Skip to contentSkip to conversation

net: ath9k: fix use-after-free in ath9k_hif_usb_rx_cb

PATCH Superseded Pavel Skripkin 2 participants 10 messages Aug 4, 2021 – Oct 2, 2021

Subsystems (3)

v1 of 6 Superseded Latest is v6 →

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 net: ath9k: fix use-after-free in ath9k_hif_usb_rx_cb

Syzbot reported use-after-free Read in ath9k_hif_usb_rx_cb(). The problem was in incorrect htc_handle->drv_priv initialization.

Probable call trace which can trigger use-after-free:

ath9k_htc_probe_device()
/* htc_handle->drv_priv = priv; */
ath9k_htc_wait_for_target() <--- Failed
ieee80211_free_hw() <--- priv pointer is freed

<IRQ>

...
ath9k_hif_usb_rx_cb()
ath9k_hif_usb_rx_stream()
RX_STAT_INC() <--- htc_handle->drv_priv access

In order to not add fancy protection for drv_priv we can move htc_handle->drv_priv initialization at the end of the ath9k_htc_probe_device() and add helper macro to make all *_STAT_* macros NULL save.

Also, I made whitespaces clean ups in *_STAT_* macros definitions to make checkpatch.pl happy.

Fixes fb9987d0f748 ("ath9k_htc: Support for AR9271 chipset.")
Reported-and-tested-by [email protected]

---

Hi, ath9k maintainer/developers!

I know, that you do not like changes, that wasn't tested on real hardware. I really don't access to this one, so I'd like you to test it on real hardware piece, if you have one. At least, this patch was tested by syzbot [1]

[1] https://syzkaller.appspot.com/bug?id=6ead44e37afb6866ac0c7dd121b4ce07cb665f60

---
 drivers/net/wireless/ath/ath9k/htc.h          | 11 ++++++-----
 drivers/net/wireless/ath/ath9k/htc_drv_init.c |  3 ++-
 2 files changed, 8 insertions(+), 6 deletions(-)
drivers/net/wireless/ath/ath9k/htc.h +6 -5
index 0a1634238e67..c16b2a482e83 100644
@@ -326,11 +326,12 @@ static inline struct ath9k_htc_tx_ctl *HTC_SKB_CB(struct sk_buff *skb)
 
 #ifdef CONFIG_ATH9K_HTC_DEBUGFS
 
-#define TX_STAT_INC(c) (hif_dev->htc_handle->drv_priv->debug.tx_stats.c++)
-#define TX_STAT_ADD(c, a) (hif_dev->htc_handle->drv_priv->debug.tx_stats.c += a)
-#define RX_STAT_INC(c) (hif_dev->htc_handle->drv_priv->debug.skbrx_stats.c++)
-#define RX_STAT_ADD(c, a) (hif_dev->htc_handle->drv_priv->debug.skbrx_stats.c += a)
-#define CAB_STAT_INC   priv->debug.tx_stats.cab_queued++
+#define __STAT_SAVE(expr)	(hif_dev->htc_handle->drv_priv ? (expr) : 0)
+#define TX_STAT_INC(c)		__STAT_SAVE(hif_dev->htc_handle->drv_priv->debug.tx_stats.c++)
+#define TX_STAT_ADD(c, a)	__STAT_SAVE(hif_dev->htc_handle->drv_priv->debug.tx_stats.c += a)
+#define RX_STAT_INC(c)		__STAT_SAVE(hif_dev->htc_handle->drv_priv->debug.skbrx_stats.c++)
+#define RX_STAT_ADD(c, a)	__STAT_SAVE(hif_dev->htc_handle->drv_priv->debug.skbrx_stats.c += a)
+#define CAB_STAT_INC		(priv->debug.tx_stats.cab_queued++)
 
 #define TX_QSTAT_INC(q) (priv->debug.tx_stats.queue_stats[q]++)
 
drivers/net/wireless/ath/ath9k/htc_drv_init.c +2 -1
index ff61ae34ecdf..07ac88fb1c57 100644
@@ -944,7 +944,6 @@ int ath9k_htc_probe_device(struct htc_target *htc_handle, struct device *dev,
 	priv->hw = hw;
 	priv->htc = htc_handle;
 	priv->dev = dev;
-	htc_handle->drv_priv = priv;
 	SET_IEEE80211_DEV(hw, priv->dev);
 
 	ret = ath9k_htc_wait_for_target(priv);
@@ -965,6 +964,8 @@ int ath9k_htc_probe_device(struct htc_target *htc_handle, struct device *dev,
 	if (ret)
 		goto err_init;
 
+	htc_handle->drv_priv = priv;
+
 	return 0;
 
 err_init:
Signature · 1 line
2.32.0
#1 lore raw
Quoted 79 lines · Pavel Skripkin

Syzbot reported use-after-free Read in ath9k_hif_usb_rx_cb(). The problem was in incorrect htc_handle->drv_priv initialization.

Probable call trace which can trigger use-after-free:

ath9k_htc_probe_device()
/* htc_handle->drv_priv = priv; */
ath9k_htc_wait_for_target() <--- Failed
ieee80211_free_hw() <--- priv pointer is freed

<IRQ>
...
ath9k_hif_usb_rx_cb()
   ath9k_hif_usb_rx_stream()
    RX_STAT_INC()		<--- htc_handle->drv_priv access

In order to not add fancy protection for drv_priv we can move htc_handle->drv_priv initialization at the end of the ath9k_htc_probe_device() and add helper macro to make all *_STAT_* macros NULL save.

Also, I made whitespaces clean ups in *_STAT_* macros definitions to make checkpatch.pl happy.

Fixes fb9987d0f748 ("ath9k_htc: Support for AR9271 chipset.")
Reported-and-tested-by [email protected]

---

Hi, ath9k maintainer/developers!

I know, that you do not like changes, that wasn't tested on real hardware. I really don't access to this one, so I'd like you to test it on real hardware piece, if you have one. At least, this patch was tested by syzbot [1]

[1] https://syzkaller.appspot.com/bug?id=6ead44e37afb6866ac0c7dd121b4ce07cb665f60

---
  drivers/net/wireless/ath/ath9k/htc.h          | 11 ++++++-----
  drivers/net/wireless/ath/ath9k/htc_drv_init.c |  3 ++-
  2 files changed, 8 insertions(+), 6 deletions(-)
drivers/net/wireless/ath/ath9k/htc.h +6 -5
index 0a1634238e67..c16b2a482e83 100644
@@ -326,11 +326,12 @@ static inline struct ath9k_htc_tx_ctl *HTC_SKB_CB(struct sk_buff *skb)
 
  #ifdef CONFIG_ATH9K_HTC_DEBUGFS
 
-#define TX_STAT_INC(c) (hif_dev->htc_handle->drv_priv->debug.tx_stats.c++)
-#define TX_STAT_ADD(c, a) (hif_dev->htc_handle->drv_priv->debug.tx_stats.c += a)
-#define RX_STAT_INC(c) (hif_dev->htc_handle->drv_priv->debug.skbrx_stats.c++)
-#define RX_STAT_ADD(c, a) (hif_dev->htc_handle->drv_priv->debug.skbrx_stats.c += a)
-#define CAB_STAT_INC   priv->debug.tx_stats.cab_queued++
+#define __STAT_SAVE(expr)	(hif_dev->htc_handle->drv_priv ? (expr) : 0)
+#define TX_STAT_INC(c)		__STAT_SAVE(hif_dev->htc_handle->drv_priv->debug.tx_stats.c++)
+#define TX_STAT_ADD(c, a)	__STAT_SAVE(hif_dev->htc_handle->drv_priv->debug.tx_stats.c += a)
+#define RX_STAT_INC(c)		__STAT_SAVE(hif_dev->htc_handle->drv_priv->debug.skbrx_stats.c++)
+#define RX_STAT_ADD(c, a)	__STAT_SAVE(hif_dev->htc_handle->drv_priv->debug.skbrx_stats.c += a)
+#define CAB_STAT_INC		(priv->debug.tx_stats.cab_queued++)
 
  #define TX_QSTAT_INC(q) (priv->debug.tx_stats.queue_stats[q]++)
 
drivers/net/wireless/ath/ath9k/htc_drv_init.c +0 -1
index ff61ae34ecdf..07ac88fb1c57 100644
@@ -944,7 +944,6 @@ int ath9k_htc_probe_device(struct htc_target *htc_handle, struct device *dev,
  	priv->hw = hw;
  	priv->htc = htc_handle;
  	priv->dev = dev;
-	htc_handle->drv_priv = priv;
  	SET_IEEE80211_DEV(hw, priv->dev);
 
  	ret = ath9k_htc_wait_for_target(priv);

drivers/net/wireless/ath/ath9k/htc_drv_init.c int ath9k_htc_probe_device(struct htc_target *htc_handle, struct device *dev, · line 971

Quoted 8 lines
drivers/net/wireless/ath/ath9k/htc_drv_init.c +2 -0
@@ -965,6 +964,8 @@ int ath9k_htc_probe_device(struct htc_target *htc_handle, struct device *dev,
  	if (ret)
  		goto err_init;
 
+	htc_handle->drv_priv = priv;
+
  	return 0;
 
  err_init:

Added missing LKML in CC. Sorry for confusion.

With regards,
Pavel Skripkin

#2 lore raw
Quoted 82 lines · Pavel Skripkin
Quoted 81 lines · Pavel Skripkin

Syzbot reported use-after-free Read in ath9k_hif_usb_rx_cb(). The problem was in incorrect htc_handle->drv_priv initialization.

Probable call trace which can trigger use-after-free:

ath9k_htc_probe_device()
/* htc_handle->drv_priv = priv; */
ath9k_htc_wait_for_target() <--- Failed
ieee80211_free_hw() <--- priv pointer is freed

<IRQ>
...
ath9k_hif_usb_rx_cb()
   ath9k_hif_usb_rx_stream()
    RX_STAT_INC()		<--- htc_handle->drv_priv access

In order to not add fancy protection for drv_priv we can move htc_handle->drv_priv initialization at the end of the ath9k_htc_probe_device() and add helper macro to make all *_STAT_* macros NULL save.

Also, I made whitespaces clean ups in *_STAT_* macros definitions to make checkpatch.pl happy.

Fixes fb9987d0f748 ("ath9k_htc: Support for AR9271 chipset.")
Reported-and-tested-by [email protected]

---

Hi, ath9k maintainer/developers!

I know, that you do not like changes, that wasn't tested on real hardware. I really don't access to this one, so I'd like you to test it on real hardware piece, if you have one. At least, this patch was tested by syzbot [1]

[1] https://syzkaller.appspot.com/bug?id=6ead44e37afb6866ac0c7dd121b4ce07cb665f60

---
  drivers/net/wireless/ath/ath9k/htc.h          | 11 ++++++-----
  drivers/net/wireless/ath/ath9k/htc_drv_init.c |  3 ++-
  2 files changed, 8 insertions(+), 6 deletions(-)
drivers/net/wireless/ath/ath9k/htc.h +6 -5
index 0a1634238e67..c16b2a482e83 100644
@@ -326,11 +326,12 @@ static inline struct ath9k_htc_tx_ctl *HTC_SKB_CB(struct sk_buff *skb)
 
