Skip to contentSkip to conversation

wifi: rtw88: phy: fix warning of possible buffer overflow

PATCH In review Sasha Levin 4 participants 41 messages Oct 9, 2022 – Oct 18, 2022

Subsystems (8)
lore
Share & tools
lore

Sasha Levin

author maintainer
#0 lore raw

PATCH 1/34 wifi: rtw88: phy: fix warning of possible buffer overflow

From: Zong-Zhe Yang <[email protected]>

[ Upstream commit 86331c7e0cd819bf0c1d0dcf895e0c90b0aa9a6f ]

reported by smatch

phy.c:854 rtw_phy_linear_2_db() error: buffer overflow 'db_invert_table[i]' 8 <= 8 (assuming for loop doesn't break)

However, it seems to be a false alarm because we prevent it originally via

       if (linear >= db_invert_table[11][7])
               return 96; /* maximum 96 dB */

Still, we adjust the code to be more readable and avoid smatch warning.

---
 drivers/net/wireless/realtek/rtw88/phy.c | 21 ++++++++-------------
 1 file changed, 8 insertions(+), 13 deletions(-)
drivers/net/wireless/realtek/rtw88/phy.c +8 -13
index af8b703d11d4..0fc5a893c395 100644
@@ -604,23 +604,18 @@ static u8 rtw_phy_linear_2_db(u64 linear)
 	u8 j;
 	u32 dB;
 
-	if (linear >= db_invert_table[11][7])
-		return 96; /* maximum 96 dB */
-
 	for (i = 0; i < 12; i++) {
-		if (i <= 2 && (linear << FRAC_BITS) <= db_invert_table[i][7])
-			break;
-		else if (i > 2 && linear <= db_invert_table[i][7])
-			break;
+		for (j = 0; j < 8; j++) {
+			if (i <= 2 && (linear << FRAC_BITS) <= db_invert_table[i][j])
+				goto cnt;
+			else if (i > 2 && linear <= db_invert_table[i][j])
+				goto cnt;
+		}
 	}
 
-	for (j = 0; j < 8; j++) {
-		if (i <= 2 && (linear << FRAC_BITS) <= db_invert_table[i][j])
-			break;
-		else if (i > 2 && linear <= db_invert_table[i][j])
-			break;
-	}
+	return 96; /* maximum 96 dB */
 
+cnt:
 	if (j == 0 && i == 0)
 		goto end;
 
Signature · 1 line
2.35.1
author maintainer
#1 lore raw

PATCH 2/34 wifi: brcmfmac: fix invalid address access when enabling SCAN log level

From: Wright Feng <[email protected]>

[ Upstream commit aa666b68e73fc06d83c070d96180b9010cf5a960 ]

The variable i is changed when setting random MAC address and causes invalid address access when printing the value of pi->reqs[i]->reqid.

We replace reqs index with ri to fix the issue.

[  136.726473] Unable to handle kernel access to user memory outside uaccess routines at virtual address 0000000000000000
[  136.737365] Mem abort info:
[  136.740172]   ESR = 0x96000004
[  136.743359]   Exception class = DABT (current EL), IL = 32 bits
[  136.749294]   SET = 0, FnV = 0
[  136.752481]   EA = 0, S1PTW = 0
[  136.755635] Data abort info:
[  136.758514]   ISV = 0, ISS = 0x00000004
[  136.762487]   CM = 0, WnR = 0
[  136.765522] user pgtable: 4k pages, 48-bit VAs, pgdp = 000000005c4e2577
[  136.772265] [0000000000000000] pgd=0000000000000000
[  136.777160] Internal error: Oops: 96000004 [#1] PREEMPT SMP
[  136.782732] Modules linked in: brcmfmac(O) brcmutil(O) cfg80211(O) compat(O)
[  136.789788] Process wificond (pid: 3175, stack limit = 0x00000000053048fb)
[  136.796664] CPU: 3 PID: 3175 Comm: wificond Tainted: G           O      4.19.42-00001-g531a5f5 #1
[  136.805532] Hardware name: Freescale i.MX8MQ EVK (DT)
[  136.810584] pstate: 60400005 (nZCv daif +PAN -UAO)
[  136.815429] pc : brcmf_pno_config_sched_scans+0x6cc/0xa80 [brcmfmac]
[  136.821811] lr : brcmf_pno_config_sched_scans+0x67c/0xa80 [brcmfmac]
[  136.828162] sp : ffff00000e9a3880
[  136.831475] x29: ffff00000e9a3890 x28: ffff800020543400
[  136.836786] x27: ffff8000b1008880 x26: ffff0000012bf6a0
[  136.842098] x25: ffff80002054345c x24: ffff800088d22400
[  136.847409] x23: ffff0000012bf638 x22: ffff0000012bf6d8
[  136.852721] x21: ffff8000aced8fc0 x20: ffff8000ac164400
[  136.858032] x19: ffff00000e9a3946 x18: 0000000000000000
[  136.863343] x17: 0000000000000000 x16: 0000000000000000
[  136.868655] x15: ffff0000093f3b37 x14: 0000000000000050
[  136.873966] x13: 0000000000003135 x12: 0000000000000000
[  136.879277] x11: 0000000000000000 x10: ffff000009a61888
[  136.884589] x9 : 000000000000000f x8 : 0000000000000008
[  136.889900] x7 : 303a32303d726464 x6 : ffff00000a1f957d
[  136.895211] x5 : 0000000000000000 x4 : ffff00000e9a3942
[  136.900523] x3 : 0000000000000000 x2 : ffff0000012cead8
[  136.905834] x1 : ffff0000012bf6d8 x0 : 0000000000000000
[  136.911146] Call trace:
[  136.913623]  brcmf_pno_config_sched_scans+0x6cc/0xa80 [brcmfmac]
[  136.919658]  brcmf_pno_start_sched_scan+0xa4/0x118 [brcmfmac]
[  136.925430]  brcmf_cfg80211_sched_scan_start+0x80/0xe0 [brcmfmac]
[  136.931636]  nl80211_start_sched_scan+0x140/0x308 [cfg80211]
[  136.937298]  genl_rcv_msg+0x358/0x3f4
[  136.940960]  netlink_rcv_skb+0xb4/0x118
[  136.944795]  genl_rcv+0x34/0x48
[  136.947935]  netlink_unicast+0x264/0x300
[  136.951856]  netlink_sendmsg+0x2e4/0x33c
[  136.955781]  __sys_sendto+0x120/0x19c
---
 .../net/wireless/broadcom/brcm80211/brcmfmac/pno.c   | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)
drivers/net/wireless/broadcom/brcm80211/brcmfmac/pno.c +6 -6
index fabfbb0b40b0..d0a7465be586 100644
@@ -158,12 +158,12 @@ static int brcmf_pno_set_random(struct brcmf_if *ifp, struct brcmf_pno_info *pi)
 	struct brcmf_pno_macaddr_le pfn_mac;
 	u8 *mac_addr = NULL;
 	u8 *mac_mask = NULL;
-	int err, i;
+	int err, i, ri;
 
-	for (i = 0; i < pi->n_reqs; i++)
-		if (pi->reqs[i]->flags & NL80211_SCAN_FLAG_RANDOM_ADDR) {
-			mac_addr = pi->reqs[i]->mac_addr;
-			mac_mask = pi->reqs[i]->mac_addr_mask;
+	for (ri = 0; ri < pi->n_reqs; ri++)
+		if (pi->reqs[ri]->flags & NL80211_SCAN_FLAG_RANDOM_ADDR) {
+			mac_addr = pi->reqs[ri]->mac_addr;
+			mac_mask = pi->reqs[ri]->mac_addr_mask;
 			break;
 		}
 
@@ -185,7 +185,7 @@ static int brcmf_pno_set_random(struct brcmf_if *ifp, struct brcmf_pno_info *pi)
 	pfn_mac.mac[0] |= 0x02;
 
 	brcmf_dbg(SCAN, "enabling random mac: reqid=%llu mac=%pM\n",
-		  pi->reqs[i]->reqid, pfn_mac.mac);
+		  pi->reqs[ri]->reqid, pfn_mac.mac);
 	err = brcmf_fil_iovar_data_set(ifp, "pfn_macaddr", &pfn_mac,
 				       sizeof(pfn_mac));
 	if (err)
Signature · 1 line
2.35.1
author maintainer
#2 lore raw

PATCH 3/34 bpftool: Clear errno after libcap's checks

In reply to Sasha Levin

From: Quentin Monnet <[email protected]>

[ Upstream commit cea558855c39b7f1f02ff50dcf701ca6596bc964 ]

When bpftool is linked against libcap, the library runs a "constructor" function to compute the number of capabilities of the running kernel [0], at the beginning of the execution of the program. As part of this, it performs multiple calls to prctl(). Some of these may fail, and set errno to a non-zero value:

    # strace -e prctl ./bpftool version
    prctl(PR_CAPBSET_READ, CAP_MAC_OVERRIDE) = 1
    prctl(PR_CAPBSET_READ, 0x30 /* CAP_??? */) = -1 EINVAL (Invalid argument)
    prctl(PR_CAPBSET_READ, CAP_CHECKPOINT_RESTORE) = 1
    prctl(PR_CAPBSET_READ, 0x2c /* CAP_??? */) = -1 EINVAL (Invalid argument)
    prctl(PR_CAPBSET_READ, 0x2a /* CAP_??? */) = -1 EINVAL (Invalid argument)
    prctl(PR_CAPBSET_READ, 0x29 /* CAP_??? */) = -1 EINVAL (Invalid argument)
    ** fprintf added at the top of main(): we have errno == 1
    ./bpftool v7.0.0
    using libbpf v1.0
    features: libbfd, libbpf_strict, skeletons
    +++ exited with 0 +++

This has been addressed in libcap 2.63 [1], but until this version is available everywhere, we can fix it on bpftool side.

Let's clean errno at the beginning of the main() function, to make sure that these checks do not interfere with the batch mode, where we error out if errno is set after a bpftool command.

[0] https://git.kernel.org/pub/scm/libs/libcap/libcap.git/tree/libcap/cap_alloc.c?h=libcap-2.65#n20
[1] https://git.kernel.org/pub/scm/libs/libcap/libcap.git/commit/?id=f25a1b7e69f7b33e6afb58b3e38f3450b7d2d9a0

---
 tools/bpf/bpftool/main.c | 10 ++++++++++
 1 file changed, 10 insertions(+)
tools/bpf/bpftool/main.c +10 -0
index 1854d6b97860..4fd4e3462ebc 100644
@@ -398,6 +398,16 @@ int main(int argc, char **argv)
 
 	setlinebuf(stdout);
 
+#ifdef USE_LIBCAP
+	/* Libcap < 2.63 hooks before main() to compute the number of
+	 * capabilities of the running kernel, and doing so it calls prctl()
+	 * which may fail and set errno to non-zero.
+	 * Let's reset errno to make sure this does not interfere with the
+	 * batch mode.
+	 */
+	errno = 0;
+#endif
+
 	last_do_help = do_help;
 	pretty_output = false;
 	json_output = false;
Signature · 1 line
2.35.1
author maintainer
#3 lore raw

PATCH 4/34 openvswitch: Fix double reporting of drops in dropwatch

In reply to Sasha Levin

From: Mike Pattrick <[email protected]>

[ Upstream commit 1100248a5c5ccd57059eb8d02ec077e839a23826 ]

Frames sent to userspace can be reported as dropped in ovs_dp_process_packet, however, if they are dropped in the netlink code then netlink_attachskb will report the same frame as dropped.

This patch checks for error codes which indicate that the frame has already been freed.

---
 net/openvswitch/datapath.c | 13 ++++++++++---
 1 file changed, 10 insertions(+), 3 deletions(-)
net/openvswitch/datapath.c +10 -3
index 9d6ef6cb9b26..4d2d91d6f990 100644
@@ -241,10 +241,17 @@ void ovs_dp_process_packet(struct sk_buff *skb, struct sw_flow_key *key)
 		upcall.portid = ovs_vport_find_upcall_portid(p, skb);
 		upcall.mru = OVS_CB(skb)->mru;
 		error = ovs_dp_upcall(dp, skb, key, &upcall, 0);
-		if (unlikely(error))
-			kfree_skb(skb);
-		else
+		switch (error) {
+		case 0:
+		case -EAGAIN:
+		case -ERESTARTSYS:
+		case -EINTR:
 			consume_skb(skb);
+			break;
+		default:
+			kfree_skb(skb);
+			break;
+		}
 		stats_counter = &stats->n_missed;
 		goto out;
 	}
Signature · 1 line
2.35.1
author maintainer
#4 lore raw

PATCH 5/34 openvswitch: Fix overreporting of drops in dropwatch

In reply to Sasha Levin

From: Mike Pattrick <[email protected]>

[ Upstream commit c21ab2afa2c64896a7f0e3cbc6845ec63dcfad2e ]

Currently queue_userspace_packet will call kfree_skb for all frames, whether or not an error occurred. This can result in a single dropped frame being reported as multiple drops in dropwatch. This functions caller may also call kfree_skb in case of an error. This patch will consume the skbs instead and allow caller's to use kfree_skb.

---
 net/openvswitch/datapath.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)
net/openvswitch/datapath.c +3 -2
index 4d2d91d6f990..6b5c0abf7f1b 100644
@@ -544,8 +544,9 @@ static int queue_userspace_packet(struct datapath *dp, struct sk_buff *skb,
 out:
 	if (err)
 		skb_tx_error(skb);
-	kfree_skb(user_skb);
-	kfree_skb(nskb);
+	consume_skb(user_skb);
+	consume_skb(nskb);
+
 	return err;
 }
 
Signature · 1 line
2.35.1
author maintainer
#5 lore raw

PATCH 6/34 tcp: annotate data-race around tcp_md5sig_pool_populated

In reply to Sasha Levin

From: Eric Dumazet <[email protected]>

[ Upstream commit aacd467c0a576e5e44d2de4205855dc0fe43f6fb ]

tcp_md5sig_pool_populated can be read while another thread changes its value.

The race has no consequence because allocations are protected with tcp_md5sig_mutex.

This patch adds READ_ONCE() and WRITE_ONCE() to document the race and silence KCSAN.

---
 net/ipv4/tcp.c | 14 ++++++++++----
 1 file changed, 10 insertions(+), 4 deletions(-)
net/ipv4/tcp.c +10 -4
index bfeb05f62b94..8b527241af8f 100644
@@ -4041,12 +4041,16 @@ static void __tcp_alloc_md5sig_pool(void)
 	 * to memory. See smp_rmb() in tcp_get_md5sig_pool()
 	 */
 	smp_wmb();
-	tcp_md5sig_pool_populated = true;
+	/* Paired with READ_ONCE() from tcp_alloc_md5sig_pool()
+	 * and tcp_get_md5sig_pool().
+	*/
+	WRITE_ONCE(tcp_md5sig_pool_populated, true);
 }
 
 bool tcp_alloc_md5sig_pool(void)
 {
-	if (unlikely(!tcp_md5sig_pool_populated)) {
+	/* Paired with WRITE_ONCE() from __tcp_alloc_md5sig_pool() */
+	if (unlikely(!READ_ONCE(tcp_md5sig_pool_populated))) {
 		mutex_lock(&tcp_md5sig_mutex);
 
 		if (!tcp_md5sig_pool_populated) {
@@ -4057,7 +4061,8 @@ bool tcp_alloc_md5sig_pool(void)
 
 		mutex_unlock(&tcp_md5sig_mutex);
 	}
-	return tcp_md5sig_pool_populated;
+	/* Paired with WRITE_ONCE() from __tcp_alloc_md5sig_pool() */
+	return READ_ONCE(tcp_md5sig_pool_populated);
 }
 EXPORT_SYMBOL(tcp_alloc_md5sig_pool);
 
@@ -4073,7 +4078,8 @@ struct tcp_md5sig_pool *tcp_get_md5sig_pool(void)
 {
 	local_bh_disable();
 
-	if (tcp_md5sig_pool_populated) {
+	/* Paired with WRITE_ONCE() from __tcp_alloc_md5sig_pool() */
+	if (READ_ONCE(tcp_md5sig_pool_populated)) {
 		/* coupled with smp_wmb() in __tcp_alloc_md5sig_pool() */
 		smp_rmb();
 		return this_cpu_ptr(&tcp_md5sig_pool);
Signature · 1 line
2.35.1
author maintainer
#6 lore raw

PATCH 7/34 micrel: ksz8851: fixes struct pointer issue

In reply to Sasha Levin

From: Jerry Ray <[email protected]>

[ Upstream commit fef5de753ff01887cfa50990532c3890fccb9338 ]

Issue found during code review. This bug has no impact as long as the ks8851_net structure is the first element of the ks8851_net_spi structure. As long as the offset to the ks8851_net struct is zero, the container_of() macro is subtracting 0 and therefore no damage done. But if the ks8851_net_spi struct is ever modified such that the ks8851_net struct within it is no longer the first element of the struct, then the bug would manifest itself and cause problems.

struct ks8851_net is contained within ks8851_net_spi. ks is contained within kss.
kss is the priv_data of the netdev structure.

---
 drivers/net/ethernet/micrel/ks8851_spi.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)
drivers/net/ethernet/micrel/ks8851_spi.c +2 -3
index 4ec7f1615977..8327e7f30476 100644
@@ -415,7 +415,8 @@ static int ks8851_probe_spi(struct spi_device *spi)
 
 	spi->bits_per_word = 8;
 
-	ks = netdev_priv(netdev);
+	kss = netdev_priv(netdev);
+	ks = &kss->ks8851;
 
 	ks->lock = ks8851_lock_spi;
 	ks->unlock = ks8851_unlock_spi;
@@ -435,8 +436,6 @@ static int ks8851_probe_spi(struct spi_device *spi)
 		 IRQ_RXPSI)	/* RX process stop */
 	ks->rc_ier = STD_IRQ;
 
-	kss = to_ks8851_spi(ks);
-
 	kss->spidev = spi;
 	mutex_init(&kss->lock);
 	INIT_WORK(&kss->tx_work, ks8851_tx_work);
Signature · 1 line
2.35.1
author maintainer
#7 lore raw

PATCH 8/34 wifi: ath9k: avoid uninit memory read in ath9k_htc_rx_msg()

In reply to Sasha Levin

From: Tetsuo Handa <[email protected]>

[ Upstream commit b383e8abed41cc6ff1a3b34de75df9397fa4878c ]

syzbot is reporting uninit value at ath9k_htc_rx_msg() [1], for ioctl(USB_RAW_IOCTL_EP_WRITE) can call ath9k_hif_usb_rx_stream() with pkt_len = 0 but ath9k_hif_usb_rx_stream() uses __dev_alloc_skb(pkt_len + 32, GFP_ATOMIC) based on an assumption that pkt_len is valid. As a result, ath9k_hif_usb_rx_stream() allocates skb with uninitialized memory and ath9k_htc_rx_msg() is reading from uninitialized memory.

Since bytes accessed by ath9k_htc_rx_msg() is not known until ath9k_htc_rx_msg() is called, it would be difficult to check minimal valid pkt_len at "if (pkt_len > 2 * MAX_RX_BUF_SIZE) {" line in ath9k_hif_usb_rx_stream().

We have two choices. One is to workaround by adding __GFP_ZERO so that ath9k_htc_rx_msg() sees 0 if pkt_len is invalid. The other is to let ath9k_htc_rx_msg() validate pkt_len before accessing. This patch chose the latter.

Note that I'm not sure threshold condition is correct, for I can't find details on possible packet length used by this protocol.

---
 drivers/net/wireless/ath/ath9k/htc_hst.c | 43 +++++++++++++++---------
 1 file changed, 28 insertions(+), 15 deletions(-)
drivers/net/wireless/ath/ath9k/htc_hst.c +28 -15
index 994ec48b2f66..ca05b07a45e6 100644
@@ -364,33 +364,27 @@ void ath9k_htc_txcompletion_cb(struct htc_target *htc_handle,
 }
 
 static void ath9k_htc_fw_panic_report(struct htc_target *htc_handle,
-				      struct sk_buff *skb)
+				      struct sk_buff *skb, u32 len)
 {
 	uint32_t *pattern = (uint32_t *)skb->data;
 
-	switch (*pattern) {
-	case 0x33221199:
-		{
+	if (*pattern == 0x33221199 && len >= sizeof(struct htc_panic_bad_vaddr)) {
 		struct htc_panic_bad_vaddr *htc_panic;
 		htc_panic = (struct htc_panic_bad_vaddr *) skb->data;
 		dev_err(htc_handle->dev, "ath: firmware panic! "
 			"exccause: 0x%08x; pc: 0x%08x; badvaddr: 0x%08x.\n",
 			htc_panic->exccause, htc_panic->pc,
 			htc_panic->badvaddr);
-		break;
-		}
-	case 0x33221299:
-		{
+		return;
+	}
+	if (*pattern == 0x33221299) {
 		struct htc_panic_bad_epid *htc_panic;
 		htc_panic = (struct htc_panic_bad_epid *) skb->data;
 		dev_err(htc_handle->dev, "ath: firmware panic! "
 			"bad epid: 0x%08x\n", htc_panic->epid);
-		break;
-		}
-	default:
-		dev_err(htc_handle->dev, "ath: unknown panic pattern!\n");
-		break;
+		return;
 	}
+	dev_err(htc_handle->dev, "ath: unknown panic pattern!\n");
 }
 
 /*
@@ -411,16 +405,26 @@ void ath9k_htc_rx_msg(struct htc_target *htc_handle,
 	if (!htc_handle || !skb)
 		return;
 
+	/* A valid message requires len >= 8.
+	 *
+	 *   sizeof(struct htc_frame_hdr) == 8
+	 *   sizeof(struct htc_ready_msg) == 8
+	 *   sizeof(struct htc_panic_bad_vaddr) == 16
+	 *   sizeof(struct htc_panic_bad_epid) == 8
+	 */
+	if (unlikely(len < sizeof(struct htc_frame_hdr)))
+		goto invalid;
 	htc_hdr = (struct htc_frame_hdr *) skb->data;
 	epid = htc_hdr->endpoint_id;
 
 	if (epid == 0x99) {
-		ath9k_htc_fw_panic_report(htc_handle, skb);
+		ath9k_htc_fw_panic_report(htc_handle, skb, len);
 		kfree_skb(skb);
 		return;
 	}
 
 	if (epid < 0 || epid >= ENDPOINT_MAX) {
+invalid:
 		if (pipe_id != USB_REG_IN_PIPE)
 			dev_kfree_skb_any(skb);
 		else
@@ -432,21 +436,30 @@ void ath9k_htc_rx_msg(struct htc_target *htc_handle,
 
 		/* Handle trailer */
 		if (htc_hdr->flags & HTC_FLAGS_RECV_TRAILER) {
-			if (be32_to_cpu(*(__be32 *) skb->data) == 0x00C60000)
+			if (be32_to_cpu(*(__be32 *) skb->data) == 0x00C60000) {
 				/* Move past the Watchdog pattern */
 				htc_hdr = (struct htc_frame_hdr *)(skb->data + 4);
+				len -= 4;
+			}
 		}
 
 		/* Get the message ID */
+		if (unlikely(len < sizeof(struct htc_frame_hdr) + sizeof(__be16)))
+			goto invalid;
 		msg_id = (__be16 *) ((void *) htc_hdr +
 				     sizeof(struct htc_frame_hdr));
 
 		/* Now process HTC messages */
 		switch (be16_to_cpu(*msg_id)) {
 		case HTC_MSG_READY_ID:
+			if (unlikely(len < sizeof(struct htc_ready_msg)))
+				goto invalid;
 			htc_process_target_rdy(htc_handle, htc_hdr);
 			break;
 		case HTC_MSG_CONNECT_SERVICE_RESPONSE_ID:
+			if (unlikely(len < sizeof(struct htc_frame_hdr) +
+				     sizeof(struct htc_conn_svc_rspmsg)))
+				goto invalid;
 			htc_process_conn_rsp(htc_handle, htc_hdr);
 			break;
 		default:
Signature · 1 line
2.35.1
author maintainer
#8 lore raw

PATCH 9/34 xfrm: Update ipcomp_scratches with NULL when freed

In reply to Sasha Levin

From: Khalid Masum <[email protected]>

[ Upstream commit 8a04d2fc700f717104bfb95b0f6694e448a4537f ]

Currently if ipcomp_alloc_scratches() fails to allocate memory ipcomp_scratches holds obsolete address. So when we try to free the percpu scratches using ipcomp_free_scratches() it tries to vfree non existent vm area. Described below:

static void * __percpu *ipcomp_alloc_scratches(void)

{
        ...
        scratches = alloc_percpu(void *);
        if (!scratches)
                return NULL;
ipcomp_scratches does not know about this allocation failure.
Therefore holding the old obsolete address.
        ...
}

So when we free,

static void ipcomp_free_scratches(void)

{
        ...
        scratches = ipcomp_scratches;

Assigning obsolete address from ipcomp_scratches

        if (!scratches)
                return;

        for_each_possible_cpu(i)
               vfree(*per_cpu_ptr(scratches, i));
Trying to free non existent page, causing warning: trying to vfree
existent vm area.
        ...
}

Fix this breakage by updating ipcomp_scrtches with NULL when scratches is freed

---
 net/xfrm/xfrm_ipcomp.c | 1 +
 1 file changed, 1 insertion(+)
net/xfrm/xfrm_ipcomp.c +1 -0
index 0814320472f1..24ac6805275e 100644
@@ -212,6 +212,7 @@ static void ipcomp_free_scratches(void)
 		vfree(*per_cpu_ptr(scratches, i));
 
 	free_percpu(scratches);
+	ipcomp_scratches = NULL;
 }
 
 static void * __percpu *ipcomp_alloc_scratches(void)
Signature · 1 line
2.35.1
author maintainer
#9 lore raw

PATCH 10/34 net: xscale: Fix return type for implementation of ndo_start_xmit

In reply to Sasha Levin

From: GUO Zihua <[email protected]>

[ Upstream commit 0dbaf0fa62329d9fe452d9041a707a33f6274f1f ]

Since Linux now supports CFI, it will be a good idea to fix mismatched return type for implementation of hooks. Otherwise this might get cought out by CFI and cause a panic.

eth_xmit() would return either NETDEV_TX_BUSY or NETDEV_TX_OK, so change the return type to netdev_tx_t directly.

---
 drivers/net/ethernet/xscale/ixp4xx_eth.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/xscale/ixp4xx_eth.c +1 -1
index 403358f2c853..5775e58b0745 100644
@@ -820,7 +820,7 @@ static void eth_txdone_irq(void *unused)
 	}
 }
 
-static int eth_xmit(struct sk_buff *skb, struct net_device *dev)
+static netdev_tx_t eth_xmit(struct sk_buff *skb, struct net_device *dev)
 {
 	struct port *port = netdev_priv(dev);
 	unsigned int txreadyq = port->plat->txreadyq;
Signature · 1 line
2.35.1
author maintainer
#10 lore raw

PATCH 11/34 net: lantiq_etop: Fix return type for implementation of ndo_start_xmit

In reply to Sasha Levin

From: GUO Zihua <[email protected]>

[ Upstream commit c8ef3c94bda0e21123202d057d4a299698fa0ed9 ]

Since Linux now supports CFI, it will be a good idea to fix mismatched return type for implementation of hooks. Otherwise this might get cought out by CFI and cause a panic.

ltq_etop_tx() would return either NETDEV_TX_BUSY or NETDEV_TX_OK, so change the return type to netdev_tx_t directly.

---
 drivers/net/ethernet/lantiq_etop.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/lantiq_etop.c +1 -1
index 2d0c52f7106b..14c35ca72b75 100644
@@ -451,7 +451,7 @@ ltq_etop_stop(struct net_device *dev)
 	return 0;
 }
 
-static int
+static netdev_tx_t
 ltq_etop_tx(struct sk_buff *skb, struct net_device *dev)
 {
 	int queue = skb_get_queue_mapping(skb);
Signature · 1 line
2.35.1
author maintainer
#11 lore raw

PATCH 12/34 net: ftmac100: fix endianness-related issues from 'sparse'

In reply to Sasha Levin

From: Sergei Antonov <[email protected]>

[ Upstream commit 9df696b3b3a4c96c3219eb87c7bf03fb50e490b8 ]

Sparse found a number of endianness-related issues of these kinds:

.../ftmac100.c:192:32: warning: restricted __le32 degrades to integer

.../ftmac100.c:208:23: warning: incorrect type in assignment (different base types)
.../ftmac100.c:208:23: expected unsigned int rxdes0
.../ftmac100.c:208:23: got restricted __le32 [usertype]

.../ftmac100.c:249:23: warning: invalid assignment: &=
.../ftmac100.c:249:23: left side has type unsigned int
.../ftmac100.c:249:23: right side has type restricted __le32

.../ftmac100.c:527:16: warning: cast to restricted __le32

Change type of some fields from 'unsigned int' to '__le32' to fix it.

---
 drivers/net/ethernet/faraday/ftmac100.h | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)
drivers/net/ethernet/faraday/ftmac100.h +6 -6
index fe986f1673fc..8af32f9070f4 100644
@@ -122,9 +122,9 @@
  * Transmit descriptor, aligned to 16 bytes
  */
 struct ftmac100_txdes {
-	unsigned int	txdes0;
-	unsigned int	txdes1;
-	unsigned int	txdes2;	/* TXBUF_BADR */
+	__le32		txdes0;
+	__le32		txdes1;
+	__le32		txdes2;	/* TXBUF_BADR */
 	unsigned int	txdes3;	/* not used by HW */
 } __attribute__ ((aligned(16)));
 
@@ -143,9 +143,9 @@ struct ftmac100_txdes {
  * Receive descriptor, aligned to 16 bytes
  */
 struct ftmac100_rxdes {
-	unsigned int	rxdes0;
-	unsigned int	rxdes1;
-	unsigned int	rxdes2;	/* RXBUF_BADR */
+	__le32		rxdes0;
+	__le32		rxdes1;
+	__le32		rxdes2;	/* RXBUF_BADR */
 	unsigned int	rxdes3;	/* not used by HW */
 } __attribute__ ((aligned(16)));
 
Signature · 1 line
2.35.1
author maintainer
#12 lore raw

PATCH 13/34 wifi: brcmfmac: fix use-after-free bug in brcmf_netdev_start_xmit()

In reply to Sasha Levin

From: Alexander Coffin <[email protected]>

[ Upstream commit 3f42faf6db431e04bf942d2ebe3ae88975723478 ]

Quoted 1 line

ret = brcmf_proto_tx_queue_data(drvr, ifp->ifidx, skb);

may be schedule, and then complete before the line

Quoted 1 line

ndev->stats.tx_bytes += skb->len;

[   46.912801] ==================================================================
[   46.920552] BUG: KASAN: use-after-free in brcmf_netdev_start_xmit+0x718/0x8c8 [brcmfmac]
[   46.928673] Read of size 4 at addr ffffff803f5882e8 by task systemd-resolve/328
[   46.935991]
[   46.937514] CPU: 1 PID: 328 Comm: systemd-resolve Tainted: G           O      5.4.199-[REDACTED] #1
[   46.947255] Hardware name: [REDACTED]
[   46.954568] Call trace:
[   46.957037]  dump_backtrace+0x0/0x2b8
[   46.960719]  show_stack+0x24/0x30
[   46.964052]  dump_stack+0x128/0x194
[   46.967557]  print_address_description.isra.0+0x64/0x380
[   46.972877]  __kasan_report+0x1d4/0x240
[   46.976723]  kasan_report+0xc/0x18
[   46.980138]  __asan_report_load4_noabort+0x18/0x20
[   46.985027]  brcmf_netdev_start_xmit+0x718/0x8c8 [brcmfmac]
[   46.990613]  dev_hard_start_xmit+0x1bc/0xda0
[   46.994894]  sch_direct_xmit+0x198/0xd08
[   46.998827]  __qdisc_run+0x37c/0x1dc0
[   47.002500]  __dev_queue_xmit+0x1528/0x21f8
[   47.006692]  dev_queue_xmit+0x24/0x30
[   47.010366]  neigh_resolve_output+0x37c/0x678
[   47.014734]  ip_finish_output2+0x598/0x2458
[   47.018927]  __ip_finish_output+0x300/0x730
[   47.023118]  ip_output+0x2e0/0x430
[   47.026530]  ip_local_out+0x90/0x140
[   47.030117]  igmpv3_sendpack+0x14c/0x228
[   47.034049]  igmpv3_send_cr+0x384/0x6b8
[   47.037895]  igmp_ifc_timer_expire+0x4c/0x118
[   47.042262]  call_timer_fn+0x1cc/0xbe8
[   47.046021]  __run_timers+0x4d8/0xb28
[   47.049693]  run_timer_softirq+0x24/0x40
[   47.053626]  __do_softirq+0x2c0/0x117c
[   47.057387]  irq_exit+0x2dc/0x388
[   47.060715]  __handle_domain_irq+0xb4/0x158
[   47.064908]  gic_handle_irq+0x58/0xb0
[   47.068581]  el0_irq_naked+0x50/0x5c
[   47.072162]
[   47.073665] Allocated by task 328:
[   47.077083]  save_stack+0x24/0xb0
[   47.080410]  __kasan_kmalloc.isra.0+0xc0/0xe0
[   47.084776]  kasan_slab_alloc+0x14/0x20
[   47.088622]  kmem_cache_alloc+0x15c/0x468
[   47.092643]  __alloc_skb+0xa4/0x498
[   47.096142]  igmpv3_newpack+0x158/0xd78
[   47.099987]  add_grhead+0x210/0x288
[   47.103485]  add_grec+0x6b0/0xb70
[   47.106811]  igmpv3_send_cr+0x2e0/0x6b8
[   47.110657]  igmp_ifc_timer_expire+0x4c/0x118
[   47.115027]  call_timer_fn+0x1cc/0xbe8
[   47.118785]  __run_timers+0x4d8/0xb28
[   47.122457]  run_timer_softirq+0x24/0x40
[   47.126389]  __do_softirq+0x2c0/0x117c
[   47.130142]
[   47.131643] Freed by task 180:
[   47.134712]  save_stack+0x24/0xb0
[   47.138041]  __kasan_slab_free+0x108/0x180
[   47.142146]  kasan_slab_free+0x10/0x18
[   47.145904]  slab_free_freelist_hook+0xa4/0x1b0
[   47.150444]  kmem_cache_free+0x8c/0x528
[   47.154292]  kfree_skbmem+0x94/0x108
[   47.157880]  consume_skb+0x10c/0x5a8
[   47.161466]  __dev_kfree_skb_any+0x88/0xa0
[   47.165598]  brcmu_pkt_buf_free_skb+0x44/0x68 [brcmutil]
[   47.171023]  brcmf_txfinalize+0xec/0x190 [brcmfmac]
[   47.176016]  brcmf_proto_bcdc_txcomplete+0x1c0/0x210 [brcmfmac]
[   47.182056]  brcmf_sdio_sendfromq+0x8dc/0x1e80 [brcmfmac]
[   47.187568]  brcmf_sdio_dpc+0xb48/0x2108 [brcmfmac]
[   47.192529]  brcmf_sdio_dataworker+0xc8/0x238 [brcmfmac]
[   47.197859]  process_one_work+0x7fc/0x1a80
[   47.201965]  worker_thread+0x31c/0xc40
[   47.205726]  kthread+0x2d8/0x370
[   47.208967]  ret_from_fork+0x10/0x18
[   47.212546]
[   47.214051] The buggy address belongs to the object at ffffff803f588280
[   47.214051]  which belongs to the cache skbuff_head_cache of size 208
[   47.227086] The buggy address is located 104 bytes inside of
[   47.227086]  208-byte region [ffffff803f588280, ffffff803f588350)
[   47.238814] The buggy address belongs to the page:
[   47.243618] page:ffffffff00dd6200 refcount:1 mapcount:0 mapping:ffffff804b6bf800 index:0xffffff803f589900 compound_mapcount: 0
[   47.255007] flags: 0x10200(slab|head)
[   47.258689] raw: 0000000000010200 ffffffff00dfa980 0000000200000002 ffffff804b6bf800
[   47.266439] raw: ffffff803f589900 0000000080190018 00000001ffffffff 0000000000000000
[   47.274180] page dumped because: kasan: bad access detected
[   47.279752]
[   47.281251] Memory state around the buggy address:
[   47.286051]  ffffff803f588180: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
[   47.293277]  ffffff803f588200: fb fb fc fc fc fc fc fc fc fc fc fc fc fc fc fc
[   47.300502] >ffffff803f588280: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
[   47.307723]                                                           ^
[   47.314343]  ffffff803f588300: fb fb fb fb fb fb fb fb fb fb fc fc fc fc fc fc
[   47.321569]  ffffff803f588380: fc fc fc fc fc fc fc fc fb fb fb fb fb fb fb fb
[   47.328789] ==================================================================
---
 drivers/net/wireless/broadcom/brcm80211/brcmfmac/core.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/net/wireless/broadcom/brcm80211/brcmfmac/core.c +2 -1
index 61039538a15b..c8e1d505f7b5 100644
@@ -290,6 +290,7 @@ static netdev_tx_t brcmf_netdev_start_xmit(struct sk_buff *skb,
 	struct brcmf_pub *drvr = ifp->drvr;
 	struct ethhdr *eh;
 	int head_delta;
+	unsigned int tx_bytes = skb->len;
 
 	brcmf_dbg(DATA, "Enter, bsscfgidx=%d\n", ifp->bsscfgidx);
 
@@ -364,7 +365,7 @@ static netdev_tx_t brcmf_netdev_start_xmit(struct sk_buff *skb,
 		ndev->stats.tx_dropped++;
 	} else {
 		ndev->stats.tx_packets++;
-		ndev->stats.tx_bytes += skb->len;
+		ndev->stats.tx_bytes += tx_bytes;
 	}
 
 	/* Return ok: we always eat the packet */
Signature · 1 line
2.35.1
author maintainer
#13 lore raw

PATCH 14/34 regulator: core: Prevent integer underflow

In reply to Sasha Levin

From: Patrick Rudolph <[email protected]>

[ Upstream commit 8d8e16592022c9650df8aedfe6552ed478d7135b ]

By using a ratio of delay to poll_enabled_time that is not integer time_remaining underflows and does not exit the loop as expected. As delay could be derived from DT and poll_enabled_time is defined in the driver this can easily happen.

Use a signed iterator to make sure that the loop exits once the remaining time is negative.

---
 drivers/regulator/core.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/regulator/core.c +1 -1
index 317d701487ec..bf8ba73d6c7c 100644
@@ -2544,7 +2544,7 @@ static int _regulator_do_enable(struct regulator_dev *rdev)
 	 * expired, return -ETIMEDOUT.
 	 */
 	if (rdev->desc->poll_enabled_time) {
-		unsigned int time_remaining = delay;
+		int time_remaining = delay;
 
 		while (time_remaining > 0) {
 			_regulator_enable_delay(rdev->desc->poll_enabled_time);
Signature · 1 line
2.35.1
author maintainer
#14 lore raw

PATCH 15/34 Bluetooth: L2CAP: initialize delayed works at l2cap_chan_create()

In reply to Sasha Levin

From: Tetsuo Handa <[email protected]>

[ Upstream commit 2d2cb3066f2c90cd8ca540b36ba7a55e7f2406e0 ]

syzbot is reporting cancel_delayed_work() without INIT_DELAYED_WORK() at l2cap_chan_del() [1], for CONF_NOT_COMPLETE flag (which meant to prevent l2cap_chan_del() from calling cancel_delayed_work()) is cleared by timer which fires before l2cap_chan_del() is called by closing file descriptor created by socket(AF_BLUETOOTH, SOCK_STREAM, BTPROTO_L2CAP).

l2cap_bredr_sig_cmd(L2CAP_CONF_REQ) and l2cap_bredr_sig_cmd(L2CAP_CONF_RSP) are calling l2cap_ertm_init(chan), and they call l2cap_chan_ready() (which clears CONF_NOT_COMPLETE flag) only when l2cap_ertm_init(chan) succeeded.

l2cap_sock_init() does not call l2cap_ertm_init(chan), and it instead sets CONF_NOT_COMPLETE flag by calling l2cap_chan_set_defaults(). However, when connect() is requested, "command 0x0409 tx timeout" happens after 2 seconds
from connect() request, and CONF_NOT_COMPLETE flag is cleared after 4
seconds from connect() request, for l2cap_conn_start() from l2cap_info_timeout() callback scheduled by

  schedule_delayed_work(&conn->info_timer, L2CAP_INFO_TIMEOUT);

in l2cap_connect() is calling l2cap_chan_ready().

Fix this problem by initializing delayed works used by L2CAP_MODE_ERTM mode as soon as l2cap_chan_create() allocates a channel, like I did in commit be8597239379f0f5 ("Bluetooth: initialize skb_queue_head at l2cap_chan_create()").

---
 net/bluetooth/l2cap_core.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)
net/bluetooth/l2cap_core.c +6 -4
index 0c38af2ff209..8d5029c81ee7 100644
@@ -61,6 +61,9 @@ static void l2cap_send_disconn_req(struct l2cap_chan *chan, int err);
 
 static void l2cap_tx(struct l2cap_chan *chan, struct l2cap_ctrl *control,
 		     struct sk_buff_head *skbs, u8 event);
+static void l2cap_retrans_timeout(struct work_struct *work);
+static void l2cap_monitor_timeout(struct work_struct *work);
+static void l2cap_ack_timeout(struct work_struct *work);
 
 static inline u8 bdaddr_type(u8 link_type, u8 bdaddr_type)
 {
@@ -476,6 +479,9 @@ struct l2cap_chan *l2cap_chan_create(void)
 	write_unlock(&chan_list_lock);
 
 	INIT_DELAYED_WORK(&chan->chan_timer, l2cap_chan_timeout);
+	INIT_DELAYED_WORK(&chan->retrans_timer, l2cap_retrans_timeout);
+	INIT_DELAYED_WORK(&chan->monitor_timer, l2cap_monitor_timeout);
+	INIT_DELAYED_WORK(&chan->ack_timer, l2cap_ack_timeout);
 
 	chan->state = BT_OPEN;
 
@@ -3316,10 +3322,6 @@ int l2cap_ertm_init(struct l2cap_chan *chan)
 	chan->rx_state = L2CAP_RX_STATE_RECV;
 	chan->tx_state = L2CAP_TX_STATE_XMIT;
 
-	INIT_DELAYED_WORK(&chan->retrans_timer, l2cap_retrans_timeout);
-	INIT_DELAYED_WORK(&chan->monitor_timer, l2cap_monitor_timeout);
-	INIT_DELAYED_WORK(&chan->ack_timer, l2cap_ack_timeout);
-
 	skb_queue_head_init(&chan->srej_q);
 
 	err = l2cap_seq_list_init(&chan->srej_list, chan->tx_win);
Signature · 1 line
2.35.1
author maintainer
#15 lore raw

PATCH 16/34 net: davicom: Fix return type of dm9000_start_xmit

In reply to Sasha Levin

From: Nathan Huckleberry <[email protected]>

[ Upstream commit 0191580b000d50089a0b351f7cdbec4866e3d0d2 ]

The ndo_start_xmit field in net_device_ops is expected to be of type netdev_tx_t (*ndo_start_xmit)(struct sk_buff *skb, struct net_device *dev).

The mismatched return type breaks forward edge kCFI since the underlying function definition does not match the function hook definition.

The return type of dm9000_start_xmit should be changed from int to netdev_tx_t.

---
 drivers/net/ethernet/davicom/dm9000.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/davicom/dm9000.c +1 -1
index afc4a103c508..c9ee5185e73e 100644
@@ -1015,7 +1015,7 @@ static void dm9000_send_packet(struct net_device *dev,
  *  Hardware start transmission.
  *  Send a packet to media from the upper layer.
  */
-static int
+static netdev_tx_t
 dm9000_start_xmit(struct sk_buff *skb, struct net_device *dev)
 {
 	unsigned long flags;
Signature · 1 line
2.35.1
author maintainer
#16 lore raw

PATCH 17/34 net: ethernet: ti: davinci_emac: Fix return type of emac_dev_xmit

In reply to Sasha Levin

From: Nathan Huckleberry <[email protected]>

[ Upstream commit 5972ca946098487c5155fe13654743f9010f5ed5 ]

The ndo_start_xmit field in net_device_ops is expected to be of type netdev_tx_t (*ndo_start_xmit)(struct sk_buff *skb, struct net_device *dev).

The mismatched return type breaks forward edge kCFI since the underlying function definition does not match the function hook definition.

The return type of emac_dev_xmit should be changed from int to netdev_tx_t.

---
 drivers/net/ethernet/ti/davinci_emac.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/ti/davinci_emac.c +1 -1
index ad5293571af4..1dbba3e8ce50 100644
@@ -942,7 +942,7 @@ static void emac_tx_handler(void *token, int len, int status)
  *
  * Returns success(NETDEV_TX_OK) or error code (typically out of desc's)
  */
-static int emac_dev_xmit(struct sk_buff *skb, struct net_device *ndev)
+static netdev_tx_t emac_dev_xmit(struct sk_buff *skb, struct net_device *ndev)
 {
 	struct device *emac_dev = &ndev->dev;
 	int ret_code;
Signature · 1 line
2.35.1
author maintainer
#17 lore raw

PATCH 18/34 net: korina: Fix return type of korina_send_packet

In reply to Sasha Levin

From: Nathan Huckleberry <[email protected]>

[ Upstream commit 106c67ce46f3c82dd276e983668a91d6ed631173 ]

The ndo_start_xmit field in net_device_ops is expected to be of type netdev_tx_t (*ndo_start_xmit)(struct sk_buff *skb, struct net_device *dev).

The mismatched return type breaks forward edge kCFI since the underlying function definition does not match the function hook definition.

The return type of korina_send_packet should be changed from int to netdev_tx_t.

---
 drivers/net/ethernet/korina.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/net/ethernet/korina.c +2 -1
index 925161959b9b..7d74bb932ee1 100644
@@ -196,7 +196,8 @@ static void korina_chain_rx(struct korina_private *lp,
 }
 
 /* transmit packet */
-static int korina_send_packet(struct sk_buff *skb, struct net_device *dev)
+static netdev_tx_t korina_send_packet(struct sk_buff *skb,
+				      struct net_device *dev)
 {
 	struct korina_private *lp = netdev_priv(dev);
 	unsigned long flags;
Signature · 1 line
2.35.1
author maintainer
#18 lore raw

PATCH 19/34 net: sfp: re-implement soft state polling setup

In reply to Sasha Levin

From: "Russell King (Oracle)" <[email protected]>

[ Upstream commit 8475c4b70b040f9d8cbc308100f2c4d865f810b3 ]

Re-implement the decision making for soft state polling. Instead of generating the soft state mask in sfp_soft_start_poll() by looking at which GPIOs are available, record their availability in sfp_sm_mod_probe() in sfp->state_hw_mask.

This will then allow us to clear bits in sfp->state_hw_mask in module specific quirks when the hardware signals should not be used, thereby allowing us to switch to using the software state polling.

---
 drivers/net/phy/sfp.c | 38 ++++++++++++++++++++++++++------------
 1 file changed, 26 insertions(+), 12 deletions(-)
drivers/net/phy/sfp.c +26 -12
index dcbe278086dc..79e8806a2d39 100644
@@ -233,6 +233,7 @@ struct sfp {
 	bool need_poll;
 
 	struct mutex st_mutex;			/* Protects state */
+	unsigned int state_hw_mask;
 	unsigned int state_soft_mask;
 	unsigned int state;
 	struct delayed_work poll;
@@ -495,17 +496,18 @@ static void sfp_soft_set_state(struct sfp *sfp, unsigned int state)
 static void sfp_soft_start_poll(struct sfp *sfp)
 {
 	const struct sfp_eeprom_id *id = &sfp->id;
+	unsigned int mask = 0;
 
 	sfp->state_soft_mask = 0;
-	if (id->ext.enhopts & SFP_ENHOPTS_SOFT_TX_DISABLE &&
-	    !sfp->gpio[GPIO_TX_DISABLE])
-		sfp->state_soft_mask |= SFP_F_TX_DISABLE;
-	if (id->ext.enhopts & SFP_ENHOPTS_SOFT_TX_FAULT &&
-	    !sfp->gpio[GPIO_TX_FAULT])
-		sfp->state_soft_mask |= SFP_F_TX_FAULT;
-	if (id->ext.enhopts & SFP_ENHOPTS_SOFT_RX_LOS &&
-	    !sfp->gpio[GPIO_LOS])
-		sfp->state_soft_mask |= SFP_F_LOS;
+	if (id->ext.enhopts & SFP_ENHOPTS_SOFT_TX_DISABLE)
+		mask |= SFP_F_TX_DISABLE;
+	if (id->ext.enhopts & SFP_ENHOPTS_SOFT_TX_FAULT)
+		mask |= SFP_F_TX_FAULT;
+	if (id->ext.enhopts & SFP_ENHOPTS_SOFT_RX_LOS)
+		mask |= SFP_F_LOS;
+
+	// Poll the soft state for hardware pins we want to ignore
+	sfp->state_soft_mask = ~sfp->state_hw_mask & mask;
 
 	if (sfp->state_soft_mask & (SFP_F_LOS | SFP_F_TX_FAULT) &&
 	    !sfp->need_poll)
@@ -519,10 +521,11 @@ static void sfp_soft_stop_poll(struct sfp *sfp)
 
 static unsigned int sfp_get_state(struct sfp *sfp)
 {
-	unsigned int state = sfp->get_state(sfp);
+	unsigned int soft = sfp->state_soft_mask & (SFP_F_LOS | SFP_F_TX_FAULT);
+	unsigned int state;
 
-	if (state & SFP_F_PRESENT &&
-	    sfp->state_soft_mask & (SFP_F_LOS | SFP_F_TX_FAULT))
+	state = sfp->get_state(sfp) & sfp->state_hw_mask;
+	if (state & SFP_F_PRESENT && soft)
 		state |= sfp_soft_get_state(sfp);
 
 	return state;
@@ -1888,6 +1891,15 @@ static int sfp_sm_mod_probe(struct sfp *sfp, bool report)
 	if (ret < 0)
 		return ret;
 
+	/* Initialise state bits to use from hardware */
+	sfp->state_hw_mask = SFP_F_PRESENT;
+	if (sfp->gpio[GPIO_TX_DISABLE])
+		sfp->state_hw_mask |= SFP_F_TX_DISABLE;
+	if (sfp->gpio[GPIO_TX_FAULT])
+		sfp->state_hw_mask |= SFP_F_TX_FAULT;
+	if (sfp->gpio[GPIO_LOS])
+		sfp->state_hw_mask |= SFP_F_LOS;
+
 	if (!memcmp(id.base.vendor_name, "ALCATELLUCENT   ", 16) &&
 	    !memcmp(id.base.vendor_pn, "3FE46541AA      ", 16))
 		sfp->module_t_start_up = T_START_UP_BAD_GPON;
@@ -2488,6 +2500,8 @@ static int sfp_probe(struct platform_device *pdev)
 				return PTR_ERR(sfp->gpio[i]);
 		}
 
+	sfp->state_hw_mask = SFP_F_PRESENT;
+
 	sfp->get_state = sfp_gpio_get_state;
 	sfp->set_state = sfp_gpio_set_state;
 
Signature · 1 line
2.35.1
author maintainer
#19 lore raw

PATCH 20/34 net: sfp: move quirk handling into sfp.c

In reply to Sasha Levin

From: "Russell King (Oracle)" <[email protected]>

[ Upstream commit 23571c7b96437483d28a990c906cc81f5f66374e ]

We need to handle more quirks than just those which affect the link modes of the module. Move the quirk lookup into sfp.c, and pass the quirk to sfp-bus.c

---
 drivers/net/phy/sfp-bus.c | 98 ++-------------------------------------
 drivers/net/phy/sfp.c     | 94 ++++++++++++++++++++++++++++++++++++-
 drivers/net/phy/sfp.h     |  9 +++-
 3 files changed, 104 insertions(+), 97 deletions(-)
drivers/net/phy/sfp-bus.c +3 -95
index 850915a37f4c..7fbe462ae346 100644
@@ -10,12 +10,6 @@
 
 #include "sfp.h"
 
-struct sfp_quirk {
-	const char *vendor;
-	const char *part;
-	void (*modes)(const struct sfp_eeprom_id *id, unsigned long *modes);
-};
-
 /**
  * struct sfp_bus - internal representation of a sfp bus
  */
@@ -38,93 +32,6 @@ struct sfp_bus {
 	bool started;
 };
 
-static void sfp_quirk_2500basex(const struct sfp_eeprom_id *id,
-				unsigned long *modes)
-{
-	phylink_set(modes, 2500baseX_Full);
-}
-
-static void sfp_quirk_ubnt_uf_instant(const struct sfp_eeprom_id *id,
-				      unsigned long *modes)
-{
-	/* Ubiquiti U-Fiber Instant module claims that support all transceiver
-	 * types including 10G Ethernet which is not truth. So clear all claimed
-	 * modes and set only one mode which module supports: 1000baseX_Full.
-	 */
-	phylink_zero(modes);
-	phylink_set(modes, 1000baseX_Full);
-}
-
-static const struct sfp_quirk sfp_quirks[] = {
-	{
-		// Alcatel Lucent G-010S-P can operate at 2500base-X, but
-		// incorrectly report 2500MBd NRZ in their EEPROM
-		.vendor = "ALCATELLUCENT",
-		.part = "G010SP",
-		.modes = sfp_quirk_2500basex,
-	}, {
-		// Alcatel Lucent G-010S-A can operate at 2500base-X, but
-		// report 3.2GBd NRZ in their EEPROM
-		.vendor = "ALCATELLUCENT",
-		.part = "3FE46541AA",
-		.modes = sfp_quirk_2500basex,
-	}, {
-		// Huawei MA5671A can operate at 2500base-X, but report 1.2GBd
-		// NRZ in their EEPROM
-		.vendor = "HUAWEI",
-		.part = "MA5671A",
-		.modes = sfp_quirk_2500basex,
-	}, {
-		// Lantech 8330-262D-E can operate at 2500base-X, but
-		// incorrectly report 2500MBd NRZ in their EEPROM
-		.vendor = "Lantech",
-		.part = "8330-262D-E",
-		.modes = sfp_quirk_2500basex,
-	}, {
-		.vendor = "UBNT",
-		.part = "UF-INSTANT",
-		.modes = sfp_quirk_ubnt_uf_instant,
-	},
-};
-
-static size_t sfp_strlen(const char *str, size_t maxlen)
-{
-	size_t size, i;
-
-	/* Trailing characters should be filled with space chars */
-	for (i = 0, size = 0; i < maxlen; i++)
-		if (str[i] != ' ')
-			size = i + 1;
-
-	return size;
-}
-
-static bool sfp_match(const char *qs, const char *str, size_t len)
-{
-	if (!qs)
-		return true;
-	if (strlen(qs) != len)
-		return false;
-	return !strncmp(qs, str, len);
-}
-
-static const struct sfp_quirk *sfp_lookup_quirk(const struct sfp_eeprom_id *id)
-{
-	const struct sfp_quirk *q;
-	unsigned int i;
-	size_t vs, ps;
-
-	vs = sfp_strlen(id->base.vendor_name, ARRAY_SIZE(id->base.vendor_name));
-	ps = sfp_strlen(id->base.vendor_pn, ARRAY_SIZE(id->base.vendor_pn));
-
-	for (i = 0, q = sfp_quirks; i < ARRAY_SIZE(sfp_quirks); i++, q++)
-		if (sfp_match(q->vendor, id->base.vendor_name, vs) &&
-		    sfp_match(q->part, id->base.vendor_pn, ps))
-			return q;
-
-	return NULL;
-}
-
 /**
  * sfp_parse_port() - Parse the EEPROM base ID, setting the port type
  * @bus: a pointer to the &struct sfp_bus structure for the sfp module
@@ -744,12 +651,13 @@ void sfp_link_down(struct sfp_bus *bus)
 }
 EXPORT_SYMBOL_GPL(sfp_link_down);
 
-int sfp_module_insert(struct sfp_bus *bus, const struct sfp_eeprom_id *id)
+int sfp_module_insert(struct sfp_bus *bus, const struct sfp_eeprom_id *id,
+		      const struct sfp_quirk *quirk)
 {
 	const struct sfp_upstream_ops *ops = sfp_get_upstream_ops(bus);
 	int ret = 0;
 
-	bus->sfp_quirk = sfp_lookup_quirk(id);
+	bus->sfp_quirk = quirk;
 
 	if (ops && ops->module_insert)
 		ret = ops->module_insert(bus->upstream, id);
drivers/net/phy/sfp.c +93 -1

118 lines not on this page · raw message

drivers/net/phy/sfp.h +8 -1
index b83f70526270..37c7bbfee539 100644
@@ -6,6 +6,12 @@
 
 struct sfp;
 
+struct sfp_quirk {
+	const char *vendor;
+	const char *part;
+	void (*modes)(const struct sfp_eeprom_id *id, unsigned long *modes);
+};
+
 struct sfp_socket_ops {
 	void (*attach)(struct sfp *sfp);
 	void (*detach)(struct sfp *sfp);
@@ -20,7 +26,8 @@ int sfp_add_phy(struct sfp_bus *bus, struct phy_device *phydev);
 void sfp_remove_phy(struct sfp_bus *bus);
 void sfp_link_up(struct sfp_bus *bus);
 void sfp_link_down(struct sfp_bus *bus);
-int sfp_module_insert(struct sfp_bus *bus, const struct sfp_eeprom_id *id);
+int sfp_module_insert(struct sfp_bus *bus, const struct sfp_eeprom_id *id,
+		      const struct sfp_quirk *quirk);
 void sfp_module_remove(struct sfp_bus *bus);
 int sfp_module_start(struct sfp_bus *bus);
 void sfp_module_stop(struct sfp_bus *bus);
Signature · 1 line
2.35.1
author maintainer
#20 lore raw

PATCH 21/34 net: sfp: move Alcatel Lucent 3FE46541AA fixup

In reply to Sasha Levin

From: "Russell King (Oracle)" <[email protected]>

[ Upstream commit 275416754e9a262c97a1ad6f806a4bc6e0464aa2 ]

Add a new fixup mechanism to the SFP quirks, and use it for this module.

---
 drivers/net/phy/sfp.c | 14 +++++++++-----
 drivers/net/phy/sfp.h |  1 +
 2 files changed, 10 insertions(+), 5 deletions(-)
drivers/net/phy/sfp.c +9 -5
index b09716083af4..8789530aa016 100644
@@ -307,6 +307,11 @@ static const struct of_device_id sfp_of_match[] = {
 };
 MODULE_DEVICE_TABLE(of, sfp_of_match);
 
+static void sfp_fixup_long_startup(struct sfp *sfp)
+{
+	sfp->module_t_start_up = T_START_UP_BAD_GPON;
+}
+
 static void sfp_quirk_2500basex(const struct sfp_eeprom_id *id,
 				unsigned long *modes)
 {
@@ -337,6 +342,7 @@ static const struct sfp_quirk sfp_quirks[] = {
 		.vendor = "ALCATELLUCENT",
 		.part = "3FE46541AA",
 		.modes = sfp_quirk_2500basex,
+		.fixup = sfp_fixup_long_startup,
 	}, {
 		// Huawei MA5671A can operate at 2500base-X, but report 1.2GBd
 		// NRZ in their EEPROM
@@ -1989,11 +1995,7 @@ static int sfp_sm_mod_probe(struct sfp *sfp, bool report)
 	if (sfp->gpio[GPIO_LOS])
 		sfp->state_hw_mask |= SFP_F_LOS;
 
-	if (!memcmp(id.base.vendor_name, "ALCATELLUCENT   ", 16) &&
-	    !memcmp(id.base.vendor_pn, "3FE46541AA      ", 16))
-		sfp->module_t_start_up = T_START_UP_BAD_GPON;
-	else
-		sfp->module_t_start_up = T_START_UP;
+	sfp->module_t_start_up = T_START_UP;
 
 	if (!memcmp(id.base.vendor_name, "HUAWEI          ", 16) &&
 	    !memcmp(id.base.vendor_pn, "MA5671A         ", 16))
@@ -2002,6 +2004,8 @@ static int sfp_sm_mod_probe(struct sfp *sfp, bool report)
 		sfp->tx_fault_ignore = false;
 
 	sfp->quirk = sfp_lookup_quirk(&id);
+	if (sfp->quirk && sfp->quirk->fixup)
+		sfp->quirk->fixup(sfp);
 
 	return 0;
 }
drivers/net/phy/sfp.h +1 -0
index 37c7bbfee539..ef06d3580eea 100644
@@ -10,6 +10,7 @@ struct sfp_quirk {
 	const char *vendor;
 	const char *part;
 	void (*modes)(const struct sfp_eeprom_id *id, unsigned long *modes);
+	void (*fixup)(struct sfp *sfp);
 };
 
 struct sfp_socket_ops {
Signature · 1 line
2.35.1
#21 lore raw

Re: [PATCH AUTOSEL 5.10 21/34] net: sfp: move Alcatel Lucent 3FE46541AA fixup

Hi!

Quoted 2 lines

Add a new fixup mechanism to the SFP quirks, and use it for this module.

There are two preparation patches for this, but this does not fix anything -- it just reimplement quirk in a different way.

We should not have patches 19-21 in stable.

Best regards,
Pavel

Signature · 2 lines
DENX Software Engineering GmbH,      Managing Director: Wolfgang Denk
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
#22 lore raw

Re: [PATCH AUTOSEL 5.10 21/34] net: sfp: move Alcatel Lucent 3FE46541AA fixup

Quoted 9 lines · Pavel Machek

Hi!

Quoted 2 lines

Add a new fixup mechanism to the SFP quirks, and use it for this module.

There are two preparation patches for this, but this does not fix anything -- it just reimplement quirk in a different way.

We should not have patches 19-21 in stable.

They should be being dropped as, as a result of this ongoing madness, I've requested stable to no longer consider *any* of my patches for AUTOSEL treatment.

I'm afraid that the stable kernel is loosing its purpose, and is becoming just another development tree as long as this autosel process exists that picks up what are development patches - coupled with the "if you don't respond we're adding it to stable anyway" approach means that it's a hell of a lot of work for maintainers to be watching what stable is doing 24 hours a day 365 days a year with _no_ possibility of having a break from that.

As I understand it, the autosel stuff is using an AI to work out whether the patch is suitable or not, and no one bothers to look at the quality of its selection - that task is loaded on to each maintainer and requires said maintainer to never take any time off from that task.

It's madness.

Signature · 2 lines
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!
author maintainer
#23 lore raw

PATCH 22/34 net/sched: taprio: taprio_dump and taprio_change are protected by rtnl_mutex

In reply to Sasha Levin

From: Vladimir Oltean <[email protected]>

[ Upstream commit 18cdd2f0998a4967b1fff4c43ed9aef049e42c39 ]

Since the writer-side lock is taken here, we do not need to open an RCU read-side critical section, instead we can use rtnl_dereference() to tell lockdep we are serialized with concurrent writes.

---
 net/sched/sch_taprio.c | 14 ++++----------
 1 file changed, 4 insertions(+), 10 deletions(-)
net/sched/sch_taprio.c +4 -10
index ab8835a72cee..d70a5c4aafe5 100644
@@ -1480,10 +1480,8 @@ static int taprio_change(struct Qdisc *sch, struct nlattr *opt,
 	}
 	INIT_LIST_HEAD(&new_admin->entries);
 
-	rcu_read_lock();
-	oper = rcu_dereference(q->oper_sched);
-	admin = rcu_dereference(q->admin_sched);
-	rcu_read_unlock();
+	oper = rtnl_dereference(q->oper_sched);
+	admin = rtnl_dereference(q->admin_sched);
 
 	/* no changes - no new mqprio settings */
 	if (!taprio_mqprio_cmp(dev, mqprio))
@@ -1843,9 +1841,8 @@ static int taprio_dump(struct Qdisc *sch, struct sk_buff *skb)
 	struct nlattr *nest, *sched_nest;
 	unsigned int i;
 
-	rcu_read_lock();
-	oper = rcu_dereference(q->oper_sched);
-	admin = rcu_dereference(q->admin_sched);
+	oper = rtnl_dereference(q->oper_sched);
+	admin = rtnl_dereference(q->admin_sched);
 
 	opt.num_tc = netdev_get_num_tc(dev);
 	memcpy(opt.prio_tc_map, dev->prio_tc_map, sizeof(opt.prio_tc_map));
@@ -1889,8 +1886,6 @@ static int taprio_dump(struct Qdisc *sch, struct sk_buff *skb)
 	nla_nest_end(skb, sched_nest);
 
 done:
-	rcu_read_unlock();
-
 	return nla_nest_end(skb, nest);
 
 admin_error:
@@ -1900,7 +1895,6 @@ static int taprio_dump(struct Qdisc *sch, struct sk_buff *skb)
 	nla_nest_cancel(skb, nest);
 
 start_error:
-	rcu_read_unlock();
 	return -ENOSPC;
 }
 
Signature · 1 line
2.35.1
#24 lore raw

Re: [PATCH AUTOSEL 5.10 22/34] net/sched: taprio: taprio_dump and taprio_change are protected by rtnl_mutex

Quoted 4 lines · Sasha Levin

From: Vladimir Oltean <[email protected]>

[ Upstream commit 18cdd2f0998a4967b1fff4c43ed9aef049e42c39 ]

Quoted 8 lines

Since the writer-side lock is taken here, we do not need to open an RCU read-side critical section, instead we can use rtnl_dereference() to tell lockdep we are serialized with concurrent writes.

---

Not needed for stable kernels, please drop, thanks.

#25 lore raw

Re: [PATCH AUTOSEL 5.10 22/34] net/sched: taprio: taprio_dump and taprio_change are protected by rtnl_mutex

In reply to Sasha Levin

Hi!

Quoted 7 lines

From: Vladimir Oltean <[email protected]>

[ Upstream commit 18cdd2f0998a4967b1fff4c43ed9aef049e42c39 ]

Since the writer-side lock is taken here, we do not need to open an RCU read-side critical section, instead we can use rtnl_dereference() to tell lockdep we are serialized with concurrent writes.

This is cleanup, not a bugfix. We should not have it in 5.10.

Best regards,
Pavel

Signature · 2 lines
DENX Software Engineering GmbH,      Managing Director: Wolfgang Denk
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
#26 lore raw

Re: [PATCH AUTOSEL 5.10 22/34] net/sched: taprio: taprio_dump and taprio_change are protected by rtnl_mutex

Quoted 11 lines · Pavel Machek

Hi!

Quoted 7 lines

From: Vladimir Oltean <[email protected]>

[ Upstream commit 18cdd2f0998a4967b1fff4c43ed9aef049e42c39 ]

Since the writer-side lock is taken here, we do not need to open an RCU read-side critical section, instead we can use rtnl_dereference() to tell lockdep we are serialized with concurrent writes.

This is cleanup, not a bugfix. We should not have it in 5.10.

Agreed, looks like I missed this one when replying to Sasha for all the others.

#27 lore raw

Re: [PATCH AUTOSEL 5.10 22/34] net/sched: taprio: taprio_dump and taprio_change are protected by rtnl_mutex

Quoted 7 lines · Vladimir Oltean
Quoted 6 lines · Pavel Machek

Hi!

Quoted 4 lines

From: Vladimir Oltean <[email protected]>

[ Upstream commit 18cdd2f0998a4967b1fff4c43ed9aef049e42c39 ]

Quoted 7 lines
Quoted 5 lines
Quoted 3 lines

Since the writer-side lock is taken here, we do not need to open an RCU read-side critical section, instead we can use rtnl_dereference() to tell lockdep we are serialized with concurrent writes.

This is cleanup, not a bugfix. We should not have it in 5.10.

Agreed, looks like I missed this one when replying to Sasha for all the others.

Ah, I do see that I did in fact respond to this already.
https://lore.kernel.org/lkml/20221010133337.4q75fsa6m2v5ttk7@skbuf/

Not sure how you put your eyes on this particular patch?

author maintainer
#28 lore raw

PATCH 23/34 Bluetooth: hci_sysfs: Fix attempting to call device_add multiple times

In reply to Sasha Levin

From: Luiz Augusto von Dentz <[email protected]>

[ Upstream commit 448a496f760664d3e2e79466aa1787e6abc922b5 ]

device_add shall not be called multiple times as stated in its documentation:

'Do not call this routine or device_register() more than once for any device structure'

Syzkaller reports a bug as follows [1]:

------------[ cut here ]------------
kernel BUG at lib/list_debug.c:33!
invalid opcode: 0000 [#1] PREEMPT SMP KASAN
[...]
Call Trace:
 <TASK>
 __list_add include/linux/list.h:69 [inline]
 list_add_tail include/linux/list.h:102 [inline]
 kobj_kset_join lib/kobject.c:164 [inline]
 kobject_add_internal+0x18f/0x8f0 lib/kobject.c:214
 kobject_add_varg lib/kobject.c:358 [inline]
 kobject_add+0x150/0x1c0 lib/kobject.c:410
 device_add+0x368/0x1e90 drivers/base/core.c:3452
 hci_conn_add_sysfs+0x9b/0x1b0 net/bluetooth/hci_sysfs.c:53
 hci_le_cis_estabilished_evt+0x57c/0xae0 net/bluetooth/hci_event.c:6799
 hci_le_meta_evt+0x2b8/0x510 net/bluetooth/hci_event.c:7110
 hci_event_func net/bluetooth/hci_event.c:7440 [inline]
 hci_event_packet+0x63d/0xfd0 net/bluetooth/hci_event.c:7495
 hci_rx_work+0xae7/0x1230 net/bluetooth/hci_core.c:4007
 process_one_work+0x991/0x1610 kernel/workqueue.c:2289
 worker_thread+0x665/0x1080 kernel/workqueue.c:2436
 kthread+0x2e4/0x3a0 kernel/kthread.c:376
 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:306
 </TASK>
---
 net/bluetooth/hci_sysfs.c | 3 +++
 1 file changed, 3 insertions(+)
net/bluetooth/hci_sysfs.c +3 -0
index b69d88b88d2e..ccd2c377bf83 100644
@@ -48,6 +48,9 @@ void hci_conn_add_sysfs(struct hci_conn *conn)
 
 	BT_DBG("conn %p", conn);
 
+	if (device_is_registered(&conn->dev))
+		return;
+
 	dev_set_name(&conn->dev, "%s:%d", hdev->name, conn->handle);
 
 	if (device_add(&conn->dev) < 0) {
Signature · 1 line
2.35.1
author maintainer
#29 lore raw

PATCH 24/34 wifi: ath10k: reset pointer after memory free to avoid potential use-after-free

In reply to Sasha Levin

From: Wen Gong <[email protected]>

[ Upstream commit 1e1cb8e0b73e6f39a9d4a7a15d940b1265387eb5 ]

When running suspend test, kernel crash happened in ath10k, and it is fixed by commit b72a4aff947b ("ath10k: skip ath10k_halt during suspend for driver state RESTARTING").

Currently the crash is fixed, but as a common code style, it is better to set the pointer to NULL after memory is free.

This is to address the code style and it will avoid potential bug of use-after-free.

Tested-on: QCA6174 hw3.2 PCI WLAN.RM.4.4.1-00110-QCARMSWP-1

---
 drivers/net/wireless/ath/ath10k/htt_rx.c | 6 ++++++
 1 file changed, 6 insertions(+)
drivers/net/wireless/ath/ath10k/htt_rx.c +6 -0
index 28ec3c5b4d1f..1b34f12b7eca 100644
@@ -297,12 +297,16 @@ void ath10k_htt_rx_free(struct ath10k_htt *htt)
 			  ath10k_htt_get_vaddr_ring(htt),
 			  htt->rx_ring.base_paddr);
 
+	ath10k_htt_config_paddrs_ring(htt, NULL);
+
 	dma_free_coherent(htt->ar->dev,
 			  sizeof(*htt->rx_ring.alloc_idx.vaddr),
 			  htt->rx_ring.alloc_idx.vaddr,
 			  htt->rx_ring.alloc_idx.paddr);
+	htt->rx_ring.alloc_idx.vaddr = NULL;
 
 	kfree(htt->rx_ring.netbufs_ring);
+	htt->rx_ring.netbufs_ring = NULL;
 }
 
 static inline struct sk_buff *ath10k_htt_rx_netbuf_pop(struct ath10k_htt *htt)
@@ -823,8 +827,10 @@ int ath10k_htt_rx_alloc(struct ath10k_htt *htt)
 			  ath10k_htt_get_rx_ring_size(htt),
 			  vaddr_ring,
 			  htt->rx_ring.base_paddr);
+	ath10k_htt_config_paddrs_ring(htt, NULL);
 err_dma_ring:
 	kfree(htt->rx_ring.netbufs_ring);
+	htt->rx_ring.netbufs_ring = NULL;
 err_netbuf:
 	return -ENOMEM;
 }
Signature · 7 lines
2.35.1


_______________________________________________
ath10k mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/ath10k
author maintainer
#30 lore raw

PATCH 25/34 can: bcm: check the result of can_send() in bcm_can_tx()

In reply to Sasha Levin

From: Ziyang Xuan <[email protected]>

[ Upstream commit 3fd7bfd28cfd68ae80a2fe92ea1615722cc2ee6e ]

If can_send() fail, it should not update frames_abs counter in bcm_can_tx(). Add the result check for can_send() in bcm_can_tx().

---
 net/can/bcm.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)
net/can/bcm.c +4 -3
index e918a0f3cda2..afa82adaf6cd 100644
@@ -274,6 +274,7 @@ static void bcm_can_tx(struct bcm_op *op)
 	struct sk_buff *skb;
 	struct net_device *dev;
 	struct canfd_frame *cf = op->frames + op->cfsiz * op->currframe;
+	int err;
 
 	/* no target device? => exit */
 	if (!op->ifindex)
@@ -298,11 +299,11 @@ static void bcm_can_tx(struct bcm_op *op)
 	/* send with loopback */
 	skb->dev = dev;
 	can_skb_set_owner(skb, op->sk);
-	can_send(skb, 1);
+	err = can_send(skb, 1);
+	if (!err)
+		op->frames_abs++;
 
-	/* update statistics */
 	op->currframe++;
-	op->frames_abs++;
 
 	/* reached last frame? */
 	if (op->currframe >= op->nframes)
Signature · 1 line
2.35.1
author maintainer
#31 lore raw

PATCH 26/34 wifi: rt2x00: don't run Rt5592 IQ calibration on MT7620

In reply to Sasha Levin

From: Daniel Golle <[email protected]>

[ Upstream commit d3aad83d05aec0cfd7670cf0028f2ad4b81de92e ]

The function rt2800_iq_calibrate is intended for Rt5592 only. Don't call it for MT7620 which has it's own calibration functions.

---
 drivers/net/wireless/ralink/rt2x00/rt2800lib.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/net/wireless/ralink/rt2x00/rt2800lib.c +2 -1
index fed6d21cd6ce..3f2c10c2aaf8 100644
@@ -4352,7 +4352,8 @@ static void rt2800_config_channel(struct rt2x00_dev *rt2x00dev,
 		reg = (rf->channel <= 14 ? 0x1c : 0x24) + 2*rt2x00dev->lna_gain;
 		rt2800_bbp_write_with_rx_chain(rt2x00dev, 66, reg);
 
-		rt2800_iq_calibrate(rt2x00dev, rf->channel);
+		if (rt2x00_rt(rt2x00dev, RT5592))
+			rt2800_iq_calibrate(rt2x00dev, rf->channel);
 	}
 
 	bbp = rt2800_bbp_read(rt2x00dev, 4);
Signature · 1 line
2.35.1
author maintainer
#32 lore raw

PATCH 27/34 wifi: rt2x00: set correct TX_SW_CFG1 MAC register for MT7620

In reply to Sasha Levin

From: Daniel Golle <[email protected]>

[ Upstream commit eeb50acf15762b61921f9df18663f839f387c054 ]

Set correct TX_SW_CFG1 MAC register as it is done also in v3 of the vendor driver[1].

[1]: https://gitlab.com/dm38/padavan-ng/-/blob/master/trunk/proprietary/rt_wifi/rtpci/3.0.X.X/mt76x2/chips/rt6352.c#L531

---
 drivers/net/wireless/ralink/rt2x00/rt2800lib.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/wireless/ralink/rt2x00/rt2800lib.c +1 -1
index 3f2c10c2aaf8..327f19cae4d7 100644
@@ -5849,7 +5849,7 @@ static int rt2800_init_registers(struct rt2x00_dev *rt2x00dev)
 		rt2800_register_write(rt2x00dev, TX_SW_CFG0, 0x00000404);
 	} else if (rt2x00_rt(rt2x00dev, RT6352)) {
 		rt2800_register_write(rt2x00dev, TX_SW_CFG0, 0x00000401);
-		rt2800_register_write(rt2x00dev, TX_SW_CFG1, 0x000C0000);
+		rt2800_register_write(rt2x00dev, TX_SW_CFG1, 0x000C0001);
 		rt2800_register_write(rt2x00dev, TX_SW_CFG2, 0x00000000);
 		rt2800_register_write(rt2x00dev, TX_ALC_VGA3, 0x00000000);
 		rt2800_register_write(rt2x00dev, TX0_BB_GAIN_ATTEN, 0x0);
Signature · 1 line
2.35.1
author maintainer
#33 lore raw

PATCH 28/34 wifi: rt2x00: set VGC gain for both chains of MT7620

In reply to Sasha Levin

From: Daniel Golle <[email protected]>

[ Upstream commit 0e09768c085709e10ece3b68f6ac921d3f6a9caa ]

Set bbp66 for all chains of the MT7620.

---
 drivers/net/wireless/ralink/rt2x00/rt2800lib.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/net/wireless/ralink/rt2x00/rt2800lib.c +2 -1
index 327f19cae4d7..94e5c3c373ba 100644
@@ -5626,7 +5626,8 @@ static inline void rt2800_set_vgc(struct rt2x00_dev *rt2x00dev,
 	if (qual->vgc_level != vgc_level) {
 		if (rt2x00_rt(rt2x00dev, RT3572) ||
 		    rt2x00_rt(rt2x00dev, RT3593) ||
-		    rt2x00_rt(rt2x00dev, RT3883)) {
+		    rt2x00_rt(rt2x00dev, RT3883) ||
+		    rt2x00_rt(rt2x00dev, RT6352)) {
 			rt2800_bbp_write_with_rx_chain(rt2x00dev, 66,
 						       vgc_level);
 		} else if (rt2x00_rt(rt2x00dev, RT5592)) {
Signature · 1 line
2.35.1
author maintainer
#34 lore raw

PATCH 29/34 wifi: rt2x00: set SoC wmac clock register

In reply to Sasha Levin

From: Daniel Golle <[email protected]>

[ Upstream commit cbde6ed406a51092d9e8a2df058f5f8490f27443 ]

Instead of using the default value 33 (pci), set US_CYC_CNT init based on Programming guide:
If available, set chipset bus clock with fallback to cpu clock/3.

---
 .../net/wireless/ralink/rt2x00/rt2800lib.c    | 21 +++++++++++++++++++
 1 file changed, 21 insertions(+)
drivers/net/wireless/ralink/rt2x00/rt2800lib.c +21 -0
index 94e5c3c373ba..f237fc17dedc 100644
@@ -6112,6 +6112,27 @@ static int rt2800_init_registers(struct rt2x00_dev *rt2x00dev)
 		reg = rt2800_register_read(rt2x00dev, US_CYC_CNT);
 		rt2x00_set_field32(&reg, US_CYC_CNT_CLOCK_CYCLE, 125);
 		rt2800_register_write(rt2x00dev, US_CYC_CNT, reg);
+	} else if (rt2x00_is_soc(rt2x00dev)) {
+		struct clk *clk = clk_get_sys("bus", NULL);
+		int rate;
+
+		if (IS_ERR(clk)) {
+			clk = clk_get_sys("cpu", NULL);
+
+			if (IS_ERR(clk)) {
+				rate = 125;
+			} else {
+				rate = clk_get_rate(clk) / 3000000;
+				clk_put(clk);
+			}
+		} else {
+			rate = clk_get_rate(clk) / 1000000;
+			clk_put(clk);
+		}
+
+		reg = rt2800_register_read(rt2x00dev, US_CYC_CNT);
+		rt2x00_set_field32(&reg, US_CYC_CNT_CLOCK_CYCLE, rate);
+		rt2800_register_write(rt2x00dev, US_CYC_CNT, reg);
 	}
 
 	reg = rt2800_register_read(rt2x00dev, HT_FBK_CFG0);
Signature · 1 line
2.35.1
author maintainer
#35 lore raw

PATCH 30/34 wifi: rt2x00: correctly set BBP register 86 for MT7620

In reply to Sasha Levin

From: Daniel Golle <[email protected]>

[ Upstream commit c9aada64fe6493461127f1522d7e2f01792d2424 ]

Instead of 0 set the correct value for BBP register 86 for MT7620.

