This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [PATCH] Mark object files with "target:" filenames as OBJF_NONLOCAL_FILENAME
- From: Gary Benson <gbenson at redhat dot com>
- To: Doug Evans <dje at google dot com>
- Cc: gdb-patches <gdb-patches at sourceware dot org>
- Date: Tue, 14 Apr 2015 22:30:05 +0100
- Subject: Re: [PATCH] Mark object files with "target:" filenames as OBJF_NONLOCAL_FILENAME
- Authentication-results: sourceware.org; auth=none
- References: <1428952063-2121-1-git-send-email-gbenson at redhat dot com> <CADPb22THqPa3WzQUK13Oh_-AF0D3GNitSmUBQcj7mZoBp0pL8A at mail dot gmail dot com> <20150414114129 dot GB4660 at blade dot nx> <CADPb22S3VfMBnG1gwdr8omKicWSNOeQjw4HhodSQU44R32aiGw at mail dot gmail dot com>
Doug Evans wrote:
> On Tue, Apr 14, 2015 at 4:41 AM, Gary Benson <gbenson@redhat.com> wrote:
> > Doug Evans wrote:
> > > While I'm all for building on "foo:bar" in path names
> > > (target:foo, remote:foo, and so on), IWBN to build a library on
> > > top of that rather than have sideband tables that recorded such
> > > extra info. [Down the road I can imagine having a class for
> > > such things such that we could augment what's recorded beyond
> > > just a "foo:bar" string, but that's later, if ever.]
> > >
> > > IOW, how about having an "is non-local" predicate that is
> > > invoked on the path whenever needed? [it could be the current
> > > "is_target_filename" or if you wanted to add a layer of
> > > abstraction that might be ok, depending on how this might
> > > evolve]
> >
> > I'm happy to remake this patch using "is_target_filename". I'll
> > do that and mail a version 2 tomorrow.
> >
> > (I've been thinking we might need something more than a prefix at
> > some point, maybe something more URL-like, but like you say, we
> > don't need that right now.)
>
> I was thinking, and this is not well thought out, maybe there's
> value in replacing OBJF_NOT_FILENAME with a flag that says the
> string is "foo:bar", and then we could have another prefix for files
> that are currently marked with OBJF_NOT_FILENAME. Just food for
> thought, or not.
Yeah, if we use some form of URL then the OBJF_NOT_FILENAME ones can
fit in there too. Local files would be file://, target ones could be
target:// maybe, or target: can be a magic prefix that gets expanded
into whatever is necessary.
The reason I was thinking we might need something more that what's
there now, by the way, is that I was trying to figure out if loaded
BFDs should be keyed to the inferior they came from. If we had two
non-local inferiors both with target:/lib64/libc.so.6, would GDB
consider those the same file somehow? I don't know.
Cheers,
Gary
--
http://gbenson.net/