From: Rusty Russell <rusty@rustcorp.com.au>
To: Andreas Mohr <andi@lisas.de>, Andrew Morton <akpm@linux-foundation.org>
Cc: Bertrand Jacquin <beber@meleeweb.net>, Marco d'Itri <md@linux.it>,
linux-modules@vger.kernel.org, linux-kernel@vger.kernel.org,
"Jon Masters" <jcm@redhat.com>
Subject: Re: [PATCH] modules: CONFIG_MODULE_COMPRESS: add hint that userspace support may easily be missing.
Date: Mon, 01 Jun 2015 15:56:44 +0930 [thread overview]
Message-ID: <87r3pwgd0b.fsf@rustcorp.com.au> (raw)
In-Reply-To: <20150531152932.GA16337@rhlx01.hs-esslingen.de>
Andreas Mohr <andi@lisas.de> writes:
> Hi,
>
> I just had a not so nice experience
> when finally upgrading to a new 4.1-rc5
> with CONFIG_MODULE_COMPRESS newly enabled -
> userspace binary parts (kmod 18 or 20 in my case)
> did not have compression enabled
> (at least on Debian 8pre, vs. encountering it enabled on FC21)
> since it does not seem to be
> the default build configuration of kmod (yet?).
Sure. Let's get the maintainers to insert the actual version required
in the help text though.
Thanks,
Rusty.
> Doing a manual
> for gz in $(find . -name "*.gz"); do echo gunzip $gz; done
> depmod -a (somewhere temporarily in bootup scripts)
> manages to fix the grave problem
> introduced by erroneously having enabled CONFIG_MODULE_COMPRESS.
>
> Thus it seems that the Kconfig text was much more optimistic
> than a hapless user would want to encounter,
> thus I decided to have it updated.
>
>
> BTW: kmod and/or module-init-tools packages
> might want to provide a generic way to openly advertise
> certain build-configured "capabilities" in the binary -
> neither
> kmod --help
> nor
> kmod -V
> indicated whether or not they provided capabilities
> such as e.g. providing or not providing support of compression types
> and which ones.
> That would have been a very helpful way
> to reliably determine
> that support in fact is missing in the binary,
> rather than having to resort to clumsy hacks
> such as ldd or even strings.
>
>
> While this is a minor but useful addition
> rather than a severe fix,
> having a CC to stable@ added subsequently might be useful.
>
> Thanks!
>
> Signed-off-by: Andreas Mohr <andim2@users.sf.net>
>
> ---
> init/Kconfig | 18 +++++++++++++-----
> 1 file changed, 13 insertions(+), 5 deletions(-)
>
> diff --git a/init/Kconfig b/init/Kconfig
> index dc24dec..8e451f30 100644
> --- a/init/Kconfig
> +++ b/init/Kconfig
> @@ -1951,11 +1951,19 @@ config MODULE_COMPRESS
> This option compresses the kernel modules when 'make
> modules_install' is run.
>
> - The modules will be compressed either using gzip or xz depend on the
> - choice made in "Compression algorithm".
> -
> - module-init-tools has support for gzip format while kmod handle gzip
> - and xz compressed modules.
> + The module files will be compressed either using gzip or xz
> + depending on the choice made in "Compression algorithm".
> +
> + Obviously one will then need appropriate userspace parts
> + which are actually able to deal with compressed files, too:
> + module-init-tools has support for gzip format
> + while kmod handles gzip and xz compressed modules.
> + However, we observed that in several environments
> + module loader binaries do not have that enabled (yet?)
> + and thus bootup will fail fatally -
> + manually doing ldd on these binaries
> + to detect compression libraries
> + is a tell-tale sign of having support enabled.
>
> When a kernel module is installed from outside of the main kernel
> source and uses the Kbuild system for installing modules then that
> --
> 2.1.4
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
next prev parent reply other threads:[~2015-06-01 6:53 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-05-31 15:29 [PATCH] modules: CONFIG_MODULE_COMPRESS: add hint that userspace support may easily be missing Andreas Mohr
2015-06-01 6:26 ` Rusty Russell [this message]
2015-06-03 12:04 ` Michal Marek
2015-06-03 17:30 ` Lucas De Marchi
2015-06-03 17:36 ` Kay Sievers
2015-06-03 17:51 ` Lucas De Marchi
2015-06-04 1:30 ` Rusty Russell
2015-06-04 2:31 ` Lucas De Marchi
2015-06-04 19:53 ` Andreas Mohr
2015-06-04 20:22 ` Rusty Russell
2015-06-07 6:18 ` Lucas De Marchi
2015-06-04 2:51 ` Marco d'Itri
2015-06-04 3:19 ` Stephen Rothwell
2015-06-04 20:09 ` Andreas Mohr
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=87r3pwgd0b.fsf@rustcorp.com.au \
--to=rusty@rustcorp.com.au \
--cc=akpm@linux-foundation.org \
--cc=andi@lisas.de \
--cc=beber@meleeweb.net \
--cc=jcm@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-modules@vger.kernel.org \
--cc=md@linux.it \
/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;
as well as URLs for NNTP newsgroup(s).