| Message ID | 1608218912-28932-2-git-send-email-anshuman.khandual@arm.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | mm/hotplug: Pre-validate the address range with platform | expand |
On 17.12.20 16:28, Anshuman Khandual wrote: > This introduces memhp_range_allowed() which can be called in various memory > hotplug paths to prevalidate the address range which is being added, with > the platform. Then memhp_range_allowed() calls memhp_get_pluggable_range() > which provides applicable address range depending on whether linear mapping > is required or not. For ranges that require linear mapping, it calls a new > arch callback arch_get_mappable_range() which the platform can override. So > the new callback, in turn provides the platform an opportunity to configure > acceptable memory hotplug address ranges in case there are constraints. > > This mechanism will help prevent platform specific errors deep down during > hotplug calls. This drops now redundant check_hotplug_memory_addressable() > check in __add_pages() but instead adds a VM_BUG_ON() check which would > ensure that the range has been validated with memhp_range_allowed() earlier > in the call chain. Besides memhp_get_pluggable_range() also can be used by > potential memory hotplug callers to avail the allowed physical range which > would go through on a given platform. > > This does not really add any new range check in generic memory hotplug but > instead compensates for lost checks in arch_add_memory() where applicable > and check_hotplug_memory_addressable(), with unified memhp_range_allowed(). > Subject s/mm\/hotplug/mm\/memory_hotplug/ Everywhere in this patch: Use "true/false" for boolean values. > Cc: David Hildenbrand <david@redhat.com> > Cc: Andrew Morton <akpm@linux-foundation.org> > Cc: linux-mm@kvack.org > Cc: linux-kernel@vger.kernel.org > Suggested-by: David Hildenbrand <david@redhat.com> > Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com> > --- > include/linux/memory_hotplug.h | 10 +++++ > mm/memory_hotplug.c | 79 +++++++++++++++++++++++++--------- > mm/memremap.c | 6 +++ > 3 files changed, 75 insertions(+), 20 deletions(-) > > diff --git a/include/linux/memory_hotplug.h b/include/linux/memory_hotplug.h > index 551093b74596..8d72354758c8 100644 > --- a/include/linux/memory_hotplug.h > +++ b/include/linux/memory_hotplug.h > @@ -70,6 +70,9 @@ typedef int __bitwise mhp_t; > */ > #define MEMHP_MERGE_RESOURCE ((__force mhp_t)BIT(0)) > > +bool memhp_range_allowed(u64 start, u64 size, bool need_mapping); > +struct range memhp_get_pluggable_range(bool need_mapping); AFAIKs, all memhp_get_pluggable_range() users pass "1". What about the "add_pages()-only" path?
On Mon, Jan 11, 2021 at 11:51:47AM +0100, David Hildenbrand wrote: > AFAIKs, all memhp_get_pluggable_range() users pass "1". > > What about the "add_pages()-only" path? I guess you refer to memremap_pages(), right? If so, moving the added memhp_range_allowed() check above the if-else might do the trick
On 1/11/21 4:21 PM, David Hildenbrand wrote: > On 17.12.20 16:28, Anshuman Khandual wrote: >> This introduces memhp_range_allowed() which can be called in various memory >> hotplug paths to prevalidate the address range which is being added, with >> the platform. Then memhp_range_allowed() calls memhp_get_pluggable_range() >> which provides applicable address range depending on whether linear mapping >> is required or not. For ranges that require linear mapping, it calls a new >> arch callback arch_get_mappable_range() which the platform can override. So >> the new callback, in turn provides the platform an opportunity to configure >> acceptable memory hotplug address ranges in case there are constraints. >> >> This mechanism will help prevent platform specific errors deep down during >> hotplug calls. This drops now redundant check_hotplug_memory_addressable() >> check in __add_pages() but instead adds a VM_BUG_ON() check which would >> ensure that the range has been validated with memhp_range_allowed() earlier >> in the call chain. Besides memhp_get_pluggable_range() also can be used by >> potential memory hotplug callers to avail the allowed physical range which >> would go through on a given platform. >> >> This does not really add any new range check in generic memory hotplug but >> instead compensates for lost checks in arch_add_memory() where applicable >> and check_hotplug_memory_addressable(), with unified memhp_range_allowed(). >> > > Subject s/mm\/hotplug/mm\/memory_hotplug/ Sure, will do. > > Everywhere in this patch: Use "true/false" for boolean values. Sure, will change. > >> Cc: David Hildenbrand <david@redhat.com> >> Cc: Andrew Morton <akpm@linux-foundation.org> >> Cc: linux-mm@kvack.org >> Cc: linux-kernel@vger.kernel.org >> Suggested-by: David Hildenbrand <david@redhat.com> >> Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com> >> --- >> include/linux/memory_hotplug.h | 10 +++++ >> mm/memory_hotplug.c | 79 +++++++++++++++++++++++++--------- >> mm/memremap.c | 6 +++ >> 3 files changed, 75 insertions(+), 20 deletions(-) >> >> diff --git a/include/linux/memory_hotplug.h b/include/linux/memory_hotplug.h >> index 551093b74596..8d72354758c8 100644 >> --- a/include/linux/memory_hotplug.h >> +++ b/include/linux/memory_hotplug.h >> @@ -70,6 +70,9 @@ typedef int __bitwise mhp_t; >> */ >> #define MEMHP_MERGE_RESOURCE ((__force mhp_t)BIT(0)) >> >> +bool memhp_range_allowed(u64 start, u64 size, bool need_mapping); >> +struct range memhp_get_pluggable_range(bool need_mapping); > > AFAIKs, all memhp_get_pluggable_range() users pass "1". Right. > > What about the "add_pages()-only" path? I had dropped memhp_range_allowed() changes for add_pages() in pagemap_range() because you had mentioned not to add any new checks in the generic code. Will add it back if that is preferred.
On 1/11/21 7:13 PM, Oscar Salvador wrote: > On Mon, Jan 11, 2021 at 11:51:47AM +0100, David Hildenbrand wrote: >> AFAIKs, all memhp_get_pluggable_range() users pass "1". >> >> What about the "add_pages()-only" path? > > I guess you refer to memremap_pages(), right? Right, via pagemap_range(). > If so, moving the added memhp_range_allowed() check above the if-else might do > the trick > We had that code in the earlier version. But dropped it, as we did not want to add any new checks in the generic code. Can add it back if that is preferred.
On 12.01.21 04:51, Anshuman Khandual wrote: > > > On 1/11/21 7:13 PM, Oscar Salvador wrote: >> On Mon, Jan 11, 2021 at 11:51:47AM +0100, David Hildenbrand wrote: >>> AFAIKs, all memhp_get_pluggable_range() users pass "1". >>> >>> What about the "add_pages()-only" path? >> >> I guess you refer to memremap_pages(), right? > > Right, via pagemap_range(). > >> If so, moving the added memhp_range_allowed() check above the if-else might do >> the trick >> > We had that code in the earlier version. But dropped it, as we did > not want to add any new checks in the generic code. Can add it back > if that is preferred. I remember discussing replacing the check in __add_pages() instead. But I don't really care where the check ends up being. As discussed, at some point, we should provide versions of add_pages() and arch_add_pages() that don't immediately end in arch-code.
On 1/12/21 3:39 PM, David Hildenbrand wrote: > On 12.01.21 04:51, Anshuman Khandual wrote: >> >> >> On 1/11/21 7:13 PM, Oscar Salvador wrote: >>> On Mon, Jan 11, 2021 at 11:51:47AM +0100, David Hildenbrand wrote: >>>> AFAIKs, all memhp_get_pluggable_range() users pass "1". >>>> >>>> What about the "add_pages()-only" path? >>> >>> I guess you refer to memremap_pages(), right? >> >> Right, via pagemap_range(). >> >>> If so, moving the added memhp_range_allowed() check above the if-else might do >>> the trick >>> >> We had that code in the earlier version. But dropped it, as we did >> not want to add any new checks in the generic code. Can add it back >> if that is preferred. > > I remember discussing replacing the check in __add_pages() instead. But The proposed change for __add_pages() now seems misleading. Instead of VM_BUG_ON(), memhp_range_allowed() should be checked directly for a non linear mapping i.e with 'false' argument and return prematurely in case that check fails. s/VM_BUG_ON(!memhp_range_allowed(.., 1)/!memhp_range_allowed(.., 0)/ /* * Reasonably generic function for adding memory. It is * expected that archs that support memory hotplug will @@ -317,10 +304,7 @@ int __ref __add_pages(int nid, unsigned long pfn, unsigned long nr_pages, if (WARN_ON_ONCE(!params->pgprot.pgprot)) return -EINVAL; - err = check_hotplug_memory_addressable(pfn, nr_pages); - if (err) - return err; - + VM_BUG_ON(!memhp_range_allowed(PFN_PHYS(pfn), nr_pages * PAGE_SIZE, 1)); if (altmap) { /* * Validate altmap is within bounds of the total request @@ -1181,6 +1165,61 @@ int add_memory_driver_managed(int nid, u64 start, u64 size, > I don't really care where the check ends up being. As discussed, at some > point, we should provide versions of add_pages() and arch_add_pages() > that don't immediately end in arch-code. Sure. But for now, AFAICS the above replacement should be sufficient.
diff --git a/include/linux/memory_hotplug.h b/include/linux/memory_hotplug.h index 551093b74596..8d72354758c8 100644 --- a/include/linux/memory_hotplug.h +++ b/include/linux/memory_hotplug.h @@ -70,6 +70,9 @@ typedef int __bitwise mhp_t; */ #define MEMHP_MERGE_RESOURCE ((__force mhp_t)BIT(0)) +bool memhp_range_allowed(u64 start, u64 size, bool need_mapping); +struct range memhp_get_pluggable_range(bool need_mapping); + /* * Extended parameters for memory hotplug: * altmap: alternative allocator for memmap array (optional) @@ -281,6 +284,13 @@ static inline bool movable_node_is_enabled(void) } #endif /* ! CONFIG_MEMORY_HOTPLUG */ +/* + * Keep this declaration outside CONFIG_MEMORY_HOTPLUG as some + * platforms might override and use arch_get_mappable_range() + * for internal non memory hotplug purposes. + */ +struct range arch_get_mappable_range(void); + #if defined(CONFIG_MEMORY_HOTPLUG) || defined(CONFIG_DEFERRED_STRUCT_PAGE_INIT) /* * pgdat resizing functions diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c index 63b2e46b6555..27ee352cbe31 100644 --- a/mm/memory_hotplug.c +++ b/mm/memory_hotplug.c @@ -107,6 +107,9 @@ static struct resource *register_memory_resource(u64 start, u64 size, if (strcmp(resource_name, "System RAM")) flags |= IORESOURCE_SYSRAM_DRIVER_MANAGED; + if (!memhp_range_allowed(start, size, 1)) + return ERR_PTR(-E2BIG); + /* * Make sure value parsed from 'mem=' only restricts memory adding * while booting, so that memory hotplug won't be impacted. Please @@ -284,22 +287,6 @@ static int check_pfn_span(unsigned long pfn, unsigned long nr_pages, return 0; } -static int check_hotplug_memory_addressable(unsigned long pfn, - unsigned long nr_pages) -{ - const u64 max_addr = PFN_PHYS(pfn + nr_pages) - 1; - - if (max_addr >> MAX_PHYSMEM_BITS) { - const u64 max_allowed = (1ull << (MAX_PHYSMEM_BITS + 1)) - 1; - WARN(1, - "Hotplugged memory exceeds maximum addressable address, range=%#llx-%#llx, maximum=%#llx\n", - (u64)PFN_PHYS(pfn), max_addr, max_allowed); - return -E2BIG; - } - - return 0; -} - /* * Reasonably generic function for adding memory. It is * expected that archs that support memory hotplug will @@ -317,10 +304,7 @@ int __ref __add_pages(int nid, unsigned long pfn, unsigned long nr_pages, if (WARN_ON_ONCE(!params->pgprot.pgprot)) return -EINVAL; - err = check_hotplug_memory_addressable(pfn, nr_pages); - if (err) - return err; - + VM_BUG_ON(!memhp_range_allowed(PFN_PHYS(pfn), nr_pages * PAGE_SIZE, 1)); if (altmap) { /* * Validate altmap is within bounds of the total request @@ -1181,6 +1165,61 @@ int add_memory_driver_managed(int nid, u64 start, u64 size, } EXPORT_SYMBOL_GPL(add_memory_driver_managed); +/* + * Platforms should define arch_get_mappable_range() that provides + * maximum possible addressable physical memory range for which the + * linear mapping could be created. The platform returned address + * range must adhere to these following semantics. + * + * - range.start <= range.end + * - Range includes both end points [range.start..range.end] + * + * There is also a fallback definition provided here, allowing the + * entire possible physical address range in case any platform does + * not define arch_get_mappable_range(). + */ +struct range __weak arch_get_mappable_range(void) +{ + struct range memhp_range = { + .start = 0UL, + .end = -1ULL, + }; + return memhp_range; +} + +struct range memhp_get_pluggable_range(bool need_mapping) +{ + const u64 max_phys = (1ULL << (MAX_PHYSMEM_BITS + 1)) - 1; + struct range memhp_range; + + if (need_mapping) { + memhp_range = arch_get_mappable_range(); + if (memhp_range.start > max_phys) { + memhp_range.start = 0; + memhp_range.end = 0; + } + memhp_range.end = min_t(u64, memhp_range.end, max_phys); + } else { + memhp_range.start = 0; + memhp_range.end = max_phys; + } + return memhp_range; +} +EXPORT_SYMBOL_GPL(memhp_get_pluggable_range); + +bool memhp_range_allowed(u64 start, u64 size, bool need_mapping) +{ + struct range memhp_range = memhp_get_pluggable_range(need_mapping); + u64 end = start + size; + + if (start < end && start >= memhp_range.start && (end - 1) <= memhp_range.end) + return true; + + pr_warn("Hotplug memory [%#llx-%#llx] exceeds maximum addressable range [%#llx-%#llx]\n", + start, end, memhp_range.start, memhp_range.end); + return false; +} + #ifdef CONFIG_MEMORY_HOTREMOVE /* * Confirm all pages in a range [start, end) belong to the same zone (skipping diff --git a/mm/memremap.c b/mm/memremap.c index 16b2fb482da1..8c5500030592 100644 --- a/mm/memremap.c +++ b/mm/memremap.c @@ -253,6 +253,12 @@ static int pagemap_range(struct dev_pagemap *pgmap, struct mhp_params *params, goto err_kasan; } + if (!memhp_range_allowed(range->start, range_len(range), 1)) { + error = -ERANGE; + mem_hotplug_done(); + goto err_add_memory; + } + error = arch_add_memory(nid, range->start, range_len(range), params); }
This introduces memhp_range_allowed() which can be called in various memory hotplug paths to prevalidate the address range which is being added, with the platform. Then memhp_range_allowed() calls memhp_get_pluggable_range() which provides applicable address range depending on whether linear mapping is required or not. For ranges that require linear mapping, it calls a new arch callback arch_get_mappable_range() which the platform can override. So the new callback, in turn provides the platform an opportunity to configure acceptable memory hotplug address ranges in case there are constraints. This mechanism will help prevent platform specific errors deep down during hotplug calls. This drops now redundant check_hotplug_memory_addressable() check in __add_pages() but instead adds a VM_BUG_ON() check which would ensure that the range has been validated with memhp_range_allowed() earlier in the call chain. Besides memhp_get_pluggable_range() also can be used by potential memory hotplug callers to avail the allowed physical range which would go through on a given platform. This does not really add any new range check in generic memory hotplug but instead compensates for lost checks in arch_add_memory() where applicable and check_hotplug_memory_addressable(), with unified memhp_range_allowed(). Cc: David Hildenbrand <david@redhat.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: linux-mm@kvack.org Cc: linux-kernel@vger.kernel.org Suggested-by: David Hildenbrand <david@redhat.com> Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com> --- include/linux/memory_hotplug.h | 10 +++++ mm/memory_hotplug.c | 79 +++++++++++++++++++++++++--------- mm/memremap.c | 6 +++ 3 files changed, 75 insertions(+), 20 deletions(-)