[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <b478e564-fa10-79de-f9df-5b55b211e85d@suse.cz>
Date: Wed, 10 May 2017 15:09:37 +0200
From: Vlastimil Babka <vbabka@...e.cz>
To: SeongJae Park <sj38.park@...il.com>, akpm@...ux-foundation.org,
kirill.shutemov@...ux.intel.com
Cc: hannes@...xchg.org, linux-mm@...ck.org,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH] mm/khugepaged: Add missed tracepoint for
collapse_huge_page_swapin
On 05/07/2017 12:18 PM, SeongJae Park wrote:
> One return case of `__collapse_huge_page_swapin()` does not invoke
> tracepoint while every other return case does. This commit adds a
> tracepoint invocation for the case.
>
> Signed-off-by: SeongJae Park <sj38.park@...il.com>
Right. But extra points by turning all of the "trace+return false"
instances into some kind of "goto out".
> ---
> mm/khugepaged.c | 4 +++-
> 1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/mm/khugepaged.c b/mm/khugepaged.c
> index ba40b7f673f4..9aad377c67a8 100644
> --- a/mm/khugepaged.c
> +++ b/mm/khugepaged.c
> @@ -909,8 +909,10 @@ static bool __collapse_huge_page_swapin(struct mm_struct *mm,
> return false;
> }
> /* check if the pmd is still valid */
> - if (mm_find_pmd(mm, address) != pmd)
> + if (mm_find_pmd(mm, address) != pmd) {
> + trace_mm_collapse_huge_page_swapin(mm, swapped_in, referenced, 0);
> return false;
> + }
> }
> if (ret & VM_FAULT_ERROR) {
> trace_mm_collapse_huge_page_swapin(mm, swapped_in, referenced, 0);
>
Powered by blists - more mailing lists