* [PATCH v3] lib/stackdepot: Fix global out-of-bounds in stackdepot
@ 2020-01-30 6:44 Walter Wu
2020-01-30 12:03 ` Alexander Potapenko
0 siblings, 1 reply; 6+ messages in thread
From: Walter Wu @ 2020-01-30 6:44 UTC (permalink / raw)
To: Dmitry Vyukov, Matthias Brugger, Thomas Gleixner,
Alexander Potapenko, Josh Poimboeuf, Greg Kroah-Hartman,
Kate Stewart
Cc: linux-kernel, linux-arm-kernel, linux-mediatek, wsd_upstream, Walter Wu
If the depot_index = STACK_ALLOC_MAX_SLABS - 2 and next_slab_inited = 0,
then it will cause array out-of-bounds access, so that we should modify
the detection to avoid this array out-of-bounds bug.
Assume depot_index = STACK_ALLOC_MAX_SLABS - 3
Consider following call flow sequence:
stack_depot_save()
depot_alloc_stack()
if (unlikely(depot_index + 1 >= STACK_ALLOC_MAX_SLABS)) //pass
depot_index++ //depot_index = STACK_ALLOC_MAX_SLABS - 2
if (depot_index + 1 < STACK_ALLOC_MAX_SLABS) //enter
smp_store_release(&next_slab_inited, 0); //next_slab_inited = 0
init_stack_slab()
if (stack_slabs[depot_index] == NULL) //enter and exit
stack_depot_save()
depot_alloc_stack()
if (unlikely(depot_index + 1 >= STACK_ALLOC_MAX_SLABS)) //pass
depot_index++ //depot_index = STACK_ALLOC_MAX_SLABS - 1
init_stack_slab(&prealloc)
stack_slabs[depot_index + 1] //here get global out-of-bounds
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Matthias Brugger <matthias.bgg@gmail.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Alexander Potapenko <glider@google.com>
Cc: Josh Poimboeuf <jpoimboe@redhat.com>
Cc: Kate Stewart <kstewart@linuxfoundation.org>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Kate Stewart <kstewart@linuxfoundation.org>
Signed-off-by: Walter Wu <walter-zh.wu@mediatek.com>
---
changes in v2:
modify call flow sequence and preconditon
changes in v3:
add some reviewers
---
lib/stackdepot.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/stackdepot.c b/lib/stackdepot.c
index ed717dd08ff3..7e8a15e41600 100644
--- a/lib/stackdepot.c
+++ b/lib/stackdepot.c
@@ -106,7 +106,7 @@ static struct stack_record *depot_alloc_stack(unsigned long *entries, int size,
required_size = ALIGN(required_size, 1 << STACK_ALLOC_ALIGN);
if (unlikely(depot_offset + required_size > STACK_ALLOC_SIZE)) {
- if (unlikely(depot_index + 1 >= STACK_ALLOC_MAX_SLABS)) {
+ if (unlikely(depot_index + 2 >= STACK_ALLOC_MAX_SLABS)) {
WARN_ONCE(1, "Stack depot reached limit capacity");
return NULL;
}
--
2.18.0
^ permalink raw reply related [flat|nested] 6+ messages in thread
* Re: [PATCH v3] lib/stackdepot: Fix global out-of-bounds in stackdepot
2020-01-30 6:44 [PATCH v3] lib/stackdepot: Fix global out-of-bounds in stackdepot Walter Wu
@ 2020-01-30 12:03 ` Alexander Potapenko
2020-01-31 2:05 ` Walter Wu
0 siblings, 1 reply; 6+ messages in thread
From: Alexander Potapenko @ 2020-01-30 12:03 UTC (permalink / raw)
To: Walter Wu
Cc: Dmitry Vyukov, Matthias Brugger, Thomas Gleixner, Josh Poimboeuf,
Greg Kroah-Hartman, Kate Stewart, LKML, linux-arm-kernel,
linux-mediatek, wsd_upstream
On Thu, Jan 30, 2020 at 7:44 AM Walter Wu <walter-zh.wu@mediatek.com> wrote:
Hi Walter,
> If the depot_index = STACK_ALLOC_MAX_SLABS - 2 and next_slab_inited = 0,
> then it will cause array out-of-bounds access, so that we should modify
> the detection to avoid this array out-of-bounds bug.
>
> Assume depot_index = STACK_ALLOC_MAX_SLABS - 3
> Consider following call flow sequence:
>
> stack_depot_save()
> depot_alloc_stack()
> if (unlikely(depot_index + 1 >= STACK_ALLOC_MAX_SLABS)) //pass
> depot_index++ //depot_index = STACK_ALLOC_MAX_SLABS - 2
> if (depot_index + 1 < STACK_ALLOC_MAX_SLABS) //enter
> smp_store_release(&next_slab_inited, 0); //next_slab_inited = 0
> init_stack_slab()
> if (stack_slabs[depot_index] == NULL) //enter and exit
>
> stack_depot_save()
> depot_alloc_stack()
> if (unlikely(depot_index + 1 >= STACK_ALLOC_MAX_SLABS)) //pass
> depot_index++ //depot_index = STACK_ALLOC_MAX_SLABS - 1
> init_stack_slab(&prealloc)
> stack_slabs[depot_index + 1] //here get global out-of-bounds
>
> Cc: Dmitry Vyukov <dvyukov@google.com>
> Cc: Matthias Brugger <matthias.bgg@gmail.com>
> Cc: Thomas Gleixner <tglx@linutronix.de>
> Cc: Alexander Potapenko <glider@google.com>
> Cc: Josh Poimboeuf <jpoimboe@redhat.com>
> Cc: Kate Stewart <kstewart@linuxfoundation.org>
> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> Cc: Kate Stewart <kstewart@linuxfoundation.org>
> Signed-off-by: Walter Wu <walter-zh.wu@mediatek.com>
> ---
> changes in v2:
> modify call flow sequence and preconditon
>
> changes in v3:
> add some reviewers
> ---
> lib/stackdepot.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/lib/stackdepot.c b/lib/stackdepot.c
> index ed717dd08ff3..7e8a15e41600 100644
> --- a/lib/stackdepot.c
> +++ b/lib/stackdepot.c
> @@ -106,7 +106,7 @@ static struct stack_record *depot_alloc_stack(unsigned long *entries, int size,
> required_size = ALIGN(required_size, 1 << STACK_ALLOC_ALIGN);
>
> if (unlikely(depot_offset + required_size > STACK_ALLOC_SIZE)) {
> - if (unlikely(depot_index + 1 >= STACK_ALLOC_MAX_SLABS)) {
> + if (unlikely(depot_index + 2 >= STACK_ALLOC_MAX_SLABS)) {
I don't think this is the right way to fix the problem.
You're basically throwing away the last element of stack_slabs[], as
we won't allocate anything from it.
How about we set |next_slab_inited| to 1 here:
diff --git a/lib/stackdepot.c b/lib/stackdepot.c
index 2e7d2232ed3c..943a51eb746d 100644
--- a/lib/stackdepot.c
+++ b/lib/stackdepot.c
@@ -105,6 +105,8 @@ static bool init_stack_slab(void **prealloc)
return true;
if (stack_slabs[depot_index] == NULL) {
stack_slabs[depot_index] = *prealloc;
+ if (depot_index + 1 == STACK_ALLOC_MAX_SLABS)
+ smp_store_release(&next_slab_inited, 1);
} else {
stack_slabs[depot_index + 1] = *prealloc;
/*
This will ensure we won't be preallocating pages once |depot_index|
reaches the last element, and we won't attempt to write those pages
anywhere either.
Could you please check if this fixes the problem for you?
