RFR: 8299748: java/util/zip/Deinflate.java failing on s390x [v4]
Jaikiran Pai
jpai at openjdk.org
Thu Apr 27 06:52:22 UTC 2023
On Thu, 27 Apr 2023 03:25:53 GMT, Amit Kumar <amitkumar at openjdk.org> wrote:
>> DeInflate.java test fails on s390x platform because size for out1 array which is responsible for storing the compressed data is insufficient. And being unable to write whole compressed data on array, on s390 whole data can't be recovered after compression. This PR updates the check method in the DeInflate test to no longer rely on pre-defined lengths/sizes to determine whether deflate followed by an inflate of data worked correctly. These sizes can vary depending on the underlying zlib implementations. The updated check method now uses a `ByteArrayOutputStream` to deflate into and then inflate from.
>>
>> Thanks to @jaikiran for amazing PR description.
>
> Amit Kumar has updated the pull request incrementally with one additional commit since the last revision:
>
> comments from Lance
test/jdk/java/util/zip/DeInflate.java line 124:
> 122: * This method checks if a given Deflater and Inflater pair can correctly compress and decompress data.
> 123: * checks were performed for this functionality using various input scenarios and ByteBuffer instances.
> 124: */
Would you consider updating this comment to something like:
/**
* Uses the {@code def} deflater to deflate the input data {@code in} of length {@code len}.
* A new {@link Inflater} is then created within this method to inflate the deflated data. The
* inflated data is then compared with the {@code in} to assert that it matches the original
* input data.
* This method repeats these checks for the different overloaded methods of
* {@code Deflater.deflate(...)} and {@code Inflater.inflate(...)}
*
* @param def the deflater to use for deflating the contents in {@code in}
* @param in the input content
* @param len the length of the input content to use
* @param nowrap will be passed to the constructor of the {@code Inflater} used in this
* method
* @throws Throwable if any error occurs during the check
*/
-------------
PR Review Comment: https://git.openjdk.org/jdk/pull/12283#discussion_r1178690586
More information about the core-libs-dev
mailing list