- Kernel - mailweb.openeuler.org

[PATCH OLK-5.10] net: sched: avoid qdisc_reset_all_tx_gt() vs dequeue race for lockless qdiscs
by Dong Chenchen 27 Apr '26

27 Apr '26

From: Koichiro Den <den(a)valinux.co.jp> mainline inclusion from mainline-v7.0-rc1 commit 7f083faf59d14c04e01ec05a7507f036c965acf8 category: bugfix bugzilla: https://atomgit.com/src-openeuler/kernel/issues/13941 CVE: CVE-2026-23340 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- When shrinking the number of real tx queues, netif_set_real_num_tx_queues() calls qdisc_reset_all_tx_gt() to flush qdiscs for queues which will no longer be used. qdisc_reset_all_tx_gt() currently serializes qdisc_reset() with qdisc_lock(). However, for lockless qdiscs, the dequeue path is serialized by qdisc_run_begin/end() using qdisc->seqlock instead, so qdisc_reset() can run concurrently with __qdisc_run() and free skbs while they are still being dequeued, leading to UAF. This can easily be reproduced on e.g. virtio-net by imposing heavy traffic while frequently changing the number of queue pairs: iperf3 -ub0 -c $peer -t 0 & while :; do ethtool -L eth0 combined 1 ethtool -L eth0 combined 2 done With KASAN enabled, this leads to reports like: BUG: KASAN: slab-use-after-free in __qdisc_run+0x133f/0x1760 ... Call Trace: <TASK> ... __qdisc_run+0x133f/0x1760 __dev_queue_xmit+0x248f/0x3550 ip_finish_output2+0xa42/0x2110 ip_output+0x1a7/0x410 ip_send_skb+0x2e6/0x480 udp_send_skb+0xb0a/0x1590 udp_sendmsg+0x13c9/0x1fc0 ... </TASK> Allocated by task 1270 on cpu 5 at 44.558414s: ... alloc_skb_with_frags+0x84/0x7c0 sock_alloc_send_pskb+0x69a/0x830 __ip_append_data+0x1b86/0x48c0 ip_make_skb+0x1e8/0x2b0 udp_sendmsg+0x13a6/0x1fc0 ... Freed by task 1306 on cpu 3 at 44.558445s: ... kmem_cache_free+0x117/0x5e0 pfifo_fast_reset+0x14d/0x580 qdisc_reset+0x9e/0x5f0 netif_set_real_num_tx_queues+0x303/0x840 virtnet_set_channels+0x1bf/0x260 [virtio_net] ethnl_set_channels+0x684/0xae0 ethnl_default_set_doit+0x31a/0x890 ... Serialize qdisc_reset_all_tx_gt() against the lockless dequeue path by taking qdisc->seqlock for TCQ_F_NOLOCK qdiscs, matching the serialization model already used by dev_reset_queue(). Additionally clear QDISC_STATE_NON_EMPTY after reset so the qdisc state reflects an empty queue, avoiding needless re-scheduling. Fixes: 6b3ba9146fe6 ("net: sched: allow qdiscs to handle locking") Signed-off-by: Koichiro Den <den(a)valinux.co.jp> Link: https://patch.msgid.link/20260228145307.3955532-1-den@valinux.co.jp Signed-off-by: Jakub Kicinski <kuba(a)kernel.org> Conflicts: include/net/sch_generic.h [commit c4fef01ba479 is not backport] Signed-off-by: Dong Chenchen <dongchenchen2(a)huawei.com> --- include/net/sch_generic.h | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/include/net/sch_generic.h b/include/net/sch_generic.h index 46afe852c455..d010017d69f1 100644 --- a/include/net/sch_generic.h +++ b/include/net/sch_generic.h @@ -751,13 +751,22 @@ static inline bool skb_skip_tc_classify(struct sk_buff *skb) static inline void qdisc_reset_all_tx_gt(struct net_device *dev, unsigned int i) { struct Qdisc *qdisc; + bool nolock; for (; i < dev->num_tx_queues; i++) { qdisc = rtnl_dereference(netdev_get_tx_queue(dev, i)->qdisc); if (qdisc) { + nolock = qdisc->flags & TCQ_F_NOLOCK; + + if (nolock) + spin_lock_bh(&qdisc->seqlock); spin_lock_bh(qdisc_lock(qdisc)); qdisc_reset(qdisc); spin_unlock_bh(qdisc_lock(qdisc)); + if (nolock) { + clear_bit(__QDISC_STATE_MISSED, &qdisc->state); + spin_unlock_bh(&qdisc->seqlock); + } } } } -- 2.25.1

2 1

[PATCH OLK-5.10 0/2] netfilter: Fix CVE-2026-23351
by Dong Chenchen 27 Apr '26

27 Apr '26

Fix CVE-2026-23351 Florian Westphal (2): netfilter: nf_tables: de-constify set commit ops function argument netfilter: nft_set_pipapo: split gc into unlink and reclaim phase include/net/netfilter/nf_tables.h | 7 +++- net/netfilter/nf_tables_api.c | 5 --- net/netfilter/nft_set_pipapo.c | 54 ++++++++++++++++++++++++------- net/netfilter/nft_set_pipapo.h | 2 ++ 4 files changed, 51 insertions(+), 17 deletions(-) -- 2.25.1

2 3

[PATCH OLK-5.10 v2] bonding: limit BOND_MODE_8023AD to Ethernet devices
by Dong Chenchen 27 Apr '26

27 Apr '26

From: Eric Dumazet <edumazet(a)google.com> mainline inclusion from mainline-v6.19-rc7 commit c84fcb79e5dbde0b8d5aeeaf04282d2149aebcf6 category: bugfix bugzilla: https://atomgit.com/src-openeuler/kernel/issues/13633 CVE: CVE-2026-23099 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- BOND_MODE_8023AD makes sense for ARPHRD_ETHER only. syzbot reported: BUG: KASAN: global-out-of-bounds in __hw_addr_create net/core/dev_addr_lists.c:63 [inline] BUG: KASAN: global-out-of-bounds in __hw_addr_add_ex+0x25d/0x760 net/core/dev_addr_lists.c:118 Read of size 16 at addr ffffffff8bf94040 by task syz.1.3580/19497 CPU: 1 UID: 0 PID: 19497 Comm: syz.1.3580 Tainted: G L syzkaller #0 PREEMPT(full) Tainted: [L]=SOFTLOCKUP Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 10/25/2025 Call Trace: <TASK> dump_stack_lvl+0xe8/0x150 lib/dump_stack.c:120 print_address_description mm/kasan/report.c:378 [inline] print_report+0xca/0x240 mm/kasan/report.c:482 kasan_report+0x118/0x150 mm/kasan/report.c:595 check_region_inline mm/kasan/generic.c:-1 [inline] kasan_check_range+0x2b0/0x2c0 mm/kasan/generic.c:200 __asan_memcpy+0x29/0x70 mm/kasan/shadow.c:105 __hw_addr_create net/core/dev_addr_lists.c:63 [inline] __hw_addr_add_ex+0x25d/0x760 net/core/dev_addr_lists.c:118 __dev_mc_add net/core/dev_addr_lists.c:868 [inline] dev_mc_add+0xa1/0x120 net/core/dev_addr_lists.c:886 bond_enslave+0x2b8b/0x3ac0 drivers/net/bonding/bond_main.c:2180 do_set_master+0x533/0x6d0 net/core/rtnetlink.c:2963 do_setlink+0xcf0/0x41c0 net/core/rtnetlink.c:3165 rtnl_changelink net/core/rtnetlink.c:3776 [inline] __rtnl_newlink net/core/rtnetlink.c:3935 [inline] rtnl_newlink+0x161c/0x1c90 net/core/rtnetlink.c:4072 rtnetlink_rcv_msg+0x7cf/0xb70 net/core/rtnetlink.c:6958 netlink_rcv_skb+0x208/0x470 net/netlink/af_netlink.c:2550 netlink_unicast_kernel net/netlink/af_netlink.c:1318 [inline] netlink_unicast+0x82f/0x9e0 net/netlink/af_netlink.c:1344 netlink_sendmsg+0x805/0xb30 net/netlink/af_netlink.c:1894 sock_sendmsg_nosec net/socket.c:727 [inline] __sock_sendmsg+0x21c/0x270 net/socket.c:742 ____sys_sendmsg+0x505/0x820 net/socket.c:2592 ___sys_sendmsg+0x21f/0x2a0 net/socket.c:2646 __sys_sendmsg+0x164/0x220 net/socket.c:2678 do_syscall_32_irqs_on arch/x86/entry/syscall_32.c:83 [inline] __do_fast_syscall_32+0x1dc/0x560 arch/x86/entry/syscall_32.c:307 do_fast_syscall_32+0x34/0x80 arch/x86/entry/syscall_32.c:332 entry_SYSENTER_compat_after_hwframe+0x84/0x8e </TASK> The buggy address belongs to the variable: lacpdu_mcast_addr+0x0/0x40 Fixes: 872254dd6b1f ("net/bonding: Enable bonding to enslave non ARPHRD_ETHER") Reported-by: syzbot+9c081b17773615f24672(a)syzkaller.appspotmail.com Closes: https://lore.kernel.org/netdev/6966946b.a70a0220.245e30.0002.GAE@google.com… Signed-off-by: Eric Dumazet <edumazet(a)google.com> Cc: Andrew Lunn <andrew+netdev(a)lunn.ch> Acked-by: Jay Vosburgh <jv(a)jvosburgh.net> Link: https://patch.msgid.link/20260113191201.3970737-1-edumazet@google.com Signed-off-by: Jakub Kicinski <kuba(a)kernel.org> Conflicts: drivers/net/bonding/bond_main.c [commit 6569fa2d4e014 is not backport] Signed-off-by: Dong Chenchen <dongchenchen2(a)huawei.com> --- drivers/net/bonding/bond_main.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c index 6eb2db8d07c0..f0029195bce4 100644 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@ -1799,6 +1799,12 @@ int bond_enslave(struct net_device *bond_dev, struct net_device *slave_dev, */ if (!bond_has_slaves(bond)) { if (bond_dev->type != slave_dev->type) { + if (slave_dev->type != ARPHRD_ETHER && + BOND_MODE(bond) == BOND_MODE_8023AD) { + NL_SET_ERR_MSG(extack, "8023AD mode requires Ethernet devices"); + slave_err(bond_dev, slave_dev, "Error: 8023AD mode requires Ethernet devices\n"); + return -EINVAL; + } slave_dbg(bond_dev, slave_dev, "change device type from %d to %d\n", bond_dev->type, slave_dev->type); -- 2.25.1

2 1

[PATCH OLK-5.10] media: dvb-net: fix OOB access in ULE extension header tables
by Liu Kai 25 Apr '26

25 Apr '26

From: Ariel Silver <arielsilver77(a)gmail.com> mainline inclusion from mainline-v7.0-rc3 commit 24d87712727a5017ad142d63940589a36cd25647 category: bugfix bugzilla: https://atomgit.com/src-openeuler/kernel/issues/14132 CVE: CVE-2026-31405 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- The ule_mandatory_ext_handlers[] and ule_optional_ext_handlers[] tables in handle_one_ule_extension() are declared with 255 elements (valid indices 0-254), but the index htype is derived from network-controlled data as (ule_sndu_type & 0x00FF), giving a range of 0-255. When htype equals 255, an out-of-bounds read occurs on the function pointer table, and the OOB value may be called as a function pointer. Add a bounds check on htype against the array size before either table is accessed. Out-of-range values now cause the SNDU to be discarded. Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Reported-by: Ariel Silver <arielsilver77(a)gmail.com> Signed-off-by: Ariel Silver <arielsilver77(a)gmail.com> Cc: stable(a)vger.kernel.org Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei(a)kernel.org> Signed-off-by: Liu Kai <liukai284(a)huawei.com> --- drivers/media/dvb-core/dvb_net.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/media/dvb-core/dvb_net.c b/drivers/media/dvb-core/dvb_net.c index c594b1bdfcaa..c8cbe901bcf0 100644 --- a/drivers/media/dvb-core/dvb_net.c +++ b/drivers/media/dvb-core/dvb_net.c @@ -228,6 +228,9 @@ static int handle_one_ule_extension( struct dvb_net_priv *p ) unsigned char hlen = (p->ule_sndu_type & 0x0700) >> 8; unsigned char htype = p->ule_sndu_type & 0x00FF; + if (htype >= ARRAY_SIZE(ule_mandatory_ext_handlers)) + return -1; + /* Discriminate mandatory and optional extension headers. */ if (hlen == 0) { /* Mandatory extension header */ -- 2.34.1

2 1

[PATCH OLK-5.10] ksmbd: do not expire session on binding failure
by Li Lingfeng 25 Apr '26

25 Apr '26

From: Hyunwoo Kim <imv4bel(a)gmail.com> mainline inclusion from mainline-v7.0-rc6 commit 9bbb19d21ded7d78645506f20d8c44895e3d0fb9 category: bugfix bugzilla: https://atomgit.com/src-openeuler/kernel/issues/14217 CVE: CVE-2026-31476 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- When a multichannel session binding request fails (e.g. wrong password), the error path unconditionally sets sess->state = SMB2_SESSION_EXPIRED. However, during binding, sess points to the target session looked up via ksmbd_session_lookup_slowpath() -- which belongs to another connection's user. This allows a remote attacker to invalidate any active session by simply sending a binding request with a wrong password (DoS). Fix this by skipping session expiration when the failed request was a binding attempt, since the session does not belong to the current connection. The reference taken by ksmbd_session_lookup_slowpath() is still correctly released via ksmbd_user_session_put(). Cc: stable(a)vger.kernel.org Signed-off-by: Hyunwoo Kim <imv4bel(a)gmail.com> Acked-by: Namjae Jeon <linkinjeon(a)kernel.org> Signed-off-by: Steve French <stfrench(a)microsoft.com> Conflicts: fs/smb/server/smb2pdu.c [Commit 38c8a9a52082 ("smb: move client and server files to common directory fs/smb") move smb2pdu.c from fs/ksmbd to fs/smb/server.] Signed-off-by: Li Lingfeng <lilingfeng3(a)huawei.com> --- fs/ksmbd/smb2pdu.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/fs/ksmbd/smb2pdu.c b/fs/ksmbd/smb2pdu.c index d75de3035327..176a6a51be6d 100644 --- a/fs/ksmbd/smb2pdu.c +++ b/fs/ksmbd/smb2pdu.c @@ -1912,8 +1912,14 @@ int smb2_sess_setup(struct ksmbd_work *work) if (sess->user && sess->user->flags & KSMBD_USER_FLAG_DELAY_SESSION) try_delay = true; - sess->last_active = jiffies; - sess->state = SMB2_SESSION_EXPIRED; + /* + * For binding requests, session belongs to another + * connection. Do not expire it. + */ + if (!(req->Flags & SMB2_SESSION_REQ_FLAG_BINDING)) { + sess->last_active = jiffies; + sess->state = SMB2_SESSION_EXPIRED; + } if (try_delay) { ksmbd_conn_set_need_reconnect(conn); ssleep(5); -- 2.52.0

2 1

[PATCH OLK-5.10] nvme-pci: ensure we're polling a polled queue
by Li Lingfeng 25 Apr '26

25 Apr '26

From: Keith Busch <kbusch(a)kernel.org> stable inclusion from stable-v5.10.253 commit 965e2c943f065122f14282a88d70a8a92e12a4da category: bugfix bugzilla: https://atomgit.com/src-openeuler/kernel/issues/14256 CVE: CVE-2026-31523 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- [ Upstream commit 166e31d7dbf6aa44829b98aa446bda5c9580f12a ] A user can change the polled queue count at run time. There's a brief window during a reset where a hipri task may try to poll that queue before the block layer has updated the queue maps, which would race with the now interrupt driven queue and may cause double completions. Reviewed-by: Christoph Hellwig <hch(a)lst.de> Reviewed-by: Kanchan Joshi <joshi.k(a)samsung.com> Signed-off-by: Keith Busch <kbusch(a)kernel.org> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Li Lingfeng <lilingfeng3(a)huawei.com> --- drivers/nvme/host/pci.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c index 5591ecff6ee8..8be07d3c99a5 100644 --- a/drivers/nvme/host/pci.c +++ b/drivers/nvme/host/pci.c @@ -1102,7 +1102,8 @@ static int nvme_poll(struct blk_mq_hw_ctx *hctx) struct nvme_queue *nvmeq = hctx->driver_data; bool found; - if (!nvme_cqe_pending(nvmeq)) + if (!test_bit(NVMEQ_POLLED, &nvmeq->flags) || + !nvme_cqe_pending(nvmeq)) return 0; spin_lock(&nvmeq->cq_poll_lock); -- 2.52.0

2 1

[PATCH OLK-6.6] nvme-pci: ensure we're polling a polled queue
by Li Lingfeng 25 Apr '26

25 Apr '26

From: Keith Busch <kbusch(a)kernel.org> stable inclusion from stable-v6.6.131 commit 6f12734c4b619f923a4df0b1a46b8098b187d324 category: bugfix bugzilla: https://atomgit.com/src-openeuler/kernel/issues/14256 CVE: CVE-2026-31523 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- [ Upstream commit 166e31d7dbf6aa44829b98aa446bda5c9580f12a ] A user can change the polled queue count at run time. There's a brief window during a reset where a hipri task may try to poll that queue before the block layer has updated the queue maps, which would race with the now interrupt driven queue and may cause double completions. Reviewed-by: Christoph Hellwig <hch(a)lst.de> Reviewed-by: Kanchan Joshi <joshi.k(a)samsung.com> Signed-off-by: Keith Busch <kbusch(a)kernel.org> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Li Lingfeng <lilingfeng3(a)huawei.com> --- drivers/nvme/host/pci.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c index d01712c3b144..da4bb0da4f27 100644 --- a/drivers/nvme/host/pci.c +++ b/drivers/nvme/host/pci.c @@ -1131,7 +1131,8 @@ static int nvme_poll(struct blk_mq_hw_ctx *hctx, struct io_comp_batch *iob) struct nvme_queue *nvmeq = hctx->driver_data; bool found; - if (!nvme_cqe_pending(nvmeq)) + if (!test_bit(NVMEQ_POLLED, &nvmeq->flags) || + !nvme_cqe_pending(nvmeq)) return 0; spin_lock(&nvmeq->cq_poll_lock); -- 2.52.0

2 1

[PATCH OLK-5.10] ext4: avoid infinite loops caused by residual data
by Li Lingfeng 25 Apr '26

25 Apr '26

From: Edward Adam Davis <eadavis(a)qq.com> stable inclusion from stable-v6.6.131 commit ecc50bfca9b5c2ee6aeef998181689b80477367b category: bugfix bugzilla: https://atomgit.com/src-openeuler/kernel/issues/14191 CVE: CVE-2026-31448 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- commit 5422fe71d26d42af6c454ca9527faaad4e677d6c upstream. On the mkdir/mknod path, when mapping logical blocks to physical blocks, if inserting a new extent into the extent tree fails (in this example, because the file system disabled the huge file feature when marking the inode as dirty), ext4_ext_map_blocks() only calls ext4_free_blocks() to reclaim the physical block without deleting the corresponding data in the extent tree. This causes subsequent mkdir operations to reference the previously reclaimed physical block number again, even though this physical block is already being used by the xattr block. Therefore, a situation arises where both the directory and xattr are using the same buffer head block in memory simultaneously. The above causes ext4_xattr_block_set() to enter an infinite loop about "inserted" and cannot release the inode lock, ultimately leading to the 143s blocking problem mentioned in [1]. If the metadata is corrupted, then trying to remove some extent space can do even more harm. Also in case EXT4_GET_BLOCKS_DELALLOC_RESERVE was passed, remove space wrongly update quota information. Jan Kara suggests distinguishing between two cases: 1) The error is ENOSPC or EDQUOT - in this case the filesystem is fully consistent and we must maintain its consistency including all the accounting. However these errors can happen only early before we've inserted the extent into the extent tree. So current code works correctly for this case. 2) Some other error - this means metadata is corrupted. We should strive to do as few modifications as possible to limit damage. So I'd just skip freeing of allocated blocks. [1] INFO: task syz.0.17:5995 blocked for more than 143 seconds. Call Trace: inode_lock_nested include/linux/fs.h:1073 [inline] __start_dirop fs/namei.c:2923 [inline] start_dirop fs/namei.c:2934 [inline] Reported-by: syzbot+512459401510e2a9a39f(a)syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=1659aaaaa8d9d11265d7 Tested-by: syzbot+1659aaaaa8d9d11265d7(a)syzkaller.appspotmail.com Reported-by: syzbot+1659aaaaa8d9d11265d7(a)syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=512459401510e2a9a39f Tested-by: syzbot+1659aaaaa8d9d11265d7(a)syzkaller.appspotmail.com Signed-off-by: Edward Adam Davis <eadavis(a)qq.com> Reviewed-by: Jan Kara <jack(a)suse.cz> Tested-by: syzbot+512459401510e2a9a39f(a)syzkaller.appspotmail.com Link: https://patch.msgid.link/tencent_43696283A68450B761D76866C6F360E36705@qq.com Signed-off-by: Theodore Ts'o <tytso(a)mit.edu> Cc: stable(a)kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Conflicts: fs/ext4/extents.c [Commit fb138df7d886 ("ext4: get rid of ppath in ext4_ext_insert_extent()") change the way to judge the result of ext4_ext_insert_extent() in ext4_ext_map_blocks().] Signed-off-by: Li Lingfeng <lilingfeng3(a)huawei.com> --- fs/ext4/extents.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 4c77284be84d..7045784f9340 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -4373,9 +4373,13 @@ int ext4_ext_map_blocks(handle_t *handle, struct inode *inode, err = ext4_ext_insert_extent(handle, inode, &path, &newex, flags); if (err) { - if (allocated_clusters) { + /* + * Gracefully handle out of space conditions. If the filesystem + * is inconsistent, we'll just leak allocated blocks to avoid + * causing even more damage. + */ + if (allocated_clusters && (err == -EDQUOT || err == -ENOSPC)) { int fb_flags = 0; - /* * free data blocks we just allocated. * not a good idea to call discard here directly, -- 2.52.0

2 1

[PATCH OLK-6.6] ext4: avoid infinite loops caused by residual data
by Li Lingfeng 25 Apr '26

25 Apr '26

From: Edward Adam Davis <eadavis(a)qq.com> stable inclusion from stable-v6.6.131 commit ecc50bfca9b5c2ee6aeef998181689b80477367b category: bugfix bugzilla: https://atomgit.com/src-openeuler/kernel/issues/14191 CVE: CVE-2026-31448 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- commit 5422fe71d26d42af6c454ca9527faaad4e677d6c upstream. On the mkdir/mknod path, when mapping logical blocks to physical blocks, if inserting a new extent into the extent tree fails (in this example, because the file system disabled the huge file feature when marking the inode as dirty), ext4_ext_map_blocks() only calls ext4_free_blocks() to reclaim the physical block without deleting the corresponding data in the extent tree. This causes subsequent mkdir operations to reference the previously reclaimed physical block number again, even though this physical block is already being used by the xattr block. Therefore, a situation arises where both the directory and xattr are using the same buffer head block in memory simultaneously. The above causes ext4_xattr_block_set() to enter an infinite loop about "inserted" and cannot release the inode lock, ultimately leading to the 143s blocking problem mentioned in [1]. If the metadata is corrupted, then trying to remove some extent space can do even more harm. Also in case EXT4_GET_BLOCKS_DELALLOC_RESERVE was passed, remove space wrongly update quota information. Jan Kara suggests distinguishing between two cases: 1) The error is ENOSPC or EDQUOT - in this case the filesystem is fully consistent and we must maintain its consistency including all the accounting. However these errors can happen only early before we've inserted the extent into the extent tree. So current code works correctly for this case. 2) Some other error - this means metadata is corrupted. We should strive to do as few modifications as possible to limit damage. So I'd just skip freeing of allocated blocks. [1] INFO: task syz.0.17:5995 blocked for more than 143 seconds. Call Trace: inode_lock_nested include/linux/fs.h:1073 [inline] __start_dirop fs/namei.c:2923 [inline] start_dirop fs/namei.c:2934 [inline] Reported-by: syzbot+512459401510e2a9a39f(a)syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=1659aaaaa8d9d11265d7 Tested-by: syzbot+1659aaaaa8d9d11265d7(a)syzkaller.appspotmail.com Reported-by: syzbot+1659aaaaa8d9d11265d7(a)syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=512459401510e2a9a39f Tested-by: syzbot+1659aaaaa8d9d11265d7(a)syzkaller.appspotmail.com Signed-off-by: Edward Adam Davis <eadavis(a)qq.com> Reviewed-by: Jan Kara <jack(a)suse.cz> Tested-by: syzbot+512459401510e2a9a39f(a)syzkaller.appspotmail.com Link: https://patch.msgid.link/tencent_43696283A68450B761D76866C6F360E36705@qq.com Signed-off-by: Theodore Ts'o <tytso(a)mit.edu> Cc: stable(a)kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Conflicts: fs/ext4/extents.c [Commit fb138df7d886 ("ext4: get rid of ppath in ext4_ext_insert_extent()") change the way to judge the result of ext4_ext_insert_extent() in ext4_ext_map_blocks().] Signed-off-by: Li Lingfeng <lilingfeng3(a)huawei.com> --- fs/ext4/extents.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index dbc2154f7d4e..997c437685a0 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -4381,9 +4381,13 @@ int ext4_ext_map_blocks(handle_t *handle, struct inode *inode, err = ext4_ext_insert_extent(handle, inode, &path, &newex, flags); if (err) { - if (allocated_clusters) { + /* + * Gracefully handle out of space conditions. If the filesystem + * is inconsistent, we'll just leak allocated blocks to avoid + * causing even more damage. + */ + if (allocated_clusters && (err == -EDQUOT || err == -ENOSPC)) { int fb_flags = 0; - /* * free data blocks we just allocated. * not a good idea to call discard here directly, -- 2.52.0

