February 2025 - Kernel - mailweb.openeuler.org

[PATCH openEuler-1.0-LTS] cifs: Prevent NULL pointer dereference caused by cifs_sb->rsize is 0
by Wang Zhaolong 25 Feb '25

25 Feb '25

hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/IBO3FG CVE: NA -------------------------------- When cifs_sb->rsize is 0, cifs_send_async_read() fails to allocate the pages array using iov_iter_get_pages_alloc(), resulting in data->pages being NULL. This eventually triggers a NULL pointer dereference in the callback function uncached_fill_pages(). cifs_sb->rsize is set to the minimum value between SMB3_DEFAULT_IOSIZE and server->max_read. SMB3_DEFAULT_IOSIZE is 4MB. There are two reasons that cause cifs_sb->rsize to be 0: 1. During the negotiation process, if the server returns an rsp->MaxReadSize of 0, server->max_read is assigned 0. And the calculated value of cifs_sb->rsize becomes 0. 2. In the cifsd thread, if cifs_reconnect() is invoked, server->max_read is set to 0. Prior to the completion of the reconnection process, the calculated value of cifs_sb->rsize becomes 0. It is important to note that cifs_sb->rsize is only set to 0 during the mount phase. Therefore, a straightforward approach to prevent potential NULL pointer dereferences in the future is to implement a check for cifs_sb->rsize during the mount stage. If the value is found to be 0, the mount process should exit early to avoid any subsequent NULL pointer dereferences. Fixes: aa24d1e96924 ("CIFS: Process reconnects for SMB2 shares") Signed-off-by: Wang Zhaolong <wangzhaolong1(a)huawei.com> --- fs/cifs/connect.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index bd9365d375f9..a0a61dc752af 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -4243,6 +4243,11 @@ cifs_mount(struct cifs_sb_info *cifs_sb, struct smb_vol *volume_info) cifs_sb->wsize = server->ops->negotiate_wsize(tcon, volume_info); cifs_sb->rsize = server->ops->negotiate_rsize(tcon, volume_info); + if (cifs_sb->rsize == 0) { + cifs_dbg(VFS, "Negotiated rsize is 0, mount failed\n"); + rc = -EINVAL; + goto mount_fail_check; + } remote_path_check: #ifdef CONFIG_CIFS_DFS_UPCALL -- 2.34.3

2 1

[PATCH OLK-5.10 0/2] smb: client: Fix netns refcount imbalance
by Wang Zhaolong 25 Feb '25

25 Feb '25

*** BLURB HERE *** Enzo Matsumiya (1): smb: client: fix TCP timers deadlock after rmmod Wang Zhaolong (1): smb: client: Fix netns refcount imbalance causing leaks and use-after-free fs/cifs/connect.c | 33 ++++++++++++++++++++++++--------- 1 file changed, 24 insertions(+), 9 deletions(-) -- 2.34.3

2 3

[PATCH OLK-6.6] NFS: remove unused variable nfs_rpcstat
by Wang Zhaolong 25 Feb '25

25 Feb '25

From: Trond Myklebust <trond.myklebust(a)hammerspace.com> mainline inclusion from mainline-v6.9-rc1 commit cdbd321ac15e1e663c15fc0433024f6487531f27 category: cleanup bugzilla: https://gitee.com/openeuler/kernel/issues/IBOSCC Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- Signed-off-by: Trond Myklebust <trond.myklebust(a)hammerspace.com> Signed-off-by: Wang Zhaolong <wangzhaolong1(a)huawei.com> --- fs/nfs/client.c | 4 ---- 1 file changed, 4 deletions(-) diff --git a/fs/nfs/client.c b/fs/nfs/client.c index 62607d52bfa5..d9b10c42a554 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -76,10 +76,6 @@ const struct rpc_program nfs_program = { .pipe_dir_name = NFS_PIPE_DIRNAME, }; -struct rpc_stat nfs_rpcstat = { - .program = &nfs_program -}; - static struct nfs_subversion *find_nfs_version(unsigned int version) { struct nfs_subversion *nfs; -- 2.34.3

2 1

[PATCH OLK-6.6] cifs: Prevent NULL pointer dereference caused by cifs_sb->rsize is 0
by Wang Zhaolong 25 Feb '25

25 Feb '25

hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/IBO3FG -------------------------------- When cifs_sb->rsize is 0, cifs_send_async_read() fails to allocate the pages array using iov_iter_get_pages_alloc(), resulting in data->pages being NULL. This eventually triggers a NULL pointer dereference in the callback function uncached_fill_pages(). cifs_sb->rsize is set to the minimum value between SMB3_DEFAULT_IOSIZE and server->max_read. SMB3_DEFAULT_IOSIZE is 4MB. There are two reasons that cause cifs_sb->rsize to be 0: 1. During the negotiation process, if the server returns an rsp->MaxReadSize of 0, server->max_read is assigned 0. And the calculated value of cifs_sb->rsize becomes 0. 2. In the cifsd thread, if cifs_reconnect() is invoked, server->max_read is set to 0. Prior to the completion of the reconnection process, the calculated value of cifs_sb->rsize becomes 0. It is important to note that cifs_sb->rsize is only set to 0 during the mount phase. Therefore, a straightforward approach to prevent potential NULL pointer dereferences in the future is to implement a check for cifs_sb->rsize during the mount stage. If the value is found to be 0, the mount process should exit early to avoid any subsequent NULL pointer dereferences. Fixes: aa24d1e96924 ("CIFS: Process reconnects for SMB2 shares") Signed-off-by: Wang Zhaolong <wangzhaolong1(a)huawei.com> --- fs/smb/client/connect.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/fs/smb/client/connect.c b/fs/smb/client/connect.c index 4a15e739a436..7ba1d1b1256b 100644 --- a/fs/smb/client/connect.c +++ b/fs/smb/client/connect.c @@ -3552,6 +3552,11 @@ int cifs_mount_get_tcon(struct cifs_mount_ctx *mnt_ctx) if ((cifs_sb->ctx->rsize == 0) || (cifs_sb->ctx->rsize > server->ops->negotiate_rsize(tcon, ctx))) cifs_sb->ctx->rsize = server->ops->negotiate_rsize(tcon, ctx); + if (cifs_sb->ctx->rsize == 0) { + cifs_dbg(VFS, "Negotiated rsize is 0, get connect failed\n"); + rc = -EINVAL; + goto out; + } /* * The cookie is initialized from volume info returned above. -- 2.34.3

2 1

[PATCH OLK-5.10 0/2] smb: client: Fix netns refcount imbalance
by Wang Zhaolong 25 Feb '25

25 Feb '25

Enzo Matsumiya (1): smb: client: fix TCP timers deadlock after rmmod Wang Zhaolong (1): smb: client: Fix netns refcount imbalance causing leaks and use-after-free fs/cifs/connect.c | 33 ++++++++++++++++++++++++--------- 1 file changed, 24 insertions(+), 9 deletions(-) -- 2.34.3

2 3

[PATCH OLK-6.6 00/22] Add support for heterogeneous live migration
by yangjinqian 25 Feb '25

25 Feb '25

Patch#1 Allow userspace to get the writable masks for feature ID registers Patch#2 Document KVM_ARM_GET_REG_WRITABLE_MASKS Patch#3 Use guest ID register values for the sake of emulation Patch#4 Advertise selected DebugVer in DBGDIDR.Version Patch#5 Reject attempts to set invalid debug arch version Patch#6 Bump up the default KVM sanitised debug version to v8p8 Patch#7 Allow userspace to change ID_AA64ISAR{0-2}_EL1 Patch#8 Allow userspace to change ID_AA64MMFR{0-2}_EL1 Patch#9 Allow userspace to change ID_AA64PFR0_EL1 Patch#10 Allow userspace to change ID_AA64ZFR0_EL1 Patch#11 Document vCPU feature selection UAPIs Patch#12 Make the exposed feature bits in AA64DFR0_EL1 writable from userspace Patch#13 Disable MPAM visibility by default and ignore VMM writes Patch#14 Use read-only helper for reading VM ID registers Patch#15 Add helper for writing ID regs Patch#16 Treat CTR_EL0 as a VM feature ID register Patch#17 Rename functions for invariant sys regs Patch#18 Make the L1Ip feature bits in CTR_EL0 writable from userspace Patch#19 Show writable masks for feature registers Patch#20 Make HCX writable from userspace Patch#21 Add support for HIP09 Spectre-BHB mitigation Patch#22 add cnp verification on Hisilicon erratum 162100125 James Morse (1): KVM: arm64: Disable MPAM visibility by default and ignore VMM writes Jing Zhang (5): KVM: arm64: Allow userspace to get the writable masks for feature ID registers KVM: arm64: Document KVM_ARM_GET_REG_WRITABLE_MASKS KVM: arm64: Use guest ID register values for the sake of emulation KVM: arm64: Allow userspace to change ID_AA64MMFR{0-2}_EL1 KVM: arm64: Allow userspace to change ID_AA64PFR0_EL1 Oliver Upton (8): KVM: arm64: Advertise selected DebugVer in DBGDIDR.Version KVM: arm64: Reject attempts to set invalid debug arch version KVM: arm64: Bump up the default KVM sanitised debug version to v8p8 KVM: arm64: Allow userspace to change ID_AA64ISAR{0-2}_EL1 KVM: arm64: Allow userspace to change ID_AA64ZFR0_EL1 KVM: arm64: Document vCPU feature selection UAPIs KVM: arm64: Use read-only helper for reading VM ID registers KVM: arm64: Add helper for writing ID regs Sebastian Ott (3): KVM: arm64: Treat CTR_EL0 as a VM feature ID register KVM: arm64: rename functions for invariant sys regs KVM: arm64: show writable masks for feature registers Shameer Kolothum (2): KVM: arm64: Make the exposed feature bits in AA64DFR0_EL1 writable from userspace KVM: arm64: Make the L1Ip feature bits in CTR_EL0 writable from userspace yangjinqian (3): KVM: arm64: Make HCX writable from userspace arm64: Add support for HIP09 Spectre-BHB mitigation arm64: add cnp verification on Hisilicon erratum 162100125 Documentation/virt/kvm/api.rst | 52 ++++ Documentation/virt/kvm/arm/index.rst | 1 + Documentation/virt/kvm/arm/vcpu-features.rst | 48 +++ arch/arm64/include/asm/cpufeature.h | 14 + arch/arm64/include/asm/kvm_emulate.h | 3 +- arch/arm64/include/asm/kvm_host.h | 22 ++ arch/arm64/include/uapi/asm/kvm.h | 32 ++ arch/arm64/kernel/cpu_errata.c | 13 +- arch/arm64/kernel/proton-pack.c | 1 + arch/arm64/kvm/arm.c | 10 + arch/arm64/kvm/pmu-emul.c | 2 +- arch/arm64/kvm/sys_regs.c | 305 +++++++++++++++---- include/uapi/linux/kvm.h | 3 + 13 files changed, 446 insertions(+), 60 deletions(-) create mode 100644 Documentation/virt/kvm/arm/vcpu-features.rst -- 2.33.0

2 23

[PATCH OLK-6.6] smb: client: Fix netns refcount imbalance causing leaks and use-after-free
by Wang Zhaolong 25 Feb '25

25 Feb '25

maillist inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/IBNCLL CVE: NA Reference: https://lore.kernel.org/linux-cifs/CAH2r5muqxgPy=n4Aks_k-2goEZmmD10Dx3u3jgO… --------------------------------------------- Commit ef7134c7fc48 ("smb: client: Fix use-after-free of network namespace.") attempted to fix a netns use-after-free issue by manually adjusting reference counts via sk->sk_net_refcnt and sock_inuse_add(). However, a later commit e9f2517a3e18 ("smb: client: fix TCP timers deadlock after rmmod") pointed out that the approach of manually setting sk->sk_net_refcnt in the first commit was technically incorrect, as sk->sk_net_refcnt should only be set for user sockets. It led to issues like TCP timers not being cleared properly on close. The second commit moved to a model of just holding an extra netns reference for server->ssocket using get_net(), and dropping it when the server is torn down. But there remain some gaps in the get_net()/put_net() balancing added by these commits. The incomplete reference handling in these fixes results in two issues: 1. Netns refcount leaks[1] The problem process is as follows: ``` mount.cifs cifsd cifs_do_mount cifs_mount cifs_mount_get_session cifs_get_tcp_session get_net() /* First get net. */ ip_connect generic_ip_connect /* Try port 445 */ get_net() ->connect() /* Failed */ put_net() generic_ip_connect /* Try port 139 */ get_net() /* Missing matching put_net() for this get_net().*/ cifs_get_smb_ses cifs_negotiate_protocol smb2_negotiate SMB2_negotiate cifs_send_recv wait_for_response cifs_demultiplex_thread cifs_read_from_socket cifs_readv_from_socket cifs_reconnect cifs_abort_connection sock_release(); server->ssocket = NULL; /* Missing put_net() here. */ generic_ip_connect get_net() ->connect() /* Failed */ put_net() sock_release(); server->ssocket = NULL; free_rsp_buf ... clean_demultiplex_info /* It's only called once here. */ put_net() ``` When cifs_reconnect() is triggered, the server->ssocket is released without a corresponding put_net() for the reference acquired in generic_ip_connect() before. it ends up calling generic_ip_connect() again to retry get_net(). After that, server->ssocket is set to NULL in the error path of generic_ip_connect(), and the net count cannot be released in the final clean_demultiplex_info() function. 2. Potential use-after-free The current refcounting scheme can lead to a potential use-after-free issue in the following scenario: ``` cifs_do_mount cifs_mount cifs_mount_get_session cifs_get_tcp_session get_net() /* First get net */ ip_connect generic_ip_connect get_net() bind_socket kernel_bind /* failed */ put_net() /* after out_err_crypto_release label */ put_net() /* after out_err label */ put_net() ``` In the exception handling process where binding the socket fails, the get_net() and put_net() calls are unbalanced, which may cause the server->net reference count to drop to zero and be prematurely released. To address both issues, this patch ties the netns reference counting to the server->ssocket and server lifecycles. The extra reference is now acquired when the server or socket is created, and released when the socket is destroyed or the server is torn down. [1]: https://bugzilla.kernel.org/show_bug.cgi?id=219792 Fixes: ef7134c7fc48 ("smb: client: Fix use-after-free of network namespace.") Fixes: e9f2517a3e18 ("smb: client: fix TCP timers deadlock after rmmod") Signed-off-by: Wang Zhaolong <wangzhaolong1(a)huawei.com> --- fs/smb/client/connect.c | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/fs/smb/client/connect.c b/fs/smb/client/connect.c index 20f303f2a5d7..4a15e739a436 100644 --- a/fs/smb/client/connect.c +++ b/fs/smb/client/connect.c @@ -316,6 +316,7 @@ cifs_abort_connection(struct TCP_Server_Info *server) server->ssocket->flags); sock_release(server->ssocket); server->ssocket = NULL; + put_net(cifs_net_ns(server)); } server->sequence_number = 0; server->session_estab = false; @@ -3140,8 +3141,12 @@ generic_ip_connect(struct TCP_Server_Info *server) /* * Grab netns reference for the socket. * - * It'll be released here, on error, or in clean_demultiplex_info() upon server - * teardown. + * This reference will be released in several situations: + * - In the failure path before the cifsd thread is started. + * - In the all place where server->socket is released, it is + * also set to NULL. + * - Ultimately in clean_demultiplex_info(), during the final + * teardown. */ get_net(net); @@ -3157,10 +3162,8 @@ generic_ip_connect(struct TCP_Server_Info *server) } rc = bind_socket(server); - if (rc < 0) { - put_net(cifs_net_ns(server)); + if (rc < 0) return rc; - } /* * Eventually check for other socket options to change from @@ -3206,9 +3209,6 @@ generic_ip_connect(struct TCP_Server_Info *server) if (sport == htons(RFC1001_PORT)) rc = ip_rfc1001_connect(server); - if (rc < 0) - put_net(cifs_net_ns(server)); - return rc; } -- 2.34.3

2 1

[openeuler:openEuler-1.0-LTS 1424/1424] mm/page_alloc.c:1645:3: warning: cast from 'int (*)(unsigned long, unsigned long, struct deferred_args *)' to 'ktask_thread_func' (aka 'int (*)(void *, void *, void *)') converts to incompatible function type
by kernel test robot 25 Feb '25

25 Feb '25

Hi Daniel, FYI, the error/warning still remains. tree: https://gitee.com/openeuler/kernel.git openEuler-1.0-LTS head: 33c097cb0d6c8c1dcf8cde0f66b1c6c907ee2740 commit: eb761d6521c32c006a4987260394a61c6684fb35 [1424/1424] mm: parallelize deferred struct page initialization within each node config: x86_64-buildonly-randconfig-003-20250217 (https://download.01.org/0day-ci/archive/20250225/202502251247.78kEByd3-lkp@…) compiler: clang version 19.1.3 (https://github.com/llvm/llvm-project ab51eccf88f5321e7c60591c5546b254b6afab99) reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20250225/202502251247.78kEByd3-lkp@…) If you fix the issue in a separate patch/commit (i.e. not just a new version of the same patch/commit), kindly add following tags | Reported-by: kernel test robot <lkp(a)intel.com> | Closes: https://lore.kernel.org/oe-kbuild-all/202502251247.78kEByd3-lkp@intel.com/ All warnings (new ones prefixed by >>): In file included from mm/page_alloc.c:21: include/linux/pagemap.h:401:21: warning: cast from 'int (*)(struct file *, struct page *)' to 'filler_t *' (aka 'int (*)(void *, struct page *)') converts to incompatible function type [-Wcast-function-type-strict] 401 | filler_t *filler = (filler_t *)mapping->a_ops->readpage; | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ In file included from mm/page_alloc.c:58: include/linux/mm_inline.h:32:43: warning: arithmetic between different enumeration types ('enum node_stat_item' and 'enum lru_list') [-Wenum-enum-conversion] 32 | __mod_node_page_state(pgdat, NR_LRU_BASE + lru, nr_pages); | ~~~~~~~~~~~ ^ ~~~ include/linux/mm_inline.h:34:22: warning: arithmetic between different enumeration types ('enum zone_stat_item' and 'enum lru_list') [-Wenum-enum-conversion] 34 | NR_ZONE_LRU_BASE + lru, nr_pages); | ~~~~~~~~~~~~~~~~ ^ ~~~ mm/page_alloc.c:1412:6: warning: no previous prototype for function 'set_zone_contiguous' [-Wmissing-prototypes] 1412 | void set_zone_contiguous(struct zone *zone) | ^ mm/page_alloc.c:1412:1: note: declare 'static' if the function is not intended to be used outside of this translation unit 1412 | void set_zone_contiguous(struct zone *zone) | ^ | static mm/page_alloc.c:1433:6: warning: no previous prototype for function 'clear_zone_contiguous' [-Wmissing-prototypes] 1433 | void clear_zone_contiguous(struct zone *zone) | ^ mm/page_alloc.c:1433:1: note: declare 'static' if the function is not intended to be used outside of this translation unit 1433 | void clear_zone_contiguous(struct zone *zone) | ^ | static >> mm/page_alloc.c:1645:3: warning: cast from 'int (*)(unsigned long, unsigned long, struct deferred_args *)' to 'ktask_thread_func' (aka 'int (*)(void *, void *, void *)') converts to incompatible function type [-Wcast-function-type-strict] 1645 | DEFINE_KTASK_CTL(ctl, deferred_init_chunk, &args, | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1646 | KTASK_PTE_MINCHUNK); | ~~~~~~~~~~~~~~~~~~~ include/linux/ktask.h:139:3: note: expanded from macro 'DEFINE_KTASK_CTL' 139 | KTASK_CTL_INITIALIZER(thread_func, func_arg, min_chunk_size) \ | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ include/linux/ktask.h:123:21: note: expanded from macro 'KTASK_CTL_INITIALIZER' 123 | .kc_thread_func = (ktask_thread_func)(thread_func), \ | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ mm/page_alloc.c:1661:3: warning: cast from 'int (*)(unsigned long, unsigned long, struct deferred_args *)' to 'ktask_thread_func' (aka 'int (*)(void *, void *, void *)') converts to incompatible function type [-Wcast-function-type-strict] 1661 | DEFINE_KTASK_CTL(ctl, deferred_free_chunk, &args, | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1662 | KTASK_PTE_MINCHUNK); | ~~~~~~~~~~~~~~~~~~~ include/linux/ktask.h:139:3: note: expanded from macro 'DEFINE_KTASK_CTL' 139 | KTASK_CTL_INITIALIZER(thread_func, func_arg, min_chunk_size) \ | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ include/linux/ktask.h:123:21: note: expanded from macro 'KTASK_CTL_INITIALIZER' 123 | .kc_thread_func = (ktask_thread_func)(thread_func), \ | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ mm/page_alloc.c:3050:3: warning: arithmetic between different enumeration types ('enum vm_event_item' and 'enum zone_type') [-Wenum-enum-conversion] 3050 | __count_zid_vm_events(PGALLOC, page_zonenum(page), 1 << order); | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ include/linux/vmstat.h:126:34: note: expanded from macro '__count_zid_vm_events' 126 | __count_vm_events(item##_NORMAL - ZONE_NORMAL + zid, delta) | ~~~~~~~~~~~~~ ^ ~~~~~~~~~~~ mm/page_alloc.c:3098:2: warning: arithmetic between different enumeration types ('enum vm_event_item' and 'enum zone_type') [-Wenum-enum-conversion] 3098 | __count_zid_vm_events(PGALLOC, page_zonenum(page), 1 << order); | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ include/linux/vmstat.h:126:34: note: expanded from macro '__count_zid_vm_events' 126 | __count_vm_events(item##_NORMAL - ZONE_NORMAL + zid, delta) | ~~~~~~~~~~~~~ ^ ~~~~~~~~~~~ 9 warnings generated. vim +1645 mm/page_alloc.c 1587 1588 /* Initialise remaining memory on a node */ 1589 static int __init deferred_init_memmap(void *data) 1590 { 1591 pg_data_t *pgdat = data; 1592 int nid = pgdat->node_id; 1593 unsigned long start = jiffies; 1594 unsigned long nr_init = 0, nr_free = 0; 1595 unsigned long spfn, epfn, first_init_pfn, flags; 1596 phys_addr_t spa, epa; 1597 int zid; 1598 struct zone *zone; 1599 const struct cpumask *cpumask = cpumask_of_node(pgdat->node_id); 1600 u64 i; 1601 unsigned long nr_node_cpus; 1602 struct ktask_node kn; 1603 1604 /* Bind memory initialisation thread to a local node if possible */ 1605 if (!cpumask_empty(cpumask)) 1606 set_cpus_allowed_ptr(current, cpumask); 1607 1608 pgdat_resize_lock(pgdat, &flags); 1609 first_init_pfn = pgdat->first_deferred_pfn; 1610 if (first_init_pfn == ULONG_MAX) { 1611 pgdat_resize_unlock(pgdat, &flags); 1612 pgdat_init_report_one_done(); 1613 return 0; 1614 } 1615 1616 /* 1617 * We'd like to know the memory bandwidth of the chip to calculate the 1618 * most efficient number of threads to start, but we can't. In 1619 * testing, a good value for a variety of systems was a quarter of the 1620 * CPUs on the node. 1621 */ 1622 nr_node_cpus = DIV_ROUND_UP(cpumask_weight(cpumask), 4); 1623 1624 /* Sanity check boundaries */ 1625 BUG_ON(pgdat->first_deferred_pfn < pgdat->node_start_pfn); 1626 BUG_ON(pgdat->first_deferred_pfn > pgdat_end_pfn(pgdat)); 1627 pgdat->first_deferred_pfn = ULONG_MAX; 1628 1629 /* Only the highest zone is deferred so find it */ 1630 for (zid = 0; zid < MAX_NR_ZONES; zid++) { 1631 zone = pgdat->node_zones + zid; 1632 if (first_init_pfn < zone_end_pfn(zone)) 1633 break; 1634 } 1635 first_init_pfn = max(zone->zone_start_pfn, first_init_pfn); 1636 1637 /* 1638 * Initialize and free pages. We do it in two loops: first we initialize 1639 * struct page, than free to buddy allocator, because while we are 1640 * freeing pages we can access pages that are ahead (computing buddy 1641 * page in __free_one_page()). 1642 */ 1643 for_each_free_mem_range(i, nid, MEMBLOCK_NONE, &spa, &epa, NULL) { 1644 struct deferred_args args = { nid, zid, ATOMIC64_INIT(0) }; > 1645 DEFINE_KTASK_CTL(ctl, deferred_init_chunk, &args, 1646 KTASK_PTE_MINCHUNK); 1647 ktask_ctl_set_max_threads(&ctl, nr_node_cpus); 1648 1649 spfn = max_t(unsigned long, first_init_pfn, PFN_UP(spa)); 1650 epfn = min_t(unsigned long, zone_end_pfn(zone), PFN_DOWN(epa)); 1651 1652 kn.kn_start = (void *)spfn; 1653 kn.kn_task_size = (spfn < epfn) ? epfn - spfn : 0; 1654 kn.kn_nid = nid; 1655 (void) ktask_run_numa(&kn, 1, &ctl); 1656 1657 nr_init += atomic64_read(&args.nr_pages); 1658 } 1659 for_each_free_mem_range(i, nid, MEMBLOCK_NONE, &spa, &epa, NULL) { 1660 struct deferred_args args = { nid, zid, ATOMIC64_INIT(0) }; 1661 DEFINE_KTASK_CTL(ctl, deferred_free_chunk, &args, 1662 KTASK_PTE_MINCHUNK); 1663 ktask_ctl_set_max_threads(&ctl, nr_node_cpus); 1664 1665 spfn = max_t(unsigned long, first_init_pfn, PFN_UP(spa)); 1666 epfn = min_t(unsigned long, zone_end_pfn(zone), PFN_DOWN(epa)); 1667 1668 kn.kn_start = (void *)spfn; 1669 kn.kn_task_size = (spfn < epfn) ? epfn - spfn : 0; 1670 kn.kn_nid = nid; 1671 (void) ktask_run_numa(&kn, 1, &ctl); 1672 1673 nr_free += atomic64_read(&args.nr_pages); 1674 } 1675 pgdat_resize_unlock(pgdat, &flags); 1676 1677 /* Sanity check that the next zone really is unpopulated */ 1678 WARN_ON(++zid < MAX_NR_ZONES && populated_zone(++zone)); 1679 VM_BUG_ON(nr_init != nr_free); 1680 1681 zone->managed_pages += nr_free; 1682 1683 pr_info("node %d initialised, %lu pages in %ums\n", nid, nr_free, 1684 jiffies_to_msecs(jiffies - start)); 1685 1686 pgdat_init_report_one_done(); 1687 return 0; 1688 } 1689 -- 0-DAY CI Kernel Test Service https://github.com/intel/lkp-tests/wiki

1 0

[PATCH openEuler-1.0-LTS 0/2] smb: client: Fix netns refcount imbalance
by Wang Zhaolong 25 Feb '25

25 Feb '25

Enzo Matsumiya (1): smb: client: fix TCP timers deadlock after rmmod Wang Zhaolong (1): smb: client: Fix netns refcount imbalance causing leaks and use-after-free fs/cifs/connect.c | 33 ++++++++++++++++++++++++--------- 1 file changed, 24 insertions(+), 9 deletions(-) -- 2.39.2

2 3

[PATCH openEuler-25.03 v3 0/4] Add TrIO support in EROFS
by Hongbo Li 25 Feb '25

25 Feb '25

TrIO can accelerate the cold start of containers during on-demand loading. It aggregates the read I/O operations required for container runtime during the first container launch. In the following startups, TrIO pulls the necessary I/O data to the container node in a single large I/O operation and uses this I/O information to construct the runtime rootfs. By improving the efficiency of network I/O, TrIO speeds up container startup in on-demand loading scenarios. TrIO consists of both kernel-space and user-space code. The kernel-space code has been adapted at the overlayfs layer, introducing the CONFIG_EROFS_TRIO configuration to provide isolation. The user-space code requires adaptation by the user, and detailed usage methods are introduced in the tools/trio/README.md section. Patches 1~2 correspond to the kernel adaptations, while patches 3~4 are the scripts and best practices that TrIO relies on for its operation. v3: - Obtain and keep the trio_object in open. - Only close readahead in trio case. Hongbo Li (4): erofs:trio: Add trio_manager in erofs erofs: trio: Support TrIO feature in erofs TrIO: Add tools for using TrIO TrIO: Add README.md fs/erofs/Kconfig | 11 + fs/erofs/Makefile | 1 + fs/erofs/data.c | 17 + fs/erofs/fscache.c | 22 +- fs/erofs/internal.h | 50 ++ fs/erofs/super.c | 47 +- fs/erofs/trio_manager.c | 348 +++++++++++++ tools/trio/README.md | 507 +++++++++++++++++++ tools/trio/bpf/iotracker/Makefile | 99 ++++ tools/trio/bpf/iotracker/iotracker.bpf.c | 59 +++ tools/trio/bpf/iotracker/iotracker.c | 57 +++ tools/trio/bpf/rio_tracker_mod/Makefile | 9 + tools/trio/bpf/rio_tracker_mod/rio_tracker.c | 278 ++++++++++ tools/trio/scripts/trace_parser.py | 277 ++++++++++ 14 files changed, 1780 insertions(+), 2 deletions(-) create mode 100644 fs/erofs/trio_manager.c create mode 100644 tools/trio/README.md create mode 100644 tools/trio/bpf/iotracker/Makefile create mode 100644 tools/trio/bpf/iotracker/iotracker.bpf.c create mode 100644 tools/trio/bpf/iotracker/iotracker.c create mode 100644 tools/trio/bpf/rio_tracker_mod/Makefile create mode 100644 tools/trio/bpf/rio_tracker_mod/rio_tracker.c create mode 100644 tools/trio/scripts/trace_parser.py -- 2.34.1

2 5