Re: [PATCH 2/2] hw/arm/vexpress: Avoid trivial memory leak of 'flashalias'

["Philippe Mathieu-Daudé" <philmd@linaro.org>]

On 12/5/23 19:02, Peter Maydell wrote:
> In the vexpress board code, we allocate a new MemoryRegion at the top
> of vexpress_common_init() but only set it up and use it inside the
> "if (map[VE_NORFLASHALIAS] != -1)" conditional, so we leak it if not.
> This isn't a very interesting leak as it's a tiny amount of memory
> once at startup, but it's easy to fix.
> 
> We could silence Coverity simply by moving the g_new() into the
> if() block, but this use of g_new(MemoryRegion, 1) is a legacy from
> when this board model was originally written; we wouldn't do that
> if we wrote it today. The MemoryRegions are conceptually a part of
> the board and must not go away until the whole board is done with
> (at the end of the simulation), so they belong in its state struct.
> 
> This machine already has a VexpressMachineState struct that extends
> MachineState, so statically put the MemoryRegions in there instead of
> dynamically allocating them separately at runtime.
> 
> Spotted by Coverity (CID 1509083).
> 
> Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
> ---
>   hw/arm/vexpress.c | 40 ++++++++++++++++++++--------------------
>   1 file changed, 20 insertions(+), 20 deletions(-)

Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>