Re: [PATCH net] tun/tap & vhost-net: make qdisc backpressure opt-in via IFF_BACKPRESSURE
From: "Michael S. Tsirkin" <mst@redhat.com>
Date: 2026-07-04 12:52:17
Also in:
linux-doc, lkml
On Sat, Jul 04, 2026 at 01:20:58PM +0200, Simon Schippers wrote:
Commit 1d6e569b7d0c ("tun/tap & vhost-net: avoid ptr_ring tail-drop
when a qdisc is present") did not show a relevant performance regression
in my testing but on Brett Sheffield's librecast testbed it shows a
significant performance drop. The regression can be pinpointed when
multiple iperf3 TCP threads are sending. For 8 threads the performance
dropped from 13.5 Gbit/s to 9.13 Gbit/s. This is the reason why this
patch makes the qdisc backpressure behavior opt-in.
One option to accomplish the opt-in would be to set the default qdisc to
noqueue at init. However this may also break userspace as users might
have chosen a custom qdisc even though most of the qdiscs did nothing
for tun/tap in the past due to missing backpressure...
This is the reason why in this patch, the flag IFF_BACKPRESSURE is
introduced instead which is required to enable the backpressure logic.
This means the stopping logic in tun_net_xmit() and the waking logic in
__tun_wake_queue() are skipped if the flag is disabled.
In tun_set_iff(), netif_tx_wake_all_queues() is replaced with looping
over all tfiles in which the netdev queues are woken and cons_cnt is
reset while the consumer_lock and producer_lock are held. This is to
ensure that tun_net_xmit() can not stop the queue concurrently, avoiding
a possible stall.
The documentation in tuntap.rst is updated accordingly.
Fixes: 1d6e569b7d0c ("tun/tap & vhost-net: avoid ptr_ring tail-drop when a qdisc is present")
Reported-by: Brett Sheffield <redacted>
Closes: https://lore.kernel.org/netdev/akVnoOYQOrt8k-Gu@karahi.librecast.net/T/#u (local)
Signed-off-by: Simon Schippers <redacted>the patch itself makes sense Acked-by: Michael S. Tsirkin <mst@redhat.com> The issue is it would ideally be in next, but we need it now to fix the regression introduced by 1d6e569b7d0c.
quoted hunk ↗ jump to hunk
--- Documentation/networking/tuntap.rst | 17 +++++++++++++++++ drivers/net/tun.c | 29 +++++++++++++++++++++++------ include/uapi/linux/if_tun.h | 1 + tools/include/uapi/linux/if_tun.h | 1 + 4 files changed, 42 insertions(+), 6 deletions(-)diff --git a/Documentation/networking/tuntap.rst b/Documentation/networking/tuntap.rst index 4d7087f727be..599264825dd2 100644 --- a/Documentation/networking/tuntap.rst +++ b/Documentation/networking/tuntap.rst@@ -206,6 +206,23 @@ enable is true we enable it, otherwise we disable it:: return ioctl(fd, TUNSETQUEUE, (void *)&ifr); } +3.4 qdisc backpressure +---------------------- + +Starting with Linux 7.2, IFF_BACKPRESSURE can be set to enable qdisc +backpressure. Without it, TX drops occur when the internal ring buffer is +full. With it, the kernel stops the TX queue instead, letting the qdisc +hold packets. Drops only occur as a rare race. This can benefit protocols +like TCP that react to drops. Backpressure requires a qdisc to be +attached and has no effect with noqueue. + +The TUN/TAP ring buffer size can be reduced alongside this flag to +further shift buffering into the qdisc and reduce bufferbloat, but comes +at possible performance cost. + +When running multiple network streams in parallel, the flag may reduce +performance due to the extra overhead of the backpressure mechanism. + Universal TUN/TAP device driver Frequently Asked Question =========================================================diff --git a/drivers/net/tun.c b/drivers/net/tun.c index ffbe6f13fb1f..3bf8a73a0816 100644 --- a/drivers/net/tun.c +++ b/drivers/net/tun.c@@ -98,7 +98,8 @@ static void tun_default_link_ksettings(struct net_device *dev, #define TUN_FASYNC IFF_ATTACH_QUEUE #define TUN_FEATURES (IFF_NO_PI | IFF_ONE_QUEUE | IFF_VNET_HDR | \ - IFF_MULTI_QUEUE | IFF_NAPI | IFF_NAPI_FRAGS) + IFF_MULTI_QUEUE | IFF_NAPI | IFF_NAPI_FRAGS | \ + IFF_BACKPRESSURE) #define GOODCOPY_LEN 128@@ -1077,7 +1078,8 @@ static netdev_tx_t tun_net_xmit(struct sk_buff *skb, struct net_device *dev) spin_lock(&tfile->tx_ring.producer_lock); ret = __ptr_ring_produce(&tfile->tx_ring, skb); - if (!qdisc_txq_has_no_queue(queue) && + if ((tun->flags & IFF_BACKPRESSURE) && + !qdisc_txq_has_no_queue(queue) && __ptr_ring_check_produce(&tfile->tx_ring) == -ENOSPC) { netif_tx_stop_queue(queue); /* Paired with smp_mb() in __tun_wake_queue() */@@ -2151,8 +2153,12 @@ static ssize_t tun_put_user(struct tun_struct *tun, static void __tun_wake_queue(struct tun_struct *tun, struct tun_file *tfile, int consumed) { - struct netdev_queue *txq = netdev_get_tx_queue(tun->dev, - tfile->queue_index); + struct netdev_queue *txq; + + if (!(tun->flags & IFF_BACKPRESSURE)) + return; + + txq = netdev_get_tx_queue(tun->dev, tfile->queue_index); /* Paired with smp_mb__after_atomic() in tun_net_xmit() */ smp_mb();@@ -2893,8 +2899,19 @@ static int tun_set_iff(struct net *net, struct file *file, struct ifreq *ifr) /* Make sure persistent devices do not get stuck in * xoff state. */ - if (netif_running(tun->dev)) - netif_tx_wake_all_queues(tun->dev); + if (netif_running(tun->dev)) { + for (int i = 0; i < tun->numqueues; i++) { + struct tun_file *i_tfile; + + i_tfile = rtnl_dereference(tun->tfiles[i]); + spin_lock_bh(&i_tfile->tx_ring.consumer_lock); + spin_lock(&i_tfile->tx_ring.producer_lock); + netif_wake_subqueue(tun->dev, i_tfile->queue_index); + i_tfile->cons_cnt = 0; + spin_unlock(&i_tfile->tx_ring.producer_lock); + spin_unlock_bh(&i_tfile->tx_ring.consumer_lock); + } + } strscpy(ifr->ifr_name, tun->dev->name); return 0;diff --git a/include/uapi/linux/if_tun.h b/include/uapi/linux/if_tun.h index 79d53c7a1ebd..73a77141315c 100644 --- a/include/uapi/linux/if_tun.h +++ b/include/uapi/linux/if_tun.h@@ -69,6 +69,7 @@ #define IFF_NAPI_FRAGS 0x0020 /* Used in TUNSETIFF to bring up tun/tap without carrier */ #define IFF_NO_CARRIER 0x0040 +#define IFF_BACKPRESSURE 0x0080 #define IFF_NO_PI 0x1000 /* This flag has no real effect */ #define IFF_ONE_QUEUE 0x2000diff --git a/tools/include/uapi/linux/if_tun.h b/tools/include/uapi/linux/if_tun.h index 2ec07de1d73b..97b670f5bc0a 100644 --- a/tools/include/uapi/linux/if_tun.h +++ b/tools/include/uapi/linux/if_tun.h@@ -67,6 +67,7 @@ #define IFF_TAP 0x0002 #define IFF_NAPI 0x0010 #define IFF_NAPI_FRAGS 0x0020 +#define IFF_BACKPRESSURE 0x0080 #define IFF_NO_PI 0x1000 /* This flag has no real effect */ #define IFF_ONE_QUEUE 0x2000-- 2.43.0