This is the mail archive of the
mailing list for the GDB project.
Re: Weird gdb problem
- From: Andrew Haley <aph at redhat dot com>
- To: gdb at sourceware dot org, java at gcc dot gnu dot org
- Date: Tue, 18 Jul 2006 17:06:50 +0100
- Subject: Re: Weird gdb problem
- References: <email@example.com>
Andrew Haley writes:
> I seem to have broken the debuginfo from gcj.
> When I step into a Java method I don't get info about args passed to
> methods. But this is only for debuginfo compiled into libgcj -- when
> I compile a test case myself I get correct args displayed. As in this
> one, where I step into p.p(), the argument "this" is displayed:
> (gdb) s
> p.p() (this=@2aaaad588e30) at Hello.java:7
> 0x00002aaaac057820 in java.lang.Object.Object() ()
> at /mnt/zebedee/aph/gcc/gcj-eclipse/libjava/java/lang/Object.java:360
> Note that the first step gave the value of the "this" argument, and
> the second one (in libgcj) didn't. Continuing:
> java.io.PrintStream.println(java.lang.String)void ()
> at /mnt/zebedee/aph/gcc/gcj-eclipse/libjava/java/io/PrintStream.java:482
> (gdb) info args
> No arguments.
> So, to recap: I get values of args to methods displayed in gdb, but
> only when I compile test cases. When stepping into libgcj itself, the
> values of args aren't displayed.
> So ... might anyone in gdb land like to guess how this might have broken?
We might have found the bug. Sorry for the noise...