Re: [PATCH 1/2] dp264: use pci_create() to initialise the cmd646 device

[Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>]

On 14/02/2020 11:47, Philippe Mathieu-Daudé wrote:

> Hi Mark,
> 
> On Fri, Feb 14, 2020 at 9:48 AM Mark Cave-Ayland
> <mark.cave-ayland@ilande.co.uk> wrote:
>>
>> Remove the call to pci_cmd646_ide_init() since global device init functions
>> are deprecated in preference of using qdev directly.
>>
>> Signed-off-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
>> ---
>>  hw/alpha/dp264.c | 8 +++++++-
>>  1 file changed, 7 insertions(+), 1 deletion(-)
>>
>> diff --git a/hw/alpha/dp264.c b/hw/alpha/dp264.c
>> index a8f9a89cc4..e91989bf9a 100644
>> --- a/hw/alpha/dp264.c
>> +++ b/hw/alpha/dp264.c
>> @@ -16,6 +16,7 @@
>>  #include "sysemu/sysemu.h"
>>  #include "hw/rtc/mc146818rtc.h"
>>  #include "hw/ide.h"
>> +#include "hw/ide/pci.h"
>>  #include "hw/timer/i8254.h"
>>  #include "hw/isa/superio.h"
>>  #include "hw/dma/i8257.h"
>> @@ -100,9 +101,14 @@ static void clipper_init(MachineState *machine)
>>      /* IDE disk setup.  */
>>      {
>>          DriveInfo *hd[MAX_IDE_BUS * MAX_IDE_DEVS];
>> +        PCIDevice *pci_dev;
>> +
>>          ide_drive_get(hd, ARRAY_SIZE(hd));
>>
>> -        pci_cmd646_ide_init(pci_bus, hd, 0);
>> +        pci_dev = pci_create(pci_bus, -1, "cmd646-ide");
> 
> Not this patch problem, but it would be nice to have a TYPE_CMD646_IDE.

Yeah - there are quite a few places where this could be improved, but certainly a
patch for another day.

>> +        qdev_prop_set_uint32(DEVICE(pci_dev), "secondary", 0);
> 
> Secondary_ide disabled is the default in cmd646_ide_properties[], can
> we avoid this call?

Sure. The patch in its current form is a very simple like-for-like conversion, but it
can be easily be removed.

If we're going to do this then another thing to consider is if we want to replace the
-1 in pci_create() with a specific PCI_DEVFN value to make things a bit more robust.


ATB,

Mark.