2 1

[PATCH OLK-6.6] ksmbd: fix OOB write in QUERY_INFO for compound requests
by Li Lingfeng 24 Apr '26

24 Apr '26

From: Asim Viladi Oglu Manizada <manizada(a)pm.me> mainline inclusion from mainline-v7.0-rc7 commit fda9522ed6afaec45cabc198d8492270c394c7bc category: bugfix bugzilla: https://atomgit.com/src-openeuler/kernel/issues/14175 CVE: CVE-2026-31402 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- When a compound request such as READ + QUERY_INFO(Security) is received, and the first command (READ) consumes most of the response buffer, ksmbd could write beyond the allocated buffer while building a security descriptor. The root cause was that smb2_get_info_sec() checked buffer space using ppntsd_size from xattr, while build_sec_desc() often synthesized a significantly larger descriptor from POSIX ACLs. This patch introduces smb_acl_sec_desc_scratch_len() to accurately compute the final descriptor size beforehand, performs proper buffer checking with smb2_calc_max_out_buf_len(), and uses exact-sized allocation + iov pinning. Cc: stable(a)vger.kernel.org Fixes: e2b76ab8b5c9 ("ksmbd: add support for read compound") Signed-off-by: Asim Viladi Oglu Manizada <manizada(a)pm.me> Signed-off-by: Namjae Jeon <linkinjeon(a)kernel.org> Signed-off-by: Steve French <stfrench(a)microsoft.com> Conflicts: fs/smb/server/smb2pdu.c [Commit 0066f623bce8 ("ksmbd: use __GFP_RETRY_MAYFAIL") change the way to alloc smb_ntsd.] Signed-off-by: Li Lingfeng <lilingfeng3(a)huawei.com> --- fs/smb/server/smb2pdu.c | 121 +++++++++++++++++++++++++++++----------- fs/smb/server/smbacl.c | 43 ++++++++++++++ fs/smb/server/smbacl.h | 2 + 3 files changed, 134 insertions(+), 32 deletions(-) diff --git a/fs/smb/server/smb2pdu.c b/fs/smb/server/smb2pdu.c index 3681006c7ac8..ea7d467bd5cf 100644 --- a/fs/smb/server/smb2pdu.c +++ b/fs/smb/server/smb2pdu.c @@ -3369,20 +3369,24 @@ int smb2_open(struct ksmbd_work *work) KSMBD_SHARE_FLAG_ACL_XATTR)) { struct smb_fattr fattr; struct smb_ntsd *pntsd; - int pntsd_size, ace_num = 0; + int pntsd_size; + size_t scratch_len; ksmbd_acls_fattr(&fattr, idmap, inode); - if (fattr.cf_acls) - ace_num = fattr.cf_acls->a_count; - if (fattr.cf_dacls) - ace_num += fattr.cf_dacls->a_count; - - pntsd = kmalloc(sizeof(struct smb_ntsd) + - sizeof(struct smb_sid) * 3 + - sizeof(struct smb_acl) + - sizeof(struct smb_ace) * ace_num * 2, - GFP_KERNEL); + scratch_len = smb_acl_sec_desc_scratch_len(&fattr, + NULL, 0, + OWNER_SECINFO | GROUP_SECINFO | + DACL_SECINFO); + if (!scratch_len || scratch_len == SIZE_MAX) { + rc = -EFBIG; + posix_acl_release(fattr.cf_acls); + posix_acl_release(fattr.cf_dacls); + goto err_out; + } + + pntsd = kvzalloc(scratch_len, GFP_KERNEL); if (!pntsd) { + rc = -ENOMEM; posix_acl_release(fattr.cf_acls); posix_acl_release(fattr.cf_dacls); goto err_out; @@ -3397,7 +3401,7 @@ int smb2_open(struct ksmbd_work *work) posix_acl_release(fattr.cf_acls); posix_acl_release(fattr.cf_dacls); if (rc) { - kfree(pntsd); + kvfree(pntsd); goto err_out; } @@ -3407,7 +3411,7 @@ int smb2_open(struct ksmbd_work *work) pntsd, pntsd_size, false); - kfree(pntsd); + kvfree(pntsd); if (rc) pr_err("failed to store ntacl in xattr : %d\n", rc); @@ -5301,8 +5305,9 @@ static int smb2_get_info_file(struct ksmbd_work *work, if (test_share_config_flag(work->tcon->share_conf, KSMBD_SHARE_FLAG_PIPE)) { /* smb2 info file called for pipe */ - return smb2_get_info_file_pipe(work->sess, req, rsp, + rc = smb2_get_info_file_pipe(work->sess, req, rsp, work->response_buf); + goto iov_pin_out; } if (work->next_smb2_rcv_hdr_off) { @@ -5402,6 +5407,12 @@ static int smb2_get_info_file(struct ksmbd_work *work, rc = buffer_check_err(le32_to_cpu(req->OutputBufferLength), rsp, work->response_buf); ksmbd_fd_put(work, fp); + +iov_pin_out: + if (!rc) + rc = ksmbd_iov_pin_rsp(work, (void *)rsp, + offsetof(struct smb2_query_info_rsp, Buffer) + + le32_to_cpu(rsp->OutputBufferLength)); return rc; } @@ -5621,6 +5632,11 @@ static int smb2_get_info_filesystem(struct ksmbd_work *work, rc = buffer_check_err(le32_to_cpu(req->OutputBufferLength), rsp, work->response_buf); path_put(&path); + + if (!rc) + rc = ksmbd_iov_pin_rsp(work, (void *)rsp, + offsetof(struct smb2_query_info_rsp, Buffer) + + le32_to_cpu(rsp->OutputBufferLength)); return rc; } @@ -5630,13 +5646,14 @@ static int smb2_get_info_sec(struct ksmbd_work *work, { struct ksmbd_file *fp; struct mnt_idmap *idmap; - struct smb_ntsd *pntsd = (struct smb_ntsd *)rsp->Buffer, *ppntsd = NULL; + struct smb_ntsd *pntsd = NULL, *ppntsd = NULL; struct smb_fattr fattr = {{0}}; struct inode *inode; __u32 secdesclen = 0; unsigned int id = KSMBD_NO_FID, pid = KSMBD_NO_FID; int addition_info = le32_to_cpu(req->AdditionalInformation); - int rc = 0, ppntsd_size = 0; + int rc = 0, ppntsd_size = 0, max_len; + size_t scratch_len = 0; if (addition_info & ~(OWNER_SECINFO | GROUP_SECINFO | DACL_SECINFO | PROTECTED_DACL_SECINFO | @@ -5644,6 +5661,11 @@ static int smb2_get_info_sec(struct ksmbd_work *work, ksmbd_debug(SMB, "Unsupported addition info: 0x%x)\n", addition_info); + pntsd = kzalloc(ALIGN(sizeof(struct smb_ntsd), 8), + GFP_KERNEL); + if (!pntsd) + return -ENOMEM; + pntsd->revision = cpu_to_le16(1); pntsd->type = cpu_to_le16(SELF_RELATIVE | DACL_PROTECTED); pntsd->osidoffset = 0; @@ -5652,9 +5674,7 @@ static int smb2_get_info_sec(struct ksmbd_work *work, pntsd->dacloffset = 0; secdesclen = sizeof(struct smb_ntsd); - rsp->OutputBufferLength = cpu_to_le32(secdesclen); - - return 0; + goto iov_pin; } if (work->next_smb2_rcv_hdr_off) { @@ -5686,18 +5706,58 @@ static int smb2_get_info_sec(struct ksmbd_work *work, &ppntsd); /* Check if sd buffer size exceeds response buffer size */ - if (smb2_resp_buf_len(work, 8) > ppntsd_size) - rc = build_sec_desc(idmap, pntsd, ppntsd, ppntsd_size, - addition_info, &secdesclen, &fattr); + max_len = smb2_calc_max_out_buf_len(work, + offsetof(struct smb2_query_info_rsp, Buffer), + le32_to_cpu(req->OutputBufferLength)); + if (max_len < 0) { + rc = -EINVAL; + goto release_acl; + } + + scratch_len = smb_acl_sec_desc_scratch_len(&fattr, ppntsd, + ppntsd_size, addition_info); + if (!scratch_len || scratch_len == SIZE_MAX) { + rc = -EFBIG; + goto release_acl; + } + + pntsd = kvzalloc(scratch_len, GFP_KERNEL); + if (!pntsd) { + rc = -ENOMEM; + goto release_acl; + } + + rc = build_sec_desc(idmap, pntsd, ppntsd, ppntsd_size, + addition_info, &secdesclen, &fattr); + +release_acl: posix_acl_release(fattr.cf_acls); posix_acl_release(fattr.cf_dacls); kfree(ppntsd); ksmbd_fd_put(work, fp); + + if (!rc && ALIGN(secdesclen, 8) > scratch_len) + rc = -EFBIG; if (rc) - return rc; + goto err_out; +iov_pin: rsp->OutputBufferLength = cpu_to_le32(secdesclen); - return 0; + rc = buffer_check_err(le32_to_cpu(req->OutputBufferLength), + rsp, work->response_buf); + if (rc) + goto err_out; + + rc = ksmbd_iov_pin_rsp_read(work, (void *)rsp, + offsetof(struct smb2_query_info_rsp, Buffer), + pntsd, secdesclen); +err_out: + if (rc) { + rsp->OutputBufferLength = 0; + kvfree(pntsd); + } + + return rc; } /** @@ -5721,6 +5781,9 @@ int smb2_query_info(struct ksmbd_work *work) goto err_out; } + rsp->StructureSize = cpu_to_le16(9); + rsp->OutputBufferOffset = cpu_to_le16(72); + switch (req->InfoType) { case SMB2_O_INFO_FILE: ksmbd_debug(SMB, "GOT SMB2_O_INFO_FILE\n"); @@ -5741,14 +5804,6 @@ int smb2_query_info(struct ksmbd_work *work) } ksmbd_revert_fsids(work); - if (!rc) { - rsp->StructureSize = cpu_to_le16(9); - rsp->OutputBufferOffset = cpu_to_le16(72); - rc = ksmbd_iov_pin_rsp(work, (void *)rsp, - offsetof(struct smb2_query_info_rsp, Buffer) + - le32_to_cpu(rsp->OutputBufferLength)); - } - err_out: if (rc < 0) { if (rc == -EACCES) @@ -5759,6 +5814,8 @@ int smb2_query_info(struct ksmbd_work *work) rsp->hdr.Status = STATUS_UNEXPECTED_IO_ERROR; else if (rc == -ENOMEM) rsp->hdr.Status = STATUS_INSUFFICIENT_RESOURCES; + else if (rc == -EINVAL && rsp->hdr.Status == 0) + rsp->hdr.Status = STATUS_INVALID_PARAMETER; else if (rc == -EOPNOTSUPP || rsp->hdr.Status == 0) rsp->hdr.Status = STATUS_INVALID_INFO_CLASS; smb2_set_err_rsp(work); diff --git a/fs/smb/server/smbacl.c b/fs/smb/server/smbacl.c index 07d31dee99c0..b7818241e295 100644 --- a/fs/smb/server/smbacl.c +++ b/fs/smb/server/smbacl.c @@ -915,6 +915,49 @@ int parse_sec_desc(struct mnt_idmap *idmap, struct smb_ntsd *pntsd, return 0; } +size_t smb_acl_sec_desc_scratch_len(struct smb_fattr *fattr, + struct smb_ntsd *ppntsd, int ppntsd_size, int addition_info) +{ + size_t len = sizeof(struct smb_ntsd); + size_t tmp; + + if (addition_info & OWNER_SECINFO) + len += sizeof(struct smb_sid); + if (addition_info & GROUP_SECINFO) + len += sizeof(struct smb_sid); + if (!(addition_info & DACL_SECINFO)) + return len; + + len += sizeof(struct smb_acl); + if (ppntsd && ppntsd_size > 0) { + unsigned int dacl_offset = le32_to_cpu(ppntsd->dacloffset); + + if (dacl_offset < ppntsd_size && + check_add_overflow(len, ppntsd_size - dacl_offset, &len)) + return 0; + } + + if (fattr->cf_acls) { + if (check_mul_overflow((size_t)fattr->cf_acls->a_count, + 2 * sizeof(struct smb_ace), &tmp) || + check_add_overflow(len, tmp, &len)) + return 0; + } else { + /* default/minimum DACL */ + if (check_add_overflow(len, 5 * sizeof(struct smb_ace), &len)) + return 0; + } + + if (fattr->cf_dacls) { + if (check_mul_overflow((size_t)fattr->cf_dacls->a_count, + sizeof(struct smb_ace), &tmp) || + check_add_overflow(len, tmp, &len)) + return 0; + } + + return len; +} + /* Convert permission bits from mode to equivalent CIFS ACL */ int build_sec_desc(struct mnt_idmap *idmap, struct smb_ntsd *pntsd, struct smb_ntsd *ppntsd, diff --git a/fs/smb/server/smbacl.h b/fs/smb/server/smbacl.h index 2b52861707d8..64e6f8ebb68e 100644 --- a/fs/smb/server/smbacl.h +++ b/fs/smb/server/smbacl.h @@ -210,6 +210,8 @@ int set_info_sec(struct ksmbd_conn *conn, struct ksmbd_tree_connect *tcon, bool type_check, bool get_write); void id_to_sid(unsigned int cid, uint sidtype, struct smb_sid *ssid); void ksmbd_init_domain(u32 *sub_auth); +size_t smb_acl_sec_desc_scratch_len(struct smb_fattr *fattr, + struct smb_ntsd *ppntsd, int ppntsd_size, int addition_info); static inline uid_t posix_acl_uid_translate(struct mnt_idmap *idmap, struct posix_acl_entry *pace) -- 2.52.0

2 1