This is the mail archive of the
cygwin
mailing list for the Cygwin project.
Re: Public key authorization problem with latest snapshot
- From: Christopher Faylor <cgf-use-the-mailinglist-please at cygwin dot com>
- To: cygwin at cygwin dot com
- Date: Sat, 29 Mar 2014 15:29:28 -0400
- Subject: Re: Public key authorization problem with latest snapshot
- Authentication-results: sourceware.org; auth=none
- References: <5329B9D0 dot 6070703 at cornell dot edu> <20140319160914 dot GD2715 at calimero dot vinschen dot de> <5329C9EF dot 1000305 at cornell dot edu> <20140319170508 dot GF2715 at calimero dot vinschen dot de> <885290127 dot 20140320015843 at yandex dot ru> <20140320150258 dot GD3729 at calimero dot vinschen dot de> <5336F3E3 dot 5000402 at cornell dot edu>
- Reply-to: cygwin at cygwin dot com
On Sat, Mar 29, 2014 at 12:25:07PM -0400, Ken Brown wrote:
>On 3/20/2014 11:02 AM, Corinna Vinschen wrote:
>> On Mar 20 01:58, Andrey Repin wrote:
>>> Greetings, Corinna Vinschen!
>>>
>>>> The code is now practically equivalent to what is in 1.7.28. Only the
>>>> VectoredContinueHandler, which was the reason Cygwin's exception handler
>>>> could be called twice, is not called anymore. Instead there's a vectored
>>>> exception handler which is only called during debugging.
>>>
>>>> Before:
>>>
>>>> if (!handler_installed)
>>>> {
>>>> handler_installed = true;
>>>> SetUnhandledExceptionFilter (handle);
>>>> AddVectoredContinueHandler (1, handle);
>>>> }
>>>
>>>> After:
>>>
>>>> if (!handler_installed)
>>>> {
>>>> handler_installed = true;
>>>> SetUnhandledExceptionFilter (handle);
>>>> AddVectoredExceptionHandler (1, handle_while_being_debugged);
>>>> }
>>>
>>>> If anybody can explain this weird behaviour, please educate me.
>>>
>>> I can't explain the behavior, but I could say, that setting
>>> "handler_installed = true;" before the handler is actually installed is not
>>> quite right.
>>> Unless that variable is used inside either of two functions called afterward,
>>> I would move it down to the end of `if' block.
>>
>> BTDT. This isn't the problem. I *may* have found the culprit today,
>> but I ripped apart a lot of the code so I'm not really sure yet. Stay
>> tuned.
>
>The problems I've reported seem to all be fixed in the latest snapshot
>(2014-03-29 15:21:43 UTC). Thanks!
I'm sure Corinna will be happy to hear that. She put in LONG hours
getting that issue sorted out.
I helped too, of course, by offering important "I don't like that
implementation" style feedback. It was one of those 50/50 collaborations
where one person does all the work and the other person mentions it on
a mailing list.
cgf
--
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