Fixing RangeError with code from 1.7 branch#3200
Conversation
|
would be nice if it compiled and could we please have a test/spec this time as well, thanks! |
d840ad5 to
f34200c
Compare
|
This compiles now. My bad for the missing }. Also I added a spec in my previous commit that should have broken the spec suite with the RangeError in question. I've compiled locally successfully but am unable to run specs under specs/ruby from my working copy directory. Do you have any tips on how to do that? Is there a doc describing how the ruby specs are meant to be used? |
|
@kares, i can't seem to understand the Travis failure above. Can you help with that? Also please see my previous question as it relates to the best way to run specs. |
|
@r6p obviously previous spec did not broke the test suite - so we're still need a correct spec here, thanks!
|
|
@kares the fact that the ruby specs on master aren't failing is concerning to me. The same code path exercised via jirb fails. Thoughts? |
With my previous commit, RangeError: bignum too big to convert into `long' was the result when running (0...2**64).max This commit ports handling for non-RubyFixnums from the 1.7.20 branch. Addresses, jruby#3118
f34200c to
34f5c8f
Compare
|
@kares, I finally figured out where to put the test so that it would fail as a part of the build when running the following suite. I added the spec and amended the commit on this branch. You'll see that the test passes below. |
|
excellent, thank you! |
Fixing RangeError with code from 1.7 branch
With my previous commit, RangeError: bignum too big to convert into `long' was the result when running (0...2**64).max
This commit ports handling for non-RubyFixnums from the 1.7.20 branch.
Addresses, #3118
.. follow-up on PR #3196