After upgrading to latest Cygwin recently we have noticed that one of
our scripts started to hang randomly. When, as a part of an
investigation, we have downgraded tcsh from 6.18.01-1 to  6.18.00-3,
the hangs became rare, but still were happening. Only when we has
downgraded to a version 6.17.0 of the tcsh, the hangs has stopped to
occur.

A colleague of mine has created a test scripts that reliably
reproduces the fault on our boxes.

The hang is provoked by running multiple copies of this script, giving
it a different output filename each time. The important bit was having
the pipe inside backquotes, it didn't fail just with pipes.

#!/bin/csh -f
echo "0_test" > $1
while 1
  set fred=`cat $1 | cut -f1 -d_ | awk '{c=$1+1 ; print c "_test"}'`
  echo "$fred" > $1
end

The hangs happens within 10-15 minutes of running 5 scripts at the
same time against 6.18.01-1.

It takes couple hours to get a hang on 6.18.00-3.

There were no problems with 6.17.00, even running 15 copies of that
script in parallel for 5 hours (almost a million interactions).
After reading threads like
http://cygwin.com/ml/cygwin/2010-08/msg00145.html, I realise it is
hard to recreate this sort of problems but can anybody give it a try?

So, for time being for people experiencing similar problem the
workaround seems to be to dig out 6.17.0 from somewhere.

Any other useful information we can provide to help to uncover the
underlying cause?

The cygcheck for the box where the latest cygwin was running is attached.

Thank you in advance
Yuriy