May 2025 - Linux-stable-mirror

[PATCH v2] net: mlx5: vport: Add error handling in mlx5_query_nic_vport_node_guid()

by Wentao Liang

The function mlx5_query_nic_vport_node_guid() calls the function mlx5_query_nic_vport_context() but does not check its return value. A proper implementation can be found in mlx5_nic_vport_query_local_lb(). Add error handling for mlx5_query_nic_vport_context(). If it fails, free the out buffer via kvfree() and return error code. Fixes: 9efa75254593 ("net/mlx5_core: Introduce access functions to query vport RoCE fields") Cc: stable(a)vger.kernel.org # v4.5 Signed-off-by: Wentao Liang <vulab(a)iscas.ac.cn> --- v2: Remove redundant reassignment. Fix typo error. drivers/net/ethernet/mellanox/mlx5/core/vport.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/vport.c b/drivers/net/ethernet/mellanox/mlx5/core/vport.c index ded086ffe8ac..8235fa6add03 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/vport.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/vport.c @@ -465,19 +465,22 @@ int mlx5_query_nic_vport_node_guid(struct mlx5_core_dev *mdev, u64 *node_guid) { u32 *out; int outlen = MLX5_ST_SZ_BYTES(query_nic_vport_context_out); + int ret; out = kvzalloc(outlen, GFP_KERNEL); if (!out) return -ENOMEM; - mlx5_query_nic_vport_context(mdev, 0, out); + ret = mlx5_query_nic_vport_context(mdev, 0, out); + if (ret) + goto err; *node_guid = MLX5_GET64(query_nic_vport_context_out, out, nic_vport_context.node_guid); - +err: kvfree(out); - return 0; + return ret; } EXPORT_SYMBOL_GPL(mlx5_query_nic_vport_node_guid); -- 2.42.0.windows.2

3 months, 3 weeks

2
1
0 0

[PATCH 6.1.y] selinux: avoid dereference of garbage after mount failure

by jianqi.ren.cn＠windriver.com

From: Christian Göttsche <cgzones(a)googlemail.com> commit 37801a36b4d68892ce807264f784d818f8d0d39b upstream. In case kern_mount() fails and returns an error pointer return in the error branch instead of continuing and dereferencing the error pointer. While on it drop the never read static variable selinuxfs_mount. Cc: stable(a)vger.kernel.org Fixes: 0619f0f5e36f ("selinux: wrap selinuxfs state") Signed-off-by: Christian Göttsche <cgzones(a)googlemail.com> Signed-off-by: Paul Moore <paul(a)paul-moore.com> Signed-off-by: Jianqi Ren <jianqi.ren.cn(a)windriver.com> Signed-off-by: He Zhe <zhe.he(a)windriver.com> --- Verified the build test --- security/selinux/selinuxfs.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) diff --git a/security/selinux/selinuxfs.c b/security/selinux/selinuxfs.c index ab804d4ea911..c236f3cd2dd7 100644 --- a/security/selinux/selinuxfs.c +++ b/security/selinux/selinuxfs.c @@ -2210,7 +2210,6 @@ static struct file_system_type sel_fs_type = { .kill_sb = sel_kill_sb, }; -static struct vfsmount *selinuxfs_mount __ro_after_init; struct path selinux_null __ro_after_init; static int __init init_sel_fs(void) @@ -2232,18 +2231,21 @@ static int __init init_sel_fs(void) return err; } - selinux_null.mnt = selinuxfs_mount = kern_mount(&sel_fs_type); - if (IS_ERR(selinuxfs_mount)) { + selinux_null.mnt = kern_mount(&sel_fs_type); + if (IS_ERR(selinux_null.mnt)) { pr_err("selinuxfs: could not mount!\n"); - err = PTR_ERR(selinuxfs_mount); - selinuxfs_mount = NULL; + err = PTR_ERR(selinux_null.mnt); + selinux_null.mnt = NULL; + return err; } + selinux_null.dentry = d_hash_and_lookup(selinux_null.mnt->mnt_root, &null_name); if (IS_ERR(selinux_null.dentry)) { pr_err("selinuxfs: could not lookup null!\n"); err = PTR_ERR(selinux_null.dentry); selinux_null.dentry = NULL; + return err; } return err; -- 2.34.1

3 months, 3 weeks

3
2
0 0

[PATCH v2 2/2] Input: adp5588-keys - Prevent buffer overflow

by Denis Arefev

If the value of 'key_val' is less than 1 or greater than 80, a buffer overflow may occur. Add a check for valid values 'key_val'. Found by Linux Verification Center (linuxtesting.org) with SVACE. Fixes: 69a4af606ed4 ("Input: adp5588-keys - support GPI events for ADP5588 devices") Cc: stable(a)vger.kernel.org Signed-off-by: Denis Arefev <arefev(a)swemel.ru> --- V1 -> V2: Added tag Fixes drivers/input/keyboard/adp5588-keys.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/drivers/input/keyboard/adp5588-keys.c b/drivers/input/keyboard/adp5588-keys.c index 13136f863270..91f00d6e2413 100644 --- a/drivers/input/keyboard/adp5588-keys.c +++ b/drivers/input/keyboard/adp5588-keys.c @@ -164,6 +164,9 @@ #define KEY_EV_PRESSED BIT(7) #define KEY_EV_MASK GENMASK(6, 0) +#define KEY_EVENT_MIN 1 +#define KEY_EVENT_MAX 80 + #define KP_SEL(x) (BIT(x) - 1) /* 2^x-1 */ #define KEYP_MAX_EVENT 10 @@ -531,7 +534,7 @@ static void adp5588_report_events(struct adp5588_kpad *kpad, int ev_cnt) if (key_val >= GPI_PIN_BASE && key_val <= GPI_PIN_END) { /* gpio line used as IRQ source */ adp5588_gpio_irq_handle(kpad, key_val, key_press); - } else { + } else if (key_val >= KEY_EVENT_MIN && key_val <= KEY_EVENT_MAX) { int row = (key_val - 1) / ADP5588_COLS_MAX; int col = (key_val - 1) % ADP5588_COLS_MAX; int code = MATRIX_SCAN_CODE(row, col, kpad->row_shift); @@ -542,6 +545,8 @@ static void adp5588_report_events(struct adp5588_kpad *kpad, int ev_cnt) input_report_key(kpad->input, kpad->keycode[code], key_press); + } else { + dev_err_ratelimited(&kpad->client->dev, "invalid report key value %d", key); } } } -- 2.43.0

3 months, 3 weeks

1
0
0 0

[PATCH v2 1/2] Input: adp5588-keys - Add check on return code

by Denis Arefev

Function 'adp5588_read()' can return a negative value, which after calculations will be used as an index to access the array 'kpad->keycode'. Add a check for the return value. Found by Linux Verification Center (linuxtesting.org) with SVACE. Fixes: 69a4af606ed4 ("Input: adp5588-keys - support GPI events for ADP5588 devices") Cc: stable(a)vger.kernel.org Signed-off-by: Denis Arefev <arefev(a)swemel.ru> --- V1 -> V2: Added tag Fixes drivers/input/keyboard/adp5588-keys.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/drivers/input/keyboard/adp5588-keys.c b/drivers/input/keyboard/adp5588-keys.c index dc734974ce06..13136f863270 100644 --- a/drivers/input/keyboard/adp5588-keys.c +++ b/drivers/input/keyboard/adp5588-keys.c @@ -519,9 +519,14 @@ static void adp5588_report_events(struct adp5588_kpad *kpad, int ev_cnt) int i; for (i = 0; i < ev_cnt; i++) { - int key = adp5588_read(kpad->client, KEY_EVENTA + i); - int key_val = key & KEY_EV_MASK; - int key_press = key & KEY_EV_PRESSED; + int key, key_val, key_press; + + key = adp5588_read(kpad->client, KEY_EVENTA + i); + if (key < 0) + continue; + + key_val = key & KEY_EV_MASK; + key_press = key & KEY_EV_PRESSED; if (key_val >= GPI_PIN_BASE && key_val <= GPI_PIN_END) { /* gpio line used as IRQ source */ -- 2.43.0

3 months, 3 weeks

1
0
0 0

[PATCH 6.1.y 1/2] drm/msm/disp/dpu: use atomic enable/disable callbacks for encoder functions

by jianqi.ren.cn＠windriver.com

From: Vinod Polimera <quic_vpolimer(a)quicinc.com> [ Upstream commit c0cd12a5d29fa36a8e2ebac7b8bec50c1a41fb57 ] Use atomic variants for encoder callback functions such that certain states like self-refresh can be accessed as part of enable/disable sequence. Signed-off-by: Kalyan Thota <quic_kalyant(a)quicinc.com> Signed-off-by: Vinod Polimera <quic_vpolimer(a)quicinc.com> Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov(a)linaro.org> Patchwork: https://patchwork.freedesktop.org/patch/524738/ Link: https://lore.kernel.org/r/1677774797-31063-12-git-send-email-quic_vpolimer@… Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov(a)linaro.org> Signed-off-by: Jianqi Ren <jianqi.ren.cn(a)windriver.com> Signed-off-by: He Zhe <zhe.he(a)windriver.com> --- Verified the build test --- drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c index 5f8345016ffe..c7fcd617b48c 100644 --- a/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c +++ b/drivers/gpu/drm/msm/disp/dpu1/dpu_encoder.c @@ -1182,7 +1182,8 @@ void dpu_encoder_virt_runtime_resume(struct drm_encoder *drm_enc) mutex_unlock(&dpu_enc->enc_lock); } -static void dpu_encoder_virt_enable(struct drm_encoder *drm_enc) +static void dpu_encoder_virt_atomic_enable(struct drm_encoder *drm_enc, + struct drm_atomic_state *state) { struct dpu_encoder_virt *dpu_enc = NULL; int ret = 0; @@ -1218,7 +1219,8 @@ static void dpu_encoder_virt_enable(struct drm_encoder *drm_enc) mutex_unlock(&dpu_enc->enc_lock); } -static void dpu_encoder_virt_disable(struct drm_encoder *drm_enc) +static void dpu_encoder_virt_atomic_disable(struct drm_encoder *drm_enc, + struct drm_atomic_state *state) { struct dpu_encoder_virt *dpu_enc = NULL; int i = 0; @@ -2407,8 +2409,8 @@ static void dpu_encoder_frame_done_timeout(struct timer_list *t) static const struct drm_encoder_helper_funcs dpu_encoder_helper_funcs = { .atomic_mode_set = dpu_encoder_virt_atomic_mode_set, - .disable = dpu_encoder_virt_disable, - .enable = dpu_encoder_virt_enable, + .atomic_disable = dpu_encoder_virt_atomic_disable, + .atomic_enable = dpu_encoder_virt_atomic_enable, .atomic_check = dpu_encoder_virt_atomic_check, }; -- 2.34.1

3 months, 3 weeks

2
1
0 0

Request for backporting accel/ivpu FW log patches to 6.12

by Jacek Lawrynowicz

Hi, Please cherry-pick following 4 patches to 6.12: 3a3fb8110c65d361cd9d750c9e16520f740c93f2 accel/ivpu: Rename ivpu_log_level to fw_log_level 4b4d9e394b6f45ac26ac6144b31604c76b7e3705 accel/ivpu: Reset fw log on cold boot 1fc1251149a76d3b75d7f4c94d9c4e081b7df6b4 accel/ivpu: Refactor functions in ivpu_fw_log.c 4bc988b47019536b3b1f7d9c5b83893c712d94d6 accel/ivpu: Fix fw log printing These are fixing some firmware log corner cases that allow us to get reliable output in case of a failure. They should apply without conflicts. Thanks, Jacek

3 months, 3 weeks

2
1
0 0

[PATCH net 3/3] can: bcm: add missing rcu read protection for procfs content

by Marc Kleine-Budde

From: Oliver Hartkopp <socketcan(a)hartkopp.net> When the procfs content is generated for a bcm_op which is in the process to be removed the procfs output might show unreliable data (UAF). As the removal of bcm_op's is already implemented with rcu handling this patch adds the missing rcu_read_lock() and makes sure the list entries are properly removed under rcu protection. Fixes: f1b4e32aca08 ("can: bcm: use call_rcu() instead of costly synchronize_rcu()") Reported-by: Anderson Nascimento <anderson(a)allelesecurity.com> Suggested-by: Anderson Nascimento <anderson(a)allelesecurity.com> Tested-by: Anderson Nascimento <anderson(a)allelesecurity.com> Signed-off-by: Oliver Hartkopp <socketcan(a)hartkopp.net> Link: https://patch.msgid.link/20250519125027.11900-2-socketcan@hartkopp.net Cc: stable(a)vger.kernel.org # >= 5.4 Signed-off-by: Marc Kleine-Budde <mkl(a)pengutronix.de> --- net/can/bcm.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/net/can/bcm.c b/net/can/bcm.c index 871707dab7db..6bc1cc4c94c5 100644 --- a/net/can/bcm.c +++ b/net/can/bcm.c @@ -219,7 +219,9 @@ static int bcm_proc_show(struct seq_file *m, void *v) seq_printf(m, " / bound %s", bcm_proc_getifname(net, ifname, bo->ifindex)); seq_printf(m, " <<<\n"); - list_for_each_entry(op, &bo->rx_ops, list) { + rcu_read_lock(); + + list_for_each_entry_rcu(op, &bo->rx_ops, list) { unsigned long reduction; @@ -275,6 +277,9 @@ static int bcm_proc_show(struct seq_file *m, void *v) seq_printf(m, "# sent %ld\n", op->frames_abs); } seq_putc(m, '\n'); + + rcu_read_unlock(); + return 0; } #endif /* CONFIG_PROC_FS */ @@ -858,7 +863,7 @@ static int bcm_delete_rx_op(struct list_head *ops, struct bcm_msg_head *mh, REGMASK(op->can_id), bcm_rx_handler, op); - list_del(&op->list); + list_del_rcu(&op->list); bcm_remove_op(op); return 1; /* done */ } @@ -878,7 +883,7 @@ static int bcm_delete_tx_op(struct list_head *ops, struct bcm_msg_head *mh, list_for_each_entry_safe(op, n, ops, list) { if ((op->can_id == mh->can_id) && (op->ifindex == ifindex) && (op->flags & CAN_FD_FRAME) == (mh->flags & CAN_FD_FRAME)) { - list_del(&op->list); + list_del_rcu(&op->list); bcm_remove_op(op); return 1; /* done */ } @@ -1296,7 +1301,7 @@ static int bcm_rx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg, bcm_rx_handler, op, "bcm", sk); if (err) { /* this bcm rx op is broken -> remove it */ - list_del(&op->list); + list_del_rcu(&op->list); bcm_remove_op(op); return err; } -- 2.47.2

3 months, 3 weeks

1
0
0 0

[PATCH net 2/3] can: bcm: add locking for bcm_op runtime updates

by Marc Kleine-Budde

From: Oliver Hartkopp <socketcan(a)hartkopp.net> The CAN broadcast manager (CAN BCM) can send a sequence of CAN frames via hrtimer. The content and also the length of the sequence can be changed resp reduced at runtime where the 'currframe' counter is then set to zero. Although this appeared to be a safe operation the updates of 'currframe' can be triggered from user space and hrtimer context in bcm_can_tx(). Anderson Nascimento created a proof of concept that triggered a KASAN slab-out-of-bounds read access which can be prevented with a spin_lock_bh. At the rework of bcm_can_tx() the 'count' variable has been moved into the protected section as this variable can be modified from both contexts too. Fixes: ffd980f976e7 ("[CAN]: Add broadcast manager (bcm) protocol") Reported-by: Anderson Nascimento <anderson(a)allelesecurity.com> Tested-by: Anderson Nascimento <anderson(a)allelesecurity.com> Reviewed-by: Marc Kleine-Budde <mkl(a)pengutronix.de> Signed-off-by: Oliver Hartkopp <socketcan(a)hartkopp.net> Link: https://patch.msgid.link/20250519125027.11900-1-socketcan@hartkopp.net Cc: stable(a)vger.kernel.org Signed-off-by: Marc Kleine-Budde <mkl(a)pengutronix.de> --- net/can/bcm.c | 66 +++++++++++++++++++++++++++++++++++---------------- 1 file changed, 45 insertions(+), 21 deletions(-) diff --git a/net/can/bcm.c b/net/can/bcm.c index 0bca1b9b3f70..871707dab7db 100644 --- a/net/can/bcm.c +++ b/net/can/bcm.c @@ -58,6 +58,7 @@ #include <linux/can/skb.h> #include <linux/can/bcm.h> #include <linux/slab.h> +#include <linux/spinlock.h> #include <net/sock.h> #include <net/net_namespace.h> @@ -122,6 +123,7 @@ struct bcm_op { struct canfd_frame last_sframe; struct sock *sk; struct net_device *rx_reg_dev; + spinlock_t bcm_tx_lock; /* protect currframe/count in runtime updates */ }; struct bcm_sock { @@ -285,13 +287,18 @@ static void bcm_can_tx(struct bcm_op *op) { struct sk_buff *skb; struct net_device *dev; - struct canfd_frame *cf = op->frames + op->cfsiz * op->currframe; + struct canfd_frame *cf; int err; /* no target device? => exit */ if (!op->ifindex) return; + /* read currframe under lock protection */ + spin_lock_bh(&op->bcm_tx_lock); + cf = op->frames + op->cfsiz * op->currframe; + spin_unlock_bh(&op->bcm_tx_lock); + dev = dev_get_by_index(sock_net(op->sk), op->ifindex); if (!dev) { /* RFC: should this bcm_op remove itself here? */ @@ -312,6 +319,10 @@ static void bcm_can_tx(struct bcm_op *op) skb->dev = dev; can_skb_set_owner(skb, op->sk); err = can_send(skb, 1); + + /* update currframe and count under lock protection */ + spin_lock_bh(&op->bcm_tx_lock); + if (!err) op->frames_abs++; @@ -320,6 +331,11 @@ static void bcm_can_tx(struct bcm_op *op) /* reached last frame? */ if (op->currframe >= op->nframes) op->currframe = 0; + + if (op->count > 0) + op->count--; + + spin_unlock_bh(&op->bcm_tx_lock); out: dev_put(dev); } @@ -430,7 +446,7 @@ static enum hrtimer_restart bcm_tx_timeout_handler(struct hrtimer *hrtimer) struct bcm_msg_head msg_head; if (op->kt_ival1 && (op->count > 0)) { - op->count--; + bcm_can_tx(op); if (!op->count && (op->flags & TX_COUNTEVT)) { /* create notification to user */ @@ -445,7 +461,6 @@ static enum hrtimer_restart bcm_tx_timeout_handler(struct hrtimer *hrtimer) bcm_send_to_user(op, &msg_head, NULL, 0); } - bcm_can_tx(op); } else if (op->kt_ival2) { bcm_can_tx(op); @@ -956,6 +971,27 @@ static int bcm_tx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg, } op->flags = msg_head->flags; + /* only lock for unlikely count/nframes/currframe changes */ + if (op->nframes != msg_head->nframes || + op->flags & TX_RESET_MULTI_IDX || + op->flags & SETTIMER) { + + spin_lock_bh(&op->bcm_tx_lock); + + if (op->nframes != msg_head->nframes || + op->flags & TX_RESET_MULTI_IDX) { + /* potentially update changed nframes */ + op->nframes = msg_head->nframes; + /* restart multiple frame transmission */ + op->currframe = 0; + } + + if (op->flags & SETTIMER) + op->count = msg_head->count; + + spin_unlock_bh(&op->bcm_tx_lock); + } + } else { /* insert new BCM operation for the given can_id */ @@ -963,9 +999,14 @@ static int bcm_tx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg, if (!op) return -ENOMEM; + spin_lock_init(&op->bcm_tx_lock); op->can_id = msg_head->can_id; op->cfsiz = CFSIZ(msg_head->flags); op->flags = msg_head->flags; + op->nframes = msg_head->nframes; + + if (op->flags & SETTIMER) + op->count = msg_head->count; /* create array for CAN frames and copy the data */ if (msg_head->nframes > 1) { @@ -1023,22 +1064,8 @@ static int bcm_tx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg, } /* if ((op = bcm_find_op(&bo->tx_ops, msg_head->can_id, ifindex))) */ - if (op->nframes != msg_head->nframes) { - op->nframes = msg_head->nframes; - /* start multiple frame transmission with index 0 */ - op->currframe = 0; - } - - /* check flags */ - - if (op->flags & TX_RESET_MULTI_IDX) { - /* start multiple frame transmission with index 0 */ - op->currframe = 0; - } - if (op->flags & SETTIMER) { /* set timer values */ - op->count = msg_head->count; op->ival1 = msg_head->ival1; op->ival2 = msg_head->ival2; op->kt_ival1 = bcm_timeval_to_ktime(msg_head->ival1); @@ -1055,11 +1082,8 @@ static int bcm_tx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg, op->flags |= TX_ANNOUNCE; } - if (op->flags & TX_ANNOUNCE) { + if (op->flags & TX_ANNOUNCE) bcm_can_tx(op); - if (op->count) - op->count--; - } if (op->flags & STARTTIMER) bcm_tx_start_timer(op); -- 2.47.2

3 months, 3 weeks

1
0
0 0

Request for backporting accel/ivpu MMU IRQ patches to 6.14

by Jacek Lawrynowicz

Hi, Please cherry-pick following 6 patches to 6.14: bc3e5f48b7ee021371dc37297678f7089be6ce28 accel/ivpu: Use workqueue for IRQ handling 0240fa18d247c99a1967f2fed025296a89a1c5f5 accel/ivpu: Dump only first MMU fault from single context 4480912f3f8b8a1fbb5ae12c5c547fd094ec4197 accel/ivpu: Move parts of MMU event IRQ handling to thread handler 353b8f48390d36b39276ff6af61464ec64cd4d5c accel/ivpu: Fix missing MMU events from reserved SSID 2f5bbea1807a064a1e4c1b385c8cea4f37bb4b17 accel/ivpu: Fix missing MMU events if file_priv is unbound 683e9fa1c885a0cffbc10b459a7eee9df92af1c1 accel/ivpu: Flush pending jobs of device's workqueues These are fixing an issue where host can be overloaded with MMU faults from NPU causing other IRQs to be missed and host to be slowed down significantly. They should apply without conflicts. Thanks, Jacek

3 months, 3 weeks

2
1
0 0

Re: Patch "sched_ext: Fix missing rq lock in scx_bpf_cpuperf_set()" has been added to the 6.14-stable tree

by Andrea Righi

Hi, On Sun, May 18, 2025 at 06:35:28AM -0400, Sasha Levin wrote: > This is a note to let you know that I've just added the patch titled > > sched_ext: Fix missing rq lock in scx_bpf_cpuperf_set() > > to the 6.14-stable tree which can be found at: > http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… > > The filename of the patch is: > sched_ext-fix-missing-rq-lock-in-scx_bpf_cpuperf_set.patch > and it can be found in the queue-6.14 subdirectory. > > If you, or anyone else, feels it should not be added to the stable tree, > please let <stable(a)vger.kernel.org> know about it. This requires upstream commit 18853ba782bef ("sched_ext: Track currently locked rq"). Thanks, -Andrea > > > > commit e0dd90f92931fd4040aee0bf75b348a402464821 > Author: Andrea Righi <arighi(a)nvidia.com> > Date: Tue Apr 22 10:26:33 2025 +0200 > > sched_ext: Fix missing rq lock in scx_bpf_cpuperf_set() > > [ Upstream commit a11d6784d7316a6c77ca9f14fb1a698ebbb3c1fb ] > > scx_bpf_cpuperf_set() can be used to set a performance target level on > any CPU. However, it doesn't correctly acquire the corresponding rq > lock, which may lead to unsafe behavior and trigger the following > warning, due to the lockdep_assert_rq_held() check: > > [ 51.713737] WARNING: CPU: 3 PID: 3899 at kernel/sched/sched.h:1512 scx_bpf_cpuperf_set+0x1a0/0x1e0 > ... > [ 51.713836] Call trace: > [ 51.713837] scx_bpf_cpuperf_set+0x1a0/0x1e0 (P) > [ 51.713839] bpf_prog_62d35beb9301601f_bpfland_init+0x168/0x440 > [ 51.713841] bpf__sched_ext_ops_init+0x54/0x8c > [ 51.713843] scx_ops_enable.constprop.0+0x2c0/0x10f0 > [ 51.713845] bpf_scx_reg+0x18/0x30 > [ 51.713847] bpf_struct_ops_link_create+0x154/0x1b0 > [ 51.713849] __sys_bpf+0x1934/0x22a0 > > Fix by properly acquiring the rq lock when possible or raising an error > if we try to operate on a CPU that is not the one currently locked. > > Fixes: d86adb4fc0655 ("sched_ext: Add cpuperf support") > Signed-off-by: Andrea Righi <arighi(a)nvidia.com> > Acked-by: Changwoo Min <changwoo(a)igalia.com> > Signed-off-by: Tejun Heo <tj(a)kernel.org> > Signed-off-by: Sasha Levin <sashal(a)kernel.org> > > diff --git a/kernel/sched/ext.c b/kernel/sched/ext.c > index 77cdff0d9f348..0067f540a3f0f 100644 > --- a/kernel/sched/ext.c > +++ b/kernel/sched/ext.c > @@ -7459,13 +7459,32 @@ __bpf_kfunc void scx_bpf_cpuperf_set(s32 cpu, u32 perf) > } > > if (ops_cpu_valid(cpu, NULL)) { > - struct rq *rq = cpu_rq(cpu); > + struct rq *rq = cpu_rq(cpu), *locked_rq = scx_locked_rq(); > + struct rq_flags rf; > + > + /* > + * When called with an rq lock held, restrict the operation > + * to the corresponding CPU to prevent ABBA deadlocks. > + */ > + if (locked_rq && rq != locked_rq) { > + scx_ops_error("Invalid target CPU %d", cpu); > + return; > + } > + > + /* > + * If no rq lock is held, allow to operate on any CPU by > + * acquiring the corresponding rq lock. > + */ > + if (!locked_rq) { > + rq_lock_irqsave(rq, &rf); > + update_rq_clock(rq); > + } > > rq->scx.cpuperf_target = perf; > + cpufreq_update_util(rq, 0); > > - rcu_read_lock_sched_notrace(); > - cpufreq_update_util(cpu_rq(cpu), 0); > - rcu_read_unlock_sched_notrace(); > + if (!locked_rq) > + rq_unlock_irqrestore(rq, &rf); > } > } >

3 months, 3 weeks

2
1
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror May 2025