[syzbot] [rdma?] KASAN: slab-use-after-free Read in ucma_create_uevent

[syzbot]

Hello,

syzbot found the following issue on:

HEAD commit: d006330be3f7 Merge tag 'sound-6.16-rc6' of git://git.kerne..
git tree: upstream
console output: https://syzkaller.appspot.com/x/log.txt?x=13e4bf70580000
kernel config: https://syzkaller.appspot.com/x/.config?x=8fa6c6703a4b2315
dashboard link: https://syzkaller.appspot.com/bug?extid=a6ffe86390c8a6afc818
compiler: gcc (Debian 12.2.0-14) 12.2.0, GNU ld (GNU Binutils for Debian) 2.40
syz repro: https://syzkaller.appspot.com/x/repro.syz?x=17e4bf70580000
C reproducer: https://syzkaller.appspot.com/x/repro.c?x=14f6ca8c580000

Downloadable assets:
disk image (non-bootable): https://storage.googleapis.com/syzbot-assets/d900f083ada3/non_bootable_disk-d006330b.raw.xz
vmlinux: https://storage.googleapis.com/syzbot-assets/d47418563f9c/vmlinux-d006330b.xz
kernel image: https://storage.googleapis.com/syzbot-assets/677ac864031c/bzImage-d006330b.xz

IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: syzbot+a6ffe8...@syzkaller.appspotmail.com

==================================================================
BUG: KASAN: slab-use-after-free in ucma_create_uevent+0xadb/0xb30 drivers/infiniband/core/ucma.c:275
Read of size 8 at addr ffff888029155410 by task kworker/u32:9/1147

CPU: 3 UID: 0 PID: 1147 Comm: kworker/u32:9 Not tainted 6.16.0-rc5-syzkaller-00025-gd006330be3f7 #0 PREEMPT(full)
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.3-debian-1.16.3-2~bpo12+1 04/01/2014
Workqueue: rdma_cm cma_iboe_join_work_handler
Call Trace:
<TASK>
__dump_stack lib/dump_stack.c:94 [inline]
dump_stack_lvl+0x116/0x1f0 lib/dump_stack.c:120
print_address_description mm/kasan/report.c:408 [inline]
print_report+0xcd/0x680 mm/kasan/report.c:521
kasan_report+0xe0/0x110 mm/kasan/report.c:634
ucma_create_uevent+0xadb/0xb30 drivers/infiniband/core/ucma.c:275
ucma_event_handler+0x102/0x940 drivers/infiniband/core/ucma.c:351
cma_cm_event_handler+0x97/0x300 drivers/infiniband/core/cma.c:2173
cma_iboe_join_work_handler+0xca/0x170 drivers/infiniband/core/cma.c:3008
process_one_work+0x9cf/0x1b70 kernel/workqueue.c:3238
process_scheduled_works kernel/workqueue.c:3321 [inline]
worker_thread+0x6c8/0xf10 kernel/workqueue.c:3402
kthread+0x3c5/0x780 kernel/kthread.c:464
ret_from_fork+0x5d7/0x6f0 arch/x86/kernel/process.c:148
ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:245
</TASK>

Allocated by task 6107:
kasan_save_stack+0x33/0x60 mm/kasan/common.c:47
kasan_save_track+0x14/0x30 mm/kasan/common.c:68
poison_kmalloc_redzone mm/kasan/common.c:377 [inline]
__kasan_kmalloc+0xaa/0xb0 mm/kasan/common.c:394
kmalloc_noprof include/linux/slab.h:905 [inline]
kzalloc_noprof include/linux/slab.h:1039 [inline]
ucma_process_join+0x237/0xa30 drivers/infiniband/core/ucma.c:1465
ucma_join_multicast+0xe8/0x160 drivers/infiniband/core/ucma.c:1557
ucma_write+0x1fb/0x330 drivers/infiniband/core/ucma.c:1738
vfs_write+0x2a0/0x1150 fs/read_write.c:684
ksys_write+0x1f8/0x250 fs/read_write.c:738
do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
do_syscall_64+0xcd/0x4c0 arch/x86/entry/syscall_64.c:94
entry_SYSCALL_64_after_hwframe+0x77/0x7f

Freed by task 6107:
kasan_save_stack+0x33/0x60 mm/kasan/common.c:47
kasan_save_track+0x14/0x30 mm/kasan/common.c:68
kasan_save_free_info+0x3b/0x60 mm/kasan/generic.c:576
poison_slab_object mm/kasan/common.c:247 [inline]
__kasan_slab_free+0x51/0x70 mm/kasan/common.c:264
kasan_slab_free include/linux/kasan.h:233 [inline]
slab_free_hook mm/slub.c:2381 [inline]
slab_free mm/slub.c:4643 [inline]
kfree+0x2b4/0x4d0 mm/slub.c:4842
ucma_process_join+0x3b9/0xa30 drivers/infiniband/core/ucma.c:1516
ucma_join_multicast+0xe8/0x160 drivers/infiniband/core/ucma.c:1557
ucma_write+0x1fb/0x330 drivers/infiniband/core/ucma.c:1738
vfs_write+0x2a0/0x1150 fs/read_write.c:684
ksys_write+0x1f8/0x250 fs/read_write.c:738
do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
do_syscall_64+0xcd/0x4c0 arch/x86/entry/syscall_64.c:94
entry_SYSCALL_64_after_hwframe+0x77/0x7f

The buggy address belongs to the object at ffff888029155400
which belongs to the cache kmalloc-192 of size 192
The buggy address is located 16 bytes inside of
freed 192-byte region [ffff888029155400, ffff8880291554c0)

The buggy address belongs to the physical page:
page: refcount:0 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x29155
flags: 0xfff00000000000(node=0|zone=1|lastcpupid=0x7ff)
page_type: f5(slab)
raw: 00fff00000000000 ffff88801b8423c0 ffffea0000a60c80 dead000000000002
raw: 0000000000000000 0000000080100010 00000000f5000000 0000000000000000
page dumped because: kasan: bad access detected
page_owner tracks the page as allocated
page last allocated via order 0, migratetype Unmovable, gfp_mask 0x52cc0(GFP_KERNEL|__GFP_NOWARN|__GFP_NORETRY|__GFP_COMP), pid 1, tgid 1 (swapper/0), ts 12897812541, free_ts 0
set_page_owner include/linux/page_owner.h:32 [inline]
post_alloc_hook+0x1c0/0x230 mm/page_alloc.c:1704
prep_new_page mm/page_alloc.c:1712 [inline]
get_page_from_freelist+0x1321/0x3890 mm/page_alloc.c:3669
__alloc_frozen_pages_noprof+0x261/0x23f0 mm/page_alloc.c:4959
alloc_pages_mpol+0x1fb/0x550 mm/mempolicy.c:2419
alloc_slab_page mm/slub.c:2451 [inline]
allocate_slab mm/slub.c:2619 [inline]
new_slab+0x23b/0x330 mm/slub.c:2673
___slab_alloc+0xd9c/0x1940 mm/slub.c:3859
__slab_alloc.constprop.0+0x56/0xb0 mm/slub.c:3949
__slab_alloc_node mm/slub.c:4024 [inline]
slab_alloc_node mm/slub.c:4185 [inline]
__kmalloc_cache_noprof+0xfb/0x3e0 mm/slub.c:4354
kmalloc_noprof include/linux/slab.h:905 [inline]
kzalloc_noprof include/linux/slab.h:1039 [inline]
call_usermodehelper_setup+0xaf/0x360 kernel/umh.c:362
kobject_uevent_env+0x1690/0x1870 lib/kobject_uevent.c:628
driver_bound+0x164/0x230 drivers/base/dd.c:421
really_probe+0x62b/0xa90 drivers/base/dd.c:707
__driver_probe_device+0x1de/0x440 drivers/base/dd.c:799
driver_probe_device+0x4c/0x1b0 drivers/base/dd.c:829
__device_attach_driver+0x1df/0x310 drivers/base/dd.c:957
bus_for_each_drv+0x156/0x1e0 drivers/base/bus.c:462
page_owner free stack trace missing

Memory state around the buggy address:
ffff888029155300: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
ffff888029155380: fb fb fb fb fb fb fb fb fc fc fc fc fc fc fc fc
>ffff888029155400: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
^
ffff888029155480: fb fb fb fb fb fb fb fb fc fc fc fc fc fc fc fc
ffff888029155500: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
==================================================================


---
This report is generated by a bot. It may contain errors.
See https://goo.gl/tpsmEJ for more information about syzbot.
syzbot engineers can be reached at syzk...@googlegroups.com.

syzbot will keep track of this issue. See:
https://goo.gl/tpsmEJ#status for how to communicate with syzbot.

If the report is already addressed, let syzbot know by replying with:
#syz fix: exact-commit-title

If you want syzbot to run the reproducer, reply with:
#syz test: git://repo/address.git branch-or-commit-hash
If you attach or paste a git patch, syzbot will apply it before testing.

If you want to overwrite report's subsystems, reply with:
#syz set subsystems: new-subsystem
(See the list of subsystem names on the web dashboard)

If the report is a duplicate of another one, reply with:
#syz dup: exact-subject-of-another-report

If you want to undo deduplication, reply with:
#syz undup

[Hillf Danton]

> Date: Tue, 08 Jul 2025 17:12:26 -0700 [thread overview]

> Hello,
>
> syzbot found the following issue on:
>
> HEAD commit: d006330be3f7 Merge tag 'sound-6.16-rc6' of git://git.kerne..
> git tree: upstream
> console output: https://syzkaller.appspot.com/x/log.txt?x=13e4bf70580000
> kernel config: https://syzkaller.appspot.com/x/.config?x=8fa6c6703a4b2315
> dashboard link: https://syzkaller.appspot.com/bug?extid=a6ffe86390c8a6afc818
> compiler: gcc (Debian 12.2.0-14) 12.2.0, GNU ld (GNU Binutils for Debian) 2.40
> syz repro: https://syzkaller.appspot.com/x/repro.syz?x=17e4bf70580000
> C reproducer: https://syzkaller.appspot.com/x/repro.c?x=14f6ca8c580000

#syz test

--- x/drivers/infiniband/core/ucma.c
+++ y/drivers/infiniband/core/ucma.c
@@ -272,8 +272,24 @@ static struct ucma_event *ucma_create_ue
case RDMA_CM_EVENT_MULTICAST_ERROR:
uevent->mc = (struct ucma_multicast *)
event->param.ud.private_data;
- uevent->resp.uid = uevent->mc->uid;
- uevent->resp.id = uevent->mc->id;
+ do {
+ struct ucma_multicast *mc;
+ int err = 1;
+
+ xa_lock(&multicast_table);
+ list_for_each_entry(mc, &ctx->mc_list, list)
+ if (mc == uevent->mc) {
+ uevent->resp.uid = uevent->mc->uid;
+ uevent->resp.id = uevent->mc->id;
+ err = 0;
+ break;
+ }
+ xa_unlock(&multicast_table);
+ if (err) {
+ kfree(uevent);
+ return NULL;
+ }
+ } while (0);
break;
default:
uevent->resp.uid = ctx->uid;
--

[Edward Adam Davis]

#syz test

diff --git a/drivers/infiniband/core/ucma.c b/drivers/infiniband/core/ucma.c
index 6e700b974033..762c96d5d9e2 100644
--- a/drivers/infiniband/core/ucma.c
+++ b/drivers/infiniband/core/ucma.c
@@ -1506,8 +1506,8 @@ static ssize_t ucma_process_join(struct ucma_file *file,
mutex_lock(&ctx->mutex);
rdma_leave_multicast(ctx->cm_id, (struct sockaddr *) &mc->addr);
mutex_unlock(&ctx->mutex);
- ucma_cleanup_mc_events(mc);
err_xa_erase:
+ ucma_cleanup_mc_events(mc);
xa_lock(&multicast_table);
list_del(&mc->list);
__xa_erase(&multicast_table, mc->id);

[syzbot]

Hello,

syzbot has tested the proposed patch and the reproducer did not trigger any issue:

Reported-by: syzbot+a6ffe8...@syzkaller.appspotmail.com
Tested-by: syzbot+a6ffe8...@syzkaller.appspotmail.com

Tested on:

commit: 73392339 Merge tag 'pwm/for-6.16-rc6-fixes' of git://g..
git tree: upstream
console output: https://syzkaller.appspot.com/x/log.txt?x=13687f70580000

kernel config: https://syzkaller.appspot.com/x/.config?x=8fa6c6703a4b2315
dashboard link: https://syzkaller.appspot.com/bug?extid=a6ffe86390c8a6afc818
compiler: gcc (Debian 12.2.0-14) 12.2.0, GNU ld (GNU Binutils for Debian) 2.40

patch: https://syzkaller.appspot.com/x/patch.diff?x=13ea3582580000

Note: testing is done by a robot and is best-effort only.

[syzbot]

Hello,

syzbot has tested the proposed patch but the reproducer is still triggering an issue:
KASAN: slab-use-after-free Read in ucma_create_uevent

==================================================================
BUG: KASAN: slab-use-after-free in ucma_create_uevent+0xadb/0xb30 drivers/infiniband/core/ucma.c:275

Read of size 8 at addr ffff888022289610 by task kworker/u32:2/46

CPU: 1 UID: 0 PID: 46 Comm: kworker/u32:2 Not tainted 6.16.0-rc5-syzkaller-g733923397fd9-dirty #0 PREEMPT(full)

Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.3-debian-1.16.3-2~bpo12+1 04/01/2014
Workqueue: rdma_cm cma_iboe_join_work_handler
Call Trace:
<TASK>
__dump_stack lib/dump_stack.c:94 [inline]
dump_stack_lvl+0x116/0x1f0 lib/dump_stack.c:120
print_address_description mm/kasan/report.c:408 [inline]
print_report+0xcd/0x680 mm/kasan/report.c:521
kasan_report+0xe0/0x110 mm/kasan/report.c:634
ucma_create_uevent+0xadb/0xb30 drivers/infiniband/core/ucma.c:275
ucma_event_handler+0x102/0x940 drivers/infiniband/core/ucma.c:351
cma_cm_event_handler+0x97/0x300 drivers/infiniband/core/cma.c:2173
cma_iboe_join_work_handler+0xca/0x170 drivers/infiniband/core/cma.c:3008
process_one_work+0x9cf/0x1b70 kernel/workqueue.c:3238
process_scheduled_works kernel/workqueue.c:3321 [inline]
worker_thread+0x6c8/0xf10 kernel/workqueue.c:3402
kthread+0x3c5/0x780 kernel/kthread.c:464
ret_from_fork+0x5d7/0x6f0 arch/x86/kernel/process.c:148
ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:245
</TASK>

Allocated by task 6542:

kasan_save_stack+0x33/0x60 mm/kasan/common.c:47
kasan_save_track+0x14/0x30 mm/kasan/common.c:68
poison_kmalloc_redzone mm/kasan/common.c:377 [inline]
__kasan_kmalloc+0xaa/0xb0 mm/kasan/common.c:394
kmalloc_noprof include/linux/slab.h:905 [inline]
kzalloc_noprof include/linux/slab.h:1039 [inline]
ucma_process_join+0x237/0xa30 drivers/infiniband/core/ucma.c:1465
ucma_join_multicast+0xe8/0x160 drivers/infiniband/core/ucma.c:1557
ucma_write+0x1fb/0x330 drivers/infiniband/core/ucma.c:1738

vfs_write+0x29d/0x1150 fs/read_write.c:684

ksys_write+0x1f8/0x250 fs/read_write.c:738
do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
do_syscall_64+0xcd/0x4c0 arch/x86/entry/syscall_64.c:94
entry_SYSCALL_64_after_hwframe+0x77/0x7f

Freed by task 6542:

kasan_save_stack+0x33/0x60 mm/kasan/common.c:47
kasan_save_track+0x14/0x30 mm/kasan/common.c:68
kasan_save_free_info+0x3b/0x60 mm/kasan/generic.c:576
poison_slab_object mm/kasan/common.c:247 [inline]
__kasan_slab_free+0x51/0x70 mm/kasan/common.c:264
kasan_slab_free include/linux/kasan.h:233 [inline]
slab_free_hook mm/slub.c:2381 [inline]
slab_free mm/slub.c:4643 [inline]
kfree+0x2b4/0x4d0 mm/slub.c:4842
ucma_process_join+0x3b9/0xa30 drivers/infiniband/core/ucma.c:1516
ucma_join_multicast+0xe8/0x160 drivers/infiniband/core/ucma.c:1557
ucma_write+0x1fb/0x330 drivers/infiniband/core/ucma.c:1738

vfs_write+0x29d/0x1150 fs/read_write.c:684

ksys_write+0x1f8/0x250 fs/read_write.c:738
do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
do_syscall_64+0xcd/0x4c0 arch/x86/entry/syscall_64.c:94
entry_SYSCALL_64_after_hwframe+0x77/0x7f

The buggy address belongs to the object at ffff888022289600

which belongs to the cache kmalloc-192 of size 192
The buggy address is located 16 bytes inside of

freed 192-byte region [ffff888022289600, ffff8880222896c0)

