[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-ID: <20230727152020.3633009-1-chengming.zhou@linux.dev>
Date: Thu, 27 Jul 2023 23:20:18 +0800
From: chengming.zhou@...ux.dev
To: axboe@...nel.dk, osandov@...com, ming.lei@...hat.com,
kbusch@...nel.org, krisman@...e.de
Cc: linux-block@...r.kernel.org, linux-kernel@...r.kernel.org,
zhouchengming@...edance.com
Subject: [PATCH v2 1/3] sbitmap: fix hint wrap in the failure case
From: Chengming Zhou <zhouchengming@...edance.com>
```
hint = nr + 1;
if (hint >= depth - 1)
hint = 0;
```
Now we wrap the hint to 0 in the failure case, but has two problems:
1. hint == depth - 1, is actually an available offset hint, in which
case we shouldn't wrap hint to 0.
2. In the strict round_robin non-wrap case, we shouldn't wrap at all.
```
wrap = wrap && hint;
```
We only need to check wrap based on the original hint ( > 0), don't need
to recheck the new hint which maybe updated in the failure case, which
may cause second wrap. We set wrap to false after we wrap once to avoid
repeated wrap, which is clearer than rechecking the hint.
Signed-off-by: Chengming Zhou <zhouchengming@...edance.com>
---
lib/sbitmap.c | 13 ++++++++++---
1 file changed, 10 insertions(+), 3 deletions(-)
diff --git a/lib/sbitmap.c b/lib/sbitmap.c
index d0a5081dfd12..ac4027884765 100644
--- a/lib/sbitmap.c
+++ b/lib/sbitmap.c
@@ -149,7 +149,8 @@ static int __sbitmap_get_word(unsigned long *word, unsigned long depth,
* offset to 0 in a failure case, so start from 0 to
* exhaust the map.
*/
- if (hint && wrap) {
+ if (wrap) {
+ wrap = false;
hint = 0;
continue;
}
@@ -160,8 +161,14 @@ static int __sbitmap_get_word(unsigned long *word, unsigned long depth,
break;
hint = nr + 1;
- if (hint >= depth - 1)
- hint = 0;
+ if (unlikely(hint >= depth)) {
+ if (wrap) {
+ wrap = false;
+ hint = 0;
+ continue;
+ }
+ return -1;
+ }
}
return nr;
--
2.41.0
Powered by blists - more mailing lists