Received-SPF: pass (google.com: domain of linux-kernel+bounces-54565-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45d1:ec00::1 as permitted sender) client-ip=2604:1380:45d1:ec00::1;
Precedence: bulk
MIME-Version: 1.0
References: <20240131094442.28834-1-fangzheng.zhang@unisoc.com>
 <2024020441-version-chihuahua-1067@gregkh> <2f43774a-16cf-412e-b4b7-840889bb796b@suse.cz>
In-Reply-To: <2f43774a-16cf-412e-b4b7-840889bb796b@suse.cz>
From: zhang fangzheng <fangzheng.zhang1003@gmail.com>
Date: Tue, 6 Feb 2024 17:08:53 +0800
Message-ID: <CA+kNDJJAQ8FP5U+SJWRAvhqEfbmmqkNpOKVk9SC66q-auCPg0A@mail.gmail.com>
Subject: Re: [PATCH] mm/slab: Add slabreclaim flag to slabinfo
To: Vlastimil Babka <vbabka@suse.cz>
Cc: Greg KH <gregkh@linuxfoundation.org>, 
	Fangzheng Zhang <fangzheng.zhang@unisoc.com>, Christoph Lameter <cl@linux.com>, 
	Pekka Enberg <penberg@kernel.org>, David Rientjes <rientjes@google.com>, 
	Joonsoo Kim <iamjoonsoo.kim@lge.com>, Andrew Morton <akpm@linux-foundation.org>, 
	Roman Gushchin <roman.gushchin@linux.dev>, Hyeonggon Yoo <42.hyeyoo@gmail.com>, linux-mm@kvack.org, 
	linux-kernel@vger.kernel.org, Yuming Han <yuming.han@unisoc.com>, 
	Chunyan Zhang <zhang.lyra@gmail.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Mon, Feb 5, 2024 at 4:50=E2=80=AFPM Vlastimil Babka <vbabka@suse.cz> wro=
te:
>
> On 2/4/24 14:09, Greg KH wrote:
> > On Wed, Jan 31, 2024 at 05:44:42PM +0800, Fangzheng Zhang wrote:
> >> In order to enhance slab debugging, we add slabreclaim flag to
> >> slabinfo. Slab type is also an important analysis point in slabinfo
> >> for per slab, when various problems such as memory leaks or memory
> >> statistics occur.
> >>
> >> Signed-off-by: Fangzheng Zhang <fangzheng.zhang@unisoc.com>
> >> ---
> >>  mm/slab_common.c | 7 ++++---
> >>  1 file changed, 4 insertions(+), 3 deletions(-)
> >>
> >> diff --git a/mm/slab_common.c b/mm/slab_common.c
> >> index 238293b1dbe1..aeeb2bfe6dda 100644
> >> --- a/mm/slab_common.c
> >> +++ b/mm/slab_common.c
> >> @@ -1038,7 +1038,7 @@ static void print_slabinfo_header(struct seq_fil=
e *m)
> >>      seq_puts(m, "slabinfo - version: 2.1\n");
> >>      seq_puts(m, "# name            <active_objs> <num_objs> <objsize>=
 <objperslab> <pagesperslab>");
> >>      seq_puts(m, " : tunables <limit> <batchcount> <sharedfactor>");
> >> -    seq_puts(m, " : slabdata <active_slabs> <num_slabs> <sharedavail>=
");
> >> +    seq_puts(m, " : slabdata <active_slabs> <num_slabs> <sharedavail>=
 <slabreclaim>");
> >
> > Doesn't this change the slabinfo version number above?  Where is this
> > change documented so that userspace knows about it?
>
Ok, I will  modify the slabinfo version number to 2.2 and I find that
there is no corresponding slabinfo output example in the proc.rst
document. Can we add an output example so that user space knows about
it?

> Yeah I was vary about this. Do the other longer-time-than-me slab
> maintainers recall how we handled this in the past?
> Also the information is already available, even if harder to gather, in t=
he
> file /sys/kernel/slab/$cache/reclaim_account
>
First  of all, thank you very much for your comments.

I would like to say, when performing slab memory information
maintenance, people often hope to see more detailed information
through a simple slabinfo command. As you mentioned the method, but it
is very unintuitive to the status of the entire slab, so we add the
slabreclaim column to slabinfo and directly output it using cmdline '
> cat proc/slabinfo'. And I think this approach will also be helpful
for future work on memory statistics.

> > thanks,
> >
> > greg k-h
>
Thanks.