On Wednesday 13 June 2007 20:40:11 Matt Mackall wrote: > On Mon, Jun 11, 2007 at 06:55:46PM -0700, Arjan van de Ven wrote: > > Andrew Morton wrote: > > >On Mon, 11 Jun 2007 18:10:40 -0700 Arjan van de Ven > > ><[EMAIL PROTECTED]> wrote: > > > > > >>Andrew Morton wrote: > > >>>>Where as resource pool is exactly opposite of mempool, where each > > >>>>time it looks for an object in the pool and if it exist then we > > >>>>return that object else we try to get the memory for OS while > > >>>>scheduling the work to grow the pool objects. In fact, the work > > >>>>is schedule to grow the pool when the low threshold point is hit. > > >>>I realise all that. But I'd have thought that the mempool approach is > > >>>actually better: use the page allocator and only deplete your reserve > > >>>pool > > >>>when the page allocator fails. > > >>the problem with that is that if anything downstream from the iommu > > >>layer ALSO needs memory, we've now eaten up the last free page and > > >>things go splat. > > > > > >If that happens, we still have the mempool reserve to fall back to. > > > > we do, except that we just ate the memory the downstream code would > > use and get ... so THAT can't get any. > > Then the downstream ought to be using a mempool?
Normally there shouldn't be a downstream. PCI IO tends to be not stacked, but at the edges (unless you're talking hypervisors with virtual devices but those definitely have separate memory pools). And the drivers I'm familiar with tend to first grab whatever resources they need and then map the DMA mappings. -Andi - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
