While the ATA specification states that a device should return command
aborted for all commands queued after the device has entered error state,
since ATA only keeps the sense data for the latest command (in non-NCQ
case), we really don't want to send block layer commands to the device
after it has entered error state. (Only ATA EH commands should be sent,
to read the sense data etc.)

Currently, scsi_queue_rq() will check if scsi_host_in_recovery()
(state is SHOST_RECOVERY), and if so, it will _not_ issue a command via:
scsi_dispatch_cmd() -> host->hostt->queuecommand() (ata_scsi_queuecmd())
-> __ata_scsi_queuecmd() -> ata_scsi_translate() -> ata_qc_issue()

Before commit e494f6a7 ("[SCSI] improved eh timeout handler"),
when receiving a TFES error IRQ, the call chain looked like this:
ahci_error_intr() -> ata_port_abort() -> ata_do_link_abort() ->
ata_qc_complete() -> ata_qc_schedule_eh() -> blk_abort_request() ->
blk_rq_timed_out() -> q->rq_timed_out_fn() (scsi_times_out()) ->
scsi_eh_scmd_add() -> scsi_host_set_state(shost, SHOST_RECOVERY)

Which meant that as soon as an error IRQ was serviced, SHOST_RECOVERY
would be set.

However, after commit e494f6a7 ("[SCSI] improved eh timeout handler"),
scsi_times_out() will instead call scsi_abort_command() which will queue
delayed work, and the worker function scmd_eh_abort_handler() will call
scsi_eh_scmd_add(), which calls scsi_host_set_state(shost, SHOST_RECOVERY).

So now, after the TFES error IRQ has been serviced, we need to wait for
the SCSI workqueue to run its work before SHOST_RECOVERY gets set.

It is worth noting that, even before commit e494f6a7 ("[SCSI] improved
eh timeout handler"), we could receive an error IRQ from the time when
scsi_queue_rq() checks scsi_host_in_recovery(), to the time when
ata_scsi_queuecmd() is actually called.

In order to handle both the delayed setting of SHOST_RECOVERY and the
window where we can receive an error IRQ, add a check against
ATA_PFLAG_EH_PENDING (which gets set when servicing the error IRQ),
inside ata_scsi_queuecmd() itself, while holding the ap->lock.
(Since the ap->lock is held while servicing IRQs.)

Fixes: e494f6a7 ("[SCSI] improved eh timeout handler")
Signed-off-by: Niklas Cassel <niklas.cassel@wdc.com>
Tested-by: John Garry <john.g.garry@oracle.com>
Signed-off-by: Damien Le Moal <damien.lemoal@opensource.wdc.com>