Messages in this thread |  | | Date | Mon, 28 Aug 2000 20:54:44 -0400 (EDT) | From | Alexander Viro <> | Subject | Re: SCO: "thread creation is about a thousand times faster than on |
| |
On Mon, 28 Aug 2000, Albert D. Cahalan wrote:
> a. The shared VM changes. If it hurts, don't do that. No bloat here. > b. All other threads get killed. This is POSIX-friendly. > c. Force an unclone(~0) to go with the exec. Can do setuid CGI. > > Doing unclone(CLONE_MM) is lame. Unclone all or none.
Screw POSIX, but why do you want to unclone everything? Components are independent. Really. I can see the point in unclone(CLONE_FILES) _if_ there was close-on-exec. I can see the point in unclone(CLONE_FS) upon personality change (had to do that during the namei rewrite - alternative root depends on personality and it's clearly a part of fs_struct). But what, in the name of Cthulhu, makes us unclone() everything, no matter whether we need it or not? E.g. would you really want to get an independent namespace upon exec()? Goodby mount(8)...
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org Please read the FAQ at http://www.tux.org/lkml/
|  |