Re: get_zone_device_page() in get_page() and page_cache_get_speculative()

[Dan Williams]

On Sun, Apr 23, 2017 at 4:31 PM, Kirill A. Shutemov
<kirill@xxxxxxxxxxxxx> wrote:
> On Thu, Apr 20, 2017 at 02:46:51PM -0700, Dan Williams wrote:
>> On Sat, Mar 18, 2017 at 2:52 AM, tip-bot for Kirill A. Shutemov
>> <tipbot@xxxxxxxxx> wrote:
>> > Commit-ID:  2947ba054a4dabbd82848728d765346886050029
>> > Gitweb:     http://git.kernel.org/tip/2947ba054a4dabbd82848728d765346886050029
>> > Author:     Kirill A. Shutemov <kirill.shutemov@xxxxxxxxxxxxxxx>
>> > AuthorDate: Fri, 17 Mar 2017 00:39:06 +0300
>> > Committer:  Ingo Molnar <mingo@xxxxxxxxxx>
>> > CommitDate: Sat, 18 Mar 2017 09:48:03 +0100
>> >
>> > x86/mm/gup: Switch GUP to the generic get_user_page_fast() implementation
>> >
>> > This patch provides all required callbacks required by the generic
>> > get_user_pages_fast() code and switches x86 over - and removes
>> > the platform specific implementation.
>> >
>> > Signed-off-by: Kirill A. Shutemov <kirill.shutemov@xxxxxxxxxxxxxxx>
>> > Cc: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
>> > Cc: Aneesh Kumar K . V <aneesh.kumar@xxxxxxxxxxxxxxxxxx>
>> > Cc: Borislav Petkov <bp@xxxxxxxxx>
>> > Cc: Catalin Marinas <catalin.marinas@xxxxxxx>
>> > Cc: Dann Frazier <dann.frazier@xxxxxxxxxxxxx>
>> > Cc: Dave Hansen <dave.hansen@xxxxxxxxx>
>> > Cc: H. Peter Anvin <hpa@xxxxxxxxx>
>> > Cc: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
>> > Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
>> > Cc: Rik van Riel <riel@xxxxxxxxxx>
>> > Cc: Steve Capper <steve.capper@xxxxxxxxxx>
>> > Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
>> > Cc: linux-arch@xxxxxxxxxxxxxxx
>> > Cc: linux-mm@xxxxxxxxx
>> > Link: http://lkml.kernel.org/r/20170316213906.89528-1-kirill.shutemov@xxxxxxxxxxxxxxx
>> > [ Minor readability edits. ]
>> > Signed-off-by: Ingo Molnar <mingo@xxxxxxxxxx>
>>
>> I'm still trying to spot the bug, but bisect points to this patch as
>> the point at which my unit tests start failing with the following
>> signature:
>>
>> [   35.423841] WARNING: CPU: 8 PID: 245 at lib/percpu-refcount.c:155
>> percpu_ref_switch_to_atomic_rcu+0x1f5/0x200
>
> Okay, I've tracked it down. The issue is triggered by replacment
> get_page() with page_cache_get_speculative().
>
> page_cache_get_speculative() doesn't have get_zone_device_page(). :-|
>
> And I think it's your bug, Dan: it's wrong to have
> get_/put_zone_device_page() in get_/put_page(). I must be handled by
> page_ref_* machinery to catch all cases where we manipulate with page
> refcount.

The page_ref conversion landed in 4.6 *after* the ZONE_DEVICE
implementation that landed in 4.5, so there was a missed conversion of
the zone-device reference counting to page_ref.

> Back to the big picture:
>
> I hate that we need to have such additional code in page refcount
> primitives. I worked hard to remove compound page ugliness from there and
> now zone_device creeping in...
>
> Is it the only option?

Not sure, I need to spend some time to understand what page_ref means
to ZONE_DEVICE.