On 12/04/2013 05:50 AM, Florian Weimer wrote:
On 12/03/2013 04:27 PM, Carlos O'Donell wrote:
I think that's certainly the right start. File a defect with the
Austin Group tracker and ask them the following things:
* What if any unintended consequences could there be if the
implementation choose to save and restore errno during
signal handling?
* Was it ever POSIX's intent to allow a signal handler to
modify the errno of the interrupted code sequence or was
that simply a consequence of being a signal handler and
modifying global state?
... and anything else you think we should get an expert opinion
on before embarking upon a change like this.
Okay, I tried to summarize the previous discussion there:
<http://austingroupbugs.net/view.php?id=807>
Interesting response from Geoff Clare:
"If we make any change as a result of this issue, I think it should
just be to make it explicit that implementations are allowed, but
not required, to restore errno on return from signal handlers."
Which would be a nice addition and clarification to the standard.