Re: Multilib Systemd Build using jhalfs?

"Thomas Trepl" ([email protected] via alfs-discuss Mailing List) <[email protected]>
Newsgroups gmane.linux.lfs.automated
Message-ID <b3d96d5e77bb966098e1e57cc5f2e8352261e63b.camel@linuxfromscratch.org>
Am Sonntag, dem 15.12.2024 um 21:48 -0500 schrieb Eric Clothier:
> Ok, so if I am reading this right, setting the branch to multilib still would not work for multilib with systemd, as that project was removed since it failed to build. I did go through and copy over the multilib instructions into the generated scripts from the systemd book, as best as I was able. That build did indeed fail, due to not having multilib instructions for the systemd package itself, which meant that util-linux failed to link when it was built for m32.
> 
> Thank you so much for the response and the info. I am going to keep working at this when I have the chance, and I will reach out to Thomas about working on multilib for systemd.
> 
> Thanks,
> Eric Clothier

Hi Eric,

Please don't top-post - if the thread gets longer, it gets very confusing

There is an option for systemd.  When you checkout the multilib branch than
you can render the systemd variant by running

$ make REV=systemd ARCH=ml_all BASEDIR=...      # or ml_32

But i have to admit that it is maintained by "best guess" - i don't have a
systemd environment here, i test only on sysv. This is what makes multilib
to be known as "sysv only" - its tested (by me) only on sysv.

If and how jhalfs can act on the multilib branch, i don't know (i run my
own toolchain), but i think to remember Pierre did some changes to jhalfs
to support ML.

If you encounter issues and have a fix for them in the systemd world, just
let me know, i'll merge them in.

-- 
Thomas

> On Sun, Dec 15, 2024 at 9:27 PM Xi Ruoyao <[email protected]> wrote:
> > On Sun, 2024-12-15 at 17:51 -0500, Eric Clothier wrote:
> > > I am using jhalfs to build a multilib Linux system, using systemd
> > > init. I chose the proper options in the jhalfs configuration,
> > > selecting the x86_64 and m32 multilib option and the systemd init
> > > option.
> > 
> > You need to also use a modified LFS book.  I.e. setting "Book version"
> > to "Branch (default to trunk) or any commit" and "Branch, tag, or any
> > commit" to "multilib".  Or, setting "Book version" to "Working Copy" and
> > "Loc of working copy" to a local copy of the LFS book source with the
> > modification.
> >  
> > 
> > /* snip */
> > 
> > > Is this being caused by the MLFS project only currently supporting
> > > SysV init?
> > 
> > *This* is not.  But the project indeed only supports SysVinit.
> > 
> > It was already reported the "MLFS systemd" did not successfully build
> > and so it was removed.  Thus it's highly unlikely it'll even
> > successfully build for you.
> > 
> > > If so, I would be interested in assisting with bringing multilib to
> > > the systemd init builds as well. Any input or recommendations would be
> > > greatly appreciated.
> > 
> > Ask Thomas (who doesn't subscribe alfs-discuss AFAIK).  LFS multilib is
> > his personal project.
> > 
> > -- 
> > Xi Ruoyao <[email protected]>
> > School of Aerospace Science and Technology, Xidian University

-- 
http://lists.linuxfromscratch.org/sympa/info/alfs-discuss
Unsubscribe: See the above information page
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.