| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <20250609173015.3851695-1-skhawaja@google.com>
Date: Mon, 9 Jun 2025 17:30:15 +0000
From: Samiullah Khawaja <skhawaja@...gle.com>
To: Jakub Kicinski <kuba@...nel.org>, "David S . Miller " <davem@...emloft.net>,
Eric Dumazet <edumazet@...gle.com>, Paolo Abeni <pabeni@...hat.com>, almasrymina@...gle.com,
willemb@...gle.com, jdamato@...tly.com, mkarsten@...terloo.ca,
weiwan@...gle.com
Cc: netdev@...r.kernel.org, skhawaja@...gle.com
Subject: [PATCH net-next v3] net: stop napi kthreads when THREADED napi is disabled
Once the THREADED napi is disabled, the napi kthread should also be
stopped. Keeping the kthread intact after disabling THREADED napi makes
the PID of this kthread show up in the output of netlink 'napi-get' and
ps -ef output.
The is discussed in the patch below:
https://lore.kernel.org/all/20250502191548.559cc416@kernel.org
NAPI kthread should stop only if,
- There are no pending napi poll scheduled for this thread.
- There are no new napi poll scheduled for this thread while it has
stopped.
- The ____napi_schedule can correctly fallback to the softirq for napi
polling.
Since napi_schedule_prep provides mutual exclusion over STATE_SCHED bit,
it is safe to unset the STATE_THREADED when SCHED_THREADED is set or the
SCHED bit is not set. SCHED_THREADED being set means that SCHED is
already set and the kthread owns this napi.
To disable threaded napi, unset STATE_THREADED bit safely if
SCHED_THREADED is set or SCHED is unset. Once STATE_THREADED is unset
safely then wait for the kthread to unset the SCHED_THREADED bit so it
safe to stop the kthread.
Add a new test in nl_netdev to verify this behaviour.
Tested:
./tools/testing/selftests/net/nl_netdev.py
TAP version 13
1..6
ok 1 nl_netdev.empty_check
ok 2 nl_netdev.lo_check
ok 3 nl_netdev.page_pool_check
ok 4 nl_netdev.napi_list_check
ok 5 nl_netdev.dev_set_threaded
ok 6 nl_netdev.nsim_rxq_reset_down
# Totals: pass:6 fail:0 xfail:0 xpass:0 skip:0 error:0
Ran neper for 300 seconds and did enable/disable of thread napi in a
loop continuously.
Signed-off-by: Samiullah Khawaja <skhawaja@...gle.com>
---
v3:
- Rework the napi threaded disable logic to unset the threaded state
when SCHED_THREADED is set or napi is idle.
v2:
- Handle the race where the STATE_THREADED is disabled and kthread is
stopped while the ____napi_schedule code has already checked the
STATE_THREADED and tries to wakeup napi kthread that is stopped.
net/core/dev.c | 45 ++++++++++++++++++++++--
tools/testing/selftests/net/nl_netdev.py | 38 ++++++++++++++++++--
2 files changed, 79 insertions(+), 4 deletions(-)
diff --git a/net/core/dev.c b/net/core/dev.c
index d0563ddff6ca..3a0165b22f77 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -6888,6 +6888,43 @@ static enum hrtimer_restart napi_watchdog(struct hrtimer *timer)
return HRTIMER_NORESTART;
}
+static void napi_stop_kthread(struct napi_struct *napi)
+{
+ unsigned long val, new;
+
+ /* Wait until the napi STATE_THREADED is unset. */
+ while (true) {
+ val = READ_ONCE(napi->state);
+
+ /* If napi kthread own this napi or the napi is idle,
+ * STATE_THREADED can be unset here.
+ */
+ if ((val & NAPIF_STATE_SCHED_THREADED) ||
+ !(val & NAPIF_STATE_SCHED)) {
+ new = val & (~NAPIF_STATE_THREADED);
+ } else {
+ msleep(20);
+ continue;
+ }
+
+ if (try_cmpxchg(&napi->state, &val, new))
+ break;
+ }
+
+ /* Once STATE_THREADED is unset, wait for SCHED_THREADED to be unset by
+ * the kthread.
