On Sat, 2006-08-12 at 19:31 +0200, Indan Zupancic wrote:

On Sat, August 12, 2006 16:14, Peter Zijlstra said:

quoted

+struct sk_buff *__alloc_skb(unsigned int size, gfp_t gfp_mask, int fclone)
+{
+	struct sk_buff *skb;
+
+	skb = ___alloc_skb(size, gfp_mask & ~__GFP_MEMALLOC, fclone);
+
+	if (!skb && (gfp_mask & __GFP_MEMALLOC) && memalloc_skbs_available())
+		skb = ___alloc_skb(size, gfp_mask, fclone);
+
+	return skb;
+}
+

I'd drop the memalloc_skbs_available() check, as that's already done by
___alloc_skb.

Right, thanks. Hmm, its the last occurence of that function, even
better.

quoted

+static DEFINE_SPINLOCK(memalloc_lock);
+static int memalloc_socks;
+static unsigned long memalloc_reserve;

Why is this a long? adjust_memalloc_reserve() takes an int.

Euhm, right :-) long comes naturaly when I think about quantities op
pages. The adjust_memalloc_reserve() argument is an increment, a delta;
perhaps I should change that to long.

Is it needed at all, considering var_free_kbytes already exists?

Having them separate would allow ajust_memalloc_reserve() to be used by
other callers too (would need some extra locking).