[
https://issues.apache.org/jira/browse/MYNEWT-542?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15813529#comment-15813529
]
William San Filippo commented on MYNEWT-542:
--------------------------------------------
After reviewing the code there were a number of places the controller was not
correctly converting microseconds to os_cputime ticks. These errors were mainly
when calculating timing based on the duration of a PDU. These bugs have been
fixed. Some caveats:
1) This was only tested with a 2 MHz os cputime; no other values were tried.
2) For the nrf51 there are still failures but these are not associated with the
code changes per se. The normal "ticks to usecs" and "usecs to ticks" functions
are not used when cputime is 1MHz as 1 usec = 1 tick and they get compiled out.
However, once you change that, the code actually gets compiled and the time it
takes to do these calculations affects timing (things take too long). This
issue will not be addressed in this ticket; when we move to a 32kHz clock they
will be addressed.
> Controller does not work if os cputime set to a value other than 1 MHz.
> -----------------------------------------------------------------------
>
> Key: MYNEWT-542
> URL: https://issues.apache.org/jira/browse/MYNEWT-542
> Project: Mynewt
> Issue Type: Bug
> Components: Nimble
> Affects Versions: v1_0_0_beta1
> Reporter: William San Filippo
> Assignee: William San Filippo
> Fix For: v1_0_0_rel
>
>
> The default os cputime was changed from 1 MHz to 2 MHz and the controller
> failed to connect. This was tried on a nrf52 platform.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
