In QEMU's 9pfs (virtio-9p) implementation, the v9fs_readdir() function does not properly constrain the memory allocation size for Treaddir responses. A guest client can send a
Treaddir request with a large 'count' parameter, and while the negotiated 'msize' provides some limit, a malicious privileged guest could negotiate a large 'msize' but provide a
small reply buffer, causing QEMU to allocate host memory proportional to 'msize' before discovering the reply cannot fit. This can lead to excessive host memory allocation for
crafted large directories, resulting in the transport device being marked as "broken" (disconnecting guest communication) while the QEMU process continues running with a potentially
large unreleased host memory allocation, negatively impacting other services on the host. This vulnerability has existed since QEMU v5.2.0-rc0 and is fixed in v11.0.3 and v10.0.12.