Skip to contentSkip to conversation

5.16.3-rc1 review

PATCH In review Greg Kroah-Hartman 12 participants 1,058 messages Jan 24, 2022 – Feb 3, 2022 Page 6 of 8

Subsystems (9)
lore
Share & tools
lore
author maintainer
#666 lore raw

PATCH 666/1039 PM: runtime: Add safety net to supplier device release

In reply to Greg Kroah-Hartman

From: Rafael J. Wysocki <[email protected]>

[ Upstream commit d1579e61192e0e686faa4208500ef4c3b529b16c ]

Because refcount_dec_not_one() returns true if the target refcount becomes saturated, it is generally unsafe to use its return value as a loop termination condition, but that is what happens when a device link's supplier device is released during runtime PM suspend operations and on device link removal.

To address this, introduce pm_runtime_release_supplier() to be used in the above cases which will check the supplier device's runtime PM usage counter in addition to the refcount_dec_not_one() return value, so the loop can be terminated in case the rpm_active refcount value becomes invalid, and update the code in question to use it as appropriate.

This change is not expected to have any visible functional impact.

---
 drivers/base/core.c          |  3 +--
 drivers/base/power/runtime.c | 41 ++++++++++++++++++++++++++----------
 include/linux/pm_runtime.h   |  3 +++
 3 files changed, 34 insertions(+), 13 deletions(-)
drivers/base/core.c +1 -2
index fd034d7424472..b191bd17de891 100644
@@ -485,8 +485,7 @@ static void device_link_release_fn(struct work_struct *work)
 	/* Ensure that all references to the link object have been dropped. */
 	device_link_synchronize_removal();
 
-	while (refcount_dec_not_one(&link->rpm_active))
-		pm_runtime_put(link->supplier);
+	pm_runtime_release_supplier(link, true);
 
 	put_device(link->consumer);
 	put_device(link->supplier);
drivers/base/power/runtime.c +30 -11

58 lines not on this page · raw message

include/linux/pm_runtime.h +3 -0
index eddd66d426caf..016de5776b6db 100644
@@ -58,6 +58,7 @@ extern void pm_runtime_get_suppliers(struct device *dev);
 extern void pm_runtime_put_suppliers(struct device *dev);
 extern void pm_runtime_new_link(struct device *dev);
 extern void pm_runtime_drop_link(struct device_link *link);
+extern void pm_runtime_release_supplier(struct device_link *link, bool check_idle);
 
 extern int devm_pm_runtime_enable(struct device *dev);
 
@@ -283,6 +284,8 @@ static inline void pm_runtime_get_suppliers(struct device *dev) {}
 static inline void pm_runtime_put_suppliers(struct device *dev) {}
 static inline void pm_runtime_new_link(struct device *dev) {}
 static inline void pm_runtime_drop_link(struct device_link *link) {}
+static inline void pm_runtime_release_supplier(struct device_link *link,
+					       bool check_idle) {}
 
 #endif /* !CONFIG_PM */
 
Signature · 1 line
2.34.1
author maintainer
#667 lore raw

PATCH 667/1039 cpufreq: Fix initialization of min and max frequency QoS requests

In reply to Greg Kroah-Hartman

From: Rafael J. Wysocki <[email protected]>

[ Upstream commit 521223d8b3ec078f670c7c35a1a04b1b2af07966 ]

The min and max frequency QoS requests in the cpufreq core are initialized to whatever the current min and max frequency values are at the init time, but if any of these values change later (for example, cpuinfo.max_freq is updated by the driver), these initial request values will be limiting the CPU frequency unnecessarily unless they are changed by user space via sysfs.

To address this, initialize min_freq_req and max_freq_req to FREQ_QOS_MIN_DEFAULT_VALUE and FREQ_QOS_MAX_DEFAULT_VALUE, respectively, so they don't really limit anything until user space updates them.

---
 drivers/cpufreq/cpufreq.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/cpufreq/cpufreq.c +2 -2
index 096c3848fa415..76ffdaf8c8b5e 100644
@@ -1403,7 +1403,7 @@ static int cpufreq_online(unsigned int cpu)
 
 		ret = freq_qos_add_request(&policy->constraints,
 					   policy->min_freq_req, FREQ_QOS_MIN,
-					   policy->min);
+					   FREQ_QOS_MIN_DEFAULT_VALUE);
 		if (ret < 0) {
 			/*
 			 * So we don't call freq_qos_remove_request() for an
@@ -1423,7 +1423,7 @@ static int cpufreq_online(unsigned int cpu)
 
 		ret = freq_qos_add_request(&policy->constraints,
 					   policy->max_freq_req, FREQ_QOS_MAX,
-					   policy->max);
+					   FREQ_QOS_MAX_DEFAULT_VALUE);
 		if (ret < 0) {
 			policy->max_freq_req = NULL;
 			goto out_destroy_policy;
Signature · 1 line
2.34.1
author maintainer
#668 lore raw

PATCH 668/1039 usb: hub: Add delay for SuperSpeed hub resume to let links transit to U0

In reply to Greg Kroah-Hartman

From: Kai-Heng Feng <[email protected]>

[ Upstream commit 00558586382891540c59c9febc671062425a6e47 ]

When a new USB device gets plugged to nested hubs, the affected hub, which connects to usb 2-1.4-port2, doesn't report there's any change, hence the nested hubs go back to runtime suspend like nothing happened:

[  281.032951] usb usb2: usb wakeup-resume
[  281.032959] usb usb2: usb auto-resume
[  281.032974] hub 2-0:1.0: hub_resume
[  281.033011] usb usb2-port1: status 0263 change 0000
[  281.033077] hub 2-0:1.0: state 7 ports 4 chg 0000 evt 0000
[  281.049797] usb 2-1: usb wakeup-resume
[  281.069800] usb 2-1: Waited 0ms for CONNECT
[  281.069810] usb 2-1: finish resume
[  281.070026] hub 2-1:1.0: hub_resume
[  281.070250] usb 2-1-port4: status 0203 change 0000
[  281.070272] usb usb2-port1: resume, status 0
[  281.070282] hub 2-1:1.0: state 7 ports 4 chg 0010 evt 0000
[  281.089813] usb 2-1.4: usb wakeup-resume
[  281.109792] usb 2-1.4: Waited 0ms for CONNECT
[  281.109801] usb 2-1.4: finish resume
[  281.109991] hub 2-1.4:1.0: hub_resume
[  281.110147] usb 2-1.4-port2: status 0263 change 0000
[  281.110234] usb 2-1-port4: resume, status 0
[  281.110239] usb 2-1-port4: status 0203, change 0000, 10.0 Gb/s
[  281.110266] hub 2-1.4:1.0: state 7 ports 4 chg 0000 evt 0000
[  281.110426] hub 2-1.4:1.0: hub_suspend
[  281.110565] usb 2-1.4: usb auto-suspend, wakeup 1
[  281.130998] hub 2-1:1.0: hub_suspend
[  281.137788] usb 2-1: usb auto-suspend, wakeup 1
[  281.142935] hub 2-0:1.0: state 7 ports 4 chg 0000 evt 0000
[  281.177828] usb 2-1: usb wakeup-resume
[  281.197839] usb 2-1: Waited 0ms for CONNECT
[  281.197850] usb 2-1: finish resume
[  281.197984] hub 2-1:1.0: hub_resume
[  281.198203] usb 2-1-port4: status 0203 change 0000
[  281.198228] usb usb2-port1: resume, status 0
[  281.198237] hub 2-1:1.0: state 7 ports 4 chg 0010 evt 0000
[  281.217835] usb 2-1.4: usb wakeup-resume
[  281.237834] usb 2-1.4: Waited 0ms for CONNECT
[  281.237845] usb 2-1.4: finish resume
[  281.237990] hub 2-1.4:1.0: hub_resume
[  281.238067] usb 2-1.4-port2: status 0263 change 0000
[  281.238148] usb 2-1-port4: resume, status 0
[  281.238152] usb 2-1-port4: status 0203, change 0000, 10.0 Gb/s
[  281.238166] hub 2-1.4:1.0: state 7 ports 4 chg 0000 evt 0000
[  281.238385] hub 2-1.4:1.0: hub_suspend
[  281.238523] usb 2-1.4: usb auto-suspend, wakeup 1
[  281.258076] hub 2-1:1.0: hub_suspend
[  281.265744] usb 2-1: usb auto-suspend, wakeup 1
[  281.285976] hub 2-0:1.0: hub_suspend
[  281.285988] usb usb2: bus auto-suspend, wakeup 1

USB 3.2 spec, 9.2.5.4 "Changing Function Suspend State" says that "If the link is in a non-U0 state, then the device must transition the link to U0 prior to sending the remote wake message", but the hub only transits the link to U0 after signaling remote wakeup.

So be more forgiving and use a 20ms delay to let the link transit to U0 for remote wakeup.

---
 drivers/usb/core/hub.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/usb/core/hub.c +4 -1
index 3bc4a86c3d0a5..ac6c5ccfe1cb7 100644
@@ -1110,7 +1110,10 @@ static void hub_activate(struct usb_hub *hub, enum hub_activation_type type)
 		} else {
 			hub_power_on(hub, true);
 		}
-	}
+	/* Give some time on remote wakeup to let links to transit to U0 */
+	} else if (hub_is_superspeed(hub->hdev))
+		msleep(20);
+
  init2:
 
 	/*
Signature · 1 line
2.34.1
author maintainer
#669 lore raw

PATCH 669/1039 mt76: mt7615: fix possible deadlock while mt7615_register_ext_phy()

In reply to Greg Kroah-Hartman

From: Peter Chiu <[email protected]>

[ Upstream commit 8c55516de3f9b76b9d9444e7890682ec2efc809f ]

ieee80211_register_hw() is called with rtnl_lock held, and this could be caused lockdep from a work item that's on a workqueue that is flushed with the rtnl held.

Move mt7615_register_ext_phy() outside the init_work().

---
 drivers/net/wireless/mediatek/mt76/mt7615/pci_init.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)
drivers/net/wireless/mediatek/mt76/mt7615/pci_init.c +6 -2
index a2465b49ecd0c..87b4aa52ee0f9 100644
@@ -28,8 +28,6 @@ static void mt7615_pci_init_work(struct work_struct *work)
 		return;
 
 	mt7615_init_work(dev);
-	if (dev->dbdc_support)
-		mt7615_register_ext_phy(dev);
 }
 
 static int mt7615_init_hardware(struct mt7615_dev *dev)
@@ -160,6 +158,12 @@ int mt7615_register_device(struct mt7615_dev *dev)
 	mt7615_init_txpower(dev, &dev->mphy.sband_2g.sband);
 	mt7615_init_txpower(dev, &dev->mphy.sband_5g.sband);
 
+	if (dev->dbdc_support) {
+		ret = mt7615_register_ext_phy(dev);
+		if (ret)
+			return ret;
+	}
+
 	return mt7615_init_debugfs(dev);
 }
 
Signature · 1 line
2.34.1
author maintainer
#670 lore raw

PATCH 670/1039 mt76: mt7915: fix SMPS operation fail

In reply to Greg Kroah-Hartman

From: Ryder Lee <[email protected]>

[ Upstream commit 8f05835425ce3f669e4b6d7c2c39a9aa22e1506c ]

TGn fails sending SM power save mode action frame to the AP to switch from dynamic SMPS mode to static mode.

---
 .../net/wireless/mediatek/mt76/mt7915/mcu.c   | 81 ++++++++++++-------
 .../net/wireless/mediatek/mt76/mt7915/mcu.h   |  8 ++
 2 files changed, 61 insertions(+), 28 deletions(-)
drivers/net/wireless/mediatek/mt76/mt7915/mcu.c +53 -28
index 852d5d97c70b1..8215b3d79bbdc 100644
@@ -1752,33 +1752,6 @@ int mt7915_mcu_sta_update_hdr_trans(struct mt7915_dev *dev,
 				     true);
 }
 
-int mt7915_mcu_add_smps(struct mt7915_dev *dev, struct ieee80211_vif *vif,
-			struct ieee80211_sta *sta)
-{
-	struct mt7915_vif *mvif = (struct mt7915_vif *)vif->drv_priv;
-	struct mt7915_sta *msta = (struct mt7915_sta *)sta->drv_priv;
-	struct wtbl_req_hdr *wtbl_hdr;
-	struct tlv *sta_wtbl;
-	struct sk_buff *skb;
-
-	skb = mt7915_mcu_alloc_sta_req(dev, mvif, msta,
-				       MT7915_STA_UPDATE_MAX_SIZE);
-	if (IS_ERR(skb))
-		return PTR_ERR(skb);
-
-	sta_wtbl = mt7915_mcu_add_tlv(skb, STA_REC_WTBL, sizeof(struct tlv));
-
-	wtbl_hdr = mt7915_mcu_alloc_wtbl_req(dev, msta, WTBL_SET, sta_wtbl,
-					     &skb);
-	if (IS_ERR(wtbl_hdr))
-		return PTR_ERR(wtbl_hdr);
-
-	mt7915_mcu_wtbl_smps_tlv(skb, sta, sta_wtbl, wtbl_hdr);
-
-	return mt76_mcu_skb_send_msg(&dev->mt76, skb,
-				     MCU_EXT_CMD(STA_REC_UPDATE), true);
-}
-
 static inline bool
 mt7915_is_ebf_supported(struct mt7915_phy *phy, struct ieee80211_vif *vif,
 			struct ieee80211_sta *sta, bool bfee)
@@ -2049,6 +2022,21 @@ mt7915_mcu_sta_bfee_tlv(struct mt7915_dev *dev, struct sk_buff *skb,
 	bfee->fb_identity_matrix = (nrow == 1 && tx_ant == 2);
 }
 
+static enum mcu_mmps_mode
+mt7915_mcu_get_mmps_mode(enum ieee80211_smps_mode smps)
+{
+	switch (smps) {
+	case IEEE80211_SMPS_OFF:
+		return MCU_MMPS_DISABLE;
+	case IEEE80211_SMPS_STATIC:
+		return MCU_MMPS_STATIC;
+	case IEEE80211_SMPS_DYNAMIC:
+		return MCU_MMPS_DYNAMIC;
+	default:
+		return MCU_MMPS_DISABLE;
+	}
+}
+
 int mt7915_mcu_set_fixed_rate_ctrl(struct mt7915_dev *dev,
 				   struct ieee80211_vif *vif,
 				   struct ieee80211_sta *sta,
@@ -2076,7 +2064,11 @@ int mt7915_mcu_set_fixed_rate_ctrl(struct mt7915_dev *dev,
 	case RATE_PARAM_FIXED_MCS:
 	case RATE_PARAM_FIXED_GI:
 	case RATE_PARAM_FIXED_HE_LTF:
-		ra->phy = *phy;
+		if (phy)
+			ra->phy = *phy;
+		break;
+	case RATE_PARAM_MMPS_UPDATE:
+		ra->mmps_mode = mt7915_mcu_get_mmps_mode(sta->smps_mode);
 		break;
 	default:
 		break;
@@ -2087,6 +2079,39 @@ int mt7915_mcu_set_fixed_rate_ctrl(struct mt7915_dev *dev,
 				     MCU_EXT_CMD(STA_REC_UPDATE), true);
 }
 
+int mt7915_mcu_add_smps(struct mt7915_dev *dev, struct ieee80211_vif *vif,
+			struct ieee80211_sta *sta)
+{
+	struct mt7915_vif *mvif = (struct mt7915_vif *)vif->drv_priv;
+	struct mt7915_sta *msta = (struct mt7915_sta *)sta->drv_priv;
+	struct wtbl_req_hdr *wtbl_hdr;
+	struct tlv *sta_wtbl;
+	struct sk_buff *skb;
+	int ret;
+
+	skb = mt7915_mcu_alloc_sta_req(dev, mvif, msta,
+				       MT7915_STA_UPDATE_MAX_SIZE);
+	if (IS_ERR(skb))
+		return PTR_ERR(skb);
+
+	sta_wtbl = mt7915_mcu_add_tlv(skb, STA_REC_WTBL, sizeof(struct tlv));
+
+	wtbl_hdr = mt7915_mcu_alloc_wtbl_req(dev, msta, WTBL_SET, sta_wtbl,
+					     &skb);
+	if (IS_ERR(wtbl_hdr))
+		return PTR_ERR(wtbl_hdr);
+
+	mt7915_mcu_wtbl_smps_tlv(skb, sta, sta_wtbl, wtbl_hdr);
+
+	ret = mt76_mcu_skb_send_msg(&dev->mt76, skb,
+				    MCU_EXT_CMD(STA_REC_UPDATE), true);
+	if (ret)
+		return ret;
+
+	return mt7915_mcu_set_fixed_rate_ctrl(dev, vif, sta, NULL,
+					      RATE_PARAM_MMPS_UPDATE);
+}
+
 static int
 mt7915_mcu_add_rate_ctrl_fixed(struct mt7915_dev *dev,
 			       struct ieee80211_vif *vif,
drivers/net/wireless/mediatek/mt76/mt7915/mcu.h +8 -0
index 1f5a64ba9b59d..628e90d0c394e 100644
@@ -365,6 +365,13 @@ enum {
 	MCU_PHY_STATE_OFDMLQ_CNINFO,
 };
 
+enum mcu_mmps_mode {
+	MCU_MMPS_STATIC,
+	MCU_MMPS_DYNAMIC,
+	MCU_MMPS_RSV,
+	MCU_MMPS_DISABLE,
+};
+
 #define STA_TYPE_STA			BIT(0)
 #define STA_TYPE_AP			BIT(1)
 #define STA_TYPE_ADHOC			BIT(2)
@@ -960,6 +967,7 @@ struct sta_rec_ra_fixed {
 
 enum {
 	RATE_PARAM_FIXED = 3,
+	RATE_PARAM_MMPS_UPDATE = 5,
 	RATE_PARAM_FIXED_HE_LTF = 7,
 	RATE_PARAM_FIXED_MCS,
 	RATE_PARAM_FIXED_GI = 11,
Signature · 1 line
2.34.1
author maintainer
#671 lore raw

PATCH 671/1039 mt76: connac: fix a theoretical NULL pointer dereference in mt76_connac_get_phy_mode

In reply to Greg Kroah-Hartman

From: Lorenzo Bianconi <[email protected]>

[ Upstream commit c9dbeac4988f6d4c4211b3747ec9c9c75ea87967 ]

Even if it is not a real bug since mt76_connac_get_phy_mode runs just for mt7921 where only STA is supported, fix a theoretical NULL pointer dereference if new added modes do not support HE

---
 drivers/net/wireless/mediatek/mt76/mt76_connac_mcu.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/net/wireless/mediatek/mt76/mt76_connac_mcu.c +2 -2
index a5857a788da50..7733c8fad2413 100644
@@ -1180,7 +1180,7 @@ mt76_connac_get_phy_mode(struct mt76_phy *phy, struct ieee80211_vif *vif,
 		if (ht_cap->ht_supported)
 			mode |= PHY_MODE_GN;
 
-		if (he_cap->has_he)
+		if (he_cap && he_cap->has_he)
 			mode |= PHY_MODE_AX_24G;
 	} else if (band == NL80211_BAND_5GHZ || band == NL80211_BAND_6GHZ) {
 		mode |= PHY_MODE_A;
@@ -1191,7 +1191,7 @@ mt76_connac_get_phy_mode(struct mt76_phy *phy, struct ieee80211_vif *vif,
 		if (vht_cap->vht_supported)
 			mode |= PHY_MODE_AC;
 
-		if (he_cap->has_he) {
+		if (he_cap && he_cap->has_he) {
 			if (band == NL80211_BAND_6GHZ)
 				mode |= PHY_MODE_AX_6G;
 			else
Signature · 1 line
2.34.1
author maintainer
#672 lore raw

PATCH 672/1039 mt76: do not pass the received frame with decryption error

In reply to Greg Kroah-Hartman

From: Xing Song <[email protected]>

[ Upstream commit dd28dea52ad9376d2b243a8981726646e1f60b1a ]

MAC80211 doesn't care any decryption error in 802.3 path, so received frame will be dropped if HW tell us that the cipher configuration is not matched as well as the header has been translated to 802.3. This case only appears when IEEE80211_FCTL_PROTECTED is 0 and cipher suit is not none in the corresponding HW entry.

The received frame is only reported to monitor interface if HW decryption block tell us there is ICV error or CCMP/BIP/WPI MIC error. Note in this case the reported frame is decrypted 802.11 frame and the payload may be malformed due to mismatched key.

---
 drivers/net/wireless/mediatek/mt76/mt7603/mac.c | 4 ++++
 drivers/net/wireless/mediatek/mt76/mt7615/mac.c | 9 ++++++++-
 drivers/net/wireless/mediatek/mt76/mt7915/mac.c | 9 ++++++++-
 drivers/net/wireless/mediatek/mt76/mt7921/mac.c | 9 ++++++++-
 4 files changed, 28 insertions(+), 3 deletions(-)
drivers/net/wireless/mediatek/mt76/mt7603/mac.c +4 -0
index fe03e31989bb1..a9ac61b9f854a 100644
@@ -525,6 +525,10 @@ mt7603_mac_fill_rx(struct mt7603_dev *dev, struct sk_buff *skb)
 	if (rxd2 & MT_RXD2_NORMAL_TKIP_MIC_ERR)
 		status->flag |= RX_FLAG_MMIC_ERROR;
 
+	/* ICV error or CCMP/BIP/WPI MIC error */
+	if (rxd2 & MT_RXD2_NORMAL_ICV_ERR)
+		status->flag |= RX_FLAG_ONLY_MONITOR;
+
 	if (FIELD_GET(MT_RXD2_NORMAL_SEC_MODE, rxd2) != 0 &&
 	    !(rxd2 & (MT_RXD2_NORMAL_CLM | MT_RXD2_NORMAL_CM))) {
 		status->flag |= RX_FLAG_DECRYPTED;
drivers/net/wireless/mediatek/mt76/mt7615/mac.c +8 -1
index 423f69015e3ec..c79abce543f3b 100644
@@ -286,9 +286,16 @@ static int mt7615_mac_fill_rx(struct mt7615_dev *dev, struct sk_buff *skb)
 	if (rxd2 & MT_RXD2_NORMAL_AMSDU_ERR)
 		return -EINVAL;
 
+	hdr_trans = rxd1 & MT_RXD1_NORMAL_HDR_TRANS;
+	if (hdr_trans && (rxd2 & MT_RXD2_NORMAL_CM))
+		return -EINVAL;
+
+	/* ICV error or CCMP/BIP/WPI MIC error */
+	if (rxd2 & MT_RXD2_NORMAL_ICV_ERR)
+		status->flag |= RX_FLAG_ONLY_MONITOR;
+
 	unicast = (rxd1 & MT_RXD1_NORMAL_ADDR_TYPE) == MT_RXD1_NORMAL_U2M;
 	idx = FIELD_GET(MT_RXD2_NORMAL_WLAN_IDX, rxd2);
-	hdr_trans = rxd1 & MT_RXD1_NORMAL_HDR_TRANS;
 	status->wcid = mt7615_rx_get_wcid(dev, idx, unicast);
 
 	if (status->wcid) {
drivers/net/wireless/mediatek/mt76/mt7915/mac.c +8 -1
index 809dc18e5083c..38d66411444a1 100644
@@ -426,9 +426,16 @@ mt7915_mac_fill_rx(struct mt7915_dev *dev, struct sk_buff *skb)
 	if (rxd2 & MT_RXD2_NORMAL_AMSDU_ERR)
 		return -EINVAL;
 
+	hdr_trans = rxd2 & MT_RXD2_NORMAL_HDR_TRANS;
+	if (hdr_trans && (rxd1 & MT_RXD1_NORMAL_CM))
+		return -EINVAL;
+
+	/* ICV error or CCMP/BIP/WPI MIC error */
+	if (rxd1 & MT_RXD1_NORMAL_ICV_ERR)
+		status->flag |= RX_FLAG_ONLY_MONITOR;
+
 	unicast = FIELD_GET(MT_RXD3_NORMAL_ADDR_TYPE, rxd3) == MT_RXD3_NORMAL_U2M;
 	idx = FIELD_GET(MT_RXD1_NORMAL_WLAN_IDX, rxd1);
-	hdr_trans = rxd2 & MT_RXD2_NORMAL_HDR_TRANS;
 	status->wcid = mt7915_rx_get_wcid(dev, idx, unicast);
 
 	if (status->wcid) {
drivers/net/wireless/mediatek/mt76/mt7921/mac.c +8 -1
index 321d9f1d3f865..7228b34c66436 100644
@@ -428,10 +428,17 @@ mt7921_mac_fill_rx(struct mt7921_dev *dev, struct sk_buff *skb)
 	if (rxd2 & MT_RXD2_NORMAL_AMSDU_ERR)
 		return -EINVAL;
 
+	hdr_trans = rxd2 & MT_RXD2_NORMAL_HDR_TRANS;
+	if (hdr_trans && (rxd1 & MT_RXD1_NORMAL_CM))
+		return -EINVAL;
+
+	/* ICV error or CCMP/BIP/WPI MIC error */
+	if (rxd1 & MT_RXD1_NORMAL_ICV_ERR)
+		status->flag |= RX_FLAG_ONLY_MONITOR;
+
 	chfreq = FIELD_GET(MT_RXD3_NORMAL_CH_FREQ, rxd3);
 	unicast = FIELD_GET(MT_RXD3_NORMAL_ADDR_TYPE, rxd3) == MT_RXD3_NORMAL_U2M;
 	idx = FIELD_GET(MT_RXD1_NORMAL_WLAN_IDX, rxd1);
-	hdr_trans = rxd2 & MT_RXD2_NORMAL_HDR_TRANS;
 	status->wcid = mt7921_rx_get_wcid(dev, idx, unicast);
 
 	if (status->wcid) {
Signature · 1 line
2.34.1
author maintainer
#673 lore raw

PATCH 673/1039 mt76: mt7615: improve wmm index allocation

In reply to Greg Kroah-Hartman

From: Felix Fietkau <[email protected]>

[ Upstream commit 70fb028707c8871ef9e56b3ffa3d895e14956cc4 ]

Typically all AP interfaces on a PHY will share the same WMM settings, while sta/mesh interfaces will usually inherit the settings from a remote device. In order minimize the likelihood of conflicting WMM settings, make all AP interfaces share one slot, and all non-AP interfaces another one.

This also fixes running multiple AP interfaces on MT7613, which only has 3 WMM slots.

---
 drivers/net/wireless/mediatek/mt76/mt7615/main.c | 8 +++-----
 1 file changed, 3 insertions(+), 5 deletions(-)
drivers/net/wireless/mediatek/mt76/mt7615/main.c +3 -5
index 24c23d4951432..1fdcada157d6b 100644
@@ -211,11 +211,9 @@ static int mt7615_add_interface(struct ieee80211_hw *hw,
 	mvif->mt76.omac_idx = idx;
 
 	mvif->mt76.band_idx = ext_phy;
-	if (mt7615_ext_phy(dev))
-		mvif->mt76.wmm_idx = ext_phy * (MT7615_MAX_WMM_SETS / 2) +
-				mvif->mt76.idx % (MT7615_MAX_WMM_SETS / 2);
-	else
-		mvif->mt76.wmm_idx = mvif->mt76.idx % MT7615_MAX_WMM_SETS;
+	mvif->mt76.wmm_idx = vif->type != NL80211_IFTYPE_AP;
+	if (ext_phy)
+		mvif->mt76.wmm_idx += 2;
 
 	dev->mt76.vif_mask |= BIT(mvif->mt76.idx);
 	dev->omac_mask |= BIT_ULL(mvif->mt76.omac_idx);
Signature · 1 line
2.34.1
author maintainer
#674 lore raw

PATCH 674/1039 mt76: mt7921: fix network buffer leak by txs missing

In reply to Greg Kroah-Hartman

From: Deren Wu <[email protected]>

[ Upstream commit e0bf699ad8e52330d848144a9624a067ad588bd6 ]

TXS in mt7921 may be forwared to tx_done event. Should try to catch TXS information in tx_done event as well.

---
 .../net/wireless/mediatek/mt76/mt7921/mac.c   |  2 +-
 .../net/wireless/mediatek/mt76/mt7921/mcu.c   | 14 ++++++++++
 .../net/wireless/mediatek/mt76/mt7921/mcu.h   | 27 +++++++++++++++++++
 .../wireless/mediatek/mt76/mt7921/mt7921.h    |  1 +
 4 files changed, 43 insertions(+), 1 deletion(-)
drivers/net/wireless/mediatek/mt76/mt7921/mac.c +1 -1
index 7228b34c66436..fc21a78b37c49 100644
@@ -1072,7 +1072,7 @@ out:
 	return !!skb;
 }
 
-static void mt7921_mac_add_txs(struct mt7921_dev *dev, void *data)
+void mt7921_mac_add_txs(struct mt7921_dev *dev, void *data)
 {
 	struct mt7921_sta *msta = NULL;
 	struct mt76_wcid *wcid;
drivers/net/wireless/mediatek/mt76/mt7921/mcu.c +14 -0
index 1cc1c32ca258e..484a8c57b862e 100644
@@ -391,6 +391,17 @@ mt7921_mcu_low_power_event(struct mt7921_dev *dev, struct sk_buff *skb)
 	trace_lp_event(dev, event->state);
 }
 
+static void
+mt7921_mcu_tx_done_event(struct mt7921_dev *dev, struct sk_buff *skb)
+{
+	struct mt7921_mcu_tx_done_event *event;
+
+	skb_pull(skb, sizeof(struct mt7921_mcu_rxd));
+	event = (struct mt7921_mcu_tx_done_event *)skb->data;
+
+	mt7921_mac_add_txs(dev, event->txs);
+}
+
 static void
 mt7921_mcu_rx_unsolicited_event(struct mt7921_dev *dev, struct sk_buff *skb)
 {
@@ -418,6 +429,9 @@ mt7921_mcu_rx_unsolicited_event(struct mt7921_dev *dev, struct sk_buff *skb)
 	case MCU_EVENT_LP_INFO:
 		mt7921_mcu_low_power_event(dev, skb);
 		break;
+	case MCU_EVENT_TX_DONE:
+		mt7921_mcu_tx_done_event(dev, skb);
+		break;
 	default:
 		break;
 	}
drivers/net/wireless/mediatek/mt76/mt7921/mcu.h +27 -0
index edc0c73f8c01c..68cb0ce013dbd 100644
@@ -91,6 +91,33 @@ enum {
 	MCU_EVENT_COREDUMP = 0xf0,
 };
 
+struct mt7921_mcu_tx_done_event {
+	u8 pid;
+	u8 status;
+	__le16 seq;
+
+	u8 wlan_idx;
+	u8 tx_cnt;
+	__le16 tx_rate;
+
+	u8 flag;
+	u8 tid;
+	u8 rsp_rate;
+	u8 mcs;
+
+	u8 bw;
+	u8 tx_pwr;
+	u8 reason;
+	u8 rsv0[1];
+
+	__le32 delay;
+	__le32 timestamp;
+	__le32 applied_flag;
+	u8 txs[28];
+
+	u8 rsv1[32];
+} __packed;
+
 /* ext event table */
 enum {
 	MCU_EXT_EVENT_RATE_REPORT = 0x87,
drivers/net/wireless/mediatek/mt76/mt7921/mt7921.h +1 -0
index d6b823713ba33..96647801850a5 100644
@@ -464,4 +464,5 @@ int mt7921s_tx_prepare_skb(struct mt76_dev *mdev, void *txwi_ptr,
 			   struct mt76_tx_info *tx_info);
 void mt7921s_tx_complete_skb(struct mt76_dev *mdev, struct mt76_queue_entry *e);
 bool mt7921s_tx_status_data(struct mt76_dev *mdev, u8 *update);
+void mt7921_mac_add_txs(struct mt7921_dev *dev, void *data);
 #endif
Signature · 1 line
2.34.1
author maintainer
#675 lore raw

PATCH 675/1039 ath9k_htc: fix NULL pointer dereference at ath9k_htc_rxep()

In reply to Greg Kroah-Hartman

From: Tetsuo Handa <[email protected]>

[ Upstream commit b0ec7e55fce65f125bd1d7f02e2dc4de62abee34 ]

syzbot is reporting lockdep warning followed by kernel panic at ath9k_htc_rxep() [1], for ath9k_htc_rxep() depends on ath9k_rx_init() being already completed.

Since ath9k_htc_rxep() is set by ath9k_htc_connect_svc(WMI_BEACON_SVC)
from ath9k_init_htc_services(), it is possible that ath9k_htc_rxep() is
called via timer interrupt before ath9k_rx_init() from ath9k_init_device() is called.

Since we can't call ath9k_init_device() before ath9k_init_htc_services(), let's hold ath9k_htc_rxep() no-op until ath9k_rx_init() completes.

---
 drivers/net/wireless/ath/ath9k/htc.h          | 1 +
 drivers/net/wireless/ath/ath9k/htc_drv_txrx.c | 8 ++++++++
 2 files changed, 9 insertions(+)
drivers/net/wireless/ath/ath9k/htc.h +1 -0
index 0a1634238e673..4f71e962279af 100644
@@ -281,6 +281,7 @@ struct ath9k_htc_rxbuf {
 struct ath9k_htc_rx {
 	struct list_head rxbuf;
 	spinlock_t rxbuflock;
+	bool initialized;
 };
 
 #define ATH9K_HTC_TX_CLEANUP_INTERVAL 50 /* ms */
drivers/net/wireless/ath/ath9k/htc_drv_txrx.c +8 -0
index 8e69e8989f6d3..e7a21eaf3a68d 100644
@@ -1130,6 +1130,10 @@ void ath9k_htc_rxep(void *drv_priv, struct sk_buff *skb,
 	struct ath9k_htc_rxbuf *rxbuf = NULL, *tmp_buf = NULL;
 	unsigned long flags;
 
+	/* Check if ath9k_rx_init() completed. */
+	if (!data_race(priv->rx.initialized))
+		goto err;
+
 	spin_lock_irqsave(&priv->rx.rxbuflock, flags);
 	list_for_each_entry(tmp_buf, &priv->rx.rxbuf, list) {
 		if (!tmp_buf->in_process) {
@@ -1185,6 +1189,10 @@ int ath9k_rx_init(struct ath9k_htc_priv *priv)
 		list_add_tail(&rxbuf->list, &priv->rx.rxbuf);
 	}
 
+	/* Allow ath9k_htc_rxep() to operate. */
+	smp_wmb();
+	priv->rx.initialized = true;
+
 	return 0;
 
 err:
Signature · 1 line
2.34.1
author maintainer
#676 lore raw

PATCH 676/1039 ath9k_htc: fix NULL pointer dereference at ath9k_htc_tx_get_packet()

In reply to Greg Kroah-Hartman

From: Tetsuo Handa <[email protected]>

[ Upstream commit 8b3046abc99eefe11438090bcc4ec3a3994b55d0 ]

syzbot is reporting lockdep warning at ath9k_wmi_event_tasklet() followed by kernel panic at get_htc_epid_queue() from ath9k_htc_tx_get_packet() from ath9k_htc_txstatus() [1], for ath9k_wmi_event_tasklet(WMI_TXSTATUS_EVENTID) depends on spin_lock_init() from ath9k_init_priv() being already completed.

Since ath9k_wmi_event_tasklet() is set by ath9k_init_wmi() from ath9k_htc_probe_device(), it is possible that ath9k_wmi_event_tasklet() is called via tasklet interrupt before spin_lock_init() from ath9k_init_priv()
from ath9k_init_device() from ath9k_htc_probe_device() is called.

Let's hold ath9k_wmi_event_tasklet(WMI_TXSTATUS_EVENTID) no-op until ath9k_tx_init() completes.

---
 drivers/net/wireless/ath/ath9k/htc.h          | 1 +
 drivers/net/wireless/ath/ath9k/htc_drv_txrx.c | 5 +++++
 drivers/net/wireless/ath/ath9k/wmi.c          | 4 ++++
 3 files changed, 10 insertions(+)
drivers/net/wireless/ath/ath9k/htc.h +1 -0
index 4f71e962279af..6b45e63fae4ba 100644
@@ -306,6 +306,7 @@ struct ath9k_htc_tx {
 	DECLARE_BITMAP(tx_slot, MAX_TX_BUF_NUM);
 	struct timer_list cleanup_timer;
 	spinlock_t tx_lock;
+	bool initialized;
 };
 
 struct ath9k_htc_tx_ctl {
drivers/net/wireless/ath/ath9k/htc_drv_txrx.c +5 -0
index e7a21eaf3a68d..6a850a0bfa8ad 100644
@@ -813,6 +813,11 @@ int ath9k_tx_init(struct ath9k_htc_priv *priv)
 	skb_queue_head_init(&priv->tx.data_vi_queue);
 	skb_queue_head_init(&priv->tx.data_vo_queue);
 	skb_queue_head_init(&priv->tx.tx_failed);
+
+	/* Allow ath9k_wmi_event_tasklet(WMI_TXSTATUS_EVENTID) to operate. */
+	smp_wmb();
+	priv->tx.initialized = true;
+
 	return 0;
 }
 
drivers/net/wireless/ath/ath9k/wmi.c +4 -0
index fe29ad4b9023c..f315c54bd3ac0 100644
@@ -169,6 +169,10 @@ void ath9k_wmi_event_tasklet(struct tasklet_struct *t)
 					     &wmi->drv_priv->fatal_work);
 			break;
 		case WMI_TXSTATUS_EVENTID:
+			/* Check if ath9k_tx_init() completed. */
+			if (!data_race(priv->tx.initialized))
+				break;
+
 			spin_lock_bh(&priv->tx.tx_lock);
 			if (priv->tx.flags & ATH9K_HTC_OP_TX_DRAIN) {
 				spin_unlock_bh(&priv->tx.tx_lock);
Signature · 1 line
2.34.1
author maintainer
#677 lore raw

PATCH 677/1039 ath9k: Fix out-of-bound memcpy in ath9k_hif_usb_rx_stream

In reply to Greg Kroah-Hartman

From: Zekun Shen <[email protected]>

[ Upstream commit 6ce708f54cc8d73beca213cec66ede5ce100a781 ]

Large pkt_len can lead to out-out-bound memcpy. Current ath9k_hif_usb_rx_stream allows combining the content of two urb inputs to one pkt. The first input can indicate the size of the pkt. Any remaining size is saved in hif_dev->rx_remain_len. While processing the next input, memcpy is used with rx_remain_len.

4-byte pkt_len can go up to 0xffff, while a single input is 0x4000 maximum in size (MAX_RX_BUF_SIZE). Thus, the patch adds a check for pkt_len which must not exceed 2 * MAX_RX_BUG_SIZE.

BUG: KASAN: slab-out-of-bounds in ath9k_hif_usb_rx_cb+0x490/0xed7 [ath9k_htc]
Read of size 46393 at addr ffff888018798000 by task kworker/0:1/23

CPU: 0 PID: 23 Comm: kworker/0:1 Not tainted 5.6.0 #63
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996),
BIOS rel-1.10.2-0-g5f4c7b1-prebuilt.qemu-project.org 04/01/2014
Workqueue: events request_firmware_work_func
Call Trace:
 <IRQ>
 dump_stack+0x76/0xa0
 print_address_description.constprop.0+0x16/0x200
 ? ath9k_hif_usb_rx_cb+0x490/0xed7 [ath9k_htc]
 ? ath9k_hif_usb_rx_cb+0x490/0xed7 [ath9k_htc]
 __kasan_report.cold+0x37/0x7c
 ? ath9k_hif_usb_rx_cb+0x490/0xed7 [ath9k_htc]
 kasan_report+0xe/0x20
 check_memory_region+0x15a/0x1d0
 memcpy+0x20/0x50
 ath9k_hif_usb_rx_cb+0x490/0xed7 [ath9k_htc]
 ? hif_usb_mgmt_cb+0x2d9/0x2d9 [ath9k_htc]
 ? _raw_spin_lock_irqsave+0x7b/0xd0
 ? _raw_spin_trylock_bh+0x120/0x120
 ? __usb_unanchor_urb+0x12f/0x210
 __usb_hcd_giveback_urb+0x1e4/0x380
 usb_giveback_urb_bh+0x241/0x4f0
 ? __hrtimer_run_queues+0x316/0x740
 ? __usb_hcd_giveback_urb+0x380/0x380
 tasklet_action_common.isra.0+0x135/0x330
 __do_softirq+0x18c/0x634
 irq_exit+0x114/0x140
 smp_apic_timer_interrupt+0xde/0x380
 apic_timer_interrupt+0xf/0x20

I found the bug using a custome USBFuzz port. It's a research work to fuzz USB stack/drivers. I modified it to fuzz ath9k driver only, providing hand-crafted usb descriptors to QEMU.

After fixing the value of pkt_tag to ATH_USB_RX_STREAM_MODE_TAG in QEMU emulation, I found the KASAN report. The bug is triggerable whenever pkt_len is above two MAX_RX_BUG_SIZE. I used the same input that crashes to test the driver works when applying the patch.

---
 drivers/net/wireless/ath/ath9k/hif_usb.c | 7 +++++++
 1 file changed, 7 insertions(+)
drivers/net/wireless/ath/ath9k/hif_usb.c +7 -0
index 860da13bfb6ac..f06eec99de688 100644
@@ -590,6 +590,13 @@ static void ath9k_hif_usb_rx_stream(struct hif_device_usb *hif_dev,
 			return;
 		}
 
+		if (pkt_len > 2 * MAX_RX_BUF_SIZE) {
+			dev_err(&hif_dev->udev->dev,
+				"ath9k_htc: invalid pkt_len (%x)\n", pkt_len);
+			RX_STAT_INC(skb_dropped);
+			return;
+		}
+
 		pad_len = 4 - (pkt_len & 0x3);
 		if (pad_len == 4)
 			pad_len = 0;
Signature · 1 line
2.34.1
author maintainer
#678 lore raw

PATCH 678/1039 rtw88: 8822c: update rx settings to prevent potential hw deadlock

In reply to Greg Kroah-Hartman

From: Po-Hao Huang <[email protected]>

[ Upstream commit c1afb26727d9e507d3e17a9890e7aaf7fc85cd55 ]

These settings enables mac to detect and recover when rx fifo circuit deadlock occurs. Previous version missed this, so we fix it.

---
 drivers/net/wireless/realtek/rtw88/main.c     | 2 +-
 drivers/net/wireless/realtek/rtw88/rtw8821c.h | 2 +-
 drivers/net/wireless/realtek/rtw88/rtw8822b.c | 2 +-
 drivers/net/wireless/realtek/rtw88/rtw8822c.c | 2 +-
 4 files changed, 4 insertions(+), 4 deletions(-)
drivers/net/wireless/realtek/rtw88/main.c +1 -1
index a0d4d6e31fb49..dbf1ce3daeb8e 100644
@@ -1868,7 +1868,7 @@ int rtw_core_init(struct rtw_dev *rtwdev)
 
 	/* default rx filter setting */
 	rtwdev->hal.rcr = BIT_APP_FCS | BIT_APP_MIC | BIT_APP_ICV |
-			  BIT_HTC_LOC_CTRL | BIT_APP_PHYSTS |
+			  BIT_PKTCTL_DLEN | BIT_HTC_LOC_CTRL | BIT_APP_PHYSTS |
 			  BIT_AB | BIT_AM | BIT_APM;
 
 	ret = rtw_load_firmware(rtwdev, RTW_NORMAL_FW);
drivers/net/wireless/realtek/rtw88/rtw8821c.h +1 -1
index 112faa60f653e..d9fbddd7b0f35 100644
@@ -131,7 +131,7 @@ _rtw_write32s_mask(struct rtw_dev *rtwdev, u32 addr, u32 mask, u32 data)
 #define WLAN_TX_FUNC_CFG2		0x30
 #define WLAN_MAC_OPT_NORM_FUNC1		0x98
 #define WLAN_MAC_OPT_LB_FUNC1		0x80
-#define WLAN_MAC_OPT_FUNC2		0x30810041
+#define WLAN_MAC_OPT_FUNC2		0xb0810041
 
 #define WLAN_SIFS_CFG	(WLAN_SIFS_CCK_CONT_TX | \
 			(WLAN_SIFS_OFDM_CONT_TX << BIT_SHIFT_SIFS_OFDM_CTX) | \
drivers/net/wireless/realtek/rtw88/rtw8822b.c +1 -1
index c409c8c29ec8b..e870ad7518342 100644
@@ -205,7 +205,7 @@ static void rtw8822b_phy_set_param(struct rtw_dev *rtwdev)
 #define WLAN_TX_FUNC_CFG2		0x30
 #define WLAN_MAC_OPT_NORM_FUNC1		0x98
 #define WLAN_MAC_OPT_LB_FUNC1		0x80
-#define WLAN_MAC_OPT_FUNC2		0x30810041
+#define WLAN_MAC_OPT_FUNC2		0xb0810041
 
 #define WLAN_SIFS_CFG	(WLAN_SIFS_CCK_CONT_TX | \
 			(WLAN_SIFS_OFDM_CONT_TX << BIT_SHIFT_SIFS_OFDM_CTX) | \
drivers/net/wireless/realtek/rtw88/rtw8822c.c +1 -1
index 46b881e8e4feb..0e1c5d95d4649 100644
@@ -1962,7 +1962,7 @@ static void rtw8822c_phy_set_param(struct rtw_dev *rtwdev)
 #define WLAN_TX_FUNC_CFG2		0x30
 #define WLAN_MAC_OPT_NORM_FUNC1		0x98
 #define WLAN_MAC_OPT_LB_FUNC1		0x80
-#define WLAN_MAC_OPT_FUNC2		0x30810041
+#define WLAN_MAC_OPT_FUNC2		0xb0810041
 #define WLAN_MAC_INT_MIG_CFG		0x33330000
 
 #define WLAN_SIFS_CFG	(WLAN_SIFS_CCK_CONT_TX | \
Signature · 1 line
2.34.1
author maintainer
#679 lore raw

PATCH 679/1039 PM: AVS: qcom-cpr: Use div64_ul instead of do_div

In reply to Greg Kroah-Hartman

From: Changcheng Deng <[email protected]>

[ Upstream commit 92c550f9ffd2884bb5def52b5c0485a35e452784 ]

do_div() does a 64-by-32 division. Here the divisor is an unsigned long which on some platforms is 64 bit wide. So use div64_ul instead of do_div to avoid a possible truncation.

---
 drivers/soc/qcom/cpr.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/soc/qcom/cpr.c +1 -1
index 1d818a8ba2089..e9b854ed1bdfd 100644
@@ -1010,7 +1010,7 @@ static int cpr_interpolate(const struct corner *corner, int step_volt,
 		return corner->uV;
 
 	temp = f_diff * (uV_high - uV_low);
-	do_div(temp, f_high - f_low);
+	temp = div64_ul(temp, f_high - f_low);
 
 	/*
 	 * max_volt_scale has units of uV/MHz while freq values
Signature · 1 line
2.34.1
author maintainer
#680 lore raw

PATCH 680/1039 iwlwifi: recognize missing PNVM data and then log filename

In reply to Greg Kroah-Hartman

From: Luca Coelho <[email protected]>

[ Upstream commit 1db385c668d3ccb04ccdb5f0f190fd17b2db29c6 ]

We can detect that a FW SYSASSERT is due to missing PNVM data by checking the assertion code. When this happens, it's is useful for the user if we print the filename where the driver is looking for the data.

Add the PNVM missing assertion code to the dump list and print out the name of the file we're looking for when this happens.

---
 drivers/net/wireless/intel/iwlwifi/fw/dump.c | 9 +++++++++
 drivers/net/wireless/intel/iwlwifi/fw/img.c  | 6 +++---
 drivers/net/wireless/intel/iwlwifi/fw/img.h  | 4 ++++
 3 files changed, 16 insertions(+), 3 deletions(-)
drivers/net/wireless/intel/iwlwifi/fw/dump.c +9 -0
index 016b3a4c5f513..6a37933a02169 100644
@@ -12,6 +12,7 @@
 #include "iwl-io.h"
 #include "iwl-prph.h"
 #include "iwl-csr.h"
+#include "pnvm.h"
 
 /*
  * Note: This structure is read from the device with IO accesses,
@@ -147,6 +148,7 @@ static void iwl_fwrt_dump_umac_error_log(struct iwl_fw_runtime *fwrt)
 	struct iwl_trans *trans = fwrt->trans;
 	struct iwl_umac_error_event_table table = {};
 	u32 base = fwrt->trans->dbg.umac_error_event_table;
+	char pnvm_name[MAX_PNVM_NAME];
 
 	if (!base &&
 	    !(fwrt->trans->dbg.error_event_table_tlv_status &
@@ -164,6 +166,13 @@ static void iwl_fwrt_dump_umac_error_log(struct iwl_fw_runtime *fwrt)
 			fwrt->trans->status, table.valid);
 	}
 
+	if ((table.error_id & ~FW_SYSASSERT_CPU_MASK) ==
+	    FW_SYSASSERT_PNVM_MISSING) {
+		iwl_pnvm_get_fs_name(trans, pnvm_name, sizeof(pnvm_name));
+		IWL_ERR(fwrt, "PNVM data is missing, please install %s\n",
+			pnvm_name);
+	}
+
 	IWL_ERR(fwrt, "0x%08X | %s\n", table.error_id,
 		iwl_fw_lookup_assert_desc(table.error_id));
 	IWL_ERR(fwrt, "0x%08X | umac branchlink1\n", table.blink1);
drivers/net/wireless/intel/iwlwifi/fw/img.c +3 -3
index 24a9666736913..530674a35eeb2 100644
@@ -1,6 +1,6 @@
 // SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause
 /*
- * Copyright(c) 2019 - 2020 Intel Corporation
+ * Copyright(c) 2019 - 2021 Intel Corporation
  */
 
 #include "img.h"
@@ -49,10 +49,9 @@ u8 iwl_fw_lookup_notif_ver(const struct iwl_fw *fw, u8 grp, u8 cmd, u8 def)
 }
 EXPORT_SYMBOL_GPL(iwl_fw_lookup_notif_ver);
 
-#define FW_SYSASSERT_CPU_MASK 0xf0000000
 static const struct {
 	const char *name;
-	u8 num;
+	u32 num;
 } advanced_lookup[] = {
 	{ "NMI_INTERRUPT_WDG", 0x34 },
 	{ "SYSASSERT", 0x35 },
@@ -73,6 +72,7 @@ static const struct {
 	{ "NMI_INTERRUPT_ACTION_PT", 0x7C },
 	{ "NMI_INTERRUPT_UNKNOWN", 0x84 },
 	{ "NMI_INTERRUPT_INST_ACTION_PT", 0x86 },
+	{ "PNVM_MISSING", FW_SYSASSERT_PNVM_MISSING },
 	{ "ADVANCED_SYSASSERT", 0 },
 };
 
drivers/net/wireless/intel/iwlwifi/fw/img.h +4 -0
index 993bda17fa309..fa7b1780064c2 100644
@@ -279,4 +279,8 @@ u8 iwl_fw_lookup_cmd_ver(const struct iwl_fw *fw, u8 grp, u8 cmd, u8 def);
 
 u8 iwl_fw_lookup_notif_ver(const struct iwl_fw *fw, u8 grp, u8 cmd, u8 def);
 const char *iwl_fw_lookup_assert_desc(u32 num);
+
+#define FW_SYSASSERT_CPU_MASK		0xf0000000
+#define FW_SYSASSERT_PNVM_MISSING	0x0010070d
+
 #endif  /* __iwl_fw_img_h__ */
Signature · 1 line
2.34.1
author maintainer
#681 lore raw

PATCH 681/1039 iwlwifi: fix leaks/bad data after failed firmware load

In reply to Greg Kroah-Hartman

From: Johannes Berg <[email protected]>

[ Upstream commit ab07506b0454bea606095951e19e72c282bfbb42 ]

If firmware load fails after having loaded some parts of the firmware, e.g. the IML image, then this would leak. For the host command list we'd end up running into a WARN on the next attempt to load another firmware image.

Fix this by calling iwl_dealloc_ucode() on failures, and make that also clear the data so we start fresh on the next round.

---
 drivers/net/wireless/intel/iwlwifi/iwl-drv.c | 8 ++++++++
 1 file changed, 8 insertions(+)
drivers/net/wireless/intel/iwlwifi/iwl-drv.c +8 -0
index 5cec467b995bb..ad94526a4e1e7 100644
@@ -130,6 +130,9 @@ static void iwl_dealloc_ucode(struct iwl_drv *drv)
 
 	for (i = 0; i < IWL_UCODE_TYPE_MAX; i++)
 		iwl_free_fw_img(drv, drv->fw.img + i);
+
+	/* clear the data for the aborted load case */
+	memset(&drv->fw, 0, sizeof(drv->fw));
 }
 
 static int iwl_alloc_fw_desc(struct iwl_drv *drv, struct fw_desc *desc,
@@ -1375,6 +1378,7 @@ static void iwl_req_fw_callback(const struct firmware *ucode_raw, void *context)
 	int i;
 	bool load_module = false;
 	bool usniffer_images = false;
+	bool failure = true;
 
 	fw->ucode_capa.max_probe_length = IWL_DEFAULT_MAX_PROBE_LENGTH;
 	fw->ucode_capa.standard_phy_calibration_size =
@@ -1644,6 +1648,7 @@ static void iwl_req_fw_callback(const struct firmware *ucode_raw, void *context)
 				op->name, err);
 #endif
 	}
+	failure = false;
 	goto free;
 
  try_again:
@@ -1659,6 +1664,9 @@ static void iwl_req_fw_callback(const struct firmware *ucode_raw, void *context)
 	complete(&drv->request_firmware_complete);
 	device_release_driver(drv->trans->dev);
  free:
+	if (failure)
+		iwl_dealloc_ucode(drv);
+
 	if (pieces) {
 		for (i = 0; i < ARRAY_SIZE(pieces->img); i++)
 			kfree(pieces->img[i].sec);
Signature · 1 line
2.34.1
author maintainer
#682 lore raw

PATCH 682/1039 iwlwifi: remove module loading failure message

In reply to Greg Kroah-Hartman

From: Johannes Berg <[email protected]>

[ Upstream commit 6518f83ffa51131daaf439b66094f684da3fb0ae ]

When CONFIG_DEBUG_TEST_DRIVER_REMOVE is set, iwlwifi crashes when the opmode module cannot be loaded, due to completing the completion before using drv->dev, which can then already be freed.

Fix this by removing the (fairly useless) message. Moving the completion later causes a deadlock instead, so that's not an option.

---
 drivers/net/wireless/intel/iwlwifi/iwl-drv.c | 9 +--------
 1 file changed, 1 insertion(+), 8 deletions(-)
drivers/net/wireless/intel/iwlwifi/iwl-drv.c +1 -8
index ad94526a4e1e7..f53ce9c086947 100644
@@ -1639,15 +1639,8 @@ static void iwl_req_fw_callback(const struct firmware *ucode_raw, void *context)
 	 * else from proceeding if the module fails to load
 	 * or hangs loading.
 	 */
-	if (load_module) {
+	if (load_module)
 		request_module("%s", op->name);
-#ifdef CONFIG_IWLWIFI_OPMODE_MODULAR
-		if (err)
-			IWL_ERR(drv,
-				"failed to load module %s (error %d), is dynamic loading enabled?\n",
-				op->name, err);
-#endif
-	}
 	failure = false;
 	goto free;
 
Signature · 1 line
2.34.1
author maintainer
#683 lore raw

PATCH 683/1039 iwlwifi: mvm: Fix calculation of frame length

In reply to Greg Kroah-Hartman

From: Ilan Peer <[email protected]>

[ Upstream commit 40a0b38d7a7f91a6027287e0df54f5f547e8d27e ]

The RADA might include in the Rx frame the MIC and CRC bytes. These bytes should be removed for non monitor interfaces and should not be passed to mac80211.

Fix the Rx processing to remove the extra bytes on non monitor cases.

---
 drivers/net/wireless/intel/iwlwifi/mvm/rxmq.c | 27 +++++++++++++++++++
 1 file changed, 27 insertions(+)
drivers/net/wireless/intel/iwlwifi/mvm/rxmq.c +27 -0
index e0601f802628c..1e2a55ccf1926 100644
@@ -121,12 +121,39 @@ static int iwl_mvm_create_skb(struct iwl_mvm *mvm, struct sk_buff *skb,
 	struct iwl_rx_mpdu_desc *desc = (void *)pkt->data;
 	unsigned int headlen, fraglen, pad_len = 0;
 	unsigned int hdrlen = ieee80211_hdrlen(hdr->frame_control);
+	u8 mic_crc_len = u8_get_bits(desc->mac_flags1,
+				     IWL_RX_MPDU_MFLG1_MIC_CRC_LEN_MASK) << 1;
 
 	if (desc->mac_flags2 & IWL_RX_MPDU_MFLG2_PAD) {
 		len -= 2;
 		pad_len = 2;
 	}
 
+	/*
+	 * For non monitor interface strip the bytes the RADA might not have
+	 * removed. As monitor interface cannot exist with other interfaces
+	 * this removal is safe.
+	 */
+	if (mic_crc_len && !ieee80211_hw_check(mvm->hw, RX_INCLUDES_FCS)) {
+		u32 pkt_flags = le32_to_cpu(pkt->len_n_flags);
+
+		/*
+		 * If RADA was not enabled then decryption was not performed so
+		 * the MIC cannot be removed.
+		 */
+		if (!(pkt_flags & FH_RSCSR_RADA_EN)) {
+			if (WARN_ON(crypt_len > mic_crc_len))
+				return -EINVAL;
+
+			mic_crc_len -= crypt_len;
+		}
+
+		if (WARN_ON(mic_crc_len > len))
+			return -EINVAL;
+
+		len -= mic_crc_len;
+	}
+
 	/* If frame is small enough to fit in skb->head, pull it completely.
 	 * If not, only pull ieee80211_hdr (including crypto if present, and
 	 * an additional 8 bytes for SNAP/ethertype, see below) so that
Signature · 1 line
2.34.1
author maintainer
#684 lore raw

PATCH 684/1039 iwlwifi: mvm: fix AUX ROC removal

In reply to Greg Kroah-Hartman

From: Avraham Stern <[email protected]>

[ Upstream commit f0337cb48f3bf5f0bbccc985d8a0a8c4aa4934b7 ]

When IWL_UCODE_TLV_CAPA_SESSION_PROT_CMD is set, removing a time event always tries to cancel session protection. However, AUX ROC does not use session protection so it was never removed. As a result, if the driver tries to allocate another AUX ROC event right after cancelling the first one, it will fail with a warning.
In addition, the time event data passed to iwl_mvm_remove_aux_roc_te() is incorrect. Fix it.

---
 drivers/net/wireless/intel/iwlwifi/mvm/time-event.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)
drivers/net/wireless/intel/iwlwifi/mvm/time-event.c +6 -3
index b8c645b9880fc..ab06dcda1462a 100644
@@ -696,11 +696,14 @@ static bool __iwl_mvm_remove_time_event(struct iwl_mvm *mvm,
 	iwl_mvm_te_clear_data(mvm, te_data);
 	spin_unlock_bh(&mvm->time_event_lock);
 
-	/* When session protection is supported, the te_data->id field
+	/* When session protection is used, the te_data->id field
 	 * is reused to save session protection's configuration.
+	 * For AUX ROC, HOT_SPOT_CMD is used and the te_data->id field is set
+	 * to HOT_SPOT_CMD.
 	 */
 	if (fw_has_capa(&mvm->fw->ucode_capa,
-			IWL_UCODE_TLV_CAPA_SESSION_PROT_CMD)) {
+			IWL_UCODE_TLV_CAPA_SESSION_PROT_CMD) &&
+	    id != HOT_SPOT_CMD) {
 		if (mvmvif && id < SESSION_PROTECT_CONF_MAX_ID) {
 			/* Session protection is still ongoing. Cancel it */
 			iwl_mvm_cancel_session_protection(mvm, mvmvif, id);
@@ -1036,7 +1039,7 @@ void iwl_mvm_stop_roc(struct iwl_mvm *mvm, struct ieee80211_vif *vif)
 			iwl_mvm_p2p_roc_finished(mvm);
 		} else {
 			iwl_mvm_remove_aux_roc_te(mvm, mvmvif,
-						  &mvmvif->time_event_data);
+						  &mvmvif->hs_time_event_data);
 			iwl_mvm_roc_finished(mvm);
 		}
 
Signature · 1 line
2.34.1
author maintainer
#685 lore raw

PATCH 685/1039 iwlwifi: pcie: make sure prph_info is set when treating wakeup IRQ

In reply to Greg Kroah-Hartman

From: Luca Coelho <[email protected]>

[ Upstream commit 459fc0f2c6b0f6e280bfa0f230c100c9dfe3a199 ]

In some rare cases when the HW is in a bad state, we may get this interrupt when prph_info is not set yet. Then we will try to dereference it to check the sleep_notif element, which will cause an oops.

Fix that by ignoring the interrupt if prph_info is not set yet.

---
 drivers/net/wireless/intel/iwlwifi/pcie/rx.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)
drivers/net/wireless/intel/iwlwifi/pcie/rx.c +6 -1
index 14602d6d6699f..8247014278f30 100644
@@ -2266,7 +2266,12 @@ irqreturn_t iwl_pcie_irq_msix_handler(int irq, void *dev_id)
 		}
 	}
 
-	if (inta_hw & MSIX_HW_INT_CAUSES_REG_WAKEUP) {
+	/*
+	 * In some rare cases when the HW is in a bad state, we may
+	 * get this interrupt too early, when prph_info is still NULL.
+	 * So make sure that it's not NULL to prevent crashing.
+	 */
+	if (inta_hw & MSIX_HW_INT_CAUSES_REG_WAKEUP && trans_pcie->prph_info) {
 		u32 sleep_notif =
 			le32_to_cpu(trans_pcie->prph_info->sleep_notif);
 		if (sleep_notif == IWL_D3_SLEEP_STATUS_SUSPEND ||
Signature · 1 line
2.34.1
author maintainer
#686 lore raw

PATCH 686/1039 mmc: sdhci-pci-gli: GL9755: Support for CD/WP inversion on OF platforms

In reply to Greg Kroah-Hartman

From: Hector Martin <[email protected]>

[ Upstream commit 189f1d9bc3a5ea3e442e119e4a5deda63da8c462 ]

This is required on some Apple ARM64 laptops using this controller. As is typical on DT platforms, pull these quirks from the device tree using the standard mmc bindings.

See Documentation/devicetree/bindings/mmc/mmc-controller.yaml

---
 drivers/mmc/host/sdhci-pci-gli.c | 11 +++++++++++
 1 file changed, 11 insertions(+)
drivers/mmc/host/sdhci-pci-gli.c +11 -0
index 4fd99c1e82ba3..ad50f16658fe2 100644
@@ -12,6 +12,7 @@
 #include <linux/pci.h>
 #include <linux/mmc/mmc.h>
 #include <linux/delay.h>
+#include <linux/of.h>
 #include "sdhci.h"
 #include "sdhci-pci.h"
 #include "cqhci.h"
@@ -116,6 +117,8 @@
 #define PCI_GLI_9755_PECONF   0x44
 #define   PCI_GLI_9755_LFCLK    GENMASK(14, 12)
 #define   PCI_GLI_9755_DMACLK   BIT(29)
+#define   PCI_GLI_9755_INVERT_CD  BIT(30)
+#define   PCI_GLI_9755_INVERT_WP  BIT(31)
 
 #define PCI_GLI_9755_CFG2          0x48
 #define   PCI_GLI_9755_CFG2_L1DLY    GENMASK(28, 24)
@@ -570,6 +573,14 @@ static void gl9755_hw_setting(struct sdhci_pci_slot *slot)
 	gl9755_wt_on(pdev);
 
 	pci_read_config_dword(pdev, PCI_GLI_9755_PECONF, &value);
+	/*
+	 * Apple ARM64 platforms using these chips may have
+	 * inverted CD/WP detection.
+	 */
+	if (of_property_read_bool(pdev->dev.of_node, "cd-inverted"))
+		value |= PCI_GLI_9755_INVERT_CD;
+	if (of_property_read_bool(pdev->dev.of_node, "wp-inverted"))
+		value |= PCI_GLI_9755_INVERT_WP;
 	value &= ~PCI_GLI_9755_LFCLK;
 	value &= ~PCI_GLI_9755_DMACLK;
 	pci_write_config_dword(pdev, PCI_GLI_9755_PECONF, value);
Signature · 1 line
2.34.1
author maintainer
#687 lore raw

PATCH 687/1039 block: check minor range in device_add_disk()

In reply to Greg Kroah-Hartman

From: Tetsuo Handa <[email protected]>

[ Upstream commit e338924bd05d6e71574bc13e310c89e10e49a8a5 ]

ioctl(fd, LOOP_CTL_ADD, 1048576) causes

sysfs: cannot create duplicate filename '/dev/block/7:0'

message because such request is treated as if ioctl(fd, LOOP_CTL_ADD, 0) due to MINORMASK == 1048575. Verify that all minor numbers for that device fit in the minor range.

---
 block/genhd.c | 2 ++
 1 file changed, 2 insertions(+)
block/genhd.c +2 -0
index 596e43764846b..5308e0920fa6f 100644
@@ -425,6 +425,8 @@ int __must_check device_add_disk(struct device *parent, struct gendisk *disk,
 				DISK_MAX_PARTS);
 			disk->minors = DISK_MAX_PARTS;
 		}
+		if (disk->first_minor + disk->minors > MINORMASK + 1)
+			return -EINVAL;
 	} else {
 		if (WARN_ON(disk->minors))
 			return -EINVAL;
Signature · 1 line
2.34.1
author maintainer
#688 lore raw

PATCH 688/1039 um: registers: Rename function names to avoid conflicts and build problems

In reply to Greg Kroah-Hartman

From: Randy Dunlap <[email protected]>

[ Upstream commit 077b7320942b64b0da182aefd83c374462a65535 ]

The function names init_registers() and restore_registers() are used in several net/ethernet/ and gpu/drm/ drivers for other purposes (not calls to UML functions), so rename them.

This fixes multiple build errors.

---
 arch/um/include/shared/registers.h | 4 ++--
 arch/um/os-Linux/registers.c       | 4 ++--
 arch/um/os-Linux/start_up.c        | 2 +-
 arch/x86/um/syscalls_64.c          | 3 ++-
 4 files changed, 7 insertions(+), 6 deletions(-)
arch/um/include/shared/registers.h +2 -2
index 0c50fa6e8a55b..fbb709a222839 100644
@@ -16,8 +16,8 @@ extern int restore_fp_registers(int pid, unsigned long *fp_regs);
 extern int save_fpx_registers(int pid, unsigned long *fp_regs);
 extern int restore_fpx_registers(int pid, unsigned long *fp_regs);
 extern int save_registers(int pid, struct uml_pt_regs *regs);
-extern int restore_registers(int pid, struct uml_pt_regs *regs);
-extern int init_registers(int pid);
+extern int restore_pid_registers(int pid, struct uml_pt_regs *regs);
+extern int init_pid_registers(int pid);
 extern void get_safe_registers(unsigned long *regs, unsigned long *fp_regs);
 extern unsigned long get_thread_reg(int reg, jmp_buf *buf);
 extern int get_fp_registers(int pid, unsigned long *regs);
arch/um/os-Linux/registers.c +2 -2
index 2d9270508e156..b123955be7acc 100644
@@ -21,7 +21,7 @@ int save_registers(int pid, struct uml_pt_regs *regs)
 	return 0;
 }
 
-int restore_registers(int pid, struct uml_pt_regs *regs)
+int restore_pid_registers(int pid, struct uml_pt_regs *regs)
 {
 	int err;
 
@@ -36,7 +36,7 @@ int restore_registers(int pid, struct uml_pt_regs *regs)
 static unsigned long exec_regs[MAX_REG_NR];
 static unsigned long exec_fp_regs[FP_SIZE];
 
-int init_registers(int pid)
+int init_pid_registers(int pid)
 {
 	int err;
 
arch/um/os-Linux/start_up.c +1 -1
index 8a72c99994eb1..e3ee4db58b40d 100644
@@ -368,7 +368,7 @@ void __init os_early_checks(void)
 	check_tmpexec();
 
 	pid = start_ptraced_child();
-	if (init_registers(pid))
+	if (init_pid_registers(pid))
 		fatal("Failed to initialize default registers");
 	stop_ptraced_child(pid, 1, 1);
 }
arch/x86/um/syscalls_64.c +2 -1
index 58f51667e2e4b..8249685b40960 100644
@@ -11,6 +11,7 @@
 #include <linux/uaccess.h>
 #include <asm/prctl.h> /* XXX This should get the constants from libc */
 #include <os.h>
+#include <registers.h>
 
 long arch_prctl(struct task_struct *task, int option,
 		unsigned long __user *arg2)
@@ -35,7 +36,7 @@ long arch_prctl(struct task_struct *task, int option,
 	switch (option) {
 	case ARCH_SET_FS:
 	case ARCH_SET_GS:
-		ret = restore_registers(pid, &current->thread.regs.regs);
+		ret = restore_pid_registers(pid, &current->thread.regs.regs);
 		if (ret)
 			return ret;
 		break;
Signature · 9 lines
2.34.1




_______________________________________________
linux-um mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-um
author maintainer
#689 lore raw

PATCH 689/1039 ath11k: Fix napi related hang

In reply to Greg Kroah-Hartman

From: Ben Greear <[email protected]>

[ Upstream commit d943fdad7589653065be0e20aadc6dff37725ed4 ]

Similar to the same bug in ath10k, a napi disable w/out it being enabled will hang forever. I believe I saw this while trying rmmod after driver had some failure on startup. Fix it by keeping state on whether napi is enabled or not.

And, remove un-used napi pointer in ath11k driver base struct.

---
 drivers/net/wireless/ath/ath11k/ahb.c  |   12 +++++++++---
 drivers/net/wireless/ath/ath11k/core.h |    2 +-
 drivers/net/wireless/ath/ath11k/pci.c  |   12 +++++++++---
 3 files changed, 19 insertions(+), 7 deletions(-)
drivers/net/wireless/ath/ath11k/ahb.c +9 -3
@@ -175,8 +175,11 @@ static void __ath11k_ahb_ext_irq_disable
 
 		ath11k_ahb_ext_grp_disable(irq_grp);
 
-		napi_synchronize(&irq_grp->napi);
-		napi_disable(&irq_grp->napi);
+		if (irq_grp->napi_enabled) {
+			napi_synchronize(&irq_grp->napi);
+			napi_disable(&irq_grp->napi);
+			irq_grp->napi_enabled = false;
+		}
 	}
 }
 
@@ -300,7 +303,10 @@ static void ath11k_ahb_ext_irq_enable(st
 	for (i = 0; i < ATH11K_EXT_IRQ_GRP_NUM_MAX; i++) {
 		struct ath11k_ext_irq_grp *irq_grp = &ab->ext_irq_grp[i];
 
-		napi_enable(&irq_grp->napi);
+		if (!irq_grp->napi_enabled) {
+			napi_enable(&irq_grp->napi);
+			irq_grp->napi_enabled = true;
+		}
 		ath11k_ahb_ext_grp_enable(irq_grp);
 	}
 }
drivers/net/wireless/ath/ath11k/core.h +1 -1
@@ -141,6 +141,7 @@ struct ath11k_ext_irq_grp {
 	u32 num_irq;
 	u32 grp_id;
 	u64 timestamp;
+	bool napi_enabled;
 	struct napi_struct napi;
 	struct net_device napi_ndev;
 };
@@ -718,7 +719,6 @@ struct ath11k_base {
 	u32 wlan_init_status;
 	int irq_num[ATH11K_IRQ_NUM_MAX];
 	struct ath11k_ext_irq_grp ext_irq_grp[ATH11K_EXT_IRQ_GRP_NUM_MAX];
-	struct napi_struct *napi;
 	struct ath11k_targ_cap target_caps;
 	u32 ext_service_bitmap[WMI_SERVICE_EXT_BM_SIZE];
 	bool pdevs_macaddr_valid;
drivers/net/wireless/ath/ath11k/pci.c +9 -3
@@ -638,8 +638,11 @@ static void __ath11k_pci_ext_irq_disable
 
 		ath11k_pci_ext_grp_disable(irq_grp);
 
-		napi_synchronize(&irq_grp->napi);
-		napi_disable(&irq_grp->napi);
+		if (irq_grp->napi_enabled) {
+			napi_synchronize(&irq_grp->napi);
+			napi_disable(&irq_grp->napi);
+			irq_grp->napi_enabled = false;
+		}
 	}
 }
 
@@ -658,7 +661,10 @@ static void ath11k_pci_ext_irq_enable(st
 	for (i = 0; i < ATH11K_EXT_IRQ_GRP_NUM_MAX; i++) {
 		struct ath11k_ext_irq_grp *irq_grp = &ab->ext_irq_grp[i];
 
-		napi_enable(&irq_grp->napi);
+		if (!irq_grp->napi_enabled) {
+			napi_enable(&irq_grp->napi);
+			irq_grp->napi_enabled = true;
+		}
 		ath11k_pci_ext_grp_enable(irq_grp);
 	}
 }
author maintainer
#690 lore raw

PATCH 690/1039 Bluetooth: btintel: Add missing quirks and msft ext for legacy bootloader

In reply to Greg Kroah-Hartman

From: Tedd Ho-Jeong An <[email protected]>

[ Upstream commit 3547a008c8962df2175db1e78b80f27e027ec549 ]

This patch add missing HCI quirks and MSFT extension for legacy bootloader when it is running in the operational firmware.

---
 drivers/bluetooth/btintel.c | 26 ++++++++++++++++++++++----
 1 file changed, 22 insertions(+), 4 deletions(-)
drivers/bluetooth/btintel.c +22 -4
index b11567b0fd9ab..851a0c9b8fae5 100644
@@ -2493,10 +2493,14 @@ static int btintel_setup_combined(struct hci_dev *hdev)
 	case 0x12:      /* ThP */
 	case 0x13:      /* HrP */
 	case 0x14:      /* CcP */
-		/* Some legacy bootloader devices from JfP supports both old
-		 * and TLV based HCI_Intel_Read_Version command. But we don't
-		 * want to use the TLV based setup routines for those legacy
-		 * bootloader device.
+		/* Some legacy bootloader devices starting from JfP,
+		 * the operational firmware supports both old and TLV based
+		 * HCI_Intel_Read_Version command based on the command
+		 * parameter.
+		 *
+		 * For upgrading firmware case, the TLV based version cannot
+		 * be used because the firmware filename for legacy bootloader
+		 * is based on the old format.
 		 *
 		 * Also, it is not easy to convert TLV based version from the
 		 * legacy version format.
@@ -2508,6 +2512,20 @@ static int btintel_setup_combined(struct hci_dev *hdev)
 		err = btintel_read_version(hdev, &ver);
 		if (err)
 			return err;
+
+		/* Apply the device specific HCI quirks
+		 *
+		 * All Legacy bootloader devices support WBS
+		 */
+		set_bit(HCI_QUIRK_WIDEBAND_SPEECH_SUPPORTED, &hdev->quirks);
+
+		/* Valid LE States quirk for JfP/ThP familiy */
+		if (ver.hw_variant == 0x11 || ver.hw_variant == 0x12)
+			set_bit(HCI_QUIRK_VALID_LE_STATES, &hdev->quirks);
+
+		/* Setup MSFT Extension support */
+		btintel_set_msft_opcode(hdev, ver.hw_variant);
+
 		err = btintel_bootloader_setup(hdev, &ver);
 		break;
 	case 0x17:
Signature · 1 line
2.34.1
author maintainer
#691 lore raw

PATCH 691/1039 cpufreq: intel_pstate: Update cpuinfo.max_freq on HWP_CAP changes

In reply to Greg Kroah-Hartman

From: Rafael J. Wysocki <[email protected]>

[ Upstream commit dfeeedc1bf5772226bddf51ed3f853e5a6707bf1 ]

With HWP enabled, when the turbo range of performance levels is disabled by the platform firmware, the CPU capacity is given by the "guaranteed performance" field in MSR_HWP_CAPABILITIES which is generally dynamic. When it changes, the kernel receives an HWP notification interrupt handled by notify_hwp_interrupt().

When the "guaranteed performance" value changes in the above configuration, the CPU performance scaling needs to be adjusted so as to use the new CPU capacity in computations, which means that the cpuinfo.max_freq value needs to be updated for that CPU.

Accordingly, modify intel_pstate_notify_work() to read MSR_HWP_CAPABILITIES and update cpuinfo.max_freq to reflect the new configuration (this update can be carried out even if the configuration doesn't actually change, because it simply doesn't matter then and it takes less time to update it than to do extra checks to decide whether or not a change has really occurred).

---
 drivers/cpufreq/intel_pstate.c | 24 +++++++++++++++++-------
 1 file changed, 17 insertions(+), 7 deletions(-)
drivers/cpufreq/intel_pstate.c +17 -7
index dec2a5649ac1a..676edc580fd4f 100644
@@ -1124,19 +1124,22 @@ static void intel_pstate_update_policies(void)
 		cpufreq_update_policy(cpu);
 }
 
+static void __intel_pstate_update_max_freq(struct cpudata *cpudata,
+					   struct cpufreq_policy *policy)
+{
+	policy->cpuinfo.max_freq = global.turbo_disabled_mf ?
+			cpudata->pstate.max_freq : cpudata->pstate.turbo_freq;
+	refresh_frequency_limits(policy);
+}
+
 static void intel_pstate_update_max_freq(unsigned int cpu)
 {
 	struct cpufreq_policy *policy = cpufreq_cpu_acquire(cpu);
-	struct cpudata *cpudata;
 
 	if (!policy)
 		return;
 
-	cpudata = all_cpu_data[cpu];
-	policy->cpuinfo.max_freq = global.turbo_disabled_mf ?
-			cpudata->pstate.max_freq : cpudata->pstate.turbo_freq;
-
-	refresh_frequency_limits(policy);
+	__intel_pstate_update_max_freq(all_cpu_data[cpu], policy);
 
 	cpufreq_cpu_release(policy);
 }
@@ -1584,8 +1587,15 @@ static void intel_pstate_notify_work(struct work_struct *work)
 {
 	struct cpudata *cpudata =
 		container_of(to_delayed_work(work), struct cpudata, hwp_notify_work);
+	struct cpufreq_policy *policy = cpufreq_cpu_acquire(cpudata->cpu);
+
+	if (policy) {
+		intel_pstate_get_hwp_cap(cpudata);
+		__intel_pstate_update_max_freq(cpudata, policy);
+
+		cpufreq_cpu_release(policy);
+	}
 
-	cpufreq_update_policy(cpudata->cpu);
 	wrmsrl_on_cpu(cpudata->cpu, MSR_HWP_STATUS, 0);
 }
 
Signature · 1 line
2.34.1
author maintainer
#692 lore raw

PATCH 692/1039 Bluetooth: vhci: Set HCI_QUIRK_VALID_LE_STATES

In reply to Greg Kroah-Hartman

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

[ Upstream commit cfb4c313be670fd4bd09650216620fa4514cdb93 ]

This set HCI_QUIRK_VALID_LE_STATES quirk which is required for the likes of experimental LE simultaneous roles.

---
 drivers/bluetooth/hci_vhci.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/bluetooth/hci_vhci.c +2 -0
index 5fd91106e853d..95af01bdd02a2 100644
@@ -239,6 +239,8 @@ static int __vhci_create_device(struct vhci_data *data, __u8 opcode)
 	if (opcode & 0x80)
 		set_bit(HCI_QUIRK_RAW_DEVICE, &hdev->quirks);
 
+	set_bit(HCI_QUIRK_VALID_LE_STATES, &hdev->quirks);
+
 	if (hci_register_dev(hdev) < 0) {
 		BT_ERR("Can't register HCI device");
 		hci_free_dev(hdev);
Signature · 1 line
2.34.1
author maintainer
#693 lore raw

PATCH 693/1039 xfrm: rate limit SA mapping change message to user space

In reply to Greg Kroah-Hartman

From: Antony Antony <[email protected]>

[ Upstream commit 4e484b3e969b52effd95c17f7a86f39208b2ccf4 ]

Kernel generates mapping change message, XFRM_MSG_MAPPING, when a source port chage is detected on a input state with UDP encapsulation set. Kernel generates a message for each IPsec packet with new source port. For a high speed flow per packet mapping change message can be excessive, and can overload the user space listener.

Introduce rate limiting for XFRM_MSG_MAPPING message to the user space.

The rate limiting is configurable via netlink, when adding a new SA or updating it. Use the new attribute XFRMA_MTIMER_THRESH in seconds.

v1->v2 change:
update xfrm_sa_len()

v2->v3 changes:
use u32 insted unsigned long to reduce size of struct xfrm_state fix xfrm_ompat size Reported-by: kernel test robot <[email protected]> accept XFRM_MSG_MAPPING only when XFRMA_ENCAP is present

---
 include/net/xfrm.h        |  5 +++++
 include/uapi/linux/xfrm.h |  1 +
 net/xfrm/xfrm_compat.c    |  6 ++++--
 net/xfrm/xfrm_state.c     | 23 ++++++++++++++++++++++-
 net/xfrm/xfrm_user.c      | 18 +++++++++++++++++-
 5 files changed, 49 insertions(+), 4 deletions(-)
include/net/xfrm.h +5 -0
index 2308210793a01..2589e4c0501bd 100644
@@ -200,6 +200,11 @@ struct xfrm_state {
 	struct xfrm_algo_aead	*aead;
 	const char		*geniv;
 
+	/* mapping change rate limiting */
+	__be16 new_mapping_sport;
+	u32 new_mapping;	/* seconds */
+	u32 mapping_maxage;	/* seconds for input SA */
+
 	/* Data for encapsulator */
 	struct xfrm_encap_tmpl	*encap;
 	struct sock __rcu	*encap_sk;
include/uapi/linux/xfrm.h +1 -0
index eda0426ec4c2b..4e29d78518902 100644
@@ -313,6 +313,7 @@ enum xfrm_attr_type_t {
 	XFRMA_SET_MARK,		/* __u32 */
 	XFRMA_SET_MARK_MASK,	/* __u32 */
 	XFRMA_IF_ID,		/* __u32 */
+	XFRMA_MTIMER_THRESH,	/* __u32 in seconds for input SA */
 	__XFRMA_MAX
 
 #define XFRMA_OUTPUT_MARK XFRMA_SET_MARK	/* Compatibility */
net/xfrm/xfrm_compat.c +4 -2
index 2bf2693901631..a0f62fa02e06e 100644
@@ -127,6 +127,7 @@ static const struct nla_policy compat_policy[XFRMA_MAX+1] = {
 	[XFRMA_SET_MARK]	= { .type = NLA_U32 },
 	[XFRMA_SET_MARK_MASK]	= { .type = NLA_U32 },
 	[XFRMA_IF_ID]		= { .type = NLA_U32 },
+	[XFRMA_MTIMER_THRESH]	= { .type = NLA_U32 },
 };
 
 static struct nlmsghdr *xfrm_nlmsg_put_compat(struct sk_buff *skb,
@@ -274,9 +275,10 @@ static int xfrm_xlate64_attr(struct sk_buff *dst, const struct nlattr *src)
 	case XFRMA_SET_MARK:
 	case XFRMA_SET_MARK_MASK:
 	case XFRMA_IF_ID:
+	case XFRMA_MTIMER_THRESH:
 		return xfrm_nla_cpy(dst, src, nla_len(src));
 	default:
-		BUILD_BUG_ON(XFRMA_MAX != XFRMA_IF_ID);
+		BUILD_BUG_ON(XFRMA_MAX != XFRMA_MTIMER_THRESH);
 		pr_warn_once("unsupported nla_type %d\n", src->nla_type);
 		return -EOPNOTSUPP;
 	}
@@ -431,7 +433,7 @@ static int xfrm_xlate32_attr(void *dst, const struct nlattr *nla,
 	int err;
 
 	if (type > XFRMA_MAX) {
-		BUILD_BUG_ON(XFRMA_MAX != XFRMA_IF_ID);
+		BUILD_BUG_ON(XFRMA_MAX != XFRMA_MTIMER_THRESH);
 		NL_SET_ERR_MSG(extack, "Bad attribute");
 		return -EOPNOTSUPP;
 	}
net/xfrm/xfrm_state.c +22 -1
index a2f4001221d16..78d51399a0f4b 100644
@@ -1593,6 +1593,9 @@ static struct xfrm_state *xfrm_state_clone(struct xfrm_state *orig,
 	x->km.seq = orig->km.seq;
 	x->replay = orig->replay;
 	x->preplay = orig->preplay;
+	x->mapping_maxage = orig->mapping_maxage;
+	x->new_mapping = 0;
+	x->new_mapping_sport = 0;
 
 	return x;
 
@@ -2242,7 +2245,7 @@ int km_query(struct xfrm_state *x, struct xfrm_tmpl *t, struct xfrm_policy *pol)
 }
 EXPORT_SYMBOL(km_query);
 
-int km_new_mapping(struct xfrm_state *x, xfrm_address_t *ipaddr, __be16 sport)
+static int __km_new_mapping(struct xfrm_state *x, xfrm_address_t *ipaddr, __be16 sport)
 {
 	int err = -EINVAL;
 	struct xfrm_mgr *km;
@@ -2257,6 +2260,24 @@ int km_new_mapping(struct xfrm_state *x, xfrm_address_t *ipaddr, __be16 sport)
 	rcu_read_unlock();
 	return err;
 }
+
+int km_new_mapping(struct xfrm_state *x, xfrm_address_t *ipaddr, __be16 sport)
+{
+	int ret = 0;
+
+	if (x->mapping_maxage) {
+		if ((jiffies / HZ - x->new_mapping) > x->mapping_maxage ||
+		    x->new_mapping_sport != sport) {
+			x->new_mapping_sport = sport;
+			x->new_mapping = jiffies / HZ;
+			ret = __km_new_mapping(x, ipaddr, sport);
+		}
+	} else {
+		ret = __km_new_mapping(x, ipaddr, sport);
+	}
+
+	return ret;
+}
 EXPORT_SYMBOL(km_new_mapping);
 
 void km_policy_expired(struct xfrm_policy *pol, int dir, int hard, u32 portid)
net/xfrm/xfrm_user.c +17 -1
index b571892fe10f5..c60441be883a8 100644
@@ -282,6 +282,10 @@ static int verify_newsa_info(struct xfrm_usersa_info *p,
 
 	err = 0;
 
+	if (attrs[XFRMA_MTIMER_THRESH])
+		if (!attrs[XFRMA_ENCAP])
+			err = -EINVAL;
+
 out:
 	return err;
 }
@@ -521,6 +525,7 @@ static void xfrm_update_ae_params(struct xfrm_state *x, struct nlattr **attrs,
 	struct nlattr *lt = attrs[XFRMA_LTIME_VAL];
 	struct nlattr *et = attrs[XFRMA_ETIMER_THRESH];
 	struct nlattr *rt = attrs[XFRMA_REPLAY_THRESH];
+	struct nlattr *mt = attrs[XFRMA_MTIMER_THRESH];
 
 	if (re) {
 		struct xfrm_replay_state_esn *replay_esn;
@@ -552,6 +557,9 @@ static void xfrm_update_ae_params(struct xfrm_state *x, struct nlattr **attrs,
 
 	if (rt)
 		x->replay_maxdiff = nla_get_u32(rt);
+
+	if (mt)
+		x->mapping_maxage = nla_get_u32(mt);
 }
 
 static void xfrm_smark_init(struct nlattr **attrs, struct xfrm_mark *m)
@@ -1029,8 +1037,13 @@ static int copy_to_user_state_extra(struct xfrm_state *x,
 		if (ret)
 			goto out;
 	}
-	if (x->security)
+	if (x->security) {
 		ret = copy_sec_ctx(x->security, skb);
+		if (ret)
+			goto out;
+	}
+	if (x->mapping_maxage)
+		ret = nla_put_u32(skb, XFRMA_MTIMER_THRESH, x->mapping_maxage);
 out:
 	return ret;
 }
@@ -3084,6 +3097,9 @@ static inline unsigned int xfrm_sa_len(struct xfrm_state *x)
 	/* Must count x->lastused as it may become non-zero behind our back. */
 	l += nla_total_size_64bit(sizeof(u64));
 
+	if (x->mapping_maxage)
+		l += nla_total_size(sizeof(x->mapping_maxage));
+
 	return l;
 }
 
Signature · 1 line
2.34.1
author maintainer
#694 lore raw

PATCH 694/1039 drm/etnaviv: consider completed fence seqno in hang check

In reply to Greg Kroah-Hartman

From: Lucas Stach <[email protected]>

[ Upstream commit cdd156955f946beaa5f3a00d8ccf90e5a197becc ]

Some GPU heavy test programs manage to trigger the hangcheck quite often. If there are no other GPU users in the system and the test program exhibits a very regular structure in the commandstreams that are being submitted, we can end up with two distinct submits managing to trigger the hangcheck with the FE in a very similar address range. This leads the hangcheck to believe that the GPU is stuck, while in reality the GPU is already busy working on a different job. To avoid those spurious GPU resets, also remember and consider the last completed fence seqno in the hang check.

---
 drivers/gpu/drm/etnaviv/etnaviv_gpu.h   | 1 +
 drivers/gpu/drm/etnaviv/etnaviv_sched.c | 4 +++-
 2 files changed, 4 insertions(+), 1 deletion(-)
drivers/gpu/drm/etnaviv/etnaviv_gpu.h +1 -0
index 1c75c8ed5bcea..85eddd492774d 100644
@@ -130,6 +130,7 @@ struct etnaviv_gpu {
 
 	/* hang detection */
 	u32 hangcheck_dma_addr;
+	u32 hangcheck_fence;
 
 	void __iomem *mmio;
 	int irq;
drivers/gpu/drm/etnaviv/etnaviv_sched.c +3 -1
index 180bb633d5c53..58f593b278c15 100644
@@ -107,8 +107,10 @@ static enum drm_gpu_sched_stat etnaviv_sched_timedout_job(struct drm_sched_job
 	 */
 	dma_addr = gpu_read(gpu, VIVS_FE_DMA_ADDRESS);
 	change = dma_addr - gpu->hangcheck_dma_addr;
-	if (change < 0 || change > 16) {
+	if (gpu->completed_fence != gpu->hangcheck_fence ||
+	    change < 0 || change > 16) {
 		gpu->hangcheck_dma_addr = dma_addr;
+		gpu->hangcheck_fence = gpu->completed_fence;
 		goto out_no_timeout;
 	}
 
Signature · 1 line
2.34.1
author maintainer
#695 lore raw

PATCH 695/1039 jffs2: GC deadlock reading a page that is used in jffs2_write_begin()

In reply to Greg Kroah-Hartman

From: Kyeong Yoo <[email protected]>

[ Upstream commit aa39cc675799bc92da153af9a13d6f969c348e82 ]

GC task can deadlock in read_cache_page() because it may attempt to release a page that is actually allocated by another task in jffs2_write_begin().
The reason is that in jffs2_write_begin() there is a small window a cache page is allocated for use but not set Uptodate yet.

This ends up with a deadlock between two tasks:
1) A task (e.g. file copy)
- jffs2_write_begin() locks a cache page
- jffs2_write_end() tries to lock "alloc_sem" from

	 jffs2_reserve_space() <-- STUCK

2) GC task (jffs2_gcd_mtd3)
- jffs2_garbage_collect_pass() locks "alloc_sem"
- try to lock the same cache page in read_cache_page() <-- STUCK

So to avoid this deadlock, hold "alloc_sem" in jffs2_write_begin() while reading data in a cache page.

---
 fs/jffs2/file.c | 40 +++++++++++++++++++++++++---------------
 1 file changed, 25 insertions(+), 15 deletions(-)
fs/jffs2/file.c +25 -15
index 4fc8cd698d1a4..bd7d58d27bfc6 100644
@@ -136,20 +136,15 @@ static int jffs2_write_begin(struct file *filp, struct address_space *mapping,
 	struct page *pg;
 	struct inode *inode = mapping->host;
 	struct jffs2_inode_info *f = JFFS2_INODE_INFO(inode);
+	struct jffs2_sb_info *c = JFFS2_SB_INFO(inode->i_sb);
 	pgoff_t index = pos >> PAGE_SHIFT;
 	uint32_t pageofs = index << PAGE_SHIFT;
 	int ret = 0;
 
-	pg = grab_cache_page_write_begin(mapping, index, flags);
-	if (!pg)
-		return -ENOMEM;
-	*pagep = pg;
-
 	jffs2_dbg(1, "%s()\n", __func__);
 
 	if (pageofs > inode->i_size) {
 		/* Make new hole frag from old EOF to new page */
-		struct jffs2_sb_info *c = JFFS2_SB_INFO(inode->i_sb);
 		struct jffs2_raw_inode ri;
 		struct jffs2_full_dnode *fn;
 		uint32_t alloc_len;
@@ -160,7 +155,7 @@ static int jffs2_write_begin(struct file *filp, struct address_space *mapping,
 		ret = jffs2_reserve_space(c, sizeof(ri), &alloc_len,
 					  ALLOC_NORMAL, JFFS2_SUMMARY_INODE_SIZE);
 		if (ret)
-			goto out_page;
+			goto out_err;
 
 		mutex_lock(&f->sem);
 		memset(&ri, 0, sizeof(ri));
@@ -190,7 +185,7 @@ static int jffs2_write_begin(struct file *filp, struct address_space *mapping,
 			ret = PTR_ERR(fn);
 			jffs2_complete_reservation(c);
 			mutex_unlock(&f->sem);
-			goto out_page;
+			goto out_err;
 		}
 		ret = jffs2_add_full_dnode_to_inode(c, f, fn);
 		if (f->metadata) {
@@ -205,13 +200,26 @@ static int jffs2_write_begin(struct file *filp, struct address_space *mapping,
 			jffs2_free_full_dnode(fn);
 			jffs2_complete_reservation(c);
 			mutex_unlock(&f->sem);
-			goto out_page;
+			goto out_err;
 		}
 		jffs2_complete_reservation(c);
 		inode->i_size = pageofs;
 		mutex_unlock(&f->sem);
 	}
 
+	/*
+	 * While getting a page and reading data in, lock c->alloc_sem until
+	 * the page is Uptodate. Otherwise GC task may attempt to read the same
+	 * page in read_cache_page(), which causes a deadlock.
+	 */
+	mutex_lock(&c->alloc_sem);
+	pg = grab_cache_page_write_begin(mapping, index, flags);
+	if (!pg) {
+		ret = -ENOMEM;
+		goto release_sem;
+	}
+	*pagep = pg;
+
 	/*
 	 * Read in the page if it wasn't already present. Cannot optimize away
 	 * the whole page write case until jffs2_write_end can handle the
@@ -221,15 +229,17 @@ static int jffs2_write_begin(struct file *filp, struct address_space *mapping,
 		mutex_lock(&f->sem);
 		ret = jffs2_do_readpage_nolock(inode, pg);
 		mutex_unlock(&f->sem);
-		if (ret)
-			goto out_page;
+		if (ret) {
+			unlock_page(pg);
+			put_page(pg);
+			goto release_sem;
+		}
 	}
 	jffs2_dbg(1, "end write_begin(). pg->flags %lx\n", pg->flags);
-	return ret;
 
-out_page:
-	unlock_page(pg);
-	put_page(pg);
+release_sem:
+	mutex_unlock(&c->alloc_sem);
+out_err:
 	return ret;
 }
 
Signature · 1 line
2.34.1
author maintainer
#696 lore raw

PATCH 696/1039 ACPICA: actypes.h: Expand the ACPI_ACCESS_ definitions

In reply to Greg Kroah-Hartman

From: Mark Langsdorf <[email protected]>

[ Upstream commit f81bdeaf816142e0729eea0cc84c395ec9673151 ]

ACPICA commit bc02c76d518135531483dfc276ed28b7ee632ce1

The current ACPI_ACCESS_*_WIDTH defines do not provide a way to test that size is small enough to not cause an overflow when applied to a 32-bit integer.

Rather than adding more magic numbers, add ACPI_ACCESS_*_SHIFT, ACPI_ACCESS_*_MAX, and ACPI_ACCESS_*_DEFAULT #defines and redefine ACPI_ACCESS_*_WIDTH in terms of the new #defines.

This was inititally reported on Linux where a size of 102 in ACPI_ACCESS_BIT_WIDTH caused an overflow error in the SPCR initialization code.

---
 include/acpi/actypes.h | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)
include/acpi/actypes.h +8 -2
index ff8b3c913f217..248242dca28d3 100644
@@ -536,8 +536,14 @@ typedef u64 acpi_integer;
  * Can be used with access_width of struct acpi_generic_address and access_size of
  * struct acpi_resource_generic_register.
  */
-#define ACPI_ACCESS_BIT_WIDTH(size)     (1 << ((size) + 2))
-#define ACPI_ACCESS_BYTE_WIDTH(size)    (1 << ((size) - 1))
+#define ACPI_ACCESS_BIT_SHIFT		2
+#define ACPI_ACCESS_BYTE_SHIFT		-1
+#define ACPI_ACCESS_BIT_MAX		(31 - ACPI_ACCESS_BIT_SHIFT)
+#define ACPI_ACCESS_BYTE_MAX		(31 - ACPI_ACCESS_BYTE_SHIFT)
+#define ACPI_ACCESS_BIT_DEFAULT		(8 - ACPI_ACCESS_BIT_SHIFT)
+#define ACPI_ACCESS_BYTE_DEFAULT	(8 - ACPI_ACCESS_BYTE_SHIFT)
+#define ACPI_ACCESS_BIT_WIDTH(size)	(1 << ((size) + ACPI_ACCESS_BIT_SHIFT))
+#define ACPI_ACCESS_BYTE_WIDTH(size)	(1 << ((size) + ACPI_ACCESS_BYTE_SHIFT))
 
 /*******************************************************************************
  *
Signature · 1 line
2.34.1
author maintainer
#697 lore raw

PATCH 697/1039 ACPICA: Utilities: Avoid deleting the same object twice in a row

In reply to Greg Kroah-Hartman

From: Rafael J. Wysocki <[email protected]>

[ Upstream commit 1cdfe9e346b4c5509ffe19ccde880fd259d9f7a3 ]

ACPICA commit c11af67d8f7e3d381068ce7771322f2b5324d687

If original_count is 0 in acpi_ut_update_ref_count (), acpi_ut_delete_internal_obj () is invoked for the target object, which is incorrect, because that object has been deleted once already and the memory allocated to store it may have been reclaimed and allocated for a different purpose by the host OS. Moreover, a confusing debug message following the "Reference Count is already zero, cannot decrement" warning is printed in that case.

To fix this issue, make acpi_ut_update_ref_count () return after finding that original_count is 0 and printing the above warning.

---
 drivers/acpi/acpica/utdelete.c | 1 +
 1 file changed, 1 insertion(+)
drivers/acpi/acpica/utdelete.c +1 -0
index e5ba9795ec696..8d7736d2d2699 100644
@@ -422,6 +422,7 @@ acpi_ut_update_ref_count(union acpi_operand_object *object, u32 action)
 			ACPI_WARNING((AE_INFO,
 				      "Obj %p, Reference Count is already zero, cannot decrement\n",
 				      object));
+			return;
 		}
 
 		ACPI_DEBUG_PRINT_RAW((ACPI_DB_ALLOCATIONS,
Signature · 1 line
2.34.1
author maintainer
#698 lore raw

PATCH 698/1039 ACPICA: Executer: Fix the REFCLASS_REFOF case in acpi_ex_opcode_1A_0T_1R()

In reply to Greg Kroah-Hartman

From: Rafael J. Wysocki <[email protected]>

[ Upstream commit 24ea5f90ec9548044a6209685c5010edd66ffe8f ]

ACPICA commit d984f12041392fa4156b52e2f7e5c5e7bc38ad9e

If Operand[0] is a reference of the ACPI_REFCLASS_REFOF class, acpi_ex_opcode_1A_0T_1R () calls acpi_ns_get_attached_object () to obtain return_desc which may require additional resolution with the help of acpi_ex_read_data_from_field (). If the latter fails, the reference counter of the original return_desc is decremented which is incorrect, because acpi_ns_get_attached_object () does not increment the reference counter of the object returned by it.

This issue may lead to premature deletion of the attached object while it is still attached and a use-after-free and crash in the host OS. For example, this may happen when on evaluation of ref_of() a local region field where there is no registered handler for the given Operation Region.

Fix it by making acpi_ex_opcode_1A_0T_1R () return Status right away after a acpi_ex_read_data_from_field () failure.

---
 drivers/acpi/acpica/exoparg1.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/acpi/acpica/exoparg1.c +2 -1
index b639e930d6429..44b7c350ed5ca 100644
@@ -1007,7 +1007,8 @@ acpi_status acpi_ex_opcode_1A_0T_1R(struct acpi_walk_state *walk_state)
 						    (walk_state, return_desc,
 						     &temp_desc);
 						if (ACPI_FAILURE(status)) {
-							goto cleanup;
+							return_ACPI_STATUS
+							    (status);
 						}
 
 						return_desc = temp_desc;
Signature · 1 line
2.34.1
author maintainer
#699 lore raw

PATCH 699/1039 ACPICA: Fix wrong interpretation of PCC address

In reply to Greg Kroah-Hartman

From: Sudeep Holla <[email protected]>

[ Upstream commit 9a3b8655db1ada31c82189ae13f40eb25da48c35 ]

ACPICA commit 41be6afacfdaec2dba3a5ed368736babc2a7aa5c

With the PCC Opregion in the firmware and we are hitting below kernel crash:

-->8

Unable to handle kernel NULL pointer dereference at virtual address 0000000000000010
 Workqueue: pm pm_runtime_work
 pstate: 80000005 (Nzcv daif -PAN -UAO -TCO -DIT -SSBS BTYPE=--)
 pc : __memcpy+0x54/0x260
 lr : acpi_ex_write_data_to_field+0xb8/0x194
 Call trace:
  __memcpy+0x54/0x260
  acpi_ex_store_object_to_node+0xa4/0x1d4
  acpi_ex_store+0x44/0x164
  acpi_ex_opcode_1A_1T_1R+0x25c/0x508
  acpi_ds_exec_end_op+0x1b4/0x44c
  acpi_ps_parse_loop+0x3a8/0x614
  acpi_ps_parse_aml+0x90/0x2f4
  acpi_ps_execute_method+0x11c/0x19c
  acpi_ns_evaluate+0x1ec/0x2b0
  acpi_evaluate_object+0x170/0x2b0
  acpi_device_set_power+0x118/0x310
  acpi_dev_suspend+0xd4/0x180
  acpi_subsys_runtime_suspend+0x28/0x38
  __rpm_callback+0x74/0x328
  rpm_suspend+0x2d8/0x624
  pm_runtime_work+0xa4/0xb8
  process_one_work+0x194/0x25c
  worker_thread+0x260/0x49c
  kthread+0x14c/0x30c
  ret_from_fork+0x10/0x20
 Code: f9000006 f81f80a7 d65f03c0 361000c2 (b9400026)
 ---[ end trace 24d8a032fa77b68a ]---

The reason for the crash is that the PCC channel index passed via region.address in acpi_ex_store_object_to_node is interpreted as the channel subtype incorrectly.

Assuming the PCC op_region support is not used by any other type, let us remove the subtype check as the AML has no access to the subtype information. Once we remove it, the kernel crash disappears and correctly complains about missing PCC Opregion handler.

ACPI Error: No handler for Region [PFRM] ((____ptrval____)) [PCC] (20210730/evregion-130) ACPI Error: Region PCC (ID=10) has no handler (20210730/exfldio-261)
ACPI Error: Aborting method \_SB.ETH0._PS3 due to previous error (AE_NOT_EXIST) (20210730/psparse-531)

---
 drivers/acpi/acpica/exfield.c | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)
drivers/acpi/acpica/exfield.c +1 -6
index 06f3c9df1e22d..8618500f23b39 100644
@@ -330,12 +330,7 @@ acpi_ex_write_data_to_field(union acpi_operand_object *source_desc,
 		       obj_desc->field.base_byte_offset,
 		       source_desc->buffer.pointer, data_length);
 
-		if ((obj_desc->field.region_obj->region.address ==
-		     PCC_MASTER_SUBSPACE
-		     && MASTER_SUBSPACE_COMMAND(obj_desc->field.
-						base_byte_offset))
-		    || GENERIC_SUBSPACE_COMMAND(obj_desc->field.
-						base_byte_offset)) {
+		if (MASTER_SUBSPACE_COMMAND(obj_desc->field.base_byte_offset)) {
 
 			/* Perform the write */
 
Signature · 1 line
2.34.1
author maintainer
#700 lore raw

PATCH 700/1039 ACPICA: Hardware: Do not flush CPU cache when entering S4 and S5

In reply to Greg Kroah-Hartman

From: Kirill A. Shutemov <[email protected]>

[ Upstream commit 1d4e0b3abb168b2ee1eca99c527cffa1b80b6161 ]

ACPICA commit 3dd7e1f3996456ef81bfe14cba29860e8d42949e

According to ACPI 6.4, Section 16.2, the CPU cache flushing is required on entering to S1, S2, and S3, but the ACPICA code flushes the CPU cache regardless of the sleep state.

Blind cache flush on entering S5 causes problems for TDX.

Flushing happens with WBINVD that is not supported in the TDX environment.

TDX only supports S5 and adjusting ACPICA code to conform to the spec more strictly fixes the issue.

---
 drivers/acpi/acpica/hwesleep.c  | 4 +++-
 drivers/acpi/acpica/hwsleep.c   | 4 +++-
 drivers/acpi/acpica/hwxfsleep.c | 2 --
 3 files changed, 6 insertions(+), 4 deletions(-)
drivers/acpi/acpica/hwesleep.c +3 -1
index 808fdf54aeebf..7ee2939c08cd4 100644
@@ -104,7 +104,9 @@ acpi_status acpi_hw_extended_sleep(u8 sleep_state)
 
 	/* Flush caches, as per ACPI specification */
 
-	ACPI_FLUSH_CPU_CACHE();
+	if (sleep_state < ACPI_STATE_S4) {
+		ACPI_FLUSH_CPU_CACHE();
+	}
 
 	status = acpi_os_enter_sleep(sleep_state, sleep_control, 0);
 	if (status == AE_CTRL_TERMINATE) {
drivers/acpi/acpica/hwsleep.c +3 -1
index 34a3825f25d37..5efa3d8e483e0 100644
@@ -110,7 +110,9 @@ acpi_status acpi_hw_legacy_sleep(u8 sleep_state)
 
 	/* Flush caches, as per ACPI specification */
 
-	ACPI_FLUSH_CPU_CACHE();
+	if (sleep_state < ACPI_STATE_S4) {
+		ACPI_FLUSH_CPU_CACHE();
+	}
 
 	status = acpi_os_enter_sleep(sleep_state, pm1a_control, pm1b_control);
 	if (status == AE_CTRL_TERMINATE) {
drivers/acpi/acpica/hwxfsleep.c +0 -2
index e4cde23a29061..ba77598ee43e8 100644
@@ -162,8 +162,6 @@ acpi_status acpi_enter_sleep_state_s4bios(void)
 		return_ACPI_STATUS(status);
 	}
 
-	ACPI_FLUSH_CPU_CACHE();
-
 	status = acpi_hw_write_port(acpi_gbl_FADT.smi_command,
 				    (u32)acpi_gbl_FADT.s4_bios_request, 8);
 	if (ACPI_FAILURE(status)) {
Signature · 1 line
2.34.1
author maintainer
#701 lore raw

PATCH 701/1039 mmc: mtk-sd: Use readl_poll_timeout instead of open-coded polling

In reply to Greg Kroah-Hartman

From: AngeloGioacchino Del Regno <[email protected]>

[ Upstream commit ffaea6ebfe9ce06ebb3a54811a47688f2b0893cd ]

Replace all instances of open-coded while loops for polling registers with calls to readl_poll_timeout() and, while at it, also fix some possible infinite loop instances.

---
 drivers/mmc/host/mtk-sd.c | 64 ++++++++++++++++++++++++---------------
 1 file changed, 39 insertions(+), 25 deletions(-)
drivers/mmc/host/mtk-sd.c +39 -25
index 632775217d35c..d5a9c269d4926 100644
@@ -636,12 +636,11 @@ static void msdc_reset_hw(struct msdc_host *host)
 	u32 val;
 
 	sdr_set_bits(host->base + MSDC_CFG, MSDC_CFG_RST);
-	while (readl(host->base + MSDC_CFG) & MSDC_CFG_RST)
-		cpu_relax();
+	readl_poll_timeout(host->base + MSDC_CFG, val, !(val & MSDC_CFG_RST), 0, 0);
 
 	sdr_set_bits(host->base + MSDC_FIFOCS, MSDC_FIFOCS_CLR);
-	while (readl(host->base + MSDC_FIFOCS) & MSDC_FIFOCS_CLR)
-		cpu_relax();
+	readl_poll_timeout(host->base + MSDC_FIFOCS, val,
+			   !(val & MSDC_FIFOCS_CLR), 0, 0);
 
 	val = readl(host->base + MSDC_INT);
 	writel(val, host->base + MSDC_INT);
@@ -814,8 +813,9 @@ static void msdc_gate_clock(struct msdc_host *host)
 	clk_disable_unprepare(host->h_clk);
 }
 
-static void msdc_ungate_clock(struct msdc_host *host)
+static int msdc_ungate_clock(struct msdc_host *host)
 {
+	u32 val;
 	int ret;
 
 	clk_prepare_enable(host->h_clk);
@@ -825,11 +825,11 @@ static void msdc_ungate_clock(struct msdc_host *host)
 	ret = clk_bulk_prepare_enable(MSDC_NR_CLOCKS, host->bulk_clks);
 	if (ret) {
 		dev_err(host->dev, "Cannot enable pclk/axi/ahb clock gates\n");
-		return;
+		return ret;
 	}
 
-	while (!(readl(host->base + MSDC_CFG) & MSDC_CFG_CKSTB))
-		cpu_relax();
+	return readl_poll_timeout(host->base + MSDC_CFG, val,
+				  (val & MSDC_CFG_CKSTB), 1, 20000);
 }
 
 static void msdc_set_mclk(struct msdc_host *host, unsigned char timing, u32 hz)
@@ -840,6 +840,7 @@ static void msdc_set_mclk(struct msdc_host *host, unsigned char timing, u32 hz)
 	u32 div;
 	u32 sclk;
 	u32 tune_reg = host->dev_comp->pad_tune_reg;
+	u32 val;
 
 	if (!hz) {
 		dev_dbg(host->dev, "set mclk to 0\n");
@@ -920,8 +921,7 @@ static void msdc_set_mclk(struct msdc_host *host, unsigned char timing, u32 hz)
 	else
 		clk_prepare_enable(clk_get_parent(host->src_clk));
 
-	while (!(readl(host->base + MSDC_CFG) & MSDC_CFG_CKSTB))
-		cpu_relax();
+	readl_poll_timeout(host->base + MSDC_CFG, val, (val & MSDC_CFG_CKSTB), 0, 0);
 	sdr_set_bits(host->base + MSDC_CFG, MSDC_CFG_CKPDN);
 	mmc->actual_clock = sclk;
 	host->mclk = hz;
@@ -1231,13 +1231,13 @@ static bool msdc_cmd_done(struct msdc_host *host, int events,
 static inline bool msdc_cmd_is_ready(struct msdc_host *host,
 		struct mmc_request *mrq, struct mmc_command *cmd)
 {
-	/* The max busy time we can endure is 20ms */
-	unsigned long tmo = jiffies + msecs_to_jiffies(20);
+	u32 val;
+	int ret;
 
-	while ((readl(host->base + SDC_STS) & SDC_STS_CMDBUSY) &&
-			time_before(jiffies, tmo))
-		cpu_relax();
-	if (readl(host->base + SDC_STS) & SDC_STS_CMDBUSY) {
+	/* The max busy time we can endure is 20ms */
+	ret = readl_poll_timeout_atomic(host->base + SDC_STS, val,
+					!(val & SDC_STS_CMDBUSY), 1, 20000);
+	if (ret) {
 		dev_err(host->dev, "CMD bus busy detected\n");
 		host->error |= REQ_CMD_BUSY;
 		msdc_cmd_done(host, MSDC_INT_CMDTMO, mrq, cmd);
@@ -1245,12 +1245,10 @@ static inline bool msdc_cmd_is_ready(struct msdc_host *host,
 	}
 
 	if (mmc_resp_type(cmd) == MMC_RSP_R1B || cmd->data) {
-		tmo = jiffies + msecs_to_jiffies(20);
 		/* R1B or with data, should check SDCBUSY */
-		while ((readl(host->base + SDC_STS) & SDC_STS_SDCBUSY) &&
-				time_before(jiffies, tmo))
-			cpu_relax();
-		if (readl(host->base + SDC_STS) & SDC_STS_SDCBUSY) {
+		ret = readl_poll_timeout_atomic(host->base + SDC_STS, val,
+						!(val & SDC_STS_SDCBUSY), 1, 20000);
+		if (ret) {
 			dev_err(host->dev, "Controller busy detected\n");
 			host->error |= REQ_CMD_BUSY;
 			msdc_cmd_done(host, MSDC_INT_CMDTMO, mrq, cmd);
@@ -1376,6 +1374,8 @@ static bool msdc_data_xfer_done(struct msdc_host *host, u32 events,
 	    (MSDC_INT_XFER_COMPL | MSDC_INT_DATCRCERR | MSDC_INT_DATTMO
 	     | MSDC_INT_DMA_BDCSERR | MSDC_INT_DMA_GPDCSERR
 	     | MSDC_INT_DMA_PROTECT);
+	u32 val;
+	int ret;
 
 	spin_lock_irqsave(&host->lock, flags);
 	done = !host->data;
@@ -1392,8 +1392,14 @@ static bool msdc_data_xfer_done(struct msdc_host *host, u32 events,
 				readl(host->base + MSDC_DMA_CFG));
 		sdr_set_field(host->base + MSDC_DMA_CTRL, MSDC_DMA_CTRL_STOP,
 				1);
-		while (readl(host->base + MSDC_DMA_CFG) & MSDC_DMA_CFG_STS)
-			cpu_relax();
+
+		ret = readl_poll_timeout_atomic(host->base + MSDC_DMA_CFG, val,
+						!(val & MSDC_DMA_CFG_STS), 1, 20000);
+		if (ret) {
+			dev_dbg(host->dev, "DMA stop timed out\n");
+			return false;
+		}
+
 		sdr_clr_bits(host->base + MSDC_INTEN, data_ints_mask);
 		dev_dbg(host->dev, "DMA stop\n");
 
@@ -2674,7 +2680,11 @@ static int msdc_drv_probe(struct platform_device *pdev)
 	spin_lock_init(&host->lock);
 
 	platform_set_drvdata(pdev, mmc);
-	msdc_ungate_clock(host);
+	ret = msdc_ungate_clock(host);
+	if (ret) {
+		dev_err(&pdev->dev, "Cannot ungate clocks!\n");
+		goto release_mem;
+	}
 	msdc_init_hw(host);
 
 	if (mmc->caps2 & MMC_CAP2_CQE) {
@@ -2833,8 +2843,12 @@ static int __maybe_unused msdc_runtime_resume(struct device *dev)
 {
 	struct mmc_host *mmc = dev_get_drvdata(dev);
 	struct msdc_host *host = mmc_priv(mmc);
+	int ret;
+
+	ret = msdc_ungate_clock(host);
+	if (ret)
+		return ret;
 
-	msdc_ungate_clock(host);
 	msdc_restore_reg(host);
 	return 0;
 }
Signature · 1 line
2.34.1
author maintainer
#702 lore raw

PATCH 702/1039 drm/amdgpu: Dont inherit GEM object VMAs in child process

In reply to Greg Kroah-Hartman

From: Rajneesh Bhardwaj <[email protected]>

[ Upstream commit fbcdbfde87509d523132b59f661a355c731139d0 ]

When an application having open file access to a node forks, its shared mappings also get reflected in the address space of child process even though it cannot access them with the object permissions applied. With the existing permission checks on the gem objects, it might be reasonable to also create the VMAs with VM_DONTCOPY flag so a user space application doesn't need to explicitly call the madvise(addr, len, MADV_DONTFORK) system call to prevent the pages in the mapped range to appear in the address space of the child process. It also prevents the memory leaks due to additional reference counts on the mapped BOs in the child process that prevented freeing the memory in the parent for which we had worked around earlier in the user space inside the thunk library.

Additionally, we faced this issue when using CRIU to checkpoint restore an application that had such inherited mappings in the child which confuse CRIU when it mmaps on restore. Having this flag set for the render node VMAs helps. VMAs mapped via KFD already take care of this so this is needed only for the render nodes.

To limit the impact of the change to user space consumers such as OpenGL etc, limit it to KFD BOs only.

---
 drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c +3 -0
index a1e63ba4c54a5..630dc99e49086 100644
@@ -264,6 +264,9 @@ static int amdgpu_gem_object_mmap(struct drm_gem_object *obj, struct vm_area_str
 	    !(vma->vm_flags & (VM_READ | VM_WRITE | VM_EXEC)))
 		vma->vm_flags &= ~VM_MAYWRITE;
 
+	if (bo->kfd_bo)
+		vma->vm_flags |= VM_DONTCOPY;
+
 	return drm_gem_ttm_mmap(obj, vma);
 }
 
Signature · 1 line
2.34.1
#703 lore raw

Re: [PATCH 5.16 0702/1039] drm/amdgpu: Dont inherit GEM object VMAs in child process

Hi Greg,

Am 24.01.22 um 19:41 schrieb Greg Kroah-Hartman:

Quoted 25 lines

From: Rajneesh Bhardwaj <[email protected]>

[ Upstream commit fbcdbfde87509d523132b59f661a355c731139d0 ]

When an application having open file access to a node forks, its shared mappings also get reflected in the address space of child process even though it cannot access them with the object permissions applied. With the existing permission checks on the gem objects, it might be reasonable to also create the VMAs with VM_DONTCOPY flag so a user space application doesn't need to explicitly call the madvise(addr, len, MADV_DONTFORK) system call to prevent the pages in the mapped range to appear in the address space of the child process. It also prevents the memory leaks due to additional reference counts on the mapped BOs in the child process that prevented freeing the memory in the parent for which we had worked around earlier in the user space inside the thunk library.

Additionally, we faced this issue when using CRIU to checkpoint restore an application that had such inherited mappings in the child which confuse CRIU when it mmaps on restore. Having this flag set for the render node VMAs helps. VMAs mapped via KFD already take care of this so this is needed only for the render nodes.

To limit the impact of the change to user space consumers such as OpenGL etc, limit it to KFD BOs only.

Quoted 6 lines

Please drop this patch from all stable versions since it was reverted from upstream later on.

Regards,
Christian.

Quoted 18 lines
---
  drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c | 3 +++
  1 file changed, 3 insertions(+)
drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c +3 -0
index a1e63ba4c54a5..630dc99e49086 100644
@@ -264,6 +264,9 @@ static int amdgpu_gem_object_mmap(struct drm_gem_object *obj, struct vm_area_str
  	    !(vma->vm_flags & (VM_READ | VM_WRITE | VM_EXEC)))
  		vma->vm_flags &= ~VM_MAYWRITE;
 
+	if (bo->kfd_bo)
+		vma->vm_flags |= VM_DONTCOPY;
+
  	return drm_gem_ttm_mmap(obj, vma);
  }
 
author maintainer
#704 lore raw

Re: [PATCH 5.16 0702/1039] drm/amdgpu: Dont inherit GEM object VMAs in child process

Quoted 35 lines · Christian König

Hi Greg,

Am 24.01.22 um 19:41 schrieb Greg Kroah-Hartman:

Quoted 25 lines

From: Rajneesh Bhardwaj <[email protected]>

[ Upstream commit fbcdbfde87509d523132b59f661a355c731139d0 ]

When an application having open file access to a node forks, its shared mappings also get reflected in the address space of child process even though it cannot access them with the object permissions applied. With the existing permission checks on the gem objects, it might be reasonable to also create the VMAs with VM_DONTCOPY flag so a user space application doesn't need to explicitly call the madvise(addr, len, MADV_DONTFORK) system call to prevent the pages in the mapped range to appear in the address space of the child process. It also prevents the memory leaks due to additional reference counts on the mapped BOs in the child process that prevented freeing the memory in the parent for which we had worked around earlier in the user space inside the thunk library.

Additionally, we faced this issue when using CRIU to checkpoint restore an application that had such inherited mappings in the child which confuse CRIU when it mmaps on restore. Having this flag set for the render node VMAs helps. VMAs mapped via KFD already take care of this so this is needed only for the render nodes.

To limit the impact of the change to user space consumers such as OpenGL etc, limit it to KFD BOs only.

Quoted 6 lines
Quoted 2 lines

Please drop this patch from all stable versions since it was reverted from upstream later on.

The revert comes later in the patch series, so all should be ok :)

thanks,

greg k-h

author maintainer
#705 lore raw

Re: [PATCH 5.16 0702/1039] drm/amdgpu: Dont inherit GEM object VMAs in child process

Quoted 36 lines · Greg Kroah-Hartman
Quoted 35 lines · Christian König

Hi Greg,

Am 24.01.22 um 19:41 schrieb Greg Kroah-Hartman:

Quoted 31 lines

From: Rajneesh Bhardwaj <[email protected]>

[ Upstream commit fbcdbfde87509d523132b59f661a355c731139d0 ]

When an application having open file access to a node forks, its shared mappings also get reflected in the address space of child process even though it cannot access them with the object permissions applied. With the existing permission checks on the gem objects, it might be reasonable to also create the VMAs with VM_DONTCOPY flag so a user space application doesn't need to explicitly call the madvise(addr, len, MADV_DONTFORK) system call to prevent the pages in the mapped range to appear in the address space of the child process. It also prevents the memory leaks due to additional reference counts on the mapped BOs in the child process that prevented freeing the memory in the parent for which we had worked around earlier in the user space inside the thunk library.

Additionally, we faced this issue when using CRIU to checkpoint restore an application that had such inherited mappings in the child which confuse CRIU when it mmaps on restore. Having this flag set for the render node VMAs helps. VMAs mapped via KFD already take care of this so this is needed only for the render nodes.

To limit the impact of the change to user space consumers such as OpenGL etc, limit it to KFD BOs only.

Quoted 4 lines
Quoted 2 lines

Please drop this patch from all stable versions since it was reverted from upstream later on.

The revert comes later in the patch series, so all should be ok :)

Looks like Sasha dropped this, so I'll go drop the revert as well.

thanks,

greg k-h

author maintainer
#706 lore raw

PATCH 703/1039 drm/amdgpu: fixup bad vram size on gmc v8

In reply to Greg Kroah-Hartman

From: Zongmin Zhou <[email protected]>

[ Upstream commit 11544d77e3974924c5a9c8a8320b996a3e9b2f8b ]

Some boards(like RX550) seem to have garbage in the upper 16 bits of the vram size register. Check for this and clamp the size properly. Fixes
boards reporting bogus amounts of vram.

after add this patch,the maximum GPU VRAM size is 64GB, otherwise only 64GB vram size will be used.

---
 drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c | 13 ++++++++++---
 1 file changed, 10 insertions(+), 3 deletions(-)
drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c +10 -3
index 492ebed2915be..63b890f1e8afb 100644
@@ -515,10 +515,10 @@ static void gmc_v8_0_mc_program(struct amdgpu_device *adev)
 static int gmc_v8_0_mc_init(struct amdgpu_device *adev)
 {
 	int r;
+	u32 tmp;
 
 	adev->gmc.vram_width = amdgpu_atombios_get_vram_width(adev);
 	if (!adev->gmc.vram_width) {
-		u32 tmp;
 		int chansize, numchan;
 
 		/* Get VRAM informations */
@@ -562,8 +562,15 @@ static int gmc_v8_0_mc_init(struct amdgpu_device *adev)
 		adev->gmc.vram_width = numchan * chansize;
 	}
 	/* size in MB on si */
-	adev->gmc.mc_vram_size = RREG32(mmCONFIG_MEMSIZE) * 1024ULL * 1024ULL;
-	adev->gmc.real_vram_size = RREG32(mmCONFIG_MEMSIZE) * 1024ULL * 1024ULL;
+	tmp = RREG32(mmCONFIG_MEMSIZE);
+	/* some boards may have garbage in the upper 16 bits */
+	if (tmp & 0xffff0000) {
+		DRM_INFO("Probable bad vram size: 0x%08x\n", tmp);
+		if (tmp & 0xffff)
+			tmp &= 0xffff;
+	}
+	adev->gmc.mc_vram_size = tmp * 1024ULL * 1024ULL;
+	adev->gmc.real_vram_size = adev->gmc.mc_vram_size;
 
 	if (!(adev->flags & AMD_IS_APU)) {
 		r = amdgpu_device_resize_fb_bar(adev);
Signature · 1 line
2.34.1
author maintainer
#707 lore raw

PATCH 704/1039 mfd: intel_soc_pmic: Use CPU-id check instead of _HRV check to differentiate variants

In reply to Greg Kroah-Hartman

From: Hans de Goede <[email protected]>

[ Upstream commit 5b78223f55a0f516a1639dbe11cd4324d4aaee20 ]

The Intel Crystal Cove PMIC has 2 different variants, one for use with Bay Trail (BYT) SoCs and one for use with Cherry Trail (CHT) SoCs.

So far we have been using an ACPI _HRV check to differentiate between the 2, but at least on the Microsoft Surface 3, which is a CHT device, the wrong _HRV value is reported by ACPI.

So instead switch to a CPU-ID check which prevents us from relying on the possibly wrong ACPI _HRV value.

---
 drivers/mfd/intel_soc_pmic_core.c | 28 +++-------------------------
 1 file changed, 3 insertions(+), 25 deletions(-)
drivers/mfd/intel_soc_pmic_core.c +3 -25
index ddd64f9e3341e..47cb7f00dfcfc 100644
@@ -14,15 +14,12 @@
 #include <linux/module.h>
 #include <linux/mfd/core.h>
 #include <linux/mfd/intel_soc_pmic.h>
+#include <linux/platform_data/x86/soc.h>
 #include <linux/pwm.h>
 #include <linux/regmap.h>
 
 #include "intel_soc_pmic_core.h"
 
-/* Crystal Cove PMIC shares same ACPI ID between different platforms */
-#define BYT_CRC_HRV		2
-#define CHT_CRC_HRV		3
-
 /* PWM consumed by the Intel GFX */
 static struct pwm_lookup crc_pwm_lookup[] = {
 	PWM_LOOKUP("crystal_cove_pwm", 0, "0000:00:02.0", "pwm_pmic_backlight", 0, PWM_POLARITY_NORMAL),
@@ -34,31 +31,12 @@ static int intel_soc_pmic_i2c_probe(struct i2c_client *i2c,
 	struct device *dev = &i2c->dev;
 	struct intel_soc_pmic_config *config;
 	struct intel_soc_pmic *pmic;
-	unsigned long long hrv;
-	acpi_status status;
 	int ret;
 
-	/*
-	 * There are 2 different Crystal Cove PMICs a Bay Trail and Cherry
-	 * Trail version, use _HRV to differentiate between the 2.
-	 */
-	status = acpi_evaluate_integer(ACPI_HANDLE(dev), "_HRV", NULL, &hrv);
-	if (ACPI_FAILURE(status)) {
-		dev_err(dev, "Failed to get PMIC hardware revision\n");
-		return -ENODEV;
-	}
-
-	switch (hrv) {
-	case BYT_CRC_HRV:
+	if (soc_intel_is_byt())
 		config = &intel_soc_pmic_config_byt_crc;
-		break;
-	case CHT_CRC_HRV:
+	else
 		config = &intel_soc_pmic_config_cht_crc;
-		break;
-	default:
-		dev_warn(dev, "Unknown hardware rev %llu, assuming BYT\n", hrv);
-		config = &intel_soc_pmic_config_byt_crc;
-	}
 
 	pmic = devm_kzalloc(dev, sizeof(*pmic), GFP_KERNEL);
 	if (!pmic)
Signature · 1 line
2.34.1
author maintainer
#708 lore raw

PATCH 705/1039 amdgpu/pm: Make sysfs pm attributes as read-only for VFs

In reply to Greg Kroah-Hartman

From: Marina Nikolic <[email protected]>

[ Upstream commit 11c9cc95f818f0f187e9b579a7f136f532b42445 ]

== Description ==
Setting values of pm attributes through sysfs should not be allowed in SRIOV mode.
These calls will not be processed by FW anyway, but error handling on sysfs level should be improved.

== Changes ==
This patch prohibits performing of all set commands in SRIOV mode on sysfs level.
It offers better error handling as calls that are not allowed will not be propagated further.

== Test ==
Writing to any sysfs file in passthrough mode will succeed. Writing to any sysfs file in ONEVF mode will yield error:
"calling process does not have sufficient permission to execute a command".

---
 drivers/gpu/drm/amd/pm/amdgpu_pm.c | 6 ++++++
 1 file changed, 6 insertions(+)
drivers/gpu/drm/amd/pm/amdgpu_pm.c +6 -0
index 41472ed992530..f8370d54100e8 100644
@@ -2123,6 +2123,12 @@ static int default_attr_update(struct amdgpu_device *adev, struct amdgpu_device_
 		}
 	}
 
+	/* setting should not be allowed from VF */
+	if (amdgpu_sriov_vf(adev)) {
+		dev_attr->attr.mode &= ~S_IWUGO;
+		dev_attr->store = NULL;
+	}
+
 #undef DEVICE_ATTR_IS
 
 	return 0;
Signature · 1 line
2.34.1
author maintainer
#709 lore raw

PATCH 706/1039 ACPI: battery: Add the ThinkPad "Not Charging" quirk

In reply to Greg Kroah-Hartman

From: Thomas Weißschuh <[email protected]>

[ Upstream commit e96c1197aca628f7d2480a1cc3214912b40b3414 ]

The EC/ACPI firmware on Lenovo ThinkPads used to report a status of "Unknown" when the battery is between the charge start and charge stop thresholds. On Windows, it reports "Not Charging" so the quirk has been added to also report correctly.

Now the "status" attribute returns "Not Charging" when the battery on ThinkPads is not physicaly charging.

---
 drivers/acpi/battery.c | 22 ++++++++++++++++++++++
 1 file changed, 22 insertions(+)
drivers/acpi/battery.c +22 -0
index 8afa85d6eb6a7..ead0114f27c9f 100644
@@ -53,6 +53,7 @@ static int battery_bix_broken_package;
 static int battery_notification_delay_ms;
 static int battery_ac_is_broken;
 static int battery_check_pmic = 1;
+static int battery_quirk_notcharging;
 static unsigned int cache_time = 1000;
 module_param(cache_time, uint, 0644);
 MODULE_PARM_DESC(cache_time, "cache time in milliseconds");
@@ -217,6 +218,8 @@ static int acpi_battery_get_property(struct power_supply *psy,
 			val->intval = POWER_SUPPLY_STATUS_CHARGING;
 		else if (acpi_battery_is_charged(battery))
 			val->intval = POWER_SUPPLY_STATUS_FULL;
+		else if (battery_quirk_notcharging)
+			val->intval = POWER_SUPPLY_STATUS_NOT_CHARGING;
 		else
 			val->intval = POWER_SUPPLY_STATUS_UNKNOWN;
 		break;
@@ -1111,6 +1114,12 @@ battery_do_not_check_pmic_quirk(const struct dmi_system_id *d)
 	return 0;
 }
 
+static int __init battery_quirk_not_charging(const struct dmi_system_id *d)
+{
+	battery_quirk_notcharging = 1;
+	return 0;
+}
+
 static const struct dmi_system_id bat_dmi_table[] __initconst = {
 	{
 		/* NEC LZ750/LS */
@@ -1155,6 +1164,19 @@ static const struct dmi_system_id bat_dmi_table[] __initconst = {
 			DMI_MATCH(DMI_PRODUCT_VERSION, "Lenovo MIIX 320-10ICR"),
 		},
 	},
+	{
+		/*
+		 * On Lenovo ThinkPads the BIOS specification defines
+		 * a state when the bits for charging and discharging
+		 * are both set to 0. That state is "Not Charging".
+		 */
+		.callback = battery_quirk_not_charging,
+		.ident = "Lenovo ThinkPad",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"),
+			DMI_MATCH(DMI_PRODUCT_VERSION, "ThinkPad"),
+		},
+	},
 	{},
 };
 
Signature · 1 line
2.34.1
author maintainer
#710 lore raw

PATCH 707/1039 ACPI: CPPC: Check present CPUs for determining _CPC is valid

In reply to Greg Kroah-Hartman

From: Mario Limonciello <[email protected]>

[ Upstream commit 2aeca6bd02776d7f56a49a32be0dd184f204d888 ]

As this is a static check, it should be based upon what is currently present on the system. This makes probeing more deterministic.

While local APIC flags field (lapic_flags) of cpu core in MADT table is 0, then the cpu core won't be enabled. In this case, _CPC won't be found in this core, and return back to _CPC invalid with walking through possible cpus (include disable cpus). This is not expected, so switch to check present CPUs instead.

---
 drivers/acpi/cppc_acpi.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/acpi/cppc_acpi.c +1 -1
index b62c87b8ce4a9..12a156d8283e6 100644
@@ -411,7 +411,7 @@ bool acpi_cpc_valid(void)
 	struct cpc_desc *cpc_ptr;
 	int cpu;
 
-	for_each_possible_cpu(cpu) {
+	for_each_present_cpu(cpu) {
 		cpc_ptr = per_cpu(cpc_desc_ptr, cpu);
 		if (!cpc_ptr)
 			return false;
Signature · 1 line
2.34.1
author maintainer
#711 lore raw

PATCH 708/1039 net/mlx5: DR, Fix error flow in creating matcher

In reply to Greg Kroah-Hartman

From: Yevgeny Kliteynik <[email protected]>

[ Upstream commit 84dfac39c61fde04126e23723138128b50cd036f ]

The error code of nic matcher init functions wasn't checked. This patch improves the matcher init function and fix error flow bug:
the handling of match parameter is moved into a separate function and error flow is simplified.

---
 .../mellanox/mlx5/core/steering/dr_matcher.c  | 53 +++++++++++--------
 1 file changed, 32 insertions(+), 21 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/steering/dr_matcher.c +32 -21
index 793365242e852..3d0cdc36a91ab 100644
@@ -872,13 +872,12 @@ uninit_nic_rx:
 	return ret;
 }
 
-static int dr_matcher_init(struct mlx5dr_matcher *matcher,
-			   struct mlx5dr_match_parameters *mask)
+static int dr_matcher_copy_param(struct mlx5dr_matcher *matcher,
+				 struct mlx5dr_match_parameters *mask)
 {
+	struct mlx5dr_domain *dmn = matcher->tbl->dmn;
 	struct mlx5dr_match_parameters consumed_mask;
-	struct mlx5dr_table *tbl = matcher->tbl;
-	struct mlx5dr_domain *dmn = tbl->dmn;
-	int i, ret;
+	int i, ret = 0;
 
 	if (matcher->match_criteria >= DR_MATCHER_CRITERIA_MAX) {
 		mlx5dr_err(dmn, "Invalid match criteria attribute\n");
@@ -898,10 +897,36 @@ static int dr_matcher_init(struct mlx5dr_matcher *matcher,
 		consumed_mask.match_sz = mask->match_sz;
 		memcpy(consumed_mask.match_buf, mask->match_buf, mask->match_sz);
 		mlx5dr_ste_copy_param(matcher->match_criteria,
-				      &matcher->mask, &consumed_mask,
-				      true);
+				      &matcher->mask, &consumed_mask, true);
+
+		/* Check that all mask data was consumed */
+		for (i = 0; i < consumed_mask.match_sz; i++) {
+			if (!((u8 *)consumed_mask.match_buf)[i])
+				continue;
+
+			mlx5dr_dbg(dmn,
+				   "Match param mask contains unsupported parameters\n");
+			ret = -EOPNOTSUPP;
+			break;
+		}
+
+		kfree(consumed_mask.match_buf);
 	}
 
+	return ret;
+}
+
+static int dr_matcher_init(struct mlx5dr_matcher *matcher,
+			   struct mlx5dr_match_parameters *mask)
+{
+	struct mlx5dr_table *tbl = matcher->tbl;
+	struct mlx5dr_domain *dmn = tbl->dmn;
+	int ret;
+
+	ret = dr_matcher_copy_param(matcher, mask);
+	if (ret)
+		return ret;
+
 	switch (dmn->type) {
 	case MLX5DR_DOMAIN_TYPE_NIC_RX:
 		matcher->rx.nic_tbl = &tbl->rx;
@@ -919,22 +944,8 @@ static int dr_matcher_init(struct mlx5dr_matcher *matcher,
 	default:
 		WARN_ON(true);
 		ret = -EINVAL;
-		goto free_consumed_mask;
-	}
-
-	/* Check that all mask data was consumed */
-	for (i = 0; i < consumed_mask.match_sz; i++) {
-		if (!((u8 *)consumed_mask.match_buf)[i])
-			continue;
-
-		mlx5dr_dbg(dmn, "Match param mask contains unsupported parameters\n");
-		ret = -EOPNOTSUPP;
-		goto free_consumed_mask;
 	}
 
-	ret =  0;
-free_consumed_mask:
-	kfree(consumed_mask.match_buf);
 	return ret;
 }
 
Signature · 1 line
2.34.1
author maintainer
#712 lore raw

PATCH 709/1039 btrfs: remove BUG_ON() in find_parent_nodes()

In reply to Greg Kroah-Hartman

From: Josef Bacik <[email protected]>

[ Upstream commit fcba0120edf88328524a4878d1d6f4ad39f2ec81 ]

We search for an extent entry with .offset = -1, which shouldn't be a thing, but corruption happens. Add an ASSERT() for the developers, return -EUCLEAN for mortals.

---
 fs/btrfs/backref.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)
fs/btrfs/backref.c +6 -1
index f735b8798ba12..6b4b0f105a572 100644
@@ -1214,7 +1214,12 @@ again:
 	ret = btrfs_search_slot(NULL, fs_info->extent_root, &key, path, 0, 0);
 	if (ret < 0)
 		goto out;
-	BUG_ON(ret == 0);
+	if (ret == 0) {
+		/* This shouldn't happen, indicates a bug or fs corruption. */
+		ASSERT(ret != 0);
+		ret = -EUCLEAN;
+		goto out;
+	}
 
 #ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
 	if (trans && likely(trans->type != __TRANS_DUMMY) &&
Signature · 1 line
2.34.1
author maintainer
#713 lore raw

PATCH 710/1039 btrfs: remove BUG_ON(!eie) in find_parent_nodes

In reply to Greg Kroah-Hartman

From: Josef Bacik <[email protected]>

[ Upstream commit 9f05c09d6baef789726346397438cca4ec43c3ee ]

If we're looking for leafs that point to a data extent we want to record the extent items that point at our bytenr. At this point we have the reference and we know for a fact that this leaf should have a reference to our bytenr. However if there's some sort of corruption we may not find any references to our leaf, and thus could end up with eie == NULL. Replace this BUG_ON() with an ASSERT() and then return -EUCLEAN for the mortals.

---
 fs/btrfs/backref.c | 14 +++++++++++---
 1 file changed, 11 insertions(+), 3 deletions(-)
fs/btrfs/backref.c +11 -3
index 6b4b0f105a572..8b090c40daf77 100644
@@ -1365,10 +1365,18 @@ again:
 				goto out;
 			if (!ret && extent_item_pos) {
 				/*
-				 * we've recorded that parent, so we must extend
-				 * its inode list here
+				 * We've recorded that parent, so we must extend
+				 * its inode list here.
+				 *
+				 * However if there was corruption we may not
+				 * have found an eie, return an error in this
+				 * case.
 				 */
-				BUG_ON(!eie);
+				ASSERT(eie);
+				if (!eie) {
+					ret = -EUCLEAN;
+					goto out;
+				}
 				while (eie->next)
 					eie = eie->next;
 				eie->next = ref->inode_list;
Signature · 1 line
2.34.1
author maintainer
#714 lore raw

PATCH 711/1039 net: mdio: Demote probed message to debug print

In reply to Greg Kroah-Hartman

From: Florian Fainelli <[email protected]>

[ Upstream commit 7590fc6f80ac2cbf23e6b42b668bbeded070850b ]

On systems with large numbers of MDIO bus/muxes the message indicating that a given MDIO bus has been successfully probed is repeated for as many buses we have, which can eat up substantial boot time for no reason, demote to a debug print.

---
 drivers/net/phy/mdio_bus.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/phy/mdio_bus.c +1 -1
index c198722e4871d..3f7b93d5c76fe 100644
@@ -594,7 +594,7 @@ int __mdiobus_register(struct mii_bus *bus, struct module *owner)
 	mdiobus_setup_mdiodev_from_board_info(bus, mdiobus_create_device);
 
 	bus->state = MDIOBUS_REGISTERED;
-	pr_info("%s: probed\n", bus->name);
+	dev_dbg(&bus->dev, "probed\n");
 	return 0;
 
 error:
Signature · 1 line
2.34.1
author maintainer
#715 lore raw

PATCH 712/1039 mac80211: allow non-standard VHT MCS-10/11

In reply to Greg Kroah-Hartman

From: Ping-Ke Shih <[email protected]>

[ Upstream commit 04be6d337d37400ad5b3d5f27ca87645ee5a18a3 ]

Some AP can possibly try non-standard VHT rate and mac80211 warns and drops packets, and leads low TCP throughput.

Rate marked as a VHT rate but data is invalid: MCS: 10, NSS: 2
WARNING: CPU: 1 PID: 7817 at net/mac80211/rx.c:4856 ieee80211_rx_list+0x223/0x2f0 [mac8021

Since commit c27aa56a72b8 ("cfg80211: add VHT rate entries for MCS-10 and MCS-11") has added, mac80211 adds this support as well.

After this patch, throughput is good and iw can get the bitrate:
rx bitrate: 975.1 MBit/s VHT-MCS 10 80MHz short GI VHT-NSS 2
or

    rx bitrate:	1083.3 MBit/s VHT-MCS 11 80MHz short GI VHT-NSS 2
---
 net/mac80211/rx.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/mac80211/rx.c +1 -1
index 0544563ede522..d2e8b84ed2836 100644
@@ -4924,7 +4924,7 @@ void ieee80211_rx_list(struct ieee80211_hw *hw, struct ieee80211_sta *pubsta,
 				goto drop;
 			break;
 		case RX_ENC_VHT:
-			if (WARN_ONCE(status->rate_idx > 9 ||
+			if (WARN_ONCE(status->rate_idx > 11 ||
 				      !status->nss ||
 				      status->nss > 8,
 				      "Rate marked as a VHT rate but data is invalid: MCS: %d, NSS: %d\n",
Signature · 1 line
2.34.1
author maintainer
#716 lore raw

PATCH 713/1039 dm btree: add a defensive bounds check to insert_at()

In reply to Greg Kroah-Hartman

From: Joe Thornber <[email protected]>

[ Upstream commit 85bca3c05b6cca31625437eedf2060e846c4bbad ]

Corrupt metadata could trigger an out of bounds write.

---
 drivers/md/persistent-data/dm-btree.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)
drivers/md/persistent-data/dm-btree.c +5 -3
index 0703ca7a7d9a4..5ce64e93aae74 100644
@@ -81,14 +81,16 @@ void inc_children(struct dm_transaction_manager *tm, struct btree_node *n,
 }
 
 static int insert_at(size_t value_size, struct btree_node *node, unsigned index,
-		      uint64_t key, void *value)
-		      __dm_written_to_disk(value)
+		     uint64_t key, void *value)
+	__dm_written_to_disk(value)
 {
 	uint32_t nr_entries = le32_to_cpu(node->header.nr_entries);
+	uint32_t max_entries = le32_to_cpu(node->header.max_entries);
 	__le64 key_le = cpu_to_le64(key);
 
 	if (index > nr_entries ||
-	    index >= le32_to_cpu(node->header.max_entries)) {
+	    index >= max_entries ||
+	    nr_entries >= max_entries) {
 		DMERR("too many entries in btree node for insert");
 		__dm_unbless_for_disk(value);
 		return -ENOMEM;
Signature · 1 line
2.34.1
author maintainer
#717 lore raw

PATCH 714/1039 dm space map common: add bounds check to sm_ll_lookup_bitmap()

In reply to Greg Kroah-Hartman

From: Joe Thornber <[email protected]>

[ Upstream commit cba23ac158db7f3cd48a923d6861bee2eb7a2978 ]

Corrupted metadata could warrant returning error from sm_ll_lookup_bitmap().

---
 drivers/md/persistent-data/dm-space-map-common.c | 5 +++++
 1 file changed, 5 insertions(+)
drivers/md/persistent-data/dm-space-map-common.c +5 -0
index 4a6a2a9b4eb49..bfbfa750e0160 100644
@@ -283,6 +283,11 @@ int sm_ll_lookup_bitmap(struct ll_disk *ll, dm_block_t b, uint32_t *result)
 	struct disk_index_entry ie_disk;
 	struct dm_block *blk;
 
+	if (b >= ll->nr_blocks) {
+		DMERR_LIMIT("metadata block out of bounds");
+		return -EINVAL;
+	}
+
 	b = do_div(index, ll->entries_per_block);
 	r = ll->load_ie(ll, index, &ie_disk);
 	if (r < 0)
Signature · 1 line
2.34.1
author maintainer
#718 lore raw

PATCH 715/1039 can: do not increase rx statistics when generating a CAN rx error message frame

In reply to Greg Kroah-Hartman

From: Vincent Mailhol <[email protected]>

[ Upstream commit 676068db69b847f06fe054fca15bf6b107bd24da ]

The CAN error message frames (i.e. error skb) are an interface specific to socket CAN. The payload of the CAN error message frames does not correspond to any actual data sent on the wire. Only an error flag and a delimiter are transmitted when an error occurs (c.f. ISO 11898-1 section 10.4.4.2 "Error flag").

For this reason, it makes no sense to increment the rx_packets and rx_bytes fields of struct net_device_stats because no actual payload were transmitted on the wire.

This patch fixes all the CAN drivers.

CC Marc Kleine-Budde <[email protected]>
CC Nicolas Ferre <[email protected]>
CC Alexandre Belloni <[email protected]>
CC Ludovic Desroches <[email protected]>
CC Chandrasekar Ramakrishnan <[email protected]>
CC Maxime Ripard <[email protected]>
CC Chen-Yu Tsai <[email protected]>
CC Jernej Skrabec <[email protected]>
CC Appana Durga Kedareswara rao <[email protected]>
CC Naga Sureshkumar Relli <[email protected]>
CC Michal Simek <[email protected]>
CC Stephane Grosjean <[email protected]>
---
 drivers/net/can/at91_can.c                        | 6 ------
 drivers/net/can/c_can/c_can_main.c                | 5 -----
 drivers/net/can/cc770/cc770.c                     | 3 ---
 drivers/net/can/dev/dev.c                         | 4 ----
 drivers/net/can/dev/rx-offload.c                  | 6 ++++--
 drivers/net/can/ifi_canfd/ifi_canfd.c             | 5 -----
 drivers/net/can/kvaser_pciefd.c                   | 5 -----
 drivers/net/can/m_can/m_can.c                     | 7 -------
 drivers/net/can/mscan/mscan.c                     | 9 +++++----
 drivers/net/can/pch_can.c                         | 3 ---
 drivers/net/can/peak_canfd/peak_canfd.c           | 4 ----
 drivers/net/can/rcar/rcar_can.c                   | 6 +-----
 drivers/net/can/rcar/rcar_canfd.c                 | 4 ----
 drivers/net/can/sja1000/sja1000.c                 | 2 --
 drivers/net/can/sun4i_can.c                       | 7 ++-----
 drivers/net/can/usb/ems_usb.c                     | 2 --
 drivers/net/can/usb/esd_usb2.c                    | 2 --
 drivers/net/can/usb/etas_es58x/es58x_core.c       | 7 -------
 drivers/net/can/usb/kvaser_usb/kvaser_usb_core.c  | 2 --
 drivers/net/can/usb/kvaser_usb/kvaser_usb_hydra.c | 8 --------
 drivers/net/can/usb/kvaser_usb/kvaser_usb_leaf.c  | 4 ----
 drivers/net/can/usb/peak_usb/pcan_usb.c           | 2 --
 drivers/net/can/usb/peak_usb/pcan_usb_fd.c        | 3 ---
 drivers/net/can/usb/peak_usb/pcan_usb_pro.c       | 2 --
 drivers/net/can/usb/ucan.c                        | 6 ++++--
 drivers/net/can/usb/usb_8dev.c                    | 2 --
 drivers/net/can/xilinx_can.c                      | 9 +--------
 27 files changed, 17 insertions(+), 108 deletions(-)
drivers/net/can/at91_can.c +0 -6
index 3aea32c9b108f..3cd872cf9be66 100644
@@ -553,8 +553,6 @@ static void at91_rx_overflow_err(struct net_device *dev)
 	cf->can_id |= CAN_ERR_CRTL;
 	cf->data[1] = CAN_ERR_CRTL_RX_OVERFLOW;
 
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
 	netif_receive_skb(skb);
 }
 
@@ -779,8 +777,6 @@ static int at91_poll_err(struct net_device *dev, int quota, u32 reg_sr)
 
 	at91_poll_err_frame(dev, cf, reg_sr);
 
-	dev->stats.rx_packets++;
-	dev->stats.rx_bytes += cf->len;
 	netif_receive_skb(skb);
 
 	return 1;
@@ -1037,8 +1033,6 @@ static void at91_irq_err(struct net_device *dev)
 
 	at91_irq_err_state(dev, cf, new_state);
 
-	dev->stats.rx_packets++;
-	dev->stats.rx_bytes += cf->len;
 	netif_rx(skb);
 
 	priv->can.state = new_state;
drivers/net/can/c_can/c_can_main.c +0 -5
index 52671d1ea17d5..670754a129846 100644
@@ -920,7 +920,6 @@ static int c_can_handle_state_change(struct net_device *dev,
 	unsigned int reg_err_counter;
 	unsigned int rx_err_passive;
 	struct c_can_priv *priv = netdev_priv(dev);
-	struct net_device_stats *stats = &dev->stats;
 	struct can_frame *cf;
 	struct sk_buff *skb;
 	struct can_berr_counter bec;
@@ -996,8 +995,6 @@ static int c_can_handle_state_change(struct net_device *dev,
 		break;
 	}
 
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
 	netif_receive_skb(skb);
 
 	return 1;
@@ -1064,8 +1061,6 @@ static int c_can_handle_bus_err(struct net_device *dev,
 		break;
 	}
 
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
 	netif_receive_skb(skb);
 	return 1;
 }
drivers/net/can/cc770/cc770.c +0 -3
index f8a130f594e2e..a5fd8ccedec21 100644
@@ -499,7 +499,6 @@ static void cc770_rx(struct net_device *dev, unsigned int mo, u8 ctrl1)
 static int cc770_err(struct net_device *dev, u8 status)
 {
 	struct cc770_priv *priv = netdev_priv(dev);
-	struct net_device_stats *stats = &dev->stats;
 	struct can_frame *cf;
 	struct sk_buff *skb;
 	u8 lec;
@@ -571,8 +570,6 @@ static int cc770_err(struct net_device *dev, u8 status)
 	}
 
 
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
 	netif_rx(skb);
 
 	return 0;
drivers/net/can/dev/dev.c +0 -4
index e3d840b81357d..4845ae6456e19 100644
@@ -136,7 +136,6 @@ EXPORT_SYMBOL_GPL(can_change_state);
 static void can_restart(struct net_device *dev)
 {
 	struct can_priv *priv = netdev_priv(dev);
-	struct net_device_stats *stats = &dev->stats;
 	struct sk_buff *skb;
 	struct can_frame *cf;
 	int err;
@@ -155,9 +154,6 @@ static void can_restart(struct net_device *dev)
 
 	cf->can_id |= CAN_ERR_RESTARTED;
 
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
-
 	netif_rx_ni(skb);
 
 restart:
drivers/net/can/dev/rx-offload.c +4 -2
index 37b0cc65237b7..7dbf46b9ca5dd 100644
@@ -54,8 +54,10 @@ static int can_rx_offload_napi_poll(struct napi_struct *napi, int quota)
 		struct can_frame *cf = (struct can_frame *)skb->data;
 
 		work_done++;
-		stats->rx_packets++;
-		stats->rx_bytes += cf->len;
+		if (!(cf->can_id & CAN_ERR_FLAG)) {
+			stats->rx_packets++;
+			stats->rx_bytes += cf->len;
+		}
 		netif_receive_skb(skb);
 	}
 
drivers/net/can/ifi_canfd/ifi_canfd.c +0 -5
index 5bb957a26bc69..e8318e984bf2f 100644
@@ -430,8 +430,6 @@ static int ifi_canfd_handle_lec_err(struct net_device *ndev)
 	       priv->base + IFI_CANFD_INTERRUPT);
 	writel(IFI_CANFD_ERROR_CTR_ER_ENABLE, priv->base + IFI_CANFD_ERROR_CTR);
 
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
 	netif_receive_skb(skb);
 
 	return 1;
@@ -456,7 +454,6 @@ static int ifi_canfd_handle_state_change(struct net_device *ndev,
 					 enum can_state new_state)
 {
 	struct ifi_canfd_priv *priv = netdev_priv(ndev);
-	struct net_device_stats *stats = &ndev->stats;
 	struct can_frame *cf;
 	struct sk_buff *skb;
 	struct can_berr_counter bec;
@@ -522,8 +519,6 @@ static int ifi_canfd_handle_state_change(struct net_device *ndev,
 		break;
 	}
 
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
 	netif_receive_skb(skb);
 
 	return 1;
drivers/net/can/kvaser_pciefd.c +0 -5
index eb74cdf26b88c..ab672c92ab078 100644
@@ -1310,9 +1310,6 @@ static int kvaser_pciefd_rx_error_frame(struct kvaser_pciefd_can *can,
 	cf->data[6] = bec.txerr;
 	cf->data[7] = bec.rxerr;
 
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
-
 	netif_rx(skb);
 	return 0;
 }
@@ -1510,8 +1507,6 @@ static void kvaser_pciefd_handle_nack_packet(struct kvaser_pciefd_can *can,
 
 	if (skb) {
 		cf->can_id |= CAN_ERR_BUSERROR;
-		stats->rx_bytes += cf->len;
-		stats->rx_packets++;
 		netif_rx(skb);
 	} else {
 		stats->rx_dropped++;
drivers/net/can/m_can/m_can.c +0 -7
index c2a8421e7845c..30d94cb43113d 100644
@@ -647,9 +647,6 @@ static int m_can_handle_lec_err(struct net_device *dev,
 		break;
 	}
 
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
-
 	if (cdev->is_peripheral)
 		timestamp = m_can_get_timestamp(cdev);
 
@@ -706,7 +703,6 @@ static int m_can_handle_state_change(struct net_device *dev,
 				     enum can_state new_state)
 {
 	struct m_can_classdev *cdev = netdev_priv(dev);
-	struct net_device_stats *stats = &dev->stats;
 	struct can_frame *cf;
 	struct sk_buff *skb;
 	struct can_berr_counter bec;
@@ -771,9 +767,6 @@ static int m_can_handle_state_change(struct net_device *dev,
 		break;
 	}
 
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
-
 	if (cdev->is_peripheral)
 		timestamp = m_can_get_timestamp(cdev);
 
drivers/net/can/mscan/mscan.c +5 -4
index fa32e418eb296..9e1cce0260da6 100644
@@ -401,13 +401,14 @@ static int mscan_rx_poll(struct napi_struct *napi, int quota)
 			continue;
 		}
 
-		if (canrflg & MSCAN_RXF)
+		if (canrflg & MSCAN_RXF) {
 			mscan_get_rx_frame(dev, frame);
-		else if (canrflg & MSCAN_ERR_IF)
+			stats->rx_packets++;
+			stats->rx_bytes += frame->len;
+		} else if (canrflg & MSCAN_ERR_IF) {
 			mscan_get_err_frame(dev, frame, canrflg);
+		}
 
-		stats->rx_packets++;
-		stats->rx_bytes += frame->len;
 		work_done++;
 		netif_receive_skb(skb);
 	}
drivers/net/can/pch_can.c +0 -3
index 964c8a09226a9..6b45840db1f9b 100644
@@ -561,9 +561,6 @@ static void pch_can_error(struct net_device *ndev, u32 status)
 
 	priv->can.state = state;
 	netif_receive_skb(skb);
-
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
 }
 
 static irqreturn_t pch_can_interrupt(int irq, void *dev_id)
drivers/net/can/peak_canfd/peak_canfd.c +0 -4
index d08718e98e110..d5b8bc6d29804 100644
@@ -409,8 +409,6 @@ static int pucan_handle_status(struct peak_canfd_priv *priv,
 		return -ENOMEM;
 	}
 
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
 	pucan_netif_rx(skb, msg->ts_low, msg->ts_high);
 
 	return 0;
@@ -438,8 +436,6 @@ static int pucan_handle_cache_critical(struct peak_canfd_priv *priv)
 	cf->data[6] = priv->bec.txerr;
 	cf->data[7] = priv->bec.rxerr;
 
-	stats->rx_bytes += cf->len;
-	stats->rx_packets++;
 	netif_rx(skb);
 
 	return 0;
drivers/net/can/rcar/rcar_can.c +1 -5
index 8999ec9455ec2..f408ed9a6ccd1 100644
@@ -223,7 +223,6 @@ static void tx_failure_cleanup(struct net_device *ndev)
 static void rcar_can_error(struct net_device *ndev)
 {
 	struct rcar_can_priv *priv = netdev_priv(ndev);
-	struct net_device_stats *stats = &ndev->stats;
 	struct can_frame *cf;
 	struct sk_buff *skb;
 	u8 eifr, txerr = 0, rxerr = 0;
@@ -362,11 +361,8 @@ static void rcar_can_error(struct net_device *ndev)
 		}
 	}
 
-	if (skb) {
-		stats->rx_packets++;
-		stats->rx_bytes += cf->len;
+	if (skb)
 		netif_rx(skb);
-	}
 }
 
 static void rcar_can_tx_done(struct net_device *ndev)
drivers/net/can/rcar/rcar_canfd.c +0 -4
index 388521e70837f..137eea4c7bad8 100644
@@ -1033,8 +1033,6 @@ static void rcar_canfd_error(struct net_device *ndev, u32 cerfl,
 	/* Clear channel error interrupts that are handled */
 	rcar_canfd_write(priv->base, RCANFD_CERFL(ch),
 			 RCANFD_CERFL_ERR(~cerfl));
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
 	netif_rx(skb);
 }
 
@@ -1174,8 +1172,6 @@ static void rcar_canfd_state_change(struct net_device *ndev,
 		rx_state = txerr <= rxerr ? state : 0;
 
 		can_change_state(ndev, cf, tx_state, rx_state);
-		stats->rx_packets++;
-		stats->rx_bytes += cf->len;
 		netif_rx(skb);
 	}
 }
drivers/net/can/sja1000/sja1000.c +0 -2
index 3fad546467461..a65546ca94610 100644
@@ -487,8 +487,6 @@ static int sja1000_err(struct net_device *dev, uint8_t isrc, uint8_t status)
 			can_bus_off(dev);
 	}
 
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
 	netif_rx(skb);
 
 	return 0;
drivers/net/can/sun4i_can.c +2 -5
index 54aa7c25c4de1..599174098883d 100644
@@ -622,13 +622,10 @@ static int sun4i_can_err(struct net_device *dev, u8 isrc, u8 status)
 			can_bus_off(dev);
 	}
 
-	if (likely(skb)) {
-		stats->rx_packets++;
-		stats->rx_bytes += cf->len;
+	if (likely(skb))
 		netif_rx(skb);
-	} else {
+	else
 		return -ENOMEM;
-	}
 
 	return 0;
 }
drivers/net/can/usb/ems_usb.c +0 -2
index 2b5302e724353..7cf65936d02e5 100644
@@ -397,8 +397,6 @@ static void ems_usb_rx_err(struct ems_usb *dev, struct ems_cpc_msg *msg)
 		stats->rx_errors++;
 	}
 
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
 	netif_rx(skb);
 }
 
drivers/net/can/usb/esd_usb2.c +0 -2
index c6068a251fbed..5f6915a27b3d9 100644
@@ -293,8 +293,6 @@ static void esd_usb2_rx_event(struct esd_usb2_net_priv *priv,
 		priv->bec.txerr = txerr;
 		priv->bec.rxerr = rxerr;
 
-		stats->rx_packets++;
-		stats->rx_bytes += cf->len;
 		netif_rx(skb);
 	}
 }
drivers/net/can/usb/etas_es58x/es58x_core.c +0 -7
index 24627ab146261..fb07c33ba0c3c 100644
@@ -849,13 +849,6 @@ int es58x_rx_err_msg(struct net_device *netdev, enum es58x_err error,
 		break;
 	}
 
-	/* driver/net/can/dev.c:can_restart() takes in account error
-	 * messages in the RX stats. Doing the same here for
-	 * consistency.
-	 */
-	netdev->stats.rx_packets++;
-	netdev->stats.rx_bytes += CAN_ERR_DLC;
-
 	if (cf) {
 		if (cf->data[1])
 			cf->can_id |= CAN_ERR_CRTL;
drivers/net/can/usb/kvaser_usb/kvaser_usb_core.c +0 -2
index 0cc0fc866a2a9..3e682ef43f8ef 100644
@@ -279,8 +279,6 @@ int kvaser_usb_can_rx_over_error(struct net_device *netdev)
 	cf->can_id |= CAN_ERR_CRTL;
 	cf->data[1] = CAN_ERR_CRTL_RX_OVERFLOW;
 
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
 	netif_rx(skb);
 
 	return 0;
drivers/net/can/usb/kvaser_usb/kvaser_usb_hydra.c +0 -8
index dcee8dc828ecc..3398da323126c 100644
@@ -869,7 +869,6 @@ static void kvaser_usb_hydra_update_state(struct kvaser_usb_net_priv *priv,
 	struct net_device *netdev = priv->netdev;
 	struct can_frame *cf;
 	struct sk_buff *skb;
-	struct net_device_stats *stats;
 	enum can_state new_state, old_state;
 
 	old_state = priv->can.state;
@@ -919,9 +918,6 @@ static void kvaser_usb_hydra_update_state(struct kvaser_usb_net_priv *priv,
 	cf->data[6] = bec->txerr;
 	cf->data[7] = bec->rxerr;
 
-	stats = &netdev->stats;
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
 	netif_rx(skb);
 }
 
@@ -1074,8 +1070,6 @@ kvaser_usb_hydra_error_frame(struct kvaser_usb_net_priv *priv,
 	cf->data[6] = bec.txerr;
 	cf->data[7] = bec.rxerr;
 
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
 	netif_rx(skb);
 
 	priv->bec.txerr = bec.txerr;
@@ -1109,8 +1103,6 @@ static void kvaser_usb_hydra_one_shot_fail(struct kvaser_usb_net_priv *priv,
 	}
 
 	stats->tx_errors++;
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
 	netif_rx(skb);
 }
 
drivers/net/can/usb/kvaser_usb/kvaser_usb_leaf.c +0 -4
index f7af1bf5ab46d..5434b7386a51e 100644
@@ -641,8 +641,6 @@ static void kvaser_usb_leaf_tx_acknowledge(const struct kvaser_usb *dev,
 		if (skb) {
 			cf->can_id |= CAN_ERR_RESTARTED;
 
-			stats->rx_packets++;
-			stats->rx_bytes += cf->len;
 			netif_rx(skb);
 		} else {
 			netdev_err(priv->netdev,
@@ -843,8 +841,6 @@ static void kvaser_usb_leaf_rx_error(const struct kvaser_usb *dev,
 	cf->data[6] = es->txerr;
 	cf->data[7] = es->rxerr;
 
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
 	netif_rx(skb);
 }
 
drivers/net/can/usb/peak_usb/pcan_usb.c +0 -2
index 8762187527669..21b06a7385959 100644
@@ -520,8 +520,6 @@ static int pcan_usb_decode_error(struct pcan_usb_msg_context *mc, u8 n,
 				     &hwts->hwtstamp);
 	}
 
-	mc->netdev->stats.rx_packets++;
-	mc->netdev->stats.rx_bytes += cf->len;
 	netif_rx(skb);
 
 	return 0;
drivers/net/can/usb/peak_usb/pcan_usb_fd.c +0 -3
index 6bd12549f1014..185f5a98d2177 100644
@@ -577,9 +577,6 @@ static int pcan_usb_fd_decode_status(struct pcan_usb_fd_if *usb_if,
 	if (!skb)
 		return -ENOMEM;
 
-	netdev->stats.rx_packets++;
-	netdev->stats.rx_bytes += cf->len;
-
 	peak_usb_netif_rx_64(skb, le32_to_cpu(sm->ts_low),
 			     le32_to_cpu(sm->ts_high));
 
drivers/net/can/usb/peak_usb/pcan_usb_pro.c +0 -2
index 858ab22708fcd..f6d19879bf404 100644
@@ -660,8 +660,6 @@ static int pcan_usb_pro_handle_error(struct pcan_usb_pro_interface *usb_if,
 
 	hwts = skb_hwtstamps(skb);
 	peak_usb_get_ts_time(&usb_if->time_ref, le32_to_cpu(er->ts32), &hwts->hwtstamp);
-	netdev->stats.rx_packets++;
-	netdev->stats.rx_bytes += can_frame->len;
 	netif_rx(skb);
 
 	return 0;
drivers/net/can/usb/ucan.c +4 -2
index 1679cbe45ded2..d582c39fc8d0e 100644
@@ -621,8 +621,10 @@ static void ucan_rx_can_msg(struct ucan_priv *up, struct ucan_message_in *m)
 		memcpy(cf->data, m->msg.can_msg.data, cf->len);
 
 	/* don't count error frames as real packets */
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
+	if (!(cf->can_id & CAN_ERR_FLAG)) {
+		stats->rx_packets++;
+		stats->rx_bytes += cf->len;
+	}
 
 	/* pass it to Linux */
 	netif_rx(skb);
drivers/net/can/usb/usb_8dev.c +0 -2
index d1b83bd1b3cb9..040324362b260 100644
@@ -449,8 +449,6 @@ static void usb_8dev_rx_err_msg(struct usb_8dev_priv *priv,
 	priv->bec.txerr = txerr;
 	priv->bec.rxerr = rxerr;
 
-	stats->rx_packets++;
-	stats->rx_bytes += cf->len;
 	netif_rx(skb);
 }
 
drivers/net/can/xilinx_can.c +1 -8
index af4a2adc85983..e0525541b32a7 100644
@@ -965,13 +965,8 @@ static void xcan_update_error_state_after_rxtx(struct net_device *ndev)
 
 		xcan_set_error_state(ndev, new_state, skb ? cf : NULL);
 
-		if (skb) {
-			struct net_device_stats *stats = &ndev->stats;
-
-			stats->rx_packets++;
-			stats->rx_bytes += cf->len;
+		if (skb)
 			netif_rx(skb);
-		}
 	}
 }
 
@@ -1095,8 +1090,6 @@ static void xcan_err_interrupt(struct net_device *ndev, u32 isr)
 		if (skb) {
 			skb_cf->can_id |= cf.can_id;
 			memcpy(skb_cf->data, cf.data, CAN_ERR_DLC);
-			stats->rx_packets++;
-			stats->rx_bytes += CAN_ERR_DLC;
 			netif_rx(skb);
 		}
 	}
Signature · 1 line
2.34.1
author maintainer
#719 lore raw

PATCH 716/1039 bpf/selftests: Fix namespace mount setup in tc_redirect

In reply to Greg Kroah-Hartman

From: Jiri Olsa <[email protected]>

[ Upstream commit 5e22dd18626726028a93ff1350a8a71a00fd843d ]

The tc_redirect umounts /sys in the new namespace, which can be mounted as shared and cause global umount. The lazy umount also takes down mounted trees under /sys like debugfs, which won't be available after sysfs mounts again and could cause fails in other tests.

  # cat /proc/self/mountinfo | grep debugfs
  34 23 0:7 / /sys/kernel/debug rw,nosuid,nodev,noexec,relatime shared:14 - debugfs debugfs rw
  # cat /proc/self/mountinfo | grep sysfs
  23 86 0:22 / /sys rw,nosuid,nodev,noexec,relatime shared:2 - sysfs sysfs rw
  # mount | grep debugfs
  debugfs on /sys/kernel/debug type debugfs (rw,nosuid,nodev,noexec,relatime)
  # ./test_progs -t tc_redirect
  #164 tc_redirect:OK
  Summary: 1/4 PASSED, 0 SKIPPED, 0 FAILED
  # mount | grep debugfs
  # cat /proc/self/mountinfo | grep debugfs
  # cat /proc/self/mountinfo | grep sysfs
  25 86 0:22 / /sys rw,relatime shared:2 - sysfs sysfs rw

Making the sysfs private under the new namespace so the umount won't trigger the global sysfs umount.

---
 tools/testing/selftests/bpf/prog_tests/tc_redirect.c | 7 +++++++
 1 file changed, 7 insertions(+)
tools/testing/selftests/bpf/prog_tests/tc_redirect.c +7 -0
index 4b18b73df10b6..c2426df58e172 100644
@@ -105,6 +105,13 @@ static int setns_by_fd(int nsfd)
 	if (!ASSERT_OK(err, "unshare"))
 		return err;
 
+	/* Make our /sys mount private, so the following umount won't
+	 * trigger the global umount in case it's shared.
+	 */
+	err = mount("none", "/sys", NULL, MS_PRIVATE, NULL);
+	if (!ASSERT_OK(err, "remount private /sys"))
+		return err;
+
 	err = umount2("/sys", MNT_DETACH);
 	if (!ASSERT_OK(err, "umount2 /sys"))
 		return err;
Signature · 1 line
2.34.1
author maintainer
#720 lore raw

PATCH 717/1039 mlxsw: pci: Avoid flow control for EMAD packets

In reply to Greg Kroah-Hartman

From: Danielle Ratson <[email protected]>

[ Upstream commit d43e4271747ace01a27a49a97a397cb4219f6487 ]

Locally generated packets ingress the device through its CPU port. When the CPU port is congested and there are not enough credits in its headroom buffer, packets can be dropped.

While this might be acceptable for data packets that traverse the network, configuration packets exchanged between the host and the device (EMADs) should not be subjected to this flow control.

The "sdq_lp" bit in the SDQ (Send Descriptor Queue) context allows the host to instruct the device to treat packets sent on this queue as "local processing" and always process them, regardless of the state of the CPU port's headroom.

Add the definition of this bit and set it for the dedicated SDQ reserved for the transmission of EMAD packets. This makes the "local processing" bit in the WQE (Work Queue Element) redundant, so clear it.

---
 drivers/net/ethernet/mellanox/mlxsw/cmd.h | 12 ++++++++++++
 drivers/net/ethernet/mellanox/mlxsw/pci.c |  6 +++++-
 2 files changed, 17 insertions(+), 1 deletion(-)
drivers/net/ethernet/mellanox/mlxsw/cmd.h +12 -0
index 392ce3cb27f72..51b260d54237e 100644
@@ -935,6 +935,18 @@ static inline int mlxsw_cmd_sw2hw_rdq(struct mlxsw_core *mlxsw_core,
  */
 MLXSW_ITEM32(cmd_mbox, sw2hw_dq, cq, 0x00, 24, 8);
 
+enum mlxsw_cmd_mbox_sw2hw_dq_sdq_lp {
+	MLXSW_CMD_MBOX_SW2HW_DQ_SDQ_LP_WQE,
+	MLXSW_CMD_MBOX_SW2HW_DQ_SDQ_LP_IGNORE_WQE,
+};
+
+/* cmd_mbox_sw2hw_dq_sdq_lp
+ * SDQ local Processing
+ * 0: local processing by wqe.lp
+ * 1: local processing (ignoring wqe.lp)
+ */
+MLXSW_ITEM32(cmd_mbox, sw2hw_dq, sdq_lp, 0x00, 23, 1);
+
 /* cmd_mbox_sw2hw_dq_sdq_tclass
  * SDQ: CPU Egress TClass
  * RDQ: Reserved
drivers/net/ethernet/mellanox/mlxsw/pci.c +5 -1
index cd3331a077bbf..f91dde4df152b 100644
@@ -285,6 +285,7 @@ static int mlxsw_pci_sdq_init(struct mlxsw_pci *mlxsw_pci, char *mbox,
 			      struct mlxsw_pci_queue *q)
 {
 	int tclass;
+	int lp;
 	int i;
 	int err;
 
@@ -292,9 +293,12 @@ static int mlxsw_pci_sdq_init(struct mlxsw_pci *mlxsw_pci, char *mbox,
 	q->consumer_counter = 0;
 	tclass = q->num == MLXSW_PCI_SDQ_EMAD_INDEX ? MLXSW_PCI_SDQ_EMAD_TC :
 						      MLXSW_PCI_SDQ_CTL_TC;
+	lp = q->num == MLXSW_PCI_SDQ_EMAD_INDEX ? MLXSW_CMD_MBOX_SW2HW_DQ_SDQ_LP_IGNORE_WQE :
+						  MLXSW_CMD_MBOX_SW2HW_DQ_SDQ_LP_WQE;
 
 	/* Set CQ of same number of this SDQ. */
 	mlxsw_cmd_mbox_sw2hw_dq_cq_set(mbox, q->num);
+	mlxsw_cmd_mbox_sw2hw_dq_sdq_lp_set(mbox, lp);
 	mlxsw_cmd_mbox_sw2hw_dq_sdq_tclass_set(mbox, tclass);
 	mlxsw_cmd_mbox_sw2hw_dq_log2_dq_sz_set(mbox, 3); /* 8 pages */
 	for (i = 0; i < MLXSW_PCI_AQ_PAGES; i++) {
@@ -1678,7 +1682,7 @@ static int mlxsw_pci_skb_transmit(void *bus_priv, struct sk_buff *skb,
 
 	wqe = elem_info->elem;
 	mlxsw_pci_wqe_c_set(wqe, 1); /* always report completion */
-	mlxsw_pci_wqe_lp_set(wqe, !!tx_info->is_emad);
+	mlxsw_pci_wqe_lp_set(wqe, 0);
 	mlxsw_pci_wqe_type_set(wqe, MLXSW_PCI_WQE_TYPE_ETHERNET);
 
 	err = mlxsw_pci_wqe_frag_map(mlxsw_pci, wqe, 0, skb->data,
Signature · 1 line
2.34.1
author maintainer
#721 lore raw

PATCH 718/1039 net: phy: marvell: configure RGMII delays for 88E1118

In reply to Greg Kroah-Hartman

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

[ Upstream commit f22725c95ececb703c3f741e8f946d23705630b7 ]

Corentin Labbe reports that the SSI 1328 does not work when allowing the PHY to operate at gigabit speeds, but does work with the generic PHY driver.

This appears to be because m88e1118_config_init() writes a fixed value to the MSCR register, claiming that this is to enable 1G speeds. However, this always sets bits 4 and 5, enabling RGMII transmit and receive delays. The suspicion is that the original board this was added for required the delays to make 1G speeds work.

Add the necessary configuration for RGMII delays for the 88E1118 to bring this into line with the requirements for RGMII support, and thus make the SSI 1328 work.

Corentin Labbe has tested this on gemini-ssi1328 and gemini-ns2502.

---
 drivers/net/phy/marvell.c |    6 ++++++
 1 file changed, 6 insertions(+)
drivers/net/phy/marvell.c +6 -0
@@ -1244,6 +1244,12 @@ static int m88e1118_config_init(struct p
 	if (err < 0)
 		return err;
 
+	if (phy_interface_is_rgmii(phydev)) {
+		err = m88e1121_config_aneg_rgmii_delays(phydev);
+		if (err < 0)
+			return err;
+	}
+
 	/* Adjust LED Control */
 	if (phydev->dev_flags & MARVELL_PHY_M1118_DNS323_LEDS)
 		err = phy_write(phydev, 0x10, 0x1100);
author maintainer
#722 lore raw

PATCH 719/1039 net: gemini: allow any RGMII interface mode

In reply to Greg Kroah-Hartman

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

[ Upstream commit 4e4f325a0a55907b14f579e6b1a38c53755e3de2 ]

The four RGMII interface modes take care of the required RGMII delay configuration at the PHY and should not be limited by the network MAC driver. Sadly, gemini was only permitting RGMII mode with no delays, which would require the required delay to be inserted via PCB tracking or by the MAC.

However, there are designs that require the PHY to add the delay, which is impossible without Gemini permitting the other three PHY interface modes. Fix the driver to allow these.

---
 drivers/net/ethernet/cortina/gemini.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)
drivers/net/ethernet/cortina/gemini.c +6 -3
index 941f175fb911e..0ff40a9b06cec 100644
@@ -305,21 +305,21 @@ static void gmac_speed_set(struct net_device *netdev)
 	switch (phydev->speed) {
 	case 1000:
 		status.bits.speed = GMAC_SPEED_1000;
-		if (phydev->interface == PHY_INTERFACE_MODE_RGMII)
+		if (phy_interface_mode_is_rgmii(phydev->interface))
 			status.bits.mii_rmii = GMAC_PHY_RGMII_1000;
 		netdev_dbg(netdev, "connect %s to RGMII @ 1Gbit\n",
 			   phydev_name(phydev));
 		break;
 	case 100:
 		status.bits.speed = GMAC_SPEED_100;
-		if (phydev->interface == PHY_INTERFACE_MODE_RGMII)
+		if (phy_interface_mode_is_rgmii(phydev->interface))
 			status.bits.mii_rmii = GMAC_PHY_RGMII_100_10;
 		netdev_dbg(netdev, "connect %s to RGMII @ 100 Mbit\n",
 			   phydev_name(phydev));
 		break;
 	case 10:
 		status.bits.speed = GMAC_SPEED_10;
-		if (phydev->interface == PHY_INTERFACE_MODE_RGMII)
+		if (phy_interface_mode_is_rgmii(phydev->interface))
 			status.bits.mii_rmii = GMAC_PHY_RGMII_100_10;
 		netdev_dbg(netdev, "connect %s to RGMII @ 10 Mbit\n",
 			   phydev_name(phydev));
@@ -389,6 +389,9 @@ static int gmac_setup_phy(struct net_device *netdev)
 		status.bits.mii_rmii = GMAC_PHY_GMII;
 		break;
 	case PHY_INTERFACE_MODE_RGMII:
+	case PHY_INTERFACE_MODE_RGMII_ID:
+	case PHY_INTERFACE_MODE_RGMII_TXID:
+	case PHY_INTERFACE_MODE_RGMII_RXID:
 		netdev_dbg(netdev,
 			   "RGMII: set GMAC0 and GMAC1 to MII/RGMII mode\n");
 		status.bits.mii_rmii = GMAC_PHY_RGMII_100_10;
Signature · 1 line
2.34.1
author maintainer
#723 lore raw

PATCH 720/1039 regulator: qcom_smd: Align probe function with rpmh-regulator

In reply to Greg Kroah-Hartman

From: Konrad Dybcio <[email protected]>

[ Upstream commit 14e2976fbabdacb01335d7f91eeebbc89c67ddb1 ]

The RPMh regulator driver is much newer and gets more attention, which in consequence makes it do a few things better. Update qcom_smd-regulator's probe function to mimic what rpmh-regulator does to address a couple of issues:

- Probe defer now works correctly, before it used to, well, kinda just die.. This fixes reliable probing on (at least) PM8994, because Linux apparently cannot deal with supply map dependencies yet..

- Regulator data is now matched more sanely: regulator data is matched against each individual regulator node name and throwing an -EINVAL if data is missing, instead of just assuming everything is fine and iterating over all subsequent array members.

- status = "disabled" will now work for disabling individual regulators in DT. Previously it didn't seem to do much if anything at all.

---
 drivers/regulator/qcom_smd-regulator.c | 100 +++++++++++++++++--------
 1 file changed, 70 insertions(+), 30 deletions(-)
drivers/regulator/qcom_smd-regulator.c +70 -30
index 8bac024dde8b4..9fc666107a06c 100644
@@ -9,6 +9,7 @@
 #include <linux/of_device.h>
 #include <linux/platform_device.h>
 #include <linux/regulator/driver.h>
+#include <linux/regulator/of_regulator.h>
 #include <linux/soc/qcom/smd-rpm.h>
 
 struct qcom_rpm_reg {
@@ -1239,52 +1240,91 @@ static const struct of_device_id rpm_of_match[] = {
 };
 MODULE_DEVICE_TABLE(of, rpm_of_match);
 
-static int rpm_reg_probe(struct platform_device *pdev)
+/**
+ * rpm_regulator_init_vreg() - initialize all attributes of a qcom_smd-regulator
+ * @vreg:		Pointer to the individual qcom_smd-regulator resource
+ * @dev:		Pointer to the top level qcom_smd-regulator PMIC device
+ * @node:		Pointer to the individual qcom_smd-regulator resource
+ *			device node
+ * @rpm:		Pointer to the rpm bus node
+ * @pmic_rpm_data:	Pointer to a null-terminated array of qcom_smd-regulator
+ *			resources defined for the top level PMIC device
+ *
+ * Return: 0 on success, errno on failure
+ */
+static int rpm_regulator_init_vreg(struct qcom_rpm_reg *vreg, struct device *dev,
+				   struct device_node *node, struct qcom_smd_rpm *rpm,
+				   const struct rpm_regulator_data *pmic_rpm_data)
 {
-	const struct rpm_regulator_data *reg;
-	const struct of_device_id *match;
-	struct regulator_config config = { };
+	struct regulator_config config = {};
+	const struct rpm_regulator_data *rpm_data;
 	struct regulator_dev *rdev;
+	int ret;
+
+	for (rpm_data = pmic_rpm_data; rpm_data->name; rpm_data++)
+		if (of_node_name_eq(node, rpm_data->name))
+			break;
+
+	if (!rpm_data->name) {
+		dev_err(dev, "Unknown regulator %pOFn\n", node);
+		return -EINVAL;
+	}
+
+	vreg->dev	= dev;
+	vreg->rpm	= rpm;
+	vreg->type	= rpm_data->type;
+	vreg->id	= rpm_data->id;
+
+	memcpy(&vreg->desc, rpm_data->desc, sizeof(vreg->desc));
+	vreg->desc.name = rpm_data->name;
+	vreg->desc.supply_name = rpm_data->supply;
+	vreg->desc.owner = THIS_MODULE;
+	vreg->desc.type = REGULATOR_VOLTAGE;
+	vreg->desc.of_match = rpm_data->name;
+
+	config.dev		= dev;
+	config.of_node		= node;
+	config.driver_data	= vreg;
+
+	rdev = devm_regulator_register(dev, &vreg->desc, &config);
+	if (IS_ERR(rdev)) {
+		ret = PTR_ERR(rdev);
+		dev_err(dev, "%pOFn: devm_regulator_register() failed, ret=%d\n", node, ret);
+		return ret;
+	}
+
+	return 0;
+}
+
+static int rpm_reg_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	const struct rpm_regulator_data *vreg_data;
+	struct device_node *node;
 	struct qcom_rpm_reg *vreg;
 	struct qcom_smd_rpm *rpm;
+	int ret;
 
 	rpm = dev_get_drvdata(pdev->dev.parent);
 	if (!rpm) {
-		dev_err(&pdev->dev, "unable to retrieve handle to rpm\n");
+		dev_err(&pdev->dev, "Unable to retrieve handle to rpm\n");
 		return -ENODEV;
 	}
 
-	match = of_match_device(rpm_of_match, &pdev->dev);
-	if (!match) {
-		dev_err(&pdev->dev, "failed to match device\n");
+	vreg_data = of_device_get_match_data(dev);
+	if (!vreg_data)
 		return -ENODEV;
-	}
 
-	for (reg = match->data; reg->name; reg++) {
+	for_each_available_child_of_node(dev->of_node, node) {
 		vreg = devm_kzalloc(&pdev->dev, sizeof(*vreg), GFP_KERNEL);
 		if (!vreg)
 			return -ENOMEM;
 
-		vreg->dev = &pdev->dev;
-		vreg->type = reg->type;
-		vreg->id = reg->id;
-		vreg->rpm = rpm;
-
-		memcpy(&vreg->desc, reg->desc, sizeof(vreg->desc));
-
-		vreg->desc.id = -1;
-		vreg->desc.owner = THIS_MODULE;
-		vreg->desc.type = REGULATOR_VOLTAGE;
-		vreg->desc.name = reg->name;
-		vreg->desc.supply_name = reg->supply;
-		vreg->desc.of_match = reg->name;
-
-		config.dev = &pdev->dev;
-		config.driver_data = vreg;
-		rdev = devm_regulator_register(&pdev->dev, &vreg->desc, &config);
-		if (IS_ERR(rdev)) {
-			dev_err(&pdev->dev, "failed to register %s\n", reg->name);
-			return PTR_ERR(rdev);
+		ret = rpm_regulator_init_vreg(vreg, dev, node, rpm, vreg_data);
+
+		if (ret < 0) {
+			of_node_put(node);
+			return ret;
 		}
 	}
 
Signature · 1 line
2.34.1
author maintainer
#724 lore raw

PATCH 721/1039 serial: pl010: Drop CR register reset on set_termios

In reply to Greg Kroah-Hartman

From: Lukas Wunner <[email protected]>

[ Upstream commit 08a0c6dff91c965e39905cf200d22db989203ccb ]

pl010_set_termios() briefly resets the CR register to zero.

Where does this register write come from?

The PL010 driver's IRQ handler ambauart_int() originally modified the CR register without holding the port spinlock. ambauart_set_termios() also modified that register. To prevent concurrent read-modify-writes by the IRQ handler and to prevent transmission while changing baudrate, ambauart_set_termios() had to disable interrupts. That is achieved by writing zero to the CR register.

However in 2004 the PL010 driver was amended to acquire the port spinlock in the IRQ handler, obviating the need to disable interrupts in ->set_termios():
https://git.kernel.org/history/history/c/157c0342e591

That rendered the CR register write obsolete. Drop it.

---
 drivers/tty/serial/amba-pl010.c | 3 ---
 1 file changed, 3 deletions(-)
drivers/tty/serial/amba-pl010.c +0 -3
index e744b953ca346..47654073123d6 100644
@@ -446,14 +446,11 @@ pl010_set_termios(struct uart_port *port, struct ktermios *termios,
 	if ((termios->c_cflag & CREAD) == 0)
 		uap->port.ignore_status_mask |= UART_DUMMY_RSR_RX;
 
-	/* first, disable everything */
 	old_cr = readb(uap->port.membase + UART010_CR) & ~UART010_CR_MSIE;
 
 	if (UART_ENABLE_MS(port, termios->c_cflag))
 		old_cr |= UART010_CR_MSIE;
 
-	writel(0, uap->port.membase + UART010_CR);
-
 	/* Set baud rate */
 	quot -= 1;
 	writel((quot & 0xf00) >> 8, uap->port.membase + UART010_LCRM);
Signature · 1 line
2.34.1
author maintainer
#725 lore raw

PATCH 722/1039 serial: pl011: Drop CR register reset on set_termios

In reply to Greg Kroah-Hartman

From: Lukas Wunner <[email protected]>

[ Upstream commit e368cc656fd6d0075f1c3ab9676e2001451e3e04 ]

pl011_set_termios() briefly resets the CR register to zero, thereby glitching DTR/RTS signals. With rs485 this may result in the bus being occupied for no reason.

Where does this register write originate from?

The PL011 driver was forked from the PL010 driver in 2004:
https://git.kernel.org/history/history/c/157c0342e591

Until this commit, the PL010 driver's IRQ handler ambauart_int() modified the CR register without holding the port spinlock.

ambauart_set_termios() also modified that register. To prevent concurrent read-modify-writes by the IRQ handler and to prevent transmission while changing baudrate, ambauart_set_termios() had to disable interrupts. On the PL010, that is achieved by writing zero to the CR register.

However, on the PL011, interrupts are disabled in the IMSC register, not in the CR register.

Additionally, the commit amended both the PL010 and PL011 driver to acquire the port spinlock in the IRQ handler, obviating the need to disable interrupts in ->set_termios().

So the CR register write is obsolete for two reasons. Drop it.

---
 drivers/tty/serial/amba-pl011.c | 2 --
 1 file changed, 2 deletions(-)
drivers/tty/serial/amba-pl011.c +0 -2
index b831d4d64c0a2..6ec34260d6b18 100644
@@ -2105,9 +2105,7 @@ pl011_set_termios(struct uart_port *port, struct ktermios *termios,
 	if (port->rs485.flags & SER_RS485_ENABLED)
 		termios->c_cflag &= ~CRTSCTS;
 
-	/* first, disable everything */
 	old_cr = pl011_read(uap, REG_CR);
-	pl011_write(0, uap, REG_CR);
 
 	if (termios->c_cflag & CRTSCTS) {
 		if (old_cr & UART011_CR_RTS)
Signature · 1 line
2.34.1
author maintainer
#726 lore raw

PATCH 723/1039 serial: core: Keep mctrl register state and cached copy in sync

In reply to Greg Kroah-Hartman

From: Lukas Wunner <[email protected]>

[ Upstream commit 93a770b7e16772530196674ffc79bb13fa927dc6 ]

struct uart_port contains a cached copy of the Modem Control signals. It is used to skip register writes in uart_update_mctrl() if the new signal state equals the old signal state. It also avoids a register read to obtain the current state of output signals.

When a uart_port is registered, uart_configure_port() changes signal state but neglects to keep the cached copy in sync. That may cause a subsequent register write to be incorrectly skipped. Fix it before it trips somebody up.

This behavior has been present ever since the serial core was introduced in 2002:
https://git.kernel.org/history/history/c/33c0d1b0c3eb

So far it was never an issue because the cached copy is initialized to 0 by kzalloc() and when uart_configure_port() is executed, at most DTR has been set by uart_set_options() or sunsu_console_setup(). Therefore, a stable designation seems unnecessary.

---
 drivers/tty/serial/serial_core.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/tty/serial/serial_core.c +2 -1
index 61e3dd0222af1..9e7e624a6c9db 100644
@@ -2393,7 +2393,8 @@ uart_configure_port(struct uart_driver *drv, struct uart_state *state,
 		 * We probably don't need a spinlock around this, but
 		 */
 		spin_lock_irqsave(&port->lock, flags);
-		port->ops->set_mctrl(port, port->mctrl & TIOCM_DTR);
+		port->mctrl &= TIOCM_DTR;
+		port->ops->set_mctrl(port, port->mctrl);
 		spin_unlock_irqrestore(&port->lock, flags);
 
 		/*
Signature · 1 line
2.34.1
author maintainer
#727 lore raw

PATCH 724/1039 random: do not throw away excess input to crng_fast_load

In reply to Greg Kroah-Hartman

From: Jason A. Donenfeld <[email protected]>

[ Upstream commit 73c7733f122e8d0107f88655a12011f68f69e74b ]

When crng_fast_load() is called by add_hwgenerator_randomness(), we currently will advance to crng_init==1 once we've acquired 64 bytes, and then throw away the rest of the buffer. Usually, that is not a problem:
When add_hwgenerator_randomness() gets called via EFI or DT during setup_arch(), there won't be any IRQ randomness. Therefore, the 64 bytes passed by EFI exactly matches what is needed to advance to crng_init==1. Usually, DT seems to pass 64 bytes as well -- with one notable exception being kexec, which hands over 128 bytes of entropy to the kexec'd kernel. In that case, we'll advance to crng_init==1 once 64 of those bytes are consumed by crng_fast_load(), but won't continue onward feeding in bytes to progress to crng_init==2. This commit fixes the issue by feeding any leftover bytes into the next phase in add_hwgenerator_randomness().

---
 drivers/char/random.c | 19 ++++++++++++-------
 1 file changed, 12 insertions(+), 7 deletions(-)
drivers/char/random.c +12 -7
index 7470ee24db2f9..a27ae3999ff32 100644
@@ -912,12 +912,14 @@ static struct crng_state *select_crng(void)
 
 /*
  * crng_fast_load() can be called by code in the interrupt service
- * path.  So we can't afford to dilly-dally.
+ * path.  So we can't afford to dilly-dally. Returns the number of
+ * bytes processed from cp.
  */
-static int crng_fast_load(const char *cp, size_t len)
+static size_t crng_fast_load(const char *cp, size_t len)
 {
 	unsigned long flags;
 	char *p;
+	size_t ret = 0;
 
 	if (!spin_trylock_irqsave(&primary_crng.lock, flags))
 		return 0;
@@ -928,7 +930,7 @@ static int crng_fast_load(const char *cp, size_t len)
 	p = (unsigned char *) &primary_crng.state[4];
 	while (len > 0 && crng_init_cnt < CRNG_INIT_CNT_THRESH) {
 		p[crng_init_cnt % CHACHA_KEY_SIZE] ^= *cp;
-		cp++; crng_init_cnt++; len--;
+		cp++; crng_init_cnt++; len--; ret++;
 	}
 	spin_unlock_irqrestore(&primary_crng.lock, flags);
 	if (crng_init_cnt >= CRNG_INIT_CNT_THRESH) {
@@ -936,7 +938,7 @@ static int crng_fast_load(const char *cp, size_t len)
 		crng_init = 1;
 		pr_notice("fast init done\n");
 	}
-	return 1;
+	return ret;
 }
 
 /*
@@ -1287,7 +1289,7 @@ void add_interrupt_randomness(int irq, int irq_flags)
 	if (unlikely(crng_init == 0)) {
 		if ((fast_pool->count >= 64) &&
 		    crng_fast_load((char *) fast_pool->pool,
-				   sizeof(fast_pool->pool))) {
+				   sizeof(fast_pool->pool)) > 0) {
 			fast_pool->count = 0;
 			fast_pool->last = now;
 		}
@@ -2295,8 +2297,11 @@ void add_hwgenerator_randomness(const char *buffer, size_t count,
 	struct entropy_store *poolp = &input_pool;
 
 	if (unlikely(crng_init == 0)) {
-		crng_fast_load(buffer, count);
-		return;
+		size_t ret = crng_fast_load(buffer, count);
+		count -= ret;
+		buffer += ret;
+		if (!count || crng_init == 0)
+			return;
 	}
 
 	/* Suspend writing if we're above the trickle threshold.
Signature · 1 line
2.34.1
author maintainer
#728 lore raw

PATCH 725/1039 net/mlx5: Update log_max_qp value to FW max capability

In reply to Greg Kroah-Hartman

From: Maher Sanalla <[email protected]>

[ Upstream commit f79a609ea6bf54ad2d2c24e4de4524288b221666 ]

log_max_qp in driver's default profile #2 was set to 18, but FW actually supports 17 at the most - a situation that led to the concerning print when the driver is loaded:
"log_max_qp value in current profile is 18, changing to HCA capabaility limit (17)"

The expected behavior from mlx5_profile #2 is to match the maximum FW capability in regards to log_max_qp. Thus, log_max_qp in profile #2 is initialized to a defined static value (0xff) - which basically means that when loading this profile, log_max_qp value will be what the currently installed FW supports at most.

---
 drivers/net/ethernet/mellanox/mlx5/core/main.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/main.c +6 -2
index 65083496f9131..6e381111f1d2f 100644
@@ -98,6 +98,8 @@ enum {
 	MLX5_ATOMIC_REQ_MODE_HOST_ENDIANNESS = 0x1,
 };
 
+#define LOG_MAX_SUPPORTED_QPS 0xff
+
 static struct mlx5_profile profile[] = {
 	[0] = {
 		.mask           = 0,
@@ -109,7 +111,7 @@ static struct mlx5_profile profile[] = {
 	[2] = {
 		.mask		= MLX5_PROF_MASK_QP_SIZE |
 				  MLX5_PROF_MASK_MR_CACHE,
-		.log_max_qp	= 18,
+		.log_max_qp	= LOG_MAX_SUPPORTED_QPS,
 		.mr_cache[0]	= {
 			.size	= 500,
 			.limit	= 250
@@ -507,7 +509,9 @@ static int handle_hca_cap(struct mlx5_core_dev *dev, void *set_ctx)
 		 to_fw_pkey_sz(dev, 128));
 
 	/* Check log_max_qp from HCA caps to set in current profile */
-	if (MLX5_CAP_GEN_MAX(dev, log_max_qp) < prof->log_max_qp) {
+	if (prof->log_max_qp == LOG_MAX_SUPPORTED_QPS) {
+		prof->log_max_qp = MLX5_CAP_GEN_MAX(dev, log_max_qp);
+	} else if (MLX5_CAP_GEN_MAX(dev, log_max_qp) < prof->log_max_qp) {
 		mlx5_core_warn(dev, "log_max_qp value in current profile is %d, changing it to HCA capability limit (%d)\n",
 			       prof->log_max_qp,
 			       MLX5_CAP_GEN_MAX(dev, log_max_qp));
Signature · 1 line
2.34.1
author maintainer
#729 lore raw

PATCH 726/1039 net/mlx5e: Unblock setting vid 0 for VF in case PF isnt eswitch manager

In reply to Greg Kroah-Hartman

From: Maor Dickman <[email protected]>

[ Upstream commit 7846665d3504812acaebf920d1141851379a7f37 ]

When using libvirt to passthrough VF to VM it will always set the VF vlan to 0 even if user didn’t request it, this will cause libvirt to fail to boot in case the PF isn't eswitch owner.

Example of such case is the DPU host PF which isn't eswitch manager, so any attempt to passthrough VF of it using libvirt will fail.

Fix it by not returning error in case set VF vlan is called with vid 0.

---
 drivers/net/ethernet/mellanox/mlx5/core/esw/legacy.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/mellanox/mlx5/core/esw/legacy.c +1 -1
index df277a6cddc0b..0c4c743ca31e1 100644
@@ -431,7 +431,7 @@ int mlx5_eswitch_set_vport_vlan(struct mlx5_eswitch *esw,
 	int err = 0;
 
 	if (!mlx5_esw_allowed(esw))
-		return -EPERM;
+		return vlan ? -EPERM : 0;
 
 	if (vlan || qos)
 		set_flags = SET_VLAN_STRIP | SET_VLAN_INSERT;
Signature · 1 line
2.34.1
author maintainer
#730 lore raw

PATCH 727/1039 parisc: Avoid calling faulthandler_disabled() twice

In reply to Greg Kroah-Hartman

From: John David Anglin <[email protected]>

[ Upstream commit 9e9d4b460f23bab61672eae397417d03917d116c ]

In handle_interruption(), we call faulthandler_disabled() to check whether the fault handler is not disabled. If the fault handler is disabled, we immediately call do_page_fault(). It then calls faulthandler_disabled(). If disabled, do_page_fault() attempts to fixup the exception by jumping to no_context:

no_context:

        if (!user_mode(regs) && fixup_exception(regs)) {
                return;
        }

        parisc_terminate("Bad Address (null pointer deref?)", regs, code, address);

Apart from the error messages, the two blocks of code perform the same function.

We can avoid two calls to faulthandler_disabled() by a simple revision to the code in handle_interruption().

Note: I didn't try to fix the formatting of this code block.

---
 arch/parisc/kernel/traps.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/parisc/kernel/traps.c +1 -1
index 892b7fc8f3c45..eb41fece19104 100644
@@ -785,7 +785,7 @@ void notrace handle_interruption(int code, struct pt_regs *regs)
 	     * unless pagefault_disable() was called before.
 	     */
 
-	    if (fault_space == 0 && !faulthandler_disabled())
+	    if (faulthandler_disabled() || fault_space == 0)
 	    {
 		/* Clean up and return if in exception table. */
 		if (fixup_exception(regs))
Signature · 1 line
2.34.1
author maintainer
#731 lore raw

PATCH 728/1039 scripts: sphinx-pre-install: Fix ctex support on Debian

In reply to Greg Kroah-Hartman

From: Mauro Carvalho Chehab <[email protected]>

[ Upstream commit 87d6576ddf8ac25f36597bc93ca17f6628289c16 ]

The name of the package with ctexhook.sty is different on Debian/Ubuntu.

---
 scripts/sphinx-pre-install | 3 +++
 1 file changed, 3 insertions(+)
scripts/sphinx-pre-install +3 -0
index 288e86a9d1e58..61a79ce705ccf 100755
@@ -369,6 +369,9 @@ sub give_debian_hints()
 	);
 
 	if ($pdf) {
+		check_missing_file(["/usr/share/texlive/texmf-dist/tex/latex/ctex/ctexhook.sty"],
+				   "texlive-lang-chinese", 2);
+
 		check_missing_file(["/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf"],
 				   "fonts-dejavu", 2);
 
Signature · 1 line
2.34.1
#732 lore raw

Re: [PATCH 5.16 0728/1039] scripts: sphinx-pre-install: Fix ctex support on Debian

Hi Greg,

Quoted 7 lines · Greg Kroah-Hartman

From: Mauro Carvalho Chehab <[email protected]>

[ Upstream commit 87d6576ddf8ac25f36597bc93ca17f6628289c16 ]

The name of the package with ctexhook.sty is different on Debian/Ubuntu.

Quoted 6 lines

This "Fix" is against upstream commit 7baab965896e ("scripts:
sphinx-pre-install: add required ctex dependency") which is also new to v5.17-rc1.

So I don't think this is worth backporting to stable branches.

Thanks, Akira

Quoted 18 lines
---
 scripts/sphinx-pre-install | 3 +++
 1 file changed, 3 insertions(+)
scripts/sphinx-pre-install +3 -0
index 288e86a9d1e58..61a79ce705ccf 100755
@@ -369,6 +369,9 @@ sub give_debian_hints()
 	);
 
 	if ($pdf) {
+		check_missing_file(["/usr/share/texlive/texmf-dist/tex/latex/ctex/ctexhook.sty"],
+				   "texlive-lang-chinese", 2);
+
 		check_missing_file(["/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf"],
 				   "fonts-dejavu", 2);
 
#733 lore raw

Re: [PATCH 5.16 0728/1039] scripts: sphinx-pre-install: Fix ctex support on Debian

Em Tue, 25 Jan 2022 07:48:39 +0900
Akira Yokosawa <[email protected]> escreveu:

Quoted 18 lines

Hi Greg,

Quoted 7 lines · Greg Kroah-Hartman

From: Mauro Carvalho Chehab <[email protected]>

[ Upstream commit 87d6576ddf8ac25f36597bc93ca17f6628289c16 ]

The name of the package with ctexhook.sty is different on Debian/Ubuntu.

Quoted 6 lines
Quoted 5 lines

This "Fix" is against upstream commit 7baab965896e ("scripts:
sphinx-pre-install: add required ctex dependency") which is also new to v5.17-rc1.

So I don't think this is worth backporting to stable branches.

Well, either both patches should be backported or none, IMHO. I guess I would just backport also commit 7baab965896e.

Quoted 14 lines

Thanks, Akira

Quoted 11 lines
---
 scripts/sphinx-pre-install | 3 +++
 1 file changed, 3 insertions(+)
scripts/sphinx-pre-install +0 -0
index 288e86a9d1e58..61a79ce705ccf 100755
@@ -369,6 +369,9 @@ sub give_debian_hints()
 	);
 
Quoted 7 lines

scripts/sphinx-pre-install sub give_debian_hints() · line 377

Quoted 7 lines
scripts/sphinx-pre-install +3 -0
 	if ($pdf) {
+		check_missing_file(["/usr/share/texlive/texmf-dist/tex/latex/ctex/ctexhook.sty"],
+				   "texlive-lang-chinese", 2);
+
 		check_missing_file(["/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf"],
 				   "fonts-dejavu", 2);
 

Thanks,
Mauro

author maintainer
#734 lore raw

Re: [PATCH 5.16 0728/1039] scripts: sphinx-pre-install: Fix ctex support on Debian

Quoted 21 lines · Mauro Carvalho Chehab

Em Tue, 25 Jan 2022 07:48:39 +0900
Akira Yokosawa <[email protected]> escreveu:

Quoted 18 lines

Hi Greg,

Quoted 13 lines · Greg Kroah-Hartman

From: Mauro Carvalho Chehab <[email protected]>

[ Upstream commit 87d6576ddf8ac25f36597bc93ca17f6628289c16 ]

The name of the package with ctexhook.sty is different on Debian/Ubuntu.

Quoted 8 lines
Quoted 5 lines

This "Fix" is against upstream commit 7baab965896e ("scripts:
sphinx-pre-install: add required ctex dependency") which is also new to v5.17-rc1.

So I don't think this is worth backporting to stable branches.

Well, either both patches should be backported or none, IMHO. I guess I would just backport also commit 7baab965896e.

Ok, in looking at that report, it makes sense, I'll add both of these now.

thanks,

greg k-h

author maintainer
#735 lore raw

Re: [PATCH 5.16 0728/1039] scripts: sphinx-pre-install: Fix ctex support on Debian

In reply to Akira Yokosawa
Quoted 18 lines · Akira Yokosawa

Hi Greg,

Quoted 7 lines · Greg Kroah-Hartman

From: Mauro Carvalho Chehab <[email protected]>

[ Upstream commit 87d6576ddf8ac25f36597bc93ca17f6628289c16 ]

The name of the package with ctexhook.sty is different on Debian/Ubuntu.

Quoted 6 lines
Quoted 5 lines

This "Fix" is against upstream commit 7baab965896e ("scripts:
sphinx-pre-install: add required ctex dependency") which is also new to v5.17-rc1.

So I don't think this is worth backporting to stable branches.

Now dropped from all trees, thanks!

greg k-h

author maintainer
#736 lore raw

PATCH 729/1039 can: flexcan: allow to change quirks at runtime

In reply to Greg Kroah-Hartman

From: Dario Binacchi <[email protected]>

[ Upstream commit 01bb4dccd92b4dc21f6af3312e5696924e371111 ]

This is a preparation patch for the upcoming support to change the rx-rtr capability via the ethtool API.

---
 drivers/net/can/flexcan.c | 54 +++++++++++++++++++--------------------
 1 file changed, 27 insertions(+), 27 deletions(-)
drivers/net/can/flexcan.c +27 -27
index 12b60ad95b02a..26bf0a0a72f1d 100644
@@ -369,7 +369,7 @@ struct flexcan_priv {
 
 	struct clk *clk_ipg;
 	struct clk *clk_per;
-	const struct flexcan_devtype_data *devtype_data;
+	struct flexcan_devtype_data devtype_data;
 	struct regulator *reg_xceiver;
 	struct flexcan_stop_mode stm;
 
@@ -600,7 +600,7 @@ static inline int flexcan_enter_stop_mode(struct flexcan_priv *priv)
 	priv->write(reg_mcr, &regs->mcr);
 
 	/* enable stop request */
-	if (priv->devtype_data->quirks & FLEXCAN_QUIRK_SETUP_STOP_MODE_SCFW) {
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_SETUP_STOP_MODE_SCFW) {
 		ret = flexcan_stop_mode_enable_scfw(priv, true);
 		if (ret < 0)
 			return ret;
@@ -619,7 +619,7 @@ static inline int flexcan_exit_stop_mode(struct flexcan_priv *priv)
 	int ret;
 
 	/* remove stop request */
-	if (priv->devtype_data->quirks & FLEXCAN_QUIRK_SETUP_STOP_MODE_SCFW) {
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_SETUP_STOP_MODE_SCFW) {
 		ret = flexcan_stop_mode_enable_scfw(priv, false);
 		if (ret < 0)
 			return ret;
@@ -1022,7 +1022,7 @@ static struct sk_buff *flexcan_mailbox_read(struct can_rx_offload *offload,
 
 	mb = flexcan_get_mb(priv, n);
 
-	if (priv->devtype_data->quirks & FLEXCAN_QUIRK_USE_OFF_TIMESTAMP) {
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_USE_OFF_TIMESTAMP) {
 		u32 code;
 
 		do {
@@ -1087,7 +1087,7 @@ static struct sk_buff *flexcan_mailbox_read(struct can_rx_offload *offload,
 	}
 
  mark_as_read:
-	if (priv->devtype_data->quirks & FLEXCAN_QUIRK_USE_OFF_TIMESTAMP)
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_USE_OFF_TIMESTAMP)
 		flexcan_write64(priv, FLEXCAN_IFLAG_MB(n), &regs->iflag1);
 	else
 		priv->write(FLEXCAN_IFLAG_RX_FIFO_AVAILABLE, &regs->iflag1);
@@ -1113,7 +1113,7 @@ static irqreturn_t flexcan_irq(int irq, void *dev_id)
 	enum can_state last_state = priv->can.state;
 
 	/* reception interrupt */
-	if (priv->devtype_data->quirks & FLEXCAN_QUIRK_USE_OFF_TIMESTAMP) {
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_USE_OFF_TIMESTAMP) {
 		u64 reg_iflag_rx;
 		int ret;
 
@@ -1173,7 +1173,7 @@ static irqreturn_t flexcan_irq(int irq, void *dev_id)
 
 	/* state change interrupt or broken error state quirk fix is enabled */
 	if ((reg_esr & FLEXCAN_ESR_ERR_STATE) ||
-	    (priv->devtype_data->quirks & (FLEXCAN_QUIRK_BROKEN_WERR_STATE |
+	    (priv->devtype_data.quirks & (FLEXCAN_QUIRK_BROKEN_WERR_STATE |
 					   FLEXCAN_QUIRK_BROKEN_PERR_STATE)))
 		flexcan_irq_state(dev, reg_esr);
 
@@ -1195,11 +1195,11 @@ static irqreturn_t flexcan_irq(int irq, void *dev_id)
 	 * (1): enabled if FLEXCAN_QUIRK_BROKEN_WERR_STATE is enabled
 	 */
 	if ((last_state != priv->can.state) &&
-	    (priv->devtype_data->quirks & FLEXCAN_QUIRK_BROKEN_PERR_STATE) &&
+	    (priv->devtype_data.quirks & FLEXCAN_QUIRK_BROKEN_PERR_STATE) &&
 	    !(priv->can.ctrlmode & CAN_CTRLMODE_BERR_REPORTING)) {
 		switch (priv->can.state) {
 		case CAN_STATE_ERROR_ACTIVE:
-			if (priv->devtype_data->quirks &
+			if (priv->devtype_data.quirks &
 			    FLEXCAN_QUIRK_BROKEN_WERR_STATE)
 				flexcan_error_irq_enable(priv);
 			else
@@ -1423,13 +1423,13 @@ static int flexcan_rx_offload_setup(struct net_device *dev)
 	else
 		priv->mb_size = sizeof(struct flexcan_mb) + CAN_MAX_DLEN;
 
-	if (priv->devtype_data->quirks & FLEXCAN_QUIRK_NR_MB_16)
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_NR_MB_16)
 		priv->mb_count = 16;
 	else
 		priv->mb_count = (sizeof(priv->regs->mb[0]) / priv->mb_size) +
 				 (sizeof(priv->regs->mb[1]) / priv->mb_size);
 
-	if (priv->devtype_data->quirks & FLEXCAN_QUIRK_USE_OFF_TIMESTAMP)
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_USE_OFF_TIMESTAMP)
 		priv->tx_mb_reserved =
 			flexcan_get_mb(priv, FLEXCAN_TX_MB_RESERVED_OFF_TIMESTAMP);
 	else
@@ -1441,7 +1441,7 @@ static int flexcan_rx_offload_setup(struct net_device *dev)
 
 	priv->offload.mailbox_read = flexcan_mailbox_read;
 
-	if (priv->devtype_data->quirks & FLEXCAN_QUIRK_USE_OFF_TIMESTAMP) {
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_USE_OFF_TIMESTAMP) {
 		priv->offload.mb_first = FLEXCAN_RX_MB_OFF_TIMESTAMP_FIRST;
 		priv->offload.mb_last = priv->mb_count - 2;
 
@@ -1506,7 +1506,7 @@ static int flexcan_chip_start(struct net_device *dev)
 	if (err)
 		goto out_chip_disable;
 
-	if (priv->devtype_data->quirks & FLEXCAN_QUIRK_SUPPORT_ECC)
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_SUPPORT_ECC)
 		flexcan_ram_init(dev);
 
 	flexcan_set_bittiming(dev);
@@ -1535,7 +1535,7 @@ static int flexcan_chip_start(struct net_device *dev)
 	 * - disable for timestamp mode
 	 * - enable for FIFO mode
 	 */
-	if (priv->devtype_data->quirks & FLEXCAN_QUIRK_USE_OFF_TIMESTAMP)
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_USE_OFF_TIMESTAMP)
 		reg_mcr &= ~FLEXCAN_MCR_FEN;
 	else
 		reg_mcr |= FLEXCAN_MCR_FEN;
@@ -1586,7 +1586,7 @@ static int flexcan_chip_start(struct net_device *dev)
 	 * on most Flexcan cores, too. Otherwise we don't get
 	 * any error warning or passive interrupts.
 	 */
-	if (priv->devtype_data->quirks & FLEXCAN_QUIRK_BROKEN_WERR_STATE ||
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_BROKEN_WERR_STATE ||
 	    priv->can.ctrlmode & CAN_CTRLMODE_BERR_REPORTING)
 		reg_ctrl |= FLEXCAN_CTRL_ERR_MSK;
 	else
@@ -1599,7 +1599,7 @@ static int flexcan_chip_start(struct net_device *dev)
 	netdev_dbg(dev, "%s: writing ctrl=0x%08x", __func__, reg_ctrl);
 	priv->write(reg_ctrl, &regs->ctrl);
 
-	if ((priv->devtype_data->quirks & FLEXCAN_QUIRK_ENABLE_EACEN_RRS)) {
+	if ((priv->devtype_data.quirks & FLEXCAN_QUIRK_ENABLE_EACEN_RRS)) {
 		reg_ctrl2 = priv->read(&regs->ctrl2);
 		reg_ctrl2 |= FLEXCAN_CTRL2_EACEN | FLEXCAN_CTRL2_RRS;
 		priv->write(reg_ctrl2, &regs->ctrl2);
@@ -1631,7 +1631,7 @@ static int flexcan_chip_start(struct net_device *dev)
 		priv->write(reg_fdctrl, &regs->fdctrl);
 	}
 
-	if (priv->devtype_data->quirks & FLEXCAN_QUIRK_USE_OFF_TIMESTAMP) {
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_USE_OFF_TIMESTAMP) {
 		for (i = priv->offload.mb_first; i <= priv->offload.mb_last; i++) {
 			mb = flexcan_get_mb(priv, i);
 			priv->write(FLEXCAN_MB_CODE_RX_EMPTY,
@@ -1659,7 +1659,7 @@ static int flexcan_chip_start(struct net_device *dev)
 	priv->write(0x0, &regs->rx14mask);
 	priv->write(0x0, &regs->rx15mask);
 
-	if (priv->devtype_data->quirks & FLEXCAN_QUIRK_DISABLE_RXFG)
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_DISABLE_RXFG)
 		priv->write(0x0, &regs->rxfgmask);
 
 	/* clear acceptance filters */
@@ -1673,7 +1673,7 @@ static int flexcan_chip_start(struct net_device *dev)
 	 * This also works around errata e5295 which generates false
 	 * positive memory errors and put the device in freeze mode.
 	 */
-	if (priv->devtype_data->quirks & FLEXCAN_QUIRK_DISABLE_MECR) {
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_DISABLE_MECR) {
 		/* Follow the protocol as described in "Detection
 		 * and Correction of Memory Errors" to write to
 		 * MECR register (step 1 - 5)
@@ -1799,7 +1799,7 @@ static int flexcan_open(struct net_device *dev)
 	if (err)
 		goto out_can_rx_offload_disable;
 
-	if (priv->devtype_data->quirks & FLEXCAN_QUIRK_NR_IRQ_3) {
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_NR_IRQ_3) {
 		err = request_irq(priv->irq_boff,
 				  flexcan_irq, IRQF_SHARED, dev->name, dev);
 		if (err)
@@ -1845,7 +1845,7 @@ static int flexcan_close(struct net_device *dev)
 	netif_stop_queue(dev);
 	flexcan_chip_interrupts_disable(dev);
 
-	if (priv->devtype_data->quirks & FLEXCAN_QUIRK_NR_IRQ_3) {
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_NR_IRQ_3) {
 		free_irq(priv->irq_err, dev);
 		free_irq(priv->irq_boff, dev);
 	}
@@ -2051,9 +2051,9 @@ static int flexcan_setup_stop_mode(struct platform_device *pdev)
 
 	priv = netdev_priv(dev);
 
-	if (priv->devtype_data->quirks & FLEXCAN_QUIRK_SETUP_STOP_MODE_SCFW)
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_SETUP_STOP_MODE_SCFW)
 		ret = flexcan_setup_stop_mode_scfw(pdev);
-	else if (priv->devtype_data->quirks & FLEXCAN_QUIRK_SETUP_STOP_MODE_GPR)
+	else if (priv->devtype_data.quirks & FLEXCAN_QUIRK_SETUP_STOP_MODE_GPR)
 		ret = flexcan_setup_stop_mode_gpr(pdev);
 	else
 		/* return 0 directly if doesn't support stop mode feature */
@@ -2181,9 +2181,10 @@ static int flexcan_probe(struct platform_device *pdev)
 	dev->flags |= IFF_ECHO;
 
 	priv = netdev_priv(dev);
+	priv->devtype_data = *devtype_data;
 
 	if (of_property_read_bool(pdev->dev.of_node, "big-endian") ||
-	    devtype_data->quirks & FLEXCAN_QUIRK_DEFAULT_BIG_ENDIAN) {
+	    priv->devtype_data.quirks & FLEXCAN_QUIRK_DEFAULT_BIG_ENDIAN) {
 		priv->read = flexcan_read_be;
 		priv->write = flexcan_write_be;
 	} else {
@@ -2202,10 +2203,9 @@ static int flexcan_probe(struct platform_device *pdev)
 	priv->clk_ipg = clk_ipg;
 	priv->clk_per = clk_per;
 	priv->clk_src = clk_src;
-	priv->devtype_data = devtype_data;
 	priv->reg_xceiver = reg_xceiver;
 
-	if (devtype_data->quirks & FLEXCAN_QUIRK_NR_IRQ_3) {
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_NR_IRQ_3) {
 		priv->irq_boff = platform_get_irq(pdev, 1);
 		if (priv->irq_boff <= 0) {
 			err = -ENODEV;
@@ -2218,7 +2218,7 @@ static int flexcan_probe(struct platform_device *pdev)
 		}
 	}
 
-	if (priv->devtype_data->quirks & FLEXCAN_QUIRK_SUPPORT_FD) {
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_SUPPORT_FD) {
 		priv->can.ctrlmode_supported |= CAN_CTRLMODE_FD |
 			CAN_CTRLMODE_FD_NON_ISO;
 		priv->can.bittiming_const = &flexcan_fd_bittiming_const;
Signature · 1 line
2.34.1
author maintainer
#737 lore raw

PATCH 730/1039 can: flexcan: rename RX modes

In reply to Greg Kroah-Hartman

From: Marc Kleine-Budde <[email protected]>

[ Upstream commit 34ea4e1c99f1f177f87e4ae7896caef238dd741a ]

Most flexcan IP cores support 2 RX modes:
- FIFO
- mailbox

The names for these modes were chosen to reflect the name of the rx-offload mode they are using.

The name of the RX modes should better reflect their difference with regards the flexcan IP core. So this patch renames the various occurrences of OFF_FIFO to RX_FIFO and OFF_TIMESTAMP to RX_MAILBOX:

| FLEXCAN_TX_MB_RESERVED_OFF_FIFO -> FLEXCAN_TX_MB_RESERVED_RX_FIFO
| FLEXCAN_TX_MB_RESERVED_OFF_TIMESTAMP -> FLEXCAN_TX_MB_RESERVED_RX_MAILBOX
| FLEXCAN_QUIRK_USE_OFF_TIMESTAMP -> FLEXCAN_QUIRK_USE_RX_MAILBOX
---
 drivers/net/can/flexcan.c | 48 +++++++++++++++++++--------------------
 1 file changed, 24 insertions(+), 24 deletions(-)
drivers/net/can/flexcan.c +24 -24
index 26bf0a0a72f1d..ba52e70d6a166 100644
@@ -173,9 +173,9 @@
 
 /* FLEXCAN interrupt flag register (IFLAG) bits */
 /* Errata ERR005829 step7: Reserve first valid MB */
-#define FLEXCAN_TX_MB_RESERVED_OFF_FIFO		8
-#define FLEXCAN_TX_MB_RESERVED_OFF_TIMESTAMP	0
-#define FLEXCAN_RX_MB_OFF_TIMESTAMP_FIRST	(FLEXCAN_TX_MB_RESERVED_OFF_TIMESTAMP + 1)
+#define FLEXCAN_TX_MB_RESERVED_RX_FIFO	8
+#define FLEXCAN_TX_MB_RESERVED_RX_MAILBOX	0
+#define FLEXCAN_RX_MB_RX_MAILBOX_FIRST	(FLEXCAN_TX_MB_RESERVED_RX_MAILBOX + 1)
 #define FLEXCAN_IFLAG_MB(x)		BIT_ULL(x)
 #define FLEXCAN_IFLAG_RX_FIFO_OVERFLOW	BIT(7)
 #define FLEXCAN_IFLAG_RX_FIFO_WARN	BIT(6)
@@ -234,8 +234,8 @@
 #define FLEXCAN_QUIRK_ENABLE_EACEN_RRS  BIT(3)
 /* Disable non-correctable errors interrupt and freeze mode */
 #define FLEXCAN_QUIRK_DISABLE_MECR BIT(4)
-/* Use timestamp based offloading */
-#define FLEXCAN_QUIRK_USE_OFF_TIMESTAMP BIT(5)
+/* Use mailboxes (not FIFO) for RX path */
+#define FLEXCAN_QUIRK_USE_RX_MAILBOX BIT(5)
 /* No interrupt for error passive */
 #define FLEXCAN_QUIRK_BROKEN_PERR_STATE BIT(6)
 /* default to BE register access */
@@ -406,38 +406,38 @@ static const struct flexcan_devtype_data fsl_imx28_devtype_data = {
 
 static const struct flexcan_devtype_data fsl_imx6q_devtype_data = {
 	.quirks = FLEXCAN_QUIRK_DISABLE_RXFG | FLEXCAN_QUIRK_ENABLE_EACEN_RRS |
-		FLEXCAN_QUIRK_USE_OFF_TIMESTAMP | FLEXCAN_QUIRK_BROKEN_PERR_STATE |
+		FLEXCAN_QUIRK_USE_RX_MAILBOX | FLEXCAN_QUIRK_BROKEN_PERR_STATE |
 		FLEXCAN_QUIRK_SETUP_STOP_MODE_GPR,
 };
 
 static const struct flexcan_devtype_data fsl_imx8qm_devtype_data = {
 	.quirks = FLEXCAN_QUIRK_DISABLE_RXFG | FLEXCAN_QUIRK_ENABLE_EACEN_RRS |
-		FLEXCAN_QUIRK_USE_OFF_TIMESTAMP | FLEXCAN_QUIRK_BROKEN_PERR_STATE |
+		FLEXCAN_QUIRK_USE_RX_MAILBOX | FLEXCAN_QUIRK_BROKEN_PERR_STATE |
 		FLEXCAN_QUIRK_SUPPORT_FD | FLEXCAN_QUIRK_SETUP_STOP_MODE_SCFW,
 };
 
 static struct flexcan_devtype_data fsl_imx8mp_devtype_data = {
 	.quirks = FLEXCAN_QUIRK_DISABLE_RXFG | FLEXCAN_QUIRK_ENABLE_EACEN_RRS |
-		FLEXCAN_QUIRK_DISABLE_MECR | FLEXCAN_QUIRK_USE_OFF_TIMESTAMP |
+		FLEXCAN_QUIRK_DISABLE_MECR | FLEXCAN_QUIRK_USE_RX_MAILBOX |
 		FLEXCAN_QUIRK_BROKEN_PERR_STATE | FLEXCAN_QUIRK_SETUP_STOP_MODE_GPR |
 		FLEXCAN_QUIRK_SUPPORT_FD | FLEXCAN_QUIRK_SUPPORT_ECC,
 };
 
 static const struct flexcan_devtype_data fsl_vf610_devtype_data = {
 	.quirks = FLEXCAN_QUIRK_DISABLE_RXFG | FLEXCAN_QUIRK_ENABLE_EACEN_RRS |
-		FLEXCAN_QUIRK_DISABLE_MECR | FLEXCAN_QUIRK_USE_OFF_TIMESTAMP |
+		FLEXCAN_QUIRK_DISABLE_MECR | FLEXCAN_QUIRK_USE_RX_MAILBOX |
 		FLEXCAN_QUIRK_BROKEN_PERR_STATE | FLEXCAN_QUIRK_SUPPORT_ECC,
 };
 
 static const struct flexcan_devtype_data fsl_ls1021a_r2_devtype_data = {
 	.quirks = FLEXCAN_QUIRK_DISABLE_RXFG | FLEXCAN_QUIRK_ENABLE_EACEN_RRS |
-		FLEXCAN_QUIRK_BROKEN_PERR_STATE | FLEXCAN_QUIRK_USE_OFF_TIMESTAMP,
+		FLEXCAN_QUIRK_BROKEN_PERR_STATE | FLEXCAN_QUIRK_USE_RX_MAILBOX,
 };
 
 static const struct flexcan_devtype_data fsl_lx2160a_r1_devtype_data = {
 	.quirks = FLEXCAN_QUIRK_DISABLE_RXFG | FLEXCAN_QUIRK_ENABLE_EACEN_RRS |
 		FLEXCAN_QUIRK_DISABLE_MECR | FLEXCAN_QUIRK_BROKEN_PERR_STATE |
-		FLEXCAN_QUIRK_USE_OFF_TIMESTAMP | FLEXCAN_QUIRK_SUPPORT_FD |
+		FLEXCAN_QUIRK_USE_RX_MAILBOX | FLEXCAN_QUIRK_SUPPORT_FD |
 		FLEXCAN_QUIRK_SUPPORT_ECC,
 };
 
@@ -1022,7 +1022,7 @@ static struct sk_buff *flexcan_mailbox_read(struct can_rx_offload *offload,
 
 	mb = flexcan_get_mb(priv, n);
 
-	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_USE_OFF_TIMESTAMP) {
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_USE_RX_MAILBOX) {
 		u32 code;
 
 		do {
@@ -1087,7 +1087,7 @@ static struct sk_buff *flexcan_mailbox_read(struct can_rx_offload *offload,
 	}
 
  mark_as_read:
-	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_USE_OFF_TIMESTAMP)
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_USE_RX_MAILBOX)
 		flexcan_write64(priv, FLEXCAN_IFLAG_MB(n), &regs->iflag1);
 	else
 		priv->write(FLEXCAN_IFLAG_RX_FIFO_AVAILABLE, &regs->iflag1);
@@ -1113,7 +1113,7 @@ static irqreturn_t flexcan_irq(int irq, void *dev_id)
 	enum can_state last_state = priv->can.state;
 
 	/* reception interrupt */
-	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_USE_OFF_TIMESTAMP) {
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_USE_RX_MAILBOX) {
 		u64 reg_iflag_rx;
 		int ret;
 
@@ -1429,20 +1429,20 @@ static int flexcan_rx_offload_setup(struct net_device *dev)
 		priv->mb_count = (sizeof(priv->regs->mb[0]) / priv->mb_size) +
 				 (sizeof(priv->regs->mb[1]) / priv->mb_size);
 
-	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_USE_OFF_TIMESTAMP)
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_USE_RX_MAILBOX)
 		priv->tx_mb_reserved =
-			flexcan_get_mb(priv, FLEXCAN_TX_MB_RESERVED_OFF_TIMESTAMP);
+			flexcan_get_mb(priv, FLEXCAN_TX_MB_RESERVED_RX_MAILBOX);
 	else
 		priv->tx_mb_reserved =
-			flexcan_get_mb(priv, FLEXCAN_TX_MB_RESERVED_OFF_FIFO);
+			flexcan_get_mb(priv, FLEXCAN_TX_MB_RESERVED_RX_FIFO);
 	priv->tx_mb_idx = priv->mb_count - 1;
 	priv->tx_mb = flexcan_get_mb(priv, priv->tx_mb_idx);
 	priv->tx_mask = FLEXCAN_IFLAG_MB(priv->tx_mb_idx);
 
 	priv->offload.mailbox_read = flexcan_mailbox_read;
 
-	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_USE_OFF_TIMESTAMP) {
-		priv->offload.mb_first = FLEXCAN_RX_MB_OFF_TIMESTAMP_FIRST;
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_USE_RX_MAILBOX) {
+		priv->offload.mb_first = FLEXCAN_RX_MB_RX_MAILBOX_FIRST;
 		priv->offload.mb_last = priv->mb_count - 2;
 
 		priv->rx_mask = GENMASK_ULL(priv->offload.mb_last,
@@ -1532,10 +1532,10 @@ static int flexcan_chip_start(struct net_device *dev)
 	/* MCR
 	 *
 	 * FIFO:
-	 * - disable for timestamp mode
+	 * - disable for mailbox mode
 	 * - enable for FIFO mode
 	 */
-	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_USE_OFF_TIMESTAMP)
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_USE_RX_MAILBOX)
 		reg_mcr &= ~FLEXCAN_MCR_FEN;
 	else
 		reg_mcr |= FLEXCAN_MCR_FEN;
@@ -1631,7 +1631,7 @@ static int flexcan_chip_start(struct net_device *dev)
 		priv->write(reg_fdctrl, &regs->fdctrl);
 	}
 
-	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_USE_OFF_TIMESTAMP) {
+	if (priv->devtype_data.quirks & FLEXCAN_QUIRK_USE_RX_MAILBOX) {
 		for (i = priv->offload.mb_first; i <= priv->offload.mb_last; i++) {
 			mb = flexcan_get_mb(priv, i);
 			priv->write(FLEXCAN_MB_CODE_RX_EMPTY,
@@ -1639,7 +1639,7 @@ static int flexcan_chip_start(struct net_device *dev)
 		}
 	} else {
 		/* clear and invalidate unused mailboxes first */
-		for (i = FLEXCAN_TX_MB_RESERVED_OFF_FIFO; i < priv->mb_count; i++) {
+		for (i = FLEXCAN_TX_MB_RESERVED_RX_FIFO; i < priv->mb_count; i++) {
 			mb = flexcan_get_mb(priv, i);
 			priv->write(FLEXCAN_MB_CODE_RX_INACTIVE,
 				    &mb->can_ctrl);
@@ -2164,7 +2164,7 @@ static int flexcan_probe(struct platform_device *pdev)
 		return -ENODEV;
 
 	if ((devtype_data->quirks & FLEXCAN_QUIRK_SUPPORT_FD) &&
-	    !(devtype_data->quirks & FLEXCAN_QUIRK_USE_OFF_TIMESTAMP)) {
+	    !(devtype_data->quirks & FLEXCAN_QUIRK_USE_RX_MAILBOX)) {
 		dev_err(&pdev->dev, "CAN-FD mode doesn't work with FIFO mode!\n");
 		return -EINVAL;
 	}
Signature · 1 line
2.34.1
author maintainer
#738 lore raw

PATCH 731/1039 can: flexcan: add more quirks to describe RX path capabilities

In reply to Greg Kroah-Hartman

From: Marc Kleine-Budde <[email protected]>

[ Upstream commit c5c88591040ee7d84d037328eed9019d3ffab821 ]

Most flexcan IP cores support 2 RX modes:
- FIFO
- mailbox

Some IP core versions cannot receive CAN RTR messages via mailboxes. This patch adds quirks to document this.

This information will be used in a later patch to switch from FIFO to more performant mailbox mode at the expense of losing the ability to receive RTR messages. This trade off is beneficial in certain use cases.

---
 drivers/net/can/flexcan.c | 66 ++++++++++++++++++++++++++++++++-------
 1 file changed, 54 insertions(+), 12 deletions(-)
drivers/net/can/flexcan.c +54 -12
index ba52e70d6a166..57cebc4156616 100644
@@ -252,6 +252,12 @@
 #define FLEXCAN_QUIRK_NR_IRQ_3 BIT(12)
 /* Setup 16 mailboxes */
 #define FLEXCAN_QUIRK_NR_MB_16 BIT(13)
+/* Device supports RX via mailboxes */
+#define FLEXCAN_QUIRK_SUPPPORT_RX_MAILBOX BIT(14)
+/* Device supports RTR reception via mailboxes */
+#define FLEXCAN_QUIRK_SUPPPORT_RX_MAILBOX_RTR BIT(15)
+/* Device supports RX via FIFO */
+#define FLEXCAN_QUIRK_SUPPPORT_RX_FIFO BIT(16)
 
 /* Structure of the message buffer */
 struct flexcan_mb {
@@ -386,59 +392,78 @@ struct flexcan_priv {
 
 static const struct flexcan_devtype_data fsl_mcf5441x_devtype_data = {
 	.quirks = FLEXCAN_QUIRK_BROKEN_PERR_STATE |
-		FLEXCAN_QUIRK_NR_IRQ_3 | FLEXCAN_QUIRK_NR_MB_16,
+		FLEXCAN_QUIRK_NR_IRQ_3 | FLEXCAN_QUIRK_NR_MB_16 |
+		FLEXCAN_QUIRK_SUPPPORT_RX_FIFO,
 };
 
 static const struct flexcan_devtype_data fsl_p1010_devtype_data = {
 	.quirks = FLEXCAN_QUIRK_BROKEN_WERR_STATE |
 		FLEXCAN_QUIRK_BROKEN_PERR_STATE |
-		FLEXCAN_QUIRK_DEFAULT_BIG_ENDIAN,
+		FLEXCAN_QUIRK_DEFAULT_BIG_ENDIAN |
+		FLEXCAN_QUIRK_SUPPPORT_RX_MAILBOX |
+		FLEXCAN_QUIRK_SUPPPORT_RX_FIFO,
 };
 
 static const struct flexcan_devtype_data fsl_imx25_devtype_data = {
 	.quirks = FLEXCAN_QUIRK_BROKEN_WERR_STATE |
-		FLEXCAN_QUIRK_BROKEN_PERR_STATE,
+		FLEXCAN_QUIRK_BROKEN_PERR_STATE |
+		FLEXCAN_QUIRK_SUPPPORT_RX_MAILBOX |
+		FLEXCAN_QUIRK_SUPPPORT_RX_FIFO,
 };
 
 static const struct flexcan_devtype_data fsl_imx28_devtype_data = {
-	.quirks = FLEXCAN_QUIRK_BROKEN_PERR_STATE,
+	.quirks = FLEXCAN_QUIRK_BROKEN_PERR_STATE |
+		FLEXCAN_QUIRK_SUPPPORT_RX_MAILBOX |
+		FLEXCAN_QUIRK_SUPPPORT_RX_FIFO,
 };
 
 static const struct flexcan_devtype_data fsl_imx6q_devtype_data = {
 	.quirks = FLEXCAN_QUIRK_DISABLE_RXFG | FLEXCAN_QUIRK_ENABLE_EACEN_RRS |
 		FLEXCAN_QUIRK_USE_RX_MAILBOX | FLEXCAN_QUIRK_BROKEN_PERR_STATE |
-		FLEXCAN_QUIRK_SETUP_STOP_MODE_GPR,
+		FLEXCAN_QUIRK_SETUP_STOP_MODE_GPR |
+		FLEXCAN_QUIRK_SUPPPORT_RX_MAILBOX |
+		FLEXCAN_QUIRK_SUPPPORT_RX_MAILBOX_RTR,
 };
 
 static const struct flexcan_devtype_data fsl_imx8qm_devtype_data = {
 	.quirks = FLEXCAN_QUIRK_DISABLE_RXFG | FLEXCAN_QUIRK_ENABLE_EACEN_RRS |
 		FLEXCAN_QUIRK_USE_RX_MAILBOX | FLEXCAN_QUIRK_BROKEN_PERR_STATE |
-		FLEXCAN_QUIRK_SUPPORT_FD | FLEXCAN_QUIRK_SETUP_STOP_MODE_SCFW,
+		FLEXCAN_QUIRK_SUPPORT_FD | FLEXCAN_QUIRK_SETUP_STOP_MODE_SCFW |
+		FLEXCAN_QUIRK_SUPPPORT_RX_MAILBOX |
+		FLEXCAN_QUIRK_SUPPPORT_RX_MAILBOX_RTR,
 };
 
 static struct flexcan_devtype_data fsl_imx8mp_devtype_data = {
 	.quirks = FLEXCAN_QUIRK_DISABLE_RXFG | FLEXCAN_QUIRK_ENABLE_EACEN_RRS |
 		FLEXCAN_QUIRK_DISABLE_MECR | FLEXCAN_QUIRK_USE_RX_MAILBOX |
 		FLEXCAN_QUIRK_BROKEN_PERR_STATE | FLEXCAN_QUIRK_SETUP_STOP_MODE_GPR |
-		FLEXCAN_QUIRK_SUPPORT_FD | FLEXCAN_QUIRK_SUPPORT_ECC,
+		FLEXCAN_QUIRK_SUPPORT_FD | FLEXCAN_QUIRK_SUPPORT_ECC |
+		FLEXCAN_QUIRK_SUPPPORT_RX_MAILBOX |
+		FLEXCAN_QUIRK_SUPPPORT_RX_MAILBOX_RTR,
 };
 
 static const struct flexcan_devtype_data fsl_vf610_devtype_data = {
 	.quirks = FLEXCAN_QUIRK_DISABLE_RXFG | FLEXCAN_QUIRK_ENABLE_EACEN_RRS |
 		FLEXCAN_QUIRK_DISABLE_MECR | FLEXCAN_QUIRK_USE_RX_MAILBOX |
-		FLEXCAN_QUIRK_BROKEN_PERR_STATE | FLEXCAN_QUIRK_SUPPORT_ECC,
+		FLEXCAN_QUIRK_BROKEN_PERR_STATE | FLEXCAN_QUIRK_SUPPORT_ECC |
+		FLEXCAN_QUIRK_SUPPPORT_RX_MAILBOX |
+		FLEXCAN_QUIRK_SUPPPORT_RX_MAILBOX_RTR,
 };
 
 static const struct flexcan_devtype_data fsl_ls1021a_r2_devtype_data = {
 	.quirks = FLEXCAN_QUIRK_DISABLE_RXFG | FLEXCAN_QUIRK_ENABLE_EACEN_RRS |
-		FLEXCAN_QUIRK_BROKEN_PERR_STATE | FLEXCAN_QUIRK_USE_RX_MAILBOX,
+		FLEXCAN_QUIRK_BROKEN_PERR_STATE | FLEXCAN_QUIRK_USE_RX_MAILBOX |
+		FLEXCAN_QUIRK_SUPPPORT_RX_MAILBOX |
+		FLEXCAN_QUIRK_SUPPPORT_RX_MAILBOX_RTR,
 };
 
 static const struct flexcan_devtype_data fsl_lx2160a_r1_devtype_data = {
 	.quirks = FLEXCAN_QUIRK_DISABLE_RXFG | FLEXCAN_QUIRK_ENABLE_EACEN_RRS |
 		FLEXCAN_QUIRK_DISABLE_MECR | FLEXCAN_QUIRK_BROKEN_PERR_STATE |
 		FLEXCAN_QUIRK_USE_RX_MAILBOX | FLEXCAN_QUIRK_SUPPORT_FD |
-		FLEXCAN_QUIRK_SUPPORT_ECC,
+		FLEXCAN_QUIRK_SUPPORT_ECC |
+		FLEXCAN_QUIRK_SUPPPORT_RX_MAILBOX |
+		FLEXCAN_QUIRK_SUPPPORT_RX_MAILBOX_RTR,
 };
 
 static const struct can_bittiming_const flexcan_bittiming_const = {
@@ -2164,8 +2189,25 @@ static int flexcan_probe(struct platform_device *pdev)
 		return -ENODEV;
 
 	if ((devtype_data->quirks & FLEXCAN_QUIRK_SUPPORT_FD) &&
-	    !(devtype_data->quirks & FLEXCAN_QUIRK_USE_RX_MAILBOX)) {
-		dev_err(&pdev->dev, "CAN-FD mode doesn't work with FIFO mode!\n");
+	    !((devtype_data->quirks &
+	       (FLEXCAN_QUIRK_USE_RX_MAILBOX |
+		FLEXCAN_QUIRK_SUPPPORT_RX_MAILBOX |
+		FLEXCAN_QUIRK_SUPPPORT_RX_MAILBOX_RTR |
+		FLEXCAN_QUIRK_SUPPPORT_RX_FIFO)) ==
+	      (FLEXCAN_QUIRK_USE_RX_MAILBOX |
+	       FLEXCAN_QUIRK_SUPPPORT_RX_MAILBOX |
+	       FLEXCAN_QUIRK_SUPPPORT_RX_MAILBOX_RTR))) {
+		dev_err(&pdev->dev, "CAN-FD mode doesn't work in RX-FIFO mode!\n");
+		return -EINVAL;
+	}
+
+	if ((devtype_data->quirks &
+	     (FLEXCAN_QUIRK_SUPPPORT_RX_MAILBOX |
+	      FLEXCAN_QUIRK_SUPPPORT_RX_MAILBOX_RTR)) ==
+	    FLEXCAN_QUIRK_SUPPPORT_RX_MAILBOX_RTR) {
+		dev_err(&pdev->dev,
+			"Quirks (0x%08x) inconsistent: RX_MAILBOX_RX supported but not RX_MAILBOX\n",
+			devtype_data->quirks);
 		return -EINVAL;
 	}
 
Signature · 1 line
2.34.1
author maintainer
#739 lore raw

PATCH 732/1039 x86/kbuild: Enable CONFIG_KALLSYMS_ALL=y in the defconfigs

In reply to Greg Kroah-Hartman

From: Ingo Molnar <[email protected]>

[ Upstream commit b6aa86cff44cf099299d3a5e66348cb709cd7964 ]

Most distro kernels have this option enabled, to improve debug output.

Lockdep also selects it.

Enable this in the defconfig kernel as well, to make it more representative of what people are using on x86.

---
 arch/x86/configs/i386_defconfig   | 1 +
 arch/x86/configs/x86_64_defconfig | 1 +
 2 files changed, 2 insertions(+)
arch/x86/configs/i386_defconfig +1 -0
index e81885384f604..99398cbdae434 100644
@@ -262,3 +262,4 @@ CONFIG_BLK_DEV_IO_TRACE=y
 CONFIG_PROVIDE_OHCI1394_DMA_INIT=y
 CONFIG_EARLY_PRINTK_DBGP=y
 CONFIG_DEBUG_BOOT_PARAMS=y
+CONFIG_KALLSYMS_ALL=y
arch/x86/configs/x86_64_defconfig +1 -0
index e8a7a0af2bdaa..d7298b104a456 100644
@@ -258,3 +258,4 @@ CONFIG_BLK_DEV_IO_TRACE=y
 CONFIG_PROVIDE_OHCI1394_DMA_INIT=y
 CONFIG_EARLY_PRINTK_DBGP=y
 CONFIG_DEBUG_BOOT_PARAMS=y
+CONFIG_KALLSYMS_ALL=y
Signature · 1 line
2.34.1
author maintainer
#740 lore raw

PATCH 733/1039 clk: samsung: exynos850: Register clocks early

In reply to Greg Kroah-Hartman

From: Sam Protsenko <[email protected]>

[ Upstream commit bcda841f9bf2cddcf2f000cba96f2e27f6f2bdbf ]

Some clocks must be registered before init calls. For example MCT clock (from CMU_PERI) is needed for MCT timer driver, which is registered with TIMER_OF_DECLARE(). By the time we get to core_initcall() used for clk-exynos850 platform driver init, it's already too late. Inability to get "mct" clock in MCT driver leads to kernel panic, as functions registered with *_OF_DECLARE() can't do deferred calls. MCT timer driver can't be fixed either, as it's acting as a clock source and it's essential to register it in start_kernel() -> time_init().

Let's register CMU_PERI clocks early, using CLK_OF_DECLARE(). CMU_TOP generates clocks needed for CMU_PERI, but it's already registered early.

While at it, let's cleanup the code a bit, by extracting everything related to CMU initialization and registration to the separate function.

Similar issue was discussed at [1] and addressed in commit 1f7db7bbf031 ("clk: renesas: cpg-mssr: Add early clock support"), as well as in drivers/clk/mediatek/clk-mt2712.c.

[1] https://patchwork.kernel.org/project/linux-renesas-soc/patch/[email protected]/

---
 drivers/clk/samsung/clk-exynos850.c | 70 ++++++++++++++++++++---------
 1 file changed, 49 insertions(+), 21 deletions(-)
drivers/clk/samsung/clk-exynos850.c +49 -21
index 2294989e244c5..79cce8ba88831 100644
@@ -60,6 +60,43 @@ static void __init exynos850_init_clocks(struct device_node *np,
 	iounmap(reg_base);
 }
 
+/**
+ * exynos850_register_cmu - Register specified Exynos850 CMU domain
+ * @dev:	Device object; may be NULL if this function is not being
+ *		called from platform driver probe function
+ * @np:		CMU device tree node
+ * @cmu:	CMU data
+ *
+ * Register specified CMU domain, which includes next steps:
+ *
+ * 1. Enable parent clock of @cmu CMU
+ * 2. Set initial registers configuration for @cmu CMU clocks
+ * 3. Register @cmu CMU clocks using Samsung clock framework API
+ */
+static void __init exynos850_register_cmu(struct device *dev,
+		struct device_node *np, const struct samsung_cmu_info *cmu)
+{
+	/* Keep CMU parent clock running (needed for CMU registers access) */
+	if (cmu->clk_name) {
+		struct clk *parent_clk;
+
+		if (dev)
+			parent_clk = clk_get(dev, cmu->clk_name);
+		else
+			parent_clk = of_clk_get_by_name(np, cmu->clk_name);
+
+		if (IS_ERR(parent_clk)) {
+			pr_err("%s: could not find bus clock %s; err = %ld\n",
+			       __func__, cmu->clk_name, PTR_ERR(parent_clk));
+		} else {
+			clk_prepare_enable(parent_clk);
+		}
+	}
+
+	exynos850_init_clocks(np, cmu->clk_regs, cmu->nr_clk_regs);
+	samsung_cmu_register_one(np, cmu);
+}
+
 /* ---- CMU_TOP ------------------------------------------------------------- */
 
 /* Register Offset definitions for CMU_TOP (0x120e0000) */
@@ -347,10 +384,10 @@ static const struct samsung_cmu_info top_cmu_info __initconst = {
 
 static void __init exynos850_cmu_top_init(struct device_node *np)
 {
-	exynos850_init_clocks(np, top_clk_regs, ARRAY_SIZE(top_clk_regs));
-	samsung_cmu_register_one(np, &top_cmu_info);
+	exynos850_register_cmu(NULL, np, &top_cmu_info);
 }
 
+/* Register CMU_TOP early, as it's a dependency for other early domains */
 CLK_OF_DECLARE(exynos850_cmu_top, "samsung,exynos850-cmu-top",
 	       exynos850_cmu_top_init);
 
@@ -615,6 +652,15 @@ static const struct samsung_cmu_info peri_cmu_info __initconst = {
 	.clk_name		= "dout_peri_bus",
 };
 
+static void __init exynos850_cmu_peri_init(struct device_node *np)
+{
+	exynos850_register_cmu(NULL, np, &peri_cmu_info);
+}
+
+/* Register CMU_PERI early, as it's needed for MCT timer */
+CLK_OF_DECLARE(exynos850_cmu_peri, "samsung,exynos850-cmu-peri",
+	       exynos850_cmu_peri_init);
+
 /* ---- CMU_CORE ------------------------------------------------------------ */
 
 /* Register Offset definitions for CMU_CORE (0x12000000) */
@@ -779,24 +825,9 @@ static int __init exynos850_cmu_probe(struct platform_device *pdev)
 {
 	const struct samsung_cmu_info *info;
 	struct device *dev = &pdev->dev;
-	struct device_node *np = dev->of_node;
 
 	info = of_device_get_match_data(dev);
-	exynos850_init_clocks(np, info->clk_regs, info->nr_clk_regs);
-	samsung_cmu_register_one(np, info);
-
-	/* Keep bus clock running, so it's possible to access CMU registers */
-	if (info->clk_name) {
-		struct clk *bus_clk;
-
-		bus_clk = clk_get(dev, info->clk_name);
-		if (IS_ERR(bus_clk)) {
-			pr_err("%s: could not find bus clock %s; err = %ld\n",
-			       __func__, info->clk_name, PTR_ERR(bus_clk));
-		} else {
-			clk_prepare_enable(bus_clk);
-		}
-	}
+	exynos850_register_cmu(dev, dev->of_node, info);
 
 	return 0;
 }
@@ -806,9 +837,6 @@ static const struct of_device_id exynos850_cmu_of_match[] = {
 	{
 		.compatible = "samsung,exynos850-cmu-hsi",
 		.data = &hsi_cmu_info,
-	}, {
-		.compatible = "samsung,exynos850-cmu-peri",
-		.data = &peri_cmu_info,
 	}, {
 		.compatible = "samsung,exynos850-cmu-core",
 		.data = &core_cmu_info,
Signature · 1 line
2.34.1
author maintainer
#741 lore raw

PATCH 734/1039 powerpc/6xx: add missing of_node_put

In reply to Greg Kroah-Hartman

From: Julia Lawall <[email protected]>

[ Upstream commit f6e82647ff71d427d4148964b71f239fba9d7937 ]

for_each_compatible_node performs an of_node_get on each iteration, so a break out of the loop requires an of_node_put.

A simplified version of the semantic patch that fixes this problem is as follows (http://coccinelle.lip6.fr):

// <smpl>
@@
expression e;
local idexpression n;
@@

@@
local idexpression n;
expression e;
@@

for_each_compatible_node(n,...) {
...
(
of_node_put(n);
|
e = n
|
+ of_node_put(n);
? break;
)
...
}
... when != n
// </smpl>

---
 arch/powerpc/platforms/embedded6xx/hlwd-pic.c | 1 +
 1 file changed, 1 insertion(+)
arch/powerpc/platforms/embedded6xx/hlwd-pic.c +1 -0
index 15396333a90bd..a4b020e4b6af0 100644
@@ -214,6 +214,7 @@ void hlwd_pic_probe(void)
 			irq_set_chained_handler(cascade_virq,
 						hlwd_pic_irq_cascade);
 			hlwd_irq_host = host;
+			of_node_put(np);
 			break;
 		}
 	}
Signature · 1 line
2.34.1
author maintainer
#742 lore raw

PATCH 735/1039 powerpc/powernv: add missing of_node_put

In reply to Greg Kroah-Hartman

From: Julia Lawall <[email protected]>

[ Upstream commit 7d405a939ca960162eb30c1475759cb2fdf38f8c ]

for_each_compatible_node performs an of_node_get on each iteration, so a break out of the loop requires an of_node_put.

A simplified version of the semantic patch that fixes this problem is as follows (http://coccinelle.lip6.fr):

// <smpl>
@@
local idexpression n;
expression e;
@@

for_each_compatible_node(n,...) {
...
(
of_node_put(n);
|
e = n
|
+ of_node_put(n);
? break;
)
...
}
... when != n
// </smpl>

---
 arch/powerpc/platforms/powernv/opal-lpc.c | 1 +
 1 file changed, 1 insertion(+)
arch/powerpc/platforms/powernv/opal-lpc.c +1 -0
index 1e5d51db40f84..5390c888db162 100644
@@ -396,6 +396,7 @@ void __init opal_lpc_init(void)
 		if (!of_get_property(np, "primary", NULL))
 			continue;
 		opal_lpc_chip_id = of_get_ibm_chip_id(np);
+		of_node_put(np);
 		break;
 	}
 	if (opal_lpc_chip_id < 0)
Signature · 1 line
2.34.1
author maintainer
#743 lore raw

PATCH 736/1039 powerpc/cell: add missing of_node_put

In reply to Greg Kroah-Hartman

From: Julia Lawall <[email protected]>

[ Upstream commit a841fd009e51c8c0a8f07c942e9ab6bb48da8858 ]

for_each_node_by_name performs an of_node_get on each iteration, so a break out of the loop requires an of_node_put.

A simplified version of the semantic patch that fixes this problem is as follows (http://coccinelle.lip6.fr):

// <smpl>
@@
expression e,e1;
local idexpression n;
@@

for_each_node_by_name(n, e1) {
... when != of_node_put(n)

       when != e = n

(
return n;
|
+ of_node_put(n);
? return ...;
)
...
}
// </smpl>

---
 arch/powerpc/platforms/cell/iommu.c | 1 +
 1 file changed, 1 insertion(+)
arch/powerpc/platforms/cell/iommu.c +1 -0
index fa08699aedeb8..d32f24de84798 100644
@@ -977,6 +977,7 @@ static int __init cell_iommu_fixed_mapping_init(void)
 			if (hbase < dbase || (hend > (dbase + dsize))) {
 				pr_debug("iommu: hash window doesn't fit in"
 					 "real DMA window\n");
+				of_node_put(np);
 				return -1;
 			}
 		}
Signature · 1 line
2.34.1
author maintainer
#744 lore raw

PATCH 737/1039 powerpc/btext: add missing of_node_put

In reply to Greg Kroah-Hartman

From: Julia Lawall <[email protected]>

[ Upstream commit a1d2b210ffa52d60acabbf7b6af3ef7e1e69cda0 ]

for_each_node_by_type performs an of_node_get on each iteration, so a break out of the loop requires an of_node_put.

A simplified version of the semantic patch that fixes this problem is as follows (http://coccinelle.lip6.fr):

// <smpl>
@@
local idexpression n;
expression e;
@@

for_each_node_by_type(n,...) {
...
(
of_node_put(n);
|
e = n
|
+ of_node_put(n);
? break;
)
...
}
... when != n
// </smpl>

---
 arch/powerpc/kernel/btext.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
arch/powerpc/kernel/btext.c +3 -1
index 803c2a45b22ac..1cffb5e7c38d6 100644
@@ -241,8 +241,10 @@ int __init btext_find_display(int allow_nonstdout)
 			rc = btext_initialize(np);
 			printk("result: %d\n", rc);
 		}
-		if (rc == 0)
+		if (rc == 0) {
+			of_node_put(np);
 			break;
+		}
 	}
 	return rc;
 }
Signature · 1 line
2.34.1
author maintainer
#745 lore raw

PATCH 738/1039 powerpc/watchdog: Fix missed watchdog reset due to memory ordering race

In reply to Greg Kroah-Hartman

From: Nicholas Piggin <[email protected]>

[ Upstream commit 5dad4ba68a2483fc80d70b9dc90bbe16e1f27263 ]

It is possible for all CPUs to miss the pending cpumask becoming clear, and then nobody resetting it, which will cause the lockup detector to stop working. It will eventually expire, but watchdog_smp_panic will avoid doing anything if the pending mask is clear and it will never be reset.

Order the cpumask clear vs the subsequent test to close this race.

Add an extra check for an empty pending mask when the watchdog fires and finds its bit still clear, to try to catch any other possible races or bugs here and keep the watchdog working. The extra test in arch_touch_nmi_watchdog is required to prevent the new warning from firing off.

---
 arch/powerpc/kernel/watchdog.c | 41 +++++++++++++++++++++++++++++++++-
 1 file changed, 40 insertions(+), 1 deletion(-)
arch/powerpc/kernel/watchdog.c +40 -1
index 3fa6d240bade2..ad94a2c6b7337 100644
@@ -135,6 +135,10 @@ static void set_cpumask_stuck(const struct cpumask *cpumask, u64 tb)
 {
 	cpumask_or(&wd_smp_cpus_stuck, &wd_smp_cpus_stuck, cpumask);
 	cpumask_andnot(&wd_smp_cpus_pending, &wd_smp_cpus_pending, cpumask);
+	/*
+	 * See wd_smp_clear_cpu_pending()
+	 */
+	smp_mb();
 	if (cpumask_empty(&wd_smp_cpus_pending)) {
 		wd_smp_last_reset_tb = tb;
 		cpumask_andnot(&wd_smp_cpus_pending,
@@ -221,13 +225,44 @@ static void wd_smp_clear_cpu_pending(int cpu, u64 tb)
 
 			cpumask_clear_cpu(cpu, &wd_smp_cpus_stuck);
 			wd_smp_unlock(&flags);
+		} else {
+			/*
+			 * The last CPU to clear pending should have reset the
+			 * watchdog so we generally should not find it empty
+			 * here if our CPU was clear. However it could happen
+			 * due to a rare race with another CPU taking the
+			 * last CPU out of the mask concurrently.
+			 *
+			 * We can't add a warning for it. But just in case
+			 * there is a problem with the watchdog that is causing
+			 * the mask to not be reset, try to kick it along here.
+			 */
+			if (unlikely(cpumask_empty(&wd_smp_cpus_pending)))
+				goto none_pending;
 		}
 		return;
 	}
+
 	cpumask_clear_cpu(cpu, &wd_smp_cpus_pending);
+
+	/*
+	 * Order the store to clear pending with the load(s) to check all
+	 * words in the pending mask to check they are all empty. This orders
+	 * with the same barrier on another CPU. This prevents two CPUs
+	 * clearing the last 2 pending bits, but neither seeing the other's
+	 * store when checking if the mask is empty, and missing an empty
+	 * mask, which ends with a false positive.
+	 */
+	smp_mb();
 	if (cpumask_empty(&wd_smp_cpus_pending)) {
 		unsigned long flags;
 
+none_pending:
+		/*
+		 * Double check under lock because more than one CPU could see
+		 * a clear mask with the lockless check after clearing their
+		 * pending bits.
+		 */
 		wd_smp_lock(&flags);
 		if (cpumask_empty(&wd_smp_cpus_pending)) {
 			wd_smp_last_reset_tb = tb;
@@ -318,8 +353,12 @@ void arch_touch_nmi_watchdog(void)
 {
 	unsigned long ticks = tb_ticks_per_usec * wd_timer_period_ms * 1000;
 	int cpu = smp_processor_id();
-	u64 tb = get_tb();
+	u64 tb;
 
+	if (!cpumask_test_cpu(cpu, &watchdog_cpumask))
+		return;
+
+	tb = get_tb();
 	if (tb - per_cpu(wd_timer_tb, cpu) >= ticks) {
 		per_cpu(wd_timer_tb, cpu) = tb;
 		wd_smp_clear_cpu_pending(cpu, tb);
Signature · 1 line
2.34.1
author maintainer
#746 lore raw

PATCH 739/1039 ASoC: imx-hdmi: add put_device() after of_find_device_by_node()

In reply to Greg Kroah-Hartman

From: Ye Guojin <[email protected]>

[ Upstream commit f670b274f7f6f4b2722d7f08d0fddf606a727e92 ]

This was found by coccicheck:
./sound/soc/fsl/imx-hdmi.c,209,1-7,ERROR missing put_device; call of_find_device_by_node on line 119, but without a corresponding object release within this function.

---
 sound/soc/fsl/imx-hdmi.c | 2 ++
 1 file changed, 2 insertions(+)
sound/soc/fsl/imx-hdmi.c +2 -0
index f10359a288005..929f69b758af4 100644
@@ -145,6 +145,8 @@ static int imx_hdmi_probe(struct platform_device *pdev)
 	data->dai.capture_only = false;
 	data->dai.init = imx_hdmi_init;
 
+	put_device(&cpu_pdev->dev);
+
 	if (of_node_name_eq(cpu_np, "sai")) {
 		data->cpu_priv.sysclk_id[1] = FSL_SAI_CLK_MAST1;
 		data->cpu_priv.sysclk_id[0] = FSL_SAI_CLK_MAST1;
Signature · 1 line
2.34.1
author maintainer
#747 lore raw

PATCH 740/1039 i2c: i801: Dont silently correct invalid transfer size

In reply to Greg Kroah-Hartman

From: Heiner Kallweit <[email protected]>

[ Upstream commit effa453168a7eeb8a562ff4edc1dbf9067360a61 ]

If an invalid block size is provided, reject it instead of silently changing it to a supported value. Especially critical I see the case of a write transfer with block length 0. In this case we have no guarantee that the byte we would write is valid. When silently reducing a read to 32 bytes then we don't return an error and the caller may falsely assume that we returned the full requested data.

If this change should break any (broken) caller, then I think we should fix the caller.

---
 drivers/i2c/busses/i2c-i801.c | 15 +++++----------
 1 file changed, 5 insertions(+), 10 deletions(-)
drivers/i2c/busses/i2c-i801.c +5 -10
index 41446f9cc52da..c87ea470eba98 100644
@@ -775,6 +775,11 @@ static int i801_block_transaction(struct i801_priv *priv, union i2c_smbus_data *
 	int result = 0;
 	unsigned char hostc;
 
+	if (read_write == I2C_SMBUS_READ && command == I2C_SMBUS_BLOCK_DATA)
+		data->block[0] = I2C_SMBUS_BLOCK_MAX;
+	else if (data->block[0] < 1 || data->block[0] > I2C_SMBUS_BLOCK_MAX)
+		return -EPROTO;
+
 	if (command == I2C_SMBUS_I2C_BLOCK_DATA) {
 		if (read_write == I2C_SMBUS_WRITE) {
 			/* set I2C_EN bit in configuration register */
@@ -788,16 +793,6 @@ static int i801_block_transaction(struct i801_priv *priv, union i2c_smbus_data *
 		}
 	}
 
-	if (read_write == I2C_SMBUS_WRITE
-	 || command == I2C_SMBUS_I2C_BLOCK_DATA) {
-		if (data->block[0] < 1)
-			data->block[0] = 1;
-		if (data->block[0] > I2C_SMBUS_BLOCK_MAX)
-			data->block[0] = I2C_SMBUS_BLOCK_MAX;
-	} else {
-		data->block[0] = 32;	/* max for SMBus block reads */
-	}
-
 	/* Experience has shown that the block buffer can only be used for
 	   SMBus (not I2C) block transactions, even though the datasheet
 	   doesn't mention this limitation. */
Signature · 1 line
2.34.1
author maintainer
#748 lore raw

PATCH 741/1039 powerpc/smp: Move setup_profiling_timer() under CONFIG_PROFILING

In reply to Greg Kroah-Hartman

From: Michael Ellerman <[email protected]>

[ Upstream commit a4ac0d249a5db80e79d573db9e4ad29354b643a8 ]

setup_profiling_timer() is only needed when CONFIG_PROFILING is enabled.

Fixes the following W=1 warning when CONFIG_PROFILING=n:
linux/arch/powerpc/kernel/smp.c:1638:5: error: no previous prototype for ‘setup_profiling_timer’

---
 arch/powerpc/kernel/smp.c | 2 ++
 1 file changed, 2 insertions(+)
arch/powerpc/kernel/smp.c +2 -0
index c23ee842c4c33..aee3a7119f977 100644
@@ -1635,10 +1635,12 @@ void start_secondary(void *unused)
 	BUG();
 }
 
+#ifdef CONFIG_PROFILING
 int setup_profiling_timer(unsigned int multiplier)
 {
 	return 0;
 }
+#endif
 
 static void fixup_topology(void)
 {
Signature · 1 line
2.34.1
author maintainer
#749 lore raw

PATCH 742/1039 i2c: mpc: Correct I2C reset procedure

In reply to Greg Kroah-Hartman

From: Joakim Tjernlund <[email protected]>

[ Upstream commit ebe82cf92cd4825c3029434cabfcd2f1780e64be ]

Current I2C reset procedure is broken in two ways:
1) It only generate 1 START instead of 9 STARTs and STOP.
2) It leaves the bus Busy so every I2C xfer after the first fixup calls the reset routine again, for every xfer there after.

This fixes both errors.

---
 drivers/i2c/busses/i2c-mpc.c | 23 +++++++++++++++--------
 1 file changed, 15 insertions(+), 8 deletions(-)
drivers/i2c/busses/i2c-mpc.c +15 -8
index db26cc36e13fe..6c698c10d3cdb 100644
@@ -119,23 +119,30 @@ static inline void writeccr(struct mpc_i2c *i2c, u32 x)
 /* Sometimes 9th clock pulse isn't generated, and slave doesn't release
  * the bus, because it wants to send ACK.
  * Following sequence of enabling/disabling and sending start/stop generates
- * the 9 pulses, so it's all OK.
+ * the 9 pulses, each with a START then ending with STOP, so it's all OK.
  */
 static void mpc_i2c_fixup(struct mpc_i2c *i2c)
 {
 	int k;
-	u32 delay_val = 1000000 / i2c->real_clk + 1;
-
-	if (delay_val < 2)
-		delay_val = 2;
+	unsigned long flags;
 
 	for (k = 9; k; k--) {
 		writeccr(i2c, 0);
-		writeccr(i2c, CCR_MSTA | CCR_MTX | CCR_MEN);
+		writeb(0, i2c->base + MPC_I2C_SR); /* clear any status bits */
+		writeccr(i2c, CCR_MEN | CCR_MSTA); /* START */
+		readb(i2c->base + MPC_I2C_DR); /* init xfer */
+		udelay(15); /* let it hit the bus */
+		local_irq_save(flags); /* should not be delayed further */
+		writeccr(i2c, CCR_MEN | CCR_MSTA | CCR_RSTA); /* delay SDA */
 		readb(i2c->base + MPC_I2C_DR);
-		writeccr(i2c, CCR_MEN);
-		udelay(delay_val << 1);
+		if (k != 1)
+			udelay(5);
+		local_irq_restore(flags);
 	}
+	writeccr(i2c, CCR_MEN); /* Initiate STOP */
+	readb(i2c->base + MPC_I2C_DR);
+	udelay(15); /* Let STOP propagate */
+	writeccr(i2c, 0);
 }
 
 static int i2c_mpc_wait_sr(struct mpc_i2c *i2c, int mask)
Signature · 1 line
2.34.1
author maintainer
#750 lore raw

PATCH 743/1039 clk: meson: gxbb: Fix the SDM_EN bit for MPLL0 on GXBB

In reply to Greg Kroah-Hartman

From: Martin Blumenstingl <[email protected]>

[ Upstream commit ff54938dd190d85f740b9bf9dde59b550936b621 ]

There are reports that 48kHz audio does not work on the WeTek Play 2 (which uses a GXBB SoC), while 44.1kHz audio works fine on the same board. There are also reports of 48kHz audio working fine on GXL and GXM SoCs, which are using an (almost) identical AIU (audio controller).

Experimenting has shown that MPLL0 is causing this problem. In the .dts we have by default:

	assigned-clocks = <&clkc CLKID_MPLL0>,
			  <&clkc CLKID_MPLL1>,
			  <&clkc CLKID_MPLL2>;
	assigned-clock-rates = <294912000>,
			       <270950400>,
			       <393216000>;

The MPLL0 rate is divisible by 48kHz without remainder and the MPLL1 rate is divisible by 44.1kHz without remainder. Swapping these two clock rates "fixes" 48kHz audio but breaks 44.1kHz audio.

Everything looks normal when looking at the info provided by the common clock framework while playing 48kHz audio (via I2S with mclk-fs = 256):

        mpll_prediv                 1        1        0  2000000000
           mpll0_div                1        1        0   294909641
              mpll0                 1        1        0   294909641
                 cts_amclk_sel       1        1        0   294909641
                    cts_amclk_div       1        1        0    12287902
                       cts_amclk       1        1        0    12287902

meson-clk-msr however shows that the actual MPLL0 clock is off by more than 38MHz:

        mp0_out               333322917    +/-10416Hz

The rate seen by meson-clk-msr is very close to what we would get when SDM (the fractional part) was ignored:
(2000000000Hz * 16384) / ((16384 * 6) = 333.33MHz
If SDM was considered the we should get close to:
(2000000000Hz * 16384) / ((16384 * 6) + 12808) = 294.9MHz

Further experimenting shows that HHI_MPLL_CNTL7[15] does not have any effect on the rate of MPLL0 as seen my meson-clk-msr (regardless of whether that bit is zero or one the rate is always the same according to meson-clk-msr). Using HHI_MPLL_CNTL[25] on the other hand as SDM_EN results in SDM being considered for the rate output by the hardware. The rate - as seen by meson-clk-msr - matches with what we expect when SDM_EN is enabled (fractional part is being considered, resulting in a 294.9MHz output) or disable (fractional part being ignored, resulting in a 333.33MHz output).

---
 drivers/clk/meson/gxbb.c | 44 +++++++++++++++++++++++++++++++++++++---
 1 file changed, 41 insertions(+), 3 deletions(-)
drivers/clk/meson/gxbb.c +41 -3
index d6eed760327d0..608e0e8ca49a8 100644
@@ -713,6 +713,35 @@ static struct clk_regmap gxbb_mpll_prediv = {
 };
 
 static struct clk_regmap gxbb_mpll0_div = {
+	.data = &(struct meson_clk_mpll_data){
+		.sdm = {
+			.reg_off = HHI_MPLL_CNTL7,
+			.shift   = 0,
+			.width   = 14,
+		},
+		.sdm_en = {
+			.reg_off = HHI_MPLL_CNTL,
+			.shift   = 25,
+			.width	 = 1,
+		},
+		.n2 = {
+			.reg_off = HHI_MPLL_CNTL7,
+			.shift   = 16,
+			.width   = 9,
+		},
+		.lock = &meson_clk_lock,
+	},
+	.hw.init = &(struct clk_init_data){
+		.name = "mpll0_div",
+		.ops = &meson_clk_mpll_ops,
+		.parent_hws = (const struct clk_hw *[]) {
+			&gxbb_mpll_prediv.hw
+		},
+		.num_parents = 1,
+	},
+};
+
+static struct clk_regmap gxl_mpll0_div = {
 	.data = &(struct meson_clk_mpll_data){
 		.sdm = {
 			.reg_off = HHI_MPLL_CNTL7,
@@ -749,7 +778,16 @@ static struct clk_regmap gxbb_mpll0 = {
 	.hw.init = &(struct clk_init_data){
 		.name = "mpll0",
 		.ops = &clk_regmap_gate_ops,
-		.parent_hws = (const struct clk_hw *[]) { &gxbb_mpll0_div.hw },
+		.parent_data = &(const struct clk_parent_data) {
+			/*
+			 * Note:
+			 * GXL and GXBB have different SDM_EN registers. We
+			 * fallback to the global naming string mechanism so
+			 * mpll0_div picks up the appropriate one.
+			 */
+			.name = "mpll0_div",
+			.index = -1,
+		},
 		.num_parents = 1,
 		.flags = CLK_SET_RATE_PARENT,
 	},
@@ -3044,7 +3082,7 @@ static struct clk_hw_onecell_data gxl_hw_onecell_data = {
 		[CLKID_VAPB_1]		    = &gxbb_vapb_1.hw,
 		[CLKID_VAPB_SEL]	    = &gxbb_vapb_sel.hw,
 		[CLKID_VAPB]		    = &gxbb_vapb.hw,
-		[CLKID_MPLL0_DIV]	    = &gxbb_mpll0_div.hw,
+		[CLKID_MPLL0_DIV]	    = &gxl_mpll0_div.hw,
 		[CLKID_MPLL1_DIV]	    = &gxbb_mpll1_div.hw,
 		[CLKID_MPLL2_DIV]	    = &gxbb_mpll2_div.hw,
 		[CLKID_MPLL_PREDIV]	    = &gxbb_mpll_prediv.hw,
@@ -3439,7 +3477,7 @@ static struct clk_regmap *const gxl_clk_regmaps[] = {
 	&gxbb_mpll0,
 	&gxbb_mpll1,
 	&gxbb_mpll2,
-	&gxbb_mpll0_div,
+	&gxl_mpll0_div,
 	&gxbb_mpll1_div,
 	&gxbb_mpll2_div,
 	&gxbb_cts_amclk_div,
Signature · 1 line
2.34.1
author maintainer
#751 lore raw

PATCH 744/1039 powerpc/powermac: Add missing lockdep_register_key()

In reply to Greg Kroah-Hartman

From: Christophe Leroy <[email protected]>

[ Upstream commit df1f679d19edb9eeb67cc2f96b29375f21991945 ]

KeyWest i2c @0xf8001003 irq 42 /uni-n@f8000000/i2c@f8001000

BUG: key c2d00cbc has not been registered!
------------[ cut here ]------------
DEBUG_LOCKS_WARN_ON(1)
WARNING: CPU: 0 PID: 1 at kernel/locking/lockdep.c:4801 lockdep_init_map_type+0x4c0/0xb4c
Modules linked in:
CPU: 0 PID: 1 Comm: swapper/0 Not tainted 5.15.5-gentoo-PowerMacG4 #9
NIP:  c01a9428 LR: c01a9428 CTR: 00000000
REGS: e1033cf0 TRAP: 0700   Not tainted  (5.15.5-gentoo-PowerMacG4)
MSR:  00029032 <EE,ME,IR,DR,RI>  CR: 24002002  XER: 00000000

GPR00: c01a9428 e1033db0 c2d1cf20 00000016 00000004 00000001 c01c0630 e1033a73
GPR08: 00000000 00000000 00000000 e1033db0 24002004 00000000 f8729377 00000003
GPR16: c1829a9c 00000000 18305357 c1416fc0 c1416f80 c006ac60 c2d00ca8 c1416f00
GPR24: 00000000 c21586f0 c2160000 00000000 c2d00cbc c2170000 c216e1a0 c2160000 NIP [c01a9428] lockdep_init_map_type+0x4c0/0xb4c
LR [c01a9428] lockdep_init_map_type+0x4c0/0xb4c

Call Trace:
[e1033db0] [c01a9428] lockdep_init_map_type+0x4c0/0xb4c (unreliable)
[e1033df0] [c1c177b8] kw_i2c_add+0x334/0x424
[e1033e20] [c1c18294] pmac_i2c_init+0x9ec/0xa9c
[e1033e80] [c1c1a790] smp_core99_probe+0xbc/0x35c
[e1033eb0] [c1c03cb0] kernel_init_freeable+0x190/0x5a4
[e1033f10] [c000946c] kernel_init+0x28/0x154
[e1033f30] [c0035148] ret_from_kernel_thread+0x14/0x1c

Add missing lockdep_register_key()

---
 arch/powerpc/platforms/powermac/low_i2c.c | 1 +
 1 file changed, 1 insertion(+)
arch/powerpc/platforms/powermac/low_i2c.c +1 -0
index 09bfe4b8f25aa..df89d916236d9 100644
@@ -582,6 +582,7 @@ static void __init kw_i2c_add(struct pmac_i2c_host_kw *host,
 	bus->close = kw_i2c_close;
 	bus->xfer = kw_i2c_xfer;
 	mutex_init(&bus->mutex);
+	lockdep_register_key(&bus->lock_key);
 	lockdep_set_class(&bus->mutex, &bus->lock_key);
 	if (controller == busnode)
 		bus->flags = pmac_i2c_multibus;
Signature · 1 line
2.34.1
author maintainer
#752 lore raw

PATCH 745/1039 KVM: PPC: Book3S: Suppress warnings when allocating too big memory slots

In reply to Greg Kroah-Hartman

From: Alexey Kardashevskiy <[email protected]>

[ Upstream commit 511d25d6b789fffcb20a3eb71899cf974a31bd9d ]

The userspace can trigger "vmalloc size %lu allocation failure: exceeds total pages" via the KVM_SET_USER_MEMORY_REGION ioctl.

This silences the warning by checking the limit before calling vzalloc() and returns ENOMEM if failed.

This does not call underlying valloc helpers as __vmalloc_node() is only exported when CONFIG_TEST_VMALLOC_MODULE and __vmalloc_node_range() is not exported at all.

Spotted by syzkaller.

---
 arch/powerpc/kvm/book3s_hv.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)
arch/powerpc/kvm/book3s_hv.c +6 -2
index 7b74fc0a986b8..94da0d25eb125 100644
@@ -4861,8 +4861,12 @@ static int kvmppc_core_prepare_memory_region_hv(struct kvm *kvm,
 	unsigned long npages = mem->memory_size >> PAGE_SHIFT;
 
 	if (change == KVM_MR_CREATE) {
-		slot->arch.rmap = vzalloc(array_size(npages,
-					  sizeof(*slot->arch.rmap)));
+		unsigned long size = array_size(npages, sizeof(*slot->arch.rmap));
+
+		if ((size >> PAGE_SHIFT) > totalram_pages())
+			return -ENOMEM;
+
+		slot->arch.rmap = vzalloc(size);
 		if (!slot->arch.rmap)
 			return -ENOMEM;
 	}
Signature · 1 line
2.34.1
author maintainer
#753 lore raw

PATCH 746/1039 KVM: PPC: Book3S: Suppress failed alloc warning in H_COPY_TOFROM_GUEST

In reply to Greg Kroah-Hartman

From: Alexey Kardashevskiy <[email protected]>

[ Upstream commit 792020907b11c6f9246c21977cab3bad985ae4b6 ]

H_COPY_TOFROM_GUEST is an hcall for an upper level VM to access its nested VMs memory. The userspace can trigger WARN_ON_ONCE(!(gfp & __GFP_NOWARN)) in __alloc_pages() by constructing a tiny VM which only does H_COPY_TOFROM_GUEST with a too big GPR9 (number of bytes to copy).

This silences the warning by adding __GFP_NOWARN.

Spotted by syzkaller.

---
 arch/powerpc/kvm/book3s_hv_nested.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/powerpc/kvm/book3s_hv_nested.c +1 -1
index ed8a2c9f56299..89295b52a97c3 100644
@@ -582,7 +582,7 @@ long kvmhv_copy_tofrom_guest_nested(struct kvm_vcpu *vcpu)
 	if (eaddr & (0xFFFUL << 52))
 		return H_PARAMETER;
 
-	buf = kzalloc(n, GFP_KERNEL);
+	buf = kzalloc(n, GFP_KERNEL | __GFP_NOWARN);
 	if (!buf)
 		return H_NO_MEM;
 
Signature · 1 line
2.34.1
author maintainer
#754 lore raw

PATCH 747/1039 w1: Misuse of get_user()/put_user() reported by sparse

In reply to Greg Kroah-Hartman

From: Christophe Leroy <[email protected]>

[ Upstream commit 33dc3e3e99e626ce51f462d883b05856c6c30b1d ]

sparse warnings: (new ones prefixed by >>)

Quoted 1 line
Quoted 1 line

drivers/w1/slaves/w1_ds28e04.c:342:13: sparse: sparse: incorrect type in initializer (different address spaces) @@ expected char [noderef] __user *_pu_addr @@ got char *buf @@

drivers/w1/slaves/w1_ds28e04.c:342:13: sparse: expected char [noderef] __user *_pu_addr
drivers/w1/slaves/w1_ds28e04.c:342:13: sparse: got char *buf

Quoted 1 line
Quoted 1 line

drivers/w1/slaves/w1_ds28e04.c:356:13: sparse: sparse: incorrect type in initializer (different address spaces) @@ expected char const [noderef] __user *_gu_addr @@ got char const *buf @@

drivers/w1/slaves/w1_ds28e04.c:356:13: sparse: expected char const [noderef] __user *_gu_addr
drivers/w1/slaves/w1_ds28e04.c:356:13: sparse: got char const *buf

The buffer buf is a failsafe buffer in kernel space, it's not user memory hence doesn't deserve the use of get_user() or put_user().

Access 'buf' content directly.

---
 drivers/w1/slaves/w1_ds28e04.c | 26 ++++++--------------------
 1 file changed, 6 insertions(+), 20 deletions(-)
drivers/w1/slaves/w1_ds28e04.c +6 -20
index e4f336111edc6..6cef6e2edb892 100644
@@ -32,7 +32,7 @@ static int w1_strong_pullup = 1;
 module_param_named(strong_pullup, w1_strong_pullup, int, 0);
 
 /* enable/disable CRC checking on DS28E04-100 memory accesses */
-static char w1_enable_crccheck = 1;
+static bool w1_enable_crccheck = true;
 
 #define W1_EEPROM_SIZE		512
 #define W1_PAGE_COUNT		16
@@ -339,32 +339,18 @@ static BIN_ATTR_RW(pio, 1);
 static ssize_t crccheck_show(struct device *dev, struct device_attribute *attr,
 			     char *buf)
 {
-	if (put_user(w1_enable_crccheck + 0x30, buf))
-		return -EFAULT;
-
-	return sizeof(w1_enable_crccheck);
+	return sysfs_emit(buf, "%d\n", w1_enable_crccheck);
 }
 
 static ssize_t crccheck_store(struct device *dev, struct device_attribute *attr,
 			      const char *buf, size_t count)
 {
-	char val;
-
-	if (count != 1 || !buf)
-		return -EINVAL;
+	int err = kstrtobool(buf, &w1_enable_crccheck);
 
-	if (get_user(val, buf))
-		return -EFAULT;
+	if (err)
+		return err;
 
-	/* convert to decimal */
-	val = val - 0x30;
-	if (val != 0 && val != 1)
-		return -EINVAL;
-
-	/* set the new value */
-	w1_enable_crccheck = val;
-
-	return sizeof(w1_enable_crccheck);
+	return count;
 }
 
 static DEVICE_ATTR_RW(crccheck);
Signature · 1 line
2.34.1
author maintainer
#755 lore raw

PATCH 748/1039 nvmem: core: set size for sysfs bin file

In reply to Greg Kroah-Hartman

From: Srinivas Kandagatla <[email protected]>

[ Upstream commit 86192251033308bb42f1e9813c962989d8ed07ec ]

For some reason we never set the size for nvmem sysfs binary file. Set this.

---
 drivers/nvmem/core.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/nvmem/core.c +2 -0
index e765d3d0542e5..23a38dcf0fc4d 100644
@@ -312,6 +312,8 @@ static umode_t nvmem_bin_attr_is_visible(struct kobject *kobj,
 	struct device *dev = kobj_to_dev(kobj);
 	struct nvmem_device *nvmem = to_nvmem_device(dev);
 
+	attr->size = nvmem->size;
+
 	return nvmem_bin_attr_get_umode(nvmem);
 }
 
Signature · 1 line
2.34.1
author maintainer
#756 lore raw

PATCH 749/1039 dm: fix alloc_dax error handling in alloc_dev

In reply to Greg Kroah-Hartman

From: Christoph Hellwig <[email protected]>

[ Upstream commit d751939235b9b7bc4af15f90a3e99288a8b844a7 ]

Make sure ->dax_dev is NULL on error so that the cleanup path doesn't trip over an ERR_PTR.

---
 drivers/md/dm.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/md/dm.c +3 -1
index 662742a310cbb..acc84dc1bded5 100644
@@ -1786,8 +1786,10 @@ static struct mapped_device *alloc_dev(int minor)
 	if (IS_ENABLED(CONFIG_DAX_DRIVER)) {
 		md->dax_dev = alloc_dax(md, md->disk->disk_name,
 					&dm_dax_ops, 0);
-		if (IS_ERR(md->dax_dev))
+		if (IS_ERR(md->dax_dev)) {
+			md->dax_dev = NULL;
 			goto bad;
+		}
 	}
 
 	format_dev_t(md->name, MKDEV(_major, minor));
Signature · 1 line
2.34.1
author maintainer
#757 lore raw

PATCH 750/1039 dm: make the DAX support depend on CONFIG_FS_DAX

In reply to Greg Kroah-Hartman

From: Christoph Hellwig <[email protected]>

[ Upstream commit 5d2a228b9e1319ff188f9ea89006fbe575561921 ]

The device mapper DAX support is all hanging off a block device and thus can't be used with device dax. Make it depend on CONFIG_FS_DAX instead of CONFIG_DAX_DRIVER. This also means that bdev_dax_pgoff only needs to be built under CONFIG_FS_DAX now.

---
 drivers/dax/super.c        | 6 ++----
 drivers/md/dm-linear.c     | 2 +-
 drivers/md/dm-log-writes.c | 2 +-
 drivers/md/dm-stripe.c     | 2 +-
 drivers/md/dm-writecache.c | 2 +-
 drivers/md/dm.c            | 2 +-
 6 files changed, 7 insertions(+), 9 deletions(-)
drivers/dax/super.c +2 -4
index b882cf8106ea3..e20d0cef10a18 100644
@@ -63,7 +63,7 @@ static int dax_host_hash(const char *host)
 	return hashlen_hash(hashlen_string("DAX", host)) % DAX_HASH_SIZE;
 }
 
-#ifdef CONFIG_BLOCK
+#if defined(CONFIG_BLOCK) && defined(CONFIG_FS_DAX)
 #include <linux/blkdev.h>
 
 int bdev_dax_pgoff(struct block_device *bdev, sector_t sector, size_t size,
@@ -80,7 +80,6 @@ int bdev_dax_pgoff(struct block_device *bdev, sector_t sector, size_t size,
 }
 EXPORT_SYMBOL(bdev_dax_pgoff);
 
-#if IS_ENABLED(CONFIG_FS_DAX)
 /**
  * dax_get_by_host() - temporary lookup mechanism for filesystem-dax
  * @host: alternate name for the device registered by a dax driver
@@ -219,8 +218,7 @@ bool dax_supported(struct dax_device *dax_dev, struct block_device *bdev,
 	return ret;
 }
 EXPORT_SYMBOL_GPL(dax_supported);
-#endif /* CONFIG_FS_DAX */
-#endif /* CONFIG_BLOCK */
+#endif /* CONFIG_BLOCK && CONFIG_FS_DAX */
 
 enum dax_device_flags {
 	/* !alive + rcu grace period == no new operations / mappings */
drivers/md/dm-linear.c +1 -1
index 66ba16713f696..0a260c35aeeed 100644
@@ -162,7 +162,7 @@ static int linear_iterate_devices(struct dm_target *ti,
 	return fn(ti, lc->dev, lc->start, ti->len, data);
 }
 
-#if IS_ENABLED(CONFIG_DAX_DRIVER)
+#if IS_ENABLED(CONFIG_FS_DAX)
 static long linear_dax_direct_access(struct dm_target *ti, pgoff_t pgoff,
 		long nr_pages, void **kaddr, pfn_t *pfn)
 {
drivers/md/dm-log-writes.c +1 -1
index 0b3ef977ceeba..3155875d4e5b0 100644
@@ -901,7 +901,7 @@ static void log_writes_io_hints(struct dm_target *ti, struct queue_limits *limit
 	limits->io_min = limits->physical_block_size;
 }
 
-#if IS_ENABLED(CONFIG_DAX_DRIVER)
+#if IS_ENABLED(CONFIG_FS_DAX)
 static int log_dax(struct log_writes_c *lc, sector_t sector, size_t bytes,
 		   struct iov_iter *i)
 {
drivers/md/dm-stripe.c +1 -1
index 6660b6b53d5bf..f084607220293 100644
@@ -300,7 +300,7 @@ static int stripe_map(struct dm_target *ti, struct bio *bio)
 	return DM_MAPIO_REMAPPED;
 }
 
-#if IS_ENABLED(CONFIG_DAX_DRIVER)
+#if IS_ENABLED(CONFIG_FS_DAX)
 static long stripe_dax_direct_access(struct dm_target *ti, pgoff_t pgoff,
 		long nr_pages, void **kaddr, pfn_t *pfn)
 {
drivers/md/dm-writecache.c +1 -1
index 4b8991cde223d..4f31591d2d25e 100644
@@ -38,7 +38,7 @@
 #define BITMAP_GRANULARITY	PAGE_SIZE
 #endif
 
-#if IS_ENABLED(CONFIG_ARCH_HAS_PMEM_API) && IS_ENABLED(CONFIG_DAX_DRIVER)
+#if IS_ENABLED(CONFIG_ARCH_HAS_PMEM_API) && IS_ENABLED(CONFIG_FS_DAX)
 #define DM_WRITECACHE_HAS_PMEM
 #endif
 
drivers/md/dm.c +1 -1
index acc84dc1bded5..b93fcc91176e5 100644
@@ -1783,7 +1783,7 @@ static struct mapped_device *alloc_dev(int minor)
 	md->disk->private_data = md;
 	sprintf(md->disk->disk_name, "dm-%d", minor);
 
-	if (IS_ENABLED(CONFIG_DAX_DRIVER)) {
+	if (IS_ENABLED(CONFIG_FS_DAX)) {
 		md->dax_dev = alloc_dax(md, md->disk->disk_name,
 					&dm_dax_ops, 0);
 		if (IS_ERR(md->dax_dev)) {
Signature · 1 line
2.34.1
author maintainer
#758 lore raw

PATCH 751/1039 ASoC: test-component: fix null pointer dereference.

In reply to Greg Kroah-Hartman

From: Ameer Hamza <[email protected]>

[ Upstream commit c686316ec1210d43653c91e104c1e4cd0156dc89 ]

Dereferncing of_id pointer will result in exception in current implementation since of_match_device() will assign it to NULL. Adding NULL check for protection.

---
 sound/soc/generic/test-component.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
sound/soc/generic/test-component.c +4 -1
index 85385a771d807..8fc97d3ff0110 100644
@@ -532,13 +532,16 @@ static int test_driver_probe(struct platform_device *pdev)
 	struct device_node *node = dev->of_node;
 	struct device_node *ep;
 	const struct of_device_id *of_id = of_match_device(test_of_match, &pdev->dev);
-	const struct test_adata *adata = of_id->data;
+	const struct test_adata *adata;
 	struct snd_soc_component_driver *cdriv;
 	struct snd_soc_dai_driver *ddriv;
 	struct test_dai_name *dname;
 	struct test_priv *priv;
 	int num, ret, i;
 
+	if (!of_id)
+		return -EINVAL;
+	adata = of_id->data;
 	num = of_graph_get_endpoint_count(node);
 	if (!num) {
 		dev_err(dev, "no port exits\n");
Signature · 1 line
2.34.1
author maintainer
#759 lore raw

PATCH 752/1039 interconnect: qcom: rpm: Prevent integer overflow in rate

In reply to Greg Kroah-Hartman

From: Stephan Gerhold <[email protected]>

[ Upstream commit a7d9436a6c85fcb8843c910fd323dcd7f839bf63 ]

Using icc-rpm on ARM32 currently results in clk_set_rate() errors during boot, e.g. "bus clk_set_rate error: -22". This is very similar to commit 7381e27b1e56 ("interconnect: qcom: msm8974: Prevent integer overflow in rate") where the u64 is converted to a signed long during clock rate rounding, resulting in an overflow on 32-bit platforms.

Let's fix it similarly by making sure that the rate does not exceed LONG_MAX. Such high clock rates will surely result in the maximum frequency of the bus anyway.

---
 drivers/interconnect/qcom/icc-rpm.c | 1 +
 1 file changed, 1 insertion(+)
drivers/interconnect/qcom/icc-rpm.c +1 -0
index ef7999a08c8bf..8114295a83129 100644
@@ -239,6 +239,7 @@ static int qcom_icc_set(struct icc_node *src, struct icc_node *dst)
 	rate = max(sum_bw, max_peak_bw);
 
 	do_div(rate, qn->buswidth);
+	rate = min_t(u64, rate, LONG_MAX);
 
 	if (qn->rate == rate)
 		return 0;
Signature · 1 line
2.34.1
author maintainer
#760 lore raw

PATCH 753/1039 scsi: ufs: Fix a kernel crash during shutdown

In reply to Greg Kroah-Hartman

From: Bart Van Assche <[email protected]>

[ Upstream commit 3489c34bd02b73a72646037d673a122a53cee174 ]

Fix the following kernel crash:

Unable to handle kernel paging request at virtual address ffffffc91e735000
Call trace:
 __queue_work+0x26c/0x624
 queue_work_on+0x6c/0xf0
 ufshcd_hold+0x12c/0x210
 __ufshcd_wl_suspend+0xc0/0x400
 ufshcd_wl_shutdown+0xb8/0xcc
 device_shutdown+0x184/0x224
 kernel_restart+0x4c/0x124
 __arm64_sys_reboot+0x194/0x264
 el0_svc_common+0xc8/0x1d4
 do_el0_svc+0x30/0x8c
 el0_svc+0x20/0x30
 el0_sync_handler+0x84/0xe4
 el0_sync+0x1bc/0x1c0

Fix this crash by ungating the clock before destroying the work queue on which clock gating work is queued.

---
 drivers/scsi/ufs/ufshcd.c | 15 ++++++++++-----
 1 file changed, 10 insertions(+), 5 deletions(-)
drivers/scsi/ufs/ufshcd.c +10 -5
index 4394081806978..c94377aa82739 100644
@@ -1666,7 +1666,8 @@ int ufshcd_hold(struct ufs_hba *hba, bool async)
 	bool flush_result;
 	unsigned long flags;
 
-	if (!ufshcd_is_clkgating_allowed(hba))
+	if (!ufshcd_is_clkgating_allowed(hba) ||
+	    !hba->clk_gating.is_initialized)
 		goto out;
 	spin_lock_irqsave(hba->host->host_lock, flags);
 	hba->clk_gating.active_reqs++;
@@ -1826,7 +1827,7 @@ static void __ufshcd_release(struct ufs_hba *hba)
 
 	if (hba->clk_gating.active_reqs || hba->clk_gating.is_suspended ||
 	    hba->ufshcd_state != UFSHCD_STATE_OPERATIONAL ||
-	    hba->outstanding_tasks ||
+	    hba->outstanding_tasks || !hba->clk_gating.is_initialized ||
 	    hba->active_uic_cmd || hba->uic_async_done ||
 	    hba->clk_gating.state == CLKS_OFF)
 		return;
@@ -1961,11 +1962,15 @@ static void ufshcd_exit_clk_gating(struct ufs_hba *hba)
 {
 	if (!hba->clk_gating.is_initialized)
 		return;
+
 	ufshcd_remove_clk_gating_sysfs(hba);
-	cancel_work_sync(&hba->clk_gating.ungate_work);
-	cancel_delayed_work_sync(&hba->clk_gating.gate_work);
-	destroy_workqueue(hba->clk_gating.clk_gating_workq);
+
+	/* Ungate the clock if necessary. */
+	ufshcd_hold(hba, false);
 	hba->clk_gating.is_initialized = false;
+	ufshcd_release(hba);
+
+	destroy_workqueue(hba->clk_gating.clk_gating_workq);
 }
 
 /* Must be called with host lock acquired */
Signature · 1 line
2.34.1
author maintainer
#761 lore raw

PATCH 754/1039 scsi: lpfc: Fix leaked lpfc_dmabuf mbox allocations with NPIV

In reply to Greg Kroah-Hartman

From: James Smart <[email protected]>

[ Upstream commit f0d3919697492950f57a26a1093aee53880d669d ]

During rmmod testing, messages appeared indicating lpfc_mbuf_pool entries were still busy. This situation was only seen doing rmmod after at least 1 vport (NPIV) instance was created and destroyed. The number of messages scaled with the number of vports created.

When a vport is created, it can receive a PLOGI from another initiator Nport. When this happens, the driver prepares to ack the PLOGI and prepares an RPI for registration (via mbx cmd) which includes an mbuf allocation. During the unsolicited PLOGI processing and after the RPI preparation, the driver recognizes it is one of the vport instances and decides to reject the PLOGI. During the LS_RJT preparation for the PLOGI, the mailbox struct allocated for RPI registration is freed, but the mbuf that was also allocated is not released.

Fix by freeing the mbuf with the mailbox struct in the LS_RJT path.

As part of the code review to figure the issue out a couple of other areas where found that also would not have released the mbuf. Those are cleaned up as well.

---
 drivers/scsi/lpfc/lpfc_els.c       | 6 +++++-
 drivers/scsi/lpfc/lpfc_init.c      | 8 ++++++--
 drivers/scsi/lpfc/lpfc_nportdisc.c | 6 ++++++
 3 files changed, 17 insertions(+), 3 deletions(-)
drivers/scsi/lpfc/lpfc_els.c +5 -1
index e83453bea2aee..5c10416c1c75a 100644
@@ -6899,6 +6899,7 @@ static int
 lpfc_get_rdp_info(struct lpfc_hba *phba, struct lpfc_rdp_context *rdp_context)
 {
 	LPFC_MBOXQ_t *mbox = NULL;
+	struct lpfc_dmabuf *mp;
 	int rc;
 
 	mbox = mempool_alloc(phba->mbox_mem_pool, GFP_KERNEL);
@@ -6914,8 +6915,11 @@ lpfc_get_rdp_info(struct lpfc_hba *phba, struct lpfc_rdp_context *rdp_context)
 	mbox->mbox_cmpl = lpfc_mbx_cmpl_rdp_page_a0;
 	mbox->ctx_ndlp = (struct lpfc_rdp_context *)rdp_context;
 	rc = lpfc_sli_issue_mbox(phba, mbox, MBX_NOWAIT);
-	if (rc == MBX_NOT_FINISHED)
+	if (rc == MBX_NOT_FINISHED) {
+		mp = (struct lpfc_dmabuf *)mbox->ctx_buf;
+		lpfc_mbuf_free(phba, mp->virt, mp->phys);
 		goto issue_mbox_fail;
+	}
 
 	return 0;
 
drivers/scsi/lpfc/lpfc_init.c +6 -2
index ba17a8f740a95..7628b0634c57a 100644
@@ -5373,8 +5373,10 @@ lpfc_sli4_async_link_evt(struct lpfc_hba *phba,
 	 */
 	if (!(phba->hba_flag & HBA_FCOE_MODE)) {
 		rc = lpfc_sli_issue_mbox(phba, pmb, MBX_NOWAIT);
-		if (rc == MBX_NOT_FINISHED)
+		if (rc == MBX_NOT_FINISHED) {
+			lpfc_mbuf_free(phba, mp->virt, mp->phys);
 			goto out_free_dmabuf;
+		}
 		return;
 	}
 	/*
@@ -6337,8 +6339,10 @@ lpfc_sli4_async_fc_evt(struct lpfc_hba *phba, struct lpfc_acqe_fc_la *acqe_fc)
 	}
 
 	rc = lpfc_sli_issue_mbox(phba, pmb, MBX_NOWAIT);
-	if (rc == MBX_NOT_FINISHED)
+	if (rc == MBX_NOT_FINISHED) {
+		lpfc_mbuf_free(phba, mp->virt, mp->phys);
 		goto out_free_dmabuf;
+	}
 	return;
 
 out_free_dmabuf:
drivers/scsi/lpfc/lpfc_nportdisc.c +6 -0
index 27263f02ab9f6..7d717a4ac14d1 100644
@@ -322,6 +322,7 @@ lpfc_rcv_plogi(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
 {
 	struct lpfc_hba    *phba = vport->phba;
 	struct lpfc_dmabuf *pcmd;
+	struct lpfc_dmabuf *mp;
 	uint64_t nlp_portwwn = 0;
 	uint32_t *lp;
 	IOCB_t *icmd;
@@ -571,6 +572,11 @@ lpfc_rcv_plogi(struct lpfc_vport *vport, struct lpfc_nodelist *ndlp,
 		 * a default RPI.
 		 */
 		if (phba->sli_rev == LPFC_SLI_REV4) {
+			mp = (struct lpfc_dmabuf *)login_mbox->ctx_buf;
+			if (mp) {
+				lpfc_mbuf_free(phba, mp->virt, mp->phys);
+				kfree(mp);
+			}
 			mempool_free(login_mbox, phba->mbox_mem_pool);
 			login_mbox = NULL;
 		} else {
Signature · 1 line
2.34.1
author maintainer
#762 lore raw

PATCH 755/1039 scsi: lpfc: Trigger SLI4 firmware dump before doing driver cleanup

In reply to Greg Kroah-Hartman

From: James Smart <[email protected]>

[ Upstream commit 7dd2e2a923173d637c272e483966be8e96a72b64 ]

Extraneous teardown routines are present in the firmware dump path causing altered states in firmware captures.

When a firmware dump is requested via sysfs, trigger the dump immediately without tearing down structures and changing adapter state.

The driver shall rely on pre-existing firmware error state clean up handlers to restore the adapter.

---
 drivers/scsi/lpfc/lpfc.h         |  2 +-
 drivers/scsi/lpfc/lpfc_attr.c    | 62 ++++++++++++++++++++------------
 drivers/scsi/lpfc/lpfc_hbadisc.c |  8 ++++-
 drivers/scsi/lpfc/lpfc_sli.c     |  6 ----
 4 files changed, 48 insertions(+), 30 deletions(-)
drivers/scsi/lpfc/lpfc.h +1 -1
index 2f8e6d0a926fe..54c58392fd5d0 100644
@@ -1023,7 +1023,6 @@ struct lpfc_hba {
 #define HBA_DEVLOSS_TMO         0x2000 /* HBA in devloss timeout */
 #define HBA_RRQ_ACTIVE		0x4000 /* process the rrq active list */
 #define HBA_IOQ_FLUSH		0x8000 /* FCP/NVME I/O queues being flushed */
-#define HBA_FW_DUMP_OP		0x10000 /* Skips fn reset before FW dump */
 #define HBA_RECOVERABLE_UE	0x20000 /* Firmware supports recoverable UE */
 #define HBA_FORCED_LINK_SPEED	0x40000 /*
 					 * Firmware supports Forced Link Speed
@@ -1040,6 +1039,7 @@ struct lpfc_hba {
 #define HBA_HBEAT_TMO		0x8000000 /* HBEAT initiated after timeout */
 #define HBA_FLOGI_OUTSTANDING	0x10000000 /* FLOGI is outstanding */
 
+	struct completion *fw_dump_cmpl; /* cmpl event tracker for fw_dump */
 	uint32_t fcp_ring_in_use; /* When polling test if intr-hndlr active*/
 	struct lpfc_dmabuf slim2p;
 
drivers/scsi/lpfc/lpfc_attr.c +40 -22

89 lines not on this page · raw message

drivers/scsi/lpfc/lpfc_hbadisc.c +7 -1
index 9fe6e5b386ce3..5e54ec503f18b 100644
@@ -869,10 +869,16 @@ lpfc_work_done(struct lpfc_hba *phba)
 	if (phba->pci_dev_grp == LPFC_PCI_DEV_OC)
 		lpfc_sli4_post_async_mbox(phba);
 
-	if (ha_copy & HA_ERATT)
+	if (ha_copy & HA_ERATT) {
 		/* Handle the error attention event */
 		lpfc_handle_eratt(phba);
 
+		if (phba->fw_dump_cmpl) {
+			complete(phba->fw_dump_cmpl);
+			phba->fw_dump_cmpl = NULL;
+		}
+	}
+
 	if (ha_copy & HA_MBATT)
 		lpfc_sli_handle_mb_event(phba);
 
drivers/scsi/lpfc/lpfc_sli.c +0 -6
index 5dedb3de271d8..513a78d08b1d5 100644
@@ -5046,12 +5046,6 @@ lpfc_sli4_brdreset(struct lpfc_hba *phba)
 	phba->fcf.fcf_flag = 0;
 	spin_unlock_irq(&phba->hbalock);
 
-	/* SLI4 INTF 2: if FW dump is being taken skip INIT_PORT */
-	if (phba->hba_flag & HBA_FW_DUMP_OP) {
-		phba->hba_flag &= ~HBA_FW_DUMP_OP;
-		return rc;
-	}
-
 	/* Now physically reset the device */
 	lpfc_printf_log(phba, KERN_INFO, LOG_INIT,
 			"0389 Performing PCI function reset!\n");
Signature · 1 line
2.34.1
author maintainer
#763 lore raw

PATCH 756/1039 ALSA: seq: Set upper limit of processed events

In reply to Greg Kroah-Hartman

From: Takashi Iwai <[email protected]>

[ Upstream commit 6fadb494a638d8b8a55864ecc6ac58194f03f327 ]

Currently ALSA sequencer core tries to process the queued events as much as possible when they become dispatchable. If applications try to queue too massive events to be processed at the very same timing, the sequencer core would still try to process such all events, either in the interrupt context or via some notifier; in either away, it might be a cause of RCU stall or such problems.

As a potential workaround for those problems, this patch adds the upper limit of the amount of events to be processed. The remaining events are processed in the next batch, so they won't be lost.

For the time being, it's limited up to 1000 events per queue, which should be high enough for any normal usages.

---
 sound/core/seq/seq_queue.c | 14 ++++++++++++--
 1 file changed, 12 insertions(+), 2 deletions(-)
sound/core/seq/seq_queue.c +12 -2
index d6c02dea976c8..bc933104c3eea 100644
@@ -235,12 +235,15 @@ struct snd_seq_queue *snd_seq_queue_find_name(char *name)
 
 /* -------------------------------------------------------- */
 
+#define MAX_CELL_PROCESSES_IN_QUEUE	1000
+
 void snd_seq_check_queue(struct snd_seq_queue *q, int atomic, int hop)
 {
 	unsigned long flags;
 	struct snd_seq_event_cell *cell;
 	snd_seq_tick_time_t cur_tick;
 	snd_seq_real_time_t cur_time;
+	int processed = 0;
 
 	if (q == NULL)
 		return;
@@ -263,6 +266,8 @@ void snd_seq_check_queue(struct snd_seq_queue *q, int atomic, int hop)
 		if (!cell)
 			break;
 		snd_seq_dispatch_event(cell, atomic, hop);
+		if (++processed >= MAX_CELL_PROCESSES_IN_QUEUE)
+			goto out; /* the rest processed at the next batch */
 	}
 
 	/* Process time queue... */
@@ -272,14 +277,19 @@ void snd_seq_check_queue(struct snd_seq_queue *q, int atomic, int hop)
 		if (!cell)
 			break;
 		snd_seq_dispatch_event(cell, atomic, hop);
+		if (++processed >= MAX_CELL_PROCESSES_IN_QUEUE)
+			goto out; /* the rest processed at the next batch */
 	}
 
+ out:
 	/* free lock */
 	spin_lock_irqsave(&q->check_lock, flags);
 	if (q->check_again) {
 		q->check_again = 0;
-		spin_unlock_irqrestore(&q->check_lock, flags);
-		goto __again;
+		if (processed < MAX_CELL_PROCESSES_IN_QUEUE) {
+			spin_unlock_irqrestore(&q->check_lock, flags);
+			goto __again;
+		}
 	}
 	q->check_blocked = 0;
 	spin_unlock_irqrestore(&q->check_lock, flags);
Signature · 1 line
2.34.1
author maintainer
#764 lore raw

PATCH 757/1039 MIPS: Loongson64: Use three arguments for slti

In reply to Greg Kroah-Hartman

From: Nathan Chancellor <[email protected]>

[ Upstream commit f2c6c22fa83ab2577619009057b3ebcb5305bb03 ]

LLVM's integrated assembler does not support 'slti <reg>, <imm>':

<instantiation>:16:12: error: invalid operand for instruction
slti $12, (0x6300 | 0x0008)

           ^

arch/mips/kernel/head.S:86:2: note: while in macro instantiation
kernel_entry_setup # cpu specific setup
^
<instantiation>:16:12: error: invalid operand for instruction
slti $12, (0x6300 | 0x0008)

           ^

arch/mips/kernel/head.S:150:2: note: while in macro instantiation
smp_slave_setup
^

To increase compatibility with LLVM's integrated assembler, use the full form of 'slti <reg>, <reg>, <imm>', which matches the rest of arch/mips/. This does not result in any change for GNU as.

---
 arch/mips/include/asm/mach-loongson64/kernel-entry-init.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
arch/mips/include/asm/mach-loongson64/kernel-entry-init.h +2 -2
index 13373c5144f89..efb41b3519747 100644
@@ -32,7 +32,7 @@
 	nop
 	/* Loongson-3A R2/R3 */
 	andi	t0, (PRID_IMP_MASK | PRID_REV_MASK)
-	slti	t0, (PRID_IMP_LOONGSON_64C | PRID_REV_LOONGSON3A_R2_0)
+	slti	t0, t0, (PRID_IMP_LOONGSON_64C | PRID_REV_LOONGSON3A_R2_0)
 	bnez	t0, 2f
 	nop
 1:
@@ -63,7 +63,7 @@
 	nop
 	/* Loongson-3A R2/R3 */
 	andi	t0, (PRID_IMP_MASK | PRID_REV_MASK)
-	slti	t0, (PRID_IMP_LOONGSON_64C | PRID_REV_LOONGSON3A_R2_0)
+	slti	t0, t0, (PRID_IMP_LOONGSON_64C | PRID_REV_LOONGSON3A_R2_0)
 	bnez	t0, 2f
 	nop
 1:
Signature · 1 line
2.34.1
author maintainer
#765 lore raw

PATCH 758/1039 powerpc/40x: Map 32Mbytes of memory at startup

In reply to Greg Kroah-Hartman

From: Christophe Leroy <[email protected]>

[ Upstream commit 06e7cbc29e97b4713b4ea6def04ae8501a7d1a59 ]

As reported by Carlo, 16Mbytes is not enough with modern kernels that tend to be a bit big, so map another 16M page at boot.

---
 arch/powerpc/kernel/head_40x.S | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)
arch/powerpc/kernel/head_40x.S +8 -1
index 7d72ee5ab387c..e783860bea838 100644
@@ -27,6 +27,7 @@
 
 #include <linux/init.h>
 #include <linux/pgtable.h>
+#include <linux/sizes.h>
 #include <asm/processor.h>
 #include <asm/page.h>
 #include <asm/mmu.h>
@@ -650,7 +651,7 @@ start_here:
 	b	.		/* prevent prefetch past rfi */
 
 /* Set up the initial MMU state so we can do the first level of
- * kernel initialization.  This maps the first 16 MBytes of memory 1:1
+ * kernel initialization.  This maps the first 32 MBytes of memory 1:1
  * virtual to physical and more importantly sets the cache mode.
  */
 initial_mmu:
@@ -687,6 +688,12 @@ initial_mmu:
 	tlbwe	r4,r0,TLB_DATA		/* Load the data portion of the entry */
 	tlbwe	r3,r0,TLB_TAG		/* Load the tag portion of the entry */
 
+	li	r0,62			/* TLB slot 62 */
+	addis	r4,r4,SZ_16M@h
+	addis	r3,r3,SZ_16M@h
+	tlbwe	r4,r0,TLB_DATA		/* Load the data portion of the entry */
+	tlbwe	r3,r0,TLB_TAG		/* Load the tag portion of the entry */
+
 	isync
 
 	/* Establish the exception vector base
Signature · 1 line
2.34.1
author maintainer
#766 lore raw

PATCH 759/1039 selftests/powerpc/spectre_v2: Return skip code when miss_percent is high

In reply to Greg Kroah-Hartman

From: Thadeu Lima de Souza Cascardo <[email protected]>

[ Upstream commit 3c42e9542050d49610077e083c7c3f5fd5e26820 ]

A mis-match between reported and actual mitigation is not restricted to the Vulnerable case. The guest might also report the mitigation as "Software count cache flush" and the host will still mitigate with branch cache disabled.

So, instead of skipping depending on the detected mitigation, simply skip whenever the detected miss_percent is the expected one for a fully mitigated system, that is, above 95%.

---
 tools/testing/selftests/powerpc/security/spectre_v2.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
tools/testing/selftests/powerpc/security/spectre_v2.c +1 -1
index adc2b7294e5fd..83647b8277e7d 100644
@@ -193,7 +193,7 @@ int spectre_v2_test(void)
 			 * We are not vulnerable and reporting otherwise, so
 			 * missing such a mismatch is safe.
 			 */
-			if (state == VULNERABLE)
+			if (miss_percent > 95)
 				return 4;
 
 			return 1;
Signature · 1 line
2.34.1
author maintainer
#767 lore raw

PATCH 760/1039 powerpc: handle kdump appropriately with crash_kexec_post_notifiers option

In reply to Greg Kroah-Hartman

From: Hari Bathini <[email protected]>

[ Upstream commit 219572d2fc4135b5ce65c735d881787d48b10e71 ]

Kdump can be triggered after panic_notifers since commit f06e5153f4ae2 ("kernel/panic.c: add "crash_kexec_post_notifiers" option for kdump after panic_notifers") introduced crash_kexec_post_notifiers option. But using this option would mean smp_send_stop(), that marks all other CPUs as offline, gets called before kdump is triggered. As a result, kdump routines fail to save other CPUs' registers. To fix this, kdump friendly crash_smp_send_stop() function was introduced with kernel commit 0ee59413c967 ("x86/panic: replace smp_send_stop() with kdump friendly version in panic path"). Override this kdump friendly weak function to handle crash_kexec_post_notifiers option appropriately on powerpc.

---
 arch/powerpc/kernel/smp.c | 30 ++++++++++++++++++++++++++++++
 1 file changed, 30 insertions(+)
arch/powerpc/kernel/smp.c +30 -0
index aee3a7119f977..7201fdcf02f1c 100644
@@ -620,6 +620,36 @@ void crash_send_ipi(void (*crash_ipi_callback)(struct pt_regs *))
 }
 #endif
 
+#ifdef CONFIG_NMI_IPI
+static void crash_stop_this_cpu(struct pt_regs *regs)
+#else
+static void crash_stop_this_cpu(void *dummy)
+#endif
+{
+	/*
+	 * Just busy wait here and avoid marking CPU as offline to ensure
+	 * register data is captured appropriately.
+	 */
+	while (1)
+		cpu_relax();
+}
+
+void crash_smp_send_stop(void)
+{
+	static bool stopped = false;
+
+	if (stopped)
+		return;
+
+	stopped = true;
+
+#ifdef CONFIG_NMI_IPI
+	smp_send_nmi_ipi(NMI_IPI_ALL_OTHERS, crash_stop_this_cpu, 1000000);
+#else
+	smp_call_function(crash_stop_this_cpu, NULL, 0);
+#endif /* CONFIG_NMI_IPI */
+}
+
 #ifdef CONFIG_NMI_IPI
 static void nmi_stop_this_cpu(struct pt_regs *regs)
 {
Signature · 1 line
2.34.1
author maintainer
#768 lore raw

PATCH 761/1039 powerpc/fadump: Fix inaccurate CPU state info in vmcore generated with panic

In reply to Greg Kroah-Hartman

From: Hari Bathini <[email protected]>

[ Upstream commit 06e629c25daa519be620a8c17359ae8fc7a2e903 ]

In panic path, fadump is triggered via a panic notifier function. Before calling panic notifier functions, smp_send_stop() gets called, which stops all CPUs except the panic'ing CPU. Commit 8389b37dffdc ("powerpc: stop_this_cpu: remove the cpu from the online map.") and again commit bab26238bbd4 ("powerpc: Offline CPU in stop_this_cpu()") started marking CPUs as offline while stopping them. So, if a kernel has either of the above commits, vmcore captured with fadump via panic path would not process register data for all CPUs except the panic'ing CPU. Sample output of crash-utility with such vmcore:

  # crash vmlinux vmcore
  ...
        KERNEL: vmlinux
      DUMPFILE: vmcore  [PARTIAL DUMP]
          CPUS: 1
          DATE: Wed Nov 10 09:56:34 EST 2021
        UPTIME: 00:00:42
  LOAD AVERAGE: 2.27, 0.69, 0.24
         TASKS: 183
      NODENAME: XXXXXXXXX
       RELEASE: 5.15.0+
       VERSION: #974 SMP Wed Nov 10 04:18:19 CST 2021
       MACHINE: ppc64le  (2500 Mhz)
        MEMORY: 8 GB
         PANIC: "Kernel panic - not syncing: sysrq triggered crash"
           PID: 3394
       COMMAND: "bash"
          TASK: c0000000150a5f80  [THREAD_INFO: c0000000150a5f80]
           CPU: 1
         STATE: TASK_RUNNING (PANIC)
  crash> p -x __cpu_online_mask
  __cpu_online_mask = $1 = {
    bits = {0x2, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0}
  }
  crash>
  crash>
  crash> p -x __cpu_active_mask
  __cpu_active_mask = $2 = {
    bits = {0xff, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0}
  }
  crash>

While this has been the case since fadump was introduced, the issue was not identified for two probable reasons:

- In general, the bulk of the vmcores analyzed were from crash due to exception.

- The above did change since commit 8341f2f222d7 ("sysrq: Use panic() to force a crash") started using panic() instead of deferencing NULL pointer to force a kernel crash. But then commit de6e5d38417e ("powerpc: smp_send_stop do not offline stopped CPUs") stopped marking CPUs as offline till kernel commit bab26238bbd4 ("powerpc: Offline CPU in stop_this_cpu()") reverted that change.

To ensure post processing register data of all other CPUs happens as intended, let panic() function take the crash friendly path (read crash_smp_send_stop()) with the help of crash_kexec_post_notifiers option. Also, as register data for all CPUs is captured by f/w, skip IPI callbacks here for fadump, to avoid any complications in finding the right backtraces.

---
 arch/powerpc/kernel/fadump.c |  8 ++++++++
 arch/powerpc/kernel/smp.c    | 10 ++++++++++
 2 files changed, 18 insertions(+)
arch/powerpc/kernel/fadump.c +8 -0
index b7ceb041743c9..60f5fc14aa235 100644
@@ -1641,6 +1641,14 @@ int __init setup_fadump(void)
 	else if (fw_dump.reserve_dump_area_size)
 		fw_dump.ops->fadump_init_mem_struct(&fw_dump);
 
+	/*
+	 * In case of panic, fadump is triggered via ppc_panic_event()
+	 * panic notifier. Setting crash_kexec_post_notifiers to 'true'
+	 * lets panic() function take crash friendly path before panic
+	 * notifiers are invoked.
+	 */
+	crash_kexec_post_notifiers = true;
+
 	return 1;
 }
 subsys_initcall(setup_fadump);
arch/powerpc/kernel/smp.c +10 -0
index 7201fdcf02f1c..c338f9d8ab37a 100644
@@ -61,6 +61,7 @@
 #include <asm/cpu_has_feature.h>
 #include <asm/ftrace.h>
 #include <asm/kup.h>
+#include <asm/fadump.h>
 
 #ifdef DEBUG
 #include <asm/udbg.h>
@@ -638,6 +639,15 @@ void crash_smp_send_stop(void)
 {
 	static bool stopped = false;
 
+	/*
+	 * In case of fadump, register data for all CPUs is captured by f/w
+	 * on ibm,os-term rtas call. Skip IPI callbacks to other CPUs before
+	 * this rtas call to avoid tricky post processing of those CPUs'
+	 * backtraces.
+	 */
+	if (should_fadump_crash())
+		return;
+
 	if (stopped)
 		return;
 
Signature · 1 line
2.34.1
author maintainer
#769 lore raw

PATCH 762/1039 ASoC: SOF: Intel: hda: add quirks for HDAudio DMA position information

In reply to Greg Kroah-Hartman

From: Pierre-Louis Bossart <[email protected]>

[ Upstream commit 288fad2f71fa0b989c075d4984879c26d47cfb06 ]

The code inherited from the Skylake driver does not seem to follow any known hardware recommendations.

The only two recommended options are
a) use DPIB registers if VC1 traffic is not allowed
b) use DPIB DDR update if VC1 traffic is used

In all of SOF-based updated, VC1 is not supported so we can 'safely' move to using DPIB registers only.

This patch keeps the legacy code, in case there was an undocumented issue lost to history, and adds the DPIB DDR update for additional debug.

---
 sound/soc/sof/intel/hda-pcm.c | 86 +++++++++++++++++++++++++----------
 sound/soc/sof/intel/hda.c     |  9 +++-
 sound/soc/sof/intel/hda.h     |  6 +++
 3 files changed, 75 insertions(+), 26 deletions(-)
sound/soc/sof/intel/hda-pcm.c +61 -25
index cc8ddef37f37b..41cb60955f5c1 100644
@@ -172,38 +172,74 @@ snd_pcm_uframes_t hda_dsp_pcm_pointer(struct snd_sof_dev *sdev,
 		goto found;
 	}
 
-	/*
-	 * DPIB/posbuf position mode:
-	 * For Playback, Use DPIB register from HDA space which
-	 * reflects the actual data transferred.
-	 * For Capture, Use the position buffer for pointer, as DPIB
-	 * is not accurate enough, its update may be completed
-	 * earlier than the data written to DDR.
-	 */
-	if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) {
+	switch (sof_hda_position_quirk) {
+	case SOF_HDA_POSITION_QUIRK_USE_SKYLAKE_LEGACY:
+		/*
+		 * This legacy code, inherited from the Skylake driver,
+		 * mixes DPIB registers and DPIB DDR updates and
+		 * does not seem to follow any known hardware recommendations.
+		 * It's not clear e.g. why there is a different flow
+		 * for capture and playback, the only information that matters is
+		 * what traffic class is used, and on all SOF-enabled platforms
+		 * only VC0 is supported so the work-around was likely not necessary
+		 * and quite possibly wrong.
+		 */
+
+		/* DPIB/posbuf position mode:
+		 * For Playback, Use DPIB register from HDA space which
+		 * reflects the actual data transferred.
+		 * For Capture, Use the position buffer for pointer, as DPIB
+		 * is not accurate enough, its update may be completed
+		 * earlier than the data written to DDR.
+		 */
+		if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) {
+			pos = snd_sof_dsp_read(sdev, HDA_DSP_HDA_BAR,
+					       AZX_REG_VS_SDXDPIB_XBASE +
+					       (AZX_REG_VS_SDXDPIB_XINTERVAL *
+						hstream->index));
+		} else {
+			/*
+			 * For capture stream, we need more workaround to fix the
+			 * position incorrect issue:
+			 *
+			 * 1. Wait at least 20us before reading position buffer after
+			 * the interrupt generated(IOC), to make sure position update
+			 * happens on frame boundary i.e. 20.833uSec for 48KHz.
+			 * 2. Perform a dummy Read to DPIB register to flush DMA
+			 * position value.
+			 * 3. Read the DMA Position from posbuf. Now the readback
+			 * value should be >= period boundary.
+			 */
+			usleep_range(20, 21);
+			snd_sof_dsp_read(sdev, HDA_DSP_HDA_BAR,
+					 AZX_REG_VS_SDXDPIB_XBASE +
+					 (AZX_REG_VS_SDXDPIB_XINTERVAL *
+					  hstream->index));
+			pos = snd_hdac_stream_get_pos_posbuf(hstream);
+		}
+		break;
+	case SOF_HDA_POSITION_QUIRK_USE_DPIB_REGISTERS:
+		/*
+		 * In case VC1 traffic is disabled this is the recommended option
+		 */
 		pos = snd_sof_dsp_read(sdev, HDA_DSP_HDA_BAR,
 				       AZX_REG_VS_SDXDPIB_XBASE +
 				       (AZX_REG_VS_SDXDPIB_XINTERVAL *
 					hstream->index));
-	} else {
+		break;
+	case SOF_HDA_POSITION_QUIRK_USE_DPIB_DDR_UPDATE:
 		/*
-		 * For capture stream, we need more workaround to fix the
-		 * position incorrect issue:
-		 *
-		 * 1. Wait at least 20us before reading position buffer after
-		 * the interrupt generated(IOC), to make sure position update
-		 * happens on frame boundary i.e. 20.833uSec for 48KHz.
-		 * 2. Perform a dummy Read to DPIB register to flush DMA
-		 * position value.
-		 * 3. Read the DMA Position from posbuf. Now the readback
-		 * value should be >= period boundary.
+		 * This is the recommended option when VC1 is enabled.
+		 * While this isn't needed for SOF platforms it's added for
+		 * consistency and debug.
 		 */
-		usleep_range(20, 21);
-		snd_sof_dsp_read(sdev, HDA_DSP_HDA_BAR,
-				 AZX_REG_VS_SDXDPIB_XBASE +
-				 (AZX_REG_VS_SDXDPIB_XINTERVAL *
-				  hstream->index));
 		pos = snd_hdac_stream_get_pos_posbuf(hstream);
+		break;
+	default:
+		dev_err_once(sdev->dev, "hda_position_quirk value %d not supported\n",
+			     sof_hda_position_quirk);
+		pos = 0;
+		break;
 	}
 
 	if (pos >= hstream->bufsize)
sound/soc/sof/intel/hda.c +8 -1
index 2c0d4d06ab364..25200a0e1dc9d 100644
@@ -440,6 +440,10 @@ MODULE_PARM_DESC(use_msi, "SOF HDA use PCI MSI mode");
 #define hda_use_msi	(1)
 #endif
 
+int sof_hda_position_quirk = SOF_HDA_POSITION_QUIRK_USE_DPIB_REGISTERS;
+module_param_named(position_quirk, sof_hda_position_quirk, int, 0444);
+MODULE_PARM_DESC(position_quirk, "SOF HDaudio position quirk");
+
 static char *hda_model;
 module_param(hda_model, charp, 0444);
 MODULE_PARM_DESC(hda_model, "Use the given HDA board model.");
@@ -618,7 +622,10 @@ static int hda_init(struct snd_sof_dev *sdev)
 	/* HDA bus init */
 	sof_hda_bus_init(bus, &pci->dev);
 
-	bus->use_posbuf = 1;
+	if (sof_hda_position_quirk == SOF_HDA_POSITION_QUIRK_USE_DPIB_REGISTERS)
+		bus->use_posbuf = 0;
+	else
+		bus->use_posbuf = 1;
 	bus->bdl_pos_adj = 0;
 	bus->sync_write = 1;
 
sound/soc/sof/intel/hda.h +6 -0
index 1195018a1f4f5..dba4733ccf9ae 100644
@@ -738,4 +738,10 @@ struct sof_ipc_dai_config;
 int hda_ctrl_dai_widget_setup(struct snd_soc_dapm_widget *w);
 int hda_ctrl_dai_widget_free(struct snd_soc_dapm_widget *w);
 
+#define SOF_HDA_POSITION_QUIRK_USE_SKYLAKE_LEGACY	(0) /* previous implementation */
+#define SOF_HDA_POSITION_QUIRK_USE_DPIB_REGISTERS	(1) /* recommended if VC0 only */
+#define SOF_HDA_POSITION_QUIRK_USE_DPIB_DDR_UPDATE	(2) /* recommended with VC0 or VC1 */
+
+extern int sof_hda_position_quirk;
+
 #endif
Signature · 1 line
2.34.1
author maintainer
#770 lore raw

PATCH 763/1039 udf: Fix error handling in udf_new_inode()

In reply to Greg Kroah-Hartman

From: Jan Kara <[email protected]>

[ Upstream commit f05f2429eec60851b98bdde213de31dab697c01b ]

When memory allocation of iinfo or block allocation fails, already allocated struct udf_inode_info gets freed with iput() and udf_evict_inode() may look at inode fields which are not properly initialized. Fix it by marking inode bad before dropping reference to it in udf_new_inode().

---
 fs/udf/ialloc.c | 2 ++
 1 file changed, 2 insertions(+)
fs/udf/ialloc.c +2 -0
index 2ecf0e87660e3..b5d611cee749c 100644
@@ -77,6 +77,7 @@ struct inode *udf_new_inode(struct inode *dir, umode_t mode)
 					GFP_KERNEL);
 	}
 	if (!iinfo->i_data) {
+		make_bad_inode(inode);
 		iput(inode);
 		return ERR_PTR(-ENOMEM);
 	}
@@ -86,6 +87,7 @@ struct inode *udf_new_inode(struct inode *dir, umode_t mode)
 			      dinfo->i_location.partitionReferenceNum,
 			      start, &err);
 	if (err) {
+		make_bad_inode(inode);
 		iput(inode);
 		return ERR_PTR(err);
 	}
Signature · 1 line
2.34.1
author maintainer
#771 lore raw

PATCH 764/1039 MIPS: OCTEON: add put_device() after of_find_device_by_node()

In reply to Greg Kroah-Hartman

From: Ye Guojin <[email protected]>

[ Upstream commit 858779df1c0787d3fec827fb705708df9ebdb15b ]

This was found by coccicheck:
./arch/mips/cavium-octeon/octeon-platform.c, 332, 1-7, ERROR missing put_device; call of_find_device_by_node on line 324, but without a corresponding object release within this function. ./arch/mips/cavium-octeon/octeon-platform.c, 395, 1-7, ERROR missing put_device; call of_find_device_by_node on line 387, but without a corresponding object release within this function. ./arch/mips/cavium-octeon/octeon-usb.c, 512, 3-9, ERROR missing put_device; call of_find_device_by_node on line 515, but without a corresponding object release within this function. ./arch/mips/cavium-octeon/octeon-usb.c, 543, 1-7, ERROR missing put_device; call of_find_device_by_node on line 515, but without a corresponding object release within this function.

---
 arch/mips/cavium-octeon/octeon-platform.c | 2 ++
 arch/mips/cavium-octeon/octeon-usb.c      | 1 +
 2 files changed, 3 insertions(+)
arch/mips/cavium-octeon/octeon-platform.c +2 -0
index d56e9b9d2e434..a994022e32c9f 100644
@@ -328,6 +328,7 @@ static int __init octeon_ehci_device_init(void)
 
 	pd->dev.platform_data = &octeon_ehci_pdata;
 	octeon_ehci_hw_start(&pd->dev);
+	put_device(&pd->dev);
 
 	return ret;
 }
@@ -391,6 +392,7 @@ static int __init octeon_ohci_device_init(void)
 
 	pd->dev.platform_data = &octeon_ohci_pdata;
 	octeon_ohci_hw_start(&pd->dev);
+	put_device(&pd->dev);
 
 	return ret;
 }
arch/mips/cavium-octeon/octeon-usb.c +1 -0
index 6e4d3619137af..4df919d26b082 100644
@@ -537,6 +537,7 @@ static int __init dwc3_octeon_device_init(void)
 			devm_iounmap(&pdev->dev, base);
 			devm_release_mem_region(&pdev->dev, res->start,
 						resource_size(res));
+			put_device(&pdev->dev);
 		}
 	} while (node != NULL);
 
Signature · 1 line
2.34.1
author maintainer
#772 lore raw

PATCH 765/1039 irqchip/gic-v4: Disable redistributors view of the VPE table at boot time

In reply to Greg Kroah-Hartman

From: Marc Zyngier <[email protected]>

[ Upstream commit 79a7f77b9b154d572bd9d2f1eecf58c4d018d8e2 ]

Jay Chen reported that using a kdump kernel on a GICv4.1 system results in a RAS error being delivered when the secondary kernel configures the ITS's view of the new VPE table.

As it turns out, that's because each RD still has a pointer to the previous instance of the VPE table, and that particular implementation is very upset by seeing two bits of the HW that should point to the same table with different values.

To solve this, let's invalidate any reference that any RD has to the VPE table when discovering the RDs. The ITS can then be programmed as expected.

---
 drivers/irqchip/irq-gic-v3.c | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)
drivers/irqchip/irq-gic-v3.c +16 -0
index daec3309b014d..86397522e7864 100644
@@ -920,6 +920,22 @@ static int __gic_update_rdist_properties(struct redist_region *region,
 {
 	u64 typer = gic_read_typer(ptr + GICR_TYPER);
 
+	/* Boot-time cleanip */
+	if ((typer & GICR_TYPER_VLPIS) && (typer & GICR_TYPER_RVPEID)) {
+		u64 val;
+
+		/* Deactivate any present vPE */
+		val = gicr_read_vpendbaser(ptr + SZ_128K + GICR_VPENDBASER);
+		if (val & GICR_VPENDBASER_Valid)
+			gicr_write_vpendbaser(GICR_VPENDBASER_PendingLast,
+					      ptr + SZ_128K + GICR_VPENDBASER);
+
+		/* Mark the VPE table as invalid */
+		val = gicr_read_vpropbaser(ptr + SZ_128K + GICR_VPROPBASER);
+		val &= ~GICR_VPROPBASER_4_1_VALID;
+		gicr_write_vpropbaser(val, ptr + SZ_128K + GICR_VPROPBASER);
+	}
+
 	gic_data.rdists.has_vlpis &= !!(typer & GICR_TYPER_VLPIS);
 
 	/* RVPEID implies some form of DirectLPI, no matter what the doc says... :-/ */
Signature · 1 line
2.34.1
author maintainer
#773 lore raw

PATCH 766/1039 i2c: designware-pci: Fix to change data types of hcnt and lcnt parameters

In reply to Greg Kroah-Hartman

From: Lakshmi Sowjanya D <[email protected]>

[ Upstream commit d52097010078c1844348dc0e467305e5f90fd317 ]

The data type of hcnt and lcnt in the struct dw_i2c_dev is of type u16. It's better to have same data type in struct dw_scl_sda_cfg as well.

---
 drivers/i2c/busses/i2c-designware-pcidrv.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
drivers/i2c/busses/i2c-designware-pcidrv.c +4 -4
index 0f409a4c2da0d..5b45941bcbddc 100644
@@ -39,10 +39,10 @@ enum dw_pci_ctl_id_t {
 };
 
 struct dw_scl_sda_cfg {
-	u32 ss_hcnt;
-	u32 fs_hcnt;
-	u32 ss_lcnt;
-	u32 fs_lcnt;
+	u16 ss_hcnt;
+	u16 fs_hcnt;
+	u16 ss_lcnt;
+	u16 fs_lcnt;
 	u32 sda_hold;
 };
 
Signature · 1 line
2.34.1
author maintainer
#774 lore raw

PATCH 767/1039 scsi: hisi_sas: Prevent parallel FLR and controller reset

In reply to Greg Kroah-Hartman

From: Qi Liu <[email protected]>

[ Upstream commit 16775db613c2bdea09705dcb876942c0641a1098 ]

If we issue a controller reset command during executing a FLR a hung task may be found:

 Call trace:
  __switch_to+0x158/0x1cc
  __schedule+0x2e8/0x85c
  schedule+0x7c/0x110
  schedule_timeout+0x190/0x1cc
  __down+0x7c/0xd4
  down+0x5c/0x7c
  hisi_sas_task_exec+0x510/0x680 [hisi_sas_main]
  hisi_sas_queue_command+0x24/0x30 [hisi_sas_main]
  smp_execute_task_sg+0xf4/0x23c [libsas]
  sas_smp_phy_control+0x110/0x1e0 [libsas]
  transport_sas_phy_reset+0xc8/0x190 [libsas]
  phy_reset_work+0x2c/0x40 [libsas]
  process_one_work+0x1dc/0x48c
  worker_thread+0x15c/0x464
  kthread+0x160/0x170
  ret_from_fork+0x10/0x18

This is a race condition which occurs when the FLR completes first.

Here the host HISI_SAS_RESETTING_BIT flag out gets of sync as HISI_SAS_RESETTING_BIT is not always cleared with the hisi_hba.sem held, so now only set/unset HISI_SAS_RESETTING_BIT under hisi_hba.sem .

---
 drivers/scsi/hisi_sas/hisi_sas_main.c  | 8 +++++---
 drivers/scsi/hisi_sas/hisi_sas_v3_hw.c | 1 +
 2 files changed, 6 insertions(+), 3 deletions(-)
drivers/scsi/hisi_sas/hisi_sas_main.c +5 -3
index f206c433de325..8a13bc08d6575 100644
@@ -1581,7 +1581,6 @@ void hisi_sas_controller_reset_prepare(struct hisi_hba *hisi_hba)
 {
 	struct Scsi_Host *shost = hisi_hba->shost;
 
-	down(&hisi_hba->sem);
 	hisi_hba->phy_state = hisi_hba->hw->get_phys_state(hisi_hba);
 
 	scsi_block_requests(shost);
@@ -1606,9 +1605,9 @@ void hisi_sas_controller_reset_done(struct hisi_hba *hisi_hba)
 	if (hisi_hba->reject_stp_links_msk)
 		hisi_sas_terminate_stp_reject(hisi_hba);
 	hisi_sas_reset_init_all_devices(hisi_hba);
-	up(&hisi_hba->sem);
 	scsi_unblock_requests(shost);
 	clear_bit(HISI_SAS_RESETTING_BIT, &hisi_hba->flags);
+	up(&hisi_hba->sem);
 
 	hisi_sas_rescan_topology(hisi_hba, hisi_hba->phy_state);
 }
@@ -1619,8 +1618,11 @@ static int hisi_sas_controller_prereset(struct hisi_hba *hisi_hba)
 	if (!hisi_hba->hw->soft_reset)
 		return -1;
 
-	if (test_and_set_bit(HISI_SAS_RESETTING_BIT, &hisi_hba->flags))
+	down(&hisi_hba->sem);
+	if (test_and_set_bit(HISI_SAS_RESETTING_BIT, &hisi_hba->flags)) {
+		up(&hisi_hba->sem);
 		return -1;
+	}
 
 	if (hisi_sas_debugfs_enable && hisi_hba->debugfs_itct[0].itct)
 		hisi_hba->hw->debugfs_snapshot_regs(hisi_hba);
drivers/scsi/hisi_sas/hisi_sas_v3_hw.c +1 -0
index 0ef6c21bf0811..11a44d9dd9b2d 100644
@@ -4848,6 +4848,7 @@ static void hisi_sas_reset_prepare_v3_hw(struct pci_dev *pdev)
 	int rc;
 
 	dev_info(dev, "FLR prepare\n");
+	down(&hisi_hba->sem);
 	set_bit(HISI_SAS_RESETTING_BIT, &hisi_hba->flags);
 	hisi_sas_controller_reset_prepare(hisi_hba);
 
Signature · 1 line
2.34.1
author maintainer
#775 lore raw

PATCH 768/1039 ASoC: SOF: ipc: Add null pointer check for substream->runtime

In reply to Greg Kroah-Hartman

From: Ajit Kumar Pandey <[email protected]>

[ Upstream commit 182b682b9ab1348e07ea1bf9d8f2505cc67f9237 ]

When pcm stream is stopped "substream->runtime" pointer will be set to NULL by ALSA core. In case host received an ipc msg from firmware of type IPC_STREAM_POSITION after pcm stream is stopped, there will be kernel NULL pointer exception in ipc_period_elapsed(). This patch fixes it by adding NULL pointer check for "substream->runtime".

---
 sound/soc/sof/ipc.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
sound/soc/sof/ipc.c +2 -1
index e6c53c6c470e4..ca30c506a0fd6 100644
@@ -547,7 +547,8 @@ static void ipc_period_elapsed(struct snd_sof_dev *sdev, u32 msg_id)
 
 	if (spcm->pcm.compress)
 		snd_sof_compr_fragment_elapsed(stream->cstream);
-	else if (!stream->substream->runtime->no_period_wakeup)
+	else if (stream->substream->runtime &&
+		 !stream->substream->runtime->no_period_wakeup)
 		/* only inform ALSA for period_wakeup mode */
 		snd_sof_pcm_period_elapsed(stream->substream);
 }
Signature · 1 line
2.34.1
author maintainer
#776 lore raw

PATCH 769/1039 selftests/powerpc: Add a test of sigreturning to the kernel

In reply to Greg Kroah-Hartman

From: Michael Ellerman <[email protected]>

[ Upstream commit a8968521cfdc3e339fe69473d6632e0aa8d7202a ]

We have a general signal fuzzer, sigfuz, which can modify the MSR & NIP before sigreturn. But the chance of it hitting a kernel address and also clearing MSR_PR is fairly slim.

So add a specific test of sigreturn to a kernel address, both with and without attempting to clear MSR_PR (which the kernel must block).

---
 .../selftests/powerpc/signal/.gitignore       |   1 +
 .../testing/selftests/powerpc/signal/Makefile |   1 +
 .../powerpc/signal/sigreturn_kernel.c         | 132 ++++++++++++++++++
 3 files changed, 134 insertions(+)
 create mode 100644 tools/testing/selftests/powerpc/signal/sigreturn_kernel.c
tools/testing/selftests/powerpc/signal/.gitignore +1 -0
index ce3375cd8e73e..8f6c816099a48 100644
@@ -4,3 +4,4 @@ signal_tm
 sigfuz
 sigreturn_vdso
 sig_sc_double_restart
+sigreturn_kernel
tools/testing/selftests/powerpc/signal/Makefile +1 -0
index d6ae54663aed7..84e201572466d 100644
@@ -1,5 +1,6 @@
 # SPDX-License-Identifier: GPL-2.0
 TEST_GEN_PROGS := signal signal_tm sigfuz sigreturn_vdso sig_sc_double_restart
+TEST_GEN_PROGS += sigreturn_kernel
 
 CFLAGS += -maltivec
 $(OUTPUT)/signal_tm: CFLAGS += -mhtm
tools/testing/selftests/powerpc/signal/sigreturn_kernel.c new file +132 -0

132 lines not on this page · raw message

Signature · 1 line
2.34.1
author maintainer
#777 lore raw

PATCH 770/1039 MIPS: Octeon: Fix build errors using clang

In reply to Greg Kroah-Hartman

From: Tianjia Zhang <[email protected]>

[ Upstream commit 95339b70677dc6f9a2d669c4716058e71b8dc1c7 ]

A large number of the following errors is reported when compiling with clang:

cvmx-bootinfo.h:326:3: error: adding 'int' to a string does not append to the string [-Werror,-Wstring-plus-int]

                  ENUM_BRD_TYPE_CASE(CVMX_BOARD_TYPE_NULL)
                  ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  cvmx-bootinfo.h:321:20: note: expanded from macro 'ENUM_BRD_TYPE_CASE'
          case x: return(#x + 16);        /* Skip CVMX_BOARD_TYPE_ */
                         ~~~^~~~
  cvmx-bootinfo.h:326:3: note: use array indexing to silence this warning
  cvmx-bootinfo.h:321:20: note: expanded from macro 'ENUM_BRD_TYPE_CASE'
          case x: return(#x + 16);        /* Skip CVMX_BOARD_TYPE_ */
                          ^

Follow the prompts to use the address operator '&' to fix this error.

---
 arch/mips/include/asm/octeon/cvmx-bootinfo.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
arch/mips/include/asm/octeon/cvmx-bootinfo.h +2 -2
index 0e6bf220db618..6c61e0a639249 100644
@@ -318,7 +318,7 @@ enum cvmx_chip_types_enum {
 
 /* Functions to return string based on type */
 #define ENUM_BRD_TYPE_CASE(x) \
-	case x: return(#x + 16);	/* Skip CVMX_BOARD_TYPE_ */
+	case x: return (&#x[16]);	/* Skip CVMX_BOARD_TYPE_ */
 static inline const char *cvmx_board_type_to_string(enum
 						    cvmx_board_types_enum type)
 {
@@ -410,7 +410,7 @@ static inline const char *cvmx_board_type_to_string(enum
 }
 
 #define ENUM_CHIP_TYPE_CASE(x) \
-	case x: return(#x + 15);	/* Skip CVMX_CHIP_TYPE */
+	case x: return (&#x[15]);	/* Skip CVMX_CHIP_TYPE */
 static inline const char *cvmx_chip_type_to_string(enum
 						   cvmx_chip_types_enum type)
 {
Signature · 1 line
2.34.1
author maintainer
#778 lore raw

PATCH 771/1039 scsi: sr: Dont use GFP_DMA

In reply to Greg Kroah-Hartman

From: Christoph Hellwig <[email protected]>

[ Upstream commit d94d94969a4ba07a43d62429c60372320519c391 ]

The allocated buffers are used as a command payload, for which the block layer and/or DMA API do the proper bounce buffering if needed.

---
 drivers/scsi/sr.c        | 2 +-
 drivers/scsi/sr_vendor.c | 4 ++--
 2 files changed, 3 insertions(+), 3 deletions(-)
drivers/scsi/sr.c +1 -1
index 8e4af111c0787..f5a2eed543452 100644
@@ -856,7 +856,7 @@ static void get_capabilities(struct scsi_cd *cd)
 
 
 	/* allocate transfer buffer */
-	buffer = kmalloc(512, GFP_KERNEL | GFP_DMA);
+	buffer = kmalloc(512, GFP_KERNEL);
 	if (!buffer) {
 		sr_printk(KERN_ERR, cd, "out of memory.\n");
 		return;
drivers/scsi/sr_vendor.c +2 -2
index 1f988a1b9166f..a61635326ae0a 100644
@@ -131,7 +131,7 @@ int sr_set_blocklength(Scsi_CD *cd, int blocklength)
 	if (cd->vendor == VENDOR_TOSHIBA)
 		density = (blocklength > 2048) ? 0x81 : 0x83;
 
-	buffer = kmalloc(512, GFP_KERNEL | GFP_DMA);
+	buffer = kmalloc(512, GFP_KERNEL);
 	if (!buffer)
 		return -ENOMEM;
 
@@ -179,7 +179,7 @@ int sr_cd_check(struct cdrom_device_info *cdi)
 	if (cd->cdi.mask & CDC_MULTI_SESSION)
 		return 0;
 
-	buffer = kmalloc(512, GFP_KERNEL | GFP_DMA);
+	buffer = kmalloc(512, GFP_KERNEL);
 	if (!buffer)
 		return -ENOMEM;
 
Signature · 1 line
2.34.1
author maintainer
#779 lore raw

PATCH 772/1039 scsi: mpi3mr: Fixes around reply request queues

In reply to Greg Kroah-Hartman

From: Sreekanth Reddy <[email protected]>

[ Upstream commit 243bcc8efdb1f44b1a1d415e6821a246714c68ce ]

Set reply queue depth of 1K for B0 and 4K for A0.

While freeing the segmented request queues use the actual queue depth that is used while creating them.

---
 drivers/scsi/mpi3mr/mpi3mr.h    | 3 ++-
 drivers/scsi/mpi3mr/mpi3mr_fw.c | 4 +++-
 2 files changed, 5 insertions(+), 2 deletions(-)
drivers/scsi/mpi3mr/mpi3mr.h +2 -1
index 9787b53a2b598..2cc42432bd0c0 100644
@@ -79,7 +79,8 @@ extern int prot_mask;
 
 /* Operational queue management definitions */
 #define MPI3MR_OP_REQ_Q_QD		512
-#define MPI3MR_OP_REP_Q_QD		4096
+#define MPI3MR_OP_REP_Q_QD		1024
+#define MPI3MR_OP_REP_Q_QD4K		4096
 #define MPI3MR_OP_REQ_Q_SEG_SIZE	4096
 #define MPI3MR_OP_REP_Q_SEG_SIZE	4096
 #define MPI3MR_MAX_SEG_LIST_SIZE	4096
drivers/scsi/mpi3mr/mpi3mr_fw.c +3 -1
index aa5d877df6f83..2daf633ea2955 100644
@@ -1278,7 +1278,7 @@ static void mpi3mr_free_op_req_q_segments(struct mpi3mr_ioc *mrioc, u16 q_idx)
 			mrioc->op_reply_qinfo[q_idx].q_segment_list = NULL;
 		}
 	} else
-		size = mrioc->req_qinfo[q_idx].num_requests *
+		size = mrioc->req_qinfo[q_idx].segment_qd *
 		    mrioc->facts.op_req_sz;
 
 	for (j = 0; j < mrioc->req_qinfo[q_idx].num_segments; j++) {
@@ -1565,6 +1565,8 @@ static int mpi3mr_create_op_reply_q(struct mpi3mr_ioc *mrioc, u16 qidx)
 
 	reply_qid = qidx + 1;
 	op_reply_q->num_replies = MPI3MR_OP_REP_Q_QD;
+	if (!mrioc->pdev->revision)
+		op_reply_q->num_replies = MPI3MR_OP_REP_Q_QD4K;
 	op_reply_q->ci = 0;
 	op_reply_q->ephase = 1;
 	atomic_set(&op_reply_q->pend_ios, 0);
Signature · 1 line
2.34.1
author maintainer
#780 lore raw

PATCH 773/1039 ASoC: mediatek: mt8192-mt6359: fix device_node leak

In reply to Greg Kroah-Hartman

From: Tzung-Bi Shih <[email protected]>

[ Upstream commit 4e28491a7a198c668437f2be8a91a76aa52f20eb ]

The of_parse_phandle() document:

Quoted 1 line
Quoted 1 line
Quoted 1 line

Use of_node_put() on it when done.

The driver didn't call of_node_put(). Fixes the leak.

---
 sound/soc/mediatek/mt8192/mt8192-mt6359-rt1015-rt5682.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
sound/soc/mediatek/mt8192/mt8192-mt6359-rt1015-rt5682.c +5 -1
index a606133951b70..24a5d0adec1ba 100644
@@ -1172,7 +1172,11 @@ static int mt8192_mt6359_dev_probe(struct platform_device *pdev)
 		return ret;
 	}
 
-	return devm_snd_soc_register_card(&pdev->dev, card);
+	ret = devm_snd_soc_register_card(&pdev->dev, card);
+
+	of_node_put(platform_node);
+	of_node_put(hdmi_codec);
+	return ret;
 }
 
 #ifdef CONFIG_OF
Signature · 1 line
2.34.1
author maintainer
#781 lore raw

PATCH 774/1039 phy: phy-mtk-tphy: add support efuse setting

In reply to Greg Kroah-Hartman

From: Chunfeng Yun <[email protected]>

[ Upstream commit 6f2b033cb883f64ad084a75f13634242c7e179a6 ]

Due to some SoCs have a bit shift issue that will drop a bit for usb3 phy or pcie phy, fix it by adding software efuse reading and setting, but only support it optionally for version 2/3.

