All of lore.kernel.org
 help / color / mirror / Atom feed
From: Alan Jenkins <sourcejedi.lkml@googlemail.com>
To: linux-kbuild <linux-kbuild@vger.kernel.org>
Cc: Carmelo Amoroso <carmelo73@gmail.com>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
	Rusty Russell <rusty@rustcorp.com.au>
Subject: Re: Fast LKM symbol resolution
Date: Mon, 2 Nov 2009 16:52:47 +0000	[thread overview]
Message-ID: <9b2b86520911020852q49c55695rb05d87090fa9ad33@mail.gmail.com> (raw)

Here's my latest code using binary search for symbol resolution.  This
version has benefited from a little more testing (heh), in particular
running on a simulated ARM system.

    web: <http://github.com/sourcejedi/linux-2.6/commits/module-V2-beta1>

    git clone git://github.com/sourcejedi/linux-2.6.git module-V2-beta1

As before, it saves 0.3s boot time for a modular kernel on my netbook
(630 Mhz Celeron M).  On that system it represents something like 90%
of the maximum possible speedup.  It is claimed that slower systems
would benefit from a further speedup (e.g. using hash tables instead),
but we don't have any numbers to illustrate this yet.


Changes since the original patches were posted:

1) The sorted version of the exports is now generated in
arch-independent assembly language, as .tmp-exports-asm.S.  It uses
similar pseudo-instructions to those in the current .tmp-kallsyms.S.

Previously the sorted exports were generated in C, but the compiler
outputted them in reverse order.  This was made to work by reversing
the comparison used in binary search, but it was a brittle hack.

This means mod_export.h now includes two implementations of
EXPORT_SYMBOL; one for general use in C code, and one for use in
.tmp-exports-asm.S.

2) In vmlinux.lds.h, the unsorted exports are now discarded in the
DISCARDS macro.

Previouslly they were discarded before including the sections of
sorted exports.  This broke the ARM build by discarding some unrelated
symbols too early.  It appears that a) the position of /DISCARD/
directives is signficant; b) all /DISCARD/ directives are merged
together and applied at the point of the first /DISCARD/ directive.

3) Now lightly tested on ARM (using qemu)

To build on ARM it was necessary to remove EXPORT_ALIAS, used in
armksyms.c for the sole purpose of allowing out of tree floating point
emulation code to be loaded as a module.  I have posted this change on
the ARM list and received no comments either way.

4) Build-tested on blackfin (a MODULE_SYMBOL_PREFIX arch)

This required that MODULE_SYMBOL_PREFIX be defined via a Kconfig
option, CONFIG_HAVE_SYMBOL_PREFIX.  Up until now it has been defined
in <asm/module.h> - but this also includes C declarations, which made
it impossible to use in .tmp-exports-asm.S.

Alan

             reply	other threads:[~2009-11-02 16:52 UTC|newest]

Thread overview: 56+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-11-02 16:52 Alan Jenkins [this message]
2009-11-03  3:55 ` Greg KH
2009-11-03 10:06 ` [PATCH 0/10] module: Speed up symbol resolution during module loading (using binary search) Alan Jenkins
2009-11-03 15:58   ` Greg KH
2009-11-05 12:17     ` Rusty Russell
2009-11-03 10:06 ` [PATCH 01/10] ARM: use unified discard definition in linker script Alan Jenkins
2009-11-03 10:06 ` [PATCH 02/10] ARM: unexport symbols used to implement floating point emulation Alan Jenkins
2009-11-03 10:06 ` [PATCH 03/10] module: extract __EXPORT_SYMBOL from module.h into mod_export.h Alan Jenkins
2009-11-03 10:06 ` [PATCH 04/10] module: make MODULE_SYMBOL_PREFIX into a CONFIG option Alan Jenkins
2009-11-03 10:19   ` Mike Frysinger
2009-11-03 10:19     ` Mike Frysinger
2009-11-03 12:16     ` Alan Jenkins
2009-11-03 12:30       ` Mike Frysinger
2009-11-03 12:30         ` Mike Frysinger
2009-11-03 13:29         ` Paul Mundt
2009-11-03 13:39           ` Mike Frysinger
2009-11-03 13:39             ` Mike Frysinger
2009-11-03 13:46             ` Paul Mundt
2009-11-03 13:58               ` Mike Frysinger
2009-11-03 13:58                 ` Mike Frysinger
2009-11-03 14:07                 ` Paul Mundt
2009-11-03 10:06 ` [PATCH 05/10] kbuild: sort the list of symbols exported by the kernel (__ksymtab) Alan Jenkins
2009-11-04  8:19   ` Rusty Russell
2009-11-04 10:00     ` Alan Jenkins
2009-11-04 11:12       ` Mike Frysinger
2009-11-04 11:12         ` Mike Frysinger
2009-11-04 17:19       ` Sam Ravnborg
2009-11-05 14:24         ` Alan Jenkins
2009-11-05 16:17           ` Mike Frysinger
2009-11-05 16:17             ` Mike Frysinger
2009-11-09  3:17           ` Rusty Russell
2009-11-20 22:20             ` Tony Luck
2009-11-20 22:20               ` Tony Luck
2009-11-21  0:02               ` Alan Jenkins
2009-11-23 19:53                 ` Alex Chiang
2009-11-23 22:44                   ` Alan Jenkins
2009-11-24  0:57                   ` Rusty Russell
2009-11-24  5:39                     ` James Bottomley
2009-11-24  9:28                       ` Alan Jenkins
2009-11-24 22:43                         ` James Bottomley
2009-11-25  9:15                           ` Alan Jenkins
2009-11-25 15:08                             ` James Bottomley
2009-11-25 17:01                               ` Alan Jenkins
2009-11-27 11:03                               ` Rusty Russell
2009-11-26  0:40   ` Andrew Morton
2009-11-26 17:14     ` Alan Jenkins
2009-11-03 10:06 ` [PATCH 06/10] module: refactor symbol tables and try to reduce code size of each_symbol() Alan Jenkins
2009-11-04  8:28   ` Rusty Russell
2009-11-04  9:45     ` Alan Jenkins
2009-11-03 10:06 ` [PATCH 07/10] lib: Add generic binary search function to the kernel Alan Jenkins
2009-11-03 10:06 ` [PATCH 08/10] lib: bsearch - remove redundant special case for arrays of size 0 Alan Jenkins
2009-11-03 10:06 ` [PATCH 09/10] module: speed up find_symbol() using binary search on the builtin symbol tables Alan Jenkins
2009-11-04  8:31   ` Rusty Russell
2009-11-03 10:06 ` [PATCH 10/10] module: fix is_exported() to return true for all types of exports Alan Jenkins
2009-11-04  8:32   ` Rusty Russell
2009-11-06  5:37 ` Fast LKM symbol resolution Carmelo Amoroso

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=9b2b86520911020852q49c55695rb05d87090fa9ad33@mail.gmail.com \
    --to=sourcejedi.lkml@googlemail.com \
    --cc=carmelo73@gmail.com \
    --cc=linux-kbuild@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=rusty@rustcorp.com.au \
    --subject='Re: Fast LKM symbol resolution' \
    /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

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.