Martinsos
Segfault because of unsanitazed function parameters
Just to mention, this happens in an older version (not sure if the bug is present in the most recent one). This is the GDB's output from running my program:
Program received signal SIGSEGV, Segmentation fault. [Switching to Thread 0x7ff2f5399700 (LWP 9563)] 0x0000000000410f8d in obtainAlignment (maxNumBlocks=3, queryLength=178, targetLength=-13, W=14, bestScore=-1, position=-1, alignData=0x0, alignment=0x7ff2f5397f60, alignmentLength=0x7ff2f5397f5c) at src/alignment/myers.cpp:765
At the beginning of the obtainAlignment function, there is an allocation without checking the values of input parameters. Also, it would be good that after every malloc/new there is a check if the pointer is NULL (but please continue handling things in C-like manner, throwing exceptions is just awful :-) ).
Ah yes, good catch, thanks! I didnt do validation of input parameters yet, but I should certainly do it. I usually do not check if malloc pointer is NULL, how likely is that to happen? Is that something that is usually handled?
I usually do not check if malloc pointer is NULL, how likely is that to happen?
I think it is always a good practice, though if you are allocating small chunks of memory it is very unlikely you won't be able to actually get it. But since your library could potentially consume huge amounts of space, I think it would be prudent.
E.g. As from one of the previous examples I've sent you, even on relatively small input sizes (e.g. query sequence of length 30k) the memory consumption gets up to 100MB if alignment reconstruction was requested. Imagine if I was to run it with even a larger query sequence, and perform that in parallel on several cores. I feel you could get in the ball park of available free memory on a normal (not very-high end) laptop.
That is true! Ok, I will certainly takecare of that at some moment soon, thanks On Feb 8, 2015 8:41 AM, "Ivan Sovic" [email protected] wrote:
I usually do not check if malloc pointer is NULL, how likely is that to happen?
I think it is always a good practice, though if you are allocating small chunks of memory it is very unlikely you won't be able to actually get it. But since your library could potentially consume huge amounts of space, I think it would be prudent.
E.g. As from one of the previous examples I've sent you, even on relatively small input sizes (e.g. query sequence of length 30k) the memory consumption gets up to 100MB if alignment reconstruction was requested. Imagine if I was to run it with even a larger query sequence, and perform that in parallel on several cores. I feel you could get in the ball park of available free memory on a normal (not very-high end) laptop.
— Reply to this email directly or view it on GitHub https://github.com/Martinsos/edlib/issues/37#issuecomment-73400684.
Note: checking if memory was allocated is not such a big issue anymore, since we do not consumer large amount of memory any more (thanks to hirschbergs algorithm).