Bug 238014 - Potential NULL pointer dereference in function ata_dev_advinfo and nvme_dev_advinfo
Summary: Potential NULL pointer dereference in function ata_dev_advinfo and nvme_dev_a...
Status: New
Alias: None
Product: Base System
Classification: Unclassified
Component: kern (show other bugs)
Version: CURRENT
Hardware: Any Any
: --- Affects Many People
Assignee: freebsd-bugs mailing list
URL:
Keywords: patch
Depends on:
Blocks:
 
Reported: 2019-05-21 07:42 UTC by Young
Modified: 2019-05-25 19:50 UTC (History)
1 user (show)

See Also:


Attachments
Proposed patch (2.72 KB, application/mbox)
2019-05-21 07:42 UTC, Young
no flags Details

Note You need to log in before you can comment on or make changes to this bug.
Description Young 2019-05-21 07:42:31 UTC
Created attachment 204502 [details]
Proposed patch

There are two NULL pointer vulnerabilities in functions named ata_dev_advinfo (sys/cam/ata/ata_xpt.c) and nvme_dev_advinfo (sys/cam/nvme/nvme_xpt.c).

We take function ata_dev_advinfo for example.

                if (cdai->flags & CDAI_FLAG_STORE) {
                        if (device->physpath != NULL)
                                free(device->physpath, M_CAMXPT);
                        device->physpath_len = cdai->bufsiz;
                        /* Clear existing buffer if zero length */
                        if (cdai->bufsiz == 0)
                                break;
                        device->physpath = malloc(cdai->bufsiz, M_CAMXPT, M_NOWAIT);
                        if (device->physpath == NULL) {
                                start_ccb->ccb_h.status = CAM_REQ_ABORTED;
                                return;
                        }
                        memcpy(device->physpath, cdai->buf, cdai->bufsiz);


if the physical path is being stored and there is a malloc failure (malloc(9) is called with M_NOWAIT), we could wind up in a situation where the device's physpath_len is set to the length the user provided, but the physpath itself is NULL.

If another context then comes in to fetch the physical path value, we would wind up trying to memcpy a NULL pointer into the caller's buffer.

So, set the physpath_len to 0 when we free the physpath on entry into the store case for the physical path.  Reset the length to a non-zero value only after we've successfully malloced a buffer to hold it.

The attachment is the proposed patch.