At least two slightly different paths can lead to a NULL pointer dereference in scsi_prep_state_check and have been hit in practice. 1. Call Trace: [<ffffffff812f8c4b>] scsi_setup_blk_pc_cmnd+0x2b/0x170 [<ffffffff81305108>] sd_prep_fn+0x568/0xdd0 [<ffffffff8121c594>] blk_peek_request+0xb4/0x240 [<ffffffff812f94fe>] scsi_request_fn+0x43e/0x4a0 [<ffffffff8121621b>] __blk_run_queue+0x1b/0x20 [<ffffffff81220638>] blk_execute_rq_nowait+0x68/0xc0 [<ffffffff812206fd>] blk_execute_rq+0x6d/0x100 [<ffffffff812f8f8c>] scsi_execute+0xfc/0x160 [<ffffffff812f975a>] scsi_execute_req+0xca/0x140 [<ffffffff81305a4e>] sd_revalidate_disk+0xde/0x1a00 [<ffffffff8130749b>] sd_probe_async+0x12b/0x1d0 ... 2. Call Trace: [<ffffffff8130078b>] scsi_setup_blk_pc_cmnd+0x2b/0x170 [<ffffffff8130091d>] scsi_prep_fn+0x4d/0x60 [<ffffffff81227794>] blk_peek_request+0xb4/0x240 [<ffffffff813010cd>] scsi_request_fn+0x47d/0x4b0 [<ffffffff81223bff>] __blk_run_queue+0x1f/0x80 [<ffffffff8122b808>] blk_execute_rq_nowait+0x68/0xc0 ... This patch should fix both cases. Signed-off-by: Joern Engel <joern@xxxxxxxxx> --- drivers/scsi/scsi_lib.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c index bb216ff..5fb5a1c 100644 --- a/drivers/scsi/scsi_lib.c +++ b/drivers/scsi/scsi_lib.c @@ -1079,8 +1079,11 @@ static struct scsi_cmnd *scsi_get_cmd_from_req(struct scsi_device *sdev, int scsi_setup_blk_pc_cmnd(struct scsi_device *sdev, struct request *req) { struct scsi_cmnd *cmd; - int ret = scsi_prep_state_check(sdev, req); + int ret; + if (!sdev) + return BLKPREP_KILL; + ret = scsi_prep_state_check(sdev, req); if (ret != BLKPREP_OK) return ret; -- 1.7.10.4 -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html