5.3.11-stable review
Subsystems (5)
Greg Kroah-Hartman
maintainerPATCH 154/193 sched/topology: Dont try to build empty sched domains
In reply to Greg Kroah-HartmanFrom: Valentin Schneider <[email protected]>
[ Upstream commit cd1cb3350561d2bf544ddfef76fbf0b1c9c7178f ]
Turns out hotplugging CPUs that are in exclusive cpusets can lead to the cpuset code feeding empty cpumasks to the sched domain rebuild machinery.
This leads to the following splat:
Internal error: Oops: 96000004 [#1] PREEMPT SMP
Modules linked in:
CPU: 0 PID: 235 Comm: kworker/5:2 Not tainted 5.4.0-rc1-00005-g8d495477d62e #23
Hardware name: ARM Juno development board (r0) (DT)
Workqueue: events cpuset_hotplug_workfn
pstate: 60000005 (nZCv daif -PAN -UAO)
pc : build_sched_domains (./include/linux/arch_topology.h:23 kernel/sched/topology.c:1898 kernel/sched/topology.c:1969)
lr : build_sched_domains (kernel/sched/topology.c:1966)
Call trace:
build_sched_domains (./include/linux/arch_topology.h:23 kernel/sched/topology.c:1898 kernel/sched/topology.c:1969)
partition_sched_domains_locked (kernel/sched/topology.c:2250)
rebuild_sched_domains_locked (./include/linux/bitmap.h:370 ./include/linux/cpumask.h:538 kernel/cgroup/cpuset.c:955 kernel/cgroup/cpuset.c:978 kernel/cgroup/cpuset.c:1019)
rebuild_sched_domains (kernel/cgroup/cpuset.c:1032)
cpuset_hotplug_workfn (kernel/cgroup/cpuset.c:3205 (discriminator 2))
process_one_work (./arch/arm64/include/asm/jump_label.h:21 ./include/linux/jump_label.h:200 ./include/trace/events/workqueue.h:114 kernel/workqueue.c:2274)
worker_thread (./include/linux/compiler.h:199 ./include/linux/list.h:268 kernel/workqueue.c:2416)
kthread (kernel/kthread.c:255)
ret_from_fork (arch/arm64/kernel/entry.S:1167)
Code: f860dae2 912802d6 aa1603e1 12800000 (f8616853)The faulty line in question is:
cap = arch_scale_cpu_capacity(cpumask_first(cpu_map));
and we're not checking the return value against nr_cpu_ids (we shouldn't have to!), which leads to the above.
Prevent generate_sched_domains() from returning empty cpumasks, and add some assertion in build_sched_domains() to scream bloody murder if it happens again.
The above splat was obtained on my Juno r0 with the following reproducer:
$ cgcreate -g cpuset:asym $ cgset -r cpuset.cpus=0-3 asym $ cgset -r cpuset.mems=0 asym $ cgset -r cpuset.cpu_exclusive=1 asym
$ cgcreate -g cpuset:smp $ cgset -r cpuset.cpus=4-5 smp $ cgset -r cpuset.mems=0 smp $ cgset -r cpuset.cpu_exclusive=1 smp
$ cgset -r cpuset.sched_load_balance=0 .
$ echo 0 > /sys/devices/system/cpu/cpu4/online $ echo 0 > /sys/devices/system/cpu/cpu5/online
--- kernel/cgroup/cpuset.c | 3 ++- kernel/sched/topology.c | 5 ++++- 2 files changed, 6 insertions(+), 2 deletions(-)
kernel/cgroup/cpuset.c +2 -1
@@ -786,7 +786,8 @@ static int generate_sched_domains(cpumask_var_t **domains, cpumask_subset(cp->cpus_allowed, top_cpuset.effective_cpus)) continue; -if (is_sched_load_balance (cp)) +if (is_sched_load_balance (cp) && + !cpumask_empty (cp->effective_cpus)) csa[csn++] = cp;/* skip @cp's subtree if not a partition root */
kernel/sched/topology.c +4 -1
@@ -1927,7 +1927,7 @@ next_level:static int build_sched_domains (const struct cpumask *cpu_map,struct sched_domain_attr *attr) { -enum s_alloc alloc_state; +enum s_alloc alloc_state = sa_none;struct sched_domain *sd;struct s_data d;struct rq *rq =NULL ;
@@ -1935,6 +1935,9 @@ build_sched_domains(const struct cpumask *cpu_map, struct sched_domain_attr *attstruct sched_domain_topology_level *tl_asym;bool has_asym = false; +if (WARN_ON (cpumask_empty (cpu_map))) + goto error; + alloc_state =__visit_domain_allocation_hell (&d, cpu_map);if (alloc_state != sa_rootdomain) goto error;
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 155/193 sched/topology: Allow sched_asym_cpucapacity to be disabled
In reply to Greg Kroah-HartmanFrom: Valentin Schneider <[email protected]>
[ Upstream commit e284df705cf1eeedb5ec3a66ed82d17a64659150 ]
While the static key is correctly initialized as being disabled, it will remain forever enabled once turned on. This means that if we start with an asymmetric system and hotplug out enough CPUs to end up with an SMP system, the static key will remain set - which is obviously wrong. We should detect this and turn off things like misfit migration and capacity aware wakeups.
As Quentin pointed out, having separate root domains makes this slightly trickier. We could have exclusive cpusets that create an SMP island - IOW, the domains within this root domain will not see any asymmetry. This means we can't just disable the key on domain destruction, we need to count how many asymmetric root domains we have.
Consider the following example using Juno r0 which is 2+4 big.LITTLE, where two identical cpusets are created: they both span both big and LITTLE CPUs:
asym0 asym1
[ ][ ]
L L B L L B
$ cgcreate -g cpuset:asym0 $ cgset -r cpuset.cpus=0,1,3 asym0 $ cgset -r cpuset.mems=0 asym0 $ cgset -r cpuset.cpu_exclusive=1 asym0
$ cgcreate -g cpuset:asym1 $ cgset -r cpuset.cpus=2,4,5 asym1 $ cgset -r cpuset.mems=0 asym1 $ cgset -r cpuset.cpu_exclusive=1 asym1
$ cgset -r cpuset.sched_load_balance=0 .
(the CPU numbering may look odd because on the Juno LITTLEs are CPUs 0,3-5 and bigs are CPUs 1-2)
If we make one of those SMP (IOW remove asymmetry) by e.g. hotplugging its big core, we would end up with an SMP cpuset and an asymmetric cpuset - the static key must remain set, because we still have one asymmetric root domain.
With the above example, this could be done with:
$ echo 0 > /sys/devices/system/cpu/cpu2/online
Which would result in:
asym0 asym1
[ ][ ]
L L B L L
When both SMP and asymmetric cpusets are present, all CPUs will observe sched_asym_cpucapacity being set (it is system-wide), but not all CPUs observe asymmetry in their sched domain hierarchy:
per_cpu(sd_asym_cpucapacity, <any CPU in asym0>) == <some SD at DIE level> per_cpu(sd_asym_cpucapacity, <any CPU in asym1>) == NULL
Change the simple key enablement to an increment, and decrement the key counter when destroying domains that cover asymmetric CPUs.
--- kernel/sched/topology.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)
kernel/sched/topology.c +5 -1
@@ -2008,7 +2008,7 @@ build_sched_domains(const struct cpumask *cpu_map, struct sched_domain_attr *attrcu_read_unlock ();if (has_asym) -static_branch_enable_cpuslocked (&sched_asym_cpucapacity); +static_branch_inc_cpuslocked (&sched_asym_cpucapacity);if (rq && sched_debug_enabled) { pr_info("root domain span: %*pbl (max cpu_capacity = %lu)\n",
@@ -2103,8 +2103,12 @@ int sched_init_domains(const struct cpumask *cpu_map)*/ static void detach_destroy_domains (const struct cpumask *cpu_map) { +unsigned int cpu =cpumask_any (cpu_map);int i; +if (rcu_access_pointer (per_cpu (sd_asym_cpucapacity, cpu))) +static_branch_dec_cpuslocked (&sched_asym_cpucapacity); +rcu_read_lock ();for_each_cpu (i, cpu_map)cpu_attach_domain (NULL , &def_root_domain, i);
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 156/193 nvme-multipath: fix possible io hang after ctrl reconnect
In reply to Greg Kroah-HartmanFrom: Anton Eidelman <[email protected]>
[ Upstream commit af8fd0424713a2adb812d10d55e86718152cf656 ]
The following scenario results in an IO hang:
1) ctrl completes a request with NVME_SC_ANA_TRANSITION. NVME_NS_ANA_PENDING bit in ns->flags is set and ana_work is triggered.
2) ana_work: nvme_read_ana_log() tries to get the ANA log page from the ctrl. This fails because ctrl disconnects.
Therefore nvme_update_ns_ana_state() is not called
and NVME_NS_ANA_PENDING bit in ns->flags is not cleared.
3) ctrl reconnects: nvme_mpath_init(ctrl,...) calls nvme_read_ana_log(ctrl, groups_only=true).
However, nvme_update_ana_state() does not update namespaces because nr_nsids = 0 (due to groups_only mode).
4) scan_work calls nvme_validate_ns() finds the ns and re-validates OK.
Result:
The ctrl is now live but NVME_NS_ANA_PENDING bit in ns->flags is still set. Consequently ctrl will never be considered a viable path by __nvme_find_path(). IO will hang if ctrl is the only or the last path to the namespace.
More generally, while ctrl is reconnecting, its ANA state may change. And because nvme_mpath_init() requests ANA log in groups_only mode, these changes are not propagated to the existing ctrl namespaces. This may result in a mal-function or an IO hang.
Solution:nvme_mpath_init() will nvme_read_ana_log() with groups_only set to false. This will not harm the new ctrl case (no namespaces present), and will make sure the ANA state of namespaces gets updated after reconnect.
Note: Another option would be for nvme_mpath_init() to invoke nvme_parse_ana_log(..., nvme_set_ns_ana_state) for each existing namespace.
--- drivers/nvme/host/multipath.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/nvme/host/multipath.c +1 -1
@@ -715,7 +715,7 @@ int nvme_mpath_init(struct nvme_ctrl *ctrl, struct nvme_id_ctrl *id) goto out; } - error =nvme_read_ana_log (ctrl, true); + error =nvme_read_ana_log (ctrl, false);if (error) goto out_free_ana_log_buf;return 0 ;
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 157/193 fjes: Handle workqueue allocation failure
In reply to Greg Kroah-HartmanFrom: Will Deacon <[email protected]>
[ Upstream commit 85ac30fa2e24f628e9f4f9344460f4015d33fd7d ]
In the highly unlikely event that we fail to allocate either of the "/txrx" or "/control" workqueues, we should bail cleanly rather than blindly march on with NULL queue pointer(s) installed in the 'fjes_adapter' instance.
--- drivers/net/fjes/fjes_main.c | 15 ++++++++++++++- 1 file changed, 14 insertions(+), 1 deletion(-)
drivers/net/fjes/fjes_main.c +14 -1
@@ -1237,8 +1237,17 @@ static int fjes_probe(struct platform_device *plat_dev) adapter->open_guard = false; adapter->txrx_wq =alloc_workqueue (DRV_NAME "/txrx" ,WQ_MEM_RECLAIM ,0 ); +if (unlikely (!adapter->txrx_wq)) { + err = -ENOMEM ; + goto err_free_netdev; + } + adapter->control_wq =alloc_workqueue (DRV_NAME "/control" ,WQ_MEM_RECLAIM ,0 ); +if (unlikely (!adapter->control_wq)) { + err = -ENOMEM ; + goto err_free_txrx_wq; + }INIT_WORK (&adapter->tx_stall_task, fjes_tx_stall_task); INIT_WORK(&adapter->raise_intr_rxdata_task,
@@ -1255,7 +1264,7 @@ static int fjes_probe(struct platform_device *plat_dev) hw->hw_res.irq =platform_get_irq (plat_dev,0 ); err =fjes_hw_init (&adapter->hw);if (err) - goto err_free_netdev; + goto err_free_control_wq;/* setup MAC address (02:00:00:00:00:[epid])*/ netdev->dev_addr[0 ] =2 ;
@@ -1277,6 +1286,10 @@ static int fjes_probe(struct platform_device *plat_dev) err_hw_exit:fjes_hw_exit (&adapter->hw); +err_free_control_wq: +destroy_workqueue (adapter->control_wq); +err_free_txrx_wq: +destroy_workqueue (adapter->txrx_wq); err_free_netdev:free_netdev (netdev); err_out:
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 158/193 net: hisilicon: Fix "Trying to free already-free IRQ"
In reply to Greg Kroah-HartmanFrom: Jiangfeng Xiao <[email protected]>
[ Upstream commit 63a41746827cb16dc6ad0d4d761ab4e7dda7a0c3 ]
When rmmod hip04_eth.ko, we can get the following warning:
Task track: rmmod(1623)>bash(1591)>login(1581)>init(1)
------------[ cut here ]------------ WARNING: CPU: 0 PID: 1623 at kernel/irq/manage.c:1557 __free_irq+0xa4/0x2ac() Trying to free already-free IRQ 200 Modules linked in: ping(O) pramdisk(O) cpuinfo(O) rtos_snapshot(O) interrupt_ctrl(O) mtdblock mtd_blkdevrtfs nfs_acl nfs lockd grace sunrpc xt_tcpudp ipt_REJECT iptable_filter ip_tables x_tables nf_reject_ipv CPU: 0 PID: 1623 Comm: rmmod Tainted: G O 4.4.193 #1 Hardware name: Hisilicon A15 [<c020b408>] (rtos_unwind_backtrace) from [<c0206624>] (show_stack+0x10/0x14) [<c0206624>] (show_stack) from [<c03f2be4>] (dump_stack+0xa0/0xd8) [<c03f2be4>] (dump_stack) from [<c021a780>] (warn_slowpath_common+0x84/0xb0) [<c021a780>] (warn_slowpath_common) from [<c021a7e8>] (warn_slowpath_fmt+0x3c/0x68) [<c021a7e8>] (warn_slowpath_fmt) from [<c026876c>] (__free_irq+0xa4/0x2ac) [<c026876c>] (__free_irq) from [<c0268a14>] (free_irq+0x60/0x7c) [<c0268a14>] (free_irq) from [<c0469e80>] (release_nodes+0x1c4/0x1ec) [<c0469e80>] (release_nodes) from [<c0466924>] (__device_release_driver+0xa8/0x104) [<c0466924>] (__device_release_driver) from [<c0466a80>] (driver_detach+0xd0/0xf8) [<c0466a80>] (driver_detach) from [<c0465e18>] (bus_remove_driver+0x64/0x8c) [<c0465e18>] (bus_remove_driver) from [<c02935b0>] (SyS_delete_module+0x198/0x1e0) [<c02935b0>] (SyS_delete_module) from [<c0202ed0>] (__sys_trace_return+0x0/0x10) ---[ end trace bb25d6123d849b44 ]---
Currently "rmmod hip04_eth.ko" call free_irq more than once as devres_release_all and hip04_remove both call free_irq. This results in a 'Trying to free already-free IRQ' warning. To solve the problem free_irq has been moved out of hip04_remove.
--- drivers/net/ethernet/hisilicon/hip04_eth.c | 1 - 1 file changed, 1 deletion(-)
drivers/net/ethernet/hisilicon/hip04_eth.c +0 -1
@@ -1041,7 +1041,6 @@ static int hip04_remove(struct platform_device *pdev)hip04_free_ring (ndev, d);unregister_netdev (ndev); -free_irq (ndev->irq, ndev);of_node_put (priv->phy_node);cancel_work_sync (&priv->tx_timeout_task);free_netdev (ndev);
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 159/193 wimax: i2400: Fix memory leak in i2400m_op_rfkill_sw_toggle
In reply to Greg Kroah-HartmanFrom: Navid Emamdoost <[email protected]>
[ Upstream commit 6f3ef5c25cc762687a7341c18cbea5af54461407 ]
In the implementation of i2400m_op_rfkill_sw_toggle() the allocated buffer for cmd should be released before returning. The documentation for i2400m_msg_to_dev() says when it returns the buffer can be reused. Meaning cmd should be released in either case. Move kfree(cmd) before return to be reached by all execution paths.
--- drivers/net/wimax/i2400m/op-rfkill.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/wimax/i2400m/op-rfkill.c +1 -1
@@ -127,12 +127,12 @@ int i2400m_op_rfkill_sw_toggle(struct wimax_dev *wimax_dev, "%d\n", result); result =0 ; error_cmd: -kfree (cmd);kfree_skb (ack_skb); error_msg_to_dev: error_alloc:d_fnend (4 , dev,"(wimax_dev %p state %d) = %d\n" , wimax_dev, state, result); +kfree (cmd);return result; }
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 160/193 net: mscc: ocelot: fix vlan_filtering when enslaving to bridge before link is up
In reply to Greg Kroah-HartmanFrom: Vladimir Oltean <[email protected]>
[ Upstream commit 1c44ce560b4de639f237b458be1729489ff44d0a ]
Background information: the driver operates the hardware in a mode where a single VLAN can be transmitted as untagged on a particular egress port. That is the "native VLAN on trunk port" use case. Its value is held in port->vid.
Consider the following command sequence (no network manager, all interfaces are down, debugging prints added by me):
$ ip link add dev br0 type bridge vlan_filtering 1 $ ip link set dev swp0 master br0
Kernel code path during last command:
br_add_slave -> ocelot_netdevice_port_event (NETDEV_CHANGEUPPER):
[ 21.401901] ocelot_vlan_port_apply: port 0 vlan aware 0 pvid 0 vid 0
br_add_slave -> nbp_vlan_init -> switchdev_port_attr_set -> ocelot_port_attr_set (SWITCHDEV_ATTR_ID_BRIDGE_VLAN_FILTERING):
[ 21.413335] ocelot_vlan_port_apply: port 0 vlan aware 1 pvid 0 vid 0
br_add_slave -> nbp_vlan_init -> nbp_vlan_add -> br_switchdev_port_vlan_add -> switchdev_port_obj_add -> ocelot_port_obj_add -> ocelot_vlan_vid_add
[ 21.667421] ocelot_vlan_port_apply: port 0 vlan aware 1 pvid 1 vid 1
So far so good. The bridge has replaced the driver's default pvid used in standalone mode (0) with its own default_pvid (1). The port's vid (native VLAN) has also changed from 0 to 1.
$ ip link set dev swp0 up
[ 31.722956] 8021q: adding VLAN 0 to HW filter on device swp0 do_setlink -> dev_change_flags -> vlan_vid_add -> ocelot_vlan_rx_add_vid -> ocelot_vlan_vid_add: [ 31.728700] ocelot_vlan_port_apply: port 0 vlan aware 1 pvid 1 vid 0
The 8021q module uses the .ndo_vlan_rx_add_vid API on .ndo_open to make ports be able to transmit and receive 802.1p-tagged traffic by default. This API is supposed to offload a VLAN sub-interface, which for a switch port means to add a VLAN that is not a pvid, and tagged on egress.
But the driver implementation of .ndo_vlan_rx_add_vid is wrong: it adds back vid 0 as "egress untagged". Now back to the initial paragraph:
there is a single untagged VID that the driver keeps track of, and that has just changed from 1 (the pvid) to 0. So this breaks the bridge core's expectation, because it has changed vid 1 from untagged to tagged, when what the user sees is.
$ bridge vlan port vlan ids swp0 1 PVID Egress Untagged
br0 1 PVID Egress Untagged
But curiously, instead of manifesting itself as "untagged and pvid-tagged traffic gets sent as tagged on egress", the bug:
- is hidden when vlan_filtering=0
- manifests as dropped traffic when vlan_filtering=1, due to this setting:
if (port->vlan_aware && !port->vid) /* If port is vlan-aware and tagged, drop untagged and priority * tagged frames. */ val |= ANA_PORT_DROP_CFG_DROP_UNTAGGED_ENA | ANA_PORT_DROP_CFG_DROP_PRIO_S_TAGGED_ENA | ANA_PORT_DROP_CFG_DROP_PRIO_C_TAGGED_ENA;
which would have made sense if it weren't for this bug. The setting's intention was "this is a trunk port with no native VLAN, so don't accept untagged traffic". So the driver was never expecting to set VLAN 0 as the value of the native VLAN, 0 was just encoding for "invalid".
So the fix is to not send 802.1p traffic as untagged, because that would change the port's native vlan to 0, unbeknownst to the bridge, and trigger unexpected code paths in the driver.
--- drivers/net/ethernet/mscc/ocelot.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/mscc/ocelot.c +1 -1
@@ -877,7 +877,7 @@ end:static int ocelot_vlan_rx_add_vid (struct net_device *dev,__be16 proto,u16 vid) { -return ocelot_vlan_vid_add (dev, vid, false, true); +return ocelot_vlan_vid_add (dev, vid, false, false); } static int ocelot_vlan_rx_kill_vid(struct net_device *dev, __be16 proto,
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 161/193 net: mscc: ocelot: refuse to overwrite the ports native vlan
In reply to Greg Kroah-HartmanFrom: Vladimir Oltean <[email protected]>
[ Upstream commit b9cd75e6689560140dadaed98eb4b41aad75d55d ]
The switch driver keeps a "vid" variable per port, which signifies _the_ VLAN ID that is stripped on that port's egress (aka the native VLAN on a trunk port).
That is the way the hardware is designed (mostly). The port->vid is programmed into REW:PORT:PORT_VLAN_CFG:PORT_VID and the rewriter is told to send all traffic as tagged except the one having port->vid.
There exists a possibility of finer-grained egress untagging decisions:
using the VCAP IS1 engine, one rule can be added to match every VLAN-tagged frame whose VLAN should be untagged, and set POP_CNT=1 as action. However, the IS1 can hold at most 512 entries, and the VLANs are in the order of 6 * 4096.
So the code is fine for now. But this sequence of commands:
$ bridge vlan add dev swp0 vid 1 pvid untagged $ bridge vlan add dev swp0 vid 2 untagged
makes untagged and pvid-tagged traffic be sent out of swp0 as tagged with VID 1, despite user's request.
Prevent that from happening. The user should temporarily remove the existing untagged VLAN (1 in this case), add it back as tagged, and then add the new untagged VLAN (2 in this case).
--- drivers/net/ethernet/mscc/ocelot.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-)
drivers/net/ethernet/mscc/ocelot.c +8 -1
@@ -260,8 +260,15 @@ static int ocelot_vlan_vid_add(struct net_device *dev, u16 vid, bool pvid, port->pvid = vid;/* Untagged egress vlan clasification */ -if (untagged) +if (untagged && port->vid != vid) { +if (port->vid) { +dev_err (ocelot->dev, +"Port already has a native VLAN: %d\n" , + port->vid); +return -EBUSY ; + } port->vid = vid; + }ocelot_vlan_port_apply (ocelot, port);
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 162/193 iommu/amd: Apply the same IVRS IOAPIC workaround to Acer Aspire A315-41
In reply to Greg Kroah-HartmanFrom: Takashi Iwai <[email protected]>
[ Upstream commit ad3e8da2d422c63c13819a53d3c5ea9312cc0b9d ]
Acer Aspire A315-41 requires the very same workaround as the existing quirk for Dell Latitude 5495. Add the new entry for that.
--- drivers/iommu/amd_iommu_quirks.c | 13 +++++++++++++ 1 file changed, 13 insertions(+)
drivers/iommu/amd_iommu_quirks.c +13 -0
@@ -73,6 +73,19 @@ static const struct dmi_system_id ivrs_quirks[] __initconst = { }, .driver_data = (void *)&ivrs_ioapic_quirks[DELL_LATITUDE_5495], }, + { +/* +* Acer Aspire A315-41 requires the very same workaround as +* Dell Latitude 5495 +*/ + .callback = ivrs_ioapic_quirk_cb, + .ident = "Acer Aspire A315-41", + .matches = { +DMI_MATCH (DMI_SYS_VENDOR ,"Acer" ), + DMI_MATCH(DMI_PRODUCT_NAME, "Aspire A315-41"), + }, + .driver_data = (void *)&ivrs_ioapic_quirks[DELL_LATITUDE_5495], + }, { .callback = ivrs_ioapic_quirk_cb, .ident = "Lenovo ideapad 330S-15ARR",
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 163/193 mt76: dma: fix buffer unmap with non-linear skbs
In reply to Greg Kroah-HartmanFrom: Lorenzo Bianconi <[email protected]>
[ Upstream commit 7bd0650be63cbb9e45e394d689c81365fe48e495 ]
mt76 dma layer is supposed to unmap skb data buffers while keep txwi mapped on hw dma ring. At the moment mt76 wrongly unmap txwi or does not unmap data fragments in even positions for non-linear skbs. This issue may result in hw hangs with A-MSDU if the system relies on IOMMU or SWIOTLB. Fix this behaviour properly unmapping data fragments on non-linear skbs.
--- drivers/net/wireless/mediatek/mt76/dma.c | 6 ++++-- drivers/net/wireless/mediatek/mt76/mt76.h | 5 +++-- 2 files changed, 7 insertions(+), 4 deletions(-)
drivers/net/wireless/mediatek/mt76/dma.c +4 -2
@@ -64,8 +64,10 @@ mt76_dma_add_buf(struct mt76_dev *dev, struct mt76_queue *q,u32 ctrl;int i, idx =-1 ; -if (txwi) +if (txwi) { q->entry[q->head].txwi =DMA_DUMMY_DATA ; + q->entry[q->head].skip_buf0 = true; + }for (i =0 ; i < nbufs; i +=2 , buf +=2 ) {u32 buf0 = buf[0 ].addr, buf1 =0 ;
@@ -108,7 +110,7 @@ mt76_dma_tx_cleanup_idx(struct mt76_dev *dev, struct mt76_queue *q, int idx,__le32 __ctrl =READ_ONCE (q->desc[idx].ctrl);u32 ctrl =le32_to_cpu (__ctrl); -if (!e->txwi || !e->skb) { +if (!e->skip_buf0) {__le32 addr =READ_ONCE (q->desc[idx].buf0);u32 len =FIELD_GET (MT_DMA_CTL_SD_LEN0 , ctrl);
drivers/net/wireless/mediatek/mt76/mt76.h +3 -2
@@ -102,8 +102,9 @@ struct mt76_queue_entry {struct urb *urb; };enum mt76_txq_id qid; -bool schedule; -bool done; + bool skip_buf0:1; + bool schedule:1; + bool done:1; };struct mt76_queue_regs {
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 164/193 drm/amdgpu/sdma5: do not execute 0-sized IBs (v2)
In reply to Greg Kroah-HartmanFrom: Pierre-Eric Pelloux-Prayer <[email protected]>
[ Upstream commit 9bdf63d3579e36942f4b91d3558a90da8116bb40 ]
This seems to help with https://bugs.freedesktop.org/show_bug.cgi?id=111481.
v2: insert a NOP instead of skipping all 0-sized IBs to avoid breaking older hw
--- drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c | 1 + 1 file changed, 1 insertion(+)
drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c +1 -0
@@ -264,6 +264,7 @@ static void gmc_v10_0_flush_gpu_tlb(struct amdgpu_device *adev, job->vm_pd_addr =amdgpu_gmc_pd_addr (adev->gart.bo); job->vm_needs_flush = true; + job->ibs->ptr[job->ibs->length_dw++] = ring->funcs->nop;amdgpu_ring_pad_ib (ring, &job->ibs[0 ]); r =amdgpu_job_submit (job, &adev->mman.entity,AMDGPU_FENCE_OWNER_UNDEFINED , &fence);
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 165/193 drm/sched: Set error to s_fence if HW job submission failed.
In reply to Greg Kroah-HartmanFrom: Andrey Grodzovsky <[email protected]>
[ Upstream commit 167bf96014a095753053595f3224fcdeb49ac3c8 ]
Problem:
When run_job fails and HW fence returned is NULL we still signal the s_fence to avoid hangs but the user has no way of knowing if the actual HW job was ran and finished.
Fix:
Allow .run_job implementations to return ERR_PTR in the fence pointer returned and then set this error for s_fence->finished fence so whoever wait on this fence can inspect the signaled fence for an error.
--- drivers/gpu/drm/scheduler/sched_main.c | 19 ++++++++++++++++--- 1 file changed, 16 insertions(+), 3 deletions(-)
drivers/gpu/drm/scheduler/sched_main.c +16 -3
@@ -478,6 +478,7 @@ void drm_sched_resubmit_jobs(struct drm_gpu_scheduler *sched)struct drm_sched_job *s_job, *tmp;uint64_t guilty_context;bool found_guilty = false; +struct dma_fence *fence;list_for_each_entry_safe (s_job, tmp, &sched->ring_mirror_list, node) {struct drm_sched_fence *s_fence = s_job->s_fence;
@@ -491,7 +492,16 @@ void drm_sched_resubmit_jobs(struct drm_gpu_scheduler *sched)dma_fence_set_error (&s_fence->finished, -ECANCELED );dma_fence_put (s_job->s_fence->parent); - s_job->s_fence->parent = sched->ops->run_job (s_job); + fence = sched->ops->run_job (s_job); + +if (IS_ERR_OR_NULL (fence)) { + s_job->s_fence->parent =NULL ; +dma_fence_set_error (&s_fence->finished,PTR_ERR (fence)); + }else { + s_job->s_fence->parent = fence; + } + + } }EXPORT_SYMBOL (drm_sched_resubmit_jobs);
@@ -719,7 +729,7 @@ static int drm_sched_main(void *param) fence = sched->ops->run_job (sched_job);drm_sched_fence_scheduled (s_fence); -if (fence) { +if (!IS_ERR_OR_NULL (fence)) { s_fence->parent =dma_fence_get (fence); r =dma_fence_add_callback (fence, &sched_job->cb, drm_sched_process_job);
@@ -729,8 +739,11 @@ static int drm_sched_main(void *param)DRM_ERROR ("fence add callback failed (%d)\n" , r);dma_fence_put (fence); - }else + }else { + +dma_fence_set_error (&s_fence->finished,PTR_ERR (fence));drm_sched_process_job (NULL , &sched_job->cb); + }wake_up (&sched->job_scheduled); }
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 166/193 drm/amdgpu: If amdgpu_ib_schedule fails return back the error.
In reply to Greg Kroah-HartmanFrom: Andrey Grodzovsky <[email protected]>
[ Upstream commit 57c0f58e9f562089de5f0b60da103677d232374c ]
Use ERR_PTR to return back the error happened during amdgpu_ib_schedule.
--- drivers/gpu/drm/amd/amdgpu/amdgpu_job.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/gpu/drm/amd/amdgpu/amdgpu_job.c +3 -1
@@ -218,7 +218,7 @@ static struct dma_fence *amdgpu_job_run(struct drm_sched_job *sched_job)struct amdgpu_ring *ring =to_amdgpu_ring (sched_job->sched);struct dma_fence *fence =NULL , *finished;struct amdgpu_job *job; -int r; +int r =0 ; job =to_amdgpu_job (sched_job); finished = &job->base.s_fence->finished;
@@ -243,6 +243,8 @@ static struct dma_fence *amdgpu_job_run(struct drm_sched_job *sched_job) job->fence =dma_fence_get (fence);amdgpu_job_free_resources (job); + + fence = r ?ERR_PTR (r) : fence;return fence; }
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 167/193 drm/amd/display: do not synchronize "drr" displays
In reply to Greg Kroah-HartmanFrom: Jun Lei <[email protected]>
[ Upstream commit 8775e89fa7121535d2da738c95167b8c65aa6e90 ]
[why]
A display that supports DRR can never really be considered "synchronized" with any other display because we can dynamically enable DRR (i.e. without modeset). this will cause their relative CRTC positions to drift and lose sync. this will disrupt features such as MCLK switching that assume and depend on their permanent alignment (that can only change with modeset)
[how]
check for ignore_msa in stream when considered synchronizability this ignore_msa is basically actually implemented as "supports drr"
--- drivers/gpu/drm/amd/display/dc/core/dc_resource.c | 6 ++++++ 1 file changed, 6 insertions(+)
drivers/gpu/drm/amd/display/dc/core/dc_resource.c +6 -0
@@ -394,6 +394,9 @@ bool resource_are_streams_timing_synchronizable(if (stream1->view_format != stream2->view_format)return false; +if (stream1->ignore_msa_timing_param || stream2->ignore_msa_timing_param) +return false; +return true; } static bool is_dp_and_hdmi_sharable(
@@ -1566,6 +1569,9 @@ bool dc_is_stream_unchanged(if (!are_stream_backends_same (old_stream, stream))return false; +if (old_stream->ignore_msa_timing_param != stream->ignore_msa_timing_param) +return false; +return true; }
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 168/193 drm/amd/display: add 50us buffer as WA for pstate switch in active
In reply to Greg Kroah-HartmanFrom: Jun Lei <[email protected]>
[ Upstream commit 7c37d399c2b84d4b79de4d512a38373f1d71ab90 ]
--- drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/gpu/drm/amd/display/dc/dml/dcn20/display_mode_vba_20.c +2 -1
@@ -2577,7 +2577,8 @@ static void dml20_DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPer mode_lib->vba.MinActiveDRAMClockChangeMargin + mode_lib->vba.DRAMClockChangeLatency; -if (mode_lib->vba.MinActiveDRAMClockChangeMargin >0 ) { +if (mode_lib->vba.MinActiveDRAMClockChangeMargin >50 ) { + mode_lib->vba.DRAMClockChangeWatermark +=25 ; mode_lib->vba.DRAMClockChangeSupport[0 ][0 ] = dm_dram_clock_change_vactive; }else {if (mode_lib->vba.SynchronizedVBlank || mode_lib->vba.NumberOfActivePlanes ==1 ) {
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 169/193 drm/amd/display: Passive DP->HDMI dongle detection fix
In reply to Greg Kroah-HartmanFrom: Michael Strauss <[email protected]>
[ Upstream commit bc2fde42e2418808dbfc04de1a6da91d7d31cf1a ]
[WHY]i2c_read is called to differentiate passive DP->HDMI and DP->DVI-D dongles The call is expected to fail in DVI-D case but pass in HDMI case Some HDMI dongles have a chance to fail as well, causing misdetection as DVI-D
[HOW]
Retry i2c_read to ensure failed result is valid
--- .../gpu/drm/amd/display/dc/core/dc_link_ddc.c | 24 ++++++++++++++----- 1 file changed, 18 insertions(+), 6 deletions(-)
drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c +18 -6
@@ -374,6 +374,7 @@ void dal_ddc_service_i2c_query_dp_dual_mode_adaptor(enum display_dongle_type *dongle = &sink_cap->dongle_type;uint8_t type2_dongle_buf[DP_ADAPTOR_TYPE2_SIZE ];bool is_type2_dongle = false; +int retry_count =2 ;struct dp_hdmi_dongle_signature_data *dongle_signature;/* Assume we have no valid DP passive dongle connected */
@@ -386,13 +387,24 @@ void dal_ddc_service_i2c_query_dp_dual_mode_adaptor( DP_HDMI_DONGLE_ADDRESS, type2_dongle_buf, sizeof(type2_dongle_buf))) { - *dongle =DISPLAY_DONGLE_DP_DVI_DONGLE ; - sink_cap->max_hdmi_pixel_clock =DP_ADAPTOR_DVI_MAX_TMDS_CLK ; +/* Passive HDMI dongles can sometimes fail here without retrying*/ +while (retry_count >0 ) { +if (i2c_read (ddc, +DP_HDMI_DONGLE_ADDRESS , + type2_dongle_buf, +sizeof (type2_dongle_buf))) +break ; + retry_count--; + } +if (retry_count ==0 ) { + *dongle =DISPLAY_DONGLE_DP_DVI_DONGLE ; + sink_cap->max_hdmi_pixel_clock =DP_ADAPTOR_DVI_MAX_TMDS_CLK ; -CONN_DATA_DETECT (ddc->link, type2_dongle_buf,sizeof (type2_dongle_buf), -"DP-DVI passive dongle %dMhz: " , -DP_ADAPTOR_DVI_MAX_TMDS_CLK /1000 ); -return ; +CONN_DATA_DETECT (ddc->link, type2_dongle_buf,sizeof (type2_dongle_buf), +"DP-DVI passive dongle %dMhz: " , +DP_ADAPTOR_DVI_MAX_TMDS_CLK /1000 ); +return ; + } }/* Check if Type 2 dongle.*/
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 170/193 dc.c:use kzalloc without test
In reply to Greg Kroah-HartmanFrom: zhongshiqi <[email protected]>
[ Upstream commit 364593f3ee5fdefc6efd89475e1804c928b4e6ba ]
dc.c:583:null check is needed after using kzalloc function
--- drivers/gpu/drm/amd/display/dc/core/dc.c | 4 ++++ 1 file changed, 4 insertions(+)
drivers/gpu/drm/amd/display/dc/core/dc.c +4 -0
@@ -566,6 +566,10 @@ static bool construct(struct dc *dc,#ifdef CONFIG_DRM_AMD_DC_DCN2_0 // Allocate memory for the vm_helper dc->vm_helper =kzalloc (sizeof (struct vm_helper ),GFP_KERNEL ); +if (!dc->vm_helper) { +dm_error ("%s: failed to create dc->vm_helper\n" , __func__); + goto fail; + }#endif memcpy (&dc->bb_overrides, &init_params->bb_overrides,sizeof (dc->bb_overrides));
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 171/193 SUNRPC: The TCP back channel mustnt disappear while requests are outstanding
In reply to Greg Kroah-HartmanFrom: Trond Myklebust <[email protected]>
[ Upstream commit 875f0706accd6501c3209bb99df8573171fb5d75 ]
If there are TCP back channel requests being processed by the server threads, then we should hold a reference to the transport to ensure it doesn't get freed from underneath us.
--- net/sunrpc/backchannel_rqst.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)
net/sunrpc/backchannel_rqst.c +3 -2
@@ -307,8 +307,8 @@ void xprt_free_bc_rqst(struct rpc_rqst *req)*/ dprintk ("RPC: Last session removed req=%p\n" , req);xprt_free_allocation (req); -return ; } +xprt_put (xprt); }/*
@@ -339,7 +339,7 @@ found:spin_unlock (&xprt->bc_pa_lock);if (new) {if (req != new) -xprt_free_bc_rqst (new); +xprt_free_allocation (new);break ; }else if (req)break ;
@@ -368,6 +368,7 @@ void xprt_complete_bc_request(struct rpc_rqst *req, uint32_t copied)set_bit (RPC_BC_PA_IN_USE , &req->rq_bc_pa_state);dprintk ("RPC: add callback request to list\n" ); +xprt_get (xprt);spin_lock (&bc_serv->sv_cb_lock);list_add (&req->rq_bc_list, &bc_serv->sv_cb_list);wake_up (&bc_serv->sv_cb_waitq);
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 172/193 SUNRPC: The RDMA back channel mustnt disappear while requests are outstanding
In reply to Greg Kroah-HartmanFrom: Trond Myklebust <[email protected]>
[ Upstream commit 9edb455e6797bb50aa38ef71e62668966065ede8 ]
If there are RDMA back channel requests being processed by the server threads, then we should hold a reference to the transport to ensure it doesn't get freed from underneath us.
--- net/sunrpc/xprtrdma/backchannel.c | 2 ++ 1 file changed, 2 insertions(+)
net/sunrpc/xprtrdma/backchannel.c +2 -0
@@ -165,6 +165,7 @@ void xprt_rdma_bc_free_rqst(struct rpc_rqst *rqst)spin_lock (&xprt->bc_pa_lock);list_add_tail (&rqst->rq_bc_pa_list, &xprt->bc_pa_list);spin_unlock (&xprt->bc_pa_lock); +xprt_put (xprt); }static struct rpc_rqst *rpcrdma_bc_rqst_get (struct rpcrdma_xprt *r_xprt)
@@ -261,6 +262,7 @@ void rpcrdma_bc_receive_call(struct rpcrdma_xprt *r_xprt,/* Queue rqst for ULP's callback service */ bc_serv = xprt->bc_serv; +xprt_get (xprt);spin_lock (&bc_serv->sv_cb_lock);list_add (&rqst->rq_bc_list, &bc_serv->sv_cb_list);spin_unlock (&bc_serv->sv_cb_lock);
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 173/193 SUNRPC: Destroy the back channel when we destroy the host transport
In reply to Greg Kroah-HartmanFrom: Trond Myklebust <[email protected]>
[ Upstream commit 669996add4c92476e0f8d6b4cd2bb308d1939fd7 ]
When we're destroying the host transport mechanism, we should ensure that we do not leak memory by failing to release any back channel slots that might still exist.
--- include/linux/sunrpc/bc_xprt.h | 5 +++++ net/sunrpc/backchannel_rqst.c | 2 +- net/sunrpc/xprt.c | 5 +++++ 3 files changed, 11 insertions(+), 1 deletion(-)
include/linux/sunrpc/bc_xprt.h +5 -0
@@ -64,6 +64,11 @@ static inline int xprt_setup_backchannel(struct rpc_xprt *xprt,return 0 ; } +static inline void xprt_destroy_backchannel (struct rpc_xprt *xprt, +unsigned int max_reqs) +{ +} +static inline bool svc_is_backchannel (const struct svc_rqst *rqstp) {return false;
net/sunrpc/backchannel_rqst.c +1 -1
@@ -220,7 +220,7 @@ void xprt_destroy_bc(struct rpc_xprt *xprt, unsigned int max_reqs) goto out;spin_lock_bh (&xprt->bc_pa_lock); - xprt->bc_alloc_max -= max_reqs; + xprt->bc_alloc_max -=min (max_reqs, xprt->bc_alloc_max);list_for_each_entry_safe (req, tmp, &xprt->bc_pa_list, rq_bc_pa_list) {dprintk ("RPC: req=%p\n" , req);list_del (&req->rq_bc_pa_list);
net/sunrpc/xprt.c +5 -0
@@ -1935,6 +1935,11 @@ static void xprt_destroy_cb(struct work_struct *work)rpc_destroy_wait_queue (&xprt->sending);rpc_destroy_wait_queue (&xprt->backlog);kfree (xprt->servername); +/* +* Destroy any existing back channel +*/ +xprt_destroy_backchannel (xprt,UINT_MAX ); +/* * Tear down transport state and free the rpc_xprt */
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 174/193 hv_netvsc: Fix error handling in netvsc_attach()
In reply to Greg Kroah-HartmanFrom: Haiyang Zhang <[email protected]>
[ Upstream commit 719b85c336ed35565d0f3982269d6f684087bb00 ]
If rndis_filter_open() fails, we need to remove the rndis device created in earlier steps, before returning an error code. Otherwise, the retry of netvsc_attach() from its callers will fail and hang.
--- drivers/net/hyperv/netvsc_drv.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-)
drivers/net/hyperv/netvsc_drv.c +8 -1
@@ -982,7 +982,7 @@ static int netvsc_attach(struct net_device *ndev,if (netif_running (ndev)) { ret =rndis_filter_open (nvdev);if (ret) -return ret; + goto err; rdev = nvdev->extension;if (!rdev->link_state)
@@ -990,6 +990,13 @@ static int netvsc_attach(struct net_device *ndev, }return 0 ; + +err: +netif_device_detach (ndev); + +rndis_filter_device_remove (hdev, nvdev); + +return ret; } static int netvsc_set_channels(struct net_device *net,
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 175/193 efi/tpm: Return -EINVAL when determining tpm final events log size fails
In reply to Greg Kroah-HartmanFrom: Jerry Snitselaar <[email protected]>
[ Upstream commit 2bb6a81633cb47dcba4c9f75605cbe49e6b73d60 ]
Currently nothing checks the return value of efi_tpm_eventlog_init(), but in case that changes in the future make sure an error is returned when it fails to determine the tpm final events log size.
--- drivers/firmware/efi/tpm.c | 1 + 1 file changed, 1 insertion(+)
drivers/firmware/efi/tpm.c +1 -0
@@ -88,6 +88,7 @@ int __init efi_tpm_eventlog_init(void)if (tbl_size <0 ) {pr_err (FW_BUG "Failed to parse event in TPM Final Events Log\n" ); + ret = -EINVAL ; goto out_calc; }
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 176/193 efi: libstub/arm: Account for firmware reserved memory at the base of RAM
In reply to Greg Kroah-HartmanFrom: Ard Biesheuvel <[email protected]>
[ Upstream commit 41cd96fa149b29684ebd38759fefb07f9c7d5276 ]
The EFI stubloader for ARM starts out by allocating a 32 MB window at the base of RAM, in order to ensure that the decompressor (which blindly copies the uncompressed kernel into that window) does not overwrite other allocations that are made while running in the context of the EFI firmware.
In some cases, (e.g., U-Boot running on the Raspberry Pi 2), this is causing boot failures because this initial allocation conflicts with a page of reserved memory at the base of RAM that contains the SMP spin tables and other pieces of firmware data and which was put there by the bootloader under the assumption that the TEXT_OFFSET window right below the kernel is only used partially during early boot, and will be left alone once the memory reservations are processed and taken into account.
So let's permit reserved memory regions to exist in the region starting at the base of RAM, and ending at TEXT_OFFSET - 5 * PAGE_SIZE, which is the window below the kernel that is not touched by the early boot code.
--- drivers/firmware/efi/libstub/Makefile | 1 + drivers/firmware/efi/libstub/arm32-stub.c | 16 +++++++++++++--- 2 files changed, 14 insertions(+), 3 deletions(-)
drivers/firmware/efi/libstub/Makefile +1 -0
@@ -52,6 +52,7 @@ lib-$(CONFIG_EFI_ARMSTUB) += arm-stub.o fdt.o string.o random.o \ lib-$(CONFIG_ARM) += arm32-stub.o lib-$(CONFIG_ARM64) += arm64-stub.o +CFLAGS_arm32-stub.o := -DTEXT_OFFSET=$(TEXT_OFFSET) CFLAGS_arm64-stub.o := -DTEXT_OFFSET=$(TEXT_OFFSET) #
drivers/firmware/efi/libstub/arm32-stub.c +13 -3
@@ -195,6 +195,7 @@ efi_status_t handle_kernel_image(efi_system_table_t *sys_table,unsigned long dram_base, efi_loaded_image_t *image) { + unsigned long kernel_base;efi_status_t status;/*
@@ -204,9 +205,18 @@ efi_status_t handle_kernel_image(efi_system_table_t *sys_table,* loaded. These assumptions are made by the decompressor, * before any memory map is available. */ - dram_base =round_up (dram_base,SZ_128M ); + kernel_base =round_up (dram_base,SZ_128M ); - status =reserve_kernel_base (sys_table, dram_base, reserve_addr, +/* +* Note that some platforms (notably, the Raspberry Pi 2) put +* spin-tables and other pieces of firmware at the base of RAM, +* abusing the fact that the window of TEXT_OFFSET bytes at the +* base of the kernel image is only partially used at the moment. +* (Up to 5 pages are used for the swapper page tables) +*/ + kernel_base +=TEXT_OFFSET -5 *PAGE_SIZE ; + + status =reserve_kernel_base (sys_table, kernel_base, reserve_addr, reserve_size);if (status !=EFI_SUCCESS ) {pr_efi_err (sys_table,"Unable to allocate memory for uncompressed kernel.\n" );
@@ -220,7 +230,7 @@ efi_status_t handle_kernel_image(efi_system_table_t *sys_table, *image_size = image->image_size; status =efi_relocate_kernel (sys_table, image_addr, *image_size, *image_size, - dram_base +MAX_UNCOMP_KERNEL_SIZE ,0 ); + kernel_base +MAX_UNCOMP_KERNEL_SIZE ,0 );if (status !=EFI_SUCCESS ) {pr_efi_err (sys_table,"Failed to relocate kernel.\n" );efi_free (sys_table, *reserve_size, *reserve_addr);
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 177/193 x86, efi: Never relocate kernel below lowest acceptable address
In reply to Greg Kroah-HartmanFrom: Kairui Song <[email protected]>
[ Upstream commit 220dd7699c46d5940115bd797b01b2ab047c87b8 ]
Currently, kernel fails to boot on some HyperV VMs when using EFI. And it's a potential issue on all x86 platforms.
It's caused by broken kernel relocation on EFI systems, when below three conditions are met:
1. Kernel image is not loaded to the default address (LOAD_PHYSICAL_ADDR) by the loader.
2. There isn't enough room to contain the kernel, starting from the default load address (eg. something else occupied part the region).
3. In the memmap provided by EFI firmware, there is a memory region starts below LOAD_PHYSICAL_ADDR, and suitable for containing the kernel.
EFI stub will perform a kernel relocation when condition 1 is met. But due to condition 2, EFI stub can't relocate kernel to the preferred address, so it fallback to ask EFI firmware to alloc lowest usable memory region, got the low region mentioned in condition 3, and relocated kernel there.
It's incorrect to relocate the kernel below LOAD_PHYSICAL_ADDR. This is the lowest acceptable kernel relocation address.
The first thing goes wrong is in arch/x86/boot/compressed/head_64.S. Kernel decompression will force use LOAD_PHYSICAL_ADDR as the output address if kernel is located below it. Then the relocation before decompression, which move kernel to the end of the decompression buffer, will overwrite other memory region, as there is no enough memory there.
To fix it, just don't let EFI stub relocate the kernel to any address lower than lowest acceptable address.
[ ardb: introduce efi_low_alloc_above() to reduce the scope of the change ]
--- arch/x86/boot/compressed/eboot.c | 4 +++- drivers/firmware/efi/libstub/arm32-stub.c | 2 +- .../firmware/efi/libstub/efi-stub-helper.c | 24 ++++++++----------- include/linux/efi.h | 18 ++++++++++++-- 4 files changed, 30 insertions(+), 18 deletions(-)
arch/x86/boot/compressed/eboot.c +3 -1
@@ -13,6 +13,7 @@#include <asm/e820/types.h> #include <asm/setup.h> #include <asm/desc.h> +#include <asm/boot.h> #include "../string.h" #include "eboot.h"
@@ -813,7 +814,8 @@ efi_main(struct efi_config *c, struct boot_params *boot_params) status =efi_relocate_kernel (sys_table, &bzimage_addr, hdr->init_size, hdr->init_size, hdr->pref_address, - hdr->kernel_alignment); + hdr->kernel_alignment, +LOAD_PHYSICAL_ADDR );if (status !=EFI_SUCCESS ) {efi_printk (sys_table,"efi_relocate_kernel() failed!\n" ); goto fail;
drivers/firmware/efi/libstub/arm32-stub.c +1 -1
@@ -230,7 +230,7 @@ efi_status_t handle_kernel_image(efi_system_table_t *sys_table, *image_size = image->image_size; status =efi_relocate_kernel (sys_table, image_addr, *image_size, *image_size, - kernel_base +MAX_UNCOMP_KERNEL_SIZE ,0 ); + kernel_base +MAX_UNCOMP_KERNEL_SIZE ,0 ,0 );if (status !=EFI_SUCCESS ) {pr_efi_err (sys_table,"Failed to relocate kernel.\n" );efi_free (sys_table, *reserve_size, *reserve_addr);
drivers/firmware/efi/libstub/efi-stub-helper.c +10 -14
@@ -260,11 +260,11 @@ fail: }/* -* Allocate at the lowest possible address. +* Allocate at the lowest possible address that is not below 'min'. */ -efi_status_t efi_low_alloc (efi_system_table_t *sys_table_arg, -unsigned long size,unsigned long align, -unsigned long *addr) +efi_status_t efi_low_alloc_above (efi_system_table_t *sys_table_arg, +unsigned long size,unsigned long align, +unsigned long *addr,unsigned long min) {unsigned long map_size, desc_size, buff_size;efi_memory_desc_t *map;
@@ -311,13 +311,8 @@ efi_status_t efi_low_alloc(efi_system_table_t *sys_table_arg, start = desc->phys_addr; end = start + desc->num_pages *EFI_PAGE_SIZE ; -/* -* Don't allocate at 0x0. It will confuse code that -* checks pointers against NULL. Skip the first 8 -* bytes so we start at a nice even number. -*/ -if (start ==0x0 ) - start +=8 ; +if (start < min) + start = min; start =round_up (start, align);if ((start + size) > end)
@@ -698,7 +693,8 @@ efi_status_t efi_relocate_kernel(efi_system_table_t *sys_table_arg,unsigned long image_size, unsigned long alloc_size, unsigned long preferred_addr, - unsigned long alignment) + unsigned long alignment, + unsigned long min_addr) { unsigned long cur_image_addr;unsigned long new_addr =0 ;
@@ -731,8 +727,8 @@ efi_status_t efi_relocate_kernel(efi_system_table_t *sys_table_arg,* possible. */ if (status !=EFI_SUCCESS ) { - status =efi_low_alloc (sys_table_arg, alloc_size, alignment, - &new_addr); + status =efi_low_alloc_above (sys_table_arg, alloc_size, + alignment, &new_addr, min_addr); }if (status !=EFI_SUCCESS ) {pr_efi_err (sys_table_arg,"Failed to allocate usable memory for kernel.\n" );
include/linux/efi.h +16 -2
@@ -1585,9 +1585,22 @@ char *efi_convert_cmdline(efi_system_table_t *sys_table_arg,efi_status_t efi_get_memory_map (efi_system_table_t *sys_table_arg,struct efi_boot_memmap *map); +efi_status_t efi_low_alloc_above (efi_system_table_t *sys_table_arg, +unsigned long size,unsigned long align, +unsigned long *addr,unsigned long min); + +static inline efi_status_t efi_low_alloc (efi_system_table_t *sys_table_arg,unsigned long size,unsigned long align, -unsigned long *addr); +unsigned long *addr) +{ +/* +* Don't allocate at 0x0. It will confuse code that +* checks pointers against NULL. Skip the first 8 +* bytes so we start at a nice even number. +*/ +return efi_low_alloc_above (sys_table_arg, size, align, addr,0x8 ); +} efi_status_t efi_high_alloc(efi_system_table_t *sys_table_arg, unsigned long size, unsigned long align,
@@ -1598,7 +1611,8 @@ efi_status_t efi_relocate_kernel(efi_system_table_t *sys_table_arg,unsigned long image_size, unsigned long alloc_size, unsigned long preferred_addr, - unsigned long alignment); + unsigned long alignment, + unsigned long min_addr); efi_status_t handle_cmdline_files(efi_system_table_t *sys_table_arg, efi_loaded_image_t *image,
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 178/193 arm64: cpufeature: Enable Qualcomm Falkor errata 1009 for Kryo
In reply to Greg Kroah-HartmanFrom: Bjorn Andersson <[email protected]>
[ Upstream commit 36c602dcdd872e9f9b91aae5266b6d7d72b69b96 ]
The Kryo cores share errata 1009 with Falkor, so add their model definitions and enable it for them as well.
--- Documentation/arm64/silicon-errata.rst | 2 +- arch/arm64/kernel/cpu_errata.c | 20 ++++++++++++++------ 2 files changed, 15 insertions(+), 7 deletions(-)
Documentation/arm64/silicon-errata.rst +1 -1
@@ -124,7 +124,7 @@ stable kernels. +----------------+-----------------+-----------------+-----------------------------+ | Qualcomm Tech. | Kryo/Falkor v1 | E1003 | QCOM_FALKOR_ERRATUM_1003 | +----------------+-----------------+-----------------+-----------------------------+ -| Qualcomm Tech. | Falkor v1 | E1009 | QCOM_FALKOR_ERRATUM_1009 | +| Qualcomm Tech. | Kryo/Falkor v1 | E1009 | QCOM_FALKOR_ERRATUM_1009 | +----------------+-----------------+-----------------+-----------------------------+ | Qualcomm Tech. | QDF2400 ITS | E0065 | QCOM_QDF2400_ERRATUM_0065 | +----------------+-----------------+-----------------+-----------------------------+
arch/arm64/kernel/cpu_errata.c +14 -6
@@ -659,17 +659,23 @@ static const struct midr_range arm64_harden_el2_vectors[] = {#endif #ifdef CONFIG_ARM64_WORKAROUND_REPEAT_TLBI - -static const struct midr_range arm64_repeat_tlbi_cpus[] = { +static const struct arm64_cpu_capabilities arm64_repeat_tlbi_list[] = { #ifdef CONFIG_QCOM_FALKOR_ERRATUM_1009 -MIDR_RANGE (MIDR_QCOM_FALKOR_V1 ,0 ,0 ,0 ,0 ), + { +ERRATA_MIDR_REV (MIDR_QCOM_FALKOR_V1 ,0 ,0 ) + }, + { + .midr_range.model =MIDR_QCOM_KRYO , + .matches = is_kryo_midr, + }, #endif #ifdef CONFIG_ARM64_ERRATUM_1286807 -MIDR_RANGE (MIDR_CORTEX_A76 ,0 ,0 ,3 ,0 ), + { +ERRATA_MIDR_RANGE (MIDR_CORTEX_A76 ,0 ,0 ,3 ,0 ), + }, #endif {}, }; -#endif #ifdef CONFIG_CAVIUM_ERRATUM_27456
@@ -825,7 +831,9 @@ const struct arm64_cpu_capabilities arm64_errata[] = { { .desc = "Qualcomm erratum 1009, ARM erratum 1286807", .capability = ARM64_WORKAROUND_REPEAT_TLBI, - ERRATA_MIDR_RANGE_LIST(arm64_repeat_tlbi_cpus), + .type = ARM64_CPUCAP_LOCAL_CPU_ERRATUM, + .matches = cpucap_multi_entry_cap_matches, + .match_list = arm64_repeat_tlbi_list, }, #endif #ifdef CONFIG_ARM64_ERRATUM_858921
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 179/193 usb: dwc3: gadget: fix race when disabling ep with cancelled xfers
In reply to Greg Kroah-HartmanFrom: Felipe Balbi <[email protected]>
[ Upstream commit d8eca64eec7103ab1fbabc0a187dbf6acfb2af93 ]
When disabling an endpoint which has cancelled requests, we should make sure to giveback requests that are currently pending in the cancelled list, otherwise we may fall into a situation where command completion interrupt fires after endpoint has been disabled, therefore causing a splat.
--- drivers/usb/dwc3/gadget.c | 6 ++++++ 1 file changed, 6 insertions(+)
drivers/usb/dwc3/gadget.c +6 -0
@@ -707,6 +707,12 @@ static void dwc3_remove_requests(struct dwc3 *dwc, struct dwc3_ep *dep)dwc3_gadget_giveback (dep, req, -ESHUTDOWN ); } + +while (!list_empty (&dep->cancelled_list)) { + req =next_request (&dep->cancelled_list); + +dwc3_gadget_giveback (dep, req, -ESHUTDOWN ); + } }/**
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 180/193 arm64: apply ARM64_ERRATUM_845719 workaround for Brahma-B53 core
In reply to Greg Kroah-HartmanFrom: Doug Berger <[email protected]>
[ Upstream commit bfc97f9f199cb041cf897af3af096540948cc705 ]
The Broadcom Brahma-B53 core is susceptible to the issue described by ARM64_ERRATUM_845719 so this commit enables the workaround to be applied when executing on that core.
Since there are now multiple entries to match, we must convert the existing ARM64_ERRATUM_845719 into an erratum list.
--- Documentation/arm64/silicon-errata.rst | 3 +++ arch/arm64/include/asm/cputype.h | 2 ++ arch/arm64/kernel/cpu_errata.c | 13 +++++++++++-- 3 files changed, 16 insertions(+), 2 deletions(-)
Documentation/arm64/silicon-errata.rst +3 -0
@@ -91,6 +91,9 @@ stable kernels. | ARM | MMU-500 | #841119,826419 | N/A | +----------------+-----------------+-----------------+-----------------------------+ +----------------+-----------------+-----------------+-----------------------------+ +| Broadcom | Brahma-B53 | N/A | ARM64_ERRATUM_845719 | ++----------------+-----------------+-----------------+-----------------------------+ ++----------------+-----------------+-----------------+-----------------------------+ | Cavium | ThunderX ITS | #22375,24313 | CAVIUM_ERRATUM_22375 | +----------------+-----------------+-----------------+-----------------------------+ | Cavium | ThunderX ITS | #23144 | CAVIUM_ERRATUM_23144 |
arch/arm64/include/asm/cputype.h +2 -0
@@ -79,6 +79,7 @@#define CAVIUM_CPU_PART_THUNDERX_83XX 0x0A3#define CAVIUM_CPU_PART_THUNDERX2 0x0AF +#define BRCM_CPU_PART_BRAHMA_B53 0x100#define BRCM_CPU_PART_VULCAN 0x516#define QCOM_CPU_PART_FALKOR_V1 0x800
@@ -105,6 +106,7 @@#define MIDR_THUNDERX_81XX MIDR_CPU_MODEL(ARM_CPU_IMP_CAVIUM, CAVIUM_CPU_PART_THUNDERX_81XX)#define MIDR_THUNDERX_83XX MIDR_CPU_MODEL(ARM_CPU_IMP_CAVIUM, CAVIUM_CPU_PART_THUNDERX_83XX)#define MIDR_CAVIUM_THUNDERX2 MIDR_CPU_MODEL(ARM_CPU_IMP_CAVIUM, CAVIUM_CPU_PART_THUNDERX2) +#define MIDR_BRAHMA_B53 MIDR_CPU_MODEL(ARM_CPU_IMP_BRCM, BRCM_CPU_PART_BRAHMA_B53)#define MIDR_BRCM_VULCAN MIDR_CPU_MODEL(ARM_CPU_IMP_BRCM, BRCM_CPU_PART_VULCAN)#define MIDR_QCOM_FALKOR_V1 MIDR_CPU_MODEL(ARM_CPU_IMP_QCOM, QCOM_CPU_PART_FALKOR_V1)#define MIDR_QCOM_FALKOR MIDR_CPU_MODEL(ARM_CPU_IMP_QCOM, QCOM_CPU_PART_FALKOR)
arch/arm64/kernel/cpu_errata.c +11 -2
@@ -743,6 +743,16 @@ static const struct midr_range erratum_1418040_list[] = { };#endif +#ifdef CONFIG_ARM64_ERRATUM_845719 +static const struct midr_range erratum_845719_list[] = { +/* Cortex-A53 r0p[01234] */ +MIDR_REV_RANGE (MIDR_CORTEX_A53 ,0 ,0 ,4 ), +/* Brahma-B53 r0p[0] */ +MIDR_REV (MIDR_BRAHMA_B53 ,0 ,0 ), + {}, +}; +#endif + const struct arm64_cpu_capabilities arm64_errata[] = {#ifdef CONFIG_ARM64_WORKAROUND_CLEAN_CACHE {
@@ -783,10 +793,9 @@ const struct arm64_cpu_capabilities arm64_errata[] = {#endif #ifdef CONFIG_ARM64_ERRATUM_845719 { - /* Cortex-A53 r0p[01234] */ .desc = "ARM erratum 845719", .capability = ARM64_WORKAROUND_845719, - ERRATA_MIDR_REV_RANGE(MIDR_CORTEX_A53, 0, 0, 4), + ERRATA_MIDR_RANGE_LIST(erratum_845719_list), }, #endif #ifdef CONFIG_CAVIUM_ERRATUM_23154
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 181/193 arm64: Brahma-B53 is SSB and spectre v2 safe
In reply to Greg Kroah-HartmanFrom: Florian Fainelli <[email protected]>
[ Upstream commit e059770cb1cdfbcbe3f1748f76005861cc79dd1a ]
Add the Brahma-B53 CPU (all versions) to the whitelists of CPUs for the SSB and spectre v2 mitigations.
--- arch/arm64/kernel/cpu_errata.c | 2 ++ 1 file changed, 2 insertions(+)
arch/arm64/kernel/cpu_errata.c +2 -0
@@ -489,6 +489,7 @@ static const struct midr_range arm64_ssb_cpus[] = {MIDR_ALL_VERSIONS (MIDR_CORTEX_A35 ),MIDR_ALL_VERSIONS (MIDR_CORTEX_A53 ),MIDR_ALL_VERSIONS (MIDR_CORTEX_A55 ), + MIDR_ALL_VERSIONS(MIDR_BRAHMA_B53), {}, };
@@ -573,6 +574,7 @@ static const struct midr_range spectre_v2_safe_list[] = {MIDR_ALL_VERSIONS (MIDR_CORTEX_A35 ),MIDR_ALL_VERSIONS (MIDR_CORTEX_A53 ),MIDR_ALL_VERSIONS (MIDR_CORTEX_A55 ), + MIDR_ALL_VERSIONS(MIDR_BRAHMA_B53), { /* sentinel */ } };
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 182/193 arm64: apply ARM64_ERRATUM_843419 workaround for Brahma-B53 core
In reply to Greg Kroah-HartmanFrom: Florian Fainelli <[email protected]>
[ Upstream commit 1cf45b8fdbb87040e1d1bd793891089f4678aa41 ]
The Broadcom Brahma-B53 core is susceptible to the issue described by ARM64_ERRATUM_843419 so this commit enables the workaround to be applied when executing on that core.
Since there are now multiple entries to match, we must convert the existing ARM64_ERRATUM_843419 into an erratum list and use cpucap_multi_entry_cap_matches to match our entries.
--- Documentation/arm64/silicon-errata.rst | 2 ++ arch/arm64/kernel/cpu_errata.c | 23 ++++++++++++++++++++--- 2 files changed, 22 insertions(+), 3 deletions(-)
Documentation/arm64/silicon-errata.rst +2 -0
@@ -93,6 +93,8 @@ stable kernels. +----------------+-----------------+-----------------+-----------------------------+ | Broadcom | Brahma-B53 | N/A | ARM64_ERRATUM_845719 | +----------------+-----------------+-----------------+-----------------------------+ +| Broadcom | Brahma-B53 | N/A | ARM64_ERRATUM_843419 | ++----------------+-----------------+-----------------+-----------------------------+ +----------------+-----------------+-----------------+-----------------------------+ | Cavium | ThunderX ITS | #22375,24313 | CAVIUM_ERRATUM_22375 | +----------------+-----------------+-----------------+-----------------------------+
arch/arm64/kernel/cpu_errata.c +20 -3
@@ -755,6 +755,23 @@ static const struct midr_range erratum_845719_list[] = { };#endif +#ifdef CONFIG_ARM64_ERRATUM_843419 +static const struct arm64_cpu_capabilities erratum_843419_list[] = { + { +/* Cortex-A53 r0p[01234] */ + .matches = is_affected_midr_range, +ERRATA_MIDR_REV_RANGE (MIDR_CORTEX_A53 ,0 ,0 ,4 ), +MIDR_FIXED (0x4 ,BIT (8 )), + }, + { +/* Brahma-B53 r0p[0] */ + .matches = is_affected_midr_range, +ERRATA_MIDR_REV (MIDR_BRAHMA_B53 ,0 ,0 ), + }, + {}, +}; +#endif + const struct arm64_cpu_capabilities arm64_errata[] = {#ifdef CONFIG_ARM64_WORKAROUND_CLEAN_CACHE {
@@ -786,11 +803,11 @@ const struct arm64_cpu_capabilities arm64_errata[] = {#endif #ifdef CONFIG_ARM64_ERRATUM_843419 { - /* Cortex-A53 r0p[01234] */ .desc = "ARM erratum 843419", .capability = ARM64_WORKAROUND_843419, - ERRATA_MIDR_REV_RANGE(MIDR_CORTEX_A53, 0, 0, 4), - MIDR_FIXED(0x4, BIT(8)), + .type = ARM64_CPUCAP_LOCAL_CPU_ERRATUM, + .matches = cpucap_multi_entry_cap_matches, + .match_list = erratum_843419_list, }, #endif #ifdef CONFIG_ARM64_ERRATUM_845719
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 183/193 NFSv4: Dont allow a cached open with a revoked delegation
In reply to Greg Kroah-HartmanFrom: Trond Myklebust <[email protected]>
[ Upstream commit be3df3dd4c70ee020587a943a31b98a0fb4b6424 ]
If the delegation is marked as being revoked, we must not use it for cached opens.
--- fs/nfs/delegation.c | 10 ++++++++++ fs/nfs/delegation.h | 1 + fs/nfs/nfs4proc.c | 7 ++----- 3 files changed, 13 insertions(+), 5 deletions(-)
fs/nfs/delegation.c +10 -0
@@ -53,6 +53,16 @@ nfs4_is_valid_delegation(const struct nfs_delegation *delegation,return false; } +struct nfs_delegation *nfs4_get_valid_delegation (const struct inode *inode) +{ +struct nfs_delegation *delegation; + + delegation =rcu_dereference (NFS_I (inode)->delegation); +if (nfs4_is_valid_delegation (delegation,0 )) +return delegation; +return NULL ; +} +static int nfs4_do_check_delegation (struct inode *inode,fmode_t flags,bool mark) {
fs/nfs/delegation.h +1 -0
@@ -68,6 +68,7 @@ int nfs4_lock_delegation_recall(struct file_lock *fl, struct nfs4_state *state,bool nfs4_copy_delegation_stateid (struct inode *inode,fmode_t flags,nfs4_stateid *dst,const struct cred **cred);bool nfs4_refresh_delegation_stateid (nfs4_stateid *dst,struct inode *inode); +struct nfs_delegation *nfs4_get_valid_delegation (const struct inode *inode);void nfs_mark_delegation_referenced (struct nfs_delegation *delegation);int nfs4_have_delegation (struct inode *inode,fmode_t flags);int nfs4_check_delegation (struct inode *inode,fmode_t flags);
fs/nfs/nfs4proc.c +2 -5
@@ -1435,8 +1435,6 @@ static int can_open_delegated(struct nfs_delegation *delegation, fmode_t fmode,return 0 ;if ((delegation->type & fmode) != fmode)return 0 ; -if (test_bit (NFS_DELEGATION_RETURNING , &delegation->flags)) -return 0 ;switch (claim) {case NFS4_OPEN_CLAIM_NULL :case NFS4_OPEN_CLAIM_FH :
@@ -1805,7 +1803,6 @@ static void nfs4_return_incompatible_delegation(struct inode *inode, fmode_t fmostatic struct nfs4_state *nfs4_try_open_cached (struct nfs4_opendata *opendata) {struct nfs4_state *state = opendata->state; -struct nfs_inode *nfsi =NFS_I (state->inode);struct nfs_delegation *delegation;int open_mode = opendata->o_arg.open_flags;fmode_t fmode = opendata->o_arg.fmode;
@@ -1822,7 +1819,7 @@ static struct nfs4_state *nfs4_try_open_cached(struct nfs4_opendata *opendata) }spin_unlock (&state->owner->so_lock);rcu_read_lock (); - delegation =rcu_dereference (nfsi->delegation); + delegation =nfs4_get_valid_delegation (state->inode);if (!can_open_delegated (delegation, fmode, claim)) {rcu_read_unlock ();break ;
@@ -2366,7 +2363,7 @@ static void nfs4_open_prepare(struct rpc_task *task, void *calldata) data->o_arg.open_flags, claim)) goto out_no_action;rcu_read_lock (); - delegation =rcu_dereference (NFS_I (data->state->inode)->delegation); + delegation =nfs4_get_valid_delegation (data->state->inode);if (can_open_delegated (delegation, data->o_arg.fmode, claim)) goto unlock_no_action;rcu_read_unlock ();
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 184/193 net: ethernet: arc: add the missed clk_disable_unprepare
In reply to Greg Kroah-HartmanFrom: Chuhong Yuan <[email protected]>
[ Upstream commit 4202e219edd6cc164c042e16fa327525410705ae ]
The remove misses to disable and unprepare priv->macclk like what is done when probe fails.
Add the missed call in remove.
--- drivers/net/ethernet/arc/emac_rockchip.c | 3 +++ 1 file changed, 3 insertions(+)
drivers/net/ethernet/arc/emac_rockchip.c +3 -0
@@ -256,6 +256,9 @@ static int emac_rockchip_remove(struct platform_device *pdev)if (priv->regulator)regulator_disable (priv->regulator); +if (priv->soc_data->need_div_macclk) +clk_disable_unprepare (priv->macclk); +free_netdev (ndev);return err; }
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 185/193 igb: Fix constant media auto sense switching when no cable is connected
In reply to Greg Kroah-HartmanFrom: Manfred Rudigier <[email protected]>
[ Upstream commit 8d5cfd7f76a2414e23c74bb8858af7540365d985 ]
At least on the i350 there is an annoying behavior that is maybe also present on 82580 devices, but was probably not noticed yet as MAS is not widely used.
If no cable is connected on both fiber/copper ports the media auto sense code will constantly swap between them as part of the watchdog task and produce many unnecessary kernel log messages.
The swap code responsible for this behavior (switching to fiber) should not be executed if the current media type is copper and there is no signal detected on the fiber port. In this case we can safely wait until the AUTOSENSE_EN bit is cleared.
--- drivers/net/ethernet/intel/igb/igb_main.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/net/ethernet/intel/igb/igb_main.c +2 -1
@@ -2064,7 +2064,8 @@ static void igb_check_swap_media(struct igb_adapter *adapter)if ((hw->phy.media_type == e1000_media_type_copper) && (!(connsw &E1000_CONNSW_AUTOSENSE_EN ))) { swap_now = true; - }else if (!(connsw &E1000_CONNSW_SERDESD )) { + }else if ((hw->phy.media_type != e1000_media_type_copper) && + !(connsw &E1000_CONNSW_SERDESD )) {/* copper signal takes time to appear */ if (adapter->copper_tries <4 ) { adapter->copper_tries++;
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 186/193 e1000: fix memory leaks
In reply to Greg Kroah-HartmanFrom: Wenwen Wang <[email protected]>
[ Upstream commit 8472ba62154058b64ebb83d5f57259a352d28697 ]
In e1000_set_ringparam(), 'tx_old' and 'rx_old' are not deallocated if e1000_up() fails, leading to memory leaks. Refactor the code to fix this issue.
--- drivers/net/ethernet/intel/e1000/e1000_ethtool.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-)
drivers/net/ethernet/intel/e1000/e1000_ethtool.c +3 -4
@@ -607,6 +607,7 @@ static int e1000_set_ringparam(struct net_device *netdev,for (i =0 ; i < adapter->num_rx_queues; i++) rxdr[i].count = rxdr->count; + err =0 ;if (netif_running (adapter->netdev)) {/* Try to get new resources before deleting old */ err =e1000_setup_all_rx_resources (adapter);
@@ -627,14 +628,13 @@ static int e1000_set_ringparam(struct net_device *netdev, adapter->rx_ring = rxdr; adapter->tx_ring = txdr; err =e1000_up (adapter); -if (err) - goto err_setup; }kfree (tx_old);kfree (rx_old);clear_bit (__E1000_RESETTING, &adapter->flags); -return 0 ; +return err; + err_setup_tx:e1000_free_all_rx_resources (adapter); err_setup_rx:
@@ -646,7 +646,6 @@ err_alloc_rx: err_alloc_tx:if (netif_running (adapter->netdev))e1000_up (adapter); -err_setup:clear_bit (__E1000_RESETTING, &adapter->flags);return err; }
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 187/193 gve: Fixes DMA synchronization.
In reply to Greg Kroah-HartmanFrom: Yangchun Fu <[email protected]>
[ Upstream commit 9cfeeb576d49a7b5e643b8066ba64a55e8417c5d ]
Synces the DMA buffer properly in order for CPU and device to see the most up-to-data data.
--- drivers/net/ethernet/google/gve/gve_rx.c | 2 ++ drivers/net/ethernet/google/gve/gve_tx.c | 24 ++++++++++++++++++++++-- 2 files changed, 24 insertions(+), 2 deletions(-)
drivers/net/ethernet/google/gve/gve_rx.c +2 -0
@@ -289,6 +289,8 @@ static bool gve_rx(struct gve_rx_ring *rx, struct gve_rx_desc *rx_desc, len =be16_to_cpu (rx_desc->len) -GVE_RX_PAD ; page_info = &rx->data.page_info[idx]; +dma_sync_single_for_cpu (&priv->pdev->dev, rx->data.qpl->page_buses[idx], +PAGE_SIZE ,DMA_FROM_DEVICE );/* gvnic can only receive into registered segments. If the buffer * can't be recycled, our only choice is to copy the data out of
drivers/net/ethernet/google/gve/gve_tx.c +22 -2
@@ -390,7 +390,21 @@ static void gve_tx_fill_seg_desc(union gve_tx_desc *seg_desc, seg_desc->seg.seg_addr =cpu_to_be64 (addr); } -static int gve_tx_add_skb (struct gve_tx_ring *tx,struct sk_buff *skb) +static void gve_dma_sync_for_device (struct device *dev,dma_addr_t *page_buses, +u64 iov_offset,u64 iov_len) +{ +dma_addr_t dma; +u64 addr; + +for (addr = iov_offset; addr < iov_offset + iov_len; + addr +=PAGE_SIZE ) { + dma = page_buses[addr /PAGE_SIZE ]; +dma_sync_single_for_device (dev, dma,PAGE_SIZE ,DMA_TO_DEVICE ); + } +} + +static int gve_tx_add_skb (struct gve_tx_ring *tx,struct sk_buff *skb, +struct device *dev) {int pad_bytes, hlen, hdr_nfrags, payload_nfrags, l4_hdr_offset;union gve_tx_desc *pkt_desc, *seg_desc;
@@ -432,6 +446,9 @@ static int gve_tx_add_skb(struct gve_tx_ring *tx, struct sk_buff *skb)skb_copy_bits (skb,0 , tx->tx_fifo.base + info->iov[hdr_nfrags -1 ].iov_offset, hlen); +gve_dma_sync_for_device (dev, tx->tx_fifo.qpl->page_buses, + info->iov[hdr_nfrags -1 ].iov_offset, + info->iov[hdr_nfrags -1 ].iov_len); copy_offset = hlen;for (i = payload_iov; i < payload_nfrags + payload_iov; i++) {
@@ -445,6 +462,9 @@ static int gve_tx_add_skb(struct gve_tx_ring *tx, struct sk_buff *skb)skb_copy_bits (skb, copy_offset, tx->tx_fifo.base + info->iov[i].iov_offset, info->iov[i].iov_len); +gve_dma_sync_for_device (dev, tx->tx_fifo.qpl->page_buses, + info->iov[i].iov_offset, + info->iov[i].iov_len); copy_offset += info->iov[i].iov_len; }
@@ -473,7 +493,7 @@ netdev_tx_t gve_tx(struct sk_buff *skb, struct net_device *dev)gve_tx_put_doorbell (priv, tx->q_resources, tx->req);return NETDEV_TX_BUSY ; } - nsegs =gve_tx_add_skb (tx, skb); + nsegs =gve_tx_add_skb (tx, skb, &priv->pdev->dev);netdev_tx_sent_queue (tx->netdev_txq, skb->len);skb_tx_timestamp (skb);
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 188/193 ocfs2: protect extent tree in ocfs2_prepare_inode_for_write()
In reply to Greg Kroah-HartmanFrom: Shuning Zhang <[email protected]>
[ Upstream commit e74540b285569d2b1e14fe7aee92297078f235ce ]
When the extent tree is modified, it should be protected by inode cluster lock and ip_alloc_sem.
The extent tree is accessed and modified in the ocfs2_prepare_inode_for_write, but isn't protected by ip_alloc_sem.
The following is a case. The function ocfs2_fiemap is accessing the extent tree, which is modified at the same time.
kernel BUG at fs/ocfs2/extent_map.c:475!
invalid opcode: 0000 [#1] SMP
Modules linked in: tun ocfs2 ocfs2_nodemanager configfs ocfs2_stackglue [...]
CPU: 16 PID: 14047 Comm: o2info Not tainted 4.1.12-124.23.1.el6uek.x86_64 #2
Hardware name: Oracle Corporation ORACLE SERVER X7-2L/ASM, MB MECH, X7-2L, BIOS 42040600 10/19/2018
task: ffff88019487e200 ti: ffff88003daa4000 task.ti: ffff88003daa4000
RIP: ocfs2_get_clusters_nocache.isra.11+0x390/0x550 [ocfs2]
Call Trace:
ocfs2_fiemap+0x1e3/0x430 [ocfs2]
do_vfs_ioctl+0x155/0x510
SyS_ioctl+0x81/0xa0
system_call_fastpath+0x18/0xd8
Code: 18 48 c7 c6 60 7f 65 a0 31 c0 bb e2 ff ff ff 48 8b 4a 40 48 8b 7a 28 48 c7 c2 78 2d 66 a0 e8 38 4f 05 00 e9 28 fe ff ff 0f 1f 00 <0f> 0b 66 0f 1f 44 00 00 bb 86 ff ff ff e9 13 fe ff ff 66 0f 1f
RIP ocfs2_get_clusters_nocache.isra.11+0x390/0x550 [ocfs2]
---[ end trace c8aa0c8180e869dc ]---
Kernel panic - not syncing: Fatal exception
Kernel Offset: disabledThis issue can be reproduced every week in a production environment.
This issue is related to the usage mode. If others use ocfs2 in this mode, the kernel will panic frequently.
--- fs/ocfs2/file.c | 134 ++++++++++++++++++++++++++++++++---------------- 1 file changed, 90 insertions(+), 44 deletions(-)
fs/ocfs2/file.c +90 -44
@@ -2092,54 +2092,90 @@ static int ocfs2_is_io_unaligned(struct inode *inode, size_t count, loff_t pos)return 0 ; } -static int ocfs2_prepare_inode_for_refcount (struct inode *inode, -struct file *file, -loff_t pos,size_t count, -int *meta_level) +static int ocfs2_inode_lock_for_extent_tree (struct inode *inode, +struct buffer_head **di_bh, +int meta_level, +int overwrite_io, +int write_sem, +int wait) { -int ret; -struct buffer_head *di_bh =NULL ; -u32 cpos = pos >>OCFS2_SB (inode->i_sb)->s_clustersize_bits; -u32 clusters = -ocfs2_clusters_for_bytes (inode->i_sb, pos + count) - cpos; +int ret =0 ; - ret =ocfs2_inode_lock (inode, &di_bh,1 ); -if (ret) { -mlog_errno (ret); +if (wait) + ret =ocfs2_inode_lock (inode,NULL , meta_level); +else + ret =ocfs2_try_inode_lock (inode, + overwrite_io ?NULL : di_bh, meta_level); +if (ret <0 ) goto out; + +if (wait) { +if (write_sem) +down_write (&OCFS2_I (inode)->ip_alloc_sem); +else +down_read (&OCFS2_I (inode)->ip_alloc_sem); + }else { +if (write_sem) + ret =down_write_trylock (&OCFS2_I (inode)->ip_alloc_sem); +else + ret =down_read_trylock (&OCFS2_I (inode)->ip_alloc_sem); + +if (!ret) { + ret = -EAGAIN ; + goto out_unlock; + } } - *meta_level =1 ; +return ret; - ret =ocfs2_refcount_cow (inode, di_bh, cpos, clusters,UINT_MAX ); -if (ret) -mlog_errno (ret); +out_unlock: +brelse (*di_bh); +ocfs2_inode_unlock (inode, meta_level); out: -brelse (di_bh);return ret; } +static void ocfs2_inode_unlock_for_extent_tree (struct inode *inode, +struct buffer_head **di_bh, +int meta_level, +int write_sem) +{ +if (write_sem) +up_write (&OCFS2_I (inode)->ip_alloc_sem); +else +up_read (&OCFS2_I (inode)->ip_alloc_sem); + +brelse (*di_bh); + *di_bh =NULL ; + +if (meta_level >=0 ) +ocfs2_inode_unlock (inode, meta_level); +} +static int ocfs2_prepare_inode_for_write (struct file *file,loff_t pos,size_t count,int wait) {int ret =0 , meta_level =0 , overwrite_io =0 ; +int write_sem =0 ;struct dentry *dentry = file->f_path.dentry;struct inode *inode =d_inode (dentry);struct buffer_head *di_bh =NULL ;loff_t end; +u32 cpos; +u32 clusters;/* * We start with a read level meta lock and only jump to an ex * if we need to make modifications here. */ for (;;) { -if (wait) - ret =ocfs2_inode_lock (inode,NULL , meta_level); -else - ret =ocfs2_try_inode_lock (inode, - overwrite_io ?NULL : &di_bh, meta_level); + ret =ocfs2_inode_lock_for_extent_tree (inode, + &di_bh, + meta_level, + overwrite_io, + write_sem, + wait);if (ret <0 ) { - meta_level =-1 ;if (ret != -EAGAIN )mlog_errno (ret); goto out;
@@ -2151,15 +2187,8 @@ static int ocfs2_prepare_inode_for_write(struct file *file,*/ if (!wait && !overwrite_io) { overwrite_io =1 ; -if (!down_read_trylock (&OCFS2_I (inode)->ip_alloc_sem)) { - ret = -EAGAIN ; - goto out_unlock; - } ret =ocfs2_overwrite_io (inode, di_bh, pos, count); -brelse (di_bh); - di_bh =NULL ; -up_read (&OCFS2_I (inode)->ip_alloc_sem);if (ret <0 ) {if (ret != -EAGAIN )mlog_errno (ret);
@@ -2178,7 +2207,10 @@ static int ocfs2_prepare_inode_for_write(struct file *file,* set inode->i_size at the end of a write. */ if (should_remove_suid (dentry)) {if (meta_level ==0 ) { -ocfs2_inode_unlock (inode, meta_level); +ocfs2_inode_unlock_for_extent_tree (inode, + &di_bh, + meta_level, + write_sem); meta_level =1 ;continue ; }
@@ -2194,18 +2226,32 @@ static int ocfs2_prepare_inode_for_write(struct file *file, ret =ocfs2_check_range_for_refcount (inode, pos, count);if (ret ==1 ) { -ocfs2_inode_unlock (inode, meta_level); - meta_level =-1 ; - - ret =ocfs2_prepare_inode_for_refcount (inode, - file, - pos, - count, - &meta_level); +ocfs2_inode_unlock_for_extent_tree (inode, + &di_bh, + meta_level, + write_sem); + ret =ocfs2_inode_lock_for_extent_tree (inode, + &di_bh, + meta_level, + overwrite_io, +1 , + wait); + write_sem =1 ; +if (ret <0 ) { +if (ret != -EAGAIN ) +mlog_errno (ret); + goto out; + } + + cpos = pos >>OCFS2_SB (inode->i_sb)->s_clustersize_bits; + clusters = +ocfs2_clusters_for_bytes (inode->i_sb, pos + count) - cpos; + ret =ocfs2_refcount_cow (inode, di_bh, cpos, clusters,UINT_MAX ); }if (ret <0 ) { -mlog_errno (ret); +if (ret != -EAGAIN ) +mlog_errno (ret); goto out_unlock; }
@@ -2216,10 +2262,10 @@ out_unlock:trace_ocfs2_prepare_inode_for_write (OCFS2_I (inode)->ip_blkno, pos, count, wait); -brelse (di_bh); - -if (meta_level >=0 ) -ocfs2_inode_unlock (inode, meta_level); +ocfs2_inode_unlock_for_extent_tree (inode, + &di_bh, + meta_level, + write_sem); out:return ret;
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 189/193 pinctrl: cherryview: Fix irq_valid_mask calculation
In reply to Greg Kroah-HartmanFrom: Hans de Goede <[email protected]>
[ Upstream commit 63bdef6cd6941917c823b9cc9aa0219d19fcb716 ]
Commit 03c4749dd6c7 ("gpio / ACPI: Drop unnecessary ACPI GPIO to Linux GPIO translation") has made the cherryview gpio numbers sparse, to get a 1:1 mapping between ACPI pin numbers and gpio numbers in Linux.
This has greatly simplified things, but the code setting the irq_valid_mask was not updated for this, so the valid mask is still in the old "compressed" numbering with the gaps in the pin numbers skipped, which is wrong as irq_valid_mask needs to be expressed in gpio numbers.
This results in the following error on devices using pin 24 (0x0018) on the north GPIO controller as an ACPI event source:
[ 0.422452] cherryview-pinctrl INT33FF:01: Failed to translate GPIO to IRQ
This has been reported (by email) to be happening on a Caterpillar CAT T20 tablet and I've reproduced this myself on a Medion Akoya e2215t 2-in-1.
This commit uses the pin number instead of the compressed index into community->pins to clear the correct bits in irq_valid_mask for GPIOs using GPEs for interrupts, fixing these errors and in case of the Medion Akoya e2215t also fixing the LID switch not working.
--- drivers/pinctrl/intel/pinctrl-cherryview.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/pinctrl/intel/pinctrl-cherryview.c +1 -1
@@ -1584,7 +1584,7 @@ static int chv_gpio_probe(struct chv_pinctrl *pctrl, int irq) intsel >>=CHV_PADCTRL0_INTSEL_SHIFT ;if (need_valid_mask && intsel >= community->nirqs) -clear_bit (i, chip->irq.valid_mask); +clear_bit (desc->number, chip->irq.valid_mask); }/*
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 190/193 clk: imx8m: Use SYS_PLL1_800M as intermediate parent of CLK_ARM
In reply to Greg Kroah-HartmanFrom: Leonard Crestez <[email protected]>
[ Upstream commit b234fe9558615098d8d62516e7041ad7f99ebcea ]
During cpu frequency switching the main "CLK_ARM" is reparented to an intermediate "step" clock. On imx8mm and imx8mn the 24M oscillator is used for this purpose but it is extremely slow, increasing wakeup latencies to the point that i2c transactions can timeout and system becomes unresponsive.
Fix by switching the "step" clk to SYS_PLL1_800M, matching the behavior of imx8m cpufreq drivers in imx vendor tree.
This bug was not immediately apparent because upstream arm64 defconfig uses the "performance" governor by default so no cpufreq transitions happen.
--- drivers/clk/imx/clk-imx8mm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/clk/imx/clk-imx8mm.c +1 -1
@@ -666,7 +666,7 @@ static int __init imx8mm_clocks_init(struct device_node *ccm_node) clks[IMX8MM_CLK_A53_DIV ], clks[IMX8MM_CLK_A53_SRC ], clks[IMX8MM_ARM_PLL_OUT ], - clks[IMX8MM_CLK_24M]); + clks[IMX8MM_SYS_PLL1_800M]);imx_check_clocks (clks,ARRAY_SIZE (clks));
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 191/193 timekeeping/vsyscall: Update VDSO data unconditionally
In reply to Greg Kroah-HartmanFrom: Huacai Chen <[email protected]>
[ Upstream commit 52338415cf4d4064ae6b8dd972dadbda841da4fa ]
The update of the VDSO data is depending on __arch_use_vsyscall() returning True. This is a leftover from the attempt to map the features of various architectures 1:1 into generic code.
The usage of __arch_use_vsyscall() in the actual vsyscall implementations got dropped and replaced by the requirement for the architecture code to return U64_MAX if the global clocksource is not usable in the VDSO.
But the __arch_use_vsyscall() check in the update code stayed which causes the VDSO data to be stale or invalid when an architecture actually implements that function and returns False when the current clocksource is not usable in the VDSO.
As a consequence the VDSO implementations of clock_getres(), time(), clock_gettime(CLOCK_.*_COARSE) operate on invalid data and return bogus information.
Remove the __arch_use_vsyscall() check from the VDSO update function and update the VDSO data unconditionally.
[ tglx: Massaged changelog and removed the now useless implementations in
asm-generic/ARM64/MIPS ]
--- arch/arm64/include/asm/vdso/vsyscall.h | 7 ------- include/asm-generic/vdso/vsyscall.h | 7 ------- kernel/time/vsyscall.c | 9 +++------ 3 files changed, 3 insertions(+), 20 deletions(-)
arch/arm64/include/asm/vdso/vsyscall.h +0 -7
@@ -30,13 +30,6 @@ int __arm64_get_clock_mode(struct timekeeper *tk) }#define __arch_get_clock_mode __arm64_get_clock_mode -static __always_inline -int __arm64_use_vsyscall(struct vdso_data *vdata) -{ -return !vdata[CS_HRES_COARSE ].clock_mode; -} -#define __arch_use_vsyscall __arm64_use_vsyscall -static __always_inline void __arm64_update_vsyscall(struct vdso_data *vdata, struct timekeeper *tk) {
include/asm-generic/vdso/vsyscall.h +0 -7
@@ -25,13 +25,6 @@ static __always_inline int __arch_get_clock_mode(struct timekeeper *tk) }#endif /* __arch_get_clock_mode */ -#ifndef __arch_use_vsyscall -static __always_inline int __arch_use_vsyscall(struct vdso_data *vdata) -{ -return 1 ; -} -#endif /* __arch_use_vsyscall */ -#ifndef __arch_update_vsyscall static __always_inline void __arch_update_vsyscall(struct vdso_data *vdata,struct timekeeper *tk)
kernel/time/vsyscall.c +3 -6
@@ -110,8 +110,7 @@ void update_vsyscall(struct timekeeper *tk) nsec = nsec + tk->wall_to_monotonic.tv_nsec; vdso_ts->sec +=__iter_div_u64_rem (nsec,NSEC_PER_SEC , &vdso_ts->nsec); -if (__arch_use_vsyscall (vdata)) -update_vdso_data (vdata, tk); +update_vdso_data (vdata, tk);__arch_update_vsyscall (vdata, tk);
@@ -124,10 +123,8 @@ void update_vsyscall_tz(void) {struct vdso_data *vdata =__arch_get_k_vdso_data (); -if (__arch_use_vsyscall (vdata)) { - vdata[CS_HRES_COARSE ].tz_minuteswest = sys_tz.tz_minuteswest; - vdata[CS_HRES_COARSE ].tz_dsttime = sys_tz.tz_dsttime; - } + vdata[CS_HRES_COARSE ].tz_minuteswest = sys_tz.tz_minuteswest; + vdata[CS_HRES_COARSE ].tz_dsttime = sys_tz.tz_dsttime;__arch_sync_vdso_data (vdata); }
Signature · 1 line
2.20.1
Greg Kroah-Hartman
maintainerPATCH 192/193 mm/filemap.c: dont initiate writeback if mapping has no dirty pages
In reply to Greg Kroah-HartmanFrom: Konstantin Khlebnikov <[email protected]>
commit c3aab9a0bd91b696a852169479b7db1ece6cbf8c upstream.
Functions like filemap_write_and_wait_range() should do nothing if inode has no dirty pages or pages currently under writeback. But they anyway construct struct writeback_control and this does some atomic operations if CONFIG_CGROUP_WRITEBACK=y - on fast path it locks inode->i_lock and updates state of writeback ownership, on slow path might be more work. Current this path is safely avoided only when inode mapping has no pages.
For example generic_file_read_iter() calls filemap_write_and_wait_range() at each O_DIRECT read - pretty hot path.
This patch skips starting new writeback if mapping has no dirty tags set. If writeback is already in progress filemap_write_and_wait_range() will wait for it.
--- mm/filemap.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
mm/filemap.c +2 -1
@@ -408,7 +408,8 @@ int __filemap_fdatawrite_range(struct ad .range_end = end, }; -if (!mapping_cap_writeback_dirty (mapping)) +if (!mapping_cap_writeback_dirty (mapping) || + !mapping_tagged (mapping,PAGECACHE_TAG_DIRTY ))return 0 ;wbc_attach_fdatawrite_inode (&wbc, mapping->host);
Greg Kroah-Hartman
maintainerPATCH 193/193 cgroup,writeback: dont switch wbs immediately on dead wbs if the memcg is dead
In reply to Greg Kroah-HartmanFrom: Tejun Heo <[email protected]>
commit 65de03e251382306a4575b1779c57c87889eee49 upstream.
cgroup writeback tries to refresh the associated wb immediately if the current wb is dead. This is to avoid keeping issuing IOs on the stale wb after memcg - blkcg association has changed (ie. when blkcg got disabled / enabled higher up in the hierarchy).
Unfortunately, the logic gets triggered spuriously on inodes which are associated with dead cgroups. When the logic is triggered on dead cgroups, the attempt fails only after doing quite a bit of work allocating and initializing a new wb.
While c3aab9a0bd91 ("mm/filemap.c: don't initiate writeback if mapping has no dirty pages") alleviated the issue significantly as it now only triggers when the inode has dirty pages. However, the condition can still be triggered before the inode is switched to a different cgroup and the logic simply doesn't make sense.
Skip the immediate switching if the associated memcg is dying.
This is a simplified version of the following two patches:
* https://lore.kernel.org/linux-mm/20190513183053.GA73423@dennisz-mbp/
* http://lkml.kernel.org/r/156355839560.2063.5265687291430814589.stgit@buzz
--- fs/fs-writeback.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-)
fs/fs-writeback.c +6 -3
@@ -577,10 +577,13 @@ void wbc_attach_and_unlock_inode(structspin_unlock (&inode->i_lock);/* -* A dying wb indicates that the memcg-blkcg mapping has changed -* and a new wb is already serving the memcg. Switch immediately. +* A dying wb indicates that either the blkcg associated with the +* memcg changed or the associated memcg is dying. In the first +* case, a replacement wb should already be available and we should +* refresh the wb immediately. In the second case, trying to +* refresh will keep failing. */ -if (unlikely (wb_dying (wbc->wb))) +if (unlikely (wb_dying (wbc->wb) && !css_is_dying (wbc->wb->memcg_css)))inode_switch_wbs (inode, wbc->wb_id); }EXPORT_SYMBOL_GPL (wbc_attach_and_unlock_inode);
stable-rc/linux-5.3.y boot: 130 boots: 0 failed, 122 passed with 7 offline, 1 conflict (v5.3.10-194-gfeeefcbdbfc1)
stable-rc/linux-5.3.y boot: 130 boots: 0 failed, 122 passed with 7 offline, 1 conflict (v5.3.10-194-gfeeefcbdbfc1)
Full Boot Summary: https://kernelci.org/boot/all/job/stable-rc/branch/linux-5.3.y/kernel/v5.3.10-194-gfeeefcbdbfc1/
Full Build Summary: https://kernelci.org/build/stable-rc/branch/linux-5.3.y/kernel/v5.3.10-194-gfeeefcbdbfc1/
Tree: stable-rc
Branch: linux-5.3.y
Git Describe: v5.3.10-194-gfeeefcbdbfc1
Git Commit: feeefcbdbfc1362972ef26970aed0aafe90cc1ae
Git URL: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git
Tested: 76 unique boards, 24 SoC families, 16 builds out of 207
Offline Platforms:
arm:
sunxi_defconfig:
gcc-8
sun5i-r8-chip: 1 offline lab
sun7i-a20-bananapi: 1 offline lab
multi_v7_defconfig:
gcc-8
qcom-apq8064-cm-qs600: 1 offline lab
sun5i-r8-chip: 1 offline lab
sun7i-a20-bananapi: 1 offline lab
davinci_all_defconfig:
gcc-8
dm365evm,legacy: 1 offline lab
qcom_defconfig:
gcc-8
qcom-apq8064-cm-qs600: 1 offline labConflicting Boot Failure Detected: (These likely are not failures as other labs are reporting PASS. Needs review.)
i386:
i386_defconfig:
qemu_i386:
lab-collabora: PASS (gcc-8)
lab-baylibre: FAIL (gcc-8)
---
For more info write to <[email protected]>
Quoted 10 lines · Greg Kroah-Hartman
This is the start of the stable review cycle for the 5.3.11 release. There are 193 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 Wed, 13 Nov 2019 18:08:44 +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-5.3.11-rc1.gzor in the git tree and branch at:
git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.gitlinux-5.3.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.
Summary
------------------------------------------------------------------------
kernel: 5.3.11-rc1
git repo: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git
git branch: linux-5.3.y
git commit: feeefcbdbfc1362972ef26970aed0aafe90cc1ae
git describe: v5.3.10-194-gfeeefcbdbfc1
Test details: https://qa-reports.linaro.org/lkft/linux-stable-rc-5.3-oe/build/v5.3.10-194-gfeeefcbdbfc1
No regressions (compared to build v5.3.10)
No fixes (compared to build v5.3.10)
Ran 26052 total tests in the following environments and test suites.
Environments
--------------
- dragonboard-410c
- hi6220-hikey
- i386
- juno-r2
- qemu_arm
- qemu_arm64
- qemu_i386
- qemu_x86_64
- x15
- x86
Test Suites
-----------
* build
* install-android-platform-tools-r2600
* kselftest
* libgpiod
* libhugetlbfs
* ltp-cap_bounds-tests
* ltp-commands-tests
* ltp-containers-tests
* ltp-cpuhotplug-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-pty-tests
* ltp-sched-tests
* ltp-securebits-tests
* ltp-syscalls-tests
* perf
* spectre-meltdown-checker-test
* network-basic-tests
* v4l2-compliance
* ltp-open-posix-tests
* kvm-unit-tests
* ssuite
* kselftest-vsyscall-mode-native
* kselftest-vsyscall-mode-none
Signature · 2 lines
Linaro LKFT https://lkft.linaro.org
Greg Kroah-Hartman
maintainerQuoted 17 lines · Naresh Kamboju
Quoted 15 lines · Greg Kroah-Hartman
This is the start of the stable review cycle for the 5.3.11 release. There are 193 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 Wed, 13 Nov 2019 18:08:44 +0000. Anything received after that time might be too late.
The whole patch series can be found in one patch at:
https://www.kernel.org/pub/linux/kernel/v5.x/stable-review/patch-5.3.11-rc1.gzor in the git tree and branch at:
git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.gitlinux-5.3.y
and the diffstat can be found below.
Quoted 6 lines
Quoted 3 lines
thanks,
greg k-h
Results from Linaro’s test farm.
No regressions on arm64, arm,x86_64, and i386.
Thanks for testing all of these and letting me know.
greg k-h
Jon Hunter
maintainerQuoted 9 lines · Greg Kroah-Hartman
This is the start of the stable review cycle for the 5.3.11 release. There are 193 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 Wed, 13 Nov 2019 18:08:44 +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-5.3.11-rc1.gzor in the git tree and branch at:
git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.gitlinux-5.3.y
and the diffstat can be found below.thanks,
greg k-h
All tests for Tegra are passing ...
Test results for stable-v5.3:
12 builds: 12 pass, 0 fail
22 boots: 22 pass, 0 fail
38 tests: 38 pass, 0 failLinux version: 5.3.11-rc1-gfeeefcbdbfc1
Boards tested: tegra124-jetson-tk1, tegra186-p2771-0000,
tegra194-p2972-0000, tegra20-ventana,
tegra210-p2371-2180, tegra30-cardhu-a04Cheers
Jon
Signature · 1 line
nvpublic
Quoted 8 lines · Greg Kroah-Hartman
This is the start of the stable review cycle for the 5.3.11 release. There are 193 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 Wed, 13 Nov 2019 18:08:44 +0000. Anything received after that time might be too late.
Build results:
total: 158 pass: 158 fail: 0 Qemu test results: total: 391 pass: 391 fail: 0
Guenter
Greg Kroah-Hartman
maintainerQuoted 6 lines · Guenter Roeck
Quoted 5 lines · Greg Kroah-Hartman
This is the start of the stable review cycle for the 5.3.11 release. There are 193 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.
Quoted 8 lines
Quoted 3 lines
Responses should be made by Wed, 13 Nov 2019 18:08:44 +0000. Anything received after that time might be too late.
Build results:
total: 158 pass: 158 fail: 0 Qemu test results: total: 391 pass: 391 fail: 0
Thanks for testing all of these and letting me know.
greg k-h