From: Jakub Kicinski <kuba@kernel.org>
To: davem@davemloft.net
Cc: netdev@vger.kernel.org, edumazet@google.com, pabeni@redhat.com,
andrew+netdev@lunn.ch, horms@kernel.org, shuah@kernel.org,
linux-kselftest@vger.kernel.org, Jakub Kicinski <kuba@kernel.org>
Subject: [PATCH net 04/12] net: shaper: try to avoid violating RCU
Date: Tue, 5 May 2026 17:06:20 -0700 [thread overview]
Message-ID: <20260506000628.1501691-5-kuba@kernel.org> (raw)
In-Reply-To: <20260506000628.1501691-1-kuba@kernel.org>
net_shaper_commit() overrides nodes which may be concurrently read
under RCU. This is not a huge deal since the entries only contain
config, worst case user will see inconsistent config params. But
we should try to avoid this obvious RCU violation. Try to allocate
a new node. Since commit() can't fail fall back to overriding.
Full fix is probably not worth the complexity, struct net_shaper
is around 80B, and the allocation is with GFP_KERNEL.
Fixes: 93954b40f6a4 ("net-shapers: implement NL set and delete operations")
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
---
net/shaper/shaper.c | 28 +++++++++++++++++++++++-----
1 file changed, 23 insertions(+), 5 deletions(-)
diff --git a/net/shaper/shaper.c b/net/shaper/shaper.c
index 692e38df42cb..a2e9adca9afc 100644
--- a/net/shaper/shaper.c
+++ b/net/shaper/shaper.c
@@ -406,18 +406,36 @@ static void net_shaper_commit(struct net_shaper_binding *binding,
int i;
for (i = 0; i < nr_shapers; ++i) {
+ struct net_shaper *new;
+
index = net_shaper_handle_to_index(&shapers[i].handle);
cur = xa_load(&hierarchy->shapers, index);
if (WARN_ON_ONCE(!cur))
continue;
- /* Successful update: drop the tentative mark
- * and update the hierarchy container.
+ /* If the shaper is already visible try to allocate a new
+ * entry so that we don't cause torn reads under RCU.
+ * This tiny GFP_KERNEL alloc should never fail, really,
+ * but if it does just override, the torn read is acceptable.
*/
- *cur = shapers[i];
- smp_wmb();
- xa_set_mark(&hierarchy->shapers, index, NET_SHAPER_VALID);
+ new = NULL;
+ if (xa_get_mark(&hierarchy->shapers, index, NET_SHAPER_VALID))
+ new = kmalloc_obj(*new);
+
+ if (new) {
+ *new = shapers[i];
+ smp_wmb();
+ /* Replacing an entry, xa_store() can't fail */
+ WARN_ON(xa_err(xa_store(&hierarchy->shapers, index,
+ new, GFP_KERNEL)));
+ kfree_rcu(cur, rcu);
+ } else {
+ *cur = shapers[i];
+ smp_wmb();
+ xa_set_mark(&hierarchy->shapers, index,
+ NET_SHAPER_VALID);
+ }
}
}
--
2.54.0
next prev parent reply other threads:[~2026-05-06 0:06 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-05-06 0:06 [PATCH net 00/12] net: shaper: fix various minor bugs Jakub Kicinski
2026-05-06 0:06 ` [PATCH net 01/12] net: shaper: drop redundant xa_lock() bracketing Jakub Kicinski
2026-05-06 15:30 ` Paolo Abeni
2026-05-06 22:33 ` Jakub Kicinski
2026-05-07 7:10 ` Paolo Abeni
2026-05-06 0:06 ` [PATCH net 02/12] net: shaper: flip the polarity of the valid flag Jakub Kicinski
2026-05-06 0:06 ` [PATCH net 03/12] net: shaper: fix trivial ordering issue in net_shaper_commit() Jakub Kicinski
2026-05-06 0:06 ` Jakub Kicinski [this message]
2026-05-06 15:22 ` [PATCH net 04/12] net: shaper: try to avoid violating RCU Paolo Abeni
2026-05-06 15:32 ` Paolo Abeni
2026-05-06 22:35 ` Jakub Kicinski
2026-05-06 0:06 ` [PATCH net 05/12] net: shaper: reject duplicate leaves in GROUP request Jakub Kicinski
2026-05-06 0:06 ` [PATCH net 06/12] selftests: drv-net: add shaper test for duplicate leaves Jakub Kicinski
2026-05-06 16:40 ` Breno Leitao
2026-05-06 22:35 ` Jakub Kicinski
2026-05-06 0:06 ` [PATCH net 07/12] net: shaper: set ret to -ENOMEM when genlmsg_new() fails in group_doit Jakub Kicinski
2026-05-06 0:06 ` [PATCH net 08/12] net: shaper: fix undersized reply skb allocation in GROUP command Jakub Kicinski
2026-05-06 0:06 ` [PATCH net 09/12] tools: ynl: add scope qualifier for definitions Jakub Kicinski
2026-05-06 2:32 ` Jakub Kicinski
2026-05-06 0:06 ` [PATCH net 10/12] net: shaper: reject handle IDs exceeding internal bit-width Jakub Kicinski
2026-05-06 0:06 ` [PATCH net 11/12] net: shaper: enforce singleton NETDEV scope with id 0 Jakub Kicinski
2026-05-06 0:06 ` [PATCH net 12/12] net: shaper: reject QUEUE scope handle with missing id Jakub Kicinski
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20260506000628.1501691-5-kuba@kernel.org \
--to=kuba@kernel.org \
--cc=andrew+netdev@lunn.ch \
--cc=davem@davemloft.net \
--cc=edumazet@google.com \
--cc=horms@kernel.org \
--cc=linux-kselftest@vger.kernel.org \
--cc=netdev@vger.kernel.org \
--cc=pabeni@redhat.com \
--cc=shuah@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox