On 2020/10/30 22:35, Peter Maydell wrote:
> On Fri, 30 Oct 2020 at 14:29, Peter Maydell <peter.mayd...@linaro.org> wrote:
>>
>> On Fri, 30 Oct 2020 at 10:23, AlexChen <alex.c...@huawei.com> wrote:
>>>
>>> In omap_lcd_interrupts(), the pointer omap_lcd is dereferenced before
>>> being check if it is valid, which may lead to NULL pointer dereference.
>>> So move the assignment to surface after checking that the omap_lcd is valid.
>>>
>>> Reported-by: Euler Robot <euler.ro...@huawei.com>
>>> Signed-off-by: Alex Chen <alex.c...@huawei.com>
>>> ---
>>> hw/display/omap_lcdc.c | 3 ++-
>>> 1 file changed, 2 insertions(+), 1 deletion(-)
>
>> Applied to target-arm.next, thanks.
>
> Whoops, spoke too soon. This doesn't compile:
>
> ../../hw/display/omap_lcdc.c: In function ‘omap_update_display’:
> ../../hw/display/omap_lcdc.c:88:10: error: ‘surface’ may be used
> uninitialized in this function [-Werror=maybe-uninitialized]
> !surface_bits_per_pixel(surface)) {
> ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>
>
> because the early exit check
> if (!omap_lcd || omap_lcd->plm == 1 || !omap_lcd->enable ||
> !surface_bits_per_pixel(surface)) {
> return;
> }
>
> uses 'surface' and this patch moves the initialization of that
> variable down below its first use.
>
Oh, I apologize for this compilation error, I will fix it in my patch v2.
Thanks,
Alex
