Introduce LWM2M subsystem for Zephyr #712

mike-scott · 2017-07-07T09:37:47Z

Hello all,

I'm submitting this series as RFC to start discussion / early review process to get an LWM2M implementation merged into Zephyr mainline.

A bit of history:
After doing quite a bit of research to select a good starting point for a Zephyr-based LWM2M stack, I selected the following source: https://github.com/sics-iot/lwm2m-contiki/tree/lwm2m-standalone-dtls/apps/oma-lwm2m
Created by: Joakim Eriksson, Niclas Finne and Joel Hoglund as a stand-alone and/or Contiki based solution. The stack they've put together is very robust and looked to be a good candidate.

Here's a small summary of my changes from the original sources:

[engine] The engine portion was re-written due to the heavy reliance on ER-CoAP APIs which are not compatible to the Zephyr CoAP APIs and the other items below marked with [engine].
[engine] All LWM2M/IPSO object data is now abstracted into resource data which stores information like the data type, length, some callbacks. The engine modifies this data directly (or makes callbacks) instead of all of the logic for this living in each object's code. (This wasn't scaling well as I was implementing changes).
[engine] Related to the above change, I also added a generic set of getter/setter functions that user applications can call to change the object data instead of having to add getter/setting methods in each object.
[engine] The original sources shared the engine's context structure quite extensively causing a problem with portability. I broke up the context into it's individual parts: LWM2M path data, input data and output data and pass only the needed data into each set of APIs (for instance reader / writers only get path data and the related input or output structure).
[sample] The sample/net/lwm2m_client uses the net app API which will be utilized more in the subsys source in the future
DTLS / Bootstrap was not included and will be added in a later patch
Float32/Float64 input / output handling functions will be implemented in a later patch
JSON output handling probably needs to be debugged more. (TLV is the primary delivery format)
More IPSO Alliance objects will be implemented as we get further into review
The Temperature Sensor will be finished as review progresses. Currently, w/o a Float32 implementation it will return empty values.
CoAP URI parser functions needs to be added and instead of the sample using CONFIG_NET_APP_PEER_IPV6_ADDR / CONFIG_NET_APP_PEER_IPV4_ADDR values directly. It will assign these as a coap:// URI to the security object 0 index. Once the LWM2M client is started, it will parse the values out of index 0 and make the correct connections (port, etc) -- including possible DNS lookup.
Block transfer is only handled in select situations. A future patch to the engine will add a tracking mechanism by which larger amounts of data can be queued up for sending and/or delivered to user applications in piecemeal when receiving.

The sample for LWM2M can be run via QEMU by following the instructions here:
https://github.com/zephyrproject-rtos/zephyr/blob/master/doc/subsystems/networking/qemu_setup.rst

(Quick Instructions)
To start with you will want to use Leshan LWM2M demo server:
[terminal #1]
$ wget https://hudson.eclipse.org/leshan/job/leshan/lastSuccessfulBuild/artifact/leshan-server-demo.jar
$ java -jar ./leshan-server-demo.jar
(You can now open a web browser to http://localhost:8080/ to watch connections)

Next, checkout https://github.com/zephyrproject-rtos/net-tools and follow the build instructions referenced above. After they are built, start up the slip tools:
[terminal #2]
$ ./loop-socat.sh
[terminal #3]
$ sudo ./loop-slip-tap.sh

Now you can cd to the Zephyr sources and build the sample for QEMU target:
$ make -C samples/net/lwm2m_client qemu

You should see the sample app connect to Leshan and register twice (IPv4 connection and IPv6 connection). Feel free to navigate around Leshan and play with the QEMU LWM2M client.

I look forward to your comments.

tbursztyka

legacy code is sometime quite hard to read due to numerous if/for/while/else/... imbrication. I don't know if the plan is to backport future patches from legacy code to this port, if not maybe it will be nice to rearrange the code for better readability.

Could be done later though.

tbursztyka · 2017-07-07T09:44:09Z

subsys/net/lib/zoap/zoap.c

move it one line down, below token

tbursztyka · 2017-07-07T09:45:06Z

include/net/zoap.h

move it below token, to ensure only 1byte is lost due to struct alignment.

Done. Looks a bit odd to split token / tkl though. :/

tbursztyka · 2017-07-07T09:47:18Z

include/net/lwm2m.h

not sure if it's github's viewer fault, but could you indent and align parameters under last ( + 1 space?
Apply that everywhere.

tbursztyka · 2017-07-07T09:48:23Z

include/net/lwm2m.h

why such space between #define and the actual term?
I guess it's legacy stuff from contiki, it's worth changing this however.

This is a file I created, and I don't mind fixing spacing / readability issues like this. Done.

tbursztyka · 2017-07-07T09:50:08Z

subsys/net/lib/lwm2m/Kconfig

I think you could stick with LWM2M without _LIB
Like HTTP does etc...

Heh, I used the MQTT_LIB model. Removing the _LIB throughout.

tbursztyka · 2017-07-07T10:01:00Z

subsys/net/lib/lwm2m/lwm2m_engine.c

having a lot of for/if/else imbrication makes things hard to read. Though here it's contiki legacy I guess, could be worth to change it so if () { continue; }

This is my code and I'll fix for readability.

tbursztyka · 2017-07-07T10:04:08Z

subsys/net/lib/lwm2m/lwm2m_engine.c

ouch, full of if/while/else/... imbrication below, hard to read. (legacy thing again)

Same. This is my code and I'll fix for readability.

tbursztyka · 2017-07-07T10:07:23Z

subsys/net/lib/lwm2m/lwm2m_rw_json.c

Would there be a way for using part of zephyr's json library?

Yes, I have a TODO marked in the header:
"Research using Zephyr JSON lib for json_next_token()"

tbursztyka · 2017-07-07T10:11:26Z

subsys/net/lib/lwm2m/Kconfig.ipso

LWM2M_ prefix is needed on all sub-config options in this menu.

tbursztyka · 2017-07-07T10:12:59Z

include/misc/byteorder.h

ok but why is this patch part of lwm2m patchset? You probably want to send this patch in another PR.

I added it here so testers could pull this as a part of the PR to test LWM2M. I will definitely submit this separately.

as long as it is in its commit, I do not think it make sense to create another PR, the commit is needed here, so it should be part of this PR, lets not go into this please, it is enough to have commits separated and have PRs be more inclusive and allow coverage of major features that touch different parts of the system.

dbkinder · 2017-07-07T16:54:18Z

doc/subsystems/networking/overview.rst

Machine-to-Machine

dbkinder · 2017-07-07T16:56:36Z

samples/net/lwm2m_client/README.rst

Header underline needs to be at least as long as the title (needs one more #)

mike-scott · 2017-07-07T19:36:46Z

Thank you for the initial feedback / reviews! I've pushed a V2.
V1->V2 diff: https://hastebin.com/texubuzoru.cs

CHANGES:

Addressed review comments (Thanks @tbursztyka and @dbkinder)
Adjusted the patch date/times so they are all sequential. This helps display order in github
Use isdigit() function to check for numeric chars in atou16()
Removed extraneous copy/paste comment in lwm2m_engine_get_rd_data()
Fixed checkpatch errors (still have warnings for typedefs)
Fixed sample compile breakage due to latest master branch merges in net app APIs. (Thanks @rbtchc)
Attempted to clean up if/then/while/for nesting in string_to_path(): https://github.com/mike-scott/zephyr/blob/lwm2m-new/subsys/net/lib/lwm2m/lwm2m_engine.c#L681
Attempted to clean up if/then/while/for nesting in do_read_op(): https://github.com/mike-scott/zephyr/blob/lwm2m-new/subsys/net/lib/lwm2m/lwm2m_engine.c#L1696

DISCUSSION POINT:

Float32 / Float64 handling needs to have a standard representation of float data. Currently, I'm trying use these structures: https://github.com/mike-scott/zephyr/blob/lwm2m-new/include/net/lwm2m.h#L98 Perhaps there's a better option?

mike-scott · 2017-07-07T19:45:02Z

subsys/net/lib/zoap/zoap.c

 		int age;

-		if (r->tkl != tkl) {
+		if ((r->id == 0) &&


Apparently there's some more testing to do with this patch. It might cause a regression where we're matching erroneously. This patch fixes a RST return from Leshan on an observer notification where the user as stopped the observer. The RST message has an ID which matches the newly stored reply->id, but there's no referencing token. I should probably add this information to the patch description.

I should probably add this information to the patch description.

Yes please.

mike-scott · 2017-07-08T16:59:32Z

FYI: I'm going to be travelling this next week so I won't be as responsive to questions on this PR. But, I will check in from time to time.

jukkar

Hi Mike, lot of code to review and I probably missed something. Anyway, looks good but had some comments.

jukkar · 2017-07-11T11:11:07Z

subsys/net/lib/lwm2m/Kconfig

+	bool "support for RD LWM2M client state machine"
+	default y
+	help
+	  Client will use RD client state machine to stay connected to LWM2M


It might be good idea to open the RD here in the help text.

Ack. Providing a better description in general.

jukkar · 2017-07-11T11:13:16Z

subsys/net/lib/lwm2m/Kconfig

+	  This value sets the maximum number of error codes that the device
+	  object will store before ignoring new values.
+
+menu "IPSO Alliance Object Support"


Perhaps "IPSO Alliance Smart Object Support" here?

jukkar · 2017-07-11T11:13:56Z

subsys/net/lib/lwm2m/Kconfig.ipso

+#
+
+menuconfig LWM2M_IPSO_SUPPORT
+	bool "IPSO Alliance Object Support"


s/Object/Smart Object/

jukkar · 2017-07-11T11:18:37Z

subsys/net/lib/lwm2m/lwm2m_engine.c

+static struct k_thread engine_thread_data;
+
+/* for debugging: to print IP addresses */
+char *sprint_ip_addr(const struct sockaddr *addr)


Perhaps static here, or add a prefix?

added a lwm2m_* prefix here as this is used in other files

jukkar · 2017-07-11T11:19:27Z

subsys/net/lib/lwm2m/lwm2m_engine.c

+	return NULL;
+}
+
+char *sprint_token(const u8_t *token, u8_t tkl)


Making this a static function as it's not used in other files

jukkar · 2017-07-11T12:10:34Z

subsys/net/lib/lwm2m/lwm2m_obj_security.c

+
+	/* initialize instance resource data */
+	INIT_OBJ_RES_DATA(res[index], i, SECURITY_SERVER_URI_ID,
+		  security_uri[index], SECURITY_URI_LEN);


Indentation looks wrong here and in other init macros few lines below.

Done. Adjusted all of the smart object INIT_OBJ_* indentations. I had been lazy here during development as this data was changing fairly often.

jukkar · 2017-07-11T12:11:28Z

subsys/net/lib/lwm2m/lwm2m_obj_server.c

+
+	/* initialize instance resource data */
+	INIT_OBJ_RES_DATA(res[index], i, SERVER_SHORT_SERVER_ID,
+		&server_id[index], sizeof(*server_id));


Indentation

jukkar · 2017-07-11T12:13:15Z

subsys/net/lib/lwm2m/lwm2m_rd_client.c

+
+	s64_t last_update;
+
+	char ep_name[32];


Perhaps we could have a define instead of magic constant here?

Done. Added a #define for CLIENT_EP_LEN

jukkar · 2017-07-11T12:17:13Z

subsys/net/lib/lwm2m/lwm2m_rd_client.c

+
+static int lwm2m_rd_client_init(struct device *dev)
+{
+	k_thread_create(&lwm2m_rd_client_thread_data,


There are lot of threads created by lwm2m system, is it possible to combine them (or some of them)? These many threads use quite a lot of precious ram for stack space.

I'd like to address this in a follow up patch if possible. I can add a JIRA if needed.

Created JIRA to track memory reduction effort: https://jira.zephyrproject.org/browse/ZEP-2418

jukkar · 2017-07-11T12:18:10Z

subsys/net/lib/zoap/zoap.c

 		int age;

-		if (r->tkl != tkl) {
+		if ((r->id == 0) &&


I should probably add this information to the patch description.

