From: Zhao Li <enderaoelyther@gmail.com>
To: Ping-Ke Shih <pkshih@realtek.com>
Cc: Hemanth Selam <hemanth.selam@gmail.com>,
Morning Star <alexbestoso@gmail.com>,
linux-wireless@vger.kernel.org, linux-kernel@vger.kernel.org,
stable@vger.kernel.org, Zhao Li <enderaoelyther@gmail.com>
Subject: [PATCH rtw-next 3/3] wifi: rtlwifi: rtl8192cu: don't mark undecrypted frames as decrypted
Date: Sun, 22 Nov 2026 00:35:47 +0800 [thread overview]
Message-ID: <20261121163547.37019-4-enderaoelyther@gmail.com> (raw)
In-Reply-To: <20261121163547.37019-1-enderaoelyther@gmail.com>
The receive descriptor's SWDEC bit is set when a frame still needs
software decryption. It is also clear for a frame that was never
encrypted, and for a protected robust management frame, which the
hardware leaves for software to decrypt. rtl92cu_rx_query_desc()
nevertheless sets RX_FLAG_DECRYPTED for every frame with SWDEC clear,
in two places.
mac80211 then accepts an unprotected unicast data frame although a
pairwise key is installed, and unprotected deauthentication and
disassociation frames on a connection that uses management frame
protection. For a protected robust management frame it skips software
decryption and MIC verification, so a forged frame is accepted and a
genuine encrypted one, such as an SA Query from the peer, reaches
mac80211 still encrypted. A station in radio range can therefore inject
plaintext data into an encrypted connection, or end a protected one, by
spoofing the peer's address.
Set the flag only for a protected frame that is not a robust management
frame, as the rtl8188ee, rtl8192ee, rtl8723ae, rtl8723be and rtl8821ae
drivers do. rtl8xxxu, which supports the same chip, stopped trusting
SWDEC alone in commit cbfbb4ddbc85 ("wifi: rtl8xxxu: enable MFP support
with security flag of RX descriptor"). The header is read before the
receive path strips the driver info, so test the Protected bit first,
which needs only the frame control field, and read that only when the
URB holds it. When Protected is set, _ieee80211_is_robust_mgmt_frame()
also reads only frame_control.
Found by code review. Compile-tested with W=1.
Fixes: 29d00a3e46bb ("rtlwifi: rtl8192cu: Add routine trx")
Cc: stable@vger.kernel.org
Signed-off-by: Zhao Li <enderaoelyther@gmail.com>
---
No new W=1, sparse or smatch warnings.
AI assistance was used.
.../net/wireless/realtek/rtlwifi/rtl8192cu/trx.c | 14 ++++++++++----
1 file changed, 10 insertions(+), 4 deletions(-)
diff --git a/drivers/net/wireless/realtek/rtlwifi/rtl8192cu/trx.c b/drivers/net/wireless/realtek/rtlwifi/rtl8192cu/trx.c
index d6c35e8d02a5..aceedc37e5d2 100644
--- a/drivers/net/wireless/realtek/rtlwifi/rtl8192cu/trx.c
+++ b/drivers/net/wireless/realtek/rtlwifi/rtl8192cu/trx.c
@@ -295,11 +295,14 @@ bool rtl92cu_rx_query_desc(struct ieee80211_hw *hw,
struct rx_desc_92c *p_desc = (struct rx_desc_92c *)pdesc8;
__le32 *pdesc = (__le32 *)pdesc8;
u32 phystatus = get_rx_desc_phy_status(pdesc);
+ struct ieee80211_hdr *hdr;
+ unsigned int hdr_off;
stats->length = (u16)get_rx_desc_pkt_len(pdesc);
stats->rx_drvinfo_size = (u8)get_rx_desc_drvinfo_size(pdesc) *
RX_DRV_INFO_SIZE_UNIT;
stats->rx_bufshift = (u8)(get_rx_desc_shift(pdesc) & 0x03);
+ hdr_off = stats->rx_drvinfo_size + stats->rx_bufshift;
stats->icv = (u16)get_rx_desc_icv(pdesc);
stats->crc = (u16)get_rx_desc_crc32(pdesc);
stats->hwerror = (stats->crc | stats->icv);
@@ -316,15 +319,18 @@ bool rtl92cu_rx_query_desc(struct ieee80211_hw *hw,
rx_status->band = hw->conf.chandef.chan->band;
if (get_rx_desc_crc32(pdesc))
rx_status->flag |= RX_FLAG_FAILED_FCS_CRC;
- if (!get_rx_desc_swdec(pdesc))
- rx_status->flag |= RX_FLAG_DECRYPTED;
if (get_rx_desc_bw(pdesc))
rx_status->bw = RATE_INFO_BW_40;
if (get_rx_desc_rx_ht(pdesc))
rx_status->encoding = RX_ENC_HT;
rx_status->flag |= RX_FLAG_MACTIME_START;
- if (stats->decrypted)
- rx_status->flag |= RX_FLAG_DECRYPTED;
+ if (stats->decrypted &&
+ skb->len >= hdr_off + sizeof(hdr->frame_control)) {
+ hdr = (struct ieee80211_hdr *)(skb->data + hdr_off);
+ if (ieee80211_has_protected(hdr->frame_control) &&
+ !_ieee80211_is_robust_mgmt_frame(hdr))
+ rx_status->flag |= RX_FLAG_DECRYPTED;
+ }
rx_status->rate_idx = rtlwifi_rate_mapping(hw, stats->is_ht,
false, stats->rate);
rx_status->mactime = get_rx_desc_tsfl(pdesc);
--
2.56.1
prev parent reply other threads:[~2026-11-21 16:36 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <20261121163547.37019-1-enderaoelyther@gmail.com>
2026-11-21 16:35 ` [PATCH rtw-next 1/3] wifi: rtlwifi: rtl8192ce: don't mark unencrypted frames as decrypted Zhao Li
2026-11-21 16:35 ` [PATCH rtw-next 2/3] wifi: rtlwifi: rtl8192se: " Zhao Li
2026-11-21 16:35 ` Zhao Li [this message]
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=20261121163547.37019-4-enderaoelyther@gmail.com \
--to=enderaoelyther@gmail.com \
--cc=alexbestoso@gmail.com \
--cc=hemanth.selam@gmail.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-wireless@vger.kernel.org \
--cc=pkshih@realtek.com \
--cc=stable@vger.kernel.org \
/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