You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This was filed in Fedora against R-jsonlite because of the marked bundling, but originates in yajl. Thus it may not affect jsonlite directly if you don't call that functionality. The CVE information indicates:
The reallocation logic at yajl_buf.c#L64 may result in the need 32bit integer wrapping to 0 when need approaches a value of 0x80000000 (i.e. ~2GB of data), which results in a reallocation of buf->alloc into a small heap chunk.
These integers are declared as size_t in the 2.x branch of yajl, which practically prevents the issue from triggering on 64bit platforms, however this does not preclude this issue triggering on 32bit builds on which size_t is a 32bit integer.
Subsequent population of this under-allocated heap chunk is based on the original buffer size, leading to heap memory corruption.
This was filed in Fedora against R-jsonlite because of the marked bundling, but originates in yajl. Thus it may not affect jsonlite directly if you don't call that functionality. The CVE information indicates:
See also the upstream issue at lloyd/yajl#239 and WIP PR at lloyd/yajl#240, as well as the adjacent fix at brianmario/yajl-ruby#211.
The text was updated successfully, but these errors were encountered: