This is the mail archive of the gdb-patches@sourceware.org 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: checked in: Re: RFC: solib.c:solib_map_sections so->so_name clobbering


Hi,

It looks like this commit introduced a small regression for MI and shared libraries, and the testsuite does not cover this case.

Suppose we have GDB running on a host and a stub/gdbserver running on a separate remote target. Suppose shared libraries (for symbols) on the target are located in a different path compared to the host, say, <host_path> and <target_path>.

During debugging, eventually the shared libraries will be loaded and we used to see a shared library load notification like the following:

=library-loaded,id="<target_path>/libhello.so",target-name="<target_path>/libhello.so",host-name="<host_path>/libhello.so",symbols-loaded="0",thread-group="i1"

After this commit, this is what we see:

=library-loaded,id="<target_path>/libhello.so",target-name="<target_path>/libhello.so",host-name="<target_path>/libhello.so",symbols-loaded="0",thread-group="i1"

So it looks like we've lost information about the shared library's path on the host, which may not be a big deal for CLI GDB, but may confuse consumers of MI output.

I gave this a quick thought, but reverting the change seemed like the most obvious solution.

But since this change affects darwin, maybe Joel has a different idea?


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