* [rdma-core patch] srp_daemon: print maximum initiator to target IU size @ 2019-09-16 1:36 Honggang LI 2019-09-17 17:19 ` Bart Van Assche 0 siblings, 1 reply; 7+ messages in thread From: Honggang LI @ 2019-09-16 1:36 UTC (permalink / raw) To: linux-rdma, bvanassche; +Cc: Honggang Li From: Honggang Li <honli@redhat.com> The 'Send Message Size' field of IOControllerProfile attributes contains the maximum initiator to target IU size. When there is something wrong with SRP login to a third party SRP target, whose ib_srpt parameters can't be collected with ordinary method, dump the 'Send Message Size' may help us to diagnose the problem. Signed-off-by: Honggang Li <honli@redhat.com> --- srp_daemon/srp_daemon.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/srp_daemon/srp_daemon.c b/srp_daemon/srp_daemon.c index 337b21c7..90533c77 100644 --- a/srp_daemon/srp_daemon.c +++ b/srp_daemon/srp_daemon.c @@ -1022,6 +1022,8 @@ static int do_port(struct resources *res, uint16_t pkey, uint16_t dlid, pr_human(" vendor ID: %06x\n", be32toh(target->ioc_prof.vendor_id) >> 8); pr_human(" device ID: %06x\n", be32toh(target->ioc_prof.device_id)); pr_human(" IO class : %04hx\n", be16toh(target->ioc_prof.io_class)); + pr_human(" Maximum initiator to target IU size: %d\n", + be32toh(target->ioc_prof.send_size)); pr_human(" ID: %s\n", target->ioc_prof.id); pr_human(" service entries: %d\n", target->ioc_prof.service_entries); -- 2.21.0 ^ permalink raw reply related [flat|nested] 7+ messages in thread
* Re: [rdma-core patch] srp_daemon: print maximum initiator to target IU size 2019-09-16 1:36 [rdma-core patch] srp_daemon: print maximum initiator to target IU size Honggang LI @ 2019-09-17 17:19 ` Bart Van Assche 2019-09-18 0:55 ` Honggang LI 0 siblings, 1 reply; 7+ messages in thread From: Bart Van Assche @ 2019-09-17 17:19 UTC (permalink / raw) To: Honggang LI, linux-rdma On 9/15/19 6:36 PM, Honggang LI wrote: > From: Honggang Li <honli@redhat.com> > > The 'Send Message Size' field of IOControllerProfile attributes > contains the maximum initiator to target IU size. > > When there is something wrong with SRP login to a third party > SRP target, whose ib_srpt parameters can't be collected with > ordinary method, dump the 'Send Message Size' may help us to > diagnose the problem. > > Signed-off-by: Honggang Li <honli@redhat.com> > --- > srp_daemon/srp_daemon.c | 2 ++ > 1 file changed, 2 insertions(+) > > diff --git a/srp_daemon/srp_daemon.c b/srp_daemon/srp_daemon.c > index 337b21c7..90533c77 100644 > --- a/srp_daemon/srp_daemon.c > +++ b/srp_daemon/srp_daemon.c > @@ -1022,6 +1022,8 @@ static int do_port(struct resources *res, uint16_t pkey, uint16_t dlid, > pr_human(" vendor ID: %06x\n", be32toh(target->ioc_prof.vendor_id) >> 8); > pr_human(" device ID: %06x\n", be32toh(target->ioc_prof.device_id)); > pr_human(" IO class : %04hx\n", be16toh(target->ioc_prof.io_class)); > + pr_human(" Maximum initiator to target IU size: %d\n", > + be32toh(target->ioc_prof.send_size)); > pr_human(" ID: %s\n", target->ioc_prof.id); > pr_human(" service entries: %d\n", target->ioc_prof.service_entries); How about using the terminology from the InfiniBand Architecture Specification? This is what I found in release 1.3, table 306: "Maximum size of Send Messages in bytes" Thanks, Bart. ^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [rdma-core patch] srp_daemon: print maximum initiator to target IU size 2019-09-17 17:19 ` Bart Van Assche @ 2019-09-18 0:55 ` Honggang LI 2019-09-23 16:59 ` Bart Van Assche 0 siblings, 1 reply; 7+ messages in thread From: Honggang LI @ 2019-09-18 0:55 UTC (permalink / raw) To: Bart Van Assche; +Cc: linux-rdma On Tue, Sep 17, 2019 at 10:19:37AM -0700, Bart Van Assche wrote: > On 9/15/19 6:36 PM, Honggang LI wrote: > > From: Honggang Li <honli@redhat.com> > > > > The 'Send Message Size' field of IOControllerProfile attributes > > contains the maximum initiator to target IU size. > > > > When there is something wrong with SRP login to a third party > > SRP target, whose ib_srpt parameters can't be collected with > > ordinary method, dump the 'Send Message Size' may help us to > > diagnose the problem. > > > > Signed-off-by: Honggang Li <honli@redhat.com> > > --- > > srp_daemon/srp_daemon.c | 2 ++ > > 1 file changed, 2 insertions(+) > > > > diff --git a/srp_daemon/srp_daemon.c b/srp_daemon/srp_daemon.c > > index 337b21c7..90533c77 100644 > > --- a/srp_daemon/srp_daemon.c > > +++ b/srp_daemon/srp_daemon.c > > @@ -1022,6 +1022,8 @@ static int do_port(struct resources *res, uint16_t pkey, uint16_t dlid, > > pr_human(" vendor ID: %06x\n", be32toh(target->ioc_prof.vendor_id) >> 8); > > pr_human(" device ID: %06x\n", be32toh(target->ioc_prof.device_id)); > > pr_human(" IO class : %04hx\n", be16toh(target->ioc_prof.io_class)); > > + pr_human(" Maximum initiator to target IU size: %d\n", > > + be32toh(target->ioc_prof.send_size)); > > pr_human(" ID: %s\n", target->ioc_prof.id); > > pr_human(" service entries: %d\n", target->ioc_prof.service_entries); > > How about using the terminology from the InfiniBand Architecture As this is srp specific, so I suggest to use the terminology from srp specification 'srp2r06'. Table B.7 — IOControllerProfile attributes for SRP target ports ---------------------------------------------------------------- | Field | SRP requirement | ---------------------------------------------------------------- |(skip many lines)..... | ---------------------------------------------------------------- |Send Message Size |MAXIMUM INITIATOR TO TARGET IU LENGTH | ---------------------------------------------------------------- > Specification? This is what I found in release 1.3, table 306: > > "Maximum size of Send Messages in bytes" > > Thanks, > > Bart. ^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [rdma-core patch] srp_daemon: print maximum initiator to target IU size 2019-09-18 0:55 ` Honggang LI @ 2019-09-23 16:59 ` Bart Van Assche 2019-09-24 1:59 ` Honggang LI 0 siblings, 1 reply; 7+ messages in thread From: Bart Van Assche @ 2019-09-23 16:59 UTC (permalink / raw) To: Honggang LI; +Cc: linux-rdma On 9/17/19 5:55 PM, Honggang LI wrote: > On Tue, Sep 17, 2019 at 10:19:37AM -0700, Bart Van Assche wrote: >> On 9/15/19 6:36 PM, Honggang LI wrote: >>> From: Honggang Li <honli@redhat.com> >>> >>> The 'Send Message Size' field of IOControllerProfile attributes >>> contains the maximum initiator to target IU size. >>> >>> When there is something wrong with SRP login to a third party >>> SRP target, whose ib_srpt parameters can't be collected with >>> ordinary method, dump the 'Send Message Size' may help us to >>> diagnose the problem. >>> >>> Signed-off-by: Honggang Li <honli@redhat.com> >>> --- >>> srp_daemon/srp_daemon.c | 2 ++ >>> 1 file changed, 2 insertions(+) >>> >>> diff --git a/srp_daemon/srp_daemon.c b/srp_daemon/srp_daemon.c >>> index 337b21c7..90533c77 100644 >>> --- a/srp_daemon/srp_daemon.c >>> +++ b/srp_daemon/srp_daemon.c >>> @@ -1022,6 +1022,8 @@ static int do_port(struct resources *res, uint16_t pkey, uint16_t dlid, >>> pr_human(" vendor ID: %06x\n", be32toh(target->ioc_prof.vendor_id) >> 8); >>> pr_human(" device ID: %06x\n", be32toh(target->ioc_prof.device_id)); >>> pr_human(" IO class : %04hx\n", be16toh(target->ioc_prof.io_class)); >>> + pr_human(" Maximum initiator to target IU size: %d\n", >>> + be32toh(target->ioc_prof.send_size)); >>> pr_human(" ID: %s\n", target->ioc_prof.id); >>> pr_human(" service entries: %d\n", target->ioc_prof.service_entries); >> >> How about using the terminology from the InfiniBand Architecture > > As this is srp specific, so I suggest to use the terminology from > srp specification 'srp2r06'. > > Table B.7 — IOControllerProfile attributes for SRP target ports > ---------------------------------------------------------------- > | Field | SRP requirement | > ---------------------------------------------------------------- > |(skip many lines)..... | > ---------------------------------------------------------------- > |Send Message Size |MAXIMUM INITIATOR TO TARGET IU LENGTH | > ---------------------------------------------------------------- > >> Specification? This is what I found in release 1.3, table 306: >> >> "Maximum size of Send Messages in bytes" I don't have a strong opinion about which description to use. The latter may be easier to comprehend though. I'm not sure whether every SRP user knows what an SRP "IU" is ... Bart. ^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [rdma-core patch] srp_daemon: print maximum initiator to target IU size 2019-09-23 16:59 ` Bart Van Assche @ 2019-09-24 1:59 ` Honggang LI 2019-09-24 8:03 ` Leon Romanovsky 2019-09-24 15:36 ` Bart Van Assche 0 siblings, 2 replies; 7+ messages in thread From: Honggang LI @ 2019-09-24 1:59 UTC (permalink / raw) To: Bart Van Assche; +Cc: linux-rdma On Mon, Sep 23, 2019 at 09:59:10AM -0700, Bart Van Assche wrote: > On 9/17/19 5:55 PM, Honggang LI wrote: > > On Tue, Sep 17, 2019 at 10:19:37AM -0700, Bart Van Assche wrote: > > > On 9/15/19 6:36 PM, Honggang LI wrote: > > > > From: Honggang Li <honli@redhat.com> > > > > > > > > The 'Send Message Size' field of IOControllerProfile attributes > > > > contains the maximum initiator to target IU size. > > > > > > > > When there is something wrong with SRP login to a third party > > > > SRP target, whose ib_srpt parameters can't be collected with > > > > ordinary method, dump the 'Send Message Size' may help us to > > > > diagnose the problem. > > > > > > > > Signed-off-by: Honggang Li <honli@redhat.com> > > > > --- > > > > srp_daemon/srp_daemon.c | 2 ++ > > > > 1 file changed, 2 insertions(+) > > > > > > > > diff --git a/srp_daemon/srp_daemon.c b/srp_daemon/srp_daemon.c > > > > index 337b21c7..90533c77 100644 > > > > --- a/srp_daemon/srp_daemon.c > > > > +++ b/srp_daemon/srp_daemon.c > > > > @@ -1022,6 +1022,8 @@ static int do_port(struct resources *res, uint16_t pkey, uint16_t dlid, > > > > pr_human(" vendor ID: %06x\n", be32toh(target->ioc_prof.vendor_id) >> 8); > > > > pr_human(" device ID: %06x\n", be32toh(target->ioc_prof.device_id)); > > > > pr_human(" IO class : %04hx\n", be16toh(target->ioc_prof.io_class)); > > > > + pr_human(" Maximum initiator to target IU size: %d\n", > > > > + be32toh(target->ioc_prof.send_size)); > > > > pr_human(" ID: %s\n", target->ioc_prof.id); > > > > pr_human(" service entries: %d\n", target->ioc_prof.service_entries); > > > > > > How about using the terminology from the InfiniBand Architecture > > > > As this is srp specific, so I suggest to use the terminology from > > srp specification 'srp2r06'. > > > > Table B.7 — IOControllerProfile attributes for SRP target ports > > ---------------------------------------------------------------- > > | Field | SRP requirement | > > ---------------------------------------------------------------- > > |(skip many lines)..... | > > ---------------------------------------------------------------- > > |Send Message Size |MAXIMUM INITIATOR TO TARGET IU LENGTH | > > ---------------------------------------------------------------- > > > > > Specification? This is what I found in release 1.3, table 306: > > > > > > "Maximum size of Send Messages in bytes" > > I don't have a strong opinion about which description to use. The latter may > be easier to comprehend though. I'm not sure whether every SRP user knows > what an SRP "IU" is ... How about replacing "IU" with "information unit"? + pr_human(" Maximum initiator to target information unit size: %d\n", + be32toh(target->ioc_prof.send_size)); If you want to use "Maximum size of Send Messages in bytes", just let me know, I will send a new patch to use that. thanks ^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [rdma-core patch] srp_daemon: print maximum initiator to target IU size 2019-09-24 1:59 ` Honggang LI @ 2019-09-24 8:03 ` Leon Romanovsky 2019-09-24 15:36 ` Bart Van Assche 1 sibling, 0 replies; 7+ messages in thread From: Leon Romanovsky @ 2019-09-24 8:03 UTC (permalink / raw) To: Honggang LI; +Cc: Bart Van Assche, linux-rdma On Tue, Sep 24, 2019 at 09:59:36AM +0800, Honggang LI wrote: > On Mon, Sep 23, 2019 at 09:59:10AM -0700, Bart Van Assche wrote: > > On 9/17/19 5:55 PM, Honggang LI wrote: > > > On Tue, Sep 17, 2019 at 10:19:37AM -0700, Bart Van Assche wrote: > > > > On 9/15/19 6:36 PM, Honggang LI wrote: > > > > > From: Honggang Li <honli@redhat.com> > > > > > > > > > > The 'Send Message Size' field of IOControllerProfile attributes > > > > > contains the maximum initiator to target IU size. > > > > > > > > > > When there is something wrong with SRP login to a third party > > > > > SRP target, whose ib_srpt parameters can't be collected with > > > > > ordinary method, dump the 'Send Message Size' may help us to > > > > > diagnose the problem. > > > > > > > > > > Signed-off-by: Honggang Li <honli@redhat.com> > > > > > --- > > > > > srp_daemon/srp_daemon.c | 2 ++ > > > > > 1 file changed, 2 insertions(+) > > > > > > > > > > diff --git a/srp_daemon/srp_daemon.c b/srp_daemon/srp_daemon.c > > > > > index 337b21c7..90533c77 100644 > > > > > --- a/srp_daemon/srp_daemon.c > > > > > +++ b/srp_daemon/srp_daemon.c > > > > > @@ -1022,6 +1022,8 @@ static int do_port(struct resources *res, uint16_t pkey, uint16_t dlid, > > > > > pr_human(" vendor ID: %06x\n", be32toh(target->ioc_prof.vendor_id) >> 8); > > > > > pr_human(" device ID: %06x\n", be32toh(target->ioc_prof.device_id)); > > > > > pr_human(" IO class : %04hx\n", be16toh(target->ioc_prof.io_class)); > > > > > + pr_human(" Maximum initiator to target IU size: %d\n", > > > > > + be32toh(target->ioc_prof.send_size)); > > > > > pr_human(" ID: %s\n", target->ioc_prof.id); > > > > > pr_human(" service entries: %d\n", target->ioc_prof.service_entries); > > > > > > > > How about using the terminology from the InfiniBand Architecture > > > > > > As this is srp specific, so I suggest to use the terminology from > > > srp specification 'srp2r06'. > > > > > > Table B.7 — IOControllerProfile attributes for SRP target ports > > > ---------------------------------------------------------------- > > > | Field | SRP requirement | > > > ---------------------------------------------------------------- > > > |(skip many lines)..... | > > > ---------------------------------------------------------------- > > > |Send Message Size |MAXIMUM INITIATOR TO TARGET IU LENGTH | > > > ---------------------------------------------------------------- > > > > > > > Specification? This is what I found in release 1.3, table 306: > > > > > > > > "Maximum size of Send Messages in bytes" > > > > I don't have a strong opinion about which description to use. The latter may > > be easier to comprehend though. I'm not sure whether every SRP user knows > > what an SRP "IU" is ... > > How about replacing "IU" with "information unit"? > > > + pr_human(" Maximum initiator to target information unit size: %d\n", > + be32toh(target->ioc_prof.send_size)); > > If you want to use "Maximum size of Send Messages in bytes", > just let me know, I will send a new patch to use that. What is the resolution? Thanks > > thanks > ^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [rdma-core patch] srp_daemon: print maximum initiator to target IU size 2019-09-24 1:59 ` Honggang LI 2019-09-24 8:03 ` Leon Romanovsky @ 2019-09-24 15:36 ` Bart Van Assche 1 sibling, 0 replies; 7+ messages in thread From: Bart Van Assche @ 2019-09-24 15:36 UTC (permalink / raw) To: Honggang LI; +Cc: linux-rdma On 9/23/19 6:59 PM, Honggang LI wrote: > If you want to use "Maximum size of Send Messages in bytes", > just let me know, I will send a new patch to use that. I think that description is short and easy to read. Thanks, Bart. ^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~2019-09-24 15:36 UTC | newest] Thread overview: 7+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2019-09-16 1:36 [rdma-core patch] srp_daemon: print maximum initiator to target IU size Honggang LI 2019-09-17 17:19 ` Bart Van Assche 2019-09-18 0:55 ` Honggang LI 2019-09-23 16:59 ` Bart Van Assche 2019-09-24 1:59 ` Honggang LI 2019-09-24 8:03 ` Leon Romanovsky 2019-09-24 15:36 ` Bart Van Assche
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox