Date: Thu, 26 Jul 2018 17:01:57 +0300 From: Greg V <greg@unrelenting.technology> To: Emmanuel Vadot <manu@bidouilliste.com> Cc: Kyle Evans <kevans@freebsd.org>, "freebsd-arm@freebsd.org" <freebsd-arm@freebsd.org> Subject: Re: Allwinner dtb overlays on CURRENT. Also, flashrom SPI! Message-ID: <1532613717.49616.0@hraggstad.unrelenting.technology> In-Reply-To: <20180726120926.9648d6b691ab7be7cbb37653@bidouilliste.com> References: <1532548163.59286.0@hraggstad.unrelenting.technology> <CACNAnaGLPgEs0MVMBbit78WGDCOMxxAcTXBYFZgamPZBmf5p3g@mail.gmail.com> <20180725234000.e917566724f997319f8fe2fd@bidouilliste.com> <1532598504.2425.0@hraggstad.unrelenting.technology> <20180726120926.9648d6b691ab7be7cbb37653@bidouilliste.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, Jul 26, 2018 at 1:09 PM, Emmanuel Vadot <manu@bidouilliste.com> wrote: > On Thu, 26 Jul 2018 12:48:24 +0300 > Greg V <greg@unrelenting.technology> wrote: >> On Thu, Jul 26, 2018 at 12:40 AM, Emmanuel Vadot >> <manu@bidouilliste.com> wrote: >> > On Wed, 25 Jul 2018 15:14:29 -0500 >> > Kyle Evans <kevans@freebsd.org> wrote: >> > This is wrong bindings. Yes I know that it's working with our >> driver >> > for ths but this is wrong. >> > 1) ths module isn't in ahb but apb. >> > 2) We need nvmem framework to get the sensor calibration data, >> please >> > see https://reviews.freebsd.org/D16419 >> >> I guess it's good to have a framework, but the existing driver >> communicates with aw_sid directly (aw_sid_read_tscalib), and that >> seems >> to work fine? > > Yeah and this is really gross. > Also the bindings aren't upstreamed and can't be without nvmem-cells. It's nice that we're going to have clean upstreamable stuff, but it's a bit sad that making everything work out of the box on current *now* is not a priority. Especially since overlays are perfect for enabling temporary hacks. >> >> > SPI: https://github.com/freebsd/freebsd/pull/166 >> > >> > What is the point of the spigen interface ? >> > If aw_spi is working on H3 (I haven't check but it should), if >> you >> > have a good dtb with the spi flash described you can use it >> directly. >> >> Uhhh I'm surprised you don't know what's the point of spigen? >> The point is raw *userspace* access. > > I'm surprised that you thing I don't know what spigen is :) Hah, sorry, I guess you should've added "for accessing flash" right there before the question mark. > Having a spigen for things other than flash is good, for flash I > don't > see the point. Yeah, it should be exposed out of the box for all kinds of things, exactly! Kernel drivers are good for permanently attached flash, but if you're going to, say, connect various motherboards' flash chips to dump/recover the firmware or flash coreboot, you don't want kernel drivers, you specifically want *flashrom* with its auto detection (and probably its support for more chips…) Fun story: I recovered my PC after a bad firmware update using an RPi once. I couldn't attach all the wires because the board has a header with 2mm pins and I had thicker connectors, so I ended up not connecting ground. It didn't work at default speed, but flashrom detected the chip when using very low speed and flashing was successful :D
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1532613717.49616.0>