Re: [PATCH] dax: Change return type to vm_fault_t

[Souptick Joarder <jrdr.linux@gmail.com>]

On Tue, Apr 17, 2018 at 12:05 AM, Dan Williams <dan.j.williams@intel.com> wrote:
> On Mon, Apr 16, 2018 at 11:28 AM, Souptick Joarder <jrdr.linux@gmail.com> wrote:
>> On Mon, Apr 16, 2018 at 11:51 PM, Matthew Wilcox <willy@infradead.org> wrote:
>>> On Mon, Apr 16, 2018 at 11:00:26AM -0700, Dan Williams wrote:
>>>> On Mon, Apr 16, 2018 at 10:47 AM, Matthew Wilcox <willy@infradead.org> wrote:
>>>> > On Mon, Apr 16, 2018 at 09:14:48AM -0700, Dan Williams wrote:
>>>> >> > -       rc = vm_insert_mixed(vmf->vma, vmf->address, pfn);
>>>> >> > -
>>>> >> > -       if (rc == -ENOMEM)
>>>> >> > -               return VM_FAULT_OOM;
>>>> >> > -       if (rc < 0 && rc != -EBUSY)
>>>> >> > -               return VM_FAULT_SIGBUS;
>>>> >> > -
>>>> >> > -       return VM_FAULT_NOPAGE;
>>>> >> > +       return vmf_insert_mixed(vmf->vma, vmf->address, pfn);
>>>> >>
>>>> >> Ugh, so this change to vmf_insert_mixed() went upstream without fixing
>>>> >> the users? This changelog is now misleading as it does not mention
>>>> >> that is now an urgent standalone fix. On first read I assumed this was
>>>> >> part of a wider effort for 4.18.
>>>> >
>>>> > You read too quickly.  vmf_insert_mixed() is a *new* function which
>>>> > *replaces* vm_insert_mixed() and
>>>> > awful-mangling-of-return-values-done-per-driver.
>>>> >
>>>> > Eventually vm_insert_mixed() will be deleted.  But today is not that day.
>>>>
>>>> Ah, ok, thanks for the clarification. Then this patch should
>>>> definitely be re-titled to "dax: convert to the new vmf_insert_mixed()
>>>> helper". The vm_fault_t conversion is just a minor side-effect of that
>>>> larger change. I assume this can wait for v4.18.
>>
>> The primary objective is to change the return type to
>> vm_fault_t in all fault handlers and to support that
>> we have replace vm_insert_mixed() with vmf_insert_
>> mixed() within one fault handler function.
>>
>> Do I really need to change the patch title ?
>
> At this point, yes, or at least mention the vm_insert_mixed -->
> vmf_insert_mixed conversion in the changelog.

Ok, I will add this in change log and send v2.