Skip to contentSkip to conversation

5.2.19-stable review

PATCH In review Greg Kroah-Hartman 10 participants 324 messages Oct 3, 2019 – Jun 23, 2020 Page 2 of 3

Subsystems (8)
lore
Share & tools
lore
author maintainer
#143 lore raw

PATCH 140/313 PM / devfreq: exynos-bus: Correct clock enable sequence

In reply to Greg Kroah-Hartman

From: Kamil Konieczny <[email protected]>

[ Upstream commit 2c2b20e0da89c76759ee28c6824413ab2fa3bfc6 ]

Regulators should be enabled before clocks to avoid h/w hang. This require change in exynos_bus_probe() to move exynos_bus_parse_of() after exynos_bus_parent_parse_of() and change in error handling. Similar change is needed in exynos_bus_exit() where clock should be disabled before regulators.

---
 drivers/devfreq/exynos-bus.c | 31 +++++++++++++++++--------------
 1 file changed, 17 insertions(+), 14 deletions(-)
drivers/devfreq/exynos-bus.c +17 -14
index d9f377912c104..7c06df8bd74fe 100644
@@ -191,11 +191,10 @@ static void exynos_bus_exit(struct device *dev)
 	if (ret < 0)
 		dev_warn(dev, "failed to disable the devfreq-event devices\n");
 
-	if (bus->regulator)
-		regulator_disable(bus->regulator);
-
 	dev_pm_opp_of_remove_table(dev);
 	clk_disable_unprepare(bus->clk);
+	if (bus->regulator)
+		regulator_disable(bus->regulator);
 }
 
 /*
@@ -383,6 +382,7 @@ static int exynos_bus_probe(struct platform_device *pdev)
 	struct exynos_bus *bus;
 	int ret, max_state;
 	unsigned long min_freq, max_freq;
+	bool passive = false;
 
 	if (!np) {
 		dev_err(dev, "failed to find devicetree node\n");
@@ -396,27 +396,27 @@ static int exynos_bus_probe(struct platform_device *pdev)
 	bus->dev = &pdev->dev;
 	platform_set_drvdata(pdev, bus);
 
-	/* Parse the device-tree to get the resource information */
-	ret = exynos_bus_parse_of(np, bus);
-	if (ret < 0)
-		return ret;
-
 	profile = devm_kzalloc(dev, sizeof(*profile), GFP_KERNEL);
-	if (!profile) {
-		ret = -ENOMEM;
-		goto err;
-	}
+	if (!profile)
+		return -ENOMEM;
 
 	node = of_parse_phandle(dev->of_node, "devfreq", 0);
 	if (node) {
 		of_node_put(node);
-		goto passive;
+		passive = true;
 	} else {
 		ret = exynos_bus_parent_parse_of(np, bus);
+		if (ret < 0)
+			return ret;
 	}
 
+	/* Parse the device-tree to get the resource information */
+	ret = exynos_bus_parse_of(np, bus);
 	if (ret < 0)
-		goto err;
+		goto err_reg;
+
+	if (passive)
+		goto passive;
 
 	/* Initialize the struct profile and governor data for parent device */
 	profile->polling_ms = 50;
@@ -507,6 +507,9 @@ static int exynos_bus_probe(struct platform_device *pdev)
 err:
 	dev_pm_opp_of_remove_table(dev);
 	clk_disable_unprepare(bus->clk);
+err_reg:
+	if (!passive)
+		regulator_disable(bus->regulator);
 
 	return ret;
 }
Signature · 1 line
2.20.1
author maintainer
#144 lore raw

PATCH 141/313 media: cec-notifier: clear cec_adap in cec_notifier_unregister

In reply to Greg Kroah-Hartman

From: Hans Verkuil <[email protected]>

[ Upstream commit 14d5511691e5290103bc480998bc322e68f139d4 ]

If cec_notifier_cec_adap_unregister() is called before cec_unregister_adapter() then everything is OK (and this is the case today). But if it is the other way around, then cec_notifier_unregister() is called first, and that doesn't set n->cec_adap to NULL.

So if e.g. cec_notifier_set_phys_addr() is called after cec_notifier_unregister() but before cec_unregister_adapter() then n->cec_adap points to an unregistered and likely deleted cec adapter. So just set n->cec_adap->notifier and n->cec_adap to NULL for rubustness.

Eventually cec_notifier_unregister will disappear and this will be simplified substantially.

---
 drivers/media/cec/cec-notifier.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/media/cec/cec-notifier.c +2 -0
index 9598c7778871a..c4aa27e0c4308 100644
@@ -124,6 +124,8 @@ void cec_notifier_unregister(struct cec_notifier *n)
 {
 	mutex_lock(&n->lock);
 	n->callback = NULL;
+	n->cec_adap->notifier = NULL;
+	n->cec_adap = NULL;
 	mutex_unlock(&n->lock);
 	cec_notifier_put(n);
 }
Signature · 1 line
2.20.1
author maintainer
#145 lore raw

PATCH 142/313 media: saa7146: add cleanup in hexium_attach()

In reply to Greg Kroah-Hartman

From: Wenwen Wang <[email protected]>

[ Upstream commit 42e64117d3b4a759013f77bbcf25ab6700e55de7 ]

If saa7146_register_device() fails, no cleanup is executed, leading to memory/resource leaks. To fix this issue, perform necessary cleanup work before returning the error.

---
 drivers/media/pci/saa7146/hexium_gemini.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/media/pci/saa7146/hexium_gemini.c +3 -0
index dca20a3d98e25..f962269306707 100644
@@ -292,6 +292,9 @@ static int hexium_attach(struct saa7146_dev *dev, struct saa7146_pci_extension_d
 	ret = saa7146_register_device(&hexium->video_dev, dev, "hexium gemini", VFL_TYPE_GRABBER);
 	if (ret < 0) {
 		pr_err("cannot register capture v4l2 device. skipping.\n");
+		saa7146_vv_release(dev);
+		i2c_del_adapter(&hexium->i2c_adapter);
+		kfree(hexium);
 		return ret;
 	}
 
Signature · 1 line
2.20.1
author maintainer
#146 lore raw

PATCH 143/313 media: cpia2_usb: fix memory leaks

In reply to Greg Kroah-Hartman

From: Wenwen Wang <[email protected]>

[ Upstream commit 1c770f0f52dca1a2323c594f01f5ec6f1dddc97f ]

In submit_urbs(), 'cam->sbuf[i].data' is allocated through kmalloc_array(). However, it is not deallocated if the following allocation for urbs fails. To fix this issue, free 'cam->sbuf[i].data' if usb_alloc_urb() fails.

---
 drivers/media/usb/cpia2/cpia2_usb.c | 4 ++++
 1 file changed, 4 insertions(+)
drivers/media/usb/cpia2/cpia2_usb.c +4 -0
index 17468f7d78ed2..3ab80a7b44985 100644
@@ -676,6 +676,10 @@ static int submit_urbs(struct camera_data *cam)
 		if (!urb) {
 			for (j = 0; j < i; j++)
 				usb_free_urb(cam->sbuf[j].urb);
+			for (j = 0; j < NUM_SBUF; j++) {
+				kfree(cam->sbuf[j].data);
+				cam->sbuf[j].data = NULL;
+			}
 			return -ENOMEM;
 		}
 
Signature · 1 line
2.20.1
author maintainer
#147 lore raw

PATCH 144/313 media: saa7134: fix terminology around saa7134_i2c_eeprom_md7134_gate()

In reply to Greg Kroah-Hartman

From: Maciej S. Szmigiero <[email protected]>

[ Upstream commit 9d802222a3405599d6e1984d9324cddf592ea1f4 ]

saa7134_i2c_eeprom_md7134_gate() function and the associated comment uses an inverted i2c gate open / closed terminology.
Let's fix this.

---
 drivers/media/pci/saa7134/saa7134-i2c.c | 12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)
drivers/media/pci/saa7134/saa7134-i2c.c +8 -4
index 493b1858815fb..04e85765373ec 100644
@@ -342,7 +342,11 @@ static const struct i2c_client saa7134_client_template = {
 
 /* ----------------------------------------------------------- */
 
-/* On Medion 7134 reading EEPROM needs DVB-T demod i2c gate open */
+/*
+ * On Medion 7134 reading the SAA7134 chip config EEPROM needs DVB-T
+ * demod i2c gate closed due to an address clash between this EEPROM
+ * and the demod one.
+ */
 static void saa7134_i2c_eeprom_md7134_gate(struct saa7134_dev *dev)
 {
 	u8 subaddr = 0x7, dmdregval;
@@ -359,14 +363,14 @@ static void saa7134_i2c_eeprom_md7134_gate(struct saa7134_dev *dev)
 
 	ret = i2c_transfer(&dev->i2c_adap, i2cgatemsg_r, 2);
 	if ((ret == 2) && (dmdregval & 0x2)) {
-		pr_debug("%s: DVB-T demod i2c gate was left closed\n",
+		pr_debug("%s: DVB-T demod i2c gate was left open\n",
 			 dev->name);
 
 		data[0] = subaddr;
 		data[1] = (dmdregval & ~0x2);
 		if (i2c_transfer(&dev->i2c_adap, i2cgatemsg_w, 1) != 1)
-			pr_err("%s: EEPROM i2c gate open failure\n",
-			  dev->name);
+			pr_err("%s: EEPROM i2c gate close failure\n",
+			       dev->name);
 	}
 }
 
Signature · 1 line
2.20.1
author maintainer
#148 lore raw

PATCH 145/313 perf trace beauty ioctl: Fix off-by-one error in cmd->string table

In reply to Greg Kroah-Hartman

From: Benjamin Peterson <[email protected]>

[ Upstream commit b92675f4a9c02dd78052645597dac9e270679ddf ]

While tracing a program that calls isatty(3), I noticed that strace reported TCGETS for the request argument of the underlying ioctl(2) syscall while perf trace reported TCSETS. strace is corrrect. The bug in perf was due to the tty ioctl beauty table starting at 0x5400 rather than 0x5401.

Committer testing:

Using augmented_raw_syscalls.o and settings to make 'perf trace' use strace formatting, i.e. with this in ~/.perfconfig

  # cat ~/.perfconfig
  [trace]
	add_events = /home/acme/git/linux/tools/perf/examples/bpf/augmented_raw_syscalls.c
	show_zeros = yes
	show_duration = no
	no_inherit = yes
	show_timestamp = no
	show_arg_names = no
	args_alignment = 40
	show_prefix = yes
  # strace -e ioctl stty > /dev/null
  ioctl(0, TCGETS, {B38400 opost isig icanon echo ...}) = 0
  ioctl(1, TIOCGWINSZ, 0x7fff8a9b0860)    = -1 ENOTTY (Inappropriate ioctl for device)
  ioctl(1, TCGETS, 0x7fff8a9b0540)        = -1 ENOTTY (Inappropriate ioctl for device)
  +++ exited with 0 +++
  #

Before:

  # perf trace -e ioctl stty > /dev/null
  ioctl(0, TCSETS, 0x7fff2cf79f20)        = 0
  ioctl(1, TIOCSWINSZ, 0x7fff2cf79f40)    = -1 ENOTTY (Inappropriate ioctl for device)
  ioctl(1, TCSETS, 0x7fff2cf79c20)        = -1 ENOTTY (Inappropriate ioctl for device)
  #

After:

  # perf trace -e ioctl stty > /dev/null
  ioctl(0, TCGETS, 0x7ffed0763920)        = 0
  ioctl(1, TIOCGWINSZ, 0x7ffed0763940)    = -1 ENOTTY (Inappropriate ioctl for device)
  ioctl(1, TCGETS, 0x7ffed0763620)        = -1 ENOTTY (Inappropriate ioctl for device)
  #
Cc Alexander Shishkin <[email protected]>
Cc Jiri Olsa <[email protected]>
Cc Namhyung Kim <[email protected]>
Cc Peter Zijlstra <[email protected]>
Fixes 1cc47f2d46206d67285aea0ca7e8450af571da13 ("perf trace beauty ioctl: Improve 'cmd' beautifier")
---
 tools/perf/trace/beauty/ioctl.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
tools/perf/trace/beauty/ioctl.c +1 -1
index 52242fa4072b0..e19eb6ea361d7 100644
@@ -21,7 +21,7 @@
 static size_t ioctl__scnprintf_tty_cmd(int nr, int dir, char *bf, size_t size)
 {
 	static const char *ioctl_tty_cmd[] = {
-	"TCGETS", "TCSETS", "TCSETSW", "TCSETSF", "TCGETA", "TCSETA", "TCSETAW",
+	[_IOC_NR(TCGETS)] = "TCGETS", "TCSETS", "TCSETSW", "TCSETSF", "TCGETA", "TCSETA", "TCSETAW",
 	"TCSETAF", "TCSBRK", "TCXONC", "TCFLSH", "TIOCEXCL", "TIOCNXCL", "TIOCSCTTY",
 	"TIOCGPGRP", "TIOCSPGRP", "TIOCOUTQ", "TIOCSTI", "TIOCGWINSZ", "TIOCSWINSZ",
 	"TIOCMGET", "TIOCMBIS", "TIOCMBIC", "TIOCMSET", "TIOCGSOFTCAR", "TIOCSSOFTCAR",
Signature · 1 line
2.20.1
author maintainer
#149 lore raw

PATCH 146/313 perf report: Fix --ns time sort key output

In reply to Greg Kroah-Hartman

From: Andi Kleen <[email protected]>

[ Upstream commit 3dab6ac080dcd7f71cb9ceb84ad7dafecd6f7c07 ]

If the user specified --ns, the column to print the sort time stamp wasn't wide enough to actually print the full nanoseconds.

Widen the time key column width when --ns is specified.

Before:

  % perf record -a sleep 1
  % perf report --sort time,overhead,symbol --stdio --ns
  ...
       2.39%  187851.10000  [k] smp_call_function_single   -      -
       1.53%  187851.10000  [k] intel_idle                 -      -
       0.59%  187851.10000  [.] __wcscmp_ifunc             -      -
       0.33%  187851.10000  [.] 0000000000000000           -      -
       0.28%  187851.10000  [k] cpuidle_enter_state        -      -

After:

  % perf report --sort time,overhead,symbol --stdio --ns
  ...
       2.39%  187851.100000000  [k] smp_call_function_single   -      -
       1.53%  187851.100000000  [k] intel_idle                 -      -
       0.59%  187851.100000000  [.] __wcscmp_ifunc             -      -
       0.33%  187851.100000000  [.] 0000000000000000           -      -
       0.28%  187851.100000000  [k] cpuidle_enter_state        -      -
---
 tools/perf/util/hist.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
tools/perf/util/hist.c +4 -1
index 7ace7a10054d8..966c248d6a3a1 100644
@@ -193,7 +193,10 @@ void hists__calc_col_len(struct hists *hists, struct hist_entry *h)
 	hists__new_col_len(hists, HISTC_MEM_LVL, 21 + 3);
 	hists__new_col_len(hists, HISTC_LOCAL_WEIGHT, 12);
 	hists__new_col_len(hists, HISTC_GLOBAL_WEIGHT, 12);
-	hists__new_col_len(hists, HISTC_TIME, 12);
+	if (symbol_conf.nanosecs)
+		hists__new_col_len(hists, HISTC_TIME, 16);
+	else
+		hists__new_col_len(hists, HISTC_TIME, 12);
 
 	if (h->srcline) {
 		len = MAX(strlen(h->srcline), strlen(sort_srcline.se_header));
Signature · 1 line
2.20.1
author maintainer
#150 lore raw

PATCH 147/313 perf script: Fix memory leaks in list_scripts()

In reply to Greg Kroah-Hartman

From: Gustavo A. R. Silva <[email protected]>

[ Upstream commit 3b4acbb92dbda4829e021e5c6d5410658849fa1c ]

In case memory resources for *buf* and *paths* were allocated, jump to *out* and release them before return.

Cc Alexander Shishkin <[email protected]>
Cc Andi Kleen <[email protected]>
Cc Gustavo A. R. Silva <[email protected]>
Cc Jiri Olsa <[email protected]>
Cc Namhyung Kim <[email protected]>
Cc Peter Zijlstra <[email protected]>

Addresses-Coverity-ID: 1444328 ("Resource leak")

Fixes 6f3da20e151f ("perf report: Support builtin perf script in scripts menu")
---
 tools/perf/ui/browsers/scripts.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
tools/perf/ui/browsers/scripts.c +4 -2
index 27cf3ab88d13f..f4edb18f67ec9 100644
@@ -131,8 +131,10 @@ static int list_scripts(char *script_name, bool *custom,
 		int key = ui_browser__input_window("perf script command",
 				"Enter perf script command line (without perf script prefix)",
 				script_args, "", 0);
-		if (key != K_ENTER)
-			return -1;
+		if (key != K_ENTER) {
+			ret = -1;
+			goto out;
+		}
 		sprintf(script_name, "%s script %s", perf, script_args);
 	} else if (choice < num + max_std) {
 		strcpy(script_name, paths[choice]);
Signature · 1 line
2.20.1
author maintainer
#151 lore raw

PATCH 148/313 media: aspeed-video: address a protential usage of an unitialized var

In reply to Greg Kroah-Hartman

From: Mauro Carvalho Chehab <[email protected]>

[ Upstream commit 31b8b0bd6e55c3ea5a08bb8141fa5d3c90600e3b ]

While this might not occur in practice, if the device is doing the right thing, it would be teoretically be possible to have both hsync_counter and vsync_counter negatives.

If this ever happen, ctrl will be undefined, but the driver will still call:

	aspeed_video_update(video, VE_CTRL, 0, ctrl);

Change the code to prevent this to happen.

This was warned by cppcheck:

[drivers/media/platform/aspeed-video.c:653]: (error) Uninitialized variable: ctrl

---
 drivers/media/platform/aspeed-video.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)
drivers/media/platform/aspeed-video.c +3 -2
index de0f192afa8b1..388c32a11345d 100644
@@ -632,7 +632,7 @@ static void aspeed_video_check_and_set_polarity(struct aspeed_video *video)
 	}
 
 	if (hsync_counter < 0 || vsync_counter < 0) {
-		u32 ctrl;
+		u32 ctrl = 0;
 
 		if (hsync_counter < 0) {
 			ctrl = VE_CTRL_HSYNC_POL;
@@ -652,7 +652,8 @@ static void aspeed_video_check_and_set_polarity(struct aspeed_video *video)
 				V4L2_DV_VSYNC_POS_POL;
 		}
 
-		aspeed_video_update(video, VE_CTRL, 0, ctrl);
+		if (ctrl)
+			aspeed_video_update(video, VE_CTRL, 0, ctrl);
 	}
 }
 
Signature · 1 line
2.20.1
author maintainer
#152 lore raw

PATCH 149/313 media: ov9650: add a sanity check

In reply to Greg Kroah-Hartman

From: Mauro Carvalho Chehab <[email protected]>

[ Upstream commit 093347abc7a4e0490e3c962ecbde2dc272a8f708 ]

As pointed by cppcheck:

[drivers/media/i2c/ov9650.c:706]: (error) Shifting by a negative value is undefined behaviour [drivers/media/i2c/ov9650.c:707]: (error) Shifting by a negative value is undefined behaviour [drivers/media/i2c/ov9650.c:721]: (error) Shifting by a negative value is undefined behaviour

Prevent mangling with gains with invalid values.

As pointed by Sylvester, this should never happen in practice, as min value of V4L2_CID_GAIN control is 16 (gain is always >= 16 and m is always >= 0), but it is too hard for a static analyzer to get this, as the logic with validates control min/max is elsewhere inside V4L2 core.

---
 drivers/media/i2c/ov9650.c | 5 +++++
 1 file changed, 5 insertions(+)
drivers/media/i2c/ov9650.c +5 -0
index 30ab2225fbd0c..b350f5c1a9890 100644
@@ -703,6 +703,11 @@ static int ov965x_set_gain(struct ov965x *ov965x, int auto_gain)
 		for (m = 6; m >= 0; m--)
 			if (gain >= (1 << m) * 16)
 				break;
+
+		/* Sanity check: don't adjust the gain with a negative value */
+		if (m < 0)
+			return -EINVAL;
+
 		rgain = (gain - ((1 << m) * 16)) / (1 << m);
 		rgain |= (((1 << m) - 1) << 4);
 
Signature · 1 line
2.20.1
author maintainer
#153 lore raw

PATCH 150/313 leds: lm3532: Fixes for the driver for stability

In reply to Greg Kroah-Hartman

From: Dan Murphy <[email protected]>

[ Upstream commit 6559ac32998248182572e1ccae79dc2eb40ac7c6 ]

Fixed misspelled words, added error check during probe on the init of the registers, and fixed ALS/I2C control mode.

Fixes bc1b8492c764 ("leds: lm3532: Introduce the lm3532 LED driver")
---
 drivers/leds/leds-lm3532.c | 17 +++++++++++++----
 1 file changed, 13 insertions(+), 4 deletions(-)
drivers/leds/leds-lm3532.c +13 -4
index 180895b83b888..e55a64847fe2f 100644
@@ -40,7 +40,7 @@
 #define LM3532_REG_ZN_3_LO	0x67
 #define LM3532_REG_MAX		0x7e
 
-/* Contorl Enable */
+/* Control Enable */
 #define LM3532_CTRL_A_ENABLE	BIT(0)
 #define LM3532_CTRL_B_ENABLE	BIT(1)
 #define LM3532_CTRL_C_ENABLE	BIT(2)
@@ -302,7 +302,7 @@ static int lm3532_led_disable(struct lm3532_led *led_data)
 	int ret;
 
 	ret = regmap_update_bits(led_data->priv->regmap, LM3532_REG_ENABLE,
-					 ctrl_en_val, ~ctrl_en_val);
+					 ctrl_en_val, 0);
 	if (ret) {
 		dev_err(led_data->priv->dev, "Failed to set ctrl:%d\n", ret);
 		return ret;
@@ -321,7 +321,7 @@ static int lm3532_brightness_set(struct led_classdev *led_cdev,
 
 	mutex_lock(&led->priv->lock);
 
-	if (led->mode == LM3532_BL_MODE_ALS) {
+	if (led->mode == LM3532_ALS_CTRL) {
 		if (brt_val > LED_OFF)
 			ret = lm3532_led_enable(led);
 		else
@@ -542,11 +542,14 @@ static int lm3532_parse_node(struct lm3532_data *priv)
 		}
 
 		if (led->mode == LM3532_BL_MODE_ALS) {
+			led->mode = LM3532_ALS_CTRL;
 			ret = lm3532_parse_als(priv);
 			if (ret)
 				dev_err(&priv->client->dev, "Failed to parse als\n");
 			else
 				lm3532_als_configure(priv, led);
+		} else {
+			led->mode = LM3532_I2C_CTRL;
 		}
 
 		led->num_leds = fwnode_property_read_u32_array(child,
@@ -590,7 +593,13 @@ static int lm3532_parse_node(struct lm3532_data *priv)
 			goto child_out;
 		}
 
-		lm3532_init_registers(led);
+		ret = lm3532_init_registers(led);
+		if (ret) {
+			dev_err(&priv->client->dev, "register init err: %d\n",
+				ret);
+			fwnode_handle_put(child);
+			goto child_out;
+		}
 
 		i++;
 	}
Signature · 1 line
2.20.1
author maintainer
#154 lore raw

PATCH 151/313 ASoC: es8316: fix headphone mixer volume table

In reply to Greg Kroah-Hartman

From: Katsuhiro Suzuki <[email protected]>

[ Upstream commit f972d02fee2496024cfd6f59021c9d89d54922a6 ]

This patch fix setting table of Headphone mixer volume. Current code uses 4 ... 7 values but these values are prohibited.

Correct settings are the following:
0000 -12dB
0001 -10.5dB
0010 -9dB
0011 -7.5dB
0100 -6dB
1000 -4.5dB
1001 -3dB
1010 -1.5dB
1011 0dB

---
 sound/soc/codecs/es8316.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)
sound/soc/codecs/es8316.c +5 -2
index 6db002cc20582..96d04896193f2 100644
@@ -51,7 +51,10 @@ static const SNDRV_CTL_TLVD_DECLARE_DB_SCALE(adc_vol_tlv, -9600, 50, 1);
 static const SNDRV_CTL_TLVD_DECLARE_DB_SCALE(alc_max_gain_tlv, -650, 150, 0);
 static const SNDRV_CTL_TLVD_DECLARE_DB_SCALE(alc_min_gain_tlv, -1200, 150, 0);
 static const SNDRV_CTL_TLVD_DECLARE_DB_SCALE(alc_target_tlv, -1650, 150, 0);
-static const SNDRV_CTL_TLVD_DECLARE_DB_SCALE(hpmixer_gain_tlv, -1200, 150, 0);
+static const SNDRV_CTL_TLVD_DECLARE_DB_RANGE(hpmixer_gain_tlv,
+	0, 4, TLV_DB_SCALE_ITEM(-1200, 150, 0),
+	8, 11, TLV_DB_SCALE_ITEM(-450, 150, 0),
+);
 
 static const SNDRV_CTL_TLVD_DECLARE_DB_RANGE(adc_pga_gain_tlv,
 	0, 0, TLV_DB_SCALE_ITEM(-350, 0, 0),
@@ -89,7 +92,7 @@ static const struct snd_kcontrol_new es8316_snd_controls[] = {
 	SOC_DOUBLE_TLV("Headphone Playback Volume", ES8316_CPHP_ICAL_VOL,
 		       4, 0, 3, 1, hpout_vol_tlv),
 	SOC_DOUBLE_TLV("Headphone Mixer Volume", ES8316_HPMIX_VOL,
-		       0, 4, 7, 0, hpmixer_gain_tlv),
+		       0, 4, 11, 0, hpmixer_gain_tlv),
 
 	SOC_ENUM("Playback Polarity", dacpol),
 	SOC_DOUBLE_R_TLV("DAC Playback Volume", ES8316_DAC_VOLL,
Signature · 1 line
2.20.1
author maintainer
#155 lore raw

PATCH 152/313 ACPI / CPPC: do not require the _PSD method

In reply to Greg Kroah-Hartman

From: Al Stone <[email protected]>

[ Upstream commit 4c4cdc4c63853fee48c02e25c8605fb65a6c9924 ]

According to the ACPI 6.3 specification, the _PSD method is optional when using CPPC. The underlying assumption is that each CPU can change frequency independently from all other CPUs; _PSD is provided to tell the OS that some processors can NOT do that.

However, the acpi_get_psd() function returns ENODEV if there is no _PSD method present, or an ACPI error status if an error occurs when evaluating _PSD, if present. This makes _PSD mandatory when using CPPC, in violation of the specification, and only on Linux.

This has forced some firmware writers to provide a dummy _PSD, even though it is irrelevant, but only because Linux requires it; other OSPMs follow the spec. We really do not want to have OS specific ACPI tables, though.

So, correct acpi_get_psd() so that it does not return an error if there is no _PSD method present, but does return a failure when the method can not be executed properly. This allows _PSD to be optional as it should be.

---
 drivers/acpi/cppc_acpi.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/acpi/cppc_acpi.c +4 -2
index 15f103d7532b0..3b2525908dd8c 100644
@@ -365,8 +365,10 @@ static int acpi_get_psd(struct cpc_desc *cpc_ptr, acpi_handle handle)
 	union acpi_object  *psd = NULL;
 	struct acpi_psd_package *pdomain;
 
-	status = acpi_evaluate_object_typed(handle, "_PSD", NULL, &buffer,
-			ACPI_TYPE_PACKAGE);
+	status = acpi_evaluate_object_typed(handle, "_PSD", NULL,
+					    &buffer, ACPI_TYPE_PACKAGE);
+	if (status == AE_NOT_FOUND)	/* _PSD is optional */
+		return 0;
 	if (ACPI_FAILURE(status))
 		return -ENODEV;
 
Signature · 1 line
2.20.1
author maintainer
#156 lore raw

PATCH 153/313 sched/cpufreq: Align trace event behavior of fast switching

In reply to Greg Kroah-Hartman

From: Douglas RAILLARD <[email protected]>

[ Upstream commit 77c84dd1881d0f0176cb678d770bfbda26c54390 ]

Fast switching path only emits an event for the CPU of interest, whereas the regular path emits an event for all the CPUs that had their frequency changed, i.e. all the CPUs sharing the same policy.

With the current behavior, looking at cpu_frequency event for a given CPU that is using the fast switching path will not give the correct frequency signal.

---
 kernel/sched/cpufreq_schedutil.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)
kernel/sched/cpufreq_schedutil.c +6 -1
index ae3ec77bb92f6..e139b54716b4a 100644
@@ -117,6 +117,7 @@ static void sugov_fast_switch(struct sugov_policy *sg_policy, u64 time,
 			      unsigned int next_freq)
 {
 	struct cpufreq_policy *policy = sg_policy->policy;
+	int cpu;
 
 	if (!sugov_update_next_freq(sg_policy, time, next_freq))
 		return;
@@ -126,7 +127,11 @@ static void sugov_fast_switch(struct sugov_policy *sg_policy, u64 time,
 		return;
 
 	policy->cur = next_freq;
-	trace_cpu_frequency(next_freq, smp_processor_id());
+
+	if (trace_cpu_frequency_enabled()) {
+		for_each_cpu(cpu, policy->cpus)
+			trace_cpu_frequency(next_freq, cpu);
+	}
 }
 
 static void sugov_deferred_update(struct sugov_policy *sg_policy, u64 time,
Signature · 1 line
2.20.1
author maintainer
#157 lore raw

PATCH 154/313 x86/apic/vector: Warn when vector space exhaustion breaks affinity

In reply to Greg Kroah-Hartman

From: Neil Horman <[email protected]>

[ Upstream commit 743dac494d61d991967ebcfab92e4f80dc7583b3 ]

On x86, CPUs are limited in the number of interrupts they can have affined to them as they only support 256 interrupt vectors per CPU. 32 vectors are reserved for the CPU and the kernel reserves another 22 for internal purposes. That leaves 202 vectors for assignement to devices.

When an interrupt is set up or the affinity is changed by the kernel or the administrator, the vector assignment code attempts to honor the requested affinity mask. If the vector space on the CPUs in that affinity mask is exhausted the code falls back to a wider set of CPUs and assigns a vector on a CPU outside of the requested affinity mask silently.

While the effective affinity is reflected in the corresponding /proc/irq/$N/effective_affinity* files the silent breakage of the requested affinity can lead to unexpected behaviour for administrators.

Add a pr_warn() when this happens so that adminstrators get at least informed about it in the syslog.

[ tglx: Massaged changelog and made the pr_warn() more informative ]

---
 arch/x86/kernel/apic/vector.c | 11 +++++++++++
 1 file changed, 11 insertions(+)
arch/x86/kernel/apic/vector.c +11 -0
index fdacb864c3dd4..2c5676b0a6e7f 100644
@@ -398,6 +398,17 @@ static int activate_reserved(struct irq_data *irqd)
 		if (!irqd_can_reserve(irqd))
 			apicd->can_reserve = false;
 	}
+
+	/*
+	 * Check to ensure that the effective affinity mask is a subset
+	 * the user supplied affinity mask, and warn the user if it is not
+	 */
+	if (!cpumask_subset(irq_data_get_effective_affinity_mask(irqd),
+			    irq_data_get_affinity_mask(irqd))) {
+		pr_warn("irq %u: Affinity broken due to vector space exhaustion.\n",
+			irqd->irq);
+	}
+
 	return ret;
 }
 
Signature · 1 line
2.20.1
author maintainer
#158 lore raw

PATCH 155/313 arm64: kpti: ensure patched kernel text is fetched from PoU

In reply to Greg Kroah-Hartman

From: Mark Rutland <[email protected]>

[ Upstream commit f32c7a8e45105bd0af76872bf6eef0438ff12fb2 ]

While the MMUs is disabled, I-cache speculation can result in instructions being fetched from the PoC. During boot we may patch instructions (e.g. for alternatives and jump labels), and these may be dirty at the PoU (and stale at the PoC).

Thus, while the MMU is disabled in the KPTI pagetable fixup code we may load stale instructions into the I-cache, potentially leading to subsequent crashes when executing regions of code which have been modified at runtime.

Similarly to commit:

8ec41987436d566f ("arm64: mm: ensure patched kernel text is fetched from PoU")

... we can invalidate the I-cache after enabling the MMU to prevent such issues.

The KPTI pagetable fixup code itself should be clean to the PoC per the boot protocol, so no maintenance is required for this code.

---
 arch/arm64/mm/proc.S | 9 +++++++++
 1 file changed, 9 insertions(+)
arch/arm64/mm/proc.S +9 -0
index 7dbf2be470f6c..28a8f7b87ff06 100644
@@ -286,6 +286,15 @@ skip_pgd:
 	msr	sctlr_el1, x18
 	isb
 
+	/*
+	 * Invalidate the local I-cache so that any instructions fetched
+	 * speculatively from the PoC are discarded, since they may have
+	 * been dynamically patched at the PoU.
+	 */
+	ic	iallu
+	dsb	nsh
+	isb
+
 	/* Set the flag to zero to indicate that we're all done */
 	str	wzr, [flag_ptr]
 	ret
Signature · 1 line
2.20.1
author maintainer
#159 lore raw

PATCH 156/313 perf evlist: Use unshare(CLONE_FS) in sb threads to let setns(CLONE_NEWNS) work

In reply to Greg Kroah-Hartman

From: Arnaldo Carvalho de Melo <[email protected]>

[ Upstream commit b397f8468fa27f08b83b348ffa56a226f72453af ]

When we started using a thread to catch the PERF_RECORD_BPF_EVENT meta data events to then ask the kernel for further info (BTF, etc) for BPF programs shortly after they get loaded, we forgot to use unshare(CLONE_FS) as was done in:

868a832918f6 ("perf top: Support lookup of symbols in other mount namespaces.")

Do it so that we can enter the namespaces to read the build-ids at the end of a 'perf record' session for the DSOs that had hits.

Before:

Starting a 'stress-ng --cpus 8' inside a container and then, outside the container running:

  # perf record -a --namespaces sleep 5
  # perf buildid-list | grep stress-ng
  #

We would end up with a 'perf.data' file that had no entry in its build-id table for the /usr/bin/stress-ng binary inside the container that got tons of PERF_RECORD_SAMPLEs.

After:

  # perf buildid-list | grep stress-ng
  f2ed02c68341183a124b9b0f6e2e6c493c465b29 /usr/bin/stress-ng
  #

Then its just a matter of making sure that that binary debuginfo package gets available in a place that 'perf report' will look at build-id keyed ELF files, which, in my case, on a f30 notebook, was a matter of installing the debuginfo file for the distro used in the container, fedora 31:

  # rpm -ivh http://fedora.c3sl.ufpr.br/linux/development/31/Everything/x86_64/debug/tree/Packages/s/stress-ng-debuginfo-0.07.29-10.fc31.x86_64.rpm

Then, because perf currently looks for those debuginfo files (richer ELF symtab) inside that namespace (look at the setns calls):

openat(AT_FDCWD, "/proc/self/ns/mnt", O_RDONLY) = 137
openat(AT_FDCWD, "/proc/13169/ns/mnt", O_RDONLY) = 139
setns(139, CLONE_NEWNS) = 0
stat("/usr/bin/stress-ng", {st_mode=S_IFREG|0755, st_size=3065416, ...}) = 0
openat(AT_FDCWD, "/usr/bin/stress-ng", O_RDONLY) = 140
fcntl(140, F_GETFD) = 0
fstat(140, {st_mode=S_IFREG|0755, st_size=3065416, ...}) = 0
mmap(NULL, 3065416, PROT_READ, MAP_PRIVATE, 140, 0) = 0x7ff2fdc5b000
munmap(0x7ff2fdc5b000, 3065416) = 0
close(140) = 0
stat("stress-ng-0.07.29-10.fc31.x86_64.debug", 0x7fff45d71260) = -1 ENOENT (No such file or directory) stat("/usr/bin/stress-ng-0.07.29-10.fc31.x86_64.debug", 0x7fff45d71260) = -1 ENOENT (No such file or directory) stat("/usr/bin/.debug/stress-ng-0.07.29-10.fc31.x86_64.debug", 0x7fff45d71260) = -1 ENOENT (No such file or directory) stat("/usr/lib/debug/usr/bin/stress-ng-0.07.29-10.fc31.x86_64.debug", 0x7fff45d71260) = -1 ENOENT (No such file or directory) stat("/root/.debug/.build-id/f2/ed02c68341183a124b9b0f6e2e6c493c465b29", 0x7fff45d711e0) = -1 ENOENT (No such file or directory)

To only then go back to the "host" namespace to look just in the users's ~/.debug cache:

  setns(137, CLONE_NEWNS)                 = 0
  chdir("/root")                          = 0
  close(137)                              = 0
  close(139)                              = 0

stat("/root/.debug/.build-id/f2/ed02c68341183a124b9b0f6e2e6c493c465b29/elf", 0x7fff45d732e0) = -1 ENOENT (No such file or directory)

It continues to fail to resolve symbols:

  # perf report | grep stress-ng | head -5
     9.50%  stress-ng-cpu    stress-ng    [.] 0x0000000000021ac1
     8.58%  stress-ng-cpu    stress-ng    [.] 0x0000000000021ab4
     8.51%  stress-ng-cpu    stress-ng    [.] 0x0000000000021489
     7.17%  stress-ng-cpu    stress-ng    [.] 0x00000000000219b6
     3.93%  stress-ng-cpu    stress-ng    [.] 0x0000000000021478
  #

To overcome that we use:

  # perf buildid-cache -v --add /usr/lib/debug/usr/bin/stress-ng-0.07.29-10.fc31.x86_64.debug
  Adding f2ed02c68341183a124b9b0f6e2e6c493c465b29 /usr/lib/debug/usr/bin/stress-ng-0.07.29-10.fc31.x86_64.debug: Ok
  #
  # ls -la /root/.debug/.build-id/f2/ed02c68341183a124b9b0f6e2e6c493c465b29/elf
  -rw-r--r--. 3 root root 2401184 Jul 27 07:03 /root/.debug/.build-id/f2/ed02c68341183a124b9b0f6e2e6c493c465b29/elf
  # file /root/.debug/.build-id/f2/ed02c68341183a124b9b0f6e2e6c493c465b29/elf
  /root/.debug/.build-id/f2/ed02c68341183a124b9b0f6e2e6c493c465b29/elf: ELF 64-bit LSB shared object, x86-64, version 1 (SYSV), dynamically linked, interpreter \004, BuildID[sha1]=f2ed02c68341183a124b9b0f6e2e6c493c465b29, for GNU/Linux 3.2.0, with debug_info, not stripped, too many notes (256)
  #

Now it finally works:

  # perf report | grep stress-ng | head -5
    23.59%  stress-ng-cpu    stress-ng    [.] ackermann
    23.33%  stress-ng-cpu    stress-ng    [.] is_prime
    17.36%  stress-ng-cpu    stress-ng    [.] stress_cpu_sieve
     6.08%  stress-ng-cpu    stress-ng    [.] stress_cpu_correlate
     3.55%  stress-ng-cpu    stress-ng    [.] queens_try
  #

I'll make sure that it looks for the build-id keyed files in both the "host" namespace (the namespace the user running 'perf record' was a the time of the recording) and in the container namespace, as it shouldn't matter where a content based key lookup finds the ELF file to use in resolving symbols, etc.

Cc Alexander Shishkin <[email protected]>
Cc Alexei Starovoitov <[email protected]>
Cc Brendan Gregg <[email protected]>
Cc Daniel Borkmann <[email protected]>
Cc Krister Johansen <[email protected]>
Cc Namhyung Kim <[email protected]>
Cc Peter Zijlstra <[email protected]>
Cc Song Liu <[email protected]>
Cc Stanislav Fomichev <[email protected]>
Cc Thomas-Mich Richter <[email protected]>
Fixes 657ee5531903 ("perf evlist: Introduce side band thread")
---
 tools/perf/util/evlist.c | 9 +++++++++
 1 file changed, 9 insertions(+)
tools/perf/util/evlist.c +9 -0
index a474ede17cd6c..001bb444d2056 100644
@@ -21,6 +21,7 @@
 #include "bpf-event.h"
 #include <signal.h>
 #include <unistd.h>
+#include <sched.h>
 
 #include "parse-events.h"
 #include <subcmd/parse-options.h>
@@ -1870,6 +1871,14 @@ static void *perf_evlist__poll_thread(void *arg)
 	struct perf_evlist *evlist = arg;
 	bool draining = false;
 	int i, done = 0;
+	/*
+	 * In order to read symbols from other namespaces perf to needs to call
+	 * setns(2).  This isn't permitted if the struct_fs has multiple users.
+	 * unshare(2) the fs so that we may continue to setns into namespaces
+	 * that we're observing when, for instance, reading the build-ids at
+	 * the end of a 'perf record' session.
+	 */
+	unshare(CLONE_FS);
 
 	while (!done) {
 		bool got_data = false;
Signature · 1 line
2.20.1
author maintainer
#160 lore raw

PATCH 157/313 x86/mm/pti: Do not invoke PTI functions when PTI is disabled

In reply to Greg Kroah-Hartman

From: Thomas Gleixner <[email protected]>

[ Upstream commit 990784b57731192b7d90c8d4049e6318d81e887d ]

When PTI is disabled at boot time either because the CPU is not affected or PTI has been disabled on the command line, the boot code still calls into pti_finalize() which then unconditionally invokes:

pti_clone_entry_text()
pti_clone_kernel_text()

pti_clone_kernel_text() was called unconditionally before the 32bit support was added and 32bit added the call to pti_clone_entry_text().

The call has no side effects as cloning the page tables into the available second one, which was allocated for PTI does not create damage. But it does not make sense either and in case that this functionality would be extended later this might actually lead to hard to diagnose issues.

Neither function should be called when PTI is runtime disabled. Make the invocation conditional.

---
 arch/x86/mm/pti.c | 2 ++
 1 file changed, 2 insertions(+)
arch/x86/mm/pti.c +2 -0
index b196524759ec5..ba22b50f4eca2 100644
@@ -666,6 +666,8 @@ void __init pti_init(void)
  */
 void pti_finalize(void)
 {
+	if (!boot_cpu_has(X86_FEATURE_PTI))
+		return;
 	/*
 	 * We need to clone everything (again) that maps parts of the
 	 * kernel image.
Signature · 1 line
2.20.1
author maintainer
#161 lore raw

PATCH 158/313 ASoC: fsl_ssi: Fix clock control issue in master mode

In reply to Greg Kroah-Hartman

From: Shengjiu Wang <[email protected]>

[ Upstream commit 696d05225cebffd172008d212657be90e823eac0 ]

The test case is
arecord -Dhw:0 -d 10 -f S16_LE -r 48000 -c 2 temp.wav & aplay -Dhw:0 -d 30 -f S16_LE -r 48000 -c 2 test.wav

There will be error after end of arecord:
aplay: pcm_write:2051: write error: Input/output error

Capture and Playback work in parallel in master mode, one substream stops, the other substream is impacted, the reason is that clock is disabled wrongly.

The clock's reference count is not increased when second substream starts, the hw_param() function returns in the beginning because first substream is enabled, then in end of first substream, the hw_free() disables the clock.

This patch is to move the clock enablement to the place before checking of the device enablement in hw_param().

---
 sound/soc/fsl/fsl_ssi.c | 18 +++++++++---------
 1 file changed, 9 insertions(+), 9 deletions(-)
sound/soc/fsl/fsl_ssi.c +9 -9
index 09b2967befd96..d83be26d64467 100644
@@ -799,15 +799,6 @@ static int fsl_ssi_hw_params(struct snd_pcm_substream *substream,
 	u32 wl = SSI_SxCCR_WL(sample_size);
 	int ret;
 
-	/*
-	 * SSI is properly configured if it is enabled and running in
-	 * the synchronous mode; Note that AC97 mode is an exception
-	 * that should set separate configurations for STCCR and SRCCR
-	 * despite running in the synchronous mode.
-	 */
-	if (ssi->streams && ssi->synchronous)
-		return 0;
-
 	if (fsl_ssi_is_i2s_master(ssi)) {
 		ret = fsl_ssi_set_bclk(substream, dai, hw_params);
 		if (ret)
@@ -823,6 +814,15 @@ static int fsl_ssi_hw_params(struct snd_pcm_substream *substream,
 		}
 	}
 
+	/*
+	 * SSI is properly configured if it is enabled and running in
+	 * the synchronous mode; Note that AC97 mode is an exception
+	 * that should set separate configurations for STCCR and SRCCR
+	 * despite running in the synchronous mode.
+	 */
+	if (ssi->streams && ssi->synchronous)
+		return 0;
+
 	if (!fsl_ssi_is_ac97(ssi)) {
 		/*
 		 * Keep the ssi->i2s_net intact while having a local variable
Signature · 1 line
2.20.1
author maintainer
#162 lore raw

PATCH 159/313 x86/mm/pti: Handle unaligned address gracefully in pti_clone_pagetable()

In reply to Greg Kroah-Hartman

From: Song Liu <[email protected]>

[ Upstream commit 825d0b73cd7526b0bb186798583fae810091cbac ]

pti_clone_pmds() assumes that the supplied address is either:

- properly PUD/PMD aligned
or
- the address is actually mapped which means that independently of the mapping level (PUD/PMD/PTE) the next higher mapping exists.

If that's not the case the unaligned address can be incremented by PUD or PMD size incorrectly. All callers supply mapped and/or aligned addresses, but for the sake of robustness it's better to handle that case properly and to emit a warning.

[ tglx: Rewrote changelog and added WARN_ON_ONCE() ]

---
 arch/x86/mm/pti.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
arch/x86/mm/pti.c +4 -2
index ba22b50f4eca2..7f2140414440d 100644
@@ -330,13 +330,15 @@ pti_clone_pgtable(unsigned long start, unsigned long end,
 
 		pud = pud_offset(p4d, addr);
 		if (pud_none(*pud)) {
-			addr += PUD_SIZE;
+			WARN_ON_ONCE(addr & ~PUD_MASK);
+			addr = round_up(addr + 1, PUD_SIZE);
 			continue;
 		}
 
 		pmd = pmd_offset(pud, addr);
 		if (pmd_none(*pmd)) {
-			addr += PMD_SIZE;
+			WARN_ON_ONCE(addr & ~PMD_MASK);
+			addr = round_up(addr + 1, PMD_SIZE);
 			continue;
 		}
 
Signature · 1 line
2.20.1
author maintainer
#163 lore raw

PATCH 160/313 nvmet: fix data units read and written counters in SMART log

In reply to Greg Kroah-Hartman

From: Tom Wu <[email protected]>

[ Upstream commit 3bec2e3754becebd4c452999adb49bc62c575ea4 ]

In nvme spec 1.3 there is a definition for data write/read counters from SMART log, (See section 5.14.1.2):
This value is reported in thousands (i.e., a value of 1 corresponds to 1000 units of 512 bytes read) and is rounded up.

However, in nvme target where value is reported with actual units, but not thousands of units as the spec requires.

---
 drivers/nvme/target/admin-cmd.c | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)
drivers/nvme/target/admin-cmd.c +8 -6
index 9f72d515fc4b3..4099093a17343 100644
@@ -81,9 +81,11 @@ static u16 nvmet_get_smart_log_nsid(struct nvmet_req *req,
 		goto out;
 
 	host_reads = part_stat_read(ns->bdev->bd_part, ios[READ]);
-	data_units_read = part_stat_read(ns->bdev->bd_part, sectors[READ]);
+	data_units_read = DIV_ROUND_UP(part_stat_read(ns->bdev->bd_part,
+		sectors[READ]), 1000);
 	host_writes = part_stat_read(ns->bdev->bd_part, ios[WRITE]);
-	data_units_written = part_stat_read(ns->bdev->bd_part, sectors[WRITE]);
+	data_units_written = DIV_ROUND_UP(part_stat_read(ns->bdev->bd_part,
+		sectors[WRITE]), 1000);
 
 	put_unaligned_le64(host_reads, &slog->host_reads[0]);
 	put_unaligned_le64(data_units_read, &slog->data_units_read[0]);
@@ -111,11 +113,11 @@ static u16 nvmet_get_smart_log_all(struct nvmet_req *req,
 		if (!ns->bdev)
 			continue;
 		host_reads += part_stat_read(ns->bdev->bd_part, ios[READ]);
-		data_units_read +=
-			part_stat_read(ns->bdev->bd_part, sectors[READ]);
+		data_units_read += DIV_ROUND_UP(
+			part_stat_read(ns->bdev->bd_part, sectors[READ]), 1000);
 		host_writes += part_stat_read(ns->bdev->bd_part, ios[WRITE]);
-		data_units_written +=
-			part_stat_read(ns->bdev->bd_part, sectors[WRITE]);
+		data_units_written += DIV_ROUND_UP(
+			part_stat_read(ns->bdev->bd_part, sectors[WRITE]), 1000);
 
 	}
 	rcu_read_unlock();
Signature · 1 line
2.20.1
author maintainer
#164 lore raw

PATCH 161/313 nvme-multipath: fix ana log nsid lookup when nsid is not found

In reply to Greg Kroah-Hartman

From: Anton Eidelman <[email protected]>

[ Upstream commit e01f91dff91c7b16a6e3faf2565017d497a73f83 ]

ANA log parsing invokes nvme_update_ana_state() per ANA group desc. This updates the state of namespaces with nsids in desc->nsids[].

Both ctrl->namespaces list and desc->nsids[] array are sorted by nsid. Hence nvme_update_ana_state() performs a single walk over ctrl->namespaces:
- if current namespace matches the current desc->nsids[n],
this namespace is updated, and n is incremented.
- the process stops when it encounters the end of either ctrl->namespaces end or desc->nsids[]

In case desc->nsids[n] does not match any of ctrl->namespaces, the remaining nsids following desc->nsids[n] will not be updated. Such situation was considered abnormal and generated WARN_ON_ONCE.

However ANA log MAY contain nsids not (yet) found in ctrl->namespaces. For example, lets consider the following scenario:
- nvme0 exposes namespaces with nsids = [2, 3] to the host
- a new namespace nsid = 1 is added dynamically
- also, a ANA topology change is triggered
- NS_CHANGED aen is generated and triggers scan_work
- before scan_work discovers nsid=1 and creates a namespace, a NOTICE_ANA aen was issues and ana_work receives ANA log with nsids=[1, 2, 3]

Result: ana_work fails to update ANA state on existing namespaces [2, 3]

Solution:
Change the way nvme_update_ana_state() namespace list walk checks the current namespace against desc->nsids[n] as follows:
a) ns->head->ns_id < desc->nsids[n]: keep walking ctrl->namespaces.
b) ns->head->ns_id == desc->nsids[n]: match, update the namespace
c) ns->head->ns_id >= desc->nsids[n]: skip to desc->nsids[n+1]

This enables correct operation in the scenario described above. This also allows ANA log to contain nsids currently invisible to the host, i.e. inactive nsids.

---
 drivers/nvme/host/multipath.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)
drivers/nvme/host/multipath.c +5 -3
index 304aa8a65f2f8..f928bcfc57b5b 100644
@@ -501,14 +501,16 @@ static int nvme_update_ana_state(struct nvme_ctrl *ctrl,
 
 	down_write(&ctrl->namespaces_rwsem);
 	list_for_each_entry(ns, &ctrl->namespaces, list) {
-		if (ns->head->ns_id != le32_to_cpu(desc->nsids[n]))
+		unsigned nsid = le32_to_cpu(desc->nsids[n]);
+
+		if (ns->head->ns_id < nsid)
 			continue;
-		nvme_update_ns_ana_state(desc, ns);
+		if (ns->head->ns_id == nsid)
+			nvme_update_ns_ana_state(desc, ns);
 		if (++n == nr_nsids)
 			break;
 	}
 	up_write(&ctrl->namespaces_rwsem);
-	WARN_ON_ONCE(n < nr_nsids);
 	return 0;
 }
 
Signature · 1 line
2.20.1
author maintainer
#165 lore raw

PATCH 162/313 ALSA: firewire-motu: add support for MOTU 4pre

In reply to Greg Kroah-Hartman

From: Takashi Sakamoto <[email protected]>

[ Upstream commit 6af86bdb8ad41f4cf1292d3b10857dc322758328 ]

MOTU 4pre was launched in 2012 by MOTU, Inc. This commit allows userspace applications can transmit and receive PCM frames and MIDI messages for this model via ALSA PCM interface and RawMidi/Sequencer interfaces.

The device supports MOTU protocol version 3. Unlike the other devices, the device is simply designed. The size of data block is fixed to 10 quadlets during available sampling rates (44.1 - 96.0 kHz). Each data block includes 1 source packet header, 2 data chunks for messages, 8 data chunks for PCM samples and 2 data chunks for padding to quadlet alignment. The device has no MIDI, optical, BNC and AES/EBU interfaces.

Like support for the other MOTU devices, the quality of playback sound is not enough good with periodical noise yet.

$ python2 crpp < ~/git/am-config-rom/motu/motu-4pre.img
               ROM header and bus information block
               -----------------------------------------------------------------
400  041078cc  bus_info_length 4, crc_length 16, crc 30924
404  31333934  bus_name "1394"
408  20ff7000  irmc 0, cmc 0, isc 1, bmc 0, cyc_clk_acc 255, max_rec 7 (256)
40c  0001f200  company_id 0001f2     |
410  000a41c5  device_id 00000a41c5  | EUI-64 0001f200000a41c5
               root directory
               -----------------------------------------------------------------
414  0004ef04  directory_length 4, crc 61188
418  030001f2  vendor
41c  0c0083c0  node capabilities per IEEE 1394
420  d1000002  --> unit directory at 428
424  8d000005  --> eui-64 leaf at 438

unit directory at 428
-----------------------------------------------------------------

428  0003ceda  directory_length 3, crc 52954
42c  120001f2  specifier id
430  13000045  version
434  17103800  model
               eui-64 leaf at 438
               -----------------------------------------------------------------
438  0002d248  leaf_length 2, crc 53832
43c  0001f200  company_id 0001f2     |
440  000a41c5  device_id 00000a41c5  | EUI-64 0001f200000a41c5
---
 sound/firewire/motu/motu.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)
sound/firewire/motu/motu.c +12 -0
index 03cda2166ea3d..72908b4de77c0 100644
@@ -247,6 +247,17 @@ static const struct snd_motu_spec motu_audio_express = {
 	.analog_out_ports = 4,
 };
 
+static const struct snd_motu_spec motu_4pre = {
+	.name = "4pre",
+	.protocol = &snd_motu_protocol_v3,
+	.flags = SND_MOTU_SPEC_SUPPORT_CLOCK_X2 |
+		 SND_MOTU_SPEC_TX_MICINST_CHUNK |
+		 SND_MOTU_SPEC_TX_RETURN_CHUNK |
+		 SND_MOTU_SPEC_RX_SEPARETED_MAIN,
+	.analog_in_ports = 2,
+	.analog_out_ports = 2,
+};
+
 #define SND_MOTU_DEV_ENTRY(model, data)			\
 {							\
 	.match_flags	= IEEE1394_MATCH_VENDOR_ID |	\
@@ -265,6 +276,7 @@ static const struct ieee1394_device_id motu_id_table[] = {
 	SND_MOTU_DEV_ENTRY(0x000015, &motu_828mk3),	/* FireWire only. */
 	SND_MOTU_DEV_ENTRY(0x000035, &motu_828mk3),	/* Hybrid. */
 	SND_MOTU_DEV_ENTRY(0x000033, &motu_audio_express),
+	SND_MOTU_DEV_ENTRY(0x000045, &motu_4pre),
 	{ }
 };
 MODULE_DEVICE_TABLE(ieee1394, motu_id_table);
Signature · 1 line
2.20.1
author maintainer
#166 lore raw

PATCH 163/313 iommu/amd: Silence warnings under memory pressure

In reply to Greg Kroah-Hartman

From: Qian Cai <[email protected]>

[ Upstream commit 3d708895325b78506e8daf00ef31549476e8586a ]

When running heavy memory pressure workloads, the system is throwing endless warnings,

smartpqi 0000:23:00.0: AMD-Vi: IOMMU mapping error in map_sg (io-pages:
5 reason: -12)

Hardware name: HPE ProLiant DL385 Gen10/ProLiant DL385 Gen10, BIOS A40
07/10/2019
swapper/10: page allocation failure: order:0, mode:0xa20(GFP_ATOMIC),
nodemask=(null),cpuset=/,mems_allowed=0,4
Call Trace:
 <IRQ>
 dump_stack+0x62/0x9a
 warn_alloc.cold.43+0x8a/0x148
 __alloc_pages_nodemask+0x1a5c/0x1bb0
 get_zeroed_page+0x16/0x20
 iommu_map_page+0x477/0x540
 map_sg+0x1ce/0x2f0
 scsi_dma_map+0xc6/0x160
 pqi_raid_submit_scsi_cmd_with_io_request+0x1c3/0x470 [smartpqi]
 do_IRQ+0x81/0x170
 common_interrupt+0xf/0xf
 </IRQ>

because the allocation could fail from iommu_map_page(), and the volume of this call could be huge which may generate a lot of serial console output and cosumes all CPUs.

Fix it by silencing the warning in this call site, and there is still a dev_err() later to notify the failure.

---
 drivers/iommu/amd_iommu.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/iommu/amd_iommu.c +3 -1
index 3e687f18b203a..a0b64c43257a6 100644
@@ -2570,7 +2570,9 @@ static int map_sg(struct device *dev, struct scatterlist *sglist,
 
 			bus_addr  = address + s->dma_address + (j << PAGE_SHIFT);
 			phys_addr = (sg_phys(s) & PAGE_MASK) + (j << PAGE_SHIFT);
-			ret = iommu_map_page(domain, bus_addr, phys_addr, PAGE_SIZE, prot, GFP_ATOMIC);
+			ret = iommu_map_page(domain, bus_addr, phys_addr,
+					     PAGE_SIZE, prot,
+					     GFP_ATOMIC | __GFP_NOWARN);
 			if (ret)
 				goto out_unmap;
 
Signature · 1 line
2.20.1
author maintainer
#167 lore raw

PATCH 164/313 ASoC: Intel: Haswell: Adjust machine device private context

In reply to Greg Kroah-Hartman

From: Cezary Rojewski <[email protected]>

[ Upstream commit ca964edf0ddbfec2cb10b3d251d09598e7ca9b13 ]

Apart from Haswell machines, all other devices have their private data set to snd_soc_acpi_mach instance.

Changes for HSW/ BDW boards introduced with series:
https://patchwork.kernel.org/cover/10782035/

added support for dai_link platform_name adjustments within card probe routines. These take for granted private_data points to snd_soc_acpi_mach whereas for Haswell, it's sst_pdata instead. Change private context of platform_device - representing machine board - to address this.

Fixes e87055d732e3 ("ASoC: Intel: haswell: platform name fixup support")
Fixes 7e40ddcf974a ("ASoC: Intel: bdw-rt5677: platform name fixup support")
Fixes 2d067b2807f9 ("ASoC: Intel: broadwell: platform name fixup support")
---
 sound/soc/intel/common/sst-acpi.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
sound/soc/intel/common/sst-acpi.c +2 -1
index 0e8e0a7a11df3..5854868650b9e 100644
@@ -141,11 +141,12 @@ static int sst_acpi_probe(struct platform_device *pdev)
 	}
 
 	platform_set_drvdata(pdev, sst_acpi);
+	mach->pdata = sst_pdata;
 
 	/* register machine driver */
 	sst_acpi->pdev_mach =
 		platform_device_register_data(dev, mach->drv_name, -1,
-					      sst_pdata, sizeof(*sst_pdata));
+					      mach, sizeof(*mach));
 	if (IS_ERR(sst_acpi->pdev_mach))
 		return PTR_ERR(sst_acpi->pdev_mach);
 
Signature · 1 line
2.20.1
author maintainer
#168 lore raw

PATCH 165/313 libata/ahci: Drop PCS quirk for Denverton and beyond

In reply to Greg Kroah-Hartman

From: Dan Williams <[email protected]>

[ Upstream commit c312ef176399e04fc5f7f2809d9a589751fbf6d9 ]

The Linux ahci driver has historically implemented a configuration fixup for platforms / platform-firmware that fails to enable the ports prior to OS hand-off at boot. The fixup was originally implemented way back before ahci moved from drivers/scsi/ to drivers/ata/, and was updated in 2007 via commit 49f290903935 "ahci: update PCS programming". The quirk sets a port-enable bitmap in the PCS register at offset 0x92.

This quirk could be applied generically up until the arrival of the Denverton (DNV) platform. The DNV AHCI controller architecture supports more than 6 ports and along with that the PCS register location and format were updated to allow for more possible ports in the bitmap. DNV AHCI expands the register to 32-bits and moves it to offset 0x94.

As it stands there are no known problem reports with existing Linux trying to set bits at offset 0x92 which indicates that the quirk is not applicable. Likely it is not applicable on a wider range of platforms, but it is difficult to discern which platforms if any still depend on the quirk.

Rather than try to fix the PCS quirk to consider the DNV register layout instead require explicit opt-in. The assumption is that the OS driver need not touch this register, and platforms can be added with a new boad_ahci_pcs7 board-id when / if problematic platforms are found in the future. The logic in ahci_intel_pcs_quirk() looks for all Intel AHCI instances with "legacy" board-ids and otherwise skips the quirk if the board was matched by class-code.

---
 drivers/ata/ahci.c | 116 +++++++++++++++++++++++++++------------------
 drivers/ata/ahci.h |   2 +
 2 files changed, 71 insertions(+), 47 deletions(-)
drivers/ata/ahci.c +69 -47
index f7652baa63370..3e63294304c72 100644
@@ -65,6 +65,12 @@ enum board_ids {
 	board_ahci_sb700,	/* for SB700 and SB800 */
 	board_ahci_vt8251,
 
+	/*
+	 * board IDs for Intel chipsets that support more than 6 ports
+	 * *and* end up needing the PCS quirk.
+	 */
+	board_ahci_pcs7,
+
 	/* aliases */
 	board_ahci_mcp_linux	= board_ahci_mcp65,
 	board_ahci_mcp67	= board_ahci_mcp65,
@@ -220,6 +226,12 @@ static const struct ata_port_info ahci_port_info[] = {
 		.udma_mask	= ATA_UDMA6,
 		.port_ops	= &ahci_vt8251_ops,
 	},
+	[board_ahci_pcs7] = {
+		.flags		= AHCI_FLAG_COMMON,
+		.pio_mask	= ATA_PIO4,
+		.udma_mask	= ATA_UDMA6,
+		.port_ops	= &ahci_ops,
+	},
 };
 
 static const struct pci_device_id ahci_pci_tbl[] = {
@@ -264,26 +276,26 @@ static const struct pci_device_id ahci_pci_tbl[] = {
 	{ PCI_VDEVICE(INTEL, 0x3b2b), board_ahci }, /* PCH RAID */
 	{ PCI_VDEVICE(INTEL, 0x3b2c), board_ahci_mobile }, /* PCH M RAID */
 	{ PCI_VDEVICE(INTEL, 0x3b2f), board_ahci }, /* PCH AHCI */
-	{ PCI_VDEVICE(INTEL, 0x19b0), board_ahci }, /* DNV AHCI */
-	{ PCI_VDEVICE(INTEL, 0x19b1), board_ahci }, /* DNV AHCI */
-	{ PCI_VDEVICE(INTEL, 0x19b2), board_ahci }, /* DNV AHCI */
-	{ PCI_VDEVICE(INTEL, 0x19b3), board_ahci }, /* DNV AHCI */
-	{ PCI_VDEVICE(INTEL, 0x19b4), board_ahci }, /* DNV AHCI */
-	{ PCI_VDEVICE(INTEL, 0x19b5), board_ahci }, /* DNV AHCI */
-	{ PCI_VDEVICE(INTEL, 0x19b6), board_ahci }, /* DNV AHCI */
-	{ PCI_VDEVICE(INTEL, 0x19b7), board_ahci }, /* DNV AHCI */
-	{ PCI_VDEVICE(INTEL, 0x19bE), board_ahci }, /* DNV AHCI */
-	{ PCI_VDEVICE(INTEL, 0x19bF), board_ahci }, /* DNV AHCI */
-	{ PCI_VDEVICE(INTEL, 0x19c0), board_ahci }, /* DNV AHCI */
-	{ PCI_VDEVICE(INTEL, 0x19c1), board_ahci }, /* DNV AHCI */
-	{ PCI_VDEVICE(INTEL, 0x19c2), board_ahci }, /* DNV AHCI */
-	{ PCI_VDEVICE(INTEL, 0x19c3), board_ahci }, /* DNV AHCI */
-	{ PCI_VDEVICE(INTEL, 0x19c4), board_ahci }, /* DNV AHCI */
-	{ PCI_VDEVICE(INTEL, 0x19c5), board_ahci }, /* DNV AHCI */
-	{ PCI_VDEVICE(INTEL, 0x19c6), board_ahci }, /* DNV AHCI */
-	{ PCI_VDEVICE(INTEL, 0x19c7), board_ahci }, /* DNV AHCI */
-	{ PCI_VDEVICE(INTEL, 0x19cE), board_ahci }, /* DNV AHCI */
-	{ PCI_VDEVICE(INTEL, 0x19cF), board_ahci }, /* DNV AHCI */
+	{ PCI_VDEVICE(INTEL, 0x19b0), board_ahci_pcs7 }, /* DNV AHCI */
+	{ PCI_VDEVICE(INTEL, 0x19b1), board_ahci_pcs7 }, /* DNV AHCI */
+	{ PCI_VDEVICE(INTEL, 0x19b2), board_ahci_pcs7 }, /* DNV AHCI */
+	{ PCI_VDEVICE(INTEL, 0x19b3), board_ahci_pcs7 }, /* DNV AHCI */
+	{ PCI_VDEVICE(INTEL, 0x19b4), board_ahci_pcs7 }, /* DNV AHCI */
+	{ PCI_VDEVICE(INTEL, 0x19b5), board_ahci_pcs7 }, /* DNV AHCI */
+	{ PCI_VDEVICE(INTEL, 0x19b6), board_ahci_pcs7 }, /* DNV AHCI */
+	{ PCI_VDEVICE(INTEL, 0x19b7), board_ahci_pcs7 }, /* DNV AHCI */
+	{ PCI_VDEVICE(INTEL, 0x19bE), board_ahci_pcs7 }, /* DNV AHCI */
+	{ PCI_VDEVICE(INTEL, 0x19bF), board_ahci_pcs7 }, /* DNV AHCI */
+	{ PCI_VDEVICE(INTEL, 0x19c0), board_ahci_pcs7 }, /* DNV AHCI */
+	{ PCI_VDEVICE(INTEL, 0x19c1), board_ahci_pcs7 }, /* DNV AHCI */
+	{ PCI_VDEVICE(INTEL, 0x19c2), board_ahci_pcs7 }, /* DNV AHCI */
+	{ PCI_VDEVICE(INTEL, 0x19c3), board_ahci_pcs7 }, /* DNV AHCI */
+	{ PCI_VDEVICE(INTEL, 0x19c4), board_ahci_pcs7 }, /* DNV AHCI */
+	{ PCI_VDEVICE(INTEL, 0x19c5), board_ahci_pcs7 }, /* DNV AHCI */
+	{ PCI_VDEVICE(INTEL, 0x19c6), board_ahci_pcs7 }, /* DNV AHCI */
+	{ PCI_VDEVICE(INTEL, 0x19c7), board_ahci_pcs7 }, /* DNV AHCI */
+	{ PCI_VDEVICE(INTEL, 0x19cE), board_ahci_pcs7 }, /* DNV AHCI */
+	{ PCI_VDEVICE(INTEL, 0x19cF), board_ahci_pcs7 }, /* DNV AHCI */
 	{ PCI_VDEVICE(INTEL, 0x1c02), board_ahci }, /* CPT AHCI */
 	{ PCI_VDEVICE(INTEL, 0x1c03), board_ahci_mobile }, /* CPT M AHCI */
 	{ PCI_VDEVICE(INTEL, 0x1c04), board_ahci }, /* CPT RAID */
@@ -623,30 +635,6 @@ static void ahci_pci_save_initial_config(struct pci_dev *pdev,
 	ahci_save_initial_config(&pdev->dev, hpriv);
 }
 
-static int ahci_pci_reset_controller(struct ata_host *host)
-{
-	struct pci_dev *pdev = to_pci_dev(host->dev);
-	int rc;
-
-	rc = ahci_reset_controller(host);
-	if (rc)
-		return rc;
-
-	if (pdev->vendor == PCI_VENDOR_ID_INTEL) {
-		struct ahci_host_priv *hpriv = host->private_data;
-		u16 tmp16;
-
-		/* configure PCS */
-		pci_read_config_word(pdev, 0x92, &tmp16);
-		if ((tmp16 & hpriv->port_map) != hpriv->port_map) {
-			tmp16 |= hpriv->port_map;
-			pci_write_config_word(pdev, 0x92, tmp16);
-		}
-	}
-
-	return 0;
-}
-
 static void ahci_pci_init_controller(struct ata_host *host)
 {
 	struct ahci_host_priv *hpriv = host->private_data;
@@ -849,7 +837,7 @@ static int ahci_pci_device_runtime_resume(struct device *dev)
 	struct ata_host *host = pci_get_drvdata(pdev);
 	int rc;
 
-	rc = ahci_pci_reset_controller(host);
+	rc = ahci_reset_controller(host);
 	if (rc)
 		return rc;
 	ahci_pci_init_controller(host);
@@ -884,7 +872,7 @@ static int ahci_pci_device_resume(struct device *dev)
 		ahci_mcp89_apple_enable(pdev);
 
 	if (pdev->dev.power.power_state.event == PM_EVENT_SUSPEND) {
-		rc = ahci_pci_reset_controller(host);
+		rc = ahci_reset_controller(host);
 		if (rc)
 			return rc;
 
@@ -1619,6 +1607,34 @@ static void ahci_update_initial_lpm_policy(struct ata_port *ap,
 		ap->target_lpm_policy = policy;
 }
 
+static void ahci_intel_pcs_quirk(struct pci_dev *pdev, struct ahci_host_priv *hpriv)
+{
+	const struct pci_device_id *id = pci_match_id(ahci_pci_tbl, pdev);
+	u16 tmp16;
+
+	/*
+	 * Only apply the 6-port PCS quirk for known legacy platforms.
+	 */
+	if (!id || id->vendor != PCI_VENDOR_ID_INTEL)
+		return;
+	if (((enum board_ids) id->driver_data) < board_ahci_pcs7)
+		return;
+
+	/*
+	 * port_map is determined from PORTS_IMPL PCI register which is
+	 * implemented as write or write-once register.  If the register
+	 * isn't programmed, ahci automatically generates it from number
+	 * of ports, which is good enough for PCS programming. It is
+	 * otherwise expected that platform firmware enables the ports
+	 * before the OS boots.
+	 */
+	pci_read_config_word(pdev, PCS_6, &tmp16);
+	if ((tmp16 & hpriv->port_map) != hpriv->port_map) {
+		tmp16 |= hpriv->port_map;
+		pci_write_config_word(pdev, PCS_6, tmp16);
+	}
+}
+
 static int ahci_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 {
 	unsigned int board_id = ent->driver_data;
@@ -1731,6 +1747,12 @@ static int ahci_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 	/* save initial config */
 	ahci_pci_save_initial_config(pdev, hpriv);
 
+	/*
+	 * If platform firmware failed to enable ports, try to enable
+	 * them here.
+	 */
+	ahci_intel_pcs_quirk(pdev, hpriv);
+
 	/* prepare host */
 	if (hpriv->cap & HOST_CAP_NCQ) {
 		pi.flags |= ATA_FLAG_NCQ;
@@ -1840,7 +1862,7 @@ static int ahci_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 	if (rc)
 		return rc;
 
-	rc = ahci_pci_reset_controller(host);
+	rc = ahci_reset_controller(host);
 	if (rc)
 		return rc;
 
drivers/ata/ahci.h +2 -0
index 0570629d719d8..3dbf398c92eaf 100644
@@ -247,6 +247,8 @@ enum {
 					  ATA_FLAG_ACPI_SATA | ATA_FLAG_AN,
 
 	ICH_MAP				= 0x90, /* ICH MAP register */
+	PCS_6				= 0x92, /* 6 port PCS */
+	PCS_7				= 0x94, /* 7+ port PCS (Denverton) */
 
 	/* em constants */
 	EM_MAX_SLOTS			= 8,
Signature · 1 line
2.20.1
author maintainer
#169 lore raw

PATCH 166/313 iommu/iova: Avoid false sharing on fq_timer_on

In reply to Greg Kroah-Hartman

From: Eric Dumazet <[email protected]>

[ Upstream commit 0d87308cca2c124f9bce02383f1d9632c9be89c4 ]

In commit 14bd9a607f90 ("iommu/iova: Separate atomic variables to improve performance") Jinyu Qi identified that the atomic_cmpxchg() in queue_iova() was causing a performance loss and moved critical fields so that the false sharing would not impact them.

However, avoiding the false sharing in the first place seems easy. We should attempt the atomic_cmpxchg() no more than 100 times per second. Adding an atomic_read() will keep the cache line mostly shared.

This false sharing came with commit 9a005a800ae8
("iommu/iova: Add flush timer").

Fixes 9a005a800ae8 ('iommu/iova: Add flush timer')
Cc Jinyu Qi <[email protected]>
Cc Joerg Roedel <[email protected]>
---
 drivers/iommu/iova.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/iommu/iova.c +3 -1
index 3e1a8a6755723..41c605b0058f9 100644
@@ -577,7 +577,9 @@ void queue_iova(struct iova_domain *iovad,
 
 	spin_unlock_irqrestore(&fq->lock, flags);
 
-	if (atomic_cmpxchg(&iovad->fq_timer_on, 0, 1) == 0)
+	/* Avoid false sharing as much as possible. */
+	if (!atomic_read(&iovad->fq_timer_on) &&
+	    !atomic_cmpxchg(&iovad->fq_timer_on, 0, 1))
 		mod_timer(&iovad->fq_timer,
 			  jiffies + msecs_to_jiffies(IOVA_FQ_TIMEOUT));
 }
Signature · 1 line
2.20.1
author maintainer
#170 lore raw

PATCH 167/313 libtraceevent: Change users plugin directory

In reply to Greg Kroah-Hartman

From: Tzvetomir Stoyanov <[email protected]>

[ Upstream commit e97fd1383cd77c467d2aed7fa4e596789df83977 ]

To be compliant with XDG user directory layout, the user's plugin directory is changed from ~/.traceevent/plugins to ~/.local/lib/traceevent/plugins/

---
 tools/lib/traceevent/Makefile       | 6 +++---
 tools/lib/traceevent/event-plugin.c | 2 +-
 2 files changed, 4 insertions(+), 4 deletions(-)
tools/lib/traceevent/Makefile +3 -3
index 3292c290654f6..86ce17a1f7fb6 100644
@@ -62,15 +62,15 @@ set_plugin_dir := 1
 
 # Set plugin_dir to preffered global plugin location
 # If we install under $HOME directory we go under
-# $(HOME)/.traceevent/plugins
+# $(HOME)/.local/lib/traceevent/plugins
 #
 # We dont set PLUGIN_DIR in case we install under $HOME
 # directory, because by default the code looks under:
-# $(HOME)/.traceevent/plugins by default.
+# $(HOME)/.local/lib/traceevent/plugins by default.
 #
 ifeq ($(plugin_dir),)
 ifeq ($(prefix),$(HOME))
-override plugin_dir = $(HOME)/.traceevent/plugins
+override plugin_dir = $(HOME)/.local/lib/traceevent/plugins
 set_plugin_dir := 0
 else
 override plugin_dir = $(libdir)/traceevent/plugins
tools/lib/traceevent/event-plugin.c +1 -1
index 8ca28de9337a5..e1f7ddd5a6cf0 100644
@@ -18,7 +18,7 @@
 #include "event-utils.h"
 #include "trace-seq.h"
 
-#define LOCAL_PLUGIN_DIR ".traceevent/plugins"
+#define LOCAL_PLUGIN_DIR ".local/lib/traceevent/plugins/"
 
 static struct registered_plugin_options {
 	struct registered_plugin_options	*next;
Signature · 1 line
2.20.1
author maintainer
#171 lore raw

PATCH 168/313 ARM: dts: exynos: Mark LDO10 as always-on on Peach Pit/Pi Chromebooks

In reply to Greg Kroah-Hartman

From: Marek Szyprowski <[email protected]>

[ Upstream commit 5b0eeeaa37615df37a9a30929b73e9defe61ca84 ]

Commit aff138bf8e37 ("ARM: dts: exynos: Add TMU nodes regulator supply for Peach boards") assigned LDO10 to Exynos Thermal Measurement Unit, but it turned out that it supplies also some other critical parts and board freezes/crashes when it is turned off.

The mentioned commit made Exynos TMU a consumer of that regulator and in typical case Exynos TMU driver keeps it enabled from early boot. However there are such configurations (example is multi_v7_defconfig), in which some of the regulators are compiled as modules and are not available from early boot. In such case it may happen that LDO10 is turned off by regulator core, because it has no consumers yet (in this case consumer drivers cannot get it, because the supply regulators for it are not yet available). This in turn causes the board to crash. This patch restores 'always-on' property for the LDO10 regulator.

Fixes aff138bf8e37 ("ARM: dts: exynos: Add TMU nodes regulator supply for Peach boards")
---
 arch/arm/boot/dts/exynos5420-peach-pit.dts | 1 +
 arch/arm/boot/dts/exynos5800-peach-pi.dts  | 1 +
 2 files changed, 2 insertions(+)
arch/arm/boot/dts/exynos5420-peach-pit.dts +1 -0
index f78db6809cca4..9eb48cabcca45 100644
@@ -440,6 +440,7 @@
 				regulator-name = "vdd_ldo10";
 				regulator-min-microvolt = <1800000>;
 				regulator-max-microvolt = <1800000>;
+				regulator-always-on;
 				regulator-state-mem {
 					regulator-off-in-suspend;
 				};
arch/arm/boot/dts/exynos5800-peach-pi.dts +1 -0
index e0f470fe54c81..4398f2d1fe881 100644
@@ -440,6 +440,7 @@
 				regulator-name = "vdd_ldo10";
 				regulator-min-microvolt = <1800000>;
 				regulator-max-microvolt = <1800000>;
+				regulator-always-on;
 				regulator-state-mem {
 					regulator-off-in-suspend;
 				};
Signature · 1 line
2.20.1
author maintainer
#172 lore raw

PATCH 169/313 x86/amd_nb: Add PCI device IDs for family 17h, model 70h

In reply to Greg Kroah-Hartman

From: Marcel Bocu <[email protected]>

[ Upstream commit af4e1c5eca95bed1192d8dc45c8ed63aea2209e8 ]

The AMD Ryzen gen 3 processors came with a different PCI IDs for the function 3 & 4 which are used to access the SMN interface. The root PCI address however remained at the same address as the model 30h.

Adding the F3/F4 PCI IDs respectively to the misc and link ids appear to be sufficient for k10temp, so let's add them and follow up on the patch if other functions need more tweaking.

Vicki Pfau sent an identical patch after I checked that no-one had written this patch. I would have been happy about dropping my patch but unlike for his patch series, I had already Cc:ed the x86 people and they already reviewed the changes. Since Vicki has not answered to any email after his initial series, let's assume she is on vacation and let's avoid duplication of reviews from the maintainers and merge my series. To acknowledge Vicki's anteriority, I added her S-o-b to the patch.

v2, suggested by Guenter Roeck and Brian Woods:
- rename from 71h to 70h

---
 arch/x86/kernel/amd_nb.c | 3 +++
 include/linux/pci_ids.h  | 1 +
 2 files changed, 4 insertions(+)
arch/x86/kernel/amd_nb.c +3 -0
index 002aedc693933..8c26b696d8930 100644
@@ -21,6 +21,7 @@
 #define PCI_DEVICE_ID_AMD_17H_DF_F4	0x1464
 #define PCI_DEVICE_ID_AMD_17H_M10H_DF_F4 0x15ec
 #define PCI_DEVICE_ID_AMD_17H_M30H_DF_F4 0x1494
+#define PCI_DEVICE_ID_AMD_17H_M70H_DF_F4 0x1444
 
 /* Protect the PCI config register pairs used for SMN and DF indirect access. */
 static DEFINE_MUTEX(smn_mutex);
@@ -50,6 +51,7 @@ const struct pci_device_id amd_nb_misc_ids[] = {
 	{ PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_17H_M10H_DF_F3) },
 	{ PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_17H_M30H_DF_F3) },
 	{ PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_CNB17H_F3) },
+	{ PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_17H_M70H_DF_F3) },
 	{}
 };
 EXPORT_SYMBOL_GPL(amd_nb_misc_ids);
@@ -63,6 +65,7 @@ static const struct pci_device_id amd_nb_link_ids[] = {
 	{ PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_17H_DF_F4) },
 	{ PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_17H_M10H_DF_F4) },
 	{ PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_17H_M30H_DF_F4) },
+	{ PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_17H_M70H_DF_F4) },
 	{ PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_CNB17H_F4) },
 	{}
 };
include/linux/pci_ids.h +1 -0
index 70e86148cb1e9..862556761bbf4 100644
@@ -548,6 +548,7 @@
 #define PCI_DEVICE_ID_AMD_17H_DF_F3	0x1463
 #define PCI_DEVICE_ID_AMD_17H_M10H_DF_F3 0x15eb
 #define PCI_DEVICE_ID_AMD_17H_M30H_DF_F3 0x1493
+#define PCI_DEVICE_ID_AMD_17H_M70H_DF_F3 0x1443
 #define PCI_DEVICE_ID_AMD_CNB17H_F3	0x1703
 #define PCI_DEVICE_ID_AMD_LANCE		0x2000
 #define PCI_DEVICE_ID_AMD_LANCE_HOME	0x2001
Signature · 1 line
2.20.1
author maintainer
#173 lore raw

PATCH 170/313 ACPI: custom_method: fix memory leaks

In reply to Greg Kroah-Hartman

From: Wenwen Wang <[email protected]>

[ Upstream commit 03d1571d9513369c17e6848476763ebbd10ec2cb ]

In cm_write(), 'buf' is allocated through kzalloc(). In the following execution, if an error occurs, 'buf' is not deallocated, leading to memory leaks. To fix this issue, free 'buf' before returning the error.

Fixes 526b4af47f44 ("ACPI: Split out custom_method functionality into an own driver")
---
 drivers/acpi/custom_method.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/acpi/custom_method.c +4 -1
index b2ef4c2ec955d..fd66a736621cf 100644
@@ -49,8 +49,10 @@ static ssize_t cm_write(struct file *file, const char __user * user_buf,
 	if ((*ppos > max_size) ||
 	    (*ppos + count > max_size) ||
 	    (*ppos + count < count) ||
-	    (count > uncopied_bytes))
+	    (count > uncopied_bytes)) {
+		kfree(buf);
 		return -EINVAL;
+	}
 
 	if (copy_from_user(buf + (*ppos), user_buf, count)) {
 		kfree(buf);
@@ -70,6 +72,7 @@ static ssize_t cm_write(struct file *file, const char __user * user_buf,
 		add_taint(TAINT_OVERRIDDEN_ACPI_TABLE, LOCKDEP_NOW_UNRELIABLE);
 	}
 
+	kfree(buf);
 	return count;
 }
 
Signature · 1 line
2.20.1
author maintainer
#174 lore raw

PATCH 171/313 ACPI / PCI: fix acpi_pci_irq_enable() memory leak

In reply to Greg Kroah-Hartman

From: Wenwen Wang <[email protected]>

[ Upstream commit 29b49958cf73b439b17fa29e9a25210809a6c01c ]

In acpi_pci_irq_enable(), 'entry' is allocated by kzalloc() in acpi_pci_irq_check_entry() (invoked from acpi_pci_irq_lookup()). However, it is not deallocated if acpi_pci_irq_valid() returns false, leading to a memory leak. To fix this issue, free 'entry' before returning 0.

Fixes e237a5518425 ("x86/ACPI/PCI: Recognize that Interrupt Line 255 means "not connected"")
---
 drivers/acpi/pci_irq.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/acpi/pci_irq.c +3 -1
index d2549ae65e1b6..dea8a60e18a4c 100644
@@ -449,8 +449,10 @@ int acpi_pci_irq_enable(struct pci_dev *dev)
 		 * No IRQ known to the ACPI subsystem - maybe the BIOS /
 		 * driver reported one, then use it. Exit in any case.
 		 */
-		if (!acpi_pci_irq_valid(dev, pin))
+		if (!acpi_pci_irq_valid(dev, pin)) {
+			kfree(entry);
 			return 0;
+		}
 
 		if (acpi_isa_register_gsi(dev))
 			dev_warn(&dev->dev, "PCI INT %c: no GSI\n",
Signature · 1 line
2.20.1
author maintainer
#175 lore raw

PATCH 172/313 closures: fix a race on wakeup from closure_sync

In reply to Greg Kroah-Hartman

From: Kent Overstreet <[email protected]>

[ Upstream commit a22a9602b88fabf10847f238ff81fde5f906fef7 ]

The race was when a thread using closure_sync() notices cl->s->done == 1 before the thread calling closure_put() calls wake_up_process(). Then, it's possible for that thread to return and exit just before wake_up_process() is called - so we're trying to wake up a process that no longer exists.

rcu_read_lock() is sufficient to protect against this, as there's an rcu barrier somewhere in the process teardown path.

---
 drivers/md/bcache/closure.c | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)
drivers/md/bcache/closure.c +8 -2
index 73f5319295bc9..c12cd809ab193 100644
@@ -105,8 +105,14 @@ struct closure_syncer {
 
 static void closure_sync_fn(struct closure *cl)
 {
-	cl->s->done = 1;
-	wake_up_process(cl->s->task);
+	struct closure_syncer *s = cl->s;
+	struct task_struct *p;
+
+	rcu_read_lock();
+	p = READ_ONCE(s->task);
+	s->done = 1;
+	wake_up_process(p);
+	rcu_read_unlock();
 }
 
 void __sched __closure_sync(struct closure *cl)
Signature · 1 line
2.20.1
author maintainer
#176 lore raw

PATCH 173/313 hwmon: (k10temp) Add support for AMD family 17h, model 70h CPUs

In reply to Greg Kroah-Hartman

From: Marcel Bocu <[email protected]>

[ Upstream commit 12163cfbfc0f804cc7d27bc20e8d266ce7459260 ]

It would seem like model 70h is behaving in the same way as model 30h, so let's just add the new F3 PCI ID to the list of compatible devices.

Unlike previous Ryzen/Threadripper, Ryzen gen 3 processors do not need temperature offsets anymore. This has been reported in the press and verified on my Ryzen 3700X by checking that the idle temperature reported by k10temp is matching the temperature reported by the firmware.

Vicki Pfau sent an identical patch after I checked that no-one had written this patch. I would have been happy about dropping my patch but unlike for his patch series, I had already Cc:ed the x86 people and they already reviewed the changes. Since Vicki has not answered to any email after his initial series, let's assume she is on vacation and let's avoid duplication of reviews from the maintainers and merge my series. To acknowledge Vicki's anteriority, I added her S-o-b to the patch.

v2, suggested by Guenter Roeck and Brian Woods:
- rename from 71h to 70h

---
 drivers/hwmon/k10temp.c | 1 +
 1 file changed, 1 insertion(+)
drivers/hwmon/k10temp.c +1 -0
index c77e89239dcd9..5c1dddde193c3 100644
@@ -349,6 +349,7 @@ static const struct pci_device_id k10temp_id_table[] = {
 	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_17H_DF_F3) },
 	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_17H_M10H_DF_F3) },
 	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_17H_M30H_DF_F3) },
+	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_17H_M70H_DF_F3) },
 	{ PCI_VDEVICE(HYGON, PCI_DEVICE_ID_AMD_17H_DF_F3) },
 	{}
 };
Signature · 1 line
2.20.1
author maintainer
#177 lore raw

PATCH 174/313 hwmon: (acpi_power_meter) Change log level for unsafe software power cap

In reply to Greg Kroah-Hartman

From: Wang Shenran <[email protected]>

[ Upstream commit 6e4d91aa071810deac2cd052161aefb376ecf04e ]

At boot time, the acpi_power_meter driver logs the following error level
message: "Ignoring unsafe software power cap". Having read about it from a few sources, it seems that the error message can be quite misleading.

While the message can imply that Linux is ignoring the fact that the system is operating in potentially dangerous conditions, the truth is the driver found an ACPI_PMC object that supports software power capping. The driver simply decides not to use it, perhaps because it doesn't support the object.

The best solution is probably changing the log level from error to warning. All sources I have found, regarding the error, have downplayed its significance. There is not much of a reason for it to be on error level, while causing potential confusions or misinterpretations.

---
 drivers/hwmon/acpi_power_meter.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/hwmon/acpi_power_meter.c +2 -2
index 6ba1a08253f0a..4cf25458f0b95 100644
@@ -681,8 +681,8 @@ static int setup_attrs(struct acpi_power_meter_resource *resource)
 
 	if (resource->caps.flags & POWER_METER_CAN_CAP) {
 		if (!can_cap_in_hardware()) {
-			dev_err(&resource->acpi_dev->dev,
-				"Ignoring unsafe software power cap!\n");
+			dev_warn(&resource->acpi_dev->dev,
+				 "Ignoring unsafe software power cap!\n");
 			goto skip_unsafe_cap;
 		}
 
Signature · 1 line
2.20.1
author maintainer
#178 lore raw

PATCH 175/313 md/raid1: fail run raid1 array when active disk less than one

In reply to Greg Kroah-Hartman

From: Yufen Yu <[email protected]>

[ Upstream commit 07f1a6850c5d5a65c917c3165692b5179ac4cb6b ]

When run test case:
mdadm -CR /dev/md1 -l 1 -n 4 /dev/sd[a-d] --assume-clean --bitmap=internal mdadm -S /dev/md1
mdadm -A /dev/md1 /dev/sd[b-c] --run --force

  mdadm --zero /dev/sda
  mdadm /dev/md1 -a /dev/sda

echo offline > /sys/block/sdc/device/state
echo offline > /sys/block/sdb/device/state
sleep 5
mdadm -S /dev/md1

  echo running > /sys/block/sdb/device/state
  echo running > /sys/block/sdc/device/state
  mdadm -A /dev/md1 /dev/sd[a-c] --run --force

mdadm run fail with kernel message as follow:

[  172.986064] md: kicking non-fresh sdb from array!
[  173.004210] md: kicking non-fresh sdc from array!
[  173.022383] md/raid1:md1: active with 0 out of 4 mirrors
[  173.022406] md1: failed to create bitmap (-5)

In fact, when active disk in raid1 array less than one, we need to return fail in raid1_run().

---
 drivers/md/raid1.c | 13 ++++++++++++-
 1 file changed, 12 insertions(+), 1 deletion(-)
drivers/md/raid1.c +12 -1
index a26731a9b38e7..f393f0dc042fc 100644
@@ -3096,6 +3096,13 @@ static int raid1_run(struct mddev *mddev)
 		    !test_bit(In_sync, &conf->mirrors[i].rdev->flags) ||
 		    test_bit(Faulty, &conf->mirrors[i].rdev->flags))
 			mddev->degraded++;
+	/*
+	 * RAID1 needs at least one disk in active
+	 */
+	if (conf->raid_disks - mddev->degraded < 1) {
+		ret = -EINVAL;
+		goto abort;
+	}
 
 	if (conf->raid_disks - mddev->degraded == 1)
 		mddev->recovery_cp = MaxSector;
@@ -3129,8 +3136,12 @@ static int raid1_run(struct mddev *mddev)
 	ret =  md_integrity_register(mddev);
 	if (ret) {
 		md_unregister_thread(&mddev->thread);
-		raid1_free(mddev, conf);
+		goto abort;
 	}
+	return 0;
+
+abort:
+	raid1_free(mddev, conf);
 	return ret;
 }
 
Signature · 1 line
2.20.1
author maintainer
#179 lore raw

PATCH 176/313 dmaengine: ti: edma: Do not reset reserved paRAM slots

In reply to Greg Kroah-Hartman

From: Peter Ujfalusi <[email protected]>

[ Upstream commit c5dbe60664b3660f5ac5854e21273ea2e7ff698f ]

Skip resetting paRAM slots marked as reserved as they might be used by other cores.

---
 drivers/dma/ti/edma.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)
drivers/dma/ti/edma.c +6 -3
index ceabdea40ae0f..982631d4e1f8a 100644
@@ -2273,9 +2273,6 @@ static int edma_probe(struct platform_device *pdev)
 
 	ecc->default_queue = info->default_queue;
 
-	for (i = 0; i < ecc->num_slots; i++)
-		edma_write_slot(ecc, i, &dummy_paramset);
-
 	if (info->rsv) {
 		/* Set the reserved slots in inuse list */
 		rsv_slots = info->rsv->rsv_slots;
@@ -2288,6 +2285,12 @@ static int edma_probe(struct platform_device *pdev)
 		}
 	}
 
+	for (i = 0; i < ecc->num_slots; i++) {
+		/* Reset only unused - not reserved - paRAM slots */
+		if (!test_bit(i, ecc->slot_inuse))
+			edma_write_slot(ecc, i, &dummy_paramset);
+	}
+
 	/* Clear the xbar mapped channels in unused list */
 	xbar_chans = info->xbar_chans;
 	if (xbar_chans) {
Signature · 1 line
2.20.1
author maintainer
#180 lore raw

PATCH 177/313 kprobes: Prohibit probing on BUG() and WARN() address

In reply to Greg Kroah-Hartman

From: Masami Hiramatsu <[email protected]>

[ Upstream commit e336b4027775cb458dc713745e526fa1a1996b2a ]

Since BUG() and WARN() may use a trap (e.g. UD2 on x86) to get the address where the BUG() has occurred, kprobes can not do single-step out-of-line that instruction. So prohibit probing on such address.

Without this fix, if someone put a kprobe on WARN(), the kernel will crash with invalid opcode error instead of outputing warning message, because kernel can not find correct bug address.

---
 include/linux/bug.h | 5 +++++
 kernel/kprobes.c    | 3 ++-
 2 files changed, 7 insertions(+), 1 deletion(-)
include/linux/bug.h +5 -0
index fe5916550da8c..f639bd0122f39 100644
@@ -47,6 +47,11 @@ void generic_bug_clear_once(void);
 
 #else	/* !CONFIG_GENERIC_BUG */
 
+static inline void *find_bug(unsigned long bugaddr)
+{
+	return NULL;
+}
+
 static inline enum bug_trap_type report_bug(unsigned long bug_addr,
 					    struct pt_regs *regs)
 {
kernel/kprobes.c +2 -1
index 2504c269e6583..1010bde1146b5 100644
@@ -1514,7 +1514,8 @@ static int check_kprobe_address_safe(struct kprobe *p,
 	/* Ensure it is not in reserved area nor out of text */
 	if (!kernel_text_address((unsigned long) p->addr) ||
 	    within_kprobe_blacklist((unsigned long) p->addr) ||
-	    jump_label_text_reserved(p->addr, p->addr)) {
+	    jump_label_text_reserved(p->addr, p->addr) ||
+	    find_bug((unsigned long)p->addr)) {
 		ret = -EINVAL;
 		goto out;
 	}
Signature · 1 line
2.20.1
author maintainer
#181 lore raw

PATCH 178/313 x86/mm: Fix cpumask_of_node() error condition

In reply to Greg Kroah-Hartman

From: Peter Zijlstra <[email protected]>

[ Upstream commit bc04a049f058a472695aa22905d57e2b1f4c77d9 ]

When CONFIG_DEBUG_PER_CPU_MAPS=y we validate that the @node argument of cpumask_of_node() is a valid node_id. It however forgets to check for negative numbers. Fix this by explicitly casting to unsigned int.

(unsigned)node >= nr_node_ids

verifies: 0 <= node < nr_node_ids

Also ammend the error message to match the condition.

---
 arch/x86/mm/numa.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
arch/x86/mm/numa.c +2 -2
index e6dad600614c2..4123100e0eafe 100644
@@ -861,9 +861,9 @@ void numa_remove_cpu(int cpu)
  */
 const struct cpumask *cpumask_of_node(int node)
 {
-	if (node >= nr_node_ids) {
+	if ((unsigned)node >= nr_node_ids) {
 		printk(KERN_WARNING
-			"cpumask_of_node(%d): node > nr_node_ids(%u)\n",
+			"cpumask_of_node(%d): (unsigned)node >= nr_node_ids(%u)\n",
 			node, nr_node_ids);
 		dump_stack();
 		return cpu_none_mask;
Signature · 1 line
2.20.1
author maintainer
#182 lore raw

PATCH 179/313 irqchip/sifive-plic: set max threshold for ignored handlers

In reply to Greg Kroah-Hartman

From: Christoph Hellwig <[email protected]>

[ Upstream commit 9ce06497c2722a0f9109e4cc3ce35b7a69617886 ]

When running in M-mode, the S-mode plic handlers are still listed in the device tree. Ignore them by setting the maximum threshold.

---
 drivers/irqchip/irq-sifive-plic.c | 12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)
drivers/irqchip/irq-sifive-plic.c +10 -2
index cf755964f2f8b..c72c036aea768 100644
@@ -244,6 +244,7 @@ static int __init plic_init(struct device_node *node,
 		struct plic_handler *handler;
 		irq_hw_number_t hwirq;
 		int cpu, hartid;
+		u32 threshold = 0;
 
 		if (of_irq_parse_one(node, i, &parent)) {
 			pr_err("failed to parse parent for context %d.\n", i);
@@ -266,10 +267,16 @@ static int __init plic_init(struct device_node *node,
 			continue;
 		}
 
+		/*
+		 * When running in M-mode we need to ignore the S-mode handler.
+		 * Here we assume it always comes later, but that might be a
+		 * little fragile.
+		 */
 		handler = per_cpu_ptr(&plic_handlers, cpu);
 		if (handler->present) {
 			pr_warn("handler already present for context %d.\n", i);
-			continue;
+			threshold = 0xffffffff;
+			goto done;
 		}
 
 		handler->present = true;
@@ -279,8 +286,9 @@ static int __init plic_init(struct device_node *node,
 		handler->enable_base =
 			plic_regs + ENABLE_BASE + i * ENABLE_PER_HART;
 
+done:
 		/* priority must be > threshold to trigger an interrupt */
-		writel(0, handler->hart_base + CONTEXT_THRESHOLD);
+		writel(threshold, handler->hart_base + CONTEXT_THRESHOLD);
 		for (hwirq = 1; hwirq <= nr_irqs; hwirq++)
 			plic_toggle(handler, hwirq, 0);
 		nr_handlers++;
Signature · 1 line
2.20.1
author maintainer
#183 lore raw

PATCH 180/313 s390/crypto: xts-aes-s390 fix extra run-time crypto self tests finding

In reply to Greg Kroah-Hartman

From: Harald Freudenberger <[email protected]>

[ Upstream commit 9e323d45ba94262620a073a3f9945ca927c07c71 ]

With 'extra run-time crypto self tests' enabled, the selftest for s390-xts fails with

alg: skcipher: xts-aes-s390 encryption unexpectedly succeeded on test vector "random: len=0 klen=64"; expected_error=-22, cfg="random: inplace use_digest nosimd src_divs=[2.61%@+4006, 84.44%@+21, 1.55%@+13, 4.50%@+344, 4.26%@+21, 2.64%@+27]"

This special case with nbytes=0 is not handled correctly and this fix now makes sure that -EINVAL is returned when there is en/decrypt called with 0 bytes to en/decrypt.

---
 arch/s390/crypto/aes_s390.c | 6 ++++++
 1 file changed, 6 insertions(+)
arch/s390/crypto/aes_s390.c +6 -0
index d00f84add5f4c..6d2dbb5089d5c 100644
@@ -586,6 +586,9 @@ static int xts_aes_encrypt(struct blkcipher_desc *desc,
 	struct s390_xts_ctx *xts_ctx = crypto_blkcipher_ctx(desc->tfm);
 	struct blkcipher_walk walk;
 
+	if (!nbytes)
+		return -EINVAL;
+
 	if (unlikely(!xts_ctx->fc))
 		return xts_fallback_encrypt(desc, dst, src, nbytes);
 
@@ -600,6 +603,9 @@ static int xts_aes_decrypt(struct blkcipher_desc *desc,
 	struct s390_xts_ctx *xts_ctx = crypto_blkcipher_ctx(desc->tfm);
 	struct blkcipher_walk walk;
 
+	if (!nbytes)
+		return -EINVAL;
+
 	if (unlikely(!xts_ctx->fc))
 		return xts_fallback_decrypt(desc, dst, src, nbytes);
 
Signature · 1 line
2.20.1
author maintainer
#184 lore raw

PATCH 181/313 irqchip/gic-v3-its: Fix LPI release for Multi-MSI devices

In reply to Greg Kroah-Hartman

From: Marc Zyngier <[email protected]>

[ Upstream commit c9c96e30ecaa0aafa225aa1a5392cb7db17c7a82 ]

When allocating a range of LPIs for a Multi-MSI capable device, this allocation extended to the closest power of 2.

But on the release path, the interrupts are released one by one. This results in not releasing the "extra" range, leaking the its_device. Trying to reprobe the device will then fail.

Fix it by releasing the LPIs the same way we allocate them.

Fixes 8208d1708b88 ("irqchip/gic-v3-its: Align PCI Multi-MSI allocation on their size")
---
 drivers/irqchip/irq-gic-v3-its.c | 9 ++++-----
 1 file changed, 4 insertions(+), 5 deletions(-)
drivers/irqchip/irq-gic-v3-its.c +4 -5
index 20e5482d91b94..fca8b90028522 100644
@@ -2641,14 +2641,13 @@ static void its_irq_domain_free(struct irq_domain *domain, unsigned int virq,
 	struct its_node *its = its_dev->its;
 	int i;
 
+	bitmap_release_region(its_dev->event_map.lpi_map,
+			      its_get_event_id(irq_domain_get_irq_data(domain, virq)),
+			      get_count_order(nr_irqs));
+
 	for (i = 0; i < nr_irqs; i++) {
 		struct irq_data *data = irq_domain_get_irq_data(domain,
 								virq + i);
-		u32 event = its_get_event_id(data);
-
-		/* Mark interrupt index as unused */
-		clear_bit(event, its_dev->event_map.lpi_map);
-
 		/* Nuke the entry in the domain */
 		irq_domain_reset_irq_data(data);
 	}
Signature · 1 line
2.20.1
author maintainer
#185 lore raw

PATCH 182/313 x86/cpu: Add Tiger Lake to Intel family

In reply to Greg Kroah-Hartman

From: Gayatri Kammela <[email protected]>

[ Upstream commit 6e1c32c5dbb4b90eea8f964c2869d0bde050dbe0 ]

Add the model numbers/CPUIDs of Tiger Lake mobile and desktop to the Intel family.

---
 arch/x86/include/asm/intel-family.h | 3 +++
 1 file changed, 3 insertions(+)
arch/x86/include/asm/intel-family.h +3 -0
index f60ddd655c787..82a57d344b9bc 100644
@@ -58,6 +58,9 @@
 #define INTEL_FAM6_ICELAKE_MOBILE	0x7E
 #define INTEL_FAM6_ICELAKE_NNPI		0x9D
 
+#define INTEL_FAM6_TIGERLAKE_L		0x8C
+#define INTEL_FAM6_TIGERLAKE		0x8D
+
 /* "Small Core" Processors (Atom) */
 
 #define INTEL_FAM6_ATOM_BONNELL		0x1C /* Diamondville, Pineview */
Signature · 1 line
2.20.1
author maintainer
#186 lore raw

PATCH 183/313 platform/x86: intel_pmc_core: Do not ioremap RAM

In reply to Greg Kroah-Hartman

From: M. Vefa Bicakci <[email protected]>

[ Upstream commit 7d505758b1e556cdf65a5e451744fe0ae8063d17 ]

On a Xen-based PVH virtual machine with more than 4 GiB of RAM, intel_pmc_core fails initialization with the following warning message from the kernel, indicating that the driver is attempting to ioremap RAM:

ioremap on RAM at 0x00000000fe000000 - 0x00000000fe001fff

  WARNING: CPU: 1 PID: 434 at arch/x86/mm/ioremap.c:186 __ioremap_caller.constprop.0+0x2aa/0x2c0
...
  Call Trace:
   ? pmc_core_probe+0x87/0x2d0 [intel_pmc_core]
   pmc_core_probe+0x87/0x2d0 [intel_pmc_core]

This issue appears to manifest itself because of the following fallback mechanism in the driver:

	if (lpit_read_residency_count_address(&slp_s0_addr))
		pmcdev->base_addr = PMC_BASE_ADDR_DEFAULT;

The validity of address PMC_BASE_ADDR_DEFAULT (i.e., 0xFE000000) is not verified by the driver, which is what this patch introduces. With this patch, if address PMC_BASE_ADDR_DEFAULT is in RAM, then the driver will not attempt to ioremap the aforementioned address.

---
 drivers/platform/x86/intel_pmc_core.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)
drivers/platform/x86/intel_pmc_core.c +6 -2
index be6cda89dcf5b..01a530e2f8017 100644
@@ -882,10 +882,14 @@ static int pmc_core_probe(struct platform_device *pdev)
 	if (pmcdev->map == &spt_reg_map && !pci_dev_present(pmc_pci_ids))
 		pmcdev->map = &cnp_reg_map;
 
-	if (lpit_read_residency_count_address(&slp_s0_addr))
+	if (lpit_read_residency_count_address(&slp_s0_addr)) {
 		pmcdev->base_addr = PMC_BASE_ADDR_DEFAULT;
-	else
+
+		if (page_is_ram(PHYS_PFN(pmcdev->base_addr)))
+			return -ENODEV;
+	} else {
 		pmcdev->base_addr = slp_s0_addr - pmcdev->map->slp_s0_offset;
+	}
 
 	pmcdev->regbase = ioremap(pmcdev->base_addr,
 				  pmcdev->map->regmap_length);
Signature · 1 line
2.20.1
author maintainer
#187 lore raw

PATCH 184/313 SoC: simple-card-utils: set 0Hz to sysclk when shutdown

In reply to Greg Kroah-Hartman

From: Katsuhiro Suzuki <[email protected]>

[ Upstream commit 2458adb8f92ad4d07ef7ab27c5bafa1d3f4678d6 ]

This patch set 0Hz to sysclk when shutdown the card.

Some codecs set rate constraints that derives from sysclk. This mechanism works correctly if machine drivers give fixed frequency.

But simple-audio and audio-graph card set variable clock rate if 'mclk-fs' property exists. In this case, rate constraints will go bad scenario. For example a codec accepts three limited rates (mclk / 256, mclk / 384, mclk / 512).

Bad scenario as follows (mclk-fs = 256):
- Initialize sysclk by correct value (Ex. 12.288MHz)
- Codec set constraints of PCM rate by sysclk 48kHz (1/256), 32kHz (1/384), 24kHz (1/512)
- Play 48kHz sound, it's acceptable
- Sysclk is not changed

- Play 32kHz sound, it's acceptable
- Set sysclk to 8.192MHz (= fs * mclk-fs = 32k * 256)
- Codec set constraints of PCM rate by sysclk 32kHz (1/256), 21.33kHz (1/384), 16kHz (1/512)

- Play 48kHz again, but it's NOT acceptable because constraints do not allow 48kHz

So codecs treat 0Hz sysclk as signal of applying no constraints to avoid this problem.

---
 sound/soc/generic/simple-card-utils.c | 7 +++++++
 1 file changed, 7 insertions(+)
sound/soc/generic/simple-card-utils.c +7 -0
index f4c6375d11c7a..ef1adf87cbc8b 100644
@@ -224,10 +224,17 @@ EXPORT_SYMBOL_GPL(asoc_simple_startup);
 void asoc_simple_shutdown(struct snd_pcm_substream *substream)
 {
 	struct snd_soc_pcm_runtime *rtd = substream->private_data;
+	struct snd_soc_dai *codec_dai = rtd->codec_dai;
+	struct snd_soc_dai *cpu_dai = rtd->cpu_dai;
 	struct asoc_simple_priv *priv = snd_soc_card_get_drvdata(rtd->card);
 	struct simple_dai_props *dai_props =
 		simple_priv_to_props(priv, rtd->num);
 
+	if (dai_props->mclk_fs) {
+		snd_soc_dai_set_sysclk(codec_dai, 0, 0, SND_SOC_CLOCK_IN);
+		snd_soc_dai_set_sysclk(cpu_dai, 0, 0, SND_SOC_CLOCK_OUT);
+	}
+
 	asoc_simple_clk_disable(dai_props->cpu_dai);
 
 	asoc_simple_clk_disable(dai_props->codec_dai);
Signature · 1 line
2.20.1
author maintainer
#188 lore raw

PATCH 185/313 ASoC: dmaengine: Make the pcm->name equal to pcm->id if the name is not set

In reply to Greg Kroah-Hartman

From: Peter Ujfalusi <[email protected]>

[ Upstream commit 2ec42f3147e1610716f184b02e65d7f493eed925 ]

Some tools use the snd_pcm_info_get_name() to try to identify PCMs or for other purposes.

Currently it is left empty with the dmaengine-pcm, in this case copy the pcm->id string as pcm->name.

For example IGT is using this to find the HDMI PCM for testing audio on it.

---
 sound/soc/soc-generic-dmaengine-pcm.c | 6 ++++++
 1 file changed, 6 insertions(+)
sound/soc/soc-generic-dmaengine-pcm.c +6 -0
index 748f5f641002e..d93db2c2b5270 100644
@@ -306,6 +306,12 @@ static int dmaengine_pcm_new(struct snd_soc_pcm_runtime *rtd)
 
 		if (!dmaengine_pcm_can_report_residue(dev, pcm->chan[i]))
 			pcm->flags |= SND_DMAENGINE_PCM_FLAG_NO_RESIDUE;
+
+		if (rtd->pcm->streams[i].pcm->name[0] == '\0') {
+			strncpy(rtd->pcm->streams[i].pcm->name,
+				rtd->pcm->streams[i].pcm->id,
+				sizeof(rtd->pcm->streams[i].pcm->name));
+		}
 	}
 
 	return 0;
Signature · 1 line
2.20.1
author maintainer
#189 lore raw

PATCH 186/313 io_uring: fix wrong sequence setting logic

In reply to Greg Kroah-Hartman

From: Jackie Liu <[email protected]>

[ Upstream commit 8776f3fa15a5cd213c4dfab7ddaf557983374ea6 ]

Sqo_thread will get sqring in batches, which will cause ctx->cached_sq_head to be added in batches. if one of these sqes is set with the DRAIN flag, then he will never get a chance to process, and finally sqo_thread will not exit.

Fixes de0617e4671 ("io_uring: add support for marking commands as draining")
---
 fs/io_uring.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
fs/io_uring.c +3 -1
index 03cd8f5bba850..701936f2bde39 100644
@@ -288,6 +288,7 @@ struct io_ring_ctx {
 struct sqe_submit {
 	const struct io_uring_sqe	*sqe;
 	unsigned short			index;
+	u32				sequence;
 	bool				has_user;
 	bool				needs_lock;
 	bool				needs_fixed_file;
@@ -1894,7 +1895,7 @@ static int io_req_set_file(struct io_ring_ctx *ctx, const struct sqe_submit *s,
 
 	if (flags & IOSQE_IO_DRAIN) {
 		req->flags |= REQ_F_IO_DRAIN;
-		req->sequence = ctx->cached_sq_head - 1;
+		req->sequence = s->sequence;
 	}
 
 	if (!io_op_needs_file(s->sqe))
@@ -2050,6 +2051,7 @@ static bool io_get_sqring(struct io_ring_ctx *ctx, struct sqe_submit *s)
 	if (head < ctx->sq_entries) {
 		s->index = head;
 		s->sqe = &ctx->sq_sqes[head];
+		s->sequence = ctx->cached_sq_head;
 		ctx->cached_sq_head++;
 		return true;
 	}
Signature · 1 line
2.20.1
author maintainer
#190 lore raw

PATCH 187/313 block: make rq sector size accessible for block stats

In reply to Greg Kroah-Hartman

From: Hou Tao <[email protected]>

[ Upstream commit 3d24430694077313c75c6b89f618db09943621e4 ]

Currently rq->data_len will be decreased by partial completion or zeroed by completion, so when blk_stat_add() is invoked, data_len will be zero and there will never be samples in poll_cb because blk_mq_poll_stats_bkt() will return -1 if data_len is zero.

We could move blk_stat_add() back to __blk_mq_complete_request(), but that would make the effort of trying to call ktime_get_ns() once in vain. Instead we can reuse throtl_size field, and use it for both block stats and block throttle, and adjust the logic in blk_mq_poll_stats_bkt() accordingly.

Fixes 4bc6339a583c ("block: move blk_stat_add() to __blk_mq_end_request()")
---
 block/blk-mq.c         | 11 +++++------
 block/blk-throttle.c   |  3 ++-
 include/linux/blkdev.h | 15 ++++++++++++---
 3 files changed, 19 insertions(+), 10 deletions(-)
block/blk-mq.c +5 -6
index f934e8afe5b43..af93ca28ce4f0 100644
@@ -44,12 +44,12 @@ static void blk_mq_poll_stats_fn(struct blk_stat_callback *cb);
 
 static int blk_mq_poll_stats_bkt(const struct request *rq)
 {
-	int ddir, bytes, bucket;
+	int ddir, sectors, bucket;
 
 	ddir = rq_data_dir(rq);
-	bytes = blk_rq_bytes(rq);
+	sectors = blk_rq_stats_sectors(rq);
 
-	bucket = ddir + 2*(ilog2(bytes) - 9);
+	bucket = ddir + 2 * ilog2(sectors);
 
 	if (bucket < 0)
 		return -1;
@@ -330,6 +330,7 @@ static struct request *blk_mq_rq_ctx_init(struct blk_mq_alloc_data *data,
 	else
 		rq->start_time_ns = 0;
 	rq->io_start_time_ns = 0;
+	rq->stats_sectors = 0;
 	rq->nr_phys_segments = 0;
 #if defined(CONFIG_BLK_DEV_INTEGRITY)
 	rq->nr_integrity_segments = 0;
@@ -679,9 +680,7 @@ void blk_mq_start_request(struct request *rq)
 
 	if (test_bit(QUEUE_FLAG_STATS, &q->queue_flags)) {
 		rq->io_start_time_ns = ktime_get_ns();
-#ifdef CONFIG_BLK_DEV_THROTTLING_LOW
-		rq->throtl_size = blk_rq_sectors(rq);
-#endif
+		rq->stats_sectors = blk_rq_sectors(rq);
 		rq->rq_flags |= RQF_STATS;
 		rq_qos_issue(q, rq);
 	}
block/blk-throttle.c +2 -1
index 8ab6c81532236..ee74bffe3504d 100644
@@ -2246,7 +2246,8 @@ void blk_throtl_stat_add(struct request *rq, u64 time_ns)
 	struct request_queue *q = rq->q;
 	struct throtl_data *td = q->td;
 
-	throtl_track_latency(td, rq->throtl_size, req_op(rq), time_ns >> 10);
+	throtl_track_latency(td, blk_rq_stats_sectors(rq), req_op(rq),
+			     time_ns >> 10);
 }
 
 void blk_throtl_bio_endio(struct bio *bio)
include/linux/blkdev.h +12 -3
index 93baef66b9424..4c6754e536725 100644
@@ -202,9 +202,12 @@ struct request {
 #ifdef CONFIG_BLK_WBT
 	unsigned short wbt_flags;
 #endif
-#ifdef CONFIG_BLK_DEV_THROTTLING_LOW
-	unsigned short throtl_size;
-#endif
+	/*
+	 * rq sectors used for blk stats. It has the same value
+	 * with blk_rq_sectors(rq), except that it never be zeroed
+	 * by completion.
+	 */
+	unsigned short stats_sectors;
 
 	/*
 	 * Number of scatter-gather DMA addr+len pairs after
@@ -902,6 +905,7 @@ static inline struct request_queue *bdev_get_queue(struct block_device *bdev)
  * blk_rq_err_bytes()		: bytes left till the next error boundary
  * blk_rq_sectors()		: sectors left in the entire request
  * blk_rq_cur_sectors()		: sectors left in the current segment
+ * blk_rq_stats_sectors()	: sectors of the entire request used for stats
  */
 static inline sector_t blk_rq_pos(const struct request *rq)
 {
@@ -930,6 +934,11 @@ static inline unsigned int blk_rq_cur_sectors(const struct request *rq)
 	return blk_rq_cur_bytes(rq) >> SECTOR_SHIFT;
 }
 
+static inline unsigned int blk_rq_stats_sectors(const struct request *rq)
+{
+	return rq->stats_sectors;
+}
+
 #ifdef CONFIG_BLK_DEV_ZONED
 static inline unsigned int blk_rq_zone_no(struct request *rq)
 {
Signature · 1 line
2.20.1
author maintainer
#191 lore raw

PATCH 188/313 raid5: dont set STRIPE_HANDLE to stripe which is in batch list

In reply to Greg Kroah-Hartman

From: Guoqing Jiang <[email protected]>

[ Upstream commit 6ce220dd2f8ea71d6afc29b9a7524c12e39f374a ]

If stripe in batch list is set with STRIPE_HANDLE flag, then the stripe could be set with STRIPE_ACTIVE by the handle_stripe function. And if error happens to the batch_head at the same time, break_stripe_batch_list is called, then below warning could happen (the same report in [1]), it means a member of batch list was set with STRIPE_ACTIVE.

[7028915.431770] stripe state: 2001
[7028915.431815] ------------[ cut here ]------------
[7028915.431828] WARNING: CPU: 18 PID: 29089 at drivers/md/raid5.c:4614 break_stripe_batch_list+0x203/0x240 [raid456]
[...]
[7028915.431879] CPU: 18 PID: 29089 Comm: kworker/u82:5 Tainted: G           O    4.14.86-1-storage #4.14.86-1.2~deb9
[7028915.431881] Hardware name: Supermicro SSG-2028R-ACR24L/X10DRH-iT, BIOS 3.1 06/18/2018
[7028915.431888] Workqueue: raid5wq raid5_do_work [raid456]
[7028915.431890] task: ffff9ab0ef36d7c0 task.stack: ffffb72926f84000
[7028915.431896] RIP: 0010:break_stripe_batch_list+0x203/0x240 [raid456]
[7028915.431898] RSP: 0018:ffffb72926f87ba8 EFLAGS: 00010286
[7028915.431900] RAX: 0000000000000012 RBX: ffff9aaa84a98000 RCX: 0000000000000000
[7028915.431901] RDX: 0000000000000000 RSI: ffff9ab2bfa15458 RDI: ffff9ab2bfa15458
[7028915.431902] RBP: ffff9aaa8fb4e900 R08: 0000000000000001 R09: 0000000000002eb4
[7028915.431903] R10: 00000000ffffffff R11: 0000000000000000 R12: ffff9ab1736f1b00
[7028915.431904] R13: 0000000000000000 R14: ffff9aaa8fb4e900 R15: 0000000000000001
[7028915.431906] FS:  0000000000000000(0000) GS:ffff9ab2bfa00000(0000) knlGS:0000000000000000
[7028915.431907] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[7028915.431908] CR2: 00007ff953b9f5d8 CR3: 0000000bf4009002 CR4: 00000000003606e0
[7028915.431909] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[7028915.431910] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[7028915.431910] Call Trace:
[7028915.431923]  handle_stripe+0x8e7/0x2020 [raid456]
[7028915.431930]  ? __wake_up_common_lock+0x89/0xc0
[7028915.431935]  handle_active_stripes.isra.58+0x35f/0x560 [raid456]
[7028915.431939]  raid5_do_work+0xc6/0x1f0 [raid456]

Also commit 59fc630b8b5f9f ("RAID5: batch adjacent full stripe write") said "If a stripe is added to batch list, then only the first stripe of the list should be put to handle_list and run handle_stripe."

So don't set STRIPE_HANDLE to stripe which is already in batch list, otherwise the stripe could be put to handle_list and run handle_stripe, then the above warning could be triggered.

[1]. https://www.spinics.net/lists/raid/msg62552.html

---
 drivers/md/raid5.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/md/raid5.c +2 -1
index da94cbaa1a9ed..8d2811e436b93 100644
@@ -5719,7 +5719,8 @@ static bool raid5_make_request(struct mddev *mddev, struct bio * bi)
 				do_flush = false;
 			}
 
-			set_bit(STRIPE_HANDLE, &sh->state);
+			if (!sh->batch_head)
+				set_bit(STRIPE_HANDLE, &sh->state);
 			clear_bit(STRIPE_DELAYED, &sh->state);
 			if ((!sh->batch_head || sh == sh->batch_head) &&
 			    (bi->bi_opf & REQ_SYNC) &&
Signature · 1 line
2.20.1
author maintainer
#192 lore raw

PATCH 189/313 mmc: core: Clarify sdio_irq_pending flag for MMC_CAP2_SDIO_IRQ_NOTHREAD

In reply to Greg Kroah-Hartman

From: Ulf Hansson <[email protected]>

[ Upstream commit 36d57efb4af534dd6b442ea0b9a04aa6dfa37abe ]

The sdio_irq_pending flag is used to let host drivers indicate that it has signaled an IRQ. If that is the case and we only have a single SDIO func that have claimed an SDIO IRQ, our assumption is that we can avoid reading the SDIO_CCCR_INTx register and just call the SDIO func irq handler immediately. This makes sense, but the flag is set/cleared in a somewhat messy order, let's fix that up according to below.

First, the flag is currently set in sdio_run_irqs(), which is executed as a work that was scheduled from sdio_signal_irq(). To make it more implicit that the host have signaled an IRQ, let's instead immediately set the flag in sdio_signal_irq(). This also makes the behavior consistent with host drivers that uses the legacy, mmc_signal_sdio_irq() API. This have no functional impact, because we don't expect host drivers to call sdio_signal_irq() until after the work (sdio_run_irqs()) have been executed anyways.

Second, currently we never clears the flag when using the sdio_run_irqs() work, but only when using the sdio_irq_thread(). Let make the behavior consistent, by moving the flag to be cleared inside the common process_sdio_pending_irqs() function. Additionally, tweak the behavior of the flag slightly, by avoiding to clear it unless we processed the SDIO IRQ. The purpose with this at this point, is to keep the information about whether there have been an SDIO IRQ signaled by the host, so at system resume we can decide to process it without reading the SDIO_CCCR_INTx register.

---
 drivers/mmc/core/sdio_irq.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)
drivers/mmc/core/sdio_irq.c +6 -3
index 9f54a259a1b36..e4823ef0a0de9 100644
@@ -31,6 +31,7 @@ static int process_sdio_pending_irqs(struct mmc_host *host)
 {
 	struct mmc_card *card = host->card;
 	int i, ret, count;
+	bool sdio_irq_pending = host->sdio_irq_pending;
 	unsigned char pending;
 	struct sdio_func *func;
 
@@ -38,13 +39,16 @@ static int process_sdio_pending_irqs(struct mmc_host *host)
 	if (mmc_card_suspended(card))
 		return 0;
 
+	/* Clear the flag to indicate that we have processed the IRQ. */
+	host->sdio_irq_pending = false;
+
 	/*
 	 * Optimization, if there is only 1 function interrupt registered
 	 * and we know an IRQ was signaled then call irq handler directly.
 	 * Otherwise do the full probe.
 	 */
 	func = card->sdio_single_irq;
-	if (func && host->sdio_irq_pending) {
+	if (func && sdio_irq_pending) {
 		func->irq_handler(func);
 		return 1;
 	}
@@ -96,7 +100,6 @@ void sdio_run_irqs(struct mmc_host *host)
 {
 	mmc_claim_host(host);
 	if (host->sdio_irqs) {
-		host->sdio_irq_pending = true;
 		process_sdio_pending_irqs(host);
 		if (host->ops->ack_sdio_irq)
 			host->ops->ack_sdio_irq(host);
@@ -115,6 +118,7 @@ void sdio_irq_work(struct work_struct *work)
 
 void sdio_signal_irq(struct mmc_host *host)
 {
+	host->sdio_irq_pending = true;
 	queue_delayed_work(system_wq, &host->sdio_irq_work, 0);
 }
 EXPORT_SYMBOL_GPL(sdio_signal_irq);
@@ -160,7 +164,6 @@ static int sdio_irq_thread(void *_host)
 		if (ret)
 			break;
 		ret = process_sdio_pending_irqs(host);
-		host->sdio_irq_pending = false;
 		mmc_release_host(host);
 
 		/*
Signature · 1 line
2.20.1
author maintainer
#193 lore raw

PATCH 190/313 sched/psi: Correct overly pessimistic size calculation

In reply to Greg Kroah-Hartman

From: Miles Chen <[email protected]>

[ Upstream commit 4adcdcea717cb2d8436bef00dd689aa5bc76f11b ]

When passing a equal or more then 32 bytes long string to psi_write(), psi_write() copies 31 bytes to its buf and overwrites buf[30] with '\0'. Which makes the input string 1 byte shorter than it should be.

Fix it by copying sizeof(buf) bytes when nbytes >= sizeof(buf).

This does not cause problems in normal use case like:
"some 500000 10000000" or "full 500000 10000000" because they are less than 32 bytes in length.

	/* assuming nbytes == 35 */
	char buf[32];

	buf_size = min(nbytes, (sizeof(buf) - 1)); /* buf_size = 31 */
	if (copy_from_user(buf, user_buf, buf_size))
		return -EFAULT;

	buf[buf_size - 1] = '\0'; /* buf[30] = '\0' */

Before:

%cd /proc/pressure/
%echo "123456789|123456789|123456789|1234" > memory

 [   22.473497] nbytes=35,buf_size=31
 [   22.473775] 123456789|123456789|123456789| (print 30 chars)
 %sh: write error: Invalid argument

%echo "123456789|123456789|123456789|1" > memory

 [   64.916162] nbytes=32,buf_size=31
 [   64.916331] 123456789|123456789|123456789| (print 30 chars)
 %sh: write error: Invalid argument

After:

%cd /proc/pressure/
%echo "123456789|123456789|123456789|1234" > memory

 [  254.837863] nbytes=35,buf_size=32
 [  254.838541] 123456789|123456789|123456789|1 (print 31 chars)
 %sh: write error: Invalid argument

%echo "123456789|123456789|123456789|1" > memory

 [ 9965.714935] nbytes=32,buf_size=32
 [ 9965.715096] 123456789|123456789|123456789|1 (print 31 chars)
 %sh: write error: Invalid argument

Also remove the superfluous parentheses.

---
 kernel/sched/psi.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
kernel/sched/psi.c +1 -1
index 6e52b67b420e7..517e3719027e6 100644
@@ -1198,7 +1198,7 @@ static ssize_t psi_write(struct file *file, const char __user *user_buf,
 	if (static_branch_likely(&psi_disabled))
 		return -EOPNOTSUPP;
 
-	buf_size = min(nbytes, (sizeof(buf) - 1));
+	buf_size = min(nbytes, sizeof(buf));
 	if (copy_from_user(buf, user_buf, buf_size))
 		return -EFAULT;
 
Signature · 1 line
2.20.1
author maintainer
#194 lore raw

PATCH 191/313 mmc: sdhci: Fix incorrect switch to HS mode

In reply to Greg Kroah-Hartman

From: Al Cooper <[email protected]>

[ Upstream commit c894e33ddc1910e14d6f2a2016f60ab613fd8b37 ]

When switching from any MMC speed mode that requires 1.8v (HS200, HS400 and HS400ES) to High Speed (HS) mode, the system ends up configured for SDR12 with a 50MHz clock which is an illegal mode.

This happens because the SDHCI_CTRL_VDD_180 bit in the SDHCI_HOST_CONTROL2 register is left set and when this bit is set, the speed mode is controlled by the SDHCI_CTRL_UHS field in the SDHCI_HOST_CONTROL2 register. The SDHCI_CTRL_UHS field will end up being set to 0 (SDR12) by sdhci_set_uhs_signaling() because there is no UHS mode being set.

The fix is to change sdhci_set_uhs_signaling() to set the SDHCI_CTRL_UHS field to SDR25 (which is the same as HS) for any switch to HS mode.

This was found on a new eMMC controller that does strict checking of the speed mode and the corresponding clock rate. It caused the switch to HS400 mode to fail because part of the sequence to switch to HS400 requires a switch from HS200 to HS before going to HS400.

---
 drivers/mmc/host/sdhci.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/mmc/host/sdhci.c +3 -1
index a5dc5aae973e6..c66e66fbaeb40 100644
@@ -1849,7 +1849,9 @@ void sdhci_set_uhs_signaling(struct sdhci_host *host, unsigned timing)
 		ctrl_2 |= SDHCI_CTRL_UHS_SDR104;
 	else if (timing == MMC_TIMING_UHS_SDR12)
 		ctrl_2 |= SDHCI_CTRL_UHS_SDR12;
-	else if (timing == MMC_TIMING_UHS_SDR25)
+	else if (timing == MMC_TIMING_SD_HS ||
+		 timing == MMC_TIMING_MMC_HS ||
+		 timing == MMC_TIMING_UHS_SDR25)
 		ctrl_2 |= SDHCI_CTRL_UHS_SDR25;
 	else if (timing == MMC_TIMING_UHS_SDR50)
 		ctrl_2 |= SDHCI_CTRL_UHS_SDR50;
Signature · 1 line
2.20.1
author maintainer
#195 lore raw

PATCH 192/313 mmc: core: Add helper function to indicate if SDIO IRQs is enabled

In reply to Greg Kroah-Hartman

From: Ulf Hansson <[email protected]>

[ Upstream commit bd880b00697befb73eff7220ee20bdae4fdd487b ]

To avoid each host driver supporting SDIO IRQs, from keeping track internally about if SDIO IRQs has been claimed, let's introduce a common helper function, sdio_irq_claimed().

The function returns true if SDIO IRQs are claimed, via using the information about the number of claimed irqs. This is safe, even without any locks, as long as the helper function is called only from runtime/system suspend callbacks of the host driver.

---
 include/linux/mmc/host.h | 9 +++++++++
 1 file changed, 9 insertions(+)
include/linux/mmc/host.h +9 -0
index 7ac3755444d3d..56a8ad506072c 100644
@@ -493,6 +493,15 @@ void mmc_command_done(struct mmc_host *host, struct mmc_request *mrq);
 
 void mmc_cqe_request_done(struct mmc_host *host, struct mmc_request *mrq);
 
+/*
+ * May be called from host driver's system/runtime suspend/resume callbacks,
+ * to know if SDIO IRQs has been claimed.
+ */
+static inline bool sdio_irq_claimed(struct mmc_host *host)
+{
+	return host->sdio_irqs > 0;
+}
+
 static inline void mmc_signal_sdio_irq(struct mmc_host *host)
 {
 	host->ops->enable_sdio_irq(host, 0);
Signature · 1 line
2.20.1
author maintainer
#196 lore raw

PATCH 193/313 mmc: dw_mmc: Re-store SDIO IRQs mask at system resume

In reply to Greg Kroah-Hartman

From: Ulf Hansson <[email protected]>

[ Upstream commit 7c526608d5afb62cbc967225e2ccaacfdd142e9d ]

In cases when SDIO IRQs have been enabled, runtime suspend is prevented by the driver. However, this still means dw_mci_runtime_suspend|resume() gets called during system suspend/resume, via pm_runtime_force_suspend|resume(). This means during system suspend/resume, the register context of the dw_mmc device most likely loses its register context, even in cases when SDIO IRQs have been enabled.

To re-enable the SDIO IRQs during system resume, the dw_mmc driver currently relies on the mmc core to re-enable the SDIO IRQs when it resumes the SDIO card, but this isn't the recommended solution. Instead, it's better to deal with this locally in the dw_mmc driver, so let's do that.

---
 drivers/mmc/host/dw_mmc.c | 4 ++++
 1 file changed, 4 insertions(+)
drivers/mmc/host/dw_mmc.c +4 -0
index 60c3a06e3469a..45c3490546839 100644
@@ -3482,6 +3482,10 @@ int dw_mci_runtime_resume(struct device *dev)
 	/* Force setup bus to guarantee available clock output */
 	dw_mci_setup_bus(host->slot, true);
 
+	/* Re-enable SDIO interrupts. */
+	if (sdio_irq_claimed(host->slot->mmc))
+		__dw_mci_enable_sdio_irq(host->slot, 1);
+
 	/* Now that slots are all setup, we can enable card detect */
 	dw_mci_enable_cd(host);
 
Signature · 1 line
2.20.1
author maintainer
#197 lore raw

PATCH 194/313 raid5: dont increment read_errors on EILSEQ return

In reply to Greg Kroah-Hartman

From: Nigel Croxon <[email protected]>

[ Upstream commit b76b4715eba0d0ed574f58918b29c1b2f0fa37a8 ]

While MD continues to count read errors returned by the lower layer. If those errors are -EILSEQ, instead of -EIO, it should NOT increase the read_errors count.

When RAID6 is set up on dm-integrity target that detects massive corruption, the leg will be ejected from the array. Even if the issue is correctable with a sector re-write and the array has necessary redundancy to correct it.

The leg is ejected because it runs up the rdev->read_errors beyond conf->max_nr_stripes. The return status in dm-drypt when there is a data integrity error is -EILSEQ (BLK_STS_PROTECTION).

---
 drivers/md/raid5.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/md/raid5.c +2 -1
index 8d2811e436b93..f04e867b38211 100644
@@ -2526,7 +2526,8 @@ static void raid5_end_read_request(struct bio * bi)
 		int set_bad = 0;
 
 		clear_bit(R5_UPTODATE, &sh->dev[i].flags);
-		atomic_inc(&rdev->read_errors);
+		if (!(bi->bi_status == BLK_STS_PROTECTION))
+			atomic_inc(&rdev->read_errors);
 		if (test_bit(R5_ReadRepl, &sh->dev[i].flags))
 			pr_warn_ratelimited(
 				"md/raid:%s: read error on replacement device (sector %llu on %s).\n",
Signature · 1 line
2.20.1
author maintainer
#198 lore raw

PATCH 195/313 mmc: mtk-sd: Re-store SDIO IRQs mask at system resume

In reply to Greg Kroah-Hartman

From: Ulf Hansson <[email protected]>

[ Upstream commit 1c81d69d4c98aab56c5a7d5a810f84aefdb37e9b ]

In cases when SDIO IRQs have been enabled, runtime suspend is prevented by the driver. However, this still means msdc_runtime_suspend|resume() gets called during system suspend/resume, via pm_runtime_force_suspend|resume().

This means during system suspend/resume, the register context of the mtk-sd device most likely loses its register context, even in cases when SDIO IRQs have been enabled.

To re-enable the SDIO IRQs during system resume, the mtk-sd driver currently relies on the mmc core to re-enable the SDIO IRQs when it resumes the SDIO card, but this isn't the recommended solution. Instead, it's better to deal with this locally in the mtk-sd driver, so let's do that.

---
 drivers/mmc/host/mtk-sd.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/mmc/host/mtk-sd.c +3 -0
index 33f4b6387ef71..978c8ccce7e31 100644
@@ -2421,6 +2421,9 @@ static void msdc_restore_reg(struct msdc_host *host)
 	} else {
 		writel(host->save_para.pad_tune, host->base + tune_reg);
 	}
+
+	if (sdio_irq_claimed(host->mmc))
+		__msdc_enable_sdio_irq(host, 1);
 }
 
 static int msdc_runtime_suspend(struct device *dev)
Signature · 1 line
2.20.1
author maintainer
#199 lore raw

PATCH 196/313 libertas: Add missing sentinel at end of if_usb.c fw_table

In reply to Greg Kroah-Hartman

From: Kevin Easton <[email protected]>

[ Upstream commit 764f3f1ecffc434096e0a2b02f1a6cc964a89df6 ]

This sentinel tells the firmware loading process when to stop.

---
 drivers/net/wireless/marvell/libertas/if_usb.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/net/wireless/marvell/libertas/if_usb.c +2 -1
index f1622f0ff8c9e..fe3142d85d1e4 100644
@@ -50,7 +50,8 @@ static const struct lbs_fw_table fw_table[] = {
 	{ MODEL_8388, "libertas/usb8388_v5.bin", NULL },
 	{ MODEL_8388, "libertas/usb8388.bin", NULL },
 	{ MODEL_8388, "usb8388.bin", NULL },
-	{ MODEL_8682, "libertas/usb8682.bin", NULL }
+	{ MODEL_8682, "libertas/usb8682.bin", NULL },
+	{ 0, NULL, NULL }
 };
 
 static const struct usb_device_id if_usb_table[] = {
Signature · 1 line
2.20.1
author maintainer
#200 lore raw

PATCH 197/313 e1000e: add workaround for possible stalled packet

In reply to Greg Kroah-Hartman

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

[ Upstream commit e5e9a2ecfe780975820e157b922edee715710b66 ]

This works around a possible stalled packet issue, which may occur due to clock recovery from the PCH being too slow, when the LAN is transitioning from K1 at 1G link speed.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=204057

---
 drivers/net/ethernet/intel/e1000e/ich8lan.c | 10 ++++++++++
 drivers/net/ethernet/intel/e1000e/ich8lan.h |  2 +-
 2 files changed, 11 insertions(+), 1 deletion(-)
drivers/net/ethernet/intel/e1000e/ich8lan.c +10 -0
index cdae0efde8e64..7998a73b6a0fa 100644
@@ -1429,6 +1429,16 @@ static s32 e1000_check_for_copper_link_ich8lan(struct e1000_hw *hw)
 			else
 				phy_reg |= 0xFA;
 			e1e_wphy_locked(hw, I217_PLL_CLOCK_GATE_REG, phy_reg);
+
+			if (speed == SPEED_1000) {
+				hw->phy.ops.read_reg_locked(hw, HV_PM_CTRL,
+							    &phy_reg);
+
+				phy_reg |= HV_PM_CTRL_K1_CLK_REQ;
+
+				hw->phy.ops.write_reg_locked(hw, HV_PM_CTRL,
+							     phy_reg);
+			}
 		}
 		hw->phy.ops.release(hw);
 
drivers/net/ethernet/intel/e1000e/ich8lan.h +1 -1
index eb09c755fa172..1502895eb45dd 100644
@@ -210,7 +210,7 @@
 
 /* PHY Power Management Control */
 #define HV_PM_CTRL		PHY_REG(770, 17)
-#define HV_PM_CTRL_PLL_STOP_IN_K1_GIGA	0x100
+#define HV_PM_CTRL_K1_CLK_REQ		0x200
 #define HV_PM_CTRL_K1_ENABLE		0x4000
 
 #define I217_PLL_CLOCK_GATE_REG	PHY_REG(772, 28)
Signature · 1 line
2.20.1
author maintainer
#201 lore raw

PATCH 198/313 ALSA: hda - Add a quirk model for fixing Huawei Matebook X right speaker

In reply to Greg Kroah-Hartman

From: Tomas Espeleta <[email protected]>

[ Upstream commit a2ef03fe617a8365fb7794531b11ba587509a9b9 ]

[ This is rather a revival of the patch Tomas sent in months ago, but
applying only with the quirk model option -- tiwai ]

Hard coded coefficients to make Huawuei Matebook X right speaker work. The Matebook X has a ALC298, please refer to bug 197801 on how these numbers were reverse engineered from the Windows driver

The reversed engineered sequence represents a repeating pattern of verbs, and the only values that are changing periodically are written on indexes 0x23 and 0x25:

0x500, 0x23
0x400, VALUE1
0x500, 0x25
0x400, VALUE2

* skipped reading sequences (0x500 - 0xc00 sequences are ignored)
* static values from reverse engineering are used

NOTE: since a significant risk is still considered, this is provided as an experimental fix that isn't applied as default for now. For enabling the fix, you'll have to choose huawei-mbx-stereo via model option of snd-hda-intel module.

If we get feedback from users that this works stably, we may apply it per default.

[ Some coding style fixes and replacement with AC_VERB_* by tiwai ]

---
 Documentation/sound/hd-audio/models.rst |  3 +
 sound/pci/hda/patch_realtek.c           | 74 +++++++++++++++++++++++++
 2 files changed, 77 insertions(+)
Documentation/sound/hd-audio/models.rst +3 -0
index 7d7c191102a73..11298f0ce44db 100644
@@ -260,6 +260,9 @@ alc295-hp-x360
     HP Spectre X360 fixups
 alc-sense-combo
     Headset button support for Chrome platform
+huawei-mbx-stereo
+    Enable initialization verbs for Huawei MBX stereo speakers;
+    might be risky, try this at your own risk
 
 ALC66x/67x/892
 ==============
sound/pci/hda/patch_realtek.c +74 -0

98 lines not on this page · raw message

Signature · 1 line
2.20.1
author maintainer
#202 lore raw

PATCH 199/313 ALSA: hda - Drop unsol event handler for Intel HDMI codecs

In reply to Greg Kroah-Hartman

From: Takashi Iwai <[email protected]>

[ Upstream commit f2dbe87c5ac1f88e6007ba1f1374f4bd8a197fb6 ]

We don't need to deal with the unsol events for Intel chips that are tied with the graphics via audio component notifier. Although the presence of the audio component is checked at the beginning of hdmi_unsol_event(), better to short cut by dropping unsol_event ops.

---
 sound/pci/hda/patch_hdmi.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)
sound/pci/hda/patch_hdmi.c +8 -1
index e49c1c00f5ce1..ca0404edd939e 100644
@@ -2611,6 +2611,8 @@ static void i915_pin_cvt_fixup(struct hda_codec *codec,
 /* precondition and allocation for Intel codecs */
 static int alloc_intel_hdmi(struct hda_codec *codec)
 {
+	int err;
+
 	/* requires i915 binding */
 	if (!codec->bus->core.audio_component) {
 		codec_info(codec, "No i915 binding for Intel HDMI/DP codec\n");
@@ -2619,7 +2621,12 @@ static int alloc_intel_hdmi(struct hda_codec *codec)
 		return -ENODEV;
 	}
 
-	return alloc_generic_hdmi(codec);
+	err = alloc_generic_hdmi(codec);
+	if (err < 0)
+		return err;
+	/* no need to handle unsol events */
+	codec->patch_ops.unsol_event = NULL;
+	return 0;
 }
 
 /* parse and post-process for Intel codecs */
Signature · 1 line
2.20.1
author maintainer
#203 lore raw

PATCH 200/313 drm/amd/powerplay/smu7: enforce minimal VBITimeout (v2)

In reply to Greg Kroah-Hartman

From: Ahzo <[email protected]>

[ Upstream commit f659bb6dae58c113805f92822e4c16ddd3156b79 ]

This fixes screen corruption/flickering on 75 Hz displays.

v2: make print statement debug only (Alex)

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=102646

---
 drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c | 5 +++++
 1 file changed, 5 insertions(+)
drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c +5 -0
index 048757e8f4949..d1919d343cce4 100644
@@ -4064,6 +4064,11 @@ static int smu7_program_display_gap(struct pp_hwmgr *hwmgr)
 
 	data->frame_time_x2 = frame_time_in_us * 2 / 100;
 
+	if (data->frame_time_x2 < 280) {
+		pr_debug("%s: enforce minimal VBITimeout: %d -> 280\n", __func__, data->frame_time_x2);
+		data->frame_time_x2 = 280;
+	}
+
 	display_gap2 = pre_vbi_time_in_us * (ref_clock / 100);
 
 	cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixCG_DISPLAY_GAP_CNTL2, display_gap2);
Signature · 1 line
2.20.1
author maintainer
#204 lore raw

PATCH 201/313 media: ttusb-dec: Fix info-leak in ttusb_dec_send_command()

In reply to Greg Kroah-Hartman

From: Tomas Bortoli <[email protected]>

[ Upstream commit a10feaf8c464c3f9cfdd3a8a7ce17e1c0d498da1 ]

The function at issue does not always initialize each byte allocated for 'b' and can therefore leak uninitialized memory to a USB device in the call to usb_bulk_msg()

Use kzalloc() instead of kmalloc()

---
 drivers/media/usb/ttusb-dec/ttusb_dec.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/media/usb/ttusb-dec/ttusb_dec.c +1 -1
index 1d0afa340f47c..3198f9624b7c0 100644
@@ -319,7 +319,7 @@ static int ttusb_dec_send_command(struct ttusb_dec *dec, const u8 command,
 
 	dprintk("%s\n", __func__);
 
-	b = kmalloc(COMMAND_PACKET_SIZE + 4, GFP_KERNEL);
+	b = kzalloc(COMMAND_PACKET_SIZE + 4, GFP_KERNEL);
 	if (!b)
 		return -ENOMEM;
 
Signature · 1 line
2.20.1
author maintainer
#205 lore raw

PATCH 202/313 drm: fix module name in edid_firmware log message

In reply to Greg Kroah-Hartman

From: Jani Nikula <[email protected]>

[ Upstream commit ade925995b172f1d7410d1c665b2f47c5e99bef0 ]

The module is drm_kms_helper, not drm_kms_firmware.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=204549

Fixes ac6c35a4d8c7 ("drm: add backwards compatibility support for drm_kms_helper.edid_firmware")
---
 drivers/gpu/drm/drm_kms_helper_common.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/gpu/drm/drm_kms_helper_common.c +1 -1
index 9c5ae825c5078..69917ecd4af67 100644
@@ -39,7 +39,7 @@ MODULE_LICENSE("GPL and additional rights");
 /* Backward compatibility for drm_kms_helper.edid_firmware */
 static int edid_firmware_set(const char *val, const struct kernel_param *kp)
 {
-	DRM_NOTE("drm_kms_firmware.edid_firmware is deprecated, please use drm.edid_firmware instead.\n");
+	DRM_NOTE("drm_kms_helper.edid_firmware is deprecated, please use drm.edid_firmware instead.\n");
 
 	return __drm_set_edid_firmware_path(val);
 }
Signature · 1 line
2.20.1
author maintainer
#206 lore raw

PATCH 203/313 ALSA: hda/realtek - Blacklist PC beep for Lenovo ThinkCentre M73/93

In reply to Greg Kroah-Hartman

From: Takashi Iwai <[email protected]>

[ Upstream commit 051c78af14fcd74a22b5af45548ad9d588247cc7 ]

Lenovo ThinkCentre M73 and M93 don't seem to have a proper beep although the driver tries to probe and set up blindly. Blacklist these machines for suppressing the beep creation.

---
 sound/pci/hda/patch_realtek.c | 3 +++
 1 file changed, 3 insertions(+)
sound/pci/hda/patch_realtek.c +3 -0
index 1bec62720374d..d223a79ac934f 100644
@@ -1058,6 +1058,9 @@ static const struct snd_pci_quirk beep_white_list[] = {
 	SND_PCI_QUIRK(0x1043, 0x834a, "EeePC", 1),
 	SND_PCI_QUIRK(0x1458, 0xa002, "GA-MA790X", 1),
 	SND_PCI_QUIRK(0x8086, 0xd613, "Intel", 1),
+	/* blacklist -- no beep available */
+	SND_PCI_QUIRK(0x17aa, 0x309e, "Lenovo ThinkCentre M73", 0),
+	SND_PCI_QUIRK(0x17aa, 0x30a3, "Lenovo ThinkCentre M93", 0),
 	{}
 };
 
Signature · 1 line
2.20.1
author maintainer
#207 lore raw

PATCH 204/313 iommu/amd: Override wrong IVRS IOAPIC on Raven Ridge systems

In reply to Greg Kroah-Hartman

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

[ Upstream commit 93d051550ee02eaff9a2541d825605a7bd778027 ]

Raven Ridge systems may have malfunction touchpad or hang at boot if incorrect IVRS IOAPIC is provided by BIOS.

Users already found correct "ivrs_ioapic=" values, let's put them inside kernel to workaround buggy BIOS.

---
 drivers/iommu/Makefile           |  2 +-
 drivers/iommu/amd_iommu.h        | 14 +++++
 drivers/iommu/amd_iommu_init.c   |  5 +-
 drivers/iommu/amd_iommu_quirks.c | 92 ++++++++++++++++++++++++++++++++
 4 files changed, 111 insertions(+), 2 deletions(-)
 create mode 100644 drivers/iommu/amd_iommu.h
 create mode 100644 drivers/iommu/amd_iommu_quirks.c
drivers/iommu/Makefile +1 -1
index 8c71a15e986b5..826a5bbe0d09b 100644
@@ -10,7 +10,7 @@ obj-$(CONFIG_IOMMU_IO_PGTABLE_LPAE) += io-pgtable-arm.o
 obj-$(CONFIG_IOMMU_IOVA) += iova.o
 obj-$(CONFIG_OF_IOMMU)	+= of_iommu.o
 obj-$(CONFIG_MSM_IOMMU) += msm_iommu.o
-obj-$(CONFIG_AMD_IOMMU) += amd_iommu.o amd_iommu_init.o
+obj-$(CONFIG_AMD_IOMMU) += amd_iommu.o amd_iommu_init.o amd_iommu_quirks.o
 obj-$(CONFIG_AMD_IOMMU_DEBUGFS) += amd_iommu_debugfs.o
 obj-$(CONFIG_AMD_IOMMU_V2) += amd_iommu_v2.o
 obj-$(CONFIG_ARM_SMMU) += arm-smmu.o
drivers/iommu/amd_iommu.h new file +14 -0
new file mode 100644
index 0000000000000..12d540d9b59b0
@@ -0,0 +1,14 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+
+#ifndef AMD_IOMMU_H
+#define AMD_IOMMU_H
+
+int __init add_special_device(u8 type, u8 id, u16 *devid, bool cmd_line);
+
+#ifdef CONFIG_DMI
+void amd_iommu_apply_ivrs_quirks(void);
+#else
+static void amd_iommu_apply_ivrs_quirks(void) { }
+#endif
+
+#endif
drivers/iommu/amd_iommu_init.c +4 -1
index 07d84dbab564e..6469f51282427 100644
@@ -30,6 +30,7 @@
 #include <asm/irq_remapping.h>
 
 #include <linux/crash_dump.h>
+#include "amd_iommu.h"
 #include "amd_iommu_proto.h"
 #include "amd_iommu_types.h"
 #include "irq_remapping.h"
@@ -997,7 +998,7 @@ static void __init set_dev_entry_from_acpi(struct amd_iommu *iommu,
 	set_iommu_for_device(iommu, devid);
 }
 
-static int __init add_special_device(u8 type, u8 id, u16 *devid, bool cmd_line)
+int __init add_special_device(u8 type, u8 id, u16 *devid, bool cmd_line)
 {
 	struct devid_map *entry;
 	struct list_head *list;
@@ -1148,6 +1149,8 @@ static int __init init_iommu_from_acpi(struct amd_iommu *iommu,
 	if (ret)
 		return ret;
 
+	amd_iommu_apply_ivrs_quirks();
+
 	/*
 	 * First save the recommended feature enable bits from ACPI
 	 */
drivers/iommu/amd_iommu_quirks.c new file +92 -0

92 lines not on this page · raw message

Signature · 1 line
2.20.1
author maintainer
#208 lore raw

PATCH 205/313 zd1211rw: remove false assertion from zd_mac_clear()

In reply to Greg Kroah-Hartman

From: Oliver Neukum <[email protected]>

[ Upstream commit 7a2eb7367fdea72e448d1a847aa857f6caf8ea2f ]

The function is called before the lock which is asserted was ever used. Just remove it.

---
 drivers/net/wireless/zydas/zd1211rw/zd_mac.c | 1 -
 1 file changed, 1 deletion(-)
drivers/net/wireless/zydas/zd1211rw/zd_mac.c +0 -1
index da7e63fca9f57..a9999d10ae81f 100644
@@ -223,7 +223,6 @@ void zd_mac_clear(struct zd_mac *mac)
 {
 	flush_workqueue(zd_workqueue);
 	zd_chip_clear(&mac->chip);
-	lockdep_assert_held(&mac->lock);
 	ZD_MEMCLEAR(mac, sizeof(struct zd_mac));
 }
 
Signature · 1 line
2.20.1
author maintainer
#209 lore raw

PATCH 206/313 btrfs: delayed-inode: Kill the BUG_ON() in btrfs_delete_delayed_dir_index()

In reply to Greg Kroah-Hartman

From: Qu Wenruo <[email protected]>

[ Upstream commit 933c22a7512c5c09b1fdc46b557384efe8d03233 ]

There is one report of fuzzed image which leads to BUG_ON() in btrfs_delete_delayed_dir_index().

Although that fuzzed image can already be addressed by enhanced extent-tree error handler, it's still better to hunt down more BUG_ON().

This patch will hunt down two BUG_ON()s in btrfs_delete_delayed_dir_index():
- One for error from btrfs_delayed_item_reserve_metadata() Instead of BUG_ON(), we output an error message and free the item. And return the error.
All callers of this function handles the error by aborting current trasaction.

- One for possible EEXIST from __btrfs_add_delayed_deletion_item() That function can return -EEXIST.
We already have a good enough error message for that, only need to clean up the reserved metadata space and allocated item.

To help above cleanup, also modifiy __btrfs_remove_delayed_item() called in btrfs_release_delayed_item(), to skip unassociated item.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=203253

---
 fs/btrfs/delayed-inode.c | 13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)
fs/btrfs/delayed-inode.c +11 -2
index 43fdb2992956a..6858a05606dd3 100644
@@ -474,6 +474,9 @@ static void __btrfs_remove_delayed_item(struct btrfs_delayed_item *delayed_item)
 	struct rb_root_cached *root;
 	struct btrfs_delayed_root *delayed_root;
 
+	/* Not associated with any delayed_node */
+	if (!delayed_item->delayed_node)
+		return;
 	delayed_root = delayed_item->delayed_node->root->fs_info->delayed_root;
 
 	BUG_ON(!delayed_root);
@@ -1525,7 +1528,12 @@ int btrfs_delete_delayed_dir_index(struct btrfs_trans_handle *trans,
 	 * we have reserved enough space when we start a new transaction,
 	 * so reserving metadata failure is impossible.
 	 */
-	BUG_ON(ret);
+	if (ret < 0) {
+		btrfs_err(trans->fs_info,
+"metadata reservation failed for delayed dir item deltiona, should have been reserved");
+		btrfs_release_delayed_item(item);
+		goto end;
+	}
 
 	mutex_lock(&node->mutex);
 	ret = __btrfs_add_delayed_deletion_item(node, item);
@@ -1534,7 +1542,8 @@ int btrfs_delete_delayed_dir_index(struct btrfs_trans_handle *trans,
 			  "err add delayed dir index item(index: %llu) into the deletion tree of the delayed node(root id: %llu, inode id: %llu, errno: %d)",
 			  index, node->root->root_key.objectid,
 			  node->inode_id, ret);
-		BUG();
+		btrfs_delayed_item_release_metadata(dir->root, item);
+		btrfs_release_delayed_item(item);
 	}
 	mutex_unlock(&node->mutex);
 end:
Signature · 1 line
2.20.1
author maintainer
#210 lore raw

PATCH 207/313 btrfs: extent-tree: Make sure we only allocate extents from block groups with the same type

In reply to Greg Kroah-Hartman

From: Qu Wenruo <[email protected]>

[ Upstream commit 2a28468e525f3924efed7f29f2bc5a2926e7e19a ]

[BUG]
With fuzzed image and MIXED_GROUPS super flag, we can hit the following BUG_ON():

kernel BUG at fs/btrfs/delayed-ref.c:491!
invalid opcode: 0000 [#1] PREEMPT SMP NOPTI

  CPU: 0 PID: 1849 Comm: sync Tainted: G           O      5.2.0-custom #27
  RIP: 0010:update_existing_head_ref.cold+0x44/0x46 [btrfs]
  Call Trace:
   add_delayed_ref_head+0x20c/0x2d0 [btrfs]
   btrfs_add_delayed_tree_ref+0x1fc/0x490 [btrfs]
   btrfs_free_tree_block+0x123/0x380 [btrfs]
   __btrfs_cow_block+0x435/0x500 [btrfs]
   btrfs_cow_block+0x110/0x240 [btrfs]
   btrfs_search_slot+0x230/0xa00 [btrfs]
   ? __lock_acquire+0x105e/0x1e20
   btrfs_insert_empty_items+0x67/0xc0 [btrfs]
   alloc_reserved_file_extent+0x9e/0x340 [btrfs]
   __btrfs_run_delayed_refs+0x78e/0x1240 [btrfs]
   ? kvm_clock_read+0x18/0x30
   ? __sched_clock_gtod_offset+0x21/0x50
   btrfs_run_delayed_refs.part.0+0x4e/0x180 [btrfs]
   btrfs_run_delayed_refs+0x23/0x30 [btrfs]
   btrfs_commit_transaction+0x53/0x9f0 [btrfs]
   btrfs_sync_fs+0x7c/0x1c0 [btrfs]
   ? __ia32_sys_fdatasync+0x20/0x20
   sync_fs_one_sb+0x23/0x30
   iterate_supers+0x95/0x100
   ksys_sync+0x62/0xb0
   __ia32_sys_sync+0xe/0x20
   do_syscall_64+0x65/0x240
   entry_SYSCALL_64_after_hwframe+0x49/0xbe

[CAUSE]
This situation is caused by several factors:
- Fuzzed image
The extent tree of this fs missed one backref for extent tree root. So we can allocated space from that slot.

- MIXED_BG feature
Super block has MIXED_BG flag.

- No mixed block groups exists
All block groups are just regular ones.

This makes data space_info->block_groups[] contains metadata block groups. And when we reserve space for data, we can use space in metadata block group.

Then we hit the following file operations:

- fallocate
We need to allocate data extents.
find_free_extent() choose to use the metadata block to allocate space from, and choose the space of extent tree root, since its backref is missing.

This generate one delayed ref head with is_data = 1.

- extent tree update
We need to update extent tree at run_delayed_ref time.

This generate one delayed ref head with is_data = 0, for the same bytenr of old extent tree root.

Then we trigger the BUG_ON().

[FIX]
The quick fix here is to check block_group->flags before using it.

The problem can only happen for MIXED_GROUPS fs. Regular filesystems won't have space_info with DATA|METADATA flag, and no way to hit the bug.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=203255

---
 fs/btrfs/extent-tree.c | 8 ++++++++
 1 file changed, 8 insertions(+)
fs/btrfs/extent-tree.c +8 -0
index b8f4720879021..37865929fdc22 100644
@@ -7875,6 +7875,14 @@ static noinline int find_free_extent(struct btrfs_fs_info *fs_info,
 			 */
 			if ((flags & extra) && !(block_group->flags & extra))
 				goto loop;
+
+			/*
+			 * This block group has different flags than we want.
+			 * It's possible that we have MIXED_GROUP flag but no
+			 * block group is mixed.  Just skip such block group.
+			 */
+			btrfs_release_block_group(block_group, delalloc);
+			continue;
 		}
 
 have_block_group:
Signature · 1 line
2.20.1
author maintainer
#211 lore raw

PATCH 208/313 btrfs: tree-checker: Add ROOT_ITEM check

In reply to Greg Kroah-Hartman

From: Qu Wenruo <[email protected]>

[ Upstream commit 259ee7754b6793af8bdd77f9ca818bc41cfe9541 ]

This patch will introduce ROOT_ITEM check, which includes:
- Key->objectid and key->offset check
Currently only some easy check, e.g. 0 as rootid is invalid.

- Item size check
Root item size is fixed.

- Generation checks
Generation, generation_v2 and last_snapshot should not be greater than super generation + 1

- Level and alignment check
Level should be in [0, 7], and bytenr must be aligned to sector size.

- Flags check

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=203261

---
 fs/btrfs/tree-checker.c | 92 +++++++++++++++++++++++++++++++++++++++++
 1 file changed, 92 insertions(+)
fs/btrfs/tree-checker.c +92 -0
index ccd5706199d76..d83adda6c090a 100644
@@ -821,6 +821,95 @@ static int check_inode_item(struct extent_buffer *leaf,
 	return 0;
 }
 
+static int check_root_item(struct extent_buffer *leaf, struct btrfs_key *key,
+			   int slot)
+{
+	struct btrfs_fs_info *fs_info = leaf->fs_info;
+	struct btrfs_root_item ri;
+	const u64 valid_root_flags = BTRFS_ROOT_SUBVOL_RDONLY |
+				     BTRFS_ROOT_SUBVOL_DEAD;
+
+	/* No such tree id */
+	if (key->objectid == 0) {
+		generic_err(leaf, slot, "invalid root id 0");
+		return -EUCLEAN;
+	}
+
+	/*
+	 * Some older kernel may create ROOT_ITEM with non-zero offset, so here
+	 * we only check offset for reloc tree whose key->offset must be a
+	 * valid tree.
+	 */
+	if (key->objectid == BTRFS_TREE_RELOC_OBJECTID && key->offset == 0) {
+		generic_err(leaf, slot, "invalid root id 0 for reloc tree");
+		return -EUCLEAN;
+	}
+
+	if (btrfs_item_size_nr(leaf, slot) != sizeof(ri)) {
+		generic_err(leaf, slot,
+			    "invalid root item size, have %u expect %zu",
+			    btrfs_item_size_nr(leaf, slot), sizeof(ri));
+	}
+
+	read_extent_buffer(leaf, &ri, btrfs_item_ptr_offset(leaf, slot),
+			   sizeof(ri));
+
+	/* Generation related */
+	if (btrfs_root_generation(&ri) >
+	    btrfs_super_generation(fs_info->super_copy) + 1) {
+		generic_err(leaf, slot,
+			"invalid root generation, have %llu expect (0, %llu]",
+			    btrfs_root_generation(&ri),
+			    btrfs_super_generation(fs_info->super_copy) + 1);
+		return -EUCLEAN;
+	}
+	if (btrfs_root_generation_v2(&ri) >
+	    btrfs_super_generation(fs_info->super_copy) + 1) {
+		generic_err(leaf, slot,
+		"invalid root v2 generation, have %llu expect (0, %llu]",
+			    btrfs_root_generation_v2(&ri),
+			    btrfs_super_generation(fs_info->super_copy) + 1);
+		return -EUCLEAN;
+	}
+	if (btrfs_root_last_snapshot(&ri) >
+	    btrfs_super_generation(fs_info->super_copy) + 1) {
+		generic_err(leaf, slot,
+		"invalid root last_snapshot, have %llu expect (0, %llu]",
+			    btrfs_root_last_snapshot(&ri),
+			    btrfs_super_generation(fs_info->super_copy) + 1);
+		return -EUCLEAN;
+	}
+
+	/* Alignment and level check */
+	if (!IS_ALIGNED(btrfs_root_bytenr(&ri), fs_info->sectorsize)) {
+		generic_err(leaf, slot,
+		"invalid root bytenr, have %llu expect to be aligned to %u",
+			    btrfs_root_bytenr(&ri), fs_info->sectorsize);
+		return -EUCLEAN;
+	}
+	if (btrfs_root_level(&ri) >= BTRFS_MAX_LEVEL) {
+		generic_err(leaf, slot,
+			    "invalid root level, have %u expect [0, %u]",
+			    btrfs_root_level(&ri), BTRFS_MAX_LEVEL - 1);
+		return -EUCLEAN;
+	}
+	if (ri.drop_level >= BTRFS_MAX_LEVEL) {
+		generic_err(leaf, slot,
+			    "invalid root level, have %u expect [0, %u]",
+			    ri.drop_level, BTRFS_MAX_LEVEL - 1);
+		return -EUCLEAN;
+	}
+
+	/* Flags check */
+	if (btrfs_root_flags(&ri) & ~valid_root_flags) {
+		generic_err(leaf, slot,
+			    "invalid root flags, have 0x%llx expect mask 0x%llx",
+			    btrfs_root_flags(&ri), valid_root_flags);
+		return -EUCLEAN;
+	}
+	return 0;
+}
+
 /*
  * Common point to switch the item-specific validation.
  */
@@ -856,6 +945,9 @@ static int check_leaf_item(struct extent_buffer *leaf,
 	case BTRFS_INODE_ITEM_KEY:
 		ret = check_inode_item(leaf, key, slot);
 		break;
+	case BTRFS_ROOT_ITEM_KEY:
+		ret = check_root_item(leaf, key, slot);
+		break;
 	}
 	return ret;
 }
Signature · 1 line
2.20.1
author maintainer
#212 lore raw

PATCH 209/313 btrfs: Detect unbalanced tree with empty leaf before crashing btree operations

In reply to Greg Kroah-Hartman

From: Qu Wenruo <[email protected]>

[ Upstream commit 62fdaa52a3d00a875da771719b6dc537ca79fce1 ]

[BUG]
With crafted image, btrfs will panic at btree operations:

kernel BUG at fs/btrfs/ctree.c:3894!
invalid opcode: 0000 [#1] SMP PTI

  CPU: 0 PID: 1138 Comm: btrfs-transacti Not tainted 5.0.0-rc8+ #9
  RIP: 0010:__push_leaf_left+0x6b6/0x6e0
  RSP: 0018:ffffc0bd4128b990 EFLAGS: 00010246
  RAX: 0000000000000000 RBX: ffffa0a4ab8f0e38 RCX: 0000000000000000
  RDX: ffffa0a280000000 RSI: 0000000000000000 RDI: ffffa0a4b3814000
  RBP: ffffc0bd4128ba38 R08: 0000000000001000 R09: ffffc0bd4128b948
  R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000240
  R13: ffffa0a4b556fb60 R14: ffffa0a4ab8f0af0 R15: ffffa0a4ab8f0af0
  FS: 0000000000000000(0000) GS:ffffa0a4b7a00000(0000) knlGS:0000000000000000
  CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
  CR2: 00007f2461c80020 CR3: 000000022b32a006 CR4: 00000000000206f0
  Call Trace:
  ? _cond_resched+0x1a/0x50
  push_leaf_left+0x179/0x190
  btrfs_del_items+0x316/0x470
  btrfs_del_csums+0x215/0x3a0
  __btrfs_free_extent.isra.72+0x5a7/0xbe0
  __btrfs_run_delayed_refs+0x539/0x1120
  btrfs_run_delayed_refs+0xdb/0x1b0
  btrfs_commit_transaction+0x52/0x950
  ? start_transaction+0x94/0x450
  transaction_kthread+0x163/0x190
  kthread+0x105/0x140
  ? btrfs_cleanup_transaction+0x560/0x560
  ? kthread_destroy_worker+0x50/0x50
  ret_from_fork+0x35/0x40
  Modules linked in:
  ---[ end trace c2425e6e89b5558f ]---

[CAUSE]
The offending csum tree looks like this:

checksum tree key (CSUM_TREE ROOT_ITEM 0)
node 29741056 level 1 items 14 free 107 generation 19 owner CSUM_TREE

	  ...
	  key (EXTENT_CSUM EXTENT_CSUM 85975040) block 29630464 gen 17
	  key (EXTENT_CSUM EXTENT_CSUM 89911296) block 29642752 gen 17 <<<
	  key (EXTENT_CSUM EXTENT_CSUM 92274688) block 29646848 gen 17
	  ...

leaf 29630464 items 6 free space 1 generation 17 owner CSUM_TREE

	  item 0 key (EXTENT_CSUM EXTENT_CSUM 85975040) itemoff 3987 itemsize 8
		  range start 85975040 end 85983232 length 8192
	  ...

leaf 29642752 items 0 free space 3995 generation 17 owner 0
^ empty leaf invalid owner ^

leaf 29646848 items 1 free space 602 generation 17 owner CSUM_TREE

	  item 0 key (EXTENT_CSUM EXTENT_CSUM 92274688) itemoff 627 itemsize 3368
		  range start 92274688 end 95723520 length 3448832

So we have a corrupted csum tree where one tree leaf is completely empty, causing unbalanced btree, thus leading to unexpected btree balance error.

[FIX]
For this particular case, we handle it in two directions to catch it:
- Check if the tree block is empty through btrfs_verify_level_key() So that invalid tree blocks won't be read out through btrfs_search_slot() and its variants.

- Check 0 tree owner in tree checker
NO tree is using 0 as its tree owner, detect it and reject at tree block read time.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=202821

---
 fs/btrfs/disk-io.c      | 10 ++++++++++
 fs/btrfs/tree-checker.c |  6 ++++++
 2 files changed, 16 insertions(+)
fs/btrfs/disk-io.c +10 -0
index deb74a8c191a4..6d26d38a2e12c 100644
@@ -423,6 +423,16 @@ int btrfs_verify_level_key(struct extent_buffer *eb, int level,
 	 */
 	if (btrfs_header_generation(eb) > fs_info->last_trans_committed)
 		return 0;
+
+	/* We have @first_key, so this @eb must have at least one item */
+	if (btrfs_header_nritems(eb) == 0) {
+		btrfs_err(fs_info,
+		"invalid tree nritems, bytenr=%llu nritems=0 expect >0",
+			  eb->start);
+		WARN_ON(IS_ENABLED(CONFIG_BTRFS_DEBUG));
+		return -EUCLEAN;
+	}
+
 	if (found_level)
 		btrfs_node_key_to_cpu(eb, &found_key, 0);
 	else
fs/btrfs/tree-checker.c +6 -0
index d83adda6c090a..9634cae1e1b1d 100644
@@ -991,6 +991,12 @@ static int check_leaf(struct extent_buffer *leaf, bool check_item_data)
 				    owner);
 			return -EUCLEAN;
 		}
+		/* Unknown tree */
+		if (owner == 0) {
+			generic_err(leaf, 0,
+				"invalid owner, root 0 is not defined");
+			return -EUCLEAN;
+		}
 		return 0;
 	}
 
Signature · 1 line
2.20.1
author maintainer
#213 lore raw

PATCH 210/313 kvm: Nested KVM MMUs need PAE root too

In reply to Greg Kroah-Hartman

From: Jiří Paleček <[email protected]>

[ Upstream commit 1cfff4d9a5d01fa61e5768a6afffc81ae1c8ecb9 ]

On AMD processors, in PAE 32bit mode, nested KVM instances don't work. The L0 host get a kernel OOPS, which is related to arch.mmu->pae_root being NULL.

The reason for this is that when setting up nested KVM instance, arch.mmu is set to &arch.guest_mmu (while normally, it would be &arch.root_mmu). However, the initialization and allocation of pae_root only creates it in root_mmu. KVM code (ie. in mmu_alloc_shadow_roots) then accesses arch.mmu->pae_root, which is the unallocated arch.guest_mmu->pae_root.

This fix just allocates (and frees) pae_root in both guest_mmu and root_mmu (and also lm_root if it was allocated). The allocation is subject to previous restrictions ie. it won't allocate anything on 64-bit and AFAIK not on Intel.

---
 arch/x86/kvm/mmu.c | 30 ++++++++++++++++++++++--------
 1 file changed, 22 insertions(+), 8 deletions(-)
arch/x86/kvm/mmu.c +22 -8
index 66055ca29b6b4..9130152d5ed83 100644
@@ -5607,13 +5607,13 @@ slot_handle_leaf(struct kvm *kvm, struct kvm_memory_slot *memslot,
 				 PT_PAGE_TABLE_LEVEL, lock_flush_tlb);
 }
 
-static void free_mmu_pages(struct kvm_vcpu *vcpu)
+static void free_mmu_pages(struct kvm_mmu *mmu)
 {
-	free_page((unsigned long)vcpu->arch.mmu->pae_root);
-	free_page((unsigned long)vcpu->arch.mmu->lm_root);
+	free_page((unsigned long)mmu->pae_root);
+	free_page((unsigned long)mmu->lm_root);
 }
 
-static int alloc_mmu_pages(struct kvm_vcpu *vcpu)
+static int alloc_mmu_pages(struct kvm_vcpu *vcpu, struct kvm_mmu *mmu)
 {
 	struct page *page;
 	int i;
@@ -5634,9 +5634,9 @@ static int alloc_mmu_pages(struct kvm_vcpu *vcpu)
 	if (!page)
 		return -ENOMEM;
 
-	vcpu->arch.mmu->pae_root = page_address(page);
+	mmu->pae_root = page_address(page);
 	for (i = 0; i < 4; ++i)
-		vcpu->arch.mmu->pae_root[i] = INVALID_PAGE;
+		mmu->pae_root[i] = INVALID_PAGE;
 
 	return 0;
 }
@@ -5644,6 +5644,7 @@ static int alloc_mmu_pages(struct kvm_vcpu *vcpu)
 int kvm_mmu_create(struct kvm_vcpu *vcpu)
 {
 	uint i;
+	int ret;
 
 	vcpu->arch.mmu = &vcpu->arch.root_mmu;
 	vcpu->arch.walk_mmu = &vcpu->arch.root_mmu;
@@ -5661,7 +5662,19 @@ int kvm_mmu_create(struct kvm_vcpu *vcpu)
 		vcpu->arch.guest_mmu.prev_roots[i] = KVM_MMU_ROOT_INFO_INVALID;
 
 	vcpu->arch.nested_mmu.translate_gpa = translate_nested_gpa;
-	return alloc_mmu_pages(vcpu);
+
+	ret = alloc_mmu_pages(vcpu, &vcpu->arch.guest_mmu);
+	if (ret)
+		return ret;
+
+	ret = alloc_mmu_pages(vcpu, &vcpu->arch.root_mmu);
+	if (ret)
+		goto fail_allocate_root;
+
+	return ret;
+ fail_allocate_root:
+	free_mmu_pages(&vcpu->arch.guest_mmu);
+	return ret;
 }
 
 
@@ -6134,7 +6147,8 @@ unsigned long kvm_mmu_calculate_default_mmu_pages(struct kvm *kvm)
 void kvm_mmu_destroy(struct kvm_vcpu *vcpu)
 {
 	kvm_mmu_unload(vcpu);
-	free_mmu_pages(vcpu);
+	free_mmu_pages(&vcpu->arch.root_mmu);
+	free_mmu_pages(&vcpu->arch.guest_mmu);
 	mmu_free_memory_caches(vcpu);
 }
 
Signature · 1 line
2.20.1
author maintainer
#214 lore raw

PATCH 211/313 media: omap3isp: Set device on omap3isp subdevs

In reply to Greg Kroah-Hartman

From: Sakari Ailus <[email protected]>

[ Upstream commit e9eb103f027725053a4b02f93d7f2858b56747ce ]

The omap3isp driver registered subdevs without the dev field being set. Do that now.

---
 drivers/media/platform/omap3isp/ispccdc.c    | 1 +
 drivers/media/platform/omap3isp/ispccp2.c    | 1 +
 drivers/media/platform/omap3isp/ispcsi2.c    | 1 +
 drivers/media/platform/omap3isp/isppreview.c | 1 +
 drivers/media/platform/omap3isp/ispresizer.c | 1 +
 drivers/media/platform/omap3isp/ispstat.c    | 2 ++
 6 files changed, 7 insertions(+)
drivers/media/platform/omap3isp/ispccdc.c +1 -0
index 1ba8a5ba343f6..e2f336c715a4d 100644
@@ -2602,6 +2602,7 @@ int omap3isp_ccdc_register_entities(struct isp_ccdc_device *ccdc,
 	int ret;
 
 	/* Register the subdev and video node. */
+	ccdc->subdev.dev = vdev->mdev->dev;
 	ret = v4l2_device_register_subdev(vdev, &ccdc->subdev);
 	if (ret < 0)
 		goto error;
drivers/media/platform/omap3isp/ispccp2.c +1 -0
index efca45bb02c8b..d0a49cdfd22d2 100644
@@ -1031,6 +1031,7 @@ int omap3isp_ccp2_register_entities(struct isp_ccp2_device *ccp2,
 	int ret;
 
 	/* Register the subdev and video nodes. */
+	ccp2->subdev.dev = vdev->mdev->dev;
 	ret = v4l2_device_register_subdev(vdev, &ccp2->subdev);
 	if (ret < 0)
 		goto error;
drivers/media/platform/omap3isp/ispcsi2.c +1 -0
index e85917f4a50cc..fd493c5e4e24f 100644
@@ -1198,6 +1198,7 @@ int omap3isp_csi2_register_entities(struct isp_csi2_device *csi2,
 	int ret;
 
 	/* Register the subdev and video nodes. */
+	csi2->subdev.dev = vdev->mdev->dev;
 	ret = v4l2_device_register_subdev(vdev, &csi2->subdev);
 	if (ret < 0)
 		goto error;
drivers/media/platform/omap3isp/isppreview.c +1 -0
index 40e22400cf5ec..97d660606d984 100644
@@ -2225,6 +2225,7 @@ int omap3isp_preview_register_entities(struct isp_prev_device *prev,
 	int ret;
 
 	/* Register the subdev and video nodes. */
+	prev->subdev.dev = vdev->mdev->dev;
 	ret = v4l2_device_register_subdev(vdev, &prev->subdev);
 	if (ret < 0)
 		goto error;
drivers/media/platform/omap3isp/ispresizer.c +1 -0
index 21ca6954df72f..78d9dd7ea2da7 100644
@@ -1681,6 +1681,7 @@ int omap3isp_resizer_register_entities(struct isp_res_device *res,
 	int ret;
 
 	/* Register the subdev and video nodes. */
+	res->subdev.dev = vdev->mdev->dev;
 	ret = v4l2_device_register_subdev(vdev, &res->subdev);
 	if (ret < 0)
 		goto error;
drivers/media/platform/omap3isp/ispstat.c +2 -0
index ca7bb8497c3da..cb5841f628cff 100644
@@ -1026,6 +1026,8 @@ void omap3isp_stat_unregister_entities(struct ispstat *stat)
 int omap3isp_stat_register_entities(struct ispstat *stat,
 				    struct v4l2_device *vdev)
 {
+	stat->subdev.dev = vdev->mdev->dev;
+
 	return v4l2_device_register_subdev(vdev, &stat->subdev);
 }
 
Signature · 1 line
2.20.1
author maintainer
#215 lore raw

PATCH 212/313 PM / devfreq: passive: fix compiler warning

In reply to Greg Kroah-Hartman

From: MyungJoo Ham <[email protected]>

[ Upstream commit 0465814831a926ce2f83e8f606d067d86745234e ]

The recent commit of
PM / devfreq: passive: Use non-devm notifiers had incurred compiler warning, "unused variable 'dev'".

---
 drivers/devfreq/governor_passive.c | 1 -
 1 file changed, 1 deletion(-)
drivers/devfreq/governor_passive.c +0 -1
index da485477065c5..be6eeab9c814e 100644
@@ -149,7 +149,6 @@ static int devfreq_passive_notifier_call(struct notifier_block *nb,
 static int devfreq_passive_event_handler(struct devfreq *devfreq,
 				unsigned int event, void *data)
 {
-	struct device *dev = devfreq->dev.parent;
 	struct devfreq_passive_data *p_data
 			= (struct devfreq_passive_data *)devfreq->data;
 	struct devfreq *parent = (struct devfreq *)p_data->parent;
Signature · 1 line
2.20.1
author maintainer
#216 lore raw

PATCH 213/313 ARM: dts: logicpd-torpedo-baseboard: Fix missing video

In reply to Greg Kroah-Hartman

From: Adam Ford <[email protected]>

commit f9f5518a38684e031d913f40482721ff553f5ba2 upstream.

A previous commit removed the panel-dpi driver, which made the Torpedo video stop working because it relied on the dpi driver for setting video timings. Now that the simple-panel driver is available in omap2plus, this patch migrates the Torpedo dev kits to use a similar panel and remove the manual timing requirements.

Fixes 8bf4b1621178 ("drm/omap: Remove panel-dpi driver")
---
 arch/arm/boot/dts/logicpd-torpedo-baseboard.dtsi |   37 +++--------------------
 1 file changed, 6 insertions(+), 31 deletions(-)
arch/arm/boot/dts/logicpd-torpedo-baseboard.dtsi +6 -31
@@ -108,7 +108,6 @@
 &dss {
 	status = "ok";
 	vdds_dsi-supply = <&vpll2>;
-	vdda_video-supply = <&video_reg>;
 	pinctrl-names = "default";
 	pinctrl-0 = <&dss_dpi_pins1>;
 	port {
@@ -124,44 +123,20 @@
 		display0 = &lcd0;
 	};
 
-	video_reg: video_reg {
-		pinctrl-names = "default";
-		pinctrl-0 = <&panel_pwr_pins>;
-		compatible = "regulator-fixed";
-		regulator-name = "fixed-supply";
-		regulator-min-microvolt = <3300000>;
-		regulator-max-microvolt = <3300000>;
-		gpio = <&gpio5 27 GPIO_ACTIVE_HIGH>;	/* gpio155, lcd INI */
-	};
-
 	lcd0: display {
-		compatible = "panel-dpi";
+		/* This isn't the exact LCD, but the timings meet spec */
+		/* To make it work, set CONFIG_OMAP2_DSS_MIN_FCK_PER_PCK=4 */
+		compatible = "newhaven,nhd-4.3-480272ef-atxl";
 		label = "15";
-		status = "okay";
-		/* default-on; */
 		pinctrl-names = "default";
-
+		pinctrl-0 = <&panel_pwr_pins>;
+		backlight = <&bl>;
+		enable-gpios = <&gpio5 27 GPIO_ACTIVE_HIGH>;
 		port {
 			lcd_in: endpoint {
 				remote-endpoint = <&dpi_out>;
 			};
 		};
-
-		panel-timing {
-			clock-frequency = <9000000>;
-			hactive = <480>;
-			vactive = <272>;
-			hfront-porch = <3>;
-			hback-porch = <2>;
-			hsync-len = <42>;
-			vback-porch = <3>;
-			vfront-porch = <4>;
-			vsync-len = <11>;
-			hsync-active = <0>;
-			vsync-active = <0>;
-			de-active = <1>;
-			pixelclk-active = <1>;
-		};
 	};
 
 	bl: backlight {
author maintainer
#217 lore raw

PATCH 214/313 ARM: omap2plus_defconfig: Fix missing video

In reply to Greg Kroah-Hartman

From: Adam Ford <[email protected]>

commit 4957eccf979b025286b39388fd1a60cde601a10a upstream.

When the panel-dpi driver was removed, the simple-panels driver was never enabled, so anyone who used the panel-dpi driver lost video, and those who used it inconjunction with simple-panels would have to manually enable CONFIG_DRM_PANEL_SIMPLE.

This patch makes CONFIG_DRM_PANEL_SIMPLE a module in the same way the deprecated panel-dpi was.

Fixes 8bf4b1621178 ("drm/omap: Remove panel-dpi driver")
---
 arch/arm/configs/omap2plus_defconfig |    1 +
 1 file changed, 1 insertion(+)
arch/arm/configs/omap2plus_defconfig +1 -0
@@ -363,6 +363,7 @@ CONFIG_DRM_OMAP_PANEL_TPO_TD028TTEC1=m
 CONFIG_DRM_OMAP_PANEL_TPO_TD043MTEA1=m
 CONFIG_DRM_OMAP_PANEL_NEC_NL8048HL11=m
 CONFIG_DRM_TILCDC=m
+CONFIG_DRM_PANEL_SIMPLE=m
 CONFIG_FB=y
 CONFIG_FIRMWARE_EDID=y
 CONFIG_FB_MODE_HELPERS=y
author maintainer
#218 lore raw

PATCH 215/313 iwlwifi: fw: dont send GEO_TX_POWER_LIMIT command to FW version 36

In reply to Greg Kroah-Hartman

From: Luca Coelho <[email protected]>

commit fddbfeece9c7882cc47754c7da460fe427e3e85b upstream.

The intention was to have the GEO_TX_POWER_LIMIT command in FW version 36 as well, but not all 8000 family got this feature enabled. The 8000 family is the only one using version 36, so skip this version entirely. If we try to send this command to the firmwares that do not support it, we get a BAD_COMMAND response from the firmware.

This fixes https://bugzilla.kernel.org/show_bug.cgi?id=204151.

---
 drivers/net/wireless/intel/iwlwifi/mvm/fw.c |    8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)
drivers/net/wireless/intel/iwlwifi/mvm/fw.c +5 -3
@@ -882,11 +882,13 @@ static bool iwl_mvm_sar_geo_support(stru
 	 * firmware versions.  Unfortunately, we don't have a TLV API
 	 * flag to rely on, so rely on the major version which is in
 	 * the first byte of ucode_ver.  This was implemented
-	 * initially on version 38 and then backported to 36, 29 and
-	 * 17.
+	 * initially on version 38 and then backported to29 and 17.
+	 * The intention was to have it in 36 as well, but not all
+	 * 8000 family got this feature enabled.  The 8000 family is
+	 * the only one using version 36, so skip this version
+	 * entirely.
 	 */
 	return IWL_UCODE_SERIAL(mvm->fw->ucode_ver) >= 38 ||
-	       IWL_UCODE_SERIAL(mvm->fw->ucode_ver) == 36 ||
 	       IWL_UCODE_SERIAL(mvm->fw->ucode_ver) == 29 ||
 	       IWL_UCODE_SERIAL(mvm->fw->ucode_ver) == 17;
 }
author maintainer
#219 lore raw

PATCH 216/313 ALSA: firewire-tascam: handle error code when getting current source of clock

In reply to Greg Kroah-Hartman

From: Takashi Sakamoto <[email protected]>

commit 2617120f4de6d0423384e0e86b14c78b9de84d5a upstream.

The return value of snd_tscm_stream_get_clock() is ignored. This commit checks the value and handle error.

Fixes e453df44f0d6 ("ALSA: firewire-tascam: add PCM functionality")
Cc <[email protected]> # v4.4+
---
 sound/firewire/tascam/tascam-pcm.c |    3 +++
 1 file changed, 3 insertions(+)
sound/firewire/tascam/tascam-pcm.c +3 -0
@@ -56,6 +56,9 @@ static int pcm_open(struct snd_pcm_subst
 		goto err_locked;
 
 	err = snd_tscm_stream_get_clock(tscm, &clock);
+	if (err < 0)
+		goto err_locked;
+
 	if (clock != SND_TSCM_CLOCK_INTERNAL ||
 	    amdtp_stream_pcm_running(&tscm->rx_stream) ||
 	    amdtp_stream_pcm_running(&tscm->tx_stream)) {
author maintainer
#220 lore raw

PATCH 217/313 ALSA: firewire-tascam: check intermediate state of clock status and retry

In reply to Greg Kroah-Hartman

From: Takashi Sakamoto <[email protected]>

commit e1a00b5b253a4f97216b9a33199a863987075162 upstream.

2 bytes in MSB of register for clock status is zero during intermediate state after changing status of sampling clock in models of TASCAM FireWire series. The duration of this state differs depending on cases. During the state, it's better to retry reading the register for current status of the clock.

In current implementation, the intermediate state is checked only when getting current sampling transmission frequency, then retry reading. This care is required for the other operations to read the register.

This commit moves the codes of check and retry into helper function commonly used for operations to read the register.

Fixes e453df44f0d6 ("ALSA: firewire-tascam: add PCM functionality")
Cc <[email protected]> # v4.4+
---
 sound/firewire/tascam/tascam-stream.c |   42 ++++++++++++++++++++++------------
 1 file changed, 28 insertions(+), 14 deletions(-)
sound/firewire/tascam/tascam-stream.c +28 -14
@@ -8,20 +8,37 @@
 #include <linux/delay.h>
 #include "tascam.h"
 
+#define CLOCK_STATUS_MASK      0xffff0000
+#define CLOCK_CONFIG_MASK      0x0000ffff
+
 #define CALLBACK_TIMEOUT 500
 
 static int get_clock(struct snd_tscm *tscm, u32 *data)
 {
+	int trial = 0;
 	__be32 reg;
 	int err;
 
-	err = snd_fw_transaction(tscm->unit, TCODE_READ_QUADLET_REQUEST,
-				 TSCM_ADDR_BASE + TSCM_OFFSET_CLOCK_STATUS,
-				 &reg, sizeof(reg), 0);
-	if (err >= 0)
+	while (trial++ < 5) {
+		err = snd_fw_transaction(tscm->unit, TCODE_READ_QUADLET_REQUEST,
+				TSCM_ADDR_BASE + TSCM_OFFSET_CLOCK_STATUS,
+				&reg, sizeof(reg), 0);
+		if (err < 0)
+			return err;
+
 		*data = be32_to_cpu(reg);
+		if (*data & CLOCK_STATUS_MASK)
+			break;
+
+		// In intermediate state after changing clock status.
+		msleep(50);
+	}
 
-	return err;
+	// Still in the intermediate state.
+	if (trial >= 5)
+		return -EAGAIN;
+
+	return 0;
 }
 
 static int set_clock(struct snd_tscm *tscm, unsigned int rate,
@@ -34,7 +51,7 @@ static int set_clock(struct snd_tscm *ts
 	err = get_clock(tscm, &data);
 	if (err < 0)
 		return err;
-	data &= 0x0000ffff;
+	data &= CLOCK_CONFIG_MASK;
 
 	if (rate > 0) {
 		data &= 0x000000ff;
@@ -79,17 +96,14 @@ static int set_clock(struct snd_tscm *ts
 
 int snd_tscm_stream_get_rate(struct snd_tscm *tscm, unsigned int *rate)
 {
-	u32 data = 0x0;
-	unsigned int trials = 0;
+	u32 data;
 	int err;
 
-	while (data == 0x0 || trials++ < 5) {
-		err = get_clock(tscm, &data);
-		if (err < 0)
-			return err;
+	err = get_clock(tscm, &data);
+	if (err < 0)
+		return err;
 
-		data = (data & 0xff000000) >> 24;
-	}
+	data = (data & 0xff000000) >> 24;
 
 	/* Check base rate. */
 	if ((data & 0x0f) == 0x01)
author maintainer
#221 lore raw

PATCH 218/313 scsi: scsi_dh_rdac: zero cdb in send_mode_select()

In reply to Greg Kroah-Hartman

From: Martin Wilck <[email protected]>

commit 57adf5d4cfd3198aa480e7c94a101fc8c4e6109d upstream.

cdb in send_mode_select() is not zeroed and is only partially filled in rdac_failover_get(), which leads to some random data getting to the device. Users have reported storage responding to such commands with INVALID FIELD IN CDB. Code before commit 327825574132 was not affected, as it called blk_rq_set_block_pc().

Fix this by zeroing out the cdb first.

Identified & fix proposed by HPE.

---
 drivers/scsi/device_handler/scsi_dh_rdac.c |    2 ++
 1 file changed, 2 insertions(+)
drivers/scsi/device_handler/scsi_dh_rdac.c +2 -0
@@ -546,6 +546,8 @@ static void send_mode_select(struct work
 	spin_unlock(&ctlr->ms_lock);
 
  retry:
+	memset(cdb, 0, sizeof(cdb));
+
 	data_size = rdac_failover_get(ctlr, &list, cdb);
 
 	RDAC_LOG(RDAC_LOG_FAILOVER, sdev, "array %s, ctlr %d, "
author maintainer
#222 lore raw

PATCH 219/313 scsi: qla2xxx: Fix Relogin to prevent modifying scan_state flag

In reply to Greg Kroah-Hartman

From: Quinn Tran <[email protected]>

commit 8b5292bcfcacf15182a77a973a98d310e76fd58b upstream.

Relogin fails to move forward due to scan_state flag indicating device is not there. Before relogin process, Session delete process accidently modified the scan_state flag.

[mkp: typos plus corrected Fixes: sha as reported by sfr]

Fixes 2dee5521028c ("scsi: qla2xxx: Fix login state machine freeze")
---
 drivers/scsi/qla2xxx/qla_init.c   |   25 ++++++++++++++++++++-----
 drivers/scsi/qla2xxx/qla_os.c     |    1 +
 drivers/scsi/qla2xxx/qla_target.c |    1 -
 3 files changed, 21 insertions(+), 6 deletions(-)
drivers/scsi/qla2xxx/qla_init.c +20 -5
@@ -289,8 +289,13 @@ qla2x00_async_login(struct scsi_qla_host
 	struct srb_iocb *lio;
 	int rval = QLA_FUNCTION_FAILED;
 
-	if (!vha->flags.online)
-		goto done;
+	if (!vha->flags.online || (fcport->flags & FCF_ASYNC_SENT) ||
+	    fcport->loop_id == FC_NO_LOOP_ID) {
+		ql_log(ql_log_warn, vha, 0xffff,
+		    "%s: %8phC - not sending command.\n",
+		    __func__, fcport->port_name);
+		return rval;
+	}
 
 	sp = qla2x00_get_sp(vha, fcport, GFP_KERNEL);
 	if (!sp)
@@ -1262,8 +1267,13 @@ int qla24xx_async_gpdb(struct scsi_qla_h
 	struct port_database_24xx *pd;
 	struct qla_hw_data *ha = vha->hw;
 
-	if (!vha->flags.online || (fcport->flags & FCF_ASYNC_SENT))
+	if (!vha->flags.online || (fcport->flags & FCF_ASYNC_SENT) ||
+	    fcport->loop_id == FC_NO_LOOP_ID) {
+		ql_log(ql_log_warn, vha, 0xffff,
+		    "%s: %8phC - not sending command.\n",
+		    __func__, fcport->port_name);
 		return rval;
+	}
 
 	fcport->disc_state = DSC_GPDB;
 
@@ -1953,8 +1963,11 @@ qla24xx_handle_plogi_done_event(struct s
 		return;
 	}
 
-	if (fcport->disc_state == DSC_DELETE_PEND)
+	if ((fcport->disc_state == DSC_DELETE_PEND) ||
+	    (fcport->disc_state == DSC_DELETED)) {
+		set_bit(RELOGIN_NEEDED, &vha->dpc_flags);
 		return;
+	}
 
 	if (ea->sp->gen2 != fcport->login_gen) {
 		/* target side must have changed it. */
@@ -6699,8 +6712,10 @@ qla2x00_abort_isp_cleanup(scsi_qla_host_
 	}
 
 	/* Clear all async request states across all VPs. */
-	list_for_each_entry(fcport, &vha->vp_fcports, list)
+	list_for_each_entry(fcport, &vha->vp_fcports, list) {
 		fcport->flags &= ~(FCF_LOGIN_NEEDED | FCF_ASYNC_SENT);
+		fcport->scan_state = 0;
+	}
 	spin_lock_irqsave(&ha->vport_slock, flags);
 	list_for_each_entry(vp, &ha->vp_list, list) {
 		atomic_inc(&vp->vref_count);
drivers/scsi/qla2xxx/qla_os.c +1 -0
@@ -5087,6 +5087,7 @@ void qla24xx_create_new_sess(struct scsi
 	if (fcport) {
 		fcport->id_changed = 1;
 		fcport->scan_state = QLA_FCPORT_FOUND;
+		fcport->chip_reset = vha->hw->base_qpair->chip_reset;
 		memcpy(fcport->node_name, e->u.new_sess.node_name, WWN_SIZE);
 
 		if (pla) {
drivers/scsi/qla2xxx/qla_target.c +0 -1
@@ -1209,7 +1209,6 @@ static void qla24xx_chk_fcp_state(struct
 		sess->logout_on_delete = 0;
 		sess->logo_ack_needed = 0;
 		sess->fw_login_state = DSC_LS_PORT_UNAVAIL;
-		sess->scan_state = 0;
 	}
 }
 
author maintainer
#223 lore raw

PATCH 220/313 printk: Do not lose last line in kmsg buffer dump

In reply to Greg Kroah-Hartman

From: Vincent Whitchurch <[email protected]>

commit c9dccacfccc72c32692eedff4a27a4b0833a2afd upstream.

kmsg_dump_get_buffer() is supposed to select all the youngest log messages which fit into the provided buffer. It determines the correct start index by using msg_print_text() with a NULL buffer to calculate the size of each entry. However, when performing the actual writes, msg_print_text() only writes the entry to the buffer if the written len is lesser than the size of the buffer. So if the lengths of the selected youngest log messages happen to precisely fill up the provided buffer, the last log message is not included.

We don't want to modify msg_print_text() to fill up the buffer and start returning a length which is equal to the size of the buffer, since callers of its other users, such as kmsg_dump_get_line(), depend upon the current behaviour.

Instead, fix kmsg_dump_get_buffer() to compensate for this.

For example, with the following two final prints:

[    6.427502] AAAAAAAAAAAAA
[    6.427769] BBBBBBBB12345

A dump of a 64-byte buffer filled by kmsg_dump_get_buffer(), before this patch:

 00000000: 3c 30 3e 5b 20 20 20 20 36 2e 35 32 32 31 39 37  <0>[    6.522197
 00000010: 5d 20 41 41 41 41 41 41 41 41 41 41 41 41 41 0a  ] AAAAAAAAAAAAA.
 00000020: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
 00000030: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................

After this patch:

 00000000: 3c 30 3e 5b 20 20 20 20 36 2e 34 35 36 36 37 38  <0>[    6.456678
 00000010: 5d 20 42 42 42 42 42 42 42 42 31 32 33 34 35 0a  ] BBBBBBBB12345.
 00000020: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
 00000030: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
Fixes e2ae715d66bf4bec ("kmsg - kmsg_dump() use iterator to receive log buffer content")

To: [email protected]

---
 kernel/printk/printk.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
kernel/printk/printk.c +1 -1
@@ -3274,7 +3274,7 @@ bool kmsg_dump_get_buffer(struct kmsg_du
 	/* move first record forward until length fits into the buffer */
 	seq = dumper->cur_seq;
 	idx = dumper->cur_idx;
-	while (l > size && seq < dumper->next_seq) {
+	while (l >= size && seq < dumper->next_seq) {
 		struct printk_log *msg = log_from_idx(idx);
 
 		l -= msg_print_text(msg, true, time, NULL, 0);
author maintainer
#224 lore raw

PATCH 221/313 IB/mlx5: Free mpi in mp_slave mode

In reply to Greg Kroah-Hartman

From: Danit Goldberg <[email protected]>

commit 5d44adebbb7e785939df3db36ac360f5e8b73e44 upstream.

ib_add_slave_port() allocates a multiport struct but never frees it. Don't leak memory, free the allocated mpi struct during driver unload.

Fixes 32f69e4be269 ("{net, IB}/mlx5: Manage port association for multiport RoCE")
---
 drivers/infiniband/hw/mlx5/main.c |    1 +
 1 file changed, 1 insertion(+)
drivers/infiniband/hw/mlx5/main.c +1 -0
@@ -6829,6 +6829,7 @@ static void mlx5_ib_remove(struct mlx5_c
 			mlx5_ib_unbind_slave_port(mpi->ibdev, mpi);
 		list_del(&mpi->list);
 		mutex_unlock(&mlx5_ib_multiport_mutex);
+		kfree(mpi);
 		return;
 	}
 
author maintainer
#225 lore raw

PATCH 222/313 IB/hfi1: Define variables as unsigned long to fix KASAN warning

In reply to Greg Kroah-Hartman

From: Ira Weiny <[email protected]>

commit f8659d68e2bee5b86a1beaf7be42d942e1fc81f4 upstream.

Define the working variables to be unsigned long to be compatible with for_each_set_bit and change types as needed.

While we are at it remove unused variables from a couple of functions.

This was found because of the following KASAN warning:
==================================================================

   BUG: KASAN: stack-out-of-bounds in find_first_bit+0x19/0x70
   Read of size 8 at addr ffff888362d778d0 by task kworker/u308:2/1889

   CPU: 21 PID: 1889 Comm: kworker/u308:2 Tainted: G W         5.3.0-rc2-mm1+ #2
   Hardware name: Intel Corporation W2600CR/W2600CR, BIOS SE5C600.86B.02.04.0003.102320141138 10/23/2014
   Workqueue: ib-comp-unb-wq ib_cq_poll_work [ib_core]
   Call Trace:
    dump_stack+0x9a/0xf0
    ? find_first_bit+0x19/0x70
    print_address_description+0x6c/0x332
    ? find_first_bit+0x19/0x70
    ? find_first_bit+0x19/0x70
    __kasan_report.cold.6+0x1a/0x3b
    ? find_first_bit+0x19/0x70
    kasan_report+0xe/0x12
    find_first_bit+0x19/0x70
    pma_get_opa_portstatus+0x5cc/0xa80 [hfi1]
    ? ret_from_fork+0x3a/0x50
    ? pma_get_opa_port_ectrs+0x200/0x200 [hfi1]
    ? stack_trace_consume_entry+0x80/0x80
    hfi1_process_mad+0x39b/0x26c0 [hfi1]
    ? __lock_acquire+0x65e/0x21b0
    ? clear_linkup_counters+0xb0/0xb0 [hfi1]
    ? check_chain_key+0x1d7/0x2e0
    ? lock_downgrade+0x3a0/0x3a0
    ? match_held_lock+0x2e/0x250
    ib_mad_recv_done+0x698/0x15e0 [ib_core]
    ? clear_linkup_counters+0xb0/0xb0 [hfi1]
    ? ib_mad_send_done+0xc80/0xc80 [ib_core]
    ? mark_held_locks+0x79/0xa0
    ? _raw_spin_unlock_irqrestore+0x44/0x60
    ? rvt_poll_cq+0x1e1/0x340 [rdmavt]
    __ib_process_cq+0x97/0x100 [ib_core]
    ib_cq_poll_work+0x31/0xb0 [ib_core]
    process_one_work+0x4ee/0xa00
    ? pwq_dec_nr_in_flight+0x110/0x110
    ? do_raw_spin_lock+0x113/0x1d0
    worker_thread+0x57/0x5a0
    ? process_one_work+0xa00/0xa00
    kthread+0x1bb/0x1e0
    ? kthread_create_on_node+0xc0/0xc0
    ret_from_fork+0x3a/0x50

   The buggy address belongs to the page:
   page:ffffea000d8b5dc0 refcount:0 mapcount:0 mapping:0000000000000000 index:0x0
   flags: 0x17ffffc0000000()
   raw: 0017ffffc0000000 0000000000000000 ffffea000d8b5dc8 0000000000000000
   raw: 0000000000000000 0000000000000000 00000000ffffffff 0000000000000000
   page dumped because: kasan: bad access detected

addr ffff888362d778d0 is located in stack of task kworker/u308:2/1889 at offset 32 in frame:

    pma_get_opa_portstatus+0x0/0xa80 [hfi1]

this frame has 1 object:

    [32, 36) 'vl_select_mask'
   Memory state around the buggy address:
    ffff888362d77780: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
    ffff888362d77800: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
   >ffff888362d77880: 00 00 00 00 00 00 f1 f1 f1 f1 04 f2 f2 f2 00 00
                                                    ^
    ffff888362d77900: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
    ffff888362d77980: 00 00 00 00 00 00 00 00 f1 f1 f1 f1 04 f2 f2 f2

==================================================================

---
 drivers/infiniband/hw/hfi1/mad.c |   45 ++++++++++++++++-----------------------
 1 file changed, 19 insertions(+), 26 deletions(-)
drivers/infiniband/hw/hfi1/mad.c +19 -26
@@ -2326,7 +2326,7 @@ struct opa_port_status_req {
 	__be32 vl_select_mask;
 };
 
-#define VL_MASK_ALL		0x000080ff
+#define VL_MASK_ALL		0x00000000000080ffUL
 
 struct opa_port_status_rsp {
 	__u8 port_num;
@@ -2625,15 +2625,14 @@ static int pma_get_opa_classportinfo(str
 }
 
 static void a0_portstatus(struct hfi1_pportdata *ppd,
-			  struct opa_port_status_rsp *rsp, u32 vl_select_mask)
+			  struct opa_port_status_rsp *rsp)
 {
 	if (!is_bx(ppd->dd)) {
 		unsigned long vl;
 		u64 sum_vl_xmit_wait = 0;
-		u32 vl_all_mask = VL_MASK_ALL;
+		unsigned long vl_all_mask = VL_MASK_ALL;
 
-		for_each_set_bit(vl, (unsigned long *)&(vl_all_mask),
-				 8 * sizeof(vl_all_mask)) {
+		for_each_set_bit(vl, &vl_all_mask, BITS_PER_LONG) {
 			u64 tmp = sum_vl_xmit_wait +
 				  read_port_cntr(ppd, C_TX_WAIT_VL,
 						 idx_from_vl(vl));
@@ -2730,12 +2729,12 @@ static int pma_get_opa_portstatus(struct
 		(struct opa_port_status_req *)pmp->data;
 	struct hfi1_devdata *dd = dd_from_ibdev(ibdev);
 	struct opa_port_status_rsp *rsp;
-	u32 vl_select_mask = be32_to_cpu(req->vl_select_mask);
+	unsigned long vl_select_mask = be32_to_cpu(req->vl_select_mask);
 	unsigned long vl;
 	size_t response_data_size;
 	u32 nports = be32_to_cpu(pmp->mad_hdr.attr_mod) >> 24;
 	u8 port_num = req->port_num;
-	u8 num_vls = hweight32(vl_select_mask);
+	u8 num_vls = hweight64(vl_select_mask);
 	struct _vls_pctrs *vlinfo;
 	struct hfi1_ibport *ibp = to_iport(ibdev, port);
 	struct hfi1_pportdata *ppd = ppd_from_ibp(ibp);
@@ -2771,7 +2770,7 @@ static int pma_get_opa_portstatus(struct
 
 	hfi1_read_link_quality(dd, &rsp->link_quality_indicator);
 
-	rsp->vl_select_mask = cpu_to_be32(vl_select_mask);
+	rsp->vl_select_mask = cpu_to_be32((u32)vl_select_mask);
 	rsp->port_xmit_data = cpu_to_be64(read_dev_cntr(dd, C_DC_XMIT_FLITS,
 					  CNTR_INVALID_VL));
 	rsp->port_rcv_data = cpu_to_be64(read_dev_cntr(dd, C_DC_RCV_FLITS,
@@ -2842,8 +2841,7 @@ static int pma_get_opa_portstatus(struct
 	 * So in the for_each_set_bit() loop below, we don't need
 	 * any additional checks for vl.
 	 */
-	for_each_set_bit(vl, (unsigned long *)&(vl_select_mask),
-			 8 * sizeof(vl_select_mask)) {
+	for_each_set_bit(vl, &vl_select_mask, BITS_PER_LONG) {
 		memset(vlinfo, 0, sizeof(*vlinfo));
 
 		tmp = read_dev_cntr(dd, C_DC_RX_FLIT_VL, idx_from_vl(vl));
@@ -2884,7 +2882,7 @@ static int pma_get_opa_portstatus(struct
 		vfi++;
 	}
 
-	a0_portstatus(ppd, rsp, vl_select_mask);
+	a0_portstatus(ppd, rsp);
 
 	if (resp_len)
 		*resp_len += response_data_size;
@@ -2931,16 +2929,14 @@ static u64 get_error_counter_summary(str
 	return error_counter_summary;
 }
 
-static void a0_datacounters(struct hfi1_pportdata *ppd, struct _port_dctrs *rsp,
-			    u32 vl_select_mask)
+static void a0_datacounters(struct hfi1_pportdata *ppd, struct _port_dctrs *rsp)
 {
 	if (!is_bx(ppd->dd)) {
 		unsigned long vl;
 		u64 sum_vl_xmit_wait = 0;
-		u32 vl_all_mask = VL_MASK_ALL;
+		unsigned long vl_all_mask = VL_MASK_ALL;
 
-		for_each_set_bit(vl, (unsigned long *)&(vl_all_mask),
-				 8 * sizeof(vl_all_mask)) {
+		for_each_set_bit(vl, &vl_all_mask, BITS_PER_LONG) {
 			u64 tmp = sum_vl_xmit_wait +
 				  read_port_cntr(ppd, C_TX_WAIT_VL,
 						 idx_from_vl(vl));
@@ -2995,7 +2991,7 @@ static int pma_get_opa_datacounters(stru
 	u64 port_mask;
 	u8 port_num;
 	unsigned long vl;
-	u32 vl_select_mask;
+	unsigned long vl_select_mask;
 	int vfi;
 	u16 link_width;
 	u16 link_speed;
@@ -3073,8 +3069,7 @@ static int pma_get_opa_datacounters(stru
 	 * So in the for_each_set_bit() loop below, we don't need
 	 * any additional checks for vl.
 	 */
-	for_each_set_bit(vl, (unsigned long *)&(vl_select_mask),
-			 8 * sizeof(req->vl_select_mask)) {
+	for_each_set_bit(vl, &vl_select_mask, BITS_PER_LONG) {
 		memset(vlinfo, 0, sizeof(*vlinfo));
 
 		rsp->vls[vfi].port_vl_xmit_data =
@@ -3122,7 +3117,7 @@ static int pma_get_opa_datacounters(stru
 		vfi++;
 	}
 
-	a0_datacounters(ppd, rsp, vl_select_mask);
+	a0_datacounters(ppd, rsp);
 
 	if (resp_len)
 		*resp_len += response_data_size;
@@ -3217,7 +3212,7 @@ static int pma_get_opa_porterrors(struct
 	struct _vls_ectrs *vlinfo;
 	unsigned long vl;
 	u64 port_mask, tmp;
-	u32 vl_select_mask;
+	unsigned long vl_select_mask;
 	int vfi;
 
 	req = (struct opa_port_error_counters64_msg *)pmp->data;
@@ -3276,8 +3271,7 @@ static int pma_get_opa_porterrors(struct
 	vlinfo = &rsp->vls[0];
 	vfi = 0;
 	vl_select_mask = be32_to_cpu(req->vl_select_mask);
-	for_each_set_bit(vl, (unsigned long *)&(vl_select_mask),
-			 8 * sizeof(req->vl_select_mask)) {
+	for_each_set_bit(vl, &vl_select_mask, BITS_PER_LONG) {
 		memset(vlinfo, 0, sizeof(*vlinfo));
 		rsp->vls[vfi].port_vl_xmit_discards =
 			cpu_to_be64(read_port_cntr(ppd, C_SW_XMIT_DSCD_VL,
@@ -3488,7 +3482,7 @@ static int pma_set_opa_portstatus(struct
 	u32 nports = be32_to_cpu(pmp->mad_hdr.attr_mod) >> 24;
 	u64 portn = be64_to_cpu(req->port_select_mask[3]);
 	u32 counter_select = be32_to_cpu(req->counter_select_mask);
-	u32 vl_select_mask = VL_MASK_ALL; /* clear all per-vl cnts */
+	unsigned long vl_select_mask = VL_MASK_ALL; /* clear all per-vl cnts */
 	unsigned long vl;
 
 	if ((nports != 1) || (portn != 1 << port)) {
@@ -3582,8 +3576,7 @@ static int pma_set_opa_portstatus(struct
 	if (counter_select & CS_UNCORRECTABLE_ERRORS)
 		write_dev_cntr(dd, C_DC_UNC_ERR, CNTR_INVALID_VL, 0);
 
-	for_each_set_bit(vl, (unsigned long *)&(vl_select_mask),
-			 8 * sizeof(vl_select_mask)) {
+	for_each_set_bit(vl, &vl_select_mask, BITS_PER_LONG) {
 		if (counter_select & CS_PORT_XMIT_DATA)
 			write_port_cntr(ppd, C_TX_FLIT_VL, idx_from_vl(vl), 0);
 
author maintainer
#226 lore raw

PATCH 223/313 IB/hfi1: Do not update hcrc for a KDETH packet during fault injection

In reply to Greg Kroah-Hartman

From: Kaike Wan <[email protected]>

commit b2590bdd0b1dfb91737e6cb07ebb47bd74957f7e upstream.

When a KDETH packet is subject to fault injection during transmission, HCRC is supposed to be omitted from the packet so that the hardware on the receiver side would drop the packet. When creating pbc, the PbcInsertHcrc field is set to be PBC_IHCRC_NONE if the KDETH packet is subject to fault injection, but overwritten with PBC_IHCRC_LKDETH when update_hcrc() is called later.

This problem is fixed by not calling update_hcrc() when the packet is subject to fault injection.

---
 drivers/infiniband/hw/hfi1/verbs.c |   17 +++++++++--------
 1 file changed, 9 insertions(+), 8 deletions(-)
drivers/infiniband/hw/hfi1/verbs.c +9 -8
@@ -874,16 +874,17 @@ int hfi1_verbs_send_dma(struct rvt_qp *q
 			else
 				pbc |= (ib_is_sc5(sc5) << PBC_DC_INFO_SHIFT);
 
-			if (unlikely(hfi1_dbg_should_fault_tx(qp, ps->opcode)))
-				pbc = hfi1_fault_tx(qp, ps->opcode, pbc);
 			pbc = create_pbc(ppd,
 					 pbc,
 					 qp->srate_mbps,
 					 vl,
 					 plen);
 
-			/* Update HCRC based on packet opcode */
-			pbc = update_hcrc(ps->opcode, pbc);
+			if (unlikely(hfi1_dbg_should_fault_tx(qp, ps->opcode)))
+				pbc = hfi1_fault_tx(qp, ps->opcode, pbc);
+			else
+				/* Update HCRC based on packet opcode */
+				pbc = update_hcrc(ps->opcode, pbc);
 		}
 		tx->wqe = qp->s_wqe;
 		ret = build_verbs_tx_desc(tx->sde, len, tx, ahg_info, pbc);
@@ -1030,12 +1031,12 @@ int hfi1_verbs_send_pio(struct rvt_qp *q
 		else
 			pbc |= (ib_is_sc5(sc5) << PBC_DC_INFO_SHIFT);
 
+		pbc = create_pbc(ppd, pbc, qp->srate_mbps, vl, plen);
 		if (unlikely(hfi1_dbg_should_fault_tx(qp, ps->opcode)))
 			pbc = hfi1_fault_tx(qp, ps->opcode, pbc);
-		pbc = create_pbc(ppd, pbc, qp->srate_mbps, vl, plen);
-
-		/* Update HCRC based on packet opcode */
-		pbc = update_hcrc(ps->opcode, pbc);
+		else
+			/* Update HCRC based on packet opcode */
+			pbc = update_hcrc(ps->opcode, pbc);
 	}
 	if (cb)
 		iowait_pio_inc(&priv->s_iowait);
author maintainer
#227 lore raw

PATCH 224/313 RDMA: Fix double-free in srq creation error flow

In reply to Greg Kroah-Hartman

From: Jack Morgenstein <[email protected]>

commit 3eca7fc2d8d1275d9cf0c709f0937becbfcf6d96 upstream.

The cited commit introduced a double-free of the srq buffer in the error flow of procedure __uverbs_create_xsrq().

The problem is that ib_destroy_srq_user() called in the error flow also frees the srq buffer.

Thus, if uverbs_response() fails in __uverbs_create_srq(), the srq buffer will be freed twice.

---
 drivers/infiniband/core/uverbs_cmd.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/infiniband/core/uverbs_cmd.c +2 -1
@@ -3477,7 +3477,8 @@ static int __uverbs_create_xsrq(struct u
 
 err_copy:
 	ib_destroy_srq_user(srq, uverbs_get_cleared_udata(attrs));
-
+	/* It was released in ib_destroy_srq_user */
+	srq = NULL;
 err_free:
 	kfree(srq);
 err_put:
author maintainer
#228 lore raw

PATCH 225/313 randstruct: Check member structs in is_pure_ops_struct()

In reply to Greg Kroah-Hartman

From: Joonwon Kang <[email protected]>

commit 60f2c82ed20bde57c362e66f796cf9e0e38a6dbb upstream.

While no uses in the kernel triggered this case, it was possible to have a false negative where a struct contains other structs which contain only function pointers because of unreachable code in is_pure_ops_struct().

---
 scripts/gcc-plugins/randomize_layout_plugin.c |   10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)
scripts/gcc-plugins/randomize_layout_plugin.c +5 -5
@@ -443,13 +443,13 @@ static int is_pure_ops_struct(const_tree
 		if (node == fieldtype)
 			continue;
 
-		if (!is_fptr(fieldtype))
-			return 0;
-
-		if (code != RECORD_TYPE && code != UNION_TYPE)
+		if (code == RECORD_TYPE || code == UNION_TYPE) {
+			if (!is_pure_ops_struct(fieldtype))
+				return 0;
 			continue;
+		}
 
-		if (!is_pure_ops_struct(fieldtype))
+		if (!is_fptr(fieldtype))
 			return 0;
 	}
 
author maintainer
#229 lore raw

PATCH 226/313 ARM: dts: am3517-evm: Fix missing video

In reply to Greg Kroah-Hartman

From: Adam Ford <[email protected]>

commit 24cf23276a54dd2825d3e3965c1b1b453e2a113d upstream.

A previous commit removed the panel-dpi driver, which made the video on the AM3517-evm stop working because it relied on the dpi driver for setting video timings. Now that the simple-panel driver is available in omap2plus, this patch migrates the am3517-evm to use a similar panel and remove the manual timing requirements.

Fixes 8bf4b1621178 ("drm/omap: Remove panel-dpi driver")
---
 arch/arm/boot/dts/am3517-evm.dts |   23 ++++-------------------
 1 file changed, 4 insertions(+), 19 deletions(-)
arch/arm/boot/dts/am3517-evm.dts +4 -19
@@ -124,10 +124,11 @@
 	};
 
 	lcd0: display@0 {
-		compatible = "panel-dpi";
+		/* This isn't the exact LCD, but the timings meet spec */
+		/* To make it work, set CONFIG_OMAP2_DSS_MIN_FCK_PER_PCK=4 */
+		compatible = "newhaven,nhd-4.3-480272ef-atxl";
 		label = "15";
-		status = "okay";
-		pinctrl-names = "default";
+		backlight = <&bl>;
 		enable-gpios = <&gpio6 16 GPIO_ACTIVE_HIGH>;	/* gpio176, lcd INI */
 		vcc-supply = <&vdd_io_reg>;
 
@@ -136,22 +137,6 @@
 				remote-endpoint = <&dpi_out>;
 			};
 		};
-
-		panel-timing {
-			clock-frequency = <9000000>;
-			hactive = <480>;
-			vactive = <272>;
-			hfront-porch = <3>;
-			hback-porch = <2>;
-			hsync-len = <42>;
-			vback-porch = <3>;
-			vfront-porch = <4>;
-			vsync-len = <11>;
-			hsync-active = <0>;
-			vsync-active = <0>;
-			de-active = <1>;
-			pixelclk-active = <1>;
-		};
 	};
 
 	bl: backlight {
author maintainer
#230 lore raw

PATCH 227/313 rcu/tree: Fix SCHED_FIFO params

In reply to Greg Kroah-Hartman

From: Peter Zijlstra <[email protected]>

[ Upstream commit 130d9c331bc59a8733b47c58ef197a2b1fa3ed43 ]

A rather embarrasing mistake had us call sched_setscheduler() before initializing the parameters passed to it.

