[igt-dev] [PATCH i-g-t] i915/query: Enable QUERY_TOPOLOGY_INFO for all gens

Summers, Stuart stuart.summers at intel.com
Sat Feb 9 00:32:06 UTC 2019


On Sat, 2019-02-09 at 00:21 +0000, Chris Wilson wrote:
> Quoting Stuart Summers (2019-02-09 00:11:46)
> > Currently the QUERY_TOPOLOGY_INFO IOCTL is only being tested on
> > haswell,
> > broadwell, skylake, kabylake, and coffeelake.
> 
> No, it's tested everywhere it's supported across the other tests.
> Only
> this particular test is skipped if there is no hardcoded truth value
> as 
> only a small number of known pci-ids are supplied.
> 
> Skip is the more accurate result (since the test is not doing
> anything it
> sets out to do, and tests should be reasonably succinct), but you
> could
> move the skip down into the switch() so that we don't repeat
> ourselves.
> -Chris

My reasoning for removing this was to exercise the asserts we are
calling within i915_query_items() for those platforms which do not
implement dev_info->gt. So we are checking something here, even if it
doesn't specifically analyze the slices/subslices for each platform.

I don't think the way it is currently implemented is wrong, but feel
the test could be used to potentially catch a bug if the driver were to
miss something in its dev_info structure for some platform, for
instance.

If you feel strongly we should keep this the way it is, I can drop
this.

Thanks,
Stuart
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/x-pkcs7-signature
Size: 3270 bytes
Desc: not available
URL: <https://lists.freedesktop.org/archives/igt-dev/attachments/20190209/e6aa5bc1/attachment.bin>


More information about the igt-dev mailing list