Date: Sun, 29 Mar 2020 16:53:25 +0200 From: =?utf-8?Q?Klaus_K=C3=BCchemann?= <maciphone2@googlemail.com> To: Greg V <greg@unrelenting.technology>, freebsd-arm@freebsd.org, Mike Karels <mike@karels.net> Subject: Re: Switch to UEFI Re: CFT: alpha test of Ethernet driver for RPi4 Message-ID: <C07E205F-410E-49C3-B98A-08769D295A39@googlemail.com> In-Reply-To: <3d84dbd6acea80b04bee712b59661a86@unrelenting.technology> References: <964BBCA2-5EE1-4673-966E-63D37FEDB4EA@googlemail.com> <AA1CAD4B-2275-4988-A533-203D91808DE2@googlemail.com> <202003281622.02SGMcmi027728@mail.karels.net> <57d4ba4ef95eeaf382d2c0b2407e9dab@unrelenting.technology> <3d84dbd6acea80b04bee712b59661a86@unrelenting.technology>
next in thread | previous in thread | raw e-mail | index | archive | help
> Am 28.03.2020 um 23:39 schrieb greg@unrelenting.technology: >=20 >=20 > What allwinner? > How did you find anything allwinner-related in ACPI code, when AFAIK = no Allwinner SoC ever had ACPI tables written for it? >=20 > I'm talking about generic XHCI. Sorry, I always mix up Allwinner with Marvell, where you changed some things in generic_xhci_fdt. =E2=80=A6 Mo matter, what I wanted to say is that there perhaps could be some need = to add things for the bcm-soc what is called ACPi-glue by jmcneill in his sources=E2=80=A6= `hope you are right that it `ll be only a simple bugfix for you to get = the right IRQ setup in generic_xhci.c =E2=80=A6 > Am 28.03.2020 um 23:27 schrieb Mike Karels <mike@karels.net>: >=20 >=20 > I=E2=80=99m not ready to put the source out in public, as the = copyright notice is not yet approved. I=E2=80=99ll put it in = phabricator when it=E2=80=99s ready. >=20 > About UEFI: I don=E2=80=99t know what=E2=80=99s involved, but half of = the driver interfaces with the rest of the network stack, and I don=E2=80=99= t see how using UEFI simplifies any of that. >=20 > Mike O.K., seems I misunderstood your 'alpha test=E2=80=98- offer=E2=80=A6 I thought you wanted us to test/review your code before you put it to = phabricator. Well, jmcneill, where you derived the genet-driver-sources from, has = already=20 integrated the driver in UEFI, while it is not yet integrated in = rpi4-uefi upstream, afaik. You can dual-boot in UEFI from dt(hangs at gpioregulator at the moment) = or ACPI(hangs unrelenting at greg`s irq-code at the the moment[.. just = kidding] . They are working on netboot at the moment ..=20 I really suggest to track the UEFI - devs to understand what=E2=80=99s = going on there..=20 it could be the right way to finally get this damned board under = control.. Regards Klaus=
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?C07E205F-410E-49C3-B98A-08769D295A39>