> On 10/19/2015 09:35 AM, Daniel Gryniewicz wrote: > > On Mon, Oct 19, 2015 at 9:13 AM, Swen Schillig <s...@vnet.ibm.com> > wrote: > >> I'm not talking about recovery, > >> just about how and where to end a program. > >> ...and I still believe a log-function is not the right place to that. > > > > I've always viewed a fatal log function as an abort (and an abort as a > > fatal log function, since you shouldn't ever have one without the > > other). That said, I'd prefer not to have thousands of lines of > > untested error checking code just to allow us to exit all the way back > > to main to we can abort. I prefer to abort at the time of failure. > > It's more likely to leave useful information in a backtrace or in the > > log. Note that the only non-main() exit in Ganesha is the Fatal() > > abort function, so I consider it a single point of exit anyway. > > +1 > > A useful backtrace is a big plus in my book.
Sometime I want to resurrect the code that we used to have that would put a call stack into the log on fatal conditions. Of course in the case of a genuine code bug, it's best to abort since the log functionality itself may be compromised. Frank --- This email has been checked for viruses by Avast antivirus software. https://www.avast.com/antivirus ------------------------------------------------------------------------------ _______________________________________________ Nfs-ganesha-devel mailing list Nfs-ganesha-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nfs-ganesha-devel
