[PATCH] [gdb/testsuite] Fix gdb.ada/mi_task_arg.exp with newer gcc

[PATCH] [gdb/testsuite] Fix gdb.ada/mi_task_arg.exp with newer gcc

[Tom de Vries]

When running test-case gdb.ada/mi_task_arg.exp on openSUSE Tumbleweed using
gcc 13.2.1, I run into (layout adapted for readability):
...
-stack-list-arguments 1^M
^done,stack-args=[
  frame={level="0",args=[]},
  frame={level="1",args=[{name="<_task>",value="0x464820"},
                         {name="<_taskL>",value="129"}]},
  frame={level="2",args=[{name="self_id",value="0x464840"}]},
  frame={level="3",args=[]},
  frame={level="4",args=[]}
]^M
(gdb) ^M
FAIL: gdb.ada/mi_task_arg.exp: -stack-list-arguments 1 (unexpected output)
...

On openSUSE Leap 15.4 with gcc 7.5.0 I get instead:
...
-stack-list-arguments 1^M
^done,stack-args=[
  frame={level="0",args=[]},
  frame={level="1",args=[{name="<_task>",value="0x444830"}]},
  frame={level="2",args=[{name="self_id",value="0x444850"}]},
  frame={level="3",args=[]},
  frame={level="4",args=[]}]^M
(gdb) ^M
PASS: gdb.ada/mi_task_arg.exp: -stack-list-arguments 1
...

The difference in gdb output is due to difference in the dwarf generated by
the compiler, so I don't see a problem with gdb here.

Fix this by updating the test-case to accept this output.

Tested on x86_64-linux.
---
 gdb/testsuite/gdb.ada/mi_task_arg.exp | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/gdb/testsuite/gdb.ada/mi_task_arg.exp b/gdb/testsuite/gdb.ada/mi_task_arg.exp
index 1d84e432ad8..747f3d3a268 100644
--- a/gdb/testsuite/gdb.ada/mi_task_arg.exp
+++ b/gdb/testsuite/gdb.ada/mi_task_arg.exp
@@ -44,7 +44,8 @@ if {![mi_runto "task_switch.break_me"]} {
 set frame0 "frame=\{level=\"0\",args=\\\[\\\]\}"
 # Frame for task_switch.caller
 set frame1_args "\{name=\"<_task>\",value=\"$hex\"\}"
-set frame1 "frame=\{level=\"1\",args=\\\[$frame1_args\\\]\}"
+set frame1_opt_args "\{name=\"<_taskL>\",value=\"$decimal\"\}"
+set frame1 "frame=\{level=\"1\",args=\\\[${frame1_args}(,$frame1_opt_args)?\\\]\}"
 # Frame for system.tasking.stages.task_wrapper
 set frame2_args "(\{name=\"self_id\",value=\"($hex|<optimized out>)\"\})?"
 set frame2 "frame=\{level=\"2\",args=\\\[$frame2_args\\\]\}"

base-commit: 940bb336cbce23f1780c449a7c72ca41e8d91fa5
-- 
2.35.3

Re: [PATCH] [gdb/testsuite] Fix gdb.ada/mi_task_arg.exp with newer gcc

[Tom Tromey]

>>>>> "Tom" == Tom de Vries via Gdb-patches <gdb-patches@sourceware.org> writes:

Tom> The difference in gdb output is due to difference in the dwarf generated by
Tom> the compiler, so I don't see a problem with gdb here.

Tom> Fix this by updating the test-case to accept this output.

Thank you.

Approved-By: Tom Tromey <tom@tromey.com>

Tom