Fixes 1a763fd7c633 ("rcu/tree: Call setschedule() gp ktread to SCHED_FIFO outside of atomic region")
Cc Juri Lelli <[email protected]>
---
 kernel/rcu/tree.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
kernel/rcu/tree.c +3 -3
index 32ea75acba144..affa7aae758f3 100644
@@ -3123,13 +3123,13 @@ static int __init rcu_spawn_gp_kthread(void)
 	t = kthread_create(rcu_gp_kthread, NULL, "%s", rcu_state.name);
 	if (WARN_ONCE(IS_ERR(t), "%s: Could not start grace-period kthread, OOM is now expected behavior\n", __func__))
 		return 0;
-	if (kthread_prio)
+	if (kthread_prio) {
+		sp.sched_priority = kthread_prio;
 		sched_setscheduler_nocheck(t, SCHED_FIFO, &sp);
+	}
 	rnp = rcu_get_root();
 	raw_spin_lock_irqsave_rcu_node(rnp, flags);
 	rcu_state.gp_kthread = t;
-	if (kthread_prio)
-		sp.sched_priority = kthread_prio;
 	raw_spin_unlock_irqrestore_rcu_node(rnp, flags);
 	wake_up_process(t);
 	rcu_spawn_nocb_kthreads();
Signature · 1 line
2.20.1
author maintainer
#231 lore raw

PATCH 228/313 ALSA: hda/realtek - PCI quirk for Medion E4254

In reply to Greg Kroah-Hartman

From: Jan-Marek Glogowski <[email protected]>

[ Upstream commit bd9c10bc663dd2eaac8fe39dad0f18cd21527446 ]

The laptop has a combined jack to attach headsets on the right. The BIOS encodes them as two different colored jacks at the front, but otherwise it seems to be configured ok. But any adaption of the pins config on its own doesn't fix the jack detection to work in Linux. Still Windows works correct.

This is somehow fixed by chaining ALC256_FIXUP_ASUS_HEADSET_MODE, which seems to register the microphone jack as a headset part and also results in fixing jack sensing, visible in dmesg as:

-snd_hda_codec_realtek hdaudioC0D0: Mic=0x19
+snd_hda_codec_realtek hdaudioC0D0: Headset Mic=0x19

[ Actually the essential change is the location of the jack; the
driver created "Front Mic Jack" without the matching volume / mute control element due to its jack location, which confused PA. -- tiwai ]

---
 sound/pci/hda/patch_realtek.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)
sound/pci/hda/patch_realtek.c +13 -0
index d223a79ac934f..36aee8ad20547 100644
@@ -5870,6 +5870,7 @@ enum {
 	ALC256_FIXUP_ASUS_MIC_NO_PRESENCE,
 	ALC299_FIXUP_PREDATOR_SPK,
 	ALC294_FIXUP_ASUS_INTSPK_HEADSET_MIC,
+	ALC256_FIXUP_MEDION_HEADSET_NO_PRESENCE,
 };
 
 static const struct hda_fixup alc269_fixups[] = {
@@ -6926,6 +6927,16 @@ static const struct hda_fixup alc269_fixups[] = {
 		.chained = true,
 		.chain_id = ALC269_FIXUP_HEADSET_MODE_NO_HP_MIC
 	},
+	[ALC256_FIXUP_MEDION_HEADSET_NO_PRESENCE] = {
+		.type = HDA_FIXUP_PINS,
+		.v.pins = (const struct hda_pintbl[]) {
+			{ 0x19, 0x04a11040 },
+			{ 0x21, 0x04211020 },
+			{ }
+		},
+		.chained = true,
+		.chain_id = ALC256_FIXUP_ASUS_HEADSET_MODE
+	},
 };
 
 static const struct snd_pci_quirk alc269_fixup_tbl[] = {
@@ -7189,6 +7200,7 @@ static const struct snd_pci_quirk alc269_fixup_tbl[] = {
 	SND_PCI_QUIRK(0x17aa, 0x9e54, "LENOVO NB", ALC269_FIXUP_LENOVO_EAPD),
 	SND_PCI_QUIRK(0x19e5, 0x3204, "Huawei MACH-WX9", ALC256_FIXUP_HUAWEI_MACH_WX9_PINS),
 	SND_PCI_QUIRK(0x1b7d, 0xa831, "Ordissimo EVE2 ", ALC269VB_FIXUP_ORDISSIMO_EVE2), /* Also known as Malata PC-B1303 */
+	SND_PCI_QUIRK(0x10ec, 0x118c, "Medion EE4254 MD62100", ALC256_FIXUP_MEDION_HEADSET_NO_PRESENCE),
 
 #if 0
 	/* Below is a quirk table taken from the old code.
@@ -7357,6 +7369,7 @@ static const struct hda_model_fixup alc269_fixup_models[] = {
 	{.id = ALC295_FIXUP_CHROME_BOOK, .name = "alc-chrome-book"},
 	{.id = ALC299_FIXUP_PREDATOR_SPK, .name = "predator-spk"},
 	{.id = ALC298_FIXUP_HUAWEI_MBX_STEREO, .name = "huawei-mbx-stereo"},
+	{.id = ALC256_FIXUP_MEDION_HEADSET_NO_PRESENCE, .name = "alc256-medion-headset"},
 	{}
 };
 #define ALC225_STANDARD_PINS \
Signature · 1 line
2.20.1
author maintainer
#232 lore raw

PATCH 229/313 blk-mq: add callback of .cleanup_rq

In reply to Greg Kroah-Hartman

From: Ming Lei <[email protected]>

[ Upstream commit 226b4fc75c78f9c497c5182d939101b260cfb9f3 ]

SCSI maintains its own driver private data hooked off of each SCSI request, and the pridate data won't be freed after scsi_queue_rq() returns BLK_STS_RESOURCE or BLK_STS_DEV_RESOURCE. An upper layer driver (e.g. dm-rq) may need to retry these SCSI requests, before SCSI has fully dispatched them, due to a lower level SCSI driver's resource limitation identified in scsi_queue_rq(). Currently SCSI's per-request private data is leaked when the upper layer driver (dm-rq) frees and then retries these requests in response to BLK_STS_RESOURCE or BLK_STS_DEV_RESOURCE returns from scsi_queue_rq().

This usecase is so specialized that it doesn't warrant training an existing blk-mq interface (e.g. blk_mq_free_request) to allow SCSI to account for freeing its driver private data -- doing so would add an extra branch for handling a special case that all other consumers of SCSI (and blk-mq) won't ever need to worry about.

So the most pragmatic way forward is to delegate freeing SCSI driver private data to the upper layer driver (dm-rq). Do so by adding new .cleanup_rq callback and calling a new blk_mq_cleanup_rq() method from dm-rq. A following commit will implement the .cleanup_rq() hook in scsi_mq_ops.

Cc Ewan D. Milne <[email protected]>
Cc Bart Van Assche <[email protected]>
Cc Hannes Reinecke <[email protected]>
Cc Christoph Hellwig <[email protected]>
Cc Mike Snitzer <[email protected]>
Fixes 396eaf21ee17 ("blk-mq: improve DM's blk-mq IO merging via blk_insert_cloned_request feedback")
---
 drivers/md/dm-rq.c     |  1 +
 include/linux/blk-mq.h | 13 +++++++++++++
 2 files changed, 14 insertions(+)
drivers/md/dm-rq.c +1 -0
index 5f7063f05ae07..b41ecb451c784 100644
@@ -408,6 +408,7 @@ static int map_request(struct dm_rq_target_io *tio)
 		ret = dm_dispatch_clone_request(clone, rq);
 		if (ret == BLK_STS_RESOURCE || ret == BLK_STS_DEV_RESOURCE) {
 			blk_rq_unprep_clone(clone);
+			blk_mq_cleanup_rq(clone);
 			tio->ti->type->release_clone_rq(clone, &tio->info);
 			tio->clone = NULL;
 			return DM_MAPIO_REQUEUE;
include/linux/blk-mq.h +13 -0
index 15d1aa53d96c4..a5a99b43f68e8 100644
@@ -140,6 +140,7 @@ typedef int (poll_fn)(struct blk_mq_hw_ctx *);
 typedef int (map_queues_fn)(struct blk_mq_tag_set *set);
 typedef bool (busy_fn)(struct request_queue *);
 typedef void (complete_fn)(struct request *);
+typedef void (cleanup_rq_fn)(struct request *);
 
 
 struct blk_mq_ops {
@@ -200,6 +201,12 @@ struct blk_mq_ops {
 	/* Called from inside blk_get_request() */
 	void (*initialize_rq_fn)(struct request *rq);
 
+	/*
+	 * Called before freeing one request which isn't completed yet,
+	 * and usually for freeing the driver private data
+	 */
+	cleanup_rq_fn		*cleanup_rq;
+
 	/*
 	 * If set, returns whether or not this queue currently is busy
 	 */
@@ -366,4 +373,10 @@ static inline blk_qc_t request_to_qc_t(struct blk_mq_hw_ctx *hctx,
 			BLK_QC_T_INTERNAL;
 }
 
+static inline void blk_mq_cleanup_rq(struct request *rq)
+{
+	if (rq->q->mq_ops->cleanup_rq)
+		rq->q->mq_ops->cleanup_rq(rq);
+}
+
 #endif
Signature · 1 line
2.20.1
author maintainer
#233 lore raw

PATCH 230/313 scsi: implement .cleanup_rq callback

In reply to Greg Kroah-Hartman

From: Ming Lei <[email protected]>

[ Upstream commit b7e9e1fb7a9227be34ad4a5e778022c3164494cf ]

Implement .cleanup_rq() callback for freeing driver private part of the request. Then we can avoid to leak this part if the request isn't completed by SCSI, and freed by blk-mq or upper layer(such as dm-rq) finally.

Cc Ewan D. Milne <[email protected]>
Cc Bart Van Assche <[email protected]>
Cc Hannes Reinecke <[email protected]>
Cc Christoph Hellwig <[email protected]>
Cc Mike Snitzer <[email protected]>
Fixes 396eaf21ee17 ("blk-mq: improve DM's blk-mq IO merging via blk_insert_cloned_request feedback")
---
 drivers/scsi/scsi_lib.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)
drivers/scsi/scsi_lib.c +13 -0
index 40f392569664b..506062081d8e2 100644
@@ -1073,6 +1073,18 @@ static void scsi_initialize_rq(struct request *rq)
 	cmd->retries = 0;
 }
 
+/*
+ * Only called when the request isn't completed by SCSI, and not freed by
+ * SCSI
+ */
+static void scsi_cleanup_rq(struct request *rq)
+{
+	if (rq->rq_flags & RQF_DONTPREP) {
+		scsi_mq_uninit_cmd(blk_mq_rq_to_pdu(rq));
+		rq->rq_flags &= ~RQF_DONTPREP;
+	}
+}
+
 /* Add a command to the list used by the aacraid and dpt_i2o drivers */
 void scsi_add_cmd_to_list(struct scsi_cmnd *cmd)
 {
@@ -1800,6 +1812,7 @@ static const struct blk_mq_ops scsi_mq_ops = {
 	.init_request	= scsi_mq_init_request,
 	.exit_request	= scsi_mq_exit_request,
 	.initialize_rq_fn = scsi_initialize_rq,
+	.cleanup_rq	= scsi_cleanup_rq,
 	.busy		= scsi_mq_lld_busy,
 	.map_queues	= scsi_map_queues,
 };
Signature · 1 line
2.20.1
author maintainer
#234 lore raw

PATCH 231/313 powerpc/imc: Dont create debugfs files for cpu-less nodes

In reply to Greg Kroah-Hartman

From: Madhavan Srinivasan <[email protected]>

commit 41ba17f20ea835c489e77bd54e2da73184e22060 upstream.

Commit <684d984038aa> ('powerpc/powernv: Add debugfs interface for imc-mode and imc') added debugfs interface for the nest imc pmu devices to support changing of different ucode modes. Primarily adding this capability for debug. But when doing so, the code did not consider the case of cpu-less nodes. So when reading the _cmd_ or _mode_ file of a cpu-less node will create this crash.

Faulting instruction address: 0xc0000000000d0d58

  Oops: Kernel access of bad area, sig: 11 [#1]
  ...
  CPU: 67 PID: 5301 Comm: cat Not tainted 5.2.0-rc6-next-20190627+ #19
  NIP:  c0000000000d0d58 LR: c00000000049aa18 CTR:c0000000000d0d50
  REGS: c00020194548f9e0 TRAP: 0300   Not tainted  (5.2.0-rc6-next-20190627+)
  MSR:  9000000000009033 <SF,HV,EE,ME,IR,DR,RI,LE>  CR:28022822  XER: 00000000
  CFAR: c00000000049aa14 DAR: 000000000003fc08 DSISR:40000000 IRQMASK: 0
  ...
  NIP imc_mem_get+0x8/0x20
  LR  simple_attr_read+0x118/0x170
  Call Trace:
    simple_attr_read+0x70/0x170 (unreliable)
    debugfs_attr_read+0x6c/0xb0
    __vfs_read+0x3c/0x70
     vfs_read+0xbc/0x1a0
    ksys_read+0x7c/0x140
    system_call+0x5c/0x70

Patch fixes the issue with a more robust check for vbase to NULL.

Before patch, ls output for the debugfs imc directory

  # ls /sys/kernel/debug/powerpc/imc/
  imc_cmd_0    imc_cmd_251  imc_cmd_253  imc_cmd_255  imc_mode_0    imc_mode_251  imc_mode_253  imc_mode_255
  imc_cmd_250  imc_cmd_252  imc_cmd_254  imc_cmd_8    imc_mode_250  imc_mode_252  imc_mode_254  imc_mode_8

After patch, ls output for the debugfs imc directory

  # ls /sys/kernel/debug/powerpc/imc/
  imc_cmd_0  imc_cmd_8  imc_mode_0  imc_mode_8

Actual bug here is that, we have two loops with potentially different loop counts. That is, in imc_get_mem_addr_nest(), loop count is obtained from the dt entries. But in case of export_imc_mode_and_cmd(), loop was based on for_each_nid() count. Patch fixes the loop count in latter based on the struct mem_info. Ideally it would be better to have array size in struct imc_pmu.

Fixes 684d984038aa ('powerpc/powernv: Add debugfs interface for imc-mode and imc')
Cc Jan Stancek <[email protected]>
---
 arch/powerpc/platforms/powernv/opal-imc.c |   12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)
arch/powerpc/platforms/powernv/opal-imc.c +6 -6
@@ -53,9 +53,9 @@ static void export_imc_mode_and_cmd(stru
 				    struct imc_pmu *pmu_ptr)
 {
 	static u64 loc, *imc_mode_addr, *imc_cmd_addr;
-	int chip = 0, nid;
 	char mode[16], cmd[16];
 	u32 cb_offset;
+	struct imc_mem_info *ptr = pmu_ptr->mem_info;
 
 	imc_debugfs_parent = debugfs_create_dir("imc", powerpc_debugfs_root);
 
@@ -69,20 +69,20 @@ static void export_imc_mode_and_cmd(stru
 	if (of_property_read_u32(node, "cb_offset", &cb_offset))
 		cb_offset = IMC_CNTL_BLK_OFFSET;
 
-	for_each_node(nid) {
-		loc = (u64)(pmu_ptr->mem_info[chip].vbase) + cb_offset;
+	while (ptr->vbase != NULL) {
+		loc = (u64)(ptr->vbase) + cb_offset;
 		imc_mode_addr = (u64 *)(loc + IMC_CNTL_BLK_MODE_OFFSET);
-		sprintf(mode, "imc_mode_%d", nid);
+		sprintf(mode, "imc_mode_%d", (u32)(ptr->id));
 		if (!imc_debugfs_create_x64(mode, 0600, imc_debugfs_parent,
 					    imc_mode_addr))
 			goto err;
 
 		imc_cmd_addr = (u64 *)(loc + IMC_CNTL_BLK_CMD_OFFSET);
-		sprintf(cmd, "imc_cmd_%d", nid);
+		sprintf(cmd, "imc_cmd_%d", (u32)(ptr->id));
 		if (!imc_debugfs_create_x64(cmd, 0600, imc_debugfs_parent,
 					    imc_cmd_addr))
 			goto err;
-		chip++;
+		ptr++;
 	}
 	return;
 
author maintainer
#235 lore raw

PATCH 232/313 tpm_tis_core: Turn on the TPM before probing IRQs

In reply to Greg Kroah-Hartman

From: Stefan Berger <[email protected]>

commit 5b359c7c43727e624eac3efc7ad21bd2defea161 upstream.

The interrupt probing sequence in tpm_tis_core cannot obviously run with the TPM power gated. Power on the TPM with tpm_chip_start() before probing IRQ's. Turn it off once the probing is complete.

Fixes a3fbfae82b4c ("tpm: take TPM chip power gating out of tpm_transmit()")
---
 drivers/char/tpm/tpm_tis_core.c |    2 ++
 1 file changed, 2 insertions(+)
drivers/char/tpm/tpm_tis_core.c +2 -0
@@ -980,6 +980,7 @@ int tpm_tis_core_init(struct device *dev
 			goto out_err;
 		}
 
+		tpm_chip_start(chip);
 		if (irq) {
 			tpm_tis_probe_irq_single(chip, intmask, IRQF_SHARED,
 						 irq);
@@ -989,6 +990,7 @@ int tpm_tis_core_init(struct device *dev
 		} else {
 			tpm_tis_probe_irq(chip, intmask);
 		}
+		tpm_chip_stop(chip);
 	}
 
 	rc = tpm_chip_register(chip);
author maintainer
#236 lore raw

PATCH 233/313 tpm_tis_core: Set TPM_CHIP_FLAG_IRQ before probing for interrupts

In reply to Greg Kroah-Hartman

From: Stefan Berger <[email protected]>

commit 1ea32c83c699df32689d329b2415796b7bfc2f6e upstream.

The tpm_tis_core has to set the TPM_CHIP_FLAG_IRQ before probing for interrupts since there is no other place in the code that would set it.

Fixes 570a36097f30 ("tpm: drop 'irq' from struct tpm_vendor_specific")
---
 drivers/char/tpm/tpm_tis_core.c |    1 +
 1 file changed, 1 insertion(+)
drivers/char/tpm/tpm_tis_core.c +1 -0
@@ -981,6 +981,7 @@ int tpm_tis_core_init(struct device *dev
 		}
 
 		tpm_chip_start(chip);
+		chip->flags |= TPM_CHIP_FLAG_IRQ;
 		if (irq) {
 			tpm_tis_probe_irq_single(chip, intmask, IRQF_SHARED,
 						 irq);
author maintainer
#237 lore raw

PATCH 234/313 tpm: Wrap the buffer from the caller to tpm_buf in tpm_send()

In reply to Greg Kroah-Hartman

From: Jarkko Sakkinen <[email protected]>

commit e13cd21ffd50a07b55dcc4d8c38cedf27f28eaa1 upstream.

tpm_send() does not give anymore the result back to the caller. This would require another memcpy(), which kind of tells that the whole approach is somewhat broken. Instead, as Mimi suggested, this commit just wraps the data to the tpm_buf, and thus the result will not go to the garbage.

Obviously this assumes from the caller that it passes large enough buffer, which makes the whole API somewhat broken because it could be different size than @buflen but since trusted keys is the only module using this API right now I think that this fix is sufficient for the moment.

In the near future the plan is to replace the parameters with a tpm_buf created by the caller.

Fixes 412eb585587a ("use tpm_buf in tpm_transmit_cmd() as the IO parameter")
---
 drivers/char/tpm/tpm-interface.c |    9 ++-------
 1 file changed, 2 insertions(+), 7 deletions(-)
drivers/char/tpm/tpm-interface.c +2 -7
@@ -354,14 +354,9 @@ int tpm_send(struct tpm_chip *chip, void
 	if (!chip)
 		return -ENODEV;
 
-	rc = tpm_buf_init(&buf, 0, 0);
-	if (rc)
-		goto out;
-
-	memcpy(buf.data, cmd, buflen);
+	buf.data = cmd;
 	rc = tpm_transmit_cmd(chip, &buf, 0, "attempting to a send a command");
-	tpm_buf_destroy(&buf);
-out:
+
 	tpm_put_ops(chip);
 	return rc;
 }
author maintainer
#238 lore raw

PATCH 235/313 fuse: fix deadlock with aio poll and fuse_iqueue::waitq.lock

In reply to Greg Kroah-Hartman

From: Eric Biggers <[email protected]>

commit 76e43c8ccaa35c30d5df853013561145a0f750a5 upstream.

When IOCB_CMD_POLL is used on the FUSE device, aio_poll() disables IRQs and takes kioctx::ctx_lock, then fuse_iqueue::waitq.lock.

This may have to wait for fuse_iqueue::waitq.lock to be released by one of many places that take it with IRQs enabled. Since the IRQ handler may take kioctx::ctx_lock, lockdep reports that a deadlock is possible.

Fix it by protecting the state of struct fuse_iqueue with a separate spinlock, and only accessing fuse_iqueue::waitq using the versions of the waitqueue functions which do IRQ-safe locking internally.

Reproducer:

	#include <fcntl.h>
	#include <stdio.h>
	#include <sys/mount.h>
	#include <sys/stat.h>
	#include <sys/syscall.h>
	#include <unistd.h>
	#include <linux/aio_abi.h>

	int main()
	{
		char opts[128];
		int fd = open("/dev/fuse", O_RDWR);
		aio_context_t ctx = 0;
		struct iocb cb = { .aio_lio_opcode = IOCB_CMD_POLL, .aio_fildes = fd };
		struct iocb *cbp = &cb;

		sprintf(opts, "fd=%d,rootmode=040000,user_id=0,group_id=0", fd);
		mkdir("mnt", 0700);
		mount("foo",  "mnt", "fuse", 0, opts);
		syscall(__NR_io_setup, 1, &ctx);
		syscall(__NR_io_submit, ctx, 1, &cbp);
	}

Beginning of lockdep output:

	=====================================================
	WARNING: SOFTIRQ-safe -> SOFTIRQ-unsafe lock order detected
	5.3.0-rc5 #9 Not tainted
	-----------------------------------------------------
	syz_fuse/135 [HC0[0]:SC0[0]:HE0:SE1] is trying to acquire:
	000000003590ceda (&fiq->waitq){+.+.}, at: spin_lock include/linux/spinlock.h:338 [inline]
	000000003590ceda (&fiq->waitq){+.+.}, at: aio_poll fs/aio.c:1751 [inline]
	000000003590ceda (&fiq->waitq){+.+.}, at: __io_submit_one.constprop.0+0x203/0x5b0 fs/aio.c:1825

	and this task is already holding:
	0000000075037284 (&(&ctx->ctx_lock)->rlock){..-.}, at: spin_lock_irq include/linux/spinlock.h:363 [inline]
	0000000075037284 (&(&ctx->ctx_lock)->rlock){..-.}, at: aio_poll fs/aio.c:1749 [inline]
	0000000075037284 (&(&ctx->ctx_lock)->rlock){..-.}, at: __io_submit_one.constprop.0+0x1f4/0x5b0 fs/aio.c:1825
	which would create a new lock dependency:
	 (&(&ctx->ctx_lock)->rlock){..-.} -> (&fiq->waitq){+.+.}

	but this new dependency connects a SOFTIRQ-irq-safe lock:
	 (&(&ctx->ctx_lock)->rlock){..-.}

	[...]
Reported-by [email protected]
Reported-by [email protected]
Fixes bfe4037e722e ("aio: implement IOCB_CMD_POLL")
Cc <[email protected]> # v4.19+
Cc Christoph Hellwig <[email protected]>
---
 fs/fuse/dev.c    |   93 ++++++++++++++++++++++++++++---------------------------
 fs/fuse/fuse_i.h |    3 +
 fs/fuse/inode.c  |    1 
 3 files changed, 52 insertions(+), 45 deletions(-)
fs/fuse/dev.c +48 -45
@@ -377,7 +377,7 @@ static void queue_request(struct fuse_iq
 	req->in.h.len = sizeof(struct fuse_in_header) +
 		len_args(req->in.numargs, (struct fuse_arg *) req->in.args);
 	list_add_tail(&req->list, &fiq->pending);
-	wake_up_locked(&fiq->waitq);
+	wake_up(&fiq->waitq);
 	kill_fasync(&fiq->fasync, SIGIO, POLL_IN);
 }
 
@@ -389,16 +389,16 @@ void fuse_queue_forget(struct fuse_conn
 	forget->forget_one.nodeid = nodeid;
 	forget->forget_one.nlookup = nlookup;
 
-	spin_lock(&fiq->waitq.lock);
+	spin_lock(&fiq->lock);
 	if (fiq->connected) {
 		fiq->forget_list_tail->next = forget;
 		fiq->forget_list_tail = forget;
-		wake_up_locked(&fiq->waitq);
+		wake_up(&fiq->waitq);
 		kill_fasync(&fiq->fasync, SIGIO, POLL_IN);
 	} else {
 		kfree(forget);
 	}
-	spin_unlock(&fiq->waitq.lock);
+	spin_unlock(&fiq->lock);
 }
 
 static void flush_bg_queue(struct fuse_conn *fc)
@@ -412,10 +412,10 @@ static void flush_bg_queue(struct fuse_c
 		req = list_first_entry(&fc->bg_queue, struct fuse_req, list);
 		list_del(&req->list);
 		fc->active_background++;
-		spin_lock(&fiq->waitq.lock);
+		spin_lock(&fiq->lock);
 		req->in.h.unique = fuse_get_unique(fiq);
 		queue_request(fiq, req);
-		spin_unlock(&fiq->waitq.lock);
+		spin_unlock(&fiq->lock);
 	}
 }
 
@@ -439,9 +439,9 @@ static void request_end(struct fuse_conn
 	 * smp_mb() from queue_interrupt().
 	 */
 	if (!list_empty(&req->intr_entry)) {
-		spin_lock(&fiq->waitq.lock);
+		spin_lock(&fiq->lock);
 		list_del_init(&req->intr_entry);
-		spin_unlock(&fiq->waitq.lock);
+		spin_unlock(&fiq->lock);
 	}
 	WARN_ON(test_bit(FR_PENDING, &req->flags));
 	WARN_ON(test_bit(FR_SENT, &req->flags));
@@ -483,10 +483,10 @@ put_request:
 
 static int queue_interrupt(struct fuse_iqueue *fiq, struct fuse_req *req)
 {
-	spin_lock(&fiq->waitq.lock);
+	spin_lock(&fiq->lock);
 	/* Check for we've sent request to interrupt this req */
 	if (unlikely(!test_bit(FR_INTERRUPTED, &req->flags))) {
-		spin_unlock(&fiq->waitq.lock);
+		spin_unlock(&fiq->lock);
 		return -EINVAL;
 	}
 
@@ -499,13 +499,13 @@ static int queue_interrupt(struct fuse_i
 		smp_mb();
 		if (test_bit(FR_FINISHED, &req->flags)) {
 			list_del_init(&req->intr_entry);
-			spin_unlock(&fiq->waitq.lock);
+			spin_unlock(&fiq->lock);
 			return 0;
 		}
-		wake_up_locked(&fiq->waitq);
+		wake_up(&fiq->waitq);
 		kill_fasync(&fiq->fasync, SIGIO, POLL_IN);
 	}
-	spin_unlock(&fiq->waitq.lock);
+	spin_unlock(&fiq->lock);
 	return 0;
 }
 
@@ -535,16 +535,16 @@ static void request_wait_answer(struct f
 		if (!err)
 			return;
 
-		spin_lock(&fiq->waitq.lock);
+		spin_lock(&fiq->lock);
 		/* Request is not yet in userspace, bail out */
 		if (test_bit(FR_PENDING, &req->flags)) {
 			list_del(&req->list);
-			spin_unlock(&fiq->waitq.lock);
+			spin_unlock(&fiq->lock);
 			__fuse_put_request(req);
 			req->out.h.error = -EINTR;
 			return;
 		}
-		spin_unlock(&fiq->waitq.lock);
+		spin_unlock(&fiq->lock);
 	}
 
 	/*
@@ -559,9 +559,9 @@ static void __fuse_request_send(struct f
 	struct fuse_iqueue *fiq = &fc->iq;
 
 	BUG_ON(test_bit(FR_BACKGROUND, &req->flags));
-	spin_lock(&fiq->waitq.lock);
+	spin_lock(&fiq->lock);
 	if (!fiq->connected) {
-		spin_unlock(&fiq->waitq.lock);
+		spin_unlock(&fiq->lock);
 		req->out.h.error = -ENOTCONN;
 	} else {
 		req->in.h.unique = fuse_get_unique(fiq);
@@ -569,7 +569,7 @@ static void __fuse_request_send(struct f
 		/* acquire extra reference, since request is still needed
 		   after request_end() */
 		__fuse_get_request(req);
-		spin_unlock(&fiq->waitq.lock);
+		spin_unlock(&fiq->lock);
 
 		request_wait_answer(fc, req);
 		/* Pairs with smp_wmb() in request_end() */
@@ -700,12 +700,12 @@ static int fuse_request_send_notify_repl
 
 	__clear_bit(FR_ISREPLY, &req->flags);
 	req->in.h.unique = unique;
-	spin_lock(&fiq->waitq.lock);
+	spin_lock(&fiq->lock);
 	if (fiq->connected) {
 		queue_request(fiq, req);
 		err = 0;
 	}
-	spin_unlock(&fiq->waitq.lock);
+	spin_unlock(&fiq->lock);
 
 	return err;
 }
@@ -1149,12 +1149,12 @@ static int request_pending(struct fuse_i
  * Unlike other requests this is assembled on demand, without a need
  * to allocate a separate fuse_req structure.
  *
- * Called with fiq->waitq.lock held, releases it
+ * Called with fiq->lock held, releases it
  */
 static int fuse_read_interrupt(struct fuse_iqueue *fiq,
 			       struct fuse_copy_state *cs,
 			       size_t nbytes, struct fuse_req *req)
-__releases(fiq->waitq.lock)
+__releases(fiq->lock)
 {
 	struct fuse_in_header ih;
 	struct fuse_interrupt_in arg;
@@ -1169,7 +1169,7 @@ __releases(fiq->waitq.lock)
 	ih.unique = (req->in.h.unique | FUSE_INT_REQ_BIT);
 	arg.unique = req->in.h.unique;
 
-	spin_unlock(&fiq->waitq.lock);
+	spin_unlock(&fiq->lock);
 	if (nbytes < reqsize)
 		return -EINVAL;
 
@@ -1206,7 +1206,7 @@ static struct fuse_forget_link *dequeue_
 static int fuse_read_single_forget(struct fuse_iqueue *fiq,
 				   struct fuse_copy_state *cs,
 				   size_t nbytes)
-__releases(fiq->waitq.lock)
+__releases(fiq->lock)
 {
 	int err;
 	struct fuse_forget_link *forget = dequeue_forget(fiq, 1, NULL);
@@ -1220,7 +1220,7 @@ __releases(fiq->waitq.lock)
 		.len = sizeof(ih) + sizeof(arg),
 	};
 
-	spin_unlock(&fiq->waitq.lock);
+	spin_unlock(&fiq->lock);
 	kfree(forget);
 	if (nbytes < ih.len)
 		return -EINVAL;
@@ -1238,7 +1238,7 @@ __releases(fiq->waitq.lock)
 
 static int fuse_read_batch_forget(struct fuse_iqueue *fiq,
 				   struct fuse_copy_state *cs, size_t nbytes)
-__releases(fiq->waitq.lock)
+__releases(fiq->lock)
 {
 	int err;
 	unsigned max_forgets;
@@ -1252,13 +1252,13 @@ __releases(fiq->waitq.lock)
 	};
 
 	if (nbytes < ih.len) {
-		spin_unlock(&fiq->waitq.lock);
+		spin_unlock(&fiq->lock);
 		return -EINVAL;
 	}
 
 	max_forgets = (nbytes - ih.len) / sizeof(struct fuse_forget_one);
 	head = dequeue_forget(fiq, max_forgets, &count);
-	spin_unlock(&fiq->waitq.lock);
+	spin_unlock(&fiq->lock);
 
 	arg.count = count;
 	ih.len += count * sizeof(struct fuse_forget_one);
@@ -1288,7 +1288,7 @@ __releases(fiq->waitq.lock)
 static int fuse_read_forget(struct fuse_conn *fc, struct fuse_iqueue *fiq,
 			    struct fuse_copy_state *cs,
 			    size_t nbytes)
-__releases(fiq->waitq.lock)
+__releases(fiq->lock)
 {
 	if (fc->minor < 16 || fiq->forget_list_head.next->next == NULL)
 		return fuse_read_single_forget(fiq, cs, nbytes);
@@ -1318,16 +1318,19 @@ static ssize_t fuse_dev_do_read(struct f
 	unsigned int hash;
 
  restart:
-	spin_lock(&fiq->waitq.lock);
-	err = -EAGAIN;
-	if ((file->f_flags & O_NONBLOCK) && fiq->connected &&
-	    !request_pending(fiq))
-		goto err_unlock;
+	for (;;) {
+		spin_lock(&fiq->lock);
+		if (!fiq->connected || request_pending(fiq))
+			break;
+		spin_unlock(&fiq->lock);
 
-	err = wait_event_interruptible_exclusive_locked(fiq->waitq,
+		if (file->f_flags & O_NONBLOCK)
+			return -EAGAIN;
+		err = wait_event_interruptible_exclusive(fiq->waitq,
 				!fiq->connected || request_pending(fiq));
-	if (err)
-		goto err_unlock;
+		if (err)
+			return err;
+	}
 
 	if (!fiq->connected) {
 		err = fc->aborted ? -ECONNABORTED : -ENODEV;
@@ -1351,7 +1354,7 @@ static ssize_t fuse_dev_do_read(struct f
 	req = list_entry(fiq->pending.next, struct fuse_req, list);
 	clear_bit(FR_PENDING, &req->flags);
 	list_del_init(&req->list);
-	spin_unlock(&fiq->waitq.lock);
+	spin_unlock(&fiq->lock);
 
 	in = &req->in;
 	reqsize = in->h.len;
@@ -1409,7 +1412,7 @@ out_end:
 	return err;
 
  err_unlock:
-	spin_unlock(&fiq->waitq.lock);
+	spin_unlock(&fiq->lock);
 	return err;
 }
 
@@ -2121,12 +2124,12 @@ static __poll_t fuse_dev_poll(struct fil
 	fiq = &fud->fc->iq;
 	poll_wait(file, &fiq->waitq, wait);
 
-	spin_lock(&fiq->waitq.lock);
+	spin_lock(&fiq->lock);
 	if (!fiq->connected)
 		mask = EPOLLERR;
 	else if (request_pending(fiq))
 		mask |= EPOLLIN | EPOLLRDNORM;
-	spin_unlock(&fiq->waitq.lock);
+	spin_unlock(&fiq->lock);
 
 	return mask;
 }
@@ -2221,15 +2224,15 @@ void fuse_abort_conn(struct fuse_conn *f
 		flush_bg_queue(fc);
 		spin_unlock(&fc->bg_lock);
 
-		spin_lock(&fiq->waitq.lock);
+		spin_lock(&fiq->lock);
 		fiq->connected = 0;
 		list_for_each_entry(req, &fiq->pending, list)
 			clear_bit(FR_PENDING, &req->flags);
 		list_splice_tail_init(&fiq->pending, &to_end);
 		while (forget_pending(fiq))
 			kfree(dequeue_forget(fiq, 1, NULL));
-		wake_up_all_locked(&fiq->waitq);
-		spin_unlock(&fiq->waitq.lock);
+		wake_up_all(&fiq->waitq);
+		spin_unlock(&fiq->lock);
 		kill_fasync(&fiq->fasync, SIGIO, POLL_IN);
 		end_polls(fc);
 		wake_up_all(&fc->blocked_waitq);
fs/fuse/fuse_i.h +3 -0
@@ -450,6 +450,9 @@ struct fuse_iqueue {
 	/** Connection established */
 	unsigned connected;
 
+	/** Lock protecting accesses to members of this structure */
+	spinlock_t lock;
+
 	/** Readers of the connection are waiting on this */
 	wait_queue_head_t waitq;
 
fs/fuse/inode.c +1 -0
@@ -582,6 +582,7 @@ static int fuse_show_options(struct seq_
 static void fuse_iqueue_init(struct fuse_iqueue *fiq)
 {
 	memset(fiq, 0, sizeof(struct fuse_iqueue));
+	spin_lock_init(&fiq->lock);
 	init_waitqueue_head(&fiq->waitq);
 	INIT_LIST_HEAD(&fiq->pending);
 	INIT_LIST_HEAD(&fiq->interrupts);
author maintainer
#239 lore raw

PATCH 236/313 fuse: fix missing unlock_page in fuse_writepage()

In reply to Greg Kroah-Hartman

From: Vasily Averin <[email protected]>

commit d5880c7a8620290a6c90ced7a0e8bd0ad9419601 upstream.

unlock_page() was missing in case of an already in-flight write against the same page.

Fixes ff17be086477 ("fuse: writepage: skip already in flight")
Cc <[email protected]> # v3.13
---
 fs/fuse/file.c |    1 +
 1 file changed, 1 insertion(+)
fs/fuse/file.c +1 -0
@@ -1767,6 +1767,7 @@ static int fuse_writepage(struct page *p
 		WARN_ON(wbc->sync_mode == WB_SYNC_ALL);
 
 		redirty_page_for_writepage(wbc, page);
+		unlock_page(page);
 		return 0;
 	}
 
author maintainer
#240 lore raw

PATCH 237/313 fuse: fix beyond-end-of-page access in fuse_parse_cache()

In reply to Greg Kroah-Hartman

From: Tejun Heo <[email protected]>

commit e5854b1cdf6cb48a20e01e3bdad0476a4c60a077 upstream.

With DEBUG_PAGEALLOC on, the following triggers.

  BUG: unable to handle page fault for address: ffff88859367c000
  #PF: supervisor read access in kernel mode
  #PF: error_code(0x0000) - not-present page
  PGD 3001067 P4D 3001067 PUD 406d3a8067 PMD 406d30c067 PTE 800ffffa6c983060
  Oops: 0000 [#1] SMP DEBUG_PAGEALLOC
  CPU: 38 PID: 3110657 Comm: python2.7
  RIP: 0010:fuse_readdir+0x88f/0xe7a [fuse]
  Code: 49 8b 4d 08 49 39 4e 60 0f 84 44 04 00 00 48 8b 43 08 43 8d 1c 3c 4d 01 7e 68 49 89 dc 48 03 5c 24 38 49 89 46 60 8b 44 24 30 <8b> 4b 10 44 29 e0 48 89 ca 48 83 c1 1f 48 83 e1 f8 83 f8 17 49 89
  RSP: 0018:ffffc90035edbde0 EFLAGS: 00010286
  RAX: 0000000000001000 RBX: ffff88859367bff0 RCX: 0000000000000000
  RDX: 0000000000000000 RSI: ffff88859367bfed RDI: 0000000000920907
  RBP: ffffc90035edbe90 R08: 000000000000014b R09: 0000000000000004
  R10: ffff88859367b000 R11: 0000000000000000 R12: 0000000000000ff0
  R13: ffffc90035edbee0 R14: ffff889fb8546180 R15: 0000000000000020
  FS:  00007f80b5f4a740(0000) GS:ffff889fffa00000(0000) knlGS:0000000000000000
  CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
  CR2: ffff88859367c000 CR3: 0000001c170c2001 CR4: 00000000003606e0
  DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
  DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
  Call Trace:
   iterate_dir+0x122/0x180
   __x64_sys_getdents+0xa6/0x140
   do_syscall_64+0x42/0x100
   entry_SYSCALL_64_after_hwframe+0x44/0xa9

It's in fuse_parse_cache(). %rbx (ffff88859367bff0) is fuse_dirent pointer - addr + offset. FUSE_DIRENT_SIZE() is trying to dereference namelen off of it but that derefs into the next page which is disabled by pagealloc debug causing a PF.

This is caused by dirent->namelen being accessed before ensuring that there's enough bytes in the page for the dirent. Fix it by pushing down reclen calculation.

Fixes 5d7bc7e8680c ("fuse: allow using readdir cache")
Cc [email protected] # v4.20+
---
 fs/fuse/readdir.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
fs/fuse/readdir.c +3 -1
@@ -372,11 +372,13 @@ static enum fuse_parse_result fuse_parse
 	for (;;) {
 		struct fuse_dirent *dirent = addr + offset;
 		unsigned int nbytes = size - offset;
-		size_t reclen = FUSE_DIRENT_SIZE(dirent);
+		size_t reclen;
 
 		if (nbytes < FUSE_NAME_OFFSET || !dirent->namelen)
 			break;
 
+		reclen = FUSE_DIRENT_SIZE(dirent); /* derefs ->namelen */
+
 		if (WARN_ON(dirent->namelen > FUSE_NAME_MAX))
 			return FOUND_ERR;
 		if (WARN_ON(reclen > nbytes))
author maintainer
#241 lore raw

PATCH 238/313 parisc: Disable HP HSC-PCI Cards to prevent kernel crash

In reply to Greg Kroah-Hartman

From: Helge Deller <[email protected]>

commit 5fa1659105fac63e0f3c199b476025c2e04111ce upstream.

The HP Dino PCI controller chip can be used in two variants: as on-board controller (e.g. in B160L), or on an Add-On card ("Card-Mode") to bridge PCI components to systems without a PCI bus, e.g. to a HSC/GSC bus. One such Add-On card is the HP HSC-PCI Card which has one or more DEC Tulip PCI NIC chips connected to the on-card Dino PCI controller.

Dino in Card-Mode has a big disadvantage: All PCI memory accesses need to go through the DINO_MEM_DATA register, so Linux drivers will not be able to use the ioremap() function. Without ioremap() many drivers will not work, one example is the tulip driver which then simply crashes the kernel if it tries to access the ports on the HP HSC card.

This patch disables the HP HSC card if it finds one, and as such fixes the kernel crash on a HP D350/2 machine.

---
 drivers/parisc/dino.c |   24 ++++++++++++++++++++++++
 1 file changed, 24 insertions(+)
drivers/parisc/dino.c +24 -0
@@ -156,6 +156,15 @@ static inline struct dino_device *DINO_D
 	return container_of(hba, struct dino_device, hba);
 }
 
+/* Check if PCI device is behind a Card-mode Dino. */
+static int pci_dev_is_behind_card_dino(struct pci_dev *dev)
+{
+	struct dino_device *dino_dev;
+
+	dino_dev = DINO_DEV(parisc_walk_tree(dev->bus->bridge));
+	return is_card_dino(&dino_dev->hba.dev->id);
+}
+
 /*
  * Dino Configuration Space Accessor Functions
  */
@@ -437,6 +446,21 @@ static void quirk_cirrus_cardbus(struct
 }
 DECLARE_PCI_FIXUP_ENABLE(PCI_VENDOR_ID_CIRRUS, PCI_DEVICE_ID_CIRRUS_6832, quirk_cirrus_cardbus );
 
+#ifdef CONFIG_TULIP
+static void pci_fixup_tulip(struct pci_dev *dev)
+{
+	if (!pci_dev_is_behind_card_dino(dev))
+		return;
+	if (!(pci_resource_flags(dev, 1) & IORESOURCE_MEM))
+		return;
+	pr_warn("%s: HP HSC-PCI Cards with card-mode Dino not yet supported.\n",
+		pci_name(dev));
+	/* Disable this card by zeroing the PCI resources */
+	memset(&dev->resource[0], 0, sizeof(dev->resource[0]));
+	memset(&dev->resource[1], 0, sizeof(dev->resource[1]));
+}
+DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_DEC, PCI_ANY_ID, pci_fixup_tulip);
+#endif /* CONFIG_TULIP */
 
 static void __init
 dino_bios_init(void)
author maintainer
#242 lore raw

PATCH 239/313 platform/x86: intel_int0002_vgpio: Fix wakeups not working on Cherry Trail

In reply to Greg Kroah-Hartman

From: Hans de Goede <[email protected]>

commit 1bd43d0077b9a32a8b8059036471f3fc82dae342 upstream.

Commit 871f1f2bcb01 ("platform/x86: intel_int0002_vgpio: Only implement irq_set_wake on Bay Trail") removed the irq_set_wake method from the struct irq_chip used on Cherry Trail, but it did not set IRQCHIP_SKIP_SET_WAKE causing kernel/irq/manage.c: set_irq_wake_real() to return -ENXIO.

This causes the kernel to no longer see PME events reported through the INT0002 device as wakeup events. Which e.g. breaks wakeup by the (USB) keyboard on many Cherry Trail 2-in-1 devices.

Fixes 871f1f2bcb01 ("platform/x86: intel_int0002_vgpio: Only implement irq_set_wake on Bay Trail")
---
 drivers/platform/x86/intel_int0002_vgpio.c |    1 +
 1 file changed, 1 insertion(+)
drivers/platform/x86/intel_int0002_vgpio.c +1 -0
@@ -155,6 +155,7 @@ static struct irq_chip int0002_cht_irqch
 	 * No set_wake, on CHT the IRQ is typically shared with the ACPI SCI
 	 * and we don't want to mess with the ACPI SCI irq settings.
 	 */
+	.flags			= IRQCHIP_SKIP_SET_WAKE,
 };
 
 static int int0002_probe(struct platform_device *pdev)
author maintainer
#243 lore raw

PATCH 240/313 KVM: x86: always stop emulation on page fault

In reply to Greg Kroah-Hartman

From: Jan Dakinevich <[email protected]>

commit 8530a79c5a9f4e29e6ffb35ec1a79d81f4968ec8 upstream.

inject_emulated_exception() returns true if and only if nested page fault happens. However, page fault can come from guest page tables walk, either nested or not nested. In both cases we should stop an attempt to read under RIP and give guest to step over its own page fault handler.

This is also visible when an emulated instruction causes a #GP fault and the VMware backdoor is enabled. To handle the VMware backdoor, KVM intercepts #GP faults; with only the next patch applied, x86_emulate_instruction() injects a #GP but returns EMULATE_FAIL
instead of EMULATE_DONE. EMULATE_FAIL causes handle_exception_nmi()
(or gp_interception() for SVM) to re-inject the original #GP because it thinks emulation failed due to a non-VMware opcode. This patch prevents the issue as x86_emulate_instruction() will return EMULATE_DONE after injecting the #GP.

Fixes 6ea6e84309ca ("KVM: x86: inject exceptions produced by x86_decode_insn")
Cc Denis Lunev <[email protected]>
Cc Roman Kagan <[email protected]>
Cc Denis Plotnikov <[email protected]>
---
 arch/x86/kvm/x86.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
arch/x86/kvm/x86.c +3 -1
@@ -6481,8 +6481,10 @@ int x86_emulate_instruction(struct kvm_v
 			if (reexecute_instruction(vcpu, cr2, write_fault_to_spt,
 						emulation_type))
 				return EMULATE_DONE;
-			if (ctxt->have_exception && inject_emulated_exception(vcpu))
+			if (ctxt->have_exception) {
+				inject_emulated_exception(vcpu);
 				return EMULATE_DONE;
+			}
 			if (emulation_type & EMULTYPE_SKIP)
 				return EMULATE_FAIL;
 			return handle_emulation_failure(vcpu, emulation_type);
author maintainer
#244 lore raw

PATCH 241/313 KVM: x86: set ctxt->have_exception in x86_decode_insn()

In reply to Greg Kroah-Hartman

From: Jan Dakinevich <[email protected]>

commit c8848cee74ff05638e913582a476bde879c968ad upstream.

x86_emulate_instruction() takes into account ctxt->have_exception flag during instruction decoding, but in practice this flag is never set in x86_decode_insn().

Fixes 6ea6e84309ca ("KVM: x86: inject exceptions produced by x86_decode_insn")
Cc Denis Lunev <[email protected]>
Cc Roman Kagan <[email protected]>
Cc Denis Plotnikov <[email protected]>
---
 arch/x86/kvm/emulate.c |    2 ++
 arch/x86/kvm/x86.c     |    6 ++++++
 2 files changed, 8 insertions(+)
arch/x86/kvm/emulate.c +2 -0
@@ -5377,6 +5377,8 @@ done_prefixes:
 					ctxt->memopp->addr.mem.ea + ctxt->_eip);
 
 done:
+	if (rc == X86EMUL_PROPAGATE_FAULT)
+		ctxt->have_exception = true;
 	return (rc != X86EMUL_CONTINUE) ? EMULATION_FAILED : EMULATION_OK;
 }
 
arch/x86/kvm/x86.c +6 -0
@@ -6482,6 +6482,12 @@ int x86_emulate_instruction(struct kvm_v
 						emulation_type))
 				return EMULATE_DONE;
 			if (ctxt->have_exception) {
+				/*
+				 * #UD should result in just EMULATION_FAILED, and trap-like
+				 * exception should not be encountered during decode.
+				 */
+				WARN_ON_ONCE(ctxt->exception.vector == UD_VECTOR ||
+					     exception_type(ctxt->exception.vector) == EXCPT_TRAP);
 				inject_emulated_exception(vcpu);
 				return EMULATE_DONE;
 			}
author maintainer
#245 lore raw

PATCH 242/313 KVM: x86: Manually calculate reserved bits when loading PDPTRS

In reply to Greg Kroah-Hartman

From: Sean Christopherson <[email protected]>

commit 16cfacc8085782dab8e365979356ce1ca87fd6cc upstream.

Manually generate the PDPTR reserved bit mask when explicitly loading PDPTRs. The reserved bits that are being tracked by the MMU reflect the current paging mode, which is unlikely to be PAE paging in the vast majority of flows that use load_pdptrs(), e.g. CR0 and CR4 emulation, __set_sregs(), etc... This can cause KVM to incorrectly signal a bad PDPTR, or more likely, miss a reserved bit check and subsequently fail a VM-Enter due to a bad VMCS.GUEST_PDPTR.

Add a one off helper to generate the reserved bits instead of sharing code across the MMU's calculations and the PDPTR emulation. The PDPTR reserved bits are basically set in stone, and pushing a helper into the MMU's calculation adds unnecessary complexity without improving readability.

Oppurtunistically fix/update the comment for load_pdptrs().

Note, the buggy commit also introduced a deliberate functional change, "Also remove bit 5-6 from rsvd_bits_mask per latest SDM.", which was effectively (and correctly) reverted by commit cd9ae5fe47df ("KVM: x86:
Fix page-tables reserved bits"). A bit of SDM archaeology shows that the SDM from late 2008 had a bug (likely a copy+paste error) where it listed bits 6:5 as AVL and A for PDPTEs used for 4k entries but reserved for 2mb entries. I.e. the SDM contradicted itself, and bits 6:5 are and always have been reserved.

Fixes 20c466b56168d ("KVM: Use rsvd_bits_mask in load_pdptrs()")
Cc Nadav Amit <[email protected]>
---
 arch/x86/kvm/x86.c |   11 ++++++++---
 1 file changed, 8 insertions(+), 3 deletions(-)
arch/x86/kvm/x86.c +8 -3
@@ -669,8 +669,14 @@ static int kvm_read_nested_guest_page(st
 				       data, offset, len, access);
 }
 
+static inline u64 pdptr_rsvd_bits(struct kvm_vcpu *vcpu)
+{
+	return rsvd_bits(cpuid_maxphyaddr(vcpu), 63) | rsvd_bits(5, 8) |
+	       rsvd_bits(1, 2);
+}
+
 /*
- * Load the pae pdptrs.  Return true is they are all valid.
+ * Load the pae pdptrs.  Return 1 if they are all valid, 0 otherwise.
  */
 int load_pdptrs(struct kvm_vcpu *vcpu, struct kvm_mmu *mmu, unsigned long cr3)
 {
@@ -689,8 +695,7 @@ int load_pdptrs(struct kvm_vcpu *vcpu, s
 	}
 	for (i = 0; i < ARRAY_SIZE(pdpte); ++i) {
 		if ((pdpte[i] & PT_PRESENT_MASK) &&
-		    (pdpte[i] &
-		     vcpu->arch.mmu->guest_rsvd_check.rsvd_bits_mask[0][2])) {
+		    (pdpte[i] & pdptr_rsvd_bits(vcpu))) {
 			ret = 0;
 			goto out;
 		}
author maintainer
#246 lore raw

PATCH 243/313 KVM: x86: Disable posted interrupts for non-standard IRQs delivery modes

In reply to Greg Kroah-Hartman

From: Alexander Graf <[email protected]>

commit fdcf756213756c23b533ca4974d1f48c6a4d4281 upstream.

We can easily route hardware interrupts directly into VM context when they target the "Fixed" or "LowPriority" delivery modes.

However, on modes such as "SMI" or "Init", we need to go via KVM code to actually put the vCPU into a different mode of operation, so we can not post the interrupt

Add code in the VMX and SVM PI logic to explicitly refuse to establish posted mappings for advanced IRQ deliver modes. This reflects the logic in __apic_accept_irq() which also only ever passes Fixed and LowPriority interrupts as posted interrupts into the guest.

This fixes a bug I have with code which configures real hardware to inject virtual SMIs into my guest.

---
 arch/x86/include/asm/kvm_host.h |    7 +++++++
 arch/x86/kvm/svm.c              |    4 +++-
 arch/x86/kvm/vmx/vmx.c          |    6 +++++-
 3 files changed, 15 insertions(+), 2 deletions(-)
arch/x86/include/asm/kvm_host.h +7 -0
@@ -1576,6 +1576,13 @@ bool kvm_intr_is_single_vcpu(struct kvm
 void kvm_set_msi_irq(struct kvm *kvm, struct kvm_kernel_irq_routing_entry *e,
 		     struct kvm_lapic_irq *irq);
 
+static inline bool kvm_irq_is_postable(struct kvm_lapic_irq *irq)
+{
+	/* We can only post Fixed and LowPrio IRQs */
+	return (irq->delivery_mode == dest_Fixed ||
+		irq->delivery_mode == dest_LowestPrio);
+}
+
 static inline void kvm_arch_vcpu_blocking(struct kvm_vcpu *vcpu)
 {
 	if (kvm_x86_ops->vcpu_blocking)
arch/x86/kvm/svm.c +3 -1
@@ -5252,7 +5252,8 @@ get_pi_vcpu_info(struct kvm *kvm, struct
 
 	kvm_set_msi_irq(kvm, e, &irq);
 
-	if (!kvm_intr_is_single_vcpu(kvm, &irq, &vcpu)) {
+	if (!kvm_intr_is_single_vcpu(kvm, &irq, &vcpu) ||
+	    !kvm_irq_is_postable(&irq)) {
 		pr_debug("SVM: %s: use legacy intr remap mode for irq %u\n",
 			 __func__, irq.vector);
 		return -1;
@@ -5306,6 +5307,7 @@ static int svm_update_pi_irte(struct kvm
 		 * 1. When cannot target interrupt to a specific vcpu.
 		 * 2. Unsetting posted interrupt.
 		 * 3. APIC virtialization is disabled for the vcpu.
+		 * 4. IRQ has incompatible delivery mode (SMI, INIT, etc)
 		 */
 		if (!get_pi_vcpu_info(kvm, e, &vcpu_info, &svm) && set &&
 		    kvm_vcpu_apicv_active(&svm->vcpu)) {
arch/x86/kvm/vmx/vmx.c +5 -1
@@ -7325,10 +7325,14 @@ static int vmx_update_pi_irte(struct kvm
 		 * irqbalance to make the interrupts single-CPU.
 		 *
 		 * We will support full lowest-priority interrupt later.
+		 *
+		 * In addition, we can only inject generic interrupts using
+		 * the PI mechanism, refuse to route others through it.
 		 */
 
 		kvm_set_msi_irq(kvm, e, &irq);
-		if (!kvm_intr_is_single_vcpu(kvm, &irq, &vcpu)) {
+		if (!kvm_intr_is_single_vcpu(kvm, &irq, &vcpu) ||
+		    !kvm_irq_is_postable(&irq)) {
 			/*
 			 * Make sure the IRTE is in remapped mode if
 			 * we don't handle it in posted mode.
author maintainer
#247 lore raw

PATCH 244/313 KVM: x86/mmu: Use fast invalidate mechanism to zap MMIO sptes

In reply to Greg Kroah-Hartman

From: Sean Christopherson <[email protected]>

commit 92f58b5c0181596d9f1e317b49ada2e728fb76eb upstream.

Use the fast invalidate mechasim to zap MMIO sptes on a MMIO generation wrap. The fast invalidate flow was reintroduced to fix a livelock bug in kvm_mmu_zap_all() that can occur if kvm_mmu_zap_all() is invoked when the guest has live vCPUs. I.e. using kvm_mmu_zap_all() to handle the MMIO generation wrap is theoretically susceptible to the livelock bug.

This effectively reverts commit 4771450c345dc ("Revert "KVM: MMU: drop kvm_mmu_zap_mmio_sptes""), i.e. restores the behavior of commit a8eca9dcc656a ("KVM: MMU: drop kvm_mmu_zap_mmio_sptes").

Note, this actually fixes commit 571c5af06e303 ("KVM: x86/mmu:
Voluntarily reschedule as needed when zapping MMIO sptes"), but there is no need to incrementally revert back to using fast invalidate, e.g. doing so doesn't provide any bisection or stability benefits.

Fixes 571c5af06e303 ("KVM: x86/mmu: Voluntarily reschedule as needed when zapping MMIO sptes")
---
 arch/x86/kvm/mmu.c |   17 +++--------------
 1 file changed, 3 insertions(+), 14 deletions(-)
arch/x86/kvm/mmu.c +3 -14
@@ -389,8 +389,6 @@ static void mark_mmio_spte(struct kvm_vc
 	mask |= (gpa & shadow_nonpresent_or_rsvd_mask)
 		<< shadow_nonpresent_or_rsvd_mask_len;
 
-	page_header(__pa(sptep))->mmio_cached = true;
-
 	trace_mark_mmio_spte(sptep, gfn, access, gen);
 	mmu_spte_set(sptep, mask);
 }
@@ -5952,7 +5950,7 @@ void kvm_mmu_slot_set_dirty(struct kvm *
 }
 EXPORT_SYMBOL_GPL(kvm_mmu_slot_set_dirty);
 
-static void __kvm_mmu_zap_all(struct kvm *kvm, bool mmio_only)
+void kvm_mmu_zap_all(struct kvm *kvm)
 {
 	struct kvm_mmu_page *sp, *node;
 	LIST_HEAD(invalid_list);
@@ -5961,14 +5959,10 @@ static void __kvm_mmu_zap_all(struct kvm
 	spin_lock(&kvm->mmu_lock);
 restart:
 	list_for_each_entry_safe(sp, node, &kvm->arch.active_mmu_pages, link) {
-		if (mmio_only && !sp->mmio_cached)
-			continue;
 		if (sp->role.invalid && sp->root_count)
 			continue;
-		if (__kvm_mmu_prepare_zap_page(kvm, sp, &invalid_list, &ign)) {
-			WARN_ON_ONCE(mmio_only);
+		if (__kvm_mmu_prepare_zap_page(kvm, sp, &invalid_list, &ign))
 			goto restart;
-		}
 		if (cond_resched_lock(&kvm->mmu_lock))
 			goto restart;
 	}
@@ -5977,11 +5971,6 @@ restart:
 	spin_unlock(&kvm->mmu_lock);
 }
 
-void kvm_mmu_zap_all(struct kvm *kvm)
-{
-	return __kvm_mmu_zap_all(kvm, false);
-}
-
 void kvm_mmu_invalidate_mmio_sptes(struct kvm *kvm, u64 gen)
 {
 	WARN_ON(gen & KVM_MEMSLOT_GEN_UPDATE_IN_PROGRESS);
@@ -6003,7 +5992,7 @@ void kvm_mmu_invalidate_mmio_sptes(struc
 	 */
 	if (unlikely(gen == 0)) {
 		kvm_debug_ratelimited("kvm: zapping shadow pages for mmio generation wraparound\n");
-		__kvm_mmu_zap_all(kvm, true);
+		kvm_mmu_zap_all_fast(kvm);
 	}
 }
 
author maintainer
#248 lore raw

PATCH 245/313 media: videobuf-core.c: poll_wait needs a non-NULL buf pointer

In reply to Greg Kroah-Hartman

From: Hans Verkuil <[email protected]>

commit 6f51fdfd8229d5358c2d6e272cf73478866e8ddc upstream.

poll_wait uses &buf->done, but buf is NULL. Move the poll_wait to later in the function once buf is correctly set and only call it if it is non-NULL.

Fixes bb436cbeb918 ("media: videobuf: fix epoll() by calling poll_wait first")
Cc <[email protected]> # for v5.1 and up
---
 drivers/media/v4l2-core/videobuf-core.c |    5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)
drivers/media/v4l2-core/videobuf-core.c +3 -2
@@ -1123,7 +1123,6 @@ __poll_t videobuf_poll_stream(struct fil
 	struct videobuf_buffer *buf = NULL;
 	__poll_t rc = 0;
 
-	poll_wait(file, &buf->done, wait);
 	videobuf_queue_lock(q);
 	if (q->streaming) {
 		if (!list_empty(&q->stream))
@@ -1143,7 +1142,9 @@ __poll_t videobuf_poll_stream(struct fil
 		}
 		buf = q->read_buf;
 	}
-	if (!buf)
+	if (buf)
+		poll_wait(file, &buf->done, wait);
+	else
 		rc = EPOLLERR;
 
 	if (0 == rc) {
author maintainer
#249 lore raw

PATCH 246/313 media: sn9c20x: Add MSI MS-1039 laptop to flip_dmi_table

In reply to Greg Kroah-Hartman

From: Hans de Goede <[email protected]>

commit 7e0bb5828311f811309bed5749528ca04992af2f upstream.

Like a bunch of other MSI laptops the MS-1039 uses a 0c45:627b SN9C201 + OV7660 webcam which is mounted upside down.

Add it to the sn9c20x flip_dmi_table to deal with this.

---
 drivers/media/usb/gspca/sn9c20x.c |    7 +++++++
 1 file changed, 7 insertions(+)
drivers/media/usb/gspca/sn9c20x.c +7 -0
@@ -124,6 +124,13 @@ static const struct dmi_system_id flip_d
 		}
 	},
 	{
+		.ident = "MSI MS-1039",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "MICRO-STAR INT'L CO.,LTD."),
+			DMI_MATCH(DMI_PRODUCT_NAME, "MS-1039"),
+		}
+	},
+	{
 		.ident = "MSI MS-1632",
 		.matches = {
 			DMI_MATCH(DMI_BOARD_VENDOR, "MSI"),
author maintainer
#250 lore raw

PATCH 247/313 media: dont drop front-end reference count for ->detach

In reply to Greg Kroah-Hartman

From: Arnd Bergmann <[email protected]>

commit 14e3cdbb00a885eedc95c0cf8eda8fe28d26d6b4 upstream.

A bugfix introduce a link failure in configurations without CONFIG_MODULES:

In file included from drivers/media/usb/dvb-usb/pctv452e.c:20:0:
drivers/media/usb/dvb-usb/pctv452e.c: In function 'pctv452e_frontend_attach':
drivers/media/dvb-frontends/stb0899_drv.h:151:36: error: weak declaration of 'stb0899_attach' being applied to a already existing, static definition

The problem is that the !IS_REACHABLE() declaration of stb0899_attach() is a 'static inline' definition that clashes with the weak definition.

I further observed that the bugfix was only done for one of the five users of stb0899_attach(), the other four still have the problem. This reverts the bugfix and instead addresses the problem by not dropping the reference count when calling '->detach()', instead we call this function directly in dvb_frontend_put() before dropping the kref on the front-end.

I first submitted this in early 2018, and after some discussion it was apparently discarded. While there is a long-term plan in place, that plan is obviously not nearing completion yet, and the current kernel is still broken unless this patch is applied.

Cc Max Kellermann <[email protected]>
Cc Wolfgang Rohdewald <[email protected]>
Fixes f686c14364ad ("[media] stb0899: move code to "detach" callback")
Fixes 6cdeaed3b142 ("media: dvb_usb_pctv452e: module refcount changes were unbalanced")
---
 drivers/media/dvb-core/dvb_frontend.c |    4 +++-
 drivers/media/usb/dvb-usb/pctv452e.c  |    8 --------
 2 files changed, 3 insertions(+), 9 deletions(-)
drivers/media/dvb-core/dvb_frontend.c +3 -1
@@ -152,6 +152,9 @@ static void dvb_frontend_free(struct kre
 
 static void dvb_frontend_put(struct dvb_frontend *fe)
 {
+	/* call detach before dropping the reference count */
+	if (fe->ops.detach)
+		fe->ops.detach(fe);
 	/*
 	 * Check if the frontend was registered, as otherwise
 	 * kref was not initialized yet.
@@ -3026,7 +3029,6 @@ void dvb_frontend_detach(struct dvb_fron
 	dvb_frontend_invoke_release(fe, fe->ops.release_sec);
 	dvb_frontend_invoke_release(fe, fe->ops.tuner_ops.release);
 	dvb_frontend_invoke_release(fe, fe->ops.analog_ops.release);
-	dvb_frontend_invoke_release(fe, fe->ops.detach);
 	dvb_frontend_put(fe);
 }
 EXPORT_SYMBOL(dvb_frontend_detach);
drivers/media/usb/dvb-usb/pctv452e.c +0 -8
@@ -909,14 +909,6 @@ static int pctv452e_frontend_attach(stru
 						&a->dev->i2c_adap);
 	if (!a->fe_adap[0].fe)
 		return -ENODEV;
-
-	/*
-	 * dvb_frontend will call dvb_detach for both stb0899_detach
-	 * and stb0899_release but we only do dvb_attach(stb0899_attach).
-	 * Increment the module refcount instead.
-	 */
-	symbol_get(stb0899_attach);
-
 	if ((dvb_attach(lnbp22_attach, a->fe_adap[0].fe,
 					&a->dev->i2c_adap)) == NULL)
 		err("Cannot attach lnbp22\n");
author maintainer
#251 lore raw

PATCH 248/313 spi: ep93xx: Repair SPI CS lookup tables

In reply to Greg Kroah-Hartman

From: Alexander Sverdlin <[email protected]>

commit 4fbc485324d2975c54201091dfad0a7dd4902324 upstream.

The actual device name of the SPI controller being registered on EP93xx is "spi0" (as seen by gpiod_find_lookup_table()). This patch fixes all relevant lookup tables and the following failure (seen on EDB9302):

ep93xx-spi ep93xx-spi.0: failed to register SPI master
ep93xx-spi: probe of ep93xx-spi.0 failed with error -22

---
 arch/arm/mach-ep93xx/edb93xx.c       |    2 +-
 arch/arm/mach-ep93xx/simone.c        |    2 +-
 arch/arm/mach-ep93xx/ts72xx.c        |    4 ++--
 arch/arm/mach-ep93xx/vision_ep9307.c |    2 +-
 4 files changed, 5 insertions(+), 5 deletions(-)
arch/arm/mach-ep93xx/edb93xx.c +1 -1
@@ -103,7 +103,7 @@ static struct spi_board_info edb93xx_spi
 };
 
 static struct gpiod_lookup_table edb93xx_spi_cs_gpio_table = {
-	.dev_id = "ep93xx-spi.0",
+	.dev_id = "spi0",
 	.table = {
 		GPIO_LOOKUP("A", 6, "cs", GPIO_ACTIVE_LOW),
 		{ },
arch/arm/mach-ep93xx/simone.c +1 -1
@@ -73,7 +73,7 @@ static struct spi_board_info simone_spi_
  * v1.3 parts will still work, since the signal on SFRMOUT is automatic.
  */
 static struct gpiod_lookup_table simone_spi_cs_gpio_table = {
-	.dev_id = "ep93xx-spi.0",
+	.dev_id = "spi0",
 	.table = {
 		GPIO_LOOKUP("A", 1, "cs", GPIO_ACTIVE_LOW),
 		{ },
arch/arm/mach-ep93xx/ts72xx.c +2 -2
@@ -267,7 +267,7 @@ static struct spi_board_info bk3_spi_boa
  * goes through CPLD
  */
 static struct gpiod_lookup_table bk3_spi_cs_gpio_table = {
-	.dev_id = "ep93xx-spi.0",
+	.dev_id = "spi0",
 	.table = {
 		GPIO_LOOKUP("F", 3, "cs", GPIO_ACTIVE_LOW),
 		{ },
@@ -316,7 +316,7 @@ static struct spi_board_info ts72xx_spi_
 };
 
 static struct gpiod_lookup_table ts72xx_spi_cs_gpio_table = {
-	.dev_id = "ep93xx-spi.0",
+	.dev_id = "spi0",
 	.table = {
 		/* DIO_17 */
 		GPIO_LOOKUP("F", 2, "cs", GPIO_ACTIVE_LOW),
arch/arm/mach-ep93xx/vision_ep9307.c +1 -1
@@ -242,7 +242,7 @@ static struct spi_board_info vision_spi_
 };
 
 static struct gpiod_lookup_table vision_spi_cs_gpio_table = {
-	.dev_id = "ep93xx-spi.0",
+	.dev_id = "spi0",
 	.table = {
 		GPIO_LOOKUP_IDX("A", 6, "cs", 0, GPIO_ACTIVE_LOW),
 		GPIO_LOOKUP_IDX("A", 7, "cs", 1, GPIO_ACTIVE_LOW),
author maintainer
#252 lore raw

PATCH 249/313 spi: spi-fsl-dspi: Exit the ISR with IRQ_NONE when its not ours

In reply to Greg Kroah-Hartman

From: Vladimir Oltean <[email protected]>

commit d41f36a6464a85c06ad920703d878e4491d2c023 upstream.

The DSPI interrupt can be shared between two controllers at least on the LX2160A. In that case, the driver for one controller might misbehave and consume the other's interrupt. Fix this by actually checking if any of the bits in the status register have been asserted.

Fixes 13aed2392741 ("spi: spi-fsl-dspi: use IRQF_SHARED mode to request IRQ")
---
 drivers/spi/spi-fsl-dspi.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/spi/spi-fsl-dspi.c +3 -1
@@ -886,9 +886,11 @@ static irqreturn_t dspi_interrupt(int ir
 					trans_mode);
 			}
 		}
+
+		return IRQ_HANDLED;
 	}
 
-	return IRQ_HANDLED;
+	return IRQ_NONE;
 }
 
 static const struct of_device_id fsl_dspi_dt_ids[] = {
author maintainer
#253 lore raw

PATCH 250/313 binfmt_elf: Do not move brk for INTERP-less ET_EXEC

In reply to Greg Kroah-Hartman

From: Kees Cook <[email protected]>

commit 7be3cb019db1cbd5fd5ffe6d64a23fefa4b6f229 upstream.

When brk was moved for binaries without an interpreter, it should have been limited to ET_DYN only. In other words, the special case was an ET_DYN that lacks an INTERP, not just an executable that lacks INTERP. The bug manifested for giant static executables, where the brk would end up in the middle of the text area on 32-bit architectures.

Fixes bbdc6076d2e5 ("binfmt_elf: move brk out of mmap when doing direct loader exec")
---
 fs/binfmt_elf.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
fs/binfmt_elf.c +2 -1
@@ -1142,7 +1142,8 @@ out_free_interp:
 		 * (since it grows up, and may collide early with the stack
 		 * growing down), and into the unused ELF_ET_DYN_BASE region.
 		 */
-		if (IS_ENABLED(CONFIG_ARCH_HAS_ELF_RANDOMIZE) && !interpreter)
+		if (IS_ENABLED(CONFIG_ARCH_HAS_ELF_RANDOMIZE) &&
+		    loc->elf_ex.e_type == ET_DYN && !interpreter)
 			current->mm->brk = current->mm->start_brk =
 				ELF_ET_DYN_BASE;
 
author maintainer
#254 lore raw

PATCH 251/313 ASoC: Intel: NHLT: Fix debug print format

In reply to Greg Kroah-Hartman

From: Amadeusz Sławiński <[email protected]>

commit 855a06da37a773fd073d51023ac9d07988c87da8 upstream.

oem_table_id is 8 chars long, so we need to limit it, otherwise it may print some unprintable characters into dmesg.

---
 sound/soc/intel/skylake/skl-nhlt.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
sound/soc/intel/skylake/skl-nhlt.c +1 -1
@@ -225,7 +225,7 @@ int skl_nhlt_update_topology_bin(struct
 	struct hdac_bus *bus = skl_to_bus(skl);
 	struct device *dev = bus->dev;
 
-	dev_dbg(dev, "oem_id %.6s, oem_table_id %8s oem_revision %d\n",
+	dev_dbg(dev, "oem_id %.6s, oem_table_id %.8s oem_revision %d\n",
 		nhlt->header.oem_id, nhlt->header.oem_table_id,
 		nhlt->header.oem_revision);
 
author maintainer
#255 lore raw

PATCH 252/313 ASoC: Intel: Skylake: Use correct function to access iomem space

In reply to Greg Kroah-Hartman

From: Amadeusz Sławiński <[email protected]>

commit 17d29ff98fd4b70e9ccdac5e95e18a087e2737ef upstream.

For copying from __iomem, we should use __ioread32_copy.

reported by sparse:
sound/soc/intel/skylake/skl-debug.c:437:34: warning: incorrect type in argument 1 (different address spaces)
sound/soc/intel/skylake/skl-debug.c:437:34: expected void [noderef] <asn:2> *to
sound/soc/intel/skylake/skl-debug.c:437:34: got unsigned char *
sound/soc/intel/skylake/skl-debug.c:437:51: warning: incorrect type in argument 2 (different address spaces)
sound/soc/intel/skylake/skl-debug.c:437:51: expected void const *from
sound/soc/intel/skylake/skl-debug.c:437:51: got void [noderef] <asn:2> *[assigned] fw_reg_addr

---
 sound/soc/intel/skylake/skl-debug.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
sound/soc/intel/skylake/skl-debug.c +1 -1
@@ -188,7 +188,7 @@ static ssize_t fw_softreg_read(struct fi
 	memset(d->fw_read_buff, 0, FW_REG_BUF);
 
 	if (w0_stat_sz > 0)
-		__iowrite32_copy(d->fw_read_buff, fw_reg_addr, w0_stat_sz >> 2);
+		__ioread32_copy(d->fw_read_buff, fw_reg_addr, w0_stat_sz >> 2);
 
 	for (offset = 0; offset < FW_REG_SIZE; offset += 16) {
 		ret += snprintf(tmp + ret, FW_REG_BUF - ret, "%#.4x: ", offset);
author maintainer
#256 lore raw

PATCH 253/313 ASoC: Intel: Fix use of potentially uninitialized variable

In reply to Greg Kroah-Hartman

From: Amadeusz Sławiński <[email protected]>

commit 810f3b860850148788fc1ed8a6f5f807199fed65 upstream.

If ipc->ops.reply_msg_match is NULL, we may end up using uninitialized mask value.

reported by smatch:
sound/soc/intel/common/sst-ipc.c:266 sst_ipc_reply_find_msg() error: uninitialized symbol 'mask'.

---
 sound/soc/intel/common/sst-ipc.c |    2 ++
 1 file changed, 2 insertions(+)
sound/soc/intel/common/sst-ipc.c +2 -0
@@ -222,6 +222,8 @@ struct ipc_message *sst_ipc_reply_find_m
 
 	if (ipc->ops.reply_msg_match != NULL)
 		header = ipc->ops.reply_msg_match(header, &mask);
+	else
+		mask = (u64)-1;
 
 	if (list_empty(&ipc->rx_list)) {
 		dev_err(ipc->dev, "error: rx list empty but received 0x%llx\n",
author maintainer
#257 lore raw

PATCH 254/313 ARM: samsung: Fix system restart on S3C6410

In reply to Greg Kroah-Hartman

From: Lihua Yao <[email protected]>

commit 16986074035cc0205472882a00d404ed9d213313 upstream.

S3C6410 system restart is triggered by watchdog reset.

Fixes 9f55342cc2de ("ARM: dts: s3c64xx: Fix infinite interrupt in soft mode")
---
 arch/arm/plat-samsung/watchdog-reset.c |    1 +
 1 file changed, 1 insertion(+)
arch/arm/plat-samsung/watchdog-reset.c +1 -0
@@ -62,6 +62,7 @@ void samsung_wdt_reset(void)
 #ifdef CONFIG_OF
 static const struct of_device_id s3c2410_wdt_match[] = {
 	{ .compatible = "samsung,s3c2410-wdt" },
+	{ .compatible = "samsung,s3c6410-wdt" },
 	{},
 };
 
author maintainer
#258 lore raw

PATCH 255/313 ARM: zynq: Use memcpy_toio instead of memcpy on smp bring-up

In reply to Greg Kroah-Hartman

From: Luis Araneda <[email protected]>

commit b7005d4ef4f3aa2dc24019ffba03a322557ac43d upstream.

This fixes a kernel panic on memcpy when
FORTIFY_SOURCE is enabled.

The initial smp implementation on commit aa7eb2bb4e4a ("arm: zynq: Add smp support")
used memcpy, which worked fine until commit ee333554fed5 ("ARM: 8749/1: Kconfig: Add ARCH_HAS_FORTIFY_SOURCE") enabled overflow checks at runtime, producing a read overflow panic.

The computed size of memcpy args are:
- p_size (dst): 4294967295 = (size_t) -1
- q_size (src): 1
- size (len): 8

Additionally, the memory is marked as __iomem, so one of the memcpy_* functions should be used for read/write.

Fixes aa7eb2bb4e4a ("arm: zynq: Add smp support")
---
 arch/arm/mach-zynq/platsmp.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm/mach-zynq/platsmp.c +1 -1
@@ -57,7 +57,7 @@ int zynq_cpun_start(u32 address, int cpu
 			* 0x4: Jump by mov instruction
 			* 0x8: Jumping address
 			*/
-			memcpy((__force void *)zero, &zynq_secondary_trampoline,
+			memcpy_toio(zero, &zynq_secondary_trampoline,
 							trampoline_size);
 			writel(address, zero + trampoline_size);
 
author maintainer
#259 lore raw

PATCH 256/313 arm64: tlb: Ensure we execute an ISB following walk cache invalidation

In reply to Greg Kroah-Hartman

From: Will Deacon <[email protected]>

commit 51696d346c49c6cf4f29e9b20d6e15832a2e3408 upstream.

05f2d2f83b5a ("arm64: tlbflush: Introduce __flush_tlb_kernel_pgtable") added a new TLB invalidation helper which is used when freeing intermediate levels of page table used for kernel mappings, but is missing the required ISB instruction after completion of the TLBI instruction.

Add the missing barrier.

Fixes 05f2d2f83b5a ("arm64: tlbflush: Introduce __flush_tlb_kernel_pgtable")
---
 arch/arm64/include/asm/tlbflush.h |    1 +
 1 file changed, 1 insertion(+)
arch/arm64/include/asm/tlbflush.h +1 -0
@@ -251,6 +251,7 @@ static inline void __flush_tlb_kernel_pg
 	dsb(ishst);
 	__tlbi(vaae1is, addr);
 	dsb(ish);
+	isb();
 }
 #endif
 
author maintainer
#260 lore raw

PATCH 257/313 arm64: dts: rockchip: limit clock rate of MMC controllers for RK3328

In reply to Greg Kroah-Hartman

From: Shawn Lin <[email protected]>

commit 03e61929c0d227ed3e1c322fc3804216ea298b7e upstream.

150MHz is a fundamental limitation of RK3328 Soc, w/o this limitation, eMMC, for instance, will run into 200MHz clock rate in HS200 mode, which makes the RK3328 boards not always boot properly. By adding it in rk3328.dtsi would also obviate the worry of missing it when adding new boards.

Fixes 52e02d377a72 ("arm64: dts: rockchip: add core dtsi file for RK3328 SoCs")
Cc Robin Murphy <[email protected]>
Cc Liang Chen <[email protected]>
---
 arch/arm64/boot/dts/rockchip/rk3328.dtsi |    3 +++
 1 file changed, 3 insertions(+)
arch/arm64/boot/dts/rockchip/rk3328.dtsi +3 -0
@@ -800,6 +800,7 @@
 			 <&cru SCLK_SDMMC_DRV>, <&cru SCLK_SDMMC_SAMPLE>;
 		clock-names = "biu", "ciu", "ciu-drive", "ciu-sample";
 		fifo-depth = <0x100>;
+		max-frequency = <150000000>;
 		status = "disabled";
 	};
 
@@ -811,6 +812,7 @@
 			 <&cru SCLK_SDIO_DRV>, <&cru SCLK_SDIO_SAMPLE>;
 		clock-names = "biu", "ciu", "ciu-drive", "ciu-sample";
 		fifo-depth = <0x100>;
+		max-frequency = <150000000>;
 		status = "disabled";
 	};
 
@@ -822,6 +824,7 @@
 			 <&cru SCLK_EMMC_DRV>, <&cru SCLK_EMMC_SAMPLE>;
 		clock-names = "biu", "ciu", "ciu-drive", "ciu-sample";
 		fifo-depth = <0x100>;
+		max-frequency = <150000000>;
 		status = "disabled";
 	};
 
author maintainer
#261 lore raw

PATCH 258/313 iommu/arm-smmu-v3: Disable detection of ATS and PRI

In reply to Greg Kroah-Hartman

From: Will Deacon <[email protected]>

commit b5e86196b83fd68e065a7c811ab8925fb0dc3893 upstream.

Detecting the ATS capability of the SMMU at probe time introduces a spinlock into the ->unmap() fast path, even when ATS is not actually in use. Furthermore, the ATC invalidation that exists is broken, as it occurs before invalidation of the main SMMU TLB which leaves a window where the ATC can be repopulated with stale entries.

Given that ATS is both a new feature and a specialist sport, disable it for now whilst we fix it properly in subsequent patches. Since PRI requires ATS, disable that too.

Fixes 9ce27afc0830 ("iommu/arm-smmu-v3: Add support for PCI ATS")
---
 drivers/iommu/arm-smmu-v3.c |    2 ++
 1 file changed, 2 insertions(+)
drivers/iommu/arm-smmu-v3.c +2 -0
@@ -2817,11 +2817,13 @@ static int arm_smmu_device_hw_probe(stru
 	}
 
 	/* Boolean feature flags */
+#if 0	/* ATS invalidation is slow and broken */
 	if (IS_ENABLED(CONFIG_PCI_PRI) && reg & IDR0_PRI)
 		smmu->features |= ARM_SMMU_FEAT_PRI;
 
 	if (IS_ENABLED(CONFIG_PCI_ATS) && reg & IDR0_ATS)
 		smmu->features |= ARM_SMMU_FEAT_ATS;
+#endif
 
 	if (reg & IDR0_SEV)
 		smmu->features |= ARM_SMMU_FEAT_SEV;
author maintainer
#262 lore raw

PATCH 259/313 alarmtimer: Use EOPNOTSUPP instead of ENOTSUPP

In reply to Greg Kroah-Hartman

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

commit f18ddc13af981ce3c7b7f26925f099e7c6929aba upstream.

ENOTSUPP is not supposed to be returned to userspace. This was found on an OpenPower machine, where the RTC does not support set_alarm.

On that system, a clock_nanosleep(CLOCK_REALTIME_ALARM, ...) results in "524 Unknown error 524"

Replace it with EOPNOTSUPP which results in the expected "95 Operation not supported" error.

---
 kernel/time/alarmtimer.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
kernel/time/alarmtimer.c +2 -2
@@ -673,7 +673,7 @@ static int alarm_timer_create(struct k_i
 	enum  alarmtimer_type type;
 
 	if (!alarmtimer_get_rtcdev())
-		return -ENOTSUPP;
+		return -EOPNOTSUPP;
 
 	if (!capable(CAP_WAKE_ALARM))
 		return -EPERM;
@@ -791,7 +791,7 @@ static int alarm_timer_nsleep(const cloc
 	int ret = 0;
 
 	if (!alarmtimer_get_rtcdev())
-		return -ENOTSUPP;
+		return -EOPNOTSUPP;
 
 	if (flags & ~TIMER_ABSTIME)
 		return -EINVAL;
author maintainer
#263 lore raw

PATCH 260/313 iommu/vt-d: Fix wrong analysis whether devices share the same bus

In reply to Greg Kroah-Hartman

From: Nadav Amit <[email protected]>

commit 2c70010867f164d1b30e787e360e05d10cc40046 upstream.

set_msi_sid_cb() is used to determine whether device aliases share the same bus, but it can provide false indications that aliases use the same bus when in fact they do not. The reason is that set_msi_sid_cb() assumes that pdev is fixed, while actually pci_for_each_dma_alias() can call fn() when pdev is set to a subordinate device.

As a result, running an VM on ESX with VT-d emulation enabled can results in the log warning such as:

DMAR: [INTR-REMAP] Request device [00:11.0] fault index 3b [fault reason 38] Blocked an interrupt request due to source-id verification failure

This seems to cause additional ata errors such as:
ata3.00: qc timeout (cmd 0xa1)
ata3.00: failed to IDENTIFY (I/O error, err_mask=0x4)

These timeouts also cause boot to be much longer and other errors.

Fix it by checking comparing the alias with the previous one instead.

Fixes 3f0c625c6ae71 ("iommu/vt-d: Allow interrupts from the entire bus for aliased devices")
Cc Logan Gunthorpe <[email protected]>
Cc David Woodhouse <[email protected]>
Cc Joerg Roedel <[email protected]>
Cc Jacob Pan <[email protected]>
---
 drivers/iommu/intel_irq_remapping.c |    6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/iommu/intel_irq_remapping.c +3 -3
@@ -376,13 +376,13 @@ static int set_msi_sid_cb(struct pci_dev
 {
 	struct set_msi_sid_data *data = opaque;
 
+	if (data->count == 0 || PCI_BUS_NUM(alias) == PCI_BUS_NUM(data->alias))
+		data->busmatch_count++;
+
 	data->pdev = pdev;
 	data->alias = alias;
 	data->count++;
 
-	if (PCI_BUS_NUM(alias) == pdev->bus->number)
-		data->busmatch_count++;
-
 	return 0;
 }
 
author maintainer
#264 lore raw

PATCH 261/313 regulator: Defer init completion for a while after late_initcall

In reply to Greg Kroah-Hartman

From: Mark Brown <[email protected]>

commit 55576cf1853798e86f620766e23b604c9224c19c upstream.

The kernel has no way of knowing when we have finished instantiating drivers, between deferred probe and systems that build key drivers as modules we might be doing this long after userspace has booted. This has always been a bit of an issue with regulator_init_complete since it can power off hardware that's not had it's driver loaded which can result in user visible effects, the main case is powering off displays. Practically speaking it's not been an issue in real systems since most systems that use the regulator API are embedded and build in key drivers anyway but with Arm laptops coming on the market it's becoming more of an issue so let's do something about it.

In the absence of any better idea just defer the powering off for 30s after late_initcall(), this is obviously a hack but it should mask the issue for now and it's no more arbitrary than late_initcall() itself. Ideally we'd have some heuristics to detect if we're on an affected system and tune or skip the delay appropriately, and there may be some need for a command line option to be added.

---
 drivers/regulator/core.c |   42 +++++++++++++++++++++++++++++++-----------
 1 file changed, 31 insertions(+), 11 deletions(-)
drivers/regulator/core.c +31 -11
@@ -5486,7 +5486,7 @@ static int __init regulator_init(void)
 /* init early to allow our consumers to complete system booting */
 core_initcall(regulator_init);
 
-static int __init regulator_late_cleanup(struct device *dev, void *data)
+static int regulator_late_cleanup(struct device *dev, void *data)
 {
 	struct regulator_dev *rdev = dev_to_rdev(dev);
 	const struct regulator_ops *ops = rdev->desc->ops;
@@ -5535,18 +5535,9 @@ unlock:
 	return 0;
 }
 
-static int __init regulator_init_complete(void)
+static void regulator_init_complete_work_function(struct work_struct *work)
 {
 	/*
-	 * Since DT doesn't provide an idiomatic mechanism for
-	 * enabling full constraints and since it's much more natural
-	 * with DT to provide them just assume that a DT enabled
-	 * system has full constraints.
-	 */
-	if (of_have_populated_dt())
-		has_full_constraints = true;
-
-	/*
 	 * Regulators may had failed to resolve their input supplies
 	 * when were registered, either because the input supply was
 	 * not registered yet or because its parent device was not
@@ -5563,6 +5554,35 @@ static int __init regulator_init_complet
 	 */
 	class_for_each_device(&regulator_class, NULL, NULL,
 			      regulator_late_cleanup);
+}
+
+static DECLARE_DELAYED_WORK(regulator_init_complete_work,
+			    regulator_init_complete_work_function);
+
+static int __init regulator_init_complete(void)
+{
+	/*
+	 * Since DT doesn't provide an idiomatic mechanism for
+	 * enabling full constraints and since it's much more natural
+	 * with DT to provide them just assume that a DT enabled
+	 * system has full constraints.
+	 */
+	if (of_have_populated_dt())
+		has_full_constraints = true;
+
+	/*
+	 * We punt completion for an arbitrary amount of time since
+	 * systems like distros will load many drivers from userspace
+	 * so consumers might not always be ready yet, this is
+	 * particularly an issue with laptops where this might bounce
+	 * the display off then on.  Ideally we'd get a notification
+	 * from userspace when this happens but we don't so just wait
+	 * a bit and hope we waited long enough.  It'd be better if
+	 * we'd only do this on systems that need it, and a kernel
+	 * command line option might be useful.
+	 */
+	schedule_delayed_work(&regulator_init_complete_work,
+			      msecs_to_jiffies(30000));
 
 	return 0;
 }
author maintainer
#265 lore raw

PATCH 262/313 efifb: BGRT: Improve efifb_bgrt_sanity_check

In reply to Greg Kroah-Hartman

From: Hans de Goede <[email protected]>

commit 51677dfcc17f88ed754143df670ff064eae67f84 upstream.

For various reasons, at least with x86 EFI firmwares, the xoffset and yoffset in the BGRT info are not always reliable.

Extensive testing has shown that when the info is correct, the BGRT image is always exactly centered horizontally (the yoffset variable is more variable and not always predictable).

This commit simplifies / improves the bgrt_sanity_check to simply check that the BGRT image is exactly centered horizontally and skips (re)drawing it when it is not.

This fixes the BGRT image sometimes being drawn in the wrong place.

Fixes 88fe4ceb2447 ("efifb: BGRT: Do not copy the boot graphics for non native resolutions")
Cc Peter Jones <[email protected]>,
---
 drivers/video/fbdev/efifb.c |   27 ++++++---------------------
 1 file changed, 6 insertions(+), 21 deletions(-)
drivers/video/fbdev/efifb.c +6 -21
@@ -122,28 +122,13 @@ static void efifb_copy_bmp(u8 *src, u32
  */
 static bool efifb_bgrt_sanity_check(struct screen_info *si, u32 bmp_width)
 {
-	static const int default_resolutions[][2] = {
-		{  800,  600 },
-		{ 1024,  768 },
-		{ 1280, 1024 },
-	};
-	u32 i, right_margin;
+	/*
+	 * All x86 firmwares horizontally center the image (the yoffset
+	 * calculations differ between boards, but xoffset is predictable).
+	 */
+	u32 expected_xoffset = (si->lfb_width - bmp_width) / 2;
 
-	for (i = 0; i < ARRAY_SIZE(default_resolutions); i++) {
-		if (default_resolutions[i][0] == si->lfb_width &&
-		    default_resolutions[i][1] == si->lfb_height)
-			break;
-	}
-	/* If not a default resolution used for textmode, this should be fine */
-	if (i >= ARRAY_SIZE(default_resolutions))
-		return true;
-
-	/* If the right margin is 5 times smaller then the left one, reject */
-	right_margin = si->lfb_width - (bgrt_tab.image_offset_x + bmp_width);
-	if (right_margin < (bgrt_tab.image_offset_x / 5))
-		return false;
-
-	return true;
+	return bgrt_tab.image_offset_x == expected_xoffset;
 }
 #else
 static bool efifb_bgrt_sanity_check(struct screen_info *si, u32 bmp_width)
author maintainer
#266 lore raw

PATCH 263/313 gfs2: clear buf_in_tr when ending a transaction in sweep_bh_for_rgrps

In reply to Greg Kroah-Hartman

From: Bob Peterson <[email protected]>

commit f0b444b349e33ae0d3dd93e25ca365482a5d17d4 upstream.

In function sweep_bh_for_rgrps, which is a helper for punch_hole, it uses variable buf_in_tr to keep track of when it needs to commit pending block frees on a partial delete that overflows the transaction created for the delete. The problem is that the variable was initialized at the start of function sweep_bh_for_rgrps but it was never cleared, even when starting a new transaction.

This patch reinitializes the variable when the transaction is ended, so the next transaction starts out with it cleared.

Fixes d552a2b9b33e ("GFS2: Non-recursive delete")
Cc [email protected] # v4.12+
---
 fs/gfs2/bmap.c |    1 +
 1 file changed, 1 insertion(+)
fs/gfs2/bmap.c +1 -0
@@ -1670,6 +1670,7 @@ out_unlock:
 			brelse(dibh);
 			up_write(&ip->i_rw_mutex);
 			gfs2_trans_end(sdp);
+			buf_in_tr = false;
 		}
 		gfs2_glock_dq_uninit(rd_gh);
 		cond_resched();
author maintainer
#267 lore raw

PATCH 264/313 z3fold: fix retry mechanism in page reclaim

In reply to Greg Kroah-Hartman

From: Vitaly Wool <[email protected]>

commit 3f9d2b5766aea06042630ac60b7316fd0cebf06f upstream.

z3fold_page_reclaim()'s retry mechanism is broken: on a second iteration it will have zhdr from the first one so that zhdr is no longer in line with struct page. That leads to crashes when the system is stressed.

Fix that by moving zhdr assignment up.

While at it, protect against using already freed handles by using own local slots structure in z3fold_page_reclaim().

---
 mm/z3fold.c |   49 ++++++++++++++++++++++++++++++++++---------------
 1 file changed, 34 insertions(+), 15 deletions(-)
mm/z3fold.c +34 -15
@@ -368,9 +368,10 @@ static inline int __idx(struct z3fold_he
  * Encodes the handle of a particular buddy within a z3fold page
  * Pool lock should be held as this function accesses first_num
  */
-static unsigned long encode_handle(struct z3fold_header *zhdr, enum buddy bud)
+static unsigned long __encode_handle(struct z3fold_header *zhdr,
+				struct z3fold_buddy_slots *slots,
+				enum buddy bud)
 {
-	struct z3fold_buddy_slots *slots;
 	unsigned long h = (unsigned long)zhdr;
 	int idx = 0;
 
@@ -387,11 +388,15 @@ static unsigned long encode_handle(struc
 	if (bud == LAST)
 		h |= (zhdr->last_chunks << BUDDY_SHIFT);
 
-	slots = zhdr->slots;
 	slots->slot[idx] = h;
 	return (unsigned long)&slots->slot[idx];
 }
 
+static unsigned long encode_handle(struct z3fold_header *zhdr, enum buddy bud)
+{
+	return __encode_handle(zhdr, zhdr->slots, bud);
+}
+
 /* Returns the z3fold page where a given handle is stored */
 static inline struct z3fold_header *handle_to_z3fold_header(unsigned long h)
 {
@@ -626,6 +631,7 @@ static void do_compact_page(struct z3fol
 	}
 
 	if (unlikely(PageIsolated(page) ||
+		     test_bit(PAGE_CLAIMED, &page->private) ||
 		     test_bit(PAGE_STALE, &page->private))) {
 		z3fold_page_unlock(zhdr);
 		return;
@@ -1102,6 +1108,7 @@ static int z3fold_reclaim_page(struct z3
 	struct z3fold_header *zhdr = NULL;
 	struct page *page = NULL;
 	struct list_head *pos;
+	struct z3fold_buddy_slots slots;
 	unsigned long first_handle = 0, middle_handle = 0, last_handle = 0;
 
 	spin_lock(&pool->lock);
@@ -1120,16 +1127,22 @@ static int z3fold_reclaim_page(struct z3
 			/* this bit could have been set by free, in which case
 			 * we pass over to the next page in the pool.
 			 */
-			if (test_and_set_bit(PAGE_CLAIMED, &page->private))
+			if (test_and_set_bit(PAGE_CLAIMED, &page->private)) {
+				page = NULL;
 				continue;
+			}
 
-			if (unlikely(PageIsolated(page)))
+			if (unlikely(PageIsolated(page))) {
+				clear_bit(PAGE_CLAIMED, &page->private);
+				page = NULL;
 				continue;
+			}
+			zhdr = page_address(page);
 			if (test_bit(PAGE_HEADLESS, &page->private))
 				break;
 
-			zhdr = page_address(page);
 			if (!z3fold_page_trylock(zhdr)) {
+				clear_bit(PAGE_CLAIMED, &page->private);
 				zhdr = NULL;
 				continue; /* can't evict at this point */
 			}
@@ -1147,26 +1160,30 @@ static int z3fold_reclaim_page(struct z3
 
 		if (!test_bit(PAGE_HEADLESS, &page->private)) {
 			/*
-			 * We need encode the handles before unlocking, since
-			 * we can race with free that will set
-			 * (first|last)_chunks to 0
+			 * We need encode the handles before unlocking, and
+			 * use our local slots structure because z3fold_free
+			 * can zero out zhdr->slots and we can't do much
+			 * about that
 			 */
 			first_handle = 0;
 			last_handle = 0;
 			middle_handle = 0;
 			if (zhdr->first_chunks)
-				first_handle = encode_handle(zhdr, FIRST);
+				first_handle = __encode_handle(zhdr, &slots,
+								FIRST);
 			if (zhdr->middle_chunks)
-				middle_handle = encode_handle(zhdr, MIDDLE);
+				middle_handle = __encode_handle(zhdr, &slots,
+								MIDDLE);
 			if (zhdr->last_chunks)
-				last_handle = encode_handle(zhdr, LAST);
+				last_handle = __encode_handle(zhdr, &slots,
+								LAST);
 			/*
 			 * it's safe to unlock here because we hold a
 			 * reference to this page
 			 */
 			z3fold_page_unlock(zhdr);
 		} else {
-			first_handle = encode_handle(zhdr, HEADLESS);
+			first_handle = __encode_handle(zhdr, &slots, HEADLESS);
 			last_handle = middle_handle = 0;
 		}
 
@@ -1196,9 +1213,9 @@ next:
 			spin_lock(&pool->lock);
 			list_add(&page->lru, &pool->lru);
 			spin_unlock(&pool->lock);
+			clear_bit(PAGE_CLAIMED, &page->private);
 		} else {
 			z3fold_page_lock(zhdr);
-			clear_bit(PAGE_CLAIMED, &page->private);
 			if (kref_put(&zhdr->refcount,
 					release_z3fold_page_locked)) {
 				atomic64_dec(&pool->pages_nr);
@@ -1213,6 +1230,7 @@ next:
 			list_add(&page->lru, &pool->lru);
 			spin_unlock(&pool->lock);
 			z3fold_page_unlock(zhdr);
+			clear_bit(PAGE_CLAIMED, &page->private);
 		}
 
 		/* We started off locked to we need to lock the pool back */
@@ -1317,7 +1335,8 @@ static bool z3fold_page_isolate(struct p
 	VM_BUG_ON_PAGE(!PageMovable(page), page);
 	VM_BUG_ON_PAGE(PageIsolated(page), page);
 
-	if (test_bit(PAGE_HEADLESS, &page->private))
+	if (test_bit(PAGE_HEADLESS, &page->private) ||
+	    test_bit(PAGE_CLAIMED, &page->private))
 		return false;
 
 	zhdr = page_address(page);
author maintainer
#268 lore raw

PATCH 265/313 z3fold: fix memory leak in kmem cache

In reply to Greg Kroah-Hartman

From: Vitaly Wool <[email protected]>

commit 63398413c00c7836ea87a1fa205c91d2199b25cf upstream.

Currently there is a leak in init_z3fold_page() -- it allocates handles from kmem cache even for headless pages, but then they are never used and never freed, so eventually kmem cache may get exhausted. This patch provides a fix for that.

---
 mm/z3fold.c |   15 +++++++++------
 1 file changed, 9 insertions(+), 6 deletions(-)
mm/z3fold.c +9 -6
@@ -297,14 +297,11 @@ static void z3fold_unregister_migration(
  }
 
 /* Initializes the z3fold header of a newly allocated z3fold page */
-static struct z3fold_header *init_z3fold_page(struct page *page,
+static struct z3fold_header *init_z3fold_page(struct page *page, bool headless,
 					struct z3fold_pool *pool, gfp_t gfp)
 {
 	struct z3fold_header *zhdr = page_address(page);
-	struct z3fold_buddy_slots *slots = alloc_slots(pool, gfp);
-
-	if (!slots)
-		return NULL;
+	struct z3fold_buddy_slots *slots;
 
 	INIT_LIST_HEAD(&page->lru);
 	clear_bit(PAGE_HEADLESS, &page->private);
@@ -312,6 +309,12 @@ static struct z3fold_header *init_z3fold
 	clear_bit(NEEDS_COMPACTING, &page->private);
 	clear_bit(PAGE_STALE, &page->private);
 	clear_bit(PAGE_CLAIMED, &page->private);
+	if (headless)
+		return zhdr;
+
+	slots = alloc_slots(pool, gfp);
+	if (!slots)
+		return NULL;
 
 	spin_lock_init(&zhdr->page_lock);
 	kref_init(&zhdr->refcount);
@@ -932,7 +935,7 @@ retry:
 	if (!page)
 		return -ENOMEM;
 
-	zhdr = init_z3fold_page(page, pool, gfp);
+	zhdr = init_z3fold_page(page, bud == HEADLESS, pool, gfp);
 	if (!zhdr) {
 		__free_page(page);
 		return -ENOMEM;
author maintainer
#269 lore raw

PATCH 266/313 mm/compaction.c: clear total_{migrate,free}_scanned before scanning a new zone

In reply to Greg Kroah-Hartman

From: Yafang Shao <[email protected]>

commit a94b525241c0fff3598809131d7cfcfe1d572d8c upstream.

total_{migrate,free}_scanned will be added to COMPACTMIGRATE_SCANNED and COMPACTFREE_SCANNED in compact_zone(). We should clear them before scanning a new zone. In the proc triggered compaction, we forgot clearing them.

[[email protected]: introduce a helper compact_zone_counters_init()]
[[email protected]: expand compact_zone_counters_init() into its single callsite, per mhocko]
[[email protected]: squash compact_zone() list_head init as well]
[[email protected]: kcompactd_do_work(): avoid unnecessary initialization of cc.zone]
Fixes 7f354a548d1c ("mm, compaction: add vmstats for kcompactd work")
Cc David Rientjes <[email protected]>
Cc Yafang Shao <[email protected]>
Cc Mel Gorman <[email protected]>
Cc Michal Hocko <[email protected]>
---
 mm/compaction.c |   35 +++++++++++++----------------------
 1 file changed, 13 insertions(+), 22 deletions(-)
mm/compaction.c +13 -22
@@ -2078,6 +2078,17 @@ compact_zone(struct compact_control *cc,
 	const bool sync = cc->mode != MIGRATE_ASYNC;
 	bool update_cached;
 
+	/*
+	 * These counters track activities during zone compaction.  Initialize
+	 * them before compacting a new zone.
+	 */
+	cc->total_migrate_scanned = 0;
+	cc->total_free_scanned = 0;
+	cc->nr_migratepages = 0;
+	cc->nr_freepages = 0;
+	INIT_LIST_HEAD(&cc->freepages);
+	INIT_LIST_HEAD(&cc->migratepages);
+
 	cc->migratetype = gfpflags_to_migratetype(cc->gfp_mask);
 	ret = compaction_suitable(cc->zone, cc->order, cc->alloc_flags,
 							cc->classzone_idx);
@@ -2281,10 +2292,6 @@ static enum compact_result compact_zone_
 {
 	enum compact_result ret;
 	struct compact_control cc = {
-		.nr_freepages = 0,
-		.nr_migratepages = 0,
-		.total_migrate_scanned = 0,
-		.total_free_scanned = 0,
 		.order = order,
 		.search_order = order,
 		.gfp_mask = gfp_mask,
@@ -2305,8 +2312,6 @@ static enum compact_result compact_zone_
 
 	if (capture)
 		current->capture_control = &capc;
-	INIT_LIST_HEAD(&cc.freepages);
-	INIT_LIST_HEAD(&cc.migratepages);
 
 	ret = compact_zone(&cc, &capc);
 
@@ -2408,8 +2413,6 @@ static void compact_node(int nid)
 	struct zone *zone;
 	struct compact_control cc = {
 		.order = -1,
-		.total_migrate_scanned = 0,
-		.total_free_scanned = 0,
 		.mode = MIGRATE_SYNC,
 		.ignore_skip_hint = true,
 		.whole_zone = true,
@@ -2423,11 +2426,7 @@ static void compact_node(int nid)
 		if (!populated_zone(zone))
 			continue;
 
-		cc.nr_freepages = 0;
-		cc.nr_migratepages = 0;
 		cc.zone = zone;
-		INIT_LIST_HEAD(&cc.freepages);
-		INIT_LIST_HEAD(&cc.migratepages);
 
 		compact_zone(&cc, NULL);
 
@@ -2529,8 +2528,6 @@ static void kcompactd_do_work(pg_data_t
 	struct compact_control cc = {
 		.order = pgdat->kcompactd_max_order,
 		.search_order = pgdat->kcompactd_max_order,
-		.total_migrate_scanned = 0,
-		.total_free_scanned = 0,
 		.classzone_idx = pgdat->kcompactd_classzone_idx,
 		.mode = MIGRATE_SYNC_LIGHT,
 		.ignore_skip_hint = false,
@@ -2554,16 +2551,10 @@ static void kcompactd_do_work(pg_data_t
 							COMPACT_CONTINUE)
 			continue;
 
-		cc.nr_freepages = 0;
-		cc.nr_migratepages = 0;
-		cc.total_migrate_scanned = 0;
-		cc.total_free_scanned = 0;
-		cc.zone = zone;
-		INIT_LIST_HEAD(&cc.freepages);
-		INIT_LIST_HEAD(&cc.migratepages);
-
 		if (kthread_should_stop())
 			return;
+
+		cc.zone = zone;
 		status = compact_zone(&cc, NULL);
 
 		if (status == COMPACT_SUCCESS) {
author maintainer
#270 lore raw

PATCH 267/313 memcg, oom: dont require __GFP_FS when invoking memcg OOM killer

In reply to Greg Kroah-Hartman

From: Tetsuo Handa <[email protected]>

commit f9c645621a28e37813a1de96d9cbd89cde94a1e4 upstream.

Masoud Sharbiani noticed that commit 29ef680ae7c21110 ("memcg, oom: move out_of_memory back to the charge path") broke memcg OOM called from __xfs_filemap_fault() path. It turned out that try_charge() is retrying forever without making forward progress because mem_cgroup_oom(GFP_NOFS) cannot invoke the OOM killer due to commit 3da88fb3bacfaa33 ("mm, oom:
move GFP_NOFS check to out_of_memory").

Allowing forced charge due to being unable to invoke memcg OOM killer will lead to global OOM situation. Also, just returning -ENOMEM will be risky because OOM path is lost and some paths (e.g. get_user_pages()) will leak -ENOMEM. Therefore, invoking memcg OOM killer (despite GFP_NOFS) will be the only choice we can choose for now.

Until 29ef680ae7c21110, we were able to invoke memcg OOM killer when GFP_KERNEL reclaim failed [1]. But since 29ef680ae7c21110, we need to invoke memcg OOM killer when GFP_NOFS reclaim failed [2]. Although in the past we did invoke memcg OOM killer for GFP_NOFS [3], we might get pre-mature memcg OOM reports due to this patch.

[1]

leaker invoked oom-killer: gfp_mask=0x6200ca(GFP_HIGHUSER_MOVABLE), nodemask=(null), order=0, oom_score_adj=0

 CPU: 0 PID: 2746 Comm: leaker Not tainted 4.18.0+ #19
 Hardware name: VMware, Inc. VMware Virtual Platform/440BX Desktop Reference Platform, BIOS 6.00 04/13/2018
 Call Trace:
  dump_stack+0x63/0x88
  dump_header+0x67/0x27a
  ? mem_cgroup_scan_tasks+0x91/0xf0
  oom_kill_process+0x210/0x410
  out_of_memory+0x10a/0x2c0
  mem_cgroup_out_of_memory+0x46/0x80
  mem_cgroup_oom_synchronize+0x2e4/0x310
  ? high_work_func+0x20/0x20
  pagefault_out_of_memory+0x31/0x76
  mm_fault_error+0x55/0x115
  ? handle_mm_fault+0xfd/0x220
  __do_page_fault+0x433/0x4e0
  do_page_fault+0x22/0x30
  ? page_fault+0x8/0x30
  page_fault+0x1e/0x30
 RIP: 0033:0x4009f0
 Code: 03 00 00 00 e8 71 fd ff ff 48 83 f8 ff 49 89 c6 74 74 48 89 c6 bf c0 0c 40 00 31 c0 e8 69 fd ff ff 45 85 ff 7e 21 31 c9 66 90 <41> 0f be 14 0e 01 d3 f7 c1 ff 0f 00 00 75 05 41 c6 04 0e 2a 48 83
 RSP: 002b:00007ffe29ae96f0 EFLAGS: 00010206
 RAX: 000000000000001b RBX: 0000000000000000 RCX: 0000000001ce1000
 RDX: 0000000000000000 RSI: 000000007fffffe5 RDI: 0000000000000000
 RBP: 000000000000000c R08: 0000000000000000 R09: 00007f94be09220d
 R10: 0000000000000002 R11: 0000000000000246 R12: 00000000000186a0
 R13: 0000000000000003 R14: 00007f949d845000 R15: 0000000002800000
 Task in /leaker killed as a result of limit of /leaker
 memory: usage 524288kB, limit 524288kB, failcnt 158965
 memory+swap: usage 0kB, limit 9007199254740988kB, failcnt 0
 kmem: usage 2016kB, limit 9007199254740988kB, failcnt 0
 Memory cgroup stats for /leaker: cache:844KB rss:521136KB rss_huge:0KB shmem:0KB mapped_file:0KB dirty:132KB writeback:0KB inactive_anon:0KB active_anon:521224KB inactive_file:1012KB active_file:8KB unevictable:0KB
 Memory cgroup out of memory: Kill process 2746 (leaker) score 998 or sacrifice child
 Killed process 2746 (leaker) total-vm:536704kB, anon-rss:521176kB, file-rss:1208kB, shmem-rss:0kB
 oom_reaper: reaped process 2746 (leaker), now anon-rss:0kB, file-rss:0kB, shmem-rss:0kB

[2]

leaker invoked oom-killer: gfp_mask=0x600040(GFP_NOFS), nodemask=(null), order=0, oom_score_adj=0

 CPU: 1 PID: 2746 Comm: leaker Not tainted 4.18.0+ #20
 Hardware name: VMware, Inc. VMware Virtual Platform/440BX Desktop Reference Platform, BIOS 6.00 04/13/2018
 Call Trace:
  dump_stack+0x63/0x88
  dump_header+0x67/0x27a
  ? mem_cgroup_scan_tasks+0x91/0xf0
  oom_kill_process+0x210/0x410
  out_of_memory+0x109/0x2d0
  mem_cgroup_out_of_memory+0x46/0x80
  try_charge+0x58d/0x650
  ? __radix_tree_replace+0x81/0x100
  mem_cgroup_try_charge+0x7a/0x100
  __add_to_page_cache_locked+0x92/0x180
  add_to_page_cache_lru+0x4d/0xf0
  iomap_readpages_actor+0xde/0x1b0
  ? iomap_zero_range_actor+0x1d0/0x1d0
  iomap_apply+0xaf/0x130
  iomap_readpages+0x9f/0x150
  ? iomap_zero_range_actor+0x1d0/0x1d0
  xfs_vm_readpages+0x18/0x20 [xfs]
  read_pages+0x60/0x140
  __do_page_cache_readahead+0x193/0x1b0
  ondemand_readahead+0x16d/0x2c0
  page_cache_async_readahead+0x9a/0xd0
  filemap_fault+0x403/0x620
  ? alloc_set_pte+0x12c/0x540
  ? _cond_resched+0x14/0x30
  __xfs_filemap_fault+0x66/0x180 [xfs]
  xfs_filemap_fault+0x27/0x30 [xfs]
  __do_fault+0x19/0x40
  __handle_mm_fault+0x8e8/0xb60
  handle_mm_fault+0xfd/0x220
  __do_page_fault+0x238/0x4e0
  do_page_fault+0x22/0x30
  ? page_fault+0x8/0x30
  page_fault+0x1e/0x30
 RIP: 0033:0x4009f0
 Code: 03 00 00 00 e8 71 fd ff ff 48 83 f8 ff 49 89 c6 74 74 48 89 c6 bf c0 0c 40 00 31 c0 e8 69 fd ff ff 45 85 ff 7e 21 31 c9 66 90 <41> 0f be 14 0e 01 d3 f7 c1 ff 0f 00 00 75 05 41 c6 04 0e 2a 48 83
 RSP: 002b:00007ffda45c9290 EFLAGS: 00010206
 RAX: 000000000000001b RBX: 0000000000000000 RCX: 0000000001a1e000
 RDX: 0000000000000000 RSI: 000000007fffffe5 RDI: 0000000000000000
 RBP: 000000000000000c R08: 0000000000000000 R09: 00007f6d061ff20d
 R10: 0000000000000002 R11: 0000000000000246 R12: 00000000000186a0
 R13: 0000000000000003 R14: 00007f6ce59b2000 R15: 0000000002800000
 Task in /leaker killed as a result of limit of /leaker
 memory: usage 524288kB, limit 524288kB, failcnt 7221
 memory+swap: usage 0kB, limit 9007199254740988kB, failcnt 0
 kmem: usage 1944kB, limit 9007199254740988kB, failcnt 0
 Memory cgroup stats for /leaker: cache:3632KB rss:518232KB rss_huge:0KB shmem:0KB mapped_file:0KB dirty:0KB writeback:0KB inactive_anon:0KB active_anon:518408KB inactive_file:3908KB active_file:12KB unevictable:0KB
 Memory cgroup out of memory: Kill process 2746 (leaker) score 992 or sacrifice child
 Killed process 2746 (leaker) total-vm:536704kB, anon-rss:518264kB, file-rss:1188kB, shmem-rss:0kB
 oom_reaper: reaped process 2746 (leaker), now anon-rss:0kB, file-rss:0kB, shmem-rss:0kB

[3]

leaker invoked oom-killer: gfp_mask=0x50, order=0, oom_score_adj=0 leaker cpuset=/ mems_allowed=0

 CPU: 1 PID: 3206 Comm: leaker Not tainted 3.10.0-957.27.2.el7.x86_64 #1
 Hardware name: VMware, Inc. VMware Virtual Platform/440BX Desktop Reference Platform, BIOS 6.00 04/13/2018
 Call Trace:
  [<ffffffffaf364147>] dump_stack+0x19/0x1b
  [<ffffffffaf35eb6a>] dump_header+0x90/0x229
  [<ffffffffaedbb456>] ? find_lock_task_mm+0x56/0xc0
  [<ffffffffaee32a38>] ? try_get_mem_cgroup_from_mm+0x28/0x60
  [<ffffffffaedbb904>] oom_kill_process+0x254/0x3d0
  [<ffffffffaee36c36>] mem_cgroup_oom_synchronize+0x546/0x570
  [<ffffffffaee360b0>] ? mem_cgroup_charge_common+0xc0/0xc0
  [<ffffffffaedbc194>] pagefault_out_of_memory+0x14/0x90
  [<ffffffffaf35d072>] mm_fault_error+0x6a/0x157
  [<ffffffffaf3717c8>] __do_page_fault+0x3c8/0x4f0
  [<ffffffffaf371925>] do_page_fault+0x35/0x90
  [<ffffffffaf36d768>] page_fault+0x28/0x30
 Task in /leaker killed as a result of limit of /leaker
 memory: usage 524288kB, limit 524288kB, failcnt 20628
 memory+swap: usage 524288kB, limit 9007199254740988kB, failcnt 0
 kmem: usage 0kB, limit 9007199254740988kB, failcnt 0
 Memory cgroup stats for /leaker: cache:840KB rss:523448KB rss_huge:0KB mapped_file:0KB swap:0KB inactive_anon:0KB active_anon:523448KB inactive_file:464KB active_file:376KB unevictable:0KB
 Memory cgroup out of memory: Kill process 3206 (leaker) score 970 or sacrifice child
 Killed process 3206 (leaker) total-vm:536692kB, anon-rss:523304kB, file-rss:412kB, shmem-rss:0kB

Bisected by Masoud Sharbiani.

Fixes 3da88fb3bacfaa33 ("mm, oom: move GFP_NOFS check to out_of_memory") [necessary after 29ef680ae7c21110]
Cc David Rientjes <[email protected]>
Cc <[email protected]> [4.19+]
---
 mm/oom_kill.c |    5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)
mm/oom_kill.c +3 -2
@@ -1060,9 +1060,10 @@ bool out_of_memory(struct oom_control *o
 	 * The OOM killer does not compensate for IO-less reclaim.
 	 * pagefault_out_of_memory lost its gfp context so we have to
 	 * make sure exclude 0 mask - all other users should have at least
-	 * ___GFP_DIRECT_RECLAIM to get here.
+	 * ___GFP_DIRECT_RECLAIM to get here. But mem_cgroup_oom() has to
+	 * invoke the OOM killer even if it is a GFP_NOFS allocation.
 	 */
-	if (oc->gfp_mask && !(oc->gfp_mask & __GFP_FS))
+	if (oc->gfp_mask && !(oc->gfp_mask & __GFP_FS) && !is_memcg_oom(oc))
 		return true;
 
 	/*
author maintainer
#271 lore raw

PATCH 268/313 memcg, kmem: do not fail __GFP_NOFAIL charges

In reply to Greg Kroah-Hartman

From: Michal Hocko <[email protected]>

commit e55d9d9bfb69405bd7615c0f8d229d8fafb3e9b8 upstream.

Thomas has noticed the following NULL ptr dereference when using cgroup v1 kmem limit:

BUG: unable to handle kernel NULL pointer dereference at 0000000000000008
PGD 0
P4D 0
Oops: 0000 [#1] PREEMPT SMP PTI
CPU: 3 PID: 16923 Comm: gtk-update-icon Not tainted 4.19.51 #42
Hardware name: Gigabyte Technology Co., Ltd. Z97X-Gaming G1/Z97X-Gaming G1, BIOS F9 07/31/2015
RIP: 0010:create_empty_buffers+0x24/0x100
Code: cd 0f 1f 44 00 00 0f 1f 44 00 00 41 54 49 89 d4 ba 01 00 00 00 55 53 48 89 fb e8 97 fe ff ff 48 89 c5 48 89 c2 eb 03 48 89 ca <48> 8b 4a 08 4c 09 22 48 85 c9 75 f1 48 89 6a 08 48 8b 43 18 48 8d
RSP: 0018:ffff927ac1b37bf8 EFLAGS: 00010286
RAX: 0000000000000000 RBX: fffff2d4429fd740 RCX: 0000000100097149
RDX: 0000000000000000 RSI: 0000000000000082 RDI: ffff9075a99fbe00
RBP: 0000000000000000 R08: fffff2d440949cc8 R09: 00000000000960c0
R10: 0000000000000002 R11: 0000000000000000 R12: 0000000000000000
R13: ffff907601f18360 R14: 0000000000002000 R15: 0000000000001000
FS:  00007fb55b288bc0(0000) GS:ffff90761f8c0000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000000000008 CR3: 000000007aebc002 CR4: 00000000001606e0
Call Trace:
 create_page_buffers+0x4d/0x60
 __block_write_begin_int+0x8e/0x5a0
 ? ext4_inode_attach_jinode.part.82+0xb0/0xb0
 ? jbd2__journal_start+0xd7/0x1f0
 ext4_da_write_begin+0x112/0x3d0
 generic_perform_write+0xf1/0x1b0
 ? file_update_time+0x70/0x140
 __generic_file_write_iter+0x141/0x1a0
 ext4_file_write_iter+0xef/0x3b0
 __vfs_write+0x17e/0x1e0
 vfs_write+0xa5/0x1a0
 ksys_write+0x57/0xd0
 do_syscall_64+0x55/0x160
 entry_SYSCALL_64_after_hwframe+0x44/0xa9

Tetsuo then noticed that this is because the __memcg_kmem_charge_memcg fails __GFP_NOFAIL charge when the kmem limit is reached. This is a wrong behavior because nofail allocations are not allowed to fail. Normal charge path simply forces the charge even if that means to cross the limit. Kmem accounting should be doing the same.

---
 mm/memcontrol.c |   10 ++++++++++
 1 file changed, 10 insertions(+)
mm/memcontrol.c +10 -0
@@ -2719,6 +2719,16 @@ int __memcg_kmem_charge_memcg(struct pag
 
 	if (!cgroup_subsys_on_dfl(memory_cgrp_subsys) &&
 	    !page_counter_try_charge(&memcg->kmem, nr_pages, &counter)) {
+
+		/*
+		 * Enforce __GFP_NOFAIL allocation because callers are not
+		 * prepared to see failures and likely do not have any failure
+		 * handling code.
+		 */
+		if (gfp & __GFP_NOFAIL) {
+			page_counter_charge(&memcg->kmem, nr_pages);
+			return 0;
+		}
 		cancel_charge(memcg, nr_pages);
 		return -ENOMEM;
 	}
author maintainer
#272 lore raw

PATCH 269/313 lib/lzo/lzo1x_compress.c: fix alignment bug in lzo-rle

In reply to Greg Kroah-Hartman

From: Dave Rodgman <[email protected]>

commit 09b35b4192f6682dff96a093ab1930998cdb73b4 upstream.

Fix an unaligned access which breaks on platforms where this is not permitted (e.g., Sparc).

---
 lib/lzo/lzo1x_compress.c |   14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)
lib/lzo/lzo1x_compress.c +8 -6
@@ -83,17 +83,19 @@ next:
 					ALIGN((uintptr_t)ir, 4)) &&
 					(ir < limit) && (*ir == 0))
 				ir++;
-			for (; (ir + 4) <= limit; ir += 4) {
-				dv = *((u32 *)ir);
-				if (dv) {
+			if (IS_ALIGNED((uintptr_t)ir, 4)) {
+				for (; (ir + 4) <= limit; ir += 4) {
+					dv = *((u32 *)ir);
+					if (dv) {
 #  if defined(__LITTLE_ENDIAN)
-					ir += __builtin_ctz(dv) >> 3;
+						ir += __builtin_ctz(dv) >> 3;
 #  elif defined(__BIG_ENDIAN)
-					ir += __builtin_clz(dv) >> 3;
+						ir += __builtin_clz(dv) >> 3;
 #  else
 #    error "missing endian definition"
 #  endif
-					break;
+						break;
+					}
 				}
 			}
 #endif
author maintainer
#273 lore raw

PATCH 270/313 mt76: round up length on mt76_wr_copy

In reply to Greg Kroah-Hartman

From: Felix Fietkau <[email protected]>

commit 850e8f6fbd5d0003b0f1119d19a01c6fef1644e2 upstream.

When beacon length is not a multiple of 4, the beacon could be sent with the last 1-3 bytes corrupted. The skb data is guaranteed to have enough room for reading beyond the end, because it is always followed by skb_shared_info, so rounding up is safe.
All other callers of mt76_wr_copy have multiple-of-4 length already.

---
 drivers/net/wireless/mediatek/mt76/mmio.c |    2 +-
 drivers/net/wireless/mediatek/mt76/usb.c  |    2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)
drivers/net/wireless/mediatek/mt76/mmio.c +1 -1
@@ -43,7 +43,7 @@ static u32 mt76_mmio_rmw(struct mt76_dev
 static void mt76_mmio_copy(struct mt76_dev *dev, u32 offset, const void *data,
 			   int len)
 {
-	__iowrite32_copy(dev->mmio.regs + offset, data, len >> 2);
+	__iowrite32_copy(dev->mmio.regs + offset, data, DIV_ROUND_UP(len, 4));
 }
 
 static int mt76_mmio_wr_rp(struct mt76_dev *dev, u32 base,
drivers/net/wireless/mediatek/mt76/usb.c +1 -1
@@ -164,7 +164,7 @@ static void mt76u_copy(struct mt76_dev *
 	int i, ret;
 
 	mutex_lock(&usb->usb_ctrl_mtx);
-	for (i = 0; i < (len / 4); i++) {
+	for (i = 0; i < DIV_ROUND_UP(len, 4); i++) {
 		put_unaligned_le32(val[i], usb->data);
 		ret = __mt76u_vendor_request(dev, MT_VEND_MULTI_WRITE,
 					     USB_DIR_OUT | USB_TYPE_VENDOR,
author maintainer
#274 lore raw

PATCH 271/313 KEYS: trusted: correctly initialize digests and fix locking issue

In reply to Greg Kroah-Hartman

From: Roberto Sassu <[email protected]>

commit 9f75c82246313d4c2a6bc77e947b45655b3b5ad5 upstream.

Commit 0b6cf6b97b7e ("tpm: pass an array of tpm_extend_digest structures to tpm_pcr_extend()") modifies tpm_pcr_extend() to accept a digest for each PCR bank. After modification, tpm_pcr_extend() expects that digests are passed in the same order as the algorithms set in chip->allocated_banks.

This patch fixes two issues introduced in the last iterations of the patch
set: missing initialization of the TPM algorithm ID in the tpm_digest structures passed to tpm_pcr_extend() by the trusted key module, and unreleased locks in the TPM driver due to returning from tpm_pcr_extend() without calling tpm_put_ops().

Fixes 0b6cf6b97b7e ("tpm: pass an array of tpm_extend_digest structures to tpm_pcr_extend()")
---
 drivers/char/tpm/tpm-interface.c |   14 +++++++++-----
 security/keys/trusted.c          |    5 +++++
 2 files changed, 14 insertions(+), 5 deletions(-)
drivers/char/tpm/tpm-interface.c +9 -5
@@ -320,18 +320,22 @@ int tpm_pcr_extend(struct tpm_chip *chip
 	if (!chip)
 		return -ENODEV;
 
-	for (i = 0; i < chip->nr_allocated_banks; i++)
-		if (digests[i].alg_id != chip->allocated_banks[i].alg_id)
-			return -EINVAL;
+	for (i = 0; i < chip->nr_allocated_banks; i++) {
+		if (digests[i].alg_id != chip->allocated_banks[i].alg_id) {
+			rc = EINVAL;
+			goto out;
+		}
+	}
 
 	if (chip->flags & TPM_CHIP_FLAG_TPM2) {
 		rc = tpm2_pcr_extend(chip, pcr_idx, digests);
-		tpm_put_ops(chip);
-		return rc;
+		goto out;
 	}
 
 	rc = tpm1_pcr_extend(chip, pcr_idx, digests[0].digest,
 			     "attempting extend a PCR value");
+
+out:
 	tpm_put_ops(chip);
 	return rc;
 }
security/keys/trusted.c +5 -0
@@ -1228,11 +1228,16 @@ hashalg_fail:
 
 static int __init init_digests(void)
 {
+	int i;
+
 	digests = kcalloc(chip->nr_allocated_banks, sizeof(*digests),
 			  GFP_KERNEL);
 	if (!digests)
 		return -ENOMEM;
 
+	for (i = 0; i < chip->nr_allocated_banks; i++)
+		digests[i].alg_id = chip->allocated_banks[i].alg_id;
+
 	return 0;
 }
 
author maintainer
#275 lore raw

PATCH 272/313 rtw88: pci: Rearrange the memory usage for skb in RX ISR

In reply to Greg Kroah-Hartman

From: Jian-Hong Pan <[email protected]>

commit ee6db78f5db9bfe426c57a1ec9713827ebccd2d4 upstream.

Testing with RTL8822BE hardware, when available memory is low, we frequently see a kernel panic and system freeze.

First, rtw_pci_rx_isr encounters a memory allocation failure (trimmed):

rx routine starvation

WARNING: CPU: 7 PID: 9871 at drivers/net/wireless/realtek/rtw88/pci.c:822 rtw_pci_rx_isr.constprop.25+0x35a/0x370 [rtwpci]
[ 2356.580313] RIP: 0010:rtw_pci_rx_isr.constprop.25+0x35a/0x370 [rtwpci]

Then we see a variety of different error conditions and kernel panics, such as this one (trimmed):

rtw_pci 0000:02:00.0: pci bus timeout, check dma status
skbuff: skb_over_panic: text:00000000091b6e66 len:415 put:415 head:00000000d2880c6f data:000000007a02b1ea tail:0x1df end:0xc0 dev:<NULL>

------------[ cut here ]------------
kernel BUG at net/core/skbuff.c:105!
invalid opcode: 0000 [#1] SMP NOPTI
RIP: 0010:skb_panic+0x43/0x45

When skb allocation fails and the "rx routine starvation" is hit, the function returns immediately without updating the RX ring. At this point, the RX ring may continue referencing an old skb which was already handed off to ieee80211_rx_irqsafe(). When it comes to be used again, bad things happen.

This patch allocates a new, data-sized skb first in RX ISR. After copying the data in, we pass it to the upper layers. However, if skb allocation fails, we effectively drop the frame. In both cases, the original, full size ring skb is reused.

In addition, to fixing the kernel crash, the RX routine should now generally behave better under low memory conditions.

---
 drivers/net/wireless/realtek/rtw88/pci.c |   49 +++++++++++++------------------
 1 file changed, 22 insertions(+), 27 deletions(-)
drivers/net/wireless/realtek/rtw88/pci.c +22 -27
@@ -763,6 +763,7 @@ static void rtw_pci_rx_isr(struct rtw_de
 	u32 pkt_offset;
 	u32 pkt_desc_sz = chip->rx_pkt_desc_sz;
 	u32 buf_desc_sz = chip->rx_buf_desc_sz;
+	u32 new_len;
 	u8 *rx_desc;
 	dma_addr_t dma;
 
@@ -790,40 +791,34 @@ static void rtw_pci_rx_isr(struct rtw_de
 		pkt_offset = pkt_desc_sz + pkt_stat.drv_info_sz +
 			     pkt_stat.shift;
 
-		if (pkt_stat.is_c2h) {
-			/* keep rx_desc, halmac needs it */
-			skb_put(skb, pkt_stat.pkt_len + pkt_offset);
+		/* allocate a new skb for this frame,
+		 * discard the frame if none available
+		 */
+		new_len = pkt_stat.pkt_len + pkt_offset;
+		new = dev_alloc_skb(new_len);
+		if (WARN_ONCE(!new, "rx routine starvation\n"))
+			goto next_rp;
+
+		/* put the DMA data including rx_desc from phy to new skb */
+		skb_put_data(new, skb->data, new_len);
 
-			/* pass offset for further operation */
-			*((u32 *)skb->cb) = pkt_offset;
-			skb_queue_tail(&rtwdev->c2h_queue, skb);
+		if (pkt_stat.is_c2h) {
+			 /* pass rx_desc & offset for further operation */
+			*((u32 *)new->cb) = pkt_offset;
+			skb_queue_tail(&rtwdev->c2h_queue, new);
 			ieee80211_queue_work(rtwdev->hw, &rtwdev->c2h_work);
 		} else {
-			/* remove rx_desc, maybe use skb_pull? */
-			skb_put(skb, pkt_stat.pkt_len);
-			skb_reserve(skb, pkt_offset);
-
-			/* alloc a smaller skb to mac80211 */
-			new = dev_alloc_skb(pkt_stat.pkt_len);
-			if (!new) {
-				new = skb;
-			} else {
-				skb_put_data(new, skb->data, skb->len);
-				dev_kfree_skb_any(skb);
-			}
-			/* TODO: merge into rx.c */
-			rtw_rx_stats(rtwdev, pkt_stat.vif, skb);
+			/* remove rx_desc */
+			skb_pull(new, pkt_offset);
+
+			rtw_rx_stats(rtwdev, pkt_stat.vif, new);
 			memcpy(new->cb, &rx_status, sizeof(rx_status));
 			ieee80211_rx_irqsafe(rtwdev->hw, new);
 		}
 
-		/* skb delivered to mac80211, alloc a new one in rx ring */
-		new = dev_alloc_skb(RTK_PCI_RX_BUF_SIZE);
-		if (WARN(!new, "rx routine starvation\n"))
-			return;
-
-		ring->buf[cur_rp] = new;
-		rtw_pci_reset_rx_desc(rtwdev, new, ring, cur_rp, buf_desc_sz);
+next_rp:
+		/* new skb delivered to mac80211, re-enable original skb DMA */
+		rtw_pci_reset_rx_desc(rtwdev, skb, ring, cur_rp, buf_desc_sz);
 
 		/* host read next element in ring */
 		if (++cur_rp >= ring->r.len)
author maintainer
#276 lore raw

PATCH 273/313 rtw88: pci: Use DMA sync instead of remapping in RX ISR

In reply to Greg Kroah-Hartman

From: Jian-Hong Pan <[email protected]>

commit 29b68a920f6abb7b5ba21ab4b779f62d536bac9b upstream.

Since each skb in RX ring is reused instead of new allocation, we can treat the DMA in a more efficient way by DMA synchronization.

---
 drivers/net/wireless/realtek/rtw88/pci.c |   24 +++++++++++++++++++++---
 1 file changed, 21 insertions(+), 3 deletions(-)
drivers/net/wireless/realtek/rtw88/pci.c +21 -3
@@ -206,6 +206,23 @@ static int rtw_pci_reset_rx_desc(struct
 	return 0;
 }
 
+static void rtw_pci_sync_rx_desc_device(struct rtw_dev *rtwdev, dma_addr_t dma,
+					struct rtw_pci_rx_ring *rx_ring,
+					u32 idx, u32 desc_sz)
+{
+	struct device *dev = rtwdev->dev;
+	struct rtw_pci_rx_buffer_desc *buf_desc;
+	int buf_sz = RTK_PCI_RX_BUF_SIZE;
+
+	dma_sync_single_for_device(dev, dma, buf_sz, DMA_FROM_DEVICE);
+
+	buf_desc = (struct rtw_pci_rx_buffer_desc *)(rx_ring->r.head +
+						     idx * desc_sz);
+	memset(buf_desc, 0, sizeof(*buf_desc));
+	buf_desc->buf_size = cpu_to_le16(RTK_PCI_RX_BUF_SIZE);
+	buf_desc->dma = cpu_to_le32(dma);
+}
+
 static int rtw_pci_init_rx_ring(struct rtw_dev *rtwdev,
 				struct rtw_pci_rx_ring *rx_ring,
 				u8 desc_size, u32 len)
@@ -782,8 +799,8 @@ static void rtw_pci_rx_isr(struct rtw_de
 		rtw_pci_dma_check(rtwdev, ring, cur_rp);
 		skb = ring->buf[cur_rp];
 		dma = *((dma_addr_t *)skb->cb);
-		pci_unmap_single(rtwpci->pdev, dma, RTK_PCI_RX_BUF_SIZE,
-				 PCI_DMA_FROMDEVICE);
+		dma_sync_single_for_cpu(rtwdev->dev, dma, RTK_PCI_RX_BUF_SIZE,
+					DMA_FROM_DEVICE);
 		rx_desc = skb->data;
 		chip->ops->query_rx_desc(rtwdev, rx_desc, &pkt_stat, &rx_status);
 
@@ -818,7 +835,8 @@ static void rtw_pci_rx_isr(struct rtw_de
 
 next_rp:
 		/* new skb delivered to mac80211, re-enable original skb DMA */
-		rtw_pci_reset_rx_desc(rtwdev, skb, ring, cur_rp, buf_desc_sz);
+		rtw_pci_sync_rx_desc_device(rtwdev, dma, ring, cur_rp,
+					    buf_desc_sz);
 
 		/* host read next element in ring */
 		if (++cur_rp >= ring->r.len)
author maintainer
#277 lore raw

PATCH 274/313 ath10k: fix channel info parsing for non tlv target

In reply to Greg Kroah-Hartman

From: Rakesh Pillai <[email protected]>

commit 6be6c04bcc2e8770b8637632789ff15765124894 upstream.

The tlv targets such as WCN3990 send more data in the chan info event, which is not sent by the non tlv targets. There is a minimum size check in the wmi event for non-tlv targets and hence we cannot update the common channel info structure as it was done in commit 13104929d2ec ("ath10k: fill the channel survey results for WCN3990 correctly"). This broke channel survey results on 10.x firmware versions.

If the common channel info structure is updated, the size check for chan info event for non-tlv targets will fail and return -EPROTO and we see the below error messages

ath10k_pci 0000:01:00.0: failed to parse chan info event: -71

Add tlv specific channel info structure and restore the original size of the common channel info structure to mitigate this issue.

Tested HW: WCN3990

	   QCA9887
Tested FW: WLAN.HL.3.1-00784-QCAHLSWMTPLZ-1
	   10.2.4-1.0-00037
Fixes 13104929d2ec ("ath10k: fill the channel survey results for WCN3990 correctly")
---
 drivers/net/wireless/ath/ath10k/wmi-tlv.c |    2 +-
 drivers/net/wireless/ath/ath10k/wmi-tlv.h |   16 ++++++++++++++++
 drivers/net/wireless/ath/ath10k/wmi.h     |    8 --------
 3 files changed, 17 insertions(+), 9 deletions(-)
drivers/net/wireless/ath/ath10k/wmi-tlv.c +1 -1
@@ -810,7 +810,7 @@ static int ath10k_wmi_tlv_op_pull_ch_inf
 					     struct wmi_ch_info_ev_arg *arg)
 {
 	const void **tb;
-	const struct wmi_chan_info_event *ev;
+	const struct wmi_tlv_chan_info_event *ev;
 	int ret;
 
 	tb = ath10k_wmi_tlv_parse_alloc(ar, skb->data, skb->len, GFP_ATOMIC);
drivers/net/wireless/ath/ath10k/wmi-tlv.h +16 -0
@@ -1607,6 +1607,22 @@ struct chan_info_params {
 
 #define WMI_TLV_FLAG_MGMT_BUNDLE_TX_COMPL	BIT(9)
 
+struct wmi_tlv_chan_info_event {
+	__le32 err_code;
+	__le32 freq;
+	__le32 cmd_flags;
+	__le32 noise_floor;
+	__le32 rx_clear_count;
+	__le32 cycle_count;
+	__le32 chan_tx_pwr_range;
+	__le32 chan_tx_pwr_tp;
+	__le32 rx_frame_count;
+	__le32 my_bss_rx_cycle_count;
+	__le32 rx_11b_mode_data_duration;
+	__le32 tx_frame_cnt;
+	__le32 mac_clk_mhz;
+} __packed;
+
 struct wmi_tlv_mgmt_tx_compl_ev {
 	__le32 desc_id;
 	__le32 status;
drivers/net/wireless/ath/ath10k/wmi.h +0 -8
@@ -6524,14 +6524,6 @@ struct wmi_chan_info_event {
 	__le32 noise_floor;
 	__le32 rx_clear_count;
 	__le32 cycle_count;
-	__le32 chan_tx_pwr_range;
-	__le32 chan_tx_pwr_tp;
-	__le32 rx_frame_count;
-	__le32 my_bss_rx_cycle_count;
-	__le32 rx_11b_mode_data_duration;
-	__le32 tx_frame_cnt;
-	__le32 mac_clk_mhz;
-
 } __packed;
 
 struct wmi_10_4_chan_info_event {
author maintainer
#278 lore raw

PATCH 275/313 i40e: check __I40E_VF_DISABLE bit in i40e_sync_filters_subtask

In reply to Greg Kroah-Hartman

From: Stefan Assmann <[email protected]>

commit a7542b87607560d0b89e7ff81d870bd6ff8835cb upstream.

While testing VF spawn/destroy the following panic occurred.

BUG: unable to handle kernel NULL pointer dereference at 0000000000000029
[...]
Workqueue: i40e i40e_service_task [i40e]
RIP: 0010:i40e_sync_vsi_filters+0x6fd/0xc60 [i40e]
[...]
Call Trace:
 ? __switch_to_asm+0x35/0x70
 ? __switch_to_asm+0x41/0x70
 ? __switch_to_asm+0x35/0x70
 ? _cond_resched+0x15/0x30
 i40e_sync_filters_subtask+0x56/0x70 [i40e]
 i40e_service_task+0x382/0x11b0 [i40e]
 ? __switch_to_asm+0x41/0x70
 ? __switch_to_asm+0x41/0x70
 process_one_work+0x1a7/0x3b0
 worker_thread+0x30/0x390
 ? create_worker+0x1a0/0x1a0
 kthread+0x112/0x130
 ? kthread_bind+0x30/0x30
 ret_from_fork+0x35/0x40

Investigation revealed a race where pf->vf[vsi->vf_id].trusted may get accessed by the watchdog via i40e_sync_filters_subtask() although i40e_free_vfs() already free'd pf->vf.
To avoid this the call to i40e_sync_vsi_filters() in i40e_sync_filters_subtask() needs to be guarded by __I40E_VF_DISABLE, which is also used by i40e_free_vfs().

Note: put the __I40E_VF_DISABLE check after the __I40E_MACVLAN_SYNC_PENDING check as the latter is more likely to trigger.

---
 drivers/net/ethernet/intel/i40e/i40e_main.c |    5 +++++
 1 file changed, 5 insertions(+)
drivers/net/ethernet/intel/i40e/i40e_main.c +5 -0
@@ -2586,6 +2586,10 @@ static void i40e_sync_filters_subtask(st
 		return;
 	if (!test_and_clear_bit(__I40E_MACVLAN_SYNC_PENDING, pf->state))
 		return;
+	if (test_and_set_bit(__I40E_VF_DISABLE, pf->state)) {
+		set_bit(__I40E_MACVLAN_SYNC_PENDING, pf->state);
+		return;
+	}
 
 	for (v = 0; v < pf->num_alloc_vsi; v++) {
 		if (pf->vsi[v] &&
@@ -2600,6 +2604,7 @@ static void i40e_sync_filters_subtask(st
 			}
 		}
 	}
+	clear_bit(__I40E_VF_DISABLE, pf->state);
 }
 
 /**
author maintainer
#279 lore raw

PATCH 276/313 block: mq-deadline: Fix queue restart handling

In reply to Greg Kroah-Hartman

From: Damien Le Moal <[email protected]>

commit cb8acabbe33b110157955a7425ee876fb81e6bbc upstream.

Commit 7211aef86f79 ("block: mq-deadline: Fix write completion handling") added a call to blk_mq_sched_mark_restart_hctx() in dd_dispatch_request() to make sure that write request dispatching does not stall when all target zones are locked. This fix left a subtle race when a write completion happens during a dispatch execution on another CPU:

CPU 0: Dispatch CPU1: write completion

dd_dispatch_request()

    lock(&dd->lock);
    ...
    lock(&dd->zone_lock);	dd_finish_request()
    rq = find request		lock(&dd->zone_lock);
    unlock(&dd->zone_lock);
    				zone write unlock
				unlock(&dd->zone_lock);
				...
				__blk_mq_free_request
                                      check restart flag (not set)
				      -> queue not run
    ...
    if (!rq && have writes)
        blk_mq_sched_mark_restart_hctx()
    unlock(&dd->lock)

Since the dispatch context finishes after the write request completion handling, marking the queue as needing a restart is not seen from __blk_mq_free_request() and blk_mq_sched_restart() not executed leading to the dispatch stall under 100% write workloads.

Fix this by moving the call to blk_mq_sched_mark_restart_hctx() from dd_dispatch_request() into dd_finish_request() under the zone lock to ensure full mutual exclusion between write request dispatch selection and zone unlock on write request completion.

Fixes 7211aef86f79 ("block: mq-deadline: Fix write completion handling")
---
 block/mq-deadline.c |   19 +++++++++----------
 1 file changed, 9 insertions(+), 10 deletions(-)
block/mq-deadline.c +9 -10
@@ -377,13 +377,6 @@ done:
  * hardware queue, but we may return a request that is for a
  * different hardware queue. This is because mq-deadline has shared
  * state for all hardware queues, in terms of sorting, FIFOs, etc.
- *
- * For a zoned block device, __dd_dispatch_request() may return NULL
- * if all the queued write requests are directed at zones that are already
- * locked due to on-going write requests. In this case, make sure to mark
- * the queue as needing a restart to ensure that the queue is run again
- * and the pending writes dispatched once the target zones for the ongoing
- * write requests are unlocked in dd_finish_request().
  */
 static struct request *dd_dispatch_request(struct blk_mq_hw_ctx *hctx)
 {
@@ -392,9 +385,6 @@ static struct request *dd_dispatch_reque
 
 	spin_lock(&dd->lock);
 	rq = __dd_dispatch_request(dd);
-	if (!rq && blk_queue_is_zoned(hctx->queue) &&
-	    !list_empty(&dd->fifo_list[WRITE]))
-		blk_mq_sched_mark_restart_hctx(hctx);
 	spin_unlock(&dd->lock);
 
 	return rq;
@@ -560,6 +550,13 @@ static void dd_prepare_request(struct re
  * spinlock so that the zone is never unlocked while deadline_fifo_request()
  * or deadline_next_request() are executing. This function is called for
  * all requests, whether or not these requests complete successfully.
+ *
+ * For a zoned block device, __dd_dispatch_request() may have stopped
+ * dispatching requests if all the queued requests are write requests directed
+ * at zones that are already locked due to on-going write requests. To ensure
+ * write request dispatch progress in this case, mark the queue as needing a
+ * restart to ensure that the queue is run again after completion of the
+ * request and zones being unlocked.
  */
 static void dd_finish_request(struct request *rq)
 {
@@ -571,6 +568,8 @@ static void dd_finish_request(struct req
 
 		spin_lock_irqsave(&dd->zone_lock, flags);
 		blk_req_zone_write_unlock(rq);
+		if (!list_empty(&dd->fifo_list[WRITE]))
+			blk_mq_sched_mark_restart_hctx(rq->mq_hctx);
 		spin_unlock_irqrestore(&dd->zone_lock, flags);
 	}
 }
author maintainer
#280 lore raw

PATCH 277/313 block: fix null pointer dereference in blk_mq_rq_timed_out()

In reply to Greg Kroah-Hartman

From: Yufen Yu <[email protected]>

commit 8d6996630c03d7ceeabe2611378fea5ca1c3f1b3 upstream.

We got a null pointer deference BUG_ON in blk_mq_rq_timed_out() as following:

[  108.825472] BUG: kernel NULL pointer dereference, address: 0000000000000040
[  108.827059] PGD 0 P4D 0
[  108.827313] Oops: 0000 [#1] SMP PTI
[  108.827657] CPU: 6 PID: 198 Comm: kworker/6:1H Not tainted 5.3.0-rc8+ #431
[  108.829503] Workqueue: kblockd blk_mq_timeout_work
[  108.829913] RIP: 0010:blk_mq_check_expired+0x258/0x330
[  108.838191] Call Trace:
[  108.838406]  bt_iter+0x74/0x80
[  108.838665]  blk_mq_queue_tag_busy_iter+0x204/0x450
[  108.839074]  ? __switch_to_asm+0x34/0x70
[  108.839405]  ? blk_mq_stop_hw_queue+0x40/0x40
[  108.839823]  ? blk_mq_stop_hw_queue+0x40/0x40
[  108.840273]  ? syscall_return_via_sysret+0xf/0x7f
[  108.840732]  blk_mq_timeout_work+0x74/0x200
[  108.841151]  process_one_work+0x297/0x680
[  108.841550]  worker_thread+0x29c/0x6f0
[  108.841926]  ? rescuer_thread+0x580/0x580
[  108.842344]  kthread+0x16a/0x1a0
[  108.842666]  ? kthread_flush_work+0x170/0x170
[  108.843100]  ret_from_fork+0x35/0x40

The bug is caused by the race between timeout handle and completion for flush request.

When timeout handle function blk_mq_rq_timed_out() try to read 'req->q->mq_ops', the 'req' have completed and reinitiated by next flush request, which would call blk_rq_init() to clear 'req' as 0.

After commit 12f5b93145 ("blk-mq: Remove generation seqeunce"), normal requests lifetime are protected by refcount. Until 'rq->ref' drop to zero, the request can really be free. Thus, these requests cannot been reused before timeout handle finish.

However, flush request has defined .end_io and rq->end_io() is still called even if 'rq->ref' doesn't drop to zero. After that, the 'flush_rq' can be reused by the next flush request handle, resulting in null pointer deference BUG ON.

We fix this problem by covering flush request with 'rq->ref'. If the refcount is not zero, flush_end_io() return and wait the last holder recall it. To record the request status, we add a new entry 'rq_status', which will be used in flush_end_io().

-------
v2:
- move rq_status from struct request to struct blk_flush_queue
v3:
- remove unnecessary '{}' pair.
v4:
- let spinlock to protect 'fq->rq_status'
v5:
- move rq_status after flush_running_idx member of struct blk_flush_queue

---
 block/blk-flush.c |   10 ++++++++++
 block/blk-mq.c    |    5 ++++-
 block/blk.h       |    7 +++++++
 3 files changed, 21 insertions(+), 1 deletion(-)
block/blk-flush.c +10 -0
@@ -214,6 +214,16 @@ static void flush_end_io(struct request
 
 	/* release the tag's ownership to the req cloned from */
 	spin_lock_irqsave(&fq->mq_flush_lock, flags);
+
+	if (!refcount_dec_and_test(&flush_rq->ref)) {
+		fq->rq_status = error;
+		spin_unlock_irqrestore(&fq->mq_flush_lock, flags);
+		return;
+	}
+
+	if (fq->rq_status != BLK_STS_OK)
+		error = fq->rq_status;
+
 	hctx = flush_rq->mq_hctx;
 	if (!q->elevator) {
 		blk_mq_tag_set_rq(hctx, flush_rq->tag, fq->orig_rq);
block/blk-mq.c +4 -1
@@ -910,7 +910,10 @@ static bool blk_mq_check_expired(struct
 	 */
 	if (blk_mq_req_expired(rq, next))
 		blk_mq_rq_timed_out(rq, reserved);
-	if (refcount_dec_and_test(&rq->ref))
+
+	if (is_flush_rq(rq, hctx))
+		rq->end_io(rq, 0);
+	else if (refcount_dec_and_test(&rq->ref))
 		__blk_mq_free_request(rq);
 
 	return true;
block/blk.h +7 -0
@@ -19,6 +19,7 @@ struct blk_flush_queue {
 	unsigned int		flush_queue_delayed:1;
 	unsigned int		flush_pending_idx:1;
 	unsigned int		flush_running_idx:1;
+	blk_status_t 		rq_status;
 	unsigned long		flush_pending_since;
 	struct list_head	flush_queue[2];
 	struct list_head	flush_data_in_flight;
@@ -47,6 +48,12 @@ static inline void __blk_get_queue(struc
 	kobject_get(&q->kobj);
 }
 
+static inline bool
+is_flush_rq(struct request *req, struct blk_mq_hw_ctx *hctx)
+{
+	return hctx->fq->flush_rq == req;
+}
+
 struct blk_flush_queue *blk_alloc_flush_queue(struct request_queue *q,
 		int node, int cmd_size, gfp_t flags);
 void blk_free_flush_queue(struct blk_flush_queue *q);
author maintainer
#281 lore raw

PATCH 278/313 smb3: allow disabling requesting leases

In reply to Greg Kroah-Hartman

From: Steve French <[email protected]>

commit 3e7a02d47872081f4b6234a9f72500f1d10f060c upstream.

In some cases to work around server bugs or performance problems it can be helpful to be able to disable requesting SMB2.1/SMB3 leases on a particular mount (not to all servers and all shares we are mounted to). Add new mount parm "nolease" which turns off requesting leases on directory or file opens. Currently the only way to disable leases is globally through a module load parameter. This is more granular.

---
 fs/cifs/cifsfs.c   |    2 ++
 fs/cifs/cifsglob.h |    2 ++
 fs/cifs/connect.c  |    9 ++++++++-
 fs/cifs/smb2pdu.c  |    2 +-
 4 files changed, 13 insertions(+), 2 deletions(-)
fs/cifs/cifsfs.c +2 -0
@@ -433,6 +433,8 @@ cifs_show_options(struct seq_file *s, st
 	cifs_show_security(s, tcon->ses);
 	cifs_show_cache_flavor(s, cifs_sb);
 
+	if (tcon->no_lease)
+		seq_puts(s, ",nolease");
 	if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MULTIUSER)
 		seq_puts(s, ",multiuser");
 	else if (tcon->ses->user_name)
fs/cifs/cifsglob.h +2 -0
@@ -575,6 +575,7 @@ struct smb_vol {
 	bool noblocksnd:1;
 	bool noautotune:1;
 	bool nostrictsync:1; /* do not force expensive SMBflush on every sync */
+	bool no_lease:1;     /* disable requesting leases */
 	bool fsc:1;	/* enable fscache */
 	bool mfsymlinks:1; /* use Minshall+French Symlinks */
 	bool multiuser:1;
@@ -1079,6 +1080,7 @@ struct cifs_tcon {
 	bool need_reopen_files:1; /* need to reopen tcon file handles */
 	bool use_resilient:1; /* use resilient instead of durable handles */
 	bool use_persistent:1; /* use persistent instead of durable handles */
+	bool no_lease:1;    /* Do not request leases on files or directories */
 	__le32 capabilities;
 	__u32 share_flags;
 	__u32 maximal_access;
fs/cifs/connect.c +8 -1
@@ -74,7 +74,7 @@ enum {
 	Opt_user_xattr, Opt_nouser_xattr,
 	Opt_forceuid, Opt_noforceuid,
 	Opt_forcegid, Opt_noforcegid,
-	Opt_noblocksend, Opt_noautotune,
+	Opt_noblocksend, Opt_noautotune, Opt_nolease,
 	Opt_hard, Opt_soft, Opt_perm, Opt_noperm,
 	Opt_mapposix, Opt_nomapposix,
 	Opt_mapchars, Opt_nomapchars, Opt_sfu,
@@ -133,6 +133,7 @@ static const match_table_t cifs_mount_op
 	{ Opt_noforcegid, "noforcegid" },
 	{ Opt_noblocksend, "noblocksend" },
 	{ Opt_noautotune, "noautotune" },
+	{ Opt_nolease, "nolease" },
 	{ Opt_hard, "hard" },
 	{ Opt_soft, "soft" },
 	{ Opt_perm, "perm" },
@@ -1709,6 +1710,9 @@ cifs_parse_mount_options(const char *mou
 		case Opt_noautotune:
 			vol->noautotune = 1;
 			break;
+		case Opt_nolease:
+			vol->no_lease = 1;
+			break;
 		case Opt_hard:
 			vol->retry = 1;
 			break;
@@ -3230,6 +3234,8 @@ static int match_tcon(struct cifs_tcon *
 		return 0;
 	if (tcon->handle_timeout != volume_info->handle_timeout)
 		return 0;
+	if (tcon->no_lease != volume_info->no_lease)
+		return 0;
 	return 1;
 }
 
@@ -3444,6 +3450,7 @@ cifs_get_tcon(struct cifs_ses *ses, stru
 	tcon->nocase = volume_info->nocase;
 	tcon->nohandlecache = volume_info->nohandlecache;
 	tcon->local_lease = volume_info->local_lease;
+	tcon->no_lease = volume_info->no_lease;
 	INIT_LIST_HEAD(&tcon->pending_opens);
 
 	spin_lock(&cifs_tcp_ses_lock);
fs/cifs/smb2pdu.c +1 -1
@@ -2370,7 +2370,7 @@ SMB2_open_init(struct cifs_tcon *tcon, s
 	iov[1].iov_len = uni_path_len;
 	iov[1].iov_base = path;
 
-	if (!server->oplocks)
+	if ((!server->oplocks) || (tcon->no_lease))
 		*oplock = SMB2_OPLOCK_LEVEL_NONE;
 
 	if (!(server->capabilities & SMB2_GLOBAL_CAP_LEASING) ||
author maintainer
#282 lore raw

PATCH 279/313 smb3: fix leak in "open on server" perf counter

In reply to Greg Kroah-Hartman

From: Steve French <[email protected]>

commit d2f15428d6a0ebfc0edc364094d7c4a2de7037ed upstream.

We were not bumping up the "open on server" (num_remote_opens) counter (in some cases) on opens of the share root so could end up showing as a negative value.

---
 fs/cifs/smb2ops.c |    5 +++++
 fs/cifs/smb2pdu.c |    1 +
 2 files changed, 6 insertions(+)
fs/cifs/smb2ops.c +5 -0
@@ -743,6 +743,8 @@ int open_shroot(unsigned int xid, struct
 	if (rc)
 		goto oshr_exit;
 
+	atomic_inc(&tcon->num_remote_opens);
+
 	o_rsp = (struct smb2_create_rsp *)rsp_iov[0].iov_base;
 	oparms.fid->persistent_fid = o_rsp->PersistentFileId;
 	oparms.fid->volatile_fid = o_rsp->VolatileFileId;
@@ -1167,6 +1169,7 @@ smb2_set_ea(const unsigned int xid, stru
 
 	rc = compound_send_recv(xid, ses, flags, 3, rqst,
 				resp_buftype, rsp_iov);
+	/* no need to bump num_remote_opens because handle immediately closed */
 
  sea_exit:
 	kfree(ea);
@@ -1488,6 +1491,8 @@ smb2_ioctl_query_info(const unsigned int
 				resp_buftype, rsp_iov);
 	if (rc)
 		goto iqinf_exit;
+
+	/* No need to bump num_remote_opens since handle immediately closed */
 	if (qi.flags & PASSTHRU_FSCTL) {
 		pqi = (struct smb_query_info __user *)arg;
 		io_rsp = (struct smb2_ioctl_rsp *)rsp_iov[1].iov_base;
fs/cifs/smb2pdu.c +1 -0
@@ -2263,6 +2263,7 @@ int smb311_posix_mkdir(const unsigned in
 	rqst.rq_iov = iov;
 	rqst.rq_nvec = n_iov;
 
+	/* no need to inc num_remote_opens because we close it just below */
 	trace_smb3_posix_mkdir_enter(xid, tcon->tid, ses->Suid, CREATE_NOT_FILE,
 				    FILE_WRITE_ATTRIBUTES);
 	/* resource #4: response buffer */
author maintainer
#283 lore raw

PATCH 280/313 ovl: Fix dereferencing possible ERR_PTR()

In reply to Greg Kroah-Hartman

From: Ding Xiang <[email protected]>

commit 97f024b9171e74c4443bbe8a8dce31b917f97ac5 upstream.

if ovl_encode_real_fh() fails, no memory was allocated and the error in the error-valued pointer should be returned.

Fixes 9b6faee07470 ("ovl: check ERR_PTR() return value from ovl_encode_fh()")
Cc <[email protected]> # v4.16+
---
 fs/overlayfs/export.c |    3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
fs/overlayfs/export.c +1 -2
@@ -227,9 +227,8 @@ static int ovl_d_to_fh(struct dentry *de
 	/* Encode an upper or lower file handle */
 	fh = ovl_encode_real_fh(enc_lower ? ovl_dentry_lower(dentry) :
 				ovl_dentry_upper(dentry), !enc_lower);
-	err = PTR_ERR(fh);
 	if (IS_ERR(fh))
-		goto fail;
+		return PTR_ERR(fh);
 
 	err = -EOVERFLOW;
 	if (fh->len > buflen)
author maintainer
#284 lore raw

PATCH 281/313 ovl: filter of trusted xattr results in audit

In reply to Greg Kroah-Hartman

From: Mark Salyzyn <[email protected]>

commit 5c2e9f346b815841f9bed6029ebcb06415caf640 upstream.

When filtering xattr list for reading, presence of trusted xattr results in a security audit log. However, if there is other content no errno will be set, and if there isn't, the errno will be -ENODATA and not -EPERM as is usually associated with a lack of capability. The check does not block the request to list the xattrs present.

Switch to ns_capable_noaudit to reflect a more appropriate check.

Cc [email protected] # v3.18+
Fixes a082c6f680da ("ovl: filter trusted xattr for non-admin")
---
 fs/overlayfs/inode.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
fs/overlayfs/inode.c +2 -1
@@ -383,7 +383,8 @@ static bool ovl_can_list(const char *s)
 		return true;
 
 	/* Never list trusted.overlay, list other trusted for superuser only */
-	return !ovl_is_private_xattr(s) && capable(CAP_SYS_ADMIN);
+	return !ovl_is_private_xattr(s) &&
+	       ns_capable_noaudit(&init_user_ns, CAP_SYS_ADMIN);
 }
 
 ssize_t ovl_listxattr(struct dentry *dentry, char *list, size_t size)
author maintainer
#285 lore raw

PATCH 282/313 btrfs: fix allocation of free space cache v1 bitmap pages

In reply to Greg Kroah-Hartman

From: Christophe Leroy <[email protected]>

commit 3acd48507dc43eeeb0a1fe965b8bad91cab904a7 upstream.

Various notifications of type "BUG kmalloc-4096 () : Redzone overwritten" have been observed recently in various parts of the kernel. After some time, it has been made a relation with the use of BTRFS filesystem and with SLUB_DEBUG turned on.

[   22.809700] BUG kmalloc-4096 (Tainted: G        W        ): Redzone overwritten
[   22.810286] INFO: 0xbe1a5921-0xfbfc06cd. First byte 0x0 instead of 0xcc
[   22.810866] INFO: Allocated in __load_free_space_cache+0x588/0x780 [btrfs] age=22 cpu=0 pid=224
[   22.811193] 	__slab_alloc.constprop.26+0x44/0x70
[   22.811345] 	kmem_cache_alloc_trace+0xf0/0x2ec
[   22.811588] 	__load_free_space_cache+0x588/0x780 [btrfs]
[   22.811848] 	load_free_space_cache+0xf4/0x1b0 [btrfs]
[   22.812090] 	cache_block_group+0x1d0/0x3d0 [btrfs]
[   22.812321] 	find_free_extent+0x680/0x12a4 [btrfs]
[   22.812549] 	btrfs_reserve_extent+0xec/0x220 [btrfs]
[   22.812785] 	btrfs_alloc_tree_block+0x178/0x5f4 [btrfs]
[   22.813032] 	__btrfs_cow_block+0x150/0x5d4 [btrfs]
[   22.813262] 	btrfs_cow_block+0x194/0x298 [btrfs]
[   22.813484] 	commit_cowonly_roots+0x44/0x294 [btrfs]
[   22.813718] 	btrfs_commit_transaction+0x63c/0xc0c [btrfs]
[   22.813973] 	close_ctree+0xf8/0x2a4 [btrfs]
[   22.814107] 	generic_shutdown_super+0x80/0x110
[   22.814250] 	kill_anon_super+0x18/0x30
[   22.814437] 	btrfs_kill_super+0x18/0x90 [btrfs]
[   22.814590] INFO: Freed in proc_cgroup_show+0xc0/0x248 age=41 cpu=0 pid=83
[   22.814841] 	proc_cgroup_show+0xc0/0x248
[   22.814967] 	proc_single_show+0x54/0x98
[   22.815086] 	seq_read+0x278/0x45c
[   22.815190] 	__vfs_read+0x28/0x17c
[   22.815289] 	vfs_read+0xa8/0x14c
[   22.815381] 	ksys_read+0x50/0x94
[   22.815475] 	ret_from_syscall+0x0/0x38

Commit 69d2480456d1 ("btrfs: use copy_page for copying pages instead of memcpy") changed the way bitmap blocks are copied. But allthough bitmaps have the size of a page, they were allocated with kzalloc().

Most of the time, kzalloc() allocates aligned blocks of memory, so copy_page() can be used. But when some debug options like SLAB_DEBUG are activated, kzalloc() may return unaligned pointer.

