[
https://issues.apache.org/jira/browse/MYNEWT-630?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15884011#comment-15884011
]
ASF subversion and git services commented on MYNEWT-630:
--------------------------------------------------------
Commit f34f225ddc1b87cf49a5dde17dd5e4ea6f8169a0 in incubator-mynewt-core's
branch refs/heads/develop from [~wes3]
[ https://git-wip-us.apache.org/repos/asf?p=incubator-mynewt-core.git;h=f34f225
]
MYNEWT-630: Connection supervision timeout with high throughput
The issue was found to be an error in the os scheduler code.
If the PendSV handler was interrupted at a certain point when
we were checking to see if a context switch could occur, the
code is os_sched would think no context switch was necessary.
This is now fixed and the high throughput test has been running
for quite some time now with no disconnect.
> Connection supervision timeout with a very high throughput connection
> ---------------------------------------------------------------------
>
> Key: MYNEWT-630
> URL: https://issues.apache.org/jira/browse/MYNEWT-630
> Project: Mynewt
> Issue Type: Bug
> Components: Nimble
> Affects Versions: v1_0_0_beta2
> Reporter: William San Filippo
> Assignee: William San Filippo
> Fix For: v1_0_0_rel
>
>
> While doing throughput testing with 27-byte PDU's (no data length extension)
> using bluez and l2test the devices would disconnect after some amount of time
> due to a supervision timeout. This did not happen when we increased the size
> of the LL PDU's; it was only with 27-byte PDU's.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
