this happened on my linux box with boinc 4.43. thought somebody might be interested:

23.12.2005 18:12:16||Suspending computation and network activity - running CPU benchmarks
23.12.2005 18:12:16|climateprediction.net|Pausing result sulphur_glrk_000774704_0 (removed from memory)
23.12.2005 18:12:18|climateprediction.net|Unrecoverable error for result sulphur_glrk_000774704_0 (process got signal 11)
23.12.2005 18:12:18||request_reschedule_cpus: process exited
23.12.2005 18:12:18||Running CPU benchmarks
23.12.2005 18:13:16||Benchmark results:
23.12.2005 18:13:16|| Number of CPUs: 1
23.12.2005 18:13:16|| 634 double precision MIPS (Whetstone) per CPU
23.12.2005 18:13:16|| 1767 integer MIPS (Dhrystone) per CPU
23.12.2005 18:13:16||Finished CPU benchmarks
23.12.2005 18:13:16||Resuming computation and network activity
23.12.2005 18:13:16|climateprediction.net|Deferring communication with project for 2 seconds
23.12.2005 18:13:16|climateprediction.net|Computation for result sulphur_glrk_000774704_0 finished
23.12.2005 18:13:16||schedule_cpus: must schedule
23.12.2005 18:13:16|climateprediction.net|Starting result sulphur_grhh_000782117_0 using sulphur_cycle version 4.22

---

You may find some solace <a href=\"http://climateapps2.oucs.ox.ac.uk/cpdnboinc/forum_thread.php?id=3616\"> here</a> with others who have this problem.

You may want to upgrade to BOINC 5.2.x and set \"Leave application in memory when pre-empted\" to Yes. This may prevent, or minimize the occurrence of this error.

The problem is when Boinc (running CPU benchmarks) or you (using boincmgr) pause a result.
Boinc run a CPU benchmarks punctual every 5 days, if you kill Boinc before, and run this after that date, in that case the first work of Boinc is running CPU benchmarks and the second is starting result and the problem disappear.

---