kernel/linux.git/fs/proc, branch v6.1.168

procfs: fix missing RCU protection when reading real_parent in do_task_stat()

2026-03-04T12:19:44+00:00

[ Upstream commit 76149d53502cf17ef3ae454ff384551236fba867 ] When reading /proc/[pid]/stat, do_task_stat() accesses task->real_parent without proper RCU protection, which leads to: cpu 0 cpu 1 ----- ----- do_task_stat var = task->real_parent release_task call_rcu(delayed_put_task_struct) task_tgid_nr_ns(var) rcu_read_lock <--- Too late to protect task->real_parent! task_pid_ptr <--- UAF! rcu_read_unlock This patch uses task_ppid_nr_ns() instead of task_tgid_nr_ns() to add proper RCU protection for accessing task->real_parent. Link: https://lkml.kernel.org/r/20260128083007.3173016-1-alexjlzheng@tencent.com Fixes: 06fffb1267c9 ("do_task_stat: don't take rcu_read_lock()") Signed-off-by: Jinliang Zheng Acked-by: Oleg Nesterov Cc: David Hildenbrand Cc: Ingo Molnar Cc: Lorenzo Stoakes Cc: Mateusz Guzik Cc: ruippan Cc: Usama Arif Signed-off-by: Andrew Morton Signed-off-by: Sasha Levin

fs/proc: fix uaf in proc_readdir_de()

2025-12-06T21:12:35+00:00

commit 895b4c0c79b092d732544011c3cecaf7322c36a1 upstream. Pde is erased from subdir rbtree through rb_erase(), but not set the node to EMPTY, which may result in uaf access. We should use RB_CLEAR_NODE() set the erased node to EMPTY, then pde_subdir_next() will return NULL to avoid uaf access. We found an uaf issue while using stress-ng testing, need to run testcase getdent and tun in the same time. The steps of the issue is as follows: 1) use getdent to traverse dir /proc/pid/net/dev_snmp6/, and current pde is tun3; 2) in the [time windows] unregister netdevice tun3 and tun2, and erase them from rbtree. erase tun3 first, and then erase tun2. the pde(tun2) will be released to slab; 3) continue to getdent process, then pde_subdir_next() will return pde(tun2) which is released, it will case uaf access. CPU 0 | CPU 1 ------------------------------------------------------------------------- traverse dir /proc/pid/net/dev_snmp6/ | unregister_netdevice(tun->dev) //tun3 tun2 sys_getdents64() | iterate_dir() | proc_readdir() | proc_readdir_de() | snmp6_unregister_dev() pde_get(de); | proc_remove() read_unlock(&proc_subdir_lock); | remove_proc_subtree() | write_lock(&proc_subdir_lock); [time window] | rb_erase(&root->subdir_node, &parent->subdir); | write_unlock(&proc_subdir_lock); read_lock(&proc_subdir_lock); | next = pde_subdir_next(de); | pde_put(de); | de = next; //UAF | rbtree of dev_snmp6 | pde(tun3) / \ NULL pde(tun2) Link: https://lkml.kernel.org/r/20251025024233.158363-1-albin_yang@163.com Signed-off-by: Wei Yang Cc: Al Viro Cc: Christian Brauner Cc: wangzijie Cc: Alexey Dobriyan Cc: Signed-off-by: Andrew Morton Signed-off-by: Greg Kroah-Hartman

proc: fix type confusion in pde_set_flags()

2025-09-19T14:29:56+00:00

[ Upstream commit 0ce9398aa0830f15f92bbed73853f9861c3e74ff ] Commit 2ce3d282bd50 ("proc: fix missing pde_set_flags() for net proc files") missed a key part in the definition of proc_dir_entry: union { const struct proc_ops *proc_ops; const struct file_operations *proc_dir_ops; }; So dereference of ->proc_ops assumes it is a proc_ops structure results in type confusion and make NULL check for 'proc_ops' not work for proc dir. Add !S_ISDIR(dp->mode) test before calling pde_set_flags() to fix it. Link: https://lkml.kernel.org/r/20250904135715.3972782-1-wangzijie1@honor.com Fixes: 2ce3d282bd50 ("proc: fix missing pde_set_flags() for net proc files") Signed-off-by: wangzijie Reported-by: Brad Spengler Closes: https://lore.kernel.org/all/20250903065758.3678537-1-wangzijie1@honor.com/ Cc: Alexey Dobriyan Cc: Al Viro Cc: Christian Brauner Cc: Jiri Slaby Cc: Stefano Brivio Cc: Signed-off-by: Andrew Morton Signed-off-by: Sasha Levin

proc: fix missing pde_set_flags() for net proc files

2025-09-09T16:54:17+00:00

commit 2ce3d282bd5050fca8577defeff08ada0d55d062 upstream. To avoid potential UAF issues during module removal races, we use pde_set_flags() to save proc_ops flags in PDE itself before proc_register(), and then use pde_has_proc_*() helpers instead of directly dereferencing pde->proc_ops->*. However, the pde_set_flags() call was missing when creating net related proc files. This omission caused incorrect behavior which FMODE_LSEEK was being cleared inappropriately in proc_reg_open() for net proc files. Lars reported it in this link[1]. Fix this by ensuring pde_set_flags() is called when register proc entry, and add NULL check for proc_ops in pde_set_flags(). [wangzijie1@honor.com: stash pde->proc_ops in a local const variable, per Christian] Link: https://lkml.kernel.org/r/20250821105806.1453833-1-wangzijie1@honor.com Link: https://lkml.kernel.org/r/20250818123102.959595-1-wangzijie1@honor.com Link: https://lore.kernel.org/all/20250815195616.64497967@chagall.paradoxon.rec/ [1] Fixes: ff7ec8dc1b64 ("proc: use the same treatment to check proc_lseek as ones for proc_read_iter et.al") Signed-off-by: wangzijie Reported-by: Lars Wendler Tested-by: Stefano Brivio Tested-by: Petr Vaněk Tested by: Lars Wendler Cc: Alexei Starovoitov Cc: Alexey Dobriyan Cc: Al Viro Cc: "Edgecombe, Rick P" Cc: Greg Kroah-Hartman Cc: Jiri Slaby Cc: Kirill A. Shutemov Cc: wangzijie Cc: Signed-off-by: Andrew Morton Signed-off-by: Greg Kroah-Hartman

proc: use the same treatment to check proc_lseek as ones for proc_read_iter et.al

2025-08-15T10:05:03+00:00

[ Upstream commit ff7ec8dc1b646296f8d94c39339e8d3833d16c05 ] Check pde->proc_ops->proc_lseek directly may cause UAF in rmmod scenario. It's a gap in proc_reg_open() after commit 654b33ada4ab("proc: fix UAF in proc_get_inode()"). Followed by AI Viro's suggestion, fix it in same manner. Link: https://lkml.kernel.org/r/20250607021353.1127963-1-wangzijie1@honor.com Fixes: 3f61631d47f1 ("take care to handle NULL ->proc_lseek()") Signed-off-by: wangzijie Reviewed-by: Alexey Dobriyan Cc: Alexei Starovoitov Cc: Al Viro Cc: "Edgecombe, Rick P" Cc: Kirill A. Shuemov Signed-off-by: Andrew Morton Signed-off-by: Sasha Levin

fix proc_sys_compare() handling of in-lookup dentries

2025-07-17T16:32:05+00:00

[ Upstream commit b969f9614885c20f903e1d1f9445611daf161d6d ] There's one case where ->d_compare() can be called for an in-lookup dentry; usually that's nothing special from ->d_compare() point of view, but... proc_sys_compare() is weird. The thing is, /proc/sys subdirectories can look differently for different processes. Up to and including having the same name resolve to different dentries - all of them hashed. The way it's done is ->d_compare() refusing to admit a match unless this dentry is supposed to be visible to this caller. The information needed to discriminate between them is stored in inode; it is set during proc_sys_lookup() and until it's done d_splice_alias() we really can't tell who should that dentry be visible for. Normally there's no negative dentries in /proc/sys; we can run into a dying dentry in RCU dcache lookup, but those can be safely rejected. However, ->d_compare() is also called for in-lookup dentries, before they get positive - or hashed, for that matter. In case of match we will wait until dentry leaves in-lookup state and repeat ->d_compare() afterwards. In other words, the right behaviour is to treat the name match as sufficient for in-lookup dentries; if dentry is not for us, we'll see that when we recheck once proc_sys_lookup() is done with it. While we are at it, fix the misspelled READ_ONCE and WRITE_ONCE there. Fixes: d9171b934526 ("parallel lookups machinery, part 4 (and last)") Reported-by: NeilBrown Reviewed-by: Christian Brauner Reviewed-by: NeilBrown Signed-off-by: Al Viro Signed-off-by: Sasha Levin

fs/procfs: fix the comment above proc_pid_wchan()

2025-04-10T12:33:37+00:00

[ Upstream commit 6287fbad1cd91f0c25cdc3a580499060828a8f30 ] proc_pid_wchan() used to report kernel addresses to user space but that is no longer the case today. Bring the comment above proc_pid_wchan() in sync with the implementation. Link: https://lkml.kernel.org/r/20250319210222.1518771-1-bvanassche@acm.org Fixes: b2f73922d119 ("fs/proc, core/debug: Don't expose absolute kernel addresses via wchan") Signed-off-by: Bart Van Assche Cc: Kees Cook Cc: Eric W. Biederman Cc: Alexey Dobriyan Signed-off-by: Andrew Morton Signed-off-by: Sasha Levin

proc: fix UAF in proc_get_inode()

2025-03-28T20:59:01+00:00

commit 654b33ada4ab5e926cd9c570196fefa7bec7c1df upstream. Fix race between rmmod and /proc/XXX's inode instantiation. The bug is that pde->proc_ops don't belong to /proc, it belongs to a module, therefore dereferencing it after /proc entry has been registered is a bug unless use_pde/unuse_pde() pair has been used. use_pde/unuse_pde can be avoided (2 atomic ops!) because pde->proc_ops never changes so information necessary for inode instantiation can be saved _before_ proc_register() in PDE itself and used later, avoiding pde->proc_ops->... dereference. rmmod lookup sys_delete_module proc_lookup_de pde_get(de); proc_get_inode(dir->i_sb, de); mod->exit() proc_remove remove_proc_subtree proc_entry_rundown(de); free_module(mod); if (S_ISREG(inode->i_mode)) if (de->proc_ops->proc_read_iter) --> As module is already freed, will trigger UAF BUG: unable to handle page fault for address: fffffbfff80a702b PGD 817fc4067 P4D 817fc4067 PUD 817fc0067 PMD 102ef4067 PTE 0 Oops: Oops: 0000 [#1] PREEMPT SMP KASAN PTI CPU: 26 UID: 0 PID: 2667 Comm: ls Tainted: G Hardware name: QEMU Standard PC (i440FX + PIIX, 1996) RIP: 0010:proc_get_inode+0x302/0x6e0 RSP: 0018:ffff88811c837998 EFLAGS: 00010a06 RAX: dffffc0000000000 RBX: ffffffffc0538140 RCX: 0000000000000007 RDX: 1ffffffff80a702b RSI: 0000000000000001 RDI: ffffffffc0538158 RBP: ffff8881299a6000 R08: 0000000067bbe1e5 R09: 1ffff11023906f20 R10: ffffffffb560ca07 R11: ffffffffb2b43a58 R12: ffff888105bb78f0 R13: ffff888100518048 R14: ffff8881299a6004 R15: 0000000000000001 FS: 00007f95b9686840(0000) GS:ffff8883af100000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: fffffbfff80a702b CR3: 0000000117dd2000 CR4: 00000000000006f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: proc_lookup_de+0x11f/0x2e0 __lookup_slow+0x188/0x350 walk_component+0x2ab/0x4f0 path_lookupat+0x120/0x660 filename_lookup+0x1ce/0x560 vfs_statx+0xac/0x150 __do_sys_newstat+0x96/0x110 do_syscall_64+0x5f/0x170 entry_SYSCALL_64_after_hwframe+0x76/0x7e [adobriyan@gmail.com: don't do 2 atomic ops on the common path] Link: https://lkml.kernel.org/r/3d25ded0-1739-447e-812b-e34da7990dcf@p183 Fixes: 778f3dd5a13c ("Fix procfs compat_ioctl regression") Signed-off-by: Ye Bin Signed-off-by: Alexey Dobriyan Cc: Al Viro Cc: David S. Miller Cc: Signed-off-by: Andrew Morton Signed-off-by: Greg Kroah-Hartman

hrtimer: Use and report correct timerslack values for realtime tasks

2025-03-28T20:58:48+00:00

commit ed4fb6d7ef68111bb539283561953e5c6e9a6e38 upstream. The timerslack_ns setting is used to specify how much the hardware timers should be delayed, to potentially dispatch multiple timers in a single interrupt. This is a performance optimization. Timers of realtime tasks (having a realtime scheduling policy) should not be delayed. This logic was inconsitently applied to the hrtimers, leading to delays of realtime tasks which used timed waits for events (e.g. condition variables). Due to the downstream override of the slack for rt tasks, the procfs reported incorrect (non-zero) timerslack_ns values. This is changed by setting the timer_slack_ns task attribute to 0 for all tasks with a rt policy. By that, downstream users do not need to specially handle rt tasks (w.r.t. the slack), and the procfs entry shows the correct value of "0". Setting non-zero slack values (either via procfs or PR_SET_TIMERSLACK) on tasks with a rt policy is ignored, as stated in "man 2 PR_SET_TIMERSLACK": Timer slack is not applied to threads that are scheduled under a real-time scheduling policy (see sched_setscheduler(2)). The special handling of timerslack on rt tasks in downstream users is removed as well. Signed-off-by: Felix Moessbauer Signed-off-by: Thomas Gleixner Link: https://lore.kernel.org/all/20240814121032.368444-2-felix.moessbauer@siemens.com Signed-off-by: Greg Kroah-Hartman

fs/proc: do_task_stat: Fix ESP not readable during coredump

2025-02-21T12:49:39+00:00

commit ab251dacfbae28772c897f068a4184f478189ff2 upstream. The field "eip" (instruction pointer) and "esp" (stack pointer) of a task can be read from /proc/PID/stat. These fields can be interesting for coredump. However, these fields were disabled by commit 0a1eb2d474ed ("fs/proc: Stop reporting eip and esp in /proc/PID/stat"), because it is generally unsafe to do so. But it is safe for a coredumping process, and therefore exceptions were made: - for a coredumping thread by commit fd7d56270b52 ("fs/proc: Report eip/esp in /prod/PID/stat for coredumping"). - for all other threads in a coredumping process by commit cb8f381f1613 ("fs/proc/array.c: allow reporting eip/esp for all coredumping threads"). The above two commits check the PF_DUMPCORE flag to determine a coredump thread and the PF_EXITING flag for the other threads. Unfortunately, commit 92307383082d ("coredump: Don't perform any cleanups before dumping core") moved coredump to happen earlier and before PF_EXITING is set. Thus, checking PF_EXITING is no longer the correct way to determine threads in a coredumping process. Instead of PF_EXITING, use PF_POSTCOREDUMP to determine the other threads. Checking of PF_EXITING was added for coredumping, so it probably can now be removed. But it doesn't hurt to keep. Fixes: 92307383082d ("coredump: Don't perform any cleanups before dumping core") Cc: stable@vger.kernel.org Cc: Eric W. Biederman Acked-by: Oleg Nesterov Acked-by: Kees Cook Signed-off-by: Nam Cao Link: https://lore.kernel.org/r/d89af63d478d6c64cc46a01420b46fd6eb147d6f.1735805772.git.namcao@linutronix.de Signed-off-by: Christian Brauner Signed-off-by: Greg Kroah-Hartman