October 2020 - Kernel - mailweb.openeuler.org

[PATCH] perf/core: Fix race in the perf_mmap_close() function
by Yang Yingliang 16 Oct '20

16 Oct '20

From: Jiri Olsa <jolsa(a)redhat.com> mainline inclusion from mainline-v5.10 commit f91072ed1b7283b13ca57fcfbece5a3b92726143 category: bugfix bugzilla: NA CVE: CVE-2020-14351 -------------------------------- There's a possible race in perf_mmap_close() when checking ring buffer's mmap_count refcount value. The problem is that the mmap_count check is not atomic because we call atomic_dec() and atomic_read() separately. perf_mmap_close: ... atomic_dec(&rb->mmap_count); ... if (atomic_read(&rb->mmap_count)) goto out_put; <ring buffer detach> free_uid out_put: ring_buffer_put(rb); /* could be last */ The race can happen when we have two (or more) events sharing same ring buffer and they go through atomic_dec() and then they both see 0 as refcount value later in atomic_read(). Then both will go on and execute code which is meant to be run just once. The code that detaches ring buffer is probably fine to be executed more than once, but the problem is in calling free_uid(), which will later on demonstrate in related crashes and refcount warnings, like: refcount_t: addition on 0; use-after-free. ... RIP: 0010:refcount_warn_saturate+0x6d/0xf ... Call Trace: prepare_creds+0x190/0x1e0 copy_creds+0x35/0x172 copy_process+0x471/0x1a80 _do_fork+0x83/0x3a0 __do_sys_wait4+0x83/0x90 __do_sys_clone+0x85/0xa0 do_syscall_64+0x5b/0x1e0 entry_SYSCALL_64_after_hwframe+0x44/0xa9 Using atomic decrease and check instead of separated calls. Tested-by: Michael Petlan <mpetlan(a)redhat.com> Signed-off-by: Jiri Olsa <jolsa(a)kernel.org> Signed-off-by: Ingo Molnar <mingo(a)kernel.org> Acked-by: Peter Zijlstra <a.p.zijlstra(a)chello.nl> Acked-by: Namhyung Kim <namhyung(a)kernel.org> Acked-by: Wade Mealing <wmealing(a)redhat.com> Fixes: 9bb5d40cd93c ("perf: Fix mmap() accounting hole"); Link: https://lore.kernel.org/r/20200916115311.GE2301783@krava Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> Conflicts: kernel/events/core.c [yyl: adjust context] Reviewed-by: Jian Cheng <cj.chengjian(a)huawei.com> Reviewed-by: Jason Yan <yanaijie(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- kernel/events/core.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/kernel/events/core.c b/kernel/events/core.c index 30a81d1b7b5c..452775f68129 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -5496,11 +5496,11 @@ static void perf_pmu_output_stop(struct perf_event *event); static void perf_mmap_close(struct vm_area_struct *vma) { struct perf_event *event = vma->vm_file->private_data; - struct ring_buffer *rb = ring_buffer_get(event); struct user_struct *mmap_user = rb->mmap_user; int mmap_locked = rb->mmap_locked; unsigned long size = perf_data_size(rb); + bool detach_rest = false; if (event->pmu->event_unmapped) event->pmu->event_unmapped(event, vma->vm_mm); @@ -5531,7 +5531,8 @@ static void perf_mmap_close(struct vm_area_struct *vma) mutex_unlock(&event->mmap_mutex); } - atomic_dec(&rb->mmap_count); + if (atomic_dec_and_test(&rb->mmap_count)) + detach_rest = true; if (!atomic_dec_and_mutex_lock(&event->mmap_count, &event->mmap_mutex)) goto out_put; @@ -5540,7 +5541,7 @@ static void perf_mmap_close(struct vm_area_struct *vma) mutex_unlock(&event->mmap_mutex); /* If there's still other mmap()s of this buffer, we're done. */ - if (atomic_read(&rb->mmap_count)) + if (!detach_rest) goto out_put; /* -- 2.25.1

1 0

[PATCH] geneve: add transport ports in route lookup for geneve
by Yang Yingliang 15 Oct '20

15 Oct '20

From: Mark Gray <mark.d.gray(a)redhat.com> stable inclusion from linux-4.19.148 commit c797110d97c48054d1491251fd713900ff51615c CVE: CVE-2020-25645 -------------------------------- [ Upstream commit 34beb21594519ce64a55a498c2fe7d567bc1ca20 ] This patch adds transport ports information for route lookup so that IPsec can select Geneve tunnel traffic to do encryption. This is needed for OVS/OVN IPsec with encrypted Geneve tunnels. This can be tested by configuring a host-host VPN using an IKE daemon and specifying port numbers. For example, for an Openswan-type configuration, the following parameters should be configured on both hosts and IPsec set up as-per normal: $ cat /etc/ipsec.conf conn in ... left=$IP1 right=$IP2 ... leftprotoport=udp/6081 rightprotoport=udp ... conn out ... left=$IP1 right=$IP2 ... leftprotoport=udp rightprotoport=udp/6081 ... The tunnel can then be setup using "ip" on both hosts (but changing the relevant IP addresses): $ ip link add tun type geneve id 1000 remote $IP2 $ ip addr add 192.168.0.1/24 dev tun $ ip link set tun up This can then be tested by pinging from $IP1: $ ping 192.168.0.2 Without this patch the traffic is unencrypted on the wire. Fixes: 2d07dc79fe04 ("geneve: add initial netdev driver for GENEVE tunnels") Signed-off-by: Qiuyu Xiao <qiuyu.xiao.qyx(a)gmail.com> Signed-off-by: Mark Gray <mark.d.gray(a)redhat.com> Reviewed-by: Greg Rose <gvrose8192(a)gmail.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> Reviewed-by: Jason Yan <yanaijie(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- drivers/net/geneve.c | 37 +++++++++++++++++++++++++++---------- 1 file changed, 27 insertions(+), 10 deletions(-) diff --git a/drivers/net/geneve.c b/drivers/net/geneve.c index 1506e913cebb..fdb251e1f471 100644 --- a/drivers/net/geneve.c +++ b/drivers/net/geneve.c @@ -721,7 +721,8 @@ static struct rtable *geneve_get_v4_rt(struct sk_buff *skb, struct net_device *dev, struct geneve_sock *gs4, struct flowi4 *fl4, - const struct ip_tunnel_info *info) + const struct ip_tunnel_info *info, + __be16 dport, __be16 sport) { bool use_cache = ip_tunnel_dst_cache_usable(skb, info); struct geneve_dev *geneve = netdev_priv(dev); @@ -737,6 +738,8 @@ static struct rtable *geneve_get_v4_rt(struct sk_buff *skb, fl4->flowi4_proto = IPPROTO_UDP; fl4->daddr = info->key.u.ipv4.dst; fl4->saddr = info->key.u.ipv4.src; + fl4->fl4_dport = dport; + fl4->fl4_sport = sport; tos = info->key.tos; if ((tos == 1) && !geneve->collect_md) { @@ -771,7 +774,8 @@ static struct dst_entry *geneve_get_v6_dst(struct sk_buff *skb, struct net_device *dev, struct geneve_sock *gs6, struct flowi6 *fl6, - const struct ip_tunnel_info *info) + const struct ip_tunnel_info *info, + __be16 dport, __be16 sport) { bool use_cache = ip_tunnel_dst_cache_usable(skb, info); struct geneve_dev *geneve = netdev_priv(dev); @@ -787,6 +791,9 @@ static struct dst_entry *geneve_get_v6_dst(struct sk_buff *skb, fl6->flowi6_proto = IPPROTO_UDP; fl6->daddr = info->key.u.ipv6.dst; fl6->saddr = info->key.u.ipv6.src; + fl6->fl6_dport = dport; + fl6->fl6_sport = sport; + prio = info->key.tos; if ((prio == 1) && !geneve->collect_md) { prio = ip_tunnel_get_dsfield(ip_hdr(skb), skb); @@ -833,14 +840,15 @@ static int geneve_xmit_skb(struct sk_buff *skb, struct net_device *dev, __be16 df; int err; - rt = geneve_get_v4_rt(skb, dev, gs4, &fl4, info); + sport = udp_flow_src_port(geneve->net, skb, 1, USHRT_MAX, true); + rt = geneve_get_v4_rt(skb, dev, gs4, &fl4, info, + geneve->info.key.tp_dst, sport); if (IS_ERR(rt)) return PTR_ERR(rt); skb_tunnel_check_pmtu(skb, &rt->dst, GENEVE_IPV4_HLEN + info->options_len); - sport = udp_flow_src_port(geneve->net, skb, 1, USHRT_MAX, true); if (geneve->collect_md) { tos = ip_tunnel_ecn_encap(key->tos, ip_hdr(skb), skb); ttl = key->ttl; @@ -875,13 +883,14 @@ static int geneve6_xmit_skb(struct sk_buff *skb, struct net_device *dev, __be16 sport; int err; - dst = geneve_get_v6_dst(skb, dev, gs6, &fl6, info); + sport = udp_flow_src_port(geneve->net, skb, 1, USHRT_MAX, true); + dst = geneve_get_v6_dst(skb, dev, gs6, &fl6, info, + geneve->info.key.tp_dst, sport); if (IS_ERR(dst)) return PTR_ERR(dst); skb_tunnel_check_pmtu(skb, dst, GENEVE_IPV6_HLEN + info->options_len); - sport = udp_flow_src_port(geneve->net, skb, 1, USHRT_MAX, true); if (geneve->collect_md) { prio = ip_tunnel_ecn_encap(key->tos, ip_hdr(skb), skb); ttl = key->ttl; @@ -957,13 +966,18 @@ static int geneve_fill_metadata_dst(struct net_device *dev, struct sk_buff *skb) { struct ip_tunnel_info *info = skb_tunnel_info(skb); struct geneve_dev *geneve = netdev_priv(dev); + __be16 sport; if (ip_tunnel_info_af(info) == AF_INET) { struct rtable *rt; struct flowi4 fl4; + struct geneve_sock *gs4 = rcu_dereference(geneve->sock4); + sport = udp_flow_src_port(geneve->net, skb, + 1, USHRT_MAX, true); - rt = geneve_get_v4_rt(skb, dev, gs4, &fl4, info); + rt = geneve_get_v4_rt(skb, dev, gs4, &fl4, info, + geneve->info.key.tp_dst, sport); if (IS_ERR(rt)) return PTR_ERR(rt); @@ -973,9 +987,13 @@ static int geneve_fill_metadata_dst(struct net_device *dev, struct sk_buff *skb) } else if (ip_tunnel_info_af(info) == AF_INET6) { struct dst_entry *dst; struct flowi6 fl6; + struct geneve_sock *gs6 = rcu_dereference(geneve->sock6); + sport = udp_flow_src_port(geneve->net, skb, + 1, USHRT_MAX, true); - dst = geneve_get_v6_dst(skb, dev, gs6, &fl6, info); + dst = geneve_get_v6_dst(skb, dev, gs6, &fl6, info, + geneve->info.key.tp_dst, sport); if (IS_ERR(dst)) return PTR_ERR(dst); @@ -986,8 +1004,7 @@ static int geneve_fill_metadata_dst(struct net_device *dev, struct sk_buff *skb) return -EINVAL; } - info->key.tp_src = udp_flow_src_port(geneve->net, skb, - 1, USHRT_MAX, true); + info->key.tp_src = sport; info->key.tp_dst = geneve->info.key.tp_dst; return 0; } -- 2.25.1

1 0

[PATCH 00/78] bugfixes for openEuler-1.0-LTS
by Yang Yingliang 15 Oct '20

15 Oct '20

Al Grant (1): perf tools: Correct SNOOPX field offset Al Viro (2): do_epoll_ctl(): clean the failure exits up a bit fix regression in "epoll: Keep a reference on files added to the check list" Alex Williamson (1): vfio/type1: Add proper error unwind for vfio_iommu_replay() Alvin Šipraga (1): macvlan: validate setting of multiple remote source MAC addresses Bodo Stroesser (1): scsi: target: tcmu: Fix crash in tcmu_flush_dcache_range on ARM Brian Foster (3): xfs: acquire superblock freeze protection on eofblocks scans xfs: reset buffer write failure state on successful completion xfs: fix duplicate verification from xfs_qm_dqflush() Charan Teja Reddy (1): mm, page_alloc: fix core hung in free_pcppages_bulk() Chris Wilson (1): locking/lockdep: Fix overflow in presentation of average lock-time Chuck Lever (1): NFS: Zero-stateid SETATTR should first return delegation Daniel Borkmann (1): uaccess: Add non-pagefault user-space write function Darrick J. Wong (4): xfs: fix partially uninitialized structure in xfs_reflink_remap_extent xfs: fix inode quota reservation checks xfs: fix xfs_bmap_validate_extent_raw when checking attr fork of rt files xfs: initialize the shortform attr header padding entry Dave Chinner (1): xfs: Don't allow logging of XFS_ISTALE inodes David Milburn (2): nvme-fc: cancel async events before freeing event struct nvme-rdma: cancel async events before freeing event struct Doug Berger (1): mm: include CMA pages in lowmem_reserve at boot Eiichi Tsukata (1): xfs: Fix UBSAN null-ptr-deref in xfs_sysfs_init Eric Biggers (1): xfs: clear PF_MEMALLOC before exiting xfsaild thread George Kennedy (1): vt_ioctl: change VT_RESIZEX ioctl to check for error return from vc_resize() Gustav Wiklander (1): spi: Fix memory leak on splited transfers Heikki Krogerus (1): device property: Fix the secondary firmware node handling in set_primary_fwnode() Helge Deller (1): fs/signalfd.c: fix inconsistent return codes for signalfd4 Hou Pu (1): scsi: target: iscsi: Fix hang in iscsit_access_np() when getting tpg->np_login_sem Hugh Dickins (2): khugepaged: khugepaged_test_exit() check mmget_still_valid() khugepaged: adjust VM_BUG_ON_MM() in __khugepaged_enter() Jan Kara (5): ext4: fix checking of directory entry validity for inline directories ext4: don't BUG on inconsistent journal feature writeback: Protect inode->i_io_list with inode->i_lock writeback: Avoid skipping inode writeback writeback: Fix sync livelock due to b_dirty_time processing Jarkko Sakkinen (1): tpm: Unify the mismatching TPM space buffer sizes Jason Gunthorpe (1): include/linux/log2.h: add missing () around n in roundup_pow_of_two() Javed Hasan (1): scsi: fcoe: Memory leak fix in fcoe_sysfs_fcf_del() Jens Axboe (1): block: ensure bdi->io_pages is always initialized Jing Xiangfeng (1): scsi: iscsi: Do not put host in iscsi_set_flashnode_param() Li Heng (1): efi: add missed destroy_workqueue when efisubsys_init fails Lukas Czerner (3): jbd2: make sure jh have b_transaction set in refile/unfile_buffer ext4: handle read only external journal device ext4: handle option set by mount flags correctly Lukas Wunner (4): spi: Prevent adding devices below an unregistering controller serial: pl011: Fix oops on -EPROBE_DEFER serial: pl011: Don't leak amba_ports entry on driver register error serial: 8250: Avoid error message on reprobe Luo Meng (1): ext4: only set last error block when check system zone failed Mao Wenan (1): virtio_ring: Avoid loop when vq is broken in virtqueue_poll Marc Zyngier (1): epoll: Keep a reference on files added to the check list Masami Hiramatsu (2): perf probe: Fix memory leakage when the probe point is not found uaccess: Add non-pagefault user-space read functions Max Reitz (1): xfs: Fix tail rounding in xfs_alloc_file_space() Mikulas Patocka (2): xfs: don't update mtime on COW faults dm writecache: handle DAX to partitions on persistent memory correctly Ming Lei (1): blk-mq: order adding requests to hctx->dispatch and checking SCHED_RESTART Muchun Song (1): kprobes: fix kill kprobe which has been marked as gone Namhyung Kim (1): perf jevents: Fix suspicious code in fixregex() Peter Xu (1): mm/hugetlb: fix calculation of adjust_range_if_pmd_sharing_possible Peter Zijlstra (1): cpuidle: Fixup IRQ state Qiushi Wu (1): PCI: Fix pci_create_slot() reference count leak Rafael J. Wysocki (1): PM: sleep: core: Fix the handling of pending runtime resume requests Ralph Campbell (1): mm/thp: fix __split_huge_pmd_locked() for migration PMD Robin Murphy (1): iommu/iova: Don't BUG on invalid PFNs Selvin Xavier (1): RDMA/bnxt_re: Do not add user qps to flushlist Sergey Senozhatsky (1): serial: 8250: change lock order in serial8250_do_startup() Sunghyun Jin (1): percpu: fix first chunk size calculation for populated bitmap Tejun Heo (1): libata: implement ATA_HORKAGE_MAX_TRIM_128M and apply to Sandisks Tetsuo Handa (1): vt: defer kfree() of vc_screenbuf in vc_do_resize() Valmer Huhn (1): serial: 8250_exar: Fix number of ports for Commtech PCIe cards Varun Prakash (1): scsi: target: iscsi: Fix data digest calculation Wei Yongjun (1): kernel/relay.c: fix memleak on destroy relay channel Xianting Tian (1): fs: prevent BUG_ON in submit_bh_wbc() Xunlei Pang (1): mm: memcg: fix memcg reclaim soft lockup Yang Xu (1): KEYS: reaching the keys quotas correctly kaixuxia (1): xfs: Fix deadlock between AGI and AGF with RENAME_WHITEOUT zhangyi (F) (1): jbd2: add the missing unlock_buffer() in the error path of jbd2_write_superblock() block/blk-core.c | 2 + block/blk-mq-sched.c | 9 ++ block/blk-mq.c | 9 ++ drivers/ata/libata-core.c | 5 +- drivers/ata/libata-scsi.c | 8 +- drivers/base/core.c | 12 +- drivers/base/power/main.c | 16 ++- drivers/char/tpm/tpm-chip.c | 9 +- drivers/char/tpm/tpm.h | 6 +- drivers/char/tpm/tpm2-space.c | 26 ++-- drivers/char/tpm/tpmrm-dev.c | 2 +- drivers/cpuidle/cpuidle.c | 3 +- drivers/firmware/efi/efi.c | 2 + drivers/infiniband/hw/bnxt_re/main.c | 3 +- drivers/iommu/iova.c | 4 +- drivers/md/dm-writecache.c | 12 +- drivers/net/macvlan.c | 21 ++- drivers/nvme/host/fc.c | 1 + drivers/nvme/host/rdma.c | 1 + drivers/pci/slot.c | 6 +- drivers/scsi/fcoe/fcoe_ctlr.c | 2 +- drivers/scsi/scsi_transport_iscsi.c | 2 +- drivers/spi/Kconfig | 3 + drivers/spi/spi.c | 30 +++- drivers/target/iscsi/iscsi_target.c | 17 ++- drivers/target/iscsi/iscsi_target_login.c | 6 +- drivers/target/iscsi/iscsi_target_login.h | 3 +- drivers/target/iscsi/iscsi_target_nego.c | 3 +- drivers/target/target_core_user.c | 2 +- drivers/tty/serial/8250/8250_core.c | 11 +- drivers/tty/serial/8250/8250_exar.c | 24 +++- drivers/tty/serial/8250/8250_port.c | 9 +- drivers/tty/serial/amba-pl011.c | 16 ++- drivers/tty/vt/vt.c | 5 +- drivers/tty/vt/vt_ioctl.c | 12 +- drivers/vfio/vfio_iommu_type1.c | 71 ++++++++- drivers/virtio/virtio_ring.c | 3 + fs/buffer.c | 9 ++ fs/eventpoll.c | 23 +-- fs/ext4/block_validity.c | 3 +- fs/ext4/namei.c | 6 +- fs/ext4/super.c | 147 ++++++++++++------- fs/fs-writeback.c | 83 ++++++----- fs/jbd2/journal.c | 4 +- fs/jbd2/transaction.c | 10 ++ fs/nfs/nfs4proc.c | 4 +- fs/signalfd.c | 10 +- fs/xfs/libxfs/xfs_attr_leaf.c | 4 +- fs/xfs/libxfs/xfs_bmap.c | 2 +- fs/xfs/xfs_bmap_util.c | 4 +- fs/xfs/xfs_buf.c | 8 +- fs/xfs/xfs_dquot.c | 9 +- fs/xfs/xfs_file.c | 12 +- fs/xfs/xfs_icache.c | 13 +- fs/xfs/xfs_inode.c | 110 ++++++++------ fs/xfs/xfs_ioctl.c | 5 +- fs/xfs/xfs_reflink.c | 1 + fs/xfs/xfs_sysfs.h | 6 +- fs/xfs/xfs_trans_ail.c | 4 +- fs/xfs/xfs_trans_dquot.c | 2 +- fs/xfs/xfs_trans_inode.c | 2 + include/linux/fs.h | 8 +- include/linux/libata.h | 1 + include/linux/log2.h | 2 +- include/linux/uaccess.h | 26 ++++ include/trace/events/writeback.h | 13 +- kernel/kprobes.c | 9 +- kernel/locking/lockdep_proc.c | 2 +- kernel/relay.c | 1 + mm/huge_memory.c | 40 +++--- mm/hugetlb.c | 24 ++-- mm/khugepaged.c | 7 +- mm/maccess.c | 167 ++++++++++++++++++++-- mm/page_alloc.c | 7 +- mm/percpu.c | 2 +- mm/vmscan.c | 8 ++ security/keys/key.c | 2 +- security/keys/keyctl.c | 4 +- tools/include/uapi/linux/perf_event.h | 2 +- tools/perf/pmu-events/jevents.c | 2 +- tools/perf/util/probe-finder.c | 2 +- 81 files changed, 864 insertions(+), 322 deletions(-) -- 2.25.1

1 78

[PATCH 01/11] xfs: Fix tail rounding in xfs_alloc_file_space()
by Yang Yingliang 15 Oct '20

15 Oct '20

From: Max Reitz <mreitz(a)redhat.com> mainline inclusion from mainline-v5.4-rc3 commit e093c4be760ebf46c131ae0dd6138865a22f46fa category: bugfix bugzilla: NA CVE: NA --------------------------- To ensure that all blocks touched by the range [offset, offset + count) are allocated, we need to calculate the block count from the difference of the range end (rounded up) and the range start (rounded down). Before this patch, we just round up the byte count, which may lead to unaligned ranges not being fully allocated: $ touch test_file $ block_size=$(stat -fc '%S' test_file) $ fallocate -o $((block_size / 2)) -l $block_size test_file $ xfs_bmap test_file test_file: 0: [0..7]: 1396264..1396271 1: [8..15]: hole There should not be a hole there. Instead, the first two blocks should be fully allocated. With this patch applied, the result is something like this: $ touch test_file $ block_size=$(stat -fc '%S' test_file) $ fallocate -o $((block_size / 2)) -l $block_size test_file $ xfs_bmap test_file test_file: 0: [0..15]: 11024..11039 Signed-off-by: Max Reitz <mreitz(a)redhat.com> Reviewed-by: Carlos Maiolino <cmaiolino(a)redhat.com> Reviewed-by: Christoph Hellwig <hch(a)lst.de> Reviewed-by: Darrick J. Wong <darrick.wong(a)oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong(a)oracle.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> Reviewed-by: zhangyi (F) <yi.zhang(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- fs/xfs/xfs_bmap_util.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/fs/xfs/xfs_bmap_util.c b/fs/xfs/xfs_bmap_util.c index 3e1dd66bd676..e54f5aed30a9 100644 --- a/fs/xfs/xfs_bmap_util.c +++ b/fs/xfs/xfs_bmap_util.c @@ -869,6 +869,7 @@ xfs_alloc_file_space( xfs_filblks_t allocatesize_fsb; xfs_extlen_t extsz, temp; xfs_fileoff_t startoffset_fsb; + xfs_fileoff_t endoffset_fsb; int nimaps; int quota_flag; int rt; @@ -896,7 +897,8 @@ xfs_alloc_file_space( imapp = &imaps[0]; nimaps = 1; startoffset_fsb = XFS_B_TO_FSBT(mp, offset); - allocatesize_fsb = XFS_B_TO_FSB(mp, count); + endoffset_fsb = XFS_B_TO_FSB(mp, offset + count); + allocatesize_fsb = endoffset_fsb - startoffset_fsb; /* * Allocate file space until done or until there is an error -- 2.25.1

1 10

[PATCH kernel-5.5] RISCV-V: KVM: use __u64 define variables instead of u64 in kvm.h
by l00484210 13 Oct '20

13 Oct '20

From: MingWang Li <limingwang(a)huawei.com> euleros inclusion category: bugfix bugzilla: NA CVE: NA one error occurred while building qemu on riscv system, which is unknown type name 'u64'. Link: https://gitee.com/openeuler/kernel/issues/I1XUB0 Signed-off-by: MingWang Li <limingwang(a)huawei.com> --- arch/riscv/include/uapi/asm/kvm.h | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/arch/riscv/include/uapi/asm/kvm.h b/arch/riscv/include/uapi/asm/kvm.h index 65cd00654..c2b4ad6b1 100644 --- a/arch/riscv/include/uapi/asm/kvm.h +++ b/arch/riscv/include/uapi/asm/kvm.h @@ -75,10 +75,10 @@ struct kvm_riscv_csr { /* TIMER registers for KVM_GET_ONE_REG and KVM_SET_ONE_REG */ struct kvm_riscv_timer { - u64 frequency; - u64 time; - u64 compare; - u64 state; + __u64 frequency; + __u64 time; + __u64 compare; + __u64 state; }; /* Possible states for kvm_riscv_timer */ @@ -110,17 +110,17 @@ struct kvm_riscv_timer { /* Timer registers are mapped as type 4 */ #define KVM_REG_RISCV_TIMER (0x04 << KVM_REG_RISCV_TYPE_SHIFT) #define KVM_REG_RISCV_TIMER_REG(name) \ - (offsetof(struct kvm_riscv_timer, name) / sizeof(u64)) + (offsetof(struct kvm_riscv_timer, name) / sizeof(__u64)) /* F extension registers are mapped as type 5 */ #define KVM_REG_RISCV_FP_F (0x05 << KVM_REG_RISCV_TYPE_SHIFT) #define KVM_REG_RISCV_FP_F_REG(name) \ - (offsetof(struct __riscv_f_ext_state, name) / sizeof(u32)) + (offsetof(struct __riscv_f_ext_state, name) / sizeof(__u32)) /* D extension registers are mapped as type 6 */ #define KVM_REG_RISCV_FP_D (0x06 << KVM_REG_RISCV_TYPE_SHIFT) #define KVM_REG_RISCV_FP_D_REG(name) \ - (offsetof(struct __riscv_d_ext_state, name) / sizeof(u64)) + (offsetof(struct __riscv_d_ext_state, name) / sizeof(__u64)) #endif -- 2.19.1

2 1

[PATCH 1/2] netfilter: nf_tables: incorrect enum nft_list_attributes definition
by Yang Yingliang 13 Oct '20

13 Oct '20

From: Pablo Neira Ayuso <pablo(a)netfilter.org> stable inclusion from linux-4.19.144 commit 3f21d1dd7cafb0230dc141e64ec5da622b3b1c46 -------------------------------- [ Upstream commit da9125df854ea48a6240c66e8a67be06e2c12c03 ] This should be NFTA_LIST_UNSPEC instead of NFTA_LIST_UNPEC, all other similar attribute definitions are postfixed with _UNSPEC. Fixes: 96518518cc41 ("netfilter: add nftables") Signed-off-by: Pablo Neira Ayuso <pablo(a)netfilter.org> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> Signed-off-by: Aichun Li <liaichun(a)huawei.com> Reviewed-by: wangxiaopeng <wangxiaopeng7(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- include/uapi/linux/netfilter/nf_tables.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/uapi/linux/netfilter/nf_tables.h b/include/uapi/linux/netfilter/nf_tables.h index 44c8ba1f3652..8586c816bea0 100644 --- a/include/uapi/linux/netfilter/nf_tables.h +++ b/include/uapi/linux/netfilter/nf_tables.h @@ -132,7 +132,7 @@ enum nf_tables_msg_types { * @NFTA_LIST_ELEM: list element (NLA_NESTED) */ enum nft_list_attributes { - NFTA_LIST_UNPEC, + NFTA_LIST_UNSPEC, NFTA_LIST_ELEM, __NFTA_LIST_MAX }; -- 2.25.1

1 1

[PATCH 1/6] blk-mq: insert passthrough request into hctx->dispatch directly
by Yang Yingliang 13 Oct '20

13 Oct '20

From: Ming Lei <ming.lei(a)redhat.com> mainline inclusion from mainline-5.6-rc4 commit 01e99aeca3979600302913cef3f89076786f32c8 category: bugfix bugzilla: 42777 CVE: NA --------------------------- For some reason, device may be in one situation which can't handle FS request, so STS_RESOURCE is always returned and the FS request will be added to hctx->dispatch. However passthrough request may be required at that time for fixing the problem. If passthrough request is added to scheduler queue, there isn't any chance for blk-mq to dispatch it given we prioritize requests in hctx->dispatch. Then the FS IO request may never be completed, and IO hang is caused. So passthrough request has to be added to hctx->dispatch directly for fixing the IO hang. Fix this issue by inserting passthrough request into hctx->dispatch directly together withing adding FS request to the tail of hctx->dispatch in blk_mq_dispatch_rq_list(). Actually we add FS request to tail of hctx->dispatch at default, see blk_mq_request_bypass_insert(). Then it becomes consistent with original legacy IO request path, in which passthrough request is always added to q->queue_head. Cc: Dongli Zhang <dongli.zhang(a)oracle.com> Cc: Christoph Hellwig <hch(a)infradead.org> Cc: Ewan D. Milne <emilne(a)redhat.com> Signed-off-by: Ming Lei <ming.lei(a)redhat.com> Signed-off-by: Jens Axboe <axboe(a)kernel.dk> Conflicts: block/blk-flush.c block/blk-mq.c block/blk-mq-sched.c Signed-off-by: Yu Kuai <yukuai3(a)huawei.com> Reviewed-by: Yufen Yu <yuyufen(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- block/blk-flush.c | 2 +- block/blk-mq-sched.c | 22 +++++++++++++++------- block/blk-mq.c | 18 +++++++++++------- block/blk-mq.h | 3 ++- 4 files changed, 29 insertions(+), 16 deletions(-) diff --git a/block/blk-flush.c b/block/blk-flush.c index 256fa1ccc2bd..2a8369eb6c1c 100644 --- a/block/blk-flush.c +++ b/block/blk-flush.c @@ -495,7 +495,7 @@ void blk_insert_flush(struct request *rq) if ((policy & REQ_FSEQ_DATA) && !(policy & (REQ_FSEQ_PREFLUSH | REQ_FSEQ_POSTFLUSH))) { if (q->mq_ops) - blk_mq_request_bypass_insert(rq, false); + blk_mq_request_bypass_insert(rq, false, false); else list_add_tail(&rq->queuelist, &q->queue_head); return; diff --git a/block/blk-mq-sched.c b/block/blk-mq-sched.c index e98f51b86342..623c258d0b34 100644 --- a/block/blk-mq-sched.c +++ b/block/blk-mq-sched.c @@ -357,13 +357,19 @@ static bool blk_mq_sched_bypass_insert(struct blk_mq_hw_ctx *hctx, bool has_sched, struct request *rq) { - /* dispatch flush rq directly */ - if (rq->rq_flags & RQF_FLUSH_SEQ) { - spin_lock(&hctx->lock); - list_add(&rq->queuelist, &hctx->dispatch); - spin_unlock(&hctx->lock); + /* + * dispatch flush and passthrough rq directly + * + * passthrough request has to be added to hctx->dispatch directly. + * For some reason, device may be in one situation which can't + * handle FS request, so STS_RESOURCE is always returned and the + * FS request will be added to hctx->dispatch. However passthrough + * request may be required at that time for fixing the problem. If + * passthrough request is added to scheduler queue, there isn't any + * chance to dispatch it given we prioritize requests in hctx->dispatch. + */ + if ((rq->rq_flags & RQF_FLUSH_SEQ) || blk_rq_is_passthrough(rq)) return true; - } if (has_sched) rq->rq_flags |= RQF_SORTED; @@ -387,8 +393,10 @@ void blk_mq_sched_insert_request(struct request *rq, bool at_head, WARN_ON(e && (rq->tag != -1)); - if (blk_mq_sched_bypass_insert(hctx, !!e, rq)) + if (blk_mq_sched_bypass_insert(hctx, !!e, rq)) { + blk_mq_request_bypass_insert(rq, at_head, false); goto run; + } if (e && e->type->ops.mq.insert_requests) { LIST_HEAD(list); diff --git a/block/blk-mq.c b/block/blk-mq.c index 59c92961cc20..a9138235b870 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -721,7 +721,7 @@ static void blk_mq_requeue_work(struct work_struct *work) * merge. */ if (rq->rq_flags & RQF_DONTPREP) - blk_mq_request_bypass_insert(rq, false); + blk_mq_request_bypass_insert(rq, false, false); else blk_mq_sched_insert_request(rq, true, false, false); } @@ -1227,7 +1227,7 @@ bool blk_mq_dispatch_rq_list(struct request_queue *q, struct list_head *list, bool needs_restart; spin_lock(&hctx->lock); - list_splice_init(list, &hctx->dispatch); + list_splice_tail_init(list, &hctx->dispatch); spin_unlock(&hctx->lock); /* @@ -1590,13 +1590,17 @@ void __blk_mq_insert_request(struct blk_mq_hw_ctx *hctx, struct request *rq, * Should only be used carefully, when the caller knows we want to * bypass a potential IO scheduler on the target device. */ -void blk_mq_request_bypass_insert(struct request *rq, bool run_queue) +void blk_mq_request_bypass_insert(struct request *rq, bool at_head, + bool run_queue) { struct blk_mq_ctx *ctx = rq->mq_ctx; struct blk_mq_hw_ctx *hctx = blk_mq_map_queue(rq->q, ctx->cpu); spin_lock(&hctx->lock); - list_add_tail(&rq->queuelist, &hctx->dispatch); + if (at_head) + list_add(&rq->queuelist, &hctx->dispatch); + else + list_add_tail(&rq->queuelist, &hctx->dispatch); spin_unlock(&hctx->lock); if (run_queue) @@ -1776,7 +1780,7 @@ static blk_status_t __blk_mq_try_issue_directly(struct blk_mq_hw_ctx *hctx, if (bypass_insert) return BLK_STS_RESOURCE; - blk_mq_request_bypass_insert(rq, run_queue); + blk_mq_request_bypass_insert(rq, false, run_queue); return BLK_STS_OK; } @@ -1792,7 +1796,7 @@ static void blk_mq_try_issue_directly(struct blk_mq_hw_ctx *hctx, ret = __blk_mq_try_issue_directly(hctx, rq, cookie, false); if (ret == BLK_STS_RESOURCE || ret == BLK_STS_DEV_RESOURCE) - blk_mq_request_bypass_insert(rq, true); + blk_mq_request_bypass_insert(rq, false, true); else if (ret != BLK_STS_OK) blk_mq_end_request(rq, ret); @@ -1827,7 +1831,7 @@ void blk_mq_try_issue_list_directly(struct blk_mq_hw_ctx *hctx, if (ret != BLK_STS_OK) { if (ret == BLK_STS_RESOURCE || ret == BLK_STS_DEV_RESOURCE) { - blk_mq_request_bypass_insert(rq, + blk_mq_request_bypass_insert(rq, false, list_empty(list)); break; } diff --git a/block/blk-mq.h b/block/blk-mq.h index a6094c27b827..debc646e1bed 100644 --- a/block/blk-mq.h +++ b/block/blk-mq.h @@ -64,7 +64,8 @@ int blk_mq_alloc_rqs(struct blk_mq_tag_set *set, struct blk_mq_tags *tags, */ void __blk_mq_insert_request(struct blk_mq_hw_ctx *hctx, struct request *rq, bool at_head); -void blk_mq_request_bypass_insert(struct request *rq, bool run_queue); +void blk_mq_request_bypass_insert(struct request *rq, bool at_head, + bool run_queue); void blk_mq_insert_requests(struct blk_mq_hw_ctx *hctx, struct blk_mq_ctx *ctx, struct list_head *list); -- 2.25.1

1 5

[PATCH] arm64/ascend: Fix register_persistent_clock definition
by Yang Yingliang 13 Oct '20

13 Oct '20

From: Fang Lijun <fanglijun3(a)huawei.com> ascend inclusion category: bugfix bugzilla: NA CVE: NA ------------------------------------------------- The register_persistent_clock will be called after kernel init, so it can not be defined as __init. Fixes: 76ab899d73d6 ("arm64/ascend: Implement the read_persistend_clock64 for aarch64") Signed-off-by: Fang Lijun <fanglijun3(a)huawei.com> Reviewed-by: Hanjun Guo <guohanjun(a)huawei.com> Reviewed-by: Ding Tianhong <dingtianhong(a)huawei.com> Signed-off-by: Yang Yingliang <yangyingliang(a)huawei.com> --- arch/arm64/kernel/time.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm64/kernel/time.c b/arch/arm64/kernel/time.c index c06c3feb6772..902d0f0b4f7b 100644 --- a/arch/arm64/kernel/time.c +++ b/arch/arm64/kernel/time.c @@ -77,7 +77,7 @@ void read_persistent_clock64(struct timespec64 *ts) __read_persistent_clock(ts); } -int __init register_persistent_clock(clock_access_fn read_persistent) +int register_persistent_clock(clock_access_fn read_persistent) { /* Only allow the clockaccess functions to be registered once */ if (__read_persistent_clock == dummy_clock_access) { -- 2.25.1

1 0

[PATCH 00/38] backport stable-4.19.150
by Yang Yingliang 13 Oct '20

13 Oct '20

Al Viro (4): epoll: do not insert into poll queues until all sanity checks are done epoll: replace ->visited/visited_list with generation count epoll: EPOLL_CTL_ADD: close the race in decision to take fast path ep_create_wakeup_source(): dentry name can change under you... Bartosz Golaszewski (1): gpio: mockup: fix resource leak in error path Bryan O'Donoghue (1): USB: gadget: f_ncm: Fix NDP16 datagram validation Chaitanya Kulkarni (1): nvme-core: get/put ctrl and transport module in nvme_dev_open/release() Chris Packham (2): spi: fsl-espi: Only process interrupts for expected events pinctrl: mvebu: Fix i2c sda definition for 98DX3236 Dinh Nguyen (1): clk: socfpga: stratix10: fix the divider for the emac_ptp_free_clk Felix Fietkau (1): mac80211: do not allow bigger VHT MPDUs than the hardware supports Greg Kroah-Hartman (1): Linux 4.19.150 Hans de Goede (1): mmc: sdhci: Workaround broken command queuing on Intel GLK based IRBIS models James Smart (1): nvme-fc: fail new connections to a deleted host or remote port Jean Delvare (1): drm/amdgpu: restore proper ref count in amdgpu_display_crtc_set_config Jeffrey Mitchell (1): nfs: Fix security label length not being reset Jiri Kosina (1): Input: i8042 - add nopnp quirk for Acer Aspire 5 A515 Laurent Dufour (2): mm: replace memmap_context by meminit_context mm: don't rely on system state to detect hot-plug operations Lucy Yan (1): net: dec: de2104x: Increase receive ring size for Tulip Marek Szyprowski (1): clk: samsung: exynos4: mark 'chipid' clock as CLK_IGNORE_UNUSED Martin Cerveny (1): drm/sun4i: mixer: Extend regmap max_register Nicolas VINCENT (1): i2c: cpm: Fix i2c_ram structure Olympia Giannou (1): rndis_host: increase sleep time in the query-response loop Sebastien Boeuf (1): net: virtio_vsock: Enhance connection semantics Stefano Garzarella (3): vsock/virtio: use RCU to avoid use-after-free on the_virtio_vsock vsock/virtio: stop workers during the .remove() vsock/virtio: add transport parameter to the virtio_transport_reset_no_sock() Steven Rostedt (VMware) (1): ftrace: Move RCU is watching check after recursion check Taiping Lai (1): gpio: sprd: Clear interrupt when setting the type as edge Thibaut Sautereau (1): random32: Restore __latent_entropy attribute on net_rand_state Vincent Huang (1): Input: trackpoint - enable Synaptics trackpoints Will McVicker (1): netfilter: ctnetlink: add a range check for l3/l4 protonum Xie He (3): drivers/net/wan/hdlc_fr: Add needed_headroom for PVC devices drivers/net/wan/lapbether: Make skb->protocol consistent with the header drivers/net/wan/hdlc: Set skb->protocol before transmitting Yu Kuai (1): iommu/exynos: add missing put_device() call in exynos_iommu_of_xlate() dillon min (1): gpio: tc35894: fix up tc35894 interrupt configuration Makefile | 2 +- arch/ia64/mm/init.c | 6 +- drivers/base/node.c | 84 ++++--- drivers/clk/samsung/clk-exynos4.c | 4 +- drivers/clk/socfpga/clk-s10.c | 2 +- drivers/gpio/gpio-mockup.c | 2 + drivers/gpio/gpio-sprd.c | 3 + drivers/gpio/gpio-tc3589x.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_display.c | 2 +- drivers/gpu/drm/sun4i/sun8i_mixer.c | 2 +- drivers/i2c/busses/i2c-cpm.c | 3 + drivers/input/mouse/trackpoint.c | 2 + drivers/input/serio/i8042-x86ia64io.h | 7 + drivers/iommu/exynos-iommu.c | 8 +- drivers/mmc/host/sdhci-pci-core.c | 3 +- drivers/net/ethernet/dec/tulip/de2104x.c | 2 +- drivers/net/usb/rndis_host.c | 2 +- drivers/net/wan/hdlc_cisco.c | 1 + drivers/net/wan/hdlc_fr.c | 6 +- drivers/net/wan/hdlc_ppp.c | 1 + drivers/net/wan/lapbether.c | 4 +- drivers/nvme/host/core.c | 15 ++ drivers/nvme/host/fc.c | 6 +- drivers/pinctrl/mvebu/pinctrl-armada-xp.c | 2 +- drivers/spi/spi-fsl-espi.c | 5 +- drivers/usb/gadget/function/f_ncm.c | 30 +-- drivers/vhost/vsock.c | 94 +++---- fs/eventpoll.c | 71 +++--- fs/nfs/dir.c | 3 + include/linux/mm.h | 2 +- include/linux/mmzone.h | 11 +- include/linux/node.h | 11 +- include/linux/virtio_vsock.h | 3 +- kernel/trace/ftrace.c | 6 +- lib/random32.c | 2 +- mm/memory_hotplug.c | 5 +- mm/page_alloc.c | 11 +- net/mac80211/vht.c | 8 +- net/netfilter/nf_conntrack_netlink.c | 2 + net/vmw_vsock/virtio_transport.c | 265 +++++++++++++------- net/vmw_vsock/virtio_transport_common.c | 13 +- 41 files changed, 416 insertions(+), 297 deletions(-) -- 2.25.1

1 38

openEuler 20.09 kernel release
by Xie XiuQi 10 Oct '20

10 Oct '20

很高兴 openEuler 20.09 kernel 如期而至，该版本包含众多特性，欢迎通过 openEuler 版本来体验。 openEuler-20.09 kernel tree 位于: git@gitee.com:openeuler/kernel.git openEuler-20.09 openEuler kernel 项目位于: https://gitee.com/openeuler/kernel openEuler 20.09 kernel 依然基于社区的 Linux 4.19 LTS 内核，在 openEuler 1.0 LTS 基础上往前演进(跟踪到 4.19.140)，在这大概6个月的时间内，合入并使能了多个特性, 下面从架构支持，驱动支持，以及内核特性支持等几个方面来进行介绍。 1 架构支持 1.1 支持海光CPU-Hygon Dhyana 海光的Peng Wu向openEuler贡献了Hygon CPU的支持补丁集，通过这个补丁集，新增了 Hygon CPU 的ACPI, cpufreq，perf，KVM，NTB以及Hygon CPU的识别等支持，通过这个补丁集，openEuler完整支持海光x86 CPU。 1.2 支持昇腾芯片包括CPU和AI CPU共享虚拟地址的SVM（shared virtual memory）特性，以及芯片的硬件支持。 1.3 支持鲲鹏PC平台鲲鹏PC CPU基于鲲鹏920,8核。20.09 版本内核加入了对PC的休眠支持（架构特性），以及I2C控制器等支持。 1.4 支持 ARM v8.x 特性 - ARMv8.0 CRC32 instructions - ARMv8.0-SB Control of speculation - Armv8.2 Cache clean to Point of Deep Persistence - ARMv8.2-TTCNP, Translation table Common not private translations - ARMv8.4-TLBI, TLB maintenance and TLB range instructions - Armv8.5-RNG, Random number generator - Armv8.5-BTI, Branch target identification - Armv8.5-E0PD, Preventing EL0 access to halves of address map - ARMv8.6 Data gathering hint - ARMv8.6 Matrix multiply extension 1.5 支持 risc-v 支持 risc-v kvm. (openEuler 20.09 通过 kernel 5.5 来支持 risc-v) kernel tree 位于： git@gitee.com:openeuler/kernel.git kernel-5.5 2. 新介质以及新驱动支持 2.1 ARM64支持 persistence memory ARMv8.2引入新的指令支持 persistent memory，鲲鹏920处理器完整支持该特性，通过打开内核的 ACPI NFIT以及libnvdimm的配置文件，就能在鲲鹏上完美支持NVDIMM介质内存。 2.2 1822 网卡驱动增强 - 1822网卡支持x86架构 - 1822网卡支持128队列 2.3 3408/3416 RAID卡规避 3408iMR/3416iMRraid卡在ARM64平台上（ARM64生态问题）使用存在问题，开启SMMU后，会存在不兼容，通过SMMU 设备bypass特性，规避不兼容问题。 2.4 Huawei iBMA 驱动 iBMA 驱动是一组 BMC 与带内系统通信的驱动，系统宕机时，能辅助保存宕机时的运行状态信息，帮助定位故障原因。 2.5 hns3: update hns3 version to 1.9.38.8 2.6 TIPC: enable TIPC module by default 3. 存储/IO特性支持 3.1 bcache稳定性提升 bcache Maintainer Coly回合了大量bcache的bugfix以及特性，大幅提升bcache稳定性和可用性 4. 性能优化特性 4.1 sched Task Steal 回合社区的Task Steal特性，通过提升CPU利用率，替身MySQL数据库mix场景的性能10%+； 4.2 REFCOUNT_FULL以及CMPXCHG_LOOP()优化通过atomic_fetch_*操作来降低cmpxchg()带来的性能开销，从而提升refcount机制性能，带来文件读写等场景的性能提升，在ARM64服务器上尤其明显，空文件读写的benchmark 20+提升。 4.3 Linux's vmalloc Seeing "Large Performance Benefits" 优化 vmalloc 性能，内核针对vmalloc区的查找做了优化，从O(N)到~O(log(N) 。 See: https://www.phoronix.com/scan.php?page=news_item&px=Linux-5.2-vmalloc-Perfo… 4.4 MPAM特性增强支持通过ACPI获取系统MPAM资源； 4.5 percpu refcount for pagecache pagecache的生命周期管理是通过refcount来管理的，在多核并发压力下，读取文件cache 的atomic操作成为瓶颈，因此引入percpu refcount mechanism来提升性能，针对nginx http测试，性能有1倍+的提升 4.6 ARM64 clear_page()性能优化通过‘stnp’代替'dc zva'指令，提升clear_page()的性能，在鲲鹏920上，针对hugetlb 测试，有53%的性能提升。 4.7 调度，优化关键进程的抢占时延，提升响应速度对 SCHED_IDLE 策略的优化，之前选核的时候只选择 idle 的CPU，现在是如果那个CPU 上运行的是 SCHED_IDLE 的进程，也可以选择到，由于运行SCHED_IDLE进程的CPU并不是 idle，没有唤醒时延，能提升响应速度。 5. 安全特性 5.1 支持AppArmor AppArmor相比Selinux使用起来简单，但之前版本未开启AppArmor支持。20.09版本支持 AppArmor，但默认的安全策略依然是Selinux，需要在启动参数传入security=apparmor使能 5.2 IMA摘要列表增强相比内核社区原生 IMA 机制，IMA 摘要列表扩展从安全性、性能、易用性三个方面进行了改良，助力完整性保护机制在生产场景下落地，并具有三大优势：极致安全原生 IMA 机制要求在现网环境下预先生成并标记文件扩展属性，访问文件时将文件扩展属性作为参考值，信任链不完整。摘要列表扩展将文件参考摘要值保存在内核空间中，构建阶段通过摘要列表的形式携带在发布的 rpm 包中，安装 rpm 包的同时导入摘要列表并执行验签，确保了参考值来自于软件发行商，实现了完整的信任链。惊艳的性能表现 IMA 度量场景下，摘要列表扩展在确保安全性的前提下，减少了不必要的 PCR 扩展操作，开启度量时性能损失小于 5%，相比原生 IMA 度量性能提升高达 50%。IMA 评估场景下，摘要列表扩展将签名验证统一移动到启动阶段进行，避免每次访问文件时都执行验签，相比原生 IMA 评估场景提升运行阶段文件访问的性能约 20%。快速部署，平滑升级原生 IMA 机制在初次部署或每次更新软件包时，需要切换到 ﬁx 模式手动标记文件扩展属性后再重启进入 enforce 模式，才能正常访问安装的程序。摘要列表扩展可实现安装完成后开箱即用，且允许直接在enforce 模式下安装或升级 rpm 包，无需重启和手动标记即可使用，实现了用户感知最小化，适合现网环境下快速批量部署。 6. 功耗特性 6.1 支持 CPU idle TEO governor TEO 是新引入的CPU调频算法，更适合 tickless 系统，可以为CPU选择更合适的节能状态。 7 调测特性 7.1 support livepatch without ftrace 8. 虚拟化特性 8.1 支持双层调度双层调度就是让Hypervisor的调度器感知到VM的VCPU上跑什么应用。让VM的调度器感知到 Hypervisor 层物理CPU压力。两层调度感知，整机达到最好的业务性能。 8.2 支持 PMU nmi watchdog 8.3 支持 SmartPolling （guest idle poll) ARM 平台上 cpu idle 支持 smart polling，用来提升性能。 8.4 支持 vmtop, 用于虚拟机性能指标监测 8.5 支持ARM架构PV qspinlock功能，提升vcpu的自旋锁等锁性能 --- openEuler Kernel SIG

1 0