Re: Propose we release GDB 9.1 next weekend (Feb 01-02)

[Tom Tromey <tom@tromey.com>]

>>>>> "Joel" == Joel Brobecker <brobecker@adacore.com> writes:

Joel> FWIW, in terms of reverting, I was only thinking of reverting
Joel> in the gdb-9-branch.

I sent a reversion patch.  It's pretty straightforward, but take a quick
look anyway.

Joel> It's possible we might decide to revert
Joel> in master as well, but, at the moment, it's looking like we have
Joel> a decent chance of fixing the issue in master withing a reasonable
Joel> time frame. So I'm not advocating for a revert in master just yet.
Joel> Open to people's thoughts, though, as always!

I wasn't sure what to do here, either.

For trunk I think we should at least have a regression test for the
bug exposed in PR breakpoints/24915.

I suppose right now the name to search by and the name to display are
entangled in the code.  So maybe another approach to fixing the original
problem would be to untangle them somehow.  I recall wanting to move
this information into the source cache, and out of the symtabs; but I
looked at this and it was a little complicated because, IIRC, the search
can depend on the path to the objfile.  Maybe we could have a unified
table per objfile though, the idea being to consolidate and cache the
lookups, so that (1) it's shared across symtabs and psymtabs, and (2)
things like the name to print can be computed lazily if possible.

Hopefully this makes sense,
Tom