Skip to contentSkip to conversation

6.16.1-rc1 review

PATCH In review Greg Kroah-Hartman 24 participants 676 messages Aug 12, 2025 – Aug 17, 2025 Page 3 of 5

Subsystems (7)
lore
Share & tools
lore
author maintainer
#259 lore raw

PATCH 249/627 wifi: brcmfmac: cyw: Fix __counted_by to be LE variant

In reply to Greg Kroah-Hartman

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

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

From: Kees Cook <[email protected]>

[ Upstream commit 204bb852863bf14f343a0801b15bc2173bc318f9 ]

In brcmf_cyw_mgmt_tx() the "len" counter of the struct brcmf_mf_params_le::data flexible array is stored as little-endian via cpu_to_le16() so the __counted_by_le() variant must be used:

	struct brcmf_mf_params_le *mf_params;
	...
	mf_params_len = offsetof(struct brcmf_mf_params_le, data) +
			(len - DOT11_MGMT_HDR_LEN);
	mf_params = kzalloc(mf_params_len, GFP_KERNEL);
	...
        mf_params->len = cpu_to_le16(len - DOT11_MGMT_HDR_LEN);
Fixes 66f909308a7c ("wifi: brcmfmac: cyw: support external SAE authentication in station mode")
---
 .../net/wireless/broadcom/brcm80211/brcmfmac/cyw/fwil_types.h   | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/wireless/broadcom/brcm80211/brcmfmac/cyw/fwil_types.h +1 -1
index 08c69142495a..669564382e32 100644
@@ -80,7 +80,7 @@ struct brcmf_mf_params_le {
 	u8 da[ETH_ALEN];
 	u8 bssid[ETH_ALEN];
 	__le32 packet_id;
-	u8 data[] __counted_by(len);
+	u8 data[] __counted_by_le(len);
 };
 
 #endif /* CYW_FWIL_TYPES_H_ */
Signature · 1 line
2.39.5
author maintainer
#260 lore raw

PATCH 250/627 wifi: mac80211: fix WARN_ON for monitor mode on some devices

In reply to Greg Kroah-Hartman

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

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

From: Johannes Berg <[email protected]>

[ Upstream commit c57e5b9819dfd16d709bcd6cb633301ed0829a66 ]

On devices without WANT_MONITOR_VIF (and probably without channel context support) we get a WARN_ON for changing the per-link setting of a monitor interface.

Since we already skip AP_VLAN interfaces and MONITOR with WANT_MONITOR_VIF and/or NO_VIRTUAL_MONITOR should update the settings, catch this in the link change code instead of the warning.

Fixes c4382d5ca1af ("wifi: mac80211: update the right link for tx power")
---
 net/mac80211/main.c | 13 ++++++++++++-
 1 file changed, 12 insertions(+), 1 deletion(-)
net/mac80211/main.c +12 -1
index 6b6de43d9420..1bad353d8a77 100644
@@ -407,9 +407,20 @@ void ieee80211_link_info_change_notify(struct ieee80211_sub_if_data *sdata,
 
 	WARN_ON_ONCE(changed & BSS_CHANGED_VIF_CFG_FLAGS);
 
-	if (!changed || sdata->vif.type == NL80211_IFTYPE_AP_VLAN)
+	if (!changed)
 		return;
 
+	switch (sdata->vif.type) {
+	case NL80211_IFTYPE_AP_VLAN:
+		return;
+	case NL80211_IFTYPE_MONITOR:
+		if (!ieee80211_hw_check(&local->hw, WANT_MONITOR_VIF))
+			return;
+		break;
+	default:
+		break;
+	}
+
 	if (!check_sdata_in_driver(sdata))
 		return;
 
Signature · 1 line
2.39.5
author maintainer
#261 lore raw

PATCH 251/627 arm64/gcs: task_gcs_el0_enable() should use passed task

In reply to Greg Kroah-Hartman

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

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

From: Jeremy Linton <[email protected]>

[ Upstream commit cbbcfb94c55c02a8c4ce52b5da0770b5591a314c ]

Mark Rutland noticed that the task parameter is ignored and 'current' is being used instead. Since this is usually what its passed, it hasn't yet been causing problems but likely will as the code gets more testing.

But, once this is fixed, it creates a new bug in copy_thread_gcs() since the gcs_el_mode isn't yet set for the task before its being checked. Move gcs_alloc_thread_stack() after the new task's gcs_el0_mode initialization to avoid this.

Fixes fc84bc5378a8 ("arm64/gcs: Context switch GCS state for EL0")
---
 arch/arm64/include/asm/gcs.h | 2 +-
 arch/arm64/kernel/process.c  | 6 +++---
 2 files changed, 4 insertions(+), 4 deletions(-)
arch/arm64/include/asm/gcs.h +1 -1
index f50660603ecf..5bc432234d3a 100644
@@ -58,7 +58,7 @@ static inline u64 gcsss2(void)
 
 static inline bool task_gcs_el0_enabled(struct task_struct *task)
 {
-	return current->thread.gcs_el0_mode & PR_SHADOW_STACK_ENABLE;
+	return task->thread.gcs_el0_mode & PR_SHADOW_STACK_ENABLE;
 }
 
 void gcs_set_el0_mode(struct task_struct *task);
arch/arm64/kernel/process.c +3 -3
index 08b7042a2e2d..3e1baff5e88d 100644
@@ -307,13 +307,13 @@ static int copy_thread_gcs(struct task_struct *p,
 	p->thread.gcs_base = 0;
 	p->thread.gcs_size = 0;
 
+	p->thread.gcs_el0_mode = current->thread.gcs_el0_mode;
+	p->thread.gcs_el0_locked = current->thread.gcs_el0_locked;
+
 	gcs = gcs_alloc_thread_stack(p, args);
 	if (IS_ERR_VALUE(gcs))
 		return PTR_ERR((void *)gcs);
 
-	p->thread.gcs_el0_mode = current->thread.gcs_el0_mode;
-	p->thread.gcs_el0_locked = current->thread.gcs_el0_locked;
-
 	return 0;
 }
 
Signature · 1 line
2.39.5
author maintainer
#262 lore raw

PATCH 252/627 wifi: iwlwifi: mld: decode EOF bit for AMPDUs

In reply to Greg Kroah-Hartman

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

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

From: Benjamin Berg <[email protected]>

[ Upstream commit bc404dfddbf6817cae9b170c34556dc72ea975e5 ]

Only the EOF bit handling for single frames was ported to the MLD driver. The code to handle AMPDUs correctly was forgotten. Add it back so that the bit is reported in the radiotap headers again.

---
 drivers/net/wireless/intel/iwlwifi/mld/rx.c | 9 +++++++++
 1 file changed, 9 insertions(+)
drivers/net/wireless/intel/iwlwifi/mld/rx.c +9 -0
index ce0093d5c638..185c1a0cb47f 100644
@@ -1039,6 +1039,15 @@ static void iwl_mld_rx_eht(struct iwl_mld *mld, struct sk_buff *skb,
 			rx_status->flag |= RX_FLAG_AMPDU_EOF_BIT;
 	}
 
+	/* update aggregation data for monitor sake on default queue */
+	if (!queue && (phy_info & IWL_RX_MPDU_PHY_TSF_OVERLOAD) &&
+	    (phy_info & IWL_RX_MPDU_PHY_AMPDU) && phy_data->first_subframe) {
+		rx_status->flag |= RX_FLAG_AMPDU_EOF_BIT_KNOWN;
+		if (phy_data->data0 &
+		    cpu_to_le32(IWL_RX_PHY_DATA0_EHT_DELIM_EOF))
+			rx_status->flag |= RX_FLAG_AMPDU_EOF_BIT;
+	}
+
 	if (phy_info & IWL_RX_MPDU_PHY_TSF_OVERLOAD)
 		iwl_mld_decode_eht_phy_data(mld, phy_data, rx_status, eht, usig);
 
Signature · 1 line
2.39.5
author maintainer
#263 lore raw

PATCH 253/627 iommu/vt-d: Fix UAF on sva unbind with pending IOPFs

In reply to Greg Kroah-Hartman

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

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

From: Lu Baolu <[email protected]>

[ Upstream commit f0b9d31c6edd50a6207489cd1bd4ddac814b9cd2 ]

Commit 17fce9d2336d ("iommu/vt-d: Put iopf enablement in domain attach path") disables IOPF on device by removing the device from its IOMMU's IOPF queue when the last IOPF-capable domain is detached from the device. Unfortunately, it did this in a wrong place where there are still pending IOPFs. As a result, a use-after-free error is potentially triggered and eventually a kernel panic with a kernel trace similar to the following:

refcount_t: underflow; use-after-free.

 WARNING: CPU: 3 PID: 313 at lib/refcount.c:28 refcount_warn_saturate+0xd8/0xe0
 Workqueue: iopf_queue/dmar0-iopfq iommu_sva_handle_iopf
 Call Trace:
   <TASK>
   iopf_free_group+0xe/0x20
   process_one_work+0x197/0x3d0
   worker_thread+0x23a/0x350
   ? rescuer_thread+0x4a0/0x4a0
   kthread+0xf8/0x230
   ? finish_task_switch.isra.0+0x81/0x260
   ? kthreads_online_cpu+0x110/0x110
   ? kthreads_online_cpu+0x110/0x110
   ret_from_fork+0x13b/0x170
   ? kthreads_online_cpu+0x110/0x110
   ret_from_fork_asm+0x11/0x20
   </TASK>
  ---[ end trace 0000000000000000 ]---

The intel_pasid_tear_down_entry() function is responsible for blocking hardware from generating new page faults and flushing all in-flight ones. Therefore, moving iopf_for_domain_remove() after this function should resolve this.

---
 drivers/iommu/intel/iommu.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/iommu/intel/iommu.c +1 -1
index 72b477911fdb..c0be0b64e4c7 100644
@@ -3999,8 +3999,8 @@ static int blocking_domain_set_dev_pasid(struct iommu_domain *domain,
 {
 	struct device_domain_info *info = dev_iommu_priv_get(dev);
 
-	iopf_for_domain_remove(old, dev);
 	intel_pasid_tear_down_entry(info->iommu, dev, pasid, false);
+	iopf_for_domain_remove(old, dev);
 	domain_remove_dev_pasid(old, dev, pasid);
 
 	return 0;
Signature · 1 line
2.39.5
author maintainer
#264 lore raw

PATCH 254/627 wifi: brcmfmac: fix EXTSAE WPA3 connection failure due to AUTH TX failure

In reply to Greg Kroah-Hartman

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

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

From: Ting-Ying Li <[email protected]>

[ Upstream commit f2d7c3c380bf0c38c395d50de3a7c1a6275983cb ]

For WPA3-SAE Connection in EXTSAE mode, the userspace daemon is allowed to generate the SAE Auth frames. The driver uses the "mgmt_frame" FW IOVAR to transmit this MGMT frame.

Before sending the IOVAR, the Driver is incorrectly treating the channel number read from the FW as a frequency value and again attempts to convert this into a channel number using ieee80211_frequency_to_channel().

This added an invalid channel number as part of the IOVAR request to the FW And some FW which strictly expects a valid channel would return BAD_CHAN error, while failing to transmit the driver requested SAE Auth MGMT frame.

Fix this in the CYW vendor specific MGMT TX cfg80211 ops handler, by not treating the channel number read from the FW as frequency value and skip the attempt to convert it again into a channel number.

Also fix this in the generic MGMT TX cfg80211 ops handler.

Fixes c2ff8cad6423 ("brcm80211: make mgmt_tx in brcmfmac accept a NULL channel")
Fixes 66f909308a7c ("wifi: brcmfmac: cyw: support external SAE authentication in station mode")
---
 .../broadcom/brcm80211/brcmfmac/cfg80211.c    | 30 ++++++++++++-------
 .../broadcom/brcm80211/brcmfmac/cyw/core.c    | 26 +++++++++-------
 2 files changed, 35 insertions(+), 21 deletions(-)
drivers/net/wireless/broadcom/brcm80211/brcmfmac/cfg80211.c +19 -11
index 086b9157292a..70e8ddd3851f 100644
@@ -5527,8 +5527,7 @@ brcmf_cfg80211_mgmt_tx(struct wiphy *wiphy, struct wireless_dev *wdev,
 	struct brcmf_fil_action_frame_le *action_frame;
 	struct brcmf_fil_af_params_le *af_params;
 	bool ack;
-	s32 chan_nr;
-	u32 freq;
+	__le32 hw_ch;
 
 	brcmf_dbg(TRACE, "Enter\n");
 
@@ -5589,25 +5588,34 @@ brcmf_cfg80211_mgmt_tx(struct wiphy *wiphy, struct wireless_dev *wdev,
 		/* Add the channel. Use the one specified as parameter if any or
 		 * the current one (got from the firmware) otherwise
 		 */
-		if (chan)
-			freq = chan->center_freq;
-		else
-			brcmf_fil_cmd_int_get(vif->ifp, BRCMF_C_GET_CHANNEL,
-					      &freq);
-		chan_nr = ieee80211_frequency_to_channel(freq);
-		af_params->channel = cpu_to_le32(chan_nr);
+		if (chan) {
+			hw_ch = cpu_to_le32(chan->hw_value);
+		} else {
+			err = brcmf_fil_cmd_data_get(vif->ifp,
+						     BRCMF_C_GET_CHANNEL,
+						     &hw_ch, sizeof(hw_ch));
+			if (err) {
+				bphy_err(drvr,
+					 "unable to get current hw channel\n");
+				goto free;
+			}
+		}
+		af_params->channel = hw_ch;
+
 		af_params->dwell_time = cpu_to_le32(params->wait);
 		memcpy(action_frame->data, &buf[DOT11_MGMT_HDR_LEN],
 		       le16_to_cpu(action_frame->len));
 
-		brcmf_dbg(TRACE, "Action frame, cookie=%lld, len=%d, freq=%d\n",
-			  *cookie, le16_to_cpu(action_frame->len), freq);
+		brcmf_dbg(TRACE, "Action frame, cookie=%lld, len=%d, channel=%d\n",
+			  *cookie, le16_to_cpu(action_frame->len),
+			  le32_to_cpu(af_params->channel));
 
 		ack = brcmf_p2p_send_action_frame(cfg, cfg_to_ndev(cfg),
 						  af_params);
 
 		cfg80211_mgmt_tx_status(wdev, *cookie, buf, len, ack,
 					GFP_KERNEL);
+free:
 		kfree(af_params);
 	} else {
 		brcmf_dbg(TRACE, "Unhandled, fc=%04x!!\n", mgmt->frame_control);
drivers/net/wireless/broadcom/brcm80211/brcmfmac/cyw/core.c +16 -10
index c9537fb597ce..4f0ea4347840 100644
@@ -112,8 +112,7 @@ int brcmf_cyw_mgmt_tx(struct wiphy *wiphy, struct wireless_dev *wdev,
 	struct brcmf_cfg80211_vif *vif;
 	s32 err = 0;
 	bool ack = false;
-	s32 chan_nr;
-	u32 freq;
+	__le16 hw_ch;
 	struct brcmf_mf_params_le *mf_params;
 	u32 mf_params_len;
 	s32 ready;
@@ -143,13 +142,18 @@ int brcmf_cyw_mgmt_tx(struct wiphy *wiphy, struct wireless_dev *wdev,
 	mf_params->len = cpu_to_le16(len - DOT11_MGMT_HDR_LEN);
 	mf_params->frame_control = mgmt->frame_control;
 
-	if (chan)
-		freq = chan->center_freq;
-	else
-		brcmf_fil_cmd_int_get(vif->ifp, BRCMF_C_GET_CHANNEL,
-				      &freq);
-	chan_nr = ieee80211_frequency_to_channel(freq);
-	mf_params->channel = cpu_to_le16(chan_nr);
+	if (chan) {
+		hw_ch = cpu_to_le16(chan->hw_value);
+	} else {
+		err = brcmf_fil_cmd_data_get(vif->ifp, BRCMF_C_GET_CHANNEL,
+					     &hw_ch, sizeof(hw_ch));
+		if (err) {
+			bphy_err(drvr, "unable to get current hw channel\n");
+			goto free;
+		}
+	}
+	mf_params->channel = hw_ch;
+
 	memcpy(&mf_params->da[0], &mgmt->da[0], ETH_ALEN);
 	memcpy(&mf_params->bssid[0], &mgmt->bssid[0], ETH_ALEN);
 	mf_params->packet_id = cpu_to_le32(*cookie);
@@ -159,7 +163,8 @@ int brcmf_cyw_mgmt_tx(struct wiphy *wiphy, struct wireless_dev *wdev,
 	brcmf_dbg(TRACE, "Auth frame, cookie=%d, fc=%04x, len=%d, channel=%d\n",
 		  le32_to_cpu(mf_params->packet_id),
 		  le16_to_cpu(mf_params->frame_control),
-		  le16_to_cpu(mf_params->len), chan_nr);
+		  le16_to_cpu(mf_params->len),
+		  le16_to_cpu(mf_params->channel));
 
 	vif->mgmt_tx_id = le32_to_cpu(mf_params->packet_id);
 	set_bit(BRCMF_MGMT_TX_SEND_FRAME, &vif->mgmt_tx_status);
@@ -185,6 +190,7 @@ int brcmf_cyw_mgmt_tx(struct wiphy *wiphy, struct wireless_dev *wdev,
 tx_status:
 	cfg80211_mgmt_tx_status(wdev, *cookie, buf, len, ack,
 				GFP_KERNEL);
+free:
 	kfree(mf_params);
 	return err;
 }
Signature · 1 line
2.39.5
author maintainer
#265 lore raw

PATCH 255/627 Bluetooth: btusb: Fix potential NULL dereference on kmalloc failure

In reply to Greg Kroah-Hartman

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

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

From: Zhongqiu Han <[email protected]>

[ Upstream commit b505902c66a282dcb01bcdc015aa1fdfaaa075db ]

Avoid potential NULL pointer dereference by checking the return value of kmalloc and handling allocation failure properly.

Fixes 7d70989fcea7 ("Bluetooth: btusb: Add HCI Drv commands for configuring altsetting")
---
 drivers/bluetooth/btusb.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/bluetooth/btusb.c +2 -0
index f9eeec0aed57..db27d28e8a7e 100644
@@ -3802,6 +3802,8 @@ static int btusb_hci_drv_supported_altsettings(struct hci_dev *hdev, void *data,
 
 	/* There are at most 7 alt (0 - 6) */
 	rp = kmalloc(sizeof(*rp) + 7, GFP_KERNEL);
+	if (!rp)
+		return -ENOMEM;
 
 	rp->num = 0;
 	if (!drvdata->isoc)
Signature · 1 line
2.39.5
author maintainer
#266 lore raw

PATCH 256/627 Bluetooth: hci_sync: fix double free in hci_discovery_filter_clear()

In reply to Greg Kroah-Hartman

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

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

From: Arseniy Krasnov <[email protected]>

[ Upstream commit 2935e556850e9c94d7a00adf14d3cd7fe406ac03 ]

Function 'hci_discovery_filter_clear()' frees 'uuids' array and then sets it to NULL. There is a tiny chance of the following race:

'hci_cmd_sync_work()'

'update_passive_scan_sync()'

'hci_update_passive_scan_sync()'

     'hci_discovery_filter_clear()'
       kfree(uuids);

       <-------------------------preempted-------------------------------->
                                           'start_service_discovery()'

                                             'hci_discovery_filter_clear()'
                                               kfree(uuids); // DOUBLE FREE

       <-------------------------preempted-------------------------------->

      uuids = NULL;

To fix it let's add locking around 'kfree()' call and NULL pointer assignment. Otherwise the following backtrace fires:

[ ] ------------[ cut here ]------------
[ ] kernel BUG at mm/slub.c:547!
[ ] Internal error: Oops - BUG: 00000000f2000800 [#1] PREEMPT SMP [ ] CPU: 3 UID: 0 PID: 246 Comm: bluetoothd Tainted: G O 6.12.19-kernel #1 [ ] Tainted: [O]=OOT_MODULE
[ ] pstate: 60400005 (nZCv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--) [ ] pc : __slab_free+0xf8/0x348
[ ] lr : __slab_free+0x48/0x348
...
[ ] Call trace:

[ ]  __slab_free+0xf8/0x348
[ ]  kfree+0x164/0x27c
[ ]  start_service_discovery+0x1d0/0x2c0
[ ]  hci_sock_sendmsg+0x518/0x924
[ ]  __sock_sendmsg+0x54/0x60
[ ]  sock_write_iter+0x98/0xf8
[ ]  do_iter_readv_writev+0xe4/0x1c8
[ ]  vfs_writev+0x128/0x2b0
[ ]  do_writev+0xfc/0x118
[ ]  __arm64_sys_writev+0x20/0x2c
[ ]  invoke_syscall+0x68/0xf0
[ ]  el0_svc_common.constprop.0+0x40/0xe0
[ ]  do_el0_svc+0x1c/0x28
[ ]  el0_svc+0x30/0xd0
[ ]  el0t_64_sync_handler+0x100/0x12c
[ ]  el0t_64_sync+0x194/0x198

[ ] Code: 8b0002e6 eb17031f 54fffbe1 d503201f (d4210000) [ ] ---[ end trace 0000000000000000 ]---

Fixes ad383c2c65a5 ("Bluetooth: hci_sync: Enable advertising when LL privacy is enabled")
---
 include/net/bluetooth/hci_core.h | 6 ++++++
 1 file changed, 6 insertions(+)
include/net/bluetooth/hci_core.h +6 -0
index f79f59e67114..c371dadc6fa3 100644
@@ -29,6 +29,7 @@
 #include <linux/idr.h>
 #include <linux/leds.h>
 #include <linux/rculist.h>
+#include <linux/spinlock.h>
 #include <linux/srcu.h>
 
 #include <net/bluetooth/hci.h>
@@ -94,6 +95,7 @@ struct discovery_state {
 	u16			uuid_count;
 	u8			(*uuids)[16];
 	unsigned long		name_resolve_timeout;
+	spinlock_t		lock;
 };
 
 #define SUSPEND_NOTIFIER_TIMEOUT	msecs_to_jiffies(2000) /* 2 seconds */
@@ -889,6 +891,7 @@ static inline void iso_recv(struct hci_conn *hcon, struct sk_buff *skb,
 
 static inline void discovery_init(struct hci_dev *hdev)
 {
+	spin_lock_init(&hdev->discovery.lock);
 	hdev->discovery.state = DISCOVERY_STOPPED;
 	INIT_LIST_HEAD(&hdev->discovery.all);
 	INIT_LIST_HEAD(&hdev->discovery.unknown);
@@ -903,8 +906,11 @@ static inline void hci_discovery_filter_clear(struct hci_dev *hdev)
 	hdev->discovery.report_invalid_rssi = true;
 	hdev->discovery.rssi = HCI_RSSI_INVALID;
 	hdev->discovery.uuid_count = 0;
+
+	spin_lock(&hdev->discovery.lock);
 	kfree(hdev->discovery.uuids);
 	hdev->discovery.uuids = NULL;
+	spin_unlock(&hdev->discovery.lock);
 }
 
 bool hci_discovery_active(struct hci_dev *hdev);
Signature · 1 line
2.39.5
author maintainer
#267 lore raw

PATCH 257/627 Bluetooth: hci_devcd_dump: fix out-of-bounds via dev_coredumpv

In reply to Greg Kroah-Hartman

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

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

From: Ivan Pravdin <[email protected]>

[ Upstream commit 7af4d7b53502286c6cf946d397ab183e76d14820 ]

Currently both dev_coredumpv and skb_put_data in hci_devcd_dump use hdev->dump.head. However, dev_coredumpv can free the buffer. From dev_coredumpm_timeout documentation, which is used by dev_coredumpv:

Quoted 4 lines

Creates a new device coredump for the given device. If a previous one hasn't been read yet, the new coredump is discarded. The data lifetime is determined by the device coredump framework and when it is no longer needed the @free function will be called to free the data.

If the data has not been read by the userspace yet, dev_coredumpv will discard new buffer, freeing hdev->dump.head. This leads to vmalloc-out-of-bounds error when skb_put_data tries to access hdev->dump.head.

A crash report from syzbot illustrates this:

    ==================================================================
    BUG: KASAN: vmalloc-out-of-bounds in skb_put_data
    include/linux/skbuff.h:2752 [inline]
    BUG: KASAN: vmalloc-out-of-bounds in hci_devcd_dump+0x142/0x240
    net/bluetooth/coredump.c:258
    Read of size 140 at addr ffffc90004ed5000 by task kworker/u9:2/5844

    CPU: 1 UID: 0 PID: 5844 Comm: kworker/u9:2 Not tainted
    6.14.0-syzkaller-10892-g4e82c87058f4 #0 PREEMPT(full)
    Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS
    Google 02/12/2025
    Workqueue: hci0 hci_devcd_timeout
    Call Trace:
     <TASK>
     __dump_stack lib/dump_stack.c:94 [inline]
     dump_stack_lvl+0x116/0x1f0 lib/dump_stack.c:120
     print_address_description mm/kasan/report.c:408 [inline]
     print_report+0xc3/0x670 mm/kasan/report.c:521
     kasan_report+0xe0/0x110 mm/kasan/report.c:634
     check_region_inline mm/kasan/generic.c:183 [inline]
     kasan_check_range+0xef/0x1a0 mm/kasan/generic.c:189
     __asan_memcpy+0x23/0x60 mm/kasan/shadow.c:105
     skb_put_data include/linux/skbuff.h:2752 [inline]
     hci_devcd_dump+0x142/0x240 net/bluetooth/coredump.c:258
     hci_devcd_timeout+0xb5/0x2e0 net/bluetooth/coredump.c:413
     process_one_work+0x9cc/0x1b70 kernel/workqueue.c:3238
     process_scheduled_works kernel/workqueue.c:3319 [inline]
     worker_thread+0x6c8/0xf10 kernel/workqueue.c:3400
     kthread+0x3c2/0x780 kernel/kthread.c:464
     ret_from_fork+0x45/0x80 arch/x86/kernel/process.c:153
     ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:245
     </TASK>

    The buggy address ffffc90004ed5000 belongs to a vmalloc virtual mapping
    Memory state around the buggy address:
     ffffc90004ed4f00: f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8
     ffffc90004ed4f80: f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8
    >ffffc90004ed5000: f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8
                       ^
     ffffc90004ed5080: f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8
     ffffc90004ed5100: f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8 f8
    ==================================================================

To avoid this issue, reorder dev_coredumpv to be called after skb_put_data that does not free the data.

Reported-by [email protected]
Fixes b257e02ecc46 ("HCI: coredump: Log devcd dumps into the monitor")
---
 net/bluetooth/coredump.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
net/bluetooth/coredump.c +3 -3
index 819eacb38762..720cb79adf96 100644
@@ -249,15 +249,15 @@ static void hci_devcd_dump(struct hci_dev *hdev)
 
 	size = hdev->dump.tail - hdev->dump.head;
 
-	/* Emit a devcoredump with the available data */
-	dev_coredumpv(&hdev->dev, hdev->dump.head, size, GFP_KERNEL);
-
 	/* Send a copy to monitor as a diagnostic packet */
 	skb = bt_skb_alloc(size, GFP_ATOMIC);
 	if (skb) {
 		skb_put_data(skb, hdev->dump.head, size);
 		hci_recv_diag(hdev, skb);
 	}
+
+	/* Emit a devcoredump with the available data */
+	dev_coredumpv(&hdev->dev, hdev->dump.head, size, GFP_KERNEL);
 }
 
 static void hci_devcd_handle_pkt_complete(struct hci_dev *hdev,
Signature · 1 line
2.39.5
author maintainer
#268 lore raw

PATCH 258/627 Bluetooth: btintel: Define a macro for Intel Reset vendor command

In reply to Greg Kroah-Hartman

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

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

From: Kiran K <[email protected]>

[ Upstream commit 15843c7fdba65568704245fd3ea2aa3aa2d50825 ]

Use macro for Intel Reset command (0xfc01) instead of hard coded value.

Stable-dep-of: 69b3d3acf3db ("Bluetooth: btintel_pcie: Make driver wait for alive interrupt")

---
 drivers/bluetooth/btintel.c      |  4 ++--
 drivers/bluetooth/btintel.h      |  2 ++
 drivers/bluetooth/btintel_pcie.c | 12 ++++++------
 drivers/bluetooth/btusb.c        |  8 ++++----
 drivers/bluetooth/hci_intel.c    | 10 +++++-----
 5 files changed, 19 insertions(+), 17 deletions(-)
drivers/bluetooth/btintel.c +2 -2
index 06016ac3965c..6aceecf5a13d 100644
@@ -889,7 +889,7 @@ int btintel_send_intel_reset(struct hci_dev *hdev, u32 boot_param)
 
 	params.boot_param = cpu_to_le32(boot_param);
 
-	skb = __hci_cmd_sync(hdev, 0xfc01, sizeof(params), &params,
+	skb = __hci_cmd_sync(hdev, BTINTEL_HCI_OP_RESET, sizeof(params), &params,
 			     HCI_INIT_TIMEOUT);
 	if (IS_ERR(skb)) {
 		bt_dev_err(hdev, "Failed to send Intel Reset command");
@@ -1287,7 +1287,7 @@ static void btintel_reset_to_bootloader(struct hci_dev *hdev)
 	params.boot_option = 0x00;
 	params.boot_param = cpu_to_le32(0x00000000);
 
-	skb = __hci_cmd_sync(hdev, 0xfc01, sizeof(params),
+	skb = __hci_cmd_sync(hdev, BTINTEL_HCI_OP_RESET, sizeof(params),
 			     &params, HCI_INIT_TIMEOUT);
 	if (IS_ERR(skb)) {
 		bt_dev_err(hdev, "FW download error recovery failed (%ld)",
drivers/bluetooth/btintel.h +2 -0
index 1d12c4113c66..431998049e68 100644
@@ -52,6 +52,8 @@ struct intel_tlv {
 	u8 val[];
 } __packed;
 
+#define BTINTEL_HCI_OP_RESET	0xfc01
+
 #define BTINTEL_CNVI_BLAZARI		0x900
 #define BTINTEL_CNVI_BLAZARIW		0x901
 #define BTINTEL_CNVI_GAP		0x910
drivers/bluetooth/btintel_pcie.c +6 -6
index f4e3fb54fe76..1638be0921a3 100644
@@ -1955,12 +1955,12 @@ static int btintel_pcie_send_frame(struct hci_dev *hdev,
 			struct hci_command_hdr *cmd = (void *)skb->data;
 			__u16 opcode = le16_to_cpu(cmd->opcode);
 
-			/* When the 0xfc01 command is issued to boot into
-			 * the operational firmware, it will actually not
-			 * send a command complete event. To keep the flow
+			/* When the BTINTEL_HCI_OP_RESET command is issued to
+			 * boot into the operational firmware, it will actually
+			 * not send a command complete event. To keep the flow
 			 * control working inject that event here.
 			 */
-			if (opcode == 0xfc01)
+			if (opcode == BTINTEL_HCI_OP_RESET)
 				btintel_pcie_inject_cmd_complete(hdev, opcode);
 		}
 		/* Firmware raises alive interrupt on HCI_OP_RESET */
@@ -1995,10 +1995,10 @@ static int btintel_pcie_send_frame(struct hci_dev *hdev,
 	}
 
 	if (type == BTINTEL_PCIE_HCI_CMD_PKT &&
-	    (opcode == HCI_OP_RESET || opcode == 0xfc01)) {
+	    (opcode == HCI_OP_RESET || opcode == BTINTEL_HCI_OP_RESET)) {
 		old_ctxt = data->alive_intr_ctxt;
 		data->alive_intr_ctxt =
-			(opcode == 0xfc01 ? BTINTEL_PCIE_INTEL_HCI_RESET1 :
+			(opcode == BTINTEL_HCI_OP_RESET ? BTINTEL_PCIE_INTEL_HCI_RESET1 :
 				BTINTEL_PCIE_HCI_RESET);
 		bt_dev_dbg(data->hdev, "sent cmd: 0x%4.4x alive context changed: %s  ->  %s",
 			   opcode, btintel_pcie_alivectxt_state2str(old_ctxt),
drivers/bluetooth/btusb.c +4 -4
index db27d28e8a7e..66fd84fbbd22 100644
@@ -2594,12 +2594,12 @@ static int btusb_send_frame_intel(struct hci_dev *hdev, struct sk_buff *skb)
 			else
 				urb = alloc_ctrl_urb(hdev, skb);
 
-			/* When the 0xfc01 command is issued to boot into
-			 * the operational firmware, it will actually not
-			 * send a command complete event. To keep the flow
+			/* When the BTINTEL_HCI_OP_RESET command is issued to
+			 * boot into the operational firmware, it will actually
+			 * not send a command complete event. To keep the flow
 			 * control working inject that event here.
 			 */
-			if (opcode == 0xfc01)
+			if (opcode == BTINTEL_HCI_OP_RESET)
 				inject_cmd_complete(hdev, opcode);
 		} else {
 			urb = alloc_ctrl_urb(hdev, skb);
drivers/bluetooth/hci_intel.c +5 -5
index d22fbb7f9fc5..9b353c3d6442 100644
@@ -1029,12 +1029,12 @@ static struct sk_buff *intel_dequeue(struct hci_uart *hu)
 		struct hci_command_hdr *cmd = (void *)skb->data;
 		__u16 opcode = le16_to_cpu(cmd->opcode);
 
-		/* When the 0xfc01 command is issued to boot into
-		 * the operational firmware, it will actually not
-		 * send a command complete event. To keep the flow
-		 * control working inject that event here.
+		/* When the BTINTEL_HCI_OP_RESET command is issued to boot into
+		 * the operational firmware, it will actually not send a command
+		 * complete event. To keep the flow control working inject that
+		 * event here.
 		 */
-		if (opcode == 0xfc01)
+		if (opcode == BTINTEL_HCI_OP_RESET)
 			inject_cmd_complete(hu->hdev, opcode);
 	}
 
Signature · 1 line
2.39.5
author maintainer
#269 lore raw

PATCH 259/627 Bluetooth: btintel_pcie: Make driver wait for alive interrupt

In reply to Greg Kroah-Hartman

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

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

From: Kiran K <[email protected]>

[ Upstream commit 69b3d3acf3dba21e2abad863c80c7114eb110b3d ]

The firmware raises an alive interrupt upon receiving the HCI_RESET or BTINTEL_HCI_OP_RESET (Intel reset - 0xfc01) command. This change fixes the driver to properly wait for the alive interrupt to avoid driver sending commands to firmware before it is ready to process.

For details on the handshake between the driver and firmware, refer to commit 05c200c8f029 ("Bluetooth: btintel_pcie: Add handshake between driver and firmware").

As the driver needs to handle two interrupts for HCI_OP_RESET and BTINTEL_HCI_OP_RESET, the firmware ensures that the TX completion interrupt is always followed by the alive interrupt.

Fixes 05c200c8f029 ("Bluetooth: btintel_pcie: Add handshake between driver and firmware")
---
 drivers/bluetooth/btintel_pcie.c | 30 +++++++++++++++++-------------
 1 file changed, 17 insertions(+), 13 deletions(-)
drivers/bluetooth/btintel_pcie.c +17 -13
index 1638be0921a3..7f789937a764 100644
@@ -928,11 +928,13 @@ static void btintel_pcie_msix_gp0_handler(struct btintel_pcie_data *data)
 	case BTINTEL_PCIE_INTEL_HCI_RESET1:
 		if (btintel_pcie_in_op(data)) {
 			submit_rx = true;
+			signal_waitq = true;
 			break;
 		}
 
 		if (btintel_pcie_in_iml(data)) {
 			submit_rx = true;
+			signal_waitq = true;
 			data->alive_intr_ctxt = BTINTEL_PCIE_FW_DL;
 			break;
 		}
@@ -1963,8 +1965,11 @@ static int btintel_pcie_send_frame(struct hci_dev *hdev,
 			if (opcode == BTINTEL_HCI_OP_RESET)
 				btintel_pcie_inject_cmd_complete(hdev, opcode);
 		}
-		/* Firmware raises alive interrupt on HCI_OP_RESET */
-		if (opcode == HCI_OP_RESET)
+
+		/* Firmware raises alive interrupt on HCI_OP_RESET or
+		 * BTINTEL_HCI_OP_RESET
+		 */
+		if (opcode == HCI_OP_RESET || opcode == BTINTEL_HCI_OP_RESET)
 			data->gp0_received = false;
 
 		hdev->stat.cmd_tx++;
@@ -2003,17 +2008,16 @@ static int btintel_pcie_send_frame(struct hci_dev *hdev,
 		bt_dev_dbg(data->hdev, "sent cmd: 0x%4.4x alive context changed: %s  ->  %s",
 			   opcode, btintel_pcie_alivectxt_state2str(old_ctxt),
 			   btintel_pcie_alivectxt_state2str(data->alive_intr_ctxt));
-		if (opcode == HCI_OP_RESET) {
-			ret = wait_event_timeout(data->gp0_wait_q,
-						 data->gp0_received,
-						 msecs_to_jiffies(BTINTEL_DEFAULT_INTR_TIMEOUT_MS));
-			if (!ret) {
-				hdev->stat.err_tx++;
-				bt_dev_err(hdev, "No alive interrupt received for %s",
-					   btintel_pcie_alivectxt_state2str(data->alive_intr_ctxt));
-				ret = -ETIME;
-				goto exit_error;
-			}
+		ret = wait_event_timeout(data->gp0_wait_q,
+					 data->gp0_received,
+					 msecs_to_jiffies(BTINTEL_DEFAULT_INTR_TIMEOUT_MS));
+		if (!ret) {
+			hdev->stat.err_tx++;
+			bt_dev_err(hdev, "Timeout on alive interrupt (%u ms). Alive context: %s",
+				   BTINTEL_DEFAULT_INTR_TIMEOUT_MS,
+				   btintel_pcie_alivectxt_state2str(data->alive_intr_ctxt));
+			ret = -ETIME;
+			goto exit_error;
 		}
 	}
 	hdev->stat.byte_tx += skb->len;
Signature · 1 line
2.39.5
author maintainer
#270 lore raw

PATCH 260/627 Bluetooth: hci_event: Mask data status from LE ext adv reports

In reply to Greg Kroah-Hartman

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

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

From: Chris Down <[email protected]>

[ Upstream commit 0cadf8534f2a727bc3a01e8c583b085d25963ee0 ]

The Event_Type field in an LE Extended Advertising Report uses bits 5 and 6 for data status (e.g. truncation or fragmentation), not the PDU type itself.

The ext_evt_type_to_legacy() function fails to mask these status bits before evaluation. This causes valid advertisements with status bits set (e.g. a truncated non-connectable advertisement, which ends up showing as PDU type 0x40) to be misclassified as unknown and subsequently dropped. This is okay for most checks which use bitwise AND on the relevant event type bits, but it doesn't work for non-connectable types, which are checked with '== LE_EXT_ADV_NON_CONN_IND' (that is, zero).

In terms of behaviour, first the device sends a truncated report:

Quoted 1 line

HCI Event: LE Meta Event (0x3e) plen 26

      LE Extended Advertising Report (0x0d)
        Entry 0
          Event type: 0x0040
            Data status: Incomplete, data truncated, no more to come
          Address type: Random (0x01)
          Address: 1D:12:46:FA:F8:6E (Non-Resolvable)
          SID: 0x03
          RSSI: -98 dBm (0x9e)
          Data length: 0x00

Then, a few seconds later, it sends the subsequent complete report:

Quoted 1 line

HCI Event: LE Meta Event (0x3e) plen 122

      LE Extended Advertising Report (0x0d)
        Entry 0
          Event type: 0x0000
            Data status: Complete
          Address type: Random (0x01)
          Address: 1D:12:46:FA:F8:6E (Non-Resolvable)
          SID: 0x03
          RSSI: -97 dBm (0x9f)
          Data length: 0x60
          Service Data: Google (0xfef3)
            Data[92]: ...

These devices often send multiple truncated reports per second.

This patch introduces a PDU type mask to ensure only the relevant bits are evaluated, allowing for the correct translation of all valid extended advertising packets.

Fixes b2cc9761f144 ("Bluetooth: Handle extended ADV PDU types")
---
 include/net/bluetooth/hci.h | 1 +
 net/bluetooth/hci_event.c   | 8 ++++++--
 2 files changed, 7 insertions(+), 2 deletions(-)
include/net/bluetooth/hci.h +1 -0
index c79901f2dc2a..5796ca9fe5da 100644
@@ -2634,6 +2634,7 @@ struct hci_ev_le_conn_complete {
 #define LE_EXT_ADV_DIRECT_IND		0x0004
 #define LE_EXT_ADV_SCAN_RSP		0x0008
 #define LE_EXT_ADV_LEGACY_PDU		0x0010
+#define LE_EXT_ADV_DATA_STATUS_MASK	0x0060
 #define LE_EXT_ADV_EVT_TYPE_MASK	0x007f
 
 #define ADDR_LE_DEV_PUBLIC		0x00
net/bluetooth/hci_event.c +6 -2
index cf4b30ac9e0e..c1dd8d78701f 100644
@@ -6239,6 +6239,11 @@ static void hci_le_adv_report_evt(struct hci_dev *hdev, void *data,
 
 static u8 ext_evt_type_to_legacy(struct hci_dev *hdev, u16 evt_type)
 {
+	u16 pdu_type = evt_type & ~LE_EXT_ADV_DATA_STATUS_MASK;
+
+	if (!pdu_type)
+		return LE_ADV_NONCONN_IND;
+
 	if (evt_type & LE_EXT_ADV_LEGACY_PDU) {
 		switch (evt_type) {
 		case LE_LEGACY_ADV_IND:
@@ -6270,8 +6275,7 @@ static u8 ext_evt_type_to_legacy(struct hci_dev *hdev, u16 evt_type)
 	if (evt_type & LE_EXT_ADV_SCAN_IND)
 		return LE_ADV_SCAN_IND;
 
-	if (evt_type == LE_EXT_ADV_NON_CONN_IND ||
-	    evt_type & LE_EXT_ADV_DIRECT_IND)
+	if (evt_type & LE_EXT_ADV_DIRECT_IND)
 		return LE_ADV_NONCONN_IND;
 
 invalid:
Signature · 1 line
2.39.5
author maintainer
#271 lore raw

PATCH 261/627 bpf: Disable migration in nf_hook_run_bpf().

In reply to Greg Kroah-Hartman

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

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

From: Kuniyuki Iwashima <[email protected]>

[ Upstream commit 17ce3e5949bc37557305ad46316f41c7875d6366 ]

syzbot reported that the netfilter bpf prog can be called without migration disabled in xmit path.

Then the assertion in __bpf_prog_run() fails, triggering the splat below. [0]

Let's use bpf_prog_run_pin_on_cpu() in nf_hook_run_bpf().

[0]:

BUG: assuming non migratable context at ./include/linux/filter.h:703

in_atomic(): 0, irqs_disabled(): 0, migration_disabled() 0 pid: 5829, name: sshd-session
3 locks held by sshd-session/5829:
#0: ffff88807b4e4218 (sk_lock-AF_INET){+.+.}-{0:0}, at: lock_sock include/net/sock.h:1667 [inline] #0: ffff88807b4e4218 (sk_lock-AF_INET){+.+.}-{0:0}, at: tcp_sendmsg+0x20/0x50 net/ipv4/tcp.c:1395 #1: ffffffff8e5c4e00 (rcu_read_lock){....}-{1:3}, at: rcu_lock_acquire include/linux/rcupdate.h:331 [inline] #1: ffffffff8e5c4e00 (rcu_read_lock){....}-{1:3}, at: rcu_read_lock include/linux/rcupdate.h:841 [inline] #1: ffffffff8e5c4e00 (rcu_read_lock){....}-{1:3}, at: __ip_queue_xmit+0x69/0x26c0 net/ipv4/ip_output.c:470 #2: ffffffff8e5c4e00 (rcu_read_lock){....}-{1:3}, at: rcu_lock_acquire include/linux/rcupdate.h:331 [inline] #2: ffffffff8e5c4e00 (rcu_read_lock){....}-{1:3}, at: rcu_read_lock include/linux/rcupdate.h:841 [inline] #2: ffffffff8e5c4e00 (rcu_read_lock){....}-{1:3}, at: nf_hook+0xb2/0x680 include/linux/netfilter.h:241

CPU: 0 UID: 0 PID: 5829 Comm: sshd-session Not tainted 6.16.0-rc6-syzkaller-00002-g155a3c003e55 #0 PREEMPT(full)
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 05/07/2025
Call Trace:
 <TASK>
 __dump_stack lib/dump_stack.c:94 [inline]
 dump_stack_lvl+0x16c/0x1f0 lib/dump_stack.c:120
 __cant_migrate kernel/sched/core.c:8860 [inline]
 __cant_migrate+0x1c7/0x250 kernel/sched/core.c:8834
 __bpf_prog_run include/linux/filter.h:703 [inline]
 bpf_prog_run include/linux/filter.h:725 [inline]
 nf_hook_run_bpf+0x83/0x1e0 net/netfilter/nf_bpf_link.c:20
 nf_hook_entry_hookfn include/linux/netfilter.h:157 [inline]
 nf_hook_slow+0xbb/0x200 net/netfilter/core.c:623
 nf_hook+0x370/0x680 include/linux/netfilter.h:272
 NF_HOOK_COND include/linux/netfilter.h:305 [inline]
 ip_output+0x1bc/0x2a0 net/ipv4/ip_output.c:433
 dst_output include/net/dst.h:459 [inline]
 ip_local_out net/ipv4/ip_output.c:129 [inline]
 __ip_queue_xmit+0x1d7d/0x26c0 net/ipv4/ip_output.c:527
 __tcp_transmit_skb+0x2686/0x3e90 net/ipv4/tcp_output.c:1479
 tcp_transmit_skb net/ipv4/tcp_output.c:1497 [inline]
 tcp_write_xmit+0x1274/0x84e0 net/ipv4/tcp_output.c:2838
 __tcp_push_pending_frames+0xaf/0x390 net/ipv4/tcp_output.c:3021
 tcp_push+0x225/0x700 net/ipv4/tcp.c:759
 tcp_sendmsg_locked+0x1870/0x42b0 net/ipv4/tcp.c:1359
 tcp_sendmsg+0x2e/0x50 net/ipv4/tcp.c:1396
 inet_sendmsg+0xb9/0x140 net/ipv4/af_inet.c:851
 sock_sendmsg_nosec net/socket.c:712 [inline]
 __sock_sendmsg net/socket.c:727 [inline]
 sock_write_iter+0x4aa/0x5b0 net/socket.c:1131
 new_sync_write fs/read_write.c:593 [inline]
 vfs_write+0x6c7/0x1150 fs/read_write.c:686
 ksys_write+0x1f8/0x250 fs/read_write.c:738
 do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
 do_syscall_64+0xcd/0x4c0 arch/x86/entry/syscall_64.c:94
 entry_SYSCALL_64_after_hwframe+0x77/0x7f
RIP: 0033:0x7fe7d365d407
Code: 48 89 fa 4c 89 df e8 38 aa 00 00 8b 93 08 03 00 00 59 5e 48 83 f8 fc 74 1a 5b c3 0f 1f 84 00 00 00 00 00 48 8b 44 24 10 0f 05 <5b> c3 0f 1f 80 00 00 00 00 83 e2 39 83 fa 08 75 de e8 23 ff ff ff

RSP:

---
 net/netfilter/nf_bpf_link.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/netfilter/nf_bpf_link.c +1 -1
index 06b084844700..25bbac8986c2 100644
@@ -17,7 +17,7 @@ static unsigned int nf_hook_run_bpf(void *bpf_prog, struct sk_buff *skb,
 		.skb = skb,
 	};
 
-	return bpf_prog_run(prog, &ctx);
+	return bpf_prog_run_pin_on_cpu(prog, &ctx);
 }
 
 struct bpf_nf_link {
Signature · 1 line
2.39.5
author maintainer
#272 lore raw

PATCH 262/627 bpf: Reject narrower access to pointer ctx fields

In reply to Greg Kroah-Hartman

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

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

From: Paul Chaignon <[email protected]>

[ Upstream commit e09299225d5ba3916c91ef70565f7d2187e4cca0 ]

The following BPF program, simplified from a syzkaller repro, causes a kernel warning:

    r0 = *(u8 *)(r1 + 169);
    exit;

With pointer field sk being at offset 168 in __sk_buff. This access is detected as a narrower read in bpf_skb_is_valid_access because it doesn't match offsetof(struct __sk_buff, sk). It is therefore allowed and later proceeds to bpf_convert_ctx_access. Note that for the "is_narrower_load" case in the convert_ctx_accesses(), the insn->off is aligned, so the cnt may not be 0 because it matches the offsetof(struct __sk_buff, sk) in the bpf_convert_ctx_access. However, the target_size stays 0 and the verifier errors with a kernel warning:

verifier bug: error during ctx access conversion(1)

This patch fixes that to return a proper "invalid bpf_context access off=X size=Y" error on the load instruction.

The same issue affects multiple other fields in context structures that allow narrow access. Some other non-affected fields (for sk_msg, sk_lookup, and sockopt) were also changed to use bpf_ctx_range_ptr for consistency.

Note this syzkaller crash was reported in the "Closes" link below, which used to be about a different bug, fixed in
commit fce7bd8e385a ("bpf/verifier: Handle BPF_LOAD_ACQ instructions in insn_def_regno()"). Because syzbot somehow confused the two bugs, the new crash and repro didn't get reported to the mailing list.

---
 kernel/bpf/cgroup.c |  8 ++++----
 net/core/filter.c   | 20 ++++++++++----------
 2 files changed, 14 insertions(+), 14 deletions(-)
kernel/bpf/cgroup.c +4 -4
index f4885514f007..deb88fade249 100644
@@ -2440,22 +2440,22 @@ static bool cg_sockopt_is_valid_access(int off, int size,
 	}
 
 	switch (off) {
-	case offsetof(struct bpf_sockopt, sk):
+	case bpf_ctx_range_ptr(struct bpf_sockopt, sk):
 		if (size != sizeof(__u64))
 			return false;
 		info->reg_type = PTR_TO_SOCKET;
 		break;
-	case offsetof(struct bpf_sockopt, optval):
+	case bpf_ctx_range_ptr(struct bpf_sockopt, optval):
 		if (size != sizeof(__u64))
 			return false;
 		info->reg_type = PTR_TO_PACKET;
 		break;
-	case offsetof(struct bpf_sockopt, optval_end):
+	case bpf_ctx_range_ptr(struct bpf_sockopt, optval_end):
 		if (size != sizeof(__u64))
 			return false;
 		info->reg_type = PTR_TO_PACKET_END;
 		break;
-	case offsetof(struct bpf_sockopt, retval):
+	case bpf_ctx_range(struct bpf_sockopt, retval):
 		if (size != size_default)
 			return false;
 		return prog->expected_attach_type == BPF_CGROUP_GETSOCKOPT;
net/core/filter.c +10 -10
index 7a72f766aacf..47073a0180a4 100644
@@ -8690,7 +8690,7 @@ static bool bpf_skb_is_valid_access(int off, int size, enum bpf_access_type type
 		if (size != sizeof(__u64))
 			return false;
 		break;
-	case offsetof(struct __sk_buff, sk):
+	case bpf_ctx_range_ptr(struct __sk_buff, sk):
 		if (type == BPF_WRITE || size != sizeof(__u64))
 			return false;
 		info->reg_type = PTR_TO_SOCK_COMMON_OR_NULL;
@@ -9268,7 +9268,7 @@ static bool sock_addr_is_valid_access(int off, int size,
 				return false;
 		}
 		break;
-	case offsetof(struct bpf_sock_addr, sk):
+	case bpf_ctx_range_ptr(struct bpf_sock_addr, sk):
 		if (type != BPF_READ)
 			return false;
 		if (size != sizeof(__u64))
@@ -9318,17 +9318,17 @@ static bool sock_ops_is_valid_access(int off, int size,
 			if (size != sizeof(__u64))
 				return false;
 			break;
-		case offsetof(struct bpf_sock_ops, sk):
+		case bpf_ctx_range_ptr(struct bpf_sock_ops, sk):
 			if (size != sizeof(__u64))
 				return false;
 			info->reg_type = PTR_TO_SOCKET_OR_NULL;
 			break;
-		case offsetof(struct bpf_sock_ops, skb_data):
+		case bpf_ctx_range_ptr(struct bpf_sock_ops, skb_data):
 			if (size != sizeof(__u64))
 				return false;
 			info->reg_type = PTR_TO_PACKET;
 			break;
-		case offsetof(struct bpf_sock_ops, skb_data_end):
+		case bpf_ctx_range_ptr(struct bpf_sock_ops, skb_data_end):
 			if (size != sizeof(__u64))
 				return false;
 			info->reg_type = PTR_TO_PACKET_END;
@@ -9337,7 +9337,7 @@ static bool sock_ops_is_valid_access(int off, int size,
 			bpf_ctx_record_field_size(info, size_default);
 			return bpf_ctx_narrow_access_ok(off, size,
 							size_default);
-		case offsetof(struct bpf_sock_ops, skb_hwtstamp):
+		case bpf_ctx_range(struct bpf_sock_ops, skb_hwtstamp):
 			if (size != sizeof(__u64))
 				return false;
 			break;
@@ -9407,17 +9407,17 @@ static bool sk_msg_is_valid_access(int off, int size,
 		return false;
 
 	switch (off) {
-	case offsetof(struct sk_msg_md, data):
+	case bpf_ctx_range_ptr(struct sk_msg_md, data):
 		info->reg_type = PTR_TO_PACKET;
 		if (size != sizeof(__u64))
 			return false;
 		break;
-	case offsetof(struct sk_msg_md, data_end):
+	case bpf_ctx_range_ptr(struct sk_msg_md, data_end):
 		info->reg_type = PTR_TO_PACKET_END;
 		if (size != sizeof(__u64))
 			return false;
 		break;
-	case offsetof(struct sk_msg_md, sk):
+	case bpf_ctx_range_ptr(struct sk_msg_md, sk):
 		if (size != sizeof(__u64))
 			return false;
 		info->reg_type = PTR_TO_SOCKET;
@@ -11623,7 +11623,7 @@ static bool sk_lookup_is_valid_access(int off, int size,
 		return false;
 
 	switch (off) {
-	case offsetof(struct bpf_sk_lookup, sk):
+	case bpf_ctx_range_ptr(struct bpf_sk_lookup, sk):
 		info->reg_type = PTR_TO_SOCKET_OR_NULL;
 		return size == sizeof(__u64);
 
Signature · 1 line
2.39.5
author maintainer
#273 lore raw

PATCH 263/627 tools/rv: Do not skip idle in trace

In reply to Greg Kroah-Hartman

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

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

From: Gabriele Monaco <[email protected]>

[ Upstream commit f60227f3448911b682c45041c3fbd94f6d3b15a2 ]

Currently, the userspace RV tool skips trace events triggered by the RV tool itself, this can be changed by passing the parameter -s, which sets the variable config_my_pid to 0 (instead of the tool's PID). This has the side effect of skipping events generated by idle (PID 0).

Set config_my_pid to -1 (an invalid pid) to avoid skipping idle.

Cc Nam Cao <[email protected]>
Cc Tomas Glozar <[email protected]>
Cc Juri Lelli <[email protected]>
Cc Clark Williams <[email protected]>
Cc John Kacur <[email protected]>
Fixes 6d60f89691fc ("tools/rv: Add in-kernel monitor interface")
---
 tools/verification/rv/src/in_kernel.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
tools/verification/rv/src/in_kernel.c +2 -2
index c0dcee795c0d..4bb746ea6e17 100644
@@ -431,7 +431,7 @@ ikm_event_handler(struct trace_seq *s, struct tep_record *record,
 
 	if (config_has_id && (config_my_pid == id))
 		return 0;
-	else if (config_my_pid && (config_my_pid == pid))
+	else if (config_my_pid == pid)
 		return 0;
 
 	tep_print_event(trace_event->tep, s, record, "%16s-%-8d [%.3d] ",
@@ -734,7 +734,7 @@ static int parse_arguments(char *monitor_name, int argc, char **argv)
 			config_reactor = optarg;
 			break;
 		case 's':
-			config_my_pid = 0;
+			config_my_pid = -1;
 			break;
 		case 't':
 			config_trace = 1;
Signature · 1 line
2.39.5
author maintainer
#274 lore raw

PATCH 264/627 selftests: drv-net: Fix remote command checking in require_cmd()

In reply to Greg Kroah-Hartman

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

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

From: Gal Pressman <[email protected]>

[ Upstream commit b4d52c698210ae1a3ceb487b189701bc70551a48 ]

The require_cmd() method was checking for command availability locally even when remote=True was specified, due to a missing host parameter.

Fix by passing host=self.remote when checking remote command availability, ensuring commands are verified on the correct host.

Fixes f1e68a1a4a40 ("selftests: drv-net: add require_XYZ() helpers for validating env")
---
 tools/testing/selftests/drivers/net/lib/py/env.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
tools/testing/selftests/drivers/net/lib/py/env.py +1 -1
index 3bccddf8cbc5..1b8bd648048f 100644
@@ -259,7 +259,7 @@ class NetDrvEpEnv(NetDrvEnvBase):
             if not self._require_cmd(comm, "local"):
                 raise KsftSkipEx("Test requires command: " + comm)
         if remote:
-            if not self._require_cmd(comm, "remote"):
+            if not self._require_cmd(comm, "remote", host=self.remote):
                 raise KsftSkipEx("Test requires (remote) command: " + comm)
 
     def wait_hw_stats_settle(self):
Signature · 1 line
2.39.5
author maintainer
#275 lore raw

PATCH 265/627 selftests: drv-net: tso: enable test cases based on hw_features

In reply to Greg Kroah-Hartman

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

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

From: Daniel Zahka <[email protected]>

[ Upstream commit 266b835e5e84a0f8fec7fd988ee81925890e8d89 ]

tso.py uses the active features at the time of test execution as the set of available gso features to test. This means if a gso feature is supported but toggled off at test start, the test will be skipped with a "Device does not support {feature}" message.

Instead, we can enumerate the set of toggleable features by capturing the driver's hw_features bitmap. To avoid configuration side-effects from running the test, we also snapshot the wanted_features flag set before making any feature changes, and then attempt to restore the same set of wanted_features before test exit.

Fixes 0d0f4174f6c8 ("selftests: drv-net: add a simple TSO test")
---
 tools/testing/selftests/drivers/net/hw/tso.py | 52 ++++++++++++++-----
 1 file changed, 40 insertions(+), 12 deletions(-)
tools/testing/selftests/drivers/net/hw/tso.py +40 -12
index 3370827409aa..f8386e3d88cd 100755
@@ -119,15 +119,30 @@ def build_tunnel(cfg, outer_ipver, tun_info):
     return remote_v4, remote_v6
 
 
+def restore_wanted_features(cfg):
+    features_cmd = ""
+    for feature in cfg.hw_features:
+        setting = "on" if feature in cfg.wanted_features else "off"
+        features_cmd += f" {feature} {setting}"
+    try:
+        ethtool(f"-K {cfg.ifname} {features_cmd}")
+    except Exception as e:
+        ksft_pr(f"WARNING: failure restoring wanted features: {e}")
+
+
 def test_builder(name, cfg, outer_ipver, feature, tun=None, inner_ipver=None):
     """Construct specific tests from the common template."""
     def f(cfg):
         cfg.require_ipver(outer_ipver)
+        defer(restore_wanted_features, cfg)
 
         if not cfg.have_stat_super_count and \
            not cfg.have_stat_wire_count:
             raise KsftSkipEx(f"Device does not support LSO queue stats")
 
+        if feature not in cfg.hw_features:
+            raise KsftSkipEx(f"Device does not support {feature}")
+
         ipver = outer_ipver
         if tun:
             remote_v4, remote_v6 = build_tunnel(cfg, ipver, tun)
@@ -138,12 +153,12 @@ def test_builder(name, cfg, outer_ipver, feature, tun=None, inner_ipver=None):
 
         tun_partial = tun and tun[1]
         # Tunnel which can silently fall back to gso-partial
-        has_gso_partial = tun and 'tx-gso-partial' in cfg.features
+        has_gso_partial = tun and 'tx-gso-partial' in cfg.hw_features
 
         # For TSO4 via partial we need mangleid
         if ipver == "4" and feature in cfg.partial_features:
             ksft_pr("Testing with mangleid enabled")
-            if 'tx-tcp-mangleid-segmentation' not in cfg.features:
+            if 'tx-tcp-mangleid-segmentation' not in cfg.hw_features:
                 ethtool(f"-K {cfg.ifname} tx-tcp-mangleid-segmentation on")
                 defer(ethtool, f"-K {cfg.ifname} tx-tcp-mangleid-segmentation off")
 
@@ -161,11 +176,8 @@ def test_builder(name, cfg, outer_ipver, feature, tun=None, inner_ipver=None):
                            should_lso=tun_partial)
 
         # Full feature enabled.
-        if feature in cfg.features:
-            ethtool(f"-K {cfg.ifname} {feature} on")
-            run_one_stream(cfg, ipver, remote_v4, remote_v6, should_lso=True)
-        else:
-            raise KsftXfailEx(f"Device does not support {feature}")
+        ethtool(f"-K {cfg.ifname} {feature} on")
+        run_one_stream(cfg, ipver, remote_v4, remote_v6, should_lso=True)
 
     f.__name__ = name + ((outer_ipver + "_") if tun else "") + "ipv" + inner_ipver
     return f
@@ -176,23 +188,39 @@ def query_nic_features(cfg) -> None:
     cfg.have_stat_super_count = False
     cfg.have_stat_wire_count = False
 
-    cfg.features = set()
     features = cfg.ethnl.features_get({"header": {"dev-index": cfg.ifindex}})
-    for f in features["active"]["bits"]["bit"]:
-        cfg.features.add(f["name"])
+
+    cfg.wanted_features = set()
+    for f in features["wanted"]["bits"]["bit"]:
+        cfg.wanted_features.add(f["name"])
+
+    cfg.hw_features = set()
+    hw_all_features_cmd = ""
+    for f in features["hw"]["bits"]["bit"]:
+        if f.get("value", False):
+            feature = f["name"]
+            cfg.hw_features.add(feature)
+            hw_all_features_cmd += f" {feature} on"
+    try:
+        ethtool(f"-K {cfg.ifname} {hw_all_features_cmd}")
+    except Exception as e:
+        ksft_pr(f"WARNING: failure enabling all hw features: {e}")
+        ksft_pr("partial gso feature detection may be impacted")
 
     # Check which features are supported via GSO partial
     cfg.partial_features = set()
-    if 'tx-gso-partial' in cfg.features:
+    if 'tx-gso-partial' in cfg.hw_features:
         ethtool(f"-K {cfg.ifname} tx-gso-partial off")
 
         no_partial = set()
         features = cfg.ethnl.features_get({"header": {"dev-index": cfg.ifindex}})
         for f in features["active"]["bits"]["bit"]:
             no_partial.add(f["name"])
-        cfg.partial_features = cfg.features - no_partial
+        cfg.partial_features = cfg.hw_features - no_partial
         ethtool(f"-K {cfg.ifname} tx-gso-partial on")
 
+    restore_wanted_features(cfg)
+
     stats = cfg.netnl.qstats_get({"ifindex": cfg.ifindex}, dump=True)
     if stats:
         if 'tx-hw-gso-packets' in stats[0]:
Signature · 1 line
2.39.5
author maintainer
#276 lore raw

PATCH 266/627 selftests: drv-net: tso: fix vxlan tunnel flags to get correct gso_type

In reply to Greg Kroah-Hartman

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

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

From: Daniel Zahka <[email protected]>

[ Upstream commit 2cfbcc5d8af9199823151c21f740e476b223dd2e ]

When vxlan is used with ipv6 as the outer network header, the correct ip link parameters for acheiving the SKB_GSO_UDP_TUNNEL gso type is "udp6zerocsumtx udp6zerocsumrx". Otherwise the gso type will be SKB_GSO_UDP_TUNNEL_CSUM.

This bug was the reason for the second of the three possible invocations of run_one_stream() invocations, so that can be deleted as well. We only need to test with the feature off and on.

Fixes 0d0f4174f6c8 ("selftests: drv-net: add a simple TSO test")
---
 tools/testing/selftests/drivers/net/hw/tso.py | 37 +++++++------------
 1 file changed, 13 insertions(+), 24 deletions(-)
tools/testing/selftests/drivers/net/hw/tso.py +13 -24
index f8386e3d88cd..6461a83b3d0e 100755
@@ -102,7 +102,7 @@ def build_tunnel(cfg, outer_ipver, tun_info):
     remote_addr = cfg.remote_addr_v[outer_ipver]
 
     tun_type = tun_info[0]
-    tun_arg  = tun_info[2]
+    tun_arg  = tun_info[1]
     ip(f"link add {tun_type}-ksft type {tun_type} {tun_arg} local {local_addr} remote {remote_addr} dev {cfg.ifname}")
     defer(ip, f"link del {tun_type}-ksft")
     ip(f"link set dev {tun_type}-ksft up")
@@ -151,29 +151,17 @@ def test_builder(name, cfg, outer_ipver, feature, tun=None, inner_ipver=None):
             remote_v4 = cfg.remote_addr_v["4"]
             remote_v6 = cfg.remote_addr_v["6"]
 
-        tun_partial = tun and tun[1]
-        # Tunnel which can silently fall back to gso-partial
-        has_gso_partial = tun and 'tx-gso-partial' in cfg.hw_features
-
-        # For TSO4 via partial we need mangleid
-        if ipver == "4" and feature in cfg.partial_features:
-            ksft_pr("Testing with mangleid enabled")
-            if 'tx-tcp-mangleid-segmentation' not in cfg.hw_features:
-                ethtool(f"-K {cfg.ifname} tx-tcp-mangleid-segmentation on")
-                defer(ethtool, f"-K {cfg.ifname} tx-tcp-mangleid-segmentation off")
-
         # First test without the feature enabled.
         ethtool(f"-K {cfg.ifname} {feature} off")
-        if has_gso_partial:
-            ethtool(f"-K {cfg.ifname} tx-gso-partial off")
         run_one_stream(cfg, ipver, remote_v4, remote_v6, should_lso=False)
 
-        # Now test with the feature enabled.
-        # For compatible tunnels only - just GSO partial, not specific feature.
-        if has_gso_partial:
+        ethtool(f"-K {cfg.ifname} tx-gso-partial off")
+        ethtool(f"-K {cfg.ifname} tx-tcp-mangleid-segmentation off")
+        if feature in cfg.partial_features:
             ethtool(f"-K {cfg.ifname} tx-gso-partial on")
-            run_one_stream(cfg, ipver, remote_v4, remote_v6,
-                           should_lso=tun_partial)
+            if ipver == "4":
+                ksft_pr("Testing with mangleid enabled")
+                ethtool(f"-K {cfg.ifname} tx-tcp-mangleid-segmentation on")
 
         # Full feature enabled.
         ethtool(f"-K {cfg.ifname} {feature} on")
@@ -239,13 +227,14 @@ def main() -> None:
         query_nic_features(cfg)
 
         test_info = (
-            # name,       v4/v6  ethtool_feature              tun:(type,    partial, args)
+            # name,       v4/v6  ethtool_feature              tun:(type,     args)
             ("",            "4", "tx-tcp-segmentation",           None),
             ("",            "6", "tx-tcp6-segmentation",          None),
-            ("vxlan",        "", "tx-udp_tnl-segmentation",       ("vxlan",  True,  "id 100 dstport 4789 noudpcsum")),
-            ("vxlan_csum",   "", "tx-udp_tnl-csum-segmentation",  ("vxlan",  False, "id 100 dstport 4789 udpcsum")),
-            ("gre",         "4", "tx-gre-segmentation",           ("gre",    False,  "")),
-            ("gre",         "6", "tx-gre-segmentation",           ("ip6gre", False,  "")),
+            ("vxlan",       "4", "tx-udp_tnl-segmentation",       ("vxlan",  "id 100 dstport 4789 noudpcsum")),
+            ("vxlan",       "6", "tx-udp_tnl-segmentation",       ("vxlan",  "id 100 dstport 4789 udp6zerocsumtx udp6zerocsumrx")),
+            ("vxlan_csum",   "", "tx-udp_tnl-csum-segmentation",  ("vxlan",  "id 100 dstport 4789 udpcsum")),
+            ("gre",         "4", "tx-gre-segmentation",           ("gre",    "")),
+            ("gre",         "6", "tx-gre-segmentation",           ("ip6gre", "")),
         )
 
         cases = []
Signature · 1 line
2.39.5
author maintainer
#277 lore raw

PATCH 267/627 selftests: drv-net: tso: fix non-tunneled tso6 test case name

In reply to Greg Kroah-Hartman

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

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

From: Daniel Zahka <[email protected]>

[ Upstream commit b25b44cd178cc54277f2dc0ff3b3d5a37ae4b26b ]

The non-tunneled tso6 test case was showing up as:
ok 8 tso.ipv4

This is because of the way test_builder() uses the inner_ipver arg in test naming, and how test_info is iterated over in main(). Given that some tunnels not supported yet, e.g. ipip or sit, only support ipv4 or ipv6 as the inner network protocol, I think the best fix here is to call test_builder() in separate branches for tunneled and non-tunneled tests, and to make supported inner l3 types an explicit attribute of tunnel test cases.

# Detected qstat for LSO wire-packets
TAP version 13
1..14
ok 1 tso.ipv4
# Testing with mangleid enabled
ok 2 tso.vxlan4_ipv4
ok 3 tso.vxlan4_ipv6
# Testing with mangleid enabled
ok 4 tso.vxlan_csum4_ipv4
ok 5 tso.vxlan_csum4_ipv6
# Testing with mangleid enabled
ok 6 tso.gre4_ipv4
ok 7 tso.gre4_ipv6
ok 8 tso.ipv6
# Testing with mangleid enabled
ok 9 tso.vxlan6_ipv4
ok 10 tso.vxlan6_ipv6
# Testing with mangleid enabled
ok 11 tso.vxlan_csum6_ipv4
ok 12 tso.vxlan_csum6_ipv6
# Testing with mangleid enabled
ok 13 tso.gre6_ipv4
ok 14 tso.gre6_ipv6
# Totals: pass:14 fail:0 xfail:0 xpass:0 skip:0 error:0

Fixes 0d0f4174f6c8 ("selftests: drv-net: add a simple TSO test")
---
 tools/testing/selftests/drivers/net/hw/tso.py | 26 ++++++++++---------
 1 file changed, 14 insertions(+), 12 deletions(-)
tools/testing/selftests/drivers/net/hw/tso.py +14 -12
index 6461a83b3d0e..5fddb5056a20 100755
@@ -227,14 +227,14 @@ def main() -> None:
         query_nic_features(cfg)
 
         test_info = (
-            # name,       v4/v6  ethtool_feature              tun:(type,     args)
-            ("",            "4", "tx-tcp-segmentation",           None),
-            ("",            "6", "tx-tcp6-segmentation",          None),
-            ("vxlan",       "4", "tx-udp_tnl-segmentation",       ("vxlan",  "id 100 dstport 4789 noudpcsum")),
-            ("vxlan",       "6", "tx-udp_tnl-segmentation",       ("vxlan",  "id 100 dstport 4789 udp6zerocsumtx udp6zerocsumrx")),
-            ("vxlan_csum",   "", "tx-udp_tnl-csum-segmentation",  ("vxlan",  "id 100 dstport 4789 udpcsum")),
-            ("gre",         "4", "tx-gre-segmentation",           ("gre",    "")),
-            ("gre",         "6", "tx-gre-segmentation",           ("ip6gre", "")),
+            # name,       v4/v6  ethtool_feature               tun:(type, args, inner ip versions)
+            ("",           "4", "tx-tcp-segmentation",         None),
+            ("",           "6", "tx-tcp6-segmentation",        None),
+            ("vxlan",      "4", "tx-udp_tnl-segmentation",     ("vxlan", "id 100 dstport 4789 noudpcsum", ("4", "6"))),
+            ("vxlan",      "6", "tx-udp_tnl-segmentation",     ("vxlan", "id 100 dstport 4789 udp6zerocsumtx udp6zerocsumrx", ("4", "6"))),
+            ("vxlan_csum", "", "tx-udp_tnl-csum-segmentation", ("vxlan", "id 100 dstport 4789 udpcsum", ("4", "6"))),
+            ("gre",        "4", "tx-gre-segmentation",         ("gre",   "", ("4", "6"))),
+            ("gre",        "6", "tx-gre-segmentation",         ("ip6gre","", ("4", "6"))),
         )
 
         cases = []
@@ -244,11 +244,13 @@ def main() -> None:
                 if info[1] and outer_ipver != info[1]:
                     continue
 
-                cases.append(test_builder(info[0], cfg, outer_ipver, info[2],
-                                          tun=info[3], inner_ipver="4"))
                 if info[3]:
-                    cases.append(test_builder(info[0], cfg, outer_ipver, info[2],
-                                              tun=info[3], inner_ipver="6"))
+                    cases += [
+                        test_builder(info[0], cfg, outer_ipver, info[2], info[3], inner_ipver)
+                        for inner_ipver in info[3][2]
+                    ]
+                else:
+                    cases.append(test_builder(info[0], cfg, outer_ipver, info[2], None, outer_ipver))
 
         ksft_run(cases=cases, args=(cfg, ))
     ksft_exit()
Signature · 1 line
2.39.5
author maintainer
#278 lore raw

PATCH 268/627 can: peak_usb: fix USB FD devices potential malfunction

In reply to Greg Kroah-Hartman

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

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

From: Stephane Grosjean <[email protected]>

[ Upstream commit 788199b73b6efe4ee2ade4d7457b50bb45493488 ]

The latest firmware versions of USB CAN FD interfaces export the EP numbers to be used to dialog with the device via the "type" field of a response to a vendor request structure, particularly when its value is greater than or equal to 2.

Correct the driver's test of this field.

Fixes 4f232482467a ("can: peak_usb: include support for a new MCU")
[mkl: rephrase commit message]
---
 drivers/net/can/usb/peak_usb/pcan_usb_fd.c | 17 +++++++++--------
 1 file changed, 9 insertions(+), 8 deletions(-)
drivers/net/can/usb/peak_usb/pcan_usb_fd.c +9 -8
index 4d85b29a17b7..ebefc274b50a 100644
@@ -49,7 +49,7 @@ struct __packed pcan_ufd_fw_info {
 	__le32	ser_no;		/* S/N */
 	__le32	flags;		/* special functions */
 
-	/* extended data when type == PCAN_USBFD_TYPE_EXT */
+	/* extended data when type >= PCAN_USBFD_TYPE_EXT */
 	u8	cmd_out_ep;	/* ep for cmd */
 	u8	cmd_in_ep;	/* ep for replies */
 	u8	data_out_ep[2];	/* ep for CANx TX */
@@ -982,10 +982,11 @@ static int pcan_usb_fd_init(struct peak_usb_device *dev)
 			dev->can.ctrlmode |= CAN_CTRLMODE_FD_NON_ISO;
 		}
 
-		/* if vendor rsp is of type 2, then it contains EP numbers to
-		 * use for cmds pipes. If not, then default EP should be used.
+		/* if vendor rsp type is greater than or equal to 2, then it
+		 * contains EP numbers to use for cmds pipes. If not, then
+		 * default EP should be used.
 		 */
-		if (fw_info->type != cpu_to_le16(PCAN_USBFD_TYPE_EXT)) {
+		if (le16_to_cpu(fw_info->type) < PCAN_USBFD_TYPE_EXT) {
 			fw_info->cmd_out_ep = PCAN_USBPRO_EP_CMDOUT;
 			fw_info->cmd_in_ep = PCAN_USBPRO_EP_CMDIN;
 		}
@@ -1018,11 +1019,11 @@ static int pcan_usb_fd_init(struct peak_usb_device *dev)
 	dev->can_channel_id =
 		le32_to_cpu(pdev->usb_if->fw_info.dev_id[dev->ctrl_idx]);
 
-	/* if vendor rsp is of type 2, then it contains EP numbers to
-	 * use for data pipes. If not, then statically defined EP are used
-	 * (see peak_usb_create_dev()).
+	/* if vendor rsp type is greater than or equal to 2, then it contains EP
+	 * numbers to use for data pipes. If not, then statically defined EP are
+	 * used (see peak_usb_create_dev()).
 	 */
-	if (fw_info->type == cpu_to_le16(PCAN_USBFD_TYPE_EXT)) {
+	if (le16_to_cpu(fw_info->type) >= PCAN_USBFD_TYPE_EXT) {
 		dev->ep_msg_in = fw_info->data_in_ep;
 		dev->ep_msg_out = fw_info->data_out_ep[dev->ctrl_idx];
 	}
Signature · 1 line
2.39.5
author maintainer
#279 lore raw

PATCH 269/627 can: tscan1: Kconfig: add COMPILE_TEST

In reply to Greg Kroah-Hartman

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

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

From: Vincent Mailhol <[email protected]>

[ Upstream commit 5323af351e7524497930b7793153ff68ee5c0ec1 ]

tscan1 depends on ISA. It also has a hidden dependency on HAS_IOPORT as reported by the kernel test bot [1]. That dependency is implied by ISA which explains why this was not an issue so far.

Add both COMPILE_TEST and HAS_IOPORT to the dependency list so that this driver can also be built on other platforms.

[1] https://lore.kernel.org/linux-can/[email protected]/

Stable-dep-of: b7d012e59627 ("can: tscan1: CAN_TSCAN1 can depend on PC104")

---
 drivers/net/can/sja1000/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/can/sja1000/Kconfig +1 -1
index 2f516cc6d22c..ba16d7bc09ef 100644
@@ -105,7 +105,7 @@ config CAN_SJA1000_PLATFORM
 
 config CAN_TSCAN1
 	tristate "TS-CAN1 PC104 boards"
-	depends on ISA
+	depends on ISA || (COMPILE_TEST && HAS_IOPORT)
 	help
 	  This driver is for Technologic Systems' TSCAN-1 PC104 boards.
 	  https://www.embeddedts.com/products/TS-CAN1
Signature · 1 line
2.39.5
author maintainer
#280 lore raw

PATCH 270/627 can: tscan1: CAN_TSCAN1 can depend on PC104

In reply to Greg Kroah-Hartman

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

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

From: Randy Dunlap <[email protected]>

[ Upstream commit b7d012e59627c1d1bb2ad5d71efc69a070ef767d ]

Add a dependency on PC104 to limit (restrict) this driver kconfig prompt to kernel configs that have PC104 set.

Add COMPILE_TEST as a possibility for more complete build coverage. I tested this build config on x86_64 5 times without problems.

---
 drivers/net/can/sja1000/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/can/sja1000/Kconfig +1 -1
index ba16d7bc09ef..e061e35769bf 100644
@@ -105,7 +105,7 @@ config CAN_SJA1000_PLATFORM
 
 config CAN_TSCAN1
 	tristate "TS-CAN1 PC104 boards"
-	depends on ISA || (COMPILE_TEST && HAS_IOPORT)
+	depends on (ISA && PC104) || (COMPILE_TEST && HAS_IOPORT)
 	help
 	  This driver is for Technologic Systems' TSCAN-1 PC104 boards.
 	  https://www.embeddedts.com/products/TS-CAN1
Signature · 1 line
2.39.5
author maintainer
#281 lore raw

PATCH 271/627 can: kvaser_pciefd: Store device channel index

In reply to Greg Kroah-Hartman

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

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

From: Jimmy Assarsson <[email protected]>

[ Upstream commit d54b16b40ddadb7d0a77fff48af7b319a0cd6aae ]

Store device channel index in netdev.dev_port.

Fixes 26ad340e582d ("can: kvaser_pciefd: Add driver for Kvaser PCIEcan devices")
---
 drivers/net/can/kvaser_pciefd.c | 1 +
 1 file changed, 1 insertion(+)
drivers/net/can/kvaser_pciefd.c +1 -0
index 09510663988c..dc748797416e 100644
@@ -982,6 +982,7 @@ static int kvaser_pciefd_setup_can_ctrls(struct kvaser_pciefd *pcie)
 		can->completed_tx_bytes = 0;
 		can->bec.txerr = 0;
 		can->bec.rxerr = 0;
+		can->can.dev->dev_port = i;
 
 		init_completion(&can->start_comp);
 		init_completion(&can->flush_comp);
Signature · 1 line
2.39.5
author maintainer
#282 lore raw

PATCH 272/627 can: kvaser_usb: Assign netdev.dev_port based on device channel index

In reply to Greg Kroah-Hartman

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

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

From: Jimmy Assarsson <[email protected]>

[ Upstream commit c151b06a087a61c7a1790b75ee2f1d6edb6a8a45 ]

Assign netdev.dev_port based on the device channel index, to indicate the port number of the network device.
While this driver already uses netdev.dev_id for that purpose, dev_port is more appropriate. However, retain dev_id to avoid potential regressions.

Fixes 3e66d0138c05 ("can: populate netdev::dev_id for udev discrimination")
---
 drivers/net/can/usb/kvaser_usb/kvaser_usb_core.c | 1 +
 1 file changed, 1 insertion(+)
drivers/net/can/usb/kvaser_usb/kvaser_usb_core.c +1 -0
index daf42080f942..e863a9b0e303 100644
@@ -852,6 +852,7 @@ static int kvaser_usb_init_one(struct kvaser_usb *dev, int channel)
 	netdev->ethtool_ops = &kvaser_usb_ethtool_ops;
 	SET_NETDEV_DEV(netdev, &dev->intf->dev);
 	netdev->dev_id = channel;
+	netdev->dev_port = channel;
 
 	dev->nets[channel] = priv;
 
Signature · 1 line
2.39.5
author maintainer
#283 lore raw

PATCH 273/627 netfilter: xt_nfacct: dont assume acct name is null-terminated

In reply to Greg Kroah-Hartman

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

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

From: Florian Westphal <[email protected]>

[ Upstream commit bf58e667af7d96c8eb9411f926a0a0955f41ce21 ]

BUG: KASAN: slab-out-of-bounds in .. lib/vsprintf.c:721
Read of size 1 at addr ffff88801eac95c8 by task syz-executor183/5851
[..]
 string+0x231/0x2b0 lib/vsprintf.c:721
 vsnprintf+0x739/0xf00 lib/vsprintf.c:2874
 [..]
 nfacct_mt_checkentry+0xd2/0xe0 net/netfilter/xt_nfacct.c:41
 xt_check_match+0x3d1/0xab0 net/netfilter/x_tables.c:523

nfnl_acct_find_get() handles non-null input, but the error printk relied on its presence.

Reported-by [email protected]
Fixes ceb98d03eac5 ("netfilter: xtables: add nfacct match to support extended accounting")
---
 net/netfilter/xt_nfacct.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
net/netfilter/xt_nfacct.c +2 -2
index 7c6bf1c16813..0ca1cdfc4095 100644
@@ -38,8 +38,8 @@ nfacct_mt_checkentry(const struct xt_mtchk_param *par)
 
 	nfacct = nfnl_acct_find_get(par->net, info->name);
 	if (nfacct == NULL) {
-		pr_info_ratelimited("accounting object `%s' does not exists\n",
-				    info->name);
+		pr_info_ratelimited("accounting object `%.*s' does not exist\n",
+				    NFACCT_NAME_MAX, info->name);
 		return -ENOENT;
 	}
 	info->nfacct = nfacct;
Signature · 1 line
2.39.5
author maintainer
#284 lore raw

PATCH 274/627 selftests: netfilter: Ignore tainted kernels in interface stress test

In reply to Greg Kroah-Hartman

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

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

From: Phil Sutter <[email protected]>

[ Upstream commit 8d1c91850d064944ab214b2fbfffb7fc08a11d65 ]

Complain about kernel taint value only if it wasn't set at start already.

Fixes 73db1b5dab6f ("selftests: netfilter: Torture nftables netdev hooks")
---
 .../testing/selftests/net/netfilter/nft_interface_stress.sh  | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
tools/testing/selftests/net/netfilter/nft_interface_stress.sh +4 -1
index 5ff7be9daeee..c0fffaa6dbd9 100755
@@ -10,6 +10,8 @@ source lib.sh
 checktool "nft --version" "run test without nft tool"
 checktool "iperf3 --version" "run test without iperf3 tool"
 
+read kernel_tainted < /proc/sys/kernel/tainted
+
 # how many seconds to torture the kernel?
 # default to 80% of max run time but don't exceed 48s
 TEST_RUNTIME=$((${kselftest_timeout:-60} * 8 / 10))
@@ -135,7 +137,8 @@ else
 	wait
 fi
 
-[[ $(</proc/sys/kernel/tainted) -eq 0 ]] || {
+
+[[ $kernel_tainted -eq 0 && $(</proc/sys/kernel/tainted) -ne 0 ]] && {
 	echo "FAIL: Kernel is tainted!"
 	exit $ksft_fail
 }
Signature · 1 line
2.39.5
author maintainer
#285 lore raw

PATCH 275/627 selftests: netfilter: ipvs.sh: Explicity disable rp_filter on interface tunl0

In reply to Greg Kroah-Hartman

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

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

From: Yi Chen <[email protected]>

[ Upstream commit 8b4a1a46e84a17f5d6fde5c506cc6bb141a24772 ]

Although setup_ns() set net.ipv4.conf.default.rp_filter=0,
loading certain module such as ipip will automatically create a tunl0 interface in all netns including new created ones. In the script, this is before than default.rp_filter=0 applied, as a result tunl0.rp_filter remains set to 1 which causes the test report FAIL when ipip module is preloaded.

Before fix:
Testing DR mode...
Testing NAT mode...
Testing Tunnel mode...
ipvs.sh: FAIL

After fix:
Testing DR mode...
Testing NAT mode...
Testing Tunnel mode...
ipvs.sh: PASS

Fixes 7c8b89ec506e ("selftests: netfilter: remove rp_filter configuration")
---
 tools/testing/selftests/net/netfilter/ipvs.sh | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
tools/testing/selftests/net/netfilter/ipvs.sh +2 -2
index 6af2ea3ad6b8..9c9d5b38ab71 100755
@@ -151,7 +151,7 @@ test_nat() {
 test_tun() {
 	ip netns exec "${ns0}" ip route add "${vip_v4}" via "${gip_v4}" dev br0
 
-	ip netns exec "${ns1}" modprobe -q ipip
+	modprobe -q ipip
 	ip netns exec "${ns1}" ip link set tunl0 up
 	ip netns exec "${ns1}" sysctl -qw net.ipv4.ip_forward=0
 	ip netns exec "${ns1}" sysctl -qw net.ipv4.conf.all.send_redirects=0
@@ -160,10 +160,10 @@ test_tun() {
 	ip netns exec "${ns1}" ipvsadm -a -i -t "${vip_v4}:${port}" -r ${rip_v4}:${port}
 	ip netns exec "${ns1}" ip addr add ${vip_v4}/32 dev lo:1
 
-	ip netns exec "${ns2}" modprobe -q ipip
 	ip netns exec "${ns2}" ip link set tunl0 up
 	ip netns exec "${ns2}" sysctl -qw net.ipv4.conf.all.arp_ignore=1
 	ip netns exec "${ns2}" sysctl -qw net.ipv4.conf.all.arp_announce=2
+	ip netns exec "${ns2}" sysctl -qw net.ipv4.conf.tunl0.rp_filter=0
 	ip netns exec "${ns2}" ip addr add "${vip_v4}/32" dev lo:1
 
 	test_service
Signature · 1 line
2.39.5
author maintainer
#286 lore raw

PATCH 276/627 net/mlx5e: Clear Read-Only port buffer size in PBMC before update

In reply to Greg Kroah-Hartman

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

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

From: Alexei Lazar <[email protected]>

[ Upstream commit fd4b97246a23c1149479b88490946bcfbd28de63 ]

When updating the PBMC register, we read its current value, modify desired fields, then write it back.

The port_buffer_size field within PBMC is Read-Only (RO). If this RO field contains a non-zero value when read, attempting to write it back will cause the entire PBMC register update to fail.

This commit ensures port_buffer_size is explicitly cleared to zero after reading the PBMC register but before writing back the modified value.
This allows updates to other fields in the PBMC register to succeed.

---
 drivers/net/ethernet/mellanox/mlx5/core/en/port_buffer.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/net/ethernet/mellanox/mlx5/core/en/port_buffer.c +3 -0
index 8e25f4ef5ccc..5ae787656a7c 100644
@@ -331,6 +331,9 @@ static int port_set_buffer(struct mlx5e_priv *priv,
 	if (err)
 		goto out;
 
+	/* RO bits should be set to 0 on write */
+	MLX5_SET(pbmc_reg, in, port_buffer_size, 0);
+
 	err = mlx5e_port_set_pbmc(mdev, in);
 out:
 	kfree(in);
Signature · 1 line
2.39.5
author maintainer
#287 lore raw

PATCH 277/627 net/mlx5e: Remove skb secpath if xfrm state is not found

In reply to Greg Kroah-Hartman

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

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

From: Jianbo Liu <[email protected]>

[ Upstream commit 6d19c44b5c6dd72f9a357d0399604ec16a77de3c ]

Hardware returns a unique identifier for a decrypted packet's xfrm state, this state is looked up in an xarray. However, the state might have been freed by the time of this lookup.

Currently, if the state is not found, only a counter is incremented. The secpath (sp) extension on the skb is not removed, resulting in sp->len becoming 0.

Subsequently, functions like __xfrm_policy_check() attempt to access fields such as xfrm_input_state(skb)->xso.type (which dereferences sp->xvec[sp->len - 1]) without first validating sp->len. This leads to a crash when dereferencing an invalid state pointer.

This patch prevents the crash by explicitly removing the secpath extension from the skb if the xfrm state is not found after hardware decryption. This ensures downstream functions do not operate on a zero-length secpath.

 BUG: unable to handle page fault for address: ffffffff000002c8
 #PF: supervisor read access in kernel mode
 #PF: error_code(0x0000) - not-present page
 PGD 282e067 P4D 282e067 PUD 0
 Oops: Oops: 0000 [#1] SMP
 CPU: 12 UID: 0 PID: 0 Comm: swapper/12 Not tainted 6.15.0-rc7_for_upstream_min_debug_2025_05_27_22_44 #1 NONE
 Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.13.0-0-gf21b5a4aeb02-prebuilt.qemu.org 04/01/2014
 RIP: 0010:__xfrm_policy_check+0x61a/0xa30
 Code: b6 77 7f 83 e6 02 74 14 4d 8b af d8 00 00 00 41 0f b6 45 05 c1 e0 03 48 98 49 01 c5 41 8b 45 00 83 e8 01 48 98 49 8b 44 c5 10 <0f> b6 80 c8 02 00 00 83 e0 0c 3c 04 0f 84 0c 02 00 00 31 ff 80 fa
 RSP: 0018:ffff88885fb04918 EFLAGS: 00010297
 RAX: ffffffff00000000 RBX: 0000000000000002 RCX: 0000000000000000
 RDX: 0000000000000002 RSI: 0000000000000002 RDI: 0000000000000000
 RBP: ffffffff8311af80 R08: 0000000000000020 R09: 00000000c2eda353
 R10: ffff88812be2bbc8 R11: 000000001faab533 R12: ffff88885fb049c8
 R13: ffff88812be2bbc8 R14: 0000000000000000 R15: ffff88811896ae00
 FS:  0000000000000000(0000) GS:ffff8888dca82000(0000) knlGS:0000000000000000
 CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
 CR2: ffffffff000002c8 CR3: 0000000243050002 CR4: 0000000000372eb0
 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
 Call Trace:
  <IRQ>
  ? try_to_wake_up+0x108/0x4c0
  ? udp4_lib_lookup2+0xbe/0x150
  ? udp_lib_lport_inuse+0x100/0x100
  ? __udp4_lib_lookup+0x2b0/0x410
  __xfrm_policy_check2.constprop.0+0x11e/0x130
  udp_queue_rcv_one_skb+0x1d/0x530
  udp_unicast_rcv_skb+0x76/0x90
  __udp4_lib_rcv+0xa64/0xe90
  ip_protocol_deliver_rcu+0x20/0x130
  ip_local_deliver_finish+0x75/0xa0
  ip_local_deliver+0xc1/0xd0
  ? ip_protocol_deliver_rcu+0x130/0x130
  ip_sublist_rcv+0x1f9/0x240
  ? ip_rcv_finish_core+0x430/0x430
  ip_list_rcv+0xfc/0x130
  __netif_receive_skb_list_core+0x181/0x1e0
  netif_receive_skb_list_internal+0x200/0x360
  ? mlx5e_build_rx_skb+0x1bc/0xda0 [mlx5_core]
  gro_receive_skb+0xfd/0x210
  mlx5e_handle_rx_cqe_mpwrq+0x141/0x280 [mlx5_core]
  mlx5e_poll_rx_cq+0xcc/0x8e0 [mlx5_core]
  ? mlx5e_handle_rx_dim+0x91/0xd0 [mlx5_core]
  mlx5e_napi_poll+0x114/0xab0 [mlx5_core]
  __napi_poll+0x25/0x170
  net_rx_action+0x32d/0x3a0
  ? mlx5_eq_comp_int+0x8d/0x280 [mlx5_core]
  ? notifier_call_chain+0x33/0xa0
  handle_softirqs+0xda/0x250
  irq_exit_rcu+0x6d/0xc0
  common_interrupt+0x81/0xa0
  </IRQ>
Fixes b2ac7541e377 ("net/mlx5e: IPsec: Add Connect-X IPsec Rx data path offload")
---
 drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_rxtx.c | 4 ++++
 1 file changed, 4 insertions(+)
drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_rxtx.c +4 -0
index 727fa7c18523..6056106edcc6 100644
@@ -327,6 +327,10 @@ void mlx5e_ipsec_offload_handle_rx_skb(struct net_device *netdev,
 	if (unlikely(!sa_entry)) {
 		rcu_read_unlock();
 		atomic64_inc(&ipsec->sw_stats.ipsec_rx_drop_sadb_miss);
+		/* Clear secpath to prevent invalid dereference
+		 * in downstream XFRM policy checks.
+		 */
+		secpath_reset(skb);
 		return;
 	}
 	xfrm_state_hold(sa_entry->x);
Signature · 1 line
2.39.5
author maintainer
#288 lore raw

PATCH 278/627 net/mlx5e: Fix potential deadlock by deferring RX timeout recovery

In reply to Greg Kroah-Hartman

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

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

From: Shahar Shitrit <[email protected]>

[ Upstream commit e80d65561571db5024fbdd5ec3f5472cfc485d21 ]

mlx5e_reporter_rx_timeout() is currently invoked synchronously in the driver's open error flow. This causes the thread holding priv->state_lock to attempt acquiring the devlink lock, which can result in a circular dependency with other devlink operations.

For example:

- Devlink health diagnose flow:
- __devlink_nl_pre_doit() acquires the devlink lock.
- devlink_nl_health_reporter_diagnose_doit() invokes the driver's diagnose callback.
- mlx5e_rx_reporter_diagnose() then attempts to acquire

    priv->state_lock.

- Driver open flow:
- mlx5e_open() acquires priv->state_lock.
- If an error occurs, devlink_health_reporter may be called, attempting to acquire the devlink lock.

To prevent this circular locking scenario, defer the RX timeout recovery by scheduling it via a workqueue. This ensures that the recovery work acquires locks in a consistent order: first the devlink lock, then priv->state_lock.

Additionally, make the recovery work acquire the netdev instance lock to safely synchronize with the open/close channel flows, similar to mlx5e_tx_timeout_work. Repeatedly attempt to acquire the netdev instance lock until it is taken or the target RQ is no longer active, as indicated by the MLX5E_STATE_CHANNELS_ACTIVE bit.

---
 drivers/net/ethernet/mellanox/mlx5/core/en.h  |  1 +
 .../mellanox/mlx5/core/en/reporter_rx.c       |  7 +++++
 .../net/ethernet/mellanox/mlx5/core/en_main.c | 26 ++++++++++++++++++-
 3 files changed, 33 insertions(+), 1 deletion(-)
drivers/net/ethernet/mellanox/mlx5/core/en.h +1 -0
index 5b0d03b3efe8..48bcd6813aff 100644
@@ -728,6 +728,7 @@ struct mlx5e_rq {
 	struct xsk_buff_pool  *xsk_pool;
 
 	struct work_struct     recover_work;
+	struct work_struct     rx_timeout_work;
 
 	/* control */
 	struct mlx5_wq_ctrl    wq_ctrl;
drivers/net/ethernet/mellanox/mlx5/core/en/reporter_rx.c +7 -0
index e75759533ae0..16c44d628eda 100644
@@ -170,16 +170,23 @@ static int mlx5e_rx_reporter_err_rq_cqe_recover(void *ctx)
 static int mlx5e_rx_reporter_timeout_recover(void *ctx)
 {
 	struct mlx5_eq_comp *eq;
+	struct mlx5e_priv *priv;
 	struct mlx5e_rq *rq;
 	int err;
 
 	rq = ctx;
+	priv = rq->priv;
+
+	mutex_lock(&priv->state_lock);
+
 	eq = rq->cq.mcq.eq;
 
 	err = mlx5e_health_channel_eq_recover(rq->netdev, eq, rq->cq.ch_stats);
 	if (err && rq->icosq)
 		clear_bit(MLX5E_SQ_STATE_ENABLED, &rq->icosq->state);
 
+	mutex_unlock(&priv->state_lock);
+
 	return err;
 }
 
drivers/net/ethernet/mellanox/mlx5/core/en_main.c +25 -1
index ea822c69d137..16d818943487 100644
@@ -707,6 +707,27 @@ static void mlx5e_rq_err_cqe_work(struct work_struct *recover_work)
 	mlx5e_reporter_rq_cqe_err(rq);
 }
 
+static void mlx5e_rq_timeout_work(struct work_struct *timeout_work)
+{
+	struct mlx5e_rq *rq = container_of(timeout_work,
+					   struct mlx5e_rq,
+					   rx_timeout_work);
+
+	/* Acquire netdev instance lock to synchronize with channel close and
+	 * reopen flows. Either successfully obtain the lock, or detect that
+	 * channels are closing for another reason, making this work no longer
+	 * necessary.
+	 */
+	while (!netdev_trylock(rq->netdev)) {
+		if (!test_bit(MLX5E_STATE_CHANNELS_ACTIVE, &rq->priv->state))
+			return;
+		msleep(20);
+	}
+
+	mlx5e_reporter_rx_timeout(rq);
+	netdev_unlock(rq->netdev);
+}
+
 static int mlx5e_alloc_mpwqe_rq_drop_page(struct mlx5e_rq *rq)
 {
 	rq->wqe_overflow.page = alloc_page(GFP_KERNEL);
@@ -830,6 +851,7 @@ static int mlx5e_alloc_rq(struct mlx5e_params *params,
 
 	rqp->wq.db_numa_node = node;
 	INIT_WORK(&rq->recover_work, mlx5e_rq_err_cqe_work);
+	INIT_WORK(&rq->rx_timeout_work, mlx5e_rq_timeout_work);
 
 	if (params->xdp_prog)
 		bpf_prog_inc(params->xdp_prog);
@@ -1204,7 +1226,8 @@ int mlx5e_wait_for_min_rx_wqes(struct mlx5e_rq *rq, int wait_time)
 	netdev_warn(rq->netdev, "Failed to get min RX wqes on Channel[%d] RQN[0x%x] wq cur_sz(%d) min_rx_wqes(%d)\n",
 		    rq->ix, rq->rqn, mlx5e_rqwq_get_cur_sz(rq), min_wqes);
 
-	mlx5e_reporter_rx_timeout(rq);
+	queue_work(rq->priv->wq, &rq->rx_timeout_work);
+
 	return -ETIMEDOUT;
 }
 
@@ -1375,6 +1398,7 @@ void mlx5e_close_rq(struct mlx5e_rq *rq)
 	if (rq->dim)
 		cancel_work_sync(&rq->dim->work);
 	cancel_work_sync(&rq->recover_work);
+	cancel_work_sync(&rq->rx_timeout_work);
 	mlx5e_destroy_rq(rq);
 	mlx5e_free_rx_descs(rq);
 	mlx5e_free_rq(rq);
Signature · 1 line
2.39.5
author maintainer
#289 lore raw

PATCH 279/627 macsec: set IFF_UNICAST_FLT priv flag

In reply to Greg Kroah-Hartman

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

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

From: Stanislav Fomichev <[email protected]>

[ Upstream commit 0349659fd72f662c054ff20d432559bfaa228ce4 ]

Cosmin reports the following locking issue:

# BUG: sleeping function called from invalid context at kernel/locking/mutex.c:275

  #   dump_stack_lvl+0x4f/0x60
  #   __might_resched+0xeb/0x140
  #   mutex_lock+0x1a/0x40
  #   dev_set_promiscuity+0x26/0x90
  #   __dev_set_promiscuity+0x85/0x170
  #   __dev_set_rx_mode+0x69/0xa0
  #   dev_uc_add+0x6d/0x80
  #   vlan_dev_open+0x5f/0x120 [8021q]
  #  __dev_open+0x10c/0x2a0
  #  __dev_change_flags+0x1a4/0x210
  #  netif_change_flags+0x22/0x60
  #  do_setlink.isra.0+0xdb0/0x10f0
  #  rtnl_newlink+0x797/0xb00
  #  rtnetlink_rcv_msg+0x1cb/0x3f0
  #  netlink_rcv_skb+0x53/0x100
  #  netlink_unicast+0x273/0x3b0
  #  netlink_sendmsg+0x1f2/0x430

Which is similar to recent syzkaller reports in [0] and [1] and triggers because macsec does not advertise IFF_UNICAST_FLT although it has proper ndo_set_rx_mode callback that takes care of pushing uc/mc addresses down to the real device.

In general, dev_uc_add call path is problematic for stacking non-IFF_UNICAST_FLT because we might grab netdev instance lock under addr_list_lock spinlock, so this is not a systemic fix.

0: https://lore.kernel.org/netdev/[email protected]
1: https://lore.kernel.org/netdev/[email protected]/

---
 drivers/net/macsec.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/macsec.c +1 -1
index 7edbe76b5455..4c75d1fea552 100644
@@ -3868,7 +3868,7 @@ static void macsec_setup(struct net_device *dev)
 	ether_setup(dev);
 	dev->min_mtu = 0;
 	dev->max_mtu = ETH_MAX_MTU;
-	dev->priv_flags |= IFF_NO_QUEUE;
+	dev->priv_flags |= IFF_NO_QUEUE | IFF_UNICAST_FLT;
 	dev->netdev_ops = &macsec_netdev_ops;
 	dev->needs_free_netdev = true;
 	dev->priv_destructor = macsec_free_netdev;
Signature · 1 line
2.39.5
author maintainer
#290 lore raw

PATCH 280/627 net: dsa: microchip: Fix wrong rx drop MIB counter for KSZ8863

In reply to Greg Kroah-Hartman

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

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

From: Tristram Ha <[email protected]>

[ Upstream commit 165a7f5db919ab68a45ae755cceb751e067273ef ]

When KSZ8863 support was first added to KSZ driver the RX drop MIB counter was somehow defined as 0x105. The TX drop MIB counter starts at 0x100 for port 1, 0x101 for port 2, and 0x102 for port 3, so the RX drop MIB counter should start at 0x103 for port 1, 0x104 for port 2, and 0x105 for port 3.

There are 5 ports for KSZ8895, so its RX drop MIB counter starts at 0x105.

Fixes 4b20a07e103f ("net: dsa: microchip: ksz8795: add support for ksz88xx chips")
---
 drivers/net/dsa/microchip/ksz8.c     | 3 +++
 drivers/net/dsa/microchip/ksz8_reg.h | 4 +++-
 2 files changed, 6 insertions(+), 1 deletion(-)
drivers/net/dsa/microchip/ksz8.c +3 -0
index be433b4e2b1c..8f55be89f8bf 100644
@@ -371,6 +371,9 @@ static void ksz8863_r_mib_pkt(struct ksz_device *dev, int port, u16 addr,
 	addr -= dev->info->reg_mib_cnt;
 	ctrl_addr = addr ? KSZ8863_MIB_PACKET_DROPPED_TX_0 :
 			   KSZ8863_MIB_PACKET_DROPPED_RX_0;
+	if (ksz_is_8895_family(dev) &&
+	    ctrl_addr == KSZ8863_MIB_PACKET_DROPPED_RX_0)
+		ctrl_addr = KSZ8895_MIB_PACKET_DROPPED_RX_0;
 	ctrl_addr += port;
 	ctrl_addr |= IND_ACC_TABLE(TABLE_MIB | TABLE_READ);
 
drivers/net/dsa/microchip/ksz8_reg.h +3 -1
index 329688603a58..da80e659c648 100644
@@ -784,7 +784,9 @@
 #define KSZ8795_MIB_TOTAL_TX_1		0x105
 
 #define KSZ8863_MIB_PACKET_DROPPED_TX_0 0x100
-#define KSZ8863_MIB_PACKET_DROPPED_RX_0 0x105
+#define KSZ8863_MIB_PACKET_DROPPED_RX_0 0x103
+
+#define KSZ8895_MIB_PACKET_DROPPED_RX_0 0x105
 
 #define MIB_PACKET_DROPPED		0x0000FFFF
 
Signature · 1 line
2.39.5
author maintainer
#291 lore raw

PATCH 281/627 neighbour: Fix null-ptr-deref in neigh_flush_dev().

In reply to Greg Kroah-Hartman

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

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

From: Kuniyuki Iwashima <[email protected]>

[ Upstream commit 1bbb76a899486827394530916f01214d049931b3 ]

kernel test robot reported null-ptr-deref in neigh_flush_dev(). [0]

The cited commit introduced per-netdev neighbour list and converted neigh_flush_dev() to use it instead of the global hash table.

One thing we missed is that neigh_table_clear() calls neigh_ifdown() with NULL dev.

Let's restore the hash table iteration.

Note that IPv6 module is no longer unloadable, so neigh_table_clear() is called only when IPv6 fails to initialise, which is unlikely to happen.

[0]:
IPv6: Attempt to unregister permanent protocol 136
IPv6: Attempt to unregister permanent protocol 17

Oops: general protection fault, probably for non-canonical address 0xdffffc00000001a0: 0000 [#1] SMP KASAN
KASAN: null-ptr-deref in range [0x0000000000000d00-0x0000000000000d07]
CPU: 1 UID: 0 PID: 1 Comm: systemd Tainted: G                T  6.12.0-rc6-01246-gf7f52738637f #1
Tainted: [T]=RANDSTRUCT
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.16.2-debian-1.16.2-1 04/01/2014
RIP: 0010:neigh_flush_dev.llvm.6395807810224103582+0x52/0x570
Code: c1 e8 03 42 8a 04 38 84 c0 0f 85 15 05 00 00 31 c0 41 83 3e 0a 0f 94 c0 48 8d 1c c3 48 81 c3 f8 0c 00 00 48 89 d8 48 c1 e8 03 <42> 80 3c 38 00 74 08 48 89 df e8 f7 49 93 fe 4c 8b 3b 4d 85 ff 0f
RSP: 0000:ffff88810026f408 EFLAGS: 00010206
RAX: 00000000000001a0 RBX: 0000000000000d00 RCX: 0000000000000000
RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffffffffc0631640
RBP: ffff88810026f470 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000000
R13: ffffffffc0625250 R14: ffffffffc0631640 R15: dffffc0000000000
FS:  00007f575cb83940(0000) GS:ffff8883aee00000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007f575db40008 CR3: 00000002bf936000 CR4: 00000000000406f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
 <TASK>
 __neigh_ifdown.llvm.6395807810224103582+0x44/0x390
 neigh_table_clear+0xb1/0x268
 ndisc_cleanup+0x21/0x38 [ipv6]
 init_module+0x2f5/0x468 [ipv6]
 do_one_initcall+0x1ba/0x628
 do_init_module+0x21a/0x530
 load_module+0x2550/0x2ea0
 __se_sys_finit_module+0x3d2/0x620
 __x64_sys_finit_module+0x76/0x88
 x64_sys_call+0x7ff/0xde8
 do_syscall_64+0xfb/0x1e8
 entry_SYSCALL_64_after_hwframe+0x67/0x6f
RIP: 0033:0x7f575d6f2719
Code: 08 89 e8 5b 5d c3 66 2e 0f 1f 84 00 00 00 00 00 90 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d b7 06 0d 00 f7 d8 64 89 01 48
RSP: 002b:00007fff82a2a268 EFLAGS: 00000246 ORIG_RAX: 0000000000000139
RAX: ffffffffffffffda RBX: 0000557827b45310 RCX: 00007f575d6f2719
RDX: 0000000000000000 RSI: 00007f575d584efd RDI: 0000000000000004
RBP: 00007f575d584efd R08: 0000000000000000 R09: 0000557827b47b00
R10: 0000000000000004 R11: 0000000000000246 R12: 0000000000020000
R13: 0000000000000000 R14: 0000557827b470e0 R15: 00007f575dbb4270
 </TASK>
Modules linked in: ipv6(+)
---
 net/core/neighbour.c | 88 ++++++++++++++++++++++++++++++--------------
 1 file changed, 61 insertions(+), 27 deletions(-)
net/core/neighbour.c +61 -27
index 49dce9a82295..a8dc72eda202 100644
@@ -368,6 +368,43 @@ static void pneigh_queue_purge(struct sk_buff_head *list, struct net *net,
 	}
 }
 
+static void neigh_flush_one(struct neighbour *n)
+{
+	hlist_del_rcu(&n->hash);
+	hlist_del_rcu(&n->dev_list);
+
+	write_lock(&n->lock);
+
+	neigh_del_timer(n);
+	neigh_mark_dead(n);
+
+	if (refcount_read(&n->refcnt) != 1) {
+		/* The most unpleasant situation.
+		 * We must destroy neighbour entry,
+		 * but someone still uses it.
+		 *
+		 * The destroy will be delayed until
+		 * the last user releases us, but
+		 * we must kill timers etc. and move
+		 * it to safe state.
+		 */
+		__skb_queue_purge(&n->arp_queue);
+		n->arp_queue_len_bytes = 0;
+		WRITE_ONCE(n->output, neigh_blackhole);
+
+		if (n->nud_state & NUD_VALID)
+			n->nud_state = NUD_NOARP;
+		else
+			n->nud_state = NUD_NONE;
+
+		neigh_dbg(2, "neigh %p is stray\n", n);
+	}
+
+	write_unlock(&n->lock);
+
+	neigh_cleanup_and_release(n);
+}
+
 static void neigh_flush_dev(struct neigh_table *tbl, struct net_device *dev,
 			    bool skip_perm)
 {
@@ -381,32 +418,24 @@ static void neigh_flush_dev(struct neigh_table *tbl, struct net_device *dev,
 		if (skip_perm && n->nud_state & NUD_PERMANENT)
 			continue;
 
-		hlist_del_rcu(&n->hash);
-		hlist_del_rcu(&n->dev_list);
-		write_lock(&n->lock);
-		neigh_del_timer(n);
-		neigh_mark_dead(n);
-		if (refcount_read(&n->refcnt) != 1) {
-			/* The most unpleasant situation.
-			 * We must destroy neighbour entry,
-			 * but someone still uses it.
-			 *
-			 * The destroy will be delayed until
-			 * the last user releases us, but
-			 * we must kill timers etc. and move
-			 * it to safe state.
-			 */
-			__skb_queue_purge(&n->arp_queue);
-			n->arp_queue_len_bytes = 0;
-			WRITE_ONCE(n->output, neigh_blackhole);
-			if (n->nud_state & NUD_VALID)
-				n->nud_state = NUD_NOARP;
-			else
-				n->nud_state = NUD_NONE;
-			neigh_dbg(2, "neigh %p is stray\n", n);
-		}
-		write_unlock(&n->lock);
-		neigh_cleanup_and_release(n);
+		neigh_flush_one(n);
+	}
+}
+
+static void neigh_flush_table(struct neigh_table *tbl)
+{
+	struct neigh_hash_table *nht;
+	int i;
+
+	nht = rcu_dereference_protected(tbl->nht,
+					lockdep_is_held(&tbl->lock));
+
+	for (i = 0; i < (1 << nht->hash_shift); i++) {
+		struct hlist_node *tmp;
+		struct neighbour *n;
+
+		neigh_for_each_in_bucket_safe(n, tmp, &nht->hash_heads[i])
+			neigh_flush_one(n);
 	}
 }
 
@@ -422,7 +451,12 @@ static int __neigh_ifdown(struct neigh_table *tbl, struct net_device *dev,
 			  bool skip_perm)
 {
 	write_lock_bh(&tbl->lock);
-	neigh_flush_dev(tbl, dev, skip_perm);
+	if (likely(dev)) {
+		neigh_flush_dev(tbl, dev, skip_perm);
+	} else {
+		DEBUG_NET_WARN_ON_ONCE(skip_perm);
+		neigh_flush_table(tbl);
+	}
 	pneigh_ifdown_and_unlock(tbl, dev);
 	pneigh_queue_purge(&tbl->proxy_queue, dev ? dev_net(dev) : NULL,
 			   tbl->family);
Signature · 1 line
2.39.5
author maintainer
#292 lore raw

PATCH 282/627 stmmac: xsk: fix negative overflow of budget in zerocopy mode

In reply to Greg Kroah-Hartman

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

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

From: Jason Xing <[email protected]>

[ Upstream commit 2764ab51d5f0e8c7d3b7043af426b1883e3bde1d ]

A negative overflow can happen when the budget number of descs are consumed. as long as the budget is decreased to zero, it will again go into while (budget-- > 0) statement and get decreased by one, so the overflow issue can happen. It will lead to returning true whereas the expected value should be false.

In this case where all the budget is used up, it means zc function should return false to let the poll run again because normally we might have more data to process. Without this patch, zc function would return true instead.

Fixes 132c32ee5bc0 ("net: stmmac: Add TX via XDP zero-copy socket")
---
 drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +1 -1
index b948df1bff9a..e0fb06af1f94 100644
@@ -2596,7 +2596,7 @@ static bool stmmac_xdp_xmit_zc(struct stmmac_priv *priv, u32 queue, u32 budget)
 
 	budget = min(budget, stmmac_tx_avail(priv, queue));
 
-	while (budget-- > 0) {
+	for (; budget > 0; budget--) {
 		struct stmmac_metadata_request meta_req;
 		struct xsk_tx_metadata *meta = NULL;
 		dma_addr_t dma_addr;
Signature · 1 line
2.39.5
author maintainer
#293 lore raw

PATCH 283/627 igb: xsk: solve negative overflow of nb_pkts in zerocopy mode

In reply to Greg Kroah-Hartman

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

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

From: Jason Xing <[email protected]>

[ Upstream commit 3b7c13dfdcc26a78756cc17a23cdf4310c5a24a9 ]

There is no break time in the while() loop, so every time at the end of igb_xmit_zc(), negative overflow of nb_pkts will occur, which renders the return value always false. But theoretically, the result should be set after calling xsk_tx_peek_release_desc_batch(). We can take i40e_xmit_zc() as a good example.

Returning false means we're not done with transmission and we need one more poll, which is exactly what igb_xmit_zc() always did before this patch. After this patch, the return value depends on the nb_pkts value. Two cases might happen then:
1. if (nb_pkts < budget), it means we process all the possible data, so return true and no more necessary poll will be triggered because of this.
2. if (nb_pkts == budget), it means we might have more data, so return false to let another poll run again.

---
 drivers/net/ethernet/intel/igb/igb_xsk.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
drivers/net/ethernet/intel/igb/igb_xsk.c +1 -2
index 5cf67ba29269..30ce5fbb5b77 100644
@@ -482,7 +482,7 @@ bool igb_xmit_zc(struct igb_ring *tx_ring, struct xsk_buff_pool *xsk_pool)
 	if (!nb_pkts)
 		return true;
 
-	while (nb_pkts-- > 0) {
+	for (; i < nb_pkts; i++) {
 		dma = xsk_buff_raw_get_dma(xsk_pool, descs[i].addr);
 		xsk_buff_raw_dma_sync_for_device(xsk_pool, dma, descs[i].len);
 
@@ -512,7 +512,6 @@ bool igb_xmit_zc(struct igb_ring *tx_ring, struct xsk_buff_pool *xsk_pool)
 
 		total_bytes += descs[i].len;
 
-		i++;
 		tx_ring->next_to_use++;
 		tx_buffer_info->next_to_watch = tx_desc;
 		if (tx_ring->next_to_use == tx_ring->count)
Signature · 1 line
2.39.5
author maintainer
#294 lore raw

PATCH 284/627 selftests: rtnetlink.sh: remove esp4_offload after test

In reply to Greg Kroah-Hartman

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

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

From: Xiumei Mu <[email protected]>

[ Upstream commit 5b32321fdaf3fd1a92ec726af18765e225b0ee2b ]

The esp4_offload module, loaded during IPsec offload tests, should be reset to its default settings after testing.
Otherwise, leaving it enabled could unintentionally affect subsequence test cases by keeping offload active.

Without this fix:

$ lsmod | grep offload; ./rtnetlink.sh -t kci_test_ipsec_offload ; lsmod | grep offload;
PASS: ipsec_offload
esp4_offload           12288  0
esp4                   32768  1 esp4_offload

With this fix:

$ lsmod | grep offload; ./rtnetlink.sh -t kci_test_ipsec_offload ; lsmod | grep offload;
PASS: ipsec_offload
---
 tools/testing/selftests/net/rtnetlink.sh | 6 ++++++
 1 file changed, 6 insertions(+)
tools/testing/selftests/net/rtnetlink.sh +6 -0
index 2e8243a65b50..d2298da320a6 100755
@@ -673,6 +673,11 @@ kci_test_ipsec_offload()
 	sysfsf=$sysfsd/ipsec
 	sysfsnet=/sys/bus/netdevsim/devices/netdevsim0/net/
 	probed=false
+	esp4_offload_probed_default=false
+
+	if lsmod | grep -q esp4_offload; then
+		esp4_offload_probed_default=true
+	fi
 
 	if ! mount | grep -q debugfs; then
 		mount -t debugfs none /sys/kernel/debug/ &> /dev/null
@@ -766,6 +771,7 @@ EOF
 	fi
 
 	# clean up any leftovers
+	! "$esp4_offload_probed_default" && lsmod | grep -q esp4_offload && rmmod esp4_offload
 	echo 0 > /sys/bus/netdevsim/del_device
 	$probed && rmmod netdevsim
 
Signature · 1 line
2.39.5
author maintainer
#295 lore raw

PATCH 285/627 vrf: Drop existing dst reference in vrf_ip6_input_dst

In reply to Greg Kroah-Hartman

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

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

From: Stanislav Fomichev <[email protected]>

[ Upstream commit f388f807eca1de9e6e70f9ffb1a573c3811c4215 ]

Commit ff3fbcdd4724 ("selftests: tc: Add generic erspan_opts matching support for tc-flower") started triggering the following kmemleak warning:

unreferenced object 0xffff888015fb0e00 (size 512):
comm "softirq", pid 0, jiffies 4294679065 hex dump (first 32 bytes):

    00 00 00 00 00 00 00 00 40 d2 85 9e ff ff ff ff  ........@.......
    41 69 59 9d ff ff ff ff 00 00 00 00 00 00 00 00  AiY.............
  backtrace (crc 30b71e8b):
    __kmalloc_noprof+0x359/0x460
    metadata_dst_alloc+0x28/0x490
    erspan_rcv+0x4f1/0x1160 [ip_gre]
    gre_rcv+0x217/0x240 [ip_gre]
    gre_rcv+0x1b8/0x400 [gre]
    ip_protocol_deliver_rcu+0x31d/0x3a0
    ip_local_deliver_finish+0x37d/0x620
    ip_local_deliver+0x174/0x460
    ip_rcv+0x52b/0x6b0
    __netif_receive_skb_one_core+0x149/0x1a0
    process_backlog+0x3c8/0x1390
    __napi_poll.constprop.0+0xa1/0x390
    net_rx_action+0x59b/0xe00
    handle_softirqs+0x22b/0x630
    do_softirq+0xb1/0xf0
    __local_bh_enable_ip+0x115/0x150

vrf_ip6_input_dst unconditionally sets skb dst entry, add a call to skb_dst_drop to drop any existing entry.

Cc David Ahern <[email protected]>
Fixes 9ff74384600a ("net: vrf: Handle ipv6 multicast and link-local addresses")
---
 drivers/net/vrf.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/net/vrf.c +2 -0
index 9a4beea6ee0c..3ccd649913b5 100644
@@ -1302,6 +1302,8 @@ static void vrf_ip6_input_dst(struct sk_buff *skb, struct net_device *vrf_dev,
 	struct net *net = dev_net(vrf_dev);
 	struct rt6_info *rt6;
 
+	skb_dst_drop(skb);
+
 	rt6 = vrf_ip6_route_lookup(net, vrf_dev, &fl6, ifindex, skb,
 				   RT6_LOOKUP_F_HAS_SADDR | RT6_LOOKUP_F_IFACE);
 	if (unlikely(!rt6))
Signature · 1 line
2.39.5
author maintainer
#296 lore raw

PATCH 286/627 ipv6: add a retry logic in net6_rt_notify()

In reply to Greg Kroah-Hartman

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

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

From: Eric Dumazet <[email protected]>

[ Upstream commit ea2f921db7a483a526058c5b5b8162edd88dabe5 ]

inet6_rt_notify() can be called under RCU protection only. This means the route could be changed concurrently and rt6_fill_node() could return -EMSGSIZE.

Re-size the skb when this happens and retry, removing one WARN_ON() that syzbot was able to trigger:

WARNING: CPU: 3 PID: 6291 at net/ipv6/route.c:6342 inet6_rt_notify+0x475/0x4b0 net/ipv6/route.c:6342
Modules linked in:
CPU: 3 UID: 0 PID: 6291 Comm: syz.0.77 Not tainted 6.16.0-rc7-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:inet6_rt_notify+0x475/0x4b0 net/ipv6/route.c:6342
Code: fc ff ff e8 6d 52 ea f7 e9 47 fc ff ff 48 8b 7c 24 08 4c 89 04 24 e8 5a 52 ea f7 4c 8b 04 24 e9 94 fd ff ff e8 9c fe 84 f7 90 <0f> 0b 90 e9 bd fd ff ff e8 6e 52 ea f7 e9 bb fb ff ff 48 89 df e8
RSP: 0018:ffffc900035cf1d8 EFLAGS: 00010293
RAX: 0000000000000000 RBX: ffffc900035cf540 RCX: ffffffff8a36e790
RDX: ffff88802f7e8000 RSI: ffffffff8a36e9d4 RDI: 0000000000000005
RBP: ffff88803c230f00 R08: 0000000000000005 R09: 00000000ffffffa6
R10: 00000000ffffffa6 R11: 0000000000000001 R12: 00000000ffffffa6
R13: 0000000000000900 R14: ffff888032ea4100 R15: 0000000000000000
FS:  00007fac7b89a6c0(0000) GS:ffff8880d6a20000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007fac7b899f98 CR3: 0000000034b3f000 CR4: 0000000000352ef0
Call Trace:
 <TASK>
  ip6_route_mpath_notify+0xde/0x280 net/ipv6/route.c:5356
  ip6_route_multipath_add+0x1181/0x1bd0 net/ipv6/route.c:5536
  inet6_rtm_newroute+0xe4/0x1a0 net/ipv6/route.c:5647
  rtnetlink_rcv_msg+0x95e/0xe90 net/core/rtnetlink.c:6944
  netlink_rcv_skb+0x155/0x420 net/netlink/af_netlink.c:2552
  netlink_unicast_kernel net/netlink/af_netlink.c:1320 [inline]
  netlink_unicast+0x58d/0x850 net/netlink/af_netlink.c:1346
  netlink_sendmsg+0x8d1/0xdd0 net/netlink/af_netlink.c:1896
  sock_sendmsg_nosec net/socket.c:712 [inline]
  __sock_sendmsg net/socket.c:727 [inline]
  ____sys_sendmsg+0xa95/0xc70 net/socket.c:2566
  ___sys_sendmsg+0x134/0x1d0 net/socket.c:2620
Fixes 169fd62799e8 ("ipv6: Get rid of RTNL for SIOCADDRT and RTM_NEWROUTE.")
---
 net/ipv6/route.c | 15 ++++++++++-----
 1 file changed, 10 insertions(+), 5 deletions(-)
net/ipv6/route.c +10 -5
index 7b9e49be7164..38016f5b2291 100644
@@ -6321,8 +6321,9 @@ static int inet6_rtm_getroute(struct sk_buff *in_skb, struct nlmsghdr *nlh,
 void inet6_rt_notify(int event, struct fib6_info *rt, struct nl_info *info,
 		     unsigned int nlm_flags)
 {
-	struct sk_buff *skb;
 	struct net *net = info->nl_net;
+	struct sk_buff *skb;
+	size_t sz;
 	u32 seq;
 	int err;
 
@@ -6330,17 +6331,21 @@ void inet6_rt_notify(int event, struct fib6_info *rt, struct nl_info *info,
 	seq = info->nlh ? info->nlh->nlmsg_seq : 0;
 
 	rcu_read_lock();
-
-	skb = nlmsg_new(rt6_nlmsg_size(rt), GFP_ATOMIC);
+	sz = rt6_nlmsg_size(rt);
+retry:
+	skb = nlmsg_new(sz, GFP_ATOMIC);
 	if (!skb)
 		goto errout;
 
 	err = rt6_fill_node(net, skb, rt, NULL, NULL, NULL, 0,
 			    event, info->portid, seq, nlm_flags);
 	if (err < 0) {
-		/* -EMSGSIZE implies BUG in rt6_nlmsg_size() */
-		WARN_ON(err == -EMSGSIZE);
 		kfree_skb(skb);
+		/* -EMSGSIZE implies needed space grew under us. */
+		if (err == -EMSGSIZE) {
+			sz = max(rt6_nlmsg_size(rt), sz << 1);
+			goto retry;
+		}
 		goto errout;
 	}
 
Signature · 1 line
2.39.5
author maintainer
#297 lore raw

PATCH 287/627 ipv6: prevent infinite loop in rt6_nlmsg_size()

In reply to Greg Kroah-Hartman

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

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

From: Eric Dumazet <[email protected]>

[ Upstream commit 54e6fe9dd3b0e7c481c2228782c9494d653546da ]

While testing prior patch, I was able to trigger an infinite loop in rt6_nlmsg_size() in the following place:

list_for_each_entry_rcu(sibling, &f6i->fib6_siblings,

			fib6_siblings) {
	rt6_nh_nlmsg_size(sibling->fib6_nh, &nexthop_len);
}

This is because fib6_del_route() and fib6_add_rt2node() uses list_del_rcu(), which can confuse rcu readers, because they might no longer see the head of the list.

Restart the loop if f6i->fib6_nsiblings is zero.

Fixes d9ccb18f83ea ("ipv6: Fix soft lockups in fib6_select_path under high next hop churn")
---
 net/ipv6/ip6_fib.c |  4 ++--
 net/ipv6/route.c   | 34 ++++++++++++++++++----------------
 2 files changed, 20 insertions(+), 18 deletions(-)
net/ipv6/ip6_fib.c +2 -2
index 93578b2ec35f..af7db69d9eac 100644
@@ -1265,7 +1265,7 @@ static int fib6_add_rt2node(struct fib6_node *fn, struct fib6_info *rt,
 							 &rt->fib6_siblings,
 							 fib6_siblings)
 					sibling->fib6_nsiblings--;
-				rt->fib6_nsiblings = 0;
+				WRITE_ONCE(rt->fib6_nsiblings, 0);
 				list_del_rcu(&rt->fib6_siblings);
 				rcu_read_lock();
 				rt6_multipath_rebalance(next_sibling);
@@ -2015,7 +2015,7 @@ static void fib6_del_route(struct fib6_table *table, struct fib6_node *fn,
 		list_for_each_entry_safe(sibling, next_sibling,
 					 &rt->fib6_siblings, fib6_siblings)
 			sibling->fib6_nsiblings--;
-		rt->fib6_nsiblings = 0;
+		WRITE_ONCE(rt->fib6_nsiblings, 0);
 		list_del_rcu(&rt->fib6_siblings);
 		rt6_multipath_rebalance(next_sibling);
 	}
net/ipv6/route.c +18 -16
index 38016f5b2291..2aeca451aab3 100644
@@ -5670,32 +5670,34 @@ static int rt6_nh_nlmsg_size(struct fib6_nh *nh, void *arg)
 
 static size_t rt6_nlmsg_size(struct fib6_info *f6i)
 {
+	struct fib6_info *sibling;
+	struct fib6_nh *nh;
 	int nexthop_len;
 
 	if (f6i->nh) {
 		nexthop_len = nla_total_size(4); /* RTA_NH_ID */
 		nexthop_for_each_fib6_nh(f6i->nh, rt6_nh_nlmsg_size,
 					 &nexthop_len);
-	} else {
-		struct fib6_nh *nh = f6i->fib6_nh;
-		struct fib6_info *sibling;
-
-		nexthop_len = 0;
-		if (f6i->fib6_nsiblings) {
-			rt6_nh_nlmsg_size(nh, &nexthop_len);
-
-			rcu_read_lock();
+		goto common;
+	}
 
-			list_for_each_entry_rcu(sibling, &f6i->fib6_siblings,
-						fib6_siblings) {
-				rt6_nh_nlmsg_size(sibling->fib6_nh, &nexthop_len);
-			}
+	rcu_read_lock();
+retry:
+	nh = f6i->fib6_nh;
+	nexthop_len = 0;
+	if (READ_ONCE(f6i->fib6_nsiblings)) {
+		rt6_nh_nlmsg_size(nh, &nexthop_len);
 
-			rcu_read_unlock();
+		list_for_each_entry_rcu(sibling, &f6i->fib6_siblings,
+					fib6_siblings) {
+			rt6_nh_nlmsg_size(sibling->fib6_nh, &nexthop_len);
+			if (!READ_ONCE(f6i->fib6_nsiblings))
+				goto retry;
 		}
-		nexthop_len += lwtunnel_get_encap_size(nh->fib_nh_lws);
 	}
-
+	rcu_read_unlock();
+	nexthop_len += lwtunnel_get_encap_size(nh->fib_nh_lws);
+common:
 	return NLMSG_ALIGN(sizeof(struct rtmsg))
 	       + nla_total_size(16) /* RTA_SRC */
 	       + nla_total_size(16) /* RTA_DST */
Signature · 1 line
2.39.5
author maintainer
#298 lore raw

PATCH 288/627 ipv6: fix possible infinite loop in fib6_info_uses_dev()

In reply to Greg Kroah-Hartman

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

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

From: Eric Dumazet <[email protected]>

[ Upstream commit f8d8ce1b515a0a6af72b30502670a406cfb75073 ]

fib6_info_uses_dev() seems to rely on RCU without an explicit protection.

Like the prior fix in rt6_nlmsg_size(),
we need to make sure fib6_del_route() or fib6_add_rt2node() have not removed the anchor from the list, or we risk an infinite loop.

Fixes d9ccb18f83ea ("ipv6: Fix soft lockups in fib6_select_path under high next hop churn")
---
 net/ipv6/route.c | 17 +++++++++++------
 1 file changed, 11 insertions(+), 6 deletions(-)
net/ipv6/route.c +11 -6
index 2aeca451aab3..bf71c8ce002f 100644
@@ -5958,16 +5958,21 @@ static bool fib6_info_uses_dev(const struct fib6_info *f6i,
 	if (f6i->fib6_nh->fib_nh_dev == dev)
 		return true;
 
-	if (f6i->fib6_nsiblings) {
-		struct fib6_info *sibling, *next_sibling;
+	if (READ_ONCE(f6i->fib6_nsiblings)) {
+		const struct fib6_info *sibling;
 
-		list_for_each_entry_safe(sibling, next_sibling,
-					 &f6i->fib6_siblings, fib6_siblings) {
-			if (sibling->fib6_nh->fib_nh_dev == dev)
+		rcu_read_lock();
+		list_for_each_entry_rcu(sibling, &f6i->fib6_siblings,
+					fib6_siblings) {
+			if (sibling->fib6_nh->fib_nh_dev == dev) {
+				rcu_read_unlock();
 				return true;
+			}
+			if (!READ_ONCE(f6i->fib6_nsiblings))
+				break;
 		}
+		rcu_read_unlock();
 	}
-
 	return false;
 }
 
Signature · 1 line
2.39.5
author maintainer
#299 lore raw

PATCH 289/627 ipv6: annotate data-races around rt->fib6_nsiblings

In reply to Greg Kroah-Hartman

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

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

From: Eric Dumazet <[email protected]>

[ Upstream commit 31d7d67ba1274f42494256d52e86da80ed09f3cb ]

rt->fib6_nsiblings can be read locklessly, add corresponding READ_ONCE() and WRITE_ONCE() annotations.

Fixes 66f5d6ce53e6 ("ipv6: replace rwlock with rcu and spinlock in fib6_table")
---
 net/ipv6/ip6_fib.c | 20 +++++++++++++-------
 net/ipv6/route.c   |  5 +++--
 2 files changed, 16 insertions(+), 9 deletions(-)
net/ipv6/ip6_fib.c +13 -7
index af7db69d9eac..4d68bd853dba 100644
@@ -445,15 +445,17 @@ struct fib6_dump_arg {
 static int fib6_rt_dump(struct fib6_info *rt, struct fib6_dump_arg *arg)
 {
 	enum fib_event_type fib_event = FIB_EVENT_ENTRY_REPLACE;
+	unsigned int nsiblings;
 	int err;
 
 	if (!rt || rt == arg->net->ipv6.fib6_null_entry)
 		return 0;
 
-	if (rt->fib6_nsiblings)
+	nsiblings = READ_ONCE(rt->fib6_nsiblings);
+	if (nsiblings)
 		err = call_fib6_multipath_entry_notifier(arg->nb, fib_event,
 							 rt,
-							 rt->fib6_nsiblings,
+							 nsiblings,
 							 arg->extack);
 	else
 		err = call_fib6_entry_notifier(arg->nb, fib_event, rt,
@@ -1138,7 +1140,7 @@ static int fib6_add_rt2node(struct fib6_node *fn, struct fib6_info *rt,
 
 			if (rt6_duplicate_nexthop(iter, rt)) {
 				if (rt->fib6_nsiblings)
-					rt->fib6_nsiblings = 0;
+					WRITE_ONCE(rt->fib6_nsiblings, 0);
 				if (!(iter->fib6_flags & RTF_EXPIRES))
 					return -EEXIST;
 				if (!(rt->fib6_flags & RTF_EXPIRES)) {
@@ -1167,7 +1169,8 @@ static int fib6_add_rt2node(struct fib6_node *fn, struct fib6_info *rt,
 			 */
 			if (rt_can_ecmp &&
 			    rt6_qualify_for_ecmp(iter))
-				rt->fib6_nsiblings++;
+				WRITE_ONCE(rt->fib6_nsiblings,
+					   rt->fib6_nsiblings + 1);
 		}
 
 		if (iter->fib6_metric > rt->fib6_metric)
@@ -1217,7 +1220,8 @@ static int fib6_add_rt2node(struct fib6_node *fn, struct fib6_info *rt,
 		fib6_nsiblings = 0;
 		list_for_each_entry_safe(sibling, temp_sibling,
 					 &rt->fib6_siblings, fib6_siblings) {
-			sibling->fib6_nsiblings++;
+			WRITE_ONCE(sibling->fib6_nsiblings,
+				   sibling->fib6_nsiblings + 1);
 			BUG_ON(sibling->fib6_nsiblings != rt->fib6_nsiblings);
 			fib6_nsiblings++;
 		}
@@ -1264,7 +1268,8 @@ static int fib6_add_rt2node(struct fib6_node *fn, struct fib6_info *rt,
 				list_for_each_entry_safe(sibling, next_sibling,
 							 &rt->fib6_siblings,
 							 fib6_siblings)
-					sibling->fib6_nsiblings--;
+					WRITE_ONCE(sibling->fib6_nsiblings,
+						   sibling->fib6_nsiblings - 1);
 				WRITE_ONCE(rt->fib6_nsiblings, 0);
 				list_del_rcu(&rt->fib6_siblings);
 				rcu_read_lock();
@@ -2014,7 +2019,8 @@ static void fib6_del_route(struct fib6_table *table, struct fib6_node *fn,
 			notify_del = true;
 		list_for_each_entry_safe(sibling, next_sibling,
 					 &rt->fib6_siblings, fib6_siblings)
-			sibling->fib6_nsiblings--;
+			WRITE_ONCE(sibling->fib6_nsiblings,
+				   sibling->fib6_nsiblings - 1);
 		WRITE_ONCE(rt->fib6_nsiblings, 0);
 		list_del_rcu(&rt->fib6_siblings);
 		rt6_multipath_rebalance(next_sibling);
net/ipv6/route.c +3 -2
index bf71c8ce002f..8adae86fbe72 100644
@@ -5346,7 +5346,8 @@ static void ip6_route_mpath_notify(struct fib6_info *rt,
 	 */
 	rcu_read_lock();
 
-	if ((nlflags & NLM_F_APPEND) && rt_last && rt_last->fib6_nsiblings) {
+	if ((nlflags & NLM_F_APPEND) && rt_last &&
+	    READ_ONCE(rt_last->fib6_nsiblings)) {
 		rt = list_first_or_null_rcu(&rt_last->fib6_siblings,
 					    struct fib6_info,
 					    fib6_siblings);
@@ -5856,7 +5857,7 @@ static int rt6_fill_node(struct net *net, struct sk_buff *skb,
 		if (dst->lwtstate &&
 		    lwtunnel_fill_encap(skb, dst->lwtstate, RTA_ENCAP, RTA_ENCAP_TYPE) < 0)
 			goto nla_put_failure;
-	} else if (rt->fib6_nsiblings) {
+	} else if (READ_ONCE(rt->fib6_nsiblings)) {
 		struct fib6_info *sibling;
 		struct nlattr *mp;
 
Signature · 1 line
2.39.5
author maintainer
#300 lore raw

PATCH 290/627 bpf/preload: Dont select USERMODE_DRIVER

In reply to Greg Kroah-Hartman

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

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

From: Thomas Weißschuh <[email protected]>

[ Upstream commit 2b03164eee20eac7ce0fe3aa4fbda7efc1e5427a ]

The usermode driver framework is not used anymore by the BPF preload code.

Fixes cb80ddc67152 ("bpf: Convert bpf_preload.ko to use light skeleton.")
---
 kernel/bpf/preload/Kconfig | 1 -
 1 file changed, 1 deletion(-)
kernel/bpf/preload/Kconfig +0 -1
index c9d45c9d6918..f9b11d01c3b5 100644
@@ -10,7 +10,6 @@ menuconfig BPF_PRELOAD
 	# The dependency on !COMPILE_TEST prevents it from being enabled
 	# in allmodconfig or allyesconfig configurations
 	depends on !COMPILE_TEST
-	select USERMODE_DRIVER
 	help
 	  This builds kernel module with several embedded BPF programs that are
 	  pinned into BPF FS mount point as human readable files that are
Signature · 1 line
2.39.5
author maintainer
#301 lore raw

PATCH 291/627 bpf, arm64: Fix fp initialization for exception boundary

In reply to Greg Kroah-Hartman

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

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

From: Puranjay Mohan <[email protected]>

[ Upstream commit b114fcee766d5101eada1aca7bb5fd0a86c89b35 ]

In the ARM64 BPF JIT when prog->aux->exception_boundary is set for a BPF program, find_used_callee_regs() is not called because for a program acting as exception boundary, all callee saved registers are saved. find_used_callee_regs() sets ctx->fp_used = true; when it sees FP being used in any of the instructions.

For programs acting as exception boundary, ctx->fp_used remains false even if frame pointer is used by the program and therefore, FP is not set-up for such programs in the prologue. This can cause the kernel to crash due to a pagefault.

Fix it by setting ctx->fp_used = true for exception boundary programs as fp is always saved in such programs.

Fixes 5d4fa9ec5643 ("bpf, arm64: Avoid blindly saving/restoring all callee-saved registers")
---
 arch/arm64/net/bpf_jit_comp.c | 1 +
 1 file changed, 1 insertion(+)
arch/arm64/net/bpf_jit_comp.c +1 -0
index da8b89dd2910..58f838b310bc 100644
@@ -412,6 +412,7 @@ static void push_callee_regs(struct jit_ctx *ctx)
 		emit(A64_PUSH(A64_R(23), A64_R(24), A64_SP), ctx);
 		emit(A64_PUSH(A64_R(25), A64_R(26), A64_SP), ctx);
 		emit(A64_PUSH(A64_R(27), A64_R(28), A64_SP), ctx);
+		ctx->fp_used = true;
 	} else {
 		find_used_callee_regs(ctx);
 		for (i = 0; i + 1 < ctx->nr_used_callee_reg; i += 2) {
Signature · 1 line
2.39.5
author maintainer
#302 lore raw

PATCH 292/627 RISC-V: KVM: Fix inclusion of Smnpm in the guest ISA bitmap

In reply to Greg Kroah-Hartman

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

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

From: Samuel Holland <[email protected]>

[ Upstream commit 7826c8f37220daabf90c09fcd9a835d6763f1372 ]

The Smnpm extension requires special handling because the guest ISA extension maps to a different extension (Ssnpm) on the host side. commit 1851e7836212 ("RISC-V: KVM: Allow Smnpm and Ssnpm extensions for guests") missed that the vcpu->arch.isa bit is based only on the host extension, so currently both KVM_RISCV_ISA_EXT_{SMNPM,SSNPM} map to vcpu->arch.isa[RISCV_ISA_EXT_SSNPM]. This does not cause any problems for the guest, because both extensions are force-enabled anyway when the host supports Ssnpm, but prevents checking for (guest) Smnpm in the SBI FWFT logic.

Redefine kvm_isa_ext_arr to look up the guest extension, since only the guest -> host mapping is unambiguous. Factor out the logic for checking for host support of an extension, so this special case only needs to be handled in one place, and be explicit about which variables hold a host vs a guest ISA extension.

Fixes 1851e7836212 ("RISC-V: KVM: Allow Smnpm and Ssnpm extensions for guests")
---
 arch/riscv/kvm/vcpu_onereg.c | 83 +++++++++++++++++++++++-------------
 1 file changed, 53 insertions(+), 30 deletions(-)
arch/riscv/kvm/vcpu_onereg.c +53 -30
index 2e1b646f0d61..cce6a38ea54f 100644
@@ -23,7 +23,7 @@
 #define KVM_ISA_EXT_ARR(ext)		\
 [KVM_RISCV_ISA_EXT_##ext] = RISCV_ISA_EXT_##ext
 
-/* Mapping between KVM ISA Extension ID & Host ISA extension ID */
+/* Mapping between KVM ISA Extension ID & guest ISA extension ID */
 static const unsigned long kvm_isa_ext_arr[] = {
 	/* Single letter extensions (alphabetically sorted) */
 	[KVM_RISCV_ISA_EXT_A] = RISCV_ISA_EXT_a,
@@ -35,7 +35,7 @@ static const unsigned long kvm_isa_ext_arr[] = {
 	[KVM_RISCV_ISA_EXT_M] = RISCV_ISA_EXT_m,
 	[KVM_RISCV_ISA_EXT_V] = RISCV_ISA_EXT_v,
 	/* Multi letter extensions (alphabetically sorted) */
-	[KVM_RISCV_ISA_EXT_SMNPM] = RISCV_ISA_EXT_SSNPM,
+	KVM_ISA_EXT_ARR(SMNPM),
 	KVM_ISA_EXT_ARR(SMSTATEEN),
 	KVM_ISA_EXT_ARR(SSAIA),
 	KVM_ISA_EXT_ARR(SSCOFPMF),
@@ -112,6 +112,36 @@ static unsigned long kvm_riscv_vcpu_base2isa_ext(unsigned long base_ext)
 	return KVM_RISCV_ISA_EXT_MAX;
 }
 
+static int kvm_riscv_vcpu_isa_check_host(unsigned long kvm_ext, unsigned long *guest_ext)
+{
+	unsigned long host_ext;
+
+	if (kvm_ext >= KVM_RISCV_ISA_EXT_MAX ||
+	    kvm_ext >= ARRAY_SIZE(kvm_isa_ext_arr))
+		return -ENOENT;
+
+	*guest_ext = kvm_isa_ext_arr[kvm_ext];
+	switch (*guest_ext) {
+	case RISCV_ISA_EXT_SMNPM:
+		/*
+		 * Pointer masking effective in (H)S-mode is provided by the
+		 * Smnpm extension, so that extension is reported to the guest,
+		 * even though the CSR bits for configuring VS-mode pointer
+		 * masking on the host side are part of the Ssnpm extension.
+		 */
+		host_ext = RISCV_ISA_EXT_SSNPM;
+		break;
+	default:
+		host_ext = *guest_ext;
+		break;
+	}
+
+	if (!__riscv_isa_extension_available(NULL, host_ext))
+		return -ENOENT;
+
+	return 0;
+}
+
 static bool kvm_riscv_vcpu_isa_enable_allowed(unsigned long ext)
 {
 	switch (ext) {
@@ -219,13 +249,13 @@ static bool kvm_riscv_vcpu_isa_disable_allowed(unsigned long ext)
 
 void kvm_riscv_vcpu_setup_isa(struct kvm_vcpu *vcpu)
 {
-	unsigned long host_isa, i;
+	unsigned long guest_ext, i;
 
 	for (i = 0; i < ARRAY_SIZE(kvm_isa_ext_arr); i++) {
-		host_isa = kvm_isa_ext_arr[i];
-		if (__riscv_isa_extension_available(NULL, host_isa) &&
-		    kvm_riscv_vcpu_isa_enable_allowed(i))
-			set_bit(host_isa, vcpu->arch.isa);
+		if (kvm_riscv_vcpu_isa_check_host(i, &guest_ext))
+			continue;
+		if (kvm_riscv_vcpu_isa_enable_allowed(i))
+			set_bit(guest_ext, vcpu->arch.isa);
 	}
 }
 
@@ -607,18 +637,15 @@ static int riscv_vcpu_get_isa_ext_single(struct kvm_vcpu *vcpu,
 					 unsigned long reg_num,
 					 unsigned long *reg_val)
 {
-	unsigned long host_isa_ext;
-
-	if (reg_num >= KVM_RISCV_ISA_EXT_MAX ||
-	    reg_num >= ARRAY_SIZE(kvm_isa_ext_arr))
-		return -ENOENT;
+	unsigned long guest_ext;
+	int ret;
 
-	host_isa_ext = kvm_isa_ext_arr[reg_num];
-	if (!__riscv_isa_extension_available(NULL, host_isa_ext))
-		return -ENOENT;
+	ret = kvm_riscv_vcpu_isa_check_host(reg_num, &guest_ext);
+	if (ret)
+		return ret;
 
 	*reg_val = 0;
-	if (__riscv_isa_extension_available(vcpu->arch.isa, host_isa_ext))
+	if (__riscv_isa_extension_available(vcpu->arch.isa, guest_ext))
 		*reg_val = 1; /* Mark the given extension as available */
 
 	return 0;
@@ -628,17 +655,14 @@ static int riscv_vcpu_set_isa_ext_single(struct kvm_vcpu *vcpu,
 					 unsigned long reg_num,
 					 unsigned long reg_val)
 {
-	unsigned long host_isa_ext;
-
-	if (reg_num >= KVM_RISCV_ISA_EXT_MAX ||
-	    reg_num >= ARRAY_SIZE(kvm_isa_ext_arr))
-		return -ENOENT;
+	unsigned long guest_ext;
+	int ret;
 
-	host_isa_ext = kvm_isa_ext_arr[reg_num];
-	if (!__riscv_isa_extension_available(NULL, host_isa_ext))
-		return -ENOENT;
+	ret = kvm_riscv_vcpu_isa_check_host(reg_num, &guest_ext);
+	if (ret)
+		return ret;
 
-	if (reg_val == test_bit(host_isa_ext, vcpu->arch.isa))
+	if (reg_val == test_bit(guest_ext, vcpu->arch.isa))
 		return 0;
 
 	if (!vcpu->arch.ran_atleast_once) {
@@ -648,10 +672,10 @@ static int riscv_vcpu_set_isa_ext_single(struct kvm_vcpu *vcpu,
 		 */
 		if (reg_val == 1 &&
 		    kvm_riscv_vcpu_isa_enable_allowed(reg_num))
-			set_bit(host_isa_ext, vcpu->arch.isa);
+			set_bit(guest_ext, vcpu->arch.isa);
 		else if (!reg_val &&
 			 kvm_riscv_vcpu_isa_disable_allowed(reg_num))
-			clear_bit(host_isa_ext, vcpu->arch.isa);
+			clear_bit(guest_ext, vcpu->arch.isa);
 		else
 			return -EINVAL;
 		kvm_riscv_vcpu_fp_reset(vcpu);
@@ -1009,16 +1033,15 @@ static int copy_fp_d_reg_indices(const struct kvm_vcpu *vcpu,
 static int copy_isa_ext_reg_indices(const struct kvm_vcpu *vcpu,
 				u64 __user *uindices)
 {
+	unsigned long guest_ext;
 	unsigned int n = 0;
-	unsigned long isa_ext;
 
 	for (int i = 0; i < KVM_RISCV_ISA_EXT_MAX; i++) {
 		u64 size = IS_ENABLED(CONFIG_32BIT) ?
 			   KVM_REG_SIZE_U32 : KVM_REG_SIZE_U64;
 		u64 reg = KVM_REG_RISCV | size | KVM_REG_RISCV_ISA_EXT | i;
 
-		isa_ext = kvm_isa_ext_arr[i];
-		if (!__riscv_isa_extension_available(NULL, isa_ext))
+		if (kvm_riscv_vcpu_isa_check_host(i, &guest_ext))
 			continue;
 
 		if (uindices) {
Signature · 1 line
2.39.5
author maintainer
#303 lore raw

PATCH 293/627 Remove trailing whitespace from tracepoint string

In reply to Greg Kroah-Hartman

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

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

From: Gabriele Monaco <[email protected]>

[ Upstream commit 7b70ac4cad2b20eaf415276bbaa0d9df9abb428c ]

RV event tracepoints print a line with the format:

    "event_xyz: S0 x event -> S1 "
    "event_xyz: S1 x event -> S0 (final)"

While printing an event leading to a non-final state, the line has a trailing white space (visible above before the closing ").

Adapt the format string not to print the trailing whitespace if we are not printing "(final)".

Stable-dep-of: 7f904ff6e58d ("rv: Use strings in da monitors tracepoints")

---
 kernel/trace/rv/rv_trace.h | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
kernel/trace/rv/rv_trace.h +4 -4
index 422b75f58891..18fa0e358a30 100644
@@ -29,11 +29,11 @@ DECLARE_EVENT_CLASS(event_da_monitor,
 		__entry->final_state		= final_state;
 	),
 
-	TP_printk("%s x %s -> %s %s",
+	TP_printk("%s x %s -> %s%s",
 		__entry->state,
 		__entry->event,
 		__entry->next_state,
-		__entry->final_state ? "(final)" : "")
+		__entry->final_state ? " (final)" : "")
 );
 
 DECLARE_EVENT_CLASS(error_da_monitor,
@@ -90,12 +90,12 @@ DECLARE_EVENT_CLASS(event_da_monitor_id,
 		__entry->final_state		= final_state;
 	),
 
-	TP_printk("%d: %s x %s -> %s %s",
+	TP_printk("%d: %s x %s -> %s%s",
 		__entry->id,
 		__entry->state,
 		__entry->event,
 		__entry->next_state,
-		__entry->final_state ? "(final)" : "")
+		__entry->final_state ? " (final)" : "")
 );
 
 DECLARE_EVENT_CLASS(error_da_monitor_id,
Signature · 1 line
2.39.5
author maintainer
#304 lore raw

PATCH 294/627 Use strings in da monitors tracepoints

In reply to Greg Kroah-Hartman

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

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

From: Gabriele Monaco <[email protected]>

[ Upstream commit 7f904ff6e58d398c4336f3c19c42b338324451f7 ]

Using DA monitors tracepoints with KASAN enabled triggers the following warning:

 BUG: KASAN: global-out-of-bounds in do_trace_event_raw_event_event_da_monitor+0xd6/0x1a0
 Read of size 32 at addr ffffffffaada8980 by task ...
 Call Trace:
  <TASK>
 [...]
  do_trace_event_raw_event_event_da_monitor+0xd6/0x1a0
  ? __pfx_do_trace_event_raw_event_event_da_monitor+0x10/0x10
  ? trace_event_sncid+0x83/0x200
  trace_event_sncid+0x163/0x200
 [...]
 The buggy address belongs to the variable:
  automaton_snep+0x4e0/0x5e0

This is caused by the tracepoints reading 32 bytes __array instead of __string from the automata definition. Such strings are literals and reading 32 bytes ends up in out of bound memory accesses (e.g. the next automaton's data in this case).
The error is harmless as, while printing the string, we stop at the null terminator, but it should still be fixed.

Use the __string facilities while defining the tracepoints to avoid reading out of bound memory.

Cc Masami Hiramatsu <[email protected]>
Cc Ingo Molnar <[email protected]>
Cc Peter Zijlstra <[email protected]>
Cc Tomas Glozar <[email protected]>
Cc Juri Lelli <[email protected]>
Cc Clark Williams <[email protected]>
Cc John Kacur <[email protected]>
Fixes 792575348ff7 ("rv/include: Add deterministic automata monitor definition via C macros")
---
 kernel/trace/rv/rv_trace.h | 76 +++++++++++++++++++-------------------
 1 file changed, 38 insertions(+), 38 deletions(-)
kernel/trace/rv/rv_trace.h +38 -38
index 18fa0e358a30..01fa84824bcb 100644
@@ -16,23 +16,23 @@ DECLARE_EVENT_CLASS(event_da_monitor,
 	TP_ARGS(state, event, next_state, final_state),
 
 	TP_STRUCT__entry(
-		__array(	char,	state,		MAX_DA_NAME_LEN	)
-		__array(	char,	event,		MAX_DA_NAME_LEN	)
-		__array(	char,	next_state,	MAX_DA_NAME_LEN	)
-		__field(	bool,	final_state			)
+		__string(	state,		state		)
+		__string(	event,		event		)
+		__string(	next_state,	next_state	)
+		__field(	bool,		final_state	)
 	),
 
 	TP_fast_assign(
-		memcpy(__entry->state,		state,		MAX_DA_NAME_LEN);
-		memcpy(__entry->event,		event,		MAX_DA_NAME_LEN);
-		memcpy(__entry->next_state,	next_state,	MAX_DA_NAME_LEN);
-		__entry->final_state		= final_state;
+		__assign_str(state);
+		__assign_str(event);
+		__assign_str(next_state);
+		__entry->final_state = final_state;
 	),
 
 	TP_printk("%s x %s -> %s%s",
-		__entry->state,
-		__entry->event,
-		__entry->next_state,
+		__get_str(state),
+		__get_str(event),
+		__get_str(next_state),
 		__entry->final_state ? " (final)" : "")
 );
 
@@ -43,18 +43,18 @@ DECLARE_EVENT_CLASS(error_da_monitor,
 	TP_ARGS(state, event),
 
 	TP_STRUCT__entry(
-		__array(	char,	state,		MAX_DA_NAME_LEN	)
-		__array(	char,	event,		MAX_DA_NAME_LEN	)
+		__string(	state,	state	)
+		__string(	event,	event	)
 	),
 
 	TP_fast_assign(
-		memcpy(__entry->state,		state,		MAX_DA_NAME_LEN);
-		memcpy(__entry->event,		event,		MAX_DA_NAME_LEN);
+		__assign_str(state);
+		__assign_str(event);
 	),
 
 	TP_printk("event %s not expected in the state %s",
-		__entry->event,
-		__entry->state)
+		__get_str(event),
+		__get_str(state))
 );
 
 #include <monitors/wip/wip_trace.h>
@@ -75,26 +75,26 @@ DECLARE_EVENT_CLASS(event_da_monitor_id,
 	TP_ARGS(id, state, event, next_state, final_state),
 
 	TP_STRUCT__entry(
-		__field(	int,	id				)
-		__array(	char,	state,		MAX_DA_NAME_LEN	)
-		__array(	char,	event,		MAX_DA_NAME_LEN	)
-		__array(	char,	next_state,	MAX_DA_NAME_LEN	)
-		__field(	bool,	final_state			)
+		__field(	int,		id		)
+		__string(	state,		state		)
+		__string(	event,		event		)
+		__string(	next_state,	next_state	)
+		__field(	bool,		final_state	)
 	),
 
 	TP_fast_assign(
-		memcpy(__entry->state,		state,		MAX_DA_NAME_LEN);
-		memcpy(__entry->event,		event,		MAX_DA_NAME_LEN);
-		memcpy(__entry->next_state,	next_state,	MAX_DA_NAME_LEN);
-		__entry->id			= id;
-		__entry->final_state		= final_state;
+		__assign_str(state);
+		__assign_str(event);
+		__assign_str(next_state);
+		__entry->id		= id;
+		__entry->final_state	= final_state;
 	),
 
 	TP_printk("%d: %s x %s -> %s%s",
 		__entry->id,
-		__entry->state,
-		__entry->event,
-		__entry->next_state,
+		__get_str(state),
+		__get_str(event),
+		__get_str(next_state),
 		__entry->final_state ? " (final)" : "")
 );
 
@@ -105,21 +105,21 @@ DECLARE_EVENT_CLASS(error_da_monitor_id,
 	TP_ARGS(id, state, event),
 
 	TP_STRUCT__entry(
-		__field(	int,	id				)
-		__array(	char,	state,		MAX_DA_NAME_LEN	)
-		__array(	char,	event,		MAX_DA_NAME_LEN	)
+		__field(	int,	id	)
+		__string(	state,	state	)
+		__string(	event,	event	)
 	),
 
 	TP_fast_assign(
-		memcpy(__entry->state,		state,		MAX_DA_NAME_LEN);
-		memcpy(__entry->event,		event,		MAX_DA_NAME_LEN);
-		__entry->id			= id;
+		__assign_str(state);
+		__assign_str(event);
+		__entry->id	= id;
 	),
 
 	TP_printk("%d: event %s not expected in the state %s",
 		__entry->id,
-		__entry->event,
-		__entry->state)
+		__get_str(event),
+		__get_str(state))
 );
 
 #include <monitors/wwnr/wwnr_trace.h>
Signature · 1 line
2.39.5
author maintainer
#305 lore raw

PATCH 295/627 Adjust monitor dependencies

In reply to Greg Kroah-Hartman

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

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

From: Gabriele Monaco <[email protected]>

[ Upstream commit 79de661707a4a2dc695fd3e00529a14b4f5ec50d ]

RV monitors relying on the preemptirqs tracepoints are set as dependent on PREEMPT_TRACER and IRQSOFF_TRACER. In fact, those configurations do enable the tracepoints but are not the minimal configurations enabling them, which are TRACE_PREEMPT_TOGGLE and TRACE_IRQFLAGS (not selectable manually).

Set TRACE_PREEMPT_TOGGLE and TRACE_IRQFLAGS as dependencies for monitors.

Cc Masami Hiramatsu <[email protected]>
Cc Ingo Molnar <[email protected]>
Cc Peter Zijlstra <[email protected]>
Cc Tomas Glozar <[email protected]>
Cc Juri Lelli <[email protected]>
Cc Clark Williams <[email protected]>
Cc John Kacur <[email protected]>
Fixes fbe6c09b7eb4 ("rv: Add scpd, snep and sncid per-cpu monitors")
---
 kernel/trace/rv/monitors/scpd/Kconfig  | 2 +-
 kernel/trace/rv/monitors/sncid/Kconfig | 2 +-
 kernel/trace/rv/monitors/snep/Kconfig  | 2 +-
 kernel/trace/rv/monitors/wip/Kconfig   | 2 +-
 4 files changed, 4 insertions(+), 4 deletions(-)
kernel/trace/rv/monitors/scpd/Kconfig +1 -1
index b9114fbf680f..682d0416188b 100644
@@ -2,7 +2,7 @@
 #
 config RV_MON_SCPD
 	depends on RV
-	depends on PREEMPT_TRACER
+	depends on TRACE_PREEMPT_TOGGLE
 	depends on RV_MON_SCHED
 	default y
 	select DA_MON_EVENTS_IMPLICIT
kernel/trace/rv/monitors/sncid/Kconfig +1 -1
index 76bcfef4fd10..3a5639feaaaf 100644
@@ -2,7 +2,7 @@
 #
 config RV_MON_SNCID
 	depends on RV
-	depends on IRQSOFF_TRACER
+	depends on TRACE_IRQFLAGS
 	depends on RV_MON_SCHED
 	default y
 	select DA_MON_EVENTS_IMPLICIT
kernel/trace/rv/monitors/snep/Kconfig +1 -1
index 77527f971232..7dd54f434ff7 100644
@@ -2,7 +2,7 @@
 #
 config RV_MON_SNEP
 	depends on RV
-	depends on PREEMPT_TRACER
+	depends on TRACE_PREEMPT_TOGGLE
 	depends on RV_MON_SCHED
 	default y
 	select DA_MON_EVENTS_IMPLICIT
kernel/trace/rv/monitors/wip/Kconfig +1 -1
index e464b9294865..87a26195792b 100644
@@ -2,7 +2,7 @@
 #
 config RV_MON_WIP
 	depends on RV
-	depends on PREEMPT_TRACER
+	depends on TRACE_PREEMPT_TOGGLE
 	select DA_MON_EVENTS_IMPLICIT
 	bool "wip monitor"
 	help
Signature · 1 line
2.39.5
author maintainer
#306 lore raw

PATCH 296/627 staging: media: atomisp: Fix stack buffer overflow in gmin_get_var_int()

In reply to Greg Kroah-Hartman

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

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

From: Kees Cook <[email protected]>

[ Upstream commit ee4cf798202d285dcbe85e4467a094c44f5ed8e6 ]

When gmin_get_config_var() calls efi.get_variable() and the EFI variable is larger than the expected buffer size, two behaviors combine to create a stack buffer overflow:

1. gmin_get_config_var() does not return the proper error code when efi.get_variable() fails. It returns the stale 'ret' value from earlier operations instead of indicating the EFI failure.

2. When efi.get_variable() returns EFI_BUFFER_TOO_SMALL, it updates *out_len to the required buffer size but writes no data to the output buffer. However, due to bug #1, gmin_get_var_int() believes the call succeeded.

The caller gmin_get_var_int() then performs:
- Allocates val[CFG_VAR_NAME_MAX + 1] (65 bytes) on stack
- Calls gmin_get_config_var(dev, is_gmin, var, val, &len) with len=64
- If EFI variable is >64 bytes, efi.get_variable() sets len=required_size
- Due to bug #1, thinks call succeeded with len=required_size
- Executes val[len] = 0, writing past end of 65-byte stack buffer

This creates a stack buffer overflow when EFI variables are larger than 64 bytes. Since EFI variables can be controlled by firmware or system configuration, this could potentially be exploited for code execution.

Fix the bug by returning proper error codes from gmin_get_config_var() based on EFI status instead of stale 'ret' value.

The gmin_get_var_int() function is called during device initialization for camera sensor configuration on Intel Bay Trail and Cherry Trail platforms using the atomisp camera stack.

---
 .../staging/media/atomisp/pci/atomisp_gmin_platform.c    | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)
drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c +5 -4
index 5f59519ac8e2..964cc3bcc0ac 100644
@@ -1272,14 +1272,15 @@ static int gmin_get_config_var(struct device *maindev,
 	if (efi_rt_services_supported(EFI_RT_SUPPORTED_GET_VARIABLE))
 		status = efi.get_variable(var16, &GMIN_CFG_VAR_EFI_GUID, NULL,
 					  (unsigned long *)out_len, out);
-	if (status == EFI_SUCCESS)
+	if (status == EFI_SUCCESS) {
 		dev_info(maindev, "found EFI entry for '%s'\n", var8);
-	else if (is_gmin)
+		return 0;
+	}
+	if (is_gmin)
 		dev_info(maindev, "Failed to find EFI gmin variable %s\n", var8);
 	else
 		dev_info(maindev, "Failed to find EFI variable %s\n", var8);
-
-	return ret;
+	return -ENOENT;
 }
 
 int gmin_get_var_int(struct device *dev, bool is_gmin, const char *var, int def)
Signature · 1 line
2.39.5
author maintainer
#307 lore raw

PATCH 297/627 fortify: Fix incorrect reporting of read buffer size

In reply to Greg Kroah-Hartman

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

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

From: Kees Cook <[email protected]>

[ Upstream commit 94fd44648dae2a5b6149a41faa0b07928c3e1963 ]

When FORTIFY_SOURCE reports about a run-time buffer overread, the wrong buffer size was being shown in the error message. (The bounds checking was correct.)

Fixes 3d965b33e40d ("fortify: Improve buffer overflow reporting")
---
 include/linux/fortify-string.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
include/linux/fortify-string.h +1 -1
index e4ce1cae03bf..b3b53f8c1b28 100644
@@ -596,7 +596,7 @@ __FORTIFY_INLINE bool fortify_memcpy_chk(__kernel_size_t size,
 	if (p_size != SIZE_MAX && p_size < size)
 		fortify_panic(func, FORTIFY_WRITE, p_size, size, true);
 	else if (q_size != SIZE_MAX && q_size < size)
-		fortify_panic(func, FORTIFY_READ, p_size, size, true);
+		fortify_panic(func, FORTIFY_READ, q_size, size, true);
 
 	/*
 	 * Warn when writing beyond destination field size.
Signature · 1 line
2.39.5
author maintainer
#308 lore raw

PATCH 298/627 pinctrl: cirrus: madera-core: Use devm_pinctrl_register_mappings()

In reply to Greg Kroah-Hartman

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

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

From: Thomas Richard <[email protected]>

[ Upstream commit 90256033c11028a57437b145449c0dab196183b9 ]

Use devm_pinctrl_register_mappings(), so the mappings are automatically unregistered by the core. If pinctrl_enable() failed during the probe, pinctrl_mappings were not freed. Now it is done by the core.

---
 drivers/pinctrl/cirrus/pinctrl-madera-core.c | 14 +++-----------
 1 file changed, 3 insertions(+), 11 deletions(-)
drivers/pinctrl/cirrus/pinctrl-madera-core.c +3 -11
index 73ec5b9beb49..d19ef13224cc 100644
@@ -1061,8 +1061,9 @@ static int madera_pin_probe(struct platform_device *pdev)
 
 	/* if the configuration is provided through pdata, apply it */
 	if (pdata->gpio_configs) {
-		ret = pinctrl_register_mappings(pdata->gpio_configs,
-						pdata->n_gpio_configs);
+		ret = devm_pinctrl_register_mappings(priv->dev,
+						     pdata->gpio_configs,
+						     pdata->n_gpio_configs);
 		if (ret)
 			return dev_err_probe(priv->dev, ret,
 						"Failed to register pdata mappings\n");
@@ -1081,17 +1082,8 @@ static int madera_pin_probe(struct platform_device *pdev)
 	return 0;
 }
 
-static void madera_pin_remove(struct platform_device *pdev)
-{
-	struct madera_pin_private *priv = platform_get_drvdata(pdev);
-
-	if (priv->madera->pdata.gpio_configs)
-		pinctrl_unregister_mappings(priv->madera->pdata.gpio_configs);
-}
-
 static struct platform_driver madera_pin_driver = {
 	.probe = madera_pin_probe,
-	.remove = madera_pin_remove,
 	.driver = {
 		.name = "madera-pinctrl",
 	},
Signature · 1 line
2.39.5
author maintainer
#309 lore raw

PATCH 299/627 remoteproc: qcom: pas: Conclude the rename from adsp

In reply to Greg Kroah-Hartman

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

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

From: Bjorn Andersson <[email protected]>

[ Upstream commit 2c0c883f895f16fd9d367ec2e64bccab907d8d87 ]

The change that renamed the driver from "adsp" to "pas" didn't change any of the implementation. The result is an aesthetic eyesore, and confusing to many.

Conclude the rename of the driver, by updating function, structures and variable names to match what the driver actually is. The "Hexagon v5" is also dropped from the name and Kconfig, as this isn't correct either.

No functional change.

Fixes 9e004f97161d ("remoteproc: qcom: Rename Hexagon v5 PAS driver")
---
 drivers/remoteproc/Kconfig         |  11 +-
 drivers/remoteproc/qcom_q6v5_pas.c | 621 ++++++++++++++---------------
 2 files changed, 313 insertions(+), 319 deletions(-)
drivers/remoteproc/Kconfig +5 -6
index 83962a114dc9..48a0d3a69ed0 100644
@@ -214,7 +214,7 @@ config QCOM_Q6V5_MSS
 	  handled by QCOM_Q6V5_PAS driver.
 
 config QCOM_Q6V5_PAS
-	tristate "Qualcomm Hexagon v5 Peripheral Authentication Service support"
+	tristate "Qualcomm Peripheral Authentication Service support"
 	depends on OF && ARCH_QCOM
 	depends on QCOM_SMEM
 	depends on RPMSG_QCOM_SMD || RPMSG_QCOM_SMD=n
@@ -229,11 +229,10 @@ config QCOM_Q6V5_PAS
 	select QCOM_RPROC_COMMON
 	select QCOM_SCM
 	help
-	  Say y here to support the TrustZone based Peripheral Image Loader
-	  for the Qualcomm Hexagon v5 based remote processors. This is commonly
-	  used to control subsystems such as ADSP (Audio DSP),
-	  CDSP (Compute DSP), MPSS (Modem Peripheral SubSystem), and
-	  SLPI (Sensor Low Power Island).
+	  Say y here to support the TrustZone based Peripheral Image Loader for
+	  the Qualcomm remote processors. This is commonly used to control
+	  subsystems such as ADSP (Audio DSP), CDSP (Compute DSP), MPSS (Modem
+	  Peripheral SubSystem), and SLPI (Sensor Low Power Island).
 
 config QCOM_Q6V5_WCSS
 	tristate "Qualcomm Hexagon based WCSS Peripheral Image Loader"
drivers/remoteproc/qcom_q6v5_pas.c +308 -313

1,235 lines not on this page · raw message

Signature · 1 line
2.39.5
author maintainer
#310 lore raw

PATCH 300/627 PCI: rockchip-host: Fix "Unexpected Completion" log message

In reply to Greg Kroah-Hartman

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

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

From: Hans Zhang <[email protected]>

[ Upstream commit fcc5f586c4edbcc10de23fb9b8c0972a84e945cd ]

Fix the debug message for the PCIE_CORE_INT_UCR interrupt to clearly indicate "Unexpected Completion" instead of a duplicate "malformed TLP" message.

Fixes e77f847df54c ("PCI: rockchip: Add Rockchip PCIe controller support")
[mani: added fixes tag]
---
 drivers/pci/controller/pcie-rockchip-host.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/pci/controller/pcie-rockchip-host.c +1 -1
index b9e7a8710cf0..648b6fcb93b0 100644
@@ -439,7 +439,7 @@ static irqreturn_t rockchip_pcie_subsys_irq_handler(int irq, void *arg)
 			dev_dbg(dev, "malformed TLP received from the link\n");
 
 		if (sub_reg & PCIE_CORE_INT_UCR)
-			dev_dbg(dev, "malformed TLP received from the link\n");
+			dev_dbg(dev, "Unexpected Completion received from the link\n");
 
 		if (sub_reg & PCIE_CORE_INT_FCE)
 			dev_dbg(dev, "an error was observed in the flow control advertisements from the other side\n");
Signature · 1 line
2.39.5
author maintainer
#311 lore raw

PATCH 301/627 clk: renesas: rzv2h: Fix missing CLK_SET_RATE_PARENT flag for ddiv clocks

In reply to Greg Kroah-Hartman

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

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

From: Lad Prabhakar <[email protected]>

[ Upstream commit 715676d8418062f54d746451294ccce9786c1734 ]

Commit bc4d25fdfadf ("clk: renesas: rzv2h: Add support for dynamic switching divider clocks") missed setting the CLK_SET_RATE_PARENT flag when registering ddiv clocks.

Without this flag, rate changes to the divider clock do not propagate to its parent, potentially resulting in incorrect clock configurations.

Fix this by setting CLK_SET_RATE_PARENT in the clock init data.

Fixes bc4d25fdfadfa ("clk: renesas: rzv2h: Add support for dynamic switching divider clocks")
---
 drivers/clk/renesas/rzv2h-cpg.c | 1 +
 1 file changed, 1 insertion(+)
drivers/clk/renesas/rzv2h-cpg.c +1 -0
index bcc496e8cbcd..fb39e6446b26 100644
@@ -381,6 +381,7 @@ rzv2h_cpg_ddiv_clk_register(const struct cpg_core_clk *core,
 		init.ops = &rzv2h_ddiv_clk_divider_ops;
 	init.parent_names = &parent_name;
 	init.num_parents = 1;
+	init.flags = CLK_SET_RATE_PARENT;
 
 	ddiv->priv = priv;
 	ddiv->mon = cfg_ddiv.monbit;
Signature · 1 line
2.39.5
author maintainer
#312 lore raw

PATCH 302/627 crypto: sun8i-ce - fix nents passed to dma_unmap_sg()

In reply to Greg Kroah-Hartman

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

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

From: Ovidiu Panait <[email protected]>

[ Upstream commit b6cd3cfb5afe49952f8f6be947aeeca9ba0faebb ]

In sun8i_ce_cipher_unprepare(), dma_unmap_sg() is incorrectly called with the number of entries returned by dma_map_sg(), rather than using the original number of entries passed when mapping the scatterlist.

To fix this, stash the original number of entries passed to dma_map_sg() in the request context.

Fixes 0605fa0f7826 ("crypto: sun8i-ce - split into prepare/run/unprepare")
---
 drivers/crypto/allwinner/sun8i-ce/sun8i-ce-cipher.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/crypto/allwinner/sun8i-ce/sun8i-ce-cipher.c +2 -2
index f9cf00d690e2..7cd3b13f3bdc 100644
@@ -278,8 +278,8 @@ static int sun8i_ce_cipher_prepare(struct crypto_engine *engine, void *async_req
 	}
 
 	chan->timeout = areq->cryptlen;
-	rctx->nr_sgs = nr_sgs;
-	rctx->nr_sgd = nr_sgd;
+	rctx->nr_sgs = ns;
+	rctx->nr_sgd = nd;
 	return 0;
 
 theend_sgs:
Signature · 1 line
2.39.5
author maintainer
#313 lore raw

PATCH 303/627 crypto: qat - use unmanaged allocation for dc_data

In reply to Greg Kroah-Hartman

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

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

From: Suman Kumar Chakraborty <[email protected]>

[ Upstream commit 4cc871ad0173e8bc22f80e3609e34d546d30ef1a ]

The dc_data structure holds data required for handling compression operations, such as overflow buffers. In this context, the use of managed memory allocation APIs (devm_kzalloc() and devm_kfree()) is not necessary, as these data structures are freed and re-allocated when a device is restarted in adf_dev_down() and adf_dev_up().

Additionally, managed APIs automatically handle memory cleanup when the device is detached, which can lead to conflicts with manual cleanup processes. Specifically, if a device driver invokes the adf_dev_down() function as part of the cleanup registered with devm_add_action_or_reset(), it may attempt to free memory that is also managed by the device's resource management system, potentially leading to a double-free.

This might result in a warning similar to the following when unloading the device specific driver, for example qat_6xxx.ko:

    qat_free_dc_data+0x4f/0x60 [intel_qat]
    qat_compression_event_handler+0x3d/0x1d0 [intel_qat]
    adf_dev_shutdown+0x6d/0x1a0 [intel_qat]
    adf_dev_down+0x32/0x50 [intel_qat]
    devres_release_all+0xb8/0x110
    device_unbind_cleanup+0xe/0x70
    device_release_driver_internal+0x1c1/0x200
    driver_detach+0x48/0x90
    bus_remove_driver+0x74/0xf0
    pci_unregister_driver+0x2e/0xb0

Use unmanaged memory allocation APIs (kzalloc_node() and kfree()) for the dc_data structure. This ensures that memory is explicitly allocated and freed under the control of the driver code, preventing manual deallocation from interfering with automatic cleanup.

Fixes 1198ae56c9a5 ("crypto: qat - expose deflate through acomp api for QAT GEN2")
---
 drivers/crypto/intel/qat/qat_common/qat_compression.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/crypto/intel/qat/qat_common/qat_compression.c +2 -2
index c285b45b8679..0a77ca65c8d4 100644
@@ -196,7 +196,7 @@ static int qat_compression_alloc_dc_data(struct adf_accel_dev *accel_dev)
 	struct adf_dc_data *dc_data = NULL;
 	u8 *obuff = NULL;
 
-	dc_data = devm_kzalloc(dev, sizeof(*dc_data), GFP_KERNEL);
+	dc_data = kzalloc_node(sizeof(*dc_data), GFP_KERNEL, dev_to_node(dev));
 	if (!dc_data)
 		goto err;
 
@@ -234,7 +234,7 @@ static void qat_free_dc_data(struct adf_accel_dev *accel_dev)
 	dma_unmap_single(dev, dc_data->ovf_buff_p, dc_data->ovf_buff_sz,
 			 DMA_FROM_DEVICE);
 	kfree_sensitive(dc_data->ovf_buff);
-	devm_kfree(dev, dc_data);
+	kfree(dc_data);
 	accel_dev->dc_data = NULL;
 }
 
Signature · 1 line
2.39.5
author maintainer
#314 lore raw

PATCH 304/627 crypto: marvell/cesa - Fix engine load inaccuracy

In reply to Greg Kroah-Hartman

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

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

From: Herbert Xu <[email protected]>

[ Upstream commit 442134ab30e75b7229c4bfc1ac5641d245cffe27 ]

If an error occurs during queueing the engine load will never be decremented. Fix this by moving the engine load adjustment into the cleanup function.

Fixes bf8f91e71192 ("crypto: marvell - Add load balancing between engines")
---
 drivers/crypto/marvell/cesa/cipher.c | 4 +++-
 drivers/crypto/marvell/cesa/hash.c   | 5 +++--
 2 files changed, 6 insertions(+), 3 deletions(-)
drivers/crypto/marvell/cesa/cipher.c +3 -1
index 48c5c8ea8c43..3fe0fd9226cf 100644
@@ -75,9 +75,12 @@ mv_cesa_skcipher_dma_cleanup(struct skcipher_request *req)
 static inline void mv_cesa_skcipher_cleanup(struct skcipher_request *req)
 {
 	struct mv_cesa_skcipher_req *creq = skcipher_request_ctx(req);
+	struct mv_cesa_engine *engine = creq->base.engine;
 
 	if (mv_cesa_req_get_type(&creq->base) == CESA_DMA_REQ)
 		mv_cesa_skcipher_dma_cleanup(req);
+
+	atomic_sub(req->cryptlen, &engine->load);
 }
 
 static void mv_cesa_skcipher_std_step(struct skcipher_request *req)
@@ -212,7 +215,6 @@ mv_cesa_skcipher_complete(struct crypto_async_request *req)
 	struct mv_cesa_engine *engine = creq->base.engine;
 	unsigned int ivsize;
 
-	atomic_sub(skreq->cryptlen, &engine->load);
 	ivsize = crypto_skcipher_ivsize(crypto_skcipher_reqtfm(skreq));
 
 	if (mv_cesa_req_get_type(&creq->base) == CESA_DMA_REQ) {
drivers/crypto/marvell/cesa/hash.c +3 -2
index 6815eddc9068..e339ce7ad533 100644
@@ -110,9 +110,12 @@ static inline void mv_cesa_ahash_dma_cleanup(struct ahash_request *req)
 static inline void mv_cesa_ahash_cleanup(struct ahash_request *req)
 {
 	struct mv_cesa_ahash_req *creq = ahash_request_ctx(req);
+	struct mv_cesa_engine *engine = creq->base.engine;
 
 	if (mv_cesa_req_get_type(&creq->base) == CESA_DMA_REQ)
 		mv_cesa_ahash_dma_cleanup(req);
+
+	atomic_sub(req->nbytes, &engine->load);
 }
 
 static void mv_cesa_ahash_last_cleanup(struct ahash_request *req)
@@ -395,8 +398,6 @@ static void mv_cesa_ahash_complete(struct crypto_async_request *req)
 			}
 		}
 	}
-
-	atomic_sub(ahashreq->nbytes, &engine->load);
 }
 
 static void mv_cesa_ahash_prepare(struct crypto_async_request *req,
Signature · 1 line
2.39.5
author maintainer
#315 lore raw

PATCH 305/627 crypto: s390/hmac - Fix counter in export state

In reply to Greg Kroah-Hartman

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

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

From: Herbert Xu <[email protected]>

[ Upstream commit 1b39bc4a703a63a22c08232015540adfb31f22ba ]

The hmac export state needs to be one block-size bigger to account for the ipad.

Fixes 08811169ac01 ("crypto: s390/hmac - Use API partial block handling")
---
 arch/s390/crypto/hmac_s390.c | 12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)
arch/s390/crypto/hmac_s390.c +8 -4
index 93a1098d9f8d..58444da9b004 100644
@@ -290,6 +290,7 @@ static int s390_hmac_export(struct shash_desc *desc, void *out)
 	struct s390_kmac_sha2_ctx *ctx = shash_desc_ctx(desc);
 	unsigned int bs = crypto_shash_blocksize(desc->tfm);
 	unsigned int ds = bs / 2;
+	u64 lo = ctx->buflen[0];
 	union {
 		u8 *u8;
 		u64 *u64;
@@ -301,9 +302,10 @@ static int s390_hmac_export(struct shash_desc *desc, void *out)
 	else
 		memcpy(p.u8, ctx->param, ds);
 	p.u8 += ds;
-	put_unaligned(ctx->buflen[0], p.u64++);
+	lo += bs;
+	put_unaligned(lo, p.u64++);
 	if (ds == SHA512_DIGEST_SIZE)
-		put_unaligned(ctx->buflen[1], p.u64);
+		put_unaligned(ctx->buflen[1] + (lo < bs), p.u64);
 	return err;
 }
 
@@ -316,14 +318,16 @@ static int s390_hmac_import(struct shash_desc *desc, const void *in)
 		const u8 *u8;
 		const u64 *u64;
 	} p = { .u8 = in };
+	u64 lo;
 	int err;
 
 	err = s390_hmac_sha2_init(desc);
 	memcpy(ctx->param, p.u8, ds);
 	p.u8 += ds;
-	ctx->buflen[0] = get_unaligned(p.u64++);
+	lo = get_unaligned(p.u64++);
+	ctx->buflen[0] = lo - bs;
 	if (ds == SHA512_DIGEST_SIZE)
-		ctx->buflen[1] = get_unaligned(p.u64);
+		ctx->buflen[1] = get_unaligned(p.u64) - (lo < bs);
 	if (ctx->buflen[0] | ctx->buflen[1])
 		ctx->gr0.ikp = 1;
 	return err;
Signature · 1 line
2.39.5
author maintainer
#316 lore raw

PATCH 306/627 crypto: s390/sha3 - Use cpu byte-order when exporting

In reply to Greg Kroah-Hartman

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

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

From: Herbert Xu <[email protected]>

[ Upstream commit 73c2437109c3eab274258a6430ae5dafac1ef43e ]

The sha3 partial hash on s390 is in little-endian just like the final hash. However the generic implementation produces native or big-endian partial hashes.

Make s390 sha3 conform to that by doing the byte-swap on export and import.

Fixes 6f90ba706551 ("crypto: s390/sha3 - Use API partial block handling")
---
 arch/s390/crypto/sha.h           |  3 +++
 arch/s390/crypto/sha3_256_s390.c | 24 +++++++++++++++++-------
 arch/s390/crypto/sha3_512_s390.c | 25 +++++++++++++++++--------
 3 files changed, 37 insertions(+), 15 deletions(-)
arch/s390/crypto/sha.h +3 -0
index d757ccbce2b4..cadb4b13622a 100644
@@ -27,6 +27,9 @@ struct s390_sha_ctx {
 			u64 state[SHA512_DIGEST_SIZE / sizeof(u64)];
 			u64 count_hi;
 		} sha512;
+		struct {
+			__le64 state[SHA3_STATE_SIZE / sizeof(u64)];
+		} sha3;
 	};
 	int func;		/* KIMD function to use */
 	bool first_message_part;
arch/s390/crypto/sha3_256_s390.c +17 -7
index 4a7731ac6bcd..03bb4f4bab70 100644
@@ -35,23 +35,33 @@ static int sha3_256_init(struct shash_desc *desc)
 static int sha3_256_export(struct shash_desc *desc, void *out)
 {
 	struct s390_sha_ctx *sctx = shash_desc_ctx(desc);
-	struct sha3_state *octx = out;
+	union {
+		u8 *u8;
+		u64 *u64;
+	} p = { .u8 = out };
+	int i;
 
 	if (sctx->first_message_part) {
-		memset(sctx->state, 0, sizeof(sctx->state));
-		sctx->first_message_part = 0;
+		memset(out, 0, SHA3_STATE_SIZE);
+		return 0;
 	}
-	memcpy(octx->st, sctx->state, sizeof(octx->st));
+	for (i = 0; i < SHA3_STATE_SIZE / 8; i++)
+		put_unaligned(le64_to_cpu(sctx->sha3.state[i]), p.u64++);
 	return 0;
 }
 
 static int sha3_256_import(struct shash_desc *desc, const void *in)
 {
 	struct s390_sha_ctx *sctx = shash_desc_ctx(desc);
-	const struct sha3_state *ictx = in;
-
+	union {
+		const u8 *u8;
+		const u64 *u64;
+	} p = { .u8 = in };
+	int i;
+
+	for (i = 0; i < SHA3_STATE_SIZE / 8; i++)
+		sctx->sha3.state[i] = cpu_to_le64(get_unaligned(p.u64++));
 	sctx->count = 0;
-	memcpy(sctx->state, ictx->st, sizeof(ictx->st));
 	sctx->first_message_part = 0;
 	sctx->func = CPACF_KIMD_SHA3_256;
 
arch/s390/crypto/sha3_512_s390.c +17 -8
index 018f02fff444..a5c9690eecb1 100644
@@ -34,24 +34,33 @@ static int sha3_512_init(struct shash_desc *desc)
 static int sha3_512_export(struct shash_desc *desc, void *out)
 {
 	struct s390_sha_ctx *sctx = shash_desc_ctx(desc);
-	struct sha3_state *octx = out;
-
+	union {
+		u8 *u8;
+		u64 *u64;
+	} p = { .u8 = out };
+	int i;
 
 	if (sctx->first_message_part) {
-		memset(sctx->state, 0, sizeof(sctx->state));
-		sctx->first_message_part = 0;
+		memset(out, 0, SHA3_STATE_SIZE);
+		return 0;
 	}
-	memcpy(octx->st, sctx->state, sizeof(octx->st));
+	for (i = 0; i < SHA3_STATE_SIZE / 8; i++)
+		put_unaligned(le64_to_cpu(sctx->sha3.state[i]), p.u64++);
 	return 0;
 }
 
 static int sha3_512_import(struct shash_desc *desc, const void *in)
 {
 	struct s390_sha_ctx *sctx = shash_desc_ctx(desc);
-	const struct sha3_state *ictx = in;
-
+	union {
+		const u8 *u8;
+		const u64 *u64;
+	} p = { .u8 = in };
+	int i;
+
+	for (i = 0; i < SHA3_STATE_SIZE / 8; i++)
+		sctx->sha3.state[i] = cpu_to_le64(get_unaligned(p.u64++));
 	sctx->count = 0;
-	memcpy(sctx->state, ictx->st, sizeof(ictx->st));
 	sctx->first_message_part = 0;
 	sctx->func = CPACF_KIMD_SHA3_512;
 
Signature · 1 line
2.39.5
author maintainer
#317 lore raw

PATCH 307/627 padata: Fix pd UAF once and for all

In reply to Greg Kroah-Hartman

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

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

From: Herbert Xu <[email protected]>

[ Upstream commit 71203f68c7749609d7fc8ae6ad054bdedeb24f91 ]

There is a race condition/UAF in padata_reorder that goes back to the initial commit. A reference count is taken at the start of the process in padata_do_parallel, and released at the end in padata_serial_worker.

This reference count is (and only is) required for padata_replace to function correctly. If padata_replace is never called then there is no issue.

In the function padata_reorder which serves as the core of padata, as soon as padata is added to queue->serial.list, and the associated spin lock released, that padata may be processed and the reference count on pd would go away.

Fix this by getting the next padata before the squeue->serial lock is released.

In order to make this possible, simplify padata_reorder by only calling it once the next padata arrives.

Fixes 16295bec6398 ("padata: Generic parallelization/serialization interface")
---
 include/linux/padata.h |   3 -
 kernel/padata.c        | 132 ++++++++++++-----------------------------
 2 files changed, 37 insertions(+), 98 deletions(-)
include/linux/padata.h +0 -3
index 0146daf34430..b486c7359de2 100644
@@ -91,7 +91,6 @@ struct padata_cpumask {
  * @cpu: Next CPU to be processed.
  * @cpumask: The cpumasks in use for parallel and serial workers.
  * @reorder_work: work struct for reordering.
- * @lock: Reorder lock.
  */
 struct parallel_data {
 	struct padata_shell		*ps;
@@ -102,8 +101,6 @@ struct parallel_data {
 	unsigned int			processed;
 	int				cpu;
 	struct padata_cpumask		cpumask;
-	struct work_struct		reorder_work;
-	spinlock_t                      ____cacheline_aligned lock;
 };
 
 /**
kernel/padata.c +37 -95

198 lines not on this page · raw message

Signature · 1 line
2.39.5
author maintainer
#318 lore raw

PATCH 308/627 crypto: ccp - Fix dereferencing uninitialized error pointer

In reply to Greg Kroah-Hartman

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

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

From: Ashish Kalra <[email protected]>

[ Upstream commit 0fa766726c091ff0ec7d26874f6e4724d23ecb0e ]

Fix below smatch warnings:
drivers/crypto/ccp/sev-dev.c:1312 __sev_platform_init_locked()
error: we previously assumed 'error' could be null

Fixes 9770b428b1a2 ("crypto: ccp - Move dev_info/err messages for SEV/SNP init and shutdown")
---
 drivers/crypto/ccp/sev-dev.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)
drivers/crypto/ccp/sev-dev.c +5 -3
index 3451bada884e..8fb94c5f006a 100644
@@ -1276,9 +1276,11 @@ static int __sev_platform_init_handle_init_ex_path(struct sev_device *sev)
 
 static int __sev_platform_init_locked(int *error)
 {
-	int rc, psp_ret = SEV_RET_NO_FW_CALL;
+	int rc, psp_ret, dfflush_error;
 	struct sev_device *sev;
 
+	psp_ret = dfflush_error = SEV_RET_NO_FW_CALL;
+
 	if (!psp_master || !psp_master->sev_data)
 		return -ENODEV;
 
@@ -1320,10 +1322,10 @@ static int __sev_platform_init_locked(int *error)
 
 	/* Prepare for first SEV guest launch after INIT */
 	wbinvd_on_all_cpus();
-	rc = __sev_do_cmd_locked(SEV_CMD_DF_FLUSH, NULL, error);
+	rc = __sev_do_cmd_locked(SEV_CMD_DF_FLUSH, NULL, &dfflush_error);
 	if (rc) {
 		dev_err(sev->dev, "SEV: DF_FLUSH failed %#x, rc %d\n",
-			*error, rc);
+			dfflush_error, rc);
 		return rc;
 	}
 
Signature · 1 line
2.39.5
author maintainer
#319 lore raw

PATCH 309/627 crypto: qat - allow enabling VFs in the absence of IOMMU

In reply to Greg Kroah-Hartman

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

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

From: Ahsan Atta <[email protected]>

[ Upstream commit 53669ff591d4deb2d80eed4c07593ad0c0b45899 ]

The commit ca88a2bdd4dd ("crypto: qat - allow disabling SR-IOV VFs") introduced an unnecessary change that prevented enabling SR-IOV when IOMMU is disabled. In certain scenarios, it is desirable to enable SR-IOV even in the absence of IOMMU. Thus, restoring the previous functionality to allow VFs to be enumerated in the absence of IOMMU.

Fixes ca88a2bdd4dd ("crypto: qat - allow disabling SR-IOV VFs")
---
 drivers/crypto/intel/qat/qat_common/adf_sriov.c | 1 -
 1 file changed, 1 deletion(-)
drivers/crypto/intel/qat/qat_common/adf_sriov.c +0 -1
index c75d0b6cb0ad..31d1ef0cb1f5 100644
@@ -155,7 +155,6 @@ static int adf_do_enable_sriov(struct adf_accel_dev *accel_dev)
 	if (!device_iommu_mapped(&GET_DEV(accel_dev))) {
 		dev_warn(&GET_DEV(accel_dev),
 			 "IOMMU should be enabled for SR-IOV to work correctly\n");
-		return -EINVAL;
 	}
 
 	if (adf_dev_started(accel_dev)) {
Signature · 1 line
2.39.5
author maintainer
#320 lore raw

PATCH 310/627 crypto: qat - fix state restore for banks with exceptions

In reply to Greg Kroah-Hartman

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

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

From: Svyatoslav Pankratov <[email protected]>

[ Upstream commit 254923ca8715f623704378266815b6d14eb26194 ]

Change the logic in the restore function to properly handle bank exceptions.

The check for exceptions in the saved state should be performed before conducting any other ringstat register checks.
If a bank was saved with an exception, the ringstat will have the appropriate rp_halt/rp_exception bits set, causing the driver to exit the restore process with an error. Instead, the restore routine should first check the ringexpstat register, and if any exception was raised, it should stop further checks and return without any error. In other words, if a ring pair is in an exception state at the source, it should be restored the same way at the destination but without raising an error.

Even though this approach might lead to losing the exception state during migration, the driver will log the exception from the saved state during the restore process.

Fixes bbfdde7d195f ("crypto: qat - add bank save and restore flows")
---
 .../intel/qat/qat_common/adf_gen4_hw_data.c   | 29 ++++++++++++++-----
 1 file changed, 22 insertions(+), 7 deletions(-)
drivers/crypto/intel/qat/qat_common/adf_gen4_hw_data.c +22 -7
index 0406cb09c5bb..14d0fdd66a4b 100644
@@ -581,6 +581,28 @@ static int bank_state_restore(struct adf_hw_csr_ops *ops, void __iomem *base,
 	ops->write_csr_int_srcsel_w_val(base, bank, state->iaintflagsrcsel0);
 	ops->write_csr_exp_int_en(base, bank, state->ringexpintenable);
 	ops->write_csr_int_col_ctl(base, bank, state->iaintcolctl);
+
+	/*
+	 * Verify whether any exceptions were raised during the bank save process.
+	 * If exceptions occurred, the status and exception registers cannot
+	 * be directly restored. Consequently, further restoration is not
+	 * feasible, and the current state of the ring should be maintained.
+	 */
+	val = state->ringexpstat;
+	if (val) {
+		pr_info("QAT: Bank %u state not fully restored due to exception in saved state (%#x)\n",
+			bank, val);
+		return 0;
+	}
+
+	/* Ensure that the restoration process completed without exceptions */
+	tmp_val = ops->read_csr_exp_stat(base, bank);
+	if (tmp_val) {
+		pr_err("QAT: Bank %u restored with exception: %#x\n",
+		       bank, tmp_val);
+		return -EFAULT;
+	}
+
 	ops->write_csr_ring_srv_arb_en(base, bank, state->ringsrvarben);
 
 	/* Check that all ring statuses match the saved state. */
@@ -614,13 +636,6 @@ static int bank_state_restore(struct adf_hw_csr_ops *ops, void __iomem *base,
 	if (ret)
 		return ret;
 
-	tmp_val = ops->read_csr_exp_stat(base, bank);
-	val = state->ringexpstat;
-	if (tmp_val && !val) {
-		pr_err("QAT: Bank was restored with exception: 0x%x\n", val);
-		return -EINVAL;
-	}
-
 	return 0;
 }
 
Signature · 1 line
2.39.5
author maintainer
#321 lore raw

PATCH 311/627 mtd: fix possible integer overflow in erase_xfer()

In reply to Greg Kroah-Hartman

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

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

From: Ivan Stepchenko <[email protected]>

[ Upstream commit 9358bdb9f9f54d94ceafc650deffefd737d19fdd ]

The expression '1 << EraseUnitSize' is evaluated in int, which causes a negative result when shifting by 31 - the upper bound of the valid range [10, 31], enforced by scan_header(). This leads to incorrect extension when storing the result in 'erase->len' (uint64_t), producing a large unexpected value.

Found by Linux Verification Center (linuxtesting.org) with Svace.

Fixes 1da177e4c3f4 ("Linux-2.6.12-rc2")
---
 drivers/mtd/ftl.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/mtd/ftl.c +1 -1
index 8c22064ead38..f2bd1984609c 100644
@@ -344,7 +344,7 @@ static int erase_xfer(partition_t *part,
             return -ENOMEM;
 
     erase->addr = xfer->Offset;
-    erase->len = 1 << part->header.EraseUnitSize;
+    erase->len = 1ULL << part->header.EraseUnitSize;
 
     ret = mtd_erase(part->mbd.mtd, erase);
     if (!ret) {
Signature · 1 line
2.39.5
author maintainer
#322 lore raw

PATCH 312/627 clk: davinci: Add NULL check in davinci_lpsc_clk_register()

In reply to Greg Kroah-Hartman

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

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

From: Henry Martin <[email protected]>

[ Upstream commit 13de464f445d42738fe18c9a28bab056ba3a290a ]

devm_kasprintf() returns NULL when memory allocation fails. Currently, davinci_lpsc_clk_register() does not check for this case, which results in a NULL pointer dereference.

Add NULL check after devm_kasprintf() to prevent this issue and ensuring no resources are left allocated.

Fixes c6ed4d734bc7 ("clk: davinci: New driver for davinci PSC clocks")
---
 drivers/clk/davinci/psc.c | 5 +++++
 1 file changed, 5 insertions(+)
drivers/clk/davinci/psc.c +5 -0
index b48322176c21..f3ee9397bb0c 100644
@@ -277,6 +277,11 @@ davinci_lpsc_clk_register(struct device *dev, const char *name,
 
 	lpsc->pm_domain.name = devm_kasprintf(dev, GFP_KERNEL, "%s: %s",
 					      best_dev_name(dev), name);
+	if (!lpsc->pm_domain.name) {
+		clk_hw_unregister(&lpsc->hw);
+		kfree(lpsc);
+		return ERR_PTR(-ENOMEM);
+	}
 	lpsc->pm_domain.attach_dev = davinci_psc_genpd_attach_dev;
 	lpsc->pm_domain.detach_dev = davinci_psc_genpd_detach_dev;
 	lpsc->pm_domain.flags = GENPD_FLAG_PM_CLK;
Signature · 1 line
2.39.5
author maintainer
#323 lore raw

PATCH 313/627 media: imx-jpeg: Account for data_offset when getting image address

In reply to Greg Kroah-Hartman

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

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

From: Ming Qian <[email protected]>

[ Upstream commit 51ad3b570ea7b1916ff4db993f1aa22bb48fdac6 ]

Applications may set data_offset when it refers to an output queue. So driver need to account for it when getting the start address of input image in the plane.

Meanwhile the mxc-jpeg codec requires the address (plane address + data_offset) to be 16-aligned.

Fixes 2db16c6ed72c ("media: imx-jpeg: Add V4L2 driver for i.MX8 JPEG Encoder/Decoder")
---
 .../media/platform/nxp/imx-jpeg/mxc-jpeg.c    | 47 ++++++++++++++-----
 .../media/platform/nxp/imx-jpeg/mxc-jpeg.h    |  1 +
 2 files changed, 37 insertions(+), 11 deletions(-)
drivers/media/platform/nxp/imx-jpeg/mxc-jpeg.c +36 -11
index 5c17bc58181e..8681dd193033 100644
@@ -598,6 +598,27 @@ static void _bswap16(u16 *a)
 	*a = ((*a & 0x00FF) << 8) | ((*a & 0xFF00) >> 8);
 }
 
+static dma_addr_t mxc_jpeg_get_plane_dma_addr(struct vb2_buffer *buf, unsigned int plane_no)
+{
+	if (plane_no >= buf->num_planes)
+		return 0;
+	return vb2_dma_contig_plane_dma_addr(buf, plane_no) + buf->planes[plane_no].data_offset;
+}
+
+static void *mxc_jpeg_get_plane_vaddr(struct vb2_buffer *buf, unsigned int plane_no)
+{
+	if (plane_no >= buf->num_planes)
+		return NULL;
+	return vb2_plane_vaddr(buf, plane_no) + buf->planes[plane_no].data_offset;
+}
+
+static unsigned long mxc_jpeg_get_plane_payload(struct vb2_buffer *buf, unsigned int plane_no)
+{
+	if (plane_no >= buf->num_planes)
+		return 0;
+	return vb2_get_plane_payload(buf, plane_no) - buf->planes[plane_no].data_offset;
+}
+
 static void print_mxc_buf(struct mxc_jpeg_dev *jpeg, struct vb2_buffer *buf,
 			  unsigned long len)
 {
@@ -610,11 +631,11 @@ static void print_mxc_buf(struct mxc_jpeg_dev *jpeg, struct vb2_buffer *buf,
 		return;
 
 	for (plane_no = 0; plane_no < buf->num_planes; plane_no++) {
-		payload = vb2_get_plane_payload(buf, plane_no);
+		payload = mxc_jpeg_get_plane_payload(buf, plane_no);
 		if (len == 0)
 			len = payload;
-		dma_addr = vb2_dma_contig_plane_dma_addr(buf, plane_no);
-		vaddr = vb2_plane_vaddr(buf, plane_no);
+		dma_addr = mxc_jpeg_get_plane_dma_addr(buf, plane_no);
+		vaddr = mxc_jpeg_get_plane_vaddr(buf, plane_no);
 		v4l2_dbg(3, debug, &jpeg->v4l2_dev,
 			 "plane %d (vaddr=%p dma_addr=%x payload=%ld):",
 			  plane_no, vaddr, dma_addr, payload);
@@ -712,16 +733,15 @@ static void mxc_jpeg_addrs(struct mxc_jpeg_desc *desc,
 	struct mxc_jpeg_q_data *q_data;
 
 	q_data = mxc_jpeg_get_q_data(ctx, raw_buf->type);
-	desc->buf_base0 = vb2_dma_contig_plane_dma_addr(raw_buf, 0);
+	desc->buf_base0 = mxc_jpeg_get_plane_dma_addr(raw_buf, 0);
 	desc->buf_base1 = 0;
 	if (img_fmt == STM_CTRL_IMAGE_FORMAT(MXC_JPEG_YUV420)) {
 		if (raw_buf->num_planes == 2)
-			desc->buf_base1 = vb2_dma_contig_plane_dma_addr(raw_buf, 1);
+			desc->buf_base1 = mxc_jpeg_get_plane_dma_addr(raw_buf, 1);
 		else
 			desc->buf_base1 = desc->buf_base0 + q_data->sizeimage[0];
 	}
-	desc->stm_bufbase = vb2_dma_contig_plane_dma_addr(jpeg_buf, 0) +
-		offset;
+	desc->stm_bufbase = mxc_jpeg_get_plane_dma_addr(jpeg_buf, 0) + offset;
 }
 
 static bool mxc_jpeg_is_extended_sequential(const struct mxc_jpeg_fmt *fmt)
@@ -1029,8 +1049,8 @@ static irqreturn_t mxc_jpeg_dec_irq(int irq, void *priv)
 			vb2_set_plane_payload(&dst_buf->vb2_buf, 1, payload);
 		}
 		dev_dbg(dev, "Decoding finished, payload size: %ld + %ld\n",
-			vb2_get_plane_payload(&dst_buf->vb2_buf, 0),
-			vb2_get_plane_payload(&dst_buf->vb2_buf, 1));
+			mxc_jpeg_get_plane_payload(&dst_buf->vb2_buf, 0),
+			mxc_jpeg_get_plane_payload(&dst_buf->vb2_buf, 1));
 	}
 
 	/* short preview of the results */
@@ -1889,8 +1909,8 @@ static int mxc_jpeg_parse(struct mxc_jpeg_ctx *ctx, struct vb2_buffer *vb)
 	struct mxc_jpeg_sof *psof = NULL;
 	struct mxc_jpeg_sos *psos = NULL;
 	struct mxc_jpeg_src_buf *jpeg_src_buf = vb2_to_mxc_buf(vb);
-	u8 *src_addr = (u8 *)vb2_plane_vaddr(vb, 0);
-	u32 size = vb2_get_plane_payload(vb, 0);
+	u8 *src_addr = (u8 *)mxc_jpeg_get_plane_vaddr(vb, 0);
+	u32 size = mxc_jpeg_get_plane_payload(vb, 0);
 	int ret;
 
 	memset(&header, 0, sizeof(header));
@@ -2027,6 +2047,11 @@ static int mxc_jpeg_buf_prepare(struct vb2_buffer *vb)
 				i, vb2_plane_size(vb, i), sizeimage);
 			return -EINVAL;
 		}
+		if (!IS_ALIGNED(mxc_jpeg_get_plane_dma_addr(vb, i), MXC_JPEG_ADDR_ALIGNMENT)) {
+			dev_err(dev, "planes[%d] address is not %d aligned\n",
+				i, MXC_JPEG_ADDR_ALIGNMENT);
+			return -EINVAL;
+		}
 	}
 	if (V4L2_TYPE_IS_CAPTURE(vb->vb2_queue->type)) {
 		vb2_set_plane_payload(vb, 0, 0);
drivers/media/platform/nxp/imx-jpeg/mxc-jpeg.h +1 -0
index fdde45f7e163..44e46face6d1 100644
@@ -30,6 +30,7 @@
 #define MXC_JPEG_MAX_PLANES		2
 #define MXC_JPEG_PATTERN_WIDTH		128
 #define MXC_JPEG_PATTERN_HEIGHT		64
+#define MXC_JPEG_ADDR_ALIGNMENT		16
 
 enum mxc_jpeg_enc_state {
 	MXC_JPEG_ENCODING	= 0, /* jpeg encode phase */
Signature · 1 line
2.39.5
author maintainer
#324 lore raw

PATCH 314/627 media: v4l2-ctrls: Fix H264 SEPARATE_COLOUR_PLANE check

In reply to Greg Kroah-Hartman

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

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

From: James Cowgill <[email protected]>

[ Upstream commit 803b9eabc649c778986449eb0596e5ffeb7a8aed ]

The separate_colour_plane_flag element is only present in the SPS if chroma_format_idc == 3, so the corresponding flag should be disabled whenever that is not the case and not just on profiles where chroma_format_idc is not present.

Fixes b32e48503df0 ("media: controls: Validate H264 stateless controls")
---
 drivers/media/v4l2-core/v4l2-ctrls-core.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
drivers/media/v4l2-core/v4l2-ctrls-core.c +4 -4
index 90d25329661e..b45809a82f9a 100644
@@ -968,12 +968,12 @@ static int std_validate_compound(const struct v4l2_ctrl *ctrl, u32 idx,
 
 			p_h264_sps->flags &=
 				~V4L2_H264_SPS_FLAG_QPPRIME_Y_ZERO_TRANSFORM_BYPASS;
-
-			if (p_h264_sps->chroma_format_idc < 3)
-				p_h264_sps->flags &=
-					~V4L2_H264_SPS_FLAG_SEPARATE_COLOUR_PLANE;
 		}
 
+		if (p_h264_sps->chroma_format_idc < 3)
+			p_h264_sps->flags &=
+				~V4L2_H264_SPS_FLAG_SEPARATE_COLOUR_PLANE;
+
 		if (p_h264_sps->flags & V4L2_H264_SPS_FLAG_FRAME_MBS_ONLY)
 			p_h264_sps->flags &=
 				~V4L2_H264_SPS_FLAG_MB_ADAPTIVE_FRAME_FIELD;
Signature · 1 line
2.39.5
author maintainer
#325 lore raw

PATCH 315/627 perf parse-events: Set default GH modifier properly

In reply to Greg Kroah-Hartman

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

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

From: Namhyung Kim <[email protected]>

[ Upstream commit dcbe6e51a0bb80a40f9a8c87750c291c2364573d ]

Commit 7b100989b4f6bce7 ("perf evlist: Remove __evlist__add_default") changed to use "cycles:P" as a default event. But the problem is it cannot set other default modifiers correctly.

perf kvm needs to set attr.exclude_host by default but it didn't work because of the logic in the parse_events__modifier_list(). Also the exclude_GH_default was applied only if ":u" modifier was specified - which is strange. Move it out after handling the ":GH" and check perf_host and perf_guest properly.

Before:

  $ ./perf kvm record -vv true |& grep exclude
  (nothing)

But specifying an event (without a modifier) works:

  $ ./perf kvm record -vv -e cycles true |& grep exclude
    exclude_host                     1

After:
It now works for the both cases:

  $ ./perf kvm record -vv true |& grep exclude
    exclude_host                     1
  $ ./perf kvm record -vv -e cycles true |& grep exclude
    exclude_host                     1
Fixes 35c8d21371e9b342 ("perf tools: Don't set attr.exclude_guest by default")
---
 tools/perf/util/parse-events.c | 11 ++++++++---
 1 file changed, 8 insertions(+), 3 deletions(-)
tools/perf/util/parse-events.c +8 -3
index 2380de56a207..d07c83ba6f1a 100644
@@ -1829,13 +1829,11 @@ static int parse_events__modifier_list(struct parse_events_state *parse_state,
 		int eH = group ? evsel->core.attr.exclude_host : 0;
 		int eG = group ? evsel->core.attr.exclude_guest : 0;
 		int exclude = eu | ek | eh;
-		int exclude_GH = group ? evsel->exclude_GH : 0;
+		int exclude_GH = eG | eH;
 
 		if (mod.user) {
 			if (!exclude)
 				exclude = eu = ek = eh = 1;
-			if (!exclude_GH && !perf_guest && exclude_GH_default)
-				eG = 1;
 			eu = 0;
 		}
 		if (mod.kernel) {
@@ -1858,6 +1856,13 @@ static int parse_events__modifier_list(struct parse_events_state *parse_state,
 				exclude_GH = eG = eH = 1;
 			eH = 0;
 		}
+		if (!exclude_GH && exclude_GH_default) {
+			if (perf_host)
+				eG = 1;
+			else if (perf_guest)
+				eH = 1;
+		}
+
 		evsel->core.attr.exclude_user   = eu;
 		evsel->core.attr.exclude_kernel = ek;
 		evsel->core.attr.exclude_hv     = eh;
Signature · 1 line
2.39.5
author maintainer
#326 lore raw

PATCH 316/627 clk: xilinx: vcu: unregister pll_post only if registered correctly

In reply to Greg Kroah-Hartman

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

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

From: Rohit Visavalia <[email protected]>

[ Upstream commit 3b0abc443ac22f7d4f61ddbbbbc5dbb06c87139d ]

If registration of pll_post is failed, it will be set to NULL or ERR, unregistering same will fail with following call trace:

Unable to handle kernel NULL pointer dereference at virtual address 008
pc : clk_hw_unregister+0xc/0x20
lr : clk_hw_unregister_fixed_factor+0x18/0x30
sp : ffff800011923850
...
Call trace:
 clk_hw_unregister+0xc/0x20
 clk_hw_unregister_fixed_factor+0x18/0x30
 xvcu_unregister_clock_provider+0xcc/0xf4 [xlnx_vcu]
 xvcu_probe+0x2bc/0x53c [xlnx_vcu]
Fixes 4472e1849db7 ("soc: xilinx: vcu: make pll post divider explicit")
---
 drivers/clk/xilinx/xlnx_vcu.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/clk/xilinx/xlnx_vcu.c +2 -2
index 81501b48412e..88b3fd8250c2 100644
@@ -587,8 +587,8 @@ static void xvcu_unregister_clock_provider(struct xvcu_device *xvcu)
 		xvcu_clk_hw_unregister_leaf(hws[CLK_XVCU_ENC_MCU]);
 	if (!IS_ERR_OR_NULL(hws[CLK_XVCU_ENC_CORE]))
 		xvcu_clk_hw_unregister_leaf(hws[CLK_XVCU_ENC_CORE]);
-
-	clk_hw_unregister_fixed_factor(xvcu->pll_post);
+	if (!IS_ERR_OR_NULL(xvcu->pll_post))
+		clk_hw_unregister_fixed_factor(xvcu->pll_post);
 }
 
 /**
Signature · 1 line
2.39.5
author maintainer
#327 lore raw

PATCH 317/627 power: supply: cpcap-charger: Fix null check for power_supply_get_by_name

In reply to Greg Kroah-Hartman

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

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

From: Charles Han <[email protected]>

[ Upstream commit d9fa3aae08f99493e67fb79413c0e95d30fca5e9 ]

In the cpcap_usb_detect() function, the power_supply_get_by_name() function may return NULL instead of an error pointer. To prevent potential null pointer dereferences, Added a null check.

Fixes eab4e6d953c1 ("power: supply: cpcap-charger: get the battery inserted infomation from cpcap-battery")
---
 drivers/power/supply/cpcap-charger.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)
drivers/power/supply/cpcap-charger.c +2 -3
index 13300dc60baf..d0c3008db534 100644
@@ -689,9 +689,8 @@ static void cpcap_usb_detect(struct work_struct *work)
 		struct power_supply *battery;
 
 		battery = power_supply_get_by_name("battery");
-		if (IS_ERR_OR_NULL(battery)) {
-			dev_err(ddata->dev, "battery power_supply not available %li\n",
-					PTR_ERR(battery));
+		if (!battery) {
+			dev_err(ddata->dev, "battery power_supply not available\n");
 			return;
 		}
 
Signature · 1 line
2.39.5
author maintainer
#328 lore raw

PATCH 318/627 power: reset: POWER_RESET_TORADEX_EC should depend on ARCH_MXC

In reply to Greg Kroah-Hartman

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

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

From: Geert Uytterhoeven <[email protected]>

[ Upstream commit 22e4d29f081df8a10f1c062d3d952bb876eb9bdc ]

The Toradex Embedded Controller is currently only present on Toradex SMARC iMX8MP and iMX95 SoMs. Hence add a dependency on ARCH_MXC, to prevent asking the user about this driver when configuring a kernel without NXP i.MX SoC family support.

---
 drivers/power/reset/Kconfig | 1 +
 1 file changed, 1 insertion(+)
drivers/power/reset/Kconfig +1 -0
index e71f0af4e378..95f140ee7077 100644
@@ -218,6 +218,7 @@ config POWER_RESET_ST
 
 config POWER_RESET_TORADEX_EC
 	tristate "Toradex Embedded Controller power-off and reset driver"
+	depends on ARCH_MXC || COMPILE_TEST
 	depends on I2C
 	select REGMAP_I2C
 	help
Signature · 1 line
2.39.5
author maintainer
#329 lore raw

PATCH 319/627 power: supply: max14577: Handle NULL pdata when CONFIG_OF is not set

In reply to Greg Kroah-Hartman

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

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

From: Charles Han <[email protected]>

[ Upstream commit 2937f5d2e24eefef8cb126244caec7fe3307f724 ]

When the kernel is not configured CONFIG_OF, the max14577_charger_dt_init function returns NULL. Fix the max14577_charger_probe functionby returning -ENODATA instead of potentially passing a NULL pointer to PTR_ERR.

This fixes the below smatch warning:
max14577_charger_probe() warn: passing zero to 'PTR_ERR'

Fixes e30110e9c96f ("charger: max14577: Configure battery-dependent settings from DTS and sysfs")
---
 drivers/power/supply/max14577_charger.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/power/supply/max14577_charger.c +2 -2
index 1cef2f860b5f..63077d38ea30 100644
@@ -501,7 +501,7 @@ static struct max14577_charger_platform_data *max14577_charger_dt_init(
 static struct max14577_charger_platform_data *max14577_charger_dt_init(
 		struct platform_device *pdev)
 {
-	return NULL;
+	return ERR_PTR(-ENODATA);
 }
 #endif /* CONFIG_OF */
 
@@ -572,7 +572,7 @@ static int max14577_charger_probe(struct platform_device *pdev)
 	chg->max14577 = max14577;
 
 	chg->pdata = max14577_charger_dt_init(pdev);
-	if (IS_ERR_OR_NULL(chg->pdata))
+	if (IS_ERR(chg->pdata))
 		return PTR_ERR(chg->pdata);
 
 	ret = max14577_charger_reg_init(chg);
Signature · 1 line
2.39.5
author maintainer
#330 lore raw

PATCH 320/627 power: supply: qcom_pmi8998_charger: fix wakeirq

In reply to Greg Kroah-Hartman

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

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

From: Casey Connolly <[email protected]>

[ Upstream commit 6c5393771c50fac30f08dfb6d2f65f4f2cfeb8c7 ]

Unloading and reloading the driver (e.g. when built as a module) currently leads to errors trying to enable wake IRQ since it's already enabled.

Use devm to manage this for us so it correctly gets disabled when removing the driver.

Additionally, call device_init_wakeup() so that charger attach/remove will trigger a wakeup by default.

Fixes 8648aeb5d7b7 ("power: supply: add Qualcomm PMI8998 SMB2 Charger driver")
---
 drivers/power/supply/qcom_pmi8998_charger.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/power/supply/qcom_pmi8998_charger.c +3 -1
index c2f8f2e24398..cd3cb473c70d 100644
@@ -1016,7 +1016,9 @@ static int smb2_probe(struct platform_device *pdev)
 	if (rc < 0)
 		return rc;
 
-	rc = dev_pm_set_wake_irq(chip->dev, chip->cable_irq);
+	devm_device_init_wakeup(chip->dev);
+
+	rc = devm_pm_set_wake_irq(chip->dev, chip->cable_irq);
 	if (rc < 0)
 		return dev_err_probe(chip->dev, rc, "Couldn't set wake irq\n");
 
Signature · 1 line
2.39.5
author maintainer
#331 lore raw

PATCH 321/627 power: supply: max1720x correct capacity computation

In reply to Greg Kroah-Hartman

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

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

From: Thomas Antoine <[email protected]>

[ Upstream commit 58ae036172b5f051a19a32eba94a3e5eb37bf47e ]

Quoted 1 line

From the datasheet of the MAX17201/17205, the LSB should be "5.0μVh/RSENSE".

The current computation sets it at 0.5mAh=5.0μVh/10mOhm, which does not take into account the value of rsense (which is in 10µV steps) which can be different from 10mOhm.

Change the computation to fit the specs.

Fixes 479b6d04964b ("power: supply: add support for MAX1720x standalone fuel gauge")
---
 drivers/power/supply/max1720x_battery.c | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)
drivers/power/supply/max1720x_battery.c +6 -5
index ea3912fd1de8..68b5314ecf3a 100644
@@ -288,9 +288,10 @@ static int max172xx_voltage_to_ps(unsigned int reg)
 	return reg * 1250;	/* in uV */
 }
 
-static int max172xx_capacity_to_ps(unsigned int reg)
+static int max172xx_capacity_to_ps(unsigned int reg,
+				   struct max1720x_device_info *info)
 {
-	return reg * 500;	/* in uAh */
+	return reg * (500000 / info->rsense);	/* in uAh */
 }
 
 /*
@@ -394,11 +395,11 @@ static int max1720x_battery_get_property(struct power_supply *psy,
 		break;
 	case POWER_SUPPLY_PROP_CHARGE_FULL_DESIGN:
 		ret = regmap_read(info->regmap, MAX172XX_DESIGN_CAP, &reg_val);
-		val->intval = max172xx_capacity_to_ps(reg_val);
+		val->intval = max172xx_capacity_to_ps(reg_val, info);
 		break;
 	case POWER_SUPPLY_PROP_CHARGE_AVG:
 		ret = regmap_read(info->regmap, MAX172XX_REPCAP, &reg_val);
-		val->intval = max172xx_capacity_to_ps(reg_val);
+		val->intval = max172xx_capacity_to_ps(reg_val, info);
 		break;
 	case POWER_SUPPLY_PROP_TIME_TO_EMPTY_AVG:
 		ret = regmap_read(info->regmap, MAX172XX_TTE, &reg_val);
@@ -422,7 +423,7 @@ static int max1720x_battery_get_property(struct power_supply *psy,
 		break;
 	case POWER_SUPPLY_PROP_CHARGE_FULL:
 		ret = regmap_read(info->regmap, MAX172XX_FULL_CAP, &reg_val);
-		val->intval = max172xx_capacity_to_ps(reg_val);
+		val->intval = max172xx_capacity_to_ps(reg_val, info);
 		break;
 	case POWER_SUPPLY_PROP_MODEL_NAME:
 		ret = regmap_read(info->regmap, MAX172XX_DEV_NAME, &reg_val);
Signature · 1 line
2.39.5
author maintainer
#332 lore raw

PATCH 322/627 crypto: arm/aes-neonbs - work around gcc-15 warning

In reply to Greg Kroah-Hartman

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

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

From: Arnd Bergmann <[email protected]>

[ Upstream commit d5fa96dc5590915f060fee3209143313e4f5b03b ]

I get a very rare -Wstringop-overread warning with gcc-15 for one function in aesbs_ctr_encrypt():

arch/arm/crypto/aes-neonbs-glue.c: In function 'ctr_encrypt':
arch/arm/crypto/aes-neonbs-glue.c:212:1446: error: '__builtin_memcpy' offset [17, 2147483647] is out of the bounds [0, 16] of object 'buf' with type 'u8[16]' {aka 'unsigned char[16]'} [-Werror=array-bounds=]
212 | src = dst = memcpy(buf + sizeof(buf) - bytes,
arch/arm/crypto/aes-neonbs-glue.c: In function 'ctr_encrypt':
arch/arm/crypto/aes-neonbs-glue.c:218:17: error: 'aesbs_ctr_encrypt' reading 1 byte from a region of size 0 [-Werror=stringop-overread]

  218 |                 aesbs_ctr_encrypt(dst, src, ctx->rk, ctx->rounds, bytes, walk.iv);
      |                 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

arch/arm/crypto/aes-neonbs-glue.c:218:17: note: referencing argument 2 of type 'const u8[0]' {aka 'const unsigned char[]'} arch/arm/crypto/aes-neonbs-glue.c:218:17: note: referencing argument 3 of type 'const u8[0]' {aka 'const unsigned char[]'} arch/arm/crypto/aes-neonbs-glue.c:218:17: note: referencing argument 6 of type 'u8[0]' {aka 'unsigned char[]'} arch/arm/crypto/aes-neonbs-glue.c:36:17: note: in a call to function 'aesbs_ctr_encrypt'
36 | asmlinkage void aesbs_ctr_encrypt(u8 out[], u8 const in[], u8 const rk[],

This could happen in theory if walk.nbytes is larger than INT_MAX and gets converted to a negative local variable.

Keep the type unsigned like the orignal nbytes to be sure there is no integer overflow.

Fixes c8bf850e991a ("crypto: arm/aes-neonbs-ctr - deal with non-multiples of AES block size")
---
 arch/arm/crypto/aes-neonbs-glue.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm/crypto/aes-neonbs-glue.c +1 -1
index c60104dc1585..df5afe601e4a 100644
@@ -206,7 +206,7 @@ static int ctr_encrypt(struct skcipher_request *req)
 	while (walk.nbytes > 0) {
 		const u8 *src = walk.src.virt.addr;
 		u8 *dst = walk.dst.virt.addr;
-		int bytes = walk.nbytes;
+		unsigned int bytes = walk.nbytes;
 
 		if (unlikely(bytes < AES_BLOCK_SIZE))
 			src = dst = memcpy(buf + sizeof(buf) - bytes,
Signature · 1 line
2.39.5
author maintainer
#333 lore raw

PATCH 323/627 crypto: ahash - Add support for drivers with no fallback

In reply to Greg Kroah-Hartman

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

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

From: Herbert Xu <[email protected]>

[ Upstream commit 4ccd065a69df163cd9fe0dd8e0f609f1eeb4723d ]

Some drivers cannot have a fallback, e.g., because the key is held in hardware. Allow these to be used with ahash by adding the bit CRYPTO_ALG_NO_FALLBACK.

Stable-dep-of: 1e2b7fcd3f07 ("crypto: ahash - Stop legacy tfms from using the set_virt fallback path")

---
 crypto/ahash.c         | 10 +++++++++-
 include/linux/crypto.h |  3 +++
 2 files changed, 12 insertions(+), 1 deletion(-)
crypto/ahash.c +9 -1
index bc84a07c924c..3878b4da3cfd 100644
@@ -347,6 +347,9 @@ static int ahash_do_req_chain(struct ahash_request *req,
 	if (crypto_ahash_statesize(tfm) > HASH_MAX_STATESIZE)
 		return -ENOSYS;
 
+	if (!crypto_ahash_need_fallback(tfm))
+		return -ENOSYS;
+
 	{
 		u8 state[HASH_MAX_STATESIZE];
 
@@ -954,6 +957,10 @@ static int ahash_prepare_alg(struct ahash_alg *alg)
 	    base->cra_reqsize > MAX_SYNC_HASH_REQSIZE)
 		return -EINVAL;
 
+	if (base->cra_flags & CRYPTO_ALG_NEED_FALLBACK &&
+	    base->cra_flags & CRYPTO_ALG_NO_FALLBACK)
+		return -EINVAL;
+
 	err = hash_prepare_alg(&alg->halg);
 	if (err)
 		return err;
@@ -962,7 +969,8 @@ static int ahash_prepare_alg(struct ahash_alg *alg)
 	base->cra_flags |= CRYPTO_ALG_TYPE_AHASH;
 
 	if ((base->cra_flags ^ CRYPTO_ALG_REQ_VIRT) &
-	    (CRYPTO_ALG_ASYNC | CRYPTO_ALG_REQ_VIRT))
+	    (CRYPTO_ALG_ASYNC | CRYPTO_ALG_REQ_VIRT) &&
+	    !(base->cra_flags & CRYPTO_ALG_NO_FALLBACK))
 		base->cra_flags |= CRYPTO_ALG_NEED_FALLBACK;
 
 	if (!alg->setkey)
include/linux/crypto.h +3 -0
index b50f1954d1bb..a2137e19be7d 100644
@@ -136,6 +136,9 @@
 /* Set if the algorithm supports virtual addresses. */
 #define CRYPTO_ALG_REQ_VIRT		0x00040000
 
+/* Set if the algorithm cannot have a fallback (e.g., phmac). */
+#define CRYPTO_ALG_NO_FALLBACK		0x00080000
+
 /* The high bits 0xff000000 are reserved for type-specific flags. */
 
 /*
Signature · 1 line
2.39.5
author maintainer
#334 lore raw

PATCH 324/627 crypto: ahash - Stop legacy tfms from using the set_virt fallback path

In reply to Greg Kroah-Hartman

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

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

From: Herbert Xu <[email protected]>

[ Upstream commit 1e2b7fcd3f075ff8c5b0e4474fe145d1c685f54f ]

Ensure that drivers that have not been converted to the ahash API do not use the ahash_request_set_virt fallback path as they cannot use the software fallback.

Fixes 9d7a0ab1c753 ("crypto: ahash - Handle partial blocks in API")
---
 crypto/ahash.c                 | 3 +++
 include/crypto/internal/hash.h | 6 ++++++
 2 files changed, 9 insertions(+)
crypto/ahash.c +3 -0
index 3878b4da3cfd..2f06e6b4f601 100644
@@ -350,6 +350,9 @@ static int ahash_do_req_chain(struct ahash_request *req,
 	if (!crypto_ahash_need_fallback(tfm))
 		return -ENOSYS;
 
+	if (crypto_hash_no_export_core(tfm))
+		return -ENOSYS;
+
 	{
 		u8 state[HASH_MAX_STATESIZE];
 
include/crypto/internal/hash.h +6 -0
index 0f85c543f80b..f052afa6e7b0 100644
@@ -91,6 +91,12 @@ static inline bool crypto_hash_alg_needs_key(struct hash_alg_common *alg)
 		!(alg->base.cra_flags & CRYPTO_ALG_OPTIONAL_KEY);
 }
 
+static inline bool crypto_hash_no_export_core(struct crypto_ahash *tfm)
+{
+	return crypto_hash_alg_common(tfm)->base.cra_flags &
+	       CRYPTO_AHASH_ALG_NO_EXPORT_CORE;
+}
+
 int crypto_grab_ahash(struct crypto_ahash_spawn *spawn,
 		      struct crypto_instance *inst,
 		      const char *name, u32 type, u32 mask);
Signature · 1 line
2.39.5
author maintainer
#335 lore raw

PATCH 325/627 crypto: qat - restore ASYM service support for GEN6 devices

In reply to Greg Kroah-Hartman

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

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

From: Suman Kumar Chakraborty <[email protected]>

[ Upstream commit 4e55a929ff4d973206879a997a47a188353b3cd6 ]

Support for asymmetric crypto services was not included in the qat_6xxx by explicitly setting the asymmetric capabilities to 0 to allow for additional testing.

Enable asymmetric crypto services on QAT GEN6 devices by setting the appropriate capability flags.

Fixes 17fd7514ae68 ("crypto: qat - add qat_6xxx driver")
---
 drivers/crypto/intel/qat/qat_6xxx/adf_6xxx_hw_data.c | 10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)
drivers/crypto/intel/qat/qat_6xxx/adf_6xxx_hw_data.c +9 -1
index 359a6447ccb8..185a7ab92b7b 100644
@@ -627,7 +627,15 @@ static u32 get_accel_cap(struct adf_accel_dev *accel_dev)
 		capabilities_sym &= ~ICP_ACCEL_CAPABILITIES_CIPHER;
 	}
 
-	capabilities_asym = 0;
+	capabilities_asym = ICP_ACCEL_CAPABILITIES_CRYPTO_ASYMMETRIC |
+			    ICP_ACCEL_CAPABILITIES_SM2 |
+			    ICP_ACCEL_CAPABILITIES_ECEDMONT;
+
+	if (fusectl1 & ICP_ACCEL_GEN6_MASK_PKE_SLICE) {
+		capabilities_asym &= ~ICP_ACCEL_CAPABILITIES_CRYPTO_ASYMMETRIC;
+		capabilities_asym &= ~ICP_ACCEL_CAPABILITIES_SM2;
+		capabilities_asym &= ~ICP_ACCEL_CAPABILITIES_ECEDMONT;
+	}
 
 	capabilities_dc = ICP_ACCEL_CAPABILITIES_COMPRESSION |
 			  ICP_ACCEL_CAPABILITIES_LZ4_COMPRESSION |
Signature · 1 line
2.39.5
author maintainer
#336 lore raw

PATCH 326/627 PCI: endpoint: pci-epf-vntb: Return -ENOENT if pci_epc_get_next_free_bar() fails

In reply to Greg Kroah-Hartman

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

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

From: Jerome Brunet <[email protected]>

[ Upstream commit 7ea488cce73263231662e426639dd3e836537068 ]

According the function documentation of epf_ntb_init_epc_bar(), the function should return an error code on error. However, it returns -1 when no BAR is available i.e., when pci_epc_get_next_free_bar() fails.

Return -ENOENT instead.

Fixes e35f56bb0330 ("PCI: endpoint: Support NTB transfer between RC and EP")
[mani: changed err code to -ENOENT]
---
 drivers/pci/endpoint/functions/pci-epf-vntb.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/pci/endpoint/functions/pci-epf-vntb.c +1 -1
index e4da3fdb0007..30c6c563335a 100644
@@ -680,7 +680,7 @@ static int epf_ntb_init_epc_bar(struct epf_ntb *ntb)
 		barno = pci_epc_get_next_free_bar(epc_features, barno);
 		if (barno < 0) {
 			dev_err(dev, "Fail to get NTB function BAR\n");
-			return barno;
+			return -ENOENT;
 		}
 		ntb->epf_ntb_bar[bar] = barno;
 	}
Signature · 1 line
2.39.5
author maintainer
#337 lore raw

PATCH 327/627 pinctrl: sunxi: Fix memory leak on krealloc failure

In reply to Greg Kroah-Hartman

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

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

From: Yuan Chen <[email protected]>

[ Upstream commit e3507c56cbb208d4f160942748c527ef6a528ba1 ]

In sunxi_pctrl_dt_node_to_map(), when krealloc() fails to resize the pinctrl_map array, the function returns -ENOMEM directly without freeing the previously allocated *map buffer. This results in a memory leak of the original kmalloc_array allocation.

Fixes e11dee2e98f8 ("pinctrl: sunxi: Deal with configless pins")
---
 drivers/pinctrl/sunxi/pinctrl-sunxi.c | 11 ++++++++---
 1 file changed, 8 insertions(+), 3 deletions(-)
drivers/pinctrl/sunxi/pinctrl-sunxi.c +8 -3
index bf8612d72daa..d63859a2a64e 100644
@@ -408,6 +408,7 @@ static int sunxi_pctrl_dt_node_to_map(struct pinctrl_dev *pctldev,
 	const char *function, *pin_prop;
 	const char *group;
 	int ret, npins, nmaps, configlen = 0, i = 0;
+	struct pinctrl_map *new_map;
 
 	*map = NULL;
 	*num_maps = 0;
@@ -482,9 +483,13 @@ static int sunxi_pctrl_dt_node_to_map(struct pinctrl_dev *pctldev,
 	 * We know have the number of maps we need, we can resize our
 	 * map array
 	 */
-	*map = krealloc(*map, i * sizeof(struct pinctrl_map), GFP_KERNEL);
-	if (!*map)
-		return -ENOMEM;
+	new_map = krealloc(*map, i * sizeof(struct pinctrl_map), GFP_KERNEL);
+	if (!new_map) {
+		ret = -ENOMEM;
+		goto err_free_map;
+	}
+
+	*map = new_map;
 
 	return 0;
 
Signature · 1 line
2.39.5
author maintainer
#338 lore raw

PATCH 328/627 pinctrl: berlin: fix memory leak in berlin_pinctrl_build_state()

In reply to Greg Kroah-Hartman

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

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

From: Yuan Chen <[email protected]>

[ Upstream commit 8f6f303551100291bf2c1e1ccc66b758fffb1168 ]

In the original implementation, krealloc() failure handling incorrectly assigned the original memory pointer to NULL after kfree(), causing a memory leak when reallocation failed.

Fixes de845036f997 ("pinctrl: berlin: fix error return code of berlin_pinctrl_build_state()")
---
 drivers/pinctrl/berlin/berlin.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)
drivers/pinctrl/berlin/berlin.c +6 -2
index c372a2a24be4..9dc2da8056b7 100644
@@ -204,6 +204,7 @@ static int berlin_pinctrl_build_state(struct platform_device *pdev)
 	const struct berlin_desc_group *desc_group;
 	const struct berlin_desc_function *desc_function;
 	int i, max_functions = 0;
+	struct pinfunction *new_functions;
 
 	pctrl->nfunctions = 0;
 
@@ -229,12 +230,15 @@ static int berlin_pinctrl_build_state(struct platform_device *pdev)
 		}
 	}
 
-	pctrl->functions = krealloc(pctrl->functions,
+	new_functions = krealloc(pctrl->functions,
 				    pctrl->nfunctions * sizeof(*pctrl->functions),
 				    GFP_KERNEL);
-	if (!pctrl->functions)
+	if (!new_functions) {
+		kfree(pctrl->functions);
 		return -ENOMEM;
+	}
 
+	pctrl->functions = new_functions;
 	/* map functions to theirs groups */
 	for (i = 0; i < pctrl->desc->ngroups; i++) {
 		desc_group = pctrl->desc->groups + i;
Signature · 1 line
2.39.5
author maintainer
#339 lore raw

PATCH 329/627 pinctrl: canaan: k230: add NULL check in DT parse

In reply to Greg Kroah-Hartman

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

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

From: Ze Huang <[email protected]>

[ Upstream commit 65bd0be486390fc12a84eafaad78758c5e5a55e6 ]

Add a NULL check for the return value of of_get_property() when retrieving the "pinmux" property in the group parser. This avoids a potential NULL pointer dereference if the property is missing from the device tree node.

Also fix a typo ("sintenel") in the device ID match table comment, correcting it to "sentinel".

Fixes 545887eab6f6 ("pinctrl: canaan: Add support for k230 SoC")
---
 drivers/pinctrl/pinctrl-k230.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/pinctrl/pinctrl-k230.c +5 -1
index a9b4627b46b0..4976308e6237 100644
@@ -477,6 +477,10 @@ static int k230_pinctrl_parse_groups(struct device_node *np,
 	grp->name = np->name;
 
 	list = of_get_property(np, "pinmux", &size);
+	if (!list) {
+		dev_err(dev, "failed to get pinmux property\n");
+		return -EINVAL;
+	}
 	size /= sizeof(*list);
 
 	grp->num_pins = size;
@@ -623,7 +627,7 @@ static int k230_pinctrl_probe(struct platform_device *pdev)
 
 static const struct of_device_id k230_dt_ids[] = {
 	{ .compatible = "canaan,k230-pinctrl", },
-	{ /* sintenel */ }
+	{ /* sentinel */ }
 };
 MODULE_DEVICE_TABLE(of, k230_dt_ids);
 
Signature · 1 line
2.39.5
author maintainer
#340 lore raw

PATCH 330/627 pinctrl: canaan: k230: Fix order of DT parse and pinctrl register

In reply to Greg Kroah-Hartman

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

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

From: Ze Huang <[email protected]>

[ Upstream commit d94a32ac688f953dc9a9f12b5b4139ecad841bbb ]

Move DT parse before pinctrl register. This ensures that device tree parsing is done before calling devm_pinctrl_register() to prevent using uninitialized pin resources.

Fixes 545887eab6f6 ("pinctrl: canaan: Add support for k230 SoC")
---
 drivers/pinctrl/pinctrl-k230.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)
drivers/pinctrl/pinctrl-k230.c +5 -2
index 4976308e6237..d716f23d837f 100644
@@ -590,6 +590,7 @@ static int k230_pinctrl_probe(struct platform_device *pdev)
 	struct device *dev = &pdev->dev;
 	struct k230_pinctrl *info;
 	struct pinctrl_desc *pctl;
+	int ret;
 
 	info = devm_kzalloc(dev, sizeof(*info), GFP_KERNEL);
 	if (!info)
@@ -615,13 +616,15 @@ static int k230_pinctrl_probe(struct platform_device *pdev)
 		return dev_err_probe(dev, PTR_ERR(info->regmap_base),
 				     "failed to init regmap\n");
 
+	ret = k230_pinctrl_parse_dt(pdev, info);
+	if (ret)
+		return ret;
+
 	info->pctl_dev = devm_pinctrl_register(dev, pctl, info);
 	if (IS_ERR(info->pctl_dev))
 		return dev_err_probe(dev, PTR_ERR(info->pctl_dev),
 				     "devm_pinctrl_register failed\n");
 
-	k230_pinctrl_parse_dt(pdev, info);
-
 	return 0;
 }
 
Signature · 1 line
2.39.5
author maintainer
#341 lore raw

PATCH 331/627 PCI: Adjust the position of reading the Link Control 2 register

In reply to Greg Kroah-Hartman

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

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

From: Jiwei Sun <[email protected]>

[ Upstream commit b85af48de3ece4e5bbdb2248a5360a409991cf67 ]

In a89c82249c37 ("PCI: Work around PCIe link training failures"), if the speed limit is set to 2.5 GT/s and the retraining is successful, an attempt will be made to lift the speed limit. One condition for lifting the speed limit is to check whether the link speed field of the Link Control 2 register is PCI_EXP_LNKCTL2_TLS_2_5GT.

However, since de9a6c8d5dbf ("PCI/bwctrl: Add pcie_set_target_speed() to set PCIe Link Speed"), the lnkctl2 local variable does not undergo any changes during the speed limit setting and retraining process. As a result, the code intended to lift the speed limit is not executed.

To address this issue, adjust the position of the Link Control 2 register read operation in the code and place it before its use.

Fixes de9a6c8d5dbf ("PCI/bwctrl: Add pcie_set_target_speed() to set PCIe Link Speed")
---
 drivers/pci/quirks.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/pci/quirks.c +4 -2
index d7f4ee634263..db6e142b082d 100644
@@ -105,13 +105,13 @@ int pcie_failed_link_retrain(struct pci_dev *dev)
 	    !pcie_cap_has_lnkctl2(dev) || !dev->link_active_reporting)
 		return ret;
 
-	pcie_capability_read_word(dev, PCI_EXP_LNKCTL2, &lnkctl2);
 	pcie_capability_read_word(dev, PCI_EXP_LNKSTA, &lnksta);
 	if (!(lnksta & PCI_EXP_LNKSTA_DLLLA) && pcie_lbms_seen(dev, lnksta)) {
-		u16 oldlnkctl2 = lnkctl2;
+		u16 oldlnkctl2;
 
 		pci_info(dev, "broken device, retraining non-functional downstream link at 2.5GT/s\n");
 
+		pcie_capability_read_word(dev, PCI_EXP_LNKCTL2, &oldlnkctl2);
 		ret = pcie_set_target_speed(dev, PCIE_SPEED_2_5GT, false);
 		if (ret) {
 			pci_info(dev, "retraining failed\n");
@@ -123,6 +123,8 @@ int pcie_failed_link_retrain(struct pci_dev *dev)
 		pcie_capability_read_word(dev, PCI_EXP_LNKSTA, &lnksta);
 	}
 
+	pcie_capability_read_word(dev, PCI_EXP_LNKCTL2, &lnkctl2);
+
 	if ((lnksta & PCI_EXP_LNKSTA_DLLLA) &&
 	    (lnkctl2 & PCI_EXP_LNKCTL2_TLS) == PCI_EXP_LNKCTL2_TLS_2_5GT &&
 	    pci_match_id(ids, dev)) {
Signature · 1 line
2.39.5
author maintainer
#342 lore raw

PATCH 332/627 PCI: Rename PCIE_RESET_CONFIG_DEVICE_WAIT_MS to PCIE_RESET_CONFIG_WAIT_MS

In reply to Greg Kroah-Hartman

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

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

From: Niklas Cassel <[email protected]>

[ Upstream commit 817f989700fddefa56e5e443e7d138018ca6709d ]

Rename PCIE_RESET_CONFIG_DEVICE_WAIT_MS to PCIE_RESET_CONFIG_WAIT_MS.

Stable-dep-of: c7eb9c5e1498 ("PCI: dw-rockchip: Wait PCIE_RESET_CONFIG_WAIT_MS after link-up IRQ")

---
 drivers/pci/controller/plda/pcie-starfive.c | 2 +-
 drivers/pci/pci.h                           | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)
drivers/pci/controller/plda/pcie-starfive.c +1 -1
index e73c1b7bc8ef..3caf53c6c082 100644
@@ -368,7 +368,7 @@ static int starfive_pcie_host_init(struct plda_pcie_rp *plda)
 	 * of 100ms following exit from a conventional reset before
 	 * sending a configuration request to the device.
 	 */
-	msleep(PCIE_RESET_CONFIG_DEVICE_WAIT_MS);
+	msleep(PCIE_RESET_CONFIG_WAIT_MS);
 
 	if (starfive_pcie_host_wait_for_link(pcie))
 		dev_info(dev, "port link down\n");
drivers/pci/pci.h +1 -1
index 12215ee72afb..98d6fccb383e 100644
@@ -61,7 +61,7 @@ struct pcie_tlp_log;
  *    completes before sending a Configuration Request to the device
  *    immediately below that Port."
  */
-#define PCIE_RESET_CONFIG_DEVICE_WAIT_MS	100
+#define PCIE_RESET_CONFIG_WAIT_MS	100
 
 /* Message Routing (r[2:0]); PCIe r6.0, sec 2.2.8 */
 #define PCIE_MSG_TYPE_R_RC	0
Signature · 1 line
2.39.5
author maintainer
#343 lore raw

PATCH 333/627 PCI: dw-rockchip: Wait PCIE_RESET_CONFIG_WAIT_MS after link-up IRQ

In reply to Greg Kroah-Hartman

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

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

From: Niklas Cassel <[email protected]>

[ Upstream commit c7eb9c5e1498882951b7583c56add0b77bfc162e ]

Per PCIe r6.0, sec 6.6.1, software must generally wait a minimum of 100ms (PCIE_RESET_CONFIG_WAIT_MS) after Link training completes before sending a Configuration Request.

Prior to ec9fd499b9c6 ("PCI: dw-rockchip: Don't wait for link since we can detect Link Up"), dw-rockchip used dw_pcie_wait_for_link(), which waited between 0 and 90ms after the link came up before we enumerate the bus, and this was apparently enough for most devices.

After ec9fd499b9c6, rockchip_pcie_rc_sys_irq_thread() started enumeration immediately when handling the link-up IRQ, and devices (e.g., Laszlo Fiat's PLEXTOR PX-256M8PeGN NVMe SSD) may not be ready to handle config requests yet.

Delay PCIE_RESET_CONFIG_WAIT_MS after the link-up IRQ before starting enumeration.

Fixes 0e898eb8df4e ("PCI: rockchip-dwc: Add Rockchip RK356X host controller driver")
Cc Laszlo Fiat <[email protected]>
---
 drivers/pci/controller/dwc/pcie-dw-rockchip.c | 1 +
 1 file changed, 1 insertion(+)
drivers/pci/controller/dwc/pcie-dw-rockchip.c +1 -0
index 93171a392879..108d30637920 100644
@@ -458,6 +458,7 @@ static irqreturn_t rockchip_pcie_rc_sys_irq_thread(int irq, void *arg)
 
 	if (reg & PCIE_RDLH_LINK_UP_CHGED) {
 		if (rockchip_pcie_link_up(pci)) {
+			msleep(PCIE_RESET_CONFIG_WAIT_MS);
 			dev_dbg(dev, "Received Link up event. Starting enumeration!\n");
 			/* Rescan the bus to enumerate endpoint devices */
 			pci_lock_rescan_remove();
Signature · 1 line
2.39.5
author maintainer
#344 lore raw

PATCH 334/627 PCI: qcom: Wait PCIE_RESET_CONFIG_WAIT_MS after link-up IRQ

In reply to Greg Kroah-Hartman

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

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

From: Niklas Cassel <[email protected]>

[ Upstream commit 15b6b243cc2b1017cf89e2477aa0b4e1a306a82a ]

Per PCIe r6.0, sec 6.6.1, software must generally wait a minimum of 100ms (PCIE_RESET_CONFIG_WAIT_MS) after Link training completes before sending a Configuration Request.

Prior to 36971d6c5a9a ("PCI: qcom: Don't wait for link if we can detect Link Up"), qcom used dw_pcie_wait_for_link(), which waited between 0 and 90ms after the link came up before we enumerate the bus, and this was apparently enough for most devices.

After 36971d6c5a9a, qcom_pcie_global_irq_thread() started enumeration immediately when handling the link-up IRQ, and devices (e.g., Laszlo Fiat's PLEXTOR PX-256M8PeGN NVMe SSD) may not be ready to handle config requests yet.

Delay PCIE_RESET_CONFIG_WAIT_MS after the link-up IRQ before starting enumeration.

---
 drivers/pci/controller/dwc/pcie-qcom.c | 1 +
 1 file changed, 1 insertion(+)
drivers/pci/controller/dwc/pcie-qcom.c +1 -0
index c789e3f85655..9b12f2f02042 100644
@@ -1564,6 +1564,7 @@ static irqreturn_t qcom_pcie_global_irq_thread(int irq, void *data)
 	writel_relaxed(status, pcie->parf + PARF_INT_ALL_CLEAR);
 
 	if (FIELD_GET(PARF_INT_ALL_LINK_UP, status)) {
+		msleep(PCIE_RESET_CONFIG_WAIT_MS);
 		dev_dbg(dev, "Received Link up event. Starting enumeration!\n");
 		/* Rescan the bus to enumerate endpoint devices */
 		pci_lock_rescan_remove();
Signature · 1 line
2.39.5
author maintainer
#345 lore raw

PATCH 335/627 soundwire: Correct some property names

In reply to Greg Kroah-Hartman

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

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

From: Charles Keepax <[email protected]>

[ Upstream commit ae6a0f5b8a5b0ca2e4bf1c0380267ad83aca8401 ]

The DisCo properties should be mipi-sdw-paging-supported and mipi-sdw-bank-delay-supported, with an 'ed' on the end. Correct the property names used in sdw_slave_read_prop().

The internal flag bank_delay_support is currently unimplemented, so that being read wrong does not currently affect anything. The two existing users for this helper and the paging_support flag rt1320-sdw.c and rt721-sdca-sdw.c both manually set the flag in their slave properties, thus are not affected by this bug either.

Fixes 56d4fe31af77 ("soundwire: Add MIPI DisCo property helpers")
---
 drivers/soundwire/mipi_disco.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/soundwire/mipi_disco.c +2 -2
index 65afb28ef8fa..c69b78cd0b62 100644
@@ -451,10 +451,10 @@ int sdw_slave_read_prop(struct sdw_slave *slave)
 			"mipi-sdw-highPHY-capable");
 
 	prop->paging_support = mipi_device_property_read_bool(dev,
-			"mipi-sdw-paging-support");
+			"mipi-sdw-paging-supported");
 
 	prop->bank_delay_support = mipi_device_property_read_bool(dev,
-			"mipi-sdw-bank-delay-support");
+			"mipi-sdw-bank-delay-supported");
 
 	device_property_read_u32(dev,
 			"mipi-sdw-port15-read-behavior", &prop->p15_behave);
Signature · 1 line
2.39.5
author maintainer
#346 lore raw

PATCH 336/627 dmaengine: mmp: Fix again Wvoid-pointer-to-enum-cast warning

In reply to Greg Kroah-Hartman

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

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

From: Krzysztof Kozlowski <[email protected]>

[ Upstream commit a0b1589b62e2fcfb112996e0f4d5593bd2edf069 ]

This was fixed and re-introduced. 'type' is an enum, thus cast of pointer on 64-bit compile test with W=1 causes:

mmp_tdma.c:644:9: error: cast to smaller integer type 'enum mmp_tdma_type' from 'const void *' [-Werror,-Wvoid-pointer-to-enum-cast]

Fixes a67ba97dfb30 ("dmaengine: Use device_get_match_data()")
---
 drivers/dma/mmp_tdma.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/dma/mmp_tdma.c +1 -1
index c8dc504510f1..b7fb843c67a6 100644
@@ -641,7 +641,7 @@ static int mmp_tdma_probe(struct platform_device *pdev)
 	int chan_num = TDMA_CHANNEL_NUM;
 	struct gen_pool *pool = NULL;
 
-	type = (enum mmp_tdma_type)device_get_match_data(&pdev->dev);
+	type = (kernel_ulong_t)device_get_match_data(&pdev->dev);
 
 	/* always have couple channels */
 	tdev = devm_kzalloc(&pdev->dev, sizeof(*tdev), GFP_KERNEL);
Signature · 1 line
2.39.5
author maintainer
#347 lore raw

PATCH 337/627 soundwire: debugfs: move debug statement outside of error handling

In reply to Greg Kroah-Hartman

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

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

From: Rodrigo Gobbi <[email protected]>

[ Upstream commit 06f77ff9d852c9f2764659ea81489364d8a69a9c ]

The start_t and finish_t variables are not properly initialized if errors happens over request_firmware actions.
This was also detected by smatch:

drivers/soundwire/debugfs.c:301 cmd_go() error: uninitialized symbol 'finish_t'. drivers/soundwire/debugfs.c:301 cmd_go() error: uninitialized symbol 'start_t'.

Move the debug statement outside of firmware error handling.

---
 drivers/soundwire/debugfs.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/soundwire/debugfs.c +3 -3
index 3099ea074f10..230a51489486 100644
@@ -291,6 +291,9 @@ static int cmd_go(void *data, u64 value)
 
 	finish_t = ktime_get();
 
+	dev_dbg(&slave->dev, "command completed, num_byte %zu status %d, time %lld ms\n",
+		num_bytes, ret, div_u64(finish_t - start_t, NSEC_PER_MSEC));
+
 out:
 	if (fw)
 		release_firmware(fw);
@@ -298,9 +301,6 @@ static int cmd_go(void *data, u64 value)
 	pm_runtime_mark_last_busy(&slave->dev);
 	pm_runtime_put(&slave->dev);
 
-	dev_dbg(&slave->dev, "command completed, num_byte %zu status %d, time %lld ms\n",
-		num_bytes, ret, div_u64(finish_t - start_t, NSEC_PER_MSEC));
-
 	return ret;
 }
 DEFINE_DEBUGFS_ATTRIBUTE(cmd_go_fops, NULL,
Signature · 1 line
2.39.5
author maintainer
#348 lore raw

PATCH 338/627 phy: qualcomm: phy-qcom-eusb2-repeater: Dont zero-out registers

In reply to Greg Kroah-Hartman

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

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

From: Luca Weiss <[email protected]>

[ Upstream commit 31bc94de76026c527f82c238f414539a14f0f3e6 ]

Zeroing out registers does not happen in the downstream kernel, and will "tune" the repeater in surely unexpected ways since most registers don't have a reset value of 0x0.

Stop doing that and instead just set the registers that are in the init sequence (though long term I don't think there's actually PMIC-specific init sequences, there's board specific tuning, but that's a story for another day).

Fixes 99a517a582fc ("phy: qualcomm: phy-qcom-eusb2-repeater: Zero out untouched tuning regs")
---
 .../phy/qualcomm/phy-qcom-eusb2-repeater.c    | 87 +++++++------------
 1 file changed, 32 insertions(+), 55 deletions(-)
drivers/phy/qualcomm/phy-qcom-eusb2-repeater.c +32 -55
index 6bd1b3c75c77..d7493c2294ef 100644
@@ -37,32 +37,13 @@
 #define EUSB2_TUNE_EUSB_EQU		0x5A
 #define EUSB2_TUNE_EUSB_HS_COMP_CUR	0x5B
 
-enum eusb2_reg_layout {
-	TUNE_EUSB_HS_COMP_CUR,
-	TUNE_EUSB_EQU,
-	TUNE_EUSB_SLEW,
-	TUNE_USB2_HS_COMP_CUR,
-	TUNE_USB2_PREEM,
-	TUNE_USB2_EQU,
-	TUNE_USB2_SLEW,
-	TUNE_SQUELCH_U,
-	TUNE_HSDISC,
-	TUNE_RES_FSDIF,
-	TUNE_IUSB2,
-	TUNE_USB2_CROSSOVER,
-	NUM_TUNE_FIELDS,
-
-	FORCE_VAL_5 = NUM_TUNE_FIELDS,
-	FORCE_EN_5,
-
-	EN_CTL1,
-
-	RPTR_STATUS,
-	LAYOUT_SIZE,
+struct eusb2_repeater_init_tbl_reg {
+	unsigned int reg;
+	unsigned int value;
 };
 
 struct eusb2_repeater_cfg {
-	const u32 *init_tbl;
+	const struct eusb2_repeater_init_tbl_reg *init_tbl;
 	int init_tbl_num;
 	const char * const *vreg_list;
 	int num_vregs;
@@ -82,16 +63,16 @@ static const char * const pm8550b_vreg_l[] = {
 	"vdd18", "vdd3",
 };
 
-static const u32 pm8550b_init_tbl[NUM_TUNE_FIELDS] = {
-	[TUNE_IUSB2] = 0x8,
-	[TUNE_SQUELCH_U] = 0x3,
-	[TUNE_USB2_PREEM] = 0x5,
+static const struct eusb2_repeater_init_tbl_reg pm8550b_init_tbl[] = {
+	{ EUSB2_TUNE_IUSB2, 0x8 },
+	{ EUSB2_TUNE_SQUELCH_U, 0x3 },
+	{ EUSB2_TUNE_USB2_PREEM, 0x5 },
 };
 
-static const u32 smb2360_init_tbl[NUM_TUNE_FIELDS] = {
-	[TUNE_IUSB2] = 0x5,
-	[TUNE_SQUELCH_U] = 0x3,
-	[TUNE_USB2_PREEM] = 0x2,
+static const struct eusb2_repeater_init_tbl_reg smb2360_init_tbl[] = {
+	{ EUSB2_TUNE_IUSB2, 0x5 },
+	{ EUSB2_TUNE_SQUELCH_U, 0x3 },
+	{ EUSB2_TUNE_USB2_PREEM, 0x2 },
 };
 
 static const struct eusb2_repeater_cfg pm8550b_eusb2_cfg = {
@@ -129,17 +110,10 @@ static int eusb2_repeater_init(struct phy *phy)
 	struct eusb2_repeater *rptr = phy_get_drvdata(phy);
 	struct device_node *np = rptr->dev->of_node;
 	struct regmap *regmap = rptr->regmap;
-	const u32 *init_tbl = rptr->cfg->init_tbl;
-	u8 tune_usb2_preem = init_tbl[TUNE_USB2_PREEM];
-	u8 tune_hsdisc = init_tbl[TUNE_HSDISC];
-	u8 tune_iusb2 = init_tbl[TUNE_IUSB2];
 	u32 base = rptr->base;
-	u32 val;
+	u32 poll_val;
 	int ret;
-
-	of_property_read_u8(np, "qcom,tune-usb2-amplitude", &tune_iusb2);
-	of_property_read_u8(np, "qcom,tune-usb2-disc-thres", &tune_hsdisc);
-	of_property_read_u8(np, "qcom,tune-usb2-preem", &tune_usb2_preem);
+	u8 val;
 
 	ret = regulator_bulk_enable(rptr->cfg->num_vregs, rptr->vregs);
 	if (ret)
@@ -147,21 +121,24 @@ static int eusb2_repeater_init(struct phy *phy)
 
 	regmap_write(regmap, base + EUSB2_EN_CTL1, EUSB2_RPTR_EN);
 
-	regmap_write(regmap, base + EUSB2_TUNE_EUSB_HS_COMP_CUR, init_tbl[TUNE_EUSB_HS_COMP_CUR]);
-	regmap_write(regmap, base + EUSB2_TUNE_EUSB_EQU, init_tbl[TUNE_EUSB_EQU]);
-	regmap_write(regmap, base + EUSB2_TUNE_EUSB_SLEW, init_tbl[TUNE_EUSB_SLEW]);
-	regmap_write(regmap, base + EUSB2_TUNE_USB2_HS_COMP_CUR, init_tbl[TUNE_USB2_HS_COMP_CUR]);
-	regmap_write(regmap, base + EUSB2_TUNE_USB2_EQU, init_tbl[TUNE_USB2_EQU]);
-	regmap_write(regmap, base + EUSB2_TUNE_USB2_SLEW, init_tbl[TUNE_USB2_SLEW]);
-	regmap_write(regmap, base + EUSB2_TUNE_SQUELCH_U, init_tbl[TUNE_SQUELCH_U]);
-	regmap_write(regmap, base + EUSB2_TUNE_RES_FSDIF, init_tbl[TUNE_RES_FSDIF]);
-	regmap_write(regmap, base + EUSB2_TUNE_USB2_CROSSOVER, init_tbl[TUNE_USB2_CROSSOVER]);
-
-	regmap_write(regmap, base + EUSB2_TUNE_USB2_PREEM, tune_usb2_preem);
-	regmap_write(regmap, base + EUSB2_TUNE_HSDISC, tune_hsdisc);
-	regmap_write(regmap, base + EUSB2_TUNE_IUSB2, tune_iusb2);
-
-	ret = regmap_read_poll_timeout(regmap, base + EUSB2_RPTR_STATUS, val, val & RPTR_OK, 10, 5);
+	/* Write registers from init table */
+	for (int i = 0; i < rptr->cfg->init_tbl_num; i++)
+		regmap_write(regmap, base + rptr->cfg->init_tbl[i].reg,
+			     rptr->cfg->init_tbl[i].value);
+
+	/* Override registers from devicetree values */
+	if (!of_property_read_u8(np, "qcom,tune-usb2-amplitude", &val))
+		regmap_write(regmap, base + EUSB2_TUNE_USB2_PREEM, val);
+
+	if (!of_property_read_u8(np, "qcom,tune-usb2-disc-thres", &val))
+		regmap_write(regmap, base + EUSB2_TUNE_HSDISC, val);
+
+	if (!of_property_read_u8(np, "qcom,tune-usb2-preem", &val))
+		regmap_write(regmap, base + EUSB2_TUNE_IUSB2, val);
+
+	/* Wait for status OK */
+	ret = regmap_read_poll_timeout(regmap, base + EUSB2_RPTR_STATUS, poll_val,
+				       poll_val & RPTR_OK, 10, 5);
 	if (ret)
 		dev_err(rptr->dev, "initialization timed-out\n");
 
Signature · 1 line
2.39.5
author maintainer
#349 lore raw

PATCH 339/627 fanotify: sanitize handle_type values when reporting fid

In reply to Greg Kroah-Hartman

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

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

From: Amir Goldstein <[email protected]>

[ Upstream commit 8631e01c2c5d1fe6705bcc0d733a0b7a17d3daac ]

Unlike file_handle, type and len of struct fanotify_fh are u8. Traditionally, filesystem return handle_type < 0xff, but there is no enforecement for that in vfs.

Add a sanity check in fanotify to avoid truncating handle_type if its value is > 0xff.

Fixes 7cdafe6cc4a6 ("exportfs: check for error return value from exportfs_encode_*()")
---
 fs/notify/fanotify/fanotify.c | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)
fs/notify/fanotify/fanotify.c +7 -1
index 3083643b864b..bfe884d624e7 100644
@@ -454,7 +454,13 @@ static int fanotify_encode_fh(struct fanotify_fh *fh, struct inode *inode,
 	dwords = fh_len >> 2;
 	type = exportfs_encode_fid(inode, buf, &dwords);
 	err = -EINVAL;
-	if (type <= 0 || type == FILEID_INVALID || fh_len != dwords << 2)
+	/*
+	 * Unlike file_handle, type and len of struct fanotify_fh are u8.
+	 * Traditionally, filesystem return handle_type < 0xff, but there
+	 * is no enforecement for that in vfs.
+	 */
+	BUILD_BUG_ON(MAX_HANDLE_SZ > 0xff || FILEID_INVALID > 0xff);
+	if (type <= 0 || type >= FILEID_INVALID || fh_len != dwords << 2)
 		goto out_err;
 
 	fh->type = type;
Signature · 1 line
2.39.5
author maintainer
#350 lore raw

PATCH 340/627 clk: clk-axi-clkgen: fix fpfd_max frequency for zynq

In reply to Greg Kroah-Hartman

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

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

From: Nuno Sá <[email protected]>

[ Upstream commit ce8a9096699500e2c5bca09dde27b16edda5f636 ]

The fpfd_max frequency should be set to 450 MHz instead of 300 MHz. Well, it actually depends on the platform speed grade but we are being conservative for ultrascale so let's be consistent. In a following change we will set these limits at runtime.

---
 drivers/clk/clk-axi-clkgen.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/clk/clk-axi-clkgen.c +1 -1
index 934e53a96ddd..00bf799964c6 100644
@@ -118,7 +118,7 @@ static const struct axi_clkgen_limits axi_clkgen_zynqmp_default_limits = {
 
 static const struct axi_clkgen_limits axi_clkgen_zynq_default_limits = {
 	.fpfd_min = 10000,
-	.fpfd_max = 300000,
+	.fpfd_max = 450000,
 	.fvco_min = 600000,
 	.fvco_max = 1200000,
 };
Signature · 1 line
2.39.5
author maintainer
#351 lore raw

PATCH 341/627 RDMA/ipoib: Use parent rdma device net namespace

In reply to Greg Kroah-Hartman

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

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

From: Mark Bloch <[email protected]>

[ Upstream commit f1208b05574f63c52e88109d8c75afdf4fc6bf42 ]

Use the net namespace of the underlying rdma device. After honoring the rdma device's namespace, the ipoib netdev now also runs in the same net namespace of the rdma device.

Add an API to read the net namespace of the rdma device so that ULP such as IPoIB can use it to initialize its netdev.

Stable-dep-of: f458ccd2aa2c ("RDMA/uverbs: Check CAP_NET_RAW in user namespace for flow create")

---
 drivers/infiniband/ulp/ipoib/ipoib_main.c | 2 ++
 include/rdma/ib_verbs.h                   | 5 +++++
 2 files changed, 7 insertions(+)
drivers/infiniband/ulp/ipoib/ipoib_main.c +2 -0
index f2f5465f2a90..7acafc5c0e09 100644
@@ -2577,6 +2577,8 @@ static struct net_device *ipoib_add_port(const char *format,
 
 	ndev->rtnl_link_ops = ipoib_get_link_ops();
 
+	dev_net_set(ndev, rdma_dev_net(hca));
+
 	result = register_netdev(ndev);
 	if (result) {
 		pr_warn("%s: couldn't register ipoib port %d; error %d\n",
include/rdma/ib_verbs.h +5 -0
index af43a8d2a74a..c83e5a375cd6 100644
@@ -4855,6 +4855,11 @@ static inline int ibdev_to_node(struct ib_device *ibdev)
 bool rdma_dev_access_netns(const struct ib_device *device,
 			   const struct net *net);
 
+static inline struct net *rdma_dev_net(struct ib_device *device)
+{
+	return read_pnet(&device->coredev.rdma_net);
+}
+
 #define IB_ROCE_UDP_ENCAP_VALID_PORT_MIN (0xC000)
 #define IB_ROCE_UDP_ENCAP_VALID_PORT_MAX (0xFFFF)
 #define IB_GRH_FLOWLABEL_MASK (0x000FFFFF)
Signature · 1 line
2.39.5
author maintainer
#352 lore raw

PATCH 342/627 RDMA/uverbs: Check CAP_NET_RAW in user namespace for flow create

In reply to Greg Kroah-Hartman

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

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

From: Parav Pandit <[email protected]>

[ Upstream commit f458ccd2aa2c5a6f0129a9b1548f2825071fdc6b ]

Currently, the capability check is done in the default init_user_ns user namespace. When a process runs in a non default user namespace, such check fails. Due to this when a process is running using Podman, it fails to create the flow resource.

Since the RDMA device is a resource within a network namespace, use the network namespace associated with the RDMA device to determine its owning user namespace.

---
 drivers/infiniband/core/device.c     | 27 ++++++++++++++++++++++++++
 drivers/infiniband/core/rdma_core.c  | 29 ++++++++++++++++++++++++++++
 drivers/infiniband/core/uverbs_cmd.c |  2 +-
 include/rdma/ib_verbs.h              |  3 +++
 4 files changed, 60 insertions(+), 1 deletion(-)
drivers/infiniband/core/device.c +27 -0
index d4263385850a..792824e0ab2c 100644
@@ -145,6 +145,33 @@ bool rdma_dev_access_netns(const struct ib_device *dev, const struct net *net)
 }
 EXPORT_SYMBOL(rdma_dev_access_netns);
 
+/**
+ * rdma_dev_has_raw_cap() - Returns whether a specified rdma device has
+ *			    CAP_NET_RAW capability or not.
+ *
+ * @dev:	Pointer to rdma device whose capability to be checked
+ *
+ * Returns true if a rdma device's owning user namespace has CAP_NET_RAW
+ * capability, otherwise false. When rdma subsystem is in legacy shared network,
+ * namespace mode, the default net namespace is considered.
+ */
+bool rdma_dev_has_raw_cap(const struct ib_device *dev)
+{
+	const struct net *net;
+
+	/* Network namespace is the resource whose user namespace
+	 * to be considered. When in shared mode, there is no reliable
+	 * network namespace resource, so consider the default net namespace.
+	 */
+	if (ib_devices_shared_netns)
+		net = &init_net;
+	else
+		net = read_pnet(&dev->coredev.rdma_net);
+
+	return ns_capable(net->user_ns, CAP_NET_RAW);
+}
+EXPORT_SYMBOL(rdma_dev_has_raw_cap);
+
 /*
  * xarray has this behavior where it won't iterate over NULL values stored in
  * allocated arrays.  So we need our own iterator to see all values stored in
drivers/infiniband/core/rdma_core.c +29 -0
index 90c177edf9b0..18918f463361 100644
@@ -1019,3 +1019,32 @@ void uverbs_finalize_object(struct ib_uobject *uobj,
 		WARN_ON(true);
 	}
 }
+
+/**
+ * rdma_uattrs_has_raw_cap() - Returns whether a rdma device linked to the
+ *			       uverbs attributes file has CAP_NET_RAW
+ *			       capability or not.
+ *
+ * @attrs:       Pointer to uverbs attributes
+ *
+ * Returns true if a rdma device's owning user namespace has CAP_NET_RAW
+ * capability, otherwise false.
+ */
+bool rdma_uattrs_has_raw_cap(const struct uverbs_attr_bundle *attrs)
+{
+	struct ib_uverbs_file *ufile = attrs->ufile;
+	struct ib_ucontext *ucontext;
+	bool has_cap = false;
+	int srcu_key;
+
+	srcu_key = srcu_read_lock(&ufile->device->disassociate_srcu);
+	ucontext = ib_uverbs_get_ucontext_file(ufile);
+	if (IS_ERR(ucontext))
+		goto out;
+	has_cap = rdma_dev_has_raw_cap(ucontext->device);
+
+out:
+	srcu_read_unlock(&ufile->device->disassociate_srcu, srcu_key);
+	return has_cap;
+}
+EXPORT_SYMBOL(rdma_uattrs_has_raw_cap);
drivers/infiniband/core/uverbs_cmd.c +1 -1
index bc9fe3ceca4d..6700c2c66167 100644
@@ -3225,7 +3225,7 @@ static int ib_uverbs_ex_create_flow(struct uverbs_attr_bundle *attrs)
 	if (cmd.comp_mask)
 		return -EINVAL;
 
-	if (!capable(CAP_NET_RAW))
+	if (!rdma_uattrs_has_raw_cap(attrs))
 		return -EPERM;
 
 	if (cmd.flow_attr.flags >= IB_FLOW_ATTR_FLAGS_RESERVED)
include/rdma/ib_verbs.h +3 -0
index c83e5a375cd6..087048b75d13 100644
@@ -4801,6 +4801,8 @@ static inline int uverbs_destroy_def_handler(struct uverbs_attr_bundle *attrs)
 }
 #endif
 
+bool rdma_uattrs_has_raw_cap(const struct uverbs_attr_bundle *attrs);
+
 struct net_device *rdma_alloc_netdev(struct ib_device *device, u32 port_num,
 				     enum rdma_netdev_t type, const char *name,
 				     unsigned char name_assign_type,
@@ -4855,6 +4857,7 @@ static inline int ibdev_to_node(struct ib_device *ibdev)
 bool rdma_dev_access_netns(const struct ib_device *device,
 			   const struct net *net);
 
+bool rdma_dev_has_raw_cap(const struct ib_device *dev);
 static inline struct net *rdma_dev_net(struct ib_device *device)
 {
 	return read_pnet(&device->coredev.rdma_net);
Signature · 1 line
2.39.5
author maintainer
#353 lore raw

PATCH 343/627 RDMA/mlx5: Check CAP_NET_RAW in user namespace for flow create

In reply to Greg Kroah-Hartman

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

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

From: Parav Pandit <[email protected]>

[ Upstream commit 95a89ec304c38f7447cdbf271f2d1cbad4c3bf81 ]

Currently, the capability check is done in the default init_user_ns user namespace. When a process runs in a non default user namespace, such check fails. Due to this when a process is running using Podman, it fails to create the flow.

Since the RDMA device is a resource within a network namespace, use the network namespace associated with the RDMA device to determine its owning user namespace.

---
 drivers/infiniband/hw/mlx5/fs.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/infiniband/hw/mlx5/fs.c +1 -1
index 680627f1de33..de8114ba9c1f 100644
@@ -2458,7 +2458,7 @@ static int UVERBS_HANDLER(MLX5_IB_METHOD_CREATE_FLOW)(
 	struct mlx5_ib_dev *dev;
 	u32 flags;
 
-	if (!capable(CAP_NET_RAW))
+	if (!rdma_uattrs_has_raw_cap(attrs))
 		return -EPERM;
 
 	fs_matcher = uverbs_attr_get_obj(attrs,
Signature · 1 line
2.39.5
author maintainer
#354 lore raw

PATCH 344/627 RDMA/mlx5: Check CAP_NET_RAW in user namespace for anchor create

In reply to Greg Kroah-Hartman

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

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

From: Parav Pandit <[email protected]>

[ Upstream commit 14957e8125e767bfd40a3ac61b1d6b8e62ee0a98 ]

Currently, the capability check is done in the default init_user_ns user namespace. When a process runs in a non default user namespace, such check fails. Due to this when a process is running using Podman, it fails to create the anchor.

Since the RDMA device is a resource within a network namespace, use the network namespace associated with the RDMA device to determine its owning user namespace.

---
 drivers/infiniband/hw/mlx5/fs.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/infiniband/hw/mlx5/fs.c +1 -1
index de8114ba9c1f..eabc37f2ac19 100644
@@ -2989,7 +2989,7 @@ static int UVERBS_HANDLER(MLX5_IB_METHOD_STEERING_ANCHOR_CREATE)(
 	u32 ft_id;
 	int err;
 
-	if (!capable(CAP_NET_RAW))
+	if (!rdma_dev_has_raw_cap(&dev->ib_dev))
 		return -EPERM;
 
 	err = uverbs_get_const(&ib_uapi_ft_type, attrs,
Signature · 1 line
2.39.5
author maintainer
#355 lore raw

PATCH 345/627 RDMA/uverbs: Check CAP_NET_RAW in user namespace for QP create

In reply to Greg Kroah-Hartman

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

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

From: Parav Pandit <[email protected]>

[ Upstream commit 0498c2d9984ed2ad75b1cd5ba6abfa1226742df5 ]

Currently, the capability check is done in the default init_user_ns user namespace. When a process runs in a non default user namespace, such check fails. Due to this when a process is running using Podman, it fails to create the QP.

Since the RDMA device is a resource within a network namespace, use the network namespace associated with the RDMA device to determine its owning user namespace.

---
 drivers/infiniband/core/uverbs_cmd.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/infiniband/core/uverbs_cmd.c +1 -1
index 6700c2c66167..4d96e4a678f3 100644
@@ -1451,7 +1451,7 @@ static int create_qp(struct uverbs_attr_bundle *attrs,
 	}
 
 	if (attr.create_flags & IB_QP_CREATE_SOURCE_QPN) {
-		if (!capable(CAP_NET_RAW)) {
+		if (!rdma_uattrs_has_raw_cap(attrs)) {
 			ret = -EPERM;
 			goto err_put;
 		}
Signature · 1 line
2.39.5
author maintainer
#356 lore raw

PATCH 346/627 RDMA/uverbs: Check CAP_NET_RAW in user namespace for RAW QP create

In reply to Greg Kroah-Hartman

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

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

From: Parav Pandit <[email protected]>

[ Upstream commit a6dca091ba7646ff5304af660c94fa51b6696476 ]

Currently, the capability check is done in the default init_user_ns user namespace. When a process runs in a non default user namespace, such check fails. Due to this when a process is running using Podman, it fails to create the QP.

Since the RDMA device is a resource within a network namespace, use the network namespace associated with the RDMA device to determine its owning user namespace.

---
 drivers/infiniband/core/uverbs_std_types_qp.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/infiniband/core/uverbs_std_types_qp.c +1 -1
index 7b4773fa4bc0..be0730e8509e 100644
@@ -133,7 +133,7 @@ static int UVERBS_HANDLER(UVERBS_METHOD_QP_CREATE)(
 		device = xrcd->device;
 		break;
 	case IB_UVERBS_QPT_RAW_PACKET:
-		if (!capable(CAP_NET_RAW))
+		if (!rdma_uattrs_has_raw_cap(attrs))
 			return -EPERM;
 		fallthrough;
 	case IB_UVERBS_QPT_RC:
Signature · 1 line
2.39.5
author maintainer
#357 lore raw

PATCH 347/627 leds: tps6131x: Add V4L2_FLASH_LED_CLASS dependency

In reply to Greg Kroah-Hartman

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

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

From: Arnd Bergmann <[email protected]>

[ Upstream commit c3c38e80016548685e439b23999b4f0bd0ad7e05 ]

This driver can optionally use the v4l2_flash infrastructure, but fails to link built=in if that is in a loadable module:

ld.lld-21: error: undefined symbol: v4l2_flash_release

Quoted 1 line
Quoted 1 line
Quoted 1 line

referenced by leds-tps6131x.c:792 (drivers/leds/flash/leds-tps6131x.c:792)

Add the usual Kconfig dependency for it, still allowing it to be built when CONFIG_V4L2_FLASH_LED_CLASS is disabled.

Fixes b338a2ae9b31 ("leds: tps6131x: Add support for Texas Instruments TPS6131X flash LED driver")
---
 drivers/leds/flash/Kconfig | 1 +
 1 file changed, 1 insertion(+)
drivers/leds/flash/Kconfig +1 -0
index 55ca663ca506..5e08102a6784 100644
@@ -136,6 +136,7 @@ config LEDS_TPS6131X
 	tristate "LED support for TI TPS6131x flash LED driver"
 	depends on I2C && OF
 	depends on GPIOLIB
+	depends on V4L2_FLASH_LED_CLASS || !V4L2_FLASH_LED_CLASS
 	select REGMAP_I2C
 	help
 	  This option enables support for Texas Instruments TPS61310/TPS61311
Signature · 1 line
2.39.5
author maintainer
#358 lore raw

PATCH 348/627 RDMA/mlx5: Check CAP_NET_RAW in user namespace for devx create

In reply to Greg Kroah-Hartman

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

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

From: Parav Pandit <[email protected]>

[ Upstream commit bd82467f17e0940c6f6a5396278cda586c9cb6fd ]

Currently, the capability check is done in the default init_user_ns user namespace. When a process runs in a non default user namespace, such check fails. Due to this when a process is running using Podman, it fails to create the devx object.

Since the RDMA device is a resource within a network namespace, use the network namespace associated with the RDMA device to determine its owning user namespace.

---
 drivers/infiniband/hw/mlx5/devx.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/infiniband/hw/mlx5/devx.c +1 -1
index 843dcd312242..c369fee33562 100644
@@ -159,7 +159,7 @@ int mlx5_ib_devx_create(struct mlx5_ib_dev *dev, bool is_user, u64 req_ucaps)
 	uctx = MLX5_ADDR_OF(create_uctx_in, in, uctx);
 	if (is_user &&
 	    (MLX5_CAP_GEN(dev->mdev, uctx_cap) & MLX5_UCTX_CAP_RAW_TX) &&
-	    capable(CAP_NET_RAW))
+	    rdma_dev_has_raw_cap(&dev->ib_dev))
 		cap |= MLX5_UCTX_CAP_RAW_TX;
 	if (is_user &&
 	    (MLX5_CAP_GEN(dev->mdev, uctx_cap) &
Signature · 1 line
2.39.5
author maintainer
#359 lore raw

PATCH 349/627 RDMA/nldev: Check CAP_NET_RAW in user namespace for QP modify

In reply to Greg Kroah-Hartman

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

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

From: Parav Pandit <[email protected]>

[ Upstream commit 28ea058a2979f063d4b756c5d82d885fc16f5ca2 ]

Currently, the capability check is done in the default init_user_ns user namespace. When a process runs in a non default user namespace, such check fails. Due to this when a process is running using Podman, it fails to modify the QP.

Since the RDMA device is a resource within a network namespace, use the network namespace associated with the RDMA device to determine its owning user namespace.

---
 drivers/infiniband/core/nldev.c      | 2 +-
 drivers/infiniband/core/uverbs_cmd.c | 3 ++-
 2 files changed, 3 insertions(+), 2 deletions(-)
drivers/infiniband/core/nldev.c +1 -1
index a872643e8039..be6b2ef0ede4 100644
@@ -255,7 +255,7 @@ EXPORT_SYMBOL(rdma_nl_put_driver_u64_hex);
 
 bool rdma_nl_get_privileged_qkey(void)
 {
-	return privileged_qkey || capable(CAP_NET_RAW);
+	return privileged_qkey;
 }
 EXPORT_SYMBOL(rdma_nl_get_privileged_qkey);
 
drivers/infiniband/core/uverbs_cmd.c +2 -1
index 4d96e4a678f3..0807e9a00008 100644
@@ -1877,7 +1877,8 @@ static int modify_qp(struct uverbs_attr_bundle *attrs,
 		attr->path_mig_state = cmd->base.path_mig_state;
 	if (cmd->base.attr_mask & IB_QP_QKEY) {
 		if (cmd->base.qkey & IB_QP_SET_QKEY &&
-		    !rdma_nl_get_privileged_qkey()) {
+		    !(rdma_nl_get_privileged_qkey() ||
+		      rdma_uattrs_has_raw_cap(attrs))) {
 			ret = -EPERM;
 			goto release_qp;
 		}
Signature · 1 line
2.39.5
author maintainer
#360 lore raw

PATCH 350/627 RDMA/counter: Check CAP_NET_RAW check in user namespace for RDMA counters

In reply to Greg Kroah-Hartman

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

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

From: Parav Pandit <[email protected]>

[ Upstream commit 449728196d65fce513dbacf4d3696764be1c6524 ]

Currently, the capability check is done in the default init_user_ns user namespace. When a process runs in a non default user namespace, such check fails.

Since the RDMA device is a resource within a network namespace, use the network namespace associated with the RDMA device to determine its owning user namespace.

---
 drivers/infiniband/core/counters.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/infiniband/core/counters.c +1 -1
index e6ec7b7a40af..c3aa6d7fc66b 100644
@@ -461,7 +461,7 @@ static struct ib_qp *rdma_counter_get_qp(struct ib_device *dev, u32 qp_num)
 		return NULL;
 
 	qp = container_of(res, struct ib_qp, res);
-	if (qp->qp_type == IB_QPT_RAW_PACKET && !capable(CAP_NET_RAW))
+	if (qp->qp_type == IB_QPT_RAW_PACKET && !rdma_dev_has_raw_cap(dev))
 		goto err;
 
 	return qp;
Signature · 1 line
2.39.5
author maintainer
#361 lore raw

PATCH 351/627 Fix dma_unmap_sg() nents value

In reply to Greg Kroah-Hartman

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

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

From: Thomas Fourier <[email protected]>

[ Upstream commit 1db50f7b7a793670adcf062df9ff27798829d963 ]

The dma_unmap_sg() functions should be called with the same nents as the dma_map_sg(), not the value the map function returned.

Fixes ed10435d3583 ("RDMA/erdma: Implement hierarchical MTT")
---
 drivers/infiniband/hw/erdma/erdma_verbs.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/infiniband/hw/erdma/erdma_verbs.c +2 -1
index af36a8d2df22..ec0ad4086066 100644
@@ -629,7 +629,8 @@ static struct erdma_mtt *erdma_create_cont_mtt(struct erdma_dev *dev,
 static void erdma_destroy_mtt_buf_sg(struct erdma_dev *dev,
 				     struct erdma_mtt *mtt)
 {
-	dma_unmap_sg(&dev->pdev->dev, mtt->sglist, mtt->nsg, DMA_TO_DEVICE);
+	dma_unmap_sg(&dev->pdev->dev, mtt->sglist,
+		     DIV_ROUND_UP(mtt->size, PAGE_SIZE), DMA_TO_DEVICE);
 	vfree(mtt->sglist);
 }
 
Signature · 1 line
2.39.5
author maintainer
#362 lore raw

PATCH 352/627 leds: pca955x: Avoid potential overflow when filling default_label (take 2)

In reply to Greg Kroah-Hartman

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

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

From: Andy Shevchenko <[email protected]>

[ Upstream commit 239afba8b9f3b0fcfd464d5ffeaed0ed4441c5a4 ]

GCC compiler v8.5.0 is not happy about printing into a too short buffer (when build with make W=1):

drivers/leds/leds-pca955x.c:696:5: note: 'snprintf' output between 2 and 11 bytes into a destination of size 8

Unfortunately this is a false positive from the old GCC versions, but we may still improve the code by using '%hhu' format specifier and reduce buffer size by 4 bytes.

---
 drivers/leds/leds-pca955x.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/leds/leds-pca955x.c +2 -2
index 42fe056b1c74..70d109246088 100644
@@ -587,7 +587,7 @@ static int pca955x_probe(struct i2c_client *client)
 	struct pca955x_platform_data *pdata;
 	bool keep_psc0 = false;
 	bool set_default_label = false;
-	char default_label[8];
+	char default_label[4];
 	int bit, err, reg;
 
 	chip = i2c_get_match_data(client);
@@ -693,7 +693,7 @@ static int pca955x_probe(struct i2c_client *client)
 			}
 
 			if (set_default_label) {
-				snprintf(default_label, sizeof(default_label), "%u", i);
+				snprintf(default_label, sizeof(default_label), "%hhu", i);
 				init_data.default_label = default_label;
 			} else {
 				init_data.default_label = NULL;
Signature · 1 line
2.39.5
author maintainer
#363 lore raw

PATCH 353/627 gitignore: allow .pylintrc to be tracked

In reply to Greg Kroah-Hartman

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

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

From: WangYuli <[email protected]>

[ Upstream commit 38d573a624a54ccde1384ead8af0780fe4005c2b ]

The .pylintrc file was introduced by commit 02df8e3b333c ("docs: add a .pylintrc file with sys path for docs scripts") to provide Python path configuration for documentation scripts. However, the generic ".*" rule in .gitignore causes this tracked file to be ignored, leading to warnings during kernel builds.

Add !.pylintrc to the exception list to explicitly allow this configuration file to be tracked by git, consistent with other development tool configuration files like .clang-format and .rustfmt.toml.

This resolves the build warning:
.pylintrc: warning: ignored by one of the .gitignore files

Fixes 02df8e3b333c ("docs: add a .pylintrc file with sys path for docs scripts")
---
 .gitignore | 1 +
 1 file changed, 1 insertion(+)
.gitignore +1 -0
index bf5ee6e01cd4..929054df5212 100644
@@ -114,6 +114,7 @@ modules.order
 !.gitignore
 !.kunitconfig
 !.mailmap
+!.pylintrc
 !.rustfmt.toml
 
 #
Signature · 1 line
2.39.5
author maintainer
#364 lore raw

PATCH 354/627 perf tools: Fix use-after-free in help_unknown_cmd()

In reply to Greg Kroah-Hartman

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

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

From: Namhyung Kim <[email protected]>

[ Upstream commit 1fdf938168c4d26fa279d4f204768690d1f9c4ae ]

Currently perf aborts when it finds an invalid command. I guess it depends on the environment as I have some custom commands in the path.

  $ perf bad-command
  perf: 'bad-command' is not a perf-command. See 'perf --help'.
  Aborted (core dumped)

It's because the exclude_cmds() in libsubcmd has a use-after-free when it removes some entries. After copying one to another entry, it keeps the pointer in the both position. And the next copy operation will free the later one but it's the same entry in the previous one.

For example, let's say cmds = { A, B, C, D, E } and excludes = { B, E }.

ci cj ei cmds-name excludes
-----------+--------------------

   0   0   0 |     A         B       :    cmp < 0, ci == cj
   1   1   0 |     B         B       :    cmp == 0
   2   1   1 |     C         E       :    cmp < 0, ci != cj

At this point, it frees cmds->names[1] and cmds->names[1] is assigned to cmds->names[2].

3 2 1 | D E : cmp < 0, ci != cj

Now it frees cmds->names[2] but it's the same as cmds->names[1]. So accessing cmds->names[1] will be invalid.

This makes the subcmd tests succeed.

  $ perf test subcmd
   69: libsubcmd help tests                                            :
   69.1: Load subcmd names                                             : Ok
   69.2: Uniquify subcmd names                                         : Ok
   69.3: Exclude duplicate subcmd names                                : Ok
Fixes 4b96679170c6 ("libsubcmd: Avoid SEGV/use-after-free when commands aren't excluded")
---
 tools/lib/subcmd/help.c | 12 +++++++-----
 1 file changed, 7 insertions(+), 5 deletions(-)
tools/lib/subcmd/help.c +7 -5
index 8561b0f01a24..9ef569492560 100644
@@ -9,6 +9,7 @@
 #include <sys/stat.h>
 #include <unistd.h>
 #include <dirent.h>
+#include <assert.h>
 #include "subcmd-util.h"
 #include "help.h"
 #include "exec-cmd.h"
@@ -82,10 +83,11 @@ void exclude_cmds(struct cmdnames *cmds, struct cmdnames *excludes)
 				ci++;
 				cj++;
 			} else {
-				zfree(&cmds->names[cj]);
-				cmds->names[cj++] = cmds->names[ci++];
+				cmds->names[cj++] = cmds->names[ci];
+				cmds->names[ci++] = NULL;
 			}
 		} else if (cmp == 0) {
+			zfree(&cmds->names[ci]);
 			ci++;
 			ei++;
 		} else if (cmp > 0) {
@@ -94,12 +96,12 @@ void exclude_cmds(struct cmdnames *cmds, struct cmdnames *excludes)
 	}
 	if (ci != cj) {
 		while (ci < cmds->cnt) {
-			zfree(&cmds->names[cj]);
-			cmds->names[cj++] = cmds->names[ci++];
+			cmds->names[cj++] = cmds->names[ci];
+			cmds->names[ci++] = NULL;
 		}
 	}
 	for (ci = cj; ci < cmds->cnt; ci++)
-		zfree(&cmds->names[ci]);
+		assert(cmds->names[ci] == NULL);
 	cmds->cnt = cj;
 }
 
Signature · 1 line
2.39.5
author maintainer
#365 lore raw

PATCH 355/627 perf dso: Add missed dso__put to dso__load_kcore

In reply to Greg Kroah-Hartman

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

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

From: Ian Rogers <[email protected]>

[ Upstream commit 63a088e999de3f431f87d9a367933da894ddb613 ]

The kcore loading creates a set of list nodes that have reference counted references to maps of the kcore. The list node freeing in the success path wasn't releasing the maps, add the missing puts. It is unclear why this leak was being missed by leak sanitizer.

Fixes 83720209961f ("perf map: Move map list node into symbol")
---
 tools/perf/util/symbol.c | 1 +
 1 file changed, 1 insertion(+)
tools/perf/util/symbol.c +1 -0
index 8b30c6f16a9e..fd4583718eab 100644
@@ -1422,6 +1422,7 @@ static int dso__load_kcore(struct dso *dso, struct map *map,
 				goto out_err;
 			}
 		}
+		map__zput(new_node->map);
 		free(new_node);
 	}
 
Signature · 1 line
2.39.5
author maintainer
#366 lore raw

PATCH 356/627 mtd: spi-nor: spansion: Fixup params->set_4byte_addr_mode for SEMPER

In reply to Greg Kroah-Hartman

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

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

From: Takahiro Kuwano <[email protected]>

[ Upstream commit a45ab839f52f3f00ac3dae18a50e902efd216de2 ]

Infineon SEMPER flash family does not support E9h opcode as Exit 4-byte mode (EX4B). Therefore, params->set_4byte_addr_mode is not determined by BFPT parse. Fixup it up by introducing vendor specific EX4B opcode (B8h) and function.

---
 drivers/mtd/spi-nor/spansion.c | 31 +++++++++++++++++++++++++++++++
 1 file changed, 31 insertions(+)
drivers/mtd/spi-nor/spansion.c +31 -0
index bf08dbf5e742..b9f156c0f8bc 100644
@@ -17,6 +17,7 @@
 
 #define SPINOR_OP_CLSR		0x30	/* Clear status register 1 */
 #define SPINOR_OP_CLPEF		0x82	/* Clear program/erase failure flags */
+#define SPINOR_OP_CYPRESS_EX4B	0xB8	/* Exit 4-byte address mode */
 #define SPINOR_OP_CYPRESS_DIE_ERASE		0x61	/* Chip (die) erase */
 #define SPINOR_OP_RD_ANY_REG			0x65	/* Read any register */
 #define SPINOR_OP_WR_ANY_REG			0x71	/* Write any register */
@@ -58,6 +59,13 @@
 		   SPI_MEM_OP_DUMMY(ndummy, 0),				\
 		   SPI_MEM_OP_DATA_IN(1, buf, 0))
 
+#define CYPRESS_NOR_EN4B_EX4B_OP(enable)				\
+	SPI_MEM_OP(SPI_MEM_OP_CMD(enable ? SPINOR_OP_EN4B :		\
+					   SPINOR_OP_CYPRESS_EX4B, 0),	\
+		   SPI_MEM_OP_NO_ADDR,					\
+		   SPI_MEM_OP_NO_DUMMY,					\
+		   SPI_MEM_OP_NO_DATA)
+
 #define SPANSION_OP(opcode)						\
 	SPI_MEM_OP(SPI_MEM_OP_CMD(opcode, 0),				\
 		   SPI_MEM_OP_NO_ADDR,					\
@@ -356,6 +364,20 @@ static int cypress_nor_quad_enable_volatile(struct spi_nor *nor)
 	return 0;
 }
 
+static int cypress_nor_set_4byte_addr_mode(struct spi_nor *nor, bool enable)
+{
+	int ret;
+	struct spi_mem_op op = CYPRESS_NOR_EN4B_EX4B_OP(enable);
+
+	spi_nor_spimem_setup_op(nor, &op, nor->reg_proto);
+
+	ret = spi_mem_exec_op(nor->spimem, &op);
+	if (ret)
+		dev_dbg(nor->dev, "error %d setting 4-byte mode\n", ret);
+
+	return ret;
+}
+
 /**
  * cypress_nor_determine_addr_mode_by_sr1() - Determine current address mode
  *                                            (3 or 4-byte) by querying status
@@ -526,6 +548,9 @@ s25fs256t_post_bfpt_fixup(struct spi_nor *nor,
 	struct spi_mem_op op;
 	int ret;
 
+	/* Assign 4-byte address mode method that is not determined in BFPT */
+	nor->params->set_4byte_addr_mode = cypress_nor_set_4byte_addr_mode;
+
 	ret = cypress_nor_set_addr_mode_nbytes(nor);
 	if (ret)
 		return ret;
@@ -591,6 +616,9 @@ s25hx_t_post_bfpt_fixup(struct spi_nor *nor,
 {
 	int ret;
 
+	/* Assign 4-byte address mode method that is not determined in BFPT */
+	nor->params->set_4byte_addr_mode = cypress_nor_set_4byte_addr_mode;
+
 	ret = cypress_nor_set_addr_mode_nbytes(nor);
 	if (ret)
 		return ret;
@@ -718,6 +746,9 @@ static int s28hx_t_post_bfpt_fixup(struct spi_nor *nor,
 				   const struct sfdp_parameter_header *bfpt_header,
 				   const struct sfdp_bfpt *bfpt)
 {
+	/* Assign 4-byte address mode method that is not determined in BFPT */
+	nor->params->set_4byte_addr_mode = cypress_nor_set_4byte_addr_mode;
+
 	return cypress_nor_set_addr_mode_nbytes(nor);
 }
 
Signature · 1 line
2.39.5
author maintainer
#367 lore raw

PATCH 357/627 perf sched: Make sure it frees the usage string

In reply to Greg Kroah-Hartman

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

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

From: Namhyung Kim <[email protected]>

[ Upstream commit 10d9b89203765fb776512742c13af8dd92821842 ]

The parse_options_subcommand() allocates the usage string based on the given subcommands. So it should reach the end of the function to free the string to prevent memory leaks.

Fixes 1a5efc9e13f357ab ("libsubcmd: Don't free the usage string")
---
 tools/perf/builtin-sched.c | 25 +++++++++++++------------
 1 file changed, 13 insertions(+), 12 deletions(-)
tools/perf/builtin-sched.c +13 -12
index 26ece6e9bfd1..b7bbfad0ed60 100644
@@ -3902,9 +3902,9 @@ int cmd_sched(int argc, const char **argv)
 	 * Aliased to 'perf script' for now:
 	 */
 	if (!strcmp(argv[0], "script")) {
-		return cmd_script(argc, argv);
+		ret = cmd_script(argc, argv);
 	} else if (strlen(argv[0]) > 2 && strstarts("record", argv[0])) {
-		return __cmd_record(argc, argv);
+		ret = __cmd_record(argc, argv);
 	} else if (strlen(argv[0]) > 2 && strstarts("latency", argv[0])) {
 		sched.tp_handler = &lat_ops;
 		if (argc > 1) {
@@ -3913,7 +3913,7 @@ int cmd_sched(int argc, const char **argv)
 				usage_with_options(latency_usage, latency_options);
 		}
 		setup_sorting(&sched, latency_options, latency_usage);
-		return perf_sched__lat(&sched);
+		ret = perf_sched__lat(&sched);
 	} else if (!strcmp(argv[0], "map")) {
 		if (argc) {
 			argc = parse_options(argc, argv, map_options, map_usage, 0);
@@ -3924,13 +3924,14 @@ int cmd_sched(int argc, const char **argv)
 				sched.map.task_names = strlist__new(sched.map.task_name, NULL);
 				if (sched.map.task_names == NULL) {
 					fprintf(stderr, "Failed to parse task names\n");
-					return -1;
+					ret = -1;
+					goto out;
 				}
 			}
 		}
 		sched.tp_handler = &map_ops;
 		setup_sorting(&sched, latency_options, latency_usage);
-		return perf_sched__map(&sched);
+		ret = perf_sched__map(&sched);
 	} else if (strlen(argv[0]) > 2 && strstarts("replay", argv[0])) {
 		sched.tp_handler = &replay_ops;
 		if (argc) {
@@ -3938,7 +3939,7 @@ int cmd_sched(int argc, const char **argv)
 			if (argc)
 				usage_with_options(replay_usage, replay_options);
 		}
-		return perf_sched__replay(&sched);
+		ret = perf_sched__replay(&sched);
 	} else if (!strcmp(argv[0], "timehist")) {
 		if (argc) {
 			argc = parse_options(argc, argv, timehist_options,
@@ -3954,19 +3955,19 @@ int cmd_sched(int argc, const char **argv)
 				parse_options_usage(NULL, timehist_options, "w", true);
 			if (sched.show_next)
 				parse_options_usage(NULL, timehist_options, "n", true);
-			return -EINVAL;
+			ret = -EINVAL;
+			goto out;
 		}
 		ret = symbol__validate_sym_arguments();
-		if (ret)
-			return ret;
-
-		return perf_sched__timehist(&sched);
+		if (!ret)
+			ret = perf_sched__timehist(&sched);
 	} else {
 		usage_with_options(sched_usage, sched_options);
 	}
 
+out:
 	/* free usage string allocated by parse_options_subcommand */
 	free((void *)sched_usage[0]);
 
-	return 0;
+	return ret;
 }
Signature · 1 line
2.39.5
author maintainer
#368 lore raw

PATCH 358/627 perf sched: Free thread->priv using priv_destructor

In reply to Greg Kroah-Hartman

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

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

From: Namhyung Kim <[email protected]>

[ Upstream commit aa9fdd106bab8c478d37eba5703c0950ad5c0d4f ]

In many perf sched subcommand saves priv data structure in the thread but it forgot to free them. As it's an opaque type with 'void *', it needs to register that knows how to free the data. In this case, just regular 'free()' is fine.

Fixes 04cb4fc4d40a5bf1 ("perf thread: Allow tools to register a thread->priv destructor")
---
 tools/perf/builtin-sched.c | 2 ++
 1 file changed, 2 insertions(+)
tools/perf/builtin-sched.c +2 -0
index b7bbfad0ed60..fa4052e04020 100644
@@ -3898,6 +3898,8 @@ int cmd_sched(int argc, const char **argv)
 	if (!argc)
 		usage_with_options(sched_usage, sched_options);
 
+	thread__set_priv_destructor(free);
+
 	/*
 	 * Aliased to 'perf script' for now:
 	 */
Signature · 1 line
2.39.5
author maintainer
#369 lore raw

PATCH 359/627 perf sched: Fix memory leaks in perf sched map

In reply to Greg Kroah-Hartman

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

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

From: Namhyung Kim <[email protected]>

[ Upstream commit dc3a80c98884d86389b3b572c50ccc7f502cd41b ]

It maintains per-cpu pointers for the current thread but it doesn't release the refcounts.

Fixes 5e895278697c014e ("perf sched: Move curr_thread initialization to perf_sched__map()")
---
 tools/perf/builtin-sched.c | 31 ++++++++++++++++++++-----------
 1 file changed, 20 insertions(+), 11 deletions(-)
tools/perf/builtin-sched.c +20 -11
index fa4052e04020..b73989fb6ace 100644
@@ -1634,6 +1634,7 @@ static int map_switch_event(struct perf_sched *sched, struct evsel *evsel,
 	const char *color = PERF_COLOR_NORMAL;
 	char stimestamp[32];
 	const char *str;
+	int ret = -1;
 
 	BUG_ON(this_cpu.cpu >= MAX_CPUS || this_cpu.cpu < 0);
 
@@ -1664,17 +1665,20 @@ static int map_switch_event(struct perf_sched *sched, struct evsel *evsel,
 	sched_in = map__findnew_thread(sched, machine, -1, next_pid);
 	sched_out = map__findnew_thread(sched, machine, -1, prev_pid);
 	if (sched_in == NULL || sched_out == NULL)
-		return -1;
+		goto out;
 
 	tr = thread__get_runtime(sched_in);
-	if (tr == NULL) {
-		thread__put(sched_in);
-		return -1;
-	}
+	if (tr == NULL)
+		goto out;
+
+	thread__put(sched->curr_thread[this_cpu.cpu]);
+	thread__put(sched->curr_out_thread[this_cpu.cpu]);
 
 	sched->curr_thread[this_cpu.cpu] = thread__get(sched_in);
 	sched->curr_out_thread[this_cpu.cpu] = thread__get(sched_out);
 
+	ret = 0;
+
 	str = thread__comm_str(sched_in);
 	new_shortname = 0;
 	if (!tr->shortname[0]) {
@@ -1769,12 +1773,10 @@ static int map_switch_event(struct perf_sched *sched, struct evsel *evsel,
 	color_fprintf(stdout, color, "\n");
 
 out:
-	if (sched->map.task_name)
-		thread__put(sched_out);
-
+	thread__put(sched_out);
 	thread__put(sched_in);
 
-	return 0;
+	return ret;
 }
 
 static int process_sched_switch_event(const struct perf_tool *tool,
@@ -3556,10 +3558,10 @@ static int perf_sched__map(struct perf_sched *sched)
 
 	sched->curr_out_thread = calloc(MAX_CPUS, sizeof(*(sched->curr_out_thread)));
 	if (!sched->curr_out_thread)
-		return rc;
+		goto out_free_curr_thread;
 
 	if (setup_cpus_switch_event(sched))
-		goto out_free_curr_thread;
+		goto out_free_curr_out_thread;
 
 	if (setup_map_cpus(sched))
 		goto out_free_cpus_switch_event;
@@ -3590,7 +3592,14 @@ static int perf_sched__map(struct perf_sched *sched)
 out_free_cpus_switch_event:
 	free_cpus_switch_event(sched);
 
+out_free_curr_out_thread:
+	for (int i = 0; i < MAX_CPUS; i++)
+		thread__put(sched->curr_out_thread[i]);
+	zfree(&sched->curr_out_thread);
+
 out_free_curr_thread:
+	for (int i = 0; i < MAX_CPUS; i++)
+		thread__put(sched->curr_thread[i]);
 	zfree(&sched->curr_thread);
 	return rc;
 }
Signature · 1 line
2.39.5
author maintainer
#370 lore raw

PATCH 360/627 perf sched: Fix thread leaks in perf sched timehist

In reply to Greg Kroah-Hartman

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

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

From: Namhyung Kim <[email protected]>

[ Upstream commit e2eb59260c4f6bac403491d0112891766b8650d1 ]

Add missing thread__put() after machine__findnew_thread() or timehist_get_thread(). Also idle threads' last_thread should be refcounted properly.

Fixes 699b5b920db04a6f ("perf sched timehist: Save callchain when entering idle")
---
 tools/perf/builtin-sched.c | 48 +++++++++++++++++++++++++++++---------
 1 file changed, 37 insertions(+), 11 deletions(-)
tools/perf/builtin-sched.c +37 -11
index b73989fb6ace..83b5a85a91b7 100644
@@ -2313,8 +2313,10 @@ static void save_task_callchain(struct perf_sched *sched,
 		return;
 	}
 
-	if (!sched->show_callchain || sample->callchain == NULL)
+	if (!sched->show_callchain || sample->callchain == NULL) {
+		thread__put(thread);
 		return;
+	}
 
 	cursor = get_tls_callchain_cursor();
 
@@ -2323,10 +2325,12 @@ static void save_task_callchain(struct perf_sched *sched,
 		if (verbose > 0)
 			pr_err("Failed to resolve callchain. Skipping\n");
 
+		thread__put(thread);
 		return;
 	}
 
 	callchain_cursor_commit(cursor);
+	thread__put(thread);
 
 	while (true) {
 		struct callchain_cursor_node *node;
@@ -2403,8 +2407,17 @@ static void free_idle_threads(void)
 		return;
 
 	for (i = 0; i < idle_max_cpu; ++i) {
-		if ((idle_threads[i]))
-			thread__delete(idle_threads[i]);
+		struct thread *idle = idle_threads[i];
+
+		if (idle) {
+			struct idle_thread_runtime *itr;
+
+			itr = thread__priv(idle);
+			if (itr)
+				thread__put(itr->last_thread);
+
+			thread__delete(idle);
+		}
 	}
 
 	free(idle_threads);
@@ -2441,7 +2454,7 @@ static struct thread *get_idle_thread(int cpu)
 		}
 	}
 
-	return idle_threads[cpu];
+	return thread__get(idle_threads[cpu]);
 }
 
 static void save_idle_callchain(struct perf_sched *sched,
@@ -2496,7 +2509,8 @@ static struct thread *timehist_get_thread(struct perf_sched *sched,
 			if (itr == NULL)
 				return NULL;
 
-			itr->last_thread = thread;
+			thread__put(itr->last_thread);
+			itr->last_thread = thread__get(thread);
 
 			/* copy task callchain when entering to idle */
 			if (evsel__intval(evsel, sample, "next_pid") == 0)
@@ -2567,6 +2581,7 @@ static void timehist_print_wakeup_event(struct perf_sched *sched,
 	/* show wakeup unless both awakee and awaker are filtered */
 	if (timehist_skip_sample(sched, thread, evsel, sample) &&
 	    timehist_skip_sample(sched, awakened, evsel, sample)) {
+		thread__put(thread);
 		return;
 	}
 
@@ -2583,6 +2598,8 @@ static void timehist_print_wakeup_event(struct perf_sched *sched,
 	printf("awakened: %s", timehist_get_commstr(awakened));
 
 	printf("\n");
+
+	thread__put(thread);
 }
 
 static int timehist_sched_wakeup_ignore(const struct perf_tool *tool __maybe_unused,
@@ -2611,8 +2628,10 @@ static int timehist_sched_wakeup_event(const struct perf_tool *tool,
 		return -1;
 
 	tr = thread__get_runtime(thread);
-	if (tr == NULL)
+	if (tr == NULL) {
+		thread__put(thread);
 		return -1;
+	}
 
 	if (tr->ready_to_run == 0)
 		tr->ready_to_run = sample->time;
@@ -2622,6 +2641,7 @@ static int timehist_sched_wakeup_event(const struct perf_tool *tool,
 	    !perf_time__skip_sample(&sched->ptime, sample->time))
 		timehist_print_wakeup_event(sched, evsel, sample, machine, thread);
 
+	thread__put(thread);
 	return 0;
 }
 
@@ -2649,6 +2669,7 @@ static void timehist_print_migration_event(struct perf_sched *sched,
 
 	if (timehist_skip_sample(sched, thread, evsel, sample) &&
 	    timehist_skip_sample(sched, migrated, evsel, sample)) {
+		thread__put(thread);
 		return;
 	}
 
@@ -2676,6 +2697,7 @@ static void timehist_print_migration_event(struct perf_sched *sched,
 	printf(" cpu %d => %d", ocpu, dcpu);
 
 	printf("\n");
+	thread__put(thread);
 }
 
 static int timehist_migrate_task_event(const struct perf_tool *tool,
@@ -2695,8 +2717,10 @@ static int timehist_migrate_task_event(const struct perf_tool *tool,
 		return -1;
 
 	tr = thread__get_runtime(thread);
-	if (tr == NULL)
+	if (tr == NULL) {
+		thread__put(thread);
 		return -1;
+	}
 
 	tr->migrations++;
 	tr->migrated = sample->time;
@@ -2706,6 +2730,7 @@ static int timehist_migrate_task_event(const struct perf_tool *tool,
 		timehist_print_migration_event(sched, evsel, sample,
 							machine, thread);
 	}
+	thread__put(thread);
 
 	return 0;
 }
@@ -2728,10 +2753,10 @@ static void timehist_update_task_prio(struct evsel *evsel,
 		return;
 
 	tr = thread__get_runtime(thread);
-	if (tr == NULL)
-		return;
+	if (tr != NULL)
+		tr->prio = next_prio;
 
-	tr->prio = next_prio;
+	thread__put(thread);
 }
 
 static int timehist_sched_change_event(const struct perf_tool *tool,
@@ -2743,7 +2768,7 @@ static int timehist_sched_change_event(const struct perf_tool *tool,
 	struct perf_sched *sched = container_of(tool, struct perf_sched, tool);
 	struct perf_time_interval *ptime = &sched->ptime;
 	struct addr_location al;
-	struct thread *thread;
+	struct thread *thread = NULL;
 	struct thread_runtime *tr = NULL;
 	u64 tprev, t = sample->time;
 	int rc = 0;
@@ -2867,6 +2892,7 @@ static int timehist_sched_change_event(const struct perf_tool *tool,
 
 	evsel__save_time(evsel, sample->time, sample->cpu);
 
+	thread__put(thread);
 	addr_location__exit(&al);
 	return rc;
 }
Signature · 1 line
2.39.5
author maintainer
#371 lore raw

PATCH 361/627 perf sched: Fix memory leaks for evsel->priv in timehist

In reply to Greg Kroah-Hartman

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

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

From: Namhyung Kim <[email protected]>

[ Upstream commit 117e5c33b1c44037af016d77ce6c0b086d55535f ]

It uses evsel->priv to save per-cpu timing information. It should be freed when the evsel is released.

Add the priv destructor for evsel same as thread to handle that.

Fixes 49394a2a24c78ce0 ("perf sched timehist: Introduce timehist command")
---
 tools/perf/builtin-sched.c | 12 ++++++++++++
 tools/perf/util/evsel.c    | 11 +++++++++++
 tools/perf/util/evsel.h    |  2 ++
 3 files changed, 25 insertions(+)
tools/perf/builtin-sched.c +12 -0
index 83b5a85a91b7..a6eb0462dd5b 100644
@@ -2020,6 +2020,16 @@ static u64 evsel__get_time(struct evsel *evsel, u32 cpu)
 	return r->last_time[cpu];
 }
 
+static void timehist__evsel_priv_destructor(void *priv)
+{
+	struct evsel_runtime *r = priv;
+
+	if (r) {
+		free(r->last_time);
+		free(r);
+	}
+}
+
 static int comm_width = 30;
 
 static char *timehist_get_commstr(struct thread *thread)
@@ -3314,6 +3324,8 @@ static int perf_sched__timehist(struct perf_sched *sched)
 
 	setup_pager();
 
+	evsel__set_priv_destructor(timehist__evsel_priv_destructor);
+
 	/* prefer sched_waking if it is captured */
 	if (evlist__find_tracepoint_by_name(session->evlist, "sched:sched_waking"))
 		handlers[1].handler = timehist_sched_wakeup_ignore;
tools/perf/util/evsel.c +11 -0
index d55482f094bf..1dc1f7b3bfb8 100644
@@ -1656,6 +1656,15 @@ static void evsel__free_config_terms(struct evsel *evsel)
 	free_config_terms(&evsel->config_terms);
 }
 
+static void (*evsel__priv_destructor)(void *priv);
+
+void evsel__set_priv_destructor(void (*destructor)(void *priv))
+{
+	assert(evsel__priv_destructor == NULL);
+
+	evsel__priv_destructor = destructor;
+}
+
 void evsel__exit(struct evsel *evsel)
 {
 	assert(list_empty(&evsel->core.node));
@@ -1686,6 +1695,8 @@ void evsel__exit(struct evsel *evsel)
 	hashmap__free(evsel->per_pkg_mask);
 	evsel->per_pkg_mask = NULL;
 	zfree(&evsel->metric_events);
+	if (evsel__priv_destructor)
+		evsel__priv_destructor(evsel->priv);
 	perf_evsel__object.fini(evsel);
 	if (evsel__tool_event(evsel) == TOOL_PMU__EVENT_SYSTEM_TIME ||
 	    evsel__tool_event(evsel) == TOOL_PMU__EVENT_USER_TIME)
tools/perf/util/evsel.h +2 -0
index 6dbc9690e0c9..b84ee274602d 100644
@@ -280,6 +280,8 @@ void evsel__init(struct evsel *evsel, struct perf_event_attr *attr, int idx);
 void evsel__exit(struct evsel *evsel);
 void evsel__delete(struct evsel *evsel);
 
+void evsel__set_priv_destructor(void (*destructor)(void *priv));
+
 struct callchain_param;
 
 void evsel__config(struct evsel *evsel, struct record_opts *opts,
Signature · 1 line
2.39.5
author maintainer
#372 lore raw

PATCH 362/627 perf sched: Use RC_CHK_EQUAL() to compare pointers

In reply to Greg Kroah-Hartman

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

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

From: Namhyung Kim <[email protected]>

[ Upstream commit 7a4002ec9e0fced907179da94f67c3082d7b4162 ]

So that it can check two pointers to the same object properly when REFCNT_CHECKING is on.

---
 tools/perf/builtin-sched.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
tools/perf/builtin-sched.c +1 -1
index a6eb0462dd5b..087d4eaba5f7 100644
@@ -994,7 +994,7 @@ thread_atoms_search(struct rb_root_cached *root, struct thread *thread,
 		else if (cmp < 0)
 			node = node->rb_right;
 		else {
-			BUG_ON(thread != atoms->thread);
+			BUG_ON(!RC_CHK_EQUAL(thread, atoms->thread));
 			return atoms;
 		}
 	}
Signature · 1 line
2.39.5
author maintainer
#373 lore raw

PATCH 363/627 perf sched: Fix memory leaks in perf sched latency

In reply to Greg Kroah-Hartman

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

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

From: Namhyung Kim <[email protected]>

[ Upstream commit e68b1c0098b959cb88afce5c93dd6a9324e6da78 ]

The work_atoms should be freed after use. Add free_work_atoms() to make sure to release all. It should use list_splice_init() when merging atoms to prevent accessing invalid pointers.

Fixes b1ffe8f3e0c96f552 ("perf sched: Finish latency => atom rename and misc cleanups")
---
 tools/perf/builtin-sched.c | 27 ++++++++++++++++++++++++---
 1 file changed, 24 insertions(+), 3 deletions(-)
tools/perf/builtin-sched.c +24 -3
index 087d4eaba5f7..4bbebd6ef2e4 100644
@@ -1111,6 +1111,21 @@ add_sched_in_event(struct work_atoms *atoms, u64 timestamp)
 	atoms->nb_atoms++;
 }
 
+static void free_work_atoms(struct work_atoms *atoms)
+{
+	struct work_atom *atom, *tmp;
+
+	if (atoms == NULL)
+		return;
+
+	list_for_each_entry_safe(atom, tmp, &atoms->work_list, list) {
+		list_del(&atom->list);
+		free(atom);
+	}
+	thread__zput(atoms->thread);
+	free(atoms);
+}
+
 static int latency_switch_event(struct perf_sched *sched,
 				struct evsel *evsel,
 				struct perf_sample *sample,
@@ -3426,13 +3441,13 @@ static void __merge_work_atoms(struct rb_root_cached *root, struct work_atoms *d
 			this->total_runtime += data->total_runtime;
 			this->nb_atoms += data->nb_atoms;
 			this->total_lat += data->total_lat;
-			list_splice(&data->work_list, &this->work_list);
+			list_splice_init(&data->work_list, &this->work_list);
 			if (this->max_lat < data->max_lat) {
 				this->max_lat = data->max_lat;
 				this->max_lat_start = data->max_lat_start;
 				this->max_lat_end = data->max_lat_end;
 			}
-			zfree(&data);
+			free_work_atoms(data);
 			return;
 		}
 	}
@@ -3511,7 +3526,6 @@ static int perf_sched__lat(struct perf_sched *sched)
 		work_list = rb_entry(next, struct work_atoms, node);
 		output_lat_thread(sched, work_list);
 		next = rb_next(next);
-		thread__zput(work_list->thread);
 	}
 
 	printf(" -----------------------------------------------------------------------------------------------------------------\n");
@@ -3525,6 +3539,13 @@ static int perf_sched__lat(struct perf_sched *sched)
 
 	rc = 0;
 
+	while ((next = rb_first_cached(&sched->sorted_atom_root))) {
+		struct work_atoms *data;
+
+		data = rb_entry(next, struct work_atoms, node);
+		rb_erase_cached(next, &sched->sorted_atom_root);
+		free_work_atoms(data);
+	}
 out_free_cpus_switch_event:
 	free_cpus_switch_event(sched);
 	return rc;
Signature · 1 line
2.39.5
author maintainer
#374 lore raw

PATCH 364/627 clk: spacemit: mark K1 pll1_d8 as critical

In reply to Greg Kroah-Hartman

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

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

From: Alex Elder <[email protected]>

[ Upstream commit 7554729de27daf6d54bcf8689d863bbe267828bf ]

The pll1_d8 clock is enabled by the boot loader, and is ultimately a parent for numerous clocks, including those used by APB and AXI buses. Guodong Xu discovered that this clock got disabled while responding to getting -EPROBE_DEFER when requesting a reset controller.

The needed clock (CLK_DMA, along with its parents) had already been enabled. To respond to the probe deferral return, the CLK_DMA clock was disabled, and this led to parent clocks also reducing their enable count. When the enable count for pll1_d8 was decremented it became 0, which caused it to be disabled. This led to a system hang.

Marking that clock critical resolves this by preventing it from being disabled.

Define a new macro CCU_FACTOR_GATE_DEFINE() to allow clock flags to be supplied for a CCU_FACTOR_GATE clock.

Fixes 1b72c59db0add ("clk: spacemit: Add clock support for SpacemiT K1 SoC")
---
 drivers/clk/spacemit/ccu-k1.c  |  3 ++-
 drivers/clk/spacemit/ccu_mix.h | 11 ++++++++---
 2 files changed, 10 insertions(+), 4 deletions(-)
drivers/clk/spacemit/ccu-k1.c +2 -1
index cdde37a05235..df65009a07bb 100644
@@ -170,7 +170,8 @@ CCU_FACTOR_GATE_DEFINE(pll1_d4, CCU_PARENT_HW(pll1), APBS_PLL1_SWCR2, BIT(3), 4,
 CCU_FACTOR_GATE_DEFINE(pll1_d5, CCU_PARENT_HW(pll1), APBS_PLL1_SWCR2, BIT(4), 5, 1);
 CCU_FACTOR_GATE_DEFINE(pll1_d6, CCU_PARENT_HW(pll1), APBS_PLL1_SWCR2, BIT(5), 6, 1);
 CCU_FACTOR_GATE_DEFINE(pll1_d7, CCU_PARENT_HW(pll1), APBS_PLL1_SWCR2, BIT(6), 7, 1);
-CCU_FACTOR_GATE_DEFINE(pll1_d8, CCU_PARENT_HW(pll1), APBS_PLL1_SWCR2, BIT(7), 8, 1);
+CCU_FACTOR_GATE_FLAGS_DEFINE(pll1_d8, CCU_PARENT_HW(pll1), APBS_PLL1_SWCR2, BIT(7), 8, 1,
+		CLK_IS_CRITICAL);
 CCU_FACTOR_GATE_DEFINE(pll1_d11_223p4, CCU_PARENT_HW(pll1), APBS_PLL1_SWCR2, BIT(15), 11, 1);
 CCU_FACTOR_GATE_DEFINE(pll1_d13_189, CCU_PARENT_HW(pll1), APBS_PLL1_SWCR2, BIT(16), 13, 1);
 CCU_FACTOR_GATE_DEFINE(pll1_d23_106p8, CCU_PARENT_HW(pll1), APBS_PLL1_SWCR2, BIT(20), 23, 1);
drivers/clk/spacemit/ccu_mix.h +8 -3
index 51d19f5d6aac..54d40cd39b27 100644
@@ -101,17 +101,22 @@ static struct ccu_mix _name = {							\
 	}									\
 }
 
-#define CCU_FACTOR_GATE_DEFINE(_name, _parent, _reg_ctrl, _mask_gate, _div,	\
-			       _mul)						\
+#define CCU_FACTOR_GATE_FLAGS_DEFINE(_name, _parent, _reg_ctrl, _mask_gate, _div,	\
+			       _mul, _flags)					\
 static struct ccu_mix _name = {							\
 	.gate	= CCU_GATE_INIT(_mask_gate),					\
 	.factor	= CCU_FACTOR_INIT(_div, _mul),					\
 	.common = {								\
 		.reg_ctrl	= _reg_ctrl,					\
-		CCU_MIX_INITHW(_name, _parent, spacemit_ccu_factor_gate_ops, 0)	\
+		CCU_MIX_INITHW(_name, _parent, spacemit_ccu_factor_gate_ops, _flags)	\
 	}									\
 }
 
+#define CCU_FACTOR_GATE_DEFINE(_name, _parent, _reg_ctrl, _mask_gate, _div,	\
+			       _mul)						\
+	CCU_FACTOR_GATE_FLAGS_DEFINE(_name, _parent, _reg_ctrl, _mask_gate, _div,	\
+			       _mul, 0)
+
 #define CCU_MUX_GATE_DEFINE(_name, _parents, _reg_ctrl, _shift, _width,		\
 			    _mask_gate, _flags)					\
 static struct ccu_mix _name = {							\
Signature · 1 line
2.39.5
author maintainer
#375 lore raw

PATCH 365/627 RDMA/hns: Fix double destruction of rsv_qp

In reply to Greg Kroah-Hartman

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

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

From: wenglianfa <[email protected]>

[ Upstream commit c6957b95ecc5b63c5a4bb4ecc28af326cf8f6dc8 ]

rsv_qp may be double destroyed in error flow, first in free_mr_init(), and then in hns_roce_exit(). Fix it by moving the free_mr_init() call into hns_roce_v2_init().

list_del corruption, ffff589732eb9b50->next is LIST_POISON1 (dead000000000100)

WARNING: CPU: 8 PID: 1047115 at lib/list_debug.c:53 __list_del_entry_valid+0x148/0x240
...
Call trace:
 __list_del_entry_valid+0x148/0x240
 hns_roce_qp_remove+0x4c/0x3f0 [hns_roce_hw_v2]
 hns_roce_v2_destroy_qp_common+0x1dc/0x5f4 [hns_roce_hw_v2]
 hns_roce_v2_destroy_qp+0x22c/0x46c [hns_roce_hw_v2]
 free_mr_exit+0x6c/0x120 [hns_roce_hw_v2]
 hns_roce_v2_exit+0x170/0x200 [hns_roce_hw_v2]
 hns_roce_exit+0x118/0x350 [hns_roce_hw_v2]
 __hns_roce_hw_v2_init_instance+0x1c8/0x304 [hns_roce_hw_v2]
 hns_roce_hw_v2_reset_notify_init+0x170/0x21c [hns_roce_hw_v2]
 hns_roce_hw_v2_reset_notify+0x6c/0x190 [hns_roce_hw_v2]
 hclge_notify_roce_client+0x6c/0x160 [hclge]
 hclge_reset_rebuild+0x150/0x5c0 [hclge]
 hclge_reset+0x10c/0x140 [hclge]
 hclge_reset_subtask+0x80/0x104 [hclge]
 hclge_reset_service_task+0x168/0x3ac [hclge]
 hclge_service_task+0x50/0x100 [hclge]
 process_one_work+0x250/0x9a0
 worker_thread+0x324/0x990
 kthread+0x190/0x210
 ret_from_fork+0x10/0x18
Fixes fd8489294dd2 ("RDMA/hns: Fix Use-After-Free of rsv_qp on HIP08")
---
 drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 25 +++++++++++-----------
 drivers/infiniband/hw/hns/hns_roce_main.c  |  6 +++---
 2 files changed, 16 insertions(+), 15 deletions(-)
drivers/infiniband/hw/hns/hns_roce_hw_v2.c +13 -12
index fa8747656f25..29068be052d9 100644
@@ -2986,14 +2986,22 @@ static int hns_roce_v2_init(struct hns_roce_dev *hr_dev)
 {
 	int ret;
 
+	if (hr_dev->pci_dev->revision == PCI_REVISION_ID_HIP08) {
+		ret = free_mr_init(hr_dev);
+		if (ret) {
+			dev_err(hr_dev->dev, "failed to init free mr!\n");
+			return ret;
+		}
+	}
+
 	/* The hns ROCEE requires the extdb info to be cleared before using */
 	ret = hns_roce_clear_extdb_list_info(hr_dev);
 	if (ret)
-		return ret;
+		goto err_clear_extdb_failed;
 
 	ret = get_hem_table(hr_dev);
 	if (ret)
-		return ret;
+		goto err_clear_extdb_failed;
 
 	if (hr_dev->is_vf)
 		return 0;
@@ -3008,6 +3016,9 @@ static int hns_roce_v2_init(struct hns_roce_dev *hr_dev)
 
 err_llm_init_failed:
 	put_hem_table(hr_dev);
+err_clear_extdb_failed:
+	if (hr_dev->pci_dev->revision == PCI_REVISION_ID_HIP08)
+		free_mr_exit(hr_dev);
 
 	return ret;
 }
@@ -7044,21 +7055,11 @@ static int __hns_roce_hw_v2_init_instance(struct hnae3_handle *handle)
 		goto error_failed_roce_init;
 	}
 
-	if (hr_dev->pci_dev->revision == PCI_REVISION_ID_HIP08) {
-		ret = free_mr_init(hr_dev);
-		if (ret) {
-			dev_err(hr_dev->dev, "failed to init free mr!\n");
-			goto error_failed_free_mr_init;
-		}
-	}
 
 	handle->priv = hr_dev;
 
 	return 0;
 
-error_failed_free_mr_init:
-	hns_roce_exit(hr_dev);
-
 error_failed_roce_init:
 	kfree(hr_dev->priv);
 
drivers/infiniband/hw/hns/hns_roce_main.c +3 -3
index e7a497cc125c..623610b3e2ec 100644
@@ -965,6 +965,9 @@ static int hns_roce_setup_hca(struct hns_roce_dev *hr_dev)
 
 	spin_lock_init(&hr_dev->sm_lock);
 
+	INIT_LIST_HEAD(&hr_dev->qp_list);
+	spin_lock_init(&hr_dev->qp_list_lock);
+
 	if (hr_dev->caps.flags & HNS_ROCE_CAP_FLAG_CQ_RECORD_DB ||
 	    hr_dev->caps.flags & HNS_ROCE_CAP_FLAG_QP_RECORD_DB) {
 		INIT_LIST_HEAD(&hr_dev->pgdir_list);
@@ -1132,9 +1135,6 @@ int hns_roce_init(struct hns_roce_dev *hr_dev)
 		}
 	}
 
-	INIT_LIST_HEAD(&hr_dev->qp_list);
-	spin_lock_init(&hr_dev->qp_list_lock);
-
 	ret = hns_roce_register_device(hr_dev);
 	if (ret)
 		goto error_failed_register_device;
Signature · 1 line
2.39.5
author maintainer
#376 lore raw

PATCH 366/627 RDMA/hns: Fix HW configurations not cleared in error flow

In reply to Greg Kroah-Hartman

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

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

From: wenglianfa <[email protected]>

[ Upstream commit 998b41cb20b02c4e28ac558e4e7f8609d659ec05 ]

hns_roce_clear_extdb_list_info() will eventually do some HW configurations through FW, and they need to be cleared by calling hns_roce_function_clear() when the initialization fails.

Fixes 7e78dd816e45 ("RDMA/hns: Clear extended doorbell info before using")
---
 drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/infiniband/hw/hns/hns_roce_hw_v2.c +3 -1
index 29068be052d9..2b04e25c6b09 100644
@@ -3001,7 +3001,7 @@ static int hns_roce_v2_init(struct hns_roce_dev *hr_dev)
 
 	ret = get_hem_table(hr_dev);
 	if (ret)
-		goto err_clear_extdb_failed;
+		goto err_get_hem_table_failed;
 
 	if (hr_dev->is_vf)
 		return 0;
@@ -3016,6 +3016,8 @@ static int hns_roce_v2_init(struct hns_roce_dev *hr_dev)
 
 err_llm_init_failed:
 	put_hem_table(hr_dev);
+err_get_hem_table_failed:
+	hns_roce_function_clear(hr_dev);
 err_clear_extdb_failed:
 	if (hr_dev->pci_dev->revision == PCI_REVISION_ID_HIP08)
 		free_mr_exit(hr_dev);
Signature · 1 line
2.39.5
author maintainer
#377 lore raw

PATCH 367/627 crypto: ccp - Fix locking on alloc failure handling

In reply to Greg Kroah-Hartman

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

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

From: Alexey Kardashevskiy <[email protected]>

[ Upstream commit b4abeccb8d39db7d9b51cb0098d6458760b30a75 ]

The __snp_alloc_firmware_pages() helper allocates pages in the firmware state (alloc + rmpupdate). In case of failed rmpupdate, it tries reclaiming pages with already changed state. This requires calling the PSP firmware and since there is sev_cmd_mutex to guard such calls, the helper takes a "locked" parameter so specify if the lock needs to be held.

Most calls happen from snp_alloc_firmware_page() which executes without the lock. However

commit 24512afa4336 ("crypto: ccp: Handle the legacy TMR allocation when SNP is enabled")

switched sev_fw_alloc() from alloc_pages() (which does not call the PSP) to __snp_alloc_firmware_pages() (which does) but did not account for the fact that sev_fw_alloc() is called from __sev_platform_init_locked() (via __sev_platform_init_handle_tmr()) and executes with the lock held.

Add a "locked" parameter to __snp_alloc_firmware_pages().
Make sev_fw_alloc() use the new parameter to prevent potential deadlock in rmp_mark_pages_firmware() if rmpupdate() failed.

Fixes 24512afa4336 ("crypto: ccp: Handle the legacy TMR allocation when SNP is enabled")
---
 drivers/crypto/ccp/sev-dev.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
drivers/crypto/ccp/sev-dev.c +4 -4
index 8fb94c5f006a..539c303beb3a 100644
@@ -434,7 +434,7 @@ static int rmp_mark_pages_firmware(unsigned long paddr, unsigned int npages, boo
 	return rc;
 }
 
-static struct page *__snp_alloc_firmware_pages(gfp_t gfp_mask, int order)
+static struct page *__snp_alloc_firmware_pages(gfp_t gfp_mask, int order, bool locked)
 {
 	unsigned long npages = 1ul << order, paddr;
 	struct sev_device *sev;
@@ -453,7 +453,7 @@ static struct page *__snp_alloc_firmware_pages(gfp_t gfp_mask, int order)
 		return page;
 
 	paddr = __pa((unsigned long)page_address(page));
-	if (rmp_mark_pages_firmware(paddr, npages, false))
+	if (rmp_mark_pages_firmware(paddr, npages, locked))
 		return NULL;
 
 	return page;
@@ -463,7 +463,7 @@ void *snp_alloc_firmware_page(gfp_t gfp_mask)
 {
 	struct page *page;
 
-	page = __snp_alloc_firmware_pages(gfp_mask, 0);
+	page = __snp_alloc_firmware_pages(gfp_mask, 0, false);
 
 	return page ? page_address(page) : NULL;
 }
@@ -498,7 +498,7 @@ static void *sev_fw_alloc(unsigned long len)
 {
 	struct page *page;
 
-	page = __snp_alloc_firmware_pages(GFP_KERNEL, get_order(len));
+	page = __snp_alloc_firmware_pages(GFP_KERNEL, get_order(len), true);
 	if (!page)
 		return NULL;
 
Signature · 1 line
2.39.5
author maintainer
#378 lore raw

PATCH 368/627 crypto: inside-secure - Fix `dma_unmap_sg()` nents value

In reply to Greg Kroah-Hartman

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

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

From: Thomas Fourier <[email protected]>

[ Upstream commit cb7fa6b6fc71e0c801e271aa498e2f19e6df2931 ]

The dma_unmap_sg() functions should be called with the same nents as the dma_map_sg(), not the value the map function returned.

Fixes c957f8b3e2e5 ("crypto: inside-secure - avoid unmapping DMA memory that was not mapped")
---
 drivers/crypto/inside-secure/safexcel_hash.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)
drivers/crypto/inside-secure/safexcel_hash.c +6 -2
index d2b632193beb..5baf4bd2fcee 100644
@@ -249,7 +249,9 @@ static int safexcel_handle_req_result(struct safexcel_crypto_priv *priv,
 	safexcel_complete(priv, ring);
 
 	if (sreq->nents) {
-		dma_unmap_sg(priv->dev, areq->src, sreq->nents, DMA_TO_DEVICE);
+		dma_unmap_sg(priv->dev, areq->src,
+			     sg_nents_for_len(areq->src, areq->nbytes),
+			     DMA_TO_DEVICE);
 		sreq->nents = 0;
 	}
 
@@ -497,7 +499,9 @@ static int safexcel_ahash_send_req(struct crypto_async_request *async, int ring,
 			 DMA_FROM_DEVICE);
 unmap_sg:
 	if (req->nents) {
-		dma_unmap_sg(priv->dev, areq->src, req->nents, DMA_TO_DEVICE);
+		dma_unmap_sg(priv->dev, areq->src,
+			     sg_nents_for_len(areq->src, areq->nbytes),
+			     DMA_TO_DEVICE);
 		req->nents = 0;
 	}
 cdesc_rollback:
Signature · 1 line
2.39.5
author maintainer
#379 lore raw

PATCH 369/627 crypto: ccp - Fix crash when rebind ccp device for ccp.ko

In reply to Greg Kroah-Hartman

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

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

From: Mengbiao Xiong <[email protected]>

[ Upstream commit 181698af38d3f93381229ad89c09b5bd0496661a ]

When CONFIG_CRYPTO_DEV_CCP_DEBUGFS is enabled, rebinding the ccp device causes the following crash:

$ echo '0000:0a:00.2' > /sys/bus/pci/drivers/ccp/unbind
$ echo '0000:0a:00.2' > /sys/bus/pci/drivers/ccp/bind
[  204.976930] BUG: kernel NULL pointer dereference, address: 0000000000000098
[  204.978026] #PF: supervisor write access in kernel mode
[  204.979126] #PF: error_code(0x0002) - not-present page
[  204.980226] PGD 0 P4D 0
[  204.981317] Oops: Oops: 0002 [#1] SMP NOPTI
...
[  204.997852] Call Trace:
[  204.999074]  <TASK>
[  205.000297]  start_creating+0x9f/0x1c0
[  205.001533]  debugfs_create_dir+0x1f/0x170
[  205.002769]  ? srso_return_thunk+0x5/0x5f
[  205.004000]  ccp5_debugfs_setup+0x87/0x170 [ccp]
[  205.005241]  ccp5_init+0x8b2/0x960 [ccp]
[  205.006469]  ccp_dev_init+0xd4/0x150 [ccp]
[  205.007709]  sp_init+0x5f/0x80 [ccp]
[  205.008942]  sp_pci_probe+0x283/0x2e0 [ccp]
[  205.010165]  ? srso_return_thunk+0x5/0x5f
[  205.011376]  local_pci_probe+0x4f/0xb0
[  205.012584]  pci_device_probe+0xdb/0x230
[  205.013810]  really_probe+0xed/0x380
[  205.015024]  __driver_probe_device+0x7e/0x160
[  205.016240]  device_driver_attach+0x2f/0x60
[  205.017457]  bind_store+0x7c/0xb0
[  205.018663]  drv_attr_store+0x28/0x40
[  205.019868]  sysfs_kf_write+0x5f/0x70
[  205.021065]  kernfs_fop_write_iter+0x145/0x1d0
[  205.022267]  vfs_write+0x308/0x440
[  205.023453]  ksys_write+0x6d/0xe0
[  205.024616]  __x64_sys_write+0x1e/0x30
[  205.025778]  x64_sys_call+0x16ba/0x2150
[  205.026942]  do_syscall_64+0x56/0x1e0
[  205.028108]  entry_SYSCALL_64_after_hwframe+0x76/0x7e
[  205.029276] RIP: 0033:0x7fbc36f10104
[  205.030420] Code: 89 02 48 c7 c0 ff ff ff ff c3 66 2e 0f 1f 84 00 00 00 00 00 66 90 48 8d 05 e1 08 2e 00 8b 00 85 c0 75 13 b8 01 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 54 f3 c3 66 90 41 54 55 49 89 d4 53 48 89 f5

This patch sets ccp_debugfs_dir to NULL after destroying it in ccp5_debugfs_destroy, allowing the directory dentry to be recreated when rebinding the ccp device.

Tested on AMD Ryzen 7 1700X.

Fixes 3cdbe346ed3f ("crypto: ccp - Add debugfs entries for CCP information")
---
 drivers/crypto/ccp/ccp-debugfs.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/crypto/ccp/ccp-debugfs.c +3 -0
index a1055554b47a..dc26bc22c91d 100644
@@ -319,5 +319,8 @@ void ccp5_debugfs_setup(struct ccp_device *ccp)
 
 void ccp5_debugfs_destroy(void)
 {
+	mutex_lock(&ccp_debugfs_lock);
 	debugfs_remove_recursive(ccp_debugfs_dir);
+	ccp_debugfs_dir = NULL;
+	mutex_unlock(&ccp_debugfs_lock);
 }
Signature · 1 line
2.39.5
author maintainer
#380 lore raw

PATCH 370/627 RDMA/hns: Get message length of ack_req from FW

In reply to Greg Kroah-Hartman

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

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

From: Junxian Huang <[email protected]>

[ Upstream commit 2c2ec0106c0f1f12d4eefd11de318ac47557a750 ]

ACK_REQ_FREQ indicates the number of packets (after MTU fragmentation) HW sends before setting an ACK request. When MTU is greater than or equal to 1024, the current ACK_REQ_FREQ value causes HW to request an ACK for every MTU fragment. The processing of a large number of ACKs severely impacts HW performance when sending large size payloads.

Get message length of ack_req from FW so that we can adjust this parameter according to different situations. There are several constraints for ACK_REQ_FREQ:

1. mtu * (2 ^ ACK_REQ_FREQ) should not be too large, otherwise it may cause some unexpected retries when sending large payload.

2. ACK_REQ_FREQ should be larger than or equal to LP_PKTN_INI.

3. ACK_REQ_FREQ must be equal to LP_PKTN_INI when using LDCP or HC3 congestion control algorithm.

Fixes 56518a603fd2 ("RDMA/hns: Modify the value of long message loopback slice")
---
 drivers/infiniband/hw/hns/hns_roce_device.h |  1 +
 drivers/infiniband/hw/hns/hns_roce_hw_v2.c  | 45 ++++++++++++++++-----
 drivers/infiniband/hw/hns/hns_roce_hw_v2.h  |  8 +++-
 3 files changed, 43 insertions(+), 11 deletions(-)
drivers/infiniband/hw/hns/hns_roce_device.h +1 -0
index 1dcc9cbb4678..254fd4d6ea9f 100644
@@ -856,6 +856,7 @@ struct hns_roce_caps {
 	u16		default_ceq_arm_st;
 	u8		cong_cap;
 	enum hns_roce_cong_type default_cong_type;
+	u32             max_ack_req_msg_len;
 };
 
 enum hns_roce_device_state {
drivers/infiniband/hw/hns/hns_roce_hw_v2.c +35 -10

100 lines not on this page · raw message

drivers/infiniband/hw/hns/hns_roce_hw_v2.h +7 -1
index bc7466830eaf..1c2660305d27 100644
@@ -1168,7 +1168,8 @@ struct hns_roce_cfg_gmv_tb_b {
 #define GMV_TB_B_SMAC_H GMV_TB_B_FIELD_LOC(47, 32)
 #define GMV_TB_B_SGID_IDX GMV_TB_B_FIELD_LOC(71, 64)
 
-#define HNS_ROCE_QUERY_PF_CAPS_CMD_NUM 5
+#define HNS_ROCE_QUERY_PF_CAPS_CMD_NUM_HIP08 5
+#define HNS_ROCE_QUERY_PF_CAPS_CMD_NUM 6
 struct hns_roce_query_pf_caps_a {
 	u8 number_ports;
 	u8 local_ca_ack_delay;
@@ -1280,6 +1281,11 @@ struct hns_roce_query_pf_caps_e {
 	__le16 aeq_period;
 };
 
+struct hns_roce_query_pf_caps_f {
+	__le32 max_ack_req_msg_len;
+	__le32 rsv[5];
+};
+
 #define PF_CAPS_E_FIELD_LOC(h, l) \
 	FIELD_LOC(struct hns_roce_query_pf_caps_e, h, l)
 
Signature · 1 line
2.39.5
author maintainer
#381 lore raw

PATCH 371/627 RDMA/hns: Fix accessing uninitialized resources

In reply to Greg Kroah-Hartman

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

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

From: Junxian Huang <[email protected]>

[ Upstream commit 278c18a4a78a9a6bf529ef45ccde512a5686ea9d ]

hr_dev->pgdir_list and hr_dev->pgdir_mutex won't be initialized if CQ/QP record db are not enabled, but they are also needed when using SRQ with SRQ record db enabled. Simplified the logic by always initailizing the reosurces.

Fixes c9813b0b9992 ("RDMA/hns: Support SRQ record doorbell")
---
 drivers/infiniband/hw/hns/hns_roce_main.c | 16 ++++------------
 1 file changed, 4 insertions(+), 12 deletions(-)
drivers/infiniband/hw/hns/hns_roce_main.c +4 -12
index 623610b3e2ec..11fa64044a8d 100644
@@ -947,10 +947,7 @@ static int hns_roce_init_hem(struct hns_roce_dev *hr_dev)
 static void hns_roce_teardown_hca(struct hns_roce_dev *hr_dev)
 {
 	hns_roce_cleanup_bitmap(hr_dev);
-
-	if (hr_dev->caps.flags & HNS_ROCE_CAP_FLAG_CQ_RECORD_DB ||
-	    hr_dev->caps.flags & HNS_ROCE_CAP_FLAG_QP_RECORD_DB)
-		mutex_destroy(&hr_dev->pgdir_mutex);
+	mutex_destroy(&hr_dev->pgdir_mutex);
 }
 
 /**
@@ -968,11 +965,8 @@ static int hns_roce_setup_hca(struct hns_roce_dev *hr_dev)
 	INIT_LIST_HEAD(&hr_dev->qp_list);
 	spin_lock_init(&hr_dev->qp_list_lock);
 
-	if (hr_dev->caps.flags & HNS_ROCE_CAP_FLAG_CQ_RECORD_DB ||
-	    hr_dev->caps.flags & HNS_ROCE_CAP_FLAG_QP_RECORD_DB) {
-		INIT_LIST_HEAD(&hr_dev->pgdir_list);
-		mutex_init(&hr_dev->pgdir_mutex);
-	}
+	INIT_LIST_HEAD(&hr_dev->pgdir_list);
+	mutex_init(&hr_dev->pgdir_mutex);
 
 	hns_roce_init_uar_table(hr_dev);
 
@@ -1004,9 +998,7 @@ static int hns_roce_setup_hca(struct hns_roce_dev *hr_dev)
 
 err_uar_table_free:
 	ida_destroy(&hr_dev->uar_ida.ida);
-	if (hr_dev->caps.flags & HNS_ROCE_CAP_FLAG_CQ_RECORD_DB ||
-	    hr_dev->caps.flags & HNS_ROCE_CAP_FLAG_QP_RECORD_DB)
-		mutex_destroy(&hr_dev->pgdir_mutex);
+	mutex_destroy(&hr_dev->pgdir_mutex);
 
 	return ret;
 }
Signature · 1 line
2.39.5
author maintainer
#382 lore raw

PATCH 372/627 RDMA/hns: Drop GFP_NOWARN

In reply to Greg Kroah-Hartman

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

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

From: Junxian Huang <[email protected]>

[ Upstream commit 5338abb299f0cd764edf78a7e71a0b746af35030 ]

GFP_NOWARN silences all warnings on dma_alloc_coherent() failure, which might otherwise help with troubleshooting.

