Skip to contentSkip to conversation

6.18.40-rc1 review

PATCH In review Greg Kroah-Hartman 11 participants 1,624 messages Jul 21 – Aug 26 Page 11 of 12

Subsystems (13)
lore
Share & tools
lore
author maintainer
#1411 lore raw

PATCH 1405/1611 net: ena: clean up XDP TX queues when regular TX setup fails

In reply to Greg Kroah-Hartman

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

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

From: Dawei Feng <[email protected]>

commit 1bd6676254b4ab6acd44b662b5e92822c036463a upstream.

create_queues_with_size_backoff() creates XDP TX queues before setting up the regular TX path. If the subsequent allocation or creation of regular TX queues fails, the error handling paths omit the teardown of the XDP TX queues, leading to a resource leak.

Fix this by explicitly destroying the XDP TX queue subset at the two missing failure points.

The bug was first flagged by an experimental analysis tool we are developing for kernel memory-management bugs while analyzing v6.13-rc1. The tool is still under development and is not yet publicly available. Manual inspection confirms that the bug is still present in v7.1-rc7.

An x86_64 allyesconfig build showed no new warnings. As we do not have an ENA device to test with, no runtime testing was able to be performed.

---
 drivers/net/ethernet/amazon/ena/ena_netdev.c |   23 +++++++++++++++++++++--
 1 file changed, 21 insertions(+), 2 deletions(-)
drivers/net/ethernet/amazon/ena/ena_netdev.c +21 -2
@@ -752,6 +752,18 @@ static void ena_destroy_all_tx_queues(st
 	}
 }
 
+static void ena_destroy_xdp_tx_queues(struct ena_adapter *adapter)
+{
+	u16 ena_qid;
+	int i;
+
+	for (i = adapter->xdp_first_ring;
+	     i < adapter->xdp_first_ring + adapter->xdp_num_queues; i++) {
+		ena_qid = ENA_IO_TXQ_IDX(i);
+		ena_com_destroy_io_queue(adapter->ena_dev, ena_qid);
+	}
+}
+
 static void ena_destroy_all_rx_queues(struct ena_adapter *adapter)
 {
 	u16 ena_qid;
@@ -2078,14 +2090,21 @@ static int create_queues_with_size_backo
 		rc = ena_setup_tx_resources_in_range(adapter,
 						     0,
 						     adapter->num_io_queues);
-		if (rc)
+		if (rc) {
+			ena_destroy_xdp_tx_queues(adapter);
+			ena_free_all_io_tx_resources_in_range(adapter,
+							      adapter->xdp_first_ring,
+							      adapter->xdp_num_queues);
 			goto err_setup_tx;
+		}
 
 		rc = ena_create_io_tx_queues_in_range(adapter,
 						      0,
 						      adapter->num_io_queues);
-		if (rc)
+		if (rc) {
+			ena_destroy_xdp_tx_queues(adapter);
 			goto err_create_tx_queues;
+		}
 
 		rc = ena_setup_all_rx_resources(adapter);
 		if (rc)
author maintainer
#1412 lore raw

PATCH 1406/1611 net: ip6_vti: require CAP_NET_ADMIN in the device netns for changelink

In reply to Greg Kroah-Hartman

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

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

From: Maoyi Xie <[email protected]>

commit e2ac3b242c37dff323a964962e43854f4b1a2b79 upstream.

vti6_changelink() operates on at most two netns, dev_net(dev) and the tunnel link netns t->net. They differ once the device is created in or moved to a netns other than the one the request runs in. The rtnl changelink path checks CAP_NET_ADMIN only against dev_net(dev), so a caller privileged there but not in t->net can rewrite a tunnel that lives in t->net.

Gate vti6_changelink() on rtnl_dev_link_net_capable() at its top, before any attribute is parsed.

---
 net/ipv6/ip6_vti.c |    3 +++
 1 file changed, 3 insertions(+)
net/ipv6/ip6_vti.c +3 -0
@@ -1046,6 +1046,9 @@ static int vti6_changelink(struct net_de
 	struct __ip6_tnl_parm p;
 	struct vti6_net *ip6n;
 
+	if (!rtnl_dev_link_net_capable(dev, net))
+		return -EPERM;
+
 	ip6n = net_generic(net, vti6_net_id);
 	if (dev == ip6n->fb_tnl_dev)
 		return -EINVAL;
author maintainer
#1413 lore raw

PATCH 1407/1611 net: ip_vti: require CAP_NET_ADMIN in the device netns for changelink

In reply to Greg Kroah-Hartman

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

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

From: Maoyi Xie <[email protected]>

commit 95cceadbfd52d7239bd730afdda0655287d77425 upstream.

vti_changelink() operates on at most two netns, dev_net(dev) and the tunnel link netns t->net. They differ once the device is created in or moved to a netns other than the one the request runs in. The rtnl changelink path checks CAP_NET_ADMIN only against dev_net(dev), so a caller privileged there but not in t->net can rewrite a tunnel that lives in t->net.

Gate vti_changelink() on rtnl_dev_link_net_capable() at its top, before any attribute is parsed.

---
 net/ipv4/ip_vti.c |    3 +++
 1 file changed, 3 insertions(+)
net/ipv4/ip_vti.c +3 -0
@@ -596,6 +596,9 @@ static int vti_changelink(struct net_dev
 	struct ip_tunnel_parm_kern p;
 	__u32 fwmark = t->fwmark;
 
+	if (!rtnl_dev_link_net_capable(dev, t->net))
+		return -EPERM;
+
 	vti_netlink_parms(data, &p, &fwmark);
 	return ip_tunnel_changelink(dev, tb, &p, fwmark);
 }
author maintainer
#1414 lore raw

PATCH 1408/1611 net: ipip: require CAP_NET_ADMIN in the device netns for changelink

In reply to Greg Kroah-Hartman

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

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

From: Maoyi Xie <[email protected]>

commit 8211a26324667980a463c069469a818e71207e02 upstream.

ipip_changelink() operates on at most two netns, dev_net(dev) and the tunnel link netns t->net. They differ once the device is created in or moved to a netns other than the one the request runs in. The rtnl changelink path checks CAP_NET_ADMIN only against dev_net(dev), so a caller privileged there but not in t->net can rewrite a tunnel that lives in t->net.

Gate ipip_changelink() on rtnl_dev_link_net_capable() at its top, before any attribute is parsed.

---
 net/ipv4/ipip.c |    3 +++
 1 file changed, 3 insertions(+)
net/ipv4/ipip.c +3 -0
@@ -469,6 +469,9 @@ static int ipip_changelink(struct net_de
 	bool collect_md;
 	__u32 fwmark = t->fwmark;
 
+	if (!rtnl_dev_link_net_capable(dev, t->net))
+		return -EPERM;
+
 	if (ip_tunnel_netlink_encap_parms(data, &ipencap)) {
 		int err = ip_tunnel_encap_setup(t, &ipencap);
 
author maintainer
#1415 lore raw

PATCH 1409/1611 net: ip6_gre: require CAP_NET_ADMIN in the device netns for changelink

In reply to Greg Kroah-Hartman

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

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

From: Maoyi Xie <[email protected]>

commit f00a50876d2818bd6dc86fa98b3ef360884c53c8 upstream.

ip6gre_changelink() and ip6erspan_changelink() operate on at most two netns, dev_net(dev) and the tunnel link netns t->net. They differ once the device is created in or moved to a netns other than the one the request runs in. The rtnl changelink path checks CAP_NET_ADMIN only against dev_net(dev), so a caller privileged there but not in t->net can rewrite a tunnel that lives in t->net.

Gate both ops on rtnl_dev_link_net_capable() at their top, before any attribute is parsed.

---
 net/ipv6/ip6_gre.c |    6 ++++++
 1 file changed, 6 insertions(+)
net/ipv6/ip6_gre.c +6 -0
@@ -2046,6 +2046,9 @@ static int ip6gre_changelink(struct net_
 	struct ip6gre_net *ign = net_generic(t->net, ip6gre_net_id);
 	struct __ip6_tnl_parm p;
 
+	if (!rtnl_dev_link_net_capable(dev, t->net))
+		return -EPERM;
+
 	t = ip6gre_changelink_common(dev, tb, data, &p, extack);
 	if (IS_ERR(t))
 		return PTR_ERR(t);
@@ -2265,6 +2268,9 @@ static int ip6erspan_changelink(struct n
 	struct __ip6_tnl_parm p;
 	struct ip6gre_net *ign;
 
+	if (!rtnl_dev_link_net_capable(dev, t->net))
+		return -EPERM;
+
 	ign = net_generic(t->net, ip6gre_net_id);
 	t = ip6gre_changelink_common(dev, tb, data, &p, extack);
 	if (IS_ERR(t))
author maintainer
#1416 lore raw

PATCH 1410/1611 net: ip6_tunnel: require CAP_NET_ADMIN in the device netns for changelink

In reply to Greg Kroah-Hartman

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

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

From: Maoyi Xie <[email protected]>

commit 2496fa0b7d180b3ad356b514e7ff93bb14e6140a upstream.

ip6_tnl_changelink() operates on at most two netns, dev_net(dev) and the tunnel link netns t->net. They differ once the device is created in or moved to a netns other than the one the request runs in. The rtnl changelink path checks CAP_NET_ADMIN only against dev_net(dev), so a caller privileged there but not in t->net can rewrite a tunnel that lives in t->net.

Gate ip6_tnl_changelink() on rtnl_dev_link_net_capable() at its top, before any attribute is parsed.

---
 net/ipv6/ip6_tunnel.c |    3 +++
 1 file changed, 3 insertions(+)
net/ipv6/ip6_tunnel.c +3 -0
@@ -2068,6 +2068,9 @@ static int ip6_tnl_changelink(struct net
 	struct ip6_tnl_net *ip6n = net_generic(net, ip6_tnl_net_id);
 	struct ip_tunnel_encap ipencap;
 
+	if (!rtnl_dev_link_net_capable(dev, net))
+		return -EPERM;
+
 	if (dev == ip6n->fb_tnl_dev) {
 		if (ip_tunnel_netlink_encap_parms(data, &ipencap)) {
 			/* iproute2 always sets TUNNEL_ENCAP_FLAG_CSUM6, so
author maintainer
#1417 lore raw

PATCH 1411/1611 octeontx2-af: Free BPID bitmap on setup failure

In reply to Greg Kroah-Hartman

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

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

From: Haoxiang Li <[email protected]>

commit 36323f54cd323122a1be89ab2c316a6e55a94e30 upstream.

nix_setup_bpids() allocates bp->bpids with rvu_alloc_bitmap(), which uses a plain kcalloc(). If any of the following devm_kcalloc() allocations for the BPID mapping arrays fails, the function returns without freeing the bitmap. Free the BPID bitmap before returning from those error paths.

---
 drivers/net/ethernet/marvell/octeontx2/af/rvu_nix.c |   11 ++++++++---
 1 file changed, 8 insertions(+), 3 deletions(-)
drivers/net/ethernet/marvell/octeontx2/af/rvu_nix.c +8 -3
@@ -527,19 +527,24 @@ static int nix_setup_bpids(struct rvu *r
 	bp->fn_map = devm_kcalloc(rvu->dev, bp->bpids.max,
 				  sizeof(u16), GFP_KERNEL);
 	if (!bp->fn_map)
-		return -ENOMEM;
+		goto free_bpids;
 
 	bp->intf_map = devm_kcalloc(rvu->dev, bp->bpids.max,
 				    sizeof(u8), GFP_KERNEL);
 	if (!bp->intf_map)
-		return -ENOMEM;
+		goto free_bpids;
 
 	bp->ref_cnt = devm_kcalloc(rvu->dev, bp->bpids.max,
 				   sizeof(u8), GFP_KERNEL);
 	if (!bp->ref_cnt)
-		return -ENOMEM;
+		goto free_bpids;
 
 	return 0;
+
+free_bpids:
+	rvu_free_bitmap(&bp->bpids);
+	bp->bpids.bmap = NULL;
+	return -ENOMEM;
 }
 
 void rvu_nix_flr_free_bpids(struct rvu *rvu, u16 pcifunc)
author maintainer
#1418 lore raw

PATCH 1412/1611 ieee802154: admin-gate legacy LLSEC dump operations

In reply to Greg Kroah-Hartman

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

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

From: Michael Bommarito <[email protected]>

commit 9c1e0b6d49471a712511d23fc9d06901561135e8 upstream.

In net/ieee802154/netlink.c, the legacy IEEE802154_NL family ops table builds the LLSEC dump entries (LLSEC_LIST_KEY, LLSEC_LIST_DEV, LLSEC_LIST_DEVKEY, LLSEC_LIST_SECLEVEL) with IEEE802154_DUMP() which sets no .flags, so generic netlink runs them ungated. The modern nl802154 family admin-gates the equivalent reads via NL802154_CMD_GET_SEC_KEY and friends with .flags = GENL_ADMIN_PERM.

Any local uid that can open AF_NETLINK / NETLINK_GENERIC can resolve the "802.15.4 MAC" family and dump LLSEC_LIST_KEY on any wpan netdev that has an LLSEC key installed; the dump handler writes the raw 16-byte AES-128 key bytes (IEEE802154_ATTR_LLSEC_KEY_BYTES, copied verbatim from struct ieee802154_llsec_key.key) into the reply. Recovering the AES key compromises 802.15.4 LLSEC link confidentiality and authenticity, since LLSEC uses CCM* and the same key authenticates and encrypts frames.

Impact: any local uid with no capabilities can read the raw 16-byte AES-128 LLSEC key from the kernel keytable on any wpan netdev that has an administrator-installed LLSEC key, by issuing an LLSEC_LIST_KEY dump on the legacy IEEE802154_NL generic-netlink family.

Introduce IEEE802154_DUMP_PRIV() mirroring IEEE802154_DUMP() but setting .flags = GENL_ADMIN_PERM, and use it for the four LLSEC dump entries. LIST_PHY and LIST_IFACE retain IEEE802154_DUMP() because the modern nl802154 family exposes their equivalents to unprivileged readers by design (NL802154_CMD_GET_WPAN_PHY and NL802154_CMD_GET_INTERFACE carry "can be retrieved by unprivileged users" annotations).

Fixes 3e9c156e2c21 ("ieee802154: add netlink interfaces for llsec")
Assisted-by Claude:claude-opus-4-7
---
 net/ieee802154/ieee802154.h |    8 ++++++++
 net/ieee802154/netlink.c    |   16 ++++++++--------
 2 files changed, 16 insertions(+), 8 deletions(-)
net/ieee802154/ieee802154.h +8 -0
@@ -23,6 +23,14 @@ void ieee802154_nl_exit(void);
 		.dumpit	= _dump,			\
 	}
 
+#define IEEE802154_DUMP_PRIV(_cmd, _func, _dump)	\
+	{						\
+		.cmd	= _cmd,				\
+		.doit	= _func,			\
+		.dumpit	= _dump,			\
+		.flags	= GENL_ADMIN_PERM,		\
+	}
+
 struct genl_info;
 
 struct sk_buff *ieee802154_nl_create(int flags, u8 req);
net/ieee802154/netlink.c +8 -8
@@ -98,20 +98,20 @@ static const struct genl_small_ops ieee8
 	IEEE802154_OP(IEEE802154_SET_MACPARAMS, ieee802154_set_macparams),
 	IEEE802154_OP(IEEE802154_LLSEC_GETPARAMS, ieee802154_llsec_getparams),
 	IEEE802154_OP(IEEE802154_LLSEC_SETPARAMS, ieee802154_llsec_setparams),
-	IEEE802154_DUMP(IEEE802154_LLSEC_LIST_KEY, NULL,
-			ieee802154_llsec_dump_keys),
+	IEEE802154_DUMP_PRIV(IEEE802154_LLSEC_LIST_KEY, NULL,
+			     ieee802154_llsec_dump_keys),
 	IEEE802154_OP(IEEE802154_LLSEC_ADD_KEY, ieee802154_llsec_add_key),
 	IEEE802154_OP(IEEE802154_LLSEC_DEL_KEY, ieee802154_llsec_del_key),
-	IEEE802154_DUMP(IEEE802154_LLSEC_LIST_DEV, NULL,
-			ieee802154_llsec_dump_devs),
+	IEEE802154_DUMP_PRIV(IEEE802154_LLSEC_LIST_DEV, NULL,
+			     ieee802154_llsec_dump_devs),
 	IEEE802154_OP(IEEE802154_LLSEC_ADD_DEV, ieee802154_llsec_add_dev),
 	IEEE802154_OP(IEEE802154_LLSEC_DEL_DEV, ieee802154_llsec_del_dev),
-	IEEE802154_DUMP(IEEE802154_LLSEC_LIST_DEVKEY, NULL,
-			ieee802154_llsec_dump_devkeys),
+	IEEE802154_DUMP_PRIV(IEEE802154_LLSEC_LIST_DEVKEY, NULL,
+			     ieee802154_llsec_dump_devkeys),
 	IEEE802154_OP(IEEE802154_LLSEC_ADD_DEVKEY, ieee802154_llsec_add_devkey),
 	IEEE802154_OP(IEEE802154_LLSEC_DEL_DEVKEY, ieee802154_llsec_del_devkey),
-	IEEE802154_DUMP(IEEE802154_LLSEC_LIST_SECLEVEL, NULL,
-			ieee802154_llsec_dump_seclevels),
+	IEEE802154_DUMP_PRIV(IEEE802154_LLSEC_LIST_SECLEVEL, NULL,
+			     ieee802154_llsec_dump_seclevels),
 	IEEE802154_OP(IEEE802154_LLSEC_ADD_SECLEVEL,
 		      ieee802154_llsec_add_seclevel),
 	IEEE802154_OP(IEEE802154_LLSEC_DEL_SECLEVEL,
author maintainer
#1419 lore raw

PATCH 1413/1611 ieee802154: allow legacy LLSEC ADD/DEL ops to pass strict validation

In reply to Greg Kroah-Hartman

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

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

From: Michael Bommarito <[email protected]>

commit a6bfdfcc6711d1d5a92e98644359dedc67c0c858 upstream.

The LLSEC ADD/DEL doit handlers under the legacy IEEE802154_NL family consume IEEE802154_ATTR_LLSEC_KEY_BYTES and IEEE802154_ATTR_LLSEC_KEY_USAGE_COMMANDS, both declared in net/ieee802154/nl_policy.c as bare length entries with no .type (defaulting to NLA_UNSPEC). Generic netlink strict validation rejects all NLA_UNSPEC attributes via validate_nla(), so every LLSEC_ADD_KEY, LLSEC_DEL_KEY, LLSEC_ADD_DEV, LLSEC_DEL_DEV, LLSEC_ADD_DEVKEY, LLSEC_DEL_DEVKEY, LLSEC_ADD_SECLEVEL, and LLSEC_DEL_SECLEVEL request fails at the dispatcher with "Unsupported attribute" before reaching the handler.

The doit path has been silently dead since strict validation became the default for genl families that do not opt out. The dump path is unaffected because dump requests carry no LLSEC attributes to validate, which is why the LLSEC_LIST_KEY read remained reachable (patch 1/2). Introduce IEEE802154_OP_RELAXED() mirroring IEEE802154_OP() but with .validate = GENL_DONT_VALIDATE_STRICT, and use it for the eight legacy LLSEC mutate ops so admin-driven LLSEC configuration via the legacy interface works again.

Fixes 3e9c156e2c21 ("ieee802154: add netlink interfaces for llsec")
Assisted-by Claude:claude-opus-4-7
---
 net/ieee802154/ieee802154.h |    9 +++++++++
 net/ieee802154/netlink.c    |   20 ++++++++++----------
 2 files changed, 19 insertions(+), 10 deletions(-)
net/ieee802154/ieee802154.h +9 -0
@@ -16,6 +16,15 @@ void ieee802154_nl_exit(void);
 		.flags	= GENL_ADMIN_PERM,		\
 	}
 
+#define IEEE802154_OP_RELAXED(_cmd, _func)		\
+	{						\
+		.cmd		= _cmd,			\
+		.doit		= _func,		\
+		.dumpit		= NULL,			\
+		.flags		= GENL_ADMIN_PERM,	\
+		.validate	= GENL_DONT_VALIDATE_STRICT,\
+	}
+
 #define IEEE802154_DUMP(_cmd, _func, _dump)		\
 	{						\
 		.cmd	= _cmd,				\
net/ieee802154/netlink.c +10 -10
@@ -100,22 +100,22 @@ static const struct genl_small_ops ieee8
 	IEEE802154_OP(IEEE802154_LLSEC_SETPARAMS, ieee802154_llsec_setparams),
 	IEEE802154_DUMP_PRIV(IEEE802154_LLSEC_LIST_KEY, NULL,
 			     ieee802154_llsec_dump_keys),
-	IEEE802154_OP(IEEE802154_LLSEC_ADD_KEY, ieee802154_llsec_add_key),
-	IEEE802154_OP(IEEE802154_LLSEC_DEL_KEY, ieee802154_llsec_del_key),
+	IEEE802154_OP_RELAXED(IEEE802154_LLSEC_ADD_KEY, ieee802154_llsec_add_key),
+	IEEE802154_OP_RELAXED(IEEE802154_LLSEC_DEL_KEY, ieee802154_llsec_del_key),
 	IEEE802154_DUMP_PRIV(IEEE802154_LLSEC_LIST_DEV, NULL,
 			     ieee802154_llsec_dump_devs),
-	IEEE802154_OP(IEEE802154_LLSEC_ADD_DEV, ieee802154_llsec_add_dev),
-	IEEE802154_OP(IEEE802154_LLSEC_DEL_DEV, ieee802154_llsec_del_dev),
+	IEEE802154_OP_RELAXED(IEEE802154_LLSEC_ADD_DEV, ieee802154_llsec_add_dev),
+	IEEE802154_OP_RELAXED(IEEE802154_LLSEC_DEL_DEV, ieee802154_llsec_del_dev),
 	IEEE802154_DUMP_PRIV(IEEE802154_LLSEC_LIST_DEVKEY, NULL,
 			     ieee802154_llsec_dump_devkeys),
-	IEEE802154_OP(IEEE802154_LLSEC_ADD_DEVKEY, ieee802154_llsec_add_devkey),
-	IEEE802154_OP(IEEE802154_LLSEC_DEL_DEVKEY, ieee802154_llsec_del_devkey),
+	IEEE802154_OP_RELAXED(IEEE802154_LLSEC_ADD_DEVKEY, ieee802154_llsec_add_devkey),
+	IEEE802154_OP_RELAXED(IEEE802154_LLSEC_DEL_DEVKEY, ieee802154_llsec_del_devkey),
 	IEEE802154_DUMP_PRIV(IEEE802154_LLSEC_LIST_SECLEVEL, NULL,
 			     ieee802154_llsec_dump_seclevels),
-	IEEE802154_OP(IEEE802154_LLSEC_ADD_SECLEVEL,
-		      ieee802154_llsec_add_seclevel),
-	IEEE802154_OP(IEEE802154_LLSEC_DEL_SECLEVEL,
-		      ieee802154_llsec_del_seclevel),
+	IEEE802154_OP_RELAXED(IEEE802154_LLSEC_ADD_SECLEVEL,
+			      ieee802154_llsec_add_seclevel),
+	IEEE802154_OP_RELAXED(IEEE802154_LLSEC_DEL_SECLEVEL,
+			      ieee802154_llsec_del_seclevel),
 };
 
 static const struct genl_multicast_group ieee802154_mcgrps[] = {
author maintainer
#1420 lore raw

PATCH 1414/1611 ieee802154: ca8210: fix cas_ctl leak on spi_async failure

In reply to Greg Kroah-Hartman

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

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

From: Shitalkumar Gandhi <[email protected]>

commit e09390e439bd7cca30dd10893b1f64802961667a upstream.

ca8210_spi_transfer() allocates cas_ctl with kzalloc_obj(GFP_ATOMIC) and relies entirely on the SPI completion callback ca8210_spi_transfer_complete() to free it.

The spi_async() API only invokes the completion callback on successful submission. On failure it returns a negative error code without ever queuing the callback, which leaves cas_ctl and its embedded spi_message and spi_transfer orphaned. Every kfree(cas_ctl) in the driver is inside the completion callback, so there is no other reclamation path.

ca8210_spi_transfer() is called from ca8210_spi_exchange(), the interrupt handler ca8210_interrupt_handler(), and from the retry path inside the completion callback itself. The exchange and interrupt handler paths loop on -EBUSY, so under sustained SPI bus contention every retry iteration leaks a fresh cas_ctl (~600 bytes per occurrence).

Fix it by freeing cas_ctl on the spi_async() error path. While here, correct the misleading error string: the function calls spi_async(), not spi_sync().

---
 drivers/net/ieee802154/ca8210.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/net/ieee802154/ca8210.c +2 -1
@@ -920,9 +920,10 @@ static int ca8210_spi_transfer(
 	if (status < 0) {
 		dev_crit(
 			&spi->dev,
-			"status %d from spi_sync in write\n",
+			"status %d from spi_async in write\n",
 			status
 		);
+		kfree(cas_ctl);
 	}
 
 	return status;
author maintainer
#1421 lore raw

PATCH 1415/1611 ieee802154: ca8210: fix pointer truncation in kfifo on 64-bit

In reply to Greg Kroah-Hartman

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

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

From: Shitalkumar Gandhi <[email protected]>

commit 6d7f7bcf225b2d566176bf6229dbd1252940cb3c upstream.

ca8210_test_int_driver_write() and ca8210_test_int_user_read() exchange a kmalloc'd buffer pointer through a struct kfifo, but pass a literal '4' as the byte count to kfifo_in()/kfifo_out().

This is correct on 32-bit (pointer = 4 bytes), but on 64-bit only the low 4 bytes of the 8-byte pointer are written into the FIFO. The reader then reads back 4 bytes into an 8-byte local pointer variable, leaving the upper 4 bytes uninitialized stack data. The first dereference of the reconstructed pointer (fifo_buffer[1]) accesses an arbitrary kernel address and generally results in an oops.

Use sizeof(fifo_buffer) so the byte count matches pointer width on every architecture.

The driver has no architecture restriction in Kconfig, so any 64-bit build with CONFIG_IEEE802154_CA8210_DEBUGFS=y is exposed. Issue has been latent since the driver was added in 2017 because it is most commonly deployed on 32-bit MCUs.

Found via a custom Coccinelle semantic patch hunting for short-byte kfifo I/O on byte-mode kfifos used to shuttle pointers.

---
 drivers/net/ieee802154/ca8210.c |    6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/net/ieee802154/ca8210.c +4 -2
@@ -595,7 +595,7 @@ static int ca8210_test_int_driver_write(
 	fifo_buffer = kmemdup(buf, len, GFP_KERNEL);
 	if (!fifo_buffer)
 		return -ENOMEM;
-	kfifo_in(&test->up_fifo, &fifo_buffer, 4);
+	kfifo_in(&test->up_fifo, &fifo_buffer, sizeof(fifo_buffer));
 	wake_up_interruptible(&priv->test.readq);
 
 	return 0;
@@ -2527,6 +2527,7 @@ static ssize_t ca8210_test_int_user_read
 	struct ca8210_priv *priv = filp->private_data;
 	unsigned char *fifo_buffer;
 	unsigned long bytes_not_copied;
+	unsigned int copied;
 
 	if (filp->f_flags & O_NONBLOCK) {
 		/* Non-blocking mode */
@@ -2540,7 +2541,8 @@ static ssize_t ca8210_test_int_user_read
 		);
 	}
 
-	if (kfifo_out(&priv->test.up_fifo, &fifo_buffer, 4) != 4) {
+	copied = kfifo_out(&priv->test.up_fifo, &fifo_buffer, sizeof(fifo_buffer));
+	if (copied != sizeof(fifo_buffer)) {
 		dev_err(
 			&priv->spi->dev,
 			"test_interface: Wrong number of elements popped from upstream fifo\n"
author maintainer
#1422 lore raw

PATCH 1416/1611 gve: fix header buffer corruption with header-split and HW-GRO

In reply to Greg Kroah-Hartman

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

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

From: Ankit Garg <[email protected]>

commit d676c9a73bdcd8237425dbb826f2bd1a25c36e40 upstream.

The DQO RX datapath programs a per-buffer-queue-descriptor header_buf_addr at post time and reads the split header back at completion time. Both the post and the read currently index the header buffer by queue position rather than by the buffer's identity:

- post (gve_rx_post_buffers_dqo): header_buf_addr is computed from

    bufq->tail

- read (gve_rx_dqo): the header is read from desc_idx (the completion queue head index)

This relies on the buffer-queue index and the completion-queue index being equal for the start of every packet, i.e. on the device consuming posted buffers and returning completions in the exact same order. That assumption does not hold once HW-GRO is enabled with multiple
flows: coalesced segments are accepted and completed in an order that may differ from the order buffers were posted, and segments from different flows may interleave.

That results in two problems:

1. Wrong header slot on read. Because the read offset is derived from the completion index (desc_idx) while the device wrote the header to the address programmed for the buffer's buf_id, the driver can copy a header belonging to a different packet. This shows up as throughput drop (about 30% drop and large numbers of TCP retransmissions) with header-split and HW-GRO both enabled and many streams.

2. Header buffer reused while still owned by the device. The driver advances bufq->head by one per completion and re-posts buffers based on that. Arrival of N RX completions only guarantees that at least N RX buffer descriptors have been read by the device. It does not guarantee that the device has relinquished the ownership of all the buffers corresponding to those N descriptors. With out-of-order completions (e.g. the completion for a packet copied into buffer N arrives before the completion for a packet copied into buffer N-1), the driver can re-post and overwrite a header buffer that the device is still going to write into, corrupting the header of a packet whose completion has not yet been processed.

Fix both issues by indexing the header buffer by buf_id on both the post and read paths. Reading from buf_id's slot is therefore always correct regardless of completion ordering (fixes problem 1).

Indexing by buf_id also ties each header slot to the lifetime of its buffer state. A buffer state is only returned to the free/recycle lists when its own completion (buf_id) is processed, so its header slot can only be re-posted after the device is done with it. This makes header slot reuse safe under out-of-order completions (fixes problem 2).

Allocate (gve_rx_alloc_hdr_bufs) and free (gve_rx_free_hdr_bufs) the header buffers based on num_buf_states to match the buf_id indexing.

---
 drivers/net/ethernet/google/gve/gve_rx_dqo.c |   28 +++++++++++++++++----------
 1 file changed, 18 insertions(+), 10 deletions(-)
drivers/net/ethernet/google/gve/gve_rx_dqo.c +18 -10
@@ -21,11 +21,13 @@
 static void gve_rx_free_hdr_bufs(struct gve_priv *priv, struct gve_rx_ring *rx)
 {
 	struct device *hdev = &priv->pdev->dev;
-	int buf_count = rx->dqo.bufq.mask + 1;
 
 	if (rx->dqo.hdr_bufs.data) {
-		dma_free_coherent(hdev, priv->header_buf_size * buf_count,
-				  rx->dqo.hdr_bufs.data, rx->dqo.hdr_bufs.addr);
+		size_t size =
+			(size_t)priv->header_buf_size * rx->dqo.num_buf_states;
+
+		dma_free_coherent(hdev, size, rx->dqo.hdr_bufs.data,
+				  rx->dqo.hdr_bufs.addr);
 		rx->dqo.hdr_bufs.data = NULL;
 	}
 }
@@ -250,7 +252,7 @@ int gve_rx_alloc_ring_dqo(struct gve_pri
 
 	/* Allocate header buffers for header-split */
 	if (cfg->enable_header_split)
-		if (gve_rx_alloc_hdr_bufs(priv, rx, buffer_queue_slots))
+		if (gve_rx_alloc_hdr_bufs(priv, rx, rx->dqo.num_buf_states))
 			goto err;
 
 	/* Allocate RX completion queue */
@@ -379,10 +381,13 @@ void gve_rx_post_buffers_dqo(struct gve_
 			break;
 		}
 
-		if (rx->dqo.hdr_bufs.data)
+		if (rx->dqo.hdr_bufs.data) {
+			u16 buf_id = le16_to_cpu(desc->buf_id);
+
 			desc->header_buf_addr =
 				cpu_to_le64(rx->dqo.hdr_bufs.addr +
-					    priv->header_buf_size * bufq->tail);
+					(size_t)priv->header_buf_size * buf_id);
+		}
 
 		bufq->tail = (bufq->tail + 1) & bufq->mask;
 		complq->num_free_slots--;
@@ -799,10 +804,13 @@ static int gve_rx_dqo(struct napi_struct
 		int unsplit = 0;
 
 		if (hdr_len && !hbo) {
-			rx->ctx.skb_head = gve_rx_copy_data(priv->dev, napi,
-							    rx->dqo.hdr_bufs.data +
-							    desc_idx * priv->header_buf_size,
-							    hdr_len);
+			size_t offset =
+				(size_t)buffer_id * priv->header_buf_size;
+
+			rx->ctx.skb_head =
+				gve_rx_copy_data(priv->dev, napi,
+						 rx->dqo.hdr_bufs.data + offset,
+						 hdr_len);
 			if (unlikely(!rx->ctx.skb_head))
 				goto error;
 			rx->ctx.skb_tail = rx->ctx.skb_head;
author maintainer
#1423 lore raw

PATCH 1417/1611 platform/x86/amd/pmc: Avoid logging "(null)" for DMI values

In reply to Greg Kroah-Hartman

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

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

From: Daniel Gibson <[email protected]>

commit a0738abd042f7406edd2175a819cf2e66388ed97 upstream.

dmi_get_system_info(...) can return NULL. Using that as %s arguments of dev_info() would log "(null)" (as part of a message like '... System Vendor: "(null)", Product Name: "(null)" ...'), which may be confusing for users.

Use Elvis operator to print "(Unknown)" instead.

---
 drivers/platform/x86/amd/pmc/pmc.c |   10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)
drivers/platform/x86/amd/pmc/pmc.c +5 -5
@@ -660,11 +660,11 @@ static bool amd_pmc_want_suspend_delay(s
 	} else if (delay_suspend == 1) {
 		if (!intermediate_wakeup)
 			dev_info(pdev->dev, "Delaying suspend by 2.5s because delay_suspend=1. If this solves problems on your machine, please report this whole line to: [email protected] so it can be automatically detected as affected in the future. System Vendor: \"%s\" Product Name: \"%s\" Product Family: \"%s\" Board Vendor: \"%s\" Board Name: \"%s\"\n",
-				 dmi_get_system_info(DMI_SYS_VENDOR),
-				 dmi_get_system_info(DMI_PRODUCT_NAME),
-				 dmi_get_system_info(DMI_PRODUCT_FAMILY),
-				 dmi_get_system_info(DMI_BOARD_VENDOR),
-				 dmi_get_system_info(DMI_BOARD_NAME));
+				 dmi_get_system_info(DMI_SYS_VENDOR) ?: "(Unknown)",
+				 dmi_get_system_info(DMI_PRODUCT_NAME) ?: "(Unknown)",
+				 dmi_get_system_info(DMI_PRODUCT_FAMILY) ?: "(Unknown)",
+				 dmi_get_system_info(DMI_BOARD_VENDOR) ?: "(Unknown)",
+				 dmi_get_system_info(DMI_BOARD_NAME) ?: "(Unknown)");
 		return true;
 	}
 	return false;
author maintainer
#1424 lore raw

PATCH 1418/1611 net/sched: sch_teql: move rcu_read_lock()/spin_lock() from _bh variants

In reply to Greg Kroah-Hartman

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

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

From: Jamal Hadi Salim <[email protected]>

commit 6301f6a34ed86fe6f3b7b3211ea069f3677fc559 upstream.

This is a followup based on sashiko comments [1] on commit e5b811fe7931 ("net/sched: sch_teql: Introduce slaves_lock to avoid race condition and UAF")

Use plain rcu_read_lock()/spin_lock() in teql_master_xmit() instead of the _bh variants, since ndo_start_xmit is already invoked with BH disabled by the core stack and the _bh primitives can warn in_hardirq() when xmit is reached through netpoll or a softirq xmit path with hard IRQs disabled.

Moves rcu_read_lock() after restart: label + adds rcu_read_unlock() before goto restart (fixes the unbounded RCU hold across retries)

[1] https://sashiko.dev/#/patchset/20260628111229.669751-1-jhs%40mojatatu.com

Fixes e5b811fe7931 ("net/sched: sch_teql: Introduce slaves_lock to avoid race condition and UAF")
---
 net/sched/sch_teql.c |   27 ++++++++++++++-------------
 1 file changed, 14 insertions(+), 13 deletions(-)
net/sched/sch_teql.c +14 -13
@@ -311,14 +311,14 @@ static netdev_tx_t teql_master_xmit(stru
 	int subq = skb_get_queue_mapping(skb);
 	struct sk_buff *skb_res = NULL;
 
-	rcu_read_lock_bh();
-
-	start = rcu_dereference_bh(master->slaves);
-
 restart:
 	nores = 0;
 	busy = 0;
 
+	rcu_read_lock();
+
+	start = rcu_dereference(master->slaves);
+
 	q = start;
 	if (!q)
 		goto drop;
@@ -345,17 +345,17 @@ restart:
 				    netdev_start_xmit(skb, slave, slave_txq, false) ==
 				    NETDEV_TX_OK) {
 					__netif_tx_unlock(slave_txq);
-					spin_lock_bh(&master->slaves_lock);
+					spin_lock(&master->slaves_lock);
 					if (rcu_dereference_protected(master->slaves,
 								      lockdep_is_held(&master->slaves_lock)) == q)
 						rcu_assign_pointer(master->slaves,
 								   rcu_dereference_protected(NEXT_SLAVE(q),
 											     lockdep_is_held(&master->slaves_lock)));
-					spin_unlock_bh(&master->slaves_lock);
+					spin_unlock(&master->slaves_lock);
 					netif_wake_queue(dev);
 					master->tx_packets++;
 					master->tx_bytes += length;
-					rcu_read_unlock_bh();
+					rcu_read_unlock();
 					return NETDEV_TX_OK;
 				}
 				__netif_tx_unlock(slave_txq);
@@ -364,37 +364,38 @@ restart:
 				busy = 1;
 			break;
 		case 1:
-			spin_lock_bh(&master->slaves_lock);
+			spin_lock(&master->slaves_lock);
 			if (rcu_dereference_protected(master->slaves,
 						      lockdep_is_held(&master->slaves_lock)) == q)
 				rcu_assign_pointer(master->slaves,
 						   rcu_dereference_protected(NEXT_SLAVE(q),
 									     lockdep_is_held(&master->slaves_lock)));
-			spin_unlock_bh(&master->slaves_lock);
-			rcu_read_unlock_bh();
+			spin_unlock(&master->slaves_lock);
+			rcu_read_unlock();
 			return NETDEV_TX_OK;
 		default:
 			nores = 1;
 			break;
 		}
 		__skb_pull(skb, skb_network_offset(skb));
-	} while ((q = rcu_dereference_bh(NEXT_SLAVE(q))) != start);
+	} while ((q = rcu_dereference(NEXT_SLAVE(q))) != start);
 
 	if (nores && skb_res == NULL) {
 		skb_res = skb;
+		rcu_read_unlock();
 		goto restart;
 	}
 
 	if (busy) {
 		netif_stop_queue(dev);
-		rcu_read_unlock_bh();
+		rcu_read_unlock();
 		return NETDEV_TX_BUSY;
 	}
 	master->tx_errors++;
 
 drop:
 	master->tx_dropped++;
-	rcu_read_unlock_bh();
+	rcu_read_unlock();
 	dev_kfree_skb(skb);
 	return NETDEV_TX_OK;
 }
author maintainer
#1425 lore raw

PATCH 1419/1611 drm/xe/userptr: Stub notifier_lock helpers when DRM_GPUSVM=n

In reply to Greg Kroah-Hartman

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

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

From: Shuicheng Lin <[email protected]>

commit f5ef65adf81da3dbce4e692e48c1754c0bb95da0 upstream.

When CONFIG_DRM_GPUSVM=n (e.g. um-allyesconfig), the only caller of xe_pt_svm_userptr_notifier_lock() is compiled out, triggering:

drivers/gpu/drm/xe/xe_pt.c:1418:13: warning:
'xe_pt_svm_userptr_notifier_lock' defined but not used [-Wunused-function]

The helpers cannot simply be removed in this case: the matching xe_pt_svm_userptr_notifier_unlock() is also referenced from xe_pt_update_ops_run(), which lives outside any DRM_GPUSVM ifdef and is gated only at runtime by pt_update_ops->needs_svm_lock. The symbol must exist in all builds.

Provide empty static inline stubs for !DRM_GPUSVM, matching the pattern used by xe_svm_notifier_lock()/_unlock() in xe_svm.h.

Fixes dca6e08c923a ("drm/xe/userptr: Hold notifier_lock for write on inject test path")
(cherry picked from commit 3359422bf0a1140e96d783a19a397686e580a3ca)
---
 drivers/gpu/drm/xe/xe_pt.c |    5 +++++
 1 file changed, 5 insertions(+)
drivers/gpu/drm/xe/xe_pt.c +5 -0
@@ -1388,6 +1388,7 @@ static int xe_pt_pre_commit(struct xe_mi
 				     pt_update_ops, rftree);
 }
 
+#if IS_ENABLED(CONFIG_DRM_GPUSVM)
 /*
  * Acquire/release the svm notifier_lock around xe_pt_svm_userptr_pre_commit()
  * and the matching late release in xe_pt_update_ops_run(). Read mode by
@@ -1414,6 +1415,10 @@ static void xe_pt_svm_userptr_notifier_u
 	xe_svm_notifier_unlock(vm);
 #endif
 }
+#else
+static inline void xe_pt_svm_userptr_notifier_lock(struct xe_vm *vm) { }
+static inline void xe_pt_svm_userptr_notifier_unlock(struct xe_vm *vm) { }
+#endif
 
 #if IS_ENABLED(CONFIG_DRM_GPUSVM)
 #ifdef CONFIG_DRM_XE_USERPTR_INVAL_INJECT
author maintainer
#1426 lore raw

PATCH 1420/1611 KVM: TDX: Account all non-transient page allocations for per-TD structures

In reply to Greg Kroah-Hartman

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

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

From: Sean Christopherson <[email protected]>

[ Upstream commit a8b2924676ec42d76597f849baba31acc3ba1bfc ]

Account all non-transient allocations associated with a single TD (or its vCPUs), as KVM's ABI is that allocations that are active for the lifetime of a VM are accounted. Leave temporary allocations, i.e. allocations that are freed within a single function/ioctl, unaccounted, to again align with KVM's existing behavior, e.g. see commit dd103407ca31 ("KVM: X86: Remove unnecessary GFP_KERNEL_ACCOUNT for temporary variables").

Fixes 8d032b683c29 ("KVM: TDX: create/destroy VM structure")
Fixes a50f673f25e0 ("KVM: TDX: Do TDX specific vcpu initialization")

[ Binbin: Resolved conflicts in __tdx_td_init() due to lacking commits
eac90a5ba0aa and 69050f8d6d07, retain nr_configured_hkid increment and kcalloc() call. ]

---
 arch/x86/kvm/vmx/tdx.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)
arch/x86/kvm/vmx/tdx.c +6 -6
index d9ffc16dfefe61..7534397defa295 100644
@@ -2474,7 +2474,7 @@ static int __tdx_td_init(struct kvm *kvm, struct td_params *td_params,
 
 	atomic_inc(&nr_configured_hkid);
 
-	tdr_page = alloc_page(GFP_KERNEL);
+	tdr_page = alloc_page(GFP_KERNEL_ACCOUNT);
 	if (!tdr_page)
 		goto free_hkid;
 
@@ -2482,12 +2482,12 @@ static int __tdx_td_init(struct kvm *kvm, struct td_params *td_params,
 	/* TDVPS = TDVPR(4K page) + TDCX(multiple 4K pages), -1 for TDVPR. */
 	kvm_tdx->td.tdcx_nr_pages = tdx_sysinfo->td_ctrl.tdvps_base_size / PAGE_SIZE - 1;
 	tdcs_pages = kcalloc(kvm_tdx->td.tdcs_nr_pages, sizeof(*kvm_tdx->td.tdcs_pages),
-			     GFP_KERNEL);
+			     GFP_KERNEL_ACCOUNT);
 	if (!tdcs_pages)
 		goto free_tdr;
 
 	for (i = 0; i < kvm_tdx->td.tdcs_nr_pages; i++) {
-		tdcs_pages[i] = alloc_page(GFP_KERNEL);
+		tdcs_pages[i] = alloc_page(GFP_KERNEL_ACCOUNT);
 		if (!tdcs_pages[i])
 			goto free_tdcs;
 	}
@@ -2937,7 +2937,7 @@ static int tdx_td_vcpu_init(struct kvm_vcpu *vcpu, u64 vcpu_rcx)
 	int ret, i;
 	u64 err;
 
-	page = alloc_page(GFP_KERNEL);
+	page = alloc_page(GFP_KERNEL_ACCOUNT);
 	if (!page)
 		return -ENOMEM;
 	tdx->vp.tdvpr_page = page;
@@ -2950,14 +2950,14 @@ static int tdx_td_vcpu_init(struct kvm_vcpu *vcpu, u64 vcpu_rcx)
 	tdx->vp.tdvpr_pa = page_to_phys(tdx->vp.tdvpr_page);
 
 	tdx->vp.tdcx_pages = kcalloc(kvm_tdx->td.tdcx_nr_pages, sizeof(*tdx->vp.tdcx_pages),
-			       	     GFP_KERNEL);
+				     GFP_KERNEL_ACCOUNT);
 	if (!tdx->vp.tdcx_pages) {
 		ret = -ENOMEM;
 		goto free_tdvpr;
 	}
 
 	for (i = 0; i < kvm_tdx->td.tdcx_nr_pages; i++) {
-		page = alloc_page(GFP_KERNEL);
+		page = alloc_page(GFP_KERNEL_ACCOUNT);
 		if (!page) {
 			ret = -ENOMEM;
 			goto free_tdcx;
Signature · 1 line
2.53.0
author maintainer
#1427 lore raw

PATCH 1421/1611 selftests/bpf: Add simple strscpy() implementation

In reply to Greg Kroah-Hartman

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

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

From: Ihor Solodrai <[email protected]>

commit 63c49efc987afefc6b9bb7de083eb8748e0b1789 upstream.

Replace bpf_strlcpy() in bpf_util.h with a sized_strscpy(), which is a simplified sized_strscpy() from the kernel (lib/string.c [1]). It:
* takes a count (destination size) parameter
* guarantees NULL-termination
* returns the number of characters copied or -E2BIG

Re-define strscpy macro similar to in-kernel implementation [2]: allow the count parameter to be optional.

Add #ifdef-s to tools/include/linux/args.h, as they may be defined in other system headers (for example, __CONCAT in sys/cdefs.h).

Fixup the single existing bpf_strlcpy() call in cgroup_helpers.c

[1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/lib/string.c?h=v6.19#n113
[2] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/include/linux/string.h?h=v6.19#n91

---
 tools/include/linux/args.h                   |    4 ++
 tools/testing/selftests/bpf/bpf_util.h       |   45 +++++++++++++++++++--------
 tools/testing/selftests/bpf/cgroup_helpers.c |    2 -
 3 files changed, 37 insertions(+), 14 deletions(-)
tools/include/linux/args.h +4 -0
@@ -22,7 +22,11 @@
 #define COUNT_ARGS(X...) __COUNT_ARGS(, ##X, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0)
 
 /* Concatenate two parameters, but allow them to be expanded beforehand. */
+#ifndef __CONCAT
 #define __CONCAT(a, b) a ## b
+#endif
+#ifndef CONCATENATE
 #define CONCATENATE(a, b) __CONCAT(a, b)
+#endif
 
 #endif	/* _LINUX_ARGS_H */
tools/testing/selftests/bpf/bpf_util.h +32 -13

63 lines not on this page · raw message

tools/testing/selftests/bpf/cgroup_helpers.c +1 -1
@@ -86,7 +86,7 @@ static int __enable_controllers(const ch
 		enable[len] = 0;
 		close(fd);
 	} else {
-		bpf_strlcpy(enable, controllers, sizeof(enable));
+		strscpy(enable, controllers);
 	}
 
 	snprintf(path, sizeof(path), "%s/cgroup.subtree_control", cgroup_path);
author maintainer
#1428 lore raw

PATCH 1422/1611 pwm: rzg2l-gpt: Fix period_ticks type from u32 to u64

In reply to Greg Kroah-Hartman

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

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

From: Biju Das <[email protected]>

commit 2b40d72de9354a76f5e3bb71230a4210eaa92849 upstream.

period_ticks is used to store PWM period values that can exceed the 32-bit range, so change its type from u32 to u64 to prevent overflow.

---
 drivers/pwm/pwm-rzg2l-gpt.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/pwm/pwm-rzg2l-gpt.c +1 -1
index 4856af080e8e..c9dfa59bc1ea 100644
@@ -81,7 +81,7 @@ struct rzg2l_gpt_chip {
 	void __iomem *mmio;
 	struct mutex lock; /* lock to protect shared channel resources */
 	unsigned long rate_khz;
-	u32 period_ticks[RZG2L_MAX_HW_CHANNELS];
+	u64 period_ticks[RZG2L_MAX_HW_CHANNELS];
 	u32 channel_request_count[RZG2L_MAX_HW_CHANNELS];
 	u32 channel_enable_count[RZG2L_MAX_HW_CHANNELS];
 };
Signature · 1 line
2.55.0
author maintainer
#1429 lore raw

PATCH 1423/1611 LoongArch: Fix nr passing in set_direct_map_valid_noflush()

In reply to Greg Kroah-Hartman

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

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

From: Xuewen Wang <[email protected]>

commit 70378a710598432f13509bdc16a1c0f06b3ecb53 upstream.

set_direct_map_valid_noflush() incorrectly passes 1 to __set_memory() instead of nr. This causes only the first page's attr to be updated when nr > 1.

Other architectures all pass nr correctly.

Fixes 0c6378a71574 ("arch: introduce set_direct_map_valid_noflush()")
---
 arch/loongarch/mm/pageattr.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/loongarch/mm/pageattr.c +1 -1
@@ -234,5 +234,5 @@ int set_direct_map_valid_noflush(struct
 		clear = __pgprot(_PAGE_PRESENT | _PAGE_VALID);
 	}
 
-	return __set_memory(addr, 1, set, clear);
+	return __set_memory(addr, nr, set, clear);
 }
author maintainer
#1430 lore raw

PATCH 1424/1611 LoongArch: Fix missing dirty page tracking in {pte,pmd}_wrprotect()

In reply to Greg Kroah-Hartman

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

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

From: Hongchen Zhang <[email protected]>

commit 018e9828eb523c638fa3d9bdf0fd4956b74555b2 upstream.

When hardware page table walker (PTW) is enabled on LoongArch, the CPU may set _PAGE_DIRTY directly in the page table entry during a write TLB miss, without going through the software TLB store handler. The software TLB store handler (tlbex.S:254) sets both _PAGE_DIRTY and_PAGE_MODIFIED together:

ori t0, t0, (_PAGE_VALID | _PAGE_DIRTY | _PAGE_MODIFIED)

Since hardware PTW only sets _PAGE_DIRTY, the software-only bit, i.e. _PAGE_MODIFIED is left unchanged. This creates a window where a PTE has _PAGE_DIRTY set (hardware knows the page is dirty) but _PAGE_MODIFIED clear (software is unaware).

When fork()/clone() triggers copy-on-write, __copy_present_ptes() calls pte_wrprotect(), which unconditionally clears both the _PAGE_WRITE and _PAGE_DIRTY bits:

    pte_val(pte) &= ~(_PAGE_WRITE | _PAGE_DIRTY);

Since _PAGE_MODIFIED was never set, the dirtiness information is lost completely. Subsequently, when memory pressure triggers page reclaim, page_mkclean() / try_to_unmap() sees the page as clean (i.e. pte_dirty() returns false) and the page may be freed without writeback, causing data corruption.

Fix this by propagating the _PAGE_DIRTY bit to the _PAGE_MODIFIED bit in both pte_wrprotect() and pmd_wrprotect() before clearing writeable bits:

    if (pte_val(pte) & _PAGE_DIRTY)
        pte_val(pte) |= _PAGE_MODIFIED;

The pmd_wrprotect() fix handles the CONFIG_TRANSPARENT_HUGEPAGE case, where pmd entries need the same treatment.

This ensures the software dirty tracking bit (checked by pte_dirty() and pmd_dirty(), which read both the _PAGE_DIRTY and _PAGE_MODIFIED bits) is preserved across fork COW write-protection.

The issue was found by the LTP madvise09 test case, which exercises page reclaim after "madvise(MADV_FREE), write and fork" operation sequence on private anonymous mappings.

Fixes 09cfefb7fa70 ("LoongArch: Add memory management")
---
 arch/loongarch/include/asm/pgtable.h |    4 ++++
 1 file changed, 4 insertions(+)
arch/loongarch/include/asm/pgtable.h +4 -0
@@ -392,6 +392,8 @@ static inline pte_t pte_mkwrite_novma(pt
 
 static inline pte_t pte_wrprotect(pte_t pte)
 {
+	if (pte_val(pte) & _PAGE_DIRTY)
+		pte_val(pte) |= _PAGE_MODIFIED;
 	pte_val(pte) &= ~(_PAGE_WRITE | _PAGE_DIRTY);
 	return pte;
 }
@@ -498,6 +500,8 @@ static inline pmd_t pmd_mkwrite_novma(pm
 
 static inline pmd_t pmd_wrprotect(pmd_t pmd)
 {
+	if (pmd_val(pmd) & _PAGE_DIRTY)
+		pmd_val(pmd) |= _PAGE_MODIFIED;
 	pmd_val(pmd) &= ~(_PAGE_WRITE | _PAGE_DIRTY);
 	return pmd;
 }
author maintainer
#1431 lore raw

PATCH 1425/1611 ipmi: Fix user refcount underflow in event delivery

In reply to Greg Kroah-Hartman

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

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

From: Matt Fleming <[email protected]>

commit 6aa9e61c46465d231e9beddf56af7effd71be682 upstream.

ipmi_alloc_recv_msg(user) takes the temporary user reference owned by the receive message, and ipmi_free_recv_msg() drops it again. If event delivery fails after allocating receive messages for earlier users, handle_read_event_rsp() rolls those messages back with ipmi_free_recv_msg().

That rollback path still drops user->refcount explicitly after freeing each message. The extra put can free a user that remains linked on intf->users, so later event delivery may dereference a freed user or trip refcount_t's addition-on-zero warning when ipmi_alloc_recv_msg() tries to acquire another reference.

Remove the stale explicit put and the now-dead user assignment. Keep the list_del() and ipmi_free_recv_msg() calls; they are the required rollback operations.

Fixes b52da4054ee0 ("ipmi: Rework user message limit handling")
Cc [email protected] # v6.18+

Message-ID: <[email protected]>

---
 drivers/char/ipmi/ipmi_msghandler.c |    2 --
 1 file changed, 2 deletions(-)
drivers/char/ipmi/ipmi_msghandler.c +0 -2
@@ -4475,10 +4475,8 @@ static int handle_read_event_rsp(struct
 			mutex_unlock(&intf->users_mutex);
 			list_for_each_entry_safe(recv_msg, recv_msg2, &msgs,
 						 link) {
-				user = recv_msg->user;
 				list_del(&recv_msg->link);
 				ipmi_free_recv_msg(recv_msg);
-				kref_put(&user->refcount, free_ipmi_user);
 			}
 			/*
 			 * We couldn't allocate memory for the
author maintainer
#1432 lore raw

PATCH 1426/1611 espintcp: use sk_msg_free_partial to fix partial send

In reply to Greg Kroah-Hartman

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

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

From: Sabrina Dubroca <[email protected]>

commit 007800408002d871f5699bdb944f985896730b8f upstream.

sk_msg_free_partial() ensures consistency of the skmsg at every iteration, without having to manually handle uncharges and offsets. This simplifies the code, and fixes some bugs in skmsg accounting when we don't send the full contents.

---
 net/xfrm/espintcp.c |   34 +++++++---------------------------
 1 file changed, 7 insertions(+), 27 deletions(-)
net/xfrm/espintcp.c +7 -27
@@ -215,43 +215,23 @@ static int espintcp_sendskmsg_locked(str
 	struct sk_msg *skmsg = &emsg->skmsg;
 	bool more = flags & MSG_MORE;
 	struct scatterlist *sg;
-	int done = 0;
 	int ret;
 
-	sg = &skmsg->sg.data[skmsg->sg.start];
 	do {
 		struct bio_vec bvec;
-		size_t size = sg->length - emsg->offset;
-		int offset = sg->offset + emsg->offset;
-		struct page *p;
-
-		emsg->offset = 0;
 
+		sg = &skmsg->sg.data[skmsg->sg.start];
 		if (sg_is_last(sg) && !more)
 			msghdr.msg_flags &= ~MSG_MORE;
 
-		p = sg_page(sg);
-retry:
-		bvec_set_page(&bvec, p, size, offset);
-		iov_iter_bvec(&msghdr.msg_iter, ITER_SOURCE, &bvec, 1, size);
-		ret = tcp_sendmsg_locked(sk, &msghdr, size);
-		if (ret < 0) {
-			emsg->offset = offset - sg->offset;
-			skmsg->sg.start += done;
+		bvec_set_page(&bvec, sg_page(sg), sg->length, sg->offset);
+		iov_iter_bvec(&msghdr.msg_iter, ITER_SOURCE, &bvec, 1, sg->length);
+		ret = tcp_sendmsg_locked(sk, &msghdr, sg->length);
+		if (ret < 0)
 			return ret;
-		}
 
-		if (ret != size) {
-			offset += ret;
-			size -= ret;
-			goto retry;
-		}
-
-		done++;
-		put_page(p);
-		sk_mem_uncharge(sk, sg->length);
-		sg = sg_next(sg);
-	} while (sg);
+		sk_msg_free_partial(sk, skmsg, ret);
+	} while (skmsg->sg.size);
 
 	memset(emsg, 0, sizeof(*emsg));
 
author maintainer
#1433 lore raw

PATCH 1427/1611 ipmi: fix refcount leak in i_ipmi_request()

In reply to Greg Kroah-Hartman

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

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

From: Wentao Liang <[email protected]>

commit a3f3859cecacb64f18fd446271ece9a3b3f2d4de upstream.

When a caller provides a supplied_recv message to i_ipmi_request(), the function increments the user's nr_msgs reference count. If an error occurs later, the out_err cleanup path only frees the recv_msg if the function allocated it itself (i.e., !supplied_recv). In the supplied_recv case the cleanup is skipped, leaving the reference count elevated. The caller ipmi_request_supply_msgs() does not release the supplied_recv on error, so the reference is permanently leaked.

Fix this by explicitly reverting the reference count operations when a supplied recv_msg with a valid user pointer is present in the error
path: decrement nr_msgs and drop the user's kref.

Fixes b52da4054ee0 ("ipmi: Rework user message limit handling")

Message-ID: <[email protected]>

---
 drivers/char/ipmi/ipmi_msghandler.c |    8 ++++++++
 1 file changed, 8 insertions(+)
drivers/char/ipmi/ipmi_msghandler.c +8 -0
@@ -2349,6 +2349,10 @@ static int i_ipmi_request(struct ipmi_us
 		if (smi_msg == NULL) {
 			if (!supplied_recv)
 				ipmi_free_recv_msg(recv_msg);
+			else if (recv_msg->user) {
+				atomic_dec(&recv_msg->user->nr_msgs);
+				kref_put(&recv_msg->user->refcount, free_ipmi_user);
+			}
 			return -ENOMEM;
 		}
 	}
@@ -2422,6 +2426,10 @@ out_err:
 			ipmi_free_smi_msg(smi_msg);
 		if (!supplied_recv)
 			ipmi_free_recv_msg(recv_msg);
+		else if (recv_msg->user) {
+			atomic_dec(&recv_msg->user->nr_msgs);
+			kref_put(&recv_msg->user->refcount, free_ipmi_user);
+		}
 	}
 	return rv;
 }
author maintainer
#1434 lore raw

PATCH 1428/1611 bnx2x: fix potential memory leak in bnx2x_alloc_mem_bp()

In reply to Greg Kroah-Hartman

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

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

From: Abdun Nihaal <[email protected]>

commit a986fde914d88af47eb78fd29c5d1af7952c3500 upstream.

If the allocation of fp[i].tpa_info fails, the error path will not free the struct bnx2x_fastpath allocated earlier, as it is not linked to the bp structure yet. Fix that by linking it immediately after allocation.

---
 drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c |    3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c +1 -2
@@ -4750,6 +4750,7 @@ int bnx2x_alloc_mem_bp(struct bnx2x *bp)
 	fp = kcalloc(bp->fp_array_size, sizeof(*fp), GFP_KERNEL);
 	if (!fp)
 		goto alloc_err;
+	bp->fp = fp;
 	for (i = 0; i < bp->fp_array_size; i++) {
 		fp[i].tpa_info =
 			kcalloc(ETH_MAX_AGGREGATION_QUEUES_E1H_E2,
@@ -4758,8 +4759,6 @@ int bnx2x_alloc_mem_bp(struct bnx2x *bp)
 			goto alloc_err;
 	}
 
-	bp->fp = fp;
-
 	/* allocate sp objs */
 	bp->sp_objs = kcalloc(bp->fp_array_size, sizeof(struct bnx2x_sp_objs),
 			      GFP_KERNEL);
author maintainer
#1435 lore raw

PATCH 1429/1611 rtc: renesas-rtca3: Fix PIE clear polling condition in alarm setup error path

In reply to Greg Kroah-Hartman

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

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

From: Lad Prabhakar <[email protected]>

commit 7e342d87aa8e6b831cf6d21ca41b1f7e032d0fcf upstream.

In rtca3_set_alarm(), the setup_failed path attempts to disable the Periodic Interrupt Enable (PIE) bit and wait until it is cleared. However, the polling condition passed to readb_poll_timeout_atomic() uses an incorrect expression:

    !(tmp & ~RTCA3_RCR1_PIE)

As ~RTCA3_RCR1_PIE evaluates to a mask of all bits except PIE, the condition effectively waits for all non-PIE bits to become zero, which is unrelated to the intended operation and is unlikely to ever be true. This causes the poll to time out unnecessarily.

Fix the condition to check for the PIE bit itself being cleared:

    !(tmp & RTCA3_RCR1_PIE)

This correctly waits until PIE is deasserted after being cleared.

---
 drivers/rtc/rtc-renesas-rtca3.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/rtc/rtc-renesas-rtca3.c +1 -1
@@ -455,7 +455,7 @@ setup_failed:
 		 * specified timeout for setup.
 		 */
 		writeb(rcr1 & ~RTCA3_RCR1_PIE, priv->base + RTCA3_RCR1);
-		readb_poll_timeout_atomic(priv->base + RTCA3_RCR1, tmp, !(tmp & ~RTCA3_RCR1_PIE),
+		readb_poll_timeout_atomic(priv->base + RTCA3_RCR1, tmp, !(tmp & RTCA3_RCR1_PIE),
 					  10, RTCA3_DEFAULT_TIMEOUT_US);
 		atomic_set(&priv->alrm_sstep, RTCA3_ALRM_SSTEP_DONE);
 	}
author maintainer
#1436 lore raw

PATCH 1430/1611 rtc: mpfs: fix counter upload completion condition

In reply to Greg Kroah-Hartman

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

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

From: Conor Dooley <[email protected]>

commit 9792ff8afa9017fe14f436f3ef3cd75f41f9f145 upstream.

The condition that needs to be checked for upload completion is the UPLOAD bit in the completion register going low. The original iterations of this driver used a do-while and this was converted to a read_poll_timeout() during upstreaming without the condition being inverted as it should have been.

I suspect that this went unnoticed until now because a) the first read was done when the bit was still set, immediately completing the read_poll_timeout() and b) because the RTC doesn't hold time when power is removed from the SoC reducing its utility (I for one keep it disabled). If my first suspicion was true when the driver was upstreamed, it's not true any longer though, hence the detection of the problem.

---
 drivers/rtc/rtc-mpfs.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/rtc/rtc-mpfs.c +1 -1
@@ -112,7 +112,7 @@ static int mpfs_rtc_settime(struct devic
 	ctrl |= CONTROL_UPLOAD_BIT;
 	writel(ctrl, rtcdev->base + CONTROL_REG);
 
-	ret = read_poll_timeout(readl, prog, prog & CONTROL_UPLOAD_BIT, 0, UPLOAD_TIMEOUT_US,
+	ret = read_poll_timeout(readl, prog, !(prog & CONTROL_UPLOAD_BIT), 0, UPLOAD_TIMEOUT_US,
 				false, rtcdev->base + CONTROL_REG);
 	if (ret) {
 		dev_err(dev, "timed out uploading time to rtc");
author maintainer
#1437 lore raw

PATCH 1431/1611 hwmon: (w83627hf) remove VID sysfs files on error and remove

In reply to Greg Kroah-Hartman

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

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

From: Pengpeng Hou <[email protected]>

commit 5264b389c4e02dec214a46c400eb3ab867a7749a upstream.

w83627hf_probe() creates cpu0_vid and vrm with device_create_file() when VID information is available.

The error path and remove callback only remove the common and optional attribute groups. Those groups do not contain cpu0_vid or vrm, so the files can remain after a later probe failure or after device removal while their callbacks still expect live driver data.

Remove the standalone VID sysfs files from both the probe error path and the remove callback.

---
 drivers/hwmon/w83627hf.c |    4 ++++
 1 file changed, 4 insertions(+)
drivers/hwmon/w83627hf.c +4 -0
@@ -1823,6 +1823,8 @@ static int w83627hf_probe(struct platfor
 	return 0;
 
  error:
+	device_remove_file(dev, &dev_attr_vrm);
+	device_remove_file(dev, &dev_attr_cpu0_vid);
 	sysfs_remove_group(&dev->kobj, &w83627hf_group);
 	sysfs_remove_group(&dev->kobj, &w83627hf_group_opt);
 	return err;
@@ -1834,6 +1836,8 @@ static void w83627hf_remove(struct platf
 
 	hwmon_device_unregister(data->hwmon_dev);
 
+	device_remove_file(&pdev->dev, &dev_attr_vrm);
+	device_remove_file(&pdev->dev, &dev_attr_cpu0_vid);
 	sysfs_remove_group(&pdev->dev.kobj, &w83627hf_group);
 	sysfs_remove_group(&pdev->dev.kobj, &w83627hf_group_opt);
 }
author maintainer
#1438 lore raw

PATCH 1432/1611 hwmon: (w83793) remove vrm sysfs file on probe failure

In reply to Greg Kroah-Hartman

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

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

From: Pengpeng Hou <[email protected]>

commit 77b983757280c69b0290811669ff1d31022e5f1d upstream.

w83793_probe() creates the vrm sysfs file after creating the VID files when VID support is present.

The normal remove path deletes vrm, but the probe error path only removes the sensor, SDA, VID, fan, PWM and temperature files. A later probe failure can therefore leave vrm behind after the driver data has been freed.

Remove vrm in the probe error path next to the VID files, matching the normal remove path.

---
 drivers/hwmon/w83793.c |    1 +
 1 file changed, 1 insertion(+)
drivers/hwmon/w83793.c +1 -0
@@ -1917,6 +1917,7 @@ exit_remove:
 
 	for (i = 0; i < ARRAY_SIZE(w83793_vid); i++)
 		device_remove_file(dev, &w83793_vid[i].dev_attr);
+	device_remove_file(dev, &dev_attr_vrm);
 
 	for (i = 0; i < ARRAY_SIZE(w83793_left_fan); i++)
 		device_remove_file(dev, &w83793_left_fan[i].dev_attr);
author maintainer
#1439 lore raw

PATCH 1433/1611 net: liquidio: fix BAR resource leak on PF number failure

In reply to Greg Kroah-Hartman

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

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

From: Haoxiang Li <[email protected]>

commit c63ee62a3c4ac1a1542f4c1a4b87e2f41df5a496 upstream.

If cn23xx_get_pf_num() fails, the function returns without unmapping either BAR. Unmap both BARs before returning from the error path.

Found by manual code review.

Fixes 0c45d7fe12c7 ("liquidio: fix use of pf in pass-through mode in a virtual machine")
---
 drivers/net/ethernet/cavium/liquidio/cn23xx_pf_device.c |   18 ++++++++--------
 1 file changed, 10 insertions(+), 8 deletions(-)
drivers/net/ethernet/cavium/liquidio/cn23xx_pf_device.c +10 -8
@@ -1163,18 +1163,14 @@ int setup_cn23xx_octeon_pf_device(struct
 	if (octeon_map_pci_barx(oct, 1, MAX_BAR1_IOREMAP_SIZE)) {
 		dev_err(&oct->pci_dev->dev, "%s CN23XX BAR1 map failed\n",
 			__func__);
-		octeon_unmap_pci_barx(oct, 0);
-		return 1;
+		goto err_unmap_bar0;
 	}
 
 	if (cn23xx_get_pf_num(oct) != 0)
-		return 1;
+		goto err_unmap_bar1;
 
-	if (cn23xx_sriov_config(oct)) {
-		octeon_unmap_pci_barx(oct, 0);
-		octeon_unmap_pci_barx(oct, 1);
-		return 1;
-	}
+	if (cn23xx_sriov_config(oct))
+		goto err_unmap_bar1;
 
 	octeon_write_csr64(oct, CN23XX_SLI_MAC_CREDIT_CNT, 0x3F802080802080ULL);
 
@@ -1205,6 +1201,12 @@ int setup_cn23xx_octeon_pf_device(struct
 	oct->coproc_clock_rate = 1000000ULL * cn23xx_coprocessor_clock(oct);
 
 	return 0;
+
+err_unmap_bar1:
+	octeon_unmap_pci_barx(oct, 1);
+err_unmap_bar0:
+	octeon_unmap_pci_barx(oct, 0);
+	return 1;
 }
 EXPORT_SYMBOL_GPL(setup_cn23xx_octeon_pf_device);
 
author maintainer
#1440 lore raw

PATCH 1434/1611 hwmon: (occ) unregister sysfs devices outside occ lock

In reply to Greg Kroah-Hartman

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

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

From: Runyu Xiao <[email protected]>

commit e31408734332b8cc611342cdaaab6ba492180156 upstream.

occ_active(false) and occ_shutdown() unregister sysfs-backed devices while occ->lock is held. hwmon_device_unregister() and sysfs_remove_group() can wait for active sysfs callbacks to drain, and those callbacks can enter the OCC update path and try to take occ->lock again. That gives the unregister paths the lock ordering occ->lock -> sysfs callback drain, while a callback has the opposite edge sysfs callback -> occ->lock.

This issue was found by our static analysis tool and then manually reviewed against the current tree.

The grounded PoC kept the real unregister and callback carrier:

occ_shutdown()
hwmon_device_unregister()
occ_show_temp_1()
occ_update_response()

Lockdep reported the circular dependency with occ_shutdown() already holding the OCC mutex and hwmon_device_unregister() waiting on the sysfs side:

WARNING: possible circular locking dependency detected
... (sysfs_lock) ... at: hwmon_device_unregister+0x12/0x30 [vuln_msv] ... (&test_occ.lock) ... at: occ_shutdown.constprop.0+0xe/0x40 [vuln_msv]

  occ_update_response.isra.0+0xb/0x20 [vuln_msv]
  occ_show_temp_1.constprop.0.isra.0+0x23/0x40 [vuln_msv]
  *** DEADLOCK ***

Serialize hwmon registration and removal with a separate hwmon_lock. Under that lock, detach occ->hwmon and update occ->active while occ->lock is held so concurrent OCC state changes still see a stable state, then drop occ->lock before calling hwmon_device_unregister(). Remove the driver sysfs group before taking occ->lock in occ_shutdown(), so draining the driver attributes cannot wait while the OCC mutex is held. Also make OCC update callbacks return -ENODEV after deactivation, so callbacks that already passed sysfs active protection do not poll the hardware after teardown has detached the hwmon device.

Fixes 849b0156d996 ("hwmon: (occ) Delay hwmon registration until user request")
Fixes ac6888ac5a11 ("hwmon: (occ) Lock mutex in shutdown to prevent race with occ_active")
---
 drivers/hwmon/occ/common.c |   34 ++++++++++++++++++++++++++++------
 drivers/hwmon/occ/common.h |    1 +
 2 files changed, 29 insertions(+), 6 deletions(-)
drivers/hwmon/occ/common.c +28 -6
@@ -214,6 +214,11 @@ int occ_update_response(struct occ *occ)
 	if (rc)
 		return rc;
 
+	if (!occ->active) {
+		rc = -ENODEV;
+		goto unlock;
+	}
+
 	/* limit the maximum rate of polling the OCC */
 	if (time_after(jiffies, occ->next_update)) {
 		rc = occ_poll(occ);
@@ -222,6 +227,7 @@ int occ_update_response(struct occ *occ)
 		rc = occ->last_error;
 	}
 
+unlock:
 	mutex_unlock(&occ->lock);
 	return rc;
 }
@@ -1105,11 +1111,16 @@ static void occ_parse_poll_response(stru
 
 int occ_active(struct occ *occ, bool active)
 {
-	int rc = mutex_lock_interruptible(&occ->lock);
+	struct device *hwmon = NULL;
+	int rc = mutex_lock_interruptible(&occ->hwmon_lock);
 
 	if (rc)
 		return rc;
 
+	rc = mutex_lock_interruptible(&occ->lock);
+	if (rc)
+		goto unlock_hwmon;
+
 	if (active) {
 		if (occ->active) {
 			rc = -EALREADY;
@@ -1154,14 +1165,17 @@ int occ_active(struct occ *occ, bool act
 			goto unlock;
 		}
 
-		if (occ->hwmon)
-			hwmon_device_unregister(occ->hwmon);
+		hwmon = occ->hwmon;
 		occ->active = false;
 		occ->hwmon = NULL;
 	}
 
 unlock:
 	mutex_unlock(&occ->lock);
+	if (hwmon)
+		hwmon_device_unregister(hwmon);
+unlock_hwmon:
+	mutex_unlock(&occ->hwmon_lock);
 	return rc;
 }
 
@@ -1170,6 +1184,7 @@ int occ_setup(struct occ *occ)
 	int rc;
 
 	mutex_init(&occ->lock);
+	mutex_init(&occ->hwmon_lock);
 	occ->groups[0] = &occ->group;
 
 	rc = occ_setup_sysfs(occ);
@@ -1190,15 +1205,22 @@ EXPORT_SYMBOL_GPL(occ_setup);
 
 void occ_shutdown(struct occ *occ)
 {
-	mutex_lock(&occ->lock);
+	struct device *hwmon;
 
 	occ_shutdown_sysfs(occ);
 
-	if (occ->hwmon)
-		hwmon_device_unregister(occ->hwmon);
+	mutex_lock(&occ->hwmon_lock);
+	mutex_lock(&occ->lock);
+
+	hwmon = occ->hwmon;
+	occ->active = false;
 	occ->hwmon = NULL;
 
 	mutex_unlock(&occ->lock);
+
+	if (hwmon)
+		hwmon_device_unregister(hwmon);
+	mutex_unlock(&occ->hwmon_lock);
 }
 EXPORT_SYMBOL_GPL(occ_shutdown);
 
drivers/hwmon/occ/common.h +1 -0
@@ -101,6 +101,7 @@ struct occ {
 
 	unsigned long next_update;
 	struct mutex lock;		/* lock OCC access */
+	struct mutex hwmon_lock;		/* serialize hwmon registration/removal */
 
 	struct device *hwmon;
 	struct occ_attribute *attrs;
author maintainer
#1441 lore raw

PATCH 1435/1611 fsl/fman: Free init resources on KeyGen failure in fman_init()

In reply to Greg Kroah-Hartman

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

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

From: Haoxiang Li <[email protected]>

commit d288efa2b94abc2e45a061fceb156b4f4e5b37be upstream.

fman_muram_alloc() allocates initialization resources before initializing the KeyGen block. If keygen_init() fails, the function returns -EINVAL directly and leaves those resources allocated. Free the initialization resources before returning from the KeyGen failure path.

---
 drivers/net/ethernet/freescale/fman/fman.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/net/ethernet/freescale/fman/fman.c +3 -1
@@ -1995,8 +1995,10 @@ static int fman_init(struct fman *fman)
 
 	/* Init KeyGen */
 	fman->keygen = keygen_init(fman->kg_regs);
-	if (!fman->keygen)
+	if (!fman->keygen) {
+		free_init_resources(fman);
 		return -EINVAL;
+	}
 
 	err = enable(fman, cfg);
 	if (err != 0)
author maintainer
#1442 lore raw

PATCH 1436/1611 net: lan743x: Initialize eth_syslock spinlock before use

In reply to Greg Kroah-Hartman

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

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

From: Andrea Righi <[email protected]>

commit 39139b1c1c2b614096519b526112c726adb12ff0 upstream.

lan743x_hardware_init() calls pci11x1x_strap_get_status() during the PCI11x1x probe sequence. That helper acquires the Ethernet subsystem hardware lock via lan743x_hs_syslock_acquire(), which relies on adapter->eth_syslock_spinlock to serialize access.

The spinlock is currently initialized only after the strap status is read. With CONFIG_DEBUG_SPINLOCK enabled, taking the zeroed initialized spinlock can trip the spinlock debug check.

Fix by initializing adapter->eth_syslock_spinlock before reading the strap status so the probe path never attempts to lock an uninitialized spinlock.

---
 drivers/net/ethernet/microchip/lan743x_main.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/microchip/lan743x_main.c +1 -1
@@ -3536,8 +3536,8 @@ static int lan743x_hardware_init(struct
 		adapter->max_tx_channels = PCI11X1X_MAX_TX_CHANNELS;
 		adapter->used_tx_channels = PCI11X1X_USED_TX_CHANNELS;
 		adapter->max_vector_count = PCI11X1X_MAX_VECTOR_COUNT;
-		pci11x1x_strap_get_status(adapter);
 		spin_lock_init(&adapter->eth_syslock_spinlock);
+		pci11x1x_strap_get_status(adapter);
 		mutex_init(&adapter->sgmii_rw_lock);
 		pci11x1x_set_rfe_rd_fifo_threshold(adapter);
 		sgmii_ctl = lan743x_csr_read(adapter, SGMII_CTL);
author maintainer
#1443 lore raw

PATCH 1437/1611 net/sched: sch_multiq: Replace direct dequeue call with peek and qdisc_dequeue_peeked

In reply to Greg Kroah-Hartman

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

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

From: Bryam Vargas <[email protected]>

commit 54f6b0c843e228d499eb4b6bbb89df68cad9ad5d upstream.

multiq_dequeue() takes a packet from a band's child with a direct ->dequeue() call after multiq_peek() peeked it. When the child is non-work-conserving the peek stashes the skb in the child's gso_skb, so the direct dequeue returns a different skb and orphans the stash, desyncing the child's qlen/backlog. With a qfq child reached through a peeking parent (e.g. tbf) this re-enters the child on an emptied list and dereferences NULL, panicking the kernel from softirq on ordinary egress.

Take the packet through qdisc_dequeue_peeked(), as sch_prio already does and as sch_red and sch_sfb were just fixed to do. The helper is a no-op when the child has no stash, so a work-conserving child is unaffected.

---
 net/sched/sch_multiq.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/sched/sch_multiq.c +1 -1
@@ -103,7 +103,7 @@ static struct sk_buff *multiq_dequeue(st
 		if (!netif_xmit_stopped(
 		    netdev_get_tx_queue(qdisc_dev(sch), q->curband))) {
 			qdisc = q->queues[q->curband];
-			skb = qdisc->dequeue(qdisc);
+			skb = qdisc_dequeue_peeked(qdisc);
 			if (skb) {
 				qdisc_bstats_update(sch, skb);
 				sch->q.qlen--;
author maintainer
#1444 lore raw

PATCH 1438/1611 net/sched: sch_taprio: Replace direct dequeue call with peek and qdisc_dequeue_peeked

In reply to Greg Kroah-Hartman

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

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

From: Bryam Vargas <[email protected]>

commit e056e1dfcddca877dd46d704e8ec9860cfc9ec44 upstream.

When taprio's software path peeks a non-work-conserving child qdisc, the child stashes the peeked skb in its gso_skb; taprio_dequeue_from_txq() then takes the packet with a direct child ->dequeue() call, which ignores that stash, orphans the peeked skb and desyncs the child's qlen/backlog. With a qfq child this re-enters the child on an emptied list and dereferences NULL, panicking the kernel from softirq on ordinary egress.

Take the packet through qdisc_dequeue_peeked(), as sch_red and sch_sfb now do. The helper returns the child's stashed skb first and is a no-op when there is none, so a work-conserving child is unaffected and the gated path now consumes the skb whose length was charged to the budget.

---
 net/sched/sch_taprio.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/sched/sch_taprio.c +1 -1
@@ -748,7 +748,7 @@ static struct sk_buff *taprio_dequeue_fr
 		return NULL;
 
 skip_peek_checks:
-	skb = child->ops->dequeue(child);
+	skb = qdisc_dequeue_peeked(child);
 	if (unlikely(!skb))
 		return NULL;
 
author maintainer
#1445 lore raw

PATCH 1439/1611 fhandle: reject detached mounts in capable_wrt_mount()

In reply to Greg Kroah-Hartman

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

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

From: David Lee <[email protected]>

commit 6c732471740bc2ac9b0946134f9f551dc75f4369 upstream.

The recent fhandle RCU fix moved the mount namespace capability check into capable_wrt_mount(), so a non-NULL mnt_namespace survives the ns_capable() dereference. The helper still assumes the later READ_ONCE(mount->mnt_ns) must be non-NULL because may_decode_fh() checked is_mounted() first.

That assumption is not stable. A detached mount from open_tree(..., OPEN_TREE_CLONE) can be dissolved on fput while open_by_handle_at() is between those checks, and umount_tree() can clear mount->mnt_ns. If the helper observes NULL, it dereferences mnt_ns->user_ns and panics.

Return false when the RCU read observes a detached mount. This keeps the relaxed permission path conservative: a mount no longer attached to a namespace cannot authorize open_by_handle_at() access.

Fixes 620c266f3949 ("fhandle: relax open_by_handle_at() permission checks")
Assisted-by LLM
---
 fs/fhandle.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
fs/fhandle.c +1 -1
@@ -297,7 +297,7 @@ static bool capable_wrt_mount(struct mou
 	 */
 	guard(rcu)();
 	mnt_ns = READ_ONCE(mount->mnt_ns);
-	return ns_capable(mnt_ns->user_ns, CAP_SYS_ADMIN);
+	return mnt_ns && ns_capable(mnt_ns->user_ns, CAP_SYS_ADMIN);
 }
 
 static inline int may_decode_fh(struct handle_to_path_ctx *ctx,
author maintainer
#1446 lore raw

PATCH 1440/1611 hwmon: (max1619) add missing select REGMAP to Kconfig

In reply to Greg Kroah-Hartman

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

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

From: Joshua Crofts <[email protected]>

commit 943a749bdffdd2132fab9240db890e07d93e1fcf upstream.

The Kconfig entry for the MAX1619 sensor doesn't contain a select REGMAP parameter, causing build failures if regmap isn't selected previously during the build process.

---
 drivers/hwmon/Kconfig |    1 +
 1 file changed, 1 insertion(+)
drivers/hwmon/Kconfig +1 -0
@@ -1213,6 +1213,7 @@ config SENSORS_MAX16065
 config SENSORS_MAX1619
 	tristate "Maxim MAX1619 sensor chip"
 	depends on I2C
+	select REGMAP
 	help
 	  If you say yes here you get support for MAX1619 sensor chip.
 
author maintainer
#1447 lore raw

PATCH 1441/1611 tracing/probes: Fix double addition of offset for @+FOFFSET

In reply to Greg Kroah-Hartman

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

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

From: Masami Hiramatsu (Google) <[email protected]>

commit 9a667b7750dda88cbf1cca96a53a2163b2ee71f7 upstream.

Since commit 533059281ee5 ("tracing: probeevent: Introduce new argument
fetching code") wrongly use @offset local variable during the parsing,
the offset value is added twice when dereferencing.
Reset the @offset after setting it in FETCH_OP_FOFFS.

Fixes 533059281ee5 ("tracing: probeevent: Introduce new argument fetching code")
---
 kernel/trace/trace_probe.c |    1 +
 1 file changed, 1 insertion(+)
kernel/trace/trace_probe.c +1 -0
@@ -1248,6 +1248,7 @@ parse_probe_arg(char *arg, const struct
 
 			code->op = FETCH_OP_FOFFS;
 			code->immediate = (unsigned long)offset;  // imm64?
+			offset = 0;
 		} else {
 			/* uprobes don't support symbols */
 			if (!(ctx->flags & TPARG_FL_KERNEL)) {
author maintainer
#1448 lore raw

PATCH 1442/1611 orangefs: keep the readdir entry size 64-bit in fill_from_part()

In reply to Greg Kroah-Hartman

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

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

From: Bryam Vargas <[email protected]>

commit 18227a6bc98bd0ba96ed3ce9d5b28776a5a28dfc upstream.

fill_from_part() computes the size of a directory entry in size_t but stores it in a __u32. An entry length near U32_MAX wraps it to a small value, bypasses the bounds check, and is then used to index the entry, reading far past the directory part -- an out-of-bounds read that oopses the kernel.

Compute the size as a u64 so it cannot truncate; the bounds check then rejects the entry. The trailer is supplied by the userspace client.

---
 fs/orangefs/dir.c |    7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)
fs/orangefs/dir.c +4 -3
@@ -190,7 +190,8 @@ static int fill_from_part(struct orangef
 {
 	const int offset = sizeof(struct orangefs_readdir_response_s);
 	struct orangefs_khandle *khandle;
-	__u32 *len, padlen;
+	__u32 *len;
+	u64 padlen;
 	loff_t i;
 	char *s;
 	i = ctx->pos & ~PART_MASK;
@@ -214,8 +215,8 @@ static int fill_from_part(struct orangef
 		 * len is the size of the string itself.  padlen is the
 		 * total size of the encoded string.
 		 */
-		padlen = (sizeof *len + *len + 1) +
-		    (8 - (sizeof *len + *len + 1)%8)%8;
+		padlen = (u64)sizeof *len + *len + 1;
+		padlen += (8 - padlen % 8) % 8;
 		if (part->len < i + padlen + sizeof *khandle)
 			goto next;
 		s = (void *)part + offset + i + sizeof *len;
author maintainer
#1449 lore raw

PATCH 1443/1611 net/mlx5: HWS, fix matcher leak on resize target setup failure

In reply to Greg Kroah-Hartman

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

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

From: Dawei Feng <[email protected]>

commit bb09d0e64ecaa0aa0f7d1133a1696ed74dead295 upstream.

hws_bwc_matcher_move() allocates a replacement matcher before setting it as the resize target. If mlx5hws_matcher_resize_set_target() fails, the replacement matcher is not attached anywhere and is leaked.

Fix the leak by destroying the replacement matcher before returning from the resize-target failure path.

The bug was first flagged by an experimental analysis tool we are developing for kernel memory-management bugs while analyzing v6.13-rc1. The tool is still under development and is not yet publicly available. Manual inspection confirms that the bug is still present in v7.1.1.

An x86_64 allyesconfig build showed no new warnings. As we do not have a mlx5 HWS-capable device to test with, no runtime testing was able to be performed.

---
 drivers/net/ethernet/mellanox/mlx5/core/steering/hws/bwc.c |    1 +
 1 file changed, 1 insertion(+)
drivers/net/ethernet/mellanox/mlx5/core/steering/hws/bwc.c +1 -0
@@ -205,6 +205,7 @@ static int hws_bwc_matcher_move(struct m
 	ret = mlx5hws_matcher_resize_set_target(old_matcher, new_matcher);
 	if (ret) {
 		mlx5hws_err(ctx, "Rehash error: failed setting resize target\n");
+		mlx5hws_matcher_destroy(new_matcher);
 		return ret;
 	}
 
author maintainer
#1450 lore raw

PATCH 1444/1611 ata: pata_pxa: Fix DMA channel leak on probe error

In reply to Greg Kroah-Hartman

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

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

From: Wentao Liang <[email protected]>

commit fcaf242e7fc406e78f444a35441e3b58f5e28781 upstream.

When dmaengine_slave_config() fails, the DMA channel acquired by dma_request_chan() is not released before returning the error, leaking the channel reference.

Fix by adding dma_release_channel() in the error path.

The ata_host_activate() error path already correctly releases the DMA channel.

Fixes 88622d80af82 ("ata: pata_pxa: dmaengine conversion")
---
 drivers/ata/pata_pxa.c |    1 +
 1 file changed, 1 insertion(+)
drivers/ata/pata_pxa.c +1 -0
@@ -286,6 +286,7 @@ static int pxa_ata_probe(struct platform
 	ret = dmaengine_slave_config(data->dma_chan, &config);
 	if (ret < 0) {
 		dev_err(&pdev->dev, "dma configuration failed: %d\n", ret);
+		dma_release_channel(data->dma_chan);
 		return ret;
 	}
 
author maintainer
#1451 lore raw

PATCH 1445/1611 net: wwan: iosm: bound device offsets in the MUX downlink decoder

In reply to Greg Kroah-Hartman

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

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

From: Maoyi Xie <[email protected]>

commit 526b8ef54668780c8f69e0211c342763d5dcbad1 upstream.

mux_dl_adb_decode() walks a chain of aggregated datagram tables using offsets and lengths taken from the modem. first_table_index, next_table_index, table_length, datagram_index and datagram_length are all device supplied le values. Only first_table_index was checked, and only for being non zero. The decoder then formed adth = block + adth_index and read the table header and the datagram entries with no bound against the received skb. A modem that reports an index or a length past the downlink buffer makes the decoder read out of bounds.

The buffer is IPC_MEM_MAX_DL_MUX_LITE_BUF_SIZE and skb->len is at most that, so skb->len is the real limit, but none of these in band offsets were checked against it.

The table chain is also followed with no forward progress check. The loop takes the next table from adth->next_table_index and stops only when that reaches zero. A modem can stage two tables that point at each other, so the loop never ends. It runs in softirq and clones the skb on every pass.

Validate every device offset and length against skb->len before use. The block header must fit. Each table header, on entry and after every next_table_index, must lie inside the skb. The datagram table must fit. Each datagram index and length must stay inside the skb. The header padding must not exceed the datagram length so the receive length does not wrap. Require each next_table_index to move forward so the chain cannot cycle.

This was reproduced under KASAN as a slab out of bounds read on a normal downlink receive once the iosm net device is up.

---
 drivers/net/wwan/iosm/iosm_ipc_mux_codec.c |   40 +++++++++++++++++++++--------
 1 file changed, 30 insertions(+), 10 deletions(-)
drivers/net/wwan/iosm/iosm_ipc_mux_codec.c +30 -10
@@ -553,19 +553,21 @@ static int mux_dl_process_dg(struct iosm
 	u32 packet_offset, i, rc, dg_len;
 
 	for (i = 0; i < nr_of_dg; i++, dg++) {
-		if (le32_to_cpu(dg->datagram_index)
-				< sizeof(struct mux_adbh))
+		u32 dg_index = le32_to_cpu(dg->datagram_index);
+
+		dg_len = le16_to_cpu(dg->datagram_length);
+
+		if (dg_index < sizeof(struct mux_adbh))
 			goto dg_error;
 
-		/* Is the packet inside of the ADB */
-		if (le32_to_cpu(dg->datagram_index) >=
-					le32_to_cpu(adbh->block_length)) {
+		/* Is the packet inside of the ADB and the received skb ? */
+		if (dg_index >= le32_to_cpu(adbh->block_length) ||
+		    dg_index >= skb->len ||
+		    dg_len > skb->len - dg_index ||
+		    dl_head_pad_len >= dg_len) {
 			goto dg_error;
 		} else {
-			packet_offset =
-				le32_to_cpu(dg->datagram_index) +
-				dl_head_pad_len;
-			dg_len = le16_to_cpu(dg->datagram_length);
+			packet_offset = dg_index + dl_head_pad_len;
 			/* Pass the packet to the netif layer. */
 			rc = ipc_mux_net_receive(ipc_mux, if_id, ipc_mux->wwan,
 						 packet_offset,
@@ -589,12 +591,16 @@ static void mux_dl_adb_decode(struct ios
 	struct mux_adbh *adbh;
 	struct mux_adth *adth;
 	int nr_of_dg, if_id;
-	u32 adth_index;
+	u32 adth_index, prev_index = 0;
 	u8 *block;
 
 	block = skb->data;
 	adbh = (struct mux_adbh *)block;
 
+	/* The block header itself must fit in the received skb. */
+	if (skb->len < sizeof(struct mux_adbh))
+		goto adb_decode_err;
+
 	/* Process the aggregated datagram tables. */
 	adth_index = le32_to_cpu(adbh->first_table_index);
 
@@ -606,6 +612,16 @@ static void mux_dl_adb_decode(struct ios
 
 	/* Loop through mixed session tables. */
 	while (adth_index) {
+		/* The table header must lie within the received skb, and the
+		 * chain must move forward so a modem cannot make the loop
+		 * cycle between two tables.
+		 */
+		if (adth_index <= prev_index ||
+		    adth_index < sizeof(struct mux_adbh) ||
+		    adth_index > skb->len - sizeof(struct mux_adth))
+			goto adb_decode_err;
+		prev_index = adth_index;
+
 		/* Get the reference to the table header. */
 		adth = (struct mux_adth *)(block + adth_index);
 
@@ -629,6 +645,10 @@ static void mux_dl_adb_decode(struct ios
 		if (le16_to_cpu(adth->table_length) < sizeof(struct mux_adth))
 			goto adb_decode_err;
 
+		/* The whole datagram table must fit in the received skb. */
+		if (le16_to_cpu(adth->table_length) > skb->len - adth_index)
+			goto adb_decode_err;
+
 		/* Calculate the number of datagrams. */
 		nr_of_dg = (le16_to_cpu(adth->table_length) -
 					sizeof(struct mux_adth)) /
author maintainer
#1452 lore raw

PATCH 1446/1611 hwmon: (asus_atk0110) Check package count before accessing element

In reply to Greg Kroah-Hartman

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

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

From: HyeongJun An <[email protected]>

commit e2735b39f044bad7bf2017aef248935525bc0b97 upstream.

atk_ec_present() walks the management group package returned by the GGRP ACPI method and, for each sub-package, reads its first element:

	id = &obj->package.elements[0];
	if (id->type != ACPI_TYPE_INTEGER)

without checking that the sub-package is non-empty. ACPICA allocates the element array with exactly package.count entries, so for a sub-package with a zero count this reads past the allocation.

The sibling function atk_debugfs_ggrp_open() performs the same access but skips empty packages with a package.count check first. Add the same check to atk_ec_present() so a malformed firmware package cannot trigger an out-of-bounds read.

Fixes 9e6eba610c2e ("hwmon: (asus_atk0110) Enable the EC")
Assisted-by Claude:claude-opus-4-8
---
 drivers/hwmon/asus_atk0110.c |    3 +++
 1 file changed, 3 insertions(+)
drivers/hwmon/asus_atk0110.c +3 -0
@@ -1037,6 +1037,9 @@ static int atk_ec_present(struct atk_dat
 		if (obj->type != ACPI_TYPE_PACKAGE)
 			continue;
 
+		if (!obj->package.count)
+			continue;
+
 		id = &obj->package.elements[0];
 		if (id->type != ACPI_TYPE_INTEGER)
 			continue;
author maintainer
#1453 lore raw

PATCH 1447/1611 riscv: probes: save original sp in rethook trampoline

In reply to Greg Kroah-Hartman

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

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

From: Martin Kaiser <[email protected]>

commit bc7b086a45521a986a49045907f017e3e46c763e upstream.

Reading a word from the stack in a kretprobe crashes a risc-v kernel.

$ cd /sys/kernel/tracing/
$ echo 'r n_tty_write $stack0' > dynamic_events
$ echo 1 > events/kprobes/enable
Unable to handle kernel paging request at virtual address 0000000200000128
...
[<ffffffff80016d16>] regs_get_kernel_stack_nth+0x26/0x38
[<ffffffff80177196>] process_fetch_insn+0x3ee/0x760
[<ffffffff80177836>] kretprobe_trace_func+0x116/0x1f0
[<ffffffff8017795a>] kretprobe_dispatcher+0x4a/0x58
[<ffffffff8013572e>] kretprobe_rethook_handler+0x5e/0x90
[<ffffffff80180838>] rethook_trampoline_handler+0x70/0x108
[<ffffffff8001ba32>] arch_rethook_trampoline_callback+0x12/0x1c
[<ffffffff8001ba84>] arch_rethook_trampoline+0x48/0x94
[<ffffffff8067872a>] tty_write+0x1a/0x30

In regs_get_kernel_stack_nth, regs->sp contains an arbitrary value.

arch_rethook_trampoline saves the registers from the probed function in a struct pt_regs. sp is not saved. Instead, sp is decremented for arch_rethook_trampoline's local stack.

Fix this crash and save the original sp along with the other registers. Use a0 as a temporary register, it is overwritten anyway.

---
 arch/riscv/kernel/probes/rethook_trampoline.S |    3 +++
 1 file changed, 3 insertions(+)
arch/riscv/kernel/probes/rethook_trampoline.S +3 -0
@@ -41,6 +41,9 @@
 	REG_S x29, PT_T4(sp)
 	REG_S x30, PT_T5(sp)
 	REG_S x31, PT_T6(sp)
+	/* save original sp */
+	addi a0, sp, PT_SIZE_ON_STACK
+	REG_S a0, PT_SP(sp)
 	.endm
 
 	.macro restore_all_base_regs
author maintainer
#1454 lore raw

PATCH 1448/1611 mm/compaction: handle free_pages_prepare() properly in compaction_free()

In reply to Greg Kroah-Hartman

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

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

From: Zi Yan <[email protected]>

commit 7da7d599b8a83271c464adfd5ef160202b470570 upstream.

free_pages_prepare() can fail but compaction_free() does not handle the failure case. Failed pages should not be added back to cc->freepages for future use, since they can be either PageHWPoison or free_page_is_bad() and might cause data corruption.

---
 mm/compaction.c |    7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)
mm/compaction.c +3 -4
@@ -1858,15 +1858,14 @@ static void compaction_free(struct folio
 	int order = folio_order(dst);
 	struct page *page = &dst->page;
 
-	if (folio_put_testzero(dst)) {
-		free_pages_prepare(page, order);
+	if (folio_put_testzero(dst) && free_pages_prepare(page, order)) {
 		list_add(&dst->lru, &cc->freepages[order]);
 		cc->nr_freepages += 1 << order;
 	}
 	cc->nr_migratepages += 1 << order;
 	/*
-	 * someone else has referenced the page, we cannot take it back to our
-	 * free list.
+	 * someone else has referenced the page or free_pages_prepare() fails,
+	 * we cannot take it back to our free list.
 	 */
 }
 
author maintainer
#1455 lore raw

PATCH 1449/1611 irqchip/irq-riscv-imsic-early: Fix fwnode leak on state setup failure

In reply to Greg Kroah-Hartman

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

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

From: Haoxiang Li <[email protected]>

commit 1358126fbed104e5657955d3ba029b283687ba02 upstream.

imsic_early_acpi_init() allocates a firmware node before setting up the IMSIC state. If imsic_setup_state() fails, the function returns without freeing the allocated fwnode.

Free the fwnode and clear the global pointer on this error path, matching the cleanup already done when imsic_early_probe() fails.

[ tglx: Use a common cleanup path instead of copying code around ]

---
 drivers/irqchip/irq-riscv-imsic-early.c |   15 ++++++++-------
 1 file changed, 8 insertions(+), 7 deletions(-)
drivers/irqchip/irq-riscv-imsic-early.c +8 -7
@@ -277,16 +277,13 @@ static int __init imsic_early_acpi_init(
 	rc = imsic_setup_state(imsic_acpi_fwnode, imsic);
 	if (rc) {
 		pr_err("%pfwP: failed to setup state (error %d)\n", imsic_acpi_fwnode, rc);
-		return rc;
+		goto cleanup;
 	}
 
 	/* Do early setup of IMSIC state and IPIs */
 	rc = imsic_early_probe(imsic_acpi_fwnode);
-	if (rc) {
-		irq_domain_free_fwnode(imsic_acpi_fwnode);
-		imsic_acpi_fwnode = NULL;
-		return rc;
-	}
+	if (rc)
+		goto cleanup;
 
 	rc = imsic_platform_acpi_probe(imsic_acpi_fwnode);
 
@@ -305,8 +302,12 @@ static int __init imsic_early_acpi_init(
 	 * DT where IPI works but MSI probe fails for some reason.
 	 */
 	return 0;
-}
 
+cleanup:
+	irq_domain_free_fwnode(imsic_acpi_fwnode);
+	imsic_acpi_fwnode = NULL;
+	return rc;
+}
 IRQCHIP_ACPI_DECLARE(riscv_imsic, ACPI_MADT_TYPE_IMSIC, NULL,
 		     1, imsic_early_acpi_init);
 #endif
author maintainer
#1456 lore raw

PATCH 1450/1611 s390/monwriter: Reject buffer reuse with different data length

In reply to Greg Kroah-Hartman

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

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

From: Gerald Schaefer <[email protected]>

commit 2995ccec260caa9e85b3301a4aba1e66ed80ad74 upstream.

When data buffers are reused, e.g. for interval sample records, the first record determines the data length, and the size of the buffer for user copy. Current monwriter code does not check if the data length was changed for subsequent records, which also would never happen for valid user programs.

However, a malicious user could change the data length, resulting in out of bounds user copy to the kernel buffer, and memory corruption. By default, the monwriter misc device is created with root-only permissions, so practical impact is typically low.

Fix this by checking for changed data length and rejecting such records.

---
 drivers/s390/char/monwriter.c |    3 +++
 1 file changed, 3 insertions(+)
drivers/s390/char/monwriter.c +3 -0
@@ -123,6 +123,9 @@ static int monwrite_new_hdr(struct mon_p
 			kfree(monbuf->data);
 			kfree(monbuf);
 			monbuf = NULL;
+		} else if (monbuf->hdr.datalen != monhdr->datalen) {
+			/* Data with buffer reuse must not change its length */
+			return -EINVAL;
 		}
 	} else if (monhdr->mon_function != MONWRITE_STOP_INTERVAL) {
 		if (mon_buf_count >= mon_max_bufs)
author maintainer
#1457 lore raw

PATCH 1451/1611 mac802154: remove interfaces with RCU list deletion

In reply to Greg Kroah-Hartman

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

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

From: Yousef Alhouseen <[email protected]>

commit 539dfcf69105d8d3d4d677b71de6e5ede2e6dfa0 upstream.

Queue wake, stop, and disable paths walk local->interfaces under RCU. The bulk hardware teardown path removes entries with list_del(), so an asynchronous transmit completion can follow a poisoned list node in ieee802154_wake_queue().

Use list_del_rcu() as in the single-interface removal path. The following unregister_netdevice() waits for in-flight RCU readers before freeing the netdevice, so no separate grace-period wait is needed.

---
 net/mac802154/iface.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/mac802154/iface.c +1 -1
@@ -703,7 +703,7 @@ void ieee802154_remove_interfaces(struct
 
 	mutex_lock(&local->iflist_mtx);
 	list_for_each_entry_safe(sdata, tmp, &local->interfaces, list) {
-		list_del(&sdata->list);
+		list_del_rcu(&sdata->list);
 
 		unregister_netdevice(sdata->dev);
 	}
author maintainer
#1458 lore raw

PATCH 1452/1611 octeontx2-pf: fix SQB pointer leak on init failure

In reply to Greg Kroah-Hartman

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

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

From: Dawei Feng <[email protected]>

commit 62e7df6d042aeebd5efb581074e28865c04477be upstream.

otx2_init_hw_resources() initializes SQ aura and pool resources before several later setup steps. On failure, err_free_sq_ptrs only frees SQB pages, leaving the per-SQ sqb_ptrs arrays behind.

Use otx2_free_sq_res() for the SQ unwind path and let it free sqb_ptrs even when sq->sqe has not been allocated yet.

The bug was first flagged by an experimental analysis tool we are developing for kernel memory-management bugs while analyzing v6.13-rc1. The tool is still under development and is not yet publicly available. Manual inspection confirms that the bug is still present in v7.1.1.

An x86_64 allyesconfig build showed no new warnings. As we do not have an OcteonTX2 PF device and the corresponding AF mailbox setup to test with, no runtime testing was able to be performed.

---
 drivers/net/ethernet/marvell/octeontx2/nic/otx2_pf.c |   21 +++++++++----------
 1 file changed, 10 insertions(+), 11 deletions(-)
drivers/net/ethernet/marvell/octeontx2/nic/otx2_pf.c +10 -11
@@ -1569,15 +1569,15 @@ static void otx2_free_sq_res(struct otx2
 	otx2_sq_free_sqbs(pf);
 	for (qidx = 0; qidx < otx2_get_total_tx_queues(pf); qidx++) {
 		sq = &qset->sq[qidx];
-		/* Skip freeing Qos queues if they are not initialized */
-		if (!sq->sqe)
-			continue;
-		qmem_free(pf->dev, sq->sqe);
-		qmem_free(pf->dev, sq->sqe_ring);
-		qmem_free(pf->dev, sq->cpt_resp);
-		qmem_free(pf->dev, sq->tso_hdrs);
-		qmem_free(pf->dev, sq->timestamps);
-		kfree(sq->sg);
+		/* sq->sqe is not initialized for unused QoS queues */
+		if (sq->sqe) {
+			qmem_free(pf->dev, sq->sqe);
+			qmem_free(pf->dev, sq->sqe_ring);
+			qmem_free(pf->dev, sq->cpt_resp);
+			qmem_free(pf->dev, sq->tso_hdrs);
+			qmem_free(pf->dev, sq->timestamps);
+			kfree(sq->sg);
+		}
 		kfree(sq->sqb_ptrs);
 	}
 }
@@ -1712,13 +1712,12 @@ int otx2_init_hw_resources(struct otx2_n
 	return err;
 
 err_free_nix_queues:
-	otx2_free_sq_res(pf);
 	otx2_free_cq_res(pf);
 	otx2_ctx_disable(mbox, NIX_AQ_CTYPE_RQ, false);
 err_free_txsch:
 	otx2_txschq_stop(pf);
 err_free_sq_ptrs:
-	otx2_sq_free_sqbs(pf);
+	otx2_free_sq_res(pf);
 err_free_rq_ptrs:
 	otx2_free_aura_ptr(pf, AURA_NIX_RQ);
 	otx2_ctx_disable(mbox, NPA_AQ_CTYPE_POOL, true);
author maintainer
#1459 lore raw

PATCH 1453/1611 selftests: net: make busywait timeout clock portable

In reply to Greg Kroah-Hartman

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

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

From: Nirmoy Das <[email protected]>

commit dd6a23bac306b7aa322e0aaccb60c6e32a198fb3 upstream.

loopy_wait() expects millisecond timestamps. However, Ubuntu Resolute can use uutils date, where date -u +%s%3N returns seconds plus full nanoseconds instead of a 3-digit millisecond field. This makes busywait expire too early and can make vlan_bridge_binding.sh read a stale operstate.

---
 tools/testing/selftests/net/lib.sh |   25 +++++++++++++++++++++++--
 1 file changed, 23 insertions(+), 2 deletions(-)
tools/testing/selftests/net/lib.sh +23 -2
@@ -70,12 +70,33 @@ ksft_exit_status_merge()
 		$ksft_xfail $ksft_pass $ksft_skip $ksft_fail
 }
 
+timestamp_ms()
+{
+	local now
+	local seconds
+	local nanoseconds
+
+	now=$(date -u +%s:%N) || return
+	seconds=${now%:*}
+	nanoseconds=${now#*:}
+
+	if [[ $nanoseconds =~ ^[0-9]+$ ]]; then
+		nanoseconds=${nanoseconds:0:9}
+	else
+		nanoseconds=0
+	fi
+
+	echo $((seconds * 1000 + 10#$nanoseconds / 1000000))
+}
+
 loopy_wait()
 {
 	local sleep_cmd=$1; shift
 	local timeout_ms=$1; shift
+	local start_time
+	local current_time
 
-	local start_time="$(date -u +%s%3N)"
+	start_time=$(timestamp_ms) || return
 	while true
 	do
 		local out
@@ -84,7 +105,7 @@ loopy_wait()
 			return 0
 		fi
 
-		local current_time="$(date -u +%s%3N)"
+		current_time=$(timestamp_ms) || return
 		if ((current_time - start_time > timeout_ms)); then
 			echo -n "$out"
 			return 1
author maintainer
#1460 lore raw

PATCH 1454/1611 llc: fix SAP refcount leak in llc_ui_autobind()

In reply to Greg Kroah-Hartman

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

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

From: Shuangpeng Bai <[email protected]>

commit 660667cd406648bbaffbd5c0d897c2263a852f11 upstream.

llc_ui_autobind() opens a SAP after choosing a dynamic LSAP. llc_sap_open() returns a reference owned by the caller, and llc_sap_add_socket() takes a second reference for the socket's membership in the SAP hash tables.

llc_ui_bind() drops the caller's reference after adding the socket, but llc_ui_autobind() keeps it. When the socket is closed, llc_sap_remove_socket() releases only the socket reference, leaving the SAP on llc_sap_list with sk_count == 0.

This is user-visible because repeated autobind and close cycles can consume all dynamic SAP values and make later autobinds fail with -EUSERS.

Drop the caller's reference after a successful autobind, matching llc_ui_bind()'s ownership model.

---
 net/llc/af_llc.c |    1 +
 1 file changed, 1 insertion(+)
net/llc/af_llc.c +1 -0
@@ -317,6 +317,7 @@ static int llc_ui_autobind(struct socket
 	/* assign new connection to its SAP */
 	llc_sap_add_socket(sap, sk);
 	sock_reset_flag(sk, SOCK_ZAPPED);
+	llc_sap_put(sap);
 	rc = 0;
 out:
 	dev_put(dev);
author maintainer
#1461 lore raw

PATCH 1455/1611 ipvs: use parsed transport offset in SCTP state lookup

In reply to Greg Kroah-Hartman

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

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

From: Yizhou Zhao <[email protected]>

commit 2f75c0faa3361b28e36cc0512b3299e163e25789 upstream.

set_sctp_state() reads the SCTP chunk header again in order to drive the IPVS SCTP state table. For IPv6 it computes the offset with sizeof(struct ipv6hdr), while the surrounding IPVS code uses iph.len from ip_vs_fill_iph_skb(), where ipv6_find_hdr() has already skipped extension headers and found the real transport header.

This makes the state machine read from the wrong offset for IPv6 SCTP packets that carry extension headers. For example, an INIT packet with an 8-byte destination options header can be scheduled correctly by sctp_conn_schedule(), but set_sctp_state() reads the first byte of the SCTP verification tag as a DATA chunk type. The connection then moves from NONE to ESTABLISHED instead of INIT1, gets the longer established timeout, and updates the active/inactive destination counters incorrectly. This happens even though the SCTP handshake has not completed.

Use the parsed transport offset passed down from ip_vs_set_state() for the SCTP chunk-header lookup. For IPv4 and IPv6 packets without extension headers this preserves the existing offset.

---
 net/netfilter/ipvs/ip_vs_proto_sctp.c |   15 +++++----------
 1 file changed, 5 insertions(+), 10 deletions(-)
net/netfilter/ipvs/ip_vs_proto_sctp.c +5 -10
@@ -372,20 +372,15 @@ static const char *sctp_state_name(int s
 
 static inline void
 set_sctp_state(struct ip_vs_proto_data *pd, struct ip_vs_conn *cp,
-		int direction, const struct sk_buff *skb)
+		int direction, const struct sk_buff *skb,
+		unsigned int iph_len)
 {
 	struct sctp_chunkhdr _sctpch, *sch;
 	unsigned char chunk_type;
 	int event, next_state;
-	int ihl, cofs;
-
-#ifdef CONFIG_IP_VS_IPV6
-	ihl = cp->af == AF_INET ? ip_hdrlen(skb) : sizeof(struct ipv6hdr);
-#else
-	ihl = ip_hdrlen(skb);
-#endif
+	int cofs;
 
-	cofs = ihl + sizeof(struct sctphdr);
+	cofs = iph_len + sizeof(struct sctphdr);
 	sch = skb_header_pointer(skb, cofs, sizeof(_sctpch), &_sctpch);
 	if (sch == NULL)
 		return;
@@ -472,7 +467,7 @@ sctp_state_transition(struct ip_vs_conn
 		unsigned int iph_len)
 {
 	spin_lock_bh(&cp->lock);
-	set_sctp_state(pd, cp, direction, skb);
+	set_sctp_state(pd, cp, direction, skb, iph_len);
 	spin_unlock_bh(&cp->lock);
 }
 
author maintainer
#1462 lore raw

PATCH 1456/1611 ipvs: reset full ip_vs_seq structs in ip_vs_conn_new

In reply to Greg Kroah-Hartman

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

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

From: Yizhou Zhao <[email protected]>

commit 2975324d164c552b028632f107b567302863b7f6 upstream.

Commit 9a05475cebdd ("ipvs: avoid kmem_cache_zalloc in ip_vs_conn_new") changed ip_vs_conn_new() to allocate an ip_vs_conn object with kmem_cache_alloc(). The function then initializes many fields explicitly, but only resets in_seq.delta and out_seq.delta in the two struct ip_vs_seq members.

That leaves init_seq and previous_delta uninitialized. This is normally harmless while the corresponding IP_VS_CONN_F_IN_SEQ or IP_VS_CONN_F_OUT_SEQ flag is clear. For connections learned from a sync message, however, ip_vs_proc_conn() preserves those flags from IP_VS_CONN_F_BACKUP_MASK and passes opt=NULL when the message omits IPVS_OPT_SEQ_DATA. In that case the new connection can be hashed with SEQ flags set but with the rest of in_seq/out_seq still containing stale slab data.

When a packet for such a connection is later handled by an IPVS application helper, vs_fix_seq() and vs_fix_ack_seq() use previous_delta and init_seq to rewrite TCP sequence numbers. A malformed sync message can therefore make forwarded packets carry stale slab bytes in their TCP seq/ack numbers, and can also corrupt the forwarded TCP flow.

Reset both struct ip_vs_seq members completely before publishing the connection. This matches the existing "reset struct ip_vs_seq" comment and keeps the sequence-adjustment gates inactive unless valid sequence data is installed later.

Fixes 9a05475cebdd ("ipvs: avoid kmem_cache_zalloc in ip_vs_conn_new")
Assisted-by Claude-Code:GLM-5.2
---
 net/netfilter/ipvs/ip_vs_conn.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
net/netfilter/ipvs/ip_vs_conn.c +2 -2
@@ -996,8 +996,8 @@ ip_vs_conn_new(const struct ip_vs_conn_p
 	cp->app = NULL;
 	cp->app_data = NULL;
 	/* reset struct ip_vs_seq */
-	cp->in_seq.delta = 0;
-	cp->out_seq.delta = 0;
+	memset(&cp->in_seq, 0, sizeof(cp->in_seq));
+	memset(&cp->out_seq, 0, sizeof(cp->out_seq));
 
 	atomic_inc(&ipvs->conn_count);
 	if (flags & IP_VS_CONN_F_NO_CPORT)
author maintainer
#1463 lore raw

PATCH 1457/1611 macsec: dont read an unset MAC header in macsec_encrypt()

In reply to Greg Kroah-Hartman

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

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

From: Daehyeon Ko <[email protected]>

commit f5089008f90c0a7c5520dff3934e0af00adf322d upstream.

macsec_encrypt() reads the Ethernet header via eth_hdr(skb) (skb->head + skb->mac_header) to memmove() the 12 source/destination MAC bytes forward and make room for the SecTAG.

On the AF_PACKET SOCK_RAW + PACKET_QDISC_BYPASS transmit path the skb reaches the macsec ndo_start_xmit() with the MAC header unset, so eth_hdr(skb) resolves to skb->head + (u16)~0 and the read is out of
bounds: a 12-byte heap over-read that is also emitted on the wire as the frame's outer source/destination MAC. KASAN reports a slab-out-of-bounds read in macsec_start_xmit() on 6.0; on current mainline a CONFIG_DEBUG_NET build flags it as an unset mac header in skb_mac_header().

On the TX path the L2 header is at skb->data, so use skb_eth_hdr(), added by commit 96cc4b69581d ("macvlan: do not assume mac_header is set in macvlan_broadcast()") for exactly this purpose.

---
 drivers/net/macsec.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/macsec.c +1 -1
@@ -642,7 +642,7 @@ static struct sk_buff *macsec_encrypt(st
 	}
 
 	unprotected_len = skb->len;
-	eth = eth_hdr(skb);
+	eth = skb_eth_hdr(skb);
 	sci_present = macsec_send_sci(secy);
 	hh = skb_push(skb, macsec_extra_len(sci_present));
 	memmove(hh, eth, 2 * ETH_ALEN);
author maintainer
#1464 lore raw

PATCH 1458/1611 dibs: loopback: validate offset and size in move_data()

In reply to Greg Kroah-Hartman

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

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

From: Dust Li <[email protected]>

commit 78237e3c0720fcc6eb9b87e90fd70f63eeca886f upstream.

The loopback move_data() performs a memcpy into the registered DMB without checking whether offset + size exceeds the DMB length. Unlike real ISM hardware, which enforces memory region bounds natively, the software loopback has no such protection.

A peer-supplied out-of-bounds offset or oversized write would result in an OOB write past the allocated kernel buffer. Add an explicit bounds check before the memcpy to reject such requests with -EINVAL.

---
 drivers/dibs/dibs_loopback.c |    5 +++++
 1 file changed, 5 insertions(+)
drivers/dibs/dibs_loopback.c +5 -0
@@ -254,6 +254,11 @@ static int dibs_lo_move_data(struct dibs
 		read_unlock_bh(&ldev->dmb_ht_lock);
 		return -EINVAL;
 	}
+	if ((u64)offset + size > rmb_node->len) {
+		read_unlock_bh(&ldev->dmb_ht_lock);
+		return -EINVAL;
+	}
+
 	memcpy((char *)rmb_node->cpu_addr + offset, data, size);
 	sba_idx = rmb_node->sba_idx;
 	read_unlock_bh(&ldev->dmb_ht_lock);
author maintainer
#1465 lore raw

PATCH 1459/1611 net: macb: drop in-flight Tx SKBs on close

In reply to Greg Kroah-Hartman

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

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

From: Théo Lebrun <[email protected]>

commit 27f575836cfebbf872dec020428742b10650a955 upstream.

The MACB driver has since forever leaked the outgoing SKBs that have not yet been marked as completed. They live in queue->tx_skb which gets freed without remorse nor checking.

macb_free_consistent() gets called in a few codepaths, but only close will trigger the added expressions. In macb_open() and macb_alloc_consistent() failure cases, queues' tx_skb just got allocated and are empty.

---
 drivers/net/ethernet/cadence/macb_main.c |   21 +++++++++++++++++++--
 1 file changed, 19 insertions(+), 2 deletions(-)
drivers/net/ethernet/cadence/macb_main.c +19 -2
@@ -2589,8 +2589,25 @@ static void macb_free_consistent(struct
 	dma_free_coherent(dev, size, bp->queues[0].rx_ring, bp->queues[0].rx_ring_dma);
 
 	for (q = 0, queue = bp->queues; q < bp->num_queues; ++q, ++queue) {
-		kfree(queue->tx_skb);
-		queue->tx_skb = NULL;
+		if (queue->tx_skb) {
+			unsigned int dropped = 0, tail;
+
+			for (tail = queue->tx_tail; tail != queue->tx_head;
+			     tail++) {
+				if (macb_tx_skb(queue, tail)->skb)
+					dropped++;
+				macb_tx_unmap(bp, macb_tx_skb(queue, tail), 0);
+			}
+
+			queue->stats.tx_dropped += dropped;
+			bp->dev->stats.tx_dropped += dropped;
+
+			kfree(queue->tx_skb);
+			queue->tx_skb = NULL;
+		}
+
+		queue->tx_head = 0;
+		queue->tx_tail = 0;
 		queue->tx_ring = NULL;
 		queue->rx_ring = NULL;
 	}
author maintainer
#1466 lore raw

PATCH 1460/1611 arm64: smp: Fix hot-unplug tearing by forcing unregistration

In reply to Greg Kroah-Hartman

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

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

From: Jinjie Ruan <[email protected]>

commit 18a4e5cf633fad5c40ac9d936c51bf38db68796d upstream.

Sashiko review pointed out the following issue[1].

Commit eba4675008a6 ("arm64: arch_register_cpu() variant to check if an ACPI handle is now available.") introduced architectural safety blocks inside arch_unregister_cpu(). If a hot-unplug operation is determined to be a physical hardware removal (where _STA evaluates to !ACPI_STA_DEVICE_PRESENT), or if firmware evaluation fails, it aborts the unregistration transaction early to protect unreadied arm64 infrastructure.

However, returning early from arch_unregister_cpu() causes a catastrophic state tearing because the generic ACPI layer (acpi_processor_post_eject()) unconditionally continues its cleanup flow. This leaves the stale sysfs device leaked in the memory, deadlocking any subsequent hot-add attempts on the same CPU.

Fix it by simplifying arch_unregister_cpu() to always proceed with the unregistration, as a pr_err_once() warning is sufficient to make it more visible for currently not supported physical CPU removal. Also remove the redundant NULL check on acpi_handle as it cannot be NULL when calling arch_unregister_cpu().

Cc Catalin Marinas <[email protected]>
Cc Jonathan Cameron <[email protected]>
Cc James Morse <[email protected]>
Fixes eba4675008a6 ("arm64: arch_register_cpu() variant to check if an ACPI handle is now available.")
---
 arch/arm64/kernel/smp.c |   16 +++-------------
 1 file changed, 3 insertions(+), 13 deletions(-)
arch/arm64/kernel/smp.c +3 -13
@@ -535,23 +535,13 @@ void arch_unregister_cpu(int cpu)
 {
 	acpi_handle acpi_handle = acpi_get_processor_handle(cpu);
 	struct cpu *c = &per_cpu(cpu_devices, cpu);
-	acpi_status status;
 	unsigned long long sta;
-
-	if (!acpi_handle) {
-		pr_err_once("Removing a CPU without associated ACPI handle\n");
-		return;
-	}
+	acpi_status status;
 
 	status = acpi_evaluate_integer(acpi_handle, "_STA", NULL, &sta);
-	if (ACPI_FAILURE(status))
-		return;
-
-	/* For now do not allow anything that looks like physical CPU HP */
-	if (cpu_present(cpu) && !(sta & ACPI_STA_DEVICE_PRESENT)) {
+	if (!ACPI_FAILURE(status) &&
+	    cpu_present(cpu) && !(sta & ACPI_STA_DEVICE_PRESENT))
 		pr_err_once("Changing CPU present bit is not supported\n");
-		return;
-	}
 
 	unregister_cpu(c);
 }
author maintainer
#1467 lore raw

PATCH 1461/1611 cpu/hotplug: Fix NULL kobject warning in cpuhp_smt_enable()

In reply to Greg Kroah-Hartman

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

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

From: Jinjie Ruan <[email protected]>

commit f9a82544c7174851f5c7524622f5966dcafd3a47 upstream.

On arm64, when booting with maxcpus greater than the number of present CPUs (e.g., QEMU -smp cpus=4,maxcpus=8), some CPUs are marked as 'present' but have not yet been registered via register_cpu(). Consequently, the per-cpu device objects for these CPUs are not yet initialized.

In cpuhp_smt_enable(), the code iterates over all present CPUs. Calling _cpu_up() for these unregistered CPUs eventually leads to sysfs_create_group() being called with a NULL kobject (or a kobject without a directory), triggering the following warning in fs/sysfs/group.c:

  WARNING: fs/sysfs/group.c:137 at internal_create_group+0x41c/0x4bc, CPU#2: sh/181
  [...]
  Call trace:
    internal_create_group+0x41c/0x4bc (P)
    sysfs_create_group+0x18/0x24
    topology_add_dev+0x1c/0x28
    cpuhp_invoke_callback+0x104/0x20c
    __cpuhp_invoke_callback_range+0x94/0x11c
    _cpu_up+0x200/0x37c

When booting with ACPI, arm64 smp_prepare_cpus() currently sets all enumerated CPUs as "present" regardless of their status in the MADT. This causes issues with SMT hotplug control. For instance, with QEMU's "-smp 4,maxcpus=8" configuration, the MADT GICC entries are populated as follows:

1. The first four CPUs: Enabled set but Online Capable not set.

2. The remaining four CPUs: Online Capable set but Enabled not set to support potential hot-plugging.

Fix this by:

1. When booting with ACPI, checking the ACPI_MADT_ENABLED flag in the GICC entry before calling set_cpu_present() during SMP initialization.

2. Properly managing the present mask in acpi_map_cpu() and acpi_unmap_cpu() to support actual CPU hotplug events, This aligns with other architectures like x86 and LoongArch.

3. Update the arm64 CPU hotplug documentation to no longer state that all online-capable vCPUs are marked as present by the kernel at boot time.

This ensures that only physically available or explicitly enabled CPUs are in the present mask, keeping the SMT control logic consistent with the actual hardware state.

---
 Documentation/arch/arm64/cpu-hotplug.rst |   28 ++++++++++++++++------------
 arch/arm64/kernel/acpi.c                 |    2 ++
 arch/arm64/kernel/smp.c                  |   12 +++++++++++-
 3 files changed, 29 insertions(+), 13 deletions(-)
Documentation/arch/arm64/cpu-hotplug.rst +16 -12
@@ -47,11 +47,12 @@ ever have can be described at boot. Ther
 as such devices are emulated.
 
 CPU Hotplug on virtual systems is supported. It is distinct from physical
-CPU Hotplug as all resources are described as ``present``, but CPUs may be
-marked as disabled by firmware. Only the CPU's online/offline behaviour is
-influenced by firmware. An example is where a virtual machine boots with a
-single CPU, and additional CPUs are added once a cloud orchestrator deploys
-the workload.
+CPU Hotplug as all vCPU resources are statically described in the firmware
+configuration tables (e.g. MADT), meaning their maximum possible count is
+known at boot. However, vCPUs that are not enabled at boot are not marked
+as ``present`` by the kernel until they are hotplugged. An example is where
+a virtual machine boots with a single CPU, and additional CPUs are added
+once a cloud orchestrator deploys the workload.
 
 For a virtual machine, the VMM (e.g. Qemu) plays the part of firmware.
 
@@ -60,16 +61,19 @@ brought online. Firmware can enforce its
 ``DENIED``.
 
 The ACPI tables must describe all the resources of the virtual machine. CPUs
-that firmware wishes to disable either from boot (or later) should not be
-``enabled`` in the MADT GICC structures, but should have the ``online capable``
-bit set, to indicate they can be enabled later. The boot CPU must be marked as
-``enabled``.  The 'always on' GICR structure must be used to describe the
-redistributors.
+that are hot-pluggable must have the ``online capable`` bit set and the
+``enabled`` bit cleared in the MADT GICC structures to indicate they can be
+enabled later. The boot CPU must be marked as ``enabled`` with its
+``online capable`` bit cleared. The 'always on' GICR structure must be used
+to describe the redistributors.
 
 CPUs described as ``online capable`` but not ``enabled`` can be set to enabled
 by the DSDT's Processor object's _STA method. On virtual systems the _STA method
-must always report the CPU as ``present``. Changes to the firmware policy can
-be notified to the OS via device-check or eject-request.
+must always set the ``ACPI_STA_DEVICE_PRESENT`` bit, while toggling the
+``ACPI_STA_DEVICE_ENABLED`` bit to reflect its plug status. The kernel will
+then dynamically mark the vCPU as ``present`` within the OS when the
+``ACPI_STA_DEVICE_ENABLED`` bit becomes set during hot-add. Changes to the
+firmware policy can be notified to the OS via device-check or eject-request.
 
 CPUs described as ``enabled`` in the static table, should not have their _STA
 modified dynamically by firmware. Soft-restart features such as kexec will
arch/arm64/kernel/acpi.c +2 -0
@@ -448,12 +448,14 @@ int acpi_map_cpu(acpi_handle handle, phy
 		return *pcpu;
 	}
 
+	set_cpu_present(*pcpu, true);
 	return 0;
 }
 EXPORT_SYMBOL(acpi_map_cpu);
 
 int acpi_unmap_cpu(int cpu)
 {
+	set_cpu_present(cpu, false);
 	return 0;
 }
 EXPORT_SYMBOL(acpi_unmap_cpu);
arch/arm64/kernel/smp.c +11 -1
@@ -556,6 +556,11 @@ struct acpi_madt_generic_interrupt *acpi
 }
 EXPORT_SYMBOL_GPL(acpi_cpu_get_madt_gicc);
 
+static bool acpi_cpu_is_present(int cpu)
+{
+	return acpi_cpu_get_madt_gicc(cpu)->flags & ACPI_MADT_ENABLED;
+}
+
 /*
  * acpi_map_gic_cpu_interface - parse processor MADT entry
  *
@@ -660,6 +665,10 @@ static void __init acpi_parse_and_init_c
 		early_map_cpu_to_node(i, acpi_numa_get_nid(i));
 }
 #else
+static bool acpi_cpu_is_present(int cpu)
+{
+	return false;
+}
 #define acpi_parse_and_init_cpus(...)	do { } while (0)
 #endif
 
@@ -798,7 +807,8 @@ void __init smp_prepare_cpus(unsigned in
 		if (err)
 			continue;
 
-		set_cpu_present(cpu, true);
+		if (acpi_disabled || acpi_cpu_is_present(cpu))
+			set_cpu_present(cpu, true);
 		numa_store_cpu_info(cpu);
 	}
 }
author maintainer
#1468 lore raw

PATCH 1462/1611 fs/resctrl: Free mon_data structures on rdt_get_tree() failure

In reply to Greg Kroah-Hartman

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

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

From: Tony Luck <[email protected]>

commit ca0676ae2e1a23ac3e858224fc630b99f9e216ea upstream.

If mkdir_mondata_all() or a subsequent call in rdt_get_tree() fails, the mon_data structures allocated by mon_get_kn_priv() are leaked.

Add mon_put_kn_priv() to the out_mongrp error path to free the mon_data structures.

---
 fs/resctrl/rdtgroup.c |    3 +++
 1 file changed, 3 insertions(+)
fs/resctrl/rdtgroup.c +3 -0
@@ -73,6 +73,8 @@ static int rdtgroup_setup_root(struct rd
 
 static void rdtgroup_destroy_root(void);
 
+static void mon_put_kn_priv(void);
+
 struct dentry *debugfs_resctrl;
 
 /*
@@ -2807,6 +2809,7 @@ out_mondata:
 		kernfs_remove(kn_mondata);
 out_mongrp:
 	if (resctrl_arch_mon_capable()) {
+		mon_put_kn_priv();
 		rdtgroup_unassign_cntrs(&rdtgroup_default);
 		kernfs_remove(kn_mongrp);
 	}
author maintainer
#1469 lore raw

PATCH 1463/1611 fs/resctrl: Fix double-add of pseudo-locked regions RMID to free list

In reply to Greg Kroah-Hartman

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

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

From: Reinette Chatre <[email protected]>

commit b9f089723aee892efc77c349ae47a6b452b293c4 upstream.

A pseudo-locked group's RMID is freed when it is created. On unmount rmdir_all_sub() unconditionally frees all RMID of all groups, resulting in a double-free of the pseudo-locked group's RMID. The consequence of this is that the original free results in the pseudo-locked group's RMID being added to the rmid_free_lru linked list and the second free then attempts to add the same RMID entry to the rmid_free_lru again.

Do not double-free a pseudo-locked group's RMID.

---
 fs/resctrl/rdtgroup.c |   12 +++++++-----
 1 file changed, 7 insertions(+), 5 deletions(-)
fs/resctrl/rdtgroup.c +7 -5
@@ -2986,10 +2986,6 @@ static void rmdir_all_sub(void)
 		if (rdtgrp == &rdtgroup_default)
 			continue;
 
-		if (rdtgrp->mode == RDT_MODE_PSEUDO_LOCKSETUP ||
-		    rdtgrp->mode == RDT_MODE_PSEUDO_LOCKED)
-			rdtgroup_pseudo_lock_remove(rdtgrp);
-
 		/*
 		 * Give any CPUs back to the default group. We cannot copy
 		 * cpu_online_mask because a CPU might have executed the
@@ -3000,7 +2996,13 @@ static void rmdir_all_sub(void)
 
 		rdtgroup_unassign_cntrs(rdtgrp);
 
-		free_rmid(rdtgrp->closid, rdtgrp->mon.rmid);
+		if (rdtgrp->mode == RDT_MODE_PSEUDO_LOCKSETUP ||
+		    rdtgrp->mode == RDT_MODE_PSEUDO_LOCKED) {
+			rdtgroup_pseudo_lock_remove(rdtgrp);
+		} else {
+			/* Pseudo-locked group's RMID is freed during setup. */
+			free_rmid(rdtgrp->closid, rdtgrp->mon.rmid);
+		}
 
 		kernfs_remove(rdtgrp->kn);
 		list_del(&rdtgrp->rdtgroup_list);
author maintainer
#1470 lore raw

PATCH 1464/1611 ata: libata-core: Skip HPA resize for locked drives

In reply to Greg Kroah-Hartman

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

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

From: TJ Adams <[email protected]>

commit 917d0a4b95ea7ba01ed6296fb808f752d5d81107 upstream.

Skip HPA resize in ata_hpa_resize() if the drive is security locked. If the drive is locked, the command to read the native max address fails with -EACCES, which currently causes the sticky quirk ATA_QUIRK_BROKEN_HPA to be set on the device.

Setting this sticky quirk causes subsequent revalidations (after the drive is unlocked) to bypass HPA checks, preventing the unlocked drive from exposing its full native capacity without a reboot or device removal.

---
 drivers/ata/libata-core.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/ata/libata-core.c +1 -1
@@ -1324,7 +1324,7 @@ static int ata_hpa_resize(struct ata_dev
 	/* do we need to do it? */
 	if ((dev->class != ATA_DEV_ATA && dev->class != ATA_DEV_ZAC) ||
 	    !ata_id_has_lba(dev->id) || !ata_id_hpa_enabled(dev->id) ||
-	    (dev->quirks & ATA_QUIRK_BROKEN_HPA))
+	    (dev->quirks & ATA_QUIRK_BROKEN_HPA) || ata_id_is_locked(dev->id))
 		return 0;
 
 	/* read native max address */
author maintainer
#1471 lore raw

PATCH 1465/1611 ata: libata-core: Allow capacity transition to zero for locked drives

In reply to Greg Kroah-Hartman

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

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

From: TJ Adams <[email protected]>

commit 793bf193b18e9bff6c4280268bbffd16a5b533e5 upstream.

Commit 91842ed844a0 ("ata: libata-core: Set capacity to zero for a security locked drive") introduced setting the device capacity (n_sectors) to zero in ata_dev_configure() if the drive is security locked.

However, during runtime revalidation, ata_dev_revalidate() compares the new capacity (now 0) with the old capacity (>0) and detects a mismatch. Since it does not consider the locked status, it returns -ENODEV.

This revalidation failure can occur when doing a reset of the PHY (e.g. hard reset) for a controller that has I/Os in flight. The timed out I/Os trigger the SCSI Error Handling (EH) path, which in turn invokes libata device revalidation. If the drive is locked at runtime (e.g. it lost power during reset and relocked), revalidation sees the capacity transition to zero and fails, eventually disabling the device.

Fix this by allowing the capacity transition to zero in ata_dev_revalidate() if the drive is reported as security locked by ata_id_is_locked().

Fixes 91842ed844a0 ("ata: libata-core: Set capacity to zero for a security locked drive")
Assisted-by Antigravity:gemini-3.5-flash
---
 drivers/ata/libata-core.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/ata/libata-core.c +1 -1
@@ -3963,7 +3963,7 @@ int ata_dev_revalidate(struct ata_device
 
 	/* verify n_sectors hasn't changed */
 	if (dev->class != ATA_DEV_ATA || !n_sectors ||
-	    dev->n_sectors == n_sectors)
+	    dev->n_sectors == n_sectors || ata_id_is_locked(dev->id))
 		return 0;
 
 	/* n_sectors has changed */
author maintainer
#1472 lore raw

PATCH 1466/1611 drbd: reject data replies with an out-of-range payload size

In reply to Greg Kroah-Hartman

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

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

From: Michael Bommarito <[email protected]>

commit bd910a7660d280595ef94cb6d193951d855d330f upstream.

recv_dless_read() receives a P_DATA_REPLY from a peer into the bio of an outstanding read request. The peer-supplied payload length reaches it as the signed int data_size, and two peer-controlled inputs can make it negative. With a negotiated data-integrity-alg the digest length is subtracted first, so a reply whose payload is smaller than the digest underflows data_size. With no integrity algorithm (the default) data_size is assigned from the unsigned h95/h100 wire length and drbdd() never bounds it for a payload-carrying command, so a length above INT_MAX casts it negative; this path needs no non-default feature. The bio receive loop then computes expect = min_t(int, data_size, bv_len), which is negative, and drbd_recv_all_warn(mapped, expect) receives with a size_t of SIZE_MAX into the first mapped page.

The sibling receive path read_in_block() is not affected: it uses an unsigned size and rejects it against DRBD_MAX_BIO_SIZE before receiving. Reject a data reply whose size is negative after the optional digest subtraction, covering both triggers.

Impact: a malicious or man-in-the-middle DRBD peer copies attacker-chosen bytes past a bio page in the receiver, corrupting kernel memory. A node that reads from its peer (a diskless node, or read-balancing to the peer) is exposed in the default configuration; data-integrity-alg is not required.

Fixes b411b3637fa7 ("The DRBD driver")
Assisted-by Codex:gpt-5-5-xhigh
Assisted-by Claude:claude-opus-4-8
---
 drivers/block/drbd/drbd_receiver.c |    5 +++++
 1 file changed, 5 insertions(+)
drivers/block/drbd/drbd_receiver.c +5 -0
@@ -1810,6 +1810,11 @@ static int recv_dless_read(struct drbd_p
 		data_size -= digest_size;
 	}
 
+	if (data_size < 0) {
+		drbd_err(peer_device, "Invalid data reply size\n");
+		return -EIO;
+	}
+
 	/* optimistically update recv_cnt.  if receiving fails below,
 	 * we disconnect anyways, and counters will be reset. */
 	peer_device->device->recv_cnt += data_size>>9;
author maintainer
#1473 lore raw

PATCH 1467/1611 riscv: Prevent NULL pointer dereference in machine_kexec_prepare()

In reply to Greg Kroah-Hartman

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

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

From: Tao Liu <[email protected]>

commit 81bbcff0c053c4f5c711c31a9b72fc492bd96c3f upstream.

A NULL pointer dereference issue is noticed in riscv's machine_kexec_prepare(), where image->segment[i].buf might be NULL and copied unchecked.

The NULL buf comes from ima_add_kexec_buffer(), where kbuf is added by kexec_add_buffer(), but kbuf.buffer is NULL, then it is copied without a check in machine_kexec_prepare():

kexec_file_load

    -> kimage_file_alloc_init()
       -> kimage_file_prepare_segments()
          -> ima_add_kexec_buffer()
             -> kexec_add_buffer()
    -> machine_kexec_prepare()
       -> memcpy()

Address this by adding a check before the data copy attempt.

---
 arch/riscv/kernel/machine_kexec.c |    3 +++
 1 file changed, 3 insertions(+)
arch/riscv/kernel/machine_kexec.c +3 -0
@@ -41,6 +41,9 @@ machine_kexec_prepare(struct kimage *ima
 		if (image->segment[i].memsz <= sizeof(fdt))
 			continue;
 
+		if (!image->segment[i].buf)
+			continue;
+
 		if (image->file_mode)
 			memcpy(&fdt, image->segment[i].buf, sizeof(fdt));
 		else if (copy_from_user(&fdt, image->segment[i].buf, sizeof(fdt)))
author maintainer
#1474 lore raw

PATCH 1468/1611 tracing/osnoise: Call synchronize_rcu() when unregistering

In reply to Greg Kroah-Hartman

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

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

From: Crystal Wood <[email protected]>

commit fe58f457ad8d0a2bef4e053cfecca4b5cd266b1a upstream.

This ensures that any RCU readers traversing the instance list have finished, before releasing the reference on the tracer that the instance points to.

---
 kernel/trace/trace_osnoise.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
kernel/trace/trace_osnoise.c +3 -1
@@ -160,7 +160,9 @@ static void osnoise_unregister_instance(
 	if (!found)
 		return;
 
-	kvfree_rcu_mightsleep(inst);
+	/* Do a full sync to ensure that tr remains valid, not just inst */
+	synchronize_rcu();
+	kvfree(inst);
 }
 
 /*
author maintainer
#1475 lore raw

PATCH 1469/1611 s390/diag: Add missing array_index_nospec() call to memtop_get_page_count()

In reply to Greg Kroah-Hartman

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

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

From: Heiko Carstens <[email protected]>

commit b7577fe4c47a31ca7c99714c53244a44af03cdfe upstream.

'level' is user space controlled and used to read from an array. Add the missing array_index_nospec() call to prevent speculative execution.

Fixes 0d30871739ab ("s390/diag: Add memory topology information via diag310")
---
 arch/s390/kernel/diag/diag310.c |    7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)
arch/s390/kernel/diag/diag310.c +4 -3
@@ -190,17 +190,18 @@ static int memtop_get_stride_len(unsigne
 static int memtop_get_page_count(unsigned long *res, unsigned long level)
 {
 	static unsigned long memtop_pages[DIAG310_LEVELMAX];
-	unsigned long pages;
+	unsigned long pages, idx;
 	int rc;
 
 	if (level > DIAG310_LEVELMAX || level < DIAG310_LEVELMIN)
 		return -EINVAL;
-	pages = READ_ONCE(memtop_pages[level - 1]);
+	idx = array_index_nospec(level - 1, ARRAY_SIZE(memtop_pages));
+	pages = READ_ONCE(memtop_pages[idx]);
 	if (!pages) {
 		rc = diag310_get_memtop_size(&pages, level);
 		if (rc)
 			return rc;
-		WRITE_ONCE(memtop_pages[level - 1], pages);
+		WRITE_ONCE(memtop_pages[idx], pages);
 	}
 	*res = pages;
 	return 0;
author maintainer
#1476 lore raw

PATCH 1470/1611 s390/mm: Fix type mismatch in get_align_mask().

In reply to Greg Kroah-Hartman

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

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

From: Gerald Schaefer <[email protected]>

commit ec84aad4c3594307d103af563991b4415ac5c8ab upstream.

Commit 86f48f922ba79 ("s390/mmap: disable mmap alignment when randomize_va_space = 0") introduced get_align_mask() with return type of 'int', while the target field 'info.align_mask' in struct vm_unmapped_area_info is 'unsigned long'.

With currently used masks, this should not cause truncation issues, but fix it and return 'unsigned long' to avoid future problems.

Fixes 86f48f922ba79 ("s390/mmap: disable mmap alignment when randomize_va_space = 0")
---
 arch/s390/mm/mmap.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/s390/mm/mmap.c +1 -1
@@ -65,7 +65,7 @@ static inline unsigned long mmap_base(un
 	return PAGE_ALIGN(STACK_TOP - gap - rnd);
 }
 
-static int get_align_mask(struct file *filp, unsigned long flags)
+static unsigned long get_align_mask(struct file *filp, unsigned long flags)
 {
 	if (filp && is_file_hugepages(filp))
 		return huge_page_mask_align(filp);
author maintainer
#1477 lore raw

PATCH 1471/1611 selftests/rseq: Fix a building error for riscv arch

In reply to Greg Kroah-Hartman

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

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

From: Hui Wang <[email protected]>

commit a2ac823d8a228e392bc6a79ea63b550ab84a55b2 upstream.

RISC-V rseq selftests include asm/fence.h from tools/arch/riscv, but the rseq Makefile only adds tools/include in the CFLAGS, this results in the building failure both for native and cross build:

In file included from rseq.h:131,
from rseq.c:37:
rseq-riscv.h:11:10: fatal error: asm/fence.h: No such file or directory

To fix it, add the matching tools/arch/$(ARCH)/include path in the CFLAGS and derive ARCH from SUBARCH for standalone native builds where ARCH is not set.

Fixes c92786e179e0 ("KVM: riscv: selftests: Use the existing RISCV_FENCE macro in `rseq-riscv.h`")
---
 tools/testing/selftests/rseq/Makefile | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
tools/testing/selftests/rseq/Makefile +5 -1
index 50d69e22ee7a..aba6317f6cb8 100644
@@ -5,9 +5,13 @@ CLANG_FLAGS += -no-integrated-as
 endif
 
 top_srcdir = ../../../..
+include $(top_srcdir)/scripts/subarch.include
+ARCH ?= $(SUBARCH)
+LINUX_TOOL_ARCH_INCLUDE = $(top_srcdir)/tools/arch/$(ARCH)/include
 
 CFLAGS += -O2 -Wall -g -I./ $(KHDR_INCLUDES) -L$(OUTPUT) -Wl,-rpath=./ \
-	  $(CLANG_FLAGS) -I$(top_srcdir)/tools/include
+	  $(CLANG_FLAGS) -I$(top_srcdir)/tools/include \
+	  -I$(LINUX_TOOL_ARCH_INCLUDE)
 LDLIBS += -lpthread -ldl
 
 # Own dependencies because we only want to build against 1st prerequisite, but
Signature · 1 line
2.55.0
author maintainer
#1478 lore raw

PATCH 1472/1611 cgroup/cpuset: rebind mm mempolicy to effective_mems, not mems_allowed

In reply to Greg Kroah-Hartman

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

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

From: Farhad Alemi <[email protected]>

commit b983c56426383e4a06fa5970c4e33cee879b1482 upstream.

Creating a child cpuset where cpuset.mems is never set leads to a div/0 when a VMA mempolicy with MPOL_F_RELATIVE_NODES rebinds in response to a CPU hotplug event.

Reproduction steps:
1) Create a cgroup w/ cpuset controls (do not set cpuset.mems)
2) Move the task into the child cpuset
3) Create a VMA mempolicy for that task with MPOL_F_RELATIVE_NODES
4) unplug and hotplug a cpu
echo 0 > /sys/devices/system/cpu/cpu1/online
echo 1 > /sys/devices/system/cpu/cpu1/online
5) mempolicy rebind does a div/0 in mpol_relative_nodemask on the call to __nodes_fold()

The cpuset code passes (cs->mems_allowed) which is not guaranteed to have nodes to the rebind routine. Use cs->effective_mems instead, which is guaranteed to have a non-empty nodemask once we reach that code path.

Fixes ae1c802382f7 ("cpuset: apply cs->effective_{cpus,mems}")
Cc Andrew Morton <[email protected]>
Cc Alistair Popple <[email protected]>
Cc Byungchul Park <[email protected]>
Cc Gregory Price <[email protected]>
Cc "Huang, Ying" <[email protected]>
Cc Joshua Hahn <[email protected]>
Cc Matthew Brost <[email protected]>
Cc Rakie Kim <[email protected]>
Cc Rasmus Villemoes <[email protected]>
Cc Zi Yan <[email protected]>
Cc Tejun Heo <[email protected]>
Cc Ridong Chen <[email protected]>
Cc Johannes Weiner <[email protected]>
Cc "Michal Koutný" <[email protected]>
[ david: add a comment, slightly rephrase description ]
---
 kernel/cgroup/cpuset.c |    7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)
kernel/cgroup/cpuset.c +6 -1
@@ -2774,7 +2774,12 @@ void cpuset_update_tasks_nodemask(struct
 
 		migrate = is_memory_migrate(cs);
 
-		mpol_rebind_mm(mm, &cs->mems_allowed);
+		/*
+		 * For v1 we can have empty effective_mems, but we cannot
+		 * attach any tasks (see cpuset_can_attach_check()). For v2,
+		 * effective_mems is guaranteed to not be empty.
+		 */
+		mpol_rebind_mm(mm, &cs->effective_mems);
 		if (migrate)
 			cpuset_migrate_mm(mm, &cs->old_mems_allowed, &newmems);
 		else
author maintainer
#1479 lore raw

PATCH 1473/1611 pmdomain: imx: Fix i.MX8MP power notifier

In reply to Greg Kroah-Hartman

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

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

From: Peng Fan <[email protected]>

commit 72422525f641f68bed6ca3389d29ee3f41fdea33 upstream.

Using imx8mm_vpu_power_notifier() for i.MX8MP is wrong, as it ungates the VPU clocks to provide the ADB clock, which is necessary on i.MX8MM, but on i.MX8MP there is a separate gate (bit 3) for the NoC. So add imx8mp_vpu_power_notifier() for i.MX8MP.

Fixes a1a5f15f7f6cb ("soc: imx: imx8m-blk-ctrl: add i.MX8MP VPU blk ctrl")
---
 drivers/pmdomain/imx/imx8m-blk-ctrl.c |   27 ++++++++++++++++++++++++++-
 1 file changed, 26 insertions(+), 1 deletion(-)
drivers/pmdomain/imx/imx8m-blk-ctrl.c +26 -1
@@ -514,9 +514,34 @@ static const struct imx8m_blk_ctrl_domai
 	},
 };
 
+static int imx8mp_vpu_power_notifier(struct notifier_block *nb,
+				     unsigned long action, void *data)
+{
+	struct imx8m_blk_ctrl *bc = container_of(nb, struct imx8m_blk_ctrl,
+						 power_nb);
+
+	if (action == GENPD_NOTIFY_ON) {
+		/*
+		 * On power up we have no software backchannel to the GPC to
+		 * wait for the ADB handshake to happen, so we just delay for a
+		 * bit. On power down the GPC driver waits for the handshake.
+		 */
+
+		udelay(5);
+
+		/* set "fuse" bits to enable the VPUs */
+		regmap_set_bits(bc->regmap, 0x8, 0xffffffff);
+		regmap_set_bits(bc->regmap, 0xc, 0xffffffff);
+		regmap_set_bits(bc->regmap, 0x10, 0xffffffff);
+		regmap_set_bits(bc->regmap, 0x14, 0xffffffff);
+	}
+
+	return NOTIFY_OK;
+}
+
 static const struct imx8m_blk_ctrl_data imx8mp_vpu_blk_ctl_dev_data = {
 	.max_reg = 0x18,
-	.power_notifier_fn = imx8mm_vpu_power_notifier,
+	.power_notifier_fn = imx8mp_vpu_power_notifier,
 	.domains = imx8mp_vpu_blk_ctl_domain_data,
 	.num_domains = ARRAY_SIZE(imx8mp_vpu_blk_ctl_domain_data),
 };
author maintainer
#1480 lore raw

PATCH 1474/1611 pmdomain: imx: Fix i.MX8MP VC8000E power up sequence

In reply to Greg Kroah-Hartman

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

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

From: Peng Fan <[email protected]>

commit 25e252bcf1593b420b12a7231d9dd64b885a2ae2 upstream.

Per errata[1]:
ERR050531: VPU_NOC power down handshake may hang during VC8000E/VPUMIX power up/down cycling.
Description: VC8000E reset de-assertion edge and AXI clock may have a timing issue.
Workaround: Set bit2 (vc8000e_clk_en) of BLK_CLK_EN_CSR to 0 to gate off both AXI clock and VC8000E clock sent to VC8000E and AXI clock sent to VPU_NOC m_v_2 interface during VC8000E power up(VC8000E reset is de-asserted by HW)

Add a bool variable is_errata_err050531 in
'struct imx8m_blk_ctrl_domain_data' to represent whether the workaround is needed. If is_errata_err050531 is true, first clear the clk before powering up gpc, then enable the clk after powering up gpc.

[1] https://www.nxp.com/webapp/Download?colCode=IMX8MP_1P33A

Fixes a1a5f15f7f6cb ("soc: imx: imx8m-blk-ctrl: add i.MX8MP VPU blk ctrl")
---
 drivers/pmdomain/imx/imx8m-blk-ctrl.c |   19 ++++++++++++++++++-
 1 file changed, 18 insertions(+), 1 deletion(-)
drivers/pmdomain/imx/imx8m-blk-ctrl.c +18 -1
@@ -54,6 +54,15 @@ struct imx8m_blk_ctrl_domain_data {
 	 * register.
 	 */
 	u32 mipi_phy_rst_mask;
+
+	/*
+	 * VC8000E reset de-assertion edge and AXI clock may have a timing issue.
+	 * Workaround: Set bit2 (vc8000e_clk_en) of BLK_CLK_EN_CSR to 0 to gate off
+	 * both AXI clock and VC8000E clock sent to VC8000E and AXI clock sent to
+	 * VPU_NOC m_v_2 interface during VC8000E power up(VC8000E reset is
+	 * de-asserted by HW)
+	 */
+	bool is_errata_err050531;
 };
 
 #define DOMAIN_MAX_CLKS 4
@@ -108,7 +117,11 @@ static int imx8m_blk_ctrl_power_on(struc
 		dev_err(bc->dev, "failed to enable clocks\n");
 		goto bus_put;
 	}
-	regmap_set_bits(bc->regmap, BLK_CLK_EN, data->clk_mask);
+
+	if (data->is_errata_err050531)
+		regmap_clear_bits(bc->regmap, BLK_CLK_EN, data->clk_mask);
+	else
+		regmap_set_bits(bc->regmap, BLK_CLK_EN, data->clk_mask);
 
 	/* power up upstream GPC domain */
 	ret = pm_runtime_get_sync(domain->power_dev);
@@ -117,6 +130,9 @@ static int imx8m_blk_ctrl_power_on(struc
 		goto clk_disable;
 	}
 
+	if (data->is_errata_err050531)
+		regmap_set_bits(bc->regmap, BLK_CLK_EN, data->clk_mask);
+
 	/* wait for reset to propagate */
 	udelay(5);
 
@@ -511,6 +527,7 @@ static const struct imx8m_blk_ctrl_domai
 		.clk_mask = BIT(2),
 		.path_names = (const char *[]){"vc8000e"},
 		.num_paths = 1,
+		.is_errata_err050531 = true,
 	},
 };
 
author maintainer
#1481 lore raw

PATCH 1475/1611 selftests/landlock: Skip scoped_signal subtest with MSG_OOB if not available

In reply to Greg Kroah-Hartman

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

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

From: Thomas Huth <[email protected]>

commit 5ab1dc6d110db6bee167a32fd94c53ea0e7ad6d2 upstream.

MSG_OOB might be disabled in the kernel for unix sockets (by not selecting CONFIG_AF_UNIX_OOB), and in this case the related tests of the scoped_signal_test are currently failing. Add a runtime probe using socketpair() to detect MSG_OOB support and skip the test gracefully if it is unavailable.

Fixes f34e9ce5f479 ("selftests/landlock: Test signal created by out-of-bound message")
---
 tools/testing/selftests/landlock/scoped_signal_test.c |   21 ++++++++++++++++++
 1 file changed, 21 insertions(+)
tools/testing/selftests/landlock/scoped_signal_test.c +21 -0
@@ -400,6 +400,24 @@ static int setup_signal_handler(int sign
 	return sigaction(SIGURG, &sa, NULL);
 }
 
+/*
+ * MSG_OOB might be disabled in the kernel via the CONFIG_AF_UNIX_OOB
+ * switch, so this function can be used for probing for its availability.
+ */
+static bool has_af_unix_oob(void)
+{
+	bool available = false;
+	int sp[2];
+
+	if (socketpair(AF_UNIX, SOCK_STREAM, 0, sp) == 0) {
+		available = (send(sp[0], ".", 1, MSG_OOB) == 1);
+		close(sp[0]);
+		close(sp[1]);
+	}
+
+	return available;
+}
+
 /* clang-format off */
 FIXTURE(fown) {};
 /* clang-format on */
@@ -462,6 +480,9 @@ TEST_F(fown, sigurg_socket)
 	int pipe_parent[2], pipe_child[2];
 	pid_t child;
 
+	if (!has_af_unix_oob())
+		SKIP(return, "CONFIG_AF_UNIX_OOB / MSG_OOB not available");
+
 	memset(&server_address, 0, sizeof(server_address));
 	set_unix_address(&server_address, 0);
 
author maintainer
#1482 lore raw

PATCH 1476/1611 selftests/landlock: Fix screwed up pointers in the scoped_signal_test

In reply to Greg Kroah-Hartman

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

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

From: Thomas Huth <[email protected]>

commit d793186aa3bb833c878ae6826c87e62c843afaa3 upstream.

The scoped_signal_test uses pthread_join(..., (void **)&ret)) in a couple of places, i.e. the return value of the thread is stored in the shape of a "void *" into the memory location of &ret. Pointers are 64-bit on modern computers, but the ret variable is declared as a simple "enum thread_return" which is only 32 bits. So the pthread_join() will overflow the ret variable by 4 byte.

The problem is very visible on big endian systems like s390x where the test is failing: The least significant byte that carries the return code of the thread is not written into the ret variable here, but somewhere else in the stack frame, so the comparison for the right return code is failing here.

Fix it by getting rid of the enum and defining the THREAD_* constants and "ret" variables as proper "void *" pointers. This way we can also get rid of some ugly (void *) castings in a couple of spots.

Fixes c8994965013e ("selftests/landlock: Test signal scoping for threads")
[mic: Add clang-format markups]
---
 tools/testing/selftests/landlock/scoped_signal_test.c |   46 +++++++++---------
 1 file changed, 23 insertions(+), 23 deletions(-)
tools/testing/selftests/landlock/scoped_signal_test.c +23 -23
@@ -249,12 +249,12 @@ TEST_F(scoped_domains, check_access_sign
 		_metadata->exit_code = KSFT_FAIL;
 }
 
-enum thread_return {
-	THREAD_INVALID = 0,
-	THREAD_SUCCESS = 1,
-	THREAD_ERROR = 2,
-	THREAD_TEST_FAILED = 3,
-};
+/* clang-format off */
+#define THREAD_INVALID		((void *)0)
+#define THREAD_SUCCESS		((void *)1)
+#define THREAD_ERROR		((void *)2)
+#define THREAD_TEST_FAILED	((void *)3)
+/* clang-format on */
 
 static void *thread_sync(void *arg)
 {
@@ -262,15 +262,15 @@ static void *thread_sync(void *arg)
 	char buf;
 
 	if (read(pipe_read, &buf, 1) != 1)
-		return (void *)THREAD_ERROR;
+		return THREAD_ERROR;
 
-	return (void *)THREAD_SUCCESS;
+	return THREAD_SUCCESS;
 }
 
 TEST(signal_scoping_thread_before)
 {
 	pthread_t no_sandbox_thread;
-	enum thread_return ret = THREAD_INVALID;
+	void *ret = THREAD_INVALID;
 	int thread_pipe[2];
 
 	drop_caps(_metadata);
@@ -285,7 +285,7 @@ TEST(signal_scoping_thread_before)
 	EXPECT_EQ(0, pthread_kill(no_sandbox_thread, 0));
 	EXPECT_EQ(1, write(thread_pipe[1], ".", 1));
 
-	EXPECT_EQ(0, pthread_join(no_sandbox_thread, (void **)&ret));
+	EXPECT_EQ(0, pthread_join(no_sandbox_thread, &ret));
 	EXPECT_EQ(THREAD_SUCCESS, ret);
 
 	EXPECT_EQ(0, close(thread_pipe[0]));
@@ -295,7 +295,7 @@ TEST(signal_scoping_thread_before)
 TEST(signal_scoping_thread_after)
 {
 	pthread_t scoped_thread;
-	enum thread_return ret = THREAD_INVALID;
+	void *ret = THREAD_INVALID;
 	int thread_pipe[2];
 
 	drop_caps(_metadata);
@@ -310,7 +310,7 @@ TEST(signal_scoping_thread_after)
 	EXPECT_EQ(0, pthread_kill(scoped_thread, 0));
 	EXPECT_EQ(1, write(thread_pipe[1], ".", 1));
 
-	EXPECT_EQ(0, pthread_join(scoped_thread, (void **)&ret));
+	EXPECT_EQ(0, pthread_join(scoped_thread, &ret));
 	EXPECT_EQ(THREAD_SUCCESS, ret);
 
 	EXPECT_EQ(0, close(thread_pipe[0]));
@@ -327,20 +327,20 @@ void *thread_setuid(void *ptr)
 	char buf;
 
 	if (read(arg->pipe_read, &buf, 1) != 1)
-		return (void *)THREAD_ERROR;
+		return THREAD_ERROR;
 
 	/* libc's setuid() should update all thread's credentials. */
 	if (getuid() != arg->new_uid)
-		return (void *)THREAD_TEST_FAILED;
+		return THREAD_TEST_FAILED;
 
-	return (void *)THREAD_SUCCESS;
+	return THREAD_SUCCESS;
 }
 
 TEST(signal_scoping_thread_setuid)
 {
 	struct thread_setuid_args arg;
 	pthread_t no_sandbox_thread;
-	enum thread_return ret = THREAD_INVALID;
+	void *ret = THREAD_INVALID;
 	int pipe_parent[2];
 	int prev_uid;
 
@@ -367,7 +367,7 @@ TEST(signal_scoping_thread_setuid)
 	EXPECT_EQ(arg.new_uid, getuid());
 	EXPECT_EQ(1, write(pipe_parent[1], ".", 1));
 
-	EXPECT_EQ(0, pthread_join(no_sandbox_thread, (void **)&ret));
+	EXPECT_EQ(0, pthread_join(no_sandbox_thread, &ret));
 	EXPECT_EQ(THREAD_SUCCESS, ret);
 
 	clear_cap(_metadata, CAP_SETUID);
@@ -688,20 +688,20 @@ static void *thread_setown_scoped(void *
 	ruleset_fd =
 		landlock_create_ruleset(&ruleset_attr, sizeof(ruleset_attr), 0);
 	if (ruleset_fd < 0)
-		return (void *)THREAD_ERROR;
+		return THREAD_ERROR;
 	if (prctl(PR_SET_NO_NEW_PRIVS, 1, 0, 0, 0) ||
 	    landlock_restrict_self(ruleset_fd, 0)) {
 		close(ruleset_fd);
-		return (void *)THREAD_ERROR;
+		return THREAD_ERROR;
 	}
 	close(ruleset_fd);
 
 	/* Makes this process group own the SIGIO source. */
 	if (fcntl(fd, F_SETSIG, SIGURG) || fcntl(fd, F_SETOWN, -getpgrp()) ||
 	    fcntl(fd, F_SETFL, O_ASYNC))
-		return (void *)THREAD_ERROR;
+		return THREAD_ERROR;
 
-	return (void *)THREAD_SUCCESS;
+	return THREAD_SUCCESS;
 }
 
 /*
@@ -723,7 +723,7 @@ TEST(sigio_to_pgid_self)
 {
 	int trigger[2];
 	pthread_t thread;
-	enum thread_return ret = THREAD_INVALID;
+	void *ret = THREAD_INVALID;
 	int i;
 
 	drop_caps(_metadata);
@@ -743,7 +743,7 @@ TEST(sigio_to_pgid_self)
 	 */
 	ASSERT_EQ(0, pthread_create(&thread, NULL, thread_setown_scoped,
 				    &trigger[0]));
-	ASSERT_EQ(0, pthread_join(thread, (void **)&ret));
+	ASSERT_EQ(0, pthread_join(thread, &ret));
 	ASSERT_EQ(THREAD_SUCCESS, ret);
 
 	/* Fans SIGURG out to the process group. */
author maintainer
#1483 lore raw

PATCH 1477/1611 mmc: sdhci-esdhc-imx: restore pinctrl before restoring ios timing on resume

In reply to Greg Kroah-Hartman

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

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

From: Luke Wang <[email protected]>

commit 1db87818bde3d2295613660879378b43a70d31f8 upstream.

SDIO devices such as WiFi may keep power during suspend, so the MMC core skips full card re-initialization on resume and directly restores the host controller's ios timing to match the card. For DDR mode, pm_runtime_force_resume() sets DDR_EN before the pin configuration is restored from sleep state.

This is related to the SoC IP integration: switching pinctrl setting (changing alt from GPIO to USDHC) impacts the internal loopback path. If pinctrl configures the pad to GPIO function, once DDR_EN is set, the DLL delay will be fixed based on the GPIO function loopback path. When the pinctrl is later changed to USDHC function, the internal loopback path changes, making the original fixed sample point no longer suitable for the current loopback path. This causes persistent read CRC errors on subsequent data transfers.

SD/eMMC running in DDR mode are unaffected as they are fully re-initialized from legacy timing after resume.

Fix this by restoring the pinctrl state based on current timing mode using esdhc_change_pinstate() before pm_runtime_force_resume(). This ensures the correct pin configuration (e.g., 100/200MHz for UHS modes) is applied before DDR_EN is set. Only restore for non-wakeup devices since wakeup devices kept their active pin state during suspend.

Fixes 676a83855614 ("mmc: host: sdhci-esdhc-imx: refactor the system PM logic")
---
 drivers/mmc/host/sdhci-esdhc-imx.c |    6 ++++++
 1 file changed, 6 insertions(+)
drivers/mmc/host/sdhci-esdhc-imx.c +6 -0
@@ -2083,6 +2083,12 @@ static int sdhci_esdhc_resume(struct dev
 	struct pltfm_imx_data *imx_data = sdhci_pltfm_priv(pltfm_host);
 	int ret;
 
+	if (!device_may_wakeup(dev)) {
+		ret = esdhc_change_pinstate(host, host->timing);
+		if (ret)
+			dev_warn(dev, "Failed to restore pinctrl state\n");
+	}
+
 	pm_runtime_force_resume(dev);
 
 	ret = mmc_gpio_set_cd_wake(host->mmc, false);
author maintainer
#1484 lore raw

PATCH 1478/1611 powerpc/pseries: fix memory leak on krealloc failure in papr_init

In reply to Greg Kroah-Hartman

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

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

From: Thorsten Blum <[email protected]>

commit bd83c98b988d2c560531084e296dbfb530aff829 upstream.

When krealloc() fails, free the original esi_buf before returning to avoid a memory leak.

Fixes 3c14b73454cf ("powerpc/pseries: Interface to represent PAPR firmware attributes")
---
 arch/powerpc/platforms/pseries/papr_platform_attributes.c |    8 +++-----
 1 file changed, 3 insertions(+), 5 deletions(-)
arch/powerpc/platforms/pseries/papr_platform_attributes.c +3 -5
@@ -271,11 +271,9 @@ retry:
 		esi_buf_size = ESI_HDR_SIZE + (CURR_MAX_ESI_ATTRS * max_esi_attrs);
 
 		temp_esi_buf = krealloc(esi_buf, esi_buf_size, GFP_KERNEL);
-		if (temp_esi_buf)
-			esi_buf = temp_esi_buf;
-		else
-			return -ENOMEM;
-
+		if (!temp_esi_buf)
+			goto out_free_esi_buf;
+		esi_buf = temp_esi_buf;
 		goto retry;
 	}
 
author maintainer
#1485 lore raw

PATCH 1479/1611 net/mlx5: free mlx5_st_idx_data on final dealloc

In reply to Greg Kroah-Hartman

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

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

From: Zhiping Zhang <[email protected]>

commit df6134b527a88b3e65ba6ae5073664af091d5fd2 upstream.

Workloads that repeatedly allocate and release mkeys carrying TPH steering-tag hints (e.g. churning RDMA MRs) leak one struct mlx5_st_idx_data per cycle; kmemleak flags it as unreferenced and the kmalloc slab grows over time.

When the last reference to an ST table entry is dropped, mlx5_st_dealloc_index() removed the entry from idx_xa but the backing mlx5_st_idx_data allocation was never freed.

Free idx_data after the xa_erase() so the lifetime of the bookkeeping struct matches the lifetime of the ST entry it tracks.

---
 drivers/net/ethernet/mellanox/mlx5/core/lib/st.c |    1 +
 1 file changed, 1 insertion(+)
drivers/net/ethernet/mellanox/mlx5/core/lib/st.c +1 -0
@@ -154,6 +154,7 @@ int mlx5_st_dealloc_index(struct mlx5_co
 
 	if (refcount_dec_and_test(&idx_data->usecount)) {
 		xa_erase(&st->idx_xa, st_index);
+		kfree(idx_data);
 		/* We leave PCI config space as was before, no mkey will refer to it */
 	}
 
author maintainer
#1486 lore raw

PATCH 1480/1611 wifi: rt2x00: avoid full teardown before work setup in probe

In reply to Greg Kroah-Hartman

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

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

From: Runyu Xiao <[email protected]>

commit 536fb3d739d75a03cb318c0c6fe799425cfea501 upstream.

rt2x00lib_probe_dev() uses the full rt2x00lib_remove_dev() teardown for all probe failures. However, drv_data allocation and workqueue allocation can fail before intf_work, autowakeup_work and sleep_work have been initialized.

Do not enter the full remove path until the probe has reached the point where those work items are set up. Return directly for drv_data allocation failure, and use a small early cleanup path for workqueue allocation failure.

This issue was found by our static analysis tool and then confirmed by manual review of rt2x00lib_probe_dev() and rt2x00lib_remove_dev(). The early probe exits should not call a common teardown path that assumes the later work setup has already completed.

A QEMU PoC forced alloc_ordered_workqueue() to fail before the work initializers are reached. The resulting fail path entered rt2x00lib_remove_dev(), and DEBUG_OBJECTS reported invalid work drains with rt2x00lib_probe_dev() and rt2x00lib_remove_dev() in the stack.

Fixes 1ebbc48520a0 ("rt2x00: Introduce concept of driver data in struct rt2x00_dev.")
Fixes 0439f5367c8d ("rt2x00: Move TX/RX work into dedicated workqueue")
---
 drivers/net/wireless/ralink/rt2x00/rt2x00dev.c |   12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)
drivers/net/wireless/ralink/rt2x00/rt2x00dev.c +10 -2
@@ -1382,7 +1382,7 @@ int rt2x00lib_probe_dev(struct rt2x00_de
 			                      GFP_KERNEL);
 		if (!rt2x00dev->drv_data) {
 			retval = -ENOMEM;
-			goto exit;
+			return retval;
 		}
 	}
 
@@ -1416,7 +1416,7 @@ int rt2x00lib_probe_dev(struct rt2x00_de
 	    alloc_ordered_workqueue("%s", 0, wiphy_name(rt2x00dev->hw->wiphy));
 	if (!rt2x00dev->workqueue) {
 		retval = -ENOMEM;
-		goto exit;
+		goto exit_free_drv_data;
 	}
 
 	INIT_WORK(&rt2x00dev->intf_work, rt2x00lib_intf_scheduled);
@@ -1489,6 +1489,14 @@ exit:
 	rt2x00lib_remove_dev(rt2x00dev);
 
 	return retval;
+
+exit_free_drv_data:
+	clear_bit(DEVICE_STATE_PRESENT, &rt2x00dev->flags);
+
+	kfree(rt2x00dev->drv_data);
+	rt2x00dev->drv_data = NULL;
+
+	return retval;
 }
 EXPORT_SYMBOL_GPL(rt2x00lib_probe_dev);
 
author maintainer
#1487 lore raw

PATCH 1481/1611 wifi: mwifiex: fix roaming to different channel in host_mlme mode

In reply to Greg Kroah-Hartman

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

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

From: Rafael Beims <[email protected]>

commit a707e4127c0f893c7a7703500ab56297a5bd2d51 upstream.

When host MLME is enabled, mwifiex_cfg80211_authenticate() transmits the authentication frame on a remain-on-channel (ROC) reservation so that the frame is sent on the target BSS's channel. The ROC is only configured when priv->auth_flag is zero.

priv->auth_flag is set to HOST_MLME_AUTH_PENDING when the auth frame is queued and advances to HOST_MLME_AUTH_DONE once authentication completes. It is only cleared back to zero on a disconnect, deauth or timeout path; nothing clears it when an association succeeds. It therefore stays at HOST_MLME_AUTH_DONE for the whole connected session.

When the station later roams to a BSS on a different channel, the next authentication finds auth_flag != 0, skips the ROC setup, and the auth frame is transmitted on the currently-associated channel instead of the target's channel. Authentication times out on the new AP and the device stays connected to the original AP.

Gate the ROC setup on HOST_MLME_AUTH_PENDING instead of on auth_flag being completely clear. This re-arms the remain-on-channel for every new authentication attempt, while still suppressing a redundant ROC during the multi-frame SAE exchange, where auth_flag stays PENDING between the commit and confirm frames.

This change was tested in 3 different devices:
Verdin AM62 (IW412 SD-UART) - (16.92.21.p142)
Verdin iMX8MM (W8997 SD-SD) - (16.68.1.p197)
Verdin iMX8MP (W8997 SD-UART) - (16.92.21.p137)

There following loop tests were performed:
1) force roaming between two AP's, one 5GHz and one 2.4GHz, same SSID. Use wpa_cli to trigger the roaming behavior, sleep 2s between iterations.
2) force a disconnection to AP 1 and a connection to AP 2, test scan. Use wpa_cli to trigger the connection changes, sleep 2s between iterations.

Each test ran in each device for at least 3 hours.

Fixes 36995892c271 ("wifi: mwifiex: add host mlme for client mode")
Assisted-by Claude:claude-opus-4-7
---
 drivers/net/wireless/marvell/mwifiex/cfg80211.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/wireless/marvell/mwifiex/cfg80211.c +1 -1
@@ -4326,7 +4326,7 @@ mwifiex_cfg80211_authenticate(struct wip
 		return -EOPNOTSUPP;
 	}
 
-	if (!priv->auth_flag) {
+	if (!(priv->auth_flag & HOST_MLME_AUTH_PENDING)) {
 		ret = mwifiex_remain_on_chan_cfg(priv, HostCmd_ACT_GEN_SET,
 						 req->bss->channel,
 						 AUTH_TX_DEFAULT_WAIT_TIME);
author maintainer
#1488 lore raw

PATCH 1482/1611 wifi: mac80211: fix memory leak in ieee80211_register_hw()

In reply to Greg Kroah-Hartman

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

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

From: Dawei Feng <[email protected]>

commit 95fc02722edde02946d0d475221f2b2054d3d8ba upstream.

If kmemdup() fails while copying supported band structures, the error path jumps to fail_rate. This skips rate_control_deinitialize() and leaks the initialized local->rate_ctrl.

Fix this by adding a fail_band label that shares the rate-control cleanup path before falling through to the remaining teardown.

The bug was first flagged by an experimental analysis tool we are developing for kernel memory-management bugs while analyzing v6.13-rc1. The tool is still under development and is not yet publicly available. Manual inspection confirms that the bug is still present in v7.1-rc7.

An x86_64 allyesconfig build showed no new warnings. As we do not have a suitable mac80211 device/driver combination to test with, no runtime testing was able to be performed.

Fixes 09b4a4faf9d0 ("mac80211: introduce capability flags for VHT EXT NSS support")
---
 net/mac80211/main.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
net/mac80211/main.c +2 -1
@@ -1575,7 +1575,7 @@ int ieee80211_register_hw(struct ieee802
 		sband = kmemdup(sband, sizeof(*sband), GFP_KERNEL);
 		if (!sband) {
 			result = -ENOMEM;
-			goto fail_rate;
+			goto fail_band;
 		}
 
 		wiphy_dbg(hw->wiphy, "copying sband (band %d) due to VHT EXT NSS BW flag\n",
@@ -1642,6 +1642,7 @@ int ieee80211_register_hw(struct ieee802
 #endif
 	wiphy_unregister(local->hw.wiphy);
  fail_wiphy_register:
+ fail_band:
 	rtnl_lock();
 	rate_control_deinitialize(local);
 	ieee80211_remove_interfaces(local);
author maintainer
#1489 lore raw

PATCH 1483/1611 wifi: brcmfmac: cyw: fix heap overflow on a short auth frame

In reply to Greg Kroah-Hartman

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

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

From: Maoyi Xie <[email protected]>

commit 240c8d2c717b3f8153e7e877b22a82518d78dbdc upstream.

brcmf_notify_auth_frame_rx() takes the frame length from the firmware event and copies the frame body with the management header offset subtracted:

	u32 mgmt_frame_len = e->datalen - sizeof(struct brcmf_rx_mgmt_data);
	...
	memcpy(&mgmt_frame->u, frame,
	       mgmt_frame_len - offsetof(struct ieee80211_mgmt, u));

The only length check is e->datalen >= sizeof(*rxframe), so mgmt_frame_len can be anything from 0 up. offsetof(struct ieee80211_mgmt, u) is 24. When mgmt_frame_len is below that, the subtraction wraps as an unsigned value to a huge length. The memcpy then runs far past the kzalloc'd buffer. A malicious or malfunctioning AP can make the frame short during the external SAE auth exchange, so this is a remotely triggered heap overflow.

Reject frames shorter than the management header offset before the copy.

---
 drivers/net/wireless/broadcom/brcm80211/brcmfmac/cyw/core.c |    6 ++++++
 1 file changed, 6 insertions(+)
drivers/net/wireless/broadcom/brcm80211/brcmfmac/cyw/core.c +6 -0
@@ -294,6 +294,12 @@ brcmf_notify_auth_frame_rx(struct brcmf_
 		return -EINVAL;
 	}
 
+	if (mgmt_frame_len < offsetof(struct ieee80211_mgmt, u)) {
+		bphy_err(drvr, "Event %s (%d) frame too small. Ignore\n",
+			 brcmf_fweh_event_name(e->event_code), e->event_code);
+		return -EINVAL;
+	}
+
 	wdev = &ifp->vif->wdev;
 	WARN_ON(!wdev);
 
author maintainer
#1490 lore raw

PATCH 1484/1611 riscv: vdso: Do not use LTO for the vDSO

In reply to Greg Kroah-Hartman

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

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

From: Thomas Weißschuh <[email protected]>

commit ad6dcfa023762e37962f77ee48e752b7570e9440 upstream.

With LTO enabled the compiler assumes that the vDSO functions are not used and optimizes them away completely. Currently this happens to __vdso_clock_getres(), __vdso_clock_gettime(), __vdso_getrandom(), __vdso_gettimeofday() and __vdso_riscv_hwprobe().

Disable LTO for the vDSO, as these functions are hand-optimized anyways.

---
 arch/riscv/kernel/vdso/Makefile |    6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
arch/riscv/kernel/vdso/Makefile +3 -3
@@ -49,9 +49,9 @@ CPPFLAGS_vdso.lds += -DHAS_VGETTIMEOFDAY
 endif
 
 # Disable -pg to prevent insert call site
-CFLAGS_REMOVE_vgettimeofday.o = $(CC_FLAGS_FTRACE) $(CC_FLAGS_SCS)
-CFLAGS_REMOVE_getrandom.o = $(CC_FLAGS_FTRACE) $(CC_FLAGS_SCS)
-CFLAGS_REMOVE_hwprobe.o = $(CC_FLAGS_FTRACE) $(CC_FLAGS_SCS)
+CFLAGS_REMOVE_vgettimeofday.o = $(CC_FLAGS_FTRACE) $(CC_FLAGS_SCS) $(CC_FLAGS_LTO)
+CFLAGS_REMOVE_getrandom.o = $(CC_FLAGS_FTRACE) $(CC_FLAGS_SCS) $(CC_FLAGS_LTO)
+CFLAGS_REMOVE_hwprobe.o = $(CC_FLAGS_FTRACE) $(CC_FLAGS_SCS) $(CC_FLAGS_LTO)
 
 # Force dependency
 $(obj)/vdso.o: $(obj)/vdso.so
author maintainer
#1491 lore raw

PATCH 1485/1611 regulator: ltc3676: Fix incorrect IRQSTAT bit offsets

In reply to Greg Kroah-Hartman

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

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

From: Abhishek Ojha <[email protected]>

commit 50dce2e2f84b56d8b4b406d97a1543709e8a87f5 upstream.

The LTC3676_IRQSTAT_* bit definitions do not match the IRQSTAT (Interrupt Request Status) register layout documented in Table 15 of the LTC3676/LTC3676-1 datasheet:

bit 0 - Pushbutton Status Active
bit 1 - Hard Reset Occurred
bit 2 - PGOOD Timeout Occurred
bit 3 - Undervoltage Warning
bit 4 - Undervoltage Standby (Fault) Occurred bit 5 - Overtemperature Warning
bit 6 - Overtemperature Standby (Fault) Occurred bit 7 - Reserved

The driver instead defines these starting at bit 3, one bit higher than the datasheet specifies, which causes ltc3676_regulator_isr() to check the wrong status bits and misreport (or miss) PGOOD timeout, undervoltage and thermal warning/fault conditions.

Fix the bit offsets to match the datasheet.

---
 drivers/regulator/ltc3676.c |   10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)
drivers/regulator/ltc3676.c +5 -5
@@ -45,11 +45,11 @@
 #define LTC3676_DVBxA_REF_SELECT	BIT(5)
 #define LTC3676_DVBxB_PGOOD_MASK	BIT(5)
 
-#define LTC3676_IRQSTAT_PGOOD_TIMEOUT	BIT(3)
-#define LTC3676_IRQSTAT_UNDERVOLT_WARN	BIT(4)
-#define LTC3676_IRQSTAT_UNDERVOLT_FAULT	BIT(5)
-#define LTC3676_IRQSTAT_THERMAL_WARN	BIT(6)
-#define LTC3676_IRQSTAT_THERMAL_FAULT	BIT(7)
+#define LTC3676_IRQSTAT_PGOOD_TIMEOUT	BIT(2)
+#define LTC3676_IRQSTAT_UNDERVOLT_WARN	BIT(3)
+#define LTC3676_IRQSTAT_UNDERVOLT_FAULT	BIT(4)
+#define LTC3676_IRQSTAT_THERMAL_WARN	BIT(5)
+#define LTC3676_IRQSTAT_THERMAL_FAULT	BIT(6)
 
 enum ltc3676_reg {
 	LTC3676_SW1,
author maintainer
#1492 lore raw

PATCH 1486/1611 net: openvswitch: reject oversized nested action attrs

In reply to Greg Kroah-Hartman

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

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

From: Asim Viladi Oglu Manizada <[email protected]>

commit 3f1f755366687d051174739fb99f7d560202f60b upstream.

Open vSwitch stores generated flow actions as nlattrs, whose nla_len field is u16. Commit a1e64addf3ff ("net: openvswitch: remove misbehaving actions length check") allowed the total sw_flow_actions stream to grow beyond 64 KiB, which is valid, but also removed the last guard preventing a generated nested action attribute from exceeding U16_MAX.

An oversized generated container can thus be closed with a truncated nla_len. A later dump or teardown then walks a structurally different stream than the one that was validated. In particular, an oversized nested CLONE/CT action may cause subsequent bytes in the generated stream to be interpreted as independent actions.

Keep the larger total-action-stream behavior, but make nested action close reject generated containers that do not fit in nla_len, and return the error through all callers. For recursive SAMPLE, CLONE, DEC_TTL, and CHECK_PKT_LEN builders, trim resource-owning action-list tails in reverse construction order before discarding failed wrappers, so resources copied into the rejected tails are released before the wrappers are removed.

Most failed outer wrappers are discarded by truncating actions_len after child resources have been released. CHECK_PKT_LEN also trims its parent after branch resources are gone. SET/TUNNEL close failures unwind their known tun_dst ownership directly, and SET_TO_MASKED has no external ownership and truncates on close failure.

Fixes a1e64addf3ff ("net: openvswitch: remove misbehaving actions length check")
Assisted-by avom-custom-harness:gpt-5.5-qwen3.6-mod-mix
---
 net/openvswitch/flow_netlink.c |  201 ++++++++++++++++++++++++++++++++---------
 1 file changed, 157 insertions(+), 44 deletions(-)
net/openvswitch/flow_netlink.c +157 -44
@@ -2496,13 +2496,56 @@ static inline int add_nested_action_star
 	return used;
 }
 
-static inline void add_nested_action_end(struct sw_flow_actions *sfa,
-					 int st_offset)
+static inline int add_nested_action_end(struct sw_flow_actions *sfa,
+					int st_offset)
 {
-	struct nlattr *a = (struct nlattr *) ((unsigned char *)sfa->actions +
-							       st_offset);
+	struct nlattr *a;
+	u32 attr_len;
+
+	if (WARN_ON_ONCE(st_offset < 0 ||
+			 (u32)st_offset > sfa->actions_len))
+		return -EINVAL;
+
+	attr_len = sfa->actions_len - (u32)st_offset;
+	if (WARN_ON_ONCE(attr_len < NLA_HDRLEN))
+		return -EINVAL;
+
+	if (attr_len > U16_MAX)
+		return -EMSGSIZE;
+
+	a = (struct nlattr *)((u8 *)sfa->actions + st_offset);
+	a->nla_len = attr_len;
+	return 0;
+}
+
+/* Free the generated action-list tail at @start and truncate it.
+ * If @nested, @start points to its containing nlattr header.
+ */
+static void ovs_nla_trim(struct sw_flow_actions *sfa, int start, bool nested)
+{
+	const struct nlattr *actions;
+	u32 len;
+
+	if (start < 0)
+		return;
+
+	if (WARN_ON_ONCE((u32)start > sfa->actions_len))
+		return;
 
-	a->nla_len = sfa->actions_len - st_offset;
+	actions = (const struct nlattr *)((u8 *)sfa->actions + start);
+	len = sfa->actions_len - (u32)start;
+
+	if (nested) {
+		if (len < NLA_HDRLEN)
+			goto out;
+
+		actions = (const struct nlattr *)((u8 *)actions + NLA_HDRLEN);
+		len -= NLA_HDRLEN;
+	}
+
+	ovs_nla_free_nested_actions(actions, len);
+out:
+	sfa->actions_len = start;
 }
 
 static int __ovs_nla_copy_actions(struct net *net, const struct nlattr *attr,
@@ -2522,6 +2565,7 @@ static int validate_and_copy_sample(stru
 	const struct nlattr *attrs[OVS_SAMPLE_ATTR_MAX + 1];
 	const struct nlattr *probability, *actions;
 	const struct nlattr *a;
+	int actions_start;
 	int rem, start, err;
 	struct sample_arg arg;
 
@@ -2565,18 +2609,27 @@ static int validate_and_copy_sample(stru
 	err = ovs_nla_add_action(sfa, OVS_SAMPLE_ATTR_ARG, &arg, sizeof(arg),
 				 log);
 	if (err)
-		return err;
+		goto err;
 
+	actions_start = (*sfa)->actions_len;
 	err = __ovs_nla_copy_actions(net, actions, key, sfa,
 				     eth_type, vlan_tci, mpls_label_count, log,
 				     depth + 1);
 
 	if (err)
-		return err;
+		goto err_free;
 
-	add_nested_action_end(*sfa, start);
+	err = add_nested_action_end(*sfa, start);
+	if (err)
+		goto err_free;
 
 	return 0;
+
+err_free:
+	ovs_nla_trim(*sfa, actions_start, false);
+err:
+	(*sfa)->actions_len = start;
+	return err;
 }
 
 static int validate_and_copy_dec_ttl(struct net *net,
@@ -2624,18 +2677,31 @@ static int validate_and_copy_dec_ttl(str
 		return start;
 
 	action_start = add_nested_action_start(sfa, OVS_DEC_TTL_ATTR_ACTION, log);
-	if (action_start < 0)
-		return action_start;
+	if (action_start < 0) {
+		err = action_start;
+		goto err;
+	}
 
 	err = __ovs_nla_copy_actions(net, actions, key, sfa, eth_type,
 				     vlan_tci, mpls_label_count, log,
 				     depth + 1);
 	if (err)
-		return err;
+		goto err_free;
 
-	add_nested_action_end(*sfa, action_start);
-	add_nested_action_end(*sfa, start);
+	err = add_nested_action_end(*sfa, action_start);
+	if (err)
+		goto err_free;
+
+	err = add_nested_action_end(*sfa, start);
+	if (err)
+		goto err_free;
 	return 0;
+
+err_free:
+	ovs_nla_trim(*sfa, action_start, true);
+err:
+	(*sfa)->actions_len = start;
+	return err;
 }
 
 static int validate_and_copy_clone(struct net *net,
@@ -2646,6 +2712,7 @@ static int validate_and_copy_clone(struc
 				   u32 mpls_label_count, bool log, bool last,
 				   u32 depth)
 {
+	int actions_start;
 	int start, err;
 	u32 exec;
 
@@ -2661,17 +2728,26 @@ static int validate_and_copy_clone(struc
 	err = ovs_nla_add_action(sfa, OVS_CLONE_ATTR_EXEC, &exec,
 				 sizeof(exec), log);
 	if (err)
-		return err;
+		goto err;
 
+	actions_start = (*sfa)->actions_len;
 	err = __ovs_nla_copy_actions(net, attr, key, sfa,
 				     eth_type, vlan_tci, mpls_label_count, log,
 				     depth + 1);
 	if (err)
-		return err;
+		goto err_free;
 
-	add_nested_action_end(*sfa, start);
+	err = add_nested_action_end(*sfa, start);
+	if (err)
+		goto err_free;
 
 	return 0;
+
+err_free:
+	ovs_nla_trim(*sfa, actions_start, false);
+err:
+	(*sfa)->actions_len = start;
+	return err;
 }
 
 void ovs_match_init(struct sw_flow_match *match,
@@ -2763,20 +2839,20 @@ static int validate_and_copy_set_tun(con
 	tun_dst = metadata_dst_alloc(key.tun_opts_len, METADATA_IP_TUNNEL,
 				     GFP_KERNEL);
 
-	if (!tun_dst)
-		return -ENOMEM;
+	if (!tun_dst) {
+		err = -ENOMEM;
+		goto err;
+	}
 
 	err = dst_cache_init(&tun_dst->u.tun_info.dst_cache, GFP_KERNEL);
-	if (err) {
-		dst_release((struct dst_entry *)tun_dst);
-		return err;
-	}
+	if (err)
+		goto err_free_tun_dst;
 
 	a = __add_action(sfa, OVS_KEY_ATTR_TUNNEL_INFO, NULL,
 			 sizeof(*ovs_tun), log);
 	if (IS_ERR(a)) {
-		dst_release((struct dst_entry *)tun_dst);
-		return PTR_ERR(a);
+		err = PTR_ERR(a);
+		goto err_free_tun_dst;
 	}
 
 	ovs_tun = nla_data(a);
@@ -2797,8 +2873,16 @@ static int validate_and_copy_set_tun(con
 	ip_tunnel_info_opts_set(tun_info,
 				TUN_METADATA_OPTS(&key, key.tun_opts_len),
 				key.tun_opts_len, dst_opt_type);
-	add_nested_action_end(*sfa, start);
+	err = add_nested_action_end(*sfa, start);
+	if (WARN_ON_ONCE(err))
+		goto err_free_tun_dst;
+
+	return 0;
 
+err_free_tun_dst:
+	dst_release((struct dst_entry *)tun_dst);
+err:
+	(*sfa)->actions_len = start;
 	return err;
 }
 
@@ -2971,7 +3055,7 @@ static int validate_set(const struct nla
 
 	/* Convert non-masked non-tunnel set actions to masked set actions. */
 	if (!masked && key_type != OVS_KEY_ATTR_TUNNEL) {
-		int start, len = key_len * 2;
+		int err, start, len = key_len * 2;
 		struct nlattr *at;
 
 		*skip_copy = true;
@@ -2983,8 +3067,11 @@ static int validate_set(const struct nla
 			return start;
 
 		at = __add_action(sfa, key_type, NULL, len, log);
-		if (IS_ERR(at))
-			return PTR_ERR(at);
+		if (IS_ERR(at)) {
+			err = PTR_ERR(at);
+			(*sfa)->actions_len = start;
+			return err;
+		}
 
 		memcpy(nla_data(at), nla_data(ovs_key), key_len); /* Key. */
 		memset(nla_data(at) + key_len, 0xff, key_len);    /* Mask. */
@@ -2994,7 +3081,11 @@ static int validate_set(const struct nla
 
 			mask->ipv6_label &= htonl(0x000FFFFF);
 		}
-		add_nested_action_end(*sfa, start);
+		err = add_nested_action_end(*sfa, start);
+		if (WARN_ON_ONCE(err)) {
+			(*sfa)->actions_len = start;
+			return err;
+		}
 	}
 
 	return 0;
@@ -3040,7 +3131,8 @@ static int validate_and_copy_check_pkt_l
 	const struct nlattr *acts_if_greater, *acts_if_lesser_eq;
 	struct nlattr *a[OVS_CHECK_PKT_LEN_ATTR_MAX + 1];
 	struct check_pkt_len_arg arg;
-	int nested_acts_start;
+	int greater_acts_start = -1;
+	int lesser_acts_start = -1;
 	int start, err;
 
 	err = nla_parse_deprecated_strict(a, OVS_CHECK_PKT_LEN_ATTR_MAX,
@@ -3075,37 +3167,58 @@ static int validate_and_copy_check_pkt_l
 	err = ovs_nla_add_action(sfa, OVS_CHECK_PKT_LEN_ATTR_ARG, &arg,
 				 sizeof(arg), log);
 	if (err)
-		return err;
+		goto err_free;
 
-	nested_acts_start = add_nested_action_start(sfa,
-		OVS_CHECK_PKT_LEN_ATTR_ACTIONS_IF_LESS_EQUAL, log);
-	if (nested_acts_start < 0)
-		return nested_acts_start;
+	lesser_acts_start =
+		add_nested_action_start(sfa,
+					OVS_CHECK_PKT_LEN_ATTR_ACTIONS_IF_LESS_EQUAL,
+					log);
+	if (lesser_acts_start < 0) {
+		err = lesser_acts_start;
+		goto err_free;
+	}
 
 	err = __ovs_nla_copy_actions(net, acts_if_lesser_eq, key, sfa,
 				     eth_type, vlan_tci, mpls_label_count, log,
 				     depth + 1);
 
 	if (err)
-		return err;
+		goto err_free;
 
-	add_nested_action_end(*sfa, nested_acts_start);
+	err = add_nested_action_end(*sfa, lesser_acts_start);
+	if (err)
+		goto err_free;
 
-	nested_acts_start = add_nested_action_start(sfa,
-		OVS_CHECK_PKT_LEN_ATTR_ACTIONS_IF_GREATER, log);
-	if (nested_acts_start < 0)
-		return nested_acts_start;
+	greater_acts_start =
+		add_nested_action_start(sfa,
+					OVS_CHECK_PKT_LEN_ATTR_ACTIONS_IF_GREATER,
+					log);
+	if (greater_acts_start < 0) {
+		err = greater_acts_start;
+		goto err_free;
+	}
 
 	err = __ovs_nla_copy_actions(net, acts_if_greater, key, sfa,
 				     eth_type, vlan_tci, mpls_label_count, log,
 				     depth + 1);
 
 	if (err)
-		return err;
+		goto err_free;
+
+	err = add_nested_action_end(*sfa, greater_acts_start);
+	if (err)
+		goto err_free;
 
-	add_nested_action_end(*sfa, nested_acts_start);
-	add_nested_action_end(*sfa, start);
+	err = add_nested_action_end(*sfa, start);
+	if (err)
+		goto err_free;
 	return 0;
+
+err_free:
+	ovs_nla_trim(*sfa, greater_acts_start, true);
+	ovs_nla_trim(*sfa, lesser_acts_start, true);
+	ovs_nla_trim(*sfa, start, false);
+	return err;
 }
 
 static int validate_psample(const struct nlattr *attr)
author maintainer
#1493 lore raw

PATCH 1487/1611 Bluetooth: btrtl: validate firmware patch bounds

In reply to Greg Kroah-Hartman

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

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

From: Laxman Acharya Padhya <[email protected]>

commit 609c5b04a28dc1b0f3af6a7bc93055135b2d2059 upstream.

rtlbt_parse_firmware() copies patch_length - 4 bytes before appending the firmware version. A malformed firmware patch shorter than the version field can make this subtraction underflow and turn the copy into an oversized read and write during Bluetooth setup.

The existing patch_offset + patch_length check can also wrap on 32-bit architectures. Validate the patch length and range without arithmetic overflow before allocating or copying the patch.

Fixes db33c77dddc2 ("Bluetooth: btrtl: Create separate module for Realtek BT driver")
---
 drivers/bluetooth/btrtl.c |    5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)
drivers/bluetooth/btrtl.c +3 -2
@@ -786,8 +786,9 @@ static int rtlbt_parse_firmware(struct h
 	}
 
 	BT_DBG("length=%x offset=%x index %d", patch_length, patch_offset, i);
-	min_size = patch_offset + patch_length;
-	if (btrtl_dev->fw_len < min_size)
+	if (patch_length < sizeof(epatch_info->fw_version) ||
+	    patch_offset > btrtl_dev->fw_len ||
+	    patch_length > btrtl_dev->fw_len - patch_offset)
 		return -EINVAL;
 
 	/* Copy the firmware into a new buffer and write the version at
author maintainer
#1494 lore raw

PATCH 1488/1611 llc: fix SAP refcount leak when creating incoming sockets

In reply to Greg Kroah-Hartman

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

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

From: Xuanqiang Luo <[email protected]>

commit 2c72eb6286347d05a885412fb076993bd5286b53 upstream.

llc_sap_add_socket() takes a SAP reference for each socket added to a SAP, and llc_sap_remove_socket() releases it. llc_create_incoming_sock() takes an additional SAP reference after adding the child socket.

This extra reference was balanced by an explicit llc_sap_put() in llc_ui_release() until commit 3100aa9d74db ("llc: fix SAP reference counting w.r.t. socket handling") removed that put. The corresponding hold in the accept path was left behind.

When such a child socket is removed, only the reference taken by llc_sap_add_socket() is released. The extra reference keeps the SAP alive after its last socket is removed. Remove the obsolete hold.

Fixes 3100aa9d74db ("llc: fix SAP reference counting w.r.t. socket handling")
---
 net/llc/llc_conn.c |    1 -
 1 file changed, 1 deletion(-)
net/llc/llc_conn.c +0 -1
@@ -773,7 +773,6 @@ static struct sock *llc_create_incoming_
 	newllc->dev = dev;
 	dev_hold(dev);
 	llc_sap_add_socket(llc->sap, newsk);
-	llc_sap_hold(llc->sap);
 out:
 	return newsk;
 }
author maintainer
#1495 lore raw

PATCH 1489/1611 macsec: fix promiscuity refcount leak in macsec_dev_open()

In reply to Greg Kroah-Hartman

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

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

From: James Raphael Tiovalen <[email protected]>

commit 7410d11460eb90d6c9281162ccc6a128534d897d upstream.

When a MACsec interface with IFF_PROMISC set is brought up on top of a device that has hardware offload enabled, macsec_dev_open() first calls dev_set_promiscuity(real_dev, 1) and then propagates the open to the offload device. If that propagation fails, the error path jumps to the clear_allmulti label, which only reverts allmulti and the unicast address. The promiscuity taken on the lower device is never dropped, so real_dev is left permanently stuck in promiscuous mode. Its promiscuity count can no longer be balanced from software.

Add a clear_promisc label that drops the promiscuity reference and route the two offload failure paths to it. The dev_set_promiscuity() failure itself still jumps to clear_allmulti, since on that failure the count was not incremented.

---
 drivers/net/macsec.c |    7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)
drivers/net/macsec.c +5 -2
@@ -3609,19 +3609,22 @@ static int macsec_dev_open(struct net_de
 		ops = macsec_get_ops(netdev_priv(dev), &ctx);
 		if (!ops) {
 			err = -EOPNOTSUPP;
-			goto clear_allmulti;
+			goto clear_promisc;
 		}
 
 		ctx.secy = &macsec->secy;
 		err = macsec_offload(ops->mdo_dev_open, &ctx);
 		if (err)
-			goto clear_allmulti;
+			goto clear_promisc;
 	}
 
 	if (netif_carrier_ok(real_dev))
 		netif_carrier_on(dev);
 
 	return 0;
+clear_promisc:
+	if (dev->flags & IFF_PROMISC)
+		dev_set_promiscuity(real_dev, -1);
 clear_allmulti:
 	if (dev->flags & IFF_ALLMULTI)
 		dev_set_allmulti(real_dev, -1);
author maintainer
#1496 lore raw

PATCH 1490/1611 memstick: ms_block: reject a card that reports too many blocks

In reply to Greg Kroah-Hartman

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

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

From: Maoyi Xie <[email protected]>

commit 718178f524b98bc920d74bc771aed823c8b81425 upstream.

msb_ftl_initialize() computes the zone count from the card block count with no bound:

	msb->zone_count = msb->block_count / MS_BLOCKS_IN_ZONE;
	...
	for (i = 0; i < msb->zone_count; i++)
		msb->free_block_count[i] = MS_BLOCKS_IN_ZONE;

msb->block_count is a card value. msb_read_boot_blocks() reads number_of_blocks from the card boot page and byte swaps it. free_block_count is a fixed int[MS_MAX_ZONES]. MS_MAX_ZONES is 16, so the valid indices are 0 to 15. The init loop above indexes it by zone_count. msb_mark_block_used() and msb_mark_block_unused() index it by pba / MS_BLOCKS_IN_ZONE, for pba up to block_count - 1. A card may report up to 65535 blocks. A block_count above 8192 (MS_MAX_ZONES * MS_BLOCKS_IN_ZONE) lets the pba index reach 16. That writes past free_block_count[] and corrupts struct msb_data. A larger count runs the init loop past the end too.

A real Memory Stick has at most 16 zones. So it has at most 8192 blocks. msb_ftl_initialize() now rejects a card that reports more than MS_MAX_ZONES * MS_BLOCKS_IN_ZONE blocks.

Fixes 0ab30494bc4f ("memstick: add support for legacy memorysticks")
---
 drivers/memstick/core/ms_block.c |    4 ++++
 1 file changed, 4 insertions(+)
drivers/memstick/core/ms_block.c +4 -0
@@ -1339,6 +1339,10 @@ static int msb_ftl_initialize(struct msb
 		return 0;
 
 	msb->zone_count = msb->block_count / MS_BLOCKS_IN_ZONE;
+	if (msb->block_count > MS_MAX_ZONES * MS_BLOCKS_IN_ZONE) {
+		pr_err("Too many blocks: %d\n", msb->block_count);
+		return -EINVAL;
+	}
 	msb->logical_block_count = msb->zone_count * 496 - 2;
 
 	msb->used_blocks_bitmap = bitmap_zalloc(msb->block_count, GFP_KERNEL);
author maintainer
#1497 lore raw

PATCH 1491/1611 ipvs: fix more places with wrong ipv6 transport offsets

In reply to Greg Kroah-Hartman

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

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

From: Julian Anastasov <[email protected]>

commit b3fe4cbd583895987935a9bdad01c8f9d3a02310 upstream.

Sashiko reports for more incorrect IPv6 transport offsets.

The app code for TCP was assuming IPv4 network header even after the ipvsh argument was provided. This can cause problems with apps over IPv6. As for the only official app in the kernel tree (FTP) this problem is harmless because we use Netfilter to mangle the FTP ports and we do not adjust the TCP seq numbers.

Also, provide correct offset of the ICMPV6 header in ip_vs_out_icmp_v6() for correct checksum checks when the IPv6 packet has extension headers.

Fixes d12e12299a69 ("ipvs: add ipv6 support to ftp")
Fixes 2a3b791e6e11 ("IPVS: Add/adjust Netfilter hook functions and helpers for v6")
---
 net/netfilter/ipvs/ip_vs_app.c  |   10 ++++------
 net/netfilter/ipvs/ip_vs_core.c |    3 +--
 2 files changed, 5 insertions(+), 8 deletions(-)
net/netfilter/ipvs/ip_vs_app.c +4 -6
@@ -362,14 +362,13 @@ static inline int app_tcp_pkt_out(struct
 				  struct ip_vs_iphdr *ipvsh)
 {
 	int diff;
-	const unsigned int tcp_offset = ip_hdrlen(skb);
 	struct tcphdr *th;
 	__u32 seq;
 
-	if (skb_ensure_writable(skb, tcp_offset + sizeof(*th)))
+	if (skb_ensure_writable(skb, ipvsh->len + sizeof(*th)))
 		return 0;
 
-	th = (struct tcphdr *)(skb_network_header(skb) + tcp_offset);
+	th = (struct tcphdr *)(skb_network_header(skb) + ipvsh->len);
 
 	/*
 	 *	Remember seq number in case this pkt gets resized
@@ -439,14 +438,13 @@ static inline int app_tcp_pkt_in(struct
 				 struct ip_vs_iphdr *ipvsh)
 {
 	int diff;
-	const unsigned int tcp_offset = ip_hdrlen(skb);
 	struct tcphdr *th;
 	__u32 seq;
 
-	if (skb_ensure_writable(skb, tcp_offset + sizeof(*th)))
+	if (skb_ensure_writable(skb, ipvsh->len + sizeof(*th)))
 		return 0;
 
-	th = (struct tcphdr *)(skb_network_header(skb) + tcp_offset);
+	th = (struct tcphdr *)(skb_network_header(skb) + ipvsh->len);
 
 	/*
 	 *	Remember seq number in case this pkt gets resized
net/netfilter/ipvs/ip_vs_core.c +1 -2
@@ -1041,8 +1041,7 @@ static int ip_vs_out_icmp_v6(struct netn
 	snet.in6 = ciph.saddr.in6;
 	offset = ciph.len;
 	return handle_response_icmp(AF_INET6, skb, &snet, ciph.protocol, cp,
-				    pp, offset, sizeof(struct ipv6hdr),
-				    hooknum);
+				    pp, offset, ipvsh->len, hooknum);
 }
 #endif
 
author maintainer
#1498 lore raw

PATCH 1492/1611 ipvs: reload ip header after head reallocation

In reply to Greg Kroah-Hartman

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

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

From: Florian Westphal <[email protected]>

commit a2f57827bf7c695b8c72dc4511cae8e86582369d upstream.

__ip_vs_get_out_rt() calls skb_ensure_writable() which may reallocate skb->head.

Fixes 8d8e20e2d7bb ("ipvs: Decrement ttl")
Assisted-by Claude:claude-sonnet-4-6
---
 net/netfilter/ipvs/ip_vs_xmit.c |    6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)
net/netfilter/ipvs/ip_vs_xmit.c +2 -4
@@ -733,13 +733,11 @@ int
 ip_vs_bypass_xmit(struct sk_buff *skb, struct ip_vs_conn *cp,
 		  struct ip_vs_protocol *pp, struct ip_vs_iphdr *ipvsh)
 {
-	struct iphdr  *iph = ip_hdr(skb);
-
-	if (__ip_vs_get_out_rt(cp->ipvs, cp->af, skb, NULL, iph->daddr,
+	if (__ip_vs_get_out_rt(cp->ipvs, cp->af, skb, NULL, ip_hdr(skb)->daddr,
 			       IP_VS_RT_MODE_NON_LOCAL, NULL, ipvsh) < 0)
 		goto tx_error;
 
-	ip_send_check(iph);
+	ip_send_check(ip_hdr(skb));
 
 	/* Another hack: avoid icmp_send in ip_fragment */
 	skb->ignore_df = 1;
author maintainer
#1499 lore raw

PATCH 1493/1611 reset: imx7: Correct polarity of MIPI CSI resets on i.MX8MQ

In reply to Greg Kroah-Hartman

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

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

From: Robby Cai <[email protected]>

commit 71827776667f4e4677a4fa806bcfb24d4b8dd9d7 upstream.

On i.MX8MQ, the MIPI CSI reset lines are active-low and not self-clearing. Writing '0' asserts reset and it remains asserted until explicitly deasserted by software.

This driver previously treated the MIPI CSI reset signals as active-high, which led to incorrect reset assert/deassert sequencing. This issue was exposed by commit 6d79bb8fd2aa ("media: imx8mq-mipi-csi2: Explicitly release reset").

Fix this by reflecting the correct reset polarity and ensuring proper reset handling.

Fixes c979dbf59987 ("reset: imx7: Add support for i.MX8MQ IP block variant")
Cc [email protected] # 6d79bb8fd2aa: media: imx8mq-mipi-csi2: Explicitly release reset
---
 drivers/reset/reset-imx7.c |    6 ++++++
 1 file changed, 6 insertions(+)
drivers/reset/reset-imx7.c +6 -0
@@ -236,6 +236,12 @@ static int imx8mq_reset_set(struct reset
 
 	case IMX8MQ_RESET_PCIE_CTRL_APPS_EN:
 	case IMX8MQ_RESET_PCIE2_CTRL_APPS_EN:
+	case IMX8MQ_RESET_MIPI_CSI1_CORE_RESET:
+	case IMX8MQ_RESET_MIPI_CSI1_PHY_REF_RESET:
+	case IMX8MQ_RESET_MIPI_CSI1_ESC_RESET:
+	case IMX8MQ_RESET_MIPI_CSI2_CORE_RESET:
+	case IMX8MQ_RESET_MIPI_CSI2_PHY_REF_RESET:
+	case IMX8MQ_RESET_MIPI_CSI2_ESC_RESET:
 	case IMX8MQ_RESET_MIPI_DSI_PCLK_RESET_N:
 	case IMX8MQ_RESET_MIPI_DSI_ESC_RESET_N:
 	case IMX8MQ_RESET_MIPI_DSI_DPI_RESET_N:
author maintainer
#1500 lore raw

PATCH 1494/1611 reset: sunxi: fix memory region leak on ioremap failure

In reply to Greg Kroah-Hartman

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

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

From: Zhao Dongdong <[email protected]>

commit 1a8c89f8c112c75e84ff9a140f969e372aed0c9a upstream.

In sunxi_reset_init(), when ioremap() fails, the memory region obtained via request_mem_region() is not released, leading to a resource leak.

Add an err_mem_region label to properly release the memory region before freeing the data structure.

Fixes 8f1ae77f4666 ("reset: Add Allwinner SoCs Reset Controller Driver")
---
 drivers/reset/reset-sunxi.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/reset/reset-sunxi.c +3 -1
@@ -44,7 +44,7 @@ static int sunxi_reset_init(struct devic
 	data->membase = ioremap(res.start, size);
 	if (!data->membase) {
 		ret = -ENOMEM;
-		goto err_alloc;
+		goto err_mem_region;
 	}
 
 	spin_lock_init(&data->lock);
@@ -57,6 +57,8 @@ static int sunxi_reset_init(struct devic
 
 	return reset_controller_register(&data->rcdev);
 
+err_mem_region:
+	release_mem_region(res.start, size);
 err_alloc:
 	kfree(data);
 	return ret;
author maintainer
#1501 lore raw

PATCH 1495/1611 powerpc/spufs: fix out-of-bounds access in spufs_mem_mmap_access()

In reply to Greg Kroah-Hartman

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

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

From: Junrui Luo <[email protected]>

commit 47b87f469a35b5ffc81c16eee6b13a9b6c8d55c6 upstream.

spufs_mem_mmap_access() computes the local store offset as address - vma->vm_start, but bounds-checks it against vma->vm_end instead of the local store size. On 64-bit, offset is always well below vma->vm_end, so the clamp never fires and len stays unbounded against the LS_SIZE buffer returned by ctx->ops->get_ls().

Reject offsets at or beyond LS_SIZE and clamp len to the remaining space, mirroring the guard already used by spufs_mem_mmap_fault() and spufs_ps_fault().

---
 arch/powerpc/platforms/cell/spufs/file.c |    6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
arch/powerpc/platforms/cell/spufs/file.c +4 -2
@@ -268,10 +268,12 @@ static int spufs_mem_mmap_access(struct
 
 	if (write && !(vma->vm_flags & VM_WRITE))
 		return -EACCES;
+	if (offset >= LS_SIZE)
+		return -EFAULT;
 	if (spu_acquire(ctx))
 		return -EINTR;
-	if ((offset + len) > vma->vm_end)
-		len = vma->vm_end - offset;
+	if ((offset + len) > LS_SIZE)
+		len = LS_SIZE - offset;
 	local_store = ctx->ops->get_ls(ctx);
 	if (write)
 		memcpy_toio(local_store + offset, buf, len);
author maintainer
#1502 lore raw

PATCH 1496/1611 wifi: cfg80211: validate EHT MLE before MLD ID read

In reply to Greg Kroah-Hartman

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

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

From: Haofeng Li <[email protected]>

commit 74e27cd1d98b546fdb276008a83708d062339661 upstream.

cfg80211_gen_new_ie() copies ML probe response elements from the parent frame when the parent EHT multi-link element has an MLD ID matching the nontransmitted BSSID index.

The code only checked that the extension element had more than one byte before calling ieee80211_mle_get_mld_id(). That helper assumes a BASIC MLE with enough common info and documents that callers must first use ieee80211_mle_type_ok().

Attack chain:
malicious AP sends a short EHT MLE in an MBSSID beacon. cfg80211_inform_bss_frame_data() stores the copied IE buffer. cfg80211_parse_mbssid_data() builds the nontransmitted BSS IE. cfg80211_gen_new_ie() sees the EHT MLE in the parent frame. ieee80211_mle_get_mld_id() then reads past the IE boundary.

Validate the MLE type and size before reading the MLD ID. This matches the contract required by the MLE helper and rejects the short element before any internal MLE fields are accessed.

Fixes 61dcfa8c2a8f ("wifi: cfg80211: copy multi-link element from the multi-link probe request's frame body to the generated elements")
---
 net/wireless/scan.c |    5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
net/wireless/scan.c +4 -1
@@ -326,8 +326,11 @@ cfg80211_gen_new_ie(const u8 *ie, size_t
 		/* For ML probe response, match the MLE in the frame body with
 		 * MLD id being 'bssid_index'
 		 */
-		if (parent->id == WLAN_EID_EXTENSION && parent->datalen > 1 &&
+		if (parent->id == WLAN_EID_EXTENSION &&
 		    parent->data[0] == WLAN_EID_EXT_EHT_MULTI_LINK &&
+		    ieee80211_mle_type_ok(parent->data + 1,
+					  IEEE80211_ML_CONTROL_TYPE_BASIC,
+					  parent->datalen - 1) &&
 		    bssid_index == ieee80211_mle_get_mld_id(parent->data + 1)) {
 			if (!cfg80211_copy_elem_with_frags(parent,
 							   ie, ielen,
author maintainer
#1503 lore raw

PATCH 1497/1611 wifi: ieee80211: validate MLE common info length

In reply to Greg Kroah-Hartman

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

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

From: Zhao Li <[email protected]>

commit 293baeae9b2434a3e432629d7720b5603db2d77e upstream.

ieee80211_mle_common_size() uses the first common-info octet as the common information length for all known MLE types. However, ieee80211_mle_size_ok() only validates that octet for Basic, Probe Request, and TDLS MLEs.

Reconfiguration MLEs also skipped the length octet when calculating the minimum common size, and Priority Access MLEs skipped validation of the advertised common information length.

Account for the Reconfiguration common-info length octet and validate the advertised common information length for all known MLE types. Keep unknown-type handling unchanged.

Fixes 0f48b8b88aa9 ("wifi: ieee80211: add definitions for multi-link element")
[remove now misleading comment]
---
 include/linux/ieee80211-eht.h |   12 ++++--------
 1 file changed, 4 insertions(+), 8 deletions(-)
include/linux/ieee80211-eht.h +4 -8
@@ -833,7 +833,7 @@ static inline bool ieee80211_mle_size_ok
 	const struct ieee80211_multi_link_elem *mle = (const void *)data;
 	u8 fixed = sizeof(*mle);
 	u8 common = 0;
-	bool check_common_len = false;
+	u8 common_len;
 	u16 control;
 
 	if (!data || len < fixed)
@@ -844,7 +844,6 @@ static inline bool ieee80211_mle_size_ok
 	switch (u16_get_bits(control, IEEE80211_ML_CONTROL_TYPE)) {
 	case IEEE80211_ML_CONTROL_TYPE_BASIC:
 		common += sizeof(struct ieee80211_mle_basic_common_info);
-		check_common_len = true;
 		if (control & IEEE80211_MLC_BASIC_PRES_LINK_ID)
 			common += 1;
 		if (control & IEEE80211_MLC_BASIC_PRES_BSS_PARAM_CH_CNT)
@@ -864,9 +863,9 @@ static inline bool ieee80211_mle_size_ok
 		common += sizeof(struct ieee80211_mle_preq_common_info);
 		if (control & IEEE80211_MLC_PREQ_PRES_MLD_ID)
 			common += 1;
-		check_common_len = true;
 		break;
 	case IEEE80211_ML_CONTROL_TYPE_RECONF:
+		common += 1;
 		if (control & IEEE80211_MLC_RECONF_PRES_MLD_MAC_ADDR)
 			common += ETH_ALEN;
 		if (control & IEEE80211_MLC_RECONF_PRES_EML_CAPA)
@@ -878,7 +877,6 @@ static inline bool ieee80211_mle_size_ok
 		break;
 	case IEEE80211_ML_CONTROL_TYPE_TDLS:
 		common += sizeof(struct ieee80211_mle_tdls_common_info);
-		check_common_len = true;
 		break;
 	case IEEE80211_ML_CONTROL_TYPE_PRIO_ACCESS:
 		common = ETH_ALEN + 1;
@@ -891,11 +889,9 @@ static inline bool ieee80211_mle_size_ok
 	if (len < fixed + common)
 		return false;
 
-	if (!check_common_len)
-		return true;
+	common_len = mle->variable[0];
 
-	/* if present, common length is the first octet there */
-	return mle->variable[0] >= common;
+	return common_len >= common && common_len <= len - fixed;
 }
 
 /**
author maintainer
#1504 lore raw

PATCH 1498/1611 wifi: mac80211: free ack status frame on TX header build failure

In reply to Greg Kroah-Hartman

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

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

From: Zhiling Zou <[email protected]>

commit 2c51457d930f723e5f2903af90f5847f7df53f42 upstream.

ieee80211_build_hdr() stores an ACK status frame before it has finished all validation and header construction. If a later error path is taken, the transmit skb is freed but the stored ACK status frame remains in local->ack_status_frames.

This can happen for control port frames when the requested MLO link ID does not match the link selected for a non-MLO station. Repeated failures can fill the ACK status IDR and leave pending ACK frames until hardware teardown.

Remove any stored ACK status frame before returning an error after it has been inserted into the IDR.

---
 net/mac80211/tx.c |   17 ++++++++++++++++-
 1 file changed, 16 insertions(+), 1 deletion(-)
net/mac80211/tx.c +16 -1
@@ -2578,6 +2578,18 @@ static u16 ieee80211_store_ack_skb(struc
 	return info_id;
 }
 
+static void ieee80211_remove_ack_skb(struct ieee80211_local *local, u16 info_id)
+{
+	struct sk_buff *ack_skb;
+	unsigned long flags;
+
+	spin_lock_irqsave(&local->ack_status_lock, flags);
+	ack_skb = idr_remove(&local->ack_status_frames, info_id);
+	spin_unlock_irqrestore(&local->ack_status_lock, flags);
+
+	kfree_skb(ack_skb);
+}
+
 /**
  * ieee80211_build_hdr - build 802.11 header in the given frame
  * @sdata: virtual interface to build the header for
@@ -2934,7 +2946,8 @@ static struct sk_buff *ieee80211_build_h
 		if (ieee80211_skb_resize(sdata, skb, head_need, ENCRYPT_DATA)) {
 			ieee80211_free_txskb(&local->hw, skb);
 			skb = NULL;
-			return ERR_PTR(-ENOMEM);
+			ret = -ENOMEM;
+			goto free;
 		}
 	}
 
@@ -3002,6 +3015,8 @@ static struct sk_buff *ieee80211_build_h
 
 	return skb;
  free:
+	if (info_id)
+		ieee80211_remove_ack_skb(local, info_id);
 	kfree_skb(skb);
 	return ERR_PTR(ret);
 }
author maintainer
#1505 lore raw

PATCH 1499/1611 wifi: mwifiex: fix permanently busy scans after multiple roam iterations

In reply to Greg Kroah-Hartman

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

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

From: Rafael Beims <[email protected]>

commit d78a407bad6f500884a8606aea1a5a9207be4030 upstream.

In order for the firmware to sleep, the driver has to confirm a previously received sleep request. The normal sequence of evets goes like this:
EVENT_SLEEP -> adapter->ps_state = PS_STATE_PRE_SLEEP -> sleep-confirm -> SLEEP -> EVENT_AWAKE -> AWAKE.
Before sending the sleep-confirm command, the driver must make sure there are no commands either running or waiting to be completed.

mwifiex_ret_802_11_associate() unconditionally sets ps_state = PS_STATE_AWAKE when it processes the association command response, outside of the normal powersave management flow. If EVENT_SLEEP arrives while the association command is in flight, ps_state is PRE_SLEEP when the association command response is parsed, and the forced AWAKE overwrites it. The deferred sleep-confirm is never sent.

A subsequent scan_start command is correctly acknowledged, but the firmware doesn't generate scan_result events. The scan request never finishes, and additional requests from userspace fail with -EBUSY.

After testing on both IW412 and W8997, I could only trigger the bug on the IW412 and observed the firmwares behave differently. On the IW412 the firmware still sends EVENT_SLEEP while the authentication / association process is ongoing. A W8997 under the same conditions seems to suppress power-save for the duration of the association, so PRE_SLEEP never coincided with the association response even after extended periods of testing using the loops described below (>12hours).

On the IW412, the delay between commands that triggers an EVENT_SLEEP was empirically determined to be ~20ms. This delay can naturally occur when the driver is outputting debugging information (debug_mask = 0x00000037), in which situation the busy scans issue is repeatable while running "test 1)" as described below. If the delay between commands is less than ~20ms, the firmware stays awake and the issue was not reproducible running the same test.

The host_mlme=false path also behaves differently. In this case, the entire authentication / association transaction is executed by one command (HostCmd_CMD_802_11_ASSOCIATE), and the firmware doesn't emit EVENT_SLEEP while the command is running.

Remove the assignment so the ps_state is only manipulated in the paths that are related to powersave event handling and on the main workqueue for correct sleep confirmation.

The following loop tests were performed (with debugging output enabled):
1) force roaming between two AP's, one 5GHz and one 2.4GHz, same SSID. Use wpa_cli to trigger the roaming behavior, sleep 2s between iterations.
2) force a disconnection to AP 1 and a connection to AP 2, test scan. Use wpa_cli to trigger the connection changes, sleep 2s between iterations.

Each test ran in each device for at least 3 hours.

Fixes 5e6e3a92b9a4 ("wireless: mwifiex: initial commit for Marvell mwifiex driver")
Assisted-by Claude:claude-opus-4-7
---
 drivers/net/wireless/marvell/mwifiex/join.c |    1 -
 1 file changed, 1 deletion(-)
drivers/net/wireless/marvell/mwifiex/join.c +0 -1
@@ -736,7 +736,6 @@ int mwifiex_ret_802_11_associate(struct
 	/* Send a Media Connected event, according to the Spec */
 	priv->media_connected = true;
 
-	priv->adapter->ps_state = PS_STATE_AWAKE;
 	priv->adapter->pps_uapsd_mode = false;
 	priv->adapter->tx_lock_flag = false;
 
author maintainer
#1506 lore raw

PATCH 1500/1611 mtd: onenand: samsung: report DMA completion timeouts

In reply to Greg Kroah-Hartman

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

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

From: Pengpeng Hou <[email protected]>

commit d03a19bd6c7f86b99ca8fb61a6ec2345cee1d9d6 upstream.

The S5PC110 OneNAND DMA helpers have bounded waits for transfer completion. The polling helper falls out of its timeout loop and returns success, and the IRQ helper ignores wait_for_completion_timeout().

Return -ETIMEDOUT when the DMA transfer-done bit or completion does not arrive before the timeout so callers can treat the buffer transfer as failed.

Fixes e23abf4b7743 ("mtd: OneNAND: S5PC110: Implement DMA interrupt method")
---
 drivers/mtd/nand/onenand/onenand_samsung.c |    7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)
drivers/mtd/nand/onenand/onenand_samsung.c +6 -1
@@ -554,6 +554,9 @@ static int s5pc110_dma_poll(dma_addr_t d
 	} while (!(status & S5PC110_DMA_TRANS_STATUS_TD) &&
 		time_before(jiffies, timeout));
 
+	if (!(status & S5PC110_DMA_TRANS_STATUS_TD))
+		return -ETIMEDOUT;
+
 	writel(S5PC110_DMA_TRANS_CMD_TDC, base + S5PC110_DMA_TRANS_CMD);
 
 	return 0;
@@ -608,7 +611,9 @@ static int s5pc110_dma_irq(dma_addr_t ds
 
 	writel(S5PC110_DMA_TRANS_CMD_TR, base + S5PC110_DMA_TRANS_CMD);
 
-	wait_for_completion_timeout(&onenand->complete, msecs_to_jiffies(20));
+	if (!wait_for_completion_timeout(&onenand->complete,
+					 msecs_to_jiffies(20)))
+		return -ETIMEDOUT;
 
 	return 0;
 }
author maintainer
#1507 lore raw

PATCH 1501/1611 mtd: mchp23k256: use SPI match data for chip caps

In reply to Greg Kroah-Hartman

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

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

From: Pengpeng Hou <[email protected]>

commit d322e40f4edf92bf0ca329e5aa4ae1c0316feb38 upstream.

The driver stores chip capacity information in both the OF match table and the SPI id table. Probe currently uses of_device_get_match_data(), so a non-OF SPI modalias match falls back to mchp23k256_caps even when the SPI id table selected a different part.

Use spi_get_device_match_data() so SPI id-table driver_data is consumed when OF match data is absent. This keeps the existing default fallback while avoiding the wrong MTD geometry for id-table-only matches.

Fixes 4379075a870b ("mtd: mchp23k256: Add support for mchp23lcv1024")
---
 drivers/mtd/devices/mchp23k256.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/mtd/devices/mchp23k256.c +1 -1
@@ -188,7 +188,7 @@ static int mchp23k256_probe(struct spi_d
 
 	data = dev_get_platdata(&spi->dev);
 
-	flash->caps = of_device_get_match_data(&spi->dev);
+	flash->caps = spi_get_device_match_data(spi);
 	if (!flash->caps)
 		flash->caps = &mchp23k256_caps;
 
author maintainer
#1508 lore raw

PATCH 1502/1611 mmc: vub300: defer reset until cmd_mutex is unlocked

In reply to Greg Kroah-Hartman

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

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

From: Runyu Xiao <[email protected]>

commit ee5fb641c4ccac8406c668d3e947eb20ce44f233 upstream.

vub300_cmndwork_thread() holds cmd_mutex while it sends a command and waits for the command response. If the response wait times out, __vub300_command_response() kills the command URBs and then synchronously resets the USB device through usb_reset_device().

That reset path re-enters the driver through vub300_pre_reset(), which also takes cmd_mutex. The worker therefore tries to acquire the same mutex recursively while it is still holding it from the command path.

This issue was found by our static analysis tool and then manually reviewed against the current tree.

The grounded PoC kept the real worker and timeout/reset carrier:

vub300_cmndwork_thread()
__vub300_command_response()
usb_lock_device_for_reset()
usb_reset_device()
vub300_pre_reset()

Lockdep reported the same-task recursive acquisition on cmd_mutex:

WARNING: possible recursive locking detected
... (&test_vub300.cmd_mutex) ... at: usb_reset_device... [vuln_msv] ... (&test_vub300.cmd_mutex) ... at: vub300_cmndwork_thread+0x12/0x20 [vuln_msv]

  Workqueue: vub300_cmd_wq vub300_cmndwork_thread [vuln_msv]
  *** DEADLOCK ***

Return a flag from __vub300_command_response() when the timeout path needs a device reset, then perform the reset after vub300_cmndwork_thread() has cleared the in-flight command state and dropped cmd_mutex. The reset is still attempted before mmc_request_done(), preserving the existing request completion ordering while avoiding the recursive lock.

Fixes 88095e7b473a ("mmc: Add new VUB300 USB-to-SD/SDIO/MMC driver")
---
 drivers/mmc/host/vub300.c |   26 +++++++++++++++++---------
 1 file changed, 17 insertions(+), 9 deletions(-)
drivers/mmc/host/vub300.c +17 -9
@@ -1586,7 +1586,7 @@ static int __command_write_data(struct v
 	return linear_length;
 }
 
-static void __vub300_command_response(struct vub300_mmc_host *vub300,
+static bool __vub300_command_response(struct vub300_mmc_host *vub300,
 				      struct mmc_command *cmd,
 				      struct mmc_data *data, int data_length)
 {
@@ -1598,17 +1598,11 @@ static void __vub300_command_response(st
 					    msecs_to_jiffies(msec_timeout));
 	if (respretval == 0) { /* TIMED OUT */
 		/* we don't know which of "out" and "res" if any failed */
-		int result;
 		vub300->usb_timed_out = 1;
 		usb_kill_urb(vub300->command_out_urb);
 		usb_kill_urb(vub300->command_res_urb);
 		cmd->error = -ETIMEDOUT;
-		result = usb_lock_device_for_reset(vub300->udev,
-						   vub300->interface);
-		if (result == 0) {
-			result = usb_reset_device(vub300->udev);
-			usb_unlock_device(vub300->udev);
-		}
+		return true;
 	} else if (respretval < 0) {
 		/* we don't know which of "out" and "res" if any failed */
 		usb_kill_urb(vub300->command_out_urb);
@@ -1704,6 +1698,8 @@ static void __vub300_command_response(st
 	} else {
 		cmd->error = -EINVAL;
 	}
+
+	return false;
 }
 
 static void construct_request_response(struct vub300_mmc_host *vub300,
@@ -1749,6 +1745,7 @@ static void vub300_cmndwork_thread(struc
 		struct mmc_request *req = vub300->req;
 		struct mmc_command *cmd = vub300->cmd;
 		struct mmc_data *data = vub300->data;
+		bool reset_device;
 		int data_length;
 		mutex_lock(&vub300->cmd_mutex);
 		init_completion(&vub300->command_complete);
@@ -1771,7 +1768,8 @@ static void vub300_cmndwork_thread(struc
 			data_length = __command_read_data(vub300, cmd, data);
 		else
 			data_length = __command_write_data(vub300, cmd, data);
-		__vub300_command_response(vub300, cmd, data, data_length);
+		reset_device = __vub300_command_response(vub300, cmd,
+							 data, data_length);
 		vub300->req = NULL;
 		vub300->cmd = NULL;
 		vub300->data = NULL;
@@ -1779,6 +1777,16 @@ static void vub300_cmndwork_thread(struc
 			if (cmd->error == -ENOMEDIUM)
 				check_vub300_port_status(vub300);
 			mutex_unlock(&vub300->cmd_mutex);
+			if (reset_device) {
+				int result;
+
+				result = usb_lock_device_for_reset(vub300->udev,
+								   vub300->interface);
+				if (result == 0) {
+					result = usb_reset_device(vub300->udev);
+					usb_unlock_device(vub300->udev);
+				}
+			}
 			mmc_request_done(vub300->mmc, req);
 			kref_put(&vub300->kref, vub300_delete);
 			return;
author maintainer
#1509 lore raw

PATCH 1503/1611 mtd: rawnand: fsl_ifc: return errors for failed page reads

In reply to Greg Kroah-Hartman

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

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

From: Pengpeng Hou <[email protected]>

commit f9a13e05a327080c3a1c8165adf9e678fb68fef2 upstream.

fsl_ifc_run_command() logs controller timeout and other non-OPC completion states in ctrl->nand_stat. fsl_ifc_read_page() then only increments the ECC failure counter for non-OPC status and still returns max_bitflips, which can be zero.

Return -ETIMEDOUT when the command did not complete at all and -EIO for other non-OPC read completions so the NAND core does not treat a failed page read as a clean page.

Fixes 82771882d960 ("NAND Machine support for Integrated Flash Controller")
---
 drivers/mtd/nand/raw/fsl_ifc_nand.c |    9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)
drivers/mtd/nand/raw/fsl_ifc_nand.c +8 -1
@@ -683,8 +683,15 @@ static int fsl_ifc_read_page(struct nand
 		return check_erased_page(chip, buf);
 	}
 
-	if (ctrl->nand_stat != IFC_NAND_EVTER_STAT_OPC)
+	if (!ctrl->nand_stat) {
 		mtd->ecc_stats.failed++;
+		return -ETIMEDOUT;
+	}
+
+	if (ctrl->nand_stat != IFC_NAND_EVTER_STAT_OPC) {
+		mtd->ecc_stats.failed++;
+		return -EIO;
+	}
 
 	return nctrl->max_bitflips;
 }
author maintainer
#1510 lore raw

PATCH 1504/1611 mtd: rawnand: lpc32xx_mlc: fail DMA transfers on timeout

In reply to Greg Kroah-Hartman

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

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

From: Pengpeng Hou <[email protected]>

commit dbf590b662695b16fbf5917ef129697be4410ea9 upstream.

lpc32xx_xmit_dma() starts a DMA transfer and waits up to one second for its completion, but it ignores the wait result and returns success after unmapping the buffer.

A timed out read can therefore return success with incomplete data, and a timed out write can continue the NAND operation without proof that the DMA payload reached the controller.

Terminate the DMA channel on timeout, unmap the scatterlist through the existing cleanup path, and return -ETIMEDOUT to the NAND read/write callers. Initialize the shared cleanup-path result before using it for dmaengine_prep_slave_sg() failures.

Fixes 70f7cb78ec53 ("mtd: add LPC32xx MLC NAND driver")
---
 drivers/mtd/nand/raw/lpc32xx_mlc.c |   12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)
drivers/mtd/nand/raw/lpc32xx_mlc.c +10 -2
@@ -396,6 +396,7 @@ static int lpc32xx_xmit_dma(struct mtd_i
 	struct lpc32xx_nand_host *host = nand_get_controller_data(chip);
 	struct dma_async_tx_descriptor *desc;
 	int flags = DMA_CTRL_ACK | DMA_PREP_INTERRUPT;
+	unsigned long time_left;
 	int res;
 
 	sg_init_one(&host->sgl, mem, len);
@@ -410,6 +411,7 @@ static int lpc32xx_xmit_dma(struct mtd_i
 				       flags);
 	if (!desc) {
 		dev_err(mtd->dev.parent, "Failed to prepare slave sg\n");
+		res = -ENXIO;
 		goto out1;
 	}
 
@@ -420,7 +422,13 @@ static int lpc32xx_xmit_dma(struct mtd_i
 	dmaengine_submit(desc);
 	dma_async_issue_pending(host->dma_chan);
 
-	wait_for_completion_timeout(&host->comp_dma, msecs_to_jiffies(1000));
+	time_left = wait_for_completion_timeout(&host->comp_dma,
+						msecs_to_jiffies(1000));
+	if (!time_left) {
+		dmaengine_terminate_sync(host->dma_chan);
+		res = -ETIMEDOUT;
+		goto out1;
+	}
 
 	dma_unmap_sg(host->dma_chan->device->dev, &host->sgl, 1,
 		     DMA_BIDIRECTIONAL);
@@ -428,7 +436,7 @@ static int lpc32xx_xmit_dma(struct mtd_i
 out1:
 	dma_unmap_sg(host->dma_chan->device->dev, &host->sgl, 1,
 		     DMA_BIDIRECTIONAL);
-	return -ENXIO;
+	return res;
 }
 
 static int lpc32xx_read_page(struct nand_chip *chip, uint8_t *buf,
author maintainer
#1511 lore raw

PATCH 1505/1611 mtd: rawnand: lpc32xx_slc: fail DMA transfer on completion timeout

In reply to Greg Kroah-Hartman

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

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

From: Pengpeng Hou <[email protected]>

commit 17a8ce84964f243c8f89dc7353ac7e8d3137bc74 upstream.

lpc32xx_xmit_dma() waits for the DMA completion callback but ignores wait_for_completion_timeout(). A timed out DMA transfer is therefore unmapped and reported as successful to the NAND read/write path.

Return -ETIMEDOUT when the completion wait expires. Terminate the DMA channel before unmapping the scatterlist so the timed out transfer cannot continue to access the buffer after the error is returned.

Fixes 2944a44da09e ("mtd: add LPC32xx SLC NAND driver")
---
 drivers/mtd/nand/raw/lpc32xx_slc.c |   12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)
drivers/mtd/nand/raw/lpc32xx_slc.c +10 -2
@@ -430,6 +430,7 @@ static int lpc32xx_xmit_dma(struct mtd_i
 	struct dma_async_tx_descriptor *desc;
 	int flags = DMA_CTRL_ACK | DMA_PREP_INTERRUPT;
 	int res;
+	unsigned long time_left;
 
 	host->dma_slave_config.direction = dir;
 	host->dma_slave_config.src_addr = dma;
@@ -467,12 +468,19 @@ static int lpc32xx_xmit_dma(struct mtd_i
 	dmaengine_submit(desc);
 	dma_async_issue_pending(host->dma_chan);
 
-	wait_for_completion_timeout(&host->comp, msecs_to_jiffies(1000));
+	time_left = wait_for_completion_timeout(&host->comp,
+						msecs_to_jiffies(1000));
+	if (!time_left) {
+		dmaengine_terminate_sync(host->dma_chan);
+		res = -ETIMEDOUT;
+	} else {
+		res = 0;
+	}
 
 	dma_unmap_sg(host->dma_chan->device->dev, &host->sgl, 1,
 		     DMA_BIDIRECTIONAL);
 
-	return 0;
+	return res;
 out1:
 	dma_unmap_sg(host->dma_chan->device->dev, &host->sgl, 1,
 		     DMA_BIDIRECTIONAL);
author maintainer
#1512 lore raw

PATCH 1506/1611 mmc: block: fix RPMB device unregister ordering

In reply to Greg Kroah-Hartman

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

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

From: Ao Sun <[email protected]>

commit 86152fef52cac15cd662ed3bfc7604fbfef378f0 upstream.

Since commit 7852028a35f0 ("mmc: block: register RPMB partition with the RPMB subsystem"), each mmc RPMB partition is represented by two device objects:
- the mmc-owned device (rpmb->dev, backing the legacy /dev/mmcblkXrpmb char device) and
- the rpmb-core device (rdev, backing /dev/rpmbN).

The child RPMB device holds a reference to its parent, so the parent's release callback cannot be invoked if the child device is still registered.

Remove rpmb_dev_unregister() from the parent release handler and unregister the child RPMB device in the remove path before tearing down the parent device.

Also delete the extra blank line between mmc_blk_remove_rpmb_part() and {.

Fixes 7852028a35f0 ("mmc: block: register RPMB partition with the RPMB subsystem")
---
 drivers/mmc/core/block.c |    3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
drivers/mmc/core/block.c +1 -2
@@ -2784,7 +2784,6 @@ static void mmc_blk_rpmb_device_release(
 {
 	struct mmc_rpmb_data *rpmb = dev_get_drvdata(dev);
 
-	rpmb_dev_unregister(rpmb->rdev);
 	mmc_blk_put(rpmb->md);
 	ida_free(&mmc_rpmb_ida, rpmb->id);
 	kfree(rpmb);
@@ -2999,8 +2998,8 @@ out_put_device:
 }
 
 static void mmc_blk_remove_rpmb_part(struct mmc_rpmb_data *rpmb)
-
 {
+	rpmb_dev_unregister(rpmb->rdev);
 	cdev_device_del(&rpmb->chrdev, &rpmb->dev);
 	put_device(&rpmb->dev);
 }
author maintainer
#1513 lore raw

PATCH 1507/1611 mmc: sdhci-of-dwcmshc: check bus clock enable result in the probe() method

In reply to Greg Kroah-Hartman

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

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

From: Sergey Shtylyov <[email protected]>

commit 521f39ca93cc43ce1b3eae8d44201f8f55dd9151 upstream.

In the driver's probe() method, clk_disable_unprepare() for the bus clock is called on the error path even if the prior clk_prepare_enable() call has failed (and the same thing happens in the remove() method as well) -- that would cause the prepare/enable counter imbalance. Also, the same problem can happen in the driver's suspend() method; note that the resume() method does check the clk_prepare_enable()'s result -- let's be consistent and do that in probe() method as well. BTW, I don't know for sure what does the bus clock control -- if it affects the register accesses, the driver will likely cause (e.g. on ARM) a kernel oops if it fails to prepare/enable the bus clock in the probe() method...

Found by Linux Verification Center (linuxtesting.org) with the Svace static analysis tool.

Fixes e438cf49b305 ("mmc: sdhci-of-dwcmshc: add SDHCI OF Synopsys DWC MSHC driver")
Fixes bccce2ec7790 ("mmc: sdhci-of-dwcmshc: add suspend/resume support")
---
 drivers/mmc/host/sdhci-of-dwcmshc.c |   14 +++++++++-----
 1 file changed, 9 insertions(+), 5 deletions(-)
drivers/mmc/host/sdhci-of-dwcmshc.c +9 -5
@@ -1421,13 +1421,16 @@ static int dwcmshc_probe(struct platform
 			return err;
 
 		priv->bus_clk = devm_clk_get(dev, "bus");
-		if (!IS_ERR(priv->bus_clk))
-			clk_prepare_enable(priv->bus_clk);
+		if (!IS_ERR(priv->bus_clk)) {
+			err = clk_prepare_enable(priv->bus_clk);
+			if (err)
+				goto err_clk;
+		}
 	}
 
 	err = mmc_of_parse(host->mmc);
 	if (err)
-		goto err_clk;
+		goto err_bus_clk;
 
 	sdhci_get_of_property(pdev);
 
@@ -1441,7 +1444,7 @@ static int dwcmshc_probe(struct platform
 	if (pltfm_data->init) {
 		err = pltfm_data->init(&pdev->dev, host, priv);
 		if (err)
-			goto err_clk;
+			goto err_bus_clk;
 	}
 
 #ifdef CONFIG_ACPI
@@ -1487,9 +1490,10 @@ err_setup_host:
 err_rpm:
 	pm_runtime_disable(dev);
 	pm_runtime_put_noidle(dev);
+err_bus_clk:
+	clk_disable_unprepare(priv->bus_clk);
 err_clk:
 	clk_disable_unprepare(pltfm_host->clk);
-	clk_disable_unprepare(priv->bus_clk);
 	clk_bulk_disable_unprepare(priv->num_other_clks, priv->other_clks);
 	return err;
 }
author maintainer
#1514 lore raw

PATCH 1508/1611 mmc: sdhci-esdhc-imx: remove unnecessary mmc_card_wake_sdio_irq check for tuning save/restore

In reply to Greg Kroah-Hartman

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

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

From: Luke Wang <[email protected]>

commit 899160e2774d9952e9f2770b38f701ff1906c0b2 upstream.

The tuning save/restore during system PM is conditioned on mmc_card_wake_sdio_irq(), but this check is unrelated to whether tuning values need to be preserved. The actual requirement is that the card keeps power during suspend and the controller is a uSDHC.

SDIO devices using out-of-band GPIO wakeup maintain power during suspend but do not set the SDIO IRQ wake flag. In this case the tuning delay values are not saved/restored.

Remove the unnecessary mmc_card_wake_sdio_irq() condition from both the suspend save and resume restore paths.

Fixes c63d25cdc59a ("mmc: sdhci-esdhc-imx: Save tuning value when card stays powered in suspend")
---
 drivers/mmc/host/sdhci-esdhc-imx.c |    6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)
drivers/mmc/host/sdhci-esdhc-imx.c +2 -4
@@ -2041,8 +2041,7 @@ static int sdhci_esdhc_suspend(struct de
 	 * to save the tuning delay value just in case the usdhc
 	 * lost power during system PM.
 	 */
-	if (mmc_card_keep_power(host->mmc) && mmc_card_wake_sdio_irq(host->mmc) &&
-	    esdhc_is_usdhc(imx_data))
+	if (mmc_card_keep_power(host->mmc) && esdhc_is_usdhc(imx_data))
 		sdhc_esdhc_tuning_save(host);
 
 	if (device_may_wakeup(dev)) {
@@ -2107,8 +2106,7 @@ static int sdhci_esdhc_resume(struct dev
 	 * restore the saved tuning delay value for the device which keep
 	 * power during system PM.
 	 */
-	if (mmc_card_keep_power(host->mmc) && mmc_card_wake_sdio_irq(host->mmc) &&
-	    esdhc_is_usdhc(imx_data))
+	if (mmc_card_keep_power(host->mmc) && esdhc_is_usdhc(imx_data))
 		sdhc_esdhc_tuning_restore(host);
 
 	pm_runtime_put_autosuspend(dev);
author maintainer
#1515 lore raw

PATCH 1509/1611 mmc: sdhci-esdhc-imx: restore DLL override for DDR modes on resume

In reply to Greg Kroah-Hartman

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

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

From: Luke Wang <[email protected]>

commit 2439becd91bad6883b135044f85f83a0538b96a6 upstream.

sdhci_esdhc_imx_hwinit() unconditionally clears ESDHC_DLL_CTRL by writing zero. For SDIO devices that keep power during system suspend and operate in DDR mode, the card remains in DDR timing while the host DLL override configuration is lost.

Extract the DLL override setup from esdhc_set_uhs_signaling() into a helper esdhc_set_dll_override(), and call it on the resume path when the card kept power and is using a DDR timing mode.

Fixes 676a83855614 ("mmc: host: sdhci-esdhc-imx: refactor the system PM logic")
---
 drivers/mmc/host/sdhci-esdhc-imx.c |   39 ++++++++++++++++++++++++++-----------
 1 file changed, 28 insertions(+), 11 deletions(-)
drivers/mmc/host/sdhci-esdhc-imx.c +28 -11
@@ -1338,6 +1338,23 @@ static int esdhc_change_pinstate(struct
 	return pinctrl_select_state(imx_data->pinctrl, pinctrl);
 }
 
+static void esdhc_set_dll_override(struct sdhci_host *host)
+{
+	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
+	struct pltfm_imx_data *imx_data = sdhci_pltfm_priv(pltfm_host);
+	struct esdhc_platform_data *boarddata = &imx_data->boarddata;
+	u32 v;
+
+	if (!boarddata->delay_line)
+		return;
+
+	v = boarddata->delay_line << ESDHC_DLL_OVERRIDE_VAL_SHIFT |
+	    (1 << ESDHC_DLL_OVERRIDE_EN_SHIFT);
+	if (is_imx53_esdhc(imx_data))
+		v <<= 1;
+	writel(v, host->ioaddr + ESDHC_DLL_CTRL);
+}
+
 /*
  * For HS400 eMMC, there is a data_strobe line. This signal is generated
  * by the device and used for data output and CRC status response output
@@ -1393,7 +1410,6 @@ static void esdhc_set_uhs_signaling(stru
 	u32 m;
 	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
 	struct pltfm_imx_data *imx_data = sdhci_pltfm_priv(pltfm_host);
-	struct esdhc_platform_data *boarddata = &imx_data->boarddata;
 
 	/* disable ddr mode and disable HS400 mode */
 	m = readl(host->ioaddr + ESDHC_MIX_CTRL);
@@ -1414,15 +1430,7 @@ static void esdhc_set_uhs_signaling(stru
 		m |= ESDHC_MIX_CTRL_DDREN;
 		writel(m, host->ioaddr + ESDHC_MIX_CTRL);
 		imx_data->is_ddr = 1;
-		if (boarddata->delay_line) {
-			u32 v;
-			v = boarddata->delay_line <<
-				ESDHC_DLL_OVERRIDE_VAL_SHIFT |
-				(1 << ESDHC_DLL_OVERRIDE_EN_SHIFT);
-			if (is_imx53_esdhc(imx_data))
-				v <<= 1;
-			writel(v, host->ioaddr + ESDHC_DLL_CTRL);
-		}
+		esdhc_set_dll_override(host);
 		break;
 	case MMC_TIMING_MMC_HS400:
 		m |= ESDHC_MIX_CTRL_DDREN | ESDHC_MIX_CTRL_HS400_EN;
@@ -2106,9 +2114,18 @@ static int sdhci_esdhc_resume(struct dev
 	 * restore the saved tuning delay value for the device which keep
 	 * power during system PM.
 	 */
-	if (mmc_card_keep_power(host->mmc) && esdhc_is_usdhc(imx_data))
+	if (mmc_card_keep_power(host->mmc) && esdhc_is_usdhc(imx_data)) {
 		sdhc_esdhc_tuning_restore(host);
 
+		/*
+		 * Restore DLL override for DDR modes. hwinit unconditionally
+		 * clears ESDHC_DLL_CTRL, but the card is still in DDR mode.
+		 */
+		if (host->timing == MMC_TIMING_UHS_DDR50 ||
+		    host->timing == MMC_TIMING_MMC_DDR52)
+			esdhc_set_dll_override(host);
+	}
+
 	pm_runtime_put_autosuspend(dev);
 
 	return ret;
author maintainer
#1516 lore raw

PATCH 1510/1611 mmc: sdhci-esdhc-imx: fix esdhc_change_pinstate() to allow default state restore

In reply to Greg Kroah-Hartman

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

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

From: Luke Wang <[email protected]>

commit 5adc14cd4b905629d5b9163b3a416dcab24c7ce2 upstream.

esdhc_change_pinstate() checks for pins_100mhz and pins_200mhz at the top of the function and returns -EINVAL if either is not defined. This prevents the default case from ever being reached, which means devices with a sleep pinctrl state but without high-speed pin states (100mhz/ 200mhz) can never restore their default pin configuration.

Move the IS_ERR checks for pins_100mhz and pins_200mhz into their respective switch cases.

Fixes 676a83855614 ("mmc: host: sdhci-esdhc-imx: refactor the system PM logic")
---
 drivers/mmc/host/sdhci-esdhc-imx.c |    8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)
drivers/mmc/host/sdhci-esdhc-imx.c +5 -3
@@ -1315,19 +1315,21 @@ static int esdhc_change_pinstate(struct
 
 	dev_dbg(mmc_dev(host->mmc), "change pinctrl state for uhs %d\n", uhs);
 
-	if (IS_ERR(imx_data->pinctrl) ||
-		IS_ERR(imx_data->pins_100mhz) ||
-		IS_ERR(imx_data->pins_200mhz))
+	if (IS_ERR(imx_data->pinctrl))
 		return -EINVAL;
 
 	switch (uhs) {
 	case MMC_TIMING_UHS_SDR50:
 	case MMC_TIMING_UHS_DDR50:
+		if (IS_ERR(imx_data->pins_100mhz))
+			return -EINVAL;
 		pinctrl = imx_data->pins_100mhz;
 		break;
 	case MMC_TIMING_UHS_SDR104:
 	case MMC_TIMING_MMC_HS200:
 	case MMC_TIMING_MMC_HS400:
+		if (IS_ERR(imx_data->pins_200mhz))
+			return -EINVAL;
 		pinctrl = imx_data->pins_200mhz;
 		break;
 	default:
author maintainer
#1517 lore raw

PATCH 1511/1611 mmc: sdhci-esdhc-imx: disable irq during suspend to fix unhandled interrupt

In reply to Greg Kroah-Hartman

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

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

From: Luke Wang <[email protected]>

commit 9d87eaf985cef9581b6ed99b461b38e8cd666480 upstream.

When using WIFI out-of-band wakeup, an "irq xxx: nobody cared" warning occurs. This happens because the usdhc interrupt is not disabled during system suspend when device_may_wakeup() returns false.

The sequence of events leading to this issue:
1. System enters suspend without disabling usdhc interrupt (because device_may_wakeup() returns false for usdhc device)
2. WIFI out-of-band wakeup triggers system resume via GPIO interrupt
3. WIFI sends a Card interrupt before usdhc has fully resumed
4. usdhc is still in runtime suspend state and cannot handle the interrupt properly
5. The unhandled interrupt triggers "nobody cared" warning

Fix this by unconditionally disabling the usdhc interrupt during suspend and re-enabling it during resume, regardless of the wakeup capability. This ensures no interrupts are processed during the suspend/resume transition.

Fixes 676a83855614 ("mmc: host: sdhci-esdhc-imx: refactor the system PM logic")
---
 drivers/mmc/host/sdhci-esdhc-imx.c |   11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)
drivers/mmc/host/sdhci-esdhc-imx.c +6 -5
@@ -2054,9 +2054,10 @@ static int sdhci_esdhc_suspend(struct de
 	if (mmc_card_keep_power(host->mmc) && esdhc_is_usdhc(imx_data))
 		sdhc_esdhc_tuning_save(host);
 
+	/* The irqs of imx are not shared. It is safe to disable */
+	disable_irq(host->irq);
+
 	if (device_may_wakeup(dev)) {
-		/* The irqs of imx are not shared. It is safe to disable */
-		disable_irq(host->irq);
 		ret = sdhci_enable_irq_wakeups(host);
 		if (!ret)
 			dev_warn(dev, "Failed to enable irq wakeup\n");
@@ -2107,10 +2108,10 @@ static int sdhci_esdhc_resume(struct dev
 	/* re-initialize hw state in case it's lost in low power mode */
 	sdhci_esdhc_imx_hwinit(host);
 
-	if (host->irq_wake_enabled) {
+	if (host->irq_wake_enabled)
 		sdhci_disable_irq_wakeups(host);
-		enable_irq(host->irq);
-	}
+
+	enable_irq(host->irq);
 
 	/*
 	 * restore the saved tuning delay value for the device which keep
author maintainer
#1518 lore raw

PATCH 1512/1611 mmc: sdhci-esdhc-imx: use pm_runtime_resume_and_get() in suspend

In reply to Greg Kroah-Hartman

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

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

From: Luke Wang <[email protected]>

commit 8da5930144712412d85e7f868693d96ec5c2018c upstream.

