Intel Gen7 Graphics Mitigation Will Try To Avoid Performance Loss In Final Version
Interestingly with this new patch series for the Gen7 mitigation, which is still under a "request for comments" flag, is now a big disclaimer:
**** NOTE: ****
This series is in active development and is not intended to be merged to mainline in its current form. The intent of the RFC is simply to outline the strategy for the mitigation, as a focus for active discussion, and to openly share progress. There has been only minimal attention paid to performance thus far, as the focus is on robustness. It is not anticipated that there will be any measurable performance impact in the final version.
**** END NOTE ****
It's good if they are indeed hopeful of eliminating the huge performance hit caused by this mitigation, though wondering how long until these patches will be in their final version and ready for mainline.
One of the resources for CVE-2019-14615 pointed towards Intel being aware of this issue since at least August, so would seem a bit odd they aren't focusing on performance at all until the end given the time they've already had, and still didn't have a proper solution in time for the planned public disclosure date that happened on Tuesday.
This mitigation work is primarily concerning Ivybridge / Haswell hardware with the "Gen9" graphics mitigation already in good shape and mainlined.
With the v3 patches are more code improvements on this nearly 700 line patch to their i915 kernel graphics driver. Once the patches are in their final form, of course, I'll be running more benchmarks to see what exactly the performance penalty looks like or if they've managed to recover completely.
At least though the Gen9 graphics mitigation for current-gen hardware is spot on and I haven't found any performance or power hit from that mainlined mitigation.
9 Comments
