Re: [PATCH EDK2 v1 1/1] OvmfPkg/XenPvBlkDxe: add assert for pool allocating

["Laszlo Ersek" <lersek@redhat.com>]

On 11/23/20 07:35, Wenyi Xie wrote:
> add assert after pool allocating.
> 
> Cc: Jordan Justen <jordan.l.justen@intel.com>
> Cc: Laszlo Ersek <lersek@redhat.com>
> Cc: Ard Biesheuvel <ard.biesheuvel@arm.com>
> Signed-off-by: Wenyi Xie <xiewenyi2@huawei.com>
> ---
>  OvmfPkg/XenPvBlkDxe/BlockFront.c | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/OvmfPkg/XenPvBlkDxe/BlockFront.c b/OvmfPkg/XenPvBlkDxe/BlockFront.c
> index 122a6baed25a..bde385090832 100644
> --- a/OvmfPkg/XenPvBlkDxe/BlockFront.c
> +++ b/OvmfPkg/XenPvBlkDxe/BlockFront.c
> @@ -155,6 +155,7 @@ XenPvBlockFrontInitialization (
>    ASSERT (NodeName != NULL);
>  
>    Dev = AllocateZeroPool (sizeof (XEN_BLOCK_FRONT_DEVICE));
> +  ASSERT (Dev != NULL);
>    Dev->Signature = XEN_BLOCK_FRONT_SIGNATURE;
>    Dev->NodeName = NodeName;
>    Dev->XenBusIo = XenBusIo;
> 

(1) Please use "BaseTools/Scripts/GetMaintainer.py" for determining the
necessary CC's on a patch. On this patch, the Xen reviewers should be
CC'd as well.

(2) I agree that the AllocateZeroPool() result should be checked, but an
ASSERT() is inappropriate. The function's return type is EFI_STATUS, so
returning EFI_OUT_OF_RESOURCES should be possible. The call site in
XenPvBlkDxeDriverBindingStart() does check the return value.

Thanks
Laszlo