Fixes 9a4435375cd1 ("IB/hns: Add driver files for hns RoCE driver")
---
 drivers/infiniband/hw/hns/hns_roce_hem.c | 18 +++++-------------
 1 file changed, 5 insertions(+), 13 deletions(-)
drivers/infiniband/hw/hns/hns_roce_hem.c +5 -13
index ca0798224e56..3d479c63b117 100644
@@ -249,15 +249,12 @@ int hns_roce_calc_hem_mhop(struct hns_roce_dev *hr_dev,
 }
 
 static struct hns_roce_hem *hns_roce_alloc_hem(struct hns_roce_dev *hr_dev,
-					       unsigned long hem_alloc_size,
-					       gfp_t gfp_mask)
+					       unsigned long hem_alloc_size)
 {
 	struct hns_roce_hem *hem;
 	int order;
 	void *buf;
 
-	WARN_ON(gfp_mask & __GFP_HIGHMEM);
-
 	order = get_order(hem_alloc_size);
 	if (PAGE_SIZE << order != hem_alloc_size) {
 		dev_err(hr_dev->dev, "invalid hem_alloc_size: %lu!\n",
@@ -265,13 +262,12 @@ static struct hns_roce_hem *hns_roce_alloc_hem(struct hns_roce_dev *hr_dev,
 		return NULL;
 	}
 
-	hem = kmalloc(sizeof(*hem),
-		      gfp_mask & ~(__GFP_HIGHMEM | __GFP_NOWARN));
+	hem = kmalloc(sizeof(*hem), GFP_KERNEL);
 	if (!hem)
 		return NULL;
 
 	buf = dma_alloc_coherent(hr_dev->dev, hem_alloc_size,
-				 &hem->dma, gfp_mask);
+				 &hem->dma, GFP_KERNEL);
 	if (!buf)
 		goto fail;
 
@@ -378,7 +374,6 @@ static int alloc_mhop_hem(struct hns_roce_dev *hr_dev,
 {
 	u32 bt_size = mhop->bt_chunk_size;
 	struct device *dev = hr_dev->dev;
-	gfp_t flag;
 	u64 bt_ba;
 	u32 size;
 	int ret;
@@ -417,8 +412,7 @@ static int alloc_mhop_hem(struct hns_roce_dev *hr_dev,
 	 * alloc bt space chunk for MTT/CQE.
 	 */
 	size = table->type < HEM_TYPE_MTT ? mhop->buf_chunk_size : bt_size;
-	flag = GFP_KERNEL | __GFP_NOWARN;
-	table->hem[index->buf] = hns_roce_alloc_hem(hr_dev, size, flag);
+	table->hem[index->buf] = hns_roce_alloc_hem(hr_dev, size);
 	if (!table->hem[index->buf]) {
 		ret = -ENOMEM;
 		goto err_alloc_hem;
@@ -546,9 +540,7 @@ int hns_roce_table_get(struct hns_roce_dev *hr_dev,
 		goto out;
 	}
 
-	table->hem[i] = hns_roce_alloc_hem(hr_dev,
-				       table->table_chunk_size,
-				       GFP_KERNEL | __GFP_NOWARN);
+	table->hem[i] = hns_roce_alloc_hem(hr_dev, table->table_chunk_size);
 	if (!table->hem[i]) {
 		ret = -ENOMEM;
 		goto out;
Signature · 1 line
2.39.5
author maintainer
#383 lore raw

PATCH 373/627 RDMA/hns: Fix -Wframe-larger-than issue

In reply to Greg Kroah-Hartman

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

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

From: Junxian Huang <[email protected]>

[ Upstream commit 79d56805c5068f2bc81518043e043c3dedd1c82a ]

Fix -Wframe-larger-than issue by allocating memory for qpc struct with kzalloc() instead of using stack memory.

---
 drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 15 ++++++++++-----
 1 file changed, 10 insertions(+), 5 deletions(-)
drivers/infiniband/hw/hns/hns_roce_hw_v2.c +10 -5
index 246642859159..b30dce00f240 100644
@@ -5387,11 +5387,10 @@ static int hns_roce_v2_modify_qp(struct ib_qp *ibqp,
 {
 	struct hns_roce_dev *hr_dev = to_hr_dev(ibqp->device);
 	struct hns_roce_qp *hr_qp = to_hr_qp(ibqp);
-	struct hns_roce_v2_qp_context ctx[2];
-	struct hns_roce_v2_qp_context *context = ctx;
-	struct hns_roce_v2_qp_context *qpc_mask = ctx + 1;
+	struct hns_roce_v2_qp_context *context;
+	struct hns_roce_v2_qp_context *qpc_mask;
 	struct ib_device *ibdev = &hr_dev->ib_dev;
-	int ret;
+	int ret = -ENOMEM;
 
 	if (attr_mask & ~IB_QP_ATTR_STANDARD_BITS)
 		return -EOPNOTSUPP;
@@ -5402,7 +5401,11 @@ static int hns_roce_v2_modify_qp(struct ib_qp *ibqp,
 	 * we should set all bits of the relevant fields in context mask to
 	 * 0 at the same time, else set them to 0x1.
 	 */
-	memset(context, 0, hr_dev->caps.qpc_sz);
+	context = kvzalloc(sizeof(*context), GFP_KERNEL);
+	qpc_mask = kvzalloc(sizeof(*qpc_mask), GFP_KERNEL);
+	if (!context || !qpc_mask)
+		goto out;
+
 	memset(qpc_mask, 0xff, hr_dev->caps.qpc_sz);
 
 	ret = hns_roce_v2_set_abs_fields(ibqp, attr, attr_mask, cur_state,
@@ -5444,6 +5447,8 @@ static int hns_roce_v2_modify_qp(struct ib_qp *ibqp,
 		clear_qp(hr_qp);
 
 out:
+	kvfree(qpc_mask);
+	kvfree(context);
 	return ret;
 }
 
Signature · 1 line
2.39.5
author maintainer
#384 lore raw

PATCH 374/627 tracing: Use queue_rcu_work() to free filters

In reply to Greg Kroah-Hartman

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

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

From: Steven Rostedt <[email protected]>

[ Upstream commit 3aceaa539cfe3a2e62bd92e6697d9fae1c20c0be ]

Freeing of filters requires to wait for both an RCU grace period as well as a RCU task trace wait period after they have been detached from their lists. The trace task period can be quite large so the freeing of the filters was moved to use the call_rcu*() routines. The problem with that is that the callback functions of call_rcu*() is done from a soft irq and can cause latencies if the callback takes a bit of time.

The filters are freed per event in a system and the syscalls system contains an event per system call, which can be over 700 events. Freeing 700 filters in a bottom half is undesirable.

Instead, move the freeing to use queue_rcu_work() which is done in task context.

Cc Masami Hiramatsu <[email protected]>
Cc Mathieu Desnoyers <[email protected]>
Fixes a9d0aab5eb33 ("tracing: Fix regression of filter waiting a long time on RCU synchronization")
---
 kernel/trace/trace_events_filter.c | 28 ++++++++++++++++++++--------
 1 file changed, 20 insertions(+), 8 deletions(-)
kernel/trace/trace_events_filter.c +20 -8
index 3885aadc434d..196c8bf34970 100644
@@ -1344,13 +1344,14 @@ struct filter_list {
 
 struct filter_head {
 	struct list_head	list;
-	struct rcu_head		rcu;
+	union {
+		struct rcu_head		rcu;
+		struct rcu_work		rwork;
+	};
 };
 
-
-static void free_filter_list(struct rcu_head *rhp)
+static void free_filter_list(struct filter_head *filter_list)
 {
-	struct filter_head *filter_list = container_of(rhp, struct filter_head, rcu);
 	struct filter_list *filter_item, *tmp;
 
 	list_for_each_entry_safe(filter_item, tmp, &filter_list->list, list) {
@@ -1361,9 +1362,20 @@ static void free_filter_list(struct rcu_head *rhp)
 	kfree(filter_list);
 }
 
+static void free_filter_list_work(struct work_struct *work)
+{
+	struct filter_head *filter_list;
+
+	filter_list = container_of(to_rcu_work(work), struct filter_head, rwork);
+	free_filter_list(filter_list);
+}
+
 static void free_filter_list_tasks(struct rcu_head *rhp)
 {
-	call_rcu(rhp, free_filter_list);
+	struct filter_head *filter_list = container_of(rhp, struct filter_head, rcu);
+
+	INIT_RCU_WORK(&filter_list->rwork, free_filter_list_work);
+	queue_rcu_work(system_wq, &filter_list->rwork);
 }
 
 /*
@@ -1460,7 +1472,7 @@ static void filter_free_subsystem_filters(struct trace_subsystem_dir *dir,
 	tracepoint_synchronize_unregister();
 
 	if (head)
-		free_filter_list(&head->rcu);
+		free_filter_list(head);
 
 	list_for_each_entry(file, &tr->events, list) {
 		if (file->system != dir || !file->filter)
@@ -2305,7 +2317,7 @@ static int process_system_preds(struct trace_subsystem_dir *dir,
 	return 0;
  fail:
 	/* No call succeeded */
-	free_filter_list(&filter_list->rcu);
+	free_filter_list(filter_list);
 	parse_error(pe, FILT_ERR_BAD_SUBSYS_FILTER, 0);
 	return -EINVAL;
  fail_mem:
@@ -2315,7 +2327,7 @@ static int process_system_preds(struct trace_subsystem_dir *dir,
 	if (!fail)
 		delay_free_filter(filter_list);
 	else
-		free_filter_list(&filter_list->rcu);
+		free_filter_list(filter_list);
 
 	return -ENOMEM;
 }
Signature · 1 line
2.39.5
author maintainer
#385 lore raw

PATCH 375/627 kernel: trace: preemptirq_delay_test: use offstack cpu mask

In reply to Greg Kroah-Hartman

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

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

From: Arnd Bergmann <[email protected]>

[ Upstream commit adc353c0bfb243ebfd29b6222fa3bf149169a6de ]

A CPU mask on the stack is broken for large values of CONFIG_NR_CPUS:

kernel/trace/preemptirq_delay_test.c: In function ‘preemptirq_delay_run’:
kernel/trace/preemptirq_delay_test.c:143:1: error: the frame size of 8512 bytes is larger than 1536 bytes [-Werror=frame-larger-than=]

Fall back to dynamic allocation here.

Cc Masami Hiramatsu <[email protected]>
Cc Song Chen <[email protected]>
Cc Mathieu Desnoyers <[email protected]>
Fixes 4b9091e1c194 ("kernel: trace: preemptirq_delay_test: add cpu affinity")
---
 kernel/trace/preemptirq_delay_test.c | 13 +++++++++----
 1 file changed, 9 insertions(+), 4 deletions(-)
kernel/trace/preemptirq_delay_test.c +9 -4
index 314ffc143039..acb0c971a408 100644
@@ -117,12 +117,15 @@ static int preemptirq_delay_run(void *data)
 {
 	int i;
 	int s = MIN(burst_size, NR_TEST_FUNCS);
-	struct cpumask cpu_mask;
+	cpumask_var_t cpu_mask;
+
+	if (!alloc_cpumask_var(&cpu_mask, GFP_KERNEL))
+		return -ENOMEM;
 
 	if (cpu_affinity > -1) {
-		cpumask_clear(&cpu_mask);
-		cpumask_set_cpu(cpu_affinity, &cpu_mask);
-		if (set_cpus_allowed_ptr(current, &cpu_mask))
+		cpumask_clear(cpu_mask);
+		cpumask_set_cpu(cpu_affinity, cpu_mask);
+		if (set_cpus_allowed_ptr(current, cpu_mask))
 			pr_err("cpu_affinity:%d, failed\n", cpu_affinity);
 	}
 
@@ -139,6 +142,8 @@ static int preemptirq_delay_run(void *data)
 
 	__set_current_state(TASK_RUNNING);
 
+	free_cpumask_var(cpu_mask);
+
 	return 0;
 }
 
Signature · 1 line
2.39.5
author maintainer
#386 lore raw

PATCH 376/627 RDMA/uverbs: Add empty rdma_uattrs_has_raw_cap() declaration

In reply to Greg Kroah-Hartman

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

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

From: Leon Romanovsky <[email protected]>

[ Upstream commit 98269398c02ab20eb9ed6d77416023a2627049d8 ]

The call to rdma_uattrs_has_raw_cap() is placed in mlx5 fs.c file, which is compiled without relation to CONFIG_INFINIBAND_USER_ACCESS.

Despite the check is used only in flows with CONFIG_INFINIBAND_USER_ACCESS=y|m, the compilers generate the following error for CONFIG_INFINIBAND_USER_ACCESS=n builds.

Quoted 1 line
Quoted 1 line

ERROR: modpost: "rdma_uattrs_has_raw_cap" [drivers/infiniband/hw/mlx5/mlx5_ib.ko] undefined!

---
 include/rdma/ib_verbs.h | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)
include/rdma/ib_verbs.h +6 -2
index 087048b75d13..6353da1c0228 100644
@@ -4794,15 +4794,19 @@ struct ib_ucontext *ib_uverbs_get_ucontext_file(struct ib_uverbs_file *ufile);
 
 #if IS_ENABLED(CONFIG_INFINIBAND_USER_ACCESS)
 int uverbs_destroy_def_handler(struct uverbs_attr_bundle *attrs);
+bool rdma_uattrs_has_raw_cap(const struct uverbs_attr_bundle *attrs);
 #else
 static inline int uverbs_destroy_def_handler(struct uverbs_attr_bundle *attrs)
 {
 	return 0;
 }
+static inline bool
+rdma_uattrs_has_raw_cap(const struct uverbs_attr_bundle *attrs)
+{
+	return false;
+}
 #endif
 
-bool rdma_uattrs_has_raw_cap(const struct uverbs_attr_bundle *attrs);
-
 struct net_device *rdma_alloc_netdev(struct ib_device *device, u32 port_num,
 				     enum rdma_netdev_t type, const char *name,
 				     unsigned char name_assign_type,
Signature · 1 line
2.39.5
author maintainer
#387 lore raw

PATCH 377/627 proc: use the same treatment to check proc_lseek as ones for proc_read_iter et.al

In reply to Greg Kroah-Hartman

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

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

From: wangzijie <[email protected]>

[ Upstream commit ff7ec8dc1b646296f8d94c39339e8d3833d16c05 ]

Check pde->proc_ops->proc_lseek directly may cause UAF in rmmod scenario. It's a gap in proc_reg_open() after commit 654b33ada4ab("proc: fix UAF in proc_get_inode()"). Followed by AI Viro's suggestion, fix it in same manner.

Fixes 3f61631d47f1 ("take care to handle NULL ->proc_lseek()")
Cc Alexei Starovoitov <[email protected]>
Cc Al Viro <[email protected]>
Cc "Edgecombe, Rick P" <[email protected]>
Cc Kirill A. Shuemov <[email protected]>
---
 fs/proc/generic.c       | 2 ++
 fs/proc/inode.c         | 2 +-
 fs/proc/internal.h      | 5 +++++
 include/linux/proc_fs.h | 1 +
 4 files changed, 9 insertions(+), 1 deletion(-)
fs/proc/generic.c +2 -0
index a3e22803cddf..e0e50914ab25 100644
@@ -569,6 +569,8 @@ static void pde_set_flags(struct proc_dir_entry *pde)
 	if (pde->proc_ops->proc_compat_ioctl)
 		pde->flags |= PROC_ENTRY_proc_compat_ioctl;
 #endif
+	if (pde->proc_ops->proc_lseek)
+		pde->flags |= PROC_ENTRY_proc_lseek;
 }
 
 struct proc_dir_entry *proc_create_data(const char *name, umode_t mode,
fs/proc/inode.c +1 -1
index 3604b616311c..129490151be1 100644
@@ -473,7 +473,7 @@ static int proc_reg_open(struct inode *inode, struct file *file)
 	typeof_member(struct proc_ops, proc_open) open;
 	struct pde_opener *pdeo;
 
-	if (!pde->proc_ops->proc_lseek)
+	if (!pde_has_proc_lseek(pde))
 		file->f_mode &= ~FMODE_LSEEK;
 
 	if (pde_is_permanent(pde)) {
fs/proc/internal.h +5 -0
index 96122e91c645..3d48ffe72583 100644
@@ -99,6 +99,11 @@ static inline bool pde_has_proc_compat_ioctl(const struct proc_dir_entry *pde)
 #endif
 }
 
+static inline bool pde_has_proc_lseek(const struct proc_dir_entry *pde)
+{
+	return pde->flags & PROC_ENTRY_proc_lseek;
+}
+
 extern struct kmem_cache *proc_dir_entry_cache;
 void pde_free(struct proc_dir_entry *pde);
 
include/linux/proc_fs.h +1 -0
index ea62201c74c4..703d0c76cc9a 100644
@@ -27,6 +27,7 @@ enum {
 
 	PROC_ENTRY_proc_read_iter	= 1U << 1,
 	PROC_ENTRY_proc_compat_ioctl	= 1U << 2,
+	PROC_ENTRY_proc_lseek		= 1U << 3,
 };
 
 struct proc_ops {
Signature · 1 line
2.39.5
author maintainer
#388 lore raw

PATCH 378/627 cxl/core: Introduce a new helper cxl_resource_contains_addr()

In reply to Greg Kroah-Hartman

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

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

From: Li Ming <[email protected]>

[ Upstream commit 5b6031c832c2747d58d3f0130098d965ef050b9a ]

In CXL subsystem, many functions need to check an address availability by checking if the resource range contains the address. Providing a new helper function cxl_resource_contains_addr() to check if the resource range contains the input address.

Stable-dep-of: 03ff65c02559 ("cxl/edac: Fix wrong dpa checking for PPR operation")

---
 drivers/cxl/core/core.h | 1 +
 drivers/cxl/core/hdm.c  | 7 +++++++
 2 files changed, 8 insertions(+)
drivers/cxl/core/core.h +1 -0
index 29b61828a847..6b78b10da3e1 100644
@@ -80,6 +80,7 @@ int cxl_dpa_alloc(struct cxl_endpoint_decoder *cxled, u64 size);
 int cxl_dpa_free(struct cxl_endpoint_decoder *cxled);
 resource_size_t cxl_dpa_size(struct cxl_endpoint_decoder *cxled);
 resource_size_t cxl_dpa_resource_start(struct cxl_endpoint_decoder *cxled);
+bool cxl_resource_contains_addr(const struct resource *res, const resource_size_t addr);
 
 enum cxl_rcrb {
 	CXL_RCRB_DOWNSTREAM,
drivers/cxl/core/hdm.c +7 -0
index ab1007495f6b..088caa6b6f74 100644
@@ -547,6 +547,13 @@ resource_size_t cxl_dpa_resource_start(struct cxl_endpoint_decoder *cxled)
 	return base;
 }
 
+bool cxl_resource_contains_addr(const struct resource *res, const resource_size_t addr)
+{
+	struct resource _addr = DEFINE_RES_MEM(addr, 1);
+
+	return resource_contains(res, &_addr);
+}
+
 int cxl_dpa_free(struct cxl_endpoint_decoder *cxled)
 {
 	struct cxl_port *port = cxled_to_port(cxled);
Signature · 1 line
2.39.5
author maintainer
#389 lore raw

PATCH 379/627 cxl/edac: Fix wrong dpa checking for PPR operation

In reply to Greg Kroah-Hartman

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

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

From: Li Ming <[email protected]>

[ Upstream commit 03ff65c02559e8da32be231d7f10fe899233ceae ]

Per Table 8-143. "Get Partition Info Output Payload" in CXL r3.2 section 8.2.10.9.2.1 "Get Partition Info(Opcode 4100h)", DPA 0 is a valid address of a CXL device. However, cxl_do_ppr() considers it as an invalid address, so that user will get an -EINVAL when user calls the sysfs interface of the edac driver to trigger a Post Package Repair(PPR) operation for DPA 0 on a CXL device. The correct implementation should be checking if the input DPA is in the DPA range of the CXL device.

---
 drivers/cxl/core/edac.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/cxl/core/edac.c +4 -1
index 623aaa4439c4..991fa3e70522 100644
@@ -1923,8 +1923,11 @@ static int cxl_ppr_set_nibble_mask(struct device *dev, void *drv_data,
 static int cxl_do_ppr(struct device *dev, void *drv_data, u32 val)
 {
 	struct cxl_ppr_context *cxl_ppr_ctx = drv_data;
+	struct cxl_memdev *cxlmd = cxl_ppr_ctx->cxlmd;
+	struct cxl_dev_state *cxlds = cxlmd->cxlds;
 
-	if (!cxl_ppr_ctx->dpa || val != EDAC_DO_MEM_REPAIR)
+	if (val != EDAC_DO_MEM_REPAIR ||
+	    !cxl_resource_contains_addr(&cxlds->dpa_res, cxl_ppr_ctx->dpa))
 		return -EINVAL;
 
 	return cxl_mem_perform_ppr(cxl_ppr_ctx);
Signature · 1 line
2.39.5
author maintainer
#390 lore raw

PATCH 380/627 pinmux: fix race causing mux_owner NULL with active mux_usecount

In reply to Greg Kroah-Hartman

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

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

From: Mukesh Ojha <[email protected]>

[ Upstream commit 0b075c011032f88d1cfde3b45d6dcf08b44140eb ]

commit 5a3e85c3c397 ("pinmux: Use sequential access to access desc->pinmux data") tried to address the issue when two client of the same gpio calls pinctrl_select_state() for the same functionality, was resulting in NULL pointer issue while accessing desc->mux_owner. However, issue was not completely fixed due to the way it was handled and it can still result in the same NULL pointer.

The issue occurs due to the following interleaving:

     cpu0 (process A)                   cpu1 (process B)

      pin_request() {                   pin_free() {

                                         mutex_lock()
                                         desc->mux_usecount--; //becomes 0
                                         ..
                                         mutex_unlock()
  mutex_lock(desc->mux)
  desc->mux_usecount++; // becomes 1
  desc->mux_owner = owner;
  mutex_unlock(desc->mux)
                                         mutex_lock(desc->mux)
                                         desc->mux_owner = NULL;
                                         mutex_unlock(desc->mux)

This sequence leads to a state where the pin appears to be in use (mux_usecount == 1) but has no owner (mux_owner == NULL), which can cause NULL pointer on next pin_request on the same pin.

Ensure that updates to mux_usecount and mux_owner are performed atomically under the same lock. Only clear mux_owner when mux_usecount reaches zero and no new owner has been assigned.

Fixes 5a3e85c3c397 ("pinmux: Use sequential access to access desc->pinmux data")
---
 drivers/pinctrl/pinmux.c | 20 +++++++++-----------
 1 file changed, 9 insertions(+), 11 deletions(-)
drivers/pinctrl/pinmux.c +9 -11
index 0743190da59e..2c31e7f2a27a 100644
@@ -236,18 +236,7 @@ static const char *pin_free(struct pinctrl_dev *pctldev, int pin,
 			if (desc->mux_usecount)
 				return NULL;
 		}
-	}
-
-	/*
-	 * If there is no kind of request function for the pin we just assume
-	 * we got it by default and proceed.
-	 */
-	if (gpio_range && ops->gpio_disable_free)
-		ops->gpio_disable_free(pctldev, gpio_range, pin);
-	else if (ops->free)
-		ops->free(pctldev, pin);
 
-	scoped_guard(mutex, &desc->mux_lock) {
 		if (gpio_range) {
 			owner = desc->gpio_owner;
 			desc->gpio_owner = NULL;
@@ -258,6 +247,15 @@ static const char *pin_free(struct pinctrl_dev *pctldev, int pin,
 		}
 	}
 
+	/*
+	 * If there is no kind of request function for the pin we just assume
+	 * we got it by default and proceed.
+	 */
+	if (gpio_range && ops->gpio_disable_free)
+		ops->gpio_disable_free(pctldev, gpio_range, pin);
+	else if (ops->free)
+		ops->free(pctldev, pin);
+
 	module_put(pctldev->owner);
 
 	return owner;
Signature · 1 line
2.39.5
author maintainer
#391 lore raw

PATCH 381/627 perf tests bp_account: Fix leaked file descriptor

In reply to Greg Kroah-Hartman

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

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

From: Leo Yan <[email protected]>

[ Upstream commit 4a6cdecaa1497f1fbbd1d5307a225b6ca5a62a90 ]

Since the commit e9846f5ead26 ("perf test: In forked mode add check that fds aren't leaked"), the test "Breakpoint accounting" reports the error:

  # perf test -vvv "Breakpoint accounting"
  20: Breakpoint accounting:
  --- start ---
  test child forked, pid 373
  failed opening event 0
  failed opening event 0
  watchpoints count 4, breakpoints count 6, has_ioctl 1, share 0
  wp 0 created
  wp 1 created
  wp 2 created
  wp 3 created
  wp 0 modified to bp
  wp max created
  ---- end(0) ----
  Leak of file descriptor 7 that opened: 'anon_inode:[perf_event]'

A watchpoint's file descriptor was not properly released. This patch fixes the leak.

---
 tools/perf/tests/bp_account.c | 1 +
 1 file changed, 1 insertion(+)
tools/perf/tests/bp_account.c +1 -0
index 4cb7d486b5c1..047433c977bc 100644
@@ -104,6 +104,7 @@ static int bp_accounting(int wp_cnt, int share)
 		fd_wp = wp_event((void *)&the_var, &attr_new);
 		TEST_ASSERT_VAL("failed to create max wp\n", fd_wp != -1);
 		pr_debug("wp max created\n");
+		close(fd_wp);
 	}
 
 	for (i = 0; i < wp_cnt; i++)
Signature · 1 line
2.39.5
author maintainer
#392 lore raw

PATCH 382/627 perf hwmon_pmu: Avoid shortening hwmon PMU name

In reply to Greg Kroah-Hartman

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

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

From: Ian Rogers <[email protected]>

[ Upstream commit 28f5aa8184c9c9b8eab35fa3884c416fe75e88e4 ]

Long names like ucsi_source_psy_USBC000:001 when prefixed with hwmon_ exceed the buffer size and the last digit is lost. This causes confusion with similar names like ucsi_source_psy_USBC000:002. Extend the buffer size to avoid this.

Fixes 53cc0b351ec9 ("perf hwmon_pmu: Add a tool PMU exposing events from hwmon in sysfs")
---
 tools/perf/util/hwmon_pmu.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
tools/perf/util/hwmon_pmu.c +1 -1
index c25e7296f1c1..75683c543994 100644
@@ -344,7 +344,7 @@ static int hwmon_pmu__read_events(struct hwmon_pmu *pmu)
 
 struct perf_pmu *hwmon_pmu__new(struct list_head *pmus, int hwmon_dir, const char *sysfs_name, const char *name)
 {
-	char buf[32];
+	char buf[64];
 	struct hwmon_pmu *hwm;
 	__u32 type = PERF_PMU_TYPE_HWMON_START + strtoul(sysfs_name + 5, NULL, 10);
 
Signature · 1 line
2.39.5
author maintainer
#393 lore raw

PATCH 383/627 perf python: Fix thread check in pyrf_evsel__read

In reply to Greg Kroah-Hartman

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

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

From: Ian Rogers <[email protected]>

[ Upstream commit 64ec9b997f3a9462901a404ad60f452f76dd2d6e ]

The CPU index is incorrectly checked rather than the thread index.

Fixes 739621f65702 ("perf python: Add evsel read method")
---
 tools/perf/util/python.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
tools/perf/util/python.c +1 -1
index 321c333877fa..eb560e3f9e35 100644
@@ -909,7 +909,7 @@ static PyObject *pyrf_evsel__read(struct pyrf_evsel *pevsel,
 		return NULL;
 	}
 	thread_idx = perf_thread_map__idx(evsel->core.threads, thread);
-	if (cpu_idx < 0) {
+	if (thread_idx < 0) {
 		PyErr_Format(PyExc_TypeError, "Thread %d is not part of evsel's threads",
 			     thread);
 		return NULL;
Signature · 1 line
2.39.5
author maintainer
#394 lore raw

PATCH 384/627 perf python: Correct pyrf_evsel__read for tool PMUs

In reply to Greg Kroah-Hartman

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

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

From: Ian Rogers <[email protected]>

[ Upstream commit 6183afcba9c1c810656ddb36170106aaf3cf778c ]

Tool PMUs assume that stat's process_counter_values is being used to read the counters. Specifically they hold onto old values in evsel->prev_raw_counts and give the cumulative count based off of this value. Update pyrf_evsel__read to allocate counts and prev_raw_counts, use evsel__read_counter rather than perf_evsel__read so tool PMUs are read from not just perf_event_open events, make the returned pyrf_counts_values contain the delta value rather than the cumulative value.

Fixes 739621f65702 ("perf python: Add evsel read method")
---
 tools/perf/util/python.c | 47 +++++++++++++++++++++++++++++++++++++---
 1 file changed, 44 insertions(+), 3 deletions(-)
tools/perf/util/python.c +44 -3
index eb560e3f9e35..b9fe7f2c14af 100644
@@ -10,6 +10,7 @@
 #endif
 #include <perf/mmap.h>
 #include "callchain.h"
+#include "counts.h"
 #include "evlist.h"
 #include "evsel.h"
 #include "event.h"
@@ -888,12 +889,38 @@ static PyObject *pyrf_evsel__threads(struct pyrf_evsel *pevsel)
 	return (PyObject *)pthread_map;
 }
 
+/*
+ * Ensure evsel's counts and prev_raw_counts are allocated, the latter
+ * used by tool PMUs to compute the cumulative count as expected by
+ * stat's process_counter_values.
+ */
+static int evsel__ensure_counts(struct evsel *evsel)
+{
+	int nthreads, ncpus;
+
+	if (evsel->counts != NULL)
+		return 0;
+
+	nthreads = perf_thread_map__nr(evsel->core.threads);
+	ncpus = perf_cpu_map__nr(evsel->core.cpus);
+
+	evsel->counts = perf_counts__new(ncpus, nthreads);
+	if (evsel->counts == NULL)
+		return -ENOMEM;
+
+	evsel->prev_raw_counts = perf_counts__new(ncpus, nthreads);
+	if (evsel->prev_raw_counts == NULL)
+		return -ENOMEM;
+
+	return 0;
+}
+
 static PyObject *pyrf_evsel__read(struct pyrf_evsel *pevsel,
 				  PyObject *args, PyObject *kwargs)
 {
 	struct evsel *evsel = &pevsel->evsel;
 	int cpu = 0, cpu_idx, thread = 0, thread_idx;
-	struct perf_counts_values counts;
+	struct perf_counts_values *old_count, *new_count;
 	struct pyrf_counts_values *count_values = PyObject_New(struct pyrf_counts_values,
 							       &pyrf_counts_values__type);
 
@@ -914,8 +941,22 @@ static PyObject *pyrf_evsel__read(struct pyrf_evsel *pevsel,
 			     thread);
 		return NULL;
 	}
-	perf_evsel__read(&(evsel->core), cpu_idx, thread_idx, &counts);
-	count_values->values = counts;
+
+	if (evsel__ensure_counts(evsel))
+		return PyErr_NoMemory();
+
+	/* Set up pointers to the old and newly read counter values. */
+	old_count = perf_counts(evsel->prev_raw_counts, cpu_idx, thread_idx);
+	new_count = perf_counts(evsel->counts, cpu_idx, thread_idx);
+	/* Update the value in evsel->counts. */
+	evsel__read_counter(evsel, cpu_idx, thread_idx);
+	/* Copy the value and turn it into the delta from old_count. */
+	count_values->values = *new_count;
+	count_values->values.val -= old_count->val;
+	count_values->values.ena -= old_count->ena;
+	count_values->values.run -= old_count->run;
+	/* Save the new count over the old_count for the next read. */
+	*old_count = *new_count;
 	return (PyObject *)count_values;
 }
 
Signature · 1 line
2.39.5
author maintainer
#395 lore raw

PATCH 385/627 RDMA/mana_ib: Fix DSCP value in modify QP

In reply to Greg Kroah-Hartman

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

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

From: Shiraz Saleem <[email protected]>

[ Upstream commit 62de0e67328e9503459a24b9343c3358937cdeef ]

Convert the traffic_class in GRH to a DSCP value as required by the HW.

---
 drivers/infiniband/hw/mana/qp.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/infiniband/hw/mana/qp.c +1 -1
index 14fd7d6c54a2..a6bf4d539e67 100644
@@ -772,7 +772,7 @@ static int mana_ib_gd_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
 		req.ah_attr.dest_port = ROCE_V2_UDP_DPORT;
 		req.ah_attr.src_port = rdma_get_udp_sport(attr->ah_attr.grh.flow_label,
 							  ibqp->qp_num, attr->dest_qp_num);
-		req.ah_attr.traffic_class = attr->ah_attr.grh.traffic_class;
+		req.ah_attr.traffic_class = attr->ah_attr.grh.traffic_class >> 2;
 		req.ah_attr.hop_limit = attr->ah_attr.grh.hop_limit;
 	}
 
Signature · 1 line
2.39.5
author maintainer
#396 lore raw

PATCH 386/627 clk: thead: th1520-ap: Correctly refer the parent of osc_12m

In reply to Greg Kroah-Hartman

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

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

From: Yao Zi <[email protected]>

[ Upstream commit d274c77ffa202b70ad01d579f33b73b4de123375 ]

The "osc_12m" fixed factor clock refers the external oscillator by setting clk_parent_data.fw_name to osc_24m, which is obviously wrong since no clock-names property is allowed for compatible thead,th1520-clk-ap.

Refer the oscillator as parent by index instead.

Fixes ae81b69fd2b1 ("clk: thead: Add support for T-Head TH1520 AP_SUBSYS clocks")
---
 drivers/clk/thead/clk-th1520-ap.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)
drivers/clk/thead/clk-th1520-ap.c +8 -1
index ebfb1d59401d..42feb4bb6329 100644
@@ -582,7 +582,14 @@ static const struct clk_parent_data peri2sys_apb_pclk_pd[] = {
 	{ .hw = &peri2sys_apb_pclk.common.hw }
 };
 
-static CLK_FIXED_FACTOR_FW_NAME(osc12m_clk, "osc_12m", "osc_24m", 2, 1, 0);
+static struct clk_fixed_factor osc12m_clk = {
+	.div		= 2,
+	.mult		= 1,
+	.hw.init	= CLK_HW_INIT_PARENTS_DATA("osc_12m",
+						   osc_24m_clk,
+						   &clk_fixed_factor_ops,
+						   0),
+};
 
 static const char * const out_parents[] = { "osc_24m", "osc_12m" };
 
Signature · 1 line
2.39.5
author maintainer
#397 lore raw

PATCH 387/627 ext4: correct the reserved credits for extent conversion

In reply to Greg Kroah-Hartman

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

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

From: Zhang Yi <[email protected]>

[ Upstream commit 95ad8ee45cdbc321c135a2db895d48b374ef0f87 ]

Now, we reserve journal credits for converting extents in only one page to written state when the I/O operation is complete. This is insufficient when large folio is enabled.

Fix this by reserving credits for converting up to one extent per block in the largest 2MB folio, this calculation should only involve extents index and leaf blocks, so it should not estimate too many credits.

---
 fs/ext4/inode.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
fs/ext4/inode.c +3 -3
index be9a4cba35fd..91da3ae0bbc6 100644
@@ -2771,12 +2771,12 @@ static int ext4_do_writepages(struct mpage_da_data *mpd)
 	mpd->journalled_more_data = 0;
 
 	if (ext4_should_dioread_nolock(inode)) {
+		int bpf = ext4_journal_blocks_per_folio(inode);
 		/*
 		 * We may need to convert up to one extent per block in
-		 * the page and we may dirty the inode.
+		 * the folio and we may dirty the inode.
 		 */
-		rsv_blocks = 1 + ext4_chunk_trans_blocks(inode,
-						PAGE_SIZE >> inode->i_blkbits);
+		rsv_blocks = 1 + ext4_ext_index_trans_blocks(inode, bpf);
 	}
 
 	if (wbc->range_start == 0 && wbc->range_end == LLONG_MAX)
Signature · 1 line
2.39.5
author maintainer
#398 lore raw

PATCH 388/627 ext4: fix insufficient credits calculation in ext4_meta_trans_blocks()

In reply to Greg Kroah-Hartman

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

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

From: Zhang Yi <[email protected]>

[ Upstream commit 5137d6c8906b55b3c7b5d1aa5a549753ec8520f5 ]

The calculation of journal credits in ext4_meta_trans_blocks() should include pextents, as each extent separately may be allocated from a different group and thus need to update different bitmap and group descriptor block.

---
 fs/ext4/inode.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
fs/ext4/inode.c +2 -2
index 91da3ae0bbc6..8997a5f096b4 100644
@@ -6139,7 +6139,7 @@ int ext4_meta_trans_blocks(struct inode *inode, int lblocks, int pextents)
 	int ret;
 
 	/*
-	 * How many index and lead blocks need to touch to map @lblocks
+	 * How many index and leaf blocks need to touch to map @lblocks
 	 * logical blocks to @pextents physical extents?
 	 */
 	idxblocks = ext4_index_trans_blocks(inode, lblocks, pextents);
@@ -6148,7 +6148,7 @@ int ext4_meta_trans_blocks(struct inode *inode, int lblocks, int pextents)
 	 * Now let's see how many group bitmaps and group descriptors need
 	 * to account
 	 */
-	groups = idxblocks;
+	groups = idxblocks + pextents;
 	gdpblocks = groups;
 	if (groups > ngroups)
 		groups = ngroups;
Signature · 1 line
2.39.5
author maintainer
#399 lore raw

PATCH 389/627 clk: sunxi-ng: v3s: Fix de clock definition

In reply to Greg Kroah-Hartman

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

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

From: Paul Kocialkowski <[email protected]>

[ Upstream commit e8ab346f9907a1a3aa2f0e5decf849925c06ae2e ]

The de clock is marked with CLK_SET_RATE_PARENT, which is really not necessary (as confirmed from experimentation) and significantly restricts flexibility for other clocks using the same parent.

In addition the source selection (parent) field is marked as using 2 bits, when it the documentation reports that it uses 3.

Fix both issues in the de clock definition.

Fixes d0f11d14b0bc ("clk: sunxi-ng: add support for V3s CCU")
---
 drivers/clk/sunxi-ng/ccu-sun8i-v3s.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
drivers/clk/sunxi-ng/ccu-sun8i-v3s.c +1 -2
index 52e4369664c5..df345a620d8d 100644
@@ -347,8 +347,7 @@ static SUNXI_CCU_GATE(dram_ohci_clk,	"dram-ohci",	"dram",
 
 static const char * const de_parents[] = { "pll-video", "pll-periph0" };
 static SUNXI_CCU_M_WITH_MUX_GATE(de_clk, "de", de_parents,
-				 0x104, 0, 4, 24, 2, BIT(31),
-				 CLK_SET_RATE_PARENT);
+				 0x104, 0, 4, 24, 3, BIT(31), 0);
 
 static const char * const tcon_parents[] = { "pll-video", "pll-periph0" };
 static SUNXI_CCU_M_WITH_MUX_GATE(tcon_clk, "tcon", tcon_parents,
Signature · 1 line
2.39.5
author maintainer
#400 lore raw

PATCH 390/627 scsi: ibmvscsi_tgt: Fix dma_unmap_sg() nents value

In reply to Greg Kroah-Hartman

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

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

From: Thomas Fourier <[email protected]>

[ Upstream commit 023a293b9cd0bb86a9b50cd7688a3d9d266826db ]

The dma_unmap_sg() functions should be called with the same nents as the dma_map_sg(), not the value the map function returned.

Fixes 88a678bbc34c ("ibmvscsis: Initial commit of IBM VSCSI Tgt Driver")
---
 drivers/scsi/ibmvscsi_tgt/libsrp.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/scsi/ibmvscsi_tgt/libsrp.c +4 -2
index 8a0e28aec928..0ecad398ed3d 100644
@@ -184,7 +184,8 @@ static int srp_direct_data(struct ibmvscsis_cmd *cmd, struct srp_direct_buf *md,
 	err = rdma_io(cmd, sg, nsg, md, 1, dir, len);
 
 	if (dma_map)
-		dma_unmap_sg(iue->target->dev, sg, nsg, DMA_BIDIRECTIONAL);
+		dma_unmap_sg(iue->target->dev, sg, cmd->se_cmd.t_data_nents,
+			     DMA_BIDIRECTIONAL);
 
 	return err;
 }
@@ -256,7 +257,8 @@ static int srp_indirect_data(struct ibmvscsis_cmd *cmd, struct srp_cmd *srp_cmd,
 	err = rdma_io(cmd, sg, nsg, md, nmd, dir, len);
 
 	if (dma_map)
-		dma_unmap_sg(iue->target->dev, sg, nsg, DMA_BIDIRECTIONAL);
+		dma_unmap_sg(iue->target->dev, sg, cmd->se_cmd.t_data_nents,
+			     DMA_BIDIRECTIONAL);
 
 free_mem:
 	if (token && dma_map) {
Signature · 1 line
2.39.5
author maintainer
#401 lore raw

PATCH 391/627 scsi: core: Fix kernel doc for scsi_track_queue_full()

In reply to Greg Kroah-Hartman

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

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

From: Bagas Sanjaya <[email protected]>

[ Upstream commit 6070bd558aee1eb5114e1676165bf0ccaa08240a ]

Sphinx reports indentation warning on scsi_track_queue_full() return values:

Documentation/driver-api/scsi:101: ./drivers/scsi/scsi.c:247: ERROR: Unexpected indentation. [docutils]

Fix the warning by making the return values listing a bullet list.

Fixes eb44820c28bc ("[SCSI] Add Documentation and integrate into docbook build")
Fixes 1da177e4c3f4 ("Linux-2.6.12-rc2")
---
 drivers/scsi/scsi.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)
drivers/scsi/scsi.c +5 -3
index 518a252eb6aa..c2527dd289d9 100644
@@ -242,9 +242,11 @@ EXPORT_SYMBOL(scsi_change_queue_depth);
  * 		specific SCSI device to determine if and when there is a
  * 		need to adjust the queue depth on the device.
  *
- * Returns:	0 - No change needed, >0 - Adjust queue depth to this new depth,
- * 		-1 - Drop back to untagged operation using host->cmd_per_lun
- * 			as the untagged command depth
+ * Returns:
+ * * 0 - No change needed
+ * * >0 - Adjust queue depth to this new depth,
+ * * -1 - Drop back to untagged operation using host->cmd_per_lun as the
+ *   untagged command depth
  *
  * Lock Status:	None held on entry
  *
Signature · 1 line
2.39.5
author maintainer
#402 lore raw

PATCH 392/627 scsi: elx: efct: Fix dma_unmap_sg() nents value

In reply to Greg Kroah-Hartman

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

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

From: Thomas Fourier <[email protected]>

[ Upstream commit 3a988d0b65d7d1713ce7596eae288a293f3b938e ]

The dma_unmap_sg() functions should be called with the same nents as the dma_map_sg(), not the value the map function returned.

Fixes 692e5d73a811 ("scsi: elx: efct: LIO backend interface routines")
---
 drivers/scsi/elx/efct/efct_lio.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/scsi/elx/efct/efct_lio.c +1 -1
index 9ac69356b13e..bd3d489e56ae 100644
@@ -382,7 +382,7 @@ efct_lio_sg_unmap(struct efct_io *io)
 		return;
 
 	dma_unmap_sg(&io->efct->pci->dev, cmd->t_data_sg,
-		     ocp->seg_map_cnt, cmd->data_direction);
+		     cmd->t_data_nents, cmd->data_direction);
 	ocp->seg_map_cnt = 0;
 }
 
Signature · 1 line
2.39.5
author maintainer
#403 lore raw

PATCH 393/627 scsi: mvsas: Fix dma_unmap_sg() nents value

In reply to Greg Kroah-Hartman

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

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

From: Thomas Fourier <[email protected]>

[ Upstream commit 0141618727bc929fe868153d21797f10ce5bef3f ]

The dma_unmap_sg() functions should be called with the same nents as the dma_map_sg(), not the value the map function returned.

Fixes b5762948263d ("[SCSI] mvsas: Add Marvell 6440 SAS/SATA driver")
---
 drivers/scsi/mvsas/mv_sas.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/scsi/mvsas/mv_sas.c +2 -2
index 6c46654b9cd9..15b3d9d55a4b 100644
@@ -818,7 +818,7 @@ static int mvs_task_prep(struct sas_task *task, struct mvs_info *mvi, int is_tmf
 	dev_printk(KERN_ERR, mvi->dev, "mvsas prep failed[%d]!\n", rc);
 	if (!sas_protocol_ata(task->task_proto))
 		if (n_elem)
-			dma_unmap_sg(mvi->dev, task->scatter, n_elem,
+			dma_unmap_sg(mvi->dev, task->scatter, task->num_scatter,
 				     task->data_dir);
 prep_out:
 	return rc;
@@ -864,7 +864,7 @@ static void mvs_slot_task_free(struct mvs_info *mvi, struct sas_task *task,
 	if (!sas_protocol_ata(task->task_proto))
 		if (slot->n_elem)
 			dma_unmap_sg(mvi->dev, task->scatter,
-				     slot->n_elem, task->data_dir);
+				     task->num_scatter, task->data_dir);
 
 	switch (task->task_proto) {
 	case SAS_PROTOCOL_SMP:
Signature · 1 line
2.39.5
author maintainer
#404 lore raw

PATCH 394/627 scsi: isci: Fix dma_unmap_sg() nents value

In reply to Greg Kroah-Hartman

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

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

From: Thomas Fourier <[email protected]>

[ Upstream commit 063bec4444d54e5f35d11949c5c90eaa1ff84c11 ]

The dma_unmap_sg() functions should be called with the same nents as the dma_map_sg(), not the value the map function returned.

---
 drivers/scsi/isci/request.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/scsi/isci/request.c +1 -1
index 355a0bc0828e..bb89a2e33eb4 100644
@@ -2904,7 +2904,7 @@ static void isci_request_io_request_complete(struct isci_host *ihost,
 					 task->total_xfer_len, task->data_dir);
 		else  /* unmap the sgl dma addresses */
 			dma_unmap_sg(&ihost->pdev->dev, task->scatter,
-				     request->num_sg_entries, task->data_dir);
+				     task->num_scatter, task->data_dir);
 		break;
 	case SAS_PROTOCOL_SMP: {
 		struct scatterlist *sg = &task->smp_task.smp_req;
Signature · 1 line
2.39.5
author maintainer
#405 lore raw

PATCH 395/627 PCI: Fix driver_managed_dma check

In reply to Greg Kroah-Hartman

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

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

From: Robin Murphy <[email protected]>

[ Upstream commit 78447d4545b2ea76ee04f4e46d473639483158b2 ]

Since it's not currently safe to take device_lock() in the IOMMU probe path, that can race against really_probe() setting dev->driver before attempting to bind. The race itself isn't so bad, since we're only concerned with dereferencing dev->driver itself anyway, but sadly my attempt to implement the check with minimal churn leads to a kind of Time-of-Check to Time-of-Use (TOCTOU) issue, where dev->driver becomes valid after to_pci_driver(NULL) is already computed, and thus the check fails to work as intended.

Will and I both hit this with the platform bus, but the pattern here is the same, so fix it for correctness too.

Fixes bcb81ac6ae3c ("iommu: Get DT/ACPI parsing into the proper probe path")
---
 drivers/pci/pci-driver.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/pci/pci-driver.c +3 -3
index 67db34fd10ee..01e6aea1b0c7 100644
@@ -1628,7 +1628,7 @@ static int pci_bus_num_vf(struct device *dev)
  */
 static int pci_dma_configure(struct device *dev)
 {
-	struct pci_driver *driver = to_pci_driver(dev->driver);
+	const struct device_driver *drv = READ_ONCE(dev->driver);
 	struct device *bridge;
 	int ret = 0;
 
@@ -1645,8 +1645,8 @@ static int pci_dma_configure(struct device *dev)
 
 	pci_put_host_bridge_device(bridge);
 
-	/* @driver may not be valid when we're called from the IOMMU layer */
-	if (!ret && dev->driver && !driver->driver_managed_dma) {
+	/* @drv may not be valid when we're called from the IOMMU layer */
+	if (!ret && drv && !to_pci_driver(drv)->driver_managed_dma) {
 		ret = iommu_device_use_default_domain(dev);
 		if (ret)
 			arch_teardown_dma_ops(dev);
Signature · 1 line
2.39.5
author maintainer
#406 lore raw

PATCH 396/627 watchdog: ziirave_wdt: check record length in ziirave_firm_verify()

In reply to Greg Kroah-Hartman

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

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

From: Dan Carpenter <[email protected]>

[ Upstream commit 8b61d8ca751bc15875b50e0ff6ac3ba0cf95a529 ]

The "rec->len" value comes from the firmware. We generally do trust firmware, but it's always better to double check. If the length value is too large it would lead to memory corruption when we set "data[i] = ret;"

---
 drivers/watchdog/ziirave_wdt.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/watchdog/ziirave_wdt.c +3 -0
index fcc1ba02e75b..5c6e3fa001d8 100644
@@ -302,6 +302,9 @@ static int ziirave_firm_verify(struct watchdog_device *wdd,
 		const u16 len = be16_to_cpu(rec->len);
 		const u32 addr = be32_to_cpu(rec->addr);
 
+		if (len > sizeof(data))
+			return -EINVAL;
+
 		if (ziirave_firm_addr_readonly(addr))
 			continue;
 
Signature · 1 line
2.39.5