6.5.13-rc1 review

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: David Woodhouse <[email protected]>

commit ef5dd8ec88ac11e8e353164407d55b73c988b369 upstream.

The xencons_connect_backend() function allocates a local interdomain event channel with xenbus_alloc_evtchn(), then calls bind_interdomain_evtchn_to_irq_lateeoi() to bind to that port# on the *remote* domain.

That doesn't work very well:

(qemu) device_add xen-console,id=con1,chardev=pty0

[   44.323872] xenconsole console-1: 2 xenbus_dev_probe on device/console/1
[   44.323995] xenconsole: probe of console-1 failed with error -2

Fix it to use bind_evtchn_to_irq_lateeoi(), which does the right thing by just binding that *local* event channel to an irq. The backend will do the interdomain binding.

This didn't affect the primary console because the setup for that is special — the toolstack allocates the guest event channel and the guest discovers it with HVMOP_get_param.

---
 drivers/tty/hvc/hvc_xen.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -436,7 +436,7 @@ static int xencons_connect_backend(struc
 	if (ret)
 		return ret;
 	info->evtchn = evtchn;
-	irq = bind_interdomain_evtchn_to_irq_lateeoi(dev, evtchn);
+	irq = bind_evtchn_to_irq_lateeoi(evtchn);
 	if (irq < 0)
 		return irq;
 	info->irq = irq;

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Lukas Wunner <[email protected]>

commit 70b70a4307cccebe91388337b1c85735ce4de6ff upstream.

struct pci_dev contains two flags which govern whether the device may suspend to D3cold:

* no_d3cold provides an opt-out for drivers (e.g. if a device is known to not wake from D3cold)

* d3cold_allowed provides an opt-out for user space (default is true, user space may set to false)

Since commit 9d26d3a8f1b0 ("PCI: Put PCIe ports into D3 during suspend"), the user space setting overwrites the driver setting. Essentially user space is trusted to know better than the driver whether D3cold is working.

That feels unsafe and wrong. Assume that the change was introduced inadvertently and do not overwrite no_d3cold when d3cold_allowed is modified. Instead, consider d3cold_allowed in addition to no_d3cold when choosing a suspend state for the device.

That way, user space may opt out of D3cold if the driver hasn't, but it may no longer force an opt in if the driver has opted out.

---
 drivers/pci/pci-acpi.c  |    2 +-
 drivers/pci/pci-sysfs.c |    5 +----
 2 files changed, 2 insertions(+), 5 deletions(-)

@@ -911,7 +911,7 @@ pci_power_t acpi_pci_choose_state(struct
 {
 	int acpi_state, d_max;
 
-	if (pdev->no_d3cold)
+	if (pdev->no_d3cold || !pdev->d3cold_allowed)
 		d_max = ACPI_STATE_D3_HOT;
 	else
 		d_max = ACPI_STATE_D3_COLD;

@@ -529,10 +529,7 @@ static ssize_t d3cold_allowed_store(stru
 		return -EINVAL;
 
 	pdev->d3cold_allowed = !!val;
-	if (pdev->d3cold_allowed)
-		pci_d3cold_enable(pdev);
-	else
-		pci_d3cold_disable(pdev);
+	pci_bridge_d3_update(pdev);
 
 	pm_runtime_resume(dev);
 

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: SeongJae Park <[email protected]>

commit 19467a950b49432a84bf6dbadbbb17bdf89418b7 upstream.

damon_sysfs_set_targets(), which updates the targets of the context for online commitment, do not remove targets that removed from the corresponding sysfs files. As a result, more than intended targets of the context can exist and hence consume memory and monitoring CPU resource more than expected.

Fix it by removing all targets of the context and fill up again using the user input. This could cause unnecessary memory dealloc and realloc operations, but this is not a hot code path. Also, note that damon_target is stateless, and hence no data is lost.

---
 mm/damon/sysfs.c |   68 ++++++++++++++++++++++++++++---------------------------
 1 file changed, 35 insertions(+), 33 deletions(-)

@@ -1144,58 +1144,60 @@ destroy_targets_out:
 	return err;
 }
 
-/*
- * Search a target in a context that corresponds to the sysfs target input.
- *
- * Return: pointer to the target if found, NULL if not found, or negative
- * error code if the search failed.
- */
-static struct damon_target *damon_sysfs_existing_target(
-		struct damon_sysfs_target *sys_target, struct damon_ctx *ctx)
+static int damon_sysfs_update_target(struct damon_target *target,
+		struct damon_ctx *ctx,
+		struct damon_sysfs_target *sys_target)
 {
 	struct pid *pid;
-	struct damon_target *t;
+	struct damon_region *r, *next;
 
-	if (!damon_target_has_pid(ctx)) {
-		/* Up to only one target for paddr could exist */
-		damon_for_each_target(t, ctx)
-			return t;
-		return NULL;
-	}
+	if (!damon_target_has_pid(ctx))
+		return 0;
 
-	/* ops.id should be DAMON_OPS_VADDR or DAMON_OPS_FVADDR */
 	pid = find_get_pid(sys_target->pid);
 	if (!pid)
-		return ERR_PTR(-EINVAL);
-	damon_for_each_target(t, ctx) {
-		if (t->pid == pid) {
-			put_pid(pid);
-			return t;
-		}
+		return -EINVAL;
+
+	/* no change to the target */
+	if (pid == target->pid) {
+		put_pid(pid);
+		return 0;
 	}
-	put_pid(pid);
-	return NULL;
+
+	/* remove old monitoring results and update the target's pid */
+	damon_for_each_region_safe(r, next, target)
+		damon_destroy_region(r, target);
+	put_pid(target->pid);
+	target->pid = pid;
+	return 0;
 }
 
 static int damon_sysfs_set_targets(struct damon_ctx *ctx,
 		struct damon_sysfs_targets *sysfs_targets)
 {
-	int i, err;
+	struct damon_target *t, *next;
+	int i = 0, err;
 
 	/* Multiple physical address space monitoring targets makes no sense */
 	if (ctx->ops.id == DAMON_OPS_PADDR && sysfs_targets->nr > 1)
 		return -EINVAL;
 
-	for (i = 0; i < sysfs_targets->nr; i++) {
+	damon_for_each_target_safe(t, next, ctx) {
+		if (i < sysfs_targets->nr) {
+			damon_sysfs_update_target(t, ctx,
+					sysfs_targets->targets_arr[i]);
+		} else {
+			if (damon_target_has_pid(ctx))
+				put_pid(t->pid);
+			damon_destroy_target(t);
+		}
+		i++;
+	}
+
+	for (; i < sysfs_targets->nr; i++) {
 		struct damon_sysfs_target *st = sysfs_targets->targets_arr[i];
-		struct damon_target *t = damon_sysfs_existing_target(st, ctx);
 
-		if (IS_ERR(t))
-			return PTR_ERR(t);
-		if (!t)
-			err = damon_sysfs_add_target(st, ctx);
-		else
-			err = damon_sysfs_set_regions(t, st->regions);
+		err = damon_sysfs_add_target(st, ctx);
 		if (err)
 			return err;
 	}

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: SeongJae Park <[email protected]>

commit 9732336006764e2ee61225387e3c70eae9139035 upstream.

When user input is committed online, DAMON sysfs interface is ignoring the user input for the monitoring target regions. Such request is valid and useful for fixed monitoring target regions-based monitoring ops like 'paddr' or 'fvaddr'.

Update the region boundaries as user specified, too. Note that the monitoring results of the regions that overlap between the latest monitoring target regions and the new target regions are preserved.

Treat empty monitoring target regions user request as a request to just make no change to the monitoring target regions. Otherwise, users should set the monitoring target regions same to current one for every online input commit, and it could be challenging for dynamic monitoring target regions update DAMON ops like 'vaddr'. If the user really need to remove all monitoring target regions, they can simply remove the target and then create the target again with empty target regions.

---
 mm/damon/sysfs.c |   47 ++++++++++++++++++++++++++++++-----------------
 1 file changed, 30 insertions(+), 17 deletions(-)

@@ -1144,34 +1144,47 @@ destroy_targets_out:
 	return err;
 }
 
-static int damon_sysfs_update_target(struct damon_target *target,
-		struct damon_ctx *ctx,
-		struct damon_sysfs_target *sys_target)
+static int damon_sysfs_update_target_pid(struct damon_target *target, int pid)
 {
-	struct pid *pid;
-	struct damon_region *r, *next;
-
-	if (!damon_target_has_pid(ctx))
-		return 0;
+	struct pid *pid_new;
 
-	pid = find_get_pid(sys_target->pid);
-	if (!pid)
+	pid_new = find_get_pid(pid);
+	if (!pid_new)
 		return -EINVAL;
 
-	/* no change to the target */
-	if (pid == target->pid) {
-		put_pid(pid);
+	if (pid_new == target->pid) {
+		put_pid(pid_new);
 		return 0;
 	}
 
-	/* remove old monitoring results and update the target's pid */
-	damon_for_each_region_safe(r, next, target)
-		damon_destroy_region(r, target);
 	put_pid(target->pid);
-	target->pid = pid;
+	target->pid = pid_new;
 	return 0;
 }
 
+static int damon_sysfs_update_target(struct damon_target *target,
+		struct damon_ctx *ctx,
+		struct damon_sysfs_target *sys_target)
+{
+	int err;
+
+	if (damon_target_has_pid(ctx)) {
+		err = damon_sysfs_update_target_pid(target, sys_target->pid);
+		if (err)
+			return err;
+	}
+
+	/*
+	 * Do monitoring target region boundary update only if one or more
+	 * regions are set by the user.  This is for keeping current monitoring
+	 * target results and range easier, especially for dynamic monitoring
+	 * target regions update ops like 'vaddr'.
+	 */
+	if (sys_target->regions->nr)
+		err = damon_sysfs_set_regions(target, sys_target->regions);
+	return err;
+}
+
 static int damon_sysfs_set_targets(struct damon_ctx *ctx,
 		struct damon_sysfs_targets *sysfs_targets)
 {

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Krister Johansen <[email protected]>

commit 8b793bcda61f6c3ed4f5b2ded7530ef6749580cb upstream.

Setting softlockup_panic from do_sysctl_args() causes it to take effect later in boot. The lockup detector is enabled before SMP is brought online, but do_sysctl_args runs afterwards. If a user wants to set softlockup_panic on boot and have it trigger should a softlockup occur during onlining of the non-boot processors, they could do this prior to commit f117955a2255 ("kernel/watchdog.c: convert {soft/hard}lockup boot parameters to sysctl aliases"). However, after this commit the value of softlockup_panic is set too late to be of help for this type of problem. Restore the prior behavior.

---
 fs/proc/proc_sysctl.c |    1 -
 kernel/watchdog.c     |    7 +++++++
 2 files changed, 7 insertions(+), 1 deletion(-)

@@ -1574,7 +1574,6 @@ static const struct sysctl_alias sysctl_
 	{"hung_task_panic",			"kernel.hung_task_panic" },
 	{"numa_zonelist_order",			"vm.numa_zonelist_order" },
 	{"softlockup_all_cpu_backtrace",	"kernel.softlockup_all_cpu_backtrace" },
-	{"softlockup_panic",			"kernel.softlockup_panic" },
 	{ }
 };
 

@@ -283,6 +283,13 @@ static DEFINE_PER_CPU(struct hrtimer, wa
 static DEFINE_PER_CPU(bool, softlockup_touch_sync);
 static unsigned long soft_lockup_nmi_warn;
 
+static int __init softlockup_panic_setup(char *str)
+{
+	softlockup_panic = simple_strtoul(str, NULL, 0);
+	return 1;
+}
+__setup("softlockup_panic=", softlockup_panic_setup);
+
 static int __init nowatchdog_setup(char *str)
 {
 	watchdog_user_enabled = 0;

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Koichiro Den <[email protected]>

commit e7250ab7ca4998fe026f2149805b03e09dc32498 upstream.

In iopt_area_split(), if the original iopt_area has filled a domain and is linked to domains_itree, pages_nodes have to be properly reinserted. Otherwise the domains_itree becomes corrupted and we will UAF.

---
 drivers/iommu/iommufd/io_pagetable.c |   10 ++++++++++
 1 file changed, 10 insertions(+)

@@ -1060,6 +1060,16 @@ static int iopt_area_split(struct iopt_a
 	if (WARN_ON(rc))
 		goto err_remove_lhs;
 
+	/*
+	 * If the original area has filled a domain, domains_itree has to be
+	 * updated.
+	 */
+	if (area->storage_domain) {
+		interval_tree_remove(&area->pages_node, &pages->domains_itree);
+		interval_tree_insert(&lhs->pages_node, &pages->domains_itree);
+		interval_tree_insert(&rhs->pages_node, &pages->domains_itree);
+	}
+
 	lhs->storage_domain = area->storage_domain;
 	lhs->pages = area->pages;
 	rhs->storage_domain = area->storage_domain;

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Helge Deller <[email protected]>

commit 8a32aa17c1cd48df1ddaa78e45abcb8c7a2220d6 upstream.

The pointer to the next STI font is actually a signed 32-bit offset. With this change the 64-bit kernel will correctly subract the (signed 32-bit) offset instead of adding a (unsigned 32-bit) offset. It has no effect on 32-bit kernels.

This fixes the stifb driver with a 64-bit kernel on qemu.

---
 include/video/sticore.h |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -232,7 +232,7 @@ struct sti_rom_font {
 	 u8 height;
 	 u8 font_type;		/* language type */
 	 u8 bytes_per_char;
-	u32 next_font;
+	s32 next_font;		/* note: signed int */
 	 u8 underline_height;
 	 u8 underline_pos;
 	 u8 res008[2];

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Mikulas Patocka <[email protected]>

commit 9793c269da6cd339757de6ba5b2c8681b54c99af upstream.

The commit 5054e778fcd9c ("dm crypt: allocate compound pages if possible") changed dm-crypt to use compound pages to improve performance. Unfortunately, there was an oversight: the allocation of compound pages was not accounted at all. Normal pages are accounted in a percpu counter cc->n_allocated_pages and dm-crypt is limited to allocate at most 2% of memory. Because compound pages were not accounted at all, dm-crypt could allocate memory over the 2% limit.

Fix this by adding the accounting of compound pages, so that memory consumption of dm-crypt is properly limited.

---
 drivers/md/dm-crypt.c |   15 ++++++++++++---
 1 file changed, 12 insertions(+), 3 deletions(-)

@@ -1700,11 +1700,17 @@ retry:
 		order = min(order, remaining_order);
 
 		while (order > 0) {
+			if (unlikely(percpu_counter_read_positive(&cc->n_allocated_pages) +
+					(1 << order) > dm_crypt_pages_per_client))
+				goto decrease_order;
 			pages = alloc_pages(gfp_mask
 				| __GFP_NOMEMALLOC | __GFP_NORETRY | __GFP_NOWARN | __GFP_COMP,
 				order);
-			if (likely(pages != NULL))
+			if (likely(pages != NULL)) {
+				percpu_counter_add(&cc->n_allocated_pages, 1 << order);
 				goto have_pages;
+			}
+decrease_order:
 			order--;
 		}
 

@@ -1742,10 +1748,13 @@ static void crypt_free_buffer_pages(stru
 
 	if (clone->bi_vcnt > 0) { /* bio_for_each_folio_all crashes with an empty bio */
 		bio_for_each_folio_all(fi, clone) {
-			if (folio_test_large(fi.folio))
+			if (folio_test_large(fi.folio)) {
+				percpu_counter_sub(&cc->n_allocated_pages,
+						1 << folio_order(fi.folio));
 				folio_put(fi.folio);
-			else
+			} else {
 				mempool_free(&fi.folio->page, &cc->page_pool);
+			}
 		}
 	}
 }

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: SeongJae Park <[email protected]>

commit 44063f125af4bb4efd1d500d8091fa33a98af325 upstream.

When calculating the hotness threshold for lru_prio scheme of DAMON_LRU_SORT, the module divides some values by the maximum nr_accesses. However, due to the type of the related variables, simple division-based calculation of the divisor can return zero. As a result, divide-by-zero is possible. Fix it by using damon_max_nr_accesses(), which handles the case.

---
 mm/damon/lru_sort.c |    4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

@@ -193,9 +193,7 @@ static int damon_lru_sort_apply_paramete
 	if (err)
 		return err;
 
-	/* aggr_interval / sample_interval is the maximum nr_accesses */
-	hot_thres = damon_lru_sort_mon_attrs.aggr_interval /
-		damon_lru_sort_mon_attrs.sample_interval *
+	hot_thres = damon_max_nr_accesses(&damon_lru_sort_mon_attrs) *
 		hot_thres_access_freq / 1000;
 	scheme = damon_lru_sort_new_hot_scheme(hot_thres);
 	if (!scheme)

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: SeongJae Park <[email protected]>

commit 3bafc47d3c4a2fc4d3b382aeb3c087f8fc84d9fd upstream.

When calculating the hotness of each region for the under-quota regions prioritization, DAMON divides some values by the maximum nr_accesses. However, due to the type of the related variables, simple division-based calculation of the divisor can return zero. As a result, divide-by-zero is possible. Fix it by using damon_max_nr_accesses(), which handles the case.

---
 mm/damon/ops-common.c |    5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

@@ -73,7 +73,6 @@ void damon_pmdp_mkold(pmd_t *pmd, struct
 int damon_hot_score(struct damon_ctx *c, struct damon_region *r,
 			struct damos *s)
 {
-	unsigned int max_nr_accesses;
 	int freq_subscore;
 	unsigned int age_in_sec;
 	int age_in_log, age_subscore;

@@ -81,8 +80,8 @@ int damon_hot_score(struct damon_ctx *c,
 	unsigned int age_weight = s->quota.weight_age;
 	int hotness;
 
-	max_nr_accesses = c->attrs.aggr_interval / c->attrs.sample_interval;
-	freq_subscore = r->nr_accesses * DAMON_MAX_SUBSCORE / max_nr_accesses;
+	freq_subscore = r->nr_accesses * DAMON_MAX_SUBSCORE /
+		damon_max_nr_accesses(&c->attrs);
 
 	age_in_sec = (unsigned long)r->age * c->attrs.aggr_interval / 1000000;
 	for (age_in_log = 0; age_in_log < DAMON_MAX_AGE_IN_LOG && age_in_sec;

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: SeongJae Park <[email protected]>

commit 35f5d94187a6a3a8df2cba54beccca1c2379edb8 upstream.

Patch series "avoid divide-by-zero due to max_nr_accesses overflow".

The maximum nr_accesses of given DAMON context can be calculated by dividing the aggregation interval by the sampling interval. Some logics in DAMON uses the maximum nr_accesses as a divisor. Hence, the value shouldn't be zero. Such case is avoided since DAMON avoids setting the agregation interval as samller than the sampling interval. However, since nr_accesses is unsigned int while the intervals are unsigned long, the maximum nr_accesses could be zero while casting.

Avoid the divide-by-zero by implementing a function that handles the corner case (first patch), and replaces the vulnerable direct max nr_accesses calculations (remaining patches).

Note that the patches for the replacements are divided for broken commits, to make backporting on required tres easier. Especially, the last patch is for a patch that not yet merged into the mainline but in mm tree.

This patch (of 4):

The maximum nr_accesses of given DAMON context can be calculated by dividing the aggregation interval by the sampling interval. Some logics in DAMON uses the maximum nr_accesses as a divisor. Hence, the value shouldn't be zero. Such case is avoided since DAMON avoids setting the agregation interval as samller than the sampling interval. However, since nr_accesses is unsigned int while the intervals are unsigned long, the maximum nr_accesses could be zero while casting. Implement a function that handles the corner case.

Note that this commit is not fixing the real issue since this is only introducing the safe function that will replaces the problematic divisions. The replacements will be made by followup commits, to make backporting on stable series easier.

---
 include/linux/damon.h |    7 +++++++
 1 file changed, 7 insertions(+)

@@ -626,6 +626,13 @@ static inline bool damon_target_has_pid(
 	return ctx->ops.id == DAMON_OPS_VADDR || ctx->ops.id == DAMON_OPS_FVADDR;
 }
 
+static inline unsigned int damon_max_nr_accesses(const struct damon_attrs *attrs)
+{
+	/* {aggr,sample}_interval are unsigned long, hence could overflow */
+	return min(attrs->aggr_interval / attrs->sample_interval,
+			(unsigned long)UINT_MAX);
+}
+
 
 int damon_start(struct damon_ctx **ctxs, int nr_ctxs, bool exclusive);
 int damon_stop(struct damon_ctx **ctxs, int nr_ctxs);

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: SeongJae Park <[email protected]>

commit d35963bfb05877455228ecec6b194f624489f96a upstream.

When monitoring attributes are changed, DAMON updates access rate of the monitoring results accordingly. For that, it divides some values by the maximum nr_accesses. However, due to the type of the related variables, simple division-based calculation of the divisor can return zero. As a result, divide-by-zero is possible. Fix it by using damon_max_nr_accesses(), which handles the case.

---
 mm/damon/core.c |   10 ++--------
 1 file changed, 2 insertions(+), 8 deletions(-)

@@ -476,20 +476,14 @@ static unsigned int damon_age_for_new_at
 static unsigned int damon_accesses_bp_to_nr_accesses(
 		unsigned int accesses_bp, struct damon_attrs *attrs)
 {
-	unsigned int max_nr_accesses =
-		attrs->aggr_interval / attrs->sample_interval;
-
-	return accesses_bp * max_nr_accesses / 10000;
+	return accesses_bp * damon_max_nr_accesses(attrs) / 10000;
 }
 
 /* convert nr_accesses to access ratio in bp (per 10,000) */
 static unsigned int damon_nr_accesses_to_accesses_bp(
 		unsigned int nr_accesses, struct damon_attrs *attrs)
 {
-	unsigned int max_nr_accesses =
-		attrs->aggr_interval / attrs->sample_interval;
-
-	return nr_accesses * 10000 / max_nr_accesses;
+	return nr_accesses * 10000 / damon_max_nr_accesses(attrs);
 }
 
 static unsigned int damon_nr_accesses_for_new_attrs(unsigned int nr_accesses,

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: SeongJae Park <[email protected]>

commit ae636ae2bbfd9279f5681dbf320d1da817e52b68 upstream.

DAMON sysfs interface's before_damos_apply callback (damon_sysfs_before_damos_apply()), which creates the DAMOS tried regions for each DAMOS action applied region, is not handling the allocation failure for the sysfs directory data. As a result, NULL pointer derefeence is possible. Fix it by handling the case.

---
 mm/damon/sysfs-schemes.c |    2 ++
 1 file changed, 2 insertions(+)

@@ -1649,6 +1649,8 @@ static int damon_sysfs_before_damos_appl
 
 	sysfs_regions = sysfs_schemes->schemes_arr[schemes_idx]->tried_regions;
 	region = damon_sysfs_scheme_region_alloc(r);
+	if (!region)
+		return 0;
 	list_add_tail(&region->list, &sysfs_regions->regions_list);
 	sysfs_regions->nr_regions++;
 	if (kobject_init_and_add(&region->kobj,

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: SeongJae Park <[email protected]>

commit 84055688b6bc075c92a88e2d6c3ad26ab93919f9 upstream.

DAMOS tried regions sysfs directory allocation function (damon_sysfs_scheme_regions_alloc()) is not handling the memory allocation failure. In the case, the code will dereference NULL pointer. Handle the failure to avoid such invalid access.

---
 mm/damon/sysfs-schemes.c |    3 +++
 1 file changed, 3 insertions(+)

@@ -125,6 +125,9 @@ damon_sysfs_scheme_regions_alloc(void)
 	struct damon_sysfs_scheme_regions *regions = kmalloc(sizeof(*regions),
 			GFP_KERNEL);
 
+	if (!regions)
+		return NULL;
+
 	regions->kobj = (struct kobject){};
 	INIT_LIST_HEAD(&regions->regions_list);
 	regions->nr_regions = 0;

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: SeongJae Park <[email protected]>

commit b4936b544b08ed44949055b92bd25f77759ebafc upstream.

Patch series "mm/damon/sysfs: fix unhandled return values".

Some of DAMON sysfs interface code is not handling return values from some functions. As a result, confusing user input handling or NULL-dereference is possible. Check those properly.

This patch (of 3):

damon_sysfs_update_target() returns error code for failures, but its caller, damon_sysfs_set_targets() is ignoring that. The update function seems making no critical change in case of such failures, but the behavior will look like DAMON sysfs is silently ignoring or only partially accepting the user input. Fix it.

---
 mm/damon/sysfs.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

@@ -1197,8 +1197,10 @@ static int damon_sysfs_set_targets(struc
 
 	damon_for_each_target_safe(t, next, ctx) {
 		if (i < sysfs_targets->nr) {
-			damon_sysfs_update_target(t, ctx,
+			err = damon_sysfs_update_target(t, ctx,
 					sysfs_targets->targets_arr[i]);
+			if (err)
+				return err;
 		} else {
 			if (damon_target_has_pid(ctx))
 				put_pid(t->pid);

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: John David Anglin <[email protected]>

commit ad4aa06e1d92b06ed56c7240252927bd60632efe upstream.

An excerpt from the PA8800 ERS states:

* The PA8800 violates the seven instruction pipeline rule when performing TLB inserts or PxTLBE instructions with the PSW C bit on. The instruction will take effect by the 12th instruction after the insert or purge.

I believe we have a problem with handling TLB misses. We don't fill the pipeline following TLB inserts. As a result, we likely fault again after returning from the interruption.

The above statement indicates that we need at least seven instructions after the insert on pre PA8800 processors and we need 12 instructions on PA8800/PA8900 processors.

Here we add macros and code to provide the required number instructions after a TLB insert.

---
 arch/parisc/kernel/entry.S |   81 ++++++++++++++++++++++++++++-----------------
 1 file changed, 52 insertions(+), 29 deletions(-)

@@ -36,6 +36,24 @@
 	.level 2.0
 #endif
 
+/*
+ * We need seven instructions after a TLB insert for it to take effect.
+ * The PA8800/PA8900 processors are an exception and need 12 instructions.
+ * The RFI changes both IAOQ_Back and IAOQ_Front, so it counts as one.
+ */
+#ifdef CONFIG_64BIT
+#define NUM_PIPELINE_INSNS    12
+#else
+#define NUM_PIPELINE_INSNS    7
+#endif
+
+	/* Insert num nops */
+	.macro	insert_nops num
+	.rept \num
+	nop
+	.endr
+	.endm
+
 	/* Get aligned page_table_lock address for this mm from cr28/tr4 */
 	.macro  get_ptl reg
 	mfctl	%cr28,\reg

@@ -415,24 +433,20 @@
 3:
 	.endm
 
-	/* Release page_table_lock without reloading lock address.
-	   We use an ordered store to ensure all prior accesses are
-	   performed prior to releasing the lock. */
-	.macro		ptl_unlock0	spc,tmp,tmp2
+	/* Release page_table_lock if for user space. We use an ordered
+	   store to ensure all prior accesses are performed prior to
+	   releasing the lock. Note stw may not be executed, so we
+	   provide one extra nop when CONFIG_TLB_PTLOCK is defined. */
+	.macro		ptl_unlock	spc,tmp,tmp2
 #ifdef CONFIG_TLB_PTLOCK
-98:	ldi		__ARCH_SPIN_LOCK_UNLOCKED_VAL, \tmp2
+98:	get_ptl		\tmp
+	ldi		__ARCH_SPIN_LOCK_UNLOCKED_VAL, \tmp2
 	or,COND(=)	%r0,\spc,%r0
 	stw,ma		\tmp2,0(\tmp)
 99:	ALTERNATIVE(98b, 99b, ALT_COND_NO_SMP, INSN_NOP)
-#endif
-	.endm
-
-	/* Release page_table_lock. */
-	.macro		ptl_unlock1	spc,tmp,tmp2
-#ifdef CONFIG_TLB_PTLOCK
-98:	get_ptl		\tmp
-	ptl_unlock0	\spc,\tmp,\tmp2
-99:	ALTERNATIVE(98b, 99b, ALT_COND_NO_SMP, INSN_NOP)
+	insert_nops	NUM_PIPELINE_INSNS - 4
+#else
+	insert_nops	NUM_PIPELINE_INSNS - 1
 #endif
 	.endm
 

@@ -1124,7 +1138,7 @@ dtlb_miss_20w:
 	
 	idtlbt          pte,prot
 
-	ptl_unlock1	spc,t0,t1
+	ptl_unlock	spc,t0,t1
 	rfir
 	nop
 

@@ -1133,6 +1147,7 @@ dtlb_check_alias_20w:
 
 	idtlbt          pte,prot
 
+	insert_nops	NUM_PIPELINE_INSNS - 1
 	rfir
 	nop
 

@@ -1150,7 +1165,7 @@ nadtlb_miss_20w:
 
 	idtlbt          pte,prot
 
-	ptl_unlock1	spc,t0,t1
+	ptl_unlock	spc,t0,t1
 	rfir
 	nop
 

@@ -1159,6 +1174,7 @@ nadtlb_check_alias_20w:
 
 	idtlbt          pte,prot
 
+	insert_nops	NUM_PIPELINE_INSNS - 1
 	rfir
 	nop
 

@@ -1184,7 +1200,7 @@ dtlb_miss_11:
 
 	mtsp		t1, %sr1	/* Restore sr1 */
 
-	ptl_unlock1	spc,t0,t1
+	ptl_unlock	spc,t0,t1
 	rfir
 	nop
 

@@ -1194,6 +1210,7 @@ dtlb_check_alias_11:
 	idtlba          pte,(va)
 	idtlbp          prot,(va)
 
+	insert_nops	NUM_PIPELINE_INSNS - 1
 	rfir
 	nop
 

@@ -1217,7 +1234,7 @@ nadtlb_miss_11:
 
 	mtsp		t1, %sr1	/* Restore sr1 */
 
-	ptl_unlock1	spc,t0,t1
+	ptl_unlock	spc,t0,t1
 	rfir
 	nop
 

@@ -1227,6 +1244,7 @@ nadtlb_check_alias_11:
 	idtlba          pte,(va)
 	idtlbp          prot,(va)
 
+	insert_nops	NUM_PIPELINE_INSNS - 1
 	rfir
 	nop
 

@@ -1246,7 +1264,7 @@ dtlb_miss_20:
 
 	idtlbt          pte,prot
 
-	ptl_unlock1	spc,t0,t1
+	ptl_unlock	spc,t0,t1
 	rfir
 	nop
 

@@ -1255,6 +1273,7 @@ dtlb_check_alias_20:
 	
 	idtlbt          pte,prot
 
+	insert_nops	NUM_PIPELINE_INSNS - 1
 	rfir
 	nop
 

@@ -1274,7 +1293,7 @@ nadtlb_miss_20:
 	
 	idtlbt		pte,prot
 
-	ptl_unlock1	spc,t0,t1
+	ptl_unlock	spc,t0,t1
 	rfir
 	nop
 

@@ -1283,6 +1302,7 @@ nadtlb_check_alias_20:
 
 	idtlbt          pte,prot
 
+	insert_nops	NUM_PIPELINE_INSNS - 1
 	rfir
 	nop
 

@@ -1319,7 +1339,7 @@ itlb_miss_20w:
 	
 	iitlbt          pte,prot
 
-	ptl_unlock1	spc,t0,t1
+	ptl_unlock	spc,t0,t1
 	rfir
 	nop
 

@@ -1343,7 +1363,7 @@ naitlb_miss_20w:
 
 	iitlbt          pte,prot
 
-	ptl_unlock1	spc,t0,t1
+	ptl_unlock	spc,t0,t1
 	rfir
 	nop
 

@@ -1352,6 +1372,7 @@ naitlb_check_alias_20w:
 
 	iitlbt		pte,prot
 
+	insert_nops	NUM_PIPELINE_INSNS - 1
 	rfir
 	nop
 

@@ -1377,7 +1398,7 @@ itlb_miss_11:
 
 	mtsp		t1, %sr1	/* Restore sr1 */
 
-	ptl_unlock1	spc,t0,t1
+	ptl_unlock	spc,t0,t1
 	rfir
 	nop
 

@@ -1401,7 +1422,7 @@ naitlb_miss_11:
 
 	mtsp		t1, %sr1	/* Restore sr1 */
 
-	ptl_unlock1	spc,t0,t1
+	ptl_unlock	spc,t0,t1
 	rfir
 	nop
 

@@ -1411,6 +1432,7 @@ naitlb_check_alias_11:
 	iitlba          pte,(%sr0, va)
 	iitlbp          prot,(%sr0, va)
 
+	insert_nops	NUM_PIPELINE_INSNS - 1
 	rfir
 	nop
 

@@ -1431,7 +1453,7 @@ itlb_miss_20:
 
 	iitlbt          pte,prot
 
-	ptl_unlock1	spc,t0,t1
+	ptl_unlock	spc,t0,t1
 	rfir
 	nop
 

@@ -1451,7 +1473,7 @@ naitlb_miss_20:
 
 	iitlbt          pte,prot
 
-	ptl_unlock1	spc,t0,t1
+	ptl_unlock	spc,t0,t1
 	rfir
 	nop
 

@@ -1460,6 +1482,7 @@ naitlb_check_alias_20:
 
 	iitlbt          pte,prot
 
+	insert_nops	NUM_PIPELINE_INSNS - 1
 	rfir
 	nop
 

@@ -1481,7 +1504,7 @@ dbit_trap_20w:
 		
 	idtlbt          pte,prot
 
-	ptl_unlock0	spc,t0,t1
+	ptl_unlock	spc,t0,t1
 	rfir
 	nop
 #else

@@ -1507,7 +1530,7 @@ dbit_trap_11:
 
 	mtsp            t1, %sr1     /* Restore sr1 */
 
-	ptl_unlock0	spc,t0,t1
+	ptl_unlock	spc,t0,t1
 	rfir
 	nop
 

@@ -1527,7 +1550,7 @@ dbit_trap_20:
 	
 	idtlbt		pte,prot
 
-	ptl_unlock0	spc,t0,t1
+	ptl_unlock	spc,t0,t1
 	rfir
 	nop
 #endif

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Werner Sembach <[email protected]>

commit 0da9eccde3270b832c059ad618bf66e510c75d33 upstream.

The TongFang GMxXGxx/TUXEDO Stellaris/Pollaris Gen5 needs IRQ overriding for the keyboard to work.

Adding an entry for this laptop to the override_table makes the internal keyboard functional.

---
 drivers/acpi/resource.c |   12 ++++++++++++
 1 file changed, 12 insertions(+)

@@ -496,6 +496,18 @@ static const struct dmi_system_id mainge
 		}
 	},
 	{
+		/* TongFang GMxXGxx/TUXEDO Polaris 15 Gen5 AMD */
+		.matches = {
+			DMI_MATCH(DMI_BOARD_NAME, "GMxXGxx"),
+		},
+	},
+	{
+		/* TongFang GM6XGxX/TUXEDO Stellaris 16 Gen5 AMD */
+		.matches = {
+			DMI_MATCH(DMI_BOARD_NAME, "GM6XGxX"),
+		},
+	},
+	{
 		.ident = "MAINGEAR Vector Pro 2 17",
 		.matches = {
 			DMI_MATCH(DMI_SYS_VENDOR, "Micro Electronics Inc"),

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Mark Brown <[email protected]>

commit 0ec7731655de196bc1e4af99e495b38778109d22 upstream.

When we sync the register cache we do so with the cache bypassed in order to avoid overhead from writing the synced values back into the cache. If the regmap has ranges and the selector register for those ranges is in a register which is cached this has the unfortunate side effect of meaning that the physical and cached copies of the selector register can be out of sync after a cache sync. The cache will have whatever the selector was when the sync started and the hardware will have the selector for the register that was synced last.

Fix this by rewriting all cached selector registers after every sync, ensuring that the hardware and cache have the same content. This will result in extra writes that wouldn't otherwise be needed but is simple so hopefully robust. We don't read from the hardware since not all devices have physical read support.

Given that nobody noticed this until now it is likely that we are rarely if ever hitting this case.

---
 drivers/base/regmap/regcache.c |   30 ++++++++++++++++++++++++++++++
 1 file changed, 30 insertions(+)

@@ -334,6 +334,11 @@ static int regcache_default_sync(struct
 	return 0;
 }
 
+static int rbtree_all(const void *key, const struct rb_node *node)
+{
+	return 0;
+}
+
 /**
  * regcache_sync - Sync the register cache with the hardware.
  *

@@ -351,6 +356,7 @@ int regcache_sync(struct regmap *map)
 	unsigned int i;
 	const char *name;
 	bool bypass;
+	struct rb_node *node;
 
 	if (WARN_ON(map->cache_type == REGCACHE_NONE))
 		return -EINVAL;

@@ -392,6 +398,30 @@ out:
 	/* Restore the bypass state */
 	map->cache_bypass = bypass;
 	map->no_sync_defaults = false;
+
+	/*
+	 * If we did any paging with cache bypassed and a cached
+	 * paging register then the register and cache state might
+	 * have gone out of sync, force writes of all the paging
+	 * registers.
+	 */
+	rb_for_each(node, 0, &map->range_tree, rbtree_all) {
+		struct regmap_range_node *this =
+			rb_entry(node, struct regmap_range_node, node);
+
+		/* If there's nothing in the cache there's nothing to sync */
+		ret = regcache_read(map, this->selector_reg, &i);
+		if (ret != 0)
+			continue;
+
+		ret = _regmap_write(map, this->selector_reg, i);
+		if (ret != 0) {
+			dev_err(map->dev, "Failed to write %x = %x: %d\n",
+				this->selector_reg, i, ret);
+			break;
+		}
+	}
+
 	map->unlock(map->lock_arg);
 
 	regmap_async_complete(map);

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Johan Hovold <[email protected]>

commit 1a5352a81b4720ba43d9c899974e3bddf7ce0ce8 upstream.

The ath11k active pdevs are protected by RCU but the temperature event handling code calling ath11k_mac_get_ar_by_pdev_id() was not marked as a read-side critical section as reported by RCU lockdep:

	=============================
	WARNING: suspicious RCU usage
	6.6.0-rc6 #7 Not tainted
	-----------------------------
	drivers/net/wireless/ath/ath11k/mac.c:638 suspicious rcu_dereference_check() usage!

	other info that might help us debug this:

	rcu_scheduler_active = 2, debug_locks = 1
	no locks held by swapper/0/0.
	...
	Call trace:
	...
	 lockdep_rcu_suspicious+0x16c/0x22c
	 ath11k_mac_get_ar_by_pdev_id+0x194/0x1b0 [ath11k]
	 ath11k_wmi_tlv_op_rx+0xa84/0x2c1c [ath11k]
	 ath11k_htc_rx_completion_handler+0x388/0x510 [ath11k]

Mark the code in question as an RCU read-side critical section to avoid any potential use-after-free issues.

Tested-on: WCN6855 hw2.1 PCI WLAN.HSP.1.1-03125-QCAHSPSWPL_V1_V2_SILICONZ_LITE-3.6510.23

---
 drivers/net/wireless/ath/ath11k/wmi.c |    8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

@@ -8383,15 +8383,19 @@ ath11k_wmi_pdev_temperature_event(struct
 	ath11k_dbg(ab, ATH11K_DBG_WMI, "event pdev temperature ev temp %d pdev_id %d\n",
 		   ev->temp, ev->pdev_id);
 
+	rcu_read_lock();
+
 	ar = ath11k_mac_get_ar_by_pdev_id(ab, ev->pdev_id);
 	if (!ar) {
 		ath11k_warn(ab, "invalid pdev id in pdev temperature ev %d", ev->pdev_id);
-		kfree(tb);
-		return;
+		goto exit;
 	}
 
 	ath11k_thermal_event_temperature(ar, ev->temp);
 
+exit:
+	rcu_read_unlock();
+
 	kfree(tb);
 }
 

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Johan Hovold <[email protected]>

commit 3b6c14833165f689cc5928574ebafe52bbce5f1e upstream.

The ath11k active pdevs are protected by RCU but the DFS radar event handling code calling ath11k_mac_get_ar_by_pdev_id() was not marked as a read-side critical section.

Mark the code in question as an RCU read-side critical section to avoid any potential use-after-free issues.

Compile tested only.

---
 drivers/net/wireless/ath/ath11k/wmi.c |    4 ++++
 1 file changed, 4 insertions(+)

@@ -8337,6 +8337,8 @@ ath11k_wmi_pdev_dfs_radar_detected_event
 		   ev->detector_id, ev->segment_id, ev->timestamp, ev->is_chirp,
 		   ev->freq_offset, ev->sidx);
 
+	rcu_read_lock();
+
 	ar = ath11k_mac_get_ar_by_pdev_id(ab, ev->pdev_id);
 
 	if (!ar) {

@@ -8354,6 +8356,8 @@ ath11k_wmi_pdev_dfs_radar_detected_event
 		ieee80211_radar_detected(ar->hw);
 
 exit:
+	rcu_read_unlock();
+
 	kfree(tb);
 }
 

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Johan Hovold <[email protected]>

commit 3f77c7d605b29df277d77e9ee75d96e7ad145d2d upstream.

The ath11k active pdevs are protected by RCU but the htt pktlog handling code calling ath11k_mac_get_ar_by_pdev_id() was not marked as a read-side critical section.

Mark the code in question as an RCU read-side critical section to avoid any potential use-after-free issues.

Compile tested only.

---
 drivers/net/wireless/ath/ath11k/dp_rx.c |    8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

@@ -1621,14 +1621,20 @@ static void ath11k_htt_pktlog(struct ath
 	u8 pdev_id;
 
 	pdev_id = FIELD_GET(HTT_T2H_PPDU_STATS_INFO_PDEV_ID, data->hdr);
+
+	rcu_read_lock();
+
 	ar = ath11k_mac_get_ar_by_pdev_id(ab, pdev_id);
 	if (!ar) {
 		ath11k_warn(ab, "invalid pdev id %d on htt pktlog\n", pdev_id);
-		return;
+		goto out;
 	}
 
 	trace_ath11k_htt_pktlog(ar, data->payload, hdr->size,
 				ar->ab->pktlog_defs_checksum);
+
+out:
+	rcu_read_unlock();
 }
 
 static void ath11k_htt_backpressure_event_handler(struct ath11k_base *ab,

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Johan Hovold <[email protected]>

commit 1dea3c0720a146bd7193969f2847ccfed5be2221 upstream.

The ath11k active pdevs are protected by RCU but the gtk offload status event handling code calling ath11k_mac_get_arvif_by_vdev_id() was not marked as a read-side critical section.

Mark the code in question as an RCU read-side critical section to avoid any potential use-after-free issues.

Compile tested only.

---
 drivers/net/wireless/ath/ath11k/wmi.c |    7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

@@ -8619,12 +8619,13 @@ static void ath11k_wmi_gtk_offload_statu
 		return;
 	}
 
+	rcu_read_lock();
+
 	arvif = ath11k_mac_get_arvif_by_vdev_id(ab, ev->vdev_id);
 	if (!arvif) {
 		ath11k_warn(ab, "failed to get arvif for vdev_id:%d\n",
 			    ev->vdev_id);
-		kfree(tb);
-		return;
+		goto exit;
 	}
 
 	ath11k_dbg(ab, ATH11K_DBG_WMI, "event gtk offload refresh_cnt %d\n",

@@ -8641,6 +8642,8 @@ static void ath11k_wmi_gtk_offload_statu
 
 	ieee80211_gtk_rekey_notify(arvif->vif, arvif->bssid,
 				   (void *)&replay_ctr_be, GFP_ATOMIC);
+exit:
+	rcu_read_unlock();
 
 	kfree(tb);
 }

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Johan Hovold <[email protected]>

commit 6afc57ea315e0f660b1f870a681737bb7b71faef upstream.

The ath12k active pdevs are protected by RCU but the htt mlo-offset event handling code calling ath12k_mac_get_ar_by_pdev_id() was not marked as a read-side critical section.

Mark the code in question as an RCU read-side critical section to avoid any potential use-after-free issues.

Compile tested only.

---
 drivers/net/wireless/ath/ath12k/dp_rx.c |    7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

@@ -1658,11 +1658,12 @@ static void ath12k_htt_mlo_offset_event_
 	msg = (struct ath12k_htt_mlo_offset_msg *)skb->data;
 	pdev_id = u32_get_bits(__le32_to_cpu(msg->info),
 			       HTT_T2H_MLO_OFFSET_INFO_PDEV_ID);
-	ar = ath12k_mac_get_ar_by_pdev_id(ab, pdev_id);
 
+	rcu_read_lock();
+	ar = ath12k_mac_get_ar_by_pdev_id(ab, pdev_id);
 	if (!ar) {
 		ath12k_warn(ab, "invalid pdev id %d on htt mlo offset\n", pdev_id);
-		return;
+		goto exit;
 	}
 
 	spin_lock_bh(&ar->data_lock);

@@ -1678,6 +1679,8 @@ static void ath12k_htt_mlo_offset_event_
 	pdev->timestamp.mlo_comp_timer = __le32_to_cpu(msg->mlo_comp_timer);
 
 	spin_unlock_bh(&ar->data_lock);
+exit:
+	rcu_read_unlock();
 }
 
 void ath12k_dp_htt_htc_t2h_msg_handler(struct ath12k_base *ab,

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Johan Hovold <[email protected]>

commit 69bd216e049349886405b1c87a55dce3d35d1ba7 upstream.

The ath12k active pdevs are protected by RCU but the DFS-radar and temperature event handling code calling ath12k_mac_get_ar_by_pdev_id() was not marked as a read-side critical section.

Mark the code in question as RCU read-side critical sections to avoid any potential use-after-free issues.

Note that the temperature event handler looks like a place holder currently but would still trigger an RCU lockdep splat.

Compile tested only.

---
 drivers/net/wireless/ath/ath12k/wmi.c |   11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)

@@ -6234,6 +6234,8 @@ ath12k_wmi_pdev_dfs_radar_detected_event
 		   ev->detector_id, ev->segment_id, ev->timestamp, ev->is_chirp,
 		   ev->freq_offset, ev->sidx);
 
+	rcu_read_lock();
+
 	ar = ath12k_mac_get_ar_by_pdev_id(ab, le32_to_cpu(ev->pdev_id));
 
 	if (!ar) {

@@ -6251,6 +6253,8 @@ ath12k_wmi_pdev_dfs_radar_detected_event
 		ieee80211_radar_detected(ar->hw);
 
 exit:
+	rcu_read_unlock();
+
 	kfree(tb);
 }
 

@@ -6269,11 +6273,16 @@ ath12k_wmi_pdev_temperature_event(struct
 	ath12k_dbg(ab, ATH12K_DBG_WMI,
 		   "pdev temperature ev temp %d pdev_id %d\n", ev.temp, ev.pdev_id);
 
+	rcu_read_lock();
+
 	ar = ath12k_mac_get_ar_by_pdev_id(ab, le32_to_cpu(ev.pdev_id));
 	if (!ar) {
 		ath12k_warn(ab, "invalid pdev id in pdev temperature ev %d", ev.pdev_id);
-		return;
+		goto exit;
 	}
+
+exit:
+	rcu_read_unlock();
 }
 
 static void ath12k_fils_discovery_event(struct ath12k_base *ab,

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Rong Chen <[email protected]>

commit 57925e16c9f7d18012bcf45bfa658f92c087981a upstream.

For the t7 and older SoC families, the CMD_CFG_ERROR has no effect. Starting from SoC family C3, setting this bit without SG LINK data address will cause the controller to generate an IRQ and stop working.

To fix it, don't set the bit CMD_CFG_ERROR anymore.

---
 drivers/mmc/host/meson-gx-mmc.c |    1 -
 1 file changed, 1 deletion(-)

@@ -801,7 +801,6 @@ static void meson_mmc_start_cmd(struct m
 
 	cmd_cfg |= FIELD_PREP(CMD_CFG_CMD_INDEX_MASK, cmd->opcode);
 	cmd_cfg |= CMD_CFG_OWNER;  /* owned by CPU */
-	cmd_cfg |= CMD_CFG_ERROR; /* stop in case of error */
 
 	meson_mmc_set_response_bits(cmd, &cmd_cfg);
 

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Herve Codina <[email protected]>

commit 5e7afb2eb7b2a7c81e9f608cbdf74a07606fd1b5 upstream.

irq_remove_generic_chip() calculates the Linux interrupt number for removing the handler and interrupt chip based on gc::irq_base as a linear function of the bit positions of set bits in the @msk argument.

When the generic chip is present in an irq domain, i.e. created with a call to irq_alloc_domain_generic_chips(), gc::irq_base contains not the base Linux interrupt number. It contains the base hardware interrupt for this chip. It is set to 0 for the first chip in the domain, 0 + N for the next chip, where $N is the number of hardware interrupts per chip.

That means the Linux interrupt number cannot be calculated based on gc::irq_base for irqdomain based chips without a domain map lookup, which is currently missing.

Rework the code to take the irqdomain case into account and calculate the Linux interrupt number by a irqdomain lookup of the domain specific hardware interrupt number.

[ tglx: Massage changelog. Reshuffle the logic and add a proper comment. ]

---
 kernel/irq/generic-chip.c |   25 +++++++++++++++++++------
 1 file changed, 19 insertions(+), 6 deletions(-)

@@ -544,21 +544,34 @@ EXPORT_SYMBOL_GPL(irq_setup_alt_chip);
 void irq_remove_generic_chip(struct irq_chip_generic *gc, u32 msk,
 			     unsigned int clr, unsigned int set)
 {
-	unsigned int i = gc->irq_base;
+	unsigned int i, virq;
 
 	raw_spin_lock(&gc_lock);
 	list_del(&gc->list);
 	raw_spin_unlock(&gc_lock);
 
-	for (; msk; msk >>= 1, i++) {
+	for (i = 0; msk; msk >>= 1, i++) {
 		if (!(msk & 0x01))
 			continue;
 
+		/*
+		 * Interrupt domain based chips store the base hardware
+		 * interrupt number in gc::irq_base. Otherwise gc::irq_base
+		 * contains the base Linux interrupt number.
+		 */
+		if (gc->domain) {
+			virq = irq_find_mapping(gc->domain, gc->irq_base + i);
+			if (!virq)
+				continue;
+		} else {
+			virq = gc->irq_base + i;
+		}
+
 		/* Remove handler first. That will mask the irq line */
-		irq_set_handler(i, NULL);
-		irq_set_chip(i, &no_irq_chip);
-		irq_set_chip_data(i, NULL);
-		irq_modify_status(i, clr, set);
+		irq_set_handler(virq, NULL);
+		irq_set_chip(virq, &no_irq_chip);
+		irq_set_chip_data(virq, NULL);
+		irq_modify_status(virq, clr, set);
 	}
 }
 EXPORT_SYMBOL_GPL(irq_remove_generic_chip);

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Hao Jia <[email protected]>

commit 5ebde09d91707a4a9bec1e3d213e3c12ffde348f upstream.

Igor Raits and Bagas Sanjaya report a RQCF_ACT_SKIP leak warning.

This warning may be triggered in the following situations:

CPU0 CPU1

__schedule()
*rq->clock_update_flags <<= 1;* unregister_fair_sched_group()

  pick_next_task_fair+0x4a/0x410      destroy_cfs_bandwidth()
    newidle_balance+0x115/0x3e0       for_each_possible_cpu(i) *i=0*
      rq_unpin_lock(this_rq, rf)      __cfsb_csd_unthrottle()
      raw_spin_rq_unlock(this_rq)
                                      rq_lock(*CPU0_rq*, &rf)
                                      rq_clock_start_loop_update()
                                      rq->clock_update_flags & RQCF_ACT_SKIP <--
      raw_spin_rq_lock(this_rq)

The purpose of RQCF_ACT_SKIP is to skip the update rq clock, but the update is very early in __schedule(), but we clear RQCF_*_SKIP very late, causing it to span that gap above and triggering this warning.

In __schedule() we can clear the RQCF_*_SKIP flag immediately after update_rq_clock() to avoid this RQCF_ACT_SKIP leak warning. And set rq->clock_update_flags to RQCF_UPDATED to avoid rq->clock_update_flags < RQCF_ACT_SKIP warning that may be triggered later.

---
 kernel/sched/core.c |    5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)

@@ -5377,8 +5377,6 @@ context_switch(struct rq *rq, struct tas
 	/* switch_mm_cid() requires the memory barriers above. */
 	switch_mm_cid(rq, prev, next);
 
-	rq->clock_update_flags &= ~(RQCF_ACT_SKIP|RQCF_REQ_SKIP);
-
 	prepare_lock_switch(rq, next, rf);
 
 	/* Here we just switch the register state and the stack. */

@@ -6634,6 +6632,7 @@ static void __sched notrace __schedule(u
 	/* Promote REQ to ACT */
 	rq->clock_update_flags <<= 1;
 	update_rq_clock(rq);
+	rq->clock_update_flags = RQCF_UPDATED;
 
 	switch_count = &prev->nivcsw;
 

@@ -6713,8 +6712,6 @@ static void __sched notrace __schedule(u
 		/* Also unlocks the rq: */
 		rq = context_switch(rq, prev, next, &rf);
 	} else {
-		rq->clock_update_flags &= ~(RQCF_ACT_SKIP|RQCF_REQ_SKIP);
-
 		rq_unpin_lock(rq, &rf);
 		__balance_callbacks(rq);
 		raw_spin_rq_unlock_irq(rq);

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Sumit Garg <[email protected]>

commit c745cd1718b7825d69315fe7127e2e289e617598 upstream.

The OP-TEE driver using the old SMC based ABI permits overlapping shared buffers, but with the new FF-A based ABI each physical page may only be registered once.

As the key and blob buffer are allocated adjancently, there is no need for redundant register shared memory invocation. Also, it is incompatibile with FF-A based ABI limitation. So refactor register shared memory implementation to use only single invocation to register both key and blob buffers.

---
 security/keys/trusted-keys/trusted_tee.c |   64 +++++++++----------------------
 1 file changed, 20 insertions(+), 44 deletions(-)

@@ -65,24 +65,16 @@ static int trusted_tee_seal(struct trust
 	int ret;
 	struct tee_ioctl_invoke_arg inv_arg;
 	struct tee_param param[4];
-	struct tee_shm *reg_shm_in = NULL, *reg_shm_out = NULL;
+	struct tee_shm *reg_shm = NULL;
 
 	memset(&inv_arg, 0, sizeof(inv_arg));
 	memset(&param, 0, sizeof(param));
 
-	reg_shm_in = tee_shm_register_kernel_buf(pvt_data.ctx, p->key,
-						 p->key_len);
-	if (IS_ERR(reg_shm_in)) {
-		dev_err(pvt_data.dev, "key shm register failed\n");
-		return PTR_ERR(reg_shm_in);
-	}
-
-	reg_shm_out = tee_shm_register_kernel_buf(pvt_data.ctx, p->blob,
-						  sizeof(p->blob));
-	if (IS_ERR(reg_shm_out)) {
-		dev_err(pvt_data.dev, "blob shm register failed\n");
-		ret = PTR_ERR(reg_shm_out);
-		goto out;
+	reg_shm = tee_shm_register_kernel_buf(pvt_data.ctx, p->key,
+					      sizeof(p->key) + sizeof(p->blob));
+	if (IS_ERR(reg_shm)) {
+		dev_err(pvt_data.dev, "shm register failed\n");
+		return PTR_ERR(reg_shm);
 	}
 
 	inv_arg.func = TA_CMD_SEAL;

@@ -90,13 +82,13 @@ static int trusted_tee_seal(struct trust
 	inv_arg.num_params = 4;
 
 	param[0].attr = TEE_IOCTL_PARAM_ATTR_TYPE_MEMREF_INPUT;
-	param[0].u.memref.shm = reg_shm_in;
+	param[0].u.memref.shm = reg_shm;
 	param[0].u.memref.size = p->key_len;
 	param[0].u.memref.shm_offs = 0;
 	param[1].attr = TEE_IOCTL_PARAM_ATTR_TYPE_MEMREF_OUTPUT;
-	param[1].u.memref.shm = reg_shm_out;
+	param[1].u.memref.shm = reg_shm;
 	param[1].u.memref.size = sizeof(p->blob);
-	param[1].u.memref.shm_offs = 0;
+	param[1].u.memref.shm_offs = sizeof(p->key);
 
 	ret = tee_client_invoke_func(pvt_data.ctx, &inv_arg, param);
 	if ((ret < 0) || (inv_arg.ret != 0)) {

@@ -107,11 +99,7 @@ static int trusted_tee_seal(struct trust
 		p->blob_len = param[1].u.memref.size;
 	}
 
-out:
-	if (reg_shm_out)
-		tee_shm_free(reg_shm_out);
-	if (reg_shm_in)
-		tee_shm_free(reg_shm_in);
+	tee_shm_free(reg_shm);
 
 	return ret;
 }

@@ -124,24 +112,16 @@ static int trusted_tee_unseal(struct tru
 	int ret;
 	struct tee_ioctl_invoke_arg inv_arg;
 	struct tee_param param[4];
-	struct tee_shm *reg_shm_in = NULL, *reg_shm_out = NULL;
+	struct tee_shm *reg_shm = NULL;
 
 	memset(&inv_arg, 0, sizeof(inv_arg));
 	memset(&param, 0, sizeof(param));
 
-	reg_shm_in = tee_shm_register_kernel_buf(pvt_data.ctx, p->blob,
-						 p->blob_len);
-	if (IS_ERR(reg_shm_in)) {
-		dev_err(pvt_data.dev, "blob shm register failed\n");
-		return PTR_ERR(reg_shm_in);
-	}
-
-	reg_shm_out = tee_shm_register_kernel_buf(pvt_data.ctx, p->key,
-						  sizeof(p->key));
-	if (IS_ERR(reg_shm_out)) {
-		dev_err(pvt_data.dev, "key shm register failed\n");
-		ret = PTR_ERR(reg_shm_out);
-		goto out;
+	reg_shm = tee_shm_register_kernel_buf(pvt_data.ctx, p->key,
+					      sizeof(p->key) + sizeof(p->blob));
+	if (IS_ERR(reg_shm)) {
+		dev_err(pvt_data.dev, "shm register failed\n");
+		return PTR_ERR(reg_shm);
 	}
 
 	inv_arg.func = TA_CMD_UNSEAL;

@@ -149,11 +129,11 @@ static int trusted_tee_unseal(struct tru
 	inv_arg.num_params = 4;
 
 	param[0].attr = TEE_IOCTL_PARAM_ATTR_TYPE_MEMREF_INPUT;
-	param[0].u.memref.shm = reg_shm_in;
+	param[0].u.memref.shm = reg_shm;
 	param[0].u.memref.size = p->blob_len;
-	param[0].u.memref.shm_offs = 0;
+	param[0].u.memref.shm_offs = sizeof(p->key);
 	param[1].attr = TEE_IOCTL_PARAM_ATTR_TYPE_MEMREF_OUTPUT;
-	param[1].u.memref.shm = reg_shm_out;
+	param[1].u.memref.shm = reg_shm;
 	param[1].u.memref.size = sizeof(p->key);
 	param[1].u.memref.shm_offs = 0;
 

@@ -166,11 +146,7 @@ static int trusted_tee_unseal(struct tru
 		p->key_len = param[1].u.memref.size;
 	}
 
-out:
-	if (reg_shm_out)
-		tee_shm_free(reg_shm_out);
-	if (reg_shm_in)
-		tee_shm_free(reg_shm_in);
+	tee_shm_free(reg_shm);
 
 	return ret;
 }

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Jarkko Sakkinen <[email protected]>

commit 31de287345f41bbfaec36a5c8cbdba035cf76442 upstream.

Do bind neither static calls nor trusted_key_exit() before a successful init, in order to maintain a consistent state. In addition, depart the init_trusted() in the case of a real error (i.e. getting back something else than -ENODEV).

---
 security/keys/trusted-keys/trusted_core.c |   20 ++++++++++----------
 1 file changed, 10 insertions(+), 10 deletions(-)

@@ -358,17 +358,17 @@ static int __init init_trusted(void)
 		if (!get_random)
 			get_random = kernel_get_random;
 
-		static_call_update(trusted_key_seal,
-				   trusted_key_sources[i].ops->seal);
-		static_call_update(trusted_key_unseal,
-				   trusted_key_sources[i].ops->unseal);
-		static_call_update(trusted_key_get_random,
-				   get_random);
-		trusted_key_exit = trusted_key_sources[i].ops->exit;
-		migratable = trusted_key_sources[i].ops->migratable;
-
 		ret = trusted_key_sources[i].ops->init();
-		if (!ret)
+		if (!ret) {
+			static_call_update(trusted_key_seal, trusted_key_sources[i].ops->seal);
+			static_call_update(trusted_key_unseal, trusted_key_sources[i].ops->unseal);
+			static_call_update(trusted_key_get_random, get_random);
+
+			trusted_key_exit = trusted_key_sources[i].ops->exit;
+			migratable = trusted_key_sources[i].ops->migratable;
+		}
+
+		if (!ret || ret != -ENODEV)
 			break;
 	}
 

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Uwe Kleine-König <[email protected]>

commit 200bddbb3f5202bbce96444fdc416305de14f547 upstream.

With CONFIG_PCIE_KEYSTONE=y and ks_pcie_remove() marked with __exit, the function is discarded from the driver. In this case a bound device can still get unbound, e.g via sysfs. Then no cleanup code is run resulting in resource leaks or worse.

The right thing to do is do always have the remove callback available. Note that this driver cannot be compiled as a module, so ks_pcie_remove() was always discarded before this change and modpost couldn't warn about this issue. Furthermore the __ref annotation also prevents a warning.

---
 drivers/pci/controller/dwc/pci-keystone.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

@@ -1303,7 +1303,7 @@ err_link:
 	return ret;
 }
 
-static int __exit ks_pcie_remove(struct platform_device *pdev)
+static int ks_pcie_remove(struct platform_device *pdev)
 {
 	struct keystone_pcie *ks_pcie = platform_get_drvdata(pdev);
 	struct device_link **link = ks_pcie->link;

@@ -1321,7 +1321,7 @@ static int __exit ks_pcie_remove(struct
 
 static struct platform_driver ks_pcie_driver __refdata = {
 	.probe  = ks_pcie_probe,
-	.remove = __exit_p(ks_pcie_remove),
+	.remove = ks_pcie_remove,
 	.driver = {
 		.name	= "keystone-pcie",
 		.of_match_table = ks_pcie_of_match,

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Uwe Kleine-König <[email protected]>

commit 7994db905c0fd692cf04c527585f08a91b560144 upstream.

