Bug 1375213

Summary:	supermin doesn't seem to work
Product:	[Fedora] Fedora	Reporter:	Dan Horák <dan>
Component:	supermin	Assignee:	Richard W.M. Jones <rjones>
Status:	CLOSED RAWHIDE	QA Contact:	Fedora Extras Quality Assurance <extras-qa>
Severity:	unspecified	Docs Contact:
Priority:	unspecified
Version:	rawhide	CC:	ptoscano, rjones
Target Milestone:	---
Target Release:	---
Hardware:	s390x
OS:	Unspecified
Whiteboard:
Fixed In Version:	supermin-5.1.16-6.fc26	Doc Type:	If docs needed, set a value
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2016-10-10 07:29:26 UTC	Type:	Bug
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:
Bug Depends On:
Bug Blocks:	467765

Description Dan Horák 2016-09-12 13:32:56 UTC

Description of problem:
Although supermin builds fine on s390x with the interpreted Ocaml, it doesn't seem to work correctly. I've found it when building libguests that failed with

...
checking if we should build the appliance... yes
checking for supermin... /usr/bin/supermin
checking for --with-supermin-packager-config option... not set
checking for --with-supermin-extra-options option... not set
checking supermin is new enough... configure: error: supermin >= 5.1 must be installed, your version is too old
error: Bad exit status from /var/tmp/rpm-tmp.VHAmuM (%build)
...

with

...
configure:61220: checking for supermin
configure:61238: found /usr/bin/supermin
configure:61251: result: /usr/bin/supermin
configure:61262: checking for --with-supermin-packager-config option
configure:61271: result: not set
configure:61279: checking for --with-supermin-extra-options option
configure:61288: result: not set
configure:61303: checking supermin is new enough
Unknown option --version.
configure:61306: error: supermin >= 5.1 must be installed, your version is too old
...

in config.log

Version-Release number of selected component (if applicable):
supermin-5.1.16-4.fc25

How reproducible:
100%

Steps to Reproduce:
1. supermin --version

Actual results:
[sharkcz@devel11 ~]$ supermin --version
Unknown option --version.

Expected results:
[dan@eagle ~]$ supermin --version
supermin 5.1.16

Additional information:
Looks to me as it install only a wrapper, but no actual supermin binary, because

[sharkcz@devel11 ~]$ supermin
No bytecode file specified.

while on x86_64 I see

[dan@eagle ~]$ supermin
supermin: you must use --prepare or --build to select the mode

for koji build and logs please see http://s390.koji.fedoraproject.org/koji/taskinfo?taskID=2345310

Comment 1 Richard W.M. Jones 2016-09-12 13:50:40 UTC

"No bytecode file specified." happens because running strip on
bytecode files damages them.

  https://bugzilla.redhat.com/show_bug.cgi?id=435559

To avoid this we usually disable stripping in the RPM:

  %ifnarch %{ocaml_native_compiler}
  %global __strip /bin/true
  %endif

Comment 2 Richard W.M. Jones 2016-09-12 13:52:19 UTC

BTW there is rumoured to be an s390x backend for the OCaml compiler,
although I have not used it and it's not in our OCaml package.
http://alan.petitepomme.net/cwn/2015.07.14.html#2

Comment 3 Dan Horák 2016-09-15 09:22:39 UTC

(In reply to Richard W.M. Jones from comment #1)
> "No bytecode file specified." happens because running strip on
> bytecode files damages them.
> 
>   https://bugzilla.redhat.com/show_bug.cgi?id=435559
> 
> To avoid this we usually disable stripping in the RPM:
> 
>   %ifnarch %{ocaml_native_compiler}
>   %global __strip /bin/true
>   %endif

looks like
%global debug_package %{nil}
is also required for supermin to work ...

Comment 4 Richard W.M. Jones 2016-09-15 09:48:42 UTC

Quite likely.  Please make sure that statement is conditional
on %{ocaml_native_compiler} too.

Comment 5 Dan Horák 2016-09-15 10:54:41 UTC

another solution - use dietlibc and no workarounds are needed