Message ID | 20200518045452.2390-1-dirk.behme@de.bosch.com (mailing list archive) |
---|---|
State | Under Review |
Delegated to: | Geert Uytterhoeven |
Headers | show |
Series | [v2] ravb: On timeout disable IRQs to stop processing | expand |
+ Sasikumar Best regards, Suguni Shashikant RBEI/ECF1 Tel. +91 80 6136-4404 -----Original Message----- From: Behme Dirk (CM/ESO2) <Dirk.Behme@de.bosch.com> Sent: Monday, May 18, 2020 10:25 AM To: sergei.shtylyov@cogentembedded.com; linux-renesas-soc@vger.kernel.org Cc: Behme Dirk (CM/ESO2) <Dirk.Behme@de.bosch.com>; Shashikant Suguni (RBEI/ECF1) <Shashikant.Suguni@in.bosch.com> Subject: [PATCH v2] ravb: On timeout disable IRQs to stop processing Analyzing [1] it seems there is a race condition where ravb_start_xmit() can be called from interrupt while tx skbuffs are being torn down in the scheduled timeout handling. The actual timeout work is done in ravb_tx_timeout_work() during which the tx skbuffs are torn down via invocations of ravb_ring_free(). But there seems to be no flag to tell the driver it is shutting down so it continues to use the ring buffer when it should not. Fix this by disabling the interrupts in the timeout handler. [1] -- cut -- ravb e6800000.ethernet ethernet: transmit timed out, status 00000000, resetting... ravb e6800000.ethernet ethernet: failed to switch device to config mode Unable to handle kernel NULL pointer dereference at virtual address 00000018 Mem abort info: Exception class = DABT (current EL), IL = 32 bits SET = 0, FnV = 0 EA = 0, S1PTW = 0 Data abort info: ISV = 0, ISS = 0x00000046 CM = 0, WnR = 1 user pgtable: 4k pages, 48-bit VAs, pgd = ffff80065622f000 [0000000000000018] *pgd=00000006962a7003 , *pud=00000006962a8003 , *pmd=0000000000000000 Internal error: Oops: 96000046 [#1] PREEMPT SMP Modules linked in: ... Process Thread1 (pid: 3132, stack limit = 0xffff000027dd0000) CPU: 2 PID: 3132 Comm: Thread1 Tainted: G WC 4.14.130-ltsi-g28acae87 #1 Hardware name: Board based on r8a7796 (DT) task: ffff80064f2aaa00 task.stack: ffff000027dd0000 PC is at ravb_start_xmit+0x138/0x5a0 LR is at ravb_start_xmit+0x40/0x5a0 pc : [<ffff0000084d6324>] lr : [<ffff0000084d622c>] pstate: 600001c5 sp : ffff000027dd3550 x29: ffff000027dd3550 x28: 0000000000000076 x27: ffff80061035ff00 x26: ffff000027dd3694 x25: ffff80069624f800 x24: ffff80069624f000 x23: 0000000000000003 x22: 0000000000000001 x21: ffff80069624f000 x20: 0000000000000000 x19: ffff80069624f000 x18: 0000000000000014 x17: 0000ffff9b90ddb0 x16: ffff00000867d07c x15: 0000155107b31031 x14: 000409000c000000 x13: 0000000003000001 x12: 0100050010000001 x11: 0000000003000001 x10: 0100010010000001 x9 : 20000000000000c0 x8 : 0000000000000000 x7 : ffff8006656f9388 x6 : 0000000000000002 x5 : 0000000000000000 x4 : ffff8006656f929c x3 : ffff000027dd3694 x2 : 0000000000000018 x1 : 0000000000000000 x0 : 0000000000000003 Call trace: Exception stack(0xffff000027dd3410 to 0xffff000027dd3550) 3400: 0000000000000003 0000000000000000 3420: 0000000000000018 ffff000027dd3694 ffff8006656f929c 0000000000000000 3440: 0000000000000002 ffff8006656f9388 0000000000000000 20000000000000c0 3460: 0100010010000001 0000000003000001 0100050010000001 0000000003000001 3480: 000409000c000000 0000155107b31031 ffff00000867d07c 0000ffff9b90ddb0 34a0: 0000000000000014 ffff80069624f000 0000000000000000 ffff80069624f000 34c0: 0000000000000001 0000000000000003 ffff80069624f000 ffff80069624f800 34e0: ffff000027dd3694 ffff80061035ff00 0000000000000076 ffff000027dd3550 3500: ffff0000084d622c ffff000027dd3550 ffff0000084d6324 00000000600001c5 3520: ffff00000921d008 ffff8006159b0d00 0000ffffffffffff ffff0000084d622c 3540: ffff000027dd3550 ffff0000084d6324 [<ffff0000084d6324>] ravb_start_xmit+0x138/0x5a0 [<ffff000008699d60>] dev_hard_start_xmit+0xa8/0x24c [<ffff0000086c51c0>] sch_direct_xmit+0xb0/0x1a8 [<ffff0000086c54cc>] __qdisc_run+0x214/0x2ec [<ffff00000869a348>] __dev_queue_xmit+0x35c/0x5b4 [<ffff00000869a5b0>] dev_queue_xmit+0x10/0x18 [<ffff000000b9765c>] register_vlan_dev+0xc74/0x10f8 [8021q] [<ffff000008699d60>] dev_hard_start_xmit+0xa8/0x24c [<ffff00000869a438>] __dev_queue_xmit+0x44c/0x5b4 [<ffff00000869a5b0>] dev_queue_xmit+0x10/0x18 [<ffff0000086a5b9c>] neigh_connected_output+0xc0/0xe4 [<ffff0000086d6bd8>] ip_finish_output2+0x3c0/0x3fc [<ffff0000086d8178>] ip_finish_output+0xf8/0x1c4 [<ffff0000086d8b74>] ip_mc_output+0x258/0x308 [<ffff0000086d8400>] ip_local_out+0x44/0x54 [<ffff0000086d9594>] ip_send_skb+0x1c/0xa8 [<ffff0000086fea4c>] udp_send_skb+0x11c/0x244 [<ffff000008701084>] udp_sendmsg+0x534/0x6bc [<ffff00000870b390>] inet_sendmsg+0x40/0xe0 [<ffff00000867b588>] sock_sendmsg+0x3c/0x58 [<ffff00000867bf18>] ___sys_sendmsg+0x228/0x278 [<ffff00000867d03c>] __sys_sendmsg+0x58/0x98 [<ffff00000867d08c>] SyS_sendmsg+0x10/0x20 Exception stack(0xffff000027dd3ec0 to 0xffff000027dd4000) 3ec0: 0000000000000012 0000ffff8a7fdf18 0000000000004000 0000000000000000 3ee0: 0000ffff8a7ff258 0000ffff8a7ff150 0000ffff8a7ff840 0000000000000000 3f00: 00000000000000d3 0100010010000001 0000000003000001 0100050010000001 3f20: 0000000003000001 000409000c000000 0000000000000047 0000155107b31031 3f40: 0000ffff9b67dfb8 0000ffff9b90ddb0 0000000000000014 0000000000004000 3f60: 0000000000000012 0000ffff8a7fdf18 0000ffff780017b0 0000000000004000 3f80: 0000000000000010 0000000000000001 0000ffff8a7fdf00 0000ffff9000b770 3fa0: 0000000000000012 0000ffff8a7fde60 0000ffff9b90de10 0000ffff8a7fde60 3fc0: 0000ffff9b90de28 0000000080000000 0000000000000012 00000000000000d3 3fe0: 0000000000000000 0000000000000000 0000000000000000 0000000000000000 [<ffff0000080832c0>] el0_svc_naked+0x34/0x38 Code: d37d7c01 d37d7c02 f90037a1 f9445f01 (f822683b) ---[ end trace eabda93d178d5bcb ]--- Kernel panic - not syncing: Fatal exception in interrupt SMP: stopping secondary CPUs Kernel Offset: disabled CPU features: 0x1802004 Memory Limit: 6144 MB -- cut -- Fixes: c156633f1353 ("Renesas Ethernet AVB driver proper") Reviewed-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com> Signed-off-by: Dirk Behme <dirk.behme@de.bosch.com> --- Changes in v2: Dropped the RFC and added Sergei's Reviewed-by. No functional change. drivers/net/ethernet/renesas/ravb_main.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/net/ethernet/renesas/ravb_main.c b/drivers/net/ethernet/renesas/ravb_main.c index 067ad25553b9..0f91ab41b22b 100644 --- a/drivers/net/ethernet/renesas/ravb_main.c +++ b/drivers/net/ethernet/renesas/ravb_main.c @@ -1447,6 +1447,11 @@ static void ravb_tx_timeout_work(struct work_struct *work) netif_tx_stop_all_queues(ndev); + /* Disable interrupts by clearing the interrupt masks. */ + ravb_write(ndev, 0, RIC0); + ravb_write(ndev, 0, RIC2); + ravb_write(ndev, 0, TIC); + /* Stop PTP Clock driver */ if (priv->chip_id == RCAR_GEN2) ravb_ptp_stop(ndev); -- 2.20.0
Hello! On 18.05.2020 7:54, Dirk Behme wrote: > Analyzing [1] it seems there is a race condition where ravb_start_xmit() > can be called from interrupt while tx skbuffs are being torn down in > the scheduled timeout handling. The actual timeout work is done in > ravb_tx_timeout_work() during which the tx skbuffs are torn down via > invocations of ravb_ring_free(). But there seems to be no flag to tell > the driver it is shutting down so it continues to use the ring buffer > when it should not. > > Fix this by disabling the interrupts in the timeout handler. > > [1] > > -- cut -- > ravb e6800000.ethernet ethernet: transmit timed out, status 00000000, resetting... > ravb e6800000.ethernet ethernet: failed to switch device to config mode > Unable to handle kernel NULL pointer dereference at virtual address 00000018 > Mem abort info: > Exception class = DABT (current EL), IL = 32 bits > SET = 0, FnV = 0 > EA = 0, S1PTW = 0 > Data abort info: > ISV = 0, ISS = 0x00000046 > CM = 0, WnR = 1 > user pgtable: 4k pages, 48-bit VAs, pgd = ffff80065622f000 > [0000000000000018] *pgd=00000006962a7003 > , *pud=00000006962a8003 > , *pmd=0000000000000000 > Internal error: Oops: 96000046 [#1] PREEMPT SMP > Modules linked in: > ... > Process Thread1 (pid: 3132, stack limit = 0xffff000027dd0000) > CPU: 2 PID: 3132 Comm: Thread1 Tainted: G WC 4.14.130-ltsi-g28acae87 #1 > Hardware name: Board based on r8a7796 (DT) > task: ffff80064f2aaa00 task.stack: ffff000027dd0000 > PC is at ravb_start_xmit+0x138/0x5a0 > LR is at ravb_start_xmit+0x40/0x5a0 > pc : [<ffff0000084d6324>] lr : [<ffff0000084d622c>] pstate: 600001c5 > sp : ffff000027dd3550 > x29: ffff000027dd3550 > x28: 0000000000000076 > x27: ffff80061035ff00 > x26: ffff000027dd3694 > x25: ffff80069624f800 > x24: ffff80069624f000 > x23: 0000000000000003 > x22: 0000000000000001 > x21: ffff80069624f000 > x20: 0000000000000000 > x19: ffff80069624f000 > x18: 0000000000000014 > x17: 0000ffff9b90ddb0 > x16: ffff00000867d07c > x15: 0000155107b31031 > x14: 000409000c000000 > x13: 0000000003000001 > x12: 0100050010000001 > x11: 0000000003000001 > x10: 0100010010000001 > x9 : 20000000000000c0 > x8 : 0000000000000000 > x7 : ffff8006656f9388 > x6 : 0000000000000002 > x5 : 0000000000000000 > x4 : ffff8006656f929c > x3 : ffff000027dd3694 > x2 : 0000000000000018 > x1 : 0000000000000000 > x0 : 0000000000000003 > Call trace: > Exception stack(0xffff000027dd3410 to 0xffff000027dd3550) > 3400: 0000000000000003 0000000000000000 > 3420: 0000000000000018 ffff000027dd3694 ffff8006656f929c 0000000000000000 > 3440: 0000000000000002 ffff8006656f9388 0000000000000000 20000000000000c0 > 3460: 0100010010000001 0000000003000001 0100050010000001 0000000003000001 > 3480: 000409000c000000 0000155107b31031 ffff00000867d07c 0000ffff9b90ddb0 > 34a0: 0000000000000014 ffff80069624f000 0000000000000000 ffff80069624f000 > 34c0: 0000000000000001 0000000000000003 ffff80069624f000 ffff80069624f800 > 34e0: ffff000027dd3694 ffff80061035ff00 0000000000000076 ffff000027dd3550 > 3500: ffff0000084d622c ffff000027dd3550 ffff0000084d6324 00000000600001c5 > 3520: ffff00000921d008 ffff8006159b0d00 0000ffffffffffff ffff0000084d622c > 3540: ffff000027dd3550 ffff0000084d6324 > [<ffff0000084d6324>] ravb_start_xmit+0x138/0x5a0 > [<ffff000008699d60>] dev_hard_start_xmit+0xa8/0x24c > [<ffff0000086c51c0>] sch_direct_xmit+0xb0/0x1a8 > [<ffff0000086c54cc>] __qdisc_run+0x214/0x2ec > [<ffff00000869a348>] __dev_queue_xmit+0x35c/0x5b4 > [<ffff00000869a5b0>] dev_queue_xmit+0x10/0x18 > [<ffff000000b9765c>] register_vlan_dev+0xc74/0x10f8 [8021q] > [<ffff000008699d60>] dev_hard_start_xmit+0xa8/0x24c > [<ffff00000869a438>] __dev_queue_xmit+0x44c/0x5b4 > [<ffff00000869a5b0>] dev_queue_xmit+0x10/0x18 > [<ffff0000086a5b9c>] neigh_connected_output+0xc0/0xe4 > [<ffff0000086d6bd8>] ip_finish_output2+0x3c0/0x3fc > [<ffff0000086d8178>] ip_finish_output+0xf8/0x1c4 > [<ffff0000086d8b74>] ip_mc_output+0x258/0x308 > [<ffff0000086d8400>] ip_local_out+0x44/0x54 > [<ffff0000086d9594>] ip_send_skb+0x1c/0xa8 > [<ffff0000086fea4c>] udp_send_skb+0x11c/0x244 > [<ffff000008701084>] udp_sendmsg+0x534/0x6bc > [<ffff00000870b390>] inet_sendmsg+0x40/0xe0 > [<ffff00000867b588>] sock_sendmsg+0x3c/0x58 > [<ffff00000867bf18>] ___sys_sendmsg+0x228/0x278 > [<ffff00000867d03c>] __sys_sendmsg+0x58/0x98 > [<ffff00000867d08c>] SyS_sendmsg+0x10/0x20 > Exception stack(0xffff000027dd3ec0 to 0xffff000027dd4000) > 3ec0: 0000000000000012 0000ffff8a7fdf18 0000000000004000 0000000000000000 > 3ee0: 0000ffff8a7ff258 0000ffff8a7ff150 0000ffff8a7ff840 0000000000000000 > 3f00: 00000000000000d3 0100010010000001 0000000003000001 0100050010000001 > 3f20: 0000000003000001 000409000c000000 0000000000000047 0000155107b31031 > 3f40: 0000ffff9b67dfb8 0000ffff9b90ddb0 0000000000000014 0000000000004000 > 3f60: 0000000000000012 0000ffff8a7fdf18 0000ffff780017b0 0000000000004000 > 3f80: 0000000000000010 0000000000000001 0000ffff8a7fdf00 0000ffff9000b770 > 3fa0: 0000000000000012 0000ffff8a7fde60 0000ffff9b90de10 0000ffff8a7fde60 > 3fc0: 0000ffff9b90de28 0000000080000000 0000000000000012 00000000000000d3 > 3fe0: 0000000000000000 0000000000000000 0000000000000000 0000000000000000 > [<ffff0000080832c0>] el0_svc_naked+0x34/0x38 > Code: d37d7c01 d37d7c02 f90037a1 f9445f01 (f822683b) > ---[ end trace eabda93d178d5bcb ]--- > Kernel panic - not syncing: Fatal exception in interrupt > SMP: stopping secondary CPUs > Kernel Offset: disabled > CPU features: 0x1802004 > Memory Limit: 6144 MB > -- cut -- > > Fixes: c156633f1353 ("Renesas Ethernet AVB driver proper") > Reviewed-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com> > Signed-off-by: Dirk Behme <dirk.behme@de.bosch.com> [...] You must post this to netdev@vget.kernel.org, else the patch won't get applied, ever... MBR, Sergei
On 18.05.2020 7:54, Dirk Behme wrote: > Analyzing [1] it seems there is a race condition where ravb_start_xmit() > can be called from interrupt while tx skbuffs are being torn down in > the scheduled timeout handling. The actual timeout work is done in > ravb_tx_timeout_work() during which the tx skbuffs are torn down via > invocations of ravb_ring_free(). But there seems to be no flag to tell > the driver it is shutting down so it continues to use the ring buffer > when it should not. > > Fix this by disabling the interrupts in the timeout handler. Hm, given that we stop all TX queues prior to tearing down the buffers, it is somewhat strange that you see the driver's send path called... But disabling the interrupts seems the Right Thing anyways... > [1] > > -- cut -- > ravb e6800000.ethernet ethernet: transmit timed out, status 00000000, resetting... > ravb e6800000.ethernet ethernet: failed to switch device to config mode > Unable to handle kernel NULL pointer dereference at virtual address 00000018 > Mem abort info: > Exception class = DABT (current EL), IL = 32 bits > SET = 0, FnV = 0 > EA = 0, S1PTW = 0 > Data abort info: > ISV = 0, ISS = 0x00000046 > CM = 0, WnR = 1 > user pgtable: 4k pages, 48-bit VAs, pgd = ffff80065622f000 > [0000000000000018] *pgd=00000006962a7003 > , *pud=00000006962a8003 > , *pmd=0000000000000000 > Internal error: Oops: 96000046 [#1] PREEMPT SMP > Modules linked in: > ... > Process Thread1 (pid: 3132, stack limit = 0xffff000027dd0000) > CPU: 2 PID: 3132 Comm: Thread1 Tainted: G WC 4.14.130-ltsi-g28acae87 #1 > Hardware name: Board based on r8a7796 (DT) > task: ffff80064f2aaa00 task.stack: ffff000027dd0000 > PC is at ravb_start_xmit+0x138/0x5a0 > LR is at ravb_start_xmit+0x40/0x5a0 > pc : [<ffff0000084d6324>] lr : [<ffff0000084d622c>] pstate: 600001c5 > sp : ffff000027dd3550 > x29: ffff000027dd3550 > x28: 0000000000000076 > x27: ffff80061035ff00 > x26: ffff000027dd3694 > x25: ffff80069624f800 > x24: ffff80069624f000 > x23: 0000000000000003 > x22: 0000000000000001 > x21: ffff80069624f000 > x20: 0000000000000000 > x19: ffff80069624f000 > x18: 0000000000000014 > x17: 0000ffff9b90ddb0 > x16: ffff00000867d07c > x15: 0000155107b31031 > x14: 000409000c000000 > x13: 0000000003000001 > x12: 0100050010000001 > x11: 0000000003000001 > x10: 0100010010000001 > x9 : 20000000000000c0 > x8 : 0000000000000000 > x7 : ffff8006656f9388 > x6 : 0000000000000002 > x5 : 0000000000000000 > x4 : ffff8006656f929c > x3 : ffff000027dd3694 > x2 : 0000000000000018 > x1 : 0000000000000000 > x0 : 0000000000000003 > Call trace: > Exception stack(0xffff000027dd3410 to 0xffff000027dd3550) > 3400: 0000000000000003 0000000000000000 > 3420: 0000000000000018 ffff000027dd3694 ffff8006656f929c 0000000000000000 > 3440: 0000000000000002 ffff8006656f9388 0000000000000000 20000000000000c0 > 3460: 0100010010000001 0000000003000001 0100050010000001 0000000003000001 > 3480: 000409000c000000 0000155107b31031 ffff00000867d07c 0000ffff9b90ddb0 > 34a0: 0000000000000014 ffff80069624f000 0000000000000000 ffff80069624f000 > 34c0: 0000000000000001 0000000000000003 ffff80069624f000 ffff80069624f800 > 34e0: ffff000027dd3694 ffff80061035ff00 0000000000000076 ffff000027dd3550 > 3500: ffff0000084d622c ffff000027dd3550 ffff0000084d6324 00000000600001c5 > 3520: ffff00000921d008 ffff8006159b0d00 0000ffffffffffff ffff0000084d622c > 3540: ffff000027dd3550 ffff0000084d6324 > [<ffff0000084d6324>] ravb_start_xmit+0x138/0x5a0 > [<ffff000008699d60>] dev_hard_start_xmit+0xa8/0x24c > [<ffff0000086c51c0>] sch_direct_xmit+0xb0/0x1a8 > [<ffff0000086c54cc>] __qdisc_run+0x214/0x2ec > [<ffff00000869a348>] __dev_queue_xmit+0x35c/0x5b4 > [<ffff00000869a5b0>] dev_queue_xmit+0x10/0x18 > [<ffff000000b9765c>] register_vlan_dev+0xc74/0x10f8 [8021q] > [<ffff000008699d60>] dev_hard_start_xmit+0xa8/0x24c > [<ffff00000869a438>] __dev_queue_xmit+0x44c/0x5b4 > [<ffff00000869a5b0>] dev_queue_xmit+0x10/0x18 > [<ffff0000086a5b9c>] neigh_connected_output+0xc0/0xe4 > [<ffff0000086d6bd8>] ip_finish_output2+0x3c0/0x3fc > [<ffff0000086d8178>] ip_finish_output+0xf8/0x1c4 > [<ffff0000086d8b74>] ip_mc_output+0x258/0x308 > [<ffff0000086d8400>] ip_local_out+0x44/0x54 > [<ffff0000086d9594>] ip_send_skb+0x1c/0xa8 > [<ffff0000086fea4c>] udp_send_skb+0x11c/0x244 > [<ffff000008701084>] udp_sendmsg+0x534/0x6bc > [<ffff00000870b390>] inet_sendmsg+0x40/0xe0 > [<ffff00000867b588>] sock_sendmsg+0x3c/0x58 > [<ffff00000867bf18>] ___sys_sendmsg+0x228/0x278 > [<ffff00000867d03c>] __sys_sendmsg+0x58/0x98 > [<ffff00000867d08c>] SyS_sendmsg+0x10/0x20 > Exception stack(0xffff000027dd3ec0 to 0xffff000027dd4000) > 3ec0: 0000000000000012 0000ffff8a7fdf18 0000000000004000 0000000000000000 > 3ee0: 0000ffff8a7ff258 0000ffff8a7ff150 0000ffff8a7ff840 0000000000000000 > 3f00: 00000000000000d3 0100010010000001 0000000003000001 0100050010000001 > 3f20: 0000000003000001 000409000c000000 0000000000000047 0000155107b31031 > 3f40: 0000ffff9b67dfb8 0000ffff9b90ddb0 0000000000000014 0000000000004000 > 3f60: 0000000000000012 0000ffff8a7fdf18 0000ffff780017b0 0000000000004000 > 3f80: 0000000000000010 0000000000000001 0000ffff8a7fdf00 0000ffff9000b770 > 3fa0: 0000000000000012 0000ffff8a7fde60 0000ffff9b90de10 0000ffff8a7fde60 > 3fc0: 0000ffff9b90de28 0000000080000000 0000000000000012 00000000000000d3 > 3fe0: 0000000000000000 0000000000000000 0000000000000000 0000000000000000 > [<ffff0000080832c0>] el0_svc_naked+0x34/0x38 > Code: d37d7c01 d37d7c02 f90037a1 f9445f01 (f822683b) > ---[ end trace eabda93d178d5bcb ]--- > Kernel panic - not syncing: Fatal exception in interrupt > SMP: stopping secondary CPUs > Kernel Offset: disabled > CPU features: 0x1802004 > Memory Limit: 6144 MB > -- cut -- > > Fixes: c156633f1353 ("Renesas Ethernet AVB driver