This command line gives a segfault: cat out.pdf | RIP_MAX_CACHE=128m gs -dNOPAUSE -sstdout=%stderr -sOutputFile=%stdout -sDEVICE=cups -r600x600 -dcupsBitsPerColor=8 -dcupsColorOrder=0 -dcupsColorSpace=4 -_ > out.raster If I set 300 dpi resolution cat out.pdf | RIP_MAX_CACHE=128m gs -dNOPAUSE -sstdout=%stderr -sOutputFile=%stdout -sDEVICE=cups -r300x300 -dcupsBitsPerColor=8 -dcupsColorOrder=0 -dcupsColorSpace=4 -_ > out.raster I do not get a segfault but the output is scrambled (use Mike Sweet's rasterview to display). If CUPS calls ghostscript it sets the RIP_MAX_CACHE=128m environment variable. If I do not set this variable, I get Ghostscript also not crashing with 600 dpi but the output is still scrambled: cat out.pdf | gs -dNOPAUSE -sstdout=%stderr -sOutputFile=%stdout -sDEVICE=cups -r600x600 -dcupsBitsPerColor=8 -dcupsColorOrder=0 -dcupsColorSpace=4 -_ > out.raster Selecting other cupsColorSpace values there is no problem (like 1 for RGB, 3, for Gray, 18, 19, 20 for SGRay, SRGB, and AdobeRGB). So there are two bugs: CMY output is scrambled, and the segfault with the first command line. Sample input file, out.pdf is attached.
Created attachment 18870 [details] out.pdf
Sorry, CMY problem was already treated and fixed in bug 702133. I am rebuilding now and re-testing the segfault problem.
The fix of bug 701625 also fixes the segfault. So all working now after applying the patch. Closing. *** This bug has been marked as a duplicate of bug 701625 ***