Just a brief note: I installed 3.9.17 yesterday. Whereas 3.9.16 was doing seg fault/core dumps, 3.9.17 instead gave tcmalloc errors that were non-fatal. In other words, program execution continued, swap space was used (roughly 12GB in total) and it generated a final assembly. In 3.9.15 I had been getting tcmalloc errors that were fatal, although I did not use an identical data set and parameters. I haven't exhaustively tested 3.9.17 to see if it always handles the tcmalloc errors gracefully in all circumstances, but I'm pleased that it ran with the parameters that I used. --Matt On Fri, May 10, 2013 08:19 PM "Matthew D. Pagel" <pagel@xxxxxxx> wrote: > >Hi all, > >Finally upgraded my system RAM to the max allowable by my board: 8GB (getting >this RAM was an adventure in itself) and used the opportunity to upgrade my >harddrive. In any case, mira (mira_3.9.16_linux-gnu_x86_64_static) is still >crashing with errors like > >tcmalloc: large alloc 1116635136 bytes == 0x225f90000 @ >Segmentation fault (core dumped) > >After going into my shell, re-upping my swap space to beyond 32GB, I monitored >memory and swap usage (via top) during one of the runs until it crashed, and it >appears that my swap space is not being used by mira. The crash is occurring >when my physical memory is all used, but the swap space is not being used. >My current memory/swap lines (not running mira) read >Mem: 7971212k total, 1693792k used, 6277420k free, 9912k buffers >Swap: 43745276k total, 519148k used, 43226128k free, 667876k cached > >Am I missing a library for handling swap with mira? >swapon -s >Filename Type Size Used Priority >/dev/mapper/VolGroup00-LogVol01 partition 43745276 518952 >-1 > >cat /proc/sys/vm/swappiness >60 > > --- Logical volume --- > LV Name /dev/VolGroup00/LogVol01 > VG Name VolGroup00 > LV Write Access read/write > LV Status available > # open 3 > LV Size 41.72 GiB > Current LE 1335 > Segments 2 > Allocation inherit > Read ahead sectors auto > - currently set to 256 > Block device 252:1 > >Also, in a recent message you said that 3.9 doesn't use the traceinfo file >anymore - how then does it retrieve clipping/masking info? > > _______________________________________________________ Matt Pagel Graduate Student Penn State Biochemistry, Microbiology and Molecular Biology -- You have received this mail because you are subscribed to the mira_talk mailing list. For information on how to subscribe or unsubscribe, please visit http://www.chevreux.org/mira_mailinglists.html