+Markus/Eduardo
On 11/1/21 18:33, Peter Maydell wrote:
> On Thu, 21 Oct 2021 at 19:40, Hao Wu <wuhao...@google.com> wrote:
>>
>> The ID can be used to indicate SMBus modules when adding
>> dynamic devices to them.
>>
>> Signed-off-by: Hao Wu <wuhao...@google.com>
>> ---
>> hw/arm/npcm7xx.c | 1 +
>> 1 file changed, 1 insertion(+)
>>
>> diff --git a/hw/arm/npcm7xx.c b/hw/arm/npcm7xx.c
>> index 2ab0080e0b..72953d65ef 100644
>> --- a/hw/arm/npcm7xx.c
>> +++ b/hw/arm/npcm7xx.c
>> @@ -421,6 +421,7 @@ static void npcm7xx_init(Object *obj)
>> for (i = 0; i < ARRAY_SIZE(s->smbus); i++) {
>> object_initialize_child(obj, "smbus[*]", &s->smbus[i],
>> TYPE_NPCM7XX_SMBUS);
>> + DEVICE(&s->smbus[i])->id = g_strdup_printf("smbus[%d]", i);
>> }
>
> This one looks weird to me -- I'm pretty sure we shouldn't be messing
> about with the DeviceState id string like that. It's supposed to be
> internal to the QOM/qdev code.
I agree with you, however:
$ git grep -F -- '->id = g_strdup' hw
hw/arm/virt.c:1512: dev->id = g_strdup(TYPE_PLATFORM_BUS_DEVICE);
hw/block/xen-block.c:761: drive->id = g_strdup(id);
hw/block/xen-block.c:853: iothread->id = g_strdup(id);
hw/core/machine-qmp-cmds.c:169: m->id =
g_strdup(object_get_canonical_path_component(obj));
hw/mem/pc-dimm.c:244: di->id = g_strdup(dev->id);
hw/pci-bridge/pci_expander_bridge.c:248: bds->id =
g_strdup(dev_name);
hw/ppc/e500.c:1009: dev->id = g_strdup(TYPE_PLATFORM_BUS_DEVICE);
hw/s390x/s390-pci-bus.c:1003: dev->id =
g_strdup_printf("auto_%02x:%02x.%01x",
hw/sh4/sh7750.c:819: dev->id = g_strdup("sci");
hw/sh4/sh7750.c:836: dev->id = g_strdup("scif");
hw/virtio/virtio-mem-pci.c:69: vi->id = g_strdup(dev->id);
hw/virtio/virtio-pmem-pci.c:74: vi->id = g_strdup(dev->id);
