* [PATCH] nvme-fabrics: fix kato initialization
@ 2021-02-18 10:04 Martin George
2021-02-19 16:42 ` Hannes Reinecke
0 siblings, 1 reply; 5+ messages in thread
From: Martin George @ 2021-02-18 10:04 UTC (permalink / raw)
To: linux-nvme; +Cc: Martin George
Currently kato is initialized to NVME_DEFAULT_KATO for both
discovery & i/o controllers. This is a problem specifically
for non-persistent discovery controllers since it always ends
up with a non-zero kato value. Fix this by initializing kato
to zero instead, and ensuring various controllers are assigned
appropriate kato values as follows:
non-persistent controllers - kato set to zero
persistent controllers - kato set to NVMF_DEV_DISC_TMO
(or any positive int via nvme-cli)
i/o controllers - kato set to NVME_DEFAULT_KATO
(or any positive int via nvme-cli)
Signed-off-by: Martin George <marting@netapp.com>
---
drivers/nvme/host/fabrics.c | 5 ++++-
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/drivers/nvme/host/fabrics.c b/drivers/nvme/host/fabrics.c
index 5dfd806fc2d2..604ab0e5a2ad 100644
--- a/drivers/nvme/host/fabrics.c
+++ b/drivers/nvme/host/fabrics.c
@@ -630,7 +630,7 @@ static int nvmf_parse_options(struct nvmf_ctrl_options *opts,
opts->queue_size = NVMF_DEF_QUEUE_SIZE;
opts->nr_io_queues = num_online_cpus();
opts->reconnect_delay = NVMF_DEF_RECONNECT_DELAY;
- opts->kato = NVME_DEFAULT_KATO;
+ opts->kato = 0;
opts->duplicate_connect = false;
opts->fast_io_fail_tmo = NVMF_DEF_FAIL_FAST_TMO;
opts->hdr_digest = false;
@@ -893,6 +893,9 @@ static int nvmf_parse_options(struct nvmf_ctrl_options *opts,
opts->nr_write_queues = 0;
opts->nr_poll_queues = 0;
opts->duplicate_connect = true;
+ } else {
+ if (!opts->kato)
+ opts->kato = NVME_DEFAULT_KATO;
}
if (ctrl_loss_tmo < 0) {
opts->max_reconnects = -1;
--
2.30.0
_______________________________________________
Linux-nvme mailing list
Linux-nvme@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-nvme
^ permalink raw reply related [flat|nested] 5+ messages in thread* Re: [PATCH] nvme-fabrics: fix kato initialization 2021-02-18 10:04 [PATCH] nvme-fabrics: fix kato initialization Martin George @ 2021-02-19 16:42 ` Hannes Reinecke 2021-02-20 8:41 ` George, Martin 0 siblings, 1 reply; 5+ messages in thread From: Hannes Reinecke @ 2021-02-19 16:42 UTC (permalink / raw) To: linux-nvme On 2/18/21 11:04 AM, Martin George wrote: > Currently kato is initialized to NVME_DEFAULT_KATO for both > discovery & i/o controllers. This is a problem specifically > for non-persistent discovery controllers since it always ends > up with a non-zero kato value. Fix this by initializing kato > to zero instead, and ensuring various controllers are assigned > appropriate kato values as follows: > > non-persistent controllers - kato set to zero > persistent controllers - kato set to NVMF_DEV_DISC_TMO > (or any positive int via nvme-cli) > i/o controllers - kato set to NVME_DEFAULT_KATO > (or any positive int via nvme-cli) > > Signed-off-by: Martin George <marting@netapp.com> > --- > drivers/nvme/host/fabrics.c | 5 ++++- > 1 file changed, 4 insertions(+), 1 deletion(-) > > diff --git a/drivers/nvme/host/fabrics.c b/drivers/nvme/host/fabrics.c > index 5dfd806fc2d2..604ab0e5a2ad 100644 > --- a/drivers/nvme/host/fabrics.c > +++ b/drivers/nvme/host/fabrics.c > @@ -630,7 +630,7 @@ static int nvmf_parse_options(struct nvmf_ctrl_options *opts, > opts->queue_size = NVMF_DEF_QUEUE_SIZE; > opts->nr_io_queues = num_online_cpus(); > opts->reconnect_delay = NVMF_DEF_RECONNECT_DELAY; > - opts->kato = NVME_DEFAULT_KATO; > + opts->kato = 0; > opts->duplicate_connect = false; > opts->fast_io_fail_tmo = NVMF_DEF_FAIL_FAST_TMO; > opts->hdr_digest = false; > @@ -893,6 +893,9 @@ static int nvmf_parse_options(struct nvmf_ctrl_options *opts, > opts->nr_write_queues = 0; > opts->nr_poll_queues = 0; > opts->duplicate_connect = true; > + } else { > + if (!opts->kato) > + opts->kato = NVME_DEFAULT_KATO; > } > if (ctrl_loss_tmo < 0) { > opts->max_reconnects = -1; > Can't you merge this with the previous patch 'ensure zero-kato ..'? Cheers, Hannes -- Dr. Hannes Reinecke Kernel Storage Architect hare@suse.de +49 911 74053 688 SUSE Software Solutions GmbH, Maxfeldstr. 5, 90409 Nürnberg HRB 36809 (AG Nürnberg), Geschäftsführer: Felix Imendörffer _______________________________________________ Linux-nvme mailing list Linux-nvme@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-nvme ^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH] nvme-fabrics: fix kato initialization 2021-02-19 16:42 ` Hannes Reinecke @ 2021-02-20 8:41 ` George, Martin 2021-02-20 14:59 ` Hannes Reinecke 0 siblings, 1 reply; 5+ messages in thread From: George, Martin @ 2021-02-20 8:41 UTC (permalink / raw) To: hch@lst.de, kbusch@kernel.org, sagi@grimberg.me, hare@suse.de, linux-nvme@lists.infradead.org On Fri, 2021-02-19 at 17:42 +0100, Hannes Reinecke wrote: > On 2/18/21 11:04 AM, Martin George wrote: > > Currently kato is initialized to NVME_DEFAULT_KATO for both > > discovery & i/o controllers. This is a problem specifically > > for non-persistent discovery controllers since it always ends > > up with a non-zero kato value. Fix this by initializing kato > > to zero instead, and ensuring various controllers are assigned > > appropriate kato values as follows: > > > > non-persistent controllers - kato set to zero > > persistent controllers - kato set to NVMF_DEV_DISC_TMO > > (or any positive int via nvme-cli) > > i/o controllers - kato set to NVME_DEFAULT_KATO > > (or any positive int via nvme-cli) > > > > Signed-off-by: Martin George <marting@netapp.com> > > --- > > drivers/nvme/host/fabrics.c | 5 ++++- > > 1 file changed, 4 insertions(+), 1 deletion(-) > > > > diff --git a/drivers/nvme/host/fabrics.c > > b/drivers/nvme/host/fabrics.c > > index 5dfd806fc2d2..604ab0e5a2ad 100644 > > --- a/drivers/nvme/host/fabrics.c > > +++ b/drivers/nvme/host/fabrics.c > > @@ -630,7 +630,7 @@ static int nvmf_parse_options(struct > > nvmf_ctrl_options *opts, > > opts->queue_size = NVMF_DEF_QUEUE_SIZE; > > opts->nr_io_queues = num_online_cpus(); > > opts->reconnect_delay = NVMF_DEF_RECONNECT_DELAY; > > - opts->kato = NVME_DEFAULT_KATO; > > + opts->kato = 0; > > opts->duplicate_connect = false; > > opts->fast_io_fail_tmo = NVMF_DEF_FAIL_FAST_TMO; > > opts->hdr_digest = false; > > @@ -893,6 +893,9 @@ static int nvmf_parse_options(struct > > nvmf_ctrl_options *opts, > > opts->nr_write_queues = 0; > > opts->nr_poll_queues = 0; > > opts->duplicate_connect = true; > > + } else { > > + if (!opts->kato) > > + opts->kato = NVME_DEFAULT_KATO; > > } > > if (ctrl_loss_tmo < 0) { > > opts->max_reconnects = -1; > > > Can't you merge this with the previous patch 'ensure zero-kato ..'? > Well, you are right that both the previous patch (i.e. 'ensure zero kato for non-persistent controllers') & this current patch are jointly required for assigning proper kato values to the various controllers, so it would have made sense to merge the two. But the previous one is a nvme-cli patch, whereas this current one is a kernel patch. That's why I chose to create separate individual patches itself here. So I hope this is good enough. If not, please let me know what's the right way to proceed here. -Martin _______________________________________________ Linux-nvme mailing list Linux-nvme@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-nvme ^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH] nvme-fabrics: fix kato initialization 2021-02-20 8:41 ` George, Martin @ 2021-02-20 14:59 ` Hannes Reinecke 0 siblings, 0 replies; 5+ messages in thread From: Hannes Reinecke @ 2021-02-20 14:59 UTC (permalink / raw) To: George, Martin, hch@lst.de, kbusch@kernel.org, sagi@grimberg.me, linux-nvme@lists.infradead.org On 2/20/21 9:41 AM, George, Martin wrote: > On Fri, 2021-02-19 at 17:42 +0100, Hannes Reinecke wrote: >> On 2/18/21 11:04 AM, Martin George wrote: >>> Currently kato is initialized to NVME_DEFAULT_KATO for both >>> discovery & i/o controllers. This is a problem specifically >>> for non-persistent discovery controllers since it always ends >>> up with a non-zero kato value. Fix this by initializing kato >>> to zero instead, and ensuring various controllers are assigned >>> appropriate kato values as follows: >>> >>> non-persistent controllers - kato set to zero >>> persistent controllers - kato set to NVMF_DEV_DISC_TMO >>> (or any positive int via nvme-cli) >>> i/o controllers - kato set to NVME_DEFAULT_KATO >>> (or any positive int via nvme-cli) >>> >>> Signed-off-by: Martin George <marting@netapp.com> >>> --- >>> drivers/nvme/host/fabrics.c | 5 ++++- >>> 1 file changed, 4 insertions(+), 1 deletion(-) >>> >>> diff --git a/drivers/nvme/host/fabrics.c >>> b/drivers/nvme/host/fabrics.c >>> index 5dfd806fc2d2..604ab0e5a2ad 100644 >>> --- a/drivers/nvme/host/fabrics.c >>> +++ b/drivers/nvme/host/fabrics.c >>> @@ -630,7 +630,7 @@ static int nvmf_parse_options(struct >>> nvmf_ctrl_options *opts, >>> opts->queue_size = NVMF_DEF_QUEUE_SIZE; >>> opts->nr_io_queues = num_online_cpus(); >>> opts->reconnect_delay = NVMF_DEF_RECONNECT_DELAY; >>> - opts->kato = NVME_DEFAULT_KATO; >>> + opts->kato = 0; >>> opts->duplicate_connect = false; >>> opts->fast_io_fail_tmo = NVMF_DEF_FAIL_FAST_TMO; >>> opts->hdr_digest = false; >>> @@ -893,6 +893,9 @@ static int nvmf_parse_options(struct >>> nvmf_ctrl_options *opts, >>> opts->nr_write_queues = 0; >>> opts->nr_poll_queues = 0; >>> opts->duplicate_connect = true; >>> + } else { >>> + if (!opts->kato) >>> + opts->kato = NVME_DEFAULT_KATO; >>> } >>> if (ctrl_loss_tmo < 0) { >>> opts->max_reconnects = -1; >>> >> Can't you merge this with the previous patch 'ensure zero-kato ..'? >> > > Well, you are right that both the previous patch (i.e. 'ensure zero > kato for non-persistent controllers') & this current patch are jointly > required for assigning proper kato values to the various controllers, > so it would have made sense to merge the two. But the previous one is a > nvme-cli patch, whereas this current one is a kernel patch. That's why > I chose to create separate individual patches itself here. > > So I hope this is good enough. If not, please let me know what's the > right way to proceed here. > Argl. Of course you are correct. Cheers, Hannes -- Dr. Hannes Reinecke Kernel Storage Architect hare@suse.de +49 911 74053 688 SUSE Software Solutions GmbH, Maxfeldstr. 5, 90409 Nürnberg HRB 36809 (AG Nürnberg), Geschäftsführer: Felix Imendörffer _______________________________________________ Linux-nvme mailing list Linux-nvme@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-nvme ^ permalink raw reply [flat|nested] 5+ messages in thread
[parent not found: <20210211175826.4821-1-marting@netapp.com>]
* Re: [PATCH] nvme-fabrics: fix kato initialization [not found] <20210211175826.4821-1-marting@netapp.com> @ 2021-02-24 9:14 ` Christoph Hellwig 0 siblings, 0 replies; 5+ messages in thread From: Christoph Hellwig @ 2021-02-24 9:14 UTC (permalink / raw) To: Martin George; +Cc: kbusch, Martin George, hch, linux-nvme, sagi Thanks, applied to nvme-5.12. _______________________________________________ Linux-nvme mailing list Linux-nvme@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-nvme ^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2021-02-24 9:14 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2021-02-18 10:04 [PATCH] nvme-fabrics: fix kato initialization Martin George
2021-02-19 16:42 ` Hannes Reinecke
2021-02-20 8:41 ` George, Martin
2021-02-20 14:59 ` Hannes Reinecke
[not found] <20210211175826.4821-1-marting@netapp.com>
2021-02-24 9:14 ` Christoph Hellwig
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox