linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Sedat Dilek <sedat.dilek@googlemail.com>
To: Wu Fengguang <fengguang.wu@intel.com>
Cc: Andrew Morton <akpm@linux-foundation.org>,
	Jan Kara <jack@suse.cz>, Christoph Hellwig <hch@lst.de>,
	Trond Myklebust <Trond.Myklebust@netapp.com>,
	Dave Chinner <david@fromorbit.com>, Theodore Ts'o <tytso@mit.edu>,
	Chris Mason <chris.mason@oracle.com>,
	Peter Zijlstra <a.p.zijlstra@chello.nl>,
	Mel Gorman <mel@csn.ul.ie>, Rik van Riel <riel@redhat.com>,
	KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>,
	Greg Thelen <gthelen@google.com>,
	Minchan Kim <minchan.kim@gmail.com>,
	Vivek Goyal <vgoyal@redhat.com>,
	Andrea Righi <arighi@develer.com>,
	Balbir Singh <balbir@linux.vnet.ibm.com>,
	linux-mm <linux-mm@kvack.org>,
	"linux-fsdevel@vger.kernel.org" <linux-fsdevel@vger.kernel.org>,
	LKML <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH 00/12] IO-less dirty throttling v7
Date: Mon, 18 Apr 2011 12:22:30 +0200	[thread overview]
Message-ID: <BANLkTi=avvFU2XoFWstmNjVnqCknpO5f+Q@mail.gmail.com> (raw)
In-Reply-To: <20110418081808.GA10374@localhost>

[-- Attachment #1: Type: text/plain, Size: 11724 bytes --]

On Mon, Apr 18, 2011 at 10:18 AM, Wu Fengguang <fengguang.wu@intel.com> wrote:
>> Unfortunately, this "v2" patch still breaks with gcc-4.6 here:
>>
>>   LD      .tmp_vmlinux1
>> mm/built-in.o: In function `calc_period_shift.part.10':
>> page-writeback.c:(.text+0x6458): undefined reference to `____ilog2_NaN'
>> make[4]: *** [.tmp_vmlinux1] Error 1
>>
>> My patchset against next-20110415 looks like this:
>>
>>   (+) OK   writeback-dirty-throttling-v7/writeback-dirty-throttling-v7.patch
>>   (+) OK   writeback-dirty-throttling-post-v7/0001-writeback-i386-compile-fix.patch
>>   (+) OK   writeback-dirty-throttling-post-v7/0002-writeback-quick-CONFIG_BLK_DEV_THROTTLING-n-compile-.patch
>>   (+) OK   writeback-dirty-throttling-post-v7/0003-Revert-writeback-scale-dirty-proportions-period-with.patch
>>   (+) OK   writeback-dirty-throttling-v7-fix/writeback-scale-dirty-proportions-period-with-writeout-bandwidth-v2.patch
>>
>> Attached are the disasm of mm/page-writeback.o (v2, gcc-4.6) and the
>> disasm of yesterday's experiments with gcc-4.5.
>>
>> [ gcc-4.5 ]
>>
>> 00006574 <calc_period_shift>:
>>     6574:       a1 90 00 00 00          mov    0x90,%eax        6575:
>> R_386_32  default_backing_dev_info
>>     6579:       55                      push   %ebp
>>     657a:       89 e5                   mov    %esp,%ebp
>>     657c:       e8 02 f8 ff ff          call   5d83 <__ilog2_u32>
>>     6581:       5d                      pop    %ebp
>>     6582:       83 c0 02                add    $0x2,%eax
>>     6585:       c3                      ret
>
> gcc-4.5 is generating the right code, while the below silly macro
> expansion seems like __builtin_constant_p(n) wrongly evaluating to true.
> I wonder if the attached patch will workaround the gcc's bug. It adds a
> local variable in calc_period_shift() for passing to ilog2().
>
> Thanks,
> Fengguang
>

Thanks for the patch, that "gcc-4.6 workaround" works here.
Feel free to add:

       Tested-by: Sedat Dilek <sedat.dilek@gmail.com>.

