This is the mail archive of the gdb-patches@sources.redhat.com mailing list for the GDB 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: [PATCH] implements MI "-file-list-exec-sections" (updated)


> > Resubmitting updated (against current CVS) patch. Includes changes to
> > GDB/MI and docs. No test case yet.
> > Maybe it's worth mentioning, that GDB/CLI output for "info file" changes
> > with this patch which will probably break programs that parse this. (I
> > don't think such programs exist)

The problem here is that:
1. I have converted the CLI output system for "info file" (*_files_info()) to 
use ui_out. The MI uses the same procedure (target_info()) to process 
"-file-list-exec-sections" (but a different ui_out instance)
2.  "info file" deals with targets
3. targets have short (well machine readable) and long (well human readable, 
localizable???) names
4. all fields provided in MI output end up in CLI output (that's the way 
ui_out works)
5. it is desirable to have short names in MI output but short names are 
currently not part of the CLI output.

>
> I get the feeling that changing the CLI output of "info file" is a bad
> idea. My hunch is that some front ends to GDB are calling that function.

I don't think GDB development should be limited by keeping some CLI-standard 
output. And GDB/MI and not GDB/CLI was made for front-ends after all. In a 
perfect world...

Stefan


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