in await: Cannot invoke this object (REPR: Null; VMNull) #5624
Comments
From @zoffixznetOn occasion, I get this error when trying to run the bellow program on a 32-core box: $ time RAKUDO_MAX_THREADS=40 perl6 foo.p6 sub what-is { my @results; |
From @jnthnOn Sun Aug 28 20:09:29 2016, cpan@zoffix.com wrote:
I've fiddled with it a bit, but had trouble getting this to blow on my much-less-core-endowed box. Guess I can leave it running in a loop while I take lunch or something... :-) Any chance you could: 1) Run it with `--ll-exception` Thanks! /jnthn |
|
The RT System itself - Status changed from 'new' to 'open' |
From @zhuomingliangOn Wed Aug 31 02:58:50 2016, jnthn@jnthn.net wrote:
RAKUDO_MAX_THREADS=40 ./perl6-m --ll-exception test.pl6 |
From @zhuomingliangOn Wed Aug 31 03:13:40 2016, jimmy wrote:
FYI, I got it from a ubuntu 14.04 linux VM box with 4 cores. |
From @nwc10On Wed, Aug 31, 2016 at 02:58:50AM -0700, jnthn@jnthn.net via RT wrote:
"my" machine has almost as many cores. I can get it to bomb at times.
You mean like this? sub what-is { I got this: Cannot invoke this object (REPR: Null; VMNull) Line 9 is `print await @results;` What do I need to do better, to get actual useful details? Nicholas Clark |
From @zoffixznetAttached 'backtrace-original.txt' is the output from the added CATCH and --ll-exception. I was able to golf it down to this: await do for 3¹²³⁴⁵ … 3¹²³⁴⁵+127 -> $num { And that backtrace is also attached (backtrace-golfed.txt). I tried replacing .is-prime with a bcrypt hashing, but was unable to produce the crash in about 8 runs: use Crypt::Bcrypt; So then I went back to my golfed version and removed the ternary operator... Couldn't reproduce anything in about 10 runs. Added a ternary as `?? 1 !! 2`, still couldn't reproduce anything, Changed the ternary to this: await do for 3¹²³⁴⁵ … 3¹²³⁴⁵+127 -> $num { ... and it crashed the first time I ran it (backtrace-changed-ternary.txt attached) and then 1 more time in the next out of 7 runs. It may of course be just a coincidence, but these are my results |
From @zoffixznet in method throw at gen/moar/m-CORE.setting line 24125 Cannot invoke this object (REPR: Null; VMNull) |
From @zoffixznetcpan@perlbuild2:/tmp/tmp.AY07p77AqM$ RAKUDO_MAX_THREADS=40 perl6 --ll-exception test.p6 Cannot invoke this object (REPR: Null; VMNull) |
From @zoffixznetcpan@perlbuild2:/tmp/tmp.AY07p77AqM$ RAKUDO_MAX_THREADS=40 perl6 --ll-exception test.p6 |
From @jnthnOn Wed Aug 31 04:31:25 2016, nicholas wrote:
Yeah, I'm guessing it's a slightly higher level data race due to mis-code-gen rather than any kind of memory corruption.
Yes, that's what I meant.
Well, that's information in and that it means the exception is not happening inside of the sub `what-is`. Unfortunately, the only thing between the two is the stuff in CORE.setting implementing promises and scheduling, which means it's probably a very occasional race somewhere in there. Probably instrumenting the bit of MoarVM that spits out that error to dump a backtrace before throwing is the easiest next thing to do. I'll see if I can recreate it locally outside of my VM. /jnthn |
From @zoffixznetRan for 83 times, not a single crash. Closing. On Wed, 31 Aug 2016 05:49:09 -0700, jnthn@jnthn.net wrote:
|
|
@zoffixznet - Status changed from 'open' to 'resolved' |
Migrated from rt.perl.org#129120 (status was 'resolved')
Searchable as RT129120$
The text was updated successfully, but these errors were encountered: