2023-06-14 16:59:16

by Lee Jones

Subject: [PATCH v2 1/1] x86/mm/KASLR: Store pud_page_tramp into entry rather than page

set_pgd() expects to be passed whole pages to operate on, whereas
trampoline_pgd_entry is, as the name suggests, an entry. The
ramifications for using set_pgd() here are that the following thread of
execution will not only place the suggested value into the
trampoline_pgd_entry (8-Byte globally stored [.bss]) variable, PTI will
also attempt to replicate that value into the non-existent neighboring
user page (located +4k away), leading to the corruption of other global
[.bss] stored variables.

Suggested-by: Dave Hansen <[email protected]>
Signed-off-by: Lee Jones <[email protected]>
---
arch/x86/mm/kaslr.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/arch/x86/mm/kaslr.c b/arch/x86/mm/kaslr.c
index 557f0fe25dff4..37db264866b64 100644
--- a/arch/x86/mm/kaslr.c
+++ b/arch/x86/mm/kaslr.c
@@ -172,10 +172,10 @@ void __meminit init_trampoline_kaslr(void)
set_p4d(p4d_tramp,
__p4d(_KERNPG_TABLE | __pa(pud_page_tramp)));

- set_pgd(&trampoline_pgd_entry,
- __pgd(_KERNPG_TABLE | __pa(p4d_page_tramp)));
+ trampoline_pgd_entry =
+ __pgd(_KERNPG_TABLE | __pa(p4d_page_tramp));
} else {
- set_pgd(&trampoline_pgd_entry,
- __pgd(_KERNPG_TABLE | __pa(pud_page_tramp)));
+ trampoline_pgd_entry =
+ __pgd(_KERNPG_TABLE | __pa(pud_page_tramp));
}
}
--
2.41.0.162.gfafddb0af9-goog

2023-06-14 18:07:02

by Nick Desaulniers

Subject: Re: [PATCH v2 1/1] x86/mm/KASLR: Store pud_page_tramp into entry rather than page

On Wed, Jun 14, 2023 at 05:38:54PM +0100, Lee Jones wrote:
> set_pgd() expects to be passed whole pages to operate on, whereas
> trampoline_pgd_entry is, as the name suggests, an entry. The
> ramifications for using set_pgd() here are that the following thread of
> execution will not only place the suggested value into the
> trampoline_pgd_entry (8-Byte globally stored [.bss]) variable, PTI will
> also attempt to replicate that value into the non-existent neighboring
> user page (located +4k away), leading to the corruption of other global
> [.bss] stored variables.
>
> Suggested-by: Dave Hansen <[email protected]>
> Signed-off-by: Lee Jones <[email protected]>

Nice work tracking this one down!

Fixes: 0925dda5962e ("x86/mm/KASLR: Use only one PUD entry for real mode trampoline")
Cc: <[email protected]>

> ---
> arch/x86/mm/kaslr.c | 8 ++++----
> 1 file changed, 4 insertions(+), 4 deletions(-)
>
> diff --git a/arch/x86/mm/kaslr.c b/arch/x86/mm/kaslr.c
> index 557f0fe25dff4..37db264866b64 100644
> --- a/arch/x86/mm/kaslr.c
> +++ b/arch/x86/mm/kaslr.c
> @@ -172,10 +172,10 @@ void __meminit init_trampoline_kaslr(void)
> set_p4d(p4d_tramp,
> __p4d(_KERNPG_TABLE | __pa(pud_page_tramp)));
>
> - set_pgd(&trampoline_pgd_entry,
> - __pgd(_KERNPG_TABLE | __pa(p4d_page_tramp)));
> + trampoline_pgd_entry =
> + __pgd(_KERNPG_TABLE | __pa(p4d_page_tramp));
> } else {
> - set_pgd(&trampoline_pgd_entry,
> - __pgd(_KERNPG_TABLE | __pa(pud_page_tramp)));
> + trampoline_pgd_entry =
> + __pgd(_KERNPG_TABLE | __pa(pud_page_tramp));
> }
> }
> --
> 2.41.0.162.gfafddb0af9-goog
>

2023-06-14 19:00:24

by Lee Jones

Subject: Re: [PATCH v2 1/1] x86/mm/KASLR: Store pud_page_tramp into entry rather than page

On Wed, 14 Jun 2023, Nick Desaulniers wrote:

> On Wed, Jun 14, 2023 at 05:38:54PM +0100, Lee Jones wrote:
> > set_pgd() expects to be passed whole pages to operate on, whereas
> > trampoline_pgd_entry is, as the name suggests, an entry. The
> > ramifications for using set_pgd() here are that the following thread of
> > execution will not only place the suggested value into the
> > trampoline_pgd_entry (8-Byte globally stored [.bss]) variable, PTI will
> > also attempt to replicate that value into the non-existent neighboring
> > user page (located +4k away), leading to the corruption of other global
> > [.bss] stored variables.
> >
> > Suggested-by: Dave Hansen <[email protected]>
> > Signed-off-by: Lee Jones <[email protected]>
>
> Nice work tracking this one down!
>
> Fixes: 0925dda5962e ("x86/mm/KASLR: Use only one PUD entry for real mode trampoline")
> Cc: <[email protected]>

I was planning on backporting this myself once it hits Mainline.

Happy to submit a RESEND with those pieces added if required though.

Or perhaps someone would be kind enough to add them on merge?

> > ---
> > arch/x86/mm/kaslr.c | 8 ++++----
> > 1 file changed, 4 insertions(+), 4 deletions(-)
> >
> > diff --git a/arch/x86/mm/kaslr.c b/arch/x86/mm/kaslr.c
> > index 557f0fe25dff4..37db264866b64 100644
> > --- a/arch/x86/mm/kaslr.c
> > +++ b/arch/x86/mm/kaslr.c
> > @@ -172,10 +172,10 @@ void __meminit init_trampoline_kaslr(void)
> > set_p4d(p4d_tramp,
> > __p4d(_KERNPG_TABLE | __pa(pud_page_tramp)));
> >
> > - set_pgd(&trampoline_pgd_entry,
> > - __pgd(_KERNPG_TABLE | __pa(p4d_page_tramp)));
> > + trampoline_pgd_entry =
> > + __pgd(_KERNPG_TABLE | __pa(p4d_page_tramp));
> > } else {
> > - set_pgd(&trampoline_pgd_entry,
> > - __pgd(_KERNPG_TABLE | __pa(pud_page_tramp)));
> > + trampoline_pgd_entry =
> > + __pgd(_KERNPG_TABLE | __pa(pud_page_tramp));
> > }
> > }
> > --
> > 2.41.0.162.gfafddb0af9-goog
> >

--
Lee Jones [李琼斯]

2023-06-16 17:26:37

by Lee Jones

Subject: Re: [PATCH v2 1/1] x86/mm/KASLR: Store pud_page_tramp into entry rather than page

On Wed, 14 Jun 2023, Lee Jones wrote:

> set_pgd() expects to be passed whole pages to operate on, whereas
> trampoline_pgd_entry is, as the name suggests, an entry. The
> ramifications for using set_pgd() here are that the following thread of
> execution will not only place the suggested value into the
> trampoline_pgd_entry (8-Byte globally stored [.bss]) variable, PTI will
> also attempt to replicate that value into the non-existent neighboring
> user page (located +4k away), leading to the corruption of other global
> [.bss] stored variables.
>
> Suggested-by: Dave Hansen <[email protected]>
> Signed-off-by: Lee Jones <[email protected]>
> ---
> arch/x86/mm/kaslr.c | 8 ++++----
> 1 file changed, 4 insertions(+), 4 deletions(-)

Is there any more you need from me at this point?

Would you like me to resubmit with the Fixes: tag applied, or is
someone happy to apply it on merge?

> diff --git a/arch/x86/mm/kaslr.c b/arch/x86/mm/kaslr.c
> index 557f0fe25dff4..37db264866b64 100644
> --- a/arch/x86/mm/kaslr.c
> +++ b/arch/x86/mm/kaslr.c
> @@ -172,10 +172,10 @@ void __meminit init_trampoline_kaslr(void)
> set_p4d(p4d_tramp,
> __p4d(_KERNPG_TABLE | __pa(pud_page_tramp)));
>
> - set_pgd(&trampoline_pgd_entry,
> - __pgd(_KERNPG_TABLE | __pa(p4d_page_tramp)));
> + trampoline_pgd_entry =
> + __pgd(_KERNPG_TABLE | __pa(p4d_page_tramp));
> } else {
> - set_pgd(&trampoline_pgd_entry,
> - __pgd(_KERNPG_TABLE | __pa(pud_page_tramp)));
> + trampoline_pgd_entry =
> + __pgd(_KERNPG_TABLE | __pa(pud_page_tramp));
> }
> }
> --
> 2.41.0.162.gfafddb0af9-goog
>

--
Lee Jones [李琼斯]

2023-06-19 11:30:59

by Lee Jones

Subject: Re: [PATCH v2 1/1] x86/mm/KASLR: Store pud_page_tramp into entry rather than page

On Fri, 16 Jun 2023, Lee Jones wrote:

> On Wed, 14 Jun 2023, Lee Jones wrote:
>
> > set_pgd() expects to be passed whole pages to operate on, whereas
> > trampoline_pgd_entry is, as the name suggests, an entry. The
> > ramifications for using set_pgd() here are that the following thread of
> > execution will not only place the suggested value into the
> > trampoline_pgd_entry (8-Byte globally stored [.bss]) variable, PTI will
> > also attempt to replicate that value into the non-existent neighboring
> > user page (located +4k away), leading to the corruption of other global
> > [.bss] stored variables.
> >
> > Suggested-by: Dave Hansen <[email protected]>
> > Signed-off-by: Lee Jones <[email protected]>
> > ---
> > arch/x86/mm/kaslr.c | 8 ++++----
> > 1 file changed, 4 insertions(+), 4 deletions(-)
>
> Is there any more you need from me at this point?
>
> Would you like me to resubmit with the Fixes: tag applied, or is
> someone happy to apply it on merge?

Superstar, thanks Dave.

--
Lee Jones [李琼斯]