proper") > Reviewed-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com> > Signed-off-by: Dirk Behme <dirk.behme@de.bosch.com> > --- > > Changes in v2: Dropped the RFC and added Sergei's Reviewed-by. No functional > change. > > drivers/net/ethernet/renesas/ravb_main.c | 5 +++++ > 1 file changed, 5 insertions(+) > > diff --git a/drivers/net/ethernet/renesas/ravb_main.c b/drivers/net/ethernet/renesas/ravb_main.c > index 067ad25553b9..0f91ab41b22b 100644 > --- a/drivers/net/ethernet/renesas/ravb_main.c > +++ b/drivers/net/ethernet/renesas/ravb_main.c > @@ -1447,6 +1447,11 @@ static void ravb_tx_timeout_work(struct work_struct *work) > > netif_tx_stop_all_queues(ndev); > > + /* Disable interrupts by clearing the interrupt masks. */ > + ravb_write(ndev, 0, RIC0); > + ravb_write(ndev, 0, RIC2); > + ravb_write(ndev, 0, TIC); > + > /* Stop PTP Clock driver */ > if (priv->chip_id == RCAR_GEN2) > ravb_ptp_stop(ndev); MBR, Sergei
Hello Sergei-san, > From: Sergei Shtylyov, Sent: Monday, May 18, 2020 5:45 PM > > On 18.05.2020 7:54, Dirk Behme wrote: > > > Analyzing [1] it seems there is a race condition where ravb_start_xmit() > > can be called from interrupt while tx skbuffs are being torn down in > > the scheduled timeout handling. The actual timeout work is done in > > ravb_tx_timeout_work() during which the tx skbuffs are torn down via > > invocations of ravb_ring_free(). But there seems to be no flag to tell > > the driver it is shutting down so it continues to use the ring buffer > > when it should not. > > > > Fix this by disabling the interrupts in the timeout handler. > > Hm, given that we stop all TX queues prior to tearing down the buffers, > it is somewhat strange that you see the driver's send path called... > But disabling the interrupts seems the Right Thing anyways... I didn't reproduce this issue and I didn't know the root cause yet. But, I'm feeling to strange. Especially, "ravb_start_xmit() can be called from interrupt". I didn't find where ravb_start_xmit() can be called from interrupt for now. By the way, I'm thinking the following message is related to the issue. > > ravb e6800000.ethernet ethernet: failed to switch device to config mode The ravb_config() output the message if failed. And, ravb_tx_timeout_work() calls ravb_config() via ravb_stop_dma() and ravb_dmac_init(). --- ravb_tx_timeout_work() { ravb_stop_dma() // call ravb_config() ravb_ring_free() // *1 <snip> ravb_dmac_init() // call ravb_config() <snip> } ravb_dmac_init() { <snip> error = ravb_config() // "failed to switch device to config modes" here and -ETIMEDOUT if (error) return error // *2 ravb_ring_init() // *3 <snip> } -- If ravb_config() failed at the *2, since ravb_ring_init() was not called, any descriptors were not allocated and then the issue should happen. Note that according to the datasheet, the controller cannot change the mode from "Operation" to Configuration" when the controller is doing TX or RX. I don't know why the first ravb_config() doesn't fail for now. But, my scenario is one of functions below enables the TX and RX by calling ravb_rcv_snd_enable(): - ravb_emac_interrupt_unlocked() ... if ESCR_LCHNG && !no_avb_link && PSR_LMON - ravb_adjust_link() ... if no_avb_link && phydev->link - ravb_set_rx_csum() ... if enabling NETIF_F_RXCSUM ravb_tx_timeout_work() calls ravb_stop_dma(). And, ravb_stop_dma() calls ravb_rcv_snd_disable(). So, we assumed ravb_tx_timeout_work() didn't transfer anymore. However, if one of the above functions was called, this is possible to enable TX and RX. If this scenario is true, I'm thinking we can fix the issue by using spin_{un}lock_irq{restore,save}() between ravb_stop_dma() to ravb_dmac_init(). # ravb_ptp_init() calls spin_lock API so that we should not spin lock ravb_ptp_{init,stop}(). What do you think? I'll try to reproduce this issue on my environment again... # I'm thinking ravb_adjust_link() is possible to cause this issue. # But, Salvator-X doesn't have renesas,no-ether-link so that I'll try to add it. Best regards, Yoshihiro Shimoda
