[HADOOP-8926] hadoop.util.PureJavaCrc32 cache hit-ratio is low for static data

• Type: Improvement
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 2.0.3-alpha
• Fix Version/s: 2.0.3-alpha, 0.23.5
• Component/s: util
• Labels: optimization

Environment

Ubuntu 10.10 i386

• Release Note: Speed up Crc32 by improving the cache hit-ratio of hadoop.util.PureJavaCrc32
• Assignee: Gopal Vijayaraghavan
• Reporter: Gopal Vijayaraghavan
• Created: 13/Oct/12 09:02
• Updated: 03/Sep/14 23:19
• Resolved: 16/Oct/12 21:56

Description

While running microbenchmarks for HDFS write codepath, a significant part of the CPU fraction was consumed by the DataChecksum.update().

The attached patch converts the static arrays in CRC32 into a single linear array for a performance boost in the inner loop.

milli-seconds for 1Gig (16400 loop over a 64kb chunk)

platform	original	cache-aware	improvement
x86	3894	2304	40.83
x86_64	2131	1826	14

The performance improvement on x86 is rather larger than the 64bit case, due to the extra register/stack pressure caused by the static arrays.

A closer analysis of the PureJavaCrc32 JIT code shows the following assembly fragment

  0x40f1e345: mov    $0x184,%ecx
  0x40f1e34a: mov    0x4415b560(%ecx),%ecx  ;*getstatic T8_5
                                        ; - PureJavaCrc32::update@95 (line 61)
                                        ;   {oop('PureJavaCrc32')}
  0x40f1e350: mov    %ecx,0x2c(%esp)

Basically, the static variables T8_0 through to T8_7 are being spilled to the stack because of register pressure. The x86_64 case has a lower likelihood of such pessimistic JIT code due to the increased number of registers.

Issue Links

• is related to: HADOOP-8971 Backport: hadoop.util.PureJavaCrc32 cache hit-ratio is low for static data (HADOOP-8926)