Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Thu, 4 Jun 2020 13:22:13 -0400
From:   Daniel Jordan <daniel.m.jordan@oracle.com>
To:     David Hildenbrand <david@redhat.com>
Cc:     Daniel Jordan <daniel.m.jordan@oracle.com>, linux-mm@kvack.org,
        linux-kernel@vger.kernel.org,
        Andrew Morton <akpm@linux-foundation.org>,
        Andy Lutomirski <luto@kernel.org>,
        Dave Hansen <dave.hansen@linux.intel.com>,
        Michal Hocko <mhocko@kernel.org>,
        Pavel Tatashin <pasha.tatashin@soleen.com>,
        Peter Zijlstra <peterz@infradead.org>,
        Steven Sistare <steven.sistare@oracle.com>
Subject: Re: [PATCH] x86/mm: use max memory block size with unaligned memory
 end
Message-ID: <20200604172213.f5lufktpqvqjkv4u@ca-dmjordan1.us.oracle.com>
References: <20200604035443.3267046-1-daniel.m.jordan@oracle.com>
 <5827baaf-0eb5-bcea-5d98-727485683512@redhat.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <5827baaf-0eb5-bcea-5d98-727485683512@redhat.com>
User-Agent: NeoMutt/20180716
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Thu, Jun 04, 2020 at 09:22:03AM +0200, David Hildenbrand wrote:
> On 04.06.20 05:54, Daniel Jordan wrote:
> > Some of our servers spend 14 out of the 21 seconds of kernel boot
> > initializing memory block sysfs directories and then creating symlinks
> > between them and the corresponding nodes.  The slowness happens because
> > the machines get stuck with the smallest supported memory block size on
> > x86 (128M), which results in 16,288 directories to cover the 2T of
> > installed RAM, and each of these paths does a linear search of the
> > memory blocks for every block id, with atomic ops at each step.
> 
> With 4fb6eabf1037 ("drivers/base/memory.c: cache memory blocks in xarray
> to accelerate lookup") merged by Linus' today (strange, I thought this
> would be long upstream)

Ah, thanks for pointing this out!  It was only posted to LKML so I missed it.

> all linear searches should be gone and at least
> the performance observation in this patch no longer applies.

The performance numbers as stated, that's certainly true, but this patch on top
still improves kernel boot by 7%.  It's a savings of half a second -- I'll take
it.

IMHO the root cause of this is really the small block size.  Building a cache
on top to avoid iterating over tons of small blocks seems like papering over
the problem, especially when one of the two affected paths in boot is a
cautious check that might be ready to be removed by now[0]:

    static int init_memory_block(struct memory_block **memory,
    			     unsigned long block_id, unsigned long state)
    {
            ...
    	mem = find_memory_block_by_id(block_id);
    	if (mem) {
    		put_device(&mem->dev);
    		return -EEXIST;
    	}

Anyway, I guess I'll redo the changelog and post again.

> The memmap init should nowadays consume most time.

Yeah, but of course it's not as bad as it was now that it's fully parallelized.

[0] https://lore.kernel.org/linux-mm/a8e96df6-dc6d-037f-491c-92182d4ada8d@redhat.com/