Skip to contentSkip to conversation

5.16.19-rc1 review

PATCH In review Greg Kroah-Hartman 15 participants 1,039 messages Apr 5, 2022 – Apr 7, 2022 Page 6 of 8

Subsystems (11)
Stable branch VFS MIPS Crypto API +7 more subsystems
lore
Share & tools
lore
author maintainer
#700 lore raw

PATCH 692/1017 staging: mt7621-dts: fix LEDs and pinctrl on GB-PC1 devicetree

In reply to Greg Kroah-Hartman

From: Arınç ÜNAL <[email protected]>

[ Upstream commit 6256e18686158fa49e019297f990f1c1817aabf1 ]

Fix LED and pinctrl definitions on the GB-PC1 devicetree. Refer to the schematics of the device for more information.

LED fixes:
- Change GPIO6 LED label from system to power as GPIO6 is connected to PLED.
- Add default-on default-trigger to power LED.
- Change GPIO8 LED label from status to system as GPIO8 is connected to SYS_LED.
- Add disk-activity default-trigger to system LED.
- Switch to the color:function naming scheme.
- Remove lan1 and lan2 LEDs as they don't exist.

Pinctrl fixes:
- Claim state_default node under pinctrl node.
- Change pinctrl0 node name to state-default.
- Change gpio node name to gpio-pinmux to respect Documentation/devicetree/bindings/pinctrl/ralink,rt2880-pinmux.yaml.
- Sort pin groups alphabetically.

Misc fixes:
- Fix formatting.
- Use the status value "okay".
- Define hexadecimal addresses in lower case.
- Make hexadecimal addresses for memory easier to read.

---
 drivers/staging/mt7621-dts/gbpc1.dts | 40 +++++++++++++---------------
 1 file changed, 18 insertions(+), 22 deletions(-)
drivers/staging/mt7621-dts/gbpc1.dts +18 -22
index e38a083811e5..5ae94b1ad599 100644
@@ -12,7 +12,8 @@
 
 	memory@0 {
 		device_type = "memory";
-		reg = <0x0 0x1c000000>, <0x20000000 0x4000000>;
+		reg = <0x00000000 0x1c000000>,
+		      <0x20000000 0x04000000>;
 	};
 
 	chosen {
@@ -38,24 +39,16 @@
 	gpio-leds {
 		compatible = "gpio-leds";
 
-		system {
-			label = "gb-pc1:green:system";
+		power {
+			label = "green:power";
 			gpios = <&gpio 6 GPIO_ACTIVE_LOW>;
+			linux,default-trigger = "default-on";
 		};
 
-		status {
-			label = "gb-pc1:green:status";
+		system {
+			label = "green:system";
 			gpios = <&gpio 8 GPIO_ACTIVE_LOW>;
-		};
-
-		lan1 {
-			label = "gb-pc1:green:lan1";
-			gpios = <&gpio 24 GPIO_ACTIVE_LOW>;
-		};
-
-		lan2 {
-			label = "gb-pc1:green:lan2";
-			gpios = <&gpio 25 GPIO_ACTIVE_LOW>;
+			linux,default-trigger = "disk-activity";
 		};
 	};
 };
@@ -95,9 +88,8 @@
 
 		partition@50000 {
 			label = "firmware";
-			reg = <0x50000 0x1FB0000>;
+			reg = <0x50000 0x1fb0000>;
 		};
-
 	};
 };
 
@@ -106,9 +98,12 @@
 };
 
 &pinctrl {
-	state_default: pinctrl0 {
-		default_gpio: gpio {
-			groups = "wdt", "rgmii2", "uart3";
+	pinctrl-names = "default";
+	pinctrl-0 = <&state_default>;
+
+	state_default: state-default {
+		gpio-pinmux {
+			groups = "rgmii2", "uart3", "wdt";
 			function = "gpio";
 		};
 	};
@@ -117,12 +112,13 @@
 &switch0 {
 	ports {
 		port@0 {
+			status = "okay";
 			label = "ethblack";
-			status = "ok";
 		};
+
 		port@4 {
+			status = "okay";
 			label = "ethblue";
-			status = "ok";
 		};
 	};
 };
Signature · 1 line
2.34.1
author maintainer
#701 lore raw

PATCH 693/1017 staging: mt7621-dts: fix formatting

In reply to Greg Kroah-Hartman

From: Arınç ÜNAL <[email protected]>

[ Upstream commit 7eeec44d33f6be7caca4fe9ca4e653cf315a36c1 ]

Fix formatting on mt7621.dtsi.

---
 drivers/staging/mt7621-dts/mt7621.dtsi | 15 ++++++++++++---
 1 file changed, 12 insertions(+), 3 deletions(-)
drivers/staging/mt7621-dts/mt7621.dtsi +12 -3
index 6d158e4f4b8c..25687f4fa671 100644
@@ -44,9 +44,9 @@
 		regulator-max-microvolt = <3300000>;
 		enable-active-high;
 		regulator-always-on;
-	  };
+	};
 
-	  mmc_fixed_1v8_io: fixedregulator@1 {
+	mmc_fixed_1v8_io: fixedregulator@1 {
 		compatible = "regulator-fixed";
 		regulator-name = "mmc_io";
 		regulator-min-microvolt = <1800000>;
@@ -363,17 +363,18 @@
 
 		mediatek,ethsys = <&sysc>;
 
-
 		gmac0: mac@0 {
 			compatible = "mediatek,eth-mac";
 			reg = <0>;
 			phy-mode = "rgmii";
+
 			fixed-link {
 				speed = <1000>;
 				full-duplex;
 				pause;
 			};
 		};
+
 		gmac1: mac@1 {
 			compatible = "mediatek,eth-mac";
 			reg = <1>;
@@ -381,6 +382,7 @@
 			phy-mode = "rgmii-rxid";
 			phy-handle = <&phy_external>;
 		};
+
 		mdio-bus {
 			#address-cells = <1>;
 			#size-cells = <0>;
@@ -411,36 +413,43 @@
 					#address-cells = <1>;
 					#size-cells = <0>;
 					reg = <0>;
+
 					port@0 {
 						status = "off";
 						reg = <0>;
 						label = "lan0";
 					};
+
 					port@1 {
 						status = "off";
 						reg = <1>;
 						label = "lan1";
 					};
+
 					port@2 {
 						status = "off";
 						reg = <2>;
 						label = "lan2";
 					};
+
 					port@3 {
 						status = "off";
 						reg = <3>;
 						label = "lan3";
 					};
+
 					port@4 {
 						status = "off";
 						reg = <4>;
 						label = "lan4";
 					};
+
 					port@6 {
 						reg = <6>;
 						label = "cpu";
 						ethernet = <&gmac0>;
 						phy-mode = "trgmii";
+
 						fixed-link {
 							speed = <1000>;
 							full-duplex;
Signature · 1 line
2.34.1
author maintainer
#702 lore raw

PATCH 694/1017 staging: mt7621-dts: fix pinctrl properties for ethernet

In reply to Greg Kroah-Hartman

From: Arınç ÜNAL <[email protected]>

[ Upstream commit 0a93c0d75809582893e82039143591b9265b520e ]

Add pinctrl properties with rgmii1 & mdio pins under ethernet node which was wrongfully put under an external phy node.
GMAC1 will start working with this fix.

Move GB-PC2 specific phy_external node to its own device tree.

---
 drivers/staging/mt7621-dts/gbpc2.dts   | 16 +++++++++++-----
 drivers/staging/mt7621-dts/mt7621.dtsi | 13 +++----------
 2 files changed, 14 insertions(+), 15 deletions(-)
drivers/staging/mt7621-dts/gbpc2.dts +11 -5
index 6fe603c7711d..03d6bb6735ac 100644
@@ -13,10 +13,16 @@
 	function = "gpio";
 };
 
-&gmac1 {
-	status = "ok";
-};
+&ethernet {
+	gmac1: mac@1 {
+		status = "ok";
+		phy-handle = <&phy_external>;
+	};
 
-&phy_external {
-	status = "ok";
+	mdio-bus {
+		phy_external: ethernet-phy@5 {
+			reg = <5>;
+			phy-mode = "rgmii-rxid";
+		};
+	};
 };
drivers/staging/mt7621-dts/mt7621.dtsi +3 -10
index 25687f4fa671..7b75e4a92b2a 100644
@@ -363,6 +363,9 @@
 
 		mediatek,ethsys = <&sysc>;
 
+		pinctrl-names = "default";
+		pinctrl-0 = <&rgmii1_pins &rgmii2_pins &mdio_pins>;
+
 		gmac0: mac@0 {
 			compatible = "mediatek,eth-mac";
 			reg = <0>;
@@ -380,22 +383,12 @@
 			reg = <1>;
 			status = "off";
 			phy-mode = "rgmii-rxid";
-			phy-handle = <&phy_external>;
 		};
 
 		mdio-bus {
 			#address-cells = <1>;
 			#size-cells = <0>;
 
-			phy_external: ethernet-phy@5 {
-				status = "off";
-				reg = <5>;
-				phy-mode = "rgmii-rxid";
-
-				pinctrl-names = "default";
-				pinctrl-0 = <&rgmii2_pins>;
-			};
-
 			switch0: switch0@0 {
 				compatible = "mediatek,mt7621";
 				#address-cells = <1>;
Signature · 1 line
2.34.1
author maintainer
#703 lore raw

PATCH 695/1017 staging: mt7621-dts: fix GB-PC2 devicetree

In reply to Greg Kroah-Hartman

From: Arınç ÜNAL <[email protected]>

[ Upstream commit 5bc148649cf358d0cccf525452a4efbd4bc89a0f ]

Fix the GB-PC2 devicetree. Refer to the schematics of the device for more information.

GB-PC2 devicetree fixes:
- Include mt7621.dtsi instead of gbpc1.dts. Add the missing definitions.
- Remove gpio-leds node as the system LED is not wired to anywhere on the board and the power LED is directly wired to GND.
- Remove uart3 pin group from gpio-pinmux node as it's not used as GPIO.
- Use reg 7 for the external phy to be on par with Documentation/devicetree/bindings/net/dsa/mt7530.txt.
- Use the status value "okay".

---
 drivers/staging/mt7621-dts/gbpc2.dts | 110 +++++++++++++++++++++++++--
 1 file changed, 102 insertions(+), 8 deletions(-)
drivers/staging/mt7621-dts/gbpc2.dts +102 -8
index 03d6bb6735ac..a7fce8de6147 100644
@@ -1,28 +1,122 @@
 // SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
 /dts-v1/;
 
-#include "gbpc1.dts"
+#include "mt7621.dtsi"
+
+#include <dt-bindings/gpio/gpio.h>
+#include <dt-bindings/input/input.h>
 
 / {
 	compatible = "gnubee,gb-pc2", "mediatek,mt7621-soc";
 	model = "GB-PC2";
+
+	memory@0 {
+		device_type = "memory";
+		reg = <0x00000000 0x1c000000>,
+		      <0x20000000 0x04000000>;
+	};
+
+	chosen {
+		bootargs = "console=ttyS0,57600";
+	};
+
+	palmbus: palmbus@1e000000 {
+		i2c@900 {
+			status = "okay";
+		};
+	};
+
+	gpio-keys {
+		compatible = "gpio-keys";
+
+		reset {
+			label = "reset";
+			gpios = <&gpio 18 GPIO_ACTIVE_HIGH>;
+			linux,code = <KEY_RESTART>;
+		};
+	};
 };
 
-&default_gpio {
-	groups = "wdt", "uart3";
-	function = "gpio";
+&sdhci {
+	status = "okay";
+};
+
+&spi0 {
+	status = "okay";
+
+	m25p80@0 {
+		#address-cells = <1>;
+		#size-cells = <1>;
+		compatible = "jedec,spi-nor";
+		reg = <0>;
+		spi-max-frequency = <50000000>;
+		broken-flash-reset;
+
+		partition@0 {
+			label = "u-boot";
+			reg = <0x0 0x30000>;
+			read-only;
+		};
+
+		partition@30000 {
+			label = "u-boot-env";
+			reg = <0x30000 0x10000>;
+			read-only;
+		};
+
+		factory: partition@40000 {
+			label = "factory";
+			reg = <0x40000 0x10000>;
+			read-only;
+		};
+
+		partition@50000 {
+			label = "firmware";
+			reg = <0x50000 0x1fb0000>;
+		};
+	};
+};
+
+&pcie {
+	status = "okay";
+};
+
+&pinctrl {
+	pinctrl-names = "default";
+	pinctrl-0 = <&state_default>;
+
+	state_default: state-default {
+		gpio-pinmux {
+			groups = "wdt";
+			function = "gpio";
+		};
+	};
 };
 
 &ethernet {
 	gmac1: mac@1 {
-		status = "ok";
-		phy-handle = <&phy_external>;
+		status = "okay";
+		phy-handle = <&ethphy7>;
 	};
 
 	mdio-bus {
-		phy_external: ethernet-phy@5 {
-			reg = <5>;
+		ethphy7: ethernet-phy@7 {
+			reg = <7>;
 			phy-mode = "rgmii-rxid";
 		};
 	};
 };
+
+&switch0 {
+	ports {
+		port@0 {
+			status = "okay";
+			label = "ethblack";
+		};
+
+		port@4 {
+			status = "okay";
+			label = "ethblue";
+		};
+	};
+};
Signature · 1 line
2.34.1
author maintainer
#704 lore raw

PATCH 696/1017 pinctrl: mediatek: Fix missing of_node_put() in mtk_pctrl_init

In reply to Greg Kroah-Hartman

From: Miaoqian Lin <[email protected]>

[ Upstream commit dab4df9ca919f59e5b9dd84385eaf34d4f20dbb0 ]

The device_node pointer is returned by of_parse_phandle() with refcount incremented. We should use of_node_put() on it when done.

Fixes a6df410d420a ("pinctrl: mediatek: Add Pinctrl/GPIO driver for mt8135.")
---
 drivers/pinctrl/mediatek/pinctrl-mtk-common.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/pinctrl/mediatek/pinctrl-mtk-common.c +2 -0
index 5f7c421ab6e7..334cb85855a9 100644
@@ -1038,6 +1038,7 @@ int mtk_pctrl_init(struct platform_device *pdev,
 	node = of_parse_phandle(np, "mediatek,pctl-regmap", 0);
 	if (node) {
 		pctl->regmap1 = syscon_node_to_regmap(node);
+		of_node_put(node);
 		if (IS_ERR(pctl->regmap1))
 			return PTR_ERR(pctl->regmap1);
 	} else if (regmap) {
@@ -1051,6 +1052,7 @@ int mtk_pctrl_init(struct platform_device *pdev,
 	node = of_parse_phandle(np, "mediatek,pctl-regmap", 1);
 	if (node) {
 		pctl->regmap2 = syscon_node_to_regmap(node);
+		of_node_put(node);
 		if (IS_ERR(pctl->regmap2))
 			return PTR_ERR(pctl->regmap2);
 	}
Signature · 1 line
2.34.1
author maintainer
#705 lore raw

PATCH 697/1017 pinctrl: mediatek: paris: Fix PIN_CONFIG_BIAS_* readback

In reply to Greg Kroah-Hartman

From: Chen-Yu Tsai <[email protected]>

[ Upstream commit 3e8c6bc608480010f360c4a59578d7841726137d ]

When reading back pin bias settings, if the pin is not in the corresponding bias state, the function should return -EINVAL.

Fix this in the mediatek-paris pinctrl library so that the read back state is not littered with bogus a "input bias disabled" combined with "pull up" or "pull down" states.

Fixes 805250982bb5 ("pinctrl: mediatek: add pinctrl-paris that implements the vendor dt-bindings")
---
 drivers/pinctrl/mediatek/pinctrl-paris.c | 16 ++++++----------
 1 file changed, 6 insertions(+), 10 deletions(-)
drivers/pinctrl/mediatek/pinctrl-paris.c +6 -10
index 4c6f6d967b18..f7644530cee1 100644
@@ -96,20 +96,16 @@ static int mtk_pinconf_get(struct pinctrl_dev *pctldev,
 			err = hw->soc->bias_get_combo(hw, desc, &pullup, &ret);
 			if (err)
 				goto out;
+			if (ret == MTK_PUPD_SET_R1R0_00)
+				ret = MTK_DISABLE;
 			if (param == PIN_CONFIG_BIAS_DISABLE) {
-				if (ret == MTK_PUPD_SET_R1R0_00)
-					ret = MTK_DISABLE;
+				if (ret != MTK_DISABLE)
+					err = -EINVAL;
 			} else if (param == PIN_CONFIG_BIAS_PULL_UP) {
-				/* When desire to get pull-up value, return
-				 *  error if current setting is pull-down
-				 */
-				if (!pullup)
+				if (!pullup || ret == MTK_DISABLE)
 					err = -EINVAL;
 			} else if (param == PIN_CONFIG_BIAS_PULL_DOWN) {
-				/* When desire to get pull-down value, return
-				 *  error if current setting is pull-up
-				 */
-				if (pullup)
+				if (pullup || ret == MTK_DISABLE)
 					err = -EINVAL;
 			}
 		} else {
Signature · 1 line
2.34.1
author maintainer
#706 lore raw

PATCH 698/1017 pinctrl: mediatek: paris: Fix "argument" argument type for mtk_pinconf_get()

In reply to Greg Kroah-Hartman

From: Chen-Yu Tsai <[email protected]>

[ Upstream commit 19bce7ce0a593c7024030a0cda9e23facea3c93d ]

For mtk_pinconf_get(), the "argument" argument is typically returned by pinconf_to_config_argument(), which holds the value for a given pinconf parameter. It certainly should not have the type of "enum pin_config_param", which describes the type of the pinconf parameter itself.

Change the type to u32, which matches the return type of pinconf_to_config_argument().

Fixes 805250982bb5 ("pinctrl: mediatek: add pinctrl-paris that implements the vendor dt-bindings")
---
 drivers/pinctrl/mediatek/pinctrl-paris.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
drivers/pinctrl/mediatek/pinctrl-paris.c +1 -2
index f7644530cee1..cbaad89fbe17 100644
@@ -184,8 +184,7 @@ static int mtk_pinconf_get(struct pinctrl_dev *pctldev,
 }
 
 static int mtk_pinconf_set(struct pinctrl_dev *pctldev, unsigned int pin,
-			   enum pin_config_param param,
-			   enum pin_config_param arg)
+			   enum pin_config_param param, u32 arg)
 {
 	struct mtk_pinctrl *hw = pinctrl_dev_get_drvdata(pctldev);
 	const struct mtk_pin_desc *desc;
Signature · 1 line
2.34.1
author maintainer
#707 lore raw

PATCH 699/1017 pinctrl: mediatek: paris: Fix pingroup pin config state readback

In reply to Greg Kroah-Hartman

From: Chen-Yu Tsai <[email protected]>

[ Upstream commit 54fe55fb384ade630ef20b9a8b8f3b2a89ad97f2 ]

mtk_pconf_group_get(), used to read back pingroup pin config state, simply returns a set of configs saved from a previous invocation of mtk_pconf_group_set(). This is an unfiltered, unvalidated set passed in from the pinconf core, which does not match the current hardware state.

Since the driver library is designed to have one pin per group, pass through mtk_pconf_group_get() to mtk_pinconf_get(), to read back the current pin config state of the only pin in the group.

Also drop the assignment of pin config state to the group.

Fixes 805250982bb5 ("pinctrl: mediatek: add pinctrl-paris that implements the vendor dt-bindings")
---
 drivers/pinctrl/mediatek/pinctrl-paris.c | 8 +++-----
 1 file changed, 3 insertions(+), 5 deletions(-)
drivers/pinctrl/mediatek/pinctrl-paris.c +3 -5
index cbaad89fbe17..faf11c9e1db3 100644
@@ -732,10 +732,10 @@ static int mtk_pconf_group_get(struct pinctrl_dev *pctldev, unsigned group,
 			       unsigned long *config)
 {
 	struct mtk_pinctrl *hw = pinctrl_dev_get_drvdata(pctldev);
+	struct mtk_pinctrl_group *grp = &hw->groups[group];
 
-	*config = hw->groups[group].config;
-
-	return 0;
+	 /* One pin per group only */
+	return mtk_pinconf_get(pctldev, grp->pin, config);
 }
 
 static int mtk_pconf_group_set(struct pinctrl_dev *pctldev, unsigned group,
@@ -751,8 +751,6 @@ static int mtk_pconf_group_set(struct pinctrl_dev *pctldev, unsigned group,
 				      pinconf_to_config_argument(configs[i]));
 		if (ret < 0)
 			return ret;
-
-		grp->config = configs[i];
 	}
 
 	return 0;
Signature · 1 line
2.34.1
author maintainer
#708 lore raw

PATCH 700/1017 pinctrl: mediatek: paris: Skip custom extra pin config dump for virtual GPIOs

In reply to Greg Kroah-Hartman

From: Chen-Yu Tsai <[email protected]>

[ Upstream commit 1763933d377ecb05454f8d20e3c8922480db2ac0 ]

Virtual GPIOs do not have any hardware state associated with them. Any attempt to read back hardware state for these pins result in error codes.

Skip dumping extra pin config information for these virtual GPIOs.

Fixes 184d8e13f9b1 ("pinctrl: mediatek: Add support for pin configuration dump via debugfs.")
---
 drivers/pinctrl/mediatek/pinctrl-paris.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/pinctrl/mediatek/pinctrl-paris.c +3 -0
index faf11c9e1db3..154d55a239e8 100644
@@ -581,6 +581,9 @@ ssize_t mtk_pctrl_show_one_pin(struct mtk_pinctrl *hw,
 	if (gpio >= hw->soc->npins)
 		return -EINVAL;
 
+	if (mtk_is_virt_gpio(hw, gpio))
+		return -EINVAL;
+
 	desc = (const struct mtk_pin_desc *)&hw->soc->pins[gpio];
 	pinmux = mtk_pctrl_get_pinmux(hw, gpio);
 	if (pinmux >= hw->soc->nfuncs)
Signature · 1 line
2.34.1
author maintainer
#709 lore raw

PATCH 701/1017 pinctrl: microchip-sgpio: lock RMW access

In reply to Greg Kroah-Hartman

From: Michael Walle <[email protected]>

[ Upstream commit 7996c5f5ec7a20b3f6b8fae93fcf3cb8f1c01743 ]

Protect any RMW access to the registers by a spinlock.

Fixes 7e5ea974e61c ("pinctrl: pinctrl-microchip-sgpio: Add pinctrl driver for Microsemi Serial GPIO")
---
 drivers/pinctrl/pinctrl-microchip-sgpio.c | 15 +++++++++++++++
 1 file changed, 15 insertions(+)
drivers/pinctrl/pinctrl-microchip-sgpio.c +15 -0
index 78765faa245a..dfa374195694 100644
@@ -18,6 +18,7 @@
 #include <linux/platform_device.h>
 #include <linux/property.h>
 #include <linux/reset.h>
+#include <linux/spinlock.h>
 
 #include "core.h"
 #include "pinconf.h"
@@ -115,6 +116,7 @@ struct sgpio_priv {
 	u32 clock;
 	u32 __iomem *regs;
 	const struct sgpio_properties *properties;
+	spinlock_t lock;
 };
 
 struct sgpio_port_addr {
@@ -216,6 +218,7 @@ static void sgpio_output_set(struct sgpio_priv *priv,
 			     int value)
 {
 	unsigned int bit = SGPIO_SRC_BITS * addr->bit;
+	unsigned long flags;
 	u32 clr, set;
 
 	switch (priv->properties->arch) {
@@ -234,7 +237,10 @@ static void sgpio_output_set(struct sgpio_priv *priv,
 	default:
 		return;
 	}
+
+	spin_lock_irqsave(&priv->lock, flags);
 	sgpio_clrsetbits(priv, REG_PORT_CONFIG, addr->port, clr, set);
+	spin_unlock_irqrestore(&priv->lock, flags);
 }
 
 static int sgpio_output_get(struct sgpio_priv *priv,
@@ -562,10 +568,13 @@ static void microchip_sgpio_irq_settype(struct irq_data *data,
 	struct sgpio_bank *bank = gpiochip_get_data(chip);
 	unsigned int gpio = irqd_to_hwirq(data);
 	struct sgpio_port_addr addr;
+	unsigned long flags;
 	u32 ena;
 
 	sgpio_pin_to_addr(bank->priv, gpio, &addr);
 
+	spin_lock_irqsave(&bank->priv->lock, flags);
+
 	/* Disable interrupt while changing type */
 	ena = sgpio_readl(bank->priv, REG_INT_ENABLE, addr.bit);
 	sgpio_writel(bank->priv, ena & ~BIT(addr.port), REG_INT_ENABLE, addr.bit);
@@ -582,6 +591,8 @@ static void microchip_sgpio_irq_settype(struct irq_data *data,
 
 	/* Possibly re-enable interrupts */
 	sgpio_writel(bank->priv, ena, REG_INT_ENABLE, addr.bit);
+
+	spin_unlock_irqrestore(&bank->priv->lock, flags);
 }
 
 static void microchip_sgpio_irq_setreg(struct irq_data *data,
@@ -592,13 +603,16 @@ static void microchip_sgpio_irq_setreg(struct irq_data *data,
 	struct sgpio_bank *bank = gpiochip_get_data(chip);
 	unsigned int gpio = irqd_to_hwirq(data);
 	struct sgpio_port_addr addr;
+	unsigned long flags;
 
 	sgpio_pin_to_addr(bank->priv, gpio, &addr);
 
+	spin_lock_irqsave(&bank->priv->lock, flags);
 	if (clear)
 		sgpio_clrsetbits(bank->priv, reg, addr.bit, BIT(addr.port), 0);
 	else
 		sgpio_clrsetbits(bank->priv, reg, addr.bit, 0, BIT(addr.port));
+	spin_unlock_irqrestore(&bank->priv->lock, flags);
 }
 
 static void microchip_sgpio_irq_mask(struct irq_data *data)
@@ -814,6 +828,7 @@ static int microchip_sgpio_probe(struct platform_device *pdev)
 		return -ENOMEM;
 
 	priv->dev = dev;
+	spin_lock_init(&priv->lock);
 
 	reset = devm_reset_control_get_optional_shared(&pdev->dev, "switch");
 	if (IS_ERR(reset))
Signature · 1 line
2.34.1
author maintainer
#710 lore raw

PATCH 702/1017 pinctrl: nomadik: Add missing of_node_put() in nmk_pinctrl_probe

In reply to Greg Kroah-Hartman

From: Miaoqian Lin <[email protected]>

[ Upstream commit c09ac191b1f97cfa06f394dbfd7a5db07986cefc ]

This node pointer is returned by of_parse_phandle() with refcount incremented in this function. Calling of_node_put() to avoid the refcount leak.

Fixes 32e67eee670e ("pinctrl: nomadik: Allow prcm_base to be extracted from Device Tree")
---
 drivers/pinctrl/nomadik/pinctrl-nomadik.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/pinctrl/nomadik/pinctrl-nomadik.c +3 -1
index 39828e9c3120..4757bf964d3c 100644
@@ -1883,8 +1883,10 @@ static int nmk_pinctrl_probe(struct platform_device *pdev)
 	}
 
 	prcm_np = of_parse_phandle(np, "prcm", 0);
-	if (prcm_np)
+	if (prcm_np) {
 		npct->prcm_base = of_iomap(prcm_np, 0);
+		of_node_put(prcm_np);
+	}
 	if (!npct->prcm_base) {
 		if (version == PINCTRL_NMK_STN8815) {
 			dev_info(&pdev->dev,
Signature · 1 line
2.34.1
author maintainer
#711 lore raw

PATCH 703/1017 pinctrl/rockchip: Add missing of_node_put() in rockchip_pinctrl_probe

In reply to Greg Kroah-Hartman

From: Miaoqian Lin <[email protected]>

[ Upstream commit 89388f8730699c259f8090ec435fb43569efe4ac ]

The device_node pointer is returned by of_parse_phandle() with refcount incremented. We should use of_node_put() on it when done.

Fixes 1e747e59cc4d ("pinctrl: rockchip: base regmap supplied by a syscon")
Fixes 14dee8677e19 ("pinctrl: rockchip: let pmu registers be supplied by a syscon")
---
 drivers/pinctrl/pinctrl-rockchip.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/pinctrl/pinctrl-rockchip.c +2 -0
index dc52da94af0b..923ff21a44c0 100644
@@ -2702,6 +2702,7 @@ static int rockchip_pinctrl_probe(struct platform_device *pdev)
 	node = of_parse_phandle(np, "rockchip,grf", 0);
 	if (node) {
 		info->regmap_base = syscon_node_to_regmap(node);
+		of_node_put(node);
 		if (IS_ERR(info->regmap_base))
 			return PTR_ERR(info->regmap_base);
 	} else {
@@ -2738,6 +2739,7 @@ static int rockchip_pinctrl_probe(struct platform_device *pdev)
 	node = of_parse_phandle(np, "rockchip,pmu", 0);
 	if (node) {
 		info->regmap_pmu = syscon_node_to_regmap(node);
+		of_node_put(node);
 		if (IS_ERR(info->regmap_pmu))
 			return PTR_ERR(info->regmap_pmu);
 	}
Signature · 1 line
2.34.1
author maintainer
#712 lore raw

PATCH 704/1017 tty: hvc: fix return value of __setup handler

In reply to Greg Kroah-Hartman

From: Randy Dunlap <[email protected]>

[ Upstream commit 53819a0d97aace1425bb042829e3446952a9e8a9 ]

__setup() handlers should return 1 to indicate that the boot option has been handled or 0 to indicate that it was not handled. Add a pr_warn() message if the option value is invalid and then always return 1.

Link lore.kernel.org/r/[email protected]
Fixes 86b40567b917 ("tty: replace strict_strtoul() with kstrtoul()")
Cc Jingoo Han <[email protected]>
Cc Greg Kroah-Hartman <[email protected]>
Cc Jiri Slaby <[email protected]>
Cc Michael Ellerman <[email protected]>
Cc Julian Wiedmann <[email protected]>
Cc Vasily Gorbik <[email protected]>
---
 drivers/tty/hvc/hvc_iucv.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/tty/hvc/hvc_iucv.c +3 -1
index 82a76cac94de..32366caca662 100644
@@ -1417,7 +1417,9 @@ static int __init hvc_iucv_init(void)
  */
 static	int __init hvc_iucv_config(char *val)
 {
-	 return kstrtoul(val, 10, &hvc_iucv_devices);
+	if (kstrtoul(val, 10, &hvc_iucv_devices))
+		pr_warn("hvc_iucv= invalid parameter value '%s'\n", val);
+	return 1;
 }
 
 
Signature · 1 line
2.34.1
author maintainer
#713 lore raw

PATCH 705/1017 kgdboc: fix return value of __setup handler

In reply to Greg Kroah-Hartman

From: Randy Dunlap <[email protected]>

[ Upstream commit ab818c7aa7544bf8d2dd4bdf68878b17a02eb332 ]

__setup() handlers should return 1 to obsolete_checksetup() in init/main.c to indicate that the boot option has been handled. A return of 0 causes the boot option/value to be listed as an Unknown kernel parameter and added to init's (limited) environment strings. So return 1 from kgdboc_option_setup().

Unknown kernel command line parameters "BOOT_IMAGE=/boot/bzImage-517rc7
kgdboc=kbd kgdbts=", will be passed to user space.

Run /sbin/init as init process
with arguments:
/sbin/init
with environment:

     HOME=/
     TERM=linux
     BOOT_IMAGE=/boot/bzImage-517rc7
     kgdboc=kbd
     kgdbts=
Link lore.kernel.org/r/[email protected]
Fixes 1bd54d851f50 ("kgdboc: Passing ekgdboc to command line causes panic")
Fixes f2d937f3bf00 ("consoles: polling support, kgdboc")
Cc He Zhe <[email protected]>
Cc Greg Kroah-Hartman <[email protected]>
Cc Jiri Slaby <[email protected]>
Cc Jason Wessel <[email protected]>
Cc Daniel Thompson <[email protected]>
Cc Douglas Anderson <[email protected]>
---
 drivers/tty/serial/kgdboc.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/tty/serial/kgdboc.c +3 -3
index 49d0c7f2b29b..79b7db8580e0 100644
@@ -403,16 +403,16 @@ static int kgdboc_option_setup(char *opt)
 {
 	if (!opt) {
 		pr_err("config string not provided\n");
-		return -EINVAL;
+		return 1;
 	}
 
 	if (strlen(opt) >= MAX_CONFIG_LEN) {
 		pr_err("config string too long\n");
-		return -ENOSPC;
+		return 1;
 	}
 	strcpy(config, opt);
 
-	return 0;
+	return 1;
 }
 
 __setup("kgdboc=", kgdboc_option_setup);
Signature · 1 line
2.34.1
author maintainer
#714 lore raw

PATCH 706/1017 serial: 8250: fix XOFF/XON sending when DMA is used

In reply to Greg Kroah-Hartman

From: Ilpo Järvinen <[email protected]>

[ Upstream commit f58c252e30cf74f68b0054293adc03b5923b9f0e ]

When 8250 UART is using DMA, x_char (XON/XOFF) is never sent to the wire. After this change, x_char is injected correctly.

Create uart_xchar_out() helper for sending the x_char out and accounting related to it. It seems that almost every driver does these same steps with x_char. Except for 8250, however, almost all currently lack .serial_out so they cannot immediately take advantage of this new helper.

The downside of this patch is that it might reintroduce the problems some devices faced with mixed DMA/non-DMA transfer which caused revert f967fc8f165f (Revert "serial: 8250_dma:
don't bother DMA with small transfers"). However, the impact should be limited to cases with XON/XOFF (that didn't work with DMA capable devices to begin with so this problem is not very likely to cause a major issue, if any at all).

---
 drivers/tty/serial/8250/8250_dma.c  | 11 ++++++++++-
 drivers/tty/serial/8250/8250_port.c |  4 +---
 drivers/tty/serial/serial_core.c    | 14 ++++++++++++++
 include/linux/serial_core.h         |  2 ++
 4 files changed, 27 insertions(+), 4 deletions(-)
drivers/tty/serial/8250/8250_dma.c +10 -1
index 890fa7ddaa7f..b3c3f7e5851a 100644
@@ -64,10 +64,19 @@ int serial8250_tx_dma(struct uart_8250_port *p)
 	struct uart_8250_dma		*dma = p->dma;
 	struct circ_buf			*xmit = &p->port.state->xmit;
 	struct dma_async_tx_descriptor	*desc;
+	struct uart_port		*up = &p->port;
 	int ret;
 
-	if (dma->tx_running)
+	if (dma->tx_running) {
+		if (up->x_char) {
+			dmaengine_pause(dma->txchan);
+			uart_xchar_out(up, UART_TX);
+			dmaengine_resume(dma->txchan);
+		}
 		return 0;
+	} else if (up->x_char) {
+		uart_xchar_out(up, UART_TX);
+	}
 
 	if (uart_tx_stopped(&p->port) || uart_circ_empty(xmit)) {
 		/* We have been called from __dma_tx_complete() */
drivers/tty/serial/8250/8250_port.c +1 -3
index 344fbe7426f6..1e1750a9614e 100644
@@ -1819,9 +1819,7 @@ void serial8250_tx_chars(struct uart_8250_port *up)
 	int count;
 
 	if (port->x_char) {
-		serial_out(up, UART_TX, port->x_char);
-		port->icount.tx++;
-		port->x_char = 0;
+		uart_xchar_out(port, UART_TX);
 		return;
 	}
 	if (uart_tx_stopped(port)) {
drivers/tty/serial/serial_core.c +14 -0
index dc6129ddef85..eb15423f935a 100644
@@ -652,6 +652,20 @@ static void uart_flush_buffer(struct tty_struct *tty)
 	tty_port_tty_wakeup(&state->port);
 }
 
+/*
+ * This function performs low-level write of high-priority XON/XOFF
+ * character and accounting for it.
+ *
+ * Requires uart_port to implement .serial_out().
+ */
+void uart_xchar_out(struct uart_port *uport, int offset)
+{
+	serial_port_out(uport, offset, uport->x_char);
+	uport->icount.tx++;
+	uport->x_char = 0;
+}
+EXPORT_SYMBOL_GPL(uart_xchar_out);
+
 /*
  * This function is used to send a high-priority XON/XOFF character to
  * the device
include/linux/serial_core.h +2 -0
index c58cc142d23f..8c32935e1059 100644
@@ -458,6 +458,8 @@ extern void uart_handle_cts_change(struct uart_port *uport,
 extern void uart_insert_char(struct uart_port *port, unsigned int status,
 		 unsigned int overrun, unsigned int ch, unsigned int flag);
 
+void uart_xchar_out(struct uart_port *uport, int offset);
+
 #ifdef CONFIG_MAGIC_SYSRQ_SERIAL
 #define SYSRQ_TIMEOUT	(HZ * 5)
 
Signature · 1 line
2.34.1
author maintainer
#715 lore raw

PATCH 707/1017 virt: acrn: obtain pa from VMA with PFNMAP flag

In reply to Greg Kroah-Hartman

From: Yonghua Huang <[email protected]>

[ Upstream commit 8a6e85f75a83d16a71077e41f2720c691f432002 ]

acrn_vm_ram_map can't pin the user pages with VM_PFNMAP flag by calling get_user_pages_fast(), the PA(physical pages) may be mapped by kernel driver and set PFNMAP flag.

This patch fixes logic to setup EPT mapping for PFN mapped RAM region by checking the memory attribute before adding EPT mapping for them.

Fixes 88f537d5e8dd ("virt: acrn: Introduce EPT mapping management")
---
 drivers/virt/acrn/mm.c | 24 ++++++++++++++++++++++++
 1 file changed, 24 insertions(+)
drivers/virt/acrn/mm.c +24 -0
index c4f2e15c8a2b..3b1b1e7a844b 100644
@@ -162,10 +162,34 @@ int acrn_vm_ram_map(struct acrn_vm *vm, struct acrn_vm_memmap *memmap)
 	void *remap_vaddr;
 	int ret, pinned;
 	u64 user_vm_pa;
+	unsigned long pfn;
+	struct vm_area_struct *vma;
 
 	if (!vm || !memmap)
 		return -EINVAL;
 
+	mmap_read_lock(current->mm);
+	vma = vma_lookup(current->mm, memmap->vma_base);
+	if (vma && ((vma->vm_flags & VM_PFNMAP) != 0)) {
+		if ((memmap->vma_base + memmap->len) > vma->vm_end) {
+			mmap_read_unlock(current->mm);
+			return -EINVAL;
+		}
+
+		ret = follow_pfn(vma, memmap->vma_base, &pfn);
+		mmap_read_unlock(current->mm);
+		if (ret < 0) {
+			dev_dbg(acrn_dev.this_device,
+				"Failed to lookup PFN at VMA:%pK.\n", (void *)memmap->vma_base);
+			return ret;
+		}
+
+		return acrn_mm_region_add(vm, memmap->user_vm_pa,
+			 PFN_PHYS(pfn), memmap->len,
+			 ACRN_MEM_TYPE_WB, memmap->attr);
+	}
+	mmap_read_unlock(current->mm);
+
 	/* Get the page number of the map region */
 	nr_pages = memmap->len >> PAGE_SHIFT;
 	pages = vzalloc(nr_pages * sizeof(struct page *));
Signature · 1 line
2.34.1
author maintainer
#716 lore raw

PATCH 708/1017 virt: acrn: fix a memory leak in acrn_dev_ioctl()

In reply to Greg Kroah-Hartman

From: Xiaolong Huang <[email protected]>

[ Upstream commit ecd1735f14d6ac868ae5d8b7a2bf193fa11f388b ]

The vm_param and cpu_regs need to be freed via kfree() before return -EINVAL error.

Fixes 9c5137aedd11 ("virt: acrn: Introduce VM management interfaces")
Fixes 2ad2aaee1bc9 ("virt: acrn: Introduce an ioctl to set vCPU registers state")
---
 drivers/virt/acrn/hsm.c | 20 +++++++++++++++-----
 1 file changed, 15 insertions(+), 5 deletions(-)
drivers/virt/acrn/hsm.c +15 -5
index 5419794fccf1..423ea888d79a 100644
@@ -136,8 +136,10 @@ static long acrn_dev_ioctl(struct file *filp, unsigned int cmd,
 		if (IS_ERR(vm_param))
 			return PTR_ERR(vm_param);
 
-		if ((vm_param->reserved0 | vm_param->reserved1) != 0)
+		if ((vm_param->reserved0 | vm_param->reserved1) != 0) {
+			kfree(vm_param);
 			return -EINVAL;
+		}
 
 		vm = acrn_vm_create(vm, vm_param);
 		if (!vm) {
@@ -182,21 +184,29 @@ static long acrn_dev_ioctl(struct file *filp, unsigned int cmd,
 			return PTR_ERR(cpu_regs);
 
 		for (i = 0; i < ARRAY_SIZE(cpu_regs->reserved); i++)
-			if (cpu_regs->reserved[i])
+			if (cpu_regs->reserved[i]) {
+				kfree(cpu_regs);
 				return -EINVAL;
+			}
 
 		for (i = 0; i < ARRAY_SIZE(cpu_regs->vcpu_regs.reserved_32); i++)
-			if (cpu_regs->vcpu_regs.reserved_32[i])
+			if (cpu_regs->vcpu_regs.reserved_32[i]) {
+				kfree(cpu_regs);
 				return -EINVAL;
+			}
 
 		for (i = 0; i < ARRAY_SIZE(cpu_regs->vcpu_regs.reserved_64); i++)
-			if (cpu_regs->vcpu_regs.reserved_64[i])
+			if (cpu_regs->vcpu_regs.reserved_64[i]) {
+				kfree(cpu_regs);
 				return -EINVAL;
+			}
 
 		for (i = 0; i < ARRAY_SIZE(cpu_regs->vcpu_regs.gdt.reserved); i++)
 			if (cpu_regs->vcpu_regs.gdt.reserved[i] |
-			    cpu_regs->vcpu_regs.idt.reserved[i])
+			    cpu_regs->vcpu_regs.idt.reserved[i]) {
+				kfree(cpu_regs);
 				return -EINVAL;
+			}
 
 		ret = hcall_set_vcpu_regs(vm->vmid, virt_to_phys(cpu_regs));
 		if (ret < 0)
Signature · 1 line
2.34.1
author maintainer
#717 lore raw

PATCH 709/1017 kgdbts: fix return value of __setup handler

In reply to Greg Kroah-Hartman

From: Randy Dunlap <[email protected]>

[ Upstream commit 96c9e802c64014a7716865332d732cc9c7f24593 ]

__setup() handlers should return 1 to indicate that the boot option has been handled. A return of 0 causes the boot option/value to be listed as an Unknown kernel parameter and added to init's (limited) environment strings. So return 1 from kgdbts_option_setup().

Unknown kernel command line parameters "BOOT_IMAGE=/boot/bzImage-517rc7
kgdboc=kbd kgdbts=", will be passed to user space.

Run /sbin/init as init process
with arguments:
/sbin/init
with environment:

     HOME=/
     TERM=linux
     BOOT_IMAGE=/boot/bzImage-517rc7
     kgdboc=kbd
     kgdbts=
Link lore.kernel.org/r/[email protected]
Fixes e8d31c204e36 ("kgdb: add kgdb internal test suite")
Cc Jason Wessel <[email protected]>
Cc Daniel Thompson <[email protected]>
Cc Douglas Anderson <[email protected]>
Cc Arnd Bergmann <[email protected]>
Cc Greg Kroah-Hartman <[email protected]>
---
 drivers/misc/kgdbts.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/misc/kgdbts.c +2 -2
index 67c5b452dd35..88b91ad8e541 100644
@@ -1070,10 +1070,10 @@ static int kgdbts_option_setup(char *opt)
 {
 	if (strlen(opt) >= MAX_CONFIG_LEN) {
 		printk(KERN_ERR "kgdbts: config string too long\n");
-		return -ENOSPC;
+		return 1;
 	}
 	strcpy(config, opt);
-	return 0;
+	return 1;
 }
 
 __setup("kgdbts=", kgdbts_option_setup);
Signature · 1 line
2.34.1
author maintainer
#718 lore raw

PATCH 710/1017 firmware: google: Properly state IOMEM dependency

In reply to Greg Kroah-Hartman

From: David Gow <[email protected]>

[ Upstream commit 37fd83916da2e4cae03d350015c82a67b1b334c4 ]

The Google Coreboot implementation requires IOMEM functions (memmremap, memunmap, devm_memremap), but does not specify this is its Kconfig. This results in build errors when HAS_IOMEM is not set, such as on some UML configurations:

/usr/bin/ld: drivers/firmware/google/coreboot_table.o: in function `coreboot_table_probe':
coreboot_table.c:(.text+0x311): undefined reference to memremap' /usr/bin/ld: coreboot_table.c:(.text+0x34e): undefined reference to memunmap' /usr/bin/ld: drivers/firmware/google/memconsole-coreboot.o: in function `memconsole_probe':
memconsole-coreboot.c:(.text+0x12d): undefined reference to memremap' /usr/bin/ld: memconsole-coreboot.c:(.text+0x17e): undefined reference to devm_memremap' /usr/bin/ld: memconsole-coreboot.c:(.text+0x191): undefined reference to memunmap' /usr/bin/ld: drivers/firmware/google/vpd.o: in function vpd_section_destroy.isra.0':
vpd.c:(.text+0x300): undefined reference to `memunmap'
/usr/bin/ld: drivers/firmware/google/vpd.o: in function `vpd_section_init':
vpd.c:(.text+0x382): undefined reference to `memremap'
/usr/bin/ld: vpd.c:(.text+0x459): undefined reference to memunmap' /usr/bin/ld: drivers/firmware/google/vpd.o: in function vpd_probe':
vpd.c:(.text+0x59d): undefined reference to `memremap'
/usr/bin/ld: vpd.c:(.text+0x5d3): undefined reference to `memunmap'
collect2: error: ld returned 1 exit status

Fixes a28aad66da8b ("firmware: coreboot: Collapse platform drivers into bus core")
---
 drivers/firmware/google/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/firmware/google/Kconfig +1 -1
index 931544c9f63d..983e07dc022e 100644
@@ -21,7 +21,7 @@ config GOOGLE_SMI
 
 config GOOGLE_COREBOOT_TABLE
 	tristate "Coreboot Table Access"
-	depends on ACPI || OF
+	depends on HAS_IOMEM && (ACPI || OF)
 	help
 	  This option enables the coreboot_table module, which provides other
 	  firmware modules access to the coreboot table. The coreboot table
Signature · 1 line
2.34.1
author maintainer
#719 lore raw

PATCH 711/1017 driver core: dd: fix return value of __setup handler

In reply to Greg Kroah-Hartman

From: Randy Dunlap <[email protected]>

[ Upstream commit f2aad54703dbe630f9d8b235eb58e8c8cc78f37d ]

When "driver_async_probe=nulltty" is used on the kernel boot command line, it causes an Unknown parameter message and the string is added to init's environment strings, polluting them.

Unknown kernel command line parameters "BOOT_IMAGE=/boot/bzImage-517rc6 driver_async_probe=nulltty", will be passed to user space.

Run /sbin/init as init process
with arguments:
/sbin/init
with environment:

     HOME=/
     TERM=linux
     BOOT_IMAGE=/boot/bzImage-517rc6
     driver_async_probe=nulltty

Change the return value of the __setup function to 1 to indicate that the __setup option has been handled.

Link lore.kernel.org/r/[email protected]
Fixes 1ea61b68d0f8 ("async: Add cmdline option to specify drivers to be async probed")
Cc Feng Tang <[email protected]>
Cc Greg Kroah-Hartman <[email protected]>
Cc "Rafael J. Wysocki" <[email protected]>
---
 drivers/base/dd.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/base/dd.c +1 -1
index 6b6630693201..64ce42b6c6b6 100644
@@ -809,7 +809,7 @@ static int __init save_async_options(char *buf)
 		pr_warn("Too long list of driver names for 'driver_async_probe'!\n");
 
 	strlcpy(async_probe_drv_names, buf, ASYNC_DRV_NAMES_MAX_LEN);
-	return 0;
+	return 1;
 }
 __setup("driver_async_probe=", save_async_options);
 
Signature · 1 line
2.34.1
author maintainer
#720 lore raw

PATCH 712/1017 jfs: fix divide error in dbNextAG

In reply to Greg Kroah-Hartman

From: Pavel Skripkin <[email protected]>

[ Upstream commit 2cc7cc01c15f57d056318c33705647f87dcd4aab ]

Syzbot reported divide error in dbNextAG(). The problem was in missing validation check for malicious image.

Syzbot crafted an image with bmp->db_numag equal to 0. There wasn't any validation checks, but dbNextAG() blindly use bmp->db_numag in divide expression

Fix it by validating bmp->db_numag in dbMount() and return an error if image is malicious

Fixes 1da177e4c3f4 ("Linux-2.6.12-rc2")
Reported-and-tested-by [email protected]
---
 fs/jfs/jfs_dmap.c | 7 +++++++
 1 file changed, 7 insertions(+)
fs/jfs/jfs_dmap.c +7 -0
index 91f4ec93dab1..d8502f4989d9 100644
@@ -148,6 +148,7 @@ static const s8 budtab[256] = {
  *	0	- success
  *	-ENOMEM	- insufficient memory
  *	-EIO	- i/o error
+ *	-EINVAL - wrong bmap data
  */
 int dbMount(struct inode *ipbmap)
 {
@@ -179,6 +180,12 @@ int dbMount(struct inode *ipbmap)
 	bmp->db_nfree = le64_to_cpu(dbmp_le->dn_nfree);
 	bmp->db_l2nbperpage = le32_to_cpu(dbmp_le->dn_l2nbperpage);
 	bmp->db_numag = le32_to_cpu(dbmp_le->dn_numag);
+	if (!bmp->db_numag) {
+		release_metapage(mp);
+		kfree(bmp);
+		return -EINVAL;
+	}
+
 	bmp->db_maxlevel = le32_to_cpu(dbmp_le->dn_maxlevel);
 	bmp->db_maxag = le32_to_cpu(dbmp_le->dn_maxag);
 	bmp->db_agpref = le32_to_cpu(dbmp_le->dn_agpref);
Signature · 1 line
2.34.1
author maintainer
#721 lore raw

PATCH 713/1017 SUNRPC/call_alloc: async tasks mustnt block waiting for memory

In reply to Greg Kroah-Hartman

From: NeilBrown <[email protected]>

[ Upstream commit c487216bec83b0c5a8803e5c61433d33ad7b104d ]

When memory is short, new worker threads cannot be created and we depend on the minimum one rpciod thread to be able to handle everything. So it must not block waiting for memory.

mempools are particularly a problem as memory can only be released back to the mempool by an async rpc task running. If all available workqueue threads are waiting on the mempool, no thread is available to return anything.

rpc_malloc() can block, and this might cause deadlocks.
So check RPC_IS_ASYNC(), rather than RPC_IS_SWAPPER() to determine if blocking is acceptable.

---
 net/sunrpc/sched.c              | 4 +++-
 net/sunrpc/xprtrdma/transport.c | 4 +++-
 2 files changed, 6 insertions(+), 2 deletions(-)
net/sunrpc/sched.c +3 -1
index e2c835482791..d5b6e897f5a5 100644
@@ -1023,8 +1023,10 @@ int rpc_malloc(struct rpc_task *task)
 	struct rpc_buffer *buf;
 	gfp_t gfp = GFP_NOFS;
 
+	if (RPC_IS_ASYNC(task))
+		gfp = GFP_NOWAIT | __GFP_NOWARN;
 	if (RPC_IS_SWAPPER(task))
-		gfp = __GFP_MEMALLOC | GFP_NOWAIT | __GFP_NOWARN;
+		gfp |= __GFP_MEMALLOC;
 
 	size += sizeof(struct rpc_buffer);
 	if (size <= RPC_BUFFER_MAXSIZE)
net/sunrpc/xprtrdma/transport.c +3 -1
index 16e5696314a4..a52277115500 100644
@@ -574,8 +574,10 @@ xprt_rdma_allocate(struct rpc_task *task)
 	gfp_t flags;
 
 	flags = RPCRDMA_DEF_GFP;
+	if (RPC_IS_ASYNC(task))
+		flags = GFP_NOWAIT | __GFP_NOWARN;
 	if (RPC_IS_SWAPPER(task))
-		flags = __GFP_MEMALLOC | GFP_NOWAIT | __GFP_NOWARN;
+		flags |= __GFP_MEMALLOC;
 
 	if (!rpcrdma_check_regbuf(r_xprt, req->rl_sendbuf, rqst->rq_callsize,
 				  flags))
Signature · 1 line
2.34.1
author maintainer
#722 lore raw

PATCH 714/1017 SUNRPC: improve swap handling: scheduling and PF_MEMALLOC

In reply to Greg Kroah-Hartman

From: NeilBrown <[email protected]>

[ Upstream commit 8db55a032ac7ac1ed7b98d6b1dc980e6378c652f ]

rpc tasks can be marked as RPC_TASK_SWAPPER. This causes GFP_MEMALLOC to be used for some allocations. This is needed in some cases, but not in all where it is currently provided, and in some where it isn't provided.

Currently *all* tasks associated with a rpc_client on which swap is enabled get the flag and hence some GFP_MEMALLOC support.

GFP_MEMALLOC is provided for ->buf_alloc() but only swap-writes need it. However xdr_alloc_bvec does not get GFP_MEMALLOC - though it often does need it.

xdr_alloc_bvec is called while the XPRT_LOCK is held. If this blocks, then it blocks all other queued tasks. So this allocation needs GFP_MEMALLOC for *all* requests, not just writes, when the xprt is used for any swap writes.

Similarly, if the transport is not connected, that will block all requests including swap writes, so memory allocations should get GFP_MEMALLOC if swap writes are possible.

So with this patch:
1/ we ONLY set RPC_TASK_SWAPPER for swap writes.
2/ __rpc_execute() sets PF_MEMALLOC while handling any task
with RPC_TASK_SWAPPER set, or when handling any task that holds the XPRT_LOCKED lock on an xprt used for swap. This removes the need for the RPC_IS_SWAPPER() test in ->buf_alloc handlers.
3/ xprt_prepare_transmit() sets PF_MEMALLOC after locking
any task to a swapper xprt. __rpc_execute() will clear it.
3/ PF_MEMALLOC is set for all the connect workers.

---
 fs/nfs/write.c                  |  2 ++
 net/sunrpc/clnt.c               |  2 --
 net/sunrpc/sched.c              | 20 +++++++++++++++++---
 net/sunrpc/xprt.c               |  3 +++
 net/sunrpc/xprtrdma/transport.c |  6 ++++--
 net/sunrpc/xprtsock.c           |  8 ++++++++
 6 files changed, 34 insertions(+), 7 deletions(-)
fs/nfs/write.c +2 -0
index 7a23b4644507..e86aff429993 100644
@@ -1411,6 +1411,8 @@ static void nfs_initiate_write(struct nfs_pgio_header *hdr,
 {
 	int priority = flush_task_priority(how);
 
+	if (IS_SWAPFILE(hdr->inode))
+		task_setup_data->flags |= RPC_TASK_SWAPPER;
 	task_setup_data->priority = priority;
 	rpc_ops->write_setup(hdr, msg, &task_setup_data->rpc_client);
 	trace_nfs_initiate_write(hdr);
net/sunrpc/clnt.c +0 -2
index c83fe618767c..5985b78eddf1 100644
@@ -1085,8 +1085,6 @@ void rpc_task_set_client(struct rpc_task *task, struct rpc_clnt *clnt)
 		task->tk_flags |= RPC_TASK_TIMEOUT;
 	if (clnt->cl_noretranstimeo)
 		task->tk_flags |= RPC_TASK_NO_RETRANS_TIMEOUT;
-	if (atomic_read(&clnt->cl_swapper))
-		task->tk_flags |= RPC_TASK_SWAPPER;
 	/* Add to the client's list of all tasks */
 	spin_lock(&clnt->cl_lock);
 	list_add_tail(&task->tk_task, &clnt->cl_tasks);
net/sunrpc/sched.c +17 -3
index d5b6e897f5a5..ae295844ac55 100644
@@ -876,6 +876,15 @@ void rpc_release_calldata(const struct rpc_call_ops *ops, void *calldata)
 		ops->rpc_release(calldata);
 }
 
+static bool xprt_needs_memalloc(struct rpc_xprt *xprt, struct rpc_task *tk)
+{
+	if (!xprt)
+		return false;
+	if (!atomic_read(&xprt->swapper))
+		return false;
+	return test_bit(XPRT_LOCKED, &xprt->state) && xprt->snd_task == tk;
+}
+
 /*
  * This is the RPC `scheduler' (or rather, the finite state machine).
  */
@@ -884,6 +893,7 @@ static void __rpc_execute(struct rpc_task *task)
 	struct rpc_wait_queue *queue;
 	int task_is_async = RPC_IS_ASYNC(task);
 	int status = 0;
+	unsigned long pflags = current->flags;
 
 	WARN_ON_ONCE(RPC_IS_QUEUED(task));
 	if (RPC_IS_QUEUED(task))
@@ -906,6 +916,10 @@ static void __rpc_execute(struct rpc_task *task)
 		}
 		if (!do_action)
 			break;
+		if (RPC_IS_SWAPPER(task) ||
+		    xprt_needs_memalloc(task->tk_xprt, task))
+			current->flags |= PF_MEMALLOC;
+
 		trace_rpc_task_run_action(task, do_action);
 		do_action(task);
 
@@ -943,7 +957,7 @@ static void __rpc_execute(struct rpc_task *task)
 		rpc_clear_running(task);
 		spin_unlock(&queue->lock);
 		if (task_is_async)
-			return;
+			goto out;
 
 		/* sync task: sleep here */
 		trace_rpc_task_sync_sleep(task, task->tk_action);
@@ -967,6 +981,8 @@ static void __rpc_execute(struct rpc_task *task)
 
 	/* Release all resources associated with the task */
 	rpc_release_task(task);
+out:
+	current_restore_flags(pflags, PF_MEMALLOC);
 }
 
 /*
@@ -1025,8 +1041,6 @@ int rpc_malloc(struct rpc_task *task)
 
 	if (RPC_IS_ASYNC(task))
 		gfp = GFP_NOWAIT | __GFP_NOWARN;
-	if (RPC_IS_SWAPPER(task))
-		gfp |= __GFP_MEMALLOC;
 
 	size += sizeof(struct rpc_buffer);
 	if (size <= RPC_BUFFER_MAXSIZE)
net/sunrpc/xprt.c +3 -0
index 5388263f8fc8..396a74974f60 100644
@@ -1503,6 +1503,9 @@ bool xprt_prepare_transmit(struct rpc_task *task)
 		return false;
 
 	}
+	if (atomic_read(&xprt->swapper))
+		/* This will be clear in __rpc_execute */
+		current->flags |= PF_MEMALLOC;
 	return true;
 }
 
net/sunrpc/xprtrdma/transport.c +4 -2
index a52277115500..6268af7e0310 100644
@@ -239,8 +239,11 @@ xprt_rdma_connect_worker(struct work_struct *work)
 	struct rpcrdma_xprt *r_xprt = container_of(work, struct rpcrdma_xprt,
 						   rx_connect_worker.work);
 	struct rpc_xprt *xprt = &r_xprt->rx_xprt;
+	unsigned int pflags = current->flags;
 	int rc;
 
+	if (atomic_read(&xprt->swapper))
+		current->flags |= PF_MEMALLOC;
 	rc = rpcrdma_xprt_connect(r_xprt);
 	xprt_clear_connecting(xprt);
 	if (!rc) {
@@ -254,6 +257,7 @@ xprt_rdma_connect_worker(struct work_struct *work)
 		rpcrdma_xprt_disconnect(r_xprt);
 	xprt_unlock_connect(xprt, r_xprt);
 	xprt_wake_pending_tasks(xprt, rc);
+	current_restore_flags(pflags, PF_MEMALLOC);
 }
 
 /**
@@ -576,8 +580,6 @@ xprt_rdma_allocate(struct rpc_task *task)
 	flags = RPCRDMA_DEF_GFP;
 	if (RPC_IS_ASYNC(task))
 		flags = GFP_NOWAIT | __GFP_NOWARN;
-	if (RPC_IS_SWAPPER(task))
-		flags |= __GFP_MEMALLOC;
 
 	if (!rpcrdma_check_regbuf(r_xprt, req->rl_sendbuf, rqst->rq_callsize,
 				  flags))
net/sunrpc/xprtsock.c +8 -0
index 056fa0230359..821eeea1c83b 100644
@@ -2070,7 +2070,10 @@ static void xs_udp_setup_socket(struct work_struct *work)
 	struct rpc_xprt *xprt = &transport->xprt;
 	struct socket *sock;
 	int status = -EIO;
+	unsigned int pflags = current->flags;
 
+	if (atomic_read(&xprt->swapper))
+		current->flags |= PF_MEMALLOC;
 	sock = xs_create_sock(xprt, transport,
 			xs_addr(xprt)->sa_family, SOCK_DGRAM,
 			IPPROTO_UDP, false);
@@ -2090,6 +2093,7 @@ static void xs_udp_setup_socket(struct work_struct *work)
 	xprt_clear_connecting(xprt);
 	xprt_unlock_connect(xprt, transport);
 	xprt_wake_pending_tasks(xprt, status);
+	current_restore_flags(pflags, PF_MEMALLOC);
 }
 
 /**
@@ -2249,7 +2253,10 @@ static void xs_tcp_setup_socket(struct work_struct *work)
 	struct socket *sock = transport->sock;
 	struct rpc_xprt *xprt = &transport->xprt;
 	int status;
+	unsigned int pflags = current->flags;
 
+	if (atomic_read(&xprt->swapper))
+		current->flags |= PF_MEMALLOC;
 	if (!sock) {
 		sock = xs_create_sock(xprt, transport,
 				xs_addr(xprt)->sa_family, SOCK_STREAM,
@@ -2314,6 +2321,7 @@ static void xs_tcp_setup_socket(struct work_struct *work)
 	xprt_clear_connecting(xprt);
 out_unlock:
 	xprt_unlock_connect(xprt, transport);
+	current_restore_flags(pflags, PF_MEMALLOC);
 }
 
 /**
Signature · 1 line
2.34.1
author maintainer
#723 lore raw

PATCH 715/1017 SUNRPC: Dont call connect() more than once on a TCP socket

In reply to Greg Kroah-Hartman

From: Trond Myklebust <[email protected]>

[ Upstream commit 89f42494f92f448747bd8a7ab1ae8b5d5520577d ]

Avoid socket state races due to repeated calls to ->connect() using the same socket. If connect() returns 0 due to the connection having completed, but we are in fact in a closing state, then we may leave the XPRT_CONNECTING flag set on the transport.

Fixes 3be232f11a3c ("SUNRPC: Prevent immediate close+reconnect")
---
 include/linux/sunrpc/xprtsock.h |  1 +
 net/sunrpc/xprtsock.c           | 22 ++++++++++++----------
 2 files changed, 13 insertions(+), 10 deletions(-)
include/linux/sunrpc/xprtsock.h +1 -0
index 3c1423ee74b4..fed813ffe7db 100644
@@ -88,5 +88,6 @@ struct sock_xprt {
 #define XPRT_SOCK_WAKE_WRITE	(5)
 #define XPRT_SOCK_WAKE_PENDING	(6)
 #define XPRT_SOCK_WAKE_DISCONNECT	(7)
+#define XPRT_SOCK_CONNECT_SENT	(8)
 
 #endif /* _LINUX_SUNRPC_XPRTSOCK_H */
net/sunrpc/xprtsock.c +12 -10
index 821eeea1c83b..37d961c1a5c9 100644
@@ -2257,10 +2257,15 @@ static void xs_tcp_setup_socket(struct work_struct *work)
 
 	if (atomic_read(&xprt->swapper))
 		current->flags |= PF_MEMALLOC;
-	if (!sock) {
-		sock = xs_create_sock(xprt, transport,
-				xs_addr(xprt)->sa_family, SOCK_STREAM,
-				IPPROTO_TCP, true);
+
+	if (xprt_connected(xprt))
+		goto out;
+	if (test_and_clear_bit(XPRT_SOCK_CONNECT_SENT,
+			       &transport->sock_state) ||
+	    !sock) {
+		xs_reset_transport(transport);
+		sock = xs_create_sock(xprt, transport, xs_addr(xprt)->sa_family,
+				      SOCK_STREAM, IPPROTO_TCP, true);
 		if (IS_ERR(sock)) {
 			xprt_wake_pending_tasks(xprt, PTR_ERR(sock));
 			goto out;
@@ -2284,6 +2289,7 @@ static void xs_tcp_setup_socket(struct work_struct *work)
 		fallthrough;
 	case -EINPROGRESS:
 		/* SYN_SENT! */
+		set_bit(XPRT_SOCK_CONNECT_SENT, &transport->sock_state);
 		if (xprt->reestablish_timeout < XS_TCP_INIT_REEST_TO)
 			xprt->reestablish_timeout = XS_TCP_INIT_REEST_TO;
 		fallthrough;
@@ -2345,13 +2351,9 @@ static void xs_connect(struct rpc_xprt *xprt, struct rpc_task *task)
 
 	WARN_ON_ONCE(!xprt_lock_connect(xprt, task, transport));
 
-	if (transport->sock != NULL && !xprt_connecting(xprt)) {
+	if (transport->sock != NULL) {
 		dprintk("RPC:       xs_connect delayed xprt %p for %lu "
-				"seconds\n",
-				xprt, xprt->reestablish_timeout / HZ);
-
-		/* Start by resetting any existing state */
-		xs_reset_transport(transport);
+			"seconds\n", xprt, xprt->reestablish_timeout / HZ);
 
 		delay = xprt_reconnect_delay(xprt);
 		xprt_reconnect_backoff(xprt, XS_TCP_INIT_REEST_TO);
Signature · 1 line
2.34.1
author maintainer
#724 lore raw

PATCH 716/1017 netfilter: egress: Report interface as outgoing

In reply to Greg Kroah-Hartman

From: Phil Sutter <[email protected]>

[ Upstream commit d645552e9bd96671079b27015294ec7f9748fa2b ]

Otherwise packets in egress chains seem like they are being received by the interface, not sent out via it.

Fixes 42df6e1d221dd ("netfilter: Introduce egress hook")
---
 include/linux/netfilter_netdev.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
include/linux/netfilter_netdev.h +1 -1
index e6487a691136..8676316547cc 100644
@@ -99,7 +99,7 @@ static inline struct sk_buff *nf_hook_egress(struct sk_buff *skb, int *rc,
 		return skb;
 
 	nf_hook_state_init(&state, NF_NETDEV_EGRESS,
-			   NFPROTO_NETDEV, dev, NULL, NULL,
+			   NFPROTO_NETDEV, NULL, dev, NULL,
 			   dev_net(dev), NULL);
 
 	/* nf assumes rcu_read_lock, not just read_lock_bh */
Signature · 1 line
2.34.1
author maintainer
#725 lore raw

PATCH 717/1017 netfilter: nf_conntrack_tcp: preserve liberal flag in tcp options

In reply to Greg Kroah-Hartman

From: Pablo Neira Ayuso <[email protected]>

[ Upstream commit f2dd495a8d589371289981d5ed33e6873df94ecc ]

Do not reset IP_CT_TCP_FLAG_BE_LIBERAL flag in out-of-sync scenarios coming before the TCP window tracking, otherwise such connections will fail in the window check.

Update tcp_options() to leave this flag in place and add a new helper function to reset the tcp window state.

Based on patch from Sven Auhagen.

Fixes c4832c7bbc3f ("netfilter: nf_ct_tcp: improve out-of-sync situation in TCP tracking")
---
 net/netfilter/nf_conntrack_proto_tcp.c | 17 +++++++++++++----
 1 file changed, 13 insertions(+), 4 deletions(-)
net/netfilter/nf_conntrack_proto_tcp.c +13 -4
index af5115e127cf..3cee5d8ee702 100644
@@ -341,8 +341,8 @@ static void tcp_options(const struct sk_buff *skb,
 	if (!ptr)
 		return;
 
-	state->td_scale =
-	state->flags = 0;
+	state->td_scale = 0;
+	state->flags &= IP_CT_TCP_FLAG_BE_LIBERAL;
 
 	while (length > 0) {
 		int opcode=*ptr++;
@@ -839,6 +839,16 @@ static bool tcp_can_early_drop(const struct nf_conn *ct)
 	return false;
 }
 
+static void nf_ct_tcp_state_reset(struct ip_ct_tcp_state *state)
+{
+	state->td_end		= 0;
+	state->td_maxend	= 0;
+	state->td_maxwin	= 0;
+	state->td_maxack	= 0;
+	state->td_scale		= 0;
+	state->flags		&= IP_CT_TCP_FLAG_BE_LIBERAL;
+}
+
 /* Returns verdict for packet, or -1 for invalid. */
 int nf_conntrack_tcp_packet(struct nf_conn *ct,
 			    struct sk_buff *skb,
@@ -945,8 +955,7 @@ int nf_conntrack_tcp_packet(struct nf_conn *ct,
 			ct->proto.tcp.last_flags &= ~IP_CT_EXP_CHALLENGE_ACK;
 			ct->proto.tcp.seen[ct->proto.tcp.last_dir].flags =
 				ct->proto.tcp.last_flags;
-			memset(&ct->proto.tcp.seen[dir], 0,
-			       sizeof(struct ip_ct_tcp_state));
+			nf_ct_tcp_state_reset(&ct->proto.tcp.seen[dir]);
 			break;
 		}
 		ct->proto.tcp.last_index = index;
Signature · 1 line
2.34.1
author maintainer
#726 lore raw

PATCH 718/1017 SUNRPC dont resend a task on an offlined transport

In reply to Greg Kroah-Hartman

From: Olga Kornievskaia <[email protected]>

[ Upstream commit 82ee41b85cef16b4be1f4732650012d9baaedddd ]

When a task is being retried, due to an NFS error, if the assigned transport has been put offline and the task is relocatable pick a new transport.

Fixes 6f081693e7b2b ("sunrpc: remove an offlined xprt using sysfs")
---
 net/sunrpc/clnt.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
net/sunrpc/clnt.c +3 -1
index 5985b78eddf1..b36d235d2d6d 100644
@@ -1065,7 +1065,9 @@ rpc_task_get_next_xprt(struct rpc_clnt *clnt)
 static
 void rpc_task_set_transport(struct rpc_task *task, struct rpc_clnt *clnt)
 {
-	if (task->tk_xprt)
+	if (task->tk_xprt &&
+			!(test_bit(XPRT_OFFLINE, &task->tk_xprt->state) &&
+                        (task->tk_flags & RPC_TASK_MOVEABLE)))
 		return;
 	if (task->tk_flags & RPC_TASK_NO_ROUND_ROBIN)
 		task->tk_xprt = rpc_task_get_first_xprt(clnt);
Signature · 1 line
2.34.1
author maintainer
#727 lore raw

PATCH 719/1017 NFSv4.1: dont retry BIND_CONN_TO_SESSION on session error

In reply to Greg Kroah-Hartman

From: Olga Kornievskaia <[email protected]>

[ Upstream commit 1d15d121cc2ad4d016a7dc1493132a9696f91fc5 ]

There is no reason to retry the operation if a session error had occurred in such case result structure isn't filled out.

Fixes dff58530c4ca ("NFSv4.1: fix handling of backchannel binding in BIND_CONN_TO_SESSION")
---
 fs/nfs/nfs4proc.c | 1 +
 1 file changed, 1 insertion(+)
fs/nfs/nfs4proc.c +1 -0
index 0abbbf5d2bdf..5298c3a278c5 100644
@@ -8275,6 +8275,7 @@ nfs4_bind_one_conn_to_session_done(struct rpc_task *task, void *calldata)
 	case -NFS4ERR_DEADSESSION:
 		nfs4_schedule_session_recovery(clp->cl_session,
 				task->tk_status);
+		return;
 	}
 	if (args->dir == NFS4_CDFC4_FORE_OR_BOTH &&
 			res->dir != NFS4_CDFS4_BOTH) {
Signature · 1 line
2.34.1
author maintainer
#728 lore raw

PATCH 720/1017 kdb: Fix the putarea helper function

In reply to Greg Kroah-Hartman

From: Daniel Thompson <[email protected]>

[ Upstream commit c1cb81429df462eca1b6ba615cddd21dd3103c46 ]

Currently kdb_putarea_size() uses copy_from_kernel_nofault() to write *to* arbitrary kernel memory. This is obviously wrong and means the memory modify ('mm') command is a serious risk to debugger stability: if we poke to a bad address we'll double-fault and lose our debug session.

Fix this the (very) obvious way.

Note that there are two Fixes: tags because the API was renamed and this patch will only trivially backport as far as the rename (and this is probably enough). Nevertheless Christoph's rename did not introduce this problem so I wanted to record that!

Fixes fe557319aa06 ("maccess: rename probe_kernel_{read,write} to copy_{from,to}_kernel_nofault")
Fixes 5d5314d6795f ("kdb: core for kgdb back end (1 of 2)")
---
 kernel/debug/kdb/kdb_support.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
kernel/debug/kdb/kdb_support.c +1 -1
index df2bface866e..85cb51c4a17e 100644
@@ -291,7 +291,7 @@ int kdb_getarea_size(void *res, unsigned long addr, size_t size)
  */
 int kdb_putarea_size(unsigned long addr, void *res, size_t size)
 {
-	int ret = copy_from_kernel_nofault((char *)addr, (char *)res, size);
+	int ret = copy_to_kernel_nofault((char *)addr, (char *)res, size);
 	if (ret) {
 		if (!KDB_STATE(SUPPRESS)) {
 			kdb_func_printf("Bad address 0x%lx\n", addr);
Signature · 1 line
2.34.1
author maintainer
#729 lore raw

PATCH 721/1017 perf stat: Fix forked applications enablement of counters

In reply to Greg Kroah-Hartman

From: Thomas Richter <[email protected]>

[ Upstream commit d0a0a511493d269514fcbd852481cdca32c95350 ]

I have run into the following issue:

 # perf stat -a -e new_pmu/INSTRUCTION_7/ --  mytest -c1 7

Performance counter stats for 'system wide':

                 0      new_pmu/INSTRUCTION_7/

       0.000366428 seconds time elapsed

#

The new PMU for s390 counts the execution of certain CPU instructions. The root cause is the extremely small run time of the mytest program. It just executes some assembly instructions and then exits.

In above invocation the instruction is executed exactly one time (-c1 option). The PMU is expected to report this one time execution by a counter value of one, but fails to do so in some cases, not all.

Debugging reveals the invocation of the child process is done *before* the counter events are installed and enabled.

Tracing reveals that sometimes the child process starts and exits before the event is installed on all CPUs. The more CPUs the machine has, the more often this miscount happens.

Fix this by reversing the start of the work load after the events have been installed on the specified CPUs. Now the comment also matches the code.

Output after:

 # perf stat -a -e new_pmu/INSTRUCTION_7/ --  mytest -c1 7

Performance counter stats for 'system wide':

                 1      new_pmu/INSTRUCTION_7/

       0.000366428 seconds time elapsed

#

Now the correct result is reported rock solid all the time regardless how many CPUs are online.

Reviewers notes:

Jiri:

Right, without -a the event has enable_on_exec so the race does not matter, but it's a problem for system wide with fork.

Namhyung:

Agreed. Also we may move the enable_counters() and the clock code out of the if block to be shared with the else block.

---
 tools/perf/builtin-stat.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
tools/perf/builtin-stat.c +1 -1
index 7974933dbc77..08c024038ca7 100644
@@ -956,10 +956,10 @@ static int __run_perf_stat(int argc, const char **argv, int run_idx)
 	 * Enable counters and exec the command:
 	 */
 	if (forks) {
-		evlist__start_workload(evsel_list);
 		err = enable_counters();
 		if (err)
 			return -1;
+		evlist__start_workload(evsel_list);
 
 		t0 = rdclock();
 		clock_gettime(CLOCK_MONOTONIC, &ref_time);
Signature · 1 line
2.34.1
author maintainer
#730 lore raw

PATCH 722/1017 clk: qcom: gcc-msm8994: Fix gpll4 width

In reply to Greg Kroah-Hartman

From: Konrad Dybcio <[email protected]>

[ Upstream commit 71021db1c532c2545ae53b9ee85b37b7154f51d4 ]

The gpll4 postdiv is actually a div4, so make sure that Linux is aware of this.

This fixes the following error messages:

mmc1: Card appears overclocked; req 200000000 Hz, actual 343999999 Hz
mmc1: Card appears overclocked; req 400000000 Hz, actual 687999999 Hz

Fixes aec89f78cf01 ("clk: qcom: Add support for msm8994 global clock controller")
---
 drivers/clk/qcom/gcc-msm8994.c | 1 +
 1 file changed, 1 insertion(+)
drivers/clk/qcom/gcc-msm8994.c +1 -0
index 5df9f1ead48e..4a9eb845b050 100644
@@ -76,6 +76,7 @@ static struct clk_alpha_pll gpll4_early = {
 
 static struct clk_alpha_pll_postdiv gpll4 = {
 	.offset = 0x1dc0,
+	.width = 4,
 	.regs = clk_alpha_pll_regs[CLK_ALPHA_PLL_TYPE_DEFAULT],
 	.clkr.hw.init = &(struct clk_init_data){
 		.name = "gpll4",
Signature · 1 line
2.34.1
author maintainer
#731 lore raw

PATCH 723/1017 vsock/virtio: initialize vdev->priv before using VQs

In reply to Greg Kroah-Hartman

From: Stefano Garzarella <[email protected]>

[ Upstream commit 4b5f1ad5566ada230aaa2ce861b28d1895f1ea68 ]

When we fill VQs with empty buffers and kick the host, it may send an interrupt. vdev->priv must be initialized before this since it is used in the virtqueue callbacks.

Fixes 0deab087b16a ("vsock/virtio: use RCU to avoid use-after-free on the_virtio_vsock")
---
 net/vmw_vsock/virtio_transport.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
net/vmw_vsock/virtio_transport.c +2 -1
index dad9ca65f4f9..fb1b8f99f679 100644
@@ -622,6 +622,8 @@ static int virtio_vsock_probe(struct virtio_device *vdev)
 	INIT_WORK(&vsock->event_work, virtio_transport_event_work);
 	INIT_WORK(&vsock->send_pkt_work, virtio_transport_send_pkt_work);
 
+	vdev->priv = vsock;
+
 	mutex_lock(&vsock->tx_lock);
 	vsock->tx_run = true;
 	mutex_unlock(&vsock->tx_lock);
@@ -639,7 +641,6 @@ static int virtio_vsock_probe(struct virtio_device *vdev)
 	if (virtio_has_feature(vdev, VIRTIO_VSOCK_F_SEQPACKET))
 		vsock->seqpacket_allow = true;
 
-	vdev->priv = vsock;
 	rcu_assign_pointer(the_virtio_vsock, vsock);
 
 	mutex_unlock(&the_virtio_vsock_mutex);
Signature · 1 line
2.34.1
author maintainer
#732 lore raw

PATCH 724/1017 vsock/virtio: read the negotiated features before using VQs

In reply to Greg Kroah-Hartman

From: Stefano Garzarella <[email protected]>

[ Upstream commit c1011c0b3a9c8d2065f425407475cbcc812540b7 ]

Complete the driver configuration, reading the negotiated features, before using the VQs in the virtio_vsock_probe().

---
 net/vmw_vsock/virtio_transport.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
net/vmw_vsock/virtio_transport.c +3 -3
index fb1b8f99f679..4a8548bdf86c 100644
@@ -622,6 +622,9 @@ static int virtio_vsock_probe(struct virtio_device *vdev)
 	INIT_WORK(&vsock->event_work, virtio_transport_event_work);
 	INIT_WORK(&vsock->send_pkt_work, virtio_transport_send_pkt_work);
 
+	if (virtio_has_feature(vdev, VIRTIO_VSOCK_F_SEQPACKET))
+		vsock->seqpacket_allow = true;
+
 	vdev->priv = vsock;
 
 	mutex_lock(&vsock->tx_lock);
@@ -638,9 +641,6 @@ static int virtio_vsock_probe(struct virtio_device *vdev)
 	vsock->event_run = true;
 	mutex_unlock(&vsock->event_lock);
 
-	if (virtio_has_feature(vdev, VIRTIO_VSOCK_F_SEQPACKET))
-		vsock->seqpacket_allow = true;
-
 	rcu_assign_pointer(the_virtio_vsock, vsock);
 
 	mutex_unlock(&the_virtio_vsock_mutex);
Signature · 1 line
2.34.1
author maintainer
#733 lore raw

PATCH 725/1017 vsock/virtio: enable VQs early on probe

In reply to Greg Kroah-Hartman

From: Stefano Garzarella <[email protected]>

[ Upstream commit 88704454ef8b00ea91537ae0d47d9348077e0e72 ]

virtio spec requires drivers to set DRIVER_OK before using VQs. This is set automatically after probe returns, but virtio-vsock driver uses VQs in the probe function to fill rx and event VQs with new buffers.

Let's fix this, calling virtio_device_ready() before using VQs in the probe function.

Fixes 0ea9e1d3a9e3 ("VSOCK: Introduce virtio_transport.ko")
---
 net/vmw_vsock/virtio_transport.c | 2 ++
 1 file changed, 2 insertions(+)
net/vmw_vsock/virtio_transport.c +2 -0
index 4a8548bdf86c..c5f936fbf876 100644
@@ -627,6 +627,8 @@ static int virtio_vsock_probe(struct virtio_device *vdev)
 
 	vdev->priv = vsock;
 
+	virtio_device_ready(vdev);
+
 	mutex_lock(&vsock->tx_lock);
 	vsock->tx_run = true;
 	mutex_unlock(&vsock->tx_lock);
Signature · 1 line
2.34.1
author maintainer
#734 lore raw

PATCH 726/1017 clk: Initialize orphan req_rate

In reply to Greg Kroah-Hartman

From: Maxime Ripard <[email protected]>

[ Upstream commit 5f7e2af00807f2117650e711a58b7f0e986ce1df ]

When registering a clock that doesn't have a recalc_rate implementation, and doesn't have its parent registered yet, we initialize the clk_core rate and 'req_rate' fields to 0.

The rate field is later updated when the parent is registered in clk_core_reparent_orphans_nolock() using __clk_recalc_rates(), but the 'req_rate' field is never updated.

This leads to an issue in clk_set_rate_range() and clk_put(), since those functions will call clk_set_rate() with the content of 'req_rate' to provide drivers with the opportunity to change the rate based on the new boundaries. In this case, we would call clk_set_rate() with a rate of 0, effectively enforcing the minimum allowed for this clock whenever we would call one of those two functions, even though the actual rate might be within range.

Let's fix this by setting 'req_rate' in clk_core_reparent_orphans_nolock() with the rate field content just updated by the call to __clk_recalc_rates().

---
 drivers/clk/clk.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)
drivers/clk/clk.c +13 -0
index e55ddc12ed53..2c5db2df9d42 100644
@@ -3413,6 +3413,19 @@ static void clk_core_reparent_orphans_nolock(void)
 			__clk_set_parent_after(orphan, parent, NULL);
 			__clk_recalc_accuracies(orphan);
 			__clk_recalc_rates(orphan, 0);
+
+			/*
+			 * __clk_init_parent() will set the initial req_rate to
+			 * 0 if the clock doesn't have clk_ops::recalc_rate and
+			 * is an orphan when it's registered.
+			 *
+			 * 'req_rate' is used by clk_set_rate_range() and
+			 * clk_put() to trigger a clk_set_rate() call whenever
+			 * the boundaries are modified. Let's make sure
+			 * 'req_rate' is set to something non-zero so that
+			 * clk_set_rate_range() doesn't drop the frequency.
+			 */
+			orphan->req_rate = orphan->rate;
 		}
 	}
 }
Signature · 1 line
2.34.1
author maintainer
#735 lore raw

PATCH 727/1017 xen: fix is_xen_pmu()

In reply to Greg Kroah-Hartman

From: Juergen Gross <[email protected]>

[ Upstream commit de2ae403b4c0e79a3410e63bc448542fbb9f9bfc ]

is_xen_pmu() is taking the cpu number as parameter, but it is not using it. Instead it just tests whether the Xen PMU initialization on the current cpu did succeed. As this test is done by checking a percpu pointer, preemption needs to be disabled in order to avoid switching the cpu while doing the test. While resuming from suspend() this seems not to be the case:

[   88.082751] ACPI: PM: Low-level resume complete
[   88.087933] ACPI: EC: EC started
[   88.091464] ACPI: PM: Restoring platform NVS memory
[   88.097166] xen_acpi_processor: Uploading Xen processor PM info
[   88.103850] Enabling non-boot CPUs ...
[   88.108128] installing Xen timer for CPU 1
[   88.112763] BUG: using smp_processor_id() in preemptible [00000000] code: systemd-sleep/7138
[   88.122256] caller is is_xen_pmu+0x12/0x30
[   88.126937] CPU: 0 PID: 7138 Comm: systemd-sleep Tainted: G        W         5.16.13-2.fc32.qubes.x86_64 #1
[   88.137939] Hardware name: Star Labs StarBook/StarBook, BIOS 7.97 03/21/2022
[   88.145930] Call Trace:
[   88.148757]  <TASK>
[   88.151193]  dump_stack_lvl+0x48/0x5e
[   88.155381]  check_preemption_disabled+0xde/0xe0
[   88.160641]  is_xen_pmu+0x12/0x30
[   88.164441]  xen_smp_intr_init_pv+0x75/0x100

Fix that by replacing is_xen_pmu() by a simple boolean variable which reflects the Xen PMU initialization state on cpu 0.

Modify xen_pmu_init() to return early in case it is being called for a cpu other than cpu 0 and the boolean variable not being set.

---
 arch/x86/xen/pmu.c    | 10 ++++------
 arch/x86/xen/pmu.h    |  3 ++-
 arch/x86/xen/smp_pv.c |  2 +-
 3 files changed, 7 insertions(+), 8 deletions(-)
arch/x86/xen/pmu.c +4 -6
index e13b0b49fcdf..d7249f4c90f1 100644
@@ -512,10 +512,7 @@ irqreturn_t xen_pmu_irq_handler(int irq, void *dev_id)
 	return ret;
 }
 
-bool is_xen_pmu(int cpu)
-{
-	return (get_xenpmu_data() != NULL);
-}
+bool is_xen_pmu;
 
 void xen_pmu_init(int cpu)
 {
@@ -526,7 +523,7 @@ void xen_pmu_init(int cpu)
 
 	BUILD_BUG_ON(sizeof(struct xen_pmu_data) > PAGE_SIZE);
 
-	if (xen_hvm_domain())
+	if (xen_hvm_domain() || (cpu != 0 && !is_xen_pmu))
 		return;
 
 	xenpmu_data = (struct xen_pmu_data *)get_zeroed_page(GFP_KERNEL);
@@ -547,7 +544,8 @@ void xen_pmu_init(int cpu)
 	per_cpu(xenpmu_shared, cpu).xenpmu_data = xenpmu_data;
 	per_cpu(xenpmu_shared, cpu).flags = 0;
 
-	if (cpu == 0) {
+	if (!is_xen_pmu) {
+		is_xen_pmu = true;
 		perf_register_guest_info_callbacks(&xen_guest_cbs);
 		xen_pmu_arch_init();
 	}
arch/x86/xen/pmu.h +2 -1
index 0e83a160589b..65c58894fc79 100644
@@ -4,6 +4,8 @@
 
 #include <xen/interface/xenpmu.h>
 
+extern bool is_xen_pmu;
+
 irqreturn_t xen_pmu_irq_handler(int irq, void *dev_id);
 #ifdef CONFIG_XEN_HAVE_VPMU
 void xen_pmu_init(int cpu);
@@ -12,7 +14,6 @@ void xen_pmu_finish(int cpu);
 static inline void xen_pmu_init(int cpu) {}
 static inline void xen_pmu_finish(int cpu) {}
 #endif
-bool is_xen_pmu(int cpu);
 bool pmu_msr_read(unsigned int msr, uint64_t *val, int *err);
 bool pmu_msr_write(unsigned int msr, uint32_t low, uint32_t high, int *err);
 int pmu_apic_update(uint32_t reg);
arch/x86/xen/smp_pv.c +1 -1
index 4a6019238ee7..688aa8b6ae29 100644
@@ -129,7 +129,7 @@ int xen_smp_intr_init_pv(unsigned int cpu)
 	per_cpu(xen_irq_work, cpu).irq = rc;
 	per_cpu(xen_irq_work, cpu).name = callfunc_name;
 
-	if (is_xen_pmu(cpu)) {
+	if (is_xen_pmu) {
 		pmu_name = kasprintf(GFP_KERNEL, "pmu%d", cpu);
 		rc = bind_virq_to_irqhandler(VIRQ_XENPMU, cpu,
 					     xen_pmu_irq_handler,
Signature · 1 line
2.34.1
author maintainer
#736 lore raw

PATCH 728/1017 net: enetc: report software timestamping via SO_TIMESTAMPING

In reply to Greg Kroah-Hartman

From: Vladimir Oltean <[email protected]>

[ Upstream commit feb13dcb1818b775fbd9191f797be67cd605f03e ]

Let user space properly determine that the enetc driver provides software timestamps.

---
 drivers/net/ethernet/freescale/enetc/enetc_ethtool.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/net/ethernet/freescale/enetc/enetc_ethtool.c +4 -1
index 910b9f722504..d62c188c8748 100644
@@ -672,7 +672,10 @@ static int enetc_get_ts_info(struct net_device *ndev,
 #ifdef CONFIG_FSL_ENETC_PTP_CLOCK
 	info->so_timestamping = SOF_TIMESTAMPING_TX_HARDWARE |
 				SOF_TIMESTAMPING_RX_HARDWARE |
-				SOF_TIMESTAMPING_RAW_HARDWARE;
+				SOF_TIMESTAMPING_RAW_HARDWARE |
+				SOF_TIMESTAMPING_TX_SOFTWARE |
+				SOF_TIMESTAMPING_RX_SOFTWARE |
+				SOF_TIMESTAMPING_SOFTWARE;
 
 	info->tx_types = (1 << HWTSTAMP_TX_OFF) |
 			 (1 << HWTSTAMP_TX_ON) |
Signature · 1 line
2.34.1
author maintainer
#737 lore raw

PATCH 729/1017 net: hns3: fix bug when PF set the duplicate MAC address for VFs

In reply to Greg Kroah-Hartman

From: Jian Shen <[email protected]>

[ Upstream commit ccb18f05535c96d26e2d559d402acb87700fc5a7 ]

If the MAC address A is configured to vport A and then vport B. The MAC address of vport A in the hardware becomes invalid. If the address of vport A is changed to MAC address B, the driver needs to delete the MAC address A of vport A. Due to the MAC address A of vport A has become invalid in the hardware entry, so "-ENOENT" is returned. In this case, the "used_umv_size" value recorded in driver is not updated. As a result, the MAC entry status of the software is inconsistent with that of the hardware.

Therefore, the driver updates the umv size even if the MAC entry cannot be found. Ensure that the software and hardware status is consistent.

Fixes ee4bcd3b7ae4 ("net: hns3: refactor the MAC address configure")
---
 drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c +2 -3
index c2a58101144e..0fc8810671dc 100644
@@ -8888,12 +8888,11 @@ int hclge_rm_uc_addr_common(struct hclge_vport *vport,
 	hnae3_set_bit(req.entry_type, HCLGE_MAC_VLAN_BIT0_EN_B, 0);
 	hclge_prepare_mac_addr(&req, addr, false);
 	ret = hclge_remove_mac_vlan_tbl(vport, &req);
-	if (!ret) {
+	if (!ret || ret == -ENOENT) {
 		mutex_lock(&hdev->vport_lock);
 		hclge_update_umv_space(vport, true);
 		mutex_unlock(&hdev->vport_lock);
-	} else if (ret == -ENOENT) {
-		ret = 0;
+		return 0;
 	}
 
 	return ret;
Signature · 1 line
2.34.1
author maintainer
#738 lore raw

PATCH 730/1017 net: hns3: fix port base vlan add fail when concurrent with reset

In reply to Greg Kroah-Hartman

From: Jian Shen <[email protected]>

[ Upstream commit c0f46de30c965d4ba208b5bf1a6d3437a7556ee2 ]

Currently, Port base vlan is initiated by PF and configured to its VFs, by using command "ip link set <pf name> vf <vf id> vlan <vlan id>". When a global reset was triggered, the hardware vlan table and the soft recorded vlan information will be cleared by PF, and restored them until VFs were ready. There is a short time window between the table had been cleared and before table restored. If configured a new port base vlan tag at this moment, driver will check the soft recorded vlan information, and find there hasn't the old tag in it, which causing a warning print.

Due to the port base vlan is managed by PF, so the VFs's port base vlan restoring should be handled by PF when PF was ready.

This patch fixes it.

Fixes 039ba863e8d7 ("net: hns3: optimize the filter table entries handling when resetting")
---
 .../hisilicon/hns3/hns3pf/hclge_main.c        | 62 +++++++++++++------
 .../hisilicon/hns3/hns3pf/hclge_main.h        |  3 +
 2 files changed, 46 insertions(+), 19 deletions(-)
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c +43 -19
index 0fc8810671dc..0338b4ffa003 100644
@@ -1956,6 +1956,7 @@ static int hclge_alloc_vport(struct hclge_dev *hdev)
 		vport->vf_info.link_state = IFLA_VF_LINK_STATE_AUTO;
 		vport->mps = HCLGE_MAC_DEFAULT_FRAME;
 		vport->port_base_vlan_cfg.state = HNAE3_PORT_BASE_VLAN_DISABLE;
+		vport->port_base_vlan_cfg.tbl_sta = true;
 		vport->rxvlan_cfg.rx_vlan_offload_en = true;
 		vport->req_vlan_fltr_en = true;
 		INIT_LIST_HEAD(&vport->vlan_list);
@@ -10334,34 +10335,52 @@ void hclge_uninit_vport_vlan_table(struct hclge_dev *hdev)
 	}
 }
 
-void hclge_restore_vport_vlan_table(struct hclge_vport *vport)
+void hclge_restore_vport_port_base_vlan_config(struct hclge_dev *hdev)
 {
-	struct hclge_vport_vlan_cfg *vlan, *tmp;
-	struct hclge_dev *hdev = vport->back;
+	struct hclge_vlan_info *vlan_info;
+	struct hclge_vport *vport;
 	u16 vlan_proto;
 	u16 vlan_id;
 	u16 state;
+	int vf_id;
 	int ret;
 
-	vlan_proto = vport->port_base_vlan_cfg.vlan_info.vlan_proto;
-	vlan_id = vport->port_base_vlan_cfg.vlan_info.vlan_tag;
-	state = vport->port_base_vlan_cfg.state;
+	/* PF should restore all vfs port base vlan */
+	for (vf_id = 0; vf_id < hdev->num_alloc_vfs; vf_id++) {
+		vport = &hdev->vport[vf_id + HCLGE_VF_VPORT_START_NUM];
+		vlan_info = vport->port_base_vlan_cfg.tbl_sta ?
+			    &vport->port_base_vlan_cfg.vlan_info :
+			    &vport->port_base_vlan_cfg.old_vlan_info;
 
-	if (state != HNAE3_PORT_BASE_VLAN_DISABLE) {
-		clear_bit(vport->vport_id, hdev->vlan_table[vlan_id]);
-		hclge_set_vlan_filter_hw(hdev, htons(vlan_proto),
-					 vport->vport_id, vlan_id,
-					 false);
-		return;
+		vlan_id = vlan_info->vlan_tag;
+		vlan_proto = vlan_info->vlan_proto;
+		state = vport->port_base_vlan_cfg.state;
+
+		if (state != HNAE3_PORT_BASE_VLAN_DISABLE) {
+			clear_bit(vport->vport_id, hdev->vlan_table[vlan_id]);
+			ret = hclge_set_vlan_filter_hw(hdev, htons(vlan_proto),
+						       vport->vport_id,
+						       vlan_id, false);
+			vport->port_base_vlan_cfg.tbl_sta = ret == 0;
+		}
 	}
+}
 
-	list_for_each_entry_safe(vlan, tmp, &vport->vlan_list, node) {
-		ret = hclge_set_vlan_filter_hw(hdev, htons(ETH_P_8021Q),
-					       vport->vport_id,
-					       vlan->vlan_id, false);
-		if (ret)
-			break;
-		vlan->hd_tbl_status = true;
+void hclge_restore_vport_vlan_table(struct hclge_vport *vport)
+{
+	struct hclge_vport_vlan_cfg *vlan, *tmp;
+	struct hclge_dev *hdev = vport->back;
+	int ret;
+
+	if (vport->port_base_vlan_cfg.state == HNAE3_PORT_BASE_VLAN_DISABLE) {
+		list_for_each_entry_safe(vlan, tmp, &vport->vlan_list, node) {
+			ret = hclge_set_vlan_filter_hw(hdev, htons(ETH_P_8021Q),
+						       vport->vport_id,
+						       vlan->vlan_id, false);
+			if (ret)
+				break;
+			vlan->hd_tbl_status = true;
+		}
 	}
 }
 
@@ -10402,6 +10421,7 @@ static void hclge_restore_hw_table(struct hclge_dev *hdev)
 	struct hnae3_handle *handle = &vport->nic;
 
 	hclge_restore_mac_table_common(vport);
+	hclge_restore_vport_port_base_vlan_config(hdev);
 	hclge_restore_vport_vlan_table(vport);
 	set_bit(HCLGE_STATE_FD_USER_DEF_CHANGED, &hdev->state);
 	hclge_restore_fd_entries(handle);
@@ -10458,6 +10478,8 @@ static int hclge_update_vlan_filter_entries(struct hclge_vport *vport,
 						 false);
 	}
 
+	vport->port_base_vlan_cfg.tbl_sta = false;
+
 	/* force add VLAN 0 */
 	ret = hclge_set_vf_vlan_common(hdev, vport->vport_id, false, 0);
 	if (ret)
@@ -10544,7 +10566,9 @@ int hclge_update_port_base_vlan_cfg(struct hclge_vport *vport, u16 state,
 	else
 		nic->port_base_vlan_state = HNAE3_PORT_BASE_VLAN_ENABLE;
 
+	vport->port_base_vlan_cfg.old_vlan_info = *old_vlan_info;
 	vport->port_base_vlan_cfg.vlan_info = *vlan_info;
+	vport->port_base_vlan_cfg.tbl_sta = true;
 	hclge_set_vport_vlan_fltr_change(vport);
 
 	return 0;
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h +3 -0
index ebba603483a0..801268b61e96 100644
@@ -1030,7 +1030,9 @@ struct hclge_vlan_info {
 
 struct hclge_port_base_vlan_config {
 	u16 state;
+	bool tbl_sta;
 	struct hclge_vlan_info vlan_info;
+	struct hclge_vlan_info old_vlan_info;
 };
 
 struct hclge_vf_info {
@@ -1154,6 +1156,7 @@ void hclge_rm_vport_all_mac_table(struct hclge_vport *vport, bool is_del_list,
 void hclge_rm_vport_all_vlan_table(struct hclge_vport *vport, bool is_del_list);
 void hclge_uninit_vport_vlan_table(struct hclge_dev *hdev);
 void hclge_restore_mac_table_common(struct hclge_vport *vport);
+void hclge_restore_vport_port_base_vlan_config(struct hclge_dev *hdev);
 void hclge_restore_vport_vlan_table(struct hclge_vport *vport);
 int hclge_update_port_base_vlan_cfg(struct hclge_vport *vport, u16 state,
 				    struct hclge_vlan_info *vlan_info);
Signature · 1 line
2.34.1
author maintainer
#739 lore raw

PATCH 731/1017 net: hns3: add vlan list lock to protect vlan list

In reply to Greg Kroah-Hartman

From: Jian Shen <[email protected]>

[ Upstream commit 1932a624ab88ff407d1a1d567fe581faa15dc725 ]

When adding port base VLAN, vf VLAN need to remove from HW and modify the vlan state in vf VLAN list as false. If the periodicity task is freeing the same node, it may cause "use after free" error. This patch adds a vlan list lock to protect the vlan list.

Fixes c6075b193462 ("net: hns3: Record VF vlan tables")
---
 .../hisilicon/hns3/hns3pf/hclge_main.c        | 38 +++++++++++++++++--
 .../hisilicon/hns3/hns3pf/hclge_main.h        |  1 +
 2 files changed, 35 insertions(+), 4 deletions(-)
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c +34 -4
index 0338b4ffa003..3ba24c6d004d 100644
@@ -10237,19 +10237,28 @@ static void hclge_add_vport_vlan_table(struct hclge_vport *vport, u16 vlan_id,
 				       bool writen_to_tbl)
 {
 	struct hclge_vport_vlan_cfg *vlan, *tmp;
+	struct hclge_dev *hdev = vport->back;
 
-	list_for_each_entry_safe(vlan, tmp, &vport->vlan_list, node)
-		if (vlan->vlan_id == vlan_id)
+	mutex_lock(&hdev->vport_lock);
+
+	list_for_each_entry_safe(vlan, tmp, &vport->vlan_list, node) {
+		if (vlan->vlan_id == vlan_id) {
+			mutex_unlock(&hdev->vport_lock);
 			return;
+		}
+	}
 
 	vlan = kzalloc(sizeof(*vlan), GFP_KERNEL);
-	if (!vlan)
+	if (!vlan) {
+		mutex_unlock(&hdev->vport_lock);
 		return;
+	}
 
 	vlan->hd_tbl_status = writen_to_tbl;
 	vlan->vlan_id = vlan_id;
 
 	list_add_tail(&vlan->node, &vport->vlan_list);
+	mutex_unlock(&hdev->vport_lock);
 }
 
 static int hclge_add_vport_all_vlan_table(struct hclge_vport *vport)
@@ -10258,6 +10267,8 @@ static int hclge_add_vport_all_vlan_table(struct hclge_vport *vport)
 	struct hclge_dev *hdev = vport->back;
 	int ret;
 
+	mutex_lock(&hdev->vport_lock);
+
 	list_for_each_entry_safe(vlan, tmp, &vport->vlan_list, node) {
 		if (!vlan->hd_tbl_status) {
 			ret = hclge_set_vlan_filter_hw(hdev, htons(ETH_P_8021Q),
@@ -10267,12 +10278,16 @@ static int hclge_add_vport_all_vlan_table(struct hclge_vport *vport)
 				dev_err(&hdev->pdev->dev,
 					"restore vport vlan list failed, ret=%d\n",
 					ret);
+
+				mutex_unlock(&hdev->vport_lock);
 				return ret;
 			}
 		}
 		vlan->hd_tbl_status = true;
 	}
 
+	mutex_unlock(&hdev->vport_lock);
+
 	return 0;
 }
 
@@ -10282,6 +10297,8 @@ static void hclge_rm_vport_vlan_table(struct hclge_vport *vport, u16 vlan_id,
 	struct hclge_vport_vlan_cfg *vlan, *tmp;
 	struct hclge_dev *hdev = vport->back;
 
+	mutex_lock(&hdev->vport_lock);
+
 	list_for_each_entry_safe(vlan, tmp, &vport->vlan_list, node) {
 		if (vlan->vlan_id == vlan_id) {
 			if (is_write_tbl && vlan->hd_tbl_status)
@@ -10296,6 +10313,8 @@ static void hclge_rm_vport_vlan_table(struct hclge_vport *vport, u16 vlan_id,
 			break;
 		}
 	}
+
+	mutex_unlock(&hdev->vport_lock);
 }
 
 void hclge_rm_vport_all_vlan_table(struct hclge_vport *vport, bool is_del_list)
@@ -10303,6 +10322,8 @@ void hclge_rm_vport_all_vlan_table(struct hclge_vport *vport, bool is_del_list)
 	struct hclge_vport_vlan_cfg *vlan, *tmp;
 	struct hclge_dev *hdev = vport->back;
 
+	mutex_lock(&hdev->vport_lock);
+
 	list_for_each_entry_safe(vlan, tmp, &vport->vlan_list, node) {
 		if (vlan->hd_tbl_status)
 			hclge_set_vlan_filter_hw(hdev,
@@ -10318,6 +10339,7 @@ void hclge_rm_vport_all_vlan_table(struct hclge_vport *vport, bool is_del_list)
 		}
 	}
 	clear_bit(vport->vport_id, hdev->vf_vlan_full);
+	mutex_unlock(&hdev->vport_lock);
 }
 
 void hclge_uninit_vport_vlan_table(struct hclge_dev *hdev)
@@ -10326,6 +10348,8 @@ void hclge_uninit_vport_vlan_table(struct hclge_dev *hdev)
 	struct hclge_vport *vport;
 	int i;
 
+	mutex_lock(&hdev->vport_lock);
+
 	for (i = 0; i < hdev->num_alloc_vport; i++) {
 		vport = &hdev->vport[i];
 		list_for_each_entry_safe(vlan, tmp, &vport->vlan_list, node) {
@@ -10333,6 +10357,8 @@ void hclge_uninit_vport_vlan_table(struct hclge_dev *hdev)
 			kfree(vlan);
 		}
 	}
+
+	mutex_unlock(&hdev->vport_lock);
 }
 
 void hclge_restore_vport_port_base_vlan_config(struct hclge_dev *hdev)
@@ -10372,6 +10398,8 @@ void hclge_restore_vport_vlan_table(struct hclge_vport *vport)
 	struct hclge_dev *hdev = vport->back;
 	int ret;
 
+	mutex_lock(&hdev->vport_lock);
+
 	if (vport->port_base_vlan_cfg.state == HNAE3_PORT_BASE_VLAN_DISABLE) {
 		list_for_each_entry_safe(vlan, tmp, &vport->vlan_list, node) {
 			ret = hclge_set_vlan_filter_hw(hdev, htons(ETH_P_8021Q),
@@ -10382,6 +10410,8 @@ void hclge_restore_vport_vlan_table(struct hclge_vport *vport)
 			vlan->hd_tbl_status = true;
 		}
 	}
+
+	mutex_unlock(&hdev->vport_lock);
 }
 
 /* For global reset and imp reset, hardware will clear the mac table,
@@ -12279,8 +12309,8 @@ static void hclge_uninit_ae_dev(struct hnae3_ae_dev *ae_dev)
 	hclge_misc_irq_uninit(hdev);
 	hclge_devlink_uninit(hdev);
 	hclge_pci_uninit(hdev);
-	mutex_destroy(&hdev->vport_lock);
 	hclge_uninit_vport_vlan_table(hdev);
+	mutex_destroy(&hdev->vport_lock);
 	ae_dev->priv = NULL;
 }
 
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h +1 -0
index 801268b61e96..c841cce2d025 100644
@@ -1087,6 +1087,7 @@ struct hclge_vport {
 	spinlock_t mac_list_lock; /* protect mac address need to add/detele */
 	struct list_head uc_mac_list;   /* Store VF unicast table */
 	struct list_head mc_mac_list;   /* Store VF multicast table */
+
 	struct list_head vlan_list;     /* Store VF vlan table */
 };
 
Signature · 1 line
2.34.1
author maintainer
#740 lore raw

PATCH 732/1017 net: hns3: format the output of the MAC address

In reply to Greg Kroah-Hartman

From: Yufeng Mo <[email protected]>

[ Upstream commit 4f331fda35f1695af8ddd8180edc948880def74b ]

Printing the whole MAC addresse may bring security risks. Therefore, the MAC address is partially encrypted to improve security.

---
 drivers/net/ethernet/hisilicon/hns3/hnae3.h   | 14 +++++
 .../net/ethernet/hisilicon/hns3/hns3_enet.c   | 29 +++++++---
 .../hisilicon/hns3/hns3pf/hclge_main.c        | 55 ++++++++++++-------
 .../hisilicon/hns3/hns3vf/hclgevf_main.c      |  7 ++-
 4 files changed, 74 insertions(+), 31 deletions(-)
drivers/net/ethernet/hisilicon/hns3/hnae3.h +14 -0
index 63f5abcc6bf4..9298fbecb31a 100644
@@ -861,6 +861,20 @@ struct hnae3_handle {
 #define hnae3_get_bit(origin, shift) \
 	hnae3_get_field(origin, 0x1 << (shift), shift)
 
+#define HNAE3_FORMAT_MAC_ADDR_LEN	18
+#define HNAE3_FORMAT_MAC_ADDR_OFFSET_0	0
+#define HNAE3_FORMAT_MAC_ADDR_OFFSET_4	4
+#define HNAE3_FORMAT_MAC_ADDR_OFFSET_5	5
+
+static inline void hnae3_format_mac_addr(char *format_mac_addr,
+					 const u8 *mac_addr)
+{
+	snprintf(format_mac_addr, HNAE3_FORMAT_MAC_ADDR_LEN, "%02x:**:**:**:%02x:%02x",
+		 mac_addr[HNAE3_FORMAT_MAC_ADDR_OFFSET_0],
+		 mac_addr[HNAE3_FORMAT_MAC_ADDR_OFFSET_4],
+		 mac_addr[HNAE3_FORMAT_MAC_ADDR_OFFSET_5]);
+}
+
 int hnae3_register_ae_dev(struct hnae3_ae_dev *ae_dev);
 void hnae3_unregister_ae_dev(struct hnae3_ae_dev *ae_dev);
 
drivers/net/ethernet/hisilicon/hns3/hns3_enet.c +20 -9
index 9ccebbaa0d69..8303863774ff 100644
@@ -2255,6 +2255,8 @@ netdev_tx_t hns3_nic_net_xmit(struct sk_buff *skb, struct net_device *netdev)
 
 static int hns3_nic_net_set_mac_address(struct net_device *netdev, void *p)
 {
+	char format_mac_addr_perm[HNAE3_FORMAT_MAC_ADDR_LEN];
+	char format_mac_addr_sa[HNAE3_FORMAT_MAC_ADDR_LEN];
 	struct hnae3_handle *h = hns3_get_handle(netdev);
 	struct sockaddr *mac_addr = p;
 	int ret;
@@ -2263,8 +2265,9 @@ static int hns3_nic_net_set_mac_address(struct net_device *netdev, void *p)
 		return -EADDRNOTAVAIL;
 
 	if (ether_addr_equal(netdev->dev_addr, mac_addr->sa_data)) {
-		netdev_info(netdev, "already using mac address %pM\n",
-			    mac_addr->sa_data);
+		hnae3_format_mac_addr(format_mac_addr_sa, mac_addr->sa_data);
+		netdev_info(netdev, "already using mac address %s\n",
+			    format_mac_addr_sa);
 		return 0;
 	}
 
@@ -2273,8 +2276,10 @@ static int hns3_nic_net_set_mac_address(struct net_device *netdev, void *p)
 	 */
 	if (!hns3_is_phys_func(h->pdev) &&
 	    !is_zero_ether_addr(netdev->perm_addr)) {
-		netdev_err(netdev, "has permanent MAC %pM, user MAC %pM not allow\n",
-			   netdev->perm_addr, mac_addr->sa_data);
+		hnae3_format_mac_addr(format_mac_addr_perm, netdev->perm_addr);
+		hnae3_format_mac_addr(format_mac_addr_sa, mac_addr->sa_data);
+		netdev_err(netdev, "has permanent MAC %s, user MAC %s not allow\n",
+			   format_mac_addr_perm, format_mac_addr_sa);
 		return -EPERM;
 	}
 
@@ -2836,14 +2841,16 @@ static int hns3_nic_set_vf_rate(struct net_device *ndev, int vf,
 static int hns3_nic_set_vf_mac(struct net_device *netdev, int vf_id, u8 *mac)
 {
 	struct hnae3_handle *h = hns3_get_handle(netdev);
+	char format_mac_addr[HNAE3_FORMAT_MAC_ADDR_LEN];
 
 	if (!h->ae_algo->ops->set_vf_mac)
 		return -EOPNOTSUPP;
 
 	if (is_multicast_ether_addr(mac)) {
+		hnae3_format_mac_addr(format_mac_addr, mac);
 		netdev_err(netdev,
-			   "Invalid MAC:%pM specified. Could not set MAC\n",
-			   mac);
+			   "Invalid MAC:%s specified. Could not set MAC\n",
+			   format_mac_addr);
 		return -EINVAL;
 	}
 
@@ -4934,6 +4941,7 @@ static void hns3_uninit_all_ring(struct hns3_nic_priv *priv)
 static int hns3_init_mac_addr(struct net_device *netdev)
 {
 	struct hns3_nic_priv *priv = netdev_priv(netdev);
+	char format_mac_addr[HNAE3_FORMAT_MAC_ADDR_LEN];
 	struct hnae3_handle *h = priv->ae_handle;
 	u8 mac_addr_temp[ETH_ALEN];
 	int ret = 0;
@@ -4944,8 +4952,9 @@ static int hns3_init_mac_addr(struct net_device *netdev)
 	/* Check if the MAC address is valid, if not get a random one */
 	if (!is_valid_ether_addr(mac_addr_temp)) {
 		eth_hw_addr_random(netdev);
-		dev_warn(priv->dev, "using random MAC address %pM\n",
-			 netdev->dev_addr);
+		hnae3_format_mac_addr(format_mac_addr, netdev->dev_addr);
+		dev_warn(priv->dev, "using random MAC address %s\n",
+			 format_mac_addr);
 	} else if (!ether_addr_equal(netdev->dev_addr, mac_addr_temp)) {
 		eth_hw_addr_set(netdev, mac_addr_temp);
 		ether_addr_copy(netdev->perm_addr, mac_addr_temp);
@@ -4997,8 +5006,10 @@ static void hns3_client_stop(struct hnae3_handle *handle)
 static void hns3_info_show(struct hns3_nic_priv *priv)
 {
 	struct hnae3_knic_private_info *kinfo = &priv->ae_handle->kinfo;
+	char format_mac_addr[HNAE3_FORMAT_MAC_ADDR_LEN];
 
-	dev_info(priv->dev, "MAC address: %pM\n", priv->netdev->dev_addr);
+	hnae3_format_mac_addr(format_mac_addr, priv->netdev->dev_addr);
+	dev_info(priv->dev, "MAC address: %s\n", format_mac_addr);
 	dev_info(priv->dev, "Task queue pairs numbers: %u\n", kinfo->num_tqps);
 	dev_info(priv->dev, "RSS size: %u\n", kinfo->rss_size);
 	dev_info(priv->dev, "Allocated RSS size: %u\n", kinfo->req_rss_size);
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c +35 -20

162 lines not on this page · raw message

drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c +5 -2
index 70491e07b0ff..4d3d14b637c0 100644
@@ -1514,15 +1514,18 @@ static void hclgevf_config_mac_list(struct hclgevf_dev *hdev,
 				    struct list_head *list,
 				    enum HCLGEVF_MAC_ADDR_TYPE mac_type)
 {
+	char format_mac_addr[HNAE3_FORMAT_MAC_ADDR_LEN];
 	struct hclgevf_mac_addr_node *mac_node, *tmp;
 	int ret;
 
 	list_for_each_entry_safe(mac_node, tmp, list, node) {
 		ret = hclgevf_add_del_mac_addr(hdev, mac_node, mac_type);
 		if  (ret) {
+			hnae3_format_mac_addr(format_mac_addr,
+					      mac_node->mac_addr);
 			dev_err(&hdev->pdev->dev,
-				"failed to configure mac %pM, state = %d, ret = %d\n",
-				mac_node->mac_addr, mac_node->state, ret);
+				"failed to configure mac %s, state = %d, ret = %d\n",
+				format_mac_addr, mac_node->state, ret);
 			return;
 		}
 		if (mac_node->state == HCLGEVF_MAC_TO_ADD) {
Signature · 1 line
2.34.1
author maintainer
#741 lore raw

PATCH 733/1017 net: hns3: refine the process when PF set VF VLAN

In reply to Greg Kroah-Hartman

From: Jian Shen <[email protected]>

[ Upstream commit 190cd8a72b0181c543ecada6243be3a50636941b ]

Currently, when PF set VF VLAN, it sends notify mailbox to VF if VF alive. VF stop its traffic, and send request mailbox to PF, then PF updates VF VLAN. It's a bit complex. If VF is killed before sending request, PF will not set VF VLAN without any log.

This patch refines the process, PF can set VF VLAN direclty, and then notify the VF. If VF is resetting at that time, the notify may be dropped, so VF should query it after reset finished.

Fixes 92f11ea177cd ("net: hns3: fix set port based VLAN issue for VF")
---
 .../hisilicon/hns3/hns3pf/hclge_main.c         | 18 +++++++++++++-----
 .../hisilicon/hns3/hns3vf/hclgevf_main.c       |  5 +++++
 2 files changed, 18 insertions(+), 5 deletions(-)
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c +13 -5
index 9866e5c1a71b..cf3c798e5400 100644
@@ -9449,11 +9449,16 @@ static int hclge_set_vf_mac(struct hnae3_handle *handle, int vf,
 
 	ether_addr_copy(vport->vf_info.mac, mac_addr);
 
+	/* there is a timewindow for PF to know VF unalive, it may
+	 * cause send mailbox fail, but it doesn't matter, VF will
+	 * query it when reinit.
+	 */
 	if (test_bit(HCLGE_VPORT_STATE_ALIVE, &vport->state)) {
 		dev_info(&hdev->pdev->dev,
 			 "MAC of VF %d has been set to %s, and it will be reinitialized!\n",
 			 vf, format_mac_addr);
-		return hclge_inform_reset_assert_to_vf(vport);
+		(void)hclge_inform_reset_assert_to_vf(vport);
+		return 0;
 	}
 
 	dev_info(&hdev->pdev->dev, "MAC of VF %d has been set to %s\n",
@@ -10681,14 +10686,17 @@ static int hclge_set_vf_vlan_filter(struct hnae3_handle *handle, int vfid,
 		return ret;
 	}
 
-	/* for DEVICE_VERSION_V3, vf doesn't need to know about the port based
+	/* there is a timewindow for PF to know VF unalive, it may
+	 * cause send mailbox fail, but it doesn't matter, VF will
+	 * query it when reinit.
+	 * for DEVICE_VERSION_V3, vf doesn't need to know about the port based
 	 * VLAN state.
 	 */
 	if (ae_dev->dev_version < HNAE3_DEVICE_VERSION_V3 &&
 	    test_bit(HCLGE_VPORT_STATE_ALIVE, &vport->state))
-		hclge_push_vf_port_base_vlan_info(&hdev->vport[0],
-						  vport->vport_id, state,
-						  &vlan_info);
+		(void)hclge_push_vf_port_base_vlan_info(&hdev->vport[0],
+							vport->vport_id,
+							state, &vlan_info);
 
 	return 0;
 }
drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c +5 -0
index 4d3d14b637c0..7e30bad08356 100644
@@ -3344,6 +3344,11 @@ static int hclgevf_reset_hdev(struct hclgevf_dev *hdev)
 		return ret;
 	}
 
+	/* get current port based vlan state from PF */
+	ret = hclgevf_get_port_base_vlan_filter_state(hdev);
+	if (ret)
+		return ret;
+
 	set_bit(HCLGEVF_STATE_PROMISC_CHANGED, &hdev->state);
 
 	hclgevf_init_rxd_adv_layout(hdev);
Signature · 1 line
2.34.1
author maintainer
#742 lore raw

PATCH 734/1017 net: phy: broadcom: Fix brcm_fet_config_init()

In reply to Greg Kroah-Hartman

From: Florian Fainelli <[email protected]>

[ Upstream commit bf8bfc4336f7a34e48b3bbd19b1542bf085bdc3d ]

A Broadcom AC201 PHY (same entry as 5241) would be flagged by the Broadcom UniMAC MDIO controller as not completing the turn around properly since the PHY expects 65 MDC clock cycles to complete a write cycle, and the MDIO controller was only sending 64 MDC clock cycles as determined by looking at a scope shot.

This would make the subsequent read fail with the UniMAC MDIO controller command field having MDIO_READ_FAIL set and we would abort the brcm_fet_config_init() function and thus not probe the PHY at all.

After issuing a software reset, wait for at least 1ms which is well above the 1us reset delay advertised by the datasheet and issue a dummy read to let the PHY turn around the line properly. This read specifically ignores -EIO which would be returned by MDIO controllers checking for the line being turned around.

If we have a genuine reaad failure, the next read of the interrupt status register would pick it up anyway.

---
 drivers/net/phy/broadcom.c | 21 +++++++++++++++++++++
 1 file changed, 21 insertions(+)
drivers/net/phy/broadcom.c +21 -0
index 3c683e0e40e9..e36809aa6d30 100644
@@ -11,6 +11,7 @@
  */
 
 #include "bcm-phy-lib.h"
+#include <linux/delay.h>
 #include <linux/module.h>
 #include <linux/phy.h>
 #include <linux/brcmphy.h>
@@ -602,6 +603,26 @@ static int brcm_fet_config_init(struct phy_device *phydev)
 	if (err < 0)
 		return err;
 
+	/* The datasheet indicates the PHY needs up to 1us to complete a reset,
+	 * build some slack here.
+	 */
+	usleep_range(1000, 2000);
+
+	/* The PHY requires 65 MDC clock cycles to complete a write operation
+	 * and turnaround the line properly.
+	 *
+	 * We ignore -EIO here as the MDIO controller (e.g.: mdio-bcm-unimac)
+	 * may flag the lack of turn-around as a read failure. This is
+	 * particularly true with this combination since the MDIO controller
+	 * only used 64 MDC cycles. This is not a critical failure in this
+	 * specific case and it has no functional impact otherwise, so we let
+	 * that one go through. If there is a genuine bus error, the next read
+	 * of MII_BRCM_FET_INTREG will error out.
+	 */
+	err = phy_read(phydev, MII_BMCR);
+	if (err < 0 && err != -EIO)
+		return err;
+
 	reg = phy_read(phydev, MII_BRCM_FET_INTREG);
 	if (reg < 0)
 		return reg;
Signature · 1 line
2.34.1
author maintainer
#743 lore raw

PATCH 735/1017 selftests: test_vxlan_under_vrf: Fix broken test case

In reply to Greg Kroah-Hartman

From: Ido Schimmel <[email protected]>

[ Upstream commit b50d3b46f84282d795ae3076111acb75ae1031f3 ]

The purpose of the last test case is to test VXLAN encapsulation and decapsulation when the underlay lookup takes place in a non-default VRF. This is achieved by enslaving the physical device of the tunnel to a VRF.

The binding of the VXLAN UDP socket to the VRF happens when the VXLAN device itself is opened, not when its physical device is opened. This was also mentioned in the cited commit ("tests that moving the underlay from a VRF to another works when down/up the VXLAN interface"), but the test did something else.

Fix it by reopening the VXLAN device instead of its physical device.

Before:

 # ./test_vxlan_under_vrf.sh
 Checking HV connectivity                                           [ OK ]
 Check VM connectivity through VXLAN (underlay in the default VRF)  [ OK ]
 Check VM connectivity through VXLAN (underlay in a VRF)            [FAIL]

After:

 # ./test_vxlan_under_vrf.sh
 Checking HV connectivity                                           [ OK ]
 Check VM connectivity through VXLAN (underlay in the default VRF)  [ OK ]
 Check VM connectivity through VXLAN (underlay in a VRF)            [ OK ]
Fixes 03f1c26b1c56 ("test/net: Add script for VXLAN underlay in a VRF")
---
 tools/testing/selftests/net/test_vxlan_under_vrf.sh | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
tools/testing/selftests/net/test_vxlan_under_vrf.sh +4 -4
index ea5a7a808f12..1fd1250ebc66 100755
@@ -120,11 +120,11 @@ echo "[ OK ]"
 
 # Move the underlay to a non-default VRF
 ip -netns hv-1 link set veth0 vrf vrf-underlay
-ip -netns hv-1 link set veth0 down
-ip -netns hv-1 link set veth0 up
+ip -netns hv-1 link set vxlan0 down
+ip -netns hv-1 link set vxlan0 up
 ip -netns hv-2 link set veth0 vrf vrf-underlay
-ip -netns hv-2 link set veth0 down
-ip -netns hv-2 link set veth0 up
+ip -netns hv-2 link set vxlan0 down
+ip -netns hv-2 link set vxlan0 up
 
 echo -n "Check VM connectivity through VXLAN (underlay in a VRF)            "
 ip netns exec vm-1 ping -c 1 -W 1 10.0.0.2 &> /dev/null || (echo "[FAIL]"; false)
Signature · 1 line
2.34.1
author maintainer
#744 lore raw

PATCH 736/1017 NFS: Dont loop forever in nfs_do_recoalesce()

In reply to Greg Kroah-Hartman

From: Trond Myklebust <[email protected]>

[ Upstream commit d02d81efc7564b4d5446a02e0214a164cf00b1f3 ]

If __nfs_pageio_add_request() fails to add the request, it will return with either desc->pg_error < 0, or mirror->pg_recoalesce will be set, so we are guaranteed either to exit the function altogether, or to loop.

However if there is nothing left in mirror->pg_list to coalesce, we must exit, so make sure that we clear mirror->pg_recoalesce every time we loop.

Fixes 70536bf4eb07 ("NFS: Clean up reset of the mirror accounting variables")
---
 fs/nfs/pagelist.c | 1 +
 1 file changed, 1 insertion(+)
fs/nfs/pagelist.c +1 -0
index ad7f83dc9a2d..815d63080245 100644
@@ -1218,6 +1218,7 @@ static int nfs_do_recoalesce(struct nfs_pageio_descriptor *desc)
 
 	do {
 		list_splice_init(&mirror->pg_list, &head);
+		mirror->pg_recoalesce = 0;
 
 		while (!list_empty(&head)) {
 			struct nfs_page *req;
Signature · 1 line
2.34.1
author maintainer
#745 lore raw

PATCH 737/1017 net: hns3: clean residual vf config after disable sriov

In reply to Greg Kroah-Hartman

From: Peng Li <[email protected]>

[ Upstream commit 671cb8cbb9c9e24b681d21b1bfae991e2386ac73 ]

After disable sriov, VF still has some config and info need to be cleaned, which configured by PF. This patch clean the HW config and SW struct vport->vf_info.

Fixes fa8d82e853e8 ("net: hns3: Add support of .sriov_configure in HNS3 driver")
---
 drivers/net/ethernet/hisilicon/hns3/hnae3.h   |  3 ++
 .../net/ethernet/hisilicon/hns3/hns3_enet.c   | 18 +++++++
 .../hisilicon/hns3/hns3pf/hclge_main.c        | 50 +++++++++++++++++++
 3 files changed, 71 insertions(+)
drivers/net/ethernet/hisilicon/hns3/hnae3.h +3 -0
index 9298fbecb31a..adea528d79ad 100644
@@ -536,6 +536,8 @@ struct hnae3_ae_dev {
  *   Get 1588 rx hwstamp
  * get_ts_info
  *   Get phc info
+ * clean_vf_config
+ *   Clean residual vf info after disable sriov
  */
 struct hnae3_ae_ops {
 	int (*init_ae_dev)(struct hnae3_ae_dev *ae_dev);
@@ -729,6 +731,7 @@ struct hnae3_ae_ops {
 			   struct ethtool_ts_info *info);
 	int (*get_link_diagnosis_info)(struct hnae3_handle *handle,
 				       u32 *status_code);
+	void (*clean_vf_config)(struct hnae3_ae_dev *ae_dev, int num_vfs);
 };
 
 struct hnae3_dcb_ops {
drivers/net/ethernet/hisilicon/hns3/hns3_enet.c +18 -0
index 8303863774ff..ef75223b0b55 100644
@@ -2954,6 +2954,21 @@ static int hns3_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	return ret;
 }
 
+/**
+ * hns3_clean_vf_config
+ * @pdev: pointer to a pci_dev structure
+ * @num_vfs: number of VFs allocated
+ *
+ * Clean residual vf config after disable sriov
+ **/
+static void hns3_clean_vf_config(struct pci_dev *pdev, int num_vfs)
+{
+	struct hnae3_ae_dev *ae_dev = pci_get_drvdata(pdev);
+
+	if (ae_dev->ops->clean_vf_config)
+		ae_dev->ops->clean_vf_config(ae_dev, num_vfs);
+}
+
 /* hns3_remove - Device removal routine
  * @pdev: PCI device information struct
  */
@@ -2992,7 +3007,10 @@ static int hns3_pci_sriov_configure(struct pci_dev *pdev, int num_vfs)
 		else
 			return num_vfs;
 	} else if (!pci_vfs_assigned(pdev)) {
+		int num_vfs_pre = pci_num_vf(pdev);
+
 		pci_disable_sriov(pdev);
+		hns3_clean_vf_config(pdev, num_vfs_pre);
 	} else {
 		dev_warn(&pdev->dev,
 			 "Unable to free VFs because some are assigned to VMs.\n");
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c +50 -0

62 lines not on this page · raw message

Signature · 1 line
2.34.1
author maintainer
#746 lore raw

PATCH 738/1017 net: sparx5: depends on PTP_1588_CLOCK_OPTIONAL

In reply to Greg Kroah-Hartman

From: Randy Dunlap <[email protected]>

[ Upstream commit 08be6b13db23f68146c600dd5adfd92e99d9ec6e ]

Fix build errors when PTP_1588_CLOCK=m and SPARX5_SWTICH=y.

arc-linux-ld: drivers/net/ethernet/microchip/sparx5/sparx5_ethtool.o: in function `sparx5_get_ts_info':
sparx5_ethtool.c:(.text+0x146): undefined reference to `ptp_clock_index'
arc-linux-ld: sparx5_ethtool.c:(.text+0x146): undefined reference to `ptp_clock_index'
arc-linux-ld: drivers/net/ethernet/microchip/sparx5/sparx5_ptp.o: in function `sparx5_ptp_init':
sparx5_ptp.c:(.text+0xd56): undefined reference to `ptp_clock_register'
arc-linux-ld: sparx5_ptp.c:(.text+0xd56): undefined reference to `ptp_clock_register'
arc-linux-ld: drivers/net/ethernet/microchip/sparx5/sparx5_ptp.o: in function `sparx5_ptp_deinit':
sparx5_ptp.c:(.text+0xf30): undefined reference to `ptp_clock_unregister'
arc-linux-ld: sparx5_ptp.c:(.text+0xf30): undefined reference to `ptp_clock_unregister'
arc-linux-ld: sparx5_ptp.c:(.text+0xf38): undefined reference to `ptp_clock_unregister'
arc-linux-ld: sparx5_ptp.c:(.text+0xf46): undefined reference to `ptp_clock_unregister'
arc-linux-ld: drivers/net/ethernet/microchip/sparx5/sparx5_ptp.o:sparx5_ptp.c:(.text+0xf46): more undefined references to `ptp_clock_unregister' follow

Fixes 3cfa11bac9bb ("net: sparx5: add the basic sparx5 driver")
Cc Horatiu Vultur <[email protected]>
Cc "David S. Miller" <[email protected]>
Cc Jakub Kicinski <[email protected]>
Cc Paolo Abeni <[email protected]>
Cc Steen Hegelund <[email protected]>
Cc Bjarni Jonasson <[email protected]>
Cc Lars Povlsen <[email protected]>
---
 drivers/net/ethernet/microchip/sparx5/Kconfig | 1 +
 1 file changed, 1 insertion(+)
drivers/net/ethernet/microchip/sparx5/Kconfig +1 -0
index 7bdbb2d09a14..85b24edb65d5 100644
@@ -4,6 +4,7 @@ config SPARX5_SWITCH
 	depends on HAS_IOMEM
 	depends on OF
 	depends on ARCH_SPARX5 || COMPILE_TEST
+	depends on PTP_1588_CLOCK_OPTIONAL
 	select PHYLINK
 	select PHY_SPARX5_SERDES
 	select RESET_CONTROLLER
Signature · 1 line
2.34.1
author maintainer
#747 lore raw

PATCH 739/1017 qlcnic: dcb: default to returning -EOPNOTSUPP

In reply to Greg Kroah-Hartman

From: Tom Rix <[email protected]>

[ Upstream commit 1521db37f0d42334a88e8ff28198a27d1ed5cd7b ]

Clang static analysis reports this issue
qlcnic_dcb.c:382:10: warning: Assigned value is
garbage or undefined

  mbx_out = *val;
          ^ ~~~~

val is set in the qlcnic_dcb_query_hw_capability() wrapper. If there is no query_hw_capability op in dcp, success is returned without setting the val.

For this and similar wrappers, return -EOPNOTSUPP.

Fixes 14d385b99059 ("qlcnic: dcb: Query adapter DCB capabilities.")
---
 drivers/net/ethernet/qlogic/qlcnic/qlcnic_dcb.h | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)
drivers/net/ethernet/qlogic/qlcnic/qlcnic_dcb.h +5 -5
index 5d79ee4370bc..7519773eaca6 100644
@@ -51,7 +51,7 @@ static inline int qlcnic_dcb_get_hw_capability(struct qlcnic_dcb *dcb)
 	if (dcb && dcb->ops->get_hw_capability)
 		return dcb->ops->get_hw_capability(dcb);
 
-	return 0;
+	return -EOPNOTSUPP;
 }
 
 static inline void qlcnic_dcb_free(struct qlcnic_dcb *dcb)
@@ -65,7 +65,7 @@ static inline int qlcnic_dcb_attach(struct qlcnic_dcb *dcb)
 	if (dcb && dcb->ops->attach)
 		return dcb->ops->attach(dcb);
 
-	return 0;
+	return -EOPNOTSUPP;
 }
 
 static inline int
@@ -74,7 +74,7 @@ qlcnic_dcb_query_hw_capability(struct qlcnic_dcb *dcb, char *buf)
 	if (dcb && dcb->ops->query_hw_capability)
 		return dcb->ops->query_hw_capability(dcb, buf);
 
-	return 0;
+	return -EOPNOTSUPP;
 }
 
 static inline void qlcnic_dcb_get_info(struct qlcnic_dcb *dcb)
@@ -89,7 +89,7 @@ qlcnic_dcb_query_cee_param(struct qlcnic_dcb *dcb, char *buf, u8 type)
 	if (dcb && dcb->ops->query_cee_param)
 		return dcb->ops->query_cee_param(dcb, buf, type);
 
-	return 0;
+	return -EOPNOTSUPP;
 }
 
 static inline int qlcnic_dcb_get_cee_cfg(struct qlcnic_dcb *dcb)
@@ -97,7 +97,7 @@ static inline int qlcnic_dcb_get_cee_cfg(struct qlcnic_dcb *dcb)
 	if (dcb && dcb->ops->get_cee_cfg)
 		return dcb->ops->get_cee_cfg(dcb);
 
-	return 0;
+	return -EOPNOTSUPP;
 }
 
 static inline void qlcnic_dcb_aen_handler(struct qlcnic_dcb *dcb, void *msg)
Signature · 1 line
2.34.1
author maintainer
#748 lore raw

PATCH 740/1017 net/x25: Fix null-ptr-deref caused by x25_disconnect

In reply to Greg Kroah-Hartman

From: Duoming Zhou <[email protected]>

[ Upstream commit 7781607938c8371d4c2b243527430241c62e39c2 ]

When the link layer is terminating, x25->neighbour will be set to NULL in x25_disconnect(). As a result, it could cause null-ptr-deref bugs in x25_sendmsg(),x25_recvmsg() and x25_connect(). One of the bugs is shown below.

    (Thread 1)                 |  (Thread 2)
x25_link_terminated()          | x25_recvmsg()
 x25_kill_by_neigh()           |  ...
  x25_disconnect()             |  lock_sock(sk)
   ...                         |  ...
   x25->neighbour = NULL //(1) |
   ...                         |  x25->neighbour->extended //(2)

The code sets NULL to x25->neighbour in position (1) and dereferences x25->neighbour in position (2), which could cause null-ptr-deref bug.

This patch adds lock_sock() in x25_kill_by_neigh() in order to synchronize with x25_sendmsg(), x25_recvmsg() and x25_connect(). What`s more, the sock held by lock_sock() is not NULL, because it is extracted from x25_list and uses x25_list_lock to synchronize.

Fixes 4becb7ee5b3d ("net/x25: Fix x25_neigh refcnt leak when x25 disconnect")
---
 net/x25/af_x25.c | 11 ++++++++---
 1 file changed, 8 insertions(+), 3 deletions(-)
net/x25/af_x25.c +8 -3
index 3583354a7d7f..3a171828638b 100644
@@ -1765,10 +1765,15 @@ void x25_kill_by_neigh(struct x25_neigh *nb)
 
 	write_lock_bh(&x25_list_lock);
 
-	sk_for_each(s, &x25_list)
-		if (x25_sk(s)->neighbour == nb)
+	sk_for_each(s, &x25_list) {
+		if (x25_sk(s)->neighbour == nb) {
+			write_unlock_bh(&x25_list_lock);
+			lock_sock(s);
 			x25_disconnect(s, ENETUNREACH, 0, 0);
-
+			release_sock(s);
+			write_lock_bh(&x25_list_lock);
+		}
+	}
 	write_unlock_bh(&x25_list_lock);
 
 	/* Remove any related forwards */
Signature · 1 line
2.34.1
author maintainer
#749 lore raw

PATCH 741/1017 net: sparx5: switchdev: fix possible NULL pointer dereference

In reply to Greg Kroah-Hartman

From: Zheng Yongjun <[email protected]>

[ Upstream commit 0906f3a3df07835e37077d8971aac65347f2ed57 ]

As the possible failure of the allocation, devm_kzalloc() may return NULL pointer.
Therefore, it should be better to check the 'db' in order to prevent the dereference of NULL pointer.

Fixes 10615907e9b51 ("net: sparx5: switchdev: adding frame DMA functionality")
---
 drivers/net/ethernet/microchip/sparx5/sparx5_fdma.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/net/ethernet/microchip/sparx5/sparx5_fdma.c +2 -0
index 7436f62fa152..174ad95e746a 100644
@@ -420,6 +420,8 @@ static int sparx5_fdma_tx_alloc(struct sparx5 *sparx5)
 			db_hw->dataptr = phys;
 			db_hw->status = 0;
 			db = devm_kzalloc(sparx5->dev, sizeof(*db), GFP_KERNEL);
+			if (!db)
+				return -ENOMEM;
 			db->cpu_addr = cpu_addr;
 			list_add_tail(&db->list, &tx->db_list);
 		}
Signature · 1 line
2.34.1
author maintainer
#750 lore raw

PATCH 742/1017 octeontx2-af: initialize action variable

In reply to Greg Kroah-Hartman

From: Tom Rix <[email protected]>

[ Upstream commit 33b5bc9e703383e396f275d51fc4bafa48dbae5a ]

Clang static analysis reports this representative issue rvu_npc.c:898:15: warning: Assigned value is garbage
or undefined

  req.match_id = action.match_id;
               ^ ~~~~~~~~~~~~~~~

The initial setting of action is conditional on
if (is_mcam_entry_enabled(...))
The later check of action.op will sometimes be garbage. So initialize action.

Reduce setting of
*(u64 *)&action = 0x00;
to
*(u64 *)&action = 0;

Fixes 967db3529eca ("octeontx2-af: add support for multicast/promisc packet replication feature")
---
 .../net/ethernet/marvell/octeontx2/af/rvu_npc.c   | 15 +++++++--------
 1 file changed, 7 insertions(+), 8 deletions(-)
drivers/net/ethernet/marvell/octeontx2/af/rvu_npc.c +7 -8
index 91f86d77cd41..3a31fb8cc155 100644
@@ -605,7 +605,7 @@ void rvu_npc_install_ucast_entry(struct rvu *rvu, u16 pcifunc,
 	struct npc_install_flow_req req = { 0 };
 	struct npc_install_flow_rsp rsp = { 0 };
 	struct npc_mcam *mcam = &rvu->hw->mcam;
-	struct nix_rx_action action;
+	struct nix_rx_action action = { 0 };
 	int blkaddr, index;
 
 	/* AF's and SDP VFs work in promiscuous mode */
@@ -626,7 +626,6 @@ void rvu_npc_install_ucast_entry(struct rvu *rvu, u16 pcifunc,
 		*(u64 *)&action = npc_get_mcam_action(rvu, mcam,
 						      blkaddr, index);
 	} else {
-		*(u64 *)&action = 0x00;
 		action.op = NIX_RX_ACTIONOP_UCAST;
 		action.pf_func = pcifunc;
 	}
@@ -657,7 +656,7 @@ void rvu_npc_install_promisc_entry(struct rvu *rvu, u16 pcifunc,
 	struct npc_mcam *mcam = &rvu->hw->mcam;
 	struct rvu_hwinfo *hw = rvu->hw;
 	int blkaddr, ucast_idx, index;
-	struct nix_rx_action action;
+	struct nix_rx_action action = { 0 };
 	u64 relaxed_mask;
 
 	if (!hw->cap.nix_rx_multicast && is_cgx_vf(rvu, pcifunc))
@@ -685,14 +684,14 @@ void rvu_npc_install_promisc_entry(struct rvu *rvu, u16 pcifunc,
 						      blkaddr, ucast_idx);
 
 	if (action.op != NIX_RX_ACTIONOP_RSS) {
-		*(u64 *)&action = 0x00;
+		*(u64 *)&action = 0;
 		action.op = NIX_RX_ACTIONOP_UCAST;
 	}
 
 	/* RX_ACTION set to MCAST for CGX PF's */
 	if (hw->cap.nix_rx_multicast && pfvf->use_mce_list &&
 	    is_pf_cgxmapped(rvu, rvu_get_pf(pcifunc))) {
-		*(u64 *)&action = 0x00;
+		*(u64 *)&action = 0;
 		action.op = NIX_RX_ACTIONOP_MCAST;
 		pfvf = rvu_get_pfvf(rvu, pcifunc & ~RVU_PFVF_FUNC_MASK);
 		action.index = pfvf->promisc_mce_idx;
@@ -832,7 +831,7 @@ void rvu_npc_install_allmulti_entry(struct rvu *rvu, u16 pcifunc, int nixlf,
 	struct rvu_hwinfo *hw = rvu->hw;
 	int blkaddr, ucast_idx, index;
 	u8 mac_addr[ETH_ALEN] = { 0 };
-	struct nix_rx_action action;
+	struct nix_rx_action action = { 0 };
 	struct rvu_pfvf *pfvf;
 	u16 vf_func;
 
@@ -861,14 +860,14 @@ void rvu_npc_install_allmulti_entry(struct rvu *rvu, u16 pcifunc, int nixlf,
 							blkaddr, ucast_idx);
 
 	if (action.op != NIX_RX_ACTIONOP_RSS) {
-		*(u64 *)&action = 0x00;
+		*(u64 *)&action = 0;
 		action.op = NIX_RX_ACTIONOP_UCAST;
 		action.pf_func = pcifunc;
 	}
 
 	/* RX_ACTION set to MCAST for CGX PF's */
 	if (hw->cap.nix_rx_multicast && pfvf->use_mce_list) {
-		*(u64 *)&action = 0x00;
+		*(u64 *)&action = 0;
 		action.op = NIX_RX_ACTIONOP_MCAST;
 		action.index = pfvf->mcast_mce_idx;
 	}
Signature · 1 line
2.34.1
author maintainer
#751 lore raw

PATCH 743/1017 selftests: tls: skip cmsg_to_pipe tests with TLS=n

In reply to Greg Kroah-Hartman

From: Jakub Kicinski <[email protected]>

[ Upstream commit 5c7e49be96ea24776a5b5a07c732c477294add00 ]

These are negative tests, testing TLS code rejects certain operations. They won't pass without TLS enabled, pure TCP accepts those operations.

Fixes d87d67fd61ef ("selftests: tls: test splicing cmsgs")
---
 tools/testing/selftests/net/tls.c | 6 ++++++
 1 file changed, 6 insertions(+)
tools/testing/selftests/net/tls.c +6 -0
index 6e468e0f42f7..5d70b04c482c 100644
@@ -683,6 +683,9 @@ TEST_F(tls, splice_cmsg_to_pipe)
 	char buf[10];
 	int p[2];
 
+	if (self->notls)
+		SKIP(return, "no TLS support");
+
 	ASSERT_GE(pipe(p), 0);
 	EXPECT_EQ(tls_send_cmsg(self->fd, 100, test_str, send_len, 0), 10);
 	EXPECT_EQ(splice(self->cfd, NULL, p[1], NULL, send_len, 0), -1);
@@ -703,6 +706,9 @@ TEST_F(tls, splice_dec_cmsg_to_pipe)
 	char buf[10];
 	int p[2];
 
+	if (self->notls)
+		SKIP(return, "no TLS support");
+
 	ASSERT_GE(pipe(p), 0);
 	EXPECT_EQ(tls_send_cmsg(self->fd, 100, test_str, send_len, 0), 10);
 	EXPECT_EQ(recv(self->cfd, buf, send_len, 0), -1);
Signature · 1 line
2.34.1
author maintainer
#752 lore raw

PATCH 744/1017 net: prefer nf_ct_put instead of nf_conntrack_put

In reply to Greg Kroah-Hartman

From: Florian Westphal <[email protected]>

[ Upstream commit 408bdcfce8dfd6902f75fbcd3b99d8b24b506597 ]

Its the same as nf_conntrack_put(), but without the need for an indirect call. The downside is a module dependency on nf_conntrack, but all of these already depend on conntrack anyway.

---
 net/netfilter/nf_conntrack_core.c |  4 ++--
 net/openvswitch/conntrack.c       | 14 ++++++++++----
 net/sched/act_ct.c                |  6 +++---
 3 files changed, 15 insertions(+), 9 deletions(-)
net/netfilter/nf_conntrack_core.c +2 -2
index 7f7997460764..917e708a4561 100644
@@ -989,7 +989,7 @@ static int __nf_ct_resolve_clash(struct sk_buff *skb,
 
 		nf_ct_acct_merge(ct, ctinfo, loser_ct);
 		nf_ct_add_to_dying_list(loser_ct);
-		nf_conntrack_put(&loser_ct->ct_general);
+		nf_ct_put(loser_ct);
 		nf_ct_set(skb, ct, ctinfo);
 
 		NF_CT_STAT_INC(net, clash_resolve);
@@ -1920,7 +1920,7 @@ nf_conntrack_in(struct sk_buff *skb, const struct nf_hook_state *state)
 		/* Invalid: inverse of the return code tells
 		 * the netfilter core what to do */
 		pr_debug("nf_conntrack_in: Can't track with proto module\n");
-		nf_conntrack_put(&ct->ct_general);
+		nf_ct_put(ct);
 		skb->_nfct = 0;
 		/* Special case: TCP tracker reports an attempt to reopen a
 		 * closed/aborted connection. We have to go back and create a
net/openvswitch/conntrack.c +10 -4
index 8f47f4e78d32..f2b64cab9af7 100644
@@ -574,7 +574,7 @@ ovs_ct_expect_find(struct net *net, const struct nf_conntrack_zone *zone,
 			struct nf_conn *ct = nf_ct_tuplehash_to_ctrack(h);
 
 			nf_ct_delete(ct, 0, 0);
-			nf_conntrack_put(&ct->ct_general);
+			nf_ct_put(ct);
 		}
 	}
 
@@ -723,7 +723,7 @@ static bool skb_nfct_cached(struct net *net,
 		if (nf_ct_is_confirmed(ct))
 			nf_ct_delete(ct, 0, 0);
 
-		nf_conntrack_put(&ct->ct_general);
+		nf_ct_put(ct);
 		nf_ct_set(skb, NULL, 0);
 		return false;
 	}
@@ -967,7 +967,8 @@ static int __ovs_ct_lookup(struct net *net, struct sw_flow_key *key,
 
 		/* Associate skb with specified zone. */
 		if (tmpl) {
-			nf_conntrack_put(skb_nfct(skb));
+			ct = nf_ct_get(skb, &ctinfo);
+			nf_ct_put(ct);
 			nf_conntrack_get(&tmpl->ct_general);
 			nf_ct_set(skb, tmpl, IP_CT_NEW);
 		}
@@ -1328,7 +1329,12 @@ int ovs_ct_execute(struct net *net, struct sk_buff *skb,
 
 int ovs_ct_clear(struct sk_buff *skb, struct sw_flow_key *key)
 {
-	nf_conntrack_put(skb_nfct(skb));
+	enum ip_conntrack_info ctinfo;
+	struct nf_conn *ct;
+
+	ct = nf_ct_get(skb, &ctinfo);
+
+	nf_ct_put(ct);
 	nf_ct_set(skb, NULL, IP_CT_UNTRACKED);
 	ovs_ct_fill_key(skb, key, false);
 
net/sched/act_ct.c +3 -3
index 4ffea1290ce1..240b3c5d2eb1 100644
@@ -592,7 +592,7 @@ static bool tcf_ct_skb_nfct_cached(struct net *net, struct sk_buff *skb,
 		if (nf_ct_is_confirmed(ct))
 			nf_ct_kill(ct);
 
-		nf_conntrack_put(&ct->ct_general);
+		nf_ct_put(ct);
 		nf_ct_set(skb, NULL, IP_CT_UNTRACKED);
 
 		return false;
@@ -757,7 +757,7 @@ static void tcf_ct_params_free(struct rcu_head *head)
 	tcf_ct_flow_table_put(params);
 
 	if (params->tmpl)
-		nf_conntrack_put(&params->tmpl->ct_general);
+		nf_ct_put(params->tmpl);
 	kfree(params);
 }
 
@@ -967,7 +967,7 @@ static int tcf_ct_act(struct sk_buff *skb, const struct tc_action *a,
 		tc_skb_cb(skb)->post_ct = false;
 		ct = nf_ct_get(skb, &ctinfo);
 		if (ct) {
-			nf_conntrack_put(&ct->ct_general);
+			nf_ct_put(ct);
 			nf_ct_set(skb, NULL, IP_CT_UNTRACKED);
 		}
 
Signature · 1 line
2.34.1
author maintainer
#753 lore raw

PATCH 745/1017 net/sched: act_ct: fix ref leak when switching zones

In reply to Greg Kroah-Hartman

From: Marcelo Ricardo Leitner <[email protected]>

[ Upstream commit bcb74e132a76ce0502bb33d5b65533a4ed72d159 ]

When switching zones or network namespaces without doing a ct clear in between, it is now leaking a reference to the old ct entry. That's because tcf_ct_skb_nfct_cached() returns false and tcf_ct_flow_table_lookup() may simply overwrite it.

The fix is to, as the ct entry is not reusable, free it already at tcf_ct_skb_nfct_cached().

Fixes 2f131de361f6 ("net/sched: act_ct: Fix flow table lookup after ct clear or switching zones")
---
 net/sched/act_ct.c | 15 +++++++++------
 1 file changed, 9 insertions(+), 6 deletions(-)
net/sched/act_ct.c +9 -6
index 240b3c5d2eb1..553bf41671a6 100644
@@ -583,22 +583,25 @@ static bool tcf_ct_skb_nfct_cached(struct net *net, struct sk_buff *skb,
 	if (!ct)
 		return false;
 	if (!net_eq(net, read_pnet(&ct->ct_net)))
-		return false;
+		goto drop_ct;
 	if (nf_ct_zone(ct)->id != zone_id)
-		return false;
+		goto drop_ct;
 
 	/* Force conntrack entry direction. */
 	if (force && CTINFO2DIR(ctinfo) != IP_CT_DIR_ORIGINAL) {
 		if (nf_ct_is_confirmed(ct))
 			nf_ct_kill(ct);
 
-		nf_ct_put(ct);
-		nf_ct_set(skb, NULL, IP_CT_UNTRACKED);
-
-		return false;
+		goto drop_ct;
 	}
 
 	return true;
+
+drop_ct:
+	nf_ct_put(ct);
+	nf_ct_set(skb, NULL, IP_CT_UNTRACKED);
+
+	return false;
 }
 
 /* Trim the skb to the length specified by the IP/IPv6 header,
Signature · 1 line
2.34.1
author maintainer
#754 lore raw

PATCH 746/1017 NFSv4/pNFS: Fix another issue with a list iterator pointing to the head

In reply to Greg Kroah-Hartman

From: Trond Myklebust <[email protected]>

[ Upstream commit 7c9d845f0612e5bcd23456a2ec43be8ac43458f1 ]

In nfs4_callback_devicenotify(), if we don't find a matching entry for the deviceid, we're left with a pointer to 'struct nfs_server' that actually points to the list of super blocks associated with our struct nfs_client.
Furthermore, even if we have a valid pointer, nothing pins the super block, and so the struct nfs_server could end up getting freed while we're using it.

Since all we want is a pointer to the struct pnfs_layoutdriver_type, let's skip all the iteration over super blocks, and just use APIs to find the layout driver directly.

Fixes 1be5683b03a7 ("pnfs: CB_NOTIFY_DEVICEID")
---
 fs/nfs/callback_proc.c | 27 +++++++++------------------
 fs/nfs/pnfs.c          | 11 +++++++++++
 fs/nfs/pnfs.h          |  2 ++
 3 files changed, 22 insertions(+), 18 deletions(-)
fs/nfs/callback_proc.c +9 -18
index c343666d9a42..6464dde03705 100644
@@ -358,12 +358,11 @@ __be32 nfs4_callback_devicenotify(void *argp, void *resp,
 				  struct cb_process_state *cps)
 {
 	struct cb_devicenotifyargs *args = argp;
+	const struct pnfs_layoutdriver_type *ld = NULL;
 	uint32_t i;
 	__be32 res = 0;
-	struct nfs_client *clp = cps->clp;
-	struct nfs_server *server = NULL;
 
-	if (!clp) {
+	if (!cps->clp) {
 		res = cpu_to_be32(NFS4ERR_OP_NOT_IN_SESSION);
 		goto out;
 	}
@@ -371,23 +370,15 @@ __be32 nfs4_callback_devicenotify(void *argp, void *resp,
 	for (i = 0; i < args->ndevs; i++) {
 		struct cb_devicenotifyitem *dev = &args->devs[i];
 
-		if (!server ||
-		    server->pnfs_curr_ld->id != dev->cbd_layout_type) {
-			rcu_read_lock();
-			list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link)
-				if (server->pnfs_curr_ld &&
-				    server->pnfs_curr_ld->id == dev->cbd_layout_type) {
-					rcu_read_unlock();
-					goto found;
-				}
-			rcu_read_unlock();
-			continue;
+		if (!ld || ld->id != dev->cbd_layout_type) {
+			pnfs_put_layoutdriver(ld);
+			ld = pnfs_find_layoutdriver(dev->cbd_layout_type);
+			if (!ld)
+				continue;
 		}
-
-	found:
-		nfs4_delete_deviceid(server->pnfs_curr_ld, clp, &dev->cbd_dev_id);
+		nfs4_delete_deviceid(ld, cps->clp, &dev->cbd_dev_id);
 	}
-
+	pnfs_put_layoutdriver(ld);
 out:
 	kfree(args->devs);
 	return res;
fs/nfs/pnfs.c +11 -0
index 7c9090a28e5c..7ddd003ab8b1 100644
@@ -92,6 +92,17 @@ find_pnfs_driver(u32 id)
 	return local;
 }
 
+const struct pnfs_layoutdriver_type *pnfs_find_layoutdriver(u32 id)
+{
+	return find_pnfs_driver(id);
+}
+
+void pnfs_put_layoutdriver(const struct pnfs_layoutdriver_type *ld)
+{
+	if (ld)
+		module_put(ld->owner);
+}
+
 void
 unset_pnfs_layoutdriver(struct nfs_server *nfss)
 {
fs/nfs/pnfs.h +2 -0
index f4d7548d67b2..07f11489e4e9 100644
@@ -234,6 +234,8 @@ struct pnfs_devicelist {
 
 extern int pnfs_register_layoutdriver(struct pnfs_layoutdriver_type *);
 extern void pnfs_unregister_layoutdriver(struct pnfs_layoutdriver_type *);
+extern const struct pnfs_layoutdriver_type *pnfs_find_layoutdriver(u32 id);
+extern void pnfs_put_layoutdriver(const struct pnfs_layoutdriver_type *ld);
 
 /* nfs4proc.c */
 extern size_t max_response_pages(struct nfs_server *server);
Signature · 1 line
2.34.1
author maintainer
#755 lore raw

PATCH 747/1017 net: dsa: bcm_sf2_cfp: fix an incorrect NULL check on list iterator

In reply to Greg Kroah-Hartman

From: Xiaomeng Tong <[email protected]>

[ Upstream commit 6da69b1da130e7d96766042750cd9f902e890eba ]

The bug is here:

	return rule;

The list iterator value 'rule' will *always* be set and non-NULL by list_for_each_entry(), so it is incorrect to assume that the iterator value will be NULL if the list is empty or no element is found.

To fix the bug, return 'rule' when found, otherwise return NULL.

---
 drivers/net/dsa/bcm_sf2_cfp.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/net/dsa/bcm_sf2_cfp.c +3 -3
index a7e2fcf2df2c..edbe5e7f1cb6 100644
@@ -567,14 +567,14 @@ static void bcm_sf2_cfp_slice_ipv6(struct bcm_sf2_priv *priv,
 static struct cfp_rule *bcm_sf2_cfp_rule_find(struct bcm_sf2_priv *priv,
 					      int port, u32 location)
 {
-	struct cfp_rule *rule = NULL;
+	struct cfp_rule *rule;
 
 	list_for_each_entry(rule, &priv->cfp.rules_list, next) {
 		if (rule->port == port && rule->fs.location == location)
-			break;
+			return rule;
 	}
 
-	return rule;
+	return NULL;
 }
 
 static int bcm_sf2_cfp_rule_cmp(struct bcm_sf2_priv *priv, int port,
Signature · 1 line
2.34.1
author maintainer
#756 lore raw

PATCH 748/1017 fs: fd tables have to be multiples of BITS_PER_LONG

In reply to Greg Kroah-Hartman

From: Linus Torvalds <[email protected]>

[ Upstream commit 1c24a186398f59c80adb9a967486b65c1423a59d ]

This has always been the rule: fdtables have several bitmaps in them, and as a result they have to be sized properly for bitmaps. We walk those bitmaps in chunks of 'unsigned long' in serveral cases, but even when we don't, we use the regular kernel bitops that are defined to work on arrays of 'unsigned long', not on some byte array.

Now, the distinction between arrays of bytes and 'unsigned long' normally only really ends up being noticeable on big-endian systems, but Fedor Pchelkin and Alexey Khoroshilov reported that copy_fd_bitmaps() could be called with an argument that wasn't even a multiple of BITS_PER_BYTE. And then it fails to do the proper copy even on little-endian machines.

The bug wasn't in copy_fd_bitmap(), but in sane_fdtable_size(), which didn't actually sanitize the fdtable size sufficiently, and never made sure it had the proper BITS_PER_LONG alignment.

That's partly because the alignment historically came not from having to explicitly align things, but simply from previous fdtable sizes, and from count_open_files(), which counts the file descriptors by walking them one 'unsigned long' word at a time and thus naturally ends up doing sizing in the proper 'chunks of unsigned long'.

But with the introduction of close_range(), we now have an external source of "this is how many files we want to have", and so sane_fdtable_size() needs to do a better job.

This also adds that explicit alignment to alloc_fdtable(), although there it is mainly just for documentation at a source code level. The arithmetic we do there to pick a reasonable fdtable size already aligns the result sufficiently.

In fact,clang notices that the added ALIGN() in that function doesn't actually do anything, and does not generate any extra code for it.

It turns out that gcc ends up confusing itself by combining a previous constant-sized shift operation with the variable-sized shift operations in roundup_pow_of_two(). And probably due to that doesn't notice that the ALIGN() is a no-op. But that's a (tiny) gcc misfeature that doesn't matter. Having the explicit alignment makes sense, and would actually matter on a 128-bit architecture if we ever go there.

This also adds big comments above both functions about how fdtable sizes have to have that BITS_PER_LONG alignment.

---
 fs/file.c | 30 ++++++++++++++++++++++++++++++
 1 file changed, 30 insertions(+)
fs/file.c +30 -0
index 97d212a9b814..c01c29417ae6 100644
@@ -87,6 +87,21 @@ static void copy_fdtable(struct fdtable *nfdt, struct fdtable *ofdt)
 	copy_fd_bitmaps(nfdt, ofdt, ofdt->max_fds);
 }
 
+/*
+ * Note how the fdtable bitmap allocations very much have to be a multiple of
+ * BITS_PER_LONG. This is not only because we walk those things in chunks of
+ * 'unsigned long' in some places, but simply because that is how the Linux
+ * kernel bitmaps are defined to work: they are not "bits in an array of bytes",
+ * they are very much "bits in an array of unsigned long".
+ *
+ * The ALIGN(nr, BITS_PER_LONG) here is for clarity: since we just multiplied
+ * by that "1024/sizeof(ptr)" before, we already know there are sufficient
+ * clear low bits. Clang seems to realize that, gcc ends up being confused.
+ *
+ * On a 128-bit machine, the ALIGN() would actually matter. In the meantime,
+ * let's consider it documentation (and maybe a test-case for gcc to improve
+ * its code generation ;)
+ */
 static struct fdtable * alloc_fdtable(unsigned int nr)
 {
 	struct fdtable *fdt;
@@ -102,6 +117,7 @@ static struct fdtable * alloc_fdtable(unsigned int nr)
 	nr /= (1024 / sizeof(struct file *));
 	nr = roundup_pow_of_two(nr + 1);
 	nr *= (1024 / sizeof(struct file *));
+	nr = ALIGN(nr, BITS_PER_LONG);
 	/*
 	 * Note that this can drive nr *below* what we had passed if sysctl_nr_open
 	 * had been set lower between the check in expand_files() and here.  Deal
@@ -269,11 +285,25 @@ static unsigned int count_open_files(struct fdtable *fdt)
 	return i;
 }
 
+/*
+ * Note that a sane fdtable size always has to be a multiple of
+ * BITS_PER_LONG, since we have bitmaps that are sized by this.
+ *
+ * 'max_fds' will normally already be properly aligned, but it
+ * turns out that in the close_range() -> __close_range() ->
+ * unshare_fd() -> dup_fd() -> sane_fdtable_size() we can end
+ * up having a 'max_fds' value that isn't already aligned.
+ *
+ * Rather than make close_range() have to worry about this,
+ * just make that BITS_PER_LONG alignment be part of a sane
+ * fdtable size. Becuase that's really what it is.
+ */
 static unsigned int sane_fdtable_size(struct fdtable *fdt, unsigned int max_fds)
 {
 	unsigned int count;
 
 	count = count_open_files(fdt);
+	max_fds = ALIGN(max_fds, BITS_PER_LONG);
 	if (max_fds < NR_OPEN_DEFAULT)
 		max_fds = NR_OPEN_DEFAULT;
 	return min(count, max_fds);
Signature · 1 line
2.34.1
author maintainer
#757 lore raw

PATCH 749/1017 lib/test: use after free in register_test_dev_kmod()

In reply to Greg Kroah-Hartman

From: Dan Carpenter <[email protected]>

[ Upstream commit dc0ce6cc4b133f5f2beb8b47dacae13a7d283c2c ]

The "test_dev" pointer is freed but then returned to the caller.

Fixes d9c6a72d6fa2 ("kmod: add test driver to stress test the module loader")
---
 lib/test_kmod.c | 1 +
 1 file changed, 1 insertion(+)
lib/test_kmod.c +1 -0
index ce1589391413..cb800b1d0d99 100644
@@ -1149,6 +1149,7 @@ static struct kmod_test_device *register_test_dev_kmod(void)
 	if (ret) {
 		pr_err("could not register misc device: %d\n", ret);
 		free_test_dev_kmod(test_dev);
+		test_dev = NULL;
 		goto out;
 	}
 
Signature · 1 line
2.34.1
author maintainer
#758 lore raw

PATCH 750/1017 fs: fix fd table size alignment properly

In reply to Greg Kroah-Hartman

From: Linus Torvalds <[email protected]>

[ Upstream commit d888c83fcec75194a8a48ccd283953bdba7b2550 ]

Jason Donenfeld reports that my commit 1c24a186398f ("fs: fd tables have to be multiples of BITS_PER_LONG") doesn't work, and the reason is an embarrassing brown-paper-bag bug.

Yes, we want to align the number of fds to BITS_PER_LONG, and yes, the reason they might not be aligned is because the incoming 'max_fd' argument might not be aligned.

But aligining the argument - while simple - will cause a "infinitely big" maxfd (eg NR_OPEN_MAX) to just overflow to zero. Which most definitely isn't what we want either.

The obvious fix was always just to do the alignment last, but I had moved it earlier just to make the patch smaller and the code look simpler. Duh. It certainly made _me_ look simple.

Fixes 1c24a186398f ("fs: fd tables have to be multiples of BITS_PER_LONG")
Cc Fedor Pchelkin <[email protected]>
Cc Alexey Khoroshilov <[email protected]>
Cc Christian Brauner <[email protected]>
---
 fs/file.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
fs/file.c +1 -2
index c01c29417ae6..ee9317346702 100644
@@ -303,10 +303,9 @@ static unsigned int sane_fdtable_size(struct fdtable *fdt, unsigned int max_fds)
 	unsigned int count;
 
 	count = count_open_files(fdt);
-	max_fds = ALIGN(max_fds, BITS_PER_LONG);
 	if (max_fds < NR_OPEN_DEFAULT)
 		max_fds = NR_OPEN_DEFAULT;
-	return min(count, max_fds);
+	return ALIGN(min(count, max_fds), BITS_PER_LONG);
 }
 
 /*
Signature · 1 line
2.34.1
author maintainer
#759 lore raw

PATCH 751/1017 LSM: general protection fault in legacy_parse_param

In reply to Greg Kroah-Hartman

From: Casey Schaufler <[email protected]>

[ Upstream commit ecff30575b5ad0eda149aadad247b7f75411fd47 ]

The usual LSM hook "bail on fail" scheme doesn't work for cases where a security module may return an error code indicating that it does not recognize an input. In this particular case Smack sees a mount option that it recognizes, and returns 0. A call to a BPF hook follows, which returns -ENOPARAM, which confuses the caller because Smack has processed its data.

The SELinux hook incorrectly returns 1 on success. There was a time when this was correct, however the current expectation is that it return 0 on success. This is repaired.

---
 security/security.c      | 17 +++++++++++++++--
 security/selinux/hooks.c |  5 ++---
 2 files changed, 17 insertions(+), 5 deletions(-)
security/security.c +15 -2
index cf6615b5e820..745a8bf66d91 100644
@@ -884,9 +884,22 @@ int security_fs_context_dup(struct fs_context *fc, struct fs_context *src_fc)
 	return call_int_hook(fs_context_dup, 0, fc, src_fc);
 }
 
-int security_fs_context_parse_param(struct fs_context *fc, struct fs_parameter *param)
+int security_fs_context_parse_param(struct fs_context *fc,
+				    struct fs_parameter *param)
 {
-	return call_int_hook(fs_context_parse_param, -ENOPARAM, fc, param);
+	struct security_hook_list *hp;
+	int trc;
+	int rc = -ENOPARAM;
+
+	hlist_for_each_entry(hp, &security_hook_heads.fs_context_parse_param,
+			     list) {
+		trc = hp->hook.fs_context_parse_param(fc, param);
+		if (trc == 0)
+			rc = 0;
+		else if (trc != -ENOPARAM)
+			return trc;
+	}
+	return rc;
 }
 
 int security_sb_alloc(struct super_block *sb)
security/selinux/hooks.c +2 -3
index 93eac9464c5f..69b0709bd156 100644
@@ -2916,10 +2916,9 @@ static int selinux_fs_context_parse_param(struct fs_context *fc,
 		return opt;
 
 	rc = selinux_add_opt(opt, param->string, &fc->security);
-	if (!rc) {
+	if (!rc)
 		param->string = NULL;
-		rc = 1;
-	}
+
 	return rc;
 }
 
Signature · 1 line
2.34.1
author maintainer
#760 lore raw

PATCH 752/1017 regulator: rpi-panel: Handle I2C errors/timing to the Atmel

In reply to Greg Kroah-Hartman

From: Dave Stevenson <[email protected]>

[ Upstream commit 5665eee7a3800430e7dc3ef6f25722476b603186 ]

The Atmel is doing some things in the I2C ISR, during which period it will not respond to further commands. This is particularly true of the POWERON command.

Increase delays appropriately, and retry should I2C errors be reported.

---
 .../regulator/rpi-panel-attiny-regulator.c    | 56 +++++++++++++++----
 1 file changed, 46 insertions(+), 10 deletions(-)
drivers/regulator/rpi-panel-attiny-regulator.c +46 -10
index ee46bfbf5eee..991b4730d768 100644
@@ -37,11 +37,24 @@ static const struct regmap_config attiny_regmap_config = {
 static int attiny_lcd_power_enable(struct regulator_dev *rdev)
 {
 	unsigned int data;
+	int ret, i;
 
 	regmap_write(rdev->regmap, REG_POWERON, 1);
+	msleep(80);
+
 	/* Wait for nPWRDWN to go low to indicate poweron is done. */
-	regmap_read_poll_timeout(rdev->regmap, REG_PORTB, data,
-					data & BIT(0), 10, 1000000);
+	for (i = 0; i < 20; i++) {
+		ret = regmap_read(rdev->regmap, REG_PORTB, &data);
+		if (!ret) {
+			if (data & BIT(0))
+				break;
+		}
+		usleep_range(10000, 12000);
+	}
+	usleep_range(10000, 12000);
+
+	if (ret)
+		pr_err("%s: regmap_read_poll_timeout failed %d\n", __func__, ret);
 
 	/* Default to the same orientation as the closed source
 	 * firmware used for the panel.  Runtime rotation
@@ -57,23 +70,34 @@ static int attiny_lcd_power_disable(struct regulator_dev *rdev)
 {
 	regmap_write(rdev->regmap, REG_PWM, 0);
 	regmap_write(rdev->regmap, REG_POWERON, 0);
-	udelay(1);
+	msleep(30);
 	return 0;
 }
 
 static int attiny_lcd_power_is_enabled(struct regulator_dev *rdev)
 {
 	unsigned int data;
-	int ret;
+	int ret, i;
 
-	ret = regmap_read(rdev->regmap, REG_POWERON, &data);
+	for (i = 0; i < 10; i++) {
+		ret = regmap_read(rdev->regmap, REG_POWERON, &data);
+		if (!ret)
+			break;
+		usleep_range(10000, 12000);
+	}
 	if (ret < 0)
 		return ret;
 
 	if (!(data & BIT(0)))
 		return 0;
 
-	ret = regmap_read(rdev->regmap, REG_PORTB, &data);
+	for (i = 0; i < 10; i++) {
+		ret = regmap_read(rdev->regmap, REG_PORTB, &data);
+		if (!ret)
+			break;
+		usleep_range(10000, 12000);
+	}
+
 	if (ret < 0)
 		return ret;
 
@@ -103,20 +127,32 @@ static int attiny_update_status(struct backlight_device *bl)
 {
 	struct regmap *regmap = bl_get_data(bl);
 	int brightness = bl->props.brightness;
+	int ret, i;
 
 	if (bl->props.power != FB_BLANK_UNBLANK ||
 	    bl->props.fb_blank != FB_BLANK_UNBLANK)
 		brightness = 0;
 
-	return regmap_write(regmap, REG_PWM, brightness);
+	for (i = 0; i < 10; i++) {
+		ret = regmap_write(regmap, REG_PWM, brightness);
+		if (!ret)
+			break;
+	}
+
+	return ret;
 }
 
 static int attiny_get_brightness(struct backlight_device *bl)
 {
 	struct regmap *regmap = bl_get_data(bl);
-	int ret, brightness;
+	int ret, brightness, i;
+
+	for (i = 0; i < 10; i++) {
+		ret = regmap_read(regmap, REG_PWM, &brightness);
+		if (!ret)
+			break;
+	}
 
-	ret = regmap_read(regmap, REG_PWM, &brightness);
 	if (ret)
 		return ret;
 
@@ -166,7 +202,7 @@ static int attiny_i2c_probe(struct i2c_client *i2c,
 	}
 
 	regmap_write(regmap, REG_POWERON, 0);
-	mdelay(1);
+	msleep(30);
 
 	config.dev = &i2c->dev;
 	config.regmap = regmap;
Signature · 1 line
2.34.1
author maintainer
#761 lore raw

PATCH 753/1017 crypto: hisilicon/qm - cleanup warning in qm_vf_read_qos

In reply to Greg Kroah-Hartman

From: Kai Ye <[email protected]>

[ Upstream commit 05b3bade290d6c940701f97f3233c07cfe27205d ]

The kernel test rebot report this warning: Uninitialized variable: ret. The code flow may return value of ret directly. This value is an uninitialized variable, here is fix it.

---
 drivers/crypto/hisilicon/qm.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/crypto/hisilicon/qm.c +1 -1
index 1dc6a27ba0e0..82d2b117e162 100644
@@ -4145,7 +4145,7 @@ static void qm_vf_get_qos(struct hisi_qm *qm, u32 fun_num)
 static int qm_vf_read_qos(struct hisi_qm *qm)
 {
 	int cnt = 0;
-	int ret;
+	int ret = -EINVAL;
 
 	/* reset mailbox qos val */
 	qm->mb_qos = 0;
Signature · 1 line
2.34.1
author maintainer
#762 lore raw

PATCH 754/1017 gcc-plugins/stackleak: Exactly match strings instead of prefixes

In reply to Greg Kroah-Hartman

From: Kees Cook <[email protected]>

[ Upstream commit 27e9faf415dbf94af19b9c827842435edbc1fbbc ]

Since STRING_CST may not be NUL terminated, strncmp() was used for check for equality. However, this may lead to mismatches for longer section names where the start matches the tested-for string. Test for exact equality by checking for the presences of NUL termination.

---
 scripts/gcc-plugins/stackleak_plugin.c | 25 +++++++++++++++++++++----
 1 file changed, 21 insertions(+), 4 deletions(-)
scripts/gcc-plugins/stackleak_plugin.c +21 -4
index e9db7dcb3e5f..b04aa8e91a41 100644
@@ -429,6 +429,23 @@ static unsigned int stackleak_cleanup_execute(void)
 	return 0;
 }
 
+/*
+ * STRING_CST may or may not be NUL terminated:
+ * https://gcc.gnu.org/onlinedocs/gccint/Constant-expressions.html
+ */
+static inline bool string_equal(tree node, const char *string, int length)
+{
+	if (TREE_STRING_LENGTH(node) < length)
+		return false;
+	if (TREE_STRING_LENGTH(node) > length + 1)
+		return false;
+	if (TREE_STRING_LENGTH(node) == length + 1 &&
+	    TREE_STRING_POINTER(node)[length] != '\0')
+		return false;
+	return !memcmp(TREE_STRING_POINTER(node), string, length);
+}
+#define STRING_EQUAL(node, str)	string_equal(node, str, strlen(str))
+
 static bool stackleak_gate(void)
 {
 	tree section;
@@ -438,13 +455,13 @@ static bool stackleak_gate(void)
 	if (section && TREE_VALUE(section)) {
 		section = TREE_VALUE(TREE_VALUE(section));
 
-		if (!strncmp(TREE_STRING_POINTER(section), ".init.text", 10))
+		if (STRING_EQUAL(section, ".init.text"))
 			return false;
-		if (!strncmp(TREE_STRING_POINTER(section), ".devinit.text", 13))
+		if (STRING_EQUAL(section, ".devinit.text"))
 			return false;
-		if (!strncmp(TREE_STRING_POINTER(section), ".cpuinit.text", 13))
+		if (STRING_EQUAL(section, ".cpuinit.text"))
 			return false;
-		if (!strncmp(TREE_STRING_POINTER(section), ".meminit.text", 13))
+		if (STRING_EQUAL(section, ".meminit.text"))
 			return false;
 	}
 
Signature · 1 line
2.34.1
author maintainer
#763 lore raw

PATCH 755/1017 rcu: Kill rnp->ofl_seq and use only rcu_state.ofl_lock for exclusion

In reply to Greg Kroah-Hartman

From: David Woodhouse <[email protected]>

[ Upstream commit 82980b1622d97017053c6792382469d7dc26a486 ]

If we allow architectures to bring APs online in parallel, then we end up requiring rcu_cpu_starting() to be reentrant. But currently, the manipulation of rnp->ofl_seq is not thread-safe.

However, rnp->ofl_seq is also fairly much pointless anyway since both rcu_cpu_starting() and rcu_report_dead() hold rcu_state.ofl_lock for fairly much the whole time that rnp->ofl_seq is set to an odd number to indicate that an operation is in progress.

So drop rnp->ofl_seq completely, and use only rcu_state.ofl_lock.

This has a couple of minor complexities: lockdep will complain when we take rcu_state.ofl_lock, and currently accepts the 'excuse' of having an odd value in rnp->ofl_seq. So switch it to an arch_spinlock_t to avoid that false positive complaint. Since we're killing rnp->ofl_seq of course that 'excuse' has to be changed too, so make it check for arch_spin_is_locked(rcu_state.ofl_lock).

There's no arch_spin_lock_irqsave() so we have to manually save and restore local interrupts around the locking.

At Paul's request based on Neeraj's analysis, make rcu_gp_init not just wait but *exclude* any CPU online/offline activity, which was fairly much true already by virtue of it holding rcu_state.ofl_lock.

---
 kernel/rcu/tree.c | 71 ++++++++++++++++++++++++-----------------------
 kernel/rcu/tree.h |  4 +--
 2 files changed, 37 insertions(+), 38 deletions(-)
kernel/rcu/tree.c +36 -35
index 28fd0cef9b1f..0e70fe0ab690 100644
@@ -91,7 +91,7 @@ static struct rcu_state rcu_state = {
 	.abbr = RCU_ABBR,
 	.exp_mutex = __MUTEX_INITIALIZER(rcu_state.exp_mutex),
 	.exp_wake_mutex = __MUTEX_INITIALIZER(rcu_state.exp_wake_mutex),
-	.ofl_lock = __RAW_SPIN_LOCK_UNLOCKED(rcu_state.ofl_lock),
+	.ofl_lock = __ARCH_SPIN_LOCK_UNLOCKED,
 };
 
 /* Dump rcu_node combining tree at boot to verify correct setup. */
@@ -1168,7 +1168,15 @@ bool rcu_lockdep_current_cpu_online(void)
 	preempt_disable_notrace();
 	rdp = this_cpu_ptr(&rcu_data);
 	rnp = rdp->mynode;
-	if (rdp->grpmask & rcu_rnp_online_cpus(rnp) || READ_ONCE(rnp->ofl_seq) & 0x1)
+	/*
+	 * Strictly, we care here about the case where the current CPU is
+	 * in rcu_cpu_starting() and thus has an excuse for rdp->grpmask
+	 * not being up to date. So arch_spin_is_locked() might have a
+	 * false positive if it's held by some *other* CPU, but that's
+	 * OK because that just means a false *negative* on the warning.
+	 */
+	if (rdp->grpmask & rcu_rnp_online_cpus(rnp) ||
+	    arch_spin_is_locked(&rcu_state.ofl_lock))
 		ret = true;
 	preempt_enable_notrace();
 	return ret;
@@ -1732,7 +1740,6 @@ static void rcu_strict_gp_boundary(void *unused)
  */
 static noinline_for_stack bool rcu_gp_init(void)
 {
-	unsigned long firstseq;
 	unsigned long flags;
 	unsigned long oldmask;
 	unsigned long mask;
@@ -1775,22 +1782,17 @@ static noinline_for_stack bool rcu_gp_init(void)
 	 * of RCU's Requirements documentation.
 	 */
 	WRITE_ONCE(rcu_state.gp_state, RCU_GP_ONOFF);
+	/* Exclude CPU hotplug operations. */
 	rcu_for_each_leaf_node(rnp) {
-		// Wait for CPU-hotplug operations that might have
-		// started before this grace period did.
-		smp_mb(); // Pair with barriers used when updating ->ofl_seq to odd values.
-		firstseq = READ_ONCE(rnp->ofl_seq);
-		if (firstseq & 0x1)
-			while (firstseq == READ_ONCE(rnp->ofl_seq))
-				schedule_timeout_idle(1);  // Can't wake unless RCU is watching.
-		smp_mb(); // Pair with barriers used when updating ->ofl_seq to even values.
-		raw_spin_lock(&rcu_state.ofl_lock);
-		raw_spin_lock_irq_rcu_node(rnp);
+		local_irq_save(flags);
+		arch_spin_lock(&rcu_state.ofl_lock);
+		raw_spin_lock_rcu_node(rnp);
 		if (rnp->qsmaskinit == rnp->qsmaskinitnext &&
 		    !rnp->wait_blkd_tasks) {
 			/* Nothing to do on this leaf rcu_node structure. */
-			raw_spin_unlock_irq_rcu_node(rnp);
-			raw_spin_unlock(&rcu_state.ofl_lock);
+			raw_spin_unlock_rcu_node(rnp);
+			arch_spin_unlock(&rcu_state.ofl_lock);
+			local_irq_restore(flags);
 			continue;
 		}
 
@@ -1825,8 +1827,9 @@ static noinline_for_stack bool rcu_gp_init(void)
 				rcu_cleanup_dead_rnp(rnp);
 		}
 
-		raw_spin_unlock_irq_rcu_node(rnp);
-		raw_spin_unlock(&rcu_state.ofl_lock);
+		raw_spin_unlock_rcu_node(rnp);
+		arch_spin_unlock(&rcu_state.ofl_lock);
+		local_irq_restore(flags);
 	}
 	rcu_gp_slow(gp_preinit_delay); /* Races with CPU hotplug. */
 
@@ -4247,11 +4250,10 @@ void rcu_cpu_starting(unsigned int cpu)
 
 	rnp = rdp->mynode;
 	mask = rdp->grpmask;
-	WRITE_ONCE(rnp->ofl_seq, rnp->ofl_seq + 1);
-	WARN_ON_ONCE(!(rnp->ofl_seq & 0x1));
+	local_irq_save(flags);
+	arch_spin_lock(&rcu_state.ofl_lock);
 	rcu_dynticks_eqs_online();
-	smp_mb(); // Pair with rcu_gp_cleanup()'s ->ofl_seq barrier().
-	raw_spin_lock_irqsave_rcu_node(rnp, flags);
+	raw_spin_lock_rcu_node(rnp);
 	WRITE_ONCE(rnp->qsmaskinitnext, rnp->qsmaskinitnext | mask);
 	newcpu = !(rnp->expmaskinitnext & mask);
 	rnp->expmaskinitnext |= mask;
@@ -4264,15 +4266,18 @@ void rcu_cpu_starting(unsigned int cpu)
 
 	/* An incoming CPU should never be blocking a grace period. */
 	if (WARN_ON_ONCE(rnp->qsmask & mask)) { /* RCU waiting on incoming CPU? */
+		/* rcu_report_qs_rnp() *really* wants some flags to restore */
+		unsigned long flags2;
+
+		local_irq_save(flags2);
 		rcu_disable_urgency_upon_qs(rdp);
 		/* Report QS -after- changing ->qsmaskinitnext! */
-		rcu_report_qs_rnp(mask, rnp, rnp->gp_seq, flags);
+		rcu_report_qs_rnp(mask, rnp, rnp->gp_seq, flags2);
 	} else {
-		raw_spin_unlock_irqrestore_rcu_node(rnp, flags);
+		raw_spin_unlock_rcu_node(rnp);
 	}
-	smp_mb(); // Pair with rcu_gp_cleanup()'s ->ofl_seq barrier().
-	WRITE_ONCE(rnp->ofl_seq, rnp->ofl_seq + 1);
-	WARN_ON_ONCE(rnp->ofl_seq & 0x1);
+	arch_spin_unlock(&rcu_state.ofl_lock);
+	local_irq_restore(flags);
 	smp_mb(); /* Ensure RCU read-side usage follows above initialization. */
 }
 
@@ -4286,7 +4291,7 @@ void rcu_cpu_starting(unsigned int cpu)
  */
 void rcu_report_dead(unsigned int cpu)
 {
-	unsigned long flags;
+	unsigned long flags, seq_flags;
 	unsigned long mask;
 	struct rcu_data *rdp = per_cpu_ptr(&rcu_data, cpu);
 	struct rcu_node *rnp = rdp->mynode;  /* Outgoing CPU's rdp & rnp. */
@@ -4300,10 +4305,8 @@ void rcu_report_dead(unsigned int cpu)
 
 	/* Remove outgoing CPU from mask in the leaf rcu_node structure. */
 	mask = rdp->grpmask;
-	WRITE_ONCE(rnp->ofl_seq, rnp->ofl_seq + 1);
-	WARN_ON_ONCE(!(rnp->ofl_seq & 0x1));
-	smp_mb(); // Pair with rcu_gp_cleanup()'s ->ofl_seq barrier().
-	raw_spin_lock(&rcu_state.ofl_lock);
+	local_irq_save(seq_flags);
+	arch_spin_lock(&rcu_state.ofl_lock);
 	raw_spin_lock_irqsave_rcu_node(rnp, flags); /* Enforce GP memory-order guarantee. */
 	rdp->rcu_ofl_gp_seq = READ_ONCE(rcu_state.gp_seq);
 	rdp->rcu_ofl_gp_flags = READ_ONCE(rcu_state.gp_flags);
@@ -4314,10 +4317,8 @@ void rcu_report_dead(unsigned int cpu)
 	}
 	WRITE_ONCE(rnp->qsmaskinitnext, rnp->qsmaskinitnext & ~mask);
 	raw_spin_unlock_irqrestore_rcu_node(rnp, flags);
-	raw_spin_unlock(&rcu_state.ofl_lock);
-	smp_mb(); // Pair with rcu_gp_cleanup()'s ->ofl_seq barrier().
-	WRITE_ONCE(rnp->ofl_seq, rnp->ofl_seq + 1);
-	WARN_ON_ONCE(rnp->ofl_seq & 0x1);
+	arch_spin_unlock(&rcu_state.ofl_lock);
+	local_irq_restore(seq_flags);
 
 	rdp->cpu_started = false;
 }
kernel/rcu/tree.h +1 -3
index 305cf6aeb408..aff4cc9303fb 100644
@@ -56,8 +56,6 @@ struct rcu_node {
 				/*  Initialized from ->qsmaskinitnext at the */
 				/*  beginning of each grace period. */
 	unsigned long qsmaskinitnext;
-	unsigned long ofl_seq;	/* CPU-hotplug operation sequence count. */
-				/* Online CPUs for next grace period. */
 	unsigned long expmask;	/* CPUs or groups that need to check in */
 				/*  to allow the current expedited GP */
 				/*  to complete. */
@@ -358,7 +356,7 @@ struct rcu_state {
 	const char *name;			/* Name of structure. */
 	char abbr;				/* Abbreviated name. */
 
-	raw_spinlock_t ofl_lock ____cacheline_internodealigned_in_smp;
+	arch_spinlock_t ofl_lock ____cacheline_internodealigned_in_smp;
 						/* Synchronize offline with */
 						/*  GP pre-initialization. */
 };
Signature · 1 line
2.34.1
author maintainer
#764 lore raw

PATCH 756/1017 pinctrl: npcm: Fix broken references to chip->parent_device

In reply to Greg Kroah-Hartman

From: Marc Zyngier <[email protected]>

[ Upstream commit f7e53e2255808ca3abcc8f38d18ad0823425e771 ]

The npcm driver has a bunch of references to the irq_chip parent_device field, but never sets it.

Fix it by fishing that reference from somewhere else, but it is obvious that these debug statements were never used. Also remove an unused field in a local data structure.

---
 drivers/pinctrl/nuvoton/pinctrl-npcm7xx.c | 25 +++++++++++------------
 1 file changed, 12 insertions(+), 13 deletions(-)
drivers/pinctrl/nuvoton/pinctrl-npcm7xx.c +12 -13
index 4d81908d6725..ba536fd4d674 100644
@@ -78,7 +78,6 @@ struct npcm7xx_gpio {
 	struct gpio_chip	gc;
 	int			irqbase;
 	int			irq;
-	void			*priv;
 	struct irq_chip		irq_chip;
 	u32			pinctrl_id;
 	int (*direction_input)(struct gpio_chip *chip, unsigned offset);
@@ -226,7 +225,7 @@ static void npcmgpio_irq_handler(struct irq_desc *desc)
 	chained_irq_enter(chip, desc);
 	sts = ioread32(bank->base + NPCM7XX_GP_N_EVST);
 	en  = ioread32(bank->base + NPCM7XX_GP_N_EVEN);
-	dev_dbg(chip->parent_device, "==> got irq sts %.8x %.8x\n", sts,
+	dev_dbg(bank->gc.parent, "==> got irq sts %.8x %.8x\n", sts,
 		en);
 
 	sts &= en;
@@ -241,33 +240,33 @@ static int npcmgpio_set_irq_type(struct irq_data *d, unsigned int type)
 		gpiochip_get_data(irq_data_get_irq_chip_data(d));
 	unsigned int gpio = BIT(d->hwirq);
 
-	dev_dbg(d->chip->parent_device, "setirqtype: %u.%u = %u\n", gpio,
+	dev_dbg(bank->gc.parent, "setirqtype: %u.%u = %u\n", gpio,
 		d->irq, type);
 	switch (type) {
 	case IRQ_TYPE_EDGE_RISING:
-		dev_dbg(d->chip->parent_device, "edge.rising\n");
+		dev_dbg(bank->gc.parent, "edge.rising\n");
 		npcm_gpio_clr(&bank->gc, bank->base + NPCM7XX_GP_N_EVBE, gpio);
 		npcm_gpio_clr(&bank->gc, bank->base + NPCM7XX_GP_N_POL, gpio);
 		break;
 	case IRQ_TYPE_EDGE_FALLING:
-		dev_dbg(d->chip->parent_device, "edge.falling\n");
+		dev_dbg(bank->gc.parent, "edge.falling\n");
 		npcm_gpio_clr(&bank->gc, bank->base + NPCM7XX_GP_N_EVBE, gpio);
 		npcm_gpio_set(&bank->gc, bank->base + NPCM7XX_GP_N_POL, gpio);
 		break;
 	case IRQ_TYPE_EDGE_BOTH:
-		dev_dbg(d->chip->parent_device, "edge.both\n");
+		dev_dbg(bank->gc.parent, "edge.both\n");
 		npcm_gpio_set(&bank->gc, bank->base + NPCM7XX_GP_N_EVBE, gpio);
 		break;
 	case IRQ_TYPE_LEVEL_LOW:
-		dev_dbg(d->chip->parent_device, "level.low\n");
+		dev_dbg(bank->gc.parent, "level.low\n");
 		npcm_gpio_set(&bank->gc, bank->base + NPCM7XX_GP_N_POL, gpio);
 		break;
 	case IRQ_TYPE_LEVEL_HIGH:
-		dev_dbg(d->chip->parent_device, "level.high\n");
+		dev_dbg(bank->gc.parent, "level.high\n");
 		npcm_gpio_clr(&bank->gc, bank->base + NPCM7XX_GP_N_POL, gpio);
 		break;
 	default:
-		dev_dbg(d->chip->parent_device, "invalid irq type\n");
+		dev_dbg(bank->gc.parent, "invalid irq type\n");
 		return -EINVAL;
 	}
 
@@ -289,7 +288,7 @@ static void npcmgpio_irq_ack(struct irq_data *d)
 		gpiochip_get_data(irq_data_get_irq_chip_data(d));
 	unsigned int gpio = d->hwirq;
 
-	dev_dbg(d->chip->parent_device, "irq_ack: %u.%u\n", gpio, d->irq);
+	dev_dbg(bank->gc.parent, "irq_ack: %u.%u\n", gpio, d->irq);
 	iowrite32(BIT(gpio), bank->base + NPCM7XX_GP_N_EVST);
 }
 
@@ -301,7 +300,7 @@ static void npcmgpio_irq_mask(struct irq_data *d)
 	unsigned int gpio = d->hwirq;
 
 	/* Clear events */
-	dev_dbg(d->chip->parent_device, "irq_mask: %u.%u\n", gpio, d->irq);
+	dev_dbg(bank->gc.parent, "irq_mask: %u.%u\n", gpio, d->irq);
 	iowrite32(BIT(gpio), bank->base + NPCM7XX_GP_N_EVENC);
 }
 
@@ -313,7 +312,7 @@ static void npcmgpio_irq_unmask(struct irq_data *d)
 	unsigned int gpio = d->hwirq;
 
 	/* Enable events */
-	dev_dbg(d->chip->parent_device, "irq_unmask: %u.%u\n", gpio, d->irq);
+	dev_dbg(bank->gc.parent, "irq_unmask: %u.%u\n", gpio, d->irq);
 	iowrite32(BIT(gpio), bank->base + NPCM7XX_GP_N_EVENS);
 }
 
@@ -323,7 +322,7 @@ static unsigned int npcmgpio_irq_startup(struct irq_data *d)
 	unsigned int gpio = d->hwirq;
 
 	/* active-high, input, clear interrupt, enable interrupt */
-	dev_dbg(d->chip->parent_device, "startup: %u.%u\n", gpio, d->irq);
+	dev_dbg(gc->parent, "startup: %u.%u\n", gpio, d->irq);
 	npcmgpio_direction_input(gc, gpio);
 	npcmgpio_irq_ack(d);
 	npcmgpio_irq_unmask(d);
Signature · 1 line
2.34.1
author maintainer
#765 lore raw

PATCH 757/1017 rcu: Mark writes to the rcu_segcblist structures ->flags field

In reply to Greg Kroah-Hartman

From: Paul E. McKenney <[email protected]>

[ Upstream commit c09929031018913b5783872a8b8cdddef4a543c7 ]

KCSAN reports data races between the rcu_segcblist_clear_flags() and rcu_segcblist_set_flags() functions, though misreporting the latter as a call to rcu_segcblist_is_enabled() from call_rcu(). This commit converts the updates of this field to WRITE_ONCE(), relying on the resulting unmarked reads to continue to detect buggy concurrent writes to this field.

---
 kernel/rcu/rcu_segcblist.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
kernel/rcu/rcu_segcblist.h +2 -2
index 9a19328ff251..5d405943823e 100644
@@ -56,13 +56,13 @@ static inline long rcu_segcblist_n_cbs(struct rcu_segcblist *rsclp)
 static inline void rcu_segcblist_set_flags(struct rcu_segcblist *rsclp,
 					   int flags)
 {
-	rsclp->flags |= flags;
+	WRITE_ONCE(rsclp->flags, rsclp->flags | flags);
 }
 
 static inline void rcu_segcblist_clear_flags(struct rcu_segcblist *rsclp,
 					     int flags)
 {
-	rsclp->flags &= ~flags;
+	WRITE_ONCE(rsclp->flags, rsclp->flags & ~flags);
 }
 
 static inline bool rcu_segcblist_test_flags(struct rcu_segcblist *rsclp,
Signature · 1 line
2.34.1
author maintainer
#766 lore raw

PATCH 758/1017 block: throttle split bio in case of iops limit

In reply to Greg Kroah-Hartman

From: Ming Lei <[email protected]>

[ Upstream commit 9f5ede3c01f9951b0ae7d68b28762ad51d9bacc8 ]

Commit 111be8839817 ("block-throttle: avoid double charge") marks bio as BIO_THROTTLED unconditionally if __blk_throtl_bio() is called on this bio, then this bio won't be called into __blk_throtl_bio() any more. This way is to avoid double charge in case of bio splitting. It is reasonable for read/write throughput limit, but not reasonable for IOPS limit because block layer provides io accounting against split bio.

Chunguang Xu has already observed this issue and fixed it in commit 4f1e9630afe6 ("blk-throtl: optimize IOPS throttle for large IO scenarios"). However, that patch only covers bio splitting in __blk_queue_split(), and we have other kind of bio splitting, such as bio_split() & submit_bio_noacct() and other ways.

This patch tries to fix the issue in one generic way by always charging the bio for iops limit in blk_throtl_bio(). This way is reasonable:
re-submission & fast-cloned bio is charged if it is submitted to same disk/queue, and BIO_THROTTLED will be cleared if bio->bi_bdev is changed.

This new approach can get much more smooth/stable iops limit compared with commit 4f1e9630afe6 ("blk-throtl: optimize IOPS throttle for large IO scenarios") since that commit can't throttle current split bios actually.

Also this way won't cause new double bio iops charge in blk_throtl_dispatch_work_fn() in which blk_throtl_bio() won't be called any more.

---
 block/blk-merge.c    |  2 --
 block/blk-throttle.c | 10 +++++++---
 block/blk-throttle.h |  2 --
 3 files changed, 7 insertions(+), 7 deletions(-)
block/blk-merge.c +0 -2
index a5b73bad9462..09bf679dc132 100644
@@ -367,8 +367,6 @@ void __blk_queue_split(struct request_queue *q, struct bio **bio,
 		trace_block_split(split, (*bio)->bi_iter.bi_sector);
 		submit_bio_noacct(*bio);
 		*bio = split;
-
-		blk_throtl_charge_bio_split(*bio);
 	}
 }
 
block/blk-throttle.c +7 -3
index 39bb6e68a9a2..96573fa4edf2 100644
@@ -807,7 +807,8 @@ static bool tg_with_in_bps_limit(struct throtl_grp *tg, struct bio *bio,
 	unsigned long jiffy_elapsed, jiffy_wait, jiffy_elapsed_rnd;
 	unsigned int bio_size = throtl_bio_data_size(bio);
 
-	if (bps_limit == U64_MAX) {
+	/* no need to throttle if this bio's bytes have been accounted */
+	if (bps_limit == U64_MAX || bio_flagged(bio, BIO_THROTTLED)) {
 		if (wait)
 			*wait = 0;
 		return true;
@@ -919,9 +920,12 @@ static void throtl_charge_bio(struct throtl_grp *tg, struct bio *bio)
 	unsigned int bio_size = throtl_bio_data_size(bio);
 
 	/* Charge the bio to the group */
-	tg->bytes_disp[rw] += bio_size;
+	if (!bio_flagged(bio, BIO_THROTTLED)) {
+		tg->bytes_disp[rw] += bio_size;
+		tg->last_bytes_disp[rw] += bio_size;
+	}
+
 	tg->io_disp[rw]++;
-	tg->last_bytes_disp[rw] += bio_size;
 	tg->last_io_disp[rw]++;
 
 	/*
block/blk-throttle.h +0 -2
index 175f03abd9e4..cb43f4417d6e 100644
@@ -170,8 +170,6 @@ static inline bool blk_throtl_bio(struct bio *bio)
 {
 	struct throtl_grp *tg = blkg_to_tg(bio->bi_blkg);
 
-	if (bio_flagged(bio, BIO_THROTTLED))
-		return false;
 	if (!tg->has_rules[bio_data_dir(bio)])
 		return false;
 
Signature · 1 line
2.34.1
author maintainer
#767 lore raw

PATCH 759/1017 memstick/mspro_block: fix handling of read-only devices

In reply to Greg Kroah-Hartman

From: Christoph Hellwig <[email protected]>

[ Upstream commit 6dab421bfe06a59bf8f212a72e34673e8acf2018 ]

Use set_disk_ro to propagate the read-only state to the block layer instead of checking for it in ->open and leaking a reference in case of a read-only device.

---
 drivers/memstick/core/mspro_block.c | 10 ++++------
 1 file changed, 4 insertions(+), 6 deletions(-)
drivers/memstick/core/mspro_block.c +4 -6
index c0450397b673..7ea312f0840e 100644
@@ -186,13 +186,8 @@ static int mspro_block_bd_open(struct block_device *bdev, fmode_t mode)
 
 	mutex_lock(&mspro_block_disk_lock);
 
-	if (msb && msb->card) {
+	if (msb && msb->card)
 		msb->usage_count++;
-		if ((mode & FMODE_WRITE) && msb->read_only)
-			rc = -EROFS;
-		else
-			rc = 0;
-	}
 
 	mutex_unlock(&mspro_block_disk_lock);
 
@@ -1239,6 +1234,9 @@ static int mspro_block_init_disk(struct memstick_dev *card)
 	set_capacity(msb->disk, capacity);
 	dev_dbg(&card->dev, "capacity set %ld\n", capacity);
 
+	if (msb->read_only)
+		set_disk_ro(msb->disk, true);
+
 	rc = device_add_disk(&card->dev, msb->disk, NULL);
 	if (rc)
 		goto out_cleanup_disk;
Signature · 1 line
2.34.1
author maintainer
#768 lore raw

PATCH 760/1017 block/bfq_wf2q: correct weight to ioprio

In reply to Greg Kroah-Hartman

From: Yahu Gao <[email protected]>

[ Upstream commit bcd2be763252f3a4d5fc4d6008d4d96c601ee74b ]

The return value is ioprio * BFQ_WEIGHT_CONVERSION_COEFF or 0. What we want is ioprio or 0.
Correct this by changing the calculation.

---
 block/bfq-wf2q.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
block/bfq-wf2q.c +1 -1
index b74cc0da118e..709b901de3ca 100644
@@ -519,7 +519,7 @@ unsigned short bfq_ioprio_to_weight(int ioprio)
 static unsigned short bfq_weight_to_ioprio(int weight)
 {
 	return max_t(int, 0,
-		     IOPRIO_NR_LEVELS * BFQ_WEIGHT_CONVERSION_COEFF - weight);
+		     IOPRIO_NR_LEVELS - weight / BFQ_WEIGHT_CONVERSION_COEFF);
 }
 
 static void bfq_get_entity(struct bfq_entity *entity)
Signature · 1 line
2.34.1
author maintainer
#769 lore raw

PATCH 761/1017 crypto: xts - Add softdep on ecb

In reply to Greg Kroah-Hartman

From: Herbert Xu <[email protected]>

[ Upstream commit dfe085d8dcd0bb1fe20cc2327e81c8064cead441 ]

The xts module needs ecb to be present as it's meant to work on top of ecb. This patch adds a softdep so ecb can be included automatically into the initramfs.

---
 crypto/xts.c | 1 +
 1 file changed, 1 insertion(+)
crypto/xts.c +1 -0
index 6c12f30dbdd6..63c85b9e64e0 100644
@@ -466,3 +466,4 @@ MODULE_LICENSE("GPL");
 MODULE_DESCRIPTION("XTS block cipher mode");
 MODULE_ALIAS_CRYPTO("xts");
 MODULE_IMPORT_NS(CRYPTO_INTERNAL);
+MODULE_SOFTDEP("pre: ecb");
Signature · 1 line
2.34.1
author maintainer
#770 lore raw

PATCH 762/1017 crypto: hisilicon/sec - not need to enable sm4 extra mode at HW V3

In reply to Greg Kroah-Hartman

From: Kai Ye <[email protected]>

[ Upstream commit f8a2652826444d13181061840b96a5d975d5b6c6 ]

It is not need to enable sm4 extra mode in at HW V3. Here is fix it.

---
 drivers/crypto/hisilicon/sec2/sec_main.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)
drivers/crypto/hisilicon/sec2/sec_main.c +5 -3
index 90551bf38b52..03d239cfdf8c 100644
@@ -443,9 +443,11 @@ static int sec_engine_init(struct hisi_qm *qm)
 
 	writel(SEC_SAA_ENABLE, qm->io_base + SEC_SAA_EN_REG);
 
-	/* Enable sm4 extra mode, as ctr/ecb */
-	writel_relaxed(SEC_BD_ERR_CHK_EN0,
-		       qm->io_base + SEC_BD_ERR_CHK_EN_REG0);
+	/* HW V2 enable sm4 extra mode, as ctr/ecb */
+	if (qm->ver < QM_HW_V3)
+		writel_relaxed(SEC_BD_ERR_CHK_EN0,
+			       qm->io_base + SEC_BD_ERR_CHK_EN_REG0);
+
 	/* Enable sm4 xts mode multiple iv */
 	writel_relaxed(SEC_BD_ERR_CHK_EN1,
 		       qm->io_base + SEC_BD_ERR_CHK_EN_REG1);
Signature · 1 line
2.34.1
author maintainer
#771 lore raw

PATCH 763/1017 block, bfq: dont move oom_bfqq

In reply to Greg Kroah-Hartman

From: Yu Kuai <[email protected]>

[ Upstream commit 8410f70977734f21b8ed45c37e925d311dfda2e7 ]

Our test report a UAF:

[ 2073.019181] ==================================================================
[ 2073.019188] BUG: KASAN: use-after-free in __bfq_put_async_bfqq+0xa0/0x168
[ 2073.019191] Write of size 8 at addr ffff8000ccf64128 by task rmmod/72584
[ 2073.019192]
[ 2073.019196] CPU: 0 PID: 72584 Comm: rmmod Kdump: loaded Not tainted 4.19.90-yk #5
[ 2073.019198] Hardware name: QEMU KVM Virtual Machine, BIOS 0.0.0 02/06/2015
[ 2073.019200] Call trace:
[ 2073.019203]  dump_backtrace+0x0/0x310
[ 2073.019206]  show_stack+0x28/0x38
[ 2073.019210]  dump_stack+0xec/0x15c
[ 2073.019216]  print_address_description+0x68/0x2d0
[ 2073.019220]  kasan_report+0x238/0x2f0
[ 2073.019224]  __asan_store8+0x88/0xb0
[ 2073.019229]  __bfq_put_async_bfqq+0xa0/0x168
[ 2073.019233]  bfq_put_async_queues+0xbc/0x208
[ 2073.019236]  bfq_pd_offline+0x178/0x238
[ 2073.019240]  blkcg_deactivate_policy+0x1f0/0x420
[ 2073.019244]  bfq_exit_queue+0x128/0x178
[ 2073.019249]  blk_mq_exit_sched+0x12c/0x160
[ 2073.019252]  elevator_exit+0xc8/0xd0
[ 2073.019256]  blk_exit_queue+0x50/0x88
[ 2073.019259]  blk_cleanup_queue+0x228/0x3d8
[ 2073.019267]  null_del_dev+0xfc/0x1e0 [null_blk]
[ 2073.019274]  null_exit+0x90/0x114 [null_blk]
[ 2073.019278]  __arm64_sys_delete_module+0x358/0x5a0
[ 2073.019282]  el0_svc_common+0xc8/0x320
[ 2073.019287]  el0_svc_handler+0xf8/0x160
[ 2073.019290]  el0_svc+0x10/0x218
[ 2073.019291]
[ 2073.019294] Allocated by task 14163:
[ 2073.019301]  kasan_kmalloc+0xe0/0x190
[ 2073.019305]  kmem_cache_alloc_node_trace+0x1cc/0x418
[ 2073.019308]  bfq_pd_alloc+0x54/0x118
[ 2073.019313]  blkcg_activate_policy+0x250/0x460
[ 2073.019317]  bfq_create_group_hierarchy+0x38/0x110
[ 2073.019321]  bfq_init_queue+0x6d0/0x948
[ 2073.019325]  blk_mq_init_sched+0x1d8/0x390
[ 2073.019330]  elevator_switch_mq+0x88/0x170
[ 2073.019334]  elevator_switch+0x140/0x270
[ 2073.019338]  elv_iosched_store+0x1a4/0x2a0
[ 2073.019342]  queue_attr_store+0x90/0xe0
[ 2073.019348]  sysfs_kf_write+0xa8/0xe8
[ 2073.019351]  kernfs_fop_write+0x1f8/0x378
[ 2073.019359]  __vfs_write+0xe0/0x360
[ 2073.019363]  vfs_write+0xf0/0x270
[ 2073.019367]  ksys_write+0xdc/0x1b8
[ 2073.019371]  __arm64_sys_write+0x50/0x60
[ 2073.019375]  el0_svc_common+0xc8/0x320
[ 2073.019380]  el0_svc_handler+0xf8/0x160
[ 2073.019383]  el0_svc+0x10/0x218
[ 2073.019385]
[ 2073.019387] Freed by task 72584:
[ 2073.019391]  __kasan_slab_free+0x120/0x228
[ 2073.019394]  kasan_slab_free+0x10/0x18
[ 2073.019397]  kfree+0x94/0x368
[ 2073.019400]  bfqg_put+0x64/0xb0
[ 2073.019404]  bfqg_and_blkg_put+0x90/0xb0
[ 2073.019408]  bfq_put_queue+0x220/0x228
[ 2073.019413]  __bfq_put_async_bfqq+0x98/0x168
[ 2073.019416]  bfq_put_async_queues+0xbc/0x208
[ 2073.019420]  bfq_pd_offline+0x178/0x238
[ 2073.019424]  blkcg_deactivate_policy+0x1f0/0x420
[ 2073.019429]  bfq_exit_queue+0x128/0x178
[ 2073.019433]  blk_mq_exit_sched+0x12c/0x160
[ 2073.019437]  elevator_exit+0xc8/0xd0
[ 2073.019440]  blk_exit_queue+0x50/0x88
[ 2073.019443]  blk_cleanup_queue+0x228/0x3d8
[ 2073.019451]  null_del_dev+0xfc/0x1e0 [null_blk]
[ 2073.019459]  null_exit+0x90/0x114 [null_blk]
[ 2073.019462]  __arm64_sys_delete_module+0x358/0x5a0
[ 2073.019467]  el0_svc_common+0xc8/0x320
[ 2073.019471]  el0_svc_handler+0xf8/0x160
[ 2073.019474]  el0_svc+0x10/0x218
[ 2073.019475]
[ 2073.019479] The buggy address belongs to the object at ffff8000ccf63f00
 which belongs to the cache kmalloc-1024 of size 1024
[ 2073.019484] The buggy address is located 552 bytes inside of
 1024-byte region [ffff8000ccf63f00, ffff8000ccf64300)
[ 2073.019486] The buggy address belongs to the page:
[ 2073.019492] page:ffff7e000333d800 count:1 mapcount:0 mapping:ffff8000c0003a00 index:0x0 compound_mapcount: 0
[ 2073.020123] flags: 0x7ffff0000008100(slab|head)
[ 2073.020403] raw: 07ffff0000008100 ffff7e0003334c08 ffff7e00001f5a08 ffff8000c0003a00
[ 2073.020409] raw: 0000000000000000 00000000001c001c 00000001ffffffff 0000000000000000
[ 2073.020411] page dumped because: kasan: bad access detected
[ 2073.020412]
[ 2073.020414] Memory state around the buggy address:
[ 2073.020420]  ffff8000ccf64000: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
[ 2073.020424]  ffff8000ccf64080: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
[ 2073.020428] >ffff8000ccf64100: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
[ 2073.020430]                                   ^
[ 2073.020434]  ffff8000ccf64180: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
[ 2073.020438]  ffff8000ccf64200: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
[ 2073.020439] ==================================================================

The same problem exist in mainline as well.

This is because oom_bfqq is moved to a non-root group, thus root_group is freed earlier.

Thus fix the problem by don't move oom_bfqq.

---
 block/bfq-cgroup.c | 6 ++++++
 1 file changed, 6 insertions(+)
block/bfq-cgroup.c +6 -0
index 24a5c5329bcd..809bc612d96b 100644
@@ -646,6 +646,12 @@ void bfq_bfqq_move(struct bfq_data *bfqd, struct bfq_queue *bfqq,
 {
 	struct bfq_entity *entity = &bfqq->entity;
 
+	/*
+	 * oom_bfqq is not allowed to move, oom_bfqq will hold ref to root_group
+	 * until elevator exit.
+	 */
+	if (bfqq == &bfqd->oom_bfqq)
+		return;
 	/*
 	 * Get extra reference to prevent bfqq from being freed in
 	 * next possible expire or deactivate.
Signature · 1 line
2.34.1
author maintainer
#772 lore raw

PATCH 764/1017 selinux: use correct type for context length

In reply to Greg Kroah-Hartman

From: Christian Göttsche <[email protected]>

[ Upstream commit b97df7c098c531010e445da88d02b7bf7bf59ef6 ]

security_sid_to_context() expects a pointer to an u32 as the address where to store the length of the computed context.

Reported by sparse:

security/selinux/xfrm.c:359:39: warning: incorrect type in arg 4
(different signedness)
security/selinux/xfrm.c:359:39: expected unsigned int
[usertype] *scontext_len
security/selinux/xfrm.c:359:39: got int *

[PM: wrapped commit description]
---
 security/selinux/xfrm.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
security/selinux/xfrm.c +1 -1
index be83e5ce4469..debe15207d2b 100644
@@ -347,7 +347,7 @@ int selinux_xfrm_state_alloc_acquire(struct xfrm_state *x,
 	int rc;
 	struct xfrm_sec_ctx *ctx;
 	char *ctx_str = NULL;
-	int str_len;
+	u32 str_len;
 
 	if (!polsec)
 		return 0;
Signature · 1 line
2.34.1
author maintainer
#773 lore raw

PATCH 765/1017 powercap/dtpm_cpu: Reset per_cpu variable in the release function

In reply to Greg Kroah-Hartman

From: Daniel Lezcano <[email protected]>

[ Upstream commit 0aea2e4ec2a2bfa2d7e8820e37ba5b5ce04f20a5 ]

The release function does not reset the per cpu variable when it is called. That will prevent creation again as the variable will be already from the previous creation.

Fix it by resetting them.

---
 drivers/powercap/dtpm_cpu.c | 7 +++++++
 1 file changed, 7 insertions(+)
drivers/powercap/dtpm_cpu.c +7 -0
index b740866b228d..1e8cac699646 100644
@@ -150,10 +150,17 @@ static int update_pd_power_uw(struct dtpm *dtpm)
 static void pd_release(struct dtpm *dtpm)
 {
 	struct dtpm_cpu *dtpm_cpu = to_dtpm_cpu(dtpm);
+	struct cpufreq_policy *policy;
 
 	if (freq_qos_request_active(&dtpm_cpu->qos_req))
 		freq_qos_remove_request(&dtpm_cpu->qos_req);
 
+	policy = cpufreq_cpu_get(dtpm_cpu->cpu);
+	if (policy) {
+		for_each_cpu(dtpm_cpu->cpu, policy->related_cpus)
+			per_cpu(dtpm_per_cpu, dtpm_cpu->cpu) = NULL;
+	}
+	
 	kfree(dtpm_cpu);
 }
 
Signature · 1 line
2.34.1
author maintainer
#774 lore raw

PATCH 766/1017 arm64: module: remove (NOLOAD) from linker script

In reply to Greg Kroah-Hartman

From: Fangrui Song <[email protected]>

[ Upstream commit 4013e26670c590944abdab56c4fa797527b74325 ]

On ELF, (NOLOAD) sets the section type to SHT_NOBITS[1]. It is conceptually inappropriate for .plt and .text.* sections which are always SHT_PROGBITS.

In GNU ld, if PLT entries are needed, .plt will be SHT_PROGBITS anyway and (NOLOAD) will be essentially ignored. In ld.lld, since
https://reviews.llvm.org/D118840 ("[ELF] Support (TYPE=<value>) to customize the output section type"), ld.lld will report a section type mismatch error. Just remove (NOLOAD) to fix the error.

[1] https://lld.llvm.org/ELF/linker_script.html As of today, "The section should be marked as not loadable" on
https://sourceware.org/binutils/docs/ld/Output-Section-Type.html is outdated for ELF.

---
 arch/arm64/include/asm/module.lds.h | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
arch/arm64/include/asm/module.lds.h +3 -3
index a11ccadd47d2..094701ec5500 100644
@@ -1,8 +1,8 @@
 SECTIONS {
 #ifdef CONFIG_ARM64_MODULE_PLTS
-	.plt 0 (NOLOAD) : { BYTE(0) }
-	.init.plt 0 (NOLOAD) : { BYTE(0) }
-	.text.ftrace_trampoline 0 (NOLOAD) : { BYTE(0) }
+	.plt 0 : { BYTE(0) }
+	.init.plt 0 : { BYTE(0) }
+	.text.ftrace_trampoline 0 : { BYTE(0) }
 #endif
 
 #ifdef CONFIG_KASAN_SW_TAGS
Signature · 1 line
2.34.1
author maintainer
#775 lore raw

PATCH 767/1017 selinux: allow FIOCLEX and FIONCLEX with policy capability

In reply to Greg Kroah-Hartman

From: Richard Haines <[email protected]>

[ Upstream commit 65881e1db4e948614d9eb195b8e1197339822949 ]

These ioctls are equivalent to fcntl(fd, F_SETFD, flags), which SELinux always allows too. Furthermore, a failed FIOCLEX could result in a file descriptor being leaked to a process that should not have access to it.

As this patch removes access controls, a policy capability needs to be enabled in policy to always allow these ioctls.

---
 security/selinux/hooks.c                   | 6 ++++++
 security/selinux/include/policycap.h       | 1 +
 security/selinux/include/policycap_names.h | 3 ++-
 security/selinux/include/security.h        | 7 +++++++
 4 files changed, 16 insertions(+), 1 deletion(-)
security/selinux/hooks.c +6 -0
index 69b0709bd156..94ef617de9d0 100644
@@ -3800,6 +3800,12 @@ static int selinux_file_ioctl(struct file *file, unsigned int cmd,
 					    CAP_OPT_NONE, true);
 		break;
 
+	case FIOCLEX:
+	case FIONCLEX:
+		if (!selinux_policycap_ioctl_skip_cloexec())
+			error = ioctl_has_perm(cred, file, FILE__IOCTL, (u16) cmd);
+		break;
+
 	/* default case assumes that the command will go
 	 * to the file's ioctl() function.
 	 */
security/selinux/include/policycap.h +1 -0
index 2ec038efbb03..a9e572ca4fd9 100644
@@ -11,6 +11,7 @@ enum {
 	POLICYDB_CAPABILITY_CGROUPSECLABEL,
 	POLICYDB_CAPABILITY_NNP_NOSUID_TRANSITION,
 	POLICYDB_CAPABILITY_GENFS_SECLABEL_SYMLINKS,
+	POLICYDB_CAPABILITY_IOCTL_SKIP_CLOEXEC,
 	__POLICYDB_CAPABILITY_MAX
 };
 #define POLICYDB_CAPABILITY_MAX (__POLICYDB_CAPABILITY_MAX - 1)
security/selinux/include/policycap_names.h +2 -1
index b89289f092c9..ebd64afe1def 100644
@@ -12,7 +12,8 @@ const char *selinux_policycap_names[__POLICYDB_CAPABILITY_MAX] = {
 	"always_check_network",
 	"cgroup_seclabel",
 	"nnp_nosuid_transition",
-	"genfs_seclabel_symlinks"
+	"genfs_seclabel_symlinks",
+	"ioctl_skip_cloexec"
 };
 
 #endif /* _SELINUX_POLICYCAP_NAMES_H_ */
security/selinux/include/security.h +7 -0
index ac0ece01305a..c0d966020ebd 100644
@@ -219,6 +219,13 @@ static inline bool selinux_policycap_genfs_seclabel_symlinks(void)
 	return READ_ONCE(state->policycap[POLICYDB_CAPABILITY_GENFS_SECLABEL_SYMLINKS]);
 }
 
+static inline bool selinux_policycap_ioctl_skip_cloexec(void)
+{
+	struct selinux_state *state = &selinux_state;
+
+	return READ_ONCE(state->policycap[POLICYDB_CAPABILITY_IOCTL_SKIP_CLOEXEC]);
+}
+
 struct selinux_policy_convert_data;
 
 struct selinux_load_state {
Signature · 1 line
2.34.1
author maintainer
#776 lore raw

PATCH 768/1017 loop: use sysfs_emit() in the sysfs xxx show()

In reply to Greg Kroah-Hartman

From: Chaitanya Kulkarni <[email protected]>

[ Upstream commit b27824d31f09ea7b4a6ba2c1b18bd328df3e8bed ]

sprintf does not know the PAGE_SIZE maximum of the temporary buffer used for outputting sysfs content and it's possible to overrun the PAGE_SIZE buffer length.

Use a generic sysfs_emit function that knows the size of the temporary buffer and ensures that no overrun is done for offset attribute in loop_attr_[offset|sizelimit|autoclear|partscan|dio]_show() callbacks.

---
 drivers/block/loop.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)
drivers/block/loop.c +5 -5
index fdb4798cb006..1422967bf5c1 100644
@@ -681,33 +681,33 @@ static ssize_t loop_attr_backing_file_show(struct loop_device *lo, char *buf)
 
 static ssize_t loop_attr_offset_show(struct loop_device *lo, char *buf)
 {
-	return sprintf(buf, "%llu\n", (unsigned long long)lo->lo_offset);
+	return sysfs_emit(buf, "%llu\n", (unsigned long long)lo->lo_offset);
 }
 
 static ssize_t loop_attr_sizelimit_show(struct loop_device *lo, char *buf)
 {
-	return sprintf(buf, "%llu\n", (unsigned long long)lo->lo_sizelimit);
+	return sysfs_emit(buf, "%llu\n", (unsigned long long)lo->lo_sizelimit);
 }
 
 static ssize_t loop_attr_autoclear_show(struct loop_device *lo, char *buf)
 {
 	int autoclear = (lo->lo_flags & LO_FLAGS_AUTOCLEAR);
 
-	return sprintf(buf, "%s\n", autoclear ? "1" : "0");
+	return sysfs_emit(buf, "%s\n", autoclear ? "1" : "0");
 }
 
 static ssize_t loop_attr_partscan_show(struct loop_device *lo, char *buf)
 {
 	int partscan = (lo->lo_flags & LO_FLAGS_PARTSCAN);
 
-	return sprintf(buf, "%s\n", partscan ? "1" : "0");
+	return sysfs_emit(buf, "%s\n", partscan ? "1" : "0");
 }
 
 static ssize_t loop_attr_dio_show(struct loop_device *lo, char *buf)
 {
 	int dio = (lo->lo_flags & LO_FLAGS_DIRECT_IO);
 
-	return sprintf(buf, "%s\n", dio ? "1" : "0");
+	return sysfs_emit(buf, "%s\n", dio ? "1" : "0");
 }
 
 LOOP_ATTR_RO(backing_file);
Signature · 1 line
2.34.1
author maintainer
#777 lore raw

PATCH 769/1017 Fix incorrect type in assignment of ipv6 port for audit

In reply to Greg Kroah-Hartman

From: Casey Schaufler <[email protected]>

[ Upstream commit a5cd1ab7ab679d252a6d2f483eee7d45ebf2040c ]

Remove inappropriate use of ntohs() and assign the port value directly.

---
 security/smack/smack_lsm.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
security/smack/smack_lsm.c +1 -1
index efd35b07c7f8..95a15b77f42f 100644
@@ -2511,7 +2511,7 @@ static int smk_ipv6_check(struct smack_known *subject,
 #ifdef CONFIG_AUDIT
 	smk_ad_init_net(&ad, __func__, LSM_AUDIT_DATA_NET, &net);
 	ad.a.u.net->family = PF_INET6;
-	ad.a.u.net->dport = ntohs(address->sin6_port);
+	ad.a.u.net->dport = address->sin6_port;
 	if (act == SMK_RECEIVING)
 		ad.a.u.net->v6info.saddr = address->sin6_addr;
 	else
Signature · 1 line
2.34.1
author maintainer
#778 lore raw

PATCH 770/1017 irqchip/qcom-pdc: Fix broken locking

In reply to Greg Kroah-Hartman

From: Marc Zyngier <[email protected]>

[ Upstream commit a6aca2f460e203781dc41391913cc5b54f4bc0ce ]

pdc_enable_intr() serves as a primitive to qcom_pdc_gic_{en,dis}able, and has a raw spinlock for mutual exclusion, which is uses with interruptible primitives.

This means that this critical section can itself be interrupted. Should the interrupt also be a PDC interrupt, and the endpoint driver perform an irq_disable() on that interrupt, we end-up in a deadlock.

Fix this by using the irqsave/irqrestore variants of the locking primitives.

---
 drivers/irqchip/qcom-pdc.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)
drivers/irqchip/qcom-pdc.c +3 -2
index 173e6520e06e..c0b457f26ec4 100644
@@ -56,17 +56,18 @@ static u32 pdc_reg_read(int reg, u32 i)
 static void pdc_enable_intr(struct irq_data *d, bool on)
 {
 	int pin_out = d->hwirq;
+	unsigned long flags;
 	u32 index, mask;
 	u32 enable;
 
 	index = pin_out / 32;
 	mask = pin_out % 32;
 
-	raw_spin_lock(&pdc_lock);
+	raw_spin_lock_irqsave(&pdc_lock, flags);
 	enable = pdc_reg_read(IRQ_ENABLE_BANK, index);
 	enable = on ? ENABLE_INTR(enable, mask) : CLEAR_INTR(enable, mask);
 	pdc_reg_write(IRQ_ENABLE_BANK, index, enable);
-	raw_spin_unlock(&pdc_lock);
+	raw_spin_unlock_irqrestore(&pdc_lock, flags);
 }
 
 static void qcom_pdc_gic_disable(struct irq_data *d)
Signature · 1 line
2.34.1
author maintainer
#779 lore raw

PATCH 771/1017 irqchip/nvic: Release nvic_base upon failure

In reply to Greg Kroah-Hartman

From: Souptick Joarder (HPE) <[email protected]>

[ Upstream commit e414c25e3399b2b3d7337dc47abccab5c71b7c8f ]

smatch warning was reported as below ->

smatch warnings:
drivers/irqchip/irq-nvic.c:131 nvic_of_init()
warn: 'nvic_base' not released on lines: 97.

Release nvic_base upon failure.

---
 drivers/irqchip/irq-nvic.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/irqchip/irq-nvic.c +2 -0
index ba4759b3e269..94230306e0ee 100644
@@ -107,6 +107,7 @@ static int __init nvic_of_init(struct device_node *node,
 
 	if (!nvic_irq_domain) {
 		pr_warn("Failed to allocate irq domain\n");
+		iounmap(nvic_base);
 		return -ENOMEM;
 	}
 
@@ -116,6 +117,7 @@ static int __init nvic_of_init(struct device_node *node,
 	if (ret) {
 		pr_warn("Failed to allocate irq chips\n");
 		irq_domain_remove(nvic_irq_domain);
+		iounmap(nvic_base);
 		return ret;
 	}
 
Signature · 1 line
2.34.1
author maintainer
#780 lore raw

PATCH 772/1017 fs/binfmt_elf: Fix AT_PHDR for unusual ELF files

In reply to Greg Kroah-Hartman

From: Akira Kawata <[email protected]>

[ Upstream commit 0da1d5002745cdc721bc018b582a8a9704d56c42 ]

As pointed out in the discussion of buglink, we cannot calculate AT_PHDR as the sum of load_addr and exec->e_phoff.

: The AT_PHDR of ELF auxiliary vectors should point to the memory address : of program header. But binfmt_elf.c calculates this address as follows:
:
: NEW_AUX_ENT(AT_PHDR, load_addr + exec->e_phoff);
:
: which is wrong since e_phoff is the file offset of program header and : load_addr is the memory base address from PT_LOAD entry.
:
: The ld.so uses AT_PHDR as the memory address of program header. In normal : case, since the e_phoff is usually 64 and in the first PT_LOAD region, it : is the correct program header address.
:
: But if the address of program header isn't equal to the first PT_LOAD : address + e_phoff (e.g. Put the program header in other non-consecutive : PT_LOAD region), ld.so will try to read program header from wrong address : then crash or use incorrect program header.

This is because exec->e_phoff
is the offset of PHDRs in the file and the address of PHDRs in the memory may differ from it. This patch fixes the bug by calculating the address of program headers from PT_LOADs directly.

---
 fs/binfmt_elf.c | 24 ++++++++++++++++++------
 1 file changed, 18 insertions(+), 6 deletions(-)
fs/binfmt_elf.c +18 -6
index d19762dc90fe..c4de845f86c8 100644
@@ -170,8 +170,8 @@ static int padzero(unsigned long elf_bss)
 
 static int
 create_elf_tables(struct linux_binprm *bprm, const struct elfhdr *exec,
-		unsigned long load_addr, unsigned long interp_load_addr,
-		unsigned long e_entry)
+		unsigned long interp_load_addr,
+		unsigned long e_entry, unsigned long phdr_addr)
 {
 	struct mm_struct *mm = current->mm;
 	unsigned long p = bprm->p;
@@ -257,7 +257,7 @@ create_elf_tables(struct linux_binprm *bprm, const struct elfhdr *exec,
 	NEW_AUX_ENT(AT_HWCAP, ELF_HWCAP);
 	NEW_AUX_ENT(AT_PAGESZ, ELF_EXEC_PAGESIZE);
 	NEW_AUX_ENT(AT_CLKTCK, CLOCKS_PER_SEC);
-	NEW_AUX_ENT(AT_PHDR, load_addr + exec->e_phoff);
+	NEW_AUX_ENT(AT_PHDR, phdr_addr);
 	NEW_AUX_ENT(AT_PHENT, sizeof(struct elf_phdr));
 	NEW_AUX_ENT(AT_PHNUM, exec->e_phnum);
 	NEW_AUX_ENT(AT_BASE, interp_load_addr);
@@ -823,7 +823,7 @@ static int parse_elf_properties(struct file *f, const struct elf_phdr *phdr,
 static int load_elf_binary(struct linux_binprm *bprm)
 {
 	struct file *interpreter = NULL; /* to shut gcc up */
- 	unsigned long load_addr = 0, load_bias = 0;
+	unsigned long load_addr, load_bias = 0, phdr_addr = 0;
 	int load_addr_set = 0;
 	unsigned long error;
 	struct elf_phdr *elf_ppnt, *elf_phdata, *interp_elf_phdata = NULL;
@@ -1180,6 +1180,17 @@ static int load_elf_binary(struct linux_binprm *bprm)
 				reloc_func_desc = load_bias;
 			}
 		}
+
+		/*
+		 * Figure out which segment in the file contains the Program
+		 * Header table, and map to the associated memory address.
+		 */
+		if (elf_ppnt->p_offset <= elf_ex->e_phoff &&
+		    elf_ex->e_phoff < elf_ppnt->p_offset + elf_ppnt->p_filesz) {
+			phdr_addr = elf_ex->e_phoff - elf_ppnt->p_offset +
+				    elf_ppnt->p_vaddr;
+		}
+
 		k = elf_ppnt->p_vaddr;
 		if ((elf_ppnt->p_flags & PF_X) && k < start_code)
 			start_code = k;
@@ -1215,6 +1226,7 @@ static int load_elf_binary(struct linux_binprm *bprm)
 	}
 
 	e_entry = elf_ex->e_entry + load_bias;
+	phdr_addr += load_bias;
 	elf_bss += load_bias;
 	elf_brk += load_bias;
 	start_code += load_bias;
@@ -1278,8 +1290,8 @@ static int load_elf_binary(struct linux_binprm *bprm)
 		goto out;
 #endif /* ARCH_HAS_SETUP_ADDITIONAL_PAGES */
 
-	retval = create_elf_tables(bprm, elf_ex,
-			  load_addr, interp_load_addr, e_entry);
+	retval = create_elf_tables(bprm, elf_ex, interp_load_addr,
+				   e_entry, phdr_addr);
 	if (retval < 0)
 		goto out;
 
Signature · 1 line
2.34.1
author maintainer
#781 lore raw

PATCH 773/1017 bfq: fix use-after-free in bfq_dispatch_request

In reply to Greg Kroah-Hartman

From: Zhang Wensheng <[email protected]>

[ Upstream commit ab552fcb17cc9e4afe0e4ac4df95fc7b30e8490a ]

KASAN reports a use-after-free report when doing normal scsi-mq test

[69832.239032] ==================================================================
[69832.241810] BUG: KASAN: use-after-free in bfq_dispatch_request+0x1045/0x44b0
[69832.243267] Read of size 8 at addr ffff88802622ba88 by task kworker/3:1H/155
[69832.244656]
[69832.245007] CPU: 3 PID: 155 Comm: kworker/3:1H Not tainted 5.10.0-10295-g576c6382529e #8
[69832.246626] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.14.0-0-g155821a1990b-prebuilt.qemu.org 04/01/2014
[69832.249069] Workqueue: kblockd blk_mq_run_work_fn
[69832.250022] Call Trace:
[69832.250541]  dump_stack+0x9b/0xce
[69832.251232]  ? bfq_dispatch_request+0x1045/0x44b0
[69832.252243]  print_address_description.constprop.6+0x3e/0x60
[69832.253381]  ? __cpuidle_text_end+0x5/0x5
[69832.254211]  ? vprintk_func+0x6b/0x120
[69832.254994]  ? bfq_dispatch_request+0x1045/0x44b0
[69832.255952]  ? bfq_dispatch_request+0x1045/0x44b0
[69832.256914]  kasan_report.cold.9+0x22/0x3a
[69832.257753]  ? bfq_dispatch_request+0x1045/0x44b0
[69832.258755]  check_memory_region+0x1c1/0x1e0
[69832.260248]  bfq_dispatch_request+0x1045/0x44b0
[69832.261181]  ? bfq_bfqq_expire+0x2440/0x2440
[69832.262032]  ? blk_mq_delay_run_hw_queues+0xf9/0x170
[69832.263022]  __blk_mq_do_dispatch_sched+0x52f/0x830
[69832.264011]  ? blk_mq_sched_request_inserted+0x100/0x100
[69832.265101]  __blk_mq_sched_dispatch_requests+0x398/0x4f0
[69832.266206]  ? blk_mq_do_dispatch_ctx+0x570/0x570
[69832.267147]  ? __switch_to+0x5f4/0xee0
[69832.267898]  blk_mq_sched_dispatch_requests+0xdf/0x140
[69832.268946]  __blk_mq_run_hw_queue+0xc0/0x270
[69832.269840]  blk_mq_run_work_fn+0x51/0x60
[69832.278170]  process_one_work+0x6d4/0xfe0
[69832.278984]  worker_thread+0x91/0xc80
[69832.279726]  ? __kthread_parkme+0xb0/0x110
[69832.280554]  ? process_one_work+0xfe0/0xfe0
[69832.281414]  kthread+0x32d/0x3f0
[69832.282082]  ? kthread_park+0x170/0x170
[69832.282849]  ret_from_fork+0x1f/0x30
[69832.283573]
[69832.283886] Allocated by task 7725:
[69832.284599]  kasan_save_stack+0x19/0x40
[69832.285385]  __kasan_kmalloc.constprop.2+0xc1/0xd0
[69832.286350]  kmem_cache_alloc_node+0x13f/0x460
[69832.287237]  bfq_get_queue+0x3d4/0x1140
[69832.287993]  bfq_get_bfqq_handle_split+0x103/0x510
[69832.289015]  bfq_init_rq+0x337/0x2d50
[69832.289749]  bfq_insert_requests+0x304/0x4e10
[69832.290634]  blk_mq_sched_insert_requests+0x13e/0x390
[69832.291629]  blk_mq_flush_plug_list+0x4b4/0x760
[69832.292538]  blk_flush_plug_list+0x2c5/0x480
[69832.293392]  io_schedule_prepare+0xb2/0xd0
[69832.294209]  io_schedule_timeout+0x13/0x80
[69832.295014]  wait_for_common_io.constprop.1+0x13c/0x270
[69832.296137]  submit_bio_wait+0x103/0x1a0
[69832.296932]  blkdev_issue_discard+0xe6/0x160
[69832.297794]  blk_ioctl_discard+0x219/0x290
[69832.298614]  blkdev_common_ioctl+0x50a/0x1750
[69832.304715]  blkdev_ioctl+0x470/0x600
[69832.305474]  block_ioctl+0xde/0x120
[69832.306232]  vfs_ioctl+0x6c/0xc0
[69832.306877]  __se_sys_ioctl+0x90/0xa0
[69832.307629]  do_syscall_64+0x2d/0x40
[69832.308362]  entry_SYSCALL_64_after_hwframe+0x44/0xa9
[69832.309382]
[69832.309701] Freed by task 155:
[69832.310328]  kasan_save_stack+0x19/0x40
[69832.311121]  kasan_set_track+0x1c/0x30
[69832.311868]  kasan_set_free_info+0x1b/0x30
[69832.312699]  __kasan_slab_free+0x111/0x160
[69832.313524]  kmem_cache_free+0x94/0x460
[69832.314367]  bfq_put_queue+0x582/0x940
[69832.315112]  __bfq_bfqd_reset_in_service+0x166/0x1d0
[69832.317275]  bfq_bfqq_expire+0xb27/0x2440
[69832.318084]  bfq_dispatch_request+0x697/0x44b0
[69832.318991]  __blk_mq_do_dispatch_sched+0x52f/0x830
[69832.319984]  __blk_mq_sched_dispatch_requests+0x398/0x4f0
[69832.321087]  blk_mq_sched_dispatch_requests+0xdf/0x140
[69832.322225]  __blk_mq_run_hw_queue+0xc0/0x270
[69832.323114]  blk_mq_run_work_fn+0x51/0x60
[69832.323942]  process_one_work+0x6d4/0xfe0
[69832.324772]  worker_thread+0x91/0xc80
[69832.325518]  kthread+0x32d/0x3f0
[69832.326205]  ret_from_fork+0x1f/0x30
[69832.326932]
[69832.338297] The buggy address belongs to the object at ffff88802622b968
[69832.338297]  which belongs to the cache bfq_queue of size 512
[69832.340766] The buggy address is located 288 bytes inside of
[69832.340766]  512-byte region [ffff88802622b968, ffff88802622bb68)
[69832.343091] The buggy address belongs to the page:
[69832.344097] page:ffffea0000988a00 refcount:1 mapcount:0 mapping:0000000000000000 index:0xffff88802622a528 pfn:0x26228
[69832.346214] head:ffffea0000988a00 order:2 compound_mapcount:0 compound_pincount:0
[69832.347719] flags: 0x1fffff80010200(slab|head)
[69832.348625] raw: 001fffff80010200 ffffea0000dbac08 ffff888017a57650 ffff8880179fe840
[69832.354972] raw: ffff88802622a528 0000000000120008 00000001ffffffff 0000000000000000
[69832.356547] page dumped because: kasan: bad access detected
[69832.357652]
[69832.357970] Memory state around the buggy address:
[69832.358926]  ffff88802622b980: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
[69832.360358]  ffff88802622ba00: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
[69832.361810] >ffff88802622ba80: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
[69832.363273]                       ^
[69832.363975]  ffff88802622bb00: fb fb fb fb fb fb fb fb fb fb fb fb fb fc fc fc
[69832.375960]  ffff88802622bb80: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
[69832.377405] ==================================================================

In bfq_dispatch_requestfunction, it may have function call:

bfq_dispatch_request
__bfq_dispatch_request
bfq_select_queue
bfq_bfqq_expire
__bfq_bfqd_reset_in_service
bfq_put_queue
kmem_cache_free
In this function call, in_serv_queue has beed expired and meet the conditions to free. In the function bfq_dispatch_request, the address of in_serv_queue pointing to has been released. For getting the value of idle_timer_disabled, it will get flags value from the address which in_serv_queue pointing to, then the problem of use-after-free happens;

Fix the problem by check in_serv_queue == bfqd->in_service_queue, to get the value of idle_timer_disabled if in_serve_queue is equel to bfqd->in_service_queue. If the space of in_serv_queue pointing has been released, this judge will aviod use-after-free problem. And if in_serv_queue may be expired or finished, the idle_timer_disabled will be false which would not give effects to bfq_update_dispatch_stats.

---
 block/bfq-iosched.c | 15 ++++++++-------
 1 file changed, 8 insertions(+), 7 deletions(-)
block/bfq-iosched.c +8 -7
index 8c0950c9a1a2..7eecd3200d83 100644
@@ -5061,7 +5061,7 @@ static struct request *bfq_dispatch_request(struct blk_mq_hw_ctx *hctx)
 	struct bfq_data *bfqd = hctx->queue->elevator->elevator_data;
 	struct request *rq;
 	struct bfq_queue *in_serv_queue;
-	bool waiting_rq, idle_timer_disabled;
+	bool waiting_rq, idle_timer_disabled = false;
 
 	spin_lock_irq(&bfqd->lock);
 
@@ -5069,14 +5069,15 @@ static struct request *bfq_dispatch_request(struct blk_mq_hw_ctx *hctx)
 	waiting_rq = in_serv_queue && bfq_bfqq_wait_request(in_serv_queue);
 
 	rq = __bfq_dispatch_request(hctx);
-
-	idle_timer_disabled =
-		waiting_rq && !bfq_bfqq_wait_request(in_serv_queue);
+	if (in_serv_queue == bfqd->in_service_queue) {
+		idle_timer_disabled =
+			waiting_rq && !bfq_bfqq_wait_request(in_serv_queue);
+	}
 
 	spin_unlock_irq(&bfqd->lock);
-
-	bfq_update_dispatch_stats(hctx->queue, rq, in_serv_queue,
-				  idle_timer_disabled);
+	bfq_update_dispatch_stats(hctx->queue, rq,
+			idle_timer_disabled ? in_serv_queue : NULL,
+				idle_timer_disabled);
 
 	return rq;
 }
Signature · 1 line
2.34.1
author maintainer
#782 lore raw

PATCH 774/1017 ACPICA: Avoid walking the ACPI Namespace if it is not there

In reply to Greg Kroah-Hartman

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

[ Upstream commit 0c9992315e738e7d6e927ef36839a466b080dba6 ]

ACPICA commit b1c3656ef4950098e530be68d4b589584f06cddc

Prevent acpi_ns_walk_namespace() from crashing when called with start_node equal to ACPI_ROOT_OBJECT if the Namespace has not been instantiated yet and acpi_gbl_root_node is NULL.

For instance, this can happen if the kernel is run with "acpi=off" in the command line.

---
 drivers/acpi/acpica/nswalk.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/acpi/acpica/nswalk.c +3 -0
index 915c2433463d..e7c30ce06e18 100644
@@ -169,6 +169,9 @@ acpi_ns_walk_namespace(acpi_object_type type,
 
 	if (start_node == ACPI_ROOT_OBJECT) {
 		start_node = acpi_gbl_root_node;
+		if (!start_node) {
+			return_ACPI_STATUS(AE_NO_NAMESPACE);
+		}
 	}
 
 	/* Null child means "get first node" */
Signature · 1 line
2.34.1
author maintainer
#783 lore raw

PATCH 775/1017 lib/raid6/test/Makefile: Use $(pound) instead of \# for Make 4.3

In reply to Greg Kroah-Hartman

From: Paul Menzel <[email protected]>

[ Upstream commit 633174a7046ec3b4572bec24ef98e6ee89bce14b ]

Buidling raid6test on Ubuntu 21.10 (ppc64le) with GNU Make 4.3 shows the errors below:

    $ cd lib/raid6/test/
    $ make
    <stdin>:1:1: error: stray ‘\’ in program
    <stdin>:1:2: error: stray ‘#’ in program
    <stdin>:1:11: error: expected ‘=’, ‘,’, ‘;’, ‘asm’ or ‘__attribute__’ \
        before ‘<’ token
    [...]

The errors come from the HAS_ALTIVEC test, which fails, and the POWER optimized versions are not built. That’s also reason nobody noticed on the other architectures.

GNU Make 4.3 does not remove the backslash anymore. From the 4.3 release announcment:

Quoted 11 lines

* WARNING: Backward-incompatibility!
Number signs (#) appearing inside a macro reference or function invocation no longer introduce comments and should not be escaped with backslashes:
thus a call such as:

    foo := $(shell echo '#')
  is legal.  Previously the number sign needed to be escaped, for example:
    foo := $(shell echo '\#')
  Now this latter will resolve to "\#".  If you want to write makefiles
  portable to both versions, assign the number sign to a variable:
    H := \#
    foo := $(shell echo '$H')
Quoted 2 lines

This was claimed to be fixed in 3.81, but wasn't, for some reason. To detect this change search for 'nocomment' in the .FEATURES variable.

So, do the same as commit 9564a8cf422d ("Kbuild: fix # escaping in .cmd files for future Make") and commit 929bef467771 ("bpf: Use $(pound) instead of \# in Makefiles") and define and use a $(pound) variable.

Reference for the change in make:
https://git.savannah.gnu.org/cgit/make.git/commit/?id=c6966b323811c37acedff05b57

---
 lib/raid6/test/Makefile | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
lib/raid6/test/Makefile +3 -1
index a4c7cd74cff5..4fb7700a741b 100644
@@ -4,6 +4,8 @@
 # from userspace.
 #
 
+pound := \#
+
 CC	 = gcc
 OPTFLAGS = -O2			# Adjust as desired
 CFLAGS	 = -I.. -I ../../../include -g $(OPTFLAGS)
@@ -42,7 +44,7 @@ else ifeq ($(HAS_NEON),yes)
         OBJS   += neon.o neon1.o neon2.o neon4.o neon8.o recov_neon.o recov_neon_inner.o
         CFLAGS += -DCONFIG_KERNEL_MODE_NEON=1
 else
-        HAS_ALTIVEC := $(shell printf '\#include <altivec.h>\nvector int a;\n' |\
+        HAS_ALTIVEC := $(shell printf '$(pound)include <altivec.h>\nvector int a;\n' |\
                          gcc -c -x c - >/dev/null && rm ./-.o && echo yes)
         ifeq ($(HAS_ALTIVEC),yes)
                 CFLAGS += -I../../../arch/powerpc/include
Signature · 1 line
2.34.1
author maintainer
#784 lore raw

PATCH 776/1017 Revert "Revert "block, bfq: honor already-setup queue merges""

In reply to Greg Kroah-Hartman

From: Paolo Valente <[email protected]>

[ Upstream commit 15729ff8143f8135b03988a100a19e66d7cb7ecd ]

A crash [1] happened to be triggered in conjunction with commit 2d52c58b9c9b ("block, bfq: honor already-setup queue merges"). The latter was then reverted by commit ebc69e897e17 ("Revert "block, bfq:
honor already-setup queue merges""). Yet, the reverted commit was not the one introducing the bug. In fact, it actually triggered a UAF introduced by a different commit, and now fixed by commit d29bd41428cf ("block, bfq: reset last_bfqq_created on group change").

So, there is no point in keeping commit 2d52c58b9c9b ("block, bfq:
honor already-setup queue merges") out. This commit restores it.

[1] https://bugzilla.kernel.org/show_bug.cgi?id=214503

---
 block/bfq-iosched.c | 16 +++++++++++++---
 1 file changed, 13 insertions(+), 3 deletions(-)
block/bfq-iosched.c +13 -3
index 7eecd3200d83..342e927f213b 100644
@@ -2662,6 +2662,15 @@ bfq_setup_merge(struct bfq_queue *bfqq, struct bfq_queue *new_bfqq)
 	 * are likely to increase the throughput.
 	 */
 	bfqq->new_bfqq = new_bfqq;
+	/*
+	 * The above assignment schedules the following redirections:
+	 * each time some I/O for bfqq arrives, the process that
+	 * generated that I/O is disassociated from bfqq and
+	 * associated with new_bfqq. Here we increases new_bfqq->ref
+	 * in advance, adding the number of processes that are
+	 * expected to be associated with new_bfqq as they happen to
+	 * issue I/O.
+	 */
 	new_bfqq->ref += process_refs;
 	return new_bfqq;
 }
@@ -2724,6 +2733,10 @@ bfq_setup_cooperator(struct bfq_data *bfqd, struct bfq_queue *bfqq,
 {
 	struct bfq_queue *in_service_bfqq, *new_bfqq;
 
+	/* if a merge has already been setup, then proceed with that first */
+	if (bfqq->new_bfqq)
+		return bfqq->new_bfqq;
+
 	/*
 	 * Check delayed stable merge for rotational or non-queueing
 	 * devs. For this branch to be executed, bfqq must not be
@@ -2825,9 +2838,6 @@ bfq_setup_cooperator(struct bfq_data *bfqd, struct bfq_queue *bfqq,
 	if (bfq_too_late_for_merging(bfqq))
 		return NULL;
 
-	if (bfqq->new_bfqq)
-		return bfqq->new_bfqq;
-
 	if (!io_struct || unlikely(bfqq == &bfqd->oom_bfqq))
 		return NULL;
 
Signature · 1 line
2.34.1
author maintainer
#785 lore raw

PATCH 777/1017 ACPI/APEI: Limit printable size of BERT table data

In reply to Greg Kroah-Hartman

From: Darren Hart <[email protected]>

[ Upstream commit 3f8dec116210ca649163574ed5f8df1e3b837d07 ]

Platforms with large BERT table data can trigger soft lockup errors while attempting to print the entire BERT table data to the console at boot:

  watchdog: BUG: soft lockup - CPU#160 stuck for 23s! [swapper/0:1]

Observed on Ampere Altra systems with a single BERT record of ~250KB.

The original bert driver appears to have assumed relatively small table data. Since it is impractical to reassemble large table data from interwoven console messages, and the table data is available in

/sys/firmware/acpi/tables/data/BERT

limit the size for tables printed to the console to 1024 (for no reason other than it seemed like a good place to kick off the discussion, would appreciate feedback from existing users in terms of what size would maintain their current usage model).

Alternatively, we could make printing a CONFIG option, use the bert_disable boot arg (or something similar), or use a debug log level. However, all those solutions require extra steps or change the existing behavior for small table data. Limiting the size preserves existing behavior on existing platforms with small table data, and eliminates the soft lockups for platforms with large table data, while still making it available.

---
 drivers/acpi/apei/bert.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)
drivers/acpi/apei/bert.c +6 -2
index 86211422f4ee..598fd19b65fa 100644
@@ -29,6 +29,7 @@
 
 #undef pr_fmt
 #define pr_fmt(fmt) "BERT: " fmt
+#define ACPI_BERT_PRINT_MAX_LEN 1024
 
 static int bert_disable;
 
@@ -58,8 +59,11 @@ static void __init bert_print_all(struct acpi_bert_region *region,
 		}
 
 		pr_info_once("Error records from previous boot:\n");
-
-		cper_estatus_print(KERN_INFO HW_ERR, estatus);
+		if (region_len < ACPI_BERT_PRINT_MAX_LEN)
+			cper_estatus_print(KERN_INFO HW_ERR, estatus);
+		else
+			pr_info_once("Max print length exceeded, table data is available at:\n"
+				     "/sys/firmware/acpi/tables/data/BERT");
 
 		/*
 		 * Because the boot error source is "one-time polled" type,
Signature · 1 line
2.34.1
author maintainer
#786 lore raw

PATCH 778/1017 PM: core: keep irq flags in device_pm_check_callbacks()

In reply to Greg Kroah-Hartman

From: Dmitry Baryshkov <[email protected]>

[ Upstream commit 524bb1da785a7ae43dd413cd392b5071c6c367f8 ]

The function device_pm_check_callbacks() can be called under the spin lock (in the reported case it happens from genpd_add_device() -> dev_pm_domain_set(), when the genpd uses spinlocks rather than mutexes.

However this function uncoditionally uses spin_lock_irq() / spin_unlock_irq(), thus not preserving the CPU flags. Use the irqsave/irqrestore instead.