The buggy address belongs to the physical page:

page: refcount:0 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x22289
ksm flags: 0xfff00000000000(node=0|zone=1|lastcpupid=0x7ff)
page_type: f5(slab)
raw: 00fff00000000000 ffff88801b8423c0 ffffea0000aeb500 dead000000000003

raw: 0000000000000000 0000000080100010 00000000f5000000 0000000000000000
page dumped because: kasan: bad access detected
page_owner tracks the page as allocated

page last allocated via order 0, migratetype Unmovable, gfp_mask 0x52cc0(GFP_KERNEL|__GFP_NOWARN|__GFP_NORETRY|__GFP_COMP), pid 1, tgid 1 (swapper/0), ts 3951737580, free_ts 0

set_page_owner include/linux/page_owner.h:32 [inline]
post_alloc_hook+0x1c0/0x230 mm/page_alloc.c:1704
prep_new_page mm/page_alloc.c:1712 [inline]
get_page_from_freelist+0x1321/0x3890 mm/page_alloc.c:3669
__alloc_frozen_pages_noprof+0x261/0x23f0 mm/page_alloc.c:4959
alloc_pages_mpol+0x1fb/0x550 mm/mempolicy.c:2419
alloc_slab_page mm/slub.c:2451 [inline]
allocate_slab mm/slub.c:2619 [inline]
new_slab+0x23b/0x330 mm/slub.c:2673
___slab_alloc+0xd9c/0x1940 mm/slub.c:3859
__slab_alloc.constprop.0+0x56/0xb0 mm/slub.c:3949
__slab_alloc_node mm/slub.c:4024 [inline]
slab_alloc_node mm/slub.c:4185 [inline]
__kmalloc_cache_noprof+0xfb/0x3e0 mm/slub.c:4354
kmalloc_noprof include/linux/slab.h:905 [inline]
kzalloc_noprof include/linux/slab.h:1039 [inline]

__tty_alloc_driver drivers/tty/tty_io.c:3336 [inline]
__tty_alloc_driver+0xbb/0x4a0 drivers/tty/tty_io.c:3326
vty_init+0xdd/0x420 drivers/tty/vt/vt.c:3822
tty_init+0x179/0x1c0 drivers/tty/tty_io.c:3649
chr_dev_init+0xfb/0x170 drivers/char/mem.c:759
do_one_initcall+0x120/0x6e0 init/main.c:1274
do_initcall_level init/main.c:1336 [inline]
do_initcalls init/main.c:1352 [inline]
do_basic_setup init/main.c:1371 [inline]
kernel_init_freeable+0x5c2/0x900 init/main.c:1584
kernel_init+0x1c/0x2b0 init/main.c:1474
ret_from_fork+0x5d7/0x6f0 arch/x86/kernel/process.c:148

page_owner free stack trace missing

Memory state around the buggy address:

ffff888022289500: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
ffff888022289580: fb fb fb fb fb fb fb fb fc fc fc fc fc fc fc fc
>ffff888022289600: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
^
ffff888022289680: fb fb fb fb fb fb fb fb fc fc fc fc fc fc fc fc
ffff888022289700: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
==================================================================

Tested on:

commit: 73392339 Merge tag 'pwm/for-6.16-rc6-fixes' of git://g..
git tree: upstream

console output: https://syzkaller.appspot.com/x/log.txt?x=17363582580000

kernel config: https://syzkaller.appspot.com/x/.config?x=8fa6c6703a4b2315
dashboard link: https://syzkaller.appspot.com/bug?extid=a6ffe86390c8a6afc818
compiler: gcc (Debian 12.2.0-14) 12.2.0, GNU ld (GNU Binutils for Debian) 2.40

patch: https://syzkaller.appspot.com/x/patch.diff?x=10bfca8c580000

[Lizhi Xu]

#syz test

diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c
index 9b471548e7ae..9b3051bf5225 100644
--- a/drivers/infiniband/core/cma.c
+++ b/drivers/infiniband/core/cma.c
@@ -2030,6 +2030,7 @@ static void destroy_mc(struct rdma_id_private *id_priv,
}
dev_put(ndev);

+ printk("cma, mc: %p, %s\n", mc, __func__);
cancel_work_sync(&mc->iboe_join.work);
}
kfree(mc);
@@ -5106,6 +5107,7 @@ void rdma_leave_multicast(struct rdma_cm_id *id, struct sockaddr *addr)
spin_unlock_irq(&id_priv->lock);

WARN_ON(id_priv->cma_dev->device != id->device);
+ printk("cma, mc: %p, id: %p, %s\n", mc, id, __func__);
destroy_mc(id_priv, mc);
return;
}
diff --git a/drivers/infiniband/core/ucma.c b/drivers/infiniband/core/ucma.c
index 6e700b974033..df3cc11b1ef9 100644
--- a/drivers/infiniband/core/ucma.c
+++ b/drivers/infiniband/core/ucma.c
@@ -272,6 +272,7 @@ static struct ucma_event *ucma_create_uevent(struct ucma_context *ctx,

case RDMA_CM_EVENT_MULTICAST_ERROR:
uevent->mc = (struct ucma_multicast *)
event->param.ud.private_data;

+ printk("mc: %p, evt: %d, %s\n", uevent->mc, event->event, __func__);

uevent->resp.uid = uevent->mc->uid;

uevent->resp.id = uevent->mc->id;

break;
@@ -1486,6 +1487,7 @@ static ssize_t ucma_process_join(struct ucma_file *file,
mutex_lock(&ctx->mutex);
ret = rdma_join_multicast(ctx->cm_id, (struct sockaddr *)&mc->addr,
join_state, mc);
+ printk("ret: %d, id: %p, mc: %p, %s\n", ret, ctx->cm_id, mc, __func__);
mutex_unlock(&ctx->mutex);
if (ret)
goto err_xa_erase;
@@ -1504,6 +1506,7 @@ static ssize_t ucma_process_join(struct ucma_file *file,

err_leave_multicast:
mutex_lock(&ctx->mutex);
+ printk("id: %p, %s\n", ctx->cm_id, __func__);

rdma_leave_multicast(ctx->cm_id, (struct sockaddr *) &mc->addr);
mutex_unlock(&ctx->mutex);

ucma_cleanup_mc_events(mc);

