[PATCH v3] wifi: mwifiex: Fix OOB and integer underflow when rx packets

[PATCH v3] wifi: mwifiex: Fix OOB and integer underflow when rx packets

[pinkperfect]

Make sure mwifiex_process_mgmt_packet and its callers
mwifiex_process_sta_rx_packet and mwifiex_process_uap_rx_packet
not out-of-bounds access the skb->data buffer.

Fixes: 2dbaf751b1de ("mwifiex: report received management frames to cfg80211")

Signed-off-by: pinkperfect <pinkperfect2021@gmail.com>
---
 drivers/net/wireless/marvell/mwifiex/sta_rx.c   |  3 ++-
 drivers/net/wireless/marvell/mwifiex/uap_txrx.c | 10 +++-------
 drivers/net/wireless/marvell/mwifiex/util.c     |  5 +++++
 3 files changed, 10 insertions(+), 8 deletions(-)

diff --git a/drivers/net/wireless/marvell/mwifiex/sta_rx.c b/drivers/net/wireless/marvell/mwifiex/sta_rx.c
index 13659b02ba88..88aaec645291 100644
--- a/drivers/net/wireless/marvell/mwifiex/sta_rx.c
+++ b/drivers/net/wireless/marvell/mwifiex/sta_rx.c
@@ -194,7 +194,8 @@ int mwifiex_process_sta_rx_packet(struct mwifiex_private *priv,
 
 	rx_pkt_hdr = (void *)local_rx_pd + rx_pkt_offset;
 
-	if ((rx_pkt_offset + rx_pkt_length) > (u16) skb->len) {
+	if ((rx_pkt_offset + rx_pkt_length) > (u16)skb->len ||
+	    skb->len - rx_pkt_offset < sizeof(*rx_pkt_hdr)) {
 		mwifiex_dbg(adapter, ERROR,
 			    "wrong rx packet: len=%d, rx_pkt_offset=%d, rx_pkt_length=%d\n",
 			    skb->len, rx_pkt_offset, rx_pkt_length);
diff --git a/drivers/net/wireless/marvell/mwifiex/uap_txrx.c b/drivers/net/wireless/marvell/mwifiex/uap_txrx.c
index e495f7eaea03..f84ed22518c6 100644
--- a/drivers/net/wireless/marvell/mwifiex/uap_txrx.c
+++ b/drivers/net/wireless/marvell/mwifiex/uap_txrx.c
@@ -367,20 +367,15 @@ int mwifiex_process_uap_rx_packet(struct mwifiex_private *priv,
 	rx_pkt_type = le16_to_cpu(uap_rx_pd->rx_pkt_type);
 	rx_pkt_hdr = (void *)uap_rx_pd + le16_to_cpu(uap_rx_pd->rx_pkt_offset);
 
-	ether_addr_copy(ta, rx_pkt_hdr->eth803_hdr.h_source);
-
 	if ((le16_to_cpu(uap_rx_pd->rx_pkt_offset) +
-	     le16_to_cpu(uap_rx_pd->rx_pkt_length)) > (u16) skb->len) {
+	     le16_to_cpu(uap_rx_pd->rx_pkt_length)) > (u16)skb->len ||
+	     skb->len - le16_to_cpu(uap_rx_pd->rx_pkt_offset) < sizeof(*rx_pkt_hdr)) {
 		mwifiex_dbg(adapter, ERROR,
 			    "wrong rx packet: len=%d, offset=%d, length=%d\n",
 			    skb->len, le16_to_cpu(uap_rx_pd->rx_pkt_offset),
 			    le16_to_cpu(uap_rx_pd->rx_pkt_length));
 		priv->stats.rx_dropped++;
 
-		node = mwifiex_get_sta_entry(priv, ta);
-		if (node)
-			node->stats.tx_failed++;
-
 		dev_kfree_skb_any(skb);
 		return 0;
 	}
@@ -393,6 +388,7 @@ int mwifiex_process_uap_rx_packet(struct mwifiex_private *priv,
 		return ret;
 	}
 
+	ether_addr_copy(ta, rx_pkt_hdr->eth803_hdr.h_source);
 
 	if (rx_pkt_type != PKT_TYPE_BAR && uap_rx_pd->priority < MAX_NUM_TID) {
 		spin_lock_bh(&priv->sta_list_spinlock);
diff --git a/drivers/net/wireless/marvell/mwifiex/util.c b/drivers/net/wireless/marvell/mwifiex/util.c
index 94c2d219835d..31e1a82883e4 100644
--- a/drivers/net/wireless/marvell/mwifiex/util.c
+++ b/drivers/net/wireless/marvell/mwifiex/util.c
@@ -399,6 +399,11 @@ mwifiex_process_mgmt_packet(struct mwifiex_private *priv,
 
 	pkt_len = le16_to_cpu(rx_pd->rx_pkt_length);
 
+	if (pkt_len < sizeof(struct ieee80211_hdr) || skb->len < pkt_len) {
+		mwifiex_dbg(priv->adapter, ERROR, "invalid rx_pkt_length");
+		return -1;
+	}
+
 	ieee_hdr = (void *)skb->data;
 	if (ieee80211_is_mgmt(ieee_hdr->frame_control)) {
 		if (mwifiex_parse_mgmt_packet(priv, (u8 *)ieee_hdr,
-- 
2.25.1

Re: [PATCH v3] wifi: mwifiex: Fix OOB and integer underflow when rx packets

[Jakub Kicinski]

On Thu, 13 Jul 2023 10:44:56 +0800 Pink Perfect wrote:
> Sorry, forgot to change the signoff name, after you review this patch I can
> submit a new one

Do you mean change to "Pink Perfect", is that the name you'd sign
legal documents with?

> On Thu, Jul 13, 2023 at 10:37 AM pinkperfect <pinkperfect2021@gmail.com>
> wrote:
> 
> > Make sure mwifiex_process_mgmt_packet and its callers
> > mwifiex_process_sta_rx_packet and mwifiex_process_uap_rx_packet
> > not out-of-bounds access the skb->data buffer.
> >
> > Fixes: 2dbaf751b1de ("mwifiex: report received management frames to
> > cfg80211")
> >
> > Signed-off-by: pinkperfect <pinkperfect2021@gmail.com>

No empty lines between tags, please.

You mentioned reporting the problem to chromeos, since mwifiex
"maintainters" seem to be AWoL, would someone from their team
possibly be willing to venture a review tag for the patch?

Re: [PATCH v3] wifi: mwifiex: Fix OOB and integer underflow when rx packets

[Kalle Valo]

Jakub Kicinski <kuba@kernel.org> writes:

> On Thu, 13 Jul 2023 10:44:56 +0800 Pink Perfect wrote:
>> Sorry, forgot to change the signoff name, after you review this patch I can
>> submit a new one
>
> Do you mean change to "Pink Perfect", is that the name you'd sign
> legal documents with?
>
>> On Thu, Jul 13, 2023 at 10:37 AM pinkperfect <pinkperfect2021@gmail.com>
>> wrote:
>> 
>> > Make sure mwifiex_process_mgmt_packet and its callers
>> > mwifiex_process_sta_rx_packet and mwifiex_process_uap_rx_packet
>> > not out-of-bounds access the skb->data buffer.
>> >
>> > Fixes: 2dbaf751b1de ("mwifiex: report received management frames to
>> > cfg80211")
>> >
>> > Signed-off-by: pinkperfect <pinkperfect2021@gmail.com>
>
> No empty lines between tags, please.
>
> You mentioned reporting the problem to chromeos, since mwifiex
> "maintainters" seem to be AWoL, would someone from their team
> possibly be willing to venture a review tag for the patch?

We have four maintainers for mwifiex and total silence:

MARVELL MWIFIEX WIRELESS DRIVER
M:      Amitkumar Karwar <amitkarwar@gmail.com>
M:      Ganapathi Bhat <ganapathi017@gmail.com>
M:      Sharvari Harisangam <sharvari.harisangam@nxp.com>
M:      Xinming Hu <huxinming820@gmail.com>
L:      linux-wireless@vger.kernel.org
S:      Maintained
F:      drivers/net/wireless/marvell/mwifiex/

I'm very close of marking this driver as orphan unless anyone steps up.
This is not how to maintain a driver.

-- 
https://patchwork.kernel.org/project/linux-wireless/list/

https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches

Re: [PATCH v3] wifi: mwifiex: Fix OOB and integer underflow when rx packets

[Jakub Kicinski]

On Thu, 20 Jul 2023 09:55:38 +0300 Kalle Valo wrote:
> > No empty lines between tags, please.
> >
> > You mentioned reporting the problem to chromeos, since mwifiex
> > "maintainters" seem to be AWoL, would someone from their team
> > possibly be willing to venture a review tag for the patch?  
> 
> We have four maintainers for mwifiex and total silence:
> 
> MARVELL MWIFIEX WIRELESS DRIVER
> M:      Amitkumar Karwar <amitkarwar@gmail.com>
> M:      Ganapathi Bhat <ganapathi017@gmail.com>
> M:      Sharvari Harisangam <sharvari.harisangam@nxp.com>
> M:      Xinming Hu <huxinming820@gmail.com>
> L:      linux-wireless@vger.kernel.org
> S:      Maintained
> F:      drivers/net/wireless/marvell/mwifiex/
> 
> I'm very close of marking this driver as orphan unless anyone steps up.

