Bug 692129

Summary: Seg. fault with -dDOINTERPOLATE
Product: Ghostscript Reporter: Marcos H. Woehrmann <marcos.woehrmann>
Component: GeneralAssignee: Robin Watts <robin.watts>
Status: NOTIFIED FIXED    
Severity: normal    
Priority: P2    
Version: master   
Hardware: PC   
OS: All   
Customer: 243 Word Size: ---

Description Marcos H. Woehrmann 2011-04-07 04:05:54 UTC
The customer reports an error when processing the attached with with 9.x, I'm not able to reproduce the problem, but Ken is.  Here are the options the customer is using:

-q
-dNOPAUSE
-dSAFER
-dBATCH
-dUseCropBox
-dConvertCMYKImagesToRGB=true
-dDOINTERPOLATE
-dTextAlphaBits=4
-dGraphicsAlphaBits=1
-dProcessColorModel=/DeviceRGB
-sDEVICE=png16m
-r100
-sstdout=%stderr
-sOutputFile=-
-dPrinted=false

and Ken's analysis:

For me the file seg faults on page 4 when -dDOINTERPOLATE is true, its in gsalloc.c, i_free_object(), line 779:

   finalize = pstype->finalize;

pstype is 0x00 which doesn't really look good.

Its called from s_IScale_release() :

   gs_free_object(mem, ss->dst, "image_scale dst");

It looks like something has written 0s all over ss->dst. One for Robin ?
Comment 2 Robin Watts 2011-04-08 13:35:47 UTC
I have a simple fix for this, committed as revision 12378.