<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<style type="text/css" style="display:none;"> P {margin-top:0;margin-bottom:0;} </style>
</head>
<body dir="ltr">
<div style="font-family: Calibri, Arial, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);" class="elementToProof ContentPasted0">
> On 26/04/2023 07:24, fei.yang@intel.com wrote:
<div class="ContentPasted0">>> From: Fei Yang <fei.yang@intel.com></div>
<div class="ContentPasted0">>></div>
<div class="ContentPasted0">>> The first three patches in this series are taken from</div>
<div class="ContentPasted0">>> https://patchwork.freedesktop.org/series/116868/</div>
<div class="ContentPasted0">>> These patches are included here because the last patch</div>
<div class="ContentPasted0">>> has dependency on the pat_index refactor.</div>
<div class="ContentPasted0">>></div>
<div class="ContentPasted0">>> This series is focusing on uAPI changes,</div>
<div class="ContentPasted0">>> 1. end support for set caching ioctl [PATCH 4/5]</div>
<div class="ContentPasted0">>> 2. add set_pat extension for gem_create [PATCH 5/5]</div>
<div class="ContentPasted0">>></div>
<div class="ContentPasted0">>> v2: drop one patch that was merged separately</div>
<div class="ContentPasted0">>> 341ad0e8e254 drm/i915/mtl: Add PTE encode function</div>
<div class="ContentPasted0">></div>
<div class="ContentPasted0">> Are the re-sends for stabilizing the series, or focusing on merge?</div>
<div><br class="ContentPasted0">
</div>
<div class="ContentPasted0">v2 was sent just to drop one of patches that has already been merged.</div>
<div><br class="ContentPasted0">
</div>
<div class="ContentPasted0">> If the latter then opens are:</div>
<div class="ContentPasted0">></div>
<div class="ContentPasted0">> 1) Link to Mesa MR reviewed and ready to merge.</div>
<div class="ContentPasted0">></div>
<div class="ContentPasted0">> 2) IGT reviewed.</div>
<div class="ContentPasted0">></div>
<div class="ContentPasted0">> 3) I raised an open that get/set_caching should not "lie" but return an</div>
<div class="ContentPasted0">> error if set pat extension has been used. I don't see a good reason not</div>
<div class="ContentPasted0">> to do that.</div>
<div><br class="ContentPasted0">
</div>
<div class="ContentPasted0">I don't think it's "lying" to the userspace. the comparison is only valid</div>
<div class="ContentPasted0">for objects created by KMD because only such object uses the pat_index</div>
<div class="ContentPasted0">from the cachelevel_to_pat table.</div>
<div><br class="ContentPasted0">
</div>
<div class="ContentPasted0">> + Joonas on this one.</div>
<div class="ContentPasted0">></div>
<div class="ContentPasted0">> 4) Refactoring as done is not very pretty and I proposed an idea for a</div>
<div class="ContentPasted0">> nicer approach. Feasible or not, open for discussion.</div>
<div><br class="ContentPasted0">
</div>
<div class="ContentPasted0">Still digesting your proposal. but not sure how would you define things</div>
<div class="ContentPasted0">like PAT_UC as that is platform dependent, not a constant.</div>
<div class="ContentPasted1">- Fei</div>
<div><br class="ContentPasted1">
</div>
<div class="ContentPasted1">> At a push I can look past that and someone can attempt to tidy the</div>
<div class="ContentPasted1">> driver later.</div>
<div class="ContentPasted1">></div>
<div class="ContentPasted1">> But without 1-3 we cannot merge this.</div>
<div class="ContentPasted1">></div>
<div class="ContentPasted1">> Regards,</div>
<div class="ContentPasted1">></div>
<div class="ContentPasted1">> Tvrtko</div>
<br>
</div>
</body>
</html>