[PATCH 5.10.y 5/5] scsi: scsi_debug: Fix REPORT ZONES alloc_len underflow OOB write

From: Ibrahim Hashimov <[email protected]>

resp_report_zones() 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 alloc_len in 0xFFFFFFC1..0xFFFFFFFF, ALIGN() rounds up to
0x100000000, so arr_len is 4 GB. On 32-bit, kzalloc()'s size_t is 32-bit
and truncates 0x100000000 to 0; kzalloc(0) returns ZERO_SIZE_PTR, which
passes the !arr check, and desc = arr + 64 is then dereferenced in the
loop -> out-of-bounds write / panic.

Clamp rep_max_zones to devip->nr_zones. The loop already stops at
sdebug_capacity (after nr_zones zones), so a report can never hold more
than nr_zones descriptors; the clamp does not change the report, it only
bounds arr_len to (nr_zones + 1) * RZONES_DESC_HD, a real device
property that can never reach 0x100000000.

Fixes: 7db0e0c8190a ("scsi: scsi_debug: Fix buffer size of REPORT ZONES command")
Suggested-by: Damien Le Moal <[email protected]>
Cc: [email protected]
Signed-off-by: Ibrahim Hashimov <[email protected]>
Assisted-by: AuditCode-AI:2026.07
Reviewed-by: Damien Le Moal <[email protected]>
Reviewed-by: Bart Van Assche <[email protected]>
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Martin K. Petersen <[email protected]>
[ Adjusted for 5.10: resp_report_zones() here predates commit 4a5fc1c6d752
  ("scsi: scsi_debug: Add gap zone support"), so it still computes

	max_zones = devip->nr_zones - (zs_lba >> devip->zsize_shift);

  and bounds the descriptor loop with it. Keep that existing, tighter
  clamp instead of introducing a second one against devip->nr_zones:
  max_zones is by construction <= devip->nr_zones and is the actual
  number of descriptors the loop can emit, so it satisfies the upstream
  requirement that arr_len be bounded by a real device property while
  leaving the reported zone list unchanged.

  Without the fix 5.10 has the same class of bug from the other end of
  the range: for alloc_len in 1..63 the unsigned (alloc_len - 64)
  underflows, kzalloc(alloc_len) returns a sub-64-byte buffer, and both
  the report header at arr + 0 and desc = arr + 64 are written out of
  bounds. Sizing the allocation from rep_max_zones rather than from
  alloc_len fixes that too. ]
(cherry picked from commit 93dde0bf2f39a0f9f57fd610aa3201ce5b753433)
---
 drivers/scsi/scsi_debug.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/drivers/scsi/scsi_debug.c b/drivers/scsi/scsi_debug.c
index ad4cea4e14848..02d624dd35964 100644
--- a/drivers/scsi/scsi_debug.c
+++ b/drivers/scsi/scsi_debug.c
@@ -4320,6 +4320,7 @@ static int resp_report_zones(struct scsi_cmnd *scp,
 	u32 alloc_len, rep_opts, rep_len;
 	bool partial;
 	u64 lba, zs_lba;
+	u64 arr_len;
 	u8 *arr = NULL, *desc;
 	u8 *cmd = scp->cmnd;
 	struct sdeb_zone_state *zsp;
@@ -4343,10 +4344,12 @@ static int resp_report_zones(struct scsi_cmnd *scp,
 	}
 
 	max_zones = devip->nr_zones - (zs_lba >> devip->zsize_shift);
-	rep_max_zones = min((alloc_len - 64) >> ilog2(RZONES_DESC_HD),
-			    max_zones);
+	rep_max_zones = (ALIGN((u64)alloc_len, RZONES_DESC_HD) - RZONES_DESC_HD) >>
+			ilog2(RZONES_DESC_HD);
+	rep_max_zones = min_t(unsigned int, rep_max_zones, max_zones);
+	arr_len = (u64)RZONES_DESC_HD * (rep_max_zones + 1);
 
-	arr = kzalloc(alloc_len, GFP_ATOMIC | __GFP_NOWARN);
+	arr = kzalloc(arr_len, GFP_ATOMIC | __GFP_NOWARN);
 	if (!arr) {
 		mk_sense_buffer(scp, ILLEGAL_REQUEST, INSUFF_RES_ASC,
 				INSUFF_RES_ASCQ);
-- 
2.53.0