e as more events are added but it'd
> keep syntax backwards-compatible for scripts.
Good idea. The PARENT_DATA_SET notification is new too (added after
4.1) and I guess there is no reason to require even the original two
notifications, so change that to "cnt < 1" wit
On Mon, Dec 04, 2023 at 04:06:31PM +0100, Miroslav Lichvar wrote:
> On Sat, Dec 02, 2023 at 03:38:18PM -0800, Richard Cochran wrote:
> > --- a/config.c
> > +++ b/config.c
> > @@ -171,6 +171,7 @@ static struct config_enum delay_filter_enu[] = {
> >
> > static
PORT_ITEM_STR("cmlds.client_address", "/var/run/cmlds_cleint"),
Swapped e and i in "cleint" is not intentional, right? It's the same
in the man page and the example default config.
Other than that, the patchset looks ok to me. I didn't run any tests,
just a quick loo
On Tue, Nov 28, 2023 at 03:34:54PM -0800, Richard Cochran wrote:
> On Tue, Nov 28, 2023 at 12:53:38PM +0100, Miroslav Lichvar wrote:
> > However, timestamping only sync messages has an advantage with very
> > large number of clients as they don't have timestamp each other's
&g
ent messages is very rare, ok,
maybe it's not worth the trouble.
However, timestamping only sync messages has an advantage with very
large number of clients as they don't have timestamp each other's
delay requets and timestamping of sync messages is more reliable.
--
Miro
On Tue, Nov 21, 2023 at 08:52:30PM -0800, Richard Cochran wrote:
> linuxptp-testsuite test 20-pmc now fails with
>
> "SUBSCRIBE_EVENTS_NP SET needs 4 values"
>
> Can you provide an update?
Done. Thanks.
--
Miroslav Lichvar
_
the hardware supports the type-specific filters, I think ptp4l
should be able to set it.
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
If the string specified by the message_tag option contains "{level}",
replace it with the log level of the message as a number.
This allows users to filter printed log messages by their level.
Signed-off-by: Miroslav Lichvar
---
phc2sys.8 | 4 +++-
prin
_XXX_SYNC or
> HWTSTAMP_FILTER_PTP_XXX_DELAY_REQ rx filters for all transport levels.
With this dynamic switching of the RX filter on port state changes is
there a reason to not do it by default? Or instead of adding a new
option add it as a new enum to the existing hwts_filter option?
--
Miro
to 30
second subscription duration) to avoid unnecessarily frequent renewals
with the default update intervals of phc2sys and ts2phc, but still be
reasonably fast in detecting missing responses.
Signed-off-by: Miroslav Lichvar
---
phc2sys.c | 2 +-
pmc_agent.c | 24
Add a new field to the domain struct to mark the domain which contains
the system realtime clock instead of relying on the fact that the domain
doesn't have a subscribed pmc agent.
Signed-off-by: Miroslav Lichvar
---
phc2sys.c | 8
1 file changed, 4 insertions(+), 4 deletions(-)
diff
with ptp4l and make it evident,
if the pmc subscription runs out, log an error message and stop
synchronization in the domain until the pmc agent is subscribed again.
Signed-off-by: Miroslav Lichvar
---
phc2sys.c | 20 +---
1 file changed, 17 insertions(+), 3 deletions(-)
diff --git
of phc2sys to handle
unresponsive ptp4l, e.g. stopped, crashed or accidentaly removed socket.
The first three patches prepare the code for the actual handling
implemented in the fifth patch.
Miroslav Lichvar (4):
pmc_agent: Add function to check if still subscribed.
pmc_agent: Make update
Add function to compare the current time with the time of the last
successful pmc update to determine if ptp4l is still responding.
Signed-off-by: Miroslav Lichvar
---
pmc_agent.c | 15 +++
pmc_agent.h | 8
2 files changed, 23 insertions(+)
diff --git a/pmc_agent.c b
On Thu, Nov 09, 2023 at 10:54:31PM -0800, Richard Cochran wrote:
> On Thu, Oct 26, 2023 at 02:40:11PM +0200, Miroslav Lichvar wrote:
>
> > @@ -942,8 +957,8 @@ static int auto_init_ports(struct domain *domain)
> > }
> >
> > err = pmc_agent
Add a new field to the domain struct to mark the domain which contains
the system realtime clock instead of relying on the fact that the domain
doesn't have a subscribed pmc agent.
Signed-off-by: Miroslav Lichvar
---
phc2sys.c | 8
1 file changed, 4 insertions(+), 4 deletions(-)
diff
The pmc_agent_update() function is called once per clock update. If the
clock update rate configured with -R is smaller than 1 per minute,
adjust the subscription interval to avoid gaps in its coverage.
Signed-off-by: Miroslav Lichvar
---
phc2sys.c | 4 +++-
1 file changed, 3 insertions(+), 1
Add function to compare the current time with the time of the last
successful pmc update to determine if ptp4l is still responding.
Signed-off-by: Miroslav Lichvar
---
pmc_agent.c | 15 +++
pmc_agent.h | 8
2 files changed, 23 insertions(+)
diff --git a/pmc_agent.c b
Replace the constant 1-minute update interval and 3-minute subscription
duration with a duration parameter of pmc_agent_subscribe().
Signed-off-by: Miroslav Lichvar
---
phc2sys.c | 2 +-
pmc_agent.c | 15 +--
pmc_agent.h | 4 +++-
ts2phc.c| 2 +-
4 files changed, 10
This series improves the automatic mode of phc2sys to handle
unresponsive ptp4l, e.g. stopped, crashed or accidentaly removed socket.
The first four patches prepare the code for the actual handling
implemented in the fifth patch.
Miroslav Lichvar (5):
pmc_agent: Make subscription duration
with ptp4l and make it evident,
if the pmc subscription runs out, log an error message and stop
synchronization in the domain until the pmc agent is subscribed again.
Also shorten the minimum subscription interval from 180 seconds to 30
seconds to react faster.
Signed-off-by: Miroslav Lichvar
tmo.it_value.tv_nsec = ns % NS_PER_SEC;
The fix looks good to me. Replacing the repeated subtraction with
division also fixes a potential slowdown with large scales.
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
to disable it in the default log
level.
Signed-off-by: Miroslav Lichvar
---
clock.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/clock.c b/clock.c
index 5a64613..74911fd 100644
--- a/clock.c
+++ b/clock.c
@@ -1585,7 +1585,7 @@ static void clock_forward_mgmt_msg(struc
The domain number used for communication with ptp4l specified by the -n
option is ignored in the non-automatic mode (-w option). Set the domain
number to the last specified value.
Fixes: 417de97d098b ("phc2sys: Add multi-domain synchronization.")
Signed-off-by: Miroslav Lichvar
---
with -w. I'll submit a
patch. Sorry for the bugs.
Thanks,
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
t;(uds_remote_cnt > 0)"
to follow the coding style.
Will you submit it as a git-formatted patch with subject and commit
message, or would you prefer me to do it?
Thanks,
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.
ronization to match the behavior of phc2sys before it supported
multiple domains.
Signed-off-by: Miroslav Lichvar
---
phc2sys.c | 18 +++---
1 file changed, 15 insertions(+), 3 deletions(-)
diff --git a/phc2sys.c b/phc2sys.c
index 2750691..313cdcf 100644
--- a/phc2sys.c
+++ b/phc2sys.c
r driver is broken. It might need to decrease the frequency
limit. The only reason your change "fixes" the problem is that it
resets the frequency to a non-buggy range.
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel
On Mon, Aug 21, 2023 at 07:02:55PM +0200, Maciek Machnikowski wrote:
> On 8/21/2023 2:04 PM, Miroslav Lichvar wrote:
> > s->drift is the drift of the uncompensated clock. It's not supposed to be
> > scaled by anything.
> If that's the intended use - then the line 147:
be to adjust the timestamps from E2E and recalculate the delay before
stepping the clock.
> This patch fixes the s->drift when entering the stable state to hold
> the scaled drift.
That looks wrong to me and in my test with a clock which has a large
frequency e
ts2phc instead of the slave event monitor.
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
. But since ptp4l is free running mode,
> pmc_agent_query_port_properties will not give the phc_index.
> Ptp4l doesn’t know about PHC since it is not disciplining it.
Well, to me that would seem like the right thing to do. The PHC is not
synchronized, so phc2sys shouldn't be using it for synchroniz
describe the use case? It's not clear to me why should
be a free-running PHC adjusted or or used as a source clock.
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo
Keep configuration settings in a separate domain and copy them to the
actual domain just before creating the pmc agent. This will be needed to
share the settings between multiple domains.
Signed-off-by: Miroslav Lichvar
---
phc2sys.c | 43 ---
1 file
modes only one domain is created for all clocks.
Signed-off-by: Miroslav Lichvar
---
phc2sys.8 | 12 +-
phc2sys.c | 337 +-
2 files changed, 241 insertions(+), 108 deletions(-)
diff --git a/phc2sys.8 b/phc2sys.8
index 6c6c7e8..9a37778 100644
ble a
fallback on failures. The first three patches refactor the code to
minimize the fourth patch adding the actual support.
Miroslav Lichvar (4):
phc2sys: Rename phc2sys_private to domain.
phc2sys: Create pmc agent after processing options.
phc2sys: Shallow do_loop().
phc2sys: Add multi-dom
Rename the phc2sys_private structure to domain in preparation for
support of multiple domains.
While touching the lines of code, also rename master to source clock.
Signed-off-by: Miroslav Lichvar
---
phc2sys.c | 326 +++---
1 file changed, 163
Move the measurements and updates of clocks from do_loop() to separate
function.
Signed-off-by: Miroslav Lichvar
---
phc2sys.c | 100 +-
1 file changed, 54 insertions(+), 46 deletions(-)
diff --git a/phc2sys.c b/phc2sys.c
index d0f5adb
stats can be combined over any number of intervals as you
like.
If that is not acceptable for some reason, I'd prefer to add a new
option to not reset the stats after printing instead of duplicating
the functionality in a different place and printing the data in a
different format.
--
Mirosla
ine servo performance.
There is already the summary_interval option to print the offset
statistics. Is that not sufficient?
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo
Add a new notification event for pmc clients to receive
PARENT_DATA_SET when a change is made there.
Signed-off-by: Miroslav Lichvar
---
clock.c| 26 +++---
notification.h | 1 +
pmc.c | 10 ++
pmc_common.c | 15 +++
4 files changed
ng new
features and the rest focused on testing and stabilization?
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
In the free-running mode the stats are updated only once per the
frequency estimation interval instead of once per sync message. The
stats max_count calculation didn't account for that, which caused a
reduced rate of printed summaries. Fix the calculation for this case.
Signed-off-by: Miroslav
t value for one field, and so the test
> suite would need the following change.
The test is now updated. Thanks.
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
or add a compatible mechanism in chronyd. Or again the user
> would have to keep track of the shutdown order.
Or make the cleanup part configurable (disabled by default)?
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-
s using HW timestamping (e.g.
chronyd).
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
, while with the current solution the interval can
gradually increase with the load and stay within the limits under
worse conditions.
Instead of trying to make the interval constant I propose add extra
randomness, maybe +/- 1%. This seems to be well within the spec.
--
Miroslav Lichvar
__
e together
with the descriptor index to make it easier to debug issues like this in
future.
Signed-off-by: Miroslav Lichvar
---
clock.c | 6 --
sk.c| 14 ++
sk.h| 7 +++
3 files changed, 25 insertions(+), 2 deletions(-)
diff --git a/clock.c b/clock.c
index 75d7c4
a test report not having a round number.
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
n to collide in their transmissions times (which requires
queueing and potentially impacts timestamp accuracy), this is more
likely to repeat frequently. If anything, I'd expect more randomness
to be added to the interval.
--
Miroslav Lichvar
___
Li
linuxptp-devel@lists.sourceforge.net/msg06420.html
Thanks,
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
l":
$ git grep 'header.control'
msg.c:// m->header.control,
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
might be accepted by the
consumer.
> To prevent such behavior - invalidate the SHM data when releasing the
> SHM servo.
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net
rything is
easier.
If you reword the commit message, I'm ok with it.
Thanks,
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
n of the system clock.
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
is expected
with SW timestamping and in the free-running mode.
Fixes: afeabf3c90ed ("ptp4l: add VLAN over bond support")
Signed-off-by: Miroslav Lichvar
---
port.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/port.c b/port.c
index 3453716..9746df8 100644
--- a/port.c
+++ b/port.c
settings
- split into multiple patches
- ...
Signed-off-by: Miroslav Lichvar
---
phc2sys.c | 594 --
1 file changed, 349 insertions(+), 245 deletions(-)
diff --git a/phc2sys.c b/phc2sys.c
index 19e8012..46499a2 100644
--- a/phc2sys.c
+++ b
On Mon, Feb 20, 2023 at 12:57:48PM -0800, Richard Cochran wrote:
> ChangeLog
> ~
> v6:
> Add new configuration options into default.cfg and man page
> Make pmc SET command ready for 2106 (Miroslav)
> Use SPDX tag for new file tz.h
Looks good to me.
Thanks,
--
atop.timeOfNextJump.seconds_lsb = jump & 0x;
+ atop.timeOfNextJump.seconds_msb = jump >> 32;
pmc_send_set_action(pmc, code, , sizeof(atop));
break;
case MID_GRANDMASTER_SETTINGS_NP:
The rest of the patchset looks good to me.
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
_MAX),
> + PORT_ITEM_INT("power_profile.2011.networkTimeInaccuracy", 0, 0,
> INT_MAX),
> + PORT_ITEM_INT("power_profile.2017.totalTimeInaccuracy", 0x, 0,
> INT_MAX),
> + PORT_ITEM_INT("power_profile.grandmasterID", 0, 0, 0
The lsb field of timeOfNextJump is unsigned, so it will overflow in
year 2106 (in the TAI timescale).
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
On Sun, Feb 12, 2023 at 07:40:36PM -0800, Richard Cochran wrote:
> > > On Thu, Feb 02, 2023 at 11:52:09AM +0100, Miroslav Lichvar wrote:
> > > > It would be nice to have the new options documented in the ptp4l man
> > > > page. Please consider renaming tzto
On Sat, Feb 11, 2023 at 07:51:27AM -0800, Richard Cochran wrote:
> On Wed, Feb 01, 2023 at 04:52:15PM +0100, Miroslav Lichvar wrote:
>
> > > + if (next) {
> > > + atop.jumpSeconds = next->local_tai_offset -
> > > tz->local_tai_offset;
> >
Deny client requests and ignore server responses that have
logInterMessagePeriod outside of [-30..30] to avoid undefined
integer shifts in calculation of the interval.
Signed-off-by: Miroslav Lichvar
---
unicast_client.c | 6 ++
unicast_service.c | 4
2 files changed, 10 insertions
If chronyd is selected as the NTP program, use the SOCK refclock instead
of SHM to improve security and reduce delay in receiving of samples.
Signed-off-by: Miroslav Lichvar
---
timemaster.8 | 10 -
timemaster.c | 58 +++-
2 files changed
as a more secure alternative to the NTP SHM, and
updates timemaster to use it.
Miroslav Lichvar (4):
Remove obsolete statement in ptp4l man page.
Add refclock_sock servo.
timemaster: Replace shm_segment with refclock_id.
timemaster: Use refclock_sock servo with chrony.
config.c
with restricted access, while the
shared memory segment (using a predictable key) can be created by
untrusted users or applications if they can run before ptp4l/phc2sys and
chronyd/ntpd, similarly to the issue with binding to an unprivileged
TCP/UDP port.
Signed-off-by: Miroslav Lichvar
---
config.c
The NTP SHM number is no longer the PTP domain number. It was made
configurable and the default value is 0.
Fixes: 3760f8b6537e ("Add option to set NTP SHM segment number.")
Signed-off-by: Miroslav Lichvar
---
ptp4l.8 | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --g
Replace the shm_segment number with a more general refclock_id and shift
by first_shm_index only in the SHM-specific context to allow other types
of refclocks to be used.
Signed-off-by: Miroslav Lichvar
---
timemaster.c | 48 +++-
1 file changed, 27
een with valgrind and address
sanitizer.
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
.jumpSeconds = next->local_tai_offset -
> tz->local_tai_offset;
> + atop.timeOfNextJump.seconds_lsb = next->timestamp;
> + }
Is this intentionally not setting the _msb field, ignoring distant future?
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Replace the shm_segment number with a more general refclock_id and shift
by first_shm_index only in the SHM-specific context to allow other types
of refclocks to be used.
Signed-off-by: Miroslav Lichvar
---
timemaster.c | 48 +++-
1 file changed, 27
with restricted access, while the
shared memory segment (using a predictable key) can be created by
untrusted users or applications if they can run before ptp4l/phc2sys and
chronyd/ntpd, similarly to the issue with binding to an unprivileged
TCP/UDP port.
Signed-off-by: Miroslav Lichvar
---
config.c
v2:
- renamed servo to "refclock_sock"
This patchset fixes a bug in man page, adds a new servo for the chrony
SOCK refclock protocol as a more secure alternative to the NTP SHM, and
updates timemaster to use it.
Miroslav Lichvar (4):
Remove obsolete statement in ptp4l man p
If chronyd is selected as the NTP program, use the SOCK refclock instead
of SHM to improve security and reduce delay in receiving of samples.
Signed-off-by: Miroslav Lichvar
---
timemaster.8 | 10 -
timemaster.c | 58 +++-
2 files changed
The NTP SHM number is no longer the PTP domain number. It was made
configurable and the default value is 0.
Fixes: 3760f8b6537e ("Add option to set NTP SHM segment number.")
Signed-off-by: Miroslav Lichvar
---
ptp4l.8 | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --g
me(clkid, ) < 0)
> + if (clock_adjtime(clkid, ) < 0) {
> pr_err("failed to adjust the clock: %m");
> + return -1;
> + }
Why not follow clockadj_get_freq() and simply exit here instead of
trying to recover?
What PHC/driver is expected
ULT_DETECTED.
There was a patch submitted for this issue and a similar issue with
vclocks earlier: "port: Avoid faults with vclocks and PHC from command
line."
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists
with the recovery. I completely missed that.
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
es) cause it to
> repeatedly re-arm itself, potentially never clearing.
Do you please provide an example where this makes a difference?
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/li
VLAN over bond support")
Signed-off-by: Miroslav Lichvar
---
port.c | 25 +++--
1 file changed, 11 insertions(+), 14 deletions(-)
diff --git a/port.c b/port.c
index 756ff3a..c88f1be 100644
--- a/port.c
+++ b/port.c
@@ -2685,10 +2685,13 @@ static void port_chang
e. Thanks for submitting the patch.
I think this patch conflicts with the clockcheck patchset. I'll update
it and resend if this one is accepted earlier.
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.
o make an incompatible change, we can change
the magic number.
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
snprintf(buf, sizeof(buf), "%s.%s", section, name);
| ^~~~~~
Signed-off-by: Miroslav Lichvar
---
config.c | 9 +++--
1 file changed, 7 insertions(+), 2 deletions(-)
diff --git a/config.c b/config.c
index e454c91..e21cde7
er issues with the patches,
I can resend.
> The implementation seems fine but its using an interface that was defined by
> chrony. I suppose another application could implement the same interface
> though..
ntpsec might be interested in implementing it. We'll see.
--
Replace the shm_segment number with a more general refclock_id and shift
by first_shm_index only in the SHM-specific context to allow other types
of refclocks to be used.
Signed-off-by: Miroslav Lichvar
---
timemaster.c | 48 +++-
1 file changed, 27
This patchset fixes a bug in man page, adds a new servo for the chrony
SOCK refclock protocol as a more secure alternative to the NTP SHM
(the commit message describes the issue), and updates timemaster to
use it.
Miroslav Lichvar (4):
Remove obsolete statement in ptp4l man page.
Add sock
if
it doesn't exist yet, possibly under a non-root owner, there is no
authentication of messages, and the protocol cannot be restarted if one
side decides to remove and recreate the segment.
Signed-off-by: Miroslav Lichvar
---
chronysock.c| 163
The NTP SHM number is no longer the PTP domain number. It was made
configurable and the default value is 0.
Fixes: 3760f8b6537e ("Add option to set NTP SHM segment number.")
Signed-off-by: Miroslav Lichvar
---
ptp4l.8 | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --g
If chronyd is selected as the NTP program, use the SOCK refclock instead
of SHM to improve security and reduce delay in receiving of samples.
Signed-off-by: Miroslav Lichvar
---
timemaster.8 | 10 -
timemaster.c | 57 +++-
2 files changed
ation.
Nice feature. In a quick test, it printed the state correctly after
setting it with "testptp -L".
Do you plan to add also pin_set?
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists
, 1, UINT8_MAX),
> PORT_ITEM_INT("G.8275.portDS.localPriority", 128, 1, UINT8_MAX),
> GLOB_ITEM_INT("gmCapable", 1, 0, 1),
Looks good to me.
Thanks,
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
Exit if an error is returned from the clock_adjtime() call in
clockadj_get_freq(). No recoverable errors are expected.
Signed-off-by: Miroslav Lichvar
---
clockadj.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/clockadj.c b/clockadj.c
index 957dc57..4c920b9 100644
--- a/clockadj.c
step the clock.
Signed-off-by: Miroslav Lichvar
---
clock.c | 3 +++
clockcheck.c | 10 ++
clockcheck.h | 8
phc2sys.c| 3 +++
ptp4l.8 | 6 --
5 files changed, 28 insertions(+), 2 deletions(-)
diff --git a/clock.c b/clock.c
index 46ac9c2..8177e77 100644
to consider the last frequency set by
ptp4l/phc2sys in order to detect cases where multiple processes are
trying to control the same clock due to a misconfiguration or bug.
Miroslav Lichvar (4):
phc2sys: Add clocks after processing configuration.
Drop support for old kernels returning zero
from command line and
add them only after the configuration is complete.
This also fixes the automatic mode to not add CLOCK_REALTIME twice.
Fixes: 33ac7d25cd92 ("phc2sys: Allow multiple sink clocks")
Signed-off-by: Miroslav Lichvar
---
phc2sys.8 | 2 +-
phc2
Kernels before 3.10 had a bug in reading of the system clock frequency,
which was worked around by commit da347d7a36f2 ("ptp4l: Set clock
frequency on start").
Drop this workaround and support for the old kernels to make
clockadj_get_freq() useful.
Signed-off-by: Miroslav Lichvar
--
sanity check is disabled. The default is 2 (20%).
If not, what would you suggest?
Thanks,
--
Miroslav Lichvar
___
Linuxptp-devel mailing list
Linuxptp-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxptp-devel
Exit if an error is returned from the clock_adjtime() call in
clockadj_get_freq(). No recoverable errors are expected.
Signed-off-by: Miroslav Lichvar
---
clockadj.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/clockadj.c b/clockadj.c
index 957dc57..4c920b9 100644
--- a/clockadj.c
set by
ptp4l/phc2sys in order to detect cases where multiple processes are
trying to control the same clock due to a misconfiguration or bug.
Miroslav Lichvar (4):
phc2sys: Add clocks after processing configuration.
Drop support for old kernels returning zero frequency.
Don't accept errors
1 - 100 of 577 matches
Mail list logo
