Nginx 0.4.9 crashes using /dev/poll on Solaris 10 on Sparc

[Per Jonsson]

On Fri, Oct 20, 2006 at 06:12:53PM +0400, Igor Sysoev wrote:
> When you get segfault ? Just when run nginx or some conditions are needed ?
> Could you use gdb ?
  
We're getting it during what we believe is normal operation, but not all the
time. I haven't been able to find any certain conditions that give us the
segfault. The nice thing though is that nginx recovers in a good way, but
it's hard to transfer bigger files from the backend webservers running behind
nginx. 
  
Some times when can run for a couple of minutes without segfaults, but then get
segfaults perhaps once a minute or more often.
  
I've managed to wrangle up a gdb, and the error is happening as per my earlier
diagnosis.
  
The problem seems to be that ngx_cycle->files seems to contain null-pointers,   
and that the check for c->fd == -1 is not preceeded with a check for c == 0.
  
Please tell me if you need more info.

Best Wishes


Per
Lysator Academic Computer Society