Testing impedance layer

Check Nyah check.nyah at gmail.com
Fri Aug 29 11:22:47 PDT 2014


Hello,

I am testing the impedance layer for the Xserver and I've been running xts
and valgrind on the xserver for the past two days, and i've received alot
of errors. As for the Valgrind Errors <http://paste.kde.org/plbai4or6>, the
main problem results from the fact that the following functions( LogVWrite
(log.c:292), ErrorF (log.c:652),  _damageRegionAppend (damage.c:266)
_damageDamageBox (damage.c:386) damagePolyFillRect , (damage.c:1259)
miPaintWindow (miexpose.c:642),  miWindowExposures (miexpose.c:471)
KdXVWindowExposures (kxv.c:1012) MapWindow (window.c:2626), have uninitialized
variables, while trying to initialize some of them I was getting serious
compilation errors and I realized that there were other similar variables
which were not initialized at the point of declaration. As for the xts
tests, I even got a segmentation fault <http://paste.kde.org/pfzo6mx2f> on
the Xephyr, which I"m using the backtrace to solve.

In order to test the impedance layer, here is my current plan.

First and foremost, when i resolve the segmentation fault above, I want to
focus exclusively on the xts/xproto test suite which focuses on the
Xserver  protocol and hence the Impedance layer, then after fixing the
errors, I'll move to the xts/xiproto tests, when I finish this guess the
impedance layer should be complete.

I'll continue with separating core aspects of the objects, between the
protocol layers and driver layers while running the appropriate test suites
to check and fix errors. I'll like to know if this plan is ok? or is there
anything I'm missing.

I've ran the xts/Xproto test <http://paste.kde.org/pka59dwdf> and I'm
currently interpreting the results to fix the problems, I'll like to know
what to do in test cases where the output is unresolved, as for the failed
test which occurred for the MapWindow, setModifierMapping, I'll like to the
various test cases which run directly against the impedance layer.

Cheers!
Nyah
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.x.org/archives/xorg-devel/attachments/20140829/0090a175/attachment.html>


More information about the xorg-devel mailing list