[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <87ftfimbjy.fsf@linux.ibm.com>
Date: Mon, 10 Feb 2020 14:28:49 -0600
From: Nathan Lynch <nathanl@...ux.ibm.com>
To: Scott Cheloha <cheloha@...ux.ibm.com>
Cc: Nathan Fontenont <ndfont@...il.com>,
Rick Lindsley <ricklind@...ux.vnet.ibm.com>,
linux-kernel@...r.kernel.org, Michael Ellerman <mpe@...erman.id.au>
Subject: Re: [PATCH] pseries/hotplug-memory: remove dlpar_memory_{add,remove}_by_index() functions
Scott Cheloha <cheloha@...ux.ibm.com> writes:
> The dlpar_memory_{add,remove}_by_index() functions are just special
> cases of their dlpar_memory_{add,remove}_by_ic() counterparts where
> the LMB count is 1.
I wish that were the case, but there are (gratuitous?) differences:
- dlpar_memory_remove_by_ic() checks DRCONF_MEM_RESERVED and
DRCONF_MEM_ASSIGNED flags; dlpar_memory_remove_by_index() does not.
- dlpar_memory_remove_by_ic() attempts to roll back failed removal;
dlpar_memory_remove_by_index() does not.
I'm not sure how much either of these gets used in practice. AFAIK the
usual HMC/drmgr-driven workflow tends to exercise
dlpar_memory_remove_by_count().
I agree this code needs consolidation, but we should proceed a little
carefully because it's likely going to entail changing some user-visible
behaviors.
Powered by blists - more mailing lists