The backtrace for the reference:

[    2.752010] ------------[ cut here ]------------
[    2.756769] raw_local_irq_restore() called with IRQs enabled
[    2.762596] WARNING: CPU: 4 PID: 1 at kernel/locking/irqflag-debug.c:10 warn_bogus_irq_restore+0x34/0x50
[    2.772338] Modules linked in:
[    2.775487] CPU: 4 PID: 1 Comm: swapper/0 Tainted: G S                5.17.0-rc6-00384-ge330d0d82eff-dirty #684
[    2.781384] Freeing initrd memory: 46024K
[    2.785839] pstate: 60400005 (nZCv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--)
[    2.785841] pc : warn_bogus_irq_restore+0x34/0x50
[    2.785844] lr : warn_bogus_irq_restore+0x34/0x50
[    2.785846] sp : ffff80000805b7d0
[    2.785847] x29: ffff80000805b7d0 x28: 0000000000000000 x27: 0000000000000002
[    2.785850] x26: ffffd40e80930b18 x25: ffff7ee2329192b8 x24: ffff7edfc9f60800
[    2.785853] x23: ffffd40e80930b18 x22: ffffd40e80930d30 x21: ffff7edfc0dffa00
[    2.785856] x20: ffff7edfc09e3768 x19: 0000000000000000 x18: ffffffffffffffff
[    2.845775] x17: 6572206f74206465 x16: 6c696166203a3030 x15: ffff80008805b4f7
[    2.853108] x14: 0000000000000000 x13: ffffd40e809550b0 x12: 00000000000003d8
[    2.860441] x11: 0000000000000148 x10: ffffd40e809550b0 x9 : ffffd40e809550b0
[    2.867774] x8 : 00000000ffffefff x7 : ffffd40e809ad0b0 x6 : ffffd40e809ad0b0
[    2.875107] x5 : 000000000000bff4 x4 : 0000000000000000 x3 : 0000000000000000
[    2.882440] x2 : 0000000000000000 x1 : 0000000000000000 x0 : ffff7edfc03a8000
[    2.889774] Call trace:
[    2.892290]  warn_bogus_irq_restore+0x34/0x50
[    2.896770]  _raw_spin_unlock_irqrestore+0x94/0xa0
[    2.901690]  genpd_unlock_spin+0x20/0x30
[    2.905724]  genpd_add_device+0x100/0x2d0
[    2.909850]  __genpd_dev_pm_attach+0xa8/0x23c
[    2.914329]  genpd_dev_pm_attach_by_id+0xc4/0x190
[    2.919167]  genpd_dev_pm_attach_by_name+0x3c/0xd0
[    2.924086]  dev_pm_domain_attach_by_name+0x24/0x30
[    2.929102]  psci_dt_attach_cpu+0x24/0x90
[    2.933230]  psci_cpuidle_probe+0x2d4/0x46c
[    2.937534]  platform_probe+0x68/0xe0
[    2.941304]  really_probe.part.0+0x9c/0x2fc
[    2.945605]  __driver_probe_device+0x98/0x144
[    2.950085]  driver_probe_device+0x44/0x15c
[    2.954385]  __device_attach_driver+0xb8/0x120
[    2.958950]  bus_for_each_drv+0x78/0xd0
[    2.962896]  __device_attach+0xd8/0x180
[    2.966843]  device_initial_probe+0x14/0x20
[    2.971144]  bus_probe_device+0x9c/0xa4
[    2.975092]  device_add+0x380/0x88c
[    2.978679]  platform_device_add+0x114/0x234
[    2.983067]  platform_device_register_full+0x100/0x190
[    2.988344]  psci_idle_init+0x6c/0xb0
[    2.992113]  do_one_initcall+0x74/0x3a0
[    2.996060]  kernel_init_freeable+0x2fc/0x384
[    3.000543]  kernel_init+0x28/0x130
[    3.004132]  ret_from_fork+0x10/0x20
[    3.007817] irq event stamp: 319826
[    3.011404] hardirqs last  enabled at (319825): [<ffffd40e7eda0268>] __up_console_sem+0x78/0x84
[    3.020332] hardirqs last disabled at (319826): [<ffffd40e7fd6d9d8>] el1_dbg+0x24/0x8c
[    3.028458] softirqs last  enabled at (318312): [<ffffd40e7ec90410>] _stext+0x410/0x588
[    3.036678] softirqs last disabled at (318299): [<ffffd40e7ed1bf68>] __irq_exit_rcu+0x158/0x174
[    3.045607] ---[ end trace 0000000000000000 ]---
---
 drivers/base/power/main.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/base/power/main.c +4 -2
index 04ea92cbd9cf..08c8a69d7b81 100644
@@ -2018,7 +2018,9 @@ static bool pm_ops_is_empty(const struct dev_pm_ops *ops)
 
 void device_pm_check_callbacks(struct device *dev)
 {
-	spin_lock_irq(&dev->power.lock);
+	unsigned long flags;
+
+	spin_lock_irqsave(&dev->power.lock, flags);
 	dev->power.no_pm_callbacks =
 		(!dev->bus || (pm_ops_is_empty(dev->bus->pm) &&
 		 !dev->bus->suspend && !dev->bus->resume)) &&
@@ -2027,7 +2029,7 @@ void device_pm_check_callbacks(struct device *dev)
 		(!dev->pm_domain || pm_ops_is_empty(&dev->pm_domain->ops)) &&
 		(!dev->driver || (pm_ops_is_empty(dev->driver->pm) &&
 		 !dev->driver->suspend && !dev->driver->resume));
-	spin_unlock_irq(&dev->power.lock);
+	spin_unlock_irqrestore(&dev->power.lock, flags);
 }
 
 bool dev_pm_skip_suspend(struct device *dev)
Signature · 1 line
2.34.1
author maintainer
#787 lore raw

PATCH 779/1017 parisc: Fix non-access data TLB cache flush faults

In reply to Greg Kroah-Hartman

From: John David Anglin <[email protected]>

[ Upstream commit f839e5f1cef36ce268950c387129b1bfefdaebc9 ]

When a page is not present, we get non-access data TLB faults from the fdc and fic instructions in flush_user_dcache_range_asm and flush_user_icache_range_asm. When these occur, the cache line is not invalidated and potentially we get memory corruption. The problem was hidden by the nullification of the flush instructions.

These faults also affect performance. With pa8800/pa8900 processors, there will be 32 faults per 4 KB page since the cache line is 128 bytes. There will be more faults with earlier processors.

The problem is fixed by using flush_cache_pages(). It does the flush using a tmp alias mapping.

The flush_cache_pages() call in flush_cache_range() flushed too large a range.

V2: Remove unnecessary preempt_disable() and preempt_enable() calls.

---
 arch/parisc/kernel/cache.c | 28 +---------------------------
 1 file changed, 1 insertion(+), 27 deletions(-)
arch/parisc/kernel/cache.c +1 -27
index 94150b91c96f..bce71cefe572 100644
@@ -558,15 +558,6 @@ static void flush_cache_pages(struct vm_area_struct *vma, struct mm_struct *mm,
 	}
 }
 
-static void flush_user_cache_tlb(struct vm_area_struct *vma,
-				 unsigned long start, unsigned long end)
-{
-	flush_user_dcache_range_asm(start, end);
-	if (vma->vm_flags & VM_EXEC)
-		flush_user_icache_range_asm(start, end);
-	flush_tlb_range(vma, start, end);
-}
-
 void flush_cache_mm(struct mm_struct *mm)
 {
 	struct vm_area_struct *vma;
@@ -581,17 +572,8 @@ void flush_cache_mm(struct mm_struct *mm)
 		return;
 	}
 
-	preempt_disable();
-	if (mm->context == mfsp(3)) {
-		for (vma = mm->mmap; vma; vma = vma->vm_next)
-			flush_user_cache_tlb(vma, vma->vm_start, vma->vm_end);
-		preempt_enable();
-		return;
-	}
-
 	for (vma = mm->mmap; vma; vma = vma->vm_next)
 		flush_cache_pages(vma, mm, vma->vm_start, vma->vm_end);
-	preempt_enable();
 }
 
 void flush_cache_range(struct vm_area_struct *vma,
@@ -605,15 +587,7 @@ void flush_cache_range(struct vm_area_struct *vma,
 		return;
 	}
 
-	preempt_disable();
-	if (vma->vm_mm->context == mfsp(3)) {
-		flush_user_cache_tlb(vma, start, end);
-		preempt_enable();
-		return;
-	}
-
-	flush_cache_pages(vma, vma->vm_mm, vma->vm_start, vma->vm_end);
-	preempt_enable();
+	flush_cache_pages(vma, vma->vm_mm, start, end);
 }
 
 void
Signature · 1 line
2.34.1
author maintainer
#788 lore raw

PATCH 780/1017 parisc: Fix handling off probe non-access faults

In reply to Greg Kroah-Hartman

From: John David Anglin <[email protected]>

[ Upstream commit e00b0a2ab8ec019c344e53bfc76e31c18bb587b7 ]

Currently, the parisc kernel does not fully support non-access TLB fault handling for probe instructions. In the fast path, we set the target register to zero if it is not a shadowed register. The slow path is not implemented, so we call do_page_fault. The architecture indicates that non-access faults should not cause a page fault from disk.

This change adds to code to provide non-access fault support for probe instructions. It also modifies the handling of faults on userspace so that if the address lies in a valid VMA and the access type matches that for the VMA, the probe target register is set to one. Otherwise, the target register is set to zero.

This was done to make probe instructions more useful for userspace. Probe instructions are not very useful if they set the target register to zero whenever a page is not present in memory. Nominally, the purpose of the probe instruction is determine whether read or write access to a given address is allowed.

This fixes a problem in function pointer comparison noticed in the glibc testsuite (stdio-common/tst-vfprintf-user-type). The same problem is likely in glibc (_dl_lookup_address).

V2 adds flush and lpa instruction support to handle_nadtlb_fault.

---
 arch/parisc/include/asm/traps.h |  1 +
 arch/parisc/kernel/traps.c      |  2 +
 arch/parisc/mm/fault.c          | 89 +++++++++++++++++++++++++++++++++
 3 files changed, 92 insertions(+)
arch/parisc/include/asm/traps.h +1 -0
index 34619f010c63..0ccdb738a9a3 100644
@@ -18,6 +18,7 @@ unsigned long parisc_acctyp(unsigned long code, unsigned int inst);
 const char *trap_name(unsigned long code);
 void do_page_fault(struct pt_regs *regs, unsigned long code,
 		unsigned long address);
+int handle_nadtlb_fault(struct pt_regs *regs);
 #endif
 
 #endif
arch/parisc/kernel/traps.c +2 -0
index eb41fece1910..b56aab7141ed 100644
@@ -662,6 +662,8 @@ void notrace handle_interruption(int code, struct pt_regs *regs)
 			 by hand. Technically we need to emulate:
 			 fdc,fdce,pdc,"fic,4f",prober,probeir,probew, probeiw
 		*/
+		if (code == 17 && handle_nadtlb_fault(regs))
+			return;
 		fault_address = regs->ior;
 		fault_space = regs->isr;
 		break;
arch/parisc/mm/fault.c +89 -0

92 lines not on this page · raw message

Signature · 1 line
2.34.1
author maintainer
#789 lore raw

PATCH 781/1017 nvme-tcp: lockdep: annotate in-kernel sockets

In reply to Greg Kroah-Hartman

From: Chris Leech <[email protected]>

[ Upstream commit 841aee4d75f18fdfb53935080b03de0c65e9b92c ]

Put NVMe/TCP sockets in their own class to avoid some lockdep warnings. Sockets created by nvme-tcp are not exposed to user-space, and will not trigger certain code paths that the general socket API exposes.

Lockdep complains about a circular dependency between the socket and filesystem locks, because setsockopt can trigger a page fault with a socket lock held, but nvme-tcp sends requests on the socket while file system locks are held.

======================================================
WARNING: possible circular locking dependency detected 5.15.0-rc3 #1 Not tainted
------------------------------------------------------
fio/1496 is trying to acquire lock:
(sk_lock-AF_INET){+.+.}-{0:0}, at: tcp_sendpage+0x23/0x80

but task is already holding lock:
(&xfs_dir_ilock_class/5){+.+.}-{3:3}, at: xfs_ilock+0xcf/0x290 [xfs]

which lock already depends on the new lock.

other info that might help us debug this:

chain exists of:
sk_lock-AF_INET --> sb_internal --> &xfs_dir_ilock_class/5

  Possible unsafe locking scenario:
        CPU0                    CPU1
        ----                    ----
   lock(&xfs_dir_ilock_class/5);
                                lock(sb_internal);
                                lock(&xfs_dir_ilock_class/5);

lock(sk_lock-AF_INET);

*** DEADLOCK ***

6 locks held by fio/1496:
#0: (sb_writers#13){.+.+}-{0:0}, at: path_openat+0x9fc/0xa20
#1: (&inode->i_sb->s_type->i_mutex_dir_key){++++}-{3:3}, at: path_openat+0x296/0xa20 #2: (sb_internal){.+.+}-{0:0}, at: xfs_trans_alloc_icreate+0x41/0xd0 [xfs] #3: (&xfs_dir_ilock_class/5){+.+.}-{3:3}, at: xfs_ilock+0xcf/0x290 [xfs] #4: (hctx->srcu){....}-{0:0}, at: hctx_lock+0x51/0xd0
#5: (&queue->send_mutex){+.+.}-{3:3}, at: nvme_tcp_queue_rq+0x33e/0x380 [nvme_tcp]

This annotation lets lockdep analyze nvme-tcp controlled sockets independently of what the user-space sockets API does.

---
 drivers/nvme/host/tcp.c | 40 ++++++++++++++++++++++++++++++++++++++++
 1 file changed, 40 insertions(+)
drivers/nvme/host/tcp.c +40 -0
index 65e00c64a588..d66e2de044e0 100644
@@ -30,6 +30,44 @@ static int so_priority;
 module_param(so_priority, int, 0644);
 MODULE_PARM_DESC(so_priority, "nvme tcp socket optimize priority");
 
+#ifdef CONFIG_DEBUG_LOCK_ALLOC
+/* lockdep can detect a circular dependency of the form
+ *   sk_lock -> mmap_lock (page fault) -> fs locks -> sk_lock
+ * because dependencies are tracked for both nvme-tcp and user contexts. Using
+ * a separate class prevents lockdep from conflating nvme-tcp socket use with
+ * user-space socket API use.
+ */
+static struct lock_class_key nvme_tcp_sk_key[2];
+static struct lock_class_key nvme_tcp_slock_key[2];
+
+static void nvme_tcp_reclassify_socket(struct socket *sock)
+{
+	struct sock *sk = sock->sk;
+
+	if (WARN_ON_ONCE(!sock_allow_reclassification(sk)))
+		return;
+
+	switch (sk->sk_family) {
+	case AF_INET:
+		sock_lock_init_class_and_name(sk, "slock-AF_INET-NVME",
+					      &nvme_tcp_slock_key[0],
+					      "sk_lock-AF_INET-NVME",
+					      &nvme_tcp_sk_key[0]);
+		break;
+	case AF_INET6:
+		sock_lock_init_class_and_name(sk, "slock-AF_INET6-NVME",
+					      &nvme_tcp_slock_key[1],
+					      "sk_lock-AF_INET6-NVME",
+					      &nvme_tcp_sk_key[1]);
+		break;
+	default:
+		WARN_ON_ONCE(1);
+	}
+}
+#else
+static void nvme_tcp_reclassify_socket(struct socket *sock) { }
+#endif
+
 enum nvme_tcp_send_state {
 	NVME_TCP_SEND_CMD_PDU = 0,
 	NVME_TCP_SEND_H2C_PDU,
@@ -1469,6 +1507,8 @@ static int nvme_tcp_alloc_queue(struct nvme_ctrl *nctrl,
 		goto err_destroy_mutex;
 	}
 
+	nvme_tcp_reclassify_socket(queue->sock);
+
 	/* Single syn retry */
 	tcp_sock_set_syncnt(queue->sock->sk, 1);
 
Signature · 1 line
2.34.1
author maintainer
#790 lore raw

PATCH 782/1017 spi: tegra20: Use of_device_get_match_data()

In reply to Greg Kroah-Hartman

From: Minghao Chi <[email protected]>

[ Upstream commit c9839acfcbe20ce43d363c2a9d0772472d9921c0 ]

Use of_device_get_match_data() to simplify the code.

---
 drivers/spi/spi-tegra20-slink.c | 8 +-------
 1 file changed, 1 insertion(+), 7 deletions(-)
drivers/spi/spi-tegra20-slink.c +1 -7
index e8204e155484..746bbafa1378 100644
@@ -1003,14 +1003,8 @@ static int tegra_slink_probe(struct platform_device *pdev)
 	struct resource		*r;
 	int ret, spi_irq;
 	const struct tegra_slink_chip_data *cdata = NULL;
-	const struct of_device_id *match;
 
-	match = of_match_device(tegra_slink_of_match, &pdev->dev);
-	if (!match) {
-		dev_err(&pdev->dev, "Error: No device match found\n");
-		return -ENODEV;
-	}
-	cdata = match->data;
+	cdata = of_device_get_match_data(&pdev->dev);
 
 	master = spi_alloc_master(&pdev->dev, sizeof(*tspi));
 	if (!master) {
Signature · 1 line
2.34.1
author maintainer
#791 lore raw

PATCH 783/1017 spi: fsi: Implement a timeout for polling status

In reply to Greg Kroah-Hartman

From: Eddie James <[email protected]>

[ Upstream commit 89b35e3f28514087d3f1e28e8f5634fbfd07c554 ]

The data transfer routines must poll the status register to determine when more data can be shifted in or out. If the hardware gets into a bad state, these polling loops may never exit. Prevent this by returning an error if a timeout is exceeded.

---
 drivers/spi/spi-fsi.c | 10 ++++++++++
 1 file changed, 10 insertions(+)
drivers/spi/spi-fsi.c +10 -0
index b6c7467f0b59..d403a7a3021d 100644
@@ -25,6 +25,7 @@
 
 #define SPI_FSI_BASE			0x70000
 #define SPI_FSI_INIT_TIMEOUT_MS		1000
+#define SPI_FSI_STATUS_TIMEOUT_MS	100
 #define SPI_FSI_MAX_RX_SIZE		8
 #define SPI_FSI_MAX_TX_SIZE		40
 
@@ -299,6 +300,7 @@ static int fsi_spi_transfer_data(struct fsi_spi *ctx,
 				 struct spi_transfer *transfer)
 {
 	int rc = 0;
+	unsigned long end;
 	u64 status = 0ULL;
 
 	if (transfer->tx_buf) {
@@ -315,10 +317,14 @@ static int fsi_spi_transfer_data(struct fsi_spi *ctx,
 			if (rc)
 				return rc;
 
+			end = jiffies + msecs_to_jiffies(SPI_FSI_STATUS_TIMEOUT_MS);
 			do {
 				rc = fsi_spi_status(ctx, &status, "TX");
 				if (rc)
 					return rc;
+
+				if (time_after(jiffies, end))
+					return -ETIMEDOUT;
 			} while (status & SPI_FSI_STATUS_TDR_FULL);
 
 			sent += nb;
@@ -329,10 +335,14 @@ static int fsi_spi_transfer_data(struct fsi_spi *ctx,
 		u8 *rx = transfer->rx_buf;
 
 		while (transfer->len > recv) {
+			end = jiffies + msecs_to_jiffies(SPI_FSI_STATUS_TIMEOUT_MS);
 			do {
 				rc = fsi_spi_status(ctx, &status, "RX");
 				if (rc)
 					return rc;
+
+				if (time_after(jiffies, end))
+					return -ETIMEDOUT;
 			} while (!(status & SPI_FSI_STATUS_RDR_FULL));
 
 			rc = fsi_spi_read_reg(ctx, SPI_FSI_DATA_RX, &in);
Signature · 1 line
2.34.1
author maintainer
#792 lore raw

PATCH 784/1017 atomics: Fix atomic64_{read_acquire,set_release} fallbacks

In reply to Greg Kroah-Hartman

From: Mark Rutland <[email protected]>

[ Upstream commit dc1b4df09acdca7a89806b28f235cd6d8dcd3d24 ]

Arnd reports that on 32-bit architectures, the fallbacks for atomic64_read_acquire() and atomic64_set_release() are broken as they use smp_load_acquire() and smp_store_release() respectively, which do not work on types larger than the native word size.

Since those contain compiletime_assert_atomic_type(), any attempt to use those fallbacks will result in a build-time error. e.g. with the following added to arch/arm/kernel/setup.c:

| void test_atomic64(atomic64_t *v)
| {
|        atomic64_set_release(v, 5);
|        atomic64_read_acquire(v);
| }

The compiler will complain as follows:

| In file included from <command-line>:
| In function 'arch_atomic64_set_release',
|     inlined from 'test_atomic64' at ./include/linux/atomic/atomic-instrumented.h:669:2:
| ././include/linux/compiler_types.h:346:38: error: call to '__compiletime_assert_9' declared with attribute error: Need native word sized stores/loads for atomicity.
|   346 |  _compiletime_assert(condition, msg, __compiletime_assert_, __COUNTER__)
|       |                                      ^
| ././include/linux/compiler_types.h:327:4: note: in definition of macro '__compiletime_assert'
|   327 |    prefix ## suffix();    \
|       |    ^~~~~~
| ././include/linux/compiler_types.h:346:2: note: in expansion of macro '_compiletime_assert'
|   346 |  _compiletime_assert(condition, msg, __compiletime_assert_, __COUNTER__)
|       |  ^~~~~~~~~~~~~~~~~~~
| ././include/linux/compiler_types.h:349:2: note: in expansion of macro 'compiletime_assert'
|   349 |  compiletime_assert(__native_word(t),    \
|       |  ^~~~~~~~~~~~~~~~~~
| ./include/asm-generic/barrier.h:133:2: note: in expansion of macro 'compiletime_assert_atomic_type'
|   133 |  compiletime_assert_atomic_type(*p);    \
|       |  ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
| ./include/asm-generic/barrier.h:164:55: note: in expansion of macro '__smp_store_release'
|   164 | #define smp_store_release(p, v) do { kcsan_release(); __smp_store_release(p, v); } while (0)
|       |                                                       ^~~~~~~~~~~~~~~~~~~
| ./include/linux/atomic/atomic-arch-fallback.h:1270:2: note: in expansion of macro 'smp_store_release'
|  1270 |  smp_store_release(&(v)->counter, i);
|       |  ^~~~~~~~~~~~~~~~~
| make[2]: *** [scripts/Makefile.build:288: arch/arm/kernel/setup.o] Error 1
| make[1]: *** [scripts/Makefile.build:550: arch/arm/kernel] Error 2
| make: *** [Makefile:1831: arch/arm] Error 2

Fix this by only using smp_load_acquire() and smp_store_release() for native atomic types, and otherwise falling back to the regular barriers necessary for acquire/release semantics, as we do in the more generic acquire and release fallbacks.

Since the fallback templates are used to generate the atomic64_*() and atomic_*() operations, the __native_word() check is added to both. For the atomic_*() operations, which are always 32-bit, the __native_word() check is redundant but not harmful, as it is always true.

For the example above this works as expected on 32-bit, e.g. for arm multi_v7_defconfig:

| <test_atomic64>:
|         push    {r4, r5}
|         dmb     ish
|         pldw    [r0]
|         mov     r2, #5
|         mov     r3, #0
|         ldrexd  r4, [r0]
|         strexd  r4, r2, [r0]
|         teq     r4, #0
|         bne     484 <test_atomic64+0x14>
|         ldrexd  r2, [r0]
|         dmb     ish
|         pop     {r4, r5}
|         bx      lr

... and also on 64-bit, e.g. for arm64 defconfig:

| <test_atomic64>:
|         bti     c
|         paciasp
|         mov     x1, #0x5
|         stlr    x1, [x0]
|         ldar    x0, [x0]
|         autiasp
|         ret
---
 include/linux/atomic/atomic-arch-fallback.h | 38 ++++++++++++++++++---
 scripts/atomic/fallbacks/read_acquire       | 11 +++++-
 scripts/atomic/fallbacks/set_release        |  7 +++-
 3 files changed, 49 insertions(+), 7 deletions(-)
include/linux/atomic/atomic-arch-fallback.h +33 -5
index a3dba31df01e..6db58d180866 100644
@@ -151,7 +151,16 @@
 static __always_inline int
 arch_atomic_read_acquire(const atomic_t *v)
 {
-	return smp_load_acquire(&(v)->counter);
+	int ret;
+
+	if (__native_word(atomic_t)) {
+		ret = smp_load_acquire(&(v)->counter);
+	} else {
+		ret = arch_atomic_read(v);
+		__atomic_acquire_fence();
+	}
+
+	return ret;
 }
 #define arch_atomic_read_acquire arch_atomic_read_acquire
 #endif
@@ -160,7 +169,12 @@ arch_atomic_read_acquire(const atomic_t *v)
 static __always_inline void
 arch_atomic_set_release(atomic_t *v, int i)
 {
-	smp_store_release(&(v)->counter, i);
+	if (__native_word(atomic_t)) {
+		smp_store_release(&(v)->counter, i);
+	} else {
+		__atomic_release_fence();
+		arch_atomic_set(v, i);
+	}
 }
 #define arch_atomic_set_release arch_atomic_set_release
 #endif
@@ -1258,7 +1272,16 @@ arch_atomic_dec_if_positive(atomic_t *v)
 static __always_inline s64
 arch_atomic64_read_acquire(const atomic64_t *v)
 {
-	return smp_load_acquire(&(v)->counter);
+	s64 ret;
+
+	if (__native_word(atomic64_t)) {
+		ret = smp_load_acquire(&(v)->counter);
+	} else {
+		ret = arch_atomic64_read(v);
+		__atomic_acquire_fence();
+	}
+
+	return ret;
 }
 #define arch_atomic64_read_acquire arch_atomic64_read_acquire
 #endif
@@ -1267,7 +1290,12 @@ arch_atomic64_read_acquire(const atomic64_t *v)
 static __always_inline void
 arch_atomic64_set_release(atomic64_t *v, s64 i)
 {
-	smp_store_release(&(v)->counter, i);
+	if (__native_word(atomic64_t)) {
+		smp_store_release(&(v)->counter, i);
+	} else {
+		__atomic_release_fence();
+		arch_atomic64_set(v, i);
+	}
 }
 #define arch_atomic64_set_release arch_atomic64_set_release
 #endif
@@ -2358,4 +2386,4 @@ arch_atomic64_dec_if_positive(atomic64_t *v)
 #endif
 
 #endif /* _LINUX_ATOMIC_FALLBACK_H */
-// cca554917d7ea73d5e3e7397dd70c484cad9b2c4
+// 8e2cc06bc0d2c0967d2f8424762bd48555ee40ae
scripts/atomic/fallbacks/read_acquire +10 -1
index 803ba7561076..a0ea1d26e6b2 100755
@@ -2,6 +2,15 @@ cat <<EOF
 static __always_inline ${ret}
 arch_${atomic}_read_acquire(const ${atomic}_t *v)
 {
-	return smp_load_acquire(&(v)->counter);
+	${int} ret;
+
+	if (__native_word(${atomic}_t)) {
+		ret = smp_load_acquire(&(v)->counter);
+	} else {
+		ret = arch_${atomic}_read(v);
+		__atomic_acquire_fence();
+	}
+
+	return ret;
 }
 EOF
scripts/atomic/fallbacks/set_release +6 -1
index 86ede759f24e..05cdb7f42477 100755
@@ -2,6 +2,11 @@ cat <<EOF
 static __always_inline void
 arch_${atomic}_set_release(${atomic}_t *v, ${int} i)
 {
-	smp_store_release(&(v)->counter, i);
+	if (__native_word(${atomic}_t)) {
+		smp_store_release(&(v)->counter, i);
+	} else {
+		__atomic_release_fence();
+		arch_${atomic}_set(v, i);
+	}
 }
 EOF
Signature · 1 line
2.34.1
author maintainer
#793 lore raw

PATCH 785/1017 locking/lockdep: Iterate lock_classes directly when reading lockdep files

In reply to Greg Kroah-Hartman

From: Waiman Long <[email protected]>

[ Upstream commit fb7275acd6fb988313dddd8d3d19efa70d9015ad ]

When dumping lock_classes information via /proc/lockdep, we can't take the lockdep lock as the lock hold time is indeterminate. Iterating over all_lock_classes without holding lock can be dangerous as there is a slight chance that it may branch off to other lists leading to infinite loop or even access invalid memory if changes are made to all_lock_classes list in parallel.

To avoid this problem, iteration of lock classes is now done directly on the lock_classes array itself. The lock_classes_in_use bitmap is checked to see if the lock class is being used. To avoid iterating the full array all the times, a new max_lock_class_idx value is added to track the maximum lock_class index that is currently being used.

We can theoretically take the lockdep lock for iterating all_lock_classes when other lockdep files (lockdep_stats and lock_stat) are accessed as the lock hold time will be shorter for them. For consistency, they are also modified to iterate the lock_classes array directly.

---
 kernel/locking/lockdep.c           | 14 +++++---
 kernel/locking/lockdep_internals.h |  6 ++--
 kernel/locking/lockdep_proc.c      | 51 +++++++++++++++++++++++++-----
 3 files changed, 56 insertions(+), 15 deletions(-)
kernel/locking/lockdep.c +9 -5
index 220913a84494..55570f52de5a 100644
@@ -183,11 +183,9 @@ static DECLARE_BITMAP(list_entries_in_use, MAX_LOCKDEP_ENTRIES);
 static struct hlist_head lock_keys_hash[KEYHASH_SIZE];
 unsigned long nr_lock_classes;
 unsigned long nr_zapped_classes;
-#ifndef CONFIG_DEBUG_LOCKDEP
-static
-#endif
+unsigned long max_lock_class_idx;
 struct lock_class lock_classes[MAX_LOCKDEP_KEYS];
-static DECLARE_BITMAP(lock_classes_in_use, MAX_LOCKDEP_KEYS);
+DECLARE_BITMAP(lock_classes_in_use, MAX_LOCKDEP_KEYS);
 
 static inline struct lock_class *hlock_class(struct held_lock *hlock)
 {
@@ -338,7 +336,7 @@ static inline void lock_release_holdtime(struct held_lock *hlock)
  * elements. These elements are linked together by the lock_entry member in
  * struct lock_class.
  */
-LIST_HEAD(all_lock_classes);
+static LIST_HEAD(all_lock_classes);
 static LIST_HEAD(free_lock_classes);
 
 /**
@@ -1252,6 +1250,7 @@ register_lock_class(struct lockdep_map *lock, unsigned int subclass, int force)
 	struct lockdep_subclass_key *key;
 	struct hlist_head *hash_head;
 	struct lock_class *class;
+	int idx;
 
 	DEBUG_LOCKS_WARN_ON(!irqs_disabled());
 
@@ -1317,6 +1316,9 @@ register_lock_class(struct lockdep_map *lock, unsigned int subclass, int force)
 	 * of classes.
 	 */
 	list_move_tail(&class->lock_entry, &all_lock_classes);
+	idx = class - lock_classes;
+	if (idx > max_lock_class_idx)
+		max_lock_class_idx = idx;
 
 	if (verbose(class)) {
 		graph_unlock();
@@ -5998,6 +6000,8 @@ static void zap_class(struct pending_free *pf, struct lock_class *class)
 		WRITE_ONCE(class->name, NULL);
 		nr_lock_classes--;
 		__clear_bit(class - lock_classes, lock_classes_in_use);
+		if (class - lock_classes == max_lock_class_idx)
+			max_lock_class_idx--;
 	} else {
 		WARN_ONCE(true, "%s() failed for class %s\n", __func__,
 			  class->name);
kernel/locking/lockdep_internals.h +4 -2
index ecb8662e7a4e..bbe9000260d0 100644
@@ -121,7 +121,6 @@ static const unsigned long LOCKF_USED_IN_IRQ_READ =
 
 #define MAX_LOCKDEP_CHAIN_HLOCKS (MAX_LOCKDEP_CHAINS*5)
 
-extern struct list_head all_lock_classes;
 extern struct lock_chain lock_chains[];
 
 #define LOCK_USAGE_CHARS (2*XXX_LOCK_USAGE_STATES + 1)
@@ -151,6 +150,10 @@ extern unsigned int nr_large_chain_blocks;
 
 extern unsigned int max_lockdep_depth;
 extern unsigned int max_bfs_queue_depth;
+extern unsigned long max_lock_class_idx;
+
+extern struct lock_class lock_classes[MAX_LOCKDEP_KEYS];
+extern unsigned long lock_classes_in_use[];
 
 #ifdef CONFIG_PROVE_LOCKING
 extern unsigned long lockdep_count_forward_deps(struct lock_class *);
@@ -205,7 +208,6 @@ struct lockdep_stats {
 };
 
 DECLARE_PER_CPU(struct lockdep_stats, lockdep_stats);
-extern struct lock_class lock_classes[MAX_LOCKDEP_KEYS];
 
 #define __debug_atomic_inc(ptr)					\
 	this_cpu_inc(lockdep_stats.ptr);
kernel/locking/lockdep_proc.c +43 -8

117 lines not on this page · raw message

Signature · 1 line
2.34.1
author maintainer
#794 lore raw

PATCH 786/1017 ext4: correct cluster len and clusters changed accounting in ext4_mb_mark_bb

In reply to Greg Kroah-Hartman

From: Ritesh Harjani <[email protected]>

[ Upstream commit a5c0e2fdf7cea535ba03259894dc184e5a4c2800 ]

ext4_mb_mark_bb() currently wrongly calculates cluster len (clen) and flex_group->free_clusters. This patch fixes that.

Identified based on code review of ext4_mb_mark_bb() function.

---
 fs/ext4/mballoc.c | 19 ++++++++++++-------
 1 file changed, 12 insertions(+), 7 deletions(-)
fs/ext4/mballoc.c +12 -7
index e872ea582555..c0868b82a62f 100644
@@ -3899,10 +3899,11 @@ void ext4_mb_mark_bb(struct super_block *sb, ext4_fsblk_t block,
 	struct ext4_sb_info *sbi = EXT4_SB(sb);
 	ext4_group_t group;
 	ext4_grpblk_t blkoff;
-	int i, clen, err;
+	int i, err;
 	int already;
+	unsigned int clen, clen_changed;
 
-	clen = EXT4_B2C(sbi, len);
+	clen = EXT4_NUM_B2C(sbi, len);
 
 	ext4_get_group_no_and_offset(sb, block, &group, &blkoff);
 	bitmap_bh = ext4_read_block_bitmap(sb, group);
@@ -3923,6 +3924,7 @@ void ext4_mb_mark_bb(struct super_block *sb, ext4_fsblk_t block,
 		if (!mb_test_bit(blkoff + i, bitmap_bh->b_data) == !state)
 			already++;
 
+	clen_changed = clen - already;
 	if (state)
 		ext4_set_bits(bitmap_bh->b_data, blkoff, clen);
 	else
@@ -3935,9 +3937,9 @@ void ext4_mb_mark_bb(struct super_block *sb, ext4_fsblk_t block,
 						group, gdp));
 	}
 	if (state)
-		clen = ext4_free_group_clusters(sb, gdp) - clen + already;
+		clen = ext4_free_group_clusters(sb, gdp) - clen_changed;
 	else
-		clen = ext4_free_group_clusters(sb, gdp) + clen - already;
+		clen = ext4_free_group_clusters(sb, gdp) + clen_changed;
 
 	ext4_free_group_clusters_set(sb, gdp, clen);
 	ext4_block_bitmap_csum_set(sb, group, gdp, bitmap_bh);
@@ -3947,10 +3949,13 @@ void ext4_mb_mark_bb(struct super_block *sb, ext4_fsblk_t block,
 
 	if (sbi->s_log_groups_per_flex) {
 		ext4_group_t flex_group = ext4_flex_group(sbi, group);
+		struct flex_groups *fg = sbi_array_rcu_deref(sbi,
+					   s_flex_groups, flex_group);
 
-		atomic64_sub(len,
-			     &sbi_array_rcu_deref(sbi, s_flex_groups,
-						  flex_group)->free_clusters);
+		if (state)
+			atomic64_sub(clen_changed, &fg->free_clusters);
+		else
+			atomic64_add(clen_changed, &fg->free_clusters);
 	}
 
 	err = ext4_handle_dirty_metadata(NULL, NULL, bitmap_bh);
Signature · 1 line
2.34.1
author maintainer
#795 lore raw

PATCH 787/1017 ext4: fix ext4_mb_mark_bb() with flex_bg with fast_commit

In reply to Greg Kroah-Hartman

From: Ritesh Harjani <[email protected]>

[ Upstream commit bfdc502a4a4c058bf4cbb1df0c297761d528f54d ]

In case of flex_bg feature (which is by default enabled), extents for any given inode might span across blocks from two different block group. ext4_mb_mark_bb() only reads the buffer_head of block bitmap once for the starting block group, but it fails to read it again when the extent length boundary overflows to another block group. Then in this below loop it accesses memory beyond the block group bitmap buffer_head and results into a data abort.

	for (i = 0; i < clen; i++)
		if (!mb_test_bit(blkoff + i, bitmap_bh->b_data) == !state)
			already++;

This patch adds this functionality for checking block group boundary in ext4_mb_mark_bb() and update the buffer_head(bitmap_bh) for every different block group.

w/o this patch, I was easily able to hit a data access abort using Power platform.

<...>

[   74.327662] EXT4-fs error (device loop3): ext4_mb_generate_buddy:1141: group 11, block bitmap and bg descriptor inconsistent: 21248 vs 23294 free clusters
[   74.533214] EXT4-fs (loop3): shut down requested (2)
[   74.536705] Aborting journal on device loop3-8.
[   74.702705] BUG: Unable to handle kernel data access on read at 0xc00000005e980000
[   74.703727] Faulting instruction address: 0xc0000000007bffb8
cpu 0xd: Vector: 300 (Data Access) at [c000000015db7060]
    pc: c0000000007bffb8: ext4_mb_mark_bb+0x198/0x5a0
    lr: c0000000007bfeec: ext4_mb_mark_bb+0xcc/0x5a0
    sp: c000000015db7300
   msr: 800000000280b033
   dar: c00000005e980000
 dsisr: 40000000
  current = 0xc000000027af6880
  paca    = 0xc00000003ffd5200   irqmask: 0x03   irq_happened: 0x01
    pid   = 5167, comm = mount

<...>
enter ? for help
[c000000015db7380] c000000000782708 ext4_ext_clear_bb+0x378/0x410 [c000000015db7400] c000000000813f14 ext4_fc_replay+0x1794/0x2000 [c000000015db7580] c000000000833f7c do_one_pass+0xe9c/0x12a0 [c000000015db7710] c000000000834504 jbd2_journal_recover+0x184/0x2d0 [c000000015db77c0] c000000000841398 jbd2_journal_load+0x188/0x4a0 [c000000015db7880] c000000000804de8 ext4_fill_super+0x2638/0x3e10 [c000000015db7a40] c0000000005f8404 get_tree_bdev+0x2b4/0x350 [c000000015db7ae0] c0000000007ef058 ext4_get_tree+0x28/0x40 [c000000015db7b00] c0000000005f6344 vfs_get_tree+0x44/0x100 [c000000015db7b70] c00000000063c408 path_mount+0xdd8/0xe70 [c000000015db7c40] c00000000063c8f0 sys_mount+0x450/0x550 [c000000015db7d50] c000000000035770 system_call_exception+0x4a0/0x4e0 [c000000015db7e10] c00000000000c74c system_call_common+0xec/0x250

---
 fs/ext4/mballoc.c | 131 +++++++++++++++++++++++++++-------------------
 1 file changed, 76 insertions(+), 55 deletions(-)
fs/ext4/mballoc.c +76 -55
index c0868b82a62f..b53c07c5c760 100644
@@ -3901,72 +3901,93 @@ void ext4_mb_mark_bb(struct super_block *sb, ext4_fsblk_t block,
 	ext4_grpblk_t blkoff;
 	int i, err;
 	int already;
-	unsigned int clen, clen_changed;
+	unsigned int clen, clen_changed, thisgrp_len;
 
-	clen = EXT4_NUM_B2C(sbi, len);
-
-	ext4_get_group_no_and_offset(sb, block, &group, &blkoff);
-	bitmap_bh = ext4_read_block_bitmap(sb, group);
-	if (IS_ERR(bitmap_bh)) {
-		err = PTR_ERR(bitmap_bh);
-		bitmap_bh = NULL;
-		goto out_err;
-	}
-
-	err = -EIO;
-	gdp = ext4_get_group_desc(sb, group, &gdp_bh);
-	if (!gdp)
-		goto out_err;
+	while (len > 0) {
+		ext4_get_group_no_and_offset(sb, block, &group, &blkoff);
 
-	ext4_lock_group(sb, group);
-	already = 0;
-	for (i = 0; i < clen; i++)
-		if (!mb_test_bit(blkoff + i, bitmap_bh->b_data) == !state)
-			already++;
-
-	clen_changed = clen - already;
-	if (state)
-		ext4_set_bits(bitmap_bh->b_data, blkoff, clen);
-	else
-		mb_test_and_clear_bits(bitmap_bh->b_data, blkoff, clen);
-	if (ext4_has_group_desc_csum(sb) &&
-	    (gdp->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT))) {
-		gdp->bg_flags &= cpu_to_le16(~EXT4_BG_BLOCK_UNINIT);
-		ext4_free_group_clusters_set(sb, gdp,
-					     ext4_free_clusters_after_init(sb,
-						group, gdp));
-	}
-	if (state)
-		clen = ext4_free_group_clusters(sb, gdp) - clen_changed;
-	else
-		clen = ext4_free_group_clusters(sb, gdp) + clen_changed;
+		/*
+		 * Check to see if we are freeing blocks across a group
+		 * boundary.
+		 * In case of flex_bg, this can happen that (block, len) may
+		 * span across more than one group. In that case we need to
+		 * get the corresponding group metadata to work with.
+		 * For this we have goto again loop.
+		 */
+		thisgrp_len = min_t(unsigned int, (unsigned int)len,
+			EXT4_BLOCKS_PER_GROUP(sb) - EXT4_C2B(sbi, blkoff));
+		clen = EXT4_NUM_B2C(sbi, thisgrp_len);
 
-	ext4_free_group_clusters_set(sb, gdp, clen);
-	ext4_block_bitmap_csum_set(sb, group, gdp, bitmap_bh);
-	ext4_group_desc_csum_set(sb, group, gdp);
+		bitmap_bh = ext4_read_block_bitmap(sb, group);
+		if (IS_ERR(bitmap_bh)) {
+			err = PTR_ERR(bitmap_bh);
+			bitmap_bh = NULL;
+			break;
+		}
 
-	ext4_unlock_group(sb, group);
+		err = -EIO;
+		gdp = ext4_get_group_desc(sb, group, &gdp_bh);
+		if (!gdp)
+			break;
 
-	if (sbi->s_log_groups_per_flex) {
-		ext4_group_t flex_group = ext4_flex_group(sbi, group);
-		struct flex_groups *fg = sbi_array_rcu_deref(sbi,
-					   s_flex_groups, flex_group);
+		ext4_lock_group(sb, group);
+		already = 0;
+		for (i = 0; i < clen; i++)
+			if (!mb_test_bit(blkoff + i, bitmap_bh->b_data) ==
+					 !state)
+				already++;
 
+		clen_changed = clen - already;
 		if (state)
-			atomic64_sub(clen_changed, &fg->free_clusters);
+			ext4_set_bits(bitmap_bh->b_data, blkoff, clen);
 		else
-			atomic64_add(clen_changed, &fg->free_clusters);
+			mb_test_and_clear_bits(bitmap_bh->b_data, blkoff, clen);
+		if (ext4_has_group_desc_csum(sb) &&
+		    (gdp->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT))) {
+			gdp->bg_flags &= cpu_to_le16(~EXT4_BG_BLOCK_UNINIT);
+			ext4_free_group_clusters_set(sb, gdp,
+			     ext4_free_clusters_after_init(sb, group, gdp));
+		}
+		if (state)
+			clen = ext4_free_group_clusters(sb, gdp) - clen_changed;
+		else
+			clen = ext4_free_group_clusters(sb, gdp) + clen_changed;
+
+		ext4_free_group_clusters_set(sb, gdp, clen);
+		ext4_block_bitmap_csum_set(sb, group, gdp, bitmap_bh);
+		ext4_group_desc_csum_set(sb, group, gdp);
+
+		ext4_unlock_group(sb, group);
+
+		if (sbi->s_log_groups_per_flex) {
+			ext4_group_t flex_group = ext4_flex_group(sbi, group);
+			struct flex_groups *fg = sbi_array_rcu_deref(sbi,
+						   s_flex_groups, flex_group);
+
+			if (state)
+				atomic64_sub(clen_changed, &fg->free_clusters);
+			else
+				atomic64_add(clen_changed, &fg->free_clusters);
+
+		}
+
+		err = ext4_handle_dirty_metadata(NULL, NULL, bitmap_bh);
+		if (err)
+			break;
+		sync_dirty_buffer(bitmap_bh);
+		err = ext4_handle_dirty_metadata(NULL, NULL, gdp_bh);
+		sync_dirty_buffer(gdp_bh);
+		if (err)
+			break;
+
+		block += thisgrp_len;
+		len -= thisgrp_len;
+		brelse(bitmap_bh);
+		BUG_ON(len < 0);
 	}
 
-	err = ext4_handle_dirty_metadata(NULL, NULL, bitmap_bh);
 	if (err)
-		goto out_err;
-	sync_dirty_buffer(bitmap_bh);
-	err = ext4_handle_dirty_metadata(NULL, NULL, gdp_bh);
-	sync_dirty_buffer(gdp_bh);
-
-out_err:
-	brelse(bitmap_bh);
+		brelse(bitmap_bh);
 }
 
 /*
Signature · 1 line
2.34.1
author maintainer
#796 lore raw

PATCH 788/1017 sched/tracing: Dont re-read p->state when emitting sched_switch event

In reply to Greg Kroah-Hartman

From: Valentin Schneider <[email protected]>

[ Upstream commit fa2c3254d7cfff5f7a916ab928a562d1165f17bb ]

As of commit

c6e7bd7afaeb ("sched/core: Optimize ttwu() spinning on p->on_cpu")

the following sequence becomes possible:

		      p->__state = TASK_INTERRUPTIBLE;
		      __schedule()
			deactivate_task(p);
  ttwu()
    READ !p->on_rq
    p->__state=TASK_WAKING
			trace_sched_switch()
			  __trace_sched_switch_state()
			    task_state_index()
			      return 0;

TASK_WAKING isn't in TASK_REPORT, so the task appears as TASK_RUNNING in the trace event.

Prevent this by pushing the value read from __schedule() down the trace event.

---
 include/linux/sched.h             | 11 ++++++++---
 include/trace/events/sched.h      | 11 +++++++----
 kernel/sched/core.c               |  4 ++--
 kernel/trace/fgraph.c             |  4 +++-
 kernel/trace/ftrace.c             |  4 +++-
 kernel/trace/trace_events.c       |  8 ++++++--
 kernel/trace/trace_osnoise.c      |  4 +++-
 kernel/trace/trace_sched_switch.c |  1 +
 kernel/trace/trace_sched_wakeup.c |  1 +
 9 files changed, 34 insertions(+), 14 deletions(-)
include/linux/sched.h +8 -3
index ee5ed8821963..084de9b70a77 100644
@@ -1612,10 +1612,10 @@ static inline pid_t task_pgrp_nr(struct task_struct *tsk)
 #define TASK_REPORT_IDLE	(TASK_REPORT + 1)
 #define TASK_REPORT_MAX		(TASK_REPORT_IDLE << 1)
 
-static inline unsigned int task_state_index(struct task_struct *tsk)
+static inline unsigned int __task_state_index(unsigned int tsk_state,
+					      unsigned int tsk_exit_state)
 {
-	unsigned int tsk_state = READ_ONCE(tsk->__state);
-	unsigned int state = (tsk_state | tsk->exit_state) & TASK_REPORT;
+	unsigned int state = (tsk_state | tsk_exit_state) & TASK_REPORT;
 
 	BUILD_BUG_ON_NOT_POWER_OF_2(TASK_REPORT_MAX);
 
@@ -1625,6 +1625,11 @@ static inline unsigned int task_state_index(struct task_struct *tsk)
 	return fls(state);
 }
 
+static inline unsigned int task_state_index(struct task_struct *tsk)
+{
+	return __task_state_index(READ_ONCE(tsk->__state), tsk->exit_state);
+}
+
 static inline char task_index_to_char(unsigned int state)
 {
 	static const char state_char[] = "RSDTtXZPI";
include/trace/events/sched.h +7 -4
index 94640482cfe7..65e786756321 100644
@@ -187,7 +187,9 @@ DEFINE_EVENT(sched_wakeup_template, sched_wakeup_new,
 	     TP_ARGS(p));
 
 #ifdef CREATE_TRACE_POINTS
-static inline long __trace_sched_switch_state(bool preempt, struct task_struct *p)
+static inline long __trace_sched_switch_state(bool preempt,
+					      unsigned int prev_state,
+					      struct task_struct *p)
 {
 	unsigned int state;
 
@@ -208,7 +210,7 @@ static inline long __trace_sched_switch_state(bool preempt, struct task_struct *
 	 * it for left shift operation to get the correct task->state
 	 * mapping.
 	 */
-	state = task_state_index(p);
+	state = __task_state_index(prev_state, p->exit_state);
 
 	return state ? (1 << (state - 1)) : state;
 }
@@ -220,10 +222,11 @@ static inline long __trace_sched_switch_state(bool preempt, struct task_struct *
 TRACE_EVENT(sched_switch,
 
 	TP_PROTO(bool preempt,
+		 unsigned int prev_state,
 		 struct task_struct *prev,
 		 struct task_struct *next),
 
-	TP_ARGS(preempt, prev, next),
+	TP_ARGS(preempt, prev_state, prev, next),
 
 	TP_STRUCT__entry(
 		__array(	char,	prev_comm,	TASK_COMM_LEN	)
@@ -239,7 +242,7 @@ TRACE_EVENT(sched_switch,
 		memcpy(__entry->next_comm, next->comm, TASK_COMM_LEN);
 		__entry->prev_pid	= prev->pid;
 		__entry->prev_prio	= prev->prio;
-		__entry->prev_state	= __trace_sched_switch_state(preempt, prev);
+		__entry->prev_state	= __trace_sched_switch_state(preempt, prev_state, prev);
 		memcpy(__entry->prev_comm, prev->comm, TASK_COMM_LEN);
 		__entry->next_pid	= next->pid;
 		__entry->next_prio	= next->prio;
kernel/sched/core.c +2 -2
index 0e4b8214af31..527ee363db3f 100644
@@ -4821,7 +4821,7 @@ static struct rq *finish_task_switch(struct task_struct *prev)
 {
 	struct rq *rq = this_rq();
 	struct mm_struct *mm = rq->prev_mm;
-	long prev_state;
+	unsigned int prev_state;
 
 	/*
 	 * The previous task will have left us with a preempt_count of 2
@@ -6257,7 +6257,7 @@ static void __sched notrace __schedule(unsigned int sched_mode)
 		migrate_disable_switch(rq, prev);
 		psi_sched_switch(prev, next, !task_on_rq_queued(prev));
 
-		trace_sched_switch(sched_mode & SM_MASK_PREEMPT, prev, next);
+		trace_sched_switch(sched_mode & SM_MASK_PREEMPT, prev_state, prev, next);
 
 		/* Also unlocks the rq: */
 		rq = context_switch(rq, prev, next, &rf);
kernel/trace/fgraph.c +3 -1
index 22061d38fc00..19028e072cdb 100644
@@ -415,7 +415,9 @@ static int alloc_retstack_tasklist(struct ftrace_ret_stack **ret_stack_list)
 
 static void
 ftrace_graph_probe_sched_switch(void *ignore, bool preempt,
-			struct task_struct *prev, struct task_struct *next)
+				unsigned int prev_state,
+				struct task_struct *prev,
+				struct task_struct *next)
 {
 	unsigned long long timestamp;
 	int index;
kernel/trace/ftrace.c +3 -1
index be5f6b32a012..bbfa4606cba6 100644
@@ -7317,7 +7317,9 @@ ftrace_func_t ftrace_ops_get_func(struct ftrace_ops *ops)
 
 static void
 ftrace_filter_pid_sched_switch_probe(void *data, bool preempt,
-		    struct task_struct *prev, struct task_struct *next)
+				     unsigned int prev_state,
+				     struct task_struct *prev,
+				     struct task_struct *next)
 {
 	struct trace_array *tr = data;
 	struct trace_pid_list *pid_list;
kernel/trace/trace_events.c +6 -2
index 92be9cb1d7d4..c097d89ca946 100644
@@ -759,7 +759,9 @@ void trace_event_follow_fork(struct trace_array *tr, bool enable)
 
 static void
 event_filter_pid_sched_switch_probe_pre(void *data, bool preempt,
-		    struct task_struct *prev, struct task_struct *next)
+					unsigned int prev_state,
+					struct task_struct *prev,
+					struct task_struct *next)
 {
 	struct trace_array *tr = data;
 	struct trace_pid_list *no_pid_list;
@@ -783,7 +785,9 @@ event_filter_pid_sched_switch_probe_pre(void *data, bool preempt,
 
 static void
 event_filter_pid_sched_switch_probe_post(void *data, bool preempt,
-		    struct task_struct *prev, struct task_struct *next)
+					 unsigned int prev_state,
+					 struct task_struct *prev,
+					 struct task_struct *next)
 {
 	struct trace_array *tr = data;
 	struct trace_pid_list *no_pid_list;
kernel/trace/trace_osnoise.c +3 -1
index b6da7bd2383d..b3b8060e8bdd 100644
@@ -1168,7 +1168,9 @@ thread_exit(struct osnoise_variables *osn_var, struct task_struct *t)
  * used to record the beginning and to report the end of a thread noise window.
  */
 static void
-trace_sched_switch_callback(void *data, bool preempt, struct task_struct *p,
+trace_sched_switch_callback(void *data, bool preempt,
+			    unsigned int prev_state,
+			    struct task_struct *p,
 			    struct task_struct *n)
 {
 	struct osnoise_variables *osn_var = this_cpu_osn_var();
kernel/trace/trace_sched_switch.c +1 -0
index e304196d7c28..993b0ed10d8c 100644
@@ -22,6 +22,7 @@ static DEFINE_MUTEX(sched_register_mutex);
 
 static void
 probe_sched_switch(void *ignore, bool preempt,
+		   unsigned int prev_state,
 		   struct task_struct *prev, struct task_struct *next)
 {
 	int flags;
kernel/trace/trace_sched_wakeup.c +1 -0
index 2402de520eca..46429f9a96fa 100644
@@ -426,6 +426,7 @@ tracing_sched_wakeup_trace(struct trace_array *tr,
 
 static void notrace
 probe_wakeup_sched_switch(void *ignore, bool preempt,
+			  unsigned int prev_state,
 			  struct task_struct *prev, struct task_struct *next)
 {
 	struct trace_array_cpu *data;
Signature · 1 line
2.34.1
author maintainer
#797 lore raw

PATCH 789/1017 sched/tracing: Report TASK_RTLOCK_WAIT tasks as TASK_UNINTERRUPTIBLE

In reply to Greg Kroah-Hartman

From: Valentin Schneider <[email protected]>

[ Upstream commit 25795ef6299f07ce3838f3253a9cb34f64efcfae ]

TASK_RTLOCK_WAIT currently isn't part of TASK_REPORT, thus a task blocking on an rtlock will appear as having a task state == 0, IOW TASK_RUNNING.

