[
https://issues.apache.org/jira/browse/HADOOP-11887?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14572836#comment-14572836
]
Alan Burlison commented on HADOOP-11887:
----------------------------------------
I've pulled the ISA-L source down and built it on Solaris x86. It needs a bit
of Makefile tweaking but it builds as a LP64 library and 'make test' and 'make
perf' succeed. One caveat: on an older machine 'make test' dumps core with
SIGILL, presumably because the library isn't checking if all the instructions
it uses are actually available. This could potentially cause problems on Linux
& Windows as well, as it's a CPU feature rather than an OS one.
The machine that all the tests pass on has the following instruction set
extensions:
amd64: avx xsave pclmulqdq aes sse4.2 sse4.1 ssse3 popcnt tscp ahf cx16 sse3
sse2 sse fxsr mmx cmov amd_sysc cx8 tsc fpu efs f16c rdrand
i386: avx xsave pclmulqdq aes sse4.2 sse4.1 ssse3 popcnt tscp ahf cx16 sse3
sse2 sse fxsr mmx cmov sep cx8 tsc fpu efs f16c rdrand
The one that some of the tests fail on has these:
amd64: xsave sse4.1 ssse3 ahf cx16 sse3 sse2 sse fxsr mmx cmov amd_sysc cx8 tsc
fpu
i386: xsave sse4.1 ssse3 ahf cx16 sse3 sse2 sse fxsr mmx cmov sep cx8 tsc fpu
The failing tests are:
gf_vect_mul_avx_test
gf_vect_dot_prod_avx_test
> Introduce Intel ISA-L erasure coding library for the native support
> -------------------------------------------------------------------
>
> Key: HADOOP-11887
> URL: https://issues.apache.org/jira/browse/HADOOP-11887
> Project: Hadoop Common
> Issue Type: Sub-task
> Components: io
> Reporter: Kai Zheng
> Assignee: Kai Zheng
> Attachments: HADOOP-11887-v1.patch, HADOOP-11887-v2.patch,
> HADOOP-11887-v3.patch, HADOOP-11887-v4.patch
>
>
> This is to introduce Intel ISA-L erasure coding library for the native
> support, via dynamic loading mechanism (dynamic module, like *.so in *nix and
> *.dll on Windows).
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
