This is the mail archive of the cygwin mailing list for the Cygwin project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Issue with 20111222 snapshot

On Sat, Jan 07, 2012 at 08:25:40AM -0500, Chris Sutcliffe wrote:
>On 7 January 2012 01:21, Christopher Faylor wrote:
>> On Fri, Jan 06, 2012 at 01:57:36PM -0500, Chris Sutcliffe wrote:
>>>Sigh... ?it seems to be an issue with the CVS HEAD mksh I'm testing,
>>>bash behaves as expected.
>> You sure it isn't a regression though? ?There has been a lot of work
>> in the pipe section of code (which seems to always be the case).
>It does actually seem to be a regression.  Testing with mksh 40c and
>Cygwin 1.7.9 works as expected, testing with mksh 40c and the 20111222
>snapshot breaks.  The interesting thing is that bash works fine, so it
>seems to be specific to some interaction between mksh and the
>To recreate the issue, using mksh 40c:
>$ touch test.out
>$ tail -f test.out | grep something
>The Ctrl+C has no affect.  The only way I can get out of the tail is
>to kill it from a separate terminal window.  I've remembered to attach
>the cygcheck output this time.
>If it would be helpful, I can start testing with different snapshots
>to see if I can narrow down which snapshot introduced the issue.
>Since this is 100% reproducible using mksh for me, hopefully you'll be
>able to recreate the issue.

I can reproduce this using the release version of mksh.  It doesn't seem
to be related to my pipe changes but I'll try to fix it ASAP.


Problem reports:
Unsubscribe info:

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]