[gs-bugs] [Bug 690346] many pcl files from the customer test file directory have problems.

bugs.ghostscript.com-bugzilla-daemon at ghostscript.com bugs.ghostscript.com-bugzilla-daemon at ghostscript.com
Thu Mar 19 22:10:53 PDT 2009


http://bugs.ghostscript.com/show_bug.cgi?id=690346





------- Additional Comments From ray.johnston at artifex.com  2009-03-19 22:10 -------
I am not sure the 'no output' cases are related to the poor performance of
39002wts.pcl

This file (and probably many others) are entering clist mode because the
cluster regression has -dMaxBitmap=30000000 instead of -dMaxBitmap=300000000
(30 million instead of 300 million).

When this happens, because there is a severe memory leak that I need to
track down. It is somewhat worse with the "ialloc" memory manager than with
the "chunk" memory manager because chunks never get freed with the ialloc
manager and memory usage grows faster.

Note that the PCL build defaults to BAND_LIST_STORAGE=memory. Using 'file',
and clist mode this seems to be about the same as with BAND_LIST_STORAGE
= memory, so this is probably not related to the 'memfile' logic.

With page buffer mode, either memory manager completes this job in about
29 seconds (DEBUG build) on my laptop and used about 105Mb. I didn't let the
ialloc case finish, but the chunk case took 3160 seconds and used over 896 Mb.

I think that I can improve the chunk manager case a bit by skipping the
search of chunks when the size requested is not available in any chunk
(a largest_free value for the chain of chunks), but the memory leak when
clist is used is more serious.



------- You are receiving this mail because: -------
You are the QA contact for the bug, or are watching the QA contact.



More information about the gs-bugs mailing list