From: Lukasz Majewski <lukma@denx.de>
To: Adhemerval Zanella <adhemerval.zanella@linaro.org>
Cc: Joseph Myers <joseph@codesourcery.com>,
Florian Weimer <fweimer@redhat.com>,
Carlos O'Donell <carlos@redhat.com>,
Andreas Schwab <schwab@linux-m68k.org>,
libc-alpha <libc-alpha@sourceware.org>
Subject: Re: [PATCH] ci: Check for necessary Debian packages when running build-many-glibcs.py
Date: Tue, 9 Nov 2021 14:39:40 +0100 [thread overview]
Message-ID: <20211109143940.2d6f10d0@ktm> (raw)
In-Reply-To: <8ca53888-f0bf-a25f-88b5-d6279235c8c5@linaro.org>
[-- Attachment #1: Type: text/plain, Size: 4988 bytes --]
Hi Adhemerval,
> On 08/11/2021 16:44, Lukasz Majewski wrote:
> > Hi Adhemerval,
> >
> >> On 08/11/2021 13:59, Joseph Myers wrote:
> >>> On Mon, 8 Nov 2021, Lukasz Majewski wrote:
> >>>
> >>>> The same approach (with using the 'distro' python module) can be
> >>>> applied to Fedora or Suse.
> >>>
> >>> That module isn't part of the Python standard library. I don't
> >>> think we should introduce a dependency on it; rather, any use of
> >>> it should be appropriately conditional, so the code still runs
> >>> (without these checks) if the module is unavailable (importing
> >>> produces an ImportError).
> >>>
> >>> In particular, even if the OS Python installation includes that
> >>> module, the script should work with a separately built copy of
> >>> Python without any such modules from the OS.
> >>
> >> Maybe add a check without tying to any distribution (tool -v and
> >> some version parsing).
> >
> > This would require some extra, work but then we would avoid 'distro'
> > module as the dependency.
>
> I think the work required is that hard, something like:
>
> --
> import shutil
> import subprocess
>
> def get_version(progname):
> out = subprocess.run([progname, '--version'],
> stdout=subprocess.PIPE, check=True, universal_newlines=True).stdout
> return [int(x) for x in
> out.splitlines()[0].split()[-1].split('.')]
>
> def get_version_awk(progname):
> out = subprocess.run([progname, '--version'],
> stdout=subprocess.PIPE, check=True, universal_newlines=True).stdout
> version =
> out.splitlines()[0].split()[2].replace(',','').split('.') return
> [int(x) for x in version]
>
> def check_version(ver, req):
> for v, r in zip(ver, req):
> if v >= r:
> return True
> return False
>
> def version_str(ver):
> return '.'.join([str (x) for x in version])
>
>
> TOOLS={ 'make' : (get_version, (4,0)),
> 'makeinfo' : (get_version, (4,7)),
> 'awk' : (get_version_awk, (3,1,2)),
> 'bison' : (get_version, (2,7)),
> 'sed' : (get_version, (3,2)),
> 'flex' : (get_version, (2,6,0)),
> 'git' : (get_version, (2,32)),
> 'patch' : (get_version, (2,7,0)),
> 'tar' : (get_version, (1,3,4))}
>
> for k, v in TOOLS.items():
> version = v[0](k)
> ok = 'ok' if check_version (version, v[1]) else 'old'
> print('{:9}: {:3} (obtained=\"{}\" required=\"{}\")'.format(k, ok,
> version_str(version), version_str(v[1])))
> --
Yes, I think that this approach is the most exhaustive one, so we would
also check the version of required tools.
LGTM :-)
>
> >
> >>
> >>>
> >>>> +def check_os_requirements():
> >>>> + if distro.id() == "debian" and distro.version() == "10":
> >>>> + # List 'Debian' specific packages requirements
> >>>> (different than
> >>>> + # vanila distro) to run this test without errors.
> >>>> + debian_requirements = ['flex', 'bison', 'dnsutils',
> >>>> 'texinfo']
> >>>
> >>> Why is dnsutils needed?
> >>
> >> Also, strictly to build and check 'texinfo' is not required
> >> either.
> >
> > if makeinfo --split-size=5000000 --split-size=5000000 -I
> > "/work/wd/glibc/glibc-many-build/src/binutils/binutils/doc" -I
> > "/work/wd/glibc/glibc-many-build/src/binutils/binutils/../libiberty"
> > -I
> > "/work/wd/glibc/glibc-many-build/src/binutils/binutils/../bfd/doc"
> > -I ../../bfd/doc --no-split -I
> > /work/wd/glibc/glibc-many-build/src/binutils/binutils/doc \ -o
> > binutils.info `test -f 'binutils.texi' || echo
> > '/work/wd/glibc/glibc-many-build/src/binutils/binutils/doc/'`binutils.texi;
> > \ then \ rc=0; \ else \ rc=$?; \ $restore $backupdir/* `echo
> > "./binutils.info" | sed 's|[^/]*$||'`; \ fi; \ rm -rf $backupdir;
> > exit $rc
> > /work/wd/glibc/glibc-many-build/src/binutils/binutils/doc/binutils.texi:871:
> > warning: @ref node name should not contain `.'
> > /work/wd/glibc/glibc-many-build/src/binutils/binutils/doc/binutils.texi:1329:
> > warning: @xref node name should not contain `.' make[4]: Leaving
> > directory
> > '/work/wd/glibc/glibc-many-build/build/compilers/powerpc64-linux-gnu/binutils/binutils/doc'
> > Making info in po make[4]: Nothing to be done for 'info'. make[4]:
> > Nothing to be done for 'info-am'. make[2]: *** [Makefile:3701:
> > all-binutils] Error 2 make[2]: *** Waiting for unfinished jobs....
> >
>
> So binutils does still requires texinfo. It would be good to get rid
> of this requirement, it does help a lot in build/test total wall time.
Best regards,
Lukasz Majewski
--
DENX Software Engineering GmbH, Managing Director: Wolfgang Denk
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-59 Fax: (+49)-8142-66989-80 Email: lukma@denx.de
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]
next prev parent reply other threads:[~2021-11-09 13:39 UTC|newest]
Thread overview: 26+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-11-08 16:52 Lukasz Majewski
2021-11-08 16:59 ` Joseph Myers
2021-11-08 17:06 ` Adhemerval Zanella
2021-11-08 17:15 ` Florian Weimer
2021-11-08 19:44 ` Lukasz Majewski
2021-11-08 20:27 ` Adhemerval Zanella
2021-11-09 13:39 ` Lukasz Majewski [this message]
2021-11-09 13:43 ` Adhemerval Zanella
2021-11-09 15:32 ` Lukasz Majewski
2023-05-26 22:12 ` DJ Delorie
2023-05-26 22:34 ` Joseph Myers
2023-05-31 19:54 ` DJ Delorie
2023-05-31 20:18 ` DJ Delorie
2023-06-05 20:42 ` Adhemerval Zanella Netto
2023-06-05 21:12 ` DJ Delorie
2023-09-21 21:44 ` DJ Delorie
2023-10-09 18:23 ` Adhemerval Zanella Netto
2023-10-09 21:53 ` DJ Delorie
2021-11-08 18:15 ` Lukasz Majewski
2021-11-08 18:31 ` Joseph Myers
2021-11-08 19:51 ` Lukasz Majewski
2021-11-08 21:25 ` Joseph Myers
2021-11-08 17:17 ` Florian Weimer
2021-11-08 18:52 ` Joseph Myers
2021-11-08 19:03 ` Florian Weimer
2021-11-08 18:56 ` Adhemerval Zanella
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=20211109143940.2d6f10d0@ktm \
--to=lukma@denx.de \
--cc=adhemerval.zanella@linaro.org \
--cc=carlos@redhat.com \
--cc=fweimer@redhat.com \
--cc=joseph@codesourcery.com \
--cc=libc-alpha@sourceware.org \
--cc=schwab@linux-m68k.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;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).