Re: [kvm-unit-tests PATCH v3] s390x: Add strict mode to specification exception interpretation test

[Thomas Huth <thuth@redhat.com>]

On 25/08/2022 09.37, Janosch Frank wrote:
> On 7/5/22 13:17, Janis Schoetterl-Glausch wrote:
>> While specification exception interpretation is not required to occur,
>> it can be useful for automatic regression testing to fail the test if it
>> does not occur.
>> Add a `--strict` argument to enable this.
>> `--strict` takes a list of machine types (as reported by STIDP)
>> for which to enable strict mode, for example
>> `--strict 3931,8562,8561,3907,3906,2965,2964`
>> will enable it for models z16 - z13.
>> Alternatively, strict mode can be enabled for all but the listed machine
>> types by prefixing the list with a `!`, for example
>> `--strict 
>> !1090,1091,2064,2066,2084,2086,2094,2096,2097,2098,2817,2818,2827,2828`
>> will enable it for z/Architecture models except those older than z13.
>> `--strict !` will enable it always.
>>
>> Signed-off-by: Janis Schoetterl-Glausch <scgl@linux.ibm.com>
> 
> GCC 11.2.0 isn't happy
> 
> s390x/spec_ex-sie.c: In function ‘test_spec_ex_sie’:
> s390x/spec_ex-sie.c:70:17: error: format not a string literal and no format 
> arguments [-Werror=format-security]
>     70 |                 report(vm.sblk->gpsw.addr == 0xdeadbeee, msg);
>        |                 ^~~~~~
> s390x/spec_ex-sie.c:72:17: error: format not a string literal and no format 
> arguments [-Werror=format-security]
>     72 |                 report_info(msg);
>        |                 ^~~~~~~~~~~
> cc1: all warnings being treated as errors
> make: *** [<builtin>: s390x/spec_ex-sie.o] Error 1

Too bad that GCC isn't smart enough to see that it is a constant string... I 
guess we have to add an artifical "%s" format string here now?

> I have to page in the discussion again to know how this fits into the 
> picture. Either that or Thomas tells me it's exactly what he wants and I'll 
> add it to my queue once the compile problem has been fixed one way or another.

The point was that we wanted to use this k-u-t to automatically test a 
backport of the related kernel commit. Without the strict mode that is not 
possible since the test does not fail if the kernel backport is missing.

I just also noticed that I never replied to this v3 ... so FWIW, code looks 
fine to me (with the "%s" fixup on top of it), thus:

Reviewed-by: Thomas Huth <thuth@redhat.com>