On powerpc, memcpy(), copy_page() and other copying functions use 'dcbz' instruction which provides an entire zeroed cacheline to avoid memory read when the intention is to overwrite a full line. Functions like memcpy() are writen to care about partial cachelines at the start and end of the destination, but copy_page() assumes it gets pages. As pages are naturally cache aligned, copy_page() doesn't care about partial lines. This means that when copy_page() is called with a misaligned pointer, a few leading bytes are zeroed.

To fix it, allocate bitmaps through kmem_cache instead of using kzalloc() The cache pool is created with PAGE_SIZE alignment constraint.

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=204371

Fixes 69d2480456d1 ("btrfs: use copy_page for copying pages instead of memcpy")
[ rename to btrfs_free_space_bitmap ]
---
 fs/btrfs/ctree.h            |    1 +
 fs/btrfs/free-space-cache.c |   20 +++++++++++++-------
 fs/btrfs/inode.c            |    8 ++++++++
 3 files changed, 22 insertions(+), 7 deletions(-)
fs/btrfs/ctree.h +1 -0
@@ -40,6 +40,7 @@ extern struct kmem_cache *btrfs_trans_ha
 extern struct kmem_cache *btrfs_bit_radix_cachep;
 extern struct kmem_cache *btrfs_path_cachep;
 extern struct kmem_cache *btrfs_free_space_cachep;
+extern struct kmem_cache *btrfs_free_space_bitmap_cachep;
 struct btrfs_ordered_sum;
 struct btrfs_ref;
 
fs/btrfs/free-space-cache.c +13 -7
@@ -764,7 +764,8 @@ static int __load_free_space_cache(struc
 		} else {
 			ASSERT(num_bitmaps);
 			num_bitmaps--;
-			e->bitmap = kzalloc(PAGE_SIZE, GFP_NOFS);
+			e->bitmap = kmem_cache_zalloc(
+					btrfs_free_space_bitmap_cachep, GFP_NOFS);
 			if (!e->bitmap) {
 				kmem_cache_free(
 					btrfs_free_space_cachep, e);
@@ -1881,7 +1882,7 @@ static void free_bitmap(struct btrfs_fre
 			struct btrfs_free_space *bitmap_info)
 {
 	unlink_free_space(ctl, bitmap_info);
-	kfree(bitmap_info->bitmap);
+	kmem_cache_free(btrfs_free_space_bitmap_cachep, bitmap_info->bitmap);
 	kmem_cache_free(btrfs_free_space_cachep, bitmap_info);
 	ctl->total_bitmaps--;
 	ctl->op->recalc_thresholds(ctl);
@@ -2135,7 +2136,8 @@ new_bitmap:
 		}
 
 		/* allocate the bitmap */
-		info->bitmap = kzalloc(PAGE_SIZE, GFP_NOFS);
+		info->bitmap = kmem_cache_zalloc(btrfs_free_space_bitmap_cachep,
+						 GFP_NOFS);
 		spin_lock(&ctl->tree_lock);
 		if (!info->bitmap) {
 			ret = -ENOMEM;
@@ -2146,7 +2148,9 @@ new_bitmap:
 
 out:
 	if (info) {
-		kfree(info->bitmap);
+		if (info->bitmap)
+			kmem_cache_free(btrfs_free_space_bitmap_cachep,
+					info->bitmap);
 		kmem_cache_free(btrfs_free_space_cachep, info);
 	}
 
@@ -2802,7 +2806,8 @@ out:
 	if (entry->bytes == 0) {
 		ctl->free_extents--;
 		if (entry->bitmap) {
-			kfree(entry->bitmap);
+			kmem_cache_free(btrfs_free_space_bitmap_cachep,
+					entry->bitmap);
 			ctl->total_bitmaps--;
 			ctl->op->recalc_thresholds(ctl);
 		}
@@ -3606,7 +3611,7 @@ again:
 	}
 
 	if (!map) {
-		map = kzalloc(PAGE_SIZE, GFP_NOFS);
+		map = kmem_cache_zalloc(btrfs_free_space_bitmap_cachep, GFP_NOFS);
 		if (!map) {
 			kmem_cache_free(btrfs_free_space_cachep, info);
 			return -ENOMEM;
@@ -3635,7 +3640,8 @@ again:
 
 	if (info)
 		kmem_cache_free(btrfs_free_space_cachep, info);
-	kfree(map);
+	if (map)
+		kmem_cache_free(btrfs_free_space_bitmap_cachep, map);
 	return 0;
 }
 
fs/btrfs/inode.c +8 -0
@@ -73,6 +73,7 @@ static struct kmem_cache *btrfs_inode_ca
 struct kmem_cache *btrfs_trans_handle_cachep;
 struct kmem_cache *btrfs_path_cachep;
 struct kmem_cache *btrfs_free_space_cachep;
+struct kmem_cache *btrfs_free_space_bitmap_cachep;
 
 static int btrfs_setsize(struct inode *inode, struct iattr *attr);
 static int btrfs_truncate(struct inode *inode, bool skip_writeback);
@@ -9361,6 +9362,7 @@ void __cold btrfs_destroy_cachep(void)
 	kmem_cache_destroy(btrfs_trans_handle_cachep);
 	kmem_cache_destroy(btrfs_path_cachep);
 	kmem_cache_destroy(btrfs_free_space_cachep);
+	kmem_cache_destroy(btrfs_free_space_bitmap_cachep);
 }
 
 int __init btrfs_init_cachep(void)
@@ -9390,6 +9392,12 @@ int __init btrfs_init_cachep(void)
 	if (!btrfs_free_space_cachep)
 		goto fail;
 
+	btrfs_free_space_bitmap_cachep = kmem_cache_create("btrfs_free_space_bitmap",
+							PAGE_SIZE, PAGE_SIZE,
+							SLAB_RED_ZONE, NULL);
+	if (!btrfs_free_space_bitmap_cachep)
+		goto fail;
+
 	return 0;
 fail:
 	btrfs_destroy_cachep();
author maintainer
#286 lore raw

PATCH 283/313 Btrfs: fix use-after-free when using the tree modification log

In reply to Greg Kroah-Hartman

From: Filipe Manana <[email protected]>

commit efad8a853ad2057f96664328a0d327a05ce39c76 upstream.

At ctree.c:get_old_root(), we are accessing a root's header owner field after we have freed the respective extent buffer. This results in an use-after-free that can lead to crashes, and when CONFIG_DEBUG_PAGEALLOC is set, results in a stack trace like the following:

  [ 3876.799331] stack segment: 0000 [#1] SMP DEBUG_PAGEALLOC PTI
  [ 3876.799363] CPU: 0 PID: 15436 Comm: pool Not tainted 5.3.0-rc3-btrfs-next-54 #1
  [ 3876.799385] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.12.0-0-ga698c8995f-prebuilt.qemu.org 04/01/2014
  [ 3876.799433] RIP: 0010:btrfs_search_old_slot+0x652/0xd80 [btrfs]
  (...)
  [ 3876.799502] RSP: 0018:ffff9f08c1a2f9f0 EFLAGS: 00010286
  [ 3876.799518] RAX: ffff8dd300000000 RBX: ffff8dd85a7a9348 RCX: 000000038da26000
  [ 3876.799538] RDX: 0000000000000000 RSI: ffffe522ce368980 RDI: 0000000000000246
  [ 3876.799559] RBP: dae1922adadad000 R08: 0000000008020000 R09: ffffe522c0000000
  [ 3876.799579] R10: ffff8dd57fd788c8 R11: 000000007511b030 R12: ffff8dd781ddc000
  [ 3876.799599] R13: ffff8dd9e6240578 R14: ffff8dd6896f7a88 R15: ffff8dd688cf90b8
  [ 3876.799620] FS:  00007f23ddd97700(0000) GS:ffff8dda20200000(0000) knlGS:0000000000000000
  [ 3876.799643] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
  [ 3876.799660] CR2: 00007f23d4024000 CR3: 0000000710bb0005 CR4: 00000000003606f0
  [ 3876.799682] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
  [ 3876.799703] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
  [ 3876.799723] Call Trace:
  [ 3876.799735]  ? do_raw_spin_unlock+0x49/0xc0
  [ 3876.799749]  ? _raw_spin_unlock+0x24/0x30
  [ 3876.799779]  resolve_indirect_refs+0x1eb/0xc80 [btrfs]
  [ 3876.799810]  find_parent_nodes+0x38d/0x1180 [btrfs]
  [ 3876.799841]  btrfs_check_shared+0x11a/0x1d0 [btrfs]
  [ 3876.799870]  ? extent_fiemap+0x598/0x6e0 [btrfs]
  [ 3876.799895]  extent_fiemap+0x598/0x6e0 [btrfs]
  [ 3876.799913]  do_vfs_ioctl+0x45a/0x700
  [ 3876.799926]  ksys_ioctl+0x70/0x80
  [ 3876.799938]  ? trace_hardirqs_off_thunk+0x1a/0x20
  [ 3876.799953]  __x64_sys_ioctl+0x16/0x20
  [ 3876.799965]  do_syscall_64+0x62/0x220
  [ 3876.799977]  entry_SYSCALL_64_after_hwframe+0x49/0xbe
  [ 3876.799993] RIP: 0033:0x7f23e0013dd7
  (...)
  [ 3876.800056] RSP: 002b:00007f23ddd96ca8 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
  [ 3876.800078] RAX: ffffffffffffffda RBX: 00007f23d80210f8 RCX: 00007f23e0013dd7
  [ 3876.800099] RDX: 00007f23d80210f8 RSI: 00000000c020660b RDI: 0000000000000003
  [ 3876.800626] RBP: 000055fa2a2a2440 R08: 0000000000000000 R09: 00007f23ddd96d7c
  [ 3876.801143] R10: 00007f23d8022000 R11: 0000000000000246 R12: 00007f23ddd96d80
  [ 3876.801662] R13: 00007f23ddd96d78 R14: 00007f23d80210f0 R15: 00007f23ddd96d80
  (...)
  [ 3876.805107] ---[ end trace e53161e179ef04f9 ]---

Fix that by saving the root's header owner field into a local variable before freeing the root's extent buffer, and then use that local variable when needed.

Fixes 30b0463a9394d9 ("Btrfs: fix accessing the root pointer in tree mod log functions")
---
 fs/btrfs/ctree.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
fs/btrfs/ctree.c +3 -1
@@ -1343,6 +1343,7 @@ get_old_root(struct btrfs_root *root, u6
 	struct tree_mod_elem *tm;
 	struct extent_buffer *eb = NULL;
 	struct extent_buffer *eb_root;
+	u64 eb_root_owner = 0;
 	struct extent_buffer *old;
 	struct tree_mod_root *old_root = NULL;
 	u64 old_generation = 0;
@@ -1380,6 +1381,7 @@ get_old_root(struct btrfs_root *root, u6
 			free_extent_buffer(old);
 		}
 	} else if (old_root) {
+		eb_root_owner = btrfs_header_owner(eb_root);
 		btrfs_tree_read_unlock(eb_root);
 		free_extent_buffer(eb_root);
 		eb = alloc_dummy_extent_buffer(fs_info, logical);
@@ -1396,7 +1398,7 @@ get_old_root(struct btrfs_root *root, u6
 	if (old_root) {
 		btrfs_set_header_bytenr(eb, eb->start);
 		btrfs_set_header_backref_rev(eb, BTRFS_MIXED_BACKREF_REV);
-		btrfs_set_header_owner(eb, btrfs_header_owner(eb_root));
+		btrfs_set_header_owner(eb, eb_root_owner);
 		btrfs_set_header_level(eb, old_root->level);
 		btrfs_set_header_generation(eb, old_generation);
 	}
author maintainer
#287 lore raw

PATCH 284/313 btrfs: Relinquish CPUs in btrfs_compare_trees

In reply to Greg Kroah-Hartman

From: Nikolay Borisov <[email protected]>

commit 6af112b11a4bc1b560f60a618ac9c1dcefe9836e upstream.

When doing any form of incremental send the parent and the child trees need to be compared via btrfs_compare_trees. This can result in long loop chains without ever relinquishing the CPU. This causes softlockup detector to trigger when comparing trees with a lot of items. Example report:

watchdog: BUG: soft lockup - CPU#0 stuck for 24s! [snapperd:16153]
CPU: 0 PID: 16153 Comm: snapperd Not tainted 5.2.9-1-default #1 openSUSE Tumbleweed (unreleased)
Hardware name: QEMU KVM Virtual Machine, BIOS 0.0.0 02/06/2015
pstate: 40000005 (nZcv daif -PAN -UAO)
pc : __ll_sc_arch_atomic_sub_return+0x14/0x20
lr : btrfs_release_extent_buffer_pages+0xe0/0x1e8 [btrfs]
sp : ffff00001273b7e0
Call trace:
 __ll_sc_arch_atomic_sub_return+0x14/0x20
 release_extent_buffer+0xdc/0x120 [btrfs]
 free_extent_buffer.part.0+0xb0/0x118 [btrfs]
 free_extent_buffer+0x24/0x30 [btrfs]
 btrfs_release_path+0x4c/0xa0 [btrfs]
 btrfs_free_path.part.0+0x20/0x40 [btrfs]
 btrfs_free_path+0x24/0x30 [btrfs]
 get_inode_info+0xa8/0xf8 [btrfs]
 finish_inode_if_needed+0xe0/0x6d8 [btrfs]
 changed_cb+0x9c/0x410 [btrfs]
 btrfs_compare_trees+0x284/0x648 [btrfs]
 send_subvol+0x33c/0x520 [btrfs]
 btrfs_ioctl_send+0x8a0/0xaf0 [btrfs]
 btrfs_ioctl+0x199c/0x2288 [btrfs]
 do_vfs_ioctl+0x4b0/0x820
 ksys_ioctl+0x84/0xb8
 __arm64_sys_ioctl+0x28/0x38
 el0_svc_common.constprop.0+0x7c/0x188
 el0_svc_handler+0x34/0x90
 el0_svc+0x8/0xc

Fix this by adding a call to cond_resched at the beginning of the main loop in btrfs_compare_trees.

Fixes 7069830a9e38 ("Btrfs: add btrfs_compare_trees function")
---
 fs/btrfs/ctree.c |    1 +
 1 file changed, 1 insertion(+)
fs/btrfs/ctree.c +1 -0
@@ -5477,6 +5477,7 @@ int btrfs_compare_trees(struct btrfs_roo
 	advance_left = advance_right = 0;
 
 	while (1) {
+		cond_resched();
 		if (advance_left && !left_end_reached) {
 			ret = tree_advance(left_path, &left_level,
 					left_root_level,
author maintainer
#288 lore raw

PATCH 285/313 btrfs: adjust dirty_metadata_bytes after writeback failure of extent buffer

In reply to Greg Kroah-Hartman

From: Dennis Zhou <[email protected]>

commit eb5b64f142504a597d67e2109d603055ff765e52 upstream.

Before, if a eb failed to write out, we would end up triggering a BUG_ON(). As of f4340622e0226 ("btrfs: extent_io: Move the BUG_ON() in flush_write_bio() one level up"), we no longer BUG_ON(), so we should make life consistent and add back the unwritten bytes to dirty_metadata_bytes.

Fixes f4340622e022 ("btrfs: extent_io: Move the BUG_ON() in flush_write_bio() one level up")
---
 fs/btrfs/extent_io.c |    9 +++++++++
 1 file changed, 9 insertions(+)
fs/btrfs/extent_io.c +9 -0
@@ -3708,12 +3708,21 @@ err_unlock:
 static void set_btree_ioerr(struct page *page)
 {
 	struct extent_buffer *eb = (struct extent_buffer *)page->private;
+	struct btrfs_fs_info *fs_info;
 
 	SetPageError(page);
 	if (test_and_set_bit(EXTENT_BUFFER_WRITE_ERR, &eb->bflags))
 		return;
 
 	/*
+	 * If we error out, we should add back the dirty_metadata_bytes
+	 * to make it consistent.
+	 */
+	fs_info = eb->fs_info;
+	percpu_counter_add_batch(&fs_info->dirty_metadata_bytes,
+				 eb->len, fs_info->dirty_metadata_batch);
+
+	/*
 	 * If writeback for a btree extent that doesn't belong to a log tree
 	 * failed, increment the counter transaction->eb_write_errors.
 	 * We do this because while the transaction is running and before it's
author maintainer
#289 lore raw

PATCH 286/313 btrfs: qgroup: Fix the wrong target io_tree when freeing reserved data space

In reply to Greg Kroah-Hartman

From: Qu Wenruo <[email protected]>

commit bab32fc069ce8829c416e8737c119f62a57970f9 upstream.

[BUG]
Under the following case with qgroup enabled, if some error happened after we have reserved delalloc space, then in error handling path, we could cause qgroup data space leakage:

Quoted 1 line

From btrfs_truncate_block() in inode.c:

	ret = btrfs_delalloc_reserve_space(inode, &data_reserved,
					   block_start, blocksize);
	if (ret)
		goto out;

again:

	page = find_or_create_page(mapping, index, mask);
	if (!page) {
		btrfs_delalloc_release_space(inode, data_reserved,
					     block_start, blocksize, true);
		btrfs_delalloc_release_extents(BTRFS_I(inode), blocksize, true);
		ret = -ENOMEM;
		goto out;
	}

[CAUSE]
In the above case, btrfs_delalloc_reserve_space() will call btrfs_qgroup_reserve_data() and mark the io_tree range with EXTENT_QGROUP_RESERVED flag.

In the error handling path, we have the following call stack:
btrfs_delalloc_release_space()
|- btrfs_free_reserved_data_space()
|- btrsf_qgroup_free_data()

      |- __btrfs_qgroup_release_data(reserved=@reserved, free=1)
         |- qgroup_free_reserved_data(reserved=@reserved)
            |- clear_record_extent_bits();
            |- freed += changeset.bytes_changed;

However due to a completion bug, qgroup_free_reserved_data() will clear EXTENT_QGROUP_RESERVED flag in BTRFS_I(inode)->io_failure_tree, other than the correct BTRFS_I(inode)->io_tree.
Since io_failure_tree is never marked with that flag, btrfs_qgroup_free_data() will not free any data reserved space at all, causing a leakage.

This type of error handling can only be triggered by errors outside of qgroup code. So EDQUOT error from qgroup can't trigger it.

[FIX]
Fix the wrong target io_tree.

Fixes bc42bda22345 ("btrfs: qgroup: Fix qgroup reserved space underflow by only freeing reserved ranges")
---
 fs/btrfs/qgroup.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
fs/btrfs/qgroup.c +1 -1
@@ -3469,7 +3469,7 @@ static int qgroup_free_reserved_data(str
 		 * EXTENT_QGROUP_RESERVED, we won't double free.
 		 * So not need to rush.
 		 */
-		ret = clear_record_extent_bits(&BTRFS_I(inode)->io_failure_tree,
+		ret = clear_record_extent_bits(&BTRFS_I(inode)->io_tree,
 				free_start, free_start + free_len - 1,
 				EXTENT_QGROUP_RESERVED, &changeset);
 		if (ret < 0)
author maintainer
#290 lore raw

PATCH 287/313 btrfs: qgroup: Fix reserved data space leak if we have multiple reserve calls

In reply to Greg Kroah-Hartman

From: Qu Wenruo <[email protected]>

commit d4e204948fe3e0dc8e1fbf3f8f3290c9c2823be3 upstream.

[BUG]
The following script can cause btrfs qgroup data space leak:

mkfs.btrfs -f $dev
mount $dev -o nospace_cache $mnt

btrfs subv create $mnt/subv
btrfs quota en $mnt
btrfs quota rescan -w $mnt
btrfs qgroup limit 128m $mnt/subv

for (( i = 0; i < 3; i++)); do
# Create 3 64M holes for latter fallocate to fail truncate -s 192m $mnt/subv/file
xfs_io -c "pwrite 64m 4k" $mnt/subv/file > /dev/null xfs_io -c "pwrite 128m 4k" $mnt/subv/file > /dev/null sync

# it's supposed to fail, and each failure will leak at least 64M # data space
xfs_io -f -c "falloc 0 192m" $mnt/subv/file &> /dev/null rm $mnt/subv/file
sync
done

# Shouldn't fail after we removed the file
xfs_io -f -c "falloc 0 64m" $mnt/subv/file

[CAUSE]
Btrfs qgroup data reserve code allow multiple reservations to happen on a single extent_changeset:
E.g:

	btrfs_qgroup_reserve_data(inode, &data_reserved, 0, SZ_1M);
	btrfs_qgroup_reserve_data(inode, &data_reserved, SZ_1M, SZ_2M);
	btrfs_qgroup_reserve_data(inode, &data_reserved, 0, SZ_4M);

Btrfs qgroup code has its internal tracking to make sure we don't double-reserve in above example.

The only pattern utilizing this feature is in the main while loop of btrfs_fallocate() function.

However btrfs_qgroup_reserve_data()'s error handling has a bug in that on error it clears all ranges in the io_tree with EXTENT_QGROUP_RESERVED flag but doesn't free previously reserved bytes.

This bug has a two fold effect:
- Clearing EXTENT_QGROUP_RESERVED ranges
This is the correct behavior, but it prevents btrfs_qgroup_check_reserved_leak() to catch the leakage as the detector is purely EXTENT_QGROUP_RESERVED flag based.

- Leak the previously reserved data bytes.

The bug manifests when N calls to btrfs_qgroup_reserve_data are made and the last one fails, leaking space reserved in the previous ones.

[FIX]
Also free previously reserved data bytes when btrfs_qgroup_reserve_data fails.

Fixes 524725537023 ("btrfs: qgroup: Introduce btrfs_qgroup_reserve_data function")
---
 fs/btrfs/qgroup.c |    3 +++
 1 file changed, 3 insertions(+)
fs/btrfs/qgroup.c +3 -0
@@ -3425,6 +3425,9 @@ cleanup:
 	while ((unode = ulist_next(&reserved->range_changed, &uiter)))
 		clear_extent_bit(&BTRFS_I(inode)->io_tree, unode->val,
 				 unode->aux, EXTENT_QGROUP_RESERVED, 0, 0, NULL);
+	/* Also free data bytes of already reserved one */
+	btrfs_qgroup_free_refroot(root->fs_info, root->root_key.objectid,
+				  orig_reserved, BTRFS_QGROUP_RSV_DATA);
 	extent_changeset_release(reserved);
 	return ret;
 }
author maintainer
#291 lore raw

PATCH 288/313 Btrfs: fix race setting up and completing qgroup rescan workers

In reply to Greg Kroah-Hartman

From: Filipe Manana <[email protected]>

commit 13fc1d271a2e3ab8a02071e711add01fab9271f6 upstream.

There is a race between setting up a qgroup rescan worker and completing a qgroup rescan worker that can lead to callers of the qgroup rescan wait ioctl to either not wait for the rescan worker to complete or to hang forever due to missing wake ups. The following diagram shows a sequence of steps that illustrates the race.

CPU 1 CPU 2 CPU 3

btrfs_ioctl_quota_rescan()
btrfs_qgroup_rescan()
qgroup_rescan_init()

    mutex_lock(&fs_info->qgroup_rescan_lock)
    spin_lock(&fs_info->qgroup_lock)

    fs_info->qgroup_flags |=
      BTRFS_QGROUP_STATUS_FLAG_RESCAN

    init_completion(
      &fs_info->qgroup_rescan_completion)

    fs_info->qgroup_rescan_running = true

    mutex_unlock(&fs_info->qgroup_rescan_lock)
    spin_unlock(&fs_info->qgroup_lock)

    btrfs_init_work()
     --> starts the worker

                                                        btrfs_qgroup_rescan_worker()
                                                         mutex_lock(&fs_info->qgroup_rescan_lock)

                                                         fs_info->qgroup_flags &=
                                                           ~BTRFS_QGROUP_STATUS_FLAG_RESCAN

                                                         mutex_unlock(&fs_info->qgroup_rescan_lock)

                                                         starts transaction, updates qgroup status
                                                         item, etc

                                                                                                           btrfs_ioctl_quota_rescan()
                                                                                                            btrfs_qgroup_rescan()
                                                                                                             qgroup_rescan_init()
                                                                                                              mutex_lock(&fs_info->qgroup_rescan_lock)
                                                                                                              spin_lock(&fs_info->qgroup_lock)

                                                                                                              fs_info->qgroup_flags |=
                                                                                                                BTRFS_QGROUP_STATUS_FLAG_RESCAN

                                                                                                              init_completion(
                                                                                                                &fs_info->qgroup_rescan_completion)

                                                                                                              fs_info->qgroup_rescan_running = true

                                                                                                              mutex_unlock(&fs_info->qgroup_rescan_lock)
                                                                                                              spin_unlock(&fs_info->qgroup_lock)

                                                                                                              btrfs_init_work()
                                                                                                               --> starts another worker

                                                         mutex_lock(&fs_info->qgroup_rescan_lock)

                                                         fs_info->qgroup_rescan_running = false

                                                         mutex_unlock(&fs_info->qgroup_rescan_lock)

							 complete_all(&fs_info->qgroup_rescan_completion)

Before the rescan worker started by the task at CPU 3 completes, if another task calls btrfs_ioctl_quota_rescan(), it will get -EINPROGRESS because the flag BTRFS_QGROUP_STATUS_FLAG_RESCAN is set at fs_info->qgroup_flags, which is expected and correct behaviour.

However if other task calls btrfs_ioctl_quota_rescan_wait() before the rescan worker started by the task at CPU 3 completes, it will return immediately without waiting for the new rescan worker to complete, because fs_info->qgroup_rescan_running is set to false by CPU 2.

This race is making test case btrfs/171 (from fstests) to fail often:

btrfs/171 9s ... - output mismatch (see /home/fdmanana/git/hub/xfstests/results//btrfs/171.out.bad)

#      --- tests/btrfs/171.out     2018-09-16 21:30:48.505104287 +0100
#      +++ /home/fdmanana/git/hub/xfstests/results//btrfs/171.out.bad      2019-09-19 02:01:36.938486039 +0100
#      @@ -1,2 +1,3 @@

# QA output created by 171
# +ERROR: quota rescan failed: Operation now in progress
# Silence is golden
# ...
# (Run 'diff -u /home/fdmanana/git/hub/xfstests/tests/btrfs/171.out /home/fdmanana/git/hub/xfstests/results//btrfs/171.out.bad' to see the entire diff)

That is because the test calls the btrfs-progs commands "qgroup quota rescan -w", "qgroup assign" and "qgroup remove" in a sequence that makes calls to the rescan start ioctl fail with -EINPROGRESS (note the "btrfs" commands 'qgroup assign' and 'qgroup remove' often call the rescan start ioctl after calling the qgroup assign ioctl, btrfs_ioctl_qgroup_assign()), since previous waits didn't actually wait for a rescan worker to complete.

Another problem the race can cause is missing wake ups for waiters, since the call to complete_all() happens outside a critical section and after clearing the flag BTRFS_QGROUP_STATUS_FLAG_RESCAN. In the sequence diagram above, if we have a waiter for the first rescan task (executed by CPU 2), then fs_info->qgroup_rescan_completion.wait is not empty, and if after the rescan worker clears BTRFS_QGROUP_STATUS_FLAG_RESCAN and before it calls complete_all() against fs_info->qgroup_rescan_completion, the task at CPU 3 calls init_completion() against fs_info->qgroup_rescan_completion which re-initilizes its wait queue to an empty queue, therefore causing the rescan worker at CPU 2 to call complete_all() against an empty queue, never waking up the task waiting for that rescan worker.

Fix this by clearing BTRFS_QGROUP_STATUS_FLAG_RESCAN and setting fs_info->qgroup_rescan_running to false in the same critical section, delimited by the mutex fs_info->qgroup_rescan_lock, as well as doing the call to complete_all() in that same critical section. This gives the protection needed to avoid rescan wait ioctl callers not waiting for a running rescan worker and the lost wake ups problem, since setting that rescan flag and boolean as well as initializing the wait queue is done already in a critical section delimited by that mutex (at qgroup_rescan_init()).

Fixes 57254b6ebce4ce ("Btrfs: add ioctl to wait for qgroup rescan completion")
Fixes d2c609b834d62f ("btrfs: properly track when rescan worker is running")
---
 fs/btrfs/qgroup.c |   33 +++++++++++++++++++--------------
 1 file changed, 19 insertions(+), 14 deletions(-)
fs/btrfs/qgroup.c +19 -14
@@ -3154,9 +3154,6 @@ out:
 	btrfs_free_path(path);
 
 	mutex_lock(&fs_info->qgroup_rescan_lock);
-	if (!btrfs_fs_closing(fs_info))
-		fs_info->qgroup_flags &= ~BTRFS_QGROUP_STATUS_FLAG_RESCAN;
-
 	if (err > 0 &&
 	    fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT) {
 		fs_info->qgroup_flags &= ~BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT;
@@ -3172,16 +3169,30 @@ out:
 	trans = btrfs_start_transaction(fs_info->quota_root, 1);
 	if (IS_ERR(trans)) {
 		err = PTR_ERR(trans);
+		trans = NULL;
 		btrfs_err(fs_info,
 			  "fail to start transaction for status update: %d",
 			  err);
-		goto done;
 	}
-	ret = update_qgroup_status_item(trans);
-	if (ret < 0) {
-		err = ret;
-		btrfs_err(fs_info, "fail to update qgroup status: %d", err);
+
+	mutex_lock(&fs_info->qgroup_rescan_lock);
+	if (!btrfs_fs_closing(fs_info))
+		fs_info->qgroup_flags &= ~BTRFS_QGROUP_STATUS_FLAG_RESCAN;
+	if (trans) {
+		ret = update_qgroup_status_item(trans);
+		if (ret < 0) {
+			err = ret;
+			btrfs_err(fs_info, "fail to update qgroup status: %d",
+				  err);
+		}
 	}
+	fs_info->qgroup_rescan_running = false;
+	complete_all(&fs_info->qgroup_rescan_completion);
+	mutex_unlock(&fs_info->qgroup_rescan_lock);
+
+	if (!trans)
+		return;
+
 	btrfs_end_transaction(trans);
 
 	if (btrfs_fs_closing(fs_info)) {
@@ -3192,12 +3203,6 @@ out:
 	} else {
 		btrfs_err(fs_info, "qgroup scan failed with %d", err);
 	}
-
-done:
-	mutex_lock(&fs_info->qgroup_rescan_lock);
-	fs_info->qgroup_rescan_running = false;
-	mutex_unlock(&fs_info->qgroup_rescan_lock);
-	complete_all(&fs_info->qgroup_rescan_completion);
 }
 
 /*
author maintainer
#292 lore raw

PATCH 289/313 SUNRPC: Dequeue the request from the receive queue while were re-encoding

In reply to Greg Kroah-Hartman

From: Trond Myklebust <[email protected]>

commit cc204d01262a69218b2d0db5cdea371de85871d9 upstream.

Ensure that we dequeue the request from the transport receive queue while we're re-encoding to prevent issues like use-after-free when we release the bvec.

Fixes 7536908982047 ("SUNRPC: Ensure the bvecs are reset when we re-encode...")
Cc [email protected] # v4.20+
---
 include/linux/sunrpc/xprt.h |    1 
 net/sunrpc/clnt.c           |    6 ++--
 net/sunrpc/xprt.c           |   54 +++++++++++++++++++++++++-------------------
 3 files changed, 35 insertions(+), 26 deletions(-)
include/linux/sunrpc/xprt.h +1 -0
@@ -346,6 +346,7 @@ bool			xprt_prepare_transmit(struct rpc_
 void			xprt_request_enqueue_transmit(struct rpc_task *task);
 void			xprt_request_enqueue_receive(struct rpc_task *task);
 void			xprt_request_wait_receive(struct rpc_task *task);
+void			xprt_request_dequeue_xprt(struct rpc_task *task);
 bool			xprt_request_need_retransmit(struct rpc_task *task);
 void			xprt_transmit(struct rpc_task *task);
 void			xprt_end_transmit(struct rpc_task *task);
net/sunrpc/clnt.c +3 -3
@@ -1785,6 +1785,7 @@ rpc_xdr_encode(struct rpc_task *task)
 		     req->rq_rbuffer,
 		     req->rq_rcvsize);
 
+	req->rq_reply_bytes_recvd = 0;
 	req->rq_snd_buf.head[0].iov_len = 0;
 	xdr_init_encode(&xdr, &req->rq_snd_buf,
 			req->rq_snd_buf.head[0].iov_base, req);
@@ -1804,6 +1805,8 @@ call_encode(struct rpc_task *task)
 	if (!rpc_task_need_encode(task))
 		goto out;
 	dprint_status(task);
+	/* Dequeue task from the receive queue while we're encoding */
+	xprt_request_dequeue_xprt(task);
 	/* Encode here so that rpcsec_gss can use correct sequence number. */
 	rpc_xdr_encode(task);
 	/* Did the encode result in an error condition? */
@@ -2437,9 +2440,6 @@ call_decode(struct rpc_task *task)
 		return;
 	case -EAGAIN:
 		task->tk_status = 0;
-		xdr_free_bvec(&req->rq_rcv_buf);
-		req->rq_reply_bytes_recvd = 0;
-		req->rq_rcv_buf.len = 0;
 		if (task->tk_client->cl_discrtry)
 			xprt_conditional_disconnect(req->rq_xprt,
 						    req->rq_connect_cookie);
net/sunrpc/xprt.c +31 -23

72 lines not on this page · raw message

author maintainer
#293 lore raw

PATCH 290/313 SUNRPC: Fix buffer handling of GSS MIC without slack

In reply to Greg Kroah-Hartman

From: Benjamin Coddington <[email protected]>

commit 5f1bc39979d868a0358c683864bec3fc8395440b upstream.

The GSS Message Integrity Check data for krb5i may lie partially in the XDR reply buffer's pages and tail. If so, we try to copy the entire MIC into free space in the tail. But as the estimations of the slack space required for authentication and verification have improved there may be less free space in the tail to complete this copy -- see commit 2c94b8eca1a2 ("SUNRPC: Use au_rslack when computing reply buffer size"). In fact, there may only be room in the tail for a single copy of the MIC, and not part of the MIC and then another complete copy.

The real world failure reported is that ls of a directory on NFS may sometimes return -EIO, which can be traced back to xdr_buf_read_netobj() failing to find available free space in the tail to copy the MIC.

Fix this by checking for the case of the MIC crossing the boundaries of head, pages, and tail. If so, shift the buffer until the MIC is contained completely within the pages or tail. This allows the remainder of the function to create a sub buffer that directly address the complete MIC.

---
 net/sunrpc/xdr.c |   27 ++++++++++++++++++---------
 1 file changed, 18 insertions(+), 9 deletions(-)
net/sunrpc/xdr.c +18 -9
@@ -1237,16 +1237,29 @@ xdr_encode_word(struct xdr_buf *buf, uns
 EXPORT_SYMBOL_GPL(xdr_encode_word);
 
 /* If the netobj starting offset bytes from the start of xdr_buf is contained
- * entirely in the head or the tail, set object to point to it; otherwise
- * try to find space for it at the end of the tail, copy it there, and
- * set obj to point to it. */
+ * entirely in the head, pages, or tail, set object to point to it; otherwise
+ * shift the buffer until it is contained entirely within the pages or tail.
+ */
 int xdr_buf_read_netobj(struct xdr_buf *buf, struct xdr_netobj *obj, unsigned int offset)
 {
 	struct xdr_buf subbuf;
+	unsigned int boundary;
 
 	if (xdr_decode_word(buf, offset, &obj->len))
 		return -EFAULT;
-	if (xdr_buf_subsegment(buf, &subbuf, offset + 4, obj->len))
+	offset += 4;
+
+	/* Is the obj partially in the head? */
+	boundary = buf->head[0].iov_len;
+	if (offset < boundary && (offset + obj->len) > boundary)
+		xdr_shift_buf(buf, boundary - offset);
+
+	/* Is the obj partially in the pages? */
+	boundary += buf->page_len;
+	if (offset < boundary && (offset + obj->len) > boundary)
+		xdr_shrink_pagelen(buf, boundary - offset);
+
+	if (xdr_buf_subsegment(buf, &subbuf, offset, obj->len))
 		return -EFAULT;
 
 	/* Is the obj contained entirely in the head? */
@@ -1258,11 +1271,7 @@ int xdr_buf_read_netobj(struct xdr_buf *
 	if (subbuf.tail[0].iov_len == obj->len)
 		return 0;
 
-	/* use end of tail as storage for obj:
-	 * (We don't copy to the beginning because then we'd have
-	 * to worry about doing a potentially overlapping copy.
-	 * This assumes the object is at most half the length of the
-	 * tail.) */
+	/* Find a contiguous area in @buf to hold all of @obj */
 	if (obj->len > buf->buflen - buf->len)
 		return -ENOMEM;
 	if (buf->tail[0].iov_len != 0)
author maintainer
#294 lore raw

PATCH 291/313 ACPI / LPSS: Save/restore LPSS private registers also on Lynxpoint

In reply to Greg Kroah-Hartman

From: Jarkko Nikula <[email protected]>

commit 57b3006492a4c11b2d4a772b5b2905d544a32037 upstream.

My assumption in commit b53548f9d9e4 ("spi: pxa2xx: Remove LPSS private register restoring during resume") that Intel Lynxpoint and compatible based chipsets may not need LPSS private registers saving and restoring over suspend/resume cycle turned out to be false on Intel Broadwell.

Curtis Malainey sent a patch bringing above change back and reported the LPSS SPI Chip Select control was lost over suspend/resume cycle on Broadwell machine.

Instead of reverting above commit lets add LPSS private register saving/restoring also for all LPSS SPI, I2C and UART controllers on Lynxpoint and compatible chipset to make sure context is not lost in case nothing else preserves it like firmware or if LPSS is always on.

Fixes b53548f9d9e4 ("spi: pxa2xx: Remove LPSS private register restoring during resume")
Cc 5.0+ <[email protected]> # 5.0+
---
 drivers/acpi/acpi_lpss.c |    8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)
drivers/acpi/acpi_lpss.c +5 -3
@@ -219,12 +219,13 @@ static void bsw_pwm_setup(struct lpss_pr
 }
 
 static const struct lpss_device_desc lpt_dev_desc = {
-	.flags = LPSS_CLK | LPSS_CLK_GATE | LPSS_CLK_DIVIDER | LPSS_LTR,
+	.flags = LPSS_CLK | LPSS_CLK_GATE | LPSS_CLK_DIVIDER | LPSS_LTR
+			| LPSS_SAVE_CTX,
 	.prv_offset = 0x800,
 };
 
 static const struct lpss_device_desc lpt_i2c_dev_desc = {
-	.flags = LPSS_CLK | LPSS_CLK_GATE | LPSS_LTR,
+	.flags = LPSS_CLK | LPSS_CLK_GATE | LPSS_LTR | LPSS_SAVE_CTX,
 	.prv_offset = 0x800,
 };
 
@@ -236,7 +237,8 @@ static struct property_entry uart_proper
 };
 
 static const struct lpss_device_desc lpt_uart_dev_desc = {
-	.flags = LPSS_CLK | LPSS_CLK_GATE | LPSS_CLK_DIVIDER | LPSS_LTR,
+	.flags = LPSS_CLK | LPSS_CLK_GATE | LPSS_CLK_DIVIDER | LPSS_LTR
+			| LPSS_SAVE_CTX,
 	.clk_con_id = "baudclk",
 	.prv_offset = 0x800,
 	.setup = lpss_uart_setup,
author maintainer
#295 lore raw

PATCH 292/313 md/raid6: Set R5_ReadError when there is read failure on parity disk

In reply to Greg Kroah-Hartman

From: Xiao Ni <[email protected]>

commit 143f6e733b73051cd22dcb80951c6c929da413ce upstream.

7471fb77ce4d ("md/raid6: Fix anomily when recovering a single device in RAID6.") avoids rereading P when it can be computed from other members. However, this misses the chance to re-write the right data to P. This patch sets R5_ReadError if the re-read fails.

Also, when re-read is skipped, we also missed the chance to reset rdev->read_errors to 0. It can fail the disk when there are many read errors on P member disk (other disks don't have read error)

V2: upper layer read request don't read parity/Q data. So there is no need to consider such situation.

This is Reported-by: kbuild test robot <[email protected]>

Fixes 7471fb77ce4d ("md/raid6: Fix anomily when recovering a single device in RAID6.")
---
 drivers/md/raid5.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/md/raid5.c +3 -1
@@ -2559,7 +2559,9 @@ static void raid5_end_read_request(struc
 		    && !test_bit(R5_ReadNoMerge, &sh->dev[i].flags))
 			retry = 1;
 		if (retry)
-			if (test_bit(R5_ReadNoMerge, &sh->dev[i].flags)) {
+			if (sh->qd_idx >= 0 && sh->pd_idx == i)
+				set_bit(R5_ReadError, &sh->dev[i].flags);
+			else if (test_bit(R5_ReadNoMerge, &sh->dev[i].flags)) {
 				set_bit(R5_ReadError, &sh->dev[i].flags);
 				clear_bit(R5_ReadNoMerge, &sh->dev[i].flags);
 			} else
author maintainer
#296 lore raw

PATCH 293/313 md: dont report active array_state until after revalidate_disk() completes.

In reply to Greg Kroah-Hartman

From: NeilBrown <[email protected]>

commit 9d4b45d6af442237560d0bb5502a012baa5234b7 upstream.

Until revalidate_disk() has completed, the size of a new md array will appear to be zero.
So we shouldn't report, through array_state, that the array is active until that time.
udev rules check array_state to see if the array is ready. As soon as it appear to be zero, fsck can be run. If it find the size to be zero, it will fail.

So add a new flag to provide an interlock between do_md_run() and array_state_show(). This flag is set while do_md_run() is active and it prevents array_state_show() from reporting that the array is active.

Before do_md_run() is called, ->pers will be NULL so array is definitely not active.
After do_md_run() is called, revalidate_disk() will have run and the array will be completely ready.

We also move various sysfs_notify*() calls out of md_run() into do_md_run() after MD_NOT_READY is cleared. This ensure the information is ready before the notification is sent.

Prior to v4.12, array_state_show() was called with the mddev->reconfig_mutex held, which provided exclusion with do_md_run().

Note that MD_NOT_READY cleared twice. This is deliberate to cover both success and error paths with minimal noise.

Fixes b7b17c9b67e5 ("md: remove mddev_lock() from md_attr_show()")
Cc [email protected] (v4.12++)
---
 drivers/md/md.c |   11 +++++++----
 drivers/md/md.h |    3 +++
 2 files changed, 10 insertions(+), 4 deletions(-)
drivers/md/md.c +7 -4
@@ -4105,7 +4105,7 @@ array_state_show(struct mddev *mddev, ch
 {
 	enum array_state st = inactive;
 
-	if (mddev->pers)
+	if (mddev->pers && !test_bit(MD_NOT_READY, &mddev->flags))
 		switch(mddev->ro) {
 		case 1:
 			st = readonly;
@@ -5660,9 +5660,6 @@ int md_run(struct mddev *mddev)
 		md_update_sb(mddev, 0);
 
 	md_new_event(mddev);
-	sysfs_notify_dirent_safe(mddev->sysfs_state);
-	sysfs_notify_dirent_safe(mddev->sysfs_action);
-	sysfs_notify(&mddev->kobj, NULL, "degraded");
 	return 0;
 
 abort:
@@ -5676,6 +5673,7 @@ static int do_md_run(struct mddev *mddev
 {
 	int err;
 
+	set_bit(MD_NOT_READY, &mddev->flags);
 	err = md_run(mddev);
 	if (err)
 		goto out;
@@ -5696,9 +5694,14 @@ static int do_md_run(struct mddev *mddev
 
 	set_capacity(mddev->gendisk, mddev->array_sectors);
 	revalidate_disk(mddev->gendisk);
+	clear_bit(MD_NOT_READY, &mddev->flags);
 	mddev->changed = 1;
 	kobject_uevent(&disk_to_dev(mddev->gendisk)->kobj, KOBJ_CHANGE);
+	sysfs_notify_dirent_safe(mddev->sysfs_state);
+	sysfs_notify_dirent_safe(mddev->sysfs_action);
+	sysfs_notify(&mddev->kobj, NULL, "degraded");
 out:
+	clear_bit(MD_NOT_READY, &mddev->flags);
 	return err;
 }
 
drivers/md/md.h +3 -0
@@ -236,6 +236,9 @@ enum mddev_flags {
 	MD_UPDATING_SB,		/* md_check_recovery is updating the metadata
 				 * without explicitly holding reconfig_mutex.
 				 */
+	MD_NOT_READY,		/* do_md_run() is active, so 'array_state'
+				 * must not report that array is ready yet
+				 */
 };
 
 enum mddev_sb_flags {
author maintainer
#297 lore raw

PATCH 294/313 md: only call set_in_sync() when it is expected to succeed.

In reply to Greg Kroah-Hartman

From: NeilBrown <[email protected]>

commit 480523feae581ab714ba6610388a3b4619a2f695 upstream.

Since commit 4ad23a976413 ("MD: use per-cpu counter for writes_pending"), set_in_sync() is substantially more expensive: it can wait for a full RCU grace period which can be 10s of milliseconds.

So we should only call it when the cost is justified.

md_check_recovery() currently calls set_in_sync() every time it finds anything to do (on non-external active arrays). For an array performing resync or recovery, this will be quite often. Each call will introduce a delay to the md thread, which can noticeable affect IO submission latency.

In md_check_recovery() we only need to call set_in_sync() if 'safemode' was non-zero at entry, meaning that there has been not recent IO. So we save this "safemode was nonzero" state, and only call set_in_sync() if it was non-zero.

This measurably reduces mean and maximum IO submission latency during resync/recovery.

Reported-and-tested-by Jack Wang <[email protected]>
Fixes 4ad23a976413 ("MD: use per-cpu counter for writes_pending")
Cc [email protected] (v4.12+)
---
 drivers/md/md.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/md/md.c +2 -1
@@ -8811,6 +8811,7 @@ void md_check_recovery(struct mddev *mdd
 
 	if (mddev_trylock(mddev)) {
 		int spares = 0;
+		bool try_set_sync = mddev->safemode != 0;
 
 		if (!mddev->external && mddev->safemode == 1)
 			mddev->safemode = 0;
@@ -8856,7 +8857,7 @@ void md_check_recovery(struct mddev *mdd
 			}
 		}
 
-		if (!mddev->external && !mddev->in_sync) {
+		if (try_set_sync && !mddev->external && !mddev->in_sync) {
 			spin_lock(&mddev->lock);
 			set_in_sync(mddev);
 			spin_unlock(&mddev->lock);
author maintainer
#298 lore raw

PATCH 295/313 cfg80211: Purge frame registrations on iftype change

In reply to Greg Kroah-Hartman

From: Denis Kenzior <[email protected]>

commit c1d3ad84eae35414b6b334790048406bd6301b12 upstream.

Currently frame registrations are not purged, even when changing the interface type. This can lead to potentially weird situations where frames possibly not allowed on a given interface type remain registered due to the type switching happening after registration.

The kernel currently relies on userspace apps to actually purge the registrations themselves, this is not something that the kernel should rely on.

Add a call to cfg80211_mlme_purge_registrations() to forcefully remove any registrations left over prior to switching the iftype.

---
 net/wireless/util.c |    1 +
 1 file changed, 1 insertion(+)
net/wireless/util.c +1 -0
@@ -960,6 +960,7 @@ int cfg80211_change_iface(struct cfg8021
 		}
 
 		cfg80211_process_rdev_events(rdev);
+		cfg80211_mlme_purge_registrations(dev->ieee80211_ptr);
 	}
 
 	err = rdev_change_virtual_intf(rdev, dev, ntype, params);
author maintainer
#299 lore raw

PATCH 296/313 /dev/mem: Bail out upon SIGKILL.

In reply to Greg Kroah-Hartman

From: Tetsuo Handa <[email protected]>

commit 8619e5bdeee8b2c685d686281f2d2a6017c4bc15 upstream.

syzbot found that a thread can stall for minutes inside read_mem() or write_mem() after that thread was killed by SIGKILL [1]. Reading from iomem areas of /dev/mem can be slow, depending on the hardware. While reading 2GB at one read() is legal, delaying termination of killed thread for minutes is bad. Thus, allow reading/writing /dev/mem and /dev/kmem to be preemptible and killable.

  [ 1335.912419][T20577] read_mem: sz=4096 count=2134565632
  [ 1335.943194][T20577] read_mem: sz=4096 count=2134561536
  [ 1335.978280][T20577] read_mem: sz=4096 count=2134557440
  [ 1336.011147][T20577] read_mem: sz=4096 count=2134553344
  [ 1336.041897][T20577] read_mem: sz=4096 count=2134549248

Theoretically, reading/writing /dev/mem and /dev/kmem can become "interruptible". But this patch chose "killable". Future patch will make them "interruptible" so that we can revert to "killable" if some program regressed.

[1] https://syzkaller.appspot.com/bug?id=a0e3436829698d5824231251fad9d8e998f94f5e

---
 drivers/char/mem.c |   21 +++++++++++++++++++++
 1 file changed, 21 insertions(+)
drivers/char/mem.c +21 -0
@@ -97,6 +97,13 @@ void __weak unxlate_dev_mem_ptr(phys_add
 }
 #endif
 
+static inline bool should_stop_iteration(void)
+{
+	if (need_resched())
+		cond_resched();
+	return fatal_signal_pending(current);
+}
+
 /*
  * This funcion reads the *physical* memory. The f_pos points directly to the
  * memory location.
@@ -175,6 +182,8 @@ static ssize_t read_mem(struct file *fil
 		p += sz;
 		count -= sz;
 		read += sz;
+		if (should_stop_iteration())
+			break;
 	}
 	kfree(bounce);
 
@@ -251,6 +260,8 @@ static ssize_t write_mem(struct file *fi
 		p += sz;
 		count -= sz;
 		written += sz;
+		if (should_stop_iteration())
+			break;
 	}
 
 	*ppos += written;
@@ -468,6 +479,10 @@ static ssize_t read_kmem(struct file *fi
 			read += sz;
 			low_count -= sz;
 			count -= sz;
+			if (should_stop_iteration()) {
+				count = 0;
+				break;
+			}
 		}
 	}
 
@@ -492,6 +507,8 @@ static ssize_t read_kmem(struct file *fi
 			buf += sz;
 			read += sz;
 			p += sz;
+			if (should_stop_iteration())
+				break;
 		}
 		free_page((unsigned long)kbuf);
 	}
@@ -544,6 +561,8 @@ static ssize_t do_write_kmem(unsigned lo
 		p += sz;
 		count -= sz;
 		written += sz;
+		if (should_stop_iteration())
+			break;
 	}
 
 	*ppos += written;
@@ -595,6 +614,8 @@ static ssize_t write_kmem(struct file *f
 			buf += sz;
 			virtr += sz;
 			p += sz;
+			if (should_stop_iteration())
+				break;
 		}
 		free_page((unsigned long)kbuf);
 	}
author maintainer
#300 lore raw

PATCH 297/313 fs: Export generic_fadvise()

In reply to Greg Kroah-Hartman

From: Jan Kara <[email protected]>

commit cf1ea0592dbf109e7e7935b7d5b1a47a1ba04174 upstream.

Filesystems will need to call this function from their fadvise handlers.

---
 include/linux/fs.h |    2 ++
 mm/fadvise.c       |    4 ++--
 2 files changed, 4 insertions(+), 2 deletions(-)
include/linux/fs.h +2 -0
@@ -3544,6 +3544,8 @@ extern void inode_nohighmem(struct inode
 /* mm/fadvise.c */
 extern int vfs_fadvise(struct file *file, loff_t offset, loff_t len,
 		       int advice);
+extern int generic_fadvise(struct file *file, loff_t offset, loff_t len,
+			   int advice);
 
 #if defined(CONFIG_IO_URING)
 extern struct sock *io_uring_get_socket(struct file *file);
mm/fadvise.c +2 -2
@@ -27,8 +27,7 @@
  * deactivate the pages and clear PG_Referenced.
  */
 
-static int generic_fadvise(struct file *file, loff_t offset, loff_t len,
-			   int advice)
+int generic_fadvise(struct file *file, loff_t offset, loff_t len, int advice)
 {
 	struct inode *inode;
 	struct address_space *mapping;
@@ -178,6 +177,7 @@ static int generic_fadvise(struct file *
 	}
 	return 0;
 }
+EXPORT_SYMBOL(generic_fadvise);
 
 int vfs_fadvise(struct file *file, loff_t offset, loff_t len, int advice)
 {
author maintainer
#301 lore raw

PATCH 298/313 mm: Handle MADV_WILLNEED through vfs_fadvise()

In reply to Greg Kroah-Hartman

From: Jan Kara <[email protected]>

commit 692fe62433d4ca47605b39f7c416efd6679ba694 upstream.

Currently handling of MADV_WILLNEED hint calls directly into readahead code. Handle it by calling vfs_fadvise() instead so that filesystem can use its ->fadvise() callback to acquire necessary locks or otherwise prepare for the request.

---
 mm/madvise.c |   22 ++++++++++++++++------
 1 file changed, 16 insertions(+), 6 deletions(-)
mm/madvise.c +16 -6
@@ -14,6 +14,7 @@
 #include <linux/userfaultfd_k.h>
 #include <linux/hugetlb.h>
 #include <linux/falloc.h>
+#include <linux/fadvise.h>
 #include <linux/sched.h>
 #include <linux/ksm.h>
 #include <linux/fs.h>
@@ -275,6 +276,7 @@ static long madvise_willneed(struct vm_a
 			     unsigned long start, unsigned long end)
 {
 	struct file *file = vma->vm_file;
+	loff_t offset;
 
 	*prev = vma;
 #ifdef CONFIG_SWAP
@@ -298,12 +300,20 @@ static long madvise_willneed(struct vm_a
 		return 0;
 	}
 
-	start = ((start - vma->vm_start) >> PAGE_SHIFT) + vma->vm_pgoff;
-	if (end > vma->vm_end)
-		end = vma->vm_end;
-	end = ((end - vma->vm_start) >> PAGE_SHIFT) + vma->vm_pgoff;
-
-	force_page_cache_readahead(file->f_mapping, file, start, end - start);
+	/*
+	 * Filesystem's fadvise may need to take various locks.  We need to
+	 * explicitly grab a reference because the vma (and hence the
+	 * vma's reference to the file) can go away as soon as we drop
+	 * mmap_sem.
+	 */
+	*prev = NULL;	/* tell sys_madvise we drop mmap_sem */
+	get_file(file);
+	up_read(&current->mm->mmap_sem);
+	offset = (loff_t)(start - vma->vm_start)
+			+ ((loff_t)vma->vm_pgoff << PAGE_SHIFT);
+	vfs_fadvise(file, offset, end - start, POSIX_FADV_WILLNEED);
+	fput(file);
+	down_read(&current->mm->mmap_sem);
 	return 0;
 }