so it will swallow your crash. Besides, the case statement right after that seems to be fairly wrong. It has an all-catch clause NewMFA, and it assumes PROCESSED is a new MFA if it is nothing that matched before. That means, an EXIT tuple will be treated as a new MFA. Unless I'm missing something, this is just wrong.
> so it will swallow your crash. Besides, the case statement right after that seems to be fairly wrong. It has an all-catch clause NewMFA, and it assumes PROCESSED is a new MFA if it is nothing that matched before. That means, an EXIT tuple will be treated as a new MFA. Unless I'm missing something, this is just wrong.
-------------------------------------------------------------------------- - for hire: mac osx device driver ninja, kernel extensions and usb drivers ---------------------+------------+--------------------------------------- http://wagerlabs.com | @wagerlabs | http://www.linkedin.com/in/joelreymont ---------------------+------------+---------------------------------------
> > so it will swallow your crash. Besides, the case statement right after > that seems to be fairly wrong. It has an all-catch clause NewMFA, and it > assumes PROCESSED is a new MFA if it is nothing that matched before. That > means, an EXIT tuple will be treated as a new MFA. Unless I'm missing > something, this is just wrong.
> -------------------------------------------------------------------------- > - for hire: mac osx device driver ninja, kernel extensions and usb drivers > ---------------------+------------+--------------------------------------- > http://wagerlabs.com | @wagerlabs | http://www.linkedin.com/in/joelreymont > ---------------------+------------+---------------------------------------
And what you get in error.log is just a formatted message after catching the exception case (catch apply(Module, do, [ModData])) of {'EXIT', Reason} -> ?hdrd("traverse modules - exit", [{reason, Reason}]), String = lists:flatten( io_lib:format("traverse exit from apply: ~p:do => ~n~p", [Module, Reason])),
On Fri, Dec 2, 2011 at 11:03 AM, Ahmed Omar <spawn.th...@gmail.com> wrote: > Well, there won't be crash report anywhere because inets does catch the > exception in httpd_resonse.erl
> On Thu, Dec 1, 2011 at 2:36 PM, Joel Reymont <joe...@gmail.com> wrote:
>> You can get the crash reports in /tmp/error.log (in my case).
>> This works if the list of httpd modules is set to [mod_log, mod_crash].
>> There's still nothing shown via regular SASL channels.
>> On Dec 1, 2011, at 1:34 PM, Yurii Rashkovskii wrote:
>> > The problem is in httpd. httpd_request_handler has this:
>> > so it will swallow your crash. Besides, the case statement right after >> that seems to be fairly wrong. It has an all-catch clause NewMFA, and it >> assumes PROCESSED is a new MFA if it is nothing that matched before. That >> means, an EXIT tuple will be treated as a new MFA. Unless I'm missing >> something, this is just wrong.
>> -------------------------------------------------------------------------- >> - for hire: mac osx device driver ninja, kernel extensions and usb drivers >> ---------------------+------------+--------------------------------------- >> http://wagerlabs.com | @wagerlabs | >> http://www.linkedin.com/in/joelreymont >> ---------------------+------------+---------------------------------------