The actual state is saved in p->saved_state, but reading it after reading p->__state has a few issues:
o that could still be TASK_RUNNING in the case of e.g. rt_spin_lock
o ttwu_state_match() might have changed that to TASK_RUNNING

As pointed out by Eric, adding TASK_RTLOCK_WAIT to TASK_REPORT implies exposing a new state to userspace tools which way not know what to do with them. The only information that needs to be conveyed here is that a task is waiting on an rt_mutex, which matches TASK_UNINTERRUPTIBLE - there's no need for a new state.

---
 include/linux/sched.h | 8 ++++++++
 1 file changed, 8 insertions(+)
include/linux/sched.h +8 -0
index 084de9b70a77..46ef8c75802f 100644
@@ -1622,6 +1622,14 @@ static inline unsigned int __task_state_index(unsigned int tsk_state,
 	if (tsk_state == TASK_IDLE)
 		state = TASK_REPORT_IDLE;
 
+	/*
+	 * We're lying here, but rather than expose a completely new task state
+	 * to userspace, we can make this appear as if the task has gone through
+	 * a regular rt_mutex_lock() call.
+	 */
+	if (tsk_state == TASK_RTLOCK_WAIT)
+		state = TASK_UNINTERRUPTIBLE;
+
 	return fls(state);
 }
 
Signature · 1 line
2.34.1
author maintainer
#798 lore raw

PATCH 790/1017 ext4: dont BUG if someone dirty pages without asking ext4 first

In reply to Greg Kroah-Hartman

From: Theodore Ts'o <[email protected]>

[ Upstream commit cc5095747edfb054ca2068d01af20be3fcc3634f ]

[un]pin_user_pages_remote is dirtying pages without properly warning the file system in advance. A related race was noted by Jan Kara in 2018[1]; however, more recently instead of it being a very hard-to-hit race, it could be reliably triggered by process_vm_writev(2) which was discovered by Syzbot[2].

This is technically a bug in mm/gup.c, but arguably ext4 is fragile in that if some other kernel subsystem dirty pages without properly notifying the file system using page_mkwrite(), ext4 will BUG, while other file systems will not BUG (although data will still be lost).

So instead of crashing with a BUG, issue a warning (since there may be potential data loss) and just mark the page as clean to avoid unprivileged denial of service attacks until the problem can be properly fixed. More discussion and background can be found in the thread starting at [2].

[1] https://lore.kernel.org/linux-mm/[email protected]
[2] https://lore.kernel.org/r/[email protected]

---
 fs/ext4/inode.c | 25 +++++++++++++++++++++++++
 1 file changed, 25 insertions(+)
fs/ext4/inode.c +25 -0
index 2f5686dfa30d..a61d1e4e1026 100644
@@ -1992,6 +1992,15 @@ static int ext4_writepage(struct page *page,
 	else
 		len = PAGE_SIZE;
 
+	/* Should never happen but for bugs in other kernel subsystems */
+	if (!page_has_buffers(page)) {
+		ext4_warning_inode(inode,
+		   "page %lu does not have buffers attached", page->index);
+		ClearPageDirty(page);
+		unlock_page(page);
+		return 0;
+	}
+
 	page_bufs = page_buffers(page);
 	/*
 	 * We cannot do block allocation or other extent handling in this
@@ -2595,6 +2604,22 @@ static int mpage_prepare_extent_to_map(struct mpage_da_data *mpd)
 			wait_on_page_writeback(page);
 			BUG_ON(PageWriteback(page));
 
+			/*
+			 * Should never happen but for buggy code in
+			 * other subsystems that call
+			 * set_page_dirty() without properly warning
+			 * the file system first.  See [1] for more
+			 * information.
+			 *
+			 * [1] https://lore.kernel.org/linux-mm/[email protected]
+			 */
+			if (!page_has_buffers(page)) {
+				ext4_warning_inode(mpd->inode, "page %lu does not have buffers attached", page->index);
+				ClearPageDirty(page);
+				unlock_page(page);
+				continue;
+			}
+
 			if (mpd->map.m_len == 0)
 				mpd->first_page = page->index;
 			mpd->next_page = page->index + 1;
Signature · 1 line
2.34.1
syzbot
#799 lore raw

Re: [PATCH 5.16 0790/1017] ext4: dont BUG if someone dirty pages without asking ext4 first

I see the command but can't find the corresponding bug. The email is sent to [email protected] address but the HASH does not correspond to any known bug. Please double check the…
Quoted 81 lines

From: Theodore Ts'o <[email protected]>

[ Upstream commit cc5095747edfb054ca2068d01af20be3fcc3634f ]

[un]pin_user_pages_remote is dirtying pages without properly warning the file system in advance. A related race was noted by Jan Kara in 2018[1]; however, more recently instead of it being a very hard-to-hit race, it could be reliably triggered by process_vm_writev(2) which was discovered by Syzbot[2].

This is technically a bug in mm/gup.c, but arguably ext4 is fragile in that if some other kernel subsystem dirty pages without properly notifying the file system using page_mkwrite(), ext4 will BUG, while other file systems will not BUG (although data will still be lost).

So instead of crashing with a BUG, issue a warning (since there may be potential data loss) and just mark the page as clean to avoid unprivileged denial of service attacks until the problem can be properly fixed. More discussion and background can be found in the thread starting at [2].

[1] https://lore.kernel.org/linux-mm/[email protected]
[2] https://lore.kernel.org/r/[email protected]

---
 fs/ext4/inode.c | 25 +++++++++++++++++++++++++
 1 file changed, 25 insertions(+)
fs/ext4/inode.c +25 -0
index 2f5686dfa30d..a61d1e4e1026 100644
@@ -1992,6 +1992,15 @@ static int ext4_writepage(struct page *page,
 	else
 		len = PAGE_SIZE;
 
+	/* Should never happen but for bugs in other kernel subsystems */
+	if (!page_has_buffers(page)) {
+		ext4_warning_inode(inode,
+		   "page %lu does not have buffers attached", page->index);
+		ClearPageDirty(page);
+		unlock_page(page);
+		return 0;
+	}
+
 	page_bufs = page_buffers(page);
 	/*
 	 * We cannot do block allocation or other extent handling in this
@@ -2595,6 +2604,22 @@ static int mpage_prepare_extent_to_map(struct mpage_da_data *mpd)
 			wait_on_page_writeback(page);
 			BUG_ON(PageWriteback(page));
 
+			/*
+			 * Should never happen but for buggy code in
+			 * other subsystems that call
+			 * set_page_dirty() without properly warning
+			 * the file system first.  See [1] for more
+			 * information.
+			 *
+			 * [1] https://lore.kernel.org/linux-mm/[email protected]
+			 */
+			if (!page_has_buffers(page)) {
+				ext4_warning_inode(mpd->inode, "page %lu does not have buffers attached", page->index);
+				ClearPageDirty(page);
+				unlock_page(page);
+				continue;
+			}
+
 			if (mpd->map.m_len == 0)
 				mpd->first_page = page->index;
 			mpd->next_page = page->index + 1;
Signature · 1 line
2.34.1

I see the command but can't find the corresponding bug. The email is sent to [email protected] address but the HASH does not correspond to any known bug. Please double check the address.

author maintainer
#800 lore raw

PATCH 791/1017 f2fs: fix to do sanity check on curseg->alloc_type

In reply to Greg Kroah-Hartman

From: Chao Yu <[email protected]>

[ Upstream commit f41ee8b91c00770d718be2ff4852a80017ae9ab3 ]

As Wenqing Liu reported in bugzilla:

https://bugzilla.kernel.org/show_bug.cgi?id=215657

- Overview

UBSAN: array-index-out-of-bounds in fs/f2fs/segment.c:3460:2 when mount and operate a corrupted image

- Reproduce
tested on kernel 5.17-rc4, 5.17-rc6

1. mkdir test_crash
2. cd test_crash
3. unzip tmp2.zip
4. mkdir mnt
5. ./single_test.sh f2fs 2

- Kernel dump

[   46.434454] loop0: detected capacity change from 0 to 131072
[   46.529839] F2FS-fs (loop0): Mounted with checkpoint version = 7548c2d9
[   46.738319] ================================================================================
[   46.738412] UBSAN: array-index-out-of-bounds in fs/f2fs/segment.c:3460:2
[   46.738475] index 231 is out of range for type 'unsigned int [2]'
[   46.738539] CPU: 2 PID: 939 Comm: umount Not tainted 5.17.0-rc6 #1
[   46.738547] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.13.0-1ubuntu1.1 04/01/2014
[   46.738551] Call Trace:
[   46.738556]  <TASK>
[   46.738563]  dump_stack_lvl+0x47/0x5c
[   46.738581]  ubsan_epilogue+0x5/0x50
[   46.738592]  __ubsan_handle_out_of_bounds+0x68/0x80
[   46.738604]  f2fs_allocate_data_block+0xdff/0xe60 [f2fs]
[   46.738819]  do_write_page+0xef/0x210 [f2fs]
[   46.738934]  f2fs_do_write_node_page+0x3f/0x80 [f2fs]
[   46.739038]  __write_node_page+0x2b7/0x920 [f2fs]
[   46.739162]  f2fs_sync_node_pages+0x943/0xb00 [f2fs]
[   46.739293]  f2fs_write_checkpoint+0x7bb/0x1030 [f2fs]
[   46.739405]  kill_f2fs_super+0x125/0x150 [f2fs]
[   46.739507]  deactivate_locked_super+0x60/0xc0
[   46.739517]  deactivate_super+0x70/0xb0
[   46.739524]  cleanup_mnt+0x11a/0x200
[   46.739532]  __cleanup_mnt+0x16/0x20
[   46.739538]  task_work_run+0x67/0xa0
[   46.739547]  exit_to_user_mode_prepare+0x18c/0x1a0
[   46.739559]  syscall_exit_to_user_mode+0x26/0x40
[   46.739568]  do_syscall_64+0x46/0xb0
[   46.739584]  entry_SYSCALL_64_after_hwframe+0x44/0xae

The root cause is we missed to do sanity check on curseg->alloc_type, result in out-of-bound accessing on sbi->block_count[] array, fix it.

---
 fs/f2fs/segment.c | 7 +++++++
 1 file changed, 7 insertions(+)
fs/f2fs/segment.c +7 -0
index df9ed75f0b7a..5fd33aabd0fc 100644
@@ -4792,6 +4792,13 @@ static int sanity_check_curseg(struct f2fs_sb_info *sbi)
 
 		sanity_check_seg_type(sbi, curseg->seg_type);
 
+		if (curseg->alloc_type != LFS && curseg->alloc_type != SSR) {
+			f2fs_err(sbi,
+				 "Current segment has invalid alloc_type:%d",
+				 curseg->alloc_type);
+			return -EFSCORRUPTED;
+		}
+
 		if (f2fs_test_bit(blkofs, se->cur_valid_map))
 			goto out;
 
Signature · 1 line
2.34.1
author maintainer
#801 lore raw

PATCH 792/1017 NFSD: Fix nfsd_breaker_owns_lease() return values

In reply to Greg Kroah-Hartman

From: Chuck Lever <[email protected]>

[ Upstream commit 50719bf3442dd6cd05159e9c98d020b3919ce978 ]

These have been incorrect since the function was introduced.

A proper kerneldoc comment is added since this function, though static, is part of an external interface.

---
 fs/nfsd/nfs4state.c | 12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)
fs/nfsd/nfs4state.c +10 -2
index ac0ddde8beef..6075a328d756 100644
@@ -4693,6 +4693,14 @@ nfsd_break_deleg_cb(struct file_lock *fl)
 	return ret;
 }
 
+/**
+ * nfsd_breaker_owns_lease - Check if lease conflict was resolved
+ * @fl: Lock state to check
+ *
+ * Return values:
+ *   %true: Lease conflict was resolved
+ *   %false: Lease conflict was not resolved.
+ */
 static bool nfsd_breaker_owns_lease(struct file_lock *fl)
 {
 	struct nfs4_delegation *dl = fl->fl_owner;
@@ -4700,11 +4708,11 @@ static bool nfsd_breaker_owns_lease(struct file_lock *fl)
 	struct nfs4_client *clp;
 
 	if (!i_am_nfsd())
-		return NULL;
+		return false;
 	rqst = kthread_data(current);
 	/* Note rq_prog == NFS_ACL_PROGRAM is also possible: */
 	if (rqst->rq_prog != NFS_PROGRAM || rqst->rq_vers < 4)
-		return NULL;
+		return false;
 	clp = *(rqst->rq_lease_breaker);
 	return dl->dl_stid.sc_client == clp;
 }
Signature · 1 line
2.34.1
author maintainer
#802 lore raw

PATCH 793/1017 f2fs: dont get FREEZE lock in f2fs_evict_inode in frozen fs

In reply to Greg Kroah-Hartman

From: Jaegeuk Kim <[email protected]>

[ Upstream commit ba900534f807f0b327c92d5141c85d2313e2d55c ]

Let's purge inode cache in order to avoid the below deadlock.

[freeze test] shrinkder
freeze_super
- pwercpu_down_write(SB_FREEZE_FS)
- super_cache_scan
- down_read(&sb->s_umount)
- prune_icache_sb
- dispose_list
- evict
- f2fs_evict_inode
thaw_super
- down_write(&sb->s_umount);
- __percpu_down_read(SB_FREEZE_FS)

---
 Documentation/ABI/testing/sysfs-fs-f2fs | 1 +
 fs/f2fs/debug.c                         | 1 +
 fs/f2fs/f2fs.h                          | 1 +
 fs/f2fs/inode.c                         | 6 ++++--
 fs/f2fs/super.c                         | 4 ++++
 5 files changed, 11 insertions(+), 2 deletions(-)
Documentation/ABI/testing/sysfs-fs-f2fs +1 -0
index b268e3e18b4a..91e2b549f817 100644
@@ -425,6 +425,7 @@ Description:	Show status of f2fs superblock in real time.
 		0x800  SBI_QUOTA_SKIP_FLUSH  skip flushing quota in current CP
 		0x1000 SBI_QUOTA_NEED_REPAIR quota file may be corrupted
 		0x2000 SBI_IS_RESIZEFS       resizefs is in process
+		0x4000 SBI_IS_FREEZING       freefs is in process
 		====== ===================== =================================
 
 What:		/sys/fs/f2fs/<disk>/ckpt_thread_ioprio
fs/f2fs/debug.c +1 -0
index 8c50518475a9..07ad0d81f0c5 100644
@@ -338,6 +338,7 @@ static char *s_flag[] = {
 	[SBI_QUOTA_SKIP_FLUSH]	= " quota_skip_flush",
 	[SBI_QUOTA_NEED_REPAIR]	= " quota_need_repair",
 	[SBI_IS_RESIZEFS]	= " resizefs",
+	[SBI_IS_FREEZING]	= " freezefs",
 };
 
 static int stat_show(struct seq_file *s, void *v)
fs/f2fs/f2fs.h +1 -0
index d753094a4919..980c0c2b6350 100644
@@ -1259,6 +1259,7 @@ enum {
 	SBI_QUOTA_SKIP_FLUSH,			/* skip flushing quota in current CP */
 	SBI_QUOTA_NEED_REPAIR,			/* quota file may be corrupted */
 	SBI_IS_RESIZEFS,			/* resizefs is in process */
+	SBI_IS_FREEZING,			/* freezefs is in process */
 };
 
 enum {
fs/f2fs/inode.c +4 -2
index 5a10f98dbbc3..a40e52ba5ec8 100644
@@ -769,7 +769,8 @@ void f2fs_evict_inode(struct inode *inode)
 	f2fs_remove_ino_entry(sbi, inode->i_ino, UPDATE_INO);
 	f2fs_remove_ino_entry(sbi, inode->i_ino, FLUSH_INO);
 
-	sb_start_intwrite(inode->i_sb);
+	if (!is_sbi_flag_set(sbi, SBI_IS_FREEZING))
+		sb_start_intwrite(inode->i_sb);
 	set_inode_flag(inode, FI_NO_ALLOC);
 	i_size_write(inode, 0);
 retry:
@@ -800,7 +801,8 @@ void f2fs_evict_inode(struct inode *inode)
 		if (dquot_initialize_needed(inode))
 			set_sbi_flag(sbi, SBI_QUOTA_NEED_REPAIR);
 	}
-	sb_end_intwrite(inode->i_sb);
+	if (!is_sbi_flag_set(sbi, SBI_IS_FREEZING))
+		sb_end_intwrite(inode->i_sb);
 no_delete:
 	dquot_drop(inode);
 
fs/f2fs/super.c +4 -0
index 45106b426e8d..9402d5b2c7fd 100644
@@ -1668,11 +1668,15 @@ static int f2fs_freeze(struct super_block *sb)
 	/* ensure no checkpoint required */
 	if (!llist_empty(&F2FS_SB(sb)->cprc_info.issue_list))
 		return -EINVAL;
+
+	/* to avoid deadlock on f2fs_evict_inode->SB_FREEZE_FS */
+	set_sbi_flag(F2FS_SB(sb), SBI_IS_FREEZING);
 	return 0;
 }
 
 static int f2fs_unfreeze(struct super_block *sb)
 {
+	clear_sbi_flag(F2FS_SB(sb), SBI_IS_FREEZING);
 	return 0;
 }
 
Signature · 1 line
2.34.1
author maintainer
#803 lore raw

PATCH 794/1017 btrfs: harden identification of a stale device

In reply to Greg Kroah-Hartman

From: Anand Jain <[email protected]>

[ Upstream commit 770c79fb65506fc7c16459855c3839429f46cb32 ]

Identifying and removing the stale device from the fs_uuids list is done by btrfs_free_stale_devices(). btrfs_free_stale_devices() in turn depends on device_path_matched() to check if the device appears in more than one btrfs_device structure.

The matching of the device happens by its path, the device path. However, when device mapper is in use, the dm device paths are nothing but a link to the actual block device, which leads to the device_path_matched() failing to match.

Fix this by matching the dev_t as provided by lookup_bdev() instead of plain string compare of the device paths.

---
 fs/btrfs/volumes.c | 45 ++++++++++++++++++++++++++++++++++++++-------
 1 file changed, 38 insertions(+), 7 deletions(-)
fs/btrfs/volumes.c +38 -7
index 42391d4aeb11..02ee42d461be 100644
@@ -530,15 +530,48 @@ btrfs_get_bdev_and_sb(const char *device_path, fmode_t flags, void *holder,
 	return ret;
 }
 
-static bool device_path_matched(const char *path, struct btrfs_device *device)
+/*
+ * Check if the device in the path matches the device in the given struct device.
+ *
+ * Returns:
+ *   true  If it is the same device.
+ *   false If it is not the same device or on error.
+ */
+static bool device_matched(const struct btrfs_device *device, const char *path)
 {
-	int found;
+	char *device_name;
+	dev_t dev_old;
+	dev_t dev_new;
+	int ret;
+
+	/*
+	 * If we are looking for a device with the matching dev_t, then skip
+	 * device without a name (a missing device).
+	 */
+	if (!device->name)
+		return false;
+
+	device_name = kzalloc(BTRFS_PATH_NAME_MAX, GFP_KERNEL);
+	if (!device_name)
+		return false;
 
 	rcu_read_lock();
-	found = strcmp(rcu_str_deref(device->name), path);
+	scnprintf(device_name, BTRFS_PATH_NAME_MAX, "%s", rcu_str_deref(device->name));
 	rcu_read_unlock();
 
-	return found == 0;
+	ret = lookup_bdev(device_name, &dev_old);
+	kfree(device_name);
+	if (ret)
+		return false;
+
+	ret = lookup_bdev(path, &dev_new);
+	if (ret)
+		return false;
+
+	if (dev_old == dev_new)
+		return true;
+
+	return false;
 }
 
 /*
@@ -571,9 +604,7 @@ static int btrfs_free_stale_devices(const char *path,
 					 &fs_devices->devices, dev_list) {
 			if (skip_device && skip_device == device)
 				continue;
-			if (path && !device->name)
-				continue;
-			if (path && !device_path_matched(path, device))
+			if (path && !device_matched(device, path))
 				continue;
 			if (fs_devices->opened) {
 				/* for an already deleted device return 0 */
Signature · 1 line
2.34.1
author maintainer
#804 lore raw

PATCH 795/1017 btrfs: make search_csum_tree return 0 if we get -EFBIG

In reply to Greg Kroah-Hartman

From: Josef Bacik <[email protected]>

[ Upstream commit 03ddb19d2ea745228879b9334f3b550c88acb10a ]

We can either fail to find a csum entry at all and return -ENOENT, or we can find a range that is close, but return -EFBIG. In essence these both mean the same thing when we are doing a lookup for a csum in an existing range, we didn't find a csum. We want to treat both of these errors the same way, complain loudly that there wasn't a csum. This currently happens anyway because we do

	count = search_csum_tree();
	if (count <= 0) {
		// reloc and error handling
	}

However it forces us to incorrectly treat EIO or ENOMEM errors as on disk corruption. Fix this by returning 0 if we get either -ENOENT or -EFBIG from btrfs_lookup_csum() so we can do proper error handling.

---
 fs/btrfs/file-item.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
fs/btrfs/file-item.c +1 -1
index d1cbb64a78f3..91ae1caa1bdb 100644
@@ -303,7 +303,7 @@ static int search_csum_tree(struct btrfs_fs_info *fs_info,
 	read_extent_buffer(path->nodes[0], dst, (unsigned long)item,
 			ret * csum_size);
 out:
-	if (ret == -ENOENT)
+	if (ret == -ENOENT || ret == -EFBIG)
 		ret = 0;
 	return ret;
 }
Signature · 1 line
2.34.1
author maintainer
#805 lore raw

PATCH 796/1017 btrfs: handle csum lookup errors properly on reads

In reply to Greg Kroah-Hartman

From: Josef Bacik <[email protected]>

[ Upstream commit 1784b7d502a94b561eae58249adde5f72c26eb3c ]

Currently any error we get while trying to lookup csums during reads shows up as a missing csum, and then on the read completion side we print an error saying there was a csum mismatch and we increase the device corruption count.

However we could have gotten an EIO from the lookup. We could also be inside of a memory constrained container and gotten a ENOMEM while trying to do the read. In either case we don't want to make this look like a file system corruption problem, we want to make it look like the actual error it is. Capture any negative value, convert it to the appropriate blk_status_t, free the csum array if we have one and bail.

Note: a possible improvement would be to make the relocation code look up the owning inode and see if it's marked as NODATASUM and set EXTENT_NODATASUM there, that way if there's corruption and there isn't a checksum when we want it we can fail here rather than later.

---
 fs/btrfs/file-item.c | 36 ++++++++++++++++++++++--------------
 1 file changed, 22 insertions(+), 14 deletions(-)
fs/btrfs/file-item.c +22 -14
index 91ae1caa1bdb..a68d9cab12ea 100644
@@ -366,6 +366,7 @@ blk_status_t btrfs_lookup_bio_sums(struct inode *inode, struct bio *bio, u8 *dst
 {
 	struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
 	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
+	struct btrfs_bio *bbio = NULL;
 	struct btrfs_path *path;
 	const u32 sectorsize = fs_info->sectorsize;
 	const u32 csum_size = fs_info->csum_size;
@@ -375,6 +376,7 @@ blk_status_t btrfs_lookup_bio_sums(struct inode *inode, struct bio *bio, u8 *dst
 	u8 *csum;
 	const unsigned int nblocks = orig_len >> fs_info->sectorsize_bits;
 	int count = 0;
+	blk_status_t ret = BLK_STS_OK;
 
 	if (!fs_info->csum_root || (BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM))
 		return BLK_STS_OK;
@@ -397,7 +399,7 @@ blk_status_t btrfs_lookup_bio_sums(struct inode *inode, struct bio *bio, u8 *dst
 		return BLK_STS_RESOURCE;
 
 	if (!dst) {
-		struct btrfs_bio *bbio = btrfs_bio(bio);
+		bbio = btrfs_bio(bio);
 
 		if (nblocks * csum_size > BTRFS_BIO_INLINE_CSUM_SIZE) {
 			bbio->csum = kmalloc_array(nblocks, csum_size, GFP_NOFS);
@@ -453,21 +455,27 @@ blk_status_t btrfs_lookup_bio_sums(struct inode *inode, struct bio *bio, u8 *dst
 
 		count = search_csum_tree(fs_info, path, cur_disk_bytenr,
 					 search_len, csum_dst);
-		if (count <= 0) {
-			/*
-			 * Either we hit a critical error or we didn't find
-			 * the csum.
-			 * Either way, we put zero into the csums dst, and skip
-			 * to the next sector.
-			 */
+		if (count < 0) {
+			ret = errno_to_blk_status(count);
+			if (bbio)
+				btrfs_bio_free_csum(bbio);
+			break;
+		}
+
+		/*
+		 * We didn't find a csum for this range.  We need to make sure
+		 * we complain loudly about this, because we are not NODATASUM.
+		 *
+		 * However for the DATA_RELOC inode we could potentially be
+		 * relocating data extents for a NODATASUM inode, so the inode
+		 * itself won't be marked with NODATASUM, but the extent we're
+		 * copying is in fact NODATASUM.  If we don't find a csum we
+		 * assume this is the case.
+		 */
+		if (count == 0) {
 			memset(csum_dst, 0, csum_size);
 			count = 1;
 
-			/*
-			 * For data reloc inode, we need to mark the range
-			 * NODATASUM so that balance won't report false csum
-			 * error.
-			 */
 			if (BTRFS_I(inode)->root->root_key.objectid ==
 			    BTRFS_DATA_RELOC_TREE_OBJECTID) {
 				u64 file_offset;
@@ -488,7 +496,7 @@ blk_status_t btrfs_lookup_bio_sums(struct inode *inode, struct bio *bio, u8 *dst
 	}
 
 	btrfs_free_path(path);
-	return BLK_STS_OK;
+	return ret;
 }
 
 int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end,
Signature · 1 line
2.34.1
author maintainer
#806 lore raw

PATCH 797/1017 btrfs: do not double complete bio on errors during compressed reads

In reply to Greg Kroah-Hartman

From: Josef Bacik <[email protected]>

[ Upstream commit f9f15de85d74e7eef021af059ca53a15f041cdd8 ]

I hit some weird panics while fixing up the error handling from btrfs_lookup_bio_sums(). Turns out the compression path will complete the bio we use if we set up any of the compression bios and then return an error, and then btrfs_submit_data_bio() will also call bio_endio() on the bio.

Fix this by making btrfs_submit_compressed_read() responsible for calling bio_endio() on the bio if there are any errors. Currently it was only doing it if we created the compression bios, otherwise it was depending on btrfs_submit_data_bio() to do the right thing. This creates the above problem, so fix up btrfs_submit_compressed_read() to always call bio_endio() in case of an error, and then simply return from btrfs_submit_data_bio() if we had to call btrfs_submit_compressed_read().

---
 fs/btrfs/compression.c | 20 ++++++++++++++------
 fs/btrfs/inode.c       |  8 +++++++-
 2 files changed, 21 insertions(+), 7 deletions(-)
fs/btrfs/compression.c +14 -6
index 32da97c3c19d..66d6a414b2ca 100644
@@ -807,7 +807,7 @@ blk_status_t btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
 	u64 em_len;
 	u64 em_start;
 	struct extent_map *em;
-	blk_status_t ret = BLK_STS_RESOURCE;
+	blk_status_t ret;
 	int faili = 0;
 	u8 *sums;
 
@@ -820,14 +820,18 @@ blk_status_t btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
 	read_lock(&em_tree->lock);
 	em = lookup_extent_mapping(em_tree, file_offset, fs_info->sectorsize);
 	read_unlock(&em_tree->lock);
-	if (!em)
-		return BLK_STS_IOERR;
+	if (!em) {
+		ret = BLK_STS_IOERR;
+		goto out;
+	}
 
 	ASSERT(em->compress_type != BTRFS_COMPRESS_NONE);
 	compressed_len = em->block_len;
 	cb = kmalloc(compressed_bio_size(fs_info, compressed_len), GFP_NOFS);
-	if (!cb)
+	if (!cb) {
+		ret = BLK_STS_RESOURCE;
 		goto out;
+	}
 
 	refcount_set(&cb->pending_sectors, compressed_len >> fs_info->sectorsize_bits);
 	cb->errors = 0;
@@ -850,8 +854,10 @@ blk_status_t btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
 	nr_pages = DIV_ROUND_UP(compressed_len, PAGE_SIZE);
 	cb->compressed_pages = kcalloc(nr_pages, sizeof(struct page *),
 				       GFP_NOFS);
-	if (!cb->compressed_pages)
+	if (!cb->compressed_pages) {
+		ret = BLK_STS_RESOURCE;
 		goto fail1;
+	}
 
 	for (pg_index = 0; pg_index < nr_pages; pg_index++) {
 		cb->compressed_pages[pg_index] = alloc_page(GFP_NOFS);
@@ -937,7 +943,7 @@ blk_status_t btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
 			comp_bio = NULL;
 		}
 	}
-	return 0;
+	return BLK_STS_OK;
 
 fail2:
 	while (faili >= 0) {
@@ -950,6 +956,8 @@ blk_status_t btrfs_submit_compressed_read(struct inode *inode, struct bio *bio,
 	kfree(cb);
 out:
 	free_extent_map(em);
+	bio->bi_status = ret;
+	bio_endio(bio);
 	return ret;
 finish_cb:
 	if (comp_bio) {
fs/btrfs/inode.c +7 -1
index cc957cce23a1..68f5a94c82b7 100644
@@ -2536,10 +2536,15 @@ blk_status_t btrfs_submit_data_bio(struct inode *inode, struct bio *bio,
 			goto out;
 
 		if (bio_flags & EXTENT_BIO_COMPRESSED) {
+			/*
+			 * btrfs_submit_compressed_read will handle completing
+			 * the bio if there were any errors, so just return
+			 * here.
+			 */
 			ret = btrfs_submit_compressed_read(inode, bio,
 							   mirror_num,
 							   bio_flags);
-			goto out;
+			goto out_no_endio;
 		} else {
 			/*
 			 * Lookup bio sums does extra checks around whether we
@@ -2573,6 +2578,7 @@ blk_status_t btrfs_submit_data_bio(struct inode *inode, struct bio *bio,
 		bio->bi_status = ret;
 		bio_endio(bio);
 	}
+out_no_endio:
 	return ret;
 }
 
Signature · 1 line
2.34.1
author maintainer
#807 lore raw

PATCH 798/1017 btrfs: do not clean up repair bio if submit fails

In reply to Greg Kroah-Hartman

From: Josef Bacik <[email protected]>

[ Upstream commit 8cbc3001a3264d998d6b6db3e23f935c158abd4d ]

The submit helper will always run bio_endio() on the bio if it fails to submit, so cleaning up the bio just leads to a variety of use-after-free and NULL pointer dereference bugs because we race with the endio function that is cleaning up the bio. Instead just return BLK_STS_OK as the repair function has to continue to process the rest of the pages, and the endio for the repair bio will do the appropriate cleanup for the page that it was given.

---
 fs/btrfs/extent_io.c | 15 +++++++--------
 1 file changed, 7 insertions(+), 8 deletions(-)
fs/btrfs/extent_io.c +7 -8
index c3f466362103..ced0195f3390 100644
@@ -2640,7 +2640,6 @@ int btrfs_repair_one_sector(struct inode *inode,
 	const int icsum = bio_offset >> fs_info->sectorsize_bits;
 	struct bio *repair_bio;
 	struct btrfs_bio *repair_bbio;
-	blk_status_t status;
 
 	btrfs_debug(fs_info,
 		   "repair read error: read error at %llu", start);
@@ -2679,13 +2678,13 @@ int btrfs_repair_one_sector(struct inode *inode,
 		    "repair read error: submitting new read to mirror %d",
 		    failrec->this_mirror);
 
-	status = submit_bio_hook(inode, repair_bio, failrec->this_mirror,
-				 failrec->bio_flags);
-	if (status) {
-		free_io_failure(failure_tree, tree, failrec);
-		bio_put(repair_bio);
-	}
-	return blk_status_to_errno(status);
+	/*
+	 * At this point we have a bio, so any errors from submit_bio_hook()
+	 * will be handled by the endio on the repair_bio, so we can't return an
+	 * error here.
+	 */
+	submit_bio_hook(inode, repair_bio, failrec->this_mirror, failrec->bio_flags);
+	return BLK_STS_OK;
 }
 
 static void end_page_read(struct page *page, bool uptodate, u64 start, u32 len)
Signature · 1 line
2.34.1
author maintainer
#808 lore raw

PATCH 799/1017 f2fs: use spin_lock to avoid hang

In reply to Greg Kroah-Hartman

From: Jaegeuk Kim <[email protected]>

[ Upstream commit 98237fcda4a24e67b0a4498c17d5aa4ad4537bc7 ]

[14696.634553] task:cat             state:D stack:    0 pid:1613738 ppid:1613735 flags:0x00000004
[14696.638285] Call Trace:
[14696.639038]  <TASK>
[14696.640032]  __schedule+0x302/0x930
[14696.640969]  schedule+0x58/0xd0
[14696.641799]  schedule_preempt_disabled+0x18/0x30
[14696.642890]  __mutex_lock.constprop.0+0x2fb/0x4f0
[14696.644035]  ? mod_objcg_state+0x10c/0x310
[14696.645040]  ? obj_cgroup_charge+0xe1/0x170
[14696.646067]  __mutex_lock_slowpath+0x13/0x20
[14696.647126]  mutex_lock+0x34/0x40
[14696.648070]  stat_show+0x25/0x17c0 [f2fs]
[14696.649218]  seq_read_iter+0x120/0x4b0
[14696.650289]  ? aa_file_perm+0x12a/0x500
[14696.651357]  ? lru_cache_add+0x1c/0x20
[14696.652470]  seq_read+0xfd/0x140
[14696.653445]  full_proxy_read+0x5c/0x80
[14696.654535]  vfs_read+0xa0/0x1a0
[14696.655497]  ksys_read+0x67/0xe0
[14696.656502]  __x64_sys_read+0x1a/0x20
[14696.657580]  do_syscall_64+0x3b/0xc0
[14696.658671]  entry_SYSCALL_64_after_hwframe+0x44/0xae
[14696.660068] RIP: 0033:0x7efe39df1cb2
[14696.661133] RSP: 002b:00007ffc8badd948 EFLAGS: 00000246 ORIG_RAX: 0000000000000000
[14696.662958] RAX: ffffffffffffffda RBX: 0000000000020000 RCX: 00007efe39df1cb2
[14696.664757] RDX: 0000000000020000 RSI: 00007efe399df000 RDI: 0000000000000003
[14696.666542] RBP: 00007efe399df000 R08: 00007efe399de010 R09: 00007efe399de010
[14696.668363] R10: 0000000000000022 R11: 0000000000000246 R12: 0000000000000000
[14696.670155] R13: 0000000000000003 R14: 0000000000020000 R15: 0000000000020000
[14696.671965]  </TASK>
[14696.672826] task:umount          state:D stack:    0 pid:1614985 ppid:1614984 flags:0x00004000
[14696.674930] Call Trace:
[14696.675903]  <TASK>
[14696.676780]  __schedule+0x302/0x930
[14696.677927]  schedule+0x58/0xd0
[14696.679019]  schedule_preempt_disabled+0x18/0x30
[14696.680412]  __mutex_lock.constprop.0+0x2fb/0x4f0
[14696.681783]  ? destroy_inode+0x65/0x80
[14696.683006]  __mutex_lock_slowpath+0x13/0x20
[14696.684305]  mutex_lock+0x34/0x40
[14696.685442]  f2fs_destroy_stats+0x1e/0x60 [f2fs]
[14696.686803]  f2fs_put_super+0x158/0x390 [f2fs]
[14696.688238]  generic_shutdown_super+0x7a/0x120
[14696.689621]  kill_block_super+0x27/0x50
[14696.690894]  kill_f2fs_super+0x7f/0x100 [f2fs]
[14696.692311]  deactivate_locked_super+0x35/0xa0
[14696.693698]  deactivate_super+0x40/0x50
[14696.694985]  cleanup_mnt+0x139/0x190
[14696.696209]  __cleanup_mnt+0x12/0x20
[14696.697390]  task_work_run+0x64/0xa0
[14696.698587]  exit_to_user_mode_prepare+0x1b7/0x1c0
[14696.700053]  syscall_exit_to_user_mode+0x27/0x50
[14696.701418]  do_syscall_64+0x48/0xc0
[14696.702630]  entry_SYSCALL_64_after_hwframe+0x44/0xae
---
 fs/f2fs/debug.c | 17 ++++++++++-------
 1 file changed, 10 insertions(+), 7 deletions(-)
fs/f2fs/debug.c +10 -7
index 07ad0d81f0c5..b449c7a372a4 100644
@@ -21,7 +21,7 @@
 #include "gc.h"
 
 static LIST_HEAD(f2fs_stat_list);
-static DEFINE_MUTEX(f2fs_stat_mutex);
+static DEFINE_RAW_SPINLOCK(f2fs_stat_lock);
 #ifdef CONFIG_DEBUG_FS
 static struct dentry *f2fs_debugfs_root;
 #endif
@@ -345,8 +345,9 @@ static int stat_show(struct seq_file *s, void *v)
 {
 	struct f2fs_stat_info *si;
 	int i = 0, j = 0;
+	unsigned long flags;
 
-	mutex_lock(&f2fs_stat_mutex);
+	raw_spin_lock_irqsave(&f2fs_stat_lock, flags);
 	list_for_each_entry(si, &f2fs_stat_list, stat_list) {
 		update_general_status(si->sbi);
 
@@ -574,7 +575,7 @@ static int stat_show(struct seq_file *s, void *v)
 		seq_printf(s, "  - paged : %llu KB\n",
 				si->page_mem >> 10);
 	}
-	mutex_unlock(&f2fs_stat_mutex);
+	raw_spin_unlock_irqrestore(&f2fs_stat_lock, flags);
 	return 0;
 }
 
@@ -585,6 +586,7 @@ int f2fs_build_stats(struct f2fs_sb_info *sbi)
 {
 	struct f2fs_super_block *raw_super = F2FS_RAW_SUPER(sbi);
 	struct f2fs_stat_info *si;
+	unsigned long flags;
 	int i;
 
 	si = f2fs_kzalloc(sbi, sizeof(struct f2fs_stat_info), GFP_KERNEL);
@@ -620,9 +622,9 @@ int f2fs_build_stats(struct f2fs_sb_info *sbi)
 	atomic_set(&sbi->max_aw_cnt, 0);
 	atomic_set(&sbi->max_vw_cnt, 0);
 
-	mutex_lock(&f2fs_stat_mutex);
+	raw_spin_lock_irqsave(&f2fs_stat_lock, flags);
 	list_add_tail(&si->stat_list, &f2fs_stat_list);
-	mutex_unlock(&f2fs_stat_mutex);
+	raw_spin_unlock_irqrestore(&f2fs_stat_lock, flags);
 
 	return 0;
 }
@@ -630,10 +632,11 @@ int f2fs_build_stats(struct f2fs_sb_info *sbi)
 void f2fs_destroy_stats(struct f2fs_sb_info *sbi)
 {
 	struct f2fs_stat_info *si = F2FS_STAT(sbi);
+	unsigned long flags;
 
-	mutex_lock(&f2fs_stat_mutex);
+	raw_spin_lock_irqsave(&f2fs_stat_lock, flags);
 	list_del(&si->stat_list);
-	mutex_unlock(&f2fs_stat_mutex);
+	raw_spin_unlock_irqrestore(&f2fs_stat_lock, flags);
 
 	kfree(si);
 }
Signature · 1 line
2.34.1
author maintainer
#809 lore raw

PATCH 800/1017 f2fs: compress: fix to print raw data size in error path of lz4 decompression

In reply to Greg Kroah-Hartman

From: Chao Yu <[email protected]>

[ Upstream commit d284af43f703760e261b1601378a0c13a19d5f1f ]

In lz4_decompress_pages(), if size of decompressed data is not equal to expected one, we should print the size rather than size of target buffer for decompressed data, fix it.

---
 fs/f2fs/compress.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)
fs/f2fs/compress.c +2 -3
index 190a3c4d4c91..ced0fd54658c 100644
@@ -313,10 +313,9 @@ static int lz4_decompress_pages(struct decompress_io_ctx *dic)
 	}
 
 	if (ret != PAGE_SIZE << dic->log_cluster_size) {
-		printk_ratelimited("%sF2FS-fs (%s): lz4 invalid rlen:%zu, "
+		printk_ratelimited("%sF2FS-fs (%s): lz4 invalid ret:%d, "
 					"expected:%lu\n", KERN_ERR,
-					F2FS_I_SB(dic->inode)->sb->s_id,
-					dic->rlen,
+					F2FS_I_SB(dic->inode)->sb->s_id, ret,
 					PAGE_SIZE << dic->log_cluster_size);
 		return -EIO;
 	}
Signature · 1 line
2.34.1
author maintainer
#810 lore raw

PATCH 801/1017 Adjust cifssb maximum read size

In reply to Greg Kroah-Hartman

From: Rohith Surabattula <[email protected]>

[ Upstream commit 06a466565d54a1a42168f9033a062a3f5c40e73b ]

When session gets reconnected during mount then read size in super block fs context gets set to zero and after negotiate, rsize is not modified which results in incorrect read with requested bytes as zero. Fixes intermittent failure of xfstest generic/240

Note that stable requires a different version of this patch which will be sent to the stable mailing list.

---
 fs/cifs/cifsfs.c |  3 +++
 fs/cifs/file.c   | 10 ++++++++++
 2 files changed, 13 insertions(+)
fs/cifs/cifsfs.c +3 -0
index 9dd5f89557b7..02b762b7e3fd 100644
@@ -209,6 +209,9 @@ cifs_read_super(struct super_block *sb)
 	if (rc)
 		goto out_no_root;
 	/* tune readahead according to rsize if readahead size not set on mount */
+	if (cifs_sb->ctx->rsize == 0)
+		cifs_sb->ctx->rsize =
+			tcon->ses->server->ops->negotiate_rsize(tcon, cifs_sb->ctx);
 	if (cifs_sb->ctx->rasize)
 		sb->s_bdi->ra_pages = cifs_sb->ctx->rasize / PAGE_SIZE;
 	else
fs/cifs/file.c +10 -0
index 9fee3af83a73..abadc2f86dea 100644
@@ -3734,6 +3734,11 @@ cifs_send_async_read(loff_t offset, size_t len, struct cifsFileInfo *open_file,
 				break;
 		}
 
+		if (cifs_sb->ctx->rsize == 0)
+			cifs_sb->ctx->rsize =
+				server->ops->negotiate_rsize(tlink_tcon(open_file->tlink),
+							     cifs_sb->ctx);
+
 		rc = server->ops->wait_mtu_credits(server, cifs_sb->ctx->rsize,
 						   &rsize, credits);
 		if (rc)
@@ -4512,6 +4517,11 @@ static int cifs_readpages(struct file *file, struct address_space *mapping,
 				break;
 		}
 
+		if (cifs_sb->ctx->rsize == 0)
+			cifs_sb->ctx->rsize =
+				server->ops->negotiate_rsize(tlink_tcon(open_file->tlink),
+							     cifs_sb->ctx);
+
 		rc = server->ops->wait_mtu_credits(server, cifs_sb->ctx->rsize,
 						   &rsize, credits);
 		if (rc)
Signature · 1 line
2.34.1
author maintainer
#811 lore raw

PATCH 802/1017 ntfs: add sanity check on allocation size

In reply to Greg Kroah-Hartman

From: Dongliang Mu <[email protected]>

[ Upstream commit 714fbf2647b1a33d914edd695d4da92029c7e7c0 ]

ntfs_read_inode_mount invokes ntfs_malloc_nofs with zero allocation size. It triggers one BUG in the __ntfs_malloc function.

Fix this by adding sanity check on ni->attr_list_size.

---
 fs/ntfs/inode.c | 4 ++++
 1 file changed, 4 insertions(+)
fs/ntfs/inode.c +4 -0
index 4474adb393ca..517b71c73aa9 100644
@@ -1881,6 +1881,10 @@ int ntfs_read_inode_mount(struct inode *vi)
 		}
 		/* Now allocate memory for the attribute list. */
 		ni->attr_list_size = (u32)ntfs_attr_size(a);
+		if (!ni->attr_list_size) {
+			ntfs_error(sb, "Attr_list_size is zero");
+			goto put_err_out;
+		}
 		ni->attr_list = ntfs_malloc_nofs(ni->attr_list_size);
 		if (!ni->attr_list) {
 			ntfs_error(sb, "Not enough memory to allocate buffer "
Signature · 1 line
2.34.1
author maintainer
#812 lore raw

PATCH 803/1017 media: staging: media: zoran: move videodev alloc

In reply to Greg Kroah-Hartman

From: Corentin Labbe <[email protected]>

[ Upstream commit 82e3a496eb56da0b9f29fdc5b63cedb3289e91de ]

Move some code out of zr36057_init() and create new functions for handling zr->video_dev. This permit to ease code reading and fix a zr->video_dev memory leak.

---
 drivers/staging/media/zoran/zoran.h        |  2 +-
 drivers/staging/media/zoran/zoran_card.c   | 80 ++++++++++++++--------
 drivers/staging/media/zoran/zoran_driver.c |  5 +-
 3 files changed, 54 insertions(+), 33 deletions(-)
drivers/staging/media/zoran/zoran.h +1 -1
index b1ad2a2b914c..50d5a7acfab6 100644
@@ -313,6 +313,6 @@ static inline struct zoran *to_zoran(struct v4l2_device *v4l2_dev)
 
 #endif
 
-int zoran_queue_init(struct zoran *zr, struct vb2_queue *vq);
+int zoran_queue_init(struct zoran *zr, struct vb2_queue *vq, int dir);
 void zoran_queue_exit(struct zoran *zr);
 int zr_set_buf(struct zoran *zr);
drivers/staging/media/zoran/zoran_card.c +50 -30

115 lines not on this page · raw message

drivers/staging/media/zoran/zoran_driver.c +3 -2
index 46382e43f1bf..551db338c7f7 100644
@@ -1008,7 +1008,7 @@ static const struct vb2_ops zr_video_qops = {
 	.wait_finish            = vb2_ops_wait_finish,
 };
 
-int zoran_queue_init(struct zoran *zr, struct vb2_queue *vq)
+int zoran_queue_init(struct zoran *zr, struct vb2_queue *vq, int dir)
 {
 	int err;
 
@@ -1016,7 +1016,8 @@ int zoran_queue_init(struct zoran *zr, struct vb2_queue *vq)
 	INIT_LIST_HEAD(&zr->queued_bufs);
 
 	vq->dev = &zr->pci_dev->dev;
-	vq->type = V4L2_BUF_TYPE_VIDEO_CAPTURE;
+	vq->type = dir;
+
 	vq->io_modes = VB2_USERPTR | VB2_DMABUF | VB2_MMAP | VB2_READ | VB2_WRITE;
 	vq->drv_priv = zr;
 	vq->buf_struct_size = sizeof(struct zr_buffer);
Signature · 1 line
2.34.1
author maintainer
#813 lore raw

PATCH 804/1017 media: staging: media: zoran: calculate the right buffer number for zoran_reap_stat_com

In reply to Greg Kroah-Hartman

From: Corentin Labbe <[email protected]>

[ Upstream commit e3b86f4e558cea9eed71d894df2f19b10d60a207 ]

On the case tmp_dcim=1, the index of buffer is miscalculated. This generate a NULL pointer dereference later.

