Re: [Linuxptp-devel] [PATCH] add report of PTP_SYS_OFFSET_PRECISE support to phc_ctl
On Sat, Aug 18, 2018 at 10:07:59AM -0400, Richard Cochran wrote: > On Thu, Aug 16, 2018 at 03:22:51PM +, FUSTE Emmanuel wrote: > > diff --git a/phc_ctl.c b/phc_ctl.c > > index 4a78a19..f6234da 100644 > > --- a/phc_ctl.c > > +++ b/phc_ctl.c > > @@ -353,12 +353,14 @@ static int do_caps(clockid_t clkid, int cmdc, char > > *cmdv[]) > > " %d programable alarms\n" > > " %d external time stamp channels\n" > > " %d programmable periodic signals\n" > > - " %s pulse per second support", > > + " %s pulse per second support\n" > > + " %s precise system-device cross timestamps support", > > This patch looks okay, but we want a newline here --^ Do we? In this case the string is printed by pr_notice(), which should add a newline automatically when printing to terminal. -- Miroslav Lichvar -- Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot ___ Linuxptp-devel mailing list Linuxptp-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
Re: [Linuxptp-devel] [PATCH RFC 0/3] Basis for SNMP support
On Mon, Aug 20, 2018 at 08:03:58AM +, Anders Selhammer wrote: > I guess you want me to update the serie with the updates before, right? Yes, please. IIRC, you had some additional changes... Thanks, Richard -- Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot ___ Linuxptp-devel mailing list Linuxptp-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
Re: [Linuxptp-devel] [PATCH] add report of PTP_SYS_OFFSET_PRECISE support to phc_ctl
On Mon, Aug 20, 2018 at 09:00:33AM +0200, Miroslav Lichvar wrote: > Do we? In this case the string is printed by pr_notice(), which should > add a newline automatically when printing to terminal. Oh, you're right. Never mind! Thanks, Richard -- Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot ___ Linuxptp-devel mailing list Linuxptp-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
Re: [Linuxptp-devel] [PATCH RFC 0/3] Basis for SNMP support
Thursday, July 26, 2018 9:34 AM >Thursday, July 26, 2018 2:39 AM >> Sounds good. >You will apply the series without updates and I include it in next series then? >Then I do not need to poll the mail the next few days 😊 Hi Im back from vacation and since you have not applied the patches yet, I guess you want me to update the serie with the updates before, right? /Anders -- Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot ___ Linuxptp-devel mailing list Linuxptp-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
[Linuxptp-devel] Master still sends out DELAY_REQ
Hi,
We found a tricky scenario is that after port goes from SLAVE->MASTER state, it
still sends out DELAY_REQ continuously by default interval. Given that we are
running interval as -6, the delay interval is random(0, 2*2pow interval), the
timer could be short.
Could the reason be like this?
1. Event we cancel the FD_DELAY_TIMER?
static void port_e2e_transition(struct port *p, enum port_state next)
{
port_clr_tmo(p->fda.fd[FD_ANNOUNCE_TIMER]);
port_clr_tmo(p->fda.fd[FD_SYNC_RX_TIMER]);
port_clr_tmo(p->fda.fd[FD_DELAY_TIMER]);
1. But there is already timeout event on epoll?
2. Soon epoll event comes, we re-arm the FD_DELAY_TIMER again.
enum fsm_event port_event(struct port *p, int fd_index)
{
...
case FD_DELAY_TIMER:
pr_debug("port %hu: delay timeout",
portnum(p));
port_set_delay_tmo(p);
return port_delay_request(p) ?
EV_FAULT_DETECTED : EV_NONE;
1. Then it repeats.
BR,
Vincent
--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
Re: [Linuxptp-devel] Master still sends out DELAY_REQ
On Mon, Aug 20, 2018 at 03:20:12PM +, Vincent Li X wrote: > Could the reason be like this? > > 1. Event we cancel the FD_DELAY_TIMER? ... > 1. But there is already timeout event on epoll? > 2. Soon epoll event comes, we re-arm the FD_DELAY_TIMER again. This was changed in: commit a8b66ce7f24a3e031712b4352dd48780540018ba Author: Richard Cochran Date: Wed May 16 06:44:56 2018 -0700 Fix crash in hybrid E2E mode. Now the delay timer event is handled before the announce receipt timeout. What version are you running? Thanks, Richard -- Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot ___ Linuxptp-devel mailing list Linuxptp-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
Re: [Linuxptp-devel] [RFC v2 0/5] Adding support for AVnu Automotive Profile
Thanks for the feedback Richard. On Fri, 2018-08-17 at 23:33 -0400, Richard Cochran wrote: > On Thu, Aug 16, 2018 at 10:42:10AM -0700, Vedang Patel wrote: > > > > > > Changes in V2: > > ~~~ > This looks a lot better, thanks! I'll have a few small comments on > the patches... > > > > > Some Opens: > > - Currently, we are using masterOnly (a per-port config option) and > > slaveOnly > > (a global config option) to determine role of the devices. This > > configuration > > option might be a little confusing to a new comer. One idea is to > > change > > slaveOnly to per-port config option. Are there any other ideas > > for this or > > the current way is not as confusing as I think? > Most users only have a single port. Probably automotive will also > mainly be single port devices. Anyone making a BC or TC already has > a > super confusing swamp full of options to wade through. In other > words, I expect such "power users" to know what they are doing. > makes sense. > > > > - In port_p2p_transition(), we are setting up the delay timer when > > BMCA is set > > as ‘noop’. Usually it is initialized then the device transitions > > to > > PS_LISTENING. But, we are skipping the LISTENING state. > These exceptional cases in the series don't seem too bad to me. > > > > > Another alternative > > is to transition to PS_LISTENING and then unconditionally > > transfer to > > PS_MASTER/PS_SLAVE. > What do you mean by "unconditionally"? State transitions are > triggered by events. > Yeah, now that I think of it, I don't think triggering anything unconditionally is going to work. -Vedang > Thanks, > Richard -- Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot ___ Linuxptp-devel mailing list Linuxptp-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
Re: [Linuxptp-devel] [RFC v2 0/5] Adding support for AVnu Automotive Profile
On Fri, 2018-08-17 at 23:50 -0400, Richard Cochran wrote: > On Thu, Aug 16, 2018 at 10:42:10AM -0700, Vedang Patel wrote: > > > > - In port_p2p_transition(), we are setting up the delay timer when > > BMCA is set > > as ‘noop’. Usually it is initialized then the device transitions > > to > > PS_LISTENING. But, we are skipping the LISTENING state. Another > > alternative > > is to transition to PS_LISTENING and then unconditionally > > transfer to > > PS_MASTER/PS_SLAVE. But, that seems more of an hack than what is > > currently > > being done. Any other alternatives? > Maybe we can reset the delay timer unconditionally when entering > SLAVE > or MASTER. It will cause a minor glitch in the delay timing, but > will > it matter? > This will just mean that the interval between path delay calculation might be atmost 2* delay_interval when you calculate delay the first time after transition to master/slave state. I don't any immediate issue with that. Will have to run some tests and see what really happens. Thanks, Vedang > Thanks, > Richard -- Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot ___ Linuxptp-devel mailing list Linuxptp-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
Re: [Linuxptp-devel] [RFC v2 0/5] Adding support for AVnu Automotive Profile
On Sat, 2018-08-18 at 00:03 -0400, Richard Cochran wrote:
> On Thu, Aug 16, 2018 at 10:42:10AM -0700, Vedang Patel wrote:
> >
> > - In port_p2p_transition(), we are setting up the delay timer when
> > BMCA is set
> > as ‘noop’. Usually it is initialized then the device transitions
> > to
> > PS_LISTENING. But, we are skipping the LISTENING state. Another
> > alternative
> > is to transition to PS_LISTENING and then unconditionally
> > transfer to
> > PS_MASTER/PS_SLAVE. But, that seems more of an hack than what is
> > currently
> > being done. Any other alternatives?
> Another idea:
>
> Can you have INITIALIZING -> LISTENING, and arrange for
> EV_ANNOUNCE_RECEIPT_TIMEOUT_EXPIRES to happen right away?
>
> Then your fsm does LISTENING -> SLAVE on the
> EV_ANNOUNCE_RECEIPT_TIMEOUT_EXPIRES event.
>
> (haven't looked at arranging that timeout, just brainstorming...)
>
We can do something on the lines of the following (completely untested)
snippet:
diff --git a/port.c b/port.c
index f4a9fbb5a8d7..214a8329d27e 100644
--- a/port.c
+++ b/port.c
@@ -1623,7 +1623,7 @@ int port_initialize(struct port *p)
p->fda.fd[FD_FIRST_TIMER + i] = fd[i];
}
- if (port_set_announce_tmo(p)) {
+ if (set_tmo_log(p->fda.fd[FD_ANNOUNCE_TIMER], 1, -10)) {
goto no_tmo;
}
if (unicast_client_enabled(p) && unicast_client_set_tmo(p)) {
This way, the timer will fire once (immediately after it's setup) even
if inhibit_announce is set.
But, there is going to be a likely effect on the normal working of
linuxptp. We can fix this by only doing this when BMCA == noop. Again,
will have to actually test this to see what issues might pop up.
Thanks,
Vedang
> Thanks,
> Richard
>
--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
Re: [Linuxptp-devel] [RFC v2 2/5] Add BMCA config option.
On Sat, 2018-08-18 at 00:07 -0400, Richard Cochran wrote:
> On Thu, Aug 16, 2018 at 10:42:12AM -0700, Vedang Patel wrote:
> >
> > diff --git a/fsm.c b/fsm.c
> > index ce6efad30937..a0c98891e39f 100644
> > --- a/fsm.c
> > +++ b/fsm.c
> > @@ -335,3 +335,103 @@ enum port_state ptp_slave_fsm(enum port_state
> > state, enum fsm_event event,
> >
> > return next;
> > }
> > +
> > +enum port_state designated_master_fsm(enum port_state state,
> > + enum fsm_event event,
> > + int mdiff)
> > +{
> ...
> >
> > +}
> > +
> > +enum port_state designated_slave_fsm(enum port_state state,
> > + enum fsm_event event,
> > + int mdiff)
> > +{
> ...
> >
> > +}
> Can these two go into their own, new file?
>
Yeah will move them out and name the file designated_fsm.c/.h.
> >
> > @@ -2853,10 +2868,24 @@ struct port *port_open(int phc_index,
> > goto err_port;
> > }
> >
> > - p->state_machine = clock_slave_only(clock) ? ptp_slave_fsm
> > : ptp_fsm;
> > p->phc_index = phc_index;
> > p->jbod = config_get_int(cfg, interface->name,
> > "boundary_clock_jbod");
> > transport = config_get_int(cfg, interface->name,
> > "network_transport");
> > + p->master_only = config_get_int(cfg, p->name,
> > "masterOnly");
> > + p->bmca = config_get_int(cfg, p->name, "BMCA");
> > +
> > + if (p->bmca == BMCA_NOOP && transport != TRANS_UDS) {
> > + if (p->master_only) {
> > + p->state_machine = designated_master_fsm;
> > + } else if (clock_slave_only(clock)) {
> > + p->state_machine = designated_slave_fsm;
> > + } else {
> > + pr_err("Please enable atleast one of
> > masterOnly or slaveOnly when BMCA == noop.\n");
> "at least" ---^
>
Will fix this.
Thanks,
Vedang
> Thanks,
> Richard
--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
Re: [Linuxptp-devel] [RFC v2 4/5] Add ignore_source_id config option.
On Sat, 2018-08-18 at 00:11 -0400, Richard Cochran wrote:
> On Thu, Aug 16, 2018 at 10:42:14AM -0700, Vedang Patel wrote:
>
> >
> > +static int check_source_identity(struct port *p, struct
> > ptp_message *m)
> > +{
> > + struct PortIdentity master;
> > +
> > + if (!p->ignore_source_id) {
> > + master = clock_parent_identity(p->clock);
> > + if (!pid_eq(&master, &m-
> > >header.sourcePortIdentity)) {
> > + return -1;
> > + }
> > + }
> > + return 0;
> > +}
> I prefer an "early out" pattern, like:
>
> if (p->ignore_source_id) {
> return 0;
> }
> master = clock_parent_identity(p->clock);
> if (!pid_eq(&master, &m->header.sourcePortIdentity)) {
> return -1;
> }
> return 0;
>
> or even:
>
> if (p->ignore_source_id) {
> return 0;
> }
> master = clock_parent_identity(p->clock);
> return pid_eq(&master, &m->header.sourcePortIdentity) ? 0 : -1;
>
Will include the second version in v3.
Thanks,
Vedang
> Thanks,
> Richard
>
>
--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
Re: [Linuxptp-devel] [RFC v2 5/5] Add example configuration for AVnu Automotive Profile.
On Sat, 2018-08-18 at 00:22 -0400, Richard Cochran wrote:
> On Thu, Aug 16, 2018 at 10:42:15AM -0700, Vedang Patel wrote:
> >
> > create mode 100644 configs/AVnu-master.cfg
> > create mode 100644 configs/AVnu-slave.cfg
> IIRC, AVnu had some earlier best practice paper about using 802.1-AS.
> So maybe these file names should include the word "automotive" in
> them.
>
> Actually, considering this
>
> https://trademarks.justia.com/778/57/avnu-77857554.html
>
> we should avoid the word AVnu altogether.
>
Yeah. Renaming files to automotive-{master,slave}.cfg.
Also, do you want me to remove AVnu from the whole series code/commit
message or just the config file names?
Thanks,
Vedang
> Thanks,
> Richard
--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
Re: [Linuxptp-devel] [RFC v2 5/5] Add example configuration for AVnu Automotive Profile.
On Tue, Aug 21, 2018 at 12:51:18AM +, Patel, Vedang wrote: > Also, do you want me to remove AVnu from the whole series code/commit > message or just the config file names? I would just remove the word "AVnu" everywhere and just say, the automotive profile. I want to avoid making the impression that AVnu somehow sponsors or endorsesour implementation. You can keep the link to the PDF in the cover letter. Thanks, Richard -- Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot ___ Linuxptp-devel mailing list Linuxptp-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
