svn commit: r331591 - stable/11/sys/dev/mlx5/mlx5_core

[Hans Petter Selasky]

On 03/27/18 00:15, Hans Petter Selasky wrote:
> On 03/27/18 00:04, John Baldwin wrote:
>> On Monday, March 26, 2018 09:06:23 PM Hans Petter Selasky wrote:
>>> Author: hselasky
>>> Date: Mon Mar 26 21:06:23 2018
>>> New Revision: 331591
>>> URL: https://svnweb.freebsd.org/changeset/base/331591
>>>
>>> Log:
>>>    MFC r330660:
>>>    Add call to setup firmware data dump structure during device load in
>>>    mlx5core.
>>>    Do not consider the inability to create a firmware dump fatal, but
>>>    inform about the situation and allow the driver to attach. The device
>>>    might not implement the needed VSC, or we might not know the 
>>> layout of
>>>    the registers map. In either case, only firmware dump 
>>> functionality is
>>>    limited, the network operations should be fine.
>>>    Submitted by:    kib@
>>>    Sponsored by:    Mellanox Technologies
>>
>> Please batch up MFC commits at least enough to ensure they compile.  We
>> really should not be breaking the build on stable branches.  If you are
>> merging a commit that required followup fixes in head, the MFC needs to
>> include those followup fixes.
>>
>> I'm getting this failure on a tree at this revision:
>>
>> /usr/src/usr.sbin/mlx5tool/mlx5tool.c:31:10: fatal error: 
>> 'dev/mlx5/mlx5io.h' file not found
>> #include <dev/mlx5/mlx5io.h>
>>           ^~~~~~~~~~~~~~~~~~~
>> 1 error generated.
>> --- mlx5tool.o ---
>> *** [mlx5tool.o] Error code 1
>>
> 
> Hi,
> 
> All the bits and pieces needed are part of:
> 
> https://svnweb.freebsd.org/changeset/base/331586
> 
> I'll dig into it.
> 
>> See also:
>>
>> https://ci.freebsd.org/job/FreeBSD-stable-11-amd64-build/2930/console
>>

Hi,

11-stable amd64 buildworld should be back to normal as of:
https://svnweb.freebsd.org/changeset/base/331601

Tested both buildworld and installworld.

It might have been that the mlx5io.h header file was already installed 
locally on my build box.

--HPS