Bug#826765: alljoyn-services-1504: FTBFS on arm and i386, patch attached

2016-06-10 Thread Adam Conrad
On Fri, Jun 10, 2016 at 08:58:20PM +0200, Thorsten Alteholz wrote:
> 
> thanks for your report. Is there really an i386 architecture
> available in Ubuntu?

There is indeed.  It happens to also be i686 right now, as Debian is,
but from the dpkg archtable point of view, all i?86 is i386, hence my
regex.

> > * Drop build_core/conf/linux/arm to stop misconfiguration on arm{el,hf}
> 
> But if you just remove those files during the clean target, you can
> not build this package twice in a row, right?

Why would you say that?  You always clean before a first build, and
once the files are gone, they're gone.

Like I said in the bug report, that arm target has no relevance at all
to Debian, so shipping those bits in the -dev packages (as you do), is
asking for trouble.  People configuring with CPU=arm should get a sane
result on Debian, IMO.

I'm not hugely picky on the matter, though.  If you prefer to keep the
incorrect cross/ARMv6t2 configs there for reasons unexplained, so long
as everything still builds and works, meh.  It's just "wrong". :P

... Adam



Bug#826765: alljoyn-services-1504: FTBFS on arm and i386, patch attached

2016-06-10 Thread Thorsten Alteholz

Hi Adam,

thanks for your report. Is there really an i386 architecture available in 
Ubuntu?


On Wed, 8 Jun 2016, Adam Conrad wrote:

 * Drop build_core/conf/linux/arm to stop misconfiguration on arm{el,hf}


But if you just remove those files during the clean target, you can not 
build this package twice in a row, right?



The arm target in the upstream source has no relevance at all to the
arm ports in Debian (not only does it require cross-compiling, but it
also targets the raspberry pi specifically).


I think the arm target is for OpenWRT. Does my remapping do any harm on 
the other arm architectures?


  Thorsten



Bug#826765: alljoyn-services-1504: FTBFS on arm and i386, patch attached

2016-06-08 Thread Adam Conrad
Package: alljoyn-services-1504
Version: 15.04-1
Severity: important
Tags: patch
User: ubuntu-de...@lists.ubuntu.com
Usertags: origin-ubuntu yakkety ubuntu-patch



In Ubuntu, the attached patch was applied to achieve the following:

  * Set CPU to x86 when dpkg-architecture -qDEB_HOST_GNU_CPU reports i?86
  * Drop build_core/conf/linux/arm to stop misconfiguration on arm{el,hf}

The arm target in the upstream source has no relevance at all to the
arm ports in Debian (not only does it require cross-compiling, but it
also targets the raspberry pi specifically).  Given that, it's not
reasonable to even ship those bits in Debian, so I took a different
approach than you did in fixing this in Ubuntu.

My fix depends on the same fix being made to alljoyn-core first, which
I will submit as another bug and reference this one.

... Adam


-- System Information:
Debian Release: stretch/sid
  APT prefers yakkety
  APT policy: (500, 'yakkety')
Architecture: amd64 (x86_64)
Foreign Architectures: i386

Kernel: Linux 4.4.0-21-lowlatency (SMP w/4 CPU cores; PREEMPT)
Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash
Init: systemd (via /run/systemd/system)
diff -Nru alljoyn-services-1504-15.04/debian/rules alljoyn-services-1504-15.04/debian/rules
--- alljoyn-services-1504-15.04/debian/rules	2016-06-04 14:46:12.0 -0600
+++ alljoyn-services-1504-15.04/debian/rules	2016-06-07 09:14:40.0 -0600
@@ -10,7 +10,7 @@
 
 DEB_HOST_MULTIARCH ?= $(shell dpkg-architecture -qDEB_HOST_MULTIARCH)
 
-CPU=$(shell dpkg-architecture -qDEB_HOST_GNU_CPU)
+CPU=$(shell dpkg-architecture -qDEB_HOST_GNU_CPU | sed -e 's/i.86/x86/')
 OS=linux
 
 export VERSION := 15.04b
@@ -59,6 +59,7 @@
 	dh_auto_build
 
 override_dh_auto_clean:
+	rm -rf build_core/conf/linux/arm
 	rm -f sample_apps/.sconsign.dblite
 	rm -rf sample_apps/.sconf_temp
 	rm -f services_common/.sconsign.dblite