Type confusion in string addition in QuickJS

An integer overflow vulnerability exists in the QuickJS regular expression engine (libregexp) due to an inconsistent representation of the bytecode buffer size.

• The regular expression bytecode is stored in a DynBuf structure, which correctly uses a $\text{size}\_\text{t}$ (an unsigned type, typically 64-bit) for its size member.

• However, several functions, such as re_emit_op_u32 and other internal parsing routines, incorrectly cast or store this DynBuf $\text{size}\_\text{t}$ value into a signed int (typically 32-bit).

• When a large or complex regular expression (such as those generated by a recursive pattern in a Proof-of-Concept) causes the bytecode size to exceed $2^{31}$ bytes (the maximum positive value for a signed 32-bit integer), the size value wraps around, resulting in a negative integer when stored in the int variable (Integer Overflow).

• This negative value is subsequently used in offset calculations. For example, within functions like re_parse_disjunction, the negative size is used to compute an offset (pos) for patching a jump instruction.

• This negative offset is then incorrectly added to the buffer pointer (s->byte\_code.buf + pos), leading to an out-of-bounds write on the first line of the snippet below:

put_u32(s->byte_code.buf + pos, len);