From 23d11a58a9a60dcb52c8fc6494efce908b24c295 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Fri, 29 Jan 2016 05:59:46 -0500 Subject: workqueue: skip flush dependency checks for legacy workqueues fca839c00a12 ("workqueue: warn if memory reclaim tries to flush !WQ_MEM_RECLAIM workqueue") implemented flush dependency warning which triggers if a PF_MEMALLOC task or WQ_MEM_RECLAIM workqueue tries to flush a !WQ_MEM_RECLAIM workquee. This assumes that workqueues marked with WQ_MEM_RECLAIM sit in memory reclaim path and making it depend on something which may need more memory to make forward progress can lead to deadlocks. Unfortunately, workqueues created with the legacy create*_workqueue() interface always have WQ_MEM_RECLAIM regardless of whether they are depended upon memory reclaim or not. These spurious WQ_MEM_RECLAIM markings cause spurious triggering of the flush dependency checks. WARNING: CPU: 0 PID: 6 at kernel/workqueue.c:2361 check_flush_dependency+0x138/0x144() workqueue: WQ_MEM_RECLAIM deferwq:deferred_probe_work_func is flushing !WQ_MEM_RECLAIM events:lru_add_drain_per_cpu ... Workqueue: deferwq deferred_probe_work_func [] (unwind_backtrace) from [] (show_stack+0x10/0x14) [] (show_stack) from [] (dump_stack+0x94/0xd4) [] (dump_stack) from [] (warn_slowpath_common+0x80/0xb0) [] (warn_slowpath_common) from [] (warn_slowpath_fmt+0x30/0x40) [] (warn_slowpath_fmt) from [] (check_flush_dependency+0x138/0x144) [] (check_flush_dependency) from [] (flush_work+0x50/0x15c) [] (flush_work) from [] (lru_add_drain_all+0x130/0x180) [] (lru_add_drain_all) from [] (migrate_prep+0x8/0x10) [] (migrate_prep) from [] (alloc_contig_range+0xd8/0x338) [] (alloc_contig_range) from [] (cma_alloc+0xe0/0x1ac) [] (cma_alloc) from [] (__alloc_from_contiguous+0x38/0xd8) [] (__alloc_from_contiguous) from [] (__dma_alloc+0x240/0x278) [] (__dma_alloc) from [] (arm_dma_alloc+0x54/0x5c) [] (arm_dma_alloc) from [] (dmam_alloc_coherent+0xc0/0xec) [] (dmam_alloc_coherent) from [] (ahci_port_start+0x150/0x1dc) [] (ahci_port_start) from [] (ata_host_start.part.3+0xc8/0x1c8) [] (ata_host_start.part.3) from [] (ata_host_activate+0x50/0x148) [] (ata_host_activate) from [] (ahci_host_activate+0x44/0x114) [] (ahci_host_activate) from [] (ahci_platform_init_host+0x1d8/0x3c8) [] (ahci_platform_init_host) from [] (tegra_ahci_probe+0x448/0x4e8) [] (tegra_ahci_probe) from [] (platform_drv_probe+0x50/0xac) [] (platform_drv_probe) from [] (driver_probe_device+0x214/0x2c0) [] (driver_probe_device) from [] (bus_for_each_drv+0x60/0x94) [] (bus_for_each_drv) from [] (__device_attach+0xb0/0x114) [] (__device_attach) from [] (bus_probe_device+0x84/0x8c) [] (bus_probe_device) from [] (deferred_probe_work_func+0x68/0x98) [] (deferred_probe_work_func) from [] (process_one_work+0x120/0x3f8) [] (process_one_work) from [] (worker_thread+0x38/0x55c) [] (worker_thread) from [] (kthread+0xdc/0xf4) [] (kthread) from [] (ret_from_fork+0x14/0x3c) Fix it by marking workqueues created via create*_workqueue() with __WQ_LEGACY and disabling flush dependency checks on them. Signed-off-by: Tejun Heo Reported-and-tested-by: Thierry Reding Link: http://lkml.kernel.org/g/20160126173843.GA11115@ulmo.nvidia.com Fixes: fca839c00a12 ("workqueue: warn if memory reclaim tries to flush !WQ_MEM_RECLAIM workqueue") diff --git a/include/linux/workqueue.h b/include/linux/workqueue.h index 0e32bc7..ca73c50 100644 --- a/include/linux/workqueue.h +++ b/include/linux/workqueue.h @@ -311,6 +311,7 @@ enum { __WQ_DRAINING = 1 << 16, /* internal: workqueue is draining */ __WQ_ORDERED = 1 << 17, /* internal: workqueue is ordered */ + __WQ_LEGACY = 1 << 18, /* internal: create*_workqueue() */ WQ_MAX_ACTIVE = 512, /* I like 512, better ideas? */ WQ_MAX_UNBOUND_PER_CPU = 4, /* 4 * #cpus for unbound wq */ @@ -411,12 +412,12 @@ __alloc_workqueue_key(const char *fmt, unsigned int flags, int max_active, alloc_workqueue(fmt, WQ_UNBOUND | __WQ_ORDERED | (flags), 1, ##args) #define create_workqueue(name) \ - alloc_workqueue("%s", WQ_MEM_RECLAIM, 1, (name)) + alloc_workqueue("%s", __WQ_LEGACY | WQ_MEM_RECLAIM, 1, (name)) #define create_freezable_workqueue(name) \ - alloc_workqueue("%s", WQ_FREEZABLE | WQ_UNBOUND | WQ_MEM_RECLAIM, \ - 1, (name)) + alloc_workqueue("%s", __WQ_LEGACY | WQ_FREEZABLE | WQ_UNBOUND | \ + WQ_MEM_RECLAIM, 1, (name)) #define create_singlethread_workqueue(name) \ - alloc_ordered_workqueue("%s", WQ_MEM_RECLAIM, name) + alloc_ordered_workqueue("%s", __WQ_LEGACY | WQ_MEM_RECLAIM, name) extern void destroy_workqueue(struct workqueue_struct *wq); diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 61a0264..dc7faad 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -2355,7 +2355,8 @@ static void check_flush_dependency(struct workqueue_struct *target_wq, WARN_ONCE(current->flags & PF_MEMALLOC, "workqueue: PF_MEMALLOC task %d(%s) is flushing !WQ_MEM_RECLAIM %s:%pf", current->pid, current->comm, target_wq->name, target_func); - WARN_ONCE(worker && (worker->current_pwq->wq->flags & WQ_MEM_RECLAIM), + WARN_ONCE(worker && ((worker->current_pwq->wq->flags & + (WQ_MEM_RECLAIM | __WQ_LEGACY)) == WQ_MEM_RECLAIM), "workqueue: WQ_MEM_RECLAIM %s:%pf is flushing !WQ_MEM_RECLAIM %s:%pf", worker->current_pwq->wq->name, worker->current_func, target_wq->name, target_func); -- cgit v0.10.2 From 041bd12e272c53a35c54c13875839bcb98c999ce Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Tue, 9 Feb 2016 16:11:26 -0500 Subject: Revert "workqueue: make sure delayed work run in local cpu" This reverts commit 874bbfe600a660cba9c776b3957b1ce393151b76. Workqueue used to implicity guarantee that work items queued without explicit CPU specified are put on the local CPU. Recent changes in timer broke the guarantee and led to vmstat breakage which was fixed by 176bed1de5bf ("vmstat: explicitly schedule per-cpu work on the CPU we need it to run on"). vmstat is the most likely to expose the issue and it's quite possible that there are other similar problems which are a lot more difficult to trigger. As a preventive measure, 874bbfe600a6 ("workqueue: make sure delayed work run in local cpu") was applied to restore the local CPU guarnatee. Unfortunately, the change exposed a bug in timer code which got fixed by 22b886dd1018 ("timers: Use proper base migration in add_timer_on()"). Due to code restructuring, the commit couldn't be backported beyond certain point and stable kernels which only had 874bbfe600a6 started crashing. The local CPU guarantee was accidental more than anything else and we want to get rid of it anyway. As, with the vmstat case fixed, 874bbfe600a6 is causing more problems than it's fixing, it has been decided to take the chance and officially break the guarantee by reverting the commit. A debug feature will be added to force foreign CPU assignment to expose cases relying on the guarantee and fixes for the individual cases will be backported to stable as necessary. Signed-off-by: Tejun Heo Fixes: 874bbfe600a6 ("workqueue: make sure delayed work run in local cpu") Link: http://lkml.kernel.org/g/20160120211926.GJ10810@quack.suse.cz Cc: stable@vger.kernel.org Cc: Mike Galbraith Cc: Henrique de Moraes Holschuh Cc: Daniel Bilik Cc: Jan Kara Cc: Shaohua Li Cc: Sasha Levin Cc: Ben Hutchings Cc: Thomas Gleixner Cc: Daniel Bilik Cc: Jiri Slaby Cc: Michal Hocko diff --git a/kernel/workqueue.c b/kernel/workqueue.c index dc7faad..5e63d3b 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -1464,13 +1464,13 @@ static void __queue_delayed_work(int cpu, struct workqueue_struct *wq, timer_stats_timer_set_start_info(&dwork->timer); dwork->wq = wq; - /* timer isn't guaranteed to run in this cpu, record earlier */ - if (cpu == WORK_CPU_UNBOUND) - cpu = raw_smp_processor_id(); dwork->cpu = cpu; timer->expires = jiffies + delay; - add_timer_on(timer, cpu); + if (unlikely(cpu != WORK_CPU_UNBOUND)) + add_timer_on(timer, cpu); + else + add_timer(timer); } /** -- cgit v0.10.2 From ef557180447fa9a7a0affd3abb21ecceb4b5e125 Mon Sep 17 00:00:00 2001 From: Mike Galbraith Date: Tue, 9 Feb 2016 17:59:38 -0500 Subject: workqueue: schedule WORK_CPU_UNBOUND work on wq_unbound_cpumask CPUs WORK_CPU_UNBOUND work items queued to a bound workqueue always run locally. This is a good thing normally, but not when the user has asked us to keep unbound work away from certain CPUs. Round robin these to wq_unbound_cpumask CPUs instead, as perturbation avoidance trumps performance. tj: Cosmetic and comment changes. WARN_ON_ONCE() dropped from empty (wq_unbound_cpumask AND cpu_online_mask). If we want that, it should be done when config changes. Signed-off-by: Mike Galbraith Signed-off-by: Tejun Heo diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 5e63d3b..0547746 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -301,7 +301,11 @@ static DEFINE_SPINLOCK(wq_mayday_lock); /* protects wq->maydays list */ static LIST_HEAD(workqueues); /* PR: list of all workqueues */ static bool workqueue_freezing; /* PL: have wqs started freezing? */ -static cpumask_var_t wq_unbound_cpumask; /* PL: low level cpumask for all unbound wqs */ +/* PL: allowable cpus for unbound wqs and work items */ +static cpumask_var_t wq_unbound_cpumask; + +/* CPU where unbound work was last round robin scheduled from this CPU */ +static DEFINE_PER_CPU(int, wq_rr_cpu_last); /* the per-cpu worker pools */ static DEFINE_PER_CPU_SHARED_ALIGNED(struct worker_pool [NR_STD_WORKER_POOLS], @@ -1298,6 +1302,32 @@ static bool is_chained_work(struct workqueue_struct *wq) return worker && worker->current_pwq->wq == wq; } +/* + * When queueing an unbound work item to a wq, prefer local CPU if allowed + * by wq_unbound_cpumask. Otherwise, round robin among the allowed ones to + * avoid perturbing sensitive tasks. + */ +static int wq_select_unbound_cpu(int cpu) +{ + int new_cpu; + + if (cpumask_test_cpu(cpu, wq_unbound_cpumask)) + return cpu; + if (cpumask_empty(wq_unbound_cpumask)) + return cpu; + + new_cpu = __this_cpu_read(wq_rr_cpu_last); + new_cpu = cpumask_next_and(new_cpu, wq_unbound_cpumask, cpu_online_mask); + if (unlikely(new_cpu >= nr_cpu_ids)) { + new_cpu = cpumask_first_and(wq_unbound_cpumask, cpu_online_mask); + if (unlikely(new_cpu >= nr_cpu_ids)) + return cpu; + } + __this_cpu_write(wq_rr_cpu_last, new_cpu); + + return new_cpu; +} + static void __queue_work(int cpu, struct workqueue_struct *wq, struct work_struct *work) { @@ -1323,7 +1353,7 @@ static void __queue_work(int cpu, struct workqueue_struct *wq, return; retry: if (req_cpu == WORK_CPU_UNBOUND) - cpu = raw_smp_processor_id(); + cpu = wq_select_unbound_cpu(raw_smp_processor_id()); /* pwq which will be used unless @work is executing elsewhere */ if (!(wq->flags & WQ_UNBOUND)) -- cgit v0.10.2 From f303fccb82928790ec58eea82722bd5c54d300b3 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Tue, 9 Feb 2016 17:59:38 -0500 Subject: workqueue: implement "workqueue.debug_force_rr_cpu" debug feature Workqueue used to guarantee local execution for work items queued without explicit target CPU. The guarantee is gone now which can break some usages in subtle ways. To flush out those cases, this patch implements a debug feature which forces round-robin CPU selection for all such work items. The debug feature defaults to off and can be enabled with a kernel parameter. The default can be flipped with a debug config option. If you hit this commit during bisection, please refer to 041bd12e272c ("Revert "workqueue: make sure delayed work run in local cpu"") for more information and ping me. Signed-off-by: Tejun Heo diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt index 87d40a7..cda2ead 100644 --- a/Documentation/kernel-parameters.txt +++ b/Documentation/kernel-parameters.txt @@ -4230,6 +4230,17 @@ bytes respectively. Such letter suffixes can also be entirely omitted. The default value of this parameter is determined by the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT. + workqueue.debug_force_rr_cpu + Workqueue used to implicitly guarantee that work + items queued without explicit CPU specified are put + on the local CPU. This guarantee is no longer true + and while local CPU is still preferred work items + may be put on foreign CPUs. This debug option + forces round-robin CPU selection to flush out + usages which depend on the now broken guarantee. + When enabled, memory and cache locality will be + impacted. + x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of default x2apic cluster mode on platforms supporting x2apic. diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 0547746..51d77e7 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -307,6 +307,18 @@ static cpumask_var_t wq_unbound_cpumask; /* CPU where unbound work was last round robin scheduled from this CPU */ static DEFINE_PER_CPU(int, wq_rr_cpu_last); +/* + * Local execution of unbound work items is no longer guaranteed. The + * following always forces round-robin CPU selection on unbound work items + * to uncover usages which depend on it. + */ +#ifdef CONFIG_DEBUG_WQ_FORCE_RR_CPU +static bool wq_debug_force_rr_cpu = true; +#else +static bool wq_debug_force_rr_cpu = false; +#endif +module_param_named(debug_force_rr_cpu, wq_debug_force_rr_cpu, bool, 0644); + /* the per-cpu worker pools */ static DEFINE_PER_CPU_SHARED_ALIGNED(struct worker_pool [NR_STD_WORKER_POOLS], cpu_worker_pools); @@ -1309,10 +1321,17 @@ static bool is_chained_work(struct workqueue_struct *wq) */ static int wq_select_unbound_cpu(int cpu) { + static bool printed_dbg_warning; int new_cpu; - if (cpumask_test_cpu(cpu, wq_unbound_cpumask)) - return cpu; + if (likely(!wq_debug_force_rr_cpu)) { + if (cpumask_test_cpu(cpu, wq_unbound_cpumask)) + return cpu; + } else if (!printed_dbg_warning) { + pr_warn("workqueue: round-robin CPU selection forced, expect performance impact\n"); + printed_dbg_warning = true; + } + if (cpumask_empty(wq_unbound_cpumask)) return cpu; diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index ecb9e75..8bfd1ac 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -1400,6 +1400,21 @@ config RCU_EQS_DEBUG endmenu # "RCU Debugging" +config DEBUG_WQ_FORCE_RR_CPU + bool "Force round-robin CPU selection for unbound work items" + depends on DEBUG_KERNEL + default n + help + Workqueue used to implicitly guarantee that work items queued + without explicit CPU specified are put on the local CPU. This + guarantee is no longer true and while local CPU is still + preferred work items may be put on foreign CPUs. Kernel + parameter "workqueue.debug_force_rr_cpu" is added to force + round-robin CPU selection to flush out usages which depend on the + now broken guarantee. This config option enables the debug + feature by default. When enabled, memory and cache locality will + be impacted. + config DEBUG_BLOCK_EXT_DEVT bool "Force extended block device numbers and spread them" depends on DEBUG_KERNEL -- cgit v0.10.2 From d6e022f1d207a161cd88e08ef0371554680ffc46 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 3 Feb 2016 13:54:25 -0500 Subject: workqueue: handle NUMA_NO_NODE for unbound pool_workqueue lookup When looking up the pool_workqueue to use for an unbound workqueue, workqueue assumes that the target CPU is always bound to a valid NUMA node. However, currently, when a CPU goes offline, the mapping is destroyed and cpu_to_node() returns NUMA_NO_NODE. This has always been broken but hasn't triggered often enough before 874bbfe600a6 ("workqueue: make sure delayed work run in local cpu"). After the commit, workqueue forcifully assigns the local CPU for delayed work items without explicit target CPU to fix a different issue. This widens the window where CPU can go offline while a delayed work item is pending causing delayed work items dispatched with target CPU set to an already offlined CPU. The resulting NUMA_NO_NODE mapping makes workqueue try to queue the work item on a NULL pool_workqueue and thus crash. While 874bbfe600a6 has been reverted for a different reason making the bug less visible again, it can still happen. Fix it by mapping NUMA_NO_NODE to the default pool_workqueue from unbound_pwq_by_node(). This is a temporary workaround. The long term solution is keeping CPU -> NODE mapping stable across CPU off/online cycles which is being worked on. Signed-off-by: Tejun Heo Reported-by: Mike Galbraith Cc: Tang Chen Cc: Rafael J. Wysocki Cc: Len Brown Cc: stable@vger.kernel.org Link: http://lkml.kernel.org/g/1454424264.11183.46.camel@gmail.com Link: http://lkml.kernel.org/g/1453702100-2597-1-git-send-email-tangchen@cn.fujitsu.com diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 51d77e7..7ff5dc7 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -586,6 +586,16 @@ static struct pool_workqueue *unbound_pwq_by_node(struct workqueue_struct *wq, int node) { assert_rcu_or_wq_mutex_or_pool_mutex(wq); + + /* + * XXX: @node can be NUMA_NO_NODE if CPU goes offline while a + * delayed item is pending. The plan is to keep CPU -> NODE + * mapping valid and stable across CPU on/offlines. Once that + * happens, this workaround can be removed. + */ + if (unlikely(node == NUMA_NO_NODE)) + return wq->dfl_pwq; + return rcu_dereference_raw(wq->numa_pwq_tbl[node]); } -- cgit v0.10.2