When using SCSI passthrough in combination with the iSCSI target driver then cmd->t_state_lock may be obtained from interrupt context. Hence, all code that obtains cmd->t_state_lock from thread context must disable interrupts first. This patch avoids that lockdep reports the following:
WARNING: inconsistent lock state 4.18.0-dbg+ #1 Not tainted -------------------------------- inconsistent {HARDIRQ-ON-W} -> {IN-HARDIRQ-W} usage. iscsi_ttx/1800 [HC1[1]:SC0[2]:HE0:SE0] takes: 000000006e7b0ceb (&(&cmd->t_state_lock)->rlock){?...}, at: target_complete_cmd+0x47/0x2c0 [target_core_mod] {HARDIRQ-ON-W} state was registered at: lock_acquire+0xd2/0x260 _raw_spin_lock+0x32/0x50 iscsit_close_connection+0x97e/0x1020 [iscsi_target_mod] iscsit_take_action_for_connection_exit+0x108/0x200 [iscsi_target_mod] iscsi_target_rx_thread+0x180/0x190 [iscsi_target_mod] kthread+0x1cf/0x1f0 ret_from_fork+0x24/0x30 irq event stamp: 1281 hardirqs last enabled at (1279): [<ffffffff970ade79>] __local_bh_enable_ip+0xa9/0x160 hardirqs last disabled at (1281): [<ffffffff97a008a5>] interrupt_entry+0xb5/0xd0 softirqs last enabled at (1278): [<ffffffff977cd9a1>] lock_sock_nested+0x51/0xc0 softirqs last disabled at (1280): [<ffffffffc07a6e04>] ip6_finish_output2+0x124/0xe40 [ipv6]
other info that might help us debug this: Possible unsafe locking scenario:
CPU0 ---- lock(&(&cmd->t_state_lock)->rlock); <Interrupt> lock(&(&cmd->t_state_lock)->rlock);
*** DEADLOCK ***
3 locks held by iscsi_ttx/1800: *0: 00000000c3b711b7 (sk_lock-AF_INET6){+.+.}, at: tcp_sendmsg+0x1e/0x50 *1: 00000000fa81046f (rcu_read_lock){....}, at: inet6_csk_xmit+0xc7/0x2e0 [ipv6] *2: 00000000c091d70d (rcu_read_lock_bh){....}, at: ip6_finish_output2+0x124/0xe40 [ipv6]
stack backtrace: CPU: 0 PID: 1800 Comm: iscsi_ttx Not tainted 4.18.0-dbg+ #1 Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.0.0-prebuilt.qemu-project.org 04/01/2014 Call Trace: <IRQ> dump_stack+0xa4/0xf5 print_usage_bug+0x25b/0x27b mark_lock+0x70f/0x7b0 __lock_acquire+0xbc2/0x1b50 lock_acquire+0xd2/0x260 _raw_spin_lock_irqsave+0x4a/0x60 target_complete_cmd+0x47/0x2c0 [target_core_mod] target_complete_cmd_with_length+0x70/0xa0 [target_core_mod] pscsi_req_done+0x335/0x530 [target_core_pscsi] __blk_mq_end_request+0xa5/0x140 scsi_end_request+0x112/0x320 [scsi_mod] scsi_io_completion+0x183/0xa30 [scsi_mod] scsi_finish_command+0x1c0/0x280 [scsi_mod] scsi_softirq_done+0x19a/0x230 [scsi_mod] __blk_mq_complete_request_remote+0x2f/0x40 flush_smp_call_function_queue+0x12a/0x220 generic_smp_call_function_single_interrupt+0x13/0x30 smp_call_function_single_interrupt+0x7a/0x350 call_function_single_interrupt+0xf/0x20 </IRQ> RIP: 0010:__asan_load4+0x1e/0x80 debug_lockdep_rcu_enabled+0x26/0x40 ip6_finish_output2+0x15a/0xe40 [ipv6] ip6_finish_output+0x308/0x440 [ipv6] ip6_output+0x11d/0x3b0 [ipv6] ip6_xmit+0x639/0xc50 [ipv6] inet6_csk_xmit+0x198/0x2e0 [ipv6] __tcp_transmit_skb+0xc1b/0x15b0 tcp_write_xmit+0x42e/0x1f20 __tcp_push_pending_frames+0x59/0x150 tcp_push+0x189/0x270 tcp_sendmsg_locked+0x7b9/0x1680 tcp_sendmsg+0x2c/0x50 inet_sendmsg+0x71/0x250 sock_sendmsg+0x4c/0x60 tx_data+0x12b/0x1f0 [iscsi_target_mod] iscsit_send_tx_data+0x77/0xe0 [iscsi_target_mod] iscsit_xmit_pdu+0x2c5/0x740 [iscsi_target_mod] iscsit_response_queue+0x941/0xd40 [iscsi_target_mod] iscsi_target_tx_thread+0x23b/0x350 [iscsi_target_mod] kthread+0x1cf/0x1f0 ret_from_fork+0x24/0x30
Fixes: 064cdd2d91c2 ("target: Fix race between iscsi-target connection shutdown + ABORT_TASK") Cc: Nicholas Bellinger nab@linux-iscsi.org Cc: Mike Christie mchristi@redhat.com Cc: Christoph Hellwig hch@lst.de Cc: Hannes Reinecke hare@suse.de Cc: stable@vger.kernel.org Signed-off-by: Bart Van Assche bvanassche@acm.org --- drivers/target/iscsi/iscsi_target.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/target/iscsi/iscsi_target.c b/drivers/target/iscsi/iscsi_target.c index 7059ffdb0c20..408d77eb717c 100644 --- a/drivers/target/iscsi/iscsi_target.c +++ b/drivers/target/iscsi/iscsi_target.c @@ -4038,9 +4038,9 @@ static void iscsit_release_commands_from_conn(struct iscsi_conn *conn) struct se_cmd *se_cmd = &cmd->se_cmd;
if (se_cmd->se_tfo != NULL) { - spin_lock(&se_cmd->t_state_lock); + spin_lock_irq(&se_cmd->t_state_lock); se_cmd->transport_state |= CMD_T_FABRIC_STOP; - spin_unlock(&se_cmd->t_state_lock); + spin_unlock_irq(&se_cmd->t_state_lock); } } spin_unlock_bh(&conn->cmd_lock);