Previously, we incorrectly cast irq_domain->host_data directly to
mvebu_icu_msi_data. However, host_data actually stores a structure of
type msi_domain_info.

This incorrect assumption caused issues such as the thermal sensors of
the CP110 platform malfunctioning. Specifically, the translation of the
SEI interrupt to IRQ_TYPE_EDGE_RISING failed, preventing proper
interrupt handling. The following error was observed:
genirq: Setting trigger mode 4 for irq 85 failed (irq_chip_set_type_parent+0x0/0x34)
armada_thermal f2400000.system-controller:thermal-sensor@70: Cannot request threaded IRQ 85

This commit resolves the issue by first casting host_data to
msi_domain_info and then accessing the mvebu_icu_msi_data through
msi_domain_info->chip_data.

Cc: stable@vger.kernel.org
Fixes: d929e4db22b6 ("irqchip/irq-mvebu-icu: Prepare for real per device MSI")
Signed-off-by: Stefan Eichenberger <redacted>
---
Changes in v2:
* This patch is a v2 because it addresses the same issue as this patch:
  https://lore.kernel.org/all/20241217111623.92625-1-eichest@gmail.com/ (local)
* This time it addresses the underlying issue instead of adding a
  workaround (thanks to Thomas)

 drivers/irqchip/irq-mvebu-icu.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/irqchip/irq-mvebu-icu.c b/drivers/irqchip/irq-mvebu-icu.c
index b337f6c05f18..4eebed39880a 100644
--- a/drivers/irqchip/irq-mvebu-icu.c
+++ b/drivers/irqchip/irq-mvebu-icu.c

@@ -68,7 +68,8 @@ static int mvebu_icu_translate(struct irq_domain *d, struct irq_fwspec *fwspec,
 			       unsigned long *hwirq, unsigned int *type)
 {
 	unsigned int param_count = static_branch_unlikely(&legacy_bindings) ? 3 : 2;
-	struct mvebu_icu_msi_data *msi_data = d->host_data;
+	struct msi_domain_info *info = d->host_data;
+	struct mvebu_icu_msi_data *msi_data = info->chip_data;
 	struct mvebu_icu *icu = msi_data->icu;
 
 	/* Check the count of the parameters in dt */

-- 
2.45.2