[Boost-users] [ASIO] Problem with reading of serial ports using ASIO

Subject: [Boost-users] [ASIO] Problem with reading of serial ports using ASIO
From: Jim Donaldson (jim.donaldson_at_[hidden])
Date: 2010-10-19 14:08:34

  Hi All,

I'm having some problems using Boost ASIO on Windows XPSP3, 32 bit.

I'm trying to set-up an asynchronous read from a serial port on the 
computer (actually a USB port masquerading as a serial port). The code 
is at

http://gist.github.com/634618

We regularly call the function
PCTimeoutUart::wait( iseSize_t len, uint32 timeoutInMS ) which wraps the 
calls to the boost asio libraries.

This creates a deadline_timer, and binds it to our timeout callback 
(PCTimeoutUart::finishedTimer)
It also calls async_read_some on the boost::asio::serial_port object 
m_port to read the required length, and binds that to the 
PCTimeoutUart::finishedRead callback.

We call the runone() function on m_io, which is declared in the header 
as a boost::asio::io_service and has been passed to both the timer and 
serial port's constructors.

We continually call this in a loop until either m_timedOut has been set 
by the asynchronous call back to PCTimeoutUart::finishedTimer
or m_finishedRead has been set via the call back to 
PCTimeoutUart::finishedRead

Normally, this works well. However, there are certain scenarios when it 
fails to write anything to the buffer, despite the fact that 
finishedRead() is called with error = 0 and the correct length . I've 
checked by initializing the buffers with a couple of different values 
and each time, it's clear that the buffer is being passed back without 
those values being altered. Strangely, it is always when the string sent 
by the external device is '\r\nNACK:XX\r\n'   where XX is a 2 character 
ascii number.

I've got portmon running on the serial port and it confirms that the 
data has been correctly received from the external device.

We've debugged it as far as putting a break-point on line 71 and at that 
point can see that the system thinks that a read is complete, n bytes 
have been returned, but only some of those bytes are in buffer.
If anyone could offer any views on why we may be having problems, or 
places that we could look, then that would be very much appreciated.

Thanks in advance,

Jim