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: [win32] wrong solib from/to addresses


> > I had a similar patch here, that I dumped in favor of a rewrite
> > of win32 solibs on top of Daniel's pending solib-target.c.
> 
> That sounds like the right way to go to me.  Can we hold of on this
> until Daniel's changes are in?

As far as I'm concerned, I'm happy to withdraw this patch. I don't
think that the consequences of these wrong addresses are that bad,
so there is certainly no huge hurry.

At the AdaCore level, we have a tiny local change in the x86 unwinder
where we decided to take a different choice than Mark did, and that
choice involves a heuristic that needs to determine  whether we're
in a DLL or not. So this patch is a bit more urgent for us, but we
have it installed in our local tree, so we're good to go.

As far as the FSF tree is concerned, apart from the fact that
the addresses are wrong when the users asks for that information,
the only other visible symptom is a minor error in the backtrace.
For frames that involve a DLL symbol, the name of the DLL where
this symbol comes from can be either printed wrong, or not printed
at all. I had never noticed this until I hit this issue with our
local patch when testing on Vista.

So patch withdrawn for now, and the puck moves to Daniel and Pedro.

-- 
Joel


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