From 0cce06ba859a515bd06224085d3addb870608b6d Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 25 Apr 2023 17:03:13 +0200 Subject: debugobjects,locking: Annotate debug_object_fill_pool() wait type violation There is an explicit wait-type violation in debug_object_fill_pool() for PREEMPT_RT=n kernels which allows them to more easily fill the object pool and reduce the chance of allocation failures. Lockdep's wait-type checks are designed to check the PREEMPT_RT locking rules even for PREEMPT_RT=n kernels and object to this, so create a lockdep annotation to allow this to stand. Specifically, create a 'lock' type that overrides the inner wait-type while it is held -- allowing one to temporarily raise it, such that the violation is hidden. Reported-by: Vlastimil Babka Reported-by: Qi Zheng Signed-off-by: Peter Zijlstra (Intel) Tested-by: Qi Zheng Link: https://lkml.kernel.org/r/20230429100614.GA1489784@hirez.programming.kicks-ass.net --- kernel/locking/lockdep.c | 28 +++++++++++++++++++++------- 1 file changed, 21 insertions(+), 7 deletions(-) (limited to 'kernel') diff --git a/kernel/locking/lockdep.c b/kernel/locking/lockdep.c index 50d4863974e7..62ef295e07e6 100644 --- a/kernel/locking/lockdep.c +++ b/kernel/locking/lockdep.c @@ -2253,6 +2253,9 @@ static inline bool usage_match(struct lock_list *entry, void *mask) static inline bool usage_skip(struct lock_list *entry, void *mask) { + if (entry->class->lock_type == LD_LOCK_NORMAL) + return false; + /* * Skip local_lock() for irq inversion detection. * @@ -2279,14 +2282,16 @@ static inline bool usage_skip(struct lock_list *entry, void *mask) * As a result, we will skip local_lock(), when we search for irq * inversion bugs. */ - if (entry->class->lock_type == LD_LOCK_PERCPU) { - if (DEBUG_LOCKS_WARN_ON(entry->class->wait_type_inner < LD_WAIT_CONFIG)) - return false; + if (entry->class->lock_type == LD_LOCK_PERCPU && + DEBUG_LOCKS_WARN_ON(entry->class->wait_type_inner < LD_WAIT_CONFIG)) + return false; - return true; - } + /* + * Skip WAIT_OVERRIDE for irq inversion detection -- it's not actually + * a lock and only used to override the wait_type. + */ - return false; + return true; } /* @@ -4752,7 +4757,8 @@ static int check_wait_context(struct task_struct *curr, struct held_lock *next) for (; depth < curr->lockdep_depth; depth++) { struct held_lock *prev = curr->held_locks + depth; - u8 prev_inner = hlock_class(prev)->wait_type_inner; + struct lock_class *class = hlock_class(prev); + u8 prev_inner = class->wait_type_inner; if (prev_inner) { /* @@ -4762,6 +4768,14 @@ static int check_wait_context(struct task_struct *curr, struct held_lock *next) * Also due to trylocks. */ curr_inner = min(curr_inner, prev_inner); + + /* + * Allow override for annotations -- this is typically + * only valid/needed for code that only exists when + * CONFIG_PREEMPT_RT=n. + */ + if (unlikely(class->lock_type == LD_LOCK_WAIT_OVERRIDE)) + curr_inner = prev_inner; } } -- cgit v1.2.3 From e1505c1cc8d527fcc5bcaf9c1ad82eed817e3e10 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Fri, 5 May 2023 14:58:36 -0700 Subject: bpf: netdev: init the offload table earlier Some netdevices may get unregistered before late_initcall(), we have to move the hashtable init earlier. Fixes: f1fc43d03946 ("bpf: Move offload initialization into late_initcall") Closes: https://bugzilla.kernel.org/show_bug.cgi?id=217399 Signed-off-by: Jakub Kicinski Acked-by: Stanislav Fomichev Link: https://lore.kernel.org/r/20230505215836.491485-1-kuba@kernel.org Signed-off-by: Alexei Starovoitov --- kernel/bpf/offload.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/bpf/offload.c b/kernel/bpf/offload.c index d9c9f45e3529..8a26cd8814c1 100644 --- a/kernel/bpf/offload.c +++ b/kernel/bpf/offload.c @@ -859,4 +859,4 @@ static int __init bpf_offload_init(void) return rhashtable_init(&offdevs, &offdevs_params); } -late_initcall(bpf_offload_init); +core_initcall(bpf_offload_init); -- cgit v1.2.3 From 6049674b5720edbbb13a7cb3e2f3d2affaa40c19 Mon Sep 17 00:00:00 2001 From: "Masami Hiramatsu (Google)" Date: Sun, 9 Apr 2023 11:28:31 +0900 Subject: tracing: fprobe: Initialize ret valiable to fix smatch error The commit 39d954200bf6 ("fprobe: Skip exit_handler if entry_handler returns !0") introduced a hidden dependency of 'ret' local variable in the fprobe_handler(), Smatch warns the `ret` can be accessed without initialization. kernel/trace/fprobe.c:59 fprobe_handler() error: uninitialized symbol 'ret'. kernel/trace/fprobe.c 49 fpr->entry_ip = ip; 50 if (fp->entry_data_size) 51 entry_data = fpr->data; 52 } 53 54 if (fp->entry_handler) 55 ret = fp->entry_handler(fp, ip, ftrace_get_regs(fregs), entry_data); ret is only initialized if there is an ->entry_handler 56 57 /* If entry_handler returns !0, nmissed is not counted. */ 58 if (rh) { rh is only true if there is an ->exit_handler. Presumably if you have and ->exit_handler that means you also have a ->entry_handler but Smatch is not smart enough to figure it out. --> 59 if (ret) ^^^ Warning here. 60 rethook_recycle(rh); 61 else 62 rethook_hook(rh, ftrace_get_regs(fregs), true); 63 } 64 out: 65 ftrace_test_recursion_unlock(bit); 66 } Link: https://lore.kernel.org/all/168100731160.79534.374827110083836722.stgit@devnote2/ Reported-by: Dan Carpenter Link: https://lore.kernel.org/all/85429a5c-a4b9-499e-b6c0-cbd313291c49@kili.mountain Fixes: 39d954200bf6 ("fprobe: Skip exit_handler if entry_handler returns !0") Acked-by: Dan Carpenter Signed-off-by: Masami Hiramatsu (Google) --- kernel/trace/fprobe.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/trace/fprobe.c b/kernel/trace/fprobe.c index 9abb3905bc8e..293184227394 100644 --- a/kernel/trace/fprobe.c +++ b/kernel/trace/fprobe.c @@ -27,7 +27,7 @@ static void fprobe_handler(unsigned long ip, unsigned long parent_ip, struct rethook_node *rh = NULL; struct fprobe *fp; void *entry_data = NULL; - int bit, ret; + int bit, ret = 0; fp = container_of(ops, struct fprobe, ops); if (fprobe_disabled(fp)) -- cgit v1.2.3 From be243bacfb25f5219f2396d787408e8cf1301dd1 Mon Sep 17 00:00:00 2001 From: Ze Gao Date: Wed, 17 May 2023 11:45:06 +0800 Subject: rethook: use preempt_{disable, enable}_notrace in rethook_trampoline_handler This patch replaces preempt_{disable, enable} with its corresponding notrace version in rethook_trampoline_handler so no worries about stack recursion or overflow introduced by preempt_count_{add, sub} under fprobe + rethook context. Link: https://lore.kernel.org/all/20230517034510.15639-2-zegao@tencent.com/ Fixes: 54ecbe6f1ed5 ("rethook: Add a generic return hook") Signed-off-by: Ze Gao Acked-by: Masami Hiramatsu (Google) Cc: Signed-off-by: Masami Hiramatsu (Google) --- kernel/trace/rethook.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/rethook.c b/kernel/trace/rethook.c index 32c3dfdb4d6a..60f6cb2b486b 100644 --- a/kernel/trace/rethook.c +++ b/kernel/trace/rethook.c @@ -288,7 +288,7 @@ unsigned long rethook_trampoline_handler(struct pt_regs *regs, * These loops must be protected from rethook_free_rcu() because those * are accessing 'rhn->rethook'. */ - preempt_disable(); + preempt_disable_notrace(); /* * Run the handler on the shadow stack. Do not unlink the list here because @@ -321,7 +321,7 @@ unsigned long rethook_trampoline_handler(struct pt_regs *regs, first = first->next; rethook_recycle(rhn); } - preempt_enable(); + preempt_enable_notrace(); return correct_ret_addr; } -- cgit v1.2.3 From 3cc4e2c5fbae84e5033723fb7e350bc6c164e3a2 Mon Sep 17 00:00:00 2001 From: Ze Gao Date: Wed, 17 May 2023 11:45:07 +0800 Subject: fprobe: make fprobe_kprobe_handler recursion free Current implementation calls kprobe related functions before doing ftrace recursion check in fprobe_kprobe_handler, which opens door to kernel crash due to stack recursion if preempt_count_{add, sub} is traceable in kprobe_busy_{begin, end}. Things goes like this without this patch quoted from Steven: " fprobe_kprobe_handler() { kprobe_busy_begin() { preempt_disable() { preempt_count_add() { <-- trace fprobe_kprobe_handler() { [ wash, rinse, repeat, CRASH!!! ] " By refactoring the common part out of fprobe_kprobe_handler and fprobe_handler and call ftrace recursion detection at the very beginning, the whole fprobe_kprobe_handler is free from recursion. [ Fix the indentation of __fprobe_handler() parameters. ] Link: https://lore.kernel.org/all/20230517034510.15639-3-zegao@tencent.com/ Fixes: ab51e15d535e ("fprobe: Introduce FPROBE_FL_KPROBE_SHARED flag for fprobe") Signed-off-by: Ze Gao Acked-by: Masami Hiramatsu (Google) Cc: stable@vger.kernel.org Signed-off-by: Masami Hiramatsu (Google) --- kernel/trace/fprobe.c | 59 ++++++++++++++++++++++++++++++++++++++------------- 1 file changed, 44 insertions(+), 15 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/fprobe.c b/kernel/trace/fprobe.c index 293184227394..7a692c02f787 100644 --- a/kernel/trace/fprobe.c +++ b/kernel/trace/fprobe.c @@ -20,30 +20,22 @@ struct fprobe_rethook_node { char data[]; }; -static void fprobe_handler(unsigned long ip, unsigned long parent_ip, - struct ftrace_ops *ops, struct ftrace_regs *fregs) +static inline void __fprobe_handler(unsigned long ip, unsigned long parent_ip, + struct ftrace_ops *ops, struct ftrace_regs *fregs) { struct fprobe_rethook_node *fpr; struct rethook_node *rh = NULL; struct fprobe *fp; void *entry_data = NULL; - int bit, ret = 0; + int ret = 0; fp = container_of(ops, struct fprobe, ops); - if (fprobe_disabled(fp)) - return; - - bit = ftrace_test_recursion_trylock(ip, parent_ip); - if (bit < 0) { - fp->nmissed++; - return; - } if (fp->exit_handler) { rh = rethook_try_get(fp->rethook); if (!rh) { fp->nmissed++; - goto out; + return; } fpr = container_of(rh, struct fprobe_rethook_node, node); fpr->entry_ip = ip; @@ -61,23 +53,60 @@ static void fprobe_handler(unsigned long ip, unsigned long parent_ip, else rethook_hook(rh, ftrace_get_regs(fregs), true); } -out: +} + +static void fprobe_handler(unsigned long ip, unsigned long parent_ip, + struct ftrace_ops *ops, struct ftrace_regs *fregs) +{ + struct fprobe *fp; + int bit; + + fp = container_of(ops, struct fprobe, ops); + if (fprobe_disabled(fp)) + return; + + /* recursion detection has to go before any traceable function and + * all functions before this point should be marked as notrace + */ + bit = ftrace_test_recursion_trylock(ip, parent_ip); + if (bit < 0) { + fp->nmissed++; + return; + } + __fprobe_handler(ip, parent_ip, ops, fregs); ftrace_test_recursion_unlock(bit); + } NOKPROBE_SYMBOL(fprobe_handler); static void fprobe_kprobe_handler(unsigned long ip, unsigned long parent_ip, struct ftrace_ops *ops, struct ftrace_regs *fregs) { - struct fprobe *fp = container_of(ops, struct fprobe, ops); + struct fprobe *fp; + int bit; + + fp = container_of(ops, struct fprobe, ops); + if (fprobe_disabled(fp)) + return; + + /* recursion detection has to go before any traceable function and + * all functions called before this point should be marked as notrace + */ + bit = ftrace_test_recursion_trylock(ip, parent_ip); + if (bit < 0) { + fp->nmissed++; + return; + } if (unlikely(kprobe_running())) { fp->nmissed++; return; } + kprobe_busy_begin(); - fprobe_handler(ip, parent_ip, ops, fregs); + __fprobe_handler(ip, parent_ip, ops, fregs); kprobe_busy_end(); + ftrace_test_recursion_unlock(bit); } static void fprobe_exit_handler(struct rethook_node *rh, void *data, -- cgit v1.2.3 From 2752741080f84f9b2fc93fa92735315d10a415bf Mon Sep 17 00:00:00 2001 From: Ze Gao Date: Wed, 17 May 2023 11:45:08 +0800 Subject: fprobe: add recursion detection in fprobe_exit_handler fprobe_hander and fprobe_kprobe_handler has guarded ftrace recursion detection but fprobe_exit_handler has not, which possibly introduce recursive calls if the fprobe exit callback calls any traceable functions. Checking in fprobe_hander or fprobe_kprobe_handler is not enough and misses this case. So add recursion free guard the same way as fprobe_hander. Since ftrace recursion check does not employ ip(s), so here use entry_ip and entry_parent_ip the same as fprobe_handler. Link: https://lore.kernel.org/all/20230517034510.15639-4-zegao@tencent.com/ Fixes: 5b0ab78998e3 ("fprobe: Add exit_handler support") Signed-off-by: Ze Gao Cc: stable@vger.kernel.org Acked-by: Masami Hiramatsu (Google) Signed-off-by: Masami Hiramatsu (Google) --- kernel/trace/fprobe.c | 14 ++++++++++++++ 1 file changed, 14 insertions(+) (limited to 'kernel') diff --git a/kernel/trace/fprobe.c b/kernel/trace/fprobe.c index 7a692c02f787..18d36842faf5 100644 --- a/kernel/trace/fprobe.c +++ b/kernel/trace/fprobe.c @@ -17,6 +17,7 @@ struct fprobe_rethook_node { struct rethook_node node; unsigned long entry_ip; + unsigned long entry_parent_ip; char data[]; }; @@ -39,6 +40,7 @@ static inline void __fprobe_handler(unsigned long ip, unsigned long parent_ip, } fpr = container_of(rh, struct fprobe_rethook_node, node); fpr->entry_ip = ip; + fpr->entry_parent_ip = parent_ip; if (fp->entry_data_size) entry_data = fpr->data; } @@ -114,14 +116,26 @@ static void fprobe_exit_handler(struct rethook_node *rh, void *data, { struct fprobe *fp = (struct fprobe *)data; struct fprobe_rethook_node *fpr; + int bit; if (!fp || fprobe_disabled(fp)) return; fpr = container_of(rh, struct fprobe_rethook_node, node); + /* + * we need to assure no calls to traceable functions in-between the + * end of fprobe_handler and the beginning of fprobe_exit_handler. + */ + bit = ftrace_test_recursion_trylock(fpr->entry_ip, fpr->entry_parent_ip); + if (bit < 0) { + fp->nmissed++; + return; + } + fp->exit_handler(fp, fpr->entry_ip, regs, fp->entry_data_size ? (void *)fpr->data : NULL); + ftrace_test_recursion_unlock(bit); } NOKPROBE_SYMBOL(fprobe_exit_handler); -- cgit v1.2.3 From 0613d8ca9ab382caabe9ed2dceb429e9781e443f Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Thu, 18 May 2023 11:25:28 +0100 Subject: bpf: Fix mask generation for 32-bit narrow loads of 64-bit fields A narrow load from a 64-bit context field results in a 64-bit load followed potentially by a 64-bit right-shift and then a bitwise AND operation to extract the relevant data. In the case of a 32-bit access, an immediate mask of 0xffffffff is used to construct a 64-bit BPP_AND operation which then sign-extends the mask value and effectively acts as a glorified no-op. For example: 0: 61 10 00 00 00 00 00 00 r0 = *(u32 *)(r1 + 0) results in the following code generation for a 64-bit field: ldr x7, [x7] // 64-bit load mov x10, #0xffffffffffffffff and x7, x7, x10 Fix the mask generation so that narrow loads always perform a 32-bit AND operation: ldr x7, [x7] // 64-bit load mov w10, #0xffffffff and w7, w7, w10 Cc: Alexei Starovoitov Cc: Daniel Borkmann Cc: John Fastabend Cc: Krzesimir Nowak Cc: Andrey Ignatov Acked-by: Yonghong Song Fixes: 31fd85816dbe ("bpf: permits narrower load from bpf program context fields") Signed-off-by: Will Deacon Link: https://lore.kernel.org/r/20230518102528.1341-1-will@kernel.org Signed-off-by: Alexei Starovoitov --- kernel/bpf/verifier.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index fbcf5a4e2fcd..5871aa78d01a 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -17033,7 +17033,7 @@ static int convert_ctx_accesses(struct bpf_verifier_env *env) insn_buf[cnt++] = BPF_ALU64_IMM(BPF_RSH, insn->dst_reg, shift); - insn_buf[cnt++] = BPF_ALU64_IMM(BPF_AND, insn->dst_reg, + insn_buf[cnt++] = BPF_ALU32_IMM(BPF_AND, insn->dst_reg, (1ULL << size * 8) - 1); } } -- cgit v1.2.3 From b34ffb0c6d23583830f9327864b9c1f486003305 Mon Sep 17 00:00:00 2001 From: Anton Protopopov Date: Mon, 22 May 2023 15:45:58 +0000 Subject: bpf: fix a memory leak in the LRU and LRU_PERCPU hash maps The LRU and LRU_PERCPU maps allocate a new element on update before locking the target hash table bucket. Right after that the maps try to lock the bucket. If this fails, then maps return -EBUSY to the caller without releasing the allocated element. This makes the element untracked: it doesn't belong to either of free lists, and it doesn't belong to the hash table, so can't be re-used; this eventually leads to the permanent -ENOMEM on LRU map updates, which is unexpected. Fix this by returning the element to the local free list if bucket locking fails. Fixes: 20b6cc34ea74 ("bpf: Avoid hashtab deadlock with map_locked") Signed-off-by: Anton Protopopov Link: https://lore.kernel.org/r/20230522154558.2166815-1-aspsk@isovalent.com Signed-off-by: Martin KaFai Lau --- kernel/bpf/hashtab.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c index 00c253b84bf5..9901efee4339 100644 --- a/kernel/bpf/hashtab.c +++ b/kernel/bpf/hashtab.c @@ -1215,7 +1215,7 @@ static long htab_lru_map_update_elem(struct bpf_map *map, void *key, void *value ret = htab_lock_bucket(htab, b, hash, &flags); if (ret) - return ret; + goto err_lock_bucket; l_old = lookup_elem_raw(head, hash, key, key_size); @@ -1236,6 +1236,7 @@ static long htab_lru_map_update_elem(struct bpf_map *map, void *key, void *value err: htab_unlock_bucket(htab, b, hash, flags); +err_lock_bucket: if (ret) htab_lru_push_free(htab, l_new); else if (l_old) @@ -1338,7 +1339,7 @@ static long __htab_lru_percpu_map_update_elem(struct bpf_map *map, void *key, ret = htab_lock_bucket(htab, b, hash, &flags); if (ret) - return ret; + goto err_lock_bucket; l_old = lookup_elem_raw(head, hash, key, key_size); @@ -1361,6 +1362,7 @@ static long __htab_lru_percpu_map_update_elem(struct bpf_map *map, void *key, ret = 0; err: htab_unlock_bucket(htab, b, hash, flags); +err_lock_bucket: if (l_new) bpf_lru_push_free(&htab->lru, &l_new->lru_node); return ret; -- cgit v1.2.3 From ab1de7ead871ebe6d12a774c3c25de0388cde082 Mon Sep 17 00:00:00 2001 From: Qi Zheng Date: Wed, 17 May 2023 07:45:45 +0000 Subject: cgroup: fix missing cpus_read_{lock,unlock}() in cgroup_transfer_tasks() The commit 4f7e7236435c ("cgroup: Fix threadgroup_rwsem <-> cpus_read_lock() deadlock") fixed the deadlock between cgroup_threadgroup_rwsem and cpus_read_lock() by introducing cgroup_attach_{lock,unlock}() and removing cpus_read_{lock,unlock}() from cpuset_attach(). But cgroup_transfer_tasks() was missed and not handled, which will cause th following warning: WARNING: CPU: 0 PID: 589 at kernel/cpu.c:526 lockdep_assert_cpus_held+0x32/0x40 CPU: 0 PID: 589 Comm: kworker/1:4 Not tainted 6.4.0-rc2-next-20230517 #50 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.14.0-2 04/01/2014 Workqueue: events cpuset_hotplug_workfn RIP: 0010:lockdep_assert_cpus_held+0x32/0x40 <...> Call Trace: cpuset_attach+0x40/0x240 cgroup_migrate_execute+0x452/0x5e0 ? _raw_spin_unlock_irq+0x28/0x40 cgroup_transfer_tasks+0x1f3/0x360 ? find_held_lock+0x32/0x90 ? cpuset_hotplug_workfn+0xc81/0xed0 cpuset_hotplug_workfn+0xcb1/0xed0 ? process_one_work+0x248/0x5b0 process_one_work+0x2b9/0x5b0 worker_thread+0x56/0x3b0 ? process_one_work+0x5b0/0x5b0 kthread+0xf1/0x120 ? kthread_complete_and_exit+0x20/0x20 ret_from_fork+0x1f/0x30 So just use the cgroup_attach_{lock,unlock}() helper to fix it. Reported-by: Zhao Gongyi Signed-off-by: Qi Zheng Acked-by: Muchun Song Fixes: 05c7b7a92cc8 ("cgroup/cpuset: Fix a race between cpuset_attach() and cpu hotplug") Cc: stable@vger.kernel.org # v5.17+ Signed-off-by: Tejun Heo --- kernel/cgroup/cgroup-v1.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/cgroup/cgroup-v1.c b/kernel/cgroup/cgroup-v1.c index aeef06c465ef..5407241dbb45 100644 --- a/kernel/cgroup/cgroup-v1.c +++ b/kernel/cgroup/cgroup-v1.c @@ -108,7 +108,7 @@ int cgroup_transfer_tasks(struct cgroup *to, struct cgroup *from) cgroup_lock(); - percpu_down_write(&cgroup_threadgroup_rwsem); + cgroup_attach_lock(true); /* all tasks in @from are being moved, all csets are source */ spin_lock_irq(&css_set_lock); @@ -144,7 +144,7 @@ int cgroup_transfer_tasks(struct cgroup *to, struct cgroup *from) } while (task && !ret); out_err: cgroup_migrate_finish(&mgctx); - percpu_up_write(&cgroup_threadgroup_rwsem); + cgroup_attach_unlock(true); cgroup_unlock(); return ret; } -- cgit v1.2.3 From d36f6efbe0cb422fe1e4475717d75f3737088832 Mon Sep 17 00:00:00 2001 From: Harshit Mogalapalli Date: Thu, 27 Apr 2023 22:59:33 -0700 Subject: module: Fix use-after-free bug in read_file_mod_stats() Smatch warns: kernel/module/stats.c:394 read_file_mod_stats() warn: passing freed memory 'buf' We are passing 'buf' to simple_read_from_buffer() after freeing it. Fix this by changing the order of 'simple_read_from_buffer' and 'kfree'. Fixes: df3e764d8e5c ("module: add debug stats to help identify memory pressure") Signed-off-by: Harshit Mogalapalli Signed-off-by: Luis Chamberlain --- kernel/module/stats.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/module/stats.c b/kernel/module/stats.c index ad7b6ada29f2..6ab2c94d6bc3 100644 --- a/kernel/module/stats.c +++ b/kernel/module/stats.c @@ -276,6 +276,7 @@ static ssize_t read_file_mod_stats(struct file *file, char __user *user_buf, struct mod_fail_load *mod_fail; unsigned int len, size, count_failed = 0; char *buf; + int ret; u32 live_mod_count, fkreads, fdecompress, fbecoming, floads; unsigned long total_size, text_size, ikread_bytes, ibecoming_bytes, idecompress_bytes, imod_bytes, total_virtual_lost; @@ -390,8 +391,9 @@ static ssize_t read_file_mod_stats(struct file *file, char __user *user_buf, out_unlock: mutex_unlock(&module_mutex); out: + ret = simple_read_from_buffer(user_buf, count, ppos, buf, len); kfree(buf); - return simple_read_from_buffer(user_buf, count, ppos, buf, len); + return ret; } #undef MAX_PREAMBLE #undef MAX_FAILED_MOD_PRINT -- cgit v1.2.3 From 2bd110339288c18823dcace602b63b0d8627e520 Mon Sep 17 00:00:00 2001 From: John Sperbeck Date: Sun, 21 May 2023 19:29:53 +0000 Subject: cgroup: always put cset in cgroup_css_set_put_fork A successful call to cgroup_css_set_fork() will always have taken a ref on kargs->cset (regardless of CLONE_INTO_CGROUP), so always do a corresponding put in cgroup_css_set_put_fork(). Without this, a cset and its contained css structures will be leaked for some fork failures. The following script reproduces the leak for a fork failure due to exceeding pids.max in the pids controller. A similar thing can happen if we jump to the bad_fork_cancel_cgroup label in copy_process(). [ -z "$1" ] && echo "Usage $0 pids-root" && exit 1 PID_ROOT=$1 CGROUP=$PID_ROOT/foo [ -e $CGROUP ] && rmdir -f $CGROUP mkdir $CGROUP echo 5 > $CGROUP/pids.max echo $$ > $CGROUP/cgroup.procs fork_bomb() { set -e for i in $(seq 10); do /bin/sleep 3600 & done } (fork_bomb) & wait echo $$ > $PID_ROOT/cgroup.procs kill $(cat $CGROUP/cgroup.procs) rmdir $CGROUP Fixes: ef2c41cf38a7 ("clone3: allow spawning processes into cgroups") Cc: stable@vger.kernel.org # v5.7+ Signed-off-by: John Sperbeck Signed-off-by: Tejun Heo --- kernel/cgroup/cgroup.c | 17 ++++++++--------- 1 file changed, 8 insertions(+), 9 deletions(-) (limited to 'kernel') diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c index 625d7483951c..245cf62ce85a 100644 --- a/kernel/cgroup/cgroup.c +++ b/kernel/cgroup/cgroup.c @@ -6486,19 +6486,18 @@ err: static void cgroup_css_set_put_fork(struct kernel_clone_args *kargs) __releases(&cgroup_threadgroup_rwsem) __releases(&cgroup_mutex) { + struct cgroup *cgrp = kargs->cgrp; + struct css_set *cset = kargs->cset; + cgroup_threadgroup_change_end(current); - if (kargs->flags & CLONE_INTO_CGROUP) { - struct cgroup *cgrp = kargs->cgrp; - struct css_set *cset = kargs->cset; + if (cset) { + put_css_set(cset); + kargs->cset = NULL; + } + if (kargs->flags & CLONE_INTO_CGROUP) { cgroup_unlock(); - - if (cset) { - put_css_set(cset); - kargs->cset = NULL; - } - if (cgrp) { cgroup_put(cgrp); kargs->cgrp = NULL; -- cgit v1.2.3 From ee7751b564a90f337330efc1221df40647d68756 Mon Sep 17 00:00:00 2001 From: Beau Belgrave Date: Fri, 5 May 2023 13:58:55 -0700 Subject: tracing/user_events: Use long vs int for atomic bit ops Each event stores a int to track which bit to set/clear when enablement changes. On big endian 64-bit configurations, it's possible this could cause memory corruption when it's used for atomic bit operations. Use unsigned long for enablement values to ensure any possible corruption cannot occur. Downcast to int after mask for the bit target. Link: https://lore.kernel.org/all/6f758683-4e5e-41c3-9b05-9efc703e827c@kili.mountain/ Link: https://lore.kernel.org/linux-trace-kernel/20230505205855.6407-1-beaub@linux.microsoft.com Fixes: dcb8177c1395 ("tracing/user_events: Add ioctl for disabling addresses") Reported-by: Dan Carpenter Signed-off-by: Beau Belgrave Signed-off-by: Steven Rostedt (Google) --- kernel/trace/trace_events_user.c | 15 ++++++++------- 1 file changed, 8 insertions(+), 7 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/trace_events_user.c b/kernel/trace/trace_events_user.c index b1ecd7677642..e37c7f168c44 100644 --- a/kernel/trace/trace_events_user.c +++ b/kernel/trace/trace_events_user.c @@ -101,7 +101,7 @@ struct user_event_enabler { unsigned long addr; /* Track enable bit, flags, etc. Aligned for bitops. */ - unsigned int values; + unsigned long values; }; /* Bits 0-5 are for the bit to update upon enable/disable (0-63 allowed) */ @@ -116,7 +116,9 @@ struct user_event_enabler { /* Only duplicate the bit value */ #define ENABLE_VAL_DUP_MASK ENABLE_VAL_BIT_MASK -#define ENABLE_BITOPS(e) ((unsigned long *)&(e)->values) +#define ENABLE_BITOPS(e) (&(e)->values) + +#define ENABLE_BIT(e) ((int)((e)->values & ENABLE_VAL_BIT_MASK)) /* Used for asynchronous faulting in of pages */ struct user_event_enabler_fault { @@ -423,9 +425,9 @@ static int user_event_enabler_write(struct user_event_mm *mm, /* Update bit atomically, user tracers must be atomic as well */ if (enabler->event && enabler->event->status) - set_bit(enabler->values & ENABLE_VAL_BIT_MASK, ptr); + set_bit(ENABLE_BIT(enabler), ptr); else - clear_bit(enabler->values & ENABLE_VAL_BIT_MASK, ptr); + clear_bit(ENABLE_BIT(enabler), ptr); kunmap_local(kaddr); unpin_user_pages_dirty_lock(&page, 1, true); @@ -440,8 +442,7 @@ static bool user_event_enabler_exists(struct user_event_mm *mm, struct user_event_enabler *next; list_for_each_entry_safe(enabler, next, &mm->enablers, link) { - if (enabler->addr == uaddr && - (enabler->values & ENABLE_VAL_BIT_MASK) == bit) + if (enabler->addr == uaddr && ENABLE_BIT(enabler) == bit) return true; } @@ -2272,7 +2273,7 @@ static long user_events_ioctl_unreg(unsigned long uarg) list_for_each_entry_safe(enabler, next, &mm->enablers, link) if (enabler->addr == reg.disable_addr && - (enabler->values & ENABLE_VAL_BIT_MASK) == reg.disable_bit) { + ENABLE_BIT(enabler) == reg.disable_bit) { set_bit(ENABLE_VAL_FREEING_BIT, ENABLE_BITOPS(enabler)); if (!test_bit(ENABLE_VAL_FAULTING_BIT, ENABLE_BITOPS(enabler))) -- cgit v1.2.3 From 632478a05821bc1c9b55c3a1dd0fb1be7bfa1acc Mon Sep 17 00:00:00 2001 From: Daniel Bristot de Oliveira Date: Thu, 11 May 2023 18:32:01 +0200 Subject: tracing/timerlat: Always wakeup the timerlat thread While testing rtla timerlat auto analysis, I reach a condition where the interface was not receiving tracing data. I was able to manually reproduce the problem with these steps: # echo 0 > tracing_on # disable trace # echo 1 > osnoise/stop_tracing_us # stop trace if timerlat irq > 1 us # echo timerlat > current_tracer # enable timerlat tracer # sleep 1 # wait... that is the time when rtla # apply configs like prio or cgroup # echo 1 > tracing_on # start tracing # cat trace # tracer: timerlat # # _-----=> irqs-off # / _----=> need-resched # | / _---=> hardirq/softirq # || / _--=> preempt-depth # ||| / _-=> migrate-disable # |||| / delay # ||||| ACTIVATION # TASK-PID CPU# ||||| TIMESTAMP ID CONTEXT LATENCY # | | | ||||| | | | | NOTHING! Then, trying to enable tracing again with echo 1 > tracing_on resulted in no change: the trace was still not tracing. This problem happens because the timerlat IRQ hits the stop tracing condition while tracing is off, and do not wake up the timerlat thread, so the timerlat threads are kept sleeping forever, resulting in no trace, even after re-enabling the tracer. Avoid this condition by always waking up the threads, even after stopping tracing, allowing the tracer to return to its normal operating after a new tracing on. Link: https://lore.kernel.org/linux-trace-kernel/1ed8f830638b20a39d535d27d908e319a9a3c4e2.1683822622.git.bristot@kernel.org Cc: Juri Lelli Cc: stable@vger.kernel.org Fixes: a955d7eac177 ("trace: Add timerlat tracer") Signed-off-by: Daniel Bristot de Oliveira Signed-off-by: Steven Rostedt (Google) --- kernel/trace/trace_osnoise.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'kernel') diff --git a/kernel/trace/trace_osnoise.c b/kernel/trace/trace_osnoise.c index efbbec2caff8..e97e3fa5cbed 100644 --- a/kernel/trace/trace_osnoise.c +++ b/kernel/trace/trace_osnoise.c @@ -1652,6 +1652,8 @@ static enum hrtimer_restart timerlat_irq(struct hrtimer *timer) osnoise_stop_tracing(); notify_new_max_latency(diff); + wake_up_process(tlat->kthread); + return HRTIMER_NORESTART; } } -- cgit v1.2.3 From 3e0fea09b17fa2255f6cb0108bbffd4a505f8925 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Fri, 19 May 2023 16:07:38 -0700 Subject: tracing/user_events: Split up mm alloc and attach When a new mm is being created in a fork() path it currently is allocated and then attached in one go. This leaves the mm exposed out to the tracing register callbacks while any parent enabler locations are copied in. This should not happen. Split up mm alloc and attach as unique operations. When duplicating enablers, first alloc, then duplicate, and only upon success, attach. This prevents any timing window outside of the event_reg mutex for enablement walking. This allows for dropping RCU requirement for enablement walking in later patches. Link: https://lkml.kernel.org/r/20230519230741.669-2-beaub@linux.microsoft.com Link: https://lore.kernel.org/linux-trace-kernel/CAHk-=whTBvXJuoi_kACo3qi5WZUmRrhyA-_=rRFsycTytmB6qw@mail.gmail.com/ Signed-off-by: Linus Torvalds [ change log written by Beau Belgrave ] Signed-off-by: Beau Belgrave Signed-off-by: Steven Rostedt (Google) --- kernel/trace/trace_events_user.c | 29 ++++++++++++++++++----------- 1 file changed, 18 insertions(+), 11 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/trace_events_user.c b/kernel/trace/trace_events_user.c index e37c7f168c44..599aab46a94b 100644 --- a/kernel/trace/trace_events_user.c +++ b/kernel/trace/trace_events_user.c @@ -539,10 +539,9 @@ static struct user_event_mm *user_event_mm_get_all(struct user_event *user) return found; } -static struct user_event_mm *user_event_mm_create(struct task_struct *t) +static struct user_event_mm *user_event_mm_alloc(struct task_struct *t) { struct user_event_mm *user_mm; - unsigned long flags; user_mm = kzalloc(sizeof(*user_mm), GFP_KERNEL_ACCOUNT); @@ -554,12 +553,6 @@ static struct user_event_mm *user_event_mm_create(struct task_struct *t) refcount_set(&user_mm->refcnt, 1); refcount_set(&user_mm->tasks, 1); - spin_lock_irqsave(&user_event_mms_lock, flags); - list_add_rcu(&user_mm->link, &user_event_mms); - spin_unlock_irqrestore(&user_event_mms_lock, flags); - - t->user_event_mm = user_mm; - /* * The lifetime of the memory descriptor can slightly outlast * the task lifetime if a ref to the user_event_mm is taken @@ -573,6 +566,17 @@ static struct user_event_mm *user_event_mm_create(struct task_struct *t) return user_mm; } +static void user_event_mm_attach(struct user_event_mm *user_mm, struct task_struct *t) +{ + unsigned long flags; + + spin_lock_irqsave(&user_event_mms_lock, flags); + list_add_rcu(&user_mm->link, &user_event_mms); + spin_unlock_irqrestore(&user_event_mms_lock, flags); + + t->user_event_mm = user_mm; +} + static struct user_event_mm *current_user_event_mm(void) { struct user_event_mm *user_mm = current->user_event_mm; @@ -580,10 +584,12 @@ static struct user_event_mm *current_user_event_mm(void) if (user_mm) goto inc; - user_mm = user_event_mm_create(current); + user_mm = user_event_mm_alloc(current); if (!user_mm) goto error; + + user_event_mm_attach(user_mm, current); inc: refcount_inc(&user_mm->refcnt); error: @@ -671,7 +677,7 @@ void user_event_mm_remove(struct task_struct *t) void user_event_mm_dup(struct task_struct *t, struct user_event_mm *old_mm) { - struct user_event_mm *mm = user_event_mm_create(t); + struct user_event_mm *mm = user_event_mm_alloc(t); struct user_event_enabler *enabler; if (!mm) @@ -685,10 +691,11 @@ void user_event_mm_dup(struct task_struct *t, struct user_event_mm *old_mm) rcu_read_unlock(); + user_event_mm_attach(mm, t); return; error: rcu_read_unlock(); - user_event_mm_remove(t); + user_event_mm_destroy(mm); } static bool current_user_event_enabler_exists(unsigned long uaddr, -- cgit v1.2.3 From aaecdaf922835ed9a8ce56cdd9a8d40fe630257a Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Fri, 19 May 2023 16:07:39 -0700 Subject: tracing/user_events: Remove RCU lock while pinning pages pin_user_pages_remote() can reschedule which means we cannot hold any RCU lock while using it. Now that enablers are not exposed out to the tracing register callbacks during fork(), there is clearly no need to require the RCU lock as event_mutex is enough to protect changes. Remove unneeded RCU usages when pinning pages and walking enablers with event_mutex held. Cleanup a misleading "safe" list walk that is not needed. During fork() duplication, remove unneeded RCU list add, since the list is not exposed yet. Link: https://lkml.kernel.org/r/20230519230741.669-3-beaub@linux.microsoft.com Link: https://lore.kernel.org/linux-trace-kernel/CAHk-=wiiBfT4zNS29jA0XEsy8EmbqTH1hAPdRJCDAJMD8Gxt5A@mail.gmail.com/ Fixes: 7235759084a4 ("tracing/user_events: Use remote writes for event enablement") Signed-off-by: Linus Torvalds [ change log written by Beau Belgrave ] Signed-off-by: Beau Belgrave Signed-off-by: Steven Rostedt (Google) --- kernel/trace/trace_events_user.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/trace_events_user.c b/kernel/trace/trace_events_user.c index 599aab46a94b..d34a59630e70 100644 --- a/kernel/trace/trace_events_user.c +++ b/kernel/trace/trace_events_user.c @@ -439,9 +439,8 @@ static bool user_event_enabler_exists(struct user_event_mm *mm, unsigned long uaddr, unsigned char bit) { struct user_event_enabler *enabler; - struct user_event_enabler *next; - list_for_each_entry_safe(enabler, next, &mm->enablers, link) { + list_for_each_entry(enabler, &mm->enablers, link) { if (enabler->addr == uaddr && ENABLE_BIT(enabler) == bit) return true; } @@ -456,19 +455,19 @@ static void user_event_enabler_update(struct user_event *user) struct user_event_mm *next; int attempt; + lockdep_assert_held(&event_mutex); + while (mm) { next = mm->next; mmap_read_lock(mm->mm); - rcu_read_lock(); - list_for_each_entry_rcu(enabler, &mm->enablers, link) { + list_for_each_entry(enabler, &mm->enablers, link) { if (enabler->event == user) { attempt = 0; user_event_enabler_write(mm, enabler, true, &attempt); } } - rcu_read_unlock(); mmap_read_unlock(mm->mm); user_event_mm_put(mm); mm = next; @@ -496,7 +495,9 @@ static bool user_event_enabler_dup(struct user_event_enabler *orig, enabler->values = orig->values & ENABLE_VAL_DUP_MASK; refcount_inc(&enabler->event->refcnt); - list_add_rcu(&enabler->link, &mm->enablers); + + /* Enablers not exposed yet, RCU not required */ + list_add(&enabler->link, &mm->enablers); return true; } -- cgit v1.2.3 From dcbd1ac2668b5fa02069ea96d581ca3f70a7543c Mon Sep 17 00:00:00 2001 From: Beau Belgrave Date: Fri, 19 May 2023 16:07:40 -0700 Subject: tracing/user_events: Rename link fields for clarity Currently most list_head fields of various structs within user_events are simply named link. This causes folks to keep additional context in their head when working with the code, which can be confusing. Instead of using link, describe what the actual link is, for example: list_del_rcu(&mm->link); Changes into: list_del_rcu(&mm->mms_link); The reader now is given a hint the link is to the mms global list instead of having to remember or spot check within the code. Link: https://lkml.kernel.org/r/20230519230741.669-4-beaub@linux.microsoft.com Link: https://lore.kernel.org/linux-trace-kernel/CAHk-=wicngggxVpbnrYHjRTwGE0WYscPRM+L2HO2BF8ia1EXgQ@mail.gmail.com/ Suggested-by: Linus Torvalds Signed-off-by: Beau Belgrave Signed-off-by: Steven Rostedt (Google) --- kernel/trace/trace_events_user.c | 40 ++++++++++++++++++++++------------------ 1 file changed, 22 insertions(+), 18 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/trace_events_user.c b/kernel/trace/trace_events_user.c index d34a59630e70..238c7a0615fa 100644 --- a/kernel/trace/trace_events_user.c +++ b/kernel/trace/trace_events_user.c @@ -96,7 +96,7 @@ struct user_event { * these to track enablement sites that are tied to an event. */ struct user_event_enabler { - struct list_head link; + struct list_head mm_enablers_link; struct user_event *event; unsigned long addr; @@ -155,7 +155,7 @@ struct user_event_file_info { #define VALIDATOR_REL (1 << 1) struct user_event_validator { - struct list_head link; + struct list_head user_event_link; int offset; int flags; }; @@ -261,7 +261,7 @@ error: static void user_event_enabler_destroy(struct user_event_enabler *enabler) { - list_del_rcu(&enabler->link); + list_del_rcu(&enabler->mm_enablers_link); /* No longer tracking the event via the enabler */ refcount_dec(&enabler->event->refcnt); @@ -440,7 +440,7 @@ static bool user_event_enabler_exists(struct user_event_mm *mm, { struct user_event_enabler *enabler; - list_for_each_entry(enabler, &mm->enablers, link) { + list_for_each_entry(enabler, &mm->enablers, mm_enablers_link) { if (enabler->addr == uaddr && ENABLE_BIT(enabler) == bit) return true; } @@ -461,7 +461,7 @@ static void user_event_enabler_update(struct user_event *user) next = mm->next; mmap_read_lock(mm->mm); - list_for_each_entry(enabler, &mm->enablers, link) { + list_for_each_entry(enabler, &mm->enablers, mm_enablers_link) { if (enabler->event == user) { attempt = 0; user_event_enabler_write(mm, enabler, true, &attempt); @@ -497,7 +497,7 @@ static bool user_event_enabler_dup(struct user_event_enabler *orig, refcount_inc(&enabler->event->refcnt); /* Enablers not exposed yet, RCU not required */ - list_add(&enabler->link, &mm->enablers); + list_add(&enabler->mm_enablers_link, &mm->enablers); return true; } @@ -527,13 +527,15 @@ static struct user_event_mm *user_event_mm_get_all(struct user_event *user) */ rcu_read_lock(); - list_for_each_entry_rcu(mm, &user_event_mms, link) - list_for_each_entry_rcu(enabler, &mm->enablers, link) + list_for_each_entry_rcu(mm, &user_event_mms, mms_link) { + list_for_each_entry_rcu(enabler, &mm->enablers, mm_enablers_link) { if (enabler->event == user) { mm->next = found; found = user_event_mm_get(mm); break; } + } + } rcu_read_unlock(); @@ -572,7 +574,7 @@ static void user_event_mm_attach(struct user_event_mm *user_mm, struct task_stru unsigned long flags; spin_lock_irqsave(&user_event_mms_lock, flags); - list_add_rcu(&user_mm->link, &user_event_mms); + list_add_rcu(&user_mm->mms_link, &user_event_mms); spin_unlock_irqrestore(&user_event_mms_lock, flags); t->user_event_mm = user_mm; @@ -601,7 +603,7 @@ static void user_event_mm_destroy(struct user_event_mm *mm) { struct user_event_enabler *enabler, *next; - list_for_each_entry_safe(enabler, next, &mm->enablers, link) + list_for_each_entry_safe(enabler, next, &mm->enablers, mm_enablers_link) user_event_enabler_destroy(enabler); mmdrop(mm->mm); @@ -638,7 +640,7 @@ void user_event_mm_remove(struct task_struct *t) /* Remove the mm from the list, so it can no longer be enabled */ spin_lock_irqsave(&user_event_mms_lock, flags); - list_del_rcu(&mm->link); + list_del_rcu(&mm->mms_link); spin_unlock_irqrestore(&user_event_mms_lock, flags); /* @@ -686,9 +688,10 @@ void user_event_mm_dup(struct task_struct *t, struct user_event_mm *old_mm) rcu_read_lock(); - list_for_each_entry_rcu(enabler, &old_mm->enablers, link) + list_for_each_entry_rcu(enabler, &old_mm->enablers, mm_enablers_link) { if (!user_event_enabler_dup(enabler, mm)) goto error; + } rcu_read_unlock(); @@ -757,7 +760,7 @@ retry: */ if (!*write_result) { refcount_inc(&enabler->event->refcnt); - list_add_rcu(&enabler->link, &user_mm->enablers); + list_add_rcu(&enabler->mm_enablers_link, &user_mm->enablers); } mutex_unlock(&event_mutex); @@ -913,8 +916,8 @@ static void user_event_destroy_validators(struct user_event *user) struct user_event_validator *validator, *next; struct list_head *head = &user->validators; - list_for_each_entry_safe(validator, next, head, link) { - list_del(&validator->link); + list_for_each_entry_safe(validator, next, head, user_event_link) { + list_del(&validator->user_event_link); kfree(validator); } } @@ -968,7 +971,7 @@ add_validator: validator->offset = offset; /* Want sequential access when validating */ - list_add_tail(&validator->link, &user->validators); + list_add_tail(&validator->user_event_link, &user->validators); add_field: field->type = type; @@ -1358,7 +1361,7 @@ static int user_event_validate(struct user_event *user, void *data, int len) void *pos, *end = data + len; u32 loc, offset, size; - list_for_each_entry(validator, head, link) { + list_for_each_entry(validator, head, user_event_link) { pos = data + validator->offset; /* Already done min_size check, no bounds check here */ @@ -2279,7 +2282,7 @@ static long user_events_ioctl_unreg(unsigned long uarg) */ mutex_lock(&event_mutex); - list_for_each_entry_safe(enabler, next, &mm->enablers, link) + list_for_each_entry_safe(enabler, next, &mm->enablers, mm_enablers_link) { if (enabler->addr == reg.disable_addr && ENABLE_BIT(enabler) == reg.disable_bit) { set_bit(ENABLE_VAL_FREEING_BIT, ENABLE_BITOPS(enabler)); @@ -2290,6 +2293,7 @@ static long user_events_ioctl_unreg(unsigned long uarg) /* Removed at least one */ ret = 0; } + } mutex_unlock(&event_mutex); -- cgit v1.2.3 From ff9e1632d69e596d8ca256deb07433a8f3565038 Mon Sep 17 00:00:00 2001 From: Beau Belgrave Date: Fri, 19 May 2023 16:07:41 -0700 Subject: tracing/user_events: Document user_event_mm one-shot list usage During 6.4 development it became clear that the one-shot list used by the user_event_mm's next field was confusing to others. It is not clear how this list is protected or what the next field usage is for unless you are familiar with the code. Add comments into the user_event_mm struct indicating lock requirement and usage. Also document how and why this approach was used via comments in both user_event_enabler_update() and user_event_mm_get_all() and the rules to properly use it. Link: https://lkml.kernel.org/r/20230519230741.669-5-beaub@linux.microsoft.com Link: https://lore.kernel.org/linux-trace-kernel/CAHk-=wicngggxVpbnrYHjRTwGE0WYscPRM+L2HO2BF8ia1EXgQ@mail.gmail.com/ Suggested-by: Linus Torvalds Signed-off-by: Beau Belgrave Signed-off-by: Steven Rostedt (Google) --- kernel/trace/trace_events_user.c | 23 ++++++++++++++++++++++- 1 file changed, 22 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/trace/trace_events_user.c b/kernel/trace/trace_events_user.c index 238c7a0615fa..dbb14705d0d3 100644 --- a/kernel/trace/trace_events_user.c +++ b/kernel/trace/trace_events_user.c @@ -451,12 +451,25 @@ static bool user_event_enabler_exists(struct user_event_mm *mm, static void user_event_enabler_update(struct user_event *user) { struct user_event_enabler *enabler; - struct user_event_mm *mm = user_event_mm_get_all(user); struct user_event_mm *next; + struct user_event_mm *mm; int attempt; lockdep_assert_held(&event_mutex); + /* + * We need to build a one-shot list of all the mms that have an + * enabler for the user_event passed in. This list is only valid + * while holding the event_mutex. The only reason for this is due + * to the global mm list being RCU protected and we use methods + * which can wait (mmap_read_lock and pin_user_pages_remote). + * + * NOTE: user_event_mm_get_all() increments the ref count of each + * mm that is added to the list to prevent removal timing windows. + * We must always put each mm after they are used, which may wait. + */ + mm = user_event_mm_get_all(user); + while (mm) { next = mm->next; mmap_read_lock(mm->mm); @@ -515,6 +528,14 @@ static struct user_event_mm *user_event_mm_get_all(struct user_event *user) struct user_event_enabler *enabler; struct user_event_mm *mm; + /* + * We use the mm->next field to build a one-shot list from the global + * RCU protected list. To build this list the event_mutex must be held. + * This lets us build a list without requiring allocs that could fail + * when user based events are most wanted for diagnostics. + */ + lockdep_assert_held(&event_mutex); + /* * We do not want to block fork/exec while enablements are being * updated, so we use RCU to walk the current tasks that have used -- cgit v1.2.3 From e30fbc618e97b38dbb49f1d44dcd0778d3f23b8c Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (Google)" Date: Tue, 23 May 2023 22:11:08 -0400 Subject: tracing/histograms: Allow variables to have some modifiers Modifiers are used to change the behavior of keys. For instance, they can grouped into buckets, converted to syscall names (from the syscall identifier), show task->comm of the current pid, be an array of longs that represent a stacktrace, and more. It was found that nothing stopped a value from taking a modifier. As values are simple counters. If this happened, it would call code that was not expecting a modifier and crash the kernel. This was fixed by having the ___create_val_field() function test if a modifier was present and fail if one was. This fixed the crash. Now there's a problem with variables. Variables are used to pass fields from one event to another. Variables are allowed to have some modifiers, as the processing may need to happen at the time of the event (like stacktraces and comm names of the current pid). The issue is that it too uses __create_val_field(). Now that fails on modifiers, variables can no longer use them (this is a regression). As not all modifiers are for variables, have them use a separate check. Link: https://lore.kernel.org/linux-trace-kernel/20230523221108.064a5d82@rorschach.local.home Cc: stable@vger.kernel.org Cc: Masami Hiramatsu Cc: Tom Zanussi Cc: Mark Rutland Fixes: e0213434fe3e4 ("tracing: Do not let histogram values have some modifiers") Signed-off-by: Steven Rostedt (Google) --- kernel/trace/trace_events_hist.c | 23 ++++++++++++++++------- 1 file changed, 16 insertions(+), 7 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/trace_events_hist.c b/kernel/trace/trace_events_hist.c index 486cca3c2b75..543cb7dc84ad 100644 --- a/kernel/trace/trace_events_hist.c +++ b/kernel/trace/trace_events_hist.c @@ -4238,13 +4238,19 @@ static int __create_val_field(struct hist_trigger_data *hist_data, goto out; } - /* Some types cannot be a value */ - if (hist_field->flags & (HIST_FIELD_FL_GRAPH | HIST_FIELD_FL_PERCENT | - HIST_FIELD_FL_BUCKET | HIST_FIELD_FL_LOG2 | - HIST_FIELD_FL_SYM | HIST_FIELD_FL_SYM_OFFSET | - HIST_FIELD_FL_SYSCALL | HIST_FIELD_FL_STACKTRACE)) { - hist_err(file->tr, HIST_ERR_BAD_FIELD_MODIFIER, errpos(field_str)); - ret = -EINVAL; + /* values and variables should not have some modifiers */ + if (hist_field->flags & HIST_FIELD_FL_VAR) { + /* Variable */ + if (hist_field->flags & (HIST_FIELD_FL_GRAPH | HIST_FIELD_FL_PERCENT | + HIST_FIELD_FL_BUCKET | HIST_FIELD_FL_LOG2)) + goto err; + } else { + /* Value */ + if (hist_field->flags & (HIST_FIELD_FL_GRAPH | HIST_FIELD_FL_PERCENT | + HIST_FIELD_FL_BUCKET | HIST_FIELD_FL_LOG2 | + HIST_FIELD_FL_SYM | HIST_FIELD_FL_SYM_OFFSET | + HIST_FIELD_FL_SYSCALL | HIST_FIELD_FL_STACKTRACE)) + goto err; } hist_data->fields[val_idx] = hist_field; @@ -4256,6 +4262,9 @@ static int __create_val_field(struct hist_trigger_data *hist_data, ret = -EINVAL; out: return ret; + err: + hist_err(file->tr, HIST_ERR_BAD_FIELD_MODIFIER, errpos(field_str)); + return -EINVAL; } static int create_val_field(struct hist_trigger_data *hist_data, -- cgit v1.2.3 From 4b512860bdbdddcf41467ebd394f27cb8dfb528c Mon Sep 17 00:00:00 2001 From: "Steven Rostedt (Google)" Date: Tue, 23 May 2023 23:09:13 -0400 Subject: tracing: Rename stacktrace field to common_stacktrace The histogram and synthetic events can use a pseudo event called "stacktrace" that will create a stacktrace at the time of the event and use it just like it was a normal field. We have other pseudo events such as "common_cpu" and "common_timestamp". To stay consistent with that, convert "stacktrace" to "common_stacktrace". As this was used in older kernels, to keep backward compatibility, this will act just like "common_cpu" did with "cpu". That is, "cpu" will be the same as "common_cpu" unless the event has a "cpu" field. In which case, the event's field is used. The same is true with "stacktrace". Also update the documentation to reflect this change. Link: https://lore.kernel.org/linux-trace-kernel/20230523230913.6860e28d@rorschach.local.home Cc: Masami Hiramatsu Cc: Tom Zanussi Cc: Mark Rutland Signed-off-by: Steven Rostedt (Google) --- kernel/trace/trace.c | 2 +- kernel/trace/trace_events.c | 2 ++ kernel/trace/trace_events_hist.c | 16 ++++++++++------ 3 files changed, 13 insertions(+), 7 deletions(-) (limited to 'kernel') diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index ebc59781456a..81801dc31784 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -5752,7 +5752,7 @@ static const char readme_msg[] = "\t table using the key(s) and value(s) named, and the value of a\n" "\t sum called 'hitcount' is incremented. Keys and values\n" "\t correspond to fields in the event's format description. Keys\n" - "\t can be any field, or the special string 'stacktrace'.\n" + "\t can be any field, or the special string 'common_stacktrace'.\n" "\t Compound keys consisting of up to two fields can be specified\n" "\t by the 'keys' keyword. Values must correspond to numeric\n" "\t fields. Sort keys consisting of up to two fields can be\n" diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index 654ffa40457a..57e539d47989 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -194,6 +194,8 @@ static int trace_define_generic_fields(void) __generic_field(int, common_cpu, FILTER_CPU); __generic_field(char *, COMM, FILTER_COMM); __generic_field(char *, comm, FILTER_COMM); + __generic_field(char *, stacktrace, FILTER_STACKTRACE); + __generic_field(char *, STACKTRACE, FILTER_STACKTRACE); return ret; } diff --git a/kernel/trace/trace_events_hist.c b/kernel/trace/trace_events_hist.c index 543cb7dc84ad..b97d3ad832f1 100644 --- a/kernel/trace/trace_events_hist.c +++ b/kernel/trace/trace_events_hist.c @@ -1364,7 +1364,7 @@ static const char *hist_field_name(struct hist_field *field, if (field->field) field_name = field->field->name; else - field_name = "stacktrace"; + field_name = "common_stacktrace"; } else if (field->flags & HIST_FIELD_FL_HITCOUNT) field_name = "hitcount"; @@ -2367,7 +2367,7 @@ parse_field(struct hist_trigger_data *hist_data, struct trace_event_file *file, hist_data->enable_timestamps = true; if (*flags & HIST_FIELD_FL_TIMESTAMP_USECS) hist_data->attrs->ts_in_usecs = true; - } else if (strcmp(field_name, "stacktrace") == 0) { + } else if (strcmp(field_name, "common_stacktrace") == 0) { *flags |= HIST_FIELD_FL_STACKTRACE; } else if (strcmp(field_name, "common_cpu") == 0) *flags |= HIST_FIELD_FL_CPU; @@ -2378,11 +2378,15 @@ parse_field(struct hist_trigger_data *hist_data, struct trace_event_file *file, if (!field || !field->size) { /* * For backward compatibility, if field_name - * was "cpu", then we treat this the same as - * common_cpu. This also works for "CPU". + * was "cpu" or "stacktrace", then we treat this + * the same as common_cpu and common_stacktrace + * respectively. This also works for "CPU", and + * "STACKTRACE". */ if (field && field->filter_type == FILTER_CPU) { *flags |= HIST_FIELD_FL_CPU; + } else if (field && field->filter_type == FILTER_STACKTRACE) { + *flags |= HIST_FIELD_FL_STACKTRACE; } else { hist_err(tr, HIST_ERR_FIELD_NOT_FOUND, errpos(field_name)); @@ -5394,7 +5398,7 @@ static void hist_trigger_print_key(struct seq_file *m, if (key_field->field) seq_printf(m, "%s.stacktrace", key_field->field->name); else - seq_puts(m, "stacktrace:\n"); + seq_puts(m, "common_stacktrace:\n"); hist_trigger_stacktrace_print(m, key + key_field->offset, HIST_STACKTRACE_DEPTH); @@ -5977,7 +5981,7 @@ static int event_hist_trigger_print(struct seq_file *m, if (field->field) seq_printf(m, "%s.stacktrace", field->field->name); else - seq_puts(m, "stacktrace"); + seq_puts(m, "common_stacktrace"); } else hist_field_print(m, field); } -- cgit v1.2.3 From 335b4223466dd75f9f3ea4918187afbadd22e5c8 Mon Sep 17 00:00:00 2001 From: Maximilian Heyne Date: Wed, 3 May 2023 13:16:53 +0000 Subject: x86/pci/xen: populate MSI sysfs entries Commit bf5e758f02fc ("genirq/msi: Simplify sysfs handling") reworked the creation of sysfs entries for MSI IRQs. The creation used to be in msi_domain_alloc_irqs_descs_locked after calling ops->domain_alloc_irqs. Then it moved into __msi_domain_alloc_irqs which is an implementation of domain_alloc_irqs. However, Xen comes with the only other implementation of domain_alloc_irqs and hence doesn't run the sysfs population code anymore. Commit 6c796996ee70 ("x86/pci/xen: Fixup fallout from the PCI/MSI overhaul") set the flag MSI_FLAG_DEV_SYSFS for the xen msi_domain_info but that doesn't actually have an effect because Xen uses it's own domain_alloc_irqs implementation. Fix this by making use of the fallback functions for sysfs population. Fixes: bf5e758f02fc ("genirq/msi: Simplify sysfs handling") Signed-off-by: Maximilian Heyne Reviewed-by: Juergen Gross Link: https://lore.kernel.org/r/20230503131656.15928-1-mheyne@amazon.de Signed-off-by: Juergen Gross --- kernel/irq/msi.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/irq/msi.c b/kernel/irq/msi.c index 7a97bcb086bf..b4c31a5c1147 100644 --- a/kernel/irq/msi.c +++ b/kernel/irq/msi.c @@ -542,7 +542,7 @@ fail: return ret; } -#ifdef CONFIG_PCI_MSI_ARCH_FALLBACKS +#if defined(CONFIG_PCI_MSI_ARCH_FALLBACKS) || defined(CONFIG_PCI_XEN) /** * msi_device_populate_sysfs - Populate msi_irqs sysfs entries for a device * @dev: The device (PCI, platform etc) which will get sysfs entries @@ -574,7 +574,7 @@ void msi_device_destroy_sysfs(struct device *dev) msi_for_each_desc(desc, dev, MSI_DESC_ALL) msi_sysfs_remove_desc(dev, desc); } -#endif /* CONFIG_PCI_MSI_ARCH_FALLBACK */ +#endif /* CONFIG_PCI_MSI_ARCH_FALLBACK || CONFIG_PCI_XEN */ #else /* CONFIG_SYSFS */ static inline int msi_sysfs_create_group(struct device *dev) { return 0; } static inline int msi_sysfs_populate_desc(struct device *dev, struct msi_desc *desc) { return 0; } -- cgit v1.2.3 From 9828ed3f695a138f7add89fa2a186ababceb8006 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Thu, 25 May 2023 09:32:25 -0700 Subject: module: error out early on concurrent load of the same module file It turns out that udev under certain circumstances will concurrently try to load the same modules over-and-over excessively. This isn't a kernel bug, but it ends up affecting the kernel, to the point that under certain circumstances we can fail to boot, because the kernel uses a lot of memory to read all the module data all at once. Note that it isn't a memory leak, it's just basically a thundering herd problem happening at bootup with a lot of CPUs, with the worst cases then being pretty bad. Admittedly the worst situations are somewhat contrived: lots and lots of CPUs, not a lot of memory, and KASAN enabled to make it all slower and as such (unintentionally) exacerbate the problem. Luis explains: [1] "My best assessment of the situation is that each CPU in udev ends up triggering a load of duplicate set of modules, not just one, but *a lot*. Not sure what heuristics udev uses to load a set of modules per CPU." Petr Pavlu chimes in: [2] "My understanding is that udev workers are forked. An initial kmod context is created by the main udevd process but no sharing happens after the fork. It means that the mentioned memory pool logic doesn't really kick in. Multiple parallel load requests come from multiple udev workers, for instance, each handling an udev event for one CPU device and making the exactly same requests as all others are doing at the same time. The optimization idea would be to recognize these duplicate requests at the udevd/kmod level and converge them" Note that module loading has tried to mitigate this issue before, see for example commit 064f4536d139 ("module: avoid allocation if module is already present and ready"), which has a few ASCII graphs on memory use due to this same issue. However, while that noticed that the module was already loaded, and exited with an error early before spending any more time on setting up the module, it didn't handle the case of multiple concurrent module loads all being active - but not complete - at the same time. Yes, one of them will eventually win the race and finalize its copy, and the others will then notice that the module already exists and error out, but while this all happens, we have tons of unnecessary concurrent work being done. Again, the real fix is for udev to not do that (maybe it should use threads instead of fork, and have actual shared data structures and not cause duplicate work). That real fix is apparently not trivial. But it turns out that the kernel already has a pretty good model for dealing with concurrent access to the same file: the i_writecount of the inode. In fact, the module loading already indirectly uses 'i_writecount' , because 'kernel_file_read()' will in fact do ret = deny_write_access(file); if (ret) return ret; ... allow_write_access(file); around the read of the file data. We do not allow concurrent writes to the file, and return -ETXTBUSY if the file was open for writing at the same time as the module data is loaded from it. And the solution to the reader concurrency problem is to simply extend this "no concurrent writers" logic to simply be "exclusive access". Note that "exclusive" in this context isn't really some absolute thing: it's only exclusion from writers and from other "special readers" that do this writer denial. So we simply introduce a variation of that "deny_write_access()" logic that not only denies write access, but also requires that this is the _only_ such access that denies write access. Which means that you can't start loading a module that is already being loaded as a module by somebody else, or you will get the same -ETXTBSY error that you would get if there were writers around. [ It also means that you can't try to load a currently executing executable as a module, for the same reason: executables do that same "deny_write_access()" thing, and that's obviously where the whole ETXTBSY logic traditionally came from. This is not a problem for kernel modules, since the set of normal executable files and kernel module files is entirely disjoint. ] This new function is called "exclusive_deny_write_access()", and the implementation is trivial, in that it's just an atomic decrement of i_writecount if it was 0 before. To use that new exclusivity check, all we then do is wrap the module loading with that exclusive_deny_write_access()() / allow_write_access() pair. The actual patch is a bit bigger than that, because we want to surround not just the "load file data" part, but the whole module setup, to get maximum exclusion. So this ends up splitting up "finit_module()" into a few helper functions to make it all very clear and legible. In Luis' test-case (bringing up 255 vcpu's in a virtual machine [3]), the "wasted vmalloc" space (ie module data read into a vmalloc'ed area in order to be loaded as a module, but then discarded because somebody else loaded the same module instead) dropped from 1.8GiB to 474kB. Yes, that's gigabytes to kilobytes. It doesn't drop completely to zero, because even with this change, you can still end up having completely ser