Join GitHub today
GitHub is home to over 36 million developers working together to host and review code, manage projects, and build software together.
Sign upbenchmark: fix buffer-base64-decode.js #27260
Conversation
This comment has been minimized.
This comment has been minimized.
nodejs-github-bot
added
benchmark
buffer
labels
Apr 16, 2019
Trott
requested a review
from
BridgeAR
Apr 16, 2019
Trott
changed the title
buffer: fix buffer-base64--decode.js
buffer: fix buffer-base64-decode.js
Apr 16, 2019
Trott
force-pushed the
Trott:fix-buffer-benchmark
branch
from
46d2dfc
to
3db71ca
Apr 16, 2019
This comment has been minimized.
This comment has been minimized.
|
Still need a fix for |
addaleax
approved these changes
Apr 16, 2019
jasnell
approved these changes
Apr 16, 2019
This comment has been minimized.
This comment has been minimized.
|
I think this should be using the |
This comment has been minimized.
This comment has been minimized.
Trott
added some commits
Apr 16, 2019
Trott
force-pushed the
Trott:fix-buffer-benchmark
branch
from
2d01e7d
to
eab7d50
Apr 16, 2019
This comment has been minimized.
This comment has been minimized.
Fixed and force-pushed. |
Trott
changed the title
buffer: fix buffer-base64-decode.js
benchmark: fix buffer-base64-decode.js
Apr 16, 2019
BridgeAR
reviewed
Apr 16, 2019
This comment has been minimized.
This comment has been minimized.
Please add an assertion that the benchmark fails in case faulty |
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
I'm not sure I understand. The benchmark test was failing. Adding an assertion for it wouldn't have detected it because it was already a failing test. (The problem is that the test is not run in CI except nightly, which is where i saw the failure.) The benchmark tests are intentionally minimal. I'm reluctant to test buffer functionality in them. |
This comment has been minimized.
This comment has been minimized.
The test would have failed earlier that way and we have some very basic assertions in some benchmarks to validate the input parameters. Adding one here as well just seemed right to me but that's not a blocker (my other comment is). |
This comment has been minimized.
This comment has been minimized.
BridgeAR
approved these changes
Apr 16, 2019
lpinca
approved these changes
Apr 16, 2019
This comment has been minimized.
This comment has been minimized.
addaleax
approved these changes
Apr 16, 2019
Trott
added
the
author ready
label
Apr 16, 2019
This comment has been minimized.
This comment has been minimized.
|
I'd like to fast-track this to unbreak the nightly CI that runs the benchmark tests. |
Trott
added
the
fast-track
label
Apr 17, 2019
refack
approved these changes
Apr 18, 2019
Trott
removed
the
fast-track
label
Apr 18, 2019
Trott
closed this
Apr 18, 2019
Trott
added a commit
to Trott/io.js
that referenced
this pull request
Apr 18, 2019
Trott
added a commit
to Trott/io.js
that referenced
this pull request
Apr 18, 2019
This comment has been minimized.
This comment has been minimized.
|
Landed in f98679f...d5bb500 |
Trott commentedApr 16, 2019
693401d added stricter range checking
for buffer operations and that apparently seems to have uncovered the
fact that one of our benchmarks was overflowing a buffer. Increase the
buffer size so the benchmark doesn't throw an error anymore.
Checklist
make -j4 test(UNIX), orvcbuild test(Windows) passes