On 06/26/2012 12:45 PM, David Rientjes wrote:
On Tue, 26 Jun 2012, Glauber Costa wrote:
diff --git a/include/linux/thread_info.h b/include/linux/thread_info.h
index ccc1899..914ec07 100644
--- a/include/linux/thread_info.h
+++ b/include/linux/thread_info.h
@@ -61,6 +61,12 @@ extern long do_no_restart_syscall(struct restart_block
*parm);
# define THREADINFO_GFP (GFP_KERNEL | __GFP_NOTRACK)
#endif
+#ifdef CONFIG_CGROUP_MEM_RES_CTLR_KMEM
+# define THREADINFO_GFP_ACCOUNTED (THREADINFO_GFP | __GFP_KMEMCG)
+#else
+# define THREADINFO_GFP_ACCOUNTED (THREADINFO_GFP)
+#endif
+
This type of requirement is going to become nasty very quickly if nobody
can use __GFP_KMEMCG without testing for CONFIG_CGROUP_MEM_RES_CTLR_KMEM.
Perhaps define __GFP_KMEMCG to be 0x0 if it's not enabled, similar to how
kmemcheck does?
That is what I've done in my first version of this patch. At that time,
Christoph wanted it to be this way so we would make sure it would never be
used with #CONFIG_CGROUP_MEM_RES_CTLR_KMEM defined. A value of zero will
generate no errors. Undefined value will.
Now, if you ask me, I personally prefer following what kmemcheck does here...
Right, because I'm sure that __GFP_KMEMCG will be used in additional
places outside of this patchset and it will be a shame if we have to
always add #ifdef's. I see no reason why we would care if __GFP_KMEMCG
was used when CONFIG_CGROUP_MEM_RES_CTLR_KMEM=n with the semantics that it
as in this patchset. It's much cleaner by making it 0x0 when disabled.
What I can do, instead, is to WARN_ON conditionally to the config option
in the page allocator, and make sure no one is actually passing the flag
in that case.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxx. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>