Hi again, > From: Yoshihiro Shimoda, Sent: Friday, May 22, 2020 2:58 PM > > Hello Sergei-san, > > > From: Sergei Shtylyov, Sent: Monday, May 18, 2020 5:45 PM > > > > On 18.05.2020 7:54, Dirk Behme wrote: > > > > > Analyzing [1] it seems there is a race condition where ravb_start_xmit() > > > can be called from interrupt while tx skbuffs are being torn down in > > > the scheduled timeout handling. The actual timeout work is done in > > > ravb_tx_timeout_work() during which the tx skbuffs are torn down via > > > invocations of ravb_ring_free(). But there seems to be no flag to tell > > > the driver it is shutting down so it continues to use the ring buffer > > > when it should not. > > > > > > Fix this by disabling the interrupts in the timeout handler. > > > > Hm, given that we stop all TX queues prior to tearing down the buffers, > > it is somewhat strange that you see the driver's send path called... > > But disabling the interrupts seems the Right Thing anyways... > > I didn't reproduce this issue and I didn't know the root cause yet. > But, I'm feeling to strange. Especially, "ravb_start_xmit() can be called from interrupt". > I didn't find where ravb_start_xmit() can be called from interrupt for now. > > By the way, I'm thinking the following message is related to the issue. > > > ravb e6800000.ethernet ethernet: failed to switch device to config mode > > The ravb_config() output the message if failed. And, ravb_tx_timeout_work() > calls ravb_config() via ravb_stop_dma() and ravb_dmac_init(). > --- > ravb_tx_timeout_work() { > ravb_stop_dma() // call ravb_config() > > ravb_ring_free() // *1 > <snip> > ravb_dmac_init() // call ravb_config() > <snip> > } > > ravb_dmac_init() > { > <snip> > error = ravb_config() // "failed to switch device to config modes" here and -ETIMEDOUT > if (error) > return error // *2 > ravb_ring_init() // *3 > <snip> > } > -- > > If ravb_config() failed at the *2, since ravb_ring_init() was not called, > any descriptors were not allocated and then the issue should happen. > Note that according to the datasheet, the controller cannot change the > mode from "Operation" to Configuration" when the controller is doing > TX or RX. I'm afraid but, this scenario seems wrong because if the controller enters "Configuration" once, the driver needs to change the mode to "Operation" for starting any transfer again. > I don't know why the first ravb_config() doesn't fail for now. > But, my scenario is one of functions below enables the TX and RX > by calling ravb_rcv_snd_enable(): > - ravb_emac_interrupt_unlocked() ... if ESCR_LCHNG && !no_avb_link && PSR_LMON > - ravb_adjust_link() ... if no_avb_link && phydev->link > - ravb_set_rx_csum() ... if enabling NETIF_F_RXCSUM > > ravb_tx_timeout_work() calls ravb_stop_dma(). And, ravb_stop_dma() calls > ravb_rcv_snd_disable(). So, we assumed ravb_tx_timeout_work() didn't transfer > anymore. However, if one of the above functions was called, this is possible > to enable TX and RX. > > If this scenario is true, I'm thinking we can fix the issue by using > spin_{un}lock_irq{restore,save}() between ravb_stop_dma() to ravb_dmac_init(). This is also wrong because ravb_ring_init() call kcalloc() with GFP_KERNEL. I'll try to reproduce this issue again in next week... Best regards, Yoshihiro Shimoda > # ravb_ptp_init() calls spin_lock API so that we should not spin lock ravb_ptp_{init,stop}(). > > What do you think? > > I'll try to reproduce this issue on my environment again... > # I'm thinking ravb_adjust_link() is possible to cause this issue. > # But, Salvator-X doesn't have renesas,no-ether-link so that I'll try to add it. > > Best regards, > Yoshihiro Shimoda
Hi again and again, > From: Yoshihiro Shimoda, Sent: Friday, May 22, 2020 8:17 PM > > Hi again, > > > From: Yoshihiro Shimoda, Sent: Friday, May 22, 2020 2:58 PM > > > > Hello Sergei-san, > > > > > From: Sergei Shtylyov, Sent: Monday, May 18, 2020 5:45 PM > > > > > > On 18.05.2020 7:54, Dirk Behme wrote: > > > > > > > Analyzing [1] it seems there is a race condition where ravb_start_xmit() > > > > can be called from interrupt while tx skbuffs are being torn down in > > > > the scheduled timeout handling. The actual timeout work is done in > > > > ravb_tx_timeout_work() during which the tx skbuffs are torn down via > > > > invocations of ravb_ring_free(). But there seems to be no flag to tell > > > > the driver it is shutting down so it continues to use the ring buffer > > > > when it should not. > > > > > > > > Fix this by disabling the interrupts in the timeout handler. > > > > > > Hm, given that we stop all TX queues prior to tearing down the buffers, > > > it is somewhat strange that you see the driver's send path called... > > > But disabling the interrupts seems the Right Thing anyways... > > > > I didn't reproduce this issue and I didn't know the root cause yet. > > But, I'm feeling to strange. Especially, "ravb_start_xmit() can be called from interrupt". > > I didn't find where ravb_start_xmit() can be called from interrupt for now. > > > > By the way, I'm thinking the following message is related to the issue. > > > > ravb e6800000.ethernet ethernet: failed to switch device to config mode > > > > The ravb_config() output the message if failed. And, ravb_tx_timeout_work() > > calls ravb_config() via ravb_stop_dma() and ravb_dmac_init(). > > --- > > ravb_tx_timeout_work() { > > ravb_stop_dma() // call ravb_config() > > > > ravb_ring_free() // *1 > > <snip> > > ravb_dmac_init() // call ravb_config() > > <snip> > > } > > > > ravb_dmac_init() > > { > > <snip> > > error = ravb_config() // "failed to switch device to config modes" here and -ETIMEDOUT > > if (error) > > return error // *2 > > ravb_ring_init() // *3 > > <snip> > > } > > -- > > > > If ravb_config() failed at the *2, since ravb_ring_init() was not called, > > any descriptors were not allocated and then the issue should happen. > > Note that according to the datasheet, the controller cannot change the > > mode from "Operation" to Configuration" when the controller is doing > > TX or RX. > > I'm afraid but, this scenario seems wrong because if the controller > enters "Configuration" once, the driver needs to change the mode to > "Operation" for starting any transfer again. I realized that ravb_stop_dma() is possible to return without ravb_rcv_snd_disable() and ravb_config() calling if TCCR or CSR register indicates transmit is started. > > I don't know why the first ravb_config() doesn't fail for now. > > But, my scenario is one of functions below enables the TX and RX > > by calling ravb_rcv_snd_enable(): > > - ravb_emac_interrupt_unlocked() ... if ESCR_LCHNG && !no_avb_link && PSR_LMON > > - ravb_adjust_link() ... if no_avb_link && phydev->link > > - ravb_set_rx_csum() ... if enabling NETIF_F_RXCSUM In such the case (ravb_stop_dma() returns without any message), I think my scenario is still alive without any these functions calling because ravb_rcv_snd_disable() is not called. I'll check the TCCR and CSR registers in next week. Best regards, Yoshihiro Shimoda
Hi again, > From: Yoshihiro Shimoda, Sent: Friday, May 22, 2020 8:59 PM > In such the case (ravb_stop_dma() returns without any message), > I think my scenario is still alive without any these functions calling > because ravb_rcv_snd_disable() is not called. > > I'll check the TCCR and CSR registers in next week. I investigated these registers' feature and it is possible to return as my expectation. So, I have submitted a patch as RFC like below. https://patchwork.kernel.org/patch/11570217/ If you have any feedback about the patch, please let me know on the email. Best regards, Yoshihiro Shimoda
On 18.05.2020 10:36, Sergei Shtylyov wrote: > Hello! > > On 18.05.2020 7:54, Dirk Behme wrote: > >> Analyzing [1] it seems there is a race condition where ravb_start_xmit() >> can be called from interrupt while tx skbuffs are being torn down in >> the scheduled timeout handling. The actual timeout work is done in >> ravb_tx_timeout_work() during which the tx skbuffs are torn down via >> invocations of ravb_ring_free(). But there seems to be no flag to tell >> the driver it is shutting down so it continues to use the ring buffer >> when it should not. >> >> Fix this by disabling the interrupts in the timeout handler. >> >> [1] >> >> -- cut -- >> ravb e6800000.ethernet ethernet: transmit timed out, status 00000000, >> resetting... >> ravb e6800000.ethernet ethernet: failed to switch device to config mode >> Unable to handle kernel NULL pointer dereference at virtual address >> 00000018 >> Mem abort info: >> Exception class = DABT (current EL), IL = 32 bits >> SET = 0, FnV = 0 >> EA = 0, S1PTW = 0 >> Data abort info: >> ISV = 0, ISS = 0x00000046 >> CM = 0, WnR = 1 >> user pgtable: 4k pages, 48-bit VAs, pgd = ffff80065622f000 >> [0000000000000018] *pgd=00000006962a7003 >> , *pud=00000006962a8003 >> , *pmd=0000000000000000 >> Internal error: Oops: 96000046 [#1] PREEMPT SMP >> Modules linked in: >> ... >> Process Thread1 (pid: 3132, stack limit = 0xffff000027dd0000) >> CPU: 2 PID: 3132 Comm: Thread1 Tainted: G WC >> 4.14.130-ltsi-g28acae87 #1 >> Hardware name: Board based on r8a7796 (DT) >> task: ffff80064f2aaa00 task.stack: ffff000027dd0000 >> PC is at ravb_start_xmit+0x138/0x5a0 >> LR is at ravb_start_xmit+0x40/0x5a0 >> pc : [<ffff0000084d6324>] lr : [<ffff0000084d622c>] pstate: 600001c5 >> sp : ffff000027dd3550 >> x29: ffff000027dd3550 >> x28: 0000000000000076 >> x27: ffff80061035ff00 >> x26: ffff000027dd3694 >> x25: ffff80069624f800 >> x24: ffff80069624f000 >> x23: 0000000000000003 >> x22: 0000000000000001 >> x21: ffff80069624f000 >> x20: 0000000000000000 >> x19: ffff80069624f000 >> x18: 0000000000000014 >> x17: 0000ffff9b90ddb0 >> x16: ffff00000867d07c >> x15: 0000155107b31031 >> x14: 000409000c000000 >> x13: 0000000003000001 >> x12: 0100050010000001 >> x11: 0000000003000001 >> x10: 0100010010000001 >> x9 : 20000000000000c0 >> x8 : 0000000000000000 >> x7 : ffff8006656f9388 >> x6 : 0000000000000002 >> x5 : 0000000000000000 >> x4 : ffff8006656f929c >> x3 : ffff000027dd3694 >> x2 : 0000000000000018 >> x1 : 0000000000000000 >> x0 : 0000000000000003 >> Call trace: >> Exception stack(0xffff000027dd3410 to 0xffff000027dd3550) >> 3400: 0000000000000003 0000000000000000 >> 3420: 0000000000000018 ffff000027dd3694 ffff8006656f929c 0000000000000000 >> 3440: 0000000000000002 ffff8006656f9388 0000000000000000 20000000000000c0 >> 3460: 0100010010000001 0000000003000001 0100050010000001 0000000003000001 >> 3480: 000409000c000000 0000155107b31031 ffff00000867d07c 0000ffff9b90ddb0 >> 34a0: 0000000000000014 ffff80069624f000 0000000000000000 ffff80069624f000 >> 34c0: 0000000000000001 0000000000000003 ffff80069624f000 ffff80069624f800 >> 34e0: ffff000027dd3694 ffff80061035ff00 0000000000000076 ffff000027dd3550 >> 3500: ffff0000084d622c ffff000027dd3550 ffff0000084d6324 00000000600001c5 >> 3520: ffff00000921d008 ffff8006159b0d00 0000ffffffffffff ffff0000084d622c >> 3540: ffff000027dd3550 ffff0000084d6324 >> [<ffff0000084d6324>] ravb_start_xmit+0x138/0x5a0 >> [<ffff000008699d60>] dev_hard_start_xmit+0xa8/0x24c >> [<ffff0000086c51c0>] sch_direct_xmit+0xb0/0x1a8 >> [<ffff0000086c54cc>] __qdisc_run+0x214/0x2ec >> [<ffff00000869a348>] __dev_queue_xmit+0x35c/0x5b4 >> [<ffff00000869a5b0>] dev_queue_xmit+0x10/0x18 >> [<ffff000000b9765c>] register_vlan_dev+0xc74/0x10f8 [8021q] >> [<ffff000008699d60>] dev_hard_start_xmit+0xa8/0x24c >> [<ffff00000869a438>] __dev_queue_xmit+0x44c/0x5b4 >> [<ffff00000869a5b0>] dev_queue_xmit+0x10/0x18 >> [<ffff0000086a5b9c>] neigh_connected_output+0xc0/0xe4 >> [<ffff0000086d6bd8>] ip_finish_output2+0x3c0/0x3fc >> [<ffff0000086d8178>] ip_finish_output+0xf8/0x1c4 >> [<ffff0000086d8b74>] ip_mc_output+0x258/0x308 >> [<ffff0000086d8400>] ip_local_out+0x44/0x54 >> [<ffff0000086d9594>] ip_send_skb+0x1c/0xa8 >> [<ffff0000086fea4c>] udp_send_skb+0x11c/0x244 >> [<ffff000008701084>] udp_sendmsg+0x534/0x6bc >> [<ffff00000870b390>] inet_sendmsg+0x40/0xe0 >> [<ffff00000867b588>] sock_sendmsg+0x3c/0x58 >> [<ffff00000867bf18>] ___sys_sendmsg+0x228/0x278 >> [<ffff00000867d03c>] __sys_sendmsg+0x58/0x98 >> [<ffff00000867d08c>] SyS_sendmsg+0x10/0x20 >> Exception stack(0xffff000027dd3ec0 to 0xffff000027dd4000) >> 3ec0: 0000000000000012 0000ffff8a7fdf18 0000000000004000 0000000000000000 >> 3ee0: 0000ffff8a7ff258 0000ffff8a7ff150 0000ffff8a7ff840 0000000000000000 >> 3f00: 00000000000000d3 0100010010000001 0000000003000001 0100050010000001 >> 3f20: 0000000003000001 000409000c000000 0000000000000047 0000155107b31031 >> 3f40: 0000ffff9b67dfb8 0000ffff9b90ddb0 0000000000000014 0000000000004000 >> 3f60: 0000000000000012 0000ffff8a7fdf18 0000ffff780017b0 0000000000004000 >> 3f80: 0000000000000010 0000000000000001 0000ffff8a7fdf00 0000ffff9000b770 >> 3fa0: 0000000000000012 0000ffff8a7fde60 0000ffff9b90de10 0000ffff8a7fde60 >> 3fc0: 0000ffff9b90de28 0000000080000000 0000000000000012 00000000000000d3 >> 3fe0: 0000000000000000 0000000000000000 0000000000000000 0000000000000000 >> [<ffff0000080832c0>] el0_svc_naked+0x34/0x38 >> Code: d37d7c01 d37d7c02 f90037a1 f9445f01 (f822683b) >> ---[ end trace eabda93d178d5bcb ]--- >> Kernel panic - not syncing: Fatal exception in interrupt >> SMP: stopping secondary CPUs >> Kernel Offset: disabled >> CPU features: 0x1802004 >> Memory Limit: 6144 MB >> -- cut -- >> >> Fixes: c156633f1353 ("Renesas Ethernet AVB driver proper") >> Reviewed-by: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com> >> Signed-off-by: Dirk Behme <dirk.behme@de.bosch.com> > [...] > > You must post this to netdev@vget.kernel.org, else the patch won't > get applied, ever... If I understood correctly, this patch is replaced by https://patchwork.kernel.org/patch/11570217/ Best regards Dirk
diff --git a/drivers/net/ethernet/renesas/ravb_main.c b/drivers/net/ethernet/renesas/ravb_main.c index 067ad25553b9..0f91ab41b22b 100644 --- a/drivers/net/ethernet/renesas/ravb_main.c +++ b/drivers/net/ethernet/renesas/ravb_main.c @@ -1447,6 +1447,11 @@ static void ravb_tx_timeout_work(struct work_struct *work) netif_tx_stop_all_queues(ndev); + /* Disable interrupts by clearing the interrupt masks. */ + ravb_write(ndev, 0, RIC0); + ravb_write(ndev, 0, RIC2); + ravb_write(ndev, 0, TIC); + /* Stop PTP Clock driver */ if (priv->chip_id == RCAR_GEN2) ravb_ptp_stop(ndev);