Project

General

Profile

Actions

Bug #13343

closed

some mlxcx devices don't support temperature sensors

Added by Robert Mustacchi 8 months ago. Updated 8 months ago.

Status:
Closed
Priority:
Normal
Category:
driver - device drivers
Start date:
Due date:
% Done:

100%

Estimated time:
Difficulty:
Medium
Tags:
Gerrit CR:

Description

The Mellanox ConnectX-4 lx (15b3,1015), unlike other parts, appears not to have a built-in temperature sensor. Unfortunately, there's no actual way for us to tell this in the hardware through what documentation we have (what does exist like the warning event is supported by folks who have the device). Because of the way the driver is structured, this causes attach to fail. Instead, we should just not create the temperature nodes and drive on when we hit this case.

This was tested by folks at RackTop who had the ConnectX-4 Lx and hit this.

Actions #1

Updated by Electric Monk 8 months ago

  • Gerrit CR set to 1088
Actions #2

Updated by Robert Mustacchi 8 months ago

This was tested in two ways. Paul at Racktop verified that this caused us to correctly attach regardless of presence. Alex verified that the temperature sensors still work.

Actions #3

Updated by Electric Monk 8 months ago

  • Status changed from New to Closed
  • % Done changed from 90 to 100

git commit 15174c59605d00b59ea3aae8ba85744a42aa1e96

commit  15174c59605d00b59ea3aae8ba85744a42aa1e96
Author: Robert Mustacchi <rm@fingolfin.org>
Date:   2020-12-11T00:52:18.000Z

    13343 some mlxcx devices don't support temperature sensors
    Reviewed by: C Fraire <cfraire@me.com>
    Reviewed by: Jerry Jelinek <gjelinek@gmail.com>
    Reviewed by: Andy Fiddaman <andy@omnios.org>
    Reviewed by: Paul Winder <paul@winder.uk.net>
    Approved by: Dan McDonald <danmcd@joyent.com>

Actions

Also available in: Atom PDF