> On 2011-01-06 11:42:18, Nathan Binkert wrote: > > SConstruct, line 379 > > <http://reviews.m5sim.org/r/366/diff/3/?file=8481#file8481line379> > > > > Can you derive from object? I really hate classic classes as they make > > various things annoying and the cause is not always obvious.
No problem, it was just an oversight. > On 2011-01-06 11:42:18, Nathan Binkert wrote: > > SConstruct, line 418 > > <http://reviews.m5sim.org/r/366/diff/3/?file=8481#file8481line418> > > > > 99 seems like a lot and goes against the idea of being concise. I > > don't care much though. > > > > Would anyone object to [%-8s] instead? the right justified seems odd. The 99 is just a practical approximation of infinity, not meant to be hit in practice... the main idea is to give a parameter you can explicitly set if you don't want all (or any) of the sources listed, so I can cleanly replace the TRANSFORM_NOSRC and TRANSFORM_1SRC things I had in the earlier version. > On 2011-01-06 11:42:18, Nathan Binkert wrote: > > SConstruct, line 481 > > <http://reviews.m5sim.org/r/366/diff/3/?file=8481#file8481line481> > > > > out of curiosity, can you do the following instead? > > main['CCCOMSTR'] = TRANSFORM("CC") > > > > Certainly, I am pretty sure that the parameter to MakeAction/Action > > could be the callable and not the string. (String substitution just slows > > down SCons) TRANSFORM("CC") doesn't work, but Transform("CC") does. I had tried this before and it didn't work so I gave up on it, but I must have had some other error at the time. > On 2011-01-06 11:42:18, Nathan Binkert wrote: > > SConstruct, line 406 > > <http://reviews.m5sim.org/r/366/diff/3/?file=8481#file8481line406> > > > > This is probably pedantic, but these could be @classmethod (and self > > should be cls) Yea, I knew that, but I didn't see any point to it, so I figured why bother with the extra verbiage. If there's an advantage to it I'd be glad to do it though. The one thing I would like would be to define color() in a way where I could use it immediately, so I could do 'Arrow = color(arrow_color, " -> ")' instead of the current expression, but I couldn't find a reasonable way to get that to work. > On 2011-01-06 11:42:18, Nathan Binkert wrote: > > SConstruct, line 383 > > <http://reviews.m5sim.org/r/366/diff/3/?file=8481#file8481line383> > > > > I hate lines like this. I'll register my annoyance, but won't continue > > further discussion. This is one of the few places I feel like lines like this are useful... this whole colorization thing already chews up way too many lines for the value it brings IMO :-). I'm also tempted to do this: def color_pfx(self, s): return self.color(self.pfx_color, s) def color_srcs(self, s): return self.color(self.srcs_color, s) def color_tgts(self, s): return self.color(self.tgts_color, s) just because I hate repetitive stuff chewing up vertical screen space. - Steve ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: http://reviews.m5sim.org/r/366/#review632 ----------------------------------------------------------- On 2011-01-06 11:15:30, Steve Reinhardt wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > http://reviews.m5sim.org/r/366/ > ----------------------------------------------------------- > > (Updated 2011-01-06 11:15:30) > > > Review request for Default, Ali Saidi, Gabe Black, Steve Reinhardt, and > Nathan Binkert. > > > Summary > ------- > > scons: show sources and targets when building. > > I like the brevity of Ali's recent change, but the ambiguity of > sometimes showing the source and sometimes the target is a little > confusing. This patch makes scons typically list all sources and > all targets for each action, with the common path prefix factored > out for brevity. It's a little more verbose now but also more > informative. > > > Diffs > ----- > > SConstruct 9f9e10967912 > src/SConscript 9f9e10967912 > src/arch/SConscript 9f9e10967912 > src/arch/isa_parser.py 9f9e10967912 > > Diff: http://reviews.m5sim.org/r/366/diff > > > Testing > ------- > > quick regressions pass > > > Screenshots > ----------- > > sample colorized output > http://reviews.m5sim.org/r/366/s/1/ > > > Thanks, > > Steve > >
_______________________________________________ m5-dev mailing list m5-dev@m5sim.org http://m5sim.org/mailman/listinfo/m5-dev
