Re: [PATCH] gdb/cli: fixes to newly added "list ." command

[Eli Zaretskii <eliz@gnu.org>]

> Date: Tue, 18 Jul 2023 15:40:24 +0200
> Cc: gdb-patches@sourceware.org
> From: Bruno Larsen <blarsen@redhat.com>
> 
> >    * The 'list' command now accepts '.' as an argument, which tells GDB
> >      to print source code around the default location.  The default
> >      location is where the inferior stopped;
> 
> The problem with "where the inferior is stopped", as mentioned by pedro 
> in an earlier email, that there are a few different interpretations, and 
> only one is correct.
> 
> 1. list around the PC that triggered a breakpoint. So if Thread 1 hit a 
> breakpoint, but a user switched to thread 2, they could expect that 
> "list ." would list around thread 1, but that's not what would happen
> 2. list the lowermost frame of a given thread. This was Pedro's initial 
> interpretation of the command, so if the inferior stopped and the user 
> went up a few frames, technically the place where it is stopped is still 
> the same PC, but we're not listing around those lines
> 3. List around the frame that is being examined. This is what I meant 
> with the command, so if you go up some frames in a different thread, 
> "list ." will list the same thing as the first call to "list". Its 
> effectively an alias for "(gdb) frame; (gdb) list", as pedro put it.

We are splitting hair, I think.  To make this more accurate, I can
suggest

  The default location is where the inferior stopped in the current
  stack frame.