Re: [PP?] Re: [PATCH] gdb/testsuite: remove perror calls when failing to run

[Simon Marchi <simark@simark.ca>]

On 12/5/22 15:59, Tom Tromey wrote:
>>>>>> "Simon" == Simon Marchi via Gdb-patches <gdb-patches@sourceware.org> writes:
> 
> Simon> The problem is that arrayptr.exp calls perror when failing to run to
> Simon> main, then returns.  perror makes it so that the next test (as in
> Simon> pass/fail) will be recorded as UNRESOLVED.  However, here, the next test
> Simon> (as in pass/fail) is in the next test (as in .exp).  Hence the spurious
> Simon> UNRESOLVED in assign_1.exp.
> 
> Simon> These perror when failing to run to X are not really useful, especially
> Simon> since runto records a FAIL on error, by default.  Remove all the
> Simon> perrors on runto failure I could find.
> 
> Thanks for doing this.  I didn't read the patch, but the idea seems
> sensible to me.

Thanks, I pushed it.

> From the dejagnu description of perror, I wonder if the gdb test suite
> should ever call it:
> 
>     Declares a severe error in the testing framework itself.

I don't reallt understand the point of perror.  If we find a severe
error, why keep running the tests (marking them as UNRESOLVED).  Just
throw an error / abort.

> Simon> I thought of adding a check ran between test (in gdb_finish
> Simon> probably) where we would emit a warning if errcnt > 0, meaning a test
> Simon> quit and left a perror "active".  However, reading that variable would
> Simon> poke into the DejaGNU internals, not sure it's a good idea.
> 
> Perhaps dejagnu itself should be doing this.

Indeed, that would make sense.  I have enough things on my plate though,
I won't pursue this :).

Simon