1 // SPDX-License-Identifier: GPL-2.0
2 /* Copyright(c) 2016-2018 Intel Corporation. All rights reserved. */
3 #include <linux/memremap.h>
4 #include <linux/pagemap.h>
5 #include <linux/module.h>
6 #include <linux/device.h>
7 #include <linux/pfn_t.h>
8 #include <linux/cdev.h>
9 #include <linux/slab.h>
10 #include <linux/dax.h>
13 #include <linux/mman.h>
14 #include "dax-private.h"
17 static int check_vma(struct dev_dax
*dev_dax
, struct vm_area_struct
*vma
,
20 struct device
*dev
= &dev_dax
->dev
;
23 if (!dax_alive(dev_dax
->dax_dev
))
26 /* prevent private mappings from being established */
27 if ((vma
->vm_flags
& VM_MAYSHARE
) != VM_MAYSHARE
) {
28 dev_info_ratelimited(dev
,
29 "%s: %s: fail, attempted private mapping\n",
34 mask
= dev_dax
->align
- 1;
35 if (vma
->vm_start
& mask
|| vma
->vm_end
& mask
) {
36 dev_info_ratelimited(dev
,
37 "%s: %s: fail, unaligned vma (%#lx - %#lx, %#lx)\n",
38 current
->comm
, func
, vma
->vm_start
, vma
->vm_end
,
43 if (!vma_is_dax(vma
)) {
44 dev_info_ratelimited(dev
,
45 "%s: %s: fail, vma is not DAX capable\n",
53 /* see "strong" declaration in tools/testing/nvdimm/dax-dev.c */
54 __weak phys_addr_t
dax_pgoff_to_phys(struct dev_dax
*dev_dax
, pgoff_t pgoff
,
59 for (i
= 0; i
< dev_dax
->nr_range
; i
++) {
60 struct dev_dax_range
*dax_range
= &dev_dax
->ranges
[i
];
61 struct range
*range
= &dax_range
->range
;
62 unsigned long long pgoff_end
;
65 pgoff_end
= dax_range
->pgoff
+ PHYS_PFN(range_len(range
)) - 1;
66 if (pgoff
< dax_range
->pgoff
|| pgoff
> pgoff_end
)
68 phys
= PFN_PHYS(pgoff
- dax_range
->pgoff
) + range
->start
;
69 if (phys
+ size
- 1 <= range
->end
)
76 static vm_fault_t
__dev_dax_pte_fault(struct dev_dax
*dev_dax
,
77 struct vm_fault
*vmf
, pfn_t
*pfn
)
79 struct device
*dev
= &dev_dax
->dev
;
81 unsigned int fault_size
= PAGE_SIZE
;
83 if (check_vma(dev_dax
, vmf
->vma
, __func__
))
84 return VM_FAULT_SIGBUS
;
86 if (dev_dax
->align
> PAGE_SIZE
) {
87 dev_dbg(dev
, "alignment (%#x) > fault size (%#x)\n",
88 dev_dax
->align
, fault_size
);
89 return VM_FAULT_SIGBUS
;
92 if (fault_size
!= dev_dax
->align
)
93 return VM_FAULT_SIGBUS
;
95 phys
= dax_pgoff_to_phys(dev_dax
, vmf
->pgoff
, PAGE_SIZE
);
97 dev_dbg(dev
, "pgoff_to_phys(%#lx) failed\n", vmf
->pgoff
);
98 return VM_FAULT_SIGBUS
;
101 *pfn
= phys_to_pfn_t(phys
, PFN_DEV
|PFN_MAP
);
103 return vmf_insert_mixed(vmf
->vma
, vmf
->address
, *pfn
);
106 static vm_fault_t
__dev_dax_pmd_fault(struct dev_dax
*dev_dax
,
107 struct vm_fault
*vmf
, pfn_t
*pfn
)
109 unsigned long pmd_addr
= vmf
->address
& PMD_MASK
;
110 struct device
*dev
= &dev_dax
->dev
;
113 unsigned int fault_size
= PMD_SIZE
;
115 if (check_vma(dev_dax
, vmf
->vma
, __func__
))
116 return VM_FAULT_SIGBUS
;
118 if (dev_dax
->align
> PMD_SIZE
) {
119 dev_dbg(dev
, "alignment (%#x) > fault size (%#x)\n",
120 dev_dax
->align
, fault_size
);
121 return VM_FAULT_SIGBUS
;
124 if (fault_size
< dev_dax
->align
)
125 return VM_FAULT_SIGBUS
;
126 else if (fault_size
> dev_dax
->align
)
127 return VM_FAULT_FALLBACK
;
129 /* if we are outside of the VMA */
130 if (pmd_addr
< vmf
->vma
->vm_start
||
131 (pmd_addr
+ PMD_SIZE
) > vmf
->vma
->vm_end
)
132 return VM_FAULT_SIGBUS
;
134 pgoff
= linear_page_index(vmf
->vma
, pmd_addr
);
135 phys
= dax_pgoff_to_phys(dev_dax
, pgoff
, PMD_SIZE
);
137 dev_dbg(dev
, "pgoff_to_phys(%#lx) failed\n", pgoff
);
138 return VM_FAULT_SIGBUS
;
141 *pfn
= phys_to_pfn_t(phys
, PFN_DEV
|PFN_MAP
);
143 return vmf_insert_pfn_pmd(vmf
, *pfn
, vmf
->flags
& FAULT_FLAG_WRITE
);
146 #ifdef CONFIG_HAVE_ARCH_TRANSPARENT_HUGEPAGE_PUD
147 static vm_fault_t
__dev_dax_pud_fault(struct dev_dax
*dev_dax
,
148 struct vm_fault
*vmf
, pfn_t
*pfn
)
150 unsigned long pud_addr
= vmf
->address
& PUD_MASK
;
151 struct device
*dev
= &dev_dax
->dev
;
154 unsigned int fault_size
= PUD_SIZE
;
157 if (check_vma(dev_dax
, vmf
->vma
, __func__
))
158 return VM_FAULT_SIGBUS
;
160 if (dev_dax
->align
> PUD_SIZE
) {
161 dev_dbg(dev
, "alignment (%#x) > fault size (%#x)\n",
162 dev_dax
->align
, fault_size
);
163 return VM_FAULT_SIGBUS
;
166 if (fault_size
< dev_dax
->align
)
167 return VM_FAULT_SIGBUS
;
168 else if (fault_size
> dev_dax
->align
)
169 return VM_FAULT_FALLBACK
;
171 /* if we are outside of the VMA */
172 if (pud_addr
< vmf
->vma
->vm_start
||
173 (pud_addr
+ PUD_SIZE
) > vmf
->vma
->vm_end
)
174 return VM_FAULT_SIGBUS
;
176 pgoff
= linear_page_index(vmf
->vma
, pud_addr
);
177 phys
= dax_pgoff_to_phys(dev_dax
, pgoff
, PUD_SIZE
);
179 dev_dbg(dev
, "pgoff_to_phys(%#lx) failed\n", pgoff
);
180 return VM_FAULT_SIGBUS
;
183 *pfn
= phys_to_pfn_t(phys
, PFN_DEV
|PFN_MAP
);
185 return vmf_insert_pfn_pud(vmf
, *pfn
, vmf
->flags
& FAULT_FLAG_WRITE
);
188 static vm_fault_t
__dev_dax_pud_fault(struct dev_dax
*dev_dax
,
189 struct vm_fault
*vmf
, pfn_t
*pfn
)
191 return VM_FAULT_FALLBACK
;
193 #endif /* !CONFIG_HAVE_ARCH_TRANSPARENT_HUGEPAGE_PUD */
195 static vm_fault_t
dev_dax_huge_fault(struct vm_fault
*vmf
,
196 enum page_entry_size pe_size
)
198 struct file
*filp
= vmf
->vma
->vm_file
;
199 unsigned long fault_size
;
200 vm_fault_t rc
= VM_FAULT_SIGBUS
;
203 struct dev_dax
*dev_dax
= filp
->private_data
;
205 dev_dbg(&dev_dax
->dev
, "%s: %s (%#lx - %#lx) size = %d\n", current
->comm
,
206 (vmf
->flags
& FAULT_FLAG_WRITE
) ? "write" : "read",
207 vmf
->vma
->vm_start
, vmf
->vma
->vm_end
, pe_size
);
209 id
= dax_read_lock();
212 fault_size
= PAGE_SIZE
;
213 rc
= __dev_dax_pte_fault(dev_dax
, vmf
, &pfn
);
216 fault_size
= PMD_SIZE
;
217 rc
= __dev_dax_pmd_fault(dev_dax
, vmf
, &pfn
);
220 fault_size
= PUD_SIZE
;
221 rc
= __dev_dax_pud_fault(dev_dax
, vmf
, &pfn
);
224 rc
= VM_FAULT_SIGBUS
;
227 if (rc
== VM_FAULT_NOPAGE
) {
232 * In the device-dax case the only possibility for a
233 * VM_FAULT_NOPAGE result is when device-dax capacity is
234 * mapped. No need to consider the zero page, or racing
235 * conflicting mappings.
237 pgoff
= linear_page_index(vmf
->vma
, vmf
->address
238 & ~(fault_size
- 1));
239 for (i
= 0; i
< fault_size
/ PAGE_SIZE
; i
++) {
242 page
= pfn_to_page(pfn_t_to_pfn(pfn
) + i
);
245 page
->mapping
= filp
->f_mapping
;
246 page
->index
= pgoff
+ i
;
254 static vm_fault_t
dev_dax_fault(struct vm_fault
*vmf
)
256 return dev_dax_huge_fault(vmf
, PE_SIZE_PTE
);
259 static int dev_dax_may_split(struct vm_area_struct
*vma
, unsigned long addr
)
261 struct file
*filp
= vma
->vm_file
;
262 struct dev_dax
*dev_dax
= filp
->private_data
;
264 if (!IS_ALIGNED(addr
, dev_dax
->align
))
269 static unsigned long dev_dax_pagesize(struct vm_area_struct
*vma
)
271 struct file
*filp
= vma
->vm_file
;
272 struct dev_dax
*dev_dax
= filp
->private_data
;
274 return dev_dax
->align
;
277 static const struct vm_operations_struct dax_vm_ops
= {
278 .fault
= dev_dax_fault
,
279 .huge_fault
= dev_dax_huge_fault
,
280 .may_split
= dev_dax_may_split
,
281 .pagesize
= dev_dax_pagesize
,
284 static int dax_mmap(struct file
*filp
, struct vm_area_struct
*vma
)
286 struct dev_dax
*dev_dax
= filp
->private_data
;
289 dev_dbg(&dev_dax
->dev
, "trace\n");
292 * We lock to check dax_dev liveness and will re-check at
295 id
= dax_read_lock();
296 rc
= check_vma(dev_dax
, vma
, __func__
);
301 vma
->vm_ops
= &dax_vm_ops
;
302 vma
->vm_flags
|= VM_HUGEPAGE
;
306 /* return an unmapped area aligned to the dax region specified alignment */
307 static unsigned long dax_get_unmapped_area(struct file
*filp
,
308 unsigned long addr
, unsigned long len
, unsigned long pgoff
,
311 unsigned long off
, off_end
, off_align
, len_align
, addr_align
, align
;
312 struct dev_dax
*dev_dax
= filp
? filp
->private_data
: NULL
;
314 if (!dev_dax
|| addr
)
317 align
= dev_dax
->align
;
318 off
= pgoff
<< PAGE_SHIFT
;
320 off_align
= round_up(off
, align
);
322 if ((off_end
<= off_align
) || ((off_end
- off_align
) < align
))
325 len_align
= len
+ align
;
326 if ((off
+ len_align
) < off
)
329 addr_align
= current
->mm
->get_unmapped_area(filp
, addr
, len_align
,
331 if (!IS_ERR_VALUE(addr_align
)) {
332 addr_align
+= (off
- addr_align
) & (align
- 1);
336 return current
->mm
->get_unmapped_area(filp
, addr
, len
, pgoff
, flags
);
339 static const struct address_space_operations dev_dax_aops
= {
340 .set_page_dirty
= noop_set_page_dirty
,
341 .invalidatepage
= noop_invalidatepage
,
344 static int dax_open(struct inode
*inode
, struct file
*filp
)
346 struct dax_device
*dax_dev
= inode_dax(inode
);
347 struct inode
*__dax_inode
= dax_inode(dax_dev
);
348 struct dev_dax
*dev_dax
= dax_get_private(dax_dev
);
350 dev_dbg(&dev_dax
->dev
, "trace\n");
351 inode
->i_mapping
= __dax_inode
->i_mapping
;
352 inode
->i_mapping
->host
= __dax_inode
;
353 inode
->i_mapping
->a_ops
= &dev_dax_aops
;
354 filp
->f_mapping
= inode
->i_mapping
;
355 filp
->f_wb_err
= filemap_sample_wb_err(filp
->f_mapping
);
356 filp
->f_sb_err
= file_sample_sb_err(filp
);
357 filp
->private_data
= dev_dax
;
358 inode
->i_flags
= S_DAX
;
363 static int dax_release(struct inode
*inode
, struct file
*filp
)
365 struct dev_dax
*dev_dax
= filp
->private_data
;
367 dev_dbg(&dev_dax
->dev
, "trace\n");
371 static const struct file_operations dax_fops
= {
372 .llseek
= noop_llseek
,
373 .owner
= THIS_MODULE
,
375 .release
= dax_release
,
376 .get_unmapped_area
= dax_get_unmapped_area
,
378 .mmap_supported_flags
= MAP_SYNC
,
381 static void dev_dax_cdev_del(void *cdev
)
386 static void dev_dax_kill(void *dev_dax
)
388 kill_dev_dax(dev_dax
);
391 int dev_dax_probe(struct dev_dax
*dev_dax
)
393 struct dax_device
*dax_dev
= dev_dax
->dax_dev
;
394 struct device
*dev
= &dev_dax
->dev
;
395 struct dev_pagemap
*pgmap
;
401 pgmap
= dev_dax
->pgmap
;
402 if (dev_WARN_ONCE(dev
, pgmap
&& dev_dax
->nr_range
> 1,
403 "static pgmap / multi-range device conflict\n"))
407 pgmap
= devm_kzalloc(dev
, sizeof(*pgmap
) + sizeof(struct range
)
408 * (dev_dax
->nr_range
- 1), GFP_KERNEL
);
411 pgmap
->nr_range
= dev_dax
->nr_range
;
414 for (i
= 0; i
< dev_dax
->nr_range
; i
++) {
415 struct range
*range
= &dev_dax
->ranges
[i
].range
;
417 if (!devm_request_mem_region(dev
, range
->start
,
418 range_len(range
), dev_name(dev
))) {
419 dev_warn(dev
, "mapping%d: %#llx-%#llx could not reserve range\n",
420 i
, range
->start
, range
->end
);
423 /* don't update the range for static pgmap */
425 pgmap
->ranges
[i
] = *range
;
428 pgmap
->type
= MEMORY_DEVICE_GENERIC
;
429 addr
= devm_memremap_pages(dev
, pgmap
);
431 return PTR_ERR(addr
);
433 inode
= dax_inode(dax_dev
);
434 cdev
= inode
->i_cdev
;
435 cdev_init(cdev
, &dax_fops
);
437 /* for the CONFIG_DEV_DAX_PMEM_COMPAT case */
438 cdev
->owner
= dev
->parent
->driver
->owner
;
440 cdev
->owner
= dev
->driver
->owner
;
441 cdev_set_parent(cdev
, &dev
->kobj
);
442 rc
= cdev_add(cdev
, dev
->devt
, 1);
446 rc
= devm_add_action_or_reset(dev
, dev_dax_cdev_del
, cdev
);
451 return devm_add_action_or_reset(dev
, dev_dax_kill
, dev_dax
);
453 EXPORT_SYMBOL_GPL(dev_dax_probe
);
455 static int dev_dax_remove(struct dev_dax
*dev_dax
)
457 /* all probe actions are unwound by devm */
461 static struct dax_device_driver device_dax_driver
= {
462 .probe
= dev_dax_probe
,
463 .remove
= dev_dax_remove
,
467 static int __init
dax_init(void)
469 return dax_driver_register(&device_dax_driver
);
472 static void __exit
dax_exit(void)
474 dax_driver_unregister(&device_dax_driver
);
477 MODULE_AUTHOR("Intel Corporation");
478 MODULE_LICENSE("GPL v2");
479 module_init(dax_init
);
480 module_exit(dax_exit
);
481 MODULE_ALIAS_DAX_DEVICE(0);