public inbox for kexec@lists.infradead.org
 help / color / mirror / Atom feed
From: ebiederm@xmission.com (Eric W. Biederman)
To: HATAYAMA Daisuke <d.hatayama@jp.fujitsu.com>
Cc: kexec@lists.infradead.org, heiko.carstens@de.ibm.com,
	linux-kernel@vger.kernel.org, lisa.mitchell@hp.com,
	kumagai-atsushi@mxc.nes.nec.co.jp, zhangyanfei@cn.fujitsu.com,
	akpm@linux-foundation.org, cpw@sgi.com, vgoyal@redhat.com
Subject: Re: [PATCH v3 00/21] kdump, vmcore: support mmap() on /proc/vmcore
Date: Tue, 19 Mar 2013 16:16:25 -0700	[thread overview]
Message-ID: <87ip4nj7zq.fsf@xmission.com> (raw)
In-Reply-To: <20130316040003.15064.62308.stgit@localhost6.localdomain6> (HATAYAMA Daisuke's message of "Sat, 16 Mar 2013 13:00:47 +0900")

HATAYAMA Daisuke <d.hatayama@jp.fujitsu.com> writes:

> Currently, read to /proc/vmcore is done by read_oldmem() that uses
> ioremap/iounmap per a single page. For example, if memory is 1GB,
> ioremap/iounmap is called (1GB / 4KB)-times, that is, 262144
> times. This causes big performance degradation.
>
> In particular, the current main user of this mmap() is makedumpfile,
> which not only reads memory from /proc/vmcore but also does other
> processing like filtering, compression and IO work. Update of page
> table and the following TLB flush makes such processing much slow;
> though I have yet to make patch for makedumpfile and yet to confirm
> how it's improved.
>
> To address the issue, this patch implements mmap() on /proc/vmcore to
> improve read performance. My simple benchmark shows the improvement
> from 200 [MiB/sec] to over 50.0 [GiB/sec].

I am in favor of this direction and the performance and other gains look
good.

I am not in favor of the ABI changes nor of the nearly order of
magnitude memory usage increase for elf notes by rounding everything up
to a page size boundary.

As a general note it is possible to support mmaping any partial page
by just rounding inside of your mmap function so you should not need to
copy partial pages.

If you don't want the memory overhead of merging the ELF notes in memory
in the second kernel you can simply require that the ELF header, the ELF
program header, and the PT_NOTE section be read from /proc/vmcore
instead of mmaped.

I did the math and with your changes to note generation in the worst
case you are reserving 20MiB in the first kernel to replace a 1.6MiB
with a 240KiB allocation in the second kernel.  That is the wrong
tradeoff, especially when you require an ABI change at the same time,
and the 5120+ entries in vmcore_list will likely measurably slow down
setting up your mappings with mmap.

Eric

_______________________________________________
kexec mailing list
kexec@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/kexec

      parent reply	other threads:[~2013-03-19 23:16 UTC|newest]

Thread overview: 76+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-03-16  4:00 [PATCH v3 00/21] kdump, vmcore: support mmap() on /proc/vmcore HATAYAMA Daisuke
2013-03-16  4:00 ` [PATCH v3 01/21] vmcore: reference e_phoff member explicitly to get position of program header table HATAYAMA Daisuke
2013-03-19 21:44   ` Eric W. Biederman
2013-03-21  2:50     ` HATAYAMA Daisuke
2013-03-21  6:11       ` Eric W. Biederman
2013-03-21 14:12       ` Vivek Goyal
2013-03-22  0:25         ` HATAYAMA Daisuke
2013-03-16  4:00 ` [PATCH v3 02/21] vmcore: clean up by removing unnecessary variable HATAYAMA Daisuke
2013-03-16  4:01 ` [PATCH v3 03/21] vmcore: rearrange program headers without assuming consequtive PT_NOTE entries HATAYAMA Daisuke
2013-03-19 21:59   ` Eric W. Biederman
2013-03-16  4:01 ` [PATCH v3 04/21] vmcore, sysfs: export ELF note segment size instead of vmcoreinfo data size HATAYAMA Daisuke
2013-03-16  4:01 ` [PATCH v3 05/21] vmcore: allocate buffer for ELF headers on page-size alignment HATAYAMA Daisuke
2013-03-16  4:01 ` [PATCH v3 06/21] vmcore: round up buffer size of ELF headers by PAGE_SIZE HATAYAMA Daisuke
2013-03-19 22:07   ` Eric W. Biederman
2013-03-16  4:01 ` [PATCH v3 07/21] vmcore, procfs: introduce a flag to distinguish objects copied in 2nd kernel HATAYAMA Daisuke
2013-03-19 19:35   ` Andrew Morton
2013-03-16  4:01 ` [PATCH v3 08/21] vmcore: copy non page-size aligned head and tail pages " HATAYAMA Daisuke
2013-03-19 19:37   ` Andrew Morton
2013-03-19 20:59   ` Eric W. Biederman
2013-03-19 21:22     ` Vivek Goyal
2013-03-19 23:35       ` Eric W. Biederman
2013-03-16  4:01 ` [PATCH v3 09/21] vmcore: modify vmcore clean-up function to free buffer on " HATAYAMA Daisuke
2013-03-16  4:01 ` [PATCH v3 10/21] vmcore: clean up read_vmcore() HATAYAMA Daisuke
2013-03-16  4:01 ` [PATCH v3 11/21] vmcore: read buffers for vmcore objects copied from old memory HATAYAMA Daisuke
2013-03-16  4:01 ` [PATCH v3 12/21] vmcore: allocate per-cpu crash_notes objects on page-size boundary HATAYAMA Daisuke
2013-03-19 21:06   ` Eric W. Biederman
2013-03-19 22:12   ` Eric W. Biederman
2013-03-20 13:48     ` Vivek Goyal
2013-03-20 20:48       ` Eric W. Biederman
2013-03-16  4:02 ` [PATCH v3 13/21] kexec: allocate vmcoreinfo note buffer " HATAYAMA Daisuke
2013-03-19 21:07   ` Eric W. Biederman
2013-03-19 22:12   ` Eric W. Biederman
2013-03-16  4:02 ` [PATCH v3 14/21] kexec, elf: introduce NT_VMCORE_DEBUGINFO note type HATAYAMA Daisuke
2013-03-16  4:02 ` [PATCH v3 15/21] elf: introduce NT_VMCORE_PAD type HATAYAMA Daisuke
2013-03-16  4:02 ` [PATCH v3 16/21] kexec: fill note buffers by NT_VMCORE_PAD notes in page-size boundary HATAYAMA Daisuke
2013-03-19 22:17   ` Eric W. Biederman
2013-03-16  4:02 ` [PATCH v3 17/21] vmcore: check NT_VMCORE_PAD as a mark indicating the end of ELF note buffer HATAYAMA Daisuke
2013-03-19 21:11   ` Eric W. Biederman
2013-03-21  2:59     ` HATAYAMA Daisuke
2013-03-21  3:54       ` Eric W. Biederman
2013-03-21 14:36         ` Vivek Goyal
2013-03-22  0:30           ` HATAYAMA Daisuke
2013-03-22  0:41             ` Eric W. Biederman
2013-03-19 22:20   ` Eric W. Biederman
2013-03-16  4:02 ` [PATCH v3 18/21] vmcore: check if vmcore objects satify mmap()'s page-size boundary requirement HATAYAMA Daisuke
2013-03-19 20:02   ` Andrew Morton
2013-03-19 21:22     ` Eric W. Biederman
2013-03-20 13:51     ` Vivek Goyal
2013-03-19 22:38   ` Eric W. Biederman
2013-03-20 13:57     ` Vivek Goyal
2013-03-20 20:55       ` Eric W. Biederman
2013-03-21  3:25         ` HATAYAMA Daisuke
2013-03-21  4:18           ` Eric W. Biederman
2013-03-21  6:14             ` HATAYAMA Daisuke
2013-03-21  6:29               ` Eric W. Biederman
2013-03-21  6:46                 ` HATAYAMA Daisuke
2013-03-21  7:07                   ` Eric W. Biederman
2013-03-21 15:21                     ` Vivek Goyal
2013-03-21 15:27                       ` Vivek Goyal
2013-03-22  0:43                         ` HATAYAMA Daisuke
2013-03-22  0:54                         ` Eric W. Biederman
2013-03-22  2:30                           ` HATAYAMA Daisuke
2013-03-21 14:57                 ` Vivek Goyal
2013-03-21  7:22               ` Eric W. Biederman
2013-03-21 14:49                 ` Vivek Goyal
2013-03-22  7:11                   ` HATAYAMA Daisuke
2013-03-21 13:50         ` Vivek Goyal
2013-03-16  4:02 ` [PATCH v3 19/21] vmcore: round-up offset of vmcore object in page-size boundary HATAYAMA Daisuke
2013-03-16  4:02 ` [PATCH v3 20/21] vmcore: count holes generated by round-up operation for vmcore size HATAYAMA Daisuke
2013-03-16  4:02 ` [PATCH v3 21/21] vmcore: introduce mmap_vmcore() HATAYAMA Daisuke
2013-03-19 19:30 ` [PATCH v3 00/21] kdump, vmcore: support mmap() on /proc/vmcore Andrew Morton
2013-03-21  3:52   ` HATAYAMA Daisuke
2013-03-21  6:16     ` Eric W. Biederman
2013-03-21  6:35       ` HATAYAMA Daisuke
2013-03-21  7:14         ` Eric W. Biederman
2013-03-19 23:16 ` Eric W. Biederman [this message]

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=87ip4nj7zq.fsf@xmission.com \
    --to=ebiederm@xmission.com \
    --cc=akpm@linux-foundation.org \
    --cc=cpw@sgi.com \
    --cc=d.hatayama@jp.fujitsu.com \
    --cc=heiko.carstens@de.ibm.com \
    --cc=kexec@lists.infradead.org \
    --cc=kumagai-atsushi@mxc.nes.nec.co.jp \
    --cc=linux-kernel@vger.kernel.org \
    --cc=lisa.mitchell@hp.com \
    --cc=vgoyal@redhat.com \
    --cc=zhangyanfei@cn.fujitsu.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