Are you referring to V8's developer shell (d8)? This has been fixed. The reason is that v8::String::Utf8Value hides exceptions thrown when evaluating the object's toString() method, e.g. stack overflow. This has now been fixed in d8 (http://code.google.com/p/v8/issues/detail?id=2317).
Yang On Tuesday, September 4, 2012 11:43:58 AM UTC+2, marcel wrote: > > Hey I'm seeing an issue where v8 starts behaving oddly when my embedded JS > is reaching high stack space. I put together a simple example of this: > > https://gist.github.com/3619095 > > When I run this the print() function calls start getting "(null)" for the > first parameter, but the correct value from "ii.toString()". I posted the > output of the executable on my system as a comment on the gist, but at some > point print() starts getting null before a throwing undefined shortly after: > > 0 0 > 1 1 > 2 2 > 3 3 > 4 4 > 5 5 > ... > 15708 15708 > 15709 15709 > (null) 15710 > (null) 15711 > (null) 15712 > ... > (null) 15737 > (null) 15738 > (null) 15739 > caught: > undefined > > Where my print() call looks like this: > print(ii, ii.toString()); > > Is this behavior expected? I'm currently troubleshooting an issue in > node-fibers where I'm seeing segfaults at high stack space instead of the > expected RangeError and I'm concerned that these issues are related. > > On the other hand this is happening on bleeding_edge, as well as very old > versions (I tested 3.6.6.25 for instance). > > gcc version 4.5.3 both 32 and 64 bit; Linux 3.2.1 > -- v8-users mailing list [email protected] http://groups.google.com/group/v8-users
