<html> <head> <meta http-equiv="Content-Type" content="text/html; charset=us-ascii"> <meta name="Generator" content="Microsoft Exchange Server"> <style></style> </head> <body> <div> <div dir="auto"> <div>Hi Ray,<br> <div class="x_gmail_extra"><br> <div class="x_gmail_quote">Am 23.04.2018 14:08 schrieb Huang Rui <ray.huang@amd</div> </div> </div> </div> </div> <font size="2"><span style="font-size:11pt;"> <div class="PlainText">On Mon, Apr 23, 2018 at 05:57:06PM +0800, Koenig, Christian wrote:<br> > Hi Ray,<br> > <br> > Am 23.04.2018 11:47 schrieb Huang Rui <ray.huang@amd.com>:<br> > <br> > On Fri, Apr 20, 2018 at 05:59:16PM +0800, Koenig, Christian wrote:<br> > > Am 20.04.2018 um 11:40 schrieb Huang Rui:<br> > > > "aaabaf4 drm/amdgpu: defer test IBs on the rings at boot (V3)"<br> > > > Above patch defers the execution of gfx/compute ib tests. However, at<br> > that time,<br> > > > the gfx may already go into idle state. If "idle" gfx receives command<br> > > > submission, it will get hang in the system. So we must add is_gfx_on<br> > checking at<br> > > > start of ib tests.<br> > ><br> > > Do I see that right that you just skip the IB test when the GFX block is<br> > > already turned of? In this case that would be a clear NAK.<br> > ><br> > > BTW: How do you detect that we need to turn GFX on again?<br> > <br> > Christian, I know point. But there is a hang issue if we would like try to<br> > disable/enable gfxoff with SMC message at runtime. Actually, I am trying to<br> > find a good sequence to fix it. After that, I can even expose an debugfs<br> > interface to configure that. So I have to skip the test for the moment when<br> > gfx is in "idle".<br> > <br> > <br> > Working around that issue for the moment is ok, but please note that explicitly<br> > in both the commit message and a code comment.<br> <br> OK. Will add it at V2.<br> <br> > <br> > But don't you run into the same problem when the UMD starts to submit commands?<br> <br> When UMD starts, RLC firmware will detect the "draw" command, then it will<br> power up gfx. So it won't have problem at that time. The mainly state<br> machine doesn't expose to driver side yet.<br> <br> > <br> > I mean the idea of the IB test is that you "simulate" an userspace command<br> > submission and see if it works.<br> > <br> <br> Yes, agree. Any idea to "simulate" the "draw" command? Or please wait for<br> my fix for enabling/disabling gfxoff at runtime.<br> <br> Thanks,<br> Ray<br> </div> </span></font> </body> </html>