Re: [PATCH v19 6/7] firmware: arm_rmm: Ensure the RMM has GPT entries for memory

On 24/09/2026 22:38, Jonathan Cameron wrote:

On Thu, 24 Sep 2026 14:52:00 +0100
Suzuki K Poulose [off-list ref] wrote:

quoted

From: Steven Price <steven.price@arm.com>

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 or TRACKING_INTERMEDIATE). The region size depends on the

...

quoted

If the firmware requires us to manage the tracking or GPT memory,
deactivate the RMM and reclaim any memory donated at RMM activation.

Apply the same checks when hotplugged memory is brought online.

Signed-off-by: Steven Price <steven.price@arm.com>
[ Switch to RMI_GPT_L1_INFO for checking GPTs and deactivate RMM ]
Co-developed-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>

A few trivial comments. Either way on those
Reviewed-by: Jonathan Cameron <redacted>

Thank you !

quoted

---
  drivers/firmware/arm_rmm/rmi.c | 218 ++++++++++++++++++++++++++++++++-
  include/linux/arm-rmi-cmds.h   |   2 +
  2 files changed, 219 insertions(+), 1 deletion(-)

diff --git a/drivers/firmware/arm_rmm/rmi.c b/drivers/firmware/arm_rmm/rmi.c
index 0859f256e192b..1a8f3debd844a 100644
--- a/drivers/firmware/arm_rmm/rmi.c
+++ b/drivers/firmware/arm_rmm/rmi.c

quoted

+
+/*
+ * rmi_gpt_info - Query the GPT info for the given PAR.
+ * @start: Base of the physical address region
+ * @end: Top of the physical address region
+ * @out_top: Top of the physical address region for which
+ *		the GPT @out_gpt_par_state is valid

wrap to 80.

Ack

quoted

+ * @out_gpt_par_state: State of the GPT covered by [start, out_top)
+ */
+static long rmi_gpt_info(unsigned long start, unsigned long end,
+			 unsigned long *out_top,
+			 unsigned long *out_gpt_par_state)
+{
+	struct arm_smccc_1_2_regs regs = {
+		SMC_RMI_GPT_INFO, start, end,
+	};
+
+	rmi_smccc_invoke(&regs);
+	if (regs.a0 != RMI_SUCCESS)
+		return regs.a0;
+
+	if (out_top)
+		*out_top = regs.a1;
+	if (out_gpt_par_state)
+		*out_gpt_par_state = regs.a2;
+
+	return RMI_SUCCESS;
+}
+
+/*
+ * We do not support creating L1 GPTs yet. So, make sure that
+ * all the regions are managed by the firmware.

Wrap to 80 chars.

Not relevant to this patch, but when are you thinking we will support that?

Once we have a full firmware stack that implements the L1GPT (Firme in tf-a and tf-RMM) we will take it up. That won't come before the first
drop of KVM is merged.

quoted

+
+static int rmi_init_metadata(void)
+{
+	phys_addr_t start, end;
+	struct memblock_region *r;
+
+	for_each_mem_region(r) {
+		int ret;
+
+		/* Firmware-reserved NOMAP regions are not usable system RAM */
+		if (memblock_is_nomap(r))
+			continue;
+
+		start = PAGE_ALIGN(r->base);
+		end = PAGE_ALIGN_DOWN(r->base + r->size);

Add a comment on why rounding down.  Doe we expect that to ever be relevant?

Do we really need a comment here ? We always deal with PAGE_SIZE and if
in the odd case this region boundary is not PAGE aligned we stick to the
page that we can use from the region. I can stick in a comment if you
really think so.


...

quoted

@@ -843,9 +1049,19 @@ static int __init arm64_init_rmi(void)

...

quoted

-	return ret;
+	ret = rmi_init_memory();
+	if (ret) {
+		/* Deactivate the RMM */
+		WARN_ON(rmi_sro_memxfer_cmd(sro, GFP_KERNEL, SMC_RMI_RMM_DEACTIVATE));
+		return ret;
+	}
+
+	arm64_rmi_is_available = true;
+	pr_info("RMI configured\n");

Blank line nice for the eye to spot the return.

Ack

Thanks for your review !

Suzuki