- Kernel - mailweb.openeuler.org

[PATCH openEuler-22.03-LTS-SP2] bpf: Protect against int overflow for stack access size
by Pu Lehui 24 May '24

24 May '24

From: Andrei Matei <andreimatei1(a)gmail.com> stable inclusion from stable-v5.10.215 commit 9970e059af471478455f9534e8c3db82f8c5496d category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I9QG3A CVE: CVE-2024-35905 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- [ Upstream commit ecc6a2101840177e57c925c102d2d29f260d37c8 ] This patch re-introduces protection against the size of access to stack memory being negative; the access size can appear negative as a result of overflowing its signed int representation. This should not actually happen, as there are other protections along the way, but we should protect against it anyway. One code path was missing such protections (fixed in the previous patch in the series), causing out-of-bounds array accesses in check_stack_range_initialized(). This patch causes the verification of a program with such a non-sensical access size to fail. This check used to exist in a more indirect way, but was inadvertendly removed in a833a17aeac7. Fixes: a833a17aeac7 ("bpf: Fix verification of indirect var-off stack access") Reported-by: syzbot+33f4297b5f927648741a(a)syzkaller.appspotmail.com Reported-by: syzbot+aafd0513053a1cbf52ef(a)syzkaller.appspotmail.com Closes: https://lore.kernel.org/bpf/CAADnVQLORV5PT0iTAhRER+iLBTkByCYNBYyvBSgjN1T31K… Acked-by: Andrii Nakryiko <andrii(a)kernel.org> Signed-off-by: Andrei Matei <andreimatei1(a)gmail.com> Link: https://lore.kernel.org/r/20240327024245.318299-3-andreimatei1@gmail.com Signed-off-by: Alexei Starovoitov <ast(a)kernel.org> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Pu Lehui <pulehui(a)huawei.com> --- kernel/bpf/verifier.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 222aa218f826..03e9f319f77a 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -3725,6 +3725,11 @@ static int check_stack_access_within_bounds( err = check_stack_slot_within_bounds(env, min_off, state, type); if (!err && max_off > 0) err = -EINVAL; /* out of stack access into non-negative offsets */ + if (!err && access_size < 0) + /* access_size should not be negative (or overflow an int); others checks + * along the way should have prevented such an access. + */ + err = -EFAULT; /* invalid negative access size; integer overflow? */ if (err) { if (tnum_is_const(reg->var_off)) { -- 2.34.1

2 1

[PATCH openEuler-22.03-LTS] bpf: Protect against int overflow for stack access size
by Pu Lehui 24 May '24

24 May '24

From: Andrei Matei <andreimatei1(a)gmail.com> stable inclusion from stable-v5.10.215 commit 9970e059af471478455f9534e8c3db82f8c5496d category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I9QG3A CVE: CVE-2024-35905 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- [ Upstream commit ecc6a2101840177e57c925c102d2d29f260d37c8 ] This patch re-introduces protection against the size of access to stack memory being negative; the access size can appear negative as a result of overflowing its signed int representation. This should not actually happen, as there are other protections along the way, but we should protect against it anyway. One code path was missing such protections (fixed in the previous patch in the series), causing out-of-bounds array accesses in check_stack_range_initialized(). This patch causes the verification of a program with such a non-sensical access size to fail. This check used to exist in a more indirect way, but was inadvertendly removed in a833a17aeac7. Fixes: a833a17aeac7 ("bpf: Fix verification of indirect var-off stack access") Reported-by: syzbot+33f4297b5f927648741a(a)syzkaller.appspotmail.com Reported-by: syzbot+aafd0513053a1cbf52ef(a)syzkaller.appspotmail.com Closes: https://lore.kernel.org/bpf/CAADnVQLORV5PT0iTAhRER+iLBTkByCYNBYyvBSgjN1T31K… Acked-by: Andrii Nakryiko <andrii(a)kernel.org> Signed-off-by: Andrei Matei <andreimatei1(a)gmail.com> Link: https://lore.kernel.org/r/20240327024245.318299-3-andreimatei1@gmail.com Signed-off-by: Alexei Starovoitov <ast(a)kernel.org> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Pu Lehui <pulehui(a)huawei.com> --- kernel/bpf/verifier.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 222aa218f826..03e9f319f77a 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -3725,6 +3725,11 @@ static int check_stack_access_within_bounds( err = check_stack_slot_within_bounds(env, min_off, state, type); if (!err && max_off > 0) err = -EINVAL; /* out of stack access into non-negative offsets */ + if (!err && access_size < 0) + /* access_size should not be negative (or overflow an int); others checks + * along the way should have prevented such an access. + */ + err = -EFAULT; /* invalid negative access size; integer overflow? */ if (err) { if (tnum_is_const(reg->var_off)) { -- 2.34.1

2 1

[PATCH OLK-5.10] bpf: Protect against int overflow for stack access size
by Pu Lehui 24 May '24

24 May '24

From: Andrei Matei <andreimatei1(a)gmail.com> stable inclusion from stable-v5.10.215 commit 9970e059af471478455f9534e8c3db82f8c5496d category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I9QG3A CVE: CVE-2024-35905 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… -------------------------------- [ Upstream commit ecc6a2101840177e57c925c102d2d29f260d37c8 ] This patch re-introduces protection against the size of access to stack memory being negative; the access size can appear negative as a result of overflowing its signed int representation. This should not actually happen, as there are other protections along the way, but we should protect against it anyway. One code path was missing such protections (fixed in the previous patch in the series), causing out-of-bounds array accesses in check_stack_range_initialized(). This patch causes the verification of a program with such a non-sensical access size to fail. This check used to exist in a more indirect way, but was inadvertendly removed in a833a17aeac7. Fixes: a833a17aeac7 ("bpf: Fix verification of indirect var-off stack access") Reported-by: syzbot+33f4297b5f927648741a(a)syzkaller.appspotmail.com Reported-by: syzbot+aafd0513053a1cbf52ef(a)syzkaller.appspotmail.com Closes: https://lore.kernel.org/bpf/CAADnVQLORV5PT0iTAhRER+iLBTkByCYNBYyvBSgjN1T31K… Acked-by: Andrii Nakryiko <andrii(a)kernel.org> Signed-off-by: Andrei Matei <andreimatei1(a)gmail.com> Link: https://lore.kernel.org/r/20240327024245.318299-3-andreimatei1@gmail.com Signed-off-by: Alexei Starovoitov <ast(a)kernel.org> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Pu Lehui <pulehui(a)huawei.com> --- kernel/bpf/verifier.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 34292a48e59c..b5d219e0319b 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -3946,6 +3946,11 @@ static int check_stack_access_within_bounds( err = check_stack_slot_within_bounds(env, min_off, state, type); if (!err && max_off > 0) err = -EINVAL; /* out of stack access into non-negative offsets */ + if (!err && access_size < 0) + /* access_size should not be negative (or overflow an int); others checks + * along the way should have prevented such an access. + */ + err = -EFAULT; /* invalid negative access size; integer overflow? */ if (err) { if (tnum_is_const(reg->var_off)) { -- 2.34.1

2 1

[PATCH OLK-5.10 v5 0/4] arm64: Allow CPU0 to be nohz_full
by Xiongfeng Wang 24 May '24

24 May '24

Changelog: v1->v2: change __setup() to early_param() v2->v3: fix kabi because #include <linux/sched/isolation.h> modify defconfig v3->v4: fix riscv compile error: move support_cpu0_nohz_full to tick.h v4->v5: fix kabi change Xiongfeng Wang (4): isolation: Do not check whether housekeeping CPUs are present isolation: Check whether all the housekeeping CPUs are online arm64: Kconfig: select ARCH_SUSPEND_NONZERO_CPU by default arm64: defconfig: Enable CONFIG_ARCH_SUSPEND_NONZERO_CPU arch/arm64/Kconfig | 3 +++ arch/arm64/configs/openeuler_defconfig | 2 ++ include/linux/cpu.h | 5 ++++- include/linux/sched/isolation.h | 2 ++ include/linux/tick.h | 2 ++ kernel/sched/isolation.c | 31 ++++++++++++++++++++++---- kernel/smp.c | 4 ++++ kernel/time/tick-sched.c | 11 ++++++++- 8 files changed, 54 insertions(+), 6 deletions(-) -- 2.30.0

2 5

[PATCH OLK-5.10 0/3] fix CVE-2024-35839
by Zhengchao Shao 24 May '24

24 May '24

Fix CVE-2024-35839. Pavel Tikhomirov (3): netfilter: nfnetlink_log: use proper helper for fetching physinif netfilter: propagate net to nf_bridge_get_physindev netfilter: bridge: replace physindev with physinif in nf_bridge_info include/linux/netfilter_bridge.h | 6 ++-- include/linux/skbuff.h | 2 +- include/net/netfilter/nf_log.h | 2 +- net/bridge/br_netfilter_hooks.c | 41 +++++++++++++++++----- net/bridge/br_netfilter_ipv6.c | 14 +++++--- net/ipv4/netfilter/nf_log_arp.c | 2 +- net/ipv4/netfilter/nf_log_ipv4.c | 2 +- net/ipv4/netfilter/nf_reject_ipv4.c | 9 +++-- net/ipv6/netfilter/nf_log_ipv6.c | 2 +- net/ipv6/netfilter/nf_reject_ipv6.c | 11 ++++-- net/netfilter/ipset/ip_set_hash_netiface.c | 8 ++--- net/netfilter/nf_log_common.c | 5 +-- net/netfilter/nf_queue.c | 2 +- net/netfilter/nfnetlink_log.c | 8 ++--- net/netfilter/xt_physdev.c | 2 +- 15 files changed, 78 insertions(+), 38 deletions(-) -- 2.34.1

2 4

[PATCH OLK-5.10 0/1] Fix allmodconfig build error
by Shuai Wu 24 May '24

24 May '24

Fix allmodconfig build error in hiroce3 driver. Fixes: ae52f9c522dd ("infiniband/hw/hiroce3: Add Huawei Intelligent Network Card RDMA Driver") Shuai Wu (1): net/ethernet/huawei/hiroce3: Fix allmodconfig build error .../hw/hiroce3/bond/roce_bond_common.c | 38 +- .../hw/hiroce3/cq/roce_cq_destroy.c | 15 +- .../hw/hiroce3/host/hmm/hmm_comp_mtt.c | 1 + .../hw/hiroce3/host/hmm/hmm_comp_mw_mr.c | 1 + .../hw/hiroce3/host/hmm/hmm_comp_res.c | 1 + .../infiniband/hw/hiroce3/host/hmm/hmm_mr.c | 98 +---- .../infiniband/hw/hiroce3/host/hmm/hmm_umem.c | 4 +- .../infiniband/hw/hiroce3/host/hmm/hmm_umem.h | 2 +- .../include/rdma/rdma_context_format.h | 4 +- .../include/rdma/rdma_ext_ctx_format.h | 5 +- .../hw/hiroce3/include/rdma/roce_ccf_format.h | 4 +- .../hw/hiroce3/include/rdma/roce_verbs_attr.h | 4 +- .../include/rdma/roce_verbs_attr_qpc_chip.h | 4 +- .../hiroce3/include/rdma/roce_verbs_cq_attr.h | 4 +- .../include/rdma/roce_verbs_ext_attr.h | 4 +- .../include/rdma/roce_verbs_gid_attr.h | 4 +- .../hiroce3/include/rdma/roce_verbs_mr_attr.h | 4 +- .../hw/hiroce3/include/rdma/roce_verbs_pub.h | 4 +- .../include/rdma/roce_verbs_srq_attr.h | 4 +- drivers/infiniband/hw/hiroce3/mr/roce_mr.c | 37 -- drivers/infiniband/hw/hiroce3/roce.h | 1 - drivers/infiniband/hw/hiroce3/roce_main.c | 7 +- .../ethernet/huawei/hinic3/bond/hinic3_bond.c | 1 - .../ethernet/huawei/hinic3/bond/hinic3_bond.h | 4 +- .../ethernet/huawei/hinic3/cqm/cqm_bat_cla.c | 198 +++------ .../huawei/hinic3/cqm/cqm_bitmap_table.c | 10 +- .../huawei/hinic3/cqm/cqm_bloomfilter.c | 69 +--- .../net/ethernet/huawei/hinic3/cqm/cqm_cmd.c | 139 ++----- .../net/ethernet/huawei/hinic3/cqm/cqm_cmd.h | 2 - .../net/ethernet/huawei/hinic3/cqm/cqm_db.c | 192 +++------ .../ethernet/huawei/hinic3/cqm/cqm_define.h | 4 - .../net/ethernet/huawei/hinic3/cqm/cqm_main.c | 178 +++------ .../ethernet/huawei/hinic3/cqm/cqm_object.c | 378 +++++------------- .../huawei/hinic3/cqm/cqm_object_intern.c | 194 +++------ .../ethernet/huawei/hinic3/hw/hinic3_hw_cfg.c | 2 +- .../ethernet/huawei/hinic3/hw/hinic3_hwif.c | 38 -- .../huawei/hinic3/hw/hinic3_multi_host_mgmt.c | 56 ++- .../hinic3/include/bond/bond_common_defs.h | 4 +- .../huawei/hinic3/include/hinic3_common.h | 2 - .../huawei/hinic3/include/hinic3_cqm.h | 11 - .../huawei/hinic3/include/hinic3_cqm_define.h | 4 - .../include/mpu/mpu_outband_ncsi_cmd_defs.h | 4 +- 42 files changed, 474 insertions(+), 1266 deletions(-) -- 2.33.0

2 2

[PATCH OLK-5.10 v4 0/4] arm64: Allow CPU0 to be nohz_full
by Xiongfeng Wang 24 May '24

24 May '24

Changelog: v1->v2: change __setup() to early_param() v2->v3: fix kabi because #include <linux/sched/isolation.h> modify defconfig v3->v4: fix riscv compile error: move support_cpu0_nohz_full to tick.h Xiongfeng Wang (4): isolation: Do not check whether housekeeping CPUs are present isolation: Check whether all the housekeeping CPUs are online arm64: Kconfig: select ARCH_SUSPEND_NONZERO_CPU by default arm64: defconfig: Enable CONFIG_ARCH_SUSPEND_NONZERO_CPU arch/arm64/Kconfig | 3 +++ arch/arm64/configs/openeuler_defconfig | 2 ++ include/linux/cpu.h | 3 ++- include/linux/sched/isolation.h | 2 ++ include/linux/tick.h | 2 ++ kernel/sched/isolation.c | 31 ++++++++++++++++++++++---- kernel/smp.c | 4 ++++ kernel/time/tick-sched.c | 11 ++++++++- 8 files changed, 52 insertions(+), 6 deletions(-) -- 2.30.0

1 4

[PATCH OLK-5.10 0/2] sched: QOS_SCHED_DYNAMIC_AFFINITY depend on FAIR_GROUP_SCHED
by Yipeng Zou 24 May '24

24 May '24

It's typo error, should be FAIR_GROUP_SCHED but not FAIR_CGROUP_SCHED. Yipeng Zou (2): Revert "sched: QOS_SCHED_DYNAMIC_AFFINITY depend on FAIR_CGROUP_SCHED" sched: QOS_SCHED_DYNAMIC_AFFINITY depend on FAIR_GROUP_SCHED init/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) -- 2.34.1

2 3

[PATCH OLK-5.10 v4 0/4] arm64: Allow CPU0 to be nohz_full
by Xiongfeng Wang 24 May '24

24 May '24

Changelog: v1->v2: change __setup() to early_param() v2->v3: fix kabi because #include <linux/sched/isolation.h> modify defconfig v3->v4: fix riscv compile error: move support_cpu0_nohz_full to tick.h Xiongfeng Wang (3): isolation: Check whether all the housekeeping CPUs are online arm64: Kconfig: select ARCH_SUSPEND_NONZERO_CPU by default arm64: defconfig: Enable CONFIG_ARCH_SUSPEND_NONZERO_CPU arch/arm64/Kconfig | 3 +++ arch/arm64/configs/openeuler_defconfig | 2 ++ include/linux/cpu.h | 3 ++- include/linux/sched/isolation.h | 2 ++ kernel/sched/isolation.c | 13 +++++++++++++ kernel/smp.c | 4 ++++ kernel/time/tick-sched.c | 2 +- 7 files changed, 27 insertions(+), 2 deletions(-) -- 2.30.0

2 4

[PATCH OLK-5.10 0/2] sched: QOS_SCHED_DYNAMIC_AFFINITY depend on FAIR_GROUP_SCHED
by Yipeng Zou 24 May '24

24 May '24

It's typo error, should be FAIR_GROUP_SCHED but not FAIR_CGROUP_SCHED. Yipeng Zou (2): Revert "sched: QOS_SCHED_DYNAMIC_AFFINITY depend on FAIR_CGROUP_SCHED" sched: QOS_SCHED_DYNAMIC_AFFINITY depend on FAIR_GROUP_SCHED init/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) -- 2.34.1

2 4