Originally posted by: RAVI KALASAPUR
Hi,
I am not sure whether my question is appropriate to this forum. Sorry if it wasn't appropriate. :)
We are constantly facing core dumps on AIX for our application servers. The stack trace always gives :-
Segmentation fault in __DoThrowV6 at 0x9000000006ad004 ($t16)
0x9000000006ad004 (__DoThrowV6+0x64) f821fce1 stdu r1,-800(r1)
(dbx) where
__DoThrowV6() at 0x9000000006ad1ac
vtcpconnb._read__14VISTCPConnBaseFUcT1PcUiT4UL(0x11056a370, 0x100000000000001, 0x100000000000001, 0x110480b10, 0x0, 0xc0000000c, 0x0) at 0x9000000012f38bc
Looks like there is a patch for this problem :-
http://www-01.ibm.com/support/docview.wss?uid=swg24023226 What I understand is to install the patch in the build machine and then rebuild the binaries. (C++ is the language)
But, this means we have to rebuild the entire application ?
We observed that whenever we had a block like below, it was dumping :-
try
{
...
}
catch(xx) // some suggestions were to change this to pass by reference
{
throw; // when we changed this to throw xx; (Explicit throw) it worked.
}
Is there any workaround (like above) through which we can handle that in code ?
Any help would be great.
#AIX-Forum