[syzbot]

Hello,

syzbot has tested the proposed patch and the reproducer did not trigger any issue:

Reported-by: syzbot+a6ffe8...@syzkaller.appspotmail.com
Tested-by: syzbot+a6ffe8...@syzkaller.appspotmail.com

Tested on:

commit: 8c2e52eb eventpoll: don't decrement ep refcount while ..
git tree: upstream
console output: https://syzkaller.appspot.com/x/log.txt?x=17ae7582580000

kernel config: https://syzkaller.appspot.com/x/.config?x=8fa6c6703a4b2315
dashboard link: https://syzkaller.appspot.com/bug?extid=a6ffe86390c8a6afc818
compiler: gcc (Debian 12.2.0-14) 12.2.0, GNU ld (GNU Binutils for Debian) 2.40

patch: https://syzkaller.appspot.com/x/patch.diff?x=14f67582580000

[Lizhi Xu]

#syz test

diff --git a/drivers/infiniband/core/ucma.c b/drivers/infiniband/core/ucma.c
index 6e700b974033..89c444c6f317 100644
--- a/drivers/infiniband/core/ucma.c
+++ b/drivers/infiniband/core/ucma.c
@@ -109,6 +109,7 @@ struct ucma_multicast {
u8 join_state;
struct list_head list;
struct sockaddr_storage addr;
+ atomic_t ref;
};

struct ucma_event {
@@ -257,6 +258,12 @@ static void ucma_copy_ud_event(struct ib_device *device,
dst->qkey = src->qkey;
}

+static void ucma_put_mc(struct ucma_multicast *mc)
+{
+ if (mc && atomic_dec_and_test(&mc->ref))
+ kfree(mc);
+}
+

static struct ucma_event *ucma_create_uevent(struct ucma_context *ctx,

struct rdma_cm_event *event)
{
@@ -274,6 +281,7 @@ static struct ucma_event *ucma_create_uevent(struct ucma_context *ctx,
event->param.ud.private_data;

uevent->resp.uid = uevent->mc->uid;
uevent->resp.id = uevent->mc->id;

+ ucma_put_mc(uevent->mc);

break;
default:
uevent->resp.uid = ctx->uid;

@@ -1471,6 +1479,7 @@ static ssize_t ucma_process_join(struct ucma_file *file,
mc->ctx = ctx;
mc->join_state = join_state;
mc->uid = cmd->uid;
+ atomic_set(&mc->ref, 1);
memcpy(&mc->addr, addr, cmd->addr_size);

xa_lock(&multicast_table);
@@ -1489,6 +1498,7 @@ static ssize_t ucma_process_join(struct ucma_file *file,

mutex_unlock(&ctx->mutex);
if (ret)
goto err_xa_erase;

+ atomic_inc(&mc->ref);

resp.id = mc->id;
if (copy_to_user(u64_to_user_ptr(cmd->response),
@@ -1513,7 +1523,7 @@ static ssize_t ucma_process_join(struct ucma_file *file,
__xa_erase(&multicast_table, mc->id);
err_free_mc:
xa_unlock(&multicast_table);
- kfree(mc);
+ ucma_put_mc(mc);
err_put_ctx:
ucma_put_ctx(ctx);
return ret;

[syzbot]

Hello,

syzbot has tested the proposed patch but the reproducer is still triggering an issue:

general protection fault in ucma_cleanup_mc_events

syz1: rxe_newlink: already configured on syz_tun
Oops: general protection fault, probably for non-canonical address 0xdffffc0000000083: 0000 [#1] SMP KASAN NOPTI
KASAN: null-ptr-deref in range [0x0000000000000418-0x000000000000041f]
CPU: 1 UID: 0 PID: 6579 Comm: syz.0.24 Not tainted 6.16.0-rc7-syzkaller-g2942242dde89-dirty #0 PREEMPT(full)

Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.3-debian-1.16.3-2~bpo12+1 04/01/2014

RIP: 0010:__mutex_lock_common kernel/locking/mutex.c:581 [inline]
RIP: 0010:__mutex_lock+0x14f/0xb90 kernel/locking/mutex.c:747
Code: d0 7c 08 84 d2 0f 85 ab 07 00 00 8b 35 1a 59 56 0f 85 f6 75 29 48 b8 00 00 00 00 00 fc ff df 48 8d 7b 60 48 89 fa 48 c1 ea 03 <80> 3c 02 00 0f 85 93 07 00 00 48 3b 5b 60 0f 85 e0 01 00 00 bf 01
RSP: 0018:ffffc900016df8b0 EFLAGS: 00010202

RAX: dffffc0000000000 RBX: 00000000000003b8 RCX: 1ffffffff35c0a00
RDX: 0000000000000083 RSI: 0000000000000000 RDI: 0000000000000418
RBP: ffffc900016df9f0 R08: ffffffff889fb8fd R09: fffff520002dbf35
R10: ffffc900016dfa10 R11: 0000000000000001 R12: dffffc0000000000
R13: ffffc900016df930 R14: 0000000000000000 R15: 1ffff920002dbf20
FS: 00007f6bbe81d6c0(0000) GS:ffff8880d6812000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000000000000 CR3: 000000002a054000 CR4: 0000000000352ef0
Call Trace:
<TASK>
ucma_cleanup_mc_events+0x5d/0x490 drivers/infiniband/core/ucma.c:518
ucma_process_join+0x966/0xad0 drivers/infiniband/core/ucma.c:1519
ucma_join_multicast+0xe8/0x160 drivers/infiniband/core/ucma.c:1567
ucma_write+0x1fb/0x330 drivers/infiniband/core/ucma.c:1748

vfs_write+0x29d/0x1150 fs/read_write.c:684
ksys_write+0x1f8/0x250 fs/read_write.c:738
do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
do_syscall_64+0xcd/0x4c0 arch/x86/entry/syscall_64.c:94
entry_SYSCALL_64_after_hwframe+0x77/0x7f

RIP: 0033:0x7f6bbd98e929
Code: ff ff c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 40 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 a8 ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007f6bbe81d038 EFLAGS: 00000246
ORIG_RAX: 0000000000000001
RAX: ffffffffffffffda RBX: 00007f6bbdbb5fa0 RCX: 00007f6bbd98e929
RDX: 00000000000000a0 RSI: 0000200000000900 RDI: 0000000000000004
RBP: 00007f6bbda10b39 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000
R13: 0000000000000000 R14: 00007f6bbdbb5fa0 R15: 00007fff295a4748
</TASK>
Modules linked in:

---[ end trace 0000000000000000 ]---
RIP: 0010:__mutex_lock_common kernel/locking/mutex.c:581 [inline]
RIP: 0010:__mutex_lock+0x14f/0xb90 kernel/locking/mutex.c:747
Code: d0 7c 08 84 d2 0f 85 ab 07 00 00 8b 35 1a 59 56 0f 85 f6 75 29 48 b8 00 00 00 00 00 fc ff df 48 8d 7b 60 48 89 fa 48 c1 ea 03 <80> 3c 02 00 0f 85 93 07 00 00 48 3b 5b 60 0f 85 e0 01 00 00 bf 01
RSP: 0018:ffffc900016df8b0 EFLAGS: 00010202

RAX: dffffc0000000000 RBX: 00000000000003b8 RCX: 1ffffffff35c0a00
----------------
Code disassembly (best guess):
0: d0 7c 08 84 sarb -0x7c(%rax,%rcx,1)
4: d2 0f rorb %cl,(%rdi)
6: 85 ab 07 00 00 8b test %ebp,-0x74fffff9(%rbx)
c: 35 1a 59 56 0f xor $0xf56591a,%eax
11: 85 f6 test %esi,%esi
13: 75 29 jne 0x3e
15: 48 b8 00 00 00 00 00 movabs $0xdffffc0000000000,%rax
1c: fc ff df
1f: 48 8d 7b 60 lea 0x60(%rbx),%rdi
23: 48 89 fa mov %rdi,%rdx
26: 48 c1 ea 03 shr $0x3,%rdx
* 2a: 80 3c 02 00 cmpb $0x0,(%rdx,%rax,1) <-- trapping instruction
2e: 0f 85 93 07 00 00 jne 0x7c7
34: 48 3b 5b 60 cmp 0x60(%rbx),%rbx
38: 0f 85 e0 01 00 00 jne 0x21e
3e: bf .byte 0xbf
3f: 01 .byte 0x1


Tested on:

commit: 2942242d Merge tag 'mm-hotfixes-stable-2025-07-24-18-0..
git tree: upstream
console output: https://syzkaller.appspot.com/x/log.txt?x=133490a2580000
kernel config: https://syzkaller.appspot.com/x/.config?x=c445d2891161cc81

dashboard link: https://syzkaller.appspot.com/bug?extid=a6ffe86390c8a6afc818
compiler: gcc (Debian 12.2.0-14) 12.2.0, GNU ld (GNU Binutils for Debian) 2.40

patch: https://syzkaller.appspot.com/x/patch.diff?x=17f994f0580000

[Edward Adam Davis]

#syz test

diff --git a/drivers/infiniband/core/ucma.c b/drivers/infiniband/core/ucma.c

index 6e700b974033..7fa02963a309 100644

--- a/drivers/infiniband/core/ucma.c
+++ b/drivers/infiniband/core/ucma.c
@@ -109,6 +109,7 @@ struct ucma_multicast {
u8 join_state;
struct list_head list;
struct sockaddr_storage addr;
+ atomic_t ref;
};

struct ucma_event {

@@ -257,6 +258,17 @@ static void ucma_copy_ud_event(struct ib_device *device,

dst->qkey = src->qkey;
}

+static bool ucma_get_mc(struct ucma_multicast *mc)
+{
+ return mc && atomic_inc_not_zero(&mc->ref);
+}
+

+static void ucma_put_mc(struct ucma_multicast *mc)
+{
+ if (mc && atomic_dec_and_test(&mc->ref))
+ kfree(mc);
+}
+
static struct ucma_event *ucma_create_uevent(struct ucma_context *ctx,
struct rdma_cm_event *event)
{

@@ -272,9 +284,12 @@ static struct ucma_event *ucma_create_uevent(struct ucma_context *ctx,

case RDMA_CM_EVENT_MULTICAST_ERROR:
uevent->mc = (struct ucma_multicast *)
event->param.ud.private_data;

- uevent->resp.uid = uevent->mc->uid;
- uevent->resp.id = uevent->mc->id;
- break;
+ if (ucma_get_mc(uevent->mc)) {
+ uevent->resp.uid = uevent->mc->uid;
+ uevent->resp.id = uevent->mc->id;
+ ucma_put_mc(uevent->mc);
+ break;
+ }

default:
uevent->resp.uid = ctx->uid;

uevent->resp.id = ctx->id;
@@ -498,7 +513,7 @@ static void ucma_cleanup_multicast(struct ucma_context *ctx)
* lock on the reader and this is enough serialization
*/
__xa_erase(&multicast_table, mc->id);
- kfree(mc);
+ ucma_put_mc(mc);
}
xa_unlock(&multicast_table);
}
@@ -1471,6 +1486,7 @@ static ssize_t ucma_process_join(struct ucma_file *file,

mc->ctx = ctx;
mc->join_state = join_state;
mc->uid = cmd->uid;
+ atomic_set(&mc->ref, 1);
memcpy(&mc->addr, addr, cmd->addr_size);

xa_lock(&multicast_table);

@@ -1490,6 +1506,7 @@ static ssize_t ucma_process_join(struct ucma_file *file,
if (ret)
goto err_xa_erase;

+ ucma_get_mc(mc);

resp.id = mc->id;
if (copy_to_user(u64_to_user_ptr(cmd->response),

&resp, sizeof(resp))) {
@@ -1506,6 +1523,7 @@ static ssize_t ucma_process_join(struct ucma_file *file,
mutex_lock(&ctx->mutex);

rdma_leave_multicast(ctx->cm_id, (struct sockaddr *) &mc->addr);
mutex_unlock(&ctx->mutex);

+ ucma_put_mc(mc);
ucma_cleanup_mc_events(mc);
err_xa_erase:
xa_lock(&multicast_table);
@@ -1513,7 +1531,7 @@ static ssize_t ucma_process_join(struct ucma_file *file,

__xa_erase(&multicast_table, mc->id);
err_free_mc:
xa_unlock(&multicast_table);
- kfree(mc);
+ ucma_put_mc(mc);
err_put_ctx:
ucma_put_ctx(ctx);
return ret;

@@ -1599,7 +1617,7 @@ static ssize_t ucma_leave_multicast(struct ucma_file *file,

ucma_put_ctx(mc->ctx);
resp.events_reported = mc->events_reported;
- kfree(mc);
+ ucma_put_mc(mc);

if (copy_to_user(u64_to_user_ptr(cmd.response),
&resp, sizeof(resp)))

[syzbot]

Hello,

syzbot has tested the proposed patch but the reproducer is still triggering an issue:

KASAN: slab-use-after-free Write in ucma_create_uevent

==================================================================
BUG: KASAN: slab-use-after-free in instrument_atomic_read_write include/linux/instrumented.h:96 [inline]
BUG: KASAN: slab-use-after-free in atomic_inc_not_zero include/linux/atomic/atomic-instrumented.h:1536 [inline]
BUG: KASAN: slab-use-after-free in ucma_get_mc drivers/infiniband/core/ucma.c:263 [inline]
BUG: KASAN: slab-use-after-free in ucma_create_uevent+0x8d1/0xd70 drivers/infiniband/core/ucma.c:287
Write of size 4 at addr ffff88802a9566b0 by task kworker/u32:5/1140

CPU: 2 UID: 0 PID: 1140 Comm: kworker/u32:5 Not tainted 6.16.0-rc7-syzkaller-g5f33ebd2018c-dirty #0 PREEMPT(full)

Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.3-debian-1.16.3-2~bpo12+1 04/01/2014

Workqueue: rdma_cm cma_iboe_join_work_handler
Call Trace:
<TASK>
__dump_stack lib/dump_stack.c:94 [inline]
dump_stack_lvl+0x116/0x1f0 lib/dump_stack.c:120

print_address_description mm/kasan/report.c:378 [inline]
print_report+0xcd/0x630 mm/kasan/report.c:482
kasan_report+0xe0/0x110 mm/kasan/report.c:595
check_region_inline mm/kasan/generic.c:183 [inline]
kasan_check_range+0x100/0x1b0 mm/kasan/generic.c:189
instrument_atomic_read_write include/linux/instrumented.h:96 [inline]
atomic_inc_not_zero include/linux/atomic/atomic-instrumented.h:1536 [inline]
ucma_get_mc drivers/infiniband/core/ucma.c:263 [inline]
ucma_create_uevent+0x8d1/0xd70 drivers/infiniband/core/ucma.c:287
ucma_event_handler+0x102/0x940 drivers/infiniband/core/ucma.c:366

cma_cm_event_handler+0x97/0x300 drivers/infiniband/core/cma.c:2173
cma_iboe_join_work_handler+0xca/0x170 drivers/infiniband/core/cma.c:3008

process_one_work+0x9cc/0x1b70 kernel/workqueue.c:3238

process_scheduled_works kernel/workqueue.c:3321 [inline]
worker_thread+0x6c8/0xf10 kernel/workqueue.c:3402

kthread+0x3c2/0x780 kernel/kthread.c:464
ret_from_fork+0x5d4/0x6f0 arch/x86/kernel/process.c:148
ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:245
</TASK>

Allocated by task 6548:

kasan_save_stack+0x33/0x60 mm/kasan/common.c:47
kasan_save_track+0x14/0x30 mm/kasan/common.c:68
poison_kmalloc_redzone mm/kasan/common.c:377 [inline]
__kasan_kmalloc+0xaa/0xb0 mm/kasan/common.c:394
kmalloc_noprof include/linux/slab.h:905 [inline]
kzalloc_noprof include/linux/slab.h:1039 [inline]

ucma_process_join+0x233/0xb90 drivers/infiniband/core/ucma.c:1480
ucma_join_multicast+0xe8/0x160 drivers/infiniband/core/ucma.c:1575
ucma_write+0x1fb/0x330 drivers/infiniband/core/ucma.c:1756

vfs_write+0x29d/0x1150 fs/read_write.c:684
ksys_write+0x1f8/0x250 fs/read_write.c:738
do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
do_syscall_64+0xcd/0x4c0 arch/x86/entry/syscall_64.c:94
entry_SYSCALL_64_after_hwframe+0x77/0x7f

Freed by task 6548:

kasan_save_stack+0x33/0x60 mm/kasan/common.c:47
kasan_save_track+0x14/0x30 mm/kasan/common.c:68
kasan_save_free_info+0x3b/0x60 mm/kasan/generic.c:576
poison_slab_object mm/kasan/common.c:247 [inline]
__kasan_slab_free+0x51/0x70 mm/kasan/common.c:264
kasan_slab_free include/linux/kasan.h:233 [inline]
slab_free_hook mm/slub.c:2381 [inline]
slab_free mm/slub.c:4643 [inline]
kfree+0x2b4/0x4d0 mm/slub.c:4842

ucma_put_mc drivers/infiniband/core/ucma.c:269 [inline]
ucma_put_mc drivers/infiniband/core/ucma.c:266 [inline]
ucma_process_join+0x763/0xb90 drivers/infiniband/core/ucma.c:1534
ucma_join_multicast+0xe8/0x160 drivers/infiniband/core/ucma.c:1575
ucma_write+0x1fb/0x330 drivers/infiniband/core/ucma.c:1756

vfs_write+0x29d/0x1150 fs/read_write.c:684
ksys_write+0x1f8/0x250 fs/read_write.c:738
do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
do_syscall_64+0xcd/0x4c0 arch/x86/entry/syscall_64.c:94
entry_SYSCALL_64_after_hwframe+0x77/0x7f

