From patchwork Wed Mar 15 20:21:10 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Fedor Pchelkin X-Patchwork-Id: 13176624 X-Patchwork-Delegate: toke@redhat.com Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id BD4D8C6FD1D for ; Wed, 15 Mar 2023 20:22:22 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S232487AbjCOUWV (ORCPT ); Wed, 15 Mar 2023 16:22:21 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:41628 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S232278AbjCOUWU (ORCPT ); Wed, 15 Mar 2023 16:22:20 -0400 Received: from mail.ispras.ru (mail.ispras.ru [83.149.199.84]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 48D9DA0F00; Wed, 15 Mar 2023 13:22:19 -0700 (PDT) Received: from fpc.intra.ispras.ru (unknown [10.10.165.16]) by mail.ispras.ru (Postfix) with ESMTPSA id 4B84E44C1010; Wed, 15 Mar 2023 20:22:17 +0000 (UTC) DKIM-Filter: OpenDKIM Filter v2.11.0 mail.ispras.ru 4B84E44C1010 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=ispras.ru; s=default; t=1678911737; bh=sNJDCRIt59eJ1lrh220dzpHP4PwNZfpMpr1y7yJ9szY=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=iOW4/2FS9/ve654NMTfeFUxqHd7YR2DaT9Jk5FvnsEEYwYlMvK9nMUB5n98mSM+x1 v8PKGwldQkizuAzdHp//w98dlhiLxeIkXxLjnZpP56tYyZ/sYo5e9SfGQ2t1oAt/+q qzF12IXIcS4E4sVyRVHV6VXkaeXZLKSghIfSPNBw= From: Fedor Pchelkin To: =?utf-8?q?Toke_H=C3=B8iland-J=C3=B8rgensen?= Cc: Fedor Pchelkin , Kalle Valo , "David S. Miller" , Eric Dumazet , Jakub Kicinski , Paolo Abeni , Senthil Balasubramanian , "John W. Linville" , Vasanthakumar Thiagarajan , Sujith , linux-wireless@vger.kernel.org, netdev@vger.kernel.org, linux-kernel@vger.kernel.org, Alexey Khoroshilov , lvc-project@linuxtesting.org, syzbot+f2cb6e0ffdb961921e4d@syzkaller.appspotmail.com Subject: [PATCH 1/3] wifi: ath9k: avoid referencing uninit memory in ath9k_wmi_ctrl_rx Date: Wed, 15 Mar 2023 23:21:10 +0300 Message-Id: <20230315202112.163012-2-pchelkin@ispras.ru> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20230315202112.163012-1-pchelkin@ispras.ru> References: <20230315202112.163012-1-pchelkin@ispras.ru> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: linux-wireless@vger.kernel.org For the reasons described in commit b383e8abed41 ("wifi: ath9k: avoid uninit memory read in ath9k_htc_rx_msg()"), ath9k_htc_rx_msg() should validate pkt_len before accessing the SKB. For example, the obtained SKB may have uninitialized memory in the case of ioctl(USB_RAW_IOCTL_EP_WRITE). Implement sanity checking inside the corresponding endpoint RX handlers: ath9k_wmi_ctrl_rx() and ath9k_htc_rxep(). Otherwise, uninit memory can be referenced. Add comments briefly describing the issue. Found by Linux Verification Center (linuxtesting.org) with Syzkaller. Fixes: fb9987d0f748 ("ath9k_htc: Support for AR9271 chipset.") Reported-and-tested-by: syzbot+f2cb6e0ffdb961921e4d@syzkaller.appspotmail.com Signed-off-by: Fedor Pchelkin --- drivers/net/wireless/ath/ath9k/htc_drv_txrx.c | 6 ++++++ drivers/net/wireless/ath/ath9k/htc_hst.c | 4 ++++ drivers/net/wireless/ath/ath9k/wmi.c | 8 ++++++++ 3 files changed, 18 insertions(+) diff --git a/drivers/net/wireless/ath/ath9k/htc_drv_txrx.c b/drivers/net/wireless/ath/ath9k/htc_drv_txrx.c index 672789e3c55d..957efb26019d 100644 --- a/drivers/net/wireless/ath/ath9k/htc_drv_txrx.c +++ b/drivers/net/wireless/ath/ath9k/htc_drv_txrx.c @@ -1147,6 +1147,12 @@ void ath9k_htc_rxep(void *drv_priv, struct sk_buff *skb, if (!data_race(priv->rx.initialized)) goto err; + /* Validate the obtained SKB so that it is handled without error + * inside rx_tasklet handler. + */ + if (unlikely(skb->len < sizeof(struct ieee80211_hdr))) + goto err; + spin_lock_irqsave(&priv->rx.rxbuflock, flags); list_for_each_entry(tmp_buf, &priv->rx.rxbuf, list) { if (!tmp_buf->in_process) { diff --git a/drivers/net/wireless/ath/ath9k/htc_hst.c b/drivers/net/wireless/ath/ath9k/htc_hst.c index fe62ff668f75..9d0d9d0e1aa8 100644 --- a/drivers/net/wireless/ath/ath9k/htc_hst.c +++ b/drivers/net/wireless/ath/ath9k/htc_hst.c @@ -475,6 +475,10 @@ void ath9k_htc_rx_msg(struct htc_target *htc_handle, skb_pull(skb, sizeof(struct htc_frame_hdr)); endpoint = &htc_handle->endpoint[epid]; + + /* The endpoint RX handlers should implement their own + * additional SKB sanity checking + */ if (endpoint->ep_callbacks.rx) endpoint->ep_callbacks.rx(endpoint->ep_callbacks.priv, skb, epid); diff --git a/drivers/net/wireless/ath/ath9k/wmi.c b/drivers/net/wireless/ath/ath9k/wmi.c index 19345b8f7bfd..2e7c361b62f5 100644 --- a/drivers/net/wireless/ath/ath9k/wmi.c +++ b/drivers/net/wireless/ath/ath9k/wmi.c @@ -204,6 +204,10 @@ static void ath9k_wmi_rsp_callback(struct wmi *wmi, struct sk_buff *skb) { skb_pull(skb, sizeof(struct wmi_cmd_hdr)); + /* Once again validate the SKB. */ + if (unlikely(skb->len < wmi->cmd_rsp_len)) + return; + if (wmi->cmd_rsp_buf != NULL && wmi->cmd_rsp_len != 0) memcpy(wmi->cmd_rsp_buf, skb->data, wmi->cmd_rsp_len); @@ -221,6 +225,10 @@ static void ath9k_wmi_ctrl_rx(void *priv, struct sk_buff *skb, if (unlikely(wmi->stopped)) goto free_skb; + /* Validate the obtained SKB. */ + if (unlikely(skb->len < sizeof(struct wmi_cmd_hdr))) + goto free_skb; + hdr = (struct wmi_cmd_hdr *) skb->data; cmd_id = be16_to_cpu(hdr->command_id); From patchwork Wed Mar 15 20:21:11 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Fedor Pchelkin X-Patchwork-Id: 13176625 X-Patchwork-Delegate: toke@redhat.com Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id CA7ACC61DA4 for ; Wed, 15 Mar 2023 20:22:37 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S232550AbjCOUWg (ORCPT ); Wed, 15 Mar 2023 16:22:36 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:42100 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S232812AbjCOUWd (ORCPT ); Wed, 15 Mar 2023 16:22:33 -0400 Received: from mail.ispras.ru (mail.ispras.ru [83.149.199.84]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id C43A9A0F34; Wed, 15 Mar 2023 13:22:22 -0700 (PDT) Received: from fpc.intra.ispras.ru (unknown [10.10.165.16]) by mail.ispras.ru (Postfix) with ESMTPSA id B6F5E44C1011; Wed, 15 Mar 2023 20:22:20 +0000 (UTC) DKIM-Filter: OpenDKIM Filter v2.11.0 mail.ispras.ru B6F5E44C1011 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=ispras.ru; s=default; t=1678911740; bh=OrnpI5NnHl+YmfKjeyUc8fbO91ALa8VLak03DsUl8sY=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=qoRg1+hQUtvvsjxslFN50LmZFPjj1sU/YtGyrYoBzGx8lrHunJ5uyZfMyiP+WStSZ 68sT46/hEdpFbadO4pcLfu8UoA46W0wMGWUWDH+mrERu8NVnMgrBhhOvJAyKvABCAd a97RF7oquPObq2LzTga6ggECjDqejZla2fKe9ILI= From: Fedor Pchelkin To: =?utf-8?q?Toke_H=C3=B8iland-J=C3=B8rgensen?= Cc: Fedor Pchelkin , Kalle Valo , "David S. Miller" , Eric Dumazet , Jakub Kicinski , Paolo Abeni , Senthil Balasubramanian , "John W. Linville" , Vasanthakumar Thiagarajan , Sujith , linux-wireless@vger.kernel.org, netdev@vger.kernel.org, linux-kernel@vger.kernel.org, Alexey Khoroshilov , lvc-project@linuxtesting.org, syzbot+df61b36319e045c00a08@syzkaller.appspotmail.com Subject: [PATCH 2/3] wifi: ath9k: fix races between ath9k_wmi_cmd and ath9k_wmi_ctrl_rx Date: Wed, 15 Mar 2023 23:21:11 +0300 Message-Id: <20230315202112.163012-3-pchelkin@ispras.ru> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20230315202112.163012-1-pchelkin@ispras.ru> References: <20230315202112.163012-1-pchelkin@ispras.ru> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: linux-wireless@vger.kernel.org Currently, the synchronization between ath9k_wmi_cmd() and ath9k_wmi_ctrl_rx() is exposed to races which can result, for example, not only in wmi->cmd_rsp_buf being incorrectly initialized, but in overall invalid behaviour of ath9k_wmi_cmd(). Consider the following scenario: CPU0 CPU1 ath9k_wmi_cmd(...) mutex_lock(&wmi->op_mutex) ath9k_wmi_cmd_issue(...) wait_for_completion_timeout(...) --- timeout --- mutex_unlock(&wmi->op_mutex) return -ETIMEDOUT ath9k_wmi_cmd(...) mutex_lock(&wmi->op_mutex) ath9k_wmi_cmd_issue(...) wait_for_completion_timeout(...) /* the one left after the first * ath9k_wmi_cmd call */ ath9k_wmi_rsp_callback(...) memcpy(...) complete(&wmi->cmd_wait); /* Awakened by the bogus callback * => invalid return */ mutex_unlock(&wmi->op_mutex) return 0 This may occur even on uniprocessor machines, and in SMP case the races may be even more intricate. To fix this, move the contents of ath9k_wmi_rsp_callback() under wmi_lock inside ath9k_wmi_ctrl_rx() so that the wmi->cmd_wait can be completed only for initially designated wmi_cmd call, otherwise the path would be rejected with last_seq_id check. Also move recording the rsp buffer and length into ath9k_wmi_cmd_issue() under the same wmi_lock with last_seq_id update to avoid their racy changes. Found by Linux Verification Center (linuxtesting.org) with Syzkaller. Fixes: fb9987d0f748 ("ath9k_htc: Support for AR9271 chipset.") Reported-and-tested-by: syzbot+df61b36319e045c00a08@syzkaller.appspotmail.com Signed-off-by: Fedor Pchelkin --- This patch depends on [PATCH 1/3] wifi: ath9k: avoid referencing uninit memory in ath9k_wmi_ctrl_rx drivers/net/wireless/ath/ath9k/wmi.c | 48 ++++++++++++++-------------- 1 file changed, 24 insertions(+), 24 deletions(-) diff --git a/drivers/net/wireless/ath/ath9k/wmi.c b/drivers/net/wireless/ath/ath9k/wmi.c index 2e7c361b62f5..99a91bbaace9 100644 --- a/drivers/net/wireless/ath/ath9k/wmi.c +++ b/drivers/net/wireless/ath/ath9k/wmi.c @@ -200,20 +200,6 @@ void ath9k_fatal_work(struct work_struct *work) ath9k_htc_reset(priv); } -static void ath9k_wmi_rsp_callback(struct wmi *wmi, struct sk_buff *skb) -{ - skb_pull(skb, sizeof(struct wmi_cmd_hdr)); - - /* Once again validate the SKB. */ - if (unlikely(skb->len < wmi->cmd_rsp_len)) - return; - - if (wmi->cmd_rsp_buf != NULL && wmi->cmd_rsp_len != 0) - memcpy(wmi->cmd_rsp_buf, skb->data, wmi->cmd_rsp_len); - - complete(&wmi->cmd_wait); -} - static void ath9k_wmi_ctrl_rx(void *priv, struct sk_buff *skb, enum htc_endpoint_id epid) { @@ -242,14 +228,26 @@ static void ath9k_wmi_ctrl_rx(void *priv, struct sk_buff *skb, /* Check if there has been a timeout. */ spin_lock_irqsave(&wmi->wmi_lock, flags); - if (be16_to_cpu(hdr->seq_no) != wmi->last_seq_id) { + if (be16_to_cpu(hdr->seq_no) != wmi->last_seq_id || + be16_to_cpu(hdr->seq_no) == 0) { + spin_unlock_irqrestore(&wmi->wmi_lock, flags); + goto free_skb; + } + + /* Next, process WMI command response */ + skb_pull(skb, sizeof(struct wmi_cmd_hdr)); + + /* Once again validate the SKB. */ + if (unlikely(skb->len < wmi->cmd_rsp_len)) { spin_unlock_irqrestore(&wmi->wmi_lock, flags); goto free_skb; } - spin_unlock_irqrestore(&wmi->wmi_lock, flags); - /* WMI command response */ - ath9k_wmi_rsp_callback(wmi, skb); + if (wmi->cmd_rsp_buf != NULL && wmi->cmd_rsp_len != 0) + memcpy(wmi->cmd_rsp_buf, skb->data, wmi->cmd_rsp_len); + + complete(&wmi->cmd_wait); + spin_unlock_irqrestore(&wmi->wmi_lock, flags); free_skb: kfree_skb(skb); @@ -287,7 +285,8 @@ int ath9k_wmi_connect(struct htc_target *htc, struct wmi *wmi, static int ath9k_wmi_cmd_issue(struct wmi *wmi, struct sk_buff *skb, - enum wmi_cmd_id cmd, u16 len) + enum wmi_cmd_id cmd, u16 len, + u8 *rsp_buf, u32 rsp_len) { struct wmi_cmd_hdr *hdr; unsigned long flags; @@ -297,6 +296,11 @@ static int ath9k_wmi_cmd_issue(struct wmi *wmi, hdr->seq_no = cpu_to_be16(++wmi->tx_seq_id); spin_lock_irqsave(&wmi->wmi_lock, flags); + + /* record the rsp buffer and length */ + wmi->cmd_rsp_buf = rsp_buf; + wmi->cmd_rsp_len = rsp_len; + wmi->last_seq_id = wmi->tx_seq_id; spin_unlock_irqrestore(&wmi->wmi_lock, flags); @@ -337,11 +341,7 @@ int ath9k_wmi_cmd(struct wmi *wmi, enum wmi_cmd_id cmd_id, goto out; } - /* record the rsp buffer and length */ - wmi->cmd_rsp_buf = rsp_buf; - wmi->cmd_rsp_len = rsp_len; - - ret = ath9k_wmi_cmd_issue(wmi, skb, cmd_id, cmd_len); + ret = ath9k_wmi_cmd_issue(wmi, skb, cmd_id, cmd_len, rsp_buf, rsp_len); if (ret) goto out; From patchwork Wed Mar 15 20:21:12 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Fedor Pchelkin X-Patchwork-Id: 13176626 X-Patchwork-Delegate: toke@redhat.com Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 2AA7FC7618D for ; Wed, 15 Mar 2023 20:22:41 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S232991AbjCOUWi (ORCPT ); Wed, 15 Mar 2023 16:22:38 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:42120 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S232877AbjCOUWe (ORCPT ); Wed, 15 Mar 2023 16:22:34 -0400 Received: from mail.ispras.ru (mail.ispras.ru [83.149.199.84]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 4A0B9A21AE; Wed, 15 Mar 2023 13:22:24 -0700 (PDT) Received: from fpc.intra.ispras.ru (unknown [10.10.165.16]) by mail.ispras.ru (Postfix) with ESMTPSA id 000B144C101F; Wed, 15 Mar 2023 20:22:22 +0000 (UTC) DKIM-Filter: OpenDKIM Filter v2.11.0 mail.ispras.ru 000B144C101F DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=ispras.ru; s=default; t=1678911743; bh=KHlEKAhpUGzm2mZCMucQ8WbZWdQoP58ONPTTGbeX3MM=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=pF8+mLdy40gtvvg8OB9D1OQDILvPJaHvvHMnT7RNekOc6QhJSM3bhgNu+no7WBUPh iEbbhudO8Dzq8G8WsxsNaChuyN0EA2dyfCCBcJVUnq7sj5jBuHY/bny1AQS0INSVZj 5t3iTEBkUd36YwJFAhhKN1TBuroFwJAK3VXphddo= From: Fedor Pchelkin To: =?utf-8?q?Toke_H=C3=B8iland-J=C3=B8rgensen?= Cc: Fedor Pchelkin , Kalle Valo , "David S. Miller" , Eric Dumazet , Jakub Kicinski , Paolo Abeni , Senthil Balasubramanian , "John W. Linville" , Vasanthakumar Thiagarajan , Sujith , linux-wireless@vger.kernel.org, netdev@vger.kernel.org, linux-kernel@vger.kernel.org, Alexey Khoroshilov , lvc-project@linuxtesting.org Subject: [PATCH 3/3] wifi: ath9k: fix ath9k_wmi_cmd return value when device is unplugged Date: Wed, 15 Mar 2023 23:21:12 +0300 Message-Id: <20230315202112.163012-4-pchelkin@ispras.ru> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20230315202112.163012-1-pchelkin@ispras.ru> References: <20230315202112.163012-1-pchelkin@ispras.ru> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: linux-wireless@vger.kernel.org Return with an error code in case the USB device has been already unplugged. Otherwise the callers of ath9k_wmi_cmd() are unaware of the fact that cmd_buf and rsp_buf are not initialized or handled properly inside this function. Found by Linux Verification Center (linuxtesting.org) with Syzkaller. Fixes: a3be14b76da1 ("ath9k_htc: Handle device unplug properly") Signed-off-by: Fedor Pchelkin --- drivers/net/wireless/ath/ath9k/wmi.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/net/wireless/ath/ath9k/wmi.c b/drivers/net/wireless/ath/ath9k/wmi.c index 99a91bbaace9..3e0ad4f8f0a0 100644 --- a/drivers/net/wireless/ath/ath9k/wmi.c +++ b/drivers/net/wireless/ath/ath9k/wmi.c @@ -320,8 +320,11 @@ int ath9k_wmi_cmd(struct wmi *wmi, enum wmi_cmd_id cmd_id, unsigned long time_left; int ret = 0; - if (ah->ah_flags & AH_UNPLUGGED) - return 0; + if (ah->ah_flags & AH_UNPLUGGED) { + ath_dbg(common, WMI, "Device unplugged for WMI command: %s\n", + wmi_cmd_to_name(cmd_id)); + return -ENODEV; + } skb = alloc_skb(headroom + cmd_len, GFP_ATOMIC); if (!skb)