---
 drivers/net/wireless/ralink/rt2x00/rt2800lib.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/net/wireless/ralink/rt2x00/rt2800lib.c +4 -1
index f237fc17dedc..4bdd3a95f2d2 100644
@@ -4151,7 +4151,10 @@ static void rt2800_config_channel(struct rt2x00_dev *rt2x00dev,
 		rt2800_bbp_write(rt2x00dev, 62, 0x37 - rt2x00dev->lna_gain);
 		rt2800_bbp_write(rt2x00dev, 63, 0x37 - rt2x00dev->lna_gain);
 		rt2800_bbp_write(rt2x00dev, 64, 0x37 - rt2x00dev->lna_gain);
-		rt2800_bbp_write(rt2x00dev, 86, 0);
+		if (rt2x00_rt(rt2x00dev, RT6352))
+			rt2800_bbp_write(rt2x00dev, 86, 0x38);
+		else
+			rt2800_bbp_write(rt2x00dev, 86, 0);
 	}
 
 	if (rf->channel <= 14) {
Signature · 1 line
2.35.1
author maintainer
#36 lore raw

PATCH 31/34 net: If sock is dead don't access sock's sk_wq in sk_stream_wait_memory

In reply to Sasha Levin

From: Liu Jian <[email protected]>

[ Upstream commit 3f8ef65af927db247418d4e1db49164d7a158fc5 ]

Fixes the below NULL pointer dereference:

[...]

  [   14.471200] Call Trace:
  [   14.471562]  <TASK>
  [   14.471882]  lock_acquire+0x245/0x2e0
  [   14.472416]  ? remove_wait_queue+0x12/0x50
  [   14.473014]  ? _raw_spin_lock_irqsave+0x17/0x50
  [   14.473681]  _raw_spin_lock_irqsave+0x3d/0x50
  [   14.474318]  ? remove_wait_queue+0x12/0x50
  [   14.474907]  remove_wait_queue+0x12/0x50
  [   14.475480]  sk_stream_wait_memory+0x20d/0x340
  [   14.476127]  ? do_wait_intr_irq+0x80/0x80
  [   14.476704]  do_tcp_sendpages+0x287/0x600
  [   14.477283]  tcp_bpf_push+0xab/0x260
  [   14.477817]  tcp_bpf_sendmsg_redir+0x297/0x500
  [   14.478461]  ? __local_bh_enable_ip+0x77/0xe0
  [   14.479096]  tcp_bpf_send_verdict+0x105/0x470
  [   14.479729]  tcp_bpf_sendmsg+0x318/0x4f0
  [   14.480311]  sock_sendmsg+0x2d/0x40
  [   14.480822]  ____sys_sendmsg+0x1b4/0x1c0
  [   14.481390]  ? copy_msghdr_from_user+0x62/0x80
  [   14.482048]  ___sys_sendmsg+0x78/0xb0
  [   14.482580]  ? vmf_insert_pfn_prot+0x91/0x150
  [   14.483215]  ? __do_fault+0x2a/0x1a0
  [   14.483738]  ? do_fault+0x15e/0x5d0
  [   14.484246]  ? __handle_mm_fault+0x56b/0x1040
  [   14.484874]  ? lock_is_held_type+0xdf/0x130
  [   14.485474]  ? find_held_lock+0x2d/0x90
  [   14.486046]  ? __sys_sendmsg+0x41/0x70
  [   14.486587]  __sys_sendmsg+0x41/0x70
  [   14.487105]  ? intel_pmu_drain_pebs_core+0x350/0x350
  [   14.487822]  do_syscall_64+0x34/0x80
  [   14.488345]  entry_SYSCALL_64_after_hwframe+0x63/0xcd
  [...]

The test scenario has the following flow:

thread1 thread2
----------- ---------------
tcp_bpf_sendmsg
tcp_bpf_send_verdict
tcp_bpf_sendmsg_redir sock_close

    tcp_bpf_push_locked                 __sock_release
     tcp_bpf_push                         //inet_release
      do_tcp_sendpages                    sock->ops->release
       sk_stream_wait_memory          	   // tcp_close
          sk_wait_event                      sk->sk_prot->close
           release_sock(__sk);
            ***
                                                lock_sock(sk);
                                                  __tcp_close
                                                    sock_orphan(sk)
                                                      sk->sk_wq  = NULL
                                                release_sock
            ****
           lock_sock(__sk);
          remove_wait_queue(sk_sleep(sk), &wait);
             sk_sleep(sk)
             //NULL pointer dereference
             &rcu_dereference_raw(sk->sk_wq)->wait

While waiting for memory in thread1, the socket is released with its wait queue because thread2 has closed it. This caused by tcp_bpf_send_verdict didn't increase the f_count of psock->sk_redir->sk_socket->file in thread1.

We should check if SOCK_DEAD flag is set on wakeup in sk_stream_wait_memory before accessing the wait queue.

---
 net/core/stream.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
net/core/stream.c +2 -1
index a166a32b411f..a61130504827 100644
@@ -159,7 +159,8 @@ int sk_stream_wait_memory(struct sock *sk, long *timeo_p)
 		*timeo_p = current_timeo;
 	}
 out:
-	remove_wait_queue(sk_sleep(sk), &wait);
+	if (!sock_flag(sk, SOCK_DEAD))
+		remove_wait_queue(sk_sleep(sk), &wait);
 	return err;
 
 do_error:
Signature · 1 line
2.35.1
author maintainer
#37 lore raw

PATCH 32/34 Bluetooth: L2CAP: Fix user-after-free

In reply to Sasha Levin

From: Luiz Augusto von Dentz <[email protected]>

[ Upstream commit 35fcbc4243aad7e7d020b7c1dfb14bb888b20a4f ]

This uses l2cap_chan_hold_unless_zero() after calling __l2cap_get_chan_blah() to prevent the following trace:

Bluetooth: l2cap_core.c:static void l2cap_chan_destroy(struct kref *kref)
Bluetooth: chan 0000000023c4974d
Bluetooth: parent 00000000ae861c08
==================================================================

BUG: KASAN: use-after-free in __mutex_waiter_is_first
kernel/locking/mutex.c:191 [inline]
BUG: KASAN: use-after-free in __mutex_lock_common
kernel/locking/mutex.c:671 [inline]
BUG: KASAN: use-after-free in __mutex_lock+0x278/0x400
kernel/locking/mutex.c:729
Read of size 8 at addr ffff888006a49b08 by task kworker/u3:2/389
---
 net/bluetooth/l2cap_core.c | 7 +++++++
 1 file changed, 7 insertions(+)
net/bluetooth/l2cap_core.c +7 -0
index 8d5029c81ee7..83dd76e9196f 100644
@@ -4305,6 +4305,12 @@ static int l2cap_connect_create_rsp(struct l2cap_conn *conn,
 		}
 	}
 
+	chan = l2cap_chan_hold_unless_zero(chan);
+	if (!chan) {
+		err = -EBADSLT;
+		goto unlock;
+	}
+
 	err = 0;
 
 	l2cap_chan_lock(chan);
@@ -4334,6 +4340,7 @@ static int l2cap_connect_create_rsp(struct l2cap_conn *conn,
 	}
 
 	l2cap_chan_unlock(chan);
+	l2cap_chan_put(chan);
 
 unlock:
 	mutex_unlock(&conn->chan_lock);
Signature · 1 line
2.35.1
author maintainer
#38 lore raw

PATCH 33/34 libbpf: Fix overrun in netlink attribute iteration

In reply to Sasha Levin

From: Xin Liu <[email protected]>

[ Upstream commit 51e05a8cf8eb34da7473823b7f236a77adfef0b4 ]

I accidentally found that a change in commit 1045b03e07d8 ("netlink: fix overrun in attribute iteration") was not synchronized to the function nla_ok in tools/lib/bpf/nlattr.c, I think it is necessary to modify, this patch will do it.

---
 tools/lib/bpf/nlattr.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
tools/lib/bpf/nlattr.c +1 -1
index b607fa9852b1..8f00a2ee5762 100644
@@ -32,7 +32,7 @@ static struct nlattr *nla_next(const struct nlattr *nla, int *remaining)
 
 static int nla_ok(const struct nlattr *nla, int remaining)
 {
-	return remaining >= sizeof(*nla) &&
+	return remaining >= (int)sizeof(*nla) &&
 	       nla->nla_len >= sizeof(*nla) &&
 	       nla->nla_len <= remaining;
 }
Signature · 1 line
2.35.1
author maintainer
#39 lore raw

PATCH 34/34 r8152: Rate limit overflow messages

In reply to Sasha Levin

From: Andrew Gaul <[email protected]>

[ Upstream commit 93e2be344a7db169b7119de21ac1bf253b8c6907 ]

My system shows almost 10 million of these messages over a 24-hour period which pollutes my logs.

---
 drivers/net/usb/r8152.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/net/usb/r8152.c +3 -1
index 0bb5b1c78654..a526242a3e36 100644
@@ -1689,7 +1689,9 @@ static void intr_callback(struct urb *urb)
 			   "Stop submitting intr, status %d\n", status);
 		return;
 	case -EOVERFLOW:
-		netif_info(tp, intr, tp->netdev, "intr status -EOVERFLOW\n");
+		if (net_ratelimit())
+			netif_info(tp, intr, tp->netdev,
+				   "intr status -EOVERFLOW\n");
 		goto resubmit;
 	/* -EPIPE:  should clear the halt */
 	default:
Signature · 1 line
2.35.1
#40 lore raw
In reply to Sasha Levin

Hi!

Quoted 4 lines

[ Upstream commit 86331c7e0cd819bf0c1d0dcf895e0c90b0aa9a6f ]

reported by smatch

Quoted 8 lines

phy.c:854 rtw_phy_linear_2_db() error: buffer overflow 'db_invert_table[i]' 8 <= 8 (assuming for loop doesn't break)

However, it seems to be a false alarm because we prevent it originally via

       if (linear >= db_invert_table[11][7])
               return 96; /* maximum 96 dB */

Still, we adjust the code to be more readable and avoid smatch warning.

There's no bug, it is just smatch that is confused. We should not take this to 5.10.

Best regards,
Pavel

Quoted 41 lines
 drivers/net/wireless/realtek/rtw88/phy.c | 21 ++++++++-------------
 1 file changed, 8 insertions(+), 13 deletions(-)
drivers/net/wireless/realtek/rtw88/phy.c +8 -13
index af8b703d11d4..0fc5a893c395 100644
@@ -604,23 +604,18 @@ static u8 rtw_phy_linear_2_db(u64 linear)
 	u8 j;
 	u32 dB;
 
-	if (linear >= db_invert_table[11][7])
-		return 96; /* maximum 96 dB */
-
 	for (i = 0; i < 12; i++) {
-		if (i <= 2 && (linear << FRAC_BITS) <= db_invert_table[i][7])
-			break;
-		else if (i > 2 && linear <= db_invert_table[i][7])
-			break;
+		for (j = 0; j < 8; j++) {
+			if (i <= 2 && (linear << FRAC_BITS) <= db_invert_table[i][j])
+				goto cnt;
+			else if (i > 2 && linear <= db_invert_table[i][j])
+				goto cnt;
+		}
 	}
 
-	for (j = 0; j < 8; j++) {
-		if (i <= 2 && (linear << FRAC_BITS) <= db_invert_table[i][j])
-			break;
-		else if (i > 2 && linear <= db_invert_table[i][j])
-			break;
-	}
+	return 96; /* maximum 96 dB */
 
+cnt:
 	if (j == 0 && i == 0)
 		goto end;
 
Signature · 1 line
2.35.1
Signature · 2 lines
DENX Software Engineering GmbH,      Managing Director: Wolfgang Denk
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany