CVE-2026-99242

*** Pendiente de traducción *** In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> ntfs: fix race between fallocate and mmap reads<br /> <br /> The fallocate implementation only takes invalidate_lock for punch hole,<br /> collapse range, and insert range operations. For standard allocation modes<br /> (mode == 0, FALLOC_FL_KEEP_SIZE), the lock is not held.<br /> <br /> During ntfs_attr_fallocate(), new clusters are mapped to the runlist via<br /> ntfs_attr_map_cluster() before being zeroed by ntfs_dio_zero_range(). This<br /> creates a window where concurrent mmap page faults can read uninitialized<br /> disk data.<br /> <br /> Since mmap uses filemap_fault() which takes invalidate_lock in shared mode,<br /> it can fault in pages during this window and expose old disk contents to<br /> userspace. This is an information leak and data integrity issue.<br /> <br /> Fix by taking invalidate_lock for all fallocate operations, not just for<br /> punch/collapse/insert modes. This prevents concurrent page faults from<br /> accessing unzeroed clusters during the allocation window.