5.15.5-rc1 review
Subsystems (11)
Greg Kroah-Hartman
maintainerPATCH 145/279 net/mlx5: E-Switch, Fix resetting of encap mode when entering switchdev
In reply to Greg Kroah-HartmanFrom: Paul Blakey <[email protected]>
[ Upstream commit d7751d6476185ff754b9dad2cba0c0a6e43ecadc ]
E-Switch encap mode is relevant only when in switchdev mode. The RDMA driver can query the encap configuration via mlx5_eswitch_get_encap_mode(). Make sure it returns the currently used mode and not the set one.
This reverts the cited commit which reset the encap mode on entering switchdev and fixes the original issue properly.
--- drivers/net/ethernet/mellanox/mlx5/core/eswitch.c | 9 +++++++-- .../net/ethernet/mellanox/mlx5/core/eswitch_offloads.c | 7 ------- include/linux/mlx5/eswitch.h | 4 ++-- 3 files changed, 9 insertions(+), 11 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/eswitch.c +7 -2
@@ -1572,6 +1572,11 @@ int mlx5_eswitch_init(struct mlx5_core_dev *dev) esw->enabled_vports =0 ; esw->mode =MLX5_ESWITCH_NONE ; esw->offloads.inline_mode =MLX5_INLINE_MODE_NONE ; +if (MLX5_CAP_ESW_FLOWTABLE_FDB (dev, reformat) && +MLX5_CAP_ESW_FLOWTABLE_FDB (dev, decap)) + esw->offloads.encap =DEVLINK_ESWITCH_ENCAP_MODE_BASIC ; +else + esw->offloads.encap =DEVLINK_ESWITCH_ENCAP_MODE_NONE ; dev->priv.eswitch = esw;BLOCKING_INIT_NOTIFIER_HEAD (&esw->n_head);
@@ -1934,7 +1939,7 @@ free_out:return err; } -u8 mlx5_eswitch_mode (struct mlx5_core_dev *dev) +u8 mlx5_eswitch_mode (const struct mlx5_core_dev *dev) {struct mlx5_eswitch *esw = dev->priv.eswitch;
@@ -1948,7 +1953,7 @@ mlx5_eswitch_get_encap_mode(const struct mlx5_core_dev *dev)struct mlx5_eswitch *esw; esw = dev->priv.eswitch; -return mlx5_esw_allowed (esw) ? esw->offloads.encap : +return (mlx5_eswitch_mode (dev) ==MLX5_ESWITCH_OFFLOADS ) ? esw->offloads.encap :DEVLINK_ESWITCH_ENCAP_MODE_NONE ; }EXPORT_SYMBOL (mlx5_eswitch_get_encap_mode);
drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c +0 -7
@@ -3141,12 +3141,6 @@ int esw_offloads_enable(struct mlx5_eswitch *esw)u64 mapping_id;int err; -if (MLX5_CAP_ESW_FLOWTABLE_FDB (esw->dev, reformat) && -MLX5_CAP_ESW_FLOWTABLE_FDB (esw->dev, decap)) - esw->offloads.encap =DEVLINK_ESWITCH_ENCAP_MODE_BASIC ; -else - esw->offloads.encap =DEVLINK_ESWITCH_ENCAP_MODE_NONE ; -mutex_init (&esw->offloads.termtbl_mutex);mlx5_rdma_enable_roce (esw->dev);
@@ -3244,7 +3238,6 @@ void esw_offloads_disable(struct mlx5_eswitch *esw)esw_offloads_metadata_uninit (esw);mlx5_rdma_disable_roce (esw->dev);mutex_destroy (&esw->offloads.termtbl_mutex); - esw->offloads.encap =DEVLINK_ESWITCH_ENCAP_MODE_NONE ; }static int esw_mode_from_devlink (u16 mode,u16 *mlx5_mode)
include/linux/mlx5/eswitch.h +2 -2
@@ -136,13 +136,13 @@ u32 mlx5_eswitch_get_vport_metadata_for_set(struct mlx5_eswitch *esw, ESW_TUN_OPTS_SLOW_TABLE_GOTO_VPORT)#define ESW_TUN_SLOW_TABLE_GOTO_VPORT_MARK ESW_TUN_OPTS_MASK -u8 mlx5_eswitch_mode (struct mlx5_core_dev *dev); +u8 mlx5_eswitch_mode (const struct mlx5_core_dev *dev);u16 mlx5_eswitch_get_total_vports (const struct mlx5_core_dev *dev);struct mlx5_core_dev *mlx5_eswitch_get_core_dev (struct mlx5_eswitch *esw);#else /* CONFIG_MLX5_ESWITCH */ -static inline u8 mlx5_eswitch_mode (struct mlx5_core_dev *dev) +static inline u8 mlx5_eswitch_mode (const struct mlx5_core_dev *dev) {return MLX5_ESWITCH_NONE ; }
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 146/279 net/mlx5e: nullify cq->dbg pointer in mlx5_debug_cq_remove()
In reply to Greg Kroah-HartmanFrom: Valentine Fatiev <[email protected]>
[ Upstream commit 76ded29d3fcda4928da8849ffc446ea46871c1c2 ]
Prior to this patch in case mlx5_core_destroy_cq() failed it proceeds to rest of destroy operations. mlx5_core_destroy_cq() could be called again by user and cause additional call of mlx5_debug_cq_remove(). cq->dbg was not nullify in previous call and cause the crash.
Fix it by nullify cq->dbg pointer after removal.
Also proceed to destroy operations only if FW return 0 for MLX5_CMD_OP_DESTROY_CQ command.
general protection fault, probably for non-canonical address 0x2000300004058: 0000 [#1] SMP PTI CPU: 5 PID: 1228 Comm: python Not tainted 5.15.0-rc5_for_upstream_min_debug_2021_10_14_11_06 #1 Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.13.0-0-gf21b5a4aeb02-prebuilt.qemu.org 04/01/2014 RIP: 0010:lockref_get+0x1/0x60 Code: 5d e9 53 ff ff ff 48 8d 7f 70 e8 0a 2e 48 00 c7 85 d0 00 00 00 02 00 00 00 c6 45 70 00 fb 5d c3 c3 cc cc cc cc cc cc cc cc 53 <48> 8b 17 48 89 fb 85 d2 75 3d 48 89 d0 bf 64 00 00 00 48 89 c1 48 RSP: 0018:ffff888137dd7a38 EFLAGS: 00010206 RAX: 0000000000000000 RBX: ffff888107d5f458 RCX: 00000000fffffffe RDX: 000000000002c2b0 RSI: ffffffff8155e2e0 RDI: 0002000300004058 RBP: ffff888137dd7a88 R08: 0002000300004058 R09: ffff8881144a9f88 R10: 0000000000000000 R11: 0000000000000000 R12: ffff8881141d4000 R13: ffff888137dd7c68 R14: ffff888137dd7d58 R15: ffff888137dd7cc0 FS: 00007f4644f2a4c0(0000) GS:ffff8887a2d40000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000055b4500f4380 CR3: 0000000114f7a003 CR4: 0000000000170ea0 Call Trace: simple_recursive_removal+0x33/0x2e0 ? debugfs_remove+0x60/0x60 debugfs_remove+0x40/0x60 mlx5_debug_cq_remove+0x32/0x70 [mlx5_core] mlx5_core_destroy_cq+0x41/0x1d0 [mlx5_core] devx_obj_cleanup+0x151/0x330 [mlx5_ib] ? __pollwait+0xd0/0xd0 ? xas_load+0x5/0x70 ? xa_load+0x62/0xa0 destroy_hw_idr_uobject+0x20/0x80 [ib_uverbs] uverbs_destroy_uobject+0x3b/0x360 [ib_uverbs] uobj_destroy+0x54/0xa0 [ib_uverbs] ib_uverbs_cmd_verbs+0xaf2/0x1160 [ib_uverbs] ? uverbs_finalize_object+0xd0/0xd0 [ib_uverbs] ib_uverbs_ioctl+0xc4/0x1b0 [ib_uverbs] __x64_sys_ioctl+0x3e4/0x8e0
--- drivers/net/ethernet/mellanox/mlx5/core/cq.c | 5 +++-- drivers/net/ethernet/mellanox/mlx5/core/debugfs.c | 4 +++- 2 files changed, 6 insertions(+), 3 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/cq.c +3 -2
@@ -164,13 +164,14 @@ int mlx5_core_destroy_cq(struct mlx5_core_dev *dev, struct mlx5_core_cq *cq)MLX5_SET (destroy_cq_in, in, cqn, cq->cqn);MLX5_SET (destroy_cq_in, in, uid, cq->uid); err =mlx5_cmd_exec_in (dev, destroy_cq, in); +if (err) +return err;synchronize_irq (cq->irqn); -mlx5_cq_put (cq);wait_for_completion (&cq->free); -return err; +return 0 ; }EXPORT_SYMBOL (mlx5_core_destroy_cq);
drivers/net/ethernet/mellanox/mlx5/core/debugfs.c +3 -1
@@ -507,6 +507,8 @@ void mlx5_debug_cq_remove(struct mlx5_core_dev *dev, struct mlx5_core_cq *cq)if (!mlx5_debugfs_root)return ; -if (cq->dbg) +if (cq->dbg) {rem_res_tree (cq->dbg); + cq->dbg =NULL ; + } }
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 147/279 net/mlx5: Update error handler for UCTX and UMEM
In reply to Greg Kroah-HartmanFrom: Neta Ostrovsky <[email protected]>
[ Upstream commit ba50cd9451f6c49cf0841c0a4a146ff6a2822699 ]
In the fast unload flow, the device state is set to internal error, which indicates that the driver started the destroy process. In this case, when a destroy command is being executed, it should return MLX5_CMD_STAT_OK.
Fix MLX5_CMD_OP_DESTROY_UCTX and MLX5_CMD_OP_DESTROY_UMEM to return OK instead of EIO.
This fixes a call trace in the umem release process -
[ 2633.536695] Call Trace: [ 2633.537518] ib_uverbs_remove_one+0xc3/0x140 [ib_uverbs] [ 2633.538596] remove_client_context+0x8b/0xd0 [ib_core] [ 2633.539641] disable_device+0x8c/0x130 [ib_core] [ 2633.540615] __ib_unregister_device+0x35/0xa0 [ib_core] [ 2633.541640] ib_unregister_device+0x21/0x30 [ib_core] [ 2633.542663] __mlx5_ib_remove+0x38/0x90 [mlx5_ib] [ 2633.543640] auxiliary_bus_remove+0x1e/0x30 [auxiliary] [ 2633.544661] device_release_driver_internal+0x103/0x1f0 [ 2633.545679] bus_remove_device+0xf7/0x170 [ 2633.546640] device_del+0x181/0x410 [ 2633.547606] mlx5_rescan_drivers_locked.part.10+0x63/0x160 [mlx5_core] [ 2633.548777] mlx5_unregister_device+0x27/0x40 [mlx5_core] [ 2633.549841] mlx5_uninit_one+0x21/0xc0 [mlx5_core] [ 2633.550864] remove_one+0x69/0xe0 [mlx5_core] [ 2633.551819] pci_device_remove+0x3b/0xc0 [ 2633.552731] device_release_driver_internal+0x103/0x1f0 [ 2633.553746] unbind_store+0xf6/0x130 [ 2633.554657] kernfs_fop_write+0x116/0x190 [ 2633.555567] vfs_write+0xa5/0x1a0 [ 2633.556407] ksys_write+0x4f/0xb0 [ 2633.557233] do_syscall_64+0x5b/0x1a0 [ 2633.558071] entry_SYSCALL_64_after_hwframe+0x65/0xca [ 2633.559018] RIP: 0033:0x7f9977132648 [ 2633.559821] Code: 89 02 48 c7 c0 ff ff ff ff eb b3 0f 1f 80 00 00 00 00 f3 0f 1e fa 48 8d 05 55 6f 2d 00 8b 00 85 c0 75 17 b8 01 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 58 c3 0f 1f 80 00 00 00 00 41 54 49 89 d4 55 [ 2633.562332] RSP: 002b:00007fffb1a83888 EFLAGS: 00000246 ORIG_RAX: 0000000000000001 [ 2633.563472] RAX: ffffffffffffffda RBX: 000000000000000c RCX: 00007f9977132648 [ 2633.564541] RDX: 000000000000000c RSI: 000055b90546e230 RDI: 0000000000000001 [ 2633.565596] RBP: 000055b90546e230 R08: 00007f9977406860 R09: 00007f9977a54740 [ 2633.566653] R10: 0000000000000000 R11: 0000000000000246 R12: 00007f99774056e0 [ 2633.567692] R13: 000000000000000c R14: 00007f9977400880 R15: 000000000000000c [ 2633.568725] ---[ end trace 10b4fe52945e544d ]---
--- drivers/net/ethernet/mellanox/mlx5/core/cmd.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/cmd.c +2 -2
@@ -334,6 +334,8 @@ static int mlx5_internal_err_ret_value(struct mlx5_core_dev *dev, u16 op,case MLX5_CMD_OP_PAGE_FAULT_RESUME :case MLX5_CMD_OP_QUERY_ESW_FUNCTIONS :case MLX5_CMD_OP_DEALLOC_SF : +case MLX5_CMD_OP_DESTROY_UCTX : +case MLX5_CMD_OP_DESTROY_UMEM :return MLX5_CMD_STAT_OK ;case MLX5_CMD_OP_QUERY_HCA_CAP :
@@ -459,9 +461,7 @@ static int mlx5_internal_err_ret_value(struct mlx5_core_dev *dev, u16 op,case MLX5_CMD_OP_MODIFY_GENERAL_OBJECT :case MLX5_CMD_OP_QUERY_GENERAL_OBJECT :case MLX5_CMD_OP_CREATE_UCTX : -case MLX5_CMD_OP_DESTROY_UCTX :case MLX5_CMD_OP_CREATE_UMEM : -case MLX5_CMD_OP_DESTROY_UMEM :case MLX5_CMD_OP_ALLOC_MEMIC :case MLX5_CMD_OP_MODIFY_XRQ :case MLX5_CMD_OP_RELEASE_XRQ_ERROR :
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 148/279 net/mlx5: E-Switch, rebuild lag only when needed
In reply to Greg Kroah-HartmanFrom: Mark Bloch <[email protected]>
[ Upstream commit 2eb0cb31bc4ce2ede5460cf3ef433b40cf5f040d ]
A user can enable VFs without changing E-Switch mode, this can happen when a user moves straight to switchdev mode and only once in switchdev VFs are enabled via the sysfs interface.
The cited commit assumed this isn't possible and exposed a single API function where the E-switch calls into the lag code, breaks the lag and prevents any other lag operations to take place until the E-switch update has ended.
Breaking the hardware lag when it isn't needed can make it such that hardware lag can't be enabled again.
In the sysfs call path check if the current E-Switch mode is NONE, in the context of the function it can only mean the E-Switch is moving out of NONE mode and the hardware lag should be disabled and enabled once the mode change has ended. If the mode isn't NONE it means VFs are about to be enabled and such operation doesn't require toggling the hardware lag.
--- drivers/net/ethernet/mellanox/mlx5/core/eswitch.c | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/eswitch.c +10 -2
@@ -1305,12 +1305,17 @@ abort:*/ int mlx5_eswitch_enable (struct mlx5_eswitch *esw,int num_vfs) { +bool toggle_lag;int ret;if (!mlx5_esw_allowed (esw))return 0 ; -mlx5_lag_disable_change (esw->dev); + toggle_lag = esw->mode ==MLX5_ESWITCH_NONE ; + +if (toggle_lag) +mlx5_lag_disable_change (esw->dev); +down_write (&esw->mode_lock);if (esw->mode ==MLX5_ESWITCH_NONE ) { ret =mlx5_eswitch_enable_locked (esw,MLX5_ESWITCH_LEGACY , num_vfs);
@@ -1324,7 +1329,10 @@ int mlx5_eswitch_enable(struct mlx5_eswitch *esw, int num_vfs) esw->esw_funcs.num_vfs = num_vfs; }up_write (&esw->mode_lock); -mlx5_lag_enable_change (esw->dev); + +if (toggle_lag) +mlx5_lag_enable_change (esw->dev); +return ret; }
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 149/279 net/mlx5e: CT, Fix multiple allocations and memleak of mod acts
In reply to Greg Kroah-HartmanFrom: Roi Dayan <[email protected]>
[ Upstream commit 806401c20a0f9c51b6c8fd7035671e6ca841f6c2 ]
CT clear action offload adds additional mod hdr actions to the flow's original mod actions in order to clear the registers which hold ct_state.
When such flow also includes encap action, a neigh update event can cause the driver to unoffload the flow and then reoffload it.
Each time this happens, the ct clear handling adds that same set of mod hdr actions to reset ct_state until the max of mod hdr actions is reached.
Also the driver never releases the allocated mod hdr actions and causing a memleak.
Fix above two issues by moving CT clear mod acts allocation into the parsing actions phase and only use it when offloading the rule. The release of mod acts will be done in the normal flow_put().
backtrace:
[<000000007316e2f3>] krealloc+0x83/0xd0
[<00000000ef157de1>] mlx5e_mod_hdr_alloc+0x147/0x300 [mlx5_core]
[<00000000970ce4ae>] mlx5e_tc_match_to_reg_set_and_get_id+0xd7/0x240 [mlx5_core]
[<0000000067c5fa17>] mlx5e_tc_match_to_reg_set+0xa/0x20 [mlx5_core]
[<00000000d032eb98>] mlx5_tc_ct_entry_set_registers.isra.0+0x36/0xc0 [mlx5_core]
[<00000000fd23b869>] mlx5_tc_ct_flow_offload+0x272/0x1f10 [mlx5_core]
[<000000004fc24acc>] mlx5e_tc_offload_fdb_rules.part.0+0x150/0x620 [mlx5_core]
[<00000000dc741c17>] mlx5e_tc_encap_flows_add+0x489/0x690 [mlx5_core]
[<00000000e92e49d7>] mlx5e_rep_update_flows+0x6e4/0x9b0 [mlx5_core]
[<00000000f60f5602>] mlx5e_rep_neigh_update+0x39a/0x5d0 [mlx5_core]--- .../ethernet/mellanox/mlx5/core/en/tc_ct.c | 26 ++++++++++++------- .../ethernet/mellanox/mlx5/core/en/tc_ct.h | 2 ++ .../net/ethernet/mellanox/mlx5/core/en_tc.c | 8 ++++-- 3 files changed, 25 insertions(+), 11 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.c +17 -9
@@ -1356,9 +1356,13 @@ mlx5_tc_ct_match_add(struct mlx5_tc_ct_priv *priv,int mlx5_tc_ct_parse_action (struct mlx5_tc_ct_priv *priv,struct mlx5_flow_attr *attr, +struct mlx5e_tc_mod_hdr_acts *mod_acts,const struct flow_action_entry *act,struct netlink_ext_ack *extack) { +bool clear_action = act->ct.action &TCA_CT_ACT_CLEAR ; +int err; +if (!priv) {NL_SET_ERR_MSG_MOD (extack,"offload of ct action isn't available" );
@@ -1369,6 +1373,17 @@ mlx5_tc_ct_parse_action(struct mlx5_tc_ct_priv *priv, attr->ct_attr.ct_action = act->ct.action; attr->ct_attr.nf_ft = act->ct.flow_table; +if (!clear_action) + goto out; + + err =mlx5_tc_ct_entry_set_registers (priv, mod_acts,0 ,0 ,0 ,0 ); +if (err) { +NL_SET_ERR_MSG_MOD (extack,"Failed to set registers for ct clear" ); +return err; + } + attr->action |=MLX5_FLOW_CONTEXT_ACTION_MOD_HDR ; + +out:return 0 ; }
@@ -1898,23 +1913,16 @@ __mlx5_tc_ct_flow_offload_clear(struct mlx5_tc_ct_priv *ct_priv,memcpy (pre_ct_attr, attr, attr_sz); - err =mlx5_tc_ct_entry_set_registers (ct_priv, mod_acts,0 ,0 ,0 ,0 ); -if (err) { -ct_dbg ("Failed to set register for ct clear" ); - goto err_set_registers; - } - mod_hdr =mlx5_modify_header_alloc (priv->mdev, ct_priv->ns_type, mod_acts->num_actions, mod_acts->actions);if (IS_ERR (mod_hdr)) { err =PTR_ERR (mod_hdr);ct_dbg ("Failed to add create ct clear mod hdr" ); - goto err_set_registers; + goto err_mod_hdr; } pre_ct_attr->modify_hdr = mod_hdr; - pre_ct_attr->action |=MLX5_FLOW_CONTEXT_ACTION_MOD_HDR ; rule =mlx5_tc_rule_insert (priv, orig_spec, pre_ct_attr);if (IS_ERR (rule)) {
@@ -1930,7 +1938,7 @@ __mlx5_tc_ct_flow_offload_clear(struct mlx5_tc_ct_priv *ct_priv, err_insert:mlx5_modify_header_dealloc (priv->mdev, mod_hdr); -err_set_registers: +err_mod_hdr:netdev_warn (priv->netdev,"Failed to offload ct clear flow, err %d\n" , err);kfree (pre_ct_attr);
drivers/net/ethernet/mellanox/mlx5/core/en/tc_ct.h +2 -0
@@ -110,6 +110,7 @@ int mlx5_tc_ct_add_no_trk_match(struct mlx5_flow_spec *spec);int mlx5_tc_ct_parse_action (struct mlx5_tc_ct_priv *priv,struct mlx5_flow_attr *attr, +struct mlx5e_tc_mod_hdr_acts *mod_acts,const struct flow_action_entry *act,struct netlink_ext_ack *extack);
@@ -172,6 +173,7 @@ mlx5_tc_ct_add_no_trk_match(struct mlx5_flow_spec *spec)static inline int mlx5_tc_ct_parse_action (struct mlx5_tc_ct_priv *priv,struct mlx5_flow_attr *attr, +struct mlx5e_tc_mod_hdr_acts *mod_acts,const struct flow_action_entry *act,struct netlink_ext_ack *extack) {
drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +6 -2
@@ -3458,7 +3458,9 @@ static int parse_tc_nic_actions(struct mlx5e_priv *priv, attr->dest_chain = act->chain_index;break ;case FLOW_ACTION_CT : - err =mlx5_tc_ct_parse_action (get_ct_priv (priv), attr, act, extack); + err =mlx5_tc_ct_parse_action (get_ct_priv (priv), attr, + &parse_attr->mod_hdr_acts, + act, extack);if (err)return err;
@@ -4009,7 +4011,9 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv,NL_SET_ERR_MSG_MOD (extack,"Sample action with connection tracking is not supported" );return -EOPNOTSUPP ; } - err =mlx5_tc_ct_parse_action (get_ct_priv (priv), attr, act, extack); + err =mlx5_tc_ct_parse_action (get_ct_priv (priv), attr, + &parse_attr->mod_hdr_acts, + act, extack);if (err)return err;
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 150/279 net/mlx5: Lag, update tracker when state change event received
In reply to Greg Kroah-HartmanFrom: Maher Sanalla <[email protected]>
[ Upstream commit ae396d85c01c7bdc9eeceecde1f493d03f793465 ]
Currently, In NETDEV_CHANGELOWERSTATE/NETDEV_CHANGEUPPERSTATE events handling, tracking is not fully completed if the LAG device is not ready at the time the events occur. But, we must keep track of the upper and lower states after receiving the events because RoCE needs this info in mlx5_lag_get_roce_netdev() - in order to return the corresponding port that its running on. Returning the wrong (not most recent) port will lead to gids table being incorrect.
For example: If during the attachment of a slave to the bond, the other non-attached port performs pci_reload, then the LAG device is not ready, but that should not result in dismissing attached slave tracker update automatically (which is performed in mlx5_handle_changelowerstate()), Since these events might not come later, which can lead to both bond ports having tx_enabled=0 - which is not a valid state of LAG bond.
--- drivers/net/ethernet/mellanox/mlx5/core/lag.c | 28 +++++++++---------- 1 file changed, 13 insertions(+), 15 deletions(-)
drivers/net/ethernet/mellanox/mlx5/core/lag.c +13 -15
@@ -565,6 +565,7 @@ static int mlx5_handle_changeupper_event(struct mlx5_lag *ldev,bool is_bonded, is_in_lag, mode_supported;int bond_status =0 ;int num_slaves =0 ; +int changed =0 ;int idx;if (!netif_is_lag_master (upper))
@@ -601,27 +602,27 @@ static int mlx5_handle_changeupper_event(struct mlx5_lag *ldev,*/ is_in_lag = num_slaves ==MLX5_MAX_PORTS && bond_status ==0x3 ; -if (!mlx5_lag_is_ready (ldev) && is_in_lag) { -NL_SET_ERR_MSG_MOD (info->info.extack, -"Can't activate LAG offload, PF is configured with more than 64 VFs" ); -return 0 ; - } -/* Lag mode must be activebackup or hash. */ mode_supported = tracker->tx_type ==NETDEV_LAG_TX_TYPE_ACTIVEBACKUP || tracker->tx_type ==NETDEV_LAG_TX_TYPE_HASH ; -if (is_in_lag && !mode_supported) -NL_SET_ERR_MSG_MOD (info->info.extack, -"Can't activate LAG offload, TX type isn't supported" ); - is_bonded = is_in_lag && mode_supported;if (tracker->is_bonded != is_bonded) { tracker->is_bonded = is_bonded; -return 1 ; + changed =1 ; } -return 0 ; +if (!is_in_lag) +return changed; + +if (!mlx5_lag_is_ready (ldev)) +NL_SET_ERR_MSG_MOD (info->info.extack, +"Can't activate LAG offload, PF is configured with more than 64 VFs" ); +else if (!mode_supported) +NL_SET_ERR_MSG_MOD (info->info.extack, +"Can't activate LAG offload, TX type isn't supported" ); + +return changed; } static int mlx5_handle_changelowerstate_event(struct mlx5_lag *ldev,
@@ -664,9 +665,6 @@ static int mlx5_lag_netdev_event(struct notifier_block *this, ldev = container_of(this, struct mlx5_lag, nb); -if (!mlx5_lag_is_ready (ldev) && event ==NETDEV_CHANGELOWERSTATE ) -return NOTIFY_DONE ; - tracker = ldev->tracker;switch (event) {
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 151/279 net/mlx5: E-Switch, return error if encap isnt supported
In reply to Greg Kroah-HartmanFrom: Raed Salem <[email protected]>
[ Upstream commit c4c3176739dfa6efcc5b1d1de4b3fd2b51b048c7 ]
On regular ConnectX HCAs getting encap mode isn't supported when the E-Switch is in NONE mode. Current code would return no error code when trying to get encap mode in such case which is wrong.
Fix by returning error value to indicate failure to caller in such case.
--- drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c +1 -1
@@ -3581,7 +3581,7 @@ int mlx5_devlink_eswitch_encap_mode_get(struct devlink *devlink, *encap = esw->offloads.encap; unlock:up_write (&esw->mode_lock); -return 0 ; +return err; } static bool
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 152/279 scsi: ufs: core: Improve SCSI abort handling
In reply to Greg Kroah-HartmanFrom: Bart Van Assche <[email protected]>
[ Upstream commit 3ff1f6b6ba6f97f50862aa50e79959cc8ddc2566 ]
The following has been observed on a test setup:
WARNING: CPU: 4 PID: 250 at drivers/scsi/ufs/ufshcd.c:2737 ufshcd_queuecommand+0x468/0x65c Call trace: ufshcd_queuecommand+0x468/0x65c scsi_send_eh_cmnd+0x224/0x6a0 scsi_eh_test_devices+0x248/0x418 scsi_eh_ready_devs+0xc34/0xe58 scsi_error_handler+0x204/0x80c kthread+0x150/0x1b4 ret_from_fork+0x10/0x30
That warning is triggered by the following statement:
WARN_ON(lrbp->cmd);
Fix this warning by clearing lrbp->cmd from the abort handler.
--- drivers/scsi/ufs/ufshcd.c | 1 + 1 file changed, 1 insertion(+)
drivers/scsi/ufs/ufshcd.c +1 -0
@@ -7040,6 +7040,7 @@ static int ufshcd_abort(struct scsi_cmnd *cmd) goto release; } + lrbp->cmd =NULL ; err =SUCCESS ; release:
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 153/279 scsi: core: sysfs: Fix hang when device state is set via sysfs
In reply to Greg Kroah-HartmanFrom: Mike Christie <[email protected]>
[ Upstream commit 4edd8cd4e86dd3047e5294bbefcc0a08f66a430f ]
This fixes a regression added with:
commit f0f82e2476f6 ("scsi: core: Fix capacity set to zero after offlinining device")
The problem is that after iSCSI recovery, iscsid will call into the kernel to set the dev's state to running, and with that patch we now call scsi_rescan_device() with the state_mutex held. If the SCSI error handler thread is just starting to test the device in scsi_send_eh_cmnd() then it's going to try to grab the state_mutex.
We are then stuck, because when scsi_rescan_device() tries to send its I/O scsi_queue_rq() calls -> scsi_host_queue_ready() -> scsi_host_in_recovery() which will return true (the host state is still in recovery) and I/O will just be requeued. scsi_send_eh_cmnd() will then never be able to grab the state_mutex to finish error handling.
To prevent the deadlock move the rescan-related code to after we drop the state_mutex.
This also adds a check for if we are already in the running state. This prevents extra scans and helps the iscsid case where if the transport class has already onlined the device during its recovery process then we don't need userspace to do it again plus possibly block that daemon.
--- drivers/scsi/scsi_sysfs.c | 30 +++++++++++++++++++----------- 1 file changed, 19 insertions(+), 11 deletions(-)
drivers/scsi/scsi_sysfs.c +19 -11
@@ -797,6 +797,7 @@ store_state_field(struct device *dev, struct device_attribute *attr,int i, ret;struct scsi_device *sdev =to_scsi_device (dev);enum scsi_device_state state =0 ; +bool rescan_dev = false;for (i =0 ; i <ARRAY_SIZE (sdev_states); i++) {const int len =strlen (sdev_states[i].name);
@@ -815,20 +816,27 @@ store_state_field(struct device *dev, struct device_attribute *attr, }mutex_lock (&sdev->state_mutex); - ret =scsi_device_set_state (sdev, state); -/* -* If the device state changes to SDEV_RUNNING, we need to -* run the queue to avoid I/O hang, and rescan the device -* to revalidate it. Running the queue first is necessary -* because another thread may be waiting inside -* blk_mq_freeze_queue_wait() and because that call may be -* waiting for pending I/O to finish. -*/ -if (ret ==0 && state ==SDEV_RUNNING ) { +if (sdev->sdev_state ==SDEV_RUNNING && state ==SDEV_RUNNING ) { + ret = count; + }else { + ret =scsi_device_set_state (sdev, state); +if (ret ==0 && state ==SDEV_RUNNING ) + rescan_dev = true; + } +mutex_unlock (&sdev->state_mutex); + +if (rescan_dev) { +/* +* If the device state changes to SDEV_RUNNING, we need to +* run the queue to avoid I/O hang, and rescan the device +* to revalidate it. Running the queue first is necessary +* because another thread may be waiting inside +* blk_mq_freeze_queue_wait() and because that call may be +* waiting for pending I/O to finish. +*/ blk_mq_run_hw_queues (sdev->request_queue, true);scsi_rescan_device (dev); } -mutex_unlock (&sdev->state_mutex);return ret ==0 ? count : -EINVAL ; }
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 154/279 scsi: ufs: core: Fix task management completion timeout race
In reply to Greg Kroah-HartmanFrom: Adrian Hunter <[email protected]>
[ Upstream commit 886fe2915cce6658b0fc19e64b82879325de61ea ]
__ufshcd_issue_tm_cmd() clears req->end_io_data after timing out, which races with the completion function ufshcd_tmc_handler() which expects req->end_io_data to have a value.
Note __ufshcd_issue_tm_cmd() and ufshcd_tmc_handler() are already synchronized using hba->tmf_rqs and hba->outstanding_tasks under the host_lock spinlock.
It is also not necessary (nor typical) to clear req->end_io_data because the block layer does it before allocating out requests e.g. via blk_get_request().
So fix by not clearing it.
--- drivers/scsi/ufs/ufshcd.c | 5 ----- 1 file changed, 5 deletions(-)
drivers/scsi/ufs/ufshcd.c +0 -5
@@ -6541,11 +6541,6 @@ static int __ufshcd_issue_tm_cmd(struct ufs_hba *hba, err =wait_for_completion_io_timeout (&wait,msecs_to_jiffies (TM_CMD_TIMEOUT ));if (!err) { -/* -* Make sure that ufshcd_compl_tm() does not trigger a -* use-after-free. -*/ - req->end_io_data =NULL ;ufshcd_add_tm_upiu_trace (hba, task_tag,UFS_TM_ERR );dev_err (hba->dev,"%s: task management cmd 0x%.2x timed-out\n" , __func__, tm_function);
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 155/279 scsi: ufs: core: Fix another task management completion race
In reply to Greg Kroah-HartmanFrom: Adrian Hunter <[email protected]>
[ Upstream commit 5cb37a26355d79ab290220677b1b57d28e99a895 ]
hba->outstanding_tasks, which is read under host_lock spinlock, tells the interrupt handler what task management tags are in use by the driver. The doorbell register bits indicate which tags are in use by the hardware. A doorbell bit that is 0 is because the bit has yet to be set by the driver, or because the task is complete. It is only possible to disambiguate the 2 cases, if reading/writing the doorbell register is synchronized with reading/writing hba->outstanding_tasks.
For that reason, reading REG_UTP_TASK_REQ_DOOR_BELL must be done under spinlock.
--- drivers/scsi/ufs/ufshcd.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-)
drivers/scsi/ufs/ufshcd.c +1 -2
@@ -6382,9 +6382,8 @@ static irqreturn_t ufshcd_tmc_handler(struct ufs_hba *hba)irqreturn_t ret =IRQ_NONE ;int tag; - pending =ufshcd_readl (hba,REG_UTP_TASK_REQ_DOOR_BELL ); -spin_lock_irqsave (hba->host->host_lock, flags); + pending =ufshcd_readl (hba,REG_UTP_TASK_REQ_DOOR_BELL ); issued = hba->outstanding_tasks & ~pending;for_each_set_bit (tag, &issued, hba->nutmrs) {struct request *req = hba->tmf_rqs[tag];
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 156/279 net: mvmdio: fix compilation warning
In reply to Greg Kroah-HartmanFrom: Marcin Wojtas <[email protected]>
[ Upstream commit 2460386bef0b9b98b71728d3c173e15558b78d82 ]
The kernel test robot reported a following issue:
Quoted 1 line
Quoted 1 line
drivers/net/ethernet/marvell/mvmdio.c:426:36: warning:
unused variable 'orion_mdio_acpi_match' [-Wunused-const-variable]
static const struct acpi_device_id orion_mdio_acpi_match[] = {
^1 warning generated.
Fix that by surrounding the variable by appropriate ifdef.
--- drivers/net/ethernet/marvell/mvmdio.c | 2 ++ 1 file changed, 2 insertions(+)
drivers/net/ethernet/marvell/mvmdio.c +2 -0
@@ -429,12 +429,14 @@ static const struct of_device_id orion_mdio_match[] = { };MODULE_DEVICE_TABLE (of, orion_mdio_match); +#ifdef CONFIG_ACPI static const struct acpi_device_id orion_mdio_acpi_match[] = { {"MRVL0100" ,BUS_TYPE_SMI }, {"MRVL0101" ,BUS_TYPE_XSMI }, { }, };MODULE_DEVICE_TABLE (acpi, orion_mdio_acpi_match); +#endif static struct platform_driver orion_mdio_driver = { .probe = orion_mdio_probe,
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 157/279 net: sched: act_mirred: drop dst for the direction from egress to ingress
In reply to Greg Kroah-HartmanFrom: Xin Long <[email protected]>
[ Upstream commit f799ada6bf2397c351220088b9b0980125c77280 ]
Without dropping dst, the packets sent from local mirred/redirected to ingress will may still use the old dst. ip_rcv() will drop it as the old dst is for output and its .input is dst_discard.
This patch is to fix by also dropping dst for those packets that are mirred or redirected from egress to ingress in act_mirred.
Note that we don't drop it for the direction change from ingress to egress, as on which there might be a user case attaching a metadata dst by act_tunnel_key that would be used later.
--- net/sched/act_mirred.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-)
net/sched/act_mirred.c +8 -3
@@ -19,6 +19,7 @@#include <linux/if_arp.h> #include <net/net_namespace.h> #include <net/netlink.h> +#include <net/dst.h> #include <net/pkt_sched.h> #include <net/pkt_cls.h> #include <linux/tc_act/tc_mirred.h>
@@ -228,6 +229,7 @@ static int tcf_mirred_act(struct sk_buff *skb, const struct tc_action *a,bool want_ingress;bool is_redirect;bool expects_nh; +bool at_ingress;int m_eaction;int mac_len;bool at_nh;
@@ -263,7 +265,8 @@ static int tcf_mirred_act(struct sk_buff *skb, const struct tc_action *a,* ingress - that covers the TC S/W datapath. */ is_redirect =tcf_mirred_is_act_redirect (m_eaction); - use_reinsert =skb_at_tc_ingress (skb) && is_redirect && + at_ingress =skb_at_tc_ingress (skb); + use_reinsert = at_ingress && is_redirect &&tcf_mirred_can_reinsert (retval);if (!use_reinsert) { skb2 =skb_clone (skb,GFP_ATOMIC );
@@ -271,10 +274,12 @@ static int tcf_mirred_act(struct sk_buff *skb, const struct tc_action *a, goto out; } + want_ingress =tcf_mirred_act_wants_ingress (m_eaction); +/* All mirred/redirected skbs should clear previous ct info */ nf_reset_ct (skb2); - - want_ingress =tcf_mirred_act_wants_ingress (m_eaction); +if (want_ingress && !at_ingress)/* drop dst for egress -> ingress */ +skb_dst_drop (skb2); expects_nh = want_ingress || !m_mac_header_xmit; at_nh = skb->data ==skb_network_header (skb);
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 158/279 net: dpaa2-eth: fix use-after-free in dpaa2_eth_remove
In reply to Greg Kroah-HartmanFrom: Pavel Skripkin <[email protected]>
[ Upstream commit 9b5a333272a48c2f8b30add7a874e46e8b26129c ]
Access to netdev after free_netdev() will cause use-after-free bug. Move debug log before free_netdev() call to avoid it.
--- drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.c +2 -2
@@ -4538,10 +4538,10 @@ static int dpaa2_eth_remove(struct fsl_mc_device *ls_dev)fsl_mc_portal_free (priv->mc_io); -free_netdev (net_dev); -dev_dbg (net_dev->dev.parent,"Removed interface %s\n" , net_dev->name); +free_netdev (net_dev); +return 0 ; }
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 159/279 net: virtio_net_hdr_to_skb: count transport header in UFO
In reply to Greg Kroah-HartmanFrom: Jonathan Davies <[email protected]>
[ Upstream commit cf9acc90c80ecbee00334aa85d92f4e74014bcff ]
virtio_net_hdr_to_skb does not set the skb's gso_size and gso_type correctly for UFO packets received via virtio-net that are a little over the GSO size. This can lead to problems elsewhere in the networking stack, e.g. ovs_vport_send dropping over-sized packets if gso_size is not set.
This is due to the comparison
if (skb->len - p_off > gso_size)
not properly accounting for the transport layer header.
p_off includes the size of the transport layer header (thlen), so skb->len - p_off is the size of the TCP/UDP payload.
gso_size is read from the virtio-net header. For UFO, fragmentation happens at the IP level so does not need to include the UDP header.
Hence the calculation could be comparing a TCP/UDP payload length with an IP payload length, causing legitimate virtio-net packets to have lack gso_type/gso_size information.
Example: a UDP packet with payload size 1473 has IP payload size 1481. If the guest used UFO, it is not fragmented and the virtio-net header's flags indicate that it is a GSO frame (VIRTIO_NET_HDR_GSO_UDP), with gso_size = 1480 for an MTU of 1500. skb->len will be 1515 and p_off will be 42, so skb->len - p_off = 1473. Hence the comparison fails, and shinfo->gso_size and gso_type are not set as they should be.
Instead, add the UDP header length before comparing to gso_size when using UFO. In this way, it is the size of the IP payload that is compared to gso_size.
--- include/linux/virtio_net.h | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
include/linux/virtio_net.h +6 -1
@@ -120,10 +120,15 @@ retry:if (hdr->gso_type !=VIRTIO_NET_HDR_GSO_NONE ) {u16 gso_size =__virtio16_to_cpu (little_endian, hdr->gso_size); +unsigned int nh_off = p_off;struct skb_shared_info *shinfo =skb_shinfo (skb); +/* UFO may not include transport header in gso_size. */ +if (gso_type &SKB_GSO_UDP ) + nh_off -= thlen; +/* Too small packets are not really GSO ones. */ -if (skb->len - p_off > gso_size) { +if (skb->len - nh_off > gso_size) { shinfo->gso_size = gso_size; shinfo->gso_type = gso_type;
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 160/279 i40e: Fix correct max_pkt_size on VF RX queue
In reply to Greg Kroah-HartmanFrom: Eryk Rybak <[email protected]>
[ Upstream commit 6afbd7b3c53cb7417189f476e99d431daccb85b0 ]
Setting VLAN port increasing RX queue max_pkt_size by 4 bytes to take VLAN tag into account. Trigger the VF reset when setting port VLAN for VF to renegotiate its capabilities and reinitialize.
--- .../ethernet/intel/i40e/i40e_virtchnl_pf.c | 53 ++++--------------- 1 file changed, 9 insertions(+), 44 deletions(-)
drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c +9 -44
@@ -674,14 +674,13 @@ static int i40e_config_vsi_rx_queue(struct i40e_vf *vf, u16 vsi_id,u16 vsi_queue_id, struct virtchnl_rxq_info *info) { + u16 pf_queue_id = i40e_vc_get_pf_queue_id(vf, vsi_id, vsi_queue_id);struct i40e_pf *pf = vf->pf; +struct i40e_vsi *vsi = pf->vsi[vf->lan_vsi_idx];struct i40e_hw *hw = &pf->hw;struct i40e_hmc_obj_rxq rx_ctx; -u16 pf_queue_id;int ret =0 ; - pf_queue_id =i40e_vc_get_pf_queue_id (vf, vsi_id, vsi_queue_id); -/* clear the context structure first */ memset (&rx_ctx,0 ,sizeof (struct i40e_hmc_obj_rxq ));
@@ -719,6 +718,10 @@ static int i40e_config_vsi_rx_queue(struct i40e_vf *vf, u16 vsi_id, } rx_ctx.rxmax = info->max_pkt_size; +/* if port VLAN is configured increase the max packet size */ +if (vsi->info.pvid) + rx_ctx.rxmax +=VLAN_HLEN ; +/* enable 32bytes desc always */ rx_ctx.dsize =1 ;
@@ -4169,34 +4172,6 @@ error_param:return ret; } -/** -* i40e_vsi_has_vlans - True if VSI has configured VLANs -* @vsi: pointer to the vsi -* -* Check if a VSI has configured any VLANs. False if we have a port VLAN or if -* we have no configured VLANs. Do not call while holding the -* mac_filter_hash_lock. -*/ -static bool i40e_vsi_has_vlans (struct i40e_vsi *vsi) -{ -bool have_vlans; - -/* If we have a port VLAN, then the VSI cannot have any VLANs -* configured, as all MAC/VLAN filters will be assigned to the PVID. -*/ -if (vsi->info.pvid) -return false; - -/* Since we don't have a PVID, we know that if the device is in VLAN -* mode it must be because of a VLAN filter configured on this VSI. -*/ -spin_lock_bh (&vsi->mac_filter_hash_lock); - have_vlans =i40e_is_vsi_in_vlan (vsi); -spin_unlock_bh (&vsi->mac_filter_hash_lock); - -return have_vlans; -} -/** * i40e_ndo_set_vf_port_vlan * @netdev: network interface device structure
@@ -4253,19 +4228,9 @@ int i40e_ndo_set_vf_port_vlan(struct net_device *netdev, int vf_id,/* duplicate request, so just return success */ goto error_pvid; -if (i40e_vsi_has_vlans (vsi)) { -dev_err (&pf->pdev->dev, -"VF %d has already configured VLAN filters and the administrator is requesting a port VLAN override.\nPlease unload and reload the VF driver for this change to take effect.\n" , - vf_id); -/* Administrator Error - knock the VF offline until he does -* the right thing by reconfiguring his network correctly -* and then reloading the VF driver. -*/ -i40e_vc_disable_vf (vf); -/* During reset the VF got a new VSI, so refresh the pointer. */ - vsi = pf->vsi[vf->lan_vsi_idx]; - } - +i40e_vc_disable_vf (vf); +/* During reset the VF got a new VSI, so refresh a pointer. */ + vsi = pf->vsi[vf->lan_vsi_idx];/* Locked once because multiple functions below iterate list */ spin_lock_bh (&vsi->mac_filter_hash_lock);
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 161/279 i40e: Fix NULL ptr dereference on VSI filter sync
In reply to Greg Kroah-HartmanFrom: Michal Maloszewski <[email protected]>
[ Upstream commit 37d9e304acd903a445df8208b8a13d707902dea6 ]
Remove the reason of null pointer dereference in sync VSI filters. Added new I40E_VSI_RELEASING flag to signalize deleting and releasing of VSI resources to sync this thread with sync filters subtask. Without this patch it is possible to start update the VSI filter list after VSI is removed, that's causing a kernel oops.
--- drivers/net/ethernet/intel/i40e/i40e.h | 1 + drivers/net/ethernet/intel/i40e/i40e_main.c | 5 +++-- 2 files changed, 4 insertions(+), 2 deletions(-)
drivers/net/ethernet/intel/i40e/i40e.h +1 -0
@@ -161,6 +161,7 @@ enum i40e_vsi_state_t { __I40E_VSI_OVERFLOW_PROMISC, __I40E_VSI_REINIT_REQUESTED, __I40E_VSI_DOWN_REQUESTED, + __I40E_VSI_RELEASING,/* This must be last as it determines the size of the BITMAP */ __I40E_VSI_STATE_SIZE__, };
drivers/net/ethernet/intel/i40e/i40e_main.c +3 -2
@@ -2623,7 +2623,8 @@ static void i40e_sync_filters_subtask(struct i40e_pf *pf)for (v =0 ; v < pf->num_alloc_vsi; v++) {if (pf->vsi[v] && - (pf->vsi[v]->flags &I40E_VSI_FLAG_FILTER_CHANGED )) { + (pf->vsi[v]->flags &I40E_VSI_FLAG_FILTER_CHANGED ) && + !test_bit (__I40E_VSI_RELEASING, pf->vsi[v]->state)) {int ret =i40e_sync_vsi_filters (pf->vsi[v]);if (ret) {
@@ -13771,7 +13772,7 @@ int i40e_vsi_release(struct i40e_vsi *vsi)dev_info (&pf->pdev->dev,"Can't remove PF VSI\n" );return -ENODEV ; } - +set_bit (__I40E_VSI_RELEASING, vsi->state); uplink_seid = vsi->uplink_seid;if (vsi->type !=I40E_VSI_SRIOV ) {if (vsi->netdev_registered) {
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 162/279 i40e: Fix changing previously set num_queue_pairs for PFs
In reply to Greg Kroah-HartmanFrom: Eryk Rybak <[email protected]>
[ Upstream commit d2a69fefd75683004ffe87166de5635b3267ee07 ]
Currently, the i40e_vsi_setup_queue_map is basing the count of queues in TCs on a VSI's alloc_queue_pairs member which is not changed throughout any user's action (for example via ethtool's set_channels callback).
This implies that vsi->tc_config.tc_info[n].qcount value that is given to the kernel via netdev_set_tc_queue() that notifies about the count of queues per particular traffic class is constant even if user has changed the total count of queues.
This in turn caused the kernel warning after setting the queue count to the lower value than the initial one:
$ ethtool -l ens801f0 Channel parameters for ens801f0: Pre-set maximums: RX: 0 TX: 0 Other: 1 Combined: 64 Current hardware settings: RX: 0 TX: 0 Other: 1 Combined: 64
$ ethtool -L ens801f0 combined 40
[dmesg]
Number of in use tx queues changed invalidating tc mappings. Priority traffic classification disabled!
Reason was that vsi->alloc_queue_pairs stayed at 64 value which was used to set the qcount on TC0 (by default only TC0 exists so all of the existing queues are assigned to TC0). we update the offset/qcount via netdev_set_tc_queue() back to the old value but then the netif_set_real_num_tx_queues() is using the vsi->num_queue_pairs as a value which got set to 40.
Fix it by using vsi->req_queue_pairs as a queue count that will be distributed across TCs. Do it only for non-zero values, which implies that user actually requested the new count of queues.
For VSIs other than main, stay with the vsi->alloc_queue_pairs as we only allow manipulating the queue count on main VSI.
--- drivers/net/ethernet/intel/i40e/i40e_main.c | 35 ++++++++++++++------- 1 file changed, 23 insertions(+), 12 deletions(-)
drivers/net/ethernet/intel/i40e/i40e_main.c +23 -12
@@ -1790,6 +1790,7 @@ static void i40e_vsi_setup_queue_map(struct i40e_vsi *vsi, bool is_add) {struct i40e_pf *pf = vsi->back; +u16 num_tc_qps =0 ;u16 sections =0 ;u8 netdev_tc =0 ;u16 numtc =1 ;
@@ -1797,13 +1798,29 @@ static void i40e_vsi_setup_queue_map(struct i40e_vsi *vsi,u8 offset;u16 qmap;int i; -u16 num_tc_qps =0 ; sections =I40E_AQ_VSI_PROP_QUEUE_MAP_VALID ; offset =0 ; +if (vsi->type ==I40E_VSI_MAIN ) { +/* This code helps add more queue to the VSI if we have +* more cores than RSS can support, the higher cores will +* be served by ATR or other filters. Furthermore, the +* non-zero req_queue_pairs says that user requested a new +* queue count via ethtool's set_channels, so use this +* value for queues distribution across traffic classes +*/ +if (vsi->req_queue_pairs >0 ) + vsi->num_queue_pairs = vsi->req_queue_pairs; +else if (pf->flags &I40E_FLAG_MSIX_ENABLED ) + vsi->num_queue_pairs = pf->num_lan_msix; + } +/* Number of queues per enabled TC */ - num_tc_qps = vsi->alloc_queue_pairs; +if (vsi->type ==I40E_VSI_MAIN ) + num_tc_qps = vsi->num_queue_pairs; +else + num_tc_qps = vsi->alloc_queue_pairs;if (enabled_tc && (vsi->back->flags &I40E_FLAG_DCB_ENABLED )) {/* Find numtc from enabled TC bitmap */ for (i =0 , numtc =0 ; i <I40E_MAX_TRAFFIC_CLASS ; i++) {
@@ -1881,16 +1898,10 @@ static void i40e_vsi_setup_queue_map(struct i40e_vsi *vsi, } ctxt->info.tc_mapping[i] =cpu_to_le16 (qmap); } - -/* Set actual Tx/Rx queue pairs */ - vsi->num_queue_pairs = offset; -if ((vsi->type ==I40E_VSI_MAIN ) && (numtc ==1 )) { -if (vsi->req_queue_pairs >0 ) - vsi->num_queue_pairs = vsi->req_queue_pairs; -else if (pf->flags &I40E_FLAG_MSIX_ENABLED ) - vsi->num_queue_pairs = pf->num_lan_msix; - } - +/* Do not change previously set num_queue_pairs for PFs */ +if ((vsi->type ==I40E_VSI_MAIN && numtc !=1 ) || + vsi->type !=I40E_VSI_MAIN ) + vsi->num_queue_pairs = offset;/* Scheduler section valid can only be set for ADD VSI */ if (is_add) { sections |=I40E_AQ_VSI_PROP_SCHED_VALID ;
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 163/279 i40e: Fix ping is lost after configuring ADq on VF
In reply to Greg Kroah-HartmanFrom: Eryk Rybak <[email protected]>
[ Upstream commit 9e0a603cb7dce2a19d98116d42de84b6db26d716 ]
Properly reconfigure VF VSIs after VF request ADQ.
Created new function to update queue mapping and queue pairs per TC with AQ update VSI. This sets proper RSS size on NIC.
VFs num_queue_pairs should not be changed during setup of queue maps. Previously, VF main VSI in ADQ had configured too many queues and had wrong RSS size, which lead to packets not being consumed and drops in connectivity.
--- drivers/net/ethernet/intel/i40e/i40e.h | 1 + drivers/net/ethernet/intel/i40e/i40e_main.c | 64 ++++++++++++++++++- .../ethernet/intel/i40e/i40e_virtchnl_pf.c | 17 +++-- 3 files changed, 74 insertions(+), 8 deletions(-)
drivers/net/ethernet/intel/i40e/i40e.h +1 -0
@@ -1248,6 +1248,7 @@ void i40e_ptp_restore_hw_time(struct i40e_pf *pf);void i40e_ptp_init (struct i40e_pf *pf);void i40e_ptp_stop (struct i40e_pf *pf);int i40e_ptp_alloc_pins (struct i40e_pf *pf); +int i40e_update_adq_vsi_queues (struct i40e_vsi *vsi,int vsi_offset);int i40e_is_vsi_uplink_mode_veb (struct i40e_vsi *vsi);i40e_status i40e_get_partition_bw_setting (struct i40e_pf *pf);i40e_status i40e_set_partition_bw_setting (struct i40e_pf *pf);
drivers/net/ethernet/intel/i40e/i40e_main.c +61 -3
93 lines not on this page · raw message
drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c +12 -5
@@ -2220,11 +2220,12 @@ static int i40e_vc_config_queues_msg(struct i40e_vf *vf, u8 *msg)struct virtchnl_vsi_queue_config_info *qci = (struct virtchnl_vsi_queue_config_info *)msg;struct virtchnl_queue_pair_info *qpi; -struct i40e_pf *pf = vf->pf;u16 vsi_id, vsi_queue_id =0 ; -u16 num_qps_all =0 ; +struct i40e_pf *pf = vf->pf;i40e_status aq_ret =0 ;int i, j =0 , idx =0 ; +struct i40e_vsi *vsi; +u16 num_qps_all =0 ;if (!test_bit (I40E_VF_STATE_ACTIVE , &vf->vf_states)) { aq_ret =I40E_ERR_PARAM ;
@@ -2313,9 +2314,15 @@ static int i40e_vc_config_queues_msg(struct i40e_vf *vf, u8 *msg) pf->vsi[vf->lan_vsi_idx]->num_queue_pairs = qci->num_queue_pairs; }else { -for (i =0 ; i < vf->num_tc; i++) - pf->vsi[vf->ch[i].vsi_idx]->num_queue_pairs = - vf->ch[i].num_qps; +for (i =0 ; i < vf->num_tc; i++) { + vsi = pf->vsi[vf->ch[i].vsi_idx]; + vsi->num_queue_pairs = vf->ch[i].num_qps; + +if (i40e_update_adq_vsi_queues (vsi, i)) { + aq_ret =I40E_ERR_CONFIG ; + goto error_param; + } + } } error_param:
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 164/279 RDMA/mlx4: Do not fail the registration on port stats
In reply to Greg Kroah-HartmanFrom: Jack Wang <[email protected]>
[ Upstream commit 378c67413de18b69fb3bb78d8c4f0f1192cfa973 ]
If the FW doesn't support MLX4_DEV_CAP_FLAG2_DIAG_PER_PORT, mlx4 driver will fail the ib_setup_port_attrs, which is called from ib_register_device()/enable_device_and_get(), in the end leads to device not detected[1][2]
To fix it, add a new mlx4_ib_hw_stats_ops1, w/o alloc_hw_port_stats if FW does not support MLX4_DEV_CAP_FLAG2_DIAG_PER_PORT.
[1] https://bugzilla.redhat.com/show_bug.cgi?id=2014094
[2] https://lore.kernel.org/linux-rdma/CAMGffEn2wvEnmzc0xe=xYiCLqpphiHDBxCxqAELrBofbUAMQxw@mail.gmail.com
--- drivers/infiniband/hw/mlx4/main.c | 18 +++++++++++++++--- 1 file changed, 15 insertions(+), 3 deletions(-)
drivers/infiniband/hw/mlx4/main.c +15 -3
@@ -2217,6 +2217,11 @@ static const struct ib_device_ops mlx4_ib_hw_stats_ops = { .get_hw_stats = mlx4_ib_get_hw_stats, }; +static const struct ib_device_ops mlx4_ib_hw_stats_ops1 = { + .alloc_hw_device_stats = mlx4_ib_alloc_hw_device_stats, + .get_hw_stats = mlx4_ib_get_hw_stats, +}; +static int mlx4_ib_alloc_diag_counters (struct mlx4_ib_dev *ibdev) {struct mlx4_ib_diag_counters *diag = ibdev->diag_counters;
@@ -2229,9 +2234,16 @@ static int mlx4_ib_alloc_diag_counters(struct mlx4_ib_dev *ibdev)return 0 ;for (i =0 ; i <MLX4_DIAG_COUNTERS_TYPES ; i++) { -/* i == 1 means we are building port counters */ -if (i && !per_port) -continue ; +/* +* i == 1 means we are building port counters, set a different +* stats ops without port stats callback. +*/ +if (i && !per_port) { +ib_set_device_ops (&ibdev->ib_dev, + &mlx4_ib_hw_stats_ops1); + +return 0 ; + } ret = __mlx4_ib_alloc_diag_counters(ibdev, &diag[i].name, &diag[i].offset,
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 165/279 i40e: Fix warning message and call stack during rmmod i40e driver
In reply to Greg Kroah-HartmanFrom: Karen Sornek <[email protected]>
[ Upstream commit 3a3b311e3881172fc8e019b6508f04bc40c92d9d ]
Restore part of reset functionality used when reset is called from the VF to reset itself. Without this fix warning message is displayed when VF is being removed via sysfs.
Fix the crash of the VF during reset by ensuring that the PF receives the reset message successfully. Refactor code to use one function instead of two.
--- .../ethernet/intel/i40e/i40e_virtchnl_pf.c | 53 ++++++++----------- 1 file changed, 21 insertions(+), 32 deletions(-)
drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c +21 -32
@@ -183,17 +183,18 @@ void i40e_vc_notify_vf_reset(struct i40e_vf *vf)/***********************misc routines*****************************/ /** -* i40e_vc_disable_vf +* i40e_vc_reset_vf * @vf: pointer to the VF info -* -* Disable the VF through a SW reset. +* @notify_vf: notify vf about reset or not +* Reset VF handler. **/ -static inline void i40e_vc_disable_vf (struct i40e_vf *vf) +static void i40e_vc_reset_vf (struct i40e_vf *vf,bool notify_vf) {struct i40e_pf *pf = vf->pf;int i; -i40e_vc_notify_vf_reset (vf); +if (notify_vf) +i40e_vc_notify_vf_reset (vf);/* We want to ensure that an actual reset occurs initiated after this * function was called. However, we do not want to wait forever, so
@@ -211,9 +212,14 @@ static inline void i40e_vc_disable_vf(struct i40e_vf *vf)usleep_range (10000 ,20000 ); } -dev_warn (&vf->pf->pdev->dev, -"Failed to initiate reset for VF %d after 200 milliseconds\n" , - vf->vf_id); +if (notify_vf) +dev_warn (&vf->pf->pdev->dev, +"Failed to initiate reset for VF %d after 200 milliseconds\n" , + vf->vf_id); +else +dev_dbg (&vf->pf->pdev->dev, +"Failed to initiate reset for VF %d after 200 milliseconds\n" , + vf->vf_id); }/**
@@ -2108,20 +2114,6 @@ err:return ret; } -/** -* i40e_vc_reset_vf_msg -* @vf: pointer to the VF info -* -* called from the VF to reset itself, -* unlike other virtchnl messages, PF driver -* doesn't send the response back to the VF -**/ -static void i40e_vc_reset_vf_msg (struct i40e_vf *vf) -{ -if (test_bit (I40E_VF_STATE_ACTIVE , &vf->vf_states)) -i40e_reset_vf (vf, false); -} -/** * i40e_vc_config_promiscuous_mode_msg * @vf: pointer to the VF info
@@ -2617,8 +2609,7 @@ static int i40e_vc_request_queues_msg(struct i40e_vf *vf, u8 *msg) }else {/* successful request */ vf->num_req_queues = req_pairs; -i40e_vc_notify_vf_reset (vf); -i40e_reset_vf (vf, false); +i40e_vc_reset_vf (vf, true);return 0 ; }
@@ -3813,8 +3804,7 @@ static int i40e_vc_add_qch_msg(struct i40e_vf *vf, u8 *msg) vf->num_req_queues =0 ;/* reset the VF in order to allocate resources */ -i40e_vc_notify_vf_reset (vf); -i40e_reset_vf (vf, false); +i40e_vc_reset_vf (vf, true);return I40E_SUCCESS ;
@@ -3854,8 +3844,7 @@ static int i40e_vc_del_qch_msg(struct i40e_vf *vf, u8 *msg) }/* reset the VF in order to allocate resources */ -i40e_vc_notify_vf_reset (vf); -i40e_reset_vf (vf, false); +i40e_vc_reset_vf (vf, true);return I40E_SUCCESS ;
@@ -3917,7 +3906,7 @@ int i40e_vc_process_vf_msg(struct i40e_pf *pf, s16 vf_id, u32 v_opcode,i40e_vc_notify_vf_link_state (vf);break ;case VIRTCHNL_OP_RESET_VF : -i40e_vc_reset_vf_msg (vf); +i40e_vc_reset_vf (vf, false); ret =0 ;break ;case VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE :
@@ -4171,7 +4160,7 @@ int i40e_ndo_set_vf_mac(struct net_device *netdev, int vf_id, u8 *mac)/* Force the VF interface down so it has to bring up with new MAC * address */ -i40e_vc_disable_vf (vf); +i40e_vc_reset_vf (vf, true);dev_info (&pf->pdev->dev,"Bring down and up the VF interface to make this change effective.\n" ); error_param:
@@ -4235,7 +4224,7 @@ int i40e_ndo_set_vf_port_vlan(struct net_device *netdev, int vf_id,/* duplicate request, so just return success */ goto error_pvid; -i40e_vc_disable_vf (vf); +i40e_vc_reset_vf (vf, true);/* During reset the VF got a new VSI, so refresh a pointer. */ vsi = pf->vsi[vf->lan_vsi_idx];/* Locked once because multiple functions below iterate list */
@@ -4613,7 +4602,7 @@ int i40e_ndo_set_vf_trust(struct net_device *netdev, int vf_id, bool setting) goto out; vf->trusted = setting; -i40e_vc_disable_vf (vf); +i40e_vc_reset_vf (vf, true);dev_info (&pf->pdev->dev,"VF %u is now %strusted\n" , vf_id, setting ?"" :"un" );
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 166/279 i40e: Fix creation of first queue by omitting it if is not power of two
In reply to Greg Kroah-HartmanFrom: Jedrzej Jagielski <[email protected]>
[ Upstream commit 2e6d218c1ec6fb9cd70693b78134cbc35ae0b5a9 ]
Reject TCs creation with proper message if the first queue assignment is not equal to the power of two.
The first queue number was checked too late in the second queue iteration, if second queue was configured at all. Now if first queue value is not a power of two, then trying to create qdisc will be rejected.
--- drivers/net/ethernet/intel/i40e/i40e_main.c | 59 +++++++-------------- 1 file changed, 19 insertions(+), 40 deletions(-)
drivers/net/ethernet/intel/i40e/i40e_main.c +19 -40
@@ -5786,24 +5786,6 @@ static void i40e_remove_queue_channels(struct i40e_vsi *vsi)INIT_LIST_HEAD (&vsi->ch_list); } -/** -* i40e_is_any_channel - channel exist or not -* @vsi: ptr to VSI to which channels are associated with -* -* Returns true or false if channel(s) exist for associated VSI or not -**/ -static bool i40e_is_any_channel (struct i40e_vsi *vsi) -{ -struct i40e_channel *ch, *ch_tmp; - -list_for_each_entry_safe (ch, ch_tmp, &vsi->ch_list, list) { -if (ch->initialized) -return true; - } - -return false; -} -/** * i40e_get_max_queues_for_channel * @vsi: ptr to VSI to which channels are associated with
@@ -6310,26 +6292,15 @@ int i40e_create_queue_channel(struct i40e_vsi *vsi,/* By default we are in VEPA mode, if this is the first VF/VMDq * VSI to be added switch to VEB mode. */ -if ((!(pf->flags &I40E_FLAG_VEB_MODE_ENABLED )) || - (!i40e_is_any_channel (vsi))) { -if (!is_power_of_2 (vsi->tc_config.tc_info[0 ].qcount)) { -dev_dbg (&pf->pdev->dev, -"Failed to create channel. Override queues (%u) not power of 2\n" , - vsi->tc_config.tc_info[0 ].qcount); -return -EINVAL ; - } -if (!(pf->flags &I40E_FLAG_VEB_MODE_ENABLED )) { - pf->flags |=I40E_FLAG_VEB_MODE_ENABLED ; +if (!(pf->flags &I40E_FLAG_VEB_MODE_ENABLED )) { + pf->flags |=I40E_FLAG_VEB_MODE_ENABLED ; -if (vsi->type ==I40E_VSI_MAIN ) { -if (pf->flags &I40E_FLAG_TC_MQPRIO ) -i40e_do_reset (pf,I40E_PF_RESET_FLAG , - true); -else -i40e_do_reset_safe (pf, -I40E_PF_RESET_FLAG ); - } +if (vsi->type ==I40E_VSI_MAIN ) { +if (pf->flags &I40E_FLAG_TC_MQPRIO ) +i40e_do_reset (pf,I40E_PF_RESET_FLAG , true); +else +i40e_do_reset_safe (pf,I40E_PF_RESET_FLAG ); }/* now onwards for main VSI, number of queues will be value * of TC0's queue count
@@ -7982,12 +7953,20 @@ config_tc: vsi->seid); need_reset = true; goto exit; - }else { -dev_info (&vsi->back->pdev->dev, -"Setup channel (id:%u) utilizing num_queues %d\n" , - vsi->seid, vsi->tc_config.tc_info[0 ].qcount); + }else if (enabled_tc && + (!is_power_of_2 (vsi->tc_config.tc_info[0 ].qcount))) { +netdev_info (netdev, +"Failed to create channel. Override queues (%u) not power of 2\n" , + vsi->tc_config.tc_info[0 ].qcount); + ret = -EINVAL ; + need_reset = true; + goto exit; } +dev_info (&vsi->back->pdev->dev, +"Setup channel (id:%u) utilizing num_queues %d\n" , + vsi->seid, vsi->tc_config.tc_info[0 ].qcount); +if (pf->flags &I40E_FLAG_TC_MQPRIO ) {if (vsi->mqprio_qopt.max_rate[0 ]) {u64 max_tx_rate = vsi->mqprio_qopt.max_rate[0 ];
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 167/279 i40e: Fix display error code in dmesg
In reply to Greg Kroah-HartmanFrom: Grzegorz Szczurek <[email protected]>
[ Upstream commit 5aff430d4e33a0b48a6b3d5beb06f79da23f9916 ]
Fix misleading display error in dmesg if tc filter return fail. Only i40e status error code should be converted to string, not linux error code. Otherwise, we return false information about the error.
--- drivers/net/ethernet/intel/i40e/i40e_main.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-)
drivers/net/ethernet/intel/i40e/i40e_main.c +2 -3
@@ -8531,9 +8531,8 @@ static int i40e_configure_clsflower(struct i40e_vsi *vsi, err =i40e_add_del_cloud_filter (vsi, filter, true);if (err) { -dev_err (&pf->pdev->dev, -"Failed to add cloud filter, err %s\n" , -i40e_stat_str (&pf->hw, err)); +dev_err (&pf->pdev->dev,"Failed to add cloud filter, err %d\n" , + err); goto err; }
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 168/279 NFC: reorganize the functions in nci_request
In reply to Greg Kroah-HartmanFrom: Lin Ma <[email protected]>
[ Upstream commit 86cdf8e38792545161dbe3350a7eced558ba4d15 ]
There is a possible data race as shown below:
thread-A in nci_request() | thread-B in nci_close_device()
| mutex_lock(&ndev->req_lock);
test_bit(NCI_UP, &ndev->flags); |
... | test_and_clear_bit(NCI_UP, &ndev->flags)
mutex_lock(&ndev->req_lock); |
|This race will allow __nci_request() to be awaked while the device is getting removed.
Similar to commit e2cb6b891ad2 ("bluetooth: eliminate the potential race condition when removing the HCI controller"). this patch alters the function sequence in nci_request() to prevent the data races between the nci_close_device().
--- net/nfc/nci/core.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-)
net/nfc/nci/core.c +7 -4
@@ -144,12 +144,15 @@ inline int nci_request(struct nci_dev *ndev, {int rc; -if (!test_bit (NCI_UP , &ndev->flags)) -return -ENETDOWN ; -/* Serialize all requests */ mutex_lock (&ndev->req_lock); - rc =__nci_request (ndev, req, opt, timeout); +/* check the state after obtaing the lock against any races +* from nci_close_device when the device gets removed. +*/ +if (test_bit (NCI_UP , &ndev->flags)) + rc =__nci_request (ndev, req, opt, timeout); +else + rc = -ENETDOWN ;mutex_unlock (&ndev->req_lock);return rc;
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 169/279 NFC: reorder the logic in nfc_{un,}register_device
In reply to Greg Kroah-HartmanFrom: Lin Ma <[email protected]>
[ Upstream commit 3e3b5dfcd16a3e254aab61bd1e8c417dd4503102 ]
There is a potential UAF between the unregistration routine and the NFC netlink operations.
The race that cause that UAF can be shown as below:
(FREE) | (USE)
nfcmrvl_nci_unregister_dev | nfc_genl_dev_up
nci_close_device |
nci_unregister_device | nfc_get_device
nfc_unregister_device | nfc_dev_up
rfkill_destory |
device_del | rfkill_blocked
... | ...The root cause for this race is concluded below:
1. The rfkill_blocked (USE) in nfc_dev_up is supposed to be placed after the device_is_registered check.
2. Since the netlink operations are possible just after the device_add in nfc_register_device, the nfc_dev_up() can happen anywhere during the rfkill creation process, which leads to data race.
This patch reorder these actions to permit
1. Once device_del is finished, the nfc_dev_up cannot dereference the rfkill object.
2. The rfkill_register need to be placed after the device_add of nfc_dev because the parent device need to be created first. So this patch keeps the order but inject device_lock to prevent the data race.
--- net/nfc/core.c | 32 ++++++++++++++++++-------------- 1 file changed, 18 insertions(+), 14 deletions(-)
net/nfc/core.c +18 -14
@@ -94,13 +94,13 @@ int nfc_dev_up(struct nfc_dev *dev)device_lock (&dev->dev); -if (dev->rfkill &&rfkill_blocked (dev->rfkill)) { - rc = -ERFKILL ; +if (!device_is_registered (&dev->dev)) { + rc = -ENODEV ; goto error; } -if (!device_is_registered (&dev->dev)) { - rc = -ENODEV ; +if (dev->rfkill &&rfkill_blocked (dev->rfkill)) { + rc = -ERFKILL ; goto error; }
@@ -1125,11 +1125,7 @@ int nfc_register_device(struct nfc_dev *dev)if (rc)pr_err ("Could not register llcp device\n" ); - rc =nfc_genl_device_added (dev); -if (rc) -pr_debug ("The userspace won't be notified that the device %s was added\n" , -dev_name (&dev->dev)); - +device_lock (&dev->dev); dev->rfkill =rfkill_alloc (dev_name (&dev->dev), &dev->dev,RFKILL_TYPE_NFC , &nfc_rfkill_ops, dev);if (dev->rfkill) {
@@ -1138,6 +1134,12 @@ int nfc_register_device(struct nfc_dev *dev) dev->rfkill =NULL ; } } +device_unlock (&dev->dev); + + rc =nfc_genl_device_added (dev); +if (rc) +pr_debug ("The userspace won't be notified that the device %s was added\n" , +dev_name (&dev->dev));return 0 ; }
@@ -1154,10 +1156,17 @@ void nfc_unregister_device(struct nfc_dev *dev)pr_debug ("dev_name=%s\n" ,dev_name (&dev->dev)); + rc =nfc_genl_device_removed (dev); +if (rc) +pr_debug ("The userspace won't be notified that the device %s " +"was removed\n" ,dev_name (&dev->dev)); + +device_lock (&dev->dev);if (dev->rfkill) {rfkill_unregister (dev->rfkill);rfkill_destroy (dev->rfkill); } +device_unlock (&dev->dev);if (dev->ops->check_presence) {device_lock (&dev->dev);
@@ -1167,11 +1176,6 @@ void nfc_unregister_device(struct nfc_dev *dev)cancel_work_sync (&dev->check_pres_work); } - rc =nfc_genl_device_removed (dev); -if (rc) -pr_debug ("The userspace won't be notified that the device %s " -"was removed\n" ,dev_name (&dev->dev)); -nfc_llcp_unregister_device (dev);mutex_lock (&nfc_devlist_mutex);
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 170/279 NFC: add NCI_UNREG flag to eliminate the race
In reply to Greg Kroah-HartmanFrom: Lin Ma <[email protected]>
[ Upstream commit 48b71a9e66c2eab60564b1b1c85f4928ed04e406 ]
There are two sites that calls queue_work() after the destroy_workqueue() and lead to possible UAF.
The first site is nci_send_cmd(), which can happen after the nci_close_device as below
nfcmrvl_nci_unregister_dev | nfc_genl_dev_up
nci_close_device |
flush_workqueue |
del_timer_sync |
nci_unregister_device | nfc_get_device
destroy_workqueue | nfc_dev_up
nfc_unregister_device | nci_dev_up
device_del | nci_open_device
| __nci_request
| nci_send_cmd
| queue_work !!!Another site is nci_cmd_timer, awaked by the nci_cmd_work from the nci_send_cmd.
... | ...
nci_unregister_device | queue_work
destroy_workqueue |
nfc_unregister_device | ...
device_del | nci_cmd_work
| mod_timer
| ...
| nci_cmd_timer
| queue_work !!!For the above two UAF, the root cause is that the nfc_dev_up can race between the nci_unregister_device routine. Therefore, this patch introduce NCI_UNREG flag to easily eliminate the possible race. In addition, the mutex_lock in nci_close_device can act as a barrier.
--- include/net/nfc/nci_core.h | 1 + net/nfc/nci/core.c | 19 +++++++++++++++++-- 2 files changed, 18 insertions(+), 2 deletions(-)
include/net/nfc/nci_core.h +1 -0
@@ -30,6 +30,7 @@ enum nci_flag {NCI_UP ,NCI_DATA_EXCHANGE ,NCI_DATA_EXCHANGE_TO , + NCI_UNREG, };/* NCI device states */
net/nfc/nci/core.c +17 -2
@@ -476,6 +476,11 @@ static int nci_open_device(struct nci_dev *ndev)mutex_lock (&ndev->req_lock); +if (test_bit (NCI_UNREG , &ndev->flags)) { + rc = -ENODEV ; + goto done; + } +if (test_bit (NCI_UP , &ndev->flags)) { rc = -EALREADY ; goto done;
@@ -548,6 +553,10 @@ done:static int nci_close_device (struct nci_dev *ndev) {nci_req_cancel (ndev,ENODEV ); + +/* This mutex needs to be held as a barrier for +* caller nci_unregister_device +*/ mutex_lock (&ndev->req_lock);if (!test_and_clear_bit (NCI_UP , &ndev->flags)) {
@@ -585,8 +594,8 @@ static int nci_close_device(struct nci_dev *ndev)del_timer_sync (&ndev->cmd_timer); -/* Clear flags */ - ndev->flags =0 ; +/* Clear flags except NCI_UNREG */ + ndev->flags &=BIT (NCI_UNREG );mutex_unlock (&ndev->req_lock);
@@ -1273,6 +1282,12 @@ void nci_unregister_device(struct nci_dev *ndev) {struct nci_conn_info *conn_info, *n; +/* This set_bit is not protected with specialized barrier, +* However, it is fine because the mutex_lock(&ndev->req_lock); +* in nci_close_device() will help to emit one. +*/ +set_bit (NCI_UNREG , &ndev->flags); +nci_close_device (ndev);destroy_workqueue (ndev->cmd_wq);
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 171/279 e100: fix device suspend/resume
In reply to Greg Kroah-HartmanFrom: Jesse Brandeburg <[email protected]>
[ Upstream commit 5d2ca2e12dfb2aff3388ca57b06f570fa6206ced ]
As reported in [1], e100 was no longer working for suspend/resume cycles. The previous commit mentioned in the fixes appears to have broken things and this attempts to practice best known methods for device power management and keep wake-up working while allowing suspend/resume to work. To do this, I reorder a little bit of code and fix the resume path to make sure the device is enabled.
[1] https://bugzilla.kernel.org/show_bug.cgi?id=214933
--- drivers/net/ethernet/intel/e100.c | 18 +++++++++++++----- 1 file changed, 13 insertions(+), 5 deletions(-)
drivers/net/ethernet/intel/e100.c +13 -5
@@ -3003,9 +3003,10 @@ static void __e100_shutdown(struct pci_dev *pdev, bool *enable_wake)struct net_device *netdev =pci_get_drvdata (pdev);struct nic *nic =netdev_priv (netdev); +netif_device_detach (netdev); +if (netif_running (netdev))e100_down (nic); -netif_device_detach (netdev);if ((nic->flags & wol_magic) |e100_asf (nic)) {/* enable reverse auto-negotiation */
@@ -3022,7 +3023,7 @@ static void __e100_shutdown(struct pci_dev *pdev, bool *enable_wake) *enable_wake = false; } -pci_clear_master (pdev); +pci_disable_device (pdev); }static int __e100_power_off (struct pci_dev *pdev,bool wake)
@@ -3042,8 +3043,6 @@ static int __maybe_unused e100_suspend(struct device *dev_d)__e100_shutdown (to_pci_dev (dev_d), &wake); -device_wakeup_disable (dev_d); -return 0 ; }
@@ -3051,6 +3050,14 @@ static int __maybe_unused e100_resume(struct device *dev_d) {struct net_device *netdev =dev_get_drvdata (dev_d);struct nic *nic =netdev_priv (netdev); +int err; + + err =pci_enable_device (to_pci_dev (dev_d)); +if (err) { +netdev_err (netdev,"Resume cannot enable PCI device, aborting\n" ); +return err; + } +pci_set_master (to_pci_dev (dev_d));/* disable reverse auto-negotiation */ if (nic->phy == phy_82552_v) {
@@ -3062,10 +3069,11 @@ static int __maybe_unused e100_resume(struct device *dev_d) smartspeed & ~(E100_82552_REV_ANEG)); } -netif_device_attach (netdev);if (netif_running (netdev))e100_up (nic); +netif_device_attach (netdev); +return 0 ; }
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 172/279 ptp: ocp: Fix a couple NULL vs IS_ERR() checks
In reply to Greg Kroah-HartmanFrom: Dan Carpenter <[email protected]>
[ Upstream commit c7521d3aa2fa7fc785682758c99b5bcae503f6be ]
The ptp_ocp_get_mem() function does not return NULL, it returns error pointers.
--- drivers/ptp/ptp_ocp.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-)
drivers/ptp/ptp_ocp.c +5 -4
@@ -1049,10 +1049,11 @@ ptp_ocp_register_ext(struct ptp_ocp *bp, struct ocp_resource *r)if (!ext)return -ENOMEM ; - err = -EINVAL ; ext->mem =ptp_ocp_get_mem (bp, r); -if (!ext->mem) +if (IS_ERR (ext->mem)) { + err =PTR_ERR (ext->mem); goto out; + } ext->bp = bp; ext->info = r->extra;
@@ -1122,8 +1123,8 @@ ptp_ocp_register_mem(struct ptp_ocp *bp, struct ocp_resource *r)void __iomem *mem; mem =ptp_ocp_get_mem (bp, r); -if (!mem) -return -EINVAL ; +if (IS_ERR (mem)) +return PTR_ERR (mem);bp_assign_entry (bp, r, mem);
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 173/279 perf bench: Fix two memory leaks detected with ASan
In reply to Greg Kroah-HartmanFrom: Sohaib Mohamed <[email protected]>
[ Upstream commit 92723ea0f11d92496687db8c9725248e9d1e5e1d ]
ASan reports memory leaks while running:
$ perf bench sched all
--- tools/perf/bench/sched-messaging.c | 4 ++++ 1 file changed, 4 insertions(+)
tools/perf/bench/sched-messaging.c +4 -0
@@ -223,6 +223,8 @@ static unsigned int group(pthread_t *pth, snd_ctx->out_fds[i] = fds[1 ];if (!thread_mode)close (fds[0 ]); + +free (ctx); }/* Now we have all the fds, fork the senders */
@@ -239,6 +241,8 @@ static unsigned int group(pthread_t *pth,for (i =0 ; i < num_fds; i++)close (snd_ctx->out_fds[i]); +free (snd_ctx); +/* Return number of children to reap */ return num_fds *2 ; }
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 174/279 tools build: Fix removal of feature-sync-compare-and-swap feature detection
In reply to Greg Kroah-HartmanFrom: Arnaldo Carvalho de Melo <[email protected]>
[ Upstream commit e8c04ea0fef5731dbcaabac86d65254c227aedf4 ]
The patch removing the feature-sync-compare-and-swap feature detection didn't remove the call to main_test_sync_compare_and_swap(), making the 'test-all' case fail an all the feature tests to be performed individually:
$ cat /tmp/build/perf/feature/test-all.make.output
In file included from test-all.c:18:
test-libpython-version.c:5:10: error: #error
5 | #error
| ^~~~~
test-all.c: In function ‘main’:
test-all.c:203:9: error: implicit declaration of function ‘main_test_sync_compare_and_swap’ [-Werror=implicit-function-declaration]
203 | main_test_sync_compare_and_swap(argc, argv);
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
cc1: all warnings being treated as errors
$Fix it, now to figure out what is that test-libpython-version.c problem...
--- tools/build/feature/test-all.c | 1 - 1 file changed, 1 deletion(-)
tools/build/feature/test-all.c +0 -1
@@ -200,7 +200,6 @@ int main(int argc, char *argv[])main_test_timerfd ();main_test_stackprotector_all ();main_test_libdw_dwarf_unwind (); -main_test_sync_compare_and_swap (argc, argv);main_test_zlib ();main_test_pthread_attr_setaffinity_np ();main_test_pthread_barrier ();
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 175/279 riscv: fix building external modules
In reply to Greg Kroah-HartmanFrom: Andreas Schwab <[email protected]>
[ Upstream commit 5a19c7e06236a9c55dfc001bb4d1a8f1950d23e7 ]
When building external modules, vdso_prepare should not be run. If the kernel sources are read-only, it will fail.
--- arch/riscv/Makefile | 2 ++ 1 file changed, 2 insertions(+)
arch/riscv/Makefile +2 -0
@@ -108,11 +108,13 @@ PHONY += vdso_install vdso_install: $(Q)$(MAKE) $(build)=arch/riscv/kernel/vdso $@ +ifeq ($(KBUILD_EXTMOD),) ifeq ($(CONFIG_MMU),y) prepare: vdso_prepare vdso_prepare: prepare0 $(Q)$(MAKE) $(build)=arch/riscv/kernel/vdso include/generated/vdso-offsets.h endif +endif ifneq ($(CONFIG_XIP_KERNEL),y) ifeq ($(CONFIG_RISCV_M_MODE)$(CONFIG_SOC_CANAAN),yy)
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 176/279 KVM: PPC: Book3S HV: Use GLOBAL_TOC for kvmppc_h_set_dabr/xdabr()
In reply to Greg Kroah-HartmanFrom: Michael Ellerman <[email protected]>
[ Upstream commit dae581864609d36fb58855fd59880b4941ce9d14 ]
kvmppc_h_set_dabr(), and kvmppc_h_set_xdabr() which jumps into it, need to use _GLOBAL_TOC to setup the kernel TOC pointer, because kvmppc_h_set_dabr() uses LOAD_REG_ADDR() to load dawr_force_enable.
When called from hcall_try_real_mode() we have the kernel TOC in r2, established near the start of kvmppc_interrupt_hv(), so there is no issue.
But they can also be called from kvmppc_pseries_do_hcall() which is module code, so the access ends up happening with the kvm-hv module's r2, which will not point at dawr_force_enable and could even cause a fault.
With the current code layout and compilers we haven't observed a fault in practice, the load hits somewhere in kvm-hv.ko and silently returns some bogus value.
Note that we we expect p8/p9 guests to use the DAWR, but SLOF uses h_set_dabr() to test if sc1 works correctly, see SLOF's lib/libhvcall/brokensc1.c.
--- arch/powerpc/kvm/book3s_hv_rmhandlers.S | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
arch/powerpc/kvm/book3s_hv_rmhandlers.S +2 -2
@@ -2005,7 +2005,7 @@ hcall_real_table: .globl hcall_real_table_end hcall_real_table_end: -_GLOBAL(kvmppc_h_set_xdabr) +_GLOBAL_TOC(kvmppc_h_set_xdabr) EXPORT_SYMBOL_GPL(kvmppc_h_set_xdabr) andi. r0, r5, DABRX_USER | DABRX_KERNEL beq 6f
@@ -2015,7 +2015,7 @@ EXPORT_SYMBOL_GPL(kvmppc_h_set_xdabr) 6: li r3, H_PARAMETER blr -_GLOBAL(kvmppc_h_set_dabr) +_GLOBAL_TOC(kvmppc_h_set_dabr) EXPORT_SYMBOL_GPL(kvmppc_h_set_dabr) li r5, DABRX_USER | DABRX_KERNEL 3:
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 177/279 powerpc: clean vdso32 and vdso64 directories
In reply to Greg Kroah-HartmanFrom: Masahiro Yamada <[email protected]>
[ Upstream commit 964c33cd0be621b291b5d253d8731eb2680082cb ]
Since commit bce74491c300 ("powerpc/vdso: fix unnecessary rebuilds of vgettimeofday.o"), "make ARCH=powerpc clean" does not clean up the arch/powerpc/kernel/{vdso32,vdso64} directories.
Use the subdir- trick to let "make clean" descend into them.
--- arch/powerpc/kernel/Makefile | 3 +++ 1 file changed, 3 insertions(+)
arch/powerpc/kernel/Makefile +3 -0
@@ -196,3 +196,6 @@ clean-files := vmlinux.lds # Force dependency (incbin is bad) $(obj)/vdso32_wrapper.o : $(obj)/vdso32/vdso32.so.dbg $(obj)/vdso64_wrapper.o : $(obj)/vdso64/vdso64.so.dbg + +# for cleaning +subdir- += vdso32 vdso64
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 178/279 powerpc/pseries: rename numa_dist_table to form2_distances
In reply to Greg Kroah-HartmanFrom: Nicholas Piggin <[email protected]>
[ Upstream commit 0bd81274e3f1195ee7c820ef02d62f31077c42c3 ]
The name of the local variable holding the "form2" property address conflicts with the numa_distance_table global.
This patch does 's/numa_dist_table/form2_distances/g' over the function, which also renames numa_dist_table_length to form2_distances_length.
--- arch/powerpc/mm/numa.c | 18 +++++++++--------- 1 file changed, 9 insertions(+), 9 deletions(-)
arch/powerpc/mm/numa.c +9 -9
@@ -376,9 +376,9 @@ static void initialize_form2_numa_distance_lookup_table(void) {int i, j;struct device_node *root; -const __u8 *numa_dist_table; +const __u8 *form2_distances;const __be32 *numa_lookup_index; -int numa_dist_table_length; +int form2_distances_length;int max_numa_index, distance_index;if (firmware_has_feature (FW_FEATURE_OPAL ))
@@ -392,20 +392,20 @@ static void initialize_form2_numa_distance_lookup_table(void) max_numa_index =of_read_number (&numa_lookup_index[0 ],1 );/* first element of the array is the size and is encode-int */ - numa_dist_table =of_get_property (root,"ibm,numa-distance-table" ,NULL ); - numa_dist_table_length =of_read_number ((const __be32 *)&numa_dist_table[0 ],1 ); + form2_distances =of_get_property (root,"ibm,numa-distance-table" ,NULL ); + form2_distances_length =of_read_number ((const __be32 *)&form2_distances[0 ],1 );/* Skip the size which is encoded int */ - numa_dist_table +=sizeof (__be32); + form2_distances +=sizeof (__be32); -pr_debug ("numa_dist_table_len = %d, numa_dist_indexes_len = %d\n" , - numa_dist_table_length, max_numa_index); +pr_debug ("form2_distances_len = %d, numa_dist_indexes_len = %d\n" , + form2_distances_length, max_numa_index);for (i =0 ; i < max_numa_index; i++)/* +1 skip the max_numa_index in the property */ numa_id_index_table[i] =of_read_number (&numa_lookup_index[i +1 ],1 ); -if (numa_dist_table_length != max_numa_index * max_numa_index) { +if (form2_distances_length != max_numa_index * max_numa_index) {WARN (1 ,"Wrong NUMA distance information\n" );/* consider everybody else just remote. */ for (i =0 ; i < max_numa_index; i++) {
@@ -427,7 +427,7 @@ static void initialize_form2_numa_distance_lookup_table(void)int nodeA = numa_id_index_table[i];int nodeB = numa_id_index_table[j]; - numa_distance_table[nodeA][nodeB] = numa_dist_table[distance_index++]; + numa_distance_table[nodeA][nodeB] = form2_distances[distance_index++];pr_debug ("dist[%d][%d]=%d " , nodeA, nodeB, numa_distance_table[nodeA][nodeB]); } }
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 179/279 powerpc/pseries: Fix numa FORM2 parsing fallback code
In reply to Greg Kroah-HartmanFrom: Nicholas Piggin <[email protected]>
[ Upstream commit 302039466f6a3b9421ecb9a6a2c528801dc24a86 ]
In case the FORM2 distance table from firmware is not the expected size, there is fallback code that just populates the lookup table as local vs remote.
However it then continues on to use the distance table. Fix.
--- arch/powerpc/mm/numa.c | 28 ++++++++++++---------------- 1 file changed, 12 insertions(+), 16 deletions(-)
arch/powerpc/mm/numa.c +12 -16
@@ -407,30 +407,26 @@ static void initialize_form2_numa_distance_lookup_table(void)if (form2_distances_length != max_numa_index * max_numa_index) {WARN (1 ,"Wrong NUMA distance information\n" ); -/* consider everybody else just remote. */ -for (i =0 ; i < max_numa_index; i++) { -for (j =0 ; j < max_numa_index; j++) { -int nodeA = numa_id_index_table[i]; -int nodeB = numa_id_index_table[j]; - -if (nodeA == nodeB) - numa_distance_table[nodeA][nodeB] =LOCAL_DISTANCE ; -else - numa_distance_table[nodeA][nodeB] =REMOTE_DISTANCE ; - } - } + form2_distances =NULL ;// don't use it } - distance_index =0 ;for (i =0 ; i < max_numa_index; i++) {for (j =0 ; j < max_numa_index; j++) {int nodeA = numa_id_index_table[i];int nodeB = numa_id_index_table[j]; - - numa_distance_table[nodeA][nodeB] = form2_distances[distance_index++]; -pr_debug ("dist[%d][%d]=%d " , nodeA, nodeB, numa_distance_table[nodeA][nodeB]); +int dist; + +if (form2_distances) + dist = form2_distances[distance_index++]; +else if (nodeA == nodeB) + dist =LOCAL_DISTANCE ; +else + dist =REMOTE_DISTANCE ; + numa_distance_table[nodeA][nodeB] = dist; +pr_debug ("dist[%d][%d]=%d " , nodeA, nodeB, dist); } } +of_node_put (root); }
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 180/279 pinctrl: qcom: sdm845: Enable dual edge errata
In reply to Greg Kroah-HartmanFrom: Bjorn Andersson <[email protected]>
[ Upstream commit 3a3a100473d2f6ebf9bdfe6efedd7e18de724388 ]
It has been observed that dual edge triggered wakeirq GPIOs on SDM845 doesn't trigger interrupts on the falling edge.
Enabling wakeirq_dual_edge_errata for SDM845 indicates that the PDC in SDM845 suffers from the same problem described, and worked around, by Doug in 'c3c0c2e18d94 ("pinctrl: qcom: Handle broken/missing PDC dual edge IRQs on sc7180")', so enable the workaround for SDM845 as well.
The specific problem seen without this is that gpio-keys does not detect the falling edge of the LID gpio on the Lenovo Yoga C630 and as such consistently reports the LID as closed.
--- drivers/pinctrl/qcom/pinctrl-sdm845.c | 1 + 1 file changed, 1 insertion(+)
drivers/pinctrl/qcom/pinctrl-sdm845.c +1 -0
@@ -1310,6 +1310,7 @@ static const struct msm_pinctrl_soc_data sdm845_pinctrl = { .ngpios = 151, .wakeirq_map = sdm845_pdc_map, .nwakeirq_map = ARRAY_SIZE(sdm845_pdc_map), + .wakeirq_dual_edge_errata = true, };static const struct msm_pinctrl_soc_data sdm845_acpi_pinctrl = {
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 181/279 pinctrl: qcom: sm8350: Correct UFS and SDC offsets
In reply to Greg Kroah-HartmanFrom: Bjorn Andersson <[email protected]>
[ Upstream commit 62209e805b5c68577602a5803a71d8e2e11ee0d3 ]
The downstream TLMM binding covers a group of TLMM-related hardware blocks, but the upstream binding only captures the particular block related to controlling the TLMM pins from an OS. In the translation of the driver from downstream, the offset of 0x100000 was lost for the UFS and SDC pingroups.
--- drivers/pinctrl/qcom/pinctrl-sm8350.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-)
drivers/pinctrl/qcom/pinctrl-sm8350.c +4 -4
@@ -1597,10 +1597,10 @@ static const struct msm_pingroup sm8350_groups[] = { [200] = PINGROUP(200, qdss_gpio, _, _, _, _, _, _, _, _), [201] = PINGROUP(201, _, _, _, _, _, _, _, _, _), [202] = PINGROUP(202, _, _, _, _, _, _, _, _, _), - [203] = UFS_RESET(ufs_reset, 0x1d8000), - [204] = SDC_PINGROUP(sdc2_clk, 0x1cf000, 14, 6), - [205] = SDC_PINGROUP(sdc2_cmd, 0x1cf000, 11, 3), - [206] = SDC_PINGROUP(sdc2_data, 0x1cf000, 9, 0), + [203] = UFS_RESET(ufs_reset, 0xd8000), + [204] = SDC_PINGROUP(sdc2_clk, 0xcf000, 14, 6), + [205] = SDC_PINGROUP(sdc2_cmd, 0xcf000, 11, 3), + [206] = SDC_PINGROUP(sdc2_data, 0xcf000, 9, 0), };static const struct msm_gpio_wakeirq_map sm8350_pdc_map[] = {
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 182/279 perf/x86/intel/uncore: Fix filter_tid mask for CHA events on Skylake Server
In reply to Greg Kroah-HartmanFrom: Alexander Antonov <[email protected]>
[ Upstream commit e324234e0aa881b7841c7c713306403e12b069ff ]
According Uncore Reference Manual: any of the CHA events may be filtered by Thread/Core-ID by using tid modifier in CHA Filter 0 Register. Update skx_cha_hw_config() to follow Uncore Guide.
--- arch/x86/events/intel/uncore_snbep.c | 3 +++ 1 file changed, 3 insertions(+)
arch/x86/events/intel/uncore_snbep.c +3 -0
@@ -3608,6 +3608,9 @@ static int skx_cha_hw_config(struct intel_uncore_box *box, struct perf_event *evstruct hw_perf_event_extra *reg1 = &event->hw.extra_reg;struct extra_reg *er;int idx =0 ; +/* Any of the CHA events may be filtered by Thread/Core-ID.*/ +if (event->hw.config &SNBEP_CBO_PMON_CTL_TID_EN ) + idx =SKX_CHA_MSR_PMON_BOX_FILTER_TID ;for (er = skx_uncore_cha_extra_regs; er->msr; er++) {if (er->event != (event->hw.config & er->config_mask))
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 183/279 perf/x86/intel/uncore: Fix IIO event constraints for Skylake Server
In reply to Greg Kroah-HartmanFrom: Alexander Antonov <[email protected]>
[ Upstream commit 3866ae319c846a612109c008f43cba80b8c15e86 ]
According to the latest uncore document, COMP_BUF_OCCUPANCY (0xd5) event can be collected on 2-3 counters. Update uncore IIO event constraints for Skylake Server.
--- arch/x86/events/intel/uncore_snbep.c | 1 + 1 file changed, 1 insertion(+)
arch/x86/events/intel/uncore_snbep.c +1 -0
@@ -3678,6 +3678,7 @@ static struct event_constraint skx_uncore_iio_constraints[] = {UNCORE_EVENT_CONSTRAINT (0xc0 ,0xc ),UNCORE_EVENT_CONSTRAINT (0xc5 ,0xc ),UNCORE_EVENT_CONSTRAINT (0xd4 ,0xc ), +UNCORE_EVENT_CONSTRAINT (0xd5 ,0xc ), EVENT_CONSTRAINT_END };
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 184/279 perf/x86/intel/uncore: Fix IIO event constraints for Snowridge
In reply to Greg Kroah-HartmanFrom: Alexander Antonov <[email protected]>
[ Upstream commit bdc0feee05174418dec1fa68de2af19e1750b99f ]
According to the latest uncore document, DATA_REQ_OF_CPU (0x83), DATA_REQ_BY_CPU (0xc0) and COMP_BUF_OCCUPANCY (0xd5) events have constraints. Add uncore IIO constraints for Snowridge.
--- arch/x86/events/intel/uncore_snbep.c | 8 ++++++++ 1 file changed, 8 insertions(+)
arch/x86/events/intel/uncore_snbep.c +8 -0
@@ -4529,6 +4529,13 @@ static void snr_iio_cleanup_mapping(struct intel_uncore_type *type)pmu_iio_cleanup_mapping (type, &snr_iio_mapping_group); } +static struct event_constraint snr_uncore_iio_constraints[] = { +UNCORE_EVENT_CONSTRAINT (0x83 ,0x3 ), +UNCORE_EVENT_CONSTRAINT (0xc0 ,0xc ), +UNCORE_EVENT_CONSTRAINT (0xd5 ,0xc ), +EVENT_CONSTRAINT_END +}; +static struct intel_uncore_type snr_uncore_iio = { .name ="iio" , .num_counters =4 ,
@@ -4540,6 +4547,7 @@ static struct intel_uncore_type snr_uncore_iio = { .event_mask_ext = SNR_IIO_PMON_RAW_EVENT_MASK_EXT, .box_ctl = SNR_IIO_MSR_PMON_BOX_CTL, .msr_offset = SNR_IIO_MSR_OFFSET, + .constraints = snr_uncore_iio_constraints, .ops = &ivbep_uncore_msr_ops, .format_group = &snr_uncore_iio_format_group, .attr_update = snr_iio_attr_update,
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 185/279 s390/kexec: fix return code handling
In reply to Greg Kroah-HartmanFrom: Heiko Carstens <[email protected]>
[ Upstream commit 20c76e242e7025bd355619ba67beb243ba1a1e95 ]
kexec_file_add_ipl_report ignores that ipl_report_finish may fail and can return an error pointer instead of a valid pointer. Fix this and simplify by returning NULL in case of an error and let the only caller handle this case.
--- arch/s390/kernel/ipl.c | 3 ++- arch/s390/kernel/machine_kexec_file.c | 8 +++++++- 2 files changed, 9 insertions(+), 2 deletions(-)
arch/s390/kernel/ipl.c +2 -1
@@ -2156,7 +2156,7 @@ void *ipl_report_finish(struct ipl_report *report) buf =vzalloc (report->size);if (!buf) -return ERR_PTR (-ENOMEM ); + goto out; ptr = buf;memcpy (ptr, report->ipib, report->ipib->hdr.len);
@@ -2195,6 +2195,7 @@ void *ipl_report_finish(struct ipl_report *report) }BUG_ON (ptr > buf + report->size); +out:return buf; }
arch/s390/kernel/machine_kexec_file.c +7 -1
@@ -170,6 +170,7 @@ static int kexec_file_add_ipl_report(struct kimage *image,struct kexec_buf buf;unsigned long addr;void *ptr, *end; +int ret; buf.image = image;
@@ -199,7 +200,10 @@ static int kexec_file_add_ipl_report(struct kimage *image, ptr += len; } + ret = -ENOMEM ; buf.buffer =ipl_report_finish (data->report); +if (!buf.buffer) + goto out; buf.bufsz = data->report->size; buf.memsz = buf.bufsz;
@@ -209,7 +213,9 @@ static int kexec_file_add_ipl_report(struct kimage *image, data->kernel_buf + offsetof(struct lowcore , ipl_parmblock_ptr); *lc_ipl_parmblock_ptr = (__u32 )buf.mem; -return kexec_add_buffer (&buf); + ret =kexec_add_buffer (&buf); +out: +return ret; } void *kexec_file_add_components(struct kimage *image,
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 186/279 blk-cgroup: fix missing put device in error path from blkg_conf_pref()
In reply to Greg Kroah-HartmanFrom: Yu Kuai <[email protected]>
[ Upstream commit 15c30104965101b8e76b24d27035569d6613a7d6 ]
If blk_queue_enter() failed due to queue is dying, the blkdev_put_no_open() is needed because blkcg_conf_open_bdev() succeeded.
--- block/blk-cgroup.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-)
block/blk-cgroup.c +5 -4
@@ -639,7 +639,7 @@ int blkg_conf_prep(struct blkcg *blkcg, const struct blkcg_policy *pol,*/ ret =blk_queue_enter (q,0 );if (ret) -return ret; + goto fail;rcu_read_lock ();spin_lock_irq (&q->queue_lock);
@@ -675,13 +675,13 @@ int blkg_conf_prep(struct blkcg *blkcg, const struct blkcg_policy *pol, new_blkg =blkg_alloc (pos, q,GFP_KERNEL );if (unlikely (!new_blkg)) { ret = -ENOMEM ; - goto fail; + goto fail_exit_queue; }if (radix_tree_preload (GFP_KERNEL )) {blkg_free (new_blkg); ret = -ENOMEM ; - goto fail; + goto fail_exit_queue; }rcu_read_lock ();
@@ -721,9 +721,10 @@ fail_preloaded: fail_unlock:spin_unlock_irq (&q->queue_lock);rcu_read_unlock (); +fail_exit_queue: +blk_queue_exit (q); fail:blkdev_put_no_open (bdev); -blk_queue_exit (q);/* * If queue was bypassing, we should retry. Do so after a * short msleep(). It isn't strictly necessary but queue
Signature · 1 line
2.33.0
Greg Kroah-Hartman
maintainerPATCH 187/279 dmaengine: remove debugfs #ifdef
In reply to Greg Kroah-HartmanFrom: Arnd Bergmann <[email protected]>
commit b3b180e735409ca0c76642014304b59482e0e653 upstream.
The ptdma driver has added debugfs support, but this fails to build when debugfs is disabled:
drivers/dma/ptdma/ptdma-debugfs.c: In function 'ptdma_debugfs_setup':drivers/dma/ptdma/ptdma-debugfs.c:93:54: error: 'struct dma_device' has no member named 'dbg_dev_root'
93 | debugfs_create_file("info", 0400, pt->dma_dev.dbg_dev_root, pt,
| ^drivers/dma/ptdma/ptdma-debugfs.c:96:55: error: 'struct dma_device' has no member named 'dbg_dev_root'
96 | debugfs_create_file("stats", 0400, pt->dma_dev.dbg_dev_root, pt,
| ^drivers/dma/ptdma/ptdma-debugfs.c:102:52: error: 'struct dma_device' has no member named 'dbg_dev_root'
102 | debugfs_create_dir("q", pt->dma_dev.dbg_dev_root);
| ^
Remove the #ifdef in the header, as this only saves a few bytes, but would require ugly #ifdefs in each driver using it. Simplify the other user while we're at it.
--- drivers/dma/xilinx/xilinx_dpdma.c | 15 +-------------- include/linux/dmaengine.h | 2 -- 2 files changed, 1 insertion(+), 16 deletions(-)
drivers/dma/xilinx/xilinx_dpdma.c +1 -14
@@ -271,9 +271,6 @@ struct xilinx_dpdma_device {/* ----------------------------------------------------------------------------- * DebugFS */ - -#ifdef CONFIG_DEBUG_FS -#define XILINX_DPDMA_DEBUGFS_READ_MAX_SIZE 32#define XILINX_DPDMA_DEBUGFS_UINT16_MAX_STR "65535"
@@ -299,7 +296,7 @@ struct xilinx_dpdma_debugfs_request {static void xilinx_dpdma_debugfs_desc_done_irq (struct xilinx_dpdma_chan *chan) { -if (chan->id == dpdma_debugfs.chan_id) +if (IS_ENABLED (CONFIG_DEBUG_FS ) && chan->id == dpdma_debugfs.chan_id) dpdma_debugfs.xilinx_dpdma_irq_done_count++; }
@@ -462,16 +459,6 @@ static void xilinx_dpdma_debugfs_init(stdev_err (xdev->dev,"Failed to create debugfs testcase file\n" ); } -#else -static void xilinx_dpdma_debugfs_init (struct xilinx_dpdma_device *xdev) -{ -} - -static void xilinx_dpdma_debugfs_desc_done_irq (struct xilinx_dpdma_chan *chan) -{ -} -#endif /* CONFIG_DEBUG_FS */ -/* ----------------------------------------------------------------------------- * I/O Accessors */
include/linux/dmaengine.h +0 -2
@@ -944,10 +944,8 @@ struct dma_device {void (*device_issue_pending)(struct dma_chan *chan);void (*device_release)(struct dma_device *dev);/* debugfs support */ -#ifdef CONFIG_DEBUG_FS void (*dbg_summary_show)(struct seq_file *s,struct dma_device *dev);struct dentry *dbg_dev_root; -#endif }; static inline int dmaengine_slave_config(struct dma_chan *chan,
Greg Kroah-Hartman
maintainerPATCH 188/279 tun: fix bonding active backup with arp monitoring
In reply to Greg Kroah-HartmanFrom: Nicolas Dichtel <[email protected]>
commit a31d27fbed5d518734cb60956303eb15089a7634 upstream.
As stated in the bonding doc, trans_start must be set manually for drivers using NETIF_F_LLTX:
Drivers that use NETIF_F_LLTX flag must also update netdev_queue->trans_start. If they do not, then the ARP monitor will immediately fail any slaves using that driver, and those slaves will stay down.
--- drivers/net/tun.c | 5 +++++ 1 file changed, 5 insertions(+)
drivers/net/tun.c +5 -0
@@ -1010,6 +1010,7 @@ static netdev_tx_t tun_net_xmit(struct s {struct tun_struct *tun =netdev_priv (dev);int txq = skb->queue_mapping; +struct netdev_queue *queue;struct tun_file *tfile;int len = skb->len;
@@ -1054,6 +1055,10 @@ static netdev_tx_t tun_net_xmit(struct sif (ptr_ring_produce (&tfile->tx_ring, skb)) goto drop; +/* NETIF_F_LLTX requires to do our own update of trans_start */ + queue =netdev_get_tx_queue (dev, txq); + queue->trans_start = jiffies; +/* Notify and wake up reader process */ if (tfile->flags &TUN_FASYNC )kill_fasync (&tfile->fasync,SIGIO ,POLL_IN );
Greg Kroah-Hartman
maintainerPATCH 189/279 Revert "mark pstore-blk as broken"
In reply to Greg Kroah-HartmanFrom: Kees Cook <[email protected]>
commit d1faacbf67b1944f0e0c618dc581d929263f6fe9 upstream.
This reverts commit d07f3b081ee632268786601f55e1334d1f68b997.
pstore-blk was fixed to avoid the unwanted APIs in commit 7bb9557b48fc ("pstore/blk: Use the normal block device I/O path"), which landed in the same release as the commit adding BROKEN.
--- fs/pstore/Kconfig | 1 - 1 file changed, 1 deletion(-)
fs/pstore/Kconfig +0 -1
@@ -173,7 +173,6 @@ config PSTORE_BLK tristate "Log panic/oops to a block device" depends on PSTORE depends on BLOCK - depends on BROKEN select PSTORE_ZONE default n help
Greg Kroah-Hartman
maintainerPATCH 190/279 pstore/blk: Use "%lu" to format unsigned long
In reply to Greg Kroah-HartmanFrom: Geert Uytterhoeven <[email protected]>
commit 61eb495c83bf6ebde490992bf888ca15b9babc39 upstream.
On 32-bit:
fs/pstore/blk.c: In function ‘__best_effort_init’:
include/linux/kern_levels.h:5:18: warning: format ‘%zu’ expects argument of type ‘size_t’, but argument 3 has type ‘long unsigned int’ [-Wformat=]
5 | #define KERN_SOH "\001" /* ASCII Start Of Header */
| ^~~~~~
include/linux/kern_levels.h:14:19: note: in expansion of macro ‘KERN_SOH’
14 | #define KERN_INFO KERN_SOH "6" /* informational */
| ^~~~~~~~
include/linux/printk.h:373:9: note: in expansion of macro ‘KERN_INFO’
373 | printk(KERN_INFO pr_fmt(fmt), ##__VA_ARGS__)
| ^~~~~~~~~
fs/pstore/blk.c:314:3: note: in expansion of macro ‘pr_info’
314 | pr_info("attached %s (%zu) (no dedicated panic_write!)\n",
| ^~~~~~~
--- fs/pstore/blk.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
fs/pstore/blk.c +1 -1
@@ -311,7 +311,7 @@ static int __init __best_effort_init(voiif (ret)kfree (best_effort_dev);else -pr_info ("attached %s (%zu) (no dedicated panic_write!)\n" , +pr_info ("attached %s (%lu) (no dedicated panic_write!)\n" , blkdev, best_effort_dev->zone.total_size);return ret;
Greg Kroah-Hartman
maintainerPATCH 191/279 hexagon: export raw I/O routines for modules
In reply to Greg Kroah-HartmanFrom: Nathan Chancellor <[email protected]>
commit ffb92ce826fd801acb0f4e15b75e4ddf0d189bde upstream.
Patch series "Fixes for ARCH=hexagon allmodconfig", v2.
This series fixes some issues noticed with ARCH=hexagon allmodconfig.
This patch (of 3):
When building ARCH=hexagon allmodconfig, the following errors occur:
ERROR: modpost: "__raw_readsl" [drivers/i3c/master/svc-i3c-master.ko] undefined!
ERROR: modpost: "__raw_writesl" [drivers/i3c/master/dw-i3c-master.ko] undefined!
ERROR: modpost: "__raw_readsl" [drivers/i3c/master/dw-i3c-master.ko] undefined!
ERROR: modpost: "__raw_writesl" [drivers/i3c/master/i3c-master-cdns.ko] undefined!
ERROR: modpost: "__raw_readsl" [drivers/i3c/master/i3c-master-cdns.ko] undefined!
Export these symbols so that modules can use them without any errors.
--- arch/hexagon/lib/io.c | 4 ++++ 1 file changed, 4 insertions(+)
arch/hexagon/lib/io.c +4 -0
@@ -27,6 +27,7 @@ void __raw_readsw(const void __iomem *ad *dst++ = *src; } +EXPORT_SYMBOL (__raw_readsw);/* * __raw_writesw - read words a short at a time
@@ -47,6 +48,7 @@ void __raw_writesw(void __iomem *addr, c } +EXPORT_SYMBOL (__raw_writesw);/* Pretty sure len is pre-adjusted for the length of the access already */ void __raw_readsl (const void __iomem *addr,void *data,int len)
@@ -62,6 +64,7 @@ void __raw_readsl(const void __iomem *ad } +EXPORT_SYMBOL (__raw_readsl);void __raw_writesl (void __iomem *addr,const void *data,int len) {
@@ -76,3 +79,4 @@ void __raw_writesl(void __iomem *addr, c } +EXPORT_SYMBOL (__raw_writesl);
Greg Kroah-Hartman
maintainerPATCH 192/279 hexagon: clean up timer-regs.h
In reply to Greg Kroah-HartmanFrom: Nathan Chancellor <[email protected]>
commit 51f2ec593441d3d1ebc0d478fac3ea329c7c93ac upstream.
When building allmodconfig, there is a warning about TIMER_ENABLE being redefined:
drivers/clocksource/timer-oxnas-rps.c:39:9: error: 'TIMER_ENABLE' macro redefined [-Werror,-Wmacro-redefined]
#define TIMER_ENABLE BIT(7)
^
arch/hexagon/include/asm/timer-regs.h:13:9: note: previous definition is here
#define TIMER_ENABLE 0
^
1 error generated.The values in this header are only used in one file each, if they are used at all. Remove the header and sink all of the constants into their respective files.
TCX0_CLK_RATE is only used in arch/hexagon/include/asm/timex.h
TIMER_ENABLE, RTOS_TIMER_INT, RTOS_TIMER_REGS_ADDR are only used in arch/hexagon/kernel/time.c.
SLEEP_CLK_RATE and TIMER_CLR_ON_MATCH have both been unused since the file's introduction in commit 71e4a47f32f4 ("Hexagon: Add time and timer functions").
TIMER_ENABLE is redefined as BIT(0) so the shift is moved into the definition, rather than its use.
--- arch/hexagon/include/asm/timer-regs.h | 26 -------------------------- arch/hexagon/include/asm/timex.h | 3 +-- arch/hexagon/kernel/time.c | 12 ++++++++++-- 3 files changed, 11 insertions(+), 30 deletions(-) delete mode 100644 arch/hexagon/include/asm/timer-regs.h
arch/hexagon/include/asm/timer-regs.h deleted +0 -26
@@ -1,26 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0-only */ -/* -* Timer support for Hexagon -* -* Copyright (c) 2010-2011, The Linux Foundation. All rights reserved. -*/ - -#ifndef _ASM_TIMER_REGS_H -#define _ASM_TIMER_REGS_H - -/* This stuff should go into a platform specific file */ -#define TCX0_CLK_RATE 19200 -#define TIMER_ENABLE 0 -#define TIMER_CLR_ON_MATCH 1 - -/* -* 8x50 HDD Specs 5-8. Simulator co-sim not fixed until -* release 1.1, and then it's "adjustable" and probably not defaulted. -*/ -#define RTOS_TIMER_INT 3 -#ifdef CONFIG_HEXAGON_COMET -#define RTOS_TIMER_REGS_ADDR 0xAB000000UL -#endif -#define SLEEP_CLK_RATE 32000 - -#endif
arch/hexagon/include/asm/timex.h +1 -2
@@ -7,11 +7,10 @@#define _ASM_TIMEX_H#include <asm-generic/timex.h> -#include <asm/timer-regs.h> #include <asm/hexagon_vm.h> /* Using TCX0 as our clock. CLOCK_TICK_RATE scheduled to be removed. */ -#define CLOCK_TICK_RATE TCX0_CLK_RATE +#define CLOCK_TICK_RATE 19200#define ARCH_HAS_READ_CURRENT_TIMER
arch/hexagon/kernel/time.c +10 -2
@@ -17,9 +17,10 @@#include <linux/of_irq.h> #include <linux/module.h> -#include <asm/timer-regs.h> #include <asm/hexagon_vm.h> +#define TIMER_ENABLE BIT(0) +/* * For the clocksource we need: * pcycle frequency (600MHz)
@@ -33,6 +34,13 @@ cycles_t pcycle_freq_mhz;cycles_t thread_freq_mhz;cycles_t sleep_clk_freq; +/* +* 8x50 HDD Specs 5-8. Simulator co-sim not fixed until +* release 1.1, and then it's "adjustable" and probably not defaulted. +*/ +#define RTOS_TIMER_INT 3 +#define RTOS_TIMER_REGS_ADDR 0xAB000000UL +static struct resource rtos_timer_resources[] = { { .start =RTOS_TIMER_REGS_ADDR ,
@@ -80,7 +88,7 @@ static int set_next_event(unsigned longiowrite32 (0 , &rtos_timer->clear);iowrite32 (delta, &rtos_timer->match); -iowrite32 (1 <<TIMER_ENABLE , &rtos_timer->enable); +iowrite32 (TIMER_ENABLE , &rtos_timer->enable);return 0 ; }
Greg Kroah-Hartman
maintainerPATCH 193/279 tipc: check for null after calling kmemdup
In reply to Greg Kroah-HartmanFrom: Tadeusz Struk <[email protected]>
commit 3e6db079751afd527bf3db32314ae938dc571916 upstream.
kmemdup can return a null pointer so need to check for it, otherwise the null key will be dereferenced later in tipc_crypto_key_xmit as can be seen in the trace [1].
[1] https://syzkaller.appspot.com/bug?id=bca180abb29567b189efdbdb34cbf7ba851c2a58
--- net/tipc/crypto.c | 4 ++++ 1 file changed, 4 insertions(+)
net/tipc/crypto.c +4 -0
@@ -597,6 +597,10 @@ static int tipc_aead_init(struct tipc_ae tmp->cloned =NULL ; tmp->authsize =TIPC_AES_GCM_TAG_SIZE ; tmp->key =kmemdup (ukey,tipc_aead_key_size (ukey),GFP_KERNEL ); +if (!tmp->key) { +tipc_aead_free (&tmp->rcu); +return -ENOMEM ; + }memcpy (&tmp->salt, ukey->key + keylen,TIPC_AES_GCM_SALT_SIZE );atomic_set (&tmp->users,0 );atomic64_set (&tmp->seqno,0 );
Greg Kroah-Hartman
maintainerPATCH 194/279 ipc: WARN if trying to remove ipc object which is absent
In reply to Greg Kroah-HartmanFrom: Alexander Mikhalitsyn <[email protected]>
commit 126e8bee943e9926238c891e2df5b5573aee76bc upstream.
Patch series "shm: shm_rmid_forced feature fixes".
Some time ago I met kernel crash after CRIU restore procedure, fortunately, it was CRIU restore, so, I had dump files and could do restore many times and crash reproduced easily. After some investigation I've constructed the minimal reproducer. It was found that it's use-after-free and it happens only if sysctl kernel.shm_rmid_forced = 1.
The key of the problem is that the exit_shm() function not handles shp's object destroy when task->sysvshm.shm_clist contains items from different IPC namespaces. In most cases this list will contain only items from one IPC namespace.
How can this list contain object from different namespaces? The exit_shm() function is designed to clean up this list always when process leaves IPC namespace. But we made a mistake a long time ago and did not add a exit_shm() call into the setns() syscall procedures.
The first idea was just to add this call to setns() syscall but it obviously changes semantics of setns() syscall and that's userspace-visible change. So, I gave up on this idea.
The first real attempt to address the issue was just to omit forced destroy if we meet shp object not from current task IPC namespace [1]. But that was not the best idea because task->sysvshm.shm_clist was protected by rwsem which belongs to current task IPC namespace. It means that list corruption may occur.
Second approach is just extend exit_shm() to properly handle shp's from different IPC namespaces [2]. This is really non-trivial thing, I've put a lot of effort into that but not believed that it's possible to make it fully safe, clean and clear.
Thanks to the efforts of Manfred Spraul working an elegant solution was designed. Thanks a lot, Manfred!
Eric also suggested the way to address the issue in ("[RFC][PATCH] shm:
In shm_exit destroy all created and never attached segments") Eric's idea was to maintain a list of shm_clists one per IPC namespace, use lock-less lists. But there is some extra memory consumption-related concerns.
An alternative solution which was suggested by me was implemented in ("shm: reset shm_clist on setns but omit forced shm destroy"). The idea is pretty simple, we add exit_shm() syscall to setns() but DO NOT destroy shm segments even if sysctl kernel.shm_rmid_forced = 1, we just clean up the task->sysvshm.shm_clist list.
This chages semantics of setns() syscall a little bit but in comparision to the "naive" solution when we just add exit_shm() without any special exclusions this looks like a safer option.
[1] https://lkml.org/lkml/2021/7/6/1108
[2] https://lkml.org/lkml/2021/7/14/736
This patch (of 2):
Let's produce a warning if we trying to remove non-existing IPC object from IPC namespace kht/idr structures.
This allows us to catch possible bugs when the ipc_rmid() function was called with inconsistent struct ipc_ids*, struct kern_ipc_perm* arguments.
--- ipc/util.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
ipc/util.c +3 -3
@@ -447,8 +447,8 @@ static int ipcget_public(struct ipc_namestatic void ipc_kht_remove (struct ipc_ids *ids,struct kern_ipc_perm *ipcp) {if (ipcp->key !=IPC_PRIVATE ) -rhashtable_remove_fast (&ids->key_ht, &ipcp->khtnode, - ipc_kht_params); +WARN_ON_ONCE (rhashtable_remove_fast (&ids->key_ht, &ipcp->khtnode, + ipc_kht_params)); }/**
@@ -498,7 +498,7 @@ void ipc_rmid(struct ipc_ids *ids, struc {int idx =ipcid_to_idx (ipcp->id); -idr_remove (&ids->ipcs_idr, idx); +WARN_ON_ONCE (idr_remove (&ids->ipcs_idr, idx) != ipcp);ipc_kht_remove (ids, ipcp); ids->in_use--; ipcp->deleted = true;
Greg Kroah-Hartman
maintainerPATCH 195/279 shm: extend forced shm destroy to support objects from several IPC nses
In reply to Greg Kroah-HartmanFrom: Alexander Mikhalitsyn <[email protected]>
commit 85b6d24646e4125c591639841169baa98a2da503 upstream.
Currently, the exit_shm() function not designed to work properly when task->sysvshm.shm_clist holds shm objects from different IPC namespaces.
This is a real pain when sysctl kernel.shm_rmid_forced = 1, because it leads to use-after-free (reproducer exists).
This is an attempt to fix the problem by extending exit_shm mechanism to handle shm's destroy from several IPC ns'es.
To achieve that we do several things:
1. add a namespace (non-refcounted) pointer to the struct shmid_kernel
2. during new shm object creation (newseg()/shmget syscall) we initialize this pointer by current task IPC ns
3. exit_shm() fully reworked such that it traverses over all shp's in task->sysvshm.shm_clist and gets IPC namespace not from current task as it was before but from shp's object itself, then call shm_destroy(shp, ns).
Note: We need to be really careful here, because as it was said before (1), our pointer to IPC ns non-refcnt'ed. To be on the safe side we using special helper get_ipc_ns_not_zero() which allows to get IPC ns refcounter only if IPC ns not in the "state of destruction".
Q/A
Q: Why can we access shp->ns memory using non-refcounted pointer?
A: Because shp object lifetime is always shorther than IPC namespace
lifetime, so, if we get shp object from the task->sysvshm.shm_clist while holding task_lock(task) nobody can steal our namespace.
Q: Does this patch change semantics of unshare/setns/clone syscalls?
A: No. It's just fixes non-covered case when process may leave IPC
namespace without getting task->sysvshm.shm_clist list cleaned up.
--- include/linux/ipc_namespace.h | 15 +++ include/linux/sched/task.h | 2 ipc/shm.c | 189 +++++++++++++++++++++++++++++++----------- 3 files changed, 159 insertions(+), 47 deletions(-)
include/linux/ipc_namespace.h +15 -0
@@ -131,6 +131,16 @@ static inline struct ipc_namespace *get_return ns; } +static inline struct ipc_namespace *get_ipc_ns_not_zero (struct ipc_namespace *ns) +{ +if (ns) { +if (refcount_inc_not_zero (&ns->ns.count)) +return ns; + } + +return NULL ; +} +extern void put_ipc_ns (struct ipc_namespace *ns);#else static inline struct ipc_namespace *copy_ipcs(unsigned long flags,
@@ -146,6 +156,11 @@ static inline struct ipc_namespace *get_ {return ns; } + +static inline struct ipc_namespace *get_ipc_ns_not_zero (struct ipc_namespace *ns) +{ +return ns; +}static inline void put_ipc_ns (struct ipc_namespace *ns) {
include/linux/sched/task.h +1 -1
@@ -158,7 +158,7 @@ static inline struct vm_struct *task_sta* Protects ->fs, ->files, ->mm, ->group_info, ->comm, keyring * subscriptions and synchronises with wait4(). Also used in procfs. Also * pins the final release of task.io_context. Also protects ->cpuset and -* ->cgroup.subsys[]. And ->vfork_done. +* ->cgroup.subsys[]. And ->vfork_done. And ->sysvshm.shm_clist. * * Nests both inside and outside of read_lock(&tasklist_lock). * It must not be nested with write_lock_irq(&tasklist_lock),
ipc/shm.c +143 -46
264 lines not on this page · raw message
Greg Kroah-Hartman
maintainerPATCH 196/279 mm: kmemleak: slob: respect SLAB_NOLEAKTRACE flag
In reply to Greg Kroah-HartmanFrom: Rustam Kovhaev <[email protected]>
commit 34dbc3aaf5d9e89ba6cc5e24add9458c21ab1950 upstream.
When kmemleak is enabled for SLOB, system does not boot and does not print anything to the console. At the very early stage in the boot process we hit infinite recursion from kmemleak_init() and eventually kernel crashes.
kmemleak_init() specifies SLAB_NOLEAKTRACE for KMEM_CACHE(), but kmem_cache_create_usercopy() removes it because CACHE_CREATE_MASK is not valid for SLOB.
Let's fix CACHE_CREATE_MASK and make kmemleak work with SLOB
--- mm/slab.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
mm/slab.h +1 -1
@@ -147,7 +147,7 @@ static inline slab_flags_t kmem_cache_fl#define SLAB_CACHE_FLAGS (SLAB_NOLEAKTRACE | SLAB_RECLAIM_ACCOUNT | \ SLAB_TEMPORARY | SLAB_ACCOUNT)#else -#define SLAB_CACHE_FLAGS (0) +#define SLAB_CACHE_FLAGS (SLAB_NOLEAKTRACE)#endif /* Common flags available with current configuration */
Greg Kroah-Hartman
maintainerPATCH 197/279 hugetlb, userfaultfd: fix reservation restore on userfaultfd error
In reply to Greg Kroah-HartmanFrom: Mina Almasry <[email protected]>
commit cc30042df6fcc82ea18acf0dace831503e60a0b7 upstream.
Currently in the is_continue case in hugetlb_mcopy_atomic_pte(), if we bail out using "goto out_release_unlock;" in the cases where idx >= size, or !huge_pte_none(), the code will detect that new_pagecache_page == false, and so call restore_reserve_on_error(). In this case I see restore_reserve_on_error() delete the reservation, and the following call to remove_inode_hugepages() will increment h->resv_hugepages causing a 100% reproducible leak.
We should treat the is_continue case similar to adding a page into the pagecache and set new_pagecache_page to true, to indicate that there is no reservation to restore on the error path, and we need not call restore_reserve_on_error(). Rename new_pagecache_page to page_in_pagecache to make that clear.
--- mm/hugetlb.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-)
mm/hugetlb.c +4 -3
@@ -5236,13 +5236,14 @@ int hugetlb_mcopy_atomic_pte(struct mm_sint ret = -ENOMEM ;struct page *page;int writable; -bool new_pagecache_page = false; +bool page_in_pagecache = false;if (is_continue) { ret = -EFAULT ; page =find_lock_page (mapping, idx);if (!page) goto out; + page_in_pagecache = true; }else if (!*pagep) {/* If a page already exists, then it's UFFDIO_COPY for * a non-missing case. Return -EEXIST.
@@ -5330,7 +5331,7 @@ int hugetlb_mcopy_atomic_pte(struct mm_s ret =huge_add_to_page_cache (page, mapping, idx);if (ret) goto out_release_nounlock; - new_pagecache_page = true; + page_in_pagecache = true; } ptl =huge_pte_lockptr (h, dst_mm, dst_pte);
@@ -5394,7 +5395,7 @@ out_release_unlock:if (vm_shared || is_continue)unlock_page (page); out_release_nounlock: -if (!new_pagecache_page) +if (!page_in_pagecache)restore_reserve_on_error (h, dst_vma, dst_addr, page);put_page (page); goto out;
Greg Kroah-Hartman
maintainerPATCH 198/279 kmap_local: dont assume kmap PTEs are linear arrays in memory
In reply to Greg Kroah-HartmanFrom: Ard Biesheuvel <[email protected]>
commit 825c43f50e3aa811a291ffcb40e02fbf6d91ba86 upstream.
The kmap_local conversion broke the ARM architecture, because the new code assumes that all PTEs used for creating kmaps form a linear array in memory, and uses array indexing to look up the kmap PTE belonging to a certain kmap index.
On ARM, this cannot work, not only because the PTE pages may be non-adjacent in memory, but also because ARM/!LPAE interleaves hardware entries and extended entries (carrying software-only bits) in a way that is not compatible with array indexing.
Fortunately, this only seems to affect configurations with more than 8 CPUs, due to the way the per-CPU kmap slots are organized in memory.
Work around this by permitting an architecture to set a Kconfig symbol that signifies that the kmap PTEs do not form a lineary array in memory, and so the only way to locate the appropriate one is to walk the page tables.
--- arch/arm/Kconfig | 1 + mm/Kconfig | 3 +++ mm/highmem.c | 32 +++++++++++++++++++++----------- 3 files changed, 25 insertions(+), 11 deletions(-)
arch/arm/Kconfig +1 -0
@@ -1455,6 +1455,7 @@ config HIGHMEM bool "High Memory Support" depends on MMU select KMAP_LOCAL + select KMAP_LOCAL_NON_LINEAR_PTE_ARRAY help The address space of ARM processors is only 4 Gigabytes large and it has to accommodate user address space, kernel address
mm/Kconfig +3 -0
@@ -887,6 +887,9 @@ config MAPPING_DIRTY_HELPERS config KMAP_LOCAL bool +config KMAP_LOCAL_NON_LINEAR_PTE_ARRAY + bool + # struct io_mapping based helper. Selected by drivers that need them config IO_MAPPING bool
mm/highmem.c +21 -11
@@ -504,16 +504,22 @@ static inline int kmap_local_calc_idx(instatic pte_t *__kmap_pte; -static pte_t *kmap_get_pte (void ) +static pte_t *kmap_get_pte (unsigned long vaddr,int idx) { +if (IS_ENABLED (CONFIG_KMAP_LOCAL_NON_LINEAR_PTE_ARRAY )) +/* +* Set by the arch if __kmap_pte[-idx] does not produce +* the correct entry. +*/ +return virt_to_kpte (vaddr);if (!__kmap_pte) __kmap_pte =virt_to_kpte (__fix_to_virt (FIX_KMAP_BEGIN )); -return __kmap_pte; +return &__kmap_pte[-idx]; }void *__kmap_local_pfn_prot (unsigned long pfn,pgprot_t prot) { -pte_t pteval, *kmap_pte =kmap_get_pte (); +pte_t pteval, *kmap_pte;unsigned long vaddr;int idx;
@@ -525,9 +531,10 @@ void *__kmap_local_pfn_prot(unsigned lonpreempt_disable (); idx =arch_kmap_local_map_idx (kmap_local_idx_push (), pfn); vaddr =__fix_to_virt (FIX_KMAP_BEGIN + idx); -BUG_ON (!pte_none (*(kmap_pte - idx))); + kmap_pte =kmap_get_pte (vaddr, idx); +BUG_ON (!pte_none (*kmap_pte)); pteval =pfn_pte (pfn, prot); -arch_kmap_local_set_pte (&init_mm, vaddr, kmap_pte - idx, pteval); +arch_kmap_local_set_pte (&init_mm, vaddr, kmap_pte, pteval);arch_kmap_local_post_map (vaddr, pteval); current->kmap_ctrl.pteval[kmap_local_idx ()] = pteval;preempt_enable ();
@@ -560,7 +567,7 @@ EXPORT_SYMBOL(__kmap_local_page_prot);void kunmap_local_indexed (void *vaddr) {unsigned long addr = (unsigned long ) vaddr &PAGE_MASK ; -pte_t *kmap_pte =kmap_get_pte (); +pte_t *kmap_pte;int idx; if (addr < __fix_to_virt(FIX_KMAP_END) ||
@@ -585,8 +592,9 @@ void kunmap_local_indexed(void *vaddr) idx =arch_kmap_local_unmap_idx (kmap_local_idx (), addr);WARN_ON_ONCE (addr !=__fix_to_virt (FIX_KMAP_BEGIN + idx)); + kmap_pte =kmap_get_pte (addr, idx);arch_kmap_local_pre_unmap (addr); -pte_clear (&init_mm, addr, kmap_pte - idx); +pte_clear (&init_mm, addr, kmap_pte);arch_kmap_local_post_unmap (addr); current->kmap_ctrl.pteval[kmap_local_idx ()] =__pte (0 );kmap_local_idx_pop ();
@@ -608,7 +616,7 @@ EXPORT_SYMBOL(kunmap_local_indexed);void __kmap_local_sched_out (void ) {struct task_struct *tsk = current; -pte_t *kmap_pte =kmap_get_pte (); +pte_t *kmap_pte;int i;/* Clear kmaps */
@@ -635,8 +643,9 @@ void __kmap_local_sched_out(void) idx =arch_kmap_local_map_idx (i,pte_pfn (pteval)); addr =__fix_to_virt (FIX_KMAP_BEGIN + idx); + kmap_pte =kmap_get_pte (addr, idx);arch_kmap_local_pre_unmap (addr); -pte_clear (&init_mm, addr, kmap_pte - idx); +pte_clear (&init_mm, addr, kmap_pte);arch_kmap_local_post_unmap (addr); } }
@@ -644,7 +653,7 @@ void __kmap_local_sched_out(void)void __kmap_local_sched_in (void ) {struct task_struct *tsk = current; -pte_t *kmap_pte =kmap_get_pte (); +pte_t *kmap_pte;int i;/* Restore kmaps */
@@ -664,7 +673,8 @@ void __kmap_local_sched_in(void)/* See comment in __kmap_local_sched_out() */ idx =arch_kmap_local_map_idx (i,pte_pfn (pteval)); addr =__fix_to_virt (FIX_KMAP_BEGIN + idx); -set_pte_at (&init_mm, addr, kmap_pte - idx, pteval); + kmap_pte =kmap_get_pte (addr, idx); +set_pte_at (&init_mm, addr, kmap_pte, pteval);arch_kmap_local_post_map (addr, pteval); } }
Greg Kroah-Hartman
maintainerPATCH 199/279 mm/damon/dbgfs: use __GFP_NOWARN for user-specified size buffer allocation
In reply to Greg Kroah-HartmanFrom: SeongJae Park <[email protected]>
commit db7a347b26fe05d2e8c115bb24dfd908d0252bc3 upstream.
Patch series "DAMON fixes".
This patch (of 2):
DAMON users can trigger below warning in '__alloc_pages()' by invoking write() to some DAMON debugfs files with arbitrarily high count argument, because DAMON debugfs interface allocates some buffers based on the user-specified 'count'.
if (unlikely(order >= MAX_ORDER)) {
WARN_ON_ONCE(!(gfp & __GFP_NOWARN));
return NULL;
}Because the DAMON debugfs interface code checks failure of the 'kmalloc()', this commit simply suppresses the warnings by adding '__GFP_NOWARN' flag.
--- mm/damon/dbgfs.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
mm/damon/dbgfs.c +2 -2
@@ -32,7 +32,7 @@ static char *user_input_str(const char _if (*ppos)return ERR_PTR (-EINVAL ); - kbuf =kmalloc (count +1 ,GFP_KERNEL ); + kbuf =kmalloc (count +1 ,GFP_KERNEL | __GFP_NOWARN);if (!kbuf)return ERR_PTR (-ENOMEM );
@@ -247,7 +247,7 @@ static ssize_t dbgfs_kdamond_pid_read(stchar *kbuf;ssize_t len; - kbuf =kmalloc (count,GFP_KERNEL ); + kbuf =kmalloc (count,GFP_KERNEL | __GFP_NOWARN);if (!kbuf)return -ENOMEM ;
Greg Kroah-Hartman
maintainerPATCH 200/279 mm/damon/dbgfs: fix missed use of damon_dbgfs_lock
In reply to Greg Kroah-HartmanFrom: SeongJae Park <[email protected]>
commit d78f3853f831eee46c6dbe726debf3be9e9c0d05 upstream.
DAMON debugfs is supposed to protect dbgfs_ctxs, dbgfs_nr_ctxs, and dbgfs_dirs using damon_dbgfs_lock. However, some of the code is accessing the variables without the protection. This fixes it by protecting all such accesses.
--- mm/damon/dbgfs.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-)
mm/damon/dbgfs.c +8 -3
@@ -538,12 +538,14 @@ static ssize_t dbgfs_monitor_on_write(streturn -EINVAL ; } +mutex_lock (&damon_dbgfs_lock);if (!strncmp (kbuf,"on" , count)) err =damon_start (dbgfs_ctxs, dbgfs_nr_ctxs);else if (!strncmp (kbuf,"off" , count)) err =damon_stop (dbgfs_ctxs, dbgfs_nr_ctxs);else err = -EINVAL ; +mutex_unlock (&damon_dbgfs_lock);if (err) ret = err;
@@ -596,15 +598,16 @@ static int __init __damon_dbgfs_init(voistatic int __init damon_dbgfs_init (void ) { -int rc; +int rc = -ENOMEM ; +mutex_lock (&damon_dbgfs_lock); dbgfs_ctxs =kmalloc (sizeof (*dbgfs_ctxs),GFP_KERNEL );if (!dbgfs_ctxs) -return -ENOMEM ; + goto out; dbgfs_ctxs[0 ] =dbgfs_new_ctx ();if (!dbgfs_ctxs[0 ]) {kfree (dbgfs_ctxs); -return -ENOMEM ; + goto out; } dbgfs_nr_ctxs =1 ;
@@ -615,6 +618,8 @@ static int __init damon_dbgfs_init(void)pr_err ("%s: dbgfs init failed\n" , __func__); } +out: +mutex_unlock (&damon_dbgfs_lock);return rc; }
Greg Kroah-Hartman
maintainerPATCH 201/279 x86/boot: Pull up cmdline preparation and early param parsing
In reply to Greg Kroah-HartmanFrom: Borislav Petkov <[email protected]>
commit 8d48bf8206f77aa8687f0e241e901e5197e52423 upstream.
Dan reports that Anjaneya Chagam can no longer use the efi=nosoftreserve kernel command line parameter to suppress "soft reservation" behavior.
This is due to the fact that the following call-chain happens at boot:
early_reserve_memory
|-> efi_memblock_x86_reserve_range
|-> efi_fake_memmap_early
which does
if (!efi_soft_reserve_enabled())
return;and that would have set EFI_MEM_NO_SOFT_RESERVE after having parsed "nosoftreserve".
However, parse_early_param() gets called *after* it, leading to the boot cmdline not being taken into account.
Therefore, carve out the command line preparation into a separate function which does the early param parsing too. So that it all goes together.
And then call that function before early_reserve_memory() so that the params would have been parsed by then.
--- arch/x86/kernel/setup.c | 66 ++++++++++++++++++++++++++++-------------------- 1 file changed, 39 insertions(+), 27 deletions(-)
arch/x86/kernel/setup.c +39 -27
@@ -742,6 +742,28 @@ dump_kernel_offset(struct notifier_blockreturn 0 ; } +static char *prepare_command_line (void ) +{ +#ifdef CONFIG_CMDLINE_BOOL +#ifdef CONFIG_CMDLINE_OVERRIDE +strlcpy (boot_command_line, builtin_cmdline,COMMAND_LINE_SIZE ); +#else +if (builtin_cmdline[0 ]) { +/* append boot loader cmdline to builtin */ +strlcat (builtin_cmdline," " ,COMMAND_LINE_SIZE ); +strlcat (builtin_cmdline, boot_command_line,COMMAND_LINE_SIZE ); +strlcpy (boot_command_line, builtin_cmdline,COMMAND_LINE_SIZE ); + } +#endif +#endif + +strlcpy (command_line, boot_command_line,COMMAND_LINE_SIZE ); + +parse_early_param (); + +return command_line; +} +/* * Determine if we were loaded by an EFI loader. If so, then we have also been * passed the efi memmap, systab, etc., so we should use these data structures
@@ -831,6 +853,23 @@ void __init setup_arch(char **cmdline_p) x86_init.oem.arch_setup ();/* +* x86_configure_nx() is called before parse_early_param() (called by +* prepare_command_line()) to detect whether hardware doesn't support +* NX (so that the early EHCI debug console setup can safely call +* set_fixmap()). It may then be called again from within noexec_setup() +* during parsing early parameters to honor the respective command line +* option. +*/ +x86_configure_nx (); + +/* +* This parses early params and it needs to run before +* early_reserve_memory() because latter relies on such settings +* supplied as early params. +*/ + *cmdline_p =prepare_command_line (); + +/* * Do some memory reservations *before* memory is added to memblock, so * memblock allocations won't overwrite it. *
@@ -863,33 +902,6 @@ void __init setup_arch(char **cmdline_p) bss_resource.start =__pa_symbol (__bss_start); bss_resource.end =__pa_symbol (__bss_stop)-1 ; -#ifdef CONFIG_CMDLINE_BOOL -#ifdef CONFIG_CMDLINE_OVERRIDE -strlcpy (boot_command_line, builtin_cmdline,COMMAND_LINE_SIZE ); -#else -if (builtin_cmdline[0 ]) { -/* append boot loader cmdline to builtin */ -strlcat (builtin_cmdline," " ,COMMAND_LINE_SIZE ); -strlcat (builtin_cmdline, boot_command_line,COMMAND_LINE_SIZE ); -strlcpy (boot_command_line, builtin_cmdline,COMMAND_LINE_SIZE ); - } -#endif -#endif - -strlcpy (command_line, boot_command_line,COMMAND_LINE_SIZE ); - *cmdline_p = command_line; - -/* -* x86_configure_nx() is called before parse_early_param() to detect -* whether hardware doesn't support NX (so that the early EHCI debug -* console setup can safely call set_fixmap()). It may then be called -* again from within noexec_setup() during parsing early parameters -* to honor the respective command line option. -*/ -x86_configure_nx (); - -parse_early_param (); -#ifdef CONFIG_MEMORY_HOTPLUG /* * Memory used by the kernel cannot be hot-removed because Linux
Greg Kroah-Hartman
maintainerPATCH 202/279 x86/sgx: Fix free page accounting
In reply to Greg Kroah-HartmanFrom: Reinette Chatre <[email protected]>
commit ac5d272a0ad0419f52e08c91953356e32b075af7 upstream.
The SGX driver maintains a single global free page counter, sgx_nr_free_pages, that reflects the number of free pages available across all NUMA nodes. Correspondingly, a list of free pages is associated with each NUMA node and sgx_nr_free_pages is updated every time a page is added or removed from any of the free page lists. The main usage of sgx_nr_free_pages is by the reclaimer that runs when it (sgx_nr_free_pages) goes below a watermark to ensure that there are always some free pages available to, for example, support efficient page faults.
With sgx_nr_free_pages accessed and modified from a few places it is essential to ensure that these accesses are done safely but this is not the case. sgx_nr_free_pages is read without any protection and updated with inconsistent protection by any one of the spin locks associated with the individual NUMA nodes. For example:
CPU_A CPU_B
----- -----
spin_lock(&nodeA->lock); spin_lock(&nodeB->lock); ... ... sgx_nr_free_pages--; /* NOT SAFE */ sgx_nr_free_pages--;
spin_unlock(&nodeA->lock); spin_unlock(&nodeB->lock);
Since sgx_nr_free_pages may be protected by different spin locks while being modified from different CPUs, the following scenario is possible:
CPU_A CPU_B
----- -----
{sgx_nr_free_pages = 100}
spin_lock(&nodeA->lock); spin_lock(&nodeB->lock); sgx_nr_free_pages--; sgx_nr_free_pages--; /* LOAD sgx_nr_free_pages = 100 */ /* LOAD sgx_nr_free_pages = 100 */ /* sgx_nr_free_pages-- */ /* sgx_nr_free_pages-- */ /* STORE sgx_nr_free_pages = 99 */ /* STORE sgx_nr_free_pages = 99 */ spin_unlock(&nodeA->lock); spin_unlock(&nodeB->lock);
In the above scenario, sgx_nr_free_pages is decremented from two CPUs but instead of sgx_nr_free_pages ending with a value that is two less than it started with, it was only decremented by one while the number of free pages were actually reduced by two. The consequence of sgx_nr_free_pages not being protected is that its value may not accurately reflect the actual number of free pages on the system, impacting the availability of free pages in support of many flows.
The problematic scenario is when the reclaimer does not run because it believes there to be sufficient free pages while any attempt to allocate a page fails because there are no free pages available. In the SGX driver the reclaimer's watermark is only 32 pages so after encountering the above example scenario 32 times a user space hang is possible when there are no more free pages because of repeated page faults caused by no free pages made available.
The following flow was encountered:asm_exc_page_fault
...
sgx_vma_fault()
sgx_encl_load_page()
sgx_encl_eldu() // Encrypted page needs to be loaded from backing
// storage into newly allocated SGX memory page
sgx_alloc_epc_page() // Allocate a page of SGX memory
__sgx_alloc_epc_page() // Fails, no free SGX memory
...
if (sgx_should_reclaim(SGX_NR_LOW_PAGES)) // Wake reclaimer
wake_up(&ksgxd_waitq);
return -EBUSY; // Return -EBUSY giving reclaimer time to run
return -EBUSY;
return -EBUSY;
return VM_FAULT_NOPAGE;
The reclaimer is triggered in above flow with the following code:
static bool sgx_should_reclaim(unsigned long watermark)
{
return sgx_nr_free_pages < watermark &&
!list_empty(&sgx_active_page_list);
}In the problematic scenario there were no free pages available yet the value of sgx_nr_free_pages was above the watermark. The allocation of SGX memory thus always failed because of a lack of free pages while no free pages were made available because the reclaimer is never started because of sgx_nr_free_pages' incorrect value. The consequence was that user space kept encountering VM_FAULT_NOPAGE that caused the same address to be accessed repeatedly with the same result.
Change the global free page counter to an atomic type that ensures simultaneous updates are done safely. While doing so, move the updating of the variable outside of the spin lock critical section to which it does not belong.
--- arch/x86/kernel/cpu/sgx/main.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-)
arch/x86/kernel/cpu/sgx/main.c +6 -6
@@ -28,8 +28,7 @@ static DECLARE_WAIT_QUEUE_HEAD(ksgxd_waistatic LIST_HEAD (sgx_active_page_list);static DEFINE_SPINLOCK (sgx_reclaimer_lock); -/* The free page list lock protected variables prepend the lock. */ -static unsigned long sgx_nr_free_pages; +static atomic_long_t sgx_nr_free_pages =ATOMIC_LONG_INIT (0 );/* Nodes with one or more EPC sections. */ static nodemask_t sgx_numa_mask;
@@ -403,14 +402,15 @@ skip:spin_lock (&node->lock);list_add_tail (&epc_page->list, &node->free_page_list); - sgx_nr_free_pages++;spin_unlock (&node->lock); +atomic_long_inc (&sgx_nr_free_pages); } }static bool sgx_should_reclaim (unsigned long watermark) { -return sgx_nr_free_pages < watermark && !list_empty (&sgx_active_page_list); +return atomic_long_read (&sgx_nr_free_pages) < watermark && + !list_empty (&sgx_active_page_list); }static int ksgxd (void *p)
@@ -471,9 +471,9 @@ static struct sgx_epc_page *__sgx_alloc_ page = list_first_entry(&node->free_page_list, struct sgx_epc_page, list);list_del_init (&page->list); - sgx_nr_free_pages--;spin_unlock (&node->lock); +atomic_long_dec (&sgx_nr_free_pages);return page; }
@@ -625,9 +625,9 @@ void sgx_free_epc_page(struct sgx_epc_paspin_lock (&node->lock);list_add_tail (&page->list, &node->free_page_list); - sgx_nr_free_pages++;spin_unlock (&node->lock); +atomic_long_inc (&sgx_nr_free_pages); } static bool __init sgx_setup_epc_section(u64 phys_addr, u64 size,
Greg Kroah-Hartman
maintainerPATCH 203/279 x86/hyperv: Fix NULL deref in set_hv_tscchange_cb() if Hyper-V setup fails
In reply to Greg Kroah-HartmanFrom: Sean Christopherson <[email protected]>
commit daf972118c517b91f74ff1731417feb4270625a4 upstream.
Check for a valid hv_vp_index array prior to derefencing hv_vp_index when setting Hyper-V's TSC change callback. If Hyper-V setup failed in hyperv_init(), the kernel will still report that it's running under Hyper-V, but will have silently disabled nearly all functionality.
BUG: kernel NULL pointer dereference, address: 0000000000000010 #PF: supervisor read access in kernel mode #PF: error_code(0x0000) - not-present page PGD 0 P4D 0 Oops: 0000 [#1] SMP CPU: 4 PID: 1 Comm: swapper/0 Not tainted 5.15.0-rc2+ #75 Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 0.0.0 02/06/2015 RIP: 0010:set_hv_tscchange_cb+0x15/0xa0 Code: <8b> 04 82 8b 15 12 17 85 01 48 c1 e0 20 48 0d ee 00 01 00 f6 c6 08 ... Call Trace: kvm_arch_init+0x17c/0x280 kvm_init+0x31/0x330 vmx_init+0xba/0x13a do_one_initcall+0x41/0x1c0 kernel_init_freeable+0x1f2/0x23b kernel_init+0x16/0x120 ret_from_fork+0x22/0x30
--- arch/x86/hyperv/hv_init.c | 3 +++ 1 file changed, 3 insertions(+)
arch/x86/hyperv/hv_init.c +3 -0
@@ -147,6 +147,9 @@ void set_hv_tscchange_cb(void (*cb)(voidreturn ; } +if (!hv_vp_index) +return ; + hv_reenlightenment_cb = cb;/* Make sure callback is registered before we write to MSRs */
Greg Kroah-Hartman
maintainerPATCH 204/279 KVM: x86: Assume a 64-bit hypercall for guests with protected state
In reply to Greg Kroah-HartmanFrom: Tom Lendacky <[email protected]>
commit b5aead0064f33ae5e693a364e3204fe1c0ac9af2 upstream.
When processing a hypercall for a guest with protected state, currently SEV-ES guests, the guest CS segment register can't be checked to determine if the guest is in 64-bit mode. For an SEV-ES guest, it is expected that communication between the guest and the hypervisor is performed to shared memory using the GHCB. In order to use the GHCB, the guest must have been in long mode, otherwise writes by the guest to the GHCB would be encrypted and not be able to be comprehended by the hypervisor.
Create a new helper function, is_64_bit_hypercall(), that assumes the guest is in 64-bit mode when the guest has protected state, and returns true, otherwise invoking is_64_bit_mode() to determine the mode. Update the hypercall related routines to use is_64_bit_hypercall() instead of is_64_bit_mode().
Add a WARN_ON_ONCE() to is_64_bit_mode() to catch occurences of calls to this helper function for a guest running with protected state.
Message-Id: <e0b20c770c9d0d1403f23d83e785385104211f74.1621878537.git.thomas.lendacky@amd.com>
--- arch/x86/kvm/hyperv.c | 4 ++-- arch/x86/kvm/x86.c | 2 +- arch/x86/kvm/x86.h | 12 ++++++++++++ arch/x86/kvm/xen.c | 2 +- 4 files changed, 16 insertions(+), 4 deletions(-)
arch/x86/kvm/hyperv.c +2 -2
@@ -2022,7 +2022,7 @@ static void kvm_hv_hypercall_set_result( {bool longmode; - longmode =is_64_bit_mode (vcpu); + longmode =is_64_bit_hypercall (vcpu);if (longmode)kvm_rax_write (vcpu, result);else {
@@ -2171,7 +2171,7 @@ int kvm_hv_hypercall(struct kvm_vcpu *vc }#ifdef CONFIG_X86_64 -if (is_64_bit_mode (vcpu)) { +if (is_64_bit_hypercall (vcpu)) { hc.param =kvm_rcx_read (vcpu); hc.ingpa =kvm_rdx_read (vcpu); hc.outgpa =kvm_r8_read (vcpu);
arch/x86/kvm/x86.c +1 -1
@@ -8737,7 +8737,7 @@ int kvm_emulate_hypercall(struct kvm_vcptrace_kvm_hypercall (nr, a0, a1, a2, a3); - op_64_bit =is_64_bit_mode (vcpu); + op_64_bit =is_64_bit_hypercall (vcpu);if (!op_64_bit) { nr &=0xFFFFFFFF ; a0 &=0xFFFFFFFF ;
arch/x86/kvm/x86.h +12 -0
@@ -153,12 +153,24 @@ static inline bool is_64_bit_mode(struct {int cs_db, cs_l; +WARN_ON_ONCE (vcpu->arch.guest_state_protected); +if (!is_long_mode (vcpu))return false;static_call (kvm_x86_get_cs_db_l_bits)(vcpu, &cs_db, &cs_l);return cs_l; } +static inline bool is_64_bit_hypercall (struct kvm_vcpu *vcpu) +{ +/* +* If running with protected guest state, the CS register is not +* accessible. The hypercall register values will have had to been +* provided in 64-bit mode, so assume the guest is in 64-bit. +*/ +return vcpu->arch.guest_state_protected ||is_64_bit_mode (vcpu); +} +static inline bool x86_exception_has_error_code (unsigned int vector) { static u32 exception_has_error_code = BIT(DF_VECTOR) | BIT(TS_VECTOR) |
arch/x86/kvm/xen.c +1 -1
@@ -698,7 +698,7 @@ int kvm_xen_hypercall(struct kvm_vcpu *v kvm_hv_hypercall_enabled(vcpu))return kvm_hv_hypercall (vcpu); - longmode =is_64_bit_mode (vcpu); + longmode =is_64_bit_hypercall (vcpu);if (!longmode) { params[0 ] = (u32 )kvm_rbx_read (vcpu); params[1 ] = (u32 )kvm_rcx_read (vcpu);
Greg Kroah-Hartman
maintainerPATCH 205/279 KVM: x86: Fix uninitialized eoi_exit_bitmap usage in vcpu_load_eoi_exitmap()
In reply to Greg Kroah-HartmanFrom: 黄乐 <[email protected]>
commit c5adbb3af051079f35abfa26551107e2c653087f upstream.
In vcpu_load_eoi_exitmap(), currently the eoi_exit_bitmap[4] array is initialized only when Hyper-V context is available, in other path it is just passed to kvm_x86_ops.load_eoi_exitmap() directly from on the stack, which would cause unexpected interrupt delivery/handling issues, e.g. an *old* linux kernel that relies on PIT to do clock calibration on KVM might randomly fail to boot.
Fix it by passing ioapic_handled_vectors to load_eoi_exitmap() when Hyper-V context is not available.
Message-Id: <[email protected]>
--- arch/x86/kvm/x86.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-)
arch/x86/kvm/x86.c +6 -2
@@ -9429,12 +9429,16 @@ static void vcpu_load_eoi_exitmap(structif (!kvm_apic_hw_enabled (vcpu->arch.apic))return ; -if (to_hv_vcpu (vcpu)) +if (to_hv_vcpu (vcpu)) {bitmap_or ((ulong *)eoi_exit_bitmap, vcpu->arch.ioapic_handled_vectors,to_hv_synic (vcpu)->vec_bitmap,256 ); +static_call (kvm_x86_load_eoi_exitmap)(vcpu, eoi_exit_bitmap); +return ; + } -static_call (kvm_x86_load_eoi_exitmap)(vcpu, eoi_exit_bitmap); +static_call (kvm_x86_load_eoi_exitmap)( + vcpu, (u64 *)vcpu->arch.ioapic_handled_vectors); } void kvm_arch_mmu_notifier_invalidate_range(struct kvm *kvm,
Greg Kroah-Hartman
maintainerPATCH 206/279 KVM: x86/mmu: include EFER.LMA in extended mmu role
In reply to Greg Kroah-HartmanFrom: Maxim Levitsky <[email protected]>
commit b8453cdcf26020030da182f0156d7bf59ae5719f upstream.
Incorporate EFER.LMA into kvm_mmu_extended_role, as it used to compute the guest root level and is not reflected in kvm_mmu_page_role.level when TDP is in use. When simply running the guest, it is impossible for EFER.LMA and kvm_mmu.root_level to get out of sync, as the guest cannot transition from PAE paging to 64-bit paging without toggling CR0.PG, i.e. without first bouncing through a different MMU context. And stuffing guest state via KVM_SET_SREGS{,2} also ensures a full MMU context reset.
However, if KVM_SET_SREGS{,2} is followed by KVM_SET_NESTED_STATE, e.g. to set guest state when migrating the VM while L2 is active, the vCPU state will reflect L2, not L1. If L1 is using TDP for L2, then root_mmu will have been configured using L2's state, despite not being used for L2. If L2.EFER.LMA != L1.EFER.LMA, and L2 is using PAE paging, then root_mmu will be configured for guest PAE paging, but will match the mmu_role for 64-bit paging and cause KVM to not reconfigure root_mmu on the next nested VM-Exit.
Alternatively, the root_mmu's role could be invalidated after a successful KVM_SET_NESTED_STATE that yields vcpu->arch.mmu != vcpu->arch.root_mmu, i.e. that switches the active mmu to guest_mmu, but doing so is unnecessarily tricky, and not even needed if L1 and L2 do have the same role (e.g., they are both 64-bit guests and run with the same CR4).
Message-Id: <[email protected]>
--- arch/x86/include/asm/kvm_host.h | 1 + arch/x86/kvm/mmu/mmu.c | 1 + 2 files changed, 2 insertions(+)
arch/x86/include/asm/kvm_host.h +1 -0
@@ -364,6 +364,7 @@ union kvm_mmu_extended_role { unsigned int cr4_smap:1; unsigned int cr4_smep:1; unsigned int cr4_la57:1; + unsigned int efer_lma:1; }; };
arch/x86/kvm/mmu/mmu.c +1 -0
@@ -4679,6 +4679,7 @@ static union kvm_mmu_extended_role kvm_c/* PKEY and LA57 are active iff long mode is active. */ ext.cr4_pke =____is_efer_lma (regs) &&____is_cr4_pke (regs); ext.cr4_la57 =____is_efer_lma (regs) &&____is_cr4_la57 (regs); + ext.efer_lma =____is_efer_lma (regs); } ext.valid =1 ;
Greg Kroah-Hartman
maintainerPATCH 207/279 KVM: x86/xen: Fix get_attr of KVM_XEN_ATTR_TYPE_SHARED_INFO
In reply to Greg Kroah-HartmanFrom: David Woodhouse <[email protected]>
commit 4e8436479ad3be76a3823e6ce466ae464ce71300 upstream.
In commit 319afe68567b ("KVM: xen: do not use struct gfn_to_hva_cache") we stopped storing this in-kernel as a GPA, and started storing it as a GFN. Which means we probably should have stopped calling gpa_to_gfn() on it when userspace asks for it back.
Message-Id: <[email protected]>
--- arch/x86/kvm/xen.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/kvm/xen.c +1 -1
@@ -299,7 +299,7 @@ int kvm_xen_hvm_get_attr(struct kvm *kvmbreak ;case KVM_XEN_ATTR_TYPE_SHARED_INFO : - data->u.shared_info.gfn =gpa_to_gfn (kvm->arch.xen.shinfo_gfn); + data->u.shared_info.gfn = kvm->arch.xen.shinfo_gfn; r =0 ;break ;
Greg Kroah-Hartman
maintainerPATCH 208/279 powerpc/signal32: Fix sigset_t copy
In reply to Greg Kroah-HartmanFrom: Christophe Leroy <[email protected]>
commit 5499802b2284331788a440585869590f1bd63f7f upstream.
The conversion from __copy_from_user() to __get_user() by commit d3ccc9781560 ("powerpc/signal: Use __get_user() to copy sigset_t") introduced a regression in __get_user_sigset() for powerpc/32. The bug was subsequently moved into unsafe_get_user_sigset().
The bug is due to the copied 64 bit value being truncated to 32 bits while being assigned to dst->sig[0]
The regression was reported by users of the Xorg packages distributed in Debian/powerpc --
"The symptoms are that the fb screen goes blank, with the backlight remaining on and no errors logged in /var/log; wdm (or startx) run with no effect (I tried logging in in the blind, with no effect). And they are hard to kill, requiring 'kill -KILL ...'"
Fix the regression by copying each word of the sigset, not only the first one.
__get_user_sigset() was tentatively optimised to copy 64 bits at once in order to minimise KUAP unlock/lock impact, but the unsafe variant doesn't suffer that, so it can just copy words.
--- arch/powerpc/kernel/signal.h | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-)
arch/powerpc/kernel/signal.h +8 -2
@@ -25,8 +25,14 @@ static inline int __get_user_sigset(sigsreturn __get_user (dst->sig[0 ], (u64 __user *)&src->sig[0 ]); } -#define unsafe_get_user_sigset (dst, src, label) \ - unsafe_get_user((dst)->sig[0], (u64 __user *)&(src)->sig[0], label) +#define unsafe_get_user_sigset (dst, src, label) do { \ + sigset_t *__dst = dst; \ + const sigset_t __user *__src = src; \ + int i; \ + \ + for (i = 0; i < _NSIG_WORDS; i++) \ + unsafe_get_user(__dst->sig[i], &__src->sig[i], label); \ +} while (0)#ifdef CONFIG_VSX extern unsigned long copy_vsx_to_user(void __user *to,
Greg Kroah-Hartman
maintainerPATCH 209/279 powerpc/xive: Change IRQ domain to a tree domain
In reply to Greg Kroah-HartmanFrom: Cédric Le Goater <[email protected]>
commit 8e80a73fa9a7747e3e8255cb149c543aabf65a24 upstream.
Commit 4f86a06e2d6e ("irqdomain: Make normal and nomap irqdomains exclusive") introduced an IRQ_DOMAIN_FLAG_NO_MAP flag to isolate the 'nomap' domains still in use under the powerpc arch. With this new flag, the revmap_tree of the IRQ domain is not used anymore. This change broke the support of shared LSIs [1] in the XIVE driver because it was relying on a lookup in the revmap_tree to query previously mapped interrupts. Linux now creates two distinct IRQ mappings on the same HW IRQ which can lead to unexpected behavior in the drivers.
The XIVE IRQ domain is not a direct mapping domain and its HW IRQ interrupt number space is rather large : 1M/socket on POWER9 and POWER10, change the XIVE driver to use a 'tree' domain type instead.
[1] For instance, a linux KVM guest with virtio-rng and virtio-balloon devices.
--- arch/powerpc/sysdev/xive/Kconfig | 1 - arch/powerpc/sysdev/xive/common.c | 3 +-- 2 files changed, 1 insertion(+), 3 deletions(-)
arch/powerpc/sysdev/xive/Kconfig +0 -1
@@ -3,7 +3,6 @@ config PPC_XIVE bool select PPC_SMP_MUXED_IPI select HARDIRQS_SW_RESEND - select IRQ_DOMAIN_NOMAP config PPC_XIVE_NATIVE bool
arch/powerpc/sysdev/xive/common.c +1 -2
@@ -1443,8 +1443,7 @@ static const struct irq_domain_ops xive_static void __init xive_init_host (struct device_node *np) { - xive_irq_domain =irq_domain_add_nomap (np,XIVE_MAX_IRQ , - &xive_irq_domain_ops,NULL ); + xive_irq_domain =irq_domain_add_tree (np, &xive_irq_domain_ops,NULL );if (WARN_ON (xive_irq_domain ==NULL ))return ;irq_set_default_host (xive_irq_domain);
Greg Kroah-Hartman
maintainerPATCH 210/279 powerpc/8xx: Fix pinned TLBs with CONFIG_STRICT_KERNEL_RWX
In reply to Greg Kroah-HartmanFrom: Christophe Leroy <[email protected]>
commit 1e35eba4055149c578baf0318d2f2f89ea3c44a0 upstream.
As spotted and explained in commit c12ab8dbc492 ("powerpc/8xx: Fix Oops with STRICT_KERNEL_RWX without DEBUG_RODATA_TEST"), the selection of STRICT_KERNEL_RWX without selecting DEBUG_RODATA_TEST has spotted the lack of the DIRTY bit in the pinned kernel data TLBs.
This problem should have been detected a lot earlier if things had been working as expected. But due to an incredible level of chance or mishap, this went undetected because of a set of bugs: In fact the DTLBs were not pinned, because instead of setting the reserve bit in MD_CTR, it was set in MI_CTR that is the register for ITLBs.
But then, another huge bug was there: the physical address was reset to 0 at the boundary between RO and RW areas, leading to the same physical space being mapped at both 0xc0000000 and 0xc8000000. This had by miracle no consequence until now because the entry was not really pinned so it was overwritten soon enough to go undetected.
Of course, now that we really pin the DTLBs, it must be fixed as well.
Depends-on: c12ab8dbc492 ("powerpc/8xx: Fix Oops with STRICT_KERNEL_RWX without DEBUG_RODATA_TEST")
--- arch/powerpc/kernel/head_8xx.S | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-)
arch/powerpc/kernel/head_8xx.S +7 -6
@@ -733,6 +733,7 @@ _GLOBAL(mmu_pin_tlb) #ifdef CONFIG_PIN_TLB_DATA LOAD_REG_IMMEDIATE(r6, PAGE_OFFSET) LOAD_REG_IMMEDIATE(r7, MI_SVALID | MI_PS8MEG | _PMD_ACCESSED) + li r8, 0 #ifdef CONFIG_PIN_TLB_IMMR li r0, 3 #else
@@ -741,26 +742,26 @@ _GLOBAL(mmu_pin_tlb) mtctr r0 cmpwi r4, 0 beq 4f - LOAD_REG_IMMEDIATE(r8, 0xf0 | _PAGE_RO | _PAGE_SPS | _PAGE_SH | _PAGE_PRESENT) LOAD_REG_ADDR(r9, _sinittext) 2: ori r0, r6, MD_EVALID + ori r12, r8, 0xf0 | _PAGE_RO | _PAGE_SPS | _PAGE_SH | _PAGE_PRESENT mtspr SPRN_MD_CTR, r5 mtspr SPRN_MD_EPN, r0 mtspr SPRN_MD_TWC, r7 - mtspr SPRN_MD_RPN, r8 + mtspr SPRN_MD_RPN, r12 addi r5, r5, 0x100 addis r6, r6, SZ_8M@h addis r8, r8, SZ_8M@h cmplw r6, r9 bdnzt lt, 2b - -4: LOAD_REG_IMMEDIATE(r8, 0xf0 | _PAGE_DIRTY | _PAGE_SPS | _PAGE_SH | _PAGE_PRESENT) +4: 2: ori r0, r6, MD_EVALID + ori r12, r8, 0xf0 | _PAGE_DIRTY | _PAGE_SPS | _PAGE_SH | _PAGE_PRESENT mtspr SPRN_MD_CTR, r5 mtspr SPRN_MD_EPN, r0 mtspr SPRN_MD_TWC, r7 - mtspr SPRN_MD_RPN, r8 + mtspr SPRN_MD_RPN, r12 addi r5, r5, 0x100 addis r6, r6, SZ_8M@h addis r8, r8, SZ_8M@h
@@ -781,7 +782,7 @@ _GLOBAL(mmu_pin_tlb) #endif #if defined(CONFIG_PIN_TLB_IMMR) || defined(CONFIG_PIN_TLB_DATA) lis r0, (MD_RSV4I | MD_TWAM)@h - mtspr SPRN_MI_CTR, r0 + mtspr SPRN_MD_CTR, r0 #endif mtspr SPRN_SRR1, r10 mtspr SPRN_SRR0, r11
Greg Kroah-Hartman
maintainerPATCH 211/279 Revert "drm/i915/tgl/dsi: Gate the ddi clocks after pll mapping"
In reply to Greg Kroah-HartmanFrom: Vandita Kulkarni <[email protected]>
commit f15863b27752682bb700c21de5f83f613a0fb77e upstream.
This reverts commit 991d9557b0c4 ("drm/i915/tgl/dsi: Gate the ddi clocks after pll mapping"). The Bspec was updated recently with the pll ungate sequence similar to that of icl dsi enable sequence. Hence reverting.
Bspec: 49187
--- drivers/gpu/drm/i915/display/icl_dsi.c | 10 ++-------- 1 file changed, 2 insertions(+), 8 deletions(-)
drivers/gpu/drm/i915/display/icl_dsi.c +2 -8
@@ -711,10 +711,7 @@ static void gen11_dsi_map_pll(struct intintel_de_write (dev_priv,ICL_DPCLKA_CFGCR0 , val);for_each_dsi_phy (phy, intel_dsi->phys) { -if (DISPLAY_VER (dev_priv) >=12 ) - val |=ICL_DPCLKA_CFGCR0_DDI_CLK_OFF (phy); -else - val &= ~ICL_DPCLKA_CFGCR0_DDI_CLK_OFF (phy); + val &= ~ICL_DPCLKA_CFGCR0_DDI_CLK_OFF (phy); }intel_de_write (dev_priv,ICL_DPCLKA_CFGCR0 , val);
@@ -1150,8 +1147,6 @@ static void gen11_dsi_enable_port_and_phy(struct intel_encoder *encoder, const struct intel_crtc_state *crtc_state) { - struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); -/* step 4a: power up all lanes of the DDI used by DSI */ gen11_dsi_power_up_lanes (encoder);
@@ -1177,8 +1172,7 @@ gen11_dsi_enable_port_and_phy(struct intgen11_dsi_configure_transcoder (encoder, crtc_state);/* Step 4l: Gate DDI clocks */ -if (DISPLAY_VER (dev_priv) ==11 ) -gen11_dsi_gate_clocks (encoder); +gen11_dsi_gate_clocks (encoder); }static void gen11_dsi_powerup_panel (struct intel_encoder *encoder)
Greg Kroah-Hartman
maintainerPATCH 212/279 Revert "parisc: Reduce sigreturn trampoline to 3 instructions"
In reply to Greg Kroah-HartmanFrom: Helge Deller <[email protected]>
commit 79df39d535c7a3770856fe9f5aba8c0ad1eebdb6 upstream.
This reverts commit e4f2006f1287e7ea17660490569cff323772dac4.
This patch shows problems with signal handling. Revert it for now.
--- arch/parisc/include/asm/rt_sigframe.h | 2 +- arch/parisc/kernel/signal.c | 13 +++++++------ arch/parisc/kernel/signal32.h | 2 +- 3 files changed, 9 insertions(+), 8 deletions(-)
arch/parisc/include/asm/rt_sigframe.h +1 -1
@@ -2,7 +2,7 @@#ifndef _ASM_PARISC_RT_SIGFRAME_H#define _ASM_PARISC_RT_SIGFRAME_H -#define SIGRETURN_TRAMP 3 +#define SIGRETURN_TRAMP 4#define SIGRESTARTBLOCK_TRAMP 5#define TRAMP_SIZE (SIGRETURN_TRAMP + SIGRESTARTBLOCK_TRAMP)
arch/parisc/kernel/signal.c +7 -6
@@ -288,21 +288,22 @@ setup_rt_frame(struct ksignal *ksig, sigalready in userspace. The first words of tramp are used to save the previous sigrestartblock trampoline that might be on the stack. We start the sigreturn trampoline at -SIGRESTARTBLOCK_TRAMP. */ +SIGRESTARTBLOCK_TRAMP+X. */ err |=__put_user (in_syscall ?INSN_LDI_R25_1 :INSN_LDI_R25_0 , &frame->tramp[SIGRESTARTBLOCK_TRAMP +0 ]); - err |=__put_user (INSN_BLE_SR2_R0 , + err |=__put_user (INSN_LDI_R20 , &frame->tramp[SIGRESTARTBLOCK_TRAMP +1 ]); - err |=__put_user (INSN_LDI_R20 , + err |=__put_user (INSN_BLE_SR2_R0 , &frame->tramp[SIGRESTARTBLOCK_TRAMP +2 ]); + err |=__put_user (INSN_NOP , &frame->tramp[SIGRESTARTBLOCK_TRAMP +3 ]); - start = (unsigned long ) &frame->tramp[SIGRESTARTBLOCK_TRAMP +0 ]; - end = (unsigned long ) &frame->tramp[SIGRESTARTBLOCK_TRAMP +3 ]; + start = (unsigned long ) &frame->tramp[0 ]; + end = (unsigned long ) &frame->tramp[TRAMP_SIZE ];flush_user_dcache_range_asm (start, end);flush_user_icache_range_asm (start, end);/* TRAMP Words 0-4, Length 5 = SIGRESTARTBLOCK_TRAMP -* TRAMP Words 5-7, Length 3 = SIGRETURN_TRAMP +* TRAMP Words 5-9, Length 4 = SIGRETURN_TRAMP * So the SIGRETURN_TRAMP is at the end of SIGRESTARTBLOCK_TRAMP */ rp = (unsigned long ) &frame->tramp[SIGRESTARTBLOCK_TRAMP ];
arch/parisc/kernel/signal32.h +1 -1
@@ -36,7 +36,7 @@ struct compat_regfile {compat_int_t rf_sar; }; -#define COMPAT_SIGRETURN_TRAMP 3 +#define COMPAT_SIGRETURN_TRAMP 4#define COMPAT_SIGRESTARTBLOCK_TRAMP 5#define COMPAT_TRAMP_SIZE (COMPAT_SIGRETURN_TRAMP + \ COMPAT_SIGRESTARTBLOCK_TRAMP)
Greg Kroah-Hartman
maintainerPATCH 213/279 ata: libata: improve ata_read_log_page() error message
In reply to Greg Kroah-HartmanFrom: Damien Le Moal <[email protected]>
commit 23ef63d5e14f916c5bba39128ebef395859d7c0f upstream.
If ata_read_log_page() fails to read a log page, the ata_dev_err() error message only print the page number, omitting the log number. In case of error, facilitate debugging by also printing the log number.
--- drivers/ata/libata-core.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)
drivers/ata/libata-core.c +3 -2
@@ -2031,8 +2031,9 @@ retry: dev->horkage |=ATA_HORKAGE_NO_DMA_LOG ; goto retry; } -ata_dev_err (dev,"Read log page 0x%02x failed, Emask 0x%x\n" , - (unsigned int )page, err_mask); +ata_dev_err (dev, +"Read log 0x%02x page 0x%02x failed, Emask 0x%x\n" , + (unsigned int )log, (unsigned int )page, err_mask); }return err_mask;
Greg Kroah-Hartman
maintainerPATCH 214/279 ata: libata: add missing ata_identify_page_supported() calls
In reply to Greg Kroah-HartmanFrom: Damien Le Moal <[email protected]>
commit 06f6c4c6c3e8354dceddd77bd58f9a7a84c67246 upstream.
ata_dev_config_ncq_prio() and ata_dev_config_devslp() both access pages of the IDENTIFY DEVICE data log. Before calling ata_read_log_page(), make sure to check for the existence of the IDENTIFY DEVICE data log and of the log page accessed using ata_identify_page_supported(). This avoids useless error messages from ata_read_log_page() and failures with some LLDD scsi drivers using libsas.
