Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <158500767138.2088294.17131646259803932461.stgit@dwillia2-desk3.amr.corp.intel.com>
 <158500773552.2088294.8756587190550753100.stgit@dwillia2-desk3.amr.corp.intel.com>
 <23742bb8-831f-29ff-1463-75427eec57c7@oracle.com> <CAPcyv4jvsBSD3SELYn=oj_nvLZ_VgM2YbyswbSEeq5DmkcbXcw@mail.gmail.com>
In-Reply-To: <CAPcyv4jvsBSD3SELYn=oj_nvLZ_VgM2YbyswbSEeq5DmkcbXcw@mail.gmail.com>
From:   Dan Williams <dan.j.williams@intel.com>
Date:   Fri, 10 Jul 2020 17:47:17 -0700
Message-ID: <CAPcyv4joMJDjaTLSw=HiU_sesWrHoVt07fOuWM44726TpjWCCA@mail.gmail.com>
Subject: Re: [PATCH 11/12] device-dax: Add dis-contiguous resource support
To:     Joao Martins <joao.m.martins@oracle.com>
Cc:     Linux MM <linux-mm@kvack.org>,
        Dave Hansen <dave.hansen@linux.intel.com>,
        Christoph Hellwig <hch@lst.de>,
        linux-nvdimm <linux-nvdimm@lists.01.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Mon, Apr 6, 2020 at 1:22 PM Dan Williams <dan.j.williams@intel.com> wrote:
>
> On Mon, Apr 6, 2020 at 3:46 AM Joao Martins <joao.m.martins@oracle.com> wrote:
> >
> > On 3/23/20 11:55 PM, Dan Williams wrote:
> >
> > [...]
> >
> > >  static ssize_t dev_dax_resize(struct dax_region *dax_region,
> > >               struct dev_dax *dev_dax, resource_size_t size)
> > >  {
> > >       resource_size_t avail = dax_region_avail_size(dax_region), to_alloc;
> > > -     resource_size_t dev_size = range_len(&dev_dax->range);
> > > +     resource_size_t dev_size = dev_dax_size(dev_dax);
> > >       struct resource *region_res = &dax_region->res;
> > >       struct device *dev = &dev_dax->dev;
> > > -     const char *name = dev_name(dev);
> > >       struct resource *res, *first;
> > > +     resource_size_t alloc = 0;
> > > +     int rc;
> > >
> > >       if (dev->driver)
> > >               return -EBUSY;
> > > @@ -684,38 +766,47 @@ static ssize_t dev_dax_resize(struct dax_region *dax_region,
> > >        * allocating a new resource.
> > >        */
> > >       first = region_res->child;
> > > -     if (!first)
> > > -             return __alloc_dev_dax_range(dev_dax, dax_region->res.start,
> > > -                             to_alloc);
> > > -     for (res = first; to_alloc && res; res = res->sibling) {
> > > +retry:
> > > +     rc = -ENOSPC;
> > > +     for (res = first; res; res = res->sibling) {
> > >               struct resource *next = res->sibling;
> > > -             resource_size_t free;
> > >
> > >               /* space at the beginning of the region */
> > > -             free = 0;
> > > -             if (res == first && res->start > dax_region->res.start)
> > > -                     free = res->start - dax_region->res.start;
> > > -             if (free >= to_alloc && dev_size == 0)
> > > -                     return __alloc_dev_dax_range(dev_dax,
> > > -                                     dax_region->res.start, to_alloc);
> > > -
> > > -             free = 0;
> > > +             if (res == first && res->start > dax_region->res.start) {
> > > +                     alloc = min(res->start - dax_region->res.start,
> > > +                                     to_alloc);
> > > +                     rc = __alloc_dev_dax_range(dev_dax,
> > > +                                     dax_region->res.start, alloc);
> >
> > You might be missing:
> >
> >         first = region_res->child;
> >
> > (...) right after returning from __alloc_dev_dax_range(). Alternatively, perhaps
> > even moving the 'retry' label to right before the @first initialization.
> >
> > In the case that you pick space from the beginning, the child resource of the
> > dax region will point to first occupied region, and that changes after you pick
> > this space. So, IIUC, you want to adjust where you start searching free space
> > otherwise you end up wrongly picking that same space twice.
> >
> > If it helps, the bug can be reproduced in this unit test below, see
> > daxctl_test3() test:
>
> It definitely will, thanks. I'll be circling back to this now that
> I've settled my tree for the v5.7 window.

s/v5.7/v5.9/ whats a couple of kernel release cycles between friends?
I went ahead and moved the retry loop above the assignment of first as
you suggested.