When virtnet_close is followed by virtnet_open, some TX completions can possibly remain unconsumed, until they are finally processed during the first NAPI poll after the netdev_tx_reset_queue(), resulting in a crash [1]. Commit b96ed2c97c79 ("virtio_net: move netdev_tx_reset_queue() call before RX napi enable") was not sufficient to eliminate all BQL crash scenarios for virtio-net.
This issue can be reproduced with the latest net-next master by running: `while :; do ip l set DEV down; ip l set DEV up; done` under heavy network TX load from inside the machine.
This patch series resolves the issue and also addresses similar existing problems:
(a). Drop netdev_tx_reset_queue() from open/close path. This eliminates the BQL crashes due to the problematic open/close path.
(b). As a result of (a), netdev_tx_reset_queue() is now explicitly required in freeze/restore path. Add netdev_tx_reset_queue() immediately after free_unused_bufs() invocation.
(c). Fix missing resetting in virtnet_tx_resize(). virtnet_tx_resize() has lacked proper resetting since commit c8bd1f7f3e61 ("virtio_net: add support for Byte Queue Limits").
(d). Fix missing resetting in the XDP_SETUP_XSK_POOL path. Similar to (c), this path lacked proper resetting. Call netdev_tx_reset_queue() when virtqueue_reset() has actually recycled unused buffers.
This patch series consists of six commits: [1/6]: Resolves (a) and (b). # also -stable 6.11.y [2/6]: Minor fix to make [4/6] streamlined. [3/6]: Prerequisite for (c). # also -stable 6.11.y [4/6]: Resolves (c) (incl. Prerequisite for (d)) # also -stable 6.11.y [5/6]: Preresuisite for (d). [6/6]: Resolves (d).
Changes for v4: - move netdev_tx_reset_queue() out of free_unused_bufs() - submit to net, not net-next Changes for v3: - replace 'flushed' argument with 'recycle_done' Changes for v2: - add tx queue resetting for (b) to (d) above
v3: https://lore.kernel.org/all/20241204050724.307544-1-koichiro.den@canonical.c... v2: https://lore.kernel.org/all/20241203073025.67065-1-koichiro.den@canonical.co... v1: https://lore.kernel.org/all/20241130181744.3772632-1-koichiro.den@canonical....
[1]: ------------[ cut here ]------------ kernel BUG at lib/dynamic_queue_limits.c:99! Oops: invalid opcode: 0000 [#1] PREEMPT SMP NOPTI CPU: 7 UID: 0 PID: 1598 Comm: ip Tainted: G N 6.12.0net-next_main+ #2 Tainted: [N]=TEST Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), \ BIOS rel-1.16.3-0-ga6ed6b701f0a-prebuilt.qemu.org 04/01/2014 RIP: 0010:dql_completed+0x26b/0x290 Code: b7 c2 49 89 e9 44 89 da 89 c6 4c 89 d7 e8 ed 17 47 00 58 65 ff 0d 4d 27 90 7e 0f 85 fd fe ff ff e8 ea 53 8d ff e9 f3 fe ff ff <0f> 0b 01 d2 44 89 d1 29 d1 ba 00 00 00 00 0f 48 ca e9 28 ff ff ff RSP: 0018:ffffc900002b0d08 EFLAGS: 00010297 RAX: 0000000000000000 RBX: ffff888102398c80 RCX: 0000000080190009 RDX: 0000000000000000 RSI: 000000000000006a RDI: 0000000000000000 RBP: ffff888102398c00 R08: 0000000000000000 R09: 0000000000000000 R10: 00000000000000ca R11: 0000000000015681 R12: 0000000000000001 R13: ffffc900002b0d68 R14: ffff88811115e000 R15: ffff8881107aca40 FS: 00007f41ded69500(0000) GS:ffff888667dc0000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000556ccc2dc1a0 CR3: 0000000104fd8003 CR4: 0000000000772ef0 PKRU: 55555554 Call Trace: <IRQ> ? die+0x32/0x80 ? do_trap+0xd9/0x100 ? dql_completed+0x26b/0x290 ? dql_completed+0x26b/0x290 ? do_error_trap+0x6d/0xb0 ? dql_completed+0x26b/0x290 ? exc_invalid_op+0x4c/0x60 ? dql_completed+0x26b/0x290 ? asm_exc_invalid_op+0x16/0x20 ? dql_completed+0x26b/0x290 __free_old_xmit+0xff/0x170 [virtio_net] free_old_xmit+0x54/0xc0 [virtio_net] virtnet_poll+0xf4/0xe30 [virtio_net] ? __update_load_avg_cfs_rq+0x264/0x2d0 ? update_curr+0x35/0x260 ? reweight_entity+0x1be/0x260 __napi_poll.constprop.0+0x28/0x1c0 net_rx_action+0x329/0x420 ? enqueue_hrtimer+0x35/0x90 ? trace_hardirqs_on+0x1d/0x80 ? kvm_sched_clock_read+0xd/0x20 ? sched_clock+0xc/0x30 ? kvm_sched_clock_read+0xd/0x20 ? sched_clock+0xc/0x30 ? sched_clock_cpu+0xd/0x1a0 handle_softirqs+0x138/0x3e0 do_softirq.part.0+0x89/0xc0 </IRQ> <TASK> __local_bh_enable_ip+0xa7/0xb0 virtnet_open+0xc8/0x310 [virtio_net] __dev_open+0xfa/0x1b0 __dev_change_flags+0x1de/0x250 dev_change_flags+0x22/0x60 do_setlink.isra.0+0x2df/0x10b0 ? rtnetlink_rcv_msg+0x34f/0x3f0 ? netlink_rcv_skb+0x54/0x100 ? netlink_unicast+0x23e/0x390 ? netlink_sendmsg+0x21e/0x490 ? ____sys_sendmsg+0x31b/0x350 ? avc_has_perm_noaudit+0x67/0xf0 ? cred_has_capability.isra.0+0x75/0x110 ? __nla_validate_parse+0x5f/0xee0 ? __pfx___probestub_irq_enable+0x3/0x10 ? __create_object+0x5e/0x90 ? security_capable+0x3b/0x7[I0 rtnl_newlink+0x784/0xaf0 ? avc_has_perm_noaudit+0x67/0xf0 ? cred_has_capability.isra.0+0x75/0x110 ? stack_depot_save_flags+0x24/0x6d0 ? __pfx_rtnl_newlink+0x10/0x10 rtnetlink_rcv_msg+0x34f/0x3f0 ? do_syscall_64+0x6c/0x180 ? entry_SYSCALL_64_after_hwframe+0x76/0x7e ? __pfx_rtnetlink_rcv_msg+0x10/0x10 netlink_rcv_skb+0x54/0x100 netlink_unicast+0x23e/0x390 netlink_sendmsg+0x21e/0x490 ____sys_sendmsg+0x31b/0x350 ? copy_msghdr_from_user+0x6d/0xa0 ___sys_sendmsg+0x86/0xd0 ? __pte_offset_map+0x17/0x160 ? preempt_count_add+0x69/0xa0 ? __call_rcu_common.constprop.0+0x147/0x610 ? preempt_count_add+0x69/0xa0 ? preempt_count_add+0x69/0xa0 ? _raw_spin_trylock+0x13/0x60 ? trace_hardirqs_on+0x1d/0x80 __sys_sendmsg+0x66/0xc0 do_syscall_64+0x6c/0x180 entry_SYSCALL_64_after_hwframe+0x76/0x7e RIP: 0033:0x7f41defe5b34 Code: 15 e1 12 0f 00 f7 d8 64 89 02 b8 ff ff ff ff eb bf 0f 1f 44 00 00 f3 0f 1e fa 80 3d 35 95 0f 00 00 74 13 b8 2e 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 4c c3 0f 1f 00 55 48 89 e5 48 83 ec 20 89 55 RSP: 002b:00007ffe5336ecc8 EFLAGS: 00000202 ORIG_RAX: 000000000000002e RAX: ffffffffffffffda RBX: 0000000000000003 RCX: 00007f41defe5b34 RDX: 0000000000000000 RSI: 00007ffe5336ed30 RDI: 0000000000000003 RBP: 00007ffe5336eda0 R08: 0000000000000010 R09: 0000000000000001 R10: 00007ffe5336f6f9 R11: 0000000000000202 R12: 0000000000000003 R13: 0000000067452259 R14: 0000556ccc28b040 R15: 0000000000000000 </TASK> [...] ---[ end Kernel panic - not syncing: Fatal exception in interrupt ]---
Koichiro Den (6): virtio_net: correct netdev_tx_reset_queue() invocation point virtio_net: replace vq2rxq with vq2txq where appropriate virtio_ring: add a func argument 'recycle_done' to virtqueue_resize() virtio_net: ensure netdev_tx_reset_queue is called on tx ring resize virtio_ring: add a func argument 'recycle_done' to virtqueue_reset() virtio_net: ensure netdev_tx_reset_queue is called on bind xsk for tx
drivers/net/virtio_net.c | 31 +++++++++++++++++++++++++------ drivers/virtio/virtio_ring.c | 12 ++++++++++-- include/linux/virtio.h | 6 ++++-- 3 files changed, 39 insertions(+), 10 deletions(-)
When virtnet_close is followed by virtnet_open, some TX completions can possibly remain unconsumed, until they are finally processed during the first NAPI poll after the netdev_tx_reset_queue(), resulting in a crash [1]. Commit b96ed2c97c79 ("virtio_net: move netdev_tx_reset_queue() call before RX napi enable") was not sufficient to eliminate all BQL crash cases for virtio-net.
This issue can be reproduced with the latest net-next master by running: `while :; do ip l set DEV down; ip l set DEV up; done` under heavy network TX load from inside the machine.
netdev_tx_reset_queue() can actually be dropped from virtnet_open path; the device is not stopped in any case. For BQL core part, it's just like traffic nearly ceases to exist for some period. For stall detector added to BQL, even if virtnet_close could somehow lead to some TX completions delayed for long, followed by virtnet_open, we can just take it as stall as mentioned in commit 6025b9135f7a ("net: dqs: add NIC stall detector based on BQL"). Note also that users can still reset stall_max via sysfs.
So, drop netdev_tx_reset_queue() from virtnet_enable_queue_pair(). This eliminates the BQL crashes. As a result, netdev_tx_reset_queue() is now explicitly required in freeze/restore path. This patch adds it to immediately after free_unused_bufs(), following the rule of thumb: netdev_tx_reset_queue() should follow any SKB freeing not followed by netdev_tx_completed_queue(). This seems the most consistent and streamlined approach, and now netdev_tx_reset_queue() runs whenever free_unused_bufs() is done.
[1]: ------------[ cut here ]------------ kernel BUG at lib/dynamic_queue_limits.c:99! Oops: invalid opcode: 0000 [#1] PREEMPT SMP NOPTI CPU: 7 UID: 0 PID: 1598 Comm: ip Tainted: G N 6.12.0net-next_main+ #2 Tainted: [N]=TEST Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), \ BIOS rel-1.16.3-0-ga6ed6b701f0a-prebuilt.qemu.org 04/01/2014 RIP: 0010:dql_completed+0x26b/0x290 Code: b7 c2 49 89 e9 44 89 da 89 c6 4c 89 d7 e8 ed 17 47 00 58 65 ff 0d 4d 27 90 7e 0f 85 fd fe ff ff e8 ea 53 8d ff e9 f3 fe ff ff <0f> 0b 01 d2 44 89 d1 29 d1 ba 00 00 00 00 0f 48 ca e9 28 ff ff ff RSP: 0018:ffffc900002b0d08 EFLAGS: 00010297 RAX: 0000000000000000 RBX: ffff888102398c80 RCX: 0000000080190009 RDX: 0000000000000000 RSI: 000000000000006a RDI: 0000000000000000 RBP: ffff888102398c00 R08: 0000000000000000 R09: 0000000000000000 R10: 00000000000000ca R11: 0000000000015681 R12: 0000000000000001 R13: ffffc900002b0d68 R14: ffff88811115e000 R15: ffff8881107aca40 FS: 00007f41ded69500(0000) GS:ffff888667dc0000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000556ccc2dc1a0 CR3: 0000000104fd8003 CR4: 0000000000772ef0 PKRU: 55555554 Call Trace: <IRQ> ? die+0x32/0x80 ? do_trap+0xd9/0x100 ? dql_completed+0x26b/0x290 ? dql_completed+0x26b/0x290 ? do_error_trap+0x6d/0xb0 ? dql_completed+0x26b/0x290 ? exc_invalid_op+0x4c/0x60 ? dql_completed+0x26b/0x290 ? asm_exc_invalid_op+0x16/0x20 ? dql_completed+0x26b/0x290 __free_old_xmit+0xff/0x170 [virtio_net] free_old_xmit+0x54/0xc0 [virtio_net] virtnet_poll+0xf4/0xe30 [virtio_net] ? __update_load_avg_cfs_rq+0x264/0x2d0 ? update_curr+0x35/0x260 ? reweight_entity+0x1be/0x260 __napi_poll.constprop.0+0x28/0x1c0 net_rx_action+0x329/0x420 ? enqueue_hrtimer+0x35/0x90 ? trace_hardirqs_on+0x1d/0x80 ? kvm_sched_clock_read+0xd/0x20 ? sched_clock+0xc/0x30 ? kvm_sched_clock_read+0xd/0x20 ? sched_clock+0xc/0x30 ? sched_clock_cpu+0xd/0x1a0 handle_softirqs+0x138/0x3e0 do_softirq.part.0+0x89/0xc0 </IRQ> <TASK> __local_bh_enable_ip+0xa7/0xb0 virtnet_open+0xc8/0x310 [virtio_net] __dev_open+0xfa/0x1b0 __dev_change_flags+0x1de/0x250 dev_change_flags+0x22/0x60 do_setlink.isra.0+0x2df/0x10b0 ? rtnetlink_rcv_msg+0x34f/0x3f0 ? netlink_rcv_skb+0x54/0x100 ? netlink_unicast+0x23e/0x390 ? netlink_sendmsg+0x21e/0x490 ? ____sys_sendmsg+0x31b/0x350 ? avc_has_perm_noaudit+0x67/0xf0 ? cred_has_capability.isra.0+0x75/0x110 ? __nla_validate_parse+0x5f/0xee0 ? __pfx___probestub_irq_enable+0x3/0x10 ? __create_object+0x5e/0x90 ? security_capable+0x3b/0x70 rtnl_newlink+0x784/0xaf0 ? avc_has_perm_noaudit+0x67/0xf0 ? cred_has_capability.isra.0+0x75/0x110 ? stack_depot_save_flags+0x24/0x6d0 ? __pfx_rtnl_newlink+0x10/0x10 rtnetlink_rcv_msg+0x34f/0x3f0 ? do_syscall_64+0x6c/0x180 ? entry_SYSCALL_64_after_hwframe+0x76/0x7e ? __pfx_rtnetlink_rcv_msg+0x10/0x10 netlink_rcv_skb+0x54/0x100 netlink_unicast+0x23e/0x390 netlink_sendmsg+0x21e/0x490 ____sys_sendmsg+0x31b/0x350 ? copy_msghdr_from_user+0x6d/0xa0 ___sys_sendmsg+0x86/0xd0 ? __pte_offset_map+0x17/0x160 ? preempt_count_add+0x69/0xa0 ? __call_rcu_common.constprop.0+0x147/0x610 ? preempt_count_add+0x69/0xa0 ? preempt_count_add+0x69/0xa0 ? _raw_spin_trylock+0x13/0x60 ? trace_hardirqs_on+0x1d/0x80 __sys_sendmsg+0x66/0xc0 do_syscall_64+0x6c/0x180 entry_SYSCALL_64_after_hwframe+0x76/0x7e RIP: 0033:0x7f41defe5b34 Code: 15 e1 12 0f 00 f7 d8 64 89 02 b8 ff ff ff ff eb bf 0f 1f 44 00 00 f3 0f 1e fa 80 3d 35 95 0f 00 00 74 13 b8 2e 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 4c c3 0f 1f 00 55 48 89 e5 48 83 ec 20 89 55 RSP: 002b:00007ffe5336ecc8 EFLAGS: 00000202 ORIG_RAX: 000000000000002e RAX: ffffffffffffffda RBX: 0000000000000003 RCX: 00007f41defe5b34 RDX: 0000000000000000 RSI: 00007ffe5336ed30 RDI: 0000000000000003 RBP: 00007ffe5336eda0 R08: 0000000000000010 R09: 0000000000000001 R10: 00007ffe5336f6f9 R11: 0000000000000202 R12: 0000000000000003 R13: 0000000067452259 R14: 0000556ccc28b040 R15: 0000000000000000 </TASK> [...] ---[ end Kernel panic - not syncing: Fatal exception in interrupt ]---
Fixes: c8bd1f7f3e61 ("virtio_net: add support for Byte Queue Limits") Cc: stable@vger.kernel.org # v6.11+ Signed-off-by: Koichiro Den koichiro.den@canonical.com --- drivers/net/virtio_net.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-)
diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c index 64c87bb48a41..6e0925f7f182 100644 --- a/drivers/net/virtio_net.c +++ b/drivers/net/virtio_net.c @@ -3054,7 +3054,6 @@ static int virtnet_enable_queue_pair(struct virtnet_info *vi, int qp_index) if (err < 0) goto err_xdp_reg_mem_model;
- netdev_tx_reset_queue(netdev_get_tx_queue(vi->dev, qp_index)); virtnet_napi_enable(vi->rq[qp_index].vq, &vi->rq[qp_index].napi); virtnet_napi_tx_enable(vi, vi->sq[qp_index].vq, &vi->sq[qp_index].napi);
@@ -6966,11 +6965,20 @@ static int virtnet_probe(struct virtio_device *vdev)
static void remove_vq_common(struct virtnet_info *vi) { + int i; + virtio_reset_device(vi->vdev);
/* Free unused buffers in both send and recv, if any. */ free_unused_bufs(vi);
+ /* + * Rule of thumb is netdev_tx_reset_queue() should follow any + * skb freeing not followed by netdev_tx_completed_queue() + */ + for (i = 0; i < vi->max_queue_pairs; i++) + netdev_tx_reset_queue(netdev_get_tx_queue(vi->dev, i)); + free_receive_bufs(vi);
free_receive_page_frags(vi);
On 12/6/24 02:10, Koichiro Den wrote:
When virtnet_close is followed by virtnet_open, some TX completions can possibly remain unconsumed, until they are finally processed during the first NAPI poll after the netdev_tx_reset_queue(), resulting in a crash [1]. Commit b96ed2c97c79 ("virtio_net: move netdev_tx_reset_queue() call before RX napi enable") was not sufficient to eliminate all BQL crash cases for virtio-net.
This issue can be reproduced with the latest net-next master by running: `while :; do ip l set DEV down; ip l set DEV up; done` under heavy network TX load from inside the machine.
netdev_tx_reset_queue() can actually be dropped from virtnet_open path; the device is not stopped in any case. For BQL core part, it's just like traffic nearly ceases to exist for some period. For stall detector added to BQL, even if virtnet_close could somehow lead to some TX completions delayed for long, followed by virtnet_open, we can just take it as stall as mentioned in commit 6025b9135f7a ("net: dqs: add NIC stall detector based on BQL"). Note also that users can still reset stall_max via sysfs.
So, drop netdev_tx_reset_queue() from virtnet_enable_queue_pair(). This eliminates the BQL crashes. As a result, netdev_tx_reset_queue() is now explicitly required in freeze/restore path. This patch adds it to immediately after free_unused_bufs(), following the rule of thumb: netdev_tx_reset_queue() should follow any SKB freeing not followed by netdev_tx_completed_queue(). This seems the most consistent and streamlined approach, and now netdev_tx_reset_queue() runs whenever free_unused_bufs() is done.
[1]: ------------[ cut here ]------------ kernel BUG at lib/dynamic_queue_limits.c:99! Oops: invalid opcode: 0000 [#1] PREEMPT SMP NOPTI CPU: 7 UID: 0 PID: 1598 Comm: ip Tainted: G N 6.12.0net-next_main+ #2 Tainted: [N]=TEST Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), \ BIOS rel-1.16.3-0-ga6ed6b701f0a-prebuilt.qemu.org 04/01/2014 RIP: 0010:dql_completed+0x26b/0x290 Code: b7 c2 49 89 e9 44 89 da 89 c6 4c 89 d7 e8 ed 17 47 00 58 65 ff 0d 4d 27 90 7e 0f 85 fd fe ff ff e8 ea 53 8d ff e9 f3 fe ff ff <0f> 0b 01 d2 44 89 d1 29 d1 ba 00 00 00 00 0f 48 ca e9 28 ff ff ff RSP: 0018:ffffc900002b0d08 EFLAGS: 00010297 RAX: 0000000000000000 RBX: ffff888102398c80 RCX: 0000000080190009 RDX: 0000000000000000 RSI: 000000000000006a RDI: 0000000000000000 RBP: ffff888102398c00 R08: 0000000000000000 R09: 0000000000000000 R10: 00000000000000ca R11: 0000000000015681 R12: 0000000000000001 R13: ffffc900002b0d68 R14: ffff88811115e000 R15: ffff8881107aca40 FS: 00007f41ded69500(0000) GS:ffff888667dc0000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000556ccc2dc1a0 CR3: 0000000104fd8003 CR4: 0000000000772ef0 PKRU: 55555554 Call Trace:
<IRQ> ? die+0x32/0x80 ? do_trap+0xd9/0x100 ? dql_completed+0x26b/0x290 ? dql_completed+0x26b/0x290 ? do_error_trap+0x6d/0xb0 ? dql_completed+0x26b/0x290 ? exc_invalid_op+0x4c/0x60 ? dql_completed+0x26b/0x290 ? asm_exc_invalid_op+0x16/0x20 ? dql_completed+0x26b/0x290 __free_old_xmit+0xff/0x170 [virtio_net] free_old_xmit+0x54/0xc0 [virtio_net] virtnet_poll+0xf4/0xe30 [virtio_net] ? __update_load_avg_cfs_rq+0x264/0x2d0 ? update_curr+0x35/0x260 ? reweight_entity+0x1be/0x260 __napi_poll.constprop.0+0x28/0x1c0 net_rx_action+0x329/0x420 ? enqueue_hrtimer+0x35/0x90 ? trace_hardirqs_on+0x1d/0x80 ? kvm_sched_clock_read+0xd/0x20 ? sched_clock+0xc/0x30 ? kvm_sched_clock_read+0xd/0x20 ? sched_clock+0xc/0x30 ? sched_clock_cpu+0xd/0x1a0 handle_softirqs+0x138/0x3e0 do_softirq.part.0+0x89/0xc0 </IRQ> <TASK> __local_bh_enable_ip+0xa7/0xb0 virtnet_open+0xc8/0x310 [virtio_net] __dev_open+0xfa/0x1b0 __dev_change_flags+0x1de/0x250 dev_change_flags+0x22/0x60 do_setlink.isra.0+0x2df/0x10b0 ? rtnetlink_rcv_msg+0x34f/0x3f0 ? netlink_rcv_skb+0x54/0x100 ? netlink_unicast+0x23e/0x390 ? netlink_sendmsg+0x21e/0x490 ? ____sys_sendmsg+0x31b/0x350 ? avc_has_perm_noaudit+0x67/0xf0 ? cred_has_capability.isra.0+0x75/0x110 ? __nla_validate_parse+0x5f/0xee0 ? __pfx___probestub_irq_enable+0x3/0x10 ? __create_object+0x5e/0x90 ? security_capable+0x3b/0x70 rtnl_newlink+0x784/0xaf0 ? avc_has_perm_noaudit+0x67/0xf0 ? cred_has_capability.isra.0+0x75/0x110 ? stack_depot_save_flags+0x24/0x6d0 ? __pfx_rtnl_newlink+0x10/0x10 rtnetlink_rcv_msg+0x34f/0x3f0 ? do_syscall_64+0x6c/0x180 ? entry_SYSCALL_64_after_hwframe+0x76/0x7e ? __pfx_rtnetlink_rcv_msg+0x10/0x10 netlink_rcv_skb+0x54/0x100 netlink_unicast+0x23e/0x390 netlink_sendmsg+0x21e/0x490 ____sys_sendmsg+0x31b/0x350 ? copy_msghdr_from_user+0x6d/0xa0 ___sys_sendmsg+0x86/0xd0 ? __pte_offset_map+0x17/0x160 ? preempt_count_add+0x69/0xa0 ? __call_rcu_common.constprop.0+0x147/0x610 ? preempt_count_add+0x69/0xa0 ? preempt_count_add+0x69/0xa0 ? _raw_spin_trylock+0x13/0x60 ? trace_hardirqs_on+0x1d/0x80 __sys_sendmsg+0x66/0xc0 do_syscall_64+0x6c/0x180 entry_SYSCALL_64_after_hwframe+0x76/0x7e RIP: 0033:0x7f41defe5b34 Code: 15 e1 12 0f 00 f7 d8 64 89 02 b8 ff ff ff ff eb bf 0f 1f 44 00 00 f3 0f 1e fa 80 3d 35 95 0f 00 00 74 13 b8 2e 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 4c c3 0f 1f 00 55 48 89 e5 48 83 ec 20 89 55 RSP: 002b:00007ffe5336ecc8 EFLAGS: 00000202 ORIG_RAX: 000000000000002e RAX: ffffffffffffffda RBX: 0000000000000003 RCX: 00007f41defe5b34 RDX: 0000000000000000 RSI: 00007ffe5336ed30 RDI: 0000000000000003 RBP: 00007ffe5336eda0 R08: 0000000000000010 R09: 0000000000000001 R10: 00007ffe5336f6f9 R11: 0000000000000202 R12: 0000000000000003 R13: 0000000067452259 R14: 0000556ccc28b040 R15: 0000000000000000 </TASK> [...] ---[ end Kernel panic - not syncing: Fatal exception in interrupt ]---
IIRC, the above separator used to cause some troubles to backport tools and to 'next' tree import, trimming the SoB area below. I'm going to remove it while applying the series.
Cheers,
Paolo
While not harmful, using vq2rxq where it's always sq appears odd. Replace it with the more appropriate vq2txq for clarity and correctness.
Fixes: 89f86675cb03 ("virtio_net: xsk: tx: support xmit xsk buffer") Signed-off-by: Koichiro Den koichiro.den@canonical.com --- drivers/net/virtio_net.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c index 6e0925f7f182..fc89c5e1a207 100644 --- a/drivers/net/virtio_net.c +++ b/drivers/net/virtio_net.c @@ -6213,7 +6213,7 @@ static void virtnet_sq_free_unused_buf(struct virtqueue *vq, void *buf) { struct virtnet_info *vi = vq->vdev->priv; struct send_queue *sq; - int i = vq2rxq(vq); + int i = vq2txq(vq);
sq = &vi->sq[i];
Hi,
Thanks for your patch.
FYI: kernel test robot notices the stable kernel rule is not satisfied.
The check is based on https://www.kernel.org/doc/html/latest/process/stable-kernel-rules.html#opti...
Rule: add the tag "Cc: stable@vger.kernel.org" in the sign-off area to have the patch automatically included in the stable tree. Subject: [PATCH net v4 2/6] virtio_net: replace vq2rxq with vq2txq where appropriate Link: https://lore.kernel.org/stable/20241206011047.923923-3-koichiro.den%40canoni...
When virtqueue_resize() has actually recycled all unused buffers, additional work may be required in some cases. Relying solely on its return status is fragile, so introduce a new function argument 'recycle_done', which is invoked when the recycle really occurs.
Cc: stable@vger.kernel.org # v6.11+ Signed-off-by: Koichiro Den koichiro.den@canonical.com --- drivers/net/virtio_net.c | 4 ++-- drivers/virtio/virtio_ring.c | 6 +++++- include/linux/virtio.h | 3 ++- 3 files changed, 9 insertions(+), 4 deletions(-)
diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c index fc89c5e1a207..e10bc9e6b072 100644 --- a/drivers/net/virtio_net.c +++ b/drivers/net/virtio_net.c @@ -3331,7 +3331,7 @@ static int virtnet_rx_resize(struct virtnet_info *vi,
virtnet_rx_pause(vi, rq);
- err = virtqueue_resize(rq->vq, ring_num, virtnet_rq_unmap_free_buf); + err = virtqueue_resize(rq->vq, ring_num, virtnet_rq_unmap_free_buf, NULL); if (err) netdev_err(vi->dev, "resize rx fail: rx queue index: %d err: %d\n", qindex, err);
@@ -3394,7 +3394,7 @@ static int virtnet_tx_resize(struct virtnet_info *vi, struct send_queue *sq,
virtnet_tx_pause(vi, sq);
- err = virtqueue_resize(sq->vq, ring_num, virtnet_sq_free_unused_buf); + err = virtqueue_resize(sq->vq, ring_num, virtnet_sq_free_unused_buf, NULL); if (err) netdev_err(vi->dev, "resize tx fail: tx queue index: %d err: %d\n", qindex, err);
diff --git a/drivers/virtio/virtio_ring.c b/drivers/virtio/virtio_ring.c index 82a7d2cbc704..6af8cf6a619e 100644 --- a/drivers/virtio/virtio_ring.c +++ b/drivers/virtio/virtio_ring.c @@ -2772,6 +2772,7 @@ EXPORT_SYMBOL_GPL(vring_create_virtqueue_dma); * @_vq: the struct virtqueue we're talking about. * @num: new ring num * @recycle: callback to recycle unused buffers + * @recycle_done: callback to be invoked when recycle for all unused buffers done * * When it is really necessary to create a new vring, it will set the current vq * into the reset state. Then call the passed callback to recycle the buffer @@ -2792,7 +2793,8 @@ EXPORT_SYMBOL_GPL(vring_create_virtqueue_dma); * */ int virtqueue_resize(struct virtqueue *_vq, u32 num, - void (*recycle)(struct virtqueue *vq, void *buf)) + void (*recycle)(struct virtqueue *vq, void *buf), + void (*recycle_done)(struct virtqueue *vq)) { struct vring_virtqueue *vq = to_vvq(_vq); int err; @@ -2809,6 +2811,8 @@ int virtqueue_resize(struct virtqueue *_vq, u32 num, err = virtqueue_disable_and_recycle(_vq, recycle); if (err) return err; + if (recycle_done) + recycle_done(_vq);
if (vq->packed_ring) err = virtqueue_resize_packed(_vq, num); diff --git a/include/linux/virtio.h b/include/linux/virtio.h index 57cc4b07fd17..0aa7df4ed5ca 100644 --- a/include/linux/virtio.h +++ b/include/linux/virtio.h @@ -109,7 +109,8 @@ dma_addr_t virtqueue_get_avail_addr(const struct virtqueue *vq); dma_addr_t virtqueue_get_used_addr(const struct virtqueue *vq);
int virtqueue_resize(struct virtqueue *vq, u32 num, - void (*recycle)(struct virtqueue *vq, void *buf)); + void (*recycle)(struct virtqueue *vq, void *buf), + void (*recycle_done)(struct virtqueue *vq)); int virtqueue_reset(struct virtqueue *vq, void (*recycle)(struct virtqueue *vq, void *buf));
virtnet_tx_resize() flushes remaining tx skbs, requiring DQL counters to be reset when flushing has actually occurred. Add virtnet_sq_free_unused_buf_done() as a callback for virtqueue_reset() to handle this.
Fixes: c8bd1f7f3e61 ("virtio_net: add support for Byte Queue Limits") Cc: stable@vger.kernel.org # v6.11+ Signed-off-by: Koichiro Den koichiro.den@canonical.com --- drivers/net/virtio_net.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-)
diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c index e10bc9e6b072..3a0341cc6085 100644 --- a/drivers/net/virtio_net.c +++ b/drivers/net/virtio_net.c @@ -503,6 +503,7 @@ struct virtio_net_common_hdr { static struct virtio_net_common_hdr xsk_hdr;
static void virtnet_sq_free_unused_buf(struct virtqueue *vq, void *buf); +static void virtnet_sq_free_unused_buf_done(struct virtqueue *vq); static int virtnet_xdp_handler(struct bpf_prog *xdp_prog, struct xdp_buff *xdp, struct net_device *dev, unsigned int *xdp_xmit, @@ -3394,7 +3395,8 @@ static int virtnet_tx_resize(struct virtnet_info *vi, struct send_queue *sq,
virtnet_tx_pause(vi, sq);
- err = virtqueue_resize(sq->vq, ring_num, virtnet_sq_free_unused_buf, NULL); + err = virtqueue_resize(sq->vq, ring_num, virtnet_sq_free_unused_buf, + virtnet_sq_free_unused_buf_done); if (err) netdev_err(vi->dev, "resize tx fail: tx queue index: %d err: %d\n", qindex, err);
@@ -6233,6 +6235,14 @@ static void virtnet_sq_free_unused_buf(struct virtqueue *vq, void *buf) } }
+static void virtnet_sq_free_unused_buf_done(struct virtqueue *vq) +{ + struct virtnet_info *vi = vq->vdev->priv; + int i = vq2txq(vq); + + netdev_tx_reset_queue(netdev_get_tx_queue(vi->dev, i)); +} + static void free_unused_bufs(struct virtnet_info *vi) { void *buf;
When virtqueue_reset() has actually recycled all unused buffers, additional work may be required in some cases. Relying solely on its return status is fragile, so introduce a new function argument 'recycle_done', which is invoked when it really occurs.