--- drivers/ata/libata-core.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/ata/libata-core.c +5 -1
@@ -2167,6 +2167,9 @@ static void ata_dev_config_ncq_prio(strustruct ata_port *ap = dev->link->ap;unsigned int err_mask; +if (!ata_identify_page_supported (dev,ATA_LOG_SATA_SETTINGS )) +return ; + err_mask = ata_read_log_page(dev, ATA_LOG_IDENTIFY_DEVICE, ATA_LOG_SATA_SETTINGS,
@@ -2443,7 +2446,8 @@ static void ata_dev_config_devslp(struct* Check device sleep capability. Get DevSlp timing variables * from SATA Settings page of Identify Device Data Log. */ -if (!ata_id_has_devslp (dev->id)) +if (!ata_id_has_devslp (dev->id) || + !ata_identify_page_supported (dev,ATA_LOG_SATA_SETTINGS ))return ; err_mask = ata_read_log_page(dev,
Greg Kroah-Hartman
maintainerPATCH 215/279 scsi: qla2xxx: Fix mailbox direction flags in qla2xxx_get_adapter_id()
In reply to Greg Kroah-HartmanFrom: Ewan D. Milne <[email protected]>
commit 392006871bb26166bcfafa56faf49431c2cfaaa8 upstream.
The SCM changes set the flags in mcp->out_mb instead of mcp->in_mb so the data was not actually being read into the mcp->mb[] array from the adapter.
--- drivers/scsi/qla2xxx/qla_mbx.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-)
drivers/scsi/qla2xxx/qla_mbx.c +2 -4
@@ -1695,10 +1695,8 @@ qla2x00_get_adapter_id(scsi_qla_host_t * mcp->in_mb |=MBX_13 |MBX_12 |MBX_11 |MBX_10 ;if (IS_FWI2_CAPABLE (vha->hw)) mcp->in_mb |=MBX_19 |MBX_18 |MBX_17 |MBX_16 ; -if (IS_QLA27XX (vha->hw) ||IS_QLA28XX (vha->hw)) { - mcp->in_mb |=MBX_15 ; - mcp->out_mb |=MBX_7 |MBX_21 |MBX_22 |MBX_23 ; - } +if (IS_QLA27XX (vha->hw) ||IS_QLA28XX (vha->hw)) + mcp->in_mb |=MBX_15 |MBX_21 |MBX_22 |MBX_23 ; mcp->tov =MBX_TOV_SECONDS ; mcp->flags =0 ;
Greg Kroah-Hartman
maintainerPATCH 216/279 pinctrl: ralink: include ralink_regs.h in pinctrl-mt7620.c
In reply to Greg Kroah-HartmanFrom: Sergio Paracuellos <[email protected]>
commit a5b9703fe11cd1d6d7a60102aa2abe686dc1867f upstream.
mt7620.h, included by pinctrl-mt7620.c, mentions MT762X_SOC_MT7628AN declared in ralink_regs.h.
--- drivers/pinctrl/ralink/pinctrl-mt7620.c | 1 + 1 file changed, 1 insertion(+)
drivers/pinctrl/ralink/pinctrl-mt7620.c +1 -0
@@ -1,5 +1,6 @@// SPDX-License-Identifier: GPL-2.0-only +#include <asm/mach-ralink/ralink_regs.h> #include <asm/mach-ralink/mt7620.h> #include <linux/module.h> #include <linux/platform_device.h>
Greg Kroah-Hartman
maintainerPATCH 217/279 s390/setup: avoid reserving memory above identity mapping
In reply to Greg Kroah-HartmanFrom: Vasily Gorbik <[email protected]>
commit 420f48f636b98fd685f44a3acc4c0a7c0840910d upstream.
Such reserved memory region, if not cleaned up later causes problems when memblock_free_all() is called to release free pages to the buddy allocator and those reserved regions are carried over to reserve_bootmem_region() which marks the pages as PageReserved.
Instead use memblock_set_current_limit() to make sure memblock allocations do not go over identity mapping (which could happen when "mem=" option is used or during kdump).
--- arch/s390/kernel/setup.c | 10 +--------- 1 file changed, 1 insertion(+), 9 deletions(-)
arch/s390/kernel/setup.c +1 -9
@@ -634,14 +634,6 @@ static struct notifier_block kdump_mem_n#endif /* -* Make sure that the area above identity mapping is protected -*/ -static void __init reserve_above_ident_map (void ) -{ -memblock_reserve (ident_map_size,ULONG_MAX ); -} - -/* * Reserve memory for kdump kernel to be loaded with kexec */ static void __init reserve_crashkernel (void )
@@ -1005,11 +997,11 @@ void __init setup_arch(char **cmdline_p)setup_control_program_code ();/* Do some memory reservations *before* memory is added to memblock */ -reserve_above_ident_map ();reserve_kernel ();reserve_initrd ();reserve_certificate_list ();reserve_mem_detect_info (); +memblock_set_current_limit (ident_map_size);memblock_allow_resize ();/* Get information about *all* installed memory */
Greg Kroah-Hartman
maintainerPATCH 218/279 s390/boot: simplify and fix kernel memory layout setup
In reply to Greg Kroah-HartmanFrom: Vasily Gorbik <[email protected]>
commit 9a39abb7c9aab50eec4ac4421e9ee7f3de013d24 upstream.
Initial KASAN shadow memory range was picked to preserve original kernel modules area position. With protected execution support, which might impose addressing limitation on vmalloc area and hence affect modules area position, current fixed KASAN shadow memory range is only making kernel memory layout setup more complex. So move it to the very end of available virtual space and simplify calculations.
At the same time return to previous kernel address space split. In particular commit 0c4f2623b957 ("s390: setup kernel memory layout early") introduced precise identity map size calculation and keeping vmemmap left most starting from a fresh region table entry. This didn't take into account additional mapping region requirement for potential DCSS mapping above available physical memory. So go back to virtual space split between 1:1 mapping & vmemmap array once vmalloc area size is subtracted.
--- arch/s390/Kconfig | 2 - arch/s390/boot/startup.c | 88 ++++++++++++++++------------------------------- 2 files changed, 32 insertions(+), 58 deletions(-)
arch/s390/Kconfig +1 -1
@@ -47,7 +47,7 @@ config ARCH_SUPPORTS_UPROBES config KASAN_SHADOW_OFFSET hex depends on KASAN - default 0x18000000000000 + default 0x1C000000000000 config S390 def_bool y
arch/s390/boot/startup.c +31 -57
113 lines not on this page · raw message
Greg Kroah-Hartman
maintainerPATCH 219/279 s390/vdso: filter out -mstack-guard and -mstack-size
In reply to Greg Kroah-HartmanFrom: Sven Schnelle <[email protected]>
commit 00b55eaf45549ce26424224d069a091c7e5d8bac upstream.
When CONFIG_VMAP_STACK is disabled, the user can enable CONFIG_STACK_CHECK, which adds a stack overflow check to each C function in the kernel. This is also done for functions in the vdso page. These functions are run in user context and user stack sizes are usually different to what the kernel uses. This might trigger the stack check although the stack size is valid. Therefore filter the -mstack-guard and -mstack-size flags when compiling vdso C files.
--- arch/s390/Makefile | 10 ++++++---- arch/s390/kernel/vdso64/Makefile | 5 +++-- 2 files changed, 9 insertions(+), 6 deletions(-)
arch/s390/Makefile +6 -4
@@ -79,10 +79,12 @@ KBUILD_AFLAGS_DECOMPRESSOR += $(aflags-y KBUILD_CFLAGS_DECOMPRESSOR += $(cflags-y) ifneq ($(call cc-option,-mstack-size=8192 -mstack-guard=128),) -cflags-$(CONFIG_CHECK_STACK) += -mstack-size=$(STACK_SIZE) -ifeq ($(call cc-option,-mstack-size=8192),) -cflags-$(CONFIG_CHECK_STACK) += -mstack-guard=$(CONFIG_STACK_GUARD) -endif + CC_FLAGS_CHECK_STACK := -mstack-size=$(STACK_SIZE) + ifeq ($(call cc-option,-mstack-size=8192),) + CC_FLAGS_CHECK_STACK += -mstack-guard=$(CONFIG_STACK_GUARD) + endif + export CC_FLAGS_CHECK_STACK + cflags-$(CONFIG_CHECK_STACK) += $(CC_FLAGS_CHECK_STACK) endif ifdef CONFIG_EXPOLINE
arch/s390/kernel/vdso64/Makefile +3 -2
@@ -8,8 +8,9 @@ ARCH_REL_TYPE_ABS += R_390_GOT|R_390_PLT include $(srctree)/lib/vdso/Makefile obj-vdso64 = vdso_user_wrapper.o note.o obj-cvdso64 = vdso64_generic.o getcpu.o -CFLAGS_REMOVE_getcpu.o = -pg $(CC_FLAGS_FTRACE) $(CC_FLAGS_EXPOLINE) -CFLAGS_REMOVE_vdso64_generic.o = -pg $(CC_FLAGS_FTRACE) $(CC_FLAGS_EXPOLINE) +VDSO_CFLAGS_REMOVE := -pg $(CC_FLAGS_FTRACE) $(CC_FLAGS_EXPOLINE) $(CC_FLAGS_CHECK_STACK) +CFLAGS_REMOVE_getcpu.o = $(VDSO_CFLAGS_REMOVE) +CFLAGS_REMOVE_vdso64_generic.o = $(VDSO_CFLAGS_REMOVE) # Build rules
Greg Kroah-Hartman
maintainerPATCH 220/279 s390/kexec: fix memory leak of ipl report buffer
In reply to Greg Kroah-HartmanFrom: Baoquan He <[email protected]>
commit 4aa9340584e37debef06fa99b56d064beb723891 upstream.
unreferenced object 0x38000195000 (size 4096):
comm "kexec", pid 8548, jiffies 4294953647 (age 32443.270s) hex dump (first 32 bytes):
00 00 00 c8 20 00 00 00 00 00 00 c0 02 80 00 00 .... ...........
40 40 40 40 40 40 40 40 00 00 00 00 00 00 00 00 @@@@@@@@........
backtrace:
[<0000000011a2f199>] __vmalloc_node_range+0xc0/0x140
[<0000000081fa2752>] vzalloc+0x5a/0x70
[<0000000063a4c92d>] ipl_report_finish+0x2c/0x180
[<00000000553304da>] kexec_file_add_ipl_report+0xf4/0x150
[<00000000862d033f>] kexec_file_add_components+0x124/0x160
[<000000000d2717bb>] arch_kexec_kernel_image_load+0x62/0x90
[<000000002e0373b6>] kimage_file_alloc_init+0x1aa/0x2e0
[<0000000060f2d14f>] __do_sys_kexec_file_load+0x17c/0x2c0
[<000000008c86fe5a>] __s390x_sys_kexec_file_load+0x40/0x50
[<000000001fdb9dac>] __do_syscall+0x1bc/0x1f0
[<000000003ee4258d>] system_call+0x78/0xa0--- arch/s390/include/asm/kexec.h | 6 ++++++ arch/s390/kernel/machine_kexec_file.c | 10 ++++++++++ 2 files changed, 16 insertions(+)
arch/s390/include/asm/kexec.h +6 -0
@@ -74,6 +74,12 @@ void *kexec_file_add_components(struct kint arch_kexec_do_relocs (int r_type,void *loc,unsigned long val,unsigned long addr); +#define ARCH_HAS_KIMAGE_ARCH + +struct kimage_arch { +void *ipl_buf; +}; +extern const struct kexec_file_ops s390_kexec_image_ops;extern const struct kexec_file_ops s390_kexec_elf_ops;
arch/s390/kernel/machine_kexec_file.c +10 -0
@@ -12,6 +12,7 @@#include <linux/kexec.h> #include <linux/module_signature.h> #include <linux/verification.h> +#include <linux/vmalloc.h> #include <asm/boot_data.h> #include <asm/ipl.h> #include <asm/setup.h>
@@ -206,6 +207,7 @@ static int kexec_file_add_ipl_report(str goto out; buf.bufsz = data->report->size; buf.memsz = buf.bufsz; + image->arch.ipl_buf = buf.buffer; data->memsz += buf.memsz;
@@ -327,3 +329,11 @@ int arch_kexec_kernel_image_probe(structreturn kexec_image_probe_default (image, buf, buf_len); } + +int arch_kimage_file_post_load_cleanup (struct kimage *image) +{ +vfree (image->arch.ipl_buf); + image->arch.ipl_buf =NULL ; + +return kexec_image_post_load_cleanup_default (image); +}
Greg Kroah-Hartman
maintainerPATCH 221/279 s390/dump: fix copying to user-space of swapped kdump oldmem
In reply to Greg Kroah-HartmanFrom: Alexander Egorenkov <[email protected]>
commit 3b90954419d4c05651de9cce6d7632bcf6977678 upstream.
This commit fixes a bug introduced by commit e9e7870f90e3 ("s390/dump:
introduce boot data 'oldmem_data'").OLDMEM_BASE was mistakenly replaced by oldmem_data.size instead of oldmem_data.start.
This bug caused the following error during kdump:
kdump.sh[878]: No program header covering vaddr 0x3434f5245found kexec bug?
--- arch/s390/kernel/crash_dump.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
arch/s390/kernel/crash_dump.c +2 -2
@@ -191,8 +191,8 @@ static int copy_oldmem_user(void __userreturn rc; }else {/* Check for swapped kdump oldmem areas */ -if (oldmem_data.start && from - oldmem_data.size < oldmem_data.size) { - from -= oldmem_data.size; +if (oldmem_data.start && from - oldmem_data.start < oldmem_data.size) { + from -= oldmem_data.start; len =min (count, oldmem_data.size - from); }else if (oldmem_data.start && from < oldmem_data.size) { len =min (count, oldmem_data.size - from);
Greg Kroah-Hartman
maintainerPATCH 222/279 block: Check ADMIN before NICE for IOPRIO_CLASS_RT
In reply to Greg Kroah-HartmanFrom: Alistair Delva <[email protected]>
commit 94c4b4fd25e6c3763941bdec3ad54f2204afa992 upstream.
Booting to Android userspace on 5.14 or newer triggers the following SELinux denial:
avc: denied { sys_nice } for comm="init" capability=23
scontext=u:r:init:s0 tcontext=u:r:init:s0 tclass=capability permissive=0
Init is PID 0 running as root, so it already has CAP_SYS_ADMIN. For better compatibility with older SEPolicy, check ADMIN before NICE.
--- block/ioprio.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-)
block/ioprio.c +8 -1
@@ -69,7 +69,14 @@ int ioprio_check_cap(int ioprio)switch (class) {case IOPRIO_CLASS_RT : -if (!capable (CAP_SYS_NICE ) && !capable (CAP_SYS_ADMIN )) +/* +* Originally this only checked for CAP_SYS_ADMIN, +* which was implicitly allowed for pid 0 by security +* modules such as SELinux. Make sure we check +* CAP_SYS_ADMIN first to avoid a denial/avc for +* possibly missing CAP_SYS_NICE permission. +*/ +if (!capable (CAP_SYS_ADMIN ) && !capable (CAP_SYS_NICE ))return -EPERM ; fallthrough;/* rt has prio field too */
Greg Kroah-Hartman
maintainerPATCH 223/279 fbdev: Prevent probing generic drivers if a FB is already registered
In reply to Greg Kroah-HartmanFrom: Javier Martinez Canillas <[email protected]>
commit fb561bf9abde49f7e00fdbf9ed2ccf2d86cac8ee upstream.
The efifb and simplefb drivers just render to a pre-allocated frame buffer and rely on the display hardware being initialized before the kernel boots.
But if another driver already probed correctly and registered a fbdev, the generic drivers shouldn't be probed since an actual driver for the display hardware is already present.
This is more likely to occur after commit d391c5827107 ("drivers/firmware:
move x86 Generic System Framebuffers support") since the "efi-framebuffer" and "simple-framebuffer" platform devices are registered at a later time.
--- drivers/video/fbdev/efifb.c | 11 +++++++++++ drivers/video/fbdev/simplefb.c | 11 +++++++++++ 2 files changed, 22 insertions(+)
drivers/video/fbdev/efifb.c +11 -0
@@ -351,6 +351,17 @@ static int efifb_probe(struct platform_dchar *option =NULL ;efi_memory_desc_t md; +/* +* Generic drivers must not be registered if a framebuffer exists. +* If a native driver was probed, the display hardware was already +* taken and attempting to use the system framebuffer is dangerous. +*/ +if (num_registered_fb >0 ) { +dev_err (&dev->dev, +"efifb: a framebuffer is already registered\n" ); +return -EINVAL ; + } +if (screen_info.orig_video_isVGA !=VIDEO_TYPE_EFI || pci_dev_disabled)return -ENODEV ;
drivers/video/fbdev/simplefb.c +11 -0
@@ -407,6 +407,17 @@ static int simplefb_probe(struct platforstruct simplefb_par *par;struct resource *mem; +/* +* Generic drivers must not be registered if a framebuffer exists. +* If a native driver was probed, the display hardware was already +* taken and attempting to use the system framebuffer is dangerous. +*/ +if (num_registered_fb >0 ) { +dev_err (&pdev->dev, +"simplefb: a framebuffer is already registered\n" ); +return -EINVAL ; + } +if (fb_get_options ("simplefb" ,NULL ))return -ENODEV ;
Greg Kroah-Hartman
maintainerPATCH 224/279 KVM: SEV: Disallow COPY_ENC_CONTEXT_FROM if target has created vCPUs
In reply to Greg Kroah-HartmanFrom: Sean Christopherson <[email protected]>
commit 79b11142763791bdead8b6460052cbdde8e08e2f upstream.
Reject COPY_ENC_CONTEXT_FROM if the destination VM has created vCPUs. KVM relies on SEV activation to occur before vCPUs are created, e.g. to set VMCB flags and intercepts correctly.
Message-Id: <[email protected]>
--- arch/x86/kvm/svm/sev.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
arch/x86/kvm/svm/sev.c +6 -1
@@ -1787,7 +1787,12 @@ int svm_vm_copy_asid_from(struct kvm *kvmutex_unlock (&source_kvm->lock);mutex_lock (&kvm->lock); -if (sev_guest (kvm)) { +/* +* Disallow out-of-band SEV/SEV-ES init if the target is already an +* SEV guest, or if vCPUs have been created. KVM relies on vCPUs being +* created after SEV/SEV-ES initialization, e.g. to init intercepts. +*/ +if (sev_guest (kvm) || kvm->created_vcpus) { ret = -EINVAL ; goto e_mirror_unlock; }
Greg Kroah-Hartman
maintainerPATCH 225/279 KVM: nVMX: dont use vcpu->arch.efer when checking host state on nested state load
In reply to Greg Kroah-HartmanFrom: Maxim Levitsky <[email protected]>
commit af957eebfcc17433ee83ab85b1195a933ab5049c upstream.
When loading nested state, don't use check vcpu->arch.efer to get the L1 host's 64-bit vs. 32-bit state and don't check it for consistency with respect to VM_EXIT_HOST_ADDR_SPACE_SIZE, as register state in vCPU may be stale when KVM_SET_NESTED_STATE is called---and architecturally does not exist. When restoring L2 state in KVM, the CPU is placed in non-root where nested VMX code has no snapshot of L1 host state: VMX (conditionally) loads host state fields loaded on VM-exit, but they need not correspond to the state before entry. A simple case occurs in KVM itself, where the host RIP field points to vmx_vmexit rather than the instruction following vmlaunch/vmresume.
However, for the particular case of L1 being in 32- or 64-bit mode on entry, the exit controls can be treated instead as the source of truth regarding the state of L1 on entry, and can be used to check that vmcs12.VM_EXIT_HOST_ADDR_SPACE_SIZE matches vmcs12.HOST_EFER if vmcs12.VM_EXIT_LOAD_IA32_EFER is set. The consistency check on CPU EFER vs. vmcs12.VM_EXIT_HOST_ADDR_SPACE_SIZE, instead, happens only on VM-Enter. That's because, again, there's conceptually no "current" L1 EFER to check on KVM_SET_NESTED_STATE.
Message-Id: <[email protected]>
--- arch/x86/kvm/vmx/nested.c | 22 +++++++++++++++++----- 1 file changed, 17 insertions(+), 5 deletions(-)
arch/x86/kvm/vmx/nested.c +17 -5
@@ -2854,6 +2854,17 @@ static int nested_vmx_check_controls(strreturn 0 ; } +static int nested_vmx_check_address_space_size (struct kvm_vcpu *vcpu, +struct vmcs12 *vmcs12) +{ +#ifdef CONFIG_X86_64 +if (CC (!!(vmcs12->vm_exit_controls &VM_EXIT_HOST_ADDR_SPACE_SIZE ) != + !!(vcpu->arch.efer &EFER_LMA ))) +return -EINVAL ; +#endif +return 0 ; +} +static int nested_vmx_check_host_state (struct kvm_vcpu *vcpu,struct vmcs12 *vmcs12) {
@@ -2878,18 +2889,16 @@ static int nested_vmx_check_host_state(sreturn -EINVAL ;#ifdef CONFIG_X86_64 - ia32e = !!(vcpu->arch.efer &EFER_LMA ); + ia32e = !!(vmcs12->vm_exit_controls &VM_EXIT_HOST_ADDR_SPACE_SIZE );#else ia32e = false;#endif if (ia32e) { -if (CC (!(vmcs12->vm_exit_controls &VM_EXIT_HOST_ADDR_SPACE_SIZE )) || -CC (!(vmcs12->host_cr4 &X86_CR4_PAE ))) +if (CC (!(vmcs12->host_cr4 &X86_CR4_PAE )))return -EINVAL ; }else { -if (CC (vmcs12->vm_exit_controls &VM_EXIT_HOST_ADDR_SPACE_SIZE ) || -CC (vmcs12->vm_entry_controls &VM_ENTRY_IA32E_MODE ) || +if (CC (vmcs12->vm_entry_controls &VM_ENTRY_IA32E_MODE ) ||CC (vmcs12->host_cr4 &X86_CR4_PCIDE ) ||CC ((vmcs12->host_rip) >>32 ))return -EINVAL ;
@@ -3559,6 +3568,9 @@ static int nested_vmx_run(struct kvm_vcpif (nested_vmx_check_controls (vcpu, vmcs12))return nested_vmx_fail (vcpu,VMXERR_ENTRY_INVALID_CONTROL_FIELD ); +if (nested_vmx_check_address_space_size (vcpu, vmcs12)) +return nested_vmx_fail (vcpu,VMXERR_ENTRY_INVALID_HOST_STATE_FIELD ); +if (nested_vmx_check_host_state (vcpu, vmcs12))return nested_vmx_fail (vcpu,VMXERR_ENTRY_INVALID_HOST_STATE_FIELD );
Greg Kroah-Hartman
maintainerPATCH 226/279 drm/cma-helper: Release non-coherent memory with dma_free_noncoherent()
In reply to Greg Kroah-HartmanFrom: Thomas Zimmermann <[email protected]>
commit 995f54ea962e03ec08b8bc6a4fe11a32b420edd3 upstream.
The GEM CMA helpers allocate non-coherent (i.e., cached) backing storage with dma_alloc_noncoherent(), but release it with dma_free_wc(). Fix this with a call to dma_free_noncoherent(). Writecombining storage is still released with dma_free_wc().
--- drivers/gpu/drm/drm_gem_cma_helper.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-)
drivers/gpu/drm/drm_gem_cma_helper.c +7 -2
@@ -210,8 +210,13 @@ void drm_gem_cma_free_object(struct drm_dma_buf_vunmap (gem_obj->import_attach->dmabuf, &map);drm_prime_gem_destroy (gem_obj, cma_obj->sgt); }else if (cma_obj->vaddr) { -dma_free_wc (gem_obj->dev->dev, cma_obj->base.size, - cma_obj->vaddr, cma_obj->paddr); +if (cma_obj->map_noncoherent) +dma_free_noncoherent (gem_obj->dev->dev, cma_obj->base.size, + cma_obj->vaddr, cma_obj->paddr, +DMA_TO_DEVICE ); +else +dma_free_wc (gem_obj->dev->dev, cma_obj->base.size, + cma_obj->vaddr, cma_obj->paddr); }drm_gem_object_release (gem_obj);
Greg Kroah-Hartman
maintainerPATCH 227/279 printk: restore flushing of NMI buffers on remote CPUs after NMI backtraces
In reply to Greg Kroah-HartmanFrom: Nicholas Piggin <[email protected]>
commit 5d5e4522a7f404d1a96fd6c703989d32a9c9568d upstream.
printk from NMI context relies on irq work being raised on the local CPU to print to console. This can be a problem if the NMI was raised by a lockup detector to print lockup stack and regs, because the CPU may not enable irqs (because it is locked up).
Introduce printk_trigger_flush() that can be called another CPU to try to get those messages to the console, call that where printk_safe_flush was previously called.
--- arch/powerpc/kernel/watchdog.c | 6 ++++++ include/linux/printk.h | 4 ++++ kernel/printk/printk.c | 5 +++++ lib/nmi_backtrace.c | 6 ++++++ 4 files changed, 21 insertions(+)
arch/powerpc/kernel/watchdog.c +6 -0
@@ -187,6 +187,12 @@ static void watchdog_smp_panic(int cpu,if (sysctl_hardlockup_all_cpu_backtrace)trigger_allbutself_cpu_backtrace (); +/* +* Force flush any remote buffers that might be stuck in IRQ context +* and therefore could not run their irq_work. +*/ +printk_trigger_flush (); +if (hardlockup_panic)nmi_panic (NULL ,"Hard LOCKUP" );
include/linux/printk.h +4 -0
@@ -198,6 +198,7 @@ void dump_stack_print_info(const char *lvoid show_regs_print_info (const char *log_lvl); extern asmlinkage void dump_stack_lvl(const char *log_lvl) __cold; extern asmlinkage void dump_stack(void) __cold; +void printk_trigger_flush (void );#else static inline __printf(1, 0) int vprintk(const char *s, va_list args)
@@ -274,6 +275,9 @@ static inline void dump_stack_lvl(conststatic inline void dump_stack (void ) { } +static inline void printk_trigger_flush (void ) +{ +}#endif #ifdef CONFIG_SMP
kernel/printk/printk.c +5 -0
@@ -3252,6 +3252,11 @@ void defer_console_output(void)preempt_enable (); } +void printk_trigger_flush (void ) +{ +defer_console_output (); +} +int vprintk_deferred (const char *fmt,va_list args) {int r;
lib/nmi_backtrace.c +6 -0
@@ -75,6 +75,12 @@ void nmi_trigger_cpumask_backtrace(consttouch_softlockup_watchdog (); } +/* +* Force flush any remote buffers that might be stuck in IRQ context +* and therefore could not run their irq_work. +*/ +printk_trigger_flush (); +clear_bit_unlock (0 , &backtrace_flag);put_cpu (); }
Greg Kroah-Hartman
maintainerPATCH 228/279 udf: Fix crash after seekdir
In reply to Greg Kroah-HartmanFrom: Jan Kara <[email protected]>
commit a48fc69fe6588b48d878d69de223b91a386a7cb4 upstream.
udf_readdir() didn't validate the directory position it should start reading from. Thus when user uses lseek(2) on directory file descriptor it can trick udf_readdir() into reading from a position in the middle of directory entry which then upsets directory parsing code resulting in errors or even possible kernel crashes. Similarly when the directory is modified between two readdir calls, the directory position need not be valid anymore.
Add code to validate current offset in the directory. This is actually rather expensive for UDF as we need to read from the beginning of the directory and parse all directory entries. This is because in UDF a directory is just a stream of data containing directory entries and since file names are fully under user's control we cannot depend on detecting magic numbers and checksums in the header of directory entry as a malicious attacker could fake them. We skip this step if we detect that nothing changed since the last readdir call.
--- fs/udf/dir.c | 32 ++++++++++++++++++++++++++++++-- fs/udf/namei.c | 3 +++ fs/udf/super.c | 2 ++ 3 files changed, 35 insertions(+), 2 deletions(-)
fs/udf/dir.c +30 -2
@@ -31,6 +31,7 @@#include <linux/mm.h> #include <linux/slab.h> #include <linux/bio.h> +#include <linux/iversion.h> #include "udf_i.h" #include "udf_sb.h"
@@ -43,7 +44,7 @@ static int udf_readdir(struct file *filestruct fileIdentDesc *fi =NULL ;struct fileIdentDesc cfi;udf_pblk_t block, iblock; -loff_t nf_pos; +loff_t nf_pos, emit_pos =0 ;int flen;unsigned char *fname =NULL , *copy_name =NULL ;unsigned char *nameptr;
@@ -57,6 +58,7 @@ static int udf_readdir(struct file *fileint i, num, ret =0 ;struct extent_position epos = {NULL ,0 , {0 ,0 } };struct super_block *sb = dir->i_sb; +bool pos_valid = false;if (ctx->pos ==0 ) {if (!dir_emit_dot (file, ctx))
@@ -67,6 +69,21 @@ static int udf_readdir(struct file *fileif (nf_pos >= size) goto out; +/* +* Something changed since last readdir (either lseek was called or dir +* changed)? We need to verify the position correctly points at the +* beginning of some dir entry so that the directory parsing code does +* not get confused. Since UDF does not have any reliable way of +* identifying beginning of dir entry (names are under user control), +* we need to scan the directory from the beginning. +*/ +if (!inode_eq_iversion (dir, file->f_version)) { + emit_pos = nf_pos; + nf_pos =0 ; + }else { + pos_valid = true; + } + fname =kmalloc (UDF_NAME_LEN ,GFP_NOFS );if (!fname) { ret = -ENOMEM ;
@@ -122,13 +139,21 @@ static int udf_readdir(struct file *filewhile (nf_pos < size) {struct kernel_lb_addr tloc; +loff_t cur_pos = nf_pos; - ctx->pos = (nf_pos >>2 ) +1 ; +/* Update file position only if we got past the current one */ +if (nf_pos >= emit_pos) { + ctx->pos = (nf_pos >>2 ) +1 ; + pos_valid = true; + } fi =udf_fileident_read (dir, &nf_pos, &fibh, &cfi, &epos, &eloc, &elen, &offset);if (!fi) goto out; +/* Still not at offset where user asked us to read from? */ +if (cur_pos < emit_pos) +continue ; liu =le16_to_cpu (cfi.lengthOfImpUse); lfi = cfi.lengthFileIdent;
@@ -186,8 +211,11 @@ static int udf_readdir(struct file *file }/* end while */ ctx->pos = (nf_pos >>2 ) +1 ; + pos_valid = true; out: +if (pos_valid) + file->f_version =inode_query_iversion (dir);if (fibh.sbh != fibh.ebh)brelse (fibh.ebh);brelse (fibh.sbh);
fs/udf/namei.c +3 -0
@@ -30,6 +30,7 @@#include <linux/sched.h> #include <linux/crc-itu-t.h> #include <linux/exportfs.h> +#include <linux/iversion.h> static inline int udf_match (int len1,const unsigned char *name1,int len2,const unsigned char *name2)
@@ -134,6 +135,8 @@ int udf_write_fi(struct inode *inode, stmark_buffer_dirty_inode (fibh->ebh, inode);mark_buffer_dirty_inode (fibh->sbh, inode); } +inode_inc_iversion (inode); +return 0 ; }
fs/udf/super.c +2 -0
@@ -57,6 +57,7 @@#include <linux/crc-itu-t.h> #include <linux/log2.h> #include <asm/byteorder.h> +#include <linux/iversion.h> #include "udf_sb.h" #include "udf_i.h"
@@ -149,6 +150,7 @@ static struct inode *udf_alloc_inode(strinit_rwsem (&ei->i_data_sem); ei->cached_extent.lstart =-1 ;spin_lock_init (&ei->i_extent_cache_lock); +inode_set_iversion (&ei->vfs_inode,1 );return &ei->vfs_inode; }
Greg Kroah-Hartman
maintainerPATCH 229/279 spi: fix use-after-free of the add_lock mutex
In reply to Greg Kroah-HartmanFrom: Michael Walle <[email protected]>
commit 6c53b45c71b4920b5e62f0ea8079a1da382b9434 upstream.
Commit 6098475d4cb4 ("spi: Fix deadlock when adding SPI controllers on SPI buses") introduced a per-controller mutex. But mutex_unlock() of said lock is called after the controller is already freed:
spi_unregister_controller(ctlr)
-> put_device(&ctlr->dev)
-> spi_controller_release(dev)
-> mutex_unlock(&ctrl->add_lock)
Move the put_device() after the mutex_unlock().
--- drivers/spi/spi.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-)
drivers/spi/spi.c +6 -6
@@ -3020,12 +3020,6 @@ void spi_unregister_controller(struct spdevice_del (&ctlr->dev); -/* Release the last reference on the controller if its driver -* has not yet been converted to devm_spi_alloc_master/slave(). -*/ -if (!ctlr->devm_allocated) -put_device (&ctlr->dev); -/* free bus id */ mutex_lock (&board_lock);if (found == ctlr)
@@ -3034,6 +3028,12 @@ void spi_unregister_controller(struct spif (IS_ENABLED (CONFIG_SPI_DYNAMIC ))mutex_unlock (&ctlr->add_lock); + +/* Release the last reference on the controller if its driver +* has not yet been converted to devm_spi_alloc_master/slave(). +*/ +if (!ctlr->devm_allocated) +put_device (&ctlr->dev); }EXPORT_SYMBOL_GPL (spi_unregister_controller);
Greg Kroah-Hartman
maintainerPATCH 230/279 net: stmmac: socfpga: add runtime suspend/resume callback for stratix10 platform
In reply to Greg Kroah-HartmanFrom: Meng Li <[email protected]>
commit 9119570039481d56350af1c636f040fb300b8cf3 upstream.
According to upstream commit 5ec55823438e("net: stmmac:
add clocks management for gmac driver"), it improve clocks management for stmmac driver. So, it is necessary to implement the runtime callback in dwmac-socfpga driver because it doesn't use the common stmmac_pltfr_pm_ops instance. Otherwise, clocks are not disabled when system enters suspend status.
--- drivers/net/ethernet/stmicro/stmmac/dwmac-socfpga.c | 24 ++++++++++++++++++-- 1 file changed, 22 insertions(+), 2 deletions(-)
drivers/net/ethernet/stmicro/stmmac/dwmac-socfpga.c +22 -2
@@ -485,8 +485,28 @@ static int socfpga_dwmac_resume(struct d }#endif /* CONFIG_PM_SLEEP */ -static SIMPLE_DEV_PM_OPS(socfpga_dwmac_pm_ops, stmmac_suspend, - socfpga_dwmac_resume); +static int __maybe_unused socfpga_dwmac_runtime_suspend (struct device *dev) +{ +struct net_device *ndev =dev_get_drvdata (dev); +struct stmmac_priv *priv =netdev_priv (ndev); + +stmmac_bus_clks_config (priv, false); + +return 0 ; +} + +static int __maybe_unused socfpga_dwmac_runtime_resume (struct device *dev) +{ +struct net_device *ndev =dev_get_drvdata (dev); +struct stmmac_priv *priv =netdev_priv (ndev); + +return stmmac_bus_clks_config (priv, true); +} + +static const struct dev_pm_ops socfpga_dwmac_pm_ops = { + SET_SYSTEM_SLEEP_PM_OPS(stmmac_suspend, socfpga_dwmac_resume) +SET_RUNTIME_PM_OPS (socfpga_dwmac_runtime_suspend, socfpga_dwmac_runtime_resume,NULL ) +};static const struct socfpga_dwmac_ops socfpga_gen5_ops = { .set_phy_mode = socfpga_gen5_set_phy_mode,
Greg Kroah-Hartman
maintainerPATCH 231/279 Drivers: hv: balloon: Use VMBUS_RING_SIZE() wrapper for dm_ring_size
In reply to Greg Kroah-HartmanFrom: Boqun Feng <[email protected]>
commit 8a7eb2d476c6823cd44d8c25a6230a52417d7ef8 upstream.
Baihua reported an error when boot an ARM64 guest with PAGE_SIZE=64k and BALLOON is enabled:
hv_vmbus: registering driver hv_balloon hv_vmbus: probe failed for device 1eccfd72-4b41-45ef-b73a-4a6e44c12924 (-22)
The cause of this is that the ringbuffer size for hv_balloon is not adjusted with VMBUS_RING_SIZE(), which makes the size not large enough for ringbuffers on guest with PAGE_SIZE=64k. Therefore use VMBUS_RING_SIZE() to calculate the ringbuffer size. Note that the old size (20 * 1024) counts a 4k header in the total size, while VMBUS_RING_SIZE() expects the parameter as the payload size, so use 16 * 1024.
--- drivers/hv/hv_balloon.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/hv/hv_balloon.c +1 -1
@@ -480,7 +480,7 @@ module_param(pressure_report_delay, uintMODULE_PARM_DESC (pressure_report_delay,"Delay in secs in reporting pressure" );static atomic_t trans_id =ATOMIC_INIT (0 ); -static int dm_ring_size =20 *1024 ; +static int dm_ring_size =VMBUS_RING_SIZE (16 *1024 );/* * Driver specific state.
Greg Kroah-Hartman
maintainerPATCH 232/279 btrfs: fix memory ordering between normal and ordered work functions
In reply to Greg Kroah-HartmanFrom: Nikolay Borisov <[email protected]>
commit 45da9c1767ac31857df572f0a909fbe88fd5a7e9 upstream.
Ordered work functions aren't guaranteed to be handled by the same thread which executed the normal work functions. The only way execution between normal/ordered functions is synchronized is via the WORK_DONE_BIT, unfortunately the used bitops don't guarantee any ordering whatsoever.
This manifested as seemingly inexplicable crashes on ARM64, where async_chunk::inode is seen as non-null in async_cow_submit which causes submit_compressed_extents to be called and crash occurs because async_chunk::inode suddenly became NULL. The call trace was similar to:
pc : submit_compressed_extents+0x38/0x3d0
lr : async_cow_submit+0x50/0xd0
sp : ffff800015d4bc20 <registers omitted for brevity>
Call trace:
submit_compressed_extents+0x38/0x3d0
async_cow_submit+0x50/0xd0
run_ordered_work+0xc8/0x280
btrfs_work_helper+0x98/0x250
process_one_work+0x1f0/0x4ac
worker_thread+0x188/0x504
kthread+0x110/0x114
ret_from_fork+0x10/0x18Fix this by adding respective barrier calls which ensure that all accesses preceding setting of WORK_DONE_BIT are strictly ordered before setting the flag. At the same time add a read barrier after reading of WORK_DONE_BIT in run_ordered_work which ensures all subsequent loads would be strictly ordered after reading the bit. This in turn ensures are all accesses before WORK_DONE_BIT are going to be strictly ordered before any access that can occur in ordered_func.
--- fs/btrfs/async-thread.c | 14 ++++++++++++++ 1 file changed, 14 insertions(+)
fs/btrfs/async-thread.c +14 -0
@@ -234,6 +234,13 @@ static void run_ordered_work(struct __bt ordered_list);if (!test_bit (WORK_DONE_BIT , &work->flags))break ; +/* +* Orders all subsequent loads after reading WORK_DONE_BIT, +* paired with the smp_mb__before_atomic in btrfs_work_helper +* this guarantees that the ordered function will see all +* updates from ordinary work function. +*/ +smp_rmb ();/* * we are going to call the ordered done function, but
@@ -317,6 +324,13 @@ static void btrfs_work_helper(struct worthresh_exec_hook (wq); work->func (work);if (need_order) { +/* +* Ensures all memory accesses done in the work function are +* ordered before setting the WORK_DONE_BIT. Ensuring the thread +* which is going to executed the ordered work sees them. +* Pairs with the smp_rmb in run_ordered_work. +*/ +smp_mb__before_atomic ();set_bit (WORK_DONE_BIT , &work->flags);run_ordered_work (wq, work); }else {
Greg Kroah-Hartman
maintainerPATCH 233/279 fs: handle circular mappings correctly
In reply to Greg Kroah-HartmanFrom: Christian Brauner <[email protected]>
commit 968219708108440b23bc292e0486e3cc1d9a1bed upstream.
When calling setattr_prepare() to determine the validity of the attributes the ia_{g,u}id fields contain the value that will be written to inode->i_{g,u}id. When the {g,u}id attribute of the file isn't altered and the caller's fs{g,u}id matches the current {g,u}id attribute the attribute change is allowed.
The value in ia_{g,u}id does already account for idmapped mounts and will have taken the relevant idmapping into account. So in order to verify that the {g,u}id attribute isn't changed we simple need to compare the ia_{g,u}id value against the inode's i_{g,u}id value.
This only has any meaning for idmapped mounts as idmapping helpers are idempotent without them. And for idmapped mounts this really only has a meaning when circular idmappings are used, i.e. mappings where e.g. id 1000 is mapped to id 1001 and id 1001 is mapped to id 1000. Such ciruclar mappings can e.g. be useful when sharing the same home directory between multiple users at the same time.
As an example consider a directory with two files: /source/file1 owned by {g,u}id 1000 and /source/file2 owned by {g,u}id 1001. Assume we create an idmapped mount at /target with an idmapping that maps files owned by {g,u}id 1000 to being owned by {g,u}id 1001 and files owned by {g,u}id 1001 to being owned by {g,u}id 1000. In effect, the idmapped mount at /target switches the ownership of /source/file1 and source/file2, i.e. /target/file1 will be owned by {g,u}id 1001 and /target/file2 will be owned by {g,u}id 1000.
This means that a user with fs{g,u}id 1000 must be allowed to setattr /target/file2 from {g,u}id 1000 to {g,u}id 1000. Similar, a user with fs{g,u}id 1001 must be allowed to setattr /target/file1 from {g,u}id 1001 to {g,u}id 1001. Conversely, a user with fs{g,u}id 1000 must fail to setattr /target/file1 from {g,u}id 1001 to {g,u}id 1000. And a user with fs{g,u}id 1001 must fail to setattr /target/file2 from {g,u}id 1000 to {g,u}id 1000. Both cases must fail with EPERM for non-capable callers.
Before this patch we could end up denying legitimate attribute changes and allowing invalid attribute changes when circular mappings are used. To even get into this situation the caller must've been privileged both to create that mapping and to create that idmapped mount.
This hasn't been seen in the wild anywhere but came up when expanding the testsuite during work on a series of hardening patches. All idmapped fstests pass without any regressions and we add new tests to verify the behavior of circular mappings.
--- fs/attr.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
fs/attr.c +2 -2
@@ -35,7 +35,7 @@ static bool chown_ok(struct user_namespa kuid_t uid) {kuid_t kuid =i_uid_into_mnt (mnt_userns, inode); -if (uid_eq (current_fsuid (), kuid) &&uid_eq (uid, kuid)) +if (uid_eq (current_fsuid (), kuid) &&uid_eq (uid, inode->i_uid))return true;if (capable_wrt_inode_uidgid (mnt_userns, inode,CAP_CHOWN ))return true;
@@ -62,7 +62,7 @@ static bool chgrp_ok(struct user_namespa {kgid_t kgid =i_gid_into_mnt (mnt_userns, inode);if (uid_eq (current_fsuid (),i_uid_into_mnt (mnt_userns, inode)) && - (in_group_p (gid) ||gid_eq (gid, kgid))) + (in_group_p (gid) ||gid_eq (gid, inode->i_gid)))return true;if (capable_wrt_inode_uidgid (mnt_userns, inode,CAP_CHOWN ))return true;
Greg Kroah-Hartman
maintainerPATCH 234/279 net: stmmac: Fix signed/unsigned wreckage
In reply to Greg Kroah-HartmanFrom: Thomas Gleixner <[email protected]>
commit 3751c3d34cd5a750c86d1c8eaf217d8faf7f9325 upstream.
The recent addition of timestamp correction to compensate the CDC error introduced a subtle signed/unsigned bug in stmmac_get_tx_hwtstamp() while it managed for some obscure reason to avoid that in stmmac_get_rx_hwtstamp().
The issue is:
s64 adjust = 0;
u64 ns;
adjust += -(2 * (NSEC_PER_SEC / priv->plat->clk_ptp_rate));
ns += adjust;works by chance on 64bit, but falls apart on 32bit because the compiler knows that adjust fits into 32bit and then treats the addition as a u64 + u32 resulting in an off by ~2 seconds failure.
The RX variant uses an u64 for adjust and does the adjustment via
ns -= adjust;
because consistency is obviously overrated.
Get rid of the pointless zero initialized adjust variable and do:
ns -= (2 * NSEC_PER_SEC) / priv->plat->clk_ptp_rate;
which is obviously correct and spares the adjust obfuscation. Aside of that it yields a more accurate result because the multiplication takes place before the integer divide truncation and not afterwards.
Stick the calculation into an inline so it can't be accidentally disimproved. Return an u32 from that inline as the result is guaranteed to fit which lets the compiler optimize the substraction.
--- drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 23 +++++++++------------- 1 file changed, 10 insertions(+), 13 deletions(-)
drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +10 -13
@@ -511,6 +511,14 @@ bool stmmac_eee_init(struct stmmac_privreturn true; } +static inline u32 stmmac_cdc_adjust (struct stmmac_priv *priv) +{ +/* Correct the clk domain crossing(CDC) error */ +if (priv->plat->has_gmac4 && priv->plat->clk_ptp_rate) +return (2 *NSEC_PER_SEC ) / priv->plat->clk_ptp_rate; +return 0 ; +} +/* stmmac_get_tx_hwtstamp - get HW TX timestamps * @priv: driver private structure * @p : descriptor pointer
@@ -524,7 +532,6 @@ static void stmmac_get_tx_hwtstamp(struc {struct skb_shared_hwtstamps shhwtstamp;bool found = false; -s64 adjust =0 ;u64 ns =0 ;if (!priv->hwts_tx_en)
@@ -543,12 +550,7 @@ static void stmmac_get_tx_hwtstamp(struc }if (found) { -/* Correct the clk domain crossing(CDC) error */ -if (priv->plat->has_gmac4 && priv->plat->clk_ptp_rate) { - adjust += -(2 * (NSEC_PER_SEC / - priv->plat->clk_ptp_rate)); - ns += adjust; - } + ns -=stmmac_cdc_adjust (priv);memset (&shhwtstamp,0 ,sizeof (struct skb_shared_hwtstamps )); shhwtstamp.hwtstamp =ns_to_ktime (ns);
@@ -573,7 +575,6 @@ static void stmmac_get_rx_hwtstamp(struc {struct skb_shared_hwtstamps *shhwtstamp =NULL ;struct dma_desc *desc = p; -u64 adjust =0 ;u64 ns =0 ;if (!priv->hwts_rx_en)
@@ -586,11 +587,7 @@ static void stmmac_get_rx_hwtstamp(strucif (stmmac_get_rx_timestamp_status (priv, p, np, priv->adv_ts)) {stmmac_get_timestamp (priv, desc, priv->adv_ts, &ns); -/* Correct the clk domain crossing(CDC) error */ -if (priv->plat->has_gmac4 && priv->plat->clk_ptp_rate) { - adjust +=2 * (NSEC_PER_SEC / priv->plat->clk_ptp_rate); - ns -= adjust; - } + ns -=stmmac_cdc_adjust (priv);netdev_dbg (priv->dev,"get valid RX hw timestamp %llu\n" , ns); shhwtstamp =skb_hwtstamps (skb);
Greg Kroah-Hartman
maintainerPATCH 235/279 parisc/sticon: fix reverse colors
In reply to Greg Kroah-HartmanFrom: Sven Schnelle <[email protected]>
commit bec05f33ebc1006899c6d3e59a00c58881fe7626 upstream.
sticon_build_attr() checked the reverse argument and flipped background and foreground color, but returned the non-reverse value afterwards. Fix this and also add two local variables for foreground and background color to make the code easier to read.
--- drivers/video/console/sticon.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-)
drivers/video/console/sticon.c +6 -6
@@ -332,13 +332,13 @@ static u8 sticon_build_attr(struct vc_da bool blink, bool underline, bool reverse, bool italic) { - u8 attr = ((color & 0x70) >> 1) | ((color & 7)); + u8 fg = color & 7; +u8 bg = (color &0x70 ) >>4 ; -if (reverse) { - color = ((color >>3 ) &0x7 ) | ((color &0x7 ) <<3 ); - } - -return attr; +if (reverse) +return (fg <<3 ) | bg; +else +return (bg <<3 ) | fg; }static void sticon_invert_region (struct vc_data *conp,u16 *p,int count)
Greg Kroah-Hartman
maintainerPATCH 236/279 cfg80211: call cfg80211_stop_ap when switch from P2P_GO type
In reply to Greg Kroah-HartmanFrom: Nguyen Dinh Phi <[email protected]>
commit 563fbefed46ae4c1f70cffb8eb54c02df480b2c2 upstream.
If the userspace tools switch from NL80211_IFTYPE_P2P_GO to NL80211_IFTYPE_ADHOC via send_msg(NL80211_CMD_SET_INTERFACE), it does not call the cleanup cfg80211_stop_ap(), this leads to the initialization of in-use data. For example, this path re-init the sdata->assigned_chanctx_list while it is still an element of assigned_vifs list, and makes that linked list corrupt.
--- net/wireless/util.c | 1 + 1 file changed, 1 insertion(+)
net/wireless/util.c +1 -0
@@ -1044,6 +1044,7 @@ int cfg80211_change_iface(struct cfg8021switch (otype) {case NL80211_IFTYPE_AP : +case NL80211_IFTYPE_P2P_GO :cfg80211_stop_ap (rdev, dev, true);break ;case NL80211_IFTYPE_ADHOC :
Greg Kroah-Hartman
maintainerPATCH 237/279 mac80211: fix radiotap header generation
In reply to Greg Kroah-HartmanFrom: Johannes Berg <[email protected]>
commit c033a38a81bc539d6c0db8c5387e0b14d819a0cf upstream.
In commit 8c89f7b3d3f2 ("mac80211: Use flex-array for radiotap header bitmap") we accidentally pointed the position to the wrong place, so we overwrite a present bitmap, and thus cause all kinds of trouble.
To see the issue, note that the previous code read:
pos = (void *)(it_present + 1);
The requirement now is that we need to calculate pos via it_optional, to not trigger the compiler hardening checks, as:
pos = (void *)&rthdr->it_optional[...];
Rewriting the original expression, we get (obviously, since that just adds "+ x - x" terms):
pos = (void *)(it_present + 1 + rthdr->it_optional - rthdr->it_optional)
and moving the "+ rthdr->it_optional" outside to be used as an array:
pos = (void *)&rthdr->it_optional[it_present + 1 - rthdr->it_optional];
The original is off by one, fix it.
--- net/mac80211/rx.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
net/mac80211/rx.c +1 -1
@@ -364,7 +364,7 @@ ieee80211_add_rx_radiotap_header(struct* the compiler to think we have walked past the end of the * struct member. */ - pos = (void *)&rthdr->it_optional[it_present - rthdr->it_optional]; + pos = (void *)&rthdr->it_optional[it_present +1 - rthdr->it_optional];/* the order of the following fields is important */
Greg Kroah-Hartman
maintainerPATCH 238/279 mac80211: drop check for DONT_REORDER in __ieee80211_select_queue
In reply to Greg Kroah-HartmanFrom: Felix Fietkau <[email protected]>
commit f6ab25d41b18f3d26883cb9c20875e1a85c4f05b upstream.
When __ieee80211_select_queue is called, skb->cb has not been cleared yet, which means that info->control.flags can contain garbage.
In some cases this leads to IEEE80211_TX_CTRL_DONT_REORDER being set, causing packets marked for other queues to randomly end up in BE instead.
This flag only needs to be checked in ieee80211_select_queue_80211, since the radiotap parser is the only piece of code that sets it
--- net/mac80211/wme.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-)
net/mac80211/wme.c +1 -2
@@ -143,7 +143,6 @@ u16 ieee80211_select_queue_80211(structu16 __ieee80211_select_queue (struct ieee80211_sub_if_data *sdata,struct sta_info *sta,struct sk_buff *skb) { -struct ieee80211_tx_info *info =IEEE80211_SKB_CB (skb);struct mac80211_qos_map *qos_map;bool qos;
@@ -156,7 +155,7 @@ u16 __ieee80211_select_queue(struct ieeeelse qos = false; -if (!qos || (info->control.flags &IEEE80211_TX_CTRL_DONT_REORDER )) { +if (!qos) { skb->priority =0 ;/* required for correct WPA/11i MIC */ return IEEE80211_AC_BE ; }
Greg Kroah-Hartman
maintainerPATCH 239/279 drm/amd/display: Update swizzle mode enums
In reply to Greg Kroah-HartmanFrom: Alvin Lee <[email protected]>
commit 58065a1e524de30df9a2d8214661d5d7eed0a2d9 upstream.
[Why]
Swizzle mode enum for DC_SW_VAR_R_X was existing, but not mapped correctly.
[How]
Update mapping and conversion for DC_SW_VAR_R_X.
--- drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c | 4 +++- drivers/gpu/drm/amd/display/dc/dml/display_mode_enums.h | 4 ++-- 2 files changed, 5 insertions(+), 3 deletions(-)
drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c +3 -1
@@ -1854,7 +1854,9 @@ static void swizzle_to_dml_params(case DC_SW_VAR_D_X : *sw_mode = dm_sw_var_d_x;break ; - +case DC_SW_VAR_R_X : + *sw_mode = dm_sw_var_r_x; +break ;default :ASSERT (0 );/* Not supported */ break ;
drivers/gpu/drm/amd/display/dc/dml/display_mode_enums.h +2 -2
@@ -80,11 +80,11 @@ enum dm_swizzle_mode { dm_sw_SPARE_13 =24 , dm_sw_64kb_s_x =25 , dm_sw_64kb_d_x =26 , - dm_sw_SPARE_14 =27 , + dm_sw_64kb_r_x =27 , dm_sw_SPARE_15 =28 , dm_sw_var_s_x =29 , dm_sw_var_d_x =30 , - dm_sw_64kb_r_x, + dm_sw_var_r_x =31 , dm_sw_gfx7_2d_thin_l_vp, dm_sw_gfx7_2d_thin_gl, };
Greg Kroah-Hartman
maintainerPATCH 240/279 drm/amd/display: Limit max DSC target bpp for specific monitors
In reply to Greg Kroah-HartmanFrom: Roman Li <[email protected]>
commit 55eea8ef98641f6e1e1c202bd3a49a57c1dd4059 upstream.
[Why]
Some monitors exhibit corruption at 16bpp DSC.
[How]
- Add helpers for patching edid caps.
- Use it for limiting DSC target bitrate to 15bpp for known monitors
--- drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c | 35 ++++++++++++++ 1 file changed, 35 insertions(+)
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c +35 -0
@@ -40,6 +40,39 @@#include "dm_helpers.h" +struct monitor_patch_info { +unsigned int manufacturer_id; +unsigned int product_id; +void (*patch_func)(struct dc_edid_caps *edid_caps,unsigned int param); +unsigned int patch_param; +}; +static void set_max_dsc_bpp_limit (struct dc_edid_caps *edid_caps,unsigned int param); + +static const struct monitor_patch_info monitor_patch_table[] = { +{0x6D1E ,0x5BBF , set_max_dsc_bpp_limit,15 }, +{0x6D1E ,0x5B9A , set_max_dsc_bpp_limit,15 }, +}; + +static void set_max_dsc_bpp_limit (struct dc_edid_caps *edid_caps,unsigned int param) +{ +if (edid_caps) + edid_caps->panel_patch.max_dsc_target_bpp_limit = param; +} + +static int amdgpu_dm_patch_edid_caps (struct dc_edid_caps *edid_caps) +{ +int i, ret =0 ; + +for (i =0 ; i <ARRAY_SIZE (monitor_patch_table); i++) +if ((edid_caps->manufacturer_id == monitor_patch_table[i].manufacturer_id) + && (edid_caps->product_id == monitor_patch_table[i].product_id)) { + monitor_patch_table[i].patch_func (edid_caps, monitor_patch_table[i].patch_param); + ret++; + } + +return ret; +} +/* dm_helpers_parse_edid_caps * * Parse edid caps
@@ -125,6 +158,8 @@ enum dc_edid_status dm_helpers_parse_edikfree (sads);kfree (sadb); +amdgpu_dm_patch_edid_caps (edid_caps); +return result; }
Greg Kroah-Hartman
maintainerPATCH 241/279 drm/i915/guc: Fix outstanding G2H accounting
In reply to Greg Kroah-HartmanFrom: Matthew Brost <[email protected]>
commit 669b949c1a44d0cb2bcd18ff6ab4fd0c21e7cf6f upstream.
A small race that could result in incorrect accounting of the number of outstanding G2H. Basically prior to this patch we did not increment the number of outstanding G2H if we encoutered a GT reset while sending a H2G. This was incorrect as the context state had already been updated to anticipate a G2H response thus the counter should be incremented.
As part of this change we remove a legacy (now unused) path that was the last caller requiring a G2H response that was not guaranteed to loop. This allows us to simplify the accounting as we don't need to handle the case where the send fails due to the channel being busy.
Also always use helper when decrementing this value.
v2 (Daniele): update GEM_BUG_ON check, pull in dead code removal from later patch, remove loop param from context_deregister.
--- drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c | 79 ++++++++++------------ 1 file changed, 37 insertions(+), 42 deletions(-)
drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c +37 -42
@@ -352,20 +352,29 @@ static inline void set_lrc_desc_registerxa_unlock_irqrestore (&guc->context_lookup, flags); } +static void decr_outstanding_submission_g2h (struct intel_guc *guc) +{ +if (atomic_dec_and_test (&guc->outstanding_submission_g2h)) +wake_up_all (&guc->ct.wq); +} +static int guc_submission_send_busy_loop (struct intel_guc *guc,const u32 *action,u32 len,u32 g2h_len_dw,bool loop) { -int err; - - err =intel_guc_send_busy_loop (guc, action, len, g2h_len_dw, loop); +/* +* We always loop when a send requires a reply (i.e. g2h_len_dw > 0), +* so we don't handle the case where we don't get a reply because we +* aborted the send due to the channel being busy. +*/ +GEM_BUG_ON (g2h_len_dw && !loop); -if (!err && g2h_len_dw) +if (g2h_len_dw)atomic_inc (&guc->outstanding_submission_g2h); -return err; +return intel_guc_send_busy_loop (guc, action, len, g2h_len_dw, loop); } int intel_guc_wait_for_pending_msg(struct intel_guc *guc,
@@ -616,7 +625,7 @@ static void scrub_guc_desc_for_outstandiinit_sched_state (ce);if (pending_enable || destroyed || deregister) { -atomic_dec (&guc->outstanding_submission_g2h); +decr_outstanding_submission_g2h (guc);if (deregister)guc_signal_context_fence (ce);if (destroyed) {
@@ -635,7 +644,7 @@ static void scrub_guc_desc_for_outstandiintel_engine_signal_breadcrumbs (ce->engine); }intel_context_sched_disable_unpin (ce); -atomic_dec (&guc->outstanding_submission_g2h); +decr_outstanding_submission_g2h (guc);spin_lock_irqsave (&ce->guc_state.lock, flags);guc_blocked_fence_complete (ce);spin_unlock_irqrestore (&ce->guc_state.lock, flags);
@@ -1233,8 +1242,7 @@ static int register_context(struct intel }static int __guc_action_deregister_context (struct intel_guc *guc, -u32 guc_id, -bool loop) +u32 guc_id) {u32 action[] = {INTEL_GUC_ACTION_DEREGISTER_CONTEXT ,
@@ -1243,16 +1251,16 @@ static int __guc_action_deregister_contereturn guc_submission_send_busy_loop (guc, action,ARRAY_SIZE (action),G2H_LEN_DW_DEREGISTER_CONTEXT , - loop); + true); } -static int deregister_context (struct intel_context *ce,u32 guc_id,bool loop) +static int deregister_context (struct intel_context *ce,u32 guc_id) {struct intel_guc *guc =ce_to_guc (ce);trace_intel_context_deregister (ce); -return __guc_action_deregister_context (guc, guc_id, loop); +return __guc_action_deregister_context (guc, guc_id); }static intel_engine_mask_t adjust_engine_mask (u8 class,intel_engine_mask_t mask)
@@ -1340,26 +1348,23 @@ static int guc_lrc_desc_pin(struct intel* registering this context. */ if (context_registered) { +bool disabled; +unsigned long flags; +trace_intel_context_steal_guc_id (ce); -if (!loop) { +GEM_BUG_ON (!loop); + +/* Seal race with Reset */ +spin_lock_irqsave (&ce->guc_state.lock, flags); + disabled =submission_disabled (guc); +if (likely (!disabled)) {set_context_wait_for_deregister_to_register (ce);intel_context_get (ce); - }else { -bool disabled; -unsigned long flags; - -/* Seal race with Reset */ -spin_lock_irqsave (&ce->guc_state.lock, flags); - disabled =submission_disabled (guc); -if (likely (!disabled)) { -set_context_wait_for_deregister_to_register (ce); -intel_context_get (ce); - } -spin_unlock_irqrestore (&ce->guc_state.lock, flags); -if (unlikely (disabled)) { -reset_lrc_desc (guc, desc_idx); -return 0 ;/* Will get registered later */ - } + } +spin_unlock_irqrestore (&ce->guc_state.lock, flags); +if (unlikely (disabled)) { +reset_lrc_desc (guc, desc_idx); +return 0 ;/* Will get registered later */ }/*
@@ -1367,13 +1372,9 @@ static int guc_lrc_desc_pin(struct intel* context whose guc_id was stolen. */ with_intel_runtime_pm (runtime_pm, wakeref) - ret = deregister_context(ce, ce->guc_id, loop); -if (unlikely (ret == -EBUSY )) { -clr_context_wait_for_deregister_to_register (ce); -intel_context_put (ce); - }else if (unlikely (ret == -ENODEV )) { + ret = deregister_context(ce, ce->guc_id); +if (unlikely (ret == -ENODEV )) ret =0 ;/* Will get registered later */ - } }else {with_intel_runtime_pm (runtime_pm, wakeref) ret = register_context(ce, loop);
@@ -1730,7 +1731,7 @@ static inline void guc_lrc_desc_unpin(stGEM_BUG_ON (context_enabled (ce));clr_context_registered (ce); -deregister_context (ce, ce->guc_id, true); +deregister_context (ce, ce->guc_id); }static void __guc_context_destroy (struct intel_context *ce)
@@ -2583,12 +2584,6 @@ g2h_context_lookup(struct intel_guc *gucreturn ce; } -static void decr_outstanding_submission_g2h (struct intel_guc *guc) -{ -if (atomic_dec_and_test (&guc->outstanding_submission_g2h)) -wake_up_all (&guc->ct.wq); -} -int intel_guc_deregister_done_process_msg (struct intel_guc *guc,const u32 *msg,u32 len)
Greg Kroah-Hartman
maintainerPATCH 242/279 drm/i915/guc: Dont enable scheduling on a banned context, guc_id invalid, not registered
In reply to Greg Kroah-HartmanFrom: Matthew Brost <[email protected]>
commit 9888beaaf118b6878347e1fe2b369fc66d756d18 upstream.
When unblocking a context, do not enable scheduling if the context is banned, guc_id invalid, or not registered.
v2:
(Daniele)
- Add helper for unblock
--- drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c | 22 +++++++++++++++++++--- 1 file changed, 19 insertions(+), 3 deletions(-)
drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c +19 -3
@@ -148,6 +148,7 @@ static inline void clr_context_registere#define SCHED_STATE_BLOCKED_SHIFT 4#define SCHED_STATE_BLOCKED BIT(SCHED_STATE_BLOCKED_SHIFT)#define SCHED_STATE_BLOCKED_MASK (0xfff << SCHED_STATE_BLOCKED_SHIFT) +static inline void init_sched_state (struct intel_context *ce) {/* Only should be called from guc_lrc_desc_pin() */
@@ -1549,6 +1550,23 @@ static struct i915_sw_fence *guc_contextreturn &ce->guc_blocked; } +#define SCHED_STATE_MULTI_BLOCKED_MASK \ + (SCHED_STATE_BLOCKED_MASK & ~SCHED_STATE_BLOCKED) +#define SCHED_STATE_NO_UNBLOCK \ + (SCHED_STATE_MULTI_BLOCKED_MASK | \ + SCHED_STATE_PENDING_DISABLE | \ + SCHED_STATE_BANNED) + +static bool context_cant_unblock (struct intel_context *ce) +{ +lockdep_assert_held (&ce->guc_state.lock); + +return (ce->guc_state.sched_state &SCHED_STATE_NO_UNBLOCK ) || +context_guc_id_invalid (ce) || + !lrc_desc_registered (ce_to_guc (ce), ce->guc_id) || + !intel_context_is_pinned (ce); +} +static void guc_context_unblock (struct intel_context *ce) {struct intel_guc *guc =ce_to_guc (ce);
@@ -1563,9 +1581,7 @@ static void guc_context_unblock(struct ispin_lock_irqsave (&ce->guc_state.lock, flags);if (unlikely (submission_disabled (guc) || - !intel_context_is_pinned (ce) || -context_pending_disable (ce) || -context_blocked (ce) >1 )) { +context_cant_unblock (ce))) { enable = false; }else { enable = true;
Greg Kroah-Hartman
maintainerPATCH 243/279 drm/i915/guc: Workaround reset G2H is received after schedule done G2H
In reply to Greg Kroah-HartmanFrom: Matthew Brost <[email protected]>
commit 1ca36cff0166b0483fe3b99e711e9c800ebbfaa4 upstream.
If the context is reset as a result of the request cancellation the context reset G2H is received after schedule disable done G2H which is the wrong order. The schedule disable done G2H release the waiting request cancellation code which resubmits the context. This races with the context reset G2H which also wants to resubmit the context but in this case it really should be a NOP as request cancellation code owns the resubmit. Use some clever tricks of checking the context state to seal this race until the GuC firmware is fixed.
v2:
(Checkpatch)
- Fix typos
v3:
(Daniele)
- State that is a bug in the GuC firmware
--- drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c | 41 ++++++++++++++++++---- 1 file changed, 35 insertions(+), 6 deletions(-)
drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c +35 -6
@@ -838,17 +838,33 @@ __unwind_incomplete_requests(struct intestatic void __guc_reset_context (struct intel_context *ce,bool stalled) {struct i915_request *rq; +unsigned long flags;u32 head; +bool skip = false;intel_context_get (ce);/* -* GuC will implicitly mark the context as non-schedulable -* when it sends the reset notification. Make sure our state -* reflects this change. The context will be marked enabled -* on resubmission. +* GuC will implicitly mark the context as non-schedulable when it sends +* the reset notification. Make sure our state reflects this change. The +* context will be marked enabled on resubmission. +* +* XXX: If the context is reset as a result of the request cancellation +* this G2H is received after the schedule disable complete G2H which is +* wrong as this creates a race between the request cancellation code +* re-submitting the context and this G2H handler. This is a bug in the +* GuC but can be worked around in the meantime but converting this to a +* NOP if a pending enable is in flight as this indicates that a request +* cancellation has occurred. */ -clr_context_enabled (ce); +spin_lock_irqsave (&ce->guc_state.lock, flags); +if (likely (!context_pending_enable (ce))) +clr_context_enabled (ce); +else + skip = true; +spin_unlock_irqrestore (&ce->guc_state.lock, flags); +if (unlikely (skip)) + goto out_put; rq =intel_context_find_active_request (ce);if (!rq) {
@@ -867,6 +883,7 @@ static void __guc_reset_context(struct i out_replay:guc_reset_state (ce, head, stalled);__unwind_incomplete_requests (ce); +out_put:intel_context_put (ce); }
@@ -1618,6 +1635,13 @@ static void guc_context_cancel_request(sguc_reset_state (ce,intel_ring_wrap (ce->ring, rq->head), true); } + +/* +* XXX: Racey if context is reset, see comment in +* __guc_reset_context(). +*/ +flush_work (&ce_to_guc (ce)->ct.requests.worker); +guc_context_unblock (ce); } }
@@ -2732,7 +2756,12 @@ static void guc_handle_context_reset(str {trace_intel_context_reset (ce); -if (likely (!intel_context_is_banned (ce))) { +/* +* XXX: Racey if request cancellation has occurred, see comment in +* __guc_reset_context(). +*/ +if (likely (!intel_context_is_banned (ce) && + !context_blocked (ce))) {capture_error_state (guc, ce);guc_context_replay (ce); }
Greg Kroah-Hartman
maintainerPATCH 244/279 drm/i915/guc: Dont drop ce->guc_active.lock when unwinding context
In reply to Greg Kroah-HartmanFrom: Matthew Brost <[email protected]>
commit 88209a8ecb8b8752322908a3c3362a001bdc3a39 upstream.
Don't drop ce->guc_active.lock when unwinding a context after reset. At one point we had to drop this because of a lock inversion but that is no longer the case. It is much safer to hold the lock so let's do that.
--- drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c | 4 ---- 1 file changed, 4 deletions(-)
drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c +0 -4
@@ -814,8 +814,6 @@ __unwind_incomplete_requests(struct intecontinue ;list_del_init (&rq->sched.link); -spin_unlock (&ce->guc_active.lock); -__i915_request_unsubmit (rq);/* Push the request back into the queue for later resubmission. */
@@ -828,8 +826,6 @@ __unwind_incomplete_requests(struct intelist_add_tail (&rq->sched.link, pl);set_bit (I915_FENCE_FLAG_PQUEUE , &rq->fence.flags); - -spin_lock (&ce->guc_active.lock); }spin_unlock (&ce->guc_active.lock);spin_unlock_irqrestore (&sched_engine->lock, flags);
Greg Kroah-Hartman
maintainerPATCH 245/279 drm/i915/guc: Unwind context requests in reverse order
In reply to Greg Kroah-HartmanFrom: Matthew Brost <[email protected]>
commit c39f51cc980dd918c5b3da61d54c4725785e766e upstream.
When unwinding requests on a reset context, if other requests in the context are in the priority list the requests could be resubmitted out of seqno order. Traverse the list of active requests in reverse and append to the head of the priority list to fix this.
--- drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-)
drivers/gpu/drm/i915/gt/uc/intel_guc_submission.c +4 -4
@@ -807,9 +807,9 @@ __unwind_incomplete_requests(struct intespin_lock_irqsave (&sched_engine->lock, flags);spin_lock (&ce->guc_active.lock); -list_for_each_entry_safe (rq, rn, - &ce->guc_active.requests, - sched.link) { +list_for_each_entry_safe_reverse (rq, rn, + &ce->guc_active.requests, + sched.link) {if (i915_request_completed (rq))continue ;
@@ -824,7 +824,7 @@ __unwind_incomplete_requests(struct inte }GEM_BUG_ON (i915_sched_engine_is_empty (sched_engine)); -list_add_tail (&rq->sched.link, pl); +list_add (&rq->sched.link, pl);set_bit (I915_FENCE_FLAG_PQUEUE , &rq->fence.flags); }spin_unlock (&ce->guc_active.lock);
Greg Kroah-Hartman
maintainerPATCH 246/279 drm/udl: fix control-message timeout
In reply to Greg Kroah-HartmanFrom: Johan Hovold <[email protected]>
commit 5591c8f79db1729d9c5ac7f5b4d3a5c26e262d93 upstream.
USB control-message timeouts are specified in milliseconds and should specifically not vary with CONFIG_HZ.
--- drivers/gpu/drm/udl/udl_connector.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/gpu/drm/udl/udl_connector.c +1 -1
@@ -30,7 +30,7 @@ static int udl_get_edid_block(void *dataint bval = (i + block *EDID_LENGTH ) <<8 ; ret =usb_control_msg (udev,usb_rcvctrlpipe (udev,0 ),0x02 , (0x80 | (0x02 <<5 )), bval, -0xA1 , read_buff,2 ,HZ ); +0xA1 , read_buff,2 ,1000 );if (ret <1 ) {DRM_ERROR ("Read EDID byte %d failed err %x\n" , i, ret);kfree (read_buff);
Greg Kroah-Hartman
maintainerPATCH 247/279 drm/prime: Fix use after free in mmap with drm_gem_ttm_mmap
In reply to Greg Kroah-HartmanFrom: Anand K Mistry <[email protected]>
commit 8244a3bc27b3efd057da154b8d7e414670d5044f upstream.
drm_gem_ttm_mmap() drops a reference to the gem object on success. If the gem object's refcount == 1 on entry to drm_gem_prime_mmap(), that drop will free the gem object, and the subsequent drm_gem_object_get() will be a UAF. Fix by grabbing a reference before calling the mmap helper.
This issue was forseen when the reference dropping was adding in commit 9786b65bc61ac ("drm/ttm: fix mmap refcounting"):
"For that to work properly the drm_gem_object_get() call in drm_gem_ttm_mmap() must be moved so it happens before calling obj->funcs->mmap(), otherwise the gem refcount would go down to zero."
--- drivers/gpu/drm/drm_prime.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/gpu/drm/drm_prime.c +4 -2
@@ -719,11 +719,13 @@ int drm_gem_prime_mmap(struct drm_gem_obif (obj->funcs && obj->funcs->mmap) { vma->vm_ops = obj->funcs->vm_ops; +drm_gem_object_get (obj); ret = obj->funcs->mmap (obj, vma); -if (ret) +if (ret) { +drm_gem_object_put (obj);return ret; + } vma->vm_private_data = obj; -drm_gem_object_get (obj);return 0 ; }
Greg Kroah-Hartman
maintainerPATCH 248/279 drm/nouveau: Add a dedicated mutex for the clients list
In reply to Greg Kroah-HartmanFrom: Jeremy Cline <[email protected]>
commit abae9164a421bc4a41a3769f01ebcd1f9d955e0e upstream.
Rather than protecting the nouveau_drm clients list with the lock within the "client" nouveau_cli, add a dedicated lock to serialize access to the list. This is both clearer and necessary to avoid lockdep being upset with us when we need to iterate through all the clients in the list and potentially lock their mutex, which is the same class as the lock protecting the entire list.
--- drivers/gpu/drm/nouveau/nouveau_drm.c | 10 ++++++---- drivers/gpu/drm/nouveau/nouveau_drv.h | 5 +++++ 2 files changed, 11 insertions(+), 4 deletions(-)
drivers/gpu/drm/nouveau/nouveau_drm.c +6 -4
@@ -562,6 +562,7 @@ nouveau_drm_device_init(struct drm_devicnvkm_dbgopt (nouveau_debug,"DRM" );INIT_LIST_HEAD (&drm->clients); +mutex_init (&drm->clients_lock);spin_lock_init (&drm->tile.lock);/* workaround an odd issue on nvc1 by disabling the device's
@@ -659,6 +660,7 @@ nouveau_drm_device_fini(struct drm_devicnouveau_cli_fini (&drm->client);nouveau_cli_fini (&drm->master);nvif_parent_dtor (&drm->parent); +mutex_destroy (&drm->clients_lock);kfree (drm); }
@@ -1090,9 +1092,9 @@ nouveau_drm_open(struct drm_device *dev, fpriv->driver_priv = cli; -mutex_lock (&drm->client.mutex); +mutex_lock (&drm->clients_lock);list_add (&cli->head, &drm->clients); -mutex_unlock (&drm->client.mutex); +mutex_unlock (&drm->clients_lock); done:if (ret && cli) {
@@ -1118,9 +1120,9 @@ nouveau_drm_postclose(struct drm_devicenouveau_abi16_fini (cli->abi16);mutex_unlock (&cli->mutex); -mutex_lock (&drm->client.mutex); +mutex_lock (&drm->clients_lock);list_del (&cli->head); -mutex_unlock (&drm->client.mutex); +mutex_unlock (&drm->clients_lock);nouveau_cli_fini (cli);kfree (cli);
drivers/gpu/drm/nouveau/nouveau_drv.h +5 -0
@@ -139,6 +139,11 @@ struct nouveau_drm {struct list_head clients; +/** +* @clients_lock: Protects access to the @clients list of &struct nouveau_cli. +*/ +struct mutex clients_lock; +u8 old_pm_cap;struct {
Greg Kroah-Hartman
maintainerPATCH 249/279 drm/nouveau: use drm_dev_unplug() during device removal
In reply to Greg Kroah-HartmanFrom: Jeremy Cline <[email protected]>
commit aff2299e0d81b26304ccc6a1ec0170e437f38efc upstream.
Nouveau does not currently support hot-unplugging, but it still makes sense to switch from drm_dev_unregister() to drm_dev_unplug(). drm_dev_unplug() calls drm_dev_unregister() after marking the device as unplugged, but only after any device critical sections are finished.
Since nouveau isn't using drm_dev_enter() and drm_dev_exit(), there are no critical sections so this is nearly functionally equivalent. However, the DRM layer does check to see if the device is unplugged, and if it is returns appropriate error codes.
In the future nouveau can add critical sections in order to truly support hot-unplugging.
--- drivers/gpu/drm/nouveau/nouveau_drm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/gpu/drm/nouveau/nouveau_drm.c +1 -1
@@ -798,7 +798,7 @@ nouveau_drm_device_remove(struct drm_devstruct nvkm_client *client;struct nvkm_device *device; -drm_dev_unregister (dev); +drm_dev_unplug (dev); client =nvxx_client (&drm->client.base); device =nvkm_device_find (client->device);
Greg Kroah-Hartman
maintainerPATCH 250/279 drm/nouveau: clean up all clients on device removal
In reply to Greg Kroah-HartmanFrom: Jeremy Cline <[email protected]>
commit f55aaf63bde0d0336c3823bb3713bd4a464abbcf upstream.
The postclose handler can run after the device has been removed (or the driver has been unbound) since userspace clients are free to hold the file open as long as they want. Because the device removal callback frees the entire nouveau_drm structure, any reference to it in the postclose handler will result in a use-after-free.
To reproduce this, one must simply open the device file, unbind the driver (or physically remove the device), and then close the device file. This was found and can be reproduced easily with the IGT core_hotunplug tests.
To avoid this, all clients are cleaned up in the device finalization rather than deferring it to the postclose handler, and the postclose handler is protected by a critical section which ensures the drm_dev_unplug() and the postclose handler won't race.
This is not an ideal fix, since as I understand the proposed plan for the kernel<->userspace interface for hotplug support, destroying the client before the file is closed will cause problems. However, I believe to properly fix this issue, the lifetime of the nouveau_drm structure needs to be extended to match the drm_device, and this proved to be a rather invasive change. Thus, I've broken this out so the fix can be easily backported.
This fixes with the two previous commits CVE-2020-27820 (Karol).
--- drivers/gpu/drm/nouveau/nouveau_drm.c | 30 ++++++++++++++++++++++++++++++ 1 file changed, 30 insertions(+)
drivers/gpu/drm/nouveau/nouveau_drm.c +30 -0
@@ -633,6 +633,7 @@ fail_alloc:static void nouveau_drm_device_fini (struct drm_device *dev) { +struct nouveau_cli *cli, *temp_cli;struct nouveau_drm *drm =nouveau_drm (dev);if (nouveau_pmops_runtime ()) {
@@ -657,6 +658,24 @@ nouveau_drm_device_fini(struct drm_devicnouveau_ttm_fini (drm);nouveau_vga_fini (drm); +/* +* There may be existing clients from as-yet unclosed files. For now, +* clean them up here rather than deferring until the file is closed, +* but this likely not correct if we want to support hot-unplugging +* properly. +*/ +mutex_lock (&drm->clients_lock); +list_for_each_entry_safe (cli, temp_cli, &drm->clients, head) { +list_del (&cli->head); +mutex_lock (&cli->mutex); +if (cli->abi16) +nouveau_abi16_fini (cli->abi16); +mutex_unlock (&cli->mutex); +nouveau_cli_fini (cli); +kfree (cli); + } +mutex_unlock (&drm->clients_lock); +nouveau_cli_fini (&drm->client);nouveau_cli_fini (&drm->master);nvif_parent_dtor (&drm->parent);
@@ -1112,6 +1131,16 @@ nouveau_drm_postclose(struct drm_device {struct nouveau_cli *cli =nouveau_cli (fpriv);struct nouveau_drm *drm =nouveau_drm (dev); +int dev_index; + +/* +* The device is gone, and as it currently stands all clients are +* cleaned up in the removal codepath. In the future this may change +* so that we can support hot-unplugging, but for now we immediately +* return to avoid a double-free situation. +*/ +if (!drm_dev_enter (dev, &dev_index)) +return ;pm_runtime_get_sync (dev->dev);
@@ -1128,6 +1157,7 @@ nouveau_drm_postclose(struct drm_devicekfree (cli);pm_runtime_mark_last_busy (dev->dev);pm_runtime_put_autosuspend (dev->dev); +drm_dev_exit (dev_index); } static const struct drm_ioctl_desc
Greg Kroah-Hartman
maintainerPATCH 251/279 drm/i915/dp: Ensure sink rate values are always valid
In reply to Greg Kroah-HartmanFrom: Imre Deak <[email protected]>
commit 6c34bd4532a3f39952952ddc102737595729afc4 upstream.
Atm, there are no sink rate values set for DP (vs. eDP) sinks until the DPCD capabilities are successfully read from the sink. During this time intel_dp->num_common_rates is 0 which can lead to a
intel_dp->common_rates[-1] (*)
access, which is an undefined behaviour, in the following cases:
- In intel_dp_sync_state(), if the encoder is enabled without a sink connected to the encoder's connector (BIOS enabled a monitor, but the user unplugged the monitor until the driver loaded).
- In intel_dp_sync_state() if the encoder is enabled with a sink connected, but for some reason the DPCD read has failed.
- In intel_dp_compute_link_config() if modesetting a connector without a sink connected on it.
- In intel_dp_compute_link_config() if modesetting a connector with a a sink connected on it, but before probing the connector first.
To avoid the (*) access in all the above cases, make sure that the sink rate table - and hence the common rate table - is always valid, by setting a default minimum sink rate when registering the connector before anything could use it.
I also considered setting all the DP link rates by default, so that modesetting with higher resolution modes also succeeds in the last two cases above. However in case a sink is not connected that would stop working after the first modeset, due to the LT fallback logic. So this would need more work, beyond the scope of this fix.
As I mentioned in the previous patch, I don't think the issue this patch fixes is user visible, however it is an undefined behaviour by definition and triggers a BUG() in CONFIG_UBSAN builds, hence CC:stable.
v2: Clear the default sink rates, before initializing these for eDP.
--- drivers/gpu/drm/i915/display/intel_dp.c | 11 +++++++++++ 1 file changed, 11 insertions(+)
drivers/gpu/drm/i915/display/intel_dp.c +11 -0
@@ -111,6 +111,12 @@ bool intel_dp_is_edp(struct intel_dp *instatic void intel_dp_unset_edid (struct intel_dp *intel_dp);static int intel_dp_dsc_compute_bpp (struct intel_dp *intel_dp,u8 dsc_max_bpc); +static void intel_dp_set_default_sink_rates (struct intel_dp *intel_dp) +{ + intel_dp->sink_rates[0 ] =162000 ; + intel_dp->num_sink_rates =1 ; +} +/* update sink rates from dpcd */ static void intel_dp_set_sink_rates (struct intel_dp *intel_dp) {
@@ -2462,6 +2468,9 @@ intel_edp_init_dpcd(struct intel_dp *int*/ intel_psr_init_dpcd (intel_dp); +/* Clear the default sink rates */ + intel_dp->num_sink_rates =0 ; +/* Read the eDP 1.4+ supported link rates. */ if (intel_dp->edp_dpcd[0 ] >=DP_EDP_14 ) {__le16 sink_rates[DP_MAX_SUPPORTED_RATES ];
@@ -5296,6 +5305,8 @@ intel_dp_init_connector(struct intel_digreturn false;intel_dp_set_source_rates (intel_dp); +intel_dp_set_default_sink_rates (intel_dp); +intel_dp_set_common_rates (intel_dp); intel_dp->reset_link_params = true; intel_dp->pps.pps_pipe =INVALID_PIPE ;
Greg Kroah-Hartman
maintainerPATCH 252/279 drm/i915/dp: Ensure max link params are always valid
In reply to Greg Kroah-HartmanFrom: Imre Deak <[email protected]>
commit cc99bc62ff6902688ee7bd3a7b25eefc620fbb6a upstream.
Atm until the DPCD for a connector is read the max link rate and lane count params are invalid. If the connector is modeset, in intel_dp_compute_config(), intel_dp_common_len_rate_limit(max_link_rate) will return 0, leading to a intel_dp->common_rates[-1] access.
Fix the above by making sure the max link params are always valid.
The above access leads to an undefined behaviour by definition, though not causing a user visible problem to my best knowledge, see the previous patch why. Nevertheless it is an undefined behaviour and it triggers a BUG() in CONFIG_UBSAN builds, hence CC:stable.
--- drivers/gpu/drm/i915/display/intel_dp.c | 18 ++++++++++-------- 1 file changed, 10 insertions(+), 8 deletions(-)
drivers/gpu/drm/i915/display/intel_dp.c +10 -8
@@ -1773,6 +1773,12 @@ void intel_dp_set_link_params(struct int intel_dp->lane_count = lane_count; } +static void intel_dp_reset_max_link_params (struct intel_dp *intel_dp) +{ + intel_dp->max_link_lane_count =intel_dp_max_common_lane_count (intel_dp); + intel_dp->max_link_rate =intel_dp_max_common_rate (intel_dp); +} +/* Enable backlight PWM and backlight PP control. */ void intel_edp_backlight_on (const struct intel_crtc_state *crtc_state,const struct drm_connector_state *conn_state)
@@ -1932,8 +1938,7 @@ void intel_dp_sync_state(struct intel_enif (intel_dp->dpcd[DP_DPCD_REV ] ==0 )intel_dp_get_dpcd (intel_dp); - intel_dp->max_link_lane_count =intel_dp_max_common_lane_count (intel_dp); - intel_dp->max_link_rate =intel_dp_max_common_rate (intel_dp); +intel_dp_reset_max_link_params (intel_dp); } bool intel_dp_initial_fastset_check(struct intel_encoder *encoder,
@@ -2506,6 +2511,7 @@ intel_edp_init_dpcd(struct intel_dp *intintel_dp_set_sink_rates (intel_dp);intel_dp_set_common_rates (intel_dp); +intel_dp_reset_max_link_params (intel_dp);/* Read the eDP DSC DPCD registers */ if (DISPLAY_VER (dev_priv) >=10 )
@@ -4249,12 +4255,7 @@ intel_dp_detect(struct drm_connector *co* supports link training fallback params. */ if (intel_dp->reset_link_params || intel_dp->is_mst) { -/* Initial max link lane count */ - intel_dp->max_link_lane_count =intel_dp_max_common_lane_count (intel_dp); - -/* Initial max link rate */ - intel_dp->max_link_rate =intel_dp_max_common_rate (intel_dp); - +intel_dp_reset_max_link_params (intel_dp); intel_dp->reset_link_params = false; }
@@ -5307,6 +5308,7 @@ intel_dp_init_connector(struct intel_digintel_dp_set_source_rates (intel_dp);intel_dp_set_default_sink_rates (intel_dp);intel_dp_set_common_rates (intel_dp); +intel_dp_reset_max_link_params (intel_dp); intel_dp->reset_link_params = true; intel_dp->pps.pps_pipe =INVALID_PIPE ;
Greg Kroah-Hartman
maintainerPATCH 253/279 drm/i915: Fix type1 DVI DP dual mode adapter heuristic for modern platforms
In reply to Greg Kroah-HartmanFrom: Ville Syrjälä <[email protected]>
commit 1977e8eb40ed53f0cac7db1a78295726f4ac0b24 upstream.
Looks like we never updated intel_bios_is_port_dp_dual_mode() when the VBT port mapping became erratic on modern platforms. This is causing us to look up the wrong child device and thus throwing the heuristic off (ie. we might end looking at a child device for a genuine DP++ port when we were supposed to look at one for a native HDMI port).
Fix it up by not using the outdated port_mapping[] in intel_bios_is_port_dp_dual_mode() and rely on intel_bios_encoder_data_lookup() instead.
--- drivers/gpu/drm/i915/display/intel_bios.c | 85 ++++++++++++++++++++++-------- 1 file changed, 63 insertions(+), 22 deletions(-)
drivers/gpu/drm/i915/display/intel_bios.c +63 -22
@@ -1692,6 +1692,39 @@ static u8 map_ddc_pin(struct drm_i915_prreturn 0 ; } +static u8 dvo_port_type (u8 dvo_port) +{ +switch (dvo_port) { +case DVO_PORT_HDMIA : +case DVO_PORT_HDMIB : +case DVO_PORT_HDMIC : +case DVO_PORT_HDMID : +case DVO_PORT_HDMIE : +case DVO_PORT_HDMIF : +case DVO_PORT_HDMIG : +case DVO_PORT_HDMIH : +case DVO_PORT_HDMII : +return DVO_PORT_HDMIA ; +case DVO_PORT_DPA : +case DVO_PORT_DPB : +case DVO_PORT_DPC : +case DVO_PORT_DPD : +case DVO_PORT_DPE : +case DVO_PORT_DPF : +case DVO_PORT_DPG : +case DVO_PORT_DPH : +case DVO_PORT_DPI : +return DVO_PORT_DPA ; +case DVO_PORT_MIPIA : +case DVO_PORT_MIPIB : +case DVO_PORT_MIPIC : +case DVO_PORT_MIPID : +return DVO_PORT_MIPIA ; +default : +return dvo_port; + } +} +static enum port __dvo_port_to_port (int n_ports,int n_dvo,const int port_mapping[][3 ],u8 dvo_port) {
@@ -2622,35 +2655,17 @@ bool intel_bios_is_port_edp(struct drm_ireturn false; } -static bool child_dev_is_dp_dual_mode (const struct child_device_config *child, -enum port port) +static bool child_dev_is_dp_dual_mode (const struct child_device_config *child) { -static const struct { -u16 dp, hdmi; - } port_mapping[] = { -/* -* Buggy VBTs may declare DP ports as having -* HDMI type dvo_port :( So let's check both. -*/ - [PORT_B ] = {DVO_PORT_DPB ,DVO_PORT_HDMIB , }, - [PORT_C ] = {DVO_PORT_DPC ,DVO_PORT_HDMIC , }, - [PORT_D ] = {DVO_PORT_DPD ,DVO_PORT_HDMID , }, - [PORT_E ] = {DVO_PORT_DPE ,DVO_PORT_HDMIE , }, - [PORT_F ] = {DVO_PORT_DPF ,DVO_PORT_HDMIF , }, - }; - -if (port ==PORT_A || port >=ARRAY_SIZE (port_mapping)) -return false; -if ((child->device_type &DEVICE_TYPE_DP_DUAL_MODE_BITS ) != (DEVICE_TYPE_DP_DUAL_MODE &DEVICE_TYPE_DP_DUAL_MODE_BITS ))return false; -if (child->dvo_port == port_mapping[port].dp) +if (dvo_port_type (child->dvo_port) ==DVO_PORT_DPA )return true;/* Only accept a HDMI dvo_port as DP++ if it has an AUX channel */ -if (child->dvo_port == port_mapping[port].hdmi && +if (dvo_port_type (child->dvo_port) ==DVO_PORT_HDMIA && child->aux_channel !=0 )return true;
@@ -2660,10 +2675,36 @@ static bool child_dev_is_dp_dual_mode(cobool intel_bios_is_port_dp_dual_mode (struct drm_i915_private *i915,enum port port) { +static const struct { +u16 dp, hdmi; + } port_mapping[] = { +/* +* Buggy VBTs may declare DP ports as having +* HDMI type dvo_port :( So let's check both. +*/ + [PORT_B ] = {DVO_PORT_DPB ,DVO_PORT_HDMIB , }, + [PORT_C ] = {DVO_PORT_DPC ,DVO_PORT_HDMIC , }, + [PORT_D ] = {DVO_PORT_DPD ,DVO_PORT_HDMID , }, + [PORT_E ] = {DVO_PORT_DPE ,DVO_PORT_HDMIE , }, + [PORT_F ] = {DVO_PORT_DPF ,DVO_PORT_HDMIF , }, + };const struct intel_bios_encoder_data *devdata; +if (HAS_DDI (i915)) { +const struct intel_bios_encoder_data *devdata; + + devdata =intel_bios_encoder_data_lookup (i915, port); + +return devdata &&child_dev_is_dp_dual_mode (&devdata->child); + } + +if (port ==PORT_A || port >=ARRAY_SIZE (port_mapping)) +return false; +list_for_each_entry (devdata, &i915->vbt.display_devices, node) { -if (child_dev_is_dp_dual_mode (&devdata->child, port)) +if ((devdata->child.dvo_port == port_mapping[port].dp || + devdata->child.dvo_port == port_mapping[port].hdmi) && +child_dev_is_dp_dual_mode (&devdata->child))return true; }
Greg Kroah-Hartman
maintainerPATCH 254/279 drm/amdgpu: fix set scaling mode Full/Full aspect/Center not works on vga and dvi connectors
In reply to Greg Kroah-HartmanFrom: hongao <[email protected]>
commit bf552083916a7f8800477b5986940d1c9a31b953 upstream.
amdgpu_connector_vga_get_modes missed function amdgpu_get_native_mode which assign amdgpu_encoder->native_mode with *preferred_mode result in amdgpu_encoder->native_mode.clock always be 0. That will cause amdgpu_connector_set_property returned early on:
if ((rmx_type != DRM_MODE_SCALE_NONE) &&
(amdgpu_encoder->native_mode.clock == 0))
when we try to set scaling mode Full/Full aspect/Center.
Add the missing function to amdgpu_connector_vga_get_mode can fix this. It also works on dvi connectors because amdgpu_connector_dvi_helper_funcs.get_mode use the same method.
--- drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c | 1 + 1 file changed, 1 insertion(+)
drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c +1 -0
@@ -827,6 +827,7 @@ static int amdgpu_connector_vga_get_modeamdgpu_connector_get_edid (connector); ret =amdgpu_connector_ddc_get_modes (connector); +amdgpu_get_native_mode (connector);return ret; }
Greg Kroah-Hartman
maintainerPATCH 255/279 drm/amd/pm: avoid duplicate powergate/ungate setting
In reply to Greg Kroah-HartmanFrom: Evan Quan <[email protected]>
commit 6ee27ee27ba8b2e725886951ba2d2d87f113bece upstream.
Just bail out if the target IP block is already in the desired powergate/ungate state. This can avoid some duplicate settings which sometimes may cause unexpected issues.
--- drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 3 +++ drivers/gpu/drm/amd/include/amd_shared.h | 3 ++- drivers/gpu/drm/amd/pm/amdgpu_dpm.c | 10 ++++++++++ drivers/gpu/drm/amd/pm/inc/amdgpu_dpm.h | 8 ++++++++ 4 files changed, 23 insertions(+), 1 deletion(-)
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +3 -0
@@ -3532,6 +3532,9 @@ int amdgpu_device_init(struct amdgpu_dev adev->rmmio_size =pci_resource_len (adev->pdev,2 ); } +for (i =0 ; i <AMD_IP_BLOCK_TYPE_NUM ; i++) +atomic_set (&adev->pm.pwr_state[i],POWER_STATE_UNKNOWN ); + adev->rmmio =ioremap (adev->rmmio_base, adev->rmmio_size);if (adev->rmmio ==NULL ) {return -ENOMEM ;
drivers/gpu/drm/amd/include/amd_shared.h +2 -1
@@ -98,7 +98,8 @@ enum amd_ip_block_type {AMD_IP_BLOCK_TYPE_ACP ,AMD_IP_BLOCK_TYPE_VCN ,AMD_IP_BLOCK_TYPE_MES , -AMD_IP_BLOCK_TYPE_JPEG +AMD_IP_BLOCK_TYPE_JPEG , + AMD_IP_BLOCK_TYPE_NUM, }; enum amd_clockgating_state {
drivers/gpu/drm/amd/pm/amdgpu_dpm.c +10 -0
@@ -927,6 +927,13 @@ int amdgpu_dpm_set_powergating_by_smu(st {int ret =0 ;const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs; +enum ip_power_state pwr_state = gate ?POWER_STATE_OFF :POWER_STATE_ON ; + +if (atomic_read (&adev->pm.pwr_state[block_type]) == pwr_state) { +dev_dbg (adev->dev,"IP block%d already in the target %s state!" , + block_type, gate ?"gate" :"ungate" ); +return 0 ; + }switch (block_type) {case AMD_IP_BLOCK_TYPE_UVD :
@@ -979,6 +986,9 @@ int amdgpu_dpm_set_powergating_by_smu(stbreak ; } +if (!ret) +atomic_set (&adev->pm.pwr_state[block_type], pwr_state); +return ret; }
drivers/gpu/drm/amd/pm/inc/amdgpu_dpm.h +8 -0
@@ -417,6 +417,12 @@ struct amdgpu_dpm {enum amd_dpm_forced_level forced_level; }; +enum ip_power_state { +POWER_STATE_UNKNOWN , +POWER_STATE_ON , +POWER_STATE_OFF , +}; +struct amdgpu_pm {struct mutex mutex;u32 current_sclk;
@@ -452,6 +458,8 @@ struct amdgpu_pm {struct i2c_adapter smu_i2c;struct mutex smu_i2c_mutex;struct list_head pm_attr_list; + +atomic_t pwr_state[AMD_IP_BLOCK_TYPE_NUM ]; };#define R600_SSTU_DFLT 0
Greg Kroah-Hartman
maintainerPATCH 256/279 signal: Implement force_fatal_sig
In reply to Greg Kroah-HartmanFrom: Eric W. Biederman <[email protected]>
commit 26d5badbccddcc063dc5174a2baffd13a23322aa upstream.
Add a simple helper force_fatal_sig that causes a signal to be delivered to a process as if the signal handler was set to SIG_DFL.
Reimplement force_sigsegv based upon this new helper. This fixes force_sigsegv so that when it forces the default signal handler to be used the code now forces the signal to be unblocked as well.
Reusing the tested logic in force_sig_info_to_task that was built for force_sig_seccomp this makes the implementation trivial.
This is interesting both because it makes force_sigsegv simpler and because there are a couple of buggy places in the kernel that call do_exit(SIGILL) or do_exit(SIGSYS) because there is no straight forward way today for those places to simply force the exit of a process with the chosen signal. Creating force_fatal_sig allows those places to be implemented with normal signal exits.
--- include/linux/sched/signal.h | 1 + kernel/signal.c | 26 +++++++++++++++++--------- 2 files changed, 18 insertions(+), 9 deletions(-)
include/linux/sched/signal.h +1 -0
@@ -338,6 +338,7 @@ extern int kill_pid(struct pid *pid, intextern __must_check bool do_notify_parent (struct task_struct *,int );extern void __wake_up_parent (struct task_struct *p,struct task_struct *parent);extern void force_sig (int ); +extern void force_fatal_sig (int );extern int send_sig (int ,struct task_struct *,int );extern int zap_other_threads (struct task_struct *p);extern struct sigqueue *sigqueue_alloc (void );
kernel/signal.c +17 -9
@@ -1650,6 +1650,19 @@ void force_sig(int sig) }EXPORT_SYMBOL (force_sig); +void force_fatal_sig (int sig) +{ +struct kernel_siginfo info; + +clear_siginfo (&info); + info.si_signo = sig; + info.si_errno =0 ; + info.si_code =SI_KERNEL ; + info.si_pid =0 ; + info.si_uid =0 ; +force_sig_info_to_task (&info, current, true); +} +/* * When things go south during signal handling, we * will force a SIGSEGV. And if the signal that caused
@@ -1658,15 +1671,10 @@ EXPORT_SYMBOL(force_sig);*/ void force_sigsegv (int sig) { -struct task_struct *p = current; - -if (sig ==SIGSEGV ) { -unsigned long flags; -spin_lock_irqsave (&p->sighand->siglock, flags); - p->sighand->action[sig -1 ].sa.sa_handler =SIG_DFL ; -spin_unlock_irqrestore (&p->sighand->siglock, flags); - } -force_sig (SIGSEGV ); +if (sig ==SIGSEGV ) +force_fatal_sig (SIGSEGV ); +else +force_sig (SIGSEGV ); } int force_sig_fault_to_task(int sig, int code, void __user *addr
Greg Kroah-Hartman
maintainerPATCH 257/279 exit/syscall_user_dispatch: Send ordinary signals on failure
In reply to Greg Kroah-HartmanFrom: Eric W. Biederman <[email protected]>
commit 941edc5bf174b67f94db19817cbeab0a93e0c32a upstream.
Use force_fatal_sig instead of calling do_exit directly. This ensures the ordinary signal handling path gets invoked, core dumps as appropriate get created, and for multi-threaded processes all of the threads are terminated not just a single thread.
When asked Gabriel Krisman Bertazi <[email protected]> said [1]:
Quoted 14 lines
[email protected] (Eric W. Biederman) asked:
Quoted 6 lines
Why does
do_syscal_user_dispatchcalldo_exit(SIGSEGV)anddo_exit(SIGSYS)instead offorce_sig(SIGSEGV)andforce_sig(SIGSYS)?Looking at the code these cases are not expected to happen, so I would be surprised if userspace depends on any particular behaviour on the failure path so I think we can change this.
Hi Eric,
There is not really a good reason, and the use case that originated the feature doesn't rely on it.
Quoted 7 lines
Unless I'm missing yet another problem and others correct me, I think it makes sense to change it as you described.
Quoted 1 line
Is using
do_exitin this way something you copied from seccomp?I'm not sure, its been a while, but I think it might be just that. The first prototype of SUD was implemented as a seccomp mode.
If at some point it becomes interesting we could relax "force_fatal_sig(SIGSEGV)" to instead say "force_sig_fault(SIGSEGV, SEGV_MAPERR, sd->selector)".
I avoid doing that in this patch to avoid making it possible to catch currently uncatchable signals.
Cc: Gabriel Krisman Bertazi <[email protected]>
Cc: Thomas Gleixner <[email protected]>
Cc: Peter Zijlstra <[email protected]>
Cc: Andy Lutomirski <[email protected]>
[1] https://lkml.kernel.org/r/[email protected]
--- kernel/entry/syscall_user_dispatch.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-)
kernel/entry/syscall_user_dispatch.c +8 -4
@@ -47,14 +47,18 @@ bool syscall_user_dispatch(struct pt_reg* access_ok() is performed once, at prctl time, when * the selector is loaded by userspace. */ -if (unlikely (__get_user (state, sd->selector))) -do_exit (SIGSEGV ); +if (unlikely (__get_user (state, sd->selector))) { +force_fatal_sig (SIGSEGV ); +return true; + }if (likely (state ==SYSCALL_DISPATCH_FILTER_ALLOW ))return false; -if (state !=SYSCALL_DISPATCH_FILTER_BLOCK ) -do_exit (SIGSYS ); +if (state !=SYSCALL_DISPATCH_FILTER_BLOCK ) { +force_fatal_sig (SIGSYS ); +return true; + } } sd->on_dispatch = true;
Greg Kroah-Hartman
maintainerPATCH 258/279 signal/powerpc: On swapcontext failure force SIGSEGV
In reply to Greg Kroah-HartmanFrom: Eric W. Biederman <[email protected]>
commit 83a1f27ad773b1d8f0460d3a676114c7651918cc upstream.
If the register state may be partial and corrupted instead of calling do_exit, call force_sigsegv(SIGSEGV). Which properly kills the process with SIGSEGV and does not let any more userspace code execute, instead of just killing one thread of the process and potentially confusing everything.
Cc: Michael Ellerman <[email protected]>
Cc: Benjamin Herrenschmidt <[email protected]>
Cc: Paul Mackerras <[email protected]>
Cc: [email protected]
History-tree: git://git.kernel.org/pub/scm/linux/kernel/git/tglx/history.git
--- arch/powerpc/kernel/signal_32.c | 6 ++++-- arch/powerpc/kernel/signal_64.c | 9 ++++++--- 2 files changed, 10 insertions(+), 5 deletions(-)
arch/powerpc/kernel/signal_32.c +4 -2
@@ -1062,8 +1062,10 @@ SYSCALL_DEFINE3(swapcontext, struct ucon* or if another thread unmaps the region containing the context. * We kill the task with a SIGSEGV in this situation. */ -if (do_setcontext (new_ctx, regs,0 )) -do_exit (SIGSEGV ); +if (do_setcontext (new_ctx, regs,0 )) { +force_sigsegv (SIGSEGV ); +return -EFAULT ; + }set_thread_flag (TIF_RESTOREALL );return 0 ;
arch/powerpc/kernel/signal_64.c +6 -3
@@ -703,15 +703,18 @@ SYSCALL_DEFINE3(swapcontext, struct ucon* We kill the task with a SIGSEGV in this situation. */ -if (__get_user_sigset (&set, &new_ctx->uc_sigmask)) -do_exit (SIGSEGV ); +if (__get_user_sigset (&set, &new_ctx->uc_sigmask)) { +force_sigsegv (SIGSEGV ); +return -EFAULT ; + }set_current_blocked (&set);if (!user_read_access_begin (new_ctx, ctx_size))return -EFAULT ;if (__unsafe_restore_sigcontext (current,NULL ,0 , &new_ctx->uc_mcontext)) {user_read_access_end (); -do_exit (SIGSEGV ); +force_sigsegv (SIGSEGV ); +return -EFAULT ; }user_read_access_end ();
Greg Kroah-Hartman
maintainerPATCH 259/279 signal/s390: Use force_sigsegv in default_trap_handler
In reply to Greg Kroah-HartmanFrom: Eric W. Biederman <[email protected]>
commit 9bc508cf0791c8e5a37696de1a046d746fcbd9d8 upstream.
Reading the history it is unclear why default_trap_handler calls do_exit. It is not even menthioned in the commit where the change happened. My best guess is that because it is unknown why the exception happened it was desired to guarantee the process never returned to userspace.
Using do_exit(SIGSEGV) has the problem that it will only terminate one thread of a process, leaving the process in an undefined state.
Use force_sigsegv(SIGSEGV) instead which effectively has the same behavior except that is uses the ordinary signal mechanism and terminates all threads of a process and is generally well defined.
History Tree: https://git.kernel.org/pub/scm/linux/kernel/git/tglx/history.git
--- arch/s390/kernel/traps.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
arch/s390/kernel/traps.c +1 -1
@@ -84,7 +84,7 @@ static void default_trap_handler(struct {if (user_mode (regs)) {report_user_fault (regs,SIGSEGV ,0 ); -do_exit (SIGSEGV ); +force_sigsegv (SIGSEGV ); }else die (regs,"Unknown program exception" ); }
Greg Kroah-Hartman
maintainerPATCH 260/279 signal/sparc32: Exit with a fatal signal when try_to_clear_window_buffer fails
In reply to Greg Kroah-HartmanFrom: Eric W. Biederman <[email protected]>
commit c317d306d55079525c9610267fdaf3a8a6d2f08b upstream.
The function try_to_clear_window_buffer is only called from rtrap_32.c. After it is called the signal pending state is retested, and signals are handled if TIF_SIGPENDING is set. This allows try_to_clear_window_buffer to call force_fatal_signal and then rely on the signal being delivered to kill the process, without any danger of returning to userspace, or otherwise using possible corrupt state on failure.
The functional difference between force_fatal_sig and do_exit is that do_exit will only terminate a single thread, and will never trigger a core-dump. A multi-threaded program for which a single thread terminates unexpectedly is hard to reason about. Calling force_fatal_sig does not give userspace a chance to catch the signal, but otherwise is an ordinary fatal signal exit, and it will trigger a coredump of the offending process if core dumps are enabled.
--- arch/sparc/kernel/windows.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)
arch/sparc/kernel/windows.c +4 -2
@@ -121,8 +121,10 @@ void try_to_clear_window_buffer(struct pif ((sp &7 ) ||copy_to_user ((char __user *) sp, &tp->reg_window[window], -sizeof (struct reg_window32 ))) -do_exit (SIGILL ); +sizeof (struct reg_window32 ))) { +force_fatal_sig (SIGILL ); +return ; + } } tp->w_saved =0 ; }
Greg Kroah-Hartman
maintainerPATCH 261/279 signal/sparc32: In setup_rt_frame and setup_fram use force_fatal_sig
In reply to Greg Kroah-HartmanFrom: Eric W. Biederman <[email protected]>
commit 086ec444f86660e103de8945d0dcae9b67132ac9 upstream.
Modify the 32bit version of setup_rt_frame and setup_frame to act similar to the 64bit version of setup_rt_frame and fail with a signal instead of calling do_exit.
Replacing do_exit(SIGILL) with force_fatal_signal(SIGILL) ensures that the process will be terminated cleanly when the stack frame is invalid, instead of just killing off a single thread and leaving the process is a weird state.
--- arch/sparc/kernel/signal_32.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
arch/sparc/kernel/signal_32.c +2 -2
@@ -244,7 +244,7 @@ static int setup_frame(struct ksignal *kget_sigframe (ksig, regs, sigframe_size);if (invalid_frame_pointer (sf, sigframe_size)) { -do_exit (SIGILL ); +force_fatal_sig (SIGILL );return -EINVAL ; }
@@ -336,7 +336,7 @@ static int setup_rt_frame(struct ksignal sf = (struct rt_signal_frame __user *)get_sigframe (ksig, regs, sigframe_size);if (invalid_frame_pointer (sf, sigframe_size)) { -do_exit (SIGILL ); +force_fatal_sig (SIGILL );return -EINVAL ; }
Greg Kroah-Hartman
maintainerPATCH 262/279 signal/vm86_32: Properly send SIGSEGV when the vm86 state cannot be saved.
In reply to Greg Kroah-HartmanFrom: Eric W. Biederman <[email protected]>
commit 1fbd60df8a852d9c55de8cd3621899cf4c72a5b7 upstream.
Update save_v86_state to always complete all of it's work except possibly some of the copies to userspace even if save_v86_state takes a fault. This ensures that the kernel is always in a sane state, even if userspace has done something silly.
When save_v86_state takes a fault update it to force userspace to take a SIGSEGV and terminate the userspace application.
As Andy pointed out in review of the first version of this change there are races between sigaction and the application terinating. Now that the code has been modified to always perform all save_v86_state's work (except possibly copying to userspace) those races do not matter from a kernel perspective.
Forcing the userspace application to terminate (by resetting it's handler to SIGDFL) is there to keep everything as close to the current behavior as possible while removing the unique (and difficult to maintain) use of do_exit.
If this new SIGSEGV happens during handle_signal the next time around the exit_to_user_mode_loop, SIGSEGV will be delivered to userspace.
All of the callers of handle_vm86_trap and handle_vm86_fault run the exit_to_user_mode_loop before they return to userspace any signal sent to the current task during their execution will be delivered to the current task before that tasks exits to usermode.
Cc: Andy Lutomirski <[email protected]>
Cc: Thomas Gleixner <[email protected]>
Cc: Ingo Molnar <[email protected]>
Cc: Borislav Petkov <[email protected]>
Cc: [email protected]
Cc: H Peter Anvin <[email protected]>
v1: https://lkml.kernel.org/r/[email protected]
--- arch/x86/kernel/vm86_32.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
arch/x86/kernel/vm86_32.c +3 -1
@@ -142,6 +142,7 @@ void save_v86_state(struct kernel_vm86_ruser_access_end (); +exit_vm86:preempt_disable (); tsk->thread.sp0 = vm86->saved_sp0; tsk->thread.sysenter_cs = __KERNEL_CS;
@@ -161,7 +162,8 @@ Efault_end:user_access_end (); Efault:pr_alert ("could not access userspace vm86 info\n" ); -do_exit (SIGSEGV ); +force_sigsegv (SIGSEGV ); + goto exit_vm86; }static int do_vm86_irq_handling (int subfunction,int irqnumber);
Greg Kroah-Hartman
maintainerPATCH 263/279 signal/x86: In emulate_vsyscall force a signal instead of calling do_exit
In reply to Greg Kroah-HartmanFrom: Eric W. Biederman <[email protected]>
commit 695dd0d634df8903e5ead8aa08d326f63b23368a upstream.
Directly calling do_exit with a signal number has the problem that all of the side effects of the signal don't happen, such as killing all of the threads of a process instead of just the calling thread.
So replace do_exit(SIGSYS) with force_fatal_sig(SIGSYS) which causes the signal handling to take it's normal path and work as expected.
--- arch/x86/entry/vsyscall/vsyscall_64.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
arch/x86/entry/vsyscall/vsyscall_64.c +2 -1
@@ -226,7 +226,8 @@ bool emulate_vsyscall(unsigned long erroif ((!tmp && regs->orig_ax != syscall_nr) || regs->ip != address) {warn_bad_vsyscall (KERN_DEBUG , regs,"seccomp tried to change syscall nr or ip" ); -do_exit (SIGSYS ); +force_fatal_sig (SIGSYS ); +return true; } regs->orig_ax =-1 ;if (tmp)
Greg Kroah-Hartman
maintainerPATCH 264/279 signal: Replace force_sigsegv(SIGSEGV) with force_fatal_sig(SIGSEGV)
In reply to Greg Kroah-HartmanFrom: Eric W. Biederman <[email protected]>
commit e21294a7aaae32c5d7154b187113a04db5852e37 upstream.
Now that force_fatal_sig exists it is unnecessary and a bit confusing to use force_sigsegv in cases where the simpler force_fatal_sig is wanted. So change every instance we can to make the code clearer.
--- arch/arc/kernel/process.c | 2 +- arch/m68k/kernel/traps.c | 2 +- arch/powerpc/kernel/signal_32.c | 2 +- arch/powerpc/kernel/signal_64.c | 4 ++-- arch/s390/kernel/traps.c | 2 +- arch/um/kernel/trap.c | 2 +- arch/x86/kernel/vm86_32.c | 2 +- fs/exec.c | 2 +- 8 files changed, 9 insertions(+), 9 deletions(-)
arch/arc/kernel/process.c +1 -1
@@ -294,7 +294,7 @@ int elf_check_arch(const struct elf32_hd eflags = x->e_flags;if ((eflags &EF_ARC_OSABI_MSK ) !=EF_ARC_OSABI_CURRENT ) {pr_err ("ABI mismatch - you need newer toolchain\n" ); -force_sigsegv (SIGSEGV ); +force_fatal_sig (SIGSEGV );return 0 ; }
arch/m68k/kernel/traps.c +1 -1
@@ -1145,7 +1145,7 @@ asmlinkage void set_esp0(unsigned long s*/ asmlinkage void fpsp040_die(void) { -force_sigsegv (SIGSEGV ); +force_fatal_sig (SIGSEGV ); }#ifdef CONFIG_M68KFPU_EMU
arch/powerpc/kernel/signal_32.c +1 -1
@@ -1063,7 +1063,7 @@ SYSCALL_DEFINE3(swapcontext, struct ucon* We kill the task with a SIGSEGV in this situation. */ if (do_setcontext (new_ctx, regs,0 )) { -force_sigsegv (SIGSEGV ); +force_fatal_sig (SIGSEGV );return -EFAULT ; }
arch/powerpc/kernel/signal_64.c +2 -2
@@ -704,7 +704,7 @@ SYSCALL_DEFINE3(swapcontext, struct ucon*/ if (__get_user_sigset (&set, &new_ctx->uc_sigmask)) { -force_sigsegv (SIGSEGV ); +force_fatal_sig (SIGSEGV );return -EFAULT ; }set_current_blocked (&set);
@@ -713,7 +713,7 @@ SYSCALL_DEFINE3(swapcontext, struct uconreturn -EFAULT ;if (__unsafe_restore_sigcontext (current,NULL ,0 , &new_ctx->uc_mcontext)) {user_read_access_end (); -force_sigsegv (SIGSEGV ); +force_fatal_sig (SIGSEGV );return -EFAULT ; }user_read_access_end ();
arch/s390/kernel/traps.c +1 -1
@@ -84,7 +84,7 @@ static void default_trap_handler(struct {if (user_mode (regs)) {report_user_fault (regs,SIGSEGV ,0 ); -force_sigsegv (SIGSEGV ); +force_fatal_sig (SIGSEGV ); }else die (regs,"Unknown program exception" ); }
arch/um/kernel/trap.c +1 -1
@@ -158,7 +158,7 @@ static void bad_segv(struct faultinfo fivoid fatal_sigsegv (void ) { -force_sigsegv (SIGSEGV ); +force_fatal_sig (SIGSEGV );do_signal (¤t->thread.regs);/* * This is to tell gcc that we're not returning - do_signal
arch/x86/kernel/vm86_32.c +1 -1
@@ -162,7 +162,7 @@ Efault_end:user_access_end (); Efault:pr_alert ("could not access userspace vm86 info\n" ); -force_sigsegv (SIGSEGV ); +force_fatal_sig (SIGSEGV ); goto exit_vm86; }
fs/exec.c +1 -1
@@ -1852,7 +1852,7 @@ out:* SIGSEGV. */ if (bprm->point_of_no_return && !fatal_signal_pending (current)) -force_sigsegv (SIGSEGV ); +force_fatal_sig (SIGSEGV ); out_unmark: current->fs->in_exec =0 ;
Greg Kroah-Hartman
maintainerPATCH 265/279 signal: Dont always set SA_IMMUTABLE for forced signals
In reply to Greg Kroah-HartmanFrom: Eric W. Biederman <[email protected]>
commit e349d945fac76bddc78ae1cb92a0145b427a87ce upstream.
Recently to prevent issues with SECCOMP_RET_KILL and similar signals being changed before they are delivered SA_IMMUTABLE was added.
Unfortunately this broke debuggers[1][2] which reasonably expect to be able to trap synchronous SIGTRAP and SIGSEGV even when the target process is not configured to handle those signals.
Update force_sig_to_task to support both the case when we can allow the debugger to intercept and possibly ignore the signal and the case when it is not safe to let userspace know about the signal until the process has exited.
[1] https://lkml.kernel.org/r/CAP045AoMY4xf8aC_4QU_-j7obuEPYgTcnQQP3Yxk=2X90jtpjw@mail.gmail.com
[2] https://lkml.kernel.org/r/20211117150258.GB5403@xsang-OptiPlex-9020
--- kernel/signal.c | 23 ++++++++++++++++------- 1 file changed, 16 insertions(+), 7 deletions(-)
kernel/signal.c +16 -7
@@ -1298,6 +1298,12 @@ int do_send_sig_info(int sig, struct kerreturn ret; } +enum sig_handler { +HANDLER_CURRENT ,/* If reachable use the current handler */ +HANDLER_SIG_DFL ,/* Always use SIG_DFL handler semantics */ +HANDLER_EXIT ,/* Only visible as the process exit code */ +}; +/* * Force a signal that the process can't ignore: if necessary * we unblock the signal and change any SIG_IGN to SIG_DFL.
@@ -1310,7 +1316,8 @@ int do_send_sig_info(int sig, struct ker* that is why we also clear SIGNAL_UNKILLABLE. */ static int -force_sig_info_to_task (struct kernel_siginfo *info,struct task_struct *t,bool sigdfl) +force_sig_info_to_task (struct kernel_siginfo *info,struct task_struct *t, +enum sig_handler handler) {unsigned long int flags;int ret, blocked, ignored;
@@ -1321,9 +1328,10 @@ force_sig_info_to_task(struct kernel_sig action = &t->sighand->action[sig-1 ]; ignored = action->sa.sa_handler ==SIG_IGN ; blocked =sigismember (&t->blocked, sig); -if (blocked || ignored || sigdfl) { +if (blocked || ignored || (handler !=HANDLER_CURRENT )) { action->sa.sa_handler =SIG_DFL ; - action->sa.sa_flags |=SA_IMMUTABLE ; +if (handler ==HANDLER_EXIT ) + action->sa.sa_flags |=SA_IMMUTABLE ;if (blocked) {sigdelset (&t->blocked, sig);recalc_sigpending_and_wake (t);
@@ -1343,7 +1351,7 @@ force_sig_info_to_task(struct kernel_sigint force_sig_info (struct kernel_siginfo *info) { -return force_sig_info_to_task (info, current, false); +return force_sig_info_to_task (info, current,HANDLER_CURRENT ); }/*
@@ -1660,7 +1668,7 @@ void force_fatal_sig(int sig) info.si_code =SI_KERNEL ; info.si_pid =0 ; info.si_uid =0 ; -force_sig_info_to_task (&info, current, true); +force_sig_info_to_task (&info, current,HANDLER_SIG_DFL ); }/*
@@ -1693,7 +1701,7 @@ int force_sig_fault_to_task(int sig, int info.si_flags = flags; info.si_isr = isr;#endif -return force_sig_info_to_task (&info, t, false); +return force_sig_info_to_task (&info, t,HANDLER_CURRENT ); } int force_sig_fault(int sig, int code, void __user *addr
@@ -1813,7 +1821,8 @@ int force_sig_seccomp(int syscall, int r info.si_errno = reason; info.si_arch =syscall_get_arch (current); info.si_syscall = syscall; -return force_sig_info_to_task (&info, current, force_coredump); +return force_sig_info_to_task (&info, current, + force_coredump ?HANDLER_EXIT :HANDLER_CURRENT ); }/* For the crazy architectures that include trap information in
Greg Kroah-Hartman
maintainerPATCH 266/279 signal: Replace force_fatal_sig with force_exit_sig when in doubt
In reply to Greg Kroah-HartmanFrom: Eric W. Biederman <[email protected]>
commit fcb116bc43c8c37c052530ead79872f8b2615711 upstream.
Recently to prevent issues with SECCOMP_RET_KILL and similar signals being changed before they are delivered SA_IMMUTABLE was added.
Unfortunately this broke debuggers[1][2] which reasonably expect to be able to trap synchronous SIGTRAP and SIGSEGV even when the target process is not configured to handle those signals.
Add force_exit_sig and use it instead of force_fatal_sig where historically the code has directly called do_exit. This has the implementation benefits of going through the signal exit path (including generating core dumps) without the danger of allowing userspace to ignore or change these signals.
This avoids userspace regressions as older kernels exited with do_exit which debuggers also can not intercept.
In the future is should be possible to improve the quality of implementation of the kernel by changing some of these force_exit_sig calls to force_fatal_sig. That can be done where it matters on a case-by-case basis with careful analysis.
[1] https://lkml.kernel.org/r/CAP045AoMY4xf8aC_4QU_-j7obuEPYgTcnQQP3Yxk=2X90jtpjw@mail.gmail.com
[2] https://lkml.kernel.org/r/20211117150258.GB5403@xsang-OptiPlex-9020
--- arch/m68k/kernel/traps.c | 2 +- arch/powerpc/kernel/signal_32.c | 2 +- arch/powerpc/kernel/signal_64.c | 4 ++-- arch/s390/kernel/traps.c | 2 +- arch/sparc/kernel/signal_32.c | 4 ++-- arch/sparc/kernel/windows.c | 2 +- arch/x86/entry/vsyscall/vsyscall_64.c | 2 +- arch/x86/kernel/vm86_32.c | 2 +- include/linux/sched/signal.h | 1 + kernel/entry/syscall_user_dispatch.c | 4 ++-- kernel/signal.c | 13 +++++++++++++ 11 files changed, 26 insertions(+), 12 deletions(-)
arch/m68k/kernel/traps.c +1 -1
@@ -1145,7 +1145,7 @@ asmlinkage void set_esp0(unsigned long s*/ asmlinkage void fpsp040_die(void) { -force_fatal_sig (SIGSEGV ); +force_exit_sig (SIGSEGV ); }#ifdef CONFIG_M68KFPU_EMU
arch/powerpc/kernel/signal_32.c +1 -1
@@ -1063,7 +1063,7 @@ SYSCALL_DEFINE3(swapcontext, struct ucon* We kill the task with a SIGSEGV in this situation. */ if (do_setcontext (new_ctx, regs,0 )) { -force_fatal_sig (SIGSEGV ); +force_exit_sig (SIGSEGV );return -EFAULT ; }
arch/powerpc/kernel/signal_64.c +2 -2
@@ -704,7 +704,7 @@ SYSCALL_DEFINE3(swapcontext, struct ucon*/ if (__get_user_sigset (&set, &new_ctx->uc_sigmask)) { -force_fatal_sig (SIGSEGV ); +force_exit_sig (SIGSEGV );return -EFAULT ; }set_current_blocked (&set);
@@ -713,7 +713,7 @@ SYSCALL_DEFINE3(swapcontext, struct uconreturn -EFAULT ;if (__unsafe_restore_sigcontext (current,NULL ,0 , &new_ctx->uc_mcontext)) {user_read_access_end (); -force_fatal_sig (SIGSEGV ); +force_exit_sig (SIGSEGV );return -EFAULT ; }user_read_access_end ();
arch/s390/kernel/traps.c +1 -1
@@ -84,7 +84,7 @@ static void default_trap_handler(struct {if (user_mode (regs)) {report_user_fault (regs,SIGSEGV ,0 ); -force_fatal_sig (SIGSEGV ); +force_exit_sig (SIGSEGV ); }else die (regs,"Unknown program exception" ); }
arch/sparc/kernel/signal_32.c +2 -2
@@ -244,7 +244,7 @@ static int setup_frame(struct ksignal *kget_sigframe (ksig, regs, sigframe_size);if (invalid_frame_pointer (sf, sigframe_size)) { -force_fatal_sig (SIGILL ); +force_exit_sig (SIGILL );return -EINVAL ; }
@@ -336,7 +336,7 @@ static int setup_rt_frame(struct ksignal sf = (struct rt_signal_frame __user *)get_sigframe (ksig, regs, sigframe_size);if (invalid_frame_pointer (sf, sigframe_size)) { -force_fatal_sig (SIGILL ); +force_exit_sig (SIGILL );return -EINVAL ; }
arch/sparc/kernel/windows.c +1 -1
@@ -122,7 +122,7 @@ void try_to_clear_window_buffer(struct pif ((sp &7 ) ||copy_to_user ((char __user *) sp, &tp->reg_window[window],sizeof (struct reg_window32 ))) { -force_fatal_sig (SIGILL ); +force_exit_sig (SIGILL );return ; } }
arch/x86/entry/vsyscall/vsyscall_64.c +1 -1
@@ -226,7 +226,7 @@ bool emulate_vsyscall(unsigned long erroif ((!tmp && regs->orig_ax != syscall_nr) || regs->ip != address) {warn_bad_vsyscall (KERN_DEBUG , regs,"seccomp tried to change syscall nr or ip" ); -force_fatal_sig (SIGSYS ); +force_exit_sig (SIGSYS );return true; } regs->orig_ax =-1 ;
arch/x86/kernel/vm86_32.c +1 -1
@@ -162,7 +162,7 @@ Efault_end:user_access_end (); Efault:pr_alert ("could not access userspace vm86 info\n" ); -force_fatal_sig (SIGSEGV ); +force_exit_sig (SIGSEGV ); goto exit_vm86; }
include/linux/sched/signal.h +1 -0
@@ -339,6 +339,7 @@ extern __must_check bool do_notify_parenextern void __wake_up_parent (struct task_struct *p,struct task_struct *parent);extern void force_sig (int );extern void force_fatal_sig (int ); +extern void force_exit_sig (int );extern int send_sig (int ,struct task_struct *,int );extern int zap_other_threads (struct task_struct *p);extern struct sigqueue *sigqueue_alloc (void );
kernel/entry/syscall_user_dispatch.c +2 -2
@@ -48,7 +48,7 @@ bool syscall_user_dispatch(struct pt_reg* the selector is loaded by userspace. */ if (unlikely (__get_user (state, sd->selector))) { -force_fatal_sig (SIGSEGV ); +force_exit_sig (SIGSEGV );return true; }
@@ -56,7 +56,7 @@ bool syscall_user_dispatch(struct pt_regreturn false;if (state !=SYSCALL_DISPATCH_FILTER_BLOCK ) { -force_fatal_sig (SIGSYS ); +force_exit_sig (SIGSYS );return true; } }
kernel/signal.c +13 -0
@@ -1671,6 +1671,19 @@ void force_fatal_sig(int sig)force_sig_info_to_task (&info, current,HANDLER_SIG_DFL ); } +void force_exit_sig (int sig) +{ +struct kernel_siginfo info; + +clear_siginfo (&info); + info.si_signo = sig; + info.si_errno =0 ; + info.si_code =SI_KERNEL ; + info.si_pid =0 ; + info.si_uid =0 ; +force_sig_info_to_task (&info, current,HANDLER_EXIT ); +} +/* * When things go south during signal handling, we * will force a SIGSEGV. And if the signal that caused
Greg Kroah-Hartman
maintainerPATCH 267/279 hugetlbfs: flush TLBs correctly after huge_pmd_unshare
In reply to Greg Kroah-HartmanFrom: Nadav Amit <[email protected]>
commit a4a118f2eead1d6c49e00765de89878288d4b890 upstream.
When __unmap_hugepage_range() calls to huge_pmd_unshare() succeed, a TLB flush is missing. This TLB flush must be performed before releasing the i_mmap_rwsem, in order to prevent an unshared PMDs page from being released and reused before the TLB flush took place.
Arguably, a comprehensive solution would use mmu_gather interface to batch the TLB flushes and the PMDs page release, however it is not an easy solution: (1) try_to_unmap_one() and try_to_migrate_one() also call huge_pmd_unshare() and they cannot use the mmu_gather interface; and (2) deferring the release of the page reference for the PMDs page until after i_mmap_rwsem is dropeed can confuse huge_pmd_unshare() into thinking PMDs are shared when they are not.
Fix __unmap_hugepage_range() by adding the missing TLB flush, and forcing a flush when unshare is successful.
--- mm/hugetlb.c | 23 +++++++++++++++++++---- 1 file changed, 19 insertions(+), 4 deletions(-)
mm/hugetlb.c +19 -4
@@ -4439,6 +4439,7 @@ void __unmap_hugepage_range(struct mmu_gstruct hstate *h =hstate_vma (vma);unsigned long sz =huge_page_size (h);struct mmu_notifier_range range; +bool force_flush = false;WARN_ON (!is_vm_hugetlb_page (vma));BUG_ON (start & ~huge_page_mask (h));
@@ -4467,10 +4468,8 @@ void __unmap_hugepage_range(struct mmu_g ptl =huge_pte_lock (h, mm, ptep);if (huge_pmd_unshare (mm, vma, &address, ptep)) {spin_unlock (ptl); -/* -* We just unmapped a page of PMDs by clearing a PUD. -* The caller's TLB flush range should cover this area. -*/ +tlb_flush_pmd_range (tlb, address &PUD_MASK ,PUD_SIZE ); + force_flush = true;continue ; }
@@ -4527,6 +4526,22 @@ void __unmap_hugepage_range(struct mmu_g }mmu_notifier_invalidate_range_end (&range);tlb_end_vma (tlb, vma); + +/* +* If we unshared PMDs, the TLB flush was not recorded in mmu_gather. We +* could defer the flush until now, since by holding i_mmap_rwsem we +* guaranteed that the last refernece would not be dropped. But we must +* do the flushing before we return, as otherwise i_mmap_rwsem will be +* dropped and the last reference to the shared PMDs page might be +* dropped as well. +* +* In theory we could defer the freeing of the PMD pages as well, but +* huge_pmd_unshare() relies on the exact page_count for the PMD page to +* detect sharing, so we cannot defer the release of the page either. +* Instead, do flush now. +*/ +if (force_flush) +tlb_flush_mmu_tlbonly (tlb); } void __unmap_hugepage_range_final(struct mmu_gather *tlb,
Greg Kroah-Hartman
maintainerPATCH 268/279 RDMA/netlink: Add __maybe_unused to static inline in C file
In reply to Greg Kroah-HartmanFrom: Leon Romanovsky <[email protected]>
commit 83dde7498fefeb920b1def317421262317d178e5 upstream.
Like other commits in the tree add __maybe_unused to a static inline in a C file because some clang compilers will complain about unused code:
Quoted 1 line
Quoted 1 line
drivers/infiniband/core/nldev.c:2543:1: warning: unused function '__chk_RDMA_NL_NLDEV'
MODULE_ALIAS_RDMA_NETLINK(RDMA_NL_NLDEV, 5); ^
--- include/rdma/rdma_netlink.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
include/rdma/rdma_netlink.h +1 -1
@@ -30,7 +30,7 @@ enum rdma_nl_flags {* constant as well and the compiler checks they are the same. */ #define MODULE_ALIAS_RDMA_NETLINK (_index, _val) \ - static inline void __chk_##_index(void) \ + static inline void __maybe_unused __chk_##_index(void) \ { \ BUILD_BUG_ON(_index != _val); \ } \
Greg Kroah-Hartman
maintainerPATCH 269/279 bpf: Forbid bpf_ktime_get_coarse_ns and bpf_timer_* in tracing progs
In reply to Greg Kroah-HartmanFrom: Dmitrii Banshchikov <[email protected]>
commit 5e0bc3082e2e403ac0753e099c2b01446bb35578 upstream.
Use of bpf_ktime_get_coarse_ns() and bpf_timer_* helpers in tracing progs may result in locking issues.
bpf_ktime_get_coarse_ns() uses ktime_get_coarse_ns() time accessor that
isn't safe for any context:
======================================================
WARNING: possible circular locking dependency detected
5.15.0-syzkaller #0 Not tainted
------------------------------------------------------
syz-executor.4/14877 is trying to acquire lock:ffffffff8cb30008 (tk_core.seq.seqcount){----}-{0:0}, at: ktime_get_coarse_ts64+0x25/0x110 kernel/time/timekeeping.c:2255
but task is already holding lock:ffffffff90dbf200 (&obj_hash[i].lock){-.-.}-{2:2}, at: debug_object_deactivate+0x61/0x400 lib/debugobjects.c:735
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
-> #1 (&obj_hash[i].lock){-.-.}-{2:2}:
lock_acquire+0x19f/0x4d0 kernel/locking/lockdep.c:5625
__raw_spin_lock_irqsave include/linux/spinlock_api_smp.h:110 [inline]
_raw_spin_lock_irqsave+0xd1/0x120 kernel/locking/spinlock.c:162
__debug_object_init+0xd9/0x1860 lib/debugobjects.c:569
debug_hrtimer_init kernel/time/hrtimer.c:414 [inline]
debug_init kernel/time/hrtimer.c:468 [inline]
hrtimer_init+0x20/0x40 kernel/time/hrtimer.c:1592
ntp_init_cmos_sync kernel/time/ntp.c:676 [inline]
ntp_init+0xa1/0xad kernel/time/ntp.c:1095
timekeeping_init+0x512/0x6bf kernel/time/timekeeping.c:1639
start_kernel+0x267/0x56e init/main.c:1030
secondary_startup_64_no_verify+0xb1/0xbb-> #0 (tk_core.seq.seqcount){----}-{0:0}:
check_prev_add kernel/locking/lockdep.c:3051 [inline]
check_prevs_add kernel/locking/lockdep.c:3174 [inline]
validate_chain+0x1dfb/0x8240 kernel/locking/lockdep.c:3789
__lock_acquire+0x1382/0x2b00 kernel/locking/lockdep.c:5015
lock_acquire+0x19f/0x4d0 kernel/locking/lockdep.c:5625
seqcount_lockdep_reader_access+0xfe/0x230 include/linux/seqlock.h:103
ktime_get_coarse_ts64+0x25/0x110 kernel/time/timekeeping.c:2255
ktime_get_coarse include/linux/timekeeping.h:120 [inline]
ktime_get_coarse_ns include/linux/timekeeping.h:126 [inline]
____bpf_ktime_get_coarse_ns kernel/bpf/helpers.c:173 [inline]
bpf_ktime_get_coarse_ns+0x7e/0x130 kernel/bpf/helpers.c:171
bpf_prog_a99735ebafdda2f1+0x10/0xb50
bpf_dispatcher_nop_func include/linux/bpf.h:721 [inline]
__bpf_prog_run include/linux/filter.h:626 [inline]
bpf_prog_run include/linux/filter.h:633 [inline]
BPF_PROG_RUN_ARRAY include/linux/bpf.h:1294 [inline]
trace_call_bpf+0x2cf/0x5d0 kernel/trace/bpf_trace.c:127
perf_trace_run_bpf_submit+0x7b/0x1d0 kernel/events/core.c:9708
perf_trace_lock+0x37c/0x440 include/trace/events/lock.h:39
trace_lock_release+0x128/0x150 include/trace/events/lock.h:58
lock_release+0x82/0x810 kernel/locking/lockdep.c:5636
__raw_spin_unlock_irqrestore include/linux/spinlock_api_smp.h:149 [inline]
_raw_spin_unlock_irqrestore+0x75/0x130 kernel/locking/spinlock.c:194
debug_hrtimer_deactivate kernel/time/hrtimer.c:425 [inline]
debug_deactivate kernel/time/hrtimer.c:481 [inline]
__run_hrtimer kernel/time/hrtimer.c:1653 [inline]
__hrtimer_run_queues+0x2f9/0xa60 kernel/time/hrtimer.c:1749
hrtimer_interrupt+0x3b3/0x1040 kernel/time/hrtimer.c:1811
local_apic_timer_interrupt arch/x86/kernel/apic/apic.c:1086 [inline]
__sysvec_apic_timer_interrupt+0xf9/0x270 arch/x86/kernel/apic/apic.c:1103
sysvec_apic_timer_interrupt+0x8c/0xb0 arch/x86/kernel/apic/apic.c:1097
asm_sysvec_apic_timer_interrupt+0x12/0x20
__raw_spin_unlock_irqrestore include/linux/spinlock_api_smp.h:152 [inline]
_raw_spin_unlock_irqrestore+0xd4/0x130 kernel/locking/spinlock.c:194
try_to_wake_up+0x702/0xd20 kernel/sched/core.c:4118
wake_up_process kernel/sched/core.c:4200 [inline]
wake_up_q+0x9a/0xf0 kernel/sched/core.c:953
futex_wake+0x50f/0x5b0 kernel/futex/waitwake.c:184
do_futex+0x367/0x560 kernel/futex/syscalls.c:127
__do_sys_futex kernel/futex/syscalls.c:199 [inline]
__se_sys_futex+0x401/0x4b0 kernel/futex/syscalls.c:180
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x44/0xd0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x44/0xaeThere is a possible deadlock with bpf_timer_* set of helpers:hrtimer_start()
lock_base();
trace_hrtimer...()
perf_event()
bpf_run()
bpf_timer_start()
hrtimer_start()
lock_base() <- DEADLOCK
Forbid use of bpf_ktime_get_coarse_ns() and bpf_timer_* helpers in BPF_PROG_TYPE_KPROBE, BPF_PROG_TYPE_TRACEPOINT, BPF_PROG_TYPE_PERF_EVENT and BPF_PROG_TYPE_RAW_TRACEPOINT prog types.
--- kernel/bpf/cgroup.c | 2 ++ kernel/bpf/helpers.c | 2 -- kernel/bpf/verifier.c | 7 +++++++ kernel/trace/bpf_trace.c | 2 -- net/core/filter.c | 6 ++++++ net/ipv4/bpf_tcp_ca.c | 2 ++ 6 files changed, 17 insertions(+), 4 deletions(-)
kernel/bpf/cgroup.c +2 -0
@@ -1773,6 +1773,8 @@ sysctl_func_proto(enum bpf_func_id func_return &bpf_sysctl_get_new_value_proto;case BPF_FUNC_sysctl_set_new_value:return &bpf_sysctl_set_new_value_proto; +case BPF_FUNC_ktime_get_coarse_ns: +return &bpf_ktime_get_coarse_ns_proto;default :return cgroup_base_func_proto (func_id, prog); }
kernel/bpf/helpers.c +0 -2
@@ -1367,8 +1367,6 @@ bpf_base_func_proto(enum bpf_func_id funreturn &bpf_ktime_get_ns_proto;case BPF_FUNC_ktime_get_boot_ns:return &bpf_ktime_get_boot_ns_proto; -case BPF_FUNC_ktime_get_coarse_ns: -return &bpf_ktime_get_coarse_ns_proto;case BPF_FUNC_ringbuf_output:return &bpf_ringbuf_output_proto;case BPF_FUNC_ringbuf_reserve:
kernel/bpf/verifier.c +7 -0
@@ -11451,6 +11451,13 @@ static int check_map_prog_compatibility( } } +if (map_value_has_timer (map)) { +if (is_tracing_prog_type (prog_type)) { +verbose (env,"tracing progs cannot use bpf_timer yet\n" ); +return -EINVAL ; + } + } +if ((bpf_prog_is_dev_bound (prog->aux) ||bpf_map_is_dev_bound (map)) && !bpf_offload_prog_map_match (prog, map)) {verbose (env,"offload device mismatch between prog and map\n" );
kernel/trace/bpf_trace.c +0 -2
@@ -1037,8 +1037,6 @@ bpf_tracing_func_proto(enum bpf_func_idreturn &bpf_ktime_get_ns_proto;case BPF_FUNC_ktime_get_boot_ns:return &bpf_ktime_get_boot_ns_proto; -case BPF_FUNC_ktime_get_coarse_ns: -return &bpf_ktime_get_coarse_ns_proto;case BPF_FUNC_tail_call:return &bpf_tail_call_proto;case BPF_FUNC_get_current_pid_tgid:
net/core/filter.c +6 -0
@@ -7162,6 +7162,8 @@ sock_filter_func_proto(enum bpf_func_id#endif case BPF_FUNC_sk_storage_get:return &bpf_sk_storage_get_cg_sock_proto; +case BPF_FUNC_ktime_get_coarse_ns: +return &bpf_ktime_get_coarse_ns_proto;default :return bpf_base_func_proto (func_id); }
@@ -10306,6 +10308,8 @@ sk_reuseport_func_proto(enum bpf_func_idreturn &sk_reuseport_load_bytes_relative_proto;case BPF_FUNC_get_socket_cookie:return &bpf_get_socket_ptr_cookie_proto; +case BPF_FUNC_ktime_get_coarse_ns: +return &bpf_ktime_get_coarse_ns_proto;default :return bpf_base_func_proto (func_id); }
@@ -10787,6 +10791,8 @@ bpf_sk_base_func_proto(enum bpf_func_idcase BPF_FUNC_skc_to_udp6_sock: func = &bpf_skc_to_udp6_sock_proto;break ; +case BPF_FUNC_ktime_get_coarse_ns: +return &bpf_ktime_get_coarse_ns_proto;default :return bpf_base_func_proto (func_id); }
net/ipv4/bpf_tcp_ca.c +2 -0
@@ -212,6 +212,8 @@ bpf_tcp_ca_get_func_proto(enum bpf_func_ offsetof(struct tcp_congestion_ops, release)) return &bpf_sk_getsockopt_proto;return NULL ; +case BPF_FUNC_ktime_get_coarse_ns: +return &bpf_ktime_get_coarse_ns_proto;default :return bpf_base_func_proto (func_id); }
Greg Kroah-Hartman
maintainerPATCH 270/279 selinux: fix NULL-pointer dereference when hashtab allocation fails
In reply to Greg Kroah-HartmanFrom: Ondrej Mosnacek <[email protected]>
commit dc27f3c5d10c58069672215787a96b4fae01818b upstream.
When the hash table slot array allocation fails in hashtab_init(), h->size is left initialized with a non-zero value, but the h->htable pointer is NULL. This may then cause a NULL pointer dereference, since the policydb code relies on the assumption that even after a failed hashtab_init(), hashtab_map() and hashtab_destroy() can be safely called on it. Yet, these detect an empty hashtab only by looking at the size.
Fix this by making sure that hashtab_init() always leaves behind a valid empty hashtab when the allocation fails.
--- security/selinux/ss/hashtab.c | 17 ++++++++++++----- 1 file changed, 12 insertions(+), 5 deletions(-)
security/selinux/ss/hashtab.c +12 -5
@@ -30,13 +30,20 @@ static u32 hashtab_compute_size(u32 nel)int hashtab_init (struct hashtab *h,u32 nel_hint) { - h->size =hashtab_compute_size (nel_hint); +u32 size =hashtab_compute_size (nel_hint); + +/* should already be zeroed, but better be safe */ h->nel =0 ; -if (!h->size) -return 0 ; + h->size =0 ; + h->htable =NULL ; - h->htable =kcalloc (h->size,sizeof (*h->htable),GFP_KERNEL ); -return h->htable ?0 : -ENOMEM ; +if (size) { + h->htable =kcalloc (size,sizeof (*h->htable),GFP_KERNEL ); +if (!h->htable) +return -ENOMEM ; + h->size = size; + } +return 0 ; } int __hashtab_insert(struct hashtab *h, struct hashtab_node **dst,
Greg Kroah-Hartman
maintainerPATCH 271/279 ASoC: DAPM: Cover regression by kctl change notification fix
In reply to Greg Kroah-HartmanFrom: Takashi Iwai <[email protected]>
commit 827b0913a9d9d07a0c3e559dbb20ca4d6d285a54 upstream.
The recent fix for DAPM to correct the kctl change notification by the commit 5af82c81b2c4 ("ASoC: DAPM: Fix missing kctl change notifications") caused other regressions since it changed the behavior of snd_soc_dapm_set_pin() that is called from several API functions. Formerly it returned always 0 for success, but now it returns 0 or 1.
This patch addresses it, restoring the old behavior of snd_soc_dapm_set_pin() while keeping the fix in snd_soc_dapm_put_pin_switch().
--- sound/soc/soc-dapm.c | 29 +++++++++++++++++++++++------ 1 file changed, 23 insertions(+), 6 deletions(-)
sound/soc/soc-dapm.c +23 -6
@@ -2557,8 +2557,13 @@ static struct snd_soc_dapm_widget *dapm_return NULL ; } -static int snd_soc_dapm_set_pin (struct snd_soc_dapm_context *dapm, -const char *pin,int status) +/* +* set the DAPM pin status: +* returns 1 when the value has been updated, 0 when unchanged, or a negative +* error code; called from kcontrol put callback +*/ +static int __snd_soc_dapm_set_pin (struct snd_soc_dapm_context *dapm, +const char *pin,int status) {struct snd_soc_dapm_widget *w =dapm_find_widget (dapm, pin, true);int ret =0 ;
@@ -2584,6 +2589,18 @@ static int snd_soc_dapm_set_pin(struct sreturn ret; } +/* +* similar as __snd_soc_dapm_set_pin(), but returns 0 when successful; +* called from several API functions below +*/ +static int snd_soc_dapm_set_pin (struct snd_soc_dapm_context *dapm, +const char *pin,int status) +{ +int ret =__snd_soc_dapm_set_pin (dapm, pin, status); + +return ret <0 ? ret :0 ; +} +/** * snd_soc_dapm_sync_unlocked - scan and power dapm paths * @dapm: DAPM context
@@ -3587,10 +3604,10 @@ int snd_soc_dapm_put_pin_switch(struct sconst char *pin = (const char *)kcontrol->private_value;int ret; -if (ucontrol->value.integer.value[0 ]) - ret =snd_soc_dapm_enable_pin (&card->dapm, pin); -else - ret =snd_soc_dapm_disable_pin (&card->dapm, pin); +mutex_lock_nested (&card->dapm_mutex,SND_SOC_DAPM_CLASS_RUNTIME ); + ret =__snd_soc_dapm_set_pin (&card->dapm, pin, + !!ucontrol->value.integer.value[0 ]); +mutex_unlock (&card->dapm_mutex);snd_soc_dapm_sync (&card->dapm);return ret;
Greg Kroah-Hartman
maintainerPATCH 272/279 ASoC: rsnd: fixup DMAEngine API
In reply to Greg Kroah-HartmanFrom: Kuninori Morimoto <[email protected]>
commit 2ce1b21cb3326e12af3c72c47e1d294b19d73947 upstream.
commit d5bb69dc54ec1 ("ASoC: sh: rcar: dma: : use proper DMAENGINE API for termination") updated DMAEngine API _all() to _sync(), but it should be _async()._all() and _async() are almost same, the difference is only return error code. _sync() will call dmaengine_synchronize() and will be kernel panic.
This patch is needed for v5.15 or later.
[ 27.293264] BUG: scheduling while atomic: irq/130-ec70000/131/0x00000003
[ 27.300084] 2 locks held by irq/130-ec70000/131:
[ 27.304743] #0: ffff0004c274d908 (&group->lock){....}-{2:2}, at: _snd_pcm_stream_lock_irqsave+0x48/0x54
[ 27.314344] #1: ffff0004c1788c60 (&priv->lock#2){....}-{2:2}, at: rsnd_soc_dai_trigger+0x70/0x7bc
[ 27.323409] irq event stamp: 206
[ 27.326664] hardirqs last enabled at (205): [<ffff80001082de50>] _raw_spin_unlock_irq+0x50/0xa0
[ 27.335529] hardirqs last disabled at (206): [<ffff80001082d9e4>] _raw_spin_lock_irqsave+0xc4/0xd0
[ 27.344564] softirqs last enabled at (0): [<ffff800010037324>] copy_process+0x644/0x1b10
[ 27.352819] softirqs last disabled at (0): [<0000000000000000>] 0x0
[ 27.359142] CPU: 0 PID: 131 Comm: irq/130-ec70000 Not tainted 5.14.0-rc1+ #918
[ 27.366429] Hardware name: Renesas H3ULCB Kingfisher board based on r8a77950 (DT)
[ 27.373975] Call trace:
[ 27.376442] dump_backtrace+0x0/0x1b4
[ 27.380141] show_stack+0x24/0x30
[ 27.383488] dump_stack_lvl+0x8c/0xb8
[ 27.387184] dump_stack+0x18/0x34
[ 27.390528] __schedule_bug+0x8c/0x9c
[ 27.394224] __schedule+0x790/0x8dc
[ 27.397746] schedule+0x7c/0x110
[ 27.401003] synchronize_irq+0x94/0xd0
[ 27.404786] rcar_dmac_device_synchronize+0x20/0x2c
[ 27.409710] rsnd_dmaen_stop+0x50/0x64
[ 27.413495] rsnd_soc_dai_trigger+0x554/0x7bc
[ 27.417890] snd_soc_pcm_dai_trigger+0xe8/0x264--- sound/soc/sh/rcar/dma.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
sound/soc/sh/rcar/dma.c +1 -1
@@ -102,7 +102,7 @@ static int rsnd_dmaen_stop(struct rsnd_mstruct rsnd_dmaen *dmaen =rsnd_dma_to_dmaen (dma);if (dmaen->chan) -dmaengine_terminate_sync (dmaen->chan); +dmaengine_terminate_async (dmaen->chan);return 0 ; }
Greg Kroah-Hartman
maintainerPATCH 273/279 usb: max-3421: Use driver data instead of maintaining a list of bound devices
In reply to Greg Kroah-HartmanFrom: Uwe Kleine-König <[email protected]>
commit fc153aba3ef371d0d76eb88230ed4e0dee5b38f2 upstream.
Instead of maintaining a single-linked list of devices that must be searched linearly in .remove() just use spi_set_drvdata() to remember the link between the spi device and the driver struct. Then the global list and the next member can be dropped.
This simplifies the driver, reduces the memory footprint and the time to search the list. Also it makes obvious that there is always a corresponding driver struct for a given device in .remove(), so the error path for !max3421_hcd can be dropped, too.
As a side effect this fixes a data inconsistency when .probe() races with itself for a second max3421 device in manipulating max3421_hcd_list. A similar race is fixed in .remove(), too.
--- drivers/usb/host/max3421-hcd.c | 25 +++++-------------------- 1 file changed, 5 insertions(+), 20 deletions(-)
drivers/usb/host/max3421-hcd.c +5 -20
@@ -125,8 +125,6 @@ struct max3421_hcd {struct task_struct *spi_thread; -struct max3421_hcd *next; -enum max3421_rh_state rh_state;/* lower 16 bits contain port status, upper 16 bits the change mask: */ u32 port_status;
@@ -174,8 +172,6 @@ struct max3421_ep {u8 retransmit;/* packet needs retransmission */ }; -static struct max3421_hcd *max3421_hcd_list; -#define MAX3421_FIFO_SIZE 64#define MAX3421_SPI_DIR_RD 0/* read register from MAX3421 */
@@ -1882,9 +1878,8 @@ max3421_probe(struct spi_device *spi) }set_bit (HCD_FLAG_POLL_RH , &hcd->flags); max3421_hcd =hcd_to_max3421 (hcd); - max3421_hcd->next = max3421_hcd_list; - max3421_hcd_list = max3421_hcd;INIT_LIST_HEAD (&max3421_hcd->ep_list); +spi_set_drvdata (spi, max3421_hcd); max3421_hcd->tx =kmalloc (sizeof (*max3421_hcd->tx),GFP_KERNEL );if (!max3421_hcd->tx)
@@ -1934,28 +1929,18 @@ error:static int max3421_remove (struct spi_device *spi) { -struct max3421_hcd *max3421_hcd =NULL , **prev; -struct usb_hcd *hcd =NULL ; +struct max3421_hcd *max3421_hcd; +struct usb_hcd *hcd;unsigned long flags; -for (prev = &max3421_hcd_list; *prev; prev = &(*prev)->next) { - max3421_hcd = *prev; - hcd =max3421_to_hcd (max3421_hcd); -if (hcd->self.controller == &spi->dev) -break ; - } -if (!max3421_hcd) { -dev_err (&spi->dev,"no MAX3421 HCD found for SPI device %p\n" , - spi); -return -ENODEV ; - } + max3421_hcd =spi_get_drvdata (spi); + hcd =max3421_to_hcd (max3421_hcd);usb_remove_hcd (hcd);spin_lock_irqsave (&max3421_hcd->lock, flags);kthread_stop (max3421_hcd->spi_thread); - *prev = max3421_hcd->next;spin_unlock_irqrestore (&max3421_hcd->lock, flags);
Greg Kroah-Hartman
maintainerPATCH 274/279 ice: Fix VF true promiscuous mode
In reply to Greg Kroah-HartmanFrom: Brett Creeley <[email protected]>
commit 1a8c7778bcde5981463a5b9f9b2caa44a327ff93 upstream.
When a VF requests promiscuous mode and it's trusted and true promiscuous mode is enabled the PF driver attempts to enable unicast and/or multicast promiscuous mode filters based on the request. This is fine, but there are a couple issues with the current code.
[1] The define to configure the unicast promiscuous mode mask also includes bits to configure the multicast promiscuous mode mask, which causes multicast to be set/cleared unintentionally.
[2] All 4 cases for enable/disable unicast/multicast mode are not handled in the promiscuous mode message handler, which causes unexpected results regarding the current promiscuous mode settings.
To fix [1] make sure any promiscuous mask defines include the correct bits for each of the promiscuous modes.
To fix [2] make sure that all 4 cases are handled since there are 2 bits (FLAG_VF_UNICAST_PROMISC and FLAG_VF_MULTICAST_PROMISC) that can be either set or cleared. Also, since either unicast and/or multicast promiscuous configuration can fail, introduce two separate error values to handle each of these cases.
--- drivers/net/ethernet/intel/ice/ice.h | 5 - drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c | 78 +++++++++++------------ 2 files changed, 40 insertions(+), 43 deletions(-)
drivers/net/ethernet/intel/ice/ice.h +1 -4
@@ -139,13 +139,10 @@#define ice_for_each_q_vector (vsi, i) \ for ((i) = 0; (i) < (vsi)->num_q_vectors; (i)++) -#define ICE_UCAST_PROMISC_BITS (ICE_PROMISC_UCAST_TX | ICE_PROMISC_MCAST_TX | \ - ICE_PROMISC_UCAST_RX | ICE_PROMISC_MCAST_RX) +#define ICE_UCAST_PROMISC_BITS (ICE_PROMISC_UCAST_TX | ICE_PROMISC_UCAST_RX)#define ICE_UCAST_VLAN_PROMISC_BITS (ICE_PROMISC_UCAST_TX | \ - ICE_PROMISC_MCAST_TX | \ ICE_PROMISC_UCAST_RX | \ - ICE_PROMISC_MCAST_RX | \ ICE_PROMISC_VLAN_TX | \ ICE_PROMISC_VLAN_RX)
drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c +39 -39
97 lines not on this page · raw message
Greg Kroah-Hartman
maintainerPATCH 275/279 ice: Delete always true check of PF pointer
In reply to Greg Kroah-HartmanFrom: Leon Romanovsky <[email protected]>
commit 2ff04286a9569675948f39cec2c6ad47c3584633 upstream.
PF pointer is always valid when PCI core calls its .shutdown() and .remove() callbacks. There is no need to check it again.
--- drivers/net/ethernet/intel/ice/ice_main.c | 3 --- 1 file changed, 3 deletions(-)
drivers/net/ethernet/intel/ice/ice_main.c +0 -3
@@ -4600,9 +4600,6 @@ static void ice_remove(struct pci_dev *pstruct ice_pf *pf =pci_get_drvdata (pdev);int i; -if (!pf) -return ; -for (i =0 ; i <ICE_MAX_RESET_WAIT ; i++) {if (!ice_is_reset_in_progress (pf->state))break ;
Greg Kroah-Hartman
maintainerPATCH 276/279 fs: export an inode_update_time helper
In reply to Greg Kroah-HartmanFrom: Josef Bacik <[email protected]>
commit e60feb445fce9e51c1558a6aa7faf9dd5ded533b upstream.
If you already have an inode and need to update the time on the inode there is no way to do this properly. Export this helper to allow file systems to update time on the inode so the appropriate handler is called, either ->update_time or generic_update_time.
--- fs/inode.c | 7 ++++--- include/linux/fs.h | 2 ++ 2 files changed, 6 insertions(+), 3 deletions(-)
fs/inode.c +4 -3
@@ -1782,12 +1782,13 @@ EXPORT_SYMBOL(generic_update_time);* This does the actual work of updating an inodes time or version. Must have * had called mnt_want_write() before calling this. */ -static int update_time (struct inode *inode,struct timespec64 *time,int flags) +int inode_update_time (struct inode *inode,struct timespec64 *time,int flags) {if (inode->i_op->update_time)return inode->i_op->update_time (inode, time, flags);return generic_update_time (inode, time, flags); } +EXPORT_SYMBOL (inode_update_time);/** * atime_needs_update - update the access time
@@ -1857,7 +1858,7 @@ void touch_atime(const struct path *path* of the fs read only, e.g. subvolumes in Btrfs. */ now =current_time (inode); -update_time (inode, &now,S_ATIME ); +inode_update_time (inode, &now,S_ATIME );__mnt_drop_write (mnt); skip_update:sb_end_write (inode->i_sb);
@@ -2002,7 +2003,7 @@ int file_update_time(struct file *file)if (__mnt_want_write_file (file))return 0 ; - ret =update_time (inode, &now, sync_it); + ret =inode_update_time (inode, &now, sync_it);__mnt_drop_write_file (file);return ret;
include/linux/fs.h +2 -0
@@ -2498,6 +2498,8 @@ enum file_time_flags {extern bool atime_needs_update (const struct path *,struct inode *);extern void touch_atime (const struct path *); +int inode_update_time (struct inode *inode,struct timespec64 *time,int flags); +static inline void file_accessed (struct file *file) {if (!(file->f_flags &O_NOATIME ))
Greg Kroah-Hartman
maintainerPATCH 277/279 btrfs: update device path inode time instead of bd_inode
In reply to Greg Kroah-HartmanFrom: Josef Bacik <[email protected]>
commit 54fde91f52f515e0b1514f0f0fa146e87a672227 upstream.
Christoph pointed out that I'm updating bdev->bd_inode for the device time when we remove block devices from a btrfs file system, however this isn't actually exposed to anything. The inode we want to update is the one that's associated with the path to the device, usually on devtmpfs, so that blkid notices the difference.
We still don't want to do the blkdev_open, so use kern_path() to get the path to the given device and do the update time on that inode.
--- fs/btrfs/volumes.c | 21 +++++++++++++-------- 1 file changed, 13 insertions(+), 8 deletions(-)
fs/btrfs/volumes.c +13 -8
@@ -14,6 +14,7 @@#include <linux/semaphore.h> #include <linux/uuid.h> #include <linux/list_sort.h> +#include <linux/namei.h> #include "misc.h" #include "ctree.h" #include "extent_map.h"
@@ -1884,18 +1885,22 @@ out:/* * Function to update ctime/mtime for a given device path. * Mainly used for ctime/mtime based probe like libblkid. +* +* We don't care about errors here, this is just to be kind to userspace. */ -static void update_dev_time (struct block_device *bdev) +static void update_dev_time (const char *device_path) { -struct inode *inode = bdev->bd_inode; +struct path path;struct timespec64 now; +int ret; -/* Shouldn't happen but just in case. */ -if (!inode) + ret =kern_path (device_path,LOOKUP_FOLLOW , &path); +if (ret)return ; - now =current_time (inode); -generic_update_time (inode, &now,S_MTIME |S_CTIME ); + now =current_time (d_inode (path.dentry)); +inode_update_time (d_inode (path.dentry), &now,S_MTIME |S_CTIME ); +path_put (&path); }static int btrfs_rm_dev_item (struct btrfs_device *device)
@@ -2071,7 +2076,7 @@ void btrfs_scratch_superblocks(struct btbtrfs_kobject_uevent (bdev,KOBJ_CHANGE );/* Update ctime/mtime for device path for libblkid */ -update_dev_time (bdev); +update_dev_time (device_path); } int btrfs_rm_device(struct btrfs_fs_info *fs_info, const char *device_path,
@@ -2735,7 +2740,7 @@ int btrfs_init_new_device(struct btrfs_fbtrfs_forget_devices (device_path);/* Update ctime/mtime for blkid or udev */ -update_dev_time (bdev); +update_dev_time (device_path);return ret;
Greg Kroah-Hartman
maintainerPATCH 278/279 net: add and use skb_unclone_keeptruesize() helper
In reply to Greg Kroah-HartmanFrom: Eric Dumazet <[email protected]>
commit c4777efa751d293e369aec464ce6875e957be255 upstream.
While commit 097b9146c0e2 ("net: fix up truesize of cloned skb in skb_prepare_for_shift()") fixed immediate issues found when KFENCE was enabled/tested, there are still similar issues, when tcp_trim_head() hits KFENCE while the master skb is cloned.
This happens under heavy networking TX workloads, when the TX completion might be delayed after incoming ACK.
This patch fixes the WARNING in sk_stream_kill_queues when sk->sk_mem_queued/sk->sk_forward_alloc are not zero.
--- include/linux/skbuff.h | 16 ++++++++++++++++ net/core/skbuff.c | 14 +------------- net/ipv4/tcp_output.c | 6 +++--- 3 files changed, 20 insertions(+), 16 deletions(-)
include/linux/skbuff.h +16 -0
@@ -1671,6 +1671,22 @@ static inline int skb_unclone(struct sk_return 0 ; } +/* This variant of skb_unclone() makes sure skb->truesize is not changed */ +static inline int skb_unclone_keeptruesize (struct sk_buff *skb,gfp_t pri) +{ +might_sleep_if (gfpflags_allow_blocking (pri)); + +if (skb_cloned (skb)) { +unsigned int save = skb->truesize; +int res; + + res =pskb_expand_head (skb,0 ,0 , pri); + skb->truesize = save; +return res; + } +return 0 ; +} +/** * skb_header_cloned - is the header a clone * @skb: buffer to check
net/core/skbuff.c +1 -13
@@ -3449,19 +3449,7 @@ EXPORT_SYMBOL(skb_split);*/ static int skb_prepare_for_shift (struct sk_buff *skb) { -int ret =0 ; - -if (skb_cloned (skb)) { -/* Save and restore truesize: pskb_expand_head() may reallocate -* memory where ksize(kmalloc(S)) != ksize(kmalloc(S)), but we -* cannot change truesize at this point. -*/ -unsigned int save_truesize = skb->truesize; - - ret =pskb_expand_head (skb,0 ,0 ,GFP_ATOMIC ); - skb->truesize = save_truesize; - } -return ret; +return skb_unclone_keeptruesize (skb,GFP_ATOMIC ); }/**
net/ipv4/tcp_output.c +3 -3
@@ -1562,7 +1562,7 @@ int tcp_fragment(struct sock *sk, enum treturn -ENOMEM ; } -if (skb_unclone (skb, gfp)) +if (skb_unclone_keeptruesize (skb, gfp))return -ENOMEM ;/* Get a new skb... force flag on. */
@@ -1672,7 +1672,7 @@ int tcp_trim_head(struct sock *sk, struc {u32 delta_truesize; -if (skb_unclone (skb,GFP_ATOMIC )) +if (skb_unclone_keeptruesize (skb,GFP_ATOMIC ))return -ENOMEM ; delta_truesize =__pskb_trim_head (skb, len);
@@ -3184,7 +3184,7 @@ int __tcp_retransmit_skb(struct sock *sk cur_mss, GFP_ATOMIC))return -ENOMEM ;/* We'll try again later. */ }else { -if (skb_unclone (skb,GFP_ATOMIC )) +if (skb_unclone_keeptruesize (skb,GFP_ATOMIC ))return -ENOMEM ; diff =tcp_skb_pcount (skb);
Greg Kroah-Hartman
maintainerPATCH 279/279 x86/Kconfig: Fix an unused variable error in dell-smm-hwmon
In reply to Greg Kroah-HartmanFrom: Randy Dunlap <[email protected]>
commit ef775a0e36c6a81c5b07cb228c02f967133fe768 upstream.
When CONFIG_PROC_FS is not set, there is a build warning (turned into an error):
../drivers/hwmon/dell-smm-hwmon.c: In function 'i8k_init_procfs':
../drivers/hwmon/dell-smm-hwmon.c:624:24: error: unused variable 'data' [-Werror=unused-variable]
struct dell_smm_data *data = dev_get_drvdata(dev);
Make I8K depend on PROC_FS and HWMON (instead of selecting HWMON -- it is strongly preferred to not select entire subsystems).
Build tested in all possible combinations of SENSORS_DELL_SMM, I8K, and PROC_FS.
--- arch/x86/Kconfig | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
arch/x86/Kconfig +2 -1
@@ -1256,7 +1256,8 @@ config TOSHIBA config I8K tristate "Dell i8k legacy laptop support" - select HWMON + depends on HWMON + depends on PROC_FS select SENSORS_DELL_SMM help This option enables legacy /proc/i8k userspace interface in hwmon
Quoted 9 lines · Greg Kroah-Hartman
This is the start of the stable review cycle for the 5.15.5 release. There are 279 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know.
Responses should be made by Fri, 26 Nov 2021 11:56:36 +0000. Anything received after that time might be too late.
The whole patch series can be found in one patch at:
Quoted 9 lines
https://www.kernel.org/pub/linux/kernel/v5.x/stable-review/patch-5.15.5-rc1.gzor in the git tree and branch at:
git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.gitlinux-5.15.y
and the diffstat can be found below.thanks,
greg k-h
5.15.5-rc1 Successfully Compiled and booted on my Raspberry PI 4b (8g) (bcm2711)
Quoted 8 lines · Greg Kroah-Hartman
This is the start of the stable review cycle for the 5.15.5 release. There are 279 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know.
Responses should be made by Fri, 26 Nov 2021 11:56:36 +0000. Anything received after that time might be too late.
Build results:
total: 154 pass: 154 fail: 0 Qemu test results: total: 482 pass: 482 fail: 0
Guenter
Quoted 10 lines · Greg Kroah-Hartman
This is the start of the stable review cycle for the 5.15.5 release. There are 279 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know.
Responses should be made by Fri, 26 Nov 2021 11:56:36 +0000. Anything received after that time might be too late.
The whole patch series can be found in one patch at:
Quoted 8 lines
https://www.kernel.org/pub/linux/kernel/v5.x/stable-review/patch-5.15.5-rc1.gzor in the git tree and branch at:
git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.gitlinux-5.15.y
and the diffstat can be found below.thanks,
greg k-h
Results from Linaro’s test farm.
No regressions on arm64, arm, x86_64, and i386.
## Build
* kernel: 5.15.5-rc1
* git: https://gitlab.com/Linaro/lkft/mirrors/stable/linux-stable-rc
* git branch: linux-5.15.y
* git commit: e3bb2e602026d5d841d7a2e62bfbe579265bb5ff
* git describe: v5.15.4-280-ge3bb2e602026
* test details:
https://qa-reports.linaro.org/lkft/linux-stable-rc-linux-5.15.y/build/v5.15.4-280-ge3bb2e602026
## No regressions (compared to v5.15.3)
## No fixes (compared to v5.15.3)
## Test result summary
total: 97214, pass: 82201, fail: 1196, skip: 13115, xfail: 702
## Build Summary
* arc: 10 total, 10 passed, 0 failed
* arm: 259 total, 259 passed, 0 failed
* arm64: 37 total, 35 passed, 2 failed
* dragonboard-410c: 1 total, 1 passed, 0 failed
* hi6220-hikey: 1 total, 1 passed, 0 failed
* i386: 36 total, 36 passed, 0 failed
* juno-r2: 1 total, 1 passed, 0 failed
* mips: 34 total, 34 passed, 0 failed
* parisc: 12 total, 12 passed, 0 failed
* powerpc: 45 total, 42 passed, 3 failed
* riscv: 24 total, 24 passed, 0 failed
* s390: 18 total, 18 passed, 0 failed
* sh: 20 total, 20 passed, 0 failed
* sparc: 12 total, 12 passed, 0 failed
* x15: 1 total, 1 passed, 0 failed
* x86: 1 total, 1 passed, 0 failed
* x86_64: 37 total, 37 passed, 0 failed
## Test suites summary
* fwts
* igt-gpu-tools
* kselftest-android
* kselftest-arm64
* kselftest-arm64/arm64.btitest.bti_c_func
* kselftest-arm64/arm64.btitest.bti_j_func
* kselftest-arm64/arm64.btitest.bti_jc_func
* kselftest-arm64/arm64.btitest.bti_none_func
* kselftest-arm64/arm64.btitest.nohint_func
* kselftest-arm64/arm64.btitest.paciasp_func
* kselftest-arm64/arm64.nobtitest.bti_c_func
* kselftest-arm64/arm64.nobtitest.bti_j_func
* kselftest-arm64/arm64.nobtitest.bti_jc_func
* kselftest-arm64/arm64.nobtitest.bti_none_func
* kselftest-arm64/arm64.nobtitest.nohint_func
* kselftest-arm64/arm64.nobtitest.paciasp_func
* kselftest-bpf
* kselftest-breakpoints
* kselftest-capabilities
* kselftest-cgroup
* kselftest-clone3
* kselftest-core
* kselftest-cpu-hotplug
* kselftest-cpufreq
* kselftest-drivers
* kselftest-efivarfs
* kselftest-filesystems
* kselftest-firmware
* kselftest-fpu
* kselftest-futex
* kselftest-gpio
* kselftest-intel_pstate
* kselftest-ipc
* kselftest-ir
* kselftest-kcmp
* kselftest-kexec
* kselftest-kvm
* kselftest-lib
* kselftest-livepatch
* kselftest-lkdtm
* kselftest-membarrier
* kselftest-memfd
* kselftest-memory-hotplug
* kselftest-mincore
* kselftest-mount
* kselftest-mqueue
* kselftest-net
* kselftest-netfilter
* kselftest-nsfs
* kselftest-openat2
* kselftest-pid_namespace
* kselftest-pidfd
* kselftest-proc
* kselftest-pstore
* kselftest-ptrace
* kselftest-rseq
* kselftest-rtc
* kselftest-seccomp
* kselftest-sigaltstack
* kselftest-size
* kselftest-splice
* kselftest-static_keys
* kselftest-sync
* kselftest-sysctl
* kselftest-tc-testing
* kselftest-timens
* kselftest-timers
* kselftest-tmpfs
* kselftest-tpm2
* kselftest-user
* kselftest-vm
* kselftest-x86
* kselftest-zram
* kunit
* kvm-unit-tests
* libgpiod
* libhugetlbfs
* linux-log-parser
* ltp-cap_bounds-tests
* ltp-commands-tests
* ltp-containers-tests
* ltp-controllers-tests
* ltp-cpuhotplug-tests
* ltp-crypto-tests
* ltp-cve-tests
* ltp-dio-tests
* ltp-fcntl-locktests-tests
* ltp-filecaps-tests
* ltp-fs-tests
* ltp-fs_bind-tests
* ltp-fs_perms_simple-tests
* ltp-fsx-tests
* ltp-hugetlb-tests
* ltp-io-tests
* ltp-ipc-tests
* ltp-math-tests
* ltp-mm-tests
* ltp-nptl-tests
* ltp-open-posix-tests
* ltp-pty-tests
* ltp-sched-tests
* ltp-securebits-tests
* ltp-syscalls-tests
* ltp-tracing-tests
* network-basic-tests
* packetdrill
* perf
* perf/Zstd-perf.data-compression
* rcutorture
* ssuite
* v4l2-compliance
--
Linaro LKFT
https://lkft.linaro.org
Quoted 7 lines · Greg Kroah-Hartman
This is the start of the stable review cycle for the 5.15.5 release. There are 279 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know.
Responses should be made by Fri, 26 Nov 2021 11:56:36 +0000. Anything received after that time might be too late.
Hi Greg,
Looking good.
Run tested on:
- Intel Tiger Lake x86_64 (nuc11 i7-1165G7)
In addition build tested on:
- Allwinner A64
- Allwinner H3
- Allwinner H5
- Allwinner H6
- NXP iMX6
- NXP iMX8
- Qualcomm Dragonboard
- Rockchip RK3288
- Rockchip RK3328
- Rockchip RK3399pro
- Samsung Exynos
Signature · 1 line
Rudi
Quoted 9 lines · Greg Kroah-Hartman
This is the start of the stable review cycle for the 5.15.5 release. There are 279 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know.
Responses should be made by Fri, 26 Nov 2021 11:56:36 +0000. Anything received after that time might be too late.
The whole patch series can be found in one patch at:
Quoted 8 lines
https://www.kernel.org/pub/linux/kernel/v5.x/stable-review/patch-5.15.5-rc1.gzor in the git tree and branch at:
git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.gitlinux-5.15.y
and the diffstat can be found below.thanks,
greg k-h
No new regressions.
Test results for stable-v5.15:
10 builds: 10 pass, 0 fail
28 boots: 28 pass, 0 fail
114 tests: 108 pass, 6 failLinux version: 5.15.5-rc1-ge3bb2e602026
Boards tested: tegra124-jetson-tk1, tegra186-p2771-0000,
tegra194-p2972-0000, tegra194-p3509-0000+p3668-0000,
tegra20-ventana, tegra210-p2371-2180,
tegra210-p3450-0000, tegra30-cardhu-a04Test failures: tegra194-p2972-0000: boot.py
tegra194-p2972-0000: tegra-audio-boot-sanity.sh
tegra194-p2972-0000: tegra-audio-hda-playback.sh
tegra194-p3509-0000+p3668-0000: devices
tegra194-p3509-0000+p3668-0000: tegra-audio-boot-sanity.sh
tegra194-p3509-0000+p3668-0000: tegra-audio-hda-playback.shJon
Signature · 1 line
nvpublic