qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
* [PATCH 0/2] hw/intc/riscv_aplic: Expand AIA target[i] source handling and refactor related code
@ 2025-10-29  7:17 Nikita Novikov
  2025-10-29  7:17 ` [PATCH 1/2] hw/intc/riscv_aplic: Expand inactive source handling for AIA target[i] Nikita Novikov
                   ` (2 more replies)
  0 siblings, 3 replies; 8+ messages in thread
From: Nikita Novikov @ 2025-10-29  7:17 UTC (permalink / raw)
  To: qemu-devel
  Cc: Nikita Novikov, qemu-riscv, Palmer Dabbelt, Alistair Francis,
	Weiwei Li, Daniel Henrique Barboza, Liu Zhiwei

During the debug of APLIC I faced with the problem with target[i]
registers of source i. According to RISC-V AIA spec v1.0 ratificated on
the 12-th of March, 2025, the section 4.5.2 ("Source configurations")
says, than register target[i] shall be read-only zero when interrupt source i
is inactive in this domain. A source is inactive if it is delegated to a child
domain or its source mode is INACTIVE. Currently we checks the source status
only using sm == INACTIVE, but we do not check whether the source is
delegated.

This series of patches expands current active source checking and reduces
code the associated code dublication.

Signed-off-by: Nikita Novikov <n.novikov@syntacore.com>
---
Nikita Novikov (2):
      hw/intc/riscv_aplic: Expand inactive source handling for AIA target[i]
      hw/intc/riscv_aplic: Factor out source_active() and remove duplicate checks

 hw/intc/riscv_aplic.c | 66 ++++++++++++++++++++++-----------------------------
 1 file changed, 29 insertions(+), 37 deletions(-)
---
base-commit: d1bf06e8cc8197a2b18dbde820d957a51899d374
change-id: 20251028-n-novikov-aplic_aia_ro-1baa353cd672

Best regards,
-- 
Nikita Novikov <n.novikov@syntacore.com>



^ permalink raw reply	[flat|nested] 8+ messages in thread

* [PATCH 1/2] hw/intc/riscv_aplic: Expand inactive source handling for AIA target[i]
  2025-10-29  7:17 [PATCH 0/2] hw/intc/riscv_aplic: Expand AIA target[i] source handling and refactor related code Nikita Novikov
@ 2025-10-29  7:17 ` Nikita Novikov
  2025-10-30 13:19   ` Daniel Henrique Barboza
  2025-11-12  0:55   ` Alistair Francis
  2025-10-29  7:17 ` [PATCH 2/2] hw/intc/riscv_aplic: Factor out source_active() and remove duplicate checks Nikita Novikov
  2025-11-12  1:06 ` [PATCH 0/2] hw/intc/riscv_aplic: Expand AIA target[i] source handling and refactor related code Alistair Francis
  2 siblings, 2 replies; 8+ messages in thread
From: Nikita Novikov @ 2025-10-29  7:17 UTC (permalink / raw)
  To: qemu-devel
  Cc: Nikita Novikov, qemu-riscv, Palmer Dabbelt, Alistair Francis,
	Weiwei Li, Daniel Henrique Barboza, Liu Zhiwei

According to the RISC-V AIA v1.0, section 4.5.2 ("Source configurations"),
register target[i] shall be read-only zero when interrupt source i is inactive
in this domain. A source is inactive if it is delegated to a child domain or
its source mode is INACTIVE.

The previous implementation only checked SM == INACTIVE. This patch adds
full compliance:
- Return zero on read if D == 1 or SM == INACTIVE
- Ignore writes in both cases

Fixes: b6f1244678 ("intc/riscv_aplic: Fix target register read when source is inactive")
Signed-off-by: Nikita Novikov <n.novikov@syntacore.com>
---
 hw/intc/riscv_aplic.c | 28 +++++++++++++++++++++++++---
 1 file changed, 25 insertions(+), 3 deletions(-)

diff --git a/hw/intc/riscv_aplic.c b/hw/intc/riscv_aplic.c
index a2041e702245211ba3bcf4334301d7c86272e36f..8c3b16074cd3ca1bc3004cfaaa13f34b8860bd48 100644
--- a/hw/intc/riscv_aplic.c
+++ b/hw/intc/riscv_aplic.c
@@ -193,6 +193,26 @@ void riscv_aplic_set_kvm_msicfgaddr(RISCVAPLICState *aplic, hwaddr addr)
 #endif
 }
 
+/*
+ * APLIC target[i] must be read-only zero if the source i is inactive
+ * in this domain (delegated or SM == INACTIVE)
+ */
+static inline bool riscv_aplic_source_active(RISCVAPLICState *aplic,
+                                             uint32_t irq)
+{
+    uint32_t sc, sm;
+
+    if ((irq == 0) || (aplic->num_irqs <= irq)) {
+        return false;
+    }
+    sc = aplic->sourcecfg[irq];
+    if (sc & APLIC_SOURCECFG_D) {
+        return false;
+    }
+    sm = sc & APLIC_SOURCECFG_SM_MASK;
+    return sm != APLIC_SOURCECFG_SM_INACTIVE;
+}
+
 static bool riscv_aplic_irq_rectified_val(RISCVAPLICState *aplic,
                                           uint32_t irq)
 {
@@ -635,7 +655,7 @@ static void riscv_aplic_request(void *opaque, int irq, int level)
 
 static uint64_t riscv_aplic_read(void *opaque, hwaddr addr, unsigned size)
 {
-    uint32_t irq, word, idc, sm;
+    uint32_t irq, word, idc;
     RISCVAPLICState *aplic = opaque;
 
     /* Reads must be 4 byte words */
@@ -703,8 +723,7 @@ static uint64_t riscv_aplic_read(void *opaque, hwaddr addr, unsigned size)
     } else if ((APLIC_TARGET_BASE <= addr) &&
             (addr < (APLIC_TARGET_BASE + (aplic->num_irqs - 1) * 4))) {
         irq = ((addr - APLIC_TARGET_BASE) >> 2) + 1;
-        sm = aplic->sourcecfg[irq] & APLIC_SOURCECFG_SM_MASK;
-        if (sm == APLIC_SOURCECFG_SM_INACTIVE) {
+        if (!riscv_aplic_source_active(aplic, irq)) {
             return 0;
         }
         return aplic->target[irq];
@@ -841,6 +860,9 @@ static void riscv_aplic_write(void *opaque, hwaddr addr, uint64_t value,
     } else if ((APLIC_TARGET_BASE <= addr) &&
             (addr < (APLIC_TARGET_BASE + (aplic->num_irqs - 1) * 4))) {
         irq = ((addr - APLIC_TARGET_BASE) >> 2) + 1;
+        if (!riscv_aplic_source_active(aplic, irq)) {
+            return;
+        }
         if (aplic->msimode) {
             aplic->target[irq] = value;
         } else {

-- 
2.51.0



^ permalink raw reply related	[flat|nested] 8+ messages in thread

* [PATCH 2/2] hw/intc/riscv_aplic: Factor out source_active() and remove duplicate checks
  2025-10-29  7:17 [PATCH 0/2] hw/intc/riscv_aplic: Expand AIA target[i] source handling and refactor related code Nikita Novikov
  2025-10-29  7:17 ` [PATCH 1/2] hw/intc/riscv_aplic: Expand inactive source handling for AIA target[i] Nikita Novikov
@ 2025-10-29  7:17 ` Nikita Novikov
  2025-10-30 13:19   ` Daniel Henrique Barboza
  2025-11-12  0:57   ` Alistair Francis
  2025-11-12  1:06 ` [PATCH 0/2] hw/intc/riscv_aplic: Expand AIA target[i] source handling and refactor related code Alistair Francis
  2 siblings, 2 replies; 8+ messages in thread
From: Nikita Novikov @ 2025-10-29  7:17 UTC (permalink / raw)
  To: qemu-devel
  Cc: Nikita Novikov, qemu-riscv, Palmer Dabbelt, Alistair Francis,
	Weiwei Li, Daniel Henrique Barboza, Liu Zhiwei

Refactor the APLIC code to consolidate repeated conditions checking
whether an interrupt source is valid, delegated, or inactive.

Signed-off-by: Nikita Novikov <n.novikov@syntacore.com>
---
 hw/intc/riscv_aplic.c | 44 +++++++-------------------------------------
 1 file changed, 7 insertions(+), 37 deletions(-)

diff --git a/hw/intc/riscv_aplic.c b/hw/intc/riscv_aplic.c
index 8c3b16074cd3ca1bc3004cfaaa13f34b8860bd48..ccfbc9b4656f3e2a69eb5bcd1cee9e5762020351 100644
--- a/hw/intc/riscv_aplic.c
+++ b/hw/intc/riscv_aplic.c
@@ -216,22 +216,13 @@ static inline bool riscv_aplic_source_active(RISCVAPLICState *aplic,
 static bool riscv_aplic_irq_rectified_val(RISCVAPLICState *aplic,
                                           uint32_t irq)
 {
-    uint32_t sourcecfg, sm, raw_input, irq_inverted;
+    uint32_t sm, raw_input, irq_inverted;
 
-    if (!irq || aplic->num_irqs <= irq) {
-        return false;
-    }
-
-    sourcecfg = aplic->sourcecfg[irq];
-    if (sourcecfg & APLIC_SOURCECFG_D) {
-        return false;
-    }
-
-    sm = sourcecfg & APLIC_SOURCECFG_SM_MASK;
-    if (sm == APLIC_SOURCECFG_SM_INACTIVE) {
+    if (!riscv_aplic_source_active(aplic, irq)) {
         return false;
     }
 
+    sm = aplic->sourcecfg[irq] & APLIC_SOURCECFG_SM_MASK;
     raw_input = (aplic->state[irq] & APLIC_ISTATE_INPUT) ? 1 : 0;
     irq_inverted = (sm == APLIC_SOURCECFG_SM_LEVEL_LOW ||
                     sm == APLIC_SOURCECFG_SM_EDGE_FALL) ? 1 : 0;
@@ -284,22 +275,13 @@ static void riscv_aplic_set_pending_raw(RISCVAPLICState *aplic,
 static void riscv_aplic_set_pending(RISCVAPLICState *aplic,
                                     uint32_t irq, bool pending)
 {
-    uint32_t sourcecfg, sm;
+    uint32_t sm;
 
-    if ((irq <= 0) || (aplic->num_irqs <= irq)) {
-        return;
-    }
-
-    sourcecfg = aplic->sourcecfg[irq];
-    if (sourcecfg & APLIC_SOURCECFG_D) {
-        return;
-    }
-
-    sm = sourcecfg & APLIC_SOURCECFG_SM_MASK;
-    if (sm == APLIC_SOURCECFG_SM_INACTIVE) {
+    if (!riscv_aplic_source_active(aplic, irq)) {
         return;
     }
 
+    sm = aplic->sourcecfg[irq] & APLIC_SOURCECFG_SM_MASK;
     if ((sm == APLIC_SOURCECFG_SM_LEVEL_HIGH) ||
         (sm == APLIC_SOURCECFG_SM_LEVEL_LOW)) {
         if (!aplic->msimode) {
@@ -370,19 +352,7 @@ static void riscv_aplic_set_enabled_raw(RISCVAPLICState *aplic,
 static void riscv_aplic_set_enabled(RISCVAPLICState *aplic,
                                     uint32_t irq, bool enabled)
 {
-    uint32_t sourcecfg, sm;
-
-    if ((irq <= 0) || (aplic->num_irqs <= irq)) {
-        return;
-    }
-
-    sourcecfg = aplic->sourcecfg[irq];
-    if (sourcecfg & APLIC_SOURCECFG_D) {
-        return;
-    }
-
-    sm = sourcecfg & APLIC_SOURCECFG_SM_MASK;
-    if (sm == APLIC_SOURCECFG_SM_INACTIVE) {
+    if (!riscv_aplic_source_active(aplic, irq)) {
         return;
     }
 

-- 
2.51.0



^ permalink raw reply related	[flat|nested] 8+ messages in thread

* Re: [PATCH 1/2] hw/intc/riscv_aplic: Expand inactive source handling for AIA target[i]
  2025-10-29  7:17 ` [PATCH 1/2] hw/intc/riscv_aplic: Expand inactive source handling for AIA target[i] Nikita Novikov
@ 2025-10-30 13:19   ` Daniel Henrique Barboza
  2025-11-12  0:55   ` Alistair Francis
  1 sibling, 0 replies; 8+ messages in thread
From: Daniel Henrique Barboza @ 2025-10-30 13:19 UTC (permalink / raw)
  To: Nikita Novikov, qemu-devel
  Cc: qemu-riscv, Palmer Dabbelt, Alistair Francis, Weiwei Li,
	Liu Zhiwei



On 10/29/25 4:17 AM, Nikita Novikov wrote:
> According to the RISC-V AIA v1.0, section 4.5.2 ("Source configurations"),
> register target[i] shall be read-only zero when interrupt source i is inactive
> in this domain. A source is inactive if it is delegated to a child domain or
> its source mode is INACTIVE.
> 
> The previous implementation only checked SM == INACTIVE. This patch adds
> full compliance:
> - Return zero on read if D == 1 or SM == INACTIVE
> - Ignore writes in both cases
> 
> Fixes: b6f1244678 ("intc/riscv_aplic: Fix target register read when source is inactive")
> Signed-off-by: Nikita Novikov <n.novikov@syntacore.com>
> ---

Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com>

>   hw/intc/riscv_aplic.c | 28 +++++++++++++++++++++++++---
>   1 file changed, 25 insertions(+), 3 deletions(-)
> 
> diff --git a/hw/intc/riscv_aplic.c b/hw/intc/riscv_aplic.c
> index a2041e702245211ba3bcf4334301d7c86272e36f..8c3b16074cd3ca1bc3004cfaaa13f34b8860bd48 100644
> --- a/hw/intc/riscv_aplic.c
> +++ b/hw/intc/riscv_aplic.c
> @@ -193,6 +193,26 @@ void riscv_aplic_set_kvm_msicfgaddr(RISCVAPLICState *aplic, hwaddr addr)
>   #endif
>   }
>   
> +/*
> + * APLIC target[i] must be read-only zero if the source i is inactive
> + * in this domain (delegated or SM == INACTIVE)
> + */
> +static inline bool riscv_aplic_source_active(RISCVAPLICState *aplic,
> +                                             uint32_t irq)
> +{
> +    uint32_t sc, sm;
> +
> +    if ((irq == 0) || (aplic->num_irqs <= irq)) {
> +        return false;
> +    }
> +    sc = aplic->sourcecfg[irq];
> +    if (sc & APLIC_SOURCECFG_D) {
> +        return false;
> +    }
> +    sm = sc & APLIC_SOURCECFG_SM_MASK;
> +    return sm != APLIC_SOURCECFG_SM_INACTIVE;
> +}
> +
>   static bool riscv_aplic_irq_rectified_val(RISCVAPLICState *aplic,
>                                             uint32_t irq)
>   {
> @@ -635,7 +655,7 @@ static void riscv_aplic_request(void *opaque, int irq, int level)
>   
>   static uint64_t riscv_aplic_read(void *opaque, hwaddr addr, unsigned size)
>   {
> -    uint32_t irq, word, idc, sm;
> +    uint32_t irq, word, idc;
>       RISCVAPLICState *aplic = opaque;
>   
>       /* Reads must be 4 byte words */
> @@ -703,8 +723,7 @@ static uint64_t riscv_aplic_read(void *opaque, hwaddr addr, unsigned size)
>       } else if ((APLIC_TARGET_BASE <= addr) &&
>               (addr < (APLIC_TARGET_BASE + (aplic->num_irqs - 1) * 4))) {
>           irq = ((addr - APLIC_TARGET_BASE) >> 2) + 1;
> -        sm = aplic->sourcecfg[irq] & APLIC_SOURCECFG_SM_MASK;
> -        if (sm == APLIC_SOURCECFG_SM_INACTIVE) {
> +        if (!riscv_aplic_source_active(aplic, irq)) {
>               return 0;
>           }
>           return aplic->target[irq];
> @@ -841,6 +860,9 @@ static void riscv_aplic_write(void *opaque, hwaddr addr, uint64_t value,
>       } else if ((APLIC_TARGET_BASE <= addr) &&
>               (addr < (APLIC_TARGET_BASE + (aplic->num_irqs - 1) * 4))) {
>           irq = ((addr - APLIC_TARGET_BASE) >> 2) + 1;
> +        if (!riscv_aplic_source_active(aplic, irq)) {
> +            return;
> +        }
>           if (aplic->msimode) {
>               aplic->target[irq] = value;
>           } else {
> 



^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [PATCH 2/2] hw/intc/riscv_aplic: Factor out source_active() and remove duplicate checks
  2025-10-29  7:17 ` [PATCH 2/2] hw/intc/riscv_aplic: Factor out source_active() and remove duplicate checks Nikita Novikov
@ 2025-10-30 13:19   ` Daniel Henrique Barboza
  2025-11-12  0:57   ` Alistair Francis
  1 sibling, 0 replies; 8+ messages in thread
From: Daniel Henrique Barboza @ 2025-10-30 13:19 UTC (permalink / raw)
  To: Nikita Novikov, qemu-devel
  Cc: qemu-riscv, Palmer Dabbelt, Alistair Francis, Weiwei Li,
	Liu Zhiwei



On 10/29/25 4:17 AM, Nikita Novikov wrote:
> Refactor the APLIC code to consolidate repeated conditions checking
> whether an interrupt source is valid, delegated, or inactive.
> 
> Signed-off-by: Nikita Novikov <n.novikov@syntacore.com>
> ---


Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com>

>   hw/intc/riscv_aplic.c | 44 +++++++-------------------------------------
>   1 file changed, 7 insertions(+), 37 deletions(-)
> 
> diff --git a/hw/intc/riscv_aplic.c b/hw/intc/riscv_aplic.c
> index 8c3b16074cd3ca1bc3004cfaaa13f34b8860bd48..ccfbc9b4656f3e2a69eb5bcd1cee9e5762020351 100644
> --- a/hw/intc/riscv_aplic.c
> +++ b/hw/intc/riscv_aplic.c
> @@ -216,22 +216,13 @@ static inline bool riscv_aplic_source_active(RISCVAPLICState *aplic,
>   static bool riscv_aplic_irq_rectified_val(RISCVAPLICState *aplic,
>                                             uint32_t irq)
>   {
> -    uint32_t sourcecfg, sm, raw_input, irq_inverted;
> +    uint32_t sm, raw_input, irq_inverted;
>   
> -    if (!irq || aplic->num_irqs <= irq) {
> -        return false;
> -    }
> -
> -    sourcecfg = aplic->sourcecfg[irq];
> -    if (sourcecfg & APLIC_SOURCECFG_D) {
> -        return false;
> -    }
> -
> -    sm = sourcecfg & APLIC_SOURCECFG_SM_MASK;
> -    if (sm == APLIC_SOURCECFG_SM_INACTIVE) {
> +    if (!riscv_aplic_source_active(aplic, irq)) {
>           return false;
>       }
>   
> +    sm = aplic->sourcecfg[irq] & APLIC_SOURCECFG_SM_MASK;
>       raw_input = (aplic->state[irq] & APLIC_ISTATE_INPUT) ? 1 : 0;
>       irq_inverted = (sm == APLIC_SOURCECFG_SM_LEVEL_LOW ||
>                       sm == APLIC_SOURCECFG_SM_EDGE_FALL) ? 1 : 0;
> @@ -284,22 +275,13 @@ static void riscv_aplic_set_pending_raw(RISCVAPLICState *aplic,
>   static void riscv_aplic_set_pending(RISCVAPLICState *aplic,
>                                       uint32_t irq, bool pending)
>   {
> -    uint32_t sourcecfg, sm;
> +    uint32_t sm;
>   
> -    if ((irq <= 0) || (aplic->num_irqs <= irq)) {
> -        return;
> -    }
> -
> -    sourcecfg = aplic->sourcecfg[irq];
> -    if (sourcecfg & APLIC_SOURCECFG_D) {
> -        return;
> -    }
> -
> -    sm = sourcecfg & APLIC_SOURCECFG_SM_MASK;
> -    if (sm == APLIC_SOURCECFG_SM_INACTIVE) {
> +    if (!riscv_aplic_source_active(aplic, irq)) {
>           return;
>       }
>   
> +    sm = aplic->sourcecfg[irq] & APLIC_SOURCECFG_SM_MASK;
>       if ((sm == APLIC_SOURCECFG_SM_LEVEL_HIGH) ||
>           (sm == APLIC_SOURCECFG_SM_LEVEL_LOW)) {
>           if (!aplic->msimode) {
> @@ -370,19 +352,7 @@ static void riscv_aplic_set_enabled_raw(RISCVAPLICState *aplic,
>   static void riscv_aplic_set_enabled(RISCVAPLICState *aplic,
>                                       uint32_t irq, bool enabled)
>   {
> -    uint32_t sourcecfg, sm;
> -
> -    if ((irq <= 0) || (aplic->num_irqs <= irq)) {
> -        return;
> -    }
> -
> -    sourcecfg = aplic->sourcecfg[irq];
> -    if (sourcecfg & APLIC_SOURCECFG_D) {
> -        return;
> -    }
> -
> -    sm = sourcecfg & APLIC_SOURCECFG_SM_MASK;
> -    if (sm == APLIC_SOURCECFG_SM_INACTIVE) {
> +    if (!riscv_aplic_source_active(aplic, irq)) {
>           return;
>       }
>   
> 



^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [PATCH 1/2] hw/intc/riscv_aplic: Expand inactive source handling for AIA target[i]
  2025-10-29  7:17 ` [PATCH 1/2] hw/intc/riscv_aplic: Expand inactive source handling for AIA target[i] Nikita Novikov
  2025-10-30 13:19   ` Daniel Henrique Barboza
@ 2025-11-12  0:55   ` Alistair Francis
  1 sibling, 0 replies; 8+ messages in thread
From: Alistair Francis @ 2025-11-12  0:55 UTC (permalink / raw)
  To: Nikita Novikov
  Cc: qemu-devel, qemu-riscv, Palmer Dabbelt, Alistair Francis,
	Weiwei Li, Daniel Henrique Barboza, Liu Zhiwei

On Wed, Oct 29, 2025 at 5:19 PM Nikita Novikov <n.novikov@syntacore.com> wrote:
>
> According to the RISC-V AIA v1.0, section 4.5.2 ("Source configurations"),
> register target[i] shall be read-only zero when interrupt source i is inactive
> in this domain. A source is inactive if it is delegated to a child domain or
> its source mode is INACTIVE.
>
> The previous implementation only checked SM == INACTIVE. This patch adds
> full compliance:
> - Return zero on read if D == 1 or SM == INACTIVE
> - Ignore writes in both cases
>
> Fixes: b6f1244678 ("intc/riscv_aplic: Fix target register read when source is inactive")
> Signed-off-by: Nikita Novikov <n.novikov@syntacore.com>

Acked-by: Alistair Francis <alistair.francis@wdc.com>

Alistair

> ---
>  hw/intc/riscv_aplic.c | 28 +++++++++++++++++++++++++---
>  1 file changed, 25 insertions(+), 3 deletions(-)
>
> diff --git a/hw/intc/riscv_aplic.c b/hw/intc/riscv_aplic.c
> index a2041e702245211ba3bcf4334301d7c86272e36f..8c3b16074cd3ca1bc3004cfaaa13f34b8860bd48 100644
> --- a/hw/intc/riscv_aplic.c
> +++ b/hw/intc/riscv_aplic.c
> @@ -193,6 +193,26 @@ void riscv_aplic_set_kvm_msicfgaddr(RISCVAPLICState *aplic, hwaddr addr)
>  #endif
>  }
>
> +/*
> + * APLIC target[i] must be read-only zero if the source i is inactive
> + * in this domain (delegated or SM == INACTIVE)
> + */
> +static inline bool riscv_aplic_source_active(RISCVAPLICState *aplic,
> +                                             uint32_t irq)
> +{
> +    uint32_t sc, sm;
> +
> +    if ((irq == 0) || (aplic->num_irqs <= irq)) {
> +        return false;
> +    }
> +    sc = aplic->sourcecfg[irq];
> +    if (sc & APLIC_SOURCECFG_D) {
> +        return false;
> +    }
> +    sm = sc & APLIC_SOURCECFG_SM_MASK;
> +    return sm != APLIC_SOURCECFG_SM_INACTIVE;
> +}
> +
>  static bool riscv_aplic_irq_rectified_val(RISCVAPLICState *aplic,
>                                            uint32_t irq)
>  {
> @@ -635,7 +655,7 @@ static void riscv_aplic_request(void *opaque, int irq, int level)
>
>  static uint64_t riscv_aplic_read(void *opaque, hwaddr addr, unsigned size)
>  {
> -    uint32_t irq, word, idc, sm;
> +    uint32_t irq, word, idc;
>      RISCVAPLICState *aplic = opaque;
>
>      /* Reads must be 4 byte words */
> @@ -703,8 +723,7 @@ static uint64_t riscv_aplic_read(void *opaque, hwaddr addr, unsigned size)
>      } else if ((APLIC_TARGET_BASE <= addr) &&
>              (addr < (APLIC_TARGET_BASE + (aplic->num_irqs - 1) * 4))) {
>          irq = ((addr - APLIC_TARGET_BASE) >> 2) + 1;
> -        sm = aplic->sourcecfg[irq] & APLIC_SOURCECFG_SM_MASK;
> -        if (sm == APLIC_SOURCECFG_SM_INACTIVE) {
> +        if (!riscv_aplic_source_active(aplic, irq)) {
>              return 0;
>          }
>          return aplic->target[irq];
> @@ -841,6 +860,9 @@ static void riscv_aplic_write(void *opaque, hwaddr addr, uint64_t value,
>      } else if ((APLIC_TARGET_BASE <= addr) &&
>              (addr < (APLIC_TARGET_BASE + (aplic->num_irqs - 1) * 4))) {
>          irq = ((addr - APLIC_TARGET_BASE) >> 2) + 1;
> +        if (!riscv_aplic_source_active(aplic, irq)) {
> +            return;
> +        }
>          if (aplic->msimode) {
>              aplic->target[irq] = value;
>          } else {
>
> --
> 2.51.0
>
>


^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [PATCH 2/2] hw/intc/riscv_aplic: Factor out source_active() and remove duplicate checks
  2025-10-29  7:17 ` [PATCH 2/