Signed-off-by: Koichiro Den koichiro.den@canonical.com --- drivers/net/virtio_net.c | 4 ++-- drivers/virtio/virtio_ring.c | 6 +++++- include/linux/virtio.h | 3 ++- 3 files changed, 9 insertions(+), 4 deletions(-)
diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c index 3a0341cc6085..5cf4b2b20431 100644 --- a/drivers/net/virtio_net.c +++ b/drivers/net/virtio_net.c @@ -5711,7 +5711,7 @@ static int virtnet_rq_bind_xsk_pool(struct virtnet_info *vi, struct receive_queu
virtnet_rx_pause(vi, rq);
- err = virtqueue_reset(rq->vq, virtnet_rq_unmap_free_buf); + err = virtqueue_reset(rq->vq, virtnet_rq_unmap_free_buf, NULL); if (err) { netdev_err(vi->dev, "reset rx fail: rx queue index: %d err: %d\n", qindex, err);
@@ -5740,7 +5740,7 @@ static int virtnet_sq_bind_xsk_pool(struct virtnet_info *vi,
virtnet_tx_pause(vi, sq);
- err = virtqueue_reset(sq->vq, virtnet_sq_free_unused_buf); + err = virtqueue_reset(sq->vq, virtnet_sq_free_unused_buf, NULL); if (err) { netdev_err(vi->dev, "reset tx fail: tx queue index: %d err: %d\n", qindex, err); pool = NULL; diff --git a/drivers/virtio/virtio_ring.c b/drivers/virtio/virtio_ring.c index 6af8cf6a619e..fdd2d2b07b5a 100644 --- a/drivers/virtio/virtio_ring.c +++ b/drivers/virtio/virtio_ring.c @@ -2827,6 +2827,7 @@ EXPORT_SYMBOL_GPL(virtqueue_resize); * virtqueue_reset - detach and recycle all unused buffers * @_vq: the struct virtqueue we're talking about. * @recycle: callback to recycle unused buffers + * @recycle_done: callback to be invoked when recycle for all unused buffers done * * Caller must ensure we don't call this with other virtqueue operations * at the same time (except where noted). @@ -2838,7 +2839,8 @@ EXPORT_SYMBOL_GPL(virtqueue_resize); * -EPERM: Operation not permitted */ int virtqueue_reset(struct virtqueue *_vq, - void (*recycle)(struct virtqueue *vq, void *buf)) + void (*recycle)(struct virtqueue *vq, void *buf), + void (*recycle_done)(struct virtqueue *vq)) { struct vring_virtqueue *vq = to_vvq(_vq); int err; @@ -2846,6 +2848,8 @@ int virtqueue_reset(struct virtqueue *_vq, err = virtqueue_disable_and_recycle(_vq, recycle); if (err) return err; + if (recycle_done) + recycle_done(_vq);
if (vq->packed_ring) virtqueue_reinit_packed(vq); diff --git a/include/linux/virtio.h b/include/linux/virtio.h index 0aa7df4ed5ca..dd88682e27e3 100644 --- a/include/linux/virtio.h +++ b/include/linux/virtio.h @@ -112,7 +112,8 @@ int virtqueue_resize(struct virtqueue *vq, u32 num, void (*recycle)(struct virtqueue *vq, void *buf), void (*recycle_done)(struct virtqueue *vq)); int virtqueue_reset(struct virtqueue *vq, - void (*recycle)(struct virtqueue *vq, void *buf)); + void (*recycle)(struct virtqueue *vq, void *buf), + void (*recycle_done)(struct virtqueue *vq));
struct virtio_admin_cmd { __le16 opcode;
virtnet_sq_bind_xsk_pool() flushes tx skbs and then resets tx queue, so DQL counters need to be reset when flushing has actually occurred, Add virtnet_sq_free_unused_buf_done() as a callback for virtqueue_resize() to handle this.
Fixes: 21a4e3ce6dc7 ("virtio_net: xsk: bind/unbind xsk for tx") Signed-off-by: Koichiro Den koichiro.den@canonical.com --- drivers/net/virtio_net.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c index 5cf4b2b20431..7646ddd9bef7 100644 --- a/drivers/net/virtio_net.c +++ b/drivers/net/virtio_net.c @@ -5740,7 +5740,8 @@ static int virtnet_sq_bind_xsk_pool(struct virtnet_info *vi,
virtnet_tx_pause(vi, sq);
- err = virtqueue_reset(sq->vq, virtnet_sq_free_unused_buf, NULL); + err = virtqueue_reset(sq->vq, virtnet_sq_free_unused_buf, + virtnet_sq_free_unused_buf_done); if (err) { netdev_err(vi->dev, "reset tx fail: tx queue index: %d err: %d\n", qindex, err); pool = NULL;
On Fri, Dec 6, 2024 at 9:11 AM Koichiro Den koichiro.den@canonical.com wrote:
When virtnet_close is followed by virtnet_open, some TX completions can possibly remain unconsumed, until they are finally processed during the first NAPI poll after the netdev_tx_reset_queue(), resulting in a crash [1]. Commit b96ed2c97c79 ("virtio_net: move netdev_tx_reset_queue() call before RX napi enable") was not sufficient to eliminate all BQL crash scenarios for virtio-net.
This issue can be reproduced with the latest net-next master by running: `while :; do ip l set DEV down; ip l set DEV up; done` under heavy network TX load from inside the machine.
This patch series resolves the issue and also addresses similar existing problems:
(a). Drop netdev_tx_reset_queue() from open/close path. This eliminates the BQL crashes due to the problematic open/close path.
(b). As a result of (a), netdev_tx_reset_queue() is now explicitly required in freeze/restore path. Add netdev_tx_reset_queue() immediately after free_unused_bufs() invocation.
(c). Fix missing resetting in virtnet_tx_resize(). virtnet_tx_resize() has lacked proper resetting since commit c8bd1f7f3e61 ("virtio_net: add support for Byte Queue Limits").
(d). Fix missing resetting in the XDP_SETUP_XSK_POOL path. Similar to (c), this path lacked proper resetting. Call netdev_tx_reset_queue() when virtqueue_reset() has actually recycled unused buffers.
This patch series consists of six commits: [1/6]: Resolves (a) and (b). # also -stable 6.11.y [2/6]: Minor fix to make [4/6] streamlined. [3/6]: Prerequisite for (c). # also -stable 6.11.y [4/6]: Resolves (c) (incl. Prerequisite for (d)) # also -stable 6.11.y [5/6]: Preresuisite for (d). [6/6]: Resolves (d).
Changes for v4:
- move netdev_tx_reset_queue() out of free_unused_bufs()
- submit to net, not net-next
Changes for v3:
- replace 'flushed' argument with 'recycle_done'
Changes for v2:
- add tx queue resetting for (b) to (d) above
v3: https://lore.kernel.org/all/20241204050724.307544-1-koichiro.den@canonical.c... v2: https://lore.kernel.org/all/20241203073025.67065-1-koichiro.den@canonical.co... v1: https://lore.kernel.org/all/20241130181744.3772632-1-koichiro.den@canonical....
[1]: ------------[ cut here ]------------ kernel BUG at lib/dynamic_queue_limits.c:99! Oops: invalid opcode: 0000 [#1] PREEMPT SMP NOPTI CPU: 7 UID: 0 PID: 1598 Comm: ip Tainted: G N 6.12.0net-next_main+ #2 Tainted: [N]=TEST Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), \ BIOS rel-1.16.3-0-ga6ed6b701f0a-prebuilt.qemu.org 04/01/2014 RIP: 0010:dql_completed+0x26b/0x290 Code: b7 c2 49 89 e9 44 89 da 89 c6 4c 89 d7 e8 ed 17 47 00 58 65 ff 0d 4d 27 90 7e 0f 85 fd fe ff ff e8 ea 53 8d ff e9 f3 fe ff ff <0f> 0b 01 d2 44 89 d1 29 d1 ba 00 00 00 00 0f 48 ca e9 28 ff ff ff RSP: 0018:ffffc900002b0d08 EFLAGS: 00010297 RAX: 0000000000000000 RBX: ffff888102398c80 RCX: 0000000080190009 RDX: 0000000000000000 RSI: 000000000000006a RDI: 0000000000000000 RBP: ffff888102398c00 R08: 0000000000000000 R09: 0000000000000000 R10: 00000000000000ca R11: 0000000000015681 R12: 0000000000000001 R13: ffffc900002b0d68 R14: ffff88811115e000 R15: ffff8881107aca40 FS: 00007f41ded69500(0000) GS:ffff888667dc0000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000556ccc2dc1a0 CR3: 0000000104fd8003 CR4: 0000000000772ef0 PKRU: 55555554 Call Trace:
