[
https://issues.apache.org/jira/browse/HBASE-14940?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15048132#comment-15048132
]
stack commented on HBASE-14940:
-------------------------------
bq, if (UnsafeAccess.unaligned() == false) {
The above readst wrong but my guess is that the author was trying to make a
point given the odd name of the method.
In unsafeCopy if len > UNSAFE_COPY_THRESHOLD we will call it again to finish up
the copy?
[~ikeda] We don't want the range checking because we think we know where we are
(famous last words) and it costs. Hopefully we won't be spending too much time
debugging corrupt memory reads...
> Make our unsafe based ops more safe
> -----------------------------------
>
> Key: HBASE-14940
> URL: https://issues.apache.org/jira/browse/HBASE-14940
> Project: HBase
> Issue Type: Bug
> Reporter: Anoop Sam John
> Assignee: Anoop Sam John
> Attachments: HBASE-14940.patch
>
>
> Thanks for the nice findings [~ikeda]
> This jira solves 3 issues with Unsafe operations and ByteBufferUtils
> 1. We can do sun unsafe based reads and writes iff unsafe package is
> available and underlying platform is having unaligned-access capability. But
> we were missing the second check
> 2. Java NIO is doing a chunk based copy while doing Unsafe copyMemory. The
> max chunk size is 1 MB. This is done for "A limit is imposed to allow for
> safepoint polling during a large copy" as mentioned in comments in Bits.java.
> We are also going to do same way
> 3. In ByteBufferUtils, when Unsafe is not available and ByteBuffers are off
> heap, we were doing byte by byte operation (read/copy). We can avoid this and
> do better way.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
