CVE-2024-50222

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> iov_iter: fix copy_page_from_iter_atomic() if KMAP_LOCAL_FORCE_MAP<br /> <br /> generic/077 on x86_32 CONFIG_DEBUG_KMAP_LOCAL_FORCE_MAP=y with highmem,<br /> on huge=always tmpfs, issues a warning and then hangs (interruptibly):<br /> <br /> WARNING: CPU: 5 PID: 3517 at mm/highmem.c:622 kunmap_local_indexed+0x62/0xc9<br /> CPU: 5 UID: 0 PID: 3517 Comm: cp Not tainted 6.12.0-rc4 #2<br /> ...<br /> copy_page_from_iter_atomic+0xa6/0x5ec<br /> generic_perform_write+0xf6/0x1b4<br /> shmem_file_write_iter+0x54/0x67<br /> <br /> Fix copy_page_from_iter_atomic() by limiting it in that case<br /> (include/linux/skbuff.h skb_frag_must_loop() does similar).<br /> <br /> But going forward, perhaps CONFIG_DEBUG_KMAP_LOCAL_FORCE_MAP is too<br /> surprising, has outlived its usefulness, and should just be removed?