Messages in this thread | | | Date | Tue, 23 Sep 2003 10:58:41 -0700 | From | "David S. Miller" <> | Subject | Re: [PATCH] softirq_pending() |
| |
On Tue, 23 Sep 2003 16:48:47 +0200 Christoph Hellwig <hch@lst.de> wrote:
> x86-64 currently ignores the cpu argument to softirq_pending() and > always uses smp_processor_id(). And indeed that's the only possible > argument. So consolidate the old softirq_pending() and > local_softirq_pending() into a single one.
The problem is that, on some of the platforms that don't ignore the argument, the code generation is much better.
GCC doesn't consider smp_processor_id() like some const local variable, so multiple invocations are assumed to return different values because in many cases 'current_thread_info()' is obscured.
Your patch is going to make a lot of new code get generated on x86 for example, so I don't think it should be applied even though my own platforms are not effected by this issue. - 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/
| |