[v6,12/41] s390/mm: Introduce pmd_mkwrite_kernel()

Commit Message

Rick Edgecombe Feb. 18, 2023, 9:14 p.m. UTC

The x86 Control-flow Enforcement Technology (CET) feature includes a new
type of memory called shadow stack. This shadow stack memory has some
unusual properties, which requires some core mm changes to function
properly.

One of these changes is to allow for pmd_mkwrite() to create different
types of writable memory (the existing conventionally writable type and
also the new shadow stack type). Future patches will convert pmd_mkwrite()
to take a VMA in order to facilitate this, however there are places in the
kernel where pmd_mkwrite() is called outside of the context of a VMA.
These are for kernel memory. So create a new variant called
pmd_mkwrite_kernel() and switch the kernel users over to it. Have
pmd_mkwrite() and pmd_mkwrite_kernel() be the same for now. Future patches
will introduce changes to make pmd_mkwrite() take a VMA.

Only do this for architectures that need it because they call pmd_mkwrite()
in arch code without an associated VMA. Since it will only currently be
used in arch code, so do not include it in arch_pgtable_helpers.rst.

Cc: [e-mail]
Cc: [e-mail]
Cc: [e-mail]
Cc: [e-mail]
Tested-by: Pengfei Xu <[e-mail]>
Suggested-by: David Hildenbrand <[e-mail]>
Signed-off-by: Rick Edgecombe <[e-mail]>

---
Hi Non-x86 Arch’s,

x86 has a feature that allows for the creation of a special type of
writable memory (shadow stack) that is only writable in limited specific
ways. Previously, changes were proposed to core MM code to teach it to
decide when to create normally writable memory or the special shadow stack
writable memory, but David Hildenbrand suggested[0] to change
pXX_mkwrite() to take a VMA, so awareness of shadow stack memory can be
moved into x86 code.

Since pXX_mkwrite() is defined in every arch, it requires some tree-wide
changes. So that is why you are seeing some patches out of a big x86
series pop up in your arch mailing list. There is no functional change.
After this refactor, the shadow stack series goes on to use the arch
helpers to push shadow stack memory details inside arch/x86.

Testing was just 0-day build testing.

Hopefully that is enough context. Thanks!

[0] https://lore.kernel.org/lkml/[e-mail]/#t

v6:
 - New patch
---
 arch/s390/include/asm/pgtable.h | 7 ++++++-
 arch/s390/mm/pageattr.c         | 2 +-
 2 files changed, 7 insertions(+), 2 deletions(-)

Comments

Kees Cook Feb. 19, 2023, 8:39 p.m. UTC | #1

On Sat, Feb 18, 2023 at 01:14:04PM -0800, Rick Edgecombe wrote:
> The x86 Control-flow Enforcement Technology (CET) feature includes a new
> type of memory called shadow stack. This shadow stack memory has some
> unusual properties, which requires some core mm changes to function
> properly.
> 
> One of these changes is to allow for pmd_mkwrite() to create different
> types of writable memory (the existing conventionally writable type and
> also the new shadow stack type). Future patches will convert pmd_mkwrite()
> to take a VMA in order to facilitate this, however there are places in the
> kernel where pmd_mkwrite() is called outside of the context of a VMA.
> These are for kernel memory. So create a new variant called
> pmd_mkwrite_kernel() and switch the kernel users over to it. Have
> pmd_mkwrite() and pmd_mkwrite_kernel() be the same for now. Future patches
> will introduce changes to make pmd_mkwrite() take a VMA.
> 
> Only do this for architectures that need it because they call pmd_mkwrite()
> in arch code without an associated VMA. Since it will only currently be
> used in arch code, so do not include it in arch_pgtable_helpers.rst.
> 
> Cc: [e-mail]
> Cc: [e-mail]
> Cc: [e-mail]
> Cc: [e-mail]
> Tested-by: Pengfei Xu <[e-mail]>
> Suggested-by: David Hildenbrand <[e-mail]>
> Signed-off-by: Rick Edgecombe <[e-mail]>

Yup, 1:1 refactor.

Reviewed-by: Kees Cook <[e-mail]>

David Hildenbrand Feb. 20, 2023, 11:21 a.m. UTC | #2

On 18.02.23 22:14, Rick Edgecombe wrote:
> The x86 Control-flow Enforcement Technology (CET) feature includes a new
> type of memory called shadow stack. This shadow stack memory has some
> unusual properties, which requires some core mm changes to function
> properly.
> 
> One of these changes is to allow for pmd_mkwrite() to create different
> types of writable memory (the existing conventionally writable type and
> also the new shadow stack type). Future patches will convert pmd_mkwrite()
> to take a VMA in order to facilitate this, however there are places in the
> kernel where pmd_mkwrite() is called outside of the context of a VMA.
> These are for kernel memory. So create a new variant called
> pmd_mkwrite_kernel() and switch the kernel users over to it. Have
> pmd_mkwrite() and pmd_mkwrite_kernel() be the same for now. Future patches
> will introduce changes to make pmd_mkwrite() take a VMA.
> 
> Only do this for architectures that need it because they call pmd_mkwrite()
> in arch code without an associated VMA. Since it will only currently be
> used in arch code, so do not include it in arch_pgtable_helpers.rst.
> 
> Cc: [e-mail]
> Cc: [e-mail]
> Cc: [e-mail]
> Cc: [e-mail]
> Tested-by: Pengfei Xu <[e-mail]>
> Suggested-by: David Hildenbrand <[e-mail]>
> Signed-off-by: Rick Edgecombe <[e-mail]>
> 

