tags 123272 upstream wontfix thanks On Mon, Jul 31, 2006 at 04:03:09AM +0200, Matej Vela wrote: > I couldn't find it at first glance and gave up; the fact that no one > cared to in 4 years is telling me something.
Well, it is a program intended to produce SIGSEGV, SIGILL, and even SIGBUS on sparc64. I think we both agree on that. There are 2 major modes of working: * one with 1 process; * another with forked processes. The first mode is when crashme is invoked without the NSUB parameter. Sometimes, the SIGSEGV can be catched, but some other times, the kernel explicitely takes control to kill the process. Depending of the nature of what triggered the SIGSEGV, the signal cannot be catched via userspace. The second mode is when crashme is invoked with the NSUB parameter. Here, processes are forked and the same appears. However, we do not really see it at first glance, because there are always more than one process which keep running. It appears the fix is there is no fix. Nowadays, the Linux kernel is too clever for that old program. ;P Cheers, -- .''`. Aurélien GÉRÔME : :' : `. `'` Free Software Developer `- Unix Sys & Net Admin
signature.asc
Description: Digital signature