- 03 Jul, 2019 40 commits
-
-
Jason Gunthorpe authored
commit 641114d2 upstream. gcc 9 now does allocation size tracking and thinks that passing the member of a union and then accessing beyond that member's bounds is an overflow. Instead of using the union member, use the entire union with a cast to get to the sockaddr. gcc will now know that the memory extends the full size of the union. Signed-off-by:
Jason Gunthorpe <jgg@mellanox.com> Signed-off-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Will Deacon authored
commit 42750351 upstream. The architecture implementations of 'arch_futex_atomic_op_inuser()' and 'futex_atomic_cmpxchg_inatomic()' are permitted to return only -EFAULT, -EAGAIN or -ENOSYS in the case of failure. Update the comments in the asm-generic/ implementation and also a stray reference in the robust futex documentation. Signed-off-by:
Will Deacon <will.deacon@arm.com> Signed-off-by:
-
Daniel Borkmann authored
commit 34b8ab09 upstream. Since ARMv8.1 supplement introduced LSE atomic instructions back in 2016, lets add support for STADD and use that in favor of LDXR / STXR loop for the XADD mapping if available. STADD is encoded as an alias for LDADD with XZR as the destination register, therefore add LDADD to the instruction encoder along with STADD as special case and use it in the JIT for CPUs that advertise LSE atomics in CPUID register. If immediate offset in the BPF XADD insn is 0, then use dst register directly instead of temporary one. Signed-off-by:
Daniel Borkmann <daniel@iogearbox.net> Acked-by:
Jean-Philippe Brucker <jean-philippe.brucker@arm.com> Acked-by:
Alexei Starovoitov <ast@kernel.org> Signed-off-by:
-
Will Deacon authored
commit 8e4e0ac0 upstream. Returning an error code from futex_atomic_cmpxchg_inatomic() indicates that the caller should not make any use of *uval, and should instead act upon on the value of the error code. Although this is implemented correctly in our futex code, we needlessly copy uninitialised stack to *uval in the error case, which can easily be avoided. Signed-off-by:
-
Martin KaFai Lau authored
commit 4ac30c4b upstream. __udp6_lib_err() may be called when handling icmpv6 message. For example, the icmpv6 toobig(type=2). __udp6_lib_lookup() is then called which may call reuseport_select_sock(). reuseport_select_sock() will call into a bpf_prog (if there is one). reuseport_select_sock() is expecting the skb->data pointing to the transport header (udphdr in this case). For example, run_bpf_filter() is pulling the transport header. However, in the __udp6_lib_err() path, the skb->data is pointing to the ipv6hdr instead of the udphdr. One option is to pull and push the ipv6hdr in __udp6_lib_err(). Instead of doing this, this patch follows how the original commit 538950a1 ("soreuseport: setsockopt SO_ATTACH_REUSEPORT_[CE]BPF") was done in IPv4, which has passed a NULL skb pointer to reuseport_select_sock(). Fixes: 538950a1 ("soreuseport: setsockopt SO_ATTACH_REUSEPORT_[CE]BPF") Cc: Craig Gallek <kraig@google.com> Signed-off-by:
Martin KaFai Lau <kafai@fb.com> Acked-by:
Song Liu <songliubraving@fb.com> Acked-by:
Craig Gallek <kraig@google.com> Signed-off-by:
-
Martin KaFai Lau authored
commit 257a525f upstream. When the commit a6024562 ("udp: Add GRO functions to UDP socket") added udp[46]_lib_lookup_skb to the udp_gro code path, it broke the reuseport_select_sock() assumption that skb->data is pointing to the transport header. This patch follows an earlier __udp6_lib_err() fix by passing a NULL skb to avoid calling the reuseport's bpf_prog. Fixes: a6024562 ("udp: Add GRO functions to UDP socket") Cc: Tom Herbert <tom@herbertland.com> Signed-off-by:
-
Daniel Borkmann authored
commit 983695fa upstream. Intention of cgroup bind/connect/sendmsg BPF hooks is to act transparently to applications as also stated in original motivation in 7828f20e ("Merge branch 'bpf-cgroup-bind-connect'"). When recently integrating the latter two hooks into Cilium to enable host based load-balancing with Kubernetes, I ran into the issue that pods couldn't start up as DNS got broken. Kubernetes typically sets up DNS as a service and is thus subject to load-balancing. Upon further debugging, it turns out that the cgroupv2 sendmsg BPF hooks API is currently insufficient and thus not usable as-is for standard applications shipped with most distros. To break down the issue we ran into with a simple example: # cat /etc/resolv.conf nameserver 147.75.207.207 nameserver 147.75.207.208 For the purpose of a simple test, we set up above IPs as service IPs and transparently redirect traffic to a different DNS backend server for that node: # cilium service list ID Frontend Backend 1 147.75.207.207:53 1 => 8.8.8.8:53 2 147.75.207.208:53 1 => 8.8.8.8:53 The attached BPF program is basically selecting one of the backends if the service IP/port matches on the cgroup hook. DNS breaks here, because the hooks are not transparent enough to applications which have built-in msg_name address checks: # nslookup 1.1.1.1 ;; reply from unexpected source: 8.8.8.8#53, expected 147.75.207.207#53 ;; reply from unexpected source: 8.8.8.8#53, expected 147.75.207.208#53 ;; reply from unexpected source: 8.8.8.8#53, expected 147.75.207.207#53 [...] ;; connection timed out; no servers could be reached # dig 1.1.1.1 ;; reply from unexpected source: 8.8.8.8#53, expected 147.75.207.207#53 ;; reply from unexpected source: 8.8.8.8#53, expected 147.75.207.208#53 ;; reply from unexpected source: 8.8.8.8#53, expected 147.75.207.207#53 [...] ; <<>> DiG 9.11.3-1ubuntu1.7-Ubuntu <<>> 1.1.1.1 ;; global options: +cmd ;; connection timed out; no servers could be reached For comparison, if none of the service IPs is used, and we tell nslookup to use 8.8.8.8 directly it works just fine, of course: # nslookup 1.1.1.1 8.8.8.8 1.1.1.1.in-addr.arpa name = one.one.one.one. In order to fix this and thus act more transparent to the application, this needs reverse translation on recvmsg() side. A minimal fix for this API is to add similar recvmsg() hooks behind the BPF cgroups static key such that the program can track state and replace the current sockaddr_in{,6} with the original service IP. From BPF side, this basically tracks the service tuple plus socket cookie in an LRU map where the reverse NAT can then be retrieved via map value as one example. Side-note: the BPF cgroups static key should be converted to a per-hook static key in future. Same example after this fix: # cilium service list ID Frontend Backend 1 147.75.207.207:53 1 => 8.8.8.8:53 2 147.75.207.208:53 1 => 8.8.8.8:53 Lookups work fine now: # nslookup 1.1.1.1 1.1.1.1.in-addr.arpa name = one.one.one.one. Authoritative answers can be found from: # dig 1.1.1.1 ; <<>> DiG 9.11.3-1ubuntu1.7-Ubuntu <<>> 1.1.1.1 ;; global options: +cmd ;; Got answer: ;; ->>HEADER<<- opcode: QUERY, status: NXDOMAIN, id: 51550 ;; flags: qr rd ra ad; QUERY: 1, ANSWER: 0, AUTHORITY: 1, ADDITIONAL: 1 ;; OPT PSEUDOSECTION: ; EDNS: version: 0, flags:; udp: 512 ;; QUESTION SECTION: ;1.1.1.1. IN A ;; AUTHORITY SECTION: . 23426 IN SOA a.root-servers.net. nstld.verisign-grs.com. 2019052001 1800 900 604800 86400 ;; Query time: 17 msec ;; SERVER: 147.75.207.207#53(147.75.207.207) ;; WHEN: Tue May 21 12:59:38 UTC 2019 ;; MSG SIZE rcvd: 111 And from an actual packet level it shows that we're using the back end server when talking via 147.75.207.20{7,8} front end: # tcpdump -i any udp [...] 12:59:52.698732 IP foo.42011 > google-public-dns-a.google.com.domain: 18803+ PTR? 1.1.1.1.in-addr.arpa. (38) 12:59:52.698735 IP foo.42011 > google-public-dns-a.google.com.domain: 18803+ PTR? 1.1.1.1.in-addr.arpa. (38) 12:59:52.701208 IP google-public-dns-a.google.com.domain > foo.42011: 18803 1/0/0 PTR one.one.one.one. (67) 12:59:52.701208 IP google-public-dns-a.google.com.domain > foo.42011: 18803 1/0/0 PTR one.one.one.one. (67) [...] In order to be flexible and to have same semantics as in sendmsg BPF programs, we only allow return codes in [1,1] range. In the sendmsg case the program is called if msg->msg_name is present which can be the case in both, connected and unconnected UDP. The former only relies on the sockaddr_in{,6} passed via connect(2) if passed msg->msg_name was NULL. Therefore, on recvmsg side, we act in similar way to call into the BPF program whenever a non-NULL msg->msg_name was passed independent of sk->sk_state being TCP_ESTABLISHED or not. Note that for TCP case, the msg->msg_name is ignored in the regular recvmsg path and therefore not relevant. For the case of ip{,v6}_recv_error() paths, picked up via MSG_ERRQUEUE, the hook is not called. This is intentional as it aligns with the same semantics as in case of TCP cgroup BPF hooks right now. This might be better addressed in future through a different bpf_attach_type such that this case can be distinguished from the regular recvmsg paths, for example. Fixes: 1cedee13 ("bpf: Hooks for sys_sendmsg") Signed-off-by:
Andrey Ignatov <rdna@fb.com> Acked-by:
Martynas Pumputis <m@lambda.lt> Signed-off-by:
-
Matt Mullins authored
commit 9594dc3c upstream. BPF_PROG_TYPE_RAW_TRACEPOINTs can be executed nested on the same CPU, as they do not increment bpf_prog_active while executing. This enables three levels of nesting, to support - a kprobe or raw tp or perf event, - another one of the above that irq context happens to call, and - another one in nmi context (at most one of which may be a kprobe or perf event). Fixes: 20b9d7ac ("bpf: avoid excessive stack usage for perf_sample_data") Signed-off-by:
Matt Mullins <mmullins@fb.com> Acked-by:
Andrii Nakryiko <andriin@fb.com> Acked-by:
-
Jonathan Lemon authored
commit da2577fd upstream. If the leftmost parent node of the tree has does not have a child on the left side, then trie_get_next_key (and bpftool map dump) will not look at the child on the right. This leads to the traversal missing elements. Lookup is not affected. Update selftest to handle this case. Reproducer: bpftool map create /sys/fs/bpf/lpm type lpm_trie key 6 \ value 1 entries 256 name test_lpm flags 1 bpftool map update pinned /sys/fs/bpf/lpm key 8 0 0 0 0 0 value 1 bpftool map update pinned /sys/fs/bpf/lpm key 16 0 0 0 0 128 value 2 bpftool map dump pinned /sys/fs/bpf/lpm Returns only 1 element. (2 expected) Fixes: b471f2f1 ("bpf: implement MAP_GET_NEXT_KEY command for LPM_TRIE") Signed-off-by:
Jonathan Lemon <jonathan.lemon@gmail.com> Acked-by:
-
Martynas Pumputis authored
commit b1d6c15b upstream. Previously, the BPF_FIB_LOOKUP_{DIRECT,OUTPUT} flags in the BPF UAPI were defined with the help of BIT macro. This had the following issues: - In order to use any of the flags, a user was required to depend on <linux/bits.h>. - No other flag in bpf.h uses the macro, so it seems that an unwritten convention is to use (1 << (nr)) to define BPF-related flags. Fixes: 87f5fc7e ("bpf: Provide helper to do forwarding lookups in kernel FIB table") Signed-off-by:
-
Dmitry Bogdanov authored
[ Upstream commit 48dd73d0 ] In configuration of vlan over bridge over aquantia device it was found that vlan tagged traffic is dropped on chip. The reason is that bridge device enables promisc mode, but in atlantic chip vlan filters will still apply. So we have to corellate promisc settings with vlan configuration. The solution is to track in a separate state variable the need of vlan forced promisc. And also consider generic promisc configuration when doing vlan filter config. Fixes: 7975d2af ("net: aquantia: add support of rx-vlan-filter offload") Signed-off-by:
Dmitry Bogdanov <dmitry.bogdanov@aquantia.com> Signed-off-by:
Igor Russkikh <igor.russkikh@aquantia.com> Signed-off-by:
David S. Miller <davem@davemloft.net> Signed-off-by:
-
Fei Li authored
[ Upstream commit 72b319dc ] Currently after setting tap0 link up, the tun code wakes tx/rx waited queues up in tun_net_open() when .ndo_open() is called, however the IFF_UP flag has not been set yet. If there's already a wait queue, it would fail to transmit when checking the IFF_UP flag in tun_sendmsg(). Then the saving vhost_poll_start() will add the wq into wqh until it is waken up again. Although this works when IFF_UP flag has been set when tun_chr_poll detects; this is not true if IFF_UP flag has not been set at that time. Sadly the latter case is a fatal error, as the wq will never be waken up in future unless later manually setting link up on purpose. Fix this by moving the wakeup process into the NETDEV_UP event notifying process, this makes sure IFF_UP has been set before all waited queues been waken up. Signed-off-by:
Fei Li <lifei.shirley@bytedance.com> Acked-by:
Jason Wang <jasowang@redhat.com> Signed-off-by:
-
Xin Long authored
[ Upstream commit 4f07b80c ] This patch is to fix an uninit-value issue, reported by syzbot: BUG: KMSAN: uninit-value in memchr+0xce/0x110 lib/string.c:981 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x191/0x1f0 lib/dump_stack.c:113 kmsan_report+0x130/0x2a0 mm/kmsan/kmsan.c:622 __msan_warning+0x75/0xe0 mm/kmsan/kmsan_instr.c:310 memchr+0xce/0x110 lib/string.c:981 string_is_valid net/tipc/netlink_compat.c:176 [inline] tipc_nl_compat_bearer_disable+0x2a1/0x480 net/tipc/netlink_compat.c:449 __tipc_nl_compat_doit net/tipc/netlink_compat.c:327 [inline] tipc_nl_compat_doit+0x3ac/0xb00 net/tipc/netlink_compat.c:360 tipc_nl_compat_handle net/tipc/netlink_compat.c:1178 [inline] tipc_nl_compat_recv+0x1b1b/0x27b0 net/tipc/netlink_compat.c:1281 TLV_GET_DATA_LEN() may return a negtive int value, which will be used as size_t (becoming a big unsigned long) passed into memchr, cause this issue. Similar to what it does in tipc_nl_compat_bearer_enable(), this fix is to return -EINVAL when TLV_GET_DATA_LEN() is negtive in tipc_nl_compat_bearer_disable(), as well as in tipc_nl_compat_link_stat_dump() and tipc_nl_compat_link_reset_stats(). v1->v2: - add the missing Fixes tags per Eric's request. Fixes: 0762216c ("tipc: fix uninit-value in tipc_nl_compat_bearer_enable") Fixes: 8b66fee7 ("tipc: fix uninit-value in tipc_nl_compat_link_reset_stats") Reported-by: syzbot+30eaa8bf392f7fafffaf@syzkaller.appspotmail.com Signed-off-by:
Xin Long <lucien.xin@gmail.com> Signed-off-by:
-
Xin Long authored
[ Upstream commit c492d4c7 ] This patch is to fix a dst defcnt leak, which can be reproduced by doing: # ip net a c; ip net a s; modprobe tipc # ip net e s ip l a n eth1 type veth peer n eth1 netns c # ip net e c ip l s lo up; ip net e c ip l s eth1 up # ip net e s ip l s lo up; ip net e s ip l s eth1 up # ip net e c ip a a 1.1.1.2/8 dev eth1 # ip net e s ip a a 1.1.1.1/8 dev eth1 # ip net e c tipc b e m udp n u1 localip 1.1.1.2 # ip net e s tipc b e m udp n u1 localip 1.1.1.1 # ip net d c; ip net d s; rmmod tipc and it will get stuck and keep logging the error: unregister_netdevice: waiting for lo to become free. Usage count = 1 The cause is that a dst is held by the udp sock's sk_rx_dst set on udp rx path with udp_early_demux == 1, and this dst (eventually holding lo dev) can't be released as bearer's removal in tipc pernet .exit happens after lo dev's removal, default_device pernet .exit. "There are two distinct types of pernet_operations recognized: subsys and device. At creation all subsys init functions are called before device init functions, and at destruction all device exit functions are called before subsys exit function." So by calling register_pernet_device instead to register tipc_net_ops, the pernet .exit() will be invoked earlier than loopback dev's removal when a netns is being destroyed, as fou/gue does. Note that vxlan and geneve udp tunnels don't have this issue, as the udp sock is released in their device ndo_stop(). This fix is also necessary for tipc dst_cache, which will hold dsts on tx path and I will introduce in my next patch. Reported-by:
Li Shuang <shuali@redhat.com> Signed-off-by:
Jon Maloy <jon.maloy@ericsson.com> Signed-off-by:
-
YueHaibing authored
[ Upstream commit ee429742 ] We should rather have vlan_tci filled all the way down to the transmitting netdevice and let it do the hw/sw vlan implementation. Suggested-by:
Jiri Pirko <jiri@resnulli.us> Signed-off-by:
YueHaibing <yuehaibing@huawei.com> Signed-off-by:
-
Xin Long authored
[ Upstream commit 25bff6d5 ] Now in sctp_endpoint_init(), it holds the sk then creates auth shkey. But when the creation fails, it doesn't release the sk, which causes a sk defcnf leak, Here to fix it by only holding the sk when auth shkey is created successfully. Fixes: a29a5bd4 ("[SCTP]: Implement SCTP-AUTH initializations.") Reported-by: syzbot+afabda3890cc2f765041@syzkaller.appspotmail.com Reported-by: syzbot+276ca1c77a19977c0130@syzkaller.appspotmail.com Signed-off-by:
Neil Horman <nhorman@redhat.com> Signed-off-by:
-
Dirk van der Merwe authored
[ Upstream commit 9354544c ] With commit 94850257 ("tls: Fix tls_device handling of partial records") a new path was introduced to cleanup partial records during sk_proto_close. This path does not handle the SW KTLS tx_list cleanup. This is unnecessary though since the free_resources calls for both SW and offload paths will cleanup a partial record. The visible effect is the following warning, but this bug also causes a page double free. WARNING: CPU: 7 PID: 4000 at net/core/stream.c:206 sk_stream_kill_queues+0x103/0x110 RIP: 0010:sk_stream_kill_queues+0x103/0x110 RSP: 0018:ffffb6df87e07bd0 EFLAGS: 00010206 RAX: 0000000000000000 RBX: ffff8c21db4971c0 RCX: 0000000000000007 RDX: ffffffffffffffa0 RSI: 000000000000001d RDI: ffff8c21db497270 RBP: ffff8c21db497270 R08: ffff8c29f4748600 R09: 000000010020001a R10: ffffb6df87e07aa0 R11: ffffffff9a445600 R12: 0000000000000007 R13: 0000000000000000 R14: ffff8c21f03f2900 R15: ffff8c21f03b8df0 Call Trace: inet_csk_destroy_sock+0x55/0x100 tcp_close+0x25d/0x400 ? tcp_check_oom+0x120/0x120 tls_sk_proto_close+0x127/0x1c0 inet_release+0x3c/0x60 __sock_release+0x3d/0xb0 sock_close+0x11/0x20 __fput+0xd8/0x210 task_work_run+0x84/0xa0 do_exit+0x2dc/0xb90 ? release_sock+0x43/0x90 do_group_exit+0x3a/0xa0 get_signal+0x295/0x720 do_signal+0x36/0x610 ? SYSC_recvfrom+0x11d/0x130 exit_to_usermode_loop+0x69/0xb0 do_syscall_64+0x173/0x180 entry_SYSCALL_64_after_hwframe+0x3d/0xa2 RIP: 0033:0x7fe9b9abc10d RSP: 002b:00007fe9b19a1d48 EFLAGS: 00000246 ORIG_RAX: 00000000000000ca RAX: fffffffffffffe00 RBX: 0000000000000006 RCX: 00007fe9b9abc10d RDX: 0000000000000002 RSI: 0000000000000080 RDI: 00007fe948003430 RBP: 00007fe948003410 R08: 00007fe948003430 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 00005603739d9080 R13: 00007fe9b9ab9f90 R14: 00007fe948003430 R15: 0000000000000000 Fixes: 94850257 ("tls: Fix tls_device handling of partial records") Signed-off-by:
Dirk van der Merwe <dirk.vandermerwe@netronome.com> Signed-off-by:
Jakub Kicinski <jakub.kicinski@netronome.com> Signed-off-by:
-
Roland Hii authored
[ Upstream commit d0bb82fd ] When transmitting certain PTP frames, e.g. SYNC and DELAY_REQ, the PTP daemon, e.g. ptp4l, is polling the driver for the frame transmit hardware timestamp. The polling will most likely timeout if the tx coalesce is enabled due to the Interrupt-on-Completion (IC) bit is not set in tx descriptor for those frames. This patch will ignore the tx coalesce parameter and set the IC bit when transmitting PTP frames which need to report out the frame transmit hardware timestamp to user space. Fixes: f748be53 ("net: stmmac: Rework coalesce timer and fix multi-queue races") Signed-off-by:
Roland Hii <roland.king.guan.hii@intel.com> Signed-off-by:
Ong Boon Leong <boon.leong.ong@intel.com> Signed-off-by:
Voon Weifeng <weifeng.voon@intel.com> Signed-off-by:
-
Roland Hii authored
[ Upstream commit a1e5388b ] When ADDSUB bit is set, the system time seconds field is calculated as the complement of the seconds part of the update value. For example, if 3.000000001 seconds need to be subtracted from the system time, this field is calculated as 2^32 - 3 = 4294967296 - 3 = 0x100000000 - 3 = 0xFFFFFFFD Previously, the 0x100000000 is mistakenly written as 100000000. This is further simplified from sec = (0x100000000ULL - sec); to sec = -sec; Fixes: ba1ffd74 ("stmmac: fix PTP support for GMAC4") Signed-off-by:
-
JingYi Hou authored
[ Upstream commit d0bae4a0 ] In sock_getsockopt(), 'optlen' is fetched the first time from userspace. 'len < 0' is then checked. Then in condition 'SO_MEMINFO', 'optlen' is fetched the second time from userspace. If change it between two fetches may cause security problems or unexpected behaivor, and there is no reason to fetch it a second time. To fix this, we need to remove the second fetch. Signed-off-by:
JingYi Hou <houjingyi647@gmail.com> Signed-off-by:
-
Eric Dumazet authored
[ Upstream commit 55655e3d ] syzbot found we can leak memory in packet_set_ring(), if user application provides buggy parameters. Fixes: 7f953ab2 ("af_packet: TX_RING support for TPACKET_V3") Signed-off-by:
Eric Dumazet <edumazet@google.com> Cc: Sowmini Varadhan <sowmini.varadhan@oracle.com> Reported-by:
syzbot <syzkaller@googlegroups.com> Signed-off-by:
-
Stephen Suryaputra authored
[ Upstream commit 38c73529 ] In commit 19e4e768 ("ipv4: Fix raw socket lookup for local traffic"), the dif argument to __raw_v4_lookup() is coming from the returned value of inet_iif() but the change was done only for the first lookup. Subsequent lookups in the while loop still use skb->dev->ifIndex. Fixes: 19e4e768 ("ipv4: Fix raw socket lookup for local traffic") Signed-off-by:
Stephen Suryaputra <ssuryaextr@gmail.com> Reviewed-by:
David Ahern <dsahern@gmail.com> Signed-off-by:
-
YueHaibing authored
[ Upstream commit 30d8177e ] We build vlan on top of bonding interface, which vlan offload is off, bond mode is 802.3ad (LACP) and xmit_hash_policy is BOND_XMIT_POLICY_ENCAP34. Because vlan tx offload is off, vlan tci is cleared and skb push the vlan header in validate_xmit_vlan() while sending from vlan devices. Then in bond_xmit_hash, __skb_flow_dissect() fails to get information from protocol headers encapsulated within vlan, because 'nhoff' is points to IP header, so bond hashing is based on layer 2 info, which fails to distribute packets across slaves. This patch always enable bonding's vlan tx offload, pass the vlan packets to the slave devices with vlan tci, let them to handle vlan implementation. Fixes: 278339a4 ("bonding: propogate vlan_features to bonding master") Suggested-by:
Jiri Pirko <jiri@mellanox.com> Signed-off-by:
-
Neil Horman authored
[ Upstream commit 89ed5b51 ] When an application is run that: a) Sets its scheduler to be SCHED_FIFO and b) Opens a memory mapped AF_PACKET socket, and sends frames with the MSG_DONTWAIT flag cleared, its possible for the application to hang forever in the kernel. This occurs because when waiting, the code in tpacket_snd calls schedule, which under normal circumstances allows other tasks to run, including ksoftirqd, which in some cases is responsible for freeing the transmitted skb (which in AF_PACKET calls a destructor that flips the status bit of the transmitted frame back to available, allowing the transmitting task to complete). However, when the calling application is SCHED_FIFO, its priority is such that the schedule call immediately places the task back on the cpu, preventing ksoftirqd from freeing the skb, which in turn prevents the transmitting task from detecting that the transmission is complete. We can fix this by converting the schedule call to a completion mechanism. By using a completion queue, we force the calling task, when it detects there are no more frames to send, to schedule itself off the cpu until such time as the last transmitted skb is freed, allowing forward progress to be made. Tested by myself and the reporter, with good results Change Notes: V1->V2: Enhance the sleep logic to support being interruptible and allowing for honoring to SK_SNDTIMEO (Willem de Bruijn) V2->V3: Rearrage the point at which we wait for the completion queue, to avoid needing to check for ph/skb being null at the end of the loop. Also move the complete call to the skb destructor to avoid needing to modify __packet_set_status. Also gate calling complete on packet_read_pending returning zero to avoid multiple calls to complete. (Willem de Bruijn) Move timeo computation within loop, to re-fetch the socket timeout since we also use the timeo variable to record the return code from the wait_for_complete call (Neil Horman) V3->V4: Willem has requested that the control flow be restored to the previous state. Doing so lets us eliminate the need for the po->wait_on_complete flag variable, and lets us get rid of the packet_next_frame function, but introduces another complexity. Specifically, but using the packet pending count, we can, if an applications calls sendmsg multiple times with MSG_DONTWAIT set, each set of transmitted frames, when complete, will cause tpacket_destruct_skb to issue a complete call, for which there will never be a wait_on_completion call. This imbalance will lead to any future call to wait_for_completion here to return early, when the frames they sent may not have completed. To correct this, we need to re-init the completion queue on every call to tpacket_snd before we enter the loop so as to ensure we wait properly for the frames we send in this iteration. Change the timeout and interrupted gotos to out_put rather than out_status so that we don't try to free a non-existant skb Clean up some extra newlines (Willem de Bruijn) Reviewed-by:
Willem de Bruijn <willemb@google.com> Signed-off-by:
Neil Horman <nhorman@tuxdriver.com> Reported-by:
Matteo Croce <mcroce@redhat.com> Signed-off-by:
-
Paul Burton authored
commit 6d4d367d upstream. The MIPS GIC contains a block of registers used to map local interrupts to a particular CPU interrupt pin. Since these registers are found at a consecutive range of addresses we access them using an index, via the (read|write)_gic_v[lo]_map accessor functions. We currently use values from enum mips_gic_local_interrupt as those indices. Unfortunately whilst enum mips_gic_local_interrupt provides the correct offsets for bits in the pending & mask registers, the ordering of the map registers is subtly different... Compared with the ordering of pending & mask bits, the map registers move the FDC from the end of the list to index 3 after the timer interrupt. As a result the performance counter & software interrupts are therefore at indices 4-6 rather than indices 3-5. Notably this causes problems with performance counter interrupts being incorrectly mapped on some systems, and presumably will also cause problems for FDC interrupts. Introduce a function to map from enum mips_gic_local_interrupt to the index of the corresponding map register, and use it to ensure we access the map registers for the correct interrupts. Signed-off-by:
Paul Burton <paul.burton@mips.com> Fixes: a0dc5cb5 ("irqchip: mips-gic: Simplify gic_local_irq_domain_map()") Fixes: da61fcf9 ("irqchip: mips-gic: Use irq_cpu_online to (un)mask all-VP(E) IRQs") Reported-and-tested-by:
Archer Yan <ayan@wavecomp.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Jason Cooper <jason@lakedaemon.net> Cc: stable@vger.kernel.org # v4.14+ Signed-off-by:
Marc Zyngier <marc.zyngier@arm.com> Signed-off-by:
-
Trond Myklebust authored
commit 7e3d3620 upstream. In the case where a record marker was used, xs_sendpages() needs to return the length of the payload + record marker so that we operate correctly in the case of a partial transmission. When the callers check return value, they therefore need to take into account the record marker length. Fixes: 06b5fc3a ("Merge tag 'nfs-rdma-for-5.1-1'...") Cc: stable@vger.kernel.org # 5.1+ Signed-off-by:
Trond Myklebust <trond.myklebust@hammerspace.com> Signed-off-by:
Anna Schumaker <Anna.Schumaker@Netapp.com> Signed-off-by:
-
Geert Uytterhoeven authored
commit 1bf72720 upstream. Currently, if the user specifies an unsupported mitigation strategy on the kernel command line, it will be ignored silently. The code will fall back to the default strategy, possibly leaving the system more vulnerable than expected. This may happen due to e.g. a simple typo, or, for a stable kernel release, because not all mitigation strategies have been backported. Inform the user by printing a message. Fixes: 98af8452 ("cpu/speculation: Add 'mitigations=' cmdline option") Signed-off-by:
Geert Uytterhoeven <geert@linux-m68k.org> Signed-off-by:
Thomas Gleixner <tglx@linutronix.de> Acked-by:
Josh Poimboeuf <jpoimboe@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Jiri Kosina <jkosina@suse.cz> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: Ben Hutchings <ben@decadent.org.uk> Cc: stable@vger.kernel.org Link: https://lkml.kernel.org/r/20190516070935.22546-1-geert@linux-m68k.orgSigned-off-by:
-
Trond Myklebust authored
commit 68f46159 upstream. Fix a typo where we're confusing the default TCP retrans value (NFS_DEF_TCP_RETRANS) for the default TCP timeout value. Fixes: 15d03055 ("pNFS/flexfiles: Set reasonable default ...") Cc: stable@vger.kernel.org # 4.8+ Signed-off-by:
-
Ard Biesheuvel authored
commit 18df7577 upstream. Ensure that the EFI memreserve entries can be accessed, even if they are located in memory that the kernel (e.g., a crashkernel) omits from the linear map. Fixes: 80424b02 ("efi: Reduce the amount of memblock reservations ...") Cc: <stable@vger.kernel.org> # 5.0+ Reported-by:
Jonathan Richardson <jonathan.richardson@broadcom.com> Reviewed-by:
Ard Biesheuvel <ard.biesheuvel@linaro.org> Signed-off-by:
-
Johannes Weiner authored
commit 7b785645 upstream. Since a2833486 ("page cache: Finish XArray conversion"), on most major Linux distributions, the page cache doesn't correctly transition when the hot data set is changing, and leaves the new pages thrashing indefinitely instead of kicking out the cold ones. On a freshly booted, freshly ssh'd into virtual machine with 1G RAM running stock Arch Linux: [root@ham ~]# ./reclaimtest.sh + dd of=workingset-a bs=1M count=0 seek=600 + cat workingset-a + cat workingset-a + cat workingset-a + cat workingset-a + cat workingset-a + cat workingset-a + cat workingset-a + cat workingset-a + ./mincore workingset-a 153600/153600 workingset-a + dd of=workingset-b bs=1M count=0 seek=600 + cat workingset-b + cat workingset-b + cat workingset-b + cat workingset-b + ./mincore workingset-a workingset-b 104029/153600 workingset-a 120086/153600 workingset-b + cat workingset-b + cat workingset-b + cat workingset-b + cat workingset-b + ./mincore workingset-a workingset-b 104029/153600 workingset-a 120268/153600 workingset-b workingset-b is a 600M file on a 1G host that is otherwise entirely idle. No matter how often it's being accessed, it won't get cached. While investigating, I noticed that the non-resident information gets aggressively reclaimed - /proc/vmstat::workingset_nodereclaim. This is a problem because a workingset transition like this relies on the non-resident information tracked in the page cache tree of evicted file ranges: when the cache faults are refaults of recently evicted cache, we challenge the existing active set, and that allows a new workingset to establish itself. Tracing the shrinker that maintains this memory revealed that all page cache tree nodes were allocated to the root cgroup. This is a problem, because 1) the shrinker sizes the amount of non-resident information it keeps to the size of the cgroup's other memory and 2) on most major Linux distributions, only kernel threads live in the root cgroup and everything else gets put into services or session groups: [root@ham ~]# cat /proc/self/cgroup 0::/user.slice/user-0.slice/session-c1.scope As a result, we basically maintain no non-resident information for the workloads running on the system, thus breaking the caching algorithm. Looking through the code, I found the culprit in the above-mentioned patch: when switching from the radix tree to xarray, it dropped the __GFP_ACCOUNT flag from the tree node allocations - the flag that makes sure the allocated memory gets charged to and tracked by the cgroup of the calling process - in this case, the one doing the fault. To fix this, allow xarray users to specify per-tree flag that makes xarray allocate nodes using __GFP_ACCOUNT. Then restore the page cache tree annotation to request such cgroup tracking for the cache nodes. With this patch applied, the page cache correctly converges on new workingsets again after just a few iterations: [root@ham ~]# ./reclaimtest.sh + dd of=workingset-a bs=1M count=0 seek=600 + cat workingset-a + cat workingset-a + cat workingset-a + cat workingset-a + cat workingset-a + cat workingset-a + cat workingset-a + cat workingset-a + ./mincore workingset-a 153600/153600 workingset-a + dd of=workingset-b bs=1M count=0 seek=600 + cat workingset-b + ./mincore workingset-a workingset-b 124607/153600 workingset-a 87876/153600 workingset-b + cat workingset-b + ./mincore workingset-a workingset-b 81313/153600 workingset-a 133321/153600 workingset-b + cat workingset-b + ./mincore workingset-a workingset-b 63036/153600 workingset-a 153600/153600 workingset-b Cc: stable@vger.kernel.org # 4.20+ Signed-off-by:
Johannes Weiner <hannes@cmpxchg.org> Reviewed-by:
Shakeel Butt <shakeelb@google.com> Signed-off-by:
Matthew Wilcox (Oracle) <willy@infradead.org> Signed-off-by:
-
Reinette Chatre authored
commit 32f010de upstream. While the DOC at the beginning of lib/bitmap.c explicitly states that "The number of valid bits in a given bitmap does _not_ need to be an exact multiple of BITS_PER_LONG.", some of the bitmap operations do indeed access BITS_PER_LONG portions of the provided bitmap no matter the size of the provided bitmap. For example, if find_first_bit() is provided with an 8 bit bitmap the operation will access BITS_PER_LONG bits from the provided bitmap. While the operation ensures that these extra bits do not affect the result, the memory is still accessed. The capacity bitmasks (CBMs) are typically stored in u32 since they can never exceed 32 bits. A few instances exist where a bitmap_* operation is performed on a CBM by simply pointing the bitmap operation to the stored u32 value. The consequence of this pattern is that some bitmap_* operations will access out-of-bounds memory when interacting with the provided CBM. This same issue has previously been addressed with commit 49e00eee ("x86/intel_rdt: Fix out-of-bounds memory access in CBM tests") but at that time not all instances of the issue were fixed. Fix this by using an unsigned long to store the capacity bitmask data that is passed to bitmap functions. Fixes: e6519011 ("x86/intel_rdt: Introduce "bit_usage" to display cache allocations details") Fixes: f4e80d67 ("x86/intel_rdt: Resctrl files reflect pseudo-locked information") Fixes: 95f0b77e ("x86/intel_rdt: Initialize new resource group with sane defaults") Signed-off-by:
Reinette Chatre <reinette.chatre@intel.com> Signed-off-by:
Borislav Petkov <bp@suse.de> Cc: Fenghua Yu <fenghua.yu@intel.com> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: stable <stable@vger.kernel.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Tony Luck <tony.luck@intel.com> Cc: x86-ml <x86@kernel.org> Link: https://lkml.kernel.org/r/58c9b6081fd9bf599af0dfc01a6fdd335768efef.1560975645.git.reinette.chatre@intel.comSigned-off-by:
-
Thomas Gleixner authored
commit 5423f5ce upstream. A recent change moved the microcode loader hotplug callback into the early startup phase which is running with interrupts disabled. It missed that the callbacks invoke sysfs functions which might sleep causing nice 'might sleep' splats with proper debugging enabled. Split the callbacks and only load the microcode in the early startup phase and move the sysfs handling back into the later threaded and preemptible bringup phase where it was before. Fixes: 78f4e932 ("x86/microcode, cpuhotplug: Add a microcode loader CPU hotplug callback") Signed-off-by:
-
Alejandro Jimenez authored
commit c1f7fec1 upstream. The bits set in x86_spec_ctrl_mask are used to calculate the guest's value of SPEC_CTRL that is written to the MSR before VMENTRY, and control which mitigations the guest can enable. In the case of SSBD, unless the host has enabled SSBD always on mode (by passing "spec_store_bypass_disable=on" in the kernel parameters), the SSBD bit is not set in the mask and the guest can not properly enable the SSBD always on mitigation mode. This has been confirmed by running the SSBD PoC on a guest using the SSBD always on mitigation mode (booted with kernel parameter "spec_store_bypass_disable=on"), and verifying that the guest is vulnerable unless the host is also using SSBD always on mode. In addition, the guest OS incorrectly reports the SSB vulnerability as mitigated. Always set the SSBD bit in x86_spec_ctrl_mask when the host CPU supports it, allowing the guest to use SSBD whether or not the host has chosen to enable the mitigation in any of its modes. Fixes: be6fcb54 ("x86/bugs: Rework spec_ctrl base and mask logic") Signed-off-by:
Alejandro Jimenez <alejandro.j.jimenez@oracle.com> Signed-off-by:
Liam Merwick <liam.merwick@oracle.com> Reviewed-by:
Mark Kanda <mark.kanda@oracle.com> Reviewed-by:
Paolo Bonzini <pbonzini@redhat.com> Cc: bp@alien8.de Cc: rkrcmar@redhat.com Cc: kvm@vger.kernel.org Cc: stable@vger.kernel.org Link: https://lkml.kernel.org/r/1560187210-11054-1-git-send-email-alejandro.j.jimenez@oracle.comSigned-off-by:
-
Jan Kara authored
commit 240b4cc8 upstream. Once we unlock adapter->hw_lock in pvscsi_queue_lck() nothing prevents just queued scsi_cmnd from completing and freeing the request. Thus cmd->cmnd[0] dereference can dereference already freed request leading to kernel crashes or other issues (which one of our customers observed). Store cmd->cmnd[0] in a local variable before unlocking adapter->hw_lock to fix the issue. CC: <stable@vger.kernel.org> Signed-off-by:
Jan Kara <jack@suse.cz> Reviewed-by:
Ewan D. Milne <emilne@redhat.com> Signed-off-by:
Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by:
-
Jens Axboe authored
commit 60c112b0 upstream. Stephen reports: I hit the following General Protection Fault when testing io_uring via the io_uring engine in fio. This was on a VM running 5.2-rc5 and the latest version of fio. The issue occurs for both null_blk and fake NVMe drives. I have not tested bare metal or real NVMe SSDs. The fio script used is given below. [io_uring] time_based=1 runtime=60 filename=/dev/nvme2n1 (note /dev/nullb0 also fails) ioengine=io_uring bs=4k rw=readwrite direct=1 fixedbufs=1 sqthread_poll=1 sqthread_poll_cpu=0 general protection fault: 0000 [#1] SMP PTI CPU: 0 PID: 872 Comm: io_uring-sq Not tainted 5.2.0-rc5-cpacket-io-uring #1 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Ubuntu-1.8.2-1ubuntu1 04/01/2014 RIP: 0010:fput_many+0x7/0x90 Code: 01 48 85 ff 74 17 55 48 89 e5 53 48 8b 1f e8 a0 f9 ff ff 48 85 db 48 89 df 75 f0 5b 5d f3 c3 0f 1f 40 00 0f 1f 44 00 00 89 f6 <f0> 48 29 77 38 74 01 c3 55 48 89 e5 53 48 89 fb 65 48 \ RSP: 0018:ffffadeb817ebc50 EFLAGS: 00010246 RAX: 0000000000000004 RBX: ffff8f46ad477480 RCX: 0000000000001805 RDX: 0000000000000000 RSI: 0000000000000001 RDI: f18b51b9a39552b5 RBP: ffffadeb817ebc58 R08: ffff8f46b7a318c0 R09: 000000000000015d R10: ffffadeb817ebce8 R11: 0000000000000020 R12: ffff8f46ad4cd000 R13: 00000000fffffff7 R14: ffffadeb817ebe30 R15: 0000000000000004 FS: 0000000000000000(0000) GS:ffff8f46b7a00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055828f0bbbf0 CR3: 0000000232176004 CR4: 00000000003606f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: ? fput+0x13/0x20 io_free_req+0x20/0x40 io_put_req+0x1b/0x20 io_submit_sqe+0x40a/0x680 ? __switch_to_asm+0x34/0x70 ? __switch_to_asm+0x40/0x70 io_submit_sqes+0xb9/0x160 ? io_submit_sqes+0xb9/0x160 ? __switch_to_asm+0x40/0x70 ? __switch_to_asm+0x34/0x70 ? __schedule+0x3f2/0x6a0 ? __switch_to_asm+0x34/0x70 io_sq_thread+0x1af/0x470 ? __switch_to_asm+0x34/0x70 ? wait_woken+0x80/0x80 ? __switch_to+0x85/0x410 ? __switch_to_asm+0x40/0x70 ? __switch_to_asm+0x34/0x70 ? __schedule+0x3f2/0x6a0 kthread+0x105/0x140 ? io_submit_sqes+0x160/0x160 ? kthread+0x105/0x140 ? io_submit_sqes+0x160/0x160 ? kthread_destroy_worker+0x50/0x50 ret_from_fork+0x35/0x40 which occurs because using a kernel side submission thread isn't valid without using fixed files (registered through io_uring_register()). This causes io_uring to put the request after logging an error, but before the file field is set in the request. If it happens to be non-zero, we attempt to fput() garbage. Fix this by ensuring that req->file is initialized when the request is allocated. Cc: stable@vger.kernel.org # 5.1+ Reported-by:
Stephen Bates <sbates@raithlin.com> Tested-by:
Jens Axboe <axboe@kernel.dk> Signed-off-by:
-
zhangyi (F) authored
commit 211ad4b7 upstream. Currently, although we submit super bios in order (and super.nr_entries is incremented by each logged entry), submit_bio() is async so each super sector may not be written to log device in order and then the final nr_entries may be smaller than it should be. This problem can be reproduced by the xfstests generic/455 with ext4: QA output created by 455 -Silence is golden +mark 'end' does not exist Fix this by serializing submission of super sectors to make sure each is written to the log disk in order. Fixes: 0e9cebe7 ("dm: add log writes target") Cc: stable@vger.kernel.org Signed-off-by:
zhangyi (F) <yi.zhang@huawei.com> Suggested-by:
Josef Bacik <josef@toxicpanda.com> Reviewed-by:
Mike Snitzer <snitzer@redhat.com> Signed-off-by:
-
Gen Zhang authored
commit dec7e649 upstream. Fix 2 kstrndup() calls with incorrect argument order. Fixes: 6bbc923d ("dm: add support to directly boot to a mapped device") Cc: stable@vger.kernel.org # v5.1 Signed-off-by:
Gen Zhang <blackgod016574@gmail.com> Signed-off-by:
-
Huang Ying authored
commit 1a5f439c upstream. 0-Day test system reported some OOM regressions for several THP (Transparent Huge Page) swap test cases. These regressions are bisected to 68614289 ("block: always define BIO_MAX_PAGES as 256"). In the commit, BIO_MAX_PAGES is set to 256 even when THP swap is enabled. So the bio_alloc(gfp_flags, 512) in get_swap_bio() may fail when swapping out THP. That causes the OOM. As in the patch description of 68614289 ("block: always define BIO_MAX_PAGES as 256"), THP swap should use multi-page bvec to write THP to swap space. So the issue is fixed via doing that in get_swap_bio(). BTW: I remember I have checked the THP swap code when 68614289 ("block: always define BIO_MAX_PAGES as 256") was merged, and thought the THP swap code needn't to be changed. But apparently, I was wrong. I should have done this at that time. Link: http://lkml.kernel.org/r/20190624075515.31040-1-ying.huang@intel.com Fixes: 68614289 ("block: always define BIO_MAX_PAGES as 256") Signed-off-by:
"Huang, Ying" <ying.huang@intel.com> Reviewed-by:
Ming Lei <ming.lei@redhat.com> Cc: Michal Hocko <mhocko@kernel.org> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Hugh Dickins <hughd@google.com> Cc: Minchan Kim <minchan@kernel.org> Cc: Rik van Riel <riel@redhat.com> Cc: Daniel Jordan <daniel.m.jordan@oracle.com> Cc: Jens Axboe <axboe@kernel.dk> Cc: <stable@vger.kernel.org> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by:
-
Colin Ian King authored
commit 7298e3b0 upstream. Currently the calcuation of end_pfn can round up the pfn number to more than the actual maximum number of pfns, causing an Oops. Fix this by ensuring end_pfn is never more than max_pfn. This can be easily triggered when on systems where the end_pfn gets rounded up to more than max_pfn using the idle-page stress-ng stress test: sudo stress-ng --idle-page 0 BUG: unable to handle kernel paging request at 00000000000020d8 #PF error: [normal kernel read fault] PGD 0 P4D 0 Oops: 0000 [#1] SMP PTI CPU: 1 PID: 11039 Comm: stress-ng-idle- Not tainted 5.0.0-5-generic #6-Ubuntu Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.10.2-1ubuntu1 04/01/2014 RIP: 0010:page_idle_get_page+0xc8/0x1a0 Code: 0f b1 0a 75 7d 48 8b 03 48 89 c2 48 c1 e8 33 83 e0 07 48 c1 ea 36 48 8d 0c 40 4c 8d 24 88 49 c1 e4 07 4c 03 24 d5 00 89 c3 be <49> 8b 44 24 58 48 8d b8 80 a1 02 00 e8 07 d5 77 00 48 8b 53 08 48 RSP: 0018:ffffafd7c672fde8 EFLAGS: 00010202 RAX: 0000000000000005 RBX: ffffe36341fff700 RCX: 000000000000000f RDX: 0000000000000284 RSI: 0000000000000275 RDI: 0000000001fff700 RBP: ffffafd7c672fe00 R08: ffffa0bc34056410 R09: 0000000000000276 R10: ffffa0bc754e9b40 R11: ffffa0bc330f6400 R12: 0000000000002080 R13: ffffe36341fff700 R14: 0000000000080000 R15: ffffa0bc330f6400 FS: 00007f0ec1ea5740(0000) GS:ffffa0bc7db00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00000000000020d8 CR3: 0000000077d68000 CR4: 00000000000006e0 Call Trace: page_idle_bitmap_write+0x8c/0x140 sysfs_kf_bin_write+0x5c/0x70 kernfs_fop_write+0x12e/0x1b0 __vfs_write+0x1b/0x40 vfs_write+0xab/0x1b0 ksys_write+0x55/0xc0 __x64_sys_write+0x1a/0x20 do_syscall_64+0x5a/0x110 entry_SYSCALL_64_after_hwframe+0x44/0xa9 Link: http://lkml.kernel.org/r/20190618124352.28307-1-colin.king@canonical.com Fixes: 33c3fc71 ("mm: introduce idle page tracking") Signed-off-by:
Colin Ian King <colin.king@canonical.com> Reviewed-by:
Vladimir Davydov <vdavydov.dev@gmail.com> Cc: Michal Hocko <mhocko@suse.com> Cc: Mike Rapoport <rppt@linux.vnet.ibm.com> Cc: Mel Gorman <mgorman@techsingularity.net> Cc: Stephen Rothwell <sfr@canb.auug.org.au> Cc: Andrey Ryabinin <aryabinin@virtuozzo.com> Cc: <stable@vger.kernel.org> Signed-off-by:
-
Naoya Horiguchi authored
commit faf53def upstream. madvise(MADV_SOFT_OFFLINE) often returns -EBUSY when calling soft offline for hugepages with overcommitting enabled. That was caused by the suboptimal code in current soft-offline code. See the following part: ret = migrate_pages(&pagelist, new_page, NULL, MPOL_MF_MOVE_ALL, MIGRATE_SYNC, MR_MEMORY_FAILURE); if (ret) { ... } else { /* * We set PG_hwpoison only when the migration source hugepage * was successfully dissolved, because otherwise hwpoisoned * hugepage remains on free hugepage list, then userspace will * find it as SIGBUS by allocation failure. That's not expected * in soft-offlining. */ ret = dissolve_free_huge_page(page); if (!ret) { if (set_hwpoison_free_buddy_page(page)) num_poisoned_pages_inc(); } } return ret; Here dissolve_free_huge_page() returns -EBUSY if the migration source page was freed into buddy in migrate_pages(), but even in that case we actually has a chance that set_hwpoison_free_buddy_page() succeeds. So that means current code gives up offlining too early now. dissolve_free_huge_page() checks that a given hugepage is suitable for dissolving, where we should return success for !PageHuge() case because the given hugepage is considered as already dissolved. This change also affects other callers of dissolve_free_huge_page(), which are cleaned up together. [n-horiguchi@ah.jp.nec.com: v3] Link: http://lkml.kernel.org/r/1560761476-4651-3-git-send-email-n-horiguchi@ah.jp.nec.comLink: http://lkml.kernel.org/r/1560154686-18497-3-git-send-email-n-horiguchi@ah.jp.nec.com Fixes: 6bc9b564 ("mm: fix race on soft-offlining") Signed-off-by:
Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Reported-by:
Chen, Jerry T <jerry.t.chen@intel.com> Tested-by:
Mike Kravetz <mike.kravetz@oracle.com> Reviewed-by:
Oscar Salvador <osalvador@suse.de> Cc: Michal Hocko <mhocko@kernel.org> Cc: Xishi Qiu <xishi.qiuxishi@alibaba-inc.com> Cc: "Chen, Jerry T" <jerry.t.chen@intel.com> Cc: "Zhuo, Qiuxu" <qiuxu.zhuo@intel.com> Cc: <stable@vger.kernel.org> [4.19+] Signed-off-by:
-
