[
https://issues.apache.org/jira/browse/HDFS-9448?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Bob Hansen updated HDFS-9448:
-----------------------------
Attachment: HDFS-9448.HDFS-8707.003.patch
[~aw] - Thanks for your feedback.
{quote}
a) by default, it should definitely compile and run all the (applicable) unit
test if valgrind isn't around.
b) if -Drequire.valgrind is set to true, then the build should fail if valgrind
isn't present. This allows for REs to make sure the build environment is in a
correct and expected state.
{quote}
Sounds great. This patch reflects those changes.
bq. This patch needs to add valgrind to the dockerfile since this patch is the
one that is introducing the dependency. I'm pretty much -1 on any other outcome
but that one
I'm glad to include the Dockerfile changes; I have been _strongly_ encouraged
in the past to trim changes down into small discrete chunks, and was erring on
the side of simplicity here.
My inclination is to retain a flag to skip valgrind tests even if valgrind is
present, as valgrind tests tend to be ~100x longer, and I can see that starting
to add up for interactive development as we add more tests to the library.
It's implemented on the CMake side with SKIP_VALGRIND (which can be set via mvn
native_cmake_args property), but is not exposed as a mvn property of its own.
Does that seem like a good approach to you?
> Enable valgrind for libhdfspp unit tests
> ----------------------------------------
>
> Key: HDFS-9448
> URL: https://issues.apache.org/jira/browse/HDFS-9448
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: hdfs-client
> Reporter: Bob Hansen
> Assignee: Bob Hansen
> Attachments: HDFS-9448.HDFS-8707.000.patch,
> HDFS-9448.HDFS-8707.001.patch, HDFS-9448.HDFS-8707.002.patch,
> HDFS-9448.HDFS-8707.003.patch
>
>
> We should have a target that runs the unit tests under valgrind if it is
> available on the target machine.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