Replace pm_runtime_get_sync() with pm_runtime_resume_and_get() to simplify error handling. pm_runtime_resume_and_get() automatically drops the usage counter on failure, avoiding the need for a separate pm_runtime_put_noidle() call. If it fails, the device is unclocked and accessing hardware registers would cause a kernel panic, so return the error immediately.

Fixes 676a83855614 ("mmc: host: sdhci-esdhc-imx: refactor the system PM logic")
---
 drivers/mmc/host/sdhci-esdhc-imx.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/mmc/host/sdhci-esdhc-imx.c +3 -1
@@ -2038,7 +2038,9 @@ static int sdhci_esdhc_suspend(struct de
 	 * 2, make sure the pm_runtime_force_resume() in sdhci_esdhc_resume() really
 	 *    invoke its ->runtime_resume callback (needs_force_resume = 1).
 	 */
-	pm_runtime_get_sync(dev);
+	ret = pm_runtime_resume_and_get(dev);
+	if (ret)
+		return ret;
 
 	if ((imx_data->socdata->flags & ESDHC_FLAG_STATE_LOST_IN_LPMODE) &&
 		(host->tuning_mode != SDHCI_TUNING_MODE_1)) {
author maintainer
#1519 lore raw

PATCH 1513/1611 mmc: sdhci-esdhc-imx: make non-fatal errors non-blocking in suspend

In reply to Greg Kroah-Hartman

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

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

From: Luke Wang <[email protected]>

commit 6aa00a43bbd3d994558a55586351757cebbff236 upstream.

Make pinctrl_pm_select_sleep_state() and mmc_gpio_set_cd_wake() failures non-fatal in the suspend path. These failures only mean slightly higher power consumption or missing CD wakeup capability, but should not block system suspend.

Also change the function to always return 0 on the success path instead of propagating non-fatal warning return values.

Fixes 676a83855614 ("mmc: host: sdhci-esdhc-imx: refactor the system PM logic")
---
 drivers/mmc/host/sdhci-esdhc-imx.c |   13 ++++++-------
 1 file changed, 6 insertions(+), 7 deletions(-)
drivers/mmc/host/sdhci-esdhc-imx.c +6 -7
@@ -2060,8 +2060,7 @@ static int sdhci_esdhc_suspend(struct de
 	disable_irq(host->irq);
 
 	if (device_may_wakeup(dev)) {
-		ret = sdhci_enable_irq_wakeups(host);
-		if (!ret)
+		if (!sdhci_enable_irq_wakeups(host))
 			dev_warn(dev, "Failed to enable irq wakeup\n");
 	} else {
 		/*
@@ -2072,12 +2071,12 @@ static int sdhci_esdhc_suspend(struct de
 		 * other function like GPIO function to save power in PM,
 		 * which finally block the SDIO wakeup function.
 		 */
-		ret = pinctrl_pm_select_sleep_state(dev);
-		if (ret)
-			return ret;
+		if (pinctrl_pm_select_sleep_state(dev))
+			dev_warn(dev, "Failed to select sleep pinctrl state\n");
 	}
 
-	ret = mmc_gpio_set_cd_wake(host->mmc, true);
+	if (mmc_gpio_set_cd_wake(host->mmc, true))
+		dev_warn(dev, "Failed to enable cd wake\n");
 
 	/*
 	 * Make sure invoke runtime_suspend to gate off clock.
@@ -2085,7 +2084,7 @@ static int sdhci_esdhc_suspend(struct de
 	 */
 	pm_runtime_force_suspend(dev);
 
-	return ret;
+	return 0;
 }
 
 static int sdhci_esdhc_resume(struct device *dev)
author maintainer
#1520 lore raw

PATCH 1514/1611 mmc: sdhci-esdhc-imx: fix resume error handling

In reply to Greg Kroah-Hartman

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

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

From: Luke Wang <[email protected]>

commit e27c946b589c53520409a0956b33d52ef7a0898f upstream.

Check pm_runtime_force_resume() return value in resume. If it fails (clock enable failure), return immediately since accessing hardware registers on an unclocked device would cause a kernel panic.

The early return intentionally skips enable_irq() and sdhci_disable_irq_wakeups() because the IRQ handler reads SDHCI_INT_STATUS, which would also fault without clocks. The PM runtime usage counter leak only affects this already-broken device instance and is an acceptable tradeoff to preserve system stability.

Remove the return value check for mmc_gpio_set_cd_wake(host->mmc, false) since disable_irq_wake() called internally always returns 0.

Also return 0 explicitly on the success path instead of propagating stale return values.

Fixes 676a83855614 ("mmc: host: sdhci-esdhc-imx: refactor the system PM logic")
---
 drivers/mmc/host/sdhci-esdhc-imx.c |    8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
drivers/mmc/host/sdhci-esdhc-imx.c +4 -4
@@ -2100,12 +2100,12 @@ static int sdhci_esdhc_resume(struct dev
 			dev_warn(dev, "Failed to restore pinctrl state\n");
 	}
 
-	pm_runtime_force_resume(dev);
-
-	ret = mmc_gpio_set_cd_wake(host->mmc, false);
+	ret = pm_runtime_force_resume(dev);
 	if (ret)
 		return ret;
 
+	mmc_gpio_set_cd_wake(host->mmc, false);
+
 	/* re-initialize hw state in case it's lost in low power mode */
 	sdhci_esdhc_imx_hwinit(host);
 
@@ -2132,7 +2132,7 @@ static int sdhci_esdhc_resume(struct dev
 
 	pm_runtime_put_autosuspend(dev);
 
-	return ret;
+	return 0;
 }
 
 static int sdhci_esdhc_runtime_suspend(struct device *dev)
author maintainer
#1521 lore raw

PATCH 1515/1611 crypto: xilinx-trng - Remove crypto_rng interface

In reply to Greg Kroah-Hartman

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

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

From: Eric Biggers <[email protected]>

commit 32b4d29280ed2a991dc196d5845b892acedc63b8 upstream.

Implementing the crypto_rng interface has no purpose, as it isn't used in practice. It's being removed from other drivers too. Just remove it. This leaves hwrng, which is actually used.

Tagging with 'Cc stable' due to the bugs that this removes:

- xtrng_trng_generate() sometimes returned success even when it didn't fill in all the bytes.

- It was possible for xtrng_trng_generate() and xtrng_hwrng_trng_read() to run concurrently and interfere with each other, as the locking code in xtrng_hwrng_trng_read() was broken.

Fixes 8979744aca80 ("crypto: xilinx - Add TRNG driver for Versal")
---
 drivers/crypto/Kconfig              |    1 
 drivers/crypto/xilinx/xilinx-trng.c |   75 ------------------------------------
 2 files changed, 1 insertion(+), 75 deletions(-)
drivers/crypto/Kconfig +0 -1
@@ -728,7 +728,6 @@ config CRYPTO_DEV_TEGRA
 config CRYPTO_DEV_XILINX_TRNG
 	tristate "Support for Xilinx True Random Generator"
 	depends on ZYNQMP_FIRMWARE || COMPILE_TEST
-	select CRYPTO_RNG
 	select HW_RANDOM
 	help
 	  Xilinx Versal SoC driver provides kernel-side support for True Random Number
drivers/crypto/xilinx/xilinx-trng.c +1 -74

142 lines not on this page · raw message

author maintainer
#1522 lore raw

PATCH 1516/1611 ACPI: NFIT: core: Fix acpi_nfit_init() error cleanup

In reply to Greg Kroah-Hartman

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

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

From: "Rafael J. Wysocki" <[email protected]>

[ Upstream commit 38bf27511ef41bffebd157ec3eba41fc89ba59cd ]

If acpi_nfit_init() fails after adding the acpi_desc object to the acpi_descs list, that object is never removed from that list because the acpi_nfit_shutdown() devm action is not added for the NFIT device in that case. Next, the acpi_nfit_init() failure causes acpi_nfit_probe() to fail, the acpi_desc object is freed, and a dangling pointer is left behind in the acpi_descs. Any subsequent ACPI Machine Check Exception will trigger nfit_handle_mce() which iterates over acpi_descs and so a use-after-free will occur.

Moreover, if acpi_nfit_probe() returns 0 after installing a notify handler for the NFIT device and without allocating the acpi_desc object and setting the NFIT device's driver data pointer, the acpi_desc object will be allocated by acpi_nfit_update_notify() and acpi_nfit_init() will be called to initialize it. Regardless of whether or not acpi_nfit_init() fails in that case, the acpi_nfit_shutdown() devm action is not added for the NFIT device and acpi_desc is never removed from the acpi_descs list. If the acpi_desc object is freed subsequently on driver removal, any subsequent ACPI MCE will lead to a use-after-free like in the previous case.

To address the first issue mentioned above, make acpi_nfit_probe() call acpi_nfit_shutdown() directly on acpi_nfit_init() failures and to address the other one, add a remove callback to the driver and make it call acpi_nfit_shutdown(). Also, since it is now possible to pass NULL to acpi_nfit_shutdown() or the acpi_desc object passed to it may not have been initialized, add checks against NULL for acpi_desc and its nvdimm_bus field to that function and make acpi_nfit_unregister() clear the latter after unregistering the NVDIMM bus.

Fixes a61fe6f7902e ("nfit, tools/testing/nvdimm: unify common init for acpi_nfit_desc")
Fixes fbabd829fe76 ("acpi, nfit: fix module unload vs workqueue shutdown race")
Cc All applicable <[email protected]>
---
 drivers/acpi/nfit/core.c |   18 +++++++++++++++---
 1 file changed, 15 insertions(+), 3 deletions(-)
drivers/acpi/nfit/core.c +15 -3
@@ -3061,6 +3061,8 @@ static void acpi_nfit_unregister(void *d
 	struct acpi_nfit_desc *acpi_desc = data;
 
 	nvdimm_bus_unregister(acpi_desc->nvdimm_bus);
+	/* The nvdimm_bus object may have been freed, so clear the pointer. */
+	acpi_desc->nvdimm_bus = NULL;
 }
 
 int acpi_nfit_init(struct acpi_nfit_desc *acpi_desc, void *data, acpi_size sz)
@@ -3301,7 +3303,10 @@ static void acpi_nfit_remove_notify_hand
 void acpi_nfit_shutdown(void *data)
 {
 	struct acpi_nfit_desc *acpi_desc = data;
-	struct device *bus_dev = to_nvdimm_bus_dev(acpi_desc->nvdimm_bus);
+	struct device *bus_dev;
+
+	if (!acpi_desc || !acpi_desc->nvdimm_bus)
+		return;
 
 	/*
 	 * Destruct under acpi_desc_lock so that nfit_handle_mce does not
@@ -3316,6 +3321,7 @@ void acpi_nfit_shutdown(void *data)
 	mutex_unlock(&acpi_desc->init_mutex);
 	cancel_delayed_work_sync(&acpi_desc->dwork);
 
+	bus_dev = to_nvdimm_bus_dev(acpi_desc->nvdimm_bus);
 	/*
 	 * Bounce the nvdimm bus lock to make sure any in-flight
 	 * acpi_nfit_ars_rescan() submissions have had a chance to
@@ -3393,9 +3399,14 @@ static int acpi_nfit_add(struct acpi_dev
 				sz - sizeof(struct acpi_table_nfit));
 
 	if (rc)
-		return rc;
+		acpi_nfit_shutdown(acpi_desc);
 
-	return devm_add_action_or_reset(dev, acpi_nfit_shutdown, acpi_desc);
+	return rc;
+}
+
+static void acpi_nfit_remove(struct acpi_device *adev)
+{
+	acpi_nfit_shutdown(dev_get_drvdata(&adev->dev));
 }
 
 static void acpi_nfit_update_notify(struct device *dev, acpi_handle handle)
@@ -3482,6 +3493,7 @@ static struct acpi_driver acpi_nfit_driv
 	.ids = acpi_nfit_ids,
 	.ops = {
 		.add = acpi_nfit_add,
+		.remove = acpi_nfit_remove,
 	},
 };
 
author maintainer
#1523 lore raw

PATCH 1517/1611 ACPI: driver: Check ACPI_COMPANION() against NULL during probe

In reply to Greg Kroah-Hartman

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

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

From: "Rafael J. Wysocki" <[email protected]>

[ Upstream commit e4865a56d013e86e46ea6acea15bb6eae01898ff ]

Since every platform driver can be forced to match a device that doesn't match its list of device IDs because of device_match_driver_override(), platform drivers that rely on the existence of a device's ACPI companion object should verify its presence.

Accordingly, add requisite ACPI_COMPANION() or ACPI_HANDLE() checks against NULL to 13 platform drivers handling core ACPI devices.

Also change the value returned by the ACPI thermal zone driver when the device's ACPI companion is not present to -ENODEV for consistency with the other drivers.

Stable-dep-of: 18a00ed0e718 ("ACPI: NFIT: core: Fix possible deadlock and missing notifications")

---
 drivers/acpi/ac.c            |    6 +++++-
 drivers/acpi/acpi_pad.c      |    6 +++++-
 drivers/acpi/acpi_tad.c      |    6 +++++-
 drivers/acpi/pfr_telemetry.c |    6 +++++-
 drivers/acpi/pfr_update.c    |    6 +++++-
 drivers/acpi/thermal.c       |    2 +-
 6 files changed, 26 insertions(+), 6 deletions(-)
drivers/acpi/ac.c +5 -1
@@ -203,11 +203,15 @@ static const struct dmi_system_id ac_dmi
 
 static int acpi_ac_probe(struct platform_device *pdev)
 {
-	struct acpi_device *adev = ACPI_COMPANION(&pdev->dev);
 	struct power_supply_config psy_cfg = {};
+	struct acpi_device *adev;
 	struct acpi_ac *ac;
 	int result;
 
+	adev = ACPI_COMPANION(&pdev->dev);
+	if (!adev)
+		return -ENODEV;
+
 	ac = kzalloc(sizeof(struct acpi_ac), GFP_KERNEL);
 	if (!ac)
 		return -ENOMEM;
drivers/acpi/acpi_pad.c +5 -1
@@ -426,9 +426,13 @@ static void acpi_pad_notify(acpi_handle
 
 static int acpi_pad_probe(struct platform_device *pdev)
 {
-	struct acpi_device *adev = ACPI_COMPANION(&pdev->dev);
+	struct acpi_device *adev;
 	acpi_status status;
 
+	adev = ACPI_COMPANION(&pdev->dev);
+	if (!adev)
+		return -ENODEV;
+
 	strscpy(acpi_device_name(adev), ACPI_PROCESSOR_AGGREGATOR_DEVICE_NAME);
 	strscpy(acpi_device_class(adev), ACPI_PROCESSOR_AGGREGATOR_CLASS);
 
drivers/acpi/acpi_tad.c +5 -1
@@ -588,12 +588,16 @@ static void acpi_tad_remove(struct platf
 static int acpi_tad_probe(struct platform_device *pdev)
 {
 	struct device *dev = &pdev->dev;
-	acpi_handle handle = ACPI_HANDLE(dev);
 	struct acpi_tad_driver_data *dd;
+	acpi_handle handle;
 	acpi_status status;
 	unsigned long long caps;
 	int ret;
 
+	handle = ACPI_HANDLE(dev);
+	if (!handle)
+		return -ENODEV;
+
 	ret = acpi_install_cmos_rtc_space_handler(handle);
 	if (ret < 0) {
 		dev_info(dev, "Unable to install space handler\n");
drivers/acpi/pfr_telemetry.c +5 -1
@@ -360,10 +360,14 @@ static void pfrt_log_put_idx(void *data)
 
 static int acpi_pfrt_log_probe(struct platform_device *pdev)
 {
-	acpi_handle handle = ACPI_HANDLE(&pdev->dev);
 	struct pfrt_log_device *pfrt_log_dev;
+	acpi_handle handle;
 	int ret;
 
+	handle = ACPI_HANDLE(&pdev->dev);
+	if (!handle)
+		return -ENODEV;
+
 	if (!acpi_has_method(handle, "_DSM")) {
 		dev_dbg(&pdev->dev, "Missing _DSM\n");
 		return -ENODEV;
drivers/acpi/pfr_update.c +5 -1
@@ -538,10 +538,14 @@ static void pfru_put_idx(void *data)
 
 static int acpi_pfru_probe(struct platform_device *pdev)
 {
-	acpi_handle handle = ACPI_HANDLE(&pdev->dev);
 	struct pfru_device *pfru_dev;
+	acpi_handle handle;
 	int ret;
 
+	handle = ACPI_HANDLE(&pdev->dev);
+	if (!handle)
+		return -ENODEV;
+
 	if (!acpi_has_method(handle, "_DSM")) {
 		dev_dbg(&pdev->dev, "Missing _DSM\n");
 		return -ENODEV;
drivers/acpi/thermal.c +1 -1
@@ -789,7 +789,7 @@ static int acpi_thermal_add(struct acpi_
 	int i;
 
 	if (!device)
-		return -EINVAL;
+		return -ENODEV;
 
 	tz = kzalloc(sizeof(struct acpi_thermal), GFP_KERNEL);
 	if (!tz)
author maintainer
#1524 lore raw

PATCH 1518/1611 ACPI: bus: Introduce devm_acpi_install_notify_handler()

In reply to Greg Kroah-Hartman

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

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

From: "Rafael J. Wysocki" <[email protected]>

[ Upstream commit ca70ce555a1eb8edf5fb1d5575f1fe19c9ae17f4 ]

Introduce devm_acpi_install_notify_handler() for installing an ACPI notify handler managed by devres that will be removed automatically on driver detach.

It installs the notify handler on the device object in the ACPI namespace that corresponds to the owner device's ACPI companion, if present (an error is returned if the owner device doesn't have an ACPI companion).

Currently, there is no way to manually remove the notify handler installed by it because none of its users brought on subsequently will need to do that.

Stable-dep-of: 18a00ed0e718 ("ACPI: NFIT: core: Fix possible deadlock and missing notifications")

---
 drivers/acpi/bus.c      |   66 ++++++++++++++++++++++++++++++++++++++++++++++++
 include/acpi/acpi_bus.h |    2 +
 2 files changed, 68 insertions(+)
drivers/acpi/bus.c +66 -0
@@ -622,6 +622,72 @@ void acpi_dev_remove_notify_handler(stru
 }
 EXPORT_SYMBOL_GPL(acpi_dev_remove_notify_handler);
 
+struct acpi_notify_handler_devres {
+	acpi_notify_handler handler;
+	u32 handler_type;
+};
+
+static void devm_acpi_notify_handler_release(struct device *dev, void *res)
+{
+	struct acpi_notify_handler_devres *dr = res;
+
+	acpi_dev_remove_notify_handler(ACPI_COMPANION(dev), dr->handler_type,
+				       dr->handler);
+}
+
+/**
+ * devm_acpi_install_notify_handler - Install an ACPI notify handler for a
+ * 				      managed device
+ * @dev: Device to install a notify handler for
+ * @handler_type: Type of the notify handler
+ * @handler: Handler function to install
+ * @context: Data passed back to the handler function
+ *
+ * This function performs the same function as acpi_dev_install_notify_handler()
+ * called for the ACPI companion of @dev with the same @handler_type, @handler,
+ * and @context arguments, but the ACPI notify handler installed by it will be
+ * automatically removed on driver detach.
+ *
+ * Callers should ensure that all resources used by @handler have been allocated
+ * prior to invoking this function, in which case those resources should be
+ * devres-managed so that they won't be released before the notify handler
+ * removal.  Otherwise, special synchronization between @handler and the
+ * management of those resources is required.
+ *
+ * When the request fails, an error message is printed.  Don't add extra error
+ * messages at the call sites.
+ *
+ * Return: 0 on success or a negative error number.
+ */
+int devm_acpi_install_notify_handler(struct device *dev, u32 handler_type,
+				     acpi_notify_handler handler, void *context)
+{
+	struct acpi_notify_handler_devres *dr;
+	struct acpi_device *adev;
+	int ret;
+
+	adev = ACPI_COMPANION(dev);
+	if (!adev)
+		return dev_err_probe(dev, -ENODEV, "No ACPI companion in %s()\n", __func__);
+
+	dr = devres_alloc(devm_acpi_notify_handler_release, sizeof(*dr), GFP_KERNEL);
+	if (!dr)
+		return -ENOMEM;
+
+	ret = acpi_dev_install_notify_handler(adev, handler_type, handler, context);
+	if (ret) {
+		devres_free(dr);
+		return dev_err_probe(dev, ret, "Failed to install an ACPI notify handler\n");
+	}
+
+	dr->handler = handler;
+	dr->handler_type = handler_type;
+	devres_add(dev, dr);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(devm_acpi_install_notify_handler);
+
 /* Handle events targeting \_SB device (at present only graceful shutdown) */
 
 #define ACPI_SB_NOTIFY_SHUTDOWN_REQUEST 0x81
include/acpi/acpi_bus.h +2 -0
@@ -625,6 +625,8 @@ int acpi_dev_install_notify_handler(stru
 void acpi_dev_remove_notify_handler(struct acpi_device *adev,
 				    u32 handler_type,
 				    acpi_notify_handler handler);
+int devm_acpi_install_notify_handler(struct device *dev, u32 handler_type,
+				     acpi_notify_handler handler, void *context);
 extern int acpi_notifier_call_chain(struct acpi_device *, u32, u32);
 extern int register_acpi_notifier(struct notifier_block *);
 extern int unregister_acpi_notifier(struct notifier_block *);
author maintainer
#1525 lore raw

PATCH 1519/1611 ACPI: NFIT: core: Use devm_acpi_install_notify_handler()

In reply to Greg Kroah-Hartman

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

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

From: "Rafael J. Wysocki" <[email protected]>

[ Upstream commit 198541ad53c0d0d891fedea4098f9953a0f566c0 ]

Now that devm_acpi_install_notify_handler() is available, use it in acpi_nfit_probe() instead of a custom devm action removing an ACPI notify handler installed via acpi_dev_install_notify_handler().

Also drop the explicit ACPI_COMPANION() check against NULL that is not necessary any more becuase devm_acpi_install_notify_handler() carries out an equivalent check internally and use ACPI_HANDLE() to retrieve the platform device's ACPI handle.

No intentional functional impact.

Stable-dep-of: 18a00ed0e718 ("ACPI: NFIT: core: Fix possible deadlock and missing notifications")

---
 drivers/acpi/nfit/core.c |   27 +++++++--------------------
 1 file changed, 7 insertions(+), 20 deletions(-)
drivers/acpi/nfit/core.c +7 -20
@@ -3285,19 +3285,11 @@ static void acpi_nfit_put_table(void *ta
 
 static void acpi_nfit_notify(acpi_handle handle, u32 event, void *data)
 {
-	struct acpi_device *adev = data;
+	struct device *dev = data;
 
-	device_lock(&adev->dev);
-	__acpi_nfit_notify(&adev->dev, handle, event);
-	device_unlock(&adev->dev);
-}
-
-static void acpi_nfit_remove_notify_handler(void *data)
-{
-	struct acpi_device *adev = data;
-
-	acpi_dev_remove_notify_handler(adev, ACPI_DEVICE_NOTIFY,
-				       acpi_nfit_notify);
+	device_lock(dev);
+	__acpi_nfit_notify(dev, handle, event);
+	device_unlock(dev);
 }
 
 void acpi_nfit_shutdown(void *data)
@@ -3344,13 +3336,8 @@ static int acpi_nfit_add(struct acpi_dev
 	acpi_size sz;
 	int rc = 0;
 
-	rc = acpi_dev_install_notify_handler(adev, ACPI_DEVICE_NOTIFY,
-					     acpi_nfit_notify, adev);
-	if (rc)
-		return rc;
-
-	rc = devm_add_action_or_reset(dev, acpi_nfit_remove_notify_handler,
-					adev);
+	rc = devm_acpi_install_notify_handler(dev, ACPI_DEVICE_NOTIFY,
+					      acpi_nfit_notify, dev);
 	if (rc)
 		return rc;
 
@@ -3381,7 +3368,7 @@ static int acpi_nfit_add(struct acpi_dev
 	acpi_desc->acpi_header = *tbl;
 
 	/* Evaluate _FIT and override with that if present */
-	status = acpi_evaluate_object(adev->handle, "_FIT", NULL, &buf);
+	status = acpi_evaluate_object(ACPI_HANDLE(dev), "_FIT", NULL, &buf);
 	if (ACPI_SUCCESS(status) && buf.length > 0) {
 		union acpi_object *obj = buf.pointer;
 
author maintainer
#1526 lore raw

PATCH 1520/1611 ACPI: NFIT: core: Fix possible deadlock and missing notifications

In reply to Greg Kroah-Hartman

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

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

From: "Rafael J. Wysocki" <[email protected]>

[ Upstream commit 18a00ed0e718473f5c3fcfa49df46c944575e60c ]

After commit 9b311b7313d6 ("ACPI: NFIT: Install Notify() handler before getting NFIT table"), ACPI NFIT driver removal may deadlock if an ACPI notify on the NFIT device is triggered concurrently. A similar deadlock may occur if an ACPI notify on the NFIT device is triggered during a failing driver probe.

The deadlock is possible because acpi_dev_remove_notify_handler() calls acpi_os_wait_events_complete() after removing the notify handler and the driver core invokes it under the NFIT platform device lock which is also acquired by acpi_nfit_notify(). Thus acpi_os_wait_events_complete() may be waiting for acpi_nfit_notify() to complete, but the latter may not be able to acquire the device lock which is being held by the driver core while the former is being executed.

Moreover, after commit 03667e146f81 ("ACPI: NFIT: core: Convert the driver to a platform one"), there are no sysfs notifications regarding NVDIMM devices because __acpi_nvdimm_notify() always bails out after checking the driver data pointer of the device's parent. That parent is the ACPI companion of the platform device used for driver binding, so its driver data pointer is always NULL after the commit in question which was overlooked by it.

A remedy for the deadlock is to use a special separate lock for ACPI notify synchronization with driver probe and removal instead of the device lock of the NFIT device, while a remedy for the second issue is to populate the driver data pointer of the NFIT device's ACPI companion when the driver is ready to operate, so do both these things. However, since the new lock is not held across the entire teardown and acpi_nfit_notify() should do nothing when teardown is in progress, make it check the driver data pointer of the NFIT device's ACPI companion, in analogy with the existing check in __acpi_nvdimm_notify(), and bail out if that pointer is NULL.

Fixes 9b311b7313d6 ("ACPI: NFIT: Install Notify() handler before getting NFIT table")
Fixes 03667e146f81 ("ACPI: NFIT: core: Convert the driver to a platform one")
Cc All applicable <[email protected]> # 9995e4404ea4: ACPI: NFIT: core: Eliminate redundant local variable
---
 drivers/acpi/bus.c       |    6 +++-
 drivers/acpi/nfit/core.c |   62 +++++++++++++++++++++++++++++++++++++----------
 2 files changed, 54 insertions(+), 14 deletions(-)
drivers/acpi/bus.c +5 -1
@@ -623,6 +623,7 @@ void acpi_dev_remove_notify_handler(stru
 EXPORT_SYMBOL_GPL(acpi_dev_remove_notify_handler);
 
 struct acpi_notify_handler_devres {
+	struct acpi_device *adev;
 	acpi_notify_handler handler;
 	u32 handler_type;
 };
@@ -631,7 +632,7 @@ static void devm_acpi_notify_handler_rel
 {
 	struct acpi_notify_handler_devres *dr = res;
 
-	acpi_dev_remove_notify_handler(ACPI_COMPANION(dev), dr->handler_type,
+	acpi_dev_remove_notify_handler(dr->adev, dr->handler_type,
 				       dr->handler);
 }
 
@@ -667,6 +668,8 @@ int devm_acpi_install_notify_handler(str
 	int ret;
 
 	adev = ACPI_COMPANION(dev);
+	if (!adev && dev->bus == &acpi_bus_type)
+		adev = to_acpi_device(dev);
 	if (!adev)
 		return dev_err_probe(dev, -ENODEV, "No ACPI companion in %s()\n", __func__);
 
@@ -680,6 +683,7 @@ int devm_acpi_install_notify_handler(str
 		return dev_err_probe(dev, ret, "Failed to install an ACPI notify handler\n");
 	}
 
+	dr->adev = adev;
 	dr->handler = handler;
 	dr->handler_type = handler_type;
 	devres_add(dev, dr);
drivers/acpi/nfit/core.c +49 -13

118 lines not on this page · raw message

author maintainer
#1527 lore raw

PATCH 1521/1611 iio: hid-sensor-rotation: Fix stale or zero output when reading raw values

In reply to Greg Kroah-Hartman

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

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

From: Zhang Lixu <[email protected]>

[ Upstream commit 3ce8d099e0afc5a7da75a2007a67f67c4f5a4af1 ]

When reading the raw quaternion attribute (in_rot_quaternion_raw), the driver currently returns either all zeros (if the sensor was never enabled) or stale data (if the sensor was previously enabled) because it reads from the internal buffer without explicitly requesting a new sample from the sensor.

To fix this, power up the sensor, call sensor_hub_input_attr_read_values() to issue a synchronous GET_REPORT and receive the full quaternion data directly into a local buffer, then decode the four components.

Fixes fc18dddc0625 ("iio: hid-sensors: Added device rotation support")
---
 drivers/iio/orientation/hid-sensor-rotation.c |   40 ++++++++++++++++++++++++--
 1 file changed, 38 insertions(+), 2 deletions(-)
drivers/iio/orientation/hid-sensor-rotation.c +38 -2
@@ -74,6 +74,13 @@ static int dev_rot_read_raw(struct iio_d
 				long mask)
 {
 	struct dev_rot_state *rot_state = iio_priv(indio_dev);
+	struct hid_sensor_hub_device *hsdev = rot_state->common_attributes.hsdev;
+	struct hid_sensor_hub_attribute_info *info = &rot_state->quaternion;
+	u32 usage_id = HID_USAGE_SENSOR_ORIENT_QUATERNION;
+	union {
+		s16 val16[4];
+		s32 val32[4];
+	} raw_buf;
 	int ret_type;
 	int i;
 
@@ -83,8 +90,37 @@ static int dev_rot_read_raw(struct iio_d
 	switch (mask) {
 	case IIO_CHAN_INFO_RAW:
 		if (size >= 4) {
-			for (i = 0; i < 4; ++i)
-				vals[i] = rot_state->scan.sampled_vals[i];
+			if (info->size <= 0 || info->size > sizeof(raw_buf))
+				return -EINVAL;
+
+			hid_sensor_power_state(&rot_state->common_attributes, true);
+
+			ret_type = sensor_hub_input_attr_read_values(hsdev,
+								     hsdev->usage,
+								     usage_id,
+								     info->report_id,
+								     SENSOR_HUB_SYNC,
+								     info->size,
+								     (u8 *)&raw_buf);
+
+			hid_sensor_power_state(&rot_state->common_attributes, false);
+
+			if (ret_type < 0)
+				return ret_type;
+
+			switch (info->size) {
+			case sizeof(raw_buf.val16):
+				for (i = 0; i < ARRAY_SIZE(raw_buf.val16); i++)
+					vals[i] = raw_buf.val16[i];
+				break;
+			case sizeof(raw_buf.val32):
+				for (i = 0; i < ARRAY_SIZE(raw_buf.val32); i++)
+					vals[i] = raw_buf.val32[i];
+				break;
+			default:
+				return -EINVAL;
+			}
+
 			ret_type = IIO_VAL_INT_MULTIPLE;
 			*val_len =  4;
 		} else
author maintainer
#1528 lore raw

PATCH 1522/1611 ALSA: scarlett2: Allow selecting config_set by firmware version

In reply to Greg Kroah-Hartman

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

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

From: "Geoffrey D. Bennett" <[email protected]>

[ Upstream commit 732a6397a526c025cd29c3c9309b0db6a2c08837 ]

The Scarlett 2i2 Gen 4 firmware 2417 moved the direct monitor gain parameters, so we now need to allow each device to list multiple scarlett2_config_set entries, one per applicable firmware version range, and pick the matching one at probe time.

No functional change yet: each device gets a single config_sets entry whose from_firmware_version matches the existing min_firmware_version (0 where none was set). This both prepares for selection and lets a follow-up commit remove the now-redundant min_firmware_version field.

scarlett2_count_io() depends on the resolved config_set so it moves out of scarlett2_init_private() into snd_scarlett2_controls_create() after the firmware version has been read.

Stable-dep-of: 3ca15754b561 ("ALSA: scarlett2: Update offsets for 2i2 Gen 4 firmware 2417")

---
 sound/usb/mixer_scarlett2.c |  143 +++++++++++++++++++++++++++++++++++++-------
 1 file changed, 121 insertions(+), 22 deletions(-)
sound/usb/mixer_scarlett2.c +121 -22
@@ -612,6 +612,20 @@ struct scarlett2_config_set {
 	const struct scarlett2_config items[SCARLETT2_CONFIG_COUNT];
 };
 
+/* Map firmware versions to config sets per-device.
+ *
+ * Each device lists one or more entries, sorted in ascending order of
+ * from_firmware_version. At probe time the running firmware version
+ * is looked up against this list and the last entry whose
+ * from_firmware_version is <= the running version is selected.
+ *
+ * The list is terminated by a sentinel entry with config_set == NULL.
+ */
+struct scarlett2_config_set_entry {
+	u16 from_firmware_version;
+	const struct scarlett2_config_set *config_set;
+};
+
 /* Input gain TLV dB ranges */
 
 static const DECLARE_TLV_DB_MINMAX(
@@ -1100,8 +1114,8 @@ struct scarlett2_meter_entry {
 };
 
 struct scarlett2_device_info {
-	/* which set of configuration parameters the device uses */
-	const struct scarlett2_config_set *config_set;
+	/* which sets of configuration parameters the device uses */
+	const struct scarlett2_config_set_entry *config_sets;
 
 	/* minimum firmware version required */
 	u16 min_firmware_version;
@@ -1343,7 +1357,10 @@ struct scarlett2_data {
 /*** Model-specific data ***/
 
 static const struct scarlett2_device_info s6i6_gen2_info = {
-	.config_set = &scarlett2_config_set_gen2a,
+	.config_sets = (const struct scarlett2_config_set_entry[]) {
+		{ 0, &scarlett2_config_set_gen2a },
+		{ }
+	},
 	.level_input_count = 2,
 	.pad_input_count = 2,
 
@@ -1393,7 +1410,10 @@ static const struct scarlett2_device_inf
 };
 
 static const struct scarlett2_device_info s18i8_gen2_info = {
-	.config_set = &scarlett2_config_set_gen2a,
+	.config_sets = (const struct scarlett2_config_set_entry[]) {
+		{ 0, &scarlett2_config_set_gen2a },
+		{ }
+	},
 	.level_input_count = 2,
 	.pad_input_count = 4,
 
@@ -1446,7 +1466,10 @@ static const struct scarlett2_device_inf
 };
 
 static const struct scarlett2_device_info s18i20_gen2_info = {
-	.config_set = &scarlett2_config_set_gen2b,
+	.config_sets = (const struct scarlett2_config_set_entry[]) {
+		{ 0, &scarlett2_config_set_gen2b },
+		{ }
+	},
 
 	.line_out_descrs = {
 		"Monitor L",
@@ -1503,7 +1526,10 @@ static const struct scarlett2_device_inf
 };
 
 static const struct scarlett2_device_info solo_gen3_info = {
-	.config_set = &scarlett2_config_set_gen3a,
+	.config_sets = (const struct scarlett2_config_set_entry[]) {
+		{ 0, &scarlett2_config_set_gen3a },
+		{ }
+	},
 	.level_input_count = 1,
 	.level_input_first = 1,
 	.air_input_count = 1,
@@ -1513,7 +1539,10 @@ static const struct scarlett2_device_inf
 };
 
 static const struct scarlett2_device_info s2i2_gen3_info = {
-	.config_set = &scarlett2_config_set_gen3a,
+	.config_sets = (const struct scarlett2_config_set_entry[]) {
+		{ 0, &scarlett2_config_set_gen3a },
+		{ }
+	},
 	.level_input_count = 2,
 	.air_input_count = 2,
 	.phantom_count = 1,
@@ -1522,7 +1551,10 @@ static const struct scarlett2_device_inf
 };
 
 static const struct scarlett2_device_info s4i4_gen3_info = {
-	.config_set = &scarlett2_config_set_gen3b,
+	.config_sets = (const struct scarlett2_config_set_entry[]) {
+		{ 0, &scarlett2_config_set_gen3b },
+		{ }
+	},
 	.level_input_count = 2,
 	.pad_input_count = 2,
 	.air_input_count = 2,
@@ -1571,7 +1603,10 @@ static const struct scarlett2_device_inf
 };
 
 static const struct scarlett2_device_info s8i6_gen3_info = {
-	.config_set = &scarlett2_config_set_gen3b,
+	.config_sets = (const struct scarlett2_config_set_entry[]) {
+		{ 0, &scarlett2_config_set_gen3b },
+		{ }
+	},
 	.level_input_count = 2,
 	.pad_input_count = 2,
 	.air_input_count = 2,
@@ -1637,7 +1672,10 @@ static const char * const scarlett2_spdi
 };
 
 static const struct scarlett2_device_info s18i8_gen3_info = {
-	.config_set = &scarlett2_config_set_gen3c,
+	.config_sets = (const struct scarlett2_config_set_entry[]) {
+		{ 0, &scarlett2_config_set_gen3c },
+		{ }
+	},
 	.has_speaker_switching = 1,
 	.level_input_count = 2,
 	.pad_input_count = 4,
@@ -1729,7 +1767,10 @@ static const char * const scarlett2_spdi
 };
 
 static const struct scarlett2_device_info s18i20_gen3_info = {
-	.config_set = &scarlett2_config_set_gen3c,
+	.config_sets = (const struct scarlett2_config_set_entry[]) {
+		{ 0, &scarlett2_config_set_gen3c },
+		{ }
+	},
 	.has_speaker_switching = 1,
 	.has_talkback = 1,
 	.level_input_count = 2,
@@ -1803,7 +1844,10 @@ static const struct scarlett2_device_inf
 };
 
 static const struct scarlett2_device_info vocaster_one_info = {
-	.config_set = &scarlett2_config_set_vocaster,
+	.config_sets = (const struct scarlett2_config_set_entry[]) {
+		{ 1769, &scarlett2_config_set_vocaster },
+		{ }
+	},
 	.min_firmware_version = 1769,
 	.has_devmap = 1,
 
@@ -1847,7 +1891,10 @@ static const struct scarlett2_device_inf
 };
 
 static const struct scarlett2_device_info vocaster_two_info = {
-	.config_set = &scarlett2_config_set_vocaster,
+	.config_sets = (const struct scarlett2_config_set_entry[]) {
+		{ 1769, &scarlett2_config_set_vocaster },
+		{ }
+	},
 	.min_firmware_version = 1769,
 	.has_devmap = 1,
 
@@ -1892,7 +1939,10 @@ static const struct scarlett2_device_inf
 };
 
 static const struct scarlett2_device_info solo_gen4_info = {
-	.config_set = &scarlett2_config_set_gen4_solo,
+	.config_sets = (const struct scarlett2_config_set_entry[]) {
+		{ 2115, &scarlett2_config_set_gen4_solo },
+		{ }
+	},
 	.min_firmware_version = 2115,
 	.has_devmap = 1,
 
@@ -1947,7 +1997,10 @@ static const struct scarlett2_device_inf
 };
 
 static const struct scarlett2_device_info s2i2_gen4_info = {
-	.config_set = &scarlett2_config_set_gen4_2i2,
+	.config_sets = (const struct scarlett2_config_set_entry[]) {
+		{ 2115, &scarlett2_config_set_gen4_2i2 },
+		{ }
+	},
 	.min_firmware_version = 2115,
 	.has_devmap = 1,
 
@@ -2002,7 +2055,10 @@ static const struct scarlett2_device_inf
 };
 
 static const struct scarlett2_device_info s4i4_gen4_info = {
-	.config_set = &scarlett2_config_set_gen4_4i4,
+	.config_sets = (const struct scarlett2_config_set_entry[]) {
+		{ 2089, &scarlett2_config_set_gen4_4i4 },
+		{ }
+	},
 	.min_firmware_version = 2089,
 	.has_devmap = 1,
 
@@ -2051,7 +2107,10 @@ static const struct scarlett2_device_inf
 };
 
 static const struct scarlett2_device_info clarett_2pre_info = {
-	.config_set = &scarlett2_config_set_clarett,
+	.config_sets = (const struct scarlett2_config_set_entry[]) {
+		{ 0, &scarlett2_config_set_clarett },
+		{ }
+	},
 	.level_input_count = 2,
 	.air_input_count = 2,
 
@@ -2107,7 +2166,10 @@ static const char * const scarlett2_spdi
 };
 
 static const struct scarlett2_device_info clarett_4pre_info = {
-	.config_set = &scarlett2_config_set_clarett,
+	.config_sets = (const struct scarlett2_config_set_entry[]) {
+		{ 0, &scarlett2_config_set_clarett },
+		{ }
+	},
 	.level_input_count = 2,
 	.air_input_count = 4,
 
@@ -2163,7 +2225,10 @@ static const struct scarlett2_device_inf
 };
 
 static const struct scarlett2_device_info clarett_8pre_info = {
-	.config_set = &scarlett2_config_set_clarett,
+	.config_sets = (const struct scarlett2_config_set_entry[]) {
+		{ 0, &scarlett2_config_set_clarett },
+		{ }
+	},
 	.level_input_count = 2,
 	.air_input_count = 8,
 
@@ -8211,10 +8276,32 @@ static void scarlett2_private_suspend(st
 
 /*** Initialisation ***/
 
+/* Select the config_set matching the running firmware version.
+ *
+ * The device info's config_sets array is ordered by ascending
+ * from_firmware_version; pick the last entry whose version is <= the
+ * running firmware version. If the running firmware is older than the
+ * first entry's from_firmware_version (i.e. older than the driver's
+ * minimum supported version for this device), the first entry's
+ * config_set is selected anyway so firmware updates can still be done
+ * (requires only the ACK handler), but the usual mixer controls
+ * aren't created.
+ */
+static void scarlett2_resolve_config_set(struct scarlett2_data *private)
+{
+	const struct scarlett2_config_set_entry *entry =
+		private->info->config_sets;
+
+	private->config_set = entry->config_set;
+	for (entry++; entry->config_set; entry++)
+		if (entry->from_firmware_version <= private->firmware_version)
+			private->config_set = entry->config_set;
+}
+
 static void scarlett2_count_io(struct scarlett2_data *private)
 {
 	const struct scarlett2_device_info *info = private->info;
-	const struct scarlett2_config_set *config_set = info->config_set;
+	const struct scarlett2_config_set *config_set = private->config_set;
 	const int (*port_count)[SCARLETT2_PORT_DIRNS] = info->port_count;
 	int port_type, srcs = 0, dsts = 0, i;
 
@@ -8311,9 +8398,14 @@ static int scarlett2_init_private(struct
 	mixer->private_suspend = scarlett2_private_suspend;
 
 	private->info = entry->info;
-	private->config_set = entry->info->config_set;
+
+	/* Set config_set to the first entry's config_set so the
+	 * notify handler has a valid pointer while USB init runs; it
+	 * is re-resolved once the firmware version has been read.
+	 */
+	private->config_set = entry->info->config_sets[0].config_set;
+
 	private->series_name = entry->series_name;
-	scarlett2_count_io(private);
 	private->scarlett2_seq = 0;
 	private->mixer = mixer;
 
@@ -8717,6 +8809,13 @@ static int snd_scarlett2_controls_create
 	if (err < 0)
 		return err;
 
+	/* Now that the firmware version is known, pick the matching
+	 * config_set
+	 */
+	scarlett2_resolve_config_set(private);
+
+	scarlett2_count_io(private);
+
 	/* Get the upgrade & settings flash segment numbers */
 	err = scarlett2_get_flash_segment_nums(mixer);
 	if (err < 0)
author maintainer
#1529 lore raw

PATCH 1523/1611 ALSA: scarlett2: Update offsets for 2i2 Gen 4 firmware 2417

In reply to Greg Kroah-Hartman

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

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

From: "Geoffrey D. Bennett" <[email protected]>

[ Upstream commit 3ca15754b561483aa7a1bce51677d6389f8ff5bb ]

Firmware 2417 for the Scarlett 4th Gen 2i2 moved the direct monitor gain parameters, so add a second config_set with the shifted offset and select it for firmware versions >= 2417.

Fixes 4e809a299677 ("ALSA: scarlett2: Add support for Solo, 2i2, and 4i4 Gen 4")
Cc [email protected] # ALSA: scarlett2: Allow selecting config_set by firmware version
Cc [email protected] # ALSA: scarlett2: Fold min_firmware_version into config_sets
---
 sound/usb/mixer_scarlett2.c |   58 ++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 58 insertions(+)
sound/usb/mixer_scarlett2.c +58 -0
@@ -937,6 +937,63 @@ static const struct scarlett2_config_set
 	}
 };
 
+/* 2i2 Gen 4, firmware version 2417 and above
+ *
+ * Firmware 2417 shifted DIRECT_MONITOR_GAIN by 4 bytes; all other
+ * offsets are unchanged from scarlett2_config_set_gen4_2i2.
+ */
+static const struct scarlett2_config_set scarlett2_config_set_gen4_2i2_2417 = {
+	.notifications = scarlett4_2i2_notifications,
+	.param_buf_addr = 0xfc,
+	.input_gain_tlv = db_scale_gen4_gain,
+	.autogain_status_texts = scarlett2_autogain_status_gen4,
+	.items = {
+		[SCARLETT2_CONFIG_MSD_SWITCH] = {
+			.offset = 0x49, .size = 8, .activate = 4 },
+
+		[SCARLETT2_CONFIG_DIRECT_MONITOR] = {
+			.offset = 0x14a, .size = 8, .activate = 16, .pbuf = 1 },
+
+		[SCARLETT2_CONFIG_AUTOGAIN_SWITCH] = {
+			.offset = 0x135, .size = 8, .activate = 10, .pbuf = 1 },
+
+		[SCARLETT2_CONFIG_AUTOGAIN_STATUS] = {
+			.offset = 0x137, .size = 8 },
+
+		[SCARLETT2_CONFIG_AG_MEAN_TARGET] = {
+			.offset = 0x131, .size = 8, .activate = 29, .pbuf = 1 },
+
+		[SCARLETT2_CONFIG_AG_PEAK_TARGET] = {
+			.offset = 0x132, .size = 8, .activate = 30, .pbuf = 1 },
+
+		[SCARLETT2_CONFIG_PHANTOM_SWITCH] = {
+			.offset = 0x48, .size = 8, .activate = 11, .pbuf = 1,
+			.mute = 1 },
+
+		[SCARLETT2_CONFIG_INPUT_GAIN] = {
+			.offset = 0x4b, .size = 8, .activate = 12, .pbuf = 1 },
+
+		[SCARLETT2_CONFIG_LEVEL_SWITCH] = {
+			.offset = 0x3c, .size = 8, .activate = 13, .pbuf = 1,
+			.mute = 1 },
+
+		[SCARLETT2_CONFIG_SAFE_SWITCH] = {
+			.offset = 0x147, .size = 8, .activate = 14, .pbuf = 1 },
+
+		[SCARLETT2_CONFIG_AIR_SWITCH] = {
+			.offset = 0x3e, .size = 8, .activate = 15, .pbuf = 1 },
+
+		[SCARLETT2_CONFIG_INPUT_SELECT_SWITCH] = {
+			.offset = 0x14b, .size = 8, .activate = 17, .pbuf = 1 },
+
+		[SCARLETT2_CONFIG_INPUT_LINK_SWITCH] = {
+			.offset = 0x14e, .size = 8, .activate = 18, .pbuf = 1 },
+
+		[SCARLETT2_CONFIG_DIRECT_MONITOR_GAIN] = {
+			.offset = 0x2a4, .size = 16, .activate = 36 }
+	}
+};
+
 /* 4i4 Gen 4 */
 static const struct scarlett2_config_set scarlett2_config_set_gen4_4i4 = {
 	.notifications = scarlett4_4i4_notifications,
@@ -1999,6 +2056,7 @@ static const struct scarlett2_device_inf
 static const struct scarlett2_device_info s2i2_gen4_info = {
 	.config_sets = (const struct scarlett2_config_set_entry[]) {
 		{ 2115, &scarlett2_config_set_gen4_2i2 },
+		{ 2417, &scarlett2_config_set_gen4_2i2_2417 },
 		{ }
 	},
 	.min_firmware_version = 2115,
author maintainer
#1530 lore raw

PATCH 1524/1611 firmware_loader: Add cancel helper for async requests

In reply to Greg Kroah-Hartman

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

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

From: Cássio Gabriel <[email protected]>

[ Upstream commit b9bdd68b8b979c7e9de58b2e7d21e1d7d932c755 ]

request_firmware_nowait() keeps the callback module pinned and holds a device reference until the firmware work completes.

Callers still have no way to cancel or synchronize the queued callback before tearing down their driver-private state.

Track scheduled async firmware work in an internal list and add request_firmware_nowait_cancel(). The helper cancels work matching the device, callback context and callback function. It cancels work that has not started yet and waits for an already-running callback to return. If the request has already completed, it is a no-op.

Keep the existing request_firmware_nowait() lifetime model manual. A devres-managed variant can be layered on top separately if needed.

Stable-dep-of: 5367e2ad14f0 ("ALSA: hda/tas2781: Cancel async firmware request at unbind")

---
 drivers/base/firmware_loader/main.c |   68 +++++++++++++++++++++++++++++++++---
 include/linux/firmware.h            |   10 +++++
 2 files changed, 74 insertions(+), 4 deletions(-)
drivers/base/firmware_loader/main.c +64 -4
@@ -1141,6 +1141,7 @@ EXPORT_SYMBOL(release_firmware);
 /* Async support */
 struct firmware_work {
 	struct work_struct work;
+	struct list_head list;
 	struct module *module;
 	const char *name;
 	struct device *device;
@@ -1149,6 +1150,17 @@ struct firmware_work {
 	u32 opt_flags;
 };
 
+static LIST_HEAD(firmware_work_list);
+static DEFINE_SPINLOCK(firmware_work_lock);
+
+static void firmware_work_free(struct firmware_work *fw_work)
+{
+	put_device(fw_work->device); /* taken in request_firmware_nowait() */
+	module_put(fw_work->module);
+	kfree_const(fw_work->name);
+	kfree(fw_work);
+}
+
 static void request_firmware_work_func(struct work_struct *work)
 {
 	struct firmware_work *fw_work;
@@ -1159,11 +1171,15 @@ static void request_firmware_work_func(s
 	_request_firmware(&fw, fw_work->name, fw_work->device, NULL, 0, 0,
 			  fw_work->opt_flags);
 	fw_work->cont(fw, fw_work->context);
-	put_device(fw_work->device); /* taken in request_firmware_nowait() */
 
-	module_put(fw_work->module);
-	kfree_const(fw_work->name);
-	kfree(fw_work);
+	spin_lock_irq(&firmware_work_lock);
+	if (!list_empty(&fw_work->list)) {
+		list_del_init(&fw_work->list);
+		spin_unlock_irq(&firmware_work_lock);
+		firmware_work_free(fw_work);
+		return;
+	}
+	spin_unlock_irq(&firmware_work_lock);
 }
 
 
@@ -1173,6 +1189,7 @@ static int _request_firmware_nowait(
 	void (*cont)(const struct firmware *fw, void *context), bool nowarn)
 {
 	struct firmware_work *fw_work;
+	unsigned long flags;
 
 	fw_work = kzalloc(sizeof(struct firmware_work), gfp);
 	if (!fw_work)
@@ -1205,7 +1222,12 @@ static int _request_firmware_nowait(
 
 	get_device(fw_work->device);
 	INIT_WORK(&fw_work->work, request_firmware_work_func);
+
+	spin_lock_irqsave(&firmware_work_lock, flags);
+	list_add_tail(&fw_work->list, &firmware_work_list);
 	schedule_work(&fw_work->work);
+	spin_unlock_irqrestore(&firmware_work_lock, flags);
+
 	return 0;
 }
 
@@ -1268,6 +1290,44 @@ int firmware_request_nowait_nowarn(
 }
 EXPORT_SYMBOL_GPL(firmware_request_nowait_nowarn);
 
+/**
+ * request_firmware_nowait_cancel() - cancel an async firmware request
+ * @device: device for which the firmware is being loaded
+ * @context: context passed to request_firmware_nowait()
+ * @cont: callback passed to request_firmware_nowait()
+ *
+ * Cancel a pending request_firmware_nowait() request for @device, @context
+ * and @cont. If the associated work has already started, this function waits
+ * until the callback has returned. If the callback has already completed, this
+ * function does nothing.
+ *
+ * This function may sleep.
+ */
+void request_firmware_nowait_cancel(struct device *device, void *context,
+				    void (*cont)(const struct firmware *fw,
+						 void *context))
+{
+	struct firmware_work *fw_work = NULL;
+	struct firmware_work *tmp;
+
+	spin_lock_irq(&firmware_work_lock);
+	list_for_each_entry_reverse(tmp, &firmware_work_list, list) {
+		if (tmp->device == device && tmp->context == context &&
+		    tmp->cont == cont) {
+			fw_work = tmp;
+			list_del_init(&fw_work->list);
+			break;
+		}
+	}
+	spin_unlock_irq(&firmware_work_lock);
+
+	if (!fw_work)
+		return;
+	cancel_work_sync(&fw_work->work);
+	firmware_work_free(fw_work);
+}
+EXPORT_SYMBOL_GPL(request_firmware_nowait_cancel);
+
 #ifdef CONFIG_FW_CACHE
 static ASYNC_DOMAIN_EXCLUSIVE(fw_cache_domain);
 
include/linux/firmware.h +10 -0
@@ -110,6 +110,9 @@ int request_firmware_nowait(
 	struct module *module, bool uevent,
 	const char *name, struct device *device, gfp_t gfp, void *context,
 	void (*cont)(const struct firmware *fw, void *context));
+void request_firmware_nowait_cancel(struct device *device, void *context,
+				    void (*cont)(const struct firmware *fw,
+						 void *context));
 int request_firmware_direct(const struct firmware **fw, const char *name,
 			    struct device *device);
 int request_firmware_into_buf(const struct firmware **firmware_p,
@@ -157,6 +160,13 @@ static inline int request_firmware_nowai
 	return -EINVAL;
 }
 
+static inline void request_firmware_nowait_cancel(struct device *device,
+						  void *context,
+						  void (*cont)(const struct firmware *fw,
+							       void *context))
+{
+}
+
 static inline void release_firmware(const struct firmware *fw)
 {
 }
author maintainer
#1531 lore raw

PATCH 1525/1611 ALSA: hda/tas2781: Cancel async firmware request at unbind

In reply to Greg Kroah-Hartman

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

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

From: Cássio Gabriel <[email protected]>

[ Upstream commit 5367e2ad14f0ae9350a7aaf2e77c87de39a43ae9 ]

TAS2781 HDA I2C and SPI queue RCA firmware loading from component bind with request_firmware_nowait(). The firmware loader keeps the callback module pinned and holds a device reference, but the callback still uses driver-private HDA state.

Component unbind removes controls and DSP state immediately. Later device removal tears down the TAS2781 private data, including codec_lock. If the async firmware callback runs after unbind has started, it can operate on state that is being torn down.

Cancel or synchronize the async firmware request before removing controls and DSP state. A queued callback is cancelled, and an already-running callback is allowed to finish before unbind continues.

---
 sound/hda/codecs/side-codecs/tas2781_hda_i2c.c |    3 +++
 sound/hda/codecs/side-codecs/tas2781_hda_spi.c |    3 +++
 2 files changed, 6 insertions(+)
sound/hda/codecs/side-codecs/tas2781_hda_i2c.c +3 -0
@@ -616,6 +616,9 @@ static void tas2781_hda_unbind(struct de
 		comp->playback_hook = NULL;
 	}
 
+	request_firmware_nowait_cancel(tas_hda->priv->dev, tas_hda->priv,
+				       tasdev_fw_ready);
+
 	tas2781_hda_remove_controls(tas_hda);
 
 	tasdevice_config_info_remove(tas_hda->priv);
sound/hda/codecs/side-codecs/tas2781_hda_spi.c +3 -0
@@ -753,6 +753,9 @@ static void tas2781_hda_unbind(struct de
 		comp->playback_hook = NULL;
 	}
 
+	request_firmware_nowait_cancel(tas_priv->dev, tas_priv,
+				       tasdev_fw_ready);
+
 	tas2781_hda_remove_controls(tas_hda);
 
 	tasdevice_config_info_remove(tas_priv);
author maintainer
#1532 lore raw

PATCH 1526/1611 vfio/mlx5: Fix racy bitfields and tighten struct layout

In reply to Greg Kroah-Hartman

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

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

From: Alex Williamson <[email protected]>

[ Upstream commit f2365a63b02ddea32e7db78b742c2503ec7b81f1 ]

Bitfield operations are not atomic, they use a read-modify-write pattern, therefore we should be careful not to pack bitfields that can be concurrently updated into the same storage unit.

This split takes a binary approach: flags that are only modified pre/post open/close remain bitfields, flags modified from user action, including actions that reach across to another device (ex. reset) use dedicated storage units.

Note mlx5_vhca_page_tracker.status is relocated to fill the alignment hole this split exposes.

Bitfield justifications:

migrate_cap: written only in mlx5vf_cmd_set_migratable() at probe
chunk_mode: written only in mlx5vf_cmd_set_migratable() at probe
mig_state_cap: written only in mlx5vf_cmd_set_migratable() at probe

Dedicated storage units:

mdev_detach: written in the VF attach/detach event notifier
mlx5fv_vf_event() at runtime
log_active: written in mlx5vf_start_page_tracker()/
mlx5vf_stop_page_tracker() during runtime dirty tracking
deferred_reset: written in mlx5vf_state_mutex_unlock()/
mlx5vf_pci_aer_reset_done() during runtime reset handling
is_err: set by tracker error handling and dirty-log polling at runtime
object_changed: set by tracker event handling and cleared by dirty-log
polling at runtime

Fixes 61a2f1460fd0 ("vfio/mlx5: Manage the VF attach/detach callback from the PF")
Fixes 79c3cf279926 ("vfio/mlx5: Init QP based resources for dirty tracking")
Fixes f886473071d6 ("vfio/mlx5: Add support for tracker object change event")
Cc Yishai Hadas <[email protected]>
Assisted-by Claude:claude-opus-4-8
---
 drivers/vfio/pci/mlx5/cmd.h |   15 +++++++++------
 1 file changed, 9 insertions(+), 6 deletions(-)
drivers/vfio/pci/mlx5/cmd.h +9 -6
@@ -155,25 +155,28 @@ struct mlx5_vhca_qp {
 struct mlx5_vhca_page_tracker {
 	u32 id;
 	u32 pdn;
-	u8 is_err:1;
-	u8 object_changed:1;
+	/* Flags modified at runtime - dedicated storage unit */
+	u8 is_err;
+	u8 object_changed;
+	int status;
 	struct mlx5_uars_page *uar;
 	struct mlx5_vhca_cq cq;
 	struct mlx5_vhca_qp *host_qp;
 	struct mlx5_vhca_qp *fw_qp;
 	struct mlx5_nb nb;
-	int status;
 };
 
 struct mlx5vf_pci_core_device {
 	struct vfio_pci_core_device core_device;
 	int vf_id;
 	u16 vhca_id;
+	/* Flags only modified on setup/release - bitfield ok */
 	u8 migrate_cap:1;
-	u8 deferred_reset:1;
-	u8 mdev_detach:1;
-	u8 log_active:1;
 	u8 chunk_mode:1;
+	/* Flags modified at runtime - dedicated storage unit */
+	u8 mdev_detach;
+	u8 log_active;
+	u8 deferred_reset;
 	struct completion tracker_comp;
 	/* protect migration state */
 	struct mutex state_mutex;
author maintainer
#1533 lore raw

PATCH 1527/1611 binder: Use LIST_HEAD() to initialize on stack list head

In reply to Greg Kroah-Hartman

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

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

From: Jisheng Zhang <[email protected]>

[ Upstream commit 91f818b184b44b105b1c92859ea8d2d6f47912a9 ]

Use LIST_HEAD to initialize on stack list head. No intentional functional impact.

Change generated with below coccinelle script:

@@
identifier name;
@@
- struct list_head name;
+ LIST_HEAD(name);
... when != name
- INIT_LIST_HEAD(&name);

Stable-dep-of: b34826e55aad ("binder: cache secctx size before release zeroes it")

---
 drivers/android/binder.c |    6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)
drivers/android/binder.c +2 -4
@@ -3097,12 +3097,10 @@ static void binder_transaction(struct bi
 	int t_debug_id = atomic_inc_return(&binder_last_id);
 	ktime_t t_start_time = ktime_get();
 	struct lsm_context lsmctx = { };
-	struct list_head sgc_head;
-	struct list_head pf_head;
+	LIST_HEAD(sgc_head);
+	LIST_HEAD(pf_head);
 	const void __user *user_buffer = (const void __user *)
 				(uintptr_t)tr->data.ptr.buffer;
-	INIT_LIST_HEAD(&sgc_head);
-	INIT_LIST_HEAD(&pf_head);
 
 	e = binder_transaction_log_add(&binder_transaction_log);
 	e->debug_id = t_debug_id;
author maintainer
#1534 lore raw

PATCH 1528/1611 binder: cache secctx size before release zeroes it

In reply to Greg Kroah-Hartman

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

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

From: Chris Mason <[email protected]>

[ Upstream commit b34826e55aad3520ec813f1f367c11b24b29dc9f ]

binder_transaction() bounds the scatter-gather buffer area with sg_buf_end_offset and subtracts the aligned LSM context size because the secctx is written at the tail of that area. The subtraction reads lsmctx.len, but that field has already been cleared by the time the line runs:

security_secid_to_secctx(secid, &lsmctx) /* lsmctx.len set */
lsmctx_aligned_size = ALIGN(lsmctx.len, sizeof(u64)) extra_buffers_size += lsmctx_aligned_size
...
security_release_secctx(&lsmctx) /* memset zeroes len */
...
sg_buf_end_offset = sg_buf_offset + extra_buffers_size
- ALIGN(lsmctx.len, sizeof(u64)) /* ALIGN(0,8) */

security_release_secctx() does memset(cp, 0, sizeof(*cp)), so lsmctx.len reads back as 0 and the subtraction contributes nothing, leaving sg_buf_end_offset too large by the aligned secctx size on every transaction to a txn_security_ctx node.

Each BINDER_TYPE_PTR object then derives buf_left = sg_buf_end_offset - sg_buf_offset as the sole upper bound on its copy, so the inflated end offset lets the copy run into the bytes that already hold the secctx.

The aligned size must therefore be cached before release rather than re-read from the now-cleared field. Fix by caching it in lsmctx_aligned_size at function scope when it is first computed and subtracting lsmctx_aligned_size instead of re-reading lsmctx.len after release. Reuse the same value for the earlier buf_offset computation.

Fixes 6fba89813ccf ("lsm: ensure the correct LSM context releaser")
Cc stable <[email protected]>
Assisted-by kres:claude-opus-4-8
---
 drivers/android/binder.c |   12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)
drivers/android/binder.c +6 -6
@@ -3097,6 +3097,7 @@ static void binder_transaction(struct bi
 	int t_debug_id = atomic_inc_return(&binder_last_id);
 	ktime_t t_start_time = ktime_get();
 	struct lsm_context lsmctx = { };
+	size_t lsmctx_aligned_size = 0;
 	LIST_HEAD(sgc_head);
 	LIST_HEAD(pf_head);
 	const void __user *user_buffer = (const void __user *)
@@ -3363,7 +3364,6 @@ static void binder_transaction(struct bi
 
 	if (target_node && target_node->txn_security_ctx) {
 		u32 secid;
-		size_t added_size;
 
 		security_cred_getsecid(proc->cred, &secid);
 		ret = security_secid_to_secctx(secid, &lsmctx);
@@ -3375,9 +3375,9 @@ static void binder_transaction(struct bi
 			return_error_line = __LINE__;
 			goto err_get_secctx_failed;
 		}
-		added_size = ALIGN(lsmctx.len, sizeof(u64));
-		extra_buffers_size += added_size;
-		if (extra_buffers_size < added_size) {
+		lsmctx_aligned_size = ALIGN(lsmctx.len, sizeof(u64));
+		extra_buffers_size += lsmctx_aligned_size;
+		if (extra_buffers_size < lsmctx_aligned_size) {
 			binder_txn_error("%d:%d integer overflow of extra_buffers_size\n",
 				thread->pid, proc->pid);
 			return_error = BR_FAILED_REPLY;
@@ -3414,7 +3414,7 @@ static void binder_transaction(struct bi
 		size_t buf_offset = ALIGN(tr->data_size, sizeof(void *)) +
 				    ALIGN(tr->offsets_size, sizeof(void *)) +
 				    ALIGN(extra_buffers_size, sizeof(void *)) -
-				    ALIGN(lsmctx.len, sizeof(u64));
+				    lsmctx_aligned_size;
 
 		t->security_ctx = t->buffer->user_data + buf_offset;
 		err = binder_alloc_copy_to_buffer(&target_proc->alloc,
@@ -3469,7 +3469,7 @@ static void binder_transaction(struct bi
 	off_end_offset = off_start_offset + tr->offsets_size;
 	sg_buf_offset = ALIGN(off_end_offset, sizeof(void *));
 	sg_buf_end_offset = sg_buf_offset + extra_buffers_size -
-		ALIGN(lsmctx.len, sizeof(u64));
+		lsmctx_aligned_size;
 	off_min = 0;
 	for (buffer_offset = off_start_offset; buffer_offset < off_end_offset;
 	     buffer_offset += sizeof(binder_size_t)) {
author maintainer
#1535 lore raw

PATCH 1529/1611 PCI: imx6: Fix reference clock source selection for i.MX95

In reply to Greg Kroah-Hartman

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

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

From: Franz Schnyder <[email protected]>

[ Upstream commit 88cc4cbe08bba27bb58888d25d336774aa0ccab1 ]

In the PCIe PHY init for the i.MX95, the reference clock source selection uses a conditional instead of always passing the mask. This currently breaks functionality if the internal refclk is used.

To fix this issue, always pass IMX95_PCIE_REF_USE_PAD as the mask and clear bit if external refclk is not used. This essentially swaps the parameters.

Stable-dep-of: 0c26b1c34d12 ("PCI: imx6: Configure REF_USE_PAD before PHY reset for i.MX95")

---
 drivers/pci/controller/dwc/pci-imx6.c |   14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)
drivers/pci/controller/dwc/pci-imx6.c +8 -6
@@ -150,6 +150,7 @@ struct imx_lut_data {
 struct imx_pcie {
 	struct dw_pcie		*pci;
 	struct gpio_desc	*reset_gpiod;
+	bool			enable_ext_refclk;
 	struct clk_bulk_data	*clks;
 	int			num_clks;
 	struct regmap		*iomuxc_gpr;
@@ -244,6 +245,8 @@ static unsigned int imx_pcie_grp_offset(
 
 static int imx95_pcie_init_phy(struct imx_pcie *imx_pcie)
 {
+	bool ext = imx_pcie->enable_ext_refclk;
+
 	/*
 	 * ERR051624: The Controller Without Vaux Cannot Exit L23 Ready
 	 * Through Beacon or PERST# De-assertion
@@ -262,13 +265,12 @@ static int imx95_pcie_init_phy(struct im
 			IMX95_PCIE_PHY_CR_PARA_SEL,
 			IMX95_PCIE_PHY_CR_PARA_SEL);
 
-	regmap_update_bits(imx_pcie->iomuxc_gpr,
-			   IMX95_PCIE_PHY_GEN_CTRL,
-			   IMX95_PCIE_REF_USE_PAD, 0);
-	regmap_update_bits(imx_pcie->iomuxc_gpr,
-			   IMX95_PCIE_SS_RW_REG_0,
+	regmap_update_bits(imx_pcie->iomuxc_gpr, IMX95_PCIE_PHY_GEN_CTRL,
+			   IMX95_PCIE_REF_USE_PAD,
+			   ext ? IMX95_PCIE_REF_USE_PAD : 0);
+	regmap_update_bits(imx_pcie->iomuxc_gpr, IMX95_PCIE_SS_RW_REG_0,
 			   IMX95_PCIE_REF_CLKEN,
-			   IMX95_PCIE_REF_CLKEN);
+			   ext ? 0 : IMX95_PCIE_REF_CLKEN);
 
 	return 0;
 }
author maintainer
#1536 lore raw

PATCH 1530/1611 PCI: imx6: Configure REF_USE_PAD before PHY reset for i.MX95

In reply to Greg Kroah-Hartman

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

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

From: Richard Zhu <[email protected]>

[ Upstream commit 0c26b1c34d12d4debfb5363cc0be6cdf68e87ba2 ]

According to the i.MX95 PCIe PHY Databook, the ref_use_pad signal in the Common Block Signals section selects the reference clock source connected to the PHY pads. Per the specification, any change to this input must be followed by a PHY reset assertion to take effect.

Move the REF_USE_PAD configuration before the PHY reset toggle to comply with the required initialization sequence.

Fixes 47f54a902dcd ("PCI: imx6: Toggle the core reset for i.MX95 PCIe")
[mani: renamed the callback and helper to match the usecase]
---
 drivers/pci/controller/dwc/pci-imx6.c |   27 ++++++++++++++++++++++++---
 1 file changed, 24 insertions(+), 3 deletions(-)
drivers/pci/controller/dwc/pci-imx6.c +24 -3
@@ -135,6 +135,7 @@ struct imx_pcie_drvdata {
 	const u32 mode_off[IMX_PCIE_MAX_INSTANCES];
 	const u32 mode_mask[IMX_PCIE_MAX_INSTANCES];
 	const struct pci_epc_features *epc_features;
+	int (*select_ref_clk_src)(struct imx_pcie *pcie);
 	int (*init_phy)(struct imx_pcie *pcie);
 	int (*enable_ref_clk)(struct imx_pcie *pcie, bool enable);
 	int (*core_reset)(struct imx_pcie *pcie, bool assert);
@@ -243,6 +244,24 @@ static unsigned int imx_pcie_grp_offset(
 	return imx_pcie->controller_id == 1 ? IOMUXC_GPR16 : IOMUXC_GPR14;
 }
 
+static int imx95_pcie_select_ref_clk_src(struct imx_pcie *imx_pcie)
+{
+	bool ext = imx_pcie->enable_ext_refclk;
+
+	/*
+	 * Regarding the Signal Descriptions of i.MX95 PCIe PHY, ref_use_pad is
+	 * used to select reference clock connected to a pair of pads.
+	 *
+	 * Any change in this input must be followed by phy_reset assertion.
+	 */
+
+	regmap_update_bits(imx_pcie->iomuxc_gpr, IMX95_PCIE_PHY_GEN_CTRL,
+			   IMX95_PCIE_REF_USE_PAD,
+			   ext ? IMX95_PCIE_REF_USE_PAD : 0);
+
+	return 0;
+}
+
 static int imx95_pcie_init_phy(struct imx_pcie *imx_pcie)
 {
 	bool ext = imx_pcie->enable_ext_refclk;
@@ -265,9 +284,6 @@ static int imx95_pcie_init_phy(struct im
 			IMX95_PCIE_PHY_CR_PARA_SEL,
 			IMX95_PCIE_PHY_CR_PARA_SEL);
 
-	regmap_update_bits(imx_pcie->iomuxc_gpr, IMX95_PCIE_PHY_GEN_CTRL,
-			   IMX95_PCIE_REF_USE_PAD,
-			   ext ? IMX95_PCIE_REF_USE_PAD : 0);
 	regmap_update_bits(imx_pcie->iomuxc_gpr, IMX95_PCIE_SS_RW_REG_0,
 			   IMX95_PCIE_REF_CLKEN,
 			   ext ? 0 : IMX95_PCIE_REF_CLKEN);
@@ -1232,6 +1248,9 @@ static int imx_pcie_host_init(struct dw_
 		pp->bridge->disable_device = imx_pcie_disable_device;
 	}
 
+	if (imx_pcie->drvdata->select_ref_clk_src)
+		imx_pcie->drvdata->select_ref_clk_src(imx_pcie);
+
 	imx_pcie_assert_core_reset(imx_pcie);
 
 	if (imx_pcie->drvdata->init_phy)
@@ -1932,6 +1951,7 @@ static const struct imx_pcie_drvdata drv
 		.mode_mask[0] = IMX95_PCIE_DEVICE_TYPE,
 		.core_reset = imx95_pcie_core_reset,
 		.init_phy = imx95_pcie_init_phy,
+		.select_ref_clk_src = imx95_pcie_select_ref_clk_src,
 		.wait_pll_lock = imx95_pcie_wait_for_phy_pll_lock,
 		.enable_ref_clk = imx95_pcie_enable_ref_clk,
 	},
@@ -1986,6 +2006,7 @@ static const struct imx_pcie_drvdata drv
 		.ltssm_mask = IMX95_PCIE_LTSSM_EN,
 		.mode_off[0]  = IMX95_PE0_GEN_CTRL_1,
 		.mode_mask[0] = IMX95_PCIE_DEVICE_TYPE,
+		.select_ref_clk_src = imx95_pcie_select_ref_clk_src,
 		.init_phy = imx95_pcie_init_phy,
 		.core_reset = imx95_pcie_core_reset,
 		.wait_pll_lock = imx95_pcie_wait_for_phy_pll_lock,
author maintainer
#1537 lore raw

PATCH 1531/1611 PCI/IOV: Adjust ->barsz[] when changing BAR size

In reply to Greg Kroah-Hartman

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

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

From: Ilpo Järvinen <[email protected]>

[ Upstream commit 4687b3315a3f76647746f5b7f92684cf1045b085 ]

pci_rebar_set_size() adjusts BAR size for both normal and IOV BARs. The struct pci_sriov keeps a cached copy of BAR size in ->barsz[] which is not adjusted by pci_rebar_set_size() but by pci_iov_resource_set_size(). pci_iov_resource_set_size() is called also from pci_resize_resource_set_size().

The current arrangement is problematic once BAR resize algorithm starts to roll back changes properly in case of a failure. The normal resource fitting algorithm rolls back resource size using the struct pci_dev_resource easily but also calling pci_resize_resource_set_size() or pci_iov_resource_set_size() to roll back BAR size would be an extra burden, whereas combining ->barsz[] update with pci_rebar_set_size() naturally rolls back it when restoring the old BAR size on a different layer of the BAR resize operation.

Thus, rework pci_rebar_set_size() to also update ->barsz[].

Stable-dep-of: ee7471fe968d ("PCI: Skip Resizable BAR restore on read error")

---
 drivers/pci/iov.c       |   15 ++++-----------
 drivers/pci/pci.c       |    4 ++++
 drivers/pci/pci.h       |    5 ++---
 drivers/pci/setup-res.c |   10 ++++------
 4 files changed, 14 insertions(+), 20 deletions(-)
drivers/pci/iov.c +4 -11
@@ -158,8 +158,7 @@ resource_size_t pci_iov_resource_size(st
 	return dev->sriov->barsz[pci_resource_num_to_vf_bar(resno)];
 }
 
-void pci_iov_resource_set_size(struct pci_dev *dev, int resno,
-			       resource_size_t size)
+void pci_iov_resource_set_size(struct pci_dev *dev, int resno, int size)
 {
 	if (!pci_resource_is_iov(resno)) {
 		pci_warn(dev, "%s is not an IOV resource\n",
@@ -167,7 +166,8 @@ void pci_iov_resource_set_size(struct pc
 		return;
 	}
 
-	dev->sriov->barsz[pci_resource_num_to_vf_bar(resno)] = size;
+	resno = pci_resource_num_to_vf_bar(resno);
+	dev->sriov->barsz[resno] = pci_rebar_size_to_bytes(size);
 }
 
 bool pci_iov_is_memory_decoding_enabled(struct pci_dev *dev)
@@ -1345,7 +1345,6 @@ EXPORT_SYMBOL_GPL(pci_sriov_configure_si
 int pci_iov_vf_bar_set_size(struct pci_dev *dev, int resno, int size)
 {
 	u32 sizes;
-	int ret;
 
 	if (!pci_resource_is_iov(resno))
 		return -EINVAL;
@@ -1360,13 +1359,7 @@ int pci_iov_vf_bar_set_size(struct pci_d
 	if (!(sizes & BIT(size)))
 		return -EINVAL;
 
-	ret = pci_rebar_set_size(dev, resno, size);
-	if (ret)
-		return ret;
-
-	pci_iov_resource_set_size(dev, resno, pci_rebar_size_to_bytes(size));
-
-	return 0;
+	return pci_rebar_set_size(dev, resno, size);
 }
 EXPORT_SYMBOL_GPL(pci_iov_vf_bar_set_size);
 
drivers/pci/pci.c +4 -0
@@ -3797,6 +3797,10 @@ int pci_rebar_set_size(struct pci_dev *p
 	ctrl &= ~PCI_REBAR_CTRL_BAR_SIZE;
 	ctrl |= FIELD_PREP(PCI_REBAR_CTRL_BAR_SIZE, size);
 	pci_write_config_dword(pdev, pos + PCI_REBAR_CTRL, ctrl);
+
+	if (pci_resource_is_iov(bar))
+		pci_iov_resource_set_size(pdev, bar, size);
+
 	return 0;
 }
 
drivers/pci/pci.h +2 -3
@@ -826,8 +826,7 @@ void pci_iov_update_resource(struct pci_
 resource_size_t pci_sriov_resource_alignment(struct pci_dev *dev, int resno);
 void pci_restore_iov_state(struct pci_dev *dev);
 int pci_iov_bus_range(struct pci_bus *bus);
-void pci_iov_resource_set_size(struct pci_dev *dev, int resno,
-			       resource_size_t size);
+void pci_iov_resource_set_size(struct pci_dev *dev, int resno, int size);
 bool pci_iov_is_memory_decoding_enabled(struct pci_dev *dev);
 static inline u16 pci_iov_vf_rebar_cap(struct pci_dev *dev)
 {
@@ -869,7 +868,7 @@ static inline int pci_iov_bus_range(stru
 	return 0;
 }
 static inline void pci_iov_resource_set_size(struct pci_dev *dev, int resno,
-					     resource_size_t size) { }
+					     int size) { }
 static inline bool pci_iov_is_memory_decoding_enabled(struct pci_dev *dev)
 {
 	return false;
drivers/pci/setup-res.c +4 -6
@@ -452,12 +452,10 @@ static void pci_resize_resource_set_size
 	resource_size_t res_size = pci_rebar_size_to_bytes(size);
 	struct resource *res = pci_resource_n(dev, resno);
 
-	if (!pci_resource_is_iov(resno)) {
-		resource_set_size(res, res_size);
-	} else {
-		resource_set_size(res, res_size * pci_sriov_get_totalvfs(dev));
-		pci_iov_resource_set_size(dev, resno, res_size);
-	}
+	if (pci_resource_is_iov(resno))
+		res_size *= pci_sriov_get_totalvfs(dev);
+
+	resource_set_size(res, res_size);
 }
 
 int pci_resize_resource(struct pci_dev *dev, int resno, int size)
author maintainer
#1538 lore raw

PATCH 1532/1611 PCI: Change pci_dev variable from bridge to dev

In reply to Greg Kroah-Hartman

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

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

From: Ilpo Järvinen <[email protected]>

[ Upstream commit 34c702ea0497e092a487eacdf28a037f43e3e39f ]

Upcoming fix to BAR resize will store also device BAR resource in the saved list. Change the pci_dev variable in the loop from 'bridge' to 'dev' as the former would be misleading with non-bridges in the list.

This is in a separate change to reduce churn in the upcoming BAR resize fix.

While it appears that the logic in the loop doing pci_setup_bridge() is altered as 'bridge' variable is no longer updated, a bridge should never appear more than once in the saved list so the check can only match to the first entry. As such, the code with two distinct pci_dev variables better represents the intention of the check compared with the old code where bridge variable was reused for a different purpose.

Stable-dep-of: ee7471fe968d ("PCI: Skip Resizable BAR restore on read error")

---
 drivers/pci/setup-bus.c |   17 +++++++++--------
 1 file changed, 9 insertions(+), 8 deletions(-)
drivers/pci/setup-bus.c +9 -8
@@ -2431,12 +2431,13 @@ int pbus_reassign_bridge_resources(struc
 	}
 
 	list_for_each_entry(dev_res, &saved, list) {
+		struct pci_dev *dev = dev_res->dev;
+
 		/* Skip the bridge we just assigned resources for */
-		if (bridge == dev_res->dev)
+		if (bridge == dev)
 			continue;
 
-		bridge = dev_res->dev;
-		pci_setup_bridge(bridge->subordinate);
+		pci_setup_bridge(dev->subordinate);
 	}
 
 	free_list(&saved);
@@ -2452,19 +2453,19 @@ cleanup:
 	/* Revert to the old configuration */
 	list_for_each_entry(dev_res, &saved, list) {
 		struct resource *res = dev_res->res;
+		struct pci_dev *dev = dev_res->dev;
 
-		bridge = dev_res->dev;
-		i = pci_resource_num(bridge, res);
+		i = pci_resource_num(dev, res);
 
 		if (res->parent) {
 			release_child_resources(res);
-			pci_release_resource(bridge, i);
+			pci_release_resource(dev, i);
 		}
 
 		restore_dev_resource(dev_res);
 
-		pci_claim_resource(bridge, i);
-		pci_setup_bridge(bridge->subordinate);
+		pci_claim_resource(dev, i);
+		pci_setup_bridge(dev->subordinate);
 	}
 	free_list(&saved);
 	up_read(&pci_bus_sem);
author maintainer
#1539 lore raw

PATCH 1533/1611 PCI: Try BAR resize even when no window was released

In reply to Greg Kroah-Hartman

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

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

From: Ilpo Järvinen <[email protected]>

[ Upstream commit 121d3e9e4b217f2f23715901fb15c6a3ca2bab46 ]

Usually, resizing BARs requires releasing bridge windows in order to resize it to fit a larger BAR into the window. This is not always the case, however, FW could have made the window large enough to accommodate larger BAR as is, or the user might prefer to shrink a BAR to make more space for another Resizable BAR.

Thus, replace the check that requires that at least one bridge window was released with a check that simply ensures bridge is not NULL.

Stable-dep-of: ee7471fe968d ("PCI: Skip Resizable BAR restore on read error")

---
 drivers/pci/setup-bus.c |    6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)
drivers/pci/setup-bus.c +2 -4
@@ -2376,7 +2376,7 @@ int pbus_reassign_bridge_resources(struc
 {
 	unsigned long type = res->flags;
 	struct pci_dev_resource *dev_res;
-	struct pci_dev *bridge;
+	struct pci_dev *bridge = NULL;
 	LIST_HEAD(saved);
 	LIST_HEAD(added);
 	LIST_HEAD(failed);
@@ -2411,10 +2411,8 @@ int pbus_reassign_bridge_resources(struc
 		bus = bus->parent;
 	}
 
-	if (list_empty(&saved)) {
-		up_read(&pci_bus_sem);
+	if (!bridge)
 		return -ENOENT;
-	}
 
 	__pci_bus_size_bridges(bridge->subordinate, &added);
 	__pci_bridge_assign_resources(bridge, &added, &failed);
author maintainer
#1540 lore raw

PATCH 1534/1611 PCI: Free saved list without holding pci_bus_sem

In reply to Greg Kroah-Hartman

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

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

From: Ilpo Järvinen <[email protected]>

[ Upstream commit 1d8a0506f69895b7cfd9d5c4546761c508231a8a ]

Freeing the saved list does not require holding pci_bus_sem, so the critical section can be made shorter.

Stable-dep-of: ee7471fe968d ("PCI: Skip Resizable BAR restore on read error")

---
 drivers/pci/setup-bus.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/pci/setup-bus.c +1 -1
@@ -2465,8 +2465,8 @@ cleanup:
 		pci_claim_resource(dev, i);
 		pci_setup_bridge(dev->subordinate);
 	}
-	free_list(&saved);
 	up_read(&pci_bus_sem);
+	free_list(&saved);
 
 	return ret;
 }
author maintainer
#1541 lore raw

PATCH 1535/1611 PCI: Fix restoring BARs on BAR resize rollback path

In reply to Greg Kroah-Hartman

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

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

From: Ilpo Järvinen <[email protected]>

[ Upstream commit 337b1b566db087347194e4543ddfdfa5645275cc ]

BAR resize operation is implemented in the pci_resize_resource() and pbus_reassign_bridge_resources() functions. pci_resize_resource() can be called either from __resource_resize_store() from sysfs or directly by the driver for the Endpoint Device.

The pci_resize_resource() requires that caller has released the device resources that share the bridge window with the BAR to be resized as otherwise the bridge window is pinned in place and cannot be changed.

pbus_reassign_bridge_resources() rolls back resources if the resize operation fails, but rollback is performed only for the bridge windows. Because releasing the device resources are done by the caller of the BAR resize interface, these functions performing the BAR resize do not have access to the device resources as they were before the resize.

pbus_reassign_bridge_resources() could try __pci_bridge_assign_resources() after rolling back the bridge windows as they were, however, it will not guarantee the resource are assigned due to differences in how FW and the kernel assign the resources (alignment of the start address and tail).

To perform rollback robustly, the BAR resize interface has to be altered to also release the device resources that share the bridge window with the BAR to be resized.

Also, remove restoring from the entries failed list as saved list should now contain both the bridge windows and device resources so the extra restore is duplicated work.

Some drivers (currently only amdgpu) want to prevent releasing some resources. Add exclude_bars param to pci_resize_resource() and make amdgpu pass its register BAR (BAR 2 or 5), which should never be released during resize operation. Normally 64-bit prefetchable resources do not share a bridge window with the 32-bit only register BAR, but there are various fallbacks in the resource assignment logic which may make the resources share the bridge window in rare cases.

This change (together with the driver side changes) is to counter the resource releases that had to be done to prevent resource tree corruption in the ("PCI: Release assigned resource before restoring them") change. As such, it likely restores functionality in cases where device resources were released to avoid resource tree conflicts which appeared to be "working" when such conflicts were not correctly detected by the kernel.

[bhelgaas: squash amdgpu BAR selection from
https: //lore.kernel.org/r/[email protected]]

Stable-dep-of: ee7471fe968d ("PCI: Skip Resizable BAR restore on read error")

---
 drivers/gpu/drm/amd/amdgpu/amdgpu_device.c  |    4 -
 drivers/gpu/drm/i915/gt/intel_region_lmem.c |    2 
 drivers/gpu/drm/xe/xe_vram.c                |    2 
 drivers/pci/pci-sysfs.c                     |   17 ----
 drivers/pci/pci.h                           |    4 -
 drivers/pci/setup-bus.c                     |  100 +++++++++++++++++++++-------
 drivers/pci/setup-res.c                     |   23 +-----
 include/linux/pci.h                         |    3 
 8 files changed, 93 insertions(+), 62 deletions(-)
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +3 -1
@@ -1736,7 +1736,9 @@ int amdgpu_device_resize_fb_bar(struct a
 
 	pci_release_resource(adev->pdev, 0);
 
-	r = pci_resize_resource(adev->pdev, 0, rbar_size);
+	r = pci_resize_resource(adev->pdev, 0, rbar_size,
+				(adev->asic_type >= CHIP_BONAIRE) ? 1 << 5
+								  : 1 << 2);
 	if (r == -ENOSPC)
 		dev_info(adev->dev,
 			 "Not enough PCI address space for a large BAR.");
drivers/gpu/drm/i915/gt/intel_region_lmem.c +1 -1
@@ -37,7 +37,7 @@ _resize_bar(struct drm_i915_private *i91
 
 	_release_bars(pdev);
 
-	ret = pci_resize_resource(pdev, resno, bar_size);
+	ret = pci_resize_resource(pdev, resno, bar_size, 0);
 	if (ret) {
 		drm_info(&i915->drm, "Failed to resize BAR%d to %dM (%pe)\n",
 			 resno, 1 << bar_size, ERR_PTR(ret));
drivers/gpu/drm/xe/xe_vram.c +1 -1
@@ -56,7 +56,7 @@ static void resize_bar(struct xe_device
 
 	release_bars(pdev);
 
-	ret = pci_resize_resource(pdev, resno, bar_size);
+	ret = pci_resize_resource(pdev, resno, bar_size, 0);
 	if (ret) {
 		drm_info(&xe->drm, "Failed to resize BAR%d to %dM (%pe). Consider enabling 'Resizable BAR' support in your BIOS\n",
 			 resno, 1 << bar_size, ERR_PTR(ret));
drivers/pci/pci-sysfs.c +2 -15
@@ -1571,18 +1571,13 @@ static ssize_t __resource_resize_store(s
 {
 	struct pci_dev *pdev = to_pci_dev(dev);
 	struct pci_bus *bus = pdev->bus;
-	struct resource *b_win, *res;
 	unsigned long size;
-	int ret, i;
+	int ret;
 	u16 cmd;
 
 	if (kstrtoul(buf, 0, &size) < 0)
 		return -EINVAL;
 
-	b_win = pbus_select_window(bus, pci_resource_n(pdev, n));
-	if (!b_win)
-		return -EINVAL;
-
 	device_lock(dev);
 	if (dev->driver || pci_num_vf(pdev)) {
 		ret = -EBUSY;
@@ -1604,15 +1599,7 @@ static ssize_t __resource_resize_store(s
 
 	pci_remove_resource_files(pdev);
 
-	pci_dev_for_each_resource(pdev, res, i) {
-		if (i >= PCI_BRIDGE_RESOURCES)
-			break;
-
-		if (b_win == pbus_select_window(bus, res))
-			pci_release_resource(pdev, i);
-	}
-
-	ret = pci_resize_resource(pdev, n, size);
+	ret = pci_resize_resource(pdev, n, size, 0);
 
 	pci_assign_unassigned_bus_resources(bus);
 
drivers/pci/pci.h +3 -1
@@ -439,8 +439,10 @@ enum pci_bar_type {
 struct device *pci_get_host_bridge_device(struct pci_dev *dev);
 void pci_put_host_bridge_device(struct device *dev);
 
+void pci_resize_resource_set_size(struct pci_dev *dev, int resno, int size);
+int pci_do_resource_release_and_resize(struct pci_dev *dev, int resno, int size,
+				       int exclude_bars);
 unsigned int pci_rescan_bus_bridge_resize(struct pci_dev *bridge);
-int pbus_reassign_bridge_resources(struct pci_bus *bus, struct resource *res);
 int __must_check pci_reassign_resource(struct pci_dev *dev, int i, resource_size_t add_size, resource_size_t align);
 
 int pci_configure_extended_tags(struct pci_dev *dev, void *ign);
drivers/pci/setup-bus.c +75 -25

149 lines not on this page · raw message

drivers/pci/setup-res.c +6 -17
@@ -446,8 +446,7 @@ static bool pci_resize_is_memory_decodin
 	return cmd & PCI_COMMAND_MEMORY;
 }
 
-static void pci_resize_resource_set_size(struct pci_dev *dev, int resno,
-					 int size)
+void pci_resize_resource_set_size(struct pci_dev *dev, int resno, int size)
 {
 	resource_size_t res_size = pci_rebar_size_to_bytes(size);
 	struct resource *res = pci_resource_n(dev, resno);
@@ -458,9 +457,9 @@ static void pci_resize_resource_set_size
 	resource_set_size(res, res_size);
 }
 
-int pci_resize_resource(struct pci_dev *dev, int resno, int size)
+int pci_resize_resource(struct pci_dev *dev, int resno, int size,
+			int exclude_bars)
 {
-	struct resource *res = pci_resource_n(dev, resno);
 	struct pci_host_bridge *host;
 	int old, ret;
 	u32 sizes;
@@ -470,10 +469,6 @@ int pci_resize_resource(struct pci_dev *
 	if (host->preserve_config)
 		return -ENOTSUPP;
 
-	/* Make sure the resource isn't assigned before resizing it. */
-	if (!(res->flags & IORESOURCE_UNSET))
-		return -EBUSY;
-
 	if (pci_resize_is_memory_decoding_enabled(dev, resno))
 		return -EBUSY;
 
@@ -492,19 +487,13 @@ int pci_resize_resource(struct pci_dev *
 	if (ret)
 		return ret;
 
-	pci_resize_resource_set_size(dev, resno, size);
-
-	/* Check if the new config works by trying to assign everything. */
-	if (dev->bus->self) {
-		ret = pbus_reassign_bridge_resources(dev->bus, res);
-		if (ret)
-			goto error_resize;
-	}
+	ret = pci_do_resource_release_and_resize(dev, resno, size, exclude_bars);
+	if (ret)
+		goto error_resize;
 	return 0;
 
 error_resize:
 	pci_rebar_set_size(dev, resno, old);
-	pci_resize_resource_set_size(dev, resno, old);
 	return ret;
 }
 EXPORT_SYMBOL(pci_resize_resource);
include/linux/pci.h +2 -1
@@ -1425,7 +1425,8 @@ static inline int pci_rebar_bytes_to_siz
 }
 
 u32 pci_rebar_get_possible_sizes(struct pci_dev *pdev, int bar);
-int __must_check pci_resize_resource(struct pci_dev *dev, int i, int size);
+int __must_check pci_resize_resource(struct pci_dev *dev, int i, int size,
+				     int exclude_bars);
 int pci_select_bars(struct pci_dev *dev, unsigned long flags);
 bool pci_device_is_present(struct pci_dev *pdev);
 void pci_ignore_hotplug(struct pci_dev *dev);
author maintainer
#1542 lore raw

PATCH 1536/1611 PCI: Add kerneldoc for pci_resize_resource()

In reply to Greg Kroah-Hartman

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

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

From: Ilpo Järvinen <[email protected]>

[ Upstream commit d787018e2dfdc4c1331538e7a8717690d1b7c9b3 ]

As pci_resize_resource() is meant to be used also outside of PCI core, document the interface with kerneldoc.

Stable-dep-of: ee7471fe968d ("PCI: Skip Resizable BAR restore on read error")

---
 drivers/pci/setup-res.c |   20 ++++++++++++++++++++
 1 file changed, 20 insertions(+)
drivers/pci/setup-res.c +20 -0
@@ -457,6 +457,26 @@ void pci_resize_resource_set_size(struct
 	resource_set_size(res, res_size);
 }
 
+/**
+ * pci_resize_resource - reconfigure a Resizable BAR and resources
+ * @dev: the PCI device
+ * @resno: index of the BAR to be resized
+ * @size: new size as defined in the spec (0=1MB, 31=128TB)
+ * @exclude_bars: a mask of BARs that should not be released
+ *
+ * Reconfigure @resno to @size and re-run resource assignment algorithm
+ * with the new size.
+ *
+ * Prior to resize, release @dev resources that share a bridge window with
+ * @resno.  This unpins the bridge window resource to allow changing it.
+ *
+ * The caller may prevent releasing a particular BAR by providing
+ * @exclude_bars mask, but this may result in the resize operation failing
+ * due to insufficient space.
+ *
+ * Return: 0 on success, or negative on error. In case of an error, the
+ *         resources are restored to their original places.
+ */
 int pci_resize_resource(struct pci_dev *dev, int resno, int size,
 			int exclude_bars)
 {
author maintainer
#1543 lore raw

PATCH 1537/1611 PCI: Move Resizable BAR code to rebar.c

In reply to Greg Kroah-Hartman

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

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

From: Ilpo Järvinen <[email protected]>

[ Upstream commit 9f71938cd77f32a448f40a288e409eca60e55486 ]

For lack of a better place to put it, Resizable BAR code has been placed inside pci.c and setup-res.c that do not use it for anything. Upcoming changes are going to add more Resizable BAR related functions, increasing the code size.

As pci.c is huge as is, move the Resizable BAR related code and the BAR resize code from setup-res.c to rebar.c.

Stable-dep-of: ee7471fe968d ("PCI: Skip Resizable BAR restore on read error")

---
 Documentation/driver-api/pci/pci.rst |    3 
 drivers/pci/Makefile                 |    2 
 drivers/pci/pci.c                    |  149 ---------------------
 drivers/pci/pci.h                    |    1 
 drivers/pci/rebar.c                  |  247 +++++++++++++++++++++++++++++++++++
 drivers/pci/setup-res.c              |   85 ------------
 6 files changed, 252 insertions(+), 235 deletions(-)
 create mode 100644 drivers/pci/rebar.c
Documentation/driver-api/pci/pci.rst +3 -0
@@ -37,6 +37,9 @@ PCI Support Library
 .. kernel-doc:: drivers/pci/slot.c
    :export:
 
+.. kernel-doc:: drivers/pci/rebar.c
+   :export:
+
 .. kernel-doc:: drivers/pci/rom.c
    :export:
 
drivers/pci/Makefile +1 -1
@@ -4,7 +4,7 @@
 
 obj-$(CONFIG_PCI)		+= access.o bus.o probe.o host-bridge.o \
 				   remove.o pci.o pci-driver.o search.o \
-				   rom.o setup-res.o irq.o vpd.o \
+				   rebar.o rom.o setup-res.o irq.o vpd.o \
 				   setup-bus.o vc.o mmap.o devres.o
 
 obj-$(CONFIG_PCI)		+= msi/
drivers/pci/pci.c +0 -149

161 lines not on this page · raw message

drivers/pci/pci.h +1 -0
@@ -1042,6 +1042,7 @@ static inline int acpi_get_rc_resources(
 #endif
 
 void pci_rebar_init(struct pci_dev *pdev);
+void pci_restore_rebar_state(struct pci_dev *pdev);
 int pci_rebar_get_current_size(struct pci_dev *pdev, int bar);
 int pci_rebar_set_size(struct pci_dev *pdev, int bar, int size);
 static inline u64 pci_rebar_size_to_bytes(int size)
drivers/pci/rebar.c new file +247 -0

247 lines not on this page · raw message

drivers/pci/setup-res.c +0 -85

91 lines not on this page · raw message

author maintainer
#1544 lore raw

PATCH 1538/1611 PCI: Skip Resizable BAR restore on read error

In reply to Greg Kroah-Hartman

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

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

From: Marco Nenciarini <[email protected]>

[ Upstream commit ee7471fe968d210939be9046089a924cd23c8c3b ]

pci_restore_rebar_state() uses the Resizable BAR Control register to decide how many BARs to restore (nbars) and which BAR each iteration addresses (bar_idx).

When a device does not respond, config reads typically return PCI_ERROR_RESPONSE (~0). Both fields are 3 bits wide, so nbars and bar_idx both evaluate to 7, past the spec's valid ranges for both fields. pci_resource_n() then returns an unrelated resource slot, whose size is used to derive a nonsensical value written back to the Resizable BAR Control register.

Bail out if any Resizable BAR Control read returns PCI_ERROR_RESPONSE. No further BARs are touched, which is safe because a config read that returns PCI_ERROR_RESPONSE indicates the device is unreachable and restoration is pointless.

---
 drivers/pci/rebar.c |    6 ++++++
 1 file changed, 6 insertions(+)
drivers/pci/rebar.c +6 -0
@@ -145,6 +145,9 @@ void pci_restore_rebar_state(struct pci_
 		return;
 
 	pci_read_config_dword(pdev, pos + PCI_REBAR_CTRL, &ctrl);
+	if (PCI_POSSIBLE_ERROR(ctrl))
+		return;
+
 	nbars = FIELD_GET(PCI_REBAR_CTRL_NBAR_MASK, ctrl);
 
 	for (i = 0; i < nbars; i++, pos += 8) {
@@ -152,6 +155,9 @@ void pci_restore_rebar_state(struct pci_
 		int bar_idx, size;
 
 		pci_read_config_dword(pdev, pos + PCI_REBAR_CTRL, &ctrl);
+		if (PCI_POSSIBLE_ERROR(ctrl))
+			return;
+
 		bar_idx = ctrl & PCI_REBAR_CTRL_BAR_IDX;
 		res = pci_resource_n(pdev, bar_idx);
 		size = pci_rebar_bytes_to_size(resource_size(res));
author maintainer
#1545 lore raw

PATCH 1539/1611 staging: rtl8723bs: core: move constants to right side in comparison

In reply to Greg Kroah-Hartman

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

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

From: William Hansen-Baird <[email protected]>

[ Upstream commit cf0f2680c30d4b438a5e84b73dc70be405936b54 ]

Move constants to right side in if-statement conditions.