So let's fix the calcul and add a check to prevent this to reappear.

---
 drivers/staging/media/zoran/zoran_device.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)
drivers/staging/media/zoran/zoran_device.c +6 -1
index 5b12a730a229..fb1f0465ca87 100644
@@ -814,7 +814,7 @@ static void zoran_reap_stat_com(struct zoran *zr)
 		if (zr->jpg_settings.tmp_dcm == 1)
 			i = (zr->jpg_dma_tail - zr->jpg_err_shift) & BUZ_MASK_STAT_COM;
 		else
-			i = ((zr->jpg_dma_tail - zr->jpg_err_shift) & 1) * 2 + 1;
+			i = ((zr->jpg_dma_tail - zr->jpg_err_shift) & 1) * 2;
 
 		stat_com = le32_to_cpu(zr->stat_com[i]);
 		if ((stat_com & 1) == 0) {
@@ -826,6 +826,11 @@ static void zoran_reap_stat_com(struct zoran *zr)
 		size = (stat_com & GENMASK(22, 1)) >> 1;
 
 		buf = zr->inuse[i];
+		if (!buf) {
+			spin_unlock_irqrestore(&zr->queued_bufs_lock, flags);
+			pci_err(zr->pci_dev, "No buffer at slot %d\n", i);
+			return;
+		}
 		buf->vbuf.vb2_buf.timestamp = ktime_get_ns();
 
 		if (zr->codec_mode == BUZ_MODE_MOTION_COMPRESS) {
Signature · 1 line
2.34.1
author maintainer
#814 lore raw

PATCH 805/1017 media: staging: media: zoran: fix various V4L2 compliance errors

In reply to Greg Kroah-Hartman

From: Hans Verkuil <[email protected]>

[ Upstream commit 914941827aad5ecddf9bf3a6dee67fbec1af1fff ]

This fixes several issues found with 'v4l2-compliance -s':

1) read()/write() is supported, but not reported in the capabilities
2) S_STD(G_STD()) failed: setting the same standard should just return 0.
3) G_PARM failed to set readbuffers.
4) different field values in the format vs. what v4l2_buffer reported.
5) zero the sequence number when starting streaming.
6) drop VB_USERPTR: makes no sense with dma_contig streaming.

---
 drivers/staging/media/zoran/zoran_card.c   |  2 +-
 drivers/staging/media/zoran/zoran_driver.c | 13 ++++++++++---
 2 files changed, 11 insertions(+), 4 deletions(-)
drivers/staging/media/zoran/zoran_card.c +1 -1
index a83314b16548..11d415c0c05d 100644
@@ -811,7 +811,7 @@ static int zoran_init_video_device(struct zoran *zr, struct video_device *video_
 	*video_dev = zoran_template;
 	video_dev->v4l2_dev = &zr->v4l2_dev;
 	video_dev->lock = &zr->lock;
-	video_dev->device_caps = V4L2_CAP_STREAMING | dir;
+	video_dev->device_caps = V4L2_CAP_STREAMING | V4L2_CAP_READWRITE | dir;
 
 	strscpy(video_dev->name, ZR_DEVNAME(zr), sizeof(video_dev->name));
 	/*
drivers/staging/media/zoran/zoran_driver.c +10 -3
index 551db338c7f7..84665637ebb7 100644
@@ -255,8 +255,6 @@ static int zoran_querycap(struct file *file, void *__fh, struct v4l2_capability
 	strscpy(cap->card, ZR_DEVNAME(zr), sizeof(cap->card));
 	strscpy(cap->driver, "zoran", sizeof(cap->driver));
 	snprintf(cap->bus_info, sizeof(cap->bus_info), "PCI:%s", pci_name(zr->pci_dev));
-	cap->device_caps = zr->video_dev->device_caps;
-	cap->capabilities = cap->device_caps | V4L2_CAP_DEVICE_CAPS;
 	return 0;
 }
 
@@ -582,6 +580,9 @@ static int zoran_s_std(struct file *file, void *__fh, v4l2_std_id std)
 	struct zoran *zr = video_drvdata(file);
 	int res = 0;
 
+	if (zr->norm == std)
+		return 0;
+
 	if (zr->running != ZORAN_MAP_MODE_NONE)
 		return -EBUSY;
 
@@ -739,6 +740,7 @@ static int zoran_g_parm(struct file *file, void *priv, struct v4l2_streamparm *p
 	if (parm->type != V4L2_BUF_TYPE_VIDEO_CAPTURE)
 		return -EINVAL;
 
+	parm->parm.capture.readbuffers = 9;
 	return 0;
 }
 
@@ -869,6 +871,10 @@ int zr_set_buf(struct zoran *zr)
 		vbuf = &buf->vbuf;
 
 		buf->vbuf.field = V4L2_FIELD_INTERLACED;
+		if (BUZ_MAX_HEIGHT < (zr->v4l_settings.height * 2))
+			buf->vbuf.field = V4L2_FIELD_INTERLACED;
+		else
+			buf->vbuf.field = V4L2_FIELD_TOP;
 		vb2_set_plane_payload(&buf->vbuf.vb2_buf, 0, zr->buffer_size);
 		vb2_buffer_done(&buf->vbuf.vb2_buf, VB2_BUF_STATE_DONE);
 		zr->inuse[0] = NULL;
@@ -928,6 +934,7 @@ static int zr_vb2_start_streaming(struct vb2_queue *vq, unsigned int count)
 		zr->stat_com[j] = cpu_to_le32(1);
 		zr->inuse[j] = NULL;
 	}
+	zr->vbseq = 0;
 
 	if (zr->map_mode != ZORAN_MAP_MODE_RAW) {
 		pci_info(zr->pci_dev, "START JPG\n");
@@ -1018,7 +1025,7 @@ int zoran_queue_init(struct zoran *zr, struct vb2_queue *vq, int dir)
 	vq->dev = &zr->pci_dev->dev;
 	vq->type = dir;
 
-	vq->io_modes = VB2_USERPTR | VB2_DMABUF | VB2_MMAP | VB2_READ | VB2_WRITE;
+	vq->io_modes = VB2_DMABUF | VB2_MMAP | VB2_READ | VB2_WRITE;
 	vq->drv_priv = zr;
 	vq->buf_struct_size = sizeof(struct zr_buffer);
 	vq->ops = &zr_video_qops;
Signature · 1 line
2.34.1
author maintainer
#815 lore raw

PATCH 806/1017 media: atmel: atmel-isc-base: report frame sizes as full supported range

In reply to Greg Kroah-Hartman

From: Eugen Hristev <[email protected]>

[ Upstream commit 72802a86e11c34e819fbfb38f58c5aef668f833d ]

The ISC supports a full broad range of frame sizes.
Until now, the subdevice was queried for possible frame sizes and these were reported to the user space.
However, the ISC should not care about which frame sizes the subdev supports, as long as this frame size is supported.
Thus, report a continuous range from smallest frame size up to the max resolution.

---
 drivers/media/platform/atmel/atmel-isc-base.c | 22 +++++++++----------
 1 file changed, 10 insertions(+), 12 deletions(-)
drivers/media/platform/atmel/atmel-isc-base.c +10 -12
index 660cd0ab6749..24807782c9e5 100644
@@ -1369,14 +1369,12 @@ static int isc_enum_framesizes(struct file *file, void *fh,
 			       struct v4l2_frmsizeenum *fsize)
 {
 	struct isc_device *isc = video_drvdata(file);
-	struct v4l2_subdev_frame_size_enum fse = {
-		.code = isc->config.sd_format->mbus_code,
-		.index = fsize->index,
-		.which = V4L2_SUBDEV_FORMAT_ACTIVE,
-	};
 	int ret = -EINVAL;
 	int i;
 
+	if (fsize->index)
+		return -EINVAL;
+
 	for (i = 0; i < isc->num_user_formats; i++)
 		if (isc->user_formats[i]->fourcc == fsize->pixel_format)
 			ret = 0;
@@ -1388,14 +1386,14 @@ static int isc_enum_framesizes(struct file *file, void *fh,
 	if (ret)
 		return ret;
 
-	ret = v4l2_subdev_call(isc->current_subdev->sd, pad, enum_frame_size,
-			       NULL, &fse);
-	if (ret)
-		return ret;
+	fsize->type = V4L2_FRMSIZE_TYPE_CONTINUOUS;
 
-	fsize->type = V4L2_FRMSIZE_TYPE_DISCRETE;
-	fsize->discrete.width = fse.max_width;
-	fsize->discrete.height = fse.max_height;
+	fsize->stepwise.min_width = 16;
+	fsize->stepwise.max_width = isc->max_width;
+	fsize->stepwise.min_height = 16;
+	fsize->stepwise.max_height = isc->max_height;
+	fsize->stepwise.step_width = 1;
+	fsize->stepwise.step_height = 1;
 
 	return 0;
 }
Signature · 1 line
2.34.1
author maintainer
#816 lore raw

PATCH 807/1017 media: ir_toy: free before error exiting

In reply to Greg Kroah-Hartman

From: Peiwei Hu <[email protected]>

[ Upstream commit 52cdb013036391d9d87aba5b4fc49cdfc6ea4b23 ]

Fix leak in error path.

---
 drivers/media/rc/ir_toy.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/media/rc/ir_toy.c +1 -1
index 7e98e7e3aace..196806709259 100644
@@ -458,7 +458,7 @@ static int irtoy_probe(struct usb_interface *intf,
 	err = usb_submit_urb(irtoy->urb_in, GFP_KERNEL);
 	if (err != 0) {
 		dev_err(irtoy->dev, "fail to submit in urb: %d\n", err);
-		return err;
+		goto free_rcdev;
 	}
 
 	err = irtoy_setup(irtoy);
Signature · 1 line
2.34.1
author maintainer
#817 lore raw

PATCH 808/1017 ASoC: sh: rz-ssi: Make the data structures available before registering the handlers

In reply to Greg Kroah-Hartman

From: Lad Prabhakar <[email protected]>

[ Upstream commit 0788785c78342d422f93b1c9831c2b2b7f137937 ]

Initialize the spinlock and make the data structures available before registering the interrupt handlers.

---
 sound/soc/sh/rz-ssi.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)
sound/soc/sh/rz-ssi.c +3 -2
index 37466f65c2b0..16de2633a873 100644
@@ -977,6 +977,9 @@ static int rz_ssi_probe(struct platform_device *pdev)
 	ssi->playback.priv = ssi;
 	ssi->capture.priv = ssi;
 
+	spin_lock_init(&ssi->lock);
+	dev_set_drvdata(&pdev->dev, ssi);
+
 	/* Error Interrupt */
 	ssi->irq_int = platform_get_irq_byname(pdev, "int_req");
 	if (ssi->irq_int < 0)
@@ -1024,8 +1027,6 @@ static int rz_ssi_probe(struct platform_device *pdev)
 	pm_runtime_enable(&pdev->dev);
 	pm_runtime_resume_and_get(&pdev->dev);
 
-	spin_lock_init(&ssi->lock);
-	dev_set_drvdata(&pdev->dev, ssi);
 	ret = devm_snd_soc_register_component(&pdev->dev, &rz_ssi_soc_component,
 					      rz_ssi_soc_dai,
 					      ARRAY_SIZE(rz_ssi_soc_dai));
Signature · 1 line
2.34.1
author maintainer
#818 lore raw

PATCH 809/1017 ASoC: cs42l42: Report full jack status when plug is detected

In reply to Greg Kroah-Hartman

From: Richard Fitzgerald <[email protected]>

[ Upstream commit 8d06f797f844d04a961f201f886f7f9985edc9bf ]

When a plug event is detect report the full state of all status bits, don't assume that there will have been a previous unplug event to clear all the bits. Report the state of both HEADPHONE and MICROPHONE bits according to detected type, and clear all the button status bits. The current button status is already checked and reported at the end of the function.

During a system suspend the jack could be unplugged and plugged, possibly changing the jack type. On resume the interrupt status will indicate a plug event - there will not be an unplug event to clear the bits.

---
 sound/soc/codecs/cs42l42.c | 14 +++++++++++---
 1 file changed, 11 insertions(+), 3 deletions(-)
sound/soc/codecs/cs42l42.c +11 -3
index a63fba4e6c9c..eb170d106396 100644
@@ -1630,7 +1630,11 @@ static irqreturn_t cs42l42_irq_thread(int irq, void *data)
 
 	mutex_lock(&cs42l42->jack_detect_mutex);
 
-	/* Check auto-detect status */
+	/*
+	 * Check auto-detect status. Don't assume a previous unplug event has
+	 * cleared the flags. If the jack is unplugged and plugged during
+	 * system suspend there won't have been an unplug event.
+	 */
 	if ((~masks[5]) & irq_params_table[5].mask) {
 		if (stickies[5] & CS42L42_HSDET_AUTO_DONE_MASK) {
 			cs42l42_process_hs_type_detect(cs42l42);
@@ -1638,11 +1642,15 @@ static irqreturn_t cs42l42_irq_thread(int irq, void *data)
 			case CS42L42_PLUG_CTIA:
 			case CS42L42_PLUG_OMTP:
 				snd_soc_jack_report(cs42l42->jack, SND_JACK_HEADSET,
-						    SND_JACK_HEADSET);
+						    SND_JACK_HEADSET |
+						    SND_JACK_BTN_0 | SND_JACK_BTN_1 |
+						    SND_JACK_BTN_2 | SND_JACK_BTN_3);
 				break;
 			case CS42L42_PLUG_HEADPHONE:
 				snd_soc_jack_report(cs42l42->jack, SND_JACK_HEADPHONE,
-						    SND_JACK_HEADPHONE);
+						    SND_JACK_HEADSET |
+						    SND_JACK_BTN_0 | SND_JACK_BTN_1 |
+						    SND_JACK_BTN_2 | SND_JACK_BTN_3);
 				break;
 			default:
 				break;
Signature · 1 line
2.34.1
author maintainer
#819 lore raw

PATCH 810/1017 ASoC: SOF: Intel: match sdw version on link_slaves_found

In reply to Greg Kroah-Hartman

From: Bard Liao <[email protected]>

[ Upstream commit f67c0c0d3b9048d86ea6ae52e36a2b78c48f265d ]

Codecs with the same part id, manufacturer id and part id, but different sdw version should be treated as different codecs. For example, rt711 and rt711-sdca are different. So, we should match sdw version as well.

---
 sound/soc/sof/intel/hda.c | 15 ++++++++++-----
 1 file changed, 10 insertions(+), 5 deletions(-)
sound/soc/sof/intel/hda.c +10 -5
index 25200a0e1dc9..fc88296ab898 100644
@@ -1200,7 +1200,7 @@ static bool link_slaves_found(struct snd_sof_dev *sdev,
 	struct hdac_bus *bus = sof_to_bus(sdev);
 	struct sdw_intel_slave_id *ids = sdw->ids;
 	int num_slaves = sdw->num_slaves;
-	unsigned int part_id, link_id, unique_id, mfg_id;
+	unsigned int part_id, link_id, unique_id, mfg_id, version;
 	int i, j, k;
 
 	for (i = 0; i < link->num_adr; i++) {
@@ -1210,12 +1210,14 @@ static bool link_slaves_found(struct snd_sof_dev *sdev,
 		mfg_id = SDW_MFG_ID(adr);
 		part_id = SDW_PART_ID(adr);
 		link_id = SDW_DISCO_LINK_ID(adr);
+		version = SDW_VERSION(adr);
 
 		for (j = 0; j < num_slaves; j++) {
 			/* find out how many identical parts were reported on that link */
 			if (ids[j].link_id == link_id &&
 			    ids[j].id.part_id == part_id &&
-			    ids[j].id.mfg_id == mfg_id)
+			    ids[j].id.mfg_id == mfg_id &&
+			    ids[j].id.sdw_version == version)
 				reported_part_count++;
 		}
 
@@ -1224,21 +1226,24 @@ static bool link_slaves_found(struct snd_sof_dev *sdev,
 
 			if (ids[j].link_id != link_id ||
 			    ids[j].id.part_id != part_id ||
-			    ids[j].id.mfg_id != mfg_id)
+			    ids[j].id.mfg_id != mfg_id ||
+			    ids[j].id.sdw_version != version)
 				continue;
 
 			/* find out how many identical parts are expected */
 			for (k = 0; k < link->num_adr; k++) {
 				u64 adr2 = link->adr_d[k].adr;
-				unsigned int part_id2, link_id2, mfg_id2;
+				unsigned int part_id2, link_id2, mfg_id2, version2;
 
 				mfg_id2 = SDW_MFG_ID(adr2);
 				part_id2 = SDW_PART_ID(adr2);
 				link_id2 = SDW_DISCO_LINK_ID(adr2);
+				version2 = SDW_VERSION(adr2);
 
 				if (link_id2 == link_id &&
 				    part_id2 == part_id &&
-				    mfg_id2 == mfg_id)
+				    mfg_id2 == mfg_id &&
+				    version2 == version)
 					expected_part_count++;
 			}
 
Signature · 1 line
2.34.1
author maintainer
#820 lore raw

PATCH 811/1017 media: imx-jpeg: Prevent decoding NV12M jpegs into single-planar buffers

In reply to Greg Kroah-Hartman

From: Mirela Rabulea <[email protected]>

[ Upstream commit 417591a766b3c040c346044541ff949c0b2bb7b2 ]

If the application queues an NV12M jpeg as output buffer, but then queues a single planar capture buffer, the kernel will crash with "Unable to handle kernel NULL pointer dereference" in mxc_jpeg_addrs, prevent this by finishing the job with error.

---
 drivers/media/platform/imx-jpeg/mxc-jpeg.c | 6 ++++++
 1 file changed, 6 insertions(+)
drivers/media/platform/imx-jpeg/mxc-jpeg.c +6 -0
index 4ca96cf9def7..b249c1bbfac8 100644
@@ -947,6 +947,12 @@ static void mxc_jpeg_device_run(void *priv)
 	v4l2_m2m_buf_copy_metadata(src_buf, dst_buf, true);
 
 	jpeg_src_buf = vb2_to_mxc_buf(&src_buf->vb2_buf);
+	if (q_data_cap->fmt->colplanes != dst_buf->vb2_buf.num_planes) {
+		dev_err(dev, "Capture format %s has %d planes, but capture buffer has %d planes\n",
+			q_data_cap->fmt->name, q_data_cap->fmt->colplanes,
+			dst_buf->vb2_buf.num_planes);
+		jpeg_src_buf->jpeg_parse_error = true;
+	}
 	if (jpeg_src_buf->jpeg_parse_error) {
 		jpeg->slot_data[ctx->slot].used = false;
 		v4l2_m2m_src_buf_remove(ctx->fh.m2m_ctx);
Signature · 1 line
2.34.1
author maintainer
#821 lore raw

PATCH 812/1017 ASoC: SOF: Intel: hda: Remove link assignment limitation

In reply to Greg Kroah-Hartman

From: Ranjani Sridharan <[email protected]>

[ Upstream commit 2ce0d008dcc59f9c01f43277b9f9743af7b01dad ]

The limitation to assign a link DMA channel for a BE iff the corresponding host DMA channel is assigned to a connected FE is only applicable if the PROCEN_FMT_QUIRK is set. So, remove it for platforms that do not enable the quirk.

Complements: a792bfc1c2bc ("ASoC: SOF: Intel: hda-stream: limit PROCEN workaround")

---
 sound/soc/sof/intel/hda-dai.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)
sound/soc/sof/intel/hda-dai.c +13 -0
index 76579383d290..1b7ba7f3e99b 100644
@@ -62,6 +62,8 @@ static struct hdac_ext_stream *
 {
 	struct snd_soc_pcm_runtime *rtd = asoc_substream_to_rtd(substream);
 	struct sof_intel_hda_stream *hda_stream;
+	const struct sof_intel_dsp_desc *chip;
+	struct snd_sof_dev *sdev;
 	struct hdac_ext_stream *res = NULL;
 	struct hdac_stream *stream = NULL;
 
@@ -80,9 +82,20 @@ static struct hdac_ext_stream *
 			continue;
 
 		hda_stream = hstream_to_sof_hda_stream(hstream);
+		sdev = hda_stream->sdev;
+		chip = get_chip_info(sdev->pdata);
 
 		/* check if link is available */
 		if (!hstream->link_locked) {
+			/*
+			 * choose the first available link for platforms that do not have the
+			 * PROCEN_FMT_QUIRK set.
+			 */
+			if (!(chip->quirks & SOF_INTEL_PROCEN_FMT_QUIRK)) {
+				res = hstream;
+				break;
+			}
+
 			if (stream->opened) {
 				/*
 				 * check if the stream tag matches the stream
Signature · 1 line
2.34.1
author maintainer
#822 lore raw

PATCH 813/1017 media: iommu/mediatek-v1: Free the existed fwspec if the master dev already has

In reply to Greg Kroah-Hartman

From: Yong Wu <[email protected]>

[ Upstream commit 822a2ed8c606caf6a11b1a180b8e46292bd77d71 ]

When the iommu master device enters of_iommu_xlate, the ops may be NULL(iommu dev is defered), then it will initialize the fwspec here:

[<c0c9c5bc>] (dev_iommu_fwspec_set) from [<c06bda80>] (iommu_fwspec_init+0xbc/0xd4)
[<c06bd9c4>] (iommu_fwspec_init) from [<c06c0db4>] (of_iommu_xlate+0x7c/0x12c)
[<c06c0d38>] (of_iommu_xlate) from [<c06c10e8>] (of_iommu_configure+0x144/0x1e8)

BUT the mtk_iommu_v1.c only supports arm32, the probing flow still is a bit weird. We always expect create the fwspec internally. otherwise it will enter here and return fail.

static int mtk_iommu_create_mapping(struct device *dev,
struct of_phandle_args *args)

{
        ...
	if (!fwspec) {
	        ....
	} else if (dev_iommu_fwspec_get(dev)->ops != &mtk_iommu_ops) {
                >>>>>>>>>>Enter here. return fail.<<<<<<<<<<<<
		return -EINVAL;
	}
	...
}

Thus, Free the existed fwspec if the master device already has fwspec.

This issue is reported at:
https://lore.kernel.org/linux-mediatek/trinity-7d9ebdc9-4849-4d93-bfb5-429dcb4ee449-1626253158870@3c-app-gmx-bs01/

---
 drivers/iommu/mtk_iommu_v1.c | 9 +++++++++
 1 file changed, 9 insertions(+)
drivers/iommu/mtk_iommu_v1.c +9 -0
index be22fcf988ce..1467ba1e4417 100644
@@ -425,6 +425,15 @@ static struct iommu_device *mtk_iommu_probe_device(struct device *dev)
 	struct mtk_iommu_data *data;
 	int err, idx = 0;
 
+	/*
+	 * In the deferred case, free the existed fwspec.
+	 * Always initialize the fwspec internally.
+	 */
+	if (fwspec) {
+		iommu_fwspec_free(dev);
+		fwspec = dev_iommu_fwspec_get(dev);
+	}
+
 	while (!of_parse_phandle_with_args(dev->of_node, "iommus",
 					   "#iommu-cells",
 					   idx, &iommu_spec)) {
Signature · 1 line
2.34.1
author maintainer
#823 lore raw

PATCH 814/1017 media: iommu/mediatek: Return ENODEV if the device is NULL

In reply to Greg Kroah-Hartman

From: Yong Wu <[email protected]>

[ Upstream commit 2fb0feed51085db77606de9b9477c96894328809 ]

The platform device is created at:
of_platform_default_populate_init: arch_initcall_sync
->of_platform_populate

        ->of_platform_device_create_pdata

When entering our probe, all the devices should be already created. if it is null, means NODEV. Currently we don't get the fail case. It's a minor fix, no need add fixes tags.

---
 drivers/iommu/mtk_iommu.c    | 2 +-
 drivers/iommu/mtk_iommu_v1.c | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)
drivers/iommu/mtk_iommu.c +1 -1
index 25b834104790..77ae20ff9b35 100644
@@ -848,7 +848,7 @@ static int mtk_iommu_probe(struct platform_device *pdev)
 		plarbdev = of_find_device_by_node(larbnode);
 		if (!plarbdev) {
 			of_node_put(larbnode);
-			return -EPROBE_DEFER;
+			return -ENODEV;
 		}
 		data->larb_imu[id].dev = &plarbdev->dev;
 
drivers/iommu/mtk_iommu_v1.c +1 -1
index 1467ba1e4417..68bf02f87cfd 100644
@@ -604,7 +604,7 @@ static int mtk_iommu_probe(struct platform_device *pdev)
 		plarbdev = of_find_device_by_node(larbnode);
 		if (!plarbdev) {
 			of_node_put(larbnode);
-			return -EPROBE_DEFER;
+			return -ENODEV;
 		}
 		data->larb_imu[i].dev = &plarbdev->dev;
 
Signature · 1 line
2.34.1
author maintainer
#824 lore raw

PATCH 815/1017 media: iommu/mediatek: Add device_link between the consumer and the larb devices

In reply to Greg Kroah-Hartman

From: Yong Wu <[email protected]>

[ Upstream commit 635319a4a7444ca97124d781cd96deb277ff4d40 ]

MediaTek IOMMU-SMI diagram is like below. all the consumer connect with smi-larb, then connect with smi-common.

        M4U
         |
    smi-common
         |
  -------------
  |         |    ...
  |         |
larb1     larb2
  |         |
vdec       venc

When the consumer works, it should enable the smi-larb's power which also need enable the smi-common's power firstly.

Thus, First of all, use the device link connect the consumer and the smi-larbs. then add device link between the smi-larb and smi-common.

This patch adds device_link between the consumer and the larbs.

When device_link_add, I add the flag DL_FLAG_STATELESS to avoid calling pm_runtime_xx to keep the original status of clocks. It can avoid two issues:
1) Display HW show fastlogo abnormally reported in [1]. At the beggining, all the clocks are enabled before entering kernel, but the clocks for display HW(always in larb0) will be gated after clk_enable and clk_disable called from device_link_add(->pm_runtime_resume) and rpm_idle. The clock operation happened before display driver probe. At that time, the display HW will be abnormal.

2) A deadlock issue reported in [2]. Use DL_FLAG_STATELESS to skip pm_runtime_xx to avoid the deadlock.

Corresponding, DL_FLAG_AUTOREMOVE_CONSUMER can't be added, then device_link_removed should be added explicitly.

Meanwhile, Currently we don't have a device connect with 2 larbs at the same time. Disallow this case, print the error log.

[1] https://lore.kernel.org/linux-mediatek/1564213888.22908.4.camel@mhfsdcap03/
[2] https://lore.kernel.org/patchwork/patch/1086569/

---
 drivers/iommu/mtk_iommu.c    | 30 ++++++++++++++++++++++++++++++
 drivers/iommu/mtk_iommu_v1.c | 29 ++++++++++++++++++++++++++++-
 2 files changed, 58 insertions(+), 1 deletion(-)
drivers/iommu/mtk_iommu.c +30 -0
index 77ae20ff9b35..5971a1168666 100644
@@ -562,22 +562,52 @@ static struct iommu_device *mtk_iommu_probe_device(struct device *dev)
 {
 	struct iommu_fwspec *fwspec = dev_iommu_fwspec_get(dev);
 	struct mtk_iommu_data *data;
+	struct device_link *link;
+	struct device *larbdev;
+	unsigned int larbid, larbidx, i;
 
 	if (!fwspec || fwspec->ops != &mtk_iommu_ops)
 		return ERR_PTR(-ENODEV); /* Not a iommu client device */
 
 	data = dev_iommu_priv_get(dev);
 
+	/*
+	 * Link the consumer device with the smi-larb device(supplier).
+	 * The device that connects with each a larb is a independent HW.
+	 * All the ports in each a device should be in the same larbs.
+	 */
+	larbid = MTK_M4U_TO_LARB(fwspec->ids[0]);
+	for (i = 1; i < fwspec->num_ids; i++) {
+		larbidx = MTK_M4U_TO_LARB(fwspec->ids[i]);
+		if (larbid != larbidx) {
+			dev_err(dev, "Can only use one larb. Fail@larb%d-%d.\n",
+				larbid, larbidx);
+			return ERR_PTR(-EINVAL);
+		}
+	}
+	larbdev = data->larb_imu[larbid].dev;
+	link = device_link_add(dev, larbdev,
+			       DL_FLAG_PM_RUNTIME | DL_FLAG_STATELESS);
+	if (!link)
+		dev_err(dev, "Unable to link %s\n", dev_name(larbdev));
 	return &data->iommu;
 }
 
 static void mtk_iommu_release_device(struct device *dev)
 {
 	struct iommu_fwspec *fwspec = dev_iommu_fwspec_get(dev);
+	struct mtk_iommu_data *data;
+	struct device *larbdev;
+	unsigned int larbid;
 
 	if (!fwspec || fwspec->ops != &mtk_iommu_ops)
 		return;
 
+	data = dev_iommu_priv_get(dev);
+	larbid = MTK_M4U_TO_LARB(fwspec->ids[0]);
+	larbdev = data->larb_imu[larbid].dev;
+	device_link_remove(dev, larbdev);
+
 	iommu_fwspec_free(dev);
 }
 
drivers/iommu/mtk_iommu_v1.c +28 -1
index 68bf02f87cfd..bc7ee90b9373 100644
@@ -423,7 +423,9 @@ static struct iommu_device *mtk_iommu_probe_device(struct device *dev)
 	struct iommu_fwspec *fwspec = dev_iommu_fwspec_get(dev);
 	struct of_phandle_args iommu_spec;
 	struct mtk_iommu_data *data;
-	int err, idx = 0;
+	int err, idx = 0, larbid, larbidx;
+	struct device_link *link;
+	struct device *larbdev;
 
 	/*
 	 * In the deferred case, free the existed fwspec.
@@ -453,6 +455,23 @@ static struct iommu_device *mtk_iommu_probe_device(struct device *dev)
 
 	data = dev_iommu_priv_get(dev);
 
+	/* Link the consumer device with the smi-larb device(supplier) */
+	larbid = mt2701_m4u_to_larb(fwspec->ids[0]);
+	for (idx = 1; idx < fwspec->num_ids; idx++) {
+		larbidx = mt2701_m4u_to_larb(fwspec->ids[idx]);
+		if (larbid != larbidx) {
+			dev_err(dev, "Can only use one larb. Fail@larb%d-%d.\n",
+				larbid, larbidx);
+			return ERR_PTR(-EINVAL);
+		}
+	}
+
+	larbdev = data->larb_imu[larbid].dev;
+	link = device_link_add(dev, larbdev,
+			       DL_FLAG_PM_RUNTIME | DL_FLAG_STATELESS);
+	if (!link)
+		dev_err(dev, "Unable to link %s\n", dev_name(larbdev));
+
 	return &data->iommu;
 }
 
@@ -473,10 +492,18 @@ static void mtk_iommu_probe_finalize(struct device *dev)
 static void mtk_iommu_release_device(struct device *dev)
 {
 	struct iommu_fwspec *fwspec = dev_iommu_fwspec_get(dev);
+	struct mtk_iommu_data *data;
+	struct device *larbdev;
+	unsigned int larbid;
 
 	if (!fwspec || fwspec->ops != &mtk_iommu_ops)
 		return;
 
+	data = dev_iommu_priv_get(dev);
+	larbid = mt2701_m4u_to_larb(fwspec->ids[0]);
+	larbdev = data->larb_imu[larbid].dev;
+	device_link_remove(dev, larbdev);
+
 	iommu_fwspec_free(dev);
 }
 
Signature · 1 line
2.34.1
author maintainer
#825 lore raw

PATCH 816/1017 video: fbdev: nvidiafb: Use strscpy() to prevent buffer overflow

In reply to Greg Kroah-Hartman

From: Tim Gardner <[email protected]>

[ Upstream commit 37a1a2e6eeeb101285cd34e12e48a881524701aa ]

Coverity complains of a possible buffer overflow. However, given the 'static' scope of nvidia_setup_i2c_bus() it looks like that can't happen after examiniing the call sites.

CID 19036 (#1 of 1): Copy into fixed size buffer (STRING_OVERFLOW)
1. fixed_size_dest: You might overrun the 48-character fixed-size string chan->adapter.name by copying name without checking the length.
2. parameter_as_source: Note: This defect has an elevated risk because the source argument is a parameter of the current function.
89 strcpy(chan->adapter.name, name);

Fix this warning by using strscpy() which will silence the warning and prevent any future buffer overflows should the names used to identify the channel become much longer.

---
 drivers/video/fbdev/nvidia/nv_i2c.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/video/fbdev/nvidia/nv_i2c.c +1 -1
index d7994a173245..0b48965a6420 100644
@@ -86,7 +86,7 @@ static int nvidia_setup_i2c_bus(struct nvidia_i2c_chan *chan, const char *name,
 {
 	int rc;
 
-	strcpy(chan->adapter.name, name);
+	strscpy(chan->adapter.name, name, sizeof(chan->adapter.name));
 	chan->adapter.owner = THIS_MODULE;
 	chan->adapter.class = i2c_class;
 	chan->adapter.algo_data = &chan->algo;
Signature · 1 line
2.34.1
author maintainer
#826 lore raw

PATCH 817/1017 video: fbdev: w100fb: Reset global state

In reply to Greg Kroah-Hartman

From: Evgeny Novikov <[email protected]>

[ Upstream commit 8738ddcac644964ae128ccd3d80d48773c8d528e ]

w100fb_probe() did not reset the global state to its initial state. This can result in invocation of iounmap() even when there was not the appropriate successful call of ioremap(). For instance, this may be the case if first probe fails after two successful ioremap() while second probe fails when first ioremap() fails. The similar issue is with w100fb_remove(). The patch fixes both bugs.

Found by Linux Driver Verification project (linuxtesting.org).

---
 drivers/video/fbdev/w100fb.c | 15 ++++++++++++---
 1 file changed, 12 insertions(+), 3 deletions(-)
drivers/video/fbdev/w100fb.c +12 -3
index d96ab28f8ce4..4e641a780726 100644
@@ -770,12 +770,18 @@ static int w100fb_probe(struct platform_device *pdev)
 		fb_dealloc_cmap(&info->cmap);
 		kfree(info->pseudo_palette);
 	}
-	if (remapped_fbuf != NULL)
+	if (remapped_fbuf != NULL) {
 		iounmap(remapped_fbuf);
-	if (remapped_regs != NULL)
+		remapped_fbuf = NULL;
+	}
+	if (remapped_regs != NULL) {
 		iounmap(remapped_regs);
-	if (remapped_base != NULL)
+		remapped_regs = NULL;
+	}
+	if (remapped_base != NULL) {
 		iounmap(remapped_base);
+		remapped_base = NULL;
+	}
 	if (info)
 		framebuffer_release(info);
 	return err;
@@ -795,8 +801,11 @@ static int w100fb_remove(struct platform_device *pdev)
 	fb_dealloc_cmap(&info->cmap);
 
 	iounmap(remapped_base);
+	remapped_base = NULL;
 	iounmap(remapped_regs);
+	remapped_regs = NULL;
 	iounmap(remapped_fbuf);
+	remapped_fbuf = NULL;
 
 	framebuffer_release(info);
 
Signature · 1 line
2.34.1
author maintainer
#827 lore raw

PATCH 818/1017 video: fbdev: cirrusfb: check pixclock to avoid divide by zero

In reply to Greg Kroah-Hartman

From: George Kennedy <[email protected]>

[ Upstream commit 5c6f402bdcf9e7239c6bc7087eda71ac99b31379 ]

Do a sanity check on pixclock value to avoid divide by zero.

If the pixclock value is zero, the cirrusfb driver will round up pixclock to get the derived frequency as close to maxclock as possible.

Syzkaller reported a divide error in cirrusfb_check_pixclock.

divide error: 0000 [#1] SMP KASAN PTI

CPU: 0 PID: 14938 Comm: cirrusfb_test Not tainted 5.15.0-rc6 #1
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.11.0-2
RIP: 0010:cirrusfb_check_var+0x6f1/0x1260

Call Trace:
 fb_set_var+0x398/0xf90
 do_fb_ioctl+0x4b8/0x6f0
 fb_ioctl+0xeb/0x130
 __x64_sys_ioctl+0x19d/0x220
 do_syscall_64+0x3a/0x80
 entry_SYSCALL_64_after_hwframe+0x44/0xae
---
 drivers/video/fbdev/cirrusfb.c | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)
drivers/video/fbdev/cirrusfb.c +8 -8
index 93802abbbc72..3d47c347b897 100644
@@ -469,7 +469,7 @@ static int cirrusfb_check_mclk(struct fb_info *info, long freq)
 	return 0;
 }
 
-static int cirrusfb_check_pixclock(const struct fb_var_screeninfo *var,
+static int cirrusfb_check_pixclock(struct fb_var_screeninfo *var,
 				   struct fb_info *info)
 {
 	long freq;
@@ -478,9 +478,7 @@ static int cirrusfb_check_pixclock(const struct fb_var_screeninfo *var,
 	unsigned maxclockidx = var->bits_per_pixel >> 3;
 
 	/* convert from ps to kHz */
-	freq = PICOS2KHZ(var->pixclock);
-
-	dev_dbg(info->device, "desired pixclock: %ld kHz\n", freq);
+	freq = PICOS2KHZ(var->pixclock ? : 1);
 
 	maxclock = cirrusfb_board_info[cinfo->btype].maxclock[maxclockidx];
 	cinfo->multiplexing = 0;
@@ -488,11 +486,13 @@ static int cirrusfb_check_pixclock(const struct fb_var_screeninfo *var,
 	/* If the frequency is greater than we can support, we might be able
 	 * to use multiplexing for the video mode */
 	if (freq > maxclock) {
-		dev_err(info->device,
-			"Frequency greater than maxclock (%ld kHz)\n",
-			maxclock);
-		return -EINVAL;
+		var->pixclock = KHZ2PICOS(maxclock);
+
+		while ((freq = PICOS2KHZ(var->pixclock)) > maxclock)
+			var->pixclock++;
 	}
+	dev_dbg(info->device, "desired pixclock: %ld kHz\n", freq);
+
 	/*
 	 * Additional constraint: 8bpp uses DAC clock doubling to allow maximum
 	 * pixel clock
Signature · 1 line
2.34.1
author maintainer
#828 lore raw

PATCH 819/1017 video: fbdev: omapfb: acx565akm: replace snprintf with sysfs_emit

In reply to Greg Kroah-Hartman

From: Yang Guang <[email protected]>

[ Upstream commit 24565bc4115961db7ee64fcc7ad2a7437c0d0a49 ]

coccinelle report:
./drivers/video/fbdev/omap2/omapfb/displays/panel-sony-acx565akm.c:
479:9-17: WARNING: use scnprintf or sprintf

Use sysfs_emit instead of scnprintf or sprintf makes more sense.

---
 .../video/fbdev/omap2/omapfb/displays/panel-sony-acx565akm.c    | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/video/fbdev/omap2/omapfb/displays/panel-sony-acx565akm.c +1 -1
index 8d8b5ff7d43c..3696eb09b69b 100644
@@ -476,7 +476,7 @@ static ssize_t show_cabc_available_modes(struct device *dev,
 	int i;
 
 	if (!ddata->has_cabc)
-		return snprintf(buf, PAGE_SIZE, "%s\n", cabc_modes[0]);
+		return sysfs_emit(buf, "%s\n", cabc_modes[0]);
 
 	for (i = 0, len = 0;
 	     len < PAGE_SIZE && i < ARRAY_SIZE(cabc_modes); i++)
Signature · 1 line
2.34.1
author maintainer
#829 lore raw

PATCH 820/1017 ARM: dts: qcom: fix gic_irq_domain_translate warnings for msm8960

In reply to Greg Kroah-Hartman

From: David Heidelberg <[email protected]>

[ Upstream commit 6f7e221e7a5cfc3299616543fce42b36e631497b ]

IRQ types blindly copied from very similar APQ8064.

Fixes warnings as:

WARNING: CPU: 0 PID: 1 at drivers/irqchip/irq-gic.c:1080 gic_irq_domain_translate+0x118/0x120

...

---
 arch/arm/boot/dts/qcom-msm8960.dtsi | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)
arch/arm/boot/dts/qcom-msm8960.dtsi +5 -3
index 2a0ec97a264f..a0f9ab7f08f3 100644
@@ -146,7 +146,9 @@
 			reg		= <0x108000 0x1000>;
 			qcom,ipc	= <&l2cc 0x8 2>;
 
-			interrupts	= <0 19 0>, <0 21 0>, <0 22 0>;
+			interrupts	= <GIC_SPI 19 IRQ_TYPE_EDGE_RISING>,
+					  <GIC_SPI 21 IRQ_TYPE_EDGE_RISING>,
+					  <GIC_SPI 22 IRQ_TYPE_EDGE_RISING>;
 			interrupt-names	= "ack", "err", "wakeup";
 
 			regulators {
@@ -192,7 +194,7 @@
 				compatible = "qcom,msm-uartdm-v1.3", "qcom,msm-uartdm";
 				reg = <0x16440000 0x1000>,
 				      <0x16400000 0x1000>;
-				interrupts = <0 154 0x0>;
+				interrupts = <GIC_SPI 154 IRQ_TYPE_LEVEL_HIGH>;
 				clocks = <&gcc GSBI5_UART_CLK>, <&gcc GSBI5_H_CLK>;
 				clock-names = "core", "iface";
 				status = "disabled";
@@ -318,7 +320,7 @@
 				#address-cells = <1>;
 				#size-cells = <0>;
 				reg = <0x16080000 0x1000>;
-				interrupts = <0 147 0>;
+				interrupts = <GIC_SPI 147 IRQ_TYPE_LEVEL_HIGH>;
 				spi-max-frequency = <24000000>;
 				cs-gpios = <&msmgpio 8 0>;
 
Signature · 1 line
2.34.1
author maintainer
#830 lore raw

PATCH 821/1017 ARM: dts: bcm2837: Add the missing L1/L2 cache information

In reply to Greg Kroah-Hartman

From: Richard Schleich <[email protected]>

[ Upstream commit bdf8762da268d2a34abf517c36528413906e9cd5 ]

This patch fixes the kernel warning "cacheinfo: Unable to detect cache hierarchy for CPU 0" for the bcm2837 on newer kernel versions.

[florian: Align and remove comments matching property values]
---
 arch/arm/boot/dts/bcm2837.dtsi | 49 ++++++++++++++++++++++++++++++++++
 1 file changed, 49 insertions(+)
arch/arm/boot/dts/bcm2837.dtsi +49 -0
index 0199ec98cd61..5dbdebc46259 100644
@@ -40,12 +40,26 @@
 		#size-cells = <0>;
 		enable-method = "brcm,bcm2836-smp"; // for ARM 32-bit
 
+		/* Source for d/i-cache-line-size and d/i-cache-sets
+		 * https://developer.arm.com/documentation/ddi0500/e/level-1-memory-system
+		 * /about-the-l1-memory-system?lang=en
+		 *
+		 * Source for d/i-cache-size
+		 * https://magpi.raspberrypi.com/articles/raspberry-pi-3-specs-benchmarks
+		 */
 		cpu0: cpu@0 {
 			device_type = "cpu";
 			compatible = "arm,cortex-a53";
 			reg = <0>;
 			enable-method = "spin-table";
 			cpu-release-addr = <0x0 0x000000d8>;
+			d-cache-size = <0x8000>;
+			d-cache-line-size = <64>;
+			d-cache-sets = <128>; // 32KiB(size)/64(line-size)=512ways/4-way set
+			i-cache-size = <0x8000>;
+			i-cache-line-size = <64>;
+			i-cache-sets = <256>; // 32KiB(size)/64(line-size)=512ways/2-way set
+			next-level-cache = <&l2>;
 		};
 
 		cpu1: cpu@1 {
@@ -54,6 +68,13 @@
 			reg = <1>;
 			enable-method = "spin-table";
 			cpu-release-addr = <0x0 0x000000e0>;
+			d-cache-size = <0x8000>;
+			d-cache-line-size = <64>;
+			d-cache-sets = <128>; // 32KiB(size)/64(line-size)=512ways/4-way set
+			i-cache-size = <0x8000>;
+			i-cache-line-size = <64>;
+			i-cache-sets = <256>; // 32KiB(size)/64(line-size)=512ways/2-way set
+			next-level-cache = <&l2>;
 		};
 
 		cpu2: cpu@2 {
@@ -62,6 +83,13 @@
 			reg = <2>;
 			enable-method = "spin-table";
 			cpu-release-addr = <0x0 0x000000e8>;
+			d-cache-size = <0x8000>;
+			d-cache-line-size = <64>;
+			d-cache-sets = <128>; // 32KiB(size)/64(line-size)=512ways/4-way set
+			i-cache-size = <0x8000>;
+			i-cache-line-size = <64>;
+			i-cache-sets = <256>; // 32KiB(size)/64(line-size)=512ways/2-way set
+			next-level-cache = <&l2>;
 		};
 
 		cpu3: cpu@3 {
@@ -70,6 +98,27 @@
 			reg = <3>;
 			enable-method = "spin-table";
 			cpu-release-addr = <0x0 0x000000f0>;
+			d-cache-size = <0x8000>;
+			d-cache-line-size = <64>;
+			d-cache-sets = <128>; // 32KiB(size)/64(line-size)=512ways/4-way set
+			i-cache-size = <0x8000>;
+			i-cache-line-size = <64>;
+			i-cache-sets = <256>; // 32KiB(size)/64(line-size)=512ways/2-way set
+			next-level-cache = <&l2>;
+		};
+
+		/* Source for cache-line-size + cache-sets
+		 * https://developer.arm.com/documentation/ddi0500
+		 * /e/level-2-memory-system/about-the-l2-memory-system?lang=en
+		 * Source for cache-size
+		 * https://datasheets.raspberrypi.com/cm/cm1-and-cm3-datasheet.pdf
+		 */
+		l2: l2-cache0 {
+			compatible = "cache";
+			cache-size = <0x80000>;
+			cache-line-size = <64>;
+			cache-sets = <512>; // 512KiB(size)/64(line-size)=8192ways/16-way set
+			cache-level = <2>;
 		};
 	};
 };
Signature · 1 line
2.34.1
author maintainer
#831 lore raw

PATCH 822/1017 ASoC: madera: Add dependencies on MFD

In reply to Greg Kroah-Hartman

From: Charles Keepax <[email protected]>

[ Upstream commit ec29170c724ca30305fc3a19ba2ee73ecac65509 ]

The Madera CODECs use regmap_irq functions but nothing ensures that regmap_irq is built into the kernel. Add dependencies on the ASoC symbols for the relevant MFD component. There is no point in building the ASoC driver if the MFD doesn't support it and the MFD part contains the necessary dependencies to ensure everything is built into the kernel.

---
 sound/soc/codecs/Kconfig | 5 +++++
 1 file changed, 5 insertions(+)
sound/soc/codecs/Kconfig +5 -0
index 3a610ba183ff..0d4e1fb9befc 100644
@@ -707,6 +707,7 @@ config SND_SOC_CS4349
 
 config SND_SOC_CS47L15
 	tristate
+	depends on MFD_CS47L15
 
 config SND_SOC_CS47L24
 	tristate
@@ -714,15 +715,19 @@ config SND_SOC_CS47L24
 
 config SND_SOC_CS47L35
 	tristate
+	depends on MFD_CS47L35
 
 config SND_SOC_CS47L85
 	tristate
+	depends on MFD_CS47L85
 
 config SND_SOC_CS47L90
 	tristate
+	depends on MFD_CS47L90
 
 config SND_SOC_CS47L92
 	tristate
+	depends on MFD_CS47L92
 
 # Cirrus Logic Quad-Channel ADC
 config SND_SOC_CS53L30
Signature · 1 line
2.34.1
author maintainer
#832 lore raw

PATCH 823/1017 media: atomisp_gmin_platform: Add DMI quirk to not turn AXP ELDO2 regulator off on some boards

In reply to Greg Kroah-Hartman

From: Hans de Goede <[email protected]>

[ Upstream commit 2c39a01154ea57d596470afa1d278e3be3b37f6a ]

The TrekStor SurfTab duo W1 10.1 has a hw bug where turning eldo2 back on after having turned it off causes the CPLM3218 ambient-light-sensor on the front camera sensor's I2C bus to crash, hanging the bus.

Add a DMI quirk table for systems on which to leave eldo2 on.

Note an alternative fix is to turn off the CPLM3218 ambient-light-sensor as long as the camera sensor is being used, this is what Windows seems to do as a workaround (based on analyzing the DSDT). But that is not easy to do cleanly under Linux.

---
 .../media/atomisp/pci/atomisp_gmin_platform.c  | 18 ++++++++++++++++++
 1 file changed, 18 insertions(+)
drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c +18 -0
index 62dc06e22476..cd0a771454da 100644
@@ -729,6 +729,21 @@ static int axp_regulator_set(struct device *dev, struct gmin_subdev *gs,
 	return 0;
 }
 
+/*
+ * Some boards contain a hw-bug where turning eldo2 back on after having turned
+ * it off causes the CPLM3218 ambient-light-sensor on the image-sensor's I2C bus
+ * to crash, hanging the bus. Do not turn eldo2 off on these systems.
+ */
+static const struct dmi_system_id axp_leave_eldo2_on_ids[] = {
+	{
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "TrekStor"),
+			DMI_MATCH(DMI_PRODUCT_NAME, "SurfTab duo W1 10.1 (VT4)"),
+		},
+	},
+	{ }
+};
+
 static int axp_v1p8_on(struct device *dev, struct gmin_subdev *gs)
 {
 	int ret;
@@ -763,6 +778,9 @@ static int axp_v1p8_off(struct device *dev, struct gmin_subdev *gs)
 	if (ret)
 		return ret;
 
+	if (dmi_check_system(axp_leave_eldo2_on_ids))
+		return 0;
+
 	ret = axp_regulator_set(dev, gs, gs->eldo2_sel_reg, gs->eldo2_1p8v,
 				ELDO_CTRL_REG, gs->eldo2_ctrl_shift, false);
 	return ret;
Signature · 1 line
2.34.1