---
 drivers/phy/mediatek/phy-mtk-tphy.c | 162 ++++++++++++++++++++++++++++
 1 file changed, 162 insertions(+)
drivers/phy/mediatek/phy-mtk-tphy.c +162 -0
index cdcef865fe9e5..98a942c607a67 100644
@@ -12,6 +12,7 @@
 #include <linux/iopoll.h>
 #include <linux/mfd/syscon.h>
 #include <linux/module.h>
+#include <linux/nvmem-consumer.h>
 #include <linux/of_address.h>
 #include <linux/of_device.h>
 #include <linux/phy/phy.h>
@@ -41,6 +42,9 @@
 #define SSUSB_SIFSLV_V2_U3PHYD		0x200
 #define SSUSB_SIFSLV_V2_U3PHYA		0x400
 
+#define U3P_MISC_REG1		0x04
+#define MR1_EFUSE_AUTO_LOAD_DIS		BIT(6)
+
 #define U3P_USBPHYACR0		0x000
 #define PA0_RG_U2PLL_FORCE_ON		BIT(15)
 #define PA0_USB20_PLL_PREDIV		GENMASK(7, 6)
@@ -133,6 +137,8 @@
 #define P3C_RG_SWRST_U3_PHYD_FORCE_EN	BIT(24)
 
 #define U3P_U3_PHYA_REG0	0x000
+#define P3A_RG_IEXT_INTR		GENMASK(15, 10)
+#define P3A_RG_IEXT_INTR_VAL(x)		((0x3f & (x)) << 10)
 #define P3A_RG_CLKDRV_OFF		GENMASK(3, 2)
 #define P3A_RG_CLKDRV_OFF_VAL(x)	((0x3 & (x)) << 2)
 
@@ -187,6 +193,19 @@
 #define P3D_RG_FWAKE_TH		GENMASK(21, 16)
 #define P3D_RG_FWAKE_TH_VAL(x)	((0x3f & (x)) << 16)
 
+#define U3P_U3_PHYD_IMPCAL0		0x010
+#define P3D_RG_FORCE_TX_IMPEL		BIT(31)
+#define P3D_RG_TX_IMPEL			GENMASK(28, 24)
+#define P3D_RG_TX_IMPEL_VAL(x)		((0x1f & (x)) << 24)
+
+#define U3P_U3_PHYD_IMPCAL1		0x014
+#define P3D_RG_FORCE_RX_IMPEL		BIT(31)
+#define P3D_RG_RX_IMPEL			GENMASK(28, 24)
+#define P3D_RG_RX_IMPEL_VAL(x)		((0x1f & (x)) << 24)
+
+#define U3P_U3_PHYD_RSV			0x054
+#define P3D_RG_EFUSE_AUTO_LOAD_DIS	BIT(12)
+
 #define U3P_U3_PHYD_CDR1		0x05c
 #define P3D_RG_CDR_BIR_LTD1		GENMASK(28, 24)
 #define P3D_RG_CDR_BIR_LTD1_VAL(x)	((0x1f & (x)) << 24)
@@ -307,6 +326,11 @@ struct mtk_phy_pdata {
 	 * 48M PLL, fix it by switching PLL to 26M from default 48M
 	 */
 	bool sw_pll_48m_to_26m;
+	/*
+	 * Some SoCs (e.g. mt8195) drop a bit when use auto load efuse,
+	 * support sw way, also support it for v2/v3 optionally.
+	 */
+	bool sw_efuse_supported;
 	enum mtk_phy_version version;
 };
 
@@ -336,6 +360,10 @@ struct mtk_phy_instance {
 	struct regmap *type_sw;
 	u32 type_sw_reg;
 	u32 type_sw_index;
+	u32 efuse_sw_en;
+	u32 efuse_intr;
+	u32 efuse_tx_imp;
+	u32 efuse_rx_imp;
 	int eye_src;
 	int eye_vrt;
 	int eye_term;
@@ -1040,6 +1068,130 @@ static int phy_type_set(struct mtk_phy_instance *instance)
 	return 0;
 }
 
+static int phy_efuse_get(struct mtk_tphy *tphy, struct mtk_phy_instance *instance)
+{
+	struct device *dev = &instance->phy->dev;
+	int ret = 0;
+
+	/* tphy v1 doesn't support sw efuse, skip it */
+	if (!tphy->pdata->sw_efuse_supported) {
+		instance->efuse_sw_en = 0;
+		return 0;
+	}
+
+	/* software efuse is optional */
+	instance->efuse_sw_en = device_property_read_bool(dev, "nvmem-cells");
+	if (!instance->efuse_sw_en)
+		return 0;
+
+	switch (instance->type) {
+	case PHY_TYPE_USB2:
+		ret = nvmem_cell_read_variable_le_u32(dev, "intr", &instance->efuse_intr);
+		if (ret) {
+			dev_err(dev, "fail to get u2 intr efuse, %d\n", ret);
+			break;
+		}
+
+		/* no efuse, ignore it */
+		if (!instance->efuse_intr) {
+			dev_warn(dev, "no u2 intr efuse, but dts enable it\n");
+			instance->efuse_sw_en = 0;
+			break;
+		}
+
+		dev_dbg(dev, "u2 efuse - intr %x\n", instance->efuse_intr);
+		break;
+
+	case PHY_TYPE_USB3:
+	case PHY_TYPE_PCIE:
+		ret = nvmem_cell_read_variable_le_u32(dev, "intr", &instance->efuse_intr);
+		if (ret) {
+			dev_err(dev, "fail to get u3 intr efuse, %d\n", ret);
+			break;
+		}
+
+		ret = nvmem_cell_read_variable_le_u32(dev, "rx_imp", &instance->efuse_rx_imp);
+		if (ret) {
+			dev_err(dev, "fail to get u3 rx_imp efuse, %d\n", ret);
+			break;
+		}
+
+		ret = nvmem_cell_read_variable_le_u32(dev, "tx_imp", &instance->efuse_tx_imp);
+		if (ret) {
+			dev_err(dev, "fail to get u3 tx_imp efuse, %d\n", ret);
+			break;
+		}
+
+		/* no efuse, ignore it */
+		if (!instance->efuse_intr &&
+		    !instance->efuse_rx_imp &&
+		    !instance->efuse_rx_imp) {
+			dev_warn(dev, "no u3 intr efuse, but dts enable it\n");
+			instance->efuse_sw_en = 0;
+			break;
+		}
+
+		dev_dbg(dev, "u3 efuse - intr %x, rx_imp %x, tx_imp %x\n",
+			instance->efuse_intr, instance->efuse_rx_imp,instance->efuse_tx_imp);
+		break;
+	default:
+		dev_err(dev, "no sw efuse for type %d\n", instance->type);
+		ret = -EINVAL;
+	}
+
+	return ret;
+}
+
+static void phy_efuse_set(struct mtk_phy_instance *instance)
+{
+	struct device *dev = &instance->phy->dev;
+	struct u2phy_banks *u2_banks = &instance->u2_banks;
+	struct u3phy_banks *u3_banks = &instance->u3_banks;
+	u32 tmp;
+
+	if (!instance->efuse_sw_en)
+		return;
+
+	switch (instance->type) {
+	case PHY_TYPE_USB2:
+		tmp = readl(u2_banks->misc + U3P_MISC_REG1);
+		tmp |= MR1_EFUSE_AUTO_LOAD_DIS;
+		writel(tmp, u2_banks->misc + U3P_MISC_REG1);
+
+		tmp = readl(u2_banks->com + U3P_USBPHYACR1);
+		tmp &= ~PA1_RG_INTR_CAL;
+		tmp |= PA1_RG_INTR_CAL_VAL(instance->efuse_intr);
+		writel(tmp, u2_banks->com + U3P_USBPHYACR1);
+		break;
+	case PHY_TYPE_USB3:
+	case PHY_TYPE_PCIE:
+		tmp = readl(u3_banks->phyd + U3P_U3_PHYD_RSV);
+		tmp |= P3D_RG_EFUSE_AUTO_LOAD_DIS;
+		writel(tmp, u3_banks->phyd + U3P_U3_PHYD_RSV);
+
+		tmp = readl(u3_banks->phyd + U3P_U3_PHYD_IMPCAL0);
+		tmp &= ~P3D_RG_TX_IMPEL;
+		tmp |= P3D_RG_TX_IMPEL_VAL(instance->efuse_tx_imp);
+		tmp |= P3D_RG_FORCE_TX_IMPEL;
+		writel(tmp, u3_banks->phyd + U3P_U3_PHYD_IMPCAL0);
+
+		tmp = readl(u3_banks->phyd + U3P_U3_PHYD_IMPCAL1);
+		tmp &= ~P3D_RG_RX_IMPEL;
+		tmp |= P3D_RG_RX_IMPEL_VAL(instance->efuse_rx_imp);
+		tmp |= P3D_RG_FORCE_RX_IMPEL;
+		writel(tmp, u3_banks->phyd + U3P_U3_PHYD_IMPCAL1);
+
+		tmp = readl(u3_banks->phya + U3P_U3_PHYA_REG0);
+		tmp &= ~P3A_RG_IEXT_INTR;
+		tmp |= P3A_RG_IEXT_INTR_VAL(instance->efuse_intr);
+		writel(tmp, u3_banks->phya + U3P_U3_PHYA_REG0);
+		break;
+	default:
+		dev_warn(dev, "no sw efuse for type %d\n", instance->type);
+		break;
+	}
+}
+
 static int mtk_phy_init(struct phy *phy)
 {
 	struct mtk_phy_instance *instance = phy_get_drvdata(phy);
@@ -1050,6 +1202,8 @@ static int mtk_phy_init(struct phy *phy)
 	if (ret)
 		return ret;
 
+	phy_efuse_set(instance);
+
 	switch (instance->type) {
 	case PHY_TYPE_USB2:
 		u2_phy_instance_init(tphy, instance);
@@ -1134,6 +1288,7 @@ static struct phy *mtk_phy_xlate(struct device *dev,
 	struct mtk_phy_instance *instance = NULL;
 	struct device_node *phy_np = args->np;
 	int index;
+	int ret;
 
 	if (args->args_count != 1) {
 		dev_err(dev, "invalid number of cells in 'phy' property\n");
@@ -1174,6 +1329,10 @@ static struct phy *mtk_phy_xlate(struct device *dev,
 		return ERR_PTR(-EINVAL);
 	}
 
+	ret = phy_efuse_get(tphy, instance);
+	if (ret)
+		return ERR_PTR(ret);
+
 	phy_parse_property(tphy, instance);
 	phy_type_set(instance);
 
@@ -1196,10 +1355,12 @@ static const struct mtk_phy_pdata tphy_v1_pdata = {
 
 static const struct mtk_phy_pdata tphy_v2_pdata = {
 	.avoid_rx_sen_degradation = false,
+	.sw_efuse_supported = true,
 	.version = MTK_PHY_V2,
 };
 
 static const struct mtk_phy_pdata tphy_v3_pdata = {
+	.sw_efuse_supported = true,
 	.version = MTK_PHY_V3,
 };
 
@@ -1210,6 +1371,7 @@ static const struct mtk_phy_pdata mt8173_pdata = {
 
 static const struct mtk_phy_pdata mt8195_pdata = {
 	.sw_pll_48m_to_26m = true,
+	.sw_efuse_supported = true,
 	.version = MTK_PHY_V3,
 };
 
Signature · 1 line
2.34.1
author maintainer
#782 lore raw

PATCH 775/1039 ASoC: mediatek: mt8173: fix device_node leak

In reply to Greg Kroah-Hartman

From: Tzung-Bi Shih <[email protected]>

[ Upstream commit 493433785df0075afc0c106ab65f10a605d0b35d ]

Fixes the device_node leak.

---
 sound/soc/mediatek/mt8173/mt8173-max98090.c      | 3 +++
 sound/soc/mediatek/mt8173/mt8173-rt5650-rt5514.c | 2 ++
 sound/soc/mediatek/mt8173/mt8173-rt5650-rt5676.c | 2 ++
 sound/soc/mediatek/mt8173/mt8173-rt5650.c        | 2 ++
 4 files changed, 9 insertions(+)
sound/soc/mediatek/mt8173/mt8173-max98090.c +3 -0
index fc94314bfc02f..3bdd4931316cd 100644
@@ -180,6 +180,9 @@ static int mt8173_max98090_dev_probe(struct platform_device *pdev)
 	if (ret)
 		dev_err(&pdev->dev, "%s snd_soc_register_card fail %d\n",
 			__func__, ret);
+
+	of_node_put(codec_node);
+	of_node_put(platform_node);
 	return ret;
 }
 
sound/soc/mediatek/mt8173/mt8173-rt5650-rt5514.c +2 -0
index 0f28dc2217c09..390da5bf727eb 100644
@@ -218,6 +218,8 @@ static int mt8173_rt5650_rt5514_dev_probe(struct platform_device *pdev)
 	if (ret)
 		dev_err(&pdev->dev, "%s snd_soc_register_card fail %d\n",
 			__func__, ret);
+
+	of_node_put(platform_node);
 	return ret;
 }
 
sound/soc/mediatek/mt8173/mt8173-rt5650-rt5676.c +2 -0
index 077c6ee067806..c8e4e85e10575 100644
@@ -285,6 +285,8 @@ static int mt8173_rt5650_rt5676_dev_probe(struct platform_device *pdev)
 	if (ret)
 		dev_err(&pdev->dev, "%s snd_soc_register_card fail %d\n",
 			__func__, ret);
+
+	of_node_put(platform_node);
 	return ret;
 }
 
sound/soc/mediatek/mt8173/mt8173-rt5650.c +2 -0
index 2cbf679f5c74b..d8cf0802813a0 100644
@@ -323,6 +323,8 @@ static int mt8173_rt5650_dev_probe(struct platform_device *pdev)
 	if (ret)
 		dev_err(&pdev->dev, "%s snd_soc_register_card fail %d\n",
 			__func__, ret);
+
+	of_node_put(platform_node);
 	return ret;
 }
 
Signature · 1 line
2.34.1
author maintainer
#783 lore raw

PATCH 776/1039 ASoC: mediatek: mt8183: fix device_node leak

In reply to Greg Kroah-Hartman

From: Tzung-Bi Shih <[email protected]>

[ Upstream commit cb006006fe6221f092fadaffd3f219288304c9ad ]

Fixes the device_node leak.

---
 sound/soc/mediatek/mt8183/mt8183-da7219-max98357.c         | 6 +++++-
 sound/soc/mediatek/mt8183/mt8183-mt6358-ts3a227-max98357.c | 7 ++++++-
 2 files changed, 11 insertions(+), 2 deletions(-)
sound/soc/mediatek/mt8183/mt8183-da7219-max98357.c +5 -1
index a4d26a6fc8492..bda103211e0bd 100644
@@ -781,7 +781,11 @@ static int mt8183_da7219_max98357_dev_probe(struct platform_device *pdev)
 		return ret;
 	}
 
-	return devm_snd_soc_register_card(&pdev->dev, card);
+	ret = devm_snd_soc_register_card(&pdev->dev, card);
+
+	of_node_put(platform_node);
+	of_node_put(hdmi_codec);
+	return ret;
 }
 
 #ifdef CONFIG_OF
sound/soc/mediatek/mt8183/mt8183-mt6358-ts3a227-max98357.c +6 -1
index aeb1af86047ef..9f0bf15fe465e 100644
@@ -780,7 +780,12 @@ mt8183_mt6358_ts3a227_max98357_dev_probe(struct platform_device *pdev)
 				 __func__, ret);
 	}
 
-	return devm_snd_soc_register_card(&pdev->dev, card);
+	ret = devm_snd_soc_register_card(&pdev->dev, card);
+
+	of_node_put(platform_node);
+	of_node_put(ec_codec);
+	of_node_put(hdmi_codec);
+	return ret;
 }
 
 #ifdef CONFIG_OF
Signature · 1 line
2.34.1
author maintainer
#784 lore raw

PATCH 777/1039 habanalabs: change wait for interrupt timeout to 64 bit

In reply to Greg Kroah-Hartman

From: Dani Liberman <[email protected]>

[ Upstream commit 48f31169830f589e4c7ac475ccc7414951ded3f0 ]

In order to increase maximum wait-for-interrupt timeout, change it to 64 bit variable. This wait is used only by newer ASICs, so no problem in changing this interface at this time.

---
 .../habanalabs/common/command_submission.c    | 22 ++++++++++++++-----
 include/uapi/misc/habanalabs.h                | 18 +++++++++------
 2 files changed, 28 insertions(+), 12 deletions(-)
drivers/misc/habanalabs/common/command_submission.c +17 -5
index 4c8000fd246cd..9451e4bae05df 100644
@@ -2765,8 +2765,23 @@ static int hl_cs_wait_ioctl(struct hl_fpriv *hpriv, void *data)
 	return 0;
 }
 
+static inline unsigned long hl_usecs64_to_jiffies(const u64 usecs)
+{
+	if (usecs <= U32_MAX)
+		return usecs_to_jiffies(usecs);
+
+	/*
+	 * If the value in nanoseconds is larger than 64 bit, use the largest
+	 * 64 bit value.
+	 */
+	if (usecs >= ((u64)(U64_MAX / NSEC_PER_USEC)))
+		return nsecs_to_jiffies(U64_MAX);
+
+	return nsecs_to_jiffies(usecs * NSEC_PER_USEC);
+}
+
 static int _hl_interrupt_wait_ioctl(struct hl_device *hdev, struct hl_ctx *ctx,
-				u32 timeout_us, u64 user_address,
+				u64 timeout_us, u64 user_address,
 				u64 target_value, u16 interrupt_offset,
 				enum hl_cs_wait_status *status,
 				u64 *timestamp)
@@ -2778,10 +2793,7 @@ static int _hl_interrupt_wait_ioctl(struct hl_device *hdev, struct hl_ctx *ctx,
 	long completion_rc;
 	int rc = 0;
 
-	if (timeout_us == U32_MAX)
-		timeout = timeout_us;
-	else
-		timeout = usecs_to_jiffies(timeout_us);
+	timeout = hl_usecs64_to_jiffies(timeout_us);
 
 	hl_ctx_get(hdev, ctx);
 
include/uapi/misc/habanalabs.h +11 -7
index 00b3095904995..c5760acebdd1d 100644
@@ -911,14 +911,18 @@ struct hl_wait_cs_in {
 	 */
 	__u32 flags;
 
-	/* Multi CS API info- valid entries in multi-CS array */
-	__u8 seq_arr_len;
-	__u8 pad[3];
+	union {
+		struct {
+			/* Multi CS API info- valid entries in multi-CS array */
+			__u8 seq_arr_len;
+			__u8 pad[7];
+		};
 
-	/* Absolute timeout to wait for an interrupt in microseconds.
-	 * Relevant only when HL_WAIT_CS_FLAGS_INTERRUPT is set
-	 */
-	__u32 interrupt_timeout_us;
+		/* Absolute timeout to wait for an interrupt in microseconds.
+		 * Relevant only when HL_WAIT_CS_FLAGS_INTERRUPT is set
+		 */
+		__u64 interrupt_timeout_us;
+	};
 };
 
 #define HL_WAIT_CS_STATUS_COMPLETED	0
Signature · 1 line
2.34.1
author maintainer
#785 lore raw

PATCH 778/1039 habanalabs: skip read fw errors if dynamic descriptor invalid

In reply to Greg Kroah-Hartman

From: Ohad Sharabi <[email protected]>

[ Upstream commit 4fac990f604e6c10538026835a8a30f3c1b6fcf5 ]

Reporting FW errors involves reading of the error registers.

In case we have a corrupted FW descriptor we cannot do that since the dynamic scratchpad is potentially corrupted as well and may cause kernel crush when attempting access to a corrupted register offset.

---
 drivers/misc/habanalabs/common/firmware_if.c | 17 +++++++++++++++--
 drivers/misc/habanalabs/common/habanalabs.h  |  2 ++
 2 files changed, 17 insertions(+), 2 deletions(-)
drivers/misc/habanalabs/common/firmware_if.c +15 -2
index 4e68fb9d2a6bd..67a0be4573710 100644
@@ -1703,6 +1703,9 @@ static int hl_fw_dynamic_validate_descriptor(struct hl_device *hdev,
 		return rc;
 	}
 
+	/* here we can mark the descriptor as valid as the content has been validated */
+	fw_loader->dynamic_loader.fw_desc_valid = true;
+
 	return 0;
 }
 
@@ -1759,7 +1762,13 @@ static int hl_fw_dynamic_read_and_validate_descriptor(struct hl_device *hdev,
 		return rc;
 	}
 
-	/* extract address copy the descriptor from */
+	/*
+	 * extract address to copy the descriptor from
+	 * in addition, as the descriptor value is going to be over-ridden by new data- we mark it
+	 * as invalid.
+	 * it will be marked again as valid once validated
+	 */
+	fw_loader->dynamic_loader.fw_desc_valid = false;
 	src = hdev->pcie_bar[region->bar_id] + region->offset_in_bar +
 							response->ram_offset;
 	memcpy_fromio(fw_desc, src, sizeof(struct lkd_fw_comms_desc));
@@ -2247,6 +2256,9 @@ static int hl_fw_dynamic_init_cpu(struct hl_device *hdev,
 	dev_info(hdev->dev,
 		"Loading firmware to device, may take some time...\n");
 
+	/* initialize FW descriptor as invalid */
+	fw_loader->dynamic_loader.fw_desc_valid = false;
+
 	/*
 	 * In this stage, "cpu_dyn_regs" contains only LKD's hard coded values!
 	 * It will be updated from FW after hl_fw_dynamic_request_descriptor().
@@ -2333,7 +2345,8 @@ static int hl_fw_dynamic_init_cpu(struct hl_device *hdev,
 	return 0;
 
 protocol_err:
-	fw_read_errors(hdev, le32_to_cpu(dyn_regs->cpu_boot_err0),
+	if (fw_loader->dynamic_loader.fw_desc_valid)
+		fw_read_errors(hdev, le32_to_cpu(dyn_regs->cpu_boot_err0),
 				le32_to_cpu(dyn_regs->cpu_boot_err1),
 				le32_to_cpu(dyn_regs->cpu_boot_dev_sts0),
 				le32_to_cpu(dyn_regs->cpu_boot_dev_sts1));
drivers/misc/habanalabs/common/habanalabs.h +2 -0
index a2002cbf794b5..ba0965667b182 100644
@@ -1010,6 +1010,7 @@ struct fw_response {
  * @image_region: region to copy the FW image to
  * @fw_image_size: size of FW image to load
  * @wait_for_bl_timeout: timeout for waiting for boot loader to respond
+ * @fw_desc_valid: true if FW descriptor has been validated and hence the data can be used
  */
 struct dynamic_fw_load_mgr {
 	struct fw_response response;
@@ -1017,6 +1018,7 @@ struct dynamic_fw_load_mgr {
 	struct pci_mem_region *image_region;
 	size_t fw_image_size;
 	u32 wait_for_bl_timeout;
+	bool fw_desc_valid;
 };
 
 /**
Signature · 1 line
2.34.1
author maintainer
#786 lore raw

PATCH 779/1039 phy: mediatek: Fix missing check in mtk_mipi_tx_probe

In reply to Greg Kroah-Hartman

From: Miaoqian Lin <[email protected]>

[ Upstream commit 399c91c3f30531593e5ff6ca7b53f47092128669 ]

The of_device_get_match_data() function may return NULL. Add check to prevent potential null dereference.

---
 drivers/phy/mediatek/phy-mtk-mipi-dsi.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/phy/mediatek/phy-mtk-mipi-dsi.c +2 -0
index 28ad9403c4414..67b005d5b9e35 100644
@@ -146,6 +146,8 @@ static int mtk_mipi_tx_probe(struct platform_device *pdev)
 		return -ENOMEM;
 
 	mipi_tx->driver_data = of_device_get_match_data(dev);
+	if (!mipi_tx->driver_data)
+		return -ENODEV;
 
 	mipi_tx->regs = devm_platform_ioremap_resource(pdev, 0);
 	if (IS_ERR(mipi_tx->regs))
Signature · 1 line
2.34.1
author maintainer
#787 lore raw

PATCH 780/1039 mailbox: change mailbox-mpfs compatible string

In reply to Greg Kroah-Hartman

From: Conor Dooley <[email protected]>

[ Upstream commit f10b1fc0161cd99e54c5687fcc63368aa255e05e ]

The Polarfire SoC is currently using two different compatible string prefixes. Fix this by changing "polarfire-soc-*" strings to "mpfs-*" in its system controller in order to match the compatible string used in the soc binding and device tree.

---
 drivers/mailbox/mailbox-mpfs.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/mailbox/mailbox-mpfs.c +1 -1
index 0d6e2231a2c75..4e34854d12389 100644
@@ -232,7 +232,7 @@ static int mpfs_mbox_probe(struct platform_device *pdev)
 }
 
 static const struct of_device_id mpfs_mbox_of_match[] = {
-	{.compatible = "microchip,polarfire-soc-mailbox", },
+	{.compatible = "microchip,mpfs-mailbox", },
 	{},
 };
 MODULE_DEVICE_TABLE(of, mpfs_mbox_of_match);
Signature · 1 line
2.34.1
author maintainer
#788 lore raw

PATCH 781/1039 signal: In get_signal test for signal_group_exit every time through the loop

In reply to Greg Kroah-Hartman

From: Eric W. Biederman <[email protected]>

[ Upstream commit e7f7c99ba911f56bc338845c1cd72954ba591707 ]

Recently while investigating a problem with rr and signals I noticed that siglock is dropped in ptrace_signal and get_signal does not jump to relock.

Looking farther to see if the problem is anywhere else I see that do_signal_stop also returns if signal_group_exit is true. I believe that test can now never be true, but it is a bit hard to trace through and be certain.

Testing signal_group_exit is not expensive, so move the test for signal_group_exit into the for loop inside of get_signal to ensure the test is never skipped improperly.

This has been a potential problem since I added the test for signal_group_exit was added.

Fixes 35634ffa1751 ("signal: Always notice exiting tasks")
---
 kernel/signal.c | 20 ++++++++++----------
 1 file changed, 10 insertions(+), 10 deletions(-)
kernel/signal.c +10 -10
index dfcee3888b00e..cf97b9c4d665a 100644
@@ -2684,19 +2684,19 @@ relock:
 		goto relock;
 	}
 
-	/* Has this task already been marked for death? */
-	if (signal_group_exit(signal)) {
-		ksig->info.si_signo = signr = SIGKILL;
-		sigdelset(&current->pending.signal, SIGKILL);
-		trace_signal_deliver(SIGKILL, SEND_SIG_NOINFO,
-				&sighand->action[SIGKILL - 1]);
-		recalc_sigpending();
-		goto fatal;
-	}
-
 	for (;;) {
 		struct k_sigaction *ka;
 
+		/* Has this task already been marked for death? */
+		if (signal_group_exit(signal)) {
+			ksig->info.si_signo = signr = SIGKILL;
+			sigdelset(&current->pending.signal, SIGKILL);
+			trace_signal_deliver(SIGKILL, SEND_SIG_NOINFO,
+				&sighand->action[SIGKILL - 1]);
+			recalc_sigpending();
+			goto fatal;
+		}
+
 		if (unlikely(current->jobctl & JOBCTL_STOP_PENDING) &&
 		    do_signal_stop(0))
 			goto relock;
Signature · 1 line
2.34.1
author maintainer
#789 lore raw

PATCH 782/1039 PCI: mediatek-gen3: Disable DVFSRC voltage request

In reply to Greg Kroah-Hartman

From: Jianjun Wang <[email protected]>

[ Upstream commit ab344fd43f2958726d17d651c0cb692c67dca382 ]

When the DVFSRC (dynamic voltage and frequency scaling resource collector) feature is not implemented, the PCIe hardware will assert a voltage request signal when exit from the L1 PM Substates to request a specific Vcore voltage, but cannot receive the voltage ready signal, which will cause the link to fail to exit the L1 PM Substates.

Disable DVFSRC voltage request by default, we need to find a common way to enable it in the future.

---
 drivers/pci/controller/pcie-mediatek-gen3.c | 8 ++++++++
 1 file changed, 8 insertions(+)
drivers/pci/controller/pcie-mediatek-gen3.c +8 -0
index 17c59b0d6978b..21207df680ccf 100644
@@ -79,6 +79,9 @@
 #define PCIE_ICMD_PM_REG		0x198
 #define PCIE_TURN_OFF_LINK		BIT(4)
 
+#define PCIE_MISC_CTRL_REG		0x348
+#define PCIE_DISABLE_DVFSRC_VLT_REQ	BIT(1)
+
 #define PCIE_TRANS_TABLE_BASE_REG	0x800
 #define PCIE_ATR_SRC_ADDR_MSB_OFFSET	0x4
 #define PCIE_ATR_TRSL_ADDR_LSB_OFFSET	0x8
@@ -297,6 +300,11 @@ static int mtk_pcie_startup_port(struct mtk_pcie_port *port)
 	val &= ~PCIE_INTX_ENABLE;
 	writel_relaxed(val, port->base + PCIE_INT_ENABLE_REG);
 
+	/* Disable DVFSRC voltage request */
+	val = readl_relaxed(port->base + PCIE_MISC_CTRL_REG);
+	val |= PCIE_DISABLE_DVFSRC_VLT_REQ;
+	writel_relaxed(val, port->base + PCIE_MISC_CTRL_REG);
+
 	/* Assert all reset signals */
 	val = readl_relaxed(port->base + PCIE_RST_CTRL_REG);
 	val |= PCIE_MAC_RSTB | PCIE_PHY_RSTB | PCIE_BRG_RSTB | PCIE_PE_RSTB;
Signature · 1 line
2.34.1
author maintainer
#790 lore raw

PATCH 783/1039 PCI: qcom: Fix an error handling path in qcom_pcie_probe()

In reply to Greg Kroah-Hartman

From: Christophe JAILLET <[email protected]>

[ Upstream commit 4e0e90539bb0e6c0ca3768c642df9eed2118a8bb ]

If 'of_device_get_match_data()' fails, previous 'pm_runtime_get_sync()/ pm_runtime_enable()' should be undone.

To fix it, the easiest is to move this block of code before the memory allocations and the pm_runtime_xxx calls.

---
 drivers/pci/controller/dwc/pcie-qcom.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)
drivers/pci/controller/dwc/pcie-qcom.c +6 -6
index 1c3d1116bb60c..baae67f71ba82 100644
@@ -1534,6 +1534,12 @@ static int qcom_pcie_probe(struct platform_device *pdev)
 	const struct qcom_pcie_cfg *pcie_cfg;
 	int ret;
 
+	pcie_cfg = of_device_get_match_data(dev);
+	if (!pcie_cfg || !pcie_cfg->ops) {
+		dev_err(dev, "Invalid platform data\n");
+		return -EINVAL;
+	}
+
 	pcie = devm_kzalloc(dev, sizeof(*pcie), GFP_KERNEL);
 	if (!pcie)
 		return -ENOMEM;
@@ -1553,12 +1559,6 @@ static int qcom_pcie_probe(struct platform_device *pdev)
 
 	pcie->pci = pci;
 
-	pcie_cfg = of_device_get_match_data(dev);
-	if (!pcie_cfg || !pcie_cfg->ops) {
-		dev_err(dev, "Invalid platform data\n");
-		return -EINVAL;
-	}
-
 	pcie->ops = pcie_cfg->ops;
 	pcie->pipe_clk_need_muxing = pcie_cfg->pipe_clk_need_muxing;
 
Signature · 1 line
2.34.1
author maintainer
#791 lore raw

PATCH 784/1039 PCI: rcar: Check if device is runtime suspended instead of __clk_is_enabled()

In reply to Greg Kroah-Hartman

From: Marek Vasut <[email protected]>

[ Upstream commit d2a14b54989e9ccea8401895fdfbc213bd1f56af ]

Replace __clk_is_enabled() with pm_runtime_suspended(),
as __clk_is_enabled() was checking the wrong bus clock
and caused the following build error too:
arm-linux-gnueabi-ld: drivers/pci/controller/pcie-rcar-host.o: in function `rcar_pcie_aarch32_abort_handler':
pcie-rcar-host.c:(.text+0xdd0): undefined reference to `__clk_is_enabled'

Fixes a115b1bd3af0 ("PCI: rcar: Add L1 link state fix into data abort hook")
Cc Arnd Bergmann <[email protected]>
Cc Bjorn Helgaas <[email protected]>
Cc Geert Uytterhoeven <[email protected]>
Cc Lorenzo Pieralisi <[email protected]>
Cc Stephen Boyd <[email protected]>
Cc Wolfram Sang <[email protected]>
Cc Yoshihiro Shimoda <[email protected]>
---
 drivers/pci/controller/pcie-rcar-host.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)
drivers/pci/controller/pcie-rcar-host.c +5 -5
index e12c2d8be05a3..780e60159993c 100644
@@ -50,10 +50,10 @@ struct rcar_msi {
  */
 static void __iomem *pcie_base;
 /*
- * Static copy of bus clock pointer, so we can check whether the clock
- * is enabled or not.
+ * Static copy of PCIe device pointer, so we can check whether the
+ * device is runtime suspended or not.
  */
-static struct clk *pcie_bus_clk;
+static struct device *pcie_dev;
 #endif
 
 /* Structure representing the PCIe interface */
@@ -792,7 +792,7 @@ static int rcar_pcie_get_resources(struct rcar_pcie_host *host)
 #ifdef CONFIG_ARM
 	/* Cache static copy for L1 link state fixup hook on aarch32 */
 	pcie_base = pcie->base;
-	pcie_bus_clk = host->bus_clk;
+	pcie_dev = pcie->dev;
 #endif
 
 	return 0;
@@ -1062,7 +1062,7 @@ static int rcar_pcie_aarch32_abort_handler(unsigned long addr,
 
 	spin_lock_irqsave(&pmsr_lock, flags);
 
-	if (!pcie_base || !__clk_is_enabled(pcie_bus_clk)) {
+	if (!pcie_base || pm_runtime_suspended(pcie_dev)) {
 		ret = 1;
 		goto unlock_exit;
 	}
Signature · 1 line
2.34.1
author maintainer
#792 lore raw

PATCH 785/1039 PCI: dwc: Do not remap invalid res

In reply to Greg Kroah-Hartman

From: Tim Harvey <[email protected]>

[ Upstream commit 6e5ebc96ec651b67131f816d7e3bf286c635e749 ]

On imx6 and perhaps others when pcie probes you get a:
imx6q-pcie 33800000.pcie: invalid resource

This occurs because the atu is not specified in the DT and as such it should not be remapped.

---
 drivers/pci/controller/dwc/pcie-designware.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)
drivers/pci/controller/dwc/pcie-designware.c +4 -3
index 850b4533f4ef5..d92c8a25094fa 100644
@@ -672,10 +672,11 @@ void dw_pcie_iatu_detect(struct dw_pcie *pci)
 		if (!pci->atu_base) {
 			struct resource *res =
 				platform_get_resource_byname(pdev, IORESOURCE_MEM, "atu");
-			if (res)
+			if (res) {
 				pci->atu_size = resource_size(res);
-			pci->atu_base = devm_ioremap_resource(dev, res);
-			if (IS_ERR(pci->atu_base))
+				pci->atu_base = devm_ioremap_resource(dev, res);
+			}
+			if (!pci->atu_base || IS_ERR(pci->atu_base))
 				pci->atu_base = pci->dbi_base + DEFAULT_DBI_ATU_OFFSET;
 		}
 
Signature · 1 line
2.34.1
author maintainer
#793 lore raw

PATCH 786/1039 PCI: aardvark: Fix checking for MEM resource type

In reply to Greg Kroah-Hartman

From: Pali Rohár <[email protected]>

[ Upstream commit 2070b2ddea89f5b604fac3d27ade5cb6d19a5706 ]

IORESOURCE_MEM_64 is not a resource type but a type flag.

Remove incorrect check for type IORESOURCE_MEM_64.

Fixes 64f160e19e92 ("PCI: aardvark: Configure PCIe resources from 'ranges' DT property")
---
 drivers/pci/controller/pci-aardvark.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)
drivers/pci/controller/pci-aardvark.c +2 -4
index c3b725afa11fd..e3001b3b32930 100644
@@ -1535,8 +1535,7 @@ static int advk_pcie_probe(struct platform_device *pdev)
 		 * only PIO for issuing configuration transfers which does
 		 * not use PCIe window configuration.
 		 */
-		if (type != IORESOURCE_MEM && type != IORESOURCE_MEM_64 &&
-		    type != IORESOURCE_IO)
+		if (type != IORESOURCE_MEM && type != IORESOURCE_IO)
 			continue;
 
 		/*
@@ -1544,8 +1543,7 @@ static int advk_pcie_probe(struct platform_device *pdev)
 		 * configuration is set to transparent memory access so it
 		 * does not need window configuration.
 		 */
-		if ((type == IORESOURCE_MEM || type == IORESOURCE_MEM_64) &&
-		    entry->offset == 0)
+		if (type == IORESOURCE_MEM && entry->offset == 0)
 			continue;
 
 		/*
Signature · 1 line
2.34.1
author maintainer
#794 lore raw

PATCH 787/1039 PCI: apple: Fix REFCLK1 enable/poll logic

In reply to Greg Kroah-Hartman

From: Hector Martin <[email protected]>

[ Upstream commit 75d36df6807838f3c826c21c0fa51cdc079667d1 ]

REFCLK1 has req/ack bits that need to be programmed, just like REFCLK0.

---
 drivers/pci/controller/pcie-apple.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)
drivers/pci/controller/pcie-apple.c +4 -3
index b090924b41fee..537ac9fe2e842 100644
@@ -42,8 +42,9 @@
 #define   CORE_FABRIC_STAT_MASK		0x001F001F
 #define CORE_LANE_CFG(port)		(0x84000 + 0x4000 * (port))
 #define   CORE_LANE_CFG_REFCLK0REQ	BIT(0)
-#define   CORE_LANE_CFG_REFCLK1		BIT(1)
+#define   CORE_LANE_CFG_REFCLK1REQ	BIT(1)
 #define   CORE_LANE_CFG_REFCLK0ACK	BIT(2)
+#define   CORE_LANE_CFG_REFCLK1ACK	BIT(3)
 #define   CORE_LANE_CFG_REFCLKEN	(BIT(9) | BIT(10))
 #define CORE_LANE_CTL(port)		(0x84004 + 0x4000 * (port))
 #define   CORE_LANE_CTL_CFGACC		BIT(15)
@@ -482,9 +483,9 @@ static int apple_pcie_setup_refclk(struct apple_pcie *pcie,
 	if (res < 0)
 		return res;
 
-	rmw_set(CORE_LANE_CFG_REFCLK1, pcie->base + CORE_LANE_CFG(port->idx));
+	rmw_set(CORE_LANE_CFG_REFCLK1REQ, pcie->base + CORE_LANE_CFG(port->idx));
 	res = readl_relaxed_poll_timeout(pcie->base + CORE_LANE_CFG(port->idx),
-					 stat, stat & CORE_LANE_CFG_REFCLK1,
+					 stat, stat & CORE_LANE_CFG_REFCLK1ACK,
 					 100, 50000);
 
 	if (res < 0)
Signature · 1 line
2.34.1
author maintainer
#795 lore raw

PATCH 788/1039 KVM: VMX: Dont unblock vCPU w/ Posted IRQ if IRQs are disabled in guest

In reply to Greg Kroah-Hartman

From: Paolo Bonzini <[email protected]>

[ Upstream commit 1831fa44df743a7cdffdf1c12c799bf6f3c12b8c ]

Don't configure the wakeup handler when a vCPU is blocking with IRQs disabled, in which case any IRQ, posted or otherwise, should not be recognized and thus should not wake the vCPU.

Fixes bf9f6ac8d749 ("KVM: Update Posted-Interrupts Descriptor when vCPU is blocked")

Message-Id: <[email protected]>

---
 arch/x86/kvm/vmx/posted_intr.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
arch/x86/kvm/vmx/posted_intr.c +2 -1
index 21ea58d25771f..696ad48ab5daa 100644
@@ -147,7 +147,8 @@ int pi_pre_block(struct kvm_vcpu *vcpu)
 	struct pi_desc old, new;
 	struct pi_desc *pi_desc = vcpu_to_pi_desc(vcpu);
 
-	if (!vmx_can_use_vtd_pi(vcpu->kvm))
+	if (!vmx_can_use_vtd_pi(vcpu->kvm) ||
+	    vmx_interrupt_blocked(vcpu))
 		return 0;
 
 	WARN_ON(irqs_disabled());
Signature · 1 line
2.34.1
author maintainer
#796 lore raw

PATCH 789/1039 KVM: s390: Ensure kvm_arch_no_poll() is read once when blocking vCPU

In reply to Greg Kroah-Hartman

From: Sean Christopherson <[email protected]>

[ Upstream commit 6f390916c4fb359507d9ac4bf1b28a4f8abee5c0 ]

Wrap s390's halt_poll_max_steal with READ_ONCE and snapshot the result of kvm_arch_no_poll() in kvm_vcpu_block() to avoid a mostly-theoretical, largely benign bug on s390 where the result of kvm_arch_no_poll() could change due to userspace modifying halt_poll_max_steal while the vCPU is blocking. The bug is largely benign as it will either cause KVM to skip updating halt-polling times (no_poll toggles false=>true) or to update halt-polling times with a slightly flawed block_ns.

Note, READ_ONCE is unnecessary in the current code, add it in case the arch hook is ever inlined, and to provide a hint that userspace can change the param at will.

Fixes 8b905d28ee17 ("KVM: s390: provide kvm_arch_no_poll function")

Message-Id: <[email protected]>

---
 arch/s390/kvm/kvm-s390.c | 2 +-
 virt/kvm/kvm_main.c      | 5 +++--
 2 files changed, 4 insertions(+), 3 deletions(-)
arch/s390/kvm/kvm-s390.c +1 -1
index ef299aad40090..8fc9c79c899b5 100644
@@ -3449,7 +3449,7 @@ bool kvm_arch_no_poll(struct kvm_vcpu *vcpu)
 {
 	/* do not poll with more than halt_poll_max_steal percent of steal time */
 	if (S390_lowcore.avg_steal_timer * 100 / (TICK_USEC << 12) >=
-	    halt_poll_max_steal) {
+	    READ_ONCE(halt_poll_max_steal)) {
 		vcpu->stat.halt_no_poll_steal++;
 		return true;
 	}
virt/kvm/kvm_main.c +3 -2
index 72c4e6b393896..5bd62342c482b 100644
@@ -3153,6 +3153,7 @@ update_halt_poll_stats(struct kvm_vcpu *vcpu, u64 poll_ns, bool waited)
  */
 void kvm_vcpu_block(struct kvm_vcpu *vcpu)
 {
+	bool halt_poll_allowed = !kvm_arch_no_poll(vcpu);
 	ktime_t start, cur, poll_end;
 	bool waited = false;
 	u64 block_ns;
@@ -3160,7 +3161,7 @@ void kvm_vcpu_block(struct kvm_vcpu *vcpu)
 	kvm_arch_vcpu_blocking(vcpu);
 
 	start = cur = poll_end = ktime_get();
-	if (vcpu->halt_poll_ns && !kvm_arch_no_poll(vcpu)) {
+	if (vcpu->halt_poll_ns && halt_poll_allowed) {
 		ktime_t stop = ktime_add_ns(ktime_get(), vcpu->halt_poll_ns);
 
 		++vcpu->stat.generic.halt_attempted_poll;
@@ -3215,7 +3216,7 @@ out:
 	update_halt_poll_stats(
 		vcpu, ktime_to_ns(ktime_sub(poll_end, start)), waited);
 
-	if (!kvm_arch_no_poll(vcpu)) {
+	if (halt_poll_allowed) {
 		if (!vcpu_valid_wakeup(vcpu)) {
 			shrink_halt_poll_ns(vcpu);
 		} else if (vcpu->kvm->max_halt_poll_ns) {
Signature · 1 line
2.34.1
author maintainer
#797 lore raw

PATCH 790/1039 KVM: VMX: Read Posted Interrupt "control" exactly once per loop iteration

In reply to Greg Kroah-Hartman

From: Sean Christopherson <[email protected]>

[ Upstream commit cfb0e1306a3790eb055ebf7cdb7b0ee8a23e9b6e ]

Use READ_ONCE() when loading the posted interrupt descriptor control field to ensure "old" and "new" have the same base value. If the compiler emits separate loads, and loads into "new" before "old", KVM could theoretically drop the ON bit if it were set between the loads.

Fixes 28b835d60fcc ("KVM: Update Posted-Interrupts Descriptor when vCPU is preempted")

Message-Id: <[email protected]>

---
 arch/x86/kvm/vmx/posted_intr.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
arch/x86/kvm/vmx/posted_intr.c +3 -3
index 696ad48ab5daa..46fb83d6a286e 100644
@@ -51,7 +51,7 @@ void vmx_vcpu_pi_load(struct kvm_vcpu *vcpu, int cpu)
 
 	/* The full case.  */
 	do {
-		old.control = new.control = pi_desc->control;
+		old.control = new.control = READ_ONCE(pi_desc->control);
 
 		dest = cpu_physical_id(cpu);
 
@@ -104,7 +104,7 @@ static void __pi_post_block(struct kvm_vcpu *vcpu)
 	unsigned int dest;
 
 	do {
-		old.control = new.control = pi_desc->control;
+		old.control = new.control = READ_ONCE(pi_desc->control);
 		WARN(old.nv != POSTED_INTR_WAKEUP_VECTOR,
 		     "Wakeup handler not enabled while the VCPU is blocked\n");
 
@@ -163,7 +163,7 @@ int pi_pre_block(struct kvm_vcpu *vcpu)
 	}
 
 	do {
-		old.control = new.control = pi_desc->control;
+		old.control = new.control = READ_ONCE(pi_desc->control);
 
 		WARN((pi_desc->sn == 1),
 		     "Warning: SN field of posted-interrupts "
Signature · 1 line
2.34.1
author maintainer
#798 lore raw

PATCH 791/1039 KVM: X86: Ensure that dirty PDPTRs are loaded

In reply to Greg Kroah-Hartman

From: Lai Jiangshan <[email protected]>

[ Upstream commit 2c5653caecc4807b8abfe9c41880ac38417be7bf ]

For VMX with EPT, dirty PDPTRs need to be loaded before the next vmentry via vmx_load_mmu_pgd()

But not all paths that call load_pdptrs() will cause vmx_load_mmu_pgd() to be invoked. Normally, kvm_mmu_reset_context() is used to cause KVM_REQ_LOAD_MMU_PGD, but sometimes it is skipped:

* commit d81135a57aa6("KVM: x86: do not reset mmu if CR0.CD and CR0.NW are changed") skips kvm_mmu_reset_context() after load_pdptrs() when changing CR0.CD and CR0.NW.

* commit 21823fbda552("KVM: x86: Invalidate all PGDs for the current PCID on MOV CR3 w/ flush") skips KVM_REQ_LOAD_MMU_PGD after load_pdptrs() when rewriting the CR3 with the same value.

* commit a91a7c709600("KVM: X86: Don't reset mmu context when toggling X86_CR4_PGE") skips kvm_mmu_reset_context() after load_pdptrs() when changing CR4.PGE.

Fixes d81135a57aa6 ("KVM: x86: do not reset mmu if CR0.CD and CR0.NW are changed")
Fixes 21823fbda552 ("KVM: x86: Invalidate all PGDs for the current PCID on MOV CR3 w/ flush")
Fixes a91a7c709600 ("KVM: X86: Don't reset mmu context when toggling X86_CR4_PGE")

Message-Id: <[email protected]>

---
 arch/x86/kvm/x86.c | 1 +
 1 file changed, 1 insertion(+)
arch/x86/kvm/x86.c +1 -0
index 0b5c61bb24a17..bb87f43cdc78c 100644
@@ -830,6 +830,7 @@ int load_pdptrs(struct kvm_vcpu *vcpu, struct kvm_mmu *mmu, unsigned long cr3)
 
 	memcpy(mmu->pdptrs, pdpte, sizeof(mmu->pdptrs));
 	kvm_register_mark_dirty(vcpu, VCPU_EXREG_PDPTR);
+	kvm_make_request(KVM_REQ_LOAD_MMU_PGD, vcpu);
 	vcpu->arch.pdptrs_from_userspace = false;
 
 	return 1;
Signature · 1 line
2.34.1
author maintainer
#799 lore raw

PATCH 792/1039 KVM: x86: Handle 32-bit wrap of EIP for EMULTYPE_SKIP with flat code seg

In reply to Greg Kroah-Hartman

From: Sean Christopherson <[email protected]>

[ Upstream commit 5e854864ee4384736f27a986633bae21731a4e4e ]

Truncate the new EIP to a 32-bit value when handling EMULTYPE_SKIP as the decode phase does not truncate _eip. Wrapping the 32-bit boundary is legal if and only if CS is a flat code segment, but that check is implicitly handled in the form of limit checks in the decode phase.

Opportunstically prepare for a future fix by storing the result of any truncation in "eip" instead of "_eip".

Fixes 1957aa63be53 ("KVM: VMX: Handle single-step #DB for EMULTYPE_SKIP on EPT misconfig")

Message-Id: <093eabb1eab2965201c9b018373baf26ff256d85.1635842679.git.houwenlong93@linux.alibaba.com>

---
 arch/x86/kvm/x86.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)
arch/x86/kvm/x86.c +6 -1
index bb87f43cdc78c..5d5a8e75edcb7 100644
@@ -8134,7 +8134,12 @@ int x86_emulate_instruction(struct kvm_vcpu *vcpu, gpa_t cr2_or_gpa,
 	 * updating interruptibility state and injecting single-step #DBs.
 	 */
 	if (emulation_type & EMULTYPE_SKIP) {
-		kvm_rip_write(vcpu, ctxt->_eip);
+		if (ctxt->mode != X86EMUL_MODE_PROT64)
+			ctxt->eip = (u32)ctxt->_eip;
+		else
+			ctxt->eip = ctxt->_eip;
+
+		kvm_rip_write(vcpu, ctxt->eip);
 		if (ctxt->eflags & X86_EFLAGS_RF)
 			kvm_set_rflags(vcpu, ctxt->eflags & ~X86_EFLAGS_RF);
 		return 1;
Signature · 1 line
2.34.1
author maintainer
#800 lore raw

PATCH 793/1039 KVM: x86: Exit to userspace if emulation prepared a completion callback

In reply to Greg Kroah-Hartman

From: Hou Wenlong <[email protected]>

[ Upstream commit adbfb12d4c4517a8adde23a7fc46538953d56eea ]

em_rdmsr() and em_wrmsr() return X86EMUL_IO_NEEDED if MSR accesses required an exit to userspace. However, x86_emulate_insn() doesn't return X86EMUL_*, so x86_emulate_instruction() doesn't directly act on X86EMUL_IO_NEEDED; instead, it looks for other signals to differentiate between PIO, MMIO, etc. causing RDMSR/WRMSR emulation not to exit to userspace now.

Nevertheless, if the userspace_msr_exit_test testcase in selftests is changed to test RDMSR/WRMSR with a forced emulation prefix, the test passes. What happens is that first userspace exit information is filled but the userspace exit does not happen. Because x86_emulate_instruction() returns 1, the guest retries the instruction---but this time RIP has already been adjusted past the forced emulation prefix, so the guest executes RDMSR/WRMSR and the userspace exit finally happens.

Since the X86EMUL_IO_NEEDED path has provided a complete_userspace_io callback, x86_emulate_instruction() can just return 0 if the callback is not NULL. Then RDMSR/WRMSR instruction emulation will exit to userspace directly, without the RDMSR/WRMSR vmexit.

Fixes 1ae099540e8c7 ("KVM: x86: Allow deflecting unknown MSR accesses to user space")

Message-Id: <56f9df2ee5c05a81155e2be366c9dc1f7adc8817.1635842679.git.houwenlong93@linux.alibaba.com>

---
 arch/x86/kvm/x86.c | 3 +++
 1 file changed, 3 insertions(+)
arch/x86/kvm/x86.c +3 -0
index 5d5a8e75edcb7..00628031de98d 100644
@@ -8203,6 +8203,9 @@ restart:
 			writeback = false;
 		r = 0;
 		vcpu->arch.complete_userspace_io = complete_emulated_mmio;
+	} else if (vcpu->arch.complete_userspace_io) {
+		writeback = false;
+		r = 0;
 	} else if (r == EMULATION_RESTART)
 		goto restart;
 	else
Signature · 1 line
2.34.1