[PATCH v2] riscv: enable multi-range memory layout

[Heinrich Schuchardt]

On 9/14/23 08:48, Wu, Fei wrote:
> On 9/14/2023 2:05 PM, Heinrich Schuchardt wrote:
>>
>>
>> Am 14. September 2023 07:30:55 MESZ schrieb Fei Wu <fei2.wu at intel.com>:
>>> In order to enable PCIe passthrough on qemu riscv, the physical memory
>>> range between 3GB and 4GB is reserved. Therefore if guest has 4GB ram,
>>> two ranges are created as [2G, 3G) and [4G, 7G), currently u-boot sets
>>> ram_top to 4G - 1 if the gd->ram_top is above 4G in
>>
>> This should move to 7GiB - 1 in your example on riscv64.
>>
> I'm describing the current implementation of board_get_usable_ram_top()
> in ./arch/riscv/cpu/generic/dram.c. Do you mean this function should be
> changed? Is the comment about 32bit DMA device still valid?
>
> phys_size_t board_get_usable_ram_top(phys_size_t total_size)
> {
>          /*
>           * Ensure that we run from first 4GB so that all
>           * addresses used by U-Boot are 32bit addresses.
>           *
>           * This in-turn ensures that 32bit DMA capable
>           * devices work fine because DMA mapping APIs will
>           * provide 32bit DMA addresses only.
>           */
>          if (gd->ram_top >= SZ_4G)
>                  return SZ_4G - 1;
>
>          return gd->ram_top;
> }

The comment above says 32bit DMA is board specific and not architecture
specific. So it is wrong to have this board_get_usable_ram_top()
function on architecture level. It makes usage of devices with all
memory above 4 GiB impossible.

It tried to pass through a SATA controller but received an error:

# modprobe vfio-pci
# echo 0000:06:00.0 > /sys/bus/pci/drivers/ahci/unbind
# echo 1022 7091 > /sys/bus/pci/drivers/vfio-pci/new_id
# qemu-system-riscv64 -kernel u-boot.bin -nographic -M virt -m 4G
-device vfio-pci,host=0000:06:00.0
qemu-system-riscv64: -device vfio-pci,host=0000:06:00.0: VFIO_MAP_DMA
failed: Invalid argument
qemu-system-riscv64: -device vfio-pci,host=0000:06:00.0: vfio
0000:06:00.0: failed to setup container for group 5: memory listener
initialization failed: Region riscv_virt_board.ram:
vfio_dma_map(0x55adbde66f70, 0x80000000, 0x100000000, 0x7fcd6fe00000) =
-22 (Invalid argument)

With which version of QEMU were you able to use PCI pass through?

Best regards

Heinrich

>
>>> board_get_usable_ram_top(), but that address is not backed by ram. This
>>> patch selects the lowest range instead.
>>>
>>> Signed-off-by: Fei Wu <fei2.wu at intel.com>
>>> ---
>>> arch/riscv/cpu/generic/dram.c        | 2 +-
>>> configs/qemu-riscv64_defconfig       | 2 +-
>>> configs/qemu-riscv64_smode_defconfig | 2 +-
>>> configs/qemu-riscv64_spl_defconfig   | 2 +-
>>> 4 files changed, 4 insertions(+), 4 deletions(-)
>>>
>>> diff --git a/arch/riscv/cpu/generic/dram.c b/arch/riscv/cpu/generic/dram.c
>>> index 44e11bd56c..fb53a57b4e 100644
>>> --- a/arch/riscv/cpu/generic/dram.c
>>> +++ b/arch/riscv/cpu/generic/dram.c
>>> @@ -13,7 +13,7 @@ DECLARE_GLOBAL_DATA_PTR;
>>>
>>> int dram_init(void)
>>> {
>>> -	return fdtdec_setup_mem_size_base();
>>> +	return fdtdec_setup_mem_size_base_lowest();
>>
>> Linaro is working on allowing to download a distro image via https and installing from a RAM disk.
>>
>> We should not artificially reduce the RAM size available for U-Boot by restricting ourselfs to 1 GiB.
>>
>> We must ensure that ram top is in the upper range.
>>
> How do they handle the case of >4GB? board_get_usable_ram_top() attached
> above always returns <4GB. And it seems like
> fdtdec_setup_mem_size_base() cannot ensure the upper one is picked, it
> just picks up one, but which one is selected depending on fdt.
>
> Thanks,
> Fei.
>
>> Best regards
>>
>> Heinrich
>>
>>> }
>>>
>>> int dram_init_banksize(void)
>>> diff --git a/configs/qemu-riscv64_defconfig b/configs/qemu-riscv64_defconfig
>>> index 9a8bbef192..aa55317d26 100644
>>> --- a/configs/qemu-riscv64_defconfig
>>> +++ b/configs/qemu-riscv64_defconfig
>>> @@ -1,6 +1,6 @@
>>> CONFIG_RISCV=y
>>> CONFIG_SYS_MALLOC_LEN=0x800000
>>> -CONFIG_NR_DRAM_BANKS=1
>>> +CONFIG_NR_DRAM_BANKS=2
>>> CONFIG_HAS_CUSTOM_SYS_INIT_SP_ADDR=y
>>> CONFIG_CUSTOM_SYS_INIT_SP_ADDR=0x80200000
>>> CONFIG_ENV_SIZE=0x20000
>>> diff --git a/configs/qemu-riscv64_smode_defconfig b/configs/qemu-riscv64_smode_defconfig
>>> index 1d0f021ade..de08a49dab 100644
>>> --- a/configs/qemu-riscv64_smode_defconfig
>>> +++ b/configs/qemu-riscv64_smode_defconfig
>>> @@ -1,6 +1,6 @@
>>> CONFIG_RISCV=y
>>> CONFIG_SYS_MALLOC_LEN=0x800000
>>> -CONFIG_NR_DRAM_BANKS=1
>>> +CONFIG_NR_DRAM_BANKS=2
>>> CONFIG_HAS_CUSTOM_SYS_INIT_SP_ADDR=y
>>> CONFIG_CUSTOM_SYS_INIT_SP_ADDR=0x80200000
>>> CONFIG_ENV_SIZE=0x20000
>>> diff --git a/configs/qemu-riscv64_spl_defconfig b/configs/qemu-riscv64_spl_defconfig
>>> index bb10145e6e..66dc2a1dd9 100644
>>> --- a/configs/qemu-riscv64_spl_defconfig
>>> +++ b/configs/qemu-riscv64_spl_defconfig
>>> @@ -1,6 +1,6 @@
>>> CONFIG_RISCV=y
>>> CONFIG_SYS_MALLOC_LEN=0x800000
>>> -CONFIG_NR_DRAM_BANKS=1
>>> +CONFIG_NR_DRAM_BANKS=2
>>> CONFIG_HAS_CUSTOM_SYS_INIT_SP_ADDR=y
>>> CONFIG_CUSTOM_SYS_INIT_SP_ADDR=0x80200000
>>> CONFIG_ENV_SIZE=0x20000
>