Test case for 230.4Kbps support patches - uncovers errno propagation problem?

[Troy Curtiss]

The attached program ratchets through the bitrates defined in termios.h and
tests a given (as first argument) serial port.  This test will indicate if
tcsetattr() returns an error when invalid bitrates are attempted on a given
serial port (along with perror() to print the error-message associated w/
errno.)

In the event that an invalid bitrate is attempted on a port, two events
should occur.  1)  tcsetattr() should return -1, and 2) errno should be set
to EINVAL.  Event #1 seems to work fine w/ my latest patch, but event #2 is
failing because errno is always set to 0.  Unless I'm doing something wrong,
it appears that errno isn't being propagated back up from the call to
tcsetattr() ... can someone else take a peek at this?  Thanks,

-Troy
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: test_serial.c
URL: <http://cygwin.com/pipermail/cygwin-patches/attachments/20030122/78703fca/attachment.c>