Messages in this thread | | | Date | Tue, 17 Mar 1998 17:19:56 -0500 (EST) | From | George <> | Subject | Re: New syscall (rough draft), swap_out_pid() |
| |
On Tue, 17 Mar 1998, Rik van Riel wrote:
>On Tue, 17 Mar 1998, Balaji Srinivasan wrote: > >> I dont think this would work as you expect it to...as swap_out_process >> does not swap_out the whole process, but just one page...am I wrong here? > >No, your completely right... What would work is: >while (p->mm->rss) { > int gfp = GFP_SWAPOUT_NOW_YOU_DAMNIT; > swapout_process(p, gfp); >}
(I know it doesn't throw the whole thing out, I watched it work.) :)
Is it safe to assume that while this code is running, the program will not be run and try to be swapped back in? I thought about that but wasn't sure if the kernel could get into a swap-fight with one part (swap_out_pid) throwing the program out and the page faults bringing it back in. Doesn't sound likely but I suppose the worst that can happen is I lock my machine up. :)
I'll clean up it and see what I can do. Any idea what GFP_ it should use or is 0 adequate?
-George (who doesn't like wasted memory when it could be a cache)
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu
| |