[ros-dev] Real HW Regression. Richard, this is for you

Daniel Reimer daniel.reimer at stud-mail.uni-wuerzburg.de
Fri Feb 26 12:36:11 CET 2010


first my congratz to the great work you do recently. HAL really needed 
your attention :-D But... I stumbled over a nice regression last night. 
I wanted to install HEAD on the Notebook we will use on Chemnitzer Linux 
Tage to show ROS and failed to get to 1st stage at all. After some 
regression testing I tracked this down to this range: r45319 still works 
up to 3rd stage. r45324 fails to even get in 1st stage. The 5 revs in 
between dont build at all, because r45324 was a build fix for r45320 and 
this one was a biiig rewrite by Richard. The screen stays black and when 
I blindly type "CONT" , I get: ***STOP: 0x0000007B 
(0xF1BCDB60,0xC0000034,0x00000000,0x00000000) Its a really big problem 
to get more output from it because it has no Serial Ports at all. We 
tried to set /DEBUG DEBUGPORT=SCREEN /SOS but nothing shows up. So if 
someone finds a possible reason for that regression or a nice way to get 
the needed debuggin Information, plz tell me. It would be very important 
to get ROS run on it again.

Machine is a:
Acer Extensa 4100 wmli / Aspire 1690 wmli
Acer Crane Mainboard
Intel Pentium III Mobile CPU
ATI Mobility Radeon X700

Thx for the help.

More information about the Ros-dev mailing list