Received-SPF: pass (google.com: domain of linux-kernel+bounces-75242-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45e3:2400::1 as permitted sender) client-ip=2604:1380:45e3:2400::1;
Precedence: bulk
MIME-Version: 1.0
References: <20240220231424.126600-1-vishal.moola@gmail.com>
 <20240220231424.126600-3-vishal.moola@gmail.com> <ZdVx_Jr9tm2lV0Ku@casper.infradead.org>
 <CAOzc2pzLbiEpMuBpX7xXhZqPc0S6ZMnziT2uUFsa0tZe6yJQ0g@mail.gmail.com> <ZdY5BfO3EdcnJM99@casper.infradead.org>
In-Reply-To: <ZdY5BfO3EdcnJM99@casper.infradead.org>
From: Vishal Moola <vishal.moola@gmail.com>
Date: Wed, 21 Feb 2024 10:02:43 -0800
Message-ID: <CAOzc2py+gFdcfSebCvkWWWjht-CTBO=O4iNTQ9xs+=Wd2Kf8Hg@mail.gmail.com>
Subject: Re: [PATCH 2/3] hugetlb: Use vmf_anon_prepare() instead of anon_vma_prepare()
To: Matthew Wilcox <willy@infradead.org>
Cc: linux-mm@kvack.org, linux-kernel@vger.kernel.org, 
	akpm@linux-foundation.org, muchun.song@linux.dev
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Wed, Feb 21, 2024 at 9:55=E2=80=AFAM Matthew Wilcox <willy@infradead.org=
> wrote:
>
> On Wed, Feb 21, 2024 at 09:15:51AM -0800, Vishal Moola wrote:
> > > >       unsigned long haddr =3D address & huge_page_mask(h);
> > > >       struct mmu_notifier_range range;
> > > > +     struct vm_fault vmf =3D {
> > > > +                             .vma =3D vma,
> > > > +                             .address =3D haddr,
> > > > +                             .real_address =3D address,
> > > > +                             .flags =3D flags,
> > > > +     };
> > >
> > > We don't usually indent quite so far.  One extra tab would be enough.
> > >
> > > Also, I thought we talked about creating the vmf in hugetlb_fault(),
> > > then passing it to hugetlb_wp() hugetlb_no_page() and handle_userfaul=
t()?
> > > Was there a reason to abandon that idea?
> >
> > No I haven't abandoned that idea, I intend to have a separate patchset =
to go
> > on top of this one - just keeping them separate since they are conceptu=
ally
> > different. I'm converting each function to use struct vm_fault first, t=
hen
> > shifting it to be passed throughout as an arguement while cleaning up t=
he
> > excess variables laying around. In a sense working bottom-up instead
> > of top-down.
>
> I think you'll find it less work to create it in hugetlb_fault()
> first.  ie patch 2 could be to hoist its creation from half-way down
> hugetlb_fault to the top of hugetlb_fault.  Patch 3 could pass it
> through hugetlb_no_page() to hugetlb_handle_userfault() and remove its
> creation there.  Now you've alreedy got it, and can make use of it in
> this patch which would be the new patch 4.

Ah I see, that way would definitely be a lot less work. I'll make that
change for this patchset in v2 then.

> If you want to do a cleanup patch afterwards, you could hoist the vmf
> creation all the way to handle_mm_fault() ;-)

Yeah, I was already looking at doing that in the future patchset :)