[PATCH v4] exec: set map length to zero when returning NULL

[PATCH v4] exec: set map length to zero when returning NULL

[P J P]

From: Prasad J Pandit <pjp@fedoraproject.org>

When mapping physical memory into host's virtual address space,
'address_space_map' may return NULL if BounceBuffer is in_use.
Set and return '*plen = 0' to avoid later NULL pointer dereference.

Reported-by: Alexander Bulekov <alxndr@bu.edu>
Fixes: https://bugs.launchpad.net/qemu/+bug/1878259
Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Suggested-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Prasad J Pandit <pjp@fedoraproject.org>
---
 exec.c                | 1 +
 include/exec/memory.h | 3 ++-
 2 files changed, 3 insertions(+), 1 deletion(-)

Update v4: update address_space_map API doc text
  -> https://lists.gnu.org/archive/html/qemu-devel/2020-05/msg07281.html

diff --git a/exec.c b/exec.c
index 5162f0d12f..4eea84bf66 100644
--- a/exec.c
+++ b/exec.c
@@ -3538,6 +3538,7 @@ void *address_space_map(AddressSpace *as,
 
     if (!memory_access_is_direct(mr, is_write)) {
         if (atomic_xchg(&bounce.in_use, true)) {
+            *plen = 0;
             return NULL;
         }
         /* Avoid unbounded allocations */
diff --git a/include/exec/memory.h b/include/exec/memory.h
index e000bd2f97..8fa2d3cab2 100644
--- a/include/exec/memory.h
+++ b/include/exec/memory.h
@@ -2303,7 +2303,8 @@ bool address_space_access_valid(AddressSpace *as, hwaddr addr, hwaddr len,
 /* address_space_map: map a physical memory region into a host virtual address
  *
  * May map a subset of the requested range, given by and returned in @plen.
- * May return %NULL if resources needed to perform the mapping are exhausted.
+ * May return %NULL and set *@plen to zero(0), if resources needed to perform
+ * the mapping are exhausted.
  * Use only for reads OR writes - not for read-modify-write operations.
  * Use cpu_register_map_client() to know when retrying the map operation is
  * likely to succeed.
-- 
2.26.2

Re: [PATCH v4] exec: set map length to zero when returning NULL

[Philippe Mathieu-Daudé]

On 5/26/20 1:17 PM, P J P wrote:
> From: Prasad J Pandit <pjp@fedoraproject.org>
> 
> When mapping physical memory into host's virtual address space,
> 'address_space_map' may return NULL if BounceBuffer is in_use.
> Set and return '*plen = 0' to avoid later NULL pointer dereference.
> 
> Reported-by: Alexander Bulekov <alxndr@bu.edu>
> Fixes: https://bugs.launchpad.net/qemu/+bug/1878259
> Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
> Suggested-by: Peter Maydell <peter.maydell@linaro.org>
> Signed-off-by: Prasad J Pandit <pjp@fedoraproject.org>
> ---
>  exec.c                | 1 +
>  include/exec/memory.h | 3 ++-
>  2 files changed, 3 insertions(+), 1 deletion(-)
> 
> Update v4: update address_space_map API doc text
>   -> https://lists.gnu.org/archive/html/qemu-devel/2020-05/msg07281.html
> 
> diff --git a/exec.c b/exec.c
> index 5162f0d12f..4eea84bf66 100644
> --- a/exec.c
> +++ b/exec.c
> @@ -3538,6 +3538,7 @@ void *address_space_map(AddressSpace *as,
>  
>      if (!memory_access_is_direct(mr, is_write)) {
>          if (atomic_xchg(&bounce.in_use, true)) {
> +            *plen = 0;

Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>

>              return NULL;
>          }
>          /* Avoid unbounded allocations */
> diff --git a/include/exec/memory.h b/include/exec/memory.h
> index e000bd2f97..8fa2d3cab2 100644
> --- a/include/exec/memory.h
> +++ b/include/exec/memory.h
> @@ -2303,7 +2303,8 @@ bool address_space_access_valid(AddressSpace *as, hwaddr addr, hwaddr len,
>  /* address_space_map: map a physical memory region into a host virtual address
>   *
>   * May map a subset of the requested range, given by and returned in @plen.
> - * May return %NULL if resources needed to perform the mapping are exhausted.
> + * May return %NULL and set *@plen to zero(0), if resources needed to perform
> + * the mapping are exhausted.
>   * Use only for reads OR writes - not for read-modify-write operations.
>   * Use cpu_register_map_client() to know when retrying the map operation is
>   * likely to succeed.
> 

Re: [PATCH v4] exec: set map length to zero when returning NULL

[Paolo Bonzini]

On 26/05/20 13:17, P J P wrote:
> From: Prasad J Pandit <pjp@fedoraproject.org>
> 
> When mapping physical memory into host's virtual address space,
> 'address_space_map' may return NULL if BounceBuffer is in_use.
> Set and return '*plen = 0' to avoid later NULL pointer dereference.
> 
> Reported-by: Alexander Bulekov <alxndr@bu.edu>
> Fixes: https://bugs.launchpad.net/qemu/+bug/1878259
> Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
> Suggested-by: Peter Maydell <peter.maydell@linaro.org>
> Signed-off-by: Prasad J Pandit <pjp@fedoraproject.org>
> ---
>  exec.c                | 1 +
>  include/exec/memory.h | 3 ++-
>  2 files changed, 3 insertions(+), 1 deletion(-)
> 
> Update v4: update address_space_map API doc text
>   -> https://lists.gnu.org/archive/html/qemu-devel/2020-05/msg07281.html
> 
> diff --git a/exec.c b/exec.c
> index 5162f0d12f..4eea84bf66 100644
> --- a/exec.c
> +++ b/exec.c
> @@ -3538,6 +3538,7 @@ void *address_space_map(AddressSpace *as,
>  
>      if (!memory_access_is_direct(mr, is_write)) {
>          if (atomic_xchg(&bounce.in_use, true)) {
> +            *plen = 0;
>              return NULL;
>          }
>          /* Avoid unbounded allocations */
> diff --git a/include/exec/memory.h b/include/exec/memory.h
> index e000bd2f97..8fa2d3cab2 100644
> --- a/include/exec/memory.h
> +++ b/include/exec/memory.h
> @@ -2303,7 +2303,8 @@ bool address_space_access_valid(AddressSpace *as, hwaddr addr, hwaddr len,
>  /* address_space_map: map a physical memory region into a host virtual address
>   *
>   * May map a subset of the requested range, given by and returned in @plen.
> - * May return %NULL if resources needed to perform the mapping are exhausted.
> + * May return %NULL and set *@plen to zero(0), if resources needed to perform
> + * the mapping are exhausted.
>   * Use only for reads OR writes - not for read-modify-write operations.
>   * Use cpu_register_map_client() to know when retrying the map operation is
>   * likely to succeed.
> 

Queued, thanks.

Paolo