| Message ID | 20210309122601.5543-1-catalin.marinas@arm.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | arm64: mte: Map hotplugged memory as Normal Tagged | expand |
On 09.03.21 13:26, Catalin Marinas wrote: > In a system supporting MTE, the linear map must allow reading/writing > allocation tags by setting the memory type as Normal Tagged. Currently, > this is only handled for memory present at boot. Hotplugged memory uses > Normal non-Tagged memory. > > Introduce pgprot_mhp() for hotplugged memory and use it in > add_memory_resource(). The arm64 code maps pgprot_mhp() to > pgprot_tagged(). > > Note that ZONE_DEVICE memory should not be mapped as Tagged and > therefore setting the memory type in arch_add_memory() is not feasible. > > Signed-off-by: Catalin Marinas <catalin.marinas@arm.com> > Fixes: 0178dc761368 ("arm64: mte: Use Normal Tagged attributes for the linear map") > Reported-by: Patrick Daly <pdaly@codeaurora.org> > Link: https://lore.kernel.org/r/1614745263-27827-1-git-send-email-pdaly@codeaurora.org > Cc: <stable@vger.kernel.org> # 5.10.x > Cc: Will Deacon <will@kernel.org> > Cc: Andrew Morton <akpm@linux-foundation.org> > Cc: Vincenzo Frascino <vincenzo.frascino@arm.com> > Cc: David Hildenbrand <david@redhat.com> > --- > > Patrick, could you please give this patch a try on your platform? Thanks. > > arch/arm64/include/asm/pgtable-prot.h | 1 - > arch/arm64/include/asm/pgtable.h | 3 +++ > arch/arm64/mm/mmu.c | 3 ++- > include/linux/pgtable.h | 4 ++++ > mm/memory_hotplug.c | 2 +- > 5 files changed, 10 insertions(+), 3 deletions(-) > > diff --git a/arch/arm64/include/asm/pgtable-prot.h b/arch/arm64/include/asm/pgtable-prot.h > index 046be789fbb4..9a65fb528110 100644 > --- a/arch/arm64/include/asm/pgtable-prot.h > +++ b/arch/arm64/include/asm/pgtable-prot.h > @@ -66,7 +66,6 @@ extern bool arm64_use_ng_mappings; > #define _PAGE_DEFAULT (_PROT_DEFAULT | PTE_ATTRINDX(MT_NORMAL)) > > #define PAGE_KERNEL __pgprot(PROT_NORMAL) > -#define PAGE_KERNEL_TAGGED __pgprot(PROT_NORMAL_TAGGED) > #define PAGE_KERNEL_RO __pgprot((PROT_NORMAL & ~PTE_WRITE) | PTE_RDONLY) > #define PAGE_KERNEL_ROX __pgprot((PROT_NORMAL & ~(PTE_WRITE | PTE_PXN)) | PTE_RDONLY) > #define PAGE_KERNEL_EXEC __pgprot(PROT_NORMAL & ~PTE_PXN) > diff --git a/arch/arm64/include/asm/pgtable.h b/arch/arm64/include/asm/pgtable.h > index e17b96d0e4b5..47027796c2f9 100644 > --- a/arch/arm64/include/asm/pgtable.h > +++ b/arch/arm64/include/asm/pgtable.h > @@ -486,6 +486,9 @@ static inline pmd_t pmd_mkdevmap(pmd_t pmd) > __pgprot_modify(prot, PTE_ATTRINDX_MASK, PTE_ATTRINDX(MT_NORMAL_NC) | PTE_PXN | PTE_UXN) > #define pgprot_device(prot) \ > __pgprot_modify(prot, PTE_ATTRINDX_MASK, PTE_ATTRINDX(MT_DEVICE_nGnRE) | PTE_PXN | PTE_UXN) > +#define pgprot_tagged(prot) \ > + __pgprot_modify(prot, PTE_ATTRINDX_MASK, PTE_ATTRINDX(MT_NORMAL_TAGGED)) > +#define pgprot_mhp pgprot_tagged > /* > * DMA allocations for non-coherent devices use what the Arm architecture calls > * "Normal non-cacheable" memory, which permits speculation, unaligned accesses > diff --git a/arch/arm64/mm/mmu.c b/arch/arm64/mm/mmu.c > index 3802cfbdd20d..9c8aa1b44cd5 100644 > --- a/arch/arm64/mm/mmu.c > +++ b/arch/arm64/mm/mmu.c > @@ -512,7 +512,8 @@ static void __init map_mem(pgd_t *pgdp) > * if MTE is present. Otherwise, it has the same attributes as > * PAGE_KERNEL. > */ > - __map_memblock(pgdp, start, end, PAGE_KERNEL_TAGGED, flags); > + __map_memblock(pgdp, start, end, pgprot_tagged(PAGE_KERNEL), > + flags); > } > > /* > diff --git a/include/linux/pgtable.h b/include/linux/pgtable.h > index cdfc4e9f253e..5e772392a379 100644 > --- a/include/linux/pgtable.h > +++ b/include/linux/pgtable.h > @@ -904,6 +904,10 @@ static inline void ptep_modify_prot_commit(struct vm_area_struct *vma, > #define pgprot_device pgprot_noncached > #endif > > +#ifndef pgprot_mhp > +#define pgprot_mhp(prot) (prot) > +#endif > + > #ifdef CONFIG_MMU > #ifndef pgprot_modify > #define pgprot_modify pgprot_modify > diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c > index 5ba51a8bdaeb..0cdbbfbc5757 100644 > --- a/mm/memory_hotplug.c > +++ b/mm/memory_hotplug.c > @@ -1072,7 +1072,7 @@ static int online_memory_block(struct memory_block *mem, void *arg) > */ > int __ref add_memory_resource(int nid, struct resource *res, mhp_t mhp_flags) > { > - struct mhp_params params = { .pgprot = PAGE_KERNEL }; > + struct mhp_params params = { .pgprot = pgprot_mhp(PAGE_KERNEL) }; > u64 start, size; > bool new_node = false; > int ret; > Reviewed-by: David Hildenbrand <david@redhat.com>
On 3/9/21 12:26 PM, Catalin Marinas wrote: > In a system supporting MTE, the linear map must allow reading/writing > allocation tags by setting the memory type as Normal Tagged. Currently, > this is only handled for memory present at boot. Hotplugged memory uses > Normal non-Tagged memory. > > Introduce pgprot_mhp() for hotplugged memory and use it in > add_memory_resource(). The arm64 code maps pgprot_mhp() to > pgprot_tagged(). > > Note that ZONE_DEVICE memory should not be mapped as Tagged and > therefore setting the memory type in arch_add_memory() is not feasible. > > Signed-off-by: Catalin Marinas <catalin.marinas@arm.com> > Fixes: 0178dc761368 ("arm64: mte: Use Normal Tagged attributes for the linear map") > Reported-by: Patrick Daly <pdaly@codeaurora.org> > Link: https://lore.kernel.org/r/1614745263-27827-1-git-send-email-pdaly@codeaurora.org > Cc: <stable@vger.kernel.org> # 5.10.x > Cc: Will Deacon <will@kernel.org> > Cc: Andrew Morton <akpm@linux-foundation.org> > Cc: Vincenzo Frascino <vincenzo.frascino@arm.com> > Cc: David Hildenbrand <david@redhat.com> Reviewed-by: Vincenzo Frascino <vincenzo.frascino@arm.com> > --- > > Patrick, could you please give this patch a try on your platform? Thanks. > > arch/arm64/include/asm/pgtable-prot.h | 1 - > arch/arm64/include/asm/pgtable.h | 3 +++ > arch/arm64/mm/mmu.c | 3 ++- > include/linux/pgtable.h | 4 ++++ > mm/memory_hotplug.c | 2 +- > 5 files changed, 10 insertions(+), 3 deletions(-) > > diff --git a/arch/arm64/include/asm/pgtable-prot.h b/arch/arm64/include/asm/pgtable-prot.h > index 046be789fbb4..9a65fb528110 100644 > --- a/arch/arm64/include/asm/pgtable-prot.h > +++ b/arch/arm64/include/asm/pgtable-prot.h > @@ -66,7 +66,6 @@ extern bool arm64_use_ng_mappings; > #define _PAGE_DEFAULT (_PROT_DEFAULT | PTE_ATTRINDX(MT_NORMAL)) > > #define PAGE_KERNEL __pgprot(PROT_NORMAL) > -#define PAGE_KERNEL_TAGGED __pgprot(PROT_NORMAL_TAGGED) > #define PAGE_KERNEL_RO __pgprot((PROT_NORMAL & ~PTE_WRITE) | PTE_RDONLY) > #define PAGE_KERNEL_ROX __pgprot((PROT_NORMAL & ~(PTE_WRITE | PTE_PXN)) | PTE_RDONLY) > #define PAGE_KERNEL_EXEC __pgprot(PROT_NORMAL & ~PTE_PXN) > diff --git a/arch/arm64/include/asm/pgtable.h b/arch/arm64/include/asm/pgtable.h > index e17b96d0e4b5..47027796c2f9 100644 > --- a/arch/arm64/include/asm/pgtable.h > +++ b/arch/arm64/include/asm/pgtable.h > @@ -486,6 +486,9 @@ static inline pmd_t pmd_mkdevmap(pmd_t pmd) > __pgprot_modify(prot, PTE_ATTRINDX_MASK, PTE_ATTRINDX(MT_NORMAL_NC) | PTE_PXN | PTE_UXN) > #define pgprot_device(prot) \ > __pgprot_modify(prot, PTE_ATTRINDX_MASK, PTE_ATTRINDX(MT_DEVICE_nGnRE) | PTE_PXN | PTE_UXN) > +#define pgprot_tagged(prot) \ > + __pgprot_modify(prot, PTE_ATTRINDX_MASK, PTE_ATTRINDX(MT_NORMAL_TAGGED)) > +#define pgprot_mhp pgprot_tagged > /* > * DMA allocations for non-coherent devices use what the Arm architecture calls > * "Normal non-cacheable" memory, which permits speculation, unaligned accesses > diff --git a/arch/arm64/mm/mmu.c b/arch/arm64/mm/mmu.c > index 3802cfbdd20d..9c8aa1b44cd5 100644 > --- a/arch/arm64/mm/mmu.c > +++ b/arch/arm64/mm/mmu.c > @@ -512,7 +512,8 @@ static void __init map_mem(pgd_t *pgdp) > * if MTE is present. Otherwise, it has the same attributes as > * PAGE_KERNEL. > */ > - __map_memblock(pgdp, start, end, PAGE_KERNEL_TAGGED, flags); > + __map_memblock(pgdp, start, end, pgprot_tagged(PAGE_KERNEL), > + flags); > } > > /* > diff --git a/include/linux/pgtable.h b/include/linux/pgtable.h > index cdfc4e9f253e..5e772392a379 100644 > --- a/include/linux/pgtable.h > +++ b/include/linux/pgtable.h > @@ -904,6 +904,10 @@ static inline void ptep_modify_prot_commit(struct vm_area_struct *vma, > #define pgprot_device pgprot_noncached > #endif > > +#ifndef pgprot_mhp > +#define pgprot_mhp(prot) (prot) > +#endif > + > #ifdef CONFIG_MMU > #ifndef pgprot_modify > #define pgprot_modify pgprot_modify > diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c > index 5ba51a8bdaeb..0cdbbfbc5757 100644 > --- a/mm/memory_hotplug.c > +++ b/mm/memory_hotplug.c > @@ -1072,7 +1072,7 @@ static int online_memory_block(struct memory_block *mem, void *arg) > */ > int __ref add_memory_resource(int nid, struct resource *res, mhp_t mhp_flags) > { > - struct mhp_params params = { .pgprot = PAGE_KERNEL }; > + struct mhp_params params = { .pgprot = pgprot_mhp(PAGE_KERNEL) }; > u64 start, size; > bool new_node = false; > int ret; >
On 3/9/21 5:56 PM, Catalin Marinas wrote: > In a system supporting MTE, the linear map must allow reading/writing > allocation tags by setting the memory type as Normal Tagged. Currently, > this is only handled for memory present at boot. Hotplugged memory uses > Normal non-Tagged memory. > > Introduce pgprot_mhp() for hotplugged memory and use it in > add_memory_resource(). The arm64 code maps pgprot_mhp() to > pgprot_tagged(). > > Note that ZONE_DEVICE memory should not be mapped as Tagged and > therefore setting the memory type in arch_add_memory() is not feasible. > > Signed-off-by: Catalin Marinas <catalin.marinas@arm.com> > Fixes: 0178dc761368 ("arm64: mte: Use Normal Tagged attributes for the linear map") > Reported-by: Patrick Daly <pdaly@codeaurora.org> > Link: https://lore.kernel.org/r/1614745263-27827-1-git-send-email-pdaly@codeaurora.org > Cc: <stable@vger.kernel.org> # 5.10.x > Cc: Will Deacon <will@kernel.org> > Cc: Andrew Morton <akpm@linux-foundation.org> > Cc: Vincenzo Frascino <vincenzo.frascino@arm.com> > Cc: David Hildenbrand <david@redhat.com> Reviewed-by: Anshuman Khandual <anshuman.khandual@arm.com>
On Tue, 9 Mar 2021 12:26:01 +0000, Catalin Marinas wrote: > In a system supporting MTE, the linear map must allow reading/writing > allocation tags by setting the memory type as Normal Tagged. Currently, > this is only handled for memory present at boot. Hotplugged memory uses > Normal non-Tagged memory. > > Introduce pgprot_mhp() for hotplugged memory and use it in > add_memory_resource(). The arm64 code maps pgprot_mhp() to > pgprot_tagged(). > > [...] Applied to arm64 (for-next/fixes), thanks! [1/1] arm64: mte: Map hotplugged memory as Normal Tagged https://git.kernel.org/arm64/c/d15dfd31384b Cheers,
From: Catalin Marinas <catalin.marinas@arm.com> > In a system supporting MTE, the linear map must allow reading/writing > allocation tags by setting the memory type as Normal Tagged. Currently, > this is only handled for memory present at boot. Hotplugged memory uses > Normal non-Tagged memory. > > Introduce pgprot_mhp() for hotplugged memory and use it in > add_memory_resource(). The arm64 code maps pgprot_mhp() to > pgprot_tagged(). > > Note that ZONE_DEVICE memory should not be mapped as Tagged and > therefore setting the memory type in arch_add_memory() is not feasible. > > Signed-off-by: Catalin Marinas <catalin.marinas@arm.com> > Fixes: 0178dc761368 ("arm64: mte: Use Normal Tagged attributes for the linear map") > Reported-by: Patrick Daly <pdaly@codeaurora.org> > Link: https://lore.kernel.org/r/1614745263-27827-1-git-send-email-pdaly@codeaurora.org > Cc: <stable@vger.kernel.org> # 5.10.x > Cc: Will Deacon <will@kernel.org> > Cc: Andrew Morton <akpm@linux-foundation.org> > Cc: Vincenzo Frascino <vincenzo.frascino@arm.com> > Cc: David Hildenbrand <david@redhat.com> > --- > > Patrick, could you please give this patch a try on your platform? Thanks. It works for me. Sry for the double reply.
diff --git a/arch/arm64/include/asm/pgtable-prot.h b/arch/arm64/include/asm/pgtable-prot.h index 046be789fbb4..9a65fb528110 100644 --- a/arch/arm64/include/asm/pgtable-prot.h +++ b/arch/arm64/include/asm/pgtable-prot.h @@ -66,7 +66,6 @@ extern bool arm64_use_ng_mappings; #define _PAGE_DEFAULT (_PROT_DEFAULT | PTE_ATTRINDX(MT_NORMAL)) #define PAGE_KERNEL __pgprot(PROT_NORMAL) -#define PAGE_KERNEL_TAGGED __pgprot(PROT_NORMAL_TAGGED) #define PAGE_KERNEL_RO __pgprot((PROT_NORMAL & ~PTE_WRITE) | PTE_RDONLY) #define PAGE_KERNEL_ROX __pgprot((PROT_NORMAL & ~(PTE_WRITE | PTE_PXN)) | PTE_RDONLY) #define PAGE_KERNEL_EXEC __pgprot(PROT_NORMAL & ~PTE_PXN) diff --git a/arch/arm64/include/asm/pgtable.h b/arch/arm64/include/asm/pgtable.h index e17b96d0e4b5..47027796c2f9 100644 --- a/arch/arm64/include/asm/pgtable.h +++ b/arch/arm64/include/asm/pgtable.h @@ -486,6 +486,9 @@ static inline pmd_t pmd_mkdevmap(pmd_t pmd) __pgprot_modify(prot, PTE_ATTRINDX_MASK, PTE_ATTRINDX(MT_NORMAL_NC) | PTE_PXN | PTE_UXN) #define pgprot_device(prot) \ __pgprot_modify(prot, PTE_ATTRINDX_MASK, PTE_ATTRINDX(MT_DEVICE_nGnRE) | PTE_PXN | PTE_UXN) +#define pgprot_tagged(prot) \ + __pgprot_modify(prot, PTE_ATTRINDX_MASK, PTE_ATTRINDX(MT_NORMAL_TAGGED)) +#define pgprot_mhp pgprot_tagged /* * DMA allocations for non-coherent devices use what the Arm architecture calls * "Normal non-cacheable" memory, which permits speculation, unaligned accesses diff --git a/arch/arm64/mm/mmu.c b/arch/arm64/mm/mmu.c index 3802cfbdd20d..9c8aa1b44cd5 100644 --- a/arch/arm64/mm/mmu.c +++ b/arch/arm64/mm/mmu.c @@ -512,7 +512,8 @@ static void __init map_mem(pgd_t *pgdp) * if MTE is present. Otherwise, it has the same attributes as * PAGE_KERNEL. */ - __map_memblock(pgdp, start, end, PAGE_KERNEL_TAGGED, flags); + __map_memblock(pgdp, start, end, pgprot_tagged(PAGE_KERNEL), + flags); } /* diff --git a/include/linux/pgtable.h b/include/linux/pgtable.h index cdfc4e9f253e..5e772392a379 100644 --- a/include/linux/pgtable.h +++ b/include/linux/pgtable.h @@ -904,6 +904,10 @@ static inline void ptep_modify_prot_commit(struct vm_area_struct *vma, #define pgprot_device pgprot_noncached #endif +#ifndef pgprot_mhp +#define pgprot_mhp(prot) (prot) +#endif + #ifdef CONFIG_MMU #ifndef pgprot_modify #define pgprot_modify pgprot_modify diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c index 5ba51a8bdaeb..0cdbbfbc5757 100644 --- a/mm/memory_hotplug.c +++ b/mm/memory_hotplug.c @@ -1072,7 +1072,7 @@ static int online_memory_block(struct memory_block *mem, void *arg) */ int __ref add_memory_resource(int nid, struct resource *res, mhp_t mhp_flags) { - struct mhp_params params = { .pgprot = PAGE_KERNEL }; + struct mhp_params params = { .pgprot = pgprot_mhp(PAGE_KERNEL) }; u64 start, size; bool new_node = false; int ret;
In a system supporting MTE, the linear map must allow reading/writing allocation tags by setting the memory type as Normal Tagged. Currently, this is only handled for memory present at boot. Hotplugged memory uses Normal non-Tagged memory. Introduce pgprot_mhp() for hotplugged memory and use it in add_memory_resource(). The arm64 code maps pgprot_mhp() to pgprot_tagged(). Note that ZONE_DEVICE memory should not be mapped as Tagged and therefore setting the memory type in arch_add_memory() is not feasible. Signed-off-by: Catalin Marinas <catalin.marinas@arm.com> Fixes: 0178dc761368 ("arm64: mte: Use Normal Tagged attributes for the linear map") Reported-by: Patrick Daly <pdaly@codeaurora.org> Link: https://lore.kernel.org/r/1614745263-27827-1-git-send-email-pdaly@codeaurora.org Cc: <stable@vger.kernel.org> # 5.10.x Cc: Will Deacon <will@kernel.org> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Vincenzo Frascino <vincenzo.frascino@arm.com> Cc: David Hildenbrand <david@redhat.com> --- Patrick, could you please give this patch a try on your platform? Thanks. arch/arm64/include/asm/pgtable-prot.h | 1 - arch/arm64/include/asm/pgtable.h | 3 +++ arch/arm64/mm/mmu.c | 3 ++- include/linux/pgtable.h | 4 ++++ mm/memory_hotplug.c | 2 +- 5 files changed, 10 insertions(+), 3 deletions(-)