Mavlink: Fixed command sender race condition #12560
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
RomanBapst
changed the title
julianoes
reviewed
Jul 29, 2019
| @@ -172,7 +172,7 @@ void MavlinkCommandSender::check_timeout(mavlink_channel_t channel) | |||
| } | |||
| } | |||
|
|
|||
| if (item->num_sent_per_channel[channel] < max_sent) { | |||
| if (item->num_sent_per_channel[channel] < max_sent && item->num_sent_per_channel[channel] != -1) { | |||
There was a problem hiding this comment.
I don't exactly understand this one. Why would it still be -1?
|
As discussed with @bkueng : If a command in the queue receives an ack, set the field "num_sent_per_channel" to -2 but leave the command in the queue. When we reach the timeout for that command we only re-send the command if "num_sent_per_channel" is not -2. |
|
Yes essentially the changes here are good (hardening the logic), but don't solve the race condition. The same command is still inserted several times. To resolve it we can wait for at least 1 timeout before removing commands from the list. |
- if we receive an ack for a command through a specific mavlink channel then do not drop the corresponding command in the queue if this specific mavlink channel did not issue the command. If we don't do this we can end up in a situation where we associate an ack coming through a specific mavlink channel to a command in the queue which was not requested by this mavlink channel. Moreover, the actual command for which the ack was meant remains in the queue and eventually triggers a timeout. Signed-off-by: RomanBapst <bapstroman@gmail.com>
did not request this command Signed-off-by: RomanBapst <bapstroman@gmail.com>
- if a channel receives an ack for a command, do not immediately remove the command item from the send queue but wait until the next ack timeout occurs. This gives other mavlink channels time to try to put identical commands into the send queue. Signed-off-by: RomanBapst <bapstroman@gmail.com>
RomanBapst
force-pushed
the
pr-mavlink_race
branch
from
559d247 to
9a41d27
Compare
|
@bkueng I updated based on your suggestion. Could you please review again. I verified basic functionality. |
bkueng
approved these changes
Aug 5, 2019
| @@ -147,6 +146,23 @@ void MavlinkCommandSender::check_timeout(mavlink_channel_t channel) | |||
| continue; | |||
| } | |||
|
|
|||
| // Loop through num_sent_per_channel and check if any channel has receives an ack for this command | |||
| // (indicated by the value -2). We avoid removing the command at the time of receiving the ack | |||
| // as some channels might be lagging behind and will end up putting the same command into the buffer. | |||
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This is an attempt to fix #12553
See commit message for further explanation.
This fixes irregular camera trigger messages. I'm not aware if there is something I'm breaking with this change, e.g. if there was a particular reason why the current behavior is as is.