[HADOOP-11466] FastByteComparisons: do not use UNSAFE_COMPARER on the SPARC architecture because it is slower there

• Type: Improvement
• Status: Closed
• Priority: Minor
• Resolution: Fixed
• Fix Version/s: 2.6.1, 3.0.0-alpha1
• Component/s: io, performance, util
• Labels: patch

Environment

Linux X86 and Solaris SPARC

• Target Version/s: 2.6.0
• Assignee: Suman Somasundar
• Reporter: Suman Somasundar
• Created: 07/Jan/15 19:31
• Updated: 30/Aug/16 01:31
• Resolved: 22/Jan/15 01:06

Description

One difference between Hadoop 2.x and Hadoop 1.x is a utility to compare two byte arrays at coarser 8-byte granularity instead of at the byte-level. The discussion at HADOOP-7761 says this fast byte comparison is somewhat faster for longer arrays and somewhat slower for smaller arrays ( AVRO-939). In order to do 8-byte reads on addresses not aligned to 8-byte boundaries, the patch uses Unsafe.getLong. The problem is that this call is incredibly expensive on SPARC. The reason is that the Studio compiler detects an unaligned pointer read and handles this read in software. x86 supports unaligned reads, so there is no penalty for this call on x86.

Issue Links

• Blocked: HADOOP-12941 abort in Unsafe_GetLong when running IA64 HPUX 64bit mode