Master is updated

[drf5n]

See http://git.savannah.gnu.org/cgit/simulavr.git for recent history of git://git.savannah.nongnu.org/simulavr.git

One of the features I like is that it compiles on my Mac, but then I'm not sure how to invoke the standard dist and get the serial-in/serial-out like your fork provides with SIMINFO_* mechanism.

[Traumflug]

I just tried to rebase on more recent sources of simulavr and the bad news is, somebody rewrote the whole ELF reading stuff in a fashion which no longer allows to apply the SIMINFO patches. The mechanism would have to be rewritten from scratch.

[Traumflug]

The traumflug branch is rebased against latest upstream now (and actually required a re-write of the ELF reading part), so this issue should be solved, including compilation on the Mac.

Thanks a lot for bringing this to attention. If further issues arise, please open a new issue.

[drf5n]

Compiled on my mac, but I needed two changes:

1) Remove '-o $@' from doc/Makefile.in rule .texi.pdf to compile past an error in generating doc/simulavr.pdf

2) ./configure CPPFLAGS=-DGTEST_USE_OWN_TR1_TUPLE=1 as in https://code.google.com/p/open-vcdiff/issues/detail?id=41 to compile past an error in compiling the unit test stuff.

[Traumflug]

Excellent. These changes sound like it'd be fine to just build until the build process stops, no matter wether it completes or fails at some point. Once src/simulavr appears it should work.

For a moment I wanted to forward these bug reports to SimulAVR's bug tracker but then I stepped back, because it'd be good manners to provide error logs and to test changes, which I can't.

[Traumflug]

I've updated build instructions on the RepRap wiki SimulAVR page, hinting towards ignoring such errors.

[drf5n]

The upstream branch fails before I get there with SWIG.3.0.0 per https://savannah.nongnu.org/bugs/?42595

My first error is in Make with: echo "Error: SWIG version >= 1.3.18 is required. You have 3.0.0. You should look at http://www.swig.org" ; false -c++ -o simulavr_wrap.cxx ./simulavr.i

I see that I had earlier tried a sloppy workaround on your code to get past the SWIG error:

diff --git a/configure.ac b/configure.ac
index dcaf05d..a82cb85 100644
--- a/configure.ac
+++ b/configure.ac
@@ -58,7 +58,7 @@ if test ${CXX_PROG} = no ; then
   # AC_PROG_CXX sets g++, if nothing is found, but if not found make will fail
   AC_MSG_ERROR([C++ compiler ${CXX} not found],1)
 fi
-AC_PROG_SWIG(2.0.0)
+AC_PROG_SWIG(3.0.0)
 if test "x$SWIG_LIB" = "x"; then
   # fall back to SWIG version 1.3.x in case of SWIG 2.0.x not found!
   AC_MSG_NOTICE([Don't worry, we can try to find SWIG 1.3.x!])

I posted https://savannah.nongnu.org/bugs/index.php?42621 for the pdf build error and a comment on https://savannah.nongnu.org/bugs/?42595 for these. I didn't run into the error requiring the tuple flag on master.