From owner-freebsd-x11@freebsd.org Tue Aug 14 18:16:40 2018 Return-Path: Delivered-To: freebsd-x11@mailman.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2610:1c1:1:606c::19:1]) by mailman.ysv.freebsd.org (Postfix) with ESMTP id F18A81080993 for ; Tue, 14 Aug 2018 18:16:39 +0000 (UTC) (envelope-from clmoonriver@equinedreams.art) Received: from MTA-06-3.privateemail.com (mta-06-3.privateemail.com [198.54.127.59]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (Client CN "privateemail.com", Issuer "COMODO RSA Domain Validation Secure Server CA" (verified OK)) by mx1.freebsd.org (Postfix) with ESMTPS id 863F280679; Tue, 14 Aug 2018 18:16:39 +0000 (UTC) (envelope-from clmoonriver@equinedreams.art) Received: from MTA-06.privateemail.com (localhost [127.0.0.1]) by MTA-06.privateemail.com (Postfix) with ESMTP id 64E0F60059; Tue, 14 Aug 2018 14:16:31 -0400 (EDT) Received: from [192.168.254.71] (unknown [10.20.151.230]) by MTA-06.privateemail.com (Postfix) with ESMTPA id 875B960055; Tue, 14 Aug 2018 18:16:30 +0000 (UTC) Subject: Re: X11 not working on 11-STABLE with AMDGPU To: Jung-uk Kim , Greg V Cc: pkubaj@anongoth.pl, freebsd-x11@freebsd.org References: <20180801092849.GA75303@KGPE-D16> <95b8cfa0-0908-3321-5155-ef49b1bb0a64@equinedreams.art> <20180801141207.GA5202@smtp.iq.pl> <4266f156-3a83-a45d-fab0-cde090c139ba@equinedreams.art> <67630578-f418-b435-9bb0-b40f12e9d881@nomadlogic.org> <1533215599.23824.0@hraggstad.unrelenting.technology> <73d85e29-be1d-a8a8-0fe9-81889a4a429f@equinedreams.art> <1533331374.1685.0@hraggstad.unrelenting.technology> <521c2cec-3c39-ecf0-8b21-f67e575f39c5@FreeBSD.org> From: CL Moonriver Message-ID: Date: Tue, 14 Aug 2018 13:16:28 -0500 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101 Thunderbird/52.9.1 MIME-Version: 1.0 In-Reply-To: <521c2cec-3c39-ecf0-8b21-f67e575f39c5@FreeBSD.org> Content-Type: text/plain; charset=windows-1251; format=flowed Content-Transfer-Encoding: 8bit Content-Language: en-US X-Virus-Scanned: ClamAV using ClamSMTP X-BeenThere: freebsd-x11@freebsd.org X-Mailman-Version: 2.1.27 Precedence: list List-Id: X11 on FreeBSD -- maintaining and support List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 14 Aug 2018 18:16:40 -0000 Resending this because I want to make sure jung-uk Kim gets included in the future discussion. Sorry for the really delayed response. Yes, I am 100% sure I am booting with UEFI. And I don't seem to have any issues. Again, the only bug I still encounter is that occasionally X hangs and has to restart the driver. And if I look at the console, I get errors about "GPU not responding. Attempting to reset GPU", and so on. I don't know if it helps, but again, my CPU is an AMD A10-7860K (2016) model with onboard Radeon R7 graphics. If there's any other info I can provide, please let me know. On 08/06/2018 11:20 AM, Jung-uk Kim wrote: > On 18. 8. 3., Greg V wrote: >> On Fri, Aug 3, 2018 at 2:16 AM, CL Moonriver >> wrote: >>> Thanks for the heads up. >>> >>> Are you sure this is still relevant in drm-next? I don't see any dmesg >>> error about disabling GPU acceleration. I did what you suggested, but >>> nothing seems to have changed except I get no text until radeonkms has >>> loaded (I don't get garbage I get nothing at all). I still get dmesg >>> errors regarding ring test failures and such, so turning the EFI >>> framebuffer off doesn't appear to have actually changed anything. >> Interesting. Are you 100% sure you're booting with UEFI, not CSM/legacy? >> >> Maybe radeonkms doesn't conflict anymore. Maybe this depends on the GPU >> model or something? >> >> For amdgpu (RX 480), it's not an error about disabling acceleration, the >> amdgpu driver *completely* fails to load when EFI framebuffer is active. >> And I retested a few weeks ago. No one fixed this specifically yet, so >> I'm not expecting it to magically stop conflicting… > It wasn't fixed and some MMIO test still fails. Actually, both > radeonkms and amdgpu still do not work properly when UEFI is active. I > don't think it is too hard to fix but I didn't dig deeper at the time. > > Jung-uk Kim >