Hi Andrew,
Fix looks good.
Jennifer
On 6/13/2012 1:09 PM, Phil Race wrote:
Andrew,
Seems fine. I don't know if we ever give up any measurable performance
but we need to be correct. Is fix this verified by the SQE test suite ?
If so add "noreg-sqe" to the bug, else try for a regression test if
you can please.
Make sure you get a second reviewer.
-phil.
On 6/9/2012 1:50 AM, Andrew Brygin wrote:
Hello,
could you please review a fix for 7153693?
This problem is caused by incorrect usage of memcpy routine in
sun_jpeg_fill_suspended_buffer. The linux man for memcpy says:
"The memory areas must not overlap. Use memmove(3) if the memory
areas do overlap."
This is exactly the case of the sun_jpeg_fill_suspended_buffer, were
we move overlapped memory areas under some conditions. Some linux
systems are tolerant to the problem, but OEL 6.* Unbreakable Kernel
x64 is not, so the problem is visible there, and manifests
in occasional failures during decoding of progressive jpegs.
It is the root cause of observed test failures.
Suggested fix is just use memmove instead of memcpy.
Please take a look.
Bug: http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=7153693
Webrev: http://cr.openjdk.java.net/~bae/7153693/8/webrev.00/
Thanks,
Andrew
