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

[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+a6ffe86390c8a6afc818@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 syzkaller@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

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

[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;
--

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

[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);

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

[syzbot]

Hello,

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

Reported-by: syzbot+a6ffe86390c8a6afc818@syzkaller.appspotmail.com
Tested-by: syzbot+a6ffe86390c8a6afc818@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.

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

[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

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

[syzbot]

For archival purposes, forwarding an incoming command email to
linux-kernel@vger.kernel.org.

***

Subject: Re: [syzbot] [rdma?] KASAN: slab-use-after-free Read in ucma_create_uevent
Author: lizhi.xu@windriver.com

#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);

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

[syzbot]

Hello,

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

Reported-by: syzbot+a6ffe86390c8a6afc818@syzkaller.appspotmail.com
Tested-by: syzbot+a6ffe86390c8a6afc818@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

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

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

[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

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

[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)))

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

[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