On Thu, Mar 12, 2020 at 03:31:49AM +0200, Liran Alon wrote:
>
> On 11/03/2020 22:24, Michael S. Tsirkin wrote:
> > Notice the process as documented in ./tests/qtest/bios-tables-test.c
> >
> Thanks for explicitly pointing me to that process.
>
> I have followed the process described there (Both steps 1-3 and steps 4-7).
> On step (6), I have noted that many existing ACPI tables don't have expected
> binaries for all the execution-matrix.
> E.g. tests/data/acpi/pc/APIC.{bridge, ipmikcs, memhp, numamem} are all
> missing.
> Similar missing files exists for FACP, FACS, HPET and MCFG.
>
> I should add for WAET the expected binaries for all the execution-matrix
> right?
> Is it just an existing issue that for the existing tables some of the
> expected binaries are missing? But the tests seems to pass.
> Can you clarify this for me?
>
> Thanks,
> -Liran
It's because of this (which we should probably rewrite as a loop):
try_again:
aml_file = g_strdup_printf("%s/%s/%.4s%s", data_dir, data->machine,
sdt->aml, ext);
if (getenv("V")) {
fprintf(stderr, "Looking for expected file '%s'\n", aml_file);
}
if (g_file_test(aml_file, G_FILE_TEST_EXISTS)) {
exp_sdt.aml_file = aml_file;
} else if (*ext != '\0') {
/* try fallback to generic (extension less) expected file */
ext = "";
g_free(aml_file);
goto try_again;
}
if WAET is always added, then a single WAET will be enough for you.
--
MST
