diff options
author | Linus Torvalds <torvalds@linux-foundation.org> | 2016-05-23 11:18:01 -0700 |
---|---|---|
committer | Linus Torvalds <torvalds@linux-foundation.org> | 2016-05-23 11:18:01 -0700 |
commit | 1f40c49570eb01436786a9b5845c4469a9a1f362 (patch) | |
tree | f0a31705d5c0a65604784d9b01841c453055d62f /drivers | |
parent | 7639dad93a5564579987abded4ec05e3db13659d (diff) | |
parent | 36092ee8ba695fce023b2118ececa6c2a56b1331 (diff) |
Merge tag 'libnvdimm-for-4.7' of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm
Pull libnvdimm updates from Dan Williams:
"The bulk of this update was stabilized before the merge window and
appeared in -next. The "device dax" implementation was revised this
week in response to review feedback, and to address failures detected
by the recently expanded ndctl unit test suite.
Not included in this pull request are two dax topic branches (dax
error handling, and dax radix-tree locking). These topics were
deferred to get a few more days of -next integration testing, and to
coordinate a branch baseline with Ted and the ext4 tree. Vishal and
Ross will send the error handling and locking topics respectively in
the next few days.
This branch has received a positive build result from the kbuild robot
across 226 configs.
Summary:
- Device DAX for persistent memory: Device DAX is the device-centric
analogue of Filesystem DAX (CONFIG_FS_DAX). It allows memory
ranges to be allocated and mapped without need of an intervening
file system. Device DAX is strict, precise and predictable.
Specifically this interface:
a) Guarantees fault granularity with respect to a given page size
(pte, pmd, or pud) set at configuration time.
b) Enforces deterministic behavior by being strict about what
fault scenarios are supported.
Persistent memory is the first target, but the mechanism is also
targeted for exclusive allocations of performance/feature
differentiated memory ranges.
- Support for the HPE DSM (device specific method) command formats.
This enables management of these first generation devices until a
unified DSM specification materializes.
- Further ACPI 6.1 compliance with support for the common dimm
identifier format.
- Various fixes and cleanups across the subsystem"
* tag 'libnvdimm-for-4.7' of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm: (40 commits)
libnvdimm, dax: fix deletion
libnvdimm, dax: fix alignment validation
libnvdimm, dax: autodetect support
libnvdimm: release ida resources
Revert "block: enable dax for raw block devices"
/dev/dax, core: file operations and dax-mmap
/dev/dax, pmem: direct access to persistent memory
libnvdimm: stop requiring a driver ->remove() method
libnvdimm, dax: record the specified alignment of a dax-device instance
libnvdimm, dax: reserve space to store labels for device-dax
libnvdimm, dax: introduce device-dax infrastructure
nfit: add sysfs dimm 'family' and 'dsm_mask' attributes
tools/testing/nvdimm: ND_CMD_CALL support
nfit: disable vendor specific commands
nfit: export subsystem ids as attributes
nfit: fix format interface code byte order per ACPI6.1
nfit, libnvdimm: limited/whitelisted dimm command marshaling mechanism
nfit, libnvdimm: clarify "commands" vs "_DSMs"
libnvdimm: increase max envelope size for ioctl
acpi/nfit: Add sysfs "id" for NVDIMM ID
...
Diffstat (limited to 'drivers')
-rw-r--r-- | drivers/Kconfig | 2 | ||||
-rw-r--r-- | drivers/Makefile | 1 | ||||
-rw-r--r-- | drivers/acpi/nfit.c | 282 | ||||
-rw-r--r-- | drivers/acpi/nfit.h | 31 | ||||
-rw-r--r-- | drivers/acpi/utils.c | 4 | ||||
-rw-r--r-- | drivers/dax/Kconfig | 26 | ||||
-rw-r--r-- | drivers/dax/Makefile | 4 | ||||
-rw-r--r-- | drivers/dax/dax.c | 575 | ||||
-rw-r--r-- | drivers/dax/dax.h | 24 | ||||
-rw-r--r-- | drivers/dax/pmem.c | 158 | ||||
-rw-r--r-- | drivers/nvdimm/Kconfig | 13 | ||||
-rw-r--r-- | drivers/nvdimm/Makefile | 1 | ||||
-rw-r--r-- | drivers/nvdimm/blk.c | 208 | ||||
-rw-r--r-- | drivers/nvdimm/btt.c | 26 | ||||
-rw-r--r-- | drivers/nvdimm/btt_devs.c | 24 | ||||
-rw-r--r-- | drivers/nvdimm/bus.c | 63 | ||||
-rw-r--r-- | drivers/nvdimm/claim.c | 86 | ||||
-rw-r--r-- | drivers/nvdimm/core.c | 5 | ||||
-rw-r--r-- | drivers/nvdimm/dax_devs.c | 134 | ||||
-rw-r--r-- | drivers/nvdimm/dimm_devs.c | 23 | ||||
-rw-r--r-- | drivers/nvdimm/namespace_devs.c | 38 | ||||
-rw-r--r-- | drivers/nvdimm/nd-core.h | 6 | ||||
-rw-r--r-- | drivers/nvdimm/nd.h | 83 | ||||
-rw-r--r-- | drivers/nvdimm/pfn.h | 5 | ||||
-rw-r--r-- | drivers/nvdimm/pfn_devs.c | 315 | ||||
-rw-r--r-- | drivers/nvdimm/pmem.c | 493 | ||||
-rw-r--r-- | drivers/nvdimm/region.c | 2 | ||||
-rw-r--r-- | drivers/nvdimm/region_devs.c | 34 |
28 files changed, 2017 insertions, 649 deletions
diff --git a/drivers/Kconfig b/drivers/Kconfig index 430f761b0d8d..e1e2066cecdb 100644 --- a/drivers/Kconfig +++ b/drivers/Kconfig @@ -192,6 +192,8 @@ source "drivers/android/Kconfig" source "drivers/nvdimm/Kconfig" +source "drivers/dax/Kconfig" + source "drivers/nvmem/Kconfig" source "drivers/hwtracing/stm/Kconfig" diff --git a/drivers/Makefile b/drivers/Makefile index 8f5d076baeb0..0b6f3d60193d 100644 --- a/drivers/Makefile +++ b/drivers/Makefile @@ -66,6 +66,7 @@ obj-$(CONFIG_PARPORT) += parport/ obj-$(CONFIG_NVM) += lightnvm/ obj-y += base/ block/ misc/ mfd/ nfc/ obj-$(CONFIG_LIBNVDIMM) += nvdimm/ +obj-$(CONFIG_DEV_DAX) += dax/ obj-$(CONFIG_DMA_SHARED_BUFFER) += dma-buf/ obj-$(CONFIG_NUBUS) += nubus/ obj-y += macintosh/ diff --git a/drivers/acpi/nfit.c b/drivers/acpi/nfit.c index 63cc9dbe4f3b..2215fc847fa9 100644 --- a/drivers/acpi/nfit.c +++ b/drivers/acpi/nfit.c @@ -45,6 +45,11 @@ module_param(scrub_overflow_abort, uint, S_IRUGO|S_IWUSR); MODULE_PARM_DESC(scrub_overflow_abort, "Number of times we overflow ARS results before abort"); +static bool disable_vendor_specific; +module_param(disable_vendor_specific, bool, S_IRUGO); +MODULE_PARM_DESC(disable_vendor_specific, + "Limit commands to the publicly specified set\n"); + static struct workqueue_struct *nfit_wq; struct nfit_table_prev { @@ -171,33 +176,46 @@ static int acpi_nfit_ctl(struct nvdimm_bus_descriptor *nd_desc, unsigned int buf_len, int *cmd_rc) { struct acpi_nfit_desc *acpi_desc = to_acpi_nfit_desc(nd_desc); - const struct nd_cmd_desc *desc = NULL; union acpi_object in_obj, in_buf, *out_obj; + const struct nd_cmd_desc *desc = NULL; struct device *dev = acpi_desc->dev; + struct nd_cmd_pkg *call_pkg = NULL; const char *cmd_name, *dimm_name; - unsigned long dsm_mask; + unsigned long cmd_mask, dsm_mask; acpi_handle handle; + unsigned int func; const u8 *uuid; u32 offset; int rc, i; + func = cmd; + if (cmd == ND_CMD_CALL) { + call_pkg = buf; + func = call_pkg->nd_command; + } + if (nvdimm) { struct nfit_mem *nfit_mem = nvdimm_provider_data(nvdimm); struct acpi_device *adev = nfit_mem->adev; if (!adev) return -ENOTTY; + if (call_pkg && nfit_mem->family != call_pkg->nd_family) + return -ENOTTY; + dimm_name = nvdimm_name(nvdimm); cmd_name = nvdimm_cmd_name(cmd); + cmd_mask = nvdimm_cmd_mask(nvdimm); dsm_mask = nfit_mem->dsm_mask; desc = nd_cmd_dimm_desc(cmd); - uuid = to_nfit_uuid(NFIT_DEV_DIMM); + uuid = to_nfit_uuid(nfit_mem->family); handle = adev->handle; } else { struct acpi_device *adev = to_acpi_dev(acpi_desc); cmd_name = nvdimm_bus_cmd_name(cmd); - dsm_mask = nd_desc->dsm_mask; + cmd_mask = nd_desc->cmd_mask; + dsm_mask = cmd_mask; desc = nd_cmd_bus_desc(cmd); uuid = to_nfit_uuid(NFIT_DEV_BUS); handle = adev->handle; @@ -207,7 +225,7 @@ static int acpi_nfit_ctl(struct nvdimm_bus_descriptor *nd_desc, if (!desc || (cmd && (desc->out_num + desc->in_num == 0))) return -ENOTTY; - if (!test_bit(cmd, &dsm_mask)) + if (!test_bit(cmd, &cmd_mask) || !test_bit(func, &dsm_mask)) return -ENOTTY; in_obj.type = ACPI_TYPE_PACKAGE; @@ -222,21 +240,44 @@ static int acpi_nfit_ctl(struct nvdimm_bus_descriptor *nd_desc, in_buf.buffer.length += nd_cmd_in_size(nvdimm, cmd, desc, i, buf); + if (call_pkg) { + /* skip over package wrapper */ + in_buf.buffer.pointer = (void *) &call_pkg->nd_payload; + in_buf.buffer.length = call_pkg->nd_size_in; + } + if (IS_ENABLED(CONFIG_ACPI_NFIT_DEBUG)) { - dev_dbg(dev, "%s:%s cmd: %s input length: %d\n", __func__, - dimm_name, cmd_name, in_buf.buffer.length); - print_hex_dump_debug(cmd_name, DUMP_PREFIX_OFFSET, 4, - 4, in_buf.buffer.pointer, min_t(u32, 128, - in_buf.buffer.length), true); + dev_dbg(dev, "%s:%s cmd: %d: func: %d input length: %d\n", + __func__, dimm_name, cmd, func, + in_buf.buffer.length); + print_hex_dump_debug("nvdimm in ", DUMP_PREFIX_OFFSET, 4, 4, + in_buf.buffer.pointer, + min_t(u32, 256, in_buf.buffer.length), true); } - out_obj = acpi_evaluate_dsm(handle, uuid, 1, cmd, &in_obj); + out_obj = acpi_evaluate_dsm(handle, uuid, 1, func, &in_obj); if (!out_obj) { dev_dbg(dev, "%s:%s _DSM failed cmd: %s\n", __func__, dimm_name, cmd_name); return -EINVAL; } + if (call_pkg) { + call_pkg->nd_fw_size = out_obj->buffer.length; + memcpy(call_pkg->nd_payload + call_pkg->nd_size_in, + out_obj->buffer.pointer, + min(call_pkg->nd_fw_size, call_pkg->nd_size_out)); + + ACPI_FREE(out_obj); + /* + * Need to support FW function w/o known size in advance. + * Caller can determine required size based upon nd_fw_size. + * If we return an error (like elsewhere) then caller wouldn't + * be able to rely upon data returned to make calculation. + */ + return 0; + } + if (out_obj->package.type != ACPI_TYPE_BUFFER) { dev_dbg(dev, "%s:%s unexpected output object type cmd: %s type: %d\n", __func__, dimm_name, cmd_name, out_obj->type); @@ -658,6 +699,7 @@ static int nfit_mem_dcr_init(struct acpi_nfit_desc *acpi_desc, if (!nfit_mem) return -ENOMEM; INIT_LIST_HEAD(&nfit_mem->list); + nfit_mem->acpi_desc = acpi_desc; list_add(&nfit_mem->list, &acpi_desc->dimms); } @@ -819,7 +861,7 @@ static ssize_t vendor_show(struct device *dev, { struct acpi_nfit_control_region *dcr = to_nfit_dcr(dev); - return sprintf(buf, "%#x\n", dcr->vendor_id); + return sprintf(buf, "0x%04x\n", be16_to_cpu(dcr->vendor_id)); } static DEVICE_ATTR_RO(vendor); @@ -828,7 +870,7 @@ static ssize_t rev_id_show(struct device *dev, { struct acpi_nfit_control_region *dcr = to_nfit_dcr(dev); - return sprintf(buf, "%#x\n", dcr->revision_id); + return sprintf(buf, "0x%04x\n", be16_to_cpu(dcr->revision_id)); } static DEVICE_ATTR_RO(rev_id); @@ -837,28 +879,142 @@ static ssize_t device_show(struct device *dev, { struct acpi_nfit_control_region *dcr = to_nfit_dcr(dev); - return sprintf(buf, "%#x\n", dcr->device_id); + return sprintf(buf, "0x%04x\n", be16_to_cpu(dcr->device_id)); } static DEVICE_ATTR_RO(device); +static ssize_t subsystem_vendor_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct acpi_nfit_control_region *dcr = to_nfit_dcr(dev); + + return sprintf(buf, "0x%04x\n", be16_to_cpu(dcr->subsystem_vendor_id)); +} +static DEVICE_ATTR_RO(subsystem_vendor); + +static ssize_t subsystem_rev_id_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct acpi_nfit_control_region *dcr = to_nfit_dcr(dev); + + return sprintf(buf, "0x%04x\n", + be16_to_cpu(dcr->subsystem_revision_id)); +} +static DEVICE_ATTR_RO(subsystem_rev_id); + +static ssize_t subsystem_device_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct acpi_nfit_control_region *dcr = to_nfit_dcr(dev); + + return sprintf(buf, "0x%04x\n", be16_to_cpu(dcr->subsystem_device_id)); +} +static DEVICE_ATTR_RO(subsystem_device); + +static int num_nvdimm_formats(struct nvdimm *nvdimm) +{ + struct nfit_mem *nfit_mem = nvdimm_provider_data(nvdimm); + int formats = 0; + + if (nfit_mem->memdev_pmem) + formats++; + if (nfit_mem->memdev_bdw) + formats++; + return formats; +} + static ssize_t format_show(struct device *dev, struct device_attribute *attr, char *buf) { struct acpi_nfit_control_region *dcr = to_nfit_dcr(dev); - return sprintf(buf, "%#x\n", dcr->code); + return sprintf(buf, "0x%04x\n", be16_to_cpu(dcr->code)); } static DEVICE_ATTR_RO(format); +static ssize_t format1_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + u32 handle; + ssize_t rc = -ENXIO; + struct nfit_mem *nfit_mem; + struct nfit_memdev *nfit_memdev; + struct acpi_nfit_desc *acpi_desc; + struct nvdimm *nvdimm = to_nvdimm(dev); + struct acpi_nfit_control_region *dcr = to_nfit_dcr(dev); + + nfit_mem = nvdimm_provider_data(nvdimm); + acpi_desc = nfit_mem->acpi_desc; + handle = to_nfit_memdev(dev)->device_handle; + + /* assumes DIMMs have at most 2 published interface codes */ + mutex_lock(&acpi_desc->init_mutex); + list_for_each_entry(nfit_memdev, &acpi_desc->memdevs, list) { + struct acpi_nfit_memory_map *memdev = nfit_memdev->memdev; + struct nfit_dcr *nfit_dcr; + + if (memdev->device_handle != handle) + continue; + + list_for_each_entry(nfit_dcr, &acpi_desc->dcrs, list) { + if (nfit_dcr->dcr->region_index != memdev->region_index) + continue; + if (nfit_dcr->dcr->code == dcr->code) + continue; + rc = sprintf(buf, "%#x\n", + be16_to_cpu(nfit_dcr->dcr->code)); + break; + } + if (rc != ENXIO) + break; + } + mutex_unlock(&acpi_desc->init_mutex); + return rc; +} +static DEVICE_ATTR_RO(format1); + +static ssize_t formats_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct nvdimm *nvdimm = to_nvdimm(dev); + + return sprintf(buf, "%d\n", num_nvdimm_formats(nvdimm)); +} +static DEVICE_ATTR_RO(formats); + static ssize_t serial_show(struct device *dev, struct device_attribute *attr, char *buf) { struct acpi_nfit_control_region *dcr = to_nfit_dcr(dev); - return sprintf(buf, "%#x\n", dcr->serial_number); + return sprintf(buf, "0x%08x\n", be32_to_cpu(dcr->serial_number)); } static DEVICE_ATTR_RO(serial); +static ssize_t family_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct nvdimm *nvdimm = to_nvdimm(dev); + struct nfit_mem *nfit_mem = nvdimm_provider_data(nvdimm); + + if (nfit_mem->family < 0) + return -ENXIO; + return sprintf(buf, "%d\n", nfit_mem->family); +} +static DEVICE_ATTR_RO(family); + +static ssize_t dsm_mask_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct nvdimm *nvdimm = to_nvdimm(dev); + struct nfit_mem *nfit_mem = nvdimm_provider_data(nvdimm); + + if (nfit_mem->family < 0) + return -ENXIO; + return sprintf(buf, "%#lx\n", nfit_mem->dsm_mask); +} +static DEVICE_ATTR_RO(dsm_mask); + static ssize_t flags_show(struct device *dev, struct device_attribute *attr, char *buf) { @@ -873,15 +1029,41 @@ static ssize_t flags_show(struct device *dev, } static DEVICE_ATTR_RO(flags); +static ssize_t id_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct acpi_nfit_control_region *dcr = to_nfit_dcr(dev); + + if (dcr->valid_fields & ACPI_NFIT_CONTROL_MFG_INFO_VALID) + return sprintf(buf, "%04x-%02x-%04x-%08x\n", + be16_to_cpu(dcr->vendor_id), + dcr->manufacturing_location, + be16_to_cpu(dcr->manufacturing_date), + be32_to_cpu(dcr->serial_number)); + else + return sprintf(buf, "%04x-%08x\n", + be16_to_cpu(dcr->vendor_id), + be32_to_cpu(dcr->serial_number)); +} +static DEVICE_ATTR_RO(id); + static struct attribute *acpi_nfit_dimm_attributes[] = { &dev_attr_handle.attr, &dev_attr_phys_id.attr, &dev_attr_vendor.attr, &dev_attr_device.attr, + &dev_attr_rev_id.attr, + &dev_attr_subsystem_vendor.attr, + &dev_attr_subsystem_device.attr, + &dev_attr_subsystem_rev_id.attr, &dev_attr_format.attr, + &dev_attr_formats.attr, + &dev_attr_format1.attr, &dev_attr_serial.attr, - &dev_attr_rev_id.attr, &dev_attr_flags.attr, + &dev_attr_id.attr, + &dev_attr_family.attr, + &dev_attr_dsm_mask.attr, NULL, }; @@ -889,11 +1071,13 @@ static umode_t acpi_nfit_dimm_attr_visible(struct kobject *kobj, struct attribute *a, int n) { struct device *dev = container_of(kobj, struct device, kobj); + struct nvdimm *nvdimm = to_nvdimm(dev); - if (to_nfit_dcr(dev)) - return a->mode; - else + if (!to_nfit_dcr(dev)) return 0; + if (a == &dev_attr_format1.attr && num_nvdimm_formats(nvdimm) <= 1) + return 0; + return a->mode; } static struct attribute_group acpi_nfit_dimm_attribute_group = { @@ -926,10 +1110,13 @@ static int acpi_nfit_add_dimm(struct acpi_nfit_desc *acpi_desc, { struct acpi_device *adev, *adev_dimm; struct device *dev = acpi_desc->dev; - const u8 *uuid = to_nfit_uuid(NFIT_DEV_DIMM); + unsigned long dsm_mask; + const u8 *uuid; int i; - nfit_mem->dsm_mask = acpi_desc->dimm_dsm_force_en; + /* nfit test assumes 1:1 relationship between commands and dsms */ + nfit_mem->dsm_mask = acpi_desc->dimm_cmd_force_en; + nfit_mem->family = NVDIMM_FAMILY_INTEL; adev = to_acpi_dev(acpi_desc); if (!adev) return 0; @@ -942,7 +1129,35 @@ static int acpi_nfit_add_dimm(struct acpi_nfit_desc *acpi_desc, return force_enable_dimms ? 0 : -ENODEV; } - for (i = ND_CMD_SMART; i <= ND_CMD_VENDOR; i++) + /* + * Until standardization materializes we need to consider up to 3 + * different command sets. Note, that checking for function0 (bit0) + * tells us if any commands are reachable through this uuid. + */ + for (i = NVDIMM_FAMILY_INTEL; i <= NVDIMM_FAMILY_HPE2; i++) + if (acpi_check_dsm(adev_dimm->handle, to_nfit_uuid(i), 1, 1)) + break; + + /* limit the supported commands to those that are publicly documented */ + nfit_mem->family = i; + if (nfit_mem->family == NVDIMM_FAMILY_INTEL) { + dsm_mask = 0x3fe; + if (disable_vendor_specific) + dsm_mask &= ~(1 << ND_CMD_VENDOR); + } else if (nfit_mem->family == NVDIMM_FAMILY_HPE1) + dsm_mask = 0x1c3c76; + else if (nfit_mem->family == NVDIMM_FAMILY_HPE2) { + dsm_mask = 0x1fe; + if (disable_vendor_specific) + dsm_mask &= ~(1 << 8); + } else { + dev_err(dev, "unknown dimm command family\n"); + nfit_mem->family = -1; + return force_enable_dimms ? 0 : -ENODEV; + } + + uuid = to_nfit_uuid(nfit_mem->family); + for_each_set_bit(i, &dsm_mask, BITS_PER_LONG) if (acpi_check_dsm(adev_dimm->handle, uuid, 1, 1ULL << i)) set_bit(i, &nfit_mem->dsm_mask); @@ -955,8 +1170,8 @@ static int acpi_nfit_register_dimms(struct acpi_nfit_desc *acpi_desc) int dimm_count = 0; list_for_each_entry(nfit_mem, &acpi_desc->dimms, list) { + unsigned long flags = 0, cmd_mask; struct nvdimm *nvdimm; - unsigned long flags = 0; u32 device_handle; u16 mem_flags; int rc; @@ -979,9 +1194,18 @@ static int acpi_nfit_register_dimms(struct acpi_nfit_desc *acpi_desc) if (rc) continue; + /* + * TODO: provide translation for non-NVDIMM_FAMILY_INTEL + * devices (i.e. from nd_cmd to acpi_dsm) to standardize the + * userspace interface. + */ + cmd_mask = 1UL << ND_CMD_CALL; + if (nfit_mem->family == NVDIMM_FAMILY_INTEL) + cmd_mask |= nfit_mem->dsm_mask; + nvdimm = nvdimm_create(acpi_desc->nvdimm_bus, nfit_mem, acpi_nfit_dimm_attribute_groups, - flags, &nfit_mem->dsm_mask); + flags, cmd_mask); if (!nvdimm) return -ENOMEM; @@ -1010,14 +1234,14 @@ static void acpi_nfit_init_dsms(struct acpi_nfit_desc *acpi_desc) struct acpi_device *adev; int i; - nd_desc->dsm_mask = acpi_desc->bus_dsm_force_en; + nd_desc->cmd_mask = acpi_desc->bus_cmd_force_en; adev = to_acpi_dev(acpi_desc); if (!adev) return; for (i = ND_CMD_ARS_CAP; i <= ND_CMD_CLEAR_ERROR; i++) if (acpi_check_dsm(adev->handle, uuid, 1, 1ULL << i)) - set_bit(i, &nd_desc->dsm_mask); + set_bit(i, &nd_desc->cmd_mask); } static ssize_t range_index_show(struct device *dev, @@ -2309,7 +2533,7 @@ static int acpi_nfit_add(struct acpi_device *adev) acpi_size sz; int rc; - status = acpi_get_table_with_size("NFIT", 0, &tbl, &sz); + status = acpi_get_table_with_size(ACPI_SIG_NFIT, 0, &tbl, &sz); if (ACPI_FAILURE(status)) { /* This is ok, we could have an nvdimm hotplugged later */ dev_dbg(dev, "failed to find NFIT at startup\n"); @@ -2466,6 +2690,8 @@ static __init int nfit_init(void) acpi_str_to_uuid(UUID_PERSISTENT_VIRTUAL_CD, nfit_uuid[NFIT_SPA_PCD]); acpi_str_to_uuid(UUID_NFIT_BUS, nfit_uuid[NFIT_DEV_BUS]); acpi_str_to_uuid(UUID_NFIT_DIMM, nfit_uuid[NFIT_DEV_DIMM]); + acpi_str_to_uuid(UUID_NFIT_DIMM_N_HPE1, nfit_uuid[NFIT_DEV_DIMM_N_HPE1]); + acpi_str_to_uuid(UUID_NFIT_DIMM_N_HPE2, nfit_uuid[NFIT_DEV_DIMM_N_HPE2]); nfit_wq = create_singlethread_workqueue("nfit"); if (!nfit_wq) diff --git a/drivers/acpi/nfit.h b/drivers/acpi/nfit.h index c75576b2d50e..11cb38348aef 100644 --- a/drivers/acpi/nfit.h +++ b/drivers/acpi/nfit.h @@ -21,13 +21,25 @@ #include <linux/acpi.h> #include <acpi/acuuid.h> +/* ACPI 6.1 */ #define UUID_NFIT_BUS "2f10e7a4-9e91-11e4-89d3-123b93f75cba" + +/* http://pmem.io/documents/NVDIMM_DSM_Interface_Example.pdf */ #define UUID_NFIT_DIMM "4309ac30-0d11-11e4-9191-0800200c9a66" + +/* https://github.com/HewlettPackard/hpe-nvm/blob/master/Documentation/ */ +#define UUID_NFIT_DIMM_N_HPE1 "9002c334-acf3-4c0e-9642-a235f0d53bc6" +#define UUID_NFIT_DIMM_N_HPE2 "5008664b-b758-41a0-a03c-27c2f2d04f7e" + #define ACPI_NFIT_MEM_FAILED_MASK (ACPI_NFIT_MEM_SAVE_FAILED \ | ACPI_NFIT_MEM_RESTORE_FAILED | ACPI_NFIT_MEM_FLUSH_FAILED \ | ACPI_NFIT_MEM_NOT_ARMED) enum nfit_uuids { + /* for simplicity alias the uuid index with the family id */ + NFIT_DEV_DIMM = NVDIMM_FAMILY_INTEL, + NFIT_DEV_DIMM_N_HPE1 = NVDIMM_FAMILY_HPE1, + NFIT_DEV_DIMM_N_HPE2 = NVDIMM_FAMILY_HPE2, NFIT_SPA_VOLATILE, NFIT_SPA_PM, NFIT_SPA_DCR, @@ -37,15 +49,16 @@ enum nfit_uuids { NFIT_SPA_PDISK, NFIT_SPA_PCD, NFIT_DEV_BUS, - NFIT_DEV_DIMM, NFIT_UUID_MAX, }; -enum nfit_fic { - NFIT_FIC_BYTE = 0x101, /* byte-addressable energy backed */ - NFIT_FIC_BLK = 0x201, /* block-addressable non-energy backed */ - NFIT_FIC_BYTEN = 0x301, /* byte-addressable non-energy backed */ -}; +/* + * Region format interface codes are stored as an array of bytes in the + * NFIT DIMM Control Region structure + */ +#define NFIT_FIC_BYTE cpu_to_be16(0x101) /* byte-addressable energy backed */ +#define NFIT_FIC_BLK cpu_to_be16(0x201) /* block-addressable non-energy backed */ +#define NFIT_FIC_BYTEN cpu_to_be16(0x301) /* byte-addressable non-energy backed */ enum { NFIT_BLK_READ_FLUSH = 1, @@ -109,7 +122,9 @@ struct nfit_mem { struct nfit_flush *nfit_flush; struct list_head list; struct acpi_device *adev; + struct acpi_nfit_desc *acpi_desc; unsigned long dsm_mask; + int family; }; struct acpi_nfit_desc { @@ -132,8 +147,8 @@ struct acpi_nfit_desc { size_t ars_status_size; struct work_struct work; unsigned int cancel:1; - unsigned long dimm_dsm_force_en; - unsigned long bus_dsm_force_en; + unsigned long dimm_cmd_force_en; + unsigned long bus_cmd_force_en; int (*blk_do_io)(struct nd_blk_region *ndbr, resource_size_t dpa, void *iobuf, u64 len, int rw); }; diff --git a/drivers/acpi/utils.c b/drivers/acpi/utils.c index ac832bf6f8c9..22c09952e177 100644 --- a/drivers/acpi/utils.c +++ b/drivers/acpi/utils.c @@ -625,7 +625,7 @@ acpi_status acpi_evaluate_lck(acpi_handle handle, int lock) * some old BIOSes do expect a buffer or an integer etc. */ union acpi_object * -acpi_evaluate_dsm(acpi_handle handle, const u8 *uuid, int rev, int func, +acpi_evaluate_dsm(acpi_handle handle, const u8 *uuid, u64 rev, u64 func, union acpi_object *argv4) { acpi_status ret; @@ -674,7 +674,7 @@ EXPORT_SYMBOL(acpi_evaluate_dsm); * functions. Currently only support 64 functions at maximum, should be * enough for now. */ -bool acpi_check_dsm(acpi_handle handle, const u8 *uuid, int rev, u64 funcs) +bool acpi_check_dsm(acpi_handle handle, const u8 *uuid, u64 rev, u64 funcs) { int i; u64 mask = 0; diff --git a/drivers/dax/Kconfig b/drivers/dax/Kconfig new file mode 100644 index 000000000000..cedab7572de3 --- /dev/null +++ b/drivers/dax/Kconfig @@ -0,0 +1,26 @@ +menuconfig DEV_DAX + tristate "DAX: direct access to differentiated memory" + default m if NVDIMM_DAX + depends on TRANSPARENT_HUGEPAGE + help + Support raw access to differentiated (persistence, bandwidth, + latency...) memory via an mmap(2) capable character + device. Platform firmware or a device driver may identify a + platform memory resource that is differentiated from the + baseline memory pool. Mappings of a /dev/daxX.Y device impose + restrictions that make the mapping behavior deterministic. + +if DEV_DAX + +config DEV_DAX_PMEM + tristate "PMEM DAX: direct access to persistent memory" + depends on NVDIMM_DAX + default DEV_DAX + help + Support raw access to persistent memory. Note that this + driver consumes memory ranges allocated and exported by the + libnvdimm sub-system. + + Say Y if unsure + +endif diff --git a/drivers/dax/Makefile b/drivers/dax/Makefile new file mode 100644 index 000000000000..27c54e38478a --- /dev/null +++ b/drivers/dax/Makefile @@ -0,0 +1,4 @@ +obj-$(CONFIG_DEV_DAX) += dax.o +obj-$(CONFIG_DEV_DAX_PMEM) += dax_pmem.o + +dax_pmem-y := pmem.o diff --git a/drivers/dax/dax.c b/drivers/dax/dax.c new file mode 100644 index 000000000000..b891a129b275 --- /dev/null +++ b/drivers/dax/dax.c @@ -0,0 +1,575 @@ +/* + * Copyright(c) 2016 Intel Corporation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of version 2 of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + */ +#include <linux/pagemap.h> +#include <linux/module.h> +#include <linux/device.h> +#include <linux/pfn_t.h> +#include <linux/slab.h> +#include <linux/dax.h> +#include <linux/fs.h> +#include <linux/mm.h> + +static int dax_major; +static struct class *dax_class; +static DEFINE_IDA(dax_minor_ida); + +/** + * struct dax_region - mapping infrastructure for dax devices + * @id: kernel-wide unique region for a memory range + * @base: linear address corresponding to @res + * @kref: to pin while other agents have a need to do lookups + * @dev: parent device backing this region + * @align: allocation and mapping alignment for child dax devices + * @res: physical address range of the region + * @pfn_flags: identify whether the pfns are paged back or not + */ +struct dax_region { + int id; + struct ida ida; + void *base; + struct kref kref; + struct device *dev; + unsigned int align; + struct resource res; + unsigned long pfn_flags; +}; + +/** + * struct dax_dev - subdivision of a dax region + * @region - parent region + * @dev - device backing the character device + * @kref - enable this data to be tracked in filp->private_data + * @alive - !alive + rcu grace period == no new mappings can be established + * @id - child id in the region + * @num_resources - number of physical address extents in this device + * @res - array of physical address ranges + */ +struct dax_dev { + struct dax_region *region; + struct device *dev; + struct kref kref; + bool alive; + int id; + int num_resources; + struct resource res[0]; +}; + +static void dax_region_free(struct kref *kref) +{ + struct dax_region *dax_region; + + dax_region = container_of(kref, struct dax_region, kref); + kfree(dax_region); +} + +void dax_region_put(struct dax_region *dax_region) +{ + kref_put(&dax_region->kref, dax_region_free); +} +EXPORT_SYMBOL_GPL(dax_region_put); + +static void dax_dev_free(struct kref *kref) +{ + struct dax_dev *dax_dev; + + dax_dev = container_of(kref, struct dax_dev, kref); + dax_region_put(dax_dev->region); + kfree(dax_dev); +} + +static void dax_dev_put(struct dax_dev *dax_dev) +{ + kref_put(&dax_dev->kref, dax_dev_free); +} + +struct dax_region *alloc_dax_region(struct device *parent, int region_id, + struct resource *res, unsigned int align, void *addr, + unsigned long pfn_flags) +{ + struct dax_region *dax_region; + + dax_region = kzalloc(sizeof(*dax_region), GFP_KERNEL); + + if (!dax_region) + return NULL; + + memcpy(&dax_region->res, res, sizeof(*res)); + dax_region->pfn_flags = pfn_flags; + kref_init(&dax_region->kref); + dax_region->id = region_id; + ida_init(&dax_region->ida); + dax_region->align = align; + dax_region->dev = parent; + dax_region->base = addr; + + return dax_region; +} +EXPORT_SYMBOL_GPL(alloc_dax_region); + +static ssize_t size_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct dax_dev *dax_dev = dev_get_drvdata(dev); + unsigned long long size = 0; + int i; + + for (i = 0; i < dax_dev->num_resources; i++) + size += resource_size(&dax_dev->res[i]); + + return sprintf(buf, "%llu\n", size); +} +static DEVICE_ATTR_RO(size); + +static struct attribute *dax_device_attributes[] = { + &dev_attr_size.attr, + NULL, +}; + +static const struct attribute_group dax_device_attribute_group = { + .attrs = dax_device_attributes, +}; + +static const struct attribute_group *dax_attribute_groups[] = { + &dax_device_attribute_group, + NULL, +}; + +static void unregister_dax_dev(void *_dev) +{ + struct device *dev = _dev; + struct dax_dev *dax_dev = dev_get_drvdata(dev); + struct dax_region *dax_region = dax_dev->region; + + dev_dbg(dev, "%s\n", __func__); + + /* + * Note, rcu is not protecting the liveness of dax_dev, rcu is + * ensuring that any fault handlers that might have seen + * dax_dev->alive == true, have completed. Any fault handlers + * that start after synchronize_rcu() has started will abort + * upon seeing dax_dev->alive == false. + */ + dax_dev->alive = false; + synchronize_rcu(); + + get_device(dev); + device_unregister(dev); + ida_simple_remove(&dax_region->ida, dax_dev->id); + ida_simple_remove(&dax_minor_ida, MINOR(dev->devt)); + put_device(dev); + dax_dev_put(dax_dev); +} + +int devm_create_dax_dev(struct dax_region *dax_region, struct resource *res, + int count) +{ + struct device *parent = dax_region->dev; + struct dax_dev *dax_dev; + struct device *dev; + int rc, minor; + dev_t dev_t; + + dax_dev = kzalloc(sizeof(*dax_dev) + sizeof(*res) * count, GFP_KERNEL); + |