- Sedat -

P.S.: Attached disasm v3 with gcc-4.6

>> [ gcc-4.6 ]
>>
>> 000008c9 <calc_period_shift.part.10>:
>>      8c9:       8b 15 90 00 00 00       mov    0x90,%edx        8cb:
>> R_386_32   default_backing_dev_info
>>      8cf:       55                      push   %ebp
>>      8d0:       89 e5                   mov    %esp,%ebp
>>      8d2:       85 d2                   test   %edx,%edx
>>      8d4:       0f 88 46 01 00 00       js     a20
>> <calc_period_shift.part.10+0x157>
>>      8da:       f7 c2 00 00 00 40       test   $0x40000000,%edx
>>      8e0:       b8 20 00 00 00          mov    $0x20,%eax
>>      8e5:       0f 85 3a 01 00 00       jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      8eb:       f7 c2 00 00 00 20       test   $0x20000000,%edx
>>      8f1:       b0 1f                   mov    $0x1f,%al
>>      8f3:       0f 85 2c 01 00 00       jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      8f9:       f7 c2 00 00 00 10       test   $0x10000000,%edx
>>      8ff:       b0 1e                   mov    $0x1e,%al
>>      901:       0f 85 1e 01 00 00       jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      907:       f7 c2 00 00 00 08       test   $0x8000000,%edx
>>      90d:       b0 1d                   mov    $0x1d,%al
>>      90f:       0f 85 10 01 00 00       jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      915:       f7 c2 00 00 00 04       test   $0x4000000,%edx
>>      91b:       b0 1c                   mov    $0x1c,%al
>>      91d:       0f 85 02 01 00 00       jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      923:       f7 c2 00 00 00 02       test   $0x2000000,%edx
>>      929:       b0 1b                   mov    $0x1b,%al
>>      92b:       0f 85 f4 00 00 00       jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      931:       f7 c2 00 00 00 01       test   $0x1000000,%edx
>>      937:       b0 1a                   mov    $0x1a,%al
>>      939:       0f 85 e6 00 00 00       jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      93f:       f7 c2 00 00 80 00       test   $0x800000,%edx
>>      945:       b0 19                   mov    $0x19,%al
>>      947:       0f 85 d8 00 00 00       jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      94d:       f7 c2 00 00 40 00       test   $0x400000,%edx
>>      953:       b0 18                   mov    $0x18,%al
>>      955:       0f 85 ca 00 00 00       jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      95b:       f7 c2 00 00 20 00       test   $0x200000,%edx
>>      961:       b0 17                   mov    $0x17,%al
>>      963:       0f 85 bc 00 00 00       jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      969:       f7 c2 00 00 10 00       test   $0x100000,%edx
>>      96f:       b0 16                   mov    $0x16,%al
>>      971:       0f 85 ae 00 00 00       jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      977:       f7 c2 00 00 08 00       test   $0x80000,%edx
>>      97d:       b0 15                   mov    $0x15,%al
>>      97f:       0f 85 a0 00 00 00       jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      985:       f7 c2 00 00 04 00       test   $0x40000,%edx
>>      98b:       b0 14                   mov    $0x14,%al
>>      98d:       0f 85 92 00 00 00       jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      993:       f7 c2 00 00 02 00       test   $0x20000,%edx
>>      999:       b0 13                   mov    $0x13,%al
>>      99b:       0f 85 84 00 00 00       jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      9a1:       f7 c2 00 00 01 00       test   $0x10000,%edx
>>      9a7:       b0 12                   mov    $0x12,%al
>>      9a9:       75 7a                   jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      9ab:       f6 c6 80                test   $0x80,%dh
>>      9ae:       b0 11                   mov    $0x11,%al
>>      9b0:       75 73                   jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      9b2:       f6 c6 40                test   $0x40,%dh
>>      9b5:       b0 10                   mov    $0x10,%al
>>      9b7:       75 6c                   jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      9b9:       f6 c6 20                test   $0x20,%dh
>>      9bc:       b0 0f                   mov    $0xf,%al
>>      9be:       75 65                   jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      9c0:       f6 c6 10                test   $0x10,%dh
>>      9c3:       b0 0e                   mov    $0xe,%al
>>      9c5:       75 5e                   jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      9c7:       f6 c6 08                test   $0x8,%dh
>>      9ca:       b0 0d                   mov    $0xd,%al
>>      9cc:       75 57                   jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      9ce:       f6 c6 04                test   $0x4,%dh
>>      9d1:       b0 0c                   mov    $0xc,%al
>>      9d3:       75 50                   jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      9d5:       f6 c6 02                test   $0x2,%dh
>>      9d8:       b0 0b                   mov    $0xb,%al
>>      9da:       75 49                   jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      9dc:       f6 c6 01                test   $0x1,%dh
>>      9df:       b0 0a                   mov    $0xa,%al
>>      9e1:       75 42                   jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      9e3:       f6 c2 80                test   $0x80,%dl
>>      9e6:       b0 09                   mov    $0x9,%al
>>      9e8:       75 3b                   jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      9ea:       f6 c2 40                test   $0x40,%dl
>>      9ed:       b0 08                   mov    $0x8,%al
>>      9ef:       75 34                   jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      9f1:       f6 c2 20                test   $0x20,%dl
>>      9f4:       b0 07                   mov    $0x7,%al
>>      9f6:       75 2d                   jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      9f8:       f6 c2 10                test   $0x10,%dl
>>      9fb:       b0 06                   mov    $0x6,%al
>>      9fd:       75 26                   jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      9ff:       f6 c2 08                test   $0x8,%dl
>>      a02:       b0 05                   mov    $0x5,%al
>>      a04:       75 1f                   jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      a06:       f6 c2 04                test   $0x4,%dl
>>      a09:       b0 04                   mov    $0x4,%al
>>      a0b:       75 18                   jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      a0d:       f6 c2 02                test   $0x2,%dl
>>      a10:       b0 03                   mov    $0x3,%al
>>      a12:       75 11                   jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      a14:       80 e2 01                and    $0x1,%dl
>>      a17:       b0 02                   mov    $0x2,%al
>>      a19:       75 0a                   jne    a25
>> <calc_period_shift.part.10+0x15c>
>>      a1b:       e8 fc ff ff ff          call   a1c
>> <calc_period_shift.part.10+0x153>    a1c: R_386_PC32 ____ilog2_NaN
>>      a20:       b8 21 00 00 00          mov    $0x21,%eax
>>      a25:       5d                      pop    %ebp
>>      a26:       c3                      ret
>>
>> 00000a27 <calc_period_shift>:
>>      a27:       55                      push   %ebp
>>      a28:       83 ca ff                or     $0xffffffff,%edx
>>      a2b:       89 e5                   mov    %esp,%ebp
>>      a2d:       0f bd 05 90 00 00 00    bsr    0x90,%eax        a30:
>> R_386_32   default_backing_dev_info
>>      a34:       0f 44 c2                cmove  %edx,%eax
>>      a37:       5d                      pop    %ebp
>>      a38:       83 c0 02                add    $0x2,%eax
>>      a3b:       c3                      ret
>>
>> - EOT -
>>
>> - Sedat -
>
>
>
>

[-- Attachment #2: mm_page-writeback.o_v3.disasm.xz --]
[-- Type: application/octet-stream, Size: 19716 bytes --]

  reply	other threads:[~2011-04-18 10:22 UTC|newest]

Thread overview: 29+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-04-16 13:25 Wu Fengguang
2011-04-16 13:25 ` [PATCH 01/12] writeback: account per-bdi accumulated written pages Wu Fengguang
2011-04-16 13:25 ` [PATCH 02/12] writeback: account per-bdi accumulated dirtied pages Wu Fengguang
2011-04-16 13:25 ` [PATCH 03/12] writeback: bdi write bandwidth estimation Wu Fengguang
2011-04-16 13:25 ` [PATCH 04/12] writeback: smoothed global/bdi dirty pages Wu Fengguang
2011-04-16 13:25 ` [PATCH 05/12] writeback: smoothed dirty threshold and limit Wu Fengguang
2011-04-16 13:25 ` [PATCH 06/12] writeback: enforce 1/4 gap between the dirty/background thresholds Wu Fengguang
2011-04-16 13:25 ` [PATCH 07/12] writeback: base throttle bandwidth and position ratio Wu Fengguang
2011-04-16 13:25 ` [PATCH 08/12] writeback: IO-less balance_dirty_pages() Wu Fengguang
2011-04-16 13:25 ` [PATCH 09/12] writeback: show bdi write bandwidth in debugfs Wu Fengguang
2011-04-16 13:25 ` [PATCH 10/12] writeback: trace dirty_ratelimit Wu Fengguang
2011-04-16 13:25 ` [PATCH 11/12] writeback: trace balance_dirty_pages Wu Fengguang
2011-04-16 13:25 ` [PATCH 12/12] writeback: trace global_dirty_state Wu Fengguang
2011-04-16 16:27 ` [PATCH 00/12] IO-less dirty throttling v7 Sedat Dilek
2011-04-17  1:44   ` Wu Fengguang
2011-04-17  3:18     ` Sedat Dilek
2011-04-17  4:10       ` Wu Fengguang
2011-04-17  4:46         ` Sedat Dilek
2011-04-17  6:46           ` Sedat Dilek
2011-04-18  0:13         ` Wu Fengguang
2011-04-18  6:57           ` Sedat Dilek
2011-04-18  8:18             ` Wu Fengguang
2011-04-18 10:22               ` Sedat Dilek [this message]
2011-04-17  7:31 ` Marco Stornelli
2011-04-17  9:30   ` Wu Fengguang
2011-04-17 17:44     ` Marco Stornelli
2011-04-17 23:31       ` Wu Fengguang
2011-04-26 17:19 ` Vivek Goyal
2011-04-28 14:27   ` Wu Fengguang

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='BANLkTi=avvFU2XoFWstmNjVnqCknpO5f+Q@mail.gmail.com' \
    --to=sedat.dilek@googlemail.com \
    --cc=Trond.Myklebust@netapp.com \
    --cc=a.p.zijlstra@chello.nl \
    --cc=akpm@linux-foundation.org \
    --cc=arighi@develer.com \
    --cc=balbir@linux.vnet.ibm.com \
    --cc=chris.mason@oracle.com \
    --cc=david@fromorbit.com \
    --cc=fengguang.wu@intel.com \
    --cc=gthelen@google.com \
    --cc=hch@lst.de \
    --cc=jack@suse.cz \
    --cc=kosaki.motohiro@jp.fujitsu.com \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=mel@csn.ul.ie \
    --cc=minchan.kim@gmail.com \
    --cc=riel@redhat.com \
    --cc=sedat.dilek@gmail.com \
    --cc=tytso@mit.edu \
    --cc=vgoyal@redhat.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