[SRU Z/Y/T: CVE-2017-1000365] fs/exec.c: account for argv/envp pointers

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
9 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

[SRU Z/Y/T: CVE-2017-1000365] fs/exec.c: account for argv/envp pointers

Stefan Bader-2
From: Kees Cook <[hidden email]>

When limiting the argv/envp strings during exec to 1/4 of the stack limit,
the storage of the pointers to the strings was not included.  This means
that an exec with huge numbers of tiny strings could eat 1/4 of the stack
limit in strings and then additional space would be later used by the
pointers to the strings.

For example, on 32-bit with a 8MB stack rlimit, an exec with 1677721
single-byte strings would consume less than 2MB of stack, the max (8MB /
4) amount allowed, but the pointers to the strings would consume the
remaining additional stack space (1677721 * 4 == 6710884).

The result (1677721 + 6710884 == 8388605) would exhaust stack space
entirely.  Controlling this stack exhaustion could result in
pathological behavior in setuid binaries (CVE-2017-1000365).

[[hidden email]: additional commenting from Kees]
Fixes: b6a2fea39318 ("mm: variable length argument support")
Link: http://lkml.kernel.org/r/20170622001720.GA32173@beast
Signed-off-by: Kees Cook <[hidden email]>
Acked-by: Rik van Riel <[hidden email]>
Acked-by: Michal Hocko <[hidden email]>
Cc: Alexander Viro <[hidden email]>
Cc: Qualys Security Advisory <[hidden email]>
Cc: <[hidden email]>
Signed-off-by: Andrew Morton <[hidden email]>
Signed-off-by: Linus Torvalds <[hidden email]>

CVE-2017-1000365

(cherry-picked from commit 98da7d08850fb8bdeb395d6368ed15753304aa0c)
Signed-off-by: Stefan Bader <[hidden email]>
---

Fix already applied to Xenial via upstream stable updates. Can be
cherry-picked into Z/Y/T.

-Stefan

 fs/exec.c | 28 ++++++++++++++++++++++++----
 1 file changed, 24 insertions(+), 4 deletions(-)

diff --git a/fs/exec.c b/fs/exec.c
index 72934df..9041990 100644
--- a/fs/exec.c
+++ b/fs/exec.c
@@ -220,8 +220,26 @@ static struct page *get_arg_page(struct linux_binprm *bprm, unsigned long pos,
 
  if (write) {
  unsigned long size = bprm->vma->vm_end - bprm->vma->vm_start;
+ unsigned long ptr_size;
  struct rlimit *rlim;
 
+ /*
+ * Since the stack will hold pointers to the strings, we
+ * must account for them as well.
+ *
+ * The size calculation is the entire vma while each arg page is
+ * built, so each time we get here it's calculating how far it
+ * is currently (rather than each call being just the newly
+ * added size from the arg page).  As a result, we need to
+ * always add the entire size of the pointers, so that on the
+ * last call to get_arg_page() we'll actually have the entire
+ * correct size.
+ */
+ ptr_size = (bprm->argc + bprm->envc) * sizeof(void *);
+ if (ptr_size > ULONG_MAX - size)
+ goto fail;
+ size += ptr_size;
+
  acct_arg_size(bprm, size / PAGE_SIZE);
 
  /*
@@ -239,13 +257,15 @@ static struct page *get_arg_page(struct linux_binprm *bprm, unsigned long pos,
  *    to work from.
  */
  rlim = current->signal->rlim;
- if (size > ACCESS_ONCE(rlim[RLIMIT_STACK].rlim_cur) / 4) {
- put_page(page);
- return NULL;
- }
+ if (size > READ_ONCE(rlim[RLIMIT_STACK].rlim_cur) / 4)
+ goto fail;
  }
 
  return page;
+
+fail:
+ put_page(page);
+ return NULL;
 }
 
 static void put_arg_page(struct page *page)
--
2.7.4


--
kernel-team mailing list
[hidden email]
https://lists.ubuntu.com/mailman/listinfo/kernel-team
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

ACK: [SRU Z/Y/T: CVE-2017-1000365] fs/exec.c: account for argv/envp pointers

Thadeu Lima de Souza Cascardo-3
Acked-by: Thadeu Lima de Souza Cascardo <[hidden email]>

--
kernel-team mailing list
[hidden email]
https://lists.ubuntu.com/mailman/listinfo/kernel-team
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

NAK: [SRU T: CVE-2017-1000365] fs/exec.c: account for argv/envp pointers

Stefan Bader-2
In reply to this post by Stefan Bader-2
On 12.07.2017 15:30, Stefan Bader wrote:

> From: Kees Cook <[hidden email]>
>
> When limiting the argv/envp strings during exec to 1/4 of the stack limit,
> the storage of the pointers to the strings was not included.  This means
> that an exec with huge numbers of tiny strings could eat 1/4 of the stack
> limit in strings and then additional space would be later used by the
> pointers to the strings.
>
> For example, on 32-bit with a 8MB stack rlimit, an exec with 1677721
> single-byte strings would consume less than 2MB of stack, the max (8MB /
> 4) amount allowed, but the pointers to the strings would consume the
> remaining additional stack space (1677721 * 4 == 6710884).
>
> The result (1677721 + 6710884 == 8388605) would exhaust stack space
> entirely.  Controlling this stack exhaustion could result in
> pathological behavior in setuid binaries (CVE-2017-1000365).
>
> [[hidden email]: additional commenting from Kees]
> Fixes: b6a2fea39318 ("mm: variable length argument support")
> Link: http://lkml.kernel.org/r/20170622001720.GA32173@beast
> Signed-off-by: Kees Cook <[hidden email]>
> Acked-by: Rik van Riel <[hidden email]>
> Acked-by: Michal Hocko <[hidden email]>
> Cc: Alexander Viro <[hidden email]>
> Cc: Qualys Security Advisory <[hidden email]>
> Cc: <[hidden email]>
> Signed-off-by: Andrew Morton <[hidden email]>
> Signed-off-by: Linus Torvalds <[hidden email]>
>
> CVE-2017-1000365
>
> (cherry-picked from commit 98da7d08850fb8bdeb395d6368ed15753304aa0c)
> Signed-off-by: Stefan Bader <[hidden email]>
> ---
>
> Fix already applied to Xenial via upstream stable updates. Can be
> cherry-picked into Z/Y/T.
>
> -Stefan
>
>  fs/exec.c | 28 ++++++++++++++++++++++++----
>  1 file changed, 24 insertions(+), 4 deletions(-)
>
> diff --git a/fs/exec.c b/fs/exec.c
> index 72934df..9041990 100644
> --- a/fs/exec.c
> +++ b/fs/exec.c
> @@ -220,8 +220,26 @@ static struct page *get_arg_page(struct linux_binprm *bprm, unsigned long pos,
>  
>   if (write) {
>   unsigned long size = bprm->vma->vm_end - bprm->vma->vm_start;
> + unsigned long ptr_size;
>   struct rlimit *rlim;
>  
> + /*
> + * Since the stack will hold pointers to the strings, we
> + * must account for them as well.
> + *
> + * The size calculation is the entire vma while each arg page is
> + * built, so each time we get here it's calculating how far it
> + * is currently (rather than each call being just the newly
> + * added size from the arg page).  As a result, we need to
> + * always add the entire size of the pointers, so that on the
> + * last call to get_arg_page() we'll actually have the entire
> + * correct size.
> + */
> + ptr_size = (bprm->argc + bprm->envc) * sizeof(void *);
> + if (ptr_size > ULONG_MAX - size)
> + goto fail;
> + size += ptr_size;
> +
>   acct_arg_size(bprm, size / PAGE_SIZE);
>  
>   /*
> @@ -239,13 +257,15 @@ static struct page *get_arg_page(struct linux_binprm *bprm, unsigned long pos,
>   *    to work from.
>   */
>   rlim = current->signal->rlim;
> - if (size > ACCESS_ONCE(rlim[RLIMIT_STACK].rlim_cur) / 4) {
> - put_page(page);
> - return NULL;
> - }
> + if (size > READ_ONCE(rlim[RLIMIT_STACK].rlim_cur) / 4)
                           ^
Should have compiled first. We need to keep this as ACCESS_ONCE in Trusty. Will
send a v2 for Trusty.

-Stefan

> + goto fail;
>   }
>  
>   return page;
> +
> +fail:
> + put_page(page);
> + return NULL;
>  }
>  
>  static void put_arg_page(struct page *page)
>


--
kernel-team mailing list
[hidden email]
https://lists.ubuntu.com/mailman/listinfo/kernel-team

signature.asc (836 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

[SRU T: CVE-2017-1000365 v2] fs/exec.c: account for argv/envp pointers

Stefan Bader-2
In reply to this post by Stefan Bader-2
From: Kees Cook <[hidden email]>

When limiting the argv/envp strings during exec to 1/4 of the stack limit,
the storage of the pointers to the strings was not included.  This means
that an exec with huge numbers of tiny strings could eat 1/4 of the stack
limit in strings and then additional space would be later used by the
pointers to the strings.

For example, on 32-bit with a 8MB stack rlimit, an exec with 1677721
single-byte strings would consume less than 2MB of stack, the max (8MB /
4) amount allowed, but the pointers to the strings would consume the
remaining additional stack space (1677721 * 4 == 6710884).

The result (1677721 + 6710884 == 8388605) would exhaust stack space
entirely.  Controlling this stack exhaustion could result in
pathological behavior in setuid binaries (CVE-2017-1000365).

[[hidden email]: additional commenting from Kees]
Fixes: b6a2fea39318 ("mm: variable length argument support")
Link: http://lkml.kernel.org/r/20170622001720.GA32173@beast
Signed-off-by: Kees Cook <[hidden email]>
Acked-by: Rik van Riel <[hidden email]>
Acked-by: Michal Hocko <[hidden email]>
Cc: Alexander Viro <[hidden email]>
Cc: Qualys Security Advisory <[hidden email]>
Cc: <[hidden email]>
Signed-off-by: Andrew Morton <[hidden email]>
Signed-off-by: Linus Torvalds <[hidden email]>

CVE-2017-1000365

(backported from commit 98da7d08850fb8bdeb395d6368ed15753304aa0c)
Signed-off-by: Stefan Bader <[hidden email]>
---

This is the updated version which does not change the ACCESS_ONCE
into a READ_ONCE which does not exist for Trusty.

-Stefan

 fs/exec.c | 28 ++++++++++++++++++++++++----
 1 file changed, 24 insertions(+), 4 deletions(-)

diff --git a/fs/exec.c b/fs/exec.c
index c031add..dd7ab64 100644
--- a/fs/exec.c
+++ b/fs/exec.c
@@ -206,8 +206,26 @@ static struct page *get_arg_page(struct linux_binprm *bprm, unsigned long pos,
 
  if (write) {
  unsigned long size = bprm->vma->vm_end - bprm->vma->vm_start;
+ unsigned long ptr_size;
  struct rlimit *rlim;
 
+ /*
+ * Since the stack will hold pointers to the strings, we
+ * must account for them as well.
+ *
+ * The size calculation is the entire vma while each arg page is
+ * built, so each time we get here it's calculating how far it
+ * is currently (rather than each call being just the newly
+ * added size from the arg page).  As a result, we need to
+ * always add the entire size of the pointers, so that on the
+ * last call to get_arg_page() we'll actually have the entire
+ * correct size.
+ */
+ ptr_size = (bprm->argc + bprm->envc) * sizeof(void *);
+ if (ptr_size > ULONG_MAX - size)
+ goto fail;
+ size += ptr_size;
+
  acct_arg_size(bprm, size / PAGE_SIZE);
 
  /*
@@ -225,13 +243,15 @@ static struct page *get_arg_page(struct linux_binprm *bprm, unsigned long pos,
  *    to work from.
  */
  rlim = current->signal->rlim;
- if (size > ACCESS_ONCE(rlim[RLIMIT_STACK].rlim_cur) / 4) {
- put_page(page);
- return NULL;
- }
+ if (size > ACCESS_ONCE(rlim[RLIMIT_STACK].rlim_cur) / 4)
+ goto fail;
  }
 
  return page;
+
+fail:
+ put_page(page);
+ return NULL;
 }
 
 static void put_arg_page(struct page *page)
--
2.7.4


--
kernel-team mailing list
[hidden email]
https://lists.ubuntu.com/mailman/listinfo/kernel-team
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

ACK: [SRU T: CVE-2017-1000365 v2] fs/exec.c: account for argv/envp pointers

Thadeu Lima de Souza Cascardo-3
Acked-by: Thadeu Lima de Souza Cascardo <[hidden email]>

--
kernel-team mailing list
[hidden email]
https://lists.ubuntu.com/mailman/listinfo/kernel-team
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

ACK: [SRU Z/Y/T: CVE-2017-1000365] fs/exec.c: account for argv/envp pointers

Seth Forshee
In reply to this post by Stefan Bader-2
On Wed, Jul 12, 2017 at 03:30:39PM +0200, Stefan Bader wrote:

> From: Kees Cook <[hidden email]>
>
> When limiting the argv/envp strings during exec to 1/4 of the stack limit,
> the storage of the pointers to the strings was not included.  This means
> that an exec with huge numbers of tiny strings could eat 1/4 of the stack
> limit in strings and then additional space would be later used by the
> pointers to the strings.
>
> For example, on 32-bit with a 8MB stack rlimit, an exec with 1677721
> single-byte strings would consume less than 2MB of stack, the max (8MB /
> 4) amount allowed, but the pointers to the strings would consume the
> remaining additional stack space (1677721 * 4 == 6710884).
>
> The result (1677721 + 6710884 == 8388605) would exhaust stack space
> entirely.  Controlling this stack exhaustion could result in
> pathological behavior in setuid binaries (CVE-2017-1000365).
>
> [[hidden email]: additional commenting from Kees]
> Fixes: b6a2fea39318 ("mm: variable length argument support")
> Link: http://lkml.kernel.org/r/20170622001720.GA32173@beast
> Signed-off-by: Kees Cook <[hidden email]>
> Acked-by: Rik van Riel <[hidden email]>
> Acked-by: Michal Hocko <[hidden email]>
> Cc: Alexander Viro <[hidden email]>
> Cc: Qualys Security Advisory <[hidden email]>
> Cc: <[hidden email]>
> Signed-off-by: Andrew Morton <[hidden email]>
> Signed-off-by: Linus Torvalds <[hidden email]>
>
> CVE-2017-1000365
>
> (cherry-picked from commit 98da7d08850fb8bdeb395d6368ed15753304aa0c)
> Signed-off-by: Stefan Bader <[hidden email]>

For Y/Z:

Acked-by: Seth Forshee <[hidden email]>

Artful already got this commit from upstream stable.

--
kernel-team mailing list
[hidden email]
https://lists.ubuntu.com/mailman/listinfo/kernel-team
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

ACK: [SRU T: CVE-2017-1000365 v2] fs/exec.c: account for argv/envp pointers

Seth Forshee
In reply to this post by Stefan Bader-2
On Wed, Jul 12, 2017 at 04:04:35PM +0200, Stefan Bader wrote:

> From: Kees Cook <[hidden email]>
>
> When limiting the argv/envp strings during exec to 1/4 of the stack limit,
> the storage of the pointers to the strings was not included.  This means
> that an exec with huge numbers of tiny strings could eat 1/4 of the stack
> limit in strings and then additional space would be later used by the
> pointers to the strings.
>
> For example, on 32-bit with a 8MB stack rlimit, an exec with 1677721
> single-byte strings would consume less than 2MB of stack, the max (8MB /
> 4) amount allowed, but the pointers to the strings would consume the
> remaining additional stack space (1677721 * 4 == 6710884).
>
> The result (1677721 + 6710884 == 8388605) would exhaust stack space
> entirely.  Controlling this stack exhaustion could result in
> pathological behavior in setuid binaries (CVE-2017-1000365).
>
> [[hidden email]: additional commenting from Kees]
> Fixes: b6a2fea39318 ("mm: variable length argument support")
> Link: http://lkml.kernel.org/r/20170622001720.GA32173@beast
> Signed-off-by: Kees Cook <[hidden email]>
> Acked-by: Rik van Riel <[hidden email]>
> Acked-by: Michal Hocko <[hidden email]>
> Cc: Alexander Viro <[hidden email]>
> Cc: Qualys Security Advisory <[hidden email]>
> Cc: <[hidden email]>
> Signed-off-by: Andrew Morton <[hidden email]>
> Signed-off-by: Linus Torvalds <[hidden email]>
>
> CVE-2017-1000365
>
> (backported from commit 98da7d08850fb8bdeb395d6368ed15753304aa0c)
> Signed-off-by: Stefan Bader <[hidden email]>

Acked-by: Seth Forshee <[hidden email]>

--
kernel-team mailing list
[hidden email]
https://lists.ubuntu.com/mailman/listinfo/kernel-team
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

APPLIED: [SRU Z/Y/T: CVE-2017-1000365] fs/exec.c: account for argv/envp pointers

Thadeu Lima de Souza Cascardo-3
In reply to this post by Stefan Bader-2
Applied to yakkety and zesty master-next branches.

Thanks.
Cascardo.

--
kernel-team mailing list
[hidden email]
https://lists.ubuntu.com/mailman/listinfo/kernel-team
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

APPLIED: [SRU T: CVE-2017-1000365 v2] fs/exec.c: account for argv/envp pointers

Thadeu Lima de Souza Cascardo-3
In reply to this post by Stefan Bader-2
Applied to trusty master-next branch.

Thanks.
Cascardo.

--
kernel-team mailing list
[hidden email]
https://lists.ubuntu.com/mailman/listinfo/kernel-team
Loading...