This is the mail archive of the
mailing list for the GDB project.
Re: multiple runs with sim target
- From: Joel Brobecker <brobecker at adacore dot com>
- To: Mike Frysinger <vapier at gentoo dot org>
- Cc: gdb at sourceware dot org, toolchain-devel at blackfin dot uclinux dot org
- Date: Mon, 20 Jun 2011 21:24:45 -0700
- Subject: Re: multiple runs with sim target
- References: <firstname.lastname@example.org>
> this can be attributed to the "load" loading up the ELF contents into the
> relevant memory regions, but "run" only executes it. so if the program
> modifies memory (such as .data), things dont get reloaded and behavior is not
> so nice.
I think that this is expected. We have the same sort of issue
on VxWorks as well, where the data segment is initialized at
> based on the code, i guess this is expected behavior. the trouble
> comes in with the gdb testsuites that explicitly try and verify that
> this works. such as gdb.base/a2-run.exp.
IMO, that's a problem with the testcase. I think it should reload
for the platforms that need it.