July 2024 - Kernel - mailweb.openeuler.org

[PATCH openEuler-1.0-LTS] netrom: Fix a memory leak in nr_heartbeat_expiry()
by Liu Jian 18 Jul '24

18 Jul '24

From: Gavrilov Ilia <Ilia.Gavrilov(a)infotecs.ru> stable inclusion from stable-v4.19.317 commit d616876256b38ecf9a1a1c7d674192c5346bc69c category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/IACZLQ CVE: CVE-2024-41006 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… --------------------------- [ Upstream commit 0b9130247f3b6a1122478471ff0e014ea96bb735 ] syzbot reported a memory leak in nr_create() [0]. Commit 409db27e3a2e ("netrom: Fix use-after-free of a listening socket.") added sock_hold() to the nr_heartbeat_expiry() function, where a) a socket has a SOCK_DESTROY flag or b) a listening socket has a SOCK_DEAD flag. But in the case "a," when the SOCK_DESTROY flag is set, the file descriptor has already been closed and the nr_release() function has been called. So it makes no sense to hold the reference count because no one will call another nr_destroy_socket() and put it as in the case "b." nr_connect nr_establish_data_link nr_start_heartbeat nr_release switch (nr->state) case NR_STATE_3 nr->state = NR_STATE_2 sock_set_flag(sk, SOCK_DESTROY); nr_rx_frame nr_process_rx_frame switch (nr->state) case NR_STATE_2 nr_state2_machine() nr_disconnect() nr_sk(sk)->state = NR_STATE_0 sock_set_flag(sk, SOCK_DEAD) nr_heartbeat_expiry switch (nr->state) case NR_STATE_0 if (sock_flag(sk, SOCK_DESTROY) || (sk->sk_state == TCP_LISTEN && sock_flag(sk, SOCK_DEAD))) sock_hold() // ( !!! ) nr_destroy_socket() To fix the memory leak, let's call sock_hold() only for a listening socket. Found by InfoTeCS on behalf of Linux Verification Center (linuxtesting.org) with Syzkaller. [0]: https://syzkaller.appspot.com/bug?extid=d327a1f3b12e1e206c16 Reported-by: syzbot+d327a1f3b12e1e206c16(a)syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=d327a1f3b12e1e206c16 Fixes: 409db27e3a2e ("netrom: Fix use-after-free of a listening socket.") Signed-off-by: Gavrilov Ilia <Ilia.Gavrilov(a)infotecs.ru> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Liu Jian <liujian56(a)huawei.com> --- net/netrom/nr_timer.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/netrom/nr_timer.c b/net/netrom/nr_timer.c index 2bf99bd5be58..67d012e0bade 100644 --- a/net/netrom/nr_timer.c +++ b/net/netrom/nr_timer.c @@ -124,7 +124,8 @@ static void nr_heartbeat_expiry(struct timer_list *t) is accepted() it isn't 'dead' so doesn't get removed. */ if (sock_flag(sk, SOCK_DESTROY) || (sk->sk_state == TCP_LISTEN && sock_flag(sk, SOCK_DEAD))) { - sock_hold(sk); + if (sk->sk_state == TCP_LISTEN) + sock_hold(sk); bh_unlock_sock(sk); nr_destroy_socket(sk); goto out; -- 2.34.1

2 1

[PATCH OLK-5.10] netrom: Fix a memory leak in nr_heartbeat_expiry()
by Liu Jian 18 Jul '24

18 Jul '24

From: Gavrilov Ilia <Ilia.Gavrilov(a)infotecs.ru> stable inclusion from stable-v5.10.221 commit 5391f9db2cab5ef1cb411be1ab7dbec728078fba category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/IACZLQ CVE: CVE-2024-41006 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… --------------------------- [ Upstream commit 0b9130247f3b6a1122478471ff0e014ea96bb735 ] syzbot reported a memory leak in nr_create() [0]. Commit 409db27e3a2e ("netrom: Fix use-after-free of a listening socket.") added sock_hold() to the nr_heartbeat_expiry() function, where a) a socket has a SOCK_DESTROY flag or b) a listening socket has a SOCK_DEAD flag. But in the case "a," when the SOCK_DESTROY flag is set, the file descriptor has already been closed and the nr_release() function has been called. So it makes no sense to hold the reference count because no one will call another nr_destroy_socket() and put it as in the case "b." nr_connect nr_establish_data_link nr_start_heartbeat nr_release switch (nr->state) case NR_STATE_3 nr->state = NR_STATE_2 sock_set_flag(sk, SOCK_DESTROY); nr_rx_frame nr_process_rx_frame switch (nr->state) case NR_STATE_2 nr_state2_machine() nr_disconnect() nr_sk(sk)->state = NR_STATE_0 sock_set_flag(sk, SOCK_DEAD) nr_heartbeat_expiry switch (nr->state) case NR_STATE_0 if (sock_flag(sk, SOCK_DESTROY) || (sk->sk_state == TCP_LISTEN && sock_flag(sk, SOCK_DEAD))) sock_hold() // ( !!! ) nr_destroy_socket() To fix the memory leak, let's call sock_hold() only for a listening socket. Found by InfoTeCS on behalf of Linux Verification Center (linuxtesting.org) with Syzkaller. [0]: https://syzkaller.appspot.com/bug?extid=d327a1f3b12e1e206c16 Reported-by: syzbot+d327a1f3b12e1e206c16(a)syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=d327a1f3b12e1e206c16 Fixes: 409db27e3a2e ("netrom: Fix use-after-free of a listening socket.") Signed-off-by: Gavrilov Ilia <Ilia.Gavrilov(a)infotecs.ru> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Liu Jian <liujian56(a)huawei.com> --- net/netrom/nr_timer.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/netrom/nr_timer.c b/net/netrom/nr_timer.c index 4e7c968cde2d..5e3ca068f04e 100644 --- a/net/netrom/nr_timer.c +++ b/net/netrom/nr_timer.c @@ -121,7 +121,8 @@ static void nr_heartbeat_expiry(struct timer_list *t) is accepted() it isn't 'dead' so doesn't get removed. */ if (sock_flag(sk, SOCK_DESTROY) || (sk->sk_state == TCP_LISTEN && sock_flag(sk, SOCK_DEAD))) { - sock_hold(sk); + if (sk->sk_state == TCP_LISTEN) + sock_hold(sk); bh_unlock_sock(sk); nr_destroy_socket(sk); goto out; -- 2.34.1

2 1

[PATCH OLK-6.6] netrom: Fix a memory leak in nr_heartbeat_expiry()
by Liu Jian 18 Jul '24

18 Jul '24

From: Gavrilov Ilia <Ilia.Gavrilov(a)infotecs.ru> stable inclusion from stable-v6.6.36 commit b6ebe4fed73eedeb73f4540f8edc4871945474c8 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/IACZLQ CVE: CVE-2024-41006 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… --------------------------- [ Upstream commit 0b9130247f3b6a1122478471ff0e014ea96bb735 ] syzbot reported a memory leak in nr_create() [0]. Commit 409db27e3a2e ("netrom: Fix use-after-free of a listening socket.") added sock_hold() to the nr_heartbeat_expiry() function, where a) a socket has a SOCK_DESTROY flag or b) a listening socket has a SOCK_DEAD flag. But in the case "a," when the SOCK_DESTROY flag is set, the file descriptor has already been closed and the nr_release() function has been called. So it makes no sense to hold the reference count because no one will call another nr_destroy_socket() and put it as in the case "b." nr_connect nr_establish_data_link nr_start_heartbeat nr_release switch (nr->state) case NR_STATE_3 nr->state = NR_STATE_2 sock_set_flag(sk, SOCK_DESTROY); nr_rx_frame nr_process_rx_frame switch (nr->state) case NR_STATE_2 nr_state2_machine() nr_disconnect() nr_sk(sk)->state = NR_STATE_0 sock_set_flag(sk, SOCK_DEAD) nr_heartbeat_expiry switch (nr->state) case NR_STATE_0 if (sock_flag(sk, SOCK_DESTROY) || (sk->sk_state == TCP_LISTEN && sock_flag(sk, SOCK_DEAD))) sock_hold() // ( !!! ) nr_destroy_socket() To fix the memory leak, let's call sock_hold() only for a listening socket. Found by InfoTeCS on behalf of Linux Verification Center (linuxtesting.org) with Syzkaller. [0]: https://syzkaller.appspot.com/bug?extid=d327a1f3b12e1e206c16 Reported-by: syzbot+d327a1f3b12e1e206c16(a)syzkaller.appspotmail.com Closes: https://syzkaller.appspot.com/bug?extid=d327a1f3b12e1e206c16 Fixes: 409db27e3a2e ("netrom: Fix use-after-free of a listening socket.") Signed-off-by: Gavrilov Ilia <Ilia.Gavrilov(a)infotecs.ru> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Liu Jian <liujian56(a)huawei.com> --- net/netrom/nr_timer.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/netrom/nr_timer.c b/net/netrom/nr_timer.c index 4e7c968cde2d..5e3ca068f04e 100644 --- a/net/netrom/nr_timer.c +++ b/net/netrom/nr_timer.c @@ -121,7 +121,8 @@ static void nr_heartbeat_expiry(struct timer_list *t) is accepted() it isn't 'dead' so doesn't get removed. */ if (sock_flag(sk, SOCK_DESTROY) || (sk->sk_state == TCP_LISTEN && sock_flag(sk, SOCK_DEAD))) { - sock_hold(sk); + if (sk->sk_state == TCP_LISTEN) + sock_hold(sk); bh_unlock_sock(sk); nr_destroy_socket(sk); goto out; -- 2.34.1

2 1

[PATCH OLK-6.6] x86: resctrl: Fix illegal access by the chips not having RDT
by Zeng Heng 18 Jul '24

18 Jul '24

hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/IA7L2U --------------------------- Before dom_data_init() initializing the monitor resource, we should check if the msc has the ability of monitor. If the chip not having RDT monitor function, variable idx_limit in dom_data_init() would be set 0. rmid_ptrs points to an illegal buffer after kcalloc() allocation, which would cause illegal access fault when accessing. Fixes: 13e249bf4944 ("x86/resctrl: Move the filesystem portions of resctrl to live in '/fs/'") Signed-off-by: Zeng Heng <zengheng4(a)huawei.com> --- fs/resctrl/monitor.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/fs/resctrl/monitor.c b/fs/resctrl/monitor.c index fcf2ab18966f..68401ca0827a 100644 --- a/fs/resctrl/monitor.c +++ b/fs/resctrl/monitor.c @@ -830,13 +830,13 @@ int resctrl_mon_resource_init(void) struct rdt_resource *r = resctrl_arch_get_resource(RDT_RESOURCE_L3); int ret; + if (!r->mon_capable) + return 0; + ret = dom_data_init(r); if (ret) return ret; - if (!r->mon_capable) - return 0; - l3_mon_evt_init(r); if (resctrl_arch_is_evt_configurable(QOS_L3_MBM_TOTAL_EVENT_ID)) { -- 2.25.1

2 1

[PATCH OLK-6.6 0/1] kvm : add coda feature
by l00500167 18 Jul '24

18 Jul '24

From: JunBin Li <lijunbin4(a)huawei.com> virtCCA feature JunBin Li (1): kvm: add coda feature Signed-off-by: JunBin Li <lijunbin4(a)huawei.com> drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.h | 14 ++++++++++++++ 1 file changed, 14 insertions(+) -- 2.25.1

2 2

[PATCH OLK-6.6] x86: resctrl: Fix illegal access by the chips not having RDT
by Zeng Heng 18 Jul '24

18 Jul '24

hulk inclusion category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/IA7L2U --------------------------- Before dom_data_init() initializing the monitor resource, we should check if the msc has the ability of monitor. If the chip not having RDT monitor function, resctrl_arch_system_num_rmid_idx() would return 0. rmid_ptrs points to an illegal buffer after kcalloc() allocation, which would cause illegal access fault when accessing. Fixes: 13e249bf4944 ("x86/resctrl: Move the filesystem portions of resctrl to live in '/fs/'") Signed-off-by: Zeng Heng <zengheng4(a)huawei.com> --- fs/resctrl/monitor.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/fs/resctrl/monitor.c b/fs/resctrl/monitor.c index fcf2ab18966f..68401ca0827a 100644 --- a/fs/resctrl/monitor.c +++ b/fs/resctrl/monitor.c @@ -830,13 +830,13 @@ int resctrl_mon_resource_init(void) struct rdt_resource *r = resctrl_arch_get_resource(RDT_RESOURCE_L3); int ret; + if (!r->mon_capable) + return 0; + ret = dom_data_init(r); if (ret) return ret; - if (!r->mon_capable) - return 0; - l3_mon_evt_init(r); if (resctrl_arch_is_evt_configurable(QOS_L3_MBM_TOTAL_EVENT_ID)) { -- 2.25.1

2 1

[PATCH OLK-6.6 0/4] CVE-2024-36022
by Liu Chuang 18 Jul '24

18 Jul '24

fixing CVE-2024-36022 Ahmad Rehman (1): [Backport] drm/amdgpu: Init zone device and drm client after mode-1 reset on reload Felix Kuehling (1): [Backport] drm/amdkfd: Export DMABufs from KFD using GEM handles Flora Cui (1): [Backport] drm/amdkfd: init drm_client with funcs hook Le Ma (1): [Backport] drm/amdgpu: move the drm client creation behind drm device registration drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c | 24 ++++++++++++++ drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h | 7 ++++ .../gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c | 33 +++++++++++++++---- drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 9 ++++- drivers/gpu/drm/amd/amdkfd/kfd_chardev.c | 4 +-- 5 files changed, 67 insertions(+), 10 deletions(-) -- 2.34.1

2 5

[PATCH OLK-6.6] drop_monitor: replace spin_lock by raw_spin_lock
by Liu Jian 18 Jul '24

18 Jul '24

From: Wander Lairson Costa <wander(a)redhat.com> stable inclusion from stable-v6.6.36 commit 76ce2f9125244e1708d29c1d3f9d1d50b347bda0 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/IACS4X CVE: CVE-2024-40980 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… --------------------------- [ Upstream commit f1e197a665c2148ebc25fe09c53689e60afea195 ] trace_drop_common() is called with preemption disabled, and it acquires a spin_lock. This is problematic for RT kernels because spin_locks are sleeping locks in this configuration, which causes the following splat: BUG: sleeping function called from invalid context at kernel/locking/spinlock_rt.c:48 in_atomic(): 1, irqs_disabled(): 1, non_block: 0, pid: 449, name: rcuc/47 preempt_count: 1, expected: 0 RCU nest depth: 2, expected: 2 5 locks held by rcuc/47/449: #0: ff1100086ec30a60 ((softirq_ctrl.lock)){+.+.}-{2:2}, at: __local_bh_disable_ip+0x105/0x210 #1: ffffffffb394a280 (rcu_read_lock){....}-{1:2}, at: rt_spin_lock+0xbf/0x130 #2: ffffffffb394a280 (rcu_read_lock){....}-{1:2}, at: __local_bh_disable_ip+0x11c/0x210 #3: ffffffffb394a160 (rcu_callback){....}-{0:0}, at: rcu_do_batch+0x360/0xc70 #4: ff1100086ee07520 (&data->lock){+.+.}-{2:2}, at: trace_drop_common.constprop.0+0xb5/0x290 irq event stamp: 139909 hardirqs last enabled at (139908): [<ffffffffb1df2b33>] _raw_spin_unlock_irqrestore+0x63/0x80 hardirqs last disabled at (139909): [<ffffffffb19bd03d>] trace_drop_common.constprop.0+0x26d/0x290 softirqs last enabled at (139892): [<ffffffffb07a1083>] __local_bh_enable_ip+0x103/0x170 softirqs last disabled at (139898): [<ffffffffb0909b33>] rcu_cpu_kthread+0x93/0x1f0 Preemption disabled at: [<ffffffffb1de786b>] rt_mutex_slowunlock+0xab/0x2e0 CPU: 47 PID: 449 Comm: rcuc/47 Not tainted 6.9.0-rc2-rt1+ #7 Hardware name: Dell Inc. PowerEdge R650/0Y2G81, BIOS 1.6.5 04/15/2022 Call Trace: <TASK> dump_stack_lvl+0x8c/0xd0 dump_stack+0x14/0x20 __might_resched+0x21e/0x2f0 rt_spin_lock+0x5e/0x130 ? trace_drop_common.constprop.0+0xb5/0x290 ? skb_queue_purge_reason.part.0+0x1bf/0x230 trace_drop_common.constprop.0+0xb5/0x290 ? preempt_count_sub+0x1c/0xd0 ? _raw_spin_unlock_irqrestore+0x4a/0x80 ? __pfx_trace_drop_common.constprop.0+0x10/0x10 ? rt_mutex_slowunlock+0x26a/0x2e0 ? skb_queue_purge_reason.part.0+0x1bf/0x230 ? __pfx_rt_mutex_slowunlock+0x10/0x10 ? skb_queue_purge_reason.part.0+0x1bf/0x230 trace_kfree_skb_hit+0x15/0x20 trace_kfree_skb+0xe9/0x150 kfree_skb_reason+0x7b/0x110 skb_queue_purge_reason.part.0+0x1bf/0x230 ? __pfx_skb_queue_purge_reason.part.0+0x10/0x10 ? mark_lock.part.0+0x8a/0x520 ... trace_drop_common() also disables interrupts, but this is a minor issue because we could easily replace it with a local_lock. Replace the spin_lock with raw_spin_lock to avoid sleeping in atomic context. Signed-off-by: Wander Lairson Costa <wander(a)redhat.com> Reported-by: Hu Chunyu <chuhu(a)redhat.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Liu Jian <liujian56(a)huawei.com> --- net/core/drop_monitor.c | 20 ++++++++++---------- 1 file changed, 10 insertions(+), 10 deletions(-) diff --git a/net/core/drop_monitor.c b/net/core/drop_monitor.c index b240d9aae4a6..58843a52bad0 100644 --- a/net/core/drop_monitor.c +++ b/net/core/drop_monitor.c @@ -74,7 +74,7 @@ struct net_dm_hw_entries { }; struct per_cpu_dm_data { - spinlock_t lock; /* Protects 'skb', 'hw_entries' and + raw_spinlock_t lock; /* Protects 'skb', 'hw_entries' and * 'send_timer' */ union { @@ -168,9 +168,9 @@ static struct sk_buff *reset_per_cpu_data(struct per_cpu_dm_data *data) err: mod_timer(&data->send_timer, jiffies + HZ / 10); out: - spin_lock_irqsave(&data->lock, flags); + raw_spin_lock_irqsave(&data->lock, flags); swap(data->skb, skb); - spin_unlock_irqrestore(&data->lock, flags); + raw_spin_unlock_irqrestore(&data->lock, flags); if (skb) { struct nlmsghdr *nlh = (struct nlmsghdr *)skb->data; @@ -225,7 +225,7 @@ static void trace_drop_common(struct sk_buff *skb, void *location) local_irq_save(flags); data = this_cpu_ptr(&dm_cpu_data); - spin_lock(&data->lock); + raw_spin_lock(&data->lock); dskb = data->skb; if (!dskb) @@ -259,7 +259,7 @@ static void trace_drop_common(struct sk_buff *skb, void *location) } out: - spin_unlock_irqrestore(&data->lock, flags); + raw_spin_unlock_irqrestore(&data->lock, flags); } static void trace_kfree_skb_hit(void *ignore, struct sk_buff *skb, @@ -314,9 +314,9 @@ net_dm_hw_reset_per_cpu_data(struct per_cpu_dm_data *hw_data) mod_timer(&hw_data->send_timer, jiffies + HZ / 10); } - spin_lock_irqsave(&hw_data->lock, flags); + raw_spin_lock_irqsave(&hw_data->lock, flags); swap(hw_data->hw_entries, hw_entries); - spin_unlock_irqrestore(&hw_data->lock, flags); + raw_spin_unlock_irqrestore(&hw_data->lock, flags); return hw_entries; } @@ -448,7 +448,7 @@ net_dm_hw_trap_summary_probe(void *ignore, const struct devlink *devlink, return; hw_data = this_cpu_ptr(&dm_hw_cpu_data); - spin_lock_irqsave(&hw_data->lock, flags); + raw_spin_lock_irqsave(&hw_data->lock, flags); hw_entries = hw_data->hw_entries; if (!hw_entries) @@ -477,7 +477,7 @@ net_dm_hw_trap_summary_probe(void *ignore, const struct devlink *devlink, } out: - spin_unlock_irqrestore(&hw_data->lock, flags); + raw_spin_unlock_irqrestore(&hw_data->lock, flags); } static const struct net_dm_alert_ops net_dm_alert_summary_ops = { @@ -1673,7 +1673,7 @@ static struct notifier_block dropmon_net_notifier = { static void __net_dm_cpu_data_init(struct per_cpu_dm_data *data) { - spin_lock_init(&data->lock); + raw_spin_lock_init(&data->lock); skb_queue_head_init(&data->drop_queue); u64_stats_init(&data->stats.syncp); } -- 2.34.1

