From: Stelian Pop <stelian@popies.net>
To: "Randy.Dunlap" <rddunlap@osdl.org>
Cc: Jeff Garzik <jgarzik@pobox.com>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
Dominik Brodowski <linux@dominikbrodowski.de>,
Andrew Morton <akpm@osdl.org>
Subject: Re: [PATCH] drivers/net/pcmcia: use module_param() instead of MODULE_PARM()
Date: Wed, 10 Nov 2004 22:21:32 +0100 [thread overview]
Message-ID: <20041110212132.GK2706@deep-space-9.dsnet> (raw)
In-Reply-To: <41927055.9030306@osdl.org>
On Wed, Nov 10, 2004 at 11:47:33AM -0800, Randy.Dunlap wrote:
> >In fact, I do think that all module parameter should be exposed in
> >/sys, and that a '0' in module_param() should really mean 0400.
> >
> >It can be useful to know what parameters have been passed to a module,
> >and I cannot think of a single case where we want to hide this
> >information (and no, security doesn't really apply here. If you have
> >root rights than you can also look into the kernel memory and find
> >out the value by yourself).
[...]
>
> I don't have an argument with most of that, but I am concerned
> about how much memory each entry requires and how useful it really
> is. IOW, if I need to know the module parameters for a module,
> I can probably find that info somewhere else, like in
> /etc/modprobe.conf or a script etc., so why waste memory on it?
The problem is that the information you can get out of /etc/modprobe.conf
or some script is not necessarily consistent. Maybe the module was
hand-loaded using different parameters that the ones in the script.
This is the same issue as with the in-kernel .config. We have this
today because this way we're sure we are looking at the right one,
not an older copy which happen to be in /boot. Same thing for KKSYMOOPS.
Yes, this wastes a bit of memory (quite a lot actually for KKCONFIG
or KKSYMOOPS), but less and less people cares. And those who really
care (embedded people etc) can disable this with a config option.
If module parameters are a memory issue, maybe we should do the same
as above: put all of them into /sys unless chosen otherwise, and in
this case disable all of them, since we can go just fine without
any module parameter in /sys.
Stelian.
--
Stelian Pop <stelian@popies.net>
next prev parent reply other threads:[~2004-11-10 21:27 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2004-11-04 11:27 [PATCH RESEND] pcmcia network drivers cleanup Stelian Pop
2004-11-05 2:25 ` Jeff Garzik
2004-11-05 8:32 ` Stelian Pop
2004-11-10 15:59 ` Stelian Pop
2004-11-10 16:00 ` [PATCH] drivers/net/pcmcia: use module_param() instead of MODULE_PARM() Stelian Pop
2004-11-10 16:35 ` Randy.Dunlap
2004-11-10 19:52 ` Stelian Pop
2004-11-10 19:47 ` Randy.Dunlap
2004-11-10 21:21 ` Stelian Pop [this message]
2004-11-13 22:44 ` Dominik Brodowski
2004-11-10 16:01 ` [PATCH] drivers/net/pcmcia: rework debugging Stelian Pop
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=20041110212132.GK2706@deep-space-9.dsnet \
--to=stelian@popies.net \
--cc=akpm@osdl.org \
--cc=jgarzik@pobox.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux@dominikbrodowski.de \
--cc=rddunlap@osdl.org \
/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