linux-scsi.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Jia-Ju Bai <baijiaju@tsinghua.edu.cn>
To: linuxdrivers@attotech.com, jejb@linux.ibm.com,
	martin.petersen@oracle.com
Cc: linux-scsi@vger.kernel.org, linux-kernel@vger.kernel.org,
	Jia-Ju Bai <baijiaju@tsinghua.edu.cn>
Subject: [PATCH] scsi: esas2r: fix possible buffer overflow caused by bad DMA value in esas2r_process_fs_ioctl()
Date: Sun,  2 Aug 2020 23:21:45 +0800	[thread overview]
Message-ID: <20200802152145.4387-1-baijiaju@tsinghua.edu.cn> (raw)

In esas2r_read_fs():
  struct esas2r_ioctl_fs *fs = 
         (struct esas2r_ioctl_fs *)a->fs_api_buffer;

Because "a->fs_api_buffer" is mapped to coherent DMA (allocated in
esas2r_write_fs()), "fs" is also mapped to DMA.

Then esas2r_read_fs() calls esas2r_process_fs_ioctl() with "fs".

In esas2r_process_fs_ioctl():
  fsc = &fs->command;
  ...
  if (fsc->command >= cmdcnt) {
    fs->status = ATTO_STS_INV_FUNC;
    return false;
  }
  func = cmd_to_fls_func[fsc->command];
  if (func == 0xFF) {
    fs->status = ATTO_STS_INV_FUNC;
    return false;
  }

Because "fs" is mapped to DMA, its data can be modified at anytime by
malicious or malfunctioning hardware. In this case, the check 
"if (fsc->command >= cmdcnt)" can be passed, and then "fsc->command" 
can be modified by hardware to cause buffer overflow.

To fix this problem, "fsc->command" is assigned to a local variable, and
then this local variable is used to replace "fsc->command".

Signed-off-by: Jia-Ju Bai <baijiaju@tsinghua.edu.cn>
---
 drivers/scsi/esas2r/esas2r_flash.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/drivers/scsi/esas2r/esas2r_flash.c b/drivers/scsi/esas2r/esas2r_flash.c
index b02ac389e6c6..ca5ff0d4c7d0 100644
--- a/drivers/scsi/esas2r/esas2r_flash.c
+++ b/drivers/scsi/esas2r/esas2r_flash.c
@@ -851,6 +851,7 @@ bool esas2r_process_fs_ioctl(struct esas2r_adapter *a,
 	struct esas2r_ioctlfs_command *fsc = &fs->command;
 	u8 func = 0;
 	u32 datalen;
+	u8 command = fsc->command;
 
 	fs->status = ATTO_STS_FAILED;
 	fs->driver_error = RS_PENDING;
@@ -860,18 +861,18 @@ bool esas2r_process_fs_ioctl(struct esas2r_adapter *a,
 		return false;
 	}
 
-	if (fsc->command >= cmdcnt) {
+	if (command >= cmdcnt) {
 		fs->status = ATTO_STS_INV_FUNC;
 		return false;
 	}
 
-	func = cmd_to_fls_func[fsc->command];
+	func = cmd_to_fls_func[command];
 	if (func == 0xFF) {
 		fs->status = ATTO_STS_INV_FUNC;
 		return false;
 	}
 
-	if (fsc->command != ESAS2R_FS_CMD_CANCEL) {
+	if (command != ESAS2R_FS_CMD_CANCEL) {
 		if ((a->pcid->device != ATTO_DID_MV_88RC9580
 		     || fs->adap_type != ESAS2R_FS_AT_ESASRAID2)
 		    && (a->pcid->device != ATTO_DID_MV_88RC9580TS
-- 
2.17.1


             reply	other threads:[~2020-08-02 15:22 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-08-02 15:21 Jia-Ju Bai [this message]
2020-08-02 15:47 ` [PATCH] scsi: esas2r: fix possible buffer overflow caused by bad DMA value in esas2r_process_fs_ioctl() James Bottomley
2020-08-03  3:07   ` Jia-Ju Bai
2020-08-03  5:51     ` James Bottomley

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20200802152145.4387-1-baijiaju@tsinghua.edu.cn \
    --to=baijiaju@tsinghua.edu.cn \
    --cc=jejb@linux.ibm.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-scsi@vger.kernel.org \
    --cc=linuxdrivers@attotech.com \
    --cc=martin.petersen@oracle.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).