x86/pti: unpoison pgd for trusted boot
Subsystems (2)
Dave Hansen
maintainerPATCH 1/1 x86/pti: unpoison pgd for trusted boot
Updated to make this on top of x86/pti.
--
From: Dave Hansen <[email protected]>
The code in -tip potentially misses the pgd clearing if pud_alloc() sets a PGD. It would also be nice to have that comment back.
Note that the -tip commit probably works in *practice* because for two adjacent calls to map_tboot_page() that share a PGD entry, the first will clear NX, *then* allocate and set the PGD (without NX clear). The second call will *not* allocate but will clear the NX bit.
This just defers the NX clearing to a point after it is known that all top-level allocations have occurred. Add a comment to clarify why.
---
b/arch/x86/kernel/tboot.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-)
arch/x86/kernel/tboot.c renamed from arch/x86/kernel/tboot.c~pti-tboot-fix +11 -1
@@ -127,7 +127,6 @@ static int map_tboot_page(unsigned long p4d =p4d_alloc (&tboot_mm, pgd, vaddr);if (!p4d)return -1 ; - pgd->pgd &= ~_PAGE_NX; pud =pud_alloc (&tboot_mm, p4d, vaddr);if (!pud)return -1 ;
@@ -139,6 +138,17 @@ static int map_tboot_page(unsigned longreturn -1 ;set_pte_at (&tboot_mm, vaddr, pte,pfn_pte (pfn, prot));pte_unmap (pte); + +/* +* PTI poisons low addresses in the kernel page tables in the +* name of making them unusable for userspace. To execute +* code at such a low address, the poison must be cleared. +* +* Note: 'pgd' actually gets set in p4d_alloc() _or_ +* pud_alloc() depending on 4/5-level paging. +*/ + pgd->pgd &= ~_PAGE_NX; +return 0 ; }
_
Quoted 2 lines · Dave Hansen
Updated to make this on top of x86/pti.
tip-bot for Dave Hansen
merge bot[tip:x86/pti] x86/pti: Make unpoison of pgd for trusted boot work for real
In reply to Dave Hansenx86/pti: Make unpoison of pgd for trusted boot work for real
Commit-ID: 8a931d1e24bacf01f00a35d43bfe7917256c5c49 Gitweb: https://git.kernel.org/tip/8a931d1e24bacf01f00a35d43bfe7917256c5c49 Author: Dave Hansen <[email protected]> AuthorDate: Wed, 10 Jan 2018 14:49:39 -0800 Committer: Thomas Gleixner <[email protected]> CommitDate: Thu, 11 Jan 2018 00:30:25 +0100
x86/pti: Make unpoison of pgd for trusted boot work for real
The inital fix for trusted boot and PTI potentially misses the pgd clearing if pud_alloc() sets a PGD. It probably works in *practice* because for two adjacent calls to map_tboot_page() that share a PGD entry, the first will clear NX, *then* allocate and set the PGD (without NX clear). The second call will *not* allocate but will clear the NX bit.
Defer the NX clearing to a point after it is known that all top-level allocations have occurred. Add a comment to clarify why.
[ tglx: Massaged changelog ]
--- arch/x86/kernel/tboot.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-)
arch/x86/kernel/tboot.c +11 -1
@@ -127,7 +127,6 @@ static int map_tboot_page(unsigned long vaddr, unsigned long pfn, p4d =p4d_alloc (&tboot_mm, pgd, vaddr);if (!p4d)return -1 ; - pgd->pgd &= ~_PAGE_NX; pud =pud_alloc (&tboot_mm, p4d, vaddr);if (!pud)return -1 ;
@@ -139,6 +138,17 @@ static int map_tboot_page(unsigned long vaddr, unsigned long pfn,return -1 ;set_pte_at (&tboot_mm, vaddr, pte,pfn_pte (pfn, prot));pte_unmap (pte); + +/* +* PTI poisons low addresses in the kernel page tables in the +* name of making them unusable for userspace. To execute +* code at such a low address, the poison must be cleared. +* +* Note: 'pgd' actually gets set in p4d_alloc() _or_ +* pud_alloc() depending on 4/5-level paging. +*/ + pgd->pgd &= ~_PAGE_NX; +return 0 ; }
tip-bot for Dave Hansen
merge bot[tip:x86/pti] x86/pti: Make unpoison of pgd for trusted boot work for real
In reply to Dave Hansenx86/pti: Make unpoison of pgd for trusted boot work for real
Commit-ID: 445b69e3b75e42362a5bdc13c8b8f61599e2228a Gitweb: https://git.kernel.org/tip/445b69e3b75e42362a5bdc13c8b8f61599e2228a Author: Dave Hansen <[email protected]> AuthorDate: Wed, 10 Jan 2018 14:49:39 -0800 Committer: Thomas Gleixner <[email protected]> CommitDate: Thu, 11 Jan 2018 23:36:59 +0100
x86/pti: Make unpoison of pgd for trusted boot work for real
The inital fix for trusted boot and PTI potentially misses the pgd clearing if pud_alloc() sets a PGD. It probably works in *practice* because for two adjacent calls to map_tboot_page() that share a PGD entry, the first will clear NX, *then* allocate and set the PGD (without NX clear). The second call will *not* allocate but will clear the NX bit.
Defer the NX clearing to a point after it is known that all top-level allocations have occurred. Add a comment to clarify why.
[ tglx: Massaged changelog ]
--- arch/x86/kernel/tboot.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-)
arch/x86/kernel/tboot.c +11 -1
@@ -127,7 +127,6 @@ static int map_tboot_page(unsigned long vaddr, unsigned long pfn, p4d =p4d_alloc (&tboot_mm, pgd, vaddr);if (!p4d)return -1 ; - pgd->pgd &= ~_PAGE_NX; pud =pud_alloc (&tboot_mm, p4d, vaddr);if (!pud)return -1 ;
@@ -139,6 +138,17 @@ static int map_tboot_page(unsigned long vaddr, unsigned long pfn,return -1 ;set_pte_at (&tboot_mm, vaddr, pte,pfn_pte (pfn, prot));pte_unmap (pte); + +/* +* PTI poisons low addresses in the kernel page tables in the +* name of making them unusable for userspace. To execute +* code at such a low address, the poison must be cleared. +* +* Note: 'pgd' actually gets set in p4d_alloc() _or_ +* pud_alloc() depending on 4/5-level paging. +*/ + pgd->pgd &= ~_PAGE_NX; +return 0 ; }