From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: stable@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
patches@lists.linux.dev, Xiang Mei <xmei5@asu.edu>,
Weiming Shi <bestswngs@gmail.com>,
Jakub Kicinski <kuba@kernel.org>
Subject: [PATCH 5.15 16/95] net: qualcomm: rmnet: fix endpoint use-after-free in rmnet_dellink()
Date: Thu, 2 Jul 2026 18:19:19 +0200 [thread overview]
Message-ID: <20260702155109.547624775@linuxfoundation.org> (raw)
In-Reply-To: <20260702155109.196223802@linuxfoundation.org>
5.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Weiming Shi <bestswngs@gmail.com>
commit d00c953a8f69921f484b629801766da68f27f658 upstream.
rmnet_dellink() removes the endpoint from the hash table with
hlist_del_init_rcu() and then immediately frees it with kfree(). However,
RCU readers on the receive path (rmnet_rx_handler ->
__rmnet_map_ingress_handler) may still hold a reference to the endpoint and
dereference ep->egress_dev after the memory has been freed. The endpoint is
a kmalloc-32 object, and the stale read at offset 8 corresponds to the
egress_dev pointer.
BUG: unable to handle page fault for address: ffffffffde942eef
Oops: 0002 [#1] SMP NOPTI
CPU: 1 UID: 0 PID: 137 Comm: poc_write Not tainted 7.0.0+ #4 PREEMPTLAZY
RIP: 0010:rmnet_vnd_rx_fixup (rmnet_vnd.c:27)
Call Trace:
<TASK>
__rmnet_map_ingress_handler (rmnet_handlers.c:48 rmnet_handlers.c:101)
rmnet_rx_handler (rmnet_handlers.c:129 rmnet_handlers.c:235)
__netif_receive_skb_core.constprop.0 (net/core/dev.c:6096)
__netif_receive_skb_one_core (net/core/dev.c:6208)
netif_receive_skb (net/core/dev.c:6467)
tun_get_user (drivers/net/tun.c:1955)
tun_chr_write_iter (drivers/net/tun.c:2003)
vfs_write (fs/read_write.c:688)
ksys_write (fs/read_write.c:740)
</TASK>
Add an rcu_head field to struct rmnet_endpoint and replace kfree() with
kfree_rcu() so the endpoint memory remains valid through the RCU grace
period. Also remove the rmnet_vnd_dellink() call and inline only the
nr_rmnet_devs decrement, since rmnet_vnd_dellink() would set
ep->egress_dev to NULL during the grace period, creating a data race
with lockless readers.
Fixes: ceed73a2cf4a ("drivers: net: ethernet: qualcomm: rmnet: Initial implementation")
Reported-by: Xiang Mei <xmei5@asu.edu>
Signed-off-by: Weiming Shi <bestswngs@gmail.com>
Link: https://patch.msgid.link/20260514122511.3083479-2-bestswngs@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
drivers/net/ethernet/qualcomm/rmnet/rmnet_config.c | 8 ++++----
drivers/net/ethernet/qualcomm/rmnet/rmnet_config.h | 1 +
2 files changed, 5 insertions(+), 4 deletions(-)
--- a/drivers/net/ethernet/qualcomm/rmnet/rmnet_config.c
+++ b/drivers/net/ethernet/qualcomm/rmnet/rmnet_config.c
@@ -205,8 +205,8 @@ static void rmnet_dellink(struct net_dev
ep = rmnet_get_endpoint(real_port, mux_id);
if (ep) {
hlist_del_init_rcu(&ep->hlnode);
- rmnet_vnd_dellink(mux_id, real_port, ep);
- kfree(ep);
+ real_port->nr_rmnet_devs--;
+ kfree_rcu(ep, rcu);
}
netdev_upper_dev_unlink(real_dev, dev);
@@ -230,9 +230,9 @@ static void rmnet_force_unassociate_devi
hash_for_each_safe(port->muxed_ep, bkt_ep, tmp_ep, ep, hlnode) {
unregister_netdevice_queue(ep->egress_dev, &list);
netdev_upper_dev_unlink(real_dev, ep->egress_dev);
- rmnet_vnd_dellink(ep->mux_id, port, ep);
hlist_del_init_rcu(&ep->hlnode);
- kfree(ep);
+ port->nr_rmnet_devs--;
+ kfree_rcu(ep, rcu);
}
rmnet_unregister_real_device(real_dev);
unregister_netdevice_many(&list);
--- a/drivers/net/ethernet/qualcomm/rmnet/rmnet_config.h
+++ b/drivers/net/ethernet/qualcomm/rmnet/rmnet_config.h
@@ -17,6 +17,7 @@ struct rmnet_endpoint {
u8 mux_id;
struct net_device *egress_dev;
struct hlist_node hlnode;
+ struct rcu_head rcu;
};
/* One instance of this structure is instantiated for each real_dev associated
next prev parent reply other threads:[~2026-07-02 16:26 UTC|newest]
Thread overview: 100+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-07-02 16:19 [PATCH 5.15 00/95] 5.15.211-rc1 review Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 01/95] fuse: limit FUSE_NOTIFY_RETRIEVE to uptodate folios Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 02/95] net/sched: act_pedit: check static offsets a priori Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 03/95] net/sched: act_pedit: rate limit datapath messages Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 04/95] net/sched: fix pedit partial COW leading to page cache corruption Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 05/95] net/sched: act_pedit: free pedit keys on bail from offset check Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 06/95] drm/amd/display: Bound VBIOS record-chain walk loops Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 07/95] ip6_vti: set netns_immutable on the fallback device Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 08/95] drm/v3d: Store the active job inside the queues state Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 09/95] drm/v3d: Skip CSD when it has zeroed workgroups Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 10/95] batman-adv: tt: reject oversized local TVLV buffers Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 11/95] batman-adv: tt: prevent TVLV entry number overflow Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 12/95] iio: light: bh1780: fix PM runtime leak on error path Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 13/95] vfio/iommu_type1: replace kfree with kvfree Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 14/95] RDMA/bnxt_re: zero shared page before exposing to userspace Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 15/95] i2c: stub: Reject I2C block transfers with invalid length Greg Kroah-Hartman
2026-07-02 16:19 ` Greg Kroah-Hartman [this message]
2026-07-02 16:19 ` [PATCH 5.15 17/95] agp/amd64: Fix broken error propagation in agp_amd64_probe() Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 18/95] xhci: fix memory leak regression when freeing xhci vdev devices depth first Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 19/95] af_unix: Reject SIOCATMARK on non-stream sockets Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 20/95] regulator: core: fix locking in regulator_resolve_supply() error path Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 21/95] vc_screen: fix null-ptr-deref in vcs_notifier() during concurrent vcs_write Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 22/95] media: vidtv: fix NULL pointer dereference in vidtv_mux_push_si Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 23/95] virtiofs: fix UAF on submount umount Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 24/95] Revert "selftest/ptp: update ptp selftest to exercise the gettimex options" Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 25/95] Revert "ptp: add testptp mask test" Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 26/95] KVM: x86/mmu: Ensure hugepage is in by slot before checking max mapping level Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 27/95] kselftest/arm64: signal: Skip SVE signal test if not enough VLs supported Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 28/95] batman-adv: tp_meter: keep unacked list in ascending ordered Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 29/95] batman-adv: tp_meter: initialize dup_acks explicitly Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 30/95] batman-adv: tp_meter: initialize dec_cwnd explicitly Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 31/95] batman-adv: tp_meter: avoid window underflow Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 32/95] batman-adv: tp_meter: avoid divide-by-zero for dec_cwnd Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 33/95] batman-adv: tp_meter: fix fast recovery precondition Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 34/95] batman-adv: tp_meter: handle seqno wrap-around for fast recovery detection Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 35/95] batman-adv: tp_meter: add only finished tp_vars to lists Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 36/95] batman-adv: bla: annotate lasttime access with READ/WRITE_ONCE Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 37/95] batman-adv: prevent ELP transmission interval underflow Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 38/95] batman-adv: tp_meter: initialize last_recv_time during init Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 39/95] batman-adv: ensure bcast is writable before modifying TTL Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 40/95] batman-adv: fix (m|b)cast csum after decrementing TTL Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 41/95] batman-adv: frag: ensure fragment is writable before modifying TTL Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 42/95] batman-adv: frag: avoid underflow of TTL Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 43/95] batman-adv: v: prevent OGM aggregation on disabled hardif Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 44/95] batman-adv: tp_meter: restrict number of unacked list entries Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 45/95] batman-adv: tp_meter: annotate last_recv_time access with READ/WRITE_ONCE Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 46/95] batman-adv: tp_meter: prevent parallel modifications of last_recv Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 47/95] batman-adv: tp_meter: handle overlapping packets Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 48/95] batman-adv: tt: dont merge change entries with different VIDs Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 49/95] batman-adv: tt: track roam count per VID Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 50/95] batman-adv: dat: prevent false sharing between VLANs Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 51/95] batman-adv: tvlv: enforce 2-byte alignment Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 52/95] batman-adv: tvlv: avoid race of cifsnotfound handler state Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 53/95] ring-buffer: Remove ring_buffer_read_prepare_sync() Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 54/95] ntfs3: reject direct userspace writes to reserved $LX* xattrs Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 55/95] ext4: add bounds check for inline data length in ext4_read_inline_page Greg Kroah-Hartman
2026-07-02 16:19 ` [PATCH 5.15 56/95] crypto: af_alg - Set merge to zero early in af_alg_sendmsg Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 57/95] mac802154: llsec: add skb_cow_data() before in-place crypto Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 58/95] KEYS: fix overflow in keyctl_pkey_params_get_2() Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 59/95] keys: Pin request_key_auth payload in instantiate paths Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 60/95] wifi: mt76: mt76x2u: Add support for ELECOM WDC-867SU3S Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 61/95] wifi: ath11k: fix warning when unbinding Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 62/95] wifi: rtlwifi: rtl8821ae: Fix C2H bit location in RX descriptor Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 63/95] f2fs: validate ACL entry sizes in f2fs_acl_from_disk() Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 64/95] bpf: use kvfree() for replaced sysctl write buffer Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 65/95] MIPS: DEC: Prevent initial console buffer from landing in XKPHYS Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 66/95] exfat: fix potential use-after-free in exfat_find_dir_entry() Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 67/95] hdlc_ppp: sync per-proto timers before freeing hdlc state Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 68/95] tipc: fix slab-use-after-free Read in tipc_aead_decrypt_done Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 69/95] pNFS: Fix use-after-free in pnfs_update_layout() Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 70/95] irqchip/imgpdc: Fix resource leak, add missing chained handler cleanup on remove Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 71/95] fpga: region: fix use-after-free in child_regions_with_firmware() Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 72/95] ocfs2: reject oversized group bitmap descriptors Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 73/95] KVM: SVM: Fix page overflow in sev_dbg_crypt() for ENCRYPT path Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 74/95] power: reset: linkstation-poweroff: fix use-after-free in the linkstation_poweroff_init() Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 75/95] fbdev: Fix fb_new_modelist to prevent null-ptr-deref in fb_videomode_to_var Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 76/95] fbdev: modedb: Fix misaligned fields in the 1920x1080-60 mode Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 77/95] NFSD: Fix SECINFO_NO_NAME decode error cleanup Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 78/95] nfsd: fix posix_acl leak on SETACL decode failure Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 79/95] nfsd: check get_user() return when reading princhashlen Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 80/95] NFSv4/pNFS: reject zero-length r_addr in nfs4_decode_mp_ds_addr Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 81/95] mptcp: fix missing wakeups in edge scenarios Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 82/95] hv: utils: handle and propagate errors in kvp_register Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 83/95] misc: fastrpc: Add dma_mask to fastrpc_channel_ctx Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 84/95] misc: fastrpc: Fix NULL pointer dereference in rpmsg callback Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 85/95] Drivers: hv: vmbus: Improve the logic of reserving fb_mmio on Gen2 VMs Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 86/95] phonet: Pass ifindex to fill_addr() Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 87/95] phonet: Pass net and ifindex to phonet_address_notify() Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 88/95] net: phonet: free phonet_device after RCU grace period Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 89/95] fuse: re-lock request before replacing page cache folio Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 90/95] ksmbd: reject non-VALID session in compound request branch Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 91/95] Documentation: ioctl-number: Extend "Include File" column width Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 92/95] crypto: qat - Replace kzalloc() + copy_from_user() with memdup_user() Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 93/95] crypto: qat - Return pointer directly in adf_ctl_alloc_resources Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 94/95] crypto: qat - remove unused character device and IOCTLs Greg Kroah-Hartman
2026-07-02 16:20 ` [PATCH 5.15 95/95] dlm: prevent NPD when writing a positive value to event_done Greg Kroah-Hartman
2026-07-02 19:46 ` [PATCH 5.15 00/95] 5.15.211-rc1 review Brett A C Sheffield
2026-07-03 6:46 ` Ron Economos
2026-07-03 9:44 ` Pavel Machek
2026-07-03 13:56 ` Mark Brown
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20260702155109.547624775@linuxfoundation.org \
--to=gregkh@linuxfoundation.org \
--cc=bestswngs@gmail.com \
--cc=kuba@kernel.org \
--cc=patches@lists.linux.dev \
--cc=stable@vger.kernel.org \
--cc=xmei5@asu.edu \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox