ImageMagick before 7.1.2-35 contains an information disclosure vulnerability in the RLA decoder that returns images containing uninitialized data. Attackers can supply a crafted RLA image to cause the decoded output to include uninitialized memory, potentially disclosing heap information.