Application cpu looping due to out of memory?

Hi all,

I am using snapraid, a software like raid system, for a long time now. It uses some decent amount of RAM to do it’s work which it did fine for almost a year. From one day to the other it stopped working (related to an turris os update? I don’t know). I posted at the snapraid forum to ask for help and they tracked down the problem to not having enough memory: https://sourceforge.net/p/snapraid/discussion/1677233/thread/4ed13e54/?limit=25

However, I restarted turris and there are 1.4 GBs available (more than enough for 700mb which snapraid needs). Running free also tells me that there is enough RAM available:

root@turris:~# free
             total       used       free     shared    buffers     cached
Mem:       2070768     632748    1438020      43080       1956      74832
-/+ buffers/cache:     555960    1514808
Swap:            0          0          0

Running strace shows that this is the last call made by snapraid:

mmap2(NULL, 46112768, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x9eae1000

After that, snapraid loops with 100% cpu for hours and never terminates.

Why is mmap2 returning 2.6GB as RAM available?
Do you guys have any idea what happend? Since I am using the same tool for months without any problems I can not really believe that the problem is caused by it.

Thank you for help.

It’s not a number but a pointer to the new memory – around 44 MiB was requested and obtained.

Ah ok that happens if a non c pointer guy trys to debug applications :smiley: Thanks.
?
Is it possible that turris’ RAM is damaged?