From mboxrd@z Thu Jan 1 00:00:00 1970 From: Paul Mundt Date: Wed, 27 Jan 2010 07:27:14 +0000 Subject: Re: [PATCH] spi_sh_msiof: Fixed data sampling on the correct edge Message-Id: <20100127072713.GH2739@linux-sh.org> List-Id: References: <95F51F4B902CAC40AF459205F6322F0171E8D4959B@BMK019S01.emtrion.local> In-Reply-To: <95F51F4B902CAC40AF459205F6322F0171E8D4959B@BMK019S01.emtrion.local> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: linux-sh@vger.kernel.org On Wed, Jan 27, 2010 at 04:23:36PM +0900, Magnus Damm wrote: > On Thu, Jan 21, 2010 at 7:28 PM, Pietrek, Markus > wrote: > > I'm using spidev to communicate with a samsung oled display S6E63D6. > > This display requires the mode SPI_CPHA and SPI_CPOL. > > > > Plain writing is fine with the old register setting (REDG=!TEDG). I > > can configure the display. But when reading the chip identifier, the > > last byte (of the identified 0x63D6) is never stored in the FIFO. I > > always read 0x6363, so the FIFO returns the 2nd last byte as the last > > byte. > > Huh, is that so? > > I tested this on my Ecovec sh7724 board with MMC-over-SPI and your > patch does not break my setup... > > > With a logical analyzer I can proof that the device is sending the > > correct data. So it must be an issue of the sampling. And changing > > REDG/TEDG to the values recommended in the SH7723 HW Reference Manual > > fixed this issue. > > .. and following the data sheet sounds like a good plan. Especially > since it fixes your issue. > > Can you please send a patch without the debug printouts to > spi-devel-general and add my acked-by? > > Acked-by: Magnus Damm > This looks like something we should handle for 2.6.33 anyways, so Markus when you resubmit this make sure to provide your Signed-off-by as well, then I'll batch it up with the next round of updates, thanks.