In the Linux kernel, the following vulnerability has been resolved:
scsi: scsidebug: Fix REPORT ZONES alloclen underflow OOB write
respreportzones() sizes the reply buffer from the CDB allocation length. The v3 fix rounds alloc_len up with ALIGN() before deriving the descriptor count:
rep_max_zones = (ALIGN((u64)alloc_len, RZONES_DESC_HD) -
RZONES_DESC_HD) >> ilog2(RZONES_DESC_HD);
arr_len = (u64)RZONES_DESC_HD * (rep_max_zones + 1);
For alloclen in 0xFFFFFFC1..0xFFFFFFFF, ALIGN() rounds up to 0x100000000, so arrlen is 4 GB. On 32-bit, kzalloc()'s sizet is 32-bit and truncates 0x100000000 to 0; kzalloc(0) returns ZEROSIZE_PTR, which passes the !arr check, and desc = arr + 64 is then dereferenced in the loop -> out-of-bounds write / panic.
Clamp repmaxzones to devip->nrzones. The loop already stops at sdebugcapacity (after nrzones zones), so a report can never hold more than nrzones descriptors; the clamp does not change the report, it only bounds arrlen to (nrzones + 1) * RZONESDESCHD, a real device property that can never reach 0x100000000.