  #ifdef CONFIG_ATH9K_HTC_DEBUGFS
 
-#define TX_STAT_INC(c) (hif_dev->htc_handle->drv_priv->debug.tx_stats.c++)
-#define TX_STAT_ADD(c, a) (hif_dev->htc_handle->drv_priv->debug.tx_stats.c += a)
-#define RX_STAT_INC(c) (hif_dev->htc_handle->drv_priv->debug.skbrx_stats.c++)
-#define RX_STAT_ADD(c, a) (hif_dev->htc_handle->drv_priv->debug.skbrx_stats.c += a)
-#define CAB_STAT_INC   priv->debug.tx_stats.cab_queued++
+#define __STAT_SAVE(expr)	(hif_dev->htc_handle->drv_priv ? (expr) : 0)
+#define TX_STAT_INC(c)		__STAT_SAVE(hif_dev->htc_handle->drv_priv->debug.tx_stats.c++)
+#define TX_STAT_ADD(c, a)	__STAT_SAVE(hif_dev->htc_handle->drv_priv->debug.tx_stats.c += a)
+#define RX_STAT_INC(c)		__STAT_SAVE(hif_dev->htc_handle->drv_priv->debug.skbrx_stats.c++)
+#define RX_STAT_ADD(c, a)	__STAT_SAVE(hif_dev->htc_handle->drv_priv->debug.skbrx_stats.c += a)
+#define CAB_STAT_INC		(priv->debug.tx_stats.cab_queued++)
 
  #define TX_QSTAT_INC(q) (priv->debug.tx_stats.queue_stats[q]++)
 
drivers/net/wireless/ath/ath9k/htc_drv_init.c +0 -1
index ff61ae34ecdf..07ac88fb1c57 100644
@@ -944,7 +944,6 @@ int ath9k_htc_probe_device(struct htc_target *htc_handle, struct device *dev,
  	priv->hw = hw;
  	priv->htc = htc_handle;
  	priv->dev = dev;
-	htc_handle->drv_priv = priv;
  	SET_IEEE80211_DEV(hw, priv->dev);
 
  	ret = ath9k_htc_wait_for_target(priv);
@@ -965,6 +964,8 @@ int ath9k_htc_probe_device(struct htc_target *htc_handle, struct device *dev,
  	if (ret)
  		goto err_init;
 
Quoted 10 lines

drivers/net/wireless/ath/ath9k/htc_drv_init.c int ath9k_htc_probe_device(struct htc_target *htc_handle, struct device *dev, · line 971

Quoted 6 lines
drivers/net/wireless/ath/ath9k/htc_drv_init.c +2 -0
+	htc_handle->drv_priv = priv;
+
  	return 0;
 
  err_init:

Added missing LKML in CC. Sorry for confusion.

Any chance to review this patch? Thanks

With regards,
Pavel Skripkin

#3 lore raw
Quoted 87 lines · Pavel Skripkin
Quoted 86 lines · Pavel Skripkin
Quoted 85 lines · Pavel Skripkin

Syzbot reported use-after-free Read in ath9k_hif_usb_rx_cb(). The problem was in incorrect htc_handle->drv_priv initialization.

Probable call trace which can trigger use-after-free:

ath9k_htc_probe_device()
/* htc_handle->drv_priv = priv; */
ath9k_htc_wait_for_target() <--- Failed
ieee80211_free_hw() <--- priv pointer is freed

<IRQ>
...
ath9k_hif_usb_rx_cb()
   ath9k_hif_usb_rx_stream()
    RX_STAT_INC()		<--- htc_handle->drv_priv access

In order to not add fancy protection for drv_priv we can move htc_handle->drv_priv initialization at the end of the ath9k_htc_probe_device() and add helper macro to make all *_STAT_* macros NULL save.

Also, I made whitespaces clean ups in *_STAT_* macros definitions to make checkpatch.pl happy.

Fixes fb9987d0f748 ("ath9k_htc: Support for AR9271 chipset.")
Reported-and-tested-by [email protected]

---

Hi, ath9k maintainer/developers!

I know, that you do not like changes, that wasn't tested on real hardware. I really don't access to this one, so I'd like you to test it on real hardware piece, if you have one. At least, this patch was tested by syzbot [1]

[1] https://syzkaller.appspot.com/bug?id=6ead44e37afb6866ac0c7dd121b4ce07cb665f60

---
  drivers/net/wireless/ath/ath9k/htc.h          | 11 ++++++-----
  drivers/net/wireless/ath/ath9k/htc_drv_init.c |  3 ++-
  2 files changed, 8 insertions(+), 6 deletions(-)
drivers/net/wireless/ath/ath9k/htc.h +6 -5
index 0a1634238e67..c16b2a482e83 100644
@@ -326,11 +326,12 @@ static inline struct ath9k_htc_tx_ctl *HTC_SKB_CB(struct sk_buff *skb)
 
  #ifdef CONFIG_ATH9K_HTC_DEBUGFS
 
-#define TX_STAT_INC(c) (hif_dev->htc_handle->drv_priv->debug.tx_stats.c++)
-#define TX_STAT_ADD(c, a) (hif_dev->htc_handle->drv_priv->debug.tx_stats.c += a)
-#define RX_STAT_INC(c) (hif_dev->htc_handle->drv_priv->debug.skbrx_stats.c++)
-#define RX_STAT_ADD(c, a) (hif_dev->htc_handle->drv_priv->debug.skbrx_stats.c += a)
-#define CAB_STAT_INC   priv->debug.tx_stats.cab_queued++
+#define __STAT_SAVE(expr)	(hif_dev->htc_handle->drv_priv ? (expr) : 0)
+#define TX_STAT_INC(c)		__STAT_SAVE(hif_dev->htc_handle->drv_priv->debug.tx_stats.c++)
+#define TX_STAT_ADD(c, a)	__STAT_SAVE(hif_dev->htc_handle->drv_priv->debug.tx_stats.c += a)
+#define RX_STAT_INC(c)		__STAT_SAVE(hif_dev->htc_handle->drv_priv->debug.skbrx_stats.c++)
+#define RX_STAT_ADD(c, a)	__STAT_SAVE(hif_dev->htc_handle->drv_priv->debug.skbrx_stats.c += a)
+#define CAB_STAT_INC		(priv->debug.tx_stats.cab_queued++)
 
  #define TX_QSTAT_INC(q) (priv->debug.tx_stats.queue_stats[q]++)
 
drivers/net/wireless/ath/ath9k/htc_drv_init.c +2 -1
index ff61ae34ecdf..07ac88fb1c57 100644
@@ -944,7 +944,6 @@ int ath9k_htc_probe_device(struct htc_target *htc_handle, struct device *dev,
  	priv->hw = hw;
  	priv->htc = htc_handle;
  	priv->dev = dev;
-	htc_handle->drv_priv = priv;
  	SET_IEEE80211_DEV(hw, priv->dev);
 
  	ret = ath9k_htc_wait_for_target(priv);
@@ -965,6 +964,8 @@ int ath9k_htc_probe_device(struct htc_target *htc_handle, struct device *dev,
  	if (ret)
  		goto err_init;
 
+	htc_handle->drv_priv = priv;
+
  	return 0;
 
Quoted 9 lines
Quoted 6 lines

drivers/net/wireless/ath/ath9k/htc_drv_init.c int ath9k_htc_probe_device(struct htc_target *htc_handle, struct device *dev, · line 971

Quoted 2 lines
drivers/net/wireless/ath/ath9k/htc_drv_init.c +0 -0
  err_init:

Added missing LKML in CC. Sorry for confusion.

Any chance to review this patch? Thanks

gentle ping one more time :) Thanks

With regards,
Pavel Skripkin

#4 lore raw

PATCH RESEND 1/1 net: ath9k: fix use-after-free in ath9k_hif_usb_rx_cb

Syzbot reported use-after-free Read in ath9k_hif_usb_rx_cb(). The problem was in incorrect htc_handle->drv_priv initialization.

Probable call trace which can trigger use-after-free:

ath9k_htc_probe_device()
/* htc_handle->drv_priv = priv; */
ath9k_htc_wait_for_target() <--- Failed
ieee80211_free_hw() <--- priv pointer is freed

<IRQ>

...
ath9k_hif_usb_rx_cb()
ath9k_hif_usb_rx_stream()
RX_STAT_INC() <--- htc_handle->drv_priv access

In order to not add fancy protection for drv_priv we can move htc_handle->drv_priv initialization at the end of the ath9k_htc_probe_device() and add helper macro to make all *_STAT_* macros NULL save.

Fixes fb9987d0f748 ("ath9k_htc: Support for AR9271 chipset.")
Reported-and-tested-by [email protected]
Reported-and-tested-by [email protected]

---

Why resend?
No activity on this patch since 8/6/21, Kalle Valo has asked around, for review and no one claimed it.

Resend changes:
1. Rebased on top of v5.15-rc2
2. Removed clean ups for macros
3. Added 1 more syzbot tag, since this patch has passed 2 syzbot tests

---
 drivers/net/wireless/ath/ath9k/htc.h          | 10 +++++-----
 drivers/net/wireless/ath/ath9k/htc_drv_init.c |  3 ++-
 2 files changed, 7 insertions(+), 6 deletions(-)
drivers/net/wireless/ath/ath9k/htc.h +5 -5
index 0a1634238e67..c8cde2e05d77 100644
@@ -325,11 +325,11 @@ static inline struct ath9k_htc_tx_ctl *HTC_SKB_CB(struct sk_buff *skb)
 }
 
 #ifdef CONFIG_ATH9K_HTC_DEBUGFS
-
-#define TX_STAT_INC(c) (hif_dev->htc_handle->drv_priv->debug.tx_stats.c++)
-#define TX_STAT_ADD(c, a) (hif_dev->htc_handle->drv_priv->debug.tx_stats.c += a)
-#define RX_STAT_INC(c) (hif_dev->htc_handle->drv_priv->debug.skbrx_stats.c++)
-#define RX_STAT_ADD(c, a) (hif_dev->htc_handle->drv_priv->debug.skbrx_stats.c += a)
+#define __STAT_SAVE(expr) (hif_dev->htc_handle->drv_priv ? (expr) : 0)
+#define TX_STAT_INC(c) __STAT_SAVE(hif_dev->htc_handle->drv_priv->debug.tx_stats.c++)
+#define TX_STAT_ADD(c, a) __STAT_SAVE(hif_dev->htc_handle->drv_priv->debug.tx_stats.c += a)
+#define RX_STAT_INC(c) __STAT_SAVE(hif_dev->htc_handle->drv_priv->debug.skbrx_stats.c++)
+#define RX_STAT_ADD(c, a) __STAT_SAVE(hif_dev->htc_handle->drv_priv->debug.skbrx_stats.c += a)
 #define CAB_STAT_INC   priv->debug.tx_stats.cab_queued++
 
 #define TX_QSTAT_INC(q) (priv->debug.tx_stats.queue_stats[q]++)
drivers/net/wireless/ath/ath9k/htc_drv_init.c +2 -1
index ff61ae34ecdf..07ac88fb1c57 100644
@@ -944,7 +944,6 @@ int ath9k_htc_probe_device(struct htc_target *htc_handle, struct device *dev,
 	priv->hw = hw;
 	priv->htc = htc_handle;
 	priv->dev = dev;
-	htc_handle->drv_priv = priv;
 	SET_IEEE80211_DEV(hw, priv->dev);
 
 	ret = ath9k_htc_wait_for_target(priv);
@@ -965,6 +964,8 @@ int ath9k_htc_probe_device(struct htc_target *htc_handle, struct device *dev,
 	if (ret)
 		goto err_init;
 
+	htc_handle->drv_priv = priv;
+
 	return 0;
 
 err_init:
Signature · 1 line
2.33.0
#5 lore raw
Quoted 31 lines · Pavel Skripkin

Syzbot reported use-after-free Read in ath9k_hif_usb_rx_cb(). The problem was in incorrect htc_handle->drv_priv initialization.

Probable call trace which can trigger use-after-free:

ath9k_htc_probe_device()
/* htc_handle->drv_priv = priv; */
ath9k_htc_wait_for_target() <--- Failed
ieee80211_free_hw() <--- priv pointer is freed

<IRQ>
...
ath9k_hif_usb_rx_cb()
   ath9k_hif_usb_rx_stream()
    RX_STAT_INC()		<--- htc_handle->drv_priv access

In order to not add fancy protection for drv_priv we can move htc_handle->drv_priv initialization at the end of the ath9k_htc_probe_device() and add helper macro to make all *_STAT_* macros NULL save.

Fixes fb9987d0f748 ("ath9k_htc: Support for AR9271 chipset.")
Reported-and-tested-by [email protected]
Reported-and-tested-by [email protected]

---

Why resend?
No activity on this patch since 8/6/21, Kalle Valo has asked around, for review and no one claimed it.

Quoted 6 lines

Resend changes:
1. Rebased on top of v5.15-rc2
2. Removed clean ups for macros
3. Added 1 more syzbot tag, since this patch has passed 2 syzbot tests

Hi, ath9k maintainers!

Does this patch need any further work? I can't see any comments on it since 8/6/21 and I can't see it on wireless patchwork.

If this bug is already fixed and I've overlooked a fix commit, please, let me know. As I see syzbot hits this bug really often [1]

Thanks

[1] https://syzkaller.appspot.com/bug?id=6ead44e37afb6866ac0c7dd121b4ce07cb665f60

With regards,
Pavel Skripkin

#6 lore raw
Quoted 39 lines · Pavel Skripkin
Quoted 31 lines · Pavel Skripkin

Syzbot reported use-after-free Read in ath9k_hif_usb_rx_cb(). The problem was in incorrect htc_handle->drv_priv initialization.

Probable call trace which can trigger use-after-free:

ath9k_htc_probe_device()
/* htc_handle->drv_priv = priv; */
ath9k_htc_wait_for_target() <--- Failed
ieee80211_free_hw() <--- priv pointer is freed

<IRQ>
...
ath9k_hif_usb_rx_cb()
   ath9k_hif_usb_rx_stream()
    RX_STAT_INC()		<--- htc_handle->drv_priv access

In order to not add fancy protection for drv_priv we can move htc_handle->drv_priv initialization at the end of the ath9k_htc_probe_device() and add helper macro to make all *_STAT_* macros NULL save.

Fixes fb9987d0f748 ("ath9k_htc: Support for AR9271 chipset.")
Reported-and-tested-by [email protected]
Reported-and-tested-by [email protected]

---

Why resend?
No activity on this patch since 8/6/21, Kalle Valo has asked around, for review and no one claimed it.

Quoted 6 lines

Resend changes:
1. Rebased on top of v5.15-rc2
2. Removed clean ups for macros
3. Added 1 more syzbot tag, since this patch has passed 2 syzbot tests

Quoted 7 lines

Hi, ath9k maintainers!

Does this patch need any further work? I can't see any comments on it since 8/6/21 and I can't see it on wireless patchwork.

If this bug is already fixed and I've overlooked a fix commit, please, let me know. As I see syzbot hits this bug really often [1]

See my other mail I just sent about ath9k syzbot patches:

https://lore.kernel.org/linux-wireless/[email protected]/

In summary: please wait patiently once I'm able to test the syzbot patches on a real device.

Signature · 3 lines
https://patchwork.kernel.org/project/linux-wireless/list/

https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches
#7 lore raw
Quoted 8 lines · Kalle Valo
Quoted 7 lines

Hi, ath9k maintainers!

Does this patch need any further work? I can't see any comments on it since 8/6/21 and I can't see it on wireless patchwork.

If this bug is already fixed and I've overlooked a fix commit, please, let me know. As I see syzbot hits this bug really often [1]

Quoted 7 lines

See my other mail I just sent about ath9k syzbot patches:

https://lore.kernel.org/linux-wireless/[email protected]/

In summary: please wait patiently once I'm able to test the syzbot patches on a real device.

I see, thank you clarifying the situation. I understand, that maintaining ath9k devices alone is not easy task

Thanks again

With regards,
Pavel Skripkin

#8 lore raw
In reply to Pavel Skripkin
Quoted 16 lines · Pavel Skripkin

Syzbot reported use-after-free Read in ath9k_hif_usb_rx_cb(). The problem was in incorrect htc_handle->drv_priv initialization.

Probable call trace which can trigger use-after-free:

ath9k_htc_probe_device()
/* htc_handle->drv_priv = priv; */
ath9k_htc_wait_for_target() <--- Failed
ieee80211_free_hw() <--- priv pointer is freed

<IRQ>

...
ath9k_hif_usb_rx_cb()
ath9k_hif_usb_rx_stream()
RX_STAT_INC() <--- htc_handle->drv_priv access

Quoted 7 lines

In order to not add fancy protection for drv_priv we can move htc_handle->drv_priv initialization at the end of the ath9k_htc_probe_device() and add helper macro to make all *_STAT_* macros NULL save.

Also, I made whitespaces clean ups in *_STAT_* macros definitions to make checkpatch.pl happy.

Separate patch for cleanups, please.

Quoted 5 lines
Fixes fb9987d0f748 ("ath9k_htc: Support for AR9271 chipset.")
Reported-and-tested-by [email protected]

---

Quoted 8 lines

Hi, ath9k maintainer/developers!

I know, that you do not like changes, that wasn't tested on real hardware. I really don't access to this one, so I'd like you to test it on real hardware piece, if you have one. At least, this patch was tested by syzbot [1]

[1] https://syzkaller.appspot.com/bug?id=6ead44e37afb6866ac0c7dd121b4ce07cb665f60

syzbot does not equal testing on real hardware. Can someone test or review this, please?

Signature · 3 lines
https://patchwork.kernel.org/project/linux-wireless/list/

https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches
#9 lore raw
In reply to Pavel Skripkin
Quoted 31 lines · Pavel Skripkin

Syzbot reported use-after-free Read in ath9k_hif_usb_rx_cb(). The problem was in incorrect htc_handle->drv_priv initialization.

Probable call trace which can trigger use-after-free:

ath9k_htc_probe_device()
/* htc_handle->drv_priv = priv; */
ath9k_htc_wait_for_target() <--- Failed
ieee80211_free_hw() <--- priv pointer is freed

<IRQ>
...
ath9k_hif_usb_rx_cb()
   ath9k_hif_usb_rx_stream()
    RX_STAT_INC()		<--- htc_handle->drv_priv access

In order to not add fancy protection for drv_priv we can move htc_handle->drv_priv initialization at the end of the ath9k_htc_probe_device() and add helper macro to make all *_STAT_* macros NULL save.

Also, I made whitespaces clean ups in *_STAT_* macros definitions to make checkpatch.pl happy.

Fixes fb9987d0f748 ("ath9k_htc: Support for AR9271 chipset.")
Reported-and-tested-by [email protected]

---

Hi, ath9k maintainer/developers!

Quoted 8 lines

I know, that you do not like changes, that wasn't tested on real hardware. I really don't access to this one, so I'd like you to test it on real hardware piece, if you have one. At least, this patch was tested by syzbot [1]

[1] https://syzkaller.appspot.com/bug?id=6ead44e37afb6866ac0c7dd121b4ce07cb665f60

---

Btw, this patch also passes this syzbot test

https://syzkaller.appspot.com/bug?id=b8101ffcec107c0567a0cd8acbbacec91e9ee8de

With regards,
Pavel Skripkin