8000 fix Size computation for allocation may overflow by odaysec · Pull Request #381 · uber/aresdb · GitHub
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content

fix Size computation for allocation may overflow #381

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 1 commit into
base: master
Choose a base branch
from

Conversation

odaysec
Copy link
@odaysec odaysec commented Apr 15, 2025

fix the problem need to ensure that the size computation does not overflow. This can be achieved by validating the size of hll.SparseData before performing the arithmetic operation. Specifically, we should check if len(hll.SparseData) is within a safe range that guarantees no overflow will occur when multiplied by recordValueBytes.

Performing calculations involving the size of potentially large strings or slices can result in an overflow (for signed integer types) or a wraparound (for unsigned types). An overflow causes the result of the calculation to become negative, while a wraparound results in a small (positive) number. This can cause further issues. If, the result is then used in an allocation, it will cause a runtime panic if it is negative, and allocate an unexpectedly small buffer otherwise.

POC

In the following example, assume that there is a function encryptBuffer that encrypts byte slices whose length must be padded to be a multiple of 16. The function encryptValue provides a convenience wrapper around this function: when passed an arbitrary value, it first encodes that value as JSON, pads the resulting byte slice, and then passes it to encryptBuffer.

package main

import "encoding/json"

func encryptValue(v interface{}) ([]byte, error) {
	jsonData, err := json.Marshal(v)
	if err != nil {
		return nil, err
	}
	size := len(jsonData) + (len(jsonData) % 16)
	buffer := make([]byte, size)
	copy(buffer, jsonData)
	return encryptBuffer(buffer)
}

When passed a value whose JSON encoding is close to the maximum value of type int in length, the computation of size will overflow, producing a negative value. When that negative value is passed to make, a runtime panic will occur.

To guard against this, the function should be improved to check the length of the JSON-encoded value. For example, here is a version of encryptValue that ensures the value is no larger than 64 MB, which fits comfortably within an int and avoids the overflow:

package main

import (
	"encoding/json"
	"errors"
)

func encryptValueGood(v interface{}) ([]byte, error) {
	jsonData, err := json.Marshal(v)
	if err != nil {
		return nil, err
	}
	if len(jsonData) > 64*1024*1024 {
		return nil, errors.New("value too large")
	}
	size := len(jsonData) + (len(jsonData) % 16)
	buffer := make([]byte, size)
	copy(buffer, jsonData)
	return encryptBuffer(buffer)
}

References

Integer overflow
Making slices, maps and channels
CWE-190

@CLAassistant
Copy link
CLAassistant commented Apr 15, 2025

CLA assistant check
All committers have signed the CLA.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants
0