Peter Gebruers <peter.gebruers@gmail.com>: Author Summary

Builds triggered by Peter Gebruers <peter.gebruers@gmail.com>

Builds triggered by an author are those builds which contains changes committed by the author.
52
3 (6%)
49 (94%)

Breakages and fixes

Broken means the build has failed but the previous build was successful.
Fixed means that the build was successful but the previous build has failed.
1 (2% of all builds triggered)
2 (4% of all builds triggered)
1
Build Completed Code commits Tests
OZW › ODB › #195 6 months ago
Rename Serial Version to Firmware Version and add Communication and Chip Version (#2168)

* Rename Serial Version to Firmware Version and add Communication Version

* Add controller Chip Type and Version
716 passed
OZW › ODB › #164 8 months ago
Merge pull request #2031 from radovanbauer/master

Added config for GE (Jasco) 46203 In-Wall Smart Dimmer.
Fix comparison uint8 with -1, also test if controller node is 0.

lldb complained:
Driver.cpp:2701:43: error:
      result of comparison of constant -1 with expression of type
      'uint8' (aka 'unsigned char') is always false
      [-Werror,-Wtautological-constant-out-of-range-compare]

Add some info to debug logging
Fix handling of controller CAN (#2049)

The key change is that before this commit "CAN received...triggering resend" immediately leads to outbound data "Sending (Command) message (Attempt 2".

After this PR then "CAN received...triggering resend will handle the incoming data "Unsolicited message received while waiting for ACK."
Fix test failure on build server

-Allow ozw_vers_revision to be 0
-Add -Wsign-compare to build flags
-Fix warnings caught by -Wsign-compare
Fix config parameter 68 in ZMNHDDX.xml #1995 (#2001)
Call CheckCompletedNodeQueries in MoveMessagesToWakeUpQueue (#2034)

If a network consists of sleeping nodes (and they sleep) none of the notifications Type_AwakeNodesQueried Type_AllNodesQueriedSomeDead Type_AllNodesQueried is send because all messages are moved to the wake up queue.

Reported by Gilles Printemps, https://groups.google.com/d/msg/openzwave/THGYQ1jqKF4/cZPF5FEDBQAJ
Add ValueID::GetAsString() for debugging purposes (#2040)

* Add ValueID::GetAsString() for debugging purposes

This function allows to convert a valueID to human readable form

Example (from tests):

ValueID(static_cast<uint64>(0x1), 0x02, ValueID::ValueGenre_System, 0x04, 5, 6, ValueID::ValueType_String).GetAsString()

"HomeID: 0x00000001, ValueID: (Id 0x0006000002c10057, NodeID 2, Genre system, CC 0x04, Instance 5, Index 6, Type string)"

Related to https://github.com/OpenZWave/open-zwave/issues/2038 "Ozwcache corruption leading to memory corruption and crash"

* Add Log to Node::GetValue in case retrieved ID does not match request ID

* no need to introduce a namespace

Co-authored-by: Justin Hammond <justin@dynam.ac>
Fix a possible infinite loop with invalid list configuration data (#2047)

Before this commit, it is possible to have an infinite loop and log filling with:

"Warning, Item value  is wrong type or does not exist in xml configuration for node 9, class 0x70, instance 1, index 1"

This is because the code skipped "itemElement = itemElement->NextSiblingElement();"

IMHO only happens with corrupt or edited ozwcache file.
Add message details to log of dropped message (#2011)

Before PR

Node002, ERROR: Dropping command, expected response not received after 1 attempt(s).

After PR

ode002, ERROR: Dropping command, expected response not received after 1 attempt(s). Command: "SensorMultilevelCmd_Get (Node=2): 0x01, 0x0b, 0x00, 0x13, 0x02, 0x04, 0x31, 0x04, 0x01, 0x00, 0x25, 0x20, 0xd0"
Update openzwave.spec
Don't request Instances is Multichannel is "after mark"

IMHO if the Multichannel CC is net before, but only after the "mark" then the device is not multi channel.

Adds log:

2019-12-05 10:34:57.454 Detail, Node004, Skipping RequestInstances() because MultiChannel CC is "after mark"

Fixes "Problems with ZME_WALLC-S after Update from 1.4 to 1.6 - states missing" #2020 by not doing 17 requests that all time out...
Fix SensorBinary V1 decoding (#2006)

I noticed bogus Binary Sensor numbers in the log, for example:

Node002, Received SensorBinary report: Sensor:199 State=On

Based on serial frame:

Received: 0x01, 0x0b, 0x00, 0x04, 0x00, 0x02, 0x03, 0x30, 0x03, 0xff, 0xc7, 0x00, 0xfa

Length of the CC is 0x03, payload is 0x30, 0x03, 0xff, so OZW mistakenly takes next byte as "Sensor Type"

The length test was off by one.

After this patch:

Node002, Received SensorBinary report: State=On

V2 reports still work:

Received: 0x01, 0x0a, 0x00, 0x04, 0x00, 0x02, 0x04, 0x30, 0x03, 0xff, 0x0c, 0x37

Node002, Received SensorBinary report: Sensor:12 State=On
Add PAN06-1 In Wall Dual Relay with id 0x13, PR #2014
Make flock(m_hSerialController... actually do something (#2026)

Previously we will still try to continue if we couldn't get a LOCK on the serial port. Now we bail out and go into a loop trying to get the Lock.
Merge pull request #2059 from petergebruers/fix_xmltest

make xmltest: fix inconsistent storage of revision (int vs string)
Merge branch 'master-ozw' into fix_xmltest

# Conflicts:
#        cpp/build/testconfigversions.cfg
Updates to "OZW revision" (git describe) script on Unix (#2028)

Potential fix for git version "dirty" flag.
Might help diagnose "revision on build server always 1.6.0".

When entering support.mk a second time, git describe returns "dirty" flag although I saw no immediate cause.

According to the docs, --git-dir should be used with --work-tree, or use the -C flag, introduced in git 1.8.5-ish.

Added debugging (Makefile warnings).

Tested on macOS Catalina and Debian 8 (Jessie), released 25–26 April 2015. Not tested on Debian 7 (Wheezy), released 4 May 2013 because it does not come with C++11 compiler by default.
make xmltest: fix inconsistent storage of revision (int vs string)

Before this commit, testconfigversions.cfg would sometimes store revision as a string, sometimes as an integer

'config/2gig/ct100.xml' => {
    'Revision' => '6',

After this commit, its always an integer

'Revision' => 6,
700 passed
OZW › OO › #35 9 months ago
Merge pull request #2059 from petergebruers/fix_xmltest

make xmltest: fix inconsistent storage of revision (int vs string)
make xmltest: fix inconsistent storage of revision (int vs string)

Before this commit, testconfigversions.cfg would sometimes store revision as a string, sometimes as an integer

'config/2gig/ct100.xml' => {
    'Revision' => '6',

After this commit, its always an integer

'Revision' => 6,
Merge branch 'master-ozw' into fix_xmltest

# Conflicts:
#        cpp/build/testconfigversions.cfg
Update openzwave.spec
Testless build
OZW › OZWMAC › #744 9 months ago
Merge pull request #2059 from petergebruers/fix_xmltest

make xmltest: fix inconsistent storage of revision (int vs string)
Update openzwave.spec
Fix handling of controller CAN (#2049)

The key change is that before this commit "CAN received...triggering resend" immediately leads to outbound data "Sending (Command) message (Attempt 2".

After this PR then "CAN received...triggering resend will handle the incoming data "Unsolicited message received while waiting for ACK."
Merge branch 'master-ozw' into fix_xmltest

# Conflicts:
#        cpp/build/testconfigversions.cfg
make xmltest: fix inconsistent storage of revision (int vs string)

Before this commit, testconfigversions.cfg would sometimes store revision as a string, sometimes as an integer

'config/2gig/ct100.xml' => {
    'Revision' => '6',

After this commit, its always an integer

'Revision' => 6,
Testless build
OZW › OZWLNX › #1344 9 months ago
make xmltest: fix inconsistent storage of revision (int vs string)

Before this commit, testconfigversions.cfg would sometimes store revision as a string, sometimes as an integer

'config/2gig/ct100.xml' => {
    'Revision' => '6',

After this commit, its always an integer

'Revision' => 6,
Merge pull request #2059 from petergebruers/fix_xmltest

make xmltest: fix inconsistent storage of revision (int vs string)
Update openzwave.spec
Merge branch 'master-ozw' into fix_xmltest

# Conflicts:
#        cpp/build/testconfigversions.cfg
691 passed
OZW › OO › #30 9 months ago
Fix handling of controller CAN (#2049)

The key change is that before this commit "CAN received...triggering resend" immediately leads to outbound data "Sending (Command) message (Attempt 2".

After this PR then "CAN received...triggering resend will handle the incoming data "Unsolicited message received while waiting for ACK."
Testless build
OZW › OZWLNX › #1340 9 months ago
Fix handling of controller CAN (#2049)

The key change is that before this commit "CAN received...triggering resend" immediately leads to outbound data "Sending (Command) message (Attempt 2".

After this PR then "CAN received...triggering resend will handle the incoming data "Unsolicited message received while waiting for ACK."
690 passed
OZW › QTOZW › #77 9 months ago
Fix test failure on build server

-Allow ozw_vers_revision to be 0
-Add -Wsign-compare to build flags
-Fix warnings caught by -Wsign-compare
Updates to "OZW revision" (git describe) script on Unix (#2028)

Potential fix for git version "dirty" flag.
Might help diagnose "revision on build server always 1.6.0".

When entering support.mk a second time, git describe returns "dirty" flag although I saw no immediate cause.

According to the docs, --git-dir should be used with --work-tree, or use the -C flag, introduced in git 1.8.5-ish.

Added debugging (Makefile warnings).

Tested on macOS Catalina and Debian 8 (Jessie), released 25–26 April 2015. Not tested on Debian 7 (Wheezy), released 4 May 2013 because it does not come with C++11 compiler by default.
Call CheckCompletedNodeQueries in MoveMessagesToWakeUpQueue (#2034)

If a network consists of sleeping nodes (and they sleep) none of the notifications Type_AwakeNodesQueried Type_AllNodesQueriedSomeDead Type_AllNodesQueried is send because all messages are moved to the wake up queue.

Reported by Gilles Printemps, https://groups.google.com/d/msg/openzwave/THGYQ1jqKF4/cZPF5FEDBQAJ
Add PAN06-1 In Wall Dual Relay with id 0x13, PR #2014
Merge pull request #2031 from radovanbauer/master

Added config for GE (Jasco) 46203 In-Wall Smart Dimmer.
Fix comparison uint8 with -1, also test if controller node is 0.

lldb complained:
Driver.cpp:2701:43: error:
      result of comparison of constant -1 with expression of type
      'uint8' (aka 'unsigned char') is always false
      [-Werror,-Wtautological-constant-out-of-range-compare]

Add some info to debug logging
Fix SensorBinary V1 decoding (#2006)

I noticed bogus Binary Sensor numbers in the log, for example:

Node002, Received SensorBinary report: Sensor:199 State=On

Based on serial frame:

Received: 0x01, 0x0b, 0x00, 0x04, 0x00, 0x02, 0x03, 0x30, 0x03, 0xff, 0xc7, 0x00, 0xfa

Length of the CC is 0x03, payload is 0x30, 0x03, 0xff, so OZW mistakenly takes next byte as "Sensor Type"

The length test was off by one.

After this patch:

Node002, Received SensorBinary report: State=On

V2 reports still work:

Received: 0x01, 0x0a, 0x00, 0x04, 0x00, 0x02, 0x04, 0x30, 0x03, 0xff, 0x0c, 0x37

Node002, Received SensorBinary report: Sensor:12 State=On
Make flock(m_hSerialController... actually do something (#2026)

Previously we will still try to continue if we couldn't get a LOCK on the serial port. Now we bail out and go into a loop trying to get the Lock.
Fix config parameter 68 in ZMNHDDX.xml #1995 (#2001)
Don't request Instances is Multichannel is "after mark"

IMHO if the Multichannel CC is net before, but only after the "mark" then the device is not multi channel.

Adds log:

2019-12-05 10:34:57.454 Detail, Node004, Skipping RequestInstances() because MultiChannel CC is "after mark"

Fixes "Problems with ZME_WALLC-S after Update from 1.4 to 1.6 - states missing" #2020 by not doing 17 requests that all time out...
Fix a possible infinite loop with invalid list configuration data (#2047)

Before this commit, it is possible to have an infinite loop and log filling with:

"Warning, Item value  is wrong type or does not exist in xml configuration for node 9, class 0x70, instance 1, index 1"

This is because the code skipped "itemElement = itemElement->NextSiblingElement();"

IMHO only happens with corrupt or edited ozwcache file.
Add message details to log of dropped message (#2011)

Before PR

Node002, ERROR: Dropping command, expected response not received after 1 attempt(s).

After PR

ode002, ERROR: Dropping command, expected response not received after 1 attempt(s). Command: "SensorMultilevelCmd_Get (Node=2): 0x01, 0x0b, 0x00, 0x13, 0x02, 0x04, 0x31, 0x04, 0x01, 0x00, 0x25, 0x20, 0xd0"
Add ValueID::GetAsString() for debugging purposes (#2040)

* Add ValueID::GetAsString() for debugging purposes

This function allows to convert a valueID to human readable form

Example (from tests):

ValueID(static_cast<uint64>(0x1), 0x02, ValueID::ValueGenre_System, 0x04, 5, 6, ValueID::ValueType_String).GetAsString()

"HomeID: 0x00000001, ValueID: (Id 0x0006000002c10057, NodeID 2, Genre system, CC 0x04, Instance 5, Index 6, Type string)"

Related to https://github.com/OpenZWave/open-zwave/issues/2038 "Ozwcache corruption leading to memory corruption and crash"

* Add Log to Node::GetValue in case retrieved ID does not match request ID

* no need to introduce a namespace

Co-authored-by: Justin Hammond <justin@dynam.ac>
Testless build
OZW › OZWMAC › #735 9 months ago
Call CheckCompletedNodeQueries in MoveMessagesToWakeUpQueue (#2034)

If a network consists of sleeping nodes (and they sleep) none of the notifications Type_AwakeNodesQueried Type_AllNodesQueriedSomeDead Type_AllNodesQueried is send because all messages are moved to the wake up queue.

Reported by Gilles Printemps, https://groups.google.com/d/msg/openzwave/THGYQ1jqKF4/cZPF5FEDBQAJ
Add ValueID::GetAsString() for debugging purposes (#2040)

* Add ValueID::GetAsString() for debugging purposes

This function allows to convert a valueID to human readable form

Example (from tests):

ValueID(static_cast<uint64>(0x1), 0x02, ValueID::ValueGenre_System, 0x04, 5, 6, ValueID::ValueType_String).GetAsString()

"HomeID: 0x00000001, ValueID: (Id 0x0006000002c10057, NodeID 2, Genre system, CC 0x04, Instance 5, Index 6, Type string)"

Related to https://github.com/OpenZWave/open-zwave/issues/2038 "Ozwcache corruption leading to memory corruption and crash"

* Add Log to Node::GetValue in case retrieved ID does not match request ID

* no need to introduce a namespace

Co-authored-by: Justin Hammond <justin@dynam.ac>
Fix a possible infinite loop with invalid list configuration data (#2047)

Before this commit, it is possible to have an infinite loop and log filling with:

"Warning, Item value  is wrong type or does not exist in xml configuration for node 9, class 0x70, instance 1, index 1"

This is because the code skipped "itemElement = itemElement->NextSiblingElement();"

IMHO only happens with corrupt or edited ozwcache file.
Testless build
OZW › OZWLNX › #1331 9 months ago
Fix a possible infinite loop with invalid list configuration data (#2047)

Before this commit, it is possible to have an infinite loop and log filling with:

"Warning, Item value  is wrong type or does not exist in xml configuration for node 9, class 0x70, instance 1, index 1"

This is because the code skipped "itemElement = itemElement->NextSiblingElement();"

IMHO only happens with corrupt or edited ozwcache file.
686 passed
Build Completed Code commits Tests
OZW › OO › #35 9 months ago
Merge pull request #2059 from petergebruers/fix_xmltest

make xmltest: fix inconsistent storage of revision (int vs string)
make xmltest: fix inconsistent storage of revision (int vs string)

Before this commit, testconfigversions.cfg would sometimes store revision as a string, sometimes as an integer

'config/2gig/ct100.xml' => {
    'Revision' => '6',

After this commit, its always an integer

'Revision' => 6,
Merge branch 'master-ozw' into fix_xmltest

# Conflicts:
#        cpp/build/testconfigversions.cfg
Update openzwave.spec
Testless build
Build Completed Code commits Tests
OZW › OZWMAC › #729 10 months ago
Fix comparison uint8 with -1, also test if controller node is 0.

lldb complained:
Driver.cpp:2701:43: error:
      result of comparison of constant -1 with expression of type
      'uint8' (aka 'unsigned char') is always false
      [-Werror,-Wtautological-constant-out-of-range-compare]

Add some info to debug logging
Don't request Instances is Multichannel is "after mark"

IMHO if the Multichannel CC is net before, but only after the "mark" then the device is not multi channel.

Adds log:

2019-12-05 10:34:57.454 Detail, Node004, Skipping RequestInstances() because MultiChannel CC is "after mark"

Fixes "Problems with ZME_WALLC-S after Update from 1.4 to 1.6 - states missing" #2020 by not doing 17 requests that all time out...
Testless build
OZW › ODB › #158 1 year ago
mhp511 fix revision and distfiles
Add support for "temperature" model of NAS-PD02Z

https://github.com/OpenZWave/open-zwave/issues/1866
Bump revision
Add Sunricher RGBW wall controller SR-ZV9003T4-RGBW-EU
Merge master
Fix config file reference srzv9001k8
merge upstream
Bump revision and format file
671 passed