[PATCH v2 0/6] irqchip/ACPI: Arm GICv5 IWB ACPI IRQ probe deferral

[PATCH v2 0/6] irqchip/ACPI: Arm GICv5 IWB ACPI IRQ probe deferral

[Lorenzo Pieralisi]

On Arm GICv5 systems, the IWB (Interrupt Wire Bridge) handles wired interrupts
and it is in charge of translating interrupt wires to GICv5 ITS messages.

In ACPI systems, an IWB is backed by a platform device. Device drivers for
devices whose IRQs are routed to an IWB might probe earlier than the IWB device
driver, which can result in failures in that until the IWB driver is probed and
its IRQ domain is duly registered, IRQs routed to it cannot be resolved.

Some interrupt controllers for the RISC-V architecture suffer from the same
issue; RISC-V, in its ACPI IRQ layer solved the problem by automatically
creating device dependencies that result in the interrupt controllers drivers
being probed before any device driver whose devices IRQs are routed to them is
probed.

Instead of reinventing the wheel, this series move the aforementioned RISC-V
code to generic ACPI IRQ handling code (while fixing some bits and pieces) and
implement GICv5 IWB ACPI probe deferral on top of it.

Signed-off-by: Lorenzo Pieralisi <lpieralisi@kernel.org>
---
Changes in v2:
- Split the patchset in several logic units according to review
- Link to v1: https://patch.msgid.link/20260505-gic-v5-acpi-iwb-probe-deferral-v1-0-b37b85998362@kernel.org

---
Lorenzo Pieralisi (6):
      ACPI: RISC-V: Fix riscv_acpi_irq_get_dep() loop termination
      ACPI: RISC-V: Check acpi_get_handle() status in riscv_acpi_add_prt_dep()
      ACPI: RISC-V: Fix riscv_acpi_add_prt_dep() loop handling
      ACPI: irq: Move RISC-V interrupt controllers autodep to ACPI IRQ code
      ACPI/IORT: Implement ACPI infrastructure to enable GICv5 IWB probe deferral
      irqchip/gic-v5: Enable GICv5 IWB ACPI probe ordering detection

 arch/riscv/include/asm/acpi.h       |   1 +
 drivers/acpi/arm64/iort.c           |  22 ++++-
 drivers/acpi/irq.c                  | 172 +++++++++++++++++++++++++++++++++++-
 drivers/acpi/riscv/irq.c            | 141 +----------------------------
 drivers/acpi/scan.c                 |   1 +
 drivers/irqchip/irq-gic-v3.c        |   2 +-
 drivers/irqchip/irq-gic-v5-iwb.c    |   5 ++
 drivers/irqchip/irq-gic-v5.c        |  13 ++-
 drivers/irqchip/irq-gic.c           |   2 +-
 drivers/irqchip/irq-loongarch-cpu.c |   2 +-
 drivers/irqchip/irq-riscv-intc.c    |   3 +-
 include/linux/acpi.h                |   5 +-
 include/linux/acpi_iort.h           |   3 +-
 13 files changed, 218 insertions(+), 154 deletions(-)
---
base-commit: 5d6919055dec134de3c40167a490f33c74c12581
change-id: 20260427-gic-v5-acpi-iwb-probe-deferral-cdc849b1a854

Best regards,
--  
Lorenzo Pieralisi <lpieralisi@kernel.org>

[PATCH v2 1/6] ACPI: RISC-V: Fix riscv_acpi_irq_get_dep() loop termination

[Lorenzo Pieralisi]

In riscv_acpi_add_irq_dep() the main loop condition would currently stop
the loop if an interrupt descriptor contains an interrupt for which the
respective GSI handle is NULL, which is not correct because subsequent
interrupts in the interrupt descriptor might still have a GSI dependency
that must not be skipped.

Rework riscv_acpi_add_irq_dep() and the riscv_acpi_irq_get_dep() call chain
to fix it - by not forcing the loop to stop in order to guarantee
dependency detection for all the interrupt entries in the CRS descriptor.

Signed-off-by: Lorenzo Pieralisi <lpieralisi@kernel.org>
Cc: "Rafael J. Wysocki" <rafael@kernel.org>
Cc: Sunil V L <sunilvl@ventanamicro.com>
---
 drivers/acpi/riscv/irq.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/drivers/acpi/riscv/irq.c b/drivers/acpi/riscv/irq.c
index 9b88d0993e88..cd83c3035cf6 100644
--- a/drivers/acpi/riscv/irq.c
+++ b/drivers/acpi/riscv/irq.c
@@ -299,6 +299,7 @@ static acpi_status riscv_acpi_irq_get_parent(struct acpi_resource *ares, void *c
 			return AE_OK;
 
 		ctx->handle = riscv_acpi_get_gsi_handle(eirq->interrupts[ctx->index]);
+		ctx->rc = 0;
 		return AE_CTRL_TERMINATE;
 	}
 
@@ -314,10 +315,8 @@ static int riscv_acpi_irq_get_dep(acpi_handle handle, unsigned int index, acpi_h
 
 	acpi_walk_resources(handle, METHOD_NAME__CRS, riscv_acpi_irq_get_parent, &ctx);
 	*gsi_handle = ctx.handle;
-	if (*gsi_handle)
-		return 1;
 
-	return 0;
+	return ctx.rc;
 }
 
 static u32 riscv_acpi_add_prt_dep(acpi_handle handle)
@@ -381,8 +380,11 @@ static u32 riscv_acpi_add_irq_dep(acpi_handle handle)
 	int i;
 
 	for (i = 0;
-	     riscv_acpi_irq_get_dep(handle, i, &gsi_handle);
+	     !riscv_acpi_irq_get_dep(handle, i, &gsi_handle);
 	     i++) {
+		if (!gsi_handle)
+			continue;
+
 		dep_devices.count = 1;
 		dep_devices.handles = kzalloc_objs(*dep_devices.handles, 1);
 		if (!dep_devices.handles) {

-- 
2.54.0

Re: [PATCH v2 1/6] ACPI: RISC-V: Fix riscv_acpi_irq_get_dep() loop termination

[Sunil V L]

Hi Lorenzo,

On Wed, Jun 3, 2026 at 1:51 PM Lorenzo Pieralisi <lpieralisi@kernel.org> wrote:
>
> In riscv_acpi_add_irq_dep() the main loop condition would currently stop
> the loop if an interrupt descriptor contains an interrupt for which the
> respective GSI handle is NULL, which is not correct because subsequent
> interrupts in the interrupt descriptor might still have a GSI dependency
> that must not be skipped.
>
> Rework riscv_acpi_add_irq_dep() and the riscv_acpi_irq_get_dep() call chain
> to fix it - by not forcing the loop to stop in order to guarantee
> dependency detection for all the interrupt entries in the CRS descriptor.
>
> Signed-off-by: Lorenzo Pieralisi <lpieralisi@kernel.org>
> Cc: "Rafael J. Wysocki" <rafael@kernel.org>
> Cc: Sunil V L <sunilvl@ventanamicro.com>
> ---
>  drivers/acpi/riscv/irq.c | 10 ++++++----
>  1 file changed, 6 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/acpi/riscv/irq.c b/drivers/acpi/riscv/irq.c
> index 9b88d0993e88..cd83c3035cf6 100644
> --- a/drivers/acpi/riscv/irq.c
> +++ b/drivers/acpi/riscv/irq.c
> @@ -299,6 +299,7 @@ static acpi_status riscv_acpi_irq_get_parent(struct acpi_resource *ares, void *c
>                         return AE_OK;
>
>                 ctx->handle = riscv_acpi_get_gsi_handle(eirq->interrupts[ctx->index]);
> +               ctx->rc = 0;
>                 return AE_CTRL_TERMINATE;
>         }
>
> @@ -314,10 +315,8 @@ static int riscv_acpi_irq_get_dep(acpi_handle handle, unsigned int index, acpi_h
>
>         acpi_walk_resources(handle, METHOD_NAME__CRS, riscv_acpi_irq_get_parent, &ctx);
>         *gsi_handle = ctx.handle;
> -       if (*gsi_handle)
> -               return 1;
>
> -       return 0;
> +       return ctx.rc;
>  }
>
>  static u32 riscv_acpi_add_prt_dep(acpi_handle handle)
> @@ -381,8 +380,11 @@ static u32 riscv_acpi_add_irq_dep(acpi_handle handle)
>         int i;
>
>         for (i = 0;
> -            riscv_acpi_irq_get_dep(handle, i, &gsi_handle);
> +            !riscv_acpi_irq_get_dep(handle, i, &gsi_handle);
>              i++) {
> +               if (!gsi_handle)
> +                       continue;
> +
>                 dep_devices.count = 1;
>                 dep_devices.handles = kzalloc_objs(*dep_devices.handles, 1);
>                 if (!dep_devices.handles) {
>
Do these fixes need the Fixes tag?

Otherwise, LGTM.
Reviewed-by: Sunil V L <sunilvl@oss.qualcomm.com>

Thanks!
Sunil

Re: [PATCH v2 1/6] ACPI: RISC-V: Fix riscv_acpi_irq_get_dep() loop termination

[Lorenzo Pieralisi]

On Mon, Jun 08, 2026 at 09:54:28PM +0530, Sunil V L wrote:
> Hi Lorenzo,
> 
> On Wed, Jun 3, 2026 at 1:51 PM Lorenzo Pieralisi <lpieralisi@kernel.org> wrote:
> >
> > In riscv_acpi_add_irq_dep() the main loop condition would currently stop
> > the loop if an interrupt descriptor contains an interrupt for which the
> > respective GSI handle is NULL, which is not correct because subsequent
> > interrupts in the interrupt descriptor might still have a GSI dependency
> > that must not be skipped.
> >
> > Rework riscv_acpi_add_irq_dep() and the riscv_acpi_irq_get_dep() call chain
> > to fix it - by not forcing the loop to stop in order to guarantee
> > dependency detection for all the interrupt entries in the CRS descriptor.
> >
> > Signed-off-by: Lorenzo Pieralisi <lpieralisi@kernel.org>
> > Cc: "Rafael J. Wysocki" <rafael@kernel.org>
> > Cc: Sunil V L <sunilvl@ventanamicro.com>
> > ---
> >  drivers/acpi/riscv/irq.c | 10 ++++++----
> >  1 file changed, 6 insertions(+), 4 deletions(-)
> >
> > diff --git a/drivers/acpi/riscv/irq.c b/drivers/acpi/riscv/irq.c
> > index 9b88d0993e88..cd83c3035cf6 100644
> > --- a/drivers/acpi/riscv/irq.c
> > +++ b/drivers/acpi/riscv/irq.c
> > @@ -299,6 +299,7 @@ static acpi_status riscv_acpi_irq_get_parent(struct acpi_resource *ares, void *c
> >                         return AE_OK;
> >
> >                 ctx->handle = riscv_acpi_get_gsi_handle(eirq->interrupts[ctx->index]);
> > +               ctx->rc = 0;
> >                 return AE_CTRL_TERMINATE;
> >         }
> >
> > @@ -314,10 +315,8 @@ static int riscv_acpi_irq_get_dep(acpi_handle handle, unsigned int index, acpi_h
> >
> >         acpi_walk_resources(handle, METHOD_NAME__CRS, riscv_acpi_irq_get_parent, &ctx);
> >         *gsi_handle = ctx.handle;
> > -       if (*gsi_handle)
> > -               return 1;
> >
> > -       return 0;
> > +       return ctx.rc;
> >  }
> >
> >  static u32 riscv_acpi_add_prt_dep(acpi_handle handle)
> > @@ -381,8 +380,11 @@ static u32 riscv_acpi_add_irq_dep(acpi_handle handle)
> >         int i;
> >
> >         for (i = 0;
> > -            riscv_acpi_irq_get_dep(handle, i, &gsi_handle);
> > +            !riscv_acpi_irq_get_dep(handle, i, &gsi_handle);
> >              i++) {
> > +               if (!gsi_handle)
> > +                       continue;
> > +
> >                 dep_devices.count = 1;
> >                 dep_devices.handles = kzalloc_objs(*dep_devices.handles, 1);
> >                 if (!dep_devices.handles) {
> >
> Do these fixes need the Fixes tag?

I can add a Fixes: tag but I wanted first some help testing them, it
is code perusal that got me there.

> Otherwise, LGTM.
> Reviewed-by: Sunil V L <sunilvl@oss.qualcomm.com>

Thanks,
Lorenzo

Re: [PATCH v2 1/6] ACPI: RISC-V: Fix riscv_acpi_irq_get_dep() loop termination

[Sunil V L]

On Tue, Jun 9, 2026 at 2:49 PM Lorenzo Pieralisi <lpieralisi@kernel.org> wrote:
>
> On Mon, Jun 08, 2026 at 09:54:28PM +0530, Sunil V L wrote:
> > Hi Lorenzo,
> >
> > On Wed, Jun 3, 2026 at 1:51 PM Lorenzo Pieralisi <lpieralisi@kernel.org> wrote:
> > >
[...]
> > Do these fixes need the Fixes tag?
>
> I can add a Fixes: tag but I wanted first some help testing them, it
> is code perusal that got me there.
>
Oh, yes. I tested these changes on the RISC-V virt platform.

So for RISC-V testing,

Tested-by: Sunil V L <sunilvl@oss.qualcomm.com>

> > Otherwise, LGTM.
> > Reviewed-by: Sunil V L <sunilvl@oss.qualcomm.com>
>
> Thanks,
> Lorenzo

[PATCH v2 2/6] ACPI: RISC-V: Check acpi_get_handle() status in riscv_acpi_add_prt_dep()

[Lorenzo Pieralisi]

In riscv_acpi_add_prt_dep(), the acpi_get_handle() call can fail which
would leave link_handle uninitialized.

Fix it by checking the acpi_get_handle() return status and skip the entry
if it fails.

Signed-off-by: Lorenzo Pieralisi <lpieralisi@kernel.org>
Cc: "Rafael J. Wysocki" <rafael@kernel.org>
Cc: Sunil V L <sunilvl@ventanamicro.com>
---
 drivers/acpi/riscv/irq.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/acpi/riscv/irq.c b/drivers/acpi/riscv/irq.c
index cd83c3035cf6..75170151c614 100644
--- a/drivers/acpi/riscv/irq.c
+++ b/drivers/acpi/riscv/irq.c
@@ -339,7 +339,9 @@ static u32 riscv_acpi_add_prt_dep(acpi_handle handle)
 	entry = buffer.pointer;
 	while (entry && (entry->length > 0)) {
 		if (entry->source[0]) {
-			acpi_get_handle(handle, entry->source, &link_handle);
+			status = acpi_get_handle(handle, entry->source, &link_handle);
+			if (ACPI_FAILURE(status))
+				continue;
 			dep_devices.count = 1;
 			dep_devices.handles = kzalloc_objs(*dep_devices.handles,
 							   1);

-- 
2.54.0

Re: [PATCH v2 2/6] ACPI: RISC-V: Check acpi_get_handle() status in riscv_acpi_add_prt_dep()

[Sunil V L]

On Wed, Jun 3, 2026 at 1:52 PM Lorenzo Pieralisi <lpieralisi@kernel.org> wrote:
>
> In riscv_acpi_add_prt_dep(), the acpi_get_handle() call can fail which
> would leave link_handle uninitialized.
>
> Fix it by checking the acpi_get_handle() return status and skip the entry
> if it fails.
>
> Signed-off-by: Lorenzo Pieralisi <lpieralisi@kernel.org>
> Cc: "Rafael J. Wysocki" <rafael@kernel.org>
> Cc: Sunil V L <sunilvl@ventanamicro.com>
> ---
>  drivers/acpi/riscv/irq.c | 4 +++-
>  1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/acpi/riscv/irq.c b/drivers/acpi/riscv/irq.c
> index cd83c3035cf6..75170151c614 100644
> --- a/drivers/acpi/riscv/irq.c
> +++ b/drivers/acpi/riscv/irq.c
> @@ -339,7 +339,9 @@ static u32 riscv_acpi_add_prt_dep(acpi_handle handle)
>         entry = buffer.pointer;
>         while (entry && (entry->length > 0)) {
>                 if (entry->source[0]) {
> -                       acpi_get_handle(handle, entry->source, &link_handle);
> +                       status = acpi_get_handle(handle, entry->source, &link_handle);
> +                       if (ACPI_FAILURE(status))
> +                               continue;
>                         dep_devices.count = 1;
>
Reviewed-by: Sunil V L <sunilvl@oss.qualcomm.com>

[PATCH v2 3/6] ACPI: RISC-V: Fix riscv_acpi_add_prt_dep() loop handling

[Lorenzo Pieralisi]

The loop in riscv_acpi_add_prt_dep() includes error conditions that are
handled in a dubious - if not outright wrong - way, by continuining the
loop (which skips and misses the entry pointer update to point to the next
entry).

Rewrite the loop as a for loop (that handles the continuation correctly)
and wrap the condition and update statements using helper functions to make
it cleaner.

Signed-off-by: Lorenzo Pieralisi <lpieralisi@kernel.org>
Cc: "Rafael J. Wysocki" <rafael@kernel.org>
Cc: Sunil V L <sunilvl@ventanamicro.com>
---
 drivers/acpi/riscv/irq.c | 19 +++++++++++++++----
 1 file changed, 15 insertions(+), 4 deletions(-)

diff --git a/drivers/acpi/riscv/irq.c b/drivers/acpi/riscv/irq.c
index 75170151c614..0cdec5dd575e 100644
--- a/drivers/acpi/riscv/irq.c
+++ b/drivers/acpi/riscv/irq.c
@@ -319,6 +319,20 @@ static int riscv_acpi_irq_get_dep(acpi_handle handle, unsigned int index, acpi_h
 	return ctx.rc;
 }
 
+static bool acpi_prt_entry_valid(void *prt_entry)
+{
+	struct acpi_pci_routing_table *entry = prt_entry;
+
+	return entry && entry->length > 0;
+}
+
+static void *acpi_prt_next_entry(void *prt_entry)
+{
+	struct acpi_pci_routing_table *entry = prt_entry;
+
+	return prt_entry + entry->length;
+}
+
 static u32 riscv_acpi_add_prt_dep(acpi_handle handle)
 {
 	struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL };
@@ -337,7 +351,7 @@ static u32 riscv_acpi_add_prt_dep(acpi_handle handle)
 	}
 
 	entry = buffer.pointer;
-	while (entry && (entry->length > 0)) {
+	for (; acpi_prt_entry_valid(entry); entry = acpi_prt_next_entry(entry)) {
 		if (entry->source[0]) {
 			status = acpi_get_handle(handle, entry->source, &link_handle);
 			if (ACPI_FAILURE(status))
@@ -365,9 +379,6 @@ static u32 riscv_acpi_add_prt_dep(acpi_handle handle)
 			dep_devices.handles[0] = gsi_handle;
 			count += acpi_scan_add_dep(handle, &dep_devices);
 		}
-
-		entry = (struct acpi_pci_routing_table *)
-			((unsigned long)entry + entry->length);
 	}
 
 	kfree(buffer.pointer);