Yes please.

rbtchc · 2017-07-12T10:20:53Z

subsys/net/lib/lwm2m/lwm2m_obj_firmware_pull.c

+	ret = zoap_packet_init(&request, pkt);
+	if (ret < 0) {
+		SYS_LOG_ERR("zoap packet init error (err:%d)", ret);
+		return ret;


should be "goto cleanup;"

nice catch. Done.

mike-scott · 2017-07-14T10:07:55Z

Thank you for the reviews / feedback. I will submit V3 on Tuesday / Wednesday (once I return from holiday):

Better documentation
Better patch descriptions
Address review comments / open JIRAs for tracking outstanding items
Fix for leaking reply structs
Rebase on net branch

rbtchc · 2017-07-17T05:57:03Z

subsys/net/lib/lwm2m/lwm2m_engine.c

+	if (net_pkt_family(pkt) == AF_INET6) {
+		net_ipaddr_copy(&net_sin6(&from_addr)->sin6_addr,
+				&NET_IPV6_HDR(pkt)->src);
+		net_sin6(&from_addr)->sin6_port = NET_TCP_HDR(pkt)->src_port;


NET_TCP_HDR() -> NET_UDP_HDR()
Assuming we are using UDP.

And in latest master branch, both have been removed check 3604c39

Fixing all of these per master branch changes

rbtchc · 2017-07-17T05:57:46Z

subsys/net/lib/lwm2m/lwm2m_engine.c

+	if (net_pkt_family(pkt) == AF_INET) {
+		net_ipaddr_copy(&net_sin(&from_addr)->sin_addr,
+				&NET_IPV4_HDR(pkt)->src);
+		net_sin(&from_addr)->sin_port = NET_TCP_HDR(pkt)->src_port;


rbtchc · 2017-07-17T05:58:27Z

subsys/net/lib/lwm2m/lwm2m_obj_firmware_pull.c

+	if (net_pkt_family(pkt) == AF_INET6) {
+		net_ipaddr_copy(&net_sin6(&from_addr)->sin6_addr,
+				&NET_IPV6_HDR(pkt)->src);
+		net_sin6(&from_addr)->sin6_port = NET_TCP_HDR(pkt)->src_port;


rbtchc · 2017-07-17T05:58:37Z

subsys/net/lib/lwm2m/lwm2m_obj_firmware_pull.c

+	if (net_pkt_family(pkt) == AF_INET) {
+		net_ipaddr_copy(&net_sin(&from_addr)->sin_addr,
+				&NET_IPV4_HDR(pkt)->src);
+		net_sin(&from_addr)->sin_port = NET_TCP_HDR(pkt)->src_port;


rbtchc · 2017-07-19T08:44:41Z

subsys/net/lib/lwm2m/lwm2m_rw_oma_tlv.c

+	struct oma_tlv tlv;
+	s16_t net_value;
+
+	net_value = sys_cpu_to_be32(value);


sys_cpu_to_be32() -> sys_cpu_to_be16()

nice catch. done.

mike-scott · 2017-07-21T05:38:41Z

Hi all,

I've posted V3 changes for this PR. I apologize for the sizable difference from V2. There was a lot of renaming and refactoring due to review comments.
V2->V3 diff: https://hastebin.com/ukikeregoz.cs (LwM2M changes only, not rebase related)

HIGH LEVEL CHANGES FOR V3:

Rebased on master branch today (7/20)
Addressed review comments and added JIRA sub-tasks to LWM2M for device management #1780 for those items not fixed.
Added better patch descriptions. I think I missed a few at the beginning, I'll get those in V4.
Fixed up some warnings when compiling the sample w/ various combinations of IPV4 / IPV6
Added dummy temperature data to the sample client
Re-added float content formatter for OMA TLV which only handles the whole number portions currently. Will need to add calculations to handle the decimal portion at some point in the future (post merge). This is useful as the temperature object can now at least show some data.
Updated execute methods to handle return values (code provided by @rsalveti )
Correctly free up zoap_reply and zoap_pending indexes when errors happen in net_context_sendto()
Several master branch changes were fixed such as the removal of NET_UDP_HDR/NET_TCP_HDR macros and zoap lib changes.
Some duplicate code was replaced with common use functions
Removed check_perm() function as it was only used one in a place where later the permission was checked anyway.

OPEN QUESTIONS:

Float32 / Float64 handling needs to have a standard representation of float data. Currently, I'm trying use these structures: https://github.com/mike-scott/zephyr/blob/lwm2m-new/include/net/lwm2m.h#L98 Perhaps there's a better option?

OPEN JIRA SUB-TASKS:
GH-3844 : Fix LWM2M header calculation in lwm2m_engine.c
GH-3847 : Fix LwM2M object firmware pull block transfer mode to select size via interface type (BT, ethernet, etc)
GH-3848 : Implement DTLS/bootstrap for LwM2M library / sample
GH-3849 : Reduce the overall memory usage of the LwM2M library

Mike

mike-scott · 2017-07-24T22:16:05Z

V4 update specifically addresses the ZoAP id/token matching commit "net: zoap: use message id for reply matching"
Diff: https://hastebin.com/erehohipup.php

NOTES:

The new patch avoids matching on id 0 and breaks out the matching checks so that they happen in the right order.
Also added a nice commit description explaining the issue that the commit is fixing.
Rebased on Jul 24 master branch

jukkar

Looks really good. I had couple of comments but those can be done later. So if there is no objections from anyway this could be commited.

jukkar · 2017-07-25T10:49:27Z

subsys/net/lib/lwm2m/lwm2m_engine.c

+	 * zoap expects that buffer->data starts at the
+	 * beginning of the CoAP header
+	 */
+	header_len = net_pkt_appdata(pkt) - pkt->frags->data;


There is also a zoap specific Jira ticked that is related to this issue https://jira.zephyrproject.org/browse/ZEP-2210

jukkar · 2017-07-25T10:51:20Z

subsys/net/lib/lwm2m/lwm2m_engine.c

Is this to be implemented or just missing atm? Please add "TODO" here if this needs to be implemented later.

jukkar · 2017-07-25T10:55:30Z

subsys/net/lib/lwm2m/lwm2m_engine.c

BTW, if we want to have transparent DTLS support, then the net-app API net_app_send_pkt() should be used here, instead of direct call to net_context_sendto(). This can be changed in later PR of course.

mike-scott · 2017-07-25T17:19:10Z

I'm going to add a V5 which:

Addresses your last comments
Adds a patch to add an entry into MAINTAINERS file for subsys/net/lib/lwm2m
Add s a patch to add an entry in doc/LICENSE.rst identifying the content formatter files and rd client files which have 3-clause BSD license.

And I'll add a JIRA subtask to track net app API conversion (send pkt)

mike-scott · 2017-07-25T18:00:53Z

Apparently if the license is in the source files, then an entry in doc/LICENSING.rst is not needed (see subsys/net/ip/rpl.c)

mike-scott · 2017-07-25T20:37:04Z

Patchset V5 diff: https://hastebin.com/ajicibemij.diff

NOTE: I'm seeing a page fault currently when testing the LwM2M client on qemu x86. I'm currently debugging why CONFIG_X86_STACK_PROTECTION is causing this.

mike-scott · 2017-07-25T20:37:39Z

FYI: There were 3 edits to fix checkpatch warnings for long lines not mentioned.

mike-scott · 2017-07-26T17:43:49Z

Pushed V7 with only patch description changes to fix SHIPPABLE warnings (I hope).

jukkar · 2017-07-27T07:28:12Z

@nashif We could merge this now but @tbursztyka earlier change request prevents merging, can you override that and merge this PR (Tomasz is on vacation right now)?

mike-scott · 2017-07-27T15:46:54Z

@nashif mentioned on IRC that the "net: lwm2m: initial library support for LWM2M" commit is missing a block in the commit description similar to:

Origin: SICS-IoT / Contiki OS
URL: https://github.com/sics-iot/lwm2m-contiki/tree/lwm2m-standalone-dtls
commit: d07b0bcd77ec7e8b93787669507f3d86cfbea64a
Purpose: Introduction of LwM2M client library.
Maintained-by: Zephyr

However, this is only true for the 9 files with 3-clause BSD licenses still attached. How should I clarify that?

