6.12.64-rc1 review

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

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

From: Nicolas Ferre <[email protected]>

commit 7d5864dc5d5ea6a35983dd05295fb17f2f2f44ce upstream.

Unlike standalone spi peripherals, on sama5d2, the flexcom spi have fifo size of 32 data. Fix flexcom/spi nodes where this property is wrong.

---
 arch/arm/boot/dts/microchip/sama5d2.dtsi |   10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

@@ -568,7 +568,7 @@
 						 AT91_XDMAC_DT_PER_IF(1) |
 						 AT91_XDMAC_DT_PERID(12))>;
 					dma-names = "tx", "rx";
-					atmel,fifo-size = <16>;
+					atmel,fifo-size = <32>;
 					status = "disabled";
 				};
 

@@ -639,7 +639,7 @@
 						 AT91_XDMAC_DT_PER_IF(1) |
 						 AT91_XDMAC_DT_PERID(14))>;
 					dma-names = "tx", "rx";
-					atmel,fifo-size = <16>;
+					atmel,fifo-size = <32>;
 					status = "disabled";
 				};
 

@@ -851,7 +851,7 @@
 						 AT91_XDMAC_DT_PER_IF(1) |
 						 AT91_XDMAC_DT_PERID(16))>;
 					dma-names = "tx", "rx";
-					atmel,fifo-size = <16>;
+					atmel,fifo-size = <32>;
 					status = "disabled";
 				};
 

@@ -922,7 +922,7 @@
 						 AT91_XDMAC_DT_PER_IF(1) |
 						 AT91_XDMAC_DT_PERID(18))>;
 					dma-names = "tx", "rx";
-					atmel,fifo-size = <16>;
+					atmel,fifo-size = <32>;
 					status = "disabled";
 				};
 

@@ -994,7 +994,7 @@
 						 AT91_XDMAC_DT_PER_IF(1) |
 						 AT91_XDMAC_DT_PERID(20))>;
 					dma-names = "tx", "rx";
-					atmel,fifo-size = <16>;
+					atmel,fifo-size = <32>;
 					status = "disabled";
 				};
 

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

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

From: Thomas Gleixner <[email protected]>

[ Upstream commit 0edc78b82bea85e1b2165d8e870a5c3535919695 ]

Luigi reported that retriggering a posted MSI interrupt does not work correctly.

The reason is that the retrigger happens at the vector domain by sending an IPI to the actual vector on the target CPU. That works correctly exactly once because the posted MSI interrupt chip does not issue an EOI as that's only required for the posted MSI notification vector itself.

As a consequence the vector becomes stale in the ISR, which not only affects this vector but also any lower priority vector in the affected APIC because the ISR bit is not cleared.

Luigi proposed to set the vector in the remap PIR bitmap and raise the posted MSI notification vector. That works, but that still does not cure a related problem:

If there is ever a stray interrupt on such a vector, then the related APIC ISR bit becomes stale due to the lack of EOI as described above. Unlikely to happen, but if it happens it's not debuggable at all.

So instead of playing games with the PIR, this can be actually solved for both cases by:

1) Keeping track of the posted interrupt vector handler state

2) Implementing a posted MSI specific irq_ack() callback which checks that state. If the posted vector handler is inactive it issues an EOI, otherwise it delegates that to the posted handler.

This is correct versus affinity changes and concurrent events on the posted vector as the actual handler invocation is serialized through the interrupt descriptor lock.

---
 arch/x86/include/asm/irq_remapping.h |    7 +++++++
 arch/x86/kernel/irq.c                |   23 +++++++++++++++++++++++
 drivers/iommu/intel/irq_remapping.c  |    8 ++++----
 3 files changed, 34 insertions(+), 4 deletions(-)

@@ -72,4 +72,11 @@ static inline void panic_if_irq_remap(co
 }
 
 #endif /* CONFIG_IRQ_REMAP */
+
+#ifdef CONFIG_X86_POSTED_MSI
+void intel_ack_posted_msi_irq(struct irq_data *irqd);
+#else
+#define intel_ack_posted_msi_irq	NULL
+#endif
+
 #endif /* __X86_IRQ_REMAPPING_H */

@@ -391,6 +391,7 @@ DEFINE_IDTENTRY_SYSVEC_SIMPLE(sysvec_kvm
 
 /* Posted Interrupt Descriptors for coalesced MSIs to be posted */
 DEFINE_PER_CPU_ALIGNED(struct pi_desc, posted_msi_pi_desc);
+static DEFINE_PER_CPU(bool, posted_msi_handler_active);
 
 void intel_posted_msi_init(void)
 {

@@ -408,6 +409,25 @@ void intel_posted_msi_init(void)
 	this_cpu_write(posted_msi_pi_desc.ndst, destination);
 }
 
+void intel_ack_posted_msi_irq(struct irq_data *irqd)
+{
+	irq_move_irq(irqd);
+
+	/*
+	 * Handle the rare case that irq_retrigger() raised the actual
+	 * assigned vector on the target CPU, which means that it was not
+	 * invoked via the posted MSI handler below. In that case APIC EOI
+	 * is required as otherwise the ISR entry becomes stale and lower
+	 * priority interrupts are never going to be delivered after that.
+	 *
+	 * If the posted handler invoked the device interrupt handler then
+	 * the EOI would be premature because it would acknowledge the
+	 * posted vector.
+	 */
+	if (unlikely(!__this_cpu_read(posted_msi_handler_active)))
+		apic_eoi();
+}
+
 /*
  * De-multiplexing posted interrupts is on the performance path, the code
  * below is written to optimize the cache performance based on the following

@@ -483,6 +503,8 @@ DEFINE_IDTENTRY_SYSVEC(sysvec_posted_msi
 
 	pid = this_cpu_ptr(&posted_msi_pi_desc);
 
+	/* Mark the handler active for intel_ack_posted_msi_irq() */
+	__this_cpu_write(posted_msi_handler_active, true);
 	inc_irq_stat(posted_msi_notification_count);
 	irq_enter();
 

@@ -511,6 +533,7 @@ DEFINE_IDTENTRY_SYSVEC(sysvec_posted_msi
 
 	apic_eoi();
 	irq_exit();
+	__this_cpu_write(posted_msi_handler_active, false);
 	set_irq_regs(old_regs);
 }
 #endif /* X86_POSTED_MSI */

@@ -1309,17 +1309,17 @@ static struct irq_chip intel_ir_chip = {
  *	irq_enter();
  *		handle_edge_irq()
  *			irq_chip_ack_parent()
- *				irq_move_irq(); // No EOI
+ *				intel_ack_posted_msi_irq(); // No EOI
  *			handle_irq_event()
  *				driver_handler()
  *		handle_edge_irq()
  *			irq_chip_ack_parent()
- *				irq_move_irq(); // No EOI
+ *				intel_ack_posted_msi_irq(); // No EOI
  *			handle_irq_event()
  *				driver_handler()
  *		handle_edge_irq()
  *			irq_chip_ack_parent()
- *				irq_move_irq(); // No EOI
+ *				intel_ack_posted_msi_irq(); // No EOI
  *			handle_irq_event()
  *				driver_handler()
  *	apic_eoi()

@@ -1328,7 +1328,7 @@ static struct irq_chip intel_ir_chip = {
  */
 static struct irq_chip intel_ir_chip_post_msi = {
 	.name			= "INTEL-IR-POST",
-	.irq_ack		= irq_move_irq,
+	.irq_ack		= intel_ack_posted_msi_irq,
 	.irq_set_affinity	= intel_ir_set_affinity,
 	.irq_compose_msi_msg	= intel_ir_compose_msi_msg,
 	.irq_set_vcpu_affinity	= intel_ir_set_vcpu_affinity,

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

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

From: Johan Hovold <[email protected]>

commit de83d4617f9fe059623e97acf7e1e10d209625b5 upstream.

The driver is dropping the references taken to the larb devices during probe after successful lookup as well as on errors. This can potentially lead to a use-after-free in case a larb device has not yet been bound to its driver so that the iommu driver probe defers.

Fix this by keeping the references as expected while the iommu driver is bound.

---
 drivers/iommu/mtk_iommu.c |   25 ++++++++++++++++++-------
 1 file changed, 18 insertions(+), 7 deletions(-)

@@ -1213,16 +1213,19 @@ static int mtk_iommu_mm_dts_parse(struct
 		}
 
 		component_match_add(dev, match, component_compare_dev, &plarbdev->dev);
-		platform_device_put(plarbdev);
 	}
 
-	if (!frst_avail_smicomm_node)
-		return -EINVAL;
+	if (!frst_avail_smicomm_node) {
+		ret = -EINVAL;
+		goto err_larbdev_put;
+	}
 
 	pcommdev = of_find_device_by_node(frst_avail_smicomm_node);
 	of_node_put(frst_avail_smicomm_node);
-	if (!pcommdev)
-		return -ENODEV;
+	if (!pcommdev) {
+		ret = -ENODEV;
+		goto err_larbdev_put;
+	}
 	data->smicomm_dev = &pcommdev->dev;
 
 	link = device_link_add(data->smicomm_dev, dev,

@@ -1230,7 +1233,8 @@ static int mtk_iommu_mm_dts_parse(struct
 	platform_device_put(pcommdev);
 	if (!link) {
 		dev_err(dev, "Unable to link %s.\n", dev_name(data->smicomm_dev));
-		return -EINVAL;
+		ret = -EINVAL;
+		goto err_larbdev_put;
 	}
 	return 0;
 

@@ -1402,8 +1406,12 @@ out_sysfs_remove:
 	iommu_device_sysfs_remove(&data->iommu);
 out_list_del:
 	list_del(&data->list);
-	if (MTK_IOMMU_IS_TYPE(data->plat_data, MTK_IOMMU_TYPE_MM))
+	if (MTK_IOMMU_IS_TYPE(data->plat_data, MTK_IOMMU_TYPE_MM)) {
 		device_link_remove(data->smicomm_dev, dev);
+
+		for (i = 0; i < MTK_LARB_NR_MAX; i++)
+			put_device(data->larb_imu[i].dev);
+	}
 out_runtime_disable:
 	pm_runtime_disable(dev);
 	return ret;

@@ -1423,6 +1431,9 @@ static void mtk_iommu_remove(struct plat
 	if (MTK_IOMMU_IS_TYPE(data->plat_data, MTK_IOMMU_TYPE_MM)) {
 		device_link_remove(data->smicomm_dev, &pdev->dev);
 		component_master_del(&pdev->dev, &mtk_iommu_com_ops);
+
+		for (i = 0; i < MTK_LARB_NR_MAX; i++)
+			put_device(data->larb_imu[i].dev);
 	}
 	pm_runtime_disable(&pdev->dev);
 	for (i = 0; i < data->plat_data->banks_num; i++) {

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

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

From: Joanne Koong <[email protected]>

commit bd5603eaae0aabf527bfb3ce1bb07e979ce5bd50 upstream.

Commit e26ee4efbc79 ("fuse: allocate ff->release_args only if release is needed") skips allocating ff->release_args if the server does not implement open. However in doing so, fuse_prepare_release() now skips grabbing the reference on the inode, which makes it possible for an inode to be evicted from the dcache while there are inflight readahead requests. This causes a deadlock if the server triggers reclaim while servicing the readahead request and reclaim attempts to evict the inode of the file being read ahead. Since the folio is locked during readahead, when reclaim evicts the fuse inode and fuse_evict_inode() attempts to remove all folios associated with the inode from the page cache (truncate_inode_pages_range()), reclaim will block forever waiting for the lock since readahead cannot relinquish the lock because it is itself blocked in reclaim:

folio_wait_bit_common (mm/filemap.c:1308:4) folio_lock (./include/linux/pagemap.h:1052:3) truncate_inode_pages_range (mm/truncate.c:336:10) fuse_evict_inode (fs/fuse/inode.c:161:2) evict (fs/inode.c:704:3) dentry_unlink_inode (fs/dcache.c:412:3) __dentry_kill (fs/dcache.c:615:3) shrink_kill (fs/dcache.c:1060:12) shrink_dentry_list (fs/dcache.c:1087:3) prune_dcache_sb (fs/dcache.c:1168:2) super_cache_scan (fs/super.c:221:10) do_shrink_slab (mm/shrinker.c:435:9) shrink_slab (mm/shrinker.c:626:10) shrink_node (mm/vmscan.c:5951:2) shrink_zones (mm/vmscan.c:6195:3) do_try_to_free_pages (mm/vmscan.c:6257:3) do_swap_page (mm/memory.c:4136:11) handle_pte_fault (mm/memory.c:5562:10) handle_mm_fault (mm/memory.c:5870:9) do_user_addr_fault (arch/x86/mm/fault.c:1338:10) handle_page_fault (arch/x86/mm/fault.c:1481:3) exc_page_fault (arch/x86/mm/fault.c:1539:2) asm_exc_page_fault+0x22/0x27

Fix this deadlock by allocating ff->release_args and grabbing the reference on the inode when preparing the file for release even if the server does not implement open. The inode reference will be dropped when the last reference on the fuse file is dropped (see fuse_file_put() -> fuse_release_end()).

---
 fs/fuse/file.c |   26 +++++++++++++++++++-------
 1 file changed, 19 insertions(+), 7 deletions(-)

@@ -109,7 +109,9 @@ static void fuse_file_put(struct fuse_fi
 			fuse_file_io_release(ff, ra->inode);
 
 		if (!args) {
-			/* Do nothing when server does not implement 'open' */
+			/* Do nothing when server does not implement 'opendir' */
+		} else if (args->opcode == FUSE_RELEASE && ff->fm->fc->no_open) {
+			fuse_release_end(ff->fm, args, 0);
 		} else if (sync) {
 			fuse_simple_request(ff->fm, args);
 			fuse_release_end(ff->fm, args, 0);

@@ -130,8 +132,17 @@ struct fuse_file *fuse_file_open(struct
 	struct fuse_file *ff;
 	int opcode = isdir ? FUSE_OPENDIR : FUSE_OPEN;
 	bool open = isdir ? !fc->no_opendir : !fc->no_open;
+	bool release = !isdir || open;
 
-	ff = fuse_file_alloc(fm, open);
+	/*
+	 * ff->args->release_args still needs to be allocated (so we can hold an
+	 * inode reference while there are pending inflight file operations when
+	 * ->release() is called, see fuse_prepare_release()) even if
+	 * fc->no_open is set else it becomes possible for reclaim to deadlock
+	 * if while servicing the readahead request the server triggers reclaim
+	 * and reclaim evicts the inode of the file being read ahead.
+	 */
+	ff = fuse_file_alloc(fm, release);
 	if (!ff)
 		return ERR_PTR(-ENOMEM);
 

@@ -151,13 +162,14 @@ struct fuse_file *fuse_file_open(struct
 			fuse_file_free(ff);
 			return ERR_PTR(err);
 		} else {
-			/* No release needed */
-			kfree(ff->args);
-			ff->args = NULL;
-			if (isdir)
+			if (isdir) {
+				/* No release needed */
+				kfree(ff->args);
+				ff->args = NULL;
 				fc->no_opendir = 1;
-			else
+			} else {
 				fc->no_open = 1;
+			}
 		}
 	}
 

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

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

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

[ Upstream commit f3ccdfda345ca9a624ea425840a926b8338c1e25 ]

The indirect IO is necessary for RTL8822CS, but not necessary for other chips. Otherwiese, it throws errors and becomes unusable.

rtw88_8723cs mmc1:0001:1: WOW Firmware version 11.0.0, H2C version 0 rtw88_8723cs mmc1:0001:1: Firmware version 11.0.0, H2C version 0 rtw88_8723cs mmc1:0001:1: sdio read32 failed (0xf0): -110 rtw88_8723cs mmc1:0001:1: sdio write8 failed (0x1c): -110 rtw88_8723cs mmc1:0001:1: sdio read32 failed (0xf0): -110

By vendor driver, only RTL8822CS and RTL8822ES need indirect IO, but RTL8822ES isn't supported yet. Therefore, limit it to RTL8822CS only.

---
 drivers/net/wireless/realtek/rtw88/sdio.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

index d6bea5ec8e24..d8db341a5731 100644
@@ -144,8 +144,10 @@ static u32 rtw_sdio_to_io_address(struct rtw_dev *rtwdev, u32 addr,
 
 static bool rtw_sdio_use_direct_io(struct rtw_dev *rtwdev, u32 addr)
 {
+	bool might_indirect_under_power_off = rtwdev->chip->id == RTW_CHIP_TYPE_8822C;
+
 	if (!test_bit(RTW_FLAG_POWERON, rtwdev->flags) &&
-	    !rtw_sdio_is_bus_addr(addr))
+	    !rtw_sdio_is_bus_addr(addr) && might_indirect_under_power_off)
 		return false;
 
 	return !rtw_sdio_is_sdio30_supported(rtwdev) ||

2.51.0

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

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

From: Morning Star <[email protected]>

[ Upstream commit dd39edb445f07400e748da967a07d5dca5c5f96e ]

TID getting from ieee80211_get_tid() might be out of range of array size of sta_entry->tids[], so check TID is less than MAX_TID_COUNT. Othwerwise, UBSAN warn:

 UBSAN: array-index-out-of-bounds in drivers/net/wireless/realtek/rtlwifi/rtl8192cu/trx.c:514:30
 index 10 is out of range for type 'rtl_tid_data [9]'

---
 drivers/net/wireless/realtek/rtlwifi/rtl8192cu/trx.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

index aa702ba7c9f5..d6c35e8d02a5 100644
@@ -511,7 +511,8 @@ void rtl92cu_tx_fill_desc(struct ieee80211_hw *hw,
 	if (sta) {
 		sta_entry = (struct rtl_sta_info *)sta->drv_priv;
 		tid = ieee80211_get_tid(hdr);
-		agg_state = sta_entry->tids[tid].agg.agg_state;
+		if (tid < MAX_TID_COUNT)
+			agg_state = sta_entry->tids[tid].agg.agg_state;
 		ampdu_density = sta->deflink.ht_cap.ampdu_density;
 	}
 

2.51.0

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

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

From: Dan Carpenter <[email protected]>

[ Upstream commit 2b77b9551d1184cb5af8271ff350e6e2c1b3db0d ]

The QGenie AI code review tool says we should store the capped length to wdev->u.client.ssid_len. The AI is correct.

---
 net/wireless/sme.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index e0d3c713538b..d8250ae17d94 100644
@@ -913,7 +913,7 @@ void __cfg80211_connect_result(struct net_device *dev,
 
 			ssid_len = min(ssid->datalen, IEEE80211_MAX_SSID_LEN);
 			memcpy(wdev->u.client.ssid, ssid->data, ssid_len);
-			wdev->u.client.ssid_len = ssid->datalen;
+			wdev->u.client.ssid_len = ssid_len;
 			break;
 		}
 		rcu_read_unlock();

2.51.0

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

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

From: Aloka Dixit <[email protected]>

[ Upstream commit a519be2f5d958c5804f2cfd68f1f384291271fab ]

When userspace brings down and deletes a non-transmitted profile, it is expected to send a new updated Beacon template for the transmitted profile of that multiple BSSID (MBSSID) group which does not include the removed profile in MBSSID element. This update comes via NL80211_CMD_SET_BEACON.

Such updates work well as long as the group continues to have at least one non-transmitted profile as NL80211_ATTR_MBSSID_ELEMS is included in the new Beacon template.

But when the last non-trasmitted profile is removed, it still gets included in Beacon templates sent to driver. This happens because when no MBSSID elements are sent by the userspace, ieee80211_assign_beacon() ends up using the element stored from earlier Beacon template.

Do not copy old MBSSID elements, instead userspace should always include these when applicable.

---
 net/mac80211/cfg.c | 10 ----------
 1 file changed, 10 deletions(-)

index 2df4df75f195..0abb687fd58d 100644
@@ -1140,7 +1140,6 @@ ieee80211_assign_beacon(struct ieee80211_sub_if_data *sdata,
 
 	size = sizeof(*new) + new_head_len + new_tail_len;
 
-	/* new or old multiple BSSID elements? */
 	if (params->mbssid_ies) {
 		mbssid = params->mbssid_ies;
 		size += struct_size(new->mbssid_ies, elem, mbssid->cnt);

@@ -1150,15 +1149,6 @@ ieee80211_assign_beacon(struct ieee80211_sub_if_data *sdata,
 		}
 		size += ieee80211_get_mbssid_beacon_len(mbssid, rnr,
 							mbssid->cnt);
-	} else if (old && old->mbssid_ies) {
-		mbssid = old->mbssid_ies;
-		size += struct_size(new->mbssid_ies, elem, mbssid->cnt);
-		if (old && old->rnr_ies) {
-			rnr = old->rnr_ies;
-			size += struct_size(new->rnr_ies, elem, rnr->cnt);
-		}
-		size += ieee80211_get_mbssid_beacon_len(mbssid, rnr,
-							mbssid->cnt);
 	}
 
 	new = kzalloc(size, GFP_KERNEL);

2.51.0

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

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

From: Przemyslaw Korba <[email protected]>

[ Upstream commit be43abc5514167cc129a8d8e9727b89b8e1d9719 ]

Add service task schedule to set_rx_mode.
In some cases there are error messages printed out in PTP application (ptp4l):

ptp4l[13848.762]: port 1 (ens2f3np3): received SYNC without timestamp ptp4l[13848.825]: port 1 (ens2f3np3): received SYNC without timestamp ptp4l[13848.887]: port 1 (ens2f3np3): received SYNC without timestamp

This happens when service task would not run immediately after set_rx_mode, and we need it for setup tasks. This service task checks, if PTP RX packets are hung in firmware, and propagate correct settings such as multicast address for IEEE 1588 Precision Time Protocol. RX timestamping depends on some of these filters set. Bug happens only with high PTP packets frequency incoming, and not every run since sometimes service task is being ran from a different place immediately after starting ptp4l.

---
 drivers/net/ethernet/intel/i40e/i40e_main.c | 1 +
 1 file changed, 1 insertion(+)

index eae5923104f7..2dc737c7e3fd 100644
@@ -2256,6 +2256,7 @@ static void i40e_set_rx_mode(struct net_device *netdev)
 		vsi->flags |= I40E_VSI_FLAG_FILTER_CHANGED;
 		set_bit(__I40E_MACVLAN_SYNC_PENDING, vsi->back->state);
 	}
+	i40e_service_event_schedule(vsi->back);
 }
 
 /**

2.51.0

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

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

From: Gregory Herrero <[email protected]>

[ Upstream commit 69942834215323cd9131db557091b4dec43f19c5 ]

The maximum number of descriptors supported by the hardware is hardware-dependent and can be retrieved using i40e_get_max_num_descriptors(). Move this function to a shared header and use it when checking for valid ring_len parameter rather than using hardcoded value.

By fixing an over-acceptance issue, behavior change could be seen where ring_len could now be rejected while configuring rx and tx queues if its size is larger than the hardware-dependent maximum number of descriptors.

---
 drivers/net/ethernet/intel/i40e/i40e.h             | 11 +++++++++++
 drivers/net/ethernet/intel/i40e/i40e_ethtool.c     | 12 ------------
 drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c |  4 ++--
 3 files changed, 13 insertions(+), 14 deletions(-)

index b8de97343ad3..de3d5e5b8306 100644
@@ -1415,4 +1415,15 @@ static inline struct i40e_veb *i40e_pf_get_main_veb(struct i40e_pf *pf)
 	return (pf->lan_veb != I40E_NO_VEB) ? pf->veb[pf->lan_veb] : NULL;
 }
 
+static inline u32 i40e_get_max_num_descriptors(const struct i40e_pf *pf)
+{
+	const struct i40e_hw *hw = &pf->hw;
+
+	switch (hw->mac.type) {
+	case I40E_MAC_XL710:
+		return I40E_MAX_NUM_DESCRIPTORS_XL710;
+	default:
+		return I40E_MAX_NUM_DESCRIPTORS;
+	}
+}
 #endif /* _I40E_H_ */

index bce5b76f1e7a..9a96f67fb648 100644
@@ -2010,18 +2010,6 @@ static void i40e_get_drvinfo(struct net_device *netdev,
 		drvinfo->n_priv_flags += I40E_GL_PRIV_FLAGS_STR_LEN;
 }
 
-static u32 i40e_get_max_num_descriptors(struct i40e_pf *pf)
-{
-	struct i40e_hw *hw = &pf->hw;
-
-	switch (hw->mac.type) {
-	case I40E_MAC_XL710:
-		return I40E_MAX_NUM_DESCRIPTORS_XL710;
-	default:
-		return I40E_MAX_NUM_DESCRIPTORS;
-	}
-}
-
 static void i40e_get_ringparam(struct net_device *netdev,
 			       struct ethtool_ringparam *ring,
 			       struct kernel_ethtool_ringparam *kernel_ring,

index 646e394f5190..3251ffa7d994 100644
@@ -656,7 +656,7 @@ static int i40e_config_vsi_tx_queue(struct i40e_vf *vf, u16 vsi_id,
 
 	/* ring_len has to be multiple of 8 */
 	if (!IS_ALIGNED(info->ring_len, 8) ||
-	    info->ring_len > I40E_MAX_NUM_DESCRIPTORS_XL710) {
+	    info->ring_len > i40e_get_max_num_descriptors(pf)) {
 		ret = -EINVAL;
 		goto error_context;
 	}

@@ -726,7 +726,7 @@ static int i40e_config_vsi_rx_queue(struct i40e_vf *vf, u16 vsi_id,
 
 	/* ring_len has to be multiple of 32 */
 	if (!IS_ALIGNED(info->ring_len, 32) ||
-	    info->ring_len > I40E_MAX_NUM_DESCRIPTORS_XL710) {
+	    info->ring_len > i40e_get_max_num_descriptors(pf)) {
 		ret = -EINVAL;
 		goto error_param;
 	}

2.51.0

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

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

From: Kohei Enju <[email protected]>

[ Upstream commit 6daa2893f323981c7894c68440823326e93a7d61 ]

There are off-by-one bugs when configuring RSS hash key and lookup table, causing out-of-bounds reads to memory [1] and out-of-bounds writes to device registers.

Before commit 43a3d9ba34c9 ("i40evf: Allow PF driver to configure RSS"), the loop upper bounds were:

    i <= I40E_VFQF_{HKEY,HLUT}_MAX_INDEX

which is safe since the value is the last valid index.

That commit changed the bounds to:

    i <= adapter->rss_{key,lut}_size / 4

where rss_{key,lut}_size / 4 is the number of dwords, so the last valid index is (rss_{key,lut}_size / 4) - 1. Therefore, using <= accesses one element past the end.

Fix the issues by using < instead of <=, ensuring we do not exceed the bounds.

[1] KASAN splat about rss_key_size off-by-one

  BUG: KASAN: slab-out-of-bounds in iavf_config_rss+0x619/0x800
  Read of size 4 at addr ffff888102c50134 by task kworker/u8:6/63

  CPU: 0 UID: 0 PID: 63 Comm: kworker/u8:6 Not tainted 6.18.0-rc2-enjuk-tnguy-00378-g3005f5b77652-dirty #156 PREEMPT(voluntary)
  Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.3-debian-1.16.3-2 04/01/2014
  Workqueue: iavf iavf_watchdog_task
  Call Trace:
   <TASK>
   dump_stack_lvl+0x6f/0xb0
   print_report+0x170/0x4f3
   kasan_report+0xe1/0x1a0
   iavf_config_rss+0x619/0x800
   iavf_watchdog_task+0x2be7/0x3230
   process_one_work+0x7fd/0x1420
   worker_thread+0x4d1/0xd40
   kthread+0x344/0x660
   ret_from_fork+0x249/0x320
   ret_from_fork_asm+0x1a/0x30
   </TASK>

  Allocated by task 63:
   kasan_save_stack+0x30/0x50
   kasan_save_track+0x14/0x30
   __kasan_kmalloc+0x7f/0x90
   __kmalloc_noprof+0x246/0x6f0
   iavf_watchdog_task+0x28fc/0x3230
   process_one_work+0x7fd/0x1420
   worker_thread+0x4d1/0xd40
   kthread+0x344/0x660
   ret_from_fork+0x249/0x320
   ret_from_fork_asm+0x1a/0x30

  The buggy address belongs to the object at ffff888102c50100
   which belongs to the cache kmalloc-64 of size 64
  The buggy address is located 0 bytes to the right of
   allocated 52-byte region [ffff888102c50100, ffff888102c50134)

  The buggy address belongs to the physical page:
  page: refcount:0 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x102c50
  flags: 0x200000000000000(node=0|zone=2)
  page_type: f5(slab)
  raw: 0200000000000000 ffff8881000418c0 dead000000000122 0000000000000000
  raw: 0000000000000000 0000000080200020 00000000f5000000 0000000000000000
  page dumped because: kasan: bad access detected

  Memory state around the buggy address:
   ffff888102c50000: 00 00 00 00 00 00 00 fc fc fc fc fc fc fc fc fc
   ffff888102c50080: 00 00 00 00 00 00 00 fc fc fc fc fc fc fc fc fc
  >ffff888102c50100: 00 00 00 00 00 00 04 fc fc fc fc fc fc fc fc fc
                                       ^
   ffff888102c50180: 00 00 00 00 00 00 00 00 fc fc fc fc fc fc fc fc
   ffff888102c50200: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc

---
 drivers/net/ethernet/intel/iavf/iavf_main.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

index 5516795cc250..422af897d933 100644
@@ -1718,11 +1718,11 @@ static int iavf_config_rss_reg(struct iavf_adapter *adapter)
 	u16 i;
 
 	dw = (u32 *)adapter->rss_key;
-	for (i = 0; i <= adapter->rss_key_size / 4; i++)
+	for (i = 0; i < adapter->rss_key_size / 4; i++)
 		wr32(hw, IAVF_VFQF_HKEY(i), dw[i]);
 
 	dw = (u32 *)adapter->rss_lut;
-	for (i = 0; i <= adapter->rss_lut_size / 4; i++)
+	for (i = 0; i < adapter->rss_lut_size / 4; i++)
 		wr32(hw, IAVF_VFQF_HLUT(i), dw[i]);
 
 	iavf_flush(hw);

2.51.0

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

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

From: Brian Vazquez <[email protected]>

[ Upstream commit b3d6bbae1d6d5638a4ab702ab195476787cde857 ]

During the IDPF init phase, the mailbox runs in poll mode until it is configured to properly handle interrupts. The previous delay of 300ms is excessively long for the mailbox polling mechanism, which causes a slow initialization of ~2s:

echo 0000:06:12.4 > /sys/bus/pci/drivers/idpf/bind

[   52.444239] idpf 0000:06:12.4: enabling device (0000 -> 0002)
[   52.485005] idpf 0000:06:12.4: Device HW Reset initiated
[   54.177181] idpf 0000:06:12.4: PTP init failed, err=-EOPNOTSUPP
[   54.206177] idpf 0000:06:12.4: Minimum RX descriptor support not provided, using the default
[   54.206182] idpf 0000:06:12.4: Minimum TX descriptor support not provided, using the default

Changing the delay to 300us avoids the delays during the initial mailbox transactions, making the init phase much faster:

[   83.342590] idpf 0000:06:12.4: enabling device (0000 -> 0002)
[   83.384402] idpf 0000:06:12.4: Device HW Reset initiated
[   83.518323] idpf 0000:06:12.4: PTP init failed, err=-EOPNOTSUPP
[   83.547430] idpf 0000:06:12.4: Minimum RX descriptor support not provided, using the default
[   83.547435] idpf 0000:06:12.4: Minimum TX descriptor support not provided, using the default

---
 drivers/net/ethernet/intel/idpf/idpf_lib.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index 371fc5052420..173ddc248867 100644
@@ -1214,7 +1214,7 @@ void idpf_mbx_task(struct work_struct *work)
 		idpf_mb_irq_enable(adapter);
 	else
 		queue_delayed_work(adapter->mbx_wq, &adapter->mbx_task,
-				   msecs_to_jiffies(300));
+				   usecs_to_jiffies(300));
 
 	idpf_recv_mb_msg(adapter);
 }

2.51.0

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

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

From: Herbert Xu <[email protected]>

[ Upstream commit 50fdb78b7c0bcc550910ef69c0984e751cac72fa ]

As soon as crypto_aead_encrypt is called, the underlying request may be freed by an asynchronous completion. Thus dereferencing req->iv after it returns is invalid.

Instead of checking req->iv against info, create a new variable unaligned_info and use it for that purpose instead.

---
 crypto/seqiv.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

index 17e11d51ddc3..04928df0095b 100644
@@ -50,6 +50,7 @@ static int seqiv_aead_encrypt(struct aead_request *req)
 	struct aead_geniv_ctx *ctx = crypto_aead_ctx(geniv);
 	struct aead_request *subreq = aead_request_ctx(req);
 	crypto_completion_t compl;
+	bool unaligned_info;
 	void *data;
 	u8 *info;
 	unsigned int ivsize = 8;

@@ -79,8 +80,9 @@ static int seqiv_aead_encrypt(struct aead_request *req)
 			return err;
 	}
 
-	if (unlikely(!IS_ALIGNED((unsigned long)info,
-				 crypto_aead_alignmask(geniv) + 1))) {
+	unaligned_info = !IS_ALIGNED((unsigned long)info,
+				     crypto_aead_alignmask(geniv) + 1);
+	if (unlikely(unaligned_info)) {
 		info = kmemdup(req->iv, ivsize, req->base.flags &
 			       CRYPTO_TFM_REQ_MAY_SLEEP ? GFP_KERNEL :
 			       GFP_ATOMIC);

@@ -100,7 +102,7 @@ static int seqiv_aead_encrypt(struct aead_request *req)
 	scatterwalk_map_and_copy(info, req->dst, req->assoclen, ivsize, 1);
 
 	err = crypto_aead_encrypt(subreq);
-	if (unlikely(info != req->iv))
+	if (unlikely(unaligned_info))
 		seqiv_aead_encrypt_complete2(req, err);
 	return err;
 }

2.51.0

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

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

From: Raphael Pinsonneault-Thibeault <[email protected]>

[ Upstream commit 252714f1e8bdd542025b16321c790458014d6880 ]

This reverts commit 98921dbd00c4e ("Bluetooth: Use devm_kzalloc in btusb.c file").

In btusb_probe(), we use devm_kzalloc() to allocate the btusb data. This ties the lifetime of all the btusb data to the binding of a driver to one interface, INTF. In a driver that binds to other interfaces, ISOC and DIAG, this is an accident waiting to happen.

The issue is revealed in btusb_disconnect(), where calling usb_driver_release_interface(&btusb_driver, data->intf) will have devm free the data that is also being used by the other interfaces of the driver that may not be released yet.

To fix this, revert the use of devm and go back to freeing memory explicitly.

---
 drivers/bluetooth/btusb.c | 12 +++++++++---
 1 file changed, 9 insertions(+), 3 deletions(-)

index fc7b3e02f14b..603ff13d9f7c 100644
@@ -3835,7 +3835,7 @@ static int btusb_probe(struct usb_interface *intf,
 			return -ENODEV;
 	}
 
-	data = devm_kzalloc(&intf->dev, sizeof(*data), GFP_KERNEL);
+	data = kzalloc(sizeof(*data), GFP_KERNEL);
 	if (!data)
 		return -ENOMEM;
 

@@ -3858,8 +3858,10 @@ static int btusb_probe(struct usb_interface *intf,
 		}
 	}
 
-	if (!data->intr_ep || !data->bulk_tx_ep || !data->bulk_rx_ep)
+	if (!data->intr_ep || !data->bulk_tx_ep || !data->bulk_rx_ep) {
+		kfree(data);
 		return -ENODEV;
+	}
 
 	if (id->driver_info & BTUSB_AMP) {
 		data->cmdreq_type = USB_TYPE_CLASS | 0x01;

@@ -3914,8 +3916,10 @@ static int btusb_probe(struct usb_interface *intf,
 	data->recv_acl = hci_recv_frame;
 
 	hdev = hci_alloc_dev_priv(priv_size);
-	if (!hdev)
+	if (!hdev) {
+		kfree(data);
 		return -ENOMEM;
+	}
 
 	hdev->bus = HCI_USB;
 	hci_set_drvdata(hdev, data);

@@ -4187,6 +4191,7 @@ static int btusb_probe(struct usb_interface *intf,
 	if (data->reset_gpio)
 		gpiod_put(data->reset_gpio);
 	hci_free_dev(hdev);
+	kfree(data);
 	return err;
 }
 

@@ -4235,6 +4240,7 @@ static void btusb_disconnect(struct usb_interface *intf)
 	}
 
 	hci_free_dev(hdev);
+	kfree(data);
 }
 
 #ifdef CONFIG_PM

2.51.0

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

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

From: Jacky Chou <[email protected]>

[ Upstream commit d1a1a4bade4b20c0858d0b2f81d2611de055f675 ]

The Aspeed MDIO controller may return incorrect data when a read operation follows immediately after a write. Due to a controller bug, the subsequent read can latch stale data, causing the polling logic to terminate earlier than expected.

To work around this hardware issue, insert a dummy read after each write operation. This ensures that the next actual read returns the correct data and prevents premature polling exit.

This workaround has been verified to stabilize MDIO transactions on affected Aspeed platforms.

---
 drivers/net/mdio/mdio-aspeed.c | 7 +++++++
 1 file changed, 7 insertions(+)

index c2170650415c..4f2bd20cdc05 100644
@@ -63,6 +63,13 @@ static int aspeed_mdio_op(struct mii_bus *bus, u8 st, u8 op, u8 phyad, u8 regad,
 
 	iowrite32(ctrl, ctx->base + ASPEED_MDIO_CTRL);
 
+	/* Workaround for read-after-write issue.
+	 * The controller may return stale data if a read follows immediately
+	 * after a write. A dummy read forces the hardware to update its
+	 * internal state, ensuring that the next real read returns correct data.
+	 */
+	ioread32(ctx->base + ASPEED_MDIO_CTRL);
+
 	return readl_poll_timeout(ctx->base + ASPEED_MDIO_CTRL, ctrl,
 				!(ctrl & ASPEED_MDIO_CTRL_FIRE),
 				ASPEED_MDIO_INTERVAL_US,

2.51.0

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

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

From: Toke Høiland-Jørgensen <[email protected]>

[ Upstream commit 5498227676303e3ffa9a3a46214af96bc3e81314 ]

The openvswitch teardown code will immediately call ovs_netdev_detach_dev() in response to a NETDEV_UNREGISTER notification. It will then start the dp_notify_work workqueue, which will later end up calling the vport destroy() callback. This callback takes the RTNL to do another ovs_netdev_detach_port(), which in this case is unnecessary. This causes extra pressure on the RTNL, in some cases leading to "unregister_netdevice: waiting for XX to become free" warnings on teardown.

We can straight-forwardly avoid the extra RTNL lock acquisition by checking the device flags before taking the lock, and skip the locking altogether if the IFF_OVS_DATAPATH flag has already been unset.

---
 net/openvswitch/vport-netdev.c | 17 +++++++++++++----
 1 file changed, 13 insertions(+), 4 deletions(-)

index 91a11067e458..6574f9bcdc02 100644
@@ -160,10 +160,19 @@ void ovs_netdev_detach_dev(struct vport *vport)
 
 static void netdev_destroy(struct vport *vport)
 {
-	rtnl_lock();
-	if (netif_is_ovs_port(vport->dev))
-		ovs_netdev_detach_dev(vport);
-	rtnl_unlock();
+	/* When called from ovs_db_notify_wq() after a dp_device_event(), the
+	 * port has already been detached, so we can avoid taking the RTNL by
+	 * checking this first.
+	 */
+	if (netif_is_ovs_port(vport->dev)) {
+		rtnl_lock();
+		/* Check again while holding the lock to ensure we don't race
+		 * with the netdev notifier and detach twice.
+		 */
+		if (netif_is_ovs_port(vport->dev))
+			ovs_netdev_detach_dev(vport);
+		rtnl_unlock();
+	}
 
 	call_rcu(&vport->rcu, vport_netdev_free);
 }

2.51.0

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

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

From: Eric Dumazet <[email protected]>

[ Upstream commit db5b4e39c4e63700c68a7e65fc4e1f1375273476 ]

Over the years, syzbot found many ways to crash the kernel in ip6gre_header() [1].

This involves team or bonding drivers ability to dynamically change their dev->needed_headroom and/or dev->hard_header_len

In this particular crash mld_newpack() allocated an skb with a too small reserve/headroom, and by the time mld_sendpack() was called, syzbot managed to attach an ip6gre device.

[1]
skbuff: skb_under_panic: text:ffffffff8a1d69a8 len:136 put:40 head:ffff888059bc7000 data:ffff888059bc6fe8 tail:0x70 end:0x6c0 dev:team0

------------[ cut here ]------------
 kernel BUG at net/core/skbuff.c:213 !
 <TASK>
  skb_under_panic net/core/skbuff.c:223 [inline]
  skb_push+0xc3/0xe0 net/core/skbuff.c:2641
  ip6gre_header+0xc8/0x790 net/ipv6/ip6_gre.c:1371
  dev_hard_header include/linux/netdevice.h:3436 [inline]
  neigh_connected_output+0x286/0x460 net/core/neighbour.c:1618
  neigh_output include/net/neighbour.h:556 [inline]
  ip6_finish_output2+0xfb3/0x1480 net/ipv6/ip6_output.c:136
 __ip6_finish_output net/ipv6/ip6_output.c:-1 [inline]
  ip6_finish_output+0x234/0x7d0 net/ipv6/ip6_output.c:220
  NF_HOOK_COND include/linux/netfilter.h:307 [inline]
  ip6_output+0x340/0x550 net/ipv6/ip6_output.c:247
  NF_HOOK+0x9e/0x380 include/linux/netfilter.h:318
  mld_sendpack+0x8d4/0xe60 net/ipv6/mcast.c:1855
  mld_send_cr net/ipv6/mcast.c:2154 [inline]
  mld_ifc_work+0x83e/0xd60 net/ipv6/mcast.c:2693

---
 net/ipv6/ip6_gre.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

index 68e9a41eed49..1c186d132fe0 100644
@@ -1395,9 +1395,16 @@ static int ip6gre_header(struct sk_buff *skb, struct net_device *dev,
 {
 	struct ip6_tnl *t = netdev_priv(dev);
 	struct ipv6hdr *ipv6h;
+	int needed;
 	__be16 *p;
 
-	ipv6h = skb_push(skb, t->hlen + sizeof(*ipv6h));
+	needed = t->hlen + sizeof(*ipv6h);
+	if (skb_headroom(skb) < needed &&
+	    pskb_expand_head(skb, HH_DATA_ALIGN(needed - skb_headroom(skb)),
+			     0, GFP_ATOMIC))
+		return -needed;
+
+	ipv6h = skb_push(skb, needed);
 	ip6_flow_hdr(ipv6h, 0, ip6_make_flowlabel(dev_net(dev), skb,
 						  t->fl.u.ip6.flowlabel,
 						  true, &t->fl.u.ip6));

2.51.0

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

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

From: Shravan Kumar Ramani <[email protected]>

[ Upstream commit f13bce715d1600698310a4a7832f6a52499d5395 ]

Some event names have trailing whitespaces at the end which causes programming of counters using the name for these specific events to fail and hence need to be removed.

---
 drivers/platform/mellanox/mlxbf-pmc.c | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)

index 9a0220b4de3c..67d9b19731ed 100644
@@ -796,18 +796,18 @@ static const struct mlxbf_pmc_events mlxbf_pmc_llt_miss_events[] = {
 	{11, "GDC_MISS_MACHINE_CHI_TXDAT"},
 	{12, "GDC_MISS_MACHINE_CHI_RXDAT"},
 	{13, "GDC_MISS_MACHINE_G_FIFO_FF_EXEC0_0"},
-	{14, "GDC_MISS_MACHINE_G_FIFO_FF_EXEC0_1 "},
+	{14, "GDC_MISS_MACHINE_G_FIFO_FF_EXEC0_1"},
 	{15, "GDC_MISS_MACHINE_G_FIFO_FF_EXEC0_2"},
-	{16, "GDC_MISS_MACHINE_G_FIFO_FF_EXEC0_3 "},
-	{17, "GDC_MISS_MACHINE_G_FIFO_FF_EXEC1_0 "},
-	{18, "GDC_MISS_MACHINE_G_FIFO_FF_EXEC1_1 "},
-	{19, "GDC_MISS_MACHINE_G_FIFO_FF_EXEC1_2 "},
-	{20, "GDC_MISS_MACHINE_G_FIFO_FF_EXEC1_3 "},
+	{16, "GDC_MISS_MACHINE_G_FIFO_FF_EXEC0_3"},
+	{17, "GDC_MISS_MACHINE_G_FIFO_FF_EXEC1_0"},
+	{18, "GDC_MISS_MACHINE_G_FIFO_FF_EXEC1_1"},
+	{19, "GDC_MISS_MACHINE_G_FIFO_FF_EXEC1_2"},
+	{20, "GDC_MISS_MACHINE_G_FIFO_FF_EXEC1_3"},
 	{21, "GDC_MISS_MACHINE_G_FIFO_FF_EXEC_DONE0_0"},
 	{22, "GDC_MISS_MACHINE_G_FIFO_FF_EXEC_DONE0_1"},
 	{23, "GDC_MISS_MACHINE_G_FIFO_FF_EXEC_DONE0_2"},
 	{24, "GDC_MISS_MACHINE_G_FIFO_FF_EXEC_DONE0_3"},
-	{25, "GDC_MISS_MACHINE_G_FIFO_FF_EXEC_DONE1_0 "},
+	{25, "GDC_MISS_MACHINE_G_FIFO_FF_EXEC_DONE1_0"},
 	{26, "GDC_MISS_MACHINE_G_FIFO_FF_EXEC_DONE1_1"},
 	{27, "GDC_MISS_MACHINE_G_FIFO_FF_EXEC_DONE1_2"},
 	{28, "GDC_MISS_MACHINE_G_FIFO_FF_EXEC_DONE1_3"},

2.51.0

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

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

From: Thomas Fourier <[email protected]>

[ Upstream commit 1461209cf813b6ee6d40f29b96b544587df6d2b1 ]

A sysfs group is created in msi_init() when old_ec_model is enabled, but never removed. Remove the msipf_old_attribute_group in that case.

---
 drivers/platform/x86/msi-laptop.c | 3 +++
 1 file changed, 3 insertions(+)

index e5391a37014d..db3dadd29b29 100644
@@ -1130,6 +1130,9 @@ static void __exit msi_cleanup(void)
 	sysfs_remove_group(&msipf_device->dev.kobj, &msipf_attribute_group);
 	if (!quirks->old_ec_model && threeg_exists)
 		device_remove_file(&msipf_device->dev, &dev_attr_threeg);
+	if (quirks->old_ec_model)
+		sysfs_remove_group(&msipf_device->dev.kobj,
+				   &msipf_old_attribute_group);
 	platform_device_unregister(msipf_device);
 	platform_driver_unregister(&msipf_driver);
 	backlight_device_unregister(msibl_device);

2.51.0

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

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

From: Junrui Luo <[email protected]>

[ Upstream commit 15dd100349b8526cbdf2de0ce3e72e700eb6c208 ]

The ibm_rtl_init() function searches for the signature but has a pointer arithmetic error. The loop counter suggests searching at 4-byte intervals but the implementation only advances by 1 byte per iteration.

Fix by properly advancing the pointer by sizeof(unsigned int) bytes each iteration.

---
 drivers/platform/x86/ibm_rtl.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index 231b37909801..139956168cf9 100644
@@ -273,7 +273,7 @@ static int __init ibm_rtl_init(void) {
 	/* search for the _RTL_ signature at the start of the table */
 	for (i = 0 ; i < ebda_size/sizeof(unsigned int); i++) {
 		struct ibm_rtl_table __iomem * tmp;
-		tmp = (struct ibm_rtl_table __iomem *) (ebda_map+i);
+		tmp = (struct ibm_rtl_table __iomem *) (ebda_map + i*sizeof(unsigned int));
 		if ((readq(&tmp->signature) & RTL_MASK) == RTL_SIGNATURE) {
 			phys_addr_t addr;
 			unsigned int plen;

2.51.0

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

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

From: Jiri Pirko <[email protected]>

[ Upstream commit 932ac51d9953eaf77a1252f79b656d4ca86163c6 ]

There has been a syzkaller bug reported recently with the following trace:

list_del corruption, ffff888058bea080->prev is LIST_POISON2 (dead000000000122)

------------[ cut here ]------------
kernel BUG at lib/list_debug.c:59!
Oops: invalid opcode: 0000 [#1] SMP KASAN NOPTI
CPU: 3 UID: 0 PID: 21246 Comm: syz.0.2928 Not tainted syzkaller #0 PREEMPT(full)
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.3-debian-1.16.3-2~bpo12+1 04/01/2014
RIP: 0010:__list_del_entry_valid_or_report+0x13e/0x200 lib/list_debug.c:59
Code: 48 c7 c7 e0 71 f0 8b e8 30 08 ef fc 90 0f 0b 48 89 ef e8 a5 02 55 fd 48 89 ea 48 89 de 48 c7 c7 40 72 f0 8b e8 13 08 ef fc 90 <0f> 0b 48 89 ef e8 88 02 55 fd 48 89 ea 48 b8 00 00 00 00 00 fc ff
RSP: 0018:ffffc9000d49f370 EFLAGS: 00010286
RAX: 000000000000004e RBX: ffff888058bea080 RCX: ffffc9002817d000
RDX: 0000000000000000 RSI: ffffffff819becc6 RDI: 0000000000000005
RBP: dead000000000122 R08: 0000000000000005 R09: 0000000000000000
R10: 0000000080000000 R11: 0000000000000001 R12: ffff888039e9c230
R13: ffff888058bea088 R14: ffff888058bea080 R15: ffff888055461480
FS:  00007fbbcfe6f6c0(0000) GS:ffff8880d6d0a000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 000000110c3afcb0 CR3: 00000000382c7000 CR4: 0000000000352ef0
Call Trace:
 <TASK>
 __list_del_entry_valid include/linux/list.h:132 [inline]
 __list_del_entry include/linux/list.h:223 [inline]
 list_del_rcu include/linux/rculist.h:178 [inline]
 __team_queue_override_port_del drivers/net/team/team_core.c:826 [inline]
 __team_queue_override_port_del drivers/net/team/team_core.c:821 [inline]
 team_queue_override_port_prio_changed drivers/net/team/team_core.c:883 [inline]
 team_priority_option_set+0x171/0x2f0 drivers/net/team/team_core.c:1534
 team_option_set drivers/net/team/team_core.c:376 [inline]
 team_nl_options_set_doit+0x8ae/0xe60 drivers/net/team/team_core.c:2653
 genl_family_rcv_msg_doit+0x209/0x2f0 net/netlink/genetlink.c:1115
 genl_family_rcv_msg net/netlink/genetlink.c:1195 [inline]
 genl_rcv_msg+0x55c/0x800 net/netlink/genetlink.c:1210
 netlink_rcv_skb+0x158/0x420 net/netlink/af_netlink.c:2552
 genl_rcv+0x28/0x40 net/netlink/genetlink.c:1219
 netlink_unicast_kernel net/netlink/af_netlink.c:1320 [inline]
 netlink_unicast+0x5aa/0x870 net/netlink/af_netlink.c:1346
 netlink_sendmsg+0x8c8/0xdd0 net/netlink/af_netlink.c:1896
 sock_sendmsg_nosec net/socket.c:727 [inline]
 __sock_sendmsg net/socket.c:742 [inline]
 ____sys_sendmsg+0xa98/0xc70 net/socket.c:2630
 ___sys_sendmsg+0x134/0x1d0 net/socket.c:2684
 __sys_sendmsg+0x16d/0x220 net/socket.c:2716
 do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
 do_syscall_64+0xcd/0xfa0 arch/x86/entry/syscall_64.c:94
 entry_SYSCALL_64_after_hwframe+0x77/0x7f

The problem is in this flow:
1) Port is enabled, queue_id != 0, in qom_list
2) Port gets disabled

        -> team_port_disable()
        -> team_queue_override_port_del()
        -> del (removed from list)
3) Port is disabled, queue_id != 0, not in any list
4) Priority changes
        -> team_queue_override_port_prio_changed()
        -> checks: port disabled && queue_id != 0
        -> calls del - hits the BUG as it is removed already

To fix this, change the check in team_queue_override_port_prio_changed() so it returns early if port is not enabled.

---
 drivers/net/team/team_core.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index 94c40c5cebdd..50732f9699ee 100644
@@ -877,7 +877,7 @@ static void __team_queue_override_enabled_check(struct team *team)
 static void team_queue_override_port_prio_changed(struct team *team,
 						  struct team_port *port)
 {
-	if (!port->queue_id || team_port_enabled(port))
+	if (!port->queue_id || !team_port_enabled(port))
 		return;
 	__team_queue_override_port_del(team, port);
 	__team_queue_override_port_add(team, port);

2.51.0

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

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

From: Vladimir Oltean <[email protected]>

[ Upstream commit a9f96dc59b4a50ffbf86158f315e115969172d48 ]

of_find_net_device_by_node() searches net devices by their /sys/class/net/, entry. It is documented in its kernel-doc that:

* If successful, returns a pointer to the net_device with the embedded
* struct device refcount incremented by one, or NULL on failure. The
* refcount must be dropped when done with the net_device.

We are missing a put_device(&conduit->dev) which we could place at the end of dsa_tree_find_first_conduit(). But to explain why calling put_device() right away is safe is the same as to explain why the chosen solution is different.

The code is very poorly split: dsa_tree_find_first_conduit() was first introduced in commit 95f510d0b792 ("net: dsa: allow the DSA master to be seen and changed through rtnetlink") but was first used several commits later, in commit acc43b7bf52a ("net: dsa: allow masters to join a LAG").

Assume there is a switch with 2 CPU ports and 2 conduits, eno2 and eno3. When we create a LAG (bonding or team device) and place eno2 and eno3 beneath it, we create a 3rd conduit (the LAG device itself), but this is slightly different than the first two.

Namely, the cpu_dp->conduit pointer of the CPU ports does not change, and remains pointing towards the physical Ethernet controllers which are now LAG ports. Only 2 things change:
- the LAG device has a dev->dsa_ptr which marks it as a DSA conduit
- dsa_port_to_conduit(user port) finds the LAG and not the physical conduit, because of the dp->cpu_port_in_lag bit being set.

When the LAG device is destroyed, dsa_tree_migrate_ports_from_lag_conduit() is called and this is where dsa_tree_find_first_conduit() kicks in.

This is the logical mistake and the reason why introducing code in one patch and using it from another is bad practice. I didn't realize that I don't have to call of_find_net_device_by_node() again; the cpu_dp->conduit association was never undone, and is still available for direct (re)use. There's only one concern - maybe the conduit disappeared in the meantime, but the netdev_hold() call we made during dsa_port_parse_cpu() (see previous change) ensures that this was not the case.

Therefore, fixing the code means reimplementing it in the simplest way.

I am blaming the time of use, since this is what "git blame" would show if we were to monitor for the conduit's kobject's refcount remaining elevated instead of being freed.

Tested on the NXP LS1028A, using the steps from Documentation/networking/dsa/configuration.rst section "Affinity of user ports to CPU ports", followed by (extra prints added by me):

$ ip link del bond0
mscc_felix 0000:00:00.5 swp3: Link is Down
bond0 (unregistering): (slave eno2): Releasing backup interface
fsl_enetc 0000:00:00.2 eno2: Link is Down
mscc_felix 0000:00:00.5 swp0: bond0 disappeared, migrating to eno2
mscc_felix 0000:00:00.5 swp1: bond0 disappeared, migrating to eno2
mscc_felix 0000:00:00.5 swp2: bond0 disappeared, migrating to eno2
mscc_felix 0000:00:00.5 swp3: bond0 disappeared, migrating to eno2

---
 net/dsa/dsa.c | 8 +-------
 1 file changed, 1 insertion(+), 7 deletions(-)

index ac3a252969cb..97599e0d5a1d 100644
@@ -366,16 +366,10 @@ static struct dsa_port *dsa_tree_find_first_cpu(struct dsa_switch_tree *dst)
 
 struct net_device *dsa_tree_find_first_conduit(struct dsa_switch_tree *dst)
 {
-	struct device_node *ethernet;
-	struct net_device *conduit;
 	struct dsa_port *cpu_dp;
 
 	cpu_dp = dsa_tree_find_first_cpu(dst);
-	ethernet = of_parse_phandle(cpu_dp->dn, "ethernet", 0);
-	conduit = of_find_net_device_by_node(ethernet);
-	of_node_put(ethernet);
-
-	return conduit;
+	return cpu_dp->conduit;
 }
 
 /* Assign the default CPU port (the first one in the tree) to all ports of the

2.51.0

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

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

From: Raju Rangoju <[email protected]>

[ Upstream commit df60c332caf95d70f967aeace826e7e2f0847361 ]

During the stress tests, early RX adaptation handshakes can fail, such as missing the RX_ADAPT ACK or not receiving a coefficient update before block lock is established. Continuing to retry RX adaptation in this state is often ineffective if the current mode selection is not viable.

Resetting the RX adaptation retry counter when an RX_ADAPT request fails to receive ACK or a coefficient update prior to block lock, and clearing mode_set so the next bring-up performs a fresh mode selection rather than looping on a likely invalid configuration.

---
 drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c | 2 ++
 1 file changed, 2 insertions(+)

index 32e633d11348..6d2c401bb246 100644
@@ -2036,6 +2036,7 @@ static void xgbe_set_rx_adap_mode(struct xgbe_prv_data *pdata,
 {
 	if (pdata->rx_adapt_retries++ >= MAX_RX_ADAPT_RETRIES) {
 		pdata->rx_adapt_retries = 0;
+		pdata->mode_set = false;
 		return;
 	}
 

@@ -2082,6 +2083,7 @@ static void xgbe_rx_adaptation(struct xgbe_prv_data *pdata)
 		 */
 		netif_dbg(pdata, link, pdata->netdev, "Block_lock done");
 		pdata->rx_adapt_done = true;
+		pdata->rx_adapt_retries = 0;
 		pdata->mode_set = false;
 		return;
 	}

2.51.0

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

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

From: Deepakkumar Karn <[email protected]>

[ Upstream commit 12cab1191d9890097171156d06bfa8d31f1e39c8 ]

In async_set_registers(), when usb_submit_urb() fails, the allocated
async_req structure and URB are not freed, causing a memory leak.

The completion callback async_set_reg_cb() is responsible for freeing these allocations, but it is only called after the URB is successfully submitted and completes (successfully or with error). If submission fails, the callback never runs and the memory is leaked.

Fix this by freeing both the URB and the request structure in the error path when usb_submit_urb() fails.

---
 drivers/net/usb/rtl8150.c | 2 ++
 1 file changed, 2 insertions(+)

index 278e6cb6f4d9..e40b0669d9f4 100644
@@ -211,6 +211,8 @@ static int async_set_registers(rtl8150_t *dev, u16 indx, u16 size, u16 reg)
 		if (res == -ENODEV)
 			netif_device_detach(dev->netdev);
 		dev_err(&dev->udev->dev, "%s failed with %d\n", __func__, res);
+		kfree(req);
+		usb_free_urb(async_urb);
 	}
 	return res;
 }

2.51.0

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

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

From: Alice C. Munduruca <[email protected]>

[ Upstream commit 472c5dd6b95c02b3e5d7395acf542150e91165e7 ]

When the selftest 'tap.c' is compiled with '-D_FORTIFY_SOURCE=3', the strcpy() in rtattr_add_strsz() is replaced with a checked version which causes the test to consistently fail when compiled with toolchains for which this option is enabled by default.

TAP version 13
1..3
# Starting 3 tests from 1 test cases.
# RUN tap.test_packet_valid_udp_gso ...
*** buffer overflow detected ***: terminated
# test_packet_valid_udp_gso: Test terminated by assertion
# FAIL tap.test_packet_valid_udp_gso
not ok 1 tap.test_packet_valid_udp_gso
# RUN tap.test_packet_valid_udp_csum ...
*** buffer overflow detected ***: terminated
# test_packet_valid_udp_csum: Test terminated by assertion
# FAIL tap.test_packet_valid_udp_csum
not ok 2 tap.test_packet_valid_udp_csum
# RUN tap.test_packet_crash_tap_invalid_eth_proto ...
*** buffer overflow detected ***: terminated
# test_packet_crash_tap_invalid_eth_proto: Test terminated by assertion
# FAIL tap.test_packet_crash_tap_invalid_eth_proto
not ok 3 tap.test_packet_crash_tap_invalid_eth_proto
# FAILED: 0 / 3 tests passed.
# Totals: pass:0 fail:3 xfail:0 xpass:0 skip:0 error:0

A buffer overflow is detected by the fortified glibc __strcpy_chk() since the __builtin_object_size() of RTA_DATA(rta) is incorrectly reported as 1, even though there is ample space in its bounding buffer req.

Additionally, given that IFLA_IFNAME also expects a null-terminated string, callers of rtaddr_add_str{,sz}() could simply use the rtaddr_add_strsz() variant. (which has been renamed to remove the trailing sz) memset() has been used for this function since it is unchecked and thus circumvents the issue discussed in the previous paragraph.

---
 tools/testing/selftests/net/tap.c | 16 +++++-----------
 1 file changed, 5 insertions(+), 11 deletions(-)

index 247c3b3ac1c9..51a209014f1c 100644
@@ -56,18 +56,12 @@ static void rtattr_end(struct nlmsghdr *nh, struct rtattr *attr)
 static struct rtattr *rtattr_add_str(struct nlmsghdr *nh, unsigned short type,
 				     const char *s)
 {
-	struct rtattr *rta = rtattr_add(nh, type, strlen(s));
+	unsigned int strsz = strlen(s) + 1;
+	struct rtattr *rta;
 
-	memcpy(RTA_DATA(rta), s, strlen(s));
-	return rta;
-}
-
-static struct rtattr *rtattr_add_strsz(struct nlmsghdr *nh, unsigned short type,
-				       const char *s)
-{
-	struct rtattr *rta = rtattr_add(nh, type, strlen(s) + 1);
+	rta = rtattr_add(nh, type, strsz);
 
-	strcpy(RTA_DATA(rta), s);
+	memcpy(RTA_DATA(rta), s, strsz);
 	return rta;
 }
 

@@ -119,7 +113,7 @@ static int dev_create(const char *dev, const char *link_type,
 
 	link_info = rtattr_begin(&req.nh, IFLA_LINKINFO);
 
-	rtattr_add_strsz(&req.nh, IFLA_INFO_KIND, link_type);
+	rtattr_add_str(&req.nh, IFLA_INFO_KIND, link_type);
 
 	if (fill_info_data) {
 		info_data = rtattr_begin(&req.nh, IFLA_INFO_DATA);

2.51.0

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

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

From: Yeoreum Yun <[email protected]>

[ Upstream commit 6402078bd9d1ed46e79465e1faaa42e3458f8a33 ]

When smc91x.c is built with PREEMPT_RT, the following splat occurs in FVP_RevC:

[   13.055000] smc91x LNRO0003:00 eth0: link up, 10Mbps, half-duplex, lpa 0x0000
[   13.062137] BUG: workqueue leaked atomic, lock or RCU: kworker/2:1[106]
[   13.062137]      preempt=0x00000000 lock=0->0 RCU=0->1 workfn=mld_ifc_work
[   13.062266] C
** replaying previous printk message **
[   13.062266] CPU: 2 UID: 0 PID: 106 Comm: kworker/2:1 Not tainted 6.18.0-dirty #179 PREEMPT_{RT,(full)}
[   13.062353] Hardware name:  , BIOS
[   13.062382] Workqueue: mld mld_ifc_work
[   13.062469] Call trace:
[   13.062494]  show_stack+0x24/0x40 (C)
[   13.062602]  __dump_stack+0x28/0x48
[   13.062710]  dump_stack_lvl+0x7c/0xb0
[   13.062818]  dump_stack+0x18/0x34
[   13.062926]  process_scheduled_works+0x294/0x450
[   13.063043]  worker_thread+0x260/0x3d8
[   13.063124]  kthread+0x1c4/0x228
[   13.063235]  ret_from_fork+0x10/0x20

This happens because smc_special_trylock() disables IRQs even on PREEMPT_RT,
but smc_special_unlock() does not restore IRQs on PREEMPT_RT.
The reason is that smc_special_unlock() calls spin_unlock_irqrestore(),
and rcu_read_unlock_bh() in __dev_queue_xmit() cannot invoke
rcu_read_unlock() through __local_bh_enable_ip() when current->softirq_disable_cnt becomes zero.

To address this issue, replace smc_special_trylock() with spin_trylock_irqsave().

---
 drivers/net/ethernet/smsc/smc91x.c | 10 +---------
 1 file changed, 1 insertion(+), 9 deletions(-)

index a5e23e2da90f..953a1d22e60a 100644
@@ -516,15 +516,7 @@ static inline void  smc_rcv(struct net_device *dev)
  * any other concurrent access and C would always interrupt B. But life
  * isn't that easy in a SMP world...
  */
-#define smc_special_trylock(lock, flags)				\
-({									\
-	int __ret;							\
-	local_irq_save(flags);						\
-	__ret = spin_trylock(lock);					\
-	if (!__ret)							\
-		local_irq_restore(flags);				\
-	__ret;								\
-})
+#define smc_special_trylock(lock, flags)	spin_trylock_irqsave(lock, flags)
 #define smc_special_lock(lock, flags)		spin_lock_irqsave(lock, flags)
 #define smc_special_unlock(lock, flags) 	spin_unlock_irqrestore(lock, flags)
 #else

2.51.0

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

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

From: Andrew Morton <[email protected]>

[ Upstream commit 5393802c94e0ab1295c04c94c57bcb00222d4674 ]

WARNING: include/linux/genalloc.h:52 function parameter 'start_addr' not described in 'genpool_algo_t'

---
 include/linux/genalloc.h | 1 +
 1 file changed, 1 insertion(+)

index 0bd581003cd5..60de63e46b33 100644
@@ -44,6 +44,7 @@ struct gen_pool;
  * @nr: The number of zeroed bits we're looking for
  * @data: optional additional data used by the callback
  * @pool: the pool being allocated from
+ * @start_addr: start address of memory chunk
  */
 typedef unsigned long (*genpool_algo_t)(unsigned long *map,
 			unsigned long size,

2.51.0

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

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

From: Thomas Fourier <[email protected]>

[ Upstream commit c48c0fd0e19684b6ecdb4108a429e3a4e73f5e21 ]

It looks like the buffer allocated and mapped in add_card() is done with size RCV_BUFFER_SIZE which is 16 KB and 4KB.

---
 drivers/firewire/nosy.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

index ea31ac7ac1ca..e59053738a43 100644
@@ -36,6 +36,8 @@
 
 static char driver_name[] = KBUILD_MODNAME;
 
+#define RCV_BUFFER_SIZE (16 * 1024)
+
 /* this is the physical layout of a PCL, its size is 128 bytes */
 struct pcl {
 	__le32 next;

@@ -517,16 +519,14 @@ remove_card(struct pci_dev *dev)
 			  lynx->rcv_start_pcl, lynx->rcv_start_pcl_bus);
 	dma_free_coherent(&lynx->pci_device->dev, sizeof(struct pcl),
 			  lynx->rcv_pcl, lynx->rcv_pcl_bus);
-	dma_free_coherent(&lynx->pci_device->dev, PAGE_SIZE, lynx->rcv_buffer,
-			  lynx->rcv_buffer_bus);
+	dma_free_coherent(&lynx->pci_device->dev, RCV_BUFFER_SIZE,
+			  lynx->rcv_buffer, lynx->rcv_buffer_bus);
 
 	iounmap(lynx->registers);
 	pci_disable_device(dev);
 	lynx_put(lynx);
 }
 
-#define RCV_BUFFER_SIZE (16 * 1024)
-
 static int
 add_card(struct pci_dev *dev, const struct pci_device_id *unused)
 {

@@ -680,7 +680,7 @@ add_card(struct pci_dev *dev, const struct pci_device_id *unused)
 		dma_free_coherent(&lynx->pci_device->dev, sizeof(struct pcl),
 				  lynx->rcv_pcl, lynx->rcv_pcl_bus);
 	if (lynx->rcv_buffer)
-		dma_free_coherent(&lynx->pci_device->dev, PAGE_SIZE,
+		dma_free_coherent(&lynx->pci_device->dev, RCV_BUFFER_SIZE,
 				  lynx->rcv_buffer, lynx->rcv_buffer_bus);
 	iounmap(lynx->registers);
 

2.51.0

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

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

From: Jonas Gorski <[email protected]>

[ Upstream commit d42bce414d1c5c0b536758466a1f63ac358e613c ]

port_fdb_dump() is supposed to only add fdb entries, but we iterate over the full ARL table, which also includes multicast entries.

So check if the entry is a multicast entry before passing it on to the callback().

Additionally, the port of those entries is a bitmask, not a port number, so any included entries would have even be for the wrong port.

---
 drivers/net/dsa/b53/b53_common.c | 3 +++
 1 file changed, 3 insertions(+)

index 01eb62706412..0b666a77ea97 100644
@@ -1972,6 +1972,9 @@ static int b53_fdb_copy(int port, const struct b53_arl_entry *ent,
 	if (!ent->is_valid)
 		return 0;
 
+	if (is_multicast_ether_addr(ent->mac))
+		return 0;
+
 	if (port != ent->port)
 		return 0;
 

2.51.0

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

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

From: Thomas De Schampheleire <[email protected]>

[ Upstream commit b08fc4d0ec2466558f6d5511434efdfabbddf2a6 ]

Since commit e7e2941300d2 ("kbuild: split device tree build rules into scripts/Makefile.dtbs"), it is no longer possible to compile a device tree blob that is not specified in a make rule
like:

    dtb-$(CONFIG_FOO) += foo.dtb

Before the mentioned commit, one could copy a dts file to e.g. arch/arm64/boot/dts/ (or a new subdirectory) and then convert it to a dtb file using:
make ARCH=arm64 foo.dtb

In this scenario, both 'dtb-y' and 'dtb-' are empty, and the inclusion of scripts/Makefile.dtbs relies on 'targets' to contain the MAKECMDGOALS. The value of 'targets', however, is only final later in the code.

Move the conditional include of scripts/Makefile.dtbs down to where the value of 'targets' is final. Since Makefile.dtbs updates 'always-y' which is used as a prerequisite in the build rule, the build rule also needs to move down.

---
 scripts/Makefile.build | 26 ++++++++++++++------------
 1 file changed, 14 insertions(+), 12 deletions(-)

index 2c5c1a214f3b..6e07023b5442 100644
@@ -449,18 +449,6 @@ ifneq ($(userprogs),)
 include $(srctree)/scripts/Makefile.userprogs
 endif
 
-ifneq ($(need-dtbslist)$(dtb-y)$(dtb-)$(filter %.dtb %.dtb.o %.dtbo.o,$(targets)),)
-include $(srctree)/scripts/Makefile.dtbs
-endif
-
-# Build
-# ---------------------------------------------------------------------------
-
-$(obj)/: $(if $(KBUILD_BUILTIN), $(targets-for-builtin)) \
-	 $(if $(KBUILD_MODULES), $(targets-for-modules)) \
-	 $(subdir-ym) $(always-y)
-	@:
-
 # Single targets
 # ---------------------------------------------------------------------------
 

@@ -490,6 +478,20 @@ FORCE:
 targets += $(filter-out $(single-subdir-goals), $(MAKECMDGOALS))
 targets := $(filter-out $(PHONY), $(targets))
 
+# Now that targets is fully known, include dtb rules if needed
+ifneq ($(need-dtbslist)$(dtb-y)$(dtb-)$(filter %.dtb %.dtb.o %.dtbo.o,$(targets)),)
+include $(srctree)/scripts/Makefile.dtbs
+endif
+
+# Build
+# Needs to be after the include of Makefile.dtbs, which updates always-y
+# ---------------------------------------------------------------------------
+
+$(obj)/: $(if $(KBUILD_BUILTIN), $(targets-for-builtin)) \
+	 $(if $(KBUILD_MODULES), $(targets-for-modules)) \
+	 $(subdir-ym) $(always-y)
+	@:
+
 # Read all saved command lines and dependencies for the $(targets) we
 # may be building above, using $(if_changed{,_dep}). As an
 # optimization, we don't need to read them if the target does not

2.51.0

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

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

From: Deepanshu Kartikey <[email protected]>

[ Upstream commit a1e077a3f76eea0dc671ed6792e7d543946227e8 ]

The ASIX driver reads the PHY address from the USB device via asix_read_phy_addr(). A malicious or faulty device can return an invalid address (>= PHY_MAX_ADDR), which causes a warning in mdiobus_get_phy():

addr 207 out of range

  WARNING: drivers/net/phy/mdio_bus.c:76

Validate the PHY address in asix_read_phy_addr() and remove the now-redundant check in ax88172a.c.

---
 drivers/net/usb/asix_common.c | 5 +++++
 1 file changed, 5 insertions(+)

index 7fd763917ae2..6ab3486072cb 100644
@@ -335,6 +335,11 @@ int asix_read_phy_addr(struct usbnet *dev, bool internal)
 	offset = (internal ? 1 : 0);
 	ret = buf[offset];
 
+	if (ret >= PHY_MAX_ADDR) {
+		netdev_err(dev->net, "invalid PHY address: %d\n", ret);
+		return -ENODEV;
+	}
+
 	netdev_dbg(dev->net, "%s PHY address 0x%x\n",
 		   internal ? "internal" : "external", ret);
 

2.51.0

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

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

From: Bagas Sanjaya <[email protected]>

[ Upstream commit f79f9b7ace1713e4b83888c385f5f55519dfb687 ]

Sphinx reports kernel-doc warning:

WARNING: ./net/bridge/br_private.h:267 struct member 'tunnel_hash' not described in 'net_bridge_vlan_group'

Fix it by describing @tunnel_hash member.

---
 net/bridge/br_private.h | 1 +
 1 file changed, 1 insertion(+)

index 741b0b8c4bab..a2e59108a5dc 100644
@@ -247,6 +247,7 @@ struct net_bridge_vlan {
  * struct net_bridge_vlan_group
  *
  * @vlan_hash: VLAN entry rhashtable
+ * @tunnel_hash: Hash table to map from tunnel key ID (e.g. VXLAN VNI) to VLAN
  * @vlan_list: sorted VLAN entry list
  * @num_vlans: number of total VLAN entries
  * @pvid: PVID VLAN id

2.51.0

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

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

From: Zilin Guan <[email protected]>

[ Upstream commit 665077d78dc7941ce6a330c02023a2b469cc8cc7 ]

pds_vfio_dirty_enable() allocates memory for region_info. If interval_tree_iter_first() returns NULL, the function returns -EINVAL immediately without freeing the allocated memory, causing a memory leak.

Fix this by jumping to the out_free_region_info label to ensure region_info is freed.

---
 drivers/vfio/pci/pds/dirty.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

index 481992142f79..4915a7c1c491 100644
@@ -292,8 +292,11 @@ static int pds_vfio_dirty_enable(struct pds_vfio_pci_device *pds_vfio,
 	len = num_ranges * sizeof(*region_info);
 
 	node = interval_tree_iter_first(ranges, 0, ULONG_MAX);
-	if (!node)
-		return -EINVAL;
+	if (!node) {
+		err = -EINVAL;
+		goto out_free_region_info;
+	}
+
 	for (int i = 0; i < num_ranges; i++) {
 		struct pds_lm_dirty_region_info *ri = &region_info[i];
 		u64 region_size = node->last - node->start + 1;

2.51.0

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

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

From: Junrui Luo <[email protected]>

[ Upstream commit e44c42c830b7ab36e3a3a86321c619f24def5206 ]

The hp_populate_*_elements_from_package() functions in the hp-bioscfg driver contain out-of-bounds array access vulnerabilities.

These functions parse ACPI packages into internal data structures using a for loop with index variable 'elem' that iterates through enum_obj/integer_obj/order_obj/password_obj/string_obj arrays.

When processing multi-element fields like PREREQUISITES and ENUM_POSSIBLE_VALUES, these functions read multiple consecutive array elements using expressions like 'enum_obj[elem + reqs]' and 'enum_obj[elem + pos_values]' within nested loops.

The bug is that the bounds check only validated elem, but did not consider the additional offset when accessing elem + reqs or elem + pos_values.

The fix changes the bounds check to validate the actual accessed index.

---
 drivers/platform/x86/hp/hp-bioscfg/enum-attributes.c       | 4 ++--
 drivers/platform/x86/hp/hp-bioscfg/int-attributes.c        | 2 +-
 drivers/platform/x86/hp/hp-bioscfg/order-list-attributes.c | 5 +++++
 drivers/platform/x86/hp/hp-bioscfg/passwdobj-attributes.c  | 5 +++++
 drivers/platform/x86/hp/hp-bioscfg/string-attributes.c     | 2 +-
 5 files changed, 14 insertions(+), 4 deletions(-)

index c50ad5880503..f346aad8e9d8 100644
@@ -207,7 +207,7 @@ static int hp_populate_enumeration_elements_from_package(union acpi_object *enum
 		case PREREQUISITES:
 			size = min_t(u32, enum_data->common.prerequisites_size, MAX_PREREQUISITES_SIZE);
 			for (reqs = 0; reqs < size; reqs++) {
-				if (elem >= enum_obj_count) {
+				if (elem + reqs >= enum_obj_count) {
 					pr_err("Error enum-objects package is too small\n");
 					return -EINVAL;
 				}

@@ -255,7 +255,7 @@ static int hp_populate_enumeration_elements_from_package(union acpi_object *enum
 
 			for (pos_values = 0; pos_values < size && pos_values < MAX_VALUES_SIZE;
 			     pos_values++) {
-				if (elem >= enum_obj_count) {
+				if (elem + pos_values >= enum_obj_count) {
 					pr_err("Error enum-objects package is too small\n");
 					return -EINVAL;
 				}

index 6c7f4d5fa9cb..63b1fda2be4e 100644
@@ -227,7 +227,7 @@ static int hp_populate_integer_elements_from_package(union acpi_object *integer_
 			size = min_t(u32, integer_data->common.prerequisites_size, MAX_PREREQUISITES_SIZE);
 
 			for (reqs = 0; reqs < size; reqs++) {
-				if (elem >= integer_obj_count) {
+				if (elem + reqs >= integer_obj_count) {
 					pr_err("Error elem-objects package is too small\n");
 					return -EINVAL;
 				}

index c6e57bb9d8b7..6a31f47ce3f5 100644
@@ -216,6 +216,11 @@ static int hp_populate_ordered_list_elements_from_package(union acpi_object *ord
 			size = min_t(u32, ordered_list_data->common.prerequisites_size,
 				     MAX_PREREQUISITES_SIZE);
 			for (reqs = 0; reqs < size; reqs++) {
+				if (elem + reqs >= order_obj_count) {
+					pr_err("Error elem-objects package is too small\n");
+					return -EINVAL;
+				}
+
 				ret = hp_convert_hexstr_to_str(order_obj[elem + reqs].string.pointer,
 							       order_obj[elem + reqs].string.length,
 							       &str_value, &value_len);

index 35936c05e45b..a5c457d06b9c 100644
@@ -303,6 +303,11 @@ static int hp_populate_password_elements_from_package(union acpi_object *passwor
 				     MAX_PREREQUISITES_SIZE);
 
 			for (reqs = 0; reqs < size; reqs++) {
+				if (elem + reqs >= password_obj_count) {
+					pr_err("Error elem-objects package is too small\n");
+					return -EINVAL;
+				}
+
 				ret = hp_convert_hexstr_to_str(password_obj[elem + reqs].string.pointer,
 							       password_obj[elem + reqs].string.length,
 							       &str_value, &value_len);

index 27758b779b2d..7b885d25650c 100644
@@ -217,7 +217,7 @@ static int hp_populate_string_elements_from_package(union acpi_object *string_ob
 				     MAX_PREREQUISITES_SIZE);
 
 			for (reqs = 0; reqs < size; reqs++) {
-				if (elem >= string_obj_count) {
+				if (elem + reqs >= string_obj_count) {
 					pr_err("Error elem-objects package is too small\n");
 					return -EINVAL;
 				}

2.51.0

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

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

From: Anshumali Gaur <[email protected]>

[ Upstream commit 85f4b0c650d9f9db10bda8d3acfa1af83bf78cf7 ]

This patch ensures that the RX ring size (rx_pending) is not set below the permitted length. This avoids UBSAN shift-out-of-bounds errors when users passes small or zero ring sizes via ethtool -G.

---
 drivers/net/ethernet/marvell/octeontx2/nic/otx2_ethtool.c | 8 ++++++++
 1 file changed, 8 insertions(+)

index 5197ce816581..cc6a63e2573f 100644
@@ -432,6 +432,14 @@ static int otx2_set_ringparam(struct net_device *netdev,
 	 */
 	if (rx_count < pfvf->hw.rq_skid)
 		rx_count =  pfvf->hw.rq_skid;
+
+	if (ring->rx_pending < 16) {
+		netdev_err(netdev,
+			   "rx ring size %u invalid, min is 16\n",
+			   ring->rx_pending);
+		return -EINVAL;
+	}
+
 	rx_count = Q_COUNT(Q_SIZE(rx_count, 3));
 
 	/* Due pipelining impact minimum 2000 unused SQ CQE's

2.51.0

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

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

From: Wei Fang <[email protected]>

[ Upstream commit a48e232210009be50591fdea8ba7c07b0f566a13 ]

There is a crash issue when running zero copy XDP_TX action, the crash log is shown below.

[  216.122464] Unable to handle kernel paging request at virtual address fffeffff80000000
[  216.187524] Internal error: Oops: 0000000096000144 [#1]  SMP
[  216.301694] Call trace:
[  216.304130]  dcache_clean_poc+0x20/0x38 (P)
[  216.308308]  __dma_sync_single_for_device+0x1bc/0x1e0
[  216.313351]  stmmac_xdp_xmit_xdpf+0x354/0x400
[  216.317701]  __stmmac_xdp_run_prog+0x164/0x368
[  216.322139]  stmmac_napi_poll_rxtx+0xba8/0xf00
[  216.326576]  __napi_poll+0x40/0x218
[  216.408054] Kernel panic - not syncing: Oops: Fatal exception in interrupt

For XDP_TX action, the xdp_buff is converted to xdp_frame by xdp_convert_buff_to_frame(). The memory type of the resulting xdp_frame depends on the memory type of the xdp_buff. For page pool based xdp_buff it produces xdp_frame with memory type MEM_TYPE_PAGE_POOL. For zero copy XSK pool based xdp_buff it produces xdp_frame with memory type MEM_TYPE_PAGE_ORDER0. However, stmmac_xdp_xmit_back() does not check the memory type and always uses the page pool type, this leads to invalid mappings and causes the crash. Therefore, check the xdp_buff memory type in stmmac_xdp_xmit_back() to fix this issue.

---
 .../net/ethernet/stmicro/stmmac/stmmac_main.c   | 17 +++++++++++++++--
 1 file changed, 15 insertions(+), 2 deletions(-)

index ce35a6f12679..112287a6e9ab 100644
@@ -87,6 +87,7 @@ MODULE_PARM_DESC(phyaddr, "Physical device address");
 #define STMMAC_XDP_CONSUMED	BIT(0)
 #define STMMAC_XDP_TX		BIT(1)
 #define STMMAC_XDP_REDIRECT	BIT(2)
+#define STMMAC_XSK_CONSUMED	BIT(3)
 
 static int flow_ctrl = FLOW_AUTO;
 module_param(flow_ctrl, int, 0644);

@@ -4998,6 +4999,7 @@ static int stmmac_xdp_get_tx_queue(struct stmmac_priv *priv,
 static int stmmac_xdp_xmit_back(struct stmmac_priv *priv,
 				struct xdp_buff *xdp)
 {
+	bool zc = !!(xdp->rxq->mem.type == MEM_TYPE_XSK_BUFF_POOL);
 	struct xdp_frame *xdpf = xdp_convert_buff_to_frame(xdp);
 	int cpu = smp_processor_id();
 	struct netdev_queue *nq;

@@ -5014,9 +5016,18 @@ static int stmmac_xdp_xmit_back(struct stmmac_priv *priv,
 	/* Avoids TX time-out as we are sharing with slow path */
 	txq_trans_cond_update(nq);
 
-	res = stmmac_xdp_xmit_xdpf(priv, queue, xdpf, false);
-	if (res == STMMAC_XDP_TX)
+	/* For zero copy XDP_TX action, dma_map is true */
+	res = stmmac_xdp_xmit_xdpf(priv, queue, xdpf, zc);
+	if (res == STMMAC_XDP_TX) {
 		stmmac_flush_tx_descriptors(priv, queue);
+	} else if (res == STMMAC_XDP_CONSUMED && zc) {
+		/* xdp has been freed by xdp_convert_buff_to_frame(),
+		 * no need to call xsk_buff_free() again, so return
+		 * STMMAC_XSK_CONSUMED.
+		 */
+		res = STMMAC_XSK_CONSUMED;
+		xdp_return_frame(xdpf);
+	}
 
 	__netif_tx_unlock(nq);
 

@@ -5366,6 +5377,8 @@ static int stmmac_rx_zc(struct stmmac_priv *priv, int limit, u32 queue)
 			break;
 		case STMMAC_XDP_CONSUMED:
 			xsk_buff_free(buf->xdp);
+			fallthrough;
+		case STMMAC_XSK_CONSUMED:
 			rx_dropped++;
 			break;
 		case STMMAC_XDP_TX:

2.51.0

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

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

From: Will Rosenberg <[email protected]>

[ Upstream commit 58fc7342b529803d3c221101102fe913df7adb83 ]

There exists a kernel oops caused by a BUG_ON(nhead < 0) at net/core/skbuff.c:2232 in pskb_expand_head().
This bug is triggered as part of the calipso_skbuff_setattr() routine when skb_cow() is passed headroom > INT_MAX (i.e. (int)(skb_headroom(skb) + len_delta) < 0).

The root cause of the bug is due to an implicit integer cast in __skb_cow(). The check (headroom > skb_headroom(skb)) is meant to ensure that delta = headroom - skb_headroom(skb) is never negative, otherwise we will trigger a BUG_ON in pskb_expand_head(). However, if headroom > INT_MAX and delta <= -NET_SKB_PAD, the check passes, delta becomes negative, and pskb_expand_head() is passed a negative value for nhead.

Fix the trigger condition in calipso_skbuff_setattr(). Avoid passing "negative" headroom sizes to skb_cow() within calipso_skbuff_setattr() by only using skb_cow() to grow headroom.

PoC:

	Using `netlabelctl` tool:

        netlabelctl map del default
        netlabelctl calipso add pass doi:7
        netlabelctl map add default address:0::1/128 protocol:calipso,7

        Then run the following PoC:

        int fd = socket(AF_INET6, SOCK_DGRAM, IPPROTO_UDP);

        // setup msghdr
        int cmsg_size = 2;
        int cmsg_len = 0x60;
        struct msghdr msg;
        struct sockaddr_in6 dest_addr;
        struct cmsghdr * cmsg = (struct cmsghdr *) calloc(1,
                        sizeof(struct cmsghdr) + cmsg_len);
        msg.msg_name = &dest_addr;
        msg.msg_namelen = sizeof(dest_addr);
        msg.msg_iov = NULL;
        msg.msg_iovlen = 0;
        msg.msg_control = cmsg;
        msg.msg_controllen = cmsg_len;
        msg.msg_flags = 0;

        // setup sockaddr
        dest_addr.sin6_family = AF_INET6;
        dest_addr.sin6_port = htons(31337);
        dest_addr.sin6_flowinfo = htonl(31337);
        dest_addr.sin6_addr = in6addr_loopback;
        dest_addr.sin6_scope_id = 31337;

        // setup cmsghdr
        cmsg->cmsg_len = cmsg_len;
        cmsg->cmsg_level = IPPROTO_IPV6;
        cmsg->cmsg_type = IPV6_HOPOPTS;
        char * hop_hdr = (char *)cmsg + sizeof(struct cmsghdr);
        hop_hdr[1] = 0x9; //set hop size - (0x9 + 1) * 8 = 80

        sendmsg(fd, &msg, 0);

---
 net/ipv6/calipso.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

index a247bb93908b..f5cc02ea3092 100644
@@ -1342,7 +1342,8 @@ static int calipso_skbuff_setattr(struct sk_buff *skb,
 	/* At this point new_end aligns to 4n, so (new_end & 4) pads to 8n */
 	pad = ((new_end & 4) + (end & 7)) & 7;
 	len_delta = new_end - (int)end + pad;
-	ret_val = skb_cow(skb, skb_headroom(skb) + len_delta);
+	ret_val = skb_cow(skb,
+			  skb_headroom(skb) + (len_delta > 0 ? len_delta : 0));
 	if (ret_val < 0)
 		return ret_val;
 

2.51.0

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

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

From: Ido Schimmel <[email protected]>

[ Upstream commit ac782f4e3bfcde145b8a7f8af31d9422d94d172a ]

When a nexthop object is deleted, it is marked as dead and then fib_table_flush() is called to flush all the routes that are using the dead nexthop.

The current logic in fib_table_flush() is to only flush error routes (e.g., blackhole) when it is called as part of network namespace dismantle (i.e., with flush_all=true). Therefore, error routes are not flushed when their nexthop object is deleted:

 # ip link add name dummy1 up type dummy
 # ip nexthop add id 1 dev dummy1
 # ip route add 198.51.100.1/32 nhid 1
 # ip route add blackhole 198.51.100.2/32 nhid 1
 # ip nexthop del id 1
 # ip route show
 blackhole 198.51.100.2 nhid 1 dev dummy1

As such, they keep holding a reference on the nexthop object which in turn holds a reference on the nexthop device, resulting in a reference count leak:

 # ip link del dev dummy1
 [   70.516258] unregister_netdevice: waiting for dummy1 to become free. Usage count = 2

Fix by flushing error routes when their nexthop is marked as dead.

IPv6 does not suffer from this problem.

---
 net/ipv4/fib_trie.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

index cc86031d2050..658f26d9a9ec 100644
@@ -2059,10 +2059,11 @@ int fib_table_flush(struct net *net, struct fib_table *tb, bool flush_all)
 				continue;
 			}
 
-			/* Do not flush error routes if network namespace is
-			 * not being dismantled
+			/* When not flushing the entire table, skip error
+			 * routes that are not marked for deletion.
 			 */
-			if (!flush_all && fib_props[fa->fa_type].error) {
+			if (!flush_all && fib_props[fa->fa_type].error &&
+			    !(fi->fib_flags & RTNH_F_DEAD)) {
 				slen = fa->fa_slen;
 				continue;
 			}

2.51.0

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

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

From: Pwnverse <[email protected]>

[ Upstream commit 6595beb40fb0ec47223d3f6058ee40354694c8e4 ]

rose_kill_by_device() collects sockets into a local array[] and then iterates over them to disconnect sockets bound to a device being brought down.

The loop mistakenly indexes array[cnt] instead of array[i]. For cnt < ARRAY_SIZE(array), this reads an uninitialized entry; for cnt == ARRAY_SIZE(array), it is an out-of-bounds read. Either case can lead to an invalid socket pointer dereference and also leaks references taken via sock_hold().

Fix the index to use i.

---
 net/rose/af_rose.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index b8078b42f5de..1676c9f4ab84 100644
@@ -205,7 +205,7 @@ static void rose_kill_by_device(struct net_device *dev)
 	spin_unlock_bh(&rose_list_lock);
 
 	for (i = 0; i < cnt; i++) {
-		sk = array[cnt];
+		sk = array[i];
 		rose = rose_sk(sk);
 		lock_sock(sk);
 		spin_lock_bh(&rose_list_lock);

2.51.0

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

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

From: Jiayuan Chen <[email protected]>

[ Upstream commit 1adaea51c61b52e24e7ab38f7d3eba023b2d050d ]

On PREEMPT_RT kernels, after rt6_get_pcpu_route() returns NULL, the current task can be preempted. Another task running on the same CPU may then execute rt6_make_pcpu_route() and successfully install a pcpu_rt entry. When the first task resumes execution, its cmpxchg() in rt6_make_pcpu_route() will fail because rt6i_pcpu is no longer NULL, triggering the BUG_ON(prev). It's easy to reproduce it by adding mdelay() after rt6_get_pcpu_route().

Using preempt_disable/enable is not appropriate here because ip6_rt_pcpu_alloc() may sleep.

Fix this by handling the cmpxchg() failure gracefully on PREEMPT_RT:
free our allocation and return the existing pcpu_rt installed by another task. The BUG_ON is replaced by WARN_ON_ONCE for non-PREEMPT_RT kernels where such races should not occur.

---
 net/ipv6/route.c | 13 ++++++++++++-
 1 file changed, 12 insertions(+), 1 deletion(-)

index 22866444efc0..276fa74af206 100644
@@ -1470,7 +1470,18 @@ static struct rt6_info *rt6_make_pcpu_route(struct net *net,
 
 	p = this_cpu_ptr(res->nh->rt6i_pcpu);
 	prev = cmpxchg(p, NULL, pcpu_rt);
-	BUG_ON(prev);
+	if (unlikely(prev)) {
+		/*
+		 * Another task on this CPU already installed a pcpu_rt.
+		 * This can happen on PREEMPT_RT where preemption is possible.
+		 * Free our allocation and return the existing one.
+		 */
+		WARN_ON_ONCE(!IS_ENABLED(CONFIG_PREEMPT_RT));
+
+		dst_dev_put(&pcpu_rt->dst);
+		dst_release(&pcpu_rt->dst);
+		return prev;
+	}
 
 	if (res->f6i->fib6_destroying) {
 		struct fib6_info *from;

2.51.0

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

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

From: Michal Schmidt <[email protected]>

[ Upstream commit 6f05611728e9d0ab024832a4f1abb74a5f5d0bb0 ]

irdma_net_event() should not dereference anything from "neigh" (alias "ptr") until it has checked that the event is NETEVENT_NEIGH_UPDATE. Other events come with different structures pointed to by "ptr" and they may be smaller than struct neighbour.

Move the read of neigh->dev under the NETEVENT_NEIGH_UPDATE case.

The bug is mostly harmless, but it triggers KASAN on debug kernels:

 BUG: KASAN: stack-out-of-bounds in irdma_net_event+0x32e/0x3b0 [irdma]
 Read of size 8 at addr ffffc900075e07f0 by task kworker/27:2/542554

 CPU: 27 PID: 542554 Comm: kworker/27:2 Kdump: loaded Not tainted 5.14.0-630.el9.x86_64+debug #1
 Hardware name: [...]
 Workqueue: events rt6_probe_deferred
 Call Trace:
  <IRQ>
  dump_stack_lvl+0x60/0xb0
  print_address_description.constprop.0+0x2c/0x3f0
  print_report+0xb4/0x270
  kasan_report+0x92/0xc0
  irdma_net_event+0x32e/0x3b0 [irdma]
  notifier_call_chain+0x9e/0x180
  atomic_notifier_call_chain+0x5c/0x110
  rt6_do_redirect+0xb91/0x1080
  tcp_v6_err+0xe9b/0x13e0
  icmpv6_notify+0x2b2/0x630
  ndisc_redirect_rcv+0x328/0x530
  icmpv6_rcv+0xc16/0x1360
  ip6_protocol_deliver_rcu+0xb84/0x12e0
  ip6_input_finish+0x117/0x240
  ip6_input+0xc4/0x370
  ipv6_rcv+0x420/0x7d0
  __netif_receive_skb_one_core+0x118/0x1b0
  process_backlog+0xd1/0x5d0
  __napi_poll.constprop.0+0xa3/0x440
  net_rx_action+0x78a/0xba0
  handle_softirqs+0x2d4/0x9c0
  do_softirq+0xad/0xe0
  </IRQ>

---
 drivers/infiniband/hw/irdma/utils.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

index 0422787592d8..87a6d58663de 100644
@@ -251,7 +251,7 @@ int irdma_net_event(struct notifier_block *notifier, unsigned long event,
 		    void *ptr)
 {
 	struct neighbour *neigh = ptr;
-	struct net_device *real_dev, *netdev = (struct net_device *)neigh->dev;
+	struct net_device *real_dev, *netdev;
 	struct irdma_device *iwdev;
 	struct ib_device *ibdev;
 	__be32 *p;

@@ -260,6 +260,7 @@ int irdma_net_event(struct notifier_block *notifier, unsigned long event,
 
 	switch (event) {
 	case NETEVENT_NEIGH_UPDATE:
+		netdev = neigh->dev;
 		real_dev = rdma_vlan_dev_real_dev(netdev);
 		if (!real_dev)
 			real_dev = netdev;

2.51.0

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

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

From: Michael Margolin <[email protected]>

[ Upstream commit 85463eb6a46caf2f1e0e1a6d0731f2f3bab17780 ]

The page size used for device might in some cases be smaller than PAGE_SIZE what results in a negative shift when calculating the number of host pages in PAGE_SIZE for a debug log. Remove the debug line together with the calculation.

---
 drivers/infiniband/hw/efa/efa_verbs.c | 4 ----
 1 file changed, 4 deletions(-)

index cc13415ff7e7..46eddef7a1cc 100644
@@ -1241,13 +1241,9 @@ static int umem_to_page_list(struct efa_dev *dev,
 			     u32 hp_cnt,
 			     u8 hp_shift)
 {
-	u32 pages_in_hp = BIT(hp_shift - PAGE_SHIFT);
 	struct ib_block_iter biter;
 	unsigned int hp_idx = 0;
 
-	ibdev_dbg(&dev->ibdev, "hp_cnt[%u], pages_in_hp[%u]\n",
-		  hp_cnt, pages_in_hp);
-
 	rdma_umem_for_each_dma_block(umem, &biter, BIT(hp_shift))
 		page_list[hp_idx++] = rdma_block_iter_dma_address(&biter);
 

2.51.0

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

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

From: Jang Ingyu <[email protected]>

[ Upstream commit 8aaa848eaddd9ef8680fc6aafbd3a0646da5df40 ]

Fix missing comparison operator for RDMA_NETWORK_ROCE_V1 in the conditional statement. The constant was used directly instead of being compared with net_type, causing the condition to always evaluate to true.

---
 drivers/infiniband/core/verbs.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index dc40001072a5..8dd96dc98fd3 100644
@@ -735,7 +735,7 @@ int ib_get_gids_from_rdma_hdr(const union rdma_network_hdr *hdr,
 				       (struct in6_addr *)dgid);
 		return 0;
 	} else if (net_type == RDMA_NETWORK_IPV6 ||
-		   net_type == RDMA_NETWORK_IB || RDMA_NETWORK_ROCE_V1) {
+		   net_type == RDMA_NETWORK_IB || net_type == RDMA_NETWORK_ROCE_V1) {
 		*dgid = hdr->ibgrh.dgid;
 		*sgid = hdr->ibgrh.sgid;
 		return 0;

2.51.0

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

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

From: Alok Tiwari <[email protected]>

[ Upstream commit 145a417a39d7efbc881f52e829817376972b278c ]

RCFW_COMM_CONS_PCI_BAR_REGION is defined as BAR 2, so checking !creq_db->reg.bar_id is incorrect and always false.

pci_resource_start() returns the BAR base address, and a value of 0 indicates that the BAR is unassigned. Update the condition to test bar_base == 0 instead.

This ensures the driver detects and logs an error for an unassigned RCFW communication BAR.

---
 drivers/infiniband/hw/bnxt_re/qplib_rcfw.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index 7a099580ca8b..38ded4687122 100644
@@ -1117,7 +1117,7 @@ static int bnxt_qplib_map_creq_db(struct bnxt_qplib_rcfw *rcfw, u32 reg_offt)
 	creq_db->dbinfo.flags = 0;
 	creq_db->reg.bar_id = RCFW_COMM_CONS_PCI_BAR_REGION;
 	creq_db->reg.bar_base = pci_resource_start(pdev, creq_db->reg.bar_id);
-	if (!creq_db->reg.bar_id)
+	if (!creq_db->reg.bar_base)
 		dev_err(&pdev->dev,
 			"QPLIB: CREQ BAR region %d resc start is 0!",
 			creq_db->reg.bar_id);

2.51.0

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

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

From: Tetsuo Handa <[email protected]>

[ Upstream commit fa3c411d21ebc26ffd175c7256c37cefa35020aa ]

Since nldev_deldev() (introduced by commit 060c642b2ab8 ("RDMA/nldev: Add support to add/delete a sub IB device through netlink") grabs a reference using ib_device_get_by_index() before calling ib_del_sub_device_and_put(), we need to drop that reference before returning -EOPNOTSUPP error.

---
 drivers/infiniband/core/device.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

index df2aa15a5bc9..bbc131737378 100644
@@ -2823,8 +2823,10 @@ int ib_del_sub_device_and_put(struct ib_device *sub)
 {
 	struct ib_device *parent = sub->parent;
 
-	if (!parent)
+	if (!parent) {
+		ib_device_put(sub);
 		return -EOPNOTSUPP;
+	}
 
 	mutex_lock(&parent->subdev_lock);
 	list_del(&sub->subdev_list);

2.51.0

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

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

From: Alok Tiwari <[email protected]>

[ Upstream commit f01765a2361323e78e3d91b1cb1d5527a83c5cf7 ]

The bnxt_re SEND path checks wr->send_flags to enable features such as IP checksum offload. However, send_flags is a bitmask and may contain multiple flags (e.g. IB_SEND_SIGNALED | IB_SEND_IP_CSUM), while the existing code uses a switch() statement that only matches when send_flags is exactly IB_SEND_IP_CSUM.

As a result, checksum offload is not enabled when additional SEND flags are present.

Replace the switch() with a bitmask test:

    if (wr->send_flags & IB_SEND_IP_CSUM)

This ensures IP checksum offload is enabled correctly when multiple SEND flags are used.

---
 drivers/infiniband/hw/bnxt_re/ib_verbs.c | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)

index c2abf2bb8026..c1587845f280 100644
@@ -2823,14 +2823,9 @@ int bnxt_re_post_send(struct ib_qp *ib_qp, const struct ib_send_wr *wr,
 				wqe.rawqp1.lflags |=
 					SQ_SEND_RAWETH_QP1_LFLAGS_ROCE_CRC;
 			}
-			switch (wr->send_flags) {
-			case IB_SEND_IP_CSUM:
+			if (wr->send_flags & IB_SEND_IP_CSUM)
 				wqe.rawqp1.lflags |=
 					SQ_SEND_RAWETH_QP1_LFLAGS_IP_CHKSUM;
-				break;
-			default:
-				break;
-			}
 			fallthrough;
 		case IB_WR_SEND_WITH_INV:
 			rc = bnxt_re_build_send_wqe(qp, wr, &wqe);

2.51.0

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

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

From: Kalesh AP <[email protected]>

[ Upstream commit 3d70e0fb0f289b0c778041c5bb04d099e1aa7c1c ]

In bnxt_qplib_alloc_init_hwq(), while allocating memory for PDE table driver incorrectly is using the "pg_size" value passed to the function. Fixed to use the right value 4K. Also, fixed the allocation size for PBL table.

---
 drivers/infiniband/hw/bnxt_re/qplib_res.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

index b785d9e7774c..f1a4bce6ce64 100644
@@ -243,7 +243,7 @@ int bnxt_qplib_alloc_init_hwq(struct bnxt_qplib_hwq *hwq,
 			if (npbl % BIT(MAX_PDL_LVL_SHIFT))
 				npde++;
 			/* Alloc PDE pages */
-			sginfo.pgsize = npde * pg_size;
+			sginfo.pgsize = npde * ROCE_PG_SIZE_4K;
 			sginfo.npages = 1;
 			rc = __alloc_pbl(res, &hwq->pbl[PBL_LVL_0], &sginfo);
 			if (rc)

@@ -251,7 +251,7 @@ int bnxt_qplib_alloc_init_hwq(struct bnxt_qplib_hwq *hwq,
 
 			/* Alloc PBL pages */
 			sginfo.npages = npbl;
-			sginfo.pgsize = PAGE_SIZE;
+			sginfo.pgsize = ROCE_PG_SIZE_4K;
 			rc = __alloc_pbl(res, &hwq->pbl[PBL_LVL_1], &sginfo);
 			if (rc)
 				goto fail;

2.51.0

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

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

From: Li Nan <[email protected]>

[ Upstream commit 00f6c1b4d15d35fadb7f34768a1831c81aaa8936 ]

The following warn is reported:

drivers/md/md.c:3912 analyze_sbs()
warn: iterator 'i' not incremented

---
 drivers/md/md.c | 5 +----
 1 file changed, 1 insertion(+), 4 deletions(-)

index 5c39246c467e..26056d53f40c 100644
@@ -3729,7 +3729,6 @@ static struct md_rdev *md_import_device(dev_t newdev, int super_format, int supe
 
 static int analyze_sbs(struct mddev *mddev)
 {
-	int i;
 	struct md_rdev *rdev, *freshest, *tmp;
 
 	freshest = NULL;

@@ -3756,11 +3755,9 @@ static int analyze_sbs(struct mddev *mddev)
 	super_types[mddev->major_version].
 		validate_super(mddev, NULL/*freshest*/, freshest);
 
-	i = 0;
 	rdev_for_each_safe(rdev, tmp, mddev) {
 		if (mddev->max_disks &&
-		    (rdev->desc_nr >= mddev->max_disks ||
-		     i > mddev->max_disks)) {
+		    rdev->desc_nr >= mddev->max_disks) {
 			pr_warn("md: %s: %pg: only %d devices permitted\n",
 				mdname(mddev), rdev->bdev,
 				mddev->max_disks);

2.51.0

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

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

From: Tuo Li <[email protected]>

[ Upstream commit 7ad6ef91d8745d04aff9cce7bdbc6320d8e05fe9 ]

The variable mddev->private is first assigned to conf and then checked:

  conf = mddev->private;
  if (!conf) ...

If conf is NULL, then mddev->private is also NULL. In this case, null-pointer dereferences can occur when calling raid5_quiesce():

  raid5_quiesce(mddev, true);
  raid5_quiesce(mddev, false);

since mddev->private is assigned to conf again in raid5_quiesce(), and conf is dereferenced in several places, for example:

  conf->quiesce = 0;
  wake_up(&conf->wait_for_quiescent);

To fix this issue, the function should unlock mddev and return before invoking raid5_quiesce() when conf is NULL, following the existing pattern in raid5_change_consistency_policy().

---
 drivers/md/raid5.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

index 8e5ccca3b68b..7262b77a8e02 100644
@@ -7181,12 +7181,14 @@ raid5_store_group_thread_cnt(struct mddev *mddev, const char *page, size_t len)
 	err = mddev_suspend_and_lock(mddev);
 	if (err)
 		return err;
+	conf = mddev->private;
+	if (!conf) {
+		mddev_unlock_and_resume(mddev);
+		return -ENODEV;
+	}
 	raid5_quiesce(mddev, true);
 
-	conf = mddev->private;
-	if (!conf)
-		err = -ENODEV;
-	else if (new != conf->worker_cnt_per_group) {
+	if (new != conf->worker_cnt_per_group) {
 		old_groups = conf->worker_groups;
 		if (old_groups)
 			flush_workqueue(raid5_wq);

2.51.0

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

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

From: Zilin Guan <[email protected]>

[ Upstream commit 0c56693b06a68476ba113db6347e7897475f9e4c ]

In get_file_all_info(), if vfs_getattr() fails, the function returns immediately without freeing the allocated filename, leading to a memory leak.

Fix this by freeing the filename before returning in this error case.

---
 fs/smb/server/smb2pdu.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

index a1579f76e063..e2cde9723001 100644
@@ -4926,8 +4926,10 @@ static int get_file_all_info(struct ksmbd_work *work,
 
 	ret = vfs_getattr(&fp->filp->f_path, &stat, STATX_BASIC_STATS,
 			  AT_STATX_SYNC_AS_STAT);
-	if (ret)
+	if (ret) {
+		kfree(filename);
 		return ret;
+	}
 
 	ksmbd_debug(SMB, "filename = %s\n", filename);
 	delete_pending = ksmbd_inode_pending_delete(fp);

2.51.0

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

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

From: Honggang LI <[email protected]>

[ Upstream commit 43bd09d5b750f700499ae8ec45fd41a4c48673e6 ]

If device max_mr_size bits in the range [mr_page_shift+31:mr_page_shift] are zero, the min3 function will set clt_path::max_pages_per_mr to zero.

alloc_path_reqs will pass zero, which is invalid, as the third parameter to ib_alloc_mr.

---
 drivers/infiniband/ulp/rtrs/rtrs-clt.c | 1 +
 1 file changed, 1 insertion(+)

index 71387811b281..2b397a544cb9 100644
@@ -1464,6 +1464,7 @@ static void query_fast_reg_mode(struct rtrs_clt_path *clt_path)
 	mr_page_shift      = max(12, ffs(ib_dev->attrs.page_size_cap) - 1);
 	max_pages_per_mr   = ib_dev->attrs.max_mr_size;
 	do_div(max_pages_per_mr, (1ull << mr_page_shift));
+	max_pages_per_mr = min_not_zero((u32)max_pages_per_mr, U32_MAX);
 	clt_path->max_pages_per_mr =
 		min3(clt_path->max_pages_per_mr, (u32)max_pages_per_mr,
 		     ib_dev->attrs.max_fast_reg_page_list_len);

2.51.0

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

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

From: Thomas Fourier <[email protected]>

[ Upstream commit fcd431a9627f272b4c0bec445eba365fe2232a94 ]

The dma_alloc_coherent() allocates a dma-mapped buffer, pbl->pg_arr[i]. The dma_free_coherent() should pass the same buffer to dma_free_coherent() and not page-aligned.

---
 drivers/infiniband/hw/bnxt_re/qplib_res.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

index f1a4bce6ce64..dfb72a5adc91 100644
@@ -70,9 +70,7 @@ static void __free_pbl(struct bnxt_qplib_res *res, struct bnxt_qplib_pbl *pbl,
 		for (i = 0; i < pbl->pg_count; i++) {
 			if (pbl->pg_arr[i])
 				dma_free_coherent(&pdev->dev, pbl->pg_size,
-						  (void *)((unsigned long)
-						   pbl->pg_arr[i] &
-						  PAGE_MASK),
+						  pbl->pg_arr[i],
 						  pbl->pg_map_arr[i]);
 			else
 				dev_warn(&pdev->dev,

2.51.0

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

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

From: Cong Zhang <[email protected]>

[ Upstream commit 10845a105bbcb030647a729f1716c2309da71d33 ]

If an hctx has no software ctx mapped, blk_mq_map_swqueue() never allocates tags and leaves hctx->tags NULL. The CPU hotplug offline notifier can still run for that hctx, return early since hctx cannot hold any requests.

---
 block/blk-mq.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index db72779760d5..1891863dcba1 100644
@@ -3658,7 +3658,7 @@ static int blk_mq_hctx_notify_offline(unsigned int cpu, struct hlist_node *node)
 			struct blk_mq_hw_ctx, cpuhp_online);
 	int ret = 0;
 
-	if (blk_mq_hctx_has_online_cpu(hctx, cpu))
+	if (!hctx->nr_ctx || blk_mq_hctx_has_online_cpu(hctx, cpu))
 		return 0;
 
 	/*

2.51.0

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

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

From: Yipeng Zou <[email protected]>

[ Upstream commit b889b4fb4cbea3ca7eb9814075d6a51936394bd9 ]

The func_traceonoff_triggers.tc sometimes goes to fail on my board, Kunpeng-920.

[root@localhost]# ./ftracetest ./test.d/ftrace/func_traceonoff_triggers.tc -l fail.log
=== Ftrace unit tests ===
[1] ftrace - test for function traceon/off triggers     [FAIL]
[2] (instance)  ftrace - test for function traceon/off triggers [UNSUPPORTED]

I look up the log, and it shows that the md5sum is different between csum1 and csum2.

++ cnt=611
++ sleep .1
+++ cnt_trace
+++ grep -v '^#' trace
+++ wc -l
++ cnt2=611
++ '[' 611 -ne 611 ']'
+++ cat tracing_on
++ on=0
++ '[' 0 '!=' 0 ']'
+++ md5sum trace
++ csum1='76896aa74362fff66a6a5f3cf8a8a500 trace'
++ sleep .1
+++ md5sum trace
++ csum2='ee8625a21c058818fc26e45c1ed3f6de trace'
++ '[' '76896aa74362fff66a6a5f3cf8a8a500 trace' '!=' 'ee8625a21c058818fc26e45c1ed3f6de trace' ']' ++ fail 'Tracing file is still changing'
++ echo Tracing file is still changing
Tracing file is still changing
++ exit_fail
++ exit 1

So I directly dump the trace file before md5sum, the diff shows that:

[root@localhost]# diff trace_1.log trace_2.log -y --suppress-common-lines
dockerd-12285   [036] d.... 18385.510290: sched_stat | <...>-12285   [036] d.... 18385.510290: sched_stat
dockerd-12285   [036] d.... 18385.510291: sched_swit | <...>-12285   [036] d.... 18385.510291: sched_swit
<...>-740       [044] d.... 18385.602859: sched_stat | kworker/44:1-740 [044] d.... 18385.602859: sched_stat
<...>-740       [044] d.... 18385.602860: sched_swit | kworker/44:1-740 [044] d.... 18385.602860: sched_swit

And we can see that <...> filed be filled with names.

We can strip off the names there to fix that.

After strip off the names:

kworker/u257:0-12 [019] d..2.  2528.758910: sched_stat | -12 [019] d..2.  2528.758910: sched_stat_runtime: comm=k
kworker/u257:0-12 [019] d..2.  2528.758912: sched_swit | -12 [019] d..2.  2528.758912: sched_switch: prev_comm=kw
<idle>-0          [000] d.s5.  2528.762318: sched_waki | -0  [000] d.s5.  2528.762318: sched_waking: comm=sshd pi
<idle>-0          [037] dNh2.  2528.762326: sched_wake | -0  [037] dNh2.  2528.762326: sched_wakeup: comm=sshd pi
<idle>-0          [037] d..2.  2528.762334: sched_swit | -0  [037] d..2.  2528.762334: sched_switch: prev_comm=sw

---
 .../ftrace/test.d/ftrace/func_traceonoff_triggers.tc         | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

index aee22289536b..1b57771dbfdf 100644
@@ -90,9 +90,10 @@ if [ $on != "0" ]; then
     fail "Tracing is not off"
 fi
 
-csum1=`md5sum trace`
+# Cannot rely on names being around as they are only cached, strip them
+csum1=`cat trace | sed -e 's/^ *[^ ]*\(-[0-9][0-9]*\)/\1/' | md5sum`
 sleep $SLEEP_TIME
-csum2=`md5sum trace`
+csum2=`cat trace | sed -e 's/^ *[^ ]*\(-[0-9][0-9]*\)/\1/' | md5sum`
 
 if [ "$csum1" != "$csum2" ]; then
     fail "Tracing file is still changing"

2.51.0

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

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

From: Matthew Wilcox (Oracle) <[email protected]>

commit 68f6bd128e75a032432eda9d16676ed2969a1096 upstream.

When reading a compressed file, we may read several pages in addition to the one requested. The current code will overwrite pages in the page cache with the data from disc which can definitely result in changes that have been made being lost.

For example if we have four consecutie pages ABCD in the file compressed into a single extent, on first access, we'll bring in ABCD. Then we write to page B. Memory pressure results in the eviction of ACD. When we attempt to write to page C, we will overwrite the data in page B with the data currently on disk.

I haven't investigated the decompression code to check whether it's OK to overwrite a clean page or whether it might be possible to see corrupt data. Out of an abundance of caution, decline to overwrite uptodate pages, not just dirty pages.

---
 fs/ntfs3/frecord.c |   35 +++++++++++++++++++++++++++++------
 1 file changed, 29 insertions(+), 6 deletions(-)

@@ -2077,6 +2077,29 @@ out:
 	return err;
 }
 
+static struct page *ntfs_lock_new_page(struct address_space *mapping,
+		pgoff_t index, gfp_t gfp)
+{
+	struct folio *folio = __filemap_get_folio(mapping, index,
+			FGP_LOCK | FGP_ACCESSED | FGP_CREAT, gfp);
+	struct page *page;
+
+	if (IS_ERR(folio))
+		return ERR_CAST(folio);
+
+	if (!folio_test_uptodate(folio))
+		return folio_file_page(folio, index);
+
+	/* Use a temporary page to avoid data corruption */
+	folio_unlock(folio);
+	folio_put(folio);
+	page = alloc_page(gfp);
+	if (!page)
+		return ERR_PTR(-ENOMEM);
+	__SetPageLocked(page);
+	return page;
+}
+
 /*
  * ni_readpage_cmpr
  *

@@ -2131,9 +2154,9 @@ int ni_readpage_cmpr(struct ntfs_inode *
 		if (i == idx)
 			continue;
 
-		pg = find_or_create_page(mapping, index, gfp_mask);
-		if (!pg) {
-			err = -ENOMEM;
+		pg = ntfs_lock_new_page(mapping, index, gfp_mask);
+		if (IS_ERR(pg)) {
+			err = PTR_ERR(pg);
 			goto out1;
 		}
 		pages[i] = pg;

@@ -2232,13 +2255,13 @@ int ni_decompress_file(struct ntfs_inode
 		for (i = 0; i < pages_per_frame; i++, index++) {
 			struct page *pg;
 
-			pg = find_or_create_page(mapping, index, gfp_mask);
-			if (!pg) {
+			pg = ntfs_lock_new_page(mapping, index, gfp_mask);
+			if (IS_ERR(pg)) {
 				while (i--) {
 					unlock_page(pages[i]);
 					put_page(pages[i]);
 				}
-				err = -ENOMEM;
+				err = PTR_ERR(pg);
 				goto out;
 			}
 			pages[i] = pg;

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

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

From: Johan Hovold <[email protected]>

commit 86dc090f737953f16f8dc60c546ae7854690d4f6 upstream.

The soundwire regmap that may be allocated during probe is not freed on late probe failures.

Add the missing error handling.

---
 sound/soc/codecs/wcd939x-sdw.c |    8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

@@ -1480,12 +1480,18 @@ static int wcd9390_probe(struct sdw_slav
 
 	ret = component_add(dev, &wcd939x_sdw_component_ops);
 	if (ret)
-		return ret;
+		goto err_free_regmap;
 
 	/* Set suspended until aggregate device is bind */
 	pm_runtime_set_suspended(dev);
 
 	return 0;
+
+err_free_regmap:
+	if (wcd->regmap)
+		regmap_exit(wcd->regmap);
+
+	return ret;
 }
 
 static int wcd9390_remove(struct sdw_slave *pdev)

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

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

From: Johan Hovold <[email protected]>

commit e26ff429eaf10c4ef1bc3dabd9bf27eb54b7e1f4 upstream.

Make sure to drop the reference taken when looking up the sync provider device and its driver data during DAI probe on probe failures and on unbind.

Note that holding a reference to a device does not prevent its driver data from going away so there is no point in keeping the reference.

---
 sound/soc/stm/stm32_sai.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -127,6 +127,7 @@ static int stm32_sai_set_sync(struct stm
 	}
 
 	sai_provider = platform_get_drvdata(pdev);
+	put_device(&pdev->dev);
 	if (!sai_provider) {
 		dev_err(&sai_client->pdev->dev,
 			"SAI sync provider data not found\n");

@@ -143,7 +144,6 @@ static int stm32_sai_set_sync(struct stm
 	ret = stm32_sai_sync_conf_provider(sai_provider, synco);
 
 error:
-	put_device(&pdev->dev);
 	of_node_put(np_provider);
 	return ret;
 }

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

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

From: Johan Hovold <[email protected]>

commit 312ec2f0d9d1a5656f76d770bbf1d967e9289aa7 upstream.

Make sure to unprepare the parent clock also on probe failures (e.g. probe deferral).

---
 sound/soc/stm/stm32_sai_sub.c |   28 +++++++++++++++++++++-------
 1 file changed, 21 insertions(+), 7 deletions(-)

@@ -1501,14 +1501,21 @@ static int stm32_sai_sub_parse_of(struct
 	if (of_property_present(np, "#clock-cells")) {
 		ret = stm32_sai_add_mclk_provider(sai);
 		if (ret < 0)
-			return ret;
+			goto err_unprepare_pclk;
 	} else {
 		sai->sai_mclk = devm_clk_get_optional(&pdev->dev, "MCLK");
-		if (IS_ERR(sai->sai_mclk))
-			return PTR_ERR(sai->sai_mclk);
+		if (IS_ERR(sai->sai_mclk)) {
+			ret = PTR_ERR(sai->sai_mclk);
+			goto err_unprepare_pclk;
+		}
 	}
 
 	return 0;
+
+err_unprepare_pclk:
+	clk_unprepare(sai->pdata->pclk);
+
+	return ret;
 }
 
 static int stm32_sai_sub_probe(struct platform_device *pdev)

@@ -1548,26 +1555,33 @@ static int stm32_sai_sub_probe(struct pl
 			       IRQF_SHARED, dev_name(&pdev->dev), sai);
 	if (ret) {
 		dev_err(&pdev->dev, "IRQ request returned %d\n", ret);
-		return ret;
+		goto err_unprepare_pclk;
 	}
 
 	if (STM_SAI_PROTOCOL_IS_SPDIF(sai))
 		conf = &stm32_sai_pcm_config_spdif;
 
 	ret = snd_dmaengine_pcm_register(&pdev->dev, conf, 0);
-	if (ret)
-		return dev_err_probe(&pdev->dev, ret, "Could not register pcm dma\n");
+	if (ret) {
+		ret = dev_err_probe(&pdev->dev, ret, "Could not register pcm dma\n");
+		goto err_unprepare_pclk;
+	}
 
 	ret = snd_soc_register_component(&pdev->dev, &stm32_component,
 					 &sai->cpu_dai_drv, 1);
 	if (ret) {
 		snd_dmaengine_pcm_unregister(&pdev->dev);
-		return ret;
+		goto err_unprepare_pclk;
 	}
 
 	pm_runtime_enable(&pdev->dev);
 
 	return 0;
+
+err_unprepare_pclk:
+	clk_unprepare(sai->pdata->pclk);
+
+	return ret;
 }
 
 static void stm32_sai_sub_remove(struct platform_device *pdev)

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

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

From: Johan Hovold <[email protected]>

commit 23261f0de09427367e99f39f588e31e2856a690e upstream.

The reference taken to the sync provider OF node when probing the platform device is currently only dropped if the set_sync() callback fails during DAI probe.

Make sure to drop the reference on platform probe failures (e.g. probe deferral) and on driver unbind.

This also avoids a potential use-after-free in case the DAI is ever reprobed without first rebinding the platform driver.

---
 sound/soc/stm/stm32_sai.c     |   12 +++---------
 sound/soc/stm/stm32_sai_sub.c |   23 ++++++++++++++++-------
 2 files changed, 19 insertions(+), 16 deletions(-)

@@ -122,7 +122,6 @@ static int stm32_sai_set_sync(struct stm
 	if (!pdev) {
 		dev_err(&sai_client->pdev->dev,
 			"Device not found for node %pOFn\n", np_provider);
-		of_node_put(np_provider);
 		return -ENODEV;
 	}
 

@@ -131,21 +130,16 @@ static int stm32_sai_set_sync(struct stm
 	if (!sai_provider) {
 		dev_err(&sai_client->pdev->dev,
 			"SAI sync provider data not found\n");
-		ret = -EINVAL;
-		goto error;
+		return -EINVAL;
 	}
 
 	/* Configure sync client */
 	ret = stm32_sai_sync_conf_client(sai_client, synci);
 	if (ret < 0)
-		goto error;
+		return ret;
 
 	/* Configure sync provider */
-	ret = stm32_sai_sync_conf_provider(sai_provider, synco);
-
-error:
-	of_node_put(np_provider);
-	return ret;
+	return stm32_sai_sync_conf_provider(sai_provider, synco);
 }
 
 static int stm32_sai_probe(struct platform_device *pdev)

@@ -1453,7 +1453,8 @@ static int stm32_sai_sub_parse_of(struct
 				dev_err(&pdev->dev,
 					"External synchro not supported\n");
 				of_node_put(args.np);
-				return -EINVAL;
+				ret = -EINVAL;
+				goto err_put_sync_provider;
 			}
 			sai->sync = SAI_SYNC_EXTERNAL;
 

@@ -1462,7 +1463,8 @@ static int stm32_sai_sub_parse_of(struct
 			    (sai->synci > (SAI_GCR_SYNCIN_MAX + 1))) {
 				dev_err(&pdev->dev, "Wrong SAI index\n");
 				of_node_put(args.np);
-				return -EINVAL;
+				ret = -EINVAL;
+				goto err_put_sync_provider;
 			}
 
 			if (of_property_match_string(args.np, "compatible",

@@ -1476,7 +1478,8 @@ static int stm32_sai_sub_parse_of(struct
 			if (!sai->synco) {
 				dev_err(&pdev->dev, "Unknown SAI sub-block\n");
 				of_node_put(args.np);
-				return -EINVAL;
+				ret = -EINVAL;
+				goto err_put_sync_provider;
 			}
 		}
 

@@ -1486,13 +1489,15 @@ static int stm32_sai_sub_parse_of(struct
 
 	of_node_put(args.np);
 	sai->sai_ck = devm_clk_get(&pdev->dev, "sai_ck");
-	if (IS_ERR(sai->sai_ck))
-		return dev_err_probe(&pdev->dev, PTR_ERR(sai->sai_ck),
-				     "Missing kernel clock sai_ck\n");
+	if (IS_ERR(sai->sai_ck)) {
+		ret = dev_err_probe(&pdev->dev, PTR_ERR(sai->sai_ck),
+				    "Missing kernel clock sai_ck\n");
+		goto err_put_sync_provider;
+	}
 
 	ret = clk_prepare(sai->pdata->pclk);
 	if (ret < 0)
-		return ret;
+		goto err_put_sync_provider;
 
 	if (STM_SAI_IS_F4(sai->pdata))
 		return 0;

@@ -1514,6 +1519,8 @@ static int stm32_sai_sub_parse_of(struct
 
 err_unprepare_pclk:
 	clk_unprepare(sai->pdata->pclk);
+err_put_sync_provider:
+	of_node_put(sai->np_sync_provider);
 
 	return ret;
 }

@@ -1580,6 +1587,7 @@ static int stm32_sai_sub_probe(struct pl
 
 err_unprepare_pclk:
 	clk_unprepare(sai->pdata->pclk);
+	of_node_put(sai->np_sync_provider);
 
 	return ret;
 }

@@ -1592,6 +1600,7 @@ static void stm32_sai_sub_remove(struct
 	snd_dmaengine_pcm_unregister(&pdev->dev);
 	snd_soc_unregister_component(&pdev->dev);
 	pm_runtime_disable(&pdev->dev);
+	of_node_put(sai->np_sync_provider);
 }
 
 #ifdef CONFIG_PM_SLEEP

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

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

From: Srinivas Kandagatla <[email protected]>

commit 7c63b5a8ed972a2c8c03d984f6a43349007cea93 upstream.

SM6115 does have soundwire controller in tx. For some reason we ended up with this incorrect patch.

Fix this by adding the flag to reflect this in SoC data.

---
 sound/soc/codecs/lpass-tx-macro.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

@@ -2474,7 +2474,8 @@ static const struct tx_macro_data lpass_
 };
 
 static const struct tx_macro_data lpass_ver_10_sm6115 = {
-	.flags			= LPASS_MACRO_FLAG_HAS_NPL_CLOCK,
+	.flags			= LPASS_MACRO_FLAG_HAS_NPL_CLOCK |
+				  LPASS_MACRO_FLAG_RESET_SWR,
 	.ver			= LPASS_VER_10_0_0,
 	.extra_widgets		= tx_macro_dapm_widgets_v9_2,
 	.extra_widgets_num	= ARRAY_SIZE(tx_macro_dapm_widgets_v9_2),

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

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

From: Srinivas Kandagatla <[email protected]>

commit 950a4e5788fc7dc6e8e93614a7d4d0449c39fb8d upstream.

Driver does not expect the appl_ptr to move backward and requires explict sync. Make sure that the userspace does not do appl_ptr rewinds by specifying the correct flags in pcm_info.

Without this patch, the result could be a forever loop as current logic assumes that appl_ptr can only move forward.

---
 sound/soc/qcom/qdsp6/q6apm-dai.c |    2 ++
 1 file changed, 2 insertions(+)

@@ -85,6 +85,7 @@ static const struct snd_pcm_hardware q6a
 	.info =                 (SNDRV_PCM_INFO_MMAP | SNDRV_PCM_INFO_BLOCK_TRANSFER |
 				 SNDRV_PCM_INFO_MMAP_VALID | SNDRV_PCM_INFO_INTERLEAVED |
 				 SNDRV_PCM_INFO_PAUSE | SNDRV_PCM_INFO_RESUME |
+				 SNDRV_PCM_INFO_NO_REWINDS | SNDRV_PCM_INFO_SYNC_APPLPTR |
 				 SNDRV_PCM_INFO_BATCH),
 	.formats =              (SNDRV_PCM_FMTBIT_S16_LE | SNDRV_PCM_FMTBIT_S24_LE),
 	.rates =                SNDRV_PCM_RATE_8000_48000,

@@ -104,6 +105,7 @@ static const struct snd_pcm_hardware q6a
 	.info =                 (SNDRV_PCM_INFO_MMAP | SNDRV_PCM_INFO_BLOCK_TRANSFER |
 				 SNDRV_PCM_INFO_MMAP_VALID | SNDRV_PCM_INFO_INTERLEAVED |
 				 SNDRV_PCM_INFO_PAUSE | SNDRV_PCM_INFO_RESUME |
+				 SNDRV_PCM_INFO_NO_REWINDS | SNDRV_PCM_INFO_SYNC_APPLPTR |
 				 SNDRV_PCM_INFO_BATCH),
 	.formats =              (SNDRV_PCM_FMTBIT_S16_LE | SNDRV_PCM_FMTBIT_S24_LE),
 	.rates =                SNDRV_PCM_RATE_8000_192000,

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

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

From: Srinivas Kandagatla <[email protected]>

commit bfbb12dfa144d45575bcfe139a71360b3ce80237 upstream.

Do not stop a q6asm stream if its not started, this can result in unnecessary dsp command which will timeout anyway something like below:

q6asm-dai ab00000.remoteproc:glink-edge:apr:service@7:dais: CMD 10bcd timeout

Fix this by correctly checking the state.

---
 sound/soc/qcom/qdsp6/q6asm-dai.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

@@ -232,13 +232,14 @@ static int q6asm_dai_prepare(struct snd_
 	prtd->pcm_count = snd_pcm_lib_period_bytes(substream);
 	prtd->pcm_irq_pos = 0;
 	/* rate and channels are sent to audio driver */
-	if (prtd->state) {
+	if (prtd->state == Q6ASM_STREAM_RUNNING) {
 		/* clear the previous setup if any  */
 		q6asm_cmd(prtd->audio_client, prtd->stream_id, CMD_CLOSE);
 		q6asm_unmap_memory_regions(substream->stream,
 					   prtd->audio_client);
 		q6routing_stream_close(soc_prtd->dai_link->id,
 					 substream->stream);
+		prtd->state = Q6ASM_STREAM_STOPPED;
 	}
 
 	ret = q6asm_map_memory_regions(substream->stream, prtd->audio_client,

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

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

From: Srinivas Kandagatla <[email protected]>

commit 74cc4f3ea4e99262ba0d619c6a4ee33e2cd47f65 upstream.

A matching Common object post processing instance is normally resused across multiple streams. However currently we close this on DSP even though there is a refcount on this copp object, this can result in below error.

q6routing ab00000.remoteproc:glink-edge:apr:service@8:routing: Found Matching Copp 0x0 qcom-q6adm aprsvc:service:4:8: cmd = 0x10325 return error = 0x2 q6routing ab00000.remoteproc:glink-edge:apr:service@8:routing: DSP returned error[2] q6routing ab00000.remoteproc:glink-edge:apr:service@8:routing: Found Matching Copp 0x0 qcom-q6adm aprsvc:service:4:8: cmd = 0x10325 return error = 0x2 q6routing ab00000.remoteproc:glink-edge:apr:service@8:routing: DSP returned error[2] qcom-q6adm aprsvc:service:4:8: cmd = 0x10327 return error = 0x2 qcom-q6adm aprsvc:service:4:8: DSP returned error[2]
qcom-q6adm aprsvc:service:4:8: Failed to close copp -22
qcom-q6adm aprsvc:service:4:8: cmd = 0x10327 return error = 0x2 qcom-q6adm aprsvc:service:4:8: DSP returned error[2]
qcom-q6adm aprsvc:service:4:8: Failed to close copp -22

Fix this by addressing moving the adm_close to copp_kref destructor callback.

---
 sound/soc/qcom/qdsp6/q6adm.c |  146 ++++++++++++++++++++-----------------------
 1 file changed, 71 insertions(+), 75 deletions(-)

@@ -109,11 +109,75 @@ static struct q6copp *q6adm_find_copp(st
 
 }
 
+static int q6adm_apr_send_copp_pkt(struct q6adm *adm, struct q6copp *copp,
+				   struct apr_pkt *pkt, uint32_t rsp_opcode)
+{
+	struct device *dev = adm->dev;
+	uint32_t opcode = pkt->hdr.opcode;
+	int ret;
+
+	mutex_lock(&adm->lock);
+	copp->result.opcode = 0;
+	copp->result.status = 0;
+	ret = apr_send_pkt(adm->apr, pkt);
+	if (ret < 0) {
+		dev_err(dev, "Failed to send APR packet\n");
+		ret = -EINVAL;
+		goto err;
+	}
+
+	/* Wait for the callback with copp id */
+	if (rsp_opcode)
+		ret = wait_event_timeout(copp->wait,
+					 (copp->result.opcode == opcode) ||
+					 (copp->result.opcode == rsp_opcode),
+					 msecs_to_jiffies(TIMEOUT_MS));
+	else
+		ret = wait_event_timeout(copp->wait,
+					 (copp->result.opcode == opcode),
+					 msecs_to_jiffies(TIMEOUT_MS));
+
+	if (!ret) {
+		dev_err(dev, "ADM copp cmd timedout\n");
+		ret = -ETIMEDOUT;
+	} else if (copp->result.status > 0) {
+		dev_err(dev, "DSP returned error[%d]\n",
+			copp->result.status);
+		ret = -EINVAL;
+	}
+
+err:
+	mutex_unlock(&adm->lock);
+	return ret;
+}
+
+static int q6adm_device_close(struct q6adm *adm, struct q6copp *copp,
+			      int port_id, int copp_idx)
+{
+	struct apr_pkt close;
+
+	close.hdr.hdr_field = APR_HDR_FIELD(APR_MSG_TYPE_SEQ_CMD,
+					APR_HDR_LEN(APR_HDR_SIZE),
+					APR_PKT_VER);
+	close.hdr.pkt_size = sizeof(close);
+	close.hdr.src_port = port_id;
+	close.hdr.dest_port = copp->id;
+	close.hdr.token = port_id << 16 | copp_idx;
+	close.hdr.opcode = ADM_CMD_DEVICE_CLOSE_V5;
+
+	return q6adm_apr_send_copp_pkt(adm, copp, &close, 0);
+}
+
 static void q6adm_free_copp(struct kref *ref)
 {
 	struct q6copp *c = container_of(ref, struct q6copp, refcount);
 	struct q6adm *adm = c->adm;
 	unsigned long flags;
+	int ret;
+
+	ret = q6adm_device_close(adm, c, c->afe_port, c->copp_idx);
+	if (ret < 0)
+		dev_err(adm->dev, "Failed to close copp %d\n", ret);
 
 	spin_lock_irqsave(&adm->copps_list_lock, flags);
 	clear_bit(c->copp_idx, &adm->copp_bitmap[c->afe_port]);