[
https://issues.apache.org/jira/browse/CASSANDRA-20190?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17930810#comment-17930810
]
Dmitry Konstantinov edited comment on CASSANDRA-20190 at 2/26/25 8:36 PM:
--------------------------------------------------------------------------
[https://github.com/apache/cassandra/pull/3936]
I have implemented a draft version of changes related to {{{}MemoryUtil{}}}: I
have introduced {{NativeEndianMemoryUtil}} and moved adjusted methods from
{{MemoryUtil}} related to a particular endian (like get/setInt).
The methods from MemoryUtil class are used only in offheap memtable entities
(NativeCell/NativeClustering/NativeDecoratedKey) which are not going to
disk/network without decoding to Java primitives, so I switched the logic to
native endian to avoid unnecessary conversions. Values stored in the entities
are still exposed as BigEndian byte buffers.
Also I dropped unused {{getHollowByteBuffer()}} method and related constants.
By some reason setShort did not support aligned architecture case, I have added
the correspondent logic. {{getShort}} is related to {{getUnsignedShort}} to
reflect the reality.
I have not touched {{Memory}} class yet.
was (Author: dnk):
[https://github.com/apache/cassandra/pull/3936]
I have implemented a draft version of changes related to {{{}MemoryUtil{}}}: I
have introduced {{NativeEndianMemoryUtil}} and moved adjusted methods from
{{MemoryUtil}} related to a particular endian (like get/setInt).
The methods from MemoryUtil class are used only in offheap memtable entities
(NativeCell/NativeClustering/NativeDecoratedKey) which are not going to
disk/network without decoding to Java primitives, so I switched the logic to
native endian to avoid unnecessary conversions. Values stored in the entities
are still exposed as BigEndian byte buffers.
Also I dropped unused {{getHollowByteBuffer()}} method and related constants.
By some reason setShort did not support aligned architecture case, I have added
the correspondent logic.
I have not touched {{Memory}} class yet.
> MemoryUtil.setInt/getInt and similar use the wrong endianness
> -------------------------------------------------------------
>
> Key: CASSANDRA-20190
> URL: https://issues.apache.org/jira/browse/CASSANDRA-20190
> Project: Apache Cassandra
> Issue Type: Bug
> Components: Local/Other
> Reporter: Branimir Lambov
> Assignee: Dmitry Konstantinov
> Priority: Normal
> Time Spent: 10m
> Remaining Estimate: 0h
>
> `NativeCell`, `NativeClustering` and `NativeDecoratedKey` use the above
> methods from `MemoryUtil` to write and read data from native memory. As far
> as I can see they are meant to write data in big endian. They do not (they
> always correct to little endian).
> Moreover, they disagree with their `ByByte` versions on big-endian machines
> (which is only likely an issue on aligned-access architectures (x86 and arm
> should be fine)).
> The same is true for the methods in `Memory`, used by compression metadata as
> well as index summaries.
> We need to verify that this does not cause any problems, and to change the
> methods to behave as expected and document the behaviour by explicitly using
> `ByteOrder.LITTLE_ENDIAN` for any data that may have been persisted on disk
> with the wrong endianness.
> The current MemoryUtil behaviour:
> ||Native
> order||MemoryUtil.setX||MemoryUtil.setXByByte||MemoryUtil.getX||MemoryUtil.getXByByte||
> |BE|LE|BE|LE|BE|
> |LE|LE|LE|LE|LE|
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