The buggy address belongs to the object at ffff88802a956600

which belongs to the cache kmalloc-192 of size 192

The buggy address is located 176 bytes inside of
freed 192-byte region [ffff88802a956600, ffff88802a9566c0)

The buggy address belongs to the physical page:

page: refcount:0 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x2a956

flags: 0xfff00000000000(node=0|zone=1|lastcpupid=0x7ff)
page_type: f5(slab)

raw: 00fff00000000000 ffff88801b8423c0 dead000000000122 0000000000000000

raw: 0000000000000000 0000000080100010 00000000f5000000 0000000000000000
page dumped because: kasan: bad access detected
page_owner tracks the page as allocated

page last allocated via order 0, migratetype Unmovable, gfp_mask 0x252800(GFP_NOWAIT|__GFP_NORETRY|__GFP_COMP|__GFP_THISNODE), pid 6445, tgid 6445 (syz-executor), ts 92358145050, free_ts 92277395639

set_page_owner include/linux/page_owner.h:32 [inline]
post_alloc_hook+0x1c0/0x230 mm/page_alloc.c:1704
prep_new_page mm/page_alloc.c:1712 [inline]
get_page_from_freelist+0x1321/0x3890 mm/page_alloc.c:3669
__alloc_frozen_pages_noprof+0x261/0x23f0 mm/page_alloc.c:4959

alloc_slab_page mm/slub.c:2453 [inline]
allocate_slab mm/slub.c:2619 [inline]
new_slab+0x94/0x330 mm/slub.c:2673

___slab_alloc+0xd9c/0x1940 mm/slub.c:3859
__slab_alloc.constprop.0+0x56/0xb0 mm/slub.c:3949
__slab_alloc_node mm/slub.c:4024 [inline]
slab_alloc_node mm/slub.c:4185 [inline]

__do_kmalloc_node mm/slub.c:4327 [inline]
__kmalloc_node_noprof+0x2ed/0x500 mm/slub.c:4334
kmalloc_array_node_noprof include/linux/slab.h:1020 [inline]
alloc_slab_obj_exts+0x41/0xa0 mm/slub.c:1992
account_slab mm/slub.c:2578 [inline]
allocate_slab mm/slub.c:2638 [inline]
new_slab+0x283/0x330 mm/slub.c:2673

___slab_alloc+0xd9c/0x1940 mm/slub.c:3859
__slab_alloc.constprop.0+0x56/0xb0 mm/slub.c:3949
__slab_alloc_node mm/slub.c:4024 [inline]
slab_alloc_node mm/slub.c:4185 [inline]

kmem_cache_alloc_lru_noprof+0xf4/0x3b0 mm/slub.c:4216
sock_alloc_inode+0x25/0x1c0 net/socket.c:309
alloc_inode+0x61/0x240 fs/inode.c:346
new_inode_pseudo include/linux/fs.h:3364 [inline]
sock_alloc+0x40/0x280 net/socket.c:622
__sock_create+0xc1/0x8d0 net/socket.c:1505
page last free pid 23 tgid 23 stack trace:
reset_page_owner include/linux/page_owner.h:25 [inline]
free_pages_prepare mm/page_alloc.c:1248 [inline]
__free_frozen_pages+0x7fe/0x1180 mm/page_alloc.c:2706
pagetable_free include/linux/mm.h:2884 [inline]
pagetable_dtor_free include/linux/mm.h:2982 [inline]
__tlb_remove_table include/asm-generic/tlb.h:220 [inline]
__tlb_remove_table_free mm/mmu_gather.c:227 [inline]
tlb_remove_table_rcu+0x116/0x1a0 mm/mmu_gather.c:290
rcu_do_batch kernel/rcu/tree.c:2576 [inline]
rcu_core+0x79c/0x14e0 kernel/rcu/tree.c:2832
handle_softirqs+0x219/0x8e0 kernel/softirq.c:579
run_ksoftirqd kernel/softirq.c:968 [inline]
run_ksoftirqd+0x3a/0x60 kernel/softirq.c:960
smpboot_thread_fn+0x3f4/0xae0 kernel/smpboot.c:164
kthread+0x3c2/0x780 kernel/kthread.c:464
ret_from_fork+0x5d4/0x6f0 arch/x86/kernel/process.c:148
ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:245

Memory state around the buggy address:

ffff88802a956580: 00 00 00 00 00 fc fc fc fc fc fc fc fc fc fc fc
ffff88802a956600: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
>ffff88802a956680: fb fb fb fb fb fb fb fb fc fc fc fc fc fc fc fc
^
ffff88802a956700: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
ffff88802a956780: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
==================================================================


Tested on:

commit: 5f33ebd2 Merge tag 'drm-fixes-2025-07-26' of https://g..
git tree: upstream
console output: https://syzkaller.appspot.com/x/log.txt?x=17ea88a2580000

kernel config: https://syzkaller.appspot.com/x/.config?x=c445d2891161cc81
dashboard link: https://syzkaller.appspot.com/bug?extid=a6ffe86390c8a6afc818
compiler: gcc (Debian 12.2.0-14) 12.2.0, GNU ld (GNU Binutils for Debian) 2.40

patch: https://syzkaller.appspot.com/x/patch.diff?x=12100782580000