From: Gerald Champagne <gerald.champagne@esstech.com>
To: Zhang Fuxin <fxzhang@ict.ac.cn>
Cc: "linux-mips@oss.sgi.com" <linux-mips@oss.sgi.com>
Subject: Re: Freeing global memory used only by __init functions
Date: Tue, 18 Sep 2001 13:37:31 -0500 [thread overview]
Message-ID: <3BA7946B.4070806@esstech.com> (raw)
In-Reply-To: 200109181808.LAA05245@mail.esstech.com
Thanks for the reply. Sorry I didn't find the __initdata
attribute. Is this stuff worth patching?
I'd make the following changes as a start:
in arch/mips/mips-boards/generic/printf.c:
Add __init to putPromChar
Add __init to getPromChar
Add __initdata to buf (1k buffer)
in arch/mips/mips-boards/generic/cmdline.c:
Add __initdata to arcs_cmdline
Can someone point to a set of rules for submitting patches
for linux-mips? I'm familiar with the methods used for the
kernel. Is this the same? I'm using 2.4.3 from the mips site.
Can I patch against that, or do I have to start from a
certain cvs version?
Thanks!
Gerald
Zhang Fuxin wrote:
> hi,Gerald Champagne,
> I think __initdata attribute is the answer.You just need to
> put init data in init data section(via __initdata),there are many examples
> in kernel.
>
>
> 在 2001-09-18 12:58:00 you wrote:
>
>>I noticed that several global buffers are used by a few functions marked
>>as __init functions. I assume that the code space used by these functions
>>will be freed up when kernel initialization is completed, but can the
>>associated global memory be freed up as well?
>>
>>An example can be found in arch/mips/mips-boards/generic/cmdline.c:
>>
>>char arcs_cmdline[COMMAND_LINE_SIZE];
>>char * __init prom_getcmdline(void);
>>void __init prom_init_cmdline(void);
>>
>>arcs_cmdline is only used by these two functions and one other function
>>marked as __init.
>>
>>This buffer is small, but it can apply to larger buffers as well. For
>>example, in arch/mips/mips-boards/generic/printf.c, I think the functions
>>putPromChar and getPromChar should be marked as __init functions, and the
>>1k buffer "buf" is never used after initialization. Can this 1k be recovered?
>>
>>I know kmalloc could normally be used in kernel code, but that won't work on
>>initialization code used before kmalloc is initialized.
>>
>>Thanks.
>>
>>Gerald
>>
>
> Regards
> Zhang Fuxin
> fxzhang@ict.ac.cn
>
>
>
>
next parent reply other threads:[~2001-09-18 18:36 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <200109181808.LAA05245@mail.esstech.com>
2001-09-18 18:37 ` Gerald Champagne [this message]
2001-09-19 1:02 ` Freeing global memory used only by __init functions Ralf Baechle
2001-09-18 18:08 Zhang Fuxin
-- strict thread matches above, loose matches on Subject: below --
2001-09-18 17:58 Gerald Champagne
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=3BA7946B.4070806@esstech.com \
--to=gerald.champagne@esstech.com \
--cc=fxzhang@ict.ac.cn \
--cc=linux-mips@oss.sgi.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