linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Daniel Vacek <neelx@redhat.com>
To: Jia He <hejianet@gmail.com>
Cc: Andrew Morton <akpm@linux-foundation.org>,
	Michal Hocko <mhocko@suse.com>,
	Catalin Marinas <catalin.marinas@arm.com>,
	Mel Gorman <mgorman@suse.de>, Will Deacon <will.deacon@arm.com>,
	Mark Rutland <mark.rutland@arm.com>,
	Ard Biesheuvel <ard.biesheuvel@linaro.org>,
	Thomas Gleixner <tglx@linutronix.de>,
	Ingo Molnar <mingo@redhat.com>, "H. Peter Anvin" <hpa@zytor.com>,
	Pavel Tatashin <pasha.tatashin@oracle.com>,
	Daniel Jordan <daniel.m.jordan@oracle.com>,
	AKASHI Takahiro <takahiro.akashi@linaro.org>,
	Gioh Kim <gi-oh.kim@profitbricks.com>,
	Steven Sistare <steven.sistare@oracle.com>,
	Eugeniu Rosca <erosca@de.adit-jv.com>,
	Vlastimil Babka <vbabka@suse.cz>,
	open list <linux-kernel@vger.kernel.org>,
	linux-mm@kvack.org, James Morse <james.morse@arm.com>,
	Steve Capper <steve.capper@arm.com>,
	x86@kernel.org, Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	Kate Stewart <kstewart@linuxfoundation.org>,
	Philippe Ombredanne <pombredanne@nexb.com>,
	Johannes Weiner <hannes@cmpxchg.org>,
	Kemi Wang <kemi.wang@intel.com>, Petr Tesarik <ptesarik@suse.com>,
	YASUAKI ISHIMATSU <yasu.isimatu@gmail.com>,
	Andrey Ryabinin <aryabinin@virtuozzo.com>,
	Nikolay Borisov <nborisov@suse.com>,
	Jia He <jia.he@hxt-semitech.com>
Subject: Re: [PATCH v3 2/5] mm: page_alloc: reduce unnecessary binary search in memblock_next_valid_pfn()
Date: Tue, 27 Mar 2018 19:17:38 +0200	[thread overview]
Message-ID: <CACjP9X-+gBPy=w6YAjO_2=WEz0hS-FGFvhSJjJas8n9n3xkpew@mail.gmail.com> (raw)
In-Reply-To: <1522033340-6575-3-git-send-email-hejianet@gmail.com>

