From 076381c261374c587700b3accf410bdd2dba334e Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Mon, 3 Nov 2025 21:34:01 +0100 Subject: [PATCH 1/5] libceph: fix potential use-after-free in have_mon_and_osd_map() The wait loop in __ceph_open_session() can race with the client receiving a new monmap or osdmap shortly after the initial map is received. Both ceph_monc_handle_map() and handle_one_map() install a new map immediately after freeing the old one kfree(monc->monmap); monc->monmap = monmap; ceph_osdmap_destroy(osdc->osdmap); osdc->osdmap = newmap; under client->monc.mutex and client->osdc.lock respectively, but because neither is taken in have_mon_and_osd_map() it's possible for client->monc.monmap->epoch and client->osdc.osdmap->epoch arms in client->monc.monmap && client->monc.monmap->epoch && client->osdc.osdmap && client->osdc.osdmap->epoch; condition to dereference an already freed map. This happens to be reproducible with generic/395 and generic/397 with KASAN enabled: BUG: KASAN: slab-use-after-free in have_mon_and_osd_map+0x56/0x70 Read of size 4 at addr ffff88811012d810 by task mount.ceph/13305 CPU: 2 UID: 0 PID: 13305 Comm: mount.ceph Not tainted 6.14.0-rc2-build2+ #1266 ... Call Trace: have_mon_and_osd_map+0x56/0x70 ceph_open_session+0x182/0x290 ceph_get_tree+0x333/0x680 vfs_get_tree+0x49/0x180 do_new_mount+0x1a3/0x2d0 path_mount+0x6dd/0x730 do_mount+0x99/0xe0 __do_sys_mount+0x141/0x180 do_syscall_64+0x9f/0x100 entry_SYSCALL_64_after_hwframe+0x76/0x7e Allocated by task 13305: ceph_osdmap_alloc+0x16/0x130 ceph_osdc_init+0x27a/0x4c0 ceph_create_client+0x153/0x190 create_fs_client+0x50/0x2a0 ceph_get_tree+0xff/0x680 vfs_get_tree+0x49/0x180 do_new_mount+0x1a3/0x2d0 path_mount+0x6dd/0x730 do_mount+0x99/0xe0 __do_sys_mount+0x141/0x180 do_syscall_64+0x9f/0x100 entry_SYSCALL_64_after_hwframe+0x76/0x7e Freed by task 9475: kfree+0x212/0x290 handle_one_map+0x23c/0x3b0 ceph_osdc_handle_map+0x3c9/0x590 mon_dispatch+0x655/0x6f0 ceph_con_process_message+0xc3/0xe0 ceph_con_v1_try_read+0x614/0x760 ceph_con_workfn+0x2de/0x650 process_one_work+0x486/0x7c0 process_scheduled_works+0x73/0x90 worker_thread+0x1c8/0x2a0 kthread+0x2ec/0x300 ret_from_fork+0x24/0x40 ret_from_fork_asm+0x1a/0x30 Rewrite the wait loop to check the above condition directly with client->monc.mutex and client->osdc.lock taken as appropriate. While at it, improve the timeout handling (previously mount_timeout could be exceeded in case wait_event_interruptible_timeout() slept more than once) and access client->auth_err under client->monc.mutex to match how it's set in finish_auth(). monmap_show() and osdmap_show() now take the respective lock before accessing the map as well. Cc: stable@vger.kernel.org Reported-by: David Howells Signed-off-by: Ilya Dryomov Reviewed-by: Viacheslav Dubeyko --- net/ceph/ceph_common.c | 53 +++++++++++++++++++++++++----------------- net/ceph/debugfs.c | 16 +++++++++---- 2 files changed, 43 insertions(+), 26 deletions(-) diff --git a/net/ceph/ceph_common.c b/net/ceph/ceph_common.c index 4c6441536d55..285e981730e5 100644 --- a/net/ceph/ceph_common.c +++ b/net/ceph/ceph_common.c @@ -785,42 +785,53 @@ void ceph_reset_client_addr(struct ceph_client *client) } EXPORT_SYMBOL(ceph_reset_client_addr); -/* - * true if we have the mon map (and have thus joined the cluster) - */ -static bool have_mon_and_osd_map(struct ceph_client *client) -{ - return client->monc.monmap && client->monc.monmap->epoch && - client->osdc.osdmap && client->osdc.osdmap->epoch; -} - /* * mount: join the ceph cluster, and open root directory. */ int __ceph_open_session(struct ceph_client *client, unsigned long started) { - unsigned long timeout = client->options->mount_timeout; - long err; + DEFINE_WAIT_FUNC(wait, woken_wake_function); + long timeout = ceph_timeout_jiffies(client->options->mount_timeout); + bool have_monmap, have_osdmap; + int err; /* open session, and wait for mon and osd maps */ err = ceph_monc_open_session(&client->monc); if (err < 0) return err; - while (!have_mon_and_osd_map(client)) { - if (timeout && time_after_eq(jiffies, started + timeout)) - return -ETIMEDOUT; + add_wait_queue(&client->auth_wq, &wait); + for (;;) { + mutex_lock(&client->monc.mutex); + err = client->auth_err; + have_monmap = client->monc.monmap && client->monc.monmap->epoch; + mutex_unlock(&client->monc.mutex); + + down_read(&client->osdc.lock); + have_osdmap = client->osdc.osdmap && client->osdc.osdmap->epoch; + up_read(&client->osdc.lock); + + if (err || (have_monmap && have_osdmap)) + break; + + if (signal_pending(current)) { + err = -ERESTARTSYS; + break; + } + + if (!timeout) { + err = -ETIMEDOUT; + break; + } /* wait */ dout("mount waiting for mon_map\n"); - err = wait_event_interruptible_timeout(client->auth_wq, - have_mon_and_osd_map(client) || (client->auth_err < 0), - ceph_timeout_jiffies(timeout)); - if (err < 0) - return err; - if (client->auth_err < 0) - return client->auth_err; + timeout = wait_woken(&wait, TASK_INTERRUPTIBLE, timeout); } + remove_wait_queue(&client->auth_wq, &wait); + + if (err) + return err; pr_info("client%llu fsid %pU\n", ceph_client_gid(client), &client->fsid); diff --git a/net/ceph/debugfs.c b/net/ceph/debugfs.c index 2110439f8a24..83c270bce63c 100644 --- a/net/ceph/debugfs.c +++ b/net/ceph/debugfs.c @@ -36,8 +36,9 @@ static int monmap_show(struct seq_file *s, void *p) int i; struct ceph_client *client = s->private; + mutex_lock(&client->monc.mutex); if (client->monc.monmap == NULL) - return 0; + goto out_unlock; seq_printf(s, "epoch %d\n", client->monc.monmap->epoch); for (i = 0; i < client->monc.monmap->num_mon; i++) { @@ -48,6 +49,9 @@ static int monmap_show(struct seq_file *s, void *p) ENTITY_NAME(inst->name), ceph_pr_addr(&inst->addr)); } + +out_unlock: + mutex_unlock(&client->monc.mutex); return 0; } @@ -56,13 +60,14 @@ static int osdmap_show(struct seq_file *s, void *p) int i; struct ceph_client *client = s->private; struct ceph_osd_client *osdc = &client->osdc; - struct ceph_osdmap *map = osdc->osdmap; + struct ceph_osdmap *map; struct rb_node *n; - if (map == NULL) - return 0; - down_read(&osdc->lock); + map = osdc->osdmap; + if (map == NULL) + goto out_unlock; + seq_printf(s, "epoch %u barrier %u flags 0x%x\n", map->epoch, osdc->epoch_barrier, map->flags); @@ -131,6 +136,7 @@ static int osdmap_show(struct seq_file *s, void *p) seq_printf(s, "]\n"); } +out_unlock: up_read(&osdc->lock); return 0; } From 85f5491d9c6e9662653c8e6e7b70637b98537ecc Mon Sep 17 00:00:00 2001 From: Ilya Dryomov Date: Mon, 3 Nov 2025 21:34:01 +0100 Subject: [PATCH 2/5] libceph: drop started parameter of __ceph_open_session() With the previous commit revamping the timeout handling, started isn't used anymore. It could be taken into account by adjusting the initial value of the timeout, but there is little point as both callers capture the timestamp shortly before calling __ceph_open_session() -- the only thing of note that happens in the interim is taking client->mount_mutex and that isn't expected to take multiple seconds. Signed-off-by: Ilya Dryomov Reviewed-by: Viacheslav Dubeyko --- fs/ceph/super.c | 2 +- include/linux/ceph/libceph.h | 3 +-- net/ceph/ceph_common.c | 5 ++--- 3 files changed, 4 insertions(+), 6 deletions(-) diff --git a/fs/ceph/super.c b/fs/ceph/super.c index ad0cf177e75a..f6bf24b5c683 100644 --- a/fs/ceph/super.c +++ b/fs/ceph/super.c @@ -1149,7 +1149,7 @@ static struct dentry *ceph_real_mount(struct ceph_fs_client *fsc, const char *path = fsc->mount_options->server_path ? fsc->mount_options->server_path + 1 : ""; - err = __ceph_open_session(fsc->client, started); + err = __ceph_open_session(fsc->client); if (err < 0) goto out; diff --git a/include/linux/ceph/libceph.h b/include/linux/ceph/libceph.h index 733e7f93db66..63e0e2aa1ce9 100644 --- a/include/linux/ceph/libceph.h +++ b/include/linux/ceph/libceph.h @@ -306,8 +306,7 @@ struct ceph_entity_addr *ceph_client_addr(struct ceph_client *client); u64 ceph_client_gid(struct ceph_client *client); extern void ceph_destroy_client(struct ceph_client *client); extern void ceph_reset_client_addr(struct ceph_client *client); -extern int __ceph_open_session(struct ceph_client *client, - unsigned long started); +extern int __ceph_open_session(struct ceph_client *client); extern int ceph_open_session(struct ceph_client *client); int ceph_wait_for_latest_osdmap(struct ceph_client *client, unsigned long timeout); diff --git a/net/ceph/ceph_common.c b/net/ceph/ceph_common.c index 285e981730e5..e734e57be083 100644 --- a/net/ceph/ceph_common.c +++ b/net/ceph/ceph_common.c @@ -788,7 +788,7 @@ EXPORT_SYMBOL(ceph_reset_client_addr); /* * mount: join the ceph cluster, and open root directory. */ -int __ceph_open_session(struct ceph_client *client, unsigned long started) +int __ceph_open_session(struct ceph_client *client) { DEFINE_WAIT_FUNC(wait, woken_wake_function); long timeout = ceph_timeout_jiffies(client->options->mount_timeout); @@ -844,12 +844,11 @@ EXPORT_SYMBOL(__ceph_open_session); int ceph_open_session(struct ceph_client *client) { int ret; - unsigned long started = jiffies; /* note the start time */ dout("open_session start\n"); mutex_lock(&client->mount_mutex); - ret = __ceph_open_session(client, started); + ret = __ceph_open_session(client); mutex_unlock(&client->mount_mutex); return ret; From 43962db4a6f593903340c85591056a0cef812dfd Mon Sep 17 00:00:00 2001 From: Viacheslav Dubeyko Date: Thu, 13 Nov 2025 14:36:24 -0800 Subject: [PATCH 3/5] ceph: fix crash in process_v2_sparse_read() for encrypted directories The crash in process_v2_sparse_read() for fscrypt-encrypted directories has been reported. Issue takes place for Ceph msgr2 protocol in secure mode. It can be reproduced by the steps: sudo mount -t ceph :/ /mnt/cephfs/ -o name=admin,fs=cephfs,ms_mode=secure (1) mkdir /mnt/cephfs/fscrypt-test-3 (2) cp area_decrypted.tar /mnt/cephfs/fscrypt-test-3 (3) fscrypt encrypt --source=raw_key --key=./my.key /mnt/cephfs/fscrypt-test-3 (4) fscrypt lock /mnt/cephfs/fscrypt-test-3 (5) fscrypt unlock --key=my.key /mnt/cephfs/fscrypt-test-3 (6) cat /mnt/cephfs/fscrypt-test-3/area_decrypted.tar (7) Issue has been triggered [ 408.072247] ------------[ cut here ]------------ [ 408.072251] WARNING: CPU: 1 PID: 392 at net/ceph/messenger_v2.c:865 ceph_con_v2_try_read+0x4b39/0x72f0 [ 408.072267] Modules linked in: intel_rapl_msr intel_rapl_common intel_uncore_frequency_common intel_pmc_core pmt_telemetry pmt_discovery pmt_class intel_pmc_ssram_telemetry intel_vsec kvm_intel joydev kvm irqbypass polyval_clmulni ghash_clmulni_intel aesni_intel rapl input_leds psmouse serio_raw i2c_piix4 vga16fb bochs vgastate i2c_smbus floppy mac_hid qemu_fw_cfg pata_acpi sch_fq_codel rbd msr parport_pc ppdev lp parport efi_pstore [ 408.072304] CPU: 1 UID: 0 PID: 392 Comm: kworker/1:3 Not tainted 6.17.0-rc7+ [ 408.072307] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.17.0-5.fc42 04/01/2014 [ 408.072310] Workqueue: ceph-msgr ceph_con_workfn [ 408.072314] RIP: 0010:ceph_con_v2_try_read+0x4b39/0x72f0 [ 408.072317] Code: c7 c1 20 f0 d4 ae 50 31 d2 48 c7 c6 60 27 d5 ae 48 c7 c7 f8 8e 6f b0 68 60 38 d5 ae e8 00 47 61 fe 48 83 c4 18 e9 ac fc ff ff <0f> 0b e9 06 fe ff ff 4c 8b 9d 98 