+ */
+ while (true) {
+ if (!test_bit(NAPIF_STATE_SCHED_THREADED, &napi->state))
+ break;
+
+ msleep(20);
+ }
+
+ kthread_stop(napi->thread);
+ napi->thread = NULL;
+}
+
int dev_set_threaded(struct net_device *dev, bool threaded)
{
struct napi_struct *napi;
@@ -6923,8 +6960,12 @@ int dev_set_threaded(struct net_device *dev, bool threaded)
* softirq mode will happen in the next round of napi_schedule().
* This should not cause hiccups/stalls to the live traffic.
*/
- list_for_each_entry(napi, &dev->napi_list, dev_list)
- assign_bit(NAPI_STATE_THREADED, &napi->state, threaded);
+ list_for_each_entry(napi, &dev->napi_list, dev_list) {
+ if (!threaded && napi->thread)
+ napi_stop_kthread(napi);
+ else
+ assign_bit(NAPI_STATE_THREADED, &napi->state, threaded);
+ }
return err;
}
diff --git a/tools/testing/selftests/net/nl_netdev.py b/tools/testing/selftests/net/nl_netdev.py
index beaee5e4e2aa..c9109627a741 100755
--- a/tools/testing/selftests/net/nl_netdev.py
+++ b/tools/testing/selftests/net/nl_netdev.py
@@ -2,8 +2,9 @@
# SPDX-License-Identifier: GPL-2.0
import time
+from os import system
from lib.py import ksft_run, ksft_exit, ksft_pr
-from lib.py import ksft_eq, ksft_ge, ksft_busy_wait
+from lib.py import ksft_eq, ksft_ge, ksft_ne, ksft_busy_wait
from lib.py import NetdevFamily, NetdevSimDev, ip
@@ -34,6 +35,39 @@ def napi_list_check(nf) -> None:
ksft_eq(len(napis), 100,
comment=f"queue count after reset queue {q} mode {i}")
+def dev_set_threaded(nf) -> None:
+ """
+ Test that verifies various cases of napi threaded
+ set and unset at device level using sysfs.
+ """
+ with NetdevSimDev(queue_count=2) as nsimdev:
+ nsim = nsimdev.nsims[0]
+
+ ip(f"link set dev {nsim.ifname} up")
+
+ napis = nf.napi_get({'ifindex': nsim.ifindex}, dump=True)
+ ksft_eq(len(napis), 2)
+
+ napi0_id = napis[0]['id']
+ napi1_id = napis[1]['id']
+
+ # set threaded
+ system(f"echo 1 > /sys/class/net/{nsim.ifname}/threaded")
+
+ # check napi threaded is set for both napis
+ napi0 = nf.napi_get({'id': napi0_id})
+ ksft_ne(napi0.get('pid'), None)
+ napi1 = nf.napi_get({'id': napi1_id})
+ ksft_ne(napi1.get('pid'), None)
+
+ # unset threaded
+ system(f"echo 0 > /sys/class/net/{nsim.ifname}/threaded")
+
+ # check napi threaded is unset for both napis
+ napi0 = nf.napi_get({'id': napi0_id})
+ ksft_eq(napi0.get('pid'), None)
+ napi1 = nf.napi_get({'id': napi1_id})
+ ksft_eq(napi1.get('pid'), None)
def nsim_rxq_reset_down(nf) -> None:
"""
@@ -122,7 +156,7 @@ def page_pool_check(nf) -> None:
def main() -> None:
nf = NetdevFamily()
ksft_run([empty_check, lo_check, page_pool_check, napi_list_check,
- nsim_rxq_reset_down],
+ dev_set_threaded, nsim_rxq_reset_down],
args=(nf, ))
ksft_exit()
base-commit: b65999e7238e6f2a48dc77c8c2109c48318ff41b
--
2.49.0.1266.g31b7d2e469-goog
Powered by blists - more mailing lists