CVE-2025-38480

In the Linux kernel, the following vulnerability has been resolved:<br /> <br /> comedi: Fix use of uninitialized data in insn_rw_emulate_bits()<br /> <br /> For Comedi `INSN_READ` and `INSN_WRITE` instructions on "digital"<br /> subdevices (subdevice types `COMEDI_SUBD_DI`, `COMEDI_SUBD_DO`, and<br /> `COMEDI_SUBD_DIO`), it is common for the subdevice driver not to have<br /> `insn_read` and `insn_write` handler functions, but to have an<br /> `insn_bits` handler function for handling Comedi `INSN_BITS`<br /> instructions. In that case, the subdevice&amp;#39;s `insn_read` and/or<br /> `insn_write` function handler pointers are set to point to the<br /> `insn_rw_emulate_bits()` function by `__comedi_device_postconfig()`.<br /> <br /> For `INSN_WRITE`, `insn_rw_emulate_bits()` currently assumes that the<br /> supplied `data[0]` value is a valid copy from user memory. It will at<br /> least exist because `do_insnlist_ioctl()` and `do_insn_ioctl()` in<br /> "comedi_fops.c" ensure at lease `MIN_SAMPLES` (16) elements are<br /> allocated. However, if `insn-&gt;n` is 0 (which is allowable for<br /> `INSN_READ` and `INSN_WRITE` instructions, then `data[0]` may contain<br /> uninitialized data, and certainly contains invalid data, possibly from a<br /> different instruction in the array of instructions handled by<br /> `do_insnlist_ioctl()`. This will result in an incorrect value being<br /> written to the digital output channel (or to the digital input/output<br /> channel if configured as an output), and may be reflected in the<br /> internal saved state of the channel.<br /> <br /> Fix it by returning 0 early if `insn-&gt;n` is 0, before reaching the code<br /> that accesses `data[0]`. Previously, the function always returned 1 on<br /> success, but it is supposed to be the number of data samples actually<br /> read or written up to `insn-&gt;n`, which is 0 in this case.