shark improperly linked against llvm 2.4 (and first steps on ARM)

[Robert Schuster]

Hi Xerxes,

Xerxes Rånby schrieb:
> > Hi Robert!
> >
> > It is fun to see you taking on the task of getting shark to run on ARM.
> > My shark adventures have only been able to run java -version and then
> > crash so im thrilled to see that you have gotten your shark jvm past
> > that part!
> > I have also got a beagleboard and are trying to get shark running on it.
If you are using OE and the Jalimo overlay that should be as easy as
'bitbake openjdk-6-shark', then waiting and then installing the IPKs.  :)

> > How did you build your llvm 2.4? I have been trying to build llvm 2.4
> > using the debian eperimental sourcetree for armel and my builds of llvm
> > stops with the error undefined reference to `__register_frame' during
> > compile of TargetSelect.cpp. Since i dont know how ARM EABI
> > registerframes are supposed to look (if the even exists) i have sort of
> > removed all references to __register_frame in my llvm sourcetree and
> > hoped for the best with the thinking they are only used for debugging.
I also have no idea how to deal with that properly. Googling for the
problem resulted in finding someone who disabled it.

I am using the two attached patches. The first disables __register_frame
usage if EABI is in use and enforces OE's CFLAGS and LDFLAGs.

The static initialization patch fixes a real problem in the llvm
codebase. According to cppreference.com static initialization order in
C++ is undefined.

> > Im doing native compiles yet i suspect you are crosscompiling as you
> > usually do, please clarify if you are crosscompiling llvm2.4 for arm and
> > hints how to do so would be much appreciated.  :)
Sure I let OE cross-compile llvm. llvm 2.4 turned out to be a bit
difficult, since it insists on compiling some stuff for the build host.
I disabled the respective parts of the makefile to get around this issue
(see fix-build.patch). Furthermore I build those tools separately. If
you check out OE's git repository have a look at packages/llvm/llvm_2.4.bb.

I do a bit of hackery to get the cross-build done and also need to fix
some paths in llvm-config. It is all written down as shell script in the
llvm 2.4 bitbake file.

> > Would you ming creating a small patch to change  engine -> nativecodegen
> > and send the patch to the list for inclusion with icedtea?
> > It would be great in my humble opinion.
Sure, I only wanted to know what others say about this.  :)

>> >> root at beagleboard:~# java HelloWorld
>> >> Hello World!
>> >>
> > Cool! i have only been able to see java -version run  :)
If you want to try out the some binaries I have uploaded armv6
compatible ones here:
http://jalimo.evolvis.org/repository/bl-armv6/

>> >> Do you think that assertion is a shark or an llvm issue?
> > I think that is a llvm issue as most stuff in llvm:: namespace are.
Interesting, I will then ask over there how I can fix this.

Regards
Robert

-------------- next part --------------
A non-text attachment was scrubbed...
Name: fix-build.patch
Type: text/x-patch
Size: 2812 bytes
Desc: not available
Url : http://mail.openjdk.java.net/pipermail/distro-pkg-dev/attachments/20090126/3861599d/fix-build.patch 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: fix-static-initialization.patch
Type: text/x-patch
Size: 2109 bytes
Desc: not available
Url : http://mail.openjdk.java.net/pipermail/distro-pkg-dev/attachments/20090126/3861599d/fix-static-initialization.patch 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 260 bytes
Desc: OpenPGP digital signature
Url : http://mail.openjdk.java.net/pipermail/distro-pkg-dev/attachments/20090126/3861599d/signature.asc