-- 
2.54.0

Re: [PATCH v2 3/6] ACPI: RISC-V: Fix riscv_acpi_add_prt_dep() loop handling

[Sunil V L]

On Wed, Jun 3, 2026 at 2:18 PM Lorenzo Pieralisi <lpieralisi@kernel.org> wrote:
>
> The loop in riscv_acpi_add_prt_dep() includes error conditions that are
> handled in a dubious - if not outright wrong - way, by continuining the
> loop (which skips and misses the entry pointer update to point to the next
> entry).
>
> Rewrite the loop as a for loop (that handles the continuation correctly)
> and wrap the condition and update statements using helper functions to make
> it cleaner.
>
> Signed-off-by: Lorenzo Pieralisi <lpieralisi@kernel.org>
> Cc: "Rafael J. Wysocki" <rafael@kernel.org>
> Cc: Sunil V L <sunilvl@ventanamicro.com>
> ---
>  drivers/acpi/riscv/irq.c | 19 +++++++++++++++----
>  1 file changed, 15 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/acpi/riscv/irq.c b/drivers/acpi/riscv/irq.c
> index 75170151c614..0cdec5dd575e 100644
> --- a/drivers/acpi/riscv/irq.c
> +++ b/drivers/acpi/riscv/irq.c
> @@ -319,6 +319,20 @@ static int riscv_acpi_irq_get_dep(acpi_handle handle, unsigned int index, acpi_h
>         return ctx.rc;
>  }
>
> +static bool acpi_prt_entry_valid(void *prt_entry)
> +{
> +       struct acpi_pci_routing_table *entry = prt_entry;
> +
> +       return entry && entry->length > 0;
> +}
> +
> +static void *acpi_prt_next_entry(void *prt_entry)
> +{
> +       struct acpi_pci_routing_table *entry = prt_entry;
> +
> +       return prt_entry + entry->length;
> +}
> +
>  static u32 riscv_acpi_add_prt_dep(acpi_handle handle)
>  {
>         struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL };
> @@ -337,7 +351,7 @@ static u32 riscv_acpi_add_prt_dep(acpi_handle handle)
>         }
>
>         entry = buffer.pointer;
> -       while (entry && (entry->length > 0)) {
> +       for (; acpi_prt_entry_valid(entry); entry = acpi_prt_next_entry(entry)) {
>                 if (entry->source[0]) {
>                         status = acpi_get_handle(handle, entry->source, &link_handle);
>                         if (ACPI_FAILURE(status))
> @@ -365,9 +379,6 @@ static u32 riscv_acpi_add_prt_dep(acpi_handle handle)
>                         dep_devices.handles[0] = gsi_handle;
>                         count += acpi_scan_add_dep(handle, &dep_devices);
>                 }
> -
> -               entry = (struct acpi_pci_routing_table *)
> -                       ((unsigned long)entry + entry->length);
>         }
>
Reviewed-by: Sunil V L <sunilvl@oss.qualcomm.com>

[PATCH v2 4/6] ACPI: irq: Move RISC-V interrupt controllers autodep to ACPI IRQ code

[Lorenzo Pieralisi]

RISC-V implements arch code to detect probe dependencies for devices and
the interrupt controller the devices GSIs are routed to.

The code itself is arch agnostic apart from an arch specific helper
function required to retrieve the acpi_handle of the interrupt controller
that manages the device GSI interrupt.

In order to enable IRQ probe dependencies detection on other architectures,
move RISC-V IRQ probe dependency detection code to generic ACPI IRQ code.

Allow interrupt controller drivers to register an arch specific function to
determine the acpi_handle for a specific GSI number to use the mechanism if
needed by the respective interrupt controller drivers.

Signed-off-by: Lorenzo Pieralisi <lpieralisi@kernel.org>
Cc: Huacai Chen <chenhuacai@kernel.org>
Cc: Thomas Gleixner <tglx@kernel.org>
Cc: Anup Patel <anup@brainfault.org>
Cc: "Rafael J. Wysocki" <rafael@kernel.org>
Cc: Sunil V L <sunilvl@ventanamicro.com>
Cc: Marc Zyngier <maz@kernel.org>
---
 arch/riscv/include/asm/acpi.h       |   1 +
 drivers/acpi/irq.c                  | 172 +++++++++++++++++++++++++++++++++++-
 drivers/acpi/riscv/irq.c            | 156 +-------------------------------
 drivers/irqchip/irq-gic-v3.c        |   2 +-
 drivers/irqchip/irq-gic-v5.c        |   2 +-
 drivers/irqchip/irq-gic.c           |   2 +-
 drivers/irqchip/irq-loongarch-cpu.c |   2 +-
 drivers/irqchip/irq-riscv-intc.c    |   3 +-
 include/linux/acpi.h                |   5 +-
 9 files changed, 181 insertions(+), 164 deletions(-)

diff --git a/arch/riscv/include/asm/acpi.h b/arch/riscv/include/asm/acpi.h
index 26ab37c171bc..f598520ac903 100644
--- a/arch/riscv/include/asm/acpi.h
+++ b/arch/riscv/include/asm/acpi.h
@@ -67,6 +67,7 @@ int acpi_get_riscv_isa(struct acpi_table_header *table,
 
 void acpi_get_cbo_block_size(struct acpi_table_header *table, u32 *cbom_size,
 			     u32 *cboz_size, u32 *cbop_size);
+acpi_handle acpi_get_riscv_gsi_handle(u32 gsi);
 #else
 static inline void acpi_init_rintc_map(void) { }
 static inline struct acpi_madt_rintc *acpi_cpu_get_madt_rintc(int cpu)
diff --git a/drivers/acpi/irq.c b/drivers/acpi/irq.c
index d1595156c86a..e4293458bf61 100644
--- a/drivers/acpi/irq.c
+++ b/drivers/acpi/irq.c
@@ -13,6 +13,7 @@
 enum acpi_irq_model_id acpi_irq_model;
 
 static acpi_gsi_domain_disp_fn acpi_get_gsi_domain_id;
+static acpi_gsi_handle_disp_fn acpi_get_gsi_handle;
 static u32 (*acpi_gsi_to_irq_fallback)(u32 gsi);
 
 /**
@@ -321,15 +322,19 @@ const struct cpumask *acpi_irq_get_affinity(acpi_handle handle,
 
 /**
  * acpi_set_irq_model - Setup the GSI irqdomain information
- * @model: the value assigned to acpi_irq_model
- * @fn: a dispatcher function that will return the domain fwnode
- *	for a given GSI
+ * @model:	the value assigned to acpi_irq_model
+ * @fn:		a dispatcher function that will return the domain fwnode
+ *		for a given GSI
+ * @gsi_dep_fn: a function to retrieve the acpi_handle a GSI interrupt is
+ *		dependent on
+ *
  */
 void __init acpi_set_irq_model(enum acpi_irq_model_id model,
-			       acpi_gsi_domain_disp_fn fn)
+			       acpi_gsi_domain_disp_fn fn, acpi_gsi_handle_disp_fn gsi_dep_fn)
 {
 	acpi_irq_model = model;
 	acpi_get_gsi_domain_id = fn;
+	acpi_get_gsi_handle = gsi_dep_fn;
 }
 
 /*
@@ -385,3 +390,162 @@ struct irq_domain *acpi_irq_create_hierarchy(unsigned int flags,
 					   host_data);
 }
 EXPORT_SYMBOL_GPL(acpi_irq_create_hierarchy);
+
+struct acpi_irq_dep_ctx {
+	int		rc;
+	unsigned int	index;
+	acpi_handle	handle;
+};
+
+static acpi_status acpi_irq_get_parent(struct acpi_resource *ares, void *context)
+{
+	struct acpi_irq_dep_ctx *ctx = context;
+	struct acpi_resource_irq *irq;
+	struct acpi_resource_extended_irq *eirq;
+
+	switch (ares->type) {
+	case ACPI_RESOURCE_TYPE_IRQ:
+		irq = &ares->data.irq;
+		if (ctx->index >= irq->interrupt_count) {
+			ctx->index -= irq->interrupt_count;
+			return AE_OK;
+		}
+		ctx->handle = acpi_get_gsi_handle(irq->interrupts[ctx->index]);
+		ctx->rc = 0;
+		return AE_CTRL_TERMINATE;
+	case ACPI_RESOURCE_TYPE_EXTENDED_IRQ:
+		eirq = &ares->data.extended_irq;
+		if (eirq->producer_consumer == ACPI_PRODUCER)
+			return AE_OK;
+
+		if (ctx->index >= eirq->interrupt_count) {
+			ctx->index -= eirq->interrupt_count;
+			return AE_OK;
+		}
+
+		/* Support GSIs only */
+		if (eirq->resource_source.string_length)
+			return AE_OK;
+
+		ctx->handle = acpi_get_gsi_handle(eirq->interrupts[ctx->index]);
+		ctx->rc = 0;
+		return AE_CTRL_TERMINATE;
+	}
+
+	return AE_OK;
+}
+
+static int acpi_irq_get_dep(acpi_handle handle, unsigned int index, acpi_handle *gsi_handle)
+{
+	struct acpi_irq_dep_ctx ctx = {-EINVAL, index, NULL};
+
+	if (!gsi_handle)
+		return -EINVAL;
+
+	acpi_walk_resources(handle, METHOD_NAME__CRS, acpi_irq_get_parent, &ctx);
+	*gsi_handle = ctx.handle;
+
+	return ctx.rc;
+}
+
+static bool acpi_prt_entry_valid(void *prt_entry)
+{
+	struct acpi_pci_routing_table *entry = prt_entry;
+
+	return entry && entry->length > 0;
+}
+
+static void *acpi_prt_next_entry(void *prt_entry)
+{
+	struct acpi_pci_routing_table *entry = prt_entry;
+
+	return prt_entry + entry->length;
+}
+
+static u32 acpi_add_prt_dep(acpi_handle handle)
+{
+	struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL };
+	struct acpi_pci_routing_table *entry;
+	struct acpi_handle_list dep_devices;
+	acpi_handle gsi_handle;
+	acpi_handle link_handle;
+	acpi_status status;
+	u32 count = 0;
+
+	status = acpi_get_irq_routing_table(handle, &buffer);
+	if (ACPI_FAILURE(status)) {
+		acpi_handle_err(handle, "failed to get IRQ routing table\n");
+		kfree(buffer.pointer);
+		return 0;
+	}
+
+	entry = buffer.pointer;
+	for (; acpi_prt_entry_valid(entry); entry = acpi_prt_next_entry(entry)) {
+		if (entry->source[0]) {
+			status = acpi_get_handle(handle, entry->source, &link_handle);
+			if (ACPI_FAILURE(status))
+				continue;
+			dep_devices.count = 1;
+			dep_devices.handles = kcalloc(1, sizeof(*dep_devices.handles), GFP_KERNEL);
+			if (!dep_devices.handles) {
+				acpi_handle_err(handle, "failed to allocate memory\n");
+				continue;
+			}
+
+			dep_devices.handles[0] = link_handle;
+			count += acpi_scan_add_dep(handle, &dep_devices);
+		} else {
+			gsi_handle = acpi_get_gsi_handle(entry->source_index);
+			if (!gsi_handle)
+				continue;
+			dep_devices.count = 1;
+			dep_devices.handles = kcalloc(1, sizeof(*dep_devices.handles), GFP_KERNEL);
+			if (!dep_devices.handles) {
+				acpi_handle_err(handle, "failed to allocate memory\n");
+				continue;
+			}
+
+			dep_devices.handles[0] = gsi_handle;
+			count += acpi_scan_add_dep(handle, &dep_devices);
+		}
+	}
+
+	kfree(buffer.pointer);
+	return count;
+}
+
+static u32 acpi_add_irq_dep(acpi_handle handle)
+{
+	struct acpi_handle_list dep_devices;
+	acpi_handle gsi_handle;
+	u32 count = 0;
+	int i;
+
+	for (i = 0; !acpi_irq_get_dep(handle, i, &gsi_handle); i++) {
+		if (!gsi_handle)
+			continue;
+
+		dep_devices.count = 1;
+		dep_devices.handles = kcalloc(1, sizeof(*dep_devices.handles), GFP_KERNEL);
+		if (!dep_devices.handles) {
+			acpi_handle_err(handle, "failed to allocate memory\n");
+			continue;
+		}
+
+		dep_devices.handles[0] = gsi_handle;
+		count += acpi_scan_add_dep(handle, &dep_devices);
+	}
+
+	return count;
+}
+
+u32 acpi_irq_add_auto_dep(acpi_handle handle)
+{
+	if (!acpi_get_gsi_handle)
+		return 0;
+
+	if (acpi_has_method(handle, "_PRT"))
+		return acpi_add_prt_dep(handle);
+
+	return acpi_add_irq_dep(handle);
+}
diff --git a/drivers/acpi/riscv/irq.c b/drivers/acpi/riscv/irq.c
index 0cdec5dd575e..da2c42e0ebfd 100644
--- a/drivers/acpi/riscv/irq.c
+++ b/drivers/acpi/riscv/irq.c
@@ -23,12 +23,6 @@ struct riscv_ext_intc_list {
 	struct list_head	list;
 };
 
-struct acpi_irq_dep_ctx {
-	int		rc;
-	unsigned int	index;
-	acpi_handle	handle;
-};
-
 LIST_HEAD(ext_intc_list);
 
 static int irqchip_cmp_func(const void *in0, const void *in1)
@@ -254,7 +248,7 @@ void __init riscv_acpi_init_gsi_mapping(void)
 	acpi_get_devices("RSCV0006", riscv_acpi_create_gsi_map_smsi, NULL, NULL);
 }
 
-static acpi_handle riscv_acpi_get_gsi_handle(u32 gsi)
+acpi_handle acpi_get_riscv_gsi_handle(u32 gsi)
 {
 	struct riscv_ext_intc_list *ext_intc_element;
 	struct list_head *i;
@@ -269,153 +263,7 @@ static acpi_handle riscv_acpi_get_gsi_handle(u32 gsi)
 	return NULL;
 }
 
-static acpi_status riscv_acpi_irq_get_parent(struct acpi_resource *ares, void *context)
-{
-	struct acpi_irq_dep_ctx *ctx = context;
-	struct acpi_resource_irq *irq;
-	struct acpi_resource_extended_irq *eirq;
-
-	switch (ares->type) {
-	case ACPI_RESOURCE_TYPE_IRQ:
-		irq = &ares->data.irq;
-		if (ctx->index >= irq->interrupt_count) {
-			ctx->index -= irq->interrupt_count;
-			return AE_OK;
-		}
-		ctx->handle = riscv_acpi_get_gsi_handle(irq->interrupts[ctx->index]);
-		return AE_CTRL_TERMINATE;
-	case ACPI_RESOURCE_TYPE_EXTENDED_IRQ:
-		eirq = &ares->data.extended_irq;
-		if (eirq->producer_consumer == ACPI_PRODUCER)
-			return AE_OK;
-
-		if (ctx->index >= eirq->interrupt_count) {
-			ctx->index -= eirq->interrupt_count;
-			return AE_OK;
-		}
-
-		/* Support GSIs only */
-		if (eirq->resource_source.string_length)
-			return AE_OK;
-
-		ctx->handle = riscv_acpi_get_gsi_handle(eirq->interrupts[ctx->index]);
-		ctx->rc = 0;
-		return AE_CTRL_TERMINATE;
-	}
-
-	return AE_OK;
-}
-
-static int riscv_acpi_irq_get_dep(acpi_handle handle, unsigned int index, acpi_handle *gsi_handle)
-{
-	struct acpi_irq_dep_ctx ctx = {-EINVAL, index, NULL};
-
-	if (!gsi_handle)
-		return 0;
-
-	acpi_walk_resources(handle, METHOD_NAME__CRS, riscv_acpi_irq_get_parent, &ctx);
-	*gsi_handle = ctx.handle;
-
-	return ctx.rc;
-}
-
-static bool acpi_prt_entry_valid(void *prt_entry)
-{
-	struct acpi_pci_routing_table *entry = prt_entry;
-
-	return entry && entry->length > 0;
-}
-
-static void *acpi_prt_next_entry(void *prt_entry)
-{
-	struct acpi_pci_routing_table *entry = prt_entry;
-
-	return prt_entry + entry->length;
-}
-
-static u32 riscv_acpi_add_prt_dep(acpi_handle handle)
-{
-	struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL };
-	struct acpi_pci_routing_table *entry;
-	struct acpi_handle_list dep_devices;
-	acpi_handle gsi_handle;
-	acpi_handle link_handle;
-	acpi_status status;
-	u32 count = 0;
-
-	status = acpi_get_irq_routing_table(handle, &buffer);
-	if (ACPI_FAILURE(status)) {
-		acpi_handle_err(handle, "failed to get IRQ routing table\n");
-		kfree(buffer.pointer);
-		return 0;
-	}
-
-	entry = buffer.pointer;
-	for (; acpi_prt_entry_valid(entry); entry = acpi_prt_next_entry(entry)) {
-		if (entry->source[0]) {
-			status = acpi_get_handle(handle, entry->source, &link_handle);
-			if (ACPI_FAILURE(status))
-				continue;
-			dep_devices.count = 1;
-			dep_devices.handles = kzalloc_objs(*dep_devices.handles,
-							   1);
-			if (!dep_devices.handles) {
-				acpi_handle_err(handle, "failed to allocate memory\n");
-				continue;
-			}
-
-			dep_devices.handles[0] = link_handle;
-			count += acpi_scan_add_dep(handle, &dep_devices);
-		} else {
-			gsi_handle = riscv_acpi_get_gsi_handle(entry->source_index);
-			dep_devices.count = 1;
-			dep_devices.handles = kzalloc_objs(*dep_devices.handles,
-							   1);
-			if (!dep_devices.handles) {
-				acpi_handle_err(handle, "failed to allocate memory\n");
-				continue;
-			}
-
-			dep_devices.handles[0] = gsi_handle;
-			count += acpi_scan_add_dep(handle, &dep_devices);
-		}
-	}
-
-	kfree(buffer.pointer);
-	return count;
-}
-
-static u32 riscv_acpi_add_irq_dep(acpi_handle handle)
-{
-	struct acpi_handle_list dep_devices;
-	acpi_handle gsi_handle;
-	u32 count = 0;
-	int i;
-
-	for (i = 0;
-	     !riscv_acpi_irq_get_dep(handle, i, &gsi_handle);
-	     i++) {
-		if (!gsi_handle)
-			continue;
-
-		dep_devices.count = 1;
-		dep_devices.handles = kzalloc_objs(*dep_devices.handles, 1);
-		if (!dep_devices.handles) {
-			acpi_handle_err(handle, "failed to allocate memory\n");
-			continue;
-		}
-
-		dep_devices.handles[0] = gsi_handle;
-		count += acpi_scan_add_dep(handle, &dep_devices);
-	}
-
-	return count;
-}
-
 u32 arch_acpi_add_auto_dep(acpi_handle handle)
 {
-	if (acpi_has_method(handle, "_PRT"))
-		return riscv_acpi_add_prt_dep(handle);
-
-	return riscv_acpi_add_irq_dep(handle);
+	return acpi_irq_add_auto_dep(handle);
 }
diff --git a/drivers/irqchip/irq-gic-v3.c b/drivers/irqchip/irq-gic-v3.c
index 99444a1b2ffa..2673954d4577 100644
--- a/drivers/irqchip/irq-gic-v3.c
+++ b/drivers/irqchip/irq-gic-v3.c
@@ -2588,7 +2588,7 @@ gic_acpi_init(union acpi_subtable_headers *header, const unsigned long end)
 	if (err)
 		goto out_fwhandle_free;
 
-	acpi_set_irq_model(ACPI_IRQ_MODEL_GIC, gic_v3_get_gsi_domain_id);
+	acpi_set_irq_model(ACPI_IRQ_MODEL_GIC, gic_v3_get_gsi_domain_id, NULL);
 
 	if (static_branch_likely(&supports_deactivate_key))
 		gic_acpi_setup_kvm_info();
diff --git a/drivers/irqchip/irq-gic-v5.c b/drivers/irqchip/irq-gic-v5.c
index 6b0903be8ebf..03cc2830b260 100644
--- a/drivers/irqchip/irq-gic-v5.c
+++ b/drivers/irqchip/irq-gic-v5.c
@@ -1242,7 +1242,7 @@ static int __init gic_acpi_init(union acpi_subtable_headers *header, const unsig
 	if (ret)
 		goto out_irs;
 
-	acpi_set_irq_model(ACPI_IRQ_MODEL_GIC_V5, gic_v5_get_gsi_domain_id);
+	acpi_set_irq_model(ACPI_IRQ_MODEL_GIC_V5, gic_v5_get_gsi_domain_id, NULL);
 
 	return 0;
 
diff --git a/drivers/irqchip/irq-gic.c b/drivers/irqchip/irq-gic.c
index ec70c84e9f91..f6bc29f515fb 100644
--- a/drivers/irqchip/irq-gic.c
+++ b/drivers/irqchip/irq-gic.c
@@ -1690,7 +1690,7 @@ static int __init gic_v2_acpi_init(union acpi_subtable_headers *header,
 		return ret;
 	}
 
-	acpi_set_irq_model(ACPI_IRQ_MODEL_GIC, gic_v2_get_gsi_domain_id);
+	acpi_set_irq_model(ACPI_IRQ_MODEL_GIC, gic_v2_get_gsi_domain_id, NULL);
 
 	if (IS_ENABLED(CONFIG_ARM_GIC_V2M))
 		gicv2m_init(NULL, gic_data[0].domain);
diff --git a/drivers/irqchip/irq-loongarch-cpu.c b/drivers/irqchip/irq-loongarch-cpu.c
index 950bc087e388..84ce24889488 100644
--- a/drivers/irqchip/irq-loongarch-cpu.c
+++ b/drivers/irqchip/irq-loongarch-cpu.c
@@ -168,7 +168,7 @@ static int __init cpuintc_acpi_init(union acpi_subtable_headers *header,
 		panic("Failed to add irqdomain for LoongArch CPU");
 
 	set_handle_irq(&handle_cpu_irq);
-	acpi_set_irq_model(ACPI_IRQ_MODEL_LPIC, lpic_get_gsi_domain_id);
+	acpi_set_irq_model(ACPI_IRQ_MODEL_LPIC, lpic_get_gsi_domain_id, NULL);
 	acpi_set_gsi_to_irq_fallback(lpic_gsi_to_irq);
 	ret = acpi_cascade_irqdomain_init();
 
diff --git a/drivers/irqchip/irq-riscv-intc.c b/drivers/irqchip/irq-riscv-intc.c
index 84418dbd5a27..0595144116e2 100644
--- a/drivers/irqchip/irq-riscv-intc.c
+++ b/drivers/irqchip/irq-riscv-intc.c
@@ -384,7 +384,8 @@ static int __init riscv_intc_acpi_init(union acpi_subtable_headers *header,
 	if (rc)
 		irq_domain_free_fwnode(fn);
 	else
-		acpi_set_irq_model(ACPI_IRQ_MODEL_RINTC, riscv_acpi_get_gsi_domain_id);
+		acpi_set_irq_model(ACPI_IRQ_MODEL_RINTC, riscv_acpi_get_gsi_domain_id,
+				   acpi_get_riscv_gsi_handle);
 
 	return rc;
 }
diff --git a/include/linux/acpi.h b/include/linux/acpi.h
index 67effb91fa98..468fc6a54651 100644
--- a/include/linux/acpi.h
+++ b/include/linux/acpi.h
@@ -360,9 +360,10 @@ int acpi_gsi_to_irq (u32 gsi, unsigned int *irq);
 int acpi_isa_irq_to_gsi (unsigned isa_irq, u32 *gsi);
 
 typedef struct fwnode_handle *(*acpi_gsi_domain_disp_fn)(u32);
+typedef acpi_handle (*acpi_gsi_handle_disp_fn)(u32);
 
 void acpi_set_irq_model(enum acpi_irq_model_id model,
-			acpi_gsi_domain_disp_fn fn);
+			acpi_gsi_domain_disp_fn fn, acpi_gsi_handle_disp_fn gsi_dep_fn);
 acpi_gsi_domain_disp_fn acpi_get_gsi_dispatcher(void);
 void acpi_set_gsi_to_irq_fallback(u32 (*)(u32));
 
@@ -372,6 +373,8 @@ struct irq_domain *acpi_irq_create_hierarchy(unsigned int flags,
 					     const struct irq_domain_ops *ops,
 					     void *host_data);
 
+u32 acpi_irq_add_auto_dep(acpi_handle handle);
+
 #ifdef CONFIG_X86_IO_APIC
 extern int acpi_get_override_irq(u32 gsi, int *trigger, int *polarity);
 #else

-- 
2.54.0

Re: [PATCH v2 4/6] ACPI: irq: Move RISC-V interrupt controllers autodep to ACPI IRQ code

[Sunil V L]

On Wed, Jun 3, 2026 at 1:54 PM Lorenzo Pieralisi <lpieralisi@kernel.org> wrote:
>
> RISC-V implements arch code to detect probe dependencies for devices and
> the interrupt controller the devices GSIs are routed to.
>
> The code itself is arch agnostic apart from an arch specific helper
> function required to retrieve the acpi_handle of the interrupt controller
> that manages the device GSI interrupt.
>
> In order to enable IRQ probe dependencies detection on other architectures,
> move RISC-V IRQ probe dependency detection code to generic ACPI IRQ code.
>
> Allow interrupt controller drivers to register an arch specific function to
> determine the acpi_handle for a specific GSI number to use the mechanism if
> needed by the respective interrupt controller drivers.
>
> Signed-off-by: Lorenzo Pieralisi <lpieralisi@kernel.org>
> Cc: Huacai Chen <chenhuacai@kernel.org>
> Cc: Thomas Gleixner <tglx@kernel.org>
> Cc: Anup Patel <anup@brainfault.org>
> Cc: "Rafael J. Wysocki" <rafael@kernel.org>
> Cc: Sunil V L <sunilvl@ventanamicro.com>
> Cc: Marc Zyngier <maz@kernel.org>
> ---
>  arch/riscv/include/asm/acpi.h       |   1 +
>  drivers/acpi/irq.c                  | 172 +++++++++++++++++++++++++++++++++++-
>  drivers/acpi/riscv/irq.c            | 156 +-------------------------------
>  drivers/irqchip/irq-gic-v3.c        |   2 +-
>  drivers/irqchip/irq-gic-v5.c        |   2 +-
>  drivers/irqchip/irq-gic.c           |   2 +-
>  drivers/irqchip/irq-loongarch-cpu.c |   2 +-
>  drivers/irqchip/irq-riscv-intc.c    |   3 +-
>  include/linux/acpi.h                |   5 +-
>  9 files changed, 181 insertions(+), 164 deletions(-)
>
LGTM.

Reviewed-by: Sunil V L <sunilvl@oss.qualcomm.com>

[PATCH v2 5/6] ACPI/IORT: Implement ACPI infrastructure to enable GICv5 IWB probe deferral

[Lorenzo Pieralisi]

Implement an IORT ACPI hook to retrieve the acpi_handle of the interrupt
controller handling a specific GSI (if any, on GICv5 systems only the IWB
is represented in firmware with an ACPI device object) and add the IWB to
the list of devices whose dependencies can be detected (and cleared) in
ACPI core to guarantee that probe dependencies for the IWB can be
satisfied.

Enable autodep detection for arm64 by adding the arch_acpi_add_auto_dep()
callback in the ACPI IORT driver.

Signed-off-by: Lorenzo Pieralisi <lpieralisi@kernel.org>
Cc: Hanjun Guo <guohanjun@huawei.com>
Cc: Sudeep Holla <sudeep.holla@kernel.org>
Cc: "Rafael J. Wysocki" <rafael@kernel.org>
---
 drivers/acpi/arm64/iort.c | 22 +++++++++++++++++++---
 drivers/acpi/scan.c       |  1 +
 include/linux/acpi_iort.h |  3 ++-
 3 files changed, 22 insertions(+), 4 deletions(-)

diff --git a/drivers/acpi/arm64/iort.c b/drivers/acpi/arm64/iort.c
index af7a9b2fd5bc..34412cd697d8 100644
--- a/drivers/acpi/arm64/iort.c
+++ b/drivers/acpi/arm64/iort.c
@@ -789,11 +789,9 @@ struct irq_domain *iort_get_device_domain(struct device *dev, u32 id,
 	return irq_find_matching_fwnode(handle, bus_token);
 }
 
-struct fwnode_handle *iort_iwb_handle(u32 iwb_id)
+acpi_handle iort_iwb_handle(u32 iwb_id)
 {
-	struct fwnode_handle *fwnode;
 	struct acpi_iort_node *node;
-	struct acpi_device *device;
 	struct acpi_iort_iwb *iwb;
 	acpi_status status;
 	acpi_handle handle;
@@ -808,6 +806,19 @@ struct fwnode_handle *iort_iwb_handle(u32 iwb_id)
 	if (ACPI_FAILURE(status))
 		return NULL;
 
+	return handle;
+}
+
+struct fwnode_handle *iort_iwb_handle_fwnode(u32 iwb_id)
+{
+	struct fwnode_handle *fwnode;
+	struct acpi_device *device;
+	acpi_handle handle;
+
+	handle = iort_iwb_handle(iwb_id);
+	if (!handle)
+		return NULL;
+
 	device = acpi_get_acpi_dev(handle);
 	if (!device)
 		return NULL;
@@ -2090,6 +2101,11 @@ static void __init iort_init_platform_devices(void)
 	}
 }
 
+u32 arch_acpi_add_auto_dep(acpi_handle handle)
+{
+	return acpi_irq_add_auto_dep(handle);
+}
+
 void __init acpi_iort_init(void)
 {
 	acpi_status status;
diff --git a/drivers/acpi/scan.c b/drivers/acpi/scan.c
index 530547cda8b2..ba15da9058dc 100644
--- a/drivers/acpi/scan.c
+++ b/drivers/acpi/scan.c
@@ -856,6 +856,7 @@ static const char * const acpi_ignore_dep_ids[] = {
 
 /* List of HIDs for which we honor deps of matching ACPI devs, when checking _DEP lists. */
 static const char * const acpi_honor_dep_ids[] = {
+	"ARMH0003", /* ARM GICv5 IWB */
 	"INT3472", /* Camera sensor PMIC / clk and regulator info */
 	"INTC1059", /* IVSC (TGL) driver must be loaded to allow i2c access to camera sensors */
 	"INTC1095", /* IVSC (ADL) driver must be loaded to allow i2c access to camera sensors */
diff --git a/include/linux/acpi_iort.h b/include/linux/acpi_iort.h
index 17bb3374f4ca..931eaa7bbf6a 100644
--- a/include/linux/acpi_iort.h
+++ b/include/linux/acpi_iort.h
@@ -27,7 +27,8 @@ int iort_register_domain_token(int trans_id, phys_addr_t base,
 			       struct fwnode_handle *fw_node);
 void iort_deregister_domain_token(int trans_id);
 struct fwnode_handle *iort_find_domain_token(int trans_id);
-struct fwnode_handle *iort_iwb_handle(u32 iwb_id);
+acpi_handle iort_iwb_handle(u32 iwb_id);
+struct fwnode_handle *iort_iwb_handle_fwnode(u32 iwb_id);
 
 #ifdef CONFIG_ACPI_IORT
 u32 iort_msi_map_id(struct device *dev, u32 id);

-- 
2.54.0

Re: [PATCH v2 5/6] ACPI/IORT: Implement ACPI infrastructure to enable GICv5 IWB probe deferral

[Rafael J. Wysocki]

On Wed, Jun 3, 2026 at 10:21 AM Lorenzo Pieralisi <lpieralisi@kernel.org> wrote:
>
> Implement an IORT ACPI hook to retrieve the acpi_handle of the interrupt
> controller handling a specific GSI (if any, on GICv5 systems only the IWB
> is represented in firmware with an ACPI device object) and add the IWB to
> the list of devices whose dependencies can be detected (and cleared) in
> ACPI core to guarantee that probe dependencies for the IWB can be
> satisfied.
>
> Enable autodep detection for arm64 by adding the arch_acpi_add_auto_dep()
> callback in the ACPI IORT driver.
>
> Signed-off-by: Lorenzo Pieralisi <lpieralisi@kernel.org>
> Cc: Hanjun Guo <guohanjun@huawei.com>
> Cc: Sudeep Holla <sudeep.holla@kernel.org>
> Cc: "Rafael J. Wysocki" <rafael@kernel.org>

Acked-by: "Rafael J. Wysocki" (Intel) <rafael@kernel.org> # ACPI: scan.c

The rest of the series is kind of outside my scope and all of the
previous patches would at least require an ACK from the RISC-V people.

> ---
>  drivers/acpi/arm64/iort.c | 22 +++++++++++++++++++---
>  drivers/acpi/scan.c       |  1 +
>  include/linux/acpi_iort.h |  3 ++-
>  3 files changed, 22 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/acpi/arm64/iort.c b/drivers/acpi/arm64/iort.c
> index af7a9b2fd5bc..34412cd697d8 100644
> --- a/drivers/acpi/arm64/iort.c
> +++ b/drivers/acpi/arm64/iort.c
> @@ -789,11 +789,9 @@ struct irq_domain *iort_get_device_domain(struct device *dev, u32 id,
>         return irq_find_matching_fwnode(handle, bus_token);
>  }
>
> -struct fwnode_handle *iort_iwb_handle(u32 iwb_id)
> +acpi_handle iort_iwb_handle(u32 iwb_id)
>  {
> -       struct fwnode_handle *fwnode;
>         struct acpi_iort_node *node;
> -       struct acpi_device *device;
>         struct acpi_iort_iwb *iwb;
>         acpi_status status;
>         acpi_handle handle;
> @@ -808,6 +806,19 @@ struct fwnode_handle *iort_iwb_handle(u32 iwb_id)
>         if (ACPI_FAILURE(status))
>                 return NULL;
>
> +       return handle;
> +}
> +
> +struct fwnode_handle *iort_iwb_handle_fwnode(u32 iwb_id)
> +{
> +       struct fwnode_handle *fwnode;
> +       struct acpi_device *device;
> +       acpi_handle handle;
> +
> +       handle = iort_iwb_handle(iwb_id);
> +       if (!handle)
> +               return NULL;
> +
>         device = acpi_get_acpi_dev(handle);
>         if (!device)
>                 return NULL;
> @@ -2090,6 +2101,11 @@ static void __init iort_init_platform_devices(void)
>         }
>  }
>
> +u32 arch_acpi_add_auto_dep(acpi_handle handle)
> +{
> +       return acpi_irq_add_auto_dep(handle);
> +}
> +
>  void __init acpi_iort_init(void)
>  {
>         acpi_status status;
> diff --git a/drivers/acpi/scan.c b/drivers/acpi/scan.c
> index 530547cda8b2..ba15da9058dc 100644
> --- a/drivers/acpi/scan.c
> +++ b/drivers/acpi/scan.c
> @@ -856,6 +856,7 @@ static const char * const acpi_ignore_dep_ids[] = {
>
>  /* List of HIDs for which we honor deps of matching ACPI devs, when checking _DEP lists. */
>  static const char * const acpi_honor_dep_ids[] = {
> +       "ARMH0003", /* ARM GICv5 IWB */
>         "INT3472", /* Camera sensor PMIC / clk and regulator info */
>         "INTC1059", /* IVSC (TGL) driver must be loaded to allow i2c access to camera sensors */
>         "INTC1095", /* IVSC (ADL) driver must be loaded to allow i2c access to camera sensors */
> diff --git a/include/linux/acpi_iort.h b/include/linux/acpi_iort.h
> index 17bb3374f4ca..931eaa7bbf6a 100644
> --- a/include/linux/acpi_iort.h
> +++ b/include/linux/acpi_iort.h
> @@ -27,7 +27,8 @@ int iort_register_domain_token(int trans_id, phys_addr_t base,
>                                struct fwnode_handle *fw_node);
>  void iort_deregister_domain_token(int trans_id);
>  struct fwnode_handle *iort_find_domain_token(int trans_id);
> -struct fwnode_handle *iort_iwb_handle(u32 iwb_id);
> +acpi_handle iort_iwb_handle(u32 iwb_id);
> +struct fwnode_handle *iort_iwb_handle_fwnode(u32 iwb_id);
>
>  #ifdef CONFIG_ACPI_IORT
>  u32 iort_msi_map_id(struct device *dev, u32 id);
>
> --
> 2.54.0
>

[PATCH v2 6/6] irqchip/gic-v5: Enable GICv5 IWB ACPI probe ordering detection

[Lorenzo Pieralisi]

Register an ACPI hook in the ACPI interrupt management code for GICv5 to
retrieve the ACPI interrupt controller handle (if any) of the controller
handling a specific GSI, by updating the acpi_set_irq_model() call with
the gic_v5_get_gsi_handle() function pointer parameter.

gicv5_get_gsi_handle() allows ACPI core to detect the ACPI handle
of the controller that manages a specific GSI interrupt.

Update the IWB driver to clear device dependencies in ACPI core once the
IWB driver has probed.

Signed-off-by: Lorenzo Pieralisi <lpieralisi@kernel.org>
Cc: Thomas Gleixner <tglx@kernel.org>
Cc: Marc Zyngier <maz@kernel.org>
---
 drivers/irqchip/irq-gic-v5-iwb.c |  5 +++++
 drivers/irqchip/irq-gic-v5.c     | 13 +++++++++++--
 2 files changed, 16 insertions(+), 2 deletions(-)

diff --git a/drivers/irqchip/irq-gic-v5-iwb.c b/drivers/irqchip/irq-gic-v5-iwb.c
index 9103feb70ce8..a02cb9537b15 100644
--- a/drivers/irqchip/irq-gic-v5-iwb.c
+++ b/drivers/irqchip/irq-gic-v5-iwb.c
@@ -269,6 +269,11 @@ static int gicv5_iwb_device_probe(struct platform_device *pdev)
 	if (IS_ERR(iwb_node))
 		return PTR_ERR(iwb_node);
 
+#ifdef CONFIG_ACPI
+	if (has_acpi_companion(&pdev->dev))
+		acpi_dev_clear_dependencies(ACPI_COMPANION(&pdev->dev));
+#endif
+
 	return 0;
 }
 
diff --git a/drivers/irqchip/irq-gic-v5.c b/drivers/irqchip/irq-gic-v5.c
index 03cc2830b260..26cfaea1af41 100644
--- a/drivers/irqchip/irq-gic-v5.c
+++ b/drivers/irqchip/irq-gic-v5.c
@@ -1217,11 +1217,19 @@ static struct fwnode_handle *gsi_domain_handle;
 static struct fwnode_handle *gic_v5_get_gsi_domain_id(u32 gsi)
 {
 	if (FIELD_GET(GICV5_GSI_IC_TYPE, gsi) == GICV5_GSI_IWB_TYPE)
-		return iort_iwb_handle(FIELD_GET(GICV5_GSI_IWB_FRAME_ID, gsi));
+		return iort_iwb_handle_fwnode(FIELD_GET(GICV5_GSI_IWB_FRAME_ID, gsi));
 
 	return gsi_domain_handle;
 }
 
+static acpi_handle gic_v5_get_gsi_handle(u32 gsi)
+{
+	if (FIELD_GET(GICV5_GSI_IC_TYPE, gsi) == GICV5_GSI_IWB_TYPE)
+		return iort_iwb_handle(FIELD_GET(GICV5_GSI_IWB_FRAME_ID, gsi));
+
+	return NULL;
+}
+
 static int __init gic_acpi_init(union acpi_subtable_headers *header, const unsigned long end)
 {
 	struct acpi_madt_gicv5_irs *irs = (struct acpi_madt_gicv5_irs *)header;
@@ -1242,7 +1250,8 @@ static int __init gic_acpi_init(union acpi_subtable_headers *header, const unsig
 	if (ret)
 		goto out_irs;
 
-	acpi_set_irq_model(ACPI_IRQ_MODEL_GIC_V5, gic_v5_get_gsi_domain_id, NULL);
+	acpi_set_irq_model(ACPI_IRQ_MODEL_GIC_V5, gic_v5_get_gsi_domain_id,
+			   gic_v5_get_gsi_handle);
 
 	return 0;
 

-- 
2.54.0

Re: [PATCH v2 6/6] irqchip/gic-v5: Enable GICv5 IWB ACPI probe ordering detection

[Rafael J. Wysocki]

On Wed, Jun 3, 2026 at 10:21 AM Lorenzo Pieralisi <lpieralisi@kernel.org> wrote:
>
> Register an ACPI hook in the ACPI interrupt management code for GICv5 to
> retrieve the ACPI interrupt controller handle (if any) of the controller
> handling a specific GSI, by updating the acpi_set_irq_model() call with
> the gic_v5_get_gsi_handle() function pointer parameter.
>
> gicv5_get_gsi_handle() allows ACPI core to detect the ACPI handle
> of the controller that manages a specific GSI interrupt.
>
> Update the IWB driver to clear device dependencies in ACPI core once the
> IWB driver has probed.
>
> Signed-off-by: Lorenzo Pieralisi <lpieralisi@kernel.org>
> Cc: Thomas Gleixner <tglx@kernel.org>
> Cc: Marc Zyngier <maz@kernel.org>
> ---
>  drivers/irqchip/irq-gic-v5-iwb.c |  5 +++++
>  drivers/irqchip/irq-gic-v5.c     | 13 +++++++++++--
>  2 files changed, 16 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/irqchip/irq-gic-v5-iwb.c b/drivers/irqchip/irq-gic-v5-iwb.c
> index 9103feb70ce8..a02cb9537b15 100644
> --- a/drivers/irqchip/irq-gic-v5-iwb.c
> +++ b/drivers/irqchip/irq-gic-v5-iwb.c
> @@ -269,6 +269,11 @@ static int gicv5_iwb_device_probe(struct platform_device *pdev)
>         if (IS_ERR(iwb_node))
>                 return PTR_ERR(iwb_node);
>
> +#ifdef CONFIG_ACPI
> +       if (has_acpi_companion(&pdev->dev))
> +               acpi_dev_clear_dependencies(ACPI_COMPANION(&pdev->dev));
> +#endif

I would rather add a wrapper for this, along with an empty stub for
the !CONFIG_ACPI case.

> +
>         return 0;
>  }
>
> diff --git a/drivers/irqchip/irq-gic-v5.c b/drivers/irqchip/irq-gic-v5.c
> index 03cc2830b260..26cfaea1af41 100644
> --- a/drivers/irqchip/irq-gic-v5.c
> +++ b/drivers/irqchip/irq-gic-v5.c
> @@ -1217,11 +1217,19 @@ static struct fwnode_handle *gsi_domain_handle;
>  static struct fwnode_handle *gic_v5_get_gsi_domain_id(u32 gsi)
>  {
>         if (FIELD_GET(GICV5_GSI_IC_TYPE, gsi) == GICV5_GSI_IWB_TYPE)
> -               return iort_iwb_handle(FIELD_GET(GICV5_GSI_IWB_FRAME_ID, gsi));
> +               return iort_iwb_handle_fwnode(FIELD_GET(GICV5_GSI_IWB_FRAME_ID, gsi));

Why is this change needed?

>
>         return gsi_domain_handle;
>  }
>
> +static acpi_handle gic_v5_get_gsi_handle(u32 gsi)
> +{
> +       if (FIELD_GET(GICV5_GSI_IC_TYPE, gsi) == GICV5_GSI_IWB_TYPE)
> +               return iort_iwb_handle(FIELD_GET(GICV5_GSI_IWB_FRAME_ID, gsi));
> +
> +       return NULL;
> +}
> +
>  static int __init gic_acpi_init(union acpi_subtable_headers *header, const unsigned long end)
>  {
>         struct acpi_madt_gicv5_irs *irs = (struct acpi_madt_gicv5_irs *)header;
> @@ -1242,7 +1250,8 @@ static int __init gic_acpi_init(union acpi_subtable_headers *header, const unsig
>         if (ret)
>                 goto out_irs;
>
> -       acpi_set_irq_model(ACPI_IRQ_MODEL_GIC_V5, gic_v5_get_gsi_domain_id, NULL);
> +       acpi_set_irq_model(ACPI_IRQ_MODEL_GIC_V5, gic_v5_get_gsi_domain_id,
> +                          gic_v5_get_gsi_handle);
>
>         return 0;
>
>
> --

Re: [PATCH v2 6/6] irqchip/gic-v5: Enable GICv5 IWB ACPI probe ordering detection

[Lorenzo Pieralisi]

On Mon, Jun 08, 2026 at 07:18:15PM +0200, Rafael J. Wysocki wrote:
> On Wed, Jun 3, 2026 at 10:21 AM Lorenzo Pieralisi <lpieralisi@kernel.org> wrote:
> >
> > Register an ACPI hook in the ACPI interrupt management code for GICv5 to
> > retrieve the ACPI interrupt controller handle (if any) of the controller
> > handling a specific GSI, by updating the acpi_set_irq_model() call with
> > the gic_v5_get_gsi_handle() function pointer parameter.
> >
> > gicv5_get_gsi_handle() allows ACPI core to detect the ACPI handle
> > of the controller that manages a specific GSI interrupt.
> >
> > Update the IWB driver to clear device dependencies in ACPI core once the
> > IWB driver has probed.
> >
> > Signed-off-by: Lorenzo Pieralisi <lpieralisi@kernel.org>
> > Cc: Thomas Gleixner <tglx@kernel.org>
> > Cc: Marc Zyngier <maz@kernel.org>
> > ---
> >  drivers/irqchip/irq-gic-v5-iwb.c |  5 +++++
> >  drivers/irqchip/irq-gic-v5.c     | 13 +++++++++++--
> >  2 files changed, 16 insertions(+), 2 deletions(-)
> >
> > diff --git a/drivers/irqchip/irq-gic-v5-iwb.c b/drivers/irqchip/irq-gic-v5-iwb.c
> > index 9103feb70ce8..a02cb9537b15 100644
> > --- a/drivers/irqchip/irq-gic-v5-iwb.c
> > +++ b/drivers/irqchip/irq-gic-v5-iwb.c
> > @@ -269,6 +269,11 @@ static int gicv5_iwb_device_probe(struct platform_device *pdev)
> >         if (IS_ERR(iwb_node))
> >                 return PTR_ERR(iwb_node);
> >
> > +#ifdef CONFIG_ACPI
> > +       if (has_acpi_companion(&pdev->dev))
> > +               acpi_dev_clear_dependencies(ACPI_COMPANION(&pdev->dev));
> > +#endif
> 
> I would rather add a wrapper for this, along with an empty stub for
> the !CONFIG_ACPI case.

Ok, I will.

