| Message ID | 1456151158-25849-3-git-send-email-ard.biesheuvel@linaro.org |
|---|---|
| State | New |
| Headers | show |
On 2 March 2016 at 13:10, Matt Fleming <matt@codeblueprint.co.uk> wrote: > On Mon, 22 Feb, at 03:25:55PM, Ard Biesheuvel wrote: >> Recent UEFI versions expose permission attributes for runtime services >> memory regions, either in the UEFI memory map or in the separate memory >> attributes table. This allows the kernel to map these regions with >> stricter permissions, rather than the RWX permissions that are used by >> default. So wire this up in our mapping routine. >> >> Note that in the absence of permission attributes, we still only map >> regions of type EFI_RUNTIME_SERVICE_CODE with the executable bit set. >> Also, we base the mapping attributes of EFI_MEMORY_MAPPED_IO on the >> type directly rather than on the absence of the EFI_MEMORY_WB attribute. >> This is more correct, but is also required for compatibility with the >> upcoming support for the Memory Attributes Table, which only carries >> permission attributes, not memory type attributes. >> >> Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> >> --- >> arch/arm64/kernel/efi.c | 27 ++++++++++++++++---- >> 1 file changed, 22 insertions(+), 5 deletions(-) >> >> diff --git a/arch/arm64/kernel/efi.c b/arch/arm64/kernel/efi.c >> index b6abc852f2a1..3364408c154f 100644 >> --- a/arch/arm64/kernel/efi.c >> +++ b/arch/arm64/kernel/efi.c >> @@ -24,15 +24,32 @@ int __init efi_create_mapping(struct mm_struct *mm, efi_memory_desc_t *md) >> /* >> * Only regions of type EFI_RUNTIME_SERVICES_CODE need to be >> * executable, everything else can be mapped with the XN bits >> - * set. >> + * set. Also take the new (optional) RO/XP bits into account. >> */ >> - if ((md->attribute & EFI_MEMORY_WB) == 0) >> + if (md->type == EFI_MEMORY_MAPPED_IO) >> prot_val = PROT_DEVICE_nGnRE; >> - else if (md->type == EFI_RUNTIME_SERVICES_CODE || >> - !PAGE_ALIGNED(md->phys_addr)) >> + else if (WARN_ONCE(!PAGE_ALIGNED(md->phys_addr), >> + "UEFI Runtime regions are not aligned to 64 KB -- buggy firmware?")) >> + /* >> + * If the region is not aligned to the page size of the OS, we >> + * can not use strict permissions, since that would also affect >> + * the mapping attributes of the adjacent regions. >> + */ >> prot_val = pgprot_val(PAGE_KERNEL_EXEC); >> - else >> + else if ((md->attribute & (EFI_MEMORY_XP | EFI_MEMORY_RO)) == >> + (EFI_MEMORY_XP | EFI_MEMORY_RO)) >> + /* R-- */ >> + prot_val = pgprot_val(PAGE_KERNEL_RO); >> + else if (md->attribute & EFI_MEMORY_RO) >> + /* R-X */ >> + prot_val = pgprot_val(PAGE_KERNEL_ROX); >> + else if (md->attribute & EFI_MEMORY_XP || >> + md->type != EFI_RUNTIME_SERVICES_CODE) >> + /* RW- */ >> prot_val = pgprot_val(PAGE_KERNEL); >> + else >> + /* RWX */ >> + prot_val = pgprot_val(PAGE_KERNEL_EXEC); >> >> create_pgd_mapping(mm, md->phys_addr, md->virt_addr, >> md->num_pages << EFI_PAGE_SHIFT, > > The actual logic looks fine but it seems like there's quite a lot > going on in this function which is fairly difficult to decipher with > the if/else if clauses. > > Would you be open to splitting this out a little? It's just a > suggestion, but maybe something like this, > Sure, that looks a lot better. I will fold that into v2 > --- > > diff --git a/arch/arm64/kernel/efi.c b/arch/arm64/kernel/efi.c > index 3364408c154f..33a6da160a50 100644 > --- a/arch/arm64/kernel/efi.c > +++ b/arch/arm64/kernel/efi.c > @@ -17,39 +17,48 @@ > > #include <asm/efi.h> > > -int __init efi_create_mapping(struct mm_struct *mm, efi_memory_desc_t *md) > +/* > + * Only regions of type EFI_RUNTIME_SERVICES_CODE need to be > + * executable, everything else can be mapped with the XN bits > + * set. Also take the new (optional) RO/XP bits into account. > + */ > +static __init pteval_t create_mapping_protection(efi_memory_desc_t *md) > { > - pteval_t prot_val; > + u64 attr = md->attribute; > + u32 type = md->type; > > - /* > - * Only regions of type EFI_RUNTIME_SERVICES_CODE need to be > - * executable, everything else can be mapped with the XN bits > - * set. Also take the new (optional) RO/XP bits into account. > - */ > - if (md->type == EFI_MEMORY_MAPPED_IO) > - prot_val = PROT_DEVICE_nGnRE; > - else if (WARN_ONCE(!PAGE_ALIGNED(md->phys_addr), > - "UEFI Runtime regions are not aligned to 64 KB -- buggy firmware?")) > + if (type == EFI_MEMORY_MAPPED_IO) > + return PROT_DEVICE_nGnRE; > + > + if (WARN_ONCE(!PAGE_ALIGNED(md->phys_addr), > + "UEFI Runtime regions are not aligned to 64 KB -- buggy firmware?")) > /* > * If the region is not aligned to the page size of the OS, we > * can not use strict permissions, since that would also affect > * the mapping attributes of the adjacent regions. > */ > - prot_val = pgprot_val(PAGE_KERNEL_EXEC); > - else if ((md->attribute & (EFI_MEMORY_XP | EFI_MEMORY_RO)) == > - (EFI_MEMORY_XP | EFI_MEMORY_RO)) > - /* R-- */ > - prot_val = pgprot_val(PAGE_KERNEL_RO); > - else if (md->attribute & EFI_MEMORY_RO) > - /* R-X */ > - prot_val = pgprot_val(PAGE_KERNEL_ROX); > - else if (md->attribute & EFI_MEMORY_XP || > - md->type != EFI_RUNTIME_SERVICES_CODE) > - /* RW- */ > - prot_val = pgprot_val(PAGE_KERNEL); > - else > - /* RWX */ > - prot_val = pgprot_val(PAGE_KERNEL_EXEC); > + return pgprot_val(PAGE_KERNEL_EXEC); > + > + /* R-- */ > + if ((attr & (EFI_MEMORY_XP | EFI_MEMORY_RO)) == > + (EFI_MEMORY_XP | EFI_MEMORY_RO)) > + return pgprot_val(PAGE_KERNEL_RO); > + > + /* R-X */ > + if (attr & EFI_MEMORY_RO) > + return pgprot_val(PAGE_KERNEL_ROX); > + > + /* RW- */ > + if (attr & EFI_MEMORY_XP || type != EFI_RUNTIME_SERVICES_CODE) > + return pgprot_val(PAGE_KERNEL); > + > + /* RWX */ > + return pgprot_val(PAGE_KERNEL_EXEC); > +} > + > +int __init efi_create_mapping(struct mm_struct *mm, efi_memory_desc_t *md) > +{ > + pteval_t prot_val = create_mapping_protection(md); > > create_pgd_mapping(mm, md->phys_addr, md->virt_addr, > md->num_pages << EFI_PAGE_SHIFT, _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
diff --git a/arch/arm64/kernel/efi.c b/arch/arm64/kernel/efi.c index b6abc852f2a1..3364408c154f 100644 --- a/arch/arm64/kernel/efi.c +++ b/arch/arm64/kernel/efi.c @@ -24,15 +24,32 @@ int __init efi_create_mapping(struct mm_struct *mm, efi_memory_desc_t *md) /* * Only regions of type EFI_RUNTIME_SERVICES_CODE need to be * executable, everything else can be mapped with the XN bits - * set. + * set. Also take the new (optional) RO/XP bits into account. */ - if ((md->attribute & EFI_MEMORY_WB) == 0) + if (md->type == EFI_MEMORY_MAPPED_IO) prot_val = PROT_DEVICE_nGnRE; - else if (md->type == EFI_RUNTIME_SERVICES_CODE || - !PAGE_ALIGNED(md->phys_addr)) + else if (WARN_ONCE(!PAGE_ALIGNED(md->phys_addr), + "UEFI Runtime regions are not aligned to 64 KB -- buggy firmware?")) + /* + * If the region is not aligned to the page size of the OS, we + * can not use strict permissions, since that would also affect + * the mapping attributes of the adjacent regions. + */ prot_val = pgprot_val(PAGE_KERNEL_EXEC); - else + else if ((md->attribute & (EFI_MEMORY_XP | EFI_MEMORY_RO)) == + (EFI_MEMORY_XP | EFI_MEMORY_RO)) + /* R-- */ + prot_val = pgprot_val(PAGE_KERNEL_RO); + else if (md->attribute & EFI_MEMORY_RO) + /* R-X */ + prot_val = pgprot_val(PAGE_KERNEL_ROX); + else if (md->attribute & EFI_MEMORY_XP || + md->type != EFI_RUNTIME_SERVICES_CODE) + /* RW- */ prot_val = pgprot_val(PAGE_KERNEL); + else + /* RWX */ + prot_val = pgprot_val(PAGE_KERNEL_EXEC); create_pgd_mapping(mm, md->phys_addr, md->virt_addr, md->num_pages << EFI_PAGE_SHIFT,
Recent UEFI versions expose permission attributes for runtime services memory regions, either in the UEFI memory map or in the separate memory attributes table. This allows the kernel to map these regions with stricter permissions, rather than the RWX permissions that are used by default. So wire this up in our mapping routine. Note that in the absence of permission attributes, we still only map regions of type EFI_RUNTIME_SERVICE_CODE with the executable bit set. Also, we base the mapping attributes of EFI_MEMORY_MAPPED_IO on the type directly rather than on the absence of the EFI_MEMORY_WB attribute. This is more correct, but is also required for compatibility with the upcoming support for the Memory Attributes Table, which only carries permission attributes, not memory type attributes. Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> --- arch/arm64/kernel/efi.c | 27 ++++++++++++++++---- 1 file changed, 22 insertions(+), 5 deletions(-) -- 2.5.0 _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel