Date: Fri, 5 Apr 2019 07:15:14 -0700 From: Marcel Flores <marcel@brickporch.com> To: freebsd-arm@freebsd.org Subject: Re: Restore broken ThunderX support in 12 by MFCing r343764? Message-ID: <AEF036AB-D81E-4D46-BE04-E470A2B350D1@brickporch.com> In-Reply-To: <3996B352-ABFB-4CA5-A897-6C6CD1D2A132@brickporch.com> References: <20190218022911.72261m7v5l6g1i7b@webmail.omc.net> <CAPyFy2CB0T6GpR5=Am_CYnrbD5KehX3YjCtnT6gd0tHP0VPcow@mail.gmail.com> <3996B352-ABFB-4CA5-A897-6C6CD1D2A132@brickporch.com>
next in thread | previous in thread | raw e-mail | index | archive | help
> On Apr 3, 2019, at 10:16 PM, Marcel Flores <marcel@brickporch.com> = wrote: >=20 >=20 >=20 >> On Apr 3, 2019, at 6:55 AM, Ed Maste <emaste@freebsd.org> wrote: >>=20 >> On Sun, 17 Feb 2019 at 20:29, <kraileth@elderlinux.org> wrote: >>>=20 >>> Not being a developer though, I cannot judge if [r343764] cannot be = MFC'd >>> into 12-STABLE due to making invasive changes or if it simply never >>> was because it was thought to be an improvement for 13 only and not = an >>> actually pretty vital fix for 12. >>=20 >> I suspect jchandra@ just didn't realize it's needed also on ThunderX, >> and I did not encounter any trouble with the ThunderX systems I have. >> It could just be that our (older) ThunderX reference firmware has >> fewer regions in its ACPI info and so works fine without r343764. >> Anyhow I've now merged the change to stable/12. >>=20 >> With respect to vt_efifb the tunable is a simple workaround, but we >> really need this to work out-of-the-box. Do you have any further >> details on the failure when vt_efifb is enabled? >>=20 >> Also, if you're aware of any other ThunderX issues please let me = know.] >=20 > stable/12 now works. In fact, the hw.syscons.disable tuneable is no = longer needed either =E2=80=94 a fresh build stable/12 appears to be = working with GENERIC out of the box. Based on this thread, that = doesn=E2=80=99t totally make sense to me, but I=E2=80=99m not sure what = all else has made it into stable/12. >=20 > The same can=E2=80=99t be said for 13, which, as of about 2 weeks ago = (I can confirm later this week if that would be worthwhile), behaves as = described here without it: > = https://lists.freebsd.org/pipermail/freebsd-arm/2018-December/019130.html = <https://lists.freebsd.org/pipermail/freebsd-arm/2018-December/019130.html= ><https://lists.freebsd.org/pipermail/freebsd-arm/2018-December/019130.htm= l = <https://lists.freebsd.org/pipermail/freebsd-arm/2018-December/019130.html= >> >=20 > Since the serial console essentially hangs, I=E2=80=99m not sure how = to get any additional information, but happy to try any suggestions. >=20 >=20 > The only other issues I=E2=80=99ve had with the ThunderX Gigabyte = board I have is getting the onboard NIC to work (Seems to be about like = this, looking at dmesg: = https://lists.freebsd.org/pipermail/freebsd-arm/2018-April/017865.html = <https://lists.freebsd.org/pipermail/freebsd-arm/2018-April/017865.html>),= but I imagine this is a driver question for the Gigabyte portions, = rather than an issue with the ThunderX itself. >=20 Correction to the above =E2=80=94 just tried to boot current/13 r345863 = via USB installer, it actually gets further, sans tuneable, but gets stuck = somewhere.=20 Boot output attached. https://hastebin.com/xeteyuhome -m
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?AEF036AB-D81E-4D46-BE04-E470A2B350D1>