commit f8001e1a51 upstream.
HT_caps_handler() iterates pIE->length bytes and writes into
HT_caps.u.HT_cap[], which is a fixed 26-byte array (sizeof struct
HT_caps_element). Because pIE->length is a raw u8 from an over-the-air
802.11 AssocResponse frame and is never validated, a malicious AP can
set it up to 255, causing up to 229 bytes of out-of-bounds writes into
adjacent fields of struct mlme_ext_info.
Truncate the iteration count to the size of HT_caps.u.HT_cap using
umin() so that data from a longer-than-expected IE is silently ignored
rather than written out of bounds, preserving interoperability with APs
that pad the element. An early return on oversized IEs was considered
but rejected: it would bypass the pmlmeinfo->HT_caps_enable = 1
assignment that precedes the loop, silently disabling HT mode for APs
that append extra bytes to the HT Capabilities IE.
Fixes: 554c0a3abf ("staging: Add rtl8723bs sdio wifi driver")
Cc: stable <stable@kernel.org>
Signed-off-by: Alexandru Hossu <hossu.alexandru@gmail.com>
Reviewed-by: Luka Gejak <luka.gejak@linux.dev>
Link: https://patch.msgid.link/20260522004531.1038924-5-hossu.alexandru@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 3bf39f711f upstream.
The loop in is_ap_in_tkip() iterates over IEs without verifying that
enough bytes remain before dereferencing the IE header or its payload:
- pIE->element_id and pIE->length are read without checking that
i + sizeof(*pIE) <= ie_length, so a truncated IE at the end of the
buffer causes an OOB read.
- For WLAN_EID_VENDOR_SPECIFIC the code compares pIE->data + 12,
which requires pIE->length >= 16. For WLAN_EID_RSN it compares
pIE->data + 8, requiring pIE->length >= 12. Neither requirement
is checked.
Add the missing IE header and payload bounds checks and guard each
data access with an explicit pIE->length minimum, matching the
pattern established in update_beacon_info().
Fixes: 554c0a3abf ("staging: Add rtl8723bs sdio wifi driver")
Cc: stable <stable@kernel.org>
Signed-off-by: Alexandru Hossu <hossu.alexandru@gmail.com>
Link: https://patch.msgid.link/20260522004531.1038924-7-hossu.alexandru@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit ef61d628df upstream.
Two IE parsing loops are missing the header bounds checks before they
dereference pIE->length:
- issue_assocreq() walks pmlmeinfo->network.ies to build the
association request. If the stored IE data ends with only an
element_id byte and no length byte, pIE->length is read one byte
past the end of the buffer.
- join_cmd_hdl() walks pnetwork->ies during station join and has
the same problem under the same conditions.
Both buffers are filled from AP beacon and probe-response frames, so a
malicious AP that sends a truncated final IE can trigger the issue.
Apply the two-guard pattern established in update_beacon_info():
1. Break if fewer than sizeof(*pIE) bytes remain.
2. Break if the IE's declared data extends past the buffer end.
Fixes: 554c0a3abf ("staging: Add rtl8723bs sdio wifi driver")
Cc: stable <stable@kernel.org>
Reviewed-by: Luka Gejak <luka.gejak@linux.dev>
Signed-off-by: Alexandru Hossu <hossu.alexandru@gmail.com>
Link: https://patch.msgid.link/20260522004531.1038924-3-hossu.alexandru@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit ed51de4a86 upstream.
The IE parsing loop in update_beacon_info() advances by
(pIE->length + 2) each iteration but only guards on i < len.
When a malicious AP sends a Beacon whose last IE has only one byte
remaining in the frame (the element_id byte lands at len-1), the loop
reads pIE->length from one byte past the allocated receive buffer.
Additionally, even when the header bytes are in bounds, pIE->length
itself can extend the data window beyond len, passing a truncated IE
to the handler functions.
Add two guards at the top of the loop body:
1. Break if fewer than sizeof(*pIE) bytes remain (can't read header).
2. Break if the IE's declared data extends past len.
Also replace i += (pIE->length + 2) with i += sizeof(*pIE) + pIE->length
for consistency with the sizeof(*pIE) guards added above.
Fixes: 554c0a3abf ("staging: Add rtl8723bs sdio wifi driver")
Cc: stable <stable@kernel.org>
Reviewed-by: Luka Gejak <luka.gejak@linux.dev>
Signed-off-by: Alexandru Hossu <hossu.alexandru@gmail.com>
Link: https://patch.msgid.link/20260522004531.1038924-2-hossu.alexandru@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit f9654207e9 upstream.
The IE parsing loop in OnAssocRsp() advances by (pIE->length + 2) each
iteration but only guards on i < pkt_len. When a malicious AP sends an
AssocResponse whose last IE has only one byte remaining in the frame
(the element_id byte lands at pkt_len-1), the loop reads pIE->length
from pframe[pkt_len], which is one byte past the allocated receive buffer.
Additionally, even when the header bytes are in bounds, pIE->length
itself can extend the data window beyond pkt_len, silently passing a
truncated IE to the handler functions.
Add two guards at the top of the loop body:
1. Break if fewer than sizeof(*pIE) bytes remain (can't read header).
2. Break if the IE's declared data extends past pkt_len.
Fixes: 554c0a3abf ("staging: Add rtl8723bs sdio wifi driver")
Cc: stable <stable@kernel.org>
Signed-off-by: Alexandru Hossu <hossu.alexandru@gmail.com>
Reviewed-by: Luka Gejak <luka.gejak@linux.dev>
Link: https://patch.msgid.link/20260522004531.1038924-6-hossu.alexandru@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit a1fc19d61f upstream.
OnAuth() has two bugs in the shared-key authentication path.
When the Privacy bit is set, rtw_wep_decrypt() is called without
verifying that the frame is long enough to contain a valid WEP IV and
ICV. Inside rtw_wep_decrypt(), length is computed as:
length = len - WLAN_HDR_A3_LEN - iv_len
and then passed as (length - 4) to crc32_le(). If len is less than
WLAN_HDR_A3_LEN + iv_len + icv_len (32 bytes), length - 4 is negative
and, after the implicit cast to size_t, causes crc32_le() to read far
beyond the frame buffer. Add a minimum length check before accessing
the IV field and calling the decryption path.
When processing a seq=3 response, rtw_get_ie() stores the Challenge
Text IE length in ie_len, but the subsequent memcmp() always reads 128
bytes regardless of ie_len. IEEE 802.11 mandates a challenge text of
exactly 128 bytes; reject any IE whose length field differs, matching
the check already applied to OnAuthClient().
Fixes: 554c0a3abf ("staging: Add rtl8723bs sdio wifi driver")
Cc: stable <stable@kernel.org>
Signed-off-by: Alexandru Hossu <hossu.alexandru@gmail.com>
Link: https://patch.msgid.link/20260522004605.1039209-1-hossu.alexandru@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 5a752a616e upstream.
supplicant_ie is a 256-byte array in struct security_priv. The WPA and
WPA2 IE copy paths use:
memcpy(padapter->securitypriv.supplicant_ie, &pwpa[0], wpa_ielen + 2);
where wpa_ielen is the raw IE length field (u8, 0-255). When a local user
supplies a connect request via nl80211 with a crafted WPA IE of length 255,
wpa_ielen + 2 equals 257, overflowing the 256-byte buffer by one byte into
the adjacent last_mic_err_time field.
rtw_parse_wpa_ie() does not prevent this: its length consistency check
compares *(wpa_ie+1) against (u8)(wpa_ie_len-2), which is (u8)(255) == 255
when wpa_ie_len = 257, so the check passes silently.
Add explicit bounds checks for both the WPA and WPA2 paths before the
memcpy, rejecting any IE whose total size (wpa_ielen + 2) exceeds the
supplicant_ie buffer.
Fixes: 554c0a3abf ("staging: Add rtl8723bs sdio wifi driver")
Cc: stable <stable@kernel.org>
Reviewed-by: Luka Gejak <luka.gejak@linux.dev>
Signed-off-by: Alexandru Hossu <hossu.alexandru@gmail.com>
Link: https://patch.msgid.link/20260522004531.1038924-4-hossu.alexandru@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit d3a9a8cf2d upstream.
In both ipu7_isys_init() and ipu7_psys_init(), pdata is allocated and
then passed to ipu7_bus_initialize_device(), which stores it in
adev->pdata. The ipu7_bus_release() function frees adev->pdata when the
device's reference count drops to zero.
Two error paths incorrectly call kfree(pdata) after the device teardown
has already freed it:
1. When ipu7_mmu_init() fails: put_device() is called, which drops the
reference count to zero and triggers ipu7_bus_release() ->
kfree(pdata). The subsequent kfree(pdata) is a double-free.
2. When ipu7_bus_add_device() fails: it calls auxiliary_device_uninit()
internally, which calls put_device() -> ipu7_bus_release() ->
kfree(pdata). The subsequent kfree(pdata) is again a double-free.
Note that the kfree(pdata) when ipu7_bus_initialize_device() itself
fails is correct, because in that case auxiliary_device_init() failed
and the release function was never set up, so pdata must be freed
manually.
Additionally, the error code was not saved before calling put_device(),
causing ERR_CAST() to dereference the already-freed adev pointer when
constructing the return value. Fix this by saving the error from
dev_err_probe() before put_device() and returning ERR_PTR() instead.
Remove the redundant kfree(pdata) calls and fix the use-after-free in
the return values of the two affected error paths.
Fixes: b7fe4c0019 ("media: staging/ipu7: add Intel IPU7 PCI device driver")
Cc: stable@vger.kernel.org
Reviewed-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: Alexandru Hossu <hossu.alexandru@gmail.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit f4d51e55dd upstream.
The load_primary_binaries() function is overly complex and has som large
variables on the stack, which can cause warnings depending on CONFIG_FRAME_WARN
setting:
drivers/staging/media/atomisp/pci/sh_css.c: In function 'load_primary_binaries':
drivers/staging/media/atomisp/pci/sh_css.c:5260:1: error: the frame size of 1560 bytes is larger than 1536 bytes [-Werror=frame-larger-than=]
Half of the stack usage is for the prim_descr[] array, but only one
member of the array is used at any given time.
Reduce the stack usage by turning the array into a single structure.
Fixes: a49d25364d ("staging/atomisp: Add support for the Intel IPU v2")
Cc: stable@vger.kernel.org
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@intel.com>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit c32fe4c491 upstream.
If the driver's stripe information is invalid it can result in an integer
underflow. Add a range check to avoid this kind of error.
This patch fixes the following smatch error:
drivers/staging/media/ipu3/ipu3-css-params.c:1792 imgu_css_cfg_acc_stripe() warn: 'acc->stripe.bds_out_stripes[0]->width - 2 * f' 4294967168 can't fit into 65535 'acc->stripe.bds_out_stripes[1]->offset'
Cc: stable@vger.kernel.org
Fixes: e11110a5b7 ("media: staging/intel-ipu3: css: Compute and program ccs")
Signed-off-by: Ricardo Ribalda <ribalda@chromium.org>
Signed-off-by: Hans Verkuil <hverkuil+cisco@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 151edde741 upstream.
tsi148_probe() allocates a location monitor resource and links it into
tsi148_bridge->lm_resources. The probe error path frees this list, but
tsi148_remove() only frees the dma, slave and master resource lists, so
the location monitor resource is leaked on device unbind or module
unload.
Free the lm_resources list in tsi148_remove() as well, before
tsi148_bridge is freed.
Fixes: d22b8ed9a3 ("Staging: vme: add Tundra TSI148 VME-PCI Bridge driver")
Cc: stable <stable@kernel.org>
Cc: Martyn Welch <martyn@welchs.me.uk>
Assisted-by: Claude:claude-fable-5
Signed-off-by: Hao-Qun Huang <alvinhuang0603@gmail.com>
Link: https://patch.msgid.link/20260704065817.403111-2-alvinhuang0603@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit e8422d89e8 upstream.
fake_init() allocates a location monitor resource and links it into
fake_bridge->lm_resources. The init error path frees this list, but
fake_exit() only frees the slave and master resource lists. Loading
and unloading the module therefore triggers a kmemleak warning:
unreferenced object 0xffff8b8b82aebe40 (size 64):
comm "init", pid 1, jiffies 4294894572
backtrace (crc c1e013ef):
kmemleak_alloc+0x4e/0x90
__kmalloc_cache_noprof+0x338/0x430
0xffffffffc0602246
do_one_initcall+0x4f/0x320
do_init_module+0x68/0x270
load_module+0x2a3b/0x2d90
Free the lm_resources list in fake_exit() as well, before fake_bridge
is freed.
Fixes: 658bcdae9c ("vme: Adding Fake VME driver")
Cc: stable <stable@kernel.org>
Cc: Martyn Welch <martyn@welchs.me.uk>
Assisted-by: Claude:claude-fable-5
Signed-off-by: Hao-Qun Huang <alvinhuang0603@gmail.com>
Link: https://patch.msgid.link/20260704065817.403111-1-alvinhuang0603@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 53b7c271f0 upstream.
smb2_check_message() has a long-standing quirk that accepts a response
whose calculated length is one byte larger than the bytes actually
received ("server can return one byte more due to implied bcc[0]").
This was introduced to accommodate servers that omit the trailing bcc[0]
overlap byte when no data area is present.
However, the exemption is applied unconditionally, regardless of whether
the command actually carries a data area (has_smb2_data_area[]). When a
response with a data area is subject to the +1 exemption, the reported
data can extend one byte beyond the bytes actually received, yet
smb2_check_message() still accepts it. The subsequent decoder then reads
past the end of the receive buffer. This is reachable during NEGOTIATE
and SESSION_SETUP, before the session is established.
The resulting out-of-bounds reads are visible under KASAN when mounting
against a non-conforming server; both the SPNEGO/negTokenInit and the
NTLMSSP challenge decoders are affected:
BUG: KASAN: slab-out-of-bounds in asn1_ber_decoder+0x16a7/0x1b00
Read of size 1 at addr ffff8880084d67c0 by task mount.cifs/81
CPU: 1 UID: 0 PID: 81 Comm: mount.cifs Not tainted 7.1.0-rc6 #1
Call Trace:
<TASK>
dump_stack_lvl+0x4e/0x70
print_report+0x157/0x4c9
kasan_report+0xce/0x100
asn1_ber_decoder+0x16a7/0x1b00
decode_negTokenInit+0x19/0x30
SMB2_negotiate+0x31d9/0x4c90
cifs_negotiate_protocol+0x1f2/0x3f0
cifs_get_smb_ses+0x93f/0x17e0
cifs_mount_get_session+0x7f/0x3a0
cifs_mount+0xb4/0xcf0
cifs_smb3_do_mount+0x23a/0x1500
smb3_get_tree+0x3b0/0x630
vfs_get_tree+0x82/0x2d0
fc_mount+0x10/0x1b0
path_mount+0x50d/0x1de0
__x64_sys_mount+0x20b/0x270
do_syscall_64+0xee/0x590
entry_SYSCALL_64_after_hwframe+0x77/0x7f
</TASK>
Allocated by task 85:
kmem_cache_alloc_noprof+0x106/0x380
mempool_alloc_noprof+0x116/0x1e0
cifs_small_buf_get+0x31/0x80
allocate_buffers+0x10d/0x2b0
cifs_demultiplex_thread+0x1d5/0x1d50
kthread+0x2c6/0x390
ret_from_fork+0x36e/0x5a0
ret_from_fork_asm+0x1a/0x30
The buggy address is located 0 bytes to the right of
allocated 448-byte region [ffff8880084d6600, ffff8880084d67c0)
which belongs to the cache cifs_small_rq of size 448
BUG: KASAN: slab-out-of-bounds in kmemdup_noprof+0x36/0x50
Read of size 329 at addr ffff88800726c678 by task mount.cifs/89
CPU: 0 UID: 0 PID: 89 Comm: mount.cifs Tainted: G B 7.1.0-rc6 #1
Call Trace:
<TASK>
dump_stack_lvl+0x4e/0x70
print_report+0x157/0x4c9
kasan_report+0xce/0x100
kasan_check_range+0x10f/0x1e0
__asan_memcpy+0x23/0x60
kmemdup_noprof+0x36/0x50
decode_ntlmssp_challenge+0x457/0x680
SMB2_sess_auth_rawntlmssp_negotiate+0x6f0/0xcb0
SMB2_sess_setup+0x219/0x4f0
cifs_setup_session+0x248/0xaf0
cifs_get_smb_ses+0xf79/0x17e0
cifs_mount_get_session+0x7f/0x3a0
cifs_mount+0xb4/0xcf0
cifs_smb3_do_mount+0x23a/0x1500
smb3_get_tree+0x3b0/0x630
vfs_get_tree+0x82/0x2d0
fc_mount+0x10/0x1b0
path_mount+0x50d/0x1de0
__x64_sys_mount+0x20b/0x270
do_syscall_64+0xee/0x590
entry_SYSCALL_64_after_hwframe+0x77/0x7f
</TASK>
Allocated by task 93:
kmem_cache_alloc_noprof+0x106/0x380
mempool_alloc_noprof+0x116/0x1e0
cifs_small_buf_get+0x31/0x80
allocate_buffers+0x10d/0x2b0
cifs_demultiplex_thread+0x1d5/0x1d50
kthread+0x2c6/0x390
ret_from_fork+0x36e/0x5a0
ret_from_fork_asm+0x1a/0x30
The buggy address is located 120 bytes inside of
allocated 448-byte region [ffff88800726c600, ffff88800726c7c0)
which belongs to the cache cifs_small_rq of size 448
Restrict the +1 exemption to responses that have no data area, so that
it still covers the bcc[0] omission it was meant for. When a data area
is present, the +1 discrepancy instead means the reported data length
overruns the received buffer, so the response must be rejected.
