On Tuesday, September 9, 2014 8:09:32 AM UTC-4, [email protected] wrote: > > Yes, I can see 'rs485 v1.1'. The python code still works fine out of the > box. C code Rx and Tx lines work OK but RTS still does nothing. > When I exported the GPIO9 by hand, dmesg got: > [ 222.055685] gpio_request: gpio-9 (RS485 TXE) status -16 > [ 222.055730] omap_uart 481a8000.serial: Could not request GPIO 9 : -16 > > That means that 485 kernel support works. Without manually exporting it > says just "[ 537.089146] rs485 v1.1" and does nothing. > > BUT now I've discovered something strange with logical level flags in > struct serial_rs485, maybe I'm using different header or maybe I don't > understand it correctly. > When I changed the SER_RS485_RTS_AFTER_SEND to 0 and SER_RS485_RTS_ON_SEND > to 0,* it started working fine*. > That means, when sending data, RTS goes HIGH, after sending, it goes > immediately LOW , that means, > I expected setting ON_SEND to 1 and AFTER_SEND to 0, but with this setting > it just changed from LOW to LOW. >
Oh yeah, forgot about that. There's a couple ternary operators in the driver that I think might be backwards, which makes it the logic levels a bit counter intuitive. It's: SER_RS485_ENABLED | SER_RS485_USE_GPIO for an active high RE/DE signal and: SER_RS485_ENABLED | SER_RS485_USE_GPIO | SER_RS485_RTS_ON_SEND | SER_RS485_RTS_AFTER_SEND for an inverted, active low RE/DE signal. > > Thank you both again for Your valuable help. > I'm satisfied now :-) > > > Dne pondělí, 8. září 2014 19:12:55 UTC+2 Alexander Hiam napsal(a): > > > > Normally If you look at the log (dmsg) you should see a line rs485 > something. It's a printk that I show when the ioctl work. > > Right, it's 'rs485 v1.1'. If you see that in dmesg then the ioctl worked > and it's almost certainly a pinmux issue. > > > And of course you have to configure the pin correctly = pin mode for TX, > rx, gpio. > Le 8 sept. 2014 17:27, <[email protected]> a écrit : > > I've downloaded Robert C Nelson's code and run build_kernel.sh and > install_kernel.sh. I've controlled that sources are patched by rs485 patch. > BBB booted up normally and uname -r says that I am running new kernel. But > I don't have patched kernel headers. I have created serial.h header as > following: > 1 #include <linux/types.h> > 2 > 3 > 4 struct serial_rs485 { > 5 __u32 flags; /* RS485 feature flags */ > 6 #define SER_RS485_ENABLED (1 << 0) /* If enabled > */ > 7 #define SER_RS485_RTS_ON_SEND (1 << 1) /* Logical > level for > 8 RTS pin > when > 9 sending */ > 10 #define SER_RS485_RTS_AFTER_SEND (1 << 2) /* Logical > level for > 11 RTS pin > after sent*/ > 12 #define SER_RS485_RTS_BEFORE_SEND (1 << 3) > 13 #define SER_RS485_USE_GPIO (1 << 5) > 14 //#define SER_RS485_RX_DURING_TX (1 << 4) > 15 __u32 delay_rts_before_send; /* Delay before send > (milliseconds) */ > 16 __u32 delay_rts_after_send; /* Delay after send > (milliseconds) */ > 17 // __u32 padding[5]; /* Memory is cheap, new > structs > 18 __u32 gpio_pin; /* GPIO Pin Index */ > 19 __u32 padding[4]; /* Memory is cheap, new structs > 20 are a royal PITA .. */ > 21 }; > > > and used it in small example program: > #include <fcntl.h> > #include <termios.h> > #include <sys/ioctl.h> > #include <sys/types.h> > #include "serial.h" > //#include <include/uapi/linux/serial.h> > #include <stdio.h> > #include <unistd.h> > > int main() > { > int fd = open("/dev/ttyO4", O_RDWR | O_NOCTTY | O_NDELAY); > if (fd < 0) > { > perror("Error opening tty!"); > return 1; > } > > struct serial_rs485 rs485conf; > > rs485conf.flags |= SER_RS485_ENABLED; > > rs485conf.flags |= SER_RS485_USE_GPIO; > rs485conf.gpio_pin = 9; > rs485conf.flags |= SER_RS485_RTS_ON_SEND; > rs485conf.flags &= ~(SER_RS485_RTS_AFTER_SEND); > rs485conf.delay_rts_before_send = 0; > rs485conf.delay_rts_after_send = 0; > > if (ioctl (fd, TIOCSRS485, &rs485conf) < 0) > { > perror("Bad ioctl"); > } > > struct termios ttyc; > tcgetattr(fd, &ttyc); > > cfsetospeed(&ttyc, B9600); > cfsetispeed(&ttyc, B9600); > ttyc.c_cflag &= ~CRTSCTS; > ttyc.c_cflag |= CS8 | CLOCAL | CREAD; > ttyc.c_cflag |= CRTSCTS; > ttyc.c_cc[VMIN] = 1; > ttyc.c_cc[VTIME] = 5; > tcsetattr(fd, TCSANOW, &ttyc); > > char data = 0xAA; > > int i = 0; > while(i < 10000) > { > if(write(fd, &data, 1) < 1) > { > perror("Error sending char"); > break; > } > > i++; > } > > if (close (fd) < 0) > { > perror("Error closing tty!"); > return 1; > } > > > return 0; > } > but RTS pin stays still LOW. > > > Dne čtvrtek, 4. září 2014 21:51:58 UTC+2 Mickae1 napsal(a): > > In the Debian image from Robert C Nelson, the patch rs485 is already > included. > Le 4 sept. 2014 20:57, <[email protected]> a écrit : > > Thank you for your answers. The python code is working perfectly. > I need to use it in greater C project though. So I am compiling new kernel > including RS485 patch. I have tried /opt/scripts/tools/update_kernel.sh > hoping that it includes the RS485 patch ...unsuccessfully. > I am now following steps from http://jkridner.wordpress.com/ > 2014/06/04/yet-another-set-of-notes-on-building-beaglebone-kernel/ > > Dne středa, 3. září 2014 17:57:53 UTC+2 [email protected] napsal(a): > > Hello, > I'm writing an app for BeagleBoneBlack running debian (3.8.13-bone50). I > would like to use UART4 to communicate with RS485 transmitter over > P9.24(UART4 Tx), P9.26(UART4 Rx) and P8.33 (UART4 RTS). > I've disabled HDMI and enabled overlays BB-UART4 and BB-UART4-RTSCTS > > cat /sys/devices/bone_capemgr.9/slots > 0: 54:PF--- > 1: 55:PF--- > 2: 56:PF--- > 3: 57:PF--- > 4: ff:P-O-L Bone-LT-eMMC-2G,00A0,Texas Instrument,BB-BONE-EMMC-2G > 5: ff:P-O-- Bone-Black-HDMI,00A0,Texas Instrument,BB-BONELT-HDMI > 6: ff:P-O-- Bone-Black-HDMIN,00A0,Texas Instrument,BB-BONELT-HDMIN > 7: ff:P-O-L Override Board Name,00A0,Override Manuf,BB-UART4 > 10: ff:P-O-L Override Board Name,00A0,Override Manuf,BB-UART4-RTSCTS > > cat /proc/tty/driver/OMAP-SERIAL > serinfo:1.0 driver revision: > 0: uart:OMAP UART0 mmio:0x44E09000 irq:72 tx:345 rx:0 RTS|CTS|DTR|DSR > 4: uart:OMAP UART4 mmio:0x481A8000 irq:45 tx:61355 rx:1 brk:1 RTS|DTR|DSR > > > RS485 transmitter is connected through RS485-USB converter to PC. When I > run screen /dev/ttyO4 9600 +crtscts and periodicaly write some data to it, > PC receives it properly, but RTS line stays constantly low (I'm using scope > on Tx and RTS lines). > I've also tried to write simple C program, using *struct serial_rs485. *When > I write some data over this program, I got response: *Resource > temporarily unavailable* and dmesg says > *omap_uart 481a8000.serial: Must use GPIO for RS485 Support. *When I > tried to use: > struct serial_rs485 rs485conf; > rs485conf.flags |= SER_RS485_USE_GPIO; > rs485conf.gpio_pin = GPIO0_9; > > I got error from gcc that it does not know those macros: > ‘SER_RS485_USE_GPIO’ was not declared in this scope > rs485conf.flags |= SER_RS485_USE_GPIO; > ‘struct serial_rs485’ has no member named ‘gpio_pin’ > rs485conf.gpio_pin = GPIO0_9; > ‘GPIO0_9’ was not declared in this scope > rs485conf.gpio_pin = GPIO0_9; > > Could somebody help me? I have no more ideas. > > > > > > -- > For more options, visit http://beagleboard.org/discuss > --- > You received this message because you are subscribed to the Google Groups > "BeagleBoard" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > For more options, visit https://groups.google.com/d/optout. > > -- > For more options, visit htt <http://beagleboard.org/discuss> > > ... -- For more options, visit http://beagleboard.org/discuss --- You received this message because you are subscribed to the Google Groups "BeagleBoard" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
