/tls/i686/cmov/libpthread.so.0 [0xb62bb240]
20: /lib/tls/i686/cmov/libc.so.6(__clone+0x5e) [0xb65cd3de]

Run Multiple Threads!

Turning this on seems to be the cause of my crashiness. I should have realised that from the stack trace! Yes, I know it's a debug option but it sounded really cool

It now even passes my "Go up really high and spin the camera around till I'm sick" test. I'm so relieved, I was starting to think I'd have to go back to my Windows version when this is released.

Thanks to everyone who offered suggestions on where to look, I know a lot more about my X server now