> WARN_ONCE(1, "Stack depot reached limit capacity");
> return NULL;
> }
> --
> 2.18.0
--
Alexander Potapenko
Software Engineer
Google Germany GmbH
Erika-Mann-Straße, 33
80636 München
Geschäftsführer: Paul Manicle, Halimah DeLaine Prado
Registergericht und -nummer: Hamburg, HRB 86891
Sitz der Gesellschaft: Hamburg
^ permalink raw reply related [flat|nested] 6+ messages in thread
* Re: [PATCH v3] lib/stackdepot: Fix global out-of-bounds in stackdepot
2020-01-30 12:03 ` Alexander Potapenko
@ 2020-01-31 2:05 ` Walter Wu
2020-01-31 18:11 ` Alexander Potapenko
0 siblings, 1 reply; 6+ messages in thread
From: Walter Wu @ 2020-01-31 2:05 UTC (permalink / raw)
To: Alexander Potapenko
Cc: Dmitry Vyukov, Matthias Brugger, Thomas Gleixner, Josh Poimboeuf,
Greg Kroah-Hartman, Kate Stewart, LKML, linux-arm-kernel,
linux-mediatek, wsd_upstream
On Thu, 2020-01-30 at 13:03 +0100, Alexander Potapenko wrote:
> On Thu, Jan 30, 2020 at 7:44 AM Walter Wu <walter-zh.wu@mediatek.com> wrote:
>
> Hi Walter,
>
> > If the depot_index = STACK_ALLOC_MAX_SLABS - 2 and next_slab_inited = 0,
> > then it will cause array out-of-bounds access, so that we should modify
> > the detection to avoid this array out-of-bounds bug.
> >
> > Assume depot_index = STACK_ALLOC_MAX_SLABS - 3
> > Consider following call flow sequence:
> >
> > stack_depot_save()
> > depot_alloc_stack()
> > if (unlikely(depot_index + 1 >= STACK_ALLOC_MAX_SLABS)) //pass
> > depot_index++ //depot_index = STACK_ALLOC_MAX_SLABS - 2
> > if (depot_index + 1 < STACK_ALLOC_MAX_SLABS) //enter
> > smp_store_release(&next_slab_inited, 0); //next_slab_inited = 0
> > init_stack_slab()
> > if (stack_slabs[depot_index] == NULL) //enter and exit
> >
> > stack_depot_save()
> > depot_alloc_stack()
> > if (unlikely(depot_index + 1 >= STACK_ALLOC_MAX_SLABS)) //pass
> > depot_index++ //depot_index = STACK_ALLOC_MAX_SLABS - 1
> > init_stack_slab(&prealloc)
> > stack_slabs[depot_index + 1] //here get global out-of-bounds
> >
> > Cc: Dmitry Vyukov <dvyukov@google.com>
> > Cc: Matthias Brugger <matthias.bgg@gmail.com>
> > Cc: Thomas Gleixner <tglx@linutronix.de>
> > Cc: Alexander Potapenko <glider@google.com>
> > Cc: Josh Poimboeuf <jpoimboe@redhat.com>
> > Cc: Kate Stewart <kstewart@linuxfoundation.org>
> > Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> > Cc: Kate Stewart <kstewart@linuxfoundation.org>
> > Signed-off-by: Walter Wu <walter-zh.wu@mediatek.com>
> > ---
> > changes in v2:
> > modify call flow sequence and preconditon
> >
> > changes in v3:
> > add some reviewers
> > ---
> > lib/stackdepot.c | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/lib/stackdepot.c b/lib/stackdepot.c
> > index ed717dd08ff3..7e8a15e41600 100644
> > --- a/lib/stackdepot.c
> > +++ b/lib/stackdepot.c
> > @@ -106,7 +106,7 @@ static struct stack_record *depot_alloc_stack(unsigned long *entries, int size,
> > required_size = ALIGN(required_size, 1 << STACK_ALLOC_ALIGN);
> >
> > if (unlikely(depot_offset + required_size > STACK_ALLOC_SIZE)) {
> > - if (unlikely(depot_index + 1 >= STACK_ALLOC_MAX_SLABS)) {
> > + if (unlikely(depot_index + 2 >= STACK_ALLOC_MAX_SLABS)) {
>
> I don't think this is the right way to fix the problem.
> You're basically throwing away the last element of stack_slabs[], as
> we won't allocate anything from it.
>
ok, I agree.
> How about we set |next_slab_inited| to 1 here:
>
> diff --git a/lib/stackdepot.c b/lib/stackdepot.c
> index 2e7d2232ed3c..943a51eb746d 100644
> --- a/lib/stackdepot.c
> +++ b/lib/stackdepot.c
> @@ -105,6 +105,8 @@ static bool init_stack_slab(void **prealloc)
> return true;
> if (stack_slabs[depot_index] == NULL) {
> stack_slabs[depot_index] = *prealloc;
> + if (depot_index + 1 == STACK_ALLOC_MAX_SLABS)
> + smp_store_release(&next_slab_inited, 1);
> } else {
> stack_slabs[depot_index + 1] = *prealloc;
> /*
>
> This will ensure we won't be preallocating pages once |depot_index|
> reaches the last element, and we won't attempt to write those pages
> anywhere either.
>
> Could you please check if this fixes the problem for you?
>
Consider above call flow sequence at first stack_depot_save(),
depot_index = STACK_ALLOC_MAX_SLABS - 2 before enter init_stack_slab(),
so the fixes should be below.
--- a/lib/stackdepot.c
+++ b/lib/stackdepot.c
@@ -83,6 +83,8 @@ static bool init_stack_slab(void **prealloc)
return true;
if (stack_slabs[depot_index] == NULL) {
stack_slabs[depot_index] = *prealloc;
+ if (depot_index + 2 == STACK_ALLOC_MAX_SLABS)
+ smp_store_release(&next_slab_inited, 1);
} else {
stack_slabs[depot_index + 1] = *prealloc;
/*
> > WARN_ONCE(1, "Stack depot reached limit capacity");
> > return NULL;
> > }
> > --
> > 2.18.0
>
>
>
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH v3] lib/stackdepot: Fix global out-of-bounds in stackdepot
2020-01-31 2:05 ` Walter Wu
@ 2020-01-31 18:11 ` Alexander Potapenko
2020-02-03 2:05 ` Walter Wu
0 siblings, 1 reply; 6+ messages in thread
From: Alexander Potapenko @ 2020-01-31 18:11 UTC (permalink / raw)
To: Walter Wu
Cc: Dmitry Vyukov, Matthias Brugger, Thomas Gleixner, Josh Poimboeuf,
Greg Kroah-Hartman, Kate Stewart, LKML, linux-arm-kernel,
linux-mediatek, wsd_upstream
On Fri, Jan 31, 2020 at 3:05 AM Walter Wu <walter-zh.wu@mediatek.com> wrote:
>
> On Thu, 2020-01-30 at 13:03 +0100, Alexander Potapenko wrote:
> > On Thu, Jan 30, 2020 at 7:44 AM Walter Wu <walter-zh.wu@mediatek.com> wrote:
> >
> > Hi Walter,
> >
> > > If the depot_index = STACK_ALLOC_MAX_SLABS - 2 and next_slab_inited = 0,
> > > then it will cause array out-of-bounds access, so that we should modify
> > > the detection to avoid this array out-of-bounds bug.
> > >
> > > Assume depot_index = STACK_ALLOC_MAX_SLABS - 3
> > > Consider following call flow sequence:
> > >
> > > stack_depot_save()
> > > depot_alloc_stack()
> > > if (unlikely(depot_index + 1 >= STACK_ALLOC_MAX_SLABS)) //pass
> > > depot_index++ //depot_index = STACK_ALLOC_MAX_SLABS - 2
> > > if (depot_index + 1 < STACK_ALLOC_MAX_SLABS) //enter
> > > smp_store_release(&next_slab_inited, 0); //next_slab_inited = 0
> > > init_stack_slab()
> > > if (stack_slabs[depot_index] == NULL) //enter and exit
> > >
> > > stack_depot_save()
> > > depot_alloc_stack()
> > > if (unlikely(depot_index + 1 >= STACK_ALLOC_MAX_SLABS)) //pass
> > > depot_index++ //depot_index = STACK_ALLOC_MAX_SLABS - 1
> > > init_stack_slab(&prealloc)
> > > stack_slabs[depot_index + 1] //here get global out-of-bounds
> > >
> > > Cc: Dmitry Vyukov <dvyukov@google.com>
> > > Cc: Matthias Brugger <matthias.bgg@gmail.com>
> > > Cc: Thomas Gleixner <tglx@linutronix.de>
> > > Cc: Alexander Potapenko <glider@google.com>
> > > Cc: Josh Poimboeuf <jpoimboe@redhat.com>
> > > Cc: Kate Stewart <kstewart@linuxfoundation.org>
> > > Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> > > Cc: Kate Stewart <kstewart@linuxfoundation.org>
> > > Signed-off-by: Walter Wu <walter-zh.wu@mediatek.com>
> > > ---
> > > changes in v2:
> > > modify call flow sequence and preconditon
> > >
> > > changes in v3:
> > > add some reviewers
> > > ---
> > > lib/stackdepot.c | 2 +-
> > > 1 file changed, 1 insertion(+), 1 deletion(-)
> > >
> > > diff --git a/lib/stackdepot.c b/lib/stackdepot.c
> > > index ed717dd08ff3..7e8a15e41600 100644
> > > --- a/lib/stackdepot.c
> > > +++ b/lib/stackdepot.c
> > > @@ -106,7 +106,7 @@ static struct stack_record *depot_alloc_stack(unsigned long *entries, int size,
> > > required_size = ALIGN(required_size, 1 << STACK_ALLOC_ALIGN);
> > >
> > > if (unlikely(depot_offset + required_size > STACK_ALLOC_SIZE)) {
> > > - if (unlikely(depot_index + 1 >= STACK_ALLOC_MAX_SLABS)) {
> > > + if (unlikely(depot_index + 2 >= STACK_ALLOC_MAX_SLABS)) {
This again means stack_slabs[STACK_ALLOC_MAX_SLABS - 2] gets
initialized, but stack_slabs[STACK_ALLOC_MAX_SLABS - 1] doesn't,
because we'll be bailing out from init_stack_slab() from now on.
Does this patch actually fix the problem (do you have a reliable reproducer?)
This addition of 2 is also counterintuitive, I don't think further
readers will understand the logic behind it.
What if we just check that depot_index + 1 is a valid index before accessing it?
diff --git a/lib/stackdepot.c b/lib/stackdepot.c
index 2e7d2232ed3c..c2e6ff18d716 100644
--- a/lib/stackdepot.c
+++ b/lib/stackdepot.c
@@ -106,7 +106,9 @@ static bool init_stack_slab(void **prealloc)
if (stack_slabs[depot_index] == NULL) {
stack_slabs[depot_index] = *prealloc;
} else {
- stack_slabs[depot_index + 1] = *prealloc;
+ /* If this is the last depot slab, do not touch the next one. */
+ if (depot_index + 1 < STACK_ALLOC_MAX_SLABS)
+ stack_slabs[depot_index + 1] = *prealloc;
/*
* This smp_store_release pairs with smp_load_acquire() from
* |next_slab_inited| above and in stack_depot_save().
^ permalink raw reply related [flat|nested] 6+ messages in thread
* Re: [PATCH v3] lib/stackdepot: Fix global out-of-bounds in stackdepot
2020-01-31 18:11 ` Alexander Potapenko
@ 2020-02-03 2:05 ` Walter Wu
2020-02-03 10:30 ` Alexander Potapenko
0 siblings, 1 reply; 6+ messages in thread
From: Walter Wu @ 2020-02-03 2:05 UTC (permalink / raw)
To: Alexander Potapenko
Cc: Dmitry Vyukov, Matthias Brugger, Thomas Gleixner, Josh Poimboeuf,
Greg Kroah-Hartman, Kate Stewart, LKML, linux-arm-kernel,
linux-mediatek, wsd_upstream
On Fri, 2020-01-31 at 19:11 +0100, Alexander Potapenko wrote:
> On Fri, Jan 31, 2020 at 3:05 AM Walter Wu <walter-zh.wu@mediatek.com> wrote:
> >
> > On Thu, 2020-01-30 at 13:03 +0100, Alexander Potapenko wrote:
> > > On Thu, Jan 30, 2020 at 7:44 AM Walter Wu <walter-zh.wu@mediatek.com> wrote:
> > >
> > > Hi Walter,
> > >
> > > > If the depot_index = STACK_ALLOC_MAX_SLABS - 2 and next_slab_inited = 0,
> > > > then it will cause array out-of-bounds access, so that we should modify
> > > > the detection to avoid this array out-of-bounds bug.
> > > >
> > > > Assume depot_index = STACK_ALLOC_MAX_SLABS - 3
> > > > Consider following call flow sequence:
> > > >
> > > > stack_depot_save()
> > > > depot_alloc_stack()
> > > > if (unlikely(depot_index + 1 >= STACK_ALLOC_MAX_SLABS)) //pass
> > > > depot_index++ //depot_index = STACK_ALLOC_MAX_SLABS - 2
> > > > if (depot_index + 1 < STACK_ALLOC_MAX_SLABS) //enter
> > > > smp_store_release(&next_slab_inited, 0); //next_slab_inited = 0
> > > > init_stack_slab()
> > > > if (stack_slabs[depot_index] == NULL) //enter and exit
> > > >
> > > > stack_depot_save()
> > > > depot_alloc_stack()
> > > > if (unlikely(depot_index + 1 >= STACK_ALLOC_MAX_SLABS)) //pass
> > > > depot_index++ //depot_index = STACK_ALLOC_MAX_SLABS - 1
> > > > init_stack_slab(&prealloc)
> > > > stack_slabs[depot_index + 1] //here get global out-of-bounds
> > > >
> > > > Cc: Dmitry Vyukov <dvyukov@google.com>
> > > > Cc: Matthias Brugger <matthias.bgg@gmail.com>
> > > > Cc: Thomas Gleixner <tglx@linutronix.de>
> > > > Cc: Alexander Potapenko <glider@google.com>
> > > > Cc: Josh Poimboeuf <jpoimboe@redhat.com>
> > > > Cc: Kate Stewart <kstewart@linuxfoundation.org>
> > > > Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> > > > Cc: Kate Stewart <kstewart@linuxfoundation.org>
> > > > Signed-off-by: Walter Wu <walter-zh.wu@mediatek.com>
> > > > ---
> > > > changes in v2:
> > > > modify call flow sequence and preconditon
> > > >
> > > > changes in v3:
> > > > add some reviewers
> > > > ---
> > > > lib/stackdepot.c | 2 +-
> > > > 1 file changed, 1 insertion(+), 1 deletion(-)
> > > >
> > > > diff --git a/lib/stackdepot.c b/lib/stackdepot.c
> > > > index ed717dd08ff3..7e8a15e41600 100644
> > > > --- a/lib/stackdepot.c
> > > > +++ b/lib/stackdepot.c
> > > > @@ -106,7 +106,7 @@ static struct stack_record *depot_alloc_stack(unsigned long *entries, int size,
> > > > required_size = ALIGN(required_size, 1 << STACK_ALLOC_ALIGN);
> > > >
> > > > if (unlikely(depot_offset + required_size > STACK_ALLOC_SIZE)) {
> > > > - if (unlikely(depot_index + 1 >= STACK_ALLOC_MAX_SLABS)) {
> > > > + if (unlikely(depot_index + 2 >= STACK_ALLOC_MAX_SLABS)) {
>
> This again means stack_slabs[STACK_ALLOC_MAX_SLABS - 2] gets
> initialized, but stack_slabs[STACK_ALLOC_MAX_SLABS - 1] doesn't,
> because we'll be bailing out from init_stack_slab() from now on.
> Does this patch actually fix the problem (do you have a reliable reproducer?)
We get it by reviewing code, because Kasan doesn't scan it and we catch
another bug internally, we found it unintentionally.
> This addition of 2 is also counterintuitive, I don't think further
> readers will understand the logic behind it.
>
Yes
> What if we just check that depot_index + 1 is a valid index before accessing it?
>
It should fix the problem, do you want to send this patch?
> diff --git a/lib/stackdepot.c b/lib/stackdepot.c
> index 2e7d2232ed3c..c2e6ff18d716 100644
> --- a/lib/stackdepot.c
> +++ b/lib/stackdepot.c
> @@ -106,7 +106,9 @@ static bool init_stack_slab(void **prealloc)
> if (stack_slabs[depot_index] == NULL) {
> stack_slabs[depot_index] = *prealloc;
> } else {
> - stack_slabs[depot_index + 1] = *prealloc;
> + /* If this is the last depot slab, do not touch the next one. */
> + if (depot_index + 1 < STACK_ALLOC_MAX_SLABS)
> + stack_slabs[depot_index + 1] = *prealloc;
> /*
> * This smp_store_release pairs with smp_load_acquire() from
> * |next_slab_inited| above and in stack_depot_save().
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH v3] lib/stackdepot: Fix global out-of-bounds in stackdepot
2020-02-03 2:05 ` Walter Wu
@ 2020-02-03 10:30 ` Alexander Potapenko
0 siblings, 0 replies; 6+ messages in thread
From: Alexander Potapenko @ 2020-02-03 10:30 UTC (permalink / raw)
To: Walter Wu
Cc: Dmitry Vyukov, Matthias Brugger, Thomas Gleixner, Josh Poimboeuf,
Greg Kroah-Hartman, Kate Stewart, LKML, linux-arm-kernel,
linux-mediatek, wsd_upstream
On Mon, Feb 3, 2020 at 3:05 AM Walter Wu <walter-zh.wu@mediatek.com> wrote:
>
> On Fri, 2020-01-31 at 19:11 +0100, Alexander Potapenko wrote:
> > On Fri, Jan 31, 2020 at 3:05 AM Walter Wu <walter-zh.wu@mediatek.com> wrote:
> > >
> > > On Thu, 2020-01-30 at 13:03 +0100, Alexander Potapenko wrote:
> > > > On Thu, Jan 30, 2020 at 7:44 AM Walter Wu <walter-zh.wu@mediatek.com> wrote:
> > > >
> > > > Hi Walter,
> > > >
> > > > > If the depot_index = STACK_ALLOC_MAX_SLABS - 2 and next_slab_inited = 0,
> > > > > then it will cause array out-of-bounds access, so that we should modify
> > > > > the detection to avoid this array out-of-bounds bug.
> > > > >
> > > > > Assume depot_index = STACK_ALLOC_MAX_SLABS - 3
> > > > > Consider following call flow sequence:
> > > > >
> > > > > stack_depot_save()
> > > > > depot_alloc_stack()
> > > > > if (unlikely(depot_index + 1 >= STACK_ALLOC_MAX_SLABS)) //pass
> > > > > depot_index++ //depot_index = STACK_ALLOC_MAX_SLABS - 2
> > > > > if (depot_index + 1 < STACK_ALLOC_MAX_SLABS) //enter
> > > > > smp_store_release(&next_slab_inited, 0); //next_slab_inited = 0
> > > > > init_stack_slab()
> > > > > if (stack_slabs[depot_index] == NULL) //enter and exit
> > > > >
> > > > > stack_depot_save()
> > > > > depot_alloc_stack()
> > > > > if (unlikely(depot_index + 1 >= STACK_ALLOC_MAX_SLABS)) //pass
> > > > > depot_index++ //depot_index = STACK_ALLOC_MAX_SLABS - 1
> > > > > init_stack_slab(&prealloc)
> > > > > stack_slabs[depot_index + 1] //here get global out-of-bounds
> > > > >
> > > > > Cc: Dmitry Vyukov <dvyukov@google.com>
> > > > > Cc: Matthias Brugger <matthias.bgg@gmail.com>
> > > > > Cc: Thomas Gleixner <tglx@linutronix.de>
> > > > > Cc: Alexander Potapenko <glider@google.com>
> > > > > Cc: Josh Poimboeuf <jpoimboe@redhat.com>
> > > > > Cc: Kate Stewart <kstewart@linuxfoundation.org>
> > > > > Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> > > > > Cc: Kate Stewart <kstewart@linuxfoundation.org>
> > > > > Signed-off-by: Walter Wu <walter-zh.wu@mediatek.com>
> > > > > ---
> > > > > changes in v2:
> > > > > modify call flow sequence and preconditon
> > > > >
> > > > > changes in v3:
> > > > > add some reviewers
> > > > > ---
> > > > > lib/stackdepot.c | 2 +-
> > > > > 1 file changed, 1 insertion(+), 1 deletion(-)
> > > > >
> > > > > diff --git a/lib/stackdepot.c b/lib/stackdepot.c
> > > > > index ed717dd08ff3..7e8a15e41600 100644
> > > > > --- a/lib/stackdepot.c
> > > > > +++ b/lib/stackdepot.c
> > > > > @@ -106,7 +106,7 @@ static struct stack_record *depot_alloc_stack(unsigned long *entries, int size,
> > > > > required_size = ALIGN(required_size, 1 << STACK_ALLOC_ALIGN);
> > > > >
> > > > > if (unlikely(depot_offset + required_size > STACK_ALLOC_SIZE)) {
> > > > > - if (unlikely(depot_index + 1 >= STACK_ALLOC_MAX_SLABS)) {
> > > > > + if (unlikely(depot_index + 2 >= STACK_ALLOC_MAX_SLABS)) {
> >
> > This again means stack_slabs[STACK_ALLOC_MAX_SLABS - 2] gets
> > initialized, but stack_slabs[STACK_ALLOC_MAX_SLABS - 1] doesn't,
> > because we'll be bailing out from init_stack_slab() from now on.
> > Does this patch actually fix the problem (do you have a reliable reproducer?)
> We get it by reviewing code, because Kasan doesn't scan it and we catch
> another bug internally, we found it unintentionally.
>
> > This addition of 2 is also counterintuitive, I don't think further
> > readers will understand the logic behind it.
> >
> Yes
>
> > What if we just check that depot_index + 1 is a valid index before accessing it?
> >
> It should fix the problem, do you want to send this patch?
I've sent the patch. Thanks for the report!
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2020-02-03 10:30 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-01-30 6:44 [PATCH v3] lib/stackdepot: Fix global out-of-bounds in stackdepot Walter Wu
2020-01-30 12:03 ` Alexander Potapenko
2020-01-31 2:05 ` Walter Wu
2020-01-31 18:11 ` Alexander Potapenko
2020-02-03 2:05 ` Walter Wu
2020-02-03 10:30 ` Alexander Potapenko
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).