Hello,
I am currently looking into Rolf Pfister's patches, and I stumbled upon
the symbol blacklisting issue. As far as I understand, the blacklisting
was introduced by Burkhard Arenfeld, to prevent the symbols that were
undefined at the first pass to trigger .ifdef sections on the second
pass. In other words, compiling the following code:
== cut ==
.device ATtiny2313
.cseg
reset:
nop
nop
.ifdef test
.message "test is defined"
.else
.message "test is not defined"
.endif
nop
nop
.define test
.ifdef test
.message "test is defined"
.else
.message "test is not defined"
.endif
nop
nop
rjmp reset
== cut ==
should emit "test is not defined" first, then "test is defined". Which
seems logical to me. However, in the current master branch, the code for
checking the blacklist is disabled. So compiling the above results in
"test is defined" being printed twice. Which is not good, IMO.
So, in the current snapshot, the blacklist is populated, but not being
checked against. The Rolf's patches seem to remove blacklisting
altogether. I don't think that either of these is a good idea.
I am thinking of re-enabling the blacklist checking, and only borrow the
.undef part from the Rolf's code (I hope it does not interfere with
blacklisting) for my current working version of avra.
Opinions on this issue would be appreciated.
--
Sincerely yours,
Alexey
------------------------------------------------------------------------------
One dashboard for servers and applications across Physical-Virtual-Cloud
Widest out-of-the-box monitoring support with 50+ applications
Performance metrics, stats and reports that give you Actionable Insights
Deep dive visibility with transaction tracing using APM Insight.
http://ad.doubleclick.net/ddm/clk/290420510;117567292;y
_______________________________________________
Avra-user mailing list
Avra-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/avra-user