Fixes: 093b2bdad3 ("CIFS: Make demultiplex_thread work with SMB2 code")
Cc: stable@vger.kernel.org
Signed-off-by: Shoichiro Miyamoto <shoichiro.miyamoto@gmail.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 9f32f38265 upstream.
The SLAVE-path helpers buffer_to_user() and buffer_from_user() copy
'count' bytes into/out of the fixed-size kern_buf (size_buf ==
PCI_BUF_SIZE == 0x20000, 128 KiB) using *ppos as the offset, without
bounding *ppos + count against size_buf.
vme_user_write()/vme_user_read() only clamp count to the VME window size
(image_size = vme_get_size(resource)), which VME_SET_SLAVE sets from the
user-supplied slave.size -- validated against the VME address space (up
to VME_A32_MAX = 4 GiB), not against PCI_BUF_SIZE. When the window
exceeds 128 KiB, a write()/read() copies past the kern_buf allocation.
Clamp count against size_buf in both helpers, with an early return when
*ppos is already at/after the buffer end. *ppos is >= 0 here (the caller
rejects negative offsets), so size_buf - *ppos cannot wrap. This mirrors
the existing clamp in the MASTER-path helpers resource_to_user() /
resource_from_user(), and matches the read()/write() convention of a
short transfer at end-of-buffer.
Found by static analysis (CodeQL taint tracking + CBMC bounded model
checking) and confirmed dynamically under KASAN with the vme_fake bridge:
BUG: KASAN: slab-out-of-bounds in _copy_from_user+0x2d/0x80
Write of size 262144 at addr ffff888004100000 by task trigger/68
_copy_from_user+0x2d/0x80
vme_user_write+0x13e/0x240 [vme_user]
vfs_write+0x1b8/0x7a0
ksys_write+0xb8/0x150
Fixes: f00a86d98a ("Staging: vme: add VME userspace driver")
Cc: stable <stable@kernel.org>
Signed-off-by: Michael Tautschnig <tautschn@amazon.com>
Link: https://patch.msgid.link/20260618114709.72499-1-tautschn@amazon.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 2b66974a1b upstream.
A broadcast PROTOCOL/STATE_MSG can carry a Gap ACK blocks record in its
data area. tipc_get_gap_ack_blks() only verifies that the record's len
field is self-consistent with its ugack_cnt/bgack_cnt counts
(sz == struct_size(p, gacks, ugack_cnt + bgack_cnt)); it does not check
that the record actually fits in the message data area, msg_data_sz().
The unicast caller tipc_link_proto_rcv() bounds it ("if (glen > dlen)
break;"), but the broadcast caller tipc_bcast_sync_rcv() discards the
returned size, so tipc_link_advance_transmq() copies the record off the
receive skb with an attacker-controlled count:
this_ga = kmemdup(ga, struct_size(ga, gacks, ga->bgack_cnt),
GFP_ATOMIC);
A TIPC neighbour that negotiated TIPC_GAP_ACK_BLOCK triggers it with one
ordinary broadcast STATE_MSG (msg_bc_ack_invalid() clear), sized so its
data area is short, carrying a Gap ACK record with len = 0x400,
bgack_cnt = 0xff and ugack_cnt = 0. len then equals
struct_size(p, gacks, 255), so the consistency check passes and ga is
non-NULL; kmemdup() reads struct_size(ga, gacks, 255) = 1024 bytes out
of the much smaller skb:
BUG: KASAN: slab-out-of-bounds in kmemdup_noprof+0x48/0x60
Read of size 1024 at addr ffff0000c7030d38 by task poc864/69
Call trace:
kmemdup_noprof+0x48/0x60
tipc_link_advance_transmq+0x86c/0xb80
tipc_link_bc_ack_rcv+0x19c/0x1e0
tipc_bcast_sync_rcv+0x1c4/0x2c4
tipc_rcv+0x85c/0x1340
tipc_l2_rcv_msg+0xac/0x104
The buggy address belongs to the object at ffff0000c7030d00
which belongs to the cache skbuff_small_head of size 704
The buggy address is located 56 bytes inside of
allocated 704-byte region [ffff0000c7030d00, ffff0000c7030fc0)
The copied-out bytes are subsequently consumed as gap/ack values, but
the read is already out of bounds at the kmemdup() regardless of how
they are used.
The unicast STATE path drops such a message: "if (glen > dlen) break;"
skips the rest of STATE_MSG handling and the skb is freed. Make the
broadcast path drop it too. tipc_bcast_sync_rcv() now bounds the record
against msg_data_sz() and, when it does not fit, reports it back through
tipc_node_bc_sync_rcv() to tipc_rcv() so the skb is discarded rather than
processed. ga is not cleared on this path: ga == NULL already means
"legacy peer without Selective ACK", a distinct legitimate state.
Fixes: d7626b5acf ("tipc: introduce Gap ACK blocks for broadcast link")
Cc: stable@vger.kernel.org
Signed-off-by: Samuel Page <sam@bynar.io>
Reviewed-by: Tung Nguyen <tung.quang.nguyen@est.tech>
Link: https://patch.msgid.link/20260625143815.1525412-1-sam@bynar.io
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 1720db928e upstream.
lowpan_nhc_do_uncompression() looks up an NHC descriptor while holding
lowpan_nhc_lock. If the descriptor has no uncompress callback, the error
path drops the lock before printing nhc->name.
lowpan_nhc_del() removes descriptors under the same lock and then relies
on synchronize_net() before the owning module can be unloaded. That only
waits for net RX RCU readers. lowpan_header_decompress() is also exported
and can be reached from callers that are not necessarily covered by the net
core RX critical section, for example the Bluetooth 6LoWPAN L2CAP receive
path.
This leaves a race where one task drops lowpan_nhc_lock in the error path,
another task unregisters and frees the matching descriptor after
synchronize_net() returns, and the first task then dereferences nhc->name
for the warning.
With the post-unlock window widened, KASAN reports:
BUG: KASAN: slab-use-after-free in lowpan_nhc_do_uncompression+0x1f4/0x220
Read of size 8
lowpan_nhc_do_uncompression
lowpan_header_decompress
Fix this by printing the warning before dropping lowpan_nhc_lock, so the
descriptor name is read while unregister is still excluded. The malformed
packet is still rejected with -ENOTSUPP.
Fixes: 92aa7c65d2 ("6lowpan: add generic nhc layer interface")
Cc: stable@vger.kernel.org
Reported-by: Yizhou Zhao <zhaoyz24@mails.tsinghua.edu.cn>
Reported-by: Yuxiang Yang <yangyx22@mails.tsinghua.edu.cn>
Reported-by: Ao Wang <wangao@seu.edu.cn>
Reported-by: Xuewei Feng <fengxw06@126.com>
Reported-by: Qi Li <qli01@tsinghua.edu.cn>
Reported-by: Ke Xu <xuke@tsinghua.edu.cn>
Signed-off-by: Yizhou Zhao <zhaoyz24@mails.tsinghua.edu.cn>
Acked-by: Alexander Aring <aahringo@redhat.com>
Link: https://patch.msgid.link/20260609080054.4541-1-zhaoyz24@mails.tsinghua.edu.cn
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 010382937f upstream.
dwc3_gadget_suspend() takes dwc->lock with IRQs disabled and then calls
dwc3_disconnect_gadget(). For async callbacks that helper only uses
plain spin_unlock()/spin_lock(), so the gadget ->disconnect() callback
still runs with IRQs disabled and any sleepable callback trips Lockdep.
This issue was found by our static analysis tool and then manually
reviewed against the current tree.
The grounded PoC kept the dwc3_gadget_suspend() ->
dwc3_disconnect_gadget() -> gadget_driver->disconnect() chain, and
Lockdep reported:
BUG: sleeping function called from invalid context
gadget_disconnect+0x21/0x39 [vuln_msv]
dwc3_gadget_suspend.constprop.0+0x2b/0x42 [vuln_msv]
Keep the disconnect callback selection in one common helper, but add a
sleepable suspend-side wrapper which snapshots the callback under
dwc->lock and then runs it after spin_unlock_irqrestore(). The regular
event path still uses the existing spin_unlock()/spin_lock() window.
Fixes: c8540870af ("usb: dwc3: gadget: Improve dwc3_gadget_suspend() and dwc3_gadget_resume()")
Cc: stable <stable@kernel.org>
Signed-off-by: Runyu Xiao <runyu.xiao@seu.edu.cn>
Acked-by: Thinh Nguyen <Thinh.Nguyen@synopsys.com>
Link: https://patch.msgid.link/20260612052005.3849659-1-runyu.xiao@seu.edu.cn
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit abf76d3239 upstream.
The chaoskey driver has a use-after-free bug in its release routine.
If the user closes the device file after the USB device has been
unplugged, a debugging log statement will try to access the
usb_interface structure after it has been deallocated:
BUG: KASAN: slab-use-after-free in dev_driver_string (drivers/base/core.c:2406)
Read of size 8 at addr ffff888168e8a0b8 by task chaoskey_raw_re/10106
Hardware name: QEMU Ubuntu 24.04 PC v2 (i440FX + PIIX, arch_caps fix, 1996), BIOS 1.16.3-debian-1.16.3-2 04/01/2014
Call Trace:
<TASK>
dump_stack_lvl (lib/dump_stack.c:94 lib/dump_stack.c:120)
print_report (mm/kasan/report.c:378 mm/kasan/report.c:482)
kasan_report (mm/kasan/report.c:595)
dev_driver_string (drivers/base/core.c:2406)
__dynamic_dev_dbg (lib/dynamic_debug.c:906)
chaoskey_release (drivers/usb/misc/chaoskey.c:323)
__fput (fs/file_table.c:510)
fput_close_sync (fs/file_table.c:615)
__x64_sys_close (fs/open.c:1507 fs/open.c:1492 fs/open.c:1492)
do_syscall_64 (arch/x86/entry/syscall_64.c:63 arch/x86/entry/syscall_64.c:94)
entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:121)
The driver's last reference to the interface structure is dropped in
the chaoskey_free() routine, so the code must not use the interface --
even in a debugging statement -- after that routine returns.
(Exception: If we know that another reference is held by someone else,
such as the device core while the disconnect routine runs, there's no
problem. Thanks to Johan Hovold for pointing this out.)
Since the bad access is part of an unimportant debugging statement,
we can fix the problem simply by removing the whole statement.
Reported-by: Shuangpeng Bai <shuangpeng.kernel@gmail.com>
Closes: https://lore.kernel.org/linux-usb/20EC9664-054E-438B-B411-2145D347F97B@gmail.com/
Tested-by: Shuangpeng Bai <shuangpeng.kernel@gmail.com>
Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Fixes: 66e3e59189 ("usb: Add driver for Altus Metrum ChaosKey device (v2)")
Cc: stable <stable@kernel.org>
Reviewed-by: Johan Hovold <johan@kernel.org>
Link: https://patch.msgid.link/bb5b1dc6-eb59-43e1-8d26-51e658e88bbe@rowland.harvard.edu
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit e3046eeada upstream.
random_recv_done() stores the device-reported used.len directly into
vi->data_avail. copy_data() then indexes vi->data[] using
vi->data_idx (advanced by previous copy_data() calls) and issues a
memcpy() without re-validating either value against the posted
buffer size sizeof(vi->data) (SMP_CACHE_BYTES bytes, typically 32
or 64).
A malicious or buggy virtio-rng backend can set used.len beyond
sizeof(vi->data), steering the memcpy() past the end of the inline
array into adjacent kmalloc-1k slab bytes. hwrng_fillfn() mixes
those bytes into the guest RNG, and guest root can also observe
them directly via /dev/hwrng.
Concrete impact is inside the guest:
- Memory-safety / hardening: any virtio-rng backend that
over-reports used.len causes the driver to read past vi->data
into unrelated slab contents. hwrng_fillfn() is a kernel thread
that runs as soon as the device is probed; no guest userspace
interaction is required to first-trigger the OOB.
- Cross-boundary leak (confidential-compute threat model): a
malicious hypervisor cooperating with a malicious or compromised
guest root userspace can use /dev/hwrng as a leak channel for
guest-kernel heap data. The host sets a large used.len, guest
root reads /dev/hwrng, and the returned bytes contain guest
kernel slab contents that were adjacent to vi->data. In
practice, confidential-compute guests (SEV-SNP, TDX) usually
disable virtio-rng entirely, so this path is narrow, but the
fix is still worth carrying because the underlying
memory-safety bug contaminates the guest RNG on any host.
KASAN confirms the OOB on a 7.1-rc4 guest whose virtio-rng backend
has been patched to report used.len = 0x10000:
BUG: KASAN: slab-out-of-bounds in virtio_read+0x394/0x5d0
Read of size 64 at addr ffff88800ae0ba20 by task hwrng/52
Call Trace:
__asan_memcpy+0x23/0x60
virtio_read+0x394/0x5d0
hwrng_fillfn+0xb2/0x470
kthread+0x2cc/0x3a0
Allocated by task 1:
probe_common+0xa5/0x660
virtio_dev_probe+0x549/0xbc0
The buggy address belongs to the object at ffff88800ae0b800
which belongs to the cache kmalloc-1k of size 1024
The buggy address is located 0 bytes to the right of
allocated 544-byte region [ffff88800ae0b800, ffff88800ae0ba20)
Same class of bug as commit c04db81cd0 ("net/9p: Fix buffer
overflow in USB transport layer"), which hardened
usb9pfs_rx_complete() against unchecked device-reported length in
the USB 9p transport.
With the clamp at point of use and array_index_nospec() in place,
the same harness boots cleanly: copy_data() returns zero for the
bogus report, the device-supplied bytes after data_idx are
discarded, and the driver issues a fresh request.
Fixes: f7f510ec19 ("virtio: An entropy device, as suggested by hpa.")
Cc: stable@vger.kernel.org
Suggested-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael Bommarito <michael.bommarito@gmail.com>
Assisted-by: Claude:claude-opus-4-8
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Message-ID: <20260531142251.2792061-1-michael.bommarito@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit c687bc3569 upstream.
Driver core expects devices to be allocated dynamically and complains
loudly when a device that lacks a release function is freed.
Use __root_device_register() to allocate and register the root device
instead of open coding using a static device.
Note that root_device_register(), which also creates a link to the
module, cannot be used as the device is registered when parsing the
module parameters which happens before the module kobject has been set
up.
Fixes: 81a054ce0b ("virtio-mmio: Devices parameter parsing")
Cc: stable@vger.kernel.org # 3.5
Cc: Pawel Moll <pawel.moll@arm.com>
Signed-off-by: Johan Hovold <johan@kernel.org>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Message-ID: <20260427143710.14702-1-johan@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit f7d380fb52 upstream.
Unbinding a virtio balloon device:
echo virtio0 > /sys/bus/virtio/drivers/virtio_balloon/unbind
triggers a NULL pointer dereference. The dmesg says:
BUG: kernel NULL pointer dereference, address: 0000000000000008
[...]
RIP: 0010:__list_del_entry_valid_or_report+0x5/0xf0
Call Trace:
<TASK>
vp_del_vqs+0x121/0x230
remove_common+0x135/0x150
virtballoon_remove+0xee/0x100
virtio_dev_remove+0x3b/0x80
device_release_driver_internal+0x187/0x2c0
unbind_store+0xb9/0xe0
kernfs_fop_write_iter.llvm.11660790530567441834+0xf6/0x180
vfs_write+0x2a9/0x3b0
ksys_write+0x5c/0xd0
do_syscall_64+0x54/0x230
entry_SYSCALL_64_after_hwframe+0x29/0x31
[...]
</TASK>
The virtio_balloon device registers 5 queues (inflate, deflate, stats,
free_page, reporting) but only the first two are unconditional. The
stats, free_page and reporting queues are each conditional on their
respective feature bits. When any of these features are absent, the
corresponding vqs_info entry has name == NULL, creating holes in the
array.
The root cause is an indexing mismatch introduced when vq info storage
was changed to be passed as an argument. vp_find_vqs_msix() and
vp_find_vqs_intx() store the info pointer at vp_dev->vqs[i], where 'i'
is the caller's sparse array index. However, the virtqueue itself gets
vq->index assigned from queue_idx, a dense index that skips NULL
entries. When holes exist, 'i' and queue_idx diverge. Later,
vp_del_vqs() looks up info via vp_dev->vqs[vq->index] using the dense
index into the sparsely-populated array, and hits NULL.
Fix this by storing info at vp_dev->vqs[queue_idx] instead of
vp_dev->vqs[i], so the store index matches the lookup index
(vq->index). Apply the fix to both the MSIX and INTX paths.
Cc: Yichun Zhang <yichun@openresty.com>
Cc: Jiri Pirko <jiri@nvidia.com>
Cc: stable@vger.kernel.org # v6.11+
Tested-by: Yuka <yuka@umeyashiki.org>
Fixes: 89a1c435ae ("virtio_pci: pass vq info as an argument to vp_setup_vq()")
Signed-off-by: Ammar Faizi <ammarfaizi2@openresty.com>
Message-Id: <20260315141808.547081-1-ammarfaizi2@openresty.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 7cd9103283 upstream.
The release path of ip_set_dump_do() and ip_set_dump_done() read
inst->ip_set_list via ip_set_ref_netlink(), a plain rcu_dereference_raw()
of the array pointer. These run from netlink_recvmsg() without the nfnl
mutex and without an RCU read-side critical section.
A concurrent ip_set_create() can grow the array: it publishes the new
array, calls synchronize_net() and then kvfree()s the old one. Since the
dump paths read the array outside any RCU reader, synchronize_net() does
not wait for them and the old array can be freed while they still index
into it, causing a use-after-free.
The dumped set itself stays pinned via set->ref_netlink, so only the
array load needs protecting. Take rcu_read_lock() around it, matching
ip_set_get_byname() and __ip_set_put_byindex().
BUG: KASAN: slab-use-after-free in ip_set_dump_do (net/netfilter/ipset/ip_set_core.c:1697)
Read of size 8 at addr ffff88800b5c4018 by task exploit/150
Call Trace:
...
kasan_report (mm/kasan/report.c:595)
ip_set_dump_do (net/netfilter/ipset/ip_set_core.c:1697)
netlink_dump (net/netlink/af_netlink.c:2325)
netlink_recvmsg (net/netlink/af_netlink.c:1976)
sock_recvmsg (net/socket.c:1159)
__sys_recvfrom (net/socket.c:2315)
...
Oops: general protection fault, probably for non-canonical address ... KASAN NOPTI
KASAN: maybe wild-memory-access in range [0x02d6...d0-0x02d6...d7]
RIP: 0010:ip_set_dump_do (net/netfilter/ipset/ip_set_core.c:1698)
Kernel panic - not syncing: Fatal exception
Fixes: 8a02bdd50b ("netfilter: ipset: Fix calling ip_set() macro at dumping")
Cc: stable@vger.kernel.org
Reported-by: Weiming Shi <bestswngs@gmail.com>
Assisted-by: Claude:claude-opus-4-8
Signed-off-by: Xiang Mei <xmei5@asu.edu>
Acked-by: Jozsef Kadlecsik <kadlec@netfilter.org>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 35d4a3cf70 upstream.
Fix three issues in damon_hot_score() that comes from wrong handling of
extreme (zero or too high) monitoring intervals user setup.
When the user sets sampling interval zero, damon_max_nr_accesses(), which
is called from damon_hot_score(), causes a divide-by-zero. Needless to
say, it is a problem.
When the user sets the aggregation interval zero, the function returns
zero. It is wrong, since the real maximum nr_acceses in the setup should
be one. Worse yet, it can cause another divide-by-zero from its caller,
damon_hot_score(), since it uses damon_max_nr_accesses() return value as a
denominator.
When the user sets the aggregation interval very high, damon_hot_score()
could return a value out of [0, DAMOS_MAX_SCORE] range. Since the return
value is used as an index to the regions_score_histogram array, which is
DAMOS_MAX_SCORE+1 size, it causes out of bounds array access.
The issues can be relatively easily reproduced like below. The sysfs
write permission is required, though.
# ./damo start --damos_action lru_prio --damos_quota_space 100M \
--damos_quota_interval 1s
# cd /sys/kernel/mm/damon/admin/kdamonds/0
# echo 0 > contexts/0/monitoring_attrs/intervals/sample_us
# echo 0 > contexts/0/monitoring_attrs/intervals/aggr_us
# echo commit > state
# dmesg
[...]
[ 131.329762] Oops: divide error: 0000 [#1] SMP NOPTI
[...]
[ 131.336089] RIP: 0010:damon_hot_score+0x27/0xd0
[...]
Fix the divide-by-zero intervals problems by explicitly handling the zero
intervals in damon_max_nr_accesses(). Fix the out-of-bound array access
by applying [0, DAMOS_MAX_SCORE] bounds before returning from
damon_hot_score().
The issue was discovered [1] by Sashiko.
Link: https://lore.kernel.org/20260623135834.67189-1-sj@kernel.org
Link: https://lore.kernel.org/20260619202459.145010-1-sj@kernel.org [1]
Fixes: 198f0f4c58 ("mm/damon/vaddr,paddr: support pageout prioritization")
Signed-off-by: SeongJae Park <sj@kernel.org>
Cc: <stable@vger.kernel.org> # 5.16.x
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 8bc4d43bcc upstream.
Commit 51e547e8c8 ("tcp: Free TCP-AO/TCP-MD5 info/keys without RCU")
removed the call_rcu() callback from tcp_ao_destroy_sock(), arguing that
"the destruction of info/keys is delayed until the socket destructor"
and therefore "no one can discover it anymore".
That argument does not hold for the call site in tcp_connect()
(net/ipv4/tcp_output.c:4327-4332). At that point the socket is in
TCP_SYN_SENT, has already been inserted into the inet ehash by
inet_hash_connect() in tcp_v4_connect(), and is therefore very much
discoverable: any softirq running tcp_v4_rcv() on another CPU can take
the socket out of the ehash, walk into tcp_inbound_hash(), and load
tp->ao_info via implicit RCU before bh_lock_sock_nested() is taken on
the destroying CPU.
The reader path then enters __tcp_ao_do_lookup() (net/ipv4/tcp_ao.c:208)
which re-loads tp->ao_info via rcu_dereference_check(); the re-load can
still observe the (about-to-be-freed) pointer because there is no
synchronize_rcu() between rcu_assign_pointer(tp->ao_info, NULL) and
tcp_ao_info_free() in tcp_ao_destroy_sock(). The captured pointer is
then walked at line 223:
hlist_for_each_entry_rcu(key, &ao->head, node, ...)
The writer's synchronous kfree() is free to complete between the line
218 re-fetch and the line 223 hlist iteration. The slab is reused
(or simply LIST_POISON1-stamped if not yet reused) and the iteration
walks attacker-controlled or poison memory in softirq context.
Reproducer (no debug shim, stock x86_64 v7.1-rc2 SMP+KASAN, QEMU+KVM):
an unprivileged uid=1000 process inside CLONE_NEWUSER|CLONE_NEWNET
installs TCP_MD5SIG + TCP_AO_ADD_KEY on a TCP socket, sprays forged
TCP-AO segments toward its eventual 4-tuple via raw sockets, then
calls connect(). The md5-wins reconciliation in tcp_connect() fires
tcp_ao_destroy_sock(); the softirq backlog reader on the loopback
NAPI path crashes on the freed ao->head.first walk:
Oops: general protection fault, probably for non-canonical
address 0xfbd59c000000002f
KASAN: maybe wild-memory-access in range
[0xdead000000000178-0xdead00000000017f]
CPU: 0 UID: 1000 PID: 100 Comm: repro_userns
RIP: 0010:__tcp_ao_do_lookup+0x107/0x1c0
Call Trace: <IRQ>
__tcp_ao_do_lookup+0x107/0x1c0
tcp_ao_inbound_lookup.constprop.0+0x12a/0x200
tcp_inbound_ao_hash+0x5ea/0x1520
tcp_inbound_hash+0x7ce/0x1240
tcp_v4_rcv+0x1e7a/0x3e10
...
Restore the RCU grace period: re-add struct rcu_head to tcp_ao_info
and replace the synchronous tcp_ao_info_free() with a call_rcu()
callback. Readers that captured tp->ao_info before rcu_assign_pointer
NULLed it now see the object remain valid until rcu_read_unlock().
With the patch applied the reproducer runs cleanly for 2000 iterations
on the same kernel build.
Fixes: 51e547e8c8 ("tcp: Free TCP-AO/TCP-MD5 info/keys without RCU")
Cc: stable@vger.kernel.org # v6.18+
Reviewed-by: Dmitry Safonov <dima@arista.com>
Signed-off-by: Michael Bommarito <michael.bommarito@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Dmitry Safonov <0x7f454c46@gmail.com>
Link: https://patch.msgid.link/20260625-tcp-md5-connect-v3-1-1fd313d6c1e0@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit f34f171222 upstream.
sriov_restore_vf_rebar_state() uses the VF Resizable BAR Control register
to decide how many VF BARs to restore (nbars) and which VF BAR each
iteration addresses (bar_idx). bar_idx indexes into dev->sriov->barsz[],
which has only PCI_SRIOV_NUM_BARS (6) entries.
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. The barsz[] access then goes out of bounds. UBSAN
reports this as:
UBSAN: array-index-out-of-bounds in drivers/pci/iov.c:948:51 index 7 is out of range for type 'resource_size_t [6]'
Observed on an NVIDIA RTX PRO 1000 GPU (GB207GLM) that stopped responding
during a failed GC6 power state exit. The subsequent pci_restore_state()
invoked sriov_restore_vf_rebar_state() while config reads returned
0xffffffff, triggering the splat.
Bail out if any VF Resizable BAR Control read returns PCI_ERROR_RESPONSE.
No further VF BARs are touched, which is safe because a config read that
returns PCI_ERROR_RESPONSE indicates the device is unreachable and
restoration is pointless. This mirrors the guard in
pci_restore_rebar_state().
Fixes: 5a8f77e24a ("PCI/IOV: Restore VF resizable BAR state after reset")
Signed-off-by: Marco Nenciarini <mnencia@kcore.it>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/44a4ae53ec2825816b816c85cd378430d9a95cc6.1776429882.git.mnencia@kcore.it
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit e0779713a1 upstream.
A lockdep warning is observed during boot on a Qcom firmware-managed
platform:
INFO: trying to register non-static key.
The code is fine but needs lockdep annotation, or maybe
you didn't initialize this object before use?
turning off the locking correctness validator.
...
Call trace:
register_lock_class+0x128/0x4d8
__lock_acquire+0x110/0x1db0
lock_acquire+0x278/0x3d8
_raw_spin_lock_irq+0x6c/0xc0
dw_pcie_irq_domain_alloc+0x48/0x190
irq_domain_alloc_irqs_parent+0x2c/0x48
msi_domain_alloc+0x90/0x160
...
dw_pcie_irq_domain_alloc() takes pp->lock while allocating MSI
interrupts. pp->lock is normally initialized by dw_pcie_host_init(), but
Qcom firmware-managed hosts use the ECAM init path instead:
pci_host_common_ecam_create()
pci_ecam_create()
qcom_pcie_ecam_host_init()
dw_pcie_msi_host_init()
dw_pcie_allocate_domains()
That path constructs a fresh struct dw_pcie_rp and calls
dw_pcie_msi_host_init() directly, without going through
dw_pcie_host_init(). As a result, pp->lock was not initialized, which
triggers the warning.
Initialize pp->lock in qcom_pcie_ecam_host_init() before registering the
MSI domains so the firmware-managed ECAM path matches the normal DWC host
initialization sequence.
Fixes: 7d944c0f14 ("PCI: qcom: Add support for Qualcomm SA8255p based PCIe Root Complex")
Signed-off-by: Yadu M G <yadu.mg@oss.qualcomm.com>
[mani: added fixes tag and CCed stable]
Signed-off-by: Manivannan Sadhasivam <mani@kernel.org>
Cc: stable@kernel.org
Link: https://patch.msgid.link/20260604122418.727274-1-yadu.mg@oss.qualcomm.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit f865a57896 upstream.
When mtk_pcie_enable_port() fails, mtk_pcie_port_free() removes the port
from pcie->ports and frees the port structure. However, the IRQ domains set
up earlier by mtk_pcie_init_irq_domain() are never freed.
Fix this by refactoring mtk_pcie_irq_teardown() into a per-port helper,
mtk_pcie_irq_teardown_port(), and calling it from mtk_pcie_setup() when
mtk_pcie_enable_port() fails. Since the IRQ teardown must only happen in
the probe error path (during resume, child devices may have active MSI
mappings and the NOIRQ context prohibits sleeping locks),
mtk_pcie_enable_port() is changed to return an error code so callers can
distinguish the two paths and act accordingly.
This issue was reported by Sashiko while reviewing the EcoNet EN7528 SoC
support series.
Fixes: b099631df1 ("PCI: mediatek: Add controller support for MT2712 and MT7622")
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@oss.qualcomm.com>
Signed-off-by: Manivannan Sadhasivam <mani@kernel.org>
Cc: stable@vger.kernel.org # 5.10
Cc: Caleb James DeLisle <cjd@cjdns.fr>
Link: https://patch.msgid.link/20260521174617.17692-1-mani@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit fda8749ba7 upstream.
pci_host_common_init() is used by several generic ECAM host drivers.
After PCI core changes around pci_flags and preserve_config, these hosts
no longer opted into full bus number reassignment the way they did
before, which broke enumeration of devices on a Marvell CN106XX board.
When PCI_PROBE_ONLY is not set, add PCI_REASSIGN_ALL_BUS so
pci_scan_bridge_extend() takes the reassignment path: bus numbers can be
assigned from firmware EA data (e.g. pci_ea_fixed_busnrs()). Skip the
flag in probe-only mode so existing assignments are not overridden.
Fixes: 7246a4520b ("PCI: Use preserve_config in place of pci_flags")
Closes: https://lore.kernel.org/all/abkqm_LCd9zAM8cW@rkannoth-OptiPlex-7090/
Signed-off-by: Ratheesh Kannoth <rkannoth@marvell.com>
[mani: added stable tag]
Signed-off-by: Manivannan Sadhasivam <mani@kernel.org>
[bhelgaas: add problem report link]
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Cc: stable@vger.kernel.org
Cc: Vidya Sagar <vidyas@nvidia.com>
Link: https://patch.msgid.link/20260414081730.3864372-1-rkannoth@marvell.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 72780f7964 upstream.
Discard Vendor:Device ID matching in the PCIe failed link retraining quirk
and ignore the link status for the removal of the 2.5GT/s speed clamp,
whether applied by the quirk itself or the firmware earlier on. Revert to
the original target link speed if this final link retraining has failed.
This is so that link training noise in hot-plug scenarios does not make a
link remain clamped to the 2.5GT/s speed where an event race has led the
quirk to apply the speed clamp for one device, only to leave it in place
for a subsequent device to be plugged in.
Refer to the Link Capabilities register directly for the maximum link speed
determination so as to streamline backporting.
Fixes: a89c82249c ("PCI: Work around PCIe link training failures")
Signed-off-by: Maciej W. Rozycki <macro@orcam.me.uk>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Tested-by: Alok Tiwari <alok.a.tiwari@oracle.com>
Cc: stable@vger.kernel.org # v6.5+
Link: https://patch.msgid.link/alpine.DEB.2.21.2512080331530.49654@angie.orcam.me.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 7a94138cae upstream.
The chained IRQ handler is set during probe, but is only removed during the
driver remove(). If pci_host_probe() fails, the handler and INTx IRQ
domain remain set even though the devm-managed host bridge storage
containing struct altera_pcie will be released, leaving the handler with
a stale data pointer.
Interrupts are also enabled before pci_host_probe() is called. If probe
fails after that point, the controller interrupt source should be disabled
before the chained handler and INTx domain are removed.
So set the chained handler only after the INTx domain has been created.
Disable controller interrupts during IRQ teardown, and tear the IRQ setup
down if pci_host_probe() fails.
Fixes: c63aed7334 ("PCI: altera: Use pci_host_probe() to register host")
Signed-off-by: Mahesh Vaidya <mahesh.vaidya@altera.com>
[mani: commit log]
Signed-off-by: Manivannan Sadhasivam <mani@kernel.org>
Reviewed-by: Subhransu S. Prusty <subhransu.sekhar.prusty@altera.com>
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20260430204330.3121003-3-mahesh.vaidya@altera.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 5ef4bac021 upstream.
altera_pcie_irq_teardown() calls irq_dispose_mapping() on pcie->irq.
However, pcie->irq is the parent IRQ returned by platform_get_irq(), not
the mapping created by Altera INTx irq_domain.
The Altera driver only sets the chained handler on the parent IRQ. It
should detach that handler during teardown, but it should not dispose the
parent IRQ mapping, which belongs to the parent interrupt controller's
irq_domain.
Drop irq_dispose_mapping(pcie->irq) from the teardown path.
Note that during irqchip remove(), the child IRQs should've disposed. But
since the chained handler itself is removed, there is no way the stale
child IRQs (if exists) could fire. So it is safe here.
Fixes: ec15c4d0d5 ("PCI: altera: Allow building as module")
Signed-off-by: Mahesh Vaidya <mahesh.vaidya@altera.com>
[mani: added a note about IRQ disposal]
Signed-off-by: Manivannan Sadhasivam <mani@kernel.org>
Reviewed-by: Subhransu S. Prusty <subhransu.sekhar.prusty@altera.com>
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20260430204330.3121003-2-mahesh.vaidya@altera.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit e8da46d99d upstream.
rt1711h_probe() registers the TCPCI port before requesting the interrupt
and enabling alert interrupts. If either of those later steps fails, the
probe function returns without unregistering the TCPCI port. The explicit
unregister currently only happens from the remove callback.
Register a devres action immediately after tcpci_register_port() succeeds,
so tcpci_unregister_port() runs on later probe failures and on driver
detach. Drop the remove callback to avoid unregistering the same port
twice.
This issue was identified during our ongoing static-analysis research while
reviewing kernel code.
Fixes: 302c570bf3 ("usb: typec: tcpci_rt1711h: avoid screaming irq causing boot hangs")
Cc: stable <stable@kernel.org>
Co-developed-by: Ijae Kim <ae878000@gmail.com>
Signed-off-by: Ijae Kim <ae878000@gmail.com>
Signed-off-by: Myeonghun Pak <mhun512@gmail.com>
Link: https://patch.msgid.link/20260706145312.37260-1-mhun512@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 49f6e3c3ef upstream.
xhci_ring_to_sgtable() allocates a temporary pages array and
uses it to build the returned sg_table with
sg_alloc_table_from_pages().
The error paths free the pages array, but the success path
returns the sg_table without freeing it. This leaks the temporary
array every time a sideband client gets an endpoint or event ring
buffer.
Free the pages array after sg_alloc_table_from_pages() succeeds.
The returned sg_table has its own scatterlist entries and does not
depend on the temporary array after construction.
Fixes: de66754e9f ("xhci: sideband: add initial api to register a secondary interrupter entity")
Cc: stable <stable@kernel.org>
Signed-off-by: Xu Rao <raoxu@uniontech.com>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://patch.msgid.link/20260703144033.483286-2-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 42c37c4b75 upstream.
When a USB device with active stream endpoints is disconnected,
xhci_free_streams() is called from the hub_event workqueue to
free the stream resources. It calls xhci_free_stream_info()
while holding xhci->lock with irqs disabled.
xhci_free_stream_info() invokes xhci_free_stream_ctx(), which
calls dma_free_coherent() for large stream context arrays.
dma_free_coherent() can sleep (e.g. via vunmap), triggering
a BUG when called from atomic context.
Call trace:
dma_free_attrs+0x174/0x220
xhci_free_stream_info+0xd0/0x11c
xhci_free_streams+0x278/0x37c
usb_free_streams+0x98/0xc0
usb_unbind_interface+0x1b8/0x2f8
device_release_driver_internal+0x1d4/0x2cc
device_release_driver+0x18/0x28
bus_remove_device+0x160/0x1a4
device_del+0x1ec/0x350
usb_disable_device+0x98/0x214
usb_disconnect+0xf0/0x35c
hub_event+0xab4/0x19ec
process_one_work+0x278/0x63c
Fix this by saving the stream_info pointers and clearing the
ep references under the lock, then calling xhci_free_stream_info()
outside the lock where sleeping is allowed.
Fixes: 8df75f42f8 ("USB: xhci: Add memory allocation for USB3 bulk streams.")
Cc: stable <stable@kernel.org>
Signed-off-by: Lianqin Hu <hulianqin@vivo.com>
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://patch.msgid.link/20260703144033.483286-3-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit bc4a982889 upstream.
Generally userspace is supposed to explicitly clear freeze listeners
before they drop the refcount on the node ref to zero, but there's
nothing forcing that. Currently, in this scenario the freeze listener
remains in the freeze_listeners rbtree and in the remote node's freeze
listener list, even though the ref for which the listener is registered
is gone. This could potentially lead to a memory leak due to a refcount
cycle. Thus, remove the freeze listener in this scenario.
Cc: stable <stable@kernel.org>
Fixes: eafedbc7c0 ("rust_binder: add Rust Binder driver")
Signed-off-by: Alice Ryhl <aliceryhl@google.com>
Link: https://patch.msgid.link/20260703-remove-freeze-on-remove-node-v3-1-6e0c4547af46@google.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit eb1645bf10 upstream.
Rust Binder stats use BC_COUNT and BR_COUNT to size the command and
return counters, and use event string tables when printing debug
statistics.
The Binder protocol includes freeze-related commands and return codes,
but the Rust Binder statistics code was not updated to cover them. As a
result, those commands and return codes are not accounted for or printed
by the stats debug output.
Update the counts and event string tables so these commands and return
codes are included in the debug statistics output.
Fixes: eafedbc7c0 ("rust_binder: add Rust Binder driver")
Cc: stable <stable@kernel.org>
Acked-by: Carlos Llamas <cmllamas@google.com>
Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Signed-off-by: Keshav Verma <iganschel@gmail.com>
Link: https://patch.msgid.link/20260615211743.734-1-iganschel@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 6849cabfd3 upstream.
Rust binder resolved handle 0 to the context manager node, but it does not
reject the case where the caller owns the same node.
The C binder driver rejects transactions from the context-manager process
to handle 0 after resolving the target node. Match that behavior in Rust
Binder by rejecting handle 0 transactions when the resolved context-manager
node is owned by the calling process.
This applies to both synchronous and oneway transactions because both paths
resolve the target through Process::get_transaction_node().
Cc: stable <stable@kernel.org>
Fixes: eafedbc7c0 ("rust_binder: add Rust Binder driver")
Signed-off-by: Keshav Verma <iganschel@gmail.com>
Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Link: https://patch.msgid.link/20260625103957.730-1-iganschel@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 803c8a9502 upstream.
Allocation's Drop walks the offsets array (binder_size_t = u64 entries),
cleaning up the objects, but it used usize instead of u64 for both the
stride and the per-entry read.
On 64-bit kernels (usize == u64) this is harmless, but on 32-bit kernels
it walks the 8-byte entries in 4-byte steps, iterating an N-entry array
2N times, and reads the always-zero high word as offset 0, cleaning up
the object at offset 0 N extra times. As a result the referenced node or
handle ends up with a lower reference count than it actually has (a
refcount over-decrement), and binder's reference accounting is corrupted;
for example, the owner can be notified of a strong reference release
(BR_RELEASE) even though references still remain.
Change the stride to u64, and read each entry as a u64, narrowing it to
usize with try_into().
On 32-bit ARM, when this over-decrement would drive a count below zero,
the driver's existing refcount guard refuses it and fires:
rust_binder: Failure: refcount underflow!
Cc: stable <stable@kernel.org>
Fixes: eafedbc7c0 ("rust_binder: add Rust Binder driver")
Signed-off-by: Hyunwoo Kim <imv4bel@gmail.com>
Acked-by: Carlos Llamas <cmllamas@google.com>
Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Link: https://patch.msgid.link/ahw3tFhLz9bMMJAO@v4bel
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit f223d27a54 upstream.
In binder_free_transaction(), the t->to_proc is read under the t->lock.
However, once the t->lock is dropped, the to_proc can die in parallel.
This leads to a use-after-free error when we attempt to acquire its
inner lock right afterwards:
==================================================================
BUG: KASAN: slab-use-after-free in _raw_spin_lock+0xe4/0x1a0
Write of size 4 at addr ffff00001125da70 by task B/672
CPU: 20 UID: 0 PID: 672 Comm: B Not tainted 7.1.0-rc6-00284-g8e65320d91cd #4 PREEMPT
Hardware name: linux,dummy-virt (DT)
Call trace:
_raw_spin_lock+0xe4/0x1a0
binder_free_transaction+0x8c/0x320
binder_send_failed_reply+0x21c/0x2f8
binder_thread_release+0x488/0x7e0
binder_ioctl+0x12c0/0x29a0
[...]
Allocated by task 675:
__kmalloc_cache_noprof+0x174/0x444
binder_open+0x118/0xb70
do_dentry_open+0x374/0x1040
vfs_open+0x58/0x3bc
[...]
Freed by task 212:
__kasan_slab_free+0x58/0x80
kfree+0x1a0/0x4a4
binder_proc_dec_tmpref+0x32c/0x5e0
binder_deferred_func+0xc48/0x104c
process_one_work+0x53c/0xbc0
[...]
==================================================================
To prevent this, pin the target thread (t->to_thread) to guarantee the
target process remains alive. Undelivered transactions without a target
thread are already safe, as the target process can only be the current
context in those paths.
Cc: stable <stable@kernel.org>
Reported-by: Alice Ryhl <aliceryhl@google.com>
Closes: https://lore.kernel.org/all/aikJKVuny_eOivwN@google.com/
Fixes: a370003cc3 ("binder: fix possible UAF when freeing buffer")
Signed-off-by: Carlos Llamas <cmllamas@google.com>
Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Link: https://patch.msgid.link/20260619185233.2194678-2-cmllamas@google.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 114a116aaa upstream.
When a thread exits, binder_thread_release() walks its transaction stack
to clear the t->from and t->to_proc that correspond with the exiting
thread. However, a process dying in parallel might attempt to kfree some
of these transactions. And if one of them has no associated t->to_proc,
the t->to_proc->inner_lock will not be acquired.
This means that transaction accesses in binder_thread_release() after
t->to_proc has been cleared might race with binder_free_transaction()
and cause a use-after-free error as reported by KASAN:
==================================================================
BUG: KASAN: slab-use-after-free in binder_thread_release+0x5d0/0x798
Write of size 8 at addr ffff000016627500 by task X/715
CPU: 17 UID: 0 PID: 715 Comm: X Not tainted 7.1.0-rc5-00149-g8fde5d1d47f6 #30 PREEMPT
Hardware name: linux,dummy-virt (DT)
Call trace:
binder_thread_release+0x5d0/0x798
binder_ioctl+0x12c0/0x299c
[...]
Allocated by task 717 on cpu 18 at 67.267803s:
__kasan_kmalloc+0xa0/0xbc
__kmalloc_cache_noprof+0x174/0x444
binder_transaction+0x554/0x8150
binder_thread_write+0xa30/0x4354
binder_ioctl+0x20f0/0x299c
[...]
Freed by task 202 on cpu 18 at 90.416221s:
__kasan_slab_free+0x58/0x80
kfree+0x1a0/0x4a4
binder_free_transaction+0x150/0x294
binder_send_failed_reply+0x398/0x6d8
binder_release_work+0x3e4/0x4ec
binder_deferred_func+0xbd8/0x104c
[...]
==================================================================
In order to avoid this, make sure that binder_free_transaction() reads
the t->to_proc under the transaction lock. This will serialize the
transaction release with the accesses in binder_thread_release(). Plus,
it matches the documented locking rules for @to_proc.
Cc: stable <stable@kernel.org>
Fixes: 7a4408c6bd ("binder: make sure accesses to proc/thread are safe")
Reviewed-by: Alice Ryhl <aliceryhl@google.com>
Signed-off-by: Carlos Llamas <cmllamas@google.com>
Link: https://patch.msgid.link/20260619185233.2194678-1-cmllamas@google.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit dc7fe87de4 upstream.
VFIO device debugfs files created with debugfs_create_devm_seqfile()
store a devres allocated debugfs_devm_entry as inode private data.
vfio_unregister_group_dev() currently calls vfio_device_del() before
vfio_device_debugfs_exit(), but device_del() releases devres. This can
leave debugfs entries visible with stale inode private data while
unregister waits for userspace references to drain.
Remove the per-device debugfs tree before vfio_device_del(). The debugfs
view is diagnostic only, so losing it at the start of unregister is
preferable to preserving entries whose backing storage may already have
been released.
Complete the teardown by clearing the per-device debugfs root after
removal. This matches the global debugfs root cleanup and prevents
future users from mistaking a removed dentry for a live debugfs tree
during the remainder of unregister.
Fixes: 2202844e44 ("vfio/migration: Add debugfs to live migration driver")
Reported-by: Sashiko AI Review <sashiko-bot@kernel.org>
Link: https://lore.kernel.org/r/20260615192725.6A2221F000E9@smtp.kernel.org
Cc: stable@vger.kernel.org
Cc: Longfang Liu <liulongfang@huawei.com>
Assisted-by: OpenAI Codex:gpt-5
Signed-off-by: Alex Williamson <alex.williamson@nvidia.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Link: https://lore.kernel.org/r/20260615204717.735302-1-alex.williamson@nvidia.com
Signed-off-by: Alex Williamson <alex@shazbot.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit a26b499b75 upstream.
vfio_mig_get_next_state() walks vfio_from_fsm_table[] one step at a time,
looping to skip optional states the device does not support until
*next_fsm is supported. A blocked transition is encoded as
VFIO_DEVICE_STATE_ERROR, which the trailing return reports as -EINVAL.
The skip loop does not account for the ERROR sentinel.
state_flags_table[ERROR] is ~0U and vfio_from_fsm_table[ERROR][*] is
ERROR, so once *next_fsm becomes ERROR the loop condition stays true and
*next_fsm never changes. The blocked arcs STOP_COPY -> PRE_COPY and
STOP_COPY -> PRE_COPY_P2P map to ERROR yet pass the support check on a
precopy-capable device, causing the loop to spin forever while holding
the driver state mutex. This can result in a soft lockup, and a panic
with softlockup_panic set.
Terminate the skip loop on the ERROR sentinel so a blocked transition
falls through to the existing return and reports -EINVAL.
Fixes: 4db52602a6 ("vfio: Extend the device migration protocol with PRE_COPY")
Reported-by: Yuhao Jiang <danisjiang@gmail.com>
Cc: stable@vger.kernel.org
Signed-off-by: Junrui Luo <moonafterrain@outlook.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com>
Link: https://lore.kernel.org/r/SYBPR01MB7881290BBDE79B61AE6A017FAF122@SYBPR01MB7881.ausprd01.prod.outlook.com
Signed-off-by: Alex Williamson <alex@shazbot.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit e73638e55f upstream.
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 that the virq_disabled and bardirty flags are relocated to fill
an existing hole in the structure.
Bitfield justifications:
has_dyn_msix: written only in vfio_pci_core_enable()
pci_2_3: written only in vfio_pci_core_enable()
reset_works: written only in vfio_pci_core_enable()
extended_caps: written only in vfio_cap_len() under vfio_config_init()
has_vga: written only in vfio_pci_core_enable()
nointx: written only in vfio_pci_core_enable()
needs_pm_restore: written only in vfio_pci_probe_power_state()
disable_idle_d3: written only at .init in vfio_pci_core_init_dev()
Dedicated storage units:
virq_disabled: written by guest INTx command writes in
vfio_basic_config_write() while the device is open
bardirty: written by guest BAR writes in vfio_basic_config_write()
while the device is open
pm_intx_masked: written in the runtime-PM suspend path.
pm_runtime_engaged: written by low-power feature entry/exit paths
needs_reset: set in vfio_pci_core_disable() and cleared for devices in
the set by vfio_pci_dev_set_try_reset()
sriov_active: written by vfio_pci_core_sriov_configure() via sysfs
sriov_numvfs while bound.
Fixes: 9cd0f6d5cb ("vfio/pci: Use bitfield for struct vfio_pci_core_device flags")
Cc: stable@vger.kernel.org
Assisted-by: Claude:claude-opus-4-8
Signed-off-by: Alex Williamson <alex.williamson@nvidia.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Link: https://lore.kernel.org/r/20260615191241.688297-4-alex.williamson@nvidia.com
Signed-off-by: Alex Williamson <alex@shazbot.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Alexandru Hossu
Arnd Bergmann
Ricardo Ribalda
Hao-Qun Huang
Shoichiro Miyamoto
Michael Tautschnig
Samuel Page
Yizhou Zhao
Cen Zhang
Runyu Xiao
Alan Stern
Michael Bommarito
Johan Hovold
Ammar Faizi
Xiang Mei
SeongJae Park
Marco Nenciarini
Yadu M G
Manivannan Sadhasivam
Richard Zhu
Ratheesh Kannoth
Maciej W. Rozycki
Mahesh Vaidya
Ziyao Li
Myeonghun Pak
Xu Rao
胡连勤
Alice Ryhl
Keshav Verma
Hyunwoo Kim
Carlos Llamas
Zenm Chen
Alex Williamson
Junrui Luo