- Kernel - mailweb.openeuler.org

[PATCH OLK-6.6 0/2] mainline bpf for olk-6.6
by Pu Lehui 27 Jun '26

27 Jun '26

Yazhou Tang (2): bpf: Fix out-of-bounds read in bpf_patch_call_args() bpf: Fix s16 truncation for large bpf-to-bpf call offsets include/linux/bpf.h | 8 +++++++- include/linux/filter.h | 3 --- kernel/bpf/core.c | 27 +++++++++++++++++++-------- kernel/bpf/syscall.c | 26 ++++++++++++++++++++++++++ kernel/bpf/verifier.c | 13 +++++++++---- 5 files changed, 61 insertions(+), 16 deletions(-) -- 2.34.1

2 3

[PATCH OLK-5.10 0/2] fix null pointer in bpf_get_local_storage
by Pu Lehui 27 Jun '26

27 Jun '26

Ilya Leoshkevich (1): bpf: Check for helper calls in check_subprogs() Leon Hwang (1): bpf, verifier: Correct tail_call_reachable for bpf prog kernel/bpf/verifier.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) -- 2.34.1

2 3

[PATCH OLK-6.6 0/4] Fix CVE-2026-52975: bonding 3ad RCU data-race
by superdcc97＠163.com 26 Jun '26

26 Jun '26

From: Dong Chenchen <dongchenchen2(a)huawei.com> This series fixes CVE-2026-52975, a data-race in bonding 3ad code (bond_3ad_get_active_agg_info / bond_3ad_state_machine_handler) reported by KCSAN. The fix adds __rcu qualifier to port->aggregator and proper RCU API usage. Patches 1-3 are Stable-dep-of prerequisites that add churn state netlink support, per-port LACP actor priority, and broadcast_neighbor option. Patch 4 is the main CVE fix implementing proper RCU rules for port->aggregator. Note: Patch 3 (broadcast_neighbor) has a Class B forced-minimal adaptation conflict - the target already has bond_up_slave array style in bond_xmit_broadcast, while the original patch expected bond_for_each_slave_rcu style. CVE core logic (all_slaves parameter, bond_should_broadcast_neighbor) was safely applied on target's existing control flow. Also, OLK-specific sysctl_bond_broadcast_arp_or_nd requires signature update at both bond_xmit_broadcast call sites. Dong Chenchen (3): bonding: print churn state via netlink bonding: add support for per-port LACP actor priority net: bonding: add broadcast_neighbor option for 802.3ad Eric Dumazet (1): bonding: 3ad: implement proper RCU rules for port->aggregator Documentation/networking/bonding.rst | 15 ++++ drivers/net/bonding/bond_3ad.c | 113 ++++++++++++++----------- drivers/net/bonding/bond_main.c | 73 +++++++++++++--- drivers/net/bonding/bond_netlink.c | 37 +++++++- drivers/net/bonding/bond_options.c | 78 +++++++++++++++++ drivers/net/bonding/bond_procfs.c | 3 +- drivers/net/bonding/bond_sysfs_slave.c | 17 ++-- include/net/bond_3ad.h | 3 +- include/net/bond_options.h | 2 + include/net/bonding.h | 3 + include/uapi/linux/if_link.h | 3 + 11 files changed, 276 insertions(+), 71 deletions(-) -- 2.43.0

2 5

[PATCH OLK-6.6] perf/amd/ibs: Avoid calling perf_allow_kernel() from the IBS NMI handler
by Luo Gengkun 26 Jun '26

26 Jun '26

From: Ravi Bangoria <ravi.bangoria(a)amd.com> mainline inclusion from mainline-v7.1-rc1 commit b0a09142622a994c4f4088c3f61db5da87cfc711 category: bugfix bugzilla: https://atomgit.com/src-openeuler/kernel/issues/15810 CVE: CVE-2026-53114 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… ---------------------------------------------------------------------- Calling perf_allow_kernel() from the NMI context is unsafe and could be fatal. Capture the permission at event-initialization time by storing it in event->hw.flags, and have the NMI handler rely on that cached flag instead of making the call directly. Fixes: 50a53b60e141 ("perf/amd/ibs: Prevent leaking sensitive data to userspace") Reported-by: Sadasivan Shaiju <sadasivan.shaiju2(a)amd.com> Signed-off-by: Ravi Bangoria <ravi.bangoria(a)amd.com> Signed-off-by: Peter Zijlstra (Intel) <peterz(a)infradead.org> Acked-by: Namhyung Kim <namhyung(a)kernel.org> Link: https://patch.msgid.link/20260216042216.1440-5-ravi.bangoria@amd.com Conflicts: arch/x86/events/amd/ibs.c [Fix conflict because commit 9ec84f79c5a7 is not merged.] Signed-off-by: Luo Gengkun <luogengkun2(a)huawei.com> --- arch/x86/events/amd/ibs.c | 5 ++++- arch/x86/events/perf_event_flags.h | 1 + 2 files changed, 5 insertions(+), 1 deletion(-) diff --git a/arch/x86/events/amd/ibs.c b/arch/x86/events/amd/ibs.c index c775047a1bda..74b0c608af5d 100644 --- a/arch/x86/events/amd/ibs.c +++ b/arch/x86/events/amd/ibs.c @@ -312,6 +312,9 @@ static int perf_ibs_init(struct perf_event *event) if (ret) return ret; + if (perf_allow_kernel(&event->attr)) + hwc->flags |= PERF_X86_EVENT_UNPRIVILEGED; + if (hwc->sample_period) { if (config & perf_ibs->cnt_mask) /* raw max_cnt may not be set */ @@ -1343,7 +1346,7 @@ static int perf_ibs_handle_irq(struct perf_ibs *perf_ibs, struct pt_regs *iregs) * unprivileged users. */ if ((event->attr.sample_type & PERF_SAMPLE_RAW) && - perf_allow_kernel(&event->attr)) { + (hwc->flags & PERF_X86_EVENT_UNPRIVILEGED)) { perf_ibs_phyaddr_clear(perf_ibs, &ibs_data); } diff --git a/arch/x86/events/perf_event_flags.h b/arch/x86/events/perf_event_flags.h index 70078334e4a3..47f84ee8f540 100644 --- a/arch/x86/events/perf_event_flags.h +++ b/arch/x86/events/perf_event_flags.h @@ -23,3 +23,4 @@ PERF_ARCH(PEBS_LAT_HYBRID, 0x0020000) /* ld and st lat for hybrid */ PERF_ARCH(NEEDS_BRANCH_STACK, 0x0040000) /* require branch stack setup */ PERF_ARCH(BRANCH_COUNTERS, 0x0080000) /* logs the counters in the extra space of each branch */ PERF_ARCH(ACR, 0x0100000) /* Auto counter reload */ +PERF_ARCH(UNPRIVILEGED, 0x0200000) /* Unprivileged event (wrt perf_allow_kernel()) */ -- 2.34.1

2 1

[PATCH OLK-6.6] fuse: limit FUSE_NOTIFY_RETRIEVE to uptodate folios
by Zizhi Wo 26 Jun '26

26 Jun '26

From: Jann Horn <jannh(a)google.com> stable inclusion from stable-v6.18.36 commit 56763afa013444a9d84ca1b74e4b7130942177ba category: bugfix bugzilla: https://atomgit.com/src-openeuler/kernel/issues/15852 Reference: https://web.git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit… -------------------------------- commit 4e3d1b2c48ca6c55f1e9ca7f8dccc76f120f276c upstream. FUSE_NOTIFY_RETRIEVE must be limited to uptodate folios; !uptodate folios can contain uninitialized data. Since FUSE_NOTIFY_RETRIEVE is intended to only return data that is already in the page cache and not wait for data from the FUSE daemon, treat !uptodate folios as if they weren't present. This only has security impact on systems that don't enable automatic zero-initialization of all page allocations via CONFIG_INIT_ON_ALLOC_DEFAULT_ON or init_on_alloc=1. Cc: stable(a)kernel.org Fixes: 2d45ba381a74 ("fuse: add retrieve request") Signed-off-by: Jann Horn <jannh(a)google.com> Link: https://patch.msgid.link/20260519-fuse-retrieve-uptodate-v1-1-a7a1912a37f9@… Acked-by: Miklos Szeredi <mszeredi(a)redhat.com> Signed-off-by: Christian Brauner (Amutable) <brauner(a)kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Conflicts: fs/fuse/dev.c [page -> folio] Signed-off-by: Zizhi Wo <wozizhi(a)huawei.com> --- fs/fuse/dev.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index 7718e2504797..35b18e4c1562 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -1802,10 +1802,14 @@ static int fuse_retrieve(struct fuse_mount *fm, struct inode *inode, unsigned int this_num; page = find_get_page(mapping, index); if (!page) break; + if (!PageUptodate(page)) { + put_page(page); + break; + } this_num = min_t(unsigned, num, PAGE_SIZE - offset); ap->pages[ap->num_pages] = page; ap->descs[ap->num_pages].offset = offset; ap->descs[ap->num_pages].length = this_num; -- 2.52.0

2 1

[PATCH OLK-5.10] smb: client: require a full NFS mode SID before reading mode bits
by Long Li 26 Jun '26

26 Jun '26

From: Michael Bommarito <michael.bommarito(a)gmail.com> mainline inclusion from mainline-v7.0 commit 2757ad3e4b6f9e0fed4c7739594e702abc5cab21 category: bugfix bugzilla: https://atomgit.com/src-openeuler/kernel/issues/14908 CVE: CVE-2026-43350 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- parse_dacl() treats an ACE SID matching sid_unix_NFS_mode as an NFS mode SID and reads sid.sub_auth[2] to recover the mode bits. That assumes the ACE carries three subauthorities, but compare_sids() only compares min(a, b) subauthorities. A malicious server can return an ACE with num_subauth = 2 and sub_auth[] = {88, 3}, which still matches sid_unix_NFS_mode and then drives the sub_auth[2] read four bytes past the end of the ACE. Require num_subauth >= 3 before treating the ACE as an NFS mode SID. This keeps the fix local to the special-SID mode path without changing compare_sids() semantics for the rest of cifsacl. Fixes: e2f8fbfb8d09 ("cifs: get mode bits from special sid on stat") Cc: stable(a)vger.kernel.org Assisted-by: Claude:claude-opus-4-6 Signed-off-by: Michael Bommarito <michael.bommarito(a)gmail.com> Signed-off-by: Steve French <stfrench(a)microsoft.com> Conflicts: fs/cifs/cifsacl.c fs/smb/client/cifsacl.c [context conflicts] Signed-off-by: Long Li <leo.lilong(a)huawei.com> --- fs/cifs/cifsacl.c | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/cifs/cifsacl.c b/fs/cifs/cifsacl.c index ef4784e72b1d..472a110158ec 100644 --- a/fs/cifs/cifsacl.c +++ b/fs/cifs/cifsacl.c @@ -757,6 +757,7 @@ static void parse_dacl(struct cifs_acl *pdacl, char *end_of_acl, dump_ace(ppace[i], end_of_acl); #endif if (mode_from_special_sid && + ppace[i]->sid.num_subauth >= 3 && (compare_sids(&(ppace[i]->sid), &sid_unix_NFS_mode) == 0)) { /* -- 2.52.0

2 1

[PATCH OLK-6.6] tracing: Do not call map->ops->elt_free() if elt_alloc() fails
by Tengda Wu 25 Jun '26

25 Jun '26

From: "Masami Hiramatsu (Google)" <mhiramat(a)kernel.org> stable inclusion from stable-v6.6.142 commit f383cff9fb382139980bac1bcd3f3f5d59f68435 category: bugfix bugzilla: https://atomgit.com/openeuler/kernel/issues/9478 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- commit 8f0f5c4fb9df0e19a341e0c6ed8dc4fda9124f03 upstream. In paths where tracing_map_elt_alloc() failed to allocate objects, the map->ops->elt_alloc() call was never successful. In this case, map->ops->elt_free() should not be called. Link: https://sashiko.dev/#/patchset/20260520223101.34710-1-rosenp%40gmail.com Cc: stable(a)vger.kernel.org Cc: Tom Zanussi <tom.zanussi(a)linux.intel.com> Cc: Mathieu Desnoyers <mathieu.desnoyers(a)efficios.com> Cc: Rosen Penev <rosenp(a)gmail.com> Reported-by: Sashiko <sashiko-bot(a)kernel.org> Fixes: 2734b629525a ("tracing: Add per-element variable support to tracing_map") Link: https://patch.msgid.link/177933895460.108746.5396070821443932634.stgit@devn… Signed-off-by: Masami Hiramatsu (Google) <mhiramat(a)kernel.org> Signed-off-by: Steven Rostedt <rostedt(a)goodmis.org> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Signed-off-by: Tengda Wu <wutengda2(a)huawei.com> --- kernel/trace/tracing_map.c | 17 +++++++++++++---- 1 file changed, 13 insertions(+), 4 deletions(-) diff --git a/kernel/trace/tracing_map.c b/kernel/trace/tracing_map.c index 1921ade45be3..0e5b3ea49936 100644 --- a/kernel/trace/tracing_map.c +++ b/kernel/trace/tracing_map.c @@ -386,13 +386,11 @@ static void tracing_map_elt_init_fields(struct tracing_map_elt *elt) } } -static void tracing_map_elt_free(struct tracing_map_elt *elt) +static void __tracing_map_elt_free(struct tracing_map_elt *elt) { if (!elt) return; - if (elt->map->ops && elt->map->ops->elt_free) - elt->map->ops->elt_free(elt); kfree(elt->fields); kfree(elt->vars); kfree(elt->var_set); @@ -400,6 +398,17 @@ static void tracing_map_elt_free(struct tracing_map_elt *elt) kfree(elt); } +static void tracing_map_elt_free(struct tracing_map_elt *elt) +{ + if (!elt) + return; + + /* Only objects initialized with alloc_elt() should be passed to free_elt().*/ + if (elt->map->ops && elt->map->ops->elt_free) + elt->map->ops->elt_free(elt); + __tracing_map_elt_free(elt); +} + static struct tracing_map_elt *tracing_map_elt_alloc(struct tracing_map *map) { struct tracing_map_elt *elt; @@ -444,7 +453,7 @@ static struct tracing_map_elt *tracing_map_elt_alloc(struct tracing_map *map) } return elt; free: - tracing_map_elt_free(elt); + __tracing_map_elt_free(elt); return ERR_PTR(err); } -- 2.34.1

2 1

[PATCH OLK-5.10] bpf: Tighten cgroup storage cookie checks for prog arrays
by Pu Lehui 25 Jun '26

25 Jun '26

From: Daniel Borkmann <daniel(a)iogearbox.net> mainline inclusion from mainline-v7.2-rc1 commit 10627ddc0167aab5c1c390a10ef461e9937aba08 category: bugfix bugzilla: https://atomgit.com/openeuler/kernel/issues/9479 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- The fix in commit abad3d0bad72 ("bpf: Fix oob access in cgroup local storage") is still incomplete. The prog-array compatibility check treats a program with no cgroup storage as compatible with any stored storage cookie. This allows a storage-less program to bridge a tail call chain between an entry program and a storage-using callee even though cgroup local storage at runtime still follows the caller's context, that is, A -> B(no storage) -> C(storage) path. Requiring exact cookie equality would break the legitimate case of a storage-less leaf program being tail called from a storage-using one. Instead, only accept a zero storage cookie if the program cannot perform tail calls itself. This keeps A -> B(no storage) working while rejecting the A -> B(no storage) -> C(storage) bridge. Fixes: abad3d0bad72 ("bpf: Fix oob access in cgroup local storage") Reported-by: Lin Ma <malin89(a)huawei.com> Signed-off-by: Daniel Borkmann <daniel(a)iogearbox.net> Acked-by: Yonghong Song <yonghong.song(a)linux.dev> Link: https://lore.kernel.org/r/20260610105539.705887-1-daniel@iogearbox.net Signed-off-by: Alexei Starovoitov <ast(a)kernel.org> Conflicts: kernel/bpf/core.c [ctx conflicts] Signed-off-by: Pu Lehui <pulehui(a)huawei.com> --- kernel/bpf/core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c index c2f5ff16b742..ba56d4c55bc2 100644 --- a/kernel/bpf/core.c +++ b/kernel/bpf/core.c @@ -1814,7 +1814,7 @@ bool bpf_prog_map_compatible(struct bpf_map *map, cookie = aux->cgroup_storage[i] ? aux->cgroup_storage[i]->cookie : 0; ret = map->owner->storage_cookie[i] == cookie || - !cookie; + (!cookie && !aux->tail_call_reachable); } if (ret && map->owner->attach_func_proto != aux->attach_func_proto) { -- 2.34.1

2 1

[PATCH OLK-6.6] bpf: Tighten cgroup storage cookie checks for prog arrays
by Pu Lehui 25 Jun '26

25 Jun '26

From: Daniel Borkmann <daniel(a)iogearbox.net> mainline inclusion from mainline-v7.2-rc1 commit 10627ddc0167aab5c1c390a10ef461e9937aba08 category: bugfix bugzilla: https://atomgit.com/openeuler/kernel/issues/9479 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?… -------------------------------- The fix in commit abad3d0bad72 ("bpf: Fix oob access in cgroup local storage") is still incomplete. The prog-array compatibility check treats a program with no cgroup storage as compatible with any stored storage cookie. This allows a storage-less program to bridge a tail call chain between an entry program and a storage-using callee even though cgroup local storage at runtime still follows the caller's context, that is, A -> B(no storage) -> C(storage) path. Requiring exact cookie equality would break the legitimate case of a storage-less leaf program being tail called from a storage-using one. Instead, only accept a zero storage cookie if the program cannot perform tail calls itself. This keeps A -> B(no storage) working while rejecting the A -> B(no storage) -> C(storage) bridge. Fixes: abad3d0bad72 ("bpf: Fix oob access in cgroup local storage") Reported-by: Lin Ma <malin89(a)huawei.com> Signed-off-by: Daniel Borkmann <daniel(a)iogearbox.net> Acked-by: Yonghong Song <yonghong.song(a)linux.dev> Link: https://lore.kernel.org/r/20260610105539.705887-1-daniel@iogearbox.net Signed-off-by: Alexei Starovoitov <ast(a)kernel.org> Signed-off-by: Pu Lehui <pulehui(a)huawei.com> --- kernel/bpf/core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c index ff20567bbb78..f1dfff8dab07 100644 --- a/kernel/bpf/core.c +++ b/kernel/bpf/core.c @@ -2368,7 +2368,7 @@ bool bpf_prog_map_compatible(struct bpf_map *map, cookie = aux->cgroup_storage[i] ? aux->cgroup_storage[i]->cookie : 0; ret = map->owner->storage_cookie[i] == cookie || - !cookie; + (!cookie && !aux->tail_call_reachable); } if (ret && map->owner->attach_func_proto != aux->attach_func_proto) { -- 2.34.1

2 1

[PATCH OLK-6.6] fsnotify: fix inode reference leak in fsnotify_recalc_mask()
by Zizhi Wo 25 Jun '26

25 Jun '26

From: Amir Goldstein <amir73il(a)gmail.com> stable inclusion from stable-v6.12.91 commit 8c8afa6444e6bdc145d2bf2f3aeeca6da3e36b42 category: bugfix bugzilla: https://atomgit.com/src-openeuler/kernel/issues/15692 Reference: https://web.git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit… -------------------------------- [ Upstream commit 4aca914ac152f5d055ddcb36704d1e539ac08977 ] fsnotify_recalc_mask() fails to handle the return value of __fsnotify_recalc_mask(), which may return an inode pointer that needs to be released via fsnotify_drop_object() when the connector's HAS_IREF flag transitions from set to cleared. This manifests as a hung task with the following call trace: INFO: task umount:1234 blocked for more than 120 seconds. Call Trace: __schedule schedule fsnotify_sb_delete generic_shutdown_super kill_anon_super cleanup_mnt task_work_run do_exit do_group_exit The race window that triggers the iref leak: Thread A (adding mark) Thread B (removing mark) ────────────────────── ──────────────────────── fsnotify_add_mark_locked(): fsnotify_add_mark_list(): spin_lock(conn->lock) add mark_B(evictable) to list spin_unlock(conn->lock) return /* ---- gap: no lock held ---- */ fsnotify_detach_mark(mark_A): spin_lock(mark_A->lock) clear ATTACHED flag on mark_A spin_unlock(mark_A->lock) fsnotify_put_mark(mark_A) fsnotify_recalc_mask(): spin_lock(conn->lock) __fsnotify_recalc_mask(): /* mark_A skipped: ATTACHED cleared */ /* only mark_B(evictable) remains */ want_iref = false has_iref = true /* not yet cleared */ -> HAS_IREF transitions true -> false -> returns inode pointer spin_unlock(conn->lock) /* BUG: return value discarded! * iput() and fsnotify_put_sb_watched_objects() * are never called */ Fix this by deferring the transition true -> false of HAS_IREF flag from fsnotify_recalc_mask() (Thread A) to fsnotify_put_mark() (thread B). Fixes: c3638b5b1374 ("fsnotify: allow adding an inode mark without pinning inode") Signed-off-by: Xin Yin <yinxin.x(a)bytedance.com> Signed-off-by: Amir Goldstein <amir73il(a)gmail.com> Link: https://patch.msgid.link/CAOQ4uxiPsbHb0o5voUKyPFMvBsDkG914FYDcs4C5UpBMNm0Vc… Signed-off-by: Jan Kara <jack(a)suse.cz> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Conflicts: fs/notify/mark.c [Commit 35ceae44742e ("fsnotify: Avoid data race between fsnotify_recalc_mask() and fsnotify_object_watched()") has not mergfed, not affect to this patch.] Signed-off-by: Zizhi Wo <wozizhi(a)huawei.com> --- fs/notify/mark.c | 39 ++++++++++++++++++++++++++++++++++++--- 1 file changed, 36 insertions(+), 3 deletions(-) diff --git a/fs/notify/mark.c b/fs/notify/mark.c index b419a5ccf192..951d7de29f82 100644 --- a/fs/notify/mark.c +++ b/fs/notify/mark.c @@ -151,11 +151,16 @@ static struct inode *fsnotify_update_iref(struct fsnotify_mark_connector *conn, } return inode; } -static void *__fsnotify_recalc_mask(struct fsnotify_mark_connector *conn) +/* + * Calculate mask of events for a list of marks. + * + * Return true if any of the attached marks want to hold an inode reference. + */ +static bool __fsnotify_recalc_mask(struct fsnotify_mark_connector *conn) { u32 new_mask = 0; bool want_iref = false; struct fsnotify_mark *mark; @@ -171,10 +176,38 @@ static void *__fsnotify_recalc_mask(struct fsnotify_mark_connector *conn) !(mark->flags & FSNOTIFY_MARK_FLAG_NO_IREF)) want_iref = true; } *fsnotify_conn_mask_p(conn) = new_mask; + return want_iref; +} + +/* + * Calculate mask of events for a list of marks after attach/modify mark + * and get an inode reference for the connector if needed. + * + * A concurrent add of evictable mark and detach of non-evictable mark can + * lead to __fsnotify_recalc_mask() returning false want_iref, but in this + * case we defer clearing iref to fsnotify_recalc_mask_clear_iref() called + * from fsnotify_put_mark(). + */ +static void fsnotify_recalc_mask_set_iref(struct fsnotify_mark_connector *conn) +{ + bool has_iref = conn->flags & FSNOTIFY_CONN_FLAG_HAS_IREF; + bool want_iref = __fsnotify_recalc_mask(conn) || has_iref; + + (void) fsnotify_update_iref(conn, want_iref); +} + +/* + * Calculate mask of events for a list of marks after detach mark + * and return the inode object if its reference is no longer needed. + */ +static void *fsnotify_recalc_mask_clear_iref(struct fsnotify_mark_connector *conn) +{ + bool want_iref = __fsnotify_recalc_mask(conn); + return fsnotify_update_iref(conn, want_iref); } static bool fsnotify_conn_watches_children( struct fsnotify_mark_connector *conn) @@ -207,11 +240,11 @@ void fsnotify_recalc_mask(struct fsnotify_mark_connector *conn) if (!conn) return; spin_lock(&conn->lock); update_children = !fsnotify_conn_watches_children(conn); - __fsnotify_recalc_mask(conn); + fsnotify_recalc_mask_set_iref(conn); update_children &= fsnotify_conn_watches_children(conn); spin_unlock(&conn->lock); /* * Set children's PARENT_WATCHED flags only if parent started watching. * When parent stops watching, we clear false positive PARENT_WATCHED @@ -340,11 +373,11 @@ void fsnotify_put_mark(struct fsnotify_mark *mark) hlist_del_init_rcu(&mark->obj_list); if (hlist_empty(&conn->list)) { objp = fsnotify_detach_connector_from_object(conn, &type); free_conn = true; } else { - objp = __fsnotify_recalc_mask(conn); + objp = fsnotify_recalc_mask_clear_iref(conn); type = conn->type; } WRITE_ONCE(mark->connector, NULL); spin_unlock(&conn->lock); -- 2.52.0

2 1