Re: [PATCH v6 1/8] GDB/Guile: Don't assert that an integer value is boolean

[Simon Marchi <simon.marchi@polymtl.ca>]

On 2022-10-21 16:54, Maciej W. Rozycki wrote:
> On Fri, 21 Oct 2022, Simon Marchi wrote:
> 
>> Hmm, I see this failure now:
>>
>>     guile (set-parameter-value! test-PARAM_ZINTEGER-param #:unlimited)^M
>>     ERROR: In procedure set-parameter-value!:^M
>>     In procedure gdbscm_set_parameter_value_x: Wrong type argument in position 2 (expecting integer): #:unlimited^M
>>     Error while executing Scheme code.^M
>>     (gdb) FAIL: gdb.guile/scm-parameter.exp: kind=PARAM_ZINTEGER: test-PARAM_ZINTEGER-param: guile (set-parameter-value! test-PARAM_ZINTEGER-param #:unlimited)
>>
>> This is with Guile 3.0, if that matters.
> 
>  It's not a failure, the "ERROR: ..." message, perhaps confusingly, comes 
> from GDB under test rather than the DejaGNU test harness and therefore it 
> doesn't score as a test error or count towards test results.  The context 
> is:
> 
> (gdb) PASS: gdb.guile/scm-parameter.exp: kind=PARAM_ZINTEGER: test-PARAM_ZINTEGER-param: PARAM_ZINTEGER parameter value (2)
> guile (set-parameter-value! test-PARAM_ZINTEGER-param #:unlimited)
> ERROR: In procedure set-parameter-value!:
> ERROR: In procedure gdbscm_set_parameter_value_x: Wrong type argument in position 2 (expecting integer): #:unlimited
> Error while executing Scheme code.
> (gdb) PASS: gdb.guile/scm-parameter.exp: kind=PARAM_ZINTEGER: test-PARAM_ZINTEGER-param: guile (set-parameter-value! test-PARAM_ZINTEGER-param #:unlimited)
> 
> and the test case requires that message to be there to pass:
> 
>     set param_integer_error \
> 	"ERROR: In procedure set-parameter-value!:\r\nERROR: In procedure\
> 	 gdbscm_set_parameter_value_x: Wrong type argument in position 2\
> 	 \\(expecting integer\\): #:unlimited\r\nError while executing Scheme\
> 	 code\\."
> 
> so there's nothing to worry about.  In other words the test case verifies 
> that a particular error message is printed by GDB when supplied with a 
> bogus command.  Thanks for your meticulousness though!

But in my case, it's really a DejaGNU FAIL.

If you look carefully at the differences between our outputs, yours has
this line:

  ERROR: In procedure gdbscm_set_parameter_value_x: Wrong type argument in position 2 (expecting integer): #:unlimited

while mine has:

  In procedure gdbscm_set_parameter_value_x: Wrong type argument in position 2 (expecting integer): #:unlimited^M

In other words, mine doesn't have ERROR printed on that line, for some
reason.

Simon