Thread (29 messages) 29 messages, 6 authors, 2012-01-08

Re: [PATCH v4 4/5] slub: Only IPI CPUs that have per cpu obj to flush

From: Gilad Ben-Yossef <gilad@benyossef.com>
Date: 2012-01-01 16:12:14
Also in: lkml

On Sun, Jan 1, 2012 at 2:41 PM, Avi Kivity [off-list ref] wrote:
On 11/23/2011 08:23 AM, Pekka Enberg wrote:
quoted
On Tue, 22 Nov 2011, Gilad Ben-Yossef wrote:
quoted
static void flush_all(struct kmem_cache *s)
{
-    on_each_cpu(flush_cpu_slab, s, 1);
+    cpumask_var_t cpus;
+    struct kmem_cache_cpu *c;
+    int cpu;
+
+    if (likely(zalloc_cpumask_var(&cpus, GFP_ATOMIC))) {
__GFP_NOWARN too maybe?
quoted
+        for_each_online_cpu(cpu) {
+            c = per_cpu_ptr(s->cpu_slab, cpu);
+            if (c->page)
+                cpumask_set_cpu(cpu, cpus);
+        }
+        on_each_cpu_mask(cpus, flush_cpu_slab, s, 1);
+        free_cpumask_var(cpus);
+    } else
+        on_each_cpu(flush_cpu_slab, s, 1);
}
Since this seems to be a common pattern, how about:

  zalloc_cpumask_var_or_all_online_cpus(&cpus, GFTP_ATOMIC);
  ...
  free_cpumask_var(cpus);

The long-named function at the top of the block either returns a newly
allocated zeroed cpumask, or a static cpumask with all online cpus set.
The code in the middle is only allowed to set bits in the cpumask
(should be the common usage).  free_cpumask_var() needs to check whether
the freed object is the static variable.
Thanks for the feedback and advice! I totally agree the repeating
pattern needs abstracting.

I ended up chosing to try a different abstraction though - basically a wrapper
on_each_cpu_cond that gets a predicate function to run per CPU to
build the mask
to send the IPI to. It seems cleaner to me not having to mess with
free_cpumask_var
and it abstracts more of the general pattern.

I intend to run the new code through some more testing tomorrow and send out V5
of the patch set. I'd be delighted if you can have a look through it then.

Thanks!
Gilad

-- 
Gilad Ben-Yossef
Chief Coffee Drinker
gilad@benyossef.com
Israel Cell: +972-52-8260388
US Cell: +1-973-8260388
http://benyossef.com

"Unfortunately, cache misses are an equal opportunity pain provider."
-- Mike Galbraith, LKML

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help