NFC: nci: Add some bounds checking in nci_hci_cmd_received()
Subsystems (4)
PATCH 1/2 NFC: nci: Add some bounds checking in nci_hci_cmd_received()
This is similar to commit 674d9de02aa7 ("NFC: Fix possible memory corruption when handling SHDLC I-Frame commands").
I'm not totally sure, but I think that commit description may have overstated the danger. I was under the impression that this data came from the firmware? If you can't trust your networking firmware, then you're already in trouble.
Anyway, these days we add bounds checking where ever we can and we call it kernel hardening. Better safe than sorry.
--- net/nfc/nci/hci.c | 8 ++++++++ 1 file changed, 8 insertions(+)
net/nfc/nci/hci.c +8 -0
@@ -312,6 +312,10 @@ static void nci_hci_cmd_received(struct nci_dev *ndev, u8 pipe, create_info = (struct nci_hci_create_pipe_resp *)skb->data; dest_gate = create_info->dest_gate; new_pipe = create_info->pipe; +if (new_pipe >=NCI_HCI_MAX_PIPES ) { + status =NCI_HCI_ANY_E_NOK ; + goto exit; + }/* Save the new created pipe and bind with local gate, * the description for skb->data[3] is destination gate id
@@ -336,6 +340,10 @@ static void nci_hci_cmd_received(struct nci_dev *ndev, u8 pipe, goto exit; } delete_info = (struct nci_hci_delete_pipe_noti *)skb->data; +if (delete_info->pipe >=NCI_HCI_MAX_PIPES ) { + status =NCI_HCI_ANY_E_NOK ; + goto exit; + } ndev->hci_dev->pipes[delete_info->pipe].gate NCI_HCI_INVALID_GATE;
Signature · 1 line
2.17.1
PATCH 2/2 nfc: nci: Potential off by one in ->pipes[] array
This is similar to commit e285d5bfb7e9 ("NFC: Fix the number of pipes") where we changed NFC_HCI_MAX_PIPES from 127 to 128.
As the comment next to the define explains, the pipe identifier is 7 bits long. The highest possible pipe is 127, but the number of possible pipes is 128. As the code is now, then there is potential for an out of bounds array access:
net/nfc/nci/hci.c:297 nci_hci_cmd_received() warn: array off by one? 'ndev->hci_dev->pipes[pipe]' '0-127 = 127'
--- include/net/nfc/nci_core.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
include/net/nfc/nci_core.h +1 -1
@@ -166,7 +166,7 @@ struct nci_conn_info {* According to specification 102 622 chapter 4.4 Pipes, * the pipe identifier is 7 bits long. */ -#define NCI_HCI_MAX_PIPES 127 +#define NCI_HCI_MAX_PIPES 128struct nci_hci_gate {u8 gate;
Signature · 1 line
2.17.1
From: Dan Carpenter <[email protected]>
Date: Wed, 3 Apr 2019 10:12:48 +0300
Quoted 8 lines
This is similar to commit
674d9de02aa7("NFC: Fix possible memory corruption when handling SHDLC I-Frame commands").I'm not totally sure, but I think that commit description may have overstated the danger. I was under the impression that this data came from the firmware? If you can't trust your networking firmware, then you're already in trouble.
Quoted 5 lines
Anyway, these days we add bounds checking where ever we can and we call it kernel hardening. Better safe than sorry.
Fixes 11f54f228643 ("NFC: nci: Add HCI over NCI protocol support")Signed-off-by Dan Carpenter <[email protected]>
Applied.
Re: [PATCH 2/2 net] nfc: nci: Potential off by one in ->pipes[] array
In reply to Dan CarpenterFrom: Dan Carpenter <[email protected]>
Date: Wed, 3 Apr 2019 10:13:51 +0300
Quoted 8 lines
This is similar to commit
e285d5bfb7e9("NFC: Fix the number of pipes") where we changedNFC_HCI_MAX_PIPESfrom 127 to 128.As the comment next to the define explains, the pipe identifier is 7 bits long. The highest possible pipe is 127, but the number of possible pipes is 128. As the code is now, then there is potential for an out of bounds array access:
Quoted 5 lines
net/nfc/nci/hci.c:297nci_hci_cmd_received()warn: array off by one? 'ndev->hci_dev->pipes[pipe]' '0-127 = 127'Fixes 11f54f228643 ("NFC: nci: Add HCI over NCI protocol support")Signed-off-by Dan Carpenter <[email protected]>
Applied.