On Wed, Nov 07, 2012 at 11:56:10AM -0800, Andrew Morton wrote:
> On Wed, 7 Nov 2012 01:05:48 -0200
> Rafael Aquini <aquini@xxxxxxxxxx> wrote:
>
> > This patch introduces MIGRATEPAGE_SUCCESS as the default return code
> > for address_space_operations.migratepage() method and documents the
> > expected return code for the same method in failure cases.
>
> I hit a large number of rejects applying this against linux-next. Due
> to the increasingly irritating sched/numa code in there.
>
> I attempted to fix it up and also converted some (but not all) of the
> implicit tests of `rc' against zero.
>
> Please check the result very carefully - more changes will be needed.
>
> All those
>
> - if (rc)
> + if (rc != MIGRATEPAGE_SUCCESS)
>
> changes are a pain. Perhaps we shouldn't bother.
Thanks for doing that.
This hunk at migrate_pages(), however, is not necessary:
@@ -1001,7 +1001,7 @@ out:
if (!swapwrite)
current->flags &= ~PF_SWAPWRITE;
- if (rc)
+ if (rc != MIGRATEPAGE_SUCCESS)
return rc;
Here, migrate_pages() is not testing rc for the migration success, but it's just trying to
devise the flow if it has to return -ENOMEM, actually.
I guess, a change to make that snippet more clear could be:
diff --git a/mm/migrate.c b/mm/migrate.c
index 77ed2d7..6562aee 100644
--- a/mm/migrate.c
+++ b/mm/migrate.c
@@ -987,7 +987,7 @@ int migrate_pages(struct list_head *from,
case -EAGAIN:
retry++;
break;
- case 0:
+ case MIGRATEPAGE_SUCCESS:
break;
default:
/* Permanent failure */
@@ -996,15 +996,12 @@ int migrate_pages(struct list_head *from,
}
}
}
- rc = 0;
+ rc = nr_failed + retry;
out:
if (!swapwrite)
current->flags &= ~PF_SWAPWRITE;
- if (rc)
- return rc;
-
- return nr_failed + retry;
+ return rc;
}
I can rebase this patch and resubmit if you prefer
--
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>