Kovid Goyal

#8119of 57,305
36.9Total CVSS
Vulnerabilities · 6
Medium
5
Critical
1
PT-2026-98839
4.6
2026-09-25
Kitty · Kitty · CVE-2026-80430
**Name of the Vulnerable Software and Affected Versions** kitty versions 0.47.0 through 0.48.x **Description** Improper link resolution in the drag source staging path of the drag and drop protocol allows a program writing to the terminal to create files and directories outside the staging directory. This occurs because the `subdir data for drag()` function in `kitty/dnd.c` resolves descendants of the staged item tree by constructing a path string and opening it with `safe open(path, O DIRECTORY | O RDONLY, 0)` instead of walking the tree component by component. An attacker can declare two entries with the same name—first a symlink targeting an arbitrary absolute path and second a directory—causing `mkdirat()` to fail with EEXIST. The code ignores this failure, leading the subsequent path resolution to follow the symlink and return a directory descriptor outside the staging directory. This descriptor is then passed as the `dirfd` argument to `add payload()` and used for all subsequent create operations. While entry names are sanitized, symlink targets are not validated. Files are created with O CREAT | O WRONLY | O EXCL at mode 0644, preventing the overwriting of existing files, but directories are created with `mkdirat()` at mode 0755, allowing the creation of new intermediate directories at any path writable by the user running the application, provided the symlink target is an existing directory. **Recommendations** Update kitty to version 0.49.0 or later.
PT-2026-98931
4.6
2026-09-25
Kovid Goyal · Kitty · CVE-2026-95834
Use After Free in the drag source path of the drag and drop protocol in kitty from 0.47.0 before 0.49.0 allows a program writing to the terminal to cause the terminal to read from and write to freed heap memory, because drag remote file data() in kitty/dnd.c holds a DragRemoteItem pointer into an array it does not own, calls toplevel data for drag() or subdir data for drag(), and then continues to use that pointer. Those helpers, and add payload() and populate dir entries() which they call, report errors through the abrt() macro, which expands to cancel drag() followed by a plain return, and cancel drag() calls drag free offer(), which frees the array the pointer refers to. The helpers return void, so the caller receives no indication that the teardown happened, and proceeds to call all children complete() on the freed pointer, which dereferences it, and then to write through it. That dereference is guarded by a local flag that is set when the request carries no payload and announces no further data, which is the same condition that selects the finalisation block of add payload(), so the error paths in that block reach it: a create that fails because an entry of the same name already exists, because the client may declare two entries with one name and the create operations use O CREAT with O EXCL and symlinkat(), a mkdirat() failure other than EEXIST, and the directory entry allocation paths. Both branches reach it. In the top level branch the caller's pointer is never cleared, so clearing the owning structure's own pointers during teardown does not help. In the sub directory branch subdir data for drag() sets the caller's pointer to NULL on entry and assigns it only after its own last error path, so its own aborts leave the caller with NULL and are stopped by a null check, but it then calls add payload() with that pointer set, and an abort there leaves the caller holding a freed child node inside the item tree, which drag free offer() frees by recursion. This results in undefined behaviour in the terminal process, reachable from the byte stream of any program running in the window.