Re: [PATCH v1] spi: spi-aspeed-smc: convert devfdt_get_addr_index output to phys_addr_t

[Johan Jonker <jbx6244@gmail.com>]

On 2/25/23 10:26, Michael Nazzareno Trimarchi wrote:
> Hi Johan
> 
> On Thu, Feb 23, 2023 at 3:06 PM Johan Jonker <jbx6244@gmail.com> wrote:
>>
>> The fdt_addr_t and phys_addr_t size have been decoupled.
>> A 32bit CPU can expect 64-bit data from the device tree parser,
>> so convert devfdt_get_addr_index output to phys_addr_t in the
>> spi-aspeed-smc.c file. Also fix dev_dbg to be able to handle
>> both sizes.
>>
>> Signed-off-by: Johan Jonker <jbx6244@gmail.com>
>> ---
>>
>> Note:
>>
>> This is needed for a Rockchip patch serie to pass the test and
>> must be merged before by Rockchip maintainers:
>>
>> [PATCH v4 00/11] Fixes for Rockchip NFC driver part 1
>> https://lore.kernel.org/u-boot/f3dba231-4a55-0a94-dfab-5cab1419d132@gmail.com/
>>
>> Request for hardware test by Aspeed people.
>> ---
>>  drivers/spi/spi-aspeed-smc.c | 8 ++++----
>>  1 file changed, 4 insertions(+), 4 deletions(-)
>>
>> diff --git a/drivers/spi/spi-aspeed-smc.c b/drivers/spi/spi-aspeed-smc.c
>> index a3c96333..17696441 100644
>> --- a/drivers/spi/spi-aspeed-smc.c
>> +++ b/drivers/spi/spi-aspeed-smc.c
>> @@ -1125,14 +1125,14 @@ static int apseed_spi_of_to_plat(struct udevice *bus)
>>         int ret;
>>         struct clk hclk;
>>
>> -       priv->regs = (void __iomem *)devfdt_get_addr_index(bus, 0);
>> +       priv->regs = (void __iomem *)((phys_addr_t)devfdt_get_addr_index(bus, 0));
> 

> devfdt_get_addr_index_ptr is not the right way?

That may work.

> 
>>         if ((u32)priv->regs == FDT_ADDR_T_NONE) {
>>                 dev_err(bus, "wrong ctrl base\n");
>>                 return -ENODEV;
>>         }
>>
>>         plat->ahb_base =
>> -               (void __iomem *)devfdt_get_addr_size_index(bus, 1, &plat->ahb_sz);
>> +               (void __iomem *)((phys_addr_t)devfdt_get_addr_size_index(bus, 1, &plat->ahb_sz));
> 

> devfdt_get_addr_index_ptr, same here

fdt_addr_t devfdt_get_addr_size_index(const struct udevice *dev, int index,
				      fdt_size_t *size)
vs

void *devfdt_get_addr_index_ptr(const struct udevice *dev, int index)

===

devfdt_get_addr_index_ptr does not offer the ahb_sz size.

https://elixir.bootlin.com/u-boot/latest/source/drivers/core/fdtaddr.c#L112

This function is used for that in devfdt_get_addr_size_index:

fdtdec_get_addr_size_auto_noparent(gd->fdt_blob, dev_of_offset(dev),
					   "reg", index, size, false);

What would you suggest? Keep the cast or use fdtdec_get_addr_size_auto_noparent ?

I'm a little bit "reluctant" to poke in code that I don't have hardware for.

Johan

> 
>>         if ((u32)plat->ahb_base == FDT_ADDR_T_NONE) {
>>                 dev_err(bus, "wrong AHB base\n");
>>                 return -ENODEV;
>> @@ -1151,8 +1151,8 @@ static int apseed_spi_of_to_plat(struct udevice *bus)
>>         plat->hclk_rate = clk_get_rate(&hclk);
>>         clk_free(&hclk);
>>
>> -       dev_dbg(bus, "ctrl_base = 0x%x, ahb_base = 0x%p, size = 0x%lx\n",
>> -               (u32)priv->regs, plat->ahb_base, plat->ahb_sz);
>> +       dev_dbg(bus, "ctrl_base = 0x%x, ahb_base = 0x%p, size = 0x%llx\n",
>> +               (u32)priv->regs, plat->ahb_base, (fdt64_t)plat->ahb_sz);
>>         dev_dbg(bus, "hclk = %dMHz, max_cs = %d\n",
>>                 plat->hclk_rate / 1000000, plat->max_cs);
>>
>> --
> 
> Michael
> 
>> 2.20.1
>>
> 
>