On Mon, Mar 26, 2018 at 5:02 AM, Jia He <hejianet@gmail.com> wrote:
> Commit b92df1de5d28 ("mm: page_alloc: skip over regions of invalid pfns
> where possible") optimized the loop in memmap_init_zone(). But there is
> still some room for improvement. E.g. if pfn and pfn+1 are in the same
> memblock region, we can simply pfn++ instead of doing the binary search
> in memblock_next_valid_pfn. This patch only works when
> CONFIG_HAVE_ARCH_PFN_VALID is enable.
>
> Signed-off-by: Jia He <jia.he@hxt-semitech.com>
> ---
>  include/linux/memblock.h |  2 +-
>  mm/memblock.c            | 73 +++++++++++++++++++++++++++++-------------------
>  mm/page_alloc.c          |  3 +-
>  3 files changed, 47 insertions(+), 31 deletions(-)
>
> diff --git a/include/linux/memblock.h b/include/linux/memblock.h
> index efbbe4b..a8fb2ab 100644
> --- a/include/linux/memblock.h
> +++ b/include/linux/memblock.h
> @@ -204,7 +204,7 @@ void __next_mem_pfn_range(int *idx, int nid, unsigned long *out_start_pfn,
>  #endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */
>
>  #ifdef CONFIG_HAVE_ARCH_PFN_VALID
> -unsigned long memblock_next_valid_pfn(unsigned long pfn);
> +unsigned long memblock_next_valid_pfn(unsigned long pfn, int *idx);
>  #endif
>
>  /**
> diff --git a/mm/memblock.c b/mm/memblock.c
> index bea5a9c..06c1a08 100644
> --- a/mm/memblock.c
> +++ b/mm/memblock.c
> @@ -1102,35 +1102,6 @@ void __init_memblock __next_mem_pfn_range(int *idx, int nid,
>                 *out_nid = r->nid;
>  }
>
> -#ifdef CONFIG_HAVE_ARCH_PFN_VALID
> -unsigned long __init_memblock memblock_next_valid_pfn(unsigned long pfn)
> -{
> -       struct memblock_type *type = &memblock.memory;
> -       unsigned int right = type->cnt;
> -       unsigned int mid, left = 0;
> -       phys_addr_t addr = PFN_PHYS(++pfn);
> -
> -       do {
> -               mid = (right + left) / 2;
> -
> -               if (addr < type->regions[mid].base)
> -                       right = mid;
> -               else if (addr >= (type->regions[mid].base +
> -                                 type->regions[mid].size))
> -                       left = mid + 1;
> -               else {
> -                       /* addr is within the region, so pfn is valid */
> -                       return pfn;
> -               }
> -       } while (left < right);
> -
> -       if (right == type->cnt)
> -               return -1UL;
> -       else
> -               return PHYS_PFN(type->regions[right].base);
> -}
> -#endif /*CONFIG_HAVE_ARCH_PFN_VALID*/
> -
>  /**
>   * memblock_set_node - set node ID on memblock regions
>   * @base: base of area to set node ID for
> @@ -1162,6 +1133,50 @@ int __init_memblock memblock_set_node(phys_addr_t base, phys_addr_t size,
>  }
>  #endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */
>
> +#ifdef CONFIG_HAVE_ARCH_PFN_VALID
> +unsigned long __init_memblock memblock_next_valid_pfn(unsigned long pfn,
> +                                                       int *last_idx)
> +{
> +       struct memblock_type *type = &memblock.memory;
> +       unsigned int right = type->cnt;
> +       unsigned int mid, left = 0;
> +       unsigned long start_pfn, end_pfn;
> +       phys_addr_t addr = PFN_PHYS(++pfn);
> +
> +       /* fast path, return pfh+1 if next pfn is in the same region */
> +       if (*last_idx != -1) {
> +               start_pfn = PFN_DOWN(type->regions[*last_idx].base);
> +               end_pfn = PFN_DOWN(type->regions[*last_idx].base +
> +                               type->regions[*last_idx].size);
> +
> +               if (pfn < end_pfn && pfn > start_pfn)
> +                       return pfn;
> +       }
> +
> +       /* slow path, do the binary searching */
> +       do {
> +               mid = (right + left) / 2;
> +
> +               if (addr < type->regions[mid].base)
> +                       right = mid;
> +               else if (addr >= (type->regions[mid].base +
> +                                 type->regions[mid].size))
> +                       left = mid + 1;
> +               else {
> +                       *last_idx = mid;
> +                       return pfn;
> +               }
> +       } while (left < right);
> +
> +       if (right == type->cnt)
> +               return -1UL;
> +
> +       *last_idx = right;
> +
> +       return PHYS_PFN(type->regions[*last_idx].base);
> +}
> +#endif /*CONFIG_HAVE_ARCH_PFN_VALID*/
> +
>  static phys_addr_t __init memblock_alloc_range_nid(phys_addr_t size,
>                                         phys_addr_t align, phys_addr_t start,
>                                         phys_addr_t end, int nid, ulong flags)
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index 2a967f7..0bb0274 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -5459,6 +5459,7 @@ void __meminit memmap_init_zone(unsigned long size, int nid, unsigned long zone,
>         unsigned long end_pfn = start_pfn + size;
>         pg_data_t *pgdat = NODE_DATA(nid);
>         unsigned long pfn;
> +       int idx = -1;
>         unsigned long nr_initialised = 0;
>         struct page *page;
>  #ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP
> @@ -5490,7 +5491,7 @@ void __meminit memmap_init_zone(unsigned long size, int nid, unsigned long zone,
>                          * end_pfn), such that we hit a valid pfn (or end_pfn)
>                          * on our next iteration of the loop.
>                          */
> -                       pfn = memblock_next_valid_pfn(pfn) - 1;
> +                       pfn = memblock_next_valid_pfn(pfn, &idx) - 1;
>  #endif
>                         continue;
>                 }
> --
> 2.7.4
>

So the function is only defined with CONFIG_HAVE_ARCH_PFN_VALID but
it's called with CONFIG_HAVE_MEMBLOCK_NODE_MAP? The definition should
likely depend on both options as the function really depends on both
conditions. Otherwise it should be defined nop.

--nX

  reply	other threads:[~2018-03-27 17:17 UTC|newest]

Thread overview: 25+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-03-26  3:02 [PATCH v3 0/5] optimize memblock_next_valid_pfn and early_pfn_valid Jia He
2018-03-26  3:02 ` [PATCH v3 1/5] mm: page_alloc: remain memblock_next_valid_pfn() when CONFIG_HAVE_ARCH_PFN_VALID is enable Jia He
2018-03-28  9:18   ` Wei Yang
2018-03-28  9:49     ` Jia He
2018-04-02  8:12       ` Wei Yang
2018-04-02  9:17         ` Jia He
2018-04-03  0:14           ` Wei Yang
2018-03-26  3:02 ` [PATCH v3 2/5] mm: page_alloc: reduce unnecessary binary search in memblock_next_valid_pfn() Jia He
2018-03-27 17:17   ` Daniel Vacek [this message]
2018-03-28  2:09     ` Jia He
2018-03-28  9:26   ` Wei Yang
2018-03-29  8:06     ` Jia He
2018-03-30  1:43       ` Wei Yang
2018-03-30  2:12         ` Jia He
2018-03-26  3:02 ` [PATCH v3 3/5] mm/memblock: introduce memblock_search_pfn_regions() Jia He
2018-03-26  3:02 ` [PATCH v3 4/5] arm64: introduce pfn_valid_region() Jia He
2018-03-28  9:38   ` Wei Yang
2018-03-26  3:02 ` [PATCH v3 5/5] mm: page_alloc: reduce unnecessary binary search in early_pfn_valid() Jia He
2018-03-27 17:51   ` Daniel Vacek
2018-03-28  2:10     ` Jia He
2018-03-27  1:02 ` [PATCH v3 0/5] optimize memblock_next_valid_pfn and early_pfn_valid Wei Yang
2018-03-27  7:15   ` Jia He
2018-03-28  0:30     ` Wei Yang
2018-03-28  1:45       ` Jia He
2018-03-28  2:36         ` Wei Yang

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to='CACjP9X-+gBPy=w6YAjO_2=WEz0hS-FGFvhSJjJas8n9n3xkpew@mail.gmail.com' \
    --to=neelx@redhat.com \
    --cc=akpm@linux-foundation.org \
    --cc=ard.biesheuvel@linaro.org \
    --cc=aryabinin@virtuozzo.com \
    --cc=catalin.marinas@arm.com \
    --cc=daniel.m.jordan@oracle.com \
    --cc=erosca@de.adit-jv.com \
    --cc=gi-oh.kim@profitbricks.com \
    --cc=gregkh@linuxfoundation.org \
    --cc=hannes@cmpxchg.org \
    --cc=hejianet@gmail.com \
    --cc=hpa@zytor.com \
    --cc=james.morse@arm.com \
    --cc=jia.he@hxt-semitech.com \
    --cc=kemi.wang@intel.com \
    --cc=kstewart@linuxfoundation.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=mark.rutland@arm.com \
    --cc=mgorman@suse.de \
    --cc=mhocko@suse.com \
    --cc=mingo@redhat.com \
    --cc=nborisov@suse.com \
    --cc=pasha.tatashin@oracle.com \
    --cc=pombredanne@nexb.com \
    --cc=ptesarik@suse.com \
    --cc=steve.capper@arm.com \
    --cc=steven.sistare@oracle.com \
    --cc=takahiro.akashi@linaro.org \
    --cc=tglx@linutronix.de \
    --cc=vbabka@suse.cz \
    --cc=will.deacon@arm.com \
    --cc=x86@kernel.org \
    --cc=yasu.isimatu@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox