On Mon, Oct 23, 2023 at 11:28:33AM +0200, Hannes Reinecke wrote:
> Iterating over all possible bus number in scsi_eh_bus_reset() is
> inefficient as not all busses may be affected during SCSI EH.
> So rewrite the loop in scsi_eh_bus_reset() to match the loop
> in scsi_eh_target_reset() and only loop over failed commands.
>
> Signed-off-by: Hannes Reinecke <hare@xxxxxxx>
> Reviewed-by: Christoph Hellwig <hch@xxxxxx>
> ---
> drivers/scsi/scsi_error.c | 62 ++++++++++++++++-----------------------
> 1 file changed, 25 insertions(+), 37 deletions(-)
>
> diff --git a/drivers/scsi/scsi_error.c b/drivers/scsi/scsi_error.c
> index 42e12756d6f4..7c9c376affda 100644
> --- a/drivers/scsi/scsi_error.c
> +++ b/drivers/scsi/scsi_error.c
> @@ -1716,43 +1715,32 @@ static int scsi_eh_bus_reset(struct Scsi_Host *shost,
> + rtn = scsi_try_bus_reset(scmd);
> + if (rtn != SUCCESS && rtn != FAST_IO_FAIL) {
> SCSI_LOG_ERROR_RECOVERY(3,
> shost_printk(KERN_INFO, shost,
> "%s: BRST failed chan: %d\n",
> current->comm, channel));
> }
> + list_for_each_entry_safe(scmd, next, work_q, eh_entry) {
You probably want to iterate over the `tmp_list`, not the `work_q`.
> + if (scmd_channel(scmd) != channel)
> + continue;
> +
--
Best Regards, Benjamin Block / Linux on IBM Z Kernel Development
IBM Deutschland Research & Development GmbH / https://www.ibm.com/privacy
Vors. Aufs.-R.: Gregor Pillen / Geschäftsführung: David Faller
Sitz der Ges.: Böblingen / Registergericht: AmtsG Stuttgart, HRB 243294
