Re: [fixed] common/libs/func_download_pkgs: line 84: [: =: unary operator expected

On Thu, 2025-09-18 at 11:15 +0200, Marcin Dulak wrote:
> On Thu, Sep 18, 2025 at 10:38 AM Pierre Labastie
> <[email protected]> wrote:
> > 
> > On Mon, 2025-09-15 at 11:48 +0000, md wrote:
> > > Hi,
> > > 
> > > With
> > > 
> > > jhalfs
> > > git show --oneline --no-abbrev --shortstat
> > > 540e2dbeccefe3edf0c718c3e1341190b5ab7830 (HEAD -> trunk,
> > > origin/trunk, origin/
> > > HE  AD) Make test-mismatch configurable
> > > 
> > > lfs
> > > git show --oneline --no-abbrev --shortstat
> > > 158d47f83d68e159f05f48b07d7be584f07ec559 (HEAD -> trunk,
> > > origin/trunk, origin/
> > > HE  AD) Add a correction about tcl tests.
> > > 
> > > I'm getting:
> > > 
> > > linux-6.16.tar.xz: using cached file in
> > > /vagrant/jhalfs/mnt/build_dir/sources
> > > common/libs/func_download_pkgs: line 84: [: =: unary operator
> > > expected
> > > common/libs/func_download_pkgs: line 85: [: =: unary operator
> > > expected
> > > md5sum: 'standard input': no properly formatted checksum lines
> > > found
> > > MD5SUM did not match REMOTE copy
> > > 
> > > It looks like this does not prevent the md5 checks from
> > > functioning.
> > > 
> > > Marcin
> > 
> > Hi Marcin,
> > 
> > Sorry for late answer. You do not say which flavor of the book you
> > are
> > using (64-bit or multilib?). Looks like something gets wrong when
> > reading the md5 of the kernel. But using linux-6.16 is (was) only
> > done
> > in multilib IIUC...
> > 
> > Since the md5 is not read correctly, the syntax of lines 84 and 85
> > in
> > common/libs/func_download_pkgs is wrong.
> > 
> > Investigating why this is so, using most recent multilib version.
> > The
> > commit ref you give is in the "trunk" branch, but has been merged
> > with
> > multilib on September 14th, while your message is dated September
> > 15th.
> > --
> > http://lists.linuxfromscratch.org/sympa/info/alfs-discuss
> > Unsubscribe: See the above information page
> 
> Hi Pierre,
> 
> I'm on x86_64, and always building the latest lfs-git and jhalfs
> trunk.
> Here is the jhalfs configuration file
> https://github.com/marcindulak/jhalfs-ci/blob/7c3c5be2e027126b0f56ae3e2d257531a8536255/configuration
> 
> The kernel m5sum problems are discussed here
> https://lists.linuxfromscratch.org/sympa/arc/lfs-support/2025-09/msg00018.html

Thanks for the details. Yes the note about base+patch for kernel was
also added into the 64-bit book, and thus generates the same error. See
jhalfs commits d9e80918f and 69eb5e74e9 for technical details. Those
should fix the error (until some editor has a brilliant idea again
:)[1]

Pierre
[1] Please do have ideas: it's the role of jhalfs to adapt (although
the maintainer may whine sometimes, or suggest alternate solutions)

-- 
http://lists.linuxfromscratch.org/sympa/info/alfs-discuss
Unsubscribe: See the above information page