Stable-dep-of: 1463ca3ec660 ("staging: rtl8723bs: fix OOB reads in rtw_get_sec_ie(), rtw_get_wapi_ie(), and rtw_get_wps_attr()")

---
 drivers/staging/rtl8723bs/core/rtw_ieee80211.c |    4 ++--
 drivers/staging/rtl8723bs/core/rtw_security.c  |    2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)
drivers/staging/rtl8723bs/core/rtw_ieee80211.c +2 -2
@@ -1012,7 +1012,7 @@ static int rtw_get_cipher_info(struct wl
 	pbuf = rtw_get_wpa_ie(&pnetwork->network.ies[12], &wpa_ielen, pnetwork->network.ie_length-12);
 
 	if (pbuf && (wpa_ielen > 0)) {
-		if (_SUCCESS == rtw_parse_wpa_ie(pbuf, wpa_ielen+2, &group_cipher, &pairwise_cipher, &is8021x)) {
+		if (rtw_parse_wpa_ie(pbuf, wpa_ielen+2, &group_cipher, &pairwise_cipher, &is8021x) == _SUCCESS) {
 			pnetwork->bcn_info.pairwise_cipher = pairwise_cipher;
 			pnetwork->bcn_info.group_cipher = group_cipher;
 			pnetwork->bcn_info.is_8021x = is8021x;
@@ -1022,7 +1022,7 @@ static int rtw_get_cipher_info(struct wl
 		pbuf = rtw_get_wpa2_ie(&pnetwork->network.ies[12], &wpa_ielen, pnetwork->network.ie_length-12);
 
 		if (pbuf && (wpa_ielen > 0)) {
-			if (_SUCCESS == rtw_parse_wpa2_ie(pbuf, wpa_ielen+2, &group_cipher, &pairwise_cipher, &is8021x)) {
+			if (rtw_parse_wpa2_ie(pbuf, wpa_ielen+2, &group_cipher, &pairwise_cipher, &is8021x) == _SUCCESS) {
 				pnetwork->bcn_info.pairwise_cipher = pairwise_cipher;
 				pnetwork->bcn_info.group_cipher = group_cipher;
 				pnetwork->bcn_info.is_8021x = is8021x;
drivers/staging/rtl8723bs/core/rtw_security.c +1 -1
@@ -1495,7 +1495,7 @@ void rtw_sec_restore_wep_key(struct adap
 	struct security_priv *securitypriv = &(adapter->securitypriv);
 	signed int keyid;
 
-	if ((_WEP40_ == securitypriv->dot11PrivacyAlgrthm) || (_WEP104_ == securitypriv->dot11PrivacyAlgrthm)) {
+	if ((securitypriv->dot11PrivacyAlgrthm == _WEP40_) || (securitypriv->dot11PrivacyAlgrthm == _WEP104_)) {
 		for (keyid = 0; keyid < 4; keyid++) {
 			if (securitypriv->key_mask & BIT(keyid)) {
 				if (keyid == securitypriv->dot11PrivacyKeyIndex)
author maintainer
#1546 lore raw

PATCH 1540/1611 staging: rtl8723bs: fix spaces around binary operators

In reply to Greg Kroah-Hartman

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

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

From: Nikolay Kulikov <[email protected]>

[ Upstream commit 0c9d1b56f9af0762a5be5118e1bb962f23784dc4 ]

Add missing spaces and fix line length to comply with kernel coding style.

Stable-dep-of: 1463ca3ec660 ("staging: rtl8723bs: fix OOB reads in rtw_get_sec_ie(), rtw_get_wapi_ie(), and rtw_get_wps_attr()")

---
 drivers/staging/rtl8723bs/core/rtw_ieee80211.c |   78 +++++++++++++------------
 1 file changed, 42 insertions(+), 36 deletions(-)
drivers/staging/rtl8723bs/core/rtw_ieee80211.c +42 -36
@@ -456,10 +456,10 @@ int rtw_parse_wpa_ie(u8 *wpa_ie, int wpa
 		return _FAIL;
 	}
 
-	if ((*wpa_ie != WLAN_EID_VENDOR_SPECIFIC) || (*(wpa_ie+1) != (u8)(wpa_ie_len - 2)) ||
-	   (memcmp(wpa_ie+2, RTW_WPA_OUI_TYPE, WPA_SELECTOR_LEN))) {
+	if ((*wpa_ie != WLAN_EID_VENDOR_SPECIFIC) ||
+	    (*(wpa_ie + 1) != (u8)(wpa_ie_len - 2)) ||
+	    (memcmp(wpa_ie + 2, RTW_WPA_OUI_TYPE, WPA_SELECTOR_LEN)))
 		return _FAIL;
-	}
 
 	pos = wpa_ie;
 
@@ -519,7 +519,7 @@ int rtw_parse_wpa2_ie(u8 *rsn_ie, int rs
 		return _FAIL;
 	}
 
-	if ((*rsn_ie != WLAN_EID_RSN) || (*(rsn_ie+1) != (u8)(rsn_ie_len - 2)))
+	if ((*rsn_ie != WLAN_EID_RSN) || (*(rsn_ie + 1) != (u8)(rsn_ie_len - 2)))
 		return _FAIL;
 
 	pos = rsn_ie;
@@ -587,18 +587,18 @@ int rtw_get_wapi_ie(u8 *in_ie, uint in_l
 	while (cnt < in_len) {
 		authmode = in_ie[cnt];
 
-		/* if (authmode == WLAN_EID_BSS_AC_ACCESS_DELAY) */
-		if (authmode == WLAN_EID_BSS_AC_ACCESS_DELAY && (!memcmp(&in_ie[cnt+6], wapi_oui1, 4) ||
-					!memcmp(&in_ie[cnt+6], wapi_oui2, 4))) {
+		if (authmode == WLAN_EID_BSS_AC_ACCESS_DELAY &&
+		    (!memcmp(&in_ie[cnt + 6], wapi_oui1, 4) ||
+		     !memcmp(&in_ie[cnt + 6], wapi_oui2, 4))) {
 			if (wapi_ie)
-				memcpy(wapi_ie, &in_ie[cnt], in_ie[cnt+1]+2);
+				memcpy(wapi_ie, &in_ie[cnt], in_ie[cnt + 1] + 2);
 
 			if (wapi_len)
-				*wapi_len = in_ie[cnt+1]+2;
+				*wapi_len = in_ie[cnt + 1] + 2;
 
-			cnt += in_ie[cnt+1]+2;  /* get next */
+			cnt += in_ie[cnt + 1] + 2;  /* get next */
 		} else {
-			cnt += in_ie[cnt+1]+2;   /* get next */
+			cnt += in_ie[cnt + 1] + 2;   /* get next */
 		}
 	}
 
@@ -622,9 +622,10 @@ void rtw_get_sec_ie(u8 *in_ie, uint in_l
 	while (cnt < in_len) {
 		authmode = in_ie[cnt];
 
-		if ((authmode == WLAN_EID_VENDOR_SPECIFIC) && (!memcmp(&in_ie[cnt+2], &wpa_oui[0], 4))) {
+		if ((authmode == WLAN_EID_VENDOR_SPECIFIC) &&
+		    (!memcmp(&in_ie[cnt + 2], &wpa_oui[0], 4))) {
 			if (wpa_ie)
-				memcpy(wpa_ie, &in_ie[cnt], in_ie[cnt+1]+2);
+				memcpy(wpa_ie, &in_ie[cnt], in_ie[cnt + 1] + 2);
 
 			*wpa_len = in_ie[cnt + 1] + 2;
 			cnt += in_ie[cnt + 1] + 2;  /* get next */
@@ -633,10 +634,10 @@ void rtw_get_sec_ie(u8 *in_ie, uint in_l
 				if (rsn_ie)
 					memcpy(rsn_ie, &in_ie[cnt], in_ie[cnt + 1] + 2);
 
-				*rsn_len = in_ie[cnt+1]+2;
-				cnt += in_ie[cnt+1]+2;  /* get next */
+				*rsn_len = in_ie[cnt + 1] + 2;
+				cnt += in_ie[cnt + 1] + 2;  /* get next */
 			} else {
-				cnt += in_ie[cnt+1]+2;   /* get next */
+				cnt += in_ie[cnt + 1] + 2;   /* get next */
 			}
 		}
 	}
@@ -668,20 +669,20 @@ u8 *rtw_get_wps_ie(u8 *in_ie, uint in_le
 	while (cnt < in_len) {
 		eid = in_ie[cnt];
 
-		if ((eid == WLAN_EID_VENDOR_SPECIFIC) && (!memcmp(&in_ie[cnt+2], wps_oui, 4))) {
+		if ((eid == WLAN_EID_VENDOR_SPECIFIC) && (!memcmp(&in_ie[cnt + 2], wps_oui, 4))) {
 			wpsie_ptr = &in_ie[cnt];
 
 			if (wps_ie)
-				memcpy(wps_ie, &in_ie[cnt], in_ie[cnt+1]+2);
+				memcpy(wps_ie, &in_ie[cnt], in_ie[cnt + 1] + 2);
 
 			if (wps_ielen)
-				*wps_ielen = in_ie[cnt+1]+2;
+				*wps_ielen = in_ie[cnt + 1] + 2;
 
-			cnt += in_ie[cnt+1]+2;
+			cnt += in_ie[cnt + 1] + 2;
 
 			break;
 		}
-		cnt += in_ie[cnt+1]+2; /* goto next */
+		cnt += in_ie[cnt + 1] + 2; /* goto next */
 	}
 
 	return wpsie_ptr;
@@ -759,12 +760,12 @@ u8 *rtw_get_wps_attr_content(u8 *wps_ie,
 
 	if (attr_ptr && attr_len) {
 		if (buf_content)
-			memcpy(buf_content, attr_ptr+4, attr_len-4);
+			memcpy(buf_content, attr_ptr + 4, attr_len - 4);
 
 		if (len_content)
-			*len_content = attr_len-4;
+			*len_content = attr_len - 4;
 
-		return attr_ptr+4;
+		return attr_ptr + 4;
 	}
 
 	return NULL;
@@ -1009,20 +1010,25 @@ static int rtw_get_cipher_info(struct wl
 	int group_cipher = 0, pairwise_cipher = 0, is8021x = 0;
 	int ret = _FAIL;
 
-	pbuf = rtw_get_wpa_ie(&pnetwork->network.ies[12], &wpa_ielen, pnetwork->network.ie_length-12);
+	pbuf = rtw_get_wpa_ie(&pnetwork->network.ies[12],
+			      &wpa_ielen,
+			      pnetwork->network.ie_length - 12);
 
 	if (pbuf && (wpa_ielen > 0)) {
-		if (rtw_parse_wpa_ie(pbuf, wpa_ielen+2, &group_cipher, &pairwise_cipher, &is8021x) == _SUCCESS) {
+		if (rtw_parse_wpa_ie(pbuf, wpa_ielen + 2, &group_cipher,
+				     &pairwise_cipher, &is8021x) == _SUCCESS) {
 			pnetwork->bcn_info.pairwise_cipher = pairwise_cipher;
 			pnetwork->bcn_info.group_cipher = group_cipher;
 			pnetwork->bcn_info.is_8021x = is8021x;
 			ret = _SUCCESS;
 		}
 	} else {
-		pbuf = rtw_get_wpa2_ie(&pnetwork->network.ies[12], &wpa_ielen, pnetwork->network.ie_length-12);
+		pbuf = rtw_get_wpa2_ie(&pnetwork->network.ies[12], &wpa_ielen,
+				       pnetwork->network.ie_length - 12);
 
 		if (pbuf && (wpa_ielen > 0)) {
-			if (rtw_parse_wpa2_ie(pbuf, wpa_ielen+2, &group_cipher, &pairwise_cipher, &is8021x) == _SUCCESS) {
+			if (rtw_parse_wpa2_ie(pbuf, wpa_ielen + 2, &group_cipher,
+					      &pairwise_cipher, &is8021x) == _SUCCESS) {
 				pnetwork->bcn_info.pairwise_cipher = pairwise_cipher;
 				pnetwork->bcn_info.group_cipher = group_cipher;
 				pnetwork->bcn_info.is_8021x = is8021x;
@@ -1091,21 +1097,21 @@ u16 rtw_mcs_rate(u8 bw_40MHz, u8 short_G
 	u16 max_rate = 0;
 
 	if (MCS_rate[0] & BIT(7))
-		max_rate = (bw_40MHz) ? ((short_GI)?1500:1350):((short_GI)?722:650);
+		max_rate = (bw_40MHz) ? ((short_GI) ? 1500 : 1350) : ((short_GI) ? 722 : 650);
 	else if (MCS_rate[0] & BIT(6))
-		max_rate = (bw_40MHz) ? ((short_GI)?1350:1215):((short_GI)?650:585);
+		max_rate = (bw_40MHz) ? ((short_GI) ? 1350 : 1215) : ((short_GI) ? 650 : 585);
 	else if (MCS_rate[0] & BIT(5))
-		max_rate = (bw_40MHz) ? ((short_GI)?1200:1080):((short_GI)?578:520);
+		max_rate = (bw_40MHz) ? ((short_GI) ? 1200 : 1080) : ((short_GI) ? 578 : 520);
 	else if (MCS_rate[0] & BIT(4))
-		max_rate = (bw_40MHz) ? ((short_GI)?900:810):((short_GI)?433:390);
+		max_rate = (bw_40MHz) ? ((short_GI) ? 900 : 810) : ((short_GI) ? 433 : 390);
 	else if (MCS_rate[0] & BIT(3))
-		max_rate = (bw_40MHz) ? ((short_GI)?600:540):((short_GI)?289:260);
+		max_rate = (bw_40MHz) ? ((short_GI) ? 600 : 540) : ((short_GI) ? 289 : 260);
 	else if (MCS_rate[0] & BIT(2))
-		max_rate = (bw_40MHz) ? ((short_GI)?450:405):((short_GI)?217:195);
+		max_rate = (bw_40MHz) ? ((short_GI) ? 450 : 405) : ((short_GI) ? 217 : 195);
 	else if (MCS_rate[0] & BIT(1))
-		max_rate = (bw_40MHz) ? ((short_GI)?300:270):((short_GI)?144:130);
+		max_rate = (bw_40MHz) ? ((short_GI) ? 300 : 270) : ((short_GI) ? 144 : 130);
 	else if (MCS_rate[0] & BIT(0))
-		max_rate = (bw_40MHz) ? ((short_GI)?150:135):((short_GI)?72:65);
+		max_rate = (bw_40MHz) ? ((short_GI) ? 150 : 135) : ((short_GI) ? 72 : 65);
 
 	return max_rate;
 }
author maintainer
#1547 lore raw

PATCH 1541/1611 staging: rtl8723bs: fix OOB reads in rtw_get_sec_ie(), rtw_get_wapi_ie(), and rtw_get_wps_attr()

In reply to Greg Kroah-Hartman

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

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

From: Alexandru Hossu <[email protected]>

[ Upstream commit 1463ca3ec6601cbb097d8d87dbf5dcf1cb86a344 ]

Three IE/attribute parsing functions have missing bounds checks.

rtw_get_sec_ie() and rtw_get_wapi_ie() iterate over a raw IE buffer without verifying that the header bytes (tag + length) are within the remaining buffer before reading them. Additionally, rtw_get_sec_ie() compares the 4-byte WPA OUI at cnt+2 without checking that at least 6 bytes remain, and rtw_get_wapi_ie() compares a 4-byte WAPI OUI at cnt+6 without checking that at least 10 bytes remain.

rtw_get_wps_attr() reads wps_ie[0] and wps_ie+2 unconditionally at entry, before verifying that wps_ielen is large enough to contain the 6-byte WPS IE header (element_id + length + 4-byte OUI). Inside the attribute loop, get_unaligned_be16() is called on attr_ptr and attr_ptr+2 without checking that 4 bytes remain in the buffer.

Add a cnt+2 bounds check before each loop body in rtw_get_sec_ie() and rtw_get_wapi_ie(), guard each multi-byte comparison with a minimum IE length requirement, add a wps_ielen < 6 early return in rtw_get_wps_attr(), and add a 4-byte bounds check in its inner loop.

---
 drivers/staging/rtl8723bs/core/rtw_ieee80211.c |   15 +++++++++++++++
 1 file changed, 15 insertions(+)
drivers/staging/rtl8723bs/core/rtw_ieee80211.c +15 -0
@@ -585,9 +585,14 @@ int rtw_get_wapi_ie(u8 *in_ie, uint in_l
 	cnt = (_TIMESTAMP_ + _BEACON_ITERVAL_ + _CAPABILITY_);
 
 	while (cnt < in_len) {
+		if (cnt + 2 > in_len)
+			break;
+		if (cnt + 2 + in_ie[cnt + 1] > in_len)
+			break;
 		authmode = in_ie[cnt];
 
 		if (authmode == WLAN_EID_BSS_AC_ACCESS_DELAY &&
+		    in_ie[cnt + 1] >= 8 &&
 		    (!memcmp(&in_ie[cnt + 6], wapi_oui1, 4) ||
 		     !memcmp(&in_ie[cnt + 6], wapi_oui2, 4))) {
 			if (wapi_ie)
@@ -620,9 +625,14 @@ void rtw_get_sec_ie(u8 *in_ie, uint in_l
 	cnt = (_TIMESTAMP_ + _BEACON_ITERVAL_ + _CAPABILITY_);
 
 	while (cnt < in_len) {
+		if (cnt + 2 > in_len)
+			break;
+		if (cnt + 2 + in_ie[cnt + 1] > in_len)
+			break;
 		authmode = in_ie[cnt];
 
 		if ((authmode == WLAN_EID_VENDOR_SPECIFIC) &&
+		    in_ie[cnt + 1] >= 4 &&
 		    (!memcmp(&in_ie[cnt + 2], &wpa_oui[0], 4))) {
 			if (wpa_ie)
 				memcpy(wpa_ie, &in_ie[cnt], in_ie[cnt + 1] + 2);
@@ -707,6 +717,9 @@ u8 *rtw_get_wps_attr(u8 *wps_ie, uint wp
 	if (len_attr)
 		*len_attr = 0;
 
+	if (wps_ielen < 6)
+		return attr_ptr;
+
 	if ((wps_ie[0] != WLAN_EID_VENDOR_SPECIFIC) ||
 		(memcmp(wps_ie + 2, wps_oui, 4))) {
 		return attr_ptr;
@@ -717,6 +730,8 @@ u8 *rtw_get_wps_attr(u8 *wps_ie, uint wp
 
 	while (attr_ptr - wps_ie < wps_ielen) {
 		/*  4 = 2(Attribute ID) + 2(Length) */
+		if (attr_ptr + 4 > wps_ie + wps_ielen)
+			break;
 		u16 attr_id = get_unaligned_be16(attr_ptr);
 		u16 attr_data_len = get_unaligned_be16(attr_ptr + 2);
 		u16 attr_len = attr_data_len + 4;
author maintainer
#1548 lore raw

PATCH 1542/1611 crypto: qat - fix VF2PF work teardown race in adf_disable_sriov()

In reply to Greg Kroah-Hartman

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

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

From: Giovanni Cabiddu <[email protected]>

[ Upstream commit 277281c10c63791067d24d421f7c43a15faa9096 ]

The VF2PF interrupt handler queues PF-side response work that stores a raw pointer to per-VF state (struct adf_accel_vf_info). Currently, adf_disable_sriov() destroys per-VF mutexes and frees vf_info without stopping new VF2PF work or waiting for in-flight workers to complete. A concurrently scheduled or already queued worker can then dereference freed memory.

This manifests as a use-after-free when KASAN is enabled:

  BUG: KASAN: null-ptr-deref in mutex_lock+0x76/0xe0
  Write of size 8 at addr 0000000000000260 by task kworker/24:2/...
  Workqueue: qat_pf2vf_resp_wq adf_iov_send_resp [intel_qat]
  Call Trace:
    kasan_report+0x119/0x140
    mutex_lock+0x76/0xe0
    adf_gen4_pfvf_send+0xd4/0x1f0 [intel_qat]
    adf_recv_and_handle_vf2pf_msg+0x290/0x360 [intel_qat]
    adf_iov_send_resp+0x8c/0xe0 [intel_qat]
    process_one_work+0x6ac/0xfd0
    worker_thread+0x4dd/0xd30
    kthread+0x326/0x410
    ret_from_fork+0x33b/0x670

Add a PF-local flag, vf2pf_disabled, that gates work queueing, worker processing, and interrupt re-enabling during teardown. Set this flag atomically with the hardware interrupt mask inside adf_disable_all_vf2pf_interrupts(). After masking, synchronize the AE cluster MSI-X interrupt and flush the PF response workqueue before tearing down per-VF locks and state so all in-flight work completes before vf_info is destroyed.

Introduce adf_enable_all_vf2pf_interrupts() to clear the flag and unmask all VF2PF interrupts under the same lock when SR-IOV is re-enabled. This ensures the software flag and hardware state transition atomically on both the enable and disable paths.

Fixes ed8ccaef52fa ("crypto: qat - Add support for SRIOV")
---
 drivers/crypto/intel/qat/qat_common/adf_accel_devices.h |    2 
 drivers/crypto/intel/qat/qat_common/adf_common_drv.h    |    2 
 drivers/crypto/intel/qat/qat_common/adf_isr.c           |   39 ++++++++++++++++
 drivers/crypto/intel/qat/qat_common/adf_sriov.c         |   20 +++++++-
 4 files changed, 61 insertions(+), 2 deletions(-)
drivers/crypto/intel/qat/qat_common/adf_accel_devices.h +2 -0
@@ -472,6 +472,8 @@ struct adf_accel_dev {
 		struct {
 			/* protects VF2PF interrupts access */
 			spinlock_t vf2pf_ints_lock;
+			/* prevents VF2PF handling from racing with VF state teardown */
+			bool vf2pf_disabled;
 			/* vf_info is non-zero when SR-IOV is init'ed */
 			struct adf_accel_vf_info *vf_info;
 		} pf;
drivers/crypto/intel/qat/qat_common/adf_common_drv.h +2 -0
@@ -118,6 +118,7 @@ void qat_comp_alg_callback(void *resp);
 
 int adf_isr_resource_alloc(struct adf_accel_dev *accel_dev);
 void adf_isr_resource_free(struct adf_accel_dev *accel_dev);
+void adf_isr_sync_ae_cluster(struct adf_accel_dev *accel_dev);
 int adf_vf_isr_resource_alloc(struct adf_accel_dev *accel_dev);
 void adf_vf_isr_resource_free(struct adf_accel_dev *accel_dev);
 
@@ -191,6 +192,7 @@ int adf_sriov_configure(struct pci_dev *
 void adf_disable_sriov(struct adf_accel_dev *accel_dev);
 void adf_reenable_sriov(struct adf_accel_dev *accel_dev);
 void adf_enable_vf2pf_interrupts(struct adf_accel_dev *accel_dev, u32 vf_mask);
+void adf_enable_all_vf2pf_interrupts(struct adf_accel_dev *accel_dev, u32 num_vfs);
 void adf_disable_all_vf2pf_interrupts(struct adf_accel_dev *accel_dev);
 bool adf_recv_and_handle_pf2vf_msg(struct adf_accel_dev *accel_dev);
 bool adf_recv_and_handle_vf2pf_msg(struct adf_accel_dev *accel_dev, u32 vf_nr);
drivers/crypto/intel/qat/qat_common/adf_isr.c +39 -0
@@ -62,6 +62,23 @@ void adf_enable_vf2pf_interrupts(struct
 	unsigned long flags;
 
 	spin_lock_irqsave(&accel_dev->pf.vf2pf_ints_lock, flags);
+	if (!READ_ONCE(accel_dev->pf.vf2pf_disabled))
+		GET_PFVF_OPS(accel_dev)->enable_vf2pf_interrupts(pmisc_addr, vf_mask);
+	spin_unlock_irqrestore(&accel_dev->pf.vf2pf_ints_lock, flags);
+}
+
+void adf_enable_all_vf2pf_interrupts(struct adf_accel_dev *accel_dev, u32 num_vfs)
+{
+	void __iomem *pmisc_addr = adf_get_pmisc_base(accel_dev);
+	unsigned long flags;
+	u32 vf_mask;
+
+	vf_mask = BIT_ULL(num_vfs) - 1;
+	if (!vf_mask)
+		return;
+
+	spin_lock_irqsave(&accel_dev->pf.vf2pf_ints_lock, flags);
+	WRITE_ONCE(accel_dev->pf.vf2pf_disabled, false);
 	GET_PFVF_OPS(accel_dev)->enable_vf2pf_interrupts(pmisc_addr, vf_mask);
 	spin_unlock_irqrestore(&accel_dev->pf.vf2pf_ints_lock, flags);
 }
@@ -72,6 +89,7 @@ void adf_disable_all_vf2pf_interrupts(st
 	unsigned long flags;
 
 	spin_lock_irqsave(&accel_dev->pf.vf2pf_ints_lock, flags);
+	WRITE_ONCE(accel_dev->pf.vf2pf_disabled, true);
 	GET_PFVF_OPS(accel_dev)->disable_all_vf2pf_interrupts(pmisc_addr);
 	spin_unlock_irqrestore(&accel_dev->pf.vf2pf_ints_lock, flags);
 }
@@ -174,6 +192,27 @@ static irqreturn_t adf_msix_isr_ae(int i
 	return IRQ_NONE;
 }
 
+void adf_isr_sync_ae_cluster(struct adf_accel_dev *accel_dev)
+{
+	struct adf_accel_pci *pci_dev_info = &accel_dev->accel_pci_dev;
+	struct adf_hw_device_data *hw_data = GET_HW_DATA(accel_dev);
+	u32 num_entries = pci_dev_info->msix_entries.num_entries;
+	struct adf_irq *irqs = pci_dev_info->msix_entries.irqs;
+	u32 irq_idx;
+	int irq;
+
+	if (!test_bit(ADF_STATUS_IRQ_ALLOCATED, &accel_dev->status) || !irqs)
+		return;
+
+	irq_idx = num_entries > 1 ? hw_data->num_banks : 0;
+	if (irq_idx >= num_entries || !irqs[irq_idx].enabled)
+		return;
+
+	irq = pci_irq_vector(pci_dev_info->pci_dev, hw_data->num_banks);
+	if (irq > 0)
+		synchronize_irq(irq);
+}
+
 static void adf_free_irqs(struct adf_accel_dev *accel_dev)
 {
 	struct adf_accel_pci *pci_dev_info = &accel_dev->accel_pci_dev;
drivers/crypto/intel/qat/qat_common/adf_sriov.c +18 -2
@@ -26,6 +26,9 @@ static void adf_iov_send_resp(struct wor
 	u32 vf_nr = vf_info->vf_nr;
 	bool ret;
 
+	if (READ_ONCE(accel_dev->pf.vf2pf_disabled))
+		goto out;
+
 	mutex_lock(&vf_info->pfvf_mig_lock);
 	ret = adf_recv_and_handle_vf2pf_msg(accel_dev, vf_nr);
 	if (ret)
@@ -33,13 +36,18 @@ static void adf_iov_send_resp(struct wor
 		adf_enable_vf2pf_interrupts(accel_dev, 1 << vf_nr);
 	mutex_unlock(&vf_info->pfvf_mig_lock);
 
+out:
 	kfree(pf2vf_resp);
 }
 
 void adf_schedule_vf2pf_handler(struct adf_accel_vf_info *vf_info)
 {
+	struct adf_accel_dev *accel_dev = vf_info->accel_dev;
 	struct adf_pf2vf_resp *pf2vf_resp;
 
+	if (READ_ONCE(accel_dev->pf.vf2pf_disabled))
+		return;
+
 	pf2vf_resp = kzalloc(sizeof(*pf2vf_resp), GFP_ATOMIC);
 	if (!pf2vf_resp)
 		return;
@@ -49,6 +57,12 @@ void adf_schedule_vf2pf_handler(struct a
 	queue_work(pf2vf_resp_wq, &pf2vf_resp->pf2vf_resp_work);
 }
 
+static void adf_flush_pf2vf_resp_wq(void)
+{
+	if (pf2vf_resp_wq)
+		flush_workqueue(pf2vf_resp_wq);
+}
+
 static int adf_enable_sriov(struct adf_accel_dev *accel_dev)
 {
 	struct pci_dev *pdev = accel_to_pci_dev(accel_dev);
@@ -75,7 +89,7 @@ static int adf_enable_sriov(struct adf_a
 		hw_data->configure_iov_threads(accel_dev, true);
 
 	/* Enable VF to PF interrupts for all VFs */
-	adf_enable_vf2pf_interrupts(accel_dev, BIT_ULL(totalvfs) - 1);
+	adf_enable_all_vf2pf_interrupts(accel_dev, totalvfs);
 
 	/* Do not enable SR-IOV if already enabled */
 	if (pci_num_vf(pdev))
@@ -259,8 +273,10 @@ void adf_disable_sriov(struct adf_accel_
 	if (!test_bit(ADF_STATUS_RESTARTING, &accel_dev->status))
 		pci_disable_sriov(accel_to_pci_dev(accel_dev));
 
-	/* Disable VF to PF interrupts */
+	/* Block VF2PF work and disable VF to PF interrupts */
 	adf_disable_all_vf2pf_interrupts(accel_dev);
+	adf_isr_sync_ae_cluster(accel_dev);
+	adf_flush_pf2vf_resp_wq();
 
 	/* Clear Valid bits in AE Thread to PCIe Function Mapping */
 	if (hw_data->configure_iov_threads)
author maintainer
#1549 lore raw

PATCH 1543/1611 media: nxp: imx8-isi: use devm_pm_runtime_enable() to simplify code

In reply to Greg Kroah-Hartman

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

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

From: Frank Li <[email protected]>

[ Upstream commit 078161dd44d6f848a62a473206d69025607736ec ]

Use devm_pm_runtime_enable() to simplify code. Change to use dev_err_probe() because previous goto change to return.

No functional change.

Stable-dep-of: b670bf89824e ("media: nxp: imx8-isi: Fix use-after-free on remove")

---
 drivers/media/platform/nxp/imx8-isi/imx8-isi-core.c |   16 +++++++---------
 1 file changed, 7 insertions(+), 9 deletions(-)
drivers/media/platform/nxp/imx8-isi/imx8-isi-core.c +7 -9
@@ -488,13 +488,14 @@ static int mxc_isi_probe(struct platform
 	dma_size = isi->pdata->has_36bit_dma ? 36 : 32;
 	dma_set_mask_and_coherent(dev, DMA_BIT_MASK(dma_size));
 
-	pm_runtime_enable(dev);
+	ret = devm_pm_runtime_enable(dev);
+	if (ret)
+		return ret;
 
 	ret = mxc_isi_crossbar_init(isi);
-	if (ret) {
-		dev_err(dev, "Failed to initialize crossbar: %d\n", ret);
-		goto err_pm;
-	}
+	if (ret)
+		return dev_err_probe(dev, ret,
+				     "Failed to initialize crossbar\n");
 
 	for (i = 0; i < isi->pdata->num_channels; ++i) {
 		ret = mxc_isi_pipe_init(isi, i);
@@ -517,8 +518,7 @@ static int mxc_isi_probe(struct platform
 
 err_xbar:
 	mxc_isi_crossbar_cleanup(&isi->crossbar);
-err_pm:
-	pm_runtime_disable(isi->dev);
+
 	return ret;
 }
 
@@ -537,8 +537,6 @@ static void mxc_isi_remove(struct platfo
 
 	mxc_isi_crossbar_cleanup(&isi->crossbar);
 	mxc_isi_v4l2_cleanup(isi);
-
-	pm_runtime_disable(isi->dev);
 }
 
 static const struct of_device_id mxc_isi_of_match[] = {
author maintainer
#1550 lore raw

PATCH 1544/1611 media: nxp: imx8-isi: Fix use-after-free on remove

In reply to Greg Kroah-Hartman

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

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

From: Xiaolei Wang <[email protected]>

[ Upstream commit b670bf89824ede5d07d20bb9bfbafb754846081d ]

KASAN reports a slab-use-after-free in __media_entity_remove_link() during rmmod of imx8_isi:

  BUG: KASAN: slab-use-after-free in __media_entity_remove_link+0x608/0x650
  Read of size 2 at addr ffff0000d47cb02a by task rmmod/724

  Call trace:
   __media_entity_remove_link+0x608/0x650
   __media_entity_remove_links+0x78/0x144
   __media_device_unregister_entity+0x150/0x280
   media_device_unregister_entity+0x48/0x68
   v4l2_device_unregister_subdev+0x158/0x300
   v4l2_async_unbind_subdev_one+0x22c/0x358
   v4l2_async_nf_unbind_all_subdevs+0xfc/0x1c0
   v4l2_async_nf_unregister+0x5c/0x14c
   mxc_isi_remove+0x124/0x2a0 [imx8_isi]

  Allocated by task 249:
   __kmalloc_noprof+0x27c/0x690
   mxc_isi_crossbar_init+0x22c/0x560 [imx8_isi]

  Freed by task 724:
   kfree+0x1e4/0x5b0
   mxc_isi_crossbar_cleanup+0x34/0x80 [imx8_isi]
   mxc_isi_remove+0x11c/0x2a0 [imx8_isi]

The problem is that mxc_isi_remove() calls mxc_isi_crossbar_cleanup() before mxc_isi_v4l2_cleanup(). The crossbar cleanup frees the media entity pads, but the subsequent v4l2 cleanup still tries to remove media links that reference those pads.

Fix this by calling mxc_isi_v4l2_cleanup() before mxc_isi_crossbar_cleanup() to ensure all media entities are properly unregistered while the pads are still valid.

---
 drivers/media/platform/nxp/imx8-isi/imx8-isi-core.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/media/platform/nxp/imx8-isi/imx8-isi-core.c +1 -1
@@ -535,8 +535,8 @@ static void mxc_isi_remove(struct platfo
 		mxc_isi_pipe_cleanup(pipe);
 	}
 
-	mxc_isi_crossbar_cleanup(&isi->crossbar);
 	mxc_isi_v4l2_cleanup(isi);
+	mxc_isi_crossbar_cleanup(&isi->crossbar);
 }
 
 static const struct of_device_id mxc_isi_of_match[] = {
author maintainer
#1551 lore raw

PATCH 1545/1611 Bluetooth: 6lowpan: fix cyclic locking warning on netdev unregister

In reply to Greg Kroah-Hartman

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

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

From: Pauli Virtanen <[email protected]>

[ Upstream commit 9707a015fe8f3ba8ec7c270f3b2b8efb38823d6b ]

6lowpan.c has theoretically conflicting lock orderings, which lockdep complains about:

    a) rtnl_lock > hdev->workqueue

    from 6lowpan.c:delete_netdev -> rtnl_lock -> device_del
    -> put_device(parent) -> hci_release_dev -> destroy_workqueue

    b) hdev->workqueue > l2cap_conn->lock > chan->lock > rtnl_lock

    from hci_rx_work -> 6lowpan.c:chan_ready_cb
    -> lowpan_register_netdev, ifup -> rtnl_lock

Actual deadlock appears not possible, as hci_rx_work is disabled and l2cap_conn flushed already on hdev unregister. Hence, do minimal thing to make lockdep happy by breaking chain a) by holding hdev refcount until after netdev put in 6lowpan.c.

Fixes the lockdep complaint:
WARNING: possible circular locking dependency detected.
kworker/0:1/11 is trying to acquire lock:
ffff8880023b3940 ((wq_completion)hci0#2){+.+.}-{0:0}, at: touch_wq_lockdep_map+0x8b/0x130 but task is already holding lock:
ffffffff95e4f9c0 (rtnl_mutex){+.+.}-{4:4}, at: lowpan_unregister_netdev+0xd/0x30

Workqueue: events delete_netdev

Stable-dep-of: 6fef032af009 ("Bluetooth: L2CAP: Fix use-after-free in l2cap_sock_new_connection_cb()")

---
 net/bluetooth/6lowpan.c |   25 +++++++++++++++++++++++--
 1 file changed, 23 insertions(+), 2 deletions(-)
net/bluetooth/6lowpan.c +23 -2
@@ -760,13 +760,33 @@ static inline struct l2cap_chan *chan_ne
 	return chan;
 }
 
+static void unregister_dev(struct lowpan_btle_dev *dev)
+{
+	struct hci_dev *hdev = READ_ONCE(dev->hdev);
+
+	/* If netdev holds last reference to hci_dev (its parent device), this
+	 * leads to theoretical cyclic locking on lowpan_unregister_netdev:
+	 *
+	 * rtnl_lock -> put_device(parent) -> hci_release_dev ->
+	 * destroy_workqueue -> hci_rx_work -> l2cap_recv_acldata ->
+	 * chan_ready_cb -> ifup -> rtnl_lock
+	 *
+	 * However, hci_rx_work is disabled in hci_unregister_dev, so this
+	 * should not occur. Make lockdep happy by postponing hdev release after
+	 * netdev put.
+	 */
+	hci_dev_hold(hdev);
+	lowpan_unregister_netdev(dev->netdev);
+	hci_dev_put(hdev);
+}
+
 static void delete_netdev(struct work_struct *work)
 {
 	struct lowpan_btle_dev *entry = container_of(work,
 						     struct lowpan_btle_dev,
 						     delete_netdev);
 
-	lowpan_unregister_netdev(entry->netdev);
+	unregister_dev(entry);
 
 	/* The entry pointer is deleted by the netdev destructor. */
 }
@@ -1249,6 +1269,7 @@ static void disconnect_devices(void)
 			break;
 
 		new_dev->netdev = entry->netdev;
+		new_dev->hdev = entry->hdev;
 		INIT_LIST_HEAD(&new_dev->list);
 
 		list_add_rcu(&new_dev->list, &devices);
@@ -1260,7 +1281,7 @@ static void disconnect_devices(void)
 		ifdown(entry->netdev);
 		BT_DBG("Unregistering netdev %s %p",
 		       entry->netdev->name, entry->netdev);
-		lowpan_unregister_netdev(entry->netdev);
+		unregister_dev(entry);
 		kfree(entry);
 	}
 }
author maintainer
#1552 lore raw

PATCH 1546/1611 Bluetooth: L2CAP: Fix use-after-free in l2cap_sock_new_connection_cb()

In reply to Greg Kroah-Hartman

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

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

From: Siwei Zhang <[email protected]>

[ Upstream commit 6fef032af0092ed5ccb767239a9ac1bc38c08a40 ]

l2cap_sock_new_connection_cb() returned l2cap_pi(sk)->chan after release_sock(parent). Once the parent lock is dropped the newly enqueued child socket sk is reachable via the accept queue, so another task can accept and free it before the callback dereferences sk, resulting in a use-after-free.

Rework the ->new_connection() op so the core, rather than the callback, owns the child channel's lifetime. The op now receives a pre-allocated new_chan and returns an errno instead of allocating and returning a channel. l2cap_new_connection() allocates the child channel and links it into the conn list via __l2cap_chan_add() before invoking the callback, so the conn-list reference keeps the channel alive once release_sock(parent) exposes the socket to other tasks.

Channel configuration that was duplicated in l2cap_sock_init() and the various new_connection callbacks is consolidated into l2cap_chan_set_defaults(), which now inherits from the parent channel when one is supplied.

Fixes 8ffb929098a5 ("Bluetooth: Remove parent socket usage from l2cap_core.c")
Assisted-by Claude:claude-opus-4-8
---
 include/net/bluetooth/l2cap.h |   10 ++--
 net/bluetooth/6lowpan.c       |   18 -------
 net/bluetooth/l2cap_core.c    |   78 ++++++++++++++++++++++++++-----
 net/bluetooth/l2cap_sock.c    |  103 +++++++++++++++++++-----------------------
 net/bluetooth/smp.c           |   27 ++---------
 5 files changed, 127 insertions(+), 109 deletions(-)
include/net/bluetooth/l2cap.h +6 -4
@@ -620,7 +620,8 @@ struct l2cap_chan {
 struct l2cap_ops {
 	char			*name;
 
-	struct l2cap_chan	*(*new_connection) (struct l2cap_chan *chan);
+	int			(*new_connection)(struct l2cap_chan *chan,
+						  struct l2cap_chan *new_chan);
 	int			(*recv) (struct l2cap_chan * chan,
 					 struct sk_buff *skb);
 	void			(*teardown) (struct l2cap_chan *chan, int err);
@@ -885,9 +886,10 @@ static inline __u16 __next_seq(struct l2
 	return (seq + 1) % (chan->tx_win_max + 1);
 }
 
-static inline struct l2cap_chan *l2cap_chan_no_new_connection(struct l2cap_chan *chan)
+static inline int l2cap_chan_no_new_connection(struct l2cap_chan *chan,
+					       struct l2cap_chan *new_chan)
 {
-	return NULL;
+	return -EOPNOTSUPP;
 }
 
 static inline int l2cap_chan_no_recv(struct l2cap_chan *chan, struct sk_buff *skb)
@@ -964,7 +966,7 @@ int l2cap_chan_send(struct l2cap_chan *c
 void l2cap_chan_busy(struct l2cap_chan *chan, int busy);
 void l2cap_chan_rx_avail(struct l2cap_chan *chan, ssize_t rx_avail);
 int l2cap_chan_check_security(struct l2cap_chan *chan, bool initiator);
-void l2cap_chan_set_defaults(struct l2cap_chan *chan);
+void l2cap_chan_set_defaults(struct l2cap_chan *chan, struct l2cap_chan *pchan);
 int l2cap_ertm_init(struct l2cap_chan *chan);
 void l2cap_chan_add(struct l2cap_conn *conn, struct l2cap_chan *chan);
 void __l2cap_chan_add(struct l2cap_conn *conn, struct l2cap_chan *chan);
net/bluetooth/6lowpan.c +1 -17
@@ -632,7 +632,7 @@ static struct l2cap_chan *chan_create(vo
 	if (!chan)
 		return NULL;
 
-	l2cap_chan_set_defaults(chan);
+	l2cap_chan_set_defaults(chan, NULL);
 
 	chan->chan_type = L2CAP_CHAN_CONN_ORIENTED;
 	chan->mode = L2CAP_MODE_LE_FLOWCTL;
@@ -745,21 +745,6 @@ static inline void chan_ready_cb(struct
 	ifup(dev->netdev);
 }
 
-static inline struct l2cap_chan *chan_new_conn_cb(struct l2cap_chan *pchan)
-{
-	struct l2cap_chan *chan;
-
-	chan = chan_create();
-	if (!chan)
-		return NULL;
-
-	chan->ops = pchan->ops;
-
-	BT_DBG("chan %p pchan %p", chan, pchan);
-
-	return chan;
-}
-
 static void unregister_dev(struct lowpan_btle_dev *dev)
 {
 	struct hci_dev *hdev = READ_ONCE(dev->hdev);
@@ -889,7 +874,6 @@ static long chan_get_sndtimeo_cb(struct
 
 static const struct l2cap_ops bt_6lowpan_chan_ops = {
 	.name			= "L2CAP 6LoWPAN channel",
-	.new_connection		= chan_new_conn_cb,
 	.recv			= chan_recv_cb,
 	.close			= chan_close_cb,
 	.state_change		= chan_state_change_cb,
net/bluetooth/l2cap_core.c +65 -13

143 lines not on this page · raw message

net/bluetooth/l2cap_sock.c +49 -54

201 lines not on this page · raw message

net/bluetooth/smp.c +6 -21
@@ -3234,34 +3234,19 @@ static const struct l2cap_ops smp_chan_o
 	.get_sndtimeo		= l2cap_chan_no_get_sndtimeo,
 };
 
-static inline struct l2cap_chan *smp_new_conn_cb(struct l2cap_chan *pchan)
+static inline int smp_new_conn_cb(struct l2cap_chan *chan,
+				  struct l2cap_chan *new_chan)
 {
-	struct l2cap_chan *chan;
-
-	BT_DBG("pchan %p", pchan);
-
-	chan = l2cap_chan_create();
-	if (!chan)
-		return NULL;
-
-	chan->chan_type	= pchan->chan_type;
-	chan->ops	= &smp_chan_ops;
-	chan->scid	= pchan->scid;
-	chan->dcid	= chan->scid;
-	chan->imtu	= pchan->imtu;
-	chan->omtu	= pchan->omtu;
-	chan->mode	= pchan->mode;
+	new_chan->ops = &smp_chan_ops;
 
 	/* Other L2CAP channels may request SMP routines in order to
 	 * change the security level. This means that the SMP channel
 	 * lock must be considered in its own category to avoid lockdep
 	 * warnings.
 	 */
-	atomic_set(&chan->nesting, L2CAP_NESTING_SMP);
-
-	BT_DBG("created chan %p", chan);
+	atomic_set(&new_chan->nesting, L2CAP_NESTING_SMP);
 
-	return chan;
+	return 0;
 }
 
 static const struct l2cap_ops smp_root_chan_ops = {
@@ -3332,7 +3317,7 @@ create_chan:
 
 	l2cap_add_scid(chan, cid);
 
-	l2cap_chan_set_defaults(chan);
+	l2cap_chan_set_defaults(chan, NULL);
 
 	if (cid == L2CAP_CID_SMP) {
 		u8 bdaddr_type;
author maintainer
#1553 lore raw

PATCH 1547/1611 cifs: SMB1 split: Rename cifstransport.c

In reply to Greg Kroah-Hartman

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

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

From: David Howells <[email protected]>

[ Upstream commit b09eab52b307df58a36f34d047378053a2e13e13 ]

Rename cifstransport.c to smb1transport.c in order to give consistent names SMB1-specific files.

Stable-dep-of: c6394bcaf254 ("ksmbd: use opener credentials for FSCTL mutations")

---
 fs/smb/client/Makefile                             | 5 ++++-
 fs/smb/client/{cifstransport.c => smb1transport.c} | 0
 fs/smb/client/Makefile        |    5 
 fs/smb/client/cifstransport.c |  565 ------------------------------------------
 fs/smb/client/smb1transport.c |  565 ++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 569 insertions(+), 566 deletions(-)
 rename fs/smb/client/{cifstransport.c => smb1transport.c} (100%)
fs/smb/client/Makefile +4 -1
@@ -32,6 +32,9 @@ cifs-$(CONFIG_CIFS_SMB_DIRECT) += smbdir
 
 cifs-$(CONFIG_CIFS_ROOT) += cifsroot.o
 
-cifs-$(CONFIG_CIFS_ALLOW_INSECURE_LEGACY) += smb1ops.o cifssmb.o cifstransport.o
+cifs-$(CONFIG_CIFS_ALLOW_INSECURE_LEGACY) += \
+	cifssmb.o \
+	smb1ops.o \
+	smb1transport.o
 
 cifs-$(CONFIG_CIFS_COMPRESSION) += compress.o compress/lz77.o
fs/smb/client/cifstransport.c deleted +0 -565

565 lines not on this page · raw message

fs/smb/client/smb1transport.c new file +565 -0

565 lines not on this page · raw message

author maintainer
#1554 lore raw

PATCH 1548/1611 cifs: SMB1 split: Add some #includes

In reply to Greg Kroah-Hartman

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

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

From: David Howells <[email protected]>

[ Upstream commit a7c7f35bcf0943ba85ab14bafbfbcbc3a30ad402 ]

Add some #includes to make sure things continue to compile as splitting occurs.

Stable-dep-of: c6394bcaf254 ("ksmbd: use opener credentials for FSCTL mutations")

---
 fs/smb/client/cifssmb.c       |    5 +++--
 fs/smb/client/smb1transport.c |    4 ++--
 2 files changed, 5 insertions(+), 4 deletions(-)
fs/smb/client/cifssmb.c +3 -2
@@ -26,11 +26,12 @@
 #include <linux/uaccess.h>
 #include <linux/netfs.h>
 #include <trace/events/netfs.h>
+#include "cifsglob.h"
+#include "cifsproto.h"
+#include "../common/smbfsctl.h"
 #include "cifspdu.h"
 #include "cifsfs.h"
-#include "cifsglob.h"
 #include "cifsacl.h"
-#include "cifsproto.h"
 #include "cifs_unicode.h"
 #include "cifs_debug.h"
 #include "fscache.h"
fs/smb/client/smb1transport.c +2 -2
@@ -22,13 +22,13 @@
 #include <linux/mempool.h>
 #include <linux/sched/signal.h>
 #include <linux/task_io_accounting_ops.h>
-#include "cifspdu.h"
 #include "cifsglob.h"
 #include "cifsproto.h"
-#include "cifs_debug.h"
+#include "cifspdu.h"
 #include "smb2proto.h"
 #include "smbdirect.h"
 #include "compress.h"
+#include "cifs_debug.h"
 
 /* Max number of iovectors we can use off the stack when sending requests. */
 #define CIFS_MAX_IOV_SIZE 8
author maintainer
#1555 lore raw

PATCH 1549/1611 ksmbd: use opener credentials for FSCTL mutations

In reply to Greg Kroah-Hartman

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

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

From: Namjae Jeon <[email protected]>

[ Upstream commit c6394bcaf254c5baf9aff43376020be5db6d3316 ]

SET_SPARSE, SET_ZERO_DATA and SET_COMPRESSION operate on an open SMB handle but call VFS xattr, fallocate or fileattr helpers with the current ksmbd worker credentials. Those helpers can revalidate inode permissions, ownership and LSM policy independently of the SMB handle access mask.

Run each operation with the credentials captured in the target file when the handle was opened. Keep credential handling local to these single-file FSCTLs rather than applying session credentials to the complete IOCTL handler, which also contains handle-less and multi-handle operations.

---
 fs/smb/server/smb2pdu.c |    3 +++
 fs/smb/server/vfs.c     |   20 +++++++++++++-------
 2 files changed, 16 insertions(+), 7 deletions(-)
fs/smb/server/smb2pdu.c +3 -0
@@ -8273,6 +8273,7 @@ static inline int fsctl_set_sparse(struc
 	if (fp->f_ci->m_fattr != old_fattr &&
 	    test_share_config_flag(work->tcon->share_conf,
 				   KSMBD_SHARE_FLAG_STORE_DOS_ATTRS)) {
+		const struct cred *saved_cred;
 		struct xattr_dos_attrib da;
 
 		ret = ksmbd_vfs_get_dos_attrib_xattr(idmap,
@@ -8281,9 +8282,11 @@ static inline int fsctl_set_sparse(struc
 			goto out;
 
 		da.attr = le32_to_cpu(fp->f_ci->m_fattr);
+		saved_cred = override_creds(fp->filp->f_cred);
 		ret = ksmbd_vfs_set_dos_attrib_xattr(idmap,
 						     &fp->filp->f_path,
 						     &da, true);
+		revert_creds(saved_cred);
 		if (ret)
 			fp->f_ci->m_fattr = old_fattr;
 	}
fs/smb/server/vfs.c +13 -7
@@ -995,15 +995,21 @@ void ksmbd_vfs_set_fadvise(struct file *
 int ksmbd_vfs_zero_data(struct ksmbd_work *work, struct ksmbd_file *fp,
 			loff_t off, loff_t len)
 {
+	const struct cred *saved_cred;
+	int err;
+
 	smb_break_all_levII_oplock(work, fp, 1);
+	saved_cred = override_creds(fp->filp->f_cred);
 	if (fp->f_ci->m_fattr & FILE_ATTRIBUTE_SPARSE_FILE_LE)
-		return vfs_fallocate(fp->filp,
-				     FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE,
-				     off, len);
-
-	return vfs_fallocate(fp->filp,
-			     FALLOC_FL_ZERO_RANGE | FALLOC_FL_KEEP_SIZE,
-			     off, len);
+		err = vfs_fallocate(fp->filp,
+				    FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE,
+				    off, len);
+	else
+		err = vfs_fallocate(fp->filp,
+				    FALLOC_FL_ZERO_RANGE | FALLOC_FL_KEEP_SIZE,
+				    off, len);
+	revert_creds(saved_cred);
+	return err;
 }
 
 int ksmbd_vfs_fqar_lseek(struct ksmbd_file *fp, loff_t start, loff_t length,