From fbb1fb4ad415cb31ce944f65a5ca700aaf73a227 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Sun, 8 Oct 2017 21:44:52 -0700 Subject: net: defer call to cgroup_sk_alloc() sk_clone_lock() might run while TCP/DCCP listener already vanished. In order to prevent use after free, it is better to defer cgroup_sk_alloc() to the point we know both parent and child exist, and from process context. Fixes: e994b2f0fb92 ("tcp: do not lock listener to process SYN packets") Signed-off-by: Eric Dumazet Cc: Johannes Weiner Cc: Tejun Heo Signed-off-by: David S. Miller --- kernel/cgroup/cgroup.c | 11 ----------- 1 file changed, 11 deletions(-) (limited to 'kernel') diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c index 44857278eb8a..3380a3e49af5 100644 --- a/kernel/cgroup/cgroup.c +++ b/kernel/cgroup/cgroup.c @@ -5709,17 +5709,6 @@ void cgroup_sk_alloc(struct sock_cgroup_data *skcd) if (cgroup_sk_alloc_disabled) return; - /* Socket clone path */ - if (skcd->val) { - /* - * We might be cloning a socket which is left in an empty - * cgroup and the cgroup might have already been rmdir'd. - * Don't use cgroup_get_live(). - */ - cgroup_get(sock_cgroup_ptr(skcd)); - return; - } - rcu_read_lock(); while (true) { -- cgit v1.2.3-71-gd317 From 75cb070960ade40fba5de32138390f3c85c90941 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 10 Oct 2017 19:12:32 -0700 Subject: Revert "net: defer call to cgroup_sk_alloc()" This reverts commit fbb1fb4ad415cb31ce944f65a5ca700aaf73a227. This was not the proper fix, lets cleanly revert it, so that following patch can be carried to stable versions. sock_cgroup_ptr() callers do not expect a NULL return value. Signed-off-by: Eric Dumazet Cc: Johannes Weiner Cc: Tejun Heo Signed-off-by: David S. Miller --- kernel/cgroup/cgroup.c | 11 +++++++++++ net/core/sock.c | 3 ++- net/ipv4/inet_connection_sock.c | 5 ----- 3 files changed, 13 insertions(+), 6 deletions(-) (limited to 'kernel') diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c index 3380a3e49af5..44857278eb8a 100644 --- a/kernel/cgroup/cgroup.c +++ b/kernel/cgroup/cgroup.c @@ -5709,6 +5709,17 @@ void cgroup_sk_alloc(struct sock_cgroup_data *skcd) if (cgroup_sk_alloc_disabled) return; + /* Socket clone path */ + if (skcd->val) { + /* + * We might be cloning a socket which is left in an empty + * cgroup and the cgroup might have already been rmdir'd. + * Don't use cgroup_get_live(). + */ + cgroup_get(sock_cgroup_ptr(skcd)); + return; + } + rcu_read_lock(); while (true) { diff --git a/net/core/sock.c b/net/core/sock.c index 4499e3153813..70c6ccbdf49f 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -1680,7 +1680,6 @@ struct sock *sk_clone_lock(const struct sock *sk, const gfp_t priority) /* sk->sk_memcg will be populated at accept() time */ newsk->sk_memcg = NULL; - memset(&newsk->sk_cgrp_data, 0, sizeof(newsk->sk_cgrp_data)); atomic_set(&newsk->sk_drops, 0); newsk->sk_send_head = NULL; @@ -1719,6 +1718,8 @@ struct sock *sk_clone_lock(const struct sock *sk, const gfp_t priority) newsk->sk_incoming_cpu = raw_smp_processor_id(); atomic64_set(&newsk->sk_cookie, 0); + cgroup_sk_alloc(&newsk->sk_cgrp_data); + /* * Before updating sk_refcnt, we must commit prior changes to memory * (Documentation/RCU/rculist_nulls.txt for details) diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c index d32c74507314..67aec7a10686 100644 --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@ -26,8 +26,6 @@ #include #include #include -#include -#include #ifdef INET_CSK_DEBUG const char inet_csk_timer_bug_msg[] = "inet_csk BUG: unknown timer value\n"; @@ -478,9 +476,6 @@ struct sock *inet_csk_accept(struct sock *sk, int flags, int *err, bool kern) spin_unlock_bh(&queue->fastopenq.lock); } mem_cgroup_sk_alloc(newsk); - cgroup_sk_alloc(&newsk->sk_cgrp_data); - sock_update_classid(&newsk->sk_cgrp_data); - sock_update_netprioidx(&newsk->sk_cgrp_data); out: release_sock(sk); if (req) -- cgit v1.2.3-71-gd317 From 20608924cc2e6bdeaf6f58ccbe9ddfe12dbfa082 Mon Sep 17 00:00:00 2001 From: Doug Berger Date: Wed, 4 Oct 2017 14:26:26 +0200 Subject: genirq: generic chip: Add irq_gc_mask_disable_and_ack_set() The irq_gc_mask_disable_reg_and_ack() function name implies that it provides the combined functions of irq_gc_mask_disable_reg() and irq_gc_ack(). However, the implementation does not actually do that since it writes the mask instead of the disable register. It also does not maintain the mask cache which makes it inappropriate to use with other masking functions. In addition, commit 659fb32d1b67 ("genirq: replace irq_gc_ack() with {set,clr}_bit variants (fwd)") effectively renamed irq_gc_ack() to irq_gc_ack_set_bit() so this function probably should have also been renamed at that time. The generic chip code currently provides three functions for use with the irq_mask member of the irq_chip structure and two functions for use with the irq_ack member of the irq_chip structure. These functions could be combined into six functions for use with the irq_mask_ack member of the irq_chip structure. However, since only one of the combinations is currently used, only the function irq_gc_mask_disable_and_ack_set() is added by this commit. The '_reg' and '_bit' portions of the base function name were left out of the new combined function name in an attempt to keep the function name length manageable with the 80 character source code line length while still allowing the distinct aspects of each combination to be captured by the name. If other combinations are desired in the future please add them to the irq generic chip library at that time. Signed-off-by: Doug Berger Signed-off-by: Marc Zyngier --- include/linux/irq.h | 1 + kernel/irq/generic-chip.c | 25 +++++++++++++++++++++++++ 2 files changed, 26 insertions(+) (limited to 'kernel') diff --git a/include/linux/irq.h b/include/linux/irq.h index d4728bf6a537..494d328f7051 100644 --- a/include/linux/irq.h +++ b/include/linux/irq.h @@ -1010,6 +1010,7 @@ void irq_gc_unmask_enable_reg(struct irq_data *d); void irq_gc_ack_set_bit(struct irq_data *d); void irq_gc_ack_clr_bit(struct irq_data *d); void irq_gc_mask_disable_reg_and_ack(struct irq_data *d); +void irq_gc_mask_disable_and_ack_set(struct irq_data *d); void irq_gc_eoi(struct irq_data *d); int irq_gc_set_wake(struct irq_data *d, unsigned int on); diff --git a/kernel/irq/generic-chip.c b/kernel/irq/generic-chip.c index 5270a54b9fa4..ec5fe9a0cb05 100644 --- a/kernel/irq/generic-chip.c +++ b/kernel/irq/generic-chip.c @@ -150,6 +150,31 @@ void irq_gc_mask_disable_reg_and_ack(struct irq_data *d) irq_gc_unlock(gc); } +/** + * irq_gc_mask_disable_and_ack_set - Mask and ack pending interrupt + * @d: irq_data + * + * This generic implementation of the irq_mask_ack method is for chips + * with separate enable/disable registers instead of a single mask + * register and where a pending interrupt is acknowledged by setting a + * bit. + * + * Note: This is the only permutation currently used. Similar generic + * functions should be added here if other permutations are required. + */ +void irq_gc_mask_disable_and_ack_set(struct irq_data *d) +{ + struct irq_chip_generic *gc = irq_data_get_irq_chip_data(d); + struct irq_chip_type *ct = irq_data_get_chip_type(d); + u32 mask = d->mask; + + irq_gc_lock(gc); + irq_reg_writel(gc, mask, ct->regs.disable); + *ct->mask_cache &= ~mask; + irq_reg_writel(gc, mask, ct->regs.ack); + irq_gc_unlock(gc); +} + /** * irq_gc_eoi - EOI interrupt * @d: irq_data -- cgit v1.2.3-71-gd317 From 0d08af35f16a0cc418ad2afde3bc5f70ace82705 Mon Sep 17 00:00:00 2001 From: Doug Berger Date: Wed, 4 Oct 2017 14:28:17 +0200 Subject: genirq: generic chip: remove irq_gc_mask_disable_reg_and_ack() Any usage of the irq_gc_mask_disable_reg_and_ack() function has been replaced with the desired functionality. The incorrect and ambiguously named function is removed here to prevent accidental misuse. Signed-off-by: Doug Berger Signed-off-by: Marc Zyngier --- include/linux/irq.h | 1 - kernel/irq/generic-chip.c | 16 ---------------- 2 files changed, 17 deletions(-) (limited to 'kernel') diff --git a/include/linux/irq.h b/include/linux/irq.h index 494d328f7051..5ad10948ea95 100644 --- a/include/linux/irq.h +++ b/include/linux/irq.h @@ -1009,7 +1009,6 @@ void irq_gc_mask_clr_bit(struct irq_data *d); void irq_gc_unmask_enable_reg(struct irq_data *d); void irq_gc_ack_set_bit(struct irq_data *d); void irq_gc_ack_clr_bit(struct irq_data *d); -void irq_gc_mask_disable_reg_and_ack(struct irq_data *d); void irq_gc_mask_disable_and_ack_set(struct irq_data *d); void irq_gc_eoi(struct irq_data *d); int irq_gc_set_wake(struct irq_data *d, unsigned int on); diff --git a/kernel/irq/generic-chip.c b/kernel/irq/generic-chip.c index ec5fe9a0cb05..c26c5bb6b491 100644 --- a/kernel/irq/generic-chip.c +++ b/kernel/irq/generic-chip.c @@ -134,22 +134,6 @@ void irq_gc_ack_clr_bit(struct irq_data *d) irq_gc_unlock(gc); } -/** - * irq_gc_mask_disable_reg_and_ack - Mask and ack pending interrupt - * @d: irq_data - */ -void irq_gc_mask_disable_reg_and_ack(struct irq_data *d) -{ - struct irq_chip_generic *gc = irq_data_get_irq_chip_data(d); - struct irq_chip_type *ct = irq_data_get_chip_type(d); - u32 mask = d->mask; - - irq_gc_lock(gc); - irq_reg_writel(gc, mask, ct->regs.mask); - irq_reg_writel(gc, mask, ct->regs.ack); - irq_gc_unlock(gc); -} - /** * irq_gc_mask_disable_and_ack_set - Mask and ack pending interrupt * @d: irq_data -- cgit v1.2.3-71-gd317 From 28e33f9d78eefe98ea86673ab31e988b37a9a738 Mon Sep 17 00:00:00 2001 From: Jakub Kicinski Date: Mon, 16 Oct 2017 11:16:55 -0700 Subject: bpf: disallow arithmetic operations on context pointer Commit f1174f77b50c ("bpf/verifier: rework value tracking") removed the crafty selection of which pointer types are allowed to be modified. This is OK for most pointer types since adjust_ptr_min_max_vals() will catch operations on immutable pointers. One exception is PTR_TO_CTX which is now allowed to be offseted freely. The intent of aforementioned commit was to allow context access via modified registers. The offset passed to ->is_valid_access() verifier callback has been adjusted by the value of the variable offset. What is missing, however, is taking the variable offset into account when the context register is used. Or in terms of the code adding the offset to the value passed to the ->convert_ctx_access() callback. This leads to the following eBPF user code: r1 += 68 r0 = *(u32 *)(r1 + 8) exit being translated to this in kernel space: 0: (07) r1 += 68 1: (61) r0 = *(u32 *)(r1 +180) 2: (95) exit Offset 8 is corresponding to 180 in the kernel, but offset 76 is valid too. Verifier will "accept" access to offset 68+8=76 but then "convert" access to offset 8 as 180. Effective access to offset 248 is beyond the kernel context. (This is a __sk_buff example on a debug-heavy kernel - packet mark is 8 -> 180, 76 would be data.) Dereferencing the modified context pointer is not as easy as dereferencing other types, because we have to translate the access to reading a field in kernel structures which is usually at a different offset and often of a different size. To allow modifying the pointer we would have to make sure that given eBPF instruction will always access the same field or the fields accessed are "compatible" in terms of offset and size... Disallow dereferencing modified context pointers and add to selftests the test case described here. Fixes: f1174f77b50c ("bpf/verifier: rework value tracking") Signed-off-by: Jakub Kicinski Acked-by: Daniel Borkmann Acked-by: Alexei Starovoitov Acked-by: Edward Cree Signed-off-by: David S. Miller --- kernel/bpf/verifier.c | 8 ++++++-- tools/testing/selftests/bpf/test_verifier.c | 14 ++++++++++++++ 2 files changed, 20 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 8b8d6ba39e23..20f3889c006e 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -1116,7 +1116,12 @@ static int check_mem_access(struct bpf_verifier_env *env, int insn_idx, u32 regn /* ctx accesses must be at a fixed offset, so that we can * determine what type of data were returned. */ - if (!tnum_is_const(reg->var_off)) { + if (reg->off) { + verbose("dereference of modified ctx ptr R%d off=%d+%d, ctx+const is allowed, ctx+const+const is not\n", + regno, reg->off, off - reg->off); + return -EACCES; + } + if (!tnum_is_const(reg->var_off) || reg->var_off.value) { char tn_buf[48]; tnum_strn(tn_buf, sizeof(tn_buf), reg->var_off); @@ -1124,7 +1129,6 @@ static int check_mem_access(struct bpf_verifier_env *env, int insn_idx, u32 regn tn_buf, off, size); return -EACCES; } - off += reg->var_off.value; err = check_ctx_access(env, insn_idx, off, size, t, ®_type); if (!err && t == BPF_READ && value_regno >= 0) { /* ctx access returns either a scalar, or a diff --git a/tools/testing/selftests/bpf/test_verifier.c b/tools/testing/selftests/bpf/test_verifier.c index 26f3250bdcd2..3c7d3a45a3c5 100644 --- a/tools/testing/selftests/bpf/test_verifier.c +++ b/tools/testing/selftests/bpf/test_verifier.c @@ -6645,6 +6645,20 @@ static struct bpf_test tests[] = { .errstr = "BPF_END uses reserved fields", .result = REJECT, }, + { + "arithmetic ops make PTR_TO_CTX unusable", + .insns = { + BPF_ALU64_IMM(BPF_ADD, BPF_REG_1, + offsetof(struct __sk_buff, data) - + offsetof(struct __sk_buff, mark)), + BPF_LDX_MEM(BPF_W, BPF_REG_0, BPF_REG_1, + offsetof(struct __sk_buff, mark)), + BPF_EXIT_INSN(), + }, + .errstr = "dereference of modified ctx ptr R1 off=68+8, ctx+const is allowed, ctx+const+const is not", + .result = REJECT, + .prog_type = BPF_PROG_TYPE_SCHED_CLS, + }, }; static int probe_filter_length(const struct bpf_insn *fp) -- cgit v1.2.3-71-gd317 From 82f8dd28bd3abe181b7a66ea4ea132134d37a400 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Tue, 17 Oct 2017 16:55:53 +0200 Subject: bpf: fix splat for illegal devmap percpu allocation It was reported that syzkaller was able to trigger a splat on devmap percpu allocation due to illegal/unsupported allocation request size passed to __alloc_percpu(): [ 70.094249] illegal size (32776) or align (8) for percpu allocation [ 70.094256] ------------[ cut here ]------------ [ 70.094259] WARNING: CPU: 3 PID: 3451 at mm/percpu.c:1365 pcpu_alloc+0x96/0x630 [...] [ 70.094325] Call Trace: [ 70.094328] __alloc_percpu_gfp+0x12/0x20 [ 70.094330] dev_map_alloc+0x134/0x1e0 [ 70.094331] SyS_bpf+0x9bc/0x1610 [ 70.094333] ? selinux_task_setrlimit+0x5a/0x60 [ 70.094334] ? security_task_setrlimit+0x43/0x60 [ 70.094336] entry_SYSCALL_64_fastpath+0x1a/0xa5 This was due to too large max_entries for the map such that we surpassed the upper limit of PCPU_MIN_UNIT_SIZE. It's fine to fail naturally here, so switch to __alloc_percpu_gfp() and pass __GFP_NOWARN instead. Fixes: 11393cc9b9be ("xdp: Add batching support to redirect map") Reported-by: Mark Rutland Reported-by: Shankara Pailoor Reported-by: Richard Weinberger Signed-off-by: Daniel Borkmann Cc: John Fastabend Acked-by: Alexei Starovoitov Acked-by: John Fastabend Signed-off-by: David S. Miller --- kernel/bpf/devmap.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/devmap.c b/kernel/bpf/devmap.c index e093d9a2c4dd..920428d84da2 100644 --- a/kernel/bpf/devmap.c +++ b/kernel/bpf/devmap.c @@ -111,8 +111,9 @@ static struct bpf_map *dev_map_alloc(union bpf_attr *attr) err = -ENOMEM; /* A per cpu bitfield with a bit per possible net device */ - dtab->flush_needed = __alloc_percpu(dev_map_bitmap_size(attr), - __alignof__(unsigned long)); + dtab->flush_needed = __alloc_percpu_gfp(dev_map_bitmap_size(attr), + __alignof__(unsigned long), + GFP_KERNEL | __GFP_NOWARN); if (!dtab->flush_needed) goto free_dtab; -- cgit v1.2.3-71-gd317 From bc6d5031b43a2291de638ab9304320b4cae61689 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Tue, 17 Oct 2017 16:55:54 +0200 Subject: bpf: do not test for PCPU_MIN_UNIT_SIZE before percpu allocations PCPU_MIN_UNIT_SIZE is an implementation detail of the percpu allocator. Given we support __GFP_NOWARN now, lets just let the allocation request fail naturally instead. The two call sites from BPF mistakenly assumed __GFP_NOWARN would work, so no changes needed to their actual __alloc_percpu_gfp() calls which use the flag already. Signed-off-by: Daniel Borkmann Acked-by: Alexei Starovoitov Acked-by: John Fastabend Signed-off-by: David S. Miller --- kernel/bpf/arraymap.c | 2 +- kernel/bpf/hashtab.c | 4 ---- 2 files changed, 1 insertion(+), 5 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/arraymap.c b/kernel/bpf/arraymap.c index 98c0f00c3f5e..e2636737b69b 100644 --- a/kernel/bpf/arraymap.c +++ b/kernel/bpf/arraymap.c @@ -98,7 +98,7 @@ static struct bpf_map *array_map_alloc(union bpf_attr *attr) array_size += (u64) attr->max_entries * elem_size * num_possible_cpus(); if (array_size >= U32_MAX - PAGE_SIZE || - elem_size > PCPU_MIN_UNIT_SIZE || bpf_array_alloc_percpu(array)) { + bpf_array_alloc_percpu(array)) { bpf_map_area_free(array); return ERR_PTR(-ENOMEM); } diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c index 431126f31ea3..6533f08d1238 100644 --- a/kernel/bpf/hashtab.c +++ b/kernel/bpf/hashtab.c @@ -317,10 +317,6 @@ static struct bpf_map *htab_map_alloc(union bpf_attr *attr) */ goto free_htab; - if (percpu && round_up(htab->map.value_size, 8) > PCPU_MIN_UNIT_SIZE) - /* make sure the size for pcpu_alloc() is reasonable */ - goto free_htab; - htab->elem_size = sizeof(struct htab_elem) + round_up(htab->map.key_size, 8); if (percpu) -- cgit v1.2.3-71-gd317 From a961e40917fb14614d368d8bc9782ca4d6a8cd11 Mon Sep 17 00:00:00 2001 From: Mathieu Desnoyers Date: Thu, 19 Oct 2017 13:30:15 -0400 Subject: membarrier: Provide register expedited private command This introduces a "register private expedited" membarrier command which allows eventual removal of important memory barrier constraints on the scheduler fast-paths. It changes how the "private expedited" membarrier command (new to 4.14) is used from user-space. This new command allows processes to register their intent to use the private expedited command. This affects how the expedited private command introduced in 4.14-rc is meant to be used, and should be merged before 4.14 final. Processes are now required to register before using MEMBARRIER_CMD_PRIVATE_EXPEDITED, otherwise that command returns EPERM. This fixes a problem that arose when designing requested extensions to sys_membarrier() to allow JITs to efficiently flush old code from instruction caches. Several potential algorithms are much less painful if the user register intent to use this functionality early on, for example, before the process spawns the second thread. Registering at this time removes the need to interrupt each and every thread in that process at the first expedited sys_membarrier() system call. Signed-off-by: Mathieu Desnoyers Acked-by: Paul E. McKenney Cc: Peter Zijlstra Cc: Ingo Molnar Cc: Alexander Viro Signed-off-by: Linus Torvalds --- fs/exec.c | 1 + include/linux/mm_types.h | 3 +++ include/linux/sched/mm.h | 16 ++++++++++++++++ include/uapi/linux/membarrier.h | 23 ++++++++++++++++------- kernel/sched/membarrier.c | 34 ++++++++++++++++++++++++++++++---- 5 files changed, 66 insertions(+), 11 deletions(-) (limited to 'kernel') diff --git a/fs/exec.c b/fs/exec.c index 5470d3c1892a..3e14ba25f678 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -1802,6 +1802,7 @@ static int do_execveat_common(int fd, struct filename *filename, /* execve succeeded */ current->fs->in_exec = 0; current->in_execve = 0; + membarrier_execve(current); acct_update_integrals(current); task_numa_free(current); free_bprm(bprm); diff --git a/include/linux/mm_types.h b/include/linux/mm_types.h index 46f4ecf5479a..1861ea8dba77 100644 --- a/include/linux/mm_types.h +++ b/include/linux/mm_types.h @@ -445,6 +445,9 @@ struct mm_struct { unsigned long flags; /* Must use atomic bitops to access the bits */ struct core_state *core_state; /* coredumping support */ +#ifdef CONFIG_MEMBARRIER + atomic_t membarrier_state; +#endif #ifdef CONFIG_AIO spinlock_t ioctx_lock; struct kioctx_table __rcu *ioctx_table; diff --git a/include/linux/sched/mm.h b/include/linux/sched/mm.h index ae53e413fb13..ab9bf7b73954 100644 --- a/include/linux/sched/mm.h +++ b/include/linux/sched/mm.h @@ -211,4 +211,20 @@ static inline void memalloc_noreclaim_restore(unsigned int flags) current->flags = (current->flags & ~PF_MEMALLOC) | flags; } +#ifdef CONFIG_MEMBARRIER +enum { + MEMBARRIER_STATE_PRIVATE_EXPEDITED_READY = (1U << 0), + MEMBARRIER_STATE_SWITCH_MM = (1U << 1), +}; + +static inline void membarrier_execve(struct task_struct *t) +{ + atomic_set(&t->mm->membarrier_state, 0); +} +#else +static inline void membarrier_execve(struct task_struct *t) +{ +} +#endif + #endif /* _LINUX_SCHED_MM_H */ diff --git a/include/uapi/linux/membarrier.h b/include/uapi/linux/membarrier.h index 6d47b3249d8a..4e01ad7ffe98 100644 --- a/include/uapi/linux/membarrier.h +++ b/include/uapi/linux/membarrier.h @@ -52,21 +52,30 @@ * (non-running threads are de facto in such a * state). This only covers threads from the * same processes as the caller thread. This - * command returns 0. The "expedited" commands - * complete faster than the non-expedited ones, - * they never block, but have the downside of - * causing extra overhead. + * command returns 0 on success. The + * "expedited" commands complete faster than + * the non-expedited ones, they never block, + * but have the downside of causing extra + * overhead. A process needs to register its + * intent to use the private expedited command + * prior to using it, otherwise this command + * returns -EPERM. + * @MEMBARRIER_CMD_REGISTER_PRIVATE_EXPEDITED: + * Register the process intent to use + * MEMBARRIER_CMD_PRIVATE_EXPEDITED. Always + * returns 0. * * Command to be passed to the membarrier system call. The commands need to * be a single bit each, except for MEMBARRIER_CMD_QUERY which is assigned to * the value 0. */ enum membarrier_cmd { - MEMBARRIER_CMD_QUERY = 0, - MEMBARRIER_CMD_SHARED = (1 << 0), + MEMBARRIER_CMD_QUERY = 0, + MEMBARRIER_CMD_SHARED = (1 << 0), /* reserved for MEMBARRIER_CMD_SHARED_EXPEDITED (1 << 1) */ /* reserved for MEMBARRIER_CMD_PRIVATE (1 << 2) */ - MEMBARRIER_CMD_PRIVATE_EXPEDITED = (1 << 3), + MEMBARRIER_CMD_PRIVATE_EXPEDITED = (1 << 3), + MEMBARRIER_CMD_REGISTER_PRIVATE_EXPEDITED = (1 << 4), }; #endif /* _UAPI_LINUX_MEMBARRIER_H */ diff --git a/kernel/sched/membarrier.c b/kernel/sched/membarrier.c index a92fddc22747..dd7908743dab 100644 --- a/kernel/sched/membarrier.c +++ b/kernel/sched/membarrier.c @@ -18,6 +18,7 @@ #include #include #include +#include #include "sched.h" /* for cpu_rq(). */ @@ -26,21 +27,26 @@ * except MEMBARRIER_CMD_QUERY. */ #define MEMBARRIER_CMD_BITMASK \ - (MEMBARRIER_CMD_SHARED | MEMBARRIER_CMD_PRIVATE_EXPEDITED) + (MEMBARRIER_CMD_SHARED | MEMBARRIER_CMD_PRIVATE_EXPEDITED \ + | MEMBARRIER_CMD_REGISTER_PRIVATE_EXPEDITED) static void ipi_mb(void *info) { smp_mb(); /* IPIs should be serializing but paranoid. */ } -static void membarrier_private_expedited(void) +static int membarrier_private_expedited(void) { int cpu; bool fallback = false; cpumask_var_t tmpmask; + if (!(atomic_read(¤t->mm->membarrier_state) + & MEMBARRIER_STATE_PRIVATE_EXPEDITED_READY)) + return -EPERM; + if (num_online_cpus() == 1) - return; + return 0; /* * Matches memory barriers around rq->curr modification in @@ -94,6 +100,24 @@ static void membarrier_private_expedited(void) * rq->curr modification in scheduler. */ smp_mb(); /* exit from system call is not a mb */ + return 0; +} + +static void membarrier_register_private_expedited(void) +{ + struct task_struct *p = current; + struct mm_struct *mm = p->mm; + + /* + * We need to consider threads belonging to different thread + * groups, which use the same mm. (CLONE_VM but not + * CLONE_THREAD). + */ + if (atomic_read(&mm->membarrier_state) + & MEMBARRIER_STATE_PRIVATE_EXPEDITED_READY) + return; + atomic_or(MEMBARRIER_STATE_PRIVATE_EXPEDITED_READY, + &mm->membarrier_state); } /** @@ -144,7 +168,9 @@ SYSCALL_DEFINE2(membarrier, int, cmd, int, flags) synchronize_sched(); return 0; case MEMBARRIER_CMD_PRIVATE_EXPEDITED: - membarrier_private_expedited(); + return membarrier_private_expedited(); + case MEMBARRIER_CMD_REGISTER_PRIVATE_EXPEDITED: + membarrier_register_private_expedited(); return 0; default: return -EINVAL; -- cgit v1.2.3-71-gd317 From 27fdb35fe99011d86bcc54f62fe84712c53f4d05 Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Thu, 19 Oct 2017 14:26:21 -0700 Subject: doc: Fix various RCU docbook comment-header problems Because many of RCU's files have not been included into docbook, a number of errors have accumulated. This commit fixes them. Signed-off-by: Paul E. McKenney Signed-off-by: Linus Torvalds --- include/linux/rculist.h | 2 +- include/linux/rcupdate.h | 22 ++++++++++++++-------- include/linux/srcu.h | 1 + kernel/rcu/srcutree.c | 2 +- kernel/rcu/sync.c | 9 ++++++--- kernel/rcu/tree.c | 18 ++++++++++-------- 6 files changed, 33 insertions(+), 21 deletions(-) (limited to 'kernel') diff --git a/include/linux/rculist.h b/include/linux/rculist.h index b1fd8bf85fdc..2bea1d5e9930 100644 --- a/include/linux/rculist.h +++ b/include/linux/rculist.h @@ -276,7 +276,7 @@ static inline void list_splice_tail_init_rcu(struct list_head *list, #define list_entry_rcu(ptr, type, member) \ container_of(lockless_dereference(ptr), type, member) -/** +/* * Where are list_empty_rcu() and list_first_entry_rcu()? * * Implementing those functions following their counterparts list_empty() and diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h index de50d8a4cf41..1a9f70d44af9 100644 --- a/include/linux/rcupdate.h +++ b/include/linux/rcupdate.h @@ -523,7 +523,7 @@ static inline void rcu_preempt_sleep_check(void) { } * Return the value of the specified RCU-protected pointer, but omit * both the smp_read_barrier_depends() and the READ_ONCE(). This * is useful in cases where update-side locks prevent the value of the - * pointer from changing. Please note that this primitive does -not- + * pointer from changing. Please note that this primitive does *not* * prevent the compiler from repeating this reference or combining it * with other references, so it should not be used without protection * of appropriate locks. @@ -568,7 +568,7 @@ static inline void rcu_preempt_sleep_check(void) { } * is handed off from RCU to some other synchronization mechanism, for * example, reference counting or locking. In C11, it would map to * kill_dependency(). It could be used as follows: - * + * `` * rcu_read_lock(); * p = rcu_dereference(gp); * long_lived = is_long_lived(p); @@ -579,6 +579,7 @@ static inline void rcu_preempt_sleep_check(void) { } * p = rcu_pointer_handoff(p); * } * rcu_read_unlock(); + *`` */ #define rcu_pointer_handoff(p) (p) @@ -778,18 +779,21 @@ static inline notrace void rcu_read_unlock_sched_notrace(void) /** * RCU_INIT_POINTER() - initialize an RCU protected pointer + * @p: The pointer to be initialized. + * @v: The value to initialized the pointer to. * * Initialize an RCU-protected pointer in special cases where readers * do not need ordering constraints on the CPU or the compiler. These * special cases are: * - * 1. This use of RCU_INIT_POINTER() is NULLing out the pointer -or- + * 1. This use of RCU_INIT_POINTER() is NULLing out the pointer *or* * 2. The caller has taken whatever steps are required to prevent - * RCU readers from concurrently accessing this pointer -or- + * RCU readers from concurrently accessing this pointer *or* * 3. The referenced data structure has already been exposed to - * readers either at compile time or via rcu_assign_pointer() -and- - * a. You have not made -any- reader-visible changes to - * this structure since then -or- + * readers either at compile time or via rcu_assign_pointer() *and* + * + * a. You have not made *any* reader-visible changes to + * this structure since then *or* * b. It is OK for readers accessing this structure from its * new location to see the old state of the structure. (For * example, the changes were to statistical counters or to @@ -805,7 +809,7 @@ static inline notrace void rcu_read_unlock_sched_notrace(void) * by a single external-to-structure RCU-protected pointer, then you may * use RCU_INIT_POINTER() to initialize the internal RCU-protected * pointers, but you must use rcu_assign_pointer() to initialize the - * external-to-structure pointer -after- you have completely initialized + * external-to-structure pointer *after* you have completely initialized * the reader-accessible portions of the linked structure. * * Note that unlike rcu_assign_pointer(), RCU_INIT_POINTER() provides no @@ -819,6 +823,8 @@ static inline notrace void rcu_read_unlock_sched_notrace(void) /** * RCU_POINTER_INITIALIZER() - statically initialize an RCU protected pointer + * @p: The pointer to be initialized. + * @v: The value to initialized the pointer to. * * GCC-style initialization for an RCU-protected pointer in a structure field. */ diff --git a/include/linux/srcu.h b/include/linux/srcu.h index 39af9bc0f653..62be8966e837 100644 --- a/include/linux/srcu.h +++ b/include/linux/srcu.h @@ -78,6 +78,7 @@ void synchronize_srcu(struct srcu_struct *sp); /** * srcu_read_lock_held - might we be in SRCU read-side critical section? + * @sp: The srcu_struct structure to check * * If CONFIG_DEBUG_LOCK_ALLOC is selected, returns nonzero iff in an SRCU * read-side critical section. In absence of CONFIG_DEBUG_LOCK_ALLOC, diff --git a/kernel/rcu/srcutree.c b/kernel/rcu/srcutree.c index 729a8706751d..6d5880089ff6 100644 --- a/kernel/rcu/srcutree.c +++ b/kernel/rcu/srcutree.c @@ -854,7 +854,7 @@ void __call_srcu(struct srcu_struct *sp, struct rcu_head *rhp, /** * call_srcu() - Queue a callback for invocation after an SRCU grace period * @sp: srcu_struct in queue the callback - * @head: structure to be used for queueing the SRCU callback. + * @rhp: structure to be used for queueing the SRCU callback. * @func: function to be invoked after the SRCU grace period * * The callback function will be invoked some time after a full SRCU diff --git a/kernel/rcu/sync.c b/kernel/rcu/sync.c index 50d1861f7759..3f943efcf61c 100644 --- a/kernel/rcu/sync.c +++ b/kernel/rcu/sync.c @@ -85,6 +85,9 @@ void rcu_sync_init(struct rcu_sync *rsp, enum rcu_sync_type type) } /** + * rcu_sync_enter_start - Force readers onto slow path for multiple updates + * @rsp: Pointer to rcu_sync structure to use for synchronization + * * Must be called after rcu_sync_init() and before first use. * * Ensures rcu_sync_is_idle() returns false and rcu_sync_{enter,exit}() @@ -142,7 +145,7 @@ void rcu_sync_enter(struct rcu_sync *rsp) /** * rcu_sync_func() - Callback function managing reader access to fastpath - * @rsp: Pointer to rcu_sync structure to use for synchronization + * @rhp: Pointer to rcu_head in rcu_sync structure to use for synchronization * * This function is passed to one of the call_rcu() functions by * rcu_sync_exit(), so that it is invoked after a grace period following the @@ -158,9 +161,9 @@ void rcu_sync_enter(struct rcu_sync *rsp) * rcu_sync_exit(). Otherwise, set all state back to idle so that readers * can again use their fastpaths. */ -static void rcu_sync_func(struct rcu_head *rcu) +static void rcu_sync_func(struct rcu_head *rhp) { - struct rcu_sync *rsp = container_of(rcu, struct rcu_sync, cb_head); + struct rcu_sync *rsp = container_of(rhp, struct rcu_sync, cb_head); unsigned long flags; BUG_ON(rsp->gp_state != GP_PASSED); diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c index b0ad62b0e7b8..3e3650e94ae6 100644 --- a/kernel/rcu/tree.c +++ b/kernel/rcu/tree.c @@ -3097,9 +3097,10 @@ __call_rcu(struct rcu_head *head, rcu_callback_t func, * read-side critical sections have completed. call_rcu_sched() assumes * that the read-side critical sections end on enabling of preemption * or on voluntary preemption. - * RCU read-side critical sections are delimited by : - * - rcu_read_lock_sched() and rcu_read_unlock_sched(), OR - * - anything that disables preemption. + * RCU read-side critical sections are delimited by: + * + * - rcu_read_lock_sched() and rcu_read_unlock_sched(), OR + * - anything that disables preemption. * * These may be nested. * @@ -3124,11 +3125,12 @@ EXPORT_SYMBOL_GPL(call_rcu_sched); * handler. This means that read-side critical sections in process * context must not be interrupted by softirqs. This interface is to be * used when most of the read-side critical sections are in softirq context. - * RCU read-side critical sections are delimited by : - * - rcu_read_lock() and rcu_read_unlock(), if in interrupt context. - * OR - * - rcu_read_lock_bh() and rcu_read_unlock_bh(), if in process context. - * These may be nested. + * RCU read-side critical sections are delimited by: + * + * - rcu_read_lock() and rcu_read_unlock(), if in interrupt context, OR + * - rcu_read_lock_bh() and rcu_read_unlock_bh(), if in process context. + * + * These may be nested. * * See the description of call_rcu() for more detailed information on * memory ordering guarantees. -- cgit v1.2.3-71-gd317 From 435bf0d3f99a164df7e8c30428cef266b91d1d3b Mon Sep 17 00:00:00 2001 From: John Fastabend Date: Wed, 18 Oct 2017 07:10:15 -0700 Subject: bpf: enforce TCP only support for sockmap Only TCP sockets have been tested and at the moment the state change callback only handles TCP sockets. This adds a check to ensure that sockets actually being added are TCP sockets. For net-next we can consider UDP support. Signed-off-by: John Fastabend Acked-by: Daniel Borkmann Acked-by: Alexei Starovoitov Signed-off-by: David S. Miller --- kernel/bpf/sockmap.c | 6 ++++++ tools/testing/selftests/bpf/test_maps.c | 12 +++++++++++- 2 files changed, 17 insertions(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/bpf/sockmap.c b/kernel/bpf/sockmap.c index 6424ce0e4969..c68899d5b246 100644 --- a/kernel/bpf/sockmap.c +++ b/kernel/bpf/sockmap.c @@ -840,6 +840,12 @@ static int sock_map_update_elem(struct bpf_map *map, return -EINVAL; } + if (skops.sk->sk_type != SOCK_STREAM || + skops.sk->sk_protocol != IPPROTO_TCP) { + fput(socket->file); + return -EOPNOTSUPP; + } + err = sock_map_ctx_update_elem(&skops, map, key, flags); fput(socket->file); return err; diff --git a/tools/testing/selftests/bpf/test_maps.c b/tools/testing/selftests/bpf/test_maps.c index fe3a443a1102..50ce52d2013d 100644 --- a/tools/testing/selftests/bpf/test_maps.c +++ b/tools/testing/selftests/bpf/test_maps.c @@ -466,7 +466,7 @@ static void test_sockmap(int tasks, void *data) int one = 1, map_fd_rx, map_fd_tx, map_fd_break, s, sc, rc; struct bpf_map *bpf_map_rx, *bpf_map_tx, *bpf_map_break; int ports[] = {50200, 50201, 50202, 50204}; - int err, i, fd, sfd[6] = {0xdeadbeef}; + int err, i, fd, udp, sfd[6] = {0xdeadbeef}; u8 buf[20] = {0x0, 0x5, 0x3, 0x2, 0x1, 0x0}; int parse_prog, verdict_prog; struct sockaddr_in addr; @@ -548,6 +548,16 @@ static void test_sockmap(int tasks, void *data) goto out_sockmap; } + /* Test update with unsupported UDP socket */ + udp = socket(AF_INET, SOCK_DGRAM, 0); + i = 0; + err = bpf_map_update_elem(fd, &i, &udp, BPF_ANY); + if (!err) { + printf("Failed socket SOCK_DGRAM allowed '%i:%i'\n", + i, udp); + goto out_sockmap; + } + /* Test update without programs */ for (i = 0; i < 6; i++) { err = bpf_map_update_elem(fd, &i, &sfd[i], BPF_ANY); -- cgit v1.2.3-71-gd317 From 34f79502bbcfab659b8729da68b5e387f96eb4c1 Mon Sep 17 00:00:00 2001 From: John Fastabend Date: Wed, 18 Oct 2017 07:10:36 -0700 Subject: bpf: avoid preempt enable/disable in sockmap using tcp_skb_cb region SK_SKB BPF programs are run from the socket/tcp context but early in the stack before much of the TCP metadata is needed in tcp_skb_cb. So we can use some unused fields to place BPF metadata needed for SK_SKB programs when implementing the redirect function. This allows us to drop the preempt disable logic. It does however require an API change so sk_redirect_map() has been updated to additionally provide ctx_ptr to skb. Note, we do however continue to disable/enable preemption around actual BPF program running to account for map updates. Signed-off-by: John Fastabend Acked-by: Daniel Borkmann Acked-by: Alexei Starovoitov Signed-off-by: David S. Miller --- include/linux/filter.h | 2 +- include/net/tcp.h | 5 ++++ kernel/bpf/sockmap.c | 19 +++++++------- net/core/filter.c | 29 +++++++++++----------- samples/sockmap/sockmap_kern.c | 2 +- tools/include/uapi/linux/bpf.h | 3 ++- tools/testing/selftests/bpf/bpf_helpers.h | 2 +- tools/testing/selftests/bpf/sockmap_verdict_prog.c | 4 +-- 8 files changed, 36 insertions(+), 30 deletions(-) (limited to 'kernel') diff --git a/include/linux/filter.h b/include/linux/filter.h index d29e58fde364..818a0b26249e 100644 --- a/include/linux/filter.h +++ b/include/linux/filter.h @@ -728,7 +728,7 @@ void xdp_do_flush_map(void); void bpf_warn_invalid_xdp_action(u32 act); void bpf_warn_invalid_xdp_redirect(u32 ifindex); -struct sock *do_sk_redirect_map(void); +struct sock *do_sk_redirect_map(struct sk_buff *skb); #ifdef CONFIG_BPF_JIT extern int bpf_jit_enable; diff --git a/include/net/tcp.h b/include/net/tcp.h index 89974c5286d8..b1ef98ebce53 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -840,6 +840,11 @@ struct tcp_skb_cb { struct inet6_skb_parm h6; #endif } header; /* For incoming skbs */ + struct { + __u32 key; + __u32 flags; + struct bpf_map *map; + } bpf; }; }; diff --git a/kernel/bpf/sockmap.c b/kernel/bpf/sockmap.c index c68899d5b246..beaabb21c3a3 100644 --- a/kernel/bpf/sockmap.c +++ b/kernel/bpf/sockmap.c @@ -39,6 +39,7 @@ #include #include #include +#include struct bpf_stab { struct bpf_map map; @@ -101,9 +102,16 @@ static int smap_verdict_func(struct smap_psock *psock, struct sk_buff *skb) return SK_DROP; skb_orphan(skb); + /* We need to ensure that BPF metadata for maps is also cleared + * when we orphan the skb so that we don't have the possibility + * to reference a stale map. + */ + TCP_SKB_CB(skb)->bpf.map = NULL; skb->sk = psock->sock; bpf_compute_data_end(skb); + preempt_disable(); rc = (*prog->bpf_func)(skb, prog->insnsi); + preempt_enable(); skb->sk = NULL; return rc; @@ -114,17 +122,10 @@ static void smap_do_verdict(struct smap_psock *psock, struct sk_buff *skb) struct sock *sk; int rc; - /* Because we use per cpu values to feed input from sock redirect - * in BPF program to do_sk_redirect_map() call we need to ensure we - * are not preempted. RCU read lock is not sufficient in this case - * with CONFIG_PREEMPT_RCU enabled so we must be explicit here. - */ - preempt_disable(); rc = smap_verdict_func(psock, skb); switch (rc) { case SK_REDIRECT: - sk = do_sk_redirect_map(); - preempt_enable(); + sk = do_sk_redirect_map(skb); if (likely(sk)) { struct smap_psock *peer = smap_psock_sk(sk); @@ -141,8 +142,6 @@ static void smap_do_verdict(struct smap_psock *psock, struct sk_buff *skb) /* Fall through and free skb otherwise */ case SK_DROP: default: - if (rc != SK_REDIRECT) - preempt_enable(); kfree_skb(skb); } } diff --git a/net/core/filter.c b/net/core/filter.c index 74b8c91fb5f4..ca1ba0bbfbc2 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -1839,31 +1839,31 @@ static const struct bpf_func_proto bpf_redirect_proto = { .arg2_type = ARG_ANYTHING, }; -BPF_CALL_3(bpf_sk_redirect_map, struct bpf_map *, map, u32, key, u64, flags) +BPF_CALL_4(bpf_sk_redirect_map, struct sk_buff *, skb, + struct bpf_map *, map, u32, key, u64, flags) { - struct redirect_info *ri = this_cpu_ptr(&redirect_info); + struct tcp_skb_cb *tcb = TCP_SKB_CB(skb); if (unlikely(flags)) return SK_ABORTED; - ri->ifindex = key; - ri->flags = flags; - ri->map = map; + tcb->bpf.key = key; + tcb->bpf.flags = flags; + tcb->bpf.map = map; return SK_REDIRECT; } -struct sock *do_sk_redirect_map(void) +struct sock *do_sk_redirect_map(struct sk_buff *skb) { - struct redirect_info *ri = this_cpu_ptr(&redirect_info); + struct tcp_skb_cb *tcb = TCP_SKB_CB(skb); struct sock *sk = NULL; - if (ri->map) { - sk = __sock_map_lookup_elem(ri->map, ri->ifindex); + if (tcb->bpf.map) { + sk = __sock_map_lookup_elem(tcb->bpf.map, tcb->bpf.key); - ri->ifindex = 0; - ri->map = NULL; - /* we do not clear flags for future lookup */ + tcb->bpf.key = 0; + tcb->bpf.map = NULL; } return sk; @@ -1873,9 +1873,10 @@ static const struct bpf_func_proto bpf_sk_redirect_map_proto = { .func = bpf_sk_redirect_map, .gpl_only = false, .ret_type = RET_INTEGER, - .arg1_type = ARG_CONST_MAP_PTR, - .arg2_type = ARG_ANYTHING, + .arg1_type = ARG_PTR_TO_CTX, + .arg2_type = ARG_CONST_MAP_PTR, .arg3_type = ARG_ANYTHING, + .arg4_type = ARG_ANYTHING, }; BPF_CALL_1(bpf_get_cgroup_classid, const struct sk_buff *, skb) diff --git a/samples/sockmap/sockmap_kern.c b/samples/sockmap/sockmap_kern.c index f9b38ef82dc2..52b0053274f4 100644 --- a/samples/sockmap/sockmap_kern.c +++ b/samples/sockmap/sockmap_kern.c @@ -62,7 +62,7 @@ int bpf_prog2(struct __sk_buff *skb) ret = 1; bpf_printk("sockmap: %d -> %d @ %d\n", lport, bpf_ntohl(rport), ret); - return bpf_sk_redirect_map(&sock_map, ret, 0); + return bpf_sk_redirect_map(skb, &sock_map, ret, 0); } SEC("sockops") diff --git a/tools/include/uapi/linux/bpf.h b/tools/include/uapi/linux/bpf.h index 43ab5c402f98..be9a631a69f7 100644 --- a/tools/include/uapi/linux/bpf.h +++ b/tools/include/uapi/linux/bpf.h @@ -569,9 +569,10 @@ union bpf_attr { * @flags: reserved for future use * Return: 0 on success or negative error code * - * int bpf_sk_redirect_map(map, key, flags) + * int bpf_sk_redirect_map(skb, map, key, flags) * Redirect skb to a sock in map using key as a lookup key for the * sock in map. + * @skb: pointer to skb * @map: pointer to sockmap * @key: key to lookup sock in map * @flags: reserved for future use diff --git a/tools/testing/selftests/bpf/bpf_helpers.h b/tools/testing/selftests/bpf/bpf_helpers.h index 36fb9161b34a..b2e02bdcd098 100644 --- a/tools/testing/selftests/bpf/bpf_helpers.h +++ b/tools/testing/selftests/bpf/bpf_helpers.h @@ -65,7 +65,7 @@ static int (*bpf_xdp_adjust_head)(void *ctx, int offset) = static int (*bpf_setsockopt)(void *ctx, int level, int optname, void *optval, int optlen) = (void *) BPF_FUNC_setsockopt; -static int (*bpf_sk_redirect_map)(void *map, int key, int flags) = +static int (*bpf_sk_redirect_map)(void *ctx, void *map, int key, int flags) = (void *) BPF_FUNC_sk_redirect_map; static int (*bpf_sock_map_update)(void *map, void *key, void *value, unsigned long long flags) = diff --git a/tools/testing/selftests/bpf/sockmap_verdict_prog.c b/tools/testing/selftests/bpf/sockmap_verdict_prog.c index 9b99bd10807d..2cd2d552938b 100644 --- a/tools/testing/selftests/bpf/sockmap_verdict_prog.c +++ b/tools/testing/selftests/bpf/sockmap_verdict_prog.c @@ -61,8 +61,8 @@ int bpf_prog2(struct __sk_buff *skb) bpf_printk("verdict: data[0] = redir(%u:%u)\n", map, sk); if (!map) - return bpf_sk_redirect_map(&sock_map_rx, sk, 0); - return bpf_sk_redirect_map(&sock_map_tx, sk, 0); + return bpf_sk_redirect_map(skb, &sock_map_rx, sk, 0); + return bpf_sk_redirect_map(skb, &sock_map_tx, sk, 0); } char _license[] SEC("license") = "GPL"; -- cgit v1.2.3-71-gd317 From fb50df8d32283cd95932a182a46a10070c4a8832 Mon Sep 17 00:00:00 2001 From: John Fastabend Date: Wed, 18 Oct 2017 07:11:22 -0700 Subject: bpf: require CAP_NET_ADMIN when using sockmap maps Restrict sockmap to CAP_NET_ADMIN. Signed-off-by: John Fastabend Acked-by: Daniel Borkmann Acked-by: Alexei Starovoitov Signed-off-by: David S. Miller --- kernel/bpf/sockmap.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'kernel') diff --git a/kernel/bpf/sockmap.c b/kernel/bpf/sockmap.c index beaabb21c3a3..2b6eb35ae5d3 100644 --- a/kernel/bpf/sockmap.c +++ b/kernel/bpf/sockmap.c @@ -486,6 +486,9 @@ static struct bpf_map *sock_map_alloc(union bpf_attr *attr) int err = -EINVAL; u64 cost; + if (!capable(CAP_NET_ADMIN)) + return ERR_PTR(-EPERM); + /* check sanity of attributes */ if (attr->max_entries == 0 || attr->key_size != 4 || attr->value_size != 4 || attr->map_flags & ~BPF_F_NUMA_NODE) -- cgit v1.2.3-71-gd317 From 9ef2a8cd5c0dcb8e1f1534615c56eb13b630c363 Mon Sep 17 00:00:00 2001 From: John Fastabend Date: Wed, 18 Oct 2017 07:11:44 -0700 Subject: bpf: require CAP_NET_ADMIN when using devmap Devmap is used with XDP which requires CAP_NET_ADMIN so lets also make CAP_NET_ADMIN required to use the map. Signed-off-by: John Fastabend Acked-by: Daniel Borkmann Acked-by: Alexei Starovoitov Signed-off-by: David S. Miller --- kernel/bpf/devmap.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'kernel') diff --git a/kernel/bpf/devmap.c b/kernel/bpf/devmap.c index 920428d84da2..52e0548ba548 100644 --- a/kernel/bpf/devmap.c +++ b/kernel/bpf/devmap.c @@ -78,6 +78,9 @@ static struct bpf_map *dev_map_alloc(union bpf_attr *attr) int err = -EINVAL; u64 cost; + if (!capable(CAP_NET_ADMIN)) + return ERR_PTR(-EPERM); + /* check sanity of attributes */ if (attr->max_entries == 0 || attr->key_size != 4 || attr->value_size != 4 || attr->map_flags & ~BPF_F_NUMA_NODE) -- cgit v1.2.3-71-gd317 From 1c9fec470b81ca5e89391c20a11ead31a1e9314b Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Fri, 20 Oct 2017 07:36:05 -0700 Subject: waitid(): Avoid unbalanced user_access_end() on access_ok() error As pointed out by Linus and David, the earlier waitid() fix resulted in a (currently harmless) unbalanced user_access_end() call. This fixes it to just directly return EFAULT on access_ok() failure. Fixes: 96ca579a1ecc ("waitid(): Add missing access_ok() checks") Acked-by: David Daney Cc: Al Viro Signed-off-by: Kees Cook Signed-off-by: Linus Torvalds --- kernel/exit.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'kernel') diff --git a/kernel/exit.c b/kernel/exit.c index cf28528842bc..f6cad39f35df 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -1611,7 +1611,7 @@ SYSCALL_DEFINE5(waitid, int, which, pid_t, upid, struct siginfo __user *, return err; if (!access_ok(VERIFY_WRITE, infop, sizeof(*infop))) - goto Efault; + return -EFAULT; user_access_begin(); unsafe_put_user(signo, &infop->si_signo, Efault); @@ -1739,7 +1739,7 @@ COMPAT_SYSCALL_DEFINE5(waitid, return err; if (!access_ok(VERIFY_WRITE, infop, sizeof(*infop))) - goto Efault; + return -EFAULT; user_access_begin(); unsafe_put_user(signo, &infop->si_signo, Efault); -- cgit v1.2.3-71-gd317 From 1f7c70d6b2bc5de301f30456621e1161fddf4242 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Sat, 21 Oct 2017 16:06:52 +0200 Subject: cpu/hotplug: Reset node state after operation The recent rework of the cpu hotplug internals changed the usage of the per cpu state->node field, but missed to clean it up after usage. So subsequent hotplug operations use the stale pointer from a previous operation and hand it into the callback functions. The callbacks then dereference a pointer which either belongs to a different facility or points to freed and potentially reused memory. In either case data corruption and crashes are the obvious consequence. Reset the node and the last pointers in the per cpu state to NULL after the operation which set them has completed. Fixes: 96abb968549c ("smp/hotplug: Allow external multi-instance rollback") Reported-by: Tvrtko Ursulin Signed-off-by: Thomas Gleixner Cc: Peter Zijlstra Cc: Sebastian Andrzej Siewior Cc: Boris Ostrovsky Cc: "Paul E. McKenney" Link: https://lkml.kernel.org/r/alpine.DEB.2.20.1710211606130.3213@nanos --- kernel/cpu.c | 5 +++++ 1 file changed, 5 insertions(+) (limited to 'kernel') diff --git a/kernel/cpu.c b/kernel/cpu.c index d851df22f5c5..04892a82f6ac 100644 --- a/kernel/cpu.c +++ b/kernel/cpu.c @@ -632,6 +632,11 @@ cpuhp_invoke_ap_callback(int cpu, enum cpuhp_state state, bool bringup, __cpuhp_kick_ap(st); } + /* + * Clean up the leftovers so the next hotplug operation wont use stale + * data. + */ + st->node = st->last = NULL; return ret; } -- cgit v1.2.3-71-gd317 From 8695a5395661fbb4a4f26c97f801f3800ae4754e Mon Sep 17 00:00:00 2001 From: John Fastabend Date: Thu, 19 Oct 2017 09:03:52 -0700 Subject: bpf: devmap fix arithmetic overflow in bitmap_size calculation An integer overflow is possible in dev_map_bitmap_size() when calculating the BITS_TO_LONG logic which becomes, after macro replacement, (((n) + (d) - 1)/ (d)) where 'n' is a __u32 and 'd' is (8 * sizeof(long)). To avoid overflow cast to u64 before arithmetic. Reported-by: Richard Weinberger Acked-by: Daniel Borkmann Signed-off-by: John Fastabend Acked-by: Alexei Starovoitov Signed-off-by: David S. Miller --- kernel/bpf/devmap.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'kernel') diff --git a/kernel/bpf/devmap.c b/kernel/bpf/devmap.c index 52e0548ba548..e745d6a88224 100644 --- a/kernel/bpf/devmap.c +++ b/kernel/bpf/devmap.c @@ -69,7 +69,7 @@ static LIST_HEAD(dev_map_list); static u64 dev_map_bitmap_size(const union bpf_attr *attr) { - return BITS_TO_LONGS(attr->max_entries) * sizeof(unsigned long); + return BITS_TO_LONGS((u64) attr->max_entries) * sizeof(unsigned long); } static struct bpf_map *dev_map_alloc(union bpf_attr *attr) -- cgit v1.2.3-71-gd317 From fb2a311a31d3457fe8c3ee16f5609877e2ead9f7 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Sat, 21 Oct 2017 02:34:21 +0200 Subject: bpf: fix off by one for range markings with L{T, E} patterns During review I noticed that the current logic for direct packet access marking in check_cond_jmp_op() has an off by one for the upper right range border when marking in find_good_pkt_pointers() with BPF_JLT and BPF_JLE. It's not really harmful given access up to pkt_end is always safe, but we should nevertheless correct the range marking before it becomes ABI. If pkt_data' denotes a pkt_data derived pointer (pkt_data + X), then for pkt_data' < pkt_end in the true branch as well as for pkt_end <= pkt_data' in the false branch we mark the range with X although it should really be X - 1 in these cases. For example, X could be pkt_end - pkt_data, then when testing for pkt_data' < pkt_end the verifier simulation cannot deduce that a byte load of pkt_data' - 1 would succeed in this branch. Fixes: b4e432f1000a ("bpf: enable BPF_J{LT, LE, SLT, SLE} opcodes in verifier") Signed-off-by: Daniel Borkmann Acked-by: Alexei Starovoitov Acked-by: John Fastabend Signed-off-by: David S. Miller --- kernel/bpf/verifier.c | 33 +++++++++++++++++++++------------ 1 file changed, 21 insertions(+), 12 deletions(-) (limited to 'kernel') diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 20f3889c006e..49cb5ad14746 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -2430,12 +2430,15 @@ static int check_alu_op(struct bpf_verifier_env *env, struct bpf_insn *insn) } static void find_good_pkt_pointers(struct bpf_verifier_state *state, - struct bpf_reg_state *dst_reg) + struct bpf_reg_state *dst_reg, + bool range_right_open) { struct bpf_reg_state *regs = state->regs, *reg; + u16 new_range; int i; - if (dst_reg->off < 0) + if (dst_reg->off < 0 || + (dst_reg->off == 0 && range_right_open)) /* This doesn't give us any range */ return; @@ -2446,9 +2449,13 @@ static void find_good_pkt_pointers(struct bpf_verifier_state *state, */ return; - /* LLVM can generate four kind of checks: + new_range = dst_reg->off; + if (range_right_open) + new_range--; + + /* Examples for register markings: * - * Type 1/2: + * pkt_data in dst register: * * r2 = r3; * r2 += 8; @@ -2465,7 +2472,7 @@ static void find_good_pkt_pointers(struct bpf_verifier_state *state, * r2=pkt(id=n,off=8,r=0) * r3=pkt(id=n,off=0,r=0) * - * Type 3/4: + * pkt_data in src register: * * r2 = r3; * r2 += 8; @@ -2483,7 +2490,9 @@ static void find_good_pkt_pointers(struct bpf_verifier_state *state, * r3=pkt(id=n,off=0,r=0) * * Find register r3 and mark its range as r3=pkt(id=n,off=0,r=8) - * so that range of bytes [r3, r3 + 8) is safe to access. + * or r3=pkt(id=n,off=0,r=8-1), so that range of bytes [r3, r3 + 8) + * and [r3, r3 + 8-1) respectively is safe to access depending on + * the check. */ /* If our ids match, then we must have the same max_value. And we @@ -2494,14 +2503,14 @@ static void find_good_pkt_pointers(struct bpf_verifier_state *state, for (i = 0; i < MAX_BPF_REG; i++) if (regs[i].type == PTR_TO_PACKET && regs[i].id == dst_reg->id) /* keep the maximum range already checked */ - regs[i].range = max_t(u16, regs[i].range, dst_reg->off); + regs[i].range = max(regs[i].range, new_range); for (i = 0; i < MAX_BPF_STACK; i += BPF_REG_SIZE) { if (state->stack_slot_type[i] != STACK_SPILL) continue; reg = &state->spilled_regs[i / BPF_REG_SIZE]; if (reg->type == PTR_TO_PACKET && reg->id == dst_reg->id) - reg->range = max_t(u16, reg->range, dst_reg->off); + reg->range = max(reg->range, new_range); } } @@ -2865,19 +2874,19 @@ static int check_cond_jmp_op(struct bpf_verifier_env *env, } else if (BPF_SRC(insn->code) == BPF_X && opcode == BPF_JGT && dst_reg->type == PTR_TO_PACKET && regs[insn->src_reg].type == PTR_TO_PACKET_END) { - find_good_pkt_pointers(this_branch, dst_reg); + find_good_pkt_pointers(this_branch, dst_reg, false); } else if (BPF_SRC(insn->code) == BPF_X && opcode == BPF_JLT && dst_reg->type == PTR_TO_PACKET && regs[insn->src_reg].type == PTR_TO_PACKET_END) { - find_good_pkt_pointers(other_branch, dst_reg); + find_good_pkt_pointers(other_branch, dst_reg, true); } else if (BPF_SRC(insn->code) == BPF_X && opcode == BPF_JGE && dst_reg->type == PTR_TO_PACKET_END && regs[insn->src_reg].type == PTR_TO_PACKET) { - find_good_pkt_pointers(other_branch, ®s[insn->src_reg]); + find_good_pkt_pointers(other_branch, ®s[insn->src_reg], false); } else if (BPF_SRC(insn->code) == BPF_X && opcode == BPF_JLE && dst_reg->type == PTR_TO_PACKET_END && regs[insn->src_reg].type == PTR_TO_PACKET) { - find_good_pkt_pointers(this_branch, ®s[insn->src_reg]); + find_good_pkt_pointers(this_branch, ®s[insn->src_reg], true); } else if (is_pointer_value(env, insn->dst_reg)) { verbose("R%d pointer comparison prohibited\n", insn->dst_reg); return -EACCES; -- cgit v1.2.3-71-gd317 From 0fd4759c5515b7f2297d7fee5c45e5d9dd733001 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Sat, 21 Oct 2017 02:34:22 +0200 Subject: bpf: fix pattern matches for direct packet access Alexander had a test program with direct packet access, where the access test was in the form of data + X > data_end. In an unrelated change to the program LLVM decided to swap the branches and emitted code for the test in form of data + X <= data_end. We hadn't seen these being generated previously, thus verifier would reject the program. Therefore, fix up the verifier to detect all test cases, so we don't run into such issues in the future. Fixes: b4e432f1000a ("bpf: enable BPF_J{LT, LE, SLT, SLE} opcodes in verifier") Reported-by: Alexander Alemayhu Signed-off-by: Daniel Borkmann Acked-by: Alexei Starovoitov Acked-by: John Fastabend Signed-off-by: David S. Miller --- kernel/bpf/verifier.c | 24 ++++++++++++++++++++++++ 1 file changed, 24 insertions(+) (limited to 'kernel') diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 49cb5ad14746..c48ca2a34b5e 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -2874,18 +2874,42 @@ static int check_cond_jmp_op(struct bpf_verifier_env *env, } else if (BPF_SRC(insn->code) == BPF_X && opcode == BPF_JGT && dst_reg->type == PTR_TO_PACKET && regs[insn->src_reg].type == PTR_TO_PACKET_END) { + /* pkt_data' > pkt_end */ find_good_pkt_pointers(this_branch, dst_reg, false); + } else if (BPF_SRC(insn->code) == BPF_X && opcode == BPF_JGT && + dst_reg->type == PTR_TO_PACKET_END && + regs[insn->src_reg].type == PTR_TO_PACKET) { + /* pkt_end > pkt_data' */ + find_good_pkt_pointers(other_branch, ®s[insn->src_reg], true); } else if (BPF_SRC(insn->code) == BPF_X && opcode == BPF_JLT && dst_reg->type == PTR_TO_PACKET && regs[insn->src_reg].type == PTR_TO_PACKET_END) { + /* pkt_data' < pkt_end */ find_good_pkt_pointers(other_branch, dst_reg, true); + } else if (BPF_SRC(insn->code) == BPF_X && opcode == BPF_JLT && + dst_reg->type == PTR_TO_PACKET_END && + regs[insn->src_reg].type == PTR_TO_PACKET) { + /* pkt_end < pkt_data' */ + find_good_pkt_pointers(this_branch, ®s[insn->src_reg], false); + } else if (BPF_SRC(insn->code) == BPF_X && opcode == BPF_JGE && + dst_reg->type == PTR_TO_PACKET && + regs[insn->src_reg].type == PTR_TO_PACKET_END) { + /* pkt_data' >= pkt_end */ + find_good_pkt_pointers(this_branch, dst_reg, true); } else if (BPF_SRC(insn->code) == BPF_X && opcode == BPF_JGE && dst_reg->type == PTR_TO_PACKET_END && regs[insn->src_reg].type == PTR_TO_PACKET) { + /* pkt_end >= pkt_data' */ find_good_pkt_pointers(other_branch, ®s[insn->src_reg], false); + } else if (BPF_SRC(insn->code) == BPF_X && opcode == BPF_JLE && + dst_reg->type == PTR_TO_PACKET && + regs[insn->src_reg].type == PTR_TO_PACKET_END) { + /* pkt_data' <= pkt_end */ + find_good_pkt_pointers(other_branch, dst_reg, false); } else if (BPF_SRC(insn->code) == BPF_X && opcode == BPF_JLE && dst_reg->type == PTR_TO_PACKET_END && regs[insn->src_reg].type == PTR_TO_PACKET) { + /* pkt_end <= pkt_data' */ find_good_pkt_pointers(this_branch, ®s[insn->src_reg], true); } else if (is_pointer_value(env, insn->dst_reg)) { verbose("R%d pointer comparison prohibited\n", insn->dst_reg); -- cgit v1.2.3-71-gd317