That seems more than justified at this point.

Re: [PATCH v3] wifi: mwifiex: Fix OOB and integer underflow when rx packets

[Brian Norris]

On Thu, Jul 20, 2023 at 09:55:38AM +0300, Kalle Valo wrote:
> We have four maintainers for mwifiex and total silence:
> 
> MARVELL MWIFIEX WIRELESS DRIVER
> M:      Amitkumar Karwar <amitkarwar@gmail.com>
> M:      Ganapathi Bhat <ganapathi017@gmail.com>
> M:      Sharvari Harisangam <sharvari.harisangam@nxp.com>
> M:      Xinming Hu <huxinming820@gmail.com>
> L:      linux-wireless@vger.kernel.org
> S:      Maintained
> F:      drivers/net/wireless/marvell/mwifiex/
> 
> I'm very close of marking this driver as orphan unless anyone steps up.
> This is not how to maintain a driver.

I'd be fully on board with removing these maintainers, as I don't recall
hearing from any of them in years. (In fact, some of these addresses
don't have a single mail logged on lore.kernel.org/all/...) I just
didn't want to be the one to say it.

On the other hand, I regularly look at pretty much anything for mwifiex,
as long as the submitter is in relatively good faith. So I wouldn't mind
being a Reviewer (or Maintainer? what's the difference, when Kalle
does the committing anyway?). And that might qualify as "Odd Fixes", as
I don't plan on doing much more than keeping the lights on.

I'll submit the MAINTAINERS patch if you'd like.

Brian

Re: [PATCH v3] wifi: mwifiex: Fix OOB and integer underflow when rx packets

[Kalle Valo]

Brian Norris <briannorris@chromium.org> writes:

> On Thu, Jul 20, 2023 at 09:55:38AM +0300, Kalle Valo wrote:
>> We have four maintainers for mwifiex and total silence:
>> 
>> MARVELL MWIFIEX WIRELESS DRIVER
>> M:      Amitkumar Karwar <amitkarwar@gmail.com>
>> M:      Ganapathi Bhat <ganapathi017@gmail.com>
>> M:      Sharvari Harisangam <sharvari.harisangam@nxp.com>
>> M:      Xinming Hu <huxinming820@gmail.com>
>> L:      linux-wireless@vger.kernel.org
>> S:      Maintained
>> F:      drivers/net/wireless/marvell/mwifiex/
>> 
>> I'm very close of marking this driver as orphan unless anyone steps up.
>> This is not how to maintain a driver.
>
> I'd be fully on board with removing these maintainers, as I don't recall
> hearing from any of them in years. (In fact, some of these addresses
> don't have a single mail logged on lore.kernel.org/all/...) I just
> didn't want to be the one to say it.
>
> On the other hand, I regularly look at pretty much anything for mwifiex,
> as long as the submitter is in relatively good faith. So I wouldn't mind
> being a Reviewer (or Maintainer? what's the difference, when Kalle
> does the committing anyway?).

Heh, that's a good question. I don't know what was the original intent
for a reviewer role but in my view ideally a driver should 1-2
maintainers, no more, and if there are more people involved they should
be reviewers. And maintainers should use Acked-by, Reviewers should use
Reviewed-by. And if I see an Acked-by from a maintainer I usually don't
review the patch so closely and just take it directly (of there are
exceptions, as always). In this case I would prefer you being the
maintainer, even if you wouldn't have much time for mwifiex.

But I don't know how others see it. Jakub has been writing documentation
about maintainership which is also a good read:

https://lore.kernel.org/all/20230719183225.1827100-1-kuba@kernel.org/

> And that might qualify as "Odd Fixes", as
> I don't plan on doing much more than keeping the lights on.
>
> I'll submit the MAINTAINERS patch if you'd like.

Sounds very good to me, thank you! Please submit the patch if you can.

-- 
https://patchwork.kernel.org/project/linux-wireless/list/

https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches

Re: [PATCH v3] wifi: mwifiex: Fix OOB and integer underflow when rx packets

[Brian Norris]

On Fri, Jul 21, 2023 at 10:15:25AM +0300, Kalle Valo wrote:
> Brian Norris <briannorris@chromium.org> writes:
> > I'll submit the MAINTAINERS patch if you'd like.
> 
> Sounds very good to me, thank you! Please submit the patch if you can.

Done:
https://lore.kernel.org/linux-wireless/20230721160603.1.Idf0e8025f59c62d73c08960638249b58cf215acc@changeid/

Brian