Bug Classes/Integer overflow

From Linux Kernel Security Subsystem
Revision as of 21:47, 9 March 2016 by KeesCook (talk | contribs) (→‎Examples)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

Details

Integer overflows (or underflows) occur when a multiplication happens that exceeds the size that can be represented by the datatype, generally wrapping around. This usually results in either writing to too-small buffers, or producing out of bound array indexes. Exploitation is most common via heap overflows, since the (too-small) buffers tend to be allocated on the heap. Additionally, reference counting can overflow and wrap around, leading to use-after-free exploits.

Examples

Mitigations

  • check for refcount overflows (e.g. PAX_REFCOUNT)
  • compiler instrumentation to detect multiplication overflows at runtime (e.g. PAX_SIZE_OVERFLOW)