<IRQ> ? die+0x32/0x80 ? do_trap+0xd9/0x100 ? dql_completed+0x26b/0x290 ? dql_completed+0x26b/0x290 ? do_error_trap+0x6d/0xb0 ? dql_completed+0x26b/0x290 ? exc_invalid_op+0x4c/0x60 ? dql_completed+0x26b/0x290 ? asm_exc_invalid_op+0x16/0x20 ? dql_completed+0x26b/0x290 __free_old_xmit+0xff/0x170 [virtio_net] free_old_xmit+0x54/0xc0 [virtio_net] virtnet_poll+0xf4/0xe30 [virtio_net] ? __update_load_avg_cfs_rq+0x264/0x2d0 ? update_curr+0x35/0x260 ? reweight_entity+0x1be/0x260 __napi_poll.constprop.0+0x28/0x1c0 net_rx_action+0x329/0x420 ? enqueue_hrtimer+0x35/0x90 ? trace_hardirqs_on+0x1d/0x80 ? kvm_sched_clock_read+0xd/0x20 ? sched_clock+0xc/0x30 ? kvm_sched_clock_read+0xd/0x20 ? sched_clock+0xc/0x30 ? sched_clock_cpu+0xd/0x1a0 handle_softirqs+0x138/0x3e0 do_softirq.part.0+0x89/0xc0 </IRQ> <TASK> __local_bh_enable_ip+0xa7/0xb0 virtnet_open+0xc8/0x310 [virtio_net] __dev_open+0xfa/0x1b0 __dev_change_flags+0x1de/0x250 dev_change_flags+0x22/0x60 do_setlink.isra.0+0x2df/0x10b0 ? rtnetlink_rcv_msg+0x34f/0x3f0 ? netlink_rcv_skb+0x54/0x100 ? netlink_unicast+0x23e/0x390 ? netlink_sendmsg+0x21e/0x490 ? ____sys_sendmsg+0x31b/0x350 ? avc_has_perm_noaudit+0x67/0xf0 ? cred_has_capability.isra.0+0x75/0x110 ? __nla_validate_parse+0x5f/0xee0 ? __pfx___probestub_irq_enable+0x3/0x10 ? __create_object+0x5e/0x90 ? security_capable+0x3b/0x7 [I0 rtnl_newlink+0x784/0xaf0 ? avc_has_perm_noaudit+0x67/0xf0 ? cred_has_capability.isra.0+0x75/0x110 ? stack_depot_save_flags+0x24/0x6d0 ? __pfx_rtnl_newlink+0x10/0x10 rtnetlink_rcv_msg+0x34f/0x3f0 ? do_syscall_64+0x6c/0x180 ? entry_SYSCALL_64_after_hwframe+0x76/0x7e ? __pfx_rtnetlink_rcv_msg+0x10/0x10 netlink_rcv_skb+0x54/0x100 netlink_unicast+0x23e/0x390 netlink_sendmsg+0x21e/0x490 ____sys_sendmsg+0x31b/0x350 ? copy_msghdr_from_user+0x6d/0xa0 ___sys_sendmsg+0x86/0xd0 ? __pte_offset_map+0x17/0x160 ? preempt_count_add+0x69/0xa0 ? __call_rcu_common.constprop.0+0x147/0x610 ? preempt_count_add+0x69/0xa0 ? preempt_count_add+0x69/0xa0 ? _raw_spin_trylock+0x13/0x60 ? trace_hardirqs_on+0x1d/0x80 __sys_sendmsg+0x66/0xc0 do_syscall_64+0x6c/0x180 entry_SYSCALL_64_after_hwframe+0x76/0x7e RIP: 0033:0x7f41defe5b34 Code: 15 e1 12 0f 00 f7 d8 64 89 02 b8 ff ff ff ff eb bf 0f 1f 44 00 00 f3 0f 1e fa 80 3d 35 95 0f 00 00 74 13 b8 2e 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 4c c3 0f 1f 00 55 48 89 e5 48 83 ec 20 89 55 RSP: 002b:00007ffe5336ecc8 EFLAGS: 00000202 ORIG_RAX: 000000000000002e RAX: ffffffffffffffda RBX: 0000000000000003 RCX: 00007f41defe5b34 RDX: 0000000000000000 RSI: 00007ffe5336ed30 RDI: 0000000000000003 RBP: 00007ffe5336eda0 R08: 0000000000000010 R09: 0000000000000001 R10: 00007ffe5336f6f9 R11: 0000000000000202 R12: 0000000000000003 R13: 0000000067452259 R14: 0000556ccc28b040 R15: 0000000000000000 </TASK> [...] ---[ end Kernel panic - not syncing: Fatal exception in interrupt ]---
Koichiro Den (6): virtio_net: correct netdev_tx_reset_queue() invocation point virtio_net: replace vq2rxq with vq2txq where appropriate virtio_ring: add a func argument 'recycle_done' to virtqueue_resize() virtio_net: ensure netdev_tx_reset_queue is called on tx ring resize virtio_ring: add a func argument 'recycle_done' to virtqueue_reset() virtio_net: ensure netdev_tx_reset_queue is called on bind xsk for tx
drivers/net/virtio_net.c | 31 +++++++++++++++++++++++++------ drivers/virtio/virtio_ring.c | 12 ++++++++++-- include/linux/virtio.h | 6 ++++-- 3 files changed, 39 insertions(+), 10 deletions(-)
-- 2.43.0
For the series,
Acked-by: Jason Wang jasowang@redhat.com
Thanks
For the series,
Reviewed-by: Xuan Zhuo xuanzhuo@linux.alibaba.com
Thanks
On Fri, 6 Dec 2024 10:10:41 +0900, Koichiro Den koichiro.den@canonical.com wrote:
When virtnet_close is followed by virtnet_open, some TX completions can possibly remain unconsumed, until they are finally processed during the first NAPI poll after the netdev_tx_reset_queue(), resulting in a crash [1]. Commit b96ed2c97c79 ("virtio_net: move netdev_tx_reset_queue() call before RX napi enable") was not sufficient to eliminate all BQL crash scenarios for virtio-net.
This issue can be reproduced with the latest net-next master by running: `while :; do ip l set DEV down; ip l set DEV up; done` under heavy network TX load from inside the machine.
This patch series resolves the issue and also addresses similar existing problems:
(a). Drop netdev_tx_reset_queue() from open/close path. This eliminates the BQL crashes due to the problematic open/close path.
(b). As a result of (a), netdev_tx_reset_queue() is now explicitly required in freeze/restore path. Add netdev_tx_reset_queue() immediately after free_unused_bufs() invocation.
(c). Fix missing resetting in virtnet_tx_resize(). virtnet_tx_resize() has lacked proper resetting since commit c8bd1f7f3e61 ("virtio_net: add support for Byte Queue Limits").
(d). Fix missing resetting in the XDP_SETUP_XSK_POOL path. Similar to (c), this path lacked proper resetting. Call netdev_tx_reset_queue() when virtqueue_reset() has actually recycled unused buffers.
This patch series consists of six commits: [1/6]: Resolves (a) and (b). # also -stable 6.11.y [2/6]: Minor fix to make [4/6] streamlined. [3/6]: Prerequisite for (c). # also -stable 6.11.y [4/6]: Resolves (c) (incl. Prerequisite for (d)) # also -stable 6.11.y [5/6]: Preresuisite for (d). [6/6]: Resolves (d).
Changes for v4:
- move netdev_tx_reset_queue() out of free_unused_bufs()
- submit to net, not net-next
Changes for v3:
- replace 'flushed' argument with 'recycle_done'
Changes for v2:
- add tx queue resetting for (b) to (d) above
v3: https://lore.kernel.org/all/20241204050724.307544-1-koichiro.den@canonical.c... v2: https://lore.kernel.org/all/20241203073025.67065-1-koichiro.den@canonical.co... v1: https://lore.kernel.org/all/20241130181744.3772632-1-koichiro.den@canonical....
[1]: ------------[ cut here ]------------ kernel BUG at lib/dynamic_queue_limits.c:99! Oops: invalid opcode: 0000 [#1] PREEMPT SMP NOPTI CPU: 7 UID: 0 PID: 1598 Comm: ip Tainted: G N 6.12.0net-next_main+ #2 Tainted: [N]=TEST Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), \ BIOS rel-1.16.3-0-ga6ed6b701f0a-prebuilt.qemu.org 04/01/2014 RIP: 0010:dql_completed+0x26b/0x290 Code: b7 c2 49 89 e9 44 89 da 89 c6 4c 89 d7 e8 ed 17 47 00 58 65 ff 0d 4d 27 90 7e 0f 85 fd fe ff ff e8 ea 53 8d ff e9 f3 fe ff ff <0f> 0b 01 d2 44 89 d1 29 d1 ba 00 00 00 00 0f 48 ca e9 28 ff ff ff RSP: 0018:ffffc900002b0d08 EFLAGS: 00010297 RAX: 0000000000000000 RBX: ffff888102398c80 RCX: 0000000080190009 RDX: 0000000000000000 RSI: 000000000000006a RDI: 0000000000000000 RBP: ffff888102398c00 R08: 0000000000000000 R09: 0000000000000000 R10: 00000000000000ca R11: 0000000000015681 R12: 0000000000000001 R13: ffffc900002b0d68 R14: ffff88811115e000 R15: ffff8881107aca40 FS: 00007f41ded69500(0000) GS:ffff888667dc0000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000556ccc2dc1a0 CR3: 0000000104fd8003 CR4: 0000000000772ef0 PKRU: 55555554 Call Trace:
<IRQ> ? die+0x32/0x80 ? do_trap+0xd9/0x100 ? dql_completed+0x26b/0x290 ? dql_completed+0x26b/0x290 ? do_error_trap+0x6d/0xb0 ? dql_completed+0x26b/0x290 ? exc_invalid_op+0x4c/0x60 ? dql_completed+0x26b/0x290 ? asm_exc_invalid_op+0x16/0x20 ? dql_completed+0x26b/0x290 __free_old_xmit+0xff/0x170 [virtio_net] free_old_xmit+0x54/0xc0 [virtio_net] virtnet_poll+0xf4/0xe30 [virtio_net] ? __update_load_avg_cfs_rq+0x264/0x2d0 ? update_curr+0x35/0x260 ? reweight_entity+0x1be/0x260 __napi_poll.constprop.0+0x28/0x1c0 net_rx_action+0x329/0x420 ? enqueue_hrtimer+0x35/0x90 ? trace_hardirqs_on+0x1d/0x80 ? kvm_sched_clock_read+0xd/0x20 ? sched_clock+0xc/0x30 ? kvm_sched_clock_read+0xd/0x20 ? sched_clock+0xc/0x30 ? sched_clock_cpu+0xd/0x1a0 handle_softirqs+0x138/0x3e0 do_softirq.part.0+0x89/0xc0 </IRQ> <TASK> __local_bh_enable_ip+0xa7/0xb0 virtnet_open+0xc8/0x310 [virtio_net] __dev_open+0xfa/0x1b0 __dev_change_flags+0x1de/0x250 dev_change_flags+0x22/0x60 do_setlink.isra.0+0x2df/0x10b0 ? rtnetlink_rcv_msg+0x34f/0x3f0 ? netlink_rcv_skb+0x54/0x100 ? netlink_unicast+0x23e/0x390 ? netlink_sendmsg+0x21e/0x490 ? ____sys_sendmsg+0x31b/0x350 ? avc_has_perm_noaudit+0x67/0xf0 ? cred_has_capability.isra.0+0x75/0x110 ? __nla_validate_parse+0x5f/0xee0 ? __pfx___probestub_irq_enable+0x3/0x10 ? __create_object+0x5e/0x90 ? security_capable+0x3b/0x7[I0 rtnl_newlink+0x784/0xaf0 ? avc_has_perm_noaudit+0x67/0xf0 ? cred_has_capability.isra.0+0x75/0x110 ? stack_depot_save_flags+0x24/0x6d0 ? __pfx_rtnl_newlink+0x10/0x10 rtnetlink_rcv_msg+0x34f/0x3f0 ? do_syscall_64+0x6c/0x180 ? entry_SYSCALL_64_after_hwframe+0x76/0x7e ? __pfx_rtnetlink_rcv_msg+0x10/0x10 netlink_rcv_skb+0x54/0x100 netlink_unicast+0x23e/0x390 netlink_sendmsg+0x21e/0x490 ____sys_sendmsg+0x31b/0x350 ? copy_msghdr_from_user+0x6d/0xa0 ___sys_sendmsg+0x86/0xd0 ? __pte_offset_map+0x17/0x160 ? preempt_count_add+0x69/0xa0 ? __call_rcu_common.constprop.0+0x147/0x610 ? preempt_count_add+0x69/0xa0 ? preempt_count_add+0x69/0xa0 ? _raw_spin_trylock+0x13/0x60 ? trace_hardirqs_on+0x1d/0x80 __sys_sendmsg+0x66/0xc0 do_syscall_64+0x6c/0x180 entry_SYSCALL_64_after_hwframe+0x76/0x7e RIP: 0033:0x7f41defe5b34 Code: 15 e1 12 0f 00 f7 d8 64 89 02 b8 ff ff ff ff eb bf 0f 1f 44 00 00 f3 0f 1e fa 80 3d 35 95 0f 00 00 74 13 b8 2e 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 4c c3 0f 1f 00 55 48 89 e5 48 83 ec 20 89 55 RSP: 002b:00007ffe5336ecc8 EFLAGS: 00000202 ORIG_RAX: 000000000000002e RAX: ffffffffffffffda RBX: 0000000000000003 RCX: 00007f41defe5b34 RDX: 0000000000000000 RSI: 00007ffe5336ed30 RDI: 0000000000000003 RBP: 00007ffe5336eda0 R08: 0000000000000010 R09: 0000000000000001 R10: 00007ffe5336f6f9 R11: 0000000000000202 R12: 0000000000000003 R13: 0000000067452259 R14: 0000556ccc28b040 R15: 0000000000000000 </TASK> [...] ---[ end Kernel panic - not syncing: Fatal exception in interrupt ]---
Koichiro Den (6): virtio_net: correct netdev_tx_reset_queue() invocation point virtio_net: replace vq2rxq with vq2txq where appropriate virtio_ring: add a func argument 'recycle_done' to virtqueue_resize() virtio_net: ensure netdev_tx_reset_queue is called on tx ring resize virtio_ring: add a func argument 'recycle_done' to virtqueue_reset() virtio_net: ensure netdev_tx_reset_queue is called on bind xsk for tx
drivers/net/virtio_net.c | 31 +++++++++++++++++++++++++------ drivers/virtio/virtio_ring.c | 12 ++++++++++-- include/linux/virtio.h | 6 ++++-- 3 files changed, 39 insertions(+), 10 deletions(-)
-- 2.43.0
Hello:
This series was applied to netdev/net.git (main) by Paolo Abeni pabeni@redhat.com:
On Fri, 6 Dec 2024 10:10:41 +0900 you wrote:
When virtnet_close is followed by virtnet_open, some TX completions can possibly remain unconsumed, until they are finally processed during the first NAPI poll after the netdev_tx_reset_queue(), resulting in a crash [1]. Commit b96ed2c97c79 ("virtio_net: move netdev_tx_reset_queue() call before RX napi enable") was not sufficient to eliminate all BQL crash scenarios for virtio-net.
[...]
Here is the summary with links: - [net,v4,1/6] virtio_net: correct netdev_tx_reset_queue() invocation point https://git.kernel.org/netdev/net/c/3ddccbefebdb - [net,v4,2/6] virtio_net: replace vq2rxq with vq2txq where appropriate https://git.kernel.org/netdev/net/c/4571dc7272b2 - [net,v4,3/6] virtio_ring: add a func argument 'recycle_done' to virtqueue_resize() https://git.kernel.org/netdev/net/c/8d6712c89201 - [net,v4,4/6] virtio_net: ensure netdev_tx_reset_queue is called on tx ring resize https://git.kernel.org/netdev/net/c/1480f0f61b67 - [net,v4,5/6] virtio_ring: add a func argument 'recycle_done' to virtqueue_reset() https://git.kernel.org/netdev/net/c/8d2da07c813a - [net,v4,6/6] virtio_net: ensure netdev_tx_reset_queue is called on bind xsk for tx https://git.kernel.org/netdev/net/c/76a771ec4c9a
You are awesome, thank you!
I tested this series v4 patches with virtio-net regression tests, all cases test pass. But I hit a bug https://bugzilla.kernel.org/show_bug.cgi?id=219588, only judging from call trace info it should be a CPU issue, and not related to the current patch.
Tested-by: Lei Yang leiyang@redhat.com
On Tue, Dec 10, 2024 at 6:40 PM patchwork-bot+netdevbpf@kernel.org wrote:
Hello:
This series was applied to netdev/net.git (main) by Paolo Abeni pabeni@redhat.com:
On Fri, 6 Dec 2024 10:10:41 +0900 you wrote:
When virtnet_close is followed by virtnet_open, some TX completions can possibly remain unconsumed, until they are finally processed during the first NAPI poll after the netdev_tx_reset_queue(), resulting in a crash [1]. Commit b96ed2c97c79 ("virtio_net: move netdev_tx_reset_queue() call before RX napi enable") was not sufficient to eliminate all BQL crash scenarios for virtio-net.
[...]
Here is the summary with links:
- [net,v4,1/6] virtio_net: correct netdev_tx_reset_queue() invocation point https://git.kernel.org/netdev/net/c/3ddccbefebdb
- [net,v4,2/6] virtio_net: replace vq2rxq with vq2txq where appropriate https://git.kernel.org/netdev/net/c/4571dc7272b2
- [net,v4,3/6] virtio_ring: add a func argument 'recycle_done' to virtqueue_resize() https://git.kernel.org/netdev/net/c/8d6712c89201
- [net,v4,4/6] virtio_net: ensure netdev_tx_reset_queue is called on tx ring resize https://git.kernel.org/netdev/net/c/1480f0f61b67
- [net,v4,5/6] virtio_ring: add a func argument 'recycle_done' to virtqueue_reset() https://git.kernel.org/netdev/net/c/8d2da07c813a
- [net,v4,6/6] virtio_net: ensure netdev_tx_reset_queue is called on bind xsk for tx https://git.kernel.org/netdev/net/c/76a771ec4c9a
You are awesome, thank you!
Deet-doot-dot, I am a bot. https://korg.docs.kernel.org/patchwork/pwbot.html
On Fri, Dec 06, 2024 at 10:10:41AM +0900, Koichiro Den wrote:
When virtnet_close is followed by virtnet_open, some TX completions can possibly remain unconsumed, until they are finally processed during the first NAPI poll after the netdev_tx_reset_queue(), resulting in a crash [1]. Commit b96ed2c97c79 ("virtio_net: move netdev_tx_reset_queue() call before RX napi enable") was not sufficient to eliminate all BQL crash scenarios for virtio-net.
This issue can be reproduced with the latest net-next master by running: `while :; do ip l set DEV down; ip l set DEV up; done` under heavy network TX load from inside the machine.
This patch series resolves the issue and also addresses similar existing problems:
(a). Drop netdev_tx_reset_queue() from open/close path. This eliminates the BQL crashes due to the problematic open/close path.
(b). As a result of (a), netdev_tx_reset_queue() is now explicitly required in freeze/restore path. Add netdev_tx_reset_queue() immediately after free_unused_bufs() invocation.
(c). Fix missing resetting in virtnet_tx_resize(). virtnet_tx_resize() has lacked proper resetting since commit c8bd1f7f3e61 ("virtio_net: add support for Byte Queue Limits").
(d). Fix missing resetting in the XDP_SETUP_XSK_POOL path. Similar to (c), this path lacked proper resetting. Call netdev_tx_reset_queue() when virtqueue_reset() has actually recycled unused buffers.
thanks
Acked-by: Michael S. Tsirkin mst@redhat.com
This patch series consists of six commits: [1/6]: Resolves (a) and (b). # also -stable 6.11.y [2/6]: Minor fix to make [4/6] streamlined. [3/6]: Prerequisite for (c). # also -stable 6.11.y [4/6]: Resolves (c) (incl. Prerequisite for (d)) # also -stable 6.11.y [5/6]: Preresuisite for (d). [6/6]: Resolves (d).
Changes for v4:
- move netdev_tx_reset_queue() out of free_unused_bufs()
- submit to net, not net-next
Changes for v3:
- replace 'flushed' argument with 'recycle_done'
Changes for v2:
- add tx queue resetting for (b) to (d) above
v3: https://lore.kernel.org/all/20241204050724.307544-1-koichiro.den@canonical.c... v2: https://lore.kernel.org/all/20241203073025.67065-1-koichiro.den@canonical.co... v1: https://lore.kernel.org/all/20241130181744.3772632-1-koichiro.den@canonical....
[1]: ------------[ cut here ]------------ kernel BUG at lib/dynamic_queue_limits.c:99! Oops: invalid opcode: 0000 [#1] PREEMPT SMP NOPTI CPU: 7 UID: 0 PID: 1598 Comm: ip Tainted: G N 6.12.0net-next_main+ #2 Tainted: [N]=TEST Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), \ BIOS rel-1.16.3-0-ga6ed6b701f0a-prebuilt.qemu.org 04/01/2014 RIP: 0010:dql_completed+0x26b/0x290 Code: b7 c2 49 89 e9 44 89 da 89 c6 4c 89 d7 e8 ed 17 47 00 58 65 ff 0d 4d 27 90 7e 0f 85 fd fe ff ff e8 ea 53 8d ff e9 f3 fe ff ff <0f> 0b 01 d2 44 89 d1 29 d1 ba 00 00 00 00 0f 48 ca e9 28 ff ff ff RSP: 0018:ffffc900002b0d08 EFLAGS: 00010297 RAX: 0000000000000000 RBX: ffff888102398c80 RCX: 0000000080190009 RDX: 0000000000000000 RSI: 000000000000006a RDI: 0000000000000000 RBP: ffff888102398c00 R08: 0000000000000000 R09: 0000000000000000 R10: 00000000000000ca R11: 0000000000015681 R12: 0000000000000001 R13: ffffc900002b0d68 R14: ffff88811115e000 R15: ffff8881107aca40 FS: 00007f41ded69500(0000) GS:ffff888667dc0000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000556ccc2dc1a0 CR3: 0000000104fd8003 CR4: 0000000000772ef0 PKRU: 55555554 Call Trace:
<IRQ> ? die+0x32/0x80 ? do_trap+0xd9/0x100 ? dql_completed+0x26b/0x290 ? dql_completed+0x26b/0x290 ? do_error_trap+0x6d/0xb0 ? dql_completed+0x26b/0x290 ? exc_invalid_op+0x4c/0x60 ? dql_completed+0x26b/0x290 ? asm_exc_invalid_op+0x16/0x20 ? dql_completed+0x26b/0x290 __free_old_xmit+0xff/0x170 [virtio_net] free_old_xmit+0x54/0xc0 [virtio_net] virtnet_poll+0xf4/0xe30 [virtio_net] ? __update_load_avg_cfs_rq+0x264/0x2d0 ? update_curr+0x35/0x260 ? reweight_entity+0x1be/0x260 __napi_poll.constprop.0+0x28/0x1c0 net_rx_action+0x329/0x420 ? enqueue_hrtimer+0x35/0x90 ? trace_hardirqs_on+0x1d/0x80 ? kvm_sched_clock_read+0xd/0x20 ? sched_clock+0xc/0x30 ? kvm_sched_clock_read+0xd/0x20 ? sched_clock+0xc/0x30 ? sched_clock_cpu+0xd/0x1a0 handle_softirqs+0x138/0x3e0 do_softirq.part.0+0x89/0xc0 </IRQ> <TASK> __local_bh_enable_ip+0xa7/0xb0 virtnet_open+0xc8/0x310 [virtio_net] __dev_open+0xfa/0x1b0 __dev_change_flags+0x1de/0x250 dev_change_flags+0x22/0x60 do_setlink.isra.0+0x2df/0x10b0 ? rtnetlink_rcv_msg+0x34f/0x3f0 ? netlink_rcv_skb+0x54/0x100 ? netlink_unicast+0x23e/0x390 ? netlink_sendmsg+0x21e/0x490 ? ____sys_sendmsg+0x31b/0x350 ? avc_has_perm_noaudit+0x67/0xf0 ? cred_has_capability.isra.0+0x75/0x110 ? __nla_validate_parse+0x5f/0xee0 ? __pfx___probestub_irq_enable+0x3/0x10 ? __create_object+0x5e/0x90 ? security_capable+0x3b/0x7[I0 rtnl_newlink+0x784/0xaf0 ? avc_has_perm_noaudit+0x67/0xf0 ? cred_has_capability.isra.0+0x75/0x110 ? stack_depot_save_flags+0x24/0x6d0 ? __pfx_rtnl_newlink+0x10/0x10 rtnetlink_rcv_msg+0x34f/0x3f0 ? do_syscall_64+0x6c/0x180 ? entry_SYSCALL_64_after_hwframe+0x76/0x7e ? __pfx_rtnetlink_rcv_msg+0x10/0x10 netlink_rcv_skb+0x54/0x100 netlink_unicast+0x23e/0x390 netlink_sendmsg+0x21e/0x490 ____sys_sendmsg+0x31b/0x350 ? copy_msghdr_from_user+0x6d/0xa0 ___sys_sendmsg+0x86/0xd0 ? __pte_offset_map+0x17/0x160 ? preempt_count_add+0x69/0xa0 ? __call_rcu_common.constprop.0+0x147/0x610 ? preempt_count_add+0x69/0xa0 ? preempt_count_add+0x69/0xa0 ? _raw_spin_trylock+0x13/0x60 ? trace_hardirqs_on+0x1d/0x80 __sys_sendmsg+0x66/0xc0 do_syscall_64+0x6c/0x180 entry_SYSCALL_64_after_hwframe+0x76/0x7e RIP: 0033:0x7f41defe5b34 Code: 15 e1 12 0f 00 f7 d8 64 89 02 b8 ff ff ff ff eb bf 0f 1f 44 00 00 f3 0f 1e fa 80 3d 35 95 0f 00 00 74 13 b8 2e 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 4c c3 0f 1f 00 55 48 89 e5 48 83 ec 20 89 55 RSP: 002b:00007ffe5336ecc8 EFLAGS: 00000202 ORIG_RAX: 000000000000002e RAX: ffffffffffffffda RBX: 0000000000000003 RCX: 00007f41defe5b34 RDX: 0000000000000000 RSI: 00007ffe5336ed30 RDI: 0000000000000003 RBP: 00007ffe5336eda0 R08: 0000000000000010 R09: 0000000000000001 R10: 00007ffe5336f6f9 R11: 0000000000000202 R12: 0000000000000003 R13: 0000000067452259 R14: 0000556ccc28b040 R15: 0000000000000000 </TASK> [...] ---[ end Kernel panic - not syncing: Fatal exception in interrupt ]---
Koichiro Den (6): virtio_net: correct netdev_tx_reset_queue() invocation point virtio_net: replace vq2rxq with vq2txq where appropriate virtio_ring: add a func argument 'recycle_done' to virtqueue_resize() virtio_net: ensure netdev_tx_reset_queue is called on tx ring resize virtio_ring: add a func argument 'recycle_done' to virtqueue_reset() virtio_net: ensure netdev_tx_reset_queue is called on bind xsk for tx
drivers/net/virtio_net.c | 31 +++++++++++++++++++++++++------ drivers/virtio/virtio_ring.c | 12 ++++++++++-- include/linux/virtio.h | 6 ++++-- 3 files changed, 39 insertions(+), 10 deletions(-)
-- 2.43.0
linux-stable-mirror@lists.linaro.org
-
Jason Wang -
kernel test robot -
Koichiro Den -
Lei Yang -
Michael S. Tsirkin -
Paolo Abeni -
patchwork-bot+netdevbpf＠kernel.org -
Xuan Zhuo