This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [tree-ssa] Get rid of LOOP_EXPRs


Hello,

> > >  >> Can you put the bsi_next_in_bb change into a separate patch
> > >  >
> > >  >yes I could; but given that it is just that one line, it would be
> > >  >perhaps easier if you just imagined I did it :-)
> > > In general, it's best to send unrelated patches as separate messages.  However,
> > > since I've already got a test running with just this one patch installed,
> > > I'll go ahead and take responsibility to break it out.
> > 
> > sorry for this; I was in an extremely lazy mood yesterday :-(
> > 
> 
> Im actually the one being derelict here. I was testing a similar patch
> yesterday.  This change should be fine. It looks like a long standing
> oversight bug...
> 
> So, are we going to do anything with BIND_EXPR's? ie, wasn't there a
> thought of removing them, promoting the vars, and replacing them with
> debugging over a range info in dwarf2? We are spending a fair amount of
> time dealing with them which might be better spent eliminating them if
> we plan to get rid of all these container type constructs...

yes, this is my plan for close future; I have already made a first step
in this direction -- the patch to make variable-size variables more
independent on BIND_EXPRs.  Once they are fully independent, there imho
should not be any principial problems with the approach Richard
proposed.

Zdenek


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]