Date: Sun, 27 Jun 2021 03:51:13 +0100 From: Graham Perrin <grahamperrin@gmail.com> To: freebsd-gecko@freebsd.org Subject: Firefox 89 gfx.webrender.all and gfx.webrender.enabled Message-ID: <cde1f094-7fcc-4235-f778-8c93ff582055@gmail.com> In-Reply-To: <60E16CFD-E01C-4F54-ABBC-43997C04EAD2@unrelenting.technology> References: <02790acf-513d-ed17-7543-10fa67c3ed78@gmail.com> <60E16CFD-E01C-4F54-ABBC-43997C04EAD2@unrelenting.technology>
next in thread | previous in thread | raw e-mail | index | archive | help
Thanks, On 26/06/2021 20:16, Greg V wrote: > … >> gfx.webrender.all true > You can set it to force WebRender on, even if the hardware probing and/or qualification failed. > > If it already shows WebRender, Firefox has successfully gathered hardware info and decided that the GPU is good. So you don't strictly *need* to force it. Just hope that the updates won't ever break the detection :) Where gfx.webrender.all is false by default (as it was for me): how can a person test the pros and cons of an override to true? Whilst I'm not into benchmarking/comparisons (I'm not a speed freak), I am aware of things such as MotionMark 1.2 <https://browserbench.org/MotionMark1.2/>. >> gfx.webrender.enabled true > > This one should not be touched. What's the implication of finding it false by default? I overrode gfx.webrender.enabled to true in _89_ on FreeBSD after stumbling across <https://old.reddit.com/r/kde/comments/m9y8os/-/grqa6ad/>; (not specific to FreeBSD, from when _88_ was nightly). Now reverted to its default, false, I'll no longer override.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?cde1f094-7fcc-4235-f778-8c93ff582055>