We currently support converting from cpu format to BE for u16_t and u32_t. Let's add u64_t as well. NOTE: This will be used in LWM2M subsys later. Signed-off-by: Michael Scott <[email protected]>

In the 08 Feb 2017 V1.0 LwM2M specification page 80 mentions: in response to a "Notify" operation for which it is not interested in any more, the LwM2M Server can send a "Reset Message". Leshan server sends this CoAP RST response and it does not contain the originating message token (which is also how the packet flow looks on page 81 of the LwM2M spec). Using the current ZoAP sources, the client has no way of matching back to observation which needs to be cancelled. Let's add a match for message ID of a reply where there is no token to handle this case. Signed-off-by: Michael Scott <[email protected]> [[email protected]: Handle both piggybackend and separate response (id doesn't need to match, only token).] Signed-off-by: Ricardo Salveti <[email protected]>

Origin: SICS-IoT / Contiki OS URL: https://github.com/sics-iot/lwm2m-contiki/tree/lwm2m-standalone-dtls commit: d07b0bcd77ec7e8b93787669507f3d86cfbea64a Purpose: Introduction of LwM2M client library. Maintained-by: Zephyr Lightweight Machine-to-Machine (LwM2M) is a protocol stack extension of the Constrained Application Protocol (CoAP) which uses UDP transmission packets. This library was based on source worked on by Joakim Eriksson, Niclas Finne and Joel Hoglund which was adopted by Contiki and then later revamped to work as a stand-alone library. A VERY high level summary of the changes made: - [ALL] sources were re-formatted to Zephyr coding standards - [engine] The engine portion was re-written due to the heavy reliance on ER-CoAP APIs which are not compatible to the Zephyr CoAP APIs as well as other Zephyr specific needs. - [engine] All LWM2M/IPSO object data is now abstracted into resource data which stores information like the data type, length, callbacks to help with read/write. The engine modifies this data directly (or makes callbacks) instead of all of the logic for this living in each object's code. (This wasn't scaling well as I was implementing changes). - [engine] Related to the above change, I also added a generic set of getter/setter functions that user applications can call to change the object data instead of having to add getter/setting methods in each object. - [engine] The original sources shared the engine's context structure quite extensively causing a problem with portability. I broke up the context into it's individual parts: LWM2M path data, input data and output data and pass only the needed data into each set of APIs. - [content format read/writer] sources were re-organized into single .c/h files per content formatter. - [content format read/writer] sources were re-written where necessary to remove the sharing of the lwm2m engine's context and instead only requires the path and input or output data specific to it's function. - [LwM2M objects] re-written using the new engine's abstractions Signed-off-by: Michael Scott <[email protected]>

This sample utilizes the new LwM2M library by setting up default values for LwM2M device and firmware objects and then establisting a connection to a LwM2M server (for example Leshan Demo Server) via the registration interface. To use QEMU for this purpose please see: doc/subsystems/networking/qemu_setup.rst NOTE: This sample currently does not demonstrate DTLS/bootstrap as neither of these is supported by the LwM2M library. Signed-off-by: Michael Scott <[email protected]>

IPSO Smart Objects are a set of template objects based on the LwM2M object framework which are designed to represent standard hardware such as temperature and humidity sensors or light controls. Let's add a place for these objects to live as well as an initial temperature sensor object. Signed-off-by: Michael Scott <[email protected]>

This commit adds IPSO temperature support to the LwM2M client sample. NOTE: A dummy value of 25C is set during initialization and does not change. Signed-off-by: Michael Scott <[email protected]>

Signed-off-by: Michael Scott <[email protected]>

mike-scott · 2017-07-27T17:55:00Z

Pushed V8 to address the contribution guide requirement for the above mentioned block in the "net: lwm2m: initial library support for LWM2M" patch. That was the only change.

mike-scott · 2017-07-27T17:55:26Z

Correction: I also rebased on today's master branch.

mike-scott · 2017-07-27T18:12:58Z

Ugh, SHIPPABLE complaining about the URL line length in the new change to the patch description:
https://app.shippable.com/github/zephyrproject-rtos/zephyr/runs/2782/3/tests

NOTE: running gitlint in my local repo does not show this error. :/

jukkar · 2017-08-01T07:22:50Z

@nashif Could you force merge this as shippable complains about non-issue?

dbkinder

+1 for doc so far, but_lwm2m-client-sample is still empty :(

Signed-off-by: James Prestwood <[email protected]>

tbursztyka requested changes Jul 7, 2017

View reviewed changes

jukkar self-requested a review July 7, 2017 13:01

dbkinder reviewed Jul 7, 2017

View reviewed changes

mike-scott force-pushed the lwm2m-new branch from b7e2a0b to 756bbec Compare July 7, 2017 19:25

mike-scott commented Jul 7, 2017

View reviewed changes

jukkar requested changes Jul 11, 2017

View reviewed changes

jukkar added the net label Jul 11, 2017

rbtchc reviewed Jul 12, 2017

View reviewed changes

nashif closed this Jul 15, 2017

nashif reopened this Jul 15, 2017

rbtchc reviewed Jul 17, 2017

View reviewed changes

rbtchc reviewed Jul 19, 2017

View reviewed changes

mike-scott force-pushed the lwm2m-new branch from 756bbec to c58da60 Compare July 21, 2017 05:15

mike-scott requested review from andrewboie and galak as code owners July 21, 2017 05:15

mike-scott changed the title ~~[RFC] Introduce LWM2M subsystem for Zephyr~~ Introduce LWM2M subsystem for Zephyr Jul 21, 2017

mike-scott force-pushed the lwm2m-new branch from c58da60 to 7fd5bed Compare July 24, 2017 22:12

jukkar reviewed Jul 25, 2017

View reviewed changes

mike-scott force-pushed the lwm2m-new branch from 7fd5bed to e4ceaf9 Compare July 25, 2017 20:25

mike-scott requested a review from inakypg as a code owner July 25, 2017 20:25

mike-scott force-pushed the lwm2m-new branch from 48ed6ee to ccc312c Compare July 26, 2017 17:43

jukkar previously approved these changes Jul 27, 2017

View reviewed changes

Michael Scott added 9 commits July 27, 2017 10:52

misc/byteorder: add support for __bswap_64

10d897a

We currently support converting from cpu format to BE for u16_t and u32_t. Let's add u64_t as well. NOTE: This will be used in LWM2M subsys later. Signed-off-by: Michael Scott <[email protected]>

samples: lwm2m: add IPSO temperature object to LWM2M client

65035d1

This commit adds IPSO temperature support to the LwM2M client sample. NOTE: A dummy value of 25C is set during initialization and does not change. Signed-off-by: Michael Scott <[email protected]>

net: lwm2m: add SPX Apache-2.0 license tag w/ Linaro copyright

7970296

Signed-off-by: Michael Scott <[email protected]>

MAINTAINERS: add maintainer for LwM2M library / samples

e7ed1a0

Signed-off-by: Michael Scott <[email protected]>

CODEOWNERS: add entries for LwM2M library / samples

4f27e89

Signed-off-by: Michael Scott <[email protected]>

mike-scott dismissed jukkar’s stale review via 4f27e89 July 27, 2017 17:53

mike-scott force-pushed the lwm2m-new branch from ccc312c to 4f27e89 Compare July 27, 2017 17:53

jukkar approved these changes Jul 28, 2017

View reviewed changes

dbkinder approved these changes Aug 3, 2017

View reviewed changes

tbursztyka approved these changes Aug 7, 2017

View reviewed changes

nashif approved these changes Aug 8, 2017

View reviewed changes

jukkar merged commit 02383de into zephyrproject-rtos:master Aug 9, 2017

This was referenced Sep 18, 2017

LWM2M for device management zephyriot/zep-jira14#181

Closed

LWM2M for device management zephyriot/zephyr-issues#219

Closed

zephyrbot mentioned this pull request Sep 23, 2017

LWM2M for device management #1780

Closed

mike-scott deleted the lwm2m-new branch October 2, 2017 21:13

nagineni pushed a commit to nagineni/zephyr that referenced this pull request Nov 20, 2017

[build] Fix OCF build break (zephyrproject-rtos#712)

3f937ee

Signed-off-by: James Prestwood <[email protected]>

Introduce LWM2M subsystem for Zephyr #712

Introduce LWM2M subsystem for Zephyr #712

Uh oh!

Conversation

mike-scott commented Jul 7, 2017

Uh oh!

tbursztyka left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

mike-scott commented Jul 7, 2017

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

mike-scott commented Jul 8, 2017

Uh oh!

jukkar left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!