Messages in this thread | | | Date | Mon, 24 May 2004 10:16:57 -0700 (PDT) | From | Davide Libenzi <> | Subject | Re: scheduler: IRQs disabled over context switches |
| |
On Mon, 24 May 2004, Ingo Molnar wrote:
> > * Davide Libenzi <davidel@xmailserver.org> wrote: > > > We used to do it in 2.4. What changed to make it fragile? The > > threading (TLS) thing? > > it _should_ work, but in the past we only had trouble from such changes > (at least in the O(1) tree of scheduling - 2.4 scheduler is OK.). We > could try the patch below. It certainly boots on SMP x86. But it causes > a 3.5% slowdown in lat_ctx so i'd not do it unless there are some really > good reasons.
IMO it is fine, as long as it works with IRQ disabled. There are archs where IRQ latencies matters more than lat_ctx times (that BTW are bogus). And we already have the infrastructure in place to let the arch to choose the way better fits it. Russel reported that a guy trying it (IRQ enabled ctx switch) with MIPS was having some problem with it though.
BTW, the unlock_irq should go in prepare not finish.
- 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/
| |