Thanks for the reviews. I’ll push patch 4, and repost the others
with the suggested changes.


> On 25 Apr 2019, at 15:31, Tom Tromey <tom@tromey.com> wrote:
> 
>>>>>> "Alan" == Alan Hayward <Alan.Hayward@arm.com> writes:
> 
> Alan> @@ -1716,6 +1716,8 @@ proc default_gdb_start { } {
> Alan>  	    warning "Couldn't set the width to 0."
> Alan>  	}
> Alan>      }
> Alan> +
> Alan> +    gdb_debug_init
> 
> One question I have is how this interacts with tests that run gdb
> multiple times.  Is there just one log?

By default the log files are opened in append mode. So everything should
happily go to a single file. I’ll double check.

> 
> Alan> +    if ![info exists gdbdebug] {
> 
> I think it's preferable to brace expressions, like "if {![info exists...]}".
> I guess it's just stylistic though.

Ok.

> 
> Alan> +  # First ensure logging is off.
> Alan> +  gdb_test_no_output "set logging off"
> 
> Related to the multiple gdb invocations question -- I think this will
> result in duplicate test names, so something extra should probably be
> done here.
> 

I can switch this to use “send_gdb” in the same way "set height 0” is used.
It’ll also stop the number of passes increasing when debug is enabled.


> thanks,
> Tom

Oh, and earlier I got a mailbox full bounce back from your email when I sent
the reply to the PIE email:
"550 5.0.350 Remote server returned an error -> 550 Mailbox is full / Blocks limit exceeded / Inode limit exceeded”


Alan.�&���j�!z�޶��מz�Yb�֫rn��r