From: Andre Przywara Date: Tue, 12 Feb 2019 14:48:47 +0000 Subject: [U-Boot] [linux-sunxi] [PATCH v2 0/4] sunxi: SPI: Add SUN6I generation SPI support In-Reply-To: References: <20190212104114.225551-1-andre.przywara@arm.com> Message-ID: <20190212144847.5fd173f8@donnerap.cambridge.arm.com> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: u-boot@lists.denx.de On Tue, 12 Feb 2019 16:15:22 +0530 Jagan Teki wrote: > On Tue, Feb 12, 2019 at 4:11 PM Andre Przywara wrote: > > > > We already have a driver for the SPI devices in older Allwinner CPUs. > > Newer SoCs (SUN6I generation) have a similar, but improved IP block. > > This series adds support for that, so we can access the SPI flash that > > some newer boards come with. > > This series gathers some patches that have appeared on the list in one > > form or another before (hence the v2 tag), but brings them all together > > and updates them. > > > > The SUN4I/SUN6I SPI devices share a very similar architecure, but differ > > annoyingly in quite some details, to a point where a shared driver > > would look unnecessarily complicated. So we use an #ifdef based approach, > > which keeps the driver simple, at the cost of not being able to access > > both types of device in one binary build. But since there are no SoCs > > which feature both types, this is a theoretical disadvantage. > > > > Patch 1/4 moves the clock toggling from the probe/remove to the > > claim/release-bus stage. > > Patch 2/4 adds the SPI gate clocks and reset gates to the clock drivers, > > also describes at least the enable bit for the SPI mod clock. This is used > > in patch 3/4, which uses the new DM clock framework in the driver, helping > > to abstract differences between the SUN4I/SUN6I types. > > Patch 4/4 eventually adds support for the new SUN6I generation SPI device. > > The actual device support is determined at compile time, based on the > > CONFIG_SUNXI_GEN_SUN6I symbol. > > > > What's left out of this series is the config bits needed to actually > > activate the support for a board. At the moment this is unnecessarily > > verbose, so that Oskari is looking at simplifying this at the Kconfig > > level, which would complement this series. > > > > This has been briefly tested on the Pine64-LTS and the OrangePi PC2 board. > > Please test them if you have other devices with a SUN6I SPI generation. > > I have new series for this which I sent two days back [1]. which uses > same sun4i driver and implement enum reg set to avoid ifdef. > > [1] https://patchwork.ozlabs.org/cover/1039178/ Ah, sorry, I missed that completely. Will have a look at this later on. Cheers, Andre.