Legacy ImageMagick Discussions Archive

Nevermind... LCMS wasn't installed on the test server. Like I said, stupid

Hi everyone, This is probably a dumb question but bear with me... I'm upgrading some C software I wrote in which I make use of MagickWand. The version I'm upgrading from is 6.4.7 and I'm going to 7.0.7-11. Everything is going fine with the exception of colorspace conversion. In my code that I use in ...

Hi guys, I have an issue that's either caused by me being dumb (a strong possibility) or an issue in the latest IM (6.7.5-6). I have a small image conversion program that I wrote to help with some day-to-day tasks and I was previously using IM version 6.4.7 in which this code worked. Wanting to keep ...

There may be a memory leak in profile.c in the method 'GetProfilesFromResourceBlock' in ImageMagick 6.4.7-5. The variable 'profile' is declared and malloc'ed via the AquireStringInfo method and never freed. Here is the associated valgrind snippet: ==30403== 65,764 (32,864 direct, 32,900 indirect ...

This may or may not relate to my previous TIFF issue (http://www.imagemagick.org/discourse-server/viewtopic.php?f=3&t=12659) so in the event that it does, I apologize. File: http://wbd.wbdhost.com/test_gray.tiff When working with the above grayscale TIFF image (produced from an EPS using an ...

It seems that ImageMagick 6.4.7 crashes when it encounters problems in a TIFF file where it would previously just warn. An example file is: http://wbd.wbdhost.com/vn-0ixzcem.tif If I run 'identify -verbose vn-0ixzcern.tif' using ImageMagick 6.4.7-4 I get a segmentation fault while if I run it using ...

I believe there may be a memory leak in 6.4.7-4 in locale.c. It appears that on line 873 the variable xml is malloc'ed and never released. Here is a snippit from a valgrind report that I ran: ==16432== 49,381 bytes in 1 blocks are possibly lost in loss record 17 of 17 ==16432== at 0x4006AEE: malloc ...

Hi everyone, I have some code that I've written in C using the MagickWand API. The code does some basic operations on my input images, resizing, resampling, changing the colorspace if necessary, etc, etc. The problem that I have is that the output JPEG file seems to be corrupt in some manner. Linux ...

No problem... One of the images that produces this error is:

http://wbd.wbdtest.com/chicon-sm.eps

Ok, this one is driving me nuts and it happens to be the last leak that I've found (or at least the last one my code triggers). Valgrind reports: ==8778== 940 (220 direct, 720 indirect) bytes in 5 blocks are definitely lost in loss record 14 of 20 ==8778== at 0x4006AEE: malloc (vg_replace_malloc.c ...

In the WriteTIFFImage method, we call AcquireQuantumInfo within a do...while loop. At the end of the loop we call DestroyQuantumInfo which tears down the memory accordingly. The leak can occur because there are break statements within the do...while loop which force us out of the loop prior to the ...

On line 557 of xml-tree.c, root->processing_instructions is freed if root->processing_instructions[0] is not NULL. There is however a possibility that the wrapping IF statement is never entered due to the loop above the IF statement destroying root->processing_instructions[0]. for (i=0; root ...

In xml-tree.c on line 1483 memory is allocated for a string located in processing_instructions[0]. In the DestroyXMLTree method on line 548 we start the loop through the processing_instructions array at [1] instead of [0] so the previously allocated string never gets torn down.