Messages in this thread | | | Date | Mon, 27 Mar 2000 08:40:02 +0000 (GMT) | From | Nicholas Vinen <> | Subject | Re: Slow pthread_create() under high load |
| |
On Mon, 27 Mar 2000, Joerg Pommnitz wrote:
> Stephen C. Tweedie wrote: > > > Maybe, but although a lot of the POSIX threads are reasonable, things > > like requiring uid/gid updates to be instantly effective across all > > threads in the process are just insane. There ain't no way that's > > going into the kernel any time soon: it would require adding > semaphore > > locking to all of the credential information that is currently task- > > local. > > Isn't this more the case of a missing "CLONE_CREDENTIALS" flag? > If POSIX specifies that credential are shared among threads > (similiar to VM, files,...) and Linux does not do this, than > it's not the specification that is broken but Linux' implementation > of the same... > > Regards > Joerg
Perhaps these people worried about thread performance need to set their priority scheme to SCHED_RR. Especially if it's a database you probably want it to take a lot of your CPU time. This should end up with more aggressive allocation of time slices, as long as appropriate sleeping is going on it should fix this kind of thing. I've done some multimedia programming in Linux and appropriate thread priority makes the difference between choppy and smooth thread interactions.
Nicholas Vinen
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |