From 9ce7bc036ae4cfe3393232c86e9e1fea2153c237 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 13 Jun 2018 10:11:56 -0700 Subject: netfilter: ipv6: nf_defrag: reduce struct net memory waste It is a waste of memory to use a full "struct netns_sysctl_ipv6" while only one pointer is really used, considering netns_sysctl_ipv6 keeps growing. Also, since "struct netns_frags" has cache line alignment, it is better to move the frags_hdr pointer outside, otherwise we spend a full cache line for this pointer. This saves 192 bytes of memory per netns. Fixes: c038a767cd69 ("ipv6: add a new namespace for nf_conntrack_reasm") Signed-off-by: Eric Dumazet Signed-off-by: Pablo Neira Ayuso --- include/net/net_namespace.h | 1 + include/net/netns/ipv6.h | 1 - 2 files changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/net/net_namespace.h b/include/net/net_namespace.h index 47e35cce3b64..a71264d75d7f 100644 --- a/include/net/net_namespace.h +++ b/include/net/net_namespace.h @@ -128,6 +128,7 @@ struct net { #endif #if IS_ENABLED(CONFIG_NF_DEFRAG_IPV6) struct netns_nf_frag nf_frag; + struct ctl_table_header *nf_frag_frags_hdr; #endif struct sock *nfnl; struct sock *nfnl_stash; diff --git a/include/net/netns/ipv6.h b/include/net/netns/ipv6.h index c978a31b0f84..762ac9931b62 100644 --- a/include/net/netns/ipv6.h +++ b/include/net/netns/ipv6.h @@ -109,7 +109,6 @@ struct netns_ipv6 { #if IS_ENABLED(CONFIG_NF_DEFRAG_IPV6) struct netns_nf_frag { - struct netns_sysctl_ipv6 sysctl; struct netns_frags frags; }; #endif -- cgit v1.2.3 From 6362f0a290023bafd7f991089e81dd9278f154b8 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Tue, 19 Jun 2018 10:12:48 -0600 Subject: libata: add command iterator helpers Now that we have the internal tag as a special (higher) value tag, it gets a bit tricky to iterate the internal commands as some loops will exceed ATA_MAX_QUEUE. Add explicit helpers for iterating pending commands, both inflight and internal. Signed-off-by: Jens Axboe Signed-off-by: Tejun Heo --- include/linux/libata.h | 23 +++++++++++++++++++++++ 1 file changed, 23 insertions(+) (limited to 'include') diff --git a/include/linux/libata.h b/include/linux/libata.h index 8b8946dd63b9..a2257e380789 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -1495,6 +1495,29 @@ static inline bool ata_tag_valid(unsigned int tag) return tag < ATA_MAX_QUEUE || ata_tag_internal(tag); } +#define __ata_qc_for_each(ap, qc, tag, max_tag, fn) \ + for ((tag) = 0; (tag) < (max_tag) && \ + ({ qc = fn((ap), (tag)); 1; }); (tag)++) \ + +/* + * Internal use only, iterate commands ignoring error handling and + * status of 'qc'. + */ +#define ata_qc_for_each_raw(ap, qc, tag) \ + __ata_qc_for_each(ap, qc, tag, ATA_MAX_QUEUE, __ata_qc_from_tag) + +/* + * Iterate all potential commands that can be queued + */ +#define ata_qc_for_each(ap, qc, tag) \ + __ata_qc_for_each(ap, qc, tag, ATA_MAX_QUEUE, ata_qc_from_tag) + +/* + * Like ata_qc_for_each, but with the internal tag included + */ +#define ata_qc_for_each_with_internal(ap, qc, tag) \ + __ata_qc_for_each(ap, qc, tag, ATA_MAX_QUEUE + 1, ata_qc_from_tag) + /* * device helpers */ -- cgit v1.2.3 From 8f732850df1b2b4d8d719f7e606dfb3050e7ea11 Mon Sep 17 00:00:00 2001 From: Benjamin Tissoires Date: Thu, 31 May 2018 13:49:29 +0200 Subject: HID: core: allow concurrent registration of drivers Detected on the Dell XPS 9365. The laptop has 2 devices that benefit from the hid-generic auto-unbinding. When those 2 devices are presented to the userspace, udev loads both wacom and hid-multitouch. When this happens, the code in __hid_bus_reprobe_drivers() is called concurrently and the second device gets reprobed twice. An other bug in the power_supply subsystem prevent to remove the wacom driver if it just finished its initialization, which basically kills the wacom node. [jkosina@suse.cz: reformat changelog a bit] Fixes c17a7476e4c4 ("HID: core: rewrite the hid-generic automatic unbind") Cc: stable@vger.kernel.org # v4.17 Tested-by: Mario Limonciello Signed-off-by: Benjamin Tissoires Signed-off-by: Jiri Kosina --- drivers/hid/hid-core.c | 5 ++++- include/linux/hid.h | 3 ++- 2 files changed, 6 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/drivers/hid/hid-core.c b/drivers/hid/hid-core.c index 355dc7e49562..a460ec147aee 100644 --- a/drivers/hid/hid-core.c +++ b/drivers/hid/hid-core.c @@ -1949,6 +1949,8 @@ static int hid_device_probe(struct device *dev) } hdev->io_started = false; + clear_bit(ffs(HID_STAT_REPROBED), &hdev->status); + if (!hdev->driver) { id = hid_match_device(hdev, hdrv); if (id == NULL) { @@ -2212,7 +2214,8 @@ static int __hid_bus_reprobe_drivers(struct device *dev, void *data) struct hid_device *hdev = to_hid_device(dev); if (hdev->driver == hdrv && - !hdrv->match(hdev, hid_ignore_special_drivers)) + !hdrv->match(hdev, hid_ignore_special_drivers) && + !test_and_set_bit(ffs(HID_STAT_REPROBED), &hdev->status)) return device_reprobe(dev); return 0; diff --git a/include/linux/hid.h b/include/linux/hid.h index 41a3d5775394..773bcb1d4044 100644 --- a/include/linux/hid.h +++ b/include/linux/hid.h @@ -511,6 +511,7 @@ struct hid_output_fifo { #define HID_STAT_ADDED BIT(0) #define HID_STAT_PARSED BIT(1) #define HID_STAT_DUP_DETECTED BIT(2) +#define HID_STAT_REPROBED BIT(3) struct hid_input { struct list_head list; @@ -579,7 +580,7 @@ struct hid_device { /* device report descriptor */ bool battery_avoid_query; #endif - unsigned int status; /* see STAT flags above */ + unsigned long status; /* see STAT flags above */ unsigned claimed; /* Claimed by hidinput, hiddev? */ unsigned quirks; /* Various quirks the device can pull on us */ bool io_started; /* If IO has started */ -- cgit v1.2.3 From fdb5c4531c1e0e50e609df83f736b6f3a02896e2 Mon Sep 17 00:00:00 2001 From: Sean Young Date: Tue, 19 Jun 2018 00:04:24 +0100 Subject: bpf: fix attach type BPF_LIRC_MODE2 dependency wrt CONFIG_CGROUP_BPF If the kernel is compiled with CONFIG_CGROUP_BPF not enabled, it is not possible to attach, detach or query IR BPF programs to /dev/lircN devices, making them impossible to use. For embedded devices, it should be possible to use IR decoding without cgroups or CONFIG_CGROUP_BPF enabled. This change requires some refactoring, since bpf_prog_{attach,detach,query} functions are now always compiled, but their code paths for cgroups need moving out. Rather than a #ifdef CONFIG_CGROUP_BPF in kernel/bpf/syscall.c, moving them to kernel/bpf/cgroup.c and kernel/bpf/sockmap.c does not require #ifdefs since that is already conditionally compiled. Fixes: f4364dcfc86d ("media: rc: introduce BPF_PROG_LIRC_MODE2") Signed-off-by: Sean Young Signed-off-by: Daniel Borkmann --- drivers/media/rc/bpf-lirc.c | 14 +------ include/linux/bpf-cgroup.h | 26 ++++++++++++ include/linux/bpf.h | 8 ++++ include/linux/bpf_lirc.h | 5 ++- kernel/bpf/cgroup.c | 54 +++++++++++++++++++++++++ kernel/bpf/sockmap.c | 18 +++++++++ kernel/bpf/syscall.c | 99 ++++++++++----------------------------------- 7 files changed, 132 insertions(+), 92 deletions(-) (limited to 'include') diff --git a/drivers/media/rc/bpf-lirc.c b/drivers/media/rc/bpf-lirc.c index 40826bba06b6..fcfab6635f9c 100644 --- a/drivers/media/rc/bpf-lirc.c +++ b/drivers/media/rc/bpf-lirc.c @@ -207,29 +207,19 @@ void lirc_bpf_free(struct rc_dev *rcdev) bpf_prog_array_free(rcdev->raw->progs); } -int lirc_prog_attach(const union bpf_attr *attr) +int lirc_prog_attach(const union bpf_attr *attr, struct bpf_prog *prog) { - struct bpf_prog *prog; struct rc_dev *rcdev; int ret; if (attr->attach_flags) return -EINVAL; - prog = bpf_prog_get_type(attr->attach_bpf_fd, - BPF_PROG_TYPE_LIRC_MODE2); - if (IS_ERR(prog)) - return PTR_ERR(prog); - rcdev = rc_dev_get_from_fd(attr->target_fd); - if (IS_ERR(rcdev)) { - bpf_prog_put(prog); + if (IS_ERR(rcdev)) return PTR_ERR(rcdev); - } ret = lirc_bpf_attach(rcdev, prog); - if (ret) - bpf_prog_put(prog); put_device(&rcdev->dev); diff --git a/include/linux/bpf-cgroup.h b/include/linux/bpf-cgroup.h index 975fb4cf1bb7..79795c5fa7c3 100644 --- a/include/linux/bpf-cgroup.h +++ b/include/linux/bpf-cgroup.h @@ -188,12 +188,38 @@ int __cgroup_bpf_check_dev_permission(short dev_type, u32 major, u32 minor, \ __ret; \ }) +int cgroup_bpf_prog_attach(const union bpf_attr *attr, + enum bpf_prog_type ptype, struct bpf_prog *prog); +int cgroup_bpf_prog_detach(const union bpf_attr *attr, + enum bpf_prog_type ptype); +int cgroup_bpf_prog_query(const union bpf_attr *attr, + union bpf_attr __user *uattr); #else +struct bpf_prog; struct cgroup_bpf {}; static inline void cgroup_bpf_put(struct cgroup *cgrp) {} static inline int cgroup_bpf_inherit(struct cgroup *cgrp) { return 0; } +static inline int cgroup_bpf_prog_attach(const union bpf_attr *attr, + enum bpf_prog_type ptype, + struct bpf_prog *prog) +{ + return -EINVAL; +} + +static inline int cgroup_bpf_prog_detach(const union bpf_attr *attr, + enum bpf_prog_type ptype) +{ + return -EINVAL; +} + +static inline int cgroup_bpf_prog_query(const union bpf_attr *attr, + union bpf_attr __user *uattr) +{ + return -EINVAL; +} + #define cgroup_bpf_enabled (0) #define BPF_CGROUP_PRE_CONNECT_ENABLED(sk) (0) #define BPF_CGROUP_RUN_PROG_INET_INGRESS(sk,skb) ({ 0; }) diff --git a/include/linux/bpf.h b/include/linux/bpf.h index 7df32a3200f7..8827e797ff97 100644 --- a/include/linux/bpf.h +++ b/include/linux/bpf.h @@ -696,6 +696,8 @@ static inline void bpf_map_offload_map_free(struct bpf_map *map) struct sock *__sock_map_lookup_elem(struct bpf_map *map, u32 key); struct sock *__sock_hash_lookup_elem(struct bpf_map *map, void *key); int sock_map_prog(struct bpf_map *map, struct bpf_prog *prog, u32 type); +int sockmap_get_from_fd(const union bpf_attr *attr, int type, + struct bpf_prog *prog); #else static inline struct sock *__sock_map_lookup_elem(struct bpf_map *map, u32 key) { @@ -714,6 +716,12 @@ static inline int sock_map_prog(struct bpf_map *map, { return -EOPNOTSUPP; } + +static inline int sockmap_get_from_fd(const union bpf_attr *attr, int type, + struct bpf_prog *prog) +{ + return -EINVAL; +} #endif #if defined(CONFIG_XDP_SOCKETS) diff --git a/include/linux/bpf_lirc.h b/include/linux/bpf_lirc.h index 5f8a4283092d..9d9ff755ec29 100644 --- a/include/linux/bpf_lirc.h +++ b/include/linux/bpf_lirc.h @@ -5,11 +5,12 @@ #include #ifdef CONFIG_BPF_LIRC_MODE2 -int lirc_prog_attach(const union bpf_attr *attr); +int lirc_prog_attach(const union bpf_attr *attr, struct bpf_prog *prog); int lirc_prog_detach(const union bpf_attr *attr); int lirc_prog_query(const union bpf_attr *attr, union bpf_attr __user *uattr); #else -static inline int lirc_prog_attach(const union bpf_attr *attr) +static inline int lirc_prog_attach(const union bpf_attr *attr, + struct bpf_prog *prog) { return -EINVAL; } diff --git a/kernel/bpf/cgroup.c b/kernel/bpf/cgroup.c index f7c00bd6f8e4..3d83ee7df381 100644 --- a/kernel/bpf/cgroup.c +++ b/kernel/bpf/cgroup.c @@ -428,6 +428,60 @@ int __cgroup_bpf_query(struct cgroup *cgrp, const union bpf_attr *attr, return ret; } +int cgroup_bpf_prog_attach(const union bpf_attr *attr, + enum bpf_prog_type ptype, struct bpf_prog *prog) +{ + struct cgroup *cgrp; + int ret; + + cgrp = cgroup_get_from_fd(attr->target_fd); + if (IS_ERR(cgrp)) + return PTR_ERR(cgrp); + + ret = cgroup_bpf_attach(cgrp, prog, attr->attach_type, + attr->attach_flags); + cgroup_put(cgrp); + return ret; +} + +int cgroup_bpf_prog_detach(const union bpf_attr *attr, enum bpf_prog_type ptype) +{ + struct bpf_prog *prog; + struct cgroup *cgrp; + int ret; + + cgrp = cgroup_get_from_fd(attr->target_fd); + if (IS_ERR(cgrp)) + return PTR_ERR(cgrp); + + prog = bpf_prog_get_type(attr->attach_bpf_fd, ptype); + if (IS_ERR(prog)) + prog = NULL; + + ret = cgroup_bpf_detach(cgrp, prog, attr->attach_type, 0); + if (prog) + bpf_prog_put(prog); + + cgroup_put(cgrp); + return ret; +} + +int cgroup_bpf_prog_query(const union bpf_attr *attr, + union bpf_attr __user *uattr) +{ + struct cgroup *cgrp; + int ret; + + cgrp = cgroup_get_from_fd(attr->query.target_fd); + if (IS_ERR(cgrp)) + return PTR_ERR(cgrp); + + ret = cgroup_bpf_query(cgrp, attr, uattr); + + cgroup_put(cgrp); + return ret; +} + /** * __cgroup_bpf_run_filter_skb() - Run a program for packet filtering * @sk: The socket sending or receiving traffic diff --git a/kernel/bpf/sockmap.c b/kernel/bpf/sockmap.c index 52a91d816c0e..81d0c55a77aa 100644 --- a/kernel/bpf/sockmap.c +++ b/kernel/bpf/sockmap.c @@ -1915,6 +1915,24 @@ int sock_map_prog(struct bpf_map *map, struct bpf_prog *prog, u32 type) return 0; } +int sockmap_get_from_fd(const union bpf_attr *attr, int type, + struct bpf_prog *prog) +{ + int ufd = attr->target_fd; + struct bpf_map *map; + struct fd f; + int err; + + f = fdget(ufd); + map = __bpf_map_get(f); + if (IS_ERR(map)) + return PTR_ERR(map); + + err = sock_map_prog(map, prog, attr->attach_type); + fdput(f); + return err; +} + static void *sock_map_lookup(struct bpf_map *map, void *key) { return NULL; diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c index 35dc466641f2..d10ecd78105f 100644 --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -1483,8 +1483,6 @@ out_free_tp: return err; } -#ifdef CONFIG_CGROUP_BPF - static int bpf_prog_attach_check_attach_type(const struct bpf_prog *prog, enum bpf_attach_type attach_type) { @@ -1499,40 +1497,6 @@ static int bpf_prog_attach_check_attach_type(const struct bpf_prog *prog, #define BPF_PROG_ATTACH_LAST_FIELD attach_flags -static int sockmap_get_from_fd(const union bpf_attr *attr, - int type, bool attach) -{ - struct bpf_prog *prog = NULL; - int ufd = attr->target_fd; - struct bpf_map *map; - struct fd f; - int err; - - f = fdget(ufd); - map = __bpf_map_get(f); - if (IS_ERR(map)) - return PTR_ERR(map); - - if (attach) { - prog = bpf_prog_get_type(attr->attach_bpf_fd, type); - if (IS_ERR(prog)) { - fdput(f); - return PTR_ERR(prog); - } - } - - err = sock_map_prog(map, prog, attr->attach_type); - if (err) { - fdput(f); - if (prog) - bpf_prog_put(prog); - return err; - } - - fdput(f); - return 0; -} - #define BPF_F_ATTACH_MASK \ (BPF_F_ALLOW_OVERRIDE | BPF_F_ALLOW_MULTI) @@ -1540,7 +1504,6 @@ static int bpf_prog_attach(const union bpf_attr *attr) { enum bpf_prog_type ptype; struct bpf_prog *prog; - struct cgroup *cgrp; int ret; if (!capable(CAP_NET_ADMIN)) @@ -1577,12 +1540,15 @@ static int bpf_prog_attach(const union bpf_attr *attr) ptype = BPF_PROG_TYPE_CGROUP_DEVICE; break; case BPF_SK_MSG_VERDICT: - return sockmap_get_from_fd(attr, BPF_PROG_TYPE_SK_MSG, true); + ptype = BPF_PROG_TYPE_SK_MSG; + break; case BPF_SK_SKB_STREAM_PARSER: case BPF_SK_SKB_STREAM_VERDICT: - return sockmap_get_from_fd(attr, BPF_PROG_TYPE_SK_SKB, true); + ptype = BPF_PROG_TYPE_SK_SKB; + break; case BPF_LIRC_MODE2: - return lirc_prog_attach(attr); + ptype = BPF_PROG_TYPE_LIRC_MODE2; + break; default: return -EINVAL; } @@ -1596,18 +1562,20 @@ static int bpf_prog_attach(const union bpf_attr *attr) return -EINVAL; } - cgrp = cgroup_get_from_fd(attr->target_fd); - if (IS_ERR(cgrp)) { - bpf_prog_put(prog); - return PTR_ERR(cgrp); + switch (ptype) { + case BPF_PROG_TYPE_SK_SKB: + case BPF_PROG_TYPE_SK_MSG: + ret = sockmap_get_from_fd(attr, ptype, prog); + break; + case BPF_PROG_TYPE_LIRC_MODE2: + ret = lirc_prog_attach(attr, prog); + break; + default: + ret = cgroup_bpf_prog_attach(attr, ptype, prog); } - ret = cgroup_bpf_attach(cgrp, prog, attr->attach_type, - attr->attach_flags); if (ret) bpf_prog_put(prog); - cgroup_put(cgrp); - return ret; } @@ -1616,9 +1584,6 @@ static int bpf_prog_attach(const union bpf_attr *attr) static int bpf_prog_detach(const union bpf_attr *attr) { enum bpf_prog_type ptype; - struct bpf_prog *prog; - struct cgroup *cgrp; - int ret; if (!capable(CAP_NET_ADMIN)) return -EPERM; @@ -1651,29 +1616,17 @@ static int bpf_prog_detach(const union bpf_attr *attr) ptype = BPF_PROG_TYPE_CGROUP_DEVICE; break; case BPF_SK_MSG_VERDICT: - return sockmap_get_from_fd(attr, BPF_PROG_TYPE_SK_MSG, false); + return sockmap_get_from_fd(attr, BPF_PROG_TYPE_SK_MSG, NULL); case BPF_SK_SKB_STREAM_PARSER: case BPF_SK_SKB_STREAM_VERDICT: - return sockmap_get_from_fd(attr, BPF_PROG_TYPE_SK_SKB, false); + return sockmap_get_from_fd(attr, BPF_PROG_TYPE_SK_SKB, NULL); case BPF_LIRC_MODE2: return lirc_prog_detach(attr); default: return -EINVAL; } - cgrp = cgroup_get_from_fd(attr->target_fd); - if (IS_ERR(cgrp)) - return PTR_ERR(cgrp); - - prog = bpf_prog_get_type(attr->attach_bpf_fd, ptype); - if (IS_ERR(prog)) - prog = NULL; - - ret = cgroup_bpf_detach(cgrp, prog, attr->attach_type, 0); - if (prog) - bpf_prog_put(prog); - cgroup_put(cgrp); - return ret; + return cgroup_bpf_prog_detach(attr, ptype); } #define BPF_PROG_QUERY_LAST_FIELD query.prog_cnt @@ -1681,9 +1634,6 @@ static int bpf_prog_detach(const union bpf_attr *attr) static int bpf_prog_query(const union bpf_attr *attr, union bpf_attr __user *uattr) { - struct cgroup *cgrp; - int ret; - if (!capable(CAP_NET_ADMIN)) return -EPERM; if (CHECK_ATTR(BPF_PROG_QUERY)) @@ -1711,14 +1661,9 @@ static int bpf_prog_query(const union bpf_attr *attr, default: return -EINVAL; } - cgrp = cgroup_get_from_fd(attr->query.target_fd); - if (IS_ERR(cgrp)) - return PTR_ERR(cgrp); - ret = cgroup_bpf_query(cgrp, attr, uattr); - cgroup_put(cgrp); - return ret; + + return cgroup_bpf_prog_query(attr, uattr); } -#endif /* CONFIG_CGROUP_BPF */ #define BPF_PROG_TEST_RUN_LAST_FIELD test.duration @@ -2365,7 +2310,6 @@ SYSCALL_DEFINE3(bpf, int, cmd, union bpf_attr __user *, uattr, unsigned int, siz case BPF_OBJ_GET: err = bpf_obj_get(&attr); break; -#ifdef CONFIG_CGROUP_BPF case BPF_PROG_ATTACH: err = bpf_prog_attach(&attr); break; @@ -2375,7 +2319,6 @@ SYSCALL_DEFINE3(bpf, int, cmd, union bpf_attr __user *, uattr, unsigned int, siz case BPF_PROG_QUERY: err = bpf_prog_query(&attr, uattr); break; -#endif case BPF_PROG_TEST_RUN: err = bpf_prog_test_run(&attr, uattr); break; -- cgit v1.2.3 From 0efc8562491b7d36f6bbc4fbc8f3348cb6641e9c Mon Sep 17 00:00:00 2001 From: Or Gerlitz Date: Thu, 31 May 2018 11:16:18 +0300 Subject: net/mlx5: E-Switch, Avoid setup attempt if not being e-switch manager In smartnic env, the host (PF) driver might not be an e-switch manager, hence the FW will err on driver attempts to deal with setting/unsetting the eswitch and as a result the overall setup of sriov will fail. Fix that by avoiding the operation if e-switch management is not allowed for this driver instance. While here, move to use the correct name for the esw manager capability name. Fixes: 81848731ff40 ('net/mlx5: E-Switch, Add SR-IOV (FDB) support') Signed-off-by: Or Gerlitz Reported-by: Guy Kushnir Reviewed-by: Eli Cohen Tested-by: Eli Cohen Signed-off-by: Saeed Mahameed --- drivers/net/ethernet/mellanox/mlx5/core/en_rep.c | 2 +- drivers/net/ethernet/mellanox/mlx5/core/eswitch.c | 2 +- drivers/net/ethernet/mellanox/mlx5/core/fs_core.c | 3 ++- drivers/net/ethernet/mellanox/mlx5/core/fw.c | 5 +++-- drivers/net/ethernet/mellanox/mlx5/core/sriov.c | 7 ++++++- include/linux/mlx5/eswitch.h | 2 ++ include/linux/mlx5/mlx5_ifc.h | 2 +- 7 files changed, 16 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c index 378ad74518ec..60236f73373c 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c @@ -839,7 +839,7 @@ static bool mlx5e_is_vf_vport_rep(struct mlx5e_priv *priv) struct mlx5e_rep_priv *rpriv = priv->ppriv; struct mlx5_eswitch_rep *rep; - if (!MLX5_CAP_GEN(priv->mdev, eswitch_flow_table)) + if (!MLX5_ESWITCH_MANAGER(priv->mdev)) return false; rep = rpriv->rep; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c index f63dfbcd29fe..103fd6a0cc65 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c @@ -1604,7 +1604,7 @@ int mlx5_eswitch_enable_sriov(struct mlx5_eswitch *esw, int nvfs, int mode) if (!ESW_ALLOWED(esw)) return 0; - if (!MLX5_CAP_GEN(esw->dev, eswitch_flow_table) || + if (!MLX5_ESWITCH_MANAGER(esw->dev) || !MLX5_CAP_ESW_FLOWTABLE_FDB(esw->dev, ft_support)) { esw_warn(esw->dev, "E-Switch FDB is not supported, aborting ...\n"); return -EOPNOTSUPP; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c index 49a75d31185e..f1a86cea86a0 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c @@ -32,6 +32,7 @@ #include #include +#include #include "mlx5_core.h" #include "fs_core.h" @@ -2652,7 +2653,7 @@ int mlx5_init_fs(struct mlx5_core_dev *dev) goto err; } - if (MLX5_CAP_GEN(dev, eswitch_flow_table)) { + if (MLX5_ESWITCH_MANAGER(dev)) { if (MLX5_CAP_ESW_FLOWTABLE_FDB(dev, ft_support)) { err = init_fdb_root_ns(steering); if (err) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fw.c b/drivers/net/ethernet/mellanox/mlx5/core/fw.c index afd9f4fa22f4..41ad24f0de2c 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fw.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fw.c @@ -32,6 +32,7 @@ #include #include +#include #include #include "mlx5_core.h" #include "../../mlxfw/mlxfw.h" @@ -159,13 +160,13 @@ int mlx5_query_hca_caps(struct mlx5_core_dev *dev) } if (MLX5_CAP_GEN(dev, vport_group_manager) && - MLX5_CAP_GEN(dev, eswitch_flow_table)) { + MLX5_ESWITCH_MANAGER(dev)) { err = mlx5_core_get_caps(dev, MLX5_CAP_ESWITCH_FLOW_TABLE); if (err) return err; } - if (MLX5_CAP_GEN(dev, eswitch_flow_table)) { + if (MLX5_ESWITCH_MANAGER(dev)) { err = mlx5_core_get_caps(dev, MLX5_CAP_ESWITCH); if (err) return err; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/sriov.c b/drivers/net/ethernet/mellanox/mlx5/core/sriov.c index 2a8b529ce6dd..a0674962f02c 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/sriov.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/sriov.c @@ -88,6 +88,9 @@ static int mlx5_device_enable_sriov(struct mlx5_core_dev *dev, int num_vfs) return -EBUSY; } + if (!MLX5_ESWITCH_MANAGER(dev)) + goto enable_vfs_hca; + err = mlx5_eswitch_enable_sriov(dev->priv.eswitch, num_vfs, SRIOV_LEGACY); if (err) { mlx5_core_warn(dev, @@ -95,6 +98,7 @@ static int mlx5_device_enable_sriov(struct mlx5_core_dev *dev, int num_vfs) return err; } +enable_vfs_hca: for (vf = 0; vf < num_vfs; vf++) { err = mlx5_core_enable_hca(dev, vf + 1); if (err) { @@ -140,7 +144,8 @@ static void mlx5_device_disable_sriov(struct mlx5_core_dev *dev) } out: - mlx5_eswitch_disable_sriov(dev->priv.eswitch); + if (MLX5_ESWITCH_MANAGER(dev)) + mlx5_eswitch_disable_sriov(dev->priv.eswitch); if (mlx5_wait_for_vf_pages(dev)) mlx5_core_warn(dev, "timeout reclaiming VFs pages\n"); diff --git a/include/linux/mlx5/eswitch.h b/include/linux/mlx5/eswitch.h index d3c9db492b30..fab5121ffb8f 100644 --- a/include/linux/mlx5/eswitch.h +++ b/include/linux/mlx5/eswitch.h @@ -8,6 +8,8 @@ #include +#define MLX5_ESWITCH_MANAGER(mdev) MLX5_CAP_GEN(mdev, eswitch_manager) + enum { SRIOV_NONE, SRIOV_LEGACY, diff --git a/include/linux/mlx5/mlx5_ifc.h b/include/linux/mlx5/mlx5_ifc.h index 27134c4fcb76..ac281f5ec9b8 100644 --- a/include/linux/mlx5/mlx5_ifc.h +++ b/include/linux/mlx5/mlx5_ifc.h @@ -922,7 +922,7 @@ struct mlx5_ifc_cmd_hca_cap_bits { u8 vnic_env_queue_counters[0x1]; u8 ets[0x1]; u8 nic_flow_table[0x1]; - u8 eswitch_flow_table[0x1]; + u8 eswitch_manager[0x1]; u8 device_memory[0x1]; u8 mcam_reg[0x1]; u8 pcam_reg[0x1]; -- cgit v1.2.3 From 951a8ee6def39e25d0e60b9394e5a249ba8b2390 Mon Sep 17 00:00:00 2001 From: John Hurley Date: Mon, 25 Jun 2018 20:36:28 -0700 Subject: nfp: reject binding to shared blocks TC shared blocks allow multiple qdiscs to be grouped together and filters shared between them. Currently the chains of filters attached to a block are only flushed when the block is removed. If a qdisc is removed from a block but the block still exists, flow del messages are not passed to the callback registered for that qdisc. For the NFP, this presents the possibility of rules still existing in hw when they should be removed. Prevent binding to shared blocks until the kernel can send per qdisc del messages when block unbinds occur. tcf_block_shared() was not used outside of the core until now, so also add an empty implementation for builds with CONFIG_NET_CLS=n. Fixes: 4861738775d7 ("net: sched: introduce shared filter blocks infrastructure") Signed-off-by: John Hurley Signed-off-by: Jakub Kicinski Reviewed-by: Simon Horman Signed-off-by: David S. Miller --- drivers/net/ethernet/netronome/nfp/bpf/main.c | 3 +++ drivers/net/ethernet/netronome/nfp/flower/offload.c | 3 +++ include/net/pkt_cls.h | 5 +++++ 3 files changed, 11 insertions(+) (limited to 'include') diff --git a/drivers/net/ethernet/netronome/nfp/bpf/main.c b/drivers/net/ethernet/netronome/nfp/bpf/main.c index fcdfb8e7fdea..6b15e3b11956 100644 --- a/drivers/net/ethernet/netronome/nfp/bpf/main.c +++ b/drivers/net/ethernet/netronome/nfp/bpf/main.c @@ -202,6 +202,9 @@ static int nfp_bpf_setup_tc_block(struct net_device *netdev, if (f->binder_type != TCF_BLOCK_BINDER_TYPE_CLSACT_INGRESS) return -EOPNOTSUPP; + if (tcf_block_shared(f->block)) + return -EOPNOTSUPP; + switch (f->command) { case TC_BLOCK_BIND: return tcf_block_cb_register(f->block, diff --git a/drivers/net/ethernet/netronome/nfp/flower/offload.c b/drivers/net/ethernet/netronome/nfp/flower/offload.c index 477f584f6d28..525057bee0ed 100644 --- a/drivers/net/ethernet/netronome/nfp/flower/offload.c +++ b/drivers/net/ethernet/netronome/nfp/flower/offload.c @@ -631,6 +631,9 @@ static int nfp_flower_setup_tc_block(struct net_device *netdev, if (f->binder_type != TCF_BLOCK_BINDER_TYPE_CLSACT_INGRESS) return -EOPNOTSUPP; + if (tcf_block_shared(f->block)) + return -EOPNOTSUPP; + switch (f->command) { case TC_BLOCK_BIND: return tcf_block_cb_register(f->block, diff --git a/include/net/pkt_cls.h b/include/net/pkt_cls.h index a3c1a2c47cd4..20b059574e60 100644 --- a/include/net/pkt_cls.h +++ b/include/net/pkt_cls.h @@ -111,6 +111,11 @@ void tcf_block_put_ext(struct tcf_block *block, struct Qdisc *q, { } +static inline bool tcf_block_shared(struct tcf_block *block) +{ + return false; +} + static inline struct Qdisc *tcf_block_q(struct tcf_block *block) { return NULL; -- cgit v1.2.3 From 4c79579b44b1876444f4d04de31c1a37098a0350 Mon Sep 17 00:00:00 2001 From: David Ahern Date: Tue, 26 Jun 2018 16:21:18 -0700 Subject: bpf: Change bpf_fib_lookup to return lookup status For ACLs implemented using either FIB rules or FIB entries, the BPF program needs the FIB lookup status to be able to drop the packet. Since the bpf_fib_lookup API has not reached a released kernel yet, change the return code to contain an encoding of the FIB lookup result and return the nexthop device index in the params struct. In addition, inform the BPF program of any post FIB lookup reason as to why the packet needs to go up the stack. The fib result for unicast routes must have an egress device, so remove the check that it is non-NULL. Signed-off-by: David Ahern Signed-off-by: Daniel Borkmann --- include/uapi/linux/bpf.h | 28 ++++++++++++--- net/core/filter.c | 86 +++++++++++++++++++++++++++++----------------- samples/bpf/xdp_fwd_kern.c | 8 ++--- 3 files changed, 81 insertions(+), 41 deletions(-) (limited to 'include') diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index 59b19b6a40d7..b7db3261c62d 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -1857,7 +1857,8 @@ union bpf_attr { * is resolved), the nexthop address is returned in ipv4_dst * or ipv6_dst based on family, smac is set to mac address of * egress device, dmac is set to nexthop mac address, rt_metric - * is set to metric from route (IPv4/IPv6 only). + * is set to metric from route (IPv4/IPv6 only), and ifindex + * is set to the device index of the nexthop from the FIB lookup. * * *plen* argument is the size of the passed in struct. * *flags* argument can be a combination of one or more of the @@ -1873,9 +1874,10 @@ union bpf_attr { * *ctx* is either **struct xdp_md** for XDP programs or * **struct sk_buff** tc cls_act programs. * Return - * Egress device index on success, 0 if packet needs to continue - * up the stack for further processing or a negative error in case - * of failure. + * * < 0 if any input argument is invalid + * * 0 on success (packet is forwarded, nexthop neighbor exists) + * * > 0 one of **BPF_FIB_LKUP_RET_** codes explaining why the + * * packet is not forwarded or needs assist from full stack * * int bpf_sock_hash_update(struct bpf_sock_ops_kern *skops, struct bpf_map *map, void *key, u64 flags) * Description @@ -2612,6 +2614,18 @@ struct bpf_raw_tracepoint_args { #define BPF_FIB_LOOKUP_DIRECT BIT(0) #define BPF_FIB_LOOKUP_OUTPUT BIT(1) +enum { + BPF_FIB_LKUP_RET_SUCCESS, /* lookup successful */ + BPF_FIB_LKUP_RET_BLACKHOLE, /* dest is blackholed; can be dropped */ + BPF_FIB_LKUP_RET_UNREACHABLE, /* dest is unreachable; can be dropped */ + BPF_FIB_LKUP_RET_PROHIBIT, /* dest not allowed; can be dropped */ + BPF_FIB_LKUP_RET_NOT_FWDED, /* packet is not forwarded */ + BPF_FIB_LKUP_RET_FWD_DISABLED, /* fwding is not enabled on ingress */ + BPF_FIB_LKUP_RET_UNSUPP_LWT, /* fwd requires encapsulation */ + BPF_FIB_LKUP_RET_NO_NEIGH, /* no neighbor entry for nh */ + BPF_FIB_LKUP_RET_FRAG_NEEDED, /* fragmentation required to fwd */ +}; + struct bpf_fib_lookup { /* input: network family for lookup (AF_INET, AF_INET6) * output: network family of egress nexthop @@ -2625,7 +2639,11 @@ struct bpf_fib_lookup { /* total length of packet from network header - used for MTU check */ __u16 tot_len; - __u32 ifindex; /* L3 device index for lookup */ + + /* input: L3 device index for lookup + * output: device index from FIB lookup + */ + __u32 ifindex; union { /* inputs to lookup */ diff --git a/net/core/filter.c b/net/core/filter.c index e7f12e9f598c..0ca6907d7efe 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -4073,8 +4073,9 @@ static int bpf_fib_set_fwd_params(struct bpf_fib_lookup *params, memcpy(params->smac, dev->dev_addr, ETH_ALEN); params->h_vlan_TCI = 0; params->h_vlan_proto = 0; + params->ifindex = dev->ifindex; - return dev->ifindex; + return 0; } #endif @@ -4098,7 +4099,7 @@ static int bpf_ipv4_fib_lookup(struct net *net, struct bpf_fib_lookup *params, /* verify forwarding is enabled on this interface */ in_dev = __in_dev_get_rcu(dev); if (unlikely(!in_dev || !IN_DEV_FORWARD(in_dev))) - return 0; + return BPF_FIB_LKUP_RET_FWD_DISABLED; if (flags & BPF_FIB_LOOKUP_OUTPUT) { fl4.flowi4_iif = 1; @@ -4123,7 +4124,7 @@ static int bpf_ipv4_fib_lookup(struct net *net, struct bpf_fib_lookup *params, tb = fib_get_table(net, tbid); if (unlikely(!tb)) - return 0; + return BPF_FIB_LKUP_RET_NOT_FWDED; err = fib_table_lookup(tb, &fl4, &res, FIB_LOOKUP_NOREF); } else { @@ -4135,8 +4136,20 @@ static int bpf_ipv4_fib_lookup(struct net *net, struct bpf_fib_lookup *params, err = fib_lookup(net, &fl4, &res, FIB_LOOKUP_NOREF); } - if (err || res.type != RTN_UNICAST) - return 0; + if (err) { + /* map fib lookup errors to RTN_ type */ + if (err == -EINVAL) + return BPF_FIB_LKUP_RET_BLACKHOLE; + if (err == -EHOSTUNREACH) + return BPF_FIB_LKUP_RET_UNREACHABLE; + if (err == -EACCES) + return BPF_FIB_LKUP_RET_PROHIBIT; + + return BPF_FIB_LKUP_RET_NOT_FWDED; + } + + if (res.type != RTN_UNICAST) + return BPF_FIB_LKUP_RET_NOT_FWDED; if (res.fi->fib_nhs > 1) fib_select_path(net, &res, &fl4, NULL); @@ -4144,19 +4157,16 @@ static int bpf_ipv4_fib_lookup(struct net *net, struct bpf_fib_lookup *params, if (check_mtu) { mtu = ip_mtu_from_fib_result(&res, params->ipv4_dst); if (params->tot_len > mtu) - return 0; + return BPF_FIB_LKUP_RET_FRAG_NEEDED; } nh = &res.fi->fib_nh[res.nh_sel]; /* do not handle lwt encaps right now */ if (nh->nh_lwtstate) - return 0; + return BPF_FIB_LKUP_RET_UNSUPP_LWT; dev = nh->nh_dev; - if (unlikely(!dev)) - return 0; - if (nh->nh_gw) params->ipv4_dst = nh->nh_gw; @@ -4166,10 +4176,10 @@ static int bpf_ipv4_fib_lookup(struct net *net, struct bpf_fib_lookup *params, * rcu_read_lock_bh is not needed here */ neigh = __ipv4_neigh_lookup_noref(dev, (__force u32)params->ipv4_dst); - if (neigh) - return bpf_fib_set_fwd_params(params, neigh, dev); + if (!neigh) + return BPF_FIB_LKUP_RET_NO_NEIGH; - return 0; + return bpf_fib_set_fwd_params(params, neigh, dev); } #endif @@ -4190,7 +4200,7 @@ static int bpf_ipv6_fib_lookup(struct net *net, struct bpf_fib_lookup *params, /* link local addresses are never forwarded */ if (rt6_need_strict(dst) || rt6_need_strict(src)) - return 0; + return BPF_FIB_LKUP_RET_NOT_FWDED; dev = dev_get_by_index_rcu(net, params->ifindex); if (unlikely(!dev)) @@ -4198,7 +4208,7 @@ static int bpf_ipv6_fib_lookup(struct net *net, struct bpf_fib_lookup *params, idev = __in6_dev_get_safely(dev); if (unlikely(!idev || !net->ipv6.devconf_all->forwarding)) - return 0; + return BPF_FIB_LKUP_RET_FWD_DISABLED; if (flags & BPF_FIB_LOOKUP_OUTPUT) { fl6.flowi6_iif = 1; @@ -4225,7 +4235,7 @@ static int bpf_ipv6_fib_lookup(struct net *net, struct bpf_fib_lookup *params, tb = ipv6_stub->fib6_get_table(net, tbid); if (unlikely(!tb)) - return 0; + return BPF_FIB_LKUP_RET_NOT_FWDED; f6i = ipv6_stub->fib6_table_lookup(net, tb, oif, &fl6, strict); } else { @@ -4238,11 +4248,23 @@ static int bpf_ipv6_fib_lookup(struct net *net, struct bpf_fib_lookup *params, } if (unlikely(IS_ERR_OR_NULL(f6i) || f6i == net->ipv6.fib6_null_entry)) - return 0; + return BPF_FIB_LKUP_RET_NOT_FWDED; + + if (unlikely(f6i->fib6_flags & RTF_REJECT)) { + switch (f6i->fib6_type) { + case RTN_BLACKHOLE: + return BPF_FIB_LKUP_RET_BLACKHOLE; + case RTN_UNREACHABLE: + return BPF_FIB_LKUP_RET_UNREACHABLE; + case RTN_PROHIBIT: + return BPF_FIB_LKUP_RET_PROHIBIT; + default: + return BPF_FIB_LKUP_RET_NOT_FWDED; + } + } - if (unlikely(f6i->fib6_flags & RTF_REJECT || - f6i->fib6_type != RTN_UNICAST)) - return 0; + if (f6i->fib6_type != RTN_UNICAST) + return BPF_FIB_LKUP_RET_NOT_FWDED; if (f6i->fib6_nsiblings && fl6.flowi6_oif == 0) f6i = ipv6_stub->fib6_multipath_select(net, f6i, &fl6, @@ -4252,11 +4274,11 @@ static int bpf_ipv6_fib_lookup(struct net *net, struct bpf_fib_lookup *params, if (check_mtu) { mtu = ipv6_stub->ip6_mtu_from_fib6(f6i, dst, src); if (params->tot_len > mtu) - return 0; + return BPF_FIB_LKUP_RET_FRAG_NEEDED; } if (f6i->fib6_nh.nh_lwtstate) - return 0; + return BPF_FIB_LKUP_RET_UNSUPP_LWT; if (f6i->fib6_flags & RTF_GATEWAY) *dst = f6i->fib6_nh.nh_gw; @@ -4270,10 +4292,10 @@ static int bpf_ipv6_fib_lookup(struct net *net, struct bpf_fib_lookup *params, */ neigh = ___neigh_lookup_noref(ipv6_stub->nd_tbl, neigh_key_eq128, ndisc_hashfn, dst, dev); - if (neigh) - return bpf_fib_set_fwd_params(params, neigh, dev); + if (!neigh) + return BPF_FIB_LKUP_RET_NO_NEIGH; - return 0; + return bpf_fib_set_fwd_params(params, neigh, dev); } #endif @@ -4315,7 +4337,7 @@ BPF_CALL_4(bpf_skb_fib_lookup, struct sk_buff *, skb, struct bpf_fib_lookup *, params, int, plen, u32, flags) { struct net *net = dev_net(skb->dev); - int index = -EAFNOSUPPORT; + int rc = -EAFNOSUPPORT; if (plen < sizeof(*params)) return -EINVAL; @@ -4326,25 +4348,25 @@ BPF_CALL_4(bpf_skb_fib_lookup, struct sk_buff *, skb, switch (params->family) { #if IS_ENABLED(CONFIG_INET) case AF_INET: - index = bpf_ipv4_fib_lookup(net, params, flags, false); + rc = bpf_ipv4_fib_lookup(net, params, flags, false); break; #endif #if IS_ENABLED(CONFIG_IPV6) case AF_INET6: - index = bpf_ipv6_fib_lookup(net, params, flags, false); + rc = bpf_ipv6_fib_lookup(net, params, flags, false); break; #endif } - if (index > 0) { + if (!rc) { struct net_device *dev; - dev = dev_get_by_index_rcu(net, index); + dev = dev_get_by_index_rcu(net, params->ifindex); if (!is_skb_forwardable(dev, skb)) - index = 0; + rc = BPF_FIB_LKUP_RET_FRAG_NEEDED; } - return index; + return rc; } static const struct bpf_func_proto bpf_skb_fib_lookup_proto = { diff --git a/samples/bpf/xdp_fwd_kern.c b/samples/bpf/xdp_fwd_kern.c index 6673cdb9f55c..a7e94e7ff87d 100644 --- a/samples/bpf/xdp_fwd_kern.c +++ b/samples/bpf/xdp_fwd_kern.c @@ -48,9 +48,9 @@ static __always_inline int xdp_fwd_flags(struct xdp_md *ctx, u32 flags) struct ethhdr *eth = data; struct ipv6hdr *ip6h; struct iphdr *iph; - int out_index; u16 h_proto; u64 nh_off; + int rc; nh_off = sizeof(*eth); if (data + nh_off > data_end) @@ -101,7 +101,7 @@ static __always_inline int xdp_fwd_flags(struct xdp_md *ctx, u32 flags) fib_params.ifindex = ctx->ingress_ifindex; - out_index = bpf_fib_lookup(ctx, &fib_params, sizeof(fib_params), flags); + rc = bpf_fib_lookup(ctx, &fib_params, sizeof(fib_params), flags); /* verify egress index has xdp support * TO-DO bpf_map_lookup_elem(&tx_port, &key) fails with @@ -109,7 +109,7 @@ static __always_inline int xdp_fwd_flags(struct xdp_md *ctx, u32 flags) * NOTE: without verification that egress index supports XDP * forwarding packets are dropped. */ - if (out_index > 0) { + if (rc == 0) { if (h_proto == htons(ETH_P_IP)) ip_decrease_ttl(iph); else if (h_proto == htons(ETH_P_IPV6)) @@ -117,7 +117,7 @@ static __always_inline int xdp_fwd_flags(struct xdp_md *ctx, u32 flags) memcpy(eth->h_dest, fib_params.dmac, ETH_ALEN); memcpy(eth->h_source, fib_params.smac, ETH_ALEN); - return bpf_redirect_map(&tx_port, out_index, 0); + return bpf_redirect_map(&tx_port, fib_params.ifindex, 0); } return XDP_PASS; -- cgit v1.2.3 From 85782e037f8aba8922dadb24a1523ca0b82ab8bc Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Thu, 28 Jun 2018 23:34:59 +0200 Subject: bpf: undo prog rejection on read-only lock failure Partially undo commit 9facc336876f ("bpf: reject any prog that failed read-only lock") since it caused a regression, that is, syzkaller was able to manage to cause a panic via fault injection deep in set_memory_ro() path by letting an allocation fail: In x86's __change_page_attr_set_clr() it was able to change the attributes of the primary mapping but not in the alias mapping via cpa_process_alias(), so the second, inner call to the __change_page_attr() via __change_page_attr_set_clr() had to split a larger page and failed in the alloc_pages() with the artifically triggered allocation error which is then propagated down to the call site. Thus, for set_memory_ro() this means that it returned with an error, but from debugging a probe_kernel_write() revealed EFAULT on that memory since the primary mapping succeeded to get changed. Therefore the subsequent hdr->locked = 0 reset triggered the panic as it was performed on read-only memory, so call-site assumptions were infact wrong to assume that it would either succeed /or/ not succeed at all since there's no such rollback in set_memory_*() calls from partial change of mappings, in other words, we're left in a state that is "half done". A later undo via set_memory_rw() is succeeding though due to matching permissions on that part (aka due to the try_preserve_large_page() succeeding). While reproducing locally with explicitly triggering this error, the initial splitting only happens on rare occasions and in real world it would additionally need oom conditions, but that said, it could partially fail. Therefore, it is definitely wrong to bail out on set_memory_ro() error and reject the program with the set_memory_*() semantics we have today. Shouldn't have gone the extra mile since no other user in tree today infact checks for any set_memory_*() errors, e.g. neither module_enable_ro() / module_disable_ro() for module RO/NX handling which is mostly default these days nor kprobes core with alloc_insn_page() / free_insn_page() as examples that could be invoked long after bootup and original 314beb9bcabf ("x86: bpf_jit_comp: secure bpf jit against spraying attacks") did neither when it got first introduced to BPF so "improving" with bailing out was clearly not right when set_memory_*() cannot handle it today. Kees suggested that if set_memory_*() can fail, we should annotate it with __must_check, and all callers need to deal with it gracefully given those set_memory_*() markings aren't "advisory", but they're expected to actually do what they say. This might be an option worth to move forward in future but would at the same time require that set_memory_*() calls from supporting archs are guaranteed to be "atomic" in that they provide rollback if part of the range fails, once that happened, the transition from RW -> RO could be made more robust that way, while subsequent RO -> RW transition /must/ continue guaranteeing to always succeed the undo part. Reported-by: syzbot+a4eb8c7766952a1ca872@syzkaller.appspotmail.com Reported-by: syzbot+d866d1925855328eac3b@syzkaller.appspotmail.com Fixes: 9facc336876f ("bpf: reject any prog that failed read-only lock") Cc: Laura Abbott Cc: Kees Cook Signed-off-by: Daniel Borkmann Acked-by: Alexei Starovoitov Signed-off-by: Alexei Starovoitov --- include/linux/filter.h | 56 ++++++++------------------------------------------ kernel/bpf/core.c | 30 +-------------------------- 2 files changed, 9 insertions(+), 77 deletions(-) (limited to 'include') diff --git a/include/linux/filter.h b/include/linux/filter.h index 20f2659dd829..300baad62c88 100644 --- a/include/linux/filter.h +++ b/include/linux/filter.h @@ -470,9 +470,7 @@ struct sock_fprog_kern { }; struct bpf_binary_header { - u16 pages; - u16 locked:1; - + u32 pages; /* Some arches need word alignment for their instructions */ u8 image[] __aligned(4); }; @@ -481,7 +479,7 @@ struct bpf_prog { u16 pages; /* Number of allocated pages */ u16 jited:1, /* Is our filter JIT'ed? */ jit_requested:1,/* archs need to JIT the prog */ - locked:1, /* Program image locked? */ + undo_set_mem:1, /* Passed set_memory_ro() checkpoint */ gpl_compatible:1, /* Is filter GPL compatible? */ cb_access:1, /* Is control block accessed? */ dst_needed:1, /* Do we need dst entry? */ @@ -677,46 +675,24 @@ bpf_ctx_narrow_access_ok(u32 off, u32 size, u32 size_default) static inline void bpf_prog_lock_ro(struct bpf_prog *fp) { -#ifdef CONFIG_ARCH_HAS_SET_MEMORY - fp->locked = 1; - if (set_memory_ro((unsigned long)fp, fp->pages)) - fp->locked = 0; -#endif + fp->undo_set_mem = 1; + set_memory_ro((unsigned long)fp, fp->pages); } static inline void bpf_prog_unlock_ro(struct bpf_prog *fp) { -#ifdef CONFIG_ARCH_HAS_SET_MEMORY - if (fp->locked) { - WARN_ON_ONCE(set_memory_rw((unsigned long)fp, fp->pages)); - /* In case set_memory_rw() fails, we want to be the first - * to crash here instead of some random place later on. - */ - fp->locked = 0; - } -#endif + if (fp->undo_set_mem) + set_memory_rw((unsigned long)fp, fp->pages); } static inline void bpf_jit_binary_lock_ro(struct bpf_binary_header *hdr) { -#ifdef CONFIG_ARCH_HAS_SET_MEMORY - hdr->locked = 1; - if (set_memory_ro((unsigned long)hdr, hdr->pages)) - hdr->locked = 0; -#endif + set_memory_ro((unsigned long)hdr, hdr->pages); } static inline void bpf_jit_binary_unlock_ro(struct bpf_binary_header *hdr) { -#ifdef CONFIG_ARCH_HAS_SET_MEMORY - if (hdr->locked) { - WARN_ON_ONCE(set_memory_rw((unsigned long)hdr, hdr->pages)); - /* In case set_memory_rw() fails, we want to be the first - * to crash here instead of some random place later on. - */ - hdr->locked = 0; - } -#endif + set_memory_rw((unsigned long)hdr, hdr->pages); } static inline struct bpf_binary_header * @@ -728,22 +704,6 @@ bpf_jit_binary_hdr(const struct bpf_prog *fp) return (void *)addr; } -#ifdef CONFIG_ARCH_HAS_SET_MEMORY -static inline int bpf_prog_check_pages_ro_single(const struct bpf_prog *fp) -{ - if (!fp->locked) - return -ENOLCK; - if (fp->jited) { - const struct bpf_binary_header *hdr = bpf_jit_binary_hdr(fp); - - if (!hdr->locked) - return -ENOLCK; - } - - return 0; -} -#endif - int sk_filter_trim_cap(struct sock *sk, struct sk_buff *skb, unsigned int cap); static inline int sk_filter(struct sock *sk, struct sk_buff *skb) { diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c index a9e6c04d0f4a..1e5625d46414 100644 --- a/kernel/bpf/core.c +++ b/kernel/bpf/core.c @@ -598,8 +598,6 @@ bpf_jit_binary_alloc(unsigned int proglen, u8 **image_ptr, bpf_fill_ill_insns(hdr, size); hdr->pages = size / PAGE_SIZE; - hdr->locked = 0; - hole = min_t(unsigned int, size - (proglen + sizeof(*hdr)), PAGE_SIZE - sizeof(*hdr)); start = (get_random_int() % hole) & ~(alignment - 1); @@ -1450,22 +1448,6 @@ static int bpf_check_tail_call(const struct bpf_prog *fp) return 0; } -static int bpf_prog_check_pages_ro_locked(const struct bpf_prog *fp) -{ -#ifdef CONFIG_ARCH_HAS_SET_MEMORY - int i, err; - - for (i = 0; i < fp->aux->func_cnt; i++) { - err = bpf_prog_check_pages_ro_single(fp->aux->func[i]); - if (err) - return err; - } - - return bpf_prog_check_pages_ro_single(fp); -#endif - return 0; -} - static void bpf_prog_select_func(struct bpf_prog *fp) { #ifndef CONFIG_BPF_JIT_ALWAYS_ON @@ -1524,17 +1506,7 @@ finalize: * all eBPF JITs might immediately support all features. */ *err = bpf_check_tail_call(fp); - if (*err) - return fp; - - /* Checkpoint: at this point onwards any cBPF -> eBPF or - * native eBPF program is read-only. If we failed to change - * the page attributes (e.g. allocation failure from - * splitting large pages), then reject the whole program - * in order to guarantee not ending up with any W+X pages - * from BPF side in kernel. - */ - *err = bpf_prog_check_pages_ro_locked(fp); + return fp; } EXPORT_SYMBOL_GPL(bpf_prog_select_runtime); -- cgit v1.2.3 From 55c5e0c602c20cb6f350e5ae357cfd7e04ebb189 Mon Sep 17 00:00:00 2001 From: Fabio Estevam Date: Sat, 2 Jun 2018 11:02:02 -0300 Subject: dt-bindings: clock: imx6ul: Do not change the clock definition order Commit f5a4670de966 ("clk: imx: Add new clo01 and clo2 controlled by CCOSR") introduced the CLK_CLKO definitions, but didn't put them at the end of the list, which may cause dtb breakage when running an old dtb with a newer kernel. In order to avoid that, simply add the new CLK_CKO clock definitions at the end of the list. Fixes: f5a4670de966 ("clk: imx: Add new clo01 and clo2 controlled by CCOSR") Reported-by: Stefan Wahren Signed-off-by: Fabio Estevam Acked-by: Rob Herring Reviewed-by: Stefan Agner Signed-off-by: Stephen Boyd --- include/dt-bindings/clock/imx6ul-clock.h | 40 +++++++++++++++----------------- 1 file changed, 19 insertions(+), 21 deletions(-) (limited to 'include') diff --git a/include/dt-bindings/clock/imx6ul-clock.h b/include/dt-bindings/clock/imx6ul-clock.h index 9564597cbfac..0aa1d9c3e0b9 100644 --- a/include/dt-bindings/clock/imx6ul-clock.h +++ b/include/dt-bindings/clock/imx6ul-clock.h @@ -235,27 +235,25 @@ #define IMX6UL_CLK_CSI_PODF 222 #define IMX6UL_CLK_PLL3_120M 223 #define IMX6UL_CLK_KPP 224 -#define IMX6UL_CLK_CKO1_SEL 225 -#define IMX6UL_CLK_CKO1_PODF 226 -#define IMX6UL_CLK_CKO1 227 -#define IMX6UL_CLK_CKO2_SEL 228 -#define IMX6UL_CLK_CKO2_PODF 229 -#define IMX6UL_CLK_CKO2 230 -#define IMX6UL_CLK_CKO 231 - -/* For i.MX6ULL */ -#define IMX6ULL_CLK_ESAI_PRED 232 -#define IMX6ULL_CLK_ESAI_PODF 233 -#define IMX6ULL_CLK_ESAI_EXTAL 234 -#define IMX6ULL_CLK_ESAI_MEM 235 -#define IMX6ULL_CLK_ESAI_IPG 236 -#define IMX6ULL_CLK_DCP_CLK 237 -#define IMX6ULL_CLK_EPDC_PRE_SEL 238 -#define IMX6ULL_CLK_EPDC_SEL 239 -#define IMX6ULL_CLK_EPDC_PODF 240 -#define IMX6ULL_CLK_EPDC_ACLK 241 -#define IMX6ULL_CLK_EPDC_PIX 242 -#define IMX6ULL_CLK_ESAI_SEL 243 +#define IMX6ULL_CLK_ESAI_PRED 225 +#define IMX6ULL_CLK_ESAI_PODF 226 +#define IMX6ULL_CLK_ESAI_EXTAL 227 +#define IMX6ULL_CLK_ESAI_MEM 228 +#define IMX6ULL_CLK_ESAI_IPG 229 +#define IMX6ULL_CLK_DCP_CLK 230 +#define IMX6ULL_CLK_EPDC_PRE_SEL 231 +#define IMX6ULL_CLK_EPDC_SEL 232 +#define IMX6ULL_CLK_EPDC_PODF 233 +#define IMX6ULL_CLK_EPDC_ACLK 234 +#define IMX6ULL_CLK_EPDC_PIX 235 +#define IMX6ULL_CLK_ESAI_SEL 236 +#define IMX6UL_CLK_CKO1_SEL 237 +#define IMX6UL_CLK_CKO1_PODF 238 +#define IMX6UL_CLK_CKO1 239 +#define IMX6UL_CLK_CKO2_SEL 240 +#define IMX6UL_CLK_CKO2_PODF 241 +#define IMX6UL_CLK_CKO2 242 +#define IMX6UL_CLK_CKO 243 #define IMX6UL_CLK_END 244 #endif /* __DT_BINDINGS_CLOCK_IMX6UL_H */ -- cgit v1.2.3 From 603d4cf8fe095b1ee78f423d514427be507fb513 Mon Sep 17 00:00:00 2001 From: Sabrina Dubroca Date: Sat, 30 Jun 2018 17:38:55 +0200 Subject: net: fix use-after-free in GRO with ESP Since the addition of GRO for ESP, gro_receive can consume the skb and return -EINPROGRESS. In that case, the lower layer GRO handler cannot touch the skb anymore. Commit 5f114163f2f5 ("net: Add a skb_gro_flush_final helper.") converted some of the gro_receive handlers that can lead to ESP's gro_receive so that they wouldn't access the skb when -EINPROGRESS is returned, but missed other spots, mainly in tunneling protocols. This patch finishes the conversion to using skb_gro_flush_final(), and adds a new helper, skb_gro_flush_final_remcsum(), used in VXLAN and GUE. Fixes: 5f114163f2f5 ("net: Add a skb_gro_flush_final helper.") Signed-off-by: Sabrina Dubroca Reviewed-by: Stefano Brivio Signed-off-by: David S. Miller --- drivers/net/geneve.c | 2 +- drivers/net/vxlan.c | 4 +--- include/linux/netdevice.h | 20 ++++++++++++++++++++ net/8021q/vlan.c | 2 +- net/ipv4/fou.c | 4 +--- net/ipv4/gre_offload.c | 2 +- net/ipv4/udp_offload.c | 2 +- 7 files changed, 26 insertions(+), 10 deletions(-) (limited to 'include') diff --git a/drivers/net/geneve.c b/drivers/net/geneve.c index 750eaa53bf0c..ada33c2d9ac2 100644 --- a/drivers/net/geneve.c +++ b/drivers/net/geneve.c @@ -476,7 +476,7 @@ static struct sk_buff **geneve_gro_receive(struct sock *sk, out_unlock: rcu_read_unlock(); out: - NAPI_GRO_CB(skb)->flush |= flush; + skb_gro_flush_final(skb, pp, flush); return pp; } diff --git a/drivers/net/vxlan.c b/drivers/net/vxlan.c index aee0e60471f1..f6bb1d54d4bd 100644 --- a/drivers/net/vxlan.c +++ b/drivers/net/vxlan.c @@ -623,9 +623,7 @@ static struct sk_buff **vxlan_gro_receive(struct sock *sk, flush = 0; out: - skb_gro_remcsum_cleanup(skb, &grc); - skb->remcsum_offload = 0; - NAPI_GRO_CB(skb)->flush |= flush; + skb_gro_flush_final_remcsum(skb, pp, flush, &grc); return pp; } diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 3ec9850c7936..3d0cc0b5cec2 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -2789,11 +2789,31 @@ static inline void skb_gro_flush_final(struct sk_buff *skb, struct sk_buff **pp, if (PTR_ERR(pp) != -EINPROGRESS) NAPI_GRO_CB(skb)->flush |= flush; } +static inline void skb_gro_flush_final_remcsum(struct sk_buff *skb, + struct sk_buff **pp, + int flush, + struct gro_remcsum *grc) +{ + if (PTR_ERR(pp) != -EINPROGRESS) { + NAPI_GRO_CB(skb)->flush |= flush; + skb_gro_remcsum_cleanup(skb, grc); + skb->remcsum_offload = 0; + } +} #else static inline void skb_gro_flush_final(struct sk_buff *skb, struct sk_buff **pp, int flush) { NAPI_GRO_CB(skb)->flush |= flush; } +static inline void skb_gro_flush_final_remcsum(struct sk_buff *skb, + struct sk_buff **pp, + int flush, + struct gro_remcsum *grc) +{ + NAPI_GRO_CB(skb)->flush |= flush; + skb_gro_remcsum_cleanup(skb, grc); + skb->remcsum_offload = 0; +} #endif static inline int dev_hard_header(struct sk_buff *skb, struct net_device *dev, diff --git a/net/8021q/vlan.c b/net/8021q/vlan.c index 73a65789271b..8ccee3d01822 100644 --- a/net/8021q/vlan.c +++ b/net/8021q/vlan.c @@ -693,7 +693,7 @@ static struct sk_buff **vlan_gro_receive(struct sk_buff **head, out_unlock: rcu_read_unlock(); out: - NAPI_GRO_CB(skb)->flush |= flush; + skb_gro_flush_final(skb, pp, flush); return pp; } diff --git a/net/ipv4/fou.c b/net/ipv4/fou.c index 1540db65241a..c9ec1603666b 100644 --- a/net/ipv4/fou.c +++ b/net/ipv4/fou.c @@ -448,9 +448,7 @@ next_proto: out_unlock: rcu_read_unlock(); out: - NAPI_GRO_CB(skb)->flush |= flush; - skb_gro_remcsum_cleanup(skb, &grc); - skb->remcsum_offload = 0; + skb_gro_flush_final_remcsum(skb, pp, flush, &grc); return pp; } diff --git a/net/ipv4/gre_offload.c b/net/ipv4/gre_offload.c index 1859c473b21a..6a7d980105f6 100644 --- a/net/ipv4/gre_offload.c +++ b/net/ipv4/gre_offload.c @@ -223,7 +223,7 @@ static struct sk_buff **gre_gro_receive(struct sk_buff **head, out_unlock: rcu_read_unlock(); out: - NAPI_GRO_CB(skb)->flush |= flush; + skb_gro_flush_final(skb, pp, flush); return pp; } diff --git a/net/ipv4/udp_offload.c b/net/ipv4/udp_offload.c index 92dc9e5a7ff3..69c54540d5b4 100644 --- a/net/ipv4/udp_offload.c +++ b/net/ipv4/udp_offload.c @@ -394,7 +394,7 @@ unflush: out_unlock: rcu_read_unlock(); out: - NAPI_GRO_CB(skb)->flush |= flush; + skb_gro_flush_final(skb, pp, flush); return pp; } EXPORT_SYMBOL(udp_gro_receive); -- cgit v1.2.3 From 240630e61870e62e39a97225048f9945848fa5f5 Mon Sep 17 00:00:00 2001 From: Hans de Goede Date: Sun, 1 Jul 2018 12:15:46 +0200 Subject: ahci: Disable LPM on Lenovo 50 series laptops with a too old BIOS There have been several reports of LPM related hard freezes about once a day on multiple Lenovo 50 series models. Strange enough these reports where not disk model specific as LPM issues usually are and some users with the exact same disk + laptop where seeing them while other users where not seeing these issues. It turns out that enabling LPM triggers a firmware bug somewhere, which has been fixed in later BIOS versions. This commit adds a new ahci_broken_lpm() function and a new ATA_FLAG_NO_LPM for dealing with this. The ahci_broken_lpm() function contains DMI match info for the 4 models which are known to be affected by this and the DMI BIOS date field for known good BIOS versions. If the BIOS date is older then the one in the table LPM will be disabled and a warning will be printed. Note the BIOS dates are for known good versions, some older versions may work too, but we don't know for sure, the table is using dates from BIOS versions for which users have confirmed that upgrading to that version makes the problem go away. Unfortunately I've been unable to get hold of the reporter who reported that BIOS version 2.35 fixed the problems on the W541 for him. I've been able to verify the DMI_SYS_VENDOR and DMI_PRODUCT_VERSION from an older dmidecode, but I don't know the exact BIOS date as reported in the DMI. Lenovo keeps a changelog with dates in their release notes, but the dates there are the release dates not the build dates which are in DMI. So I've chosen to set the date to which we compare to one day past the release date of the 2.34 BIOS. I plan to fix this with a follow up commit once I've the necessary info. Cc: stable@vger.kernel.org Signed-off-by: Hans de Goede Signed-off-by: Tejun Heo --- drivers/ata/ahci.c | 59 +++++++++++++++++++++++++++++++++++++++++++++++ drivers/ata/libata-core.c | 3 +++ include/linux/libata.h | 1 + 3 files changed, 63 insertions(+) (limited to 'include') diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c index 738fb22978dd..fdeb3b4d0f4a 100644 --- a/drivers/ata/ahci.c +++ b/drivers/ata/ahci.c @@ -1280,6 +1280,59 @@ static bool ahci_broken_suspend(struct pci_dev *pdev) return strcmp(buf, dmi->driver_data) < 0; } +static bool ahci_broken_lpm(struct pci_dev *pdev) +{ + static const struct dmi_system_id sysids[] = { + /* Various Lenovo 50 series have LPM issues with older BIOSen */ + { + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"), + DMI_MATCH(DMI_PRODUCT_VERSION, "ThinkPad X250"), + }, + .driver_data = "20180406", /* 1.31 */ + }, + { + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"), + DMI_MATCH(DMI_PRODUCT_VERSION, "ThinkPad L450"), + }, + .driver_data = "20180420", /* 1.28 */ + }, + { + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"), + DMI_MATCH(DMI_PRODUCT_VERSION, "ThinkPad T450s"), + }, + .driver_data = "20180315", /* 1.33 */ + }, + { + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "LENOVO"), + DMI_MATCH(DMI_PRODUCT_VERSION, "ThinkPad W541"), + }, + /* + * Note date based on release notes, 2.35 has been + * reported to be good, but I've been unable to get + * a hold of the reporter to get the DMI BIOS date. + * TODO: fix this. + */ + .driver_data = "20180310", /* 2.35 */ + }, + { } /* terminate list */ + }; + const struct dmi_system_id *dmi = dmi_first_match(sysids); + int year, month, date; + char buf[9]; + + if (!dmi) + return false; + + dmi_get_date(DMI_BIOS_DATE, &year, &month, &date); + snprintf(buf, sizeof(buf), "%04d%02d%02d", year, month, date); + + return strcmp(buf, dmi->driver_data) < 0; +} + static bool ahci_broken_online(struct pci_dev *pdev) { #define ENCODE_BUSDEVFN(bus, slot, func) \ @@ -1694,6 +1747,12 @@ static int ahci_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) "quirky BIOS, skipping spindown on poweroff\n"); } + if (ahci_broken_lpm(pdev)) { + pi.flags |= ATA_FLAG_NO_LPM; + dev_warn(&pdev->dev, + "BIOS update required for Link Power Management support\n"); + } + if (ahci_broken_suspend(pdev)) { hpriv->flags |= AHCI_HFLAG_NO_SUSPEND; dev_warn(&pdev->dev, diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c index 27d15ed7fa3d..cc71c63df381 100644 --- a/drivers/ata/libata-core.c +++ b/drivers/ata/libata-core.c @@ -2493,6 +2493,9 @@ int ata_dev_configure(struct ata_device *dev) (id[ATA_ID_SATA_CAPABILITY] & 0xe) == 0x2) dev->horkage |= ATA_HORKAGE_NOLPM; + if (ap->flags & ATA_FLAG_NO_LPM) + dev->horkage |= ATA_HORKAGE_NOLPM; + if (dev->horkage & ATA_HORKAGE_NOLPM) { ata_dev_warn(dev, "LPM support broken, forcing max_power\n"); dev->link->ap->target_lpm_policy = ATA_LPM_MAX_POWER; diff --git a/include/linux/libata.h b/include/linux/libata.h index a2257e380789..32f247cb5e9e 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -210,6 +210,7 @@ enum { ATA_FLAG_SLAVE_POSS = (1 << 0), /* host supports slave dev */ /* (doesn't imply presence) */ ATA_FLAG_SATA = (1 << 1), + ATA_FLAG_NO_LPM = (1 << 2), /* host not happy with LPM */ ATA_FLAG_NO_LOG_PAGE = (1 << 5), /* do not issue log page read */ ATA_FLAG_NO_ATAPI = (1 << 6), /* No ATAPI support */ ATA_FLAG_PIO_DMA = (1 << 7), /* PIO cmds via DMA */ -- cgit v1.2.3 From 1cef1150ef40ec52f507436a14230cbc2623299c Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 7 Jun 2018 11:45:49 +0200 Subject: kthread, sched/core: Fix kthread_parkme() (again...) Gaurav reports that commit: 85f1abe0019f ("kthread, sched/wait: Fix kthread_parkme() completion issue") isn't working for him. Because of the following race: > controller Thread CPUHP Thread > takedown_cpu > kthread_park > kthread_parkme > Set KTHREAD_SHOULD_PARK > smpboot_thread_fn > set Task interruptible > > > wake_up_process > if (!(p->state & state)) > goto out; > > Kthread_parkme > SET TASK_PARKED > schedule > raw_spin_lock(&rq->lock) > ttwu_remote > waiting for __task_rq_lock > context_switch > > finish_lock_switch > > > > Case TASK_PARKED > kthread_park_complete > > > SET Running Furthermore, Oleg noticed that the whole scheduler TASK_PARKED handling is buggered because the TASK_DEAD thing is done with preemption disabled, the current code can still complete early on preemption :/ So basically revert that earlier fix and go with a variant of the alternative mentioned in the commit. Promote TASK_PARKED to special state to avoid the store-store issue on task->state leading to the WARN in kthread_unpark() -> __kthread_bind(). But in addition, add wait_task_inactive() to kthread_park() to ensure the task really is PARKED when we return from kthread_park(). This avoids the whole kthread still gets migrated nonsense -- although it would be really good to get this done differently. Reported-by: Gaurav Kohli Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Oleg Nesterov Cc: Peter Zijlstra Cc: Thomas Gleixner Fixes: 85f1abe0019f ("kthread, sched/wait: Fix kthread_parkme() completion issue") Signed-off-by: Ingo Molnar --- include/linux/kthread.h | 1 - include/linux/sched.h | 2 +- kernel/kthread.c | 30 ++++++++++++++++++++++++------ kernel/sched/core.c | 31 +++++++++++-------------------- 4 files changed, 36 insertions(+), 28 deletions(-) (limited to 'include') diff --git a/include/linux/kthread.h b/include/linux/kthread.h index 2803264c512f..c1961761311d 100644 --- a/include/linux/kthread.h +++ b/include/linux/kthread.h @@ -62,7 +62,6 @@ void *kthread_probe_data(struct task_struct *k); int kthread_park(struct task_struct *k); void kthread_unpark(struct task_struct *k); void kthread_parkme(void); -void kthread_park_complete(struct task_struct *k); int kthreadd(void *unused); extern struct task_struct *kthreadd_task; diff --git a/include/linux/sched.h b/include/linux/sched.h index 9256118bd40c..43731fe51c97 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -118,7 +118,7 @@ struct task_group; * the comment with set_special_state(). */ #define is_special_task_state(state) \ - ((state) & (__TASK_STOPPED | __TASK_TRACED | TASK_DEAD)) + ((state) & (__TASK_STOPPED | __TASK_TRACED | TASK_PARKED | TASK_DEAD)) #define __set_current_state(state_value) \ do { \ diff --git a/kernel/kthread.c b/kernel/kthread.c index 481951bf091d..750cb8082694 100644 --- a/kernel/kthread.c +++ b/kernel/kthread.c @@ -177,9 +177,20 @@ void *kthread_probe_data(struct task_struct *task) static void __kthread_parkme(struct kthread *self) { for (;;) { - set_current_state(TASK_PARKED); + /* + * TASK_PARKED is a special state; we must serialize against + * possible pending wakeups to avoid store-store collisions on + * task->state. + * + * Such a collision might possibly result in the task state + * changin from TASK_PARKED and us failing the + * wait_task_inactive() in kthread_park(). + */ + set_special_state(TASK_PARKED); if (!test_bit(KTHREAD_SHOULD_PARK, &self->flags)) break; + + complete_all(&self->parked); schedule(); } __set_current_state(TASK_RUNNING); @@ -191,11 +202,6 @@ void kthread_parkme(void) } EXPORT_SYMBOL_GPL(kthread_parkme); -void kthread_park_complete(struct task_struct *k) -{ - complete_all(&to_kthread(k)->parked); -} - static int kthread(void *_create) { /* Copy data: it's on kthread's stack */ @@ -461,6 +467,9 @@ void kthread_unpark(struct task_struct *k) reinit_completion(&kthread->parked); clear_bit(KTHREAD_SHOULD_PARK, &kthread->flags); + /* + * __kthread_parkme() will either see !SHOULD_PARK or get the wakeup. + */ wake_up_state(k, TASK_PARKED); } EXPORT_SYMBOL_GPL(kthread_unpark); @@ -487,7 +496,16 @@ int kthread_park(struct task_struct *k) set_bit(KTHREAD_SHOULD_PARK, &kthread->flags); if (k != current) { wake_up_process(k); + /* + * Wait for __kthread_parkme() to complete(), this means we + * _will_ have TASK_PARKED and are about to call schedule(). + */ wait_for_completion(&kthread->parked); + /* + * Now wait for that schedule() to complete and the task to + * get scheduled out. + */ + WARN_ON_ONCE(!wait_task_inactive(k, TASK_PARKED)); } return 0; diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 22fce36426c0..fe365c9a08e9 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -7,7 +7,6 @@ */ #include "sched.h" -#include #include #include @@ -2724,28 +2723,20 @@ static struct rq *finish_task_switch(struct task_struct *prev) membarrier_mm_sync_core_before_usermode(mm); mmdrop(mm); } - if (unlikely(prev_state & (TASK_DEAD|TASK_PARKED))) { - switch (prev_state) { - case TASK_DEAD: - if (prev->sched_class->task_dead) - prev->sched_class->task_dead(prev); + if (unlikely(prev_state == TASK_DEAD)) { + if (prev->sched_class->task_dead) + prev->sched_class->task_dead(prev); - /* - * Remove function-return probe instances associated with this - * task and put them back on the free list. - */ - kprobe_flush_task(prev); - - /* Task is done with its stack. */ - put_task_stack(prev); + /* + * Remove function-return probe instances associated with this + * task and put them back on the free list. + */ + kprobe_flush_task(prev); - put_task_struct(prev); - break; + /* Task is done with its stack. */ + put_task_stack(prev); - case TASK_PARKED: - kthread_park_complete(prev); - break; - } + put_task_struct(prev); } tick_nohz_task_switch(); -- cgit v1.2.3 From d03db2bc26f0e4a6849ad649a09c9c73fccdc656 Mon Sep 17 00:00:00 2001 From: Nick Desaulniers Date: Thu, 21 Jun 2018 09:23:22 -0700 Subject: compiler-gcc.h: Add __attribute__((gnu_inline)) to all inline declarations Functions marked extern inline do not emit an externally visible function when the gnu89 C standard is used. Some KBUILD Makefiles overwrite KBUILD_CFLAGS. This is an issue for GCC 5.1+ users as without an explicit C standard specified, the default is gnu11. Since c99, the semantics of extern inline have changed such that an externally visible function is always emitted. This can lead to multiple definition errors of extern inline functions at link time of compilation units whose build files have removed an explicit C standard compiler flag for users of GCC 5.1+ or Clang. Suggested-by: Arnd Bergmann Suggested-by: H. Peter Anvin Suggested-by: Joe Perches Signed-off-by: Nick Desaulniers Acked-by: Juergen Gross Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: acme@redhat.com Cc: akataria@vmware.com Cc: akpm@linux-foundation.org Cc: andrea.parri@amarulasolutions.com Cc: ard.biesheuvel@linaro.org Cc: aryabinin@virtuozzo.com Cc: astrachan@google.com Cc: boris.ostrovsky@oracle.com Cc: brijesh.singh@amd.com Cc: caoj.fnst@cn.fujitsu.com Cc: geert@linux-m68k.org Cc: ghackmann@google.com Cc: gregkh@linuxfoundation.org Cc: jan.kiszka@siemens.com Cc: jarkko.sakkinen@linux.intel.com Cc: jpoimboe@redhat.com Cc: keescook@google.com Cc: kirill.shutemov@linux.intel.com Cc: kstewart@linuxfoundation.org Cc: linux-efi@vger.kernel.org Cc: linux-kbuild@vger.kernel.org Cc: manojgupta@google.com Cc: mawilcox@microsoft.com Cc: michal.lkml@markovi.net Cc: mjg59@google.com Cc: mka@chromium.org Cc: pombredanne@nexb.com Cc: rientjes@google.com Cc: rostedt@goodmis.org Cc: sedat.dilek@gmail.com Cc: thomas.lendacky@amd.com Cc: tstellar@redhat.com Cc: tweek@google.com Cc: virtualization@lists.linux-foundation.org Cc: will.deacon@arm.com Cc: yamada.masahiro@socionext.com Link: http://lkml.kernel.org/r/20180621162324.36656-2-ndesaulniers@google.com Signed-off-by: Ingo Molnar --- include/linux/compiler-gcc.h | 29 ++++++++++++++++++++++------- 1 file changed, 22 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/include/linux/compiler-gcc.h b/include/linux/compiler-gcc.h index fd282c7d3e5e..573f5a7d42d4 100644 --- a/include/linux/compiler-gcc.h +++ b/include/linux/compiler-gcc.h @@ -65,6 +65,18 @@ #define __must_be_array(a) BUILD_BUG_ON_ZERO(__same_type((a), &(a)[0])) #endif +/* + * Feature detection for gnu_inline (gnu89 extern inline semantics). Either + * __GNUC_STDC_INLINE__ is defined (not using gnu89 extern inline semantics, + * and we opt in to the gnu89 semantics), or __GNUC_STDC_INLINE__ is not + * defined so the gnu89 semantics are the default. + */ +#ifdef __GNUC_STDC_INLINE__ +# define __gnu_inline __attribute__((gnu_inline)) +#else +# define __gnu_inline +#endif + /* * Force always-inline if the user requests it so via the .config, * or if gcc is too old. @@ -72,19 +84,22 @@ * -Wunused-function. This turns out to avoid the need for complex #ifdef * directives. Suppress the warning in clang as well by using "unused" * function attribute, which is redundant but not harmful for gcc. + * Prefer gnu_inline, so that extern inline functions do not emit an + * externally visible function. This makes extern inline behave as per gnu89 + * semantics rather than c99. This prevents multiple symbol definition errors + * of extern inline functions at link time. + * A lot of inline functions can cause havoc with function tracing. */ #if !defined(CONFIG_ARCH_SUPPORTS_OPTIMIZED_INLINING) || \ !defined(CONFIG_OPTIMIZE_INLINING) || (__GNUC__ < 4) -#define inline inline __attribute__((always_inline,unused)) notrace -#define __inline__ __inline__ __attribute__((always_inline,unused)) notrace -#define __inline __inline __attribute__((always_inline,unused)) notrace +#define inline \ + inline __attribute__((always_inline, unused)) notrace __gnu_inline #else -/* A lot of inline functions can cause havoc with function tracing */ -#define inline inline __attribute__((unused)) notrace -#define __inline__ __inline__ __attribute__((unused)) notrace -#define __inline __inline __attribute__((unused)) notrace +#define inline inline __attribute__((unused)) notrace __gnu_inline #endif +#define __inline__ inline +#define __inline inline #define __always_inline inline __attribute__((always_inline)) #define noinline __attribute__((noinline)) -- cgit v1.2.3 From 5ccba64a560fa6ca06008d4001f5d46ebeb34b41 Mon Sep 17 00:00:00 2001 From: Yisheng Xie Date: Fri, 2 Feb 2018 10:14:49 +0800 Subject: ftrace: Nuke clear_ftrace_function clear_ftrace_function is not used outside of ftrace.c and is not help to use a function, so nuke it per Steve's suggestion. Link: http://lkml.kernel.org/r/1517537689-34947-1-git-send-email-xieyisheng1@huawei.com Suggested-by: Steven Rostedt Signed-off-by: Yisheng Xie Signed-off-by: Steven Rostedt (VMware) --- include/linux/ftrace.h | 2 -- kernel/trace/ftrace.c | 13 +------------ 2 files changed, 1 insertion(+), 14 deletions(-) (limited to 'include') diff --git a/include/linux/ftrace.h b/include/linux/ftrace.h index 8154f4920fcb..ebb77674be90 100644 --- a/include/linux/ftrace.h +++ b/include/linux/ftrace.h @@ -223,7 +223,6 @@ extern enum ftrace_tracing_type_t ftrace_tracing_type; */ int register_ftrace_function(struct ftrace_ops *ops); int unregister_ftrace_function(struct ftrace_ops *ops); -void clear_ftrace_function(void); extern void ftrace_stub(unsigned long a0, unsigned long a1, struct ftrace_ops *op, struct pt_regs *regs); @@ -239,7 +238,6 @@ static inline int ftrace_nr_registered_ops(void) { return 0; } -static inline void clear_ftrace_function(void) { } static inline void ftrace_kill(void) { } static inline void ftrace_free_init_mem(void) { } static inline void ftrace_free_mem(struct module *mod, void *start, void *end) { } diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index efed9c1cfb7e..caf9cbf35816 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -192,17 +192,6 @@ static void ftrace_pid_func(unsigned long ip, unsigned long parent_ip, op->saved_func(ip, parent_ip, op, regs); } -/** - * clear_ftrace_function - reset the ftrace function - * - * This NULLs the ftrace function and in essence stops - * tracing. There may be lag - */ -void clear_ftrace_function(void) -{ - ftrace_trace_function = ftrace_stub; -} - static void ftrace_sync(struct work_struct *work) { /* @@ -6689,7 +6678,7 @@ void ftrace_kill(void) { ftrace_disabled = 1; ftrace_enabled = 0; - clear_ftrace_function(); + ftrace_trace_function = ftrace_stub; } /** -- cgit v1.2.3 From 543af5861f41af0a5d2432f6fb5976af50f9cee5 Mon Sep 17 00:00:00 2001 From: Xiubo Li Date: Fri, 6 Jul 2018 22:05:38 -0400 Subject: uio: change to use the mutex lock instead of the spin lock We are hitting a regression with the following commit: commit a93e7b331568227500186a465fee3c2cb5dffd1f Author: Hamish Martin Date: Mon May 14 13:32:23 2018 +1200 uio: Prevent device destruction while fds are open The problem is the addition of spin_lock_irqsave in uio_write. This leads to hitting uio_write -> copy_from_user -> _copy_from_user -> might_fault and the logs filling up with sleeping warnings. I also noticed some uio drivers allocate memory, sleep, grab mutexes from callouts like open() and release and uio is now doing spin_lock_irqsave while calling them. Reported-by: Mike Christie CC: Hamish Martin Reviewed-by: Hamish Martin Signed-off-by: Xiubo Li Signed-off-by: Greg Kroah-Hartman --- drivers/uio/uio.c | 32 +++++++++++++------------------- include/linux/uio_driver.h | 2 +- 2 files changed, 14 insertions(+), 20 deletions(-) (limited to 'include') diff --git a/drivers/uio/uio.c b/drivers/uio/uio.c index b4b2ae1e0473..655ade4fb3b1 100644 --- a/drivers/uio/uio.c +++ b/drivers/uio/uio.c @@ -433,7 +433,6 @@ static int uio_open(struct inode *inode, struct file *filep) struct uio_device *idev; struct uio_listener *listener; int ret = 0; - unsigned long flags; mutex_lock(&minor_lock); idev = idr_find(&uio_idr, iminor(inode)); @@ -460,10 +459,10 @@ static int uio_open(struct inode *inode, struct file *filep) listener->event_count = atomic_read(&idev->event); filep->private_data = listener; - spin_lock_irqsave(&idev->info_lock, flags); + mutex_lock(&idev->info_lock); if (idev->info && idev->info->open) ret = idev->info->open(idev->info, inode); - spin_unlock_irqrestore(&idev->info_lock, flags); + mutex_unlock(&idev->info_lock); if (ret) goto err_infoopen; @@ -495,12 +494,11 @@ static int uio_release(struct inode *inode, struct file *filep) int ret = 0; struct uio_listener *listener = filep->private_data; struct uio_device *idev = listener->dev; - unsigned long flags; - spin_lock_irqsave(&idev->info_lock, flags); + mutex_lock(&idev->info_lock); if (idev->info && idev->info->release) ret = idev->info->release(idev->info, inode); - spin_unlock_irqrestore(&idev->info_lock, flags); + mutex_unlock(&idev->info_lock); module_put(idev->owner); kfree(listener); @@ -513,12 +511,11 @@ static __poll_t uio_poll(struct file *filep, poll_table *wait) struct uio_listener *listener = filep->private_data; struct uio_device *idev = listener->dev; __poll_t ret = 0; - unsigned long flags; - spin_lock_irqsave(&idev->info_lock, flags); + mutex_lock(&idev->info_lock); if (!idev->info || !idev->info->irq) ret = -EIO; - spin_unlock_irqrestore(&idev->info_lock, flags); + mutex_unlock(&idev->info_lock); if (ret) return ret; @@ -537,12 +534,11 @@ static ssize_t uio_read(struct file *filep, char __user *buf, DECLARE_WAITQUEUE(wait, current); ssize_t retval = 0; s32 event_count; - unsigned long flags; - spin_lock_irqsave(&idev->info_lock, flags); + mutex_lock(&idev->info_lock); if (!idev->info || !idev->info->irq) retval = -EIO; - spin_unlock_irqrestore(&idev->info_lock, flags); + mutex_unlock(&idev->info_lock); if (retval) return retval; @@ -592,9 +588,8 @@ static ssize_t uio_write(struct file *filep, const char __user *buf, struct uio_device *idev = listener->dev; ssize_t retval; s32 irq_on; - unsigned long flags; - spin_lock_irqsave(&idev->info_lock, flags); + mutex_lock(&idev->info_lock); if (!idev->info || !idev->info->irq) { retval = -EIO; goto out; @@ -618,7 +613,7 @@ static ssize_t uio_write(struct file *filep, const char __user *buf, retval = idev->info->irqcontrol(idev->info, irq_on); out: - spin_unlock_irqrestore(&idev->info_lock, flags); + mutex_unlock(&idev->info_lock); return retval ? retval : sizeof(s32); } @@ -865,7 +860,7 @@ int __uio_register_device(struct module *owner, idev->owner = owner; idev->info = info; - spin_lock_init(&idev->info_lock); + mutex_init(&idev->info_lock); init_waitqueue_head(&idev->wait); atomic_set(&idev->event, 0); @@ -929,7 +924,6 @@ EXPORT_SYMBOL_GPL(__uio_register_device); void uio_unregister_device(struct uio_info *info) { struct uio_device *idev; - unsigned long flags; if (!info || !info->uio_dev) return; @@ -943,9 +937,9 @@ void uio_unregister_device(struct uio_info *info) if (info->irq && info->irq != UIO_IRQ_CUSTOM) free_irq(info->irq, idev); - spin_lock_irqsave(&idev->info_lock, flags); + mutex_lock(&idev->info_lock); idev->info = NULL; - spin_unlock_irqrestore(&idev->info_lock, flags); + mutex_unlock(&idev->info_lock); device_unregister(&idev->dev); diff --git a/include/linux/uio_driver.h b/include/linux/uio_driver.h index 6c5f2074e14f..6f8b68cd460f 100644 --- a/include/linux/uio_driver.h +++ b/include/linux/uio_driver.h @@ -75,7 +75,7 @@ struct uio_device { struct fasync_struct *async_queue; wait_queue_head_t wait; struct uio_info *info; - spinlock_t info_lock; + struct mutex info_lock; struct kobject *map_dir; struct kobject *portio_dir; }; -- cgit v1.2.3 From e96d71359e9bbea846a2111e4469a03a055dfa6f Mon Sep 17 00:00:00 2001 From: Mathieu Desnoyers Date: Mon, 9 Jul 2018 15:51:50 -0400 Subject: rseq: Use __u64 for rseq_cs fields, validate user inputs Change the rseq ABI so rseq_cs start_ip, post_commit_offset and abort_ip fields are seen as 64-bit fields by both 32-bit and 64-bit kernels rather that ignoring the 32 upper bits on 32-bit kernels. This ensures we have a consistent behavior for a 32-bit binary executed on 32-bit kernels and in compat mode on 64-bit kernels. Validating the value of abort_ip field to be below TASK_SIZE ensures the kernel don't return to an invalid address when returning to userspace after an abort. I don't fully trust each architecture code to consistently deal with invalid return addresses. Validating the value of the start_ip and post_commit_offset fields prevents overflow on arithmetic performed on those values, used to check whether abort_ip is within the rseq critical section. If validation fails, the process is killed with a segmentation fault. When the signature encountered before abort_ip does not match the expected signature, return -EINVAL rather than -EPERM to be consistent with other input validation return codes from rseq_get_rseq_cs(). Signed-off-by: Mathieu Desnoyers Signed-off-by: Thomas Gleixner Cc: linux-api@vger.kernel.org Cc: Peter Zijlstra Cc: "Paul E . McKenney" Cc: Boqun Feng Cc: Andy Lutomirski Cc: Dave Watson Cc: Paul Turner Cc: Andrew Morton Cc: Russell King Cc: "H . Peter Anvin" Cc: Andi Kleen Cc: Chris Lameter Cc: Ben Maurer Cc: Steven Rostedt Cc: Josh Triplett Cc: Linus Torvalds Cc: Catalin Marinas Cc: Will Deacon Cc: Michael Kerrisk Cc: Joel Fernandes Cc: "Paul E. McKenney" Cc: "H. Peter Anvin" Link: https://lkml.kernel.org/r/20180709195155.7654-2-mathieu.desnoyers@efficios.com --- include/uapi/linux/rseq.h | 6 +++--- kernel/rseq.c | 14 ++++++++++---- 2 files changed, 13 insertions(+), 7 deletions(-) (limited to 'include') diff --git a/include/uapi/linux/rseq.h b/include/uapi/linux/rseq.h index d620fa43756c..519ad6e176d1 100644 --- a/include/uapi/linux/rseq.h +++ b/include/uapi/linux/rseq.h @@ -52,10 +52,10 @@ struct rseq_cs { __u32 version; /* enum rseq_cs_flags */ __u32 flags; - LINUX_FIELD_u32_u64(start_ip); + __u64 start_ip; /* Offset from start_ip. */ - LINUX_FIELD_u32_u64(post_commit_offset); - LINUX_FIELD_u32_u64(abort_ip); + __u64 post_commit_offset; + __u64 abort_ip; } __attribute__((aligned(4 * sizeof(__u64)))); /* diff --git a/kernel/rseq.c b/kernel/rseq.c index 22b6acf1ad63..16b38c5342f9 100644 --- a/kernel/rseq.c +++ b/kernel/rseq.c @@ -130,14 +130,20 @@ static int rseq_get_rseq_cs(struct task_struct *t, struct rseq_cs *rseq_cs) urseq_cs = (struct rseq_cs __user *)ptr; if (copy_from_user(rseq_cs, urseq_cs, sizeof(*rseq_cs))) return -EFAULT; - if (rseq_cs->version > 0) - return -EINVAL; + if (rseq_cs->start_ip >= TASK_SIZE || + rseq_cs->start_ip + rseq_cs->post_commit_offset >= TASK_SIZE || + rseq_cs->abort_ip >= TASK_SIZE || + rseq_cs->version > 0) + return -EINVAL; + /* Check for overflow. */ + if (rseq_cs->start_ip + rseq_cs->post_commit_offset < rseq_cs->start_ip) + return -EINVAL; /* Ensure that abort_ip is not in the critical section. */ if (rseq_cs->abort_ip - rseq_cs->start_ip < rseq_cs->post_commit_offset) return -EINVAL; - usig = (u32 __user *)(rseq_cs->abort_ip - sizeof(u32)); + usig = (u32 __user *)(unsigned long)(rseq_cs->abort_ip - sizeof(u32)); ret = get_user(sig, usig); if (ret) return ret; @@ -146,7 +152,7 @@ static int rseq_get_rseq_cs(struct task_struct *t, struct rseq_cs *rseq_cs) printk_ratelimited(KERN_WARNING "Possible attack attempt. Unexpected rseq signature 0x%x, expecting 0x%x (pid=%d, addr=%p).\n", sig, current->rseq_sig, current->pid, usig); - return -EPERM; + return -EINVAL; } return 0; } -- cgit v1.2.3 From 0fb9a1abc8c97f858997e962694eb36b4517144e Mon Sep 17 00:00:00 2001 From: Mathieu Desnoyers Date: Mon, 9 Jul 2018 15:51:52 -0400 Subject: rseq: uapi: Update uapi comments Update rseq uapi header comments to reflect that user-space need to do thread-local loads/stores from/to the struct rseq fields. As a consequence of this added requirement, the kernel does not need to perform loads/stores with single-copy atomicity. Update the comment associated to the "flags" fields to describe more accurately that it's only useful to facilitate single-stepping through rseq critical sections with debuggers. Signed-off-by: Mathieu Desnoyers Signed-off-by: Thomas Gleixner Cc: linux-api@vger.kernel.org Cc: Peter Zijlstra Cc: "Paul E . McKenney" Cc: Boqun Feng Cc: Andy Lutomirski Cc: Dave Watson Cc: Paul Turner Cc: Andrew Morton Cc: Russell King Cc: "H . Peter Anvin" Cc: Andi Kleen Cc: Chris Lameter Cc: Ben Maurer Cc: Steven Rostedt Cc: Josh Triplett Cc: Linus Torvalds Cc: Catalin Marinas Cc: Will Deacon Cc: Michael Kerrisk Cc: Joel Fernandes Cc: "Paul E. McKenney" Cc: "H. Peter Anvin" Link: https://lkml.kernel.org/r/20180709195155.7654-4-mathieu.desnoyers@efficios.com --- include/uapi/linux/rseq.h | 69 ++++++++++++++++++++++++----------------------- kernel/rseq.c | 2 +- 2 files changed, 37 insertions(+), 34 deletions(-) (limited to 'include') diff --git a/include/uapi/linux/rseq.h b/include/uapi/linux/rseq.h index 519ad6e176d1..bf4188c13bec 100644 --- a/include/uapi/linux/rseq.h +++ b/include/uapi/linux/rseq.h @@ -67,28 +67,30 @@ struct rseq_cs { struct rseq { /* * Restartable sequences cpu_id_start field. Updated by the - * kernel, and read by user-space with single-copy atomicity - * semantics. Aligned on 32-bit. Always contains a value in the - * range of possible CPUs, although the value may not be the - * actual current CPU (e.g. if rseq is not initialized). This - * CPU number value should always be compared against the value - * of the cpu_id field before performing a rseq commit or - * returning a value read from a data structure indexed using - * the cpu_id_start value. + * kernel. Read by user-space with single-copy atomicity + * semantics. This field should only be read by the thread which + * registered this data structure. Aligned on 32-bit. Always + * contains a value in the range of possible CPUs, although the + * value may not be the actual current CPU (e.g. if rseq is not + * initialized). This CPU number value should always be compared + * against the value of the cpu_id field before performing a rseq + * commit or returning a value read from a data structure indexed + * using the cpu_id_start value. */ __u32 cpu_id_start; /* - * Restartable sequences cpu_id field. Updated by the kernel, - * and read by user-space with single-copy atomicity semantics. - * Aligned on 32-bit. Values RSEQ_CPU_ID_UNINITIALIZED and - * RSEQ_CPU_ID_REGISTRATION_FAILED have a special semantic: the - * former means "rseq uninitialized", and latter means "rseq - * initialization failed". This value is meant to be read within - * rseq critical sections and compared with the cpu_id_start - * value previously read, before performing the commit instruction, - * or read and compared with the cpu_id_start value before returning - * a value loaded from a data structure indexed using the - * cpu_id_start value. + * Restartable sequences cpu_id field. Updated by the kernel. + * Read by user-space with single-copy atomicity semantics. This + * field should only be read by the thread which registered this + * data structure. Aligned on 32-bit. Values + * RSEQ_CPU_ID_UNINITIALIZED and RSEQ_CPU_ID_REGISTRATION_FAILED + * have a special semantic: the former means "rseq uninitialized", + * and latter means "rseq initialization failed". This value is + * meant to be read within rseq critical sections and compared + * with the cpu_id_start value previously read, before performing + * the commit instruction, or read and compared with the + * cpu_id_start value before returning a value loaded from a data + * structure indexed using the cpu_id_start value. */ __u32 cpu_id; /* @@ -105,27 +107,28 @@ struct rseq { * targeted by the rseq_cs. Also needs to be set to NULL by user-space * before reclaiming memory that contains the targeted struct rseq_cs. * - * Read and set by the kernel with single-copy atomicity semantics. - * Set by user-space with single-copy atomicity semantics. Aligned - * on 64-bit. + * Read and set by the kernel. Set by user-space with single-copy + * atomicity semantics. This field should only be updated by the + * thread which registered this data structure. Aligned on 64-bit. */ LINUX_FIELD_u32_u64(rseq_cs); /* - * - RSEQ_DISABLE flag: + * Restartable sequences flags field. + * + * This field should only be updated by the thread which + * registered this data structure. Read by the kernel. + * Mainly used for single-stepping through rseq critical sections + * with debuggers. * - * Fallback fast-track flag for single-stepping. - * Set by user-space if lack of progress is detected. - * Cleared by user-space after rseq finish. - * Read by the kernel. * - RSEQ_CS_FLAG_NO_RESTART_ON_PREEMPT - * Inhibit instruction sequence block restart and event - * counter increment on preemption for this thread. + * Inhibit instruction sequence block restart on preemption + * for this thread. * - RSEQ_CS_FLAG_NO_RESTART_ON_SIGNAL - * Inhibit instruction sequence block restart and event - * counter increment on signal delivery for this thread. + * Inhibit instruction sequence block restart on signal + * delivery for this thread. * - RSEQ_CS_FLAG_NO_RESTART_ON_MIGRATE - * Inhibit instruction sequence block restart and event - * counter increment on migration for this thread. + * Inhibit instruction sequence block restart on migration for + * this thread. */ __u32 flags; } __attribute__((aligned(4 * sizeof(__u64)))); diff --git a/kernel/rseq.c b/kernel/rseq.c index 2c8463acb50d..2a7748675be7 100644 --- a/kernel/rseq.c +++ b/kernel/rseq.c @@ -201,7 +201,7 @@ static int clear_rseq_cs(struct task_struct *t) * of code outside of the rseq assembly block. This performs * a lazy clear of the rseq_cs field. * - * Set rseq_cs to NULL with single-copy atomicity. + * Set rseq_cs to NULL. */ return put_user(0UL, &t->rseq->rseq_cs); } -- cgit v1.2.3 From ec9c82e03a744e5698bd95eab872855861a821fa Mon Sep 17 00:00:00 2001 From: Mathieu Desnoyers Date: Mon, 9 Jul 2018 15:51:53 -0400 Subject: rseq: uapi: Declare rseq_cs field as union, update includes Declaring the rseq_cs field as a union between __u64 and two __u32 allows both 32-bit and 64-bit kernels to read the full __u64, and therefore validate that a 32-bit user-space cleared the upper 32 bits, thus ensuring a consistent behavior between native 32-bit kernels and 32-bit compat tasks on 64-bit kernels. Check that the rseq_cs value read is < TASK_SIZE. The asm/byteorder.h header needs to be included by rseq.h, now that it is not using linux/types_32_64.h anymore. Considering that only __32 and __u64 types are declared in linux/rseq.h, the linux/types.h header should always be included for both kernel and user-space code: including stdint.h is just for u64 and u32, which are not used in this header at all. Use copy_from_user()/clear_user() to interact with a 64-bit field, because arm32 does not implement 64-bit __get_user, and ppc32 does not 64-bit get_user. Considering that the rseq_cs pointer does not need to be loaded/stored with single-copy atomicity from the kernel anymore, we can simply use copy_from_user()/clear_user(). Signed-off-by: Mathieu Desnoyers Signed-off-by: Thomas Gleixner Cc: linux-api@vger.kernel.org Cc: Peter Zijlstra Cc: "Paul E . McKenney" Cc: Boqun Feng Cc: Andy Lutomirski Cc: Dave Watson Cc: Paul Turner Cc: Andrew Morton Cc: Russell King Cc: "H . Peter Anvin" Cc: Andi Kleen Cc: Chris Lameter Cc: Ben Maurer Cc: Steven Rostedt Cc: Josh Triplett Cc: Linus Torvalds Cc: Catalin Marinas Cc: Will Deacon Cc: Michael Kerrisk Cc: Joel Fernandes Cc: "Paul E. McKenney" Cc: "H. Peter Anvin" Link: https://lkml.kernel.org/r/20180709195155.7654-5-mathieu.desnoyers@efficios.com --- include/uapi/linux/rseq.h | 27 +++++++++++++++++++-------- kernel/rseq.c | 15 +++++++++------ tools/testing/selftests/rseq/rseq.h | 11 ++++++++++- 3 files changed, 38 insertions(+), 15 deletions(-) (limited to 'include') diff --git a/include/uapi/linux/rseq.h b/include/uapi/linux/rseq.h index bf4188c13bec..9a402fdb60e9 100644 --- a/include/uapi/linux/rseq.h +++ b/include/uapi/linux/rseq.h @@ -10,13 +10,8 @@ * Copyright (c) 2015-2018 Mathieu Desnoyers */ -#ifdef __KERNEL__ -# include -#else -# include -#endif - -#include +#include +#include enum rseq_cpu_id_state { RSEQ_CPU_ID_UNINITIALIZED = -1, @@ -111,7 +106,23 @@ struct rseq { * atomicity semantics. This field should only be updated by the * thread which registered this data structure. Aligned on 64-bit. */ - LINUX_FIELD_u32_u64(rseq_cs); + union { + __u64 ptr64; +#ifdef __LP64__ + __u64 ptr; +#else + struct { +#if (defined(__BYTE_ORDER) && (__BYTE_ORDER == __BIG_ENDIAN)) || defined(__BIG_ENDIAN) + __u32 padding; /* Initialized to zero. */ + __u32 ptr32; +#else /* LITTLE */ + __u32 ptr32; + __u32 padding; /* Initialized to zero. */ +#endif /* ENDIAN */ + } ptr; +#endif + } rseq_cs; + /* * Restartable sequences flags field. * diff --git a/kernel/rseq.c b/kernel/rseq.c index 2a7748675be7..c6242d8594dc 100644 --- a/kernel/rseq.c +++ b/kernel/rseq.c @@ -115,19 +115,20 @@ static int rseq_reset_rseq_cpu_id(struct task_struct *t) static int rseq_get_rseq_cs(struct task_struct *t, struct rseq_cs *rseq_cs) { struct rseq_cs __user *urseq_cs; - unsigned long ptr; + u64 ptr; u32 __user *usig; u32 sig; int ret; - ret = get_user(ptr, &t->rseq->rseq_cs); - if (ret) - return ret; + if (copy_from_user(&ptr, &t->rseq->rseq_cs.ptr64, sizeof(ptr))) + return -EFAULT; if (!ptr) { memset(rseq_cs, 0, sizeof(*rseq_cs)); return 0; } - urseq_cs = (struct rseq_cs __user *)ptr; + if (ptr >= TASK_SIZE) + return -EINVAL; + urseq_cs = (struct rseq_cs __user *)(unsigned long)ptr; if (copy_from_user(rseq_cs, urseq_cs, sizeof(*rseq_cs))) return -EFAULT; @@ -203,7 +204,9 @@ static int clear_rseq_cs(struct task_struct *t) * * Set rseq_cs to NULL. */ - return put_user(0UL, &t->rseq->rseq_cs); + if (clear_user(&t->rseq->rseq_cs.ptr64, sizeof(t->rseq->rseq_cs.ptr64))) + return -EFAULT; + return 0; } /* diff --git a/tools/testing/selftests/rseq/rseq.h b/tools/testing/selftests/rseq/rseq.h index a4684112676c..f2073cfa4448 100644 --- a/tools/testing/selftests/rseq/rseq.h +++ b/tools/testing/selftests/rseq/rseq.h @@ -133,6 +133,15 @@ static inline uint32_t rseq_current_cpu(void) return cpu; } +static inline void rseq_clear_rseq_cs(void) +{ +#ifdef __LP64__ + __rseq_abi.rseq_cs.ptr = 0; +#else + __rseq_abi.rseq_cs.ptr.ptr32 = 0; +#endif +} + /* * rseq_prepare_unload() should be invoked by each thread using rseq_finish*() * at least once between their last rseq_finish*() and library unload of the @@ -143,7 +152,7 @@ static inline uint32_t rseq_current_cpu(void) */ static inline void rseq_prepare_unload(void) { - __rseq_abi.rseq_cs = 0; + rseq_clear_rseq_cs(); } #endif /* RSEQ_H_ */ -- cgit v1.2.3 From 4f4c0acdf4652a964da869d578a3c8bf6df14ce2 Mon Sep 17 00:00:00 2001 From: Mathieu Desnoyers Date: Mon, 9 Jul 2018 15:51:54 -0400 Subject: rseq: Remove unused types_32_64.h uapi header This header was introduced in the 4.18 merge window, and rseq does not need it anymore. Nuke it before the final release. Signed-off-by: Mathieu Desnoyers Signed-off-by: Thomas Gleixner Cc: linux-api@vger.kernel.org Cc: Peter Zijlstra Cc: "Paul E . McKenney" Cc: Boqun Feng Cc: Andy Lutomirski Cc: Dave Watson Cc: Paul Turner Cc: Andrew Morton Cc: Russell King Cc: "H . Peter Anvin" Cc: Andi Kleen Cc: Chris Lameter Cc: Ben Maurer Cc: Steven Rostedt Cc: Josh Triplett Cc: Linus Torvalds Cc: Catalin Marinas Cc: Will Deacon Cc: Michael Kerrisk Cc: Joel Fernandes Cc: "Paul E. McKenney" Cc: "H. Peter Anvin" Link: https://lkml.kernel.org/r/20180709195155.7654-6-mathieu.desnoyers@efficios.com --- include/uapi/linux/types_32_64.h | 50 ---------------------------------------- 1 file changed, 50 deletions(-) delete mode 100644 include/uapi/linux/types_32_64.h (limited to 'include') diff --git a/include/uapi/linux/types_32_64.h b/include/uapi/linux/types_32_64.h deleted file mode 100644 index 0a87ace34a57..000000000000 --- a/include/uapi/linux/types_32_64.h +++ /dev/null @@ -1,50 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0+ WITH Linux-syscall-note */ -#ifndef _UAPI_LINUX_TYPES_32_64_H -#define _UAPI_LINUX_TYPES_32_64_H - -/* - * linux/types_32_64.h - * - * Integer type declaration for pointers across 32-bit and 64-bit systems. - * - * Copyright (c) 2015-2018 Mathieu Desnoyers - */ - -#ifdef __KERNEL__ -# include -#else -# include -#endif - -#include - -#ifdef __BYTE_ORDER -# if (__BYTE_ORDER == __BIG_ENDIAN) -# define LINUX_BYTE_ORDER_BIG_ENDIAN -# else -# define LINUX_BYTE_ORDER_LITTLE_ENDIAN -# endif -#else -# ifdef __BIG_ENDIAN -# define LINUX_BYTE_ORDER_BIG_ENDIAN -# else -# define LINUX_BYTE_ORDER_LITTLE_ENDIAN -# endif -#endif - -#ifdef __LP64__ -# define LINUX_FIELD_u32_u64(field) __u64 field -# define LINUX_FIELD_u32_u64_INIT_ONSTACK(field, v) field = (intptr_t)v -#else -# ifdef LINUX_BYTE_ORDER_BIG_ENDIAN -# define LINUX_FIELD_u32_u64(field) __u32 field ## _padding, field -# define LINUX_FIELD_u32_u64_INIT_ONSTACK(field, v) \ - field ## _padding = 0, field = (intptr_t)v -# else -# define LINUX_FIELD_u32_u64(field) __u32 field, field ## _padding -# define LINUX_FIELD_u32_u64_INIT_ONSTACK(field, v) \ - field = (intptr_t)v, field ## _padding = 0 -# endif -#endif - -#endif /* _UAPI_LINUX_TYPES_32_64_H */ -- cgit v1.2.3 From a90744bac57c3c07d0d4422af62f3e44549ade30 Mon Sep 17 00:00:00 2001 From: Nicholas Piggin Date: Fri, 13 Jul 2018 16:59:03 -0700 Subject: mm: allow arch to supply p??_free_tlb functions The mmu_gather APIs keep track of the invalidated address range including the span covered by invalidated page table pages. Ranges covered by page tables but not ptes (and therefore no TLBs) still need to be invalidated because some architectures (x86) can cache intermediate page table entries, and invalidate those with normal TLB invalidation instructions to be almost-backward-compatible. Architectures which don't cache intermediate page table entries, or which invalidate these caches separately from TLB invalidation, do not require TLB invalidation range expanded over page tables. Allow architectures to supply their own p??_free_tlb functions, which can avoid the __tlb_adjust_range. Link: http://lkml.kernel.org/r/20180703013131.2807-1-npiggin@gmail.com Signed-off-by: Nicholas Piggin Reviewed-by: Andrew Morton Cc: "Aneesh Kumar K. V" Cc: Minchan Kim Cc: Mel Gorman Cc: Nadav Amit Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/asm-generic/tlb.h | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'include') diff --git a/include/asm-generic/tlb.h b/include/asm-generic/tlb.h index faddde44de8c..3063125197ad 100644 --- a/include/asm-generic/tlb.h +++ b/include/asm-generic/tlb.h @@ -265,33 +265,41 @@ static inline void tlb_remove_check_page_size_change(struct mmu_gather *tlb, * For now w.r.t page table cache, mark the range_size as PAGE_SIZE */ +#ifndef pte_free_tlb #define pte_free_tlb(tlb, ptep, address) \ do { \ __tlb_adjust_range(tlb, address, PAGE_SIZE); \ __pte_free_tlb(tlb, ptep, address); \ } while (0) +#endif +#ifndef pmd_free_tlb #define pmd_free_tlb(tlb, pmdp, address) \ do { \ __tlb_adjust_range(tlb, address, PAGE_SIZE); \ __pmd_free_tlb(tlb, pmdp, address); \ } while (0) +#endif #ifndef __ARCH_HAS_4LEVEL_HACK +#ifndef pud_free_tlb #define pud_free_tlb(tlb, pudp, address) \ do { \ __tlb_adjust_range(tlb, address, PAGE_SIZE); \ __pud_free_tlb(tlb, pudp, address); \ } while (0) #endif +#endif #ifndef __ARCH_HAS_5LEVEL_HACK +#ifndef p4d_free_tlb #define p4d_free_tlb(tlb, pudp, address) \ do { \ __tlb_adjust_range(tlb, address, PAGE_SIZE); \ __p4d_free_tlb(tlb, pudp, address); \ } while (0) #endif +#endif #define tlb_migrate_finish(mm) do {} while (0) -- cgit v1.2.3