pkg: oonf_api: make patch not working

[LudwigKnuepfer]

I followed #630 (comment) and got:

make patch
make: *** No rule to make target '/home/lo/native/RIOT/pkg/oonf_api/oonf_api', needed by '/home/lo/native/RIOT/pkg/oonf_api/oonf_api/Makefile'.  Stop.

After mkdir oonf_api I get:

# Here you apply your patch.
cd "/home/lo/native/RIOT/pkg/oonf_api/oonf_api" && git am "../0001-add-RIOT-support.patch"; cd "/home/lo/native/RIOT/pkg/oonf_api/oonf_api" && git am "../0002-port-tests-to-riot.patch"; cd "/home/lo/native/RIOT/pkg/oonf_api/oonf_api" && git am "../0003-port-example-to-riot.patch";
Applying: add RIOT support
/home/lo/native/RIOT/.git/rebase-apply/patch:202: trailing whitespace.
#ifndef DONT_HAVE_SIN6_SCOPE_ID      
error: Makefile: already exists in index
error: src-api/common/autobuf.c: does not exist in index
error: src-api/common/common_types.h: does not exist in index
error: src-api/common/daemonize.c: does not exist in index
error: src-api/common/netaddr.c: does not exist in index
error: src-api/common/netaddr.h: does not exist in index
error: src-api/rfc5444/rfc5444_print.c: does not exist in index
Patch failed at 0001 add RIOT support
The copy of the patch that failed is found in:
   /home/lo/native/RIOT/.git/rebase-apply/patch
When you have resolved this problem, run "git am --continue".
If you prefer to skip this patch, run "git am --skip" instead.
To restore the original branch and stop patching, run "git am --abort".
previous rebase directory /home/lo/native/RIOT/.git/rebase-apply still exists but mbox given.
previous rebase directory /home/lo/native/RIOT/.git/rebase-apply still exists but mbox given.
Makefile:29: recipe for target '/home/lo/native/RIOT/pkg/oonf_api/oonf_api/Makefile' failed
make: *** [/home/lo/native/RIOT/pkg/oonf_api/oonf_api/Makefile] Error 1

Now:

git st
On branch master
Your branch is up-to-date with 'origin/master'.

You are in the middle of an am session.
  (fix conflicts and then run "git am --continue")
  (use "git am --skip" to skip this patch)
  (use "git am --abort" to restore the original branch)

Would be nice to add some kind of error handling..

[LudwigKnuepfer]

While you're at it, you can add this update for #880 :

diff --git a/pkg/oonf_api/0001-add-RIOT-support.patch b/pkg/oonf_api/0001-add-RIOT-support.patch
index 971a8a2..bbfe37d 100644
--- a/pkg/oonf_api/0001-add-RIOT-support.patch
+++ b/pkg/oonf_api/0001-add-RIOT-support.patch
@@ -54,7 +54,7 @@ index 0000000..cf66baa
 +
 +all:
 +      mkdir -p $(BINDIR)
-+      @for i in $(DIRS) ; do $(MAKE) -C $$i ; done ;
++      @for i in $(DIRS) ; do $(MAKE) -C $$i || exit 1; done ;
 +
 +clean:
 diff --git a/external/regex/Makefile b/external/regex/Makefile

[benpicco]

Now this is very strange, oonf_api should be checked out and created by git.

I can't reproduce this,

benpicco@rechenknecht2k7:~/riot/RIOT/pkg/oonf_api$ ls
0001-add-RIOT-support.patch  0002-port-tests-to-riot.patch  0003-port-example-to-riot.patch  Makefile  Makefile.include
benpicco@rechenknecht2k7:~/riot/RIOT/pkg/oonf_api$ make patch 
git clone git://olsr.org/oonf_api.git /home/benpicco/riot/RIOT/pkg/oonf_api/oonf_api && \
        cd /home/benpicco/riot/RIOT/pkg/oonf_api/oonf_api && git checkout -b v0.3.0
Klone nach '/home/benpicco/riot/RIOT/pkg/oonf_api/oonf_api'...
remote: Counting objects: 6528, done.
remote: Compressing objects: 100% (3976/3976), done.
remote: Total 6528 (delta 5273), reused 3038 (delta 2450)
Empfange Objekte: 100% (6528/6528), 1.63 MiB | 1009.00 KiB/s, done.
Löse Unterschiede auf: 100% (5273/5273), done.
Checking connectivity... done
Gewechselt zu einem neuen Zweig 'v0.3.0'
cd "/home/benpicco/riot/RIOT/pkg/oonf_api/oonf_api" && git am "../0001-add-RIOT-support.patch"; cd "/home/benpicco/riot/RIOT/pkg/oonf_api/oonf_api" && git am "../0002-port-tests-to-riot.patch"; cd "/home/benpicco/riot/RIOT/pkg/oonf_api/oonf_api" && git am "../0003-port-example-to-riot.patch";
Wende an: add RIOT support
Wende an: port tests to riot
Wende an: port example to riot

$ ls
0001-add-RIOT-support.patch  0002-port-tests-to-riot.patch  0003-port-example-to-riot.patch  Makefile  Makefile.include  oonf_api

afterwards I can run make in e.g. RIOT/pkg/oonf_api/oonf_api/examples/rfc5444_reader_writer and a native binary is produced.

I've created #895 to make the make loop abort on error as you suggested. I guess someone has to ack it before I can merge it? (Yay, I've got the power to merge :D)

What error handling do you suggest with git btw? I looks like it tried to apply the patches to an empty folder (and therefore the underlying RIOT git), but this should never happen to begin with.

[LudwigKnuepfer]

Could you try checking out RIOT-OS/master in a fresh dir and repeat it there?

[benpicco]

Yes, when I do

cd /tmp
git clone [email protected]:RIOT-OS/RIOT.git
cd RIOT/pkg/oonf_api/
make patch
cd oonf_api/examples/rfc5444_reader_writer/
make

it builds.

[OlegHahm]

.-(~)
`--> cd /tmp
.-(/tmp)
`--> git clone [email protected]:RIOT-OS/RIOT.git
Cloning into 'RIOT'...
remote: Reusing existing pack: 19759, done.
remote: Counting objects: 10, done.
remote: Compressing objects: 100% (10/10), done.
remote: Total 19769 (delta 3), reused 1 (delta 0)
Receiving objects: 100% (19769/19769), 6.69 MiB | 1.69 MiB/s, done.
Resolving deltas: 100% (11303/11303), done.
Checking connectivity... done.
.-(/tmp)
`--> cd RIOT/pkg/oonf_api/
.-(/tmp/RIOT/pkg/oonf_api)
`--> make patch
make: *** No rule to make target '/tmp/RIOT/pkg/oonf_api/oonf_api', needed by '/tmp/RIOT/pkg/oonf_api/oonf_api/Makefile'.  Stop.

[benpicco]

Ok, when I try to build it with make 4.0 on Arch I get the same error, are you using that version too?
I've build it with Ubuntu 13.10 which still has make 3.81, so I didn't notice that.

[benpicco]

Does it build when you change
PKG_NAME=oonf_api in the Makefile to PKG_NAME:=oonf_api ?

[LudwigKnuepfer]

No, it doesn't.

[benpicco]

sorry, I was in the wrong terminal…
But removing the / from $(PKG_NAME)/: seems to work, I guess the example Makefile.git will need this as well.

[LudwigKnuepfer]

Yup, could you include it?

[benpicco]

it's in the pull request, can I merge it?

[LudwigKnuepfer]

See discussion over at the PR