Page MenuHomeFreeBSD

powerpc/powernv: Add OPAL heartbeat thread
ClosedPublic

Authored by jhibbits on Mar 29 2019, 3:31 AM.
Tags
None
Referenced Files
F106130641: D19743.id55713.diff
Wed, Dec 25, 10:04 PM
F106099578: D19743.diff
Wed, Dec 25, 9:41 AM
Unknown Object (File)
Fri, Nov 29, 1:38 AM
Unknown Object (File)
Fri, Nov 29, 1:38 AM
Unknown Object (File)
Fri, Nov 29, 1:37 AM
Unknown Object (File)
Fri, Nov 29, 1:18 AM
Unknown Object (File)
Nov 9 2024, 7:40 AM
Unknown Object (File)
Sep 20 2024, 2:25 PM
Subscribers

Details

Summary

OPAL needs to be kicked periodically in order for the firmware to make
progress on its tasks. To do so, create a heartbeat thread to perform this task
every N milliseconds, defined by the device tree. This task is also a central
location to handle all messages received from OPAL.

Diff Detail

Repository
rS FreeBSD src repository - subversion
Lint
Lint Passed
Unit
No Test Coverage
Build Status
Buildable 23365
Build 22389: arc lint + arc unit

Event Timeline

Looks good overall, but check the comments below.

sys/powerpc/powernv/opal_dev.c
134

Shouldn't opal_call() return code be checked here?

203

Is "ibm,heartbeat-ms" property guaranteed to exist? Wouldn't it be better to check this and set a default or panic if it doesn't exist?

sys/powerpc/powernv/opal_dev.c
134

Technically, probably. The OPAL code, though, returns either OPAL_SUCCESS or OPAL_PARAMETER (invalid pointer argument), so we're likely not to see any failure. I'll add a check before committing, though.

203

If it doesn't exist, since opal_heartbeat_ms is static it's pre-initialized to 0. It'll then be caught when spawning the thread (see above at the top of opal_heartbeat()) No need to put the check in two places.

Ok, thanks for the explanation, then it LGTM.

This revision is now accepted and ready to land.Apr 1 2019, 3:51 PM
This revision was automatically updated to reflect the committed changes.