Re: [PATCH] irq_enter/irq_exit consolidation
From: Christoph Hellwig (hch_at_infradead.org)
Date: 09/14/04
- Previous message: Richard B. Johnson: "Re: Kernel stack overflow on 2.6.9-rc2"
- In reply to: Martin Schwidefsky: "Re: [PATCH] irq_enter/irq_exit consolidation"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Tue, 14 Sep 2004 19:31:35 +0100 To: Martin Schwidefsky <schwidefsky@de.ibm.com>
On Mon, Sep 13, 2004 at 03:02:40PM +0200, Martin Schwidefsky wrote:
> Hi Christoph,
>
> > s390 has an assembly wrapper around do_softirq.
> >
> > I've extended the invoke_softirq mechanism used by s390 (also called
> > by ksoftirqd) to the two arm variants, but the right thing to do is
> > probably to use the normal do_softirq call in arm and set
> > __ARCH_HAS_DO_SOFTIRQ + providing a per-arch do_softirq for all callers
> > for s390 and maybe arm26.
>
> do_call_softirq switches to the asynchronous interrupt stack,
> just what i386 does now as well. Trouble is that on s390 it is
> non-trivial to do the switch in C with inline assembly. We need
> a bit of assembly. But we could get rid of invoke_softirq, define
> __ARCH_HAS_DO_SOFTIRQ and use do_softirq to call the assembly
> wrapper.
Well, the question is do you want the direct do_softirq invocations from
the networking code use the separate stack or not? If not the current
code is fine. If yes you should sent the patch to Andrew.
-
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/
- Previous message: Richard B. Johnson: "Re: Kernel stack overflow on 2.6.9-rc2"
- In reply to: Martin Schwidefsky: "Re: [PATCH] irq_enter/irq_exit consolidation"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|
|
