From mboxrd@z Thu Jan 1 00:00:00 1970 From: mingo@elte.hu (Ingo Molnar) Date: Thu, 30 Apr 2009 16:38:19 +0200 Subject: [ltt-dev] [PATCH] Fix dirty page accounting in redirty_page_for_writepage() In-Reply-To: References: <20090429232546.GB15782@Krystal> <20090430024303.GB19875@Krystal> <20090430133859.GB8329@elte.hu> <20090430141446.GD14696@elte.hu> Message-ID: <20090430143819.GF14696@elte.hu> * Christoph Lameter wrote: > On Thu, 30 Apr 2009, Ingo Molnar wrote: > > > The patch below makes the fallback/slowpath irq safe. > > Yes but sometimes you are already irq safe and such a fallback > would create significant irq/enable/disable stack operations etc > overhead for architectures that are using the fallback. It's a fallback slowpath - non-x86 architectures should still fill in a real implementation of course. > I think we really need another __xxx op here. Especially since > these operations are often in critical code paths. That's a receipe for fragility: as using __xxx will still be irq-safe on x86, and 95% of the testing is done on x86, so this opens up the path to non-x86 bugs. So we first have to see the list of architectures that _cannot_ implement an irq-safe op here via a single machine instruction. x86, ia64 and powerpc should be fine. Ingo