Probably a core.rrb-vector bug, exercised via fipp library

Description

See the description of the test case involved on this Github issue: https://github.com/brandonbloom/fipp/issues/53

Environment

None

Activity

Show:
Andy Fingerhut
August 4, 2019, 9:40 PM

Attachment crrbv-20-fix-v1.patch is one possible fix for this issue. See the commit comments for some notes. The approach in that patch is a bit different than the one described in my previous comment.

The patch does not include a new test that fails without the fix but succeeds with it. The tests in the deftest named many-subvec-and-catvec-leads-to-exception at the link below, and the functions it calls, contain some test cases that fail without these changes, and work with them.

Andy Fingerhut
August 7, 2019, 1:56 AM

Removed crrbv-20-fix-v1.patch in favor of the slightly different crrbv-20-fix-v2.patch, which is the same except the new comments in the code are more accurate.

 

Without the patch crrbv-20-fix-v2.patch, the new test in patch crrbv-20-test-case-v1.patch fails (based on the code submitted by Mike Fikes in an earlier comment). With that patch, it succeeds.

Andy Fingerhut
August 7, 2019, 3:26 PM

Attachment crrbv-20-cljs-fix-v1.patch is the ClojureScript version of the added test case and fix.

Andy Fingerhut
September 17, 2019, 11:04 PM

Added clj and cljs failing tests, and clj and cljs fixes, with this commit:

Will close this issue after build.clojure.org confirms that the tests pass there, as well as they already have on my machine.

Andy Fingerhut
September 17, 2019, 11:42 PM

All tests pass on build.clojure.org for the full test matrix. Closing this issue as resolved.

Fixed

Assignee

Andy Fingerhut

Reporter

Andy Fingerhut

Labels

None

Approval

None

Patch

None

Priority

Major
Configure