> > +
> >         return 0;
> >  }
> >
> > diff --git a/drivers/irqchip/irq-gic-v5.c b/drivers/irqchip/irq-gic-v5.c
> > index 03cc2830b260..26cfaea1af41 100644
> > --- a/drivers/irqchip/irq-gic-v5.c
> > +++ b/drivers/irqchip/irq-gic-v5.c
> > @@ -1217,11 +1217,19 @@ static struct fwnode_handle *gsi_domain_handle;
> >  static struct fwnode_handle *gic_v5_get_gsi_domain_id(u32 gsi)
> >  {
> >         if (FIELD_GET(GICV5_GSI_IC_TYPE, gsi) == GICV5_GSI_IWB_TYPE)
> > -               return iort_iwb_handle(FIELD_GET(GICV5_GSI_IWB_FRAME_ID, gsi));
> > +               return iort_iwb_handle_fwnode(FIELD_GET(GICV5_GSI_IWB_FRAME_ID, gsi));
> 
> Why is this change needed?

This is a mistake, thanks for spotting it, it belongs in the previous patch
(IMO patch 5 and 6 are a single logical entity, I split it to try to keep
irqchip specific changes in one patch).

iort_iwb_handle() was refactored to return an acpi_handle not a fwnode_handle
(in preparation for adding gic_v5_get_gsi_handle() below - the dependency
chain requires acpi_handle retrieval not fwnode_handle), so the change above
belongs in patch 5, I will move this line change to patch 5 in v3.

Thanks,
Lorenzo

> >
> >         return gsi_domain_handle;
> >  }
> >
> > +static acpi_handle gic_v5_get_gsi_handle(u32 gsi)
> > +{
> > +       if (FIELD_GET(GICV5_GSI_IC_TYPE, gsi) == GICV5_GSI_IWB_TYPE)
> > +               return iort_iwb_handle(FIELD_GET(GICV5_GSI_IWB_FRAME_ID, gsi));
> > +
> > +       return NULL;
> > +}
> > +
> >  static int __init gic_acpi_init(union acpi_subtable_headers *header, const unsigned long end)
> >  {
> >         struct acpi_madt_gicv5_irs *irs = (struct acpi_madt_gicv5_irs *)header;
> > @@ -1242,7 +1250,8 @@ static int __init gic_acpi_init(union acpi_subtable_headers *header, const unsig
> >         if (ret)
> >                 goto out_irs;
> >
> > -       acpi_set_irq_model(ACPI_IRQ_MODEL_GIC_V5, gic_v5_get_gsi_domain_id, NULL);
> > +       acpi_set_irq_model(ACPI_IRQ_MODEL_GIC_V5, gic_v5_get_gsi_domain_id,
> > +                          gic_v5_get_gsi_handle);
> >
> >         return 0;
> >
> >
> > --

Re: [PATCH v2 6/6] irqchip/gic-v5: Enable GICv5 IWB ACPI probe ordering detection

[Lorenzo Pieralisi]

On Mon, Jun 08, 2026 at 07:18:15PM +0200, Rafael J. Wysocki wrote:
> On Wed, Jun 3, 2026 at 10:21 AM Lorenzo Pieralisi <lpieralisi@kernel.org> wrote:
> >
> > Register an ACPI hook in the ACPI interrupt management code for GICv5 to
> > retrieve the ACPI interrupt controller handle (if any) of the controller
> > handling a specific GSI, by updating the acpi_set_irq_model() call with
> > the gic_v5_get_gsi_handle() function pointer parameter.
> >
> > gicv5_get_gsi_handle() allows ACPI core to detect the ACPI handle
> > of the controller that manages a specific GSI interrupt.
> >
> > Update the IWB driver to clear device dependencies in ACPI core once the
> > IWB driver has probed.
> >
> > Signed-off-by: Lorenzo Pieralisi <lpieralisi@kernel.org>
> > Cc: Thomas Gleixner <tglx@kernel.org>
> > Cc: Marc Zyngier <maz@kernel.org>
> > ---
> >  drivers/irqchip/irq-gic-v5-iwb.c |  5 +++++
> >  drivers/irqchip/irq-gic-v5.c     | 13 +++++++++++--
> >  2 files changed, 16 insertions(+), 2 deletions(-)
> >
> > diff --git a/drivers/irqchip/irq-gic-v5-iwb.c b/drivers/irqchip/irq-gic-v5-iwb.c
> > index 9103feb70ce8..a02cb9537b15 100644
> > --- a/drivers/irqchip/irq-gic-v5-iwb.c
> > +++ b/drivers/irqchip/irq-gic-v5-iwb.c
> > @@ -269,6 +269,11 @@ static int gicv5_iwb_device_probe(struct platform_device *pdev)
> >         if (IS_ERR(iwb_node))
> >                 return PTR_ERR(iwb_node);
> >
> > +#ifdef CONFIG_ACPI
> > +       if (has_acpi_companion(&pdev->dev))
> > +               acpi_dev_clear_dependencies(ACPI_COMPANION(&pdev->dev));
> > +#endif
> 
> I would rather add a wrapper for this, along with an empty stub for
> the !CONFIG_ACPI case.

Given that this has no OF counterpart, I'd rather add an empty stub
for acpi_dev_clear_dependencies() and remove the ifdef - I am not
sure that adding a FW agnostic:

dev_clear_dependencies()

would help much but that's just my opinion. Then the question is whether I
should convert all other acpi_dev_clear_dependencies() users (with ifdeffery
included) in tree.

Thanks,
Lorenzo

> 
> > +
> >         return 0;
> >  }
> >
> > diff --git a/drivers/irqchip/irq-gic-v5.c b/drivers/irqchip/irq-gic-v5.c
> > index 03cc2830b260..26cfaea1af41 100644
> > --- a/drivers/irqchip/irq-gic-v5.c
> > +++ b/drivers/irqchip/irq-gic-v5.c
> > @@ -1217,11 +1217,19 @@ static struct fwnode_handle *gsi_domain_handle;
> >  static struct fwnode_handle *gic_v5_get_gsi_domain_id(u32 gsi)
> >  {
> >         if (FIELD_GET(GICV5_GSI_IC_TYPE, gsi) == GICV5_GSI_IWB_TYPE)
> > -               return iort_iwb_handle(FIELD_GET(GICV5_GSI_IWB_FRAME_ID, gsi));
> > +               return iort_iwb_handle_fwnode(FIELD_GET(GICV5_GSI_IWB_FRAME_ID, gsi));
> 
> Why is this change needed?
> 
> >
> >         return gsi_domain_handle;
> >  }
> >
> > +static acpi_handle gic_v5_get_gsi_handle(u32 gsi)
> > +{
> > +       if (FIELD_GET(GICV5_GSI_IC_TYPE, gsi) == GICV5_GSI_IWB_TYPE)
> > +               return iort_iwb_handle(FIELD_GET(GICV5_GSI_IWB_FRAME_ID, gsi));
> > +
> > +       return NULL;
> > +}
> > +
> >  static int __init gic_acpi_init(union acpi_subtable_headers *header, const unsigned long end)
> >  {
> >         struct acpi_madt_gicv5_irs *irs = (struct acpi_madt_gicv5_irs *)header;
> > @@ -1242,7 +1250,8 @@ static int __init gic_acpi_init(union acpi_subtable_headers *header, const unsig
> >         if (ret)
> >                 goto out_irs;
> >
> > -       acpi_set_irq_model(ACPI_IRQ_MODEL_GIC_V5, gic_v5_get_gsi_domain_id, NULL);
> > +       acpi_set_irq_model(ACPI_IRQ_MODEL_GIC_V5, gic_v5_get_gsi_domain_id,
> > +                          gic_v5_get_gsi_handle);
> >
> >         return 0;
> >
> >
> > --

Re: [PATCH v2 6/6] irqchip/gic-v5: Enable GICv5 IWB ACPI probe ordering detection

[Lorenzo Pieralisi]

On Mon, Jun 08, 2026 at 07:18:15PM +0200, Rafael J. Wysocki wrote:

[...]

> > +#ifdef CONFIG_ACPI
> > +       if (has_acpi_companion(&pdev->dev))
> > +               acpi_dev_clear_dependencies(ACPI_COMPANION(&pdev->dev));
> > +#endif
> 
> I would rather add a wrapper for this, along with an empty stub for
> the !CONFIG_ACPI case.

#ifdef CONFIG_ACPI
static inline void acpi_device_clear_dep(struct device *dev)
{
	if (has_acpi_companion(dev))
		acpi_dev_clear_dependencies(ACPI_COMPANION(dev));

}
#else
static inline void acpi_device_clear_dep(struct device *dev) {}
#endif

Something like that ?

Thanks,
Lorenzo

Re: [PATCH v2 6/6] irqchip/gic-v5: Enable GICv5 IWB ACPI probe ordering detection

[Rafael J. Wysocki (Intel)]

On Tue, Jun 30, 2026 at 12:51 PM Lorenzo Pieralisi
<lpieralisi@kernel.org> wrote:
>
> On Mon, Jun 08, 2026 at 07:18:15PM +0200, Rafael J. Wysocki wrote:
>
> [...]
>
> > > +#ifdef CONFIG_ACPI
> > > +       if (has_acpi_companion(&pdev->dev))
> > > +               acpi_dev_clear_dependencies(ACPI_COMPANION(&pdev->dev));
> > > +#endif
> >
> > I would rather add a wrapper for this, along with an empty stub for
> > the !CONFIG_ACPI case.
>
> #ifdef CONFIG_ACPI
> static inline void acpi_device_clear_dep(struct device *dev)
> {
>         if (has_acpi_companion(dev))
>                 acpi_dev_clear_dependencies(ACPI_COMPANION(dev));
>
> }
> #else
> static inline void acpi_device_clear_dep(struct device *dev) {}
> #endif
>
> Something like that ?

Yup.