This is the mail archive of the
cygwin
mailing list for the Cygwin project.
Re: cygwin64: Issue with sigaddset
- From: Richard Elberger <rich at richelberger dot com>
- To: cygwin at cygwin dot com
- Date: Mon, 23 Feb 2015 09:52:43 -0500
- Subject: Re: cygwin64: Issue with sigaddset
- Authentication-results: sourceware.org; auth=none
- References: <B7278424-8EBE-4119-9F02-F68CB98388A3 at richelberger dot com> <20150223124239 dot GN437 at calimero dot vinschen dot de> <6627A2BA-85A0-47F4-A8AC-BB8EC4A0D456 at richelberger dot com> <20150223140353 dot GQ437 at calimero dot vinschen dot de>
Corinna,
Works great! Good thing I just ran the test before blowing away the perl distribution becauseâ.
Unfortunately now I think I really screwed things up. I never did a snapshot install before. I followed the instructions in the FAQ.
So now I am getting an incompatible error from ld looking at libcygwin.a when searching -lcygwin (I wanted to rebuild perl from scratch to make absolutely sure). I looked at the FAQ to see what I have to do to recover but couldnât find any relevant notes.
Help is welcome before I use the sledgehammer and run setup.exe again.
â rich
> On Feb 23, 2015, at 9:03 AM, Corinna Vinschen <corinna-cygwin@cygwin.com> wrote:
>
> On Feb 23 07:59, Richard Elberger wrote:
>> Awesome ;) Thanks so much I will keep a look out for it and verify on
>> my end that the perl unit test works as expected.
>>
>>
>>> On Feb 23, 2015, at 7:42 AM, Corinna Vinschen <corinna-cygwin@cygwin.com> wrote:
>>>
>>> Hi Richard,
>>>
>>> On Feb 21 16:17, Richard Elberger wrote:
>>>> Hello,
>>>>
>>>> I found this while trying to compile perl 5.20.1 on latest cygwin64
>>>> (just updated yesterday).
>>>>
>>>> Effectively, sigaddset is adding a signal along these lines. If
>>>> signal x is less than 32, then two signals are added: x and x+32. If
>>>> signal x is greater than 32, then again two signals are added, x and
>>>> x-32.
>>>>
>>>> Itâs been far too many years since Iâve written C but the attached
>>>> code snippet to prove it (hopefully it proves it) â and running the
>>>> same âcodeâ (I put that in quotes because itâs like preschool C) on
>>>> any Linux works as expected - just adding the one signal.
>>>>
>>>> This is causing the sigset.t test for the POSIX module in perl core to
>>>> fail. There are a ton of other failures in testing 5.20.1 core but
>>>> hopefully theyâre not all cygwin related like this one.
>>>
>>> Thanks a lot for the testcase. I fixed the problem (an int overflow
>>> computing a bit mask) in CVS. I'll generate a developer snapshot on
>>> https://cygwin.com/snapshots/ and probably another Cygwin 1.7.35 test
>>> release soon. Both will contain that patch.
>
> Snapshots are up. Please give it a try.
>
>
> Thanks,
> Corinna
>
> --
> Corinna Vinschen Please, send mails regarding Cygwin to
> Cygwin Maintainer cygwin AT cygwin DOT com
> Red Hat
--
Problem reports: http://cygwin.com/problems.html
FAQ: http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple