| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| ... | |
| | | | | |
|
| | | | | |
|
| | | | | |
|
| | | | |
| | | |
| | | |
| | | |
| | | | |
End goal here is to remove all printf() and fprintf() calls from gPTP
source.
|
| |/ / /
| | |
| | |
| | | |
Issue #321
|
| |\ \ \
| | | |
| | | | |
Task gptp v2
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
There are probably other places that the info should be updated.
This is just the first approach to get that information in place.
Later it needs to be implemented the phaseChange calculation and
call the method that updates the TBI in other necessary places as
well
|
| | | | | |
|
| | | | |\
| | |/ /
| |/| | |
|
| | | | |\
| | | |/
| | |/| |
|
| | | |\ \
| | | | |
| | | | | |
igb_avb: sync kcompat.[ch] with upstream
|
| | | |/ /
| | | |
| | | |
| | | |
| | | |
| | | | |
Sync kernel compatibility code for older kernel versions with upstream.
Signed-off-by: Todd Fujinaka <todd.fujinaka@intel.com>
|
| | | |\ \
| | | | |
| | | | | |
daemon_cl: Fixes PPS out-of-phase issue.
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
From I210 datasheet section
7.8.3.3.3 "Synchronized Output Clock on SDP Pins":
The clock out drives initially a logical '0' level on the selected SDP.
If the TSAUXC.ST0/1 flag is cleared, it happens instantly when setting the
TSAUXC.EN_CLK0/1 bit.
Otherwise it happens when the SYSTIM is >= the TRGTTIM.
Before the fix, when stopping PPS, only the SDP pin is disabled.
Therefore when starting PPS again, as soon as the SDP pin is enabled,
the clock output starts.
This fix will correctly clear the TSAUXC.ST0/1 and TSAUXC.EN_CLK0/1 bits
when stopping PPS. This will ensure that the clock output only happen
when the SYSTIM is >= the TRGTTIM.
Signed-off-by: Roland Hii <roland.king.guan.hii@intel.com>
|
| | | |\ \ \
| | | |/ /
| | |/| | |
Fixes: Wrong output from log messages on 64bit machines.
|
| | | |/ /
| | | |
| | | |
| | | |
| | | |
| | | | |
Details: Used macros for print formating eg. PRIu64 or PRIu32 so the expected
and supplied formats match. Also changed format specifiers when they weren't matching
expected format: eg. "llu" to "lu".
|
| | | |\ \
| | | | |
| | | | | |
New: VLAN ID can be configured for no endpoint builds.
|
| | | | | | |
|
| | |_|/ /
|/| | |
| | | |
| | | |
| | | | |
Details: You can specify VLAN Identifier in talker ini file.
Syntax is vlan_id = 5; Default value is 2 for both class A and class B streams.
|
| | | |\ \
| |/ / /
|/| | | |
AVTP pipeline improvements
|
| | | | | |
|
| | | | |
| | | |
| | | |
| | | |
| | | | |
Details: Added checking if payload len from header is not bigger
than actually received frame.
|
| | |/ /
|/| |
| | |
| | |
| | |
| | |
| | | |
Rawsock Constructors fill virtual function tables.
In rawsock_impl.c just proper function from virtual table is called,
no additional checks are made, no fallback to default implementation.
|
| | | | |
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
The same idea used inside the avtp_pipeline was used here.
We are now adding the possibility to use a configuration (ini) file
to configure the internal variables of gptp.
Currently the following was added:
- 3rd party ini file parser;
- Class with callback to the parser
- some options were added as an example to the gptp_cfg.ini file
- all the current options are parsed inside linux/daemon_cl.cpp and printed out during init.
Still missing:
- use the options from the file inside gPTP
Also, the linux daemon_cl.cpp was indented since there were several indenting problems
on the file.
|
| | | | |
|
| | | | |
|
| | |\ \
| | |/
| | |
| | |
| | |
| | |
| | | |
Conflicts:
daemons/gptp/common/ieee1588.hpp
daemons/gptp/common/ieee1588port.cpp
daemons/gptp/linux/src/linux_hal_generic.cpp
|
| | | |\
| |_|/
|/| | |
|
| |\ \ \
| | | |
| | | | |
openavb-next: STC is now Harman so update the logo.
|
| |/ / / |
|
| |\ \ \
| | | |
| | | | |
Srinath88 gptp enhancements
|
| | |\ \ \
|/ / / /
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
into srinath88-gptp-enhancements
Conflicts:
examples/jackd-talker/jackd_talker.c
examples/simple_talker/simple_talker.c
|
| | | | |
| | | |
| | | |
| | | | |
listener_mrp_client.h
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
values. It defaults to 1G phy delay values of i20 card if there is no user input.
The code now uses user input for tx and rx phy delays instead of the TX_PHY_TIME and RX_PHY_TIME macros
|
| | | | |
| | | |
| | | |
| | | | |
to all the functions that need it in simple_talker and simple_listener.
|
| | | | |
| | | |
| | | |
| | | | |
clinet to a struct.
|
| | | | |
| | | |
| | | |
| | | | |
destination mac address from mrp packet instead of hardcoding it
|
| | | | |
| | | |
| | | |
| | | | |
instead of hardcoding it.
|
| | | | |
| | | |
| | | |
| | | | |
and VLAN id instead of using hardcoded values.
|
| | | | |\
| |_|_|/
|/| | | |
|
| |\ \ \ \
| | | | |
| | | | | |
Feature avtp pipeline
|
| | |\ \ \ \
| | | | | |
| | | | | | |
AVTP pipeline documentation improvements
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | | |
Change-Id: I858c2ac06165f888bb91458cbb769cb08ba86397
|
| | |/ / / /
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Added gst pipelines for gstreamer 1.0.
Change-Id: I12ca124e363f515c60c14a77c5040daede3726b2
|
| | |\ \ \ \
| | | | | |
| | | | | | |
Create pool of tx buffers from multiple dma pages.
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | | |
Added waiting for available tx buffer. Added out of tx buffers reporting.
|
| | |/ / / / |
|
| | |\ \ \ \
| | | | | |
| | | | | | |
AVTP pipeline is using gptplocaltime() function
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
RevBy: Marcin Miklas
Details: Changed the message printed when rtp payloader
cannot be found by its name in the pipeline.
Change-Id: I2318953e2059c9e2623d389810fb2ed4b1c1a26e
|
