On Saturday, June 15, 2024 2:07 AM, Hannes Reinecke <hare@xxxxxxx> wrote: > > On 6/15/24 05:47, Karan Tilak Kumar (kartilak) wrote: > > On Tuesday, June 11, 2024 11:57 PM, Hannes Reinecke <hare@xxxxxxx> wrote: > >> > >> On 6/10/24 23:50, Karan Tilak Kumar wrote: > >>> Add interfaces in fnic to use FDLS services. > >>> Modify link up and link down functionality to use FDLS. > >>> Replace existing interfaces to handle new functionality provided by > >>> FDLS. > >>> Modify data types of some data members to handle new functionality. > >>> Add processing of tports and handling of tports. > >>> > >>> Reviewed-by: Sesidhar Baddela <sebaddel@xxxxxxxxx> > >>> Reviewed-by: Arulprabhu Ponnusamy <arulponn@xxxxxxxxx> > >>> Reviewed-by: Gian Carlo Boffa <gcboffa@xxxxxxxxx> > >>> Signed-off-by: Karan Tilak Kumar <kartilak@xxxxxxxxx> > >>> --- > >>> drivers/scsi/fnic/fdls_disc.c | 74 +++++ > >>> drivers/scsi/fnic/fip.c | 27 +- > >>> drivers/scsi/fnic/fnic.h | 20 +- > >>> drivers/scsi/fnic/fnic_fcs.c | 498 ++++++++++++++++++++++++---------- > >>> drivers/scsi/fnic/fnic_main.c | 10 +- > >>> drivers/scsi/fnic/fnic_scsi.c | 127 +++++++-- > >>> 6 files changed, 587 insertions(+), 169 deletions(-) > >>> > >> This seems to not just _add_ the functionality to use FDLS, but > >> rather _replace_ the existing functionality with FDLS. > >> IE it seems that after this change the driver will always do FDLS, > >> causing a possible service interruption with existing setups. > >> Hmm? > > > > Thanks for your review comments, Hannes. > > As I mentioned in the other patch comments, Cisco has been shipping an > > async driver based on FDLS for the past six years. > > The async driver is backward compatible and supports all the adapters > > that are supported by the existing upstream driver, and more. > > The async driver in fact overrides the upstream driver on our installations. > > > Ah. Good to know. Ever more a reason to have your driver upstreamed, then... Yes Hannes. That's correct. The Cisco team feels that way too. > > On Cisco hardware, the best practice out in the field, is to update > > the driver to the async driver during OS installation itself. > > Due to this best practice, we have _not_ received any feedback from > > customers indicating an abnormal service interruption specifically due to the driver update. > > > Ah. I wasn't aware of that. > So that's fine, then, and you can disregard my comments here. > Sure Hannes. No worries. Regards, Karan
