From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: stable@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
patches@lists.linux.dev, Eric Dumazet <edumazet@google.com>,
Antoine Tenart <atenart@kernel.org>,
Jakub Kicinski <kuba@kernel.org>, Sasha Levin <sashal@kernel.org>
Subject: [PATCH 6.1 05/64] net: add exit_batch_rtnl() method
Date: Tue, 21 Jan 2025 18:52:04 +0100 [thread overview]
Message-ID: <20250121174521.774480523@linuxfoundation.org> (raw)
In-Reply-To: <20250121174521.568417761@linuxfoundation.org>
6.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Eric Dumazet <edumazet@google.com>
[ Upstream commit fd4f101edbd9f99567ab2adb1f2169579ede7c13 ]
Many (struct pernet_operations)->exit_batch() methods have
to acquire rtnl.
In presence of rtnl mutex pressure, this makes cleanup_net()
very slow.
This patch adds a new exit_batch_rtnl() method to reduce
number of rtnl acquisitions from cleanup_net().
exit_batch_rtnl() handlers are called while rtnl is locked,
and devices to be killed can be queued in a list provided
as their second argument.
A single unregister_netdevice_many() is called right
before rtnl is released.
exit_batch_rtnl() handlers are called before ->exit() and
->exit_batch() handlers.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Antoine Tenart <atenart@kernel.org>
Link: https://lore.kernel.org/r/20240206144313.2050392-2-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Stable-dep-of: 46841c7053e6 ("gtp: Use for_each_netdev_rcu() in gtp_genl_dump_pdp().")
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
include/net/net_namespace.h | 3 +++
net/core/net_namespace.c | 31 ++++++++++++++++++++++++++++++-
2 files changed, 33 insertions(+), 1 deletion(-)
diff --git a/include/net/net_namespace.h b/include/net/net_namespace.h
index 7ca4b7af57ca6..17c7a88418345 100644
--- a/include/net/net_namespace.h
+++ b/include/net/net_namespace.h
@@ -426,6 +426,9 @@ struct pernet_operations {
void (*pre_exit)(struct net *net);
void (*exit)(struct net *net);
void (*exit_batch)(struct list_head *net_exit_list);
+ /* Following method is called with RTNL held. */
+ void (*exit_batch_rtnl)(struct list_head *net_exit_list,
+ struct list_head *dev_kill_list);
unsigned int *id;
size_t size;
};
diff --git a/net/core/net_namespace.c b/net/core/net_namespace.c
index 6c6d2a785c004..abf1e1751d6c8 100644
--- a/net/core/net_namespace.c
+++ b/net/core/net_namespace.c
@@ -314,8 +314,9 @@ static __net_init int setup_net(struct net *net, struct user_namespace *user_ns)
{
/* Must be called with pernet_ops_rwsem held */
const struct pernet_operations *ops, *saved_ops;
- int error = 0;
LIST_HEAD(net_exit_list);
+ LIST_HEAD(dev_kill_list);
+ int error = 0;
refcount_set(&net->ns.count, 1);
ref_tracker_dir_init(&net->refcnt_tracker, 128);
@@ -353,6 +354,15 @@ static __net_init int setup_net(struct net *net, struct user_namespace *user_ns)
synchronize_rcu();
+ ops = saved_ops;
+ rtnl_lock();
+ list_for_each_entry_continue_reverse(ops, &pernet_list, list) {
+ if (ops->exit_batch_rtnl)
+ ops->exit_batch_rtnl(&net_exit_list, &dev_kill_list);
+ }
+ unregister_netdevice_many(&dev_kill_list);
+ rtnl_unlock();
+
ops = saved_ops;
list_for_each_entry_continue_reverse(ops, &pernet_list, list)
ops_exit_list(ops, &net_exit_list);
@@ -576,6 +586,7 @@ static void cleanup_net(struct work_struct *work)
struct net *net, *tmp, *last;
struct llist_node *net_kill_list;
LIST_HEAD(net_exit_list);
+ LIST_HEAD(dev_kill_list);
/* Atomically snapshot the list of namespaces to cleanup */
net_kill_list = llist_del_all(&cleanup_list);
@@ -616,6 +627,14 @@ static void cleanup_net(struct work_struct *work)
*/
synchronize_rcu();
+ rtnl_lock();
+ list_for_each_entry_reverse(ops, &pernet_list, list) {
+ if (ops->exit_batch_rtnl)
+ ops->exit_batch_rtnl(&net_exit_list, &dev_kill_list);
+ }
+ unregister_netdevice_many(&dev_kill_list);
+ rtnl_unlock();
+
/* Run all of the network namespace exit methods */
list_for_each_entry_reverse(ops, &pernet_list, list)
ops_exit_list(ops, &net_exit_list);
@@ -1159,7 +1178,17 @@ static void free_exit_list(struct pernet_operations *ops, struct list_head *net_
{
ops_pre_exit_list(ops, net_exit_list);
synchronize_rcu();
+
+ if (ops->exit_batch_rtnl) {
+ LIST_HEAD(dev_kill_list);
+
+ rtnl_lock();
+ ops->exit_batch_rtnl(net_exit_list, &dev_kill_list);
+ unregister_netdevice_many(&dev_kill_list);
+ rtnl_unlock();
+ }
ops_exit_list(ops, net_exit_list);
+
ops_free_list(ops, net_exit_list);
}
--
2.39.5
next prev parent reply other threads:[~2025-01-21 18:04 UTC|newest]
Thread overview: 74+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-01-21 17:51 [PATCH 6.1 00/64] 6.1.127-rc1 review Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 01/64] net: ethernet: ti: cpsw_ale: Fix cpsw_ale_get_field() Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 02/64] bpf: Fix bpf_sk_select_reuseport() memory leak Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 03/64] openvswitch: fix lockup on tx to unregistering netdev with carrier Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 04/64] pktgen: Avoid out-of-bounds access in get_imix_entries Greg Kroah-Hartman
2025-01-21 17:52 ` Greg Kroah-Hartman [this message]
2025-01-21 17:52 ` [PATCH 6.1 06/64] gtp: use exit_batch_rtnl() method Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 07/64] gtp: Use for_each_netdev_rcu() in gtp_genl_dump_pdp() Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 08/64] gtp: Destroy device along with udp sockets netns dismantle Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 09/64] nfp: bpf: prevent integer overflow in nfp_bpf_event_output() Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 10/64] net: xilinx: axienet: Fix IRQ coalescing packet count overflow Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 11/64] net/mlx5: Fix RDMA TX steering prio Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 12/64] net/mlx5: Clear port select structure when fail to create Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 13/64] drm/v3d: Ensure job pointer is set to NULL after job completion Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 14/64] hwmon: (tmp513) Fix division of negative numbers Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 15/64] Revert "mtd: spi-nor: core: replace dummy buswidth from addr to data" Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 16/64] i2c: mux: demux-pinctrl: check initial mux selection, too Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 17/64] i2c: rcar: fix NACK handling when being a target Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 18/64] nvmet: propagate npwg topology Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 19/64] mac802154: check local interfaces before deleting sdata list Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 20/64] hfs: Sanity check the root record Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 21/64] fs: fix missing declaration of init_files Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 22/64] kheaders: Ignore silly-rename files Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 23/64] cachefiles: Parse the "secctx" immediately Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 24/64] scsi: ufs: core: Honor runtime/system PM levels if set by host controller drivers Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 25/64] selftests: tc-testing: reduce rshift value Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 26/64] ACPI: resource: acpi_dev_irq_override(): Check DMI match last Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 27/64] iomap: avoid avoid truncating 64-bit offset to 32 bits Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 28/64] poll_wait: add mb() to fix theoretical race between waitqueue_active() and .poll() Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 29/64] x86/asm: Make serialize() always_inline Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 30/64] ALSA: hda/realtek: Add support for Ayaneo System using CS35L41 HDA Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 31/64] zram: fix potential UAF of zram table Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 32/64] mptcp: be sure to send ack when mptcp-level window re-opens Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 33/64] selftests: mptcp: avoid spurious errors on disconnect Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 34/64] net: ethernet: xgbe: re-add aneg to supported features in PHY quirks Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 35/64] vsock/virtio: discard packets if the transport changes Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 36/64] vsock/virtio: cancel close work in the destructor Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 37/64] vsock: reset socket state when de-assigning the transport Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 38/64] vsock: prevent null-ptr-deref in vsock_*[has_data|has_space] Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 39/64] filemap: avoid truncating 64-bit offset to 32 bits Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 40/64] fs/proc: fix softlockup in __read_vmcore (part 2) Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 41/64] gpiolib: cdev: Fix use after free in lineinfo_changed_notify Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 42/64] pmdomain: imx8mp-blk-ctrl: add missing loop break condition Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 43/64] irqchip: Plug a OF node reference leak in platform_irqchip_probe() Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 44/64] irqchip/gic-v3: Handle CPU_PM_ENTER_FAILED correctly Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 45/64] irqchip/gic-v3-its: Dont enable interrupts in its_irq_set_vcpu_affinity() Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 46/64] hrtimers: Handle CPU state correctly on hotplug Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 47/64] drm/i915/fb: Relax clear color alignment to 64 bytes Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 48/64] Revert "PCI: Use preserve_config in place of pci_flags" Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 49/64] iio: imu: inv_icm42600: fix spi burst write not supported Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 50/64] iio: imu: inv_icm42600: fix timestamps after suspend if sensor is on Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 51/64] iio: adc: rockchip_saradc: fix information leak in triggered buffer Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 52/64] drm/amd/display: Fix out-of-bounds access in dcn21_link_encoder_create Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 53/64] drm/amdgpu: fix usage slab after free Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 54/64] block: fix uaf for flush rq while iterating tags Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 55/64] Revert "drm/amdgpu: rework resume handling for display (v2)" Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 56/64] RDMA/rxe: Fix the qp flush warnings in req Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 57/64] scsi: sg: Fix slab-use-after-free read in sg_release() Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 58/64] Revert "regmap: detach regmap from dev on regmap_exit" Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 59/64] wifi: ath10k: avoid NULL pointer error during sdio remove Greg Kroah-Hartman
2025-01-21 17:52 ` [PATCH 6.1 60/64] erofs: tidy up EROFS on-disk naming Greg Kroah-Hartman
2025-01-21 17:53 ` [PATCH 6.1 61/64] erofs: handle NONHEAD !delta[1] lclusters gracefully Greg Kroah-Hartman
2025-01-21 17:53 ` [PATCH 6.1 62/64] nfsd: add list_head nf_gc to struct nfsd_file Greg Kroah-Hartman
2025-01-21 17:53 ` [PATCH 6.1 63/64] x86/xen: fix SLS mitigation in xen_hypercall_iret() Greg Kroah-Hartman
2025-01-21 17:53 ` [PATCH 6.1 64/64] net: fix data-races around sk->sk_forward_alloc Greg Kroah-Hartman
2025-01-21 18:44 ` [PATCH 6.1 00/64] 6.1.127-rc1 review Florian Fainelli
2025-01-21 19:32 ` Peter Schneider
2025-01-21 21:57 ` Salvatore Bonaccorso
2025-01-22 7:35 ` Greg Kroah-Hartman
2025-01-22 7:38 ` Greg Kroah-Hartman
2025-01-22 5:32 ` Ron Economos
2025-01-21 23:35 ` Shuah Khan
2025-01-21 23:46 ` SeongJae Park
2025-01-22 20:17 ` [PATCH 6.1] " Hardik Garg
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=20250121174521.774480523@linuxfoundation.org \
--to=gregkh@linuxfoundation.org \
--cc=atenart@kernel.org \
--cc=edumazet@google.com \
--cc=kuba@kernel.org \
--cc=patches@lists.linux.dev \
--cc=sashal@kernel.org \
--cc=stable@vger.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