Re: [RFC v1 1/2] mm/gup: fixup for 9947ea2c1e608e32 "mm/gup: track FOLL_PIN pages"
From: John Hubbard <jhubbard@nvidia.com>
Date: 2020-02-28 23:08:40
Also in:
linux-mm, linux-s390, lkml
On 2/28/20 7:43 AM, Claudio Imbrenda wrote:
quoted hunk ↗ jump to hunk
In case pin fails, we need to unpin, a simple put_page will not be enough Signed-off-by: Claudio Imbrenda <imbrenda@linux.ibm.com> --- mm/gup.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)diff --git a/mm/gup.c b/mm/gup.c index f589299b0d4a..0b9a806898f3 100644 --- a/mm/gup.c +++ b/mm/gup.c@@ -2134,7 +2134,10 @@ static int gup_pte_range(pmd_t pmd, unsigned long addr, unsigned long end, goto pte_unmap; if (unlikely(pte_val(pte) != pte_val(*ptep))) { - put_page(head); + if (flags & FOLL_GET) + put_page(head); + else if (flags & FOLL_PIN) + unpin_user_page(head);
Hi Claudio,
Instead, I think that should actually be:
put_compound_head(page, 1, flags);
which does a bit more (bug checks and /proc/vmstat instrumentation) than your diff,
but has the same basic idea: call the right "put" function.
...oh, actually, I see you have the commit hash in the subject line. Instead, it should
be in the commit description. Let's maybe change the subject line to approx:
mm/gup: Fix a missing put_compound_head() call in gup_pte_range()
And the write up...how about something like this, if you like it:
try_grab_compound_head() must be undone via put_compound_head(), not put_page().
This was missed in the original implementation of the gup/dma tracking patch, so
fix it now.
Fixes: 0ea2781c3de4 ("mm/gup: track FOLL_PIN pages")
(Aside: I'm using the linux-next commit hash. How does one get the correct hash before
it goes to mainline? I guess maintainer scripts fix all those up?)
It's also good to Cc some reviewers in case I'm overlooking something, so I'm adding
Jan and Kirill.
thanks,
--
John Hubbard
NVIDIAgoto pte_unmap; }