| Message ID | 20200817120431.32233-1-ardb@kernel.org (mailing list archive) |
|---|---|
| State | Superseded, archived |
| Headers | show |
| Series | ACPI: ioremap: avoid redundant rounding to OS page size | expand |
On Mon, Aug 17, 2020 at 02:04:31PM +0200, Ard Biesheuvel wrote: > The arm64 implementation of acpi_os_ioremap() was recently updated to > tighten the checks around which parts of memory are permitted to be > mapped by ACPI code, which generally only needs access to memory regions > that are statically described by firmware, and any attempts to access > memory that is in active use by the OS is generally a bug or a hacking > attempt. This tightening is based on the EFI memory map, which describes > all memory in the system. > > The AArch64 architecture permits page sizes of 16k and 64k in addition > to the EFI default, which is 4k, which means that the EFI memory map may > describe regions that cannot be mapped seamlessly if the OS page size is > greater than 4k. This is usually not a problem, given that the EFI spec > does not permit memory regions requiring different memory attributes to > share a 64k page frame, and so the usual rounding to page size performed > by ioremap() is sufficient to deal with this. However, this rounding does > complicate our EFI memory map permission check, due to the loss of > information that occurs when several small regions share a single 64k > page frame (where rounding each of them will result in the same 64k > single page region). > > However, due to the fact that the region check occurs *before* the call > to ioremap() where the necessary rounding is performed, we can deal > with this issue simply by removing the redundant rounding performed by > acpi_os_map_iomem(), as it appears to be the only place where the > arguments to a call to acpi_os_ioremap() are rounded up. So omit the > rounding in the call, and instead, apply the necessary offset to the > result of kmap(). > > Fixes: 1583052d111f ("arm64/acpi: disallow AML memory opregions to access kernel memory") > Signed-off-by: Ard Biesheuvel <ardb@kernel.org> To me the whole acpi_map() / acpi_os_map_iomem() concept looks bogus, especially as it mixes up iomem and RAM pages in a really bad way, and then throws in staic fixmap-like mappings as well. Also looking at the callers I see no point in keeping a list of the memory mappings. Does anyone have an idea where this craziness comes from? Which of the callers actually has to deal both with iomem and RAM mappings at the same time? It seems like we should be able to untangle the few callers and remove this mess entirely.
On Mon, 17 Aug 2020 at 14:40, Christoph Hellwig <hch@infradead.org> wrote: > > On Mon, Aug 17, 2020 at 02:04:31PM +0200, Ard Biesheuvel wrote: > > The arm64 implementation of acpi_os_ioremap() was recently updated to > > tighten the checks around which parts of memory are permitted to be > > mapped by ACPI code, which generally only needs access to memory regions > > that are statically described by firmware, and any attempts to access > > memory that is in active use by the OS is generally a bug or a hacking > > attempt. This tightening is based on the EFI memory map, which describes > > all memory in the system. > > > > The AArch64 architecture permits page sizes of 16k and 64k in addition > > to the EFI default, which is 4k, which means that the EFI memory map may > > describe regions that cannot be mapped seamlessly if the OS page size is > > greater than 4k. This is usually not a problem, given that the EFI spec > > does not permit memory regions requiring different memory attributes to > > share a 64k page frame, and so the usual rounding to page size performed > > by ioremap() is sufficient to deal with this. However, this rounding does > > complicate our EFI memory map permission check, due to the loss of > > information that occurs when several small regions share a single 64k > > page frame (where rounding each of them will result in the same 64k > > single page region). > > > > However, due to the fact that the region check occurs *before* the call > > to ioremap() where the necessary rounding is performed, we can deal > > with this issue simply by removing the redundant rounding performed by > > acpi_os_map_iomem(), as it appears to be the only place where the > > arguments to a call to acpi_os_ioremap() are rounded up. So omit the > > rounding in the call, and instead, apply the necessary offset to the > > result of kmap(). > > > > Fixes: 1583052d111f ("arm64/acpi: disallow AML memory opregions to access kernel memory") > > Signed-off-by: Ard Biesheuvel <ardb@kernel.org> > > To me the whole acpi_map() / acpi_os_map_iomem() concept looks bogus, > especially as it mixes up iomem and RAM pages in a really bad way, > and then throws in staic fixmap-like mappings as well. > > Also looking at the callers I see no point in keeping a list of the > memory mappings. Does anyone have an idea where this craziness comes > from? Which of the callers actually has to deal both with iomem and > RAM mappings at the same time? > > It seems like we should be able to untangle the few callers and remove > this mess entirely. That would be my preference as well. However, this code is used by x86, ia64 and arm64, and I'd like to get this piece in as a fix, given that the referenced patch was merged in v5.9-rc1, and causes issues on 64k kernels running on older arm64 systems with not-quite-compliant EFI firmware.
Hi Ard,
I love your patch! Perhaps something to improve:
[auto build test WARNING on pm/linux-next]
[also build test WARNING on arm64/for-next/core arm/for-next soc/for-next v5.9-rc1 next-20200817]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]
url: https://github.com/0day-ci/linux/commits/Ard-Biesheuvel/ACPI-ioremap-avoid-redundant-rounding-to-OS-page-size/20200817-200603
base: https://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm.git linux-next
config: x86_64-randconfig-a013-20200817 (attached as .config)
compiler: clang version 12.0.0 (https://github.com/llvm/llvm-project de71b46a519db014ce906a39f8a0e1b235ef1568)
reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
chmod +x ~/bin/make.cross
# install x86_64 cross compiling tool for clang build
# apt-get install binutils-x86-64-linux-gnu
# save the attached .config to linux build tree
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross ARCH=x86_64
If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@intel.com>
All warnings (new ones prefixed by >>):
>> drivers/acpi/osl.c:341:6: warning: variable 'virt' is used uninitialized whenever 'if' condition is true [-Wsometimes-uninitialized]
if (map) {
^~~
drivers/acpi/osl.c:371:9: note: uninitialized use occurs here
return virt;
^~~~
drivers/acpi/osl.c:341:2: note: remove the 'if' if its condition is always false
if (map) {
^~~~~~~~~~
drivers/acpi/osl.c:326:20: note: initialize the variable 'virt' to silence this warning
void __iomem *virt;
^
= NULL
1 warning generated.
# https://github.com/0day-ci/linux/commit/a34cc34917319aed90ebf9b0fbf4146666f5f75d
git remote add linux-review https://github.com/0day-ci/linux
git fetch --no-tags linux-review Ard-Biesheuvel/ACPI-ioremap-avoid-redundant-rounding-to-OS-page-size/20200817-200603
git checkout a34cc34917319aed90ebf9b0fbf4146666f5f75d
vim +341 drivers/acpi/osl.c
ba242d5b1a84bc Myron Stowe 2012-01-20 308
9d128ed17c672b Rafael J. Wysocki 2016-01-02 309 /**
9d128ed17c672b Rafael J. Wysocki 2016-01-02 310 * acpi_os_map_iomem - Get a virtual address for a given physical address range.
9d128ed17c672b Rafael J. Wysocki 2016-01-02 311 * @phys: Start of the physical address range to map.
9d128ed17c672b Rafael J. Wysocki 2016-01-02 312 * @size: Size of the physical address range to map.
9d128ed17c672b Rafael J. Wysocki 2016-01-02 313 *
9d128ed17c672b Rafael J. Wysocki 2016-01-02 314 * Look up the given physical address range in the list of existing ACPI memory
9d128ed17c672b Rafael J. Wysocki 2016-01-02 315 * mappings. If found, get a reference to it and return a pointer to it (its
9d128ed17c672b Rafael J. Wysocki 2016-01-02 316 * virtual address). If not found, map it, add it to that list and return a
9d128ed17c672b Rafael J. Wysocki 2016-01-02 317 * pointer to it.
9d128ed17c672b Rafael J. Wysocki 2016-01-02 318 *
8d3523fb3b7274 Lv Zheng 2016-12-14 319 * During early init (when acpi_permanent_mmap has not been set yet) this
9d128ed17c672b Rafael J. Wysocki 2016-01-02 320 * routine simply calls __acpi_map_table() to get the job done.
9d128ed17c672b Rafael J. Wysocki 2016-01-02 321 */
9fe51603d95341 Qian Cai 2019-06-03 322 void __iomem __ref
9fe51603d95341 Qian Cai 2019-06-03 323 *acpi_os_map_iomem(acpi_physical_address phys, acpi_size size)
^1da177e4c3f41 Linus Torvalds 2005-04-16 324 {
7ffd0443f25024 Rafael J. Wysocki 2011-02-08 325 struct acpi_ioremap *map;
620242ae8c3d9c Myron Stowe 2010-10-21 326 void __iomem *virt;
2d6d9fd3a54a28 Rafael J. Wysocki 2011-01-19 327 acpi_physical_address pg_off;
2d6d9fd3a54a28 Rafael J. Wysocki 2011-01-19 328 acpi_size pg_sz;
620242ae8c3d9c Myron Stowe 2010-10-21 329
^1da177e4c3f41 Linus Torvalds 2005-04-16 330 if (phys > ULONG_MAX) {
^1da177e4c3f41 Linus Torvalds 2005-04-16 331 printk(KERN_ERR PREFIX "Cannot map memory that high\n");
70c0846e430881 Randy Dunlap 2007-02-13 332 return NULL;
^1da177e4c3f41 Linus Torvalds 2005-04-16 333 }
620242ae8c3d9c Myron Stowe 2010-10-21 334
8d3523fb3b7274 Lv Zheng 2016-12-14 335 if (!acpi_permanent_mmap)
ad71860a17ba33 Alexey Starikovskiy 2007-02-02 336 return __acpi_map_table((unsigned long)phys, size);
620242ae8c3d9c Myron Stowe 2010-10-21 337
7ffd0443f25024 Rafael J. Wysocki 2011-02-08 338 mutex_lock(&acpi_ioremap_lock);
7ffd0443f25024 Rafael J. Wysocki 2011-02-08 339 /* Check if there's a suitable mapping already. */
7ffd0443f25024 Rafael J. Wysocki 2011-02-08 340 map = acpi_map_lookup(phys, size);
7ffd0443f25024 Rafael J. Wysocki 2011-02-08 @341 if (map) {
1757659d022b73 Rafael J. Wysocki 2020-07-02 342 map->track.refcount++;
7ffd0443f25024 Rafael J. Wysocki 2011-02-08 343 goto out;
7ffd0443f25024 Rafael J. Wysocki 2011-02-08 344 }
7ffd0443f25024 Rafael J. Wysocki 2011-02-08 345
620242ae8c3d9c Myron Stowe 2010-10-21 346 map = kzalloc(sizeof(*map), GFP_KERNEL);
7ffd0443f25024 Rafael J. Wysocki 2011-02-08 347 if (!map) {
7ffd0443f25024 Rafael J. Wysocki 2011-02-08 348 mutex_unlock(&acpi_ioremap_lock);
620242ae8c3d9c Myron Stowe 2010-10-21 349 return NULL;
7ffd0443f25024 Rafael J. Wysocki 2011-02-08 350 }
620242ae8c3d9c Myron Stowe 2010-10-21 351
4a3cba5e72a523 Myron Stowe 2010-10-21 352 pg_off = round_down(phys, PAGE_SIZE);
4a3cba5e72a523 Myron Stowe 2010-10-21 353 pg_sz = round_up(phys + size, PAGE_SIZE) - pg_off;
a34cc34917319a Ard Biesheuvel 2020-08-17 354 virt = acpi_map(phys, size);
620242ae8c3d9c Myron Stowe 2010-10-21 355 if (!virt) {
7ffd0443f25024 Rafael J. Wysocki 2011-02-08 356 mutex_unlock(&acpi_ioremap_lock);
620242ae8c3d9c Myron Stowe 2010-10-21 357 kfree(map);
620242ae8c3d9c Myron Stowe 2010-10-21 358 return NULL;
620242ae8c3d9c Myron Stowe 2010-10-21 359 }
620242ae8c3d9c Myron Stowe 2010-10-21 360
620242ae8c3d9c Myron Stowe 2010-10-21 361 INIT_LIST_HEAD(&map->list);
a34cc34917319a Ard Biesheuvel 2020-08-17 362 map->virt = (void *)((unsigned long)virt & PAGE_MASK);
4a3cba5e72a523 Myron Stowe 2010-10-21 363 map->phys = pg_off;
4a3cba5e72a523 Myron Stowe 2010-10-21 364 map->size = pg_sz;
1757659d022b73 Rafael J. Wysocki 2020-07-02 365 map->track.refcount = 1;
620242ae8c3d9c Myron Stowe 2010-10-21 366
78cdb3ed405379 Myron Stowe 2010-10-21 367 list_add_tail_rcu(&map->list, &acpi_ioremaps);
620242ae8c3d9c Myron Stowe 2010-10-21 368
7ffd0443f25024 Rafael J. Wysocki 2011-02-08 369 out:
7ffd0443f25024 Rafael J. Wysocki 2011-02-08 370 mutex_unlock(&acpi_ioremap_lock);
a34cc34917319a Ard Biesheuvel 2020-08-17 371 return virt;
^1da177e4c3f41 Linus Torvalds 2005-04-16 372 }
a238317ce81855 Lv Zheng 2014-05-20 373 EXPORT_SYMBOL_GPL(acpi_os_map_iomem);
a238317ce81855 Lv Zheng 2014-05-20 374
---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org
Hi Ard,
I love your patch! Perhaps something to improve:
[auto build test WARNING on pm/linux-next]
[also build test WARNING on arm64/for-next/core arm/for-next soc/for-next v5.9-rc1 next-20200817]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]
url: https://github.com/0day-ci/linux/commits/Ard-Biesheuvel/ACPI-ioremap-avoid-redundant-rounding-to-OS-page-size/20200817-200603
base: https://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm.git linux-next
config: x86_64-randconfig-s022-20200817 (attached as .config)
compiler: gcc-9 (Debian 9.3.0-15) 9.3.0
reproduce:
# apt-get install sparse
# sparse version: v0.6.2-180-g49f7e13a-dirty
# save the attached .config to linux build tree
make W=1 C=1 CF='-fdiagnostic-prefix -D__CHECK_ENDIAN__' ARCH=x86_64
If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@intel.com>
sparse warnings: (new ones prefixed by >>)
>> drivers/acpi/osl.c:362:19: sparse: sparse: incorrect type in assignment (different address spaces) @@ expected void [noderef] __iomem *virt @@ got void * @@
>> drivers/acpi/osl.c:362:19: sparse: expected void [noderef] __iomem *virt
>> drivers/acpi/osl.c:362:19: sparse: got void *
drivers/acpi/osl.c:377:17: sparse: sparse: cast removes address space '__iomem' of expression
drivers/acpi/osl.c:714:1: sparse: sparse: context imbalance in 'acpi_os_read_memory' - wrong count at exit
drivers/acpi/osl.c:747:1: sparse: sparse: context imbalance in 'acpi_os_write_memory' - wrong count at exit
# https://github.com/0day-ci/linux/commit/a34cc34917319aed90ebf9b0fbf4146666f5f75d
git remote add linux-review https://github.com/0day-ci/linux
git fetch --no-tags linux-review Ard-Biesheuvel/ACPI-ioremap-avoid-redundant-rounding-to-OS-page-size/20200817-200603
git checkout a34cc34917319aed90ebf9b0fbf4146666f5f75d
vim +362 drivers/acpi/osl.c
308
309 /**
310 * acpi_os_map_iomem - Get a virtual address for a given physical address range.
311 * @phys: Start of the physical address range to map.
312 * @size: Size of the physical address range to map.
313 *
314 * Look up the given physical address range in the list of existing ACPI memory
315 * mappings. If found, get a reference to it and return a pointer to it (its
316 * virtual address). If not found, map it, add it to that list and return a
317 * pointer to it.
318 *
319 * During early init (when acpi_permanent_mmap has not been set yet) this
320 * routine simply calls __acpi_map_table() to get the job done.
321 */
322 void __iomem __ref
323 *acpi_os_map_iomem(acpi_physical_address phys, acpi_size size)
324 {
325 struct acpi_ioremap *map;
326 void __iomem *virt;
327 acpi_physical_address pg_off;
328 acpi_size pg_sz;
329
330 if (phys > ULONG_MAX) {
331 printk(KERN_ERR PREFIX "Cannot map memory that high\n");
332 return NULL;
333 }
334
335 if (!acpi_permanent_mmap)
336 return __acpi_map_table((unsigned long)phys, size);
337
338 mutex_lock(&acpi_ioremap_lock);
339 /* Check if there's a suitable mapping already. */
340 map = acpi_map_lookup(phys, size);
341 if (map) {
342 map->track.refcount++;
343 goto out;
344 }
345
346 map = kzalloc(sizeof(*map), GFP_KERNEL);
347 if (!map) {
348 mutex_unlock(&acpi_ioremap_lock);
349 return NULL;
350 }
351
352 pg_off = round_down(phys, PAGE_SIZE);
353 pg_sz = round_up(phys + size, PAGE_SIZE) - pg_off;
354 virt = acpi_map(phys, size);
355 if (!virt) {
356 mutex_unlock(&acpi_ioremap_lock);
357 kfree(map);
358 return NULL;
359 }
360
361 INIT_LIST_HEAD(&map->list);
> 362 map->virt = (void *)((unsigned long)virt & PAGE_MASK);
363 map->phys = pg_off;
364 map->size = pg_sz;
365 map->track.refcount = 1;
366
367 list_add_tail_rcu(&map->list, &acpi_ioremaps);
368
369 out:
370 mutex_unlock(&acpi_ioremap_lock);
371 return virt;
372 }
373 EXPORT_SYMBOL_GPL(acpi_os_map_iomem);
374
---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org
Hi Ard,
I love your patch! Perhaps something to improve:
[auto build test WARNING on pm/linux-next]
[also build test WARNING on arm64/for-next/core arm/for-next soc/for-next v5.9-rc1 next-20200817]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]
url: https://github.com/0day-ci/linux/commits/Ard-Biesheuvel/ACPI-ioremap-avoid-redundant-rounding-to-OS-page-size/20200817-200603
base: https://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm.git linux-next
config: i386-randconfig-m021-20200817 (attached as .config)
compiler: gcc-9 (Debian 9.3.0-15) 9.3.0
If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@intel.com>
smatch warnings:
drivers/acpi/osl.c:371 acpi_os_map_iomem() error: uninitialized symbol 'virt'.
# https://github.com/0day-ci/linux/commit/a34cc34917319aed90ebf9b0fbf4146666f5f75d
git remote add linux-review https://github.com/0day-ci/linux
git fetch --no-tags linux-review Ard-Biesheuvel/ACPI-ioremap-avoid-redundant-rounding-to-OS-page-size/20200817-200603
git checkout a34cc34917319aed90ebf9b0fbf4146666f5f75d
vim +/virt +371 drivers/acpi/osl.c
308
309 /**
310 * acpi_os_map_iomem - Get a virtual address for a given physical address range.
311 * @phys: Start of the physical address range to map.
312 * @size: Size of the physical address range to map.
313 *
314 * Look up the given physical address range in the list of existing ACPI memory
315 * mappings. If found, get a reference to it and return a pointer to it (its
316 * virtual address). If not found, map it, add it to that list and return a
317 * pointer to it.
318 *
319 * During early init (when acpi_permanent_mmap has not been set yet) this
320 * routine simply calls __acpi_map_table() to get the job done.
321 */
322 void __iomem __ref
323 *acpi_os_map_iomem(acpi_physical_address phys, acpi_size size)
324 {
325 struct acpi_ioremap *map;
326 void __iomem *virt;
327 acpi_physical_address pg_off;
328 acpi_size pg_sz;
329
330 if (phys > ULONG_MAX) {
331 printk(KERN_ERR PREFIX "Cannot map memory that high\n");
332 return NULL;
333 }
334
335 if (!acpi_permanent_mmap)
336 return __acpi_map_table((unsigned long)phys, size);
337
338 mutex_lock(&acpi_ioremap_lock);
339 /* Check if there's a suitable mapping already. */
340 map = acpi_map_lookup(phys, size);
341 if (map) {
342 map->track.refcount++;
343 goto out;
344 }
345
346 map = kzalloc(sizeof(*map), GFP_KERNEL);
347 if (!map) {
348 mutex_unlock(&acpi_ioremap_lock);
349 return NULL;
350 }
351
352 pg_off = round_down(phys, PAGE_SIZE);
353 pg_sz = round_up(phys + size, PAGE_SIZE) - pg_off;
354 virt = acpi_map(phys, size);
355 if (!virt) {
356 mutex_unlock(&acpi_ioremap_lock);
357 kfree(map);
358 return NULL;
359 }
360
361 INIT_LIST_HEAD(&map->list);
362 map->virt = (void *)((unsigned long)virt & PAGE_MASK);
363 map->phys = pg_off;
364 map->size = pg_sz;
365 map->track.refcount = 1;
366
367 list_add_tail_rcu(&map->list, &acpi_ioremaps);
368
369 out:
370 mutex_unlock(&acpi_ioremap_lock);
> 371 return virt;
372 }
373 EXPORT_SYMBOL_GPL(acpi_os_map_iomem);
374
---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org
diff --git a/drivers/acpi/osl.c b/drivers/acpi/osl.c index 6ad8cb05f672..55dbdbbae3be 100644 --- a/drivers/acpi/osl.c +++ b/drivers/acpi/osl.c @@ -289,7 +289,8 @@ static void __iomem *acpi_map(acpi_physical_address pg_off, unsigned long pg_sz) if (should_use_kmap(pfn)) { if (pg_sz > PAGE_SIZE) return NULL; - return (void __iomem __force *)kmap(pfn_to_page(pfn)); + pg_off &= ~PAGE_MASK; + return (void __iomem __force *)(kmap(pfn_to_page(pfn)) + pg_off); } else return acpi_os_ioremap(pg_off, pg_sz); } @@ -350,7 +351,7 @@ void __iomem __ref pg_off = round_down(phys, PAGE_SIZE); pg_sz = round_up(phys + size, PAGE_SIZE) - pg_off; - virt = acpi_map(pg_off, pg_sz); + virt = acpi_map(phys, size); if (!virt) { mutex_unlock(&acpi_ioremap_lock); kfree(map); @@ -358,7 +359,7 @@ void __iomem __ref } INIT_LIST_HEAD(&map->list); - map->virt = virt; + map->virt = (void *)((unsigned long)virt & PAGE_MASK); map->phys = pg_off; map->size = pg_sz; map->track.refcount = 1; @@ -367,7 +368,7 @@ void __iomem __ref out: mutex_unlock(&acpi_ioremap_lock); - return map->virt + (phys - map->phys); + return virt; } EXPORT_SYMBOL_GPL(acpi_os_map_iomem);
The arm64 implementation of acpi_os_ioremap() was recently updated to tighten the checks around which parts of memory are permitted to be mapped by ACPI code, which generally only needs access to memory regions that are statically described by firmware, and any attempts to access memory that is in active use by the OS is generally a bug or a hacking attempt. This tightening is based on the EFI memory map, which describes all memory in the system. The AArch64 architecture permits page sizes of 16k and 64k in addition to the EFI default, which is 4k, which means that the EFI memory map may describe regions that cannot be mapped seamlessly if the OS page size is greater than 4k. This is usually not a problem, given that the EFI spec does not permit memory regions requiring different memory attributes to share a 64k page frame, and so the usual rounding to page size performed by ioremap() is sufficient to deal with this. However, this rounding does complicate our EFI memory map permission check, due to the loss of information that occurs when several small regions share a single 64k page frame (where rounding each of them will result in the same 64k single page region). However, due to the fact that the region check occurs *before* the call to ioremap() where the necessary rounding is performed, we can deal with this issue simply by removing the redundant rounding performed by acpi_os_map_iomem(), as it appears to be the only place where the arguments to a call to acpi_os_ioremap() are rounded up. So omit the rounding in the call, and instead, apply the necessary offset to the result of kmap(). Fixes: 1583052d111f ("arm64/acpi: disallow AML memory opregions to access kernel memory") Signed-off-by: Ard Biesheuvel <ardb@kernel.org> --- drivers/acpi/osl.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-)