Re: [PATCH] [gdb] Prune inferior after switching inferior

[Tom de Vries <tdevries@suse.de>]

On 3/4/24 13:21, Tom de Vries wrote:
> Usually with test-case gdb.python/py-progspace-events.exp I get:
> ...
> (gdb) inferior 1^M
> [Switching to inferior 1 [process 4116] (py-progspace-events)]^M
> [Switching to thread 1.1 (Thread 0xf77d0ce0 (LWP 4116))]^M
> 28      { /* Nothing.  */ }^M
> (gdb) PASS: gdb.python/py-progspace-events.exp: inferior 1
> step^M
> FreeProgspaceEvent: <gdb.Progspace object at 0xabf4f850>^M
> do_parent_stuff () at py-progspace-events.c:41^M
> 41        ++global_var;^M
> (gdb) PASS: gdb.python/py-progspace-events.exp: step
> ...
> 
> But occasionally I run into the following FAIL:
> ...
> (gdb) inferior 1^M
> [Switching to inferior 1 [process 5199] (py-progspace-events)]^M
> [Switching to thread 1.1 (Thread 0xf77d0ce0 (LWP 5199))]^M
> 28      { /* Nothing.  */ }^M
> (gdb) FreeProgspaceEvent: <gdb.Progspace object at 0xabaf03a0>^M
> FAIL: gdb.python/py-progspace-events.exp: inferior 1 (timeout)
> ...
> 
> This is caused by a race between the handling of an event, and the
> "inferior 1" command.
> 
> In the passing case, the event is handled first.  During which prune_inferiors
> is called, but it can't remove inferior 1, because it's still the current one.
> 

I just spotted two mistakes in the commit log.

-can't remove inferior 1
+can't remove inferior 2

> In the failing case, the "inferior 1" command is handled first.  Then during
> handling of the event, prune_inferiors is called, and it can remove inferior 1
> because it's no longer the current one.
> 

-can remove inferior 1
+can remove inferior 2

I'll update this locally.

Thanks,
- Tom

> This looks like a test-case issue to me, but ISTM that we can do better: by
> calling prune_inferiors asap, at the end of the "inferior 1" command, we
> stabilize the moment when the inferior is removed:
> ...
> (gdb) inferior 1^M
> [Switching to inferior 1 [process 5199] (py-progspace-events)]^M
> [Switching to thread 1.1 (Thread 0xf77d0ce0 (LWP 5199))]^M
> 28      { /* Nothing.  */ }^M
> FreeProgspaceEvent: <gdb.Progspace object at 0xabaf03a0>^M
> (gdb) PASS: gdb.python/py-progspace-events.exp: inferior 1
> ...
> 
> Tested on x86_64-linux.
> 
> PR gdb/31440
> Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=31440
> ---
>   gdb/inferior.c                                |  4 +++
>   .../gdb.python/py-progspace-events.exp        | 31 +++----------------
>   2 files changed, 9 insertions(+), 26 deletions(-)
> 
> diff --git a/gdb/inferior.c b/gdb/inferior.c
> index 5ff5eb98955..4e179f7fb80 100644
> --- a/gdb/inferior.c
> +++ b/gdb/inferior.c
> @@ -791,6 +791,10 @@ inferior_command (const char *args, int from_tty)
>   	  notify_user_selected_context_changed
>   	    (USER_SELECTED_INFERIOR);
>   	}
> +
> +      /* Switching current inferior may have made one of the inferiors
> +	 prunable, so prune it.  */
> +      prune_inferiors ();
>       }
>   }
>   
> diff --git a/gdb/testsuite/gdb.python/py-progspace-events.exp b/gdb/testsuite/gdb.python/py-progspace-events.exp
> index 95e4ca8da0b..9dfc7573d40 100644
> --- a/gdb/testsuite/gdb.python/py-progspace-events.exp
> +++ b/gdb/testsuite/gdb.python/py-progspace-events.exp
> @@ -79,37 +79,16 @@ gdb_test "continue" \
>   	 "\\\[Inferior $decimal \[^\r\n\]+ exited normally\\\]"] \
>       "continue until inferior 2 exits"
>   
> -gdb_test "inferior 1" "\\\[Switching to inferior 1 .*"
> -
> -# Step the inferior.  During this process GDB will prune the now
> +# Switch to inferior 1.  During this process GDB will prune the now
>   # defunct inferior, which deletes its program space, which should
>   # trigger the FreeProgspaceEvent.
>   #
> -# However, there is a slight problem.  When the target is remote, and
> -# GDB is accessing files using remote fileio, then GDB will attempt to
> -# prune the inferior at a point in time when the remote target is
> -# waiting for a stop reply.  Pruning an inferior causes GDB to close
> -# files associated with that inferior.
> -#
> -# In non-async mode we can't send fileio packets while waiting for a
> -# stop reply, so the attempts to close files fails, and this shows up
> -# as an error.
> -#
> -# As this error has nothing to do with the feature being tested here,
> -# we just accept the error message, the important part is the
> -# 'FreeProgspaceEvent' string, so long as that appears (just once)
> -# then the test is a success.
> -set warning_msg \
> -    [multi_line \
> -	 "warning: cannot close \"\[^\r\n\]+\": Cannot execute this command while the target is running\\." \
> -	 "Use the \"interrupt\" command to stop the target" \
> -	 "and then try again\\."]
>   
> -gdb_test "step" \
> +gdb_test "inferior 1" \
>       [multi_line \
> -	 "^FreeProgspaceEvent.*: <gdb.Progspace object at $hex>(?:\r\n$warning_msg)*" \
> -	 "do_parent_stuff \\(\\) at \[^\r\n\]+" \
> -	 "$decimal\\s+\[^\r\n\]+"]
> +	 "\\\[Switching to inferior 1 .*" \
> +	 ".*" \
> +	 "FreeProgspaceEvent.*: <gdb.Progspace object at $hex>"]
>   
>   # Let this inferior run to completion.
>   gdb_continue_to_end
> 
> base-commit: 1485a3fb63619cced99dd7a4a043cf01a0f423d9