Hi Tim, On Thu, 7 Dec 2006 16:06:50 +0100 Tim Dijkstra <[EMAIL PROTECTED]> wrote:
> From the strace it seems suspend-keygen exits cleanly. There are > several points in the code where the execution path jumps to clean up > if something returned an error. Unfortunately there are almost no > error messages. I could stick some in, but if you are able to handle > gdb, you could also try to find the problem yourself. I don't think I have that much knowledge to do it myself. > What bothers me a bit, is that apparently suspend-keygen bails out at > different points. If you get no prompt for the number of bits, already > initialization of the libgcrypt library didn't succeed. > In the second trace at least part of the key was generated. > > Hmm, could it be that you are running out of memory? There's enough disk space, memory, swap, tmp... I could create gpg keys with 4096 bits without problems - this demands the memory the same way, doesn't it? I figured out that I could create keys with suspend-keygen up to 3986 bits. But with higher bit rate it fails. I attached a strace output from a successful key creation. Philip.
strace_suspend_3986
Description: Binary data
