diff options
Diffstat (limited to 'drivers')
54 files changed, 10750 insertions, 296 deletions
diff --git a/drivers/acpi/thermal.c b/drivers/acpi/thermal.c index 589b98b7b21..1ada017d01e 100644 --- a/drivers/acpi/thermal.c +++ b/drivers/acpi/thermal.c @@ -59,8 +59,6 @@ #define ACPI_THERMAL_NOTIFY_CRITICAL 0xF0 #define ACPI_THERMAL_NOTIFY_HOT 0xF1 #define ACPI_THERMAL_MODE_ACTIVE 0x00 -#define ACPI_THERMAL_MODE_PASSIVE 0x01 -#define ACPI_THERMAL_MODE_CRITICAL 0xff #define ACPI_THERMAL_PATH_POWEROFF "/sbin/poweroff" #define ACPI_THERMAL_MAX_ACTIVE 10 @@ -86,9 +84,6 @@ static int acpi_thermal_resume(struct acpi_device *device); static int acpi_thermal_state_open_fs(struct inode *inode, struct file *file); static int acpi_thermal_temp_open_fs(struct inode *inode, struct file *file); static int acpi_thermal_trip_open_fs(struct inode *inode, struct file *file); -static ssize_t acpi_thermal_write_trip_points(struct file *, - const char __user *, size_t, - loff_t *); static int acpi_thermal_cooling_open_fs(struct inode *inode, struct file *file); static ssize_t acpi_thermal_write_cooling_mode(struct file *, const char __user *, size_t, @@ -167,7 +162,6 @@ struct acpi_thermal { unsigned long temperature; unsigned long last_temperature; unsigned long polling_frequency; - u8 cooling_mode; volatile u8 zombie; struct acpi_thermal_flags flags; struct acpi_thermal_state state; @@ -193,7 +187,6 @@ static const struct file_operations acpi_thermal_temp_fops = { static const struct file_operations acpi_thermal_trip_fops = { .open = acpi_thermal_trip_open_fs, .read = seq_read, - .write = acpi_thermal_write_trip_points, .llseek = seq_lseek, .release = single_release, }; @@ -297,11 +290,6 @@ static int acpi_thermal_set_cooling_mode(struct acpi_thermal *tz, int mode) if (ACPI_FAILURE(status)) return -ENODEV; - tz->cooling_mode = mode; - - ACPI_DEBUG_PRINT((ACPI_DB_INFO, "Cooling mode [%s]\n", - mode ? "passive" : "active")); - return 0; } @@ -889,67 +877,6 @@ static int acpi_thermal_trip_open_fs(struct inode *inode, struct file *file) return single_open(file, acpi_thermal_trip_seq_show, PDE(inode)->data); } -static ssize_t -acpi_thermal_write_trip_points(struct file *file, - const char __user * buffer, - size_t count, loff_t * ppos) -{ - struct seq_file *m = file->private_data; - struct acpi_thermal *tz = m->private; - - char *limit_string; - int num, critical, hot, passive; - int *active; - int i = 0; - - - limit_string = kzalloc(ACPI_THERMAL_MAX_LIMIT_STR_LEN, GFP_KERNEL); - if (!limit_string) - return -ENOMEM; - - active = kmalloc(ACPI_THERMAL_MAX_ACTIVE * sizeof(int), GFP_KERNEL); - if (!active) { - kfree(limit_string); - return -ENOMEM; - } - - if (!tz || (count > ACPI_THERMAL_MAX_LIMIT_STR_LEN - 1)) { - count = -EINVAL; - goto end; - } - - if (copy_from_user(limit_string, buffer, count)) { - count = -EFAULT; - goto end; - } - - limit_string[count] = '\0'; - - num = sscanf(limit_string, "%d:%d:%d:%d:%d:%d:%d:%d:%d:%d:%d:%d:%d", - &critical, &hot, &passive, - &active[0], &active[1], &active[2], &active[3], &active[4], - &active[5], &active[6], &active[7], &active[8], - &active[9]); - if (!(num >= 5 && num < (ACPI_THERMAL_MAX_ACTIVE + 3))) { - count = -EINVAL; - goto end; - } - - tz->trips.critical.temperature = CELSIUS_TO_KELVIN(critical); - tz->trips.hot.temperature = CELSIUS_TO_KELVIN(hot); - tz->trips.passive.temperature = CELSIUS_TO_KELVIN(passive); - for (i = 0; i < num - 3; i++) { - if (!(tz->trips.active[i].flags.valid)) - break; - tz->trips.active[i].temperature = CELSIUS_TO_KELVIN(active[i]); - } - - end: - kfree(active); - kfree(limit_string); - return count; -} - static int acpi_thermal_cooling_seq_show(struct seq_file *seq, void *offset) { struct acpi_thermal *tz = seq->private; @@ -958,15 +885,10 @@ static int acpi_thermal_cooling_seq_show(struct seq_file *seq, void *offset) if (!tz) goto end; - if (!tz->flags.cooling_mode) { + if (!tz->flags.cooling_mode) seq_puts(seq, "<setting not supported>\n"); - } - - if (tz->cooling_mode == ACPI_THERMAL_MODE_CRITICAL) - seq_printf(seq, "cooling mode: critical\n"); else - seq_printf(seq, "cooling mode: %s\n", - tz->cooling_mode ? "passive" : "active"); + seq_puts(seq, "0 - Active; 1 - Passive\n"); end: return 0; @@ -1223,28 +1145,6 @@ static int acpi_thermal_get_info(struct acpi_thermal *tz) result = acpi_thermal_set_cooling_mode(tz, ACPI_THERMAL_MODE_ACTIVE); if (!result) tz->flags.cooling_mode = 1; - else { - /* Oh,we have not _SCP method. - Generally show cooling_mode by _ACx, _PSV,spec 12.2 */ - tz->flags.cooling_mode = 0; - if (tz->trips.active[0].flags.valid - && tz->trips.passive.flags.valid) { - if (tz->trips.passive.temperature > - tz->trips.active[0].temperature) - tz->cooling_mode = ACPI_THERMAL_MODE_ACTIVE; - else - tz->cooling_mode = ACPI_THERMAL_MODE_PASSIVE; - } else if (!tz->trips.active[0].flags.valid - && tz->trips.passive.flags.valid) { - tz->cooling_mode = ACPI_THERMAL_MODE_PASSIVE; - } else if (tz->trips.active[0].flags.valid - && !tz->trips.passive.flags.valid) { - tz->cooling_mode = ACPI_THERMAL_MODE_ACTIVE; - } else { - /* _ACx and _PSV are optional, but _CRT is required */ - tz->cooling_mode = ACPI_THERMAL_MODE_CRITICAL; - } - } /* Get default polling frequency [_TZP] (optional) */ if (tzp) diff --git a/drivers/infiniband/Kconfig b/drivers/infiniband/Kconfig index 66b36de9fa6..37deaae4919 100644 --- a/drivers/infiniband/Kconfig +++ b/drivers/infiniband/Kconfig @@ -29,6 +29,11 @@ config INFINIBAND_USER_ACCESS libibverbs, libibcm and a hardware driver library from <http://www.openib.org>. +config INFINIBAND_USER_MEM + bool + depends on INFINIBAND_USER_ACCESS != n + default y + config INFINIBAND_ADDR_TRANS bool depends on INFINIBAND && INET @@ -40,6 +45,8 @@ source "drivers/infiniband/hw/ehca/Kconfig" source "drivers/infiniband/hw/amso1100/Kconfig" source "drivers/infiniband/hw/cxgb3/Kconfig" +source "drivers/infiniband/hw/mlx4/Kconfig" + source "drivers/infiniband/ulp/ipoib/Kconfig" source "drivers/infiniband/ulp/srp/Kconfig" diff --git a/drivers/infiniband/Makefile b/drivers/infiniband/Makefile index da2066c4f22..75f325e40b5 100644 --- a/drivers/infiniband/Makefile +++ b/drivers/infiniband/Makefile @@ -4,6 +4,7 @@ obj-$(CONFIG_INFINIBAND_IPATH) += hw/ipath/ obj-$(CONFIG_INFINIBAND_EHCA) += hw/ehca/ obj-$(CONFIG_INFINIBAND_AMSO1100) += hw/amso1100/ obj-$(CONFIG_INFINIBAND_CXGB3) += hw/cxgb3/ +obj-$(CONFIG_MLX4_INFINIBAND) += hw/mlx4/ obj-$(CONFIG_INFINIBAND_IPOIB) += ulp/ipoib/ obj-$(CONFIG_INFINIBAND_SRP) += ulp/srp/ obj-$(CONFIG_INFINIBAND_ISER) += ulp/iser/ diff --git a/drivers/infiniband/core/Makefile b/drivers/infiniband/core/Makefile index 189e5d4b9b1..cb1ab3ea499 100644 --- a/drivers/infiniband/core/Makefile +++ b/drivers/infiniband/core/Makefile @@ -9,6 +9,7 @@ obj-$(CONFIG_INFINIBAND_USER_ACCESS) += ib_uverbs.o ib_ucm.o \ ib_core-y := packer.o ud_header.o verbs.o sysfs.o \ device.o fmr_pool.o cache.o +ib_core-$(CONFIG_INFINIBAND_USER_MEM) += umem.o ib_mad-y := mad.o smi.o agent.o mad_rmpp.o @@ -28,5 +29,4 @@ ib_umad-y := user_mad.o ib_ucm-y := ucm.o -ib_uverbs-y := uverbs_main.o uverbs_cmd.o uverbs_mem.o \ - uverbs_marshall.o +ib_uverbs-y := uverbs_main.o uverbs_cmd.o uverbs_marshall.o diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core/device.c index 7fabb425b03..592c90aa318 100644 --- a/drivers/infiniband/core/device.c +++ b/drivers/infiniband/core/device.c @@ -613,6 +613,8 @@ static void __exit ib_core_cleanup(void) { ib_cache_cleanup(); ib_sysfs_cleanup(); + /* Make sure that any pending umem accounting work is done. */ + flush_scheduled_work(); } module_init(ib_core_init); diff --git a/drivers/infiniband/core/uverbs_mem.c b/drivers/infiniband/core/umem.c index c95fe952abd..f32ca5fbb26 100644 --- a/drivers/infiniband/core/uverbs_mem.c +++ b/drivers/infiniband/core/umem.c @@ -39,13 +39,6 @@ #include "uverbs.h" -struct ib_umem_account_work { - struct work_struct work; - struct mm_struct *mm; - unsigned long diff; -}; - - static void __ib_umem_release(struct ib_device *dev, struct ib_umem *umem, int dirty) { struct ib_umem_chunk *chunk, *tmp; @@ -64,35 +57,56 @@ static void __ib_umem_release(struct ib_device *dev, struct ib_umem *umem, int d } } -int ib_umem_get(struct ib_device *dev, struct ib_umem *mem, - void *addr, size_t size, int write) +/** + * ib_umem_get - Pin and DMA map userspace memory. + * @context: userspace context to pin memory for + * @addr: userspace virtual address to start at + * @size: length of region to pin + * @access: IB_ACCESS_xxx flags for memory being pinned + */ +struct ib_umem *ib_umem_get(struct ib_ucontext *context, unsigned long addr, + size_t size, int access) { + struct ib_umem *umem; struct page **page_list; struct ib_umem_chunk *chunk; unsigned long locked; unsigned long lock_limit; unsigned long cur_base; unsigned long npages; - int ret = 0; + int ret; int off; int i; if (!can_do_mlock()) - return -EPERM; + return ERR_PTR(-EPERM); - page_list = (struct page **) __get_free_page(GFP_KERNEL); - if (!page_list) - return -ENOMEM; + umem = kmalloc(sizeof *umem, GFP_KERNEL); + if (!umem) + return ERR_PTR(-ENOMEM); + + umem->context = context; + umem->length = size; + umem->offset = addr & ~PAGE_MASK; + umem->page_size = PAGE_SIZE; + /* + * We ask for writable memory if any access flags other than + * "remote read" are set. "Local write" and "remote write" + * obviously require write access. "Remote atomic" can do + * things like fetch and add, which will modify memory, and + * "MW bind" can change permissions by binding a window. + */ + umem->writable = !!(access & ~IB_ACCESS_REMOTE_READ); - mem->user_base = (unsigned long) addr; - mem->length = size; - mem->offset = (unsigned long) addr & ~PAGE_MASK; - mem->page_size = PAGE_SIZE; - mem->writable = write; + INIT_LIST_HEAD(&umem->chunk_list); - INIT_LIST_HEAD(&mem->chunk_list); + page_list = (struct page **) __get_free_page(GFP_KERNEL); + if (!page_list) { + kfree(umem); + return ERR_PTR(-ENOMEM); + } - npages = PAGE_ALIGN(size + mem->offset) >> PAGE_SHIFT; + npages = PAGE_ALIGN(size + umem->offset) >> PAGE_SHIFT; down_write(¤t->mm->mmap_sem); @@ -104,13 +118,13 @@ int ib_umem_get(struct ib_device *dev, struct ib_umem *mem, goto out; } - cur_base = (unsigned long) addr & PAGE_MASK; + cur_base = addr & PAGE_MASK; while (npages) { ret = get_user_pages(current, current->mm, cur_base, min_t(int, npages, PAGE_SIZE / sizeof (struct page *)), - 1, !write, page_list, NULL); + 1, !umem->writable, page_list, NULL); if (ret < 0) goto out; @@ -136,7 +150,7 @@ int ib_umem_get(struct ib_device *dev, struct ib_umem *mem, chunk->page_list[i].length = PAGE_SIZE; } - chunk->nmap = ib_dma_map_sg(dev, + chunk->nmap = ib_dma_map_sg(context->device, &chunk->page_list[0], chunk->nents, DMA_BIDIRECTIONAL); @@ -151,75 +165,94 @@ int ib_umem_get(struct ib_device *dev, struct ib_umem *mem, ret -= chunk->nents; off += chunk->nents; - list_add_tail(&chunk->list, &mem->chunk_list); + list_add_tail(&chunk->list, &umem->chunk_list); } ret = 0; } out: - if (ret < 0) - __ib_umem_release(dev, mem, 0); - else + if (ret < 0) { + __ib_umem_release(context->device, umem, 0); + kfree(umem); + } else current->mm->locked_vm = locked; up_write(¤t->mm->mmap_sem); free_page((unsigned long) page_list); - return ret; + return ret < 0 ? ERR_PTR(ret) : umem; } +EXPORT_SYMBOL(ib_umem_get); -void ib_umem_release(struct ib_device *dev, struct ib_umem *umem) +static void ib_umem_account(struct work_struct *work) { - __ib_umem_release(dev, umem, 1); - - down_write(¤t->mm->mmap_sem); - current->mm->locked_vm -= - PAGE_ALIGN(umem->length + umem->offset) >> PAGE_SHIFT; - up_write(¤t->mm->mmap_sem); -} + struct ib_umem *umem = container_of(work, struct ib_umem, work); -static void ib_umem_account(struct work_struct *_work) -{ - struct ib_umem_account_work *work = - container_of(_work, struct ib_umem_account_work, work); - - down_write(&work->mm->mmap_sem); - work->mm->locked_vm -= work->diff; - up_write(&work->mm->mmap_sem); - mmput(work->mm); - kfree(work); + down_write(&umem->mm->mmap_sem); + umem->mm->locked_vm -= umem->diff; + up_write(&umem->mm->mmap_sem); + mmput(umem->mm); + kfree(umem); } -void ib_umem_release_on_close(struct ib_device *dev, struct ib_umem *umem) +/** + * ib_umem_release - release memory pinned with ib_umem_get + * @umem: umem struct to release + */ +void ib_umem_release(struct ib_umem *umem) { - struct ib_umem_account_work *work; + struct ib_ucontext *context = umem->context; struct mm_struct *mm; + unsigned long diff; - __ib_umem_release(dev, umem, 1); + __ib_umem_release(umem->context->device, umem, 1); mm = get_task_mm(current); if (!mm) return; + diff = PAGE_ALIGN(umem->length + umem->offset) >> PAGE_SHIFT; + /* * We may be called with the mm's mmap_sem already held. This * can happen when a userspace munmap() is the call that drops * the last reference to our file and calls our release * method. If there are memory regions to destroy, we'll end - * up here and not be able to take the mmap_sem. Therefore we - * defer the vm_locked accounting to the system workqueue. + * up here and not be able to take the mmap_sem. In that case + * we defer the vm_locked accounting to the system workqueue. */ + if (context->closing && !down_write_trylock(&mm->mmap_sem)) { + INIT_WORK(&umem->work, ib_umem_account); + umem->mm = mm; + umem->diff = diff; - work = kmalloc(sizeof *work, GFP_KERNEL); - if (!work) { - mmput(mm); + schedule_work(&umem->work); return; - } + } else + down_write(&mm->mmap_sem); + + current->mm->locked_vm -= diff; + up_write(&mm->mmap_sem); + mmput(mm); + kfree(umem); +} +EXPORT_SYMBOL(ib_umem_release); + +int ib_umem_page_count(struct ib_umem *umem) +{ + struct ib_umem_chunk *chunk; + int shift; + int i; + int n; + + shift = ilog2(umem->page_size); - INIT_WORK(&work->work, ib_umem_account); - work->mm = mm; - work->diff = PAGE_ALIGN(umem->length + umem->offset) >> PAGE_SHIFT; + n = 0; + list_for_each_entry(chunk, &umem->chunk_list, list) + for (i = 0; i < chunk->nmap; ++i) + n += sg_dma_len(&chunk->page_list[i]) >> shift; - schedule_work(&work->work); + return n; } +EXPORT_SYMBOL(ib_umem_page_count); diff --git a/drivers/infiniband/core/uverbs.h b/drivers/infiniband/core/uverbs.h index 102a59c033f..c33546f9e96 100644 --- a/drivers/infiniband/core/uverbs.h +++ b/drivers/infiniband/core/uverbs.h @@ -45,6 +45,7 @@ #include <linux/completion.h> #include <rdma/ib_verbs.h> +#include <rdma/ib_umem.h> #include <rdma/ib_user_verbs.h> /* @@ -163,11 +164,6 @@ void ib_uverbs_srq_event_handler(struct ib_event *event, void *context_ptr); void ib_uverbs_event_handler(struct ib_event_handler *handler, struct ib_event *event); -int ib_umem_get(struct ib_device *dev, struct ib_umem *mem, - void *addr, size_t size, int write); -void ib_umem_release(struct ib_device *dev, struct ib_umem *umem); -void ib_umem_release_on_close(struct ib_device *dev, struct ib_umem *umem); - #define IB_UVERBS_DECLARE_CMD(name) \ ssize_t ib_uverbs_##name(struct ib_uverbs_file *file, \ const char __user *buf, int in_len, \ diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c index bab66769be1..01d70084aeb 100644 --- a/drivers/infiniband/core/uverbs_cmd.c +++ b/drivers/infiniband/core/uverbs_cmd.c @@ -1,6 +1,6 @@ /* * Copyright (c) 2005 Topspin Communications. All rights reserved. - * Copyright (c) 2005, 2006 Cisco Systems. All rights reserved. + * Copyright (c) 2005, 2006, 2007 Cisco Systems. All rights reserved. * Copyright (c) 2005 PathScale, Inc. All rights reserved. * Copyright (c) 2006 Mellanox Technologies. All rights reserved. * @@ -295,6 +295,7 @@ ssize_t ib_uverbs_get_context(struct ib_uverbs_file *file, INIT_LIST_HEAD(&ucontext->qp_list); INIT_LIST_HEAD(&ucontext->srq_list); INIT_LIST_HEAD(&ucontext->ah_list); + ucontext->closing = 0; resp.num_comp_vectors = file->device->num_comp_vectors; @@ -573,7 +574,7 @@ ssize_t ib_uverbs_reg_mr(struct ib_uverbs_file *file, struct ib_uverbs_reg_mr cmd; struct ib_uverbs_reg_mr_resp resp; struct ib_udata udata; - struct ib_umem_object *obj; + struct ib_uobject *uobj; struct ib_pd *pd; struct ib_mr *mr; int ret; @@ -599,35 +600,21 @@ ssize_t ib_uverbs_reg_mr(struct ib_uverbs_file *file, !(cmd.access_flags & IB_ACCESS_LOCAL_WRITE)) return -EINVAL; - obj = kmalloc(sizeof *obj, GFP_KERNEL); - if (!obj) + uobj = kmalloc(sizeof *uobj, GFP_KERNEL); + if (!uobj) return -ENOMEM; - init_uobj(&obj->uobject, 0, file->ucontext, &mr_lock_key); - down_write(&obj->uobject.mutex); - - /* - * We ask for writable memory if any access flags other than - * "remote read" are set. "Local write" and "remote write" - * obviously require write access. "Remote atomic" can do - * things like fetch and add, which will modify memory, and - * "MW bind" can change permissions by binding a window. - */ - ret = ib_umem_get(file->device->ib_dev, &obj->umem, - (void *) (unsigned long) cmd.start, cmd.length, - !!(cmd.access_flags & ~IB_ACCESS_REMOTE_READ)); - if (ret) - goto err_free; - - obj->umem.virt_base = cmd.hca_va; + init_uobj(uobj, 0, file->ucontext, &mr_lock_key); + down_write(&uobj->mutex); pd = idr_read_pd(cmd.pd_handle, file->ucontext); if (!pd) { ret = -EINVAL; - goto err_release; + goto err_free; } - mr = pd->device->reg_user_mr(pd, &obj->umem, cmd.access_flags, &udata); + mr = pd->device->reg_user_mr(pd, cmd.start, cmd.length, cmd.hca_va, + cmd.access_flags, &udata); if (IS_ERR(mr)) { ret = PTR_ERR(mr); goto err_put; @@ -635,19 +622,19 @@ ssize_t ib_uverbs_reg_mr(struct ib_uverbs_file *file, mr->device = pd->device; mr->pd = pd; - mr->uobject = &obj->uobject; + mr->uobject = uobj; atomic_inc(&pd->usecnt); atomic_set(&mr->usecnt, 0); - obj->uobject.object = mr; - ret = idr_add_uobj(&ib_uverbs_mr_idr, &obj->uobject); + uobj->object = mr; + ret = idr_add_uobj(&ib_uverbs_mr_idr, uobj); if (ret) goto err_unreg; memset(&resp, 0, sizeof resp); resp.lkey = mr->lkey; resp.rkey = mr->rkey; - resp.mr_handle = obj->uobject.id; + resp.mr_handle = uobj->id; if (co |