The __init annotation makes the ks_pcie_probe() function disappear after booting completes. However a device can also be bound later. In that case, we try to call ks_pcie_probe(), but the backing memory is likely already overwritten.

The right thing to do is do always have the probe callback available. Note that the (wrong) __refdata annotation prevented this issue to be noticed by modpost.

---
 drivers/pci/controller/dwc/pci-keystone.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

@@ -1101,7 +1101,7 @@ static const struct of_device_id ks_pcie
 	{ },
 };
 
-static int __init ks_pcie_probe(struct platform_device *pdev)
+static int ks_pcie_probe(struct platform_device *pdev)
 {
 	const struct dw_pcie_host_ops *host_ops;
 	const struct dw_pcie_ep_ops *ep_ops;

@@ -1319,7 +1319,7 @@ static int ks_pcie_remove(struct platfor
 	return 0;
 }
 
-static struct platform_driver ks_pcie_driver __refdata = {
+static struct platform_driver ks_pcie_driver = {
 	.probe  = ks_pcie_probe,
 	.remove = ks_pcie_remove,
 	.driver = {

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Nathan Chancellor <[email protected]>

commit 146a15b873353f8ac28dc281c139ff611a3c4848 upstream.

Prior to LLVM 15.0.0, LLVM's integrated assembler would incorrectly byte-swap NOP when compiling for big-endian, and the resulting series of bytes happened to match the encoding of FNMADD S21, S30, S0, S0.

This went unnoticed until commit:

34f66c4c4d5518c1 ("arm64: Use a positive cpucap for FP/SIMD")

Prior to that commit, the kernel would always enable the use of FPSIMD early in boot when __cpu_setup() initialized CPACR_EL1, and so usage of FNMADD within the kernel was not detected, but could result in the corruption of user or kernel FPSIMD state.

After that commit, the instructions happen to trap during boot prior to FPSIMD being detected and enabled, e.g.

| Unhandled 64-bit el1h sync exception on CPU0, ESR 0x000000001fe00000 -- ASIMD
| CPU: 0 PID: 0 Comm: swapper Not tainted 6.6.0-rc3-00013-g34f66c4c4d55 #1
| Hardware name: linux,dummy-virt (DT)
| pstate: 400000c9 (nZcv daIF -PAN -UAO -TCO -DIT -SSBS BTYPE=--)
| pc : __pi_strcmp+0x1c/0x150
| lr : populate_properties+0xe4/0x254
| sp : ffffd014173d3ad0
| x29: ffffd014173d3af0 x28: fffffbfffddffcb8 x27: 0000000000000000
| x26: 0000000000000058 x25: fffffbfffddfe054 x24: 0000000000000008
| x23: fffffbfffddfe000 x22: fffffbfffddfe000 x21: fffffbfffddfe044
| x20: ffffd014173d3b70 x19: 0000000000000001 x18: 0000000000000005
| x17: 0000000000000010 x16: 0000000000000000 x15: 00000000413e7000
| x14: 0000000000000000 x13: 0000000000001bcc x12: 0000000000000000
| x11: 00000000d00dfeed x10: ffffd414193f2cd0 x9 : 0000000000000000
| x8 : 0101010101010101 x7 : ffffffffffffffc0 x6 : 0000000000000000
| x5 : 0000000000000000 x4 : 0101010101010101 x3 : 000000000000002a
| x2 : 0000000000000001 x1 : ffffd014171f2988 x0 : fffffbfffddffcb8
| Kernel panic - not syncing: Unhandled exception
| CPU: 0 PID: 0 Comm: swapper Not tainted 6.6.0-rc3-00013-g34f66c4c4d55 #1
| Hardware name: linux,dummy-virt (DT)
| Call trace:
|  dump_backtrace+0xec/0x108
|  show_stack+0x18/0x2c
|  dump_stack_lvl+0x50/0x68
|  dump_stack+0x18/0x24
|  panic+0x13c/0x340
|  el1t_64_irq_handler+0x0/0x1c
|  el1_abort+0x0/0x5c
|  el1h_64_sync+0x64/0x68
|  __pi_strcmp+0x1c/0x150
|  unflatten_dt_nodes+0x1e8/0x2d8
|  __unflatten_device_tree+0x5c/0x15c
|  unflatten_device_tree+0x38/0x50
|  setup_arch+0x164/0x1e0
|  start_kernel+0x64/0x38c
|  __primary_switched+0xbc/0xc4

Restrict CONFIG_CPU_BIG_ENDIAN to a known good assembler, which is either GNU as or LLVM's IAS 15.0.0 and newer, which contains the linked commit.

---
 arch/arm64/Kconfig |    2 ++
 1 file changed, 2 insertions(+)

@@ -1369,6 +1369,8 @@ choice
 config CPU_BIG_ENDIAN
 	bool "Build big-endian kernel"
 	depends on !LD_IS_LLD || LLD_VERSION >= 130000
+	# https://github.com/llvm/llvm-project/commit/1379b150991f70a5782e9a143c2ba5308da1161c
+	depends on AS_IS_GNU || AS_VERSION >= 150000
 	help
 	  Say Y if you plan on running a kernel with a big-endian userspace.
 

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Maria Yu <[email protected]>

commit d35686444fc80950c731e33a2f6ad4a55822be9b upstream.

The counting of module PLTs has been broken when CONFIG_RANDOMIZE_BASE=n since commit:

3e35d303ab7d22c4 ("arm64: module: rework module VA range selection")

Prior to that commit, when CONFIG_RANDOMIZE_BASE=n, the kernel image and all modules were placed within a 128M region, and no PLTs were necessary for B or BL. Hence count_plts() and partition_branch_plt_relas() skipped handling B and BL when CONFIG_RANDOMIZE_BASE=n.

After that commit, modules can be placed anywhere within a 2G window regardless of CONFIG_RANDOMIZE_BASE, and hence PLTs may be necessary for B and BL even when CONFIG_RANDOMIZE_BASE=n. Unfortunately that commit failed to update count_plts() and partition_branch_plt_relas() accordingly.

Due to this, module_emit_plt_entry() may fail if an insufficient number of PLT entries have been reserved, resulting in modules failing to load with -ENOEXEC.

Fix this by counting PLTs regardless of CONFIG_RANDOMIZE_BASE in count_plts() and partition_branch_plt_relas().

---
 arch/arm64/kernel/module-plts.c | 6 ------
 1 file changed, 6 deletions(-)

index bd69a4e7cd60..79200f21e123 100644
@@ -167,9 +167,6 @@ static unsigned int count_plts(Elf64_Sym *syms, Elf64_Rela *rela, int num,
 		switch (ELF64_R_TYPE(rela[i].r_info)) {
 		case R_AARCH64_JUMP26:
 		case R_AARCH64_CALL26:
-			if (!IS_ENABLED(CONFIG_RANDOMIZE_BASE))
-				break;
-
 			/*
 			 * We only have to consider branch targets that resolve
 			 * to symbols that are defined in a different section.

@@ -269,9 +266,6 @@ static int partition_branch_plt_relas(Elf64_Sym *syms, Elf64_Rela *rela,
 {
 	int i = 0, j = numrels - 1;
 
-	if (!IS_ENABLED(CONFIG_RANDOMIZE_BASE))
-		return 0;
-
 	while (i < j) {
 		if (branch_rela_needs_plt(syms, &rela[i], dstidx))
 			i++;

2.43.0

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Helge Deller <[email protected]>

commit 86bb854d134f4429feb35d2e05f55c6e036770d2 upstream.

The PDIR table of the System Bus Adapter (SBA) I/O MMU uses 64-bit little-endian pointers.

---
 drivers/char/agp/parisc-agp.c | 16 +++++++++-------
 1 file changed, 9 insertions(+), 7 deletions(-)

index c6f181702b9a..edbc4d338117 100644
@@ -38,7 +38,7 @@ static struct _parisc_agp_info {
 
 	int lba_cap_offset;
 
-	u64 *gatt;
+	__le64 *gatt;
 	u64 gatt_entries;
 
 	u64 gart_base;

@@ -104,7 +104,7 @@ parisc_agp_create_gatt_table(struct agp_bridge_data *bridge)
 	int i;
 
 	for (i = 0; i < info->gatt_entries; i++) {
-		info->gatt[i] = (unsigned long)agp_bridge->scratch_page;
+		info->gatt[i] = cpu_to_le64(agp_bridge->scratch_page);
 	}
 
 	return 0;

@@ -158,9 +158,9 @@ parisc_agp_insert_memory(struct agp_memory *mem, off_t pg_start, int type)
 		for (k = 0;
 		     k < info->io_pages_per_kpage;
 		     k++, j++, paddr += info->io_page_size) {
-			info->gatt[j] =
+			info->gatt[j] = cpu_to_le64(
 				parisc_agp_mask_memory(agp_bridge,
-					paddr, type);
+					paddr, type));
 			asm_io_fdc(&info->gatt[j]);
 		}
 	}

@@ -184,7 +184,7 @@ parisc_agp_remove_memory(struct agp_memory *mem, off_t pg_start, int type)
 	io_pg_start = info->io_pages_per_kpage * pg_start;
 	io_pg_count = info->io_pages_per_kpage * mem->page_count;
 	for (i = io_pg_start; i < io_pg_count + io_pg_start; i++) {
-		info->gatt[i] = agp_bridge->scratch_page;
+		info->gatt[i] = cpu_to_le64(agp_bridge->scratch_page);
 	}
 
 	agp_bridge->driver->tlb_flush(mem);

@@ -204,7 +204,8 @@ parisc_agp_mask_memory(struct agp_bridge_data *bridge, dma_addr_t addr,
 	pa |= (ci >> PAGE_SHIFT) & 0xff;/* move CI (8 bits) into lowest byte */
 	pa |= SBA_PDIR_VALID_BIT;	/* set "valid" bit */
 
-	return cpu_to_le64(pa);
+	/* return native (big-endian) PDIR entry */
+	return pa;
 }
 
 static void

@@ -251,7 +252,8 @@ static int __init
 agp_ioc_init(void __iomem *ioc_regs)
 {
 	struct _parisc_agp_info *info = &parisc_agp_info;
-        u64 iova_base, *io_pdir, io_tlb_ps;
+        u64 iova_base, io_tlb_ps;
+	__le64 *io_pdir;
         int io_tlb_shift;
 
         printk(KERN_INFO DRVPFX "IO PDIR shared with sba_iommu\n");

2.43.0

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Helge Deller <[email protected]>

commit 6240553b52c475d9fc9674de0521b77e692f3764 upstream.

PDC2.0 specifies the additional PSW-bit field.

---
 arch/parisc/include/uapi/asm/pdc.h |    1 +
 1 file changed, 1 insertion(+)

@@ -472,6 +472,7 @@ struct pdc_model {		/* for PDC_MODEL */
 	unsigned long arch_rev;
 	unsigned long pot_key;
 	unsigned long curr_key;
+	unsigned long width;	/* default of PSW_W bit (1=enabled) */
 };
 
 struct pdc_cache_cf {		/* for PDC_CACHE  (I/D-caches) */

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Helge Deller <[email protected]>

commit d0c219472980d15f5cbc5c8aec736848bda3f235 upstream.

---
 drivers/parisc/power.c |   16 +++++++++++++++-
 1 file changed, 15 insertions(+), 1 deletion(-)

@@ -197,6 +197,14 @@ static struct notifier_block parisc_pani
 	.priority	= INT_MAX,
 };
 
+/* qemu soft power-off function */
+static int qemu_power_off(struct sys_off_data *data)
+{
+	/* this turns the system off via SeaBIOS */
+	*(int *)data->cb_data = 0;
+	pdc_soft_power_button(1);
+	return NOTIFY_DONE;
+}
 
 static int __init power_init(void)
 {

@@ -226,7 +234,13 @@ static int __init power_init(void)
 				soft_power_reg);
 	}
 
-	power_task = kthread_run(kpowerswd, (void*)soft_power_reg, KTHREAD_NAME);
+	power_task = NULL;
+	if (running_on_qemu && soft_power_reg)
+		register_sys_off_handler(SYS_OFF_MODE_POWER_OFF, SYS_OFF_PRIO_DEFAULT,
+					qemu_power_off, (void *)soft_power_reg);
+	else
+		power_task = kthread_run(kpowerswd, (void*)soft_power_reg,
+					KTHREAD_NAME);
 	if (IS_ERR(power_task)) {
 		printk(KERN_ERR DRIVER_NAME ": thread creation failed.  Driver not loaded.\n");
 		pdc_soft_power_button(0);

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Christian Marangi <[email protected]>

commit ea167a7fc2426f7685c3735e104921c1a20a6d3f upstream.

Commit 3c0897c180c6 ("cpufreq: Use scnprintf() for avoiding potential buffer overflow") switched from snprintf to the more secure scnprintf but never updated the exit condition for PAGE_SIZE.

As the commit say and as scnprintf document, what scnprintf returns what is actually written not counting the '\0' end char. This results in the case of len exceeding the size, len set to PAGE_SIZE - 1, as it can be written at max PAGE_SIZE - 1 (as '\0' is not counted)

Because of len is never set to PAGE_SIZE, the function never break early, never prints the warning and never return -EFBIG.

Fix this by changing the condition to PAGE_SIZE - 1 to correctly trigger the error.

---
 drivers/cpufreq/cpufreq_stats.c |   14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)

@@ -131,23 +131,23 @@ static ssize_t show_trans_table(struct c
 	len += sysfs_emit_at(buf, len, "   From  :    To\n");
 	len += sysfs_emit_at(buf, len, "         : ");
 	for (i = 0; i < stats->state_num; i++) {
-		if (len >= PAGE_SIZE)
+		if (len >= PAGE_SIZE - 1)
 			break;
 		len += sysfs_emit_at(buf, len, "%9u ", stats->freq_table[i]);
 	}
-	if (len >= PAGE_SIZE)
-		return PAGE_SIZE;
+	if (len >= PAGE_SIZE - 1)
+		return PAGE_SIZE - 1;
 
 	len += sysfs_emit_at(buf, len, "\n");
 
 	for (i = 0; i < stats->state_num; i++) {
-		if (len >= PAGE_SIZE)
+		if (len >= PAGE_SIZE - 1)
 			break;
 
 		len += sysfs_emit_at(buf, len, "%9u: ", stats->freq_table[i]);
 
 		for (j = 0; j < stats->state_num; j++) {
-			if (len >= PAGE_SIZE)
+			if (len >= PAGE_SIZE - 1)
 				break;
 
 			if (pending)

@@ -157,12 +157,12 @@ static ssize_t show_trans_table(struct c
 
 			len += sysfs_emit_at(buf, len, "%9u ", count);
 		}
-		if (len >= PAGE_SIZE)
+		if (len >= PAGE_SIZE - 1)
 			break;
 		len += sysfs_emit_at(buf, len, "\n");
 	}
 
-	if (len >= PAGE_SIZE) {
+	if (len >= PAGE_SIZE - 1) {
 		pr_warn_once("cpufreq transition table exceeds PAGE_SIZE. Disabling\n");
 		return -EFBIG;
 	}

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Ville Syrjälä <[email protected]>

commit a60ec4485f1c72dfece365cf95e6de82bdd74300 upstream.

Before the refactoring the pr_warn() only triggered when someone explicitly tried to write to a BIOS locked limit. After the refactoring the warning is also triggering during system resume. The user can't do anything about this so printing scary warnings doesn't make sense

Keep the printk but make it pr_debug() instead of pr_warn() to make it clear it's not a serious issue.

---
 drivers/powercap/intel_rapl_common.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index 40a2cc649c79..2feed036c1cd 100644
@@ -892,7 +892,7 @@ static int rapl_write_pl_data(struct rapl_domain *rd, int pl,
 		return -EINVAL;
 
 	if (rd->rpl[pl].locked) {
-		pr_warn("%s:%s:%s locked by BIOS\n", rd->rp->name, rd->name, pl_names[pl]);
+		pr_debug("%s:%s:%s locked by BIOS\n", rd->rp->name, rd->name, pl_names[pl]);
 		return -EACCES;
 	}
 

2.43.0

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

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

commit d761bb01c85b22d5b44abe283eb89019693f6595 upstream.

struct clk_hw_onecell_data is a flexible structure, which means that it contains flexible-array member at the bottom, in this case array hws:

include/linux/clk-provider.h:
1380 struct clk_hw_onecell_data {
1381 unsigned int num;
1382 struct clk_hw *hws[] __counted_by(num);
1383 };

This could potentially lead to an overwrite of the objects following clk_data in struct stratix10_clock_data, in this case void __iomem *base; at run-time:

drivers/clk/socfpga/stratix10-clk.h:
9 struct stratix10_clock_data {
10 struct clk_hw_onecell_data clk_data;
11 void __iomem *base;
12 };

There are currently three different places where memory is allocated for struct stratix10_clock_data, including the flex-array hws in struct clk_hw_onecell_data:

drivers/clk/socfpga/clk-agilex.c:
469 clk_data = devm_kzalloc(dev, struct_size(clk_data, clk_data.hws,
470 num_clks), GFP_KERNEL);

drivers/clk/socfpga/clk-agilex.c:
509 clk_data = devm_kzalloc(dev, struct_size(clk_data, clk_data.hws,
510 num_clks), GFP_KERNEL);

drivers/clk/socfpga/clk-s10.c:
400 clk_data = devm_kzalloc(dev, struct_size(clk_data, clk_data.hws,
401 num_clks), GFP_KERNEL);

I'll use just one of them to describe the issue. See below.

Notice that a total of 440 bytes are allocated for flexible-array member hws at line 469:

include/dt-bindings/clock/agilex-clock.h:
70 #define AGILEX_NUM_CLKS 55

drivers/clk/socfpga/clk-agilex.c:
459 struct stratix10_clock_data *clk_data;
460 void __iomem *base;
...
466
467 num_clks = AGILEX_NUM_CLKS;
468
469 clk_data = devm_kzalloc(dev, struct_size(clk_data, clk_data.hws,
470 num_clks), GFP_KERNEL);

struct_size(clk_data, clk_data.hws, num_clks) above translates to
sizeof(struct stratix10_clock_data) + sizeof(struct clk_hw *) * 55 == 16 + 8 * 55 == 16 + 440

		    ^^^
		     |
	allocated bytes for flex-array `hws`

474 for (i = 0; i < num_clks; i++)
475 clk_data->clk_data.hws[i] = ERR_PTR(-ENOENT);
476
477 clk_data->base = base;

and then some data is written into both hws and base objects.

Fix this by placing the declaration of object clk_data at the end of struct stratix10_clock_data. Also, add a comment to make it clear that this object must always be last in the structure.

-Wflex-array-member-not-at-end is coming in GCC-14, and we are getting ready to enable it globally.

---
 drivers/clk/socfpga/stratix10-clk.h |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

@@ -7,8 +7,10 @@
 #define	__STRATIX10_CLK_H
 
 struct stratix10_clock_data {
-	struct clk_hw_onecell_data	clk_data;
 	void __iomem		*base;
+
+	/* Must be last */
+	struct clk_hw_onecell_data	clk_data;
 };
 
 struct stratix10_pll_clock {

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

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

commit 5ad1e217a2b23aa046b241183bd9452d259d70d0 upstream.

struct clk_hw_onecell_data is a flexible structure, which means that it contains flexible-array member at the bottom, in this case array hws:

include/linux/clk-provider.h:
1380 struct clk_hw_onecell_data {
1381 unsigned int num;
1382 struct clk_hw *hws[] __counted_by(num);
1383 };

This could potentially lead to an overwrite of the objects following clk_data in struct visconti_pll_provider, in this case struct device_node *node;, at run-time:

drivers/clk/visconti/pll.h:
16 struct visconti_pll_provider {

 17         void __iomem *reg_base;
 18         struct clk_hw_onecell_data clk_data;
 19         struct device_node *node;

20 };

Notice that a total of 56 bytes are allocated for flexible-array hws at line 328. See below:

include/dt-bindings/clock/toshiba,tmpv770x.h:
14 #define TMPV770X_NR_PLL 7

drivers/clk/visconti/pll-tmpv770x.c:
69 ctx = visconti_init_pll(np, reg_base, TMPV770X_NR_PLL);

drivers/clk/visconti/pll.c:
321 struct visconti_pll_provider * __init visconti_init_pll(struct device_node *np,
322 void __iomem *base,
323 unsigned long nr_plls)
324 {
325 struct visconti_pll_provider *ctx;
...
328 ctx = kzalloc(struct_size(ctx, clk_data.hws, nr_plls), GFP_KERNEL);

struct_size(ctx, clk_data.hws, nr_plls) above translates to sizeof(struct visconti_pll_provider) + sizeof(struct clk_hw *) * 7 == 24 + 8 * 7 == 24 + 56

		  ^^^^
		   |
	allocated bytes for flex array `hws`

$ pahole -C visconti_pll_provider drivers/clk/visconti/pll.o
struct visconti_pll_provider {
	void *                     reg_base;             /*     0     8 */
	struct clk_hw_onecell_data clk_data;             /*     8     8 */
	struct device_node *       node;                 /*    16     8 */

	/* size: 24, cachelines: 1, members: 3 */
	/* last cacheline: 24 bytes */
};

And then, after the allocation, some data is written into all members of struct visconti_pll_provider:

332 for (i = 0; i < nr_plls; ++i)
333 ctx->clk_data.hws[i] = ERR_PTR(-ENOENT);
334

335         ctx->node = np;
336         ctx->reg_base = base;
337         ctx->clk_data.num = nr_plls;

Fix all these by placing the declaration of object clk_data at the end of struct visconti_pll_provider. Also, add a comment to make it clear that this object must always be last in the structure, and prevent this bug from being introduced again in the future.

-Wflex-array-member-not-at-end is coming in GCC-14, and we are getting ready to enable it globally.

---
 drivers/clk/visconti/pll.h |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

@@ -15,8 +15,10 @@
 
 struct visconti_pll_provider {
 	void __iomem *reg_base;
-	struct clk_hw_onecell_data clk_data;
 	struct device_node *node;
+
+	/* Must be last */
+	struct clk_hw_onecell_data clk_data;
 };
 
 #define VISCONTI_PLL_RATE(_rate, _dacen, _dsmen, \

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Kathiravan Thirumoorthy <[email protected]>

commit e641a070137dd959932c7c222e000d9d941167a2 upstream.

GPLL, NSS crypto PLL clock rates are fixed and shouldn't be scaled based on the request from dependent clocks. Doing so will result in the unexpected behaviour. So drop the CLK_SET_RATE_PARENT flag from the PLL clocks.

---
 drivers/clk/qcom/gcc-ipq8074.c |    6 ------
 1 file changed, 6 deletions(-)

@@ -76,7 +76,6 @@ static struct clk_fixed_factor gpll0_out
 				&gpll0_main.clkr.hw },
 		.num_parents = 1,
 		.ops = &clk_fixed_factor_ops,
-		.flags = CLK_SET_RATE_PARENT,
 	},
 };
 

@@ -122,7 +121,6 @@ static struct clk_alpha_pll_postdiv gpll
 				&gpll2_main.clkr.hw },
 		.num_parents = 1,
 		.ops = &clk_alpha_pll_postdiv_ro_ops,
-		.flags = CLK_SET_RATE_PARENT,
 	},
 };
 

@@ -155,7 +153,6 @@ static struct clk_alpha_pll_postdiv gpll
 				&gpll4_main.clkr.hw },
 		.num_parents = 1,
 		.ops = &clk_alpha_pll_postdiv_ro_ops,
-		.flags = CLK_SET_RATE_PARENT,
 	},
 };
 

@@ -189,7 +186,6 @@ static struct clk_alpha_pll_postdiv gpll
 				&gpll6_main.clkr.hw },
 		.num_parents = 1,
 		.ops = &clk_alpha_pll_postdiv_ro_ops,
-		.flags = CLK_SET_RATE_PARENT,
 	},
 };
 

@@ -202,7 +198,6 @@ static struct clk_fixed_factor gpll6_out
 				&gpll6_main.clkr.hw },
 		.num_parents = 1,
 		.ops = &clk_fixed_factor_ops,
-		.flags = CLK_SET_RATE_PARENT,
 	},
 };
 

@@ -267,7 +262,6 @@ static struct clk_alpha_pll_postdiv nss_
 				&nss_crypto_pll_main.clkr.hw },
 		.num_parents = 1,
 		.ops = &clk_alpha_pll_postdiv_ro_ops,
-		.flags = CLK_SET_RATE_PARENT,
 	},
 };
 

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Kathiravan Thirumoorthy <[email protected]>

commit 99cd4935cb972d0aafb16838bb2aeadbcaf196ce upstream.

GPLL, NSS crypto PLL clock rates are fixed and shouldn't be scaled based on the request from dependent clocks. Doing so will result in the unexpected behaviour. So drop the CLK_SET_RATE_PARENT flag from the PLL clocks.

---
 drivers/clk/qcom/gcc-ipq6018.c |    6 ------
 1 file changed, 6 deletions(-)

@@ -73,7 +73,6 @@ static struct clk_fixed_factor gpll0_out
 				&gpll0_main.clkr.hw },
 		.num_parents = 1,
 		.ops = &clk_fixed_factor_ops,
-		.flags = CLK_SET_RATE_PARENT,
 	},
 };
 

@@ -87,7 +86,6 @@ static struct clk_alpha_pll_postdiv gpll
 				&gpll0_main.clkr.hw },
 		.num_parents = 1,
 		.ops = &clk_alpha_pll_postdiv_ro_ops,
-		.flags = CLK_SET_RATE_PARENT,
 	},
 };
 

@@ -162,7 +160,6 @@ static struct clk_alpha_pll_postdiv gpll
 				&gpll6_main.clkr.hw },
 		.num_parents = 1,
 		.ops = &clk_alpha_pll_postdiv_ro_ops,
-		.flags = CLK_SET_RATE_PARENT,
 	},
 };
 

@@ -193,7 +190,6 @@ static struct clk_alpha_pll_postdiv gpll
 				&gpll4_main.clkr.hw },
 		.num_parents = 1,
 		.ops = &clk_alpha_pll_postdiv_ro_ops,
-		.flags = CLK_SET_RATE_PARENT,
 	},
 };
 

@@ -244,7 +240,6 @@ static struct clk_alpha_pll_postdiv gpll
 				&gpll2_main.clkr.hw },
 		.num_parents = 1,
 		.ops = &clk_alpha_pll_postdiv_ro_ops,
-		.flags = CLK_SET_RATE_PARENT,
 	},
 };
 

@@ -275,7 +270,6 @@ static struct clk_alpha_pll_postdiv nss_
 				&nss_crypto_pll_main.clkr.hw },
 		.num_parents = 1,
 		.ops = &clk_alpha_pll_postdiv_ro_ops,
-		.flags = CLK_SET_RATE_PARENT,
 	},
 };
 

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Marios Makassikis <[email protected]>

commit 807252f028c59b9a3bac4d62ad84761548c10f11 upstream.

Running smb2.rename test from Samba smbtorture suite against a kernel built with lockdep triggers a "possible recursive locking detected" warning.

This is because mnt_want_write() is called twice with no mnt_drop_write() in between:
-> ksmbd_vfs_mkdir()

    -> ksmbd_vfs_kern_path_create()
       -> kern_path_create()
          -> filename_create()
            -> mnt_want_write()
       -> mnt_want_write()

Fix this by removing the mnt_want_write/mnt_drop_write calls from vfs helpers that call kern_path_create().

Full lockdep trace below:

============================================
WARNING: possible recursive locking detected
6.6.0-rc5 #775 Not tainted
--------------------------------------------
kworker/1:1/32 is trying to acquire lock:
ffff888005ac83f8 (sb_writers#5){.+.+}-{0:0}, at: ksmbd_vfs_mkdir+0xe1/0x410

but task is already holding lock:
ffff888005ac83f8 (sb_writers#5){.+.+}-{0:0}, at: filename_create+0xb6/0x260

other info that might help us debug this:

 Possible unsafe locking scenario:

       CPU0
       ----

lock(sb_writers#5);
lock(sb_writers#5);

*** DEADLOCK ***

May be due to missing lock nesting notation

4 locks held by kworker/1:1/32:
#0: ffff8880064e4138 ((wq_completion)ksmbd-io){+.+.}-{0:0}, at: process_one_work+0x40e/0x980 #1: ffff888005b0fdd0 ((work_completion)(&work->work)){+.+.}-{0:0}, at: process_one_work+0x40e/0x980 #2: ffff888005ac83f8 (sb_writers#5){.+.+}-{0:0}, at: filename_create+0xb6/0x260
#3: ffff8880057ce760 (&type->i_mutex_dir_key#3/1){+.+.}-{3:3}, at: filename_create+0x123/0x260

---
 fs/smb/server/vfs.c |   23 +++--------------------
 1 file changed, 3 insertions(+), 20 deletions(-)

@@ -173,10 +173,6 @@ int ksmbd_vfs_create(struct ksmbd_work *
 		return err;
 	}
 
-	err = mnt_want_write(path.mnt);
-	if (err)
-		goto out_err;
-
 	mode |= S_IFREG;
 	err = vfs_create(mnt_idmap(path.mnt), d_inode(path.dentry),
 			 dentry, mode, true);

@@ -186,9 +182,7 @@ int ksmbd_vfs_create(struct ksmbd_work *
 	} else {
 		pr_err("File(%s): creation failed (err:%d)\n", name, err);
 	}
-	mnt_drop_write(path.mnt);
 
-out_err:
 	done_path_create(&path, dentry);
 	return err;
 }

@@ -219,10 +213,6 @@ int ksmbd_vfs_mkdir(struct ksmbd_work *w
 		return err;
 	}
 
-	err = mnt_want_write(path.mnt);
-	if (err)
-		goto out_err2;
-
 	idmap = mnt_idmap(path.mnt);
 	mode |= S_IFDIR;
 	err = vfs_mkdir(idmap, d_inode(path.dentry), dentry, mode);

@@ -233,21 +223,19 @@ int ksmbd_vfs_mkdir(struct ksmbd_work *w
 			       dentry->d_name.len);
 		if (IS_ERR(d)) {
 			err = PTR_ERR(d);
-			goto out_err1;
+			goto out_err;
 		}
 		if (unlikely(d_is_negative(d))) {
 			dput(d);
 			err = -ENOENT;
-			goto out_err1;
+			goto out_err;
 		}
 
 		ksmbd_vfs_inherit_owner(work, d_inode(path.dentry), d_inode(d));
 		dput(d);
 	}
 
-out_err1:
-	mnt_drop_write(path.mnt);
-out_err2:
+out_err:
 	done_path_create(&path, dentry);
 	if (err)
 		pr_err("mkdir(%s): creation failed (err:%d)\n", name, err);

@@ -665,16 +653,11 @@ int ksmbd_vfs_link(struct ksmbd_work *wo
 		goto out3;
 	}
 
-	err = mnt_want_write(newpath.mnt);
-	if (err)
-		goto out3;
-
 	err = vfs_link(oldpath.dentry, mnt_idmap(newpath.mnt),
 		       d_inode(newpath.dentry),
 		       dentry, NULL);
 	if (err)
 		ksmbd_debug(VFS, "vfs_link failed err %d\n", err);
-	mnt_drop_write(newpath.mnt);
 
 out3:
 	done_path_create(&newpath, dentry);

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Namjae Jeon <[email protected]>

commit 5a5409d90bd05f87fe5623a749ccfbf3f7c7d400 upstream.

If set_smb1_rsp_status() is not implemented, It will cause NULL pointer dereferece error when client send malformed smb1 message. This patch add set_smb1_rsp_status() to ignore malformed smb1 message.

---
 fs/smb/server/smb_common.c |   11 +++++++++++
 1 file changed, 11 insertions(+)

@@ -372,11 +372,22 @@ static int smb1_allocate_rsp_buf(struct
 	return 0;
 }
 
+/**
+ * set_smb1_rsp_status() - set error type in smb response header
+ * @work:	smb work containing smb response header
+ * @err:	error code to set in response
+ */
+static void set_smb1_rsp_status(struct ksmbd_work *work, __le32 err)
+{
+	work->send_no_response = 1;
+}
+
 static struct smb_version_ops smb1_server_ops = {
 	.get_cmd_val = get_smb1_cmd_val,
 	.init_rsp_hdr = init_smb1_rsp_hdr,
 	.allocate_rsp_buf = smb1_allocate_rsp_buf,
 	.check_user_session = smb1_check_user_session,
+	.set_rsp_status = set_smb1_rsp_status,
 };
 
 static int smb1_negotiate(struct ksmbd_work *work)

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Namjae Jeon <[email protected]>

commit eebff19acaa35820cb09ce2ccb3d21bee2156ffb upstream.

slab out-of-bounds write is caused by that offsets is bigger than pntsd allocation size. This patch add the check to validate 3 offsets using allocation size.

---
 fs/smb/server/smbacl.c |   29 ++++++++++++++++++++++++++---
 1 file changed, 26 insertions(+), 3 deletions(-)

@@ -1107,6 +1107,7 @@ pass:
 		struct smb_acl *pdacl;
 		struct smb_sid *powner_sid = NULL, *pgroup_sid = NULL;
 		int powner_sid_size = 0, pgroup_sid_size = 0, pntsd_size;
+		int pntsd_alloc_size;
 
 		if (parent_pntsd->osidoffset) {
 			powner_sid = (struct smb_sid *)((char *)parent_pntsd +

@@ -1119,9 +1120,10 @@ pass:
 			pgroup_sid_size = 1 + 1 + 6 + (pgroup_sid->num_subauth * 4);
 		}
 
-		pntsd = kzalloc(sizeof(struct smb_ntsd) + powner_sid_size +
-				pgroup_sid_size + sizeof(struct smb_acl) +
-				nt_size, GFP_KERNEL);
+		pntsd_alloc_size = sizeof(struct smb_ntsd) + powner_sid_size +
+			pgroup_sid_size + sizeof(struct smb_acl) + nt_size;
+
+		pntsd = kzalloc(pntsd_alloc_size, GFP_KERNEL);
 		if (!pntsd) {
 			rc = -ENOMEM;
 			goto free_aces_base;

@@ -1136,6 +1138,27 @@ pass:
 		pntsd->gsidoffset = parent_pntsd->gsidoffset;
 		pntsd->dacloffset = parent_pntsd->dacloffset;
 
+		if ((u64)le32_to_cpu(pntsd->osidoffset) + powner_sid_size >
+		    pntsd_alloc_size) {
+			rc = -EINVAL;
+			kfree(pntsd);
+			goto free_aces_base;
+		}
+
+		if ((u64)le32_to_cpu(pntsd->gsidoffset) + pgroup_sid_size >
+		    pntsd_alloc_size) {
+			rc = -EINVAL;
+			kfree(pntsd);
+			goto free_aces_base;
+		}
+
+		if ((u64)le32_to_cpu(pntsd->dacloffset) + sizeof(struct smb_acl) + nt_size >
+		    pntsd_alloc_size) {
+			rc = -EINVAL;
+			kfree(pntsd);
+			goto free_aces_base;
+		}
+
 		if (pntsd->osidoffset) {
 			struct smb_sid *owner_sid = (struct smb_sid *)((char *)pntsd +
 					le32_to_cpu(pntsd->osidoffset));

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Dan Carpenter <[email protected]>

commit b44f9da81783fda72632ef9b0d05ea3f3ca447a5 upstream.

This error path should return -EINVAL instead of success.

---
 drivers/mmc/host/vub300.c |    1 +
 1 file changed, 1 insertion(+)

@@ -2309,6 +2309,7 @@ static int vub300_probe(struct usb_inter
 		vub300->read_only =
 			(0x0010 & vub300->system_port_status.port_flags) ? 1 : 0;
 	} else {
+		retval = -EINVAL;
 		goto error5;
 	}
 	usb_set_intfdata(interface, vub300);

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Nitin Yadav <[email protected]>

commit 71956d0cb56c1e5f9feeb4819db87a076418e930 upstream.

ti,otap-del-sel-legacy/ti,itap-del-sel-legacy passed from DT are currently ignored for all SD/MMC and eMMC modes. Fix this by making start loop index to MMC_TIMING_LEGACY.

---
 drivers/mmc/host/sdhci_am654.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -598,7 +598,7 @@ static int sdhci_am654_get_otap_delay(st
 		return 0;
 	}
 
-	for (i = MMC_TIMING_MMC_HS; i <= MMC_TIMING_MMC_HS400; i++) {
+	for (i = MMC_TIMING_LEGACY; i <= MMC_TIMING_MMC_HS400; i++) {
 
 		ret = device_property_read_u32(dev, td[i].otap_binding,
 					       &sdhci_am654->otap_del_sel[i]);

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Bean Huo <[email protected]>

commit ed9009ad300c0f15a3ecfe9613547b1962bde02c upstream.

Micron MTFC4GACAJCN eMMC supports cache but requires that flush cache operation be allowed only after a write has occurred. Otherwise, the cache flush command or subsequent commands will time out.

---
 drivers/mmc/core/block.c  |    4 +++-
 drivers/mmc/core/card.h   |    4 ++++
 drivers/mmc/core/mmc.c    |    8 ++++++--
 drivers/mmc/core/quirks.h |    7 ++++---
 include/linux/mmc/card.h  |    2 ++
 5 files changed, 19 insertions(+), 6 deletions(-)

@@ -2389,8 +2389,10 @@ enum mmc_issued mmc_blk_mq_issue_rq(stru
 			}
 			ret = mmc_blk_cqe_issue_flush(mq, req);
 			break;
-		case REQ_OP_READ:
 		case REQ_OP_WRITE:
+			card->written_flag = true;
+			fallthrough;
+		case REQ_OP_READ:
 			if (host->cqe_enabled)
 				ret = mmc_blk_cqe_issue_rw_rq(mq, req);
 			else

@@ -280,4 +280,8 @@ static inline int mmc_card_broken_sd_cac
 	return c->quirks & MMC_QUIRK_BROKEN_SD_CACHE;
 }
 
+static inline int mmc_card_broken_cache_flush(const struct mmc_card *c)
+{
+	return c->quirks & MMC_QUIRK_BROKEN_CACHE_FLUSH;
+}
 #endif

@@ -2081,13 +2081,17 @@ static int _mmc_flush_cache(struct mmc_h
 {
 	int err = 0;
 
+	if (mmc_card_broken_cache_flush(host->card) && !host->card->written_flag)
+		return 0;
+
 	if (_mmc_cache_enabled(host)) {
 		err = mmc_switch(host->card, EXT_CSD_CMD_SET_NORMAL,
 				 EXT_CSD_FLUSH_CACHE, 1,
 				 CACHE_FLUSH_TIMEOUT_MS);
 		if (err)
-			pr_err("%s: cache flush error %d\n",
-			       mmc_hostname(host), err);
+			pr_err("%s: cache flush error %d\n", mmc_hostname(host), err);
+		else
+			host->card->written_flag = false;
 	}
 
 	return err;

@@ -110,11 +110,12 @@ static const struct mmc_fixup __maybe_un
 		  MMC_QUIRK_TRIM_BROKEN),
 
 	/*
-	 * Micron MTFC4GACAJCN-1M advertises TRIM but it does not seems to
-	 * support being used to offload WRITE_ZEROES.
+	 * Micron MTFC4GACAJCN-1M supports TRIM but does not appear to support
+	 * WRITE_ZEROES offloading. It also supports caching, but the cache can
+	 * only be flushed after a write has occurred.
 	 */
 	MMC_FIXUP("Q2J54A", CID_MANFID_MICRON, 0x014e, add_quirk_mmc,
-		  MMC_QUIRK_TRIM_BROKEN),
+		  MMC_QUIRK_TRIM_BROKEN | MMC_QUIRK_BROKEN_CACHE_FLUSH),
 
 	/*
 	 * Kingston EMMC04G-M627 advertises TRIM but it does not seems to

@@ -295,7 +295,9 @@ struct mmc_card {
 #define MMC_QUIRK_BROKEN_HPI	(1<<13)		/* Disable broken HPI support */
 #define MMC_QUIRK_BROKEN_SD_DISCARD	(1<<14)	/* Disable broken SD discard support */
 #define MMC_QUIRK_BROKEN_SD_CACHE	(1<<15)	/* Disable broken SD cache support */
+#define MMC_QUIRK_BROKEN_CACHE_FLUSH	(1<<16)	/* Don't flush cache until the write has occurred */
 
+	bool			written_flag;	/* Indicates eMMC has been written since power on */
 	bool			reenable_cmdq;	/* Re-enable Command Queue */
 
 	unsigned int		erase_size;	/* erase size in sectors */

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Heiner Kallweit <[email protected]>

commit 8e37372ad0bea4c9b4712d9943f6ae96cff9491f upstream.

aspm_attr_store_common(), which handles sysfs control of ASPM, has the same problem as fb097dcd5a28 ("PCI/ASPM: Disable only ASPM_STATE_L1 when driver disables L1"): disabling L1 adds only ASPM_L1 (but not any of the L1.x substates) to the "aspm_disable" mask.

Enabling one substate, e.g., L1.1, via sysfs removes ASPM_L1 from the disable mask. Since disabling L1 via sysfs doesn't add any of the substates to the disable mask, enabling L1.1 actually enables *all* the substates.

In this scenario:

- Write 0 to "l1_aspm" to disable L1
- Write 1 to "l1_1_aspm" to enable L1.1

the intention is to disable L1 and all L1.x substates, then enable just L1.1, but in fact, *all* L1.x substates are enabled.

Fix this by explicitly disabling all the L1.x substates when disabling L1.

---
 drivers/pci/pcie/aspm.c |    2 ++
 1 file changed, 2 insertions(+)

@@ -1248,6 +1248,8 @@ static ssize_t aspm_attr_store_common(st
 			link->aspm_disable &= ~ASPM_STATE_L1;
 	} else {
 		link->aspm_disable |= state;
+		if (state & ASPM_STATE_L1)
+			link->aspm_disable |= ASPM_STATE_L1SS;
 	}
 
 	pcie_config_aspm_link(link, policy_to_aspm_state(link));

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Uwe Kleine-König <[email protected]>

commit 3064ef2e88c1629c1e67a77d7bc20020b35846f2 upstream.

With CONFIG_PCIE_KIRIN=y and kirin_pcie_remove() marked with __exit, the function is discarded from the driver. In this case a bound device can still get unbound, e.g via sysfs. Then no cleanup code is run resulting in resource leaks or worse.

The right thing to do is do always have the remove callback available. This fixes the following warning by modpost:

drivers/pci/controller/dwc/pcie-kirin: section mismatch in reference: kirin_pcie_driver+0x8 (section: .data) -> kirin_pcie_remove (section: .exit.text)

(with ARCH=x86_64 W=1 allmodconfig).

---
 drivers/pci/controller/dwc/pcie-kirin.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

@@ -742,7 +742,7 @@ err:
 	return ret;
 }
 
-static int __exit kirin_pcie_remove(struct platform_device *pdev)
+static int kirin_pcie_remove(struct platform_device *pdev)
 {
 	struct kirin_pcie *kirin_pcie = platform_get_drvdata(pdev);
 

@@ -819,7 +819,7 @@ static int kirin_pcie_probe(struct platf
 
 static struct platform_driver kirin_pcie_driver = {
 	.probe			= kirin_pcie_probe,
-	.remove	        	= __exit_p(kirin_pcie_remove),
+	.remove	        	= kirin_pcie_remove,
 	.driver			= {
 		.name			= "kirin-pcie",
 		.of_match_table		= kirin_pcie_match,

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Uwe Kleine-König <[email protected]>

commit 83a939f0fdc208ff3639dd3d42ac9b3c35607fd2 upstream.

With CONFIG_PCI_EXYNOS=y and exynos_pcie_remove() marked with __exit, the function is discarded from the driver. In this case a bound device can still get unbound, e.g via sysfs. Then no cleanup code is run resulting in resource leaks or worse.

The right thing to do is do always have the remove callback available. This fixes the following warning by modpost:

WARNING: modpost: drivers/pci/controller/dwc/pci-exynos: section mismatch in reference: exynos_pcie_driver+0x8 (section: .data) -> exynos_pcie_remove (section: .exit.text)

(with ARCH=x86_64 W=1 allmodconfig).

---
 drivers/pci/controller/dwc/pci-exynos.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

@@ -375,7 +375,7 @@ fail_probe:
 	return ret;
 }
 
-static int __exit exynos_pcie_remove(struct platform_device *pdev)
+static int exynos_pcie_remove(struct platform_device *pdev)
 {
 	struct exynos_pcie *ep = platform_get_drvdata(pdev);
 

@@ -431,7 +431,7 @@ static const struct of_device_id exynos_
 
 static struct platform_driver exynos_pcie_driver = {
 	.probe		= exynos_pcie_probe,
-	.remove		= __exit_p(exynos_pcie_remove),
+	.remove		= exynos_pcie_remove,
 	.driver = {
 		.name	= "exynos-pcie",
 		.of_match_table = exynos_pcie_of_match,

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Ajay Singh <[email protected]>

commit 05ac1a198a63ad66bf5ae8b7321407c102d40ef3 upstream.

Enabling KASAN and running some iperf tests raises some memory issues with vmm_table:

BUG: KASAN: slab-out-of-bounds in wilc_wlan_handle_txq+0x6ac/0xdb4
Write of size 4 at addr c3a61540 by task wlan0-tx/95

KASAN detects that we are writing data beyond range allocated to vmm_table. There is indeed a mismatch between the size passed to allocator in wilc_wlan_init, and the range of possible indexes used later: allocation size is missing a multiplication by sizeof(u32)

---
 drivers/net/wireless/microchip/wilc1000/wlan.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -1492,7 +1492,7 @@ int wilc_wlan_init(struct net_device *de
 	}
 
 	if (!wilc->vmm_table)
-		wilc->vmm_table = kzalloc(WILC_VMM_TBL_SIZE, GFP_KERNEL);
+		wilc->vmm_table = kcalloc(WILC_VMM_TBL_SIZE, sizeof(u32), GFP_KERNEL);
 
 	if (!wilc->vmm_table) {
 		ret = -ENOBUFS;

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Chuck Lever <[email protected]>

commit 197115ebf358cb440c73e868b2a0a5ef728decc6 upstream.

When an RPC Call message cannot be pulled from the client, that is a message loss, by definition. Close the connection to trigger the client to resend.

---
 net/sunrpc/xprtrdma/svc_rdma_recvfrom.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

@@ -852,7 +852,8 @@ out_readfail:
 	if (ret == -EINVAL)
 		svc_rdma_send_error(rdma_xprt, ctxt, ret);
 	svc_rdma_recv_ctxt_put(rdma_xprt, ctxt);
-	return ret;
+	svc_xprt_deferred_close(xprt);
+	return -ENOTCONN;
 
 out_backchannel:
 	svc_rdma_handle_bc_reply(rqstp, ctxt);

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Joel Fernandes (Google) <[email protected]>

commit b96e7a5fa0ba9cda32888e04f8f4bac42d49a7f8 upstream.

There are instances where rcu_cpu_stall_reset() is called when jiffies did not get a chance to update for a long time. Before jiffies is updated, the CPU stall detector can go off triggering false-positives where a just-started grace period appears to be ages old. In the past, we disabled stall detection in rcu_cpu_stall_reset() however this got changed [1]. This is resulting in false-positives in KGDB usecase [2].

Fix this by deferring the update of jiffies to the third run of the FQS loop. This is more robust, as, even if rcu_cpu_stall_reset() is called just before jiffies is read, we would end up pushing out the jiffies read by 3 more FQS loops. Meanwhile the CPU stall detection will be delayed and we will not get any false positives.

[1] https://lore.kernel.org/all/[email protected]/
[2] https://lore.kernel.org/all/[email protected]/

Tested with rcutorture.cpu_stall option as well to verify stall behavior with/without patch.

---
 kernel/rcu/tree.c       |   12 ++++++++++++
 kernel/rcu/tree.h       |    4 ++++
 kernel/rcu/tree_stall.h |   20 ++++++++++++++++++--
 3 files changed, 34 insertions(+), 2 deletions(-)

@@ -1552,10 +1552,22 @@ static bool rcu_gp_fqs_check_wake(int *g
  */
 static void rcu_gp_fqs(bool first_time)
 {
+	int nr_fqs = READ_ONCE(rcu_state.nr_fqs_jiffies_stall);
 	struct rcu_node *rnp = rcu_get_root();
 
 	WRITE_ONCE(rcu_state.gp_activity, jiffies);
 	WRITE_ONCE(rcu_state.n_force_qs, rcu_state.n_force_qs + 1);
+
+	WARN_ON_ONCE(nr_fqs > 3);
+	/* Only countdown nr_fqs for stall purposes if jiffies moves. */
+	if (nr_fqs) {
+		if (nr_fqs == 1) {
+			WRITE_ONCE(rcu_state.jiffies_stall,
+				   jiffies + rcu_jiffies_till_stall_check());
+		}
+		WRITE_ONCE(rcu_state.nr_fqs_jiffies_stall, --nr_fqs);
+	}
+
 	if (first_time) {
 		/* Collect dyntick-idle snapshots. */
 		force_qs_rnp(dyntick_save_progress_counter);

@@ -386,6 +386,10 @@ struct rcu_state {
 						/*  in jiffies. */
 	unsigned long jiffies_stall;		/* Time at which to check */
 						/*  for CPU stalls. */
+	int nr_fqs_jiffies_stall;		/* Number of fqs loops after
+						 * which read jiffies and set
+						 * jiffies_stall. Stall
+						 * warnings disabled if !0. */
 	unsigned long jiffies_resched;		/* Time at which to resched */
 						/*  a reluctant CPU. */
 	unsigned long n_force_qs_gpstart;	/* Snapshot of n_force_qs at */

@@ -149,12 +149,17 @@ static void panic_on_rcu_stall(void)
 /**
  * rcu_cpu_stall_reset - restart stall-warning timeout for current grace period
  *
+ * To perform the reset request from the caller, disable stall detection until
+ * 3 fqs loops have passed. This is required to ensure a fresh jiffies is
+ * loaded.  It should be safe to do from the fqs loop as enough timer
+ * interrupts and context switches should have passed.
+ *
  * The caller must disable hard irqs.
  */
 void rcu_cpu_stall_reset(void)
 {
-	WRITE_ONCE(rcu_state.jiffies_stall,
-		   jiffies + rcu_jiffies_till_stall_check());
+	WRITE_ONCE(rcu_state.nr_fqs_jiffies_stall, 3);
+	WRITE_ONCE(rcu_state.jiffies_stall, ULONG_MAX);
 }
 
 //////////////////////////////////////////////////////////////////////////////

@@ -170,6 +175,7 @@ static void record_gp_stall_check_time(v
 	WRITE_ONCE(rcu_state.gp_start, j);
 	j1 = rcu_jiffies_till_stall_check();
 	smp_mb(); // ->gp_start before ->jiffies_stall and caller's ->gp_seq.
+	WRITE_ONCE(rcu_state.nr_fqs_jiffies_stall, 0);
 	WRITE_ONCE(rcu_state.jiffies_stall, j + j1);
 	rcu_state.jiffies_resched = j + j1 / 2;
 	rcu_state.n_force_qs_gpstart = READ_ONCE(rcu_state.n_force_qs);

@@ -725,6 +731,16 @@ static void check_cpu_stall(struct rcu_d
 	    !rcu_gp_in_progress())
 		return;
 	rcu_stall_kick_kthreads();
+
+	/*
+	 * Check if it was requested (via rcu_cpu_stall_reset()) that the FQS
+	 * loop has to set jiffies to ensure a non-stale jiffies value. This
+	 * is required to have good jiffies value after coming out of long
+	 * breaks of jiffies updates. Not doing so can cause false positives.
+	 */
+	if (READ_ONCE(rcu_state.nr_fqs_jiffies_stall) > 0)
+		return;
+
 	j = jiffies;
 
 	/*

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Vignesh Viswanathan <[email protected]>

commit 95d97b111e1e184b0c8656137033ed64f2cf21e4 upstream.

SMEM uses lock index 3 of the TCSR Mutex hwlock for allocations in SMEM region shared by the Host and FW.

Fix the SMEM hwlock index to 3 for IPQ6018.

---
 arch/arm64/boot/dts/qcom/ipq6018.dtsi |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -207,7 +207,7 @@
 	smem {
 		compatible = "qcom,smem";
 		memory-region = <&smem_region>;
-		hwlocks = <&tcsr_mutex 0>;
+		hwlocks = <&tcsr_mutex 3>;
 	};
 
 	soc: soc@0 {

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Biju Das <[email protected]>

commit b7a8f1f7a8a25e09aaefebb6251a77f44cda638b upstream.

As per R01UH0914EJ0130 Rev.1.30 HW manual the MTU3 overflow/underflow interrupt names starts with 'tci' instead of 'tgi'.

Fix this documentation issue by replacing below overflow/underflow interrupt names:
- tgiv0->tciv0
- tgiv1->tciv1
- tgiu1->tciu1
- tgiv2->tciv2
- tgiu2->tciu2
- tgiv3->tciv3
- tgiv4->tciv4
- tgiv6->tciv6
- tgiv7->tciv7
- tgiv8->tciv8
- tgiu8->tciu8

---
 .../bindings/timer/renesas,rz-mtu3.yaml       | 38 +++++++++----------
 1 file changed, 19 insertions(+), 19 deletions(-)

index bffdab0b0185..fbac40b958dd 100644
@@ -169,27 +169,27 @@ properties:
       - const: tgib0
       - const: tgic0
       - const: tgid0
-      - const: tgiv0
+      - const: tciv0
       - const: tgie0
       - const: tgif0
       - const: tgia1
       - const: tgib1
-      - const: tgiv1
-      - const: tgiu1
+      - const: tciv1
+      - const: tciu1
       - const: tgia2
       - const: tgib2
-      - const: tgiv2
-      - const: tgiu2
+      - const: tciv2
+      - const: tciu2
       - const: tgia3
       - const: tgib3
       - const: tgic3
       - const: tgid3
-      - const: tgiv3
+      - const: tciv3
       - const: tgia4
       - const: tgib4
       - const: tgic4
       - const: tgid4
-      - const: tgiv4
+      - const: tciv4
       - const: tgiu5
       - const: tgiv5
       - const: tgiw5

@@ -197,18 +197,18 @@ properties:
       - const: tgib6
       - const: tgic6
       - const: tgid6
-      - const: tgiv6
+      - const: tciv6
       - const: tgia7
       - const: tgib7
       - const: tgic7
       - const: tgid7
-      - const: tgiv7
+      - const: tciv7
       - const: tgia8
       - const: tgib8
       - const: tgic8
       - const: tgid8
-      - const: tgiv8
-      - const: tgiu8
+      - const: tciv8
+      - const: tciu8
 
   clocks:
     maxItems: 1

@@ -285,16 +285,16 @@ examples:
                    <GIC_SPI 211 IRQ_TYPE_EDGE_RISING>,
                    <GIC_SPI 212 IRQ_TYPE_EDGE_RISING>,
                    <GIC_SPI 213 IRQ_TYPE_EDGE_RISING>;
-      interrupt-names = "tgia0", "tgib0", "tgic0", "tgid0", "tgiv0", "tgie0",
+      interrupt-names = "tgia0", "tgib0", "tgic0", "tgid0", "tciv0", "tgie0",
                         "tgif0",
-                        "tgia1", "tgib1", "tgiv1", "tgiu1",
-                        "tgia2", "tgib2", "tgiv2", "tgiu2",
-                        "tgia3", "tgib3", "tgic3", "tgid3", "tgiv3",
-                        "tgia4", "tgib4", "tgic4", "tgid4", "tgiv4",
+                        "tgia1", "tgib1", "tciv1", "tciu1",
+                        "tgia2", "tgib2", "tciv2", "tciu2",
+                        "tgia3", "tgib3", "tgic3", "tgid3", "tciv3",
+                        "tgia4", "tgib4", "tgic4", "tgid4", "tciv4",
                         "tgiu5", "tgiv5", "tgiw5",
-                        "tgia6", "tgib6", "tgic6", "tgid6", "tgiv6",
-                        "tgia7", "tgib7", "tgic7", "tgid7", "tgiv7",
-                        "tgia8", "tgib8", "tgic8", "tgid8", "tgiv8", "tgiu8";
+                        "tgia6", "tgib6", "tgic6", "tgid6", "tciv6",
+                        "tgia7", "tgib7", "tgic7", "tgid7", "tciv7",
+                        "tgia8", "tgib8", "tgic8", "tgid8", "tciv8", "tciu8";
       clocks = <&cpg CPG_MOD R9A07G044_MTU_X_MCK_MTU3>;
       power-domains = <&cpg>;
       resets = <&cpg R9A07G044_MTU_X_PRESET_MTU3>;

2.43.0

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Brian Geffon <[email protected]>

commit f0c7183008b41e92fa676406d87f18773724b48b upstream.

We found at least one situation where the safe pages list was empty and get_buffer() would gladly try to use a NULL pointer.

---
 kernel/power/snapshot.c |   10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

@@ -2474,8 +2474,9 @@ static void *get_highmem_page_buffer(str
 		pbe->copy_page = tmp;
 	} else {
 		/* Copy of the page will be stored in normal memory */
-		kaddr = safe_pages_list;
-		safe_pages_list = safe_pages_list->next;
+		kaddr = __get_safe_page(ca->gfp_mask);
+		if (!kaddr)
+			return ERR_PTR(-ENOMEM);
 		pbe->copy_page = virt_to_page(kaddr);
 	}
 	pbe->next = highmem_pblist;

@@ -2655,8 +2656,9 @@ static void *get_buffer(struct memory_bi
 		return ERR_PTR(-ENOMEM);
 	}
 	pbe->orig_address = page_address(page);
-	pbe->address = safe_pages_list;
-	safe_pages_list = safe_pages_list->next;
+	pbe->address = __get_safe_page(ca->gfp_mask);
+	if (!pbe->address)
+		return ERR_PTR(-ENOMEM);
 	pbe->next = restore_pblist;
 	restore_pblist = pbe;
 	return pbe->address;

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Brian Geffon <[email protected]>

commit d08970df1980476f27936e24d452550f3e9e92e1 upstream.

In snapshot_write_next(), sync_read is set and unset in three different spots unnecessiarly. As a result there is a subtle bug where the first page after the meta data has been loaded unconditionally sets sync_read to 0. If this first PFN was actually a highmem page, then the returned buffer will be the global "buffer," and the page needs to be loaded synchronously.

That is, I'm not sure we can always assume the following to be safe:

	handle->buffer = get_buffer(&orig_bm, &ca);
	handle->sync_read = 0;

Because get_buffer() can call get_highmem_page_buffer() which can return 'buffer'.

The easiest way to address this is just set sync_read before snapshot_write_next() returns if handle->buffer == buffer.

---
 kernel/power/snapshot.c |    6 +-----
 1 file changed, 1 insertion(+), 5 deletions(-)

@@ -2689,8 +2689,6 @@ int snapshot_write_next(struct snapshot_
 	if (handle->cur > 1 && handle->cur > nr_meta_pages + nr_copy_pages)
 		return 0;
 
-	handle->sync_read = 1;
-
 	if (!handle->cur) {
 		if (!buffer)
 			/* This makes the buffer be freed by swsusp_free() */

@@ -2726,7 +2724,6 @@ int snapshot_write_next(struct snapshot_
 			memory_bm_position_reset(&orig_bm);
 			restore_pblist = NULL;
 			handle->buffer = get_buffer(&orig_bm, &ca);
-			handle->sync_read = 0;
 			if (IS_ERR(handle->buffer))
 				return PTR_ERR(handle->buffer);
 		}

@@ -2736,9 +2733,8 @@ int snapshot_write_next(struct snapshot_
 		handle->buffer = get_buffer(&orig_bm, &ca);
 		if (IS_ERR(handle->buffer))
 			return PTR_ERR(handle->buffer);
-		if (handle->buffer != buffer)
-			handle->sync_read = 0;
 	}
+	handle->sync_read = (handle->buffer == buffer);
 	handle->cur++;
 	return PAGE_SIZE;
 }

6.5-stable review patch. If anyone has any objections, please let me know.

------------------

From: Catalin Marinas <[email protected]>

commit 5f98fd034ca6fd1ab8c91a3488968a0e9caaabf6 upstream.

Since the actual slab freeing is deferred when calling kvfree_rcu(), so is the kmemleak_free() callback informing kmemleak of the object deletion. From the perspective of the kvfree_rcu() caller, the object is freed and it may remove any references to it. Since kmemleak does not scan RCU internal data storing the pointer, it will report such objects as leaks during the grace period.

Tell kmemleak to ignore such objects on the kvfree_call_rcu() path. Note that the tiny RCU implementation does not have such issue since the objects can be tracked from the rcu_ctrlblk structure.

---
 kernel/rcu/tree.c |    9 +++++++++
 1 file changed, 9 insertions(+)

@@ -31,6 +31,7 @@
 #include <linux/bitops.h>
 #include <linux/export.h>
 #include <linux/completion.h>
+#include <linux/kmemleak.h>
 #include <linux/moduleparam.h>
 #include <linux/panic.h>
 #include <linux/panic_notifier.h>

@@ -3397,6 +3398,14 @@ void kvfree_call_rcu(struct rcu_head *he
 		success = true;
 	}
 
+	/*
+	 * The kvfree_rcu() caller considers the pointer freed at this point
+	 * and likely removes any references to it. Since the actual slab
+	 * freeing (and kmemleak_free()) is deferred, tell kmemleak to ignore
+	 * this object (no scanning or false positives reporting).
+	 */
+	kmemleak_ignore(ptr);
+
 	// Set timer to drain after KFREE_DRAIN_JIFFIES.
 	if (rcu_scheduler_active == RCU_SCHEDULER_RUNNING)
 		schedule_delayed_monitor_work(krcp);