fhandler_tty_slave::tcflush() in fhandler_tty.cc
Fri Aug 5 14:17:00 GMT 2005
>From: Vaclav Haisman
>Sent: 05 August 2005 14:01
> fhandler_tty_slave::tcflush() is IMHO still wrong. The result of
> comparison is bool and bool converted to int is either 1 or 0. The
> following patch should cure it.
Just to enlarge upon that, the problem is not just that it's returning a
zero-or-one when it should be returning a zero-or-minus-one result, but that
the logical sense is inverted too: when flushing input, the test means that
it used to return zero for failure and non-zero for success. Your patch
looks good to me.
Can't think of a witty .sigline today....
More information about the Cygwin-patches