gnash-dev
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Gnash-dev] updated Android toolchain


From: Rob Savoye
Subject: [Gnash-dev] updated Android toolchain
Date: Mon, 15 Mar 2010 23:20:36 -0600
User-agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9.1.8) Gecko/20100301 Fedora/3.0.3-1.fc12 Lightning/1.0b1 Thunderbird/3.0.3

I've just uploaded a fresh Android toolchain to gnashdev.org. This one
is greatly improved over the previous one, as it now works well works
with the Android NDK. Better yet, I don't have any files copied from the
NDL or SDK in the toolchain, so it's much cleaner now, and easier to
update, and doesn't violate the SDK usage agreement.

Since I had built a new GCC 4.5 toolchain, I recompiled all the
dependencies Gnash needs for an SDL-AGG build. I managed to get boost
1.42.0 to cross compile, which is always a major miracle. It took a bit
of hacking as boost kept thinking the weird pthread implementation in
Bionic meant no pthread support. I had to do a bit of hacking to the
other libraries, as Bionic has a few definitions in different headers,
and headers in different directories. I kept notes on all the hacks, and
threw up an initial wiki page on how to port things to Android:
http://wiki.gnashdev.org/AndroidDependencies. It's amazing how poor the
support is in most packages for cross compiling.

The shared memory routines, shmctl, semget, semop, shmdt, shmat, shmget,
semctl, are undefined, so for now I stubbed them out. Digging into it,
there is syscall support for these buried deep, so I should be able to
get that fixed up soon.

I'm having an issue with "dynamic_cast' not permitted with -fno-rtti",
but there should be rtti support for the arm5, so I'm trying to track
that down as the last problem before Gnash fully builds again. Last time
I kindof hacked everything into existence, this time it's building
correctly.

        - rob -




reply via email to

[Prev in Thread] Current Thread [Next in Thread]