Re: Issue with GM through JNI

On Mon, 10 Jul 2017, Tianyu Lang wrote:
>
> Here is my locale setting:
> -------------------------------------------------------------------------------------
> tianyulang@dev-tianyulang:~/jni-test-inputs$ locale -a
> C
> C.UTF-8
> en_US.utf8
> POSIX
> -------------------------------------------------------------------------------------

Should be good.

> I can play with piping to see if it is gonna reduce the latency.
> I tried the Ram Disk solution unfortunately there is no speed boost.
>
> Could the lack of some signal handlers cause problems like heap corruption?

No.

> I guess one thing I can try is to see if my ImageMagick JNI solution shows
> the same number of handlers registered (if they have similar logging
> settings). The problem puzzles me because apparently neither using GM in a
> C/C++ program nor GM CLI tool has memory corruption issue.........

By post the output of 'gm -version' I was wanting the output (e.g. 
version, list of what is supported, and the compilation options used) 
that it prints, not the debug logs when it is run with the 
MAGICK_DEBUG variable set.

The most common performance problem when calling gm from a huge 
program like Java is that Java is slow to execute external programs. 
For this, you should investigate 'gm batch' which allows executing 
multiple commands in a single child-process invocation.  The 
performance improvement can be quite large.

Please read this article written by the author of 'gm batch':

http://kennethxu.blogspot.com/2013/02/integrate-java-and-graphicsmagick.html

Bob
-- 
Bob Friesenhahn
[email protected], http://www.simplesystems.org/users/bfriesen/
GraphicsMagick Maintainer,    http://www.GraphicsMagick.org/

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot