Bug 72680
| Summary: | (Kernel Panic) FreeBSD 5.3 Does not survive fork bombing | ||
|---|---|---|---|
| Product: | Base System | Reporter: | Nikolas Britton <freebsd> |
| Component: | kern | Assignee: | freebsd-bugs (Nobody) <bugs> |
| Status: | Closed FIXED | ||
| Severity: | Affects Only Me | ||
| Priority: | Normal | ||
| Version: | 5.3-BETA7 | ||
| Hardware: | Any | ||
| OS: | Any | ||
State Changed From-To: open->feedback This doesn't make sense to me. How is swapoff(2) related to this situation? Please build a kernel with debug support and provide a panic message and a backtrace. Sorry to have wasted your time but after 3 hours of fork bombing and trying everthing else under the sun, I could not get FreeBSD to panic with the new debug kernel. I guess it was just a fluke or I was doing the wrong thing at the right time (when it paniced the first time it was in the middle of a portupgrade -arR, X session, etc.). here is a list of the many error messages I receviced (if it helps any): maxproc limit exceeded kern.ipc.maxpipekva exceeded swap_pager: out of swap space swap_pager_getswapspace Pipe call failed: bad file descriptor kern.maxfiles limit exceeded Can't open /dev/null: Too meny open files in the system Connot fork: Resource temporarily unavailable ad0: WARNING - READ_DMA interrupt was seen but timeout fired LBA=xxxxxx Also I take back what I said about freebsd 5.3 being unstable, you guys really know how to build'em, at one time I had the system load (top) for the 1min avg. pegged in the 90s and the 15min. avg. in the 50s - 60s. At the end of my testing I did finally get the system into an unrecoverable state and was forced to reboot it, I was running it with X up and when I try to switch to the console the screen went dead. I blame this one on X. It wasn't a total waste though as I did get to thoroughly burn-in and stress test my new (Used Armada 1750) laptop. So guess this PR is closed unless someone else can crash beastie. State Changed From-To: feedback->closed Closed at submitter's request; cannot reproduce. |
FreeBSD does not survive fork bombing (cannot recover the system) and panics about 15 minutes into the event complaining about swapoff. To me this seems like a regression in system stability because 4.x can easily survive this same attack. How-To-Repeat: Run this in a sh shell :(){ :|:& };: