A quick update of where we stand on a few topics. 

Kernel Port: As you know the immediate, short-term goal is to get exec_init and 
run sh on the ODW target. The runtime loader (rtld) aka ld.so.1 is built and 
loading init and we're executing, but currently it is getting hung up in  
__lwp_mutex_lock(). Our userland locking is on the critical path at this 
moment. Brian is working his way through this but the lack of a comprehensive 
debugger has been a real productivity show stopper.  I given up on estimates to 
when we will get to sh since we are in new territory but we feel this is close.

Source Releases: As far as commits we are still at rev 97 the current debugging 
scenario has not yet yielded anything significant, once it does we will putback 
a meaningful update. Guy has done some cleanup in the shared libs and improved 
the overall build process. 

Build Machine: We are putting together a build machine from scratch. It's a 
v20z donated by the Labs that will be a publically available clone of the 
internal Labs development system. The key here is to recreate a functional 
build machine from all of the public sources mainly, opensolaris, ppc-dev 
project and blastwave. Just as important is the documentation of the process 
which is what we're focusing on this go around and making sure everything is 
current on the ppc-dev site. This is about 30% done at the moment, but I'll be 
unable to make progress on it again until the week of the 25th.
 
 
This message posted from opensolaris.org

Reply via email to