I must confess that I deduced this from observed behavior,
not actual measurement.
Running my modified leak.c under Solaris 2.6/SPARC would produce two
behaviors: malloc failure (and succeeding error msg), or a core dump
during the set-to-zero loop. The approximate crash location can be
easily determined by the fprintf() statements in that area of code.
Since the malloc calls succeeded, a core dump in the buffer init step
implied to me that memory was overcommitted, and could not be paged in
when needed.
It was compiled w/ gcc 2.8.1 '-O2 -s' if that makes any difference.
Jeff
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/