Re: [PATCH v16 06/45] firmware: arm_rmm: Ensure the RMM has GPT entries for memory

On 03/08/2026 14:43, Steven Price wrote:
> The RMM maintains the state of all the granules in the system to make
> sure that the host is abiding by the rules. This state can be maintained
> at different granularity, per page (TRACKING_FINE) or per region
> (TRACKING_COARSE). The region size depends on the underlying
> "RMI_GRANULE_SIZE". For a "coarse" region all pages in the region must
> be of the same state, this implies we need to have "fine" tracking for
> DRAM, so that we can delegate individual pages.
> 
> For now we only support a statically carved out memory for tracking
> granules for the "fine" regions. This can be extended in the future to
> allow modifying the tracking granularity and remove the need for a
> static allocation.
> 
> Similarly, the firmware may create L0 GPT entries describing the total
> address space. But if we change the "PAS" (Physical Address Space) of a
> granule then the firmware may need to create L1 tables to track the PAS
> at a finer granularity.

This sounds a bit incomplete to me. We could add:

"Again, this series do not support creation of L1 GPT tables yet. Thus 
make sure that the firmware has L1 GPTs covering the DRAM region."


> 
> Signed-off-by: Steven Price <[email protected]>
> ---
> Changes since v15:
>   * Skip firmware-reserved NOMAP memory in rmi_init_metadata()
>   * Handle negative error codes from wrappers.
> Changes since v14:
>   * Move the implementation into drivers/firmware/arm_rmm.
> Changes since v13:
>   * Moved out of KVM
> ---
>   drivers/firmware/arm_rmm/rmi.c | 101 +++++++++++++++++++++++++++++++++
>   include/linux/arm-rmi-cmds.h   |   2 +
>   2 files changed, 103 insertions(+)
> 
> diff --git a/drivers/firmware/arm_rmm/rmi.c b/drivers/firmware/arm_rmm/rmi.c
> index 51ea661cecf9..bce3304bbc1a 100644
> --- a/drivers/firmware/arm_rmm/rmi.c
> +++ b/drivers/firmware/arm_rmm/rmi.c
> @@ -12,6 +12,8 @@
>   #include <asm/memory.h>
>   #include <asm/pgtable-hwdef.h>
>   
> +static bool arm64_rmi_is_available;
> +
>   /* Currently only the first 2 registers are used by Linux */
>   #define RMI_FEAT_REG_COUNT	2
>   static __ro_after_init unsigned long rmi_feat_reg_cache[RMI_FEAT_REG_COUNT];
> @@ -647,6 +649,98 @@ static int rmi_configure(void)
>   	return ret;
>   }
>   
> +/*
> + * Make sure the area is tracked by RMM at FINE granularity.
> + * We do not support changing the tracking yet.
> + */
> +static int rmi_verify_memory_tracking(phys_addr_t start, phys_addr_t end)
> +{
> +	while (start < end) {
> +		unsigned long ret, category, state, next;
> +
> +		ret = rmi_granule_tracking_get(start, end, &category, &state, &next);
> +		if (ret != RMI_SUCCESS ||
> +		    state != RMI_TRACKING_FINE ||
> +		    category != RMI_MEM_CATEGORY_CONVENTIONAL) {
> +			/* TODO: Set granule tracking in this case */
> +			pr_err("Granule tracking for region isn't fine/conventional: %llx\n",
> +			       start);
> +			return -ENODEV;
> +		}
> +		start = next;
> +	}
> +
> +	return 0;
> +}
> +
> +static int rmi_create_gpts(phys_addr_t start, phys_addr_t end)
> +{
> +	struct rmi_sro_state *sro;
> +	unsigned long l0gpt_sz;
> +
> +	sro = kmalloc_obj(*sro, GFP_KERNEL);
> +	if (!sro)
> +		return -ENOMEM;
> +
> +	l0gpt_sz = 1UL << (30 + FIELD_GET(RMI_FEATURE_REGISTER_1_L0GPTSZ,
> +					  rmi_feat_reg(1)));
> +	start = ALIGN_DOWN(start, l0gpt_sz);
> +	end = ALIGN(end, l0gpt_sz);
> +
> +	while (start < end) {
> +		long ret = rmi_gpt_l1_create(start, sro, GFP_KERNEL);
> +
> +		/*
> +		 * Make sure the L1 GPT tables are created for the region.
> +		 * RMI_ERROR_GPT indicates the L1 table already exists.
> +		 */

minor nit: The comment could be moved down closer to the check.

> +		if (ret < 0) {
> +			kfree(sro);
> +			return ret;
> +		}
> +
> +		if (ret != RMI_SUCCESS && RMI_RETURN_STATUS(ret) != RMI_ERROR_GPT) {
> +			pr_err("GPT Level1 table missing for %llx\n", start);
> +			kfree(sro);
> +			return -ENOMEM;
> +		}
> +		start += l0gpt_sz;
> +	}
> +

Rest looks good to me

Suzuki