Heh, that answers my question to patch #11

Acked-by: David Hildenbrand <[e-mail]>

Heiko Carstens Feb. 23, 2023, 12:14 p.m. UTC | #3

On Sat, Feb 18, 2023 at 01:14:04PM -0800, Rick Edgecombe wrote:
> The x86 Control-flow Enforcement Technology (CET) feature includes a new
> type of memory called shadow stack. This shadow stack memory has some
> unusual properties, which requires some core mm changes to function
> properly.
> 
> One of these changes is to allow for pmd_mkwrite() to create different
> types of writable memory (the existing conventionally writable type and
> also the new shadow stack type). Future patches will convert pmd_mkwrite()
> to take a VMA in order to facilitate this, however there are places in the
> kernel where pmd_mkwrite() is called outside of the context of a VMA.
> These are for kernel memory. So create a new variant called
> pmd_mkwrite_kernel() and switch the kernel users over to it. Have
> pmd_mkwrite() and pmd_mkwrite_kernel() be the same for now. Future patches
> will introduce changes to make pmd_mkwrite() take a VMA.
> 
> Only do this for architectures that need it because they call pmd_mkwrite()
> in arch code without an associated VMA. Since it will only currently be
> used in arch code, so do not include it in arch_pgtable_helpers.rst.
> 
> Cc: [e-mail]
> Cc: [e-mail]
> Cc: [e-mail]
> Cc: [e-mail]
> Tested-by: Pengfei Xu <[e-mail]>
> Suggested-by: David Hildenbrand <[e-mail]>
> Signed-off-by: Rick Edgecombe <[e-mail]>
...
> ---
>  arch/s390/include/asm/pgtable.h | 7 ++++++-
>  arch/s390/mm/pageattr.c         | 2 +-
>  2 files changed, 7 insertions(+), 2 deletions(-)

Acked-by: Heiko Carstens <[e-mail]>

Rick Edgecombe Feb. 23, 2023, 5:59 p.m. UTC | #4

On Thu, 2023-02-23 at 13:14 +0100, Heiko Carstens wrote:
> Acked-by: Heiko Carstens <[e-mail]>

Thanks!

Patch

diff --git a/arch/s390/include/asm/pgtable.h b/arch/s390/include/asm/pgtable.h
index 29522418b5f4..c48a447d1432 100644
--- a/arch/s390/include/asm/pgtable.h
+++ b/arch/s390/include/asm/pgtable.h
@@ -1425,7 +1425,7 @@  static inline pmd_t pmd_wrprotect(pmd_t pmd)
 	return set_pmd_bit(pmd, __pgprot(_SEGMENT_ENTRY_PROTECT));
 }
 
-static inline pmd_t pmd_mkwrite(pmd_t pmd)
+static inline pmd_t pmd_mkwrite_kernel(pmd_t pmd)
 {
 	pmd = set_pmd_bit(pmd, __pgprot(_SEGMENT_ENTRY_WRITE));
 	if (pmd_val(pmd) & _SEGMENT_ENTRY_DIRTY)
@@ -1433,6 +1433,11 @@  static inline pmd_t pmd_mkwrite(pmd_t pmd)
 	return pmd;
 }
 
+static inline pmd_t pmd_mkwrite(pmd_t pmd)
+{
+	return pmd_mkwrite_kernel(pmd);
+}
+
 static inline pmd_t pmd_mkclean(pmd_t pmd)
 {
 	pmd = clear_pmd_bit(pmd, __pgprot(_SEGMENT_ENTRY_DIRTY));
diff --git a/arch/s390/mm/pageattr.c b/arch/s390/mm/pageattr.c
index 4ee5fe5caa23..7b6967dfacd0 100644
--- a/arch/s390/mm/pageattr.c
+++ b/arch/s390/mm/pageattr.c
@@ -146,7 +146,7 @@  static void modify_pmd_page(pmd_t *pmdp, unsigned long addr,
 	if (flags & SET_MEMORY_RO)
 		new = pmd_wrprotect(new);
 	else if (flags & SET_MEMORY_RW)
-		new = pmd_mkwrite(pmd_mkdirty(new));
+		new = pmd_mkwrite_kernel(pmd_mkdirty(new));
 	if (flags & SET_MEMORY_NX)
 		new = set_pmd_bit(new, __pgprot(_SEGMENT_ENTRY_NOEXEC));
 	else if (flags & SET_MEMORY_X)