Re: ZONE_PADDING wastes 4 bytes of the new cacheline

[Andrea Arcangeli]

On Thu, Oct 21, 2004 at 08:26:56PM -0700, Andrew Morton wrote:
> Well yes, I spose the answer as always is "show me a testcase".  But the

I already gave you a testcase ages ago. Pretty simple. Get your hands on
a 2G box. boot, swapoff -a, load 1G in pagecache. malloc 1G, bzero 1G,
then open some dozen thousand files until the machine deadlocks despite
1G is perfcetly freeable in highmem.
 
> Any halfway setting will screw everyone.

this stuff run in production for 2 years at least, if anyone would be
screwed I'd already know. The only thing I know is that I've oom
deadlocks in my queue of bugs, and we'll see if this fixes them, it's
hard to tell since I don't know how much lowmem is pinned by highmem
capable allocs (like pagetables!).
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/