Re: dev_pagemap related cleanups v4

From: Dan Williams
Date: Tue Jul 02 2019 - 20:59:32 EST

On Tue, Jul 2, 2019 at 11:42 AM Jason Gunthorpe <jgg@xxxxxxxxxxxx> wrote:
> On Mon, Jul 01, 2019 at 10:25:17AM +0200, Christoph Hellwig wrote:
> > And I've demonstrated that I can't send patch series.. While this
> > has all the right patches, it also has the extra patches already
> > in the hmm tree, and four extra patches I wanted to send once
> > this series is merged. I'll give up for now, please use the git
> > url for anything serious, as it contains the right thing.
> Okay, I sorted it all out and temporarily put it here:
> Bit involved job:
> - Took Ira's v4 patch into hmm.git and confirmed it matches what
> Andrew has in linux-next after all the fixups
> - Checked your github v4 and the v3 that hit the mailing list were
> substantially similar (I never did get a clean v4) and largely
> went with the github version
> - Based CH's v4 series on -rc7 and put back the removal hunk in swap.c
> so it compiles
> - Merge'd CH's series to hmm.git and fixed all the conflicts with Ira
> and Ralph's patches (such that swap.c remains unchanged)
> - Added Dan's ack's and tested-by's

Looks good. Test merge (with some collisions, see below) also passes
my test suite.

> I think this fairly closely follows what was posted to the mailing
> list.
> As it was more than a simple 'git am', I'll let it sit on github until
> I hear OK's then I'll move it to's hmm.git and it will hit
> linux-next. 0-day should also run on this whole thing from my github.
> What I know is outstanding:
> - The conflicting ARM patches, I understand Andrew will handle these
> post-linux-next
> - The conflict with AMD GPU in -next, I am waiting to hear from AMD

Just a heads up that this also collides with the "sub-section" patches
in Andrew's tree. The resolution is straightforward, mostly just
colliding updates to arch_{add,remove}_memory() call sites in
kernel/memremap.c and collisions with pgmap_altmap() usage.