Re: Intervention needed?

[Optikos <optikos@verizon.net>]

On Wednesday, April 3, 2019 at 10:54:04 AM UTC-4, Dmitry A. Kazakov wrote:
> On 2019-04-03 16:31, Optikos wrote:
> 
> > Which feature(s) [that other debuggers have] do you think are missing from GDB when debugging Ada?
> 
> Other debuggers work, GDB does not. If you have an Ada project of a 
> moderate size GDB stops working.

Fascinating, that might explain some of Ada's lack of increasing popularity.
1) Is this bug across all variants of GNAT?
1.1) paid-support AdaCore GNAT Pro?
1.2) AdaCore GPL-edition GNAT?
1.3) FSF GNAT?
1.4) All recent versions?
2) Is the bug observable both
2.1) in GDB-API/ABI-based Eclipse+GNATbench and
2.2) in stand-alone real FSF GDB?
3) Is the bug observable both
3.1) in normal symbols-in-the-object-code mode as well as
3.2) remote debugging with symbols only back on the mothership?
4) Has anyone reported it officially?
4.1) If so, what was the response (or lack thereof)?
5) Is it reproducible 100% of the time with Ada source code of size X?
5.1) How big is X?
5.2) Is X measured in raw size (e.g., kilobytes of source code)?
5.3) Or is X measured in density of complexity (e.g., quantity of sophisticated features intertwined within one/few localities)?

And due to the relatively narrow window of time, no one probably used the old DragonEgg with LLVM backend enough to know whether the bug is observable in a drastically different back-end (i.e., lldb).