climateprediction.net home page
Posts by old_user59213

Posts by old_user59213

1) Questions and Answers : Unix/Linux : hadsm3um_4.10 SIGSEGV on SMP Linux laptop (Message 10430)
Posted 5 Mar 2005 by old_user59213
Post:

Worked for me too! Thanks much.

Best Regards,
Ben
2) Questions and Answers : Unix/Linux : hadsm3um_4.10 SIGSEGV on SMP Linux laptop (Message 10398)
Posted 5 Mar 2005 by old_user59213
Post:

Yep, at least sometimes. My earlier post about 'new participant finds no
work available' was answered by "Tolu" as a Project Admin.

Climateprediction works fine on my single processor desktop. When I see a
new version of hadsm3um get downloaded onto it, I'll try climateprediction
again on the SMP laptop. Till then the laptop is 100% setiathome (which I
recently converted from classic to BOINC with no problem, except the power
outages SETI has been experiencing this week).
3) Questions and Answers : Unix/Linux : hadsm3um_4.10 SIGSEGV on SMP Linux laptop (Message 10270)
Posted 3 Mar 2005 by old_user59213
Post:
Having gotten boinc running on my desktop, I installed
on my Linux laptop (P4 with HT, kernel 2.6.6 compiled
with SMP, Fedora core 2 with updates).

The hadsm3um_4.10 processes (it sees the system
as a 2 CPU system) core dump with signal 11
after running for only 90 CPU seconds or so.
They show up as defunct processes in ps.

The gdb output from the core doesn't say much:

Program terminated with signal 11, Segmentation fault.
0x0857e64c in trans_source_coeff_.H()
Cannot access memory at address 0xbfff9a94

This happens consistently every time I stop and
restart the client (four times now). Well, the
crash is as the same spot although sometimes the
bad address is 0xbfff9584 instead of 0xbfff9a94.

Any suggestions?

Thanks in advance,
Ben

P.S. This same machine has run SETI Classic for 2 weeks at
a time 24x7 so I don't think it's a hardware problem.
4) Questions and Answers : Unix/Linux : boinc now working, hadsm3viz-4.10 requires libstdc++.so.6 (Message 10243)
Posted 2 Mar 2005 by old_user59213
Post:

The binary started by viz (hadsm3viz-4.10-*) on
my new boinc installation (now working on its first
model) requires libstdc++.so.6.

My RedHat 8.0 machine has no later than 5.0.1
My Fedora Core 2 machine has no later than 5.0.5

Why does hadsm3viz-4.10 require a libstdc++ that
leaves systems as recent as Fedora Core 2 (with
all updates, as of last week) unsupported?

Best Regards,
Ben
5) Questions and Answers : Unix/Linux : new participant finds no work available (Message 10242)
Posted 2 Mar 2005 by old_user59213
Post:

It's working now. Thanks!

Ben
6) Questions and Answers : Unix/Linux : new participant finds no work available (Message 10208)
Posted 2 Mar 2005 by old_user59213
Post:

Hi, I just got set up with a new account and
boinc 4.19 on my Athlon 2600XP, 512MB, Red Hat 8.0+,
disk allocated 10GB and currently 55GB free on
that file system.

Boinc connects fine to climateprediction, but
for the last 24 hours or more (since I got it up
and running) there is, it seems, no work available.
It connects fine to the cgi, which responds 'no
work available' (no error messages).

Does climateprediction not need my help any more???

Best Regards,
Ben





©2024 climateprediction.net