Hi Jones and Jin, AFAIK, the dw-spi controller HW has a feature (so-called) that the CS will be de-asserted when the TX FIFO is empty, if you use the chip select register for CS.
So if you meet problem with the chip select stuff, there are 2 ways you may try: 1. Try to keep the TX FIFO not empty 2. Use a dedicated extern GPIO based line for the chip select. There is already a "cs_control" hook in current spi-dw code, and you can tweak your device's platform data to set the right cs_control. (btw: that code is contributed by other developer, and myself haven't manually tested that). - Feng > -----Original Message----- > From: Michael Jones [mailto:[email protected]] > Sent: Tuesday, August 12, 2014 11:25 PM > To: [email protected] > Cc: Tang, Feng > Subject: does cs_change work in spi-dw.c? > > Hello, > > I'm using spi-dw.c via a userspace spidev device. I'm doing an SPI read > access via the ioctl SPI_IOC_MESSAGE(). I would like to execute two > messages, with the chip select staying active between the two. If I > understand correctly, this is what the member 'cs_change' in 'struct > spi_ioc_transfer' is suppose to provide. What I observe, though, is > that the chip select deactivates between messages regardless of how I > set 'cs_change'. > > Can anybody confirm that they're able to use the cs_change switch with > this driver? > > I'm using an older kernel (3.10.31-LTSI), but I've browsed the more > recent changes to this driver and didn't see anything that obviously > effected this feature. > > thanks, > Michael > > MATRIX VISION GmbH, Talstrasse 16, DE-71570 Oppenweiler > Registergericht: Amtsgericht Stuttgart, HRB 271090 > Geschaeftsfuehrer: Uwe Furtner, Erhard Meier -- To unsubscribe from this list: send the line "unsubscribe linux-spi" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
