#19 code dumped for no obvious reson

[Kamil S Jaron]

Hello,

I am trying to run numer to find a match between two assemblies (different individuals, same species, 500M genome) getting an:

ERROR: mummer and/or mgaps returned non-zero

Then I run just mummer to find out what non-zero returned value I get (and if there will more info...).

mummer <genome1> <genome2>

reading input file "<genome1>" of length 492200683

construct suffix tree for sequence of length 492200683

(maximum reference length is 536870908)

(maximum query length is 4294967295)

process 4922006 characters per dot

...........................................................................Segmentation fault (core dumped)

So I recompiled Mumer with -g flag and here is what gdb was reporting:

Program received signal SIGSEGV, Segmentation fault.
0x00000000004062a9 in insertbranchnode (stree=0x7ffffffdb3b0) at construct.c:482
482 SETBROTHER(stree->branchtab + GETBRANCHINDEX(stree->insertprev)

If it is there anything I could help you with (like reading a value of some variable thought gdb).

Do you have and idea what could go wrong? My genome is softmasked by the way (according to manual it should be ok).

Cheers,

Kamil

P.S. I tried to ask this quesiton thought mailing list, but I got always auto-reply that I have to sign up for the subscription of the list to post there question, even I was already subscriber.