From fc3100d64f0ae383ae8d845989103da06d62763b Mon Sep 17 00:00:00 2001 From: Pu Hou Date: Tue, 5 Sep 2017 05:17:24 +0200 Subject: [PATCH 1/8] s390/perf: fix bug when creating per-thread event A per-thread event could not be created correctly like below: perf record --per-thread -e rB0000 -- sleep 1 Error: The sys_perf_event_open() syscall returned with 19 (No such device) for event (rB0000). /bin/dmesg may provide additional information. No CONFIG_PERF_EVENTS=y kernel support configured? This bug was introduced by: commit c311c797998c1e70eade463dd60b843da4f1a203 Author: Alexey Dobriyan Date: Mon May 8 15:56:15 2017 -0700 cpumask: make "nr_cpumask_bits" unsigned If a per-thread event is not attached to any CPU, the cpu field in struct perf_event is -1. The above commit converts the CPU number to unsigned int, which result in an illegal CPU number. Fixes: c311c797998c ("cpumask: make "nr_cpumask_bits" unsigned") Cc: # v4.12+ Cc: Alexey Dobriyan Acked-by: Heiko Carstens Signed-off-by: Pu Hou Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/perf_cpum_sf.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/arch/s390/kernel/perf_cpum_sf.c b/arch/s390/kernel/perf_cpum_sf.c index c1bf75ffb875..7e1e40323b78 100644 --- a/arch/s390/kernel/perf_cpum_sf.c +++ b/arch/s390/kernel/perf_cpum_sf.c @@ -823,9 +823,12 @@ static int cpumsf_pmu_event_init(struct perf_event *event) } /* Check online status of the CPU to which the event is pinned */ - if ((unsigned int)event->cpu >= nr_cpumask_bits || - (event->cpu >= 0 && !cpu_online(event->cpu))) - return -ENODEV; + if (event->cpu >= 0) { + if ((unsigned int)event->cpu >= nr_cpumask_bits) + return -ENODEV; + if (!cpu_online(event->cpu)) + return -ENODEV; + } /* Force reset of idle/hv excludes regardless of what the * user requested. From 673514aff5d797b0cdb3f09097ae0e253b5667c3 Mon Sep 17 00:00:00 2001 From: Stefan Haberland Date: Tue, 12 Sep 2017 17:22:42 +0200 Subject: [PATCH 2/8] s390/dasd: fix race during dasd initialization Fix a panic in blk_mq_hctx_has_pending() that is caused by a racy call to blk_mq_run_hw_queues in a dasd function that might get called with the request queue not yet initialized during initialization. Signed-off-by: Stefan Haberland Signed-off-by: Martin Schwidefsky --- drivers/s390/block/dasd.c | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-) diff --git a/drivers/s390/block/dasd.c b/drivers/s390/block/dasd.c index ea19b4ff87a2..29f35e29d480 100644 --- a/drivers/s390/block/dasd.c +++ b/drivers/s390/block/dasd.c @@ -1644,7 +1644,9 @@ void dasd_generic_handle_state_change(struct dasd_device *device) dasd_schedule_device_bh(device); if (device->block) { dasd_schedule_block_bh(device->block); - blk_mq_run_hw_queues(device->block->request_queue, true); + if (device->block->request_queue) + blk_mq_run_hw_queues(device->block->request_queue, + true); } } EXPORT_SYMBOL_GPL(dasd_generic_handle_state_change); @@ -3759,7 +3761,9 @@ int dasd_generic_path_operational(struct dasd_device *device) dasd_schedule_device_bh(device); if (device->block) { dasd_schedule_block_bh(device->block); - blk_mq_run_hw_queues(device->block->request_queue, true); + if (device->block->request_queue) + blk_mq_run_hw_queues(device->block->request_queue, + true); } if (!device->stopped) @@ -4025,7 +4029,9 @@ int dasd_generic_restore_device(struct ccw_device *cdev) if (device->block) { dasd_schedule_block_bh(device->block); - blk_mq_run_hw_queues(device->block->request_queue, true); + if (device->block->request_queue) + blk_mq_run_hw_queues(device->block->request_queue, + true); } clear_bit(DASD_FLAG_SUSPENDED, &device->flags); From c8b850241559d6bad7e640075e282a83a4ad7ead Mon Sep 17 00:00:00 2001 From: Sebastian Ott Date: Mon, 18 Sep 2017 12:25:22 +0200 Subject: [PATCH 3/8] s390/scm_blk: consistently use blk_status_t as error type Fix these warnings found by sparse: drivers/s390/block/scm_blk.c:257:24: warning: incorrect type in assignment (different base types) drivers/s390/block/scm_blk.c:257:24: expected int [signed] drivers/s390/block/scm_blk.c:257:24: got restricted blk_status_t [usertype] error drivers/s390/block/scm_blk.c:420:33: warning: incorrect type in argument 2 (different base types) drivers/s390/block/scm_blk.c:420:33: expected restricted blk_status_t [usertype] error drivers/s390/block/scm_blk.c:420:33: got int [signed] Signed-off-by: Sebastian Ott Reported-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- drivers/s390/block/scm_blk.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/s390/block/scm_blk.c b/drivers/s390/block/scm_blk.c index 2e7fd966c515..eb51893c74a4 100644 --- a/drivers/s390/block/scm_blk.c +++ b/drivers/s390/block/scm_blk.c @@ -249,7 +249,7 @@ static void scm_request_requeue(struct scm_request *scmrq) static void scm_request_finish(struct scm_request *scmrq) { struct scm_blk_dev *bdev = scmrq->bdev; - int *error; + blk_status_t *error; int i; for (i = 0; i < nr_requests_per_io && scmrq->request[i]; i++) { @@ -415,7 +415,7 @@ void scm_blk_irq(struct scm_device *scmdev, void *data, blk_status_t error) static void scm_blk_request_done(struct request *req) { - int *error = blk_mq_rq_to_pdu(req); + blk_status_t *error = blk_mq_rq_to_pdu(req); blk_mq_end_request(req, *error); } @@ -450,7 +450,7 @@ int scm_blk_dev_setup(struct scm_blk_dev *bdev, struct scm_device *scmdev) atomic_set(&bdev->queued_reqs, 0); bdev->tag_set.ops = &scm_mq_ops; - bdev->tag_set.cmd_size = sizeof(int); + bdev->tag_set.cmd_size = sizeof(blk_status_t); bdev->tag_set.nr_hw_queues = nr_requests; bdev->tag_set.queue_depth = nr_requests_per_io * nr_requests; bdev->tag_set.flags = BLK_MQ_F_SHOULD_MERGE; From 55fb7347579017bdb09550b4e8019447340b7004 Mon Sep 17 00:00:00 2001 From: Sebastian Ott Date: Thu, 14 Sep 2017 13:55:22 +0200 Subject: [PATCH 4/8] s390/cio: recover from bad paths In some situations we don't receive notification from firmware that a previously unusable channelpath is usable again. Schedule recovery for devices that return from path verification without using all potentially usable paths. The recovery thread will periodically trigger a path verification on the affected devices. Signed-off-by: Sebastian Ott Suggested-by: Peter Oberparleiter Reviewed-by: Peter Oberparleiter Signed-off-by: Martin Schwidefsky --- drivers/s390/cio/device.c | 12 +++++++++--- drivers/s390/cio/device.h | 1 + drivers/s390/cio/device_fsm.c | 12 ++++++++++++ drivers/s390/cio/io_sch.h | 2 ++ 4 files changed, 24 insertions(+), 3 deletions(-) diff --git a/drivers/s390/cio/device.c b/drivers/s390/cio/device.c index 489b583f263d..e5c32f4b5287 100644 --- a/drivers/s390/cio/device.c +++ b/drivers/s390/cio/device.c @@ -1225,10 +1225,16 @@ static int device_is_disconnected(struct ccw_device *cdev) static int recovery_check(struct device *dev, void *data) { struct ccw_device *cdev = to_ccwdev(dev); + struct subchannel *sch; int *redo = data; spin_lock_irq(cdev->ccwlock); switch (cdev->private->state) { + case DEV_STATE_ONLINE: + sch = to_subchannel(cdev->dev.parent); + if ((sch->schib.pmcw.pam & sch->opm) == sch->vpm) + break; + /* fall through */ case DEV_STATE_DISCONNECTED: CIO_MSG_EVENT(3, "recovery: trigger 0.%x.%04x\n", cdev->private->dev_id.ssid, @@ -1260,7 +1266,7 @@ static void recovery_work_func(struct work_struct *unused) } spin_unlock_irq(&recovery_lock); } else - CIO_MSG_EVENT(4, "recovery: end\n"); + CIO_MSG_EVENT(3, "recovery: end\n"); } static DECLARE_WORK(recovery_work, recovery_work_func); @@ -1274,11 +1280,11 @@ static void recovery_func(unsigned long data) schedule_work(&recovery_work); } -static void ccw_device_schedule_recovery(void) +void ccw_device_schedule_recovery(void) { unsigned long flags; - CIO_MSG_EVENT(4, "recovery: schedule\n"); + CIO_MSG_EVENT(3, "recovery: schedule\n"); spin_lock_irqsave(&recovery_lock, flags); if (!timer_pending(&recovery_timer) || (recovery_phase != 0)) { recovery_phase = 0; diff --git a/drivers/s390/cio/device.h b/drivers/s390/cio/device.h index ec497af99dd8..69cb70f080a5 100644 --- a/drivers/s390/cio/device.h +++ b/drivers/s390/cio/device.h @@ -134,6 +134,7 @@ void ccw_device_set_disconnected(struct ccw_device *cdev); void ccw_device_set_notoper(struct ccw_device *cdev); void ccw_device_set_timeout(struct ccw_device *, int); +void ccw_device_schedule_recovery(void); /* Channel measurement facility related */ void retry_set_schib(struct ccw_device *cdev); diff --git a/drivers/s390/cio/device_fsm.c b/drivers/s390/cio/device_fsm.c index 12016e32e519..f98ea674c3d8 100644 --- a/drivers/s390/cio/device_fsm.c +++ b/drivers/s390/cio/device_fsm.c @@ -476,6 +476,17 @@ static void create_fake_irb(struct irb *irb, int type) } } +static void ccw_device_handle_broken_paths(struct ccw_device *cdev) +{ + struct subchannel *sch = to_subchannel(cdev->dev.parent); + u8 broken_paths = (sch->schib.pmcw.pam & sch->opm) ^ sch->vpm; + + if (broken_paths && (cdev->private->path_broken_mask != broken_paths)) + ccw_device_schedule_recovery(); + + cdev->private->path_broken_mask = broken_paths; +} + void ccw_device_verify_done(struct ccw_device *cdev, int err) { struct subchannel *sch; @@ -508,6 +519,7 @@ void ccw_device_verify_done(struct ccw_device *cdev, int err) memset(&cdev->private->irb, 0, sizeof(struct irb)); } ccw_device_report_path_events(cdev); + ccw_device_handle_broken_paths(cdev); break; case -ETIME: case -EUSERS: diff --git a/drivers/s390/cio/io_sch.h b/drivers/s390/cio/io_sch.h index 220f49145b2f..9a1b56b2df3e 100644 --- a/drivers/s390/cio/io_sch.h +++ b/drivers/s390/cio/io_sch.h @@ -131,6 +131,8 @@ struct ccw_device_private { not operable */ u8 path_gone_mask; /* mask of paths, that became unavailable */ u8 path_new_mask; /* mask of paths, that became available */ + u8 path_broken_mask; /* mask of paths, which were found to be + unusable */ struct { unsigned int fast:1; /* post with "channel end" */ unsigned int repall:1; /* report every interrupt status */ From 91c575b335766effa6103eba42a82aea560c365f Mon Sep 17 00:00:00 2001 From: Gerald Schaefer Date: Mon, 18 Sep 2017 16:10:35 +0200 Subject: [PATCH 5/8] s390/mm: make pmdp_invalidate() do invalidation only Commit 227be799c39a ("s390/mm: uninline pmdp_xxx functions from pgtable.h") inadvertently changed the behavior of pmdp_invalidate(), so that it now clears the pmd instead of just marking it as invalid. Fix this by restoring the original behavior. A possible impact of the misbehaving pmdp_invalidate() would be the MADV_DONTNEED races (see commits ced10803 and 58ceeb6b), although we should not have any negative impact on the related dirty/young flags, since those flags are not set by the hardware on s390. Fixes: 227be799c39a ("s390/mm: uninline pmdp_xxx functions from pgtable.h") Cc: # v4.6+ Signed-off-by: Gerald Schaefer Signed-off-by: Martin Schwidefsky --- arch/s390/include/asm/pgtable.h | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/arch/s390/include/asm/pgtable.h b/arch/s390/include/asm/pgtable.h index dce708e061ea..20e75a2ca93a 100644 --- a/arch/s390/include/asm/pgtable.h +++ b/arch/s390/include/asm/pgtable.h @@ -1507,7 +1507,9 @@ static inline pmd_t pmdp_huge_clear_flush(struct vm_area_struct *vma, static inline void pmdp_invalidate(struct vm_area_struct *vma, unsigned long addr, pmd_t *pmdp) { - pmdp_xchg_direct(vma->vm_mm, addr, pmdp, __pmd(_SEGMENT_ENTRY_EMPTY)); + pmd_t pmd = __pmd(pmd_val(*pmdp) | _SEGMENT_ENTRY_INVALID); + + pmdp_xchg_direct(vma->vm_mm, addr, pmdp, pmd); } #define __HAVE_ARCH_PMDP_SET_WRPROTECT From ba385c0594e723d41790ecfb12c610e6f90c7785 Mon Sep 17 00:00:00 2001 From: Gerald Schaefer Date: Mon, 18 Sep 2017 16:51:51 +0200 Subject: [PATCH 6/8] s390/mm: fix write access check in gup_huge_pmd() The check for the _SEGMENT_ENTRY_PROTECT bit in gup_huge_pmd() is the wrong way around. It must not be set for write==1, and not be checked for write==0. Fix this similar to how it was fixed for ptes long time ago in commit 25591b070336 ("[S390] fix get_user_pages_fast"). One impact of this bug would be unnecessarily using the gup slow path for write==0 on r/w mappings. A potentially more severe impact would be that gup_huge_pmd() will succeed for write==1 on r/o mappings. Cc: Signed-off-by: Gerald Schaefer Signed-off-by: Martin Schwidefsky --- arch/s390/mm/gup.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/arch/s390/mm/gup.c b/arch/s390/mm/gup.c index 8ecc25e760fa..98ffe3ee9411 100644 --- a/arch/s390/mm/gup.c +++ b/arch/s390/mm/gup.c @@ -56,13 +56,12 @@ static inline int gup_pte_range(pmd_t *pmdp, pmd_t pmd, unsigned long addr, static inline int gup_huge_pmd(pmd_t *pmdp, pmd_t pmd, unsigned long addr, unsigned long end, int write, struct page **pages, int *nr) { - unsigned long mask, result; struct page *head, *page; + unsigned long mask; int refs; - result = write ? 0 : _SEGMENT_ENTRY_PROTECT; - mask = result | _SEGMENT_ENTRY_INVALID; - if ((pmd_val(pmd) & mask) != result) + mask = (write ? _SEGMENT_ENTRY_PROTECT : 0) | _SEGMENT_ENTRY_INVALID; + if ((pmd_val(pmd) & mask) != 0) return 0; VM_BUG_ON(!pfn_valid(pmd_val(pmd) >> PAGE_SHIFT)); From 1b25fda0533462c9cee3a22e8a7bea68fa670af2 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Tue, 19 Sep 2017 12:52:22 +0200 Subject: [PATCH 7/8] s390/topology: alternative topology for topology-less machines If running on machines that do not provide topology information we currently generate a "fake" topology which defines the maximum distance between each cpu: each cpu will be put into an own drawer. Historically this used to be the best option for (virtual) machines in overcommited hypervisors. For some workloads however it is better to generate a different topology where all cpus are siblings within a package (all cpus are core siblings). This shows performance improvements of up to 10%, depending on the workload. In order to keep the current behaviour, but also allow to switch to the different core sibling topology use the existing "topology=" kernel parameter: Specifying "topology=on" on machines without topology information will generate the core siblings (fake) topology information, instead of the default topology information where all cpus have the maximum distance. On machines which provide topology information specifying "topology=on" does not have any effect. Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/early.c | 12 ------- arch/s390/kernel/topology.c | 72 +++++++++++++++++++++++++++++++------ 2 files changed, 61 insertions(+), 23 deletions(-) diff --git a/arch/s390/kernel/early.c b/arch/s390/kernel/early.c index ca8cd80e8feb..60181caf8e8a 100644 --- a/arch/s390/kernel/early.c +++ b/arch/s390/kernel/early.c @@ -404,18 +404,6 @@ static inline void save_vector_registers(void) #endif } -static int __init topology_setup(char *str) -{ - bool enabled; - int rc; - - rc = kstrtobool(str, &enabled); - if (!rc && !enabled) - S390_lowcore.machine_flags &= ~MACHINE_FLAG_TOPOLOGY; - return rc; -} -early_param("topology", topology_setup); - static int __init disable_vector_extension(char *str) { S390_lowcore.machine_flags &= ~MACHINE_FLAG_VX; diff --git a/arch/s390/kernel/topology.c b/arch/s390/kernel/topology.c index bb47c92476f0..a0ce9c83f589 100644 --- a/arch/s390/kernel/topology.c +++ b/arch/s390/kernel/topology.c @@ -29,12 +29,20 @@ #define PTF_VERTICAL (1UL) #define PTF_CHECK (2UL) +enum { + TOPOLOGY_MODE_HW, + TOPOLOGY_MODE_SINGLE, + TOPOLOGY_MODE_PACKAGE, + TOPOLOGY_MODE_UNINITIALIZED +}; + struct mask_info { struct mask_info *next; unsigned char id; cpumask_t mask; }; +static int topology_mode = TOPOLOGY_MODE_UNINITIALIZED; static void set_topology_timer(void); static void topology_work_fn(struct work_struct *work); static struct sysinfo_15_1_x *tl_info; @@ -59,11 +67,26 @@ static cpumask_t cpu_group_map(struct mask_info *info, unsigned int cpu) cpumask_t mask; cpumask_copy(&mask, cpumask_of(cpu)); - if (!MACHINE_HAS_TOPOLOGY) - return mask; - for (; info; info = info->next) { - if (cpumask_test_cpu(cpu, &info->mask)) - return info->mask; + switch (topology_mode) { + case TOPOLOGY_MODE_HW: + while (info) { + if (cpumask_test_cpu(cpu, &info->mask)) { + mask = info->mask; + break; + } + info = info->next; + } + if (cpumask_empty(&mask)) + cpumask_copy(&mask, cpumask_of(cpu)); + break; + case TOPOLOGY_MODE_PACKAGE: + cpumask_copy(&mask, cpu_present_mask); + break; + default: + /* fallthrough */ + case TOPOLOGY_MODE_SINGLE: + cpumask_copy(&mask, cpumask_of(cpu)); + break; } return mask; } @@ -74,7 +97,7 @@ static cpumask_t cpu_thread_map(unsigned int cpu) int i; cpumask_copy(&mask, cpumask_of(cpu)); - if (!MACHINE_HAS_TOPOLOGY) + if (topology_mode != TOPOLOGY_MODE_HW) return mask; cpu -= cpu % (smp_cpu_mtid + 1); for (i = 0; i <= smp_cpu_mtid; i++) @@ -223,7 +246,7 @@ int topology_set_cpu_management(int fc) static void update_cpu_masks(void) { struct cpu_topology_s390 *topo; - int cpu; + int cpu, id; for_each_possible_cpu(cpu) { topo = &cpu_topology[cpu]; @@ -231,12 +254,13 @@ static void update_cpu_masks(void) topo->core_mask = cpu_group_map(&socket_info, cpu); topo->book_mask = cpu_group_map(&book_info, cpu); topo->drawer_mask = cpu_group_map(&drawer_info, cpu); - if (!MACHINE_HAS_TOPOLOGY) { + if (topology_mode != TOPOLOGY_MODE_HW) { + id = topology_mode == TOPOLOGY_MODE_PACKAGE ? 0 : cpu; topo->thread_id = cpu; topo->core_id = cpu; - topo->socket_id = cpu; - topo->book_id = cpu; - topo->drawer_id = cpu; + topo->socket_id = id; + topo->book_id = id; + topo->drawer_id = id; if (cpu_present(cpu)) cpumask_set_cpu(cpu, &cpus_with_topology); } @@ -459,6 +483,12 @@ void __init topology_init_early(void) struct sysinfo_15_1_x *info; set_sched_topology(s390_topology); + if (topology_mode == TOPOLOGY_MODE_UNINITIALIZED) { + if (MACHINE_HAS_TOPOLOGY) + topology_mode = TOPOLOGY_MODE_HW; + else + topology_mode = TOPOLOGY_MODE_SINGLE; + } if (!MACHINE_HAS_TOPOLOGY) goto out; tl_info = memblock_virt_alloc(PAGE_SIZE, PAGE_SIZE); @@ -474,6 +504,26 @@ void __init topology_init_early(void) __arch_update_cpu_topology(); } +static inline int topology_get_mode(int enabled) +{ + if (!enabled) + return TOPOLOGY_MODE_SINGLE; + return MACHINE_HAS_TOPOLOGY ? TOPOLOGY_MODE_HW : TOPOLOGY_MODE_PACKAGE; +} + +static int __init topology_setup(char *str) +{ + bool enabled; + int rc; + + rc = kstrtobool(str, &enabled); + if (rc) + return rc; + topology_mode = topology_get_mode(enabled); + return 0; +} +early_param("topology", topology_setup); + static int __init topology_init(void) { if (MACHINE_HAS_TOPOLOGY) From 51dce3867c6c63c7500332e5448c2ba76808d6b5 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Thu, 14 Sep 2017 14:42:32 +0200 Subject: [PATCH 8/8] s390/topology: enable / disable topology dynamically Add a new sysctl file /proc/sys/s390/topology which displays if topology is on (1) or off (0) as specified by the "topology=" kernel parameter. This allows to change topology information during runtime and configuring it via /etc/sysctl.conf instead of using the kernel line parameter. Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/topology.c | 76 ++++++++++++++++++++++++++++++++++++- 1 file changed, 74 insertions(+), 2 deletions(-) diff --git a/arch/s390/kernel/topology.c b/arch/s390/kernel/topology.c index a0ce9c83f589..ed0bdd220e1a 100644 --- a/arch/s390/kernel/topology.c +++ b/arch/s390/kernel/topology.c @@ -8,6 +8,8 @@ #include #include +#include +#include #include #include #include @@ -207,10 +209,8 @@ static void topology_update_polarization_simple(void) { int cpu; - mutex_lock(&smp_cpu_state_mutex); for_each_possible_cpu(cpu) smp_cpu_set_polarization(cpu, POLARIZATION_HRZ); - mutex_unlock(&smp_cpu_state_mutex); } static int ptf(unsigned long fc) @@ -278,6 +278,7 @@ static int __arch_update_cpu_topology(void) struct sysinfo_15_1_x *info = tl_info; int rc = 0; + mutex_lock(&smp_cpu_state_mutex); cpumask_clear(&cpus_with_topology); if (MACHINE_HAS_TOPOLOGY) { rc = 1; @@ -287,6 +288,7 @@ static int __arch_update_cpu_topology(void) update_cpu_masks(); if (!MACHINE_HAS_TOPOLOGY) topology_update_polarization_simple(); + mutex_unlock(&smp_cpu_state_mutex); return rc; } @@ -313,6 +315,11 @@ void topology_schedule_update(void) schedule_work(&topology_work); } +static void topology_flush_work(void) +{ + flush_work(&topology_work); +} + static void topology_timer_fn(unsigned long ignored) { if (ptf(PTF_CHECK)) @@ -511,6 +518,11 @@ static inline int topology_get_mode(int enabled) return MACHINE_HAS_TOPOLOGY ? TOPOLOGY_MODE_HW : TOPOLOGY_MODE_PACKAGE; } +static inline int topology_is_enabled(void) +{ + return topology_mode != TOPOLOGY_MODE_SINGLE; +} + static int __init topology_setup(char *str) { bool enabled; @@ -524,12 +536,72 @@ static int __init topology_setup(char *str) } early_param("topology", topology_setup); +static int topology_ctl_handler(struct ctl_table *ctl, int write, + void __user *buffer, size_t *lenp, loff_t *ppos) +{ + unsigned int len; + int new_mode; + char buf[2]; + + if (!*lenp || *ppos) { + *lenp = 0; + return 0; + } + if (!write) { + strncpy(buf, topology_is_enabled() ? "1\n" : "0\n", + ARRAY_SIZE(buf)); + len = strnlen(buf, ARRAY_SIZE(buf)); + if (len > *lenp) + len = *lenp; + if (copy_to_user(buffer, buf, len)) + return -EFAULT; + goto out; + } + len = *lenp; + if (copy_from_user(buf, buffer, len > sizeof(buf) ? sizeof(buf) : len)) + return -EFAULT; + if (buf[0] != '0' && buf[0] != '1') + return -EINVAL; + mutex_lock(&smp_cpu_state_mutex); + new_mode = topology_get_mode(buf[0] == '1'); + if (topology_mode != new_mode) { + topology_mode = new_mode; + topology_schedule_update(); + } + mutex_unlock(&smp_cpu_state_mutex); + topology_flush_work(); +out: + *lenp = len; + *ppos += len; + return 0; +} + +static struct ctl_table topology_ctl_table[] = { + { + .procname = "topology", + .mode = 0644, + .proc_handler = topology_ctl_handler, + }, + { }, +}; + +static struct ctl_table topology_dir_table[] = { + { + .procname = "s390", + .maxlen = 0, + .mode = 0555, + .child = topology_ctl_table, + }, + { }, +}; + static int __init topology_init(void) { if (MACHINE_HAS_TOPOLOGY) set_topology_timer(); else topology_update_polarization_simple(); + register_sysctl_table(topology_dir_table); return device_create_file(cpu_subsys.dev_root, &dev_attr_dispatching); } device_initcall(topology_init);