On Sun, 21 Nov 2004, Davide Libenzi wrote:
So it seems it did not work even before, the gdb-SIGTRAP stepping. In 2.6.8 I get a straight segfault just for running it.
Ok, that at least means it's not a regression, although it may be that the altered behaviour is enough to make some program work/not-work depending on exactly what it is testing. My example is certainly not the only way to try to mess up a debugger.
I'm by no means 100% sure that we should encourage the kind of programming "skills" I showed with that example program, so in that sense this may not be worth worrying about. That said, I do hate the notion of having programs that are basically undebuggable, so from a QoI standpoint I'd really like to say that you can run my horrid little program under the debugger and see it work...
Linus