This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: set print object on should affect MI varobjs (PR mi/13393)
- From: xgsa <xgsa at yandex dot ru>
- To: Tom Tromey <tromey at redhat dot com>
- Cc: Eli Zaretskii <eliz at gnu dot org>, gdb-patches at sourceware dot org
- Date: Fri, 30 Mar 2012 23:12:17 +0300
- Subject: Re: set print object on should affect MI varobjs (PR mi/13393)
- References: <4ED92C05.9080803@yandex.ru> <m3mxanjiy0.fsf@fleche.redhat.com> <4EF22A4D.80703@yandex.ru> <20111221193630.GA28985@host2.jankratochvil.net> <4EF30389.5010008@yandex.ru> <20111223200951.GA19581@host2.jankratochvil.net> <4EF5D1C5.5010902@yandex.ru> <20120102022153.GA686@host2.jankratochvil.net> <4F07177C.5080201@yandex.ru> <m38vkfljq5.fsf@fleche.redhat.com> <4F32C009.2050103@yandex.ru> <87pqdma1mo.fsf@fleche.redhat.com> <4F412637.2010405@yandex.ru> <4F660CF6.9070008@yandex.ru> <8339953anu.fsf@gnu.org> <4F66DBD0.4000903@yandex.ru> <83vcm01obe.fsf@gnu.org> <4F6CADEF.4080700@yandex.ru> <87k422q8p7.fsf@fleche.redhat.com>
You didn't say whether you regression-tested this.
Yes, I didn't say it because I was not sure I have done it properly. I
have run the gdb.mi tests but there were a few unexpected failures.
However these failures are reproducible on gdb master branch without my
changes, so it should be a problem with my environment setup.
Unfortunately, I cannot figure out how to fix it. Moreover probably I
should run all the tests (not only gdb.mi) with "make check" but I
cannot found the overall statistics (only by every module), so it is
quite hard to compare the results. I have read the Testsuite chapter[1]
in gdb internals manual however that didn't help. Can you clarify me how
the regression testing should be done or point me where to read about it?
[1]
http://sourceware.org/gdb/current/onlinedocs/gdbint/Testsuite.html#Testsuite
Anton> + if (real_type_found)
Anton> + *real_type_found = 0;
Fixed
Too much indentation on the second line here.
Anton> + if (real_type_found)
Anton> + *real_type_found = 1;
Here too.
Fixed
Anton> + if (real_type_found)
Anton> + *real_type_found = 1;
Here too.
Fixed
Anton> + REAL_TYPE_FOUND is used to inform whether the real type was found
Anton> + (or just static type was used). The NULL may be passed if it is not
Anton> + necessary. */
Two spaces after periods.
Just write "NULL", not "The NULL".
Fixed.
Anton> + if (real_type_found)
Anton> + value = value_cast (var->type, value);
Too much indentation.
Fixed.
Anton> + int type_updated = update_type_if_necessary(existing, value);
Space before open paren.
Fixed.
Anton> + curr_type_str = varobj_get_type(var);
Space before open paren.
Fixed.
Anton> + if (update_type_if_necessary(v, new))
Anton> + r.type_changed = 1;
Too much indentation.
Fixed.
Anton.