On Tue, Jul 19, 2022 at 04:49:09PM -0700, Peter Collingbourne wrote:
On a system that implements FEAT_EPAN, read/write access to the idmap is denied because UXN is not set on the swapper PTEs. As a result, idmap_kpti_install_ng_mappings panics the kernel when accessing __idmap_kpti_flag. Fix it by setting UXN on these PTEs.
Fixes: 18107f8a2df6 ("arm64: Support execute-only permissions with Enhanced PAN") Cc: stable@vger.kernel.org # 5.15 Link: https://linux-review.googlesource.com/id/Ic452fa4b4f74753e54f71e61027e7222a0... Signed-off-by: Peter Collingbourne pcc@google.com
This fix is no longer needed since commit c3cee924bd85 ("arm64: head: cover entire kernel image in initial ID map"), which moved __idmap_kpti_flag to .data, but that commit is currently only present in next.
arch/arm64/include/asm/kernel-pgtable.h | 4 ++-- arch/arm64/kernel/head.S | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/arch/arm64/include/asm/kernel-pgtable.h b/arch/arm64/include/asm/kernel-pgtable.h index 96dc0f7da258..a971d462f531 100644 --- a/arch/arm64/include/asm/kernel-pgtable.h +++ b/arch/arm64/include/asm/kernel-pgtable.h @@ -103,8 +103,8 @@ /*
- Initial memory map attributes.
*/ -#define SWAPPER_PTE_FLAGS (PTE_TYPE_PAGE | PTE_AF | PTE_SHARED) -#define SWAPPER_PMD_FLAGS (PMD_TYPE_SECT | PMD_SECT_AF | PMD_SECT_S) +#define SWAPPER_PTE_FLAGS (PTE_TYPE_PAGE | PTE_AF | PTE_SHARED | PTE_UXN) +#define SWAPPER_PMD_FLAGS (PMD_TYPE_SECT | PMD_SECT_AF | PMD_SECT_S | PMD_SECT_UXN) #if ARM64_KERNEL_USES_PMD_MAPS #define SWAPPER_MM_MMUFLAGS (PMD_ATTRINDX(MT_NORMAL) | SWAPPER_PMD_FLAGS) diff --git a/arch/arm64/kernel/head.S b/arch/arm64/kernel/head.S index 6a98f1a38c29..8a93a0a7489b 100644 --- a/arch/arm64/kernel/head.S +++ b/arch/arm64/kernel/head.S @@ -285,7 +285,7 @@ SYM_FUNC_START_LOCAL(__create_page_tables) subs x1, x1, #64 b.ne 1b
- mov x7, SWAPPER_MM_MMUFLAGS
- mov_q x7, SWAPPER_MM_MMUFLAGS
Heh, nice find:
Acked-by: Will Deacon will@kernel.org
Catalin, I guess this needs to land for 5.19 given that it's no longer relevant after the idmap rework pending for 5.20? If it all explodes in -next, I can merge in for-next/fixes.
Will