I need some hint regarding opening a serial (ttyUSB) port using the unit serial. Minicom works fine under my user id. When i call SerOpen('/dev/ttyUSB0'), fpopen blocks forever. When i first start minicom on that port, SerOpen and all other functions of serial.pp works fine.

ttyUSB0 is:
crw-rw---- 1 root dialout 188, 0 Aug 18 09:37 /dev/ttyUSB0

changing it to
crw-rw-rw- 1 root dialout 188, 0 Aug 18 09:38 ttyUSB0

makes no difference. (my user id has the supplemental group "dialout" added) SerOpen will always block if minicom is not started on the same port, the same applies if i try it wit root.

stty with minicom started:
speed 19200 baud; rows 0; columns 0; line = 0;
intr = ^C; quit = ^\; erase = ^?; kill = ^U; eof = ^A; eol = <undef>; eol2 = <undef>; swtch = <undef>; start = ^Q; stop = ^S; susp = ^Z; rprnt = ^R; werase = ^W; lnext = ^V; flush = ^O; min = 1; time = 5;
-parenb -parodd cs8 hupcl -cstopb cread clocal crtscts ignbrk -brkint -ignpar -parmrk -inpck -istrip -inlcr -igncr -icrnl ixon ixoff -iuclc -ixany -imaxbel -iutf8
-opost -olcuc -ocrnl -onlcr -onocr -onlret -ofill -ofdel nl0 cr0 tab0 bs0 vt0 ff0
-isig -icanon -iexten -echo -echoe -echok -echonl -noflsh -xcase -tostop -echoprt -echoctl -echoke

stty without mnicom started:
speed 19200 baud; rows 0; columns 0; line = 0;
intr = ^C; quit = ^\; erase = ^?; kill = ^U; eof = ^A; eol = <undef>; eol2 = <undef>; swtch = <undef>; start = ^Q; stop = ^S; susp = ^Z; rprnt = ^R; werase = ^W; lnext = ^V; flush = ^O; min = 1; time = 0;
-parenb -parodd cs8 hupcl -cstopb cread -clocal -crtscts -ignbrk -brkint -ignpar -parmrk -inpck -istrip -inlcr -igncr -icrnl -ixon -ixoff -iuclc -ixany -imaxbel -iutf8
-opost -olcuc -ocrnl -onlcr -onocr -onlret -ofill -ofdel nl0 cr0 tab0 bs0 vt0 ff0
-isig -icanon -iexten -echo -echoe -echok -echonl -noflsh -xcase -tostop -echoprt -echoctl -echoke

diff
time = 0; -clocal -crtscts -ixon -ixof

Fedora 14, 64 Bit
2.6.35.13-92.fc14.x86_64 #1 SMP Sat May 21 17:26:25 UTC 2011 x86_64 x86_64 x86_64 GNU/Linux

Thx
Armin

Assuming that you mean the standard serial.pp, that's not something I've seen here /but/ I'm using a 32-bit kernel. I think that the first thing I'd explore is whether you need the incoming control lines (cts, dsr, cd) to be asserted before the function completes, but quite frankly if I'm reading your

When i first start minicom on that port, SerOpen and all other functions of serial.pp works fine.

correctly then it sounds like a kernel or driver issue.

Anybody: I've modified serial.pp to support Win-32 and fix a couple of issues. Bug http://62.166.198.202/view.php?id=18946

I don't use serial.pp but here is how I open a serial port for basic communication (no flow control, modem control or stuff like that) under linux. I remember this taking some experimentation to get right.

procedure tlserial.open;
var
  fd : longint;
  config : termios;
  baudrateos : longint;
begin
  fd := fpopen(device,O_RDWR or O_NOCTTY);
  if isatty(fd)=0 then begin
    writeln('not a tty');
    halt(1);
  end;
  fillchar(config,sizeof(config),#0);
  config.c_cflag := CLOCAL or CREAD;
  cfmakeraw(config);
  case baudrate of
    50:     baudrateos := B50;
    75:     baudrateos := B75;
    110:    baudrateos := B110;
    134:    baudrateos := B134;
    150:    baudrateos := B150;
    200:    baudrateos := B200;
    300:    baudrateos := B300;
    600:    baudrateos := B600;
    1200:   baudrateos := B1200;
    1800:   baudrateos := B1800;
    2400:   baudrateos := B2400;
    4800:   baudrateos := B4800;
    9600:   baudrateos := B9600;
    19200:  baudrateos := B19200;
    38400:  baudrateos := B38400;
    57600:  baudrateos := B57600;
    115200: baudrateos := B115200;
    230400: baudrateos := B230400;
    else raise exception.create('unrecognised baudrate');
  end;
  cfsetispeed(config,baudrateos);
  cfsetospeed(config,baudrateos);
  config.c_cc[VMIN]  := 1;
  config.c_cc[VTIME] := 0;
  if   tcsetattr(fd,TCSAFLUSH,config) <0 then begin
    writeln('could not set termios attributes');
    halt(3);
  end;
  dup(fd);
  closehandles := true;
end;

Hi Peter,

that is the same as in the standard serial.pp:

Result := fpopen(DeviceName, O_RDWR or O_NOCTTY);

and this fpopen will block as long as minicom is not started on that device.

In our previous episode, Armin Diehl said:

Yes. And seems that minicom turns of xon/xoff. You could try to verify that by disabling it with stty.

Try to add O_NONBLOCK to the flags with fpopen. Then set CLOCAL like Peter suggested. Then disable O_NONBLOCK again with fcntl if you want blocking I/O back.

Henry

Just tried it on a 32 bit system (Linux t41.armin.d 2.6.35.13-92.fc14.i686 #1 SMP Sat May 21 17:39:42 UTC 2011 i686 i686 i386 GNU/Linux) as root

uses baseunix;
var handle : longint;
begin
   writeln('Start: fpopen');
   handle := fpopen('/dev/ttyUSB0',O_RDWR or O_NOCTTY);
   writeln ('Result: ',Handle);
   if handle  0 then fpclose(handle);
end.

and exactly the same problem. fpopen blocks if minicom is not started on the same port. I think i have to figure out how minicom (or stty) opens the port.

btw, ttyUSB0 is a Prolific pl2303:

Aug 18 11:47:18 t41 kernel: [ 599.994081] usb 3-2: new full speed USB device using uhci_hcd and address 2
Aug 18 11:47:18 t41 kernel: [ 600.140165] usb 3-2: New USB device found, idVendor=067b, idProduct=2303
Aug 18 11:47:18 t41 kernel: [ 600.140180] usb 3-2: New USB device strings: Mfr=1, Product=2, SerialNumber=0
Aug 18 11:47:18 t41 kernel: [ 600.140191] usb 3-2: Product: USB-Serial Controller
Aug 18 11:47:18 t41 kernel: [ 600.140200] usb 3-2: Manufacturer: Prolific Technology Inc.
Aug 18 11:47:18 t41 mtp-probe: checking bus 3, device 2: "/sys/devices/pci0000:00/0000:00:1d.1/usb3/3-2"
Aug 18 11:47:18 t41 mtp-probe: bus: 3, device: 2 was not an MTP device Aug 18 11:47:18 t41 kernel: [ 600.329041] usbcore: registered new interface driver usbserial
Aug 18 11:47:18 t41 kernel: [ 600.329070] USB Serial support registered for generic
Aug 18 11:47:18 t41 kernel: [ 600.329948] usbcore: registered new interface driver usbserial_generic
Aug 18 11:47:18 t41 kernel: [ 600.329953] usbserial: USB Serial Driver core Aug 18 11:47:18 t41 kernel: [ 600.345812] USB Serial support registered for pl2303
Aug 18 11:47:18 t41 kernel: [ 600.346655] pl2303 3-2:1.0: pl2303 converter detected
Aug 18 11:47:18 t41 kernel: [ 600.358507] usb 3-2: pl2303 converter now attached to ttyUSB0
Aug 18 11:47:18 t41 kernel: [ 600.359439] usbcore: registered new interface driver pl2303
Aug 18 11:47:18 t41 kernel: [ 600.359445] pl2303: Prolific PL2303 USB to serial adaptor driver

Yes, found that a few minutes ago (strace is nice), micom does:

open("/dev/ttyUSB0", O_RDWR|O_NOCTTY|O_NONBLOCK) = 3 fcntl(3, F_GETFL) = 0x8802 (flags O_RDWR|O_NONBLOCK|O_LARGEFILE)
fcntl(3, F_SETFL, O_RDWR|O_LARGEFILE) = 0
ioctl(3, SNDCTL_TMR_TIMEBASE or SNDRV_TIMER_IOCTL_NEXT_DEVICE or TCGETS, {B19200 -opost -isig -icanon -echo ...}) = 0
ioctl(3, TIOCMGET, [TIOCM_DTR|TIOCM_RTS]) = 0
ioctl(3, SNDCTL_TMR_TIMEBASE or SNDRV_TIMER_IOCTL_NEXT_DEVICE or TCGETS, {B19200 -opost -isig -icanon -echo ...}) = 0
ioctl(3, SNDCTL_TMR_START or SNDRV_TIMER_IOCTL_TREAD or TCSETS, {B19200 -opost -isig -icanon -echo ...}) = 0
ioctl(3, SNDCTL_TMR_TIMEBASE or SNDRV_TIMER_IOCTL_NEXT_DEVICE or TCGETS, {B19200 -opost -isig -icanon -echo ...}) = 0
ioctl(3, TIOCMGET, [TIOCM_DTR|TIOCM_RTS]) = 0
ioctl(3, TIOCMSET, [TIOCM_DTR|TIOCM_RTS]) = 0
ioctl(3, SNDCTL_TMR_TIMEBASE or SNDRV_TIMER_IOCTL_NEXT_DEVICE or TCGETS, {B19200 -opost -isig -icanon -echo ...}) = 0
ioctl(3, SNDCTL_TMR_START or SNDRV_TIMER_IOCTL_TREAD or TCSETS, {B19200 -opost -isig -icanon -echo ...}) = 0
ioctl(3, SNDCTL_TMR_TIMEBASE or SNDRV_TIMER_IOCTL_NEXT_DEVICE or TCGETS, {B19200 -opost -isig -icanon -echo ...}) = 0
ioctl(3, SNDCTL_TMR_TIMEBASE or SNDRV_TIMER_IOCTL_NEXT_DEVICE or TCGETS, {B19200 -opost -isig -icanon -echo ...}) = 0
ioctl(3, SNDCTL_TMR_START or SNDRV_TIMER_IOCTL_TREAD or TCSETS, {B19200 -opost -isig -icanon -echo ...}) = 0
ioctl(3, SNDCTL_TMR_TIMEBASE or SNDRV_TIMER_IOCTL_NEXT_DEVICE or TCGETS, {B19200 -opost -isig -icanon -echo ...}) = 0
ioctl(3, SNDCTL_TMR_TIMEBASE or SNDRV_TIMER_IOCTL_NEXT_DEVICE or TCGETS, {B19200 -opost -isig -icanon -echo ...}) = 0
ioctl(3, SNDCTL_TMR_START or SNDRV_TIMER_IOCTL_TREAD or TCSETS, {B19200 -opost -isig -icanon -echo ...}) = 0
ioctl(3, SNDCTL_TMR_TIMEBASE or SNDRV_TIMER_IOCTL_NEXT_DEVICE or TCGETS, {B19200 -opost -isig -icanon -echo ...}) = 0
ioctl(3, TCFLSH, 0x2) = 0

should that be changed in the standard serial.pp ?

In our previous episode, Armin Diehl said:

should that be changed in the standard serial.pp ?

I'm not sure, since this leaves nonblocking behaviour on by default, moreover serial is pan unix, and non linuxes might react differently.

Btw do you see minicom do a sequence of "opening in nonblock, setting params and reopening without nonblock" ?

No it doesn't, the fcntl call disables it again.

Henry

Hi Marco,

minicom calls open with O_NONBLOCK and resets O_NONBLOCK directly after the open call:

open("/dev/ttyUSB0", O_RDWR|O_NOCTTY|O_NONBLOCK) = 3 fcntl(3, F_GETFL) = 0x8802 (flags O_RDWR|O_NONBLOCK|O_LARGEFILE)
fcntl(3, F_SETFL, O_RDWR|O_LARGEFILE) = 0

doing the same in serial.pp should not change the blocking behaviour of serial.pp.

function SerOpen(const DeviceName: String): TSerialHandle;
begin
   Result := fpopen(DeviceName, O_RDWR or O_NOCTTY or O_NONBLOCK);  //
AD: O_NONBLOCK was missing
   if result > -1 then
     fpfcntl(Result,F_SETFL,O_RDWR or O_NOCTTY);
//AD: non blocking off again
end;

Better use F_GETFL first, then flags &= ~O_NONBLOCK, like minicom does.

Henry

yes, that is better

function SerOpen(const DeviceName: String): TSerialHandle;
var
   flags : cint;
begin
   Result := fpopen(DeviceName, O_RDWR or O_NOCTTY or O_NONBLOCK);
   if result > -1 then
   begin
     flags := fpfcntl(Result,F_GetFl);
     fpfcntl(Result,F_SETFL,flags and (not O_NONBLOCK));
   end;
end;

results in
open("/dev/ttyUSB0", O_RDWR|O_NOCTTY|O_NONBLOCK|O_LARGEFILE) = 10 fcntl(10, F_GETFL) = 0x8802 (flags O_RDWR|O_NONBLOCK|O_LARGEFILE)
fcntl(10, F_SETFL, O_RDWR|O_LARGEFILE) = 0

i think this should be changed in serial.pp

But why have I not seen this problem on Debian despite having used this unit extensively? And why have I not seen it when testing the modified unit on Solaris with the SerOpen() function being unchanged?

Mark Morgan Lloyd wrote

But why have I not seen this problem on Debian despite having used this unit extensively? And why have I not seen it when testing the modified unit on Solaris with the SerOpen() function being unchanged?

Some serial ports seem to have clocal unset, for some reason, which causes it to block on open (waiting for the modem status lines). Like you, I've never seen it with any of my serial ports, either.

Henry

In our previous episode, Henry Vermaak said:

Might have to do with the defaulting to autobaud of some usbserial chips.

Apparently a bug in a recent usb serial kernel module pl2303:

https://bugs.launchpad.net/ubuntu/+source/linux/+bug/661321

Ludo

Well spotted :-) I was going to speculate that it might be something to do with the Prolific driver- I think that the device I use most is FTDI.

there must be a reason why minicom uses O_NDELAY (=O_NONBLOCK) in case O_NDELAY ist defined. Now i remember that i had the same problem several years ago with cat file >/dev/ttyUSB0, i think it was Fedora Core 3 or 4. I assume this does not only happen with a pl2303.

#if defined(O_NDELAY) && defined(F_SETFL)

       portfd = open(dial_tty, O_RDWR|O_NDELAY|O_NOCTTY);
       if (portfd >= 0) {
         /* Cancel the O_NDELAY flag. */
         n = fcntl(portfd, F_GETFL, 0);
         fcntl(portfd, F_SETFL, n & ~O_NDELAY);
       }

#else
if (portfd < 0)

         portfd = open(dial_tty, O_RDWR|O_NOCTTY);

#endif