2 1

[PATCH OLK-5.10] drop_monitor: replace spin_lock by raw_spin_lock
by Liu Jian 18 Jul '24

18 Jul '24

From: Wander Lairson Costa <wander(a)redhat.com> stable inclusion from stable-v5.10.221 commit 96941f29ebcc1e9cbf570dc903f30374909562f5 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/IACS4X CVE: CVE-2024-40980 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… --------------------------- [ Upstream commit f1e197a665c2148ebc25fe09c53689e60afea195 ] trace_drop_common() is called with preemption disabled, and it acquires a spin_lock. This is problematic for RT kernels because spin_locks are sleeping locks in this configuration, which causes the following splat: BUG: sleeping function called from invalid context at kernel/locking/spinlock_rt.c:48 in_atomic(): 1, irqs_disabled(): 1, non_block: 0, pid: 449, name: rcuc/47 preempt_count: 1, expected: 0 RCU nest depth: 2, expected: 2 5 locks held by rcuc/47/449: #0: ff1100086ec30a60 ((softirq_ctrl.lock)){+.+.}-{2:2}, at: __local_bh_disable_ip+0x105/0x210 #1: ffffffffb394a280 (rcu_read_lock){....}-{1:2}, at: rt_spin_lock+0xbf/0x130 #2: ffffffffb394a280 (rcu_read_lock){....}-{1:2}, at: __local_bh_disable_ip+0x11c/0x210 #3: ffffffffb394a160 (rcu_callback){....}-{0:0}, at: rcu_do_batch+0x360/0xc70 #4: ff1100086ee07520 (&data->lock){+.+.}-{2:2}, at: trace_drop_common.constprop.0+0xb5/0x290 irq event stamp: 139909 hardirqs last enabled at (139908): [<ffffffffb1df2b33>] _raw_spin_unlock_irqrestore+0x63/0x80 hardirqs last disabled at (139909): [<ffffffffb19bd03d>] trace_drop_common.constprop.0+0x26d/0x290 softirqs last enabled at (139892): [<ffffffffb07a1083>] __local_bh_enable_ip+0x103/0x170 softirqs last disabled at (139898): [<ffffffffb0909b33>] rcu_cpu_kthread+0x93/0x1f0 Preemption disabled at: [<ffffffffb1de786b>] rt_mutex_slowunlock+0xab/0x2e0 CPU: 47 PID: 449 Comm: rcuc/47 Not tainted 6.9.0-rc2-rt1+ #7 Hardware name: Dell Inc. PowerEdge R650/0Y2G81, BIOS 1.6.5 04/15/2022 Call Trace: <TASK> dump_stack_lvl+0x8c/0xd0 dump_stack+0x14/0x20 __might_resched+0x21e/0x2f0 rt_spin_lock+0x5e/0x130 ? trace_drop_common.constprop.0+0xb5/0x290 ? skb_queue_purge_reason.part.0+0x1bf/0x230 trace_drop_common.constprop.0+0xb5/0x290 ? preempt_count_sub+0x1c/0xd0 ? _raw_spin_unlock_irqrestore+0x4a/0x80 ? __pfx_trace_drop_common.constprop.0+0x10/0x10 ? rt_mutex_slowunlock+0x26a/0x2e0 ? skb_queue_purge_reason.part.0+0x1bf/0x230 ? __pfx_rt_mutex_slowunlock+0x10/0x10 ? skb_queue_purge_reason.part.0+0x1bf/0x230 trace_kfree_skb_hit+0x15/0x20 trace_kfree_skb+0xe9/0x150 kfree_skb_reason+0x7b/0x110 skb_queue_purge_reason.part.0+0x1bf/0x230 ? __pfx_skb_queue_purge_reason.part.0+0x10/0x10 ? mark_lock.part.0+0x8a/0x520 ... trace_drop_common() also disables interrupts, but this is a minor issue because we could easily replace it with a local_lock. Replace the spin_lock with raw_spin_lock to avoid sleeping in atomic context. Signed-off-by: Wander Lairson Costa <wander(a)redhat.com> Reported-by: Hu Chunyu <chuhu(a)redhat.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Liu Jian <liujian56(a)huawei.com> --- net/core/drop_monitor.c | 20 ++++++++++---------- 1 file changed, 10 insertions(+), 10 deletions(-) diff --git a/net/core/drop_monitor.c b/net/core/drop_monitor.c index a3412464cbd3..581df3596680 100644 --- a/net/core/drop_monitor.c +++ b/net/core/drop_monitor.c @@ -73,7 +73,7 @@ struct net_dm_hw_entries { }; struct per_cpu_dm_data { - spinlock_t lock; /* Protects 'skb', 'hw_entries' and + raw_spinlock_t lock; /* Protects 'skb', 'hw_entries' and * 'send_timer' */ union { @@ -168,9 +168,9 @@ static struct sk_buff *reset_per_cpu_data(struct per_cpu_dm_data *data) err: mod_timer(&data->send_timer, jiffies + HZ / 10); out: - spin_lock_irqsave(&data->lock, flags); + raw_spin_lock_irqsave(&data->lock, flags); swap(data->skb, skb); - spin_unlock_irqrestore(&data->lock, flags); + raw_spin_unlock_irqrestore(&data->lock, flags); if (skb) { struct nlmsghdr *nlh = (struct nlmsghdr *)skb->data; @@ -225,7 +225,7 @@ static void trace_drop_common(struct sk_buff *skb, void *location) local_irq_save(flags); data = this_cpu_ptr(&dm_cpu_data); - spin_lock(&data->lock); + raw_spin_lock(&data->lock); dskb = data->skb; if (!dskb) @@ -259,7 +259,7 @@ static void trace_drop_common(struct sk_buff *skb, void *location) } out: - spin_unlock_irqrestore(&data->lock, flags); + raw_spin_unlock_irqrestore(&data->lock, flags); } static void trace_kfree_skb_hit(void *ignore, struct sk_buff *skb, void *location) @@ -318,9 +318,9 @@ net_dm_hw_reset_per_cpu_data(struct per_cpu_dm_data *hw_data) mod_timer(&hw_data->send_timer, jiffies + HZ / 10); } - spin_lock_irqsave(&hw_data->lock, flags); + raw_spin_lock_irqsave(&hw_data->lock, flags); swap(hw_data->hw_entries, hw_entries); - spin_unlock_irqrestore(&hw_data->lock, flags); + raw_spin_unlock_irqrestore(&hw_data->lock, flags); return hw_entries; } @@ -452,7 +452,7 @@ net_dm_hw_trap_summary_probe(void *ignore, const struct devlink *devlink, return; hw_data = this_cpu_ptr(&dm_hw_cpu_data); - spin_lock_irqsave(&hw_data->lock, flags); + raw_spin_lock_irqsave(&hw_data->lock, flags); hw_entries = hw_data->hw_entries; if (!hw_entries) @@ -481,7 +481,7 @@ net_dm_hw_trap_summary_probe(void *ignore, const struct devlink *devlink, } out: - spin_unlock_irqrestore(&hw_data->lock, flags); + raw_spin_unlock_irqrestore(&hw_data->lock, flags); } static const struct net_dm_alert_ops net_dm_alert_summary_ops = { @@ -1669,7 +1669,7 @@ static struct notifier_block dropmon_net_notifier = { static void __net_dm_cpu_data_init(struct per_cpu_dm_data *data) { - spin_lock_init(&data->lock); + raw_spin_lock_init(&data->lock); skb_queue_head_init(&data->drop_queue); u64_stats_init(&data->stats.syncp); } -- 2.34.1

2 1

[PATCH openEuler-22.03-LTS-SP1] drop_monitor: replace spin_lock by raw_spin_lock
by Liu Jian 18 Jul '24

18 Jul '24

From: Wander Lairson Costa <wander(a)redhat.com> stable inclusion from stable-v5.10.221 commit 96941f29ebcc1e9cbf570dc903f30374909562f5 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/IACS4X CVE: CVE-2024-40980 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id… --------------------------- [ Upstream commit f1e197a665c2148ebc25fe09c53689e60afea195 ] trace_drop_common() is called with preemption disabled, and it acquires a spin_lock. This is problematic for RT kernels because spin_locks are sleeping locks in this configuration, which causes the following splat: BUG: sleeping function called from invalid context at kernel/locking/spinlock_rt.c:48 in_atomic(): 1, irqs_disabled(): 1, non_block: 0, pid: 449, name: rcuc/47 preempt_count: 1, expected: 0 RCU nest depth: 2, expected: 2 5 locks held by rcuc/47/449: #0: ff1100086ec30a60 ((softirq_ctrl.lock)){+.+.}-{2:2}, at: __local_bh_disable_ip+0x105/0x210 #1: ffffffffb394a280 (rcu_read_lock){....}-{1:2}, at: rt_spin_lock+0xbf/0x130 #2: ffffffffb394a280 (rcu_read_lock){....}-{1:2}, at: __local_bh_disable_ip+0x11c/0x210 #3: ffffffffb394a160 (rcu_callback){....}-{0:0}, at: rcu_do_batch+0x360/0xc70 #4: ff1100086ee07520 (&data->lock){+.+.}-{2:2}, at: trace_drop_common.constprop.0+0xb5/0x290 irq event stamp: 139909 hardirqs last enabled at (139908): [<ffffffffb1df2b33>] _raw_spin_unlock_irqrestore+0x63/0x80 hardirqs last disabled at (139909): [<ffffffffb19bd03d>] trace_drop_common.constprop.0+0x26d/0x290 softirqs last enabled at (139892): [<ffffffffb07a1083>] __local_bh_enable_ip+0x103/0x170 softirqs last disabled at (139898): [<ffffffffb0909b33>] rcu_cpu_kthread+0x93/0x1f0 Preemption disabled at: [<ffffffffb1de786b>] rt_mutex_slowunlock+0xab/0x2e0 CPU: 47 PID: 449 Comm: rcuc/47 Not tainted 6.9.0-rc2-rt1+ #7 Hardware name: Dell Inc. PowerEdge R650/0Y2G81, BIOS 1.6.5 04/15/2022 Call Trace: <TASK> dump_stack_lvl+0x8c/0xd0 dump_stack+0x14/0x20 __might_resched+0x21e/0x2f0 rt_spin_lock+0x5e/0x130 ? trace_drop_common.constprop.0+0xb5/0x290 ? skb_queue_purge_reason.part.0+0x1bf/0x230 trace_drop_common.constprop.0+0xb5/0x290 ? preempt_count_sub+0x1c/0xd0 ? _raw_spin_unlock_irqrestore+0x4a/0x80 ? __pfx_trace_drop_common.constprop.0+0x10/0x10 ? rt_mutex_slowunlock+0x26a/0x2e0 ? skb_queue_purge_reason.part.0+0x1bf/0x230 ? __pfx_rt_mutex_slowunlock+0x10/0x10 ? skb_queue_purge_reason.part.0+0x1bf/0x230 trace_kfree_skb_hit+0x15/0x20 trace_kfree_skb+0xe9/0x150 kfree_skb_reason+0x7b/0x110 skb_queue_purge_reason.part.0+0x1bf/0x230 ? __pfx_skb_queue_purge_reason.part.0+0x10/0x10 ? mark_lock.part.0+0x8a/0x520 ... trace_drop_common() also disables interrupts, but this is a minor issue because we could easily replace it with a local_lock. Replace the spin_lock with raw_spin_lock to avoid sleeping in atomic context. Signed-off-by: Wander Lairson Costa <wander(a)redhat.com> Reported-by: Hu Chunyu <chuhu(a)redhat.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Sasha Levin <sashal(a)kernel.org> Signed-off-by: Liu Jian <liujian56(a)huawei.com> --- net/core/drop_monitor.c | 20 ++++++++++---------- 1 file changed, 10 insertions(+), 10 deletions(-) diff --git a/net/core/drop_monitor.c b/net/core/drop_monitor.c index ed9dd17f9348..d48d3527d47a 100644 --- a/net/core/drop_monitor.c +++ b/net/core/drop_monitor.c @@ -73,7 +73,7 @@ struct net_dm_hw_entries { }; struct per_cpu_dm_data { - spinlock_t lock; /* Protects 'skb', 'hw_entries' and + raw_spinlock_t lock; /* Protects 'skb', 'hw_entries' and * 'send_timer' */ union { @@ -168,9 +168,9 @@ static struct sk_buff *reset_per_cpu_data(struct per_cpu_dm_data *data) err: mod_timer(&data->send_timer, jiffies + HZ / 10); out: - spin_lock_irqsave(&data->lock, flags); + raw_spin_lock_irqsave(&data->lock, flags); swap(data->skb, skb); - spin_unlock_irqrestore(&data->lock, flags); + raw_spin_unlock_irqrestore(&data->lock, flags); if (skb) { struct nlmsghdr *nlh = (struct nlmsghdr *)skb->data; @@ -225,7 +225,7 @@ static void trace_drop_common(struct sk_buff *skb, void *location) local_irq_save(flags); data = this_cpu_ptr(&dm_cpu_data); - spin_lock(&data->lock); + raw_spin_lock(&data->lock); dskb = data->skb; if (!dskb) @@ -259,7 +259,7 @@ static void trace_drop_common(struct sk_buff *skb, void *location) } out: - spin_unlock_irqrestore(&data->lock, flags); + raw_spin_unlock_irqrestore(&data->lock, flags); } static void trace_kfree_skb_hit(void *ignore, struct sk_buff *skb, void *location) @@ -318,9 +318,9 @@ net_dm_hw_reset_per_cpu_data(struct per_cpu_dm_data *hw_data) mod_timer(&hw_data->send_timer, jiffies + HZ / 10); } - spin_lock_irqsave(&hw_data->lock, flags); + raw_spin_lock_irqsave(&hw_data->lock, flags); swap(hw_data->hw_entries, hw_entries); - spin_unlock_irqrestore(&hw_data->lock, flags); + raw_spin_unlock_irqrestore(&hw_data->lock, flags); return hw_entries; } @@ -452,7 +452,7 @@ net_dm_hw_trap_summary_probe(void *ignore, const struct devlink *devlink, return; hw_data = this_cpu_ptr(&dm_hw_cpu_data); - spin_lock_irqsave(&hw_data->lock, flags); + raw_spin_lock_irqsave(&hw_data->lock, flags); hw_entries = hw_data->hw_entries; if (!hw_entries) @@ -481,7 +481,7 @@ net_dm_hw_trap_summary_probe(void *ignore, const struct devlink *devlink, } out: - spin_unlock_irqrestore(&hw_data->lock, flags); + raw_spin_unlock_irqrestore(&hw_data->lock, flags); } static const struct net_dm_alert_ops net_dm_alert_summary_ops = { @@ -1667,7 +1667,7 @@ static struct notifier_block dropmon_net_notifier = { static void __net_dm_cpu_data_init(struct per_cpu_dm_data *data) { - spin_lock_init(&data->lock); + raw_spin_lock_init(&data->lock); skb_queue_head_init(&data->drop_queue); u64_stats_init(&data->stats.syncp); } -- 2.34.1

2 1

2024

2023

2022

2021

2020

2019

Kernel July 2024