fd ff ff 0f 84 64 e7 ff ff 89 85 [ 408.072319] RSP: 0018:ffff88811c3e7a30 EFLAGS: 00010246 [ 408.072322] RAX: ffffed1024874c6f RBX: ffffea00042c2b40 RCX: 0000000000000f38 [ 408.072324] RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000000 [ 408.072325] RBP: ffff88811c3e7ca8 R08: 0000000000000000 R09: 00000000000000c8 [ 408.072326] R10: 00000000000000c8 R11: 0000000000000000 R12: 00000000000000c8 [ 408.072327] R13: dffffc0000000000 R14: ffff8881243a6030 R15: 0000000000003000 [ 408.072329] FS: 0000000000000000(0000) GS:ffff88823eadf000(0000) knlGS:0000000000000000 [ 408.072331] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 408.072332] CR2: 000000c0003c6000 CR3: 000000010c106005 CR4: 0000000000772ef0 [ 408.072336] PKRU: 55555554 [ 408.072337] Call Trace: [ 408.072338] [ 408.072340] ? sched_clock_noinstr+0x9/0x10 [ 408.072344] ? __pfx_ceph_con_v2_try_read+0x10/0x10 [ 408.072347] ? _raw_spin_unlock+0xe/0x40 [ 408.072349] ? finish_task_switch.isra.0+0x15d/0x830 [ 408.072353] ? __kasan_check_write+0x14/0x30 [ 408.072357] ? mutex_lock+0x84/0xe0 [ 408.072359] ? __pfx_mutex_lock+0x10/0x10 [ 408.072361] ceph_con_workfn+0x27e/0x10e0 [ 408.072364] ? metric_delayed_work+0x311/0x2c50 [ 408.072367] process_one_work+0x611/0xe20 [ 408.072371] ? __kasan_check_write+0x14/0x30 [ 408.072373] worker_thread+0x7e3/0x1580 [ 408.072375] ? __pfx__raw_spin_lock_irqsave+0x10/0x10 [ 408.072378] ? __pfx_worker_thread+0x10/0x10 [ 408.072381] kthread+0x381/0x7a0 [ 408.072383] ? __pfx__raw_spin_lock_irq+0x10/0x10 [ 408.072385] ? __pfx_kthread+0x10/0x10 [ 408.072387] ? __kasan_check_write+0x14/0x30 [ 408.072389] ? recalc_sigpending+0x160/0x220 [ 408.072392] ? _raw_spin_unlock_irq+0xe/0x50 [ 408.072394] ? calculate_sigpending+0x78/0xb0 [ 408.072395] ? __pfx_kthread+0x10/0x10 [ 408.072397] ret_from_fork+0x2b6/0x380 [ 408.072400] ? __pfx_kthread+0x10/0x10 [ 408.072402] ret_from_fork_asm+0x1a/0x30 [ 408.072406] [ 408.072407] ---[ end trace 0000000000000000 ]--- [ 408.072418] Oops: general protection fault, probably for non-canonical address 0xdffffc0000000000: 0000 [#1] SMP KASAN NOPTI [ 408.072984] KASAN: null-ptr-deref in range [0x0000000000000000- 0x0000000000000007] [ 408.073350] CPU: 1 UID: 0 PID: 392 Comm: kworker/1:3 Tainted: G W 6.17.0-rc7+ #1 PREEMPT(voluntary) [ 408.073886] Tainted: [W]=WARN [ 408.074042] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.17.0-5.fc42 04/01/2014 [ 408.074468] Workqueue: ceph-msgr ceph_con_workfn [ 408.074694] RIP: 0010:ceph_msg_data_advance+0x79/0x1a80 [ 408.074976] Code: fc ff df 49 8d 77 08 48 c1 ee 03 80 3c 16 00 0f 85 07 11 00 00 48 ba 00 00 00 00 00 fc ff df 49 8b 5f 08 48 89 de 48 c1 ee 03 <0f> b6 14 16 84 d2 74 09 80 fa 03 0f 8e 0f 0e 00 00 8b 13 83 fa 03 [ 408.075884] RSP: 0018:ffff88811c3e7990 EFLAGS: 00010246 [ 408.076305] RAX: ffff8881243a6388 RBX: 0000000000000000 RCX: 0000000000000000 [ 408.076909] RDX: dffffc0000000000 RSI: 0000000000000000 RDI: ffff8881243a6378 [ 408.077466] RBP: ffff88811c3e7a20 R08: 0000000000000000 R09: 00000000000000c8 [ 408.078034] R10: ffff8881243a6388 R11: 0000000000000000 R12: ffffed1024874c71 [ 408.078575] R13: dffffc0000000000 R14: ffff8881243a6030 R15: ffff8881243a6378 [ 408.079159] FS: 0000000000000000(0000) GS:ffff88823eadf000(0000) knlGS:0000000000000000 [ 408.079736] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 408.080039] CR2: 000000c0003c6000 CR3: 000000010c106005 CR4: 0000000000772ef0 [ 408.080376] PKRU: 55555554 [ 408.080513] Call Trace: [ 408.080630] [ 408.080729] ceph_con_v2_try_read+0x49b9/0x72f0 [ 408.081115] ? __pfx_ceph_con_v2_try_read+0x10/0x10 [ 408.081348] ? _raw_spin_unlock+0xe/0x40 [ 408.081538] ? finish_task_switch.isra.0+0x15d/0x830 [ 408.081768] ? __kasan_check_write+0x14/0x30 [ 408.081986] ? mutex_lock+0x84/0xe0 [ 408.082160] ? __pfx_mutex_lock+0x10/0x10 [ 408.082343] ceph_con_workfn+0x27e/0x10e0 [ 408.082529] ? metric_delayed_work+0x311/0x2c50 [ 408.082737] process_one_work+0x611/0xe20 [ 408.082948] ? __kasan_check_write+0x14/0x30 [ 408.083156] worker_thread+0x7e3/0x1580 [ 408.083331] ? __pfx__raw_spin_lock_irqsave+0x10/0x10 [ 408.083557] ? __pfx_worker_thread+0x10/0x10 [ 408.083751] kthread+0x381/0x7a0 [ 408.083922] ? __pfx__raw_spin_lock_irq+0x10/0x10 [ 408.084139] ? __pfx_kthread+0x10/0x10 [ 408.084310] ? __kasan_check_write+0x14/0x30 [ 408.084510] ? recalc_sigpending+0x160/0x220 [ 408.084708] ? _raw_spin_unlock_irq+0xe/0x50 [ 408.084917] ? calculate_sigpending+0x78/0xb0 [ 408.085138] ? __pfx_kthread+0x10/0x10 [ 408.085335] ret_from_fork+0x2b6/0x380 [ 408.085525] ? __pfx_kthread+0x10/0x10 [ 408.085720] ret_from_fork_asm+0x1a/0x30 [ 408.085922] [ 408.086036] Modules linked in: intel_rapl_msr intel_rapl_common intel_uncore_frequency_common intel_pmc_core pmt_telemetry pmt_discovery pmt_class intel_pmc_ssram_telemetry intel_vsec kvm_intel joydev kvm irqbypass polyval_clmulni ghash_clmulni_intel aesni_intel rapl input_leds psmouse serio_raw i2c_piix4 vga16fb bochs vgastate i2c_smbus floppy mac_hid qemu_fw_cfg pata_acpi sch_fq_codel rbd msr parport_pc ppdev lp parport efi_pstore [ 408.087778] ---[ end trace 0000000000000000 ]--- [ 408.088007] RIP: 0010:ceph_msg_data_advance+0x79/0x1a80 [ 408.088260] Code: fc ff df 49 8d 77 08 48 c1 ee 03 80 3c 16 00 0f 85 07 11 00 00 48 ba 00 00 00 00 00 fc ff df 49 8b 5f 08 48 89 de 48 c1 ee 03 <0f> b6 14 16 84 d2 74 09 80 fa 03 0f 8e 0f 0e 00 00 8b 13 83 fa 03 [ 408.089118] RSP: 0018:ffff88811c3e7990 EFLAGS: 00010246 [ 408.089357] RAX: ffff8881243a6388 RBX: 0000000000000000 RCX: 0000000000000000 [ 408.089678] RDX: dffffc0000000000 RSI: 0000000000000000 RDI: ffff8881243a6378 [ 408.090020] RBP: ffff88811c3e7a20 R08: 0000000000000000 R09: 00000000000000c8 [ 408.090360] R10: ffff8881243a6388 R11: 0000000000000000 R12: ffffed1024874c71 [ 408.090687] R13: dffffc0000000000 R14: ffff8881243a6030 R15: ffff8881243a6378 [ 408.091035] FS: 0000000000000000(0000) GS:ffff88823eadf000(0000) knlGS:0000000000000000 [ 408.091452] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 408.092015] CR2: 000000c0003c6000 CR3: 000000010c106005 CR4: 0000000000772ef0 [ 408.092530] PKRU: 55555554 [ 417.112915] ================================================================== [ 417.113491] BUG: KASAN: slab-use-after-free in __mutex_lock.constprop.0+0x1522/0x1610 [ 417.114014] Read of size 4 at addr ffff888124870034 by task kworker/2:0/4951 [ 417.114587] CPU: 2 UID: 0 PID: 4951 Comm: kworker/2:0 Tainted: G D W 6.17.0-rc7+ #1 PREEMPT(voluntary) [ 417.114592] Tainted: [D]=DIE, [W]=WARN [ 417.114593] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.17.0-5.fc42 04/01/2014 [ 417.114596] Workqueue: events handle_timeout [ 417.114601] Call Trace: [ 417.114602] [ 417.114604] dump_stack_lvl+0x5c/0x90 [ 417.114610] print_report+0x171/0x4dc [ 417.114613] ? __pfx__raw_spin_lock_irqsave+0x10/0x10 [ 417.114617] ? kasan_complete_mode_report_info+0x80/0x220 [ 417.114621] kasan_report+0xbd/0x100 [ 417.114625] ? __mutex_lock.constprop.0+0x1522/0x1610 [ 417.114628] ? __mutex_lock.constprop.0+0x1522/0x1610 [ 417.114630] __asan_report_load4_noabort+0x14/0x30 [ 417.114633] __mutex_lock.constprop.0+0x1522/0x1610 [ 417.114635] ? queue_con_delay+0x8d/0x200 [ 417.114638] ? __pfx___mutex_lock.constprop.0+0x10/0x10 [ 417.114641] ? __send_subscribe+0x529/0xb20 [ 417.114644] __mutex_lock_slowpath+0x13/0x20 [ 417.114646] mutex_lock+0xd4/0xe0 [ 417.114649] ? __pfx_mutex_lock+0x10/0x10 [ 417.114652] ? ceph_monc_renew_subs+0x2a/0x40 [ 417.114654] ceph_con_keepalive+0x22/0x110 [ 417.114656] handle_timeout+0x6b3/0x11d0 [ 417.114659] ? _raw_spin_unlock_irq+0xe/0x50 [ 417.114662] ? __pfx_handle_timeout+0x10/0x10 [ 417.114664] ? queue_delayed_work_on+0x8e/0xa0 [ 417.114669] process_one_work+0x611/0xe20 [ 417.114672] ? __kasan_check_write+0x14/0x30 [ 417.114676] worker_thread+0x7e3/0x1580 [ 417.114678] ? __pfx__raw_spin_lock_irqsave+0x10/0x10 [ 417.114682] ? __pfx_sched_setscheduler_nocheck+0x10/0x10 [ 417.114687] ? __pfx_worker_thread+0x10/0x10 [ 417.114689] kthread+0x381/0x7a0 [ 417.114692] ? __pfx__raw_spin_lock_irq+0x10/0x10 [ 417.114694] ? __pfx_kthread+0x10/0x10 [ 417.114697] ? __kasan_check_write+0x14/0x30 [ 417.114699] ? recalc_sigpending+0x160/0x220 [ 417.114703] ? _raw_spin_unlock_irq+0xe/0x50 [ 417.114705] ? calculate_sigpending+0x78/0xb0 [ 417.114707] ? __pfx_kthread+0x10/0x10 [ 417.114710] ret_from_fork+0x2b6/0x380 [ 417.114713] ? __pfx_kthread+0x10/0x10 [ 417.114715] ret_from_fork_asm+0x1a/0x30 [ 417.114720] [ 417.125171] Allocated by task 2: [ 417.125333] kasan_save_stack+0x26/0x60 [ 417.125522] kasan_save_track+0x14/0x40 [ 417.125742] kasan_save_alloc_info+0x39/0x60 [ 417.125945] __kasan_slab_alloc+0x8b/0xb0 [ 417.126133] kmem_cache_alloc_node_noprof+0x13b/0x460 [ 417.126381] copy_process+0x320/0x6250 [ 417.126595] kernel_clone+0xb7/0x840 [ 417.126792] kernel_thread+0xd6/0x120 [ 417.126995] kthreadd+0x85c/0xbe0 [ 417.127176] ret_from_fork+0x2b6/0x380 [ 417.127378] ret_from_fork_asm+0x1a/0x30 [ 417.127692] Freed by task 0: [ 417.127851] kasan_save_stack+0x26/0x60 [ 417.128057] kasan_save_track+0x14/0x40 [ 417.128267] kasan_save_free_info+0x3b/0x60 [ 417.128491] __kasan_slab_free+0x6c/0xa0 [ 417.128708] kmem_cache_free+0x182/0x550 [ 417.128906] free_task+0xeb/0x140 [ 417.129070] __put_task_struct+0x1d2/0x4f0 [ 417.129259] __put_task_struct_rcu_cb+0x15/0x20 [ 417.129480] rcu_do_batch+0x3d3/0xe70 [ 417.129681] rcu_core+0x549/0xb30 [ 417.129839] rcu_core_si+0xe/0x20 [ 417.130005] handle_softirqs+0x160/0x570 [ 417.130190] __irq_exit_rcu+0x189/0x1e0 [ 417.130369] irq_exit_rcu+0xe/0x20 [ 417.130531] sysvec_apic_timer_interrupt+0x9f/0xd0 [ 417.130768] asm_sysvec_apic_timer_interrupt+0x1b/0x20 [ 417.131082] Last potentially related work creation: [ 417.131305] kasan_save_stack+0x26/0x60 [ 417.131484] kasan_record_aux_stack+0xae/0xd0 [ 417.131695] __call_rcu_common+0xcd/0x14b0 [ 417.131909] call_rcu+0x31/0x50 [ 417.132071] delayed_put_task_struct+0x128/0x190 [ 417.132295] rcu_do_batch+0x3d3/0xe70 [ 417.132478] rcu_core+0x549/0xb30 [ 417.132658] rcu_core_si+0xe/0x20 [ 417.132808] handle_softirqs+0x160/0x570 [ 417.132993] __irq_exit_rcu+0x189/0x1e0 [ 417.133181] irq_exit_rcu+0xe/0x20 [ 417.133353] sysvec_apic_timer_interrupt+0x9f/0xd0 [ 417.133584] asm_sysvec_apic_timer_interrupt+0x1b/0x20 [ 417.133921] Second to last potentially related work creation: [ 417.134183] kasan_save_stack+0x26/0x60 [ 417.134362] kasan_record_aux_stack+0xae/0xd0 [ 417.134566] __call_rcu_common+0xcd/0x14b0 [ 417.134782] call_rcu+0x31/0x50 [ 417.134929] put_task_struct_rcu_user+0x58/0xb0 [ 417.135143] finish_task_switch.isra.0+0x5d3/0x830 [ 417.135366] __schedule+0xd30/0x5100 [ 417.135534] schedule_idle+0x5a/0x90 [ 417.135712] do_idle+0x25f/0x410 [ 417.135871] cpu_startup_entry+0x53/0x70 [ 417.136053] start_secondary+0x216/0x2c0 [ 417.136233] common_startup_64+0x13e/0x141 [ 417.136894] The buggy address belongs to the object at ffff888124870000 which belongs to the cache task_struct of size 10504 [ 417.138122] The buggy address is located 52 bytes inside of freed 10504-byte region [ffff888124870000, ffff888124872908) [ 417.139465] The buggy address belongs to the physical page: [ 417.140016] page: refcount:0 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x124870 [ 417.140789] head: order:3 mapcount:0 entire_mapcount:0 nr_pages_mapped:0 pincount:0 [ 417.141519] memcg:ffff88811aa20e01 [ 417.141874] anon flags: 0x17ffffc0000040(head|node=0|zone=2|lastcpupid=0x1fffff) [ 417.142600] page_type: f5(slab) [ 417.142922] raw: 0017ffffc0000040 ffff88810094f040 0000000000000000 dead000000000001 [ 417.143554] raw: 0000000000000000 0000000000030003 00000000f5000000 ffff88811aa20e01 [ 417.143954] head: 0017ffffc0000040 ffff88810094f040 0000000000000000 dead000000000001 [ 417.144329] head: 0000000000000000 0000000000030003 00000000f5000000 ffff88811aa20e01 [ 417.144710] head: 0017ffffc0000003 ffffea0004921c01 00000000ffffffff 00000000ffffffff [ 417.145106] head: ffffffffffffffff 0000000000000000 00000000ffffffff 0000000000000008 [ 417.145485] page dumped because: kasan: bad access detected [ 417.145859] Memory state around the buggy address: [ 417.146094] ffff88812486ff00: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc [ 417.146439] ffff88812486ff80: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc [ 417.146791] >ffff888124870000: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb [ 417.147145] ^ [ 417.147387] ffff888124870080: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb [ 417.147751] ffff888124870100: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb [ 417.148123] ================================================================== First of all, we have warning in get_bvec_at() because cursor->total_resid contains zero value. And, finally, we have crash in ceph_msg_data_advance() because cursor->data is NULL. It means that get_bvec_at() receives not initialized ceph_msg_data_cursor structure because data is NULL and total_resid contains zero. Moreover, we don't have likewise issue for the case of Ceph msgr1 protocol because ceph_msg_data_cursor_init() has been called before reading sparse data. This patch adds calling of ceph_msg_data_cursor_init() in the beginning of process_v2_sparse_read() with the goal to guarantee that logic of reading sparse data works correctly for the case of Ceph msgr2 protocol. Cc: stable@vger.kernel.org Link: https://tracker.ceph.com/issues/73152 Signed-off-by: Viacheslav Dubeyko Reviewed-by: Ilya Dryomov Signed-off-by: Ilya Dryomov --- net/ceph/messenger_v2.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) diff --git a/net/ceph/messenger_v2.c b/net/ceph/messenger_v2.c index 9e39378eda00..9e48623018a3 100644 --- a/net/ceph/messenger_v2.c +++ b/net/ceph/messenger_v2.c @@ -1061,13 +1061,16 @@ static int decrypt_control_remainder(struct ceph_connection *con) static int process_v2_sparse_read(struct ceph_connection *con, struct page **pages, int spos) { - struct ceph_msg_data_cursor *cursor = &con->v2.in_cursor; + struct ceph_msg_data_cursor cursor; int ret; + ceph_msg_data_cursor_init(&cursor, con->in_msg, + con->in_msg->sparse_read_total); + for (;;) { char *buf = NULL; - ret = con->ops->sparse_read(con, cursor, &buf); + ret = con->ops->sparse_read(con, &cursor, &buf); if (ret <= 0) return ret; @@ -1085,11 +1088,11 @@ static int process_v2_sparse_read(struct ceph_connection *con, } else { struct bio_vec bv; - get_bvec_at(cursor, &bv); + get_bvec_at(&cursor, &bv); len = min_t(int, len, bv.bv_len); memcpy_page(bv.bv_page, bv.bv_offset, spage, soff, len); - ceph_msg_data_advance(cursor, len); + ceph_msg_data_advance(&cursor, len); } spos += len; ret -= len; From ec3797f043756a94ea2d0f106022e14ac4946c02 Mon Sep 17 00:00:00 2001 From: ziming zhang Date: Mon, 17 Nov 2025 18:07:41 +0800 Subject: [PATCH 4/5] libceph: replace BUG_ON with bounds check for map->max_osd OSD indexes come from untrusted network packets. Boundary checks are added to validate these against map->max_osd. [ idryomov: drop BUG_ON in ceph_get_primary_affinity(), minor cosmetic edits ] Cc: stable@vger.kernel.org Signed-off-by: ziming zhang Reviewed-by: Ilya Dryomov Signed-off-by: Ilya Dryomov --- net/ceph/osdmap.c | 18 +++++++++++------- 1 file changed, 11 insertions(+), 7 deletions(-) diff --git a/net/ceph/osdmap.c b/net/ceph/osdmap.c index 295098873861..d245fa508e1c 100644 --- a/net/ceph/osdmap.c +++ b/net/ceph/osdmap.c @@ -1504,8 +1504,6 @@ static int decode_new_primary_temp(void **p, void *end, u32 ceph_get_primary_affinity(struct ceph_osdmap *map, int osd) { - BUG_ON(osd >= map->max_osd); - if (!map->osd_primary_affinity) return CEPH_OSD_DEFAULT_PRIMARY_AFFINITY; @@ -1514,8 +1512,6 @@ u32 ceph_get_primary_affinity(struct ceph_osdmap *map, int osd) static int set_primary_affinity(struct ceph_osdmap *map, int osd, u32 aff) { - BUG_ON(osd >= map->max_osd); - if (!map->osd_primary_affinity) { int i; @@ -1577,6 +1573,8 @@ static int decode_new_primary_affinity(void **p, void *end, ceph_decode_32_safe(p, end, osd, e_inval); ceph_decode_32_safe(p, end, aff, e_inval); + if (osd >= map->max_osd) + goto e_inval; ret = set_primary_affinity(map, osd, aff); if (ret) @@ -1879,7 +1877,9 @@ static int decode_new_up_state_weight(void **p, void *end, u8 struct_v, ceph_decode_need(p, end, 2*sizeof(u32), e_inval); osd = ceph_decode_32(p); w = ceph_decode_32(p); - BUG_ON(osd >= map->max_osd); + if (osd >= map->max_osd) + goto e_inval; + osdmap_info(map, "osd%d weight 0x%x %s\n", osd, w, w == CEPH_OSD_IN ? "(in)" : (w == CEPH_OSD_OUT ? "(out)" : "")); @@ -1905,13 +1905,15 @@ static int decode_new_up_state_weight(void **p, void *end, u8 struct_v, u32 xorstate; osd = ceph_decode_32(p); + if (osd >= map->max_osd) + goto e_inval; + if (struct_v >= 5) xorstate = ceph_decode_32(p); else xorstate = ceph_decode_8(p); if (xorstate == 0) xorstate = CEPH_OSD_UP; - BUG_ON(osd >= map->max_osd); if ((map->osd_state[osd] & CEPH_OSD_UP) && (xorstate & CEPH_OSD_UP)) osdmap_info(map, "osd%d down\n", osd); @@ -1937,7 +1939,9 @@ static int decode_new_up_state_weight(void **p, void *end, u8 struct_v, struct ceph_entity_addr addr; osd = ceph_decode_32(p); - BUG_ON(osd >= map->max_osd); + if (osd >= map->max_osd) + goto e_inval; + if (struct_v >= 7) ret = ceph_decode_entity_addrvec(p, end, msgr2, &addr); else From 7fce830ecd0a0256590ee37eb65a39cbad3d64fc Mon Sep 17 00:00:00 2001 From: ziming zhang Date: Fri, 14 Nov 2025 16:56:10 +0800 Subject: [PATCH 5/5] libceph: prevent potential out-of-bounds writes in handle_auth_session_key() The len field originates from untrusted network packets. Boundary checks have been added to prevent potential out-of-bounds writes when decrypting the connection secret or processing service tickets. [ idryomov: changelog ] Cc: stable@vger.kernel.org Signed-off-by: ziming zhang Reviewed-by: Ilya Dryomov Signed-off-by: Ilya Dryomov --- net/ceph/auth_x.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/net/ceph/auth_x.c b/net/ceph/auth_x.c index b71b1635916e..a21c157daf7d 100644 --- a/net/ceph/auth_x.c +++ b/net/ceph/auth_x.c @@ -631,6 +631,7 @@ static int handle_auth_session_key(struct ceph_auth_client *ac, u64 global_id, /* connection secret */ ceph_decode_32_safe(p, end, len, e_inval); + ceph_decode_need(p, end, len, e_inval); dout("%s connection secret blob len %d\n", __func__, len); if (len > 0) { dp = *p + ceph_x_encrypt_offset(); @@ -648,6 +649,7 @@ static int handle_auth_session_key(struct ceph_auth_client *ac, u64 global_id, /* service tickets */ ceph_decode_32_safe(p, end, len, e_inval); + ceph_decode_need(p, end, len, e_inval); dout("%s service tickets blob len %d\n", __func__, len); if (len > 0) { ret = ceph_x_proc_ticket_reply(ac, &th->session_key,