Eduardo Habkost <ehabk...@redhat.com> writes: > On Thu, Sep 24, 2015 at 08:57:21PM +0200, Markus Armbruster wrote: > [...] >> diff --git a/hw/arm/allwinner-a10.c b/hw/arm/allwinner-a10.c >> index ff249af..7692090 100644 >> --- a/hw/arm/allwinner-a10.c >> +++ b/hw/arm/allwinner-a10.c >> @@ -103,6 +103,8 @@ static void aw_a10_class_init(ObjectClass *oc, void >> *data) >> DeviceClass *dc = DEVICE_CLASS(oc); >> >> dc->realize = aw_a10_realize; >> + /* Reason: creates a CPU, thus use after free(), see cpu_class_init() */ >> + dc->cannot_even_create_with_object_new_yet = true; > > The comments at aw_a10_class_init(), digic_class_init(), > fsl_imx25_class_init(), fsl_imx31_class_init(), and > xlnx_zynqmp_class_init() are now outdated, as cpu_class_init() doesn't > set cannot_even_create_with_object_new_yet anymore. > > We could do this: > * Update the comments to "Reason: creates an ARM CPU, thus use after > free(), see arm_cpu_class_init()"
Yes. > * Add a note at arm_cpu_class_init() saying that we can probably > unset cannot_even_create_with_object_new_yet in those functions > once we fix TYPE_ARM_CPU Okay. Thanks!
