Messages in this thread | | | Date | Tue, 24 Aug 2004 08:13:39 -0700 (PDT) | From | Davide Libenzi <> | Subject | Re: [patch] lazy TSS's I/O bitmap copy ... |
| |
On Tue, 24 Aug 2004, Arjan van de Ven wrote:
> On Mon, 2004-08-23 at 23:23, Davide Libenzi wrote: > > The following patch implements a lazy I/O bitmap copy for the i386 > > architecture. With I/O bitmaps now reaching considerable sizes, if the > > switched task does not perform any I/O operation, we can save the copy > > altogether. In my box X is working fine with the following patch, even if > > more test would be required. > > the thing is that X will not hit your fault path, since it runs with > iopl() called... your patch is a nice optimisation for X as a result, > however as test, X is almost worthless ;(
Arjan, using FC1 here and X definitely hits the GPF path. A few crashes when doing The Wrong Thing confirmed it yesterday :-) Also, according to the snippet Alan posted, X does iopl() on fallback of failing ioperm() IIRC.
- Davide
- 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/
| |