4:43 a.m.
On Tue, Feb 06, 2018 at 03:18:32PM +0100, Lukas Slebodnik wrote:
On (06/02/18 10:47), Sumit Bose wrote:
>On Mon, Feb 05, 2018 at 04:54:26PM +0100, Sumit Bose wrote:
>> On Sat, Feb 03, 2018 at 07:13:35AM -0800, Richard Sharpe wrote:
>> > On Sat, Feb 3, 2018 at 3:08 AM, Sumit Bose <sbose(a)redhat.com> wrote:
>> > > On Fri, Feb 02, 2018 at 07:04:50PM -0800, Richard Sharpe wrote:
>> > >> Hi folks,
>> > >>
>> > >> After installing all the RPMs needed, I managed to get sssd to
build
>> > >> using chmake.
>> > >>
>> > >> This is what happened at the end:
>> > >>
>> > >> -------------------------
>> > >> PASS: src/tests/double_semicolon_test
>> > >> make[4]: execvp: /bin/sh: Argument list too long
>> > >> make[4]: *** [test-suite.log] Error 127
>> > >> make[4]: Leaving directory `/home/rsharpe/src/sssd/x86_64'
>> > >> make[3]: *** [check-TESTS] Error 2
>> > >> make[3]: Leaving directory `/home/rsharpe/src/sssd/x86_64'
>> > >> make[2]: *** [check-am] Error 2
>> > >> make[2]: Leaving directory `/home/rsharpe/src/sssd/x86_64'
>> > >> make[1]: *** [check-recursive] Error 1
>> > >> make[1]: Leaving directory `/home/rsharpe/src/sssd/x86_64'
>> > >> make: *** [check] Error 2
>> > >> --------------------------
>> > >>
>> > >> Is this expected?
>> > >>
>> > >> The branch is master and I cloned it today.
>> > >
>> > > I can confirm this on RHEL, I'll try to see what is going on.
>> >
>> > Curiously, it does not fail when building the 1.15.2 RPM on the same
>> > VM (after fixing the cert issues.)
>>
>> It looks like there are more tests in master which pushed the argument
>> list over the limit.
>>
>> I installed a newer version of automake from
>> https://koji.fedoraproject.org/koji/buildinfo?buildID=995648 with
>>
>> rpm -Uhv automake-1.15.1-3.fc28.noarch.rpm --nodeps
>>
>> to get around a dependency warning on RHEL7 and with this 'chmake' is
>> working for me.
>>
>> I opened https://bugzilla.redhat.com/show_bug.cgi?id=1542123 to hear if
>> this can be fixed in the RHEL7 version as well or if there is some
>> work-around for older automake versions.
>
>The RHEL package maintainer praiskup(a)redhat.com provided a work-around:
>
> $ make check am__is_gnu_make=:
>
>I tested this as
>
> $ chmake am__is_gnu_make=:
>
>and the tests are passing now.
>
>Lukas, do you know other distros/platforms which use automake-1.14 or
>older versions? Since the fix Pavel mentioned in only in automake-1.15
>and newer I wonder if it would be worth to add a configure check for
>older versions?
>
I did some testing and it does not affect el6
with even older automake 1.11.1-4.el6.
Because they don't do "test -n '$(MAKEFILE_LIST)'" in such old
version
The most interesting for me was that it does not happen within mock
or in our CI. And the explanation is quite simple.
In both cases, configure script was invoked with --disable-dependency-tracking.
And it is a little bit nicer workaround then make check am__is_gnu_make=:
I tested following oneliner on CentOS7 and it passed for me
reconfig --without-python3-bindings --disable-dependency-tracking && chmake
Thank you for checking this. --disable-dependency-tracking works for me
as well.
Given that there are now at least two workarounds and the issue will be
only seen when building git master on platforms with older automake
versions I think there is not needed to change anything in the code. As
we already talked about it would be sufficient to have a hint in the
developer documentation there this issue might happen on RHEL7 and how
to get around it.
bye,
Sumit
>
> LS