Message-ID: <CA+bK7J5GH6qR08rDCNacixLi6CE=2E1qBW64v-zCWWB1vJKtFA@mail.gmail.com>
Date:	Wed, 15 Aug 2012 11:02:32 -0700
From:	Tim Bird <tbird20d@...il.com>
To:	Thai Bui <blquythai@...il.com>
Cc:	Konrad Rzeszutek Wilk <konrad.wilk@...cle.com>,
	Randy Dunlap <rdunlap@...otime.net>, linux-doc@...r.kernel.org,
	linux-kernel@...r.kernel.org, tim.bird@...sony.com
Subject: Re: [PATCH 1/1] boot: Put initcall_debug into its own Kconfig option DEBUG_INITCALL

Thai,

How can I find out more about this project to reduce the size of the Linux
kernel?  I have been doing some research at Sony on ways to reduce
the Linux kernel to very small sizes, without introducing persistent source
code changes or config options.  The reason for this is to avoid the inevitable
resistance to mainlining hundreds or thousands of config options and associated
code changes.

In a nutshell what I do is select individual data fields or functions
to eliminate or constrain,
then let automated tools impose that constraint throughout a working copy
of the kernel source tree.  (That is, my tools are run for each build
of the kernel,
so the source changes they make are not persistent).  The compiler
then optimizes
out the fields, and code associated with them.  This allows me to impose a
constraint (similar to a config option), without interjecting it into
the Kconfig
system or the mainline kernel source tree.

The code is in very early proof-of-concept condition right now.
I am currently experimenting to see how robust this system is, in terms of
both code correctness for a specific kernel version, and applicability
of technique
over a range of kernel versions.

I'd like to describe my work to your group in more detail, and see if there's
any overlap between what I'm doing and what you're doing.  My own goal
is kernel runtime footprint, including static and dynamic memory usage of 500k.

Is there any web site describing your work, or staging area for your patches?
Would you be willing to share your .config files?

Thanks for any information you can provide about your work.
 -- Tim

On Mon, Aug 13, 2012 at 3:11 PM, Thai Bui <blquythai@...il.com> wrote:
> Hi all,
>
> I am as part of a capstone group at Portland State University is
> working to tinify the kernel as small as possible. The ultimate goal
> is to make the kernel small enough to use on micro-controller (or
> under < 200k). This patch is one of them, it saves 176 bytes on a
> minimal configuration of the kernel (the bzImage file was reduced from
> 363264 bytes to 363264 bytes applying this patch).
>
> Aside from the purpose of reducing the size of the kernel. We are also
> trying to clean up the kernel by making Kconfig options to compile out
> the stuffs that aren't used often.
>
> On Mon, Aug 13, 2012 at 5:21 PM, Konrad Rzeszutek Wilk
> <konrad.wilk@...cle.com> wrote:
>> On Mon, Aug 13, 2012 at 01:57:11PM -0700, Thai Bui wrote:
>>> Putting DEBUG_INITCALL config option to compile out the command-line option
>>> "initcall_debug".
>>
>> Can you explain the benfits of this please?
>>>
>>> Signed-off-by: Thai Bui <blquythai@...il.com>
>>> Reviewed-by: Josh Triplett <josh@...htriplett.org>
>>> ---
>>>  Documentation/kernel-parameters.txt |    3 ++-
>>>  include/linux/init.h                |    4 ++++
>>>  init/main.c                         |    2 ++
>>>  lib/Kconfig.debug                   |    9 +++++++++
>>>  4 files changed, 17 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt
>>> index d99fd9c..3dbaf15 100644
>>> --- a/Documentation/kernel-parameters.txt
>>> +++ b/Documentation/kernel-parameters.txt
>>> @@ -1026,7 +1026,8 @@ bytes respectively. Such letter suffixes can also be entirely omitted.
>>>
>>>       initcall_debug  [KNL] Trace initcalls as they are executed.  Useful
>>>                       for working out where the kernel is dying during
>>> -                     startup.
>>> +                     startup. DEBUG_INITCALL needs to be enabled in order
>>> +                     for this option to work.
>>>
>>>       initrd=         [BOOT] Specify the location of the initial ramdisk
>>>
>>> diff --git a/include/linux/init.h b/include/linux/init.h
>>> index 6b95109..d2f31f1 100644
>>> --- a/include/linux/init.h
>>> +++ b/include/linux/init.h
>>> @@ -157,7 +157,11 @@ void prepare_namespace(void);
>>>
>>>  extern void (*late_time_init)(void);
>>>
>>> +#ifdef CONFIG_DEBUG_INITCALL
>>>  extern bool initcall_debug;
>>> +#else
>>> +static const bool initcall_debug = false;
>>> +#endif /* CONFIG_DEBUG_INITCALL */
>>>
>>>  #endif
>>>
>>> diff --git a/init/main.c b/init/main.c
>>> index ff49a6d..65837f7 100644
>>> --- a/init/main.c
>>> +++ b/init/main.c
>>> @@ -648,8 +648,10 @@ static void __init do_ctors(void)
>>>  #endif
>>>  }
>>>
>>> +#ifdef CONFIG_DEBUG_INITCALL
>>>  bool initcall_debug;
>>>  core_param(initcall_debug, initcall_debug, bool, 0644);
>>> +#endif /* CONFIG_DEBUG_INITCALL */
>>>
>>>  static char msgbuf[64];
>>>
>>> diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug
>>> index 8745ac7..424ac93 100644
>>> --- a/lib/Kconfig.debug
>>> +++ b/lib/Kconfig.debug
>>> @@ -769,6 +769,15 @@ config DEBUG_WRITECOUNT
>>>
>>>         If unsure, say N.
>>>
>>> +config DEBUG_INITCALL
>>> +     bool "Debug initcalls as they are executed"
>>> +     depends on DEBUG_KERNEL
>>> +     help
>>> +       Enable this for tracing initcalls during startup. Useful for working
>>> +       out where the kernel is dying during startup.
>>> +
>>> +       If unsure, say N
>>> +
>>>  config DEBUG_MEMORY_INIT
>>>       bool "Debug memory initialisation" if EXPERT
>>>       default !EXPERT
>>> --
>>> 1.7.9.5
>>>
>>> --
>>> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
>>> the body of a message to majordomo@...r.kernel.org
>>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>> Please read the FAQ at  http://www.tux.org/lkml/
>
>
>
> --
> Thai
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@...r.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Powered by blists - more mailing lists