CVE-2019-16214

Libra Core before 2019-09-03 has an erroneous regular expression for inline comments, which makes it easier for attackers to interfere with code auditing by using a nonstandard line-break character for a comment. For example, a Move module author can enter the // sequence (which introduces a single-line comment), followed by very brief comment text, the \r character, and code that has security-critical functionality. In many popular environments, this code is displayed on a separate line, and thus a reader may infer that the code is executed. However, the code is NOT executed, because language/compiler/ir_to_bytecode/src/parser.rs allows the comment to continue after the \r character.

Libra Core before 2019-09-03 has an erroneous regular expression for inline comments, which makes it easier for attackers to interfere with code auditing by using a nonstandard line-break character for a comment. For example, a Move module author can enter the // sequence (which introduces a single-line comment), followed by very brief comment text, the \r character, and code that has security-critical functionality. In many popular environments, this code is displayed on a separate line, and thus a reader may infer that the code is executed. However, the code is NOT executed, because language/compiler/ir_to_bytecode/src/parser.rs allows the comment to continue after the \r character.

Libra Core before 2019-09-03 has an erroneous regular expression for inline comments, which makes it easier for attackers to interfere with code auditing by using a nonstandard line-break character for a comment. For example, a Move module author can enter the // sequence (which introduces a single-line comment), followed by very brief comment text, the \r character, and code that has security-critical functionality. In many popular environments, this code is displayed on a separate line, and thus a reader may infer that the code is executed. However, the code is NOT executed, because language/compiler/ir_to_bytecode/src/parser.rs allows the comment to continue after the \r character.

Libra Core before 2019-09-03 has an erroneous regular expression for inline comments, which makes it easier for attackers to interfere with code auditing by using a nonstandard line-break character for a comment. For example, a Move module author can enter the // sequence (which introduces a single-line comment), followed by very brief comment text, the \r character, and code that has security-critical functionality. In many popular environments, this code is displayed on a separate line, and thus a reader may infer that the code is executed. However, the code is NOT executed, because language/compiler/ir_to_bytecode/src/parser.rs allows the comment to continue after the \r character.

Libra Core before 2019-09-03 has an erroneous regular expression for inline comments, which makes it easier for attackers to interfere with code auditing by using a nonstandard line-break character for a comment. For example, a Move module author can enter the // sequence (which introduces a single-line comment), followed by very brief comment text, the \r character, and code that has security-critical functionality. In many popular environments, this code is displayed on a separate line, and thus a reader may infer that the code is executed. However, the code is NOT executed, because language/compiler/ir_to_bytecode/src/parser.rs allows the comment to continue after the \r character.