Messages in this thread | | | Date | Thu, 24 Feb 2005 19:02:08 -0800 | From | Chris Wright <> | Subject | Re: [PATCH] override RLIMIT_SIGPENDING for non-RT signals |
| |
* Roland McGrath (roland@redhat.com) wrote: > > Yeah, it fixes the issue, but opens the door to larger consumption of > > pending signals. Roland, what was your final preference? I'm kind of > > leaning towards Jeremy's original patch. > > It's not a matter of preference. As I said in the first place, without my > patch we are violating POSIX, and delivering unreliable results to users.
Right, and as you also mentioned, it's identical case to exhausting atomic pool, in either case we're out of resources, and in both cases the machine may recover and be functional. And sneaking around the rlimit can cost ~4k per-process, which is why I'd consider the edge case a reasonable loss. (heck, maybe 4k is fine considering task size, and mlock limits, etc).
thanks, -chris -- Linux Security Modules http://lsm.immunix.org http://lsm.bkbits.net - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |