A vulnerability exists in the QuickJS engines BigInt string conversion logic (js_bigint_to_string1) due to an incorrect calculation of the required number of digits, which in turn leads to reading memory past the allocated BigInt structure.
- The function determines the number of characters (n_digits) needed for the string representation by calculating:
$$ text{n_digits} = (text{n_bits} + text{log2_radix} - 1) / text{log2_radix}$$
$$$$This formula is off-by-one in certain edge cases when calculating the necessary memory limbs. For instance, a 127-bit BigInt using radix 32 (where $text{log2_radix}=5$) is calculated to need $text{n_digits}=26$.
-
The maximum number of bits actually stored is $text{n_bits}=127$, which requires only two 64-bit limbs ($text{JS_LIMB_BITS}=64$).
-
The conversion loop iterates $text{n_digits}=26$ times, attempting to read 5 bits in each iteration, totaling $26 times 5 = 130$ bits.
-
In the final iterations of the loop, the code attempts to read data that spans two limbs:
C
c = (r->tab[pos] » shift) | (r->tab[pos + 1] « (JS_LIMB_BITS - shift));
- Since the BigInt was only allocated two limbs, the read operation for r->tab[pos + 1] becomes an Out-of-Bounds Read when pos points to the last valid limb (e.g., $pos=1$).
This vulnerability allows an attacker to cause the engine to read and process data from the memory immediately following the BigInt buffer. This can lead to Information Disclosure of sensitive data stored on the heap adjacent to the BigInt object.
Weakness
The product reads data past the end, or before the beginning, of the intended buffer.
Potential Mitigations
- Assume all input is malicious. Use an “accept known good” input validation strategy, i.e., use a list of acceptable inputs that strictly conform to specifications. Reject any input that does not strictly conform to specifications, or transform it into something that does.
- When performing input validation, consider all potentially relevant properties, including length, type of input, the full range of acceptable values, missing or extra inputs, syntax, consistency across related fields, and conformance to business rules. As an example of business rule logic, “boat” may be syntactically valid because it only contains alphanumeric characters, but it is not valid if the input is only expected to contain colors such as “red” or “blue.”
- Do not rely exclusively on looking for malicious or malformed inputs. This is likely to miss at least one undesirable input, especially if the code’s environment changes. This can give attackers enough room to bypass the intended validation. However, denylists can be useful for detecting potential attacks or determining which inputs are so malformed that they should be rejected outright.
- To reduce the likelihood of introducing an out-of-bounds read, ensure that you validate and ensure correct calculations for any length argument, buffer size calculation, or offset. Be especially careful of relying on a sentinel (i.e. special character such as NUL) in untrusted inputs.
References