Skip to contentSkip to conversation

5.12.17-rc1 review

PATCH In review Greg Kroah-Hartman 9 participants 711 messages Jul 12, 2021 – Jul 13, 2021 Page 2 of 5

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

PATCH 139/700 hwrng: exynos - Fix runtime PM imbalance on error

In reply to Greg Kroah-Hartman

From: Łukasz Stelmach <[email protected]>

[ Upstream commit 0cdbabf8bb7a6147f5adf37dbc251e92a1bbc2c7 ]

pm_runtime_resume_and_get() wraps around pm_runtime_get_sync() and decrements the runtime PM usage counter in case the latter function fails and keeps the counter balanced.

---
 drivers/char/hw_random/exynos-trng.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/char/hw_random/exynos-trng.c +2 -2
index 8e1fe3f8dd2d..c8db62bc5ff7 100644
@@ -132,7 +132,7 @@ static int exynos_trng_probe(struct platform_device *pdev)
 		return PTR_ERR(trng->mem);
 
 	pm_runtime_enable(&pdev->dev);
-	ret = pm_runtime_get_sync(&pdev->dev);
+	ret = pm_runtime_resume_and_get(&pdev->dev);
 	if (ret < 0) {
 		dev_err(&pdev->dev, "Could not get runtime PM.\n");
 		goto err_pm_get;
@@ -165,7 +165,7 @@ err_register:
 	clk_disable_unprepare(trng->clk);
 
 err_clock:
-	pm_runtime_put_sync(&pdev->dev);
+	pm_runtime_put_noidle(&pdev->dev);
 
 err_pm_get:
 	pm_runtime_disable(&pdev->dev);
Signature · 1 line
2.30.2
author maintainer
#140 lore raw

PATCH 140/700 crypto: nx - add missing MODULE_DEVICE_TABLE

In reply to Greg Kroah-Hartman

From: Bixuan Cui <[email protected]>

[ Upstream commit 06676aa1f455c74e3ad1624cea3acb9ed2ef71ae ]

This patch adds missing MODULE_DEVICE_TABLE definition which generates correct modalias for automatic loading of this driver when it is built as an external module.

---
 drivers/crypto/nx/nx-842-pseries.c | 1 +
 1 file changed, 1 insertion(+)
drivers/crypto/nx/nx-842-pseries.c +1 -0
index cc8dd3072b8b..8ee547ee378e 100644
@@ -1069,6 +1069,7 @@ static const struct vio_device_id nx842_vio_driver_ids[] = {
 	{"ibm,compression-v1", "ibm,compression"},
 	{"", ""},
 };
+MODULE_DEVICE_TABLE(vio, nx842_vio_driver_ids);
 
 static struct vio_driver nx842_vio_driver = {
 	.name = KBUILD_MODNAME,
Signature · 1 line
2.30.2
author maintainer
#141 lore raw

PATCH 141/700 media: sti: fix obj-$(config) targets

In reply to Greg Kroah-Hartman

From: Mauro Carvalho Chehab <[email protected]>

[ Upstream commit 56c1f0876293888f686e31278d183d4af2cac3c3 ]

The right thing to do is to add a new object to the building system when a certain config option is selected, and *not* override them.

So, fix obj-$(config) logic at sti makefiles, using "+=", instead of ":=".

---
 drivers/media/platform/sti/bdisp/Makefile | 2 +-
 drivers/media/platform/sti/delta/Makefile | 2 +-
 drivers/media/platform/sti/hva/Makefile   | 2 +-
 3 files changed, 3 insertions(+), 3 deletions(-)
drivers/media/platform/sti/bdisp/Makefile +1 -1
index caf7ccd193ea..39ade0a34723 100644
@@ -1,4 +1,4 @@
 # SPDX-License-Identifier: GPL-2.0-only
-obj-$(CONFIG_VIDEO_STI_BDISP) := bdisp.o
+obj-$(CONFIG_VIDEO_STI_BDISP) += bdisp.o
 
 bdisp-objs := bdisp-v4l2.o bdisp-hw.o bdisp-debug.o
drivers/media/platform/sti/delta/Makefile +1 -1
index 92b37e216f00..32412fa4c632 100644
@@ -1,5 +1,5 @@
 # SPDX-License-Identifier: GPL-2.0-only
-obj-$(CONFIG_VIDEO_STI_DELTA_DRIVER) := st-delta.o
+obj-$(CONFIG_VIDEO_STI_DELTA_DRIVER) += st-delta.o
 st-delta-y := delta-v4l2.o delta-mem.o delta-ipc.o delta-debug.o
 
 # MJPEG support
drivers/media/platform/sti/hva/Makefile +1 -1
index 74b41ec52f97..b5a5478bdd01 100644
@@ -1,4 +1,4 @@
 # SPDX-License-Identifier: GPL-2.0-only
-obj-$(CONFIG_VIDEO_STI_HVA) := st-hva.o
+obj-$(CONFIG_VIDEO_STI_HVA) += st-hva.o
 st-hva-y := hva-v4l2.o hva-hw.o hva-mem.o hva-h264.o
 st-hva-$(CONFIG_VIDEO_STI_HVA_DEBUGFS) += hva-debugfs.o
Signature · 1 line
2.30.2
author maintainer
#142 lore raw

PATCH 142/700 sched: Make the idle task quack like a per-CPU kthread

In reply to Greg Kroah-Hartman

From: Valentin Schneider <[email protected]>

[ Upstream commit 00b89fe0197f0c55a045775c11553c0cdb7082fe ]

For all intents and purposes, the idle task is a per-CPU kthread. It isn't created via the same route as other pcpu kthreads however, and as a result it is missing a few bells and whistles: it fails kthread_is_per_cpu() and it doesn't have PF_NO_SETAFFINITY set.

Fix the former by giving the idle task a kthread struct along with the KTHREAD_IS_PER_CPU flag. This requires some extra iffery as init_idle() call be called more than once on the same idle task.

---
 include/linux/kthread.h |  2 ++
 kernel/kthread.c        | 30 ++++++++++++++++++------------
 kernel/sched/core.c     | 21 +++++++++++++++------
 3 files changed, 35 insertions(+), 18 deletions(-)
include/linux/kthread.h +2 -0
index 2484ed97e72f..d9133d6db308 100644
@@ -33,6 +33,8 @@ struct task_struct *kthread_create_on_cpu(int (*threadfn)(void *data),
 					  unsigned int cpu,
 					  const char *namefmt);
 
+void set_kthread_struct(struct task_struct *p);
+
 void kthread_set_per_cpu(struct task_struct *k, int cpu);
 bool kthread_is_per_cpu(struct task_struct *k);
 
kernel/kthread.c +18 -12
index 4fdf2bd9b558..e8da89c714f5 100644
@@ -68,16 +68,6 @@ enum KTHREAD_BITS {
 	KTHREAD_SHOULD_PARK,
 };
 
-static inline void set_kthread_struct(void *kthread)
-{
-	/*
-	 * We abuse ->set_child_tid to avoid the new member and because it
-	 * can't be wrongly copied by copy_process(). We also rely on fact
-	 * that the caller can't exec, so PF_KTHREAD can't be cleared.
-	 */
-	current->set_child_tid = (__force void __user *)kthread;
-}
-
 static inline struct kthread *to_kthread(struct task_struct *k)
 {
 	WARN_ON(!(k->flags & PF_KTHREAD));
@@ -103,6 +93,22 @@ static inline struct kthread *__to_kthread(struct task_struct *p)
 	return kthread;
 }
 
+void set_kthread_struct(struct task_struct *p)
+{
+	struct kthread *kthread;
+
+	if (__to_kthread(p))
+		return;
+
+	kthread = kzalloc(sizeof(*kthread), GFP_KERNEL);
+	/*
+	 * We abuse ->set_child_tid to avoid the new member and because it
+	 * can't be wrongly copied by copy_process(). We also rely on fact
+	 * that the caller can't exec, so PF_KTHREAD can't be cleared.
+	 */
+	p->set_child_tid = (__force void __user *)kthread;
+}
+
 void free_kthread_struct(struct task_struct *k)
 {
 	struct kthread *kthread;
@@ -272,8 +278,8 @@ static int kthread(void *_create)
 	struct kthread *self;
 	int ret;
 
-	self = kzalloc(sizeof(*self), GFP_KERNEL);
-	set_kthread_struct(self);
+	set_kthread_struct(current);
+	self = to_kthread(current);
 
 	/* If user was SIGKILLed, I release the structure. */
 	done = xchg(&create->done, NULL);
kernel/sched/core.c +15 -6
index e25b2d8ec18d..17f612045271 100644
@@ -7435,12 +7435,25 @@ void __init init_idle(struct task_struct *idle, int cpu)
 
 	__sched_fork(0, idle);
 
+	/*
+	 * The idle task doesn't need the kthread struct to function, but it
+	 * is dressed up as a per-CPU kthread and thus needs to play the part
+	 * if we want to avoid special-casing it in code that deals with per-CPU
+	 * kthreads.
+	 */
+	set_kthread_struct(idle);
+
 	raw_spin_lock_irqsave(&idle->pi_lock, flags);
 	raw_spin_lock(&rq->lock);
 
 	idle->state = TASK_RUNNING;
 	idle->se.exec_start = sched_clock();
-	idle->flags |= PF_IDLE;
+	/*
+	 * PF_KTHREAD should already be set at this point; regardless, make it
+	 * look like a proper per-CPU kthread.
+	 */
+	idle->flags |= PF_IDLE | PF_KTHREAD | PF_NO_SETAFFINITY;
+	kthread_set_per_cpu(idle, cpu);
 
 	scs_task_reset(idle);
 	kasan_unpoison_task_stack(idle);
@@ -7647,12 +7660,8 @@ static void balance_push(struct rq *rq)
 	/*
 	 * Both the cpu-hotplug and stop task are in this case and are
 	 * required to complete the hotplug process.
-	 *
-	 * XXX: the idle task does not match kthread_is_per_cpu() due to
-	 * histerical raisins.
 	 */
-	if (rq->idle == push_task ||
-	    kthread_is_per_cpu(push_task) ||
+	if (kthread_is_per_cpu(push_task) ||
 	    is_migration_disabled(push_task)) {
 
 		/*
Signature · 1 line
2.30.2
author maintainer
#143 lore raw

PATCH 143/700 media: cpia2: fix memory leak in cpia2_usb_probe

In reply to Greg Kroah-Hartman

From: Pavel Skripkin <[email protected]>

[ Upstream commit be8656e62e9e791837b606a027802b504a945c97 ]

syzbot reported leak in cpia2 usb driver. The problem was in invalid error handling.

v4l2_device_register() is called in cpia2_init_camera_struct(), but all error cases after cpia2_init_camera_struct() did not call the v4l2_device_unregister()

---
 drivers/media/usb/cpia2/cpia2.h      |  1 +
 drivers/media/usb/cpia2/cpia2_core.c | 12 ++++++++++++
 drivers/media/usb/cpia2/cpia2_usb.c  | 13 +++++++------
 3 files changed, 20 insertions(+), 6 deletions(-)
drivers/media/usb/cpia2/cpia2.h +1 -0
index 50835f5f7512..57b7f1ea68da 100644
@@ -429,6 +429,7 @@ int cpia2_send_command(struct camera_data *cam, struct cpia2_command *cmd);
 int cpia2_do_command(struct camera_data *cam,
 		     unsigned int command,
 		     unsigned char direction, unsigned char param);
+void cpia2_deinit_camera_struct(struct camera_data *cam, struct usb_interface *intf);
 struct camera_data *cpia2_init_camera_struct(struct usb_interface *intf);
 int cpia2_init_camera(struct camera_data *cam);
 int cpia2_allocate_buffers(struct camera_data *cam);
drivers/media/usb/cpia2/cpia2_core.c +12 -0
index e747548ab286..b5a2d06fb356 100644
@@ -2163,6 +2163,18 @@ static void reset_camera_struct(struct camera_data *cam)
 	cam->height = cam->params.roi.height;
 }
 
+/******************************************************************************
+ *
+ *  cpia2_init_camera_struct
+ *
+ *  Deinitialize camera struct
+ *****************************************************************************/
+void cpia2_deinit_camera_struct(struct camera_data *cam, struct usb_interface *intf)
+{
+	v4l2_device_unregister(&cam->v4l2_dev);
+	kfree(cam);
+}
+
 /******************************************************************************
  *
  *  cpia2_init_camera_struct
drivers/media/usb/cpia2/cpia2_usb.c +7 -6
index 3ab80a7b4498..76aac06f9fb8 100644
@@ -844,15 +844,13 @@ static int cpia2_usb_probe(struct usb_interface *intf,
 	ret = set_alternate(cam, USBIF_CMDONLY);
 	if (ret < 0) {
 		ERR("%s: usb_set_interface error (ret = %d)\n", __func__, ret);
-		kfree(cam);
-		return ret;
+		goto alt_err;
 	}
 
 
 	if((ret = cpia2_init_camera(cam)) < 0) {
 		ERR("%s: failed to initialize cpia2 camera (ret = %d)\n", __func__, ret);
-		kfree(cam);
-		return ret;
+		goto alt_err;
 	}
 	LOG("  CPiA Version: %d.%02d (%d.%d)\n",
 	       cam->params.version.firmware_revision_hi,
@@ -872,11 +870,14 @@ static int cpia2_usb_probe(struct usb_interface *intf,
 	ret = cpia2_register_camera(cam);
 	if (ret < 0) {
 		ERR("%s: Failed to register cpia2 camera (ret = %d)\n", __func__, ret);
-		kfree(cam);
-		return ret;
+		goto alt_err;
 	}
 
 	return 0;
+
+alt_err:
+	cpia2_deinit_camera_struct(cam, intf);
+	return ret;
 }
 
 /******************************************************************************
Signature · 1 line
2.30.2
author maintainer
#144 lore raw

PATCH 144/700 media: cobalt: fix race condition in setting HPD

In reply to Greg Kroah-Hartman

From: Hans Verkuil <[email protected]>

[ Upstream commit 3d37ef41bed0854805ab9af22c422267510e1344 ]

The cobalt_s_bit_sysctrl reads the old register value over PCI, then changes a bit and sets writes the new value to the register.

This is used among other things for setting the HPD output pin.

But if the HPD is changed for multiple inputs at the same time, then this causes a race condition where a stale value is read.

Serialize this function with a mutex.

---
 drivers/media/pci/cobalt/cobalt-driver.c | 1 +
 drivers/media/pci/cobalt/cobalt-driver.h | 7 ++++++-
 2 files changed, 7 insertions(+), 1 deletion(-)
drivers/media/pci/cobalt/cobalt-driver.c +1 -0
index 0695078ef812..1bd8bbe57a30 100644
@@ -667,6 +667,7 @@ static int cobalt_probe(struct pci_dev *pci_dev,
 		return -ENOMEM;
 	cobalt->pci_dev = pci_dev;
 	cobalt->instance = i;
+	mutex_init(&cobalt->pci_lock);
 
 	retval = v4l2_device_register(&pci_dev->dev, &cobalt->v4l2_dev);
 	if (retval) {
drivers/media/pci/cobalt/cobalt-driver.h +6 -1
index bca68572b324..12c33e035904 100644
@@ -251,6 +251,8 @@ struct cobalt {
 	int instance;
 	struct pci_dev *pci_dev;
 	struct v4l2_device v4l2_dev;
+	/* serialize PCI access in cobalt_s_bit_sysctrl() */
+	struct mutex pci_lock;
 
 	void __iomem *bar0, *bar1;
 
@@ -320,10 +322,13 @@ static inline u32 cobalt_g_sysctrl(struct cobalt *cobalt)
 static inline void cobalt_s_bit_sysctrl(struct cobalt *cobalt,
 					int bit, int val)
 {
-	u32 ctrl = cobalt_read_bar1(cobalt, COBALT_SYS_CTRL_BASE);
+	u32 ctrl;
 
+	mutex_lock(&cobalt->pci_lock);
+	ctrl = cobalt_read_bar1(cobalt, COBALT_SYS_CTRL_BASE);
 	cobalt_write_bar1(cobalt, COBALT_SYS_CTRL_BASE,
 			(ctrl & ~(1UL << bit)) | (val << bit));
+	mutex_unlock(&cobalt->pci_lock);
 }
 
 static inline u32 cobalt_g_sysstat(struct cobalt *cobalt)
Signature · 1 line
2.30.2
author maintainer
#145 lore raw

PATCH 145/700 media: hevc: Fix dependent slice segment flags

In reply to Greg Kroah-Hartman

From: Jernej Skrabec <[email protected]>

[ Upstream commit 67a7e53d5b21f3a84efc03a4e62db7caf97841ef ]

Dependent slice segment flag for PPS control is misnamed. It should have "enabled" at the end. It only tells if this flag is present in slice header or not and not the actual value.

Fix this by renaming the PPS flag and introduce another flag for slice control which tells actual value.

---
 Documentation/userspace-api/media/v4l/ext-ctrls-codec.rst | 5 ++++-
 drivers/staging/media/sunxi/cedrus/cedrus_h265.c          | 4 ++--
 include/media/hevc-ctrls.h                                | 3 ++-
 3 files changed, 8 insertions(+), 4 deletions(-)
Documentation/userspace-api/media/v4l/ext-ctrls-codec.rst +4 -1
index 00944e97d638..09f28ba60e6f 100644
@@ -3285,7 +3285,7 @@ enum v4l2_mpeg_video_hevc_size_of_length_field -
     :stub-columns: 0
     :widths:       1 1 2
 
-    * - ``V4L2_HEVC_PPS_FLAG_DEPENDENT_SLICE_SEGMENT``
+    * - ``V4L2_HEVC_PPS_FLAG_DEPENDENT_SLICE_SEGMENT_ENABLED``
       - 0x00000001
       -
     * - ``V4L2_HEVC_PPS_FLAG_OUTPUT_FLAG_PRESENT``
@@ -3493,6 +3493,9 @@ enum v4l2_mpeg_video_hevc_size_of_length_field -
     * - ``V4L2_HEVC_SLICE_PARAMS_FLAG_SLICE_LOOP_FILTER_ACROSS_SLICES_ENABLED``
       - 0x00000100
       -
+    * - ``V4L2_HEVC_SLICE_PARAMS_FLAG_DEPENDENT_SLICE_SEGMENT``
+      - 0x00000200
+      -
 
 .. c:type:: v4l2_hevc_dpb_entry
 
drivers/staging/media/sunxi/cedrus/cedrus_h265.c +2 -2
index ce497d0197df..10744fab7cea 100644
@@ -477,8 +477,8 @@ static void cedrus_h265_setup(struct cedrus_ctx *ctx,
 				slice_params->flags);
 
 	reg |= VE_DEC_H265_FLAG(VE_DEC_H265_DEC_SLICE_HDR_INFO0_FLAG_DEPENDENT_SLICE_SEGMENT,
-				V4L2_HEVC_PPS_FLAG_DEPENDENT_SLICE_SEGMENT,
-				pps->flags);
+				V4L2_HEVC_SLICE_PARAMS_FLAG_DEPENDENT_SLICE_SEGMENT,
+				slice_params->flags);
 
 	/* FIXME: For multi-slice support. */
 	reg |= VE_DEC_H265_DEC_SLICE_HDR_INFO0_FLAG_FIRST_SLICE_SEGMENT_IN_PIC;
include/media/hevc-ctrls.h +2 -1
index b4cb2ef02f17..226fcfa0e026 100644
@@ -81,7 +81,7 @@ struct v4l2_ctrl_hevc_sps {
 	__u64	flags;
 };
 
-#define V4L2_HEVC_PPS_FLAG_DEPENDENT_SLICE_SEGMENT		(1ULL << 0)
+#define V4L2_HEVC_PPS_FLAG_DEPENDENT_SLICE_SEGMENT_ENABLED	(1ULL << 0)
 #define V4L2_HEVC_PPS_FLAG_OUTPUT_FLAG_PRESENT			(1ULL << 1)
 #define V4L2_HEVC_PPS_FLAG_SIGN_DATA_HIDING_ENABLED		(1ULL << 2)
 #define V4L2_HEVC_PPS_FLAG_CABAC_INIT_PRESENT			(1ULL << 3)
@@ -160,6 +160,7 @@ struct v4l2_hevc_pred_weight_table {
 #define V4L2_HEVC_SLICE_PARAMS_FLAG_USE_INTEGER_MV		(1ULL << 6)
 #define V4L2_HEVC_SLICE_PARAMS_FLAG_SLICE_DEBLOCKING_FILTER_DISABLED (1ULL << 7)
 #define V4L2_HEVC_SLICE_PARAMS_FLAG_SLICE_LOOP_FILTER_ACROSS_SLICES_ENABLED (1ULL << 8)
+#define V4L2_HEVC_SLICE_PARAMS_FLAG_DEPENDENT_SLICE_SEGMENT	(1ULL << 9)
 
 struct v4l2_ctrl_hevc_slice_params {
 	__u32	bit_size;
Signature · 1 line
2.30.2
author maintainer
#146 lore raw

PATCH 146/700 media: pvrusb2: fix warning in pvr2_i2c_core_done

In reply to Greg Kroah-Hartman

From: Anirudh Rayabharam <[email protected]>

[ Upstream commit f8194e5e63fdcb349e8da9eef9e574d5b1d687cb ]

syzbot has reported the following warning in pvr2_i2c_done:

sysfs group 'power' not found for kobject '1-0043'

When the device is disconnected (pvr_hdw_disconnect), the i2c adapter is not unregistered along with the USB and v4l2 teardown. As part of the USB device disconnect, the sysfs files of the subdevices are also deleted. So, by the time pvr_i2c_core_done is called by pvr_context_destroy, the sysfs files have been deleted.

To fix this, unregister the i2c adapter too in pvr_hdw_disconnect. Make the device deregistration code shared by calling pvr_hdw_disconnect from pvr2_hdw_destroy.

---
 drivers/media/usb/pvrusb2/pvrusb2-hdw.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/media/usb/pvrusb2/pvrusb2-hdw.c +2 -2
index f4a727918e35..d38dee1792e4 100644
@@ -2676,9 +2676,8 @@ void pvr2_hdw_destroy(struct pvr2_hdw *hdw)
 		pvr2_stream_destroy(hdw->vid_stream);
 		hdw->vid_stream = NULL;
 	}
-	pvr2_i2c_core_done(hdw);
 	v4l2_device_unregister(&hdw->v4l2_dev);
-	pvr2_hdw_remove_usb_stuff(hdw);
+	pvr2_hdw_disconnect(hdw);
 	mutex_lock(&pvr2_unit_mtx);
 	do {
 		if ((hdw->unit_number >= 0) &&
@@ -2705,6 +2704,7 @@ void pvr2_hdw_disconnect(struct pvr2_hdw *hdw)
 {
 	pvr2_trace(PVR2_TRACE_INIT,"pvr2_hdw_disconnect(hdw=%p)",hdw);
 	LOCK_TAKE(hdw->big_lock);
+	pvr2_i2c_core_done(hdw);
 	LOCK_TAKE(hdw->ctl_lock);
 	pvr2_hdw_remove_usb_stuff(hdw);
 	LOCK_GIVE(hdw->ctl_lock);
Signature · 1 line
2.30.2
author maintainer
#147 lore raw

PATCH 147/700 media: imx: imx7_mipi_csis: Fix logging of only error event counters

In reply to Greg Kroah-Hartman

From: Laurent Pinchart <[email protected]>

[ Upstream commit d2fcc9c2de1191ea80366e3658711753738dd10a ]

The mipi_csis_events array ends with 6 non-error events, not 4. Update mipi_csis_log_counters() accordingly. While at it, log event counters in forward order, as there's no reason to log them backward.

---
 drivers/staging/media/imx/imx7-mipi-csis.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/staging/media/imx/imx7-mipi-csis.c +4 -2
index a01a7364b4b9..b365790256e4 100644
@@ -597,13 +597,15 @@ static void mipi_csis_clear_counters(struct csi_state *state)
 
 static void mipi_csis_log_counters(struct csi_state *state, bool non_errors)
 {
-	int i = non_errors ? MIPI_CSIS_NUM_EVENTS : MIPI_CSIS_NUM_EVENTS - 4;
+	unsigned int num_events = non_errors ? MIPI_CSIS_NUM_EVENTS
+				: MIPI_CSIS_NUM_EVENTS - 6;
 	struct device *dev = &state->pdev->dev;
 	unsigned long flags;
+	unsigned int i;
 
 	spin_lock_irqsave(&state->slock, flags);
 
-	for (i--; i >= 0; i--) {
+	for (i = 0; i < num_events; ++i) {
 		if (state->events[i].counter > 0 || state->debug)
 			dev_info(dev, "%s events: %d\n", state->events[i].name,
 				 state->events[i].counter);
Signature · 1 line
2.30.2
author maintainer
#148 lore raw

PATCH 148/700 crypto: qat - check return code of qat_hal_rd_rel_reg()

In reply to Greg Kroah-Hartman

From: Jack Xu <[email protected]>

[ Upstream commit 96b57229209490c8bca4335b01a426a96173dc56 ]

Check the return code of the function qat_hal_rd_rel_reg() and return it to the caller.

This is to fix the following warning when compiling the driver with clang scan-build:

drivers/crypto/qat/qat_common/qat_hal.c:1436:2: warning: 6th function call argument is an uninitialized value

---
 drivers/crypto/qat/qat_common/qat_hal.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/crypto/qat/qat_common/qat_hal.c +5 -1
index bd3028126cbe..069f51621f0e 100644
@@ -1417,7 +1417,11 @@ static int qat_hal_put_rel_wr_xfer(struct icp_qat_fw_loader_handle *handle,
 		pr_err("QAT: bad xfrAddr=0x%x\n", xfr_addr);
 		return -EINVAL;
 	}
-	qat_hal_rd_rel_reg(handle, ae, ctx, ICP_GPB_REL, gprnum, &gprval);
+	status = qat_hal_rd_rel_reg(handle, ae, ctx, ICP_GPB_REL, gprnum, &gprval);
+	if (status) {
+		pr_err("QAT: failed to read register");
+		return status;
+	}
 	gpr_addr = qat_hal_get_reg_addr(ICP_GPB_REL, gprnum);
 	data16low = 0xffff & data;
 	data16hi = 0xffff & (data >> 0x10);
Signature · 1 line
2.30.2
author maintainer
#149 lore raw

PATCH 149/700 crypto: qat - remove unused macro in FW loader

In reply to Greg Kroah-Hartman

From: Jack Xu <[email protected]>

[ Upstream commit 9afe77cf25d9670e61b489fd52cc6f75fd7f6803 ]

Remove the unused macro ICP_DH895XCC_PESRAM_BAR_SIZE in the firmware loader.

This is to fix the following warning when compiling the driver using the clang compiler with CC=clang W=2:

drivers/crypto/qat/qat_common/qat_uclo.c:345:9: warning: macro is not used [-Wunused-macros]

---
 drivers/crypto/qat/qat_common/qat_uclo.c | 1 -
 1 file changed, 1 deletion(-)
drivers/crypto/qat/qat_common/qat_uclo.c +0 -1
index 1fb5fc852f6b..6d95160e451e 100644
@@ -342,7 +342,6 @@ static int qat_uclo_init_umem_seg(struct icp_qat_fw_loader_handle *handle,
 	return 0;
 }
 
-#define ICP_DH895XCC_PESRAM_BAR_SIZE 0x80000
 static int qat_uclo_init_ae_memory(struct icp_qat_fw_loader_handle *handle,
 				   struct icp_qat_uof_initmem *init_mem)
 {
Signature · 1 line
2.30.2
author maintainer
#150 lore raw

PATCH 150/700 crypto: qce: skcipher: Fix incorrect sg count for dma transfers

In reply to Greg Kroah-Hartman

From: Thara Gopinath <[email protected]>

[ Upstream commit 1339a7c3ba05137a2d2fe75f602311bbfc6fab33 ]

Use the sg count returned by dma_map_sg to call into dmaengine_prep_slave_sg rather than using the original sg count. dma_map_sg can merge consecutive sglist entries, thus making the original sg count wrong. This is a fix for memory coruption issues observed while testing encryption/decryption of large messages using libkcapi framework.

Patch has been tested further by running full suite of tcrypt.ko tests including fuzz tests.

---
 drivers/crypto/qce/skcipher.c | 15 ++++++++-------
 1 file changed, 8 insertions(+), 7 deletions(-)
drivers/crypto/qce/skcipher.c +8 -7
index a2d3da0ad95f..5a6559131eac 100644
@@ -71,7 +71,7 @@ qce_skcipher_async_req_handle(struct crypto_async_request *async_req)
 	struct scatterlist *sg;
 	bool diff_dst;
 	gfp_t gfp;
-	int ret;
+	int dst_nents, src_nents, ret;
 
 	rctx->iv = req->iv;
 	rctx->ivsize = crypto_skcipher_ivsize(skcipher);
@@ -122,21 +122,22 @@ qce_skcipher_async_req_handle(struct crypto_async_request *async_req)
 	sg_mark_end(sg);
 	rctx->dst_sg = rctx->dst_tbl.sgl;
 
-	ret = dma_map_sg(qce->dev, rctx->dst_sg, rctx->dst_nents, dir_dst);
-	if (ret < 0)
+	dst_nents = dma_map_sg(qce->dev, rctx->dst_sg, rctx->dst_nents, dir_dst);
+	if (dst_nents < 0)
 		goto error_free;
 
 	if (diff_dst) {
-		ret = dma_map_sg(qce->dev, req->src, rctx->src_nents, dir_src);
-		if (ret < 0)
+		src_nents = dma_map_sg(qce->dev, req->src, rctx->src_nents, dir_src);
+		if (src_nents < 0)
 			goto error_unmap_dst;
 		rctx->src_sg = req->src;
 	} else {
 		rctx->src_sg = rctx->dst_sg;
+		src_nents = dst_nents - 1;
 	}
 
-	ret = qce_dma_prep_sgs(&qce->dma, rctx->src_sg, rctx->src_nents,
-			       rctx->dst_sg, rctx->dst_nents,
+	ret = qce_dma_prep_sgs(&qce->dma, rctx->src_sg, src_nents,
+			       rctx->dst_sg, dst_nents,
 			       qce_skcipher_done, async_req);
 	if (ret)
 		goto error_unmap_src;
Signature · 1 line
2.30.2
author maintainer
#151 lore raw

PATCH 151/700 arm64: perf: Convert snprintf to sysfs_emit

In reply to Greg Kroah-Hartman

From: Tian Tao <[email protected]>

[ Upstream commit a5740e955540181f4ab8f076cc9795c6bbe4d730 ]

Use sysfs_emit instead of snprintf to avoid buf overrun,because in sysfs_emit it strictly checks whether buf is null or buf whether pagesize aligned, otherwise it returns an error.

---
 arch/arm64/kernel/perf_event.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm64/kernel/perf_event.c +1 -1
index 4658fcf88c2b..3baca49fcf6b 100644
@@ -312,7 +312,7 @@ static ssize_t slots_show(struct device *dev, struct device_attribute *attr,
 	struct arm_pmu *cpu_pmu = container_of(pmu, struct arm_pmu, pmu);
 	u32 slots = cpu_pmu->reg_pmmir & ARMV8_PMU_SLOTS_MASK;
 
-	return snprintf(page, PAGE_SIZE, "0x%08x\n", slots);
+	return sysfs_emit(page, "0x%08x\n", slots);
 }
 
 static DEVICE_ATTR_RO(slots);
Signature · 1 line
2.30.2
author maintainer
#152 lore raw

PATCH 152/700 sched/fair: Fix ascii art by relpacing tabs

In reply to Greg Kroah-Hartman

From: Odin Ugedal <[email protected]>

[ Upstream commit 08f7c2f4d0e9f4283f5796b8168044c034a1bfcb ]

When using something other than 8 spaces per tab, this ascii art makes not sense, and the reader might end up wondering what this advanced equation "is".

---
 kernel/sched/fair.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
kernel/sched/fair.c +4 -4
index 47fcc3fe9dc5..272c583fc167 100644
@@ -3134,7 +3134,7 @@ void reweight_task(struct task_struct *p, int prio)
  *
  *                     tg->weight * grq->load.weight
  *   ge->load.weight = -----------------------------               (1)
- *			  \Sum grq->load.weight
+ *                       \Sum grq->load.weight
  *
  * Now, because computing that sum is prohibitively expensive to compute (been
  * there, done that) we approximate it with this average stuff. The average
@@ -3148,7 +3148,7 @@ void reweight_task(struct task_struct *p, int prio)
  *
  *                     tg->weight * grq->avg.load_avg
  *   ge->load.weight = ------------------------------              (3)
- *				tg->load_avg
+ *                             tg->load_avg
  *
  * Where: tg->load_avg ~= \Sum grq->avg.load_avg
  *
@@ -3164,7 +3164,7 @@ void reweight_task(struct task_struct *p, int prio)
  *
  *                     tg->weight * grq->load.weight
  *   ge->load.weight = ----------------------------- = tg->weight   (4)
- *			    grp->load.weight
+ *                         grp->load.weight
  *
  * That is, the sum collapses because all other CPUs are idle; the UP scenario.
  *
@@ -3183,7 +3183,7 @@ void reweight_task(struct task_struct *p, int prio)
  *
  *                     tg->weight * grq->load.weight
  *   ge->load.weight = -----------------------------		   (6)
- *				tg_load_avg'
+ *                             tg_load_avg'
  *
  * Where:
  *
Signature · 1 line
2.30.2
author maintainer
#153 lore raw

PATCH 153/700 ima: Dont remove security.ima if file must not be appraised

In reply to Greg Kroah-Hartman

From: Roberto Sassu <[email protected]>

[ Upstream commit ed1b472fc15aeaa20ddeeb93fd25190014e50d17 ]

Files might come from a remote source and might have xattrs, including security.ima. It should not be IMA task to decide whether security.ima should be kept or not. This patch removes the removexattr() system call in ima_inode_post_setattr().

---
 security/integrity/ima/ima_appraise.c | 2 --
 1 file changed, 2 deletions(-)
security/integrity/ima/ima_appraise.c +0 -2
index 565e33ff19d0..d7cc6f897746 100644
@@ -522,8 +522,6 @@ void ima_inode_post_setattr(struct user_namespace *mnt_userns,
 		return;
 
 	action = ima_must_appraise(mnt_userns, inode, MAY_ACCESS, POST_SETATTR);
-	if (!action)
-		__vfs_removexattr(&init_user_ns, dentry, XATTR_NAME_IMA);
 	iint = integrity_iint_find(inode);
 	if (iint) {
 		set_bit(IMA_CHANGE_ATTR, &iint->atomic_flags);
Signature · 1 line
2.30.2
author maintainer
#154 lore raw

PATCH 154/700 media: i2c: ov2659: Use clk_{prepare_enable,disable_unprepare}() to set xvclk on/off

In reply to Greg Kroah-Hartman

From: Dillon Min <[email protected]>

[ Upstream commit 24786ccd9c80fdb05494aa4d90fcb8f34295c193 ]

On some platform(imx6q), xvclk might not switch on in advance, also for power save purpose, xvclk should not be always on. so, add clk_prepare_enable(), clk_disable_unprepare() in driver side to set xvclk on/off at proper stage.

Add following changes:
- add 'struct clk *clk;' in 'struct ov2659 {}'
- enable xvclk in ov2659_power_on()
- disable xvclk in ov2659_power_off()

---
 drivers/media/i2c/ov2659.c | 24 ++++++++++++++++++------
 1 file changed, 18 insertions(+), 6 deletions(-)
drivers/media/i2c/ov2659.c +18 -6
index 42f64175a6df..fb78a1cedc03 100644
@@ -204,6 +204,7 @@ struct ov2659 {
 	struct i2c_client *client;
 	struct v4l2_ctrl_handler ctrls;
 	struct v4l2_ctrl *link_frequency;
+	struct clk *clk;
 	const struct ov2659_framesize *frame_size;
 	struct sensor_register *format_ctrl_regs;
 	struct ov2659_pll_ctrl pll;
@@ -1270,6 +1271,8 @@ static int ov2659_power_off(struct device *dev)
 
 	gpiod_set_value(ov2659->pwdn_gpio, 1);
 
+	clk_disable_unprepare(ov2659->clk);
+
 	return 0;
 }
 
@@ -1278,9 +1281,17 @@ static int ov2659_power_on(struct device *dev)
 	struct i2c_client *client = to_i2c_client(dev);
 	struct v4l2_subdev *sd = i2c_get_clientdata(client);
 	struct ov2659 *ov2659 = to_ov2659(sd);
+	int ret;
 
 	dev_dbg(&client->dev, "%s:\n", __func__);
 
+	ret = clk_prepare_enable(ov2659->clk);
+	if (ret) {
+		dev_err(&client->dev, "%s: failed to enable clock\n",
+			__func__);
+		return ret;
+	}
+
 	gpiod_set_value(ov2659->pwdn_gpio, 0);
 
 	if (ov2659->resetb_gpio) {
@@ -1425,7 +1436,6 @@ static int ov2659_probe(struct i2c_client *client)
 	const struct ov2659_platform_data *pdata = ov2659_get_pdata(client);
 	struct v4l2_subdev *sd;
 	struct ov2659 *ov2659;
-	struct clk *clk;
 	int ret;
 
 	if (!pdata) {
@@ -1440,11 +1450,11 @@ static int ov2659_probe(struct i2c_client *client)
 	ov2659->pdata = pdata;
 	ov2659->client = client;
 
-	clk = devm_clk_get(&client->dev, "xvclk");
-	if (IS_ERR(clk))
-		return PTR_ERR(clk);
+	ov2659->clk = devm_clk_get(&client->dev, "xvclk");
+	if (IS_ERR(ov2659->clk))
+		return PTR_ERR(ov2659->clk);
 
-	ov2659->xvclk_frequency = clk_get_rate(clk);
+	ov2659->xvclk_frequency = clk_get_rate(ov2659->clk);
 	if (ov2659->xvclk_frequency < 6000000 ||
 	    ov2659->xvclk_frequency > 27000000)
 		return -EINVAL;
@@ -1506,7 +1516,9 @@ static int ov2659_probe(struct i2c_client *client)
 	ov2659->frame_size = &ov2659_framesizes[2];
 	ov2659->format_ctrl_regs = ov2659_formats[0].format_ctrl_regs;
 
-	ov2659_power_on(&client->dev);
+	ret = ov2659_power_on(&client->dev);
+	if (ret < 0)
+		goto error;
 
 	ret = ov2659_detect(sd);
 	if (ret < 0)
Signature · 1 line
2.30.2
author maintainer
#155 lore raw

PATCH 155/700 media: bt878: do not schedule tasklet when it is not setup

In reply to Greg Kroah-Hartman

From: Tong Zhang <[email protected]>

[ Upstream commit a3a54bf4bddaecda8b5767209cfc703f0be2841d ]

There is a problem with the tasklet in bt878. bt->tasklet is set by dvb-bt8xx.ko, and bt878.ko can be loaded independently. In this case if interrupt comes it may cause null-ptr-dereference. To solve this issue, we check if the tasklet is actually set before calling tasklet_schedule.

[    1.750438] bt878(0): irq FDSR FBUS risc_pc=
[    1.750728] BUG: kernel NULL pointer dereference, address: 0000000000000000
[    1.752969] RIP: 0010:0x0
[    1.757526] Call Trace:
[    1.757659]  <IRQ>
[    1.757770]  tasklet_action_common.isra.0+0x107/0x110
[    1.758041]  tasklet_action+0x22/0x30
[    1.758237]  __do_softirq+0xe0/0x29b
[    1.758430]  irq_exit_rcu+0xa4/0xb0
[    1.758618]  common_interrupt+0x8d/0xa0
[    1.758824]  </IRQ>
---
 drivers/media/pci/bt8xx/bt878.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/media/pci/bt8xx/bt878.c +2 -1
index 78dd35c9b65d..7ca309121fb5 100644
@@ -300,7 +300,8 @@ static irqreturn_t bt878_irq(int irq, void *dev_id)
 		}
 		if (astat & BT878_ARISCI) {
 			bt->finished_block = (stat & BT878_ARISCS) >> 28;
-			tasklet_schedule(&bt->tasklet);
+			if (bt->tasklet.callback)
+				tasklet_schedule(&bt->tasklet);
 			break;
 		}
 		count++;
Signature · 1 line
2.30.2
author maintainer
#156 lore raw

PATCH 156/700 media: em28xx: Fix possible memory leak of em28xx struct

In reply to Greg Kroah-Hartman

From: Igor Matheus Andrade Torrente <[email protected]>

[ Upstream commit ac5688637144644f06ed1f3c6d4dd8bb7db96020 ]

The em28xx struct kref isn't being decreased after an error in the em28xx_ir_init, leading to a possible memory leak.

A kref_put and em28xx_shutdown_buttons is added to the error handler code.

---
 drivers/media/usb/em28xx/em28xx-input.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)
drivers/media/usb/em28xx/em28xx-input.c +6 -2
index 5aa15a7a49de..59529cbf9cd0 100644
@@ -720,7 +720,8 @@ static int em28xx_ir_init(struct em28xx *dev)
 			dev->board.has_ir_i2c = 0;
 			dev_warn(&dev->intf->dev,
 				 "No i2c IR remote control device found.\n");
-			return -ENODEV;
+			err = -ENODEV;
+			goto ref_put;
 		}
 	}
 
@@ -735,7 +736,7 @@ static int em28xx_ir_init(struct em28xx *dev)
 
 	ir = kzalloc(sizeof(*ir), GFP_KERNEL);
 	if (!ir)
-		return -ENOMEM;
+		goto ref_put;
 	rc = rc_allocate_device(RC_DRIVER_SCANCODE);
 	if (!rc)
 		goto error;
@@ -839,6 +840,9 @@ error:
 	dev->ir = NULL;
 	rc_free_device(rc);
 	kfree(ir);
+ref_put:
+	em28xx_shutdown_buttons(dev);
+	kref_put(&dev->ref, em28xx_free_device);
 	return err;
 }
 
Signature · 1 line
2.30.2
author maintainer
#157 lore raw

PATCH 157/700 media: hantro: Fix .buf_prepare

In reply to Greg Kroah-Hartman

From: Andrzej Pietrasiewicz <[email protected]>

[ Upstream commit 082aaecff35fbe1937531057911b1dd1fc6b496e ]

The driver should only set the payload on .buf_prepare if the buffer is CAPTURE type. If an OUTPUT buffer has a zero bytesused set by userspace then v4l2-core will set it to buffer length.

If we overwrite bytesused for OUTPUT buffers, too, then vb2_get_plane_payload() will return incorrect value which might be then written to hw registers by the driver in hantro_g1_h264_dec.c.

---
 drivers/staging/media/hantro/hantro_v4l2.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)
drivers/staging/media/hantro/hantro_v4l2.c +8 -1
index 1bc118e375a1..7ccc6405036a 100644
@@ -639,7 +639,14 @@ static int hantro_buf_prepare(struct vb2_buffer *vb)
 	ret = hantro_buf_plane_check(vb, pix_fmt);
 	if (ret)
 		return ret;
-	vb2_set_plane_payload(vb, 0, pix_fmt->plane_fmt[0].sizeimage);
+	/*
+	 * Buffer's bytesused must be written by driver for CAPTURE buffers.
+	 * (for OUTPUT buffers, if userspace passes 0 bytesused, v4l2-core sets
+	 * it to buffer length).
+	 */
+	if (V4L2_TYPE_IS_CAPTURE(vq->type))
+		vb2_set_plane_payload(vb, 0, pix_fmt->plane_fmt[0].sizeimage);
+
 	return 0;
 }
 
Signature · 1 line
2.30.2
author maintainer
#158 lore raw

PATCH 158/700 media: cedrus: Fix .buf_prepare

In reply to Greg Kroah-Hartman

From: Andrzej Pietrasiewicz <[email protected]>

[ Upstream commit d84b9202d712309840f8b5abee0ed272506563bd ]

The driver should only set the payload on .buf_prepare if the buffer is CAPTURE type. If an OUTPUT buffer has a zero bytesused set by userspace then v4l2-core will set it to buffer length.

If we overwrite bytesused for OUTPUT buffers, too, then vb2_get_plane_payload() will return incorrect value which might be then written to hw registers by the driver in cedrus_h264.c or cedrus_vp8.c.

---
 drivers/staging/media/sunxi/cedrus/cedrus_video.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)
drivers/staging/media/sunxi/cedrus/cedrus_video.c +7 -1
index b62eb8e84057..bf731caf2ed5 100644
@@ -457,7 +457,13 @@ static int cedrus_buf_prepare(struct vb2_buffer *vb)
 	if (vb2_plane_size(vb, 0) < pix_fmt->sizeimage)
 		return -EINVAL;
 
-	vb2_set_plane_payload(vb, 0, pix_fmt->sizeimage);
+	/*
+	 * Buffer's bytesused must be written by driver for CAPTURE buffers.
+	 * (for OUTPUT buffers, if userspace passes 0 bytesused, v4l2-core sets
+	 * it to buffer length).
+	 */
+	if (V4L2_TYPE_IS_CAPTURE(vq->type))
+		vb2_set_plane_payload(vb, 0, pix_fmt->sizeimage);
 
 	return 0;
 }
Signature · 1 line
2.30.2
author maintainer
#159 lore raw

PATCH 159/700 media: v4l2-core: Avoid the dangling pointer in v4l2_fh_release

In reply to Greg Kroah-Hartman

From: Lv Yunlong <[email protected]>

[ Upstream commit 7dd0c9e547b6924e18712b6b51aa3cba1896ee2c ]

A use after free bug caused by the dangling pointer filp->privitate_data in v4l2_fh_release.
See https://lore.kernel.org/patchwork/patch/1419058/.

My patch sets the dangling pointer to NULL to provide robust.

---
 drivers/media/v4l2-core/v4l2-fh.c | 1 +
 1 file changed, 1 insertion(+)
drivers/media/v4l2-core/v4l2-fh.c +1 -0
index 684574f58e82..90eec79ee995 100644
@@ -96,6 +96,7 @@ int v4l2_fh_release(struct file *filp)
 		v4l2_fh_del(fh);
 		v4l2_fh_exit(fh);
 		kfree(fh);
+		filp->private_data = NULL;
 	}
 	return 0;
 }
Signature · 1 line
2.30.2
author maintainer
#160 lore raw

PATCH 160/700 media: bt8xx: Fix a missing check bug in bt878_probe

In reply to Greg Kroah-Hartman

From: Zheyu Ma <[email protected]>

[ Upstream commit 1a4520090681853e6b850cbe54b27247a013e0e5 ]

In 'bt878_irq', the driver calls 'tasklet_schedule', but this tasklet is set in 'dvb_bt8xx_load_card' of another driver 'dvb-bt8xx'. However, this two drivers are separate. The user may not load the 'dvb-bt8xx' driver when loading the 'bt8xx' driver, that is, the tasklet has not been initialized when 'tasklet_schedule' is called, so it is necessary to check whether the tasklet is initialized in 'bt878_probe'.

Fix this by adding a check at the end of bt878_probe.

The KASAN's report reveals it:

BUG: unable to handle kernel NULL pointer dereference at 0000000000000000
PGD 800000006aab2067 P4D 800000006aab2067 PUD 6b2ea067 PMD 0
Oops: 0010 [#1] PREEMPT SMP KASAN PTI
CPU: 2 PID: 8724 Comm: syz-executor.0 Not tainted 4.19.177-
gdba4159c14ef-dirty #40
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.12.0-59-
gc9ba5276e321-prebuilt.qemu.org 04/01/2014
RIP: 0010:          (null)
Code: Bad RIP value.
RSP: 0018:ffff88806c287ea0 EFLAGS: 00010246
RAX: fffffbfff1b01774 RBX: dffffc0000000000 RCX: 0000000000000000
RDX: 0000000000000000 RSI: 1ffffffff1b01775 RDI: 0000000000000000
RBP: ffff88806c287f00 R08: fffffbfff1b01774 R09: fffffbfff1b01774
R10: 0000000000000001 R11: fffffbfff1b01773 R12: 0000000000000000
R13: ffff88806c29f530 R14: ffffffff8d80bb88 R15: ffffffff8d80bb90
FS:  00007f6b550e6700(0000) GS:ffff88806c280000(0000) knlGS:
0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: ffffffffffffffd6 CR3: 000000005ec98000 CR4: 00000000000006e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
 <IRQ>
 tasklet_action_common.isra.17+0x141/0x420 kernel/softirq.c:522
 tasklet_action+0x50/0x70 kernel/softirq.c:540
 __do_softirq+0x224/0x92c kernel/softirq.c:292
 invoke_softirq kernel/softirq.c:372 [inline]
 irq_exit+0x15a/0x180 kernel/softirq.c:412
 exiting_irq arch/x86/include/asm/apic.h:535 [inline]
 do_IRQ+0x123/0x1e0 arch/x86/kernel/irq.c:260
 common_interrupt+0xf/0xf arch/x86/entry/entry_64.S:670
 </IRQ>
RIP: 0010:__do_sys_interrupt kernel/sys.c:2593 [inline]
RIP: 0010:__se_sys_interrupt kernel/sys.c:2584 [inline]
RIP: 0010:__x64_sys_interrupt+0x5b/0x80 kernel/sys.c:2584
Code: ba 00 04 00 00 48 c7 c7 c0 99 31 8c e8 ae 76 5e 01 48 85 c0 75 21 e8

14 ae 24 00 48 c7 c3 c0 99 31 8c b8 0c 00 00 00 0f 01 c1 <31> db e8 fe ad 24 00 48 89 d8 5b 5d c3 48 c7 c3 ea ff ff ff eb ec
RSP: 0018:ffff888054167f10 EFLAGS: 00000212 ORIG_RAX: ffffffffffffffde
RAX: 000000000000000c RBX: ffffffff8c3199c0 RCX: ffffc90001ca6000
RDX: 000000000000001a RSI: ffffffff813478fc RDI: ffffffff8c319dc0
RBP: ffff888054167f18 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000080 R11: fffffbfff18633b7 R12: ffff888054167f58
R13: ffff88805f638000 R14: 0000000000000000 R15: 0000000000000000

 do_syscall_64+0xb0/0x4e0 arch/x86/entry/common.c:293
 entry_SYSCALL_64_after_hwframe+0x49/0xbe
RIP: 0033:0x4692a9
Code: f7 d8 64 89 02 b8 ff ff ff ff c3 66 0f 1f 44 00 00 48 89 f8 48 89 f7

48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 bc ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007f6b550e5c48 EFLAGS: 00000246 ORIG_RAX: 000000000000014f
RAX: ffffffffffffffda RBX: 000000000077bf60 RCX: 00000000004692a9
RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000020000140
RBP: 00000000004cf7eb R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 000000000077bf60
R13: 0000000000000000 R14: 000000000077bf60 R15: 00007fff55a1dca0

Modules linked in:
Dumping ftrace buffer:
   (ftrace buffer empty)
CR2: 0000000000000000
---[ end trace 68e5849c3f77cbb6 ]---
RIP: 0010:          (null)
Code: Bad RIP value.
RSP: 0018:ffff88806c287ea0 EFLAGS: 00010246
RAX: fffffbfff1b01774 RBX: dffffc0000000000 RCX: 0000000000000000
RDX: 0000000000000000 RSI: 1ffffffff1b01775 RDI: 0000000000000000
RBP: ffff88806c287f00 R08: fffffbfff1b01774 R09: fffffbfff1b01774
R10: 0000000000000001 R11: fffffbfff1b01773 R12: 0000000000000000
R13: ffff88806c29f530 R14: ffffffff8d80bb88 R15: ffffffff8d80bb90
FS:  00007f6b550e6700(0000) GS:ffff88806c280000(0000) knlGS:
0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: ffffffffffffffd6 CR3: 000000005ec98000 CR4: 00000000000006e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
---
 drivers/media/pci/bt8xx/bt878.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/media/pci/bt8xx/bt878.c +3 -0
index 7ca309121fb5..90972d6952f1 100644
@@ -478,6 +478,9 @@ static int bt878_probe(struct pci_dev *dev, const struct pci_device_id *pci_id)
 	btwrite(0, BT878_AINT_MASK);
 	bt878_num++;
 
+	if (!bt->tasklet.func)
+		tasklet_disable(&bt->tasklet);
+
 	return 0;
 
       fail2:
Signature · 1 line
2.30.2
author maintainer
#161 lore raw

PATCH 161/700 media: st-hva: Fix potential NULL pointer dereferences

In reply to Greg Kroah-Hartman

From: Evgeny Novikov <[email protected]>

[ Upstream commit b7fdd208687ba59ebfb09b2199596471c63b69e3 ]

When ctx_id >= HVA_MAX_INSTANCES in hva_hw_its_irq_thread() it tries to access fields of ctx that is NULL at that point. The patch gets rid of these accesses.

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

---
 drivers/media/platform/sti/hva/hva-hw.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
drivers/media/platform/sti/hva/hva-hw.c +1 -2
index f59811e27f51..6eeee5017fac 100644
@@ -130,8 +130,7 @@ static irqreturn_t hva_hw_its_irq_thread(int irq, void *arg)
 	ctx_id = (hva->sts_reg & 0xFF00) >> 8;
 	if (ctx_id >= HVA_MAX_INSTANCES) {
 		dev_err(dev, "%s     %s: bad context identifier: %d\n",
-			ctx->name, __func__, ctx_id);
-		ctx->hw_err = true;
+			HVA_PREFIX, __func__, ctx_id);
 		goto out;
 	}
 
Signature · 1 line
2.30.2
author maintainer
#162 lore raw

PATCH 162/700 crypto: hisilicon/sec - fixup 3des minimum key size declaration

In reply to Greg Kroah-Hartman

From: Kai Ye <[email protected]>

[ Upstream commit 6161f40c630bd7ced5f236cd5fbabec06e47afae ]

Fixup the 3des algorithm minimum key size declaration.

---
 drivers/crypto/hisilicon/sec2/sec_crypto.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/crypto/hisilicon/sec2/sec_crypto.c +2 -2
index 8adcbb327126..c86b01abd0a6 100644
@@ -1513,11 +1513,11 @@ static struct skcipher_alg sec_skciphers[] = {
 			 AES_BLOCK_SIZE, AES_BLOCK_SIZE)
 
 	SEC_SKCIPHER_ALG("ecb(des3_ede)", sec_setkey_3des_ecb,
-			 SEC_DES3_2KEY_SIZE, SEC_DES3_3KEY_SIZE,
+			 SEC_DES3_3KEY_SIZE, SEC_DES3_3KEY_SIZE,
 			 DES3_EDE_BLOCK_SIZE, 0)
 
 	SEC_SKCIPHER_ALG("cbc(des3_ede)", sec_setkey_3des_cbc,
-			 SEC_DES3_2KEY_SIZE, SEC_DES3_3KEY_SIZE,
+			 SEC_DES3_3KEY_SIZE, SEC_DES3_3KEY_SIZE,
 			 DES3_EDE_BLOCK_SIZE, DES3_EDE_BLOCK_SIZE)
 
 	SEC_SKCIPHER_ALG("xts(sm4)", sec_setkey_sm4_xts,
Signature · 1 line
2.30.2
author maintainer
#163 lore raw

PATCH 163/700 Makefile: fix GDB warning with CONFIG_RELR

In reply to Greg Kroah-Hartman

From: Nick Desaulniers <[email protected]>

[ Upstream commit 27f2a4db76e8d8a8b601fc1c6a7a17f88bd907ab ]

GDB produces the following warning when debugging kernels built with CONFIG_RELR:

BFD: /android0/linux-next/vmlinux: unknown type [0x13] section `.relr.dyn'

when loading a kernel built with CONFIG_RELR into GDB. It can also prevent debugging symbols using such relocations.

Peter sugguests:
[That flag] means that lld will use dynamic tags and section type numbers in the OS-specific range rather than the generic range. The kernel itself doesn't care about these numbers; it determines the location of the RELR section using symbols defined by a linker script.

---
 Makefile                      | 2 +-
 scripts/tools-support-relr.sh | 3 ++-
 2 files changed, 3 insertions(+), 2 deletions(-)
Makefile +1 -1
index bf6accb2328c..6682790891df 100644
@@ -1006,7 +1006,7 @@ LDFLAGS_vmlinux	+= $(call ld-option, -X,)
 endif
 
 ifeq ($(CONFIG_RELR),y)
-LDFLAGS_vmlinux	+= --pack-dyn-relocs=relr
+LDFLAGS_vmlinux	+= --pack-dyn-relocs=relr --use-android-relr-tags
 endif
 
 # We never want expected sections to be placed heuristically by the
scripts/tools-support-relr.sh +2 -1
index 45e8aa360b45..cb55878bd5b8 100755
@@ -7,7 +7,8 @@ trap "rm -f $tmp_file.o $tmp_file $tmp_file.bin" EXIT
 cat << "END" | $CC -c -x c - -o $tmp_file.o >/dev/null 2>&1
 void *p = &p;
 END
-$LD $tmp_file.o -shared -Bsymbolic --pack-dyn-relocs=relr -o $tmp_file
+$LD $tmp_file.o -shared -Bsymbolic --pack-dyn-relocs=relr \
+  --use-android-relr-tags -o $tmp_file
 
 # Despite printing an error message, GNU nm still exits with exit code 0 if it
 # sees a relr section. So we need to check that nothing is printed to stderr.
Signature · 1 line
2.30.2
author maintainer
#164 lore raw

PATCH 164/700 media: dvd_usb: memory leak in cinergyt2_fe_attach

In reply to Greg Kroah-Hartman

From: Dongliang Mu <[email protected]>

[ Upstream commit 9ad1efee086e0e913914fa2b2173efb830bad68c ]

When the driver fails to talk with the hardware with dvb_usb_generic_rw, it will return an error to dvb_usb_adapter_frontend_init. However, the driver forgets to free the resource (e.g., struct cinergyt2_fe_state), which leads to a memory leak.

Fix this by freeing struct cinergyt2_fe_state when dvb_usb_generic_rw fails in cinergyt2_frontend_attach.

backtrace:

  [<0000000056e17b1a>] kmalloc include/linux/slab.h:552 [inline]
  [<0000000056e17b1a>] kzalloc include/linux/slab.h:682 [inline]
  [<0000000056e17b1a>] cinergyt2_fe_attach+0x21/0x80 drivers/media/usb/dvb-usb/cinergyT2-fe.c:271
  [<00000000ae0b1711>] cinergyt2_frontend_attach+0x21/0x70 drivers/media/usb/dvb-usb/cinergyT2-core.c:74
  [<00000000d0254861>] dvb_usb_adapter_frontend_init+0x11b/0x1b0 drivers/media/usb/dvb-usb/dvb-usb-dvb.c:290
  [<0000000002e08ac6>] dvb_usb_adapter_init drivers/media/usb/dvb-usb/dvb-usb-init.c:84 [inline]
  [<0000000002e08ac6>] dvb_usb_init drivers/media/usb/dvb-usb/dvb-usb-init.c:173 [inline]
  [<0000000002e08ac6>] dvb_usb_device_init.cold+0x4d0/0x6ae drivers/media/usb/dvb-usb/dvb-usb-init.c:287
---
 drivers/media/usb/dvb-usb/cinergyT2-core.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/media/usb/dvb-usb/cinergyT2-core.c +2 -0
index 969a7ec71dff..4116ba5c45fc 100644
@@ -78,6 +78,8 @@ static int cinergyt2_frontend_attach(struct dvb_usb_adapter *adap)
 
 	ret = dvb_usb_generic_rw(d, st->data, 1, st->data, 3, 0);
 	if (ret < 0) {
+		if (adap->fe_adap[0].fe)
+			adap->fe_adap[0].fe->ops.release(adap->fe_adap[0].fe);
 		deb_rc("cinergyt2_power_ctrl() Failed to retrieve sleep state info\n");
 	}
 	mutex_unlock(&d->data_mutex);
Signature · 1 line
2.30.2
author maintainer
#165 lore raw

PATCH 165/700 memstick: rtsx_usb_ms: fix UAF

In reply to Greg Kroah-Hartman

From: Tong Zhang <[email protected]>

[ Upstream commit 42933c8aa14be1caa9eda41f65cde8a3a95d3e39 ]

This patch fixes the following issues:
1. memstick_free_host() will free the host, so the use of ms_dev(host) after it will be a problem. To fix this, move memstick_free_host() after when we are done with ms_dev(host).
2. In rtsx_usb_ms_drv_remove(), pm need to be disabled before we remove and free host otherwise memstick_check will be called and UAF will happen.

[   11.351173] BUG: KASAN: use-after-free in rtsx_usb_ms_drv_remove+0x94/0x140 [rtsx_usb_ms]
[   11.357077]  rtsx_usb_ms_drv_remove+0x94/0x140 [rtsx_usb_ms]
[   11.357376]  platform_remove+0x2a/0x50
[   11.367531] Freed by task 298:
[   11.368537]  kfree+0xa4/0x2a0
[   11.368711]  device_release+0x51/0xe0
[   11.368905]  kobject_put+0xa2/0x120
[   11.369090]  rtsx_usb_ms_drv_remove+0x8c/0x140 [rtsx_usb_ms]
[   11.369386]  platform_remove+0x2a/0x50
[   12.038408] BUG: KASAN: use-after-free in __mutex_lock.isra.0+0x3ec/0x7c0
[   12.045432]  mutex_lock+0xc9/0xd0
[   12.046080]  memstick_check+0x6a/0x578 [memstick]
[   12.046509]  process_one_work+0x46d/0x750
[   12.052107] Freed by task 297:
[   12.053115]  kfree+0xa4/0x2a0
[   12.053272]  device_release+0x51/0xe0
[   12.053463]  kobject_put+0xa2/0x120
[   12.053647]  rtsx_usb_ms_drv_remove+0xc4/0x140 [rtsx_usb_ms]
[   12.053939]  platform_remove+0x2a/0x50
---
 drivers/memstick/host/rtsx_usb_ms.c | 10 ++++------
 1 file changed, 4 insertions(+), 6 deletions(-)
drivers/memstick/host/rtsx_usb_ms.c +4 -6
index 102dbb8080da..29271ad4728a 100644
@@ -799,9 +799,9 @@ static int rtsx_usb_ms_drv_probe(struct platform_device *pdev)
 
 	return 0;
 err_out:
-	memstick_free_host(msh);
 	pm_runtime_disable(ms_dev(host));
 	pm_runtime_put_noidle(ms_dev(host));
+	memstick_free_host(msh);
 	return err;
 }
 
@@ -828,9 +828,6 @@ static int rtsx_usb_ms_drv_remove(struct platform_device *pdev)
 	}
 	mutex_unlock(&host->host_mutex);
 
-	memstick_remove_host(msh);
-	memstick_free_host(msh);
-
 	/* Balance possible unbalanced usage count
 	 * e.g. unconditional module removal
 	 */
@@ -838,10 +835,11 @@ static int rtsx_usb_ms_drv_remove(struct platform_device *pdev)
 		pm_runtime_put(ms_dev(host));
 
 	pm_runtime_disable(ms_dev(host));
-	platform_set_drvdata(pdev, NULL);
-
+	memstick_remove_host(msh);
 	dev_dbg(ms_dev(host),
 		": Realtek USB Memstick controller has been removed\n");
+	memstick_free_host(msh);
+	platform_set_drvdata(pdev, NULL);
 
 	return 0;
 }
Signature · 1 line
2.30.2
author maintainer
#166 lore raw

PATCH 166/700 mmc: sdhci-sprd: use sdhci_sprd_writew

In reply to Greg Kroah-Hartman

From: Krzysztof Kozlowski <[email protected]>

[ Upstream commit 961470820021e6f9d74db4837bd6831a1a30341b ]

The sdhci_sprd_writew() was defined by never used in sdhci_ops:

drivers/mmc/host/sdhci-sprd.c:134:20: warning: unused function 'sdhci_sprd_writew'

---
 drivers/mmc/host/sdhci-sprd.c | 1 +
 1 file changed, 1 insertion(+)
drivers/mmc/host/sdhci-sprd.c +1 -0
index 5dc36efff47f..11e375579cfb 100644
@@ -393,6 +393,7 @@ static void sdhci_sprd_request_done(struct sdhci_host *host,
 static struct sdhci_ops sdhci_sprd_ops = {
 	.read_l = sdhci_sprd_readl,
 	.write_l = sdhci_sprd_writel,
+	.write_w = sdhci_sprd_writew,
 	.write_b = sdhci_sprd_writeb,
 	.set_clock = sdhci_sprd_set_clock,
 	.get_max_clock = sdhci_sprd_get_max_clock,
Signature · 1 line
2.30.2
author maintainer
#167 lore raw

PATCH 167/700 mmc: via-sdmmc: add a check against NULL pointer dereference

In reply to Greg Kroah-Hartman

From: Zheyu Ma <[email protected]>

[ Upstream commit 45c8ddd06c4b729c56a6083ab311bfbd9643f4a6 ]

Before referencing 'host->data', the driver needs to check whether it is null pointer, otherwise it will cause a null pointer reference.

This log reveals it:

[   29.355199] BUG: kernel NULL pointer dereference, address:
0000000000000014
[   29.357323] #PF: supervisor write access in kernel mode
[   29.357706] #PF: error_code(0x0002) - not-present page
[   29.358088] PGD 0 P4D 0
[   29.358280] Oops: 0002 [#1] PREEMPT SMP PTI
[   29.358595] CPU: 2 PID: 0 Comm: swapper/2 Not tainted 5.12.4-
g70e7f0549188-dirty #102
[   29.359164] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009),
BIOS rel-1.12.0-59-gc9ba5276e321-prebuilt.qemu.org 04/01/2014
[   29.359978] RIP: 0010:via_sdc_isr+0x21f/0x410
[   29.360314] Code: ff ff e8 84 aa d0 fd 66 45 89 7e 28 66 41 f7 c4 00
10 75 56 e8 72 aa d0 fd 66 41 f7 c4 00 c0 74 10 e8 65 aa d0 fd 48 8b 43
18 <c7> 40 14 ac ff ff ff e8 55 aa d0 fd 48 89 df e8 ad fb ff ff e9 77
[   29.361661] RSP: 0018:ffffc90000118e98 EFLAGS: 00010046
[   29.362042] RAX: 0000000000000000 RBX: ffff888107d77880
RCX: 0000000000000000
[   29.362564] RDX: 0000000000000000 RSI: ffffffff835d20bb
RDI: 00000000ffffffff
[   29.363085] RBP: ffffc90000118ed8 R08: 0000000000000001
R09: 0000000000000001
[   29.363604] R10: 0000000000000000 R11: 0000000000000001
R12: 0000000000008600
[   29.364128] R13: ffff888107d779c8 R14: ffffc90009c00200
R15: 0000000000008000
[   29.364651] FS:  0000000000000000(0000) GS:ffff88817bc80000(0000)
knlGS:0000000000000000
[   29.365235] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[   29.365655] CR2: 0000000000000014 CR3: 0000000005a2e000
CR4: 00000000000006e0
[   29.366170] DR0: 0000000000000000 DR1: 0000000000000000
DR2: 0000000000000000
[   29.366683] DR3: 0000000000000000 DR6: 00000000fffe0ff0
DR7: 0000000000000400
[   29.367197] Call Trace:
[   29.367381]  <IRQ>
[   29.367537]  __handle_irq_event_percpu+0x53/0x3e0
[   29.367916]  handle_irq_event_percpu+0x35/0x90
[   29.368247]  handle_irq_event+0x39/0x60
[   29.368632]  handle_fasteoi_irq+0xc2/0x1d0
[   29.368950]  __common_interrupt+0x7f/0x150
[   29.369254]  common_interrupt+0xb4/0xd0
[   29.369547]  </IRQ>
[   29.369708]  asm_common_interrupt+0x1e/0x40
[   29.370016] RIP: 0010:native_safe_halt+0x17/0x20
[   29.370360] Code: 07 0f 00 2d db 80 43 00 f4 5d c3 0f 1f 84 00 00 00
00 00 8b 05 c2 37 e5 01 55 48 89 e5 85 c0 7e 07 0f 00 2d bb 80 43 00 fb
f4 <5d> c3 cc cc cc cc cc cc cc 55 48 89 e5 e8 67 53 ff ff 8b 0d f9 91
[   29.371696] RSP: 0018:ffffc9000008fe90 EFLAGS: 00000246
[   29.372079] RAX: 0000000000000000 RBX: 0000000000000002
RCX: 0000000000000000
[   29.372595] RDX: 0000000000000000 RSI: ffffffff854f67a4
RDI: ffffffff85403406
[   29.373122] RBP: ffffc9000008fe90 R08: 0000000000000001
R09: 0000000000000001
[   29.373646] R10: 0000000000000000 R11: 0000000000000001
R12: ffffffff86009188
[   29.374160] R13: 0000000000000000 R14: 0000000000000000
R15: ffff888100258000
[   29.374690]  default_idle+0x9/0x10
[   29.374944]  arch_cpu_idle+0xa/0x10
[   29.375198]  default_idle_call+0x6e/0x250
[   29.375491]  do_idle+0x1f0/0x2d0
[   29.375740]  cpu_startup_entry+0x18/0x20
[   29.376034]  start_secondary+0x11f/0x160
[   29.376328]  secondary_startup_64_no_verify+0xb0/0xbb
[   29.376705] Modules linked in:
[   29.376939] Dumping ftrace buffer:
[   29.377187]    (ftrace buffer empty)
[   29.377460] CR2: 0000000000000014
[   29.377712] ---[ end trace 51a473dffb618c47 ]---
[   29.378056] RIP: 0010:via_sdc_isr+0x21f/0x410
[   29.378380] Code: ff ff e8 84 aa d0 fd 66 45 89 7e 28 66 41 f7 c4 00
10 75 56 e8 72 aa d0 fd 66 41 f7 c4 00 c0 74 10 e8 65 aa d0 fd 48 8b 43
18 <c7> 40 14 ac ff ff ff e8 55 aa d0 fd 48 89 df e8 ad fb ff ff e9 77
[   29.379714] RSP: 0018:ffffc90000118e98 EFLAGS: 00010046
[   29.380098] RAX: 0000000000000000 RBX: ffff888107d77880
RCX: 0000000000000000
[   29.380614] RDX: 0000000000000000 RSI: ffffffff835d20bb
RDI: 00000000ffffffff
[   29.381134] RBP: ffffc90000118ed8 R08: 0000000000000001
R09: 0000000000000001
[   29.381653] R10: 0000000000000000 R11: 0000000000000001
R12: 0000000000008600
[   29.382176] R13: ffff888107d779c8 R14: ffffc90009c00200
R15: 0000000000008000
[   29.382697] FS:  0000000000000000(0000) GS:ffff88817bc80000(0000)
knlGS:0000000000000000
[   29.383277] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[   29.383697] CR2: 0000000000000014 CR3: 0000000005a2e000
CR4: 00000000000006e0
[   29.384223] DR0: 0000000000000000 DR1: 0000000000000000
DR2: 0000000000000000
[   29.384736] DR3: 0000000000000000 DR6: 00000000fffe0ff0
DR7: 0000000000000400
[   29.385260] Kernel panic - not syncing: Fatal exception in interrupt
[   29.385882] Dumping ftrace buffer:
[   29.386135]    (ftrace buffer empty)
[   29.386401] Kernel Offset: disabled
[   29.386656] Rebooting in 1 seconds..
---
 drivers/mmc/host/via-sdmmc.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/mmc/host/via-sdmmc.c +3 -0
index 4f4c0813f9fd..350e67056fa6 100644
@@ -857,6 +857,9 @@ static void via_sdc_data_isr(struct via_crdr_mmc_host *host, u16 intmask)
 {
 	BUG_ON(intmask == 0);
 
+	if (!host->data)
+		return;
+
 	if (intmask & VIA_CRDR_SDSTS_DT)
 		host->data->error = -ETIMEDOUT;
 	else if (intmask & (VIA_CRDR_SDSTS_RC | VIA_CRDR_SDSTS_WC))
Signature · 1 line
2.30.2
author maintainer
#168 lore raw

PATCH 168/700 mmc: sdhci-of-aspeed: Turn down a phase correction warning

In reply to Greg Kroah-Hartman

From: Andrew Jeffery <[email protected]>

[ Upstream commit a7ab186f60785850b5af1be183867000485ad491 ]

The card timing and the bus frequency are not changed atomically with respect to calls to the set_clock() callback in the driver. The result is the driver sees a transient state where there's a mismatch between the two and thus the inputs to the phase correction calculation formula are garbage.

Switch from dev_warn() to dev_dbg() to avoid noise in the normal case, though the change does make bad configurations less likely to be noticed.

---
 drivers/mmc/host/sdhci-of-aspeed.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/mmc/host/sdhci-of-aspeed.c +1 -1
index 7d8692e90996..b6ac2af199b8 100644
@@ -150,7 +150,7 @@ static int aspeed_sdhci_phase_to_tap(struct device *dev, unsigned long rate_hz,
 
 	tap = div_u64(phase_period_ps, prop_delay_ps);
 	if (tap > ASPEED_SDHCI_NR_TAPS) {
-		dev_warn(dev,
+		dev_dbg(dev,
 			 "Requested out of range phase tap %d for %d degrees of phase compensation at %luHz, clamping to tap %d\n",
 			 tap, phase_deg, rate_hz, ASPEED_SDHCI_NR_TAPS);
 		tap = ASPEED_SDHCI_NR_TAPS;
Signature · 1 line
2.30.2
author maintainer
#169 lore raw

PATCH 169/700 spi: meson-spicc: fix a wrong goto jump for avoiding memory leak.

In reply to Greg Kroah-Hartman

From: zpershuai <[email protected]>

[ Upstream commit 95730d5eb73170a6d225a9998c478be273598634 ]

In meson_spifc_probe function, when enable the device pclk clock is error, it should use clk_disable_unprepare to release the core clock.

---
 drivers/spi/spi-meson-spicc.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/spi/spi-meson-spicc.c +4 -2
index ecba6b4a5d85..51aef2c6e966 100644
@@ -725,7 +725,7 @@ static int meson_spicc_probe(struct platform_device *pdev)
 	ret = clk_prepare_enable(spicc->pclk);
 	if (ret) {
 		dev_err(&pdev->dev, "pclk clock enable failed\n");
-		goto out_master;
+		goto out_core_clk;
 	}
 
 	device_reset_optional(&pdev->dev);
@@ -764,9 +764,11 @@ static int meson_spicc_probe(struct platform_device *pdev)
 	return 0;
 
 out_clk:
-	clk_disable_unprepare(spicc->core);
 	clk_disable_unprepare(spicc->pclk);
 
+out_core_clk:
+	clk_disable_unprepare(spicc->core);
+
 out_master:
 	spi_master_put(master);
 
Signature · 1 line
2.30.2
author maintainer
#170 lore raw

PATCH 170/700 spi: meson-spicc: fix memory leak in meson_spicc_probe

In reply to Greg Kroah-Hartman

From: zpershuai <[email protected]>

[ Upstream commit b2d501c13470409ee7613855b17e5e5ec4111e1c ]

when meson_spicc_clk_init returns failed, it should goto the out_clk label.

---
 drivers/spi/spi-meson-spicc.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/spi/spi-meson-spicc.c +1 -1
index 51aef2c6e966..b2c4621db34d 100644
@@ -752,7 +752,7 @@ static int meson_spicc_probe(struct platform_device *pdev)
 	ret = meson_spicc_clk_init(spicc);
 	if (ret) {
 		dev_err(&pdev->dev, "clock registration failed\n");
-		goto out_master;
+		goto out_clk;
 	}
 
 	ret = devm_spi_register_master(&pdev->dev, master);
Signature · 1 line
2.30.2
author maintainer
#171 lore raw

PATCH 171/700 regulator: mt6315: Fix checking return value of devm_regmap_init_spmi_ext

In reply to Greg Kroah-Hartman

From: Axel Lin <[email protected]>

[ Upstream commit 70d654ea3de937d7754c107bb8eeb20e30262c89 ]

devm_regmap_init_spmi_ext() returns ERR_PTR() on error.

---
 drivers/regulator/mt6315-regulator.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/regulator/mt6315-regulator.c +2 -2
index 6b8be52c3772..7514702f78cf 100644
@@ -223,8 +223,8 @@ static int mt6315_regulator_probe(struct spmi_device *pdev)
 	int i;
 
 	regmap = devm_regmap_init_spmi_ext(pdev, &mt6315_regmap_config);
-	if (!regmap)
-		return -ENODEV;
+	if (IS_ERR(regmap))
+		return PTR_ERR(regmap);
 
 	chip = devm_kzalloc(dev, sizeof(struct mt6315_chip), GFP_KERNEL);
 	if (!chip)
Signature · 1 line
2.30.2
author maintainer
#172 lore raw

PATCH 172/700 crypto: shash - avoid comparing pointers to exported functions under CFI

In reply to Greg Kroah-Hartman

From: Ard Biesheuvel <[email protected]>

[ Upstream commit 22ca9f4aaf431a9413dcc115dd590123307f274f ]

crypto_shash_alg_has_setkey() is implemented by testing whether the .setkey() member of a struct shash_alg points to the default version, called shash_no_setkey(). As crypto_shash_alg_has_setkey() is a static inline, this requires shash_no_setkey() to be exported to modules.

Unfortunately, when building with CFI, function pointers are routed via CFI stubs which are private to each module (or to the kernel proper) and so this function pointer comparison may fail spuriously.

Let's fix this by turning crypto_shash_alg_has_setkey() into an out of line function.

---
 crypto/shash.c                 | 18 +++++++++++++++---
 include/crypto/internal/hash.h |  8 +-------
 2 files changed, 16 insertions(+), 10 deletions(-)
crypto/shash.c +15 -3
index 2e3433ad9762..0a0a50cb694f 100644
@@ -20,12 +20,24 @@
 
 static const struct crypto_type crypto_shash_type;
 
-int shash_no_setkey(struct crypto_shash *tfm, const u8 *key,
-		    unsigned int keylen)
+static int shash_no_setkey(struct crypto_shash *tfm, const u8 *key,
+			   unsigned int keylen)
 {
 	return -ENOSYS;
 }
-EXPORT_SYMBOL_GPL(shash_no_setkey);
+
+/*
+ * Check whether an shash algorithm has a setkey function.
+ *
+ * For CFI compatibility, this must not be an inline function.  This is because
+ * when CFI is enabled, modules won't get the same address for shash_no_setkey
+ * (if it were exported, which inlining would require) as the core kernel will.
+ */
+bool crypto_shash_alg_has_setkey(struct shash_alg *alg)
+{
+	return alg->setkey != shash_no_setkey;
+}
+EXPORT_SYMBOL_GPL(crypto_shash_alg_has_setkey);
 
 static int shash_setkey_unaligned(struct crypto_shash *tfm, const u8 *key,
 				  unsigned int keylen)
include/crypto/internal/hash.h +1 -7
index 0a288dddcf5b..25806141db59 100644
@@ -75,13 +75,7 @@ void crypto_unregister_ahashes(struct ahash_alg *algs, int count);
 int ahash_register_instance(struct crypto_template *tmpl,
 			    struct ahash_instance *inst);
 
-int shash_no_setkey(struct crypto_shash *tfm, const u8 *key,
-		    unsigned int keylen);
-
-static inline bool crypto_shash_alg_has_setkey(struct shash_alg *alg)
-{
-	return alg->setkey != shash_no_setkey;
-}
+bool crypto_shash_alg_has_setkey(struct shash_alg *alg);
 
 static inline bool crypto_shash_alg_needs_key(struct shash_alg *alg)
 {
Signature · 1 line
2.30.2
author maintainer
#173 lore raw

PATCH 173/700 media: dvb_net: avoid speculation from net slot

In reply to Greg Kroah-Hartman

From: Mauro Carvalho Chehab <[email protected]>

[ Upstream commit abc0226df64dc137b48b911c1fe4319aec5891bb ]

The risk of especulation is actually almost-non-existing here, as there are very few users of TCP/IP using the DVB stack, as, this is mainly used with DVB-S/S2 cards, and only by people that receives TCP/IP from satellite connections, which limits a lot the number of users of such feature(*).

(*) In thesis, DVB-C cards could also benefit from it, but I'm yet to see a hardware that supports it.

Yet, fixing it is trivial.

---
 drivers/media/dvb-core/dvb_net.c | 25 +++++++++++++++++++------
 1 file changed, 19 insertions(+), 6 deletions(-)
drivers/media/dvb-core/dvb_net.c +19 -6
index 89620da983ba..dddebea644bb 100644
@@ -45,6 +45,7 @@
 #include <linux/module.h>
 #include <linux/kernel.h>
 #include <linux/netdevice.h>
+#include <linux/nospec.h>
 #include <linux/etherdevice.h>
 #include <linux/dvb/net.h>
 #include <linux/uio.h>
@@ -1462,14 +1463,20 @@ static int dvb_net_do_ioctl(struct file *file,
 		struct net_device *netdev;
 		struct dvb_net_priv *priv_data;
 		struct dvb_net_if *dvbnetif = parg;
+		int if_num = dvbnetif->if_num;
 
-		if (dvbnetif->if_num >= DVB_NET_DEVICES_MAX ||
-		    !dvbnet->state[dvbnetif->if_num]) {
+		if (if_num >= DVB_NET_DEVICES_MAX) {
 			ret = -EINVAL;
 			goto ioctl_error;
 		}
+		if_num = array_index_nospec(if_num, DVB_NET_DEVICES_MAX);
 
-		netdev = dvbnet->device[dvbnetif->if_num];
+		if (!dvbnet->state[if_num]) {
+			ret = -EINVAL;
+			goto ioctl_error;
+		}
+
+		netdev = dvbnet->device[if_num];
 
 		priv_data = netdev_priv(netdev);
 		dvbnetif->pid=priv_data->pid;
@@ -1522,14 +1529,20 @@ static int dvb_net_do_ioctl(struct file *file,
 		struct net_device *netdev;
 		struct dvb_net_priv *priv_data;
 		struct __dvb_net_if_old *dvbnetif = parg;
+		int if_num = dvbnetif->if_num;
+
+		if (if_num >= DVB_NET_DEVICES_MAX) {
+			ret = -EINVAL;
+			goto ioctl_error;
+		}
+		if_num = array_index_nospec(if_num, DVB_NET_DEVICES_MAX);
 
-		if (dvbnetif->if_num >= DVB_NET_DEVICES_MAX ||
-		    !dvbnet->state[dvbnetif->if_num]) {
+		if (!dvbnet->state[if_num]) {
 			ret = -EINVAL;
 			goto ioctl_error;
 		}
 
-		netdev = dvbnet->device[dvbnetif->if_num];
+		netdev = dvbnet->device[if_num];
 
 		priv_data = netdev_priv(netdev);
 		dvbnetif->pid=priv_data->pid;
Signature · 1 line
2.30.2
author maintainer
#174 lore raw

PATCH 174/700 media: dvbdev: fix error logic at dvb_register_device()

In reply to Greg Kroah-Hartman

From: Mauro Carvalho Chehab <[email protected]>

[ Upstream commit 1fec2ecc252301110e4149e6183fa70460d29674 ]

As reported by smatch:

	drivers/media/dvb-core/dvbdev.c: drivers/media/dvb-core/dvbdev.c:510 dvb_register_device() warn: '&dvbdev->list_head' not removed from list
	drivers/media/dvb-core/dvbdev.c: drivers/media/dvb-core/dvbdev.c:530 dvb_register_device() warn: '&dvbdev->list_head' not removed from list
	drivers/media/dvb-core/dvbdev.c: drivers/media/dvb-core/dvbdev.c:545 dvb_register_device() warn: '&dvbdev->list_head' not removed from list

The error logic inside dvb_register_device() doesn't remove devices from the dvb_adapter_list in case of errors.

---
 drivers/media/dvb-core/dvbdev.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/media/dvb-core/dvbdev.c +3 -0
index 3862ddc86ec4..795d9bfaba5c 100644
@@ -506,6 +506,7 @@ int dvb_register_device(struct dvb_adapter *adap, struct dvb_device **pdvbdev,
 			break;
 
 	if (minor == MAX_DVB_MINORS) {
+		list_del (&dvbdev->list_head);
 		kfree(dvbdevfops);
 		kfree(dvbdev);
 		up_write(&minor_rwsem);
@@ -526,6 +527,7 @@ int dvb_register_device(struct dvb_adapter *adap, struct dvb_device **pdvbdev,
 		      __func__);
 
 		dvb_media_device_free(dvbdev);
+		list_del (&dvbdev->list_head);
 		kfree(dvbdevfops);
 		kfree(dvbdev);
 		mutex_unlock(&dvbdev_register_lock);
@@ -541,6 +543,7 @@ int dvb_register_device(struct dvb_adapter *adap, struct dvb_device **pdvbdev,
 		pr_err("%s: failed to create device dvb%d.%s%d (%ld)\n",
 		       __func__, adap->num, dnames[type], id, PTR_ERR(clsdev));
 		dvb_media_device_free(dvbdev);
+		list_del (&dvbdev->list_head);
 		kfree(dvbdevfops);
 		kfree(dvbdev);
 		return PTR_ERR(clsdev);
Signature · 1 line
2.30.2
author maintainer
#175 lore raw

PATCH 175/700 media: siano: fix device register error path

In reply to Greg Kroah-Hartman

From: Mauro Carvalho Chehab <[email protected]>

[ Upstream commit 5368b1ee2939961a16e74972b69088433fc52195 ]

As reported by smatch:

	drivers/media/common/siano/smsdvb-main.c:1231 smsdvb_hotplug() warn: '&client->entry' not removed from list

If an error occur at the end of the registration logic, it won't drop the device from the list.

---
 drivers/media/common/siano/smsdvb-main.c | 4 ++++
 1 file changed, 4 insertions(+)
drivers/media/common/siano/smsdvb-main.c +4 -0
index ae17407e477a..7cc654bc52d3 100644
@@ -1176,6 +1176,10 @@ static int smsdvb_hotplug(struct smscore_device_t *coredev,
 	return 0;
 
 media_graph_error:
+	mutex_lock(&g_smsdvb_clientslock);
+	list_del(&client->entry);
+	mutex_unlock(&g_smsdvb_clientslock);
+
 	smsdvb_debugfs_release(client);
 
 client_error:
Signature · 1 line
2.30.2
author maintainer
#176 lore raw

PATCH 176/700 media: imx-csi: Skip first few frames from a BT.656 source

In reply to Greg Kroah-Hartman

From: Steve Longerbeam <[email protected]>

[ Upstream commit e198be37e52551bb863d07d2edc535d0932a3c4f ]

Some BT.656 sensors (e.g. ADV718x) transmit frames with unstable BT.656 sync codes after initial power on. This confuses the imx CSI,resulting in vertical and/or horizontal sync issues. Skip the first 20 frames to avoid the unstable sync codes.

[fabio: fixed checkpatch warning and increased the frame skipping to 20]

---
 drivers/staging/media/imx/imx-media-csi.c | 14 +++++++++++++-
 1 file changed, 13 insertions(+), 1 deletion(-)
drivers/staging/media/imx/imx-media-csi.c +13 -1
index ef5add079774..7f4b967646d9 100644
@@ -753,9 +753,10 @@ static int csi_setup(struct csi_priv *priv)
 
 static int csi_start(struct csi_priv *priv)
 {
-	struct v4l2_fract *output_fi;
+	struct v4l2_fract *input_fi, *output_fi;
 	int ret;
 
+	input_fi = &priv->frame_interval[CSI_SINK_PAD];
 	output_fi = &priv->frame_interval[priv->active_output_pad];
 
 	/* start upstream */
@@ -764,6 +765,17 @@ static int csi_start(struct csi_priv *priv)
 	if (ret)
 		return ret;
 
+	/* Skip first few frames from a BT.656 source */
+	if (priv->upstream_ep.bus_type == V4L2_MBUS_BT656) {
+		u32 delay_usec, bad_frames = 20;
+
+		delay_usec = DIV_ROUND_UP_ULL((u64)USEC_PER_SEC *
+			input_fi->numerator * bad_frames,
+			input_fi->denominator);
+
+		usleep_range(delay_usec, delay_usec + 1000);
+	}
+
 	if (priv->dest == IPU_CSI_DEST_IDMAC) {
 		ret = csi_idmac_start(priv);
 		if (ret)
Signature · 1 line
2.30.2
author maintainer
#177 lore raw

PATCH 177/700 hwmon: (max31790) Report correct current pwm duty cycles

In reply to Greg Kroah-Hartman

From: Guenter Roeck <[email protected]>

[ Upstream commit 897f6339893b741a5d68ae8e2475df65946041c2 ]

The MAX31790 has two sets of registers for pwm duty cycles, one to request a duty cycle and one to read the actual current duty cycle. Both do not have to be the same.

When reporting the pwm duty cycle to the user, the actual pwm duty cycle from pwm duty cycle registers needs to be reported. When setting it, the pwm target duty cycle needs to be written. Since we don't know the actual pwm duty cycle after a target pwm duty cycle has been written, set the valid flag to false to indicate that actual pwm duty cycle should be read from the chip instead of using cached values.

---
 Documentation/hwmon/max31790.rst | 3 ++-
 drivers/hwmon/max31790.c         | 6 +++---
 2 files changed, 5 insertions(+), 4 deletions(-)
Documentation/hwmon/max31790.rst +2 -1
index f301385d8cef..54ff0f49e28f 100644
@@ -39,5 +39,6 @@ fan[1-12]_input    RO  fan tachometer speed in RPM
 fan[1-12]_fault    RO  fan experienced fault
 fan[1-6]_target    RW  desired fan speed in RPM
 pwm[1-6]_enable    RW  regulator mode, 0=disabled, 1=manual mode, 2=rpm mode
-pwm[1-6]           RW  fan target duty cycle (0-255)
+pwm[1-6]           RW  read: current pwm duty cycle,
+                       write: target pwm duty cycle (0-255)
 ================== === =======================================================
drivers/hwmon/max31790.c +3 -3
index 86e6c71db685..8ad7a45bfe68 100644
@@ -104,7 +104,7 @@ static struct max31790_data *max31790_update_device(struct device *dev)
 				data->tach[NR_CHANNEL + i] = rv;
 			} else {
 				rv = i2c_smbus_read_word_swapped(client,
-						MAX31790_REG_PWMOUT(i));
+						MAX31790_REG_PWM_DUTY_CYCLE(i));
 				if (rv < 0)
 					goto abort;
 				data->pwm[i] = rv;
@@ -299,10 +299,10 @@ static int max31790_write_pwm(struct device *dev, u32 attr, int channel,
 			err = -EINVAL;
 			break;
 		}
-		data->pwm[channel] = val << 8;
+		data->valid = false;
 		err = i2c_smbus_write_word_swapped(client,
 						   MAX31790_REG_PWMOUT(channel),
-						   data->pwm[channel]);
+						   val << 8);
 		break;
 	case hwmon_pwm_enable:
 		fan_config = data->fan_config[channel];
Signature · 1 line
2.30.2
author maintainer
#178 lore raw

PATCH 178/700 hwmon: (max31790) Fix pwmX_enable attributes

In reply to Greg Kroah-Hartman

From: Guenter Roeck <[email protected]>

[ Upstream commit 148c847c9e5a54b99850617bf9c143af9a344f92 ]

pwmX_enable supports three possible values:

0: Fan control disabled. Duty cycle is fixed to 0% 1: Fan control enabled, pwm mode. Duty cycle is determined by
values written into Target Duty Cycle registers.
2: Fan control enabled, rpm mode
Duty cycle is adjusted such that fan speed matches the values in Target Count registers

The current code does not do this; instead, it mixes pwm control configuration with fan speed monitoring configuration. Worse, it reports that pwm control would be disabled (pwmX_enable==0) when it is in fact enabled in pwm mode. Part of the problem may be that the chip sets the "TACH input enable" bit on its own whenever the mode bit is set to RPM mode, but that doesn't mean that "TACH input enable" accurately reflects the pwm mode.

Fix it up and only handle pwm control with the pwmX_enable attributes. In the documentation, clarify that disabling pwm control (pwmX_enable=0) sets the pwm duty cycle to 0%. In the code, explain why TACH_INPUT_EN is set together with RPM_MODE.

While at it, only update the configuration register if the configuration has changed, and only update the cached configuration if updating the chip configuration was successful.

---
 Documentation/hwmon/max31790.rst |  2 +-
 drivers/hwmon/max31790.c         | 41 ++++++++++++++++++++------------
 2 files changed, 27 insertions(+), 16 deletions(-)
Documentation/hwmon/max31790.rst +1 -1
index 54ff0f49e28f..7b097c3b9b90 100644
@@ -38,7 +38,7 @@ Sysfs entries
 fan[1-12]_input    RO  fan tachometer speed in RPM
 fan[1-12]_fault    RO  fan experienced fault
 fan[1-6]_target    RW  desired fan speed in RPM
-pwm[1-6]_enable    RW  regulator mode, 0=disabled, 1=manual mode, 2=rpm mode
+pwm[1-6]_enable    RW  regulator mode, 0=disabled (duty cycle=0%), 1=manual mode, 2=rpm mode
 pwm[1-6]           RW  read: current pwm duty cycle,
                        write: target pwm duty cycle (0-255)
 ================== === =======================================================
drivers/hwmon/max31790.c +26 -15

67 lines not on this page · raw message

Signature · 1 line
2.30.2
author maintainer
#179 lore raw

PATCH 179/700 sched/fair: Take thermal pressure into account while estimating energy

In reply to Greg Kroah-Hartman

From: Lukasz Luba <[email protected]>

[ Upstream commit 489f16459e0008c7a5c4c5af34bd80898aa82c2d ]

Energy Aware Scheduling (EAS) needs to be able to predict the frequency requests made by the SchedUtil governor to properly estimate energy used in the future. It has to take into account CPUs utilization and forecast Performance Domain (PD) frequency. There is a corner case when the max allowed frequency might be reduced due to thermal. SchedUtil is aware of that reduced frequency, so it should be taken into account also in EAS estimations.

SchedUtil, as a CPUFreq governor, knows the maximum allowed frequency of a CPU, thanks to cpufreq_driver_resolve_freq() and internal clamping to 'policy::max'. SchedUtil is responsible to respect that upper limit while setting the frequency through CPUFreq drivers. This effective frequency is stored internally in 'sugov_policy::next_freq' and EAS has to predict that value.

In the existing code the raw value of arch_scale_cpu_capacity() is used for clamping the returned CPU utilization from effective_cpu_util(). This patch fixes issue with too big single CPU utilization, by introducing clamping to the allowed CPU capacity. The allowed CPU capacity is a CPU capacity reduced by thermal pressure raw value.

Thanks to knowledge about allowed CPU capacity, we don't get too big value for a single CPU utilization, which is then added to the util sum. The util sum is used as a source of information for estimating whole PD energy. To avoid wrong energy estimation in EAS (due to capped frequency), make sure that the calculation of util sum is aware of allowed CPU capacity.

This thermal pressure might be visible in scenarios where the CPUs are not heavily loaded, but some other component (like GPU) drastically reduced available power budget and increased the SoC temperature. Thus, we still use EAS for task placement and CPUs are not over-utilized.

---
 kernel/sched/fair.c | 11 ++++++++---
 1 file changed, 8 insertions(+), 3 deletions(-)
kernel/sched/fair.c +8 -3
index 272c583fc167..20ac5dff9a0c 100644
@@ -6564,8 +6564,11 @@ compute_energy(struct task_struct *p, int dst_cpu, struct perf_domain *pd)
 	struct cpumask *pd_mask = perf_domain_span(pd);
 	unsigned long cpu_cap = arch_scale_cpu_capacity(cpumask_first(pd_mask));
 	unsigned long max_util = 0, sum_util = 0;
+	unsigned long _cpu_cap = cpu_cap;
 	int cpu;
 
+	_cpu_cap -= arch_scale_thermal_pressure(cpumask_first(pd_mask));
+
 	/*
 	 * The capacity state of CPUs of the current rd can be driven by CPUs
 	 * of another rd if they belong to the same pd. So, account for the
@@ -6601,8 +6604,10 @@ compute_energy(struct task_struct *p, int dst_cpu, struct perf_domain *pd)
 		 * is already enough to scale the EM reported power
 		 * consumption at the (eventually clamped) cpu_capacity.
 		 */
-		sum_util += effective_cpu_util(cpu, util_running, cpu_cap,
-					       ENERGY_UTIL, NULL);
+		cpu_util = effective_cpu_util(cpu, util_running, cpu_cap,
+					      ENERGY_UTIL, NULL);
+
+		sum_util += min(cpu_util, _cpu_cap);
 
 		/*
 		 * Performance domain frequency: utilization clamping
@@ -6613,7 +6618,7 @@ compute_energy(struct task_struct *p, int dst_cpu, struct perf_domain *pd)
 		 */
 		cpu_util = effective_cpu_util(cpu, util_freq, cpu_cap,
 					      FREQUENCY_UTIL, tsk);
-		max_util = max(max_util, cpu_util);
+		max_util = max(max_util, min(cpu_util, _cpu_cap));
 	}
 
 	return em_cpu_energy(pd->em_pd, max_util, sum_util);
Signature · 1 line
2.30.2
author maintainer
#180 lore raw

PATCH 180/700 drivers/perf: fix the missed ida_simple_remove() in ddr_perf_probe()

In reply to Greg Kroah-Hartman

From: Jing Xiangfeng <[email protected]>

[ Upstream commit d96b1b8c9f79b6bb234a31c80972a6f422079376 ]

ddr_perf_probe() misses to call ida_simple_remove() in an error path. Jump to cpuhp_state_err to fix it.

---
 drivers/perf/fsl_imx8_ddr_perf.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/perf/fsl_imx8_ddr_perf.c +4 -2
index be1f26b62ddb..4a56849f0400 100644
@@ -706,8 +706,10 @@ static int ddr_perf_probe(struct platform_device *pdev)
 
 	name = devm_kasprintf(&pdev->dev, GFP_KERNEL, DDR_PERF_DEV_NAME "%d",
 			      num);
-	if (!name)
-		return -ENOMEM;
+	if (!name) {
+		ret = -ENOMEM;
+		goto cpuhp_state_err;
+	}
 
 	pmu->devtype_data = of_device_get_match_data(&pdev->dev);
 
Signature · 1 line
2.30.2
author maintainer
#181 lore raw

PATCH 181/700 KVM: arm64: Restore PMU configuration on first run

In reply to Greg Kroah-Hartman

From: Marc Zyngier <[email protected]>

[ Upstream commit d0c94c49792cf780cbfefe29f81bb8c3b73bc76b ]

Restoring a guest with an active virtual PMU results in no perf counters being instanciated on the host side. Not quite what you'd expect from a restore.

In order to fix this, force a writeback of PMCR_EL0 on the first run of a vcpu (using a new request so that it happens once the vcpu has been loaded). This will in turn create all the host-side counters that were missing.

---
 arch/arm64/include/asm/kvm_host.h | 1 +
 arch/arm64/kvm/arm.c              | 4 ++++
 arch/arm64/kvm/pmu-emul.c         | 3 +++
 3 files changed, 8 insertions(+)
arch/arm64/include/asm/kvm_host.h +1 -0
index 858c2fcfc043..4e4356add46e 100644
@@ -46,6 +46,7 @@
 #define KVM_REQ_VCPU_RESET	KVM_ARCH_REQ(2)
 #define KVM_REQ_RECORD_STEAL	KVM_ARCH_REQ(3)
 #define KVM_REQ_RELOAD_GICv4	KVM_ARCH_REQ(4)
+#define KVM_REQ_RELOAD_PMU	KVM_ARCH_REQ(5)
 
 #define KVM_DIRTY_LOG_MANUAL_CAPS   (KVM_DIRTY_LOG_MANUAL_PROTECT_ENABLE | \
 				     KVM_DIRTY_LOG_INITIALLY_SET)
arch/arm64/kvm/arm.c +4 -0
index 7730b81aad6d..8455c5c30116 100644
@@ -684,6 +684,10 @@ static void check_vcpu_requests(struct kvm_vcpu *vcpu)
 			vgic_v4_load(vcpu);
 			preempt_enable();
 		}
+
+		if (kvm_check_request(KVM_REQ_RELOAD_PMU, vcpu))
+			kvm_pmu_handle_pmcr(vcpu,
+					    __vcpu_sys_reg(vcpu, PMCR_EL0));
 	}
 }
 
arch/arm64/kvm/pmu-emul.c +3 -0
index e32c6e139a09..e9699d10d2bd 100644
@@ -850,6 +850,9 @@ int kvm_arm_pmu_v3_enable(struct kvm_vcpu *vcpu)
 		   return -EINVAL;
 	}
 
+	/* One-off reload of the PMU on first run */
+	kvm_make_request(KVM_REQ_RELOAD_PMU, vcpu);
+
 	return 0;
 }
 
Signature · 1 line
2.30.2
author maintainer
#182 lore raw

PATCH 182/700 KVM: PPC: Book3S HV: Fix TLB management on SMT8 POWER9 and POWER10 processors

In reply to Greg Kroah-Hartman

From: Suraj Jitindar Singh <[email protected]>

[ Upstream commit 77bbbc0cf84834ed130838f7ac1988567f4d0288 ]

The POWER9 vCPU TLB management code assumes all threads in a core share a TLB, and that TLBIEL execued by one thread will invalidate TLBs for all threads. This is not the case for SMT8 capable POWER9 and POWER10 (big core) processors, where the TLB is split between groups of threads. This results in TLB multi-hits, random data corruption, etc.

Fix this by introducing cpu_first_tlb_thread_sibling etc., to determine which siblings share TLBs, and use that in the guest TLB flushing code.

[[email protected]: add changelog and comment]

---
 arch/powerpc/include/asm/cputhreads.h | 30 +++++++++++++++++++++++++++
 arch/powerpc/kvm/book3s_hv.c          | 13 ++++++------
 arch/powerpc/kvm/book3s_hv_builtin.c  |  2 +-
 arch/powerpc/kvm/book3s_hv_rm_mmu.c   |  2 +-
 4 files changed, 39 insertions(+), 8 deletions(-)
arch/powerpc/include/asm/cputhreads.h +30 -0
index 98c8bd155bf9..b167186aaee4 100644
@@ -98,6 +98,36 @@ static inline int cpu_last_thread_sibling(int cpu)
 	return cpu | (threads_per_core - 1);
 }
 
+/*
+ * tlb_thread_siblings are siblings which share a TLB. This is not
+ * architected, is not something a hypervisor could emulate and a future
+ * CPU may change behaviour even in compat mode, so this should only be
+ * used on PowerNV, and only with care.
+ */
+static inline int cpu_first_tlb_thread_sibling(int cpu)
+{
+	if (cpu_has_feature(CPU_FTR_ARCH_300) && (threads_per_core == 8))
+		return cpu & ~0x6;	/* Big Core */
+	else
+		return cpu_first_thread_sibling(cpu);
+}
+
+static inline int cpu_last_tlb_thread_sibling(int cpu)
+{
+	if (cpu_has_feature(CPU_FTR_ARCH_300) && (threads_per_core == 8))
+		return cpu | 0x6;	/* Big Core */
+	else
+		return cpu_last_thread_sibling(cpu);
+}
+
+static inline int cpu_tlb_thread_sibling_step(void)
+{
+	if (cpu_has_feature(CPU_FTR_ARCH_300) && (threads_per_core == 8))
+		return 2;		/* Big Core */
+	else
+		return 1;
+}
+
 static inline u32 get_tensr(void)
 {
 #ifdef	CONFIG_BOOKE
arch/powerpc/kvm/book3s_hv.c +7 -6
index 60c5bc0c130c..1c6e0a52fb53 100644
@@ -2619,7 +2619,7 @@ static void radix_flush_cpu(struct kvm *kvm, int cpu, struct kvm_vcpu *vcpu)
 	cpumask_t *cpu_in_guest;
 	int i;
 
-	cpu = cpu_first_thread_sibling(cpu);
+	cpu = cpu_first_tlb_thread_sibling(cpu);
 	if (nested) {
 		cpumask_set_cpu(cpu, &nested->need_tlb_flush);
 		cpu_in_guest = &nested->cpu_in_guest;
@@ -2633,9 +2633,10 @@ static void radix_flush_cpu(struct kvm *kvm, int cpu, struct kvm_vcpu *vcpu)
 	 * the other side is the first smp_mb() in kvmppc_run_core().
 	 */
 	smp_mb();
-	for (i = 0; i < threads_per_core; ++i)
-		if (cpumask_test_cpu(cpu + i, cpu_in_guest))
-			smp_call_function_single(cpu + i, do_nothing, NULL, 1);
+	for (i = cpu; i <= cpu_last_tlb_thread_sibling(cpu);
+					i += cpu_tlb_thread_sibling_step())
+		if (cpumask_test_cpu(i, cpu_in_guest))
+			smp_call_function_single(i, do_nothing, NULL, 1);
 }
 
 static void kvmppc_prepare_radix_vcpu(struct kvm_vcpu *vcpu, int pcpu)
@@ -2666,8 +2667,8 @@ static void kvmppc_prepare_radix_vcpu(struct kvm_vcpu *vcpu, int pcpu)
 	 */
 	if (prev_cpu != pcpu) {
 		if (prev_cpu >= 0 &&
-		    cpu_first_thread_sibling(prev_cpu) !=
-		    cpu_first_thread_sibling(pcpu))
+		    cpu_first_tlb_thread_sibling(prev_cpu) !=
+		    cpu_first_tlb_thread_sibling(pcpu))
 			radix_flush_cpu(kvm, prev_cpu, vcpu);
 		if (nested)
 			nested->prev_cpu[vcpu->arch.nested_vcpu_id] = pcpu;
arch/powerpc/kvm/book3s_hv_builtin.c +1 -1
index 158d309b42a3..b5e5d07cb40f 100644
@@ -797,7 +797,7 @@ void kvmppc_check_need_tlb_flush(struct kvm *kvm, int pcpu,
 	 * Thus we make all 4 threads use the same bit.
 	 */
 	if (cpu_has_feature(CPU_FTR_ARCH_300))
-		pcpu = cpu_first_thread_sibling(pcpu);
+		pcpu = cpu_first_tlb_thread_sibling(pcpu);
 
 	if (nested)
 		need_tlb_flush = &nested->need_tlb_flush;
arch/powerpc/kvm/book3s_hv_rm_mmu.c +1 -1
index 88da2764c1bb..3ddc83d2e849 100644
@@ -67,7 +67,7 @@ static int global_invalidates(struct kvm *kvm)
 		 * so use the bit for the first thread to represent the core.
 		 */
 		if (cpu_has_feature(CPU_FTR_ARCH_300))
-			cpu = cpu_first_thread_sibling(cpu);
+			cpu = cpu_first_tlb_thread_sibling(cpu);
 		cpumask_clear_cpu(cpu, &kvm->arch.need_tlb_flush);
 	}
 
Signature · 1 line
2.30.2
author maintainer
#183 lore raw

PATCH 183/700 btrfs: fix error handling in __btrfs_update_delayed_inode

In reply to Greg Kroah-Hartman

From: Josef Bacik <[email protected]>

[ Upstream commit bb385bedded3ccbd794559600de4a09448810f4a ]

If we get an error while looking up the inode item we'll simply bail without cleaning up the delayed node. This results in this style of warning happening on commit:

  WARNING: CPU: 0 PID: 76403 at fs/btrfs/delayed-inode.c:1365 btrfs_assert_delayed_root_empty+0x5b/0x90
  CPU: 0 PID: 76403 Comm: fsstress Tainted: G        W         5.13.0-rc1+ #373
  Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.13.0-2.fc32 04/01/2014
  RIP: 0010:btrfs_assert_delayed_root_empty+0x5b/0x90
  RSP: 0018:ffffb8bb815a7e50 EFLAGS: 00010286
  RAX: 0000000000000000 RBX: ffff95d6d07e1888 RCX: ffff95d6c0fa3000
  RDX: 0000000000000002 RSI: 000000000029e91c RDI: ffff95d6c0fc8060
  RBP: ffff95d6c0fc8060 R08: 00008d6d701a2c1d R09: 0000000000000000
  R10: ffff95d6d1760ea0 R11: 0000000000000001 R12: ffff95d6c15a4d00
  R13: ffff95d6c0fa3000 R14: 0000000000000000 R15: ffffb8bb815a7e90
  FS:  00007f490e8dbb80(0000) GS:ffff95d73bc00000(0000) knlGS:0000000000000000
  CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
  CR2: 00007f6e75555cb0 CR3: 00000001101ce001 CR4: 0000000000370ef0
  Call Trace:
   btrfs_commit_transaction+0x43c/0xb00
   ? finish_wait+0x80/0x80
   ? vfs_fsync_range+0x90/0x90
   iterate_supers+0x8c/0x100
   ksys_sync+0x50/0x90
   __do_sys_sync+0xa/0x10
   do_syscall_64+0x3d/0x80
   entry_SYSCALL_64_after_hwframe+0x44/0xae

Because the iref isn't dropped and this leaves an elevated node->count, so any release just re-queues it onto the delayed inodes list. Fix this by going to the out label to handle the proper cleanup of the delayed node.

---
 fs/btrfs/delayed-inode.c | 10 ++++------
 1 file changed, 4 insertions(+), 6 deletions(-)
fs/btrfs/delayed-inode.c +4 -6
index c1d2b6786129..2f73846d712f 100644
@@ -1025,12 +1025,10 @@ static int __btrfs_update_delayed_inode(struct btrfs_trans_handle *trans,
 	nofs_flag = memalloc_nofs_save();
 	ret = btrfs_lookup_inode(trans, root, path, &key, mod);
 	memalloc_nofs_restore(nofs_flag);
-	if (ret > 0) {
-		btrfs_release_path(path);
-		return -ENOENT;
-	} else if (ret < 0) {
-		return ret;
-	}
+	if (ret > 0)
+		ret = -ENOENT;
+	if (ret < 0)
+		goto out;
 
 	leaf = path->nodes[0];
 	inode_item = btrfs_item_ptr(leaf, path->slots[0],
Signature · 1 line
2.30.2
author maintainer
#184 lore raw

PATCH 184/700 btrfs: abort transaction if we fail to update the delayed inode

In reply to Greg Kroah-Hartman

From: Josef Bacik <[email protected]>

[ Upstream commit 04587ad9bef6ce9d510325b4ba9852b6129eebdb ]

If we fail to update the delayed inode we need to abort the transaction, because we could leave an inode with the improper counts or some other such corruption behind.

---
 fs/btrfs/delayed-inode.c | 8 ++++++++
 1 file changed, 8 insertions(+)
fs/btrfs/delayed-inode.c +8 -0
index 2f73846d712f..55dad12a5ce0 100644
@@ -1066,6 +1066,14 @@ err_out:
 	btrfs_delayed_inode_release_metadata(fs_info, node, (ret < 0));
 	btrfs_release_delayed_inode(node);
 
+	/*
+	 * If we fail to update the delayed inode we need to abort the
+	 * transaction, because we could leave the inode with the improper
+	 * counts behind.
+	 */
+	if (ret && ret != -ENOENT)
+		btrfs_abort_transaction(trans, ret);
+
 	return ret;
 
 search:
Signature · 1 line
2.30.2
author maintainer
#185 lore raw

PATCH 185/700 btrfs: always abort the transaction if we abort a trans handle

In reply to Greg Kroah-Hartman

From: Josef Bacik <[email protected]>

[ Upstream commit 5963ffcaf383134985a5a2d8a4baa582d3999e0a ]

While stress testing our error handling I noticed that sometimes we would still commit the transaction even though we had aborted the transaction.

Currently we track if a trans handle has dirtied any metadata, and if it hasn't we mark the filesystem as having an error (so no new transactions can be started), but we will allow the current transaction to complete as we do not mark the transaction itself as having been aborted.

This sounds good in theory, but we were not properly tracking IO errors in btrfs_finish_ordered_io, and thus committing the transaction with bogus free space data. This isn't necessarily a problem per-se with the free space cache, as the other guards in place would have kept us from accepting the free space cache as valid, but highlights a real world case where we had a bug and could have corrupted the filesystem because of it.

This "skip abort on empty trans handle" is nice in theory, but assumes we have perfect error handling everywhere, which we clearly do not. Also we do not allow further transactions to be started, so all this does is save the last transaction that was happening, which doesn't necessarily gain us anything other than the potential for real corruption.

Remove this particular bit of code, if we decide we need to abort the transaction then abort the current one and keep us from doing real harm to the file system, regardless of whether this specific trans handle dirtied anything or not.

---
 fs/btrfs/ctree.c       |  5 +----
 fs/btrfs/extent-tree.c |  1 -
 fs/btrfs/super.c       | 11 -----------
 fs/btrfs/transaction.c |  8 --------
 fs/btrfs/transaction.h |  1 -
 5 files changed, 1 insertion(+), 25 deletions(-)
fs/btrfs/ctree.c +1 -4
index f43ce82a6aed..8f48553d861e 100644
@@ -1498,7 +1498,6 @@ noinline int btrfs_cow_block(struct btrfs_trans_handle *trans,
 		       trans->transid, fs_info->generation);
 
 	if (!should_cow_block(trans, root, buf)) {
-		trans->dirty = true;
 		*cow_ret = buf;
 		return 0;
 	}
@@ -2670,10 +2669,8 @@ again:
 			 * then we don't want to set the path blocking,
 			 * so we test it here
 			 */
-			if (!should_cow_block(trans, root, b)) {
-				trans->dirty = true;
+			if (!should_cow_block(trans, root, b))
 				goto cow_done;
-			}
 
 			/*
 			 * must have write locks on this node and the
fs/btrfs/extent-tree.c +0 -1
index 27c368007481..1cde6f84f145 100644
@@ -4799,7 +4799,6 @@ btrfs_init_new_buffer(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 		set_extent_dirty(&trans->transaction->dirty_pages, buf->start,
 			 buf->start + buf->len - 1, GFP_NOFS);
 	}
-	trans->dirty = true;
 	/* this returns a buffer locked for blocking */
 	return buf;
 }
fs/btrfs/super.c +0 -11
index f7a4ad86adee..b3b7f3066cfa 100644
@@ -273,17 +273,6 @@ void __btrfs_abort_transaction(struct btrfs_trans_handle *trans,
 	struct btrfs_fs_info *fs_info = trans->fs_info;
 
 	WRITE_ONCE(trans->aborted, errno);
-	/* Nothing used. The other threads that have joined this
-	 * transaction may be able to continue. */
-	if (!trans->dirty && list_empty(&trans->new_bgs)) {
-		const char *errstr;
-
-		errstr = btrfs_decode_error(errno);
-		btrfs_warn(fs_info,
-		           "%s:%d: Aborting unused transaction(%s).",
-		           function, line, errstr);
-		return;
-	}
 	WRITE_ONCE(trans->transaction->aborted, errno);
 	/* Wake up anybody who may be waiting on this transaction */
 	wake_up(&fs_info->transaction_wait);
fs/btrfs/transaction.c +0 -8
index d678c24250a4..81c8567dffee 100644
@@ -2056,14 +2056,6 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans)
 
 	ASSERT(refcount_read(&trans->use_count) == 1);
 
-	/*
-	 * Some places just start a transaction to commit it.  We need to make
-	 * sure that if this commit fails that the abort code actually marks the
-	 * transaction as failed, so set trans->dirty to make the abort code do
-	 * the right thing.
-	 */
-	trans->dirty = true;
-
 	/* Stop the commit early if ->aborted is set */
 	if (TRANS_ABORTED(cur_trans)) {
 		ret = cur_trans->aborted;
fs/btrfs/transaction.h +0 -1
index 364cfbb4c5c5..c49e2266b28b 100644
@@ -143,7 +143,6 @@ struct btrfs_trans_handle {
 	bool allocating_chunk;
 	bool can_flush_pending_bgs;
 	bool reloc_reserved;
-	bool dirty;
 	bool in_fsync;
 	struct btrfs_root *root;
 	struct btrfs_fs_info *fs_info;
Signature · 1 line
2.30.2
author maintainer
#186 lore raw

PATCH 186/700 btrfs: sysfs: fix format string for some discard stats

In reply to Greg Kroah-Hartman

From: David Sterba <[email protected]>

[ Upstream commit 8c5ec995616f1202ab92e195fd75d6f60d86f85c ]

The type of discard_bitmap_bytes and discard_extent_bytes is u64 so the format should be %llu, though the actual values would hardly ever overflow to negative values.

---
 fs/btrfs/sysfs.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
fs/btrfs/sysfs.c +2 -2
index 6eb1c50fa98c..bd95446869d0 100644
@@ -414,7 +414,7 @@ static ssize_t btrfs_discard_bitmap_bytes_show(struct kobject *kobj,
 {
 	struct btrfs_fs_info *fs_info = discard_to_fs_info(kobj);
 
-	return scnprintf(buf, PAGE_SIZE, "%lld\n",
+	return scnprintf(buf, PAGE_SIZE, "%llu\n",
 			fs_info->discard_ctl.discard_bitmap_bytes);
 }
 BTRFS_ATTR(discard, discard_bitmap_bytes, btrfs_discard_bitmap_bytes_show);
@@ -436,7 +436,7 @@ static ssize_t btrfs_discard_extent_bytes_show(struct kobject *kobj,
 {
 	struct btrfs_fs_info *fs_info = discard_to_fs_info(kobj);
 
-	return scnprintf(buf, PAGE_SIZE, "%lld\n",
+	return scnprintf(buf, PAGE_SIZE, "%llu\n",
 			fs_info->discard_ctl.discard_extent_bytes);
 }
 BTRFS_ATTR(discard, discard_extent_bytes, btrfs_discard_extent_bytes_show);
Signature · 1 line
2.30.2
author maintainer
#187 lore raw

PATCH 187/700 btrfs: dont clear page extent mapped if were not invalidating the full page

In reply to Greg Kroah-Hartman

From: Qu Wenruo <[email protected]>

[ Upstream commit bcd77455d590eaa0422a5e84ae852007cfce574a ]

[BUG]
With current btrfs subpage rw support, the following script can lead to fs hang:

  $ mkfs.btrfs -f -s 4k $dev
  $ mount $dev -o nospace_cache $mnt
  $ fsstress -w -n 100 -p 1 -s 1608140256 -v -d $mnt

The fs will hang at btrfs_start_ordered_extent().

[CAUSE]
In above test case, btrfs_invalidate() will be called with the following parameters:

offset = 0 length = 53248 page dirty = 1 subpage dirty bitmap = 0x2000

Since @offset is 0, btrfs_invalidate() will try to invalidate the full page, and finally call clear_page_extent_mapped() which will detach subpage structure from the page.

And since the page no longer has subpage structure, the subpage dirty bitmap will be cleared, preventing the dirty range from being written back, thus no way to wake up the ordered extent.

[FIX]
Just follow other filesystems, only to invalidate the page if the range covers the full page.

There are cases like truncate_setsize() which can call btrfs_invalidatepage() with offset == 0 and length != 0 for the last page of an inode.

Although the old code will still try to invalidate the full page, we are still safe to just wait for ordered extent to finish.
So it shouldn't cause extra problems.

---
 fs/btrfs/inode.c | 14 +++++++++++++-
 1 file changed, 13 insertions(+), 1 deletion(-)
fs/btrfs/inode.c +13 -1
index 3e7173c7660d..1b22ded1a799 100644
@@ -8386,7 +8386,19 @@ static void btrfs_invalidatepage(struct page *page, unsigned int offset,
 	 */
 	wait_on_page_writeback(page);
 
-	if (offset) {
+	/*
+	 * For subpage case, we have call sites like
+	 * btrfs_punch_hole_lock_range() which passes range not aligned to
+	 * sectorsize.
+	 * If the range doesn't cover the full page, we don't need to and
+	 * shouldn't clear page extent mapped, as page->private can still
+	 * record subpage dirty bits for other part of the range.
+	 *
+	 * For cases that can invalidate the full even the range doesn't
+	 * cover the full page, like invalidating the last page, we're
+	 * still safe to wait for ordered extent to finish.
+	 */
+	if (!(offset == 0 && length == PAGE_SIZE)) {
 		btrfs_releasepage(page, GFP_NOFS);
 		return;
 	}
Signature · 1 line
2.30.2
author maintainer
#188 lore raw

PATCH 188/700 btrfs: disable build on platforms having page size 256K

In reply to Greg Kroah-Hartman

From: Christophe Leroy <[email protected]>

[ Upstream commit b05fbcc36be1f8597a1febef4892053a0b2f3f60 ]

With a config having PAGE_SIZE set to 256K, BTRFS build fails with the following message

include/linux/compiler_types.h:326:38: error: call to '__compiletime_assert_791' declared with attribute error:
BUILD_BUG_ON failed: (BTRFS_MAX_COMPRESSED % PAGE_SIZE) != 0

BTRFS_MAX_COMPRESSED being 128K, BTRFS cannot support platforms with 256K pages at the time being.

There are two platforms that can select 256K pages:
- hexagon
- powerpc

Disable BTRFS when 256K page size is selected. Supporting this would require changes to the subpage mode that's currently being developed. Given that 256K is many times larger than page sizes commonly used and for what the algorithms and structures have been tuned, it's out of scope and disabling build is a reasonable option.

---
 fs/btrfs/Kconfig | 2 ++
 1 file changed, 2 insertions(+)
fs/btrfs/Kconfig +2 -0
index 68b95ad82126..520a0f6a7d9e 100644
@@ -18,6 +18,8 @@ config BTRFS_FS
 	select RAID6_PQ
 	select XOR_BLOCKS
 	select SRCU
+	depends on !PPC_256K_PAGES	# powerpc
+	depends on !PAGE_SIZE_256KB	# hexagon
 
 	help
 	  Btrfs is a general purpose copy-on-write filesystem with extents,
Signature · 1 line
2.30.2
author maintainer
#189 lore raw

PATCH 189/700 locking/lockdep: Fix the dep path printing for backwards BFS

In reply to Greg Kroah-Hartman

From: Boqun Feng <[email protected]>

[ Upstream commit 69c7a5fb2482636f525f016c8333fdb9111ecb9d ]

We use the same code to print backwards lock dependency path as the forwards lock dependency path, and this could result into incorrect printing because for a backwards lock_list ->trace is not the call trace where the lock of ->class is acquired.

Fix this by introducing a separate function on printing the backwards dependency path. Also add a few comments about the printing while we are at it.

---
 kernel/locking/lockdep.c | 108 ++++++++++++++++++++++++++++++++++++++-
 1 file changed, 106 insertions(+), 2 deletions(-)
kernel/locking/lockdep.c +106 -2
index 5bf6b1659215..03a9d9b96045 100644
@@ -2305,7 +2305,56 @@ static void print_lock_class_header(struct lock_class *class, int depth)
 }
 
 /*
- * printk the shortest lock dependencies from @start to @end in reverse order:
+ * Dependency path printing:
+ *
+ * After BFS we get a lock dependency path (linked via ->parent of lock_list),
+ * printing out each lock in the dependency path will help on understanding how
+ * the deadlock could happen. Here are some details about dependency path
+ * printing:
+ *
+ * 1)	A lock_list can be either forwards or backwards for a lock dependency,
+ * 	for a lock dependency A -> B, there are two lock_lists:
+ *
+ * 	a)	lock_list in the ->locks_after list of A, whose ->class is B and
+ * 		->links_to is A. In this case, we can say the lock_list is
+ * 		"A -> B" (forwards case).
+ *
+ * 	b)	lock_list in the ->locks_before list of B, whose ->class is A
+ * 		and ->links_to is B. In this case, we can say the lock_list is
+ * 		"B <- A" (bacwards case).
+ *
+ * 	The ->trace of both a) and b) point to the call trace where B was
+ * 	acquired with A held.
+ *
+ * 2)	A "helper" lock_list is introduced during BFS, this lock_list doesn't
+ * 	represent a certain lock dependency, it only provides an initial entry
+ * 	for BFS. For example, BFS may introduce a "helper" lock_list whose
+ * 	->class is A, as a result BFS will search all dependencies starting with
+ * 	A, e.g. A -> B or A -> C.
+ *
+ * 	The notation of a forwards helper lock_list is like "-> A", which means
+ * 	we should search the forwards dependencies starting with "A", e.g A -> B
+ * 	or A -> C.
+ *
+ * 	The notation of a bacwards helper lock_list is like "<- B", which means
+ * 	we should search the backwards dependencies ending with "B", e.g.
+ * 	B <- A or B <- C.
+ */
+
+/*
+ * printk the shortest lock dependencies from @root to @leaf in reverse order.
+ *
+ * We have a lock dependency path as follow:
+ *
+ *    @root                                                                 @leaf
+ *      |                                                                     |
+ *      V                                                                     V
+ *	          ->parent                                   ->parent
+ * | lock_list | <--------- | lock_list | ... | lock_list  | <--------- | lock_list |
+ * |    -> L1  |            | L1 -> L2  | ... |Ln-2 -> Ln-1|            | Ln-1 -> Ln|
+ *
+ * , so it's natural that we start from @leaf and print every ->class and
+ * ->trace until we reach the @root.
  */
 static void __used
 print_shortest_lock_dependencies(struct lock_list *leaf,
@@ -2333,6 +2382,61 @@ print_shortest_lock_dependencies(struct lock_list *leaf,
 	} while (entry && (depth >= 0));
 }
 
+/*
+ * printk the shortest lock dependencies from @leaf to @root.
+ *
+ * We have a lock dependency path (from a backwards search) as follow:
+ *
+ *    @leaf                                                                 @root
+ *      |                                                                     |
+ *      V                                                                     V
+ *	          ->parent                                   ->parent
+ * | lock_list | ---------> | lock_list | ... | lock_list  | ---------> | lock_list |
+ * | L2 <- L1  |            | L3 <- L2  | ... | Ln <- Ln-1 |            |    <- Ln  |
+ *
+ * , so when we iterate from @leaf to @root, we actually print the lock
+ * dependency path L1 -> L2 -> .. -> Ln in the non-reverse order.
+ *
+ * Another thing to notice here is that ->class of L2 <- L1 is L1, while the
+ * ->trace of L2 <- L1 is the call trace of L2, in fact we don't have the call
+ * trace of L1 in the dependency path, which is alright, because most of the
+ * time we can figure out where L1 is held from the call trace of L2.
+ */
+static void __used
+print_shortest_lock_dependencies_backwards(struct lock_list *leaf,
+					   struct lock_list *root)
+{
+	struct lock_list *entry = leaf;
+	const struct lock_trace *trace = NULL;
+	int depth;
+
+	/*compute depth from generated tree by BFS*/
+	depth = get_lock_depth(leaf);
+
+	do {
+		print_lock_class_header(entry->class, depth);
+		if (trace) {
+			printk("%*s ... acquired at:\n", depth, "");
+			print_lock_trace(trace, 2);
+			printk("\n");
+		}
+
+		/*
+		 * Record the pointer to the trace for the next lock_list
+		 * entry, see the comments for the function.
+		 */
+		trace = entry->trace;
+
+		if (depth == 0 && (entry != root)) {
+			printk("lockdep:%s bad path found in chain graph\n", __func__);
+			break;
+		}
+
+		entry = get_lock_parent(entry);
+		depth--;
+	} while (entry && (depth >= 0));
+}
+
 static void
 print_irq_lock_scenario(struct lock_list *safe_entry,
 			struct lock_list *unsafe_entry,
@@ -2450,7 +2554,7 @@ print_bad_irq_dependency(struct task_struct *curr,
 	prev_root->trace = save_trace();
 	if (!prev_root->trace)
 		return;
-	print_shortest_lock_dependencies(backwards_entry, prev_root);
+	print_shortest_lock_dependencies_backwards(backwards_entry, prev_root);
 
 	pr_warn("\nthe dependencies between the lock to be acquired");
 	pr_warn(" and %s-irq-unsafe lock:\n", irqclass);
Signature · 1 line
2.30.2
author maintainer
#190 lore raw

PATCH 190/700 lockding/lockdep: Avoid to find wrong lock dep path in check_irq_usage()

In reply to Greg Kroah-Hartman

From: Boqun Feng <[email protected]>

[ Upstream commit 7b1f8c6179769af6ffa055e1169610b51d71edd5 ]

In the step #3 of check_irq_usage(), we seach backwards to find a lock whose usage conflicts the usage of @target_entry1 on safe/unsafe. However, we should only keep the irq-unsafe usage of @target_entry1 into consideration, because it could be a case where a lock is hardirq-unsafe but soft-safe, and in check_irq_usage() we find it because its hardirq-unsafe could result into a hardirq-safe-unsafe deadlock, but currently since we don't filter out the other usage bits, so we may find a lock dependency path softirq-unsafe -> softirq-safe, which in fact doesn't cause a deadlock. And this may cause misleading lockdep splats.

Fix this by only keeping LOCKF_ENABLED_IRQ_ALL bits when we try the backwards search.

---
 kernel/locking/lockdep.c | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)
kernel/locking/lockdep.c +11 -1
index 03a9d9b96045..b56c3855756e 100644
@@ -2772,8 +2772,18 @@ static int check_irq_usage(struct task_struct *curr, struct held_lock *prev,
 	 * Step 3: we found a bad match! Now retrieve a lock from the backward
 	 * list whose usage mask matches the exclusive usage mask from the
 	 * lock found on the forward list.
+	 *
+	 * Note, we should only keep the LOCKF_ENABLED_IRQ_ALL bits, considering
+	 * the follow case:
+	 *
+	 * When trying to add A -> B to the graph, we find that there is a
+	 * hardirq-safe L, that L -> ... -> A, and another hardirq-unsafe M,
+	 * that B -> ... -> M. However M is **softirq-safe**, if we use exact
+	 * invert bits of M's usage_mask, we will find another lock N that is
+	 * **softirq-unsafe** and N -> ... -> A, however N -> .. -> M will not
+	 * cause a inversion deadlock.
 	 */
-	backward_mask = original_mask(target_entry1->class->usage_mask);
+	backward_mask = original_mask(target_entry1->class->usage_mask & LOCKF_ENABLED_IRQ_ALL);
 
 	ret = find_usage_backwards(&this, backward_mask, &target_entry);
 	if (bfs_error(ret)) {
Signature · 1 line
2.30.2
author maintainer
#191 lore raw

PATCH 191/700 KVM: s390: get rid of register asm usage

In reply to Greg Kroah-Hartman

From: Heiko Carstens <[email protected]>

[ Upstream commit 4fa3b91bdee1b08348c82660668ca0ca34e271ad ]

Using register asm statements has been proven to be very error prone, especially when using code instrumentation where gcc may add function calls, which clobbers register contents in an unexpected way.

Therefore get rid of register asm statements in kvm code, even though there is currently nothing wrong with them. This way we know for sure that this bug class won't be introduced here.

---
 arch/s390/kvm/kvm-s390.c | 18 +++++++++---------
 1 file changed, 9 insertions(+), 9 deletions(-)
arch/s390/kvm/kvm-s390.c +9 -9
index 24ad447e648c..dd7136b3ed9a 100644
@@ -323,31 +323,31 @@ static void allow_cpu_feat(unsigned long nr)
 
 static inline int plo_test_bit(unsigned char nr)
 {
-	register unsigned long r0 asm("0") = (unsigned long) nr | 0x100;
+	unsigned long function = (unsigned long)nr | 0x100;
 	int cc;
 
 	asm volatile(
+		"	lgr	0,%[function]\n"
 		/* Parameter registers are ignored for "test bit" */
 		"	plo	0,0,0,0(0)\n"
 		"	ipm	%0\n"
 		"	srl	%0,28\n"
 		: "=d" (cc)
-		: "d" (r0)
-		: "cc");
+		: [function] "d" (function)
+		: "cc", "0");
 	return cc == 0;
 }
 
 static __always_inline void __insn32_query(unsigned int opcode, u8 *query)
 {
-	register unsigned long r0 asm("0") = 0;	/* query function */
-	register unsigned long r1 asm("1") = (unsigned long) query;
-
 	asm volatile(
-		/* Parameter regs are ignored */
+		"	lghi	0,0\n"
+		"	lgr	1,%[query]\n"
+		/* Parameter registers are ignored */
 		"	.insn	rrf,%[opc] << 16,2,4,6,0\n"
 		:
-		: "d" (r0), "a" (r1), [opc] "i" (opcode)
-		: "cc", "memory");
+		: [query] "d" ((unsigned long)query), [opc] "i" (opcode)
+		: "cc", "memory", "0", "1");
 }
 
 #define INSN_SORTL 0xb938
Signature · 1 line
2.30.2
author maintainer
#192 lore raw

PATCH 192/700 regulator: mt6358: Fix vdram2 .vsel_mask

In reply to Greg Kroah-Hartman

From: Hsin-Hsiung Wang <[email protected]>

[ Upstream commit 50c9462edcbf900f3d5097ca3ad60171346124de ]

The valid vsel value are 0 and 12, so the .vsel_mask should be 0xf.

---
 drivers/regulator/mt6358-regulator.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/regulator/mt6358-regulator.c +1 -1
index 13cb6ac9a892..1d4eb5dc4fac 100644
@@ -457,7 +457,7 @@ static struct mt6358_regulator_info mt6358_regulators[] = {
 	MT6358_REG_FIXED("ldo_vaud28", VAUD28,
 			 MT6358_LDO_VAUD28_CON0, 0, 2800000),
 	MT6358_LDO("ldo_vdram2", VDRAM2, vdram2_voltages, vdram2_idx,
-		   MT6358_LDO_VDRAM2_CON0, 0, MT6358_LDO_VDRAM2_ELR0, 0x10, 0),
+		   MT6358_LDO_VDRAM2_CON0, 0, MT6358_LDO_VDRAM2_ELR0, 0xf, 0),
 	MT6358_LDO("ldo_vsim1", VSIM1, vsim_voltages, vsim_idx,
 		   MT6358_LDO_VSIM1_CON0, 0, MT6358_VSIM1_ANA_CON0, 0xf00, 8),
 	MT6358_LDO("ldo_vibr", VIBR, vibr_voltages, vibr_idx,
Signature · 1 line
2.30.2
author maintainer
#193 lore raw

PATCH 193/700 regulator: da9052: Ensure enough delay time for .set_voltage_time_sel

In reply to Greg Kroah-Hartman

From: Axel Lin <[email protected]>

[ Upstream commit a336dc8f683e5be794186b5643cd34cb28dd2c53 ]

Use DIV_ROUND_UP to prevent truncation by integer division issue. This ensures we return enough delay time.

Also fix returning negative value when new_sel < old_sel.

---
 drivers/regulator/da9052-regulator.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/regulator/da9052-regulator.c +2 -1
index e18d291c7f21..23fa429ebe76 100644
@@ -250,7 +250,8 @@ static int da9052_regulator_set_voltage_time_sel(struct regulator_dev *rdev,
 	case DA9052_ID_BUCK3:
 	case DA9052_ID_LDO2:
 	case DA9052_ID_LDO3:
-		ret = (new_sel - old_sel) * info->step_uV / 6250;
+		ret = DIV_ROUND_UP(abs(new_sel - old_sel) * info->step_uV,
+				   6250);
 		break;
 	}
 
Signature · 1 line
2.30.2
author maintainer
#194 lore raw

PATCH 194/700 media: Fix Media Controller API config checks

In reply to Greg Kroah-Hartman

From: Shuah Khan <[email protected]>

[ Upstream commit 50e7a31d30e8221632675abed3be306382324ca2 ]

Smatch static checker warns that "mdev" can be null:

sound/usb/media.c:287 snd_media_device_create()
warn: 'mdev' can also be NULL

If CONFIG_MEDIA_CONTROLLER is disabled, this file should not be included in the build.

The below conditions in the sound/usb/Makefile are in place to ensure that media.c isn't included in the build.

sound/usb/Makefile:
snd-usb-audio-$(CONFIG_SND_USB_AUDIO_USE_MEDIA_CONTROLLER) += media.o

select SND_USB_AUDIO_USE_MEDIA_CONTROLLER if MEDIA_CONTROLLER &&

       (MEDIA_SUPPORT=y || MEDIA_SUPPORT=SND_USB_AUDIO)

The following config check in include/media/media-dev-allocator.h is in place to enable the API only when CONFIG_MEDIA_CONTROLLER and CONFIG_USB are enabled.

#if defined(CONFIG_MEDIA_CONTROLLER) && defined(CONFIG_USB)

This check doesn't work as intended when CONFIG_USB=m. When CONFIG_USB=m, CONFIG_USB_MODULE is defined and CONFIG_USB is not. The above config check doesn't catch that CONFIG_USB is defined as a module and disables the API. This results in sound/usb enabling Media Controller specific ALSA driver code, while Media disables the Media Controller API.

Fix the problem requires two changes:

1. Change the check to use IS_ENABLED to detect when CONFIG_USB is enabled as a module or static. Since CONFIG_MEDIA_CONTROLLER is a bool, leave the check unchanged to be consistent with drivers/media/Makefile.

2. Change the drivers/media/mc/Makefile to include mc-dev-allocator.o in mc-objs when CONFIG_USB is enabled.

---
 drivers/media/mc/Makefile           | 2 +-
 include/media/media-dev-allocator.h | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)
drivers/media/mc/Makefile +1 -1
index 119037f0e686..2b7af42ba59c 100644
@@ -3,7 +3,7 @@
 mc-objs	:= mc-device.o mc-devnode.o mc-entity.o \
 	   mc-request.o
 
-ifeq ($(CONFIG_USB),y)
+ifneq ($(CONFIG_USB),)
 	mc-objs += mc-dev-allocator.o
 endif
 
include/media/media-dev-allocator.h +1 -1
index b35ea6062596..2ab54d426c64 100644
@@ -19,7 +19,7 @@
 
 struct usb_device;
 
-#if defined(CONFIG_MEDIA_CONTROLLER) && defined(CONFIG_USB)
+#if defined(CONFIG_MEDIA_CONTROLLER) && IS_ENABLED(CONFIG_USB)
 /**
  * media_device_usb_allocate() - Allocate and return struct &media device
  *
Signature · 1 line
2.30.2
author maintainer
#195 lore raw

PATCH 195/700 ACPI: video: use native backlight for GA401/GA502/GA503

In reply to Greg Kroah-Hartman

From: Luke D Jones <[email protected]>

[ Upstream commit 2dfbacc65d1d2eae587ccb6b93f6280542641858 ]

Force backlight control in these models to use the native interface at /sys/class/backlight/amdgpu_bl0.

---
 drivers/acpi/video_detect.c | 24 ++++++++++++++++++++++++
 1 file changed, 24 insertions(+)
drivers/acpi/video_detect.c +24 -0
index 83cd4c95faf0..33474fd96991 100644
@@ -385,6 +385,30 @@ static const struct dmi_system_id video_detect_dmi_table[] = {
 		DMI_MATCH(DMI_BOARD_NAME, "BA51_MV"),
 		},
 	},
+	{
+	.callback = video_detect_force_native,
+	.ident = "ASUSTeK COMPUTER INC. GA401",
+	.matches = {
+		DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
+		DMI_MATCH(DMI_PRODUCT_NAME, "GA401"),
+		},
+	},
+	{
+	.callback = video_detect_force_native,
+	.ident = "ASUSTeK COMPUTER INC. GA502",
+	.matches = {
+		DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
+		DMI_MATCH(DMI_PRODUCT_NAME, "GA502"),
+		},
+	},
+	{
+	.callback = video_detect_force_native,
+	.ident = "ASUSTeK COMPUTER INC. GA503",
+	.matches = {
+		DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
+		DMI_MATCH(DMI_PRODUCT_NAME, "GA503"),
+		},
+	},
 
 	/*
 	 * Desktops which falsely report a backlight and which our heuristics
Signature · 1 line
2.30.2
author maintainer
#196 lore raw

PATCH 196/700 HID: do not use down_interruptible() when unbinding devices

In reply to Greg Kroah-Hartman

From: Dmitry Torokhov <[email protected]>

[ Upstream commit f2145f8dc566c4f3b5a8deb58dcd12bed4e20194 ]

Action of unbinding driver from a device is not cancellable and should not fail, and driver core does not pay attention to the result of "remove" method, therefore using down_interruptible() in hid_device_remove() does not make sense.

---
 drivers/hid/hid-core.c | 10 +++-------
 1 file changed, 3 insertions(+), 7 deletions(-)
drivers/hid/hid-core.c +3 -7
index 0f69f35f2957..5550c943f985 100644
@@ -2306,12 +2306,8 @@ static int hid_device_remove(struct device *dev)
 {
 	struct hid_device *hdev = to_hid_device(dev);
 	struct hid_driver *hdrv;
-	int ret = 0;
 
-	if (down_interruptible(&hdev->driver_input_lock)) {
-		ret = -EINTR;
-		goto end;
-	}
+	down(&hdev->driver_input_lock);
 	hdev->io_started = false;
 
 	hdrv = hdev->driver;
@@ -2326,8 +2322,8 @@ static int hid_device_remove(struct device *dev)
 
 	if (!hdev->io_started)
 		up(&hdev->driver_input_lock);
-end:
-	return ret;
+
+	return 0;
 }
 
 static ssize_t modalias_show(struct device *dev, struct device_attribute *a,
Signature · 1 line
2.30.2
author maintainer
#197 lore raw

PATCH 197/700 EDAC/ti: Add missing MODULE_DEVICE_TABLE

In reply to Greg Kroah-Hartman

From: Bixuan Cui <[email protected]>

[ Upstream commit 0a37f32ba5272b2d4ec8c8d0f6b212b81b578f7e ]

The module misses MODULE_DEVICE_TABLE() for of_device_id tables and thus never autoloads on ID matches.

Add the missing declaration.

---
 drivers/edac/ti_edac.c | 1 +
 1 file changed, 1 insertion(+)
drivers/edac/ti_edac.c +1 -0
index e7eae20f83d1..169f96e51c29 100644
@@ -197,6 +197,7 @@ static const struct of_device_id ti_edac_of_match[] = {
 	{ .compatible = "ti,emif-dra7xx", .data = (void *)EMIF_TYPE_DRA7 },
 	{},
 };
+MODULE_DEVICE_TABLE(of, ti_edac_of_match);
 
 static int _emif_get_id(struct device_node *node)
 {
Signature · 1 line
2.30.2
author maintainer
#198 lore raw

PATCH 198/700 ACPI: PM: s2idle: Add missing LPS0 functions for AMD

In reply to Greg Kroah-Hartman

From: Alex Deucher <[email protected]>

[ Upstream commit f59a905b962c34642e862b5edec35c0eda72d70d ]

These are supposedly not required for AMD platforms, but at least some HP laptops seem to require it to properly turn off the keyboard backlight.

Based on a patch from Marcin Bachry <[email protected]>.

---
 drivers/acpi/x86/s2idle.c | 4 ++++
 1 file changed, 4 insertions(+)
drivers/acpi/x86/s2idle.c +4 -0
index 2b69536cdccb..2d7ddb8a8cb6 100644
@@ -42,6 +42,8 @@ static const struct acpi_device_id lps0_device_ids[] = {
 
 /* AMD */
 #define ACPI_LPS0_DSM_UUID_AMD      "e3f32452-febc-43ce-9039-932122d37721"
+#define ACPI_LPS0_ENTRY_AMD         2
+#define ACPI_LPS0_EXIT_AMD          3
 #define ACPI_LPS0_SCREEN_OFF_AMD    4
 #define ACPI_LPS0_SCREEN_ON_AMD     5
 
@@ -408,6 +410,7 @@ int acpi_s2idle_prepare_late(void)
 
 	if (acpi_s2idle_vendor_amd()) {
 		acpi_sleep_run_lps0_dsm(ACPI_LPS0_SCREEN_OFF_AMD);
+		acpi_sleep_run_lps0_dsm(ACPI_LPS0_ENTRY_AMD);
 	} else {
 		acpi_sleep_run_lps0_dsm(ACPI_LPS0_SCREEN_OFF);
 		acpi_sleep_run_lps0_dsm(ACPI_LPS0_ENTRY);
@@ -422,6 +425,7 @@ void acpi_s2idle_restore_early(void)
 		return;
 
 	if (acpi_s2idle_vendor_amd()) {
+		acpi_sleep_run_lps0_dsm(ACPI_LPS0_EXIT_AMD);
 		acpi_sleep_run_lps0_dsm(ACPI_LPS0_SCREEN_ON_AMD);
 	} else {
 		acpi_sleep_run_lps0_dsm(ACPI_LPS0_EXIT);
Signature · 1 line
2.30.2
author maintainer
#199 lore raw

PATCH 199/700 ACPI: processor idle: Fix up C-state latency if not ordered

In reply to Greg Kroah-Hartman

From: Mario Limonciello <[email protected]>

[ Upstream commit 65ea8f2c6e230bdf71fed0137cf9e9d1b307db32 ]

Generally, the C-state latency is provided by the _CST method or FADT, but some OEM platforms using AMD Picasso, Renoir, Van Gogh, and Cezanne set the C2 latency greater than C3's which causes the C2 state to be skipped.

That will block the core entering PC6, which prevents S0ix working properly on Linux systems.

In other operating systems, the latency values are not validated and this does not cause problems by skipping states.

To avoid this issue on Linux, detect when latencies are not an arithmetic progression and sort them.

---
 drivers/acpi/processor_idle.c | 40 +++++++++++++++++++++++++++++++++++
 1 file changed, 40 insertions(+)
drivers/acpi/processor_idle.c +40 -0
index 4e2d76b8b697..6790df5a2462 100644
@@ -16,6 +16,7 @@
 #include <linux/acpi.h>
 #include <linux/dmi.h>
 #include <linux/sched.h>       /* need_resched() */
+#include <linux/sort.h>
 #include <linux/tick.h>
 #include <linux/cpuidle.h>
 #include <linux/cpu.h>
@@ -388,10 +389,37 @@ static void acpi_processor_power_verify_c3(struct acpi_processor *pr,
 	return;
 }
 
+static int acpi_cst_latency_cmp(const void *a, const void *b)
+{
+	const struct acpi_processor_cx *x = a, *y = b;
+
+	if (!(x->valid && y->valid))
+		return 0;
+	if (x->latency > y->latency)
+		return 1;
+	if (x->latency < y->latency)
+		return -1;
+	return 0;
+}
+static void acpi_cst_latency_swap(void *a, void *b, int n)
+{
+	struct acpi_processor_cx *x = a, *y = b;
+	u32 tmp;
+
+	if (!(x->valid && y->valid))
+		return;
+	tmp = x->latency;
+	x->latency = y->latency;
+	y->latency = tmp;
+}
+
 static int acpi_processor_power_verify(struct acpi_processor *pr)
 {
 	unsigned int i;
 	unsigned int working = 0;
+	unsigned int last_latency = 0;
+	unsigned int last_type = 0;
+	bool buggy_latency = false;
 
 	pr->power.timer_broadcast_on_state = INT_MAX;
 
@@ -415,12 +443,24 @@ static int acpi_processor_power_verify(struct acpi_processor *pr)
 		}
 		if (!cx->valid)
 			continue;
+		if (cx->type >= last_type && cx->latency < last_latency)
+			buggy_latency = true;
+		last_latency = cx->latency;
+		last_type = cx->type;
 
 		lapic_timer_check_state(i, pr, cx);
 		tsc_check_state(cx->type);
 		working++;
 	}
 
+	if (buggy_latency) {
+		pr_notice("FW issue: working around C-state latencies out of order\n");
+		sort(&pr->power.states[1], max_cstate,
+		     sizeof(struct acpi_processor_cx),
+		     acpi_cst_latency_cmp,
+		     acpi_cst_latency_swap);
+	}
+
 	lapic_timer_propagate_broadcast(pr);
 
 	return (working);
Signature · 1 line
2.30.2
author maintainer
#200 lore raw

PATCH 200/700 hv_utils: Fix passing zero to PTR_ERR warning

In reply to Greg Kroah-Hartman

From: YueHaibing <[email protected]>

[ Upstream commit c6a8625fa4c6b0a97860d053271660ccedc3d1b3 ]

Sparse warn this:

drivers/hv/hv_util.c:753 hv_timesync_init() warn:
passing zero to 'PTR_ERR'

Use PTR_ERR_OR_ZERO instead of PTR_ERR to fix this.

---
 drivers/hv/hv_util.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/hv/hv_util.c +2 -2
index e4aefeb330da..136576cba26f 100644
@@ -750,8 +750,8 @@ static int hv_timesync_init(struct hv_util_service *srv)
 	 */
 	hv_ptp_clock = ptp_clock_register(&ptp_hyperv_info, NULL);
 	if (IS_ERR_OR_NULL(hv_ptp_clock)) {
-		pr_err("cannot register PTP clock: %ld\n",
-		       PTR_ERR(hv_ptp_clock));
+		pr_err("cannot register PTP clock: %d\n",
+		       PTR_ERR_OR_ZERO(hv_ptp_clock));
 		hv_ptp_clock = NULL;
 	}
 
Signature · 1 line
2.30.2
author maintainer
#201 lore raw

PATCH 201/700 lib: vsprintf: Fix handling of number field widths in vsscanf

In reply to Greg Kroah-Hartman

From: Richard Fitzgerald <[email protected]>

[ Upstream commit 900fdc4573766dd43b847b4f54bd4a1ee2bc7360 ]

The existing code attempted to handle numbers by doing a strto[u]l(), ignoring the field width, and then repeatedly dividing to extract the field out of the full converted value. If the string contains a run of valid digits longer than will fit in a long or long long, this would overflow and no amount of dividing can recover the correct value.

This patch fixes vsscanf() to obey number field widths when parsing the number.

A new _parse_integer_limit() is added that takes a limit for the number of characters to parse. The number field conversion in vsscanf is changed to use this new function.

If a number starts with a radix prefix, the field width must be long enough for at last one digit after the prefix. If not, it will be handled like this:

sscanf("0x4", "%1i", &i): i=0, scanning continues with the 'x' sscanf("0x4", "%2i", &i): i=0, scanning continues with the '4'

This is consistent with the observed behaviour of userland sscanf.

Note that this patch does NOT fix the problem of a single field value overflowing the target type. So for example:

  sscanf("123456789abcdef", "%x", &i);

Will not produce the correct result because the value obviously overflows INT_MAX. But sscanf will report a successful conversion.

Note that where a very large number is used to mean "unlimited", the value INT_MAX is used for consistency with the behaviour of vsnprintf().

---
 lib/kstrtox.c  | 13 ++++++--
 lib/kstrtox.h  |  2 ++
 lib/vsprintf.c | 82 +++++++++++++++++++++++++++++---------------------
 3 files changed, 60 insertions(+), 37 deletions(-)
lib/kstrtox.c +10 -3
index a118b0b1e9b2..0b5fe8b41173 100644
@@ -39,20 +39,22 @@ const char *_parse_integer_fixup_radix(const char *s, unsigned int *base)
 
 /*
  * Convert non-negative integer string representation in explicitly given radix
- * to an integer.
+ * to an integer. A maximum of max_chars characters will be converted.
+ *
  * Return number of characters consumed maybe or-ed with overflow bit.
  * If overflow occurs, result integer (incorrect) is still returned.
  *
  * Don't you dare use this function.
  */
-unsigned int _parse_integer(const char *s, unsigned int base, unsigned long long *p)
+unsigned int _parse_integer_limit(const char *s, unsigned int base, unsigned long long *p,
+				  size_t max_chars)
 {
 	unsigned long long res;
 	unsigned int rv;
 
 	res = 0;
 	rv = 0;
-	while (1) {
+	while (max_chars--) {
 		unsigned int c = *s;
 		unsigned int lc = c | 0x20; /* don't tolower() this line */
 		unsigned int val;
@@ -82,6 +84,11 @@ unsigned int _parse_integer(const char *s, unsigned int base, unsigned long long
 	return rv;
 }
 
+unsigned int _parse_integer(const char *s, unsigned int base, unsigned long long *p)
+{
+	return _parse_integer_limit(s, base, p, INT_MAX);
+}
+
 static int _kstrtoull(const char *s, unsigned int base, unsigned long long *res)
 {
 	unsigned long long _res;
lib/kstrtox.h +2 -0
index 3b4637bcd254..158c400ca865 100644
@@ -4,6 +4,8 @@
 
 #define KSTRTOX_OVERFLOW	(1U << 31)
 const char *_parse_integer_fixup_radix(const char *s, unsigned int *base);
+unsigned int _parse_integer_limit(const char *s, unsigned int base, unsigned long long *res,
+				  size_t max_chars);
 unsigned int _parse_integer(const char *s, unsigned int base, unsigned long long *res);
 
 #endif
lib/vsprintf.c +48 -34

113 lines not on this page · raw message

Signature · 1 line
2.30.2
author maintainer
#202 lore raw

PATCH 202/700 Input: goodix - platform/x86: touchscreen_dmi - Move upside down quirks to touchscreen_dmi.c

In reply to Greg Kroah-Hartman

From: Hans de Goede <[email protected]>

[ Upstream commit 5a6f0dbe621a5c20dc912ac474debf9f11129e03 ]

Move the DMI quirks for upside-down mounted Goodix touchscreens from drivers/input/touchscreen/goodix.c to drivers/platform/x86/touchscreen_dmi.c,
where all the other x86 touchscreen quirks live.

Note the touchscreen_dmi.c code attaches standard touchscreen device-properties to an i2c-client device based on a combination of a DMI match + a device-name match. I've verified that the: Teclast X98 Pro, WinBook TW100 and WinBook TW700 uses an ACPI devicename of "GDIX1001:00" based on acpidumps and/or dmesg output available on the web.

This patch was tested on a Teclast X89 tablet.

---
 drivers/input/touchscreen/goodix.c     | 52 ------------------------
 drivers/platform/x86/touchscreen_dmi.c | 56 ++++++++++++++++++++++++++
 2 files changed, 56 insertions(+), 52 deletions(-)
drivers/input/touchscreen/goodix.c +0 -52
index c682b028f0a2..4f53d3c57e69 100644
@@ -178,51 +178,6 @@ static const unsigned long goodix_irq_flags[] = {
 	IRQ_TYPE_LEVEL_HIGH,
 };
 
-/*
- * Those tablets have their coordinates origin at the bottom right
- * of the tablet, as if rotated 180 degrees
- */
-static const struct dmi_system_id rotated_screen[] = {
-#if defined(CONFIG_DMI) && defined(CONFIG_X86)
-	{
-		.ident = "Teclast X89",
-		.matches = {
-			/* tPAD is too generic, also match on bios date */
-			DMI_MATCH(DMI_BOARD_VENDOR, "TECLAST"),
-			DMI_MATCH(DMI_BOARD_NAME, "tPAD"),
-			DMI_MATCH(DMI_BIOS_DATE, "12/19/2014"),
-		},
-	},
-	{
-		.ident = "Teclast X98 Pro",
-		.matches = {
-			/*
-			 * Only match BIOS date, because the manufacturers
-			 * BIOS does not report the board name at all
-			 * (sometimes)...
-			 */
-			DMI_MATCH(DMI_BOARD_VENDOR, "TECLAST"),
-			DMI_MATCH(DMI_BIOS_DATE, "10/28/2015"),
-		},
-	},
-	{
-		.ident = "WinBook TW100",
-		.matches = {
-			DMI_MATCH(DMI_SYS_VENDOR, "WinBook"),
-			DMI_MATCH(DMI_PRODUCT_NAME, "TW100")
-		}
-	},
-	{
-		.ident = "WinBook TW700",
-		.matches = {
-			DMI_MATCH(DMI_SYS_VENDOR, "WinBook"),
-			DMI_MATCH(DMI_PRODUCT_NAME, "TW700")
-		},
-	},
-#endif
-	{}
-};
-
 static const struct dmi_system_id nine_bytes_report[] = {
 #if defined(CONFIG_DMI) && defined(CONFIG_X86)
 	{
@@ -1123,13 +1078,6 @@ static int goodix_configure_dev(struct goodix_ts_data *ts)
 				  ABS_MT_POSITION_Y, ts->prop.max_y);
 	}
 
-	if (dmi_check_system(rotated_screen)) {
-		ts->prop.invert_x = true;
-		ts->prop.invert_y = true;
-		dev_dbg(&ts->client->dev,
-			"Applying '180 degrees rotated screen' quirk\n");
-	}
-
 	if (dmi_check_system(nine_bytes_report)) {
 		ts->contact_size = 9;
 
drivers/platform/x86/touchscreen_dmi.c +56 -0

80 lines not on this page · raw message

Signature · 1 line
2.30.2
author maintainer
#203 lore raw

PATCH 203/700 platform/x86: touchscreen_dmi: Add an extra entry for the upside down Goodix touchscreen on Teclast X89 tablets

In reply to Greg Kroah-Hartman

From: Hans de Goede <[email protected]>

[ Upstream commit a22e3803f2a4d947ff0083a9448a169269ea0f62 ]

Teclast X89 tablets come in 2 versions, with Windows pre-installed and with Android pre-installed. These 2 versions have different DMI strings.

Add a match for the DMI strings used by the Android version BIOS.

Note the Android version BIOS has a bug in the DSDT where no IRQ is provided, so for the touchscreen to work a DSDT override fixing this is necessary as well.

---
 drivers/platform/x86/touchscreen_dmi.c | 8 ++++++++
 1 file changed, 8 insertions(+)
drivers/platform/x86/touchscreen_dmi.c +8 -0
index 222f6c9f0b45..bbd8d80230cd 100644
@@ -1312,6 +1312,14 @@ const struct dmi_system_id touchscreen_dmi_table[] = {
 			DMI_MATCH(DMI_BOARD_NAME, "X3 Plus"),
 		},
 	},
+	{
+		/* Teclast X89 (Android version / BIOS) */
+		.driver_data = (void *)&gdix1001_00_upside_down_data,
+		.matches = {
+			DMI_MATCH(DMI_BOARD_VENDOR, "WISKY"),
+			DMI_MATCH(DMI_BOARD_NAME, "3G062i"),
+		},
+	},
 	{
 		/* Teclast X89 (Windows version / BIOS) */
 		.driver_data = (void *)&gdix1001_01_upside_down_data,
Signature · 1 line
2.30.2
author maintainer
#204 lore raw

PATCH 204/700 platform/x86: touchscreen_dmi: Add info for the Goodix GT912 panel of TM800A550L tablets

In reply to Greg Kroah-Hartman

From: Hans de Goede <[email protected]>

[ Upstream commit fcd8cf0e3e48f4c66af82c8e799c37cb0cccffe0 ]

The Bay Trail Glavey TM800A550L tablet, which ships with Android installed from the factory, uses a GT912 touchscreen controller which needs to have its firmware uploaded by the OS to work (this is a first for a x86 based device with a Goodix touchscreen controller).

Add a touchscreen_dmi entry for this which specifies the filenames to use for the firmware and config files needed for this.

Note this matches on a GDIX1001 ACPI HID, while the original DSDT uses a HID of GODX0911. For the touchscreen to work on these devices a DSDT override is necessary to fix a missing IRQ and broken GPIO settings in the ACPI-resources for the touchscreen. This override also changes the HID to the standard GDIX1001 id typically used for Goodix touchscreens. The DSDT override is available here:
https://fedorapeople.org/~jwrdegoede/glavey-tm800a550l-dsdt-override/

---
 drivers/platform/x86/touchscreen_dmi.c | 21 +++++++++++++++++++++
 1 file changed, 21 insertions(+)
drivers/platform/x86/touchscreen_dmi.c +21 -0
index bbd8d80230cd..b47f6821615e 100644
@@ -316,6 +316,18 @@ static const struct ts_dmi_data gdix1001_01_upside_down_data = {
 	.properties	= gdix1001_upside_down_props,
 };
 
+static const struct property_entry glavey_tm800a550l_props[] = {
+	PROPERTY_ENTRY_STRING("firmware-name", "gt912-glavey-tm800a550l.fw"),
+	PROPERTY_ENTRY_STRING("goodix,config-name", "gt912-glavey-tm800a550l.cfg"),
+	PROPERTY_ENTRY_U32("goodix,main-clk", 54),
+	{ }
+};
+
+static const struct ts_dmi_data glavey_tm800a550l_data = {
+	.acpi_name	= "GDIX1001:00",
+	.properties	= glavey_tm800a550l_props,
+};
+
 static const struct property_entry gp_electronic_t701_props[] = {
 	PROPERTY_ENTRY_U32("touchscreen-size-x", 960),
 	PROPERTY_ENTRY_U32("touchscreen-size-y", 640),
@@ -1029,6 +1041,15 @@ const struct dmi_system_id touchscreen_dmi_table[] = {
 			DMI_MATCH(DMI_PRODUCT_NAME, "eSTAR BEAUTY HD Intel Quad core"),
 		},
 	},
+	{	/* Glavey TM800A550L */
+		.driver_data = (void *)&glavey_tm800a550l_data,
+		.matches = {
+			DMI_MATCH(DMI_BOARD_VENDOR, "AMI Corporation"),
+			DMI_MATCH(DMI_BOARD_NAME, "Aptio CRB"),
+			/* Above strings are too generic, also match on BIOS version */
+			DMI_MATCH(DMI_BIOS_VERSION, "ZY-8-BI-PX4S70VTR400-X423B-005-D"),
+		},
+	},
 	{
 		/* GP-electronic T701 */
 		.driver_data = (void *)&gp_electronic_t701_data,
Signature · 1 line
2.30.2
author maintainer
#205 lore raw

PATCH 205/700 ACPI: EC: Make more Asus laptops use ECDT _GPE

In reply to Greg Kroah-Hartman

From: Chris Chiu <[email protected]>

[ Upstream commit 6306f0431914beaf220634ad36c08234006571d5 ]

More ASUS laptops have the _GPE define in the DSDT table with a different value than the _GPE number in the ECDT.

This is causing media keys not working on ASUS X505BA/BP, X542BA/BP

Add model info to the quirks list.

---
 drivers/acpi/ec.c | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)
drivers/acpi/ec.c +16 -0
index 13565629ce0a..e8c5da2b964a 100644
@@ -1846,6 +1846,22 @@ static const struct dmi_system_id ec_dmi_table[] __initconst = {
 	DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
 	DMI_MATCH(DMI_PRODUCT_NAME, "GL702VMK"),}, NULL},
 	{
+	ec_honor_ecdt_gpe, "ASUSTeK COMPUTER INC. X505BA", {
+	DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
+	DMI_MATCH(DMI_PRODUCT_NAME, "X505BA"),}, NULL},
+	{
+	ec_honor_ecdt_gpe, "ASUSTeK COMPUTER INC. X505BP", {
+	DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
+	DMI_MATCH(DMI_PRODUCT_NAME, "X505BP"),}, NULL},
+	{
+	ec_honor_ecdt_gpe, "ASUSTeK COMPUTER INC. X542BA", {
+	DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
+	DMI_MATCH(DMI_PRODUCT_NAME, "X542BA"),}, NULL},
+	{
+	ec_honor_ecdt_gpe, "ASUSTeK COMPUTER INC. X542BP", {
+	DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
+	DMI_MATCH(DMI_PRODUCT_NAME, "X542BP"),}, NULL},
+	{
 	ec_honor_ecdt_gpe, "ASUS X550VXK", {
 	DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
 	DMI_MATCH(DMI_PRODUCT_NAME, "X550VXK"),}, NULL},
Signature · 1 line
2.30.2
author maintainer
#206 lore raw

PATCH 206/700 block_dump: remove block_dump feature in mark_inode_dirty()

In reply to Greg Kroah-Hartman

From: zhangyi (F) <[email protected]>

[ Upstream commit 12e0613715e1cf305fffafaf0e89d810d9a85cc0 ]

block_dump is an old debugging interface, one of it's functions is used to print the information about who write which file on disk. If we enable block_dump through /proc/sys/vm/block_dump and turn on debug log level, we can gather information about write process name, target file name and disk from kernel message. This feature is realized in block_dump___mark_inode_dirty(), it print above information into kernel message directly when marking inode dirty, so it is noisy and can easily trigger log storm. At the same time, get the dentry refcount is also not safe, we found it will lead to deadlock on ext4 file system with data=journal mode.

After tracepoints has been introduced into the kernel, we got a tracepoint in __mark_inode_dirty(), which is a better replacement of block_dump___mark_inode_dirty(). The only downside is that it only trace the inode number and not a file name, but it probably doesn't matter because the original printed file name in block_dump is not accurate in some cases, and we can still find it through the inode number and device id. So this patch delete the dirting inode part of block_dump feature.

---
 fs/fs-writeback.c | 25 -------------------------
 1 file changed, 25 deletions(-)
fs/fs-writeback.c +0 -25
index e91980f49388..7c46d1588a19 100644
@@ -2205,28 +2205,6 @@ int dirtytime_interval_handler(struct ctl_table *table, int write,
 	return ret;
 }
 
-static noinline void block_dump___mark_inode_dirty(struct inode *inode)
-{
-	if (inode->i_ino || strcmp(inode->i_sb->s_id, "bdev")) {
-		struct dentry *dentry;
-		const char *name = "?";
-
-		dentry = d_find_alias(inode);
-		if (dentry) {
-			spin_lock(&dentry->d_lock);
-			name = (const char *) dentry->d_name.name;
-		}
-		printk(KERN_DEBUG
-		       "%s(%d): dirtied inode %lu (%s) on %s\n",
-		       current->comm, task_pid_nr(current), inode->i_ino,
-		       name, inode->i_sb->s_id);
-		if (dentry) {
-			spin_unlock(&dentry->d_lock);
-			dput(dentry);
-		}
-	}
-}
-
 /**
  * __mark_inode_dirty -	internal function to mark an inode dirty
  *
@@ -2296,9 +2274,6 @@ void __mark_inode_dirty(struct inode *inode, int flags)
 	    (dirtytime && (inode->i_state & I_DIRTY_INODE)))
 		return;
 
-	if (unlikely(block_dump))
-		block_dump___mark_inode_dirty(inode);
-
 	spin_lock(&inode->i_lock);
 	if (dirtytime && (inode->i_state & I_DIRTY_INODE))
 		goto out_unlock_inode;
Signature · 1 line
2.30.2
author maintainer
#207 lore raw

PATCH 207/700 blk-mq: grab rq->refcount before calling ->fn in blk_mq_tagset_busy_iter

In reply to Greg Kroah-Hartman

From: Ming Lei <[email protected]>

[ Upstream commit 2e315dc07df009c3e29d6926871f62a30cfae394 ]

Grab rq->refcount before calling ->fn in blk_mq_tagset_busy_iter(), and this way will prevent the request from being re-used when ->fn is running. The approach is same as what we do during handling timeout.

Fix request use-after-free(UAF) related with completion race or queue releasing:

- If one rq is referred before rq->q is frozen, then queue won't be frozen before the request is released during iteration.

- If one rq is referred after rq->q is frozen, refcount_inc_not_zero() will return false, and we won't iterate over this request.

However, still one request UAF not covered: refcount_inc_not_zero() may read one freed request, and it will be handled in next patch.

---
 block/blk-mq-tag.c | 44 +++++++++++++++++++++++++++++++++-----------
 block/blk-mq.c     | 14 +++++++++-----
 block/blk-mq.h     |  1 +
 3 files changed, 43 insertions(+), 16 deletions(-)
block/blk-mq-tag.c +33 -11
index 9c92053e704d..6772c3728865 100644
@@ -199,6 +199,16 @@ struct bt_iter_data {
 	bool reserved;
 };
 
+static struct request *blk_mq_find_and_get_req(struct blk_mq_tags *tags,
+		unsigned int bitnr)
+{
+	struct request *rq = tags->rqs[bitnr];
+
+	if (!rq || !refcount_inc_not_zero(&rq->ref))
+		return NULL;
+	return rq;
+}
+
 static bool bt_iter(struct sbitmap *bitmap, unsigned int bitnr, void *data)
 {
 	struct bt_iter_data *iter_data = data;
@@ -206,18 +216,22 @@ static bool bt_iter(struct sbitmap *bitmap, unsigned int bitnr, void *data)
 	struct blk_mq_tags *tags = hctx->tags;
 	bool reserved = iter_data->reserved;
 	struct request *rq;
+	bool ret = true;
 
 	if (!reserved)
 		bitnr += tags->nr_reserved_tags;
-	rq = tags->rqs[bitnr];
-
 	/*
 	 * We can hit rq == NULL here, because the tagging functions
 	 * test and set the bit before assigning ->rqs[].
 	 */
-	if (rq && rq->q == hctx->queue && rq->mq_hctx == hctx)
-		return iter_data->fn(hctx, rq, iter_data->data, reserved);
-	return true;
+	rq = blk_mq_find_and_get_req(tags, bitnr);
+	if (!rq)
+		return true;
+
+	if (rq->q == hctx->queue && rq->mq_hctx == hctx)
+		ret = iter_data->fn(hctx, rq, iter_data->data, reserved);
+	blk_mq_put_rq_ref(rq);
+	return ret;
 }
 
 /**
@@ -264,6 +278,8 @@ static bool bt_tags_iter(struct sbitmap *bitmap, unsigned int bitnr, void *data)
 	struct blk_mq_tags *tags = iter_data->tags;
 	bool reserved = iter_data->flags & BT_TAG_ITER_RESERVED;
 	struct request *rq;
+	bool ret = true;
+	bool iter_static_rqs = !!(iter_data->flags & BT_TAG_ITER_STATIC_RQS);
 
 	if (!reserved)
 		bitnr += tags->nr_reserved_tags;
@@ -272,16 +288,19 @@ static bool bt_tags_iter(struct sbitmap *bitmap, unsigned int bitnr, void *data)
 	 * We can hit rq == NULL here, because the tagging functions
 	 * test and set the bit before assigning ->rqs[].
 	 */
-	if (iter_data->flags & BT_TAG_ITER_STATIC_RQS)
+	if (iter_static_rqs)
 		rq = tags->static_rqs[bitnr];
 	else
-		rq = tags->rqs[bitnr];
+		rq = blk_mq_find_and_get_req(tags, bitnr);
 	if (!rq)
 		return true;
-	if ((iter_data->flags & BT_TAG_ITER_STARTED) &&
-	    !blk_mq_request_started(rq))
-		return true;
-	return iter_data->fn(rq, iter_data->data, reserved);
+
+	if (!(iter_data->flags & BT_TAG_ITER_STARTED) ||
+	    blk_mq_request_started(rq))
+		ret = iter_data->fn(rq, iter_data->data, reserved);
+	if (!iter_static_rqs)
+		blk_mq_put_rq_ref(rq);
+	return ret;
 }
 
 /**
@@ -348,6 +367,9 @@ void blk_mq_all_tag_iter(struct blk_mq_tags *tags, busy_tag_iter_fn *fn,
  *		indicates whether or not @rq is a reserved request. Return
  *		true to continue iterating tags, false to stop.
  * @priv:	Will be passed as second argument to @fn.
+ *
+ * We grab one request reference before calling @fn and release it after
+ * @fn returns.
  */
 void blk_mq_tagset_busy_iter(struct blk_mq_tag_set *tagset,
 		busy_tag_iter_fn *fn, void *priv)
block/blk-mq.c +9 -5
index 0e120547ccb7..d5370ab2eb31 100644
@@ -908,6 +908,14 @@ static bool blk_mq_req_expired(struct request *rq, unsigned long *next)
 	return false;
 }
 
+void blk_mq_put_rq_ref(struct request *rq)
+{
+	if (is_flush_rq(rq, rq->mq_hctx))
+		rq->end_io(rq, 0);
+	else if (refcount_dec_and_test(&rq->ref))
+		__blk_mq_free_request(rq);
+}
+
 static bool blk_mq_check_expired(struct blk_mq_hw_ctx *hctx,
 		struct request *rq, void *priv, bool reserved)
 {
@@ -941,11 +949,7 @@ static bool blk_mq_check_expired(struct blk_mq_hw_ctx *hctx,
 	if (blk_mq_req_expired(rq, next))
 		blk_mq_rq_timed_out(rq, reserved);
 
-	if (is_flush_rq(rq, hctx))
-		rq->end_io(rq, 0);
-	else if (refcount_dec_and_test(&rq->ref))
-		__blk_mq_free_request(rq);
-
+	blk_mq_put_rq_ref(rq);
 	return true;
 }
 
block/blk-mq.h +1 -0
index 3616453ca28c..143afe42c63a 100644
@@ -47,6 +47,7 @@ void blk_mq_add_to_requeue_list(struct request *rq, bool at_head,
 void blk_mq_flush_busy_ctxs(struct blk_mq_hw_ctx *hctx, struct list_head *list);
 struct request *blk_mq_dequeue_from_ctx(struct blk_mq_hw_ctx *hctx,
 					struct blk_mq_ctx *start);
+void blk_mq_put_rq_ref(struct request *rq);
 
 /*
  * Internal helpers for allocating/freeing the request map
Signature · 1 line
2.30.2
author maintainer
#208 lore raw

PATCH 208/700 blk-mq: clear stale request in tags->rq[] before freeing one request pool

In reply to Greg Kroah-Hartman

From: Ming Lei <[email protected]>

[ Upstream commit bd63141d585bef14f4caf111f6d0e27fe2300ec6 ]

refcount_inc_not_zero() in bt_tags_iter() still may read one freed request.

Fix the issue by the following approach:

1) hold a per-tags spinlock when reading ->rqs[tag] and calling refcount_inc_not_zero in bt_tags_iter()

2) clearing stale request referred via ->rqs[tag] before freeing request pool, the per-tags spinlock is held for clearing stale ->rq[tag]

So after we cleared stale requests, bt_tags_iter() won't observe freed request any more, also the clearing will wait for pending request reference.

The idea of clearing ->rqs[] is borrowed from John Garry's previous patch and one recent David's patch.

---
 block/blk-mq-tag.c |  9 +++++++--
 block/blk-mq-tag.h |  6 ++++++
 block/blk-mq.c     | 46 +++++++++++++++++++++++++++++++++++++++++-----
 3 files changed, 54 insertions(+), 7 deletions(-)
block/blk-mq-tag.c +7 -2
index 6772c3728865..c4f2f6c123ae 100644
@@ -202,10 +202,14 @@ struct bt_iter_data {
 static struct request *blk_mq_find_and_get_req(struct blk_mq_tags *tags,
 		unsigned int bitnr)
 {
-	struct request *rq = tags->rqs[bitnr];
+	struct request *rq;
+	unsigned long flags;
 
+	spin_lock_irqsave(&tags->lock, flags);
+	rq = tags->rqs[bitnr];
 	if (!rq || !refcount_inc_not_zero(&rq->ref))
-		return NULL;
+		rq = NULL;
+	spin_unlock_irqrestore(&tags->lock, flags);
 	return rq;
 }
 
@@ -538,6 +542,7 @@ struct blk_mq_tags *blk_mq_init_tags(unsigned int total_tags,
 
 	tags->nr_tags = total_tags;
 	tags->nr_reserved_tags = reserved_tags;
+	spin_lock_init(&tags->lock);
 
 	if (flags & BLK_MQ_F_TAG_HCTX_SHARED)
 		return tags;
block/blk-mq-tag.h +6 -0
index 7d3e6b333a4a..f887988e5ef6 100644
@@ -20,6 +20,12 @@ struct blk_mq_tags {
 	struct request **rqs;
 	struct request **static_rqs;
 	struct list_head page_list;
+
+	/*
+	 * used to clear request reference in rqs[] before freeing one
+	 * request pool
+	 */
+	spinlock_t lock;
 };
 
 extern struct blk_mq_tags *blk_mq_init_tags(unsigned int nr_tags,
block/blk-mq.c +41 -5

64 lines not on this page · raw message

Signature · 1 line
2.30.2
author maintainer
#209 lore raw

PATCH 209/700 fs: dlm: reconnect if socket error report occurs

In reply to Greg Kroah-Hartman

From: Alexander Aring <[email protected]>

[ Upstream commit ba868d9deaab2bb1c09e50650127823925154802 ]

This patch will change the reconnect handling that if an error occurs if a socket error callback is occurred. This will also handle reconnects in a non blocking connecting case which is currently missing. If error ECONNREFUSED is reported we delay the reconnect by one second.

---
 fs/dlm/lowcomms.c | 60 ++++++++++++++++++++++++++++++-----------------
 1 file changed, 39 insertions(+), 21 deletions(-)
fs/dlm/lowcomms.c +39 -21
index 45c2fdaf34c4..01b672cee783 100644
@@ -79,6 +79,8 @@ struct connection {
 #define CF_CLOSING 8
 #define CF_SHUTDOWN 9
 #define CF_CONNECTED 10
+#define CF_RECONNECT 11
+#define CF_DELAY_CONNECT 12
 	struct list_head writequeue;  /* List of outgoing writequeue_entries */
 	spinlock_t writequeue_lock;
 	void (*connect_action) (struct connection *);	/* What to do to connect */
@@ -87,6 +89,7 @@ struct connection {
 #define MAX_CONNECT_RETRIES 3
 	struct hlist_node list;
 	struct connection *othercon;
+	struct connection *sendcon;
 	struct work_struct rwork; /* Receive workqueue */
 	struct work_struct swork; /* Send workqueue */
 	wait_queue_head_t shutdown_wait; /* wait for graceful shutdown */
@@ -584,6 +587,22 @@ static void lowcomms_error_report(struct sock *sk)
 				   dlm_config.ci_tcp_port, sk->sk_err,
 				   sk->sk_err_soft);
 	}
+
+	/* below sendcon only handling */
+	if (test_bit(CF_IS_OTHERCON, &con->flags))
+		con = con->sendcon;
+
+	switch (sk->sk_err) {
+	case ECONNREFUSED:
+		set_bit(CF_DELAY_CONNECT, &con->flags);
+		break;
+	default:
+		break;
+	}
+
+	if (!test_and_set_bit(CF_RECONNECT, &con->flags))
+		queue_work(send_workqueue, &con->swork);
+
 out:
 	read_unlock_bh(&sk->sk_callback_lock);
 	if (orig_report)
@@ -701,6 +720,8 @@ static void close_connection(struct connection *con, bool and_other,
 	con->rx_leftover = 0;
 	con->retries = 0;
 	clear_bit(CF_CONNECTED, &con->flags);
+	clear_bit(CF_DELAY_CONNECT, &con->flags);
+	clear_bit(CF_RECONNECT, &con->flags);
 	mutex_unlock(&con->sock_mutex);
 	clear_bit(CF_CLOSING, &con->flags);
 }
@@ -839,18 +860,15 @@ out_resched:
 
 out_close:
 	mutex_unlock(&con->sock_mutex);
-	if (ret != -EAGAIN) {
-		/* Reconnect when there is something to send */
+	if (ret == 0) {
 		close_connection(con, false, true, false);
-		if (ret == 0) {
-			log_print("connection %p got EOF from %d",
-				  con, con->nodeid);
-			/* handling for tcp shutdown */
-			clear_bit(CF_SHUTDOWN, &con->flags);
-			wake_up(&con->shutdown_wait);
-			/* signal to breaking receive worker */
-			ret = -1;
-		}
+		log_print("connection %p got EOF from %d",
+			  con, con->nodeid);
+		/* handling for tcp shutdown */
+		clear_bit(CF_SHUTDOWN, &con->flags);
+		wake_up(&con->shutdown_wait);
+		/* signal to breaking receive worker */
+		ret = -1;
 	}
 	return ret;
 }
@@ -933,6 +951,7 @@ static int accept_from_sock(struct listen_connection *con)
 			}
 
 			newcon->othercon = othercon;
+			othercon->sendcon = newcon;
 		} else {
 			/* close other sock con if we have something new */
 			close_connection(othercon, false, true, false);
@@ -1478,7 +1497,7 @@ static void send_to_sock(struct connection *con)
 				cond_resched();
 				goto out;
 			} else if (ret < 0)
-				goto send_error;
+				goto out;
 		}
 
 		/* Don't starve people filling buffers */
@@ -1495,14 +1514,6 @@ out:
 	mutex_unlock(&con->sock_mutex);
 	return;
 
-send_error:
-	mutex_unlock(&con->sock_mutex);
-	close_connection(con, false, false, true);
-	/* Requeue the send work. When the work daemon runs again, it will try
-	   a new connection, then call this function again. */
-	queue_work(send_workqueue, &con->swork);
-	return;
-
 out_connect:
 	mutex_unlock(&con->sock_mutex);
 	queue_work(send_workqueue, &con->swork);
@@ -1574,8 +1585,15 @@ static void process_send_sockets(struct work_struct *work)
 	struct connection *con = container_of(work, struct connection, swork);
 
 	clear_bit(CF_WRITE_PENDING, &con->flags);
-	if (con->sock == NULL) /* not mutex protected so check it inside too */
+
+	if (test_and_clear_bit(CF_RECONNECT, &con->flags))
+		close_connection(con, false, false, true);
+
+	if (con->sock == NULL) { /* not mutex protected so check it inside too */
+		if (test_and_clear_bit(CF_DELAY_CONNECT, &con->flags))
+			msleep(1000);
 		con->connect_action(con);
+	}
 	if (!list_empty(&con->writequeue))
 		send_to_sock(con);
 }
Signature · 1 line
2.30.2
author maintainer
#210 lore raw

PATCH 210/700 fs: dlm: cancel work sync othercon

In reply to Greg Kroah-Hartman

From: Alexander Aring <[email protected]>

[ Upstream commit c6aa00e3d20c2767ba3f57b64eb862572b9744b3 ]

These rx tx flags arguments are for signaling close_connection() from which worker they are called. Obviously the receive worker cannot cancel itself and vice versa for swork. For the othercon the receive worker should only be used, however to avoid deadlocks we should pass the same flags as the original close_connection() was called.

---
 fs/dlm/lowcomms.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
fs/dlm/lowcomms.c +1 -1
index 01b672cee783..7e6736c70e11 100644
@@ -714,7 +714,7 @@ static void close_connection(struct connection *con, bool and_other,
 
 	if (con->othercon && and_other) {
 		/* Will only re-enter once. */
-		close_connection(con->othercon, false, true, true);
+		close_connection(con->othercon, false, tx, rx);
 	}
 
 	con->rx_leftover = 0;
Signature · 1 line
2.30.2
author maintainer
#211 lore raw

PATCH 211/700 random32: Fix implicit truncation warning in prandom_seed_state()

In reply to Greg Kroah-Hartman

From: Richard Fitzgerald <[email protected]>

[ Upstream commit d327ea15a305024ef0085252fa3657bbb1ce25f5 ]

sparse generates the following warning:

include/linux/prandom.h:114:45: sparse: sparse: cast truncates bits from constant value

This is because the 64-bit seed value is manipulated and then placed in a u32, causing an implicit cast and truncation. A forced cast to u32 doesn't prevent this warning, which is reasonable because a typecast doesn't prove that truncation was expected.

Logical-AND the value with 0xffffffff to make explicit that truncation to 32-bit is intended.

---
 include/linux/prandom.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
include/linux/prandom.h +1 -1
index bbf4b4ad61df..056d31317e49 100644
@@ -111,7 +111,7 @@ static inline u32 __seed(u32 x, u32 m)
  */
 static inline void prandom_seed_state(struct rnd_state *state, u64 seed)
 {
-	u32 i = (seed >> 32) ^ (seed << 10) ^ seed;
+	u32 i = ((seed >> 32) ^ (seed << 10) ^ seed) & 0xffffffffUL;
 
 	state->s1 = __seed(i,   2U);
 	state->s2 = __seed(i,   8U);
Signature · 1 line
2.30.2
author maintainer
#212 lore raw

PATCH 212/700 open: dont silently ignore unknown O-flags in openat2()

In reply to Greg Kroah-Hartman

From: Christian Brauner <[email protected]>

[ Upstream commit cfe80306a0dd6d363934913e47c3f30d71b721e5 ]

The new openat2() syscall verifies that no unknown O-flag values are set and returns an error to userspace if they are while the older open syscalls like open() and openat() simply ignore unknown flag values:

  #define O_FLAG_CURRENTLY_INVALID (1 << 31)
  struct open_how how = {
          .flags = O_RDONLY | O_FLAG_CURRENTLY_INVALID,
          .resolve = 0,
  };

  /* fails */
  fd = openat2(-EBADF, "/dev/null", &how, sizeof(how));

  /* succeeds */
  fd = openat(-EBADF, "/dev/null", O_RDONLY | O_FLAG_CURRENTLY_INVALID);

However, openat2() silently truncates the upper 32 bits meaning:

  #define O_FLAG_CURRENTLY_INVALID_LOWER32 (1 << 31)
  #define O_FLAG_CURRENTLY_INVALID_UPPER32 (1 << 40)

  struct open_how how_lowe32 = {
          .flags = O_RDONLY | O_FLAG_CURRENTLY_INVALID_LOWER32,
  };

  struct open_how how_upper32 = {
          .flags = O_RDONLY | O_FLAG_CURRENTLY_INVALID_UPPER32,
  };

  /* fails */
  fd = openat2(-EBADF, "/dev/null", &how_lower32, sizeof(how_lower32));

  /* succeeds */
  fd = openat2(-EBADF, "/dev/null", &how_upper32, sizeof(how_upper32));

Fix this by preventing the immediate truncation in build_open_flags().

There's a snafu here though stripping FMODE_* directly from flags would cause the upper 32 bits to be truncated as well due to integer promotion rules since FMODE_* is unsigned int, O_* are signed ints (yuck).

In addition, struct open_flags currently defines flags to be 32 bit which is reasonable. If we simply were to bump it to 64 bit we would need to change a lot of code preemptively which doesn't seem worth it. So simply add a compile-time check verifying that all currently known O_* flags are within the 32 bit range and fail to build if they aren't anymore.

This change shouldn't regress old open syscalls since they silently truncate any unknown values anyway. It is a tiny semantic change for openat2() but it is very unlikely people pass ing > 32 bit unknown flags and the syscall is relatively new too.

---
 fs/open.c | 14 +++++++++++---
 1 file changed, 11 insertions(+), 3 deletions(-)
fs/open.c +11 -3
index e53af13b5835..53bc0573c0ec 100644
@@ -1002,12 +1002,20 @@ inline struct open_how build_open_how(int flags, umode_t mode)
 
 inline int build_open_flags(const struct open_how *how, struct open_flags *op)
 {
-	int flags = how->flags;
+	u64 flags = how->flags;
+	u64 strip = FMODE_NONOTIFY | O_CLOEXEC;
 	int lookup_flags = 0;
 	int acc_mode = ACC_MODE(flags);
 
-	/* Must never be set by userspace */
-	flags &= ~(FMODE_NONOTIFY | O_CLOEXEC);
+	BUILD_BUG_ON_MSG(upper_32_bits(VALID_OPEN_FLAGS),
+			 "struct open_flags doesn't yet handle flags > 32 bits");
+
+	/*
+	 * Strip flags that either shouldn't be set by userspace like
+	 * FMODE_NONOTIFY or that aren't relevant in determining struct
+	 * open_flags like O_CLOEXEC.
+	 */
+	flags &= ~strip;
 
 	/*
 	 * Older syscalls implicitly clear all of the invalid flags or argument
Signature · 1 line
2.30.2
author maintainer
#213 lore raw

PATCH 213/700 drivers: hv: Fix missing error code in vmbus_connect()

In reply to Greg Kroah-Hartman

From: Jiapeng Chong <[email protected]>

[ Upstream commit 9de6655cc5a6a1febc514465c87c24a0e96d8dba ]

Eliminate the follow smatch warning:

drivers/hv/connection.c:236 vmbus_connect() warn: missing error code 'ret'.

---
 drivers/hv/connection.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/hv/connection.c +3 -1
index c83612cddb99..425bf85ed1a0 100644
@@ -229,8 +229,10 @@ int vmbus_connect(void)
 	 */
 
 	for (i = 0; ; i++) {
-		if (i == ARRAY_SIZE(vmbus_versions))
+		if (i == ARRAY_SIZE(vmbus_versions)) {
+			ret = -EDOM;
 			goto cleanup;
+		}
 
 		version = vmbus_versions[i];
 		if (version > max_version)
Signature · 1 line
2.30.2
author maintainer
#214 lore raw

PATCH 214/700 fs: dlm: fix lowcomms_start error case

In reply to Greg Kroah-Hartman

From: Alexander Aring <[email protected]>

[ Upstream commit fcef0e6c27ce109d2c617aa12f0bfd9f7ff47d38 ]

This patch fixes the error path handling in lowcomms_start(). We need to cleanup some static allocated data structure and cleanup possible workqueue if these have started.

---
 fs/dlm/lowcomms.c | 15 ++++++++++++---
 1 file changed, 12 insertions(+), 3 deletions(-)
fs/dlm/lowcomms.c +12 -3
index 7e6736c70e11..d2a0ea0acca3 100644
@@ -1600,10 +1600,15 @@ static void process_send_sockets(struct work_struct *work)
 
 static void work_stop(void)
 {
-	if (recv_workqueue)
+	if (recv_workqueue) {
 		destroy_workqueue(recv_workqueue);
-	if (send_workqueue)
+		recv_workqueue = NULL;
+	}
+
+	if (send_workqueue) {
 		destroy_workqueue(send_workqueue);
+		send_workqueue = NULL;
+	}
 }
 
 static int work_start(void)
@@ -1620,6 +1625,7 @@ static int work_start(void)
 	if (!send_workqueue) {
 		log_print("can't start dlm_send");
 		destroy_workqueue(recv_workqueue);
+		recv_workqueue = NULL;
 		return -ENOMEM;
 	}
 
@@ -1751,7 +1757,7 @@ int dlm_lowcomms_start(void)
 
 	error = work_start();
 	if (error)
-		goto fail;
+		goto fail_local;
 
 	dlm_allow_conn = 1;
 
@@ -1768,6 +1774,9 @@ int dlm_lowcomms_start(void)
 fail_unlisten:
 	dlm_allow_conn = 0;
 	dlm_close_sock(&listen_con.sock);
+	work_stop();
+fail_local:
+	deinit_local();
 fail:
 	return error;
 }
Signature · 1 line
2.30.2
author maintainer
#215 lore raw

PATCH 215/700 fs: dlm: fix memory leak when fenced

In reply to Greg Kroah-Hartman

From: Alexander Aring <[email protected]>

[ Upstream commit 700ab1c363c7b54c9ea3222379b33fc00ab02f7b ]

I got some kmemleak report when a node was fenced. The user space tool dlm_controld will therefore run some rmdir() in dlm configfs which was triggering some memleaks. This patch stores the sps and cms attributes which stores some handling for subdirectories of the configfs cluster entry and free them if they get released as the parent directory gets freed.

unreferenced object 0xffff88810d9e3e00 (size 192):
comm "dlm_controld", pid 342, jiffies 4294698126 (age 55438.801s) hex dump (first 32 bytes):

    00 00 00 00 00 00 00 00 73 70 61 63 65 73 00 00  ........spaces..
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
  backtrace:
    [<00000000db8b640b>] make_cluster+0x5d/0x360
    [<000000006a571db4>] configfs_mkdir+0x274/0x730
    [<00000000b094501c>] vfs_mkdir+0x27e/0x340
    [<0000000058b0adaf>] do_mkdirat+0xff/0x1b0
    [<00000000d1ffd156>] do_syscall_64+0x40/0x80
    [<00000000ab1408c8>] entry_SYSCALL_64_after_hwframe+0x44/0xae
unreferenced object 0xffff88810d9e3a00 (size 192):
  comm "dlm_controld", pid 342, jiffies 4294698126 (age 55438.801s)
  hex dump (first 32 bytes):
    00 00 00 00 00 00 00 00 63 6f 6d 6d 73 00 00 00  ........comms...
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
  backtrace:
    [<00000000a7ef6ad2>] make_cluster+0x82/0x360
    [<000000006a571db4>] configfs_mkdir+0x274/0x730
    [<00000000b094501c>] vfs_mkdir+0x27e/0x340
    [<0000000058b0adaf>] do_mkdirat+0xff/0x1b0
    [<00000000d1ffd156>] do_syscall_64+0x40/0x80
    [<00000000ab1408c8>] entry_SYSCALL_64_after_hwframe+0x44/0xae
---
 fs/dlm/config.c | 9 +++++++++
 1 file changed, 9 insertions(+)
fs/dlm/config.c +9 -0
index 88d95d96e36c..52bcda64172a 100644
@@ -79,6 +79,9 @@ struct dlm_cluster {
 	unsigned int cl_new_rsb_count;
 	unsigned int cl_recover_callbacks;
 	char cl_cluster_name[DLM_LOCKSPACE_LEN];
+
+	struct dlm_spaces *sps;
+	struct dlm_comms *cms;
 };
 
 static struct dlm_cluster *config_item_to_cluster(struct config_item *i)
@@ -409,6 +412,9 @@ static struct config_group *make_cluster(struct config_group *g,
 	if (!cl || !sps || !cms)
 		goto fail;
 
+	cl->sps = sps;
+	cl->cms = cms;
+
 	config_group_init_type_name(&cl->group, name, &cluster_type);
 	config_group_init_type_name(&sps->ss_group, "spaces", &spaces_type);
 	config_group_init_type_name(&cms->cs_group, "comms", &comms_type);
@@ -458,6 +464,9 @@ static void drop_cluster(struct config_group *g, struct config_item *i)
 static void release_cluster(struct config_item *i)
 {
 	struct dlm_cluster *cl = config_item_to_cluster(i);
+
+	kfree(cl->sps);
+	kfree(cl->cms);
 	kfree(cl);
 }
 
Signature · 1 line
2.30.2
author maintainer
#216 lore raw

PATCH 216/700 ACPICA: Fix memory leak caused by _CID repair function

In reply to Greg Kroah-Hartman

From: Erik Kaneda <[email protected]>

[ Upstream commit c27bac0314131b11bccd735f7e8415ac6444b667 ]

ACPICA commit 180cb53963aa876c782a6f52cc155d951b26051a

According to the ACPI spec, _CID returns a package containing hardware ID's. Each element of an ASL package contains a reference count from the parent package as well as the element itself.

Name (TEST, Package() {
"String object" // this package element has a reference count of 2
})

A memory leak was caused in the _CID repair function because it did not decrement the reference count created by the package. Fix the memory leak by calling acpi_ut_remove_reference on _CID package elements that represent a hardware ID (_HID).

---
 drivers/acpi/acpica/nsrepair2.c | 7 +++++++
 1 file changed, 7 insertions(+)
drivers/acpi/acpica/nsrepair2.c +7 -0
index 14b71b41e845..38e10ab976e6 100644
@@ -379,6 +379,13 @@ acpi_ns_repair_CID(struct acpi_evaluate_info *info,
 
 			(*element_ptr)->common.reference_count =
 			    original_ref_count;
+
+			/*
+			 * The original_element holds a reference from the package object
+			 * that represents _HID. Since a new element was created by _HID,
+			 * remove the reference from the _CID package.
+			 */
+			acpi_ut_remove_reference(original_element);
 		}
 
 		element_ptr++;
Signature · 1 line
2.30.2
author maintainer
#217 lore raw

PATCH 217/700 ACPI: bus: Call kobject_put() in acpi_init() error path

In reply to Greg Kroah-Hartman

From: Hanjun Guo <[email protected]>

[ Upstream commit 4ac7a817f1992103d4e68e9837304f860b5e7300 ]

Although the system will not be in a good condition or it will not boot if acpi_bus_init() fails, it is still necessary to put the kobject in the error path before returning to avoid leaking memory.

[ rjw: Subject and changelog edits ]
---
 drivers/acpi/bus.c | 1 +
 1 file changed, 1 insertion(+)
drivers/acpi/bus.c +1 -0
index a4bd673934c0..44b4f02e2c6d 100644
@@ -1321,6 +1321,7 @@ static int __init acpi_init(void)
 
 	result = acpi_bus_init();
 	if (result) {
+		kobject_put(acpi_kobj);
 		disable_acpi();
 		return result;
 	}
Signature · 1 line
2.30.2
author maintainer
#218 lore raw

PATCH 218/700 ACPI: resources: Add checks for ACPI IRQ override

In reply to Greg Kroah-Hartman

From: Hui Wang <[email protected]>

[ Upstream commit 0ec4e55e9f571f08970ed115ec0addc691eda613 ]

The laptop keyboard doesn't work on many MEDION notebooks, but the keyboard works well under Windows and Unix.

Through debugging, we found this log in the dmesg:

ACPI: IRQ 1 override to edge, high
pnp 00:03: Plug and Play ACPI device, IDs PNP0303 (active)

And we checked the IRQ definition in the DSDT, it is:

IRQ (Level, ActiveLow, Exclusive, )
{1}

So the BIOS defines the keyboard IRQ to Level_Low, but the Linux kernel override it to Edge_High. If the Linux kernel is modified to skip the IRQ override, the keyboard will work normally.

Quoted 1 line

From the existing comment in acpi_dev_get_irqresource(), the override

function only needs to be called when IRQ() or IRQNoFlags() is used to populate the resource descriptor, and according to Section 6.4.2.1 of ACPI 6.4 [1], if IRQ() is empty or IRQNoFlags() is used, the IRQ is High true, edge sensitive and non-shareable. ACPICA also assumes that to be the case (see acpi_rs_set_irq[] in rsirq.c).

In accordance with the above, check 3 additional conditions (EdgeSensitive, ActiveHigh and Exclusive) when deciding whether or not to treat an ACPI_RESOURCE_TYPE_IRQ resource as "legacy", in which case the IRQ override is applicable to it.

---
 drivers/acpi/resource.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)
drivers/acpi/resource.c +8 -1
index 20a7892c6d3f..f0b2c3791253 100644
@@ -423,6 +423,13 @@ static void acpi_dev_get_irqresource(struct resource *res, u32 gsi,
 	}
 }
 
+static bool irq_is_legacy(struct acpi_resource_irq *irq)
+{
+	return irq->triggering == ACPI_EDGE_SENSITIVE &&
+		irq->polarity == ACPI_ACTIVE_HIGH &&
+		irq->shareable == ACPI_EXCLUSIVE;
+}
+
 /**
  * acpi_dev_resource_interrupt - Extract ACPI interrupt resource information.
  * @ares: Input ACPI resource object.
@@ -461,7 +468,7 @@ bool acpi_dev_resource_interrupt(struct acpi_resource *ares, int index,
 		}
 		acpi_dev_get_irqresource(res, irq->interrupts[index],
 					 irq->triggering, irq->polarity,
-					 irq->shareable, true);
+					 irq->shareable, irq_is_legacy(irq));
 		break;
 	case ACPI_RESOURCE_TYPE_EXTENDED_IRQ:
 		ext_irq = &ares->data.extended_irq;
Signature · 1 line
2.30.2
author maintainer
#219 lore raw

PATCH 219/700 HID: hid-input: add Surface Go battery quirk

In reply to Greg Kroah-Hartman

From: Zoltan Tamas Vajda <[email protected]>

[ Upstream commit b5539722eb832441f309642fe5102cc3536f92b8 ]

The Elantech touchscreen/digitizer in the Surface Go mistakenly reports having a battery. This results in a low battery message every time you try to use the pen.

This patch adds a quirk to ignore the non-existent battery and gets rid of the false low battery messages.

---
 drivers/hid/hid-ids.h   | 1 +
 drivers/hid/hid-input.c | 2 ++
 2 files changed, 3 insertions(+)
drivers/hid/hid-ids.h +1 -0
index 03978111d944..06168f485722 100644
@@ -397,6 +397,7 @@
 #define USB_DEVICE_ID_HP_X2_10_COVER	0x0755
 #define I2C_DEVICE_ID_HP_SPECTRE_X360_15	0x2817
 #define USB_DEVICE_ID_ASUS_UX550_TOUCHSCREEN	0x2706
+#define I2C_DEVICE_ID_SURFACE_GO_TOUCHSCREEN	0x261A
 
 #define USB_VENDOR_ID_ELECOM		0x056e
 #define USB_DEVICE_ID_ELECOM_BM084	0x0061
drivers/hid/hid-input.c +2 -0
index e982d8173c9c..bf5e728258c1 100644
@@ -326,6 +326,8 @@ static const struct hid_device_id hid_battery_quirks[] = {
 	  HID_BATTERY_QUIRK_IGNORE },
 	{ HID_I2C_DEVICE(USB_VENDOR_ID_ELAN, I2C_DEVICE_ID_HP_SPECTRE_X360_15),
 	  HID_BATTERY_QUIRK_IGNORE },
+	{ HID_I2C_DEVICE(USB_VENDOR_ID_ELAN, I2C_DEVICE_ID_SURFACE_GO_TOUCHSCREEN),
+	  HID_BATTERY_QUIRK_IGNORE },
 	{}
 };
 
Signature · 1 line
2.30.2
author maintainer
#220 lore raw

PATCH 220/700 HID: sony: fix freeze when inserting ghlive ps3/wii dongles

In reply to Greg Kroah-Hartman

From: Pascal Giard <[email protected]>

[ Upstream commit fb1a79a6b6e1223ddb18f12aa35e36f832da2290 ]

This commit fixes a freeze on insertion of a Guitar Hero Live PS3/WiiU USB dongle. Indeed, with the current implementation, inserting one of those USB dongles will lead to a hard freeze. I apologize for not catching this earlier, it didn't occur on my old laptop.

While the issue was isolated to memory alloc/free, I could not figure out why it causes a freeze. So this patch fixes this issue by simplifying memory allocation and usage.

We remind that for the dongle to work properly, a control URB needs to be sent periodically. We used to alloc/free the URB each time this URB needed to be sent.

With this patch, the memory for the URB is allocated on the probe, reused for as long as the dongle is plugged in, and freed once the dongle is unplugged.

---
 drivers/hid/hid-sony.c | 98 +++++++++++++++++++++---------------------
 1 file changed, 49 insertions(+), 49 deletions(-)
drivers/hid/hid-sony.c +49 -49
index 8319b0ce385a..b3722c51ec78 100644
@@ -597,9 +597,8 @@ struct sony_sc {
 	/* DS4 calibration data */
 	struct ds4_calibration_data ds4_calib_data[6];
 	/* GH Live */
+	struct urb *ghl_urb;
 	struct timer_list ghl_poke_timer;
-	struct usb_ctrlrequest *ghl_cr;
-	u8 *ghl_databuf;
 };
 
 static void sony_set_leds(struct sony_sc *sc);
@@ -625,66 +624,54 @@ static inline void sony_schedule_work(struct sony_sc *sc,
 
 static void ghl_magic_poke_cb(struct urb *urb)
 {
-	if (urb) {
-		/* Free sc->ghl_cr and sc->ghl_databuf allocated in
-		 * ghl_magic_poke()
-		 */
-		kfree(urb->setup_packet);
-		kfree(urb->transfer_buffer);
-	}
+	struct sony_sc *sc = urb->context;
+
+	if (urb->status < 0)
+		hid_err(sc->hdev, "URB transfer failed : %d", urb->status);
+
+	mod_timer(&sc->ghl_poke_timer, jiffies + GHL_GUITAR_POKE_INTERVAL*HZ);
 }
 
 static void ghl_magic_poke(struct timer_list *t)
 {
+	int ret;
 	struct sony_sc *sc = from_timer(sc, t, ghl_poke_timer);
 
-	int ret;
+	ret = usb_submit_urb(sc->ghl_urb, GFP_ATOMIC);
+	if (ret < 0)
+		hid_err(sc->hdev, "usb_submit_urb failed: %d", ret);
+}
+
+static int ghl_init_urb(struct sony_sc *sc, struct usb_device *usbdev)
+{
+	struct usb_ctrlrequest *cr;
+	u16 poke_size;
+	u8 *databuf;
 	unsigned int pipe;
-	struct urb *urb;
-	struct usb_device *usbdev = to_usb_device(sc->hdev->dev.parent->parent);
-	const u16 poke_size =
-		ARRAY_SIZE(ghl_ps3wiiu_magic_data);
 
+	poke_size = ARRAY_SIZE(ghl_ps3wiiu_magic_data);
 	pipe = usb_sndctrlpipe(usbdev, 0);
 
-	if (!sc->ghl_cr) {
-		sc->ghl_cr = kzalloc(sizeof(*sc->ghl_cr), GFP_ATOMIC);
-		if (!sc->ghl_cr)
-			goto resched;
-	}
-
-	if (!sc->ghl_databuf) {
-		sc->ghl_databuf = kzalloc(poke_size, GFP_ATOMIC);
-		if (!sc->ghl_databuf)
-			goto resched;
-	}
+	cr = devm_kzalloc(&sc->hdev->dev, sizeof(*cr), GFP_ATOMIC);
+	if (cr == NULL)
+		return -ENOMEM;
 
-	urb = usb_alloc_urb(0, GFP_ATOMIC);
-	if (!urb)
-		goto resched;
+	databuf = devm_kzalloc(&sc->hdev->dev, poke_size, GFP_ATOMIC);
+	if (databuf == NULL)
+		return -ENOMEM;
 
-	sc->ghl_cr->bRequestType =
+	cr->bRequestType =
 		USB_RECIP_INTERFACE | USB_TYPE_CLASS | USB_DIR_OUT;
-	sc->ghl_cr->bRequest = USB_REQ_SET_CONFIGURATION;
-	sc->ghl_cr->wValue = cpu_to_le16(ghl_ps3wiiu_magic_value);
-	sc->ghl_cr->wIndex = 0;
-	sc->ghl_cr->wLength = cpu_to_le16(poke_size);
-	memcpy(sc->ghl_databuf, ghl_ps3wiiu_magic_data, poke_size);
-
+	cr->bRequest = USB_REQ_SET_CONFIGURATION;
+	cr->wValue = cpu_to_le16(ghl_ps3wiiu_magic_value);
+	cr->wIndex = 0;
+	cr->wLength = cpu_to_le16(poke_size);
+	memcpy(databuf, ghl_ps3wiiu_magic_data, poke_size);
 	usb_fill_control_urb(
-		urb, usbdev, pipe,
-		(unsigned char *) sc->ghl_cr, sc->ghl_databuf,
-		poke_size, ghl_magic_poke_cb, NULL);
-	ret = usb_submit_urb(urb, GFP_ATOMIC);
-	if (ret < 0) {
-		kfree(sc->ghl_databuf);
-		kfree(sc->ghl_cr);
-	}
-	usb_free_urb(urb);
-
-resched:
-	/* Reschedule for next time */
-	mod_timer(&sc->ghl_poke_timer, jiffies + GHL_GUITAR_POKE_INTERVAL*HZ);
+		sc->ghl_urb, usbdev, pipe,
+		(unsigned char *) cr, databuf, poke_size,
+		ghl_magic_poke_cb, sc);
+	return 0;
 }
 
 static int guitar_mapping(struct hid_device *hdev, struct hid_input *hi,
@@ -2981,6 +2968,7 @@ static int sony_probe(struct hid_device *hdev, const struct hid_device_id *id)
 	int ret;
 	unsigned long quirks = id->driver_data;
 	struct sony_sc *sc;
+	struct usb_device *usbdev;
 	unsigned int connect_mask = HID_CONNECT_DEFAULT;
 
 	if (!strcmp(hdev->name, "FutureMax Dance Mat"))
@@ -3000,6 +2988,7 @@ static int sony_probe(struct hid_device *hdev, const struct hid_device_id *id)
 	sc->quirks = quirks;
 	hid_set_drvdata(hdev, sc);
 	sc->hdev = hdev;
+	usbdev = to_usb_device(sc->hdev->dev.parent->parent);
 
 	ret = hid_parse(hdev);
 	if (ret) {
@@ -3042,6 +3031,15 @@ static int sony_probe(struct hid_device *hdev, const struct hid_device_id *id)
 	}
 
 	if (sc->quirks & GHL_GUITAR_PS3WIIU) {
+		sc->ghl_urb = usb_alloc_urb(0, GFP_ATOMIC);
+		if (!sc->ghl_urb)
+			return -ENOMEM;
+		ret = ghl_init_urb(sc, usbdev);
+		if (ret) {
+			hid_err(hdev, "error preparing URB\n");
+			return ret;
+		}
+
 		timer_setup(&sc->ghl_poke_timer, ghl_magic_poke, 0);
 		mod_timer(&sc->ghl_poke_timer,
 			  jiffies + GHL_GUITAR_POKE_INTERVAL*HZ);
@@ -3054,8 +3052,10 @@ static void sony_remove(struct hid_device *hdev)
 {
 	struct sony_sc *sc = hid_get_drvdata(hdev);
 
-	if (sc->quirks & GHL_GUITAR_PS3WIIU)
+	if (sc->quirks & GHL_GUITAR_PS3WIIU) {
 		del_timer_sync(&sc->ghl_poke_timer);
+		usb_free_urb(sc->ghl_urb);
+	}
 
 	hid_hw_close(hdev);
 
Signature · 1 line
2.30.2
author maintainer
#221 lore raw

PATCH 221/700 block: fix race between adding/removing rq qos and normal IO

In reply to Greg Kroah-Hartman

From: Ming Lei <[email protected]>

[ Upstream commit 2cafe29a8d03f02a3d16193bdaae2f3e82a423f9 ]

Yi reported several kernel panics on:

[16687.001777] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000008
...
[16687.163549] pc : __rq_qos_track+0x38/0x60

or

[  997.690455] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000020
...
[  997.850347] pc : __rq_qos_done+0x2c/0x50

Turns out it is caused by race between adding rq qos(wbt) and normal IO because rq_qos_add can be run when IO is being submitted, fix this issue by freezing queue before adding/deleting rq qos to queue.

rq_qos_exit() needn't to freeze queue because it is called after queue has been frozen.

iolatency calls rq_qos_add() during allocating queue, so freezing won't add delay because queue usage refcount works at atomic mode at that time.

iocost calls rq_qos_add() when writing cgroup attribute file, that is fine to freeze queue at that time since we usually freeze queue when storing to queue sysfs attribute, meantime iocost only exists on the root cgroup.

wbt_init calls it in blk_register_queue() and queue sysfs attribute store(queue_wb_lat_store() when write it 1st time in case of !BLK_WBT_MQ), the following patch will speedup the queue freezing in wbt_init.

---
 block/blk-rq-qos.h | 24 ++++++++++++++++++++++++
 1 file changed, 24 insertions(+)
block/blk-rq-qos.h +24 -0
index 2bc43e94f4c4..2bcb3495e376 100644
@@ -7,6 +7,7 @@
 #include <linux/blk_types.h>
 #include <linux/atomic.h>
 #include <linux/wait.h>
+#include <linux/blk-mq.h>
 
 #include "blk-mq-debugfs.h"
 
@@ -99,8 +100,21 @@ static inline void rq_wait_init(struct rq_wait *rq_wait)
 
 static inline void rq_qos_add(struct request_queue *q, struct rq_qos *rqos)
 {
+	/*
+	 * No IO can be in-flight when adding rqos, so freeze queue, which
+	 * is fine since we only support rq_qos for blk-mq queue.
+	 *
+	 * Reuse ->queue_lock for protecting against other concurrent
+	 * rq_qos adding/deleting
+	 */
+	blk_mq_freeze_queue(q);
+
+	spin_lock_irq(&q->queue_lock);
 	rqos->next = q->rq_qos;
 	q->rq_qos = rqos;
+	spin_unlock_irq(&q->queue_lock);
+
+	blk_mq_unfreeze_queue(q);
 
 	if (rqos->ops->debugfs_attrs)
 		blk_mq_debugfs_register_rqos(rqos);
@@ -110,12 +124,22 @@ static inline void rq_qos_del(struct request_queue *q, struct rq_qos *rqos)
 {
 	struct rq_qos **cur;
 
+	/*
+	 * See comment in rq_qos_add() about freezing queue & using
+	 * ->queue_lock.
+	 */
+	blk_mq_freeze_queue(q);
+
+	spin_lock_irq(&q->queue_lock);
 	for (cur = &q->rq_qos; *cur; cur = &(*cur)->next) {
 		if (*cur == rqos) {
 			*cur = rqos->next;
 			break;
 		}
 	}
+	spin_unlock_irq(&q->queue_lock);
+
+	blk_mq_unfreeze_queue(q);
 
 	blk_mq_debugfs_unregister_rqos(rqos);
 }
Signature · 1 line
2.30.2
author maintainer
#222 lore raw

PATCH 222/700 platform/x86: asus-nb-wmi: Revert "Drop duplicate DMI quirk structures"

In reply to Greg Kroah-Hartman

From: Luke D. Jones <[email protected]>

[ Upstream commit 98c0c85b1040db24f0d04d3e1d315c6c7b05cc07 ]

This is a preparation revert for reverting the "add support for ASUS ROG Zephyrus G14 and G15" change. This reverts
commit 67186653c903 ("platform/x86: asus-nb-wmi: Drop duplicate DMI quirk structures")

---
 drivers/platform/x86/asus-nb-wmi.c | 23 ++++++++++++++---------
 1 file changed, 14 insertions(+), 9 deletions(-)
drivers/platform/x86/asus-nb-wmi.c +14 -9
index d41d7ad14be0..b07b1288346e 100644
@@ -110,7 +110,12 @@ static struct quirk_entry quirk_asus_forceals = {
 	.wmi_force_als_set = true,
 };
 
-static struct quirk_entry quirk_asus_vendor_backlight = {
+static struct quirk_entry quirk_asus_ga401i = {
+	.wmi_backlight_power = true,
+	.wmi_backlight_set_devstate = true,
+};
+
+static struct quirk_entry quirk_asus_ga502i = {
 	.wmi_backlight_power = true,
 	.wmi_backlight_set_devstate = true,
 };
@@ -432,7 +437,7 @@ static const struct dmi_system_id asus_quirks[] = {
 			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
 			DMI_MATCH(DMI_PRODUCT_NAME, "GA401IH"),
 		},
-		.driver_data = &quirk_asus_vendor_backlight,
+		.driver_data = &quirk_asus_ga401i,
 	},
 	{
 		.callback = dmi_matched,
@@ -441,7 +446,7 @@ static const struct dmi_system_id asus_quirks[] = {
 			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
 			DMI_MATCH(DMI_PRODUCT_NAME, "GA401II"),
 		},
-		.driver_data = &quirk_asus_vendor_backlight,
+		.driver_data = &quirk_asus_ga401i,
 	},
 	{
 		.callback = dmi_matched,
@@ -450,7 +455,7 @@ static const struct dmi_system_id asus_quirks[] = {
 			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
 			DMI_MATCH(DMI_PRODUCT_NAME, "GA401IU"),
 		},
-		.driver_data = &quirk_asus_vendor_backlight,
+		.driver_data = &quirk_asus_ga401i,
 	},
 	{
 		.callback = dmi_matched,
@@ -459,7 +464,7 @@ static const struct dmi_system_id asus_quirks[] = {
 			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
 			DMI_MATCH(DMI_PRODUCT_NAME, "GA401IV"),
 		},
-		.driver_data = &quirk_asus_vendor_backlight,
+		.driver_data = &quirk_asus_ga401i,
 	},
 	{
 		.callback = dmi_matched,
@@ -468,7 +473,7 @@ static const struct dmi_system_id asus_quirks[] = {
 			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
 			DMI_MATCH(DMI_PRODUCT_NAME, "GA401IVC"),
 		},
-		.driver_data = &quirk_asus_vendor_backlight,
+		.driver_data = &quirk_asus_ga401i,
 	},
 		{
 		.callback = dmi_matched,
@@ -477,7 +482,7 @@ static const struct dmi_system_id asus_quirks[] = {
 			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
 			DMI_MATCH(DMI_PRODUCT_NAME, "GA502II"),
 		},
-		.driver_data = &quirk_asus_vendor_backlight,
+		.driver_data = &quirk_asus_ga502i,
 	},
 	{
 		.callback = dmi_matched,
@@ -486,7 +491,7 @@ static const struct dmi_system_id asus_quirks[] = {
 			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
 			DMI_MATCH(DMI_PRODUCT_NAME, "GA502IU"),
 		},
-		.driver_data = &quirk_asus_vendor_backlight,
+		.driver_data = &quirk_asus_ga502i,
 	},
 	{
 		.callback = dmi_matched,
@@ -495,7 +500,7 @@ static const struct dmi_system_id asus_quirks[] = {
 			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
 			DMI_MATCH(DMI_PRODUCT_NAME, "GA502IV"),
 		},
-		.driver_data = &quirk_asus_vendor_backlight,
+		.driver_data = &quirk_asus_ga502i,
 	},
 	{
 		.callback = dmi_matched,
Signature · 1 line
2.30.2
author maintainer
#223 lore raw

PATCH 223/700 platform/x86: asus-nb-wmi: Revert "add support for ASUS ROG Zephyrus G14 and G15"

In reply to Greg Kroah-Hartman

From: Luke D. Jones <[email protected]>

[ Upstream commit 28117f3a5c3c8375a3304af76357d5bf9cf30f0b ]

The quirks added to asus-nb-wmi for the ASUS ROG Zephyrus G14 and G15 are wrong, they tell the asus-wmi code to use the vendor specific WMI backlight interface. But there is no such interface on these laptops.

As a side effect, these quirks stop the acpi_video driver to register since they make acpi_video_get_backlight_type() return acpi_backlight_vendor, leaving only the native AMD backlight driver in place, which is the one we want. This happy coincidence is being replaced with a new quirk in drivers/acpi/video_detect.c which actually sets the backlight_type to acpi_backlight_native fixinf this properly. This reverts
commit 13bceda68fb9 ("platform/x86: asus-nb-wmi: add support for ASUS ROG Zephyrus G14 and G15").

---
 drivers/platform/x86/asus-nb-wmi.c | 82 ------------------------------
 1 file changed, 82 deletions(-)
drivers/platform/x86/asus-nb-wmi.c +0 -82
index b07b1288346e..0cb927f0f301 100644
@@ -110,16 +110,6 @@ static struct quirk_entry quirk_asus_forceals = {
 	.wmi_force_als_set = true,
 };
 
-static struct quirk_entry quirk_asus_ga401i = {
-	.wmi_backlight_power = true,
-	.wmi_backlight_set_devstate = true,
-};
-
-static struct quirk_entry quirk_asus_ga502i = {
-	.wmi_backlight_power = true,
-	.wmi_backlight_set_devstate = true,
-};
-
 static struct quirk_entry quirk_asus_use_kbd_dock_devid = {
 	.use_kbd_dock_devid = true,
 };
@@ -430,78 +420,6 @@ static const struct dmi_system_id asus_quirks[] = {
 		},
 		.driver_data = &quirk_asus_forceals,
 	},
-	{
-		.callback = dmi_matched,
-		.ident = "ASUSTeK COMPUTER INC. GA401IH",
-		.matches = {
-			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
-			DMI_MATCH(DMI_PRODUCT_NAME, "GA401IH"),
-		},
-		.driver_data = &quirk_asus_ga401i,
-	},
-	{
-		.callback = dmi_matched,
-		.ident = "ASUSTeK COMPUTER INC. GA401II",
-		.matches = {
-			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
-			DMI_MATCH(DMI_PRODUCT_NAME, "GA401II"),
-		},
-		.driver_data = &quirk_asus_ga401i,
-	},
-	{
-		.callback = dmi_matched,
-		.ident = "ASUSTeK COMPUTER INC. GA401IU",
-		.matches = {
-			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
-			DMI_MATCH(DMI_PRODUCT_NAME, "GA401IU"),
-		},
-		.driver_data = &quirk_asus_ga401i,
-	},
-	{
-		.callback = dmi_matched,
-		.ident = "ASUSTeK COMPUTER INC. GA401IV",
-		.matches = {
-			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
-			DMI_MATCH(DMI_PRODUCT_NAME, "GA401IV"),
-		},
-		.driver_data = &quirk_asus_ga401i,
-	},
-	{
-		.callback = dmi_matched,
-		.ident = "ASUSTeK COMPUTER INC. GA401IVC",
-		.matches = {
-			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
-			DMI_MATCH(DMI_PRODUCT_NAME, "GA401IVC"),
-		},
-		.driver_data = &quirk_asus_ga401i,
-	},
-		{
-		.callback = dmi_matched,
-		.ident = "ASUSTeK COMPUTER INC. GA502II",
-		.matches = {
-			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
-			DMI_MATCH(DMI_PRODUCT_NAME, "GA502II"),
-		},
-		.driver_data = &quirk_asus_ga502i,
-	},
-	{
-		.callback = dmi_matched,
-		.ident = "ASUSTeK COMPUTER INC. GA502IU",
-		.matches = {
-			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
-			DMI_MATCH(DMI_PRODUCT_NAME, "GA502IU"),
-		},
-		.driver_data = &quirk_asus_ga502i,
-	},
-	{
-		.callback = dmi_matched,
-		.ident = "ASUSTeK COMPUTER INC. GA502IV",
-		.matches = {
-			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
-			DMI_MATCH(DMI_PRODUCT_NAME, "GA502IV"),
-		},
-		.driver_data = &quirk_asus_ga502i,
-	},
 	{
 		.callback = dmi_matched,
 		.ident = "Asus Transformer T100TA / T100HA / T100CHI",
Signature · 1 line
2.30.2
author maintainer
#224 lore raw

PATCH 224/700 platform/x86: toshiba_acpi: Fix missing error code in toshiba_acpi_setup_keyboard()

In reply to Greg Kroah-Hartman

From: Jiapeng Chong <[email protected]>

[ Upstream commit 28e367127718a9cb85d615a71e152f7acee41bfc ]

The error code is missing in this code scenario, add the error code '-EINVAL' to the return value 'error'.

Eliminate the follow smatch warning:

drivers/platform/x86/toshiba_acpi.c:2834 toshiba_acpi_setup_keyboard()
warn: missing error code 'error'.

---
 drivers/platform/x86/toshiba_acpi.c | 1 +
 1 file changed, 1 insertion(+)
drivers/platform/x86/toshiba_acpi.c +1 -0
index fa7232ad8c39..352508d30467 100644
@@ -2831,6 +2831,7 @@ static int toshiba_acpi_setup_keyboard(struct toshiba_acpi_dev *dev)
 
 	if (!dev->info_supported && !dev->system_event_supported) {
 		pr_warn("No hotkey query interface found\n");
+		error = -EINVAL;
 		goto err_remove_filter;
 	}
 
Signature · 1 line
2.30.2
author maintainer
#225 lore raw

PATCH 225/700 nvme-pci: fix var. type for increasing cq_head

In reply to Greg Kroah-Hartman

From: JK Kim <[email protected]>

[ Upstream commit a0aac973a26d1ac814b9e131e209eb39472a67ce ]

nvmeq->cq_head is compared with nvmeq->q_depth and changed the value and cq_phase for handling the next cq db.

but, nvmeq->q_depth's type is u32 and max. value is 0x10000 when CQP.MSQE is 0xffff and io_queue_depth is 0x10000.

current temp. variable for comparing with nvmeq->q_depth is overflowed when previous nvmeq->cq_head is 0xffff.

in this case, nvmeq->cq_phase is not updated.
so, fix data type for temp. variable to u32.

---
 drivers/nvme/host/pci.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/nvme/host/pci.c +1 -1
index c92a15c3fbc5..4555e9202851 100644
@@ -1027,7 +1027,7 @@ static inline void nvme_handle_cqe(struct nvme_queue *nvmeq, u16 idx)
 
 static inline void nvme_update_cq_head(struct nvme_queue *nvmeq)
 {
-	u16 tmp = nvmeq->cq_head + 1;
+	u32 tmp = nvmeq->cq_head + 1;
 
 	if (tmp == nvmeq->q_depth) {
 		nvmeq->cq_head = 0;
Signature · 1 line
2.30.2
author maintainer
#226 lore raw

PATCH 226/700 nvmet-fc: do not check for invalid target port in nvmet_fc_handle_fcp_rqst()

In reply to Greg Kroah-Hartman

From: Hannes Reinecke <[email protected]>

[ Upstream commit 2a4a910aa4f0acc428dc8d10227c42e14ed21d10 ]

When parsing a request in nvmet_fc_handle_fcp_rqst() we should not check for invalid target ports; if we do the command is aborted from the fcp layer, causing the host to assume a transport error. Rather we should still forward this request to the nvmet layer, which will then correctly fail the command with an appropriate error status.

---
 drivers/nvme/target/fc.c | 10 ++--------
 1 file changed, 2 insertions(+), 8 deletions(-)
drivers/nvme/target/fc.c +2 -8
index d375745fc4ed..b81db5270018 100644
@@ -2494,13 +2494,6 @@ nvmet_fc_handle_fcp_rqst(struct nvmet_fc_tgtport *tgtport,
 	u32 xfrlen = be32_to_cpu(cmdiu->data_len);
 	int ret;
 
-	/*
-	 * if there is no nvmet mapping to the targetport there
-	 * shouldn't be requests. just terminate them.
-	 */
-	if (!tgtport->pe)
-		goto transport_error;
-
 	/*
 	 * Fused commands are currently not supported in the linux
 	 * implementation.
@@ -2528,7 +2521,8 @@ nvmet_fc_handle_fcp_rqst(struct nvmet_fc_tgtport *tgtport,
 
 	fod->req.cmd = &fod->cmdiubuf.sqe;
 	fod->req.cqe = &fod->rspiubuf.cqe;
-	fod->req.port = tgtport->pe->port;
+	if (tgtport->pe)
+		fod->req.port = tgtport->pe->port;
 
 	/* clear any response payload */
 	memset(&fod->rspiubuf, 0, sizeof(fod->rspiubuf));
Signature · 1 line
2.30.2
author maintainer
#227 lore raw

PATCH 227/700 EDAC/Intel: Do not load EDAC driver when running as a guest

In reply to Greg Kroah-Hartman

From: Luck, Tony <[email protected]>

[ Upstream commit f0a029fff4a50eb01648810a77ba1873e829fdd4 ]

There's little to no point in loading an EDAC driver running in a guest:
1) The CPU model reported by CPUID may not represent actual h/w
2) The hypervisor likely does not pass in access to memory controller devices
3) Hypervisors generally do not pass corrected error details to guests

Add a check in each of the Intel EDAC drivers for X86_FEATURE_HYPERVISOR and simply return -ENODEV in the init routine.

---
 drivers/edac/i10nm_base.c | 3 +++
 drivers/edac/pnd2_edac.c  | 3 +++
 drivers/edac/sb_edac.c    | 3 +++
 drivers/edac/skx_base.c   | 3 +++
 4 files changed, 12 insertions(+)
drivers/edac/i10nm_base.c +3 -0
index 238a4ad1e526..37b4e875420e 100644
@@ -278,6 +278,9 @@ static int __init i10nm_init(void)
 	if (owner && strncmp(owner, EDAC_MOD_STR, sizeof(EDAC_MOD_STR)))
 		return -EBUSY;
 
+	if (cpu_feature_enabled(X86_FEATURE_HYPERVISOR))
+		return -ENODEV;
+
 	id = x86_match_cpu(i10nm_cpuids);
 	if (!id)
 		return -ENODEV;
drivers/edac/pnd2_edac.c +3 -0
index 928f63a374c7..c94ca1f790c4 100644
@@ -1554,6 +1554,9 @@ static int __init pnd2_init(void)
 	if (owner && strncmp(owner, EDAC_MOD_STR, sizeof(EDAC_MOD_STR)))
 		return -EBUSY;
 
+	if (cpu_feature_enabled(X86_FEATURE_HYPERVISOR))
+		return -ENODEV;
+
 	id = x86_match_cpu(pnd2_cpuids);
 	if (!id)
 		return -ENODEV;
drivers/edac/sb_edac.c +3 -0
index 93daa4297f2e..4c626fcd4dcb 100644
@@ -3510,6 +3510,9 @@ static int __init sbridge_init(void)
 	if (owner && strncmp(owner, EDAC_MOD_STR, sizeof(EDAC_MOD_STR)))
 		return -EBUSY;
 
+	if (cpu_feature_enabled(X86_FEATURE_HYPERVISOR))
+		return -ENODEV;
+
 	id = x86_match_cpu(sbridge_cpuids);
 	if (!id)
 		return -ENODEV;
drivers/edac/skx_base.c +3 -0
index 6a4f0b27c654..4dbd46575bfb 100644
@@ -656,6 +656,9 @@ static int __init skx_init(void)
 	if (owner && strncmp(owner, EDAC_MOD_STR, sizeof(EDAC_MOD_STR)))
 		return -EBUSY;
 
+	if (cpu_feature_enabled(X86_FEATURE_HYPERVISOR))
+		return -ENODEV;
+
 	id = x86_match_cpu(skx_cpuids);
 	if (!id)
 		return -ENODEV;
Signature · 1 line
2.30.2
author maintainer
#228 lore raw

PATCH 228/700 tools/power/x86/intel-speed-select: Fix uncore memory frequency display

In reply to Greg Kroah-Hartman

From: Srinivas Pandruvada <[email protected]>

[ Upstream commit 159f130f60f402273b235801d1fde3fc115c6795 ]

The uncore memory frequency value from the mailbox command CONFIG_TDP_GET_MEM_FREQ needs to be scaled based on the platform for display. There is no single constant multiplier.

This change introduces CPU model specific memory frequency multiplier.

---
 tools/power/x86/intel-speed-select/isst-config.c | 16 ++++++++++++++++
 tools/power/x86/intel-speed-select/isst-core.c   | 15 +++++++++++++++
 .../power/x86/intel-speed-select/isst-display.c  |  2 +-
 tools/power/x86/intel-speed-select/isst.h        |  2 ++
 4 files changed, 34 insertions(+), 1 deletion(-)
tools/power/x86/intel-speed-select/isst-config.c +16 -0
index 582feb88eca3..3ff8d64369d7 100644
@@ -106,6 +106,22 @@ int is_skx_based_platform(void)
 	return 0;
 }
 
+int is_spr_platform(void)
+{
+	if (cpu_model == 0x8F)
+		return 1;
+
+	return 0;
+}
+
+int is_icx_platform(void)
+{
+	if (cpu_model == 0x6A || cpu_model == 0x6C)
+		return 1;
+
+	return 0;
+}
+
 static int update_cpu_model(void)
 {
 	unsigned int ebx, ecx, edx;
tools/power/x86/intel-speed-select/isst-core.c +15 -0
index 6a26d5769984..4431c8a0d40a 100644
@@ -201,6 +201,7 @@ void isst_get_uncore_mem_freq(int cpu, int config_index,
 {
 	unsigned int resp;
 	int ret;
+
 	ret = isst_send_mbox_command(cpu, CONFIG_TDP, CONFIG_TDP_GET_MEM_FREQ,
 				     0, config_index, &resp);
 	if (ret) {
@@ -209,6 +210,20 @@ void isst_get_uncore_mem_freq(int cpu, int config_index,
 	}
 
 	ctdp_level->mem_freq = resp & GENMASK(7, 0);
+	if (is_spr_platform()) {
+		ctdp_level->mem_freq *= 200;
+	} else if (is_icx_platform()) {
+		if (ctdp_level->mem_freq < 7) {
+			ctdp_level->mem_freq = (12 - ctdp_level->mem_freq) * 133.33 * 2 * 10;
+			ctdp_level->mem_freq /= 10;
+			if (ctdp_level->mem_freq % 10 > 5)
+				ctdp_level->mem_freq++;
+		} else {
+			ctdp_level->mem_freq = 0;
+		}
+	} else {
+		ctdp_level->mem_freq = 0;
+	}
 	debug_printf(
 		"cpu:%d ctdp:%d CONFIG_TDP_GET_MEM_FREQ resp:%x uncore mem_freq:%d\n",
 		cpu, config_index, resp, ctdp_level->mem_freq);
tools/power/x86/intel-speed-select/isst-display.c +1 -1
index 3bf1820c0da1..f97d8859ada7 100644
@@ -446,7 +446,7 @@ void isst_ctdp_display_information(int cpu, FILE *outf, int tdp_level,
 		if (ctdp_level->mem_freq) {
 			snprintf(header, sizeof(header), "mem-frequency(MHz)");
 			snprintf(value, sizeof(value), "%d",
-				 ctdp_level->mem_freq * DISP_FREQ_MULTIPLIER);
+				 ctdp_level->mem_freq);
 			format_and_print(outf, level + 2, header, value);
 		}
 
tools/power/x86/intel-speed-select/isst.h +2 -0
index 0cac6c54be87..1aa15d5ea57c 100644
@@ -257,5 +257,7 @@ extern int get_cpufreq_base_freq(int cpu);
 extern int isst_read_pm_config(int cpu, int *cp_state, int *cp_cap);
 extern void isst_display_error_info_message(int error, char *msg, int arg_valid, int arg);
 extern int is_skx_based_platform(void);
+extern int is_spr_platform(void);
+extern int is_icx_platform(void);
 extern void isst_trl_display_information(int cpu, FILE *outf, unsigned long long trl);
 #endif
Signature · 1 line
2.30.2
author maintainer
#229 lore raw

PATCH 229/700 PCI: hv: Add check for hyperv_initialized in init_hv_pci_drv()

In reply to Greg Kroah-Hartman

From: Haiyang Zhang <[email protected]>

[ Upstream commit 7d815f4afa87f2032b650ae1bba7534b550a6b8b ]

Add check for hv_is_hyperv_initialized() at the top of init_hv_pci_drv(), so if the pci-hyperv driver is force-loaded on non Hyper-V platforms, the init_hv_pci_drv() will exit immediately, without any side effects, like assignments to hvpci_block_ops, etc.

---
 drivers/pci/controller/pci-hyperv.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/pci/controller/pci-hyperv.c +3 -0
index 27a17a1e4a7c..7479edf3676c 100644
@@ -3480,6 +3480,9 @@ static void __exit exit_hv_pci_drv(void)
 
 static int __init init_hv_pci_drv(void)
 {
+	if (!hv_is_hyperv_initialized())
+		return -ENODEV;
+
 	/* Set the invalid domain number's bit, so it will not be used */
 	set_bit(HVPCI_DOM_INVALID, hvpci_dom_map);
 
Signature · 1 line
2.30.2
author maintainer
#230 lore raw

PATCH 230/700 cifs: improve fallocate emulation

In reply to Greg Kroah-Hartman

From: Ronnie Sahlberg <[email protected]>

[ Upstream commit 966a3cb7c7db786452a87afdc3b48858fc4d4d6b ]

RHBZ: 1866684

We don't have a real fallocate in the SMB2 protocol so we used to emulate fallocate by simply switching the file to become non-sparse. But as that could potantially consume a lot more data than we intended to fallocate (large sparse file and fallocating a thin slice in the middle) we would only do this IFF the fallocate request was for virtually the entire file.

This patch improves this and starts allowing us to fallocate smaller chunks of a file by overwriting the region with 0, for the parts that are unallocated.

The method used is to first query the server for FSCTL_QUERY_ALLOCATED_RANGES to find what is unallocated in the fallocate range and then to only overwrite-with-zero the unallocated ranges to fill in the holes.

As overwriting-with-zero is different from just allocating blocks, and potentially much more expensive, we limit this to only allow fallocate ranges up to 1Mb in size.

---
 fs/cifs/smb2ops.c | 133 ++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 133 insertions(+)
fs/cifs/smb2ops.c +133 -0
index e9a530da4255..ea5e958fd6b0 100644
@@ -3561,6 +3561,119 @@ static long smb3_punch_hole(struct file *file, struct cifs_tcon *tcon,
 	return rc;
 }
 
+static int smb3_simple_fallocate_write_range(unsigned int xid,
+					     struct cifs_tcon *tcon,
+					     struct cifsFileInfo *cfile,
+					     loff_t off, loff_t len,
+					     char *buf)
+{
+	struct cifs_io_parms io_parms = {0};
+	int nbytes;
+	struct kvec iov[2];
+
+	io_parms.netfid = cfile->fid.netfid;
+	io_parms.pid = current->tgid;
+	io_parms.tcon = tcon;
+	io_parms.persistent_fid = cfile->fid.persistent_fid;
+	io_parms.volatile_fid = cfile->fid.volatile_fid;
+	io_parms.offset = off;
+	io_parms.length = len;
+
+	/* iov[0] is reserved for smb header */
+	iov[1].iov_base = buf;
+	iov[1].iov_len = io_parms.length;
+	return SMB2_write(xid, &io_parms, &nbytes, iov, 1);
+}
+
+static int smb3_simple_fallocate_range(unsigned int xid,
+				       struct cifs_tcon *tcon,
+				       struct cifsFileInfo *cfile,
+				       loff_t off, loff_t len)
+{
+	struct file_allocated_range_buffer in_data, *out_data = NULL, *tmp_data;
+	u32 out_data_len;
+	char *buf = NULL;
+	loff_t l;
+	int rc;
+
+	in_data.file_offset = cpu_to_le64(off);
+	in_data.length = cpu_to_le64(len);
+	rc = SMB2_ioctl(xid, tcon, cfile->fid.persistent_fid,
+			cfile->fid.volatile_fid,
+			FSCTL_QUERY_ALLOCATED_RANGES, true,
+			(char *)&in_data, sizeof(in_data),
+			1024 * sizeof(struct file_allocated_range_buffer),
+			(char **)&out_data, &out_data_len);
+	if (rc)
+		goto out;
+	/*
+	 * It is already all allocated
+	 */
+	if (out_data_len == 0)
+		goto out;
+
+	buf = kzalloc(1024 * 1024, GFP_KERNEL);
+	if (buf == NULL) {
+		rc = -ENOMEM;
+		goto out;
+	}
+
+	tmp_data = out_data;
+	while (len) {
+		/*
+		 * The rest of the region is unmapped so write it all.
+		 */
+		if (out_data_len == 0) {
+			rc = smb3_simple_fallocate_write_range(xid, tcon,
+					       cfile, off, len, buf);
+			goto out;
+		}
+
+		if (out_data_len < sizeof(struct file_allocated_range_buffer)) {
+			rc = -EINVAL;
+			goto out;
+		}
+
+		if (off < le64_to_cpu(tmp_data->file_offset)) {
+			/*
+			 * We are at a hole. Write until the end of the region
+			 * or until the next allocated data,
+			 * whichever comes next.
+			 */
+			l = le64_to_cpu(tmp_data->file_offset) - off;
+			if (len < l)
+				l = len;
+			rc = smb3_simple_fallocate_write_range(xid, tcon,
+					       cfile, off, l, buf);
+			if (rc)
+				goto out;
+			off = off + l;
+			len = len - l;
+			if (len == 0)
+				goto out;
+		}
+		/*
+		 * We are at a section of allocated data, just skip forward
+		 * until the end of the data or the end of the region
+		 * we are supposed to fallocate, whichever comes first.
+		 */
+		l = le64_to_cpu(tmp_data->length);
+		if (len < l)
+			l = len;
+		off += l;
+		len -= l;
+
+		tmp_data = &tmp_data[1];
+		out_data_len -= sizeof(struct file_allocated_range_buffer);
+	}
+
+ out:
+	kfree(out_data);
+	kfree(buf);
+	return rc;
+}
+
+
 static long smb3_simple_falloc(struct file *file, struct cifs_tcon *tcon,
 			    loff_t off, loff_t len, bool keep_size)
 {
@@ -3621,6 +3734,26 @@ static long smb3_simple_falloc(struct file *file, struct cifs_tcon *tcon,
 	}
 
 	if ((keep_size == true) || (i_size_read(inode) >= off + len)) {
+		/*
+		 * At this point, we are trying to fallocate an internal
+		 * regions of a sparse file. Since smb2 does not have a
+		 * fallocate command we have two otions on how to emulate this.
+		 * We can either turn the entire file to become non-sparse
+		 * which we only do if the fallocate is for virtually
+		 * the whole file,  or we can overwrite the region with zeroes
+		 * using SMB2_write, which could be prohibitevly expensive
+		 * if len is large.
+		 */
+		/*
+		 * We are only trying to fallocate a small region so
+		 * just write it with zero.
+		 */
+		if (len <= 1024 * 1024) {
+			rc = smb3_simple_fallocate_range(xid, tcon, cfile,
+							 off, len);
+			goto out;
+		}
+
 		/*
 		 * Check if falloc starts within first few pages of file
 		 * and ends within a few pages of the end of file to
Signature · 1 line
2.30.2
author maintainer
#231 lore raw

PATCH 231/700 cifs: fix check of dfs interlinks

In reply to Greg Kroah-Hartman

From: Paulo Alcantara <[email protected]>

[ Upstream commit 889c2a700799f3b6f82210925e1faf4a9b833c4a ]

Interlink is a special type of DFS link that resolves to a different DFS domain-based namespace. To determine whether it is an interlink or not, check if ReferralServers and StorageServers bits are set to 1 and 0 respectively in ReferralHeaderFlags, as specified in MS-DFSC 3.1.5.4.5 Determining Whether a Referral Response is an Interlink.

---
 fs/cifs/dfs_cache.c | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)
fs/cifs/dfs_cache.c +3 -4
index 098b4bc8da59..d2d686ee10a3 100644
@@ -25,8 +25,7 @@
 #define CACHE_HTABLE_SIZE 32
 #define CACHE_MAX_ENTRIES 64
 
-#define IS_INTERLINK_SET(v) ((v) & (DFSREF_REFERRAL_SERVER | \
-				    DFSREF_STORAGE_SERVER))
+#define IS_DFS_INTERLINK(v) (((v) & DFSREF_REFERRAL_SERVER) && !((v) & DFSREF_STORAGE_SERVER))
 
 struct cache_dfs_tgt {
 	char *name;
@@ -170,7 +169,7 @@ static int dfscache_proc_show(struct seq_file *m, void *v)
 				   "cache entry: path=%s,type=%s,ttl=%d,etime=%ld,hdr_flags=0x%x,ref_flags=0x%x,interlink=%s,path_consumed=%d,expired=%s\n",
 				   ce->path, ce->srvtype == DFS_TYPE_ROOT ? "root" : "link",
 				   ce->ttl, ce->etime.tv_nsec, ce->ref_flags, ce->hdr_flags,
-				   IS_INTERLINK_SET(ce->hdr_flags) ? "yes" : "no",
+				   IS_DFS_INTERLINK(ce->hdr_flags) ? "yes" : "no",
 				   ce->path_consumed, cache_entry_expired(ce) ? "yes" : "no");
 
 			list_for_each_entry(t, &ce->tlist, list) {
@@ -239,7 +238,7 @@ static inline void dump_ce(const struct cache_entry *ce)
 		 ce->srvtype == DFS_TYPE_ROOT ? "root" : "link", ce->ttl,
 		 ce->etime.tv_nsec,
 		 ce->hdr_flags, ce->ref_flags,
-		 IS_INTERLINK_SET(ce->hdr_flags) ? "yes" : "no",
+		 IS_DFS_INTERLINK(ce->hdr_flags) ? "yes" : "no",
 		 ce->path_consumed,
 		 cache_entry_expired(ce) ? "yes" : "no");
 	dump_tgts(ce);
Signature · 1 line
2.30.2
author maintainer
#232 lore raw

PATCH 232/700 smb3: fix uninitialized value for port in witness protocol move

In reply to Greg Kroah-Hartman

From: Steve French <[email protected]>

[ Upstream commit ff93b71a3eff25fe9d4364ef13b6e01d935600c6 ]

Although in practice this can not occur (since IPv4 and IPv6 are the only two cases currently supported), it is cleaner to avoid uninitialized variable warnings.

Addresses smatch warning:
fs/cifs/cifs_swn.c:468 cifs_swn_store_swn_addr() error: uninitialized symbol 'port'.

---
 fs/cifs/cifs_swn.c | 10 +++-------
 1 file changed, 3 insertions(+), 7 deletions(-)
fs/cifs/cifs_swn.c +3 -7
index d829b8bf833e..93b47818c6c2 100644
@@ -447,15 +447,13 @@ static int cifs_swn_store_swn_addr(const struct sockaddr_storage *new,
 				   const struct sockaddr_storage *old,
 				   struct sockaddr_storage *dst)
 {
-	__be16 port;
+	__be16 port = cpu_to_be16(CIFS_PORT);
 
 	if (old->ss_family == AF_INET) {
 		struct sockaddr_in *ipv4 = (struct sockaddr_in *)old;
 
 		port = ipv4->sin_port;
-	}
-
-	if (old->ss_family == AF_INET6) {
+	} else if (old->ss_family == AF_INET6) {
 		struct sockaddr_in6 *ipv6 = (struct sockaddr_in6 *)old;
 
 		port = ipv6->sin6_port;
@@ -465,9 +463,7 @@ static int cifs_swn_store_swn_addr(const struct sockaddr_storage *new,
 		struct sockaddr_in *ipv4 = (struct sockaddr_in *)new;
 
 		ipv4->sin_port = port;
-	}
-
-	if (new->ss_family == AF_INET6) {
+	} else if (new->ss_family == AF_INET6) {
 		struct sockaddr_in6 *ipv6 = (struct sockaddr_in6 *)new;
 
 		ipv6->sin6_port = port;
Signature · 1 line
2.30.2
author maintainer
#233 lore raw

PATCH 233/700 ACPI: EC: trust DSDT GPE for certain HP laptop

In reply to Greg Kroah-Hartman

From: Zhang Rui <[email protected]>

[ Upstream commit 4370cbf350dbaca984dbda9f9ce3fac45d6949d5 ]

On HP Pavilion Gaming Laptop 15-cx0xxx, the ECDT EC and DSDT EC share the same port addresses but different GPEs. And the DSDT GPE is the right one to use.

The current code duplicates DSDT EC with ECDT EC if the port addresses are the same, and uses ECDT GPE as a result, which breaks this machine.

Introduce a new quirk for the HP laptop to trust the DSDT GPE, and avoid duplicating even if the port addresses are the same.

---
 drivers/acpi/ec.c | 21 ++++++++++++++++++++-
 1 file changed, 20 insertions(+), 1 deletion(-)
drivers/acpi/ec.c +20 -1
index e8c5da2b964a..87c3b4a099b9 100644
@@ -183,6 +183,7 @@ static struct workqueue_struct *ec_query_wq;
 
 static int EC_FLAGS_CORRECT_ECDT; /* Needs ECDT port address correction */
 static int EC_FLAGS_IGNORE_DSDT_GPE; /* Needs ECDT GPE as correction setting */
+static int EC_FLAGS_TRUST_DSDT_GPE; /* Needs DSDT GPE as correction setting */
 static int EC_FLAGS_CLEAR_ON_RESUME; /* Needs acpi_ec_clear() on boot/resume */
 
 /* --------------------------------------------------------------------------
@@ -1593,7 +1594,8 @@ static int acpi_ec_add(struct acpi_device *device)
 		}
 
 		if (boot_ec && ec->command_addr == boot_ec->command_addr &&
-		    ec->data_addr == boot_ec->data_addr) {
+		    ec->data_addr == boot_ec->data_addr &&
+		    !EC_FLAGS_TRUST_DSDT_GPE) {
 			/*
 			 * Trust PNP0C09 namespace location rather than
 			 * ECDT ID. But trust ECDT GPE rather than _GPE
@@ -1816,6 +1818,18 @@ static int ec_correct_ecdt(const struct dmi_system_id *id)
 	return 0;
 }
 
+/*
+ * Some ECDTs contain wrong GPE setting, but they share the same port addresses
+ * with DSDT EC, don't duplicate the DSDT EC with ECDT EC in this case.
+ * https://bugzilla.kernel.org/show_bug.cgi?id=209989
+ */
+static int ec_honor_dsdt_gpe(const struct dmi_system_id *id)
+{
+	pr_debug("Detected system needing DSDT GPE setting.\n");
+	EC_FLAGS_TRUST_DSDT_GPE = 1;
+	return 0;
+}
+
 /*
  * Some DSDTs contain wrong GPE setting.
  * Asus FX502VD/VE, GL702VMK, X550VXK, X580VD
@@ -1870,6 +1884,11 @@ static const struct dmi_system_id ec_dmi_table[] __initconst = {
 	DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
 	DMI_MATCH(DMI_PRODUCT_NAME, "X580VD"),}, NULL},
 	{
+	/* https://bugzilla.kernel.org/show_bug.cgi?id=209989 */
+	ec_honor_dsdt_gpe, "HP Pavilion Gaming Laptop 15-cx0xxx", {
+	DMI_MATCH(DMI_SYS_VENDOR, "HP"),
+	DMI_MATCH(DMI_PRODUCT_NAME, "HP Pavilion Gaming Laptop 15-cx0xxx"),}, NULL},
+	{
 	ec_clear_on_resume, "Samsung hardware", {
 	DMI_MATCH(DMI_SYS_VENDOR, "SAMSUNG ELECTRONICS CO., LTD.")}, NULL},
 	{},
Signature · 1 line
2.30.2
author maintainer
#234 lore raw

PATCH 234/700 clocksource: Retry clock read if long delays detected

In reply to Greg Kroah-Hartman

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

[ Upstream commit db3a34e17433de2390eb80d436970edcebd0ca3e ]

When the clocksource watchdog marks a clock as unstable, this might be due to that clock being unstable or it might be due to delays that happen to occur between the reads of the two clocks. Yes, interrupts are disabled across those two reads, but there are no shortage of things that can delay interrupts-disabled regions of code ranging from SMI handlers to vCPU preemption. It would be good to have some indication as to why the clock was marked unstable.

Therefore, re-read the watchdog clock on either side of the read from the clock under test. If the watchdog clock shows an excessive time delta between its pair of reads, the reads are retried.

The maximum number of retries is specified by a new kernel boot parameter clocksource.max_cswd_read_retries, which defaults to three, that is, up to four reads, one initial and up to three retries. If more than one retry was required, a message is printed on the console (the occasional single retry is expected behavior, especially in guest OSes). If the maximum number of retries is exceeded, the clock under test will be marked unstable. However, the probability of this happening due to various sorts of delays is quite small. In addition, the reason (clock-read delays) for the unstable marking will be apparent.

---
 .../admin-guide/kernel-parameters.txt         |  6 +++
 kernel/time/clocksource.c                     | 53 ++++++++++++++++---
 2 files changed, 53 insertions(+), 6 deletions(-)
Documentation/admin-guide/kernel-parameters.txt +6 -0
index 835f810f2f26..c08e174e6ff4 100644
@@ -583,6 +583,12 @@
 			loops can be debugged more effectively on production
 			systems.
 
+	clocksource.max_cswd_read_retries= [KNL]
+			Number of clocksource_watchdog() retries due to
+			external delays before the clock will be marked
+			unstable.  Defaults to three retries, that is,
+			four attempts to read the clock under test.
+
 	clearcpuid=BITNUM[,BITNUM...] [X86]
 			Disable CPUID feature X for the kernel. See
 			arch/x86/include/asm/cpufeatures.h for the valid bit
kernel/time/clocksource.c +47 -6

75 lines not on this page · raw message

Signature · 1 line
2.30.2
author maintainer
#235 lore raw

PATCH 235/700 clocksource: Check per-CPU clock synchronization when marked unstable

In reply to Greg Kroah-Hartman

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

[ Upstream commit 7560c02bdffb7c52d1457fa551b9e745d4b9e754 ]

Some sorts of per-CPU clock sources have a history of going out of synchronization with each other. However, this problem has purportedy been solved in the past ten years. Except that it is all too possible that the problem has instead simply been made less likely, which might mean that some of the occasional "Marking clocksource 'tsc' as unstable" messages might be due to desynchronization. How would anyone know?

Therefore apply CPU-to-CPU synchronization checking to newly unstable clocksource that are marked with the new CLOCK_SOURCE_VERIFY_PERCPU flag. Lists of desynchronized CPUs are printed, with the caveat that if it is the reporting CPU that is itself desynchronized, it will appear that all the other clocks are wrong. Just like in real life.

---
 arch/x86/kernel/tsc.c       |  3 +-
 include/linux/clocksource.h |  2 +-
 kernel/time/clocksource.c   | 60 +++++++++++++++++++++++++++++++++++++
 3 files changed, 63 insertions(+), 2 deletions(-)
arch/x86/kernel/tsc.c +2 -1
index f70dffc2771f..56289170753c 100644
@@ -1151,7 +1151,8 @@ static struct clocksource clocksource_tsc = {
 	.mask			= CLOCKSOURCE_MASK(64),
 	.flags			= CLOCK_SOURCE_IS_CONTINUOUS |
 				  CLOCK_SOURCE_VALID_FOR_HRES |
-				  CLOCK_SOURCE_MUST_VERIFY,
+				  CLOCK_SOURCE_MUST_VERIFY |
+				  CLOCK_SOURCE_VERIFY_PERCPU,
 	.vdso_clock_mode	= VDSO_CLOCKMODE_TSC,
 	.enable			= tsc_cs_enable,
 	.resume			= tsc_resume,
include/linux/clocksource.h +1 -1
index 86d143db6523..83a3ebff7456 100644
@@ -131,7 +131,7 @@ struct clocksource {
 #define CLOCK_SOURCE_UNSTABLE			0x40
 #define CLOCK_SOURCE_SUSPEND_NONSTOP		0x80
 #define CLOCK_SOURCE_RESELECT			0x100
-
+#define CLOCK_SOURCE_VERIFY_PERCPU		0x200
 /* simplify initialization of mask field */
 #define CLOCKSOURCE_MASK(bits) GENMASK_ULL((bits) - 1, 0)
 
kernel/time/clocksource.c +60 -0

72 lines not on this page · raw message

Signature · 1 line
2.30.2
author maintainer
#236 lore raw

PATCH 236/700 tpm_tis_spi: add missing SPI device ID entries

In reply to Greg Kroah-Hartman

From: Javier Martinez Canillas <[email protected]>

[ Upstream commit c46ed2281bbe4b84e6f3d4bdfb0e4e9ab813fa9d ]

The SPI core always reports a "MODALIAS=spi:<foo>", even if the device was registered via OF. This means that this module won't auto-load if a DT has for example has a node with a compatible "infineon,slb9670" string.

In that case kmod will expect a "MODALIAS=of:N*T*Cinfineon,slb9670" uevent but instead will get a "MODALIAS=spi:slb9670", which is not present in the kernel module aliases:

$ modinfo drivers/char/tpm/tpm_tis_spi.ko | grep alias
alias:          of:N*T*Cgoogle,cr50C*
alias:          of:N*T*Cgoogle,cr50
alias:          of:N*T*Ctcg,tpm_tis-spiC*
alias:          of:N*T*Ctcg,tpm_tis-spi
alias:          of:N*T*Cinfineon,slb9670C*
alias:          of:N*T*Cinfineon,slb9670
alias:          of:N*T*Cst,st33htpm-spiC*
alias:          of:N*T*Cst,st33htpm-spi
alias:          spi:cr50
alias:          spi:tpm_tis_spi
alias:          acpi*:SMO0768:*

To workaround this issue, add in the SPI device ID table all the entries that are present in the OF device ID table.

---
 drivers/char/tpm/tpm_tis_spi_main.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/char/tpm/tpm_tis_spi_main.c +2 -0
index 3856f6ebcb34..de4209003a44 100644
@@ -260,6 +260,8 @@ static int tpm_tis_spi_remove(struct spi_device *dev)
 }
 
 static const struct spi_device_id tpm_tis_spi_id[] = {
+	{ "st33htpm-spi", (unsigned long)tpm_tis_spi_probe },
+	{ "slb9670", (unsigned long)tpm_tis_spi_probe },
 	{ "tpm_tis_spi", (unsigned long)tpm_tis_spi_probe },
 	{ "cr50", (unsigned long)cr50_spi_probe },
 	{}
Signature · 1 line
2.30.2
author maintainer
#237 lore raw

PATCH 237/700 ACPI: tables: Add custom DSDT file as makefile prerequisite

In reply to Greg Kroah-Hartman

From: Richard Fitzgerald <[email protected]>

[ Upstream commit d1059c1b1146870c52f3dac12cb7b6cbf39ed27f ]

A custom DSDT file is mostly used during development or debugging, and in that case it is quite likely to want to rebuild the kernel after changing ONLY the content of the DSDT.

This patch adds the custom DSDT as a prerequisite to tables.o to ensure a rebuild if the DSDT file is updated. Make will merge the prerequisites from multiple rules for the same target.

---
 drivers/acpi/Makefile | 5 +++++
 1 file changed, 5 insertions(+)
drivers/acpi/Makefile +5 -0
index 700b41adf2db..9aa82d527272 100644
@@ -8,6 +8,11 @@ ccflags-$(CONFIG_ACPI_DEBUG)	+= -DACPI_DEBUG_OUTPUT
 #
 # ACPI Boot-Time Table Parsing
 #
+ifeq ($(CONFIG_ACPI_CUSTOM_DSDT),y)
+tables.o: $(src)/../../include/$(subst $\",,$(CONFIG_ACPI_CUSTOM_DSDT_FILE)) ;
+
+endif
+
 obj-$(CONFIG_ACPI)		+= tables.o
 obj-$(CONFIG_X86)		+= blacklist.o
 
Signature · 1 line
2.30.2
author maintainer
#238 lore raw

PATCH 238/700 smb3: fix possible access to uninitialized pointer to DACL

In reply to Greg Kroah-Hartman

From: Steve French <[email protected]>

[ Upstream commit a5628263a9f8d47d9a1548fe9d5d75ba4423a735 ]

dacl_ptr can be null so we must check for it everywhere it is used in build_sec_desc.

Addresses-Coverity: 1475598 ("Explicit null dereference")

---
 fs/cifs/cifsacl.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
fs/cifs/cifsacl.c +1 -1
index d178cf85e926..b80b6ba232aa 100644
@@ -1310,7 +1310,7 @@ static int build_sec_desc(struct cifs_ntsd *pntsd, struct cifs_ntsd *pnntsd,
 		ndacl_ptr = (struct cifs_acl *)((char *)pnntsd + ndacloffset);
 		ndacl_ptr->revision =
 			dacloffset ? dacl_ptr->revision : cpu_to_le16(ACL_REVISION);
-		ndacl_ptr->num_aces = dacl_ptr->num_aces;
+		ndacl_ptr->num_aces = dacl_ptr ? dacl_ptr->num_aces : 0;
 
 		if (uid_valid(uid)) { /* chown */
 			uid_t id;
Signature · 1 line
2.30.2
author maintainer
#239 lore raw

PATCH 239/700 HID: wacom: Correct base usage for capacitive ExpressKey status bits

In reply to Greg Kroah-Hartman

From: Jason Gerecke <[email protected]>

[ Upstream commit 424d8237945c6c448c8b3f23885d464fb5685c97 ]

The capacitive status of ExpressKeys is reported with usages beginning at 0x940, not 0x950. Bring our driver into alignment with reality.

---
 drivers/hid/wacom_wac.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/hid/wacom_wac.h +1 -1
index 195910dd2154..e3835407e8d2 100644
@@ -122,7 +122,7 @@
 #define WACOM_HID_WD_TOUCHONOFF         (WACOM_HID_UP_WACOMDIGITIZER | 0x0454)
 #define WACOM_HID_WD_BATTERY_LEVEL      (WACOM_HID_UP_WACOMDIGITIZER | 0x043b)
 #define WACOM_HID_WD_EXPRESSKEY00       (WACOM_HID_UP_WACOMDIGITIZER | 0x0910)
-#define WACOM_HID_WD_EXPRESSKEYCAP00    (WACOM_HID_UP_WACOMDIGITIZER | 0x0950)
+#define WACOM_HID_WD_EXPRESSKEYCAP00    (WACOM_HID_UP_WACOMDIGITIZER | 0x0940)
 #define WACOM_HID_WD_MODE_CHANGE        (WACOM_HID_UP_WACOMDIGITIZER | 0x0980)
 #define WACOM_HID_WD_MUTE_DEVICE        (WACOM_HID_UP_WACOMDIGITIZER | 0x0981)
 #define WACOM_HID_WD_CONTROLPANEL       (WACOM_HID_UP_WACOMDIGITIZER | 0x0982)
Signature · 1 line
2.30.2
author maintainer
#240 lore raw

PATCH 240/700 cifs: fix missing spinlock around update to ses->status

In reply to Greg Kroah-Hartman

From: Steve French <[email protected]>

[ Upstream commit 0060a4f28a9ef45ae8163c0805e944a2b1546762 ]

In the other places where we update ses->status we protect the updates via GlobalMid_Lock. So to be consistent add the same locking around it in cifs_put_smb_ses where it was missing.

Addresses-Coverity: 1268904 ("Data race condition")

---
 fs/cifs/cifsglob.h | 3 ++-
 fs/cifs/connect.c  | 5 ++++-
 2 files changed, 6 insertions(+), 2 deletions(-)
fs/cifs/cifsglob.h +2 -1
index ec824ab8c5ca..723b97002d8d 100644
@@ -896,7 +896,7 @@ struct cifs_ses {
 	struct mutex session_mutex;
 	struct TCP_Server_Info *server;	/* pointer to server info */
 	int ses_count;		/* reference counter */
-	enum statusEnum status;
+	enum statusEnum status;  /* updates protected by GlobalMid_Lock */
 	unsigned overrideSecFlg;  /* if non-zero override global sec flags */
 	char *serverOS;		/* name of operating system underlying server */
 	char *serverNOS;	/* name of network operating system of server */
@@ -1783,6 +1783,7 @@ require use of the stronger protocol */
  *	list operations on pending_mid_q and oplockQ
  *      updates to XID counters, multiplex id  and SMB sequence numbers
  *      list operations on global DnotifyReqList
+ *      updates to ses->status
  *  tcp_ses_lock protects:
  *	list operations on tcp and SMB session lists
  *  tcon->open_file_lock protects the list of open files hanging off the tcon
fs/cifs/connect.c +4 -1
index 3d62d52d730b..09a3939f25bf 100644
@@ -1639,9 +1639,12 @@ void cifs_put_smb_ses(struct cifs_ses *ses)
 		spin_unlock(&cifs_tcp_ses_lock);
 		return;
 	}
+	spin_unlock(&cifs_tcp_ses_lock);
+
+	spin_lock(&GlobalMid_Lock);
 	if (ses->status == CifsGood)
 		ses->status = CifsExiting;
-	spin_unlock(&cifs_tcp_ses_lock);
+	spin_unlock(&GlobalMid_Lock);
 
 	cifs_free_ipc(ses);
 
Signature · 1 line
2.30.2
author maintainer
#241 lore raw

PATCH 241/700 bfq: Remove merged request already in bfq_requests_merged()

In reply to Greg Kroah-Hartman

From: Jan Kara <[email protected]>

[ Upstream commit a921c655f2033dd1ce1379128efe881dda23ea37 ]

Currently, bfq does very little in bfq_requests_merged() and handles all the request cleanup in bfq_finish_requeue_request() called from blk_mq_free_request(). That is currently safe only because blk_mq_free_request() is called shortly after bfq_requests_merged() while bfqd->lock is still held. However to fix a lock inversion between bfqd->lock and ioc->lock, we need to call blk_mq_free_request() after dropping bfqd->lock. That would mean that already merged request could be seen by other processes inside bfq queues and possibly dispatched to the device which is wrong. So move cleanup of the request from bfq_finish_requeue_request() to bfq_requests_merged().

---
 block/bfq-iosched.c | 41 +++++++++++++----------------------------
 1 file changed, 13 insertions(+), 28 deletions(-)
block/bfq-iosched.c +13 -28
index bc319931d2b3..9a180c4b9605 100644
@@ -2376,7 +2376,7 @@ static void bfq_requests_merged(struct request_queue *q, struct request *rq,
 		*next_bfqq = bfq_init_rq(next);
 
 	if (!bfqq)
-		return;
+		goto remove;
 
 	/*
 	 * If next and rq belong to the same bfq_queue and next is older
@@ -2399,6 +2399,14 @@ static void bfq_requests_merged(struct request_queue *q, struct request *rq,
 		bfqq->next_rq = rq;
 
 	bfqg_stats_update_io_merged(bfqq_group(bfqq), next->cmd_flags);
+remove:
+	/* Merged request may be in the IO scheduler. Remove it. */
+	if (!RB_EMPTY_NODE(&next->rb_node)) {
+		bfq_remove_request(next->q, next);
+		if (next_bfqq)
+			bfqg_stats_update_io_remove(bfqq_group(next_bfqq),
+						    next->cmd_flags);
+	}
 }
 
 /* Must be called with bfqq != NULL */
@@ -6015,6 +6023,7 @@ static void bfq_finish_requeue_request(struct request *rq)
 {
 	struct bfq_queue *bfqq = RQ_BFQQ(rq);
 	struct bfq_data *bfqd;
+	unsigned long flags;
 
 	/*
 	 * rq either is not associated with any icq, or is an already
@@ -6032,39 +6041,15 @@ static void bfq_finish_requeue_request(struct request *rq)
 					     rq->io_start_time_ns,
 					     rq->cmd_flags);
 
+	spin_lock_irqsave(&bfqd->lock, flags);
 	if (likely(rq->rq_flags & RQF_STARTED)) {
-		unsigned long flags;
-
-		spin_lock_irqsave(&bfqd->lock, flags);
-
 		if (rq == bfqd->waited_rq)
 			bfq_update_inject_limit(bfqd, bfqq);
 
 		bfq_completed_request(bfqq, bfqd);
-		bfq_finish_requeue_request_body(bfqq);
-
-		spin_unlock_irqrestore(&bfqd->lock, flags);
-	} else {
-		/*
-		 * Request rq may be still/already in the scheduler,
-		 * in which case we need to remove it (this should
-		 * never happen in case of requeue). And we cannot
-		 * defer such a check and removal, to avoid
-		 * inconsistencies in the time interval from the end
-		 * of this function to the start of the deferred work.
-		 * This situation seems to occur only in process
-		 * context, as a consequence of a merge. In the
-		 * current version of the code, this implies that the
-		 * lock is held.
-		 */
-
-		if (!RB_EMPTY_NODE(&rq->rb_node)) {
-			bfq_remove_request(rq->q, rq);
-			bfqg_stats_update_io_remove(bfqq_group(bfqq),
-						    rq->cmd_flags);
-		}
-		bfq_finish_requeue_request_body(bfqq);
 	}
+	bfq_finish_requeue_request_body(bfqq);
+	spin_unlock_irqrestore(&bfqd->lock, flags);
 
 	/*
 	 * Reset private fields. In case of a requeue, this allows
Signature · 1 line
2.30.2
author maintainer
#242 lore raw

PATCH 242/700 mailbox: qcom: Use PLATFORM_DEVID_AUTO to register platform device

In reply to Greg Kroah-Hartman

From: Shawn Guo <[email protected]>

[ Upstream commit 96e39e95c01283ff5695dafe659df88ada802159 ]

In adding APCS clock support for MSM8939, the second clock registration fails due to duplicate device name like below.

[    0.519657] sysfs: cannot create duplicate filename '/bus/platform/devices/qcom-apcs-msm8916-clk'
...
[    0.661158] qcom_apcs_ipc b111000.mailbox: failed to register APCS clk

This is because MSM8939 has 3 APCS instances for Cluster0 (little cores), Cluster1 (big cores) and CCI (Cache Coherent Interconnect). Although only APCS of Cluster0 and Cluster1 have IPC bits, each of 3 APCS has A53PLL clock control bits. That said, 3 'qcom-apcs-msm8916-clk' devices need to be registered to instantiate all 3 clocks. Use PLATFORM_DEVID_AUTO rather than PLATFORM_DEVID_NONE for platform_device_register_data() call to fix the issue above.

---
 drivers/mailbox/qcom-apcs-ipc-mailbox.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/mailbox/qcom-apcs-ipc-mailbox.c +1 -1
index f25324d03842..15236d729625 100644
@@ -132,7 +132,7 @@ static int qcom_apcs_ipc_probe(struct platform_device *pdev)
 	if (apcs_data->clk_name) {
 		apcs->clk = platform_device_register_data(&pdev->dev,
 							  apcs_data->clk_name,
-							  PLATFORM_DEVID_NONE,
+							  PLATFORM_DEVID_AUTO,
 							  NULL, 0);
 		if (IS_ERR(apcs->clk))
 			dev_err(&pdev->dev, "failed to register APCS clk\n");
Signature · 1 line
2.30.2
author maintainer
#243 lore raw

PATCH 243/700 block: fix discard request merge

In reply to Greg Kroah-Hartman

From: Ming Lei <[email protected]>

[ Upstream commit 2705dfb2094777e405e065105e307074af8965c1 ]

ll_new_hw_segment() is reached only in case of single range discard merge, and we don't have max discard segment size limit actually, so it is wrong to run the following check:

if (req->nr_phys_segments + nr_phys_segs > blk_rq_get_max_segments(req))

it may be always false since req->nr_phys_segments is initialized as one, and bio's segment count is still 1, blk_rq_get_max_segments(reg) is 1 too.

Fix the issue by not doing the check and bypassing the calculation of discard request's nr_phys_segments.

Based on analysis from Wang Shanker.

---
 block/blk-merge.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)
block/blk-merge.c +6 -2
index 4d97fb6dd226..bcdff1879c34 100644
@@ -559,10 +559,14 @@ static inline unsigned int blk_rq_get_max_segments(struct request *rq)
 static inline int ll_new_hw_segment(struct request *req, struct bio *bio,
 		unsigned int nr_phys_segs)
 {
-	if (req->nr_phys_segments + nr_phys_segs > blk_rq_get_max_segments(req))
+	if (blk_integrity_merge_bio(req->q, req, bio) == false)
 		goto no_merge;
 
-	if (blk_integrity_merge_bio(req->q, req, bio) == false)
+	/* discard request merge won't add new segment */
+	if (req_op(req) == REQ_OP_DISCARD)
+		return 1;
+
+	if (req->nr_phys_segments + nr_phys_segs > blk_rq_get_max_segments(req))
 		goto no_merge;
 
 	/*
Signature · 1 line
2.30.2
author maintainer
#244 lore raw

PATCH 244/700 kthread_worker: fix return value when kthread_mod_delayed_work() races with kthread_cancel_delayed_work_sync()

In reply to Greg Kroah-Hartman

From: Petr Mladek <[email protected]>

[ Upstream commit d71ba1649fa3c464c51ec7163e4b817345bff2c7 ]

kthread_mod_delayed_work() might race with kthread_cancel_delayed_work_sync() or another kthread_mod_delayed_work() call. The function lets the other operation win when it sees work->canceling counter set. And it returns @false.

But it should return @true as it is done by the related workqueue API, see mod_delayed_work_on().

The reason is that the return value might be used for reference counting. It has to distinguish the case when the number of queued works has changed or stayed the same.

The change is safe. kthread_mod_delayed_work() return value is not checked anywhere at the moment.

---
 kernel/kthread.c | 19 ++++++++++++-------
 1 file changed, 12 insertions(+), 7 deletions(-)
kernel/kthread.c +12 -7
index e8da89c714f5..44f89a602b00 100644
@@ -1161,14 +1161,14 @@ static bool __kthread_cancel_work(struct kthread_work *work)
  * modify @dwork's timer so that it expires after @delay. If @delay is zero,
  * @work is guaranteed to be queued immediately.
  *
- * Return: %true if @dwork was pending and its timer was modified,
- * %false otherwise.
+ * Return: %false if @dwork was idle and queued, %true otherwise.
  *
  * A special case is when the work is being canceled in parallel.
  * It might be caused either by the real kthread_cancel_delayed_work_sync()
  * or yet another kthread_mod_delayed_work() call. We let the other command
- * win and return %false here. The caller is supposed to synchronize these
- * operations a reasonable way.
+ * win and return %true here. The return value can be used for reference
+ * counting and the number of queued works stays the same. Anyway, the caller
+ * is supposed to synchronize these operations a reasonable way.
  *
  * This function is safe to call from any context including IRQ handler.
  * See __kthread_cancel_work() and kthread_delayed_work_timer_fn()
@@ -1180,13 +1180,15 @@ bool kthread_mod_delayed_work(struct kthread_worker *worker,
 {
 	struct kthread_work *work = &dwork->work;
 	unsigned long flags;
-	int ret = false;
+	int ret;
 
 	raw_spin_lock_irqsave(&worker->lock, flags);
 
 	/* Do not bother with canceling when never queued. */
-	if (!work->worker)
+	if (!work->worker) {
+		ret = false;
 		goto fast_queue;
+	}
 
 	/* Work must not be used with >1 worker, see kthread_queue_work() */
 	WARN_ON_ONCE(work->worker != worker);
@@ -1204,8 +1206,11 @@ bool kthread_mod_delayed_work(struct kthread_worker *worker,
 	 * be used for reference counting.
 	 */
 	kthread_cancel_delayed_work_timer(work, &flags);
-	if (work->canceling)
+	if (work->canceling) {
+		/* The number of works in the queue does not change. */
+		ret = true;
 		goto out;
+	}
 	ret = __kthread_cancel_work(work);
 
 fast_queue:
Signature · 1 line
2.30.2
author maintainer
#245 lore raw

PATCH 245/700 ia64: mca_drv: fix incorrect array size calculation

In reply to Greg Kroah-Hartman

From: Arnd Bergmann <[email protected]>

[ Upstream commit c5f320ff8a79501bb59338278336ec43acb9d7e2 ]

gcc points out a mistake in the mca driver that goes back to before the git history:

arch/ia64/kernel/mca_drv.c: In function 'init_record_index_pools':
arch/ia64/kernel/mca_drv.c:346:54: error: expression does not compute the number of elements in this array; element typ e is 'int', not 'size_t' {aka 'long unsigned int'} [-Werror=sizeof-array-div]
346 | for (i = 1; i < sizeof sal_log_sect_min_sizes/sizeof(size_t); i++)
| ^

This is the same as sizeof(size_t), which is two shorter than the actual array. Use the ARRAY_SIZE() macro to get the correct calculation instead.

---
 arch/ia64/kernel/mca_drv.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/ia64/kernel/mca_drv.c +1 -1
index 36a69b4e6169..5bfc79be4cef 100644
@@ -343,7 +343,7 @@ init_record_index_pools(void)
 
 	/* - 2 - */
 	sect_min_size = sal_log_sect_min_sizes[0];
-	for (i = 1; i < sizeof sal_log_sect_min_sizes/sizeof(size_t); i++)
+	for (i = 1; i < ARRAY_SIZE(sal_log_sect_min_sizes); i++)
 		if (sect_min_size > sal_log_sect_min_sizes[i])
 			sect_min_size = sal_log_sect_min_sizes[i];
 
Signature · 1 line
2.30.2
author maintainer
#246 lore raw

PATCH 246/700 writeback, cgroup: increment isw_nr_in_flight before grabbing an inode

In reply to Greg Kroah-Hartman

From: Roman Gushchin <[email protected]>

[ Upstream commit 8826ee4fe75051f8cbfa5d4a9aa70565938e724c ]

isw_nr_in_flight is used to determine whether the inode switch queue should be flushed from the umount path. Currently it's increased after grabbing an inode and even scheduling the switch work. It means the umount path can walk past cleanup_offline_cgwb() with active inode references, which can result in a "Busy inodes after unmount." message and use-after-free issues (with inode->i_sb which gets freed).

Fix it by incrementing isw_nr_in_flight before doing anything with the inode and decrementing in the case when switching wasn't scheduled.

The problem hasn't yet been seen in the real life and was discovered by Jan Kara by looking into the code.

---
 fs/fs-writeback.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)
fs/fs-writeback.c +3 -2
index 7c46d1588a19..d684f541af48 100644
@@ -505,6 +505,8 @@ static void inode_switch_wbs(struct inode *inode, int new_wb_id)
 	if (!isw)
 		return;
 
+	atomic_inc(&isw_nr_in_flight);
+
 	/* find and pin the new wb */
 	rcu_read_lock();
 	memcg_css = css_from_id(new_wb_id, &memory_cgrp_subsys);
@@ -535,11 +537,10 @@ static void inode_switch_wbs(struct inode *inode, int new_wb_id)
 	 * Let's continue after I_WB_SWITCH is guaranteed to be visible.
 	 */
 	call_rcu(&isw->rcu_head, inode_switch_wbs_rcu_fn);
-
-	atomic_inc(&isw_nr_in_flight);
 	return;
 
 out_free:
+	atomic_dec(&isw_nr_in_flight);
 	if (isw->new_wb)
 		wb_put(isw->new_wb);
 	kfree(isw);
Signature · 1 line
2.30.2
author maintainer
#247 lore raw

PATCH 247/700 mm: define default MAX_PTRS_PER_* in include/pgtable.h

In reply to Greg Kroah-Hartman

From: Daniel Axtens <[email protected]>

[ Upstream commit c0f8aa4fa815daacb6eca52cae04820d6aecb7c2 ]

Commit c65e774fb3f6 ("x86/mm: Make PGDIR_SHIFT and PTRS_PER_P4D variable") made PTRS_PER_P4D variable on x86 and introduced MAX_PTRS_PER_P4D as a constant for cases which need a compile-time constant (e.g. fixed-size arrays).

powerpc likewise has boot-time selectable MMU features which can cause other mm "constants" to vary. For KASAN, we have some static PTE/PMD/PUD/P4D arrays so we need compile-time maximums for all these constants. Extend the MAX_PTRS_PER_ idiom, and place default definitions in include/pgtable.h. These define MAX_PTRS_PER_x to be PTRS_PER_x unless an architecture has defined MAX_PTRS_PER_x in its arch headers.

Clean up pgtable-nop4d.h and s390's MAX_PTRS_PER_P4D definitions while we're at it: both can just pick up the default now.

---
 arch/s390/include/asm/pgtable.h     |  2 --
 include/asm-generic/pgtable-nop4d.h |  1 -
 include/linux/pgtable.h             | 22 ++++++++++++++++++++++
 3 files changed, 22 insertions(+), 3 deletions(-)
arch/s390/include/asm/pgtable.h +0 -2
index 29c7ecd5ad1d..b38f7b781564 100644
@@ -344,8 +344,6 @@ static inline int is_module_addr(void *addr)
 #define PTRS_PER_P4D	_CRST_ENTRIES
 #define PTRS_PER_PGD	_CRST_ENTRIES
 
-#define MAX_PTRS_PER_P4D	PTRS_PER_P4D
-
 /*
  * Segment table and region3 table entry encoding
  * (R = read-only, I = invalid, y = young bit):
include/asm-generic/pgtable-nop4d.h +0 -1
index ce2cbb3c380f..2f6b1befb129 100644
@@ -9,7 +9,6 @@
 typedef struct { pgd_t pgd; } p4d_t;
 
 #define P4D_SHIFT		PGDIR_SHIFT
-#define MAX_PTRS_PER_P4D	1
 #define PTRS_PER_P4D		1
 #define P4D_SIZE		(1UL << P4D_SHIFT)
 #define P4D_MASK		(~(P4D_SIZE-1))
include/linux/pgtable.h +22 -0
index 136b1d996075..6fbbd620f6db 100644
@@ -1580,4 +1580,26 @@ typedef unsigned int pgtbl_mod_mask;
 #define pte_leaf_size(x) PAGE_SIZE
 #endif
 
+/*
+ * Some architectures have MMUs that are configurable or selectable at boot
+ * time. These lead to variable PTRS_PER_x. For statically allocated arrays it
+ * helps to have a static maximum value.
+ */
+
+#ifndef MAX_PTRS_PER_PTE
+#define MAX_PTRS_PER_PTE PTRS_PER_PTE
+#endif
+
+#ifndef MAX_PTRS_PER_PMD
+#define MAX_PTRS_PER_PMD PTRS_PER_PMD
+#endif
+
+#ifndef MAX_PTRS_PER_PUD
+#define MAX_PTRS_PER_PUD PTRS_PER_PUD
+#endif
+
+#ifndef MAX_PTRS_PER_P4D
+#define MAX_PTRS_PER_P4D PTRS_PER_P4D
+#endif
+
 #endif /* _LINUX_PGTABLE_H */
Signature · 1 line
2.30.2
author maintainer
#248 lore raw

PATCH 248/700 spi: Allow to have all native CSs in use along with GPIOs

In reply to Greg Kroah-Hartman

From: Andy Shevchenko <[email protected]>

[ Upstream commit dbaca8e56ea3f23fa215f48c2d46dd03ede06e02 ]

The commit 7d93aecdb58d ("spi: Add generic support for unused native cs with cs-gpios") excludes the valid case for the controllers that doesn't need to switch native CS in order to perform the transfer, i.e. when

0 native
... ...
<n> - 1 native
<n> GPIO
<n> + 1 GPIO
... ...

where <n> defines maximum of native CSs supported by the controller.

To allow this, bail out from spi_get_gpio_descs() conditionally for the controllers which explicitly marked with SPI_MASTER_GPIO_SS.

Fixes 7d93aecdb58d ("spi: Add generic support for unused native cs with cs-gpios")
---
 drivers/spi/spi.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)
drivers/spi/spi.c +3 -2
index 789354ee6a11..af126053213f 100644
@@ -2631,8 +2631,9 @@ static int spi_get_gpio_descs(struct spi_controller *ctlr)
 	}
 
 	ctlr->unused_native_cs = ffz(native_cs_mask);
-	if (num_cs_gpios && ctlr->max_native_cs &&
-	    ctlr->unused_native_cs >= ctlr->max_native_cs) {
+
+	if ((ctlr->flags & SPI_MASTER_GPIO_SS) && num_cs_gpios &&
+	    ctlr->max_native_cs && ctlr->unused_native_cs >= ctlr->max_native_cs) {
 		dev_err(dev, "No unused native chip select available\n");
 		return -EINVAL;
 	}
Signature · 1 line
2.30.2
author maintainer
#249 lore raw

PATCH 249/700 spi: Avoid undefined behaviour when counting unused native CSs

In reply to Greg Kroah-Hartman

From: Andy Shevchenko <[email protected]>

[ Upstream commit f60d7270c8a3d2beb1c23ae0da42497afa3584c2 ]

ffz(), that has been used to count unused native CSs, might cause undefined behaviour when called against ~0U. To fix that, open code it with ffs(~value) - 1.

Fixes 7d93aecdb58d ("spi: Add generic support for unused native cs with cs-gpios")
---
 drivers/spi/spi.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/spi/spi.c +1 -1
index af126053213f..2350463bfb8f 100644
@@ -2630,7 +2630,7 @@ static int spi_get_gpio_descs(struct spi_controller *ctlr)
 		native_cs_mask |= BIT(i);
 	}
 
-	ctlr->unused_native_cs = ffz(native_cs_mask);
+	ctlr->unused_native_cs = ffs(~native_cs_mask) - 1;
 
 	if ((ctlr->flags & SPI_MASTER_GPIO_SS) && num_cs_gpios &&
 	    ctlr->max_native_cs && ctlr->unused_native_cs >= ctlr->max_native_cs) {
Signature · 1 line
2.30.2
author maintainer
#250 lore raw

PATCH 250/700 media: venus: Rework error fail recover logic

In reply to Greg Kroah-Hartman

From: Mauro Carvalho Chehab <[email protected]>

[ Upstream commit 4cba5473c5ce0f1389d316c5dc6f83a0259df5eb ]

The Venus code has a sort of watchdog that attempts to recover from IP errors, implemented as a delayed work job, which calls venus_sys_error_handler().

Right now, it has several issues:

1. It assumes that PM runtime resume never fails

2. It internally runs two while() loops that also assume that PM runtime will never fail to go idle:

	while (pm_runtime_active(core->dev_dec) || pm_runtime_active(core->dev_enc))
		msleep(10);

...

	while (core->pmdomains[0] && pm_runtime_active(core->pmdomains[0]))
		usleep_range(1000, 1500);

3. It uses an OR to merge all return codes and then report to the user

4. If the hardware never recovers, it keeps running on every 10ms, flooding the syslog with 2 messages (so, up to 200 messages per second).

Rework the code, in order to prevent that, by:

1. check the return code from PM runtime resume;
2. don't let the while() loops run forever;
3. store the failed event;
4. use warn ratelimited when it fails to recover.

Fixes af2c3834c8ca ("[media] media: venus: adding core part and helper functions")
---
 drivers/media/platform/qcom/venus/core.c | 60 +++++++++++++++++++-----
 1 file changed, 47 insertions(+), 13 deletions(-)
drivers/media/platform/qcom/venus/core.c +47 -13
index ae374bb2a48f..28443547ae8f 100644
@@ -76,22 +76,32 @@ static const struct hfi_core_ops venus_core_ops = {
 	.event_notify = venus_event_notify,
 };
 
+#define RPM_WAIT_FOR_IDLE_MAX_ATTEMPTS 10
+
 static void venus_sys_error_handler(struct work_struct *work)
 {
 	struct venus_core *core =
 			container_of(work, struct venus_core, work.work);
-	int ret = 0;
-
-	pm_runtime_get_sync(core->dev);
+	int ret, i, max_attempts = RPM_WAIT_FOR_IDLE_MAX_ATTEMPTS;
+	const char *err_msg = "";
+	bool failed = false;
+
+	ret = pm_runtime_get_sync(core->dev);
+	if (ret < 0) {
+		err_msg = "resume runtime PM";
+		max_attempts = 0;
+		failed = true;
+	}
 
 	hfi_core_deinit(core, true);
 
-	dev_warn(core->dev, "system error has occurred, starting recovery!\n");
-
 	mutex_lock(&core->lock);
 
-	while (pm_runtime_active(core->dev_dec) || pm_runtime_active(core->dev_enc))
+	for (i = 0; i < max_attempts; i++) {
+		if (!pm_runtime_active(core->dev_dec) && !pm_runtime_active(core->dev_enc))
+			break;
 		msleep(10);
+	}
 
 	venus_shutdown(core);
 
@@ -99,31 +109,55 @@ static void venus_sys_error_handler(struct work_struct *work)
 
 	pm_runtime_put_sync(core->dev);
 
-	while (core->pmdomains[0] && pm_runtime_active(core->pmdomains[0]))
+	for (i = 0; i < max_attempts; i++) {
+		if (!core->pmdomains[0] || !pm_runtime_active(core->pmdomains[0]))
+			break;
 		usleep_range(1000, 1500);
+	}
 
 	hfi_reinit(core);
 
-	pm_runtime_get_sync(core->dev);
+	ret = pm_runtime_get_sync(core->dev);
+	if (ret < 0) {
+		err_msg = "resume runtime PM";
+		failed = true;
+	}
 
-	ret |= venus_boot(core);
-	ret |= hfi_core_resume(core, true);
+	ret = venus_boot(core);
+	if (ret && !failed) {
+		err_msg = "boot Venus";
+		failed = true;
+	}
+
+	ret = hfi_core_resume(core, true);
+	if (ret && !failed) {
+		err_msg = "resume HFI";
+		failed = true;
+	}
 
 	enable_irq(core->irq);
 
 	mutex_unlock(&core->lock);
 
-	ret |= hfi_core_init(core);
+	ret = hfi_core_init(core);
+	if (ret && !failed) {
+		err_msg = "init HFI";
+		failed = true;
+	}
 
 	pm_runtime_put_sync(core->dev);
 
-	if (ret) {
+	if (failed) {
 		disable_irq_nosync(core->irq);
-		dev_warn(core->dev, "recovery failed (%d)\n", ret);
+		dev_warn_ratelimited(core->dev,
+				     "System error has occurred, recovery failed to %s\n",
+				     err_msg);
 		schedule_delayed_work(&core->work, msecs_to_jiffies(10));
 		return;
 	}
 
+	dev_warn(core->dev, "system error has occurred (recovered)\n");
+
 	mutex_lock(&core->lock);
 	core->sys_error = false;
 	mutex_unlock(&core->lock);
Signature · 1 line
2.30.2
author maintainer
#251 lore raw

PATCH 251/700 media: s5p_cec: decrement usage count if disabled

In reply to Greg Kroah-Hartman

From: Mauro Carvalho Chehab <[email protected]>

[ Upstream commit 747bad54a677d8633ec14b39dfbeb859c821d7f2 ]

There's a bug at s5p_cec_adap_enable(): if called to disable the device, it should call pm_runtime_put() instead of pm_runtime_disable(), as the goal here is to decrement the usage_count and not to disable PM runtime.

Fixes 1bcbf6f4b6b0 ("[media] cec: s5p-cec: Add s5p-cec driver")
---
 drivers/media/cec/platform/s5p/s5p_cec.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/media/cec/platform/s5p/s5p_cec.c +1 -1
index 2250c1cbc64e..028a09a7531e 100644
@@ -54,7 +54,7 @@ static int s5p_cec_adap_enable(struct cec_adapter *adap, bool enable)
 	} else {
 		s5p_cec_mask_tx_interrupts(cec);
 		s5p_cec_mask_rx_interrupts(cec);
-		pm_runtime_disable(cec->dev);
+		pm_runtime_put(cec->dev);
 	}
 
 	return 0;
Signature · 1 line
2.30.2
author maintainer
#252 lore raw

PATCH 252/700 media: i2c: ccs-core: return the right error code at suspend

In reply to Greg Kroah-Hartman

From: Mauro Carvalho Chehab <[email protected]>

[ Upstream commit 6005a8e955e4e451e4bf6000affaab566d4cab5e ]

If pm_runtime resume logic fails, return the error code provided by it, instead of -EAGAIN, as, depending on what caused it to fail, it may not be something that would be recovered.

Fixes cbba45d43631 ("[media] smiapp: Use runtime PM")
---
 drivers/media/i2c/ccs/ccs-core.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/media/i2c/ccs/ccs-core.c +1 -1
index 4505594996bd..fde0c51f0406 100644
@@ -3093,7 +3093,7 @@ static int __maybe_unused ccs_suspend(struct device *dev)
 	if (rval < 0) {
 		pm_runtime_put_noidle(dev);
 
-		return -EAGAIN;
+		return rval;
 	}
 
 	if (sensor->streaming)
Signature · 1 line
2.30.2
author maintainer
#253 lore raw

PATCH 253/700 media: hantro: do a PM resume earlier

In reply to Greg Kroah-Hartman

From: Mauro Carvalho Chehab <[email protected]>

[ Upstream commit 892bb6ecead9b834ba7ad1d07513e9eba1baa3a4 ]

The device_run() first enables the clock and then
tries to resume PM runtime, checking for errors.

Well, if for some reason the pm_runtime can not resume, it would be better to detect it beforehand.

So, change the order inside device_run().

Fixes 775fec69008d ("media: add Rockchip VPU JPEG encoder driver")
---
 drivers/staging/media/hantro/hantro_drv.c | 33 +++++++++++++++--------
 1 file changed, 22 insertions(+), 11 deletions(-)
drivers/staging/media/hantro/hantro_drv.c +22 -11
index e5f200e64993..2d6e0056be62 100644
@@ -56,16 +56,12 @@ dma_addr_t hantro_get_ref(struct hantro_ctx *ctx, u64 ts)
 	return hantro_get_dec_buf_addr(ctx, buf);
 }
 
-static void hantro_job_finish(struct hantro_dev *vpu,
-			      struct hantro_ctx *ctx,
-			      enum vb2_buffer_state result)
+static void hantro_job_finish_no_pm(struct hantro_dev *vpu,
+				    struct hantro_ctx *ctx,
+				    enum vb2_buffer_state result)
 {
 	struct vb2_v4l2_buffer *src, *dst;
 
-	pm_runtime_mark_last_busy(vpu->dev);
-	pm_runtime_put_autosuspend(vpu->dev);
-	clk_bulk_disable(vpu->variant->num_clocks, vpu->clocks);
-
 	src = v4l2_m2m_next_src_buf(ctx->fh.m2m_ctx);
 	dst = v4l2_m2m_next_dst_buf(ctx->fh.m2m_ctx);
 
@@ -81,6 +77,18 @@ static void hantro_job_finish(struct hantro_dev *vpu,
 					 result);
 }
 
+static void hantro_job_finish(struct hantro_dev *vpu,
+			      struct hantro_ctx *ctx,
+			      enum vb2_buffer_state result)
+{
+	pm_runtime_mark_last_busy(vpu->dev);
+	pm_runtime_put_autosuspend(vpu->dev);
+
+	clk_bulk_disable(vpu->variant->num_clocks, vpu->clocks);
+
+	hantro_job_finish_no_pm(vpu, ctx, result);
+}
+
 void hantro_irq_done(struct hantro_dev *vpu,
 		     enum vb2_buffer_state result)
 {
@@ -152,12 +160,15 @@ static void device_run(void *priv)
 	src = hantro_get_src_buf(ctx);
 	dst = hantro_get_dst_buf(ctx);
 
+	ret = pm_runtime_get_sync(ctx->dev->dev);
+	if (ret < 0) {
+		pm_runtime_put_noidle(ctx->dev->dev);
+		goto err_cancel_job;
+	}
+
 	ret = clk_bulk_enable(ctx->dev->variant->num_clocks, ctx->dev->clocks);
 	if (ret)
 		goto err_cancel_job;
-	ret = pm_runtime_get_sync(ctx->dev->dev);
-	if (ret < 0)
-		goto err_cancel_job;
 
 	v4l2_m2m_buf_copy_metadata(src, dst, true);
 
@@ -165,7 +176,7 @@ static void device_run(void *priv)
 	return;
 
 err_cancel_job:
-	hantro_job_finish(ctx->dev, ctx, VB2_BUF_STATE_ERROR);
+	hantro_job_finish_no_pm(ctx->dev, ctx, VB2_BUF_STATE_ERROR);
 }
 
 static struct v4l2_m2m_ops vpu_m2m_ops = {
Signature · 1 line
2.30.2
author maintainer
#254 lore raw

PATCH 254/700 crypto: ixp4xx - dma_unmap the correct address

In reply to Greg Kroah-Hartman

From: Corentin Labbe <[email protected]>

[ Upstream commit 9395c58fdddd79cdd3882132cdd04e8ac7ad525f ]

Testing ixp4xx_crypto with CONFIG_DMA_API_DEBUG lead to the following error:
DMA-API: platform ixp4xx_crypto.0: device driver tries to free DMA memory it has not allocated [device address=0x0000000000000000] [size=24 bytes]

This is due to dma_unmap using the wrong address.

Fixes 0d44dc59b2b4 ("crypto: ixp4xx - Fix handling of chained sg buffers")
---
 drivers/crypto/ixp4xx_crypto.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/crypto/ixp4xx_crypto.c +1 -1
index 8b0f17fc09fb..9e330e93e340 100644
@@ -330,7 +330,7 @@ static void free_buf_chain(struct device *dev, struct buffer_desc *buf,
 
 		buf1 = buf->next;
 		phys1 = buf->phys_next;
-		dma_unmap_single(dev, buf->phys_next, buf->buf_len, buf->dir);
+		dma_unmap_single(dev, buf->phys_addr, buf->buf_len, buf->dir);
 		dma_pool_free(buffer_pool, buf, phys);
 		buf = buf1;
 		phys = phys1;
Signature · 1 line
2.30.2
author maintainer
#255 lore raw

PATCH 255/700 crypto: ixp4xx - update IV after requests

In reply to Greg Kroah-Hartman

From: Corentin Labbe <[email protected]>

[ Upstream commit e8acf011f2e7e21a7e2fae47cbaa06598e533d40 ]

Crypto selftests fail on ixp4xx since it do not update IV after skcipher requests.

Fixes 81bef0150074 ("crypto: ixp4xx - Hardware crypto support for IXP4xx CPUs")
---
 drivers/crypto/ixp4xx_crypto.c | 22 ++++++++++++++++++++++
 1 file changed, 22 insertions(+)
drivers/crypto/ixp4xx_crypto.c +22 -0
index 9e330e93e340..7567456a21a0 100644
@@ -149,6 +149,8 @@ struct crypt_ctl {
 struct ablk_ctx {
 	struct buffer_desc *src;
 	struct buffer_desc *dst;
+	u8 iv[MAX_IVLEN];
+	bool encrypt;
 };
 
 struct aead_ctx {
@@ -381,6 +383,20 @@ static void one_packet(dma_addr_t phys)
 	case CTL_FLAG_PERFORM_ABLK: {
 		struct skcipher_request *req = crypt->data.ablk_req;
 		struct ablk_ctx *req_ctx = skcipher_request_ctx(req);
+		struct crypto_skcipher *tfm = crypto_skcipher_reqtfm(req);
+		unsigned int ivsize = crypto_skcipher_ivsize(tfm);
+		unsigned int offset;
+
+		if (ivsize > 0) {
+			offset = req->cryptlen - ivsize;
+			if (req_ctx->encrypt) {
+				scatterwalk_map_and_copy(req->iv, req->dst,
+							 offset, ivsize, 0);
+			} else {
+				memcpy(req->iv, req_ctx->iv, ivsize);
+				memzero_explicit(req_ctx->iv, ivsize);
+			}
+		}
 
 		if (req_ctx->dst) {
 			free_buf_chain(dev, req_ctx->dst, crypt->dst_buf);
@@ -876,6 +892,7 @@ static int ablk_perform(struct skcipher_request *req, int encrypt)
 	struct ablk_ctx *req_ctx = skcipher_request_ctx(req);
 	struct buffer_desc src_hook;
 	struct device *dev = &pdev->dev;
+	unsigned int offset;
 	gfp_t flags = req->base.flags & CRYPTO_TFM_REQ_MAY_SLEEP ?
 				GFP_KERNEL : GFP_ATOMIC;
 
@@ -885,6 +902,7 @@ static int ablk_perform(struct skcipher_request *req, int encrypt)
 		return -EAGAIN;
 
 	dir = encrypt ? &ctx->encrypt : &ctx->decrypt;
+	req_ctx->encrypt = encrypt;
 
 	crypt = get_crypt_desc();
 	if (!crypt)
@@ -900,6 +918,10 @@ static int ablk_perform(struct skcipher_request *req, int encrypt)
 
 	BUG_ON(ivsize && !req->iv);
 	memcpy(crypt->iv, req->iv, ivsize);
+	if (ivsize > 0 && !encrypt) {
+		offset = req->cryptlen - ivsize;
+		scatterwalk_map_and_copy(req_ctx->iv, req->src, offset, ivsize, 0);
+	}
 	if (req->src != req->dst) {
 		struct buffer_desc dst_hook;
 		crypt->mode |= NPE_OP_NOT_IN_PLACE;
Signature · 1 line
2.30.2
author maintainer
#256 lore raw

PATCH 256/700 crypto: ux500 - Fix error return code in hash_hw_final()

In reply to Greg Kroah-Hartman

From: Zhen Lei <[email protected]>

[ Upstream commit b01360384009ab066940b45f34880991ea7ccbfb ]

Fix to return a negative error code from the error handling case instead of 0, as done elsewhere in this function.

Fixes 8a63b1994c50 ("crypto: ux500 - Add driver for HASH hardware")
---
 drivers/crypto/ux500/hash/hash_core.c | 1 +
 1 file changed, 1 insertion(+)
drivers/crypto/ux500/hash/hash_core.c +1 -0
index da284b0ea1b2..243515df609b 100644
@@ -1010,6 +1010,7 @@ static int hash_hw_final(struct ahash_request *req)
 			goto out;
 		}
 	} else if (req->nbytes == 0 && ctx->keylen > 0) {
+		ret = -EPERM;
 		dev_err(device_data->dev, "%s: Empty message with keylength > 0, NOT supported\n",
 			__func__);
 		goto out;
Signature · 1 line
2.30.2
author maintainer
#257 lore raw

PATCH 257/700 sata_highbank: fix deferred probing

In reply to Greg Kroah-Hartman

From: Sergey Shtylyov <[email protected]>

[ Upstream commit 4a24efa16e7db02306fb5db84518bb0a7ada5a46 ]

The driver overrides the error codes returned by platform_get_irq() to -EINVAL, so if it returns -EPROBE_DEFER, the driver would fail the probe permanently instead of the deferred probing. Switch to propagating the error code upstream, still checking/overriding IRQ0 as libata regards it as "no IRQ" (thus polling) anyway...

Fixes 9ec36cafe43b ("of/irq: do irq resolution in platform_get_irq")
---
 drivers/ata/sata_highbank.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/ata/sata_highbank.c +4 -2
index 64b2ef15ec19..8440203e835e 100644
@@ -469,10 +469,12 @@ static int ahci_highbank_probe(struct platform_device *pdev)
 	}
 
 	irq = platform_get_irq(pdev, 0);
-	if (irq <= 0) {
+	if (irq < 0) {
 		dev_err(dev, "no irq\n");
-		return -EINVAL;
+		return irq;
 	}
+	if (!irq)
+		return -EINVAL;
 
 	hpriv = devm_kzalloc(dev, sizeof(*hpriv), GFP_KERNEL);
 	if (!hpriv) {
Signature · 1 line
2.30.2
author maintainer
#258 lore raw

PATCH 258/700 pata_rb532_cf: fix deferred probing

In reply to Greg Kroah-Hartman

From: Sergey Shtylyov <[email protected]>

[ Upstream commit 2d3a62fbae8e5badc2342388f65ab2191c209cc0 ]

The driver overrides the error codes returned by platform_get_irq() to -ENOENT, so if it returns -EPROBE_DEFER, the driver would fail the probe permanently instead of the deferred probing. Switch to propagating the error code upstream, still checking/overriding IRQ0 as libata regards it as "no IRQ" (thus polling) anyway...

Fixes 9ec36cafe43b ("of/irq: do irq resolution in platform_get_irq")
---
 drivers/ata/pata_rb532_cf.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/ata/pata_rb532_cf.c +4 -2
index 479c4b29b856..303f8c375b3a 100644
@@ -115,10 +115,12 @@ static int rb532_pata_driver_probe(struct platform_device *pdev)
 	}
 
 	irq = platform_get_irq(pdev, 0);
-	if (irq <= 0) {
+	if (irq < 0) {
 		dev_err(&pdev->dev, "no IRQ resource found\n");
-		return -ENOENT;
+		return irq;
 	}
+	if (!irq)
+		return -EINVAL;
 
 	gpiod = devm_gpiod_get(&pdev->dev, NULL, GPIOD_IN);
 	if (IS_ERR(gpiod)) {
Signature · 1 line
2.30.2
author maintainer
#259 lore raw

PATCH 259/700 media: I2C: change RST to "RSET" to fix multiple build errors

In reply to Greg Kroah-Hartman

From: Randy Dunlap <[email protected]>

[ Upstream commit 8edcb5049ac29aa3c8acc5ef15dd4036543d747e ]

The use of an enum named 'RST' conflicts with a #define macro named 'RST' in arch/mips/include/asm/mach-rc32434/rb.h.

The MIPS use of RST was there first (AFAICT), so change the media/i2c/ uses of RST to be named 'RSET'.
'git grep -w RSET' does not report any naming conflicts with the new name.

This fixes multiple build errors:

arch/mips/include/asm/mach-rc32434/rb.h:15:14: error: expected identifier before '(' token

   15 | #define RST  (1 << 15)
      |              ^
drivers/media/i2c/s5c73m3/s5c73m3.h:356:2: note: in expansion of macro 'RST'
  356 |  RST,
      |  ^~~

../arch/mips/include/asm/mach-rc32434/rb.h:15:14: error: expected identifier before '(' token

   15 | #define RST  (1 << 15)
      |              ^
../drivers/media/i2c/s5k6aa.c:180:2: note: in expansion of macro 'RST'
  180 |  RST,
      |  ^~~

../arch/mips/include/asm/mach-rc32434/rb.h:15:14: error: expected identifier before '(' token

   15 | #define RST  (1 << 15)
      |              ^
../drivers/media/i2c/s5k5baf.c:238:2: note: in expansion of macro 'RST'
  238 |  RST,
      |  ^~~

and some others that I have trimmed.

Fixes cac47f1822fc ("[media] V4L: Add S5C73M3 camera driver")
Fixes 8b99312b7214 ("[media] Add v4l2 subdev driver for S5K4ECGX sensor")
Fixes 7d459937dc09 ("[media] Add driver for Samsung S5K5BAF camera sensor")
Fixes bfa8dd3a0524 ("[media] v4l: Add v4l2 subdev driver for S5K6AAFX sensor")
Cc Shawn Guo <[email protected]>
Cc Sascha Hauer <[email protected]>
Cc Pengutronix Kernel Team <[email protected]>
Cc Fabio Estevam <[email protected]>
Cc NXP Linux Team <[email protected]>
Cc [email protected] (moderated for non-subscribers)
Cc Andrzej Hajda <[email protected]>
Cc Sylwester Nawrocki <[email protected]>
Cc Sangwook Lee <[email protected]>
---
 drivers/media/i2c/s5c73m3/s5c73m3-core.c |  6 +++---
 drivers/media/i2c/s5c73m3/s5c73m3.h      |  2 +-
 drivers/media/i2c/s5k4ecgx.c             | 10 +++++-----
 drivers/media/i2c/s5k5baf.c              |  6 +++---
 drivers/media/i2c/s5k6aa.c               | 10 +++++-----
 5 files changed, 17 insertions(+), 17 deletions(-)
drivers/media/i2c/s5c73m3/s5c73m3-core.c +3 -3
index 5b4c4a3547c9..71804a70bc6d 100644
@@ -1386,7 +1386,7 @@ static int __s5c73m3_power_on(struct s5c73m3 *state)
 	s5c73m3_gpio_deassert(state, STBY);
 	usleep_range(100, 200);
 
-	s5c73m3_gpio_deassert(state, RST);
+	s5c73m3_gpio_deassert(state, RSET);
 	usleep_range(50, 100);
 
 	return 0;
@@ -1401,7 +1401,7 @@ static int __s5c73m3_power_off(struct s5c73m3 *state)
 {
 	int i, ret;
 
-	if (s5c73m3_gpio_assert(state, RST))
+	if (s5c73m3_gpio_assert(state, RSET))
 		usleep_range(10, 50);
 
 	if (s5c73m3_gpio_assert(state, STBY))
@@ -1606,7 +1606,7 @@ static int s5c73m3_get_platform_data(struct s5c73m3 *state)
 
 		state->mclk_frequency = pdata->mclk_frequency;
 		state->gpio[STBY] = pdata->gpio_stby;
-		state->gpio[RST] = pdata->gpio_reset;
+		state->gpio[RSET] = pdata->gpio_reset;
 		return 0;
 	}
 
drivers/media/i2c/s5c73m3/s5c73m3.h +1 -1
index ef7e85b34263..c3fcfdd3ea66 100644
@@ -353,7 +353,7 @@ struct s5c73m3_ctrls {
 
 enum s5c73m3_gpio_id {
 	STBY,
-	RST,
+	RSET,
 	GPIO_NUM,
 };
 
drivers/media/i2c/s5k4ecgx.c +5 -5
index b2d53417badf..4e97309a67f4 100644
@@ -173,7 +173,7 @@ static const char * const s5k4ecgx_supply_names[] = {
 
 enum s5k4ecgx_gpio_id {
 	STBY,
-	RST,
+	RSET,
 	GPIO_NUM,
 };
 
@@ -476,7 +476,7 @@ static int __s5k4ecgx_power_on(struct s5k4ecgx *priv)
 	if (s5k4ecgx_gpio_set_value(priv, STBY, priv->gpio[STBY].level))
 		usleep_range(30, 50);
 
-	if (s5k4ecgx_gpio_set_value(priv, RST, priv->gpio[RST].level))
+	if (s5k4ecgx_gpio_set_value(priv, RSET, priv->gpio[RSET].level))
 		usleep_range(30, 50);
 
 	return 0;
@@ -484,7 +484,7 @@ static int __s5k4ecgx_power_on(struct s5k4ecgx *priv)
 
 static int __s5k4ecgx_power_off(struct s5k4ecgx *priv)
 {
-	if (s5k4ecgx_gpio_set_value(priv, RST, !priv->gpio[RST].level))
+	if (s5k4ecgx_gpio_set_value(priv, RSET, !priv->gpio[RSET].level))
 		usleep_range(30, 50);
 
 	if (s5k4ecgx_gpio_set_value(priv, STBY, !priv->gpio[STBY].level))
@@ -872,7 +872,7 @@ static int s5k4ecgx_config_gpios(struct s5k4ecgx *priv,
 	int ret;
 
 	priv->gpio[STBY].gpio = -EINVAL;
-	priv->gpio[RST].gpio  = -EINVAL;
+	priv->gpio[RSET].gpio  = -EINVAL;
 
 	ret = s5k4ecgx_config_gpio(gpio->gpio, gpio->level, "S5K4ECGX_STBY");
 
@@ -891,7 +891,7 @@ static int s5k4ecgx_config_gpios(struct s5k4ecgx *priv,
 		s5k4ecgx_free_gpios(priv);
 		return ret;
 	}
-	priv->gpio[RST] = *gpio;
+	priv->gpio[RSET] = *gpio;
 	if (gpio_is_valid(gpio->gpio))
 		gpio_set_value(gpio->gpio, 0);
 
drivers/media/i2c/s5k5baf.c +3 -3
index ec6f22efe19a..ec65a8e084c6 100644
@@ -235,7 +235,7 @@ struct s5k5baf_gpio {
 
 enum s5k5baf_gpio_id {
 	STBY,
-	RST,
+	RSET,
 	NUM_GPIOS,
 };
 
@@ -969,7 +969,7 @@ static int s5k5baf_power_on(struct s5k5baf *state)
 
 	s5k5baf_gpio_deassert(state, STBY);
 	usleep_range(50, 100);
-	s5k5baf_gpio_deassert(state, RST);
+	s5k5baf_gpio_deassert(state, RSET);
 	return 0;
 
 err_reg_dis:
@@ -987,7 +987,7 @@ static int s5k5baf_power_off(struct s5k5baf *state)
 	state->apply_cfg = 0;
 	state->apply_crop = 0;
 
-	s5k5baf_gpio_assert(state, RST);
+	s5k5baf_gpio_assert(state, RSET);
 	s5k5baf_gpio_assert(state, STBY);
 
 	if (!IS_ERR(state->clock))
drivers/media/i2c/s5k6aa.c +5 -5
index 72439fae7968..6516e205e9a3 100644
@@ -177,7 +177,7 @@ static const char * const s5k6aa_supply_names[] = {
 
 enum s5k6aa_gpio_id {
 	STBY,
-	RST,
+	RSET,
 	GPIO_NUM,
 };
 
@@ -841,7 +841,7 @@ static int __s5k6aa_power_on(struct s5k6aa *s5k6aa)
 		ret = s5k6aa->s_power(1);
 	usleep_range(4000, 5000);
 
-	if (s5k6aa_gpio_deassert(s5k6aa, RST))
+	if (s5k6aa_gpio_deassert(s5k6aa, RSET))
 		msleep(20);
 
 	return ret;
@@ -851,7 +851,7 @@ static int __s5k6aa_power_off(struct s5k6aa *s5k6aa)
 {
 	int ret;
 
-	if (s5k6aa_gpio_assert(s5k6aa, RST))
+	if (s5k6aa_gpio_assert(s5k6aa, RSET))
 		usleep_range(100, 150);
 
 	if (s5k6aa->s_power) {
@@ -1510,7 +1510,7 @@ static int s5k6aa_configure_gpios(struct s5k6aa *s5k6aa,
 	int ret;
 
 	s5k6aa->gpio[STBY].gpio = -EINVAL;
-	s5k6aa->gpio[RST].gpio  = -EINVAL;
+	s5k6aa->gpio[RSET].gpio  = -EINVAL;
 
 	gpio = &pdata->gpio_stby;
 	if (gpio_is_valid(gpio->gpio)) {
@@ -1533,7 +1533,7 @@ static int s5k6aa_configure_gpios(struct s5k6aa *s5k6aa,
 		if (ret < 0)
 			return ret;
 
-		s5k6aa->gpio[RST] = *gpio;
+		s5k6aa->gpio[RSET] = *gpio;
 	}
 
 	return 0;
Signature · 9 lines
2.30.2




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

PATCH 260/700 sched/uclamp: Fix wrong implementation of cpu.uclamp.min

In reply to Greg Kroah-Hartman

From: Qais Yousef <[email protected]>

[ Upstream commit 0c18f2ecfcc274a4bcc1d122f79ebd4001c3b445 ]

cpu.uclamp.min is a protection as described in cgroup-v2 Resource Distribution Model

Documentation/admin-guide/cgroup-v2.rst

which means we try our best to preserve the minimum performance point of tasks in this group. See full description of cpu.uclamp.min in the cgroup-v2.rst.

But the current implementation makes it a limit, which is not what was intended.

For example:

	tg->cpu.uclamp.min = 20%

	p0->uclamp[UCLAMP_MIN] = 0
	p1->uclamp[UCLAMP_MIN] = 50%

	Previous Behavior (limit):

		p0->effective_uclamp = 0
		p1->effective_uclamp = 20%

	New Behavior (Protection):

		p0->effective_uclamp = 20%
		p1->effective_uclamp = 50%

Which is inline with how protections should work.

With this change the cgroup and per-task behaviors are the same, as expected.

Additionally, we remove the confusing relationship between cgroup and !user_defined flag.

We don't want for example RT tasks that are boosted by default to max to change their boost value when they attach to a cgroup. If a cgroup wants to limit the max performance point of tasks attached to it, then cpu.uclamp.max must be set accordingly.

Or if they want to set different boost value based on cgroup, then sysctl_sched_util_clamp_min_rt_default must be used to NOT boost to max and set the right cpu.uclamp.min for each group to let the RT tasks obtain the desired boost value when attached to that group.

As it stands the dependency on !user_defined flag adds an extra layer of complexity that is not required now cpu.uclamp.min behaves properly as a protection.

The propagation model of effective cpu.uclamp.min in child cgroups as implemented by cpu_util_update_eff() is still correct. The parent protection sets an upper limit of what the child cgroups will effectively get.

Fixes 3eac870a3247 (sched/uclamp: Use TG's clamps to restrict TASK's clamps)
---
 kernel/sched/core.c | 21 +++++++++++++++++----
 1 file changed, 17 insertions(+), 4 deletions(-)
kernel/sched/core.c +17 -4
index 17f612045271..3fe7daf9d31d 100644
@@ -1057,7 +1057,6 @@ uclamp_tg_restrict(struct task_struct *p, enum uclamp_id clamp_id)
 {
 	struct uclamp_se uc_req = p->uclamp_req[clamp_id];
 #ifdef CONFIG_UCLAMP_TASK_GROUP
-	struct uclamp_se uc_max;
 
 	/*
 	 * Tasks in autogroups or root task group will be
@@ -1068,9 +1067,23 @@ uclamp_tg_restrict(struct task_struct *p, enum uclamp_id clamp_id)
 	if (task_group(p) == &root_task_group)
 		return uc_req;
 
-	uc_max = task_group(p)->uclamp[clamp_id];
-	if (uc_req.value > uc_max.value || !uc_req.user_defined)
-		return uc_max;
+	switch (clamp_id) {
+	case UCLAMP_MIN: {
+		struct uclamp_se uc_min = task_group(p)->uclamp[clamp_id];
+		if (uc_req.value < uc_min.value)
+			return uc_min;
+		break;
+	}
+	case UCLAMP_MAX: {
+		struct uclamp_se uc_max = task_group(p)->uclamp[clamp_id];
+		if (uc_req.value > uc_max.value)
+			return uc_max;
+		break;
+	}
+	default:
+		WARN_ON_ONCE(1);
+		break;
+	}
 #endif
 
 	return uc_req;
Signature · 1 line
2.30.2
author maintainer
#261 lore raw

PATCH 261/700 sched/uclamp: Fix locking around cpu_util_update_eff()

In reply to Greg Kroah-Hartman

From: Qais Yousef <[email protected]>

[ Upstream commit 93b73858701fd01de26a4a874eb95f9b7156fd4b ]

cpu_cgroup_css_online() calls cpu_util_update_eff() without holding the uclamp_mutex or rcu_read_lock() like other call sites, which is a mistake.

The uclamp_mutex is required to protect against concurrent reads and writes that could update the cgroup hierarchy.

The rcu_read_lock() is required to traverse the cgroup data structures in cpu_util_update_eff().

Surround the caller with the required locks and add some asserts to better document the dependency in cpu_util_update_eff().

Fixes 7226017ad37a ("sched/uclamp: Fix a bug in propagating uclamp value in new cgroups")
---
 kernel/sched/core.c | 7 +++++++
 1 file changed, 7 insertions(+)
kernel/sched/core.c +7 -0
index 3fe7daf9d31d..f59166fe499a 100644
@@ -8693,7 +8693,11 @@ static int cpu_cgroup_css_online(struct cgroup_subsys_state *css)
 
 #ifdef CONFIG_UCLAMP_TASK_GROUP
 	/* Propagate the effective uclamp value for the new group */
+	mutex_lock(&uclamp_mutex);
+	rcu_read_lock();
 	cpu_util_update_eff(css);
+	rcu_read_unlock();
+	mutex_unlock(&uclamp_mutex);
 #endif
 
 	return 0;
@@ -8783,6 +8787,9 @@ static void cpu_util_update_eff(struct cgroup_subsys_state *css)
 	enum uclamp_id clamp_id;
 	unsigned int clamps;
 
+	lockdep_assert_held(&uclamp_mutex);
+	SCHED_WARN_ON(!rcu_read_lock_held());
+
 	css_for_each_descendant_pre(css, top_css) {
 		uc_parent = css_tg(css)->parent
 			? css_tg(css)->parent->uclamp : NULL;
Signature · 1 line
2.30.2
author maintainer
#262 lore raw

PATCH 262/700 kbuild: Fix objtool dependency for OBJECT_FILES_NON_STANDARD_<obj> := n

In reply to Greg Kroah-Hartman

From: Josh Poimboeuf <[email protected]>

[ Upstream commit 8852c552402979508fdc395ae07aa8761aa46045 ]

"OBJECT_FILES_NON_STANDARD_vma.o := n" has a dependency bug. When objtool source is updated, the affected object doesn't get re-analyzed by objtool.

Peter's new variable-sized jump label feature relies on objtool rewriting the object file. Otherwise the system can fail to boot. That effectively upgrades this minor dependency issue to a major bug.

The problem is that variables in prerequisites are expanded early, during the read-in phase. The '$(objtool_dep)' variable indirectly uses '$@', which isn't yet available when the target prerequisites are evaluated.

Use '.SECONDEXPANSION:' which causes '$(objtool_dep)' to be expanded in a later phase, after the target-specific '$@' variable has been defined.

Fixes b9ab5ebb14ec ("objtool: Add CONFIG_STACK_VALIDATION option")
Fixes ab3257042c26 ("jump_label, x86: Allow short NOPs")
---
 scripts/Makefile.build | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)
scripts/Makefile.build +3 -2
index 1b6094a13034..73701d637ed5 100644
@@ -267,7 +267,8 @@ define rule_as_o_S
 endef
 
 # Built-in and composite module parts
-$(obj)/%.o: $(src)/%.c $(recordmcount_source) $(objtool_dep) FORCE
+.SECONDEXPANSION:
+$(obj)/%.o: $(src)/%.c $(recordmcount_source) $$(objtool_dep) FORCE
 	$(call if_changed_rule,cc_o_c)
 	$(call cmd,force_checksrc)
 
@@ -348,7 +349,7 @@ cmd_modversions_S =								\
 	fi
 endif
 
-$(obj)/%.o: $(src)/%.S $(objtool_dep) FORCE
+$(obj)/%.o: $(src)/%.S $$(objtool_dep) FORCE
 	$(call if_changed_rule,as_o_S)
 
 targets += $(filter-out $(subdir-builtin), $(real-obj-y))
Signature · 1 line
2.30.2
author maintainer
#263 lore raw

PATCH 263/700 pata_octeon_cf: avoid WARN_ON() in ata_host_activate()

In reply to Greg Kroah-Hartman

From: Sergey Shtylyov <[email protected]>

[ Upstream commit bfc1f378c8953e68ccdbfe0a8c20748427488b80 ]

Iff platform_get_irq() fails (or returns IRQ0) and thus the polling mode has to be used, ata_host_activate() hits the WARN_ON() due to 'irq_handler' parameter being non-NULL if the polling mode is selected. Let's only set the pointer to the driver's IRQ handler if platform_get_irq() returns a valid IRQ # -- this should avoid the unnecessary WARN_ON()...

Fixes 43f01da0f279 ("MIPS/OCTEON/ata: Convert pata_octeon_cf.c to use device tree.")
---
 drivers/ata/pata_octeon_cf.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)
drivers/ata/pata_octeon_cf.c +3 -2
index bd87476ab481..b5a3f710d76d 100644
@@ -898,10 +898,11 @@ static int octeon_cf_probe(struct platform_device *pdev)
 					return -EINVAL;
 				}
 
-				irq_handler = octeon_cf_interrupt;
 				i = platform_get_irq(dma_dev, 0);
-				if (i > 0)
+				if (i > 0) {
 					irq = i;
+					irq_handler = octeon_cf_interrupt;
+				}
 			}
 			of_node_put(dma_node);
 		}
Signature · 1 line
2.30.2
author maintainer
#264 lore raw

PATCH 264/700 evm: fix writing <securityfs>/evm overflow

In reply to Greg Kroah-Hartman

From: Mimi Zohar <[email protected]>

[ Upstream commit 49219d9b8785ba712575c40e48ce0f7461254626 ]

EVM_SETUP_COMPLETE is defined as 0x80000000, which is larger than INT_MAX. The "-fno-strict-overflow" compiler option properly prevents signaling EVM that the EVM policy setup is complete. Define and read an unsigned int.

Fixes f00d79750712 ("EVM: Allow userspace to signal an RSA key has been loaded")
---
 security/integrity/evm/evm_secfs.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)
security/integrity/evm/evm_secfs.c +3 -2
index c175e2b659e4..5f0da41bccd0 100644
@@ -66,12 +66,13 @@ static ssize_t evm_read_key(struct file *filp, char __user *buf,
 static ssize_t evm_write_key(struct file *file, const char __user *buf,
 			     size_t count, loff_t *ppos)
 {
-	int i, ret;
+	unsigned int i;
+	int ret;
 
 	if (!capable(CAP_SYS_ADMIN) || (evm_initialized & EVM_SETUP_COMPLETE))
 		return -EPERM;
 
-	ret = kstrtoint_from_user(buf, count, 0, &i);
+	ret = kstrtouint_from_user(buf, count, 0, &i);
 
 	if (ret)
 		return ret;
Signature · 1 line
2.30.2
author maintainer
#265 lore raw

PATCH 265/700 x86/elf: Use _BITUL() macro in UAPI headers

In reply to Greg Kroah-Hartman

From: Joe Richey <[email protected]>

[ Upstream commit d06aca989c243dd9e5d3e20aa4e5c2ecfdd07050 ]

Replace BIT() in x86's UAPI header with _BITUL(). BIT() is not defined in the UAPI headers and its usage may cause userspace build errors.

Fixes 742c45c3ecc9 ("x86/elf: Enumerate kernel FSGSBASE capability in AT_HWCAP2")
---
 arch/x86/include/uapi/asm/hwcap2.h | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
arch/x86/include/uapi/asm/hwcap2.h +4 -2
index 5fdfcb47000f..054604aba9f0 100644
@@ -2,10 +2,12 @@
 #ifndef _ASM_X86_HWCAP2_H
 #define _ASM_X86_HWCAP2_H
 
+#include <linux/const.h>
+
 /* MONITOR/MWAIT enabled in Ring 3 */
-#define HWCAP2_RING3MWAIT		(1 << 0)
+#define HWCAP2_RING3MWAIT		_BITUL(0)
 
 /* Kernel allows FSGSBASE instructions available in Ring 3 */
-#define HWCAP2_FSGSBASE			BIT(1)
+#define HWCAP2_FSGSBASE			_BITUL(1)
 
 #endif
Signature · 1 line
2.30.2
author maintainer
#266 lore raw

PATCH 266/700 crypto: sa2ul - Fix leaks on failure paths with sa_dma_init()

In reply to Greg Kroah-Hartman

From: Suman Anna <[email protected]>

[ Upstream commit 4c0716ee1d973f6504d13f0e8d4d10350c85ad37 ]

The sa_dma_init() function doesn't release the requested dma channels on all failure paths. Any failure in this function also ends up leaking the dma pool created in sa_init_mem() in the sa_ul_probe() function. Fix all of these issues.

Fixes 7694b6ca649f ("crypto: sa2ul - Add crypto driver")
---
 drivers/crypto/sa2ul.c | 21 ++++++++++++---------
 1 file changed, 12 insertions(+), 9 deletions(-)
drivers/crypto/sa2ul.c +12 -9
index b0f0502a5bb0..efde3e96f62d 100644
@@ -2275,9 +2275,9 @@ static int sa_dma_init(struct sa_crypto_data *dd)
 
 	dd->dma_rx2 = dma_request_chan(dd->dev, "rx2");
 	if (IS_ERR(dd->dma_rx2)) {
-		dma_release_channel(dd->dma_rx1);
-		return dev_err_probe(dd->dev, PTR_ERR(dd->dma_rx2),
-				     "Unable to request rx2 DMA channel\n");
+		ret = dev_err_probe(dd->dev, PTR_ERR(dd->dma_rx2),
+				    "Unable to request rx2 DMA channel\n");
+		goto err_dma_rx2;
 	}
 
 	dd->dma_tx = dma_request_chan(dd->dev, "tx");
@@ -2298,28 +2298,31 @@ static int sa_dma_init(struct sa_crypto_data *dd)
 	if (ret) {
 		dev_err(dd->dev, "can't configure IN dmaengine slave: %d\n",
 			ret);
-		return ret;
+		goto err_dma_config;
 	}
 
 	ret = dmaengine_slave_config(dd->dma_rx2, &cfg);
 	if (ret) {
 		dev_err(dd->dev, "can't configure IN dmaengine slave: %d\n",
 			ret);
-		return ret;
+		goto err_dma_config;
 	}
 
 	ret = dmaengine_slave_config(dd->dma_tx, &cfg);
 	if (ret) {
 		dev_err(dd->dev, "can't configure OUT dmaengine slave: %d\n",
 			ret);
-		return ret;
+		goto err_dma_config;
 	}
 
 	return 0;
 
+err_dma_config:
+	dma_release_channel(dd->dma_tx);
 err_dma_tx:
-	dma_release_channel(dd->dma_rx1);
 	dma_release_channel(dd->dma_rx2);
+err_dma_rx2:
+	dma_release_channel(dd->dma_rx1);
 
 	return ret;
 }
@@ -2364,7 +2367,7 @@ static int sa_ul_probe(struct platform_device *pdev)
 	sa_init_mem(dev_data);
 	ret = sa_dma_init(dev_data);
 	if (ret)
-		goto disable_pm_runtime;
+		goto destroy_dma_pool;
 
 	spin_lock_init(&dev_data->scid_lock);
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
@@ -2394,9 +2397,9 @@ release_dma:
 	dma_release_channel(dev_data->dma_rx1);
 	dma_release_channel(dev_data->dma_tx);
 
+destroy_dma_pool:
 	dma_pool_destroy(dev_data->sc_pool);
 
-disable_pm_runtime:
 	pm_runtime_put_sync(&pdev->dev);
 	pm_runtime_disable(&pdev->dev);
 
Signature · 1 line
2.30.2
author maintainer
#267 lore raw

PATCH 267/700 crypto: sa2ul - Fix pm_runtime enable in sa_ul_probe()

In reply to Greg Kroah-Hartman

From: Suman Anna <[email protected]>

[ Upstream commit 5c8552325e013cbdabc443cd1f1b4d03c4a2e64e ]

The pm_runtime APIs added first in commit 7694b6ca649f ("crypto: sa2ul - Add crypto driver") are not unwound properly and was fixed up partially in commit 13343badae09 ("crypto: sa2ul - Fix PM reference leak in sa_ul_probe()"). This fixed up the pm_runtime usage count but not the state. Fix this properly.

Fixes 13343badae09 ("crypto: sa2ul - Fix PM reference leak in sa_ul_probe()")
---
 drivers/crypto/sa2ul.c | 1 +
 1 file changed, 1 insertion(+)
drivers/crypto/sa2ul.c +1 -0
index efde3e96f62d..ba116670ef8c 100644
@@ -2361,6 +2361,7 @@ static int sa_ul_probe(struct platform_device *pdev)
 	if (ret < 0) {
 		dev_err(&pdev->dev, "%s: failed to get sync: %d\n", __func__,
 			ret);
+		pm_runtime_disable(dev);
 		return ret;
 	}
 
Signature · 1 line
2.30.2
author maintainer
#268 lore raw

PATCH 268/700 crypto: ccp - Fix a resource leak in an error handling path

In reply to Greg Kroah-Hartman

From: Christophe JAILLET <[email protected]>

[ Upstream commit a6f8e68e238a15bb15f1726b35c695136c64eaba ]

If an error occurs after calling 'sp_get_irqs()', 'sp_free_irqs()' must be called as already done in the error handling path.

Fixes f4d18d656f88 ("crypto: ccp - Abstract interrupt registeration")
---
 drivers/crypto/ccp/sp-pci.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/crypto/ccp/sp-pci.c +4 -2
index f471dbaef1fb..7d346d842a39 100644
@@ -222,7 +222,7 @@ static int sp_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 		if (ret) {
 			dev_err(dev, "dma_set_mask_and_coherent failed (%d)\n",
 				ret);
-			goto e_err;
+			goto free_irqs;
 		}
 	}
 
@@ -230,10 +230,12 @@ static int sp_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 
 	ret = sp_init(sp);
 	if (ret)
-		goto e_err;
+		goto free_irqs;
 
 	return 0;
 
+free_irqs:
+	sp_free_irqs(sp);
 e_err:
 	dev_notice(dev, "initialization failed\n");
 	return ret;
Signature · 1 line
2.30.2
author maintainer
#269 lore raw

PATCH 269/700 media: rc: i2c: Fix an error message

In reply to Greg Kroah-Hartman

From: Christophe JAILLET <[email protected]>

[ Upstream commit 9c87ae1a0dbeb5794957421157fd266d38a869b4 ]

'ret' is known to be 1 here. In fact 'i' is expected instead. Store the return value of 'i2c_master_recv()' in 'ret' so that the error message print the correct error code.

Fixes acaa34bf06e9 ("media: rc: implement zilog transmitter")
---
 drivers/media/i2c/ir-kbd-i2c.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/media/i2c/ir-kbd-i2c.c +2 -2
index e8119ad0bc71..92376592455e 100644
@@ -678,8 +678,8 @@ static int zilog_tx(struct rc_dev *rcdev, unsigned int *txbuf,
 		goto out_unlock;
 	}
 
-	i = i2c_master_recv(ir->tx_c, buf, 1);
-	if (i != 1) {
+	ret = i2c_master_recv(ir->tx_c, buf, 1);
+	if (ret != 1) {
 		dev_err(&ir->rc->dev, "i2c_master_recv failed with %d\n", ret);
 		ret = -EIO;
 		goto out_unlock;
Signature · 1 line
2.30.2
author maintainer
#270 lore raw

PATCH 270/700 pata_ep93xx: fix deferred probing

In reply to Greg Kroah-Hartman

From: Sergey Shtylyov <[email protected]>

[ Upstream commit 5c8121262484d99bffb598f39a0df445cecd8efb ]

The driver overrides the error codes returned by platform_get_irq() to -ENXIO, so if it returns -EPROBE_DEFER, the driver would fail the probe permanently instead of the deferred probing. Propagate the error code upstream, as it should have been done from the start...

Fixes 2fff27512600 ("PATA host controller driver for ep93xx")
---
 drivers/ata/pata_ep93xx.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/ata/pata_ep93xx.c +1 -1
index badab6708893..46208ececbb6 100644
@@ -928,7 +928,7 @@ static int ep93xx_pata_probe(struct platform_device *pdev)
 	/* INT[3] (IRQ_EP93XX_EXT3) line connected as pull down */
 	irq = platform_get_irq(pdev, 0);
 	if (irq < 0) {
-		err = -ENXIO;
+		err = irq;
 		goto err_rel_gpio;
 	}
 
Signature · 1 line
2.30.2
author maintainer
#271 lore raw

PATCH 271/700 locking/lockdep: Reduce LOCKDEP dependency list

In reply to Greg Kroah-Hartman

From: Randy Dunlap <[email protected]>

[ Upstream commit b8e00abe7d9fe21dd13609e2e3a707e38902b105 ]

Some arches (um, sparc64, riscv, xtensa) cause a Kconfig warning for LOCKDEP.
These arch-es select LOCKDEP_SUPPORT but they are not listed as one of the arch-es that LOCKDEP depends on.

Since (16) arch-es define the Kconfig symbol LOCKDEP_SUPPORT if they intend to have LOCKDEP support, replace the awkward list of arch-es that LOCKDEP depends on with the LOCKDEP_SUPPORT symbol.

But wait. LOCKDEP_SUPPORT is included in LOCK_DEBUGGING_SUPPORT, which is already a dependency here, so LOCKDEP_SUPPORT is redundant and not needed.
That leaves the FRAME_POINTER dependency, but it is part of an expression like this:

	depends on (A && B) && (FRAME_POINTER || B')

where B' is a dependency of B so if B is true then B' is true and the value of FRAME_POINTER does not matter. Thus we can also delete the FRAME_POINTER dependency.

Fixes this kconfig warning: (for um, sparc64, riscv, xtensa)

WARNING: unmet direct dependencies detected for LOCKDEP
Depends on [n]: DEBUG_KERNEL [=y] && LOCK_DEBUGGING_SUPPORT [=y] && (FRAME_POINTER [=n] || MIPS || PPC || S390 || MICROBLAZE || ARM || ARC || X86) Selected by [y]:
- PROVE_LOCKING [=y] && DEBUG_KERNEL [=y] && LOCK_DEBUGGING_SUPPORT [=y]
- LOCK_STAT [=y] && DEBUG_KERNEL [=y] && LOCK_DEBUGGING_SUPPORT [=y]
- DEBUG_LOCK_ALLOC [=y] && DEBUG_KERNEL [=y] && LOCK_DEBUGGING_SUPPORT [=y]

Fixes 7d37cb2c912d ("lib: fix kconfig dependency on ARCH_WANT_FRAME_POINTERS")
---
 lib/Kconfig.debug | 1 -
 1 file changed, 1 deletion(-)
lib/Kconfig.debug +0 -1
index 417c3d3e521b..5c9f528dd46d 100644
@@ -1363,7 +1363,6 @@ config LOCKDEP
 	bool
 	depends on DEBUG_KERNEL && LOCK_DEBUGGING_SUPPORT
 	select STACKTRACE
-	depends on FRAME_POINTER || MIPS || PPC || S390 || MICROBLAZE || ARM || ARC || X86
 	select KALLSYMS
 	select KALLSYMS_ALL
 
Signature · 1 line
2.30.2
author maintainer
#272 lore raw

PATCH 272/700 sched: Dont defer CPU pick to migration_cpu_stop()

In reply to Greg Kroah-Hartman

From: Valentin Schneider <[email protected]>

[ Upstream commit 475ea6c60279e9f2ddf7e4cf2648cd8ae0608361 ]

Will reported that the 'XXX __migrate_task() can fail' in migration_cpu_stop() can happen, and it *is* sort of a big deal. Looking at it some more, one will note there is a glaring hole in the deferred CPU selection:

(w/ CONFIG_CPUSET=n, so that the affinity mask passed via taskset doesn't get AND'd with cpu_online_mask)

  $ taskset -pc 0-2 $PID
  # offline CPUs 3-4
  $ taskset -pc 3-5 $PID
    `\
      $PID may stay on 0-2 due to the cpumask_any_distribute() picking an
      offline CPU and __migrate_task() refusing to do anything due to
      cpu_is_allowed().

set_cpus_allowed_ptr() goes to some length to pick a dest_cpu that matches the right constraints vs affinity and the online/active state of the CPUs. Reuse that instead of discarding it in the affine_move_task() case.

Fixes 6d337eab041d ("sched: Fix migrate_disable() vs set_cpus_allowed_ptr()")
---
 kernel/sched/core.c | 20 ++++++++++++--------
 1 file changed, 12 insertions(+), 8 deletions(-)
kernel/sched/core.c +12 -8
index f59166fe499a..fe5da692dd7a 100644
@@ -1919,7 +1919,6 @@ static int migration_cpu_stop(void *data)
 	struct migration_arg *arg = data;
 	struct set_affinity_pending *pending = arg->pending;
 	struct task_struct *p = arg->task;
-	int dest_cpu = arg->dest_cpu;
 	struct rq *rq = this_rq();
 	bool complete = false;
 	struct rq_flags rf;
@@ -1952,19 +1951,15 @@ static int migration_cpu_stop(void *data)
 			if (p->migration_pending == pending)
 				p->migration_pending = NULL;
 			complete = true;
-		}
 
-		if (dest_cpu < 0) {
 			if (cpumask_test_cpu(task_cpu(p), &p->cpus_mask))
 				goto out;
-
-			dest_cpu = cpumask_any_distribute(&p->cpus_mask);
 		}
 
 		if (task_on_rq_queued(p))
-			rq = __migrate_task(rq, &rf, p, dest_cpu);
+			rq = __migrate_task(rq, &rf, p, arg->dest_cpu);
 		else
-			p->wake_cpu = dest_cpu;
+			p->wake_cpu = arg->dest_cpu;
 
 		/*
 		 * XXX __migrate_task() can fail, at which point we might end
@@ -2243,7 +2238,7 @@ static int affine_move_task(struct rq *rq, struct task_struct *p, struct rq_flag
 			init_completion(&my_pending.done);
 			my_pending.arg = (struct migration_arg) {
 				.task = p,
-				.dest_cpu = -1,		/* any */
+				.dest_cpu = dest_cpu,
 				.pending = &my_pending,
 			};
 
@@ -2251,6 +2246,15 @@ static int affine_move_task(struct rq *rq, struct task_struct *p, struct rq_flag
 		} else {
 			pending = p->migration_pending;
 			refcount_inc(&pending->refs);
+			/*
+			 * Affinity has changed, but we've already installed a
+			 * pending. migration_cpu_stop() *must* see this, else
+			 * we risk a completion of the pending despite having a
+			 * task on a disallowed CPU.
+			 *
+			 * Serialized by p->pi_lock, so this is safe.
+			 */
+			pending->arg.dest_cpu = dest_cpu;
 		}
 	}
 	pending = p->migration_pending;
Signature · 1 line
2.30.2
author maintainer
#273 lore raw

PATCH 273/700 media: ipu3-cio2: Fix reference counting when looping over ACPI devices

In reply to Greg Kroah-Hartman

From: Andy Shevchenko <[email protected]>

[ Upstream commit 2cb2705cf7ffe41dc5bd81290e4241bfb7f031cc ]

When we continue, due to device is disabled, loop we have to drop reference count. When we have an array full of devices we have to also drop the reference count. Note, in this case the cio2_bridge_unregister_sensors() is called by the caller.

Fixes 803abec64ef9 ("media: ipu3-cio2: Add cio2-bridge to ipu3-cio2 driver")
---
 drivers/media/pci/intel/ipu3/cio2-bridge.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)
drivers/media/pci/intel/ipu3/cio2-bridge.c +5 -5
index c2199042d3db..85f8b587405e 100644
@@ -173,14 +173,15 @@ static int cio2_bridge_connect_sensor(const struct cio2_sensor_config *cfg,
 	int ret;
 
 	for_each_acpi_dev_match(adev, cfg->hid, NULL, -1) {
-		if (!adev->status.enabled)
+		if (!adev->status.enabled) {
+			acpi_dev_put(adev);
 			continue;
+		}
 
 		if (bridge->n_sensors >= CIO2_NUM_PORTS) {
+			acpi_dev_put(adev);
 			dev_err(&cio2->dev, "Exceeded available CIO2 ports\n");
-			cio2_bridge_unregister_sensors(bridge);
-			ret = -EINVAL;
-			goto err_out;
+			return -EINVAL;
 		}
 
 		sensor = &bridge->sensors[bridge->n_sensors];
@@ -228,7 +229,6 @@ err_free_swnodes:
 	software_node_unregister_nodes(sensor->swnodes);
 err_put_adev:
 	acpi_dev_put(sensor->adev);
-err_out:
 	return ret;
 }
 
Signature · 1 line
2.30.2
author maintainer
#274 lore raw

PATCH 274/700 media: rkvdec: Fix .buf_prepare

In reply to Greg Kroah-Hartman

From: Ezequiel Garcia <[email protected]>

[ Upstream commit ba1ed4ae760a81caf39f54232e089d95157a0dba ]

The driver should only set the payload on .buf_prepare if the buffer is CAPTURE type. If an OUTPUT buffer has a zero bytesused set by userspace then v4l2-core will set it to buffer length.

If we overwrite bytesused for OUTPUT buffers, too, then vb2_get_plane_payload() will return incorrect value which might be then written to hw registers by the driver in rkvdec-h264.c.

[Changed the comment and used V4L2_TYPE_IS_CAPTURE macro]

---
 drivers/staging/media/rkvdec/rkvdec.c | 10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)
drivers/staging/media/rkvdec/rkvdec.c +9 -1
index 8c17615f3a7a..7131156c1f2c 100644
@@ -481,7 +481,15 @@ static int rkvdec_buf_prepare(struct vb2_buffer *vb)
 		if (vb2_plane_size(vb, i) < sizeimage)
 			return -EINVAL;
 	}
-	vb2_set_plane_payload(vb, 0, f->fmt.pix_mp.plane_fmt[0].sizeimage);
+
+	/*
+	 * Buffer's bytesused must be written by driver for CAPTURE buffers.
+	 * (for OUTPUT buffers, if userspace passes 0 bytesused, v4l2-core sets
+	 * it to buffer length).
+	 */
+	if (V4L2_TYPE_IS_CAPTURE(vq->type))
+		vb2_set_plane_payload(vb, 0, f->fmt.pix_mp.plane_fmt[0].sizeimage);
+
 	return 0;
 }
 
Signature · 1 line
2.30.2
author maintainer
#275 lore raw

PATCH 275/700 media: exynos4-is: Fix a use after free in isp_video_release

In reply to Greg Kroah-Hartman

From: Lv Yunlong <[email protected]>

[ Upstream commit 01fe904c9afd26e79c1f73aa0ca2e3d785e5e319 ]

In isp_video_release, file->private_data is freed via _vb2_fop_release()->v4l2_fh_release(). But the freed file->private_data is still used in v4l2_fh_is_singular_file() ->v4l2_fh_is_singular(file->private_data), which is a use after free bug.

My patch uses a variable 'is_singular_file' to avoid the uaf.
v3: https://lore.kernel.org/patchwork/patch/1419058/

Fixes 34947b8aebe3f ("[media] exynos4-is: Add the FIMC-IS ISP capture DMA driver")
---
 drivers/media/platform/exynos4-is/fimc-isp-video.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)
drivers/media/platform/exynos4-is/fimc-isp-video.c +5 -2
index 8d9dc597deaa..83688a7982f7 100644
@@ -305,17 +305,20 @@ static int isp_video_release(struct file *file)
 	struct fimc_is_video *ivc = &isp->video_capture;
 	struct media_entity *entity = &ivc->ve.vdev.entity;
 	struct media_device *mdev = entity->graph_obj.mdev;
+	bool is_singular_file;
 
 	mutex_lock(&isp->video_lock);
 
-	if (v4l2_fh_is_singular_file(file) && ivc->streaming) {
+	is_singular_file = v4l2_fh_is_singular_file(file);
+
+	if (is_singular_file && ivc->streaming) {
 		media_pipeline_stop(entity);
 		ivc->streaming = 0;
 	}
 
 	_vb2_fop_release(file, NULL);
 
-	if (v4l2_fh_is_singular_file(file)) {
+	if (is_singular_file) {
 		fimc_pipeline_call(&ivc->ve, close);
 
 		mutex_lock(&mdev->graph_mutex);
Signature · 1 line
2.30.2
author maintainer
#276 lore raw

PATCH 276/700 media: au0828: fix a NULL vs IS_ERR() check

In reply to Greg Kroah-Hartman

From: Dan Carpenter <[email protected]>

[ Upstream commit 8f2e452730d2bcd59fe05246f0e19a4c52e0012d ]

The media_device_usb_allocate() function returns error pointers when it's enabled and something goes wrong. It can return NULL as well, but only if CONFIG_MEDIA_CONTROLLER is disabled so that doesn't apply here.

Fixes 812658d88d26 ("media: change au0828 to use Media Device Allocator API")
---
 drivers/media/usb/au0828/au0828-core.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/media/usb/au0828/au0828-core.c +2 -2
index a8a72d5fbd12..caefac07af92 100644
@@ -199,8 +199,8 @@ static int au0828_media_device_init(struct au0828_dev *dev,
 	struct media_device *mdev;
 
 	mdev = media_device_usb_allocate(udev, KBUILD_MODNAME, THIS_MODULE);
-	if (!mdev)
-		return -ENOMEM;
+	if (IS_ERR(mdev))
+		return PTR_ERR(mdev);
 
 	dev->media_dev = mdev;
 #endif
Signature · 1 line
2.30.2
author maintainer
#277 lore raw

PATCH 277/700 media: tc358743: Fix error return code in tc358743_probe_of()

In reply to Greg Kroah-Hartman

From: Zhen Lei <[email protected]>

[ Upstream commit a6b1e7093f0a099571fc8836ab4a589633f956a8 ]

When the CSI bps per lane is not in the valid range, an appropriate error code -EINVAL should be returned. However, we currently do not explicitly assign this error code to 'ret'. As a result, 0 was incorrectly returned.

Fixes 256148246852 ("[media] tc358743: support probe from device tree")
---
 drivers/media/i2c/tc358743.c | 1 +
 1 file changed, 1 insertion(+)
drivers/media/i2c/tc358743.c +1 -0
index 1b309bb743c7..f21da11caf22 100644
@@ -1974,6 +1974,7 @@ static int tc358743_probe_of(struct tc358743_state *state)
 	bps_pr_lane = 2 * endpoint.link_frequencies[0];
 	if (bps_pr_lane < 62500000U || bps_pr_lane > 1000000000U) {
 		dev_err(dev, "unsupported bps per lane: %u bps\n", bps_pr_lane);
+		ret = -EINVAL;
 		goto disable_clk;
 	}
 
Signature · 1 line
2.30.2
author maintainer
#278 lore raw

PATCH 278/700 media: vicodec: Use _BITUL() macro in UAPI headers

In reply to Greg Kroah-Hartman

From: Joe Richey <[email protected]>

[ Upstream commit ce67eaca95f8ab5c6aae41a10adfe9a6e8efa58c ]

Replace BIT() in v4l2's UPAI header with _BITUL(). BIT() is not defined in the UAPI headers and its usage may cause userspace build errors.

Fixes 206bc0f6fb94 ("media: vicodec: mark the stateless FWHT API as stable")
---
 include/uapi/linux/v4l2-controls.h | 23 ++++++++++++-----------
 1 file changed, 12 insertions(+), 11 deletions(-)
include/uapi/linux/v4l2-controls.h +12 -11
index 039c0d7add1b..c7fe032df185 100644
@@ -50,6 +50,7 @@
 #ifndef __LINUX_V4L2_CONTROLS_H
 #define __LINUX_V4L2_CONTROLS_H
 
+#include <linux/const.h>
 #include <linux/types.h>
 
 /* Control classes */
@@ -1593,30 +1594,30 @@ struct v4l2_ctrl_h264_decode_params {
 #define V4L2_FWHT_VERSION			3
 
 /* Set if this is an interlaced format */
-#define V4L2_FWHT_FL_IS_INTERLACED		BIT(0)
+#define V4L2_FWHT_FL_IS_INTERLACED		_BITUL(0)
 /* Set if this is a bottom-first (NTSC) interlaced format */
-#define V4L2_FWHT_FL_IS_BOTTOM_FIRST		BIT(1)
+#define V4L2_FWHT_FL_IS_BOTTOM_FIRST		_BITUL(1)
 /* Set if each 'frame' contains just one field */
-#define V4L2_FWHT_FL_IS_ALTERNATE		BIT(2)
+#define V4L2_FWHT_FL_IS_ALTERNATE		_BITUL(2)
 /*
  * If V4L2_FWHT_FL_IS_ALTERNATE was set, then this is set if this
  * 'frame' is the bottom field, else it is the top field.
  */
-#define V4L2_FWHT_FL_IS_BOTTOM_FIELD		BIT(3)
+#define V4L2_FWHT_FL_IS_BOTTOM_FIELD		_BITUL(3)
 /* Set if the Y' plane is uncompressed */
-#define V4L2_FWHT_FL_LUMA_IS_UNCOMPRESSED	BIT(4)
+#define V4L2_FWHT_FL_LUMA_IS_UNCOMPRESSED	_BITUL(4)
 /* Set if the Cb plane is uncompressed */
-#define V4L2_FWHT_FL_CB_IS_UNCOMPRESSED		BIT(5)
+#define V4L2_FWHT_FL_CB_IS_UNCOMPRESSED		_BITUL(5)
 /* Set if the Cr plane is uncompressed */
-#define V4L2_FWHT_FL_CR_IS_UNCOMPRESSED		BIT(6)
+#define V4L2_FWHT_FL_CR_IS_UNCOMPRESSED		_BITUL(6)
 /* Set if the chroma plane is full height, if cleared it is half height */
-#define V4L2_FWHT_FL_CHROMA_FULL_HEIGHT		BIT(7)
+#define V4L2_FWHT_FL_CHROMA_FULL_HEIGHT		_BITUL(7)
 /* Set if the chroma plane is full width, if cleared it is half width */
-#define V4L2_FWHT_FL_CHROMA_FULL_WIDTH		BIT(8)
+#define V4L2_FWHT_FL_CHROMA_FULL_WIDTH		_BITUL(8)
 /* Set if the alpha plane is uncompressed */
-#define V4L2_FWHT_FL_ALPHA_IS_UNCOMPRESSED	BIT(9)
+#define V4L2_FWHT_FL_ALPHA_IS_UNCOMPRESSED	_BITUL(9)
 /* Set if this is an I Frame */
-#define V4L2_FWHT_FL_I_FRAME			BIT(10)
+#define V4L2_FWHT_FL_I_FRAME			_BITUL(10)
 
 /* A 4-values flag - the number of components - 1 */
 #define V4L2_FWHT_FL_COMPONENTS_NUM_MSK		GENMASK(18, 16)
Signature · 1 line
2.30.2
author maintainer
#279 lore raw

PATCH 279/700 media: gspca/gl860: fix zero-length control requests

In reply to Greg Kroah-Hartman

From: Johan Hovold <[email protected]>

[ Upstream commit 8ed339f23d41e21660a389adf2e7b2966d457ff6 ]

The direction of the pipe argument must match the request-type direction bit or control requests may fail depending on the host-controller-driver implementation.

Control transfers without a data stage are treated as OUT requests by the USB stack and should be using usb_sndctrlpipe(). Failing to do so will now trigger a warning.

Fix the gl860_RTx() helper so that zero-length control reads fail with an error message instead. Note that there are no current callers that would trigger this.

Fixes 4f7cb8837cec ("V4L/DVB (12954): gspca - gl860: Addition of GL860 based webcams")
---
 drivers/media/usb/gspca/gl860/gl860.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/media/usb/gspca/gl860/gl860.c +2 -2
index 2c05ea2598e7..ce4ee8bc75c8 100644
@@ -561,8 +561,8 @@ int gl860_RTx(struct gspca_dev *gspca_dev,
 					len, 400 + 200 * (len > 1));
 			memcpy(pdata, gspca_dev->usb_buf, len);
 		} else {
-			r = usb_control_msg(udev, usb_rcvctrlpipe(udev, 0),
-					req, pref, val, index, NULL, len, 400);
+			gspca_err(gspca_dev, "zero-length read request\n");
+			r = -EINVAL;
 		}
 	}
 
Signature · 1 line
2.30.2
author maintainer
#280 lore raw

PATCH 280/700 m68k: atari: Fix ATARI_KBD_CORE kconfig unmet dependency warning

In reply to Greg Kroah-Hartman

From: Randy Dunlap <[email protected]>

[ Upstream commit c1367ee016e3550745315fb9a2dd1e4ce02cdcf6 ]

Since the code for ATARI_KBD_CORE does not use drivers/input/keyboard/ code, just move ATARI_KBD_CORE to arch/m68k/Kconfig.machine to remove the dependency on INPUT_KEYBOARD.

Removes this kconfig warning:

WARNING: unmet direct dependencies detected for ATARI_KBD_CORE
Depends on [n]: !UML && INPUT [=y] && INPUT_KEYBOARD [=n] Selected by [y]:
- MOUSE_ATARI [=y] && !UML && INPUT [=y] && INPUT_MOUSE [=y] && ATARI [=y]

---
 arch/m68k/Kconfig.machine      | 3 +++
 drivers/input/keyboard/Kconfig | 3 ---
 2 files changed, 3 insertions(+), 3 deletions(-)
arch/m68k/Kconfig.machine +3 -0
index 4d59ec2f5b8d..d964c1f27399 100644
@@ -25,6 +25,9 @@ config ATARI
 	  this kernel on an Atari, say Y here and browse the material
 	  available in <file:Documentation/m68k>; otherwise say N.
 
+config ATARI_KBD_CORE
+	bool
+
 config MAC
 	bool "Macintosh support"
 	depends on MMU
drivers/input/keyboard/Kconfig +0 -3
index 32d15809ae58..40a070a2e7f5 100644
@@ -67,9 +67,6 @@ config KEYBOARD_AMIGA
 	  To compile this driver as a module, choose M here: the
 	  module will be called amikbd.
 
-config ATARI_KBD_CORE
-	bool
-
 config KEYBOARD_APPLESPI
 	tristate "Apple SPI keyboard and trackpad"
 	depends on ACPI && EFI
Signature · 1 line
2.30.2
author maintainer
#281 lore raw

PATCH 281/700 media: siano: Fix out-of-bounds warnings in smscore_load_firmware_family2()

In reply to Greg Kroah-Hartman

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

[ Upstream commit 13dfead49db07225335d4f587a560a2210391a1a ]

Rename struct sms_msg_data4 to sms_msg_data5 and increase the size of its msg_data array from 4 to 5 elements. Notice that at some point the 5th element of msg_data is being accessed in function smscore_load_firmware_family2():

1006 trigger_msg->msg_data[4] = 4; /* Task ID */

Also, there is no need for the object _trigger_msg_ of type struct sms_msg_data *, when _msg_ can be used, directly. Notice that msg_data in struct sms_msg_data is a one-element array, which causes multiple out-of-bounds warnings when accessing beyond its first element in function smscore_load_firmware_family2():

992 struct sms_msg_data *trigger_msg =
993 (struct sms_msg_data *) msg;
994
995 pr_debug("sending MSG_SMS_SWDOWNLOAD_TRIGGER_REQ\n");
996 SMS_INIT_MSG(&msg->x_msg_header,

 997                                 MSG_SMS_SWDOWNLOAD_TRIGGER_REQ,
 998                                 sizeof(struct sms_msg_hdr) +
 999                                 sizeof(u32) * 5);
1000
1001                 trigger_msg->msg_data[0] = firmware->start_address;
1002                                         /* Entry point */
1003                 trigger_msg->msg_data[1] = 6; /* Priority */
1004                 trigger_msg->msg_data[2] = 0x200; /* Stack size */
1005                 trigger_msg->msg_data[3] = 0; /* Parameter */
1006                 trigger_msg->msg_data[4] = 4; /* Task ID */

even when enough dynamic memory is allocated for _msg_:

929 /* PAGE_SIZE buffer shall be enough and dma aligned */
930 msg = kmalloc(PAGE_SIZE, GFP_KERNEL | coredev->gfp_buf_flags);

but as _msg_ is casted to (struct sms_msg_data *):

992 struct sms_msg_data *trigger_msg =
993 (struct sms_msg_data *) msg;

the out-of-bounds warnings are actually valid and should be addressed.

Fix this by declaring object _msg_ of type struct sms_msg_data5 *, which contains a 5-elements array, instead of just 4. And use _msg_ directly, instead of creating object trigger_msg.

This helps with the ongoing efforts to enable -Warray-bounds by fixing the following warnings:

CC [M] drivers/media/common/siano/smscoreapi.o
drivers/media/common/siano/smscoreapi.c: In function ‘smscore_load_firmware_family2’:
drivers/media/common/siano/smscoreapi.c:1003:24: warning: array subscript 1 is above array bounds of ‘u32[1]’ {aka ‘unsigned int[1]’} [-Warray-bounds]

 1003 |   trigger_msg->msg_data[1] = 6; /* Priority */
      |   ~~~~~~~~~~~~~~~~~~~~~^~~
In file included from drivers/media/common/siano/smscoreapi.c:12:
drivers/media/common/siano/smscoreapi.h:619:6: note: while referencing ‘msg_data’
  619 |  u32 msg_data[1];
      |      ^~~~~~~~
drivers/media/common/siano/smscoreapi.c:1004:24: warning: array subscript 2 is above array bounds of ‘u32[1]’ {aka ‘unsigned int[1]’} [-Warray-bounds]
 1004 |   trigger_msg->msg_data[2] = 0x200; /* Stack size */
      |   ~~~~~~~~~~~~~~~~~~~~~^~~
In file included from drivers/media/common/siano/smscoreapi.c:12:
drivers/media/common/siano/smscoreapi.h:619:6: note: while referencing ‘msg_data’
  619 |  u32 msg_data[1];
      |      ^~~~~~~~
drivers/media/common/siano/smscoreapi.c:1005:24: warning: array subscript 3 is above array bounds of ‘u32[1]’ {aka ‘unsigned int[1]’} [-Warray-bounds]
 1005 |   trigger_msg->msg_data[3] = 0; /* Parameter */
      |   ~~~~~~~~~~~~~~~~~~~~~^~~
In file included from drivers/media/common/siano/smscoreapi.c:12:
drivers/media/common/siano/smscoreapi.h:619:6: note: while referencing ‘msg_data’
  619 |  u32 msg_data[1];
      |      ^~~~~~~~
drivers/media/common/siano/smscoreapi.c:1006:24: warning: array subscript 4 is above array bounds of ‘u32[1]’ {aka ‘unsigned int[1]’} [-Warray-bounds]
 1006 |   trigger_msg->msg_data[4] = 4; /* Task ID */
      |   ~~~~~~~~~~~~~~~~~~~~~^~~
In file included from drivers/media/common/siano/smscoreapi.c:12:
drivers/media/common/siano/smscoreapi.h:619:6: note: while referencing ‘msg_data’
  619 |  u32 msg_data[1];
      |      ^~~~~~~~
Fixes 018b0c6f8acb ("[media] siano: make load firmware logic to work with newer firmwares")
---
 drivers/media/common/siano/smscoreapi.c | 22 +++++++++-------------
 drivers/media/common/siano/smscoreapi.h |  4 ++--
 2 files changed, 11 insertions(+), 15 deletions(-)
drivers/media/common/siano/smscoreapi.c +9 -13
index c1511094fdc7..b735e2370137 100644
@@ -908,7 +908,7 @@ static int smscore_load_firmware_family2(struct smscore_device_t *coredev,
 					 void *buffer, size_t size)
 {
 	struct sms_firmware *firmware = (struct sms_firmware *) buffer;
-	struct sms_msg_data4 *msg;
+	struct sms_msg_data5 *msg;
 	u32 mem_address,  calc_checksum = 0;
 	u32 i, *ptr;
 	u8 *payload = firmware->payload;
@@ -989,24 +989,20 @@ static int smscore_load_firmware_family2(struct smscore_device_t *coredev,
 		goto exit_fw_download;
 
 	if (coredev->mode == DEVICE_MODE_NONE) {
-		struct sms_msg_data *trigger_msg =
-			(struct sms_msg_data *) msg;
-
 		pr_debug("sending MSG_SMS_SWDOWNLOAD_TRIGGER_REQ\n");
 		SMS_INIT_MSG(&msg->x_msg_header,
 				MSG_SMS_SWDOWNLOAD_TRIGGER_REQ,
-				sizeof(struct sms_msg_hdr) +
-				sizeof(u32) * 5);
+				sizeof(*msg));
 
-		trigger_msg->msg_data[0] = firmware->start_address;
+		msg->msg_data[0] = firmware->start_address;
 					/* Entry point */
-		trigger_msg->msg_data[1] = 6; /* Priority */
-		trigger_msg->msg_data[2] = 0x200; /* Stack size */
-		trigger_msg->msg_data[3] = 0; /* Parameter */
-		trigger_msg->msg_data[4] = 4; /* Task ID */
+		msg->msg_data[1] = 6; /* Priority */
+		msg->msg_data[2] = 0x200; /* Stack size */
+		msg->msg_data[3] = 0; /* Parameter */
+		msg->msg_data[4] = 4; /* Task ID */
 
-		rc = smscore_sendrequest_and_wait(coredev, trigger_msg,
-					trigger_msg->x_msg_header.msg_length,
+		rc = smscore_sendrequest_and_wait(coredev, msg,
+					msg->x_msg_header.msg_length,
 					&coredev->trigger_done);
 	} else {
 		SMS_INIT_MSG(&msg->x_msg_header, MSG_SW_RELOAD_EXEC_REQ,
drivers/media/common/siano/smscoreapi.h +2 -2
index b3b793b5caf3..16c45afabc53 100644
@@ -629,9 +629,9 @@ struct sms_msg_data2 {
 	u32 msg_data[2];
 };
 
-struct sms_msg_data4 {
+struct sms_msg_data5 {
 	struct sms_msg_hdr x_msg_header;
-	u32 msg_data[4];
+	u32 msg_data[5];
 };
 
 struct sms_data_download {
Signature · 1 line
2.30.2
author maintainer
#282 lore raw

PATCH 282/700 regulator: fan53880: Fix vsel_mask setting for FAN53880_BUCK

In reply to Greg Kroah-Hartman

From: Axel Lin <[email protected]>

[ Upstream commit 2e11737a772b95c6587df73f216eec1762431432 ]

According to the datasheet:
REGISTER DETAILS − 0x02 BUCK, BUCK_OUT is BIT0 ~ BIT7.

So vsel_mask for FAN53880_BUCK should be 0xFF.

Fixes e6dea51e2d41 ("regulator: fan53880: Add initial support")
---
 drivers/regulator/fan53880.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/regulator/fan53880.c +1 -1
index 1684faf82ed2..94f02f3099dd 100644
@@ -79,7 +79,7 @@ static const struct regulator_desc fan53880_regulators[] = {
 		.n_linear_ranges = 2,
 		.n_voltages =	   0xf8,
 		.vsel_reg =	   FAN53880_BUCKVOUT,
-		.vsel_mask =	   0x7f,
+		.vsel_mask =	   0xff,
 		.enable_reg =	   FAN53880_ENABLE,
 		.enable_mask =	   0x10,
 		.enable_time =	   480,
Signature · 1 line
2.30.2
author maintainer
#283 lore raw

PATCH 283/700 crypto: nitrox - fix unchecked variable in nitrox_register_interrupts

In reply to Greg Kroah-Hartman

From: Tong Tiangen <[email protected]>

[ Upstream commit 57c126661f50b884d3812e7db6e00f2e778eccfb ]

Function nitrox_register_interrupts leaves variable 'nr_vecs' unchecked, which would be use as kcalloc parameter later.

Fixes 5155e118dda9 ("crypto: cavium/nitrox - use pci_alloc_irq_vectors() while enabling MSI-X.")
---
 drivers/crypto/cavium/nitrox/nitrox_isr.c | 4 ++++
 1 file changed, 4 insertions(+)
drivers/crypto/cavium/nitrox/nitrox_isr.c +4 -0
index 99b053094f5a..b16689b48f5a 100644
@@ -307,6 +307,10 @@ int nitrox_register_interrupts(struct nitrox_device *ndev)
 	 * Entry 192: NPS_CORE_INT_ACTIVE
 	 */
 	nr_vecs = pci_msix_vec_count(pdev);
+	if (nr_vecs < 0) {
+		dev_err(DEV(ndev), "Error in getting vec count %d\n", nr_vecs);
+		return nr_vecs;
+	}
 
 	/* Enable MSI-X */
 	ret = pci_alloc_irq_vectors(pdev, nr_vecs, nr_vecs, PCI_IRQ_MSIX);
Signature · 1 line
2.30.2
author maintainer
#284 lore raw

PATCH 284/700 crypto: omap-sham - Fix PM reference leak in omap sham ops

In reply to Greg Kroah-Hartman

From: Zhang Qilong <[email protected]>

[ Upstream commit ca323b2c61ec321eb9f2179a405b9c34cdb4f553 ]

pm_runtime_get_sync will increment pm usage counter even it failed. Forgetting to putting operation will result in reference leak here. We fix it by replacing it with pm_runtime_resume_and_get to keep usage counter balanced.

Fixes 604c31039dae4 ("crypto: omap-sham - Check for return value from pm_runtime_get_sync")
---
 drivers/crypto/omap-sham.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/crypto/omap-sham.c +2 -2
index ae0d320d3c60..dd53ad9987b0 100644
@@ -372,7 +372,7 @@ static int omap_sham_hw_init(struct omap_sham_dev *dd)
 {
 	int err;
 
-	err = pm_runtime_get_sync(dd->dev);
+	err = pm_runtime_resume_and_get(dd->dev);
 	if (err < 0) {
 		dev_err(dd->dev, "failed to get sync: %d\n", err);
 		return err;
@@ -2244,7 +2244,7 @@ static int omap_sham_suspend(struct device *dev)
 
 static int omap_sham_resume(struct device *dev)
 {
-	int err = pm_runtime_get_sync(dev);
+	int err = pm_runtime_resume_and_get(dev);
 	if (err < 0) {
 		dev_err(dev, "failed to get sync: %d\n", err);
 		return err;
Signature · 1 line
2.30.2
author maintainer
#285 lore raw

PATCH 285/700 crypto: x86/curve25519 - fix cpu feature checking logic in mod_exit

In reply to Greg Kroah-Hartman

From: Hangbin Liu <[email protected]>

[ Upstream commit 1b82435d17774f3eaab35dce239d354548aa9da2 ]

In curve25519_mod_init() the curve25519_alg will be registered only when (X86_FEATURE_BMI2 && X86_FEATURE_ADX). But in curve25519_mod_exit() it still checks (X86_FEATURE_BMI2 || X86_FEATURE_ADX) when do crypto unregister. This will trigger a BUG_ON in crypto_unregister_alg() as alg->cra_refcnt is 0 if the cpu only supports one of X86_FEATURE_BMI2 and X86_FEATURE_ADX.

Fixes 07b586fe0662 ("crypto: x86/curve25519 - replace with formally verified implementation")
---
 arch/x86/crypto/curve25519-x86_64.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/crypto/curve25519-x86_64.c +1 -1
index 5af8021b98ce..11b4c83c715e 100644
@@ -1500,7 +1500,7 @@ static int __init curve25519_mod_init(void)
 static void __exit curve25519_mod_exit(void)
 {
 	if (IS_REACHABLE(CONFIG_CRYPTO_KPP) &&
-	    (boot_cpu_has(X86_FEATURE_BMI2) || boot_cpu_has(X86_FEATURE_ADX)))
+	    static_branch_likely(&curve25519_use_bmi2_adx))
 		crypto_unregister_kpp(&curve25519_alg);
 }
 
Signature · 1 line
2.30.2
author maintainer
#286 lore raw

PATCH 286/700 crypto: sm2 - fix a memory leak in sm2

In reply to Greg Kroah-Hartman

From: Hongbo Li <[email protected]>

[ Upstream commit 5cd259ca5d466f65ffd21e2e2fa00fb648a8c555 ]

SM2 module alloc ec->Q in sm2_set_pub_key(), when doing alg test in test_akcipher_one(), it will set public key for every test vector, and don't free ec->Q. This will cause a memory leak.

This patch alloc ec->Q in sm2_ec_ctx_init().

Fixes ea7ecb66440b ("crypto: sm2 - introduce OSCCA SM2 asymmetric cipher algorithm")
---
 crypto/sm2.c | 24 ++++++++++--------------
 1 file changed, 10 insertions(+), 14 deletions(-)
crypto/sm2.c +10 -14
index b21addc3ac06..db8a4a265669 100644
@@ -79,10 +79,17 @@ static int sm2_ec_ctx_init(struct mpi_ec_ctx *ec)
 		goto free;
 
 	rc = -ENOMEM;
+
+	ec->Q = mpi_point_new(0);
+	if (!ec->Q)
+		goto free;
+
 	/* mpi_ec_setup_elliptic_curve */
 	ec->G = mpi_point_new(0);
-	if (!ec->G)
+	if (!ec->G) {
+		mpi_point_release(ec->Q);
 		goto free;
+	}
 
 	mpi_set(ec->G->x, x);
 	mpi_set(ec->G->y, y);
@@ -91,6 +98,7 @@ static int sm2_ec_ctx_init(struct mpi_ec_ctx *ec)
 	rc = -EINVAL;
 	ec->n = mpi_scanval(ecp->n);
 	if (!ec->n) {
+		mpi_point_release(ec->Q);
 		mpi_point_release(ec->G);
 		goto free;
 	}
@@ -386,27 +394,15 @@ static int sm2_set_pub_key(struct crypto_akcipher *tfm,
 	MPI a;
 	int rc;
 
-	ec->Q = mpi_point_new(0);
-	if (!ec->Q)
-		return -ENOMEM;
-
 	/* include the uncompressed flag '0x04' */
-	rc = -ENOMEM;
 	a = mpi_read_raw_data(key, keylen);
 	if (!a)
-		goto error;
+		return -ENOMEM;
 
 	mpi_normalize(a);
 	rc = sm2_ecc_os2ec(ec->Q, a);
 	mpi_free(a);
-	if (rc)
-		goto error;
-
-	return 0;
 
-error:
-	mpi_point_release(ec->Q);
-	ec->Q = NULL;
 	return rc;
 }
 
Signature · 1 line
2.30.2
author maintainer
#287 lore raw

PATCH 287/700 mmc: usdhi6rol0: fix error return code in usdhi6_probe()

In reply to Greg Kroah-Hartman

From: Zhen Lei <[email protected]>

[ Upstream commit 2f9ae69e5267f53e89e296fccee291975a85f0eb ]

Fix to return a negative error code from the error handling case instead of 0, as done elsewhere in this function.

Fixes 75fa9ea6e3c0 ("mmc: add a driver for the Renesas usdhi6rol0 SD/SDIO host controller")
---
 drivers/mmc/host/usdhi6rol0.c | 1 +
 1 file changed, 1 insertion(+)
drivers/mmc/host/usdhi6rol0.c +1 -0
index 615f3d008af1..b9b79b1089a0 100644
@@ -1801,6 +1801,7 @@ static int usdhi6_probe(struct platform_device *pdev)
 
 	version = usdhi6_read(host, USDHI6_VERSION);
 	if ((version & 0xfff) != 0xa0d) {
+		ret = -EPERM;
 		dev_err(dev, "Version not recognized %x\n", version);
 		goto e_clk_off;
 	}
Signature · 1 line
2.30.2
author maintainer
#288 lore raw

PATCH 288/700 arm64/mm: Fix ttbr0 values stored in struct thread_info for software-pan

In reply to Greg Kroah-Hartman

From: Anshuman Khandual <[email protected]>

[ Upstream commit 9163f01130304fab1f74683d7d44632da7bda637 ]

When using CONFIG_ARM64_SW_TTBR0_PAN, a task's thread_info::ttbr0 must be the TTBR0_EL1 value used to run userspace. With 52-bit PAs, the PA must be packed into the TTBR using phys_to_ttbr(), but we forget to do this in some of the SW PAN code. Thus, if the value is installed into TTBR0_EL1 (as may happen in the uaccess routines), this could result in UNPREDICTABLE behaviour.

Since hardware with 52-bit PA support almost certainly has HW PAN, which will be used in preference, this shouldn't be a practical issue, but let's fix this for consistency.

Cc Catalin Marinas <[email protected]>
Cc Will Deacon <[email protected]>
Cc Mark Rutland <[email protected]>
Cc James Morse <[email protected]>
Fixes 529c4b05a3cb ("arm64: handle 52-bit addresses in TTBR")
---
 arch/arm64/include/asm/mmu_context.h | 4 ++--
 arch/arm64/kernel/setup.c            | 2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)
arch/arm64/include/asm/mmu_context.h +2 -2
index bd02e99b1a4c..44dceac442fc 100644
@@ -177,9 +177,9 @@ static inline void update_saved_ttbr0(struct task_struct *tsk,
 		return;
 
 	if (mm == &init_mm)
-		ttbr = __pa_symbol(reserved_pg_dir);
+		ttbr = phys_to_ttbr(__pa_symbol(reserved_pg_dir));
 	else
-		ttbr = virt_to_phys(mm->pgd) | ASID(mm) << 48;
+		ttbr = phys_to_ttbr(virt_to_phys(mm->pgd)) | ASID(mm) << 48;
 
 	WRITE_ONCE(task_thread_info(tsk)->ttbr0, ttbr);
 }
arch/arm64/kernel/setup.c +1 -1
index 61845c0821d9..68b30e8c22db 100644
@@ -381,7 +381,7 @@ void __init __no_sanitize_address setup_arch(char **cmdline_p)
 	 * faults in case uaccess_enable() is inadvertently called by the init
 	 * thread.
 	 */
-	init_task.thread_info.ttbr0 = __pa_symbol(reserved_pg_dir);
+	init_task.thread_info.ttbr0 = phys_to_ttbr(__pa_symbol(reserved_pg_dir));
 #endif
 
 	if (boot_args[1] || boot_args[2] || boot_args[3]) {
Signature · 9 lines
2.30.2




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

PATCH 289/700 media: v4l2-core: ignore native time32 ioctls on 64-bit

In reply to Greg Kroah-Hartman

From: Arnd Bergmann <[email protected]>

[ Upstream commit c344f07aa1b4ba38ca8fabe407a2afe2f436323c ]

Syzbot found that passing ioctl command 0xc0505609 into a 64-bit kernel from a 32-bit process causes uninitialized kernel memory to get passed to drivers instead of the user space data:

BUG: KMSAN: uninit-value in check_array_args drivers/media/v4l2-core/v4l2-ioctl.c:3041 [inline]
BUG: KMSAN: uninit-value in video_usercopy+0x1631/0x3d30 drivers/media/v4l2-core/v4l2-ioctl.c:3315
CPU: 0 PID: 19595 Comm: syz-executor.4 Not tainted 5.11.0-rc7-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Call Trace:
 __dump_stack lib/dump_stack.c:79 [inline]
 dump_stack+0x21c/0x280 lib/dump_stack.c:120
 kmsan_report+0xfb/0x1e0 mm/kmsan/kmsan_report.c:118
 __msan_warning+0x5f/0xa0 mm/kmsan/kmsan_instr.c:197
 check_array_args drivers/media/v4l2-core/v4l2-ioctl.c:3041 [inline]
 video_usercopy+0x1631/0x3d30 drivers/media/v4l2-core/v4l2-ioctl.c:3315
 video_ioctl2+0x9f/0xb0 drivers/media/v4l2-core/v4l2-ioctl.c:3391
 v4l2_ioctl+0x255/0x290 drivers/media/v4l2-core/v4l2-dev.c:360
 v4l2_compat_ioctl32+0x2c6/0x370 drivers/media/v4l2-core/v4l2-compat-ioctl32.c:1248
 __do_compat_sys_ioctl fs/ioctl.c:842 [inline]
 __se_compat_sys_ioctl+0x53d/0x1100 fs/ioctl.c:793
 __ia32_compat_sys_ioctl+0x4a/0x70 fs/ioctl.c:793
 do_syscall_32_irqs_on arch/x86/entry/common.c:79 [inline]
 __do_fast_syscall_32+0x102/0x160 arch/x86/entry/common.c:141
 do_fast_syscall_32+0x6a/0xc0 arch/x86/entry/common.c:166
 do_SYSENTER_32+0x73/0x90 arch/x86/entry/common.c:209
 entry_SYSENTER_compat_after_hwframe+0x4d/0x5c

The time32 commands are defined but were never meant to be called on 64-bit machines, as those have always used time64 interfaces. I missed this in my patch that introduced the time64 handling on 32-bit platforms.

The problem in this case is the mismatch of one function checking for the numeric value of the command and another function checking for the type of process (native vs compat) instead, with the result being that for this combination, nothing gets copied into the buffer at all.

Avoid this by only trying to convert the time32 commands when running on a 32-bit kernel where these are defined in a meaningful way.

[hverkuil: fix 3 warnings: switch with no cases]

Fixes 577c89b0ce72 ("media: v4l2-core: fix v4l2_buffer handling for time64 ABI")
Reported-by [email protected]
---
 drivers/media/v4l2-core/v4l2-ioctl.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)
drivers/media/v4l2-core/v4l2-ioctl.c +6 -6
index 31d1342e61e8..7e8bf4b1ab2e 100644
@@ -3114,8 +3114,8 @@ static int check_array_args(unsigned int cmd, void *parg, size_t *array_size,
 
 static unsigned int video_translate_cmd(unsigned int cmd)
 {
+#if !defined(CONFIG_64BIT) && defined(CONFIG_COMPAT_32BIT_TIME)
 	switch (cmd) {
-#ifdef CONFIG_COMPAT_32BIT_TIME
 	case VIDIOC_DQEVENT_TIME32:
 		return VIDIOC_DQEVENT;
 	case VIDIOC_QUERYBUF_TIME32:
@@ -3126,8 +3126,8 @@ static unsigned int video_translate_cmd(unsigned int cmd)
 		return VIDIOC_DQBUF;
 	case VIDIOC_PREPARE_BUF_TIME32:
 		return VIDIOC_PREPARE_BUF;
-#endif
 	}
+#endif
 	if (in_compat_syscall())
 		return v4l2_compat_translate_cmd(cmd);
 
@@ -3168,8 +3168,8 @@ static int video_get_user(void __user *arg, void *parg,
 	} else if (in_compat_syscall()) {
 		err = v4l2_compat_get_user(arg, parg, cmd);
 	} else {
+#if !defined(CONFIG_64BIT) && defined(CONFIG_COMPAT_32BIT_TIME)
 		switch (cmd) {
-#ifdef CONFIG_COMPAT_32BIT_TIME
 		case VIDIOC_QUERYBUF_TIME32:
 		case VIDIOC_QBUF_TIME32:
 		case VIDIOC_DQBUF_TIME32:
@@ -3197,8 +3197,8 @@ static int video_get_user(void __user *arg, void *parg,
 			};
 			break;
 		}
-#endif
 		}
+#endif
 	}
 
 	/* zero out anything we don't copy from userspace */
@@ -3223,8 +3223,8 @@ static int video_put_user(void __user *arg, void *parg,
 	if (in_compat_syscall())
 		return v4l2_compat_put_user(arg, parg, cmd);
 
+#if !defined(CONFIG_64BIT) && defined(CONFIG_COMPAT_32BIT_TIME)
 	switch (cmd) {
-#ifdef CONFIG_COMPAT_32BIT_TIME
 	case VIDIOC_DQEVENT_TIME32: {
 		struct v4l2_event *ev = parg;
 		struct v4l2_event_time32 ev32;
@@ -3272,8 +3272,8 @@ static int video_put_user(void __user *arg, void *parg,
 			return -EFAULT;
 		break;
 	}
-#endif
 	}
+#endif
 
 	return 0;
 }
Signature · 1 line
2.30.2
author maintainer
#290 lore raw

PATCH 290/700 media: subdev: remove VIDIOC_DQEVENT_TIME32 handling

In reply to Greg Kroah-Hartman

From: Arnd Bergmann <[email protected]>

[ Upstream commit 765ba251d2522e2a0daa2f0793fd0f0ce34816ec ]

Converting the VIDIOC_DQEVENT_TIME32/VIDIOC_DQEVENT32/ VIDIOC_DQEVENT32_TIME32 arguments to the canonical form is done in common code, but for some reason I ended up adding another conversion helper to subdev_do_ioctl() as well. I must have concluded that this does not go through the common conversion, but it has done that since the ioctl handler was first added.

I assume this one is harmless as there should be no way to arrive here from user space if CONFIG_COMPAT_32BIT_TIME is set, but since it is dead code, it should just get removed.

On a 64-bit architecture, as well as a 32-bit architecture without CONFIG_COMPAT_32BIT_TIME, handling this command is a mistake, and the kernel should return an error.

Fixes 1a6c0b36dd19 ("media: v4l2-core: fix VIDIOC_DQEVENT for time64 ABI")
---
 drivers/media/v4l2-core/v4l2-subdev.c | 24 ------------------------
 1 file changed, 24 deletions(-)
drivers/media/v4l2-core/v4l2-subdev.c +0 -24
index 956dafab43d4..bf3aa9252458 100644
@@ -428,30 +428,6 @@ static long subdev_do_ioctl(struct file *file, unsigned int cmd, void *arg)
 
 		return v4l2_event_dequeue(vfh, arg, file->f_flags & O_NONBLOCK);
 
-	case VIDIOC_DQEVENT_TIME32: {
-		struct v4l2_event_time32 *ev32 = arg;
-		struct v4l2_event ev = { };
-
-		if (!(sd->flags & V4L2_SUBDEV_FL_HAS_EVENTS))
-			return -ENOIOCTLCMD;
-
-		rval = v4l2_event_dequeue(vfh, &ev, file->f_flags & O_NONBLOCK);
-
-		*ev32 = (struct v4l2_event_time32) {
-			.type		= ev.type,
-			.pending	= ev.pending,
-			.sequence	= ev.sequence,
-			.timestamp.tv_sec  = ev.timestamp.tv_sec,
-			.timestamp.tv_nsec = ev.timestamp.tv_nsec,
-			.id		= ev.id,
-		};
-
-		memcpy(&ev32->u, &ev.u, sizeof(ev.u));
-		memcpy(&ev32->reserved, &ev.reserved, sizeof(ev.reserved));
-
-		return rval;
-	}
-
 	case VIDIOC_SUBSCRIBE_EVENT:
 		return v4l2_subdev_call(sd, core, subscribe_event, vfh, arg);
 
Signature · 1 line
2.30.2
author maintainer
#291 lore raw

PATCH 291/700 media: s5p-g2d: Fix a memory leak on ctx->fh.m2m_ctx

In reply to Greg Kroah-Hartman

From: Dillon Min <[email protected]>

[ Upstream commit 5d11e6aad1811ea293ee2996cec9124f7fccb661 ]

The m2m_ctx resources was allocated by v4l2_m2m_ctx_init() in g2d_open() should be freed from g2d_release() when it's not used.

Fix it

Fixes 918847341af0 ("[media] v4l: add G2D driver for s5p device family")
---
 drivers/media/platform/s5p-g2d/g2d.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/media/platform/s5p-g2d/g2d.c +3 -0
index 15bcb7f6e113..1cb5eaabf340 100644
@@ -276,6 +276,9 @@ static int g2d_release(struct file *file)
 	struct g2d_dev *dev = video_drvdata(file);
 	struct g2d_ctx *ctx = fh2ctx(file->private_data);
 
+	mutex_lock(&dev->mutex);
+	v4l2_m2m_ctx_release(ctx->fh.m2m_ctx);
+	mutex_unlock(&dev->mutex);
 	v4l2_ctrl_handler_free(&ctx->ctrl_handler);
 	v4l2_fh_del(&ctx->fh);
 	v4l2_fh_exit(&ctx->fh);
Signature · 1 line
2.30.2
author maintainer
#292 lore raw

PATCH 292/700 media: i2c: rdacm21: Fix OV10640 powerup

In reply to Greg Kroah-Hartman

From: Jacopo Mondi <[email protected]>

[ Upstream commit ff75332b260cd33cc19000fdb5d256d9db4470d1 ]

The OV10640 image sensor powerdown signal is controlled by the first line of the OV490 GPIO pad #1, but the pad #0 identifier OV490_GPIO_OUTPUT_VALUE0 was erroneously used. As a result the image sensor powerdown signal was never asserted but was left floating and kept high by an internal pull-up resistor, causing sporadic failures during the image sensor startup phase.

Fix this by using the correct GPIO pad identifier and wait the mandatory 1.5 millisecond delay after the powerup lane is asserted. The reset delay is not characterized in the chip manual if not as "255 XVCLK + initialization". Wait for at least 3 milliseconds to guarantee the SCCB bus is available.

While at it also fix the reset sequence, as the reset line was released before the powerdown one, and the line was not cycled.

This commit fixes a sporadic start-up error triggered by a failure to read the OV10640 chip ID:
rdacm21 8-0054: OV10640 ID mismatch: (0x01)

Fixes a59f853b3b4b ("media: i2c: Add driver for RDACM21 camera module")
---
 drivers/media/i2c/rdacm21.c | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)
drivers/media/i2c/rdacm21.c +8 -2
index 179d107f494c..4b0dfd0a75e1 100644
@@ -333,13 +333,19 @@ static int ov10640_initialize(struct rdacm21_device *dev)
 {
 	u8 val;
 
-	/* Power-up OV10640 by setting RESETB and PWDNB pins high. */
+	/* Enable GPIO0#0 (reset) and GPIO1#0 (pwdn) as output lines. */
 	ov490_write_reg(dev, OV490_GPIO_SEL0, OV490_GPIO0);
 	ov490_write_reg(dev, OV490_GPIO_SEL1, OV490_SPWDN0);
 	ov490_write_reg(dev, OV490_GPIO_DIRECTION0, OV490_GPIO0);
 	ov490_write_reg(dev, OV490_GPIO_DIRECTION1, OV490_SPWDN0);
+
+	/* Power up OV10640 and then reset it. */
+	ov490_write_reg(dev, OV490_GPIO_OUTPUT_VALUE1, OV490_SPWDN0);
+	usleep_range(1500, 3000);
+
+	ov490_write_reg(dev, OV490_GPIO_OUTPUT_VALUE0, 0x00);
+	usleep_range(1500, 3000);
 	ov490_write_reg(dev, OV490_GPIO_OUTPUT_VALUE0, OV490_GPIO0);
-	ov490_write_reg(dev, OV490_GPIO_OUTPUT_VALUE0, OV490_SPWDN0);
 	usleep_range(3000, 5000);
 
 	/* Read OV10640 ID to test communications. */
Signature · 1 line
2.30.2
author maintainer
#293 lore raw

PATCH 293/700 media: i2c: rdacm21: Power up OV10640 before OV490

In reply to Greg Kroah-Hartman

From: Jacopo Mondi <[email protected]>

[ Upstream commit 2b821698dc73c00719e3dc367db712f727bbda85 ]

The current RDACM21 initialization routine powers up the OV10640 image sensor after the OV490 ISP. The ISP is programmed with a firmware loaded from an embedded serial flash that (most probably) tries to interact and program also the image sensor connected to the ISP.

As described in commit "media: i2c: rdacm21: Fix OV10640 powerup" the image sensor powerdown signal is kept high by an internal pull up resistor and occasionally fails to startup correctly if the powerdown line is not asserted explicitly. Failures in the OV10640 startup causes the OV490 firmware to fail to boot correctly resulting in the camera module initialization to fail consequentially.

Fix this by powering up the OV10640 image sensor before testing the OV490 firmware boot completion, by splitting the ov10640_initialize() function in an ov10640_power_up() one and an ov10640_check_id() one.

Also make sure the OV10640 identification procedure gives enough time to the image sensor to resume after the programming phase performed by the OV490 firmware by repeating the ID read procedure.

This commit fixes a sporadic start-up error triggered by a failure to detect the OV490 firmware boot completion:
rdacm21 8-0054: Timeout waiting for firmware boot

[hverkuil: fixed two typos in commit log]

Fixes a59f853b3b4b ("media: i2c: Add driver for RDACM21 camera module")
---
 drivers/media/i2c/rdacm21.c | 46 ++++++++++++++++++++++++++-----------
 1 file changed, 32 insertions(+), 14 deletions(-)
drivers/media/i2c/rdacm21.c +32 -14
index 4b0dfd0a75e1..50e2af522760 100644
@@ -69,6 +69,7 @@
 #define OV490_ISP_VSIZE_LOW		0x80820062
 #define OV490_ISP_VSIZE_HIGH		0x80820063
 
+#define OV10640_PID_TIMEOUT		20
 #define OV10640_ID_HIGH			0xa6
 #define OV10640_CHIP_ID			0x300a
 #define OV10640_PIXEL_RATE		55000000
@@ -329,10 +330,8 @@ static const struct v4l2_subdev_ops rdacm21_subdev_ops = {
 	.pad		= &rdacm21_subdev_pad_ops,
 };
 
-static int ov10640_initialize(struct rdacm21_device *dev)
+static void ov10640_power_up(struct rdacm21_device *dev)
 {
-	u8 val;
-
 	/* Enable GPIO0#0 (reset) and GPIO1#0 (pwdn) as output lines. */
 	ov490_write_reg(dev, OV490_GPIO_SEL0, OV490_GPIO0);
 	ov490_write_reg(dev, OV490_GPIO_SEL1, OV490_SPWDN0);
@@ -347,18 +346,35 @@ static int ov10640_initialize(struct rdacm21_device *dev)
 	usleep_range(1500, 3000);
 	ov490_write_reg(dev, OV490_GPIO_OUTPUT_VALUE0, OV490_GPIO0);
 	usleep_range(3000, 5000);
+}
 
-	/* Read OV10640 ID to test communications. */
-	ov490_write_reg(dev, OV490_SCCB_SLAVE0_DIR, OV490_SCCB_SLAVE_READ);
-	ov490_write_reg(dev, OV490_SCCB_SLAVE0_ADDR_HIGH, OV10640_CHIP_ID >> 8);
-	ov490_write_reg(dev, OV490_SCCB_SLAVE0_ADDR_LOW, OV10640_CHIP_ID & 0xff);
-
-	/* Trigger SCCB slave transaction and give it some time to complete. */
-	ov490_write_reg(dev, OV490_HOST_CMD, OV490_HOST_CMD_TRIGGER);
-	usleep_range(1000, 1500);
+static int ov10640_check_id(struct rdacm21_device *dev)
+{
+	unsigned int i;
+	u8 val;
 
-	ov490_read_reg(dev, OV490_SCCB_SLAVE0_DIR, &val);
-	if (val != OV10640_ID_HIGH) {
+	/* Read OV10640 ID to test communications. */
+	for (i = 0; i < OV10640_PID_TIMEOUT; ++i) {
+		ov490_write_reg(dev, OV490_SCCB_SLAVE0_DIR,
+				OV490_SCCB_SLAVE_READ);
+		ov490_write_reg(dev, OV490_SCCB_SLAVE0_ADDR_HIGH,
+				OV10640_CHIP_ID >> 8);
+		ov490_write_reg(dev, OV490_SCCB_SLAVE0_ADDR_LOW,
+				OV10640_CHIP_ID & 0xff);
+
+		/*
+		 * Trigger SCCB slave transaction and give it some time
+		 * to complete.
+		 */
+		ov490_write_reg(dev, OV490_HOST_CMD, OV490_HOST_CMD_TRIGGER);
+		usleep_range(1000, 1500);
+
+		ov490_read_reg(dev, OV490_SCCB_SLAVE0_DIR, &val);
+		if (val == OV10640_ID_HIGH)
+			break;
+		usleep_range(1000, 1500);
+	}
+	if (i == OV10640_PID_TIMEOUT) {
 		dev_err(dev->dev, "OV10640 ID mismatch: (0x%02x)\n", val);
 		return -ENODEV;
 	}
@@ -374,6 +390,8 @@ static int ov490_initialize(struct rdacm21_device *dev)
 	unsigned int i;
 	int ret;
 
+	ov10640_power_up(dev);
+
 	/*
 	 * Read OV490 Id to test communications. Give it up to 40msec to
 	 * exit from reset.
@@ -411,7 +429,7 @@ static int ov490_initialize(struct rdacm21_device *dev)
 		return -ENODEV;
 	}
 
-	ret = ov10640_initialize(dev);
+	ret = ov10640_check_id(dev);
 	if (ret)
 		return ret;
 
Signature · 1 line
2.30.2
author maintainer
#294 lore raw

PATCH 294/700 hwmon: (lm70) Revert "hwmon: (lm70) Add support for ACPI"

In reply to Greg Kroah-Hartman

From: Guenter Roeck <[email protected]>

[ Upstream commit ac61c8aae446b9c0fe18981fe721d4a43e283ad6 ]

This reverts commit b58bd4c6dfe709646ed9efcbba2a70643f9bc873.

None of the ACPI IDs introduced with the reverted patch is a valid ACPI device ID. Any ACPI users of this driver are advised to use PRP0001 and a devicetree-compatible device identification.

Fixes b58bd4c6dfe7 ("hwmon: (lm70) Add support for ACPI")
Cc Andrej Picej <[email protected]>
---
 drivers/hwmon/lm70.c | 26 +-------------------------
 1 file changed, 1 insertion(+), 25 deletions(-)
drivers/hwmon/lm70.c +1 -25
index 40eab3349904..6b884ea00987 100644
@@ -22,10 +22,10 @@
 #include <linux/hwmon.h>
 #include <linux/mutex.h>
 #include <linux/mod_devicetable.h>
+#include <linux/of.h>
 #include <linux/property.h>
 #include <linux/spi/spi.h>
 #include <linux/slab.h>
-#include <linux/acpi.h>
 
 #define DRVNAME		"lm70"
 
@@ -148,29 +148,6 @@ static const struct of_device_id lm70_of_ids[] = {
 MODULE_DEVICE_TABLE(of, lm70_of_ids);
 #endif
 
-#ifdef CONFIG_ACPI
-static const struct acpi_device_id lm70_acpi_ids[] = {
-	{
-		.id = "LM000070",
-		.driver_data = LM70_CHIP_LM70,
-	},
-	{
-		.id = "TMP00121",
-		.driver_data = LM70_CHIP_TMP121,
-	},
-	{
-		.id = "LM000071",
-		.driver_data = LM70_CHIP_LM71,
-	},
-	{
-		.id = "LM000074",
-		.driver_data = LM70_CHIP_LM74,
-	},
-	{},
-};
-MODULE_DEVICE_TABLE(acpi, lm70_acpi_ids);
-#endif
-
 static int lm70_probe(struct spi_device *spi)
 {
 	struct device *hwmon_dev;
@@ -217,7 +194,6 @@ static struct spi_driver lm70_driver = {
 	.driver = {
 		.name	= "lm70",
 		.of_match_table	= of_match_ptr(lm70_of_ids),
-		.acpi_match_table = ACPI_PTR(lm70_acpi_ids),
 	},
 	.id_table = lm70_ids,
 	.probe	= lm70_probe,
Signature · 1 line
2.30.2