Re: [PATCH 6/6] gdb: don't share aspace/pspace on fork with "detach-on-fork on" and "follow-fork-mode child"

[Pedro Alves <pedro@palves.net>]

On 2021-09-28 8:12 p.m., Simon Marchi via Gdb-patches wrote:

> 
> So, what do you think of the patch below?  The regexp accepts (1|2) for
> both lines.  If you have a simple way to say "either 1 then 2 or 2 then
> 1" (so that it rejects 1 then 1 or 2 then 2), I would take it, but
> otherwise I think this is sufficient.

There's always plain alternatives.  Something like:

       set any "\[^\r\n\]*"

       set loc1_inf1 "$bpnum\\.1 $any inf 1\r\n$bpnum\\.2 $any inf 1"
       set loc1_inf2 "$bpnum\\.1 $any inf 1\r\n$bpnum\\.2 $any inf 2"

       set loc2_inf1 "$bpnum\\.2 $any inf 1\r\n$bpnum\\.2 $any inf 1"
       set loc2_inf2 "$bpnum\\.2 $any inf 1\r\n$bpnum\\.2 $any inf 2"

       gdb_test "info breakpoints $bpnum" \
 	    "($loc1_inf1\r\n$loc2_inf2|$loc1_inf2\r\n$loc2_inf1)" \
  	    "info breakpoints"

Alternatively, something like:

  	gdb_test_multiple "info breakpoints $bpnum" "info breakpoints" {
	  -wrap -re "$bpnum\\.1 $any inf ($::decimal)\r\n$bpnum\\.2 $any inf ($::decimal)" {
		set inf_a $expect_out(1,string)
		set inf_b $expect_out(2,string)

		gdb_assert {($inf_a == 1 && $inf_b == 2) \
                            || ($inf_a == 2 && $inf_b == 1)} \
			$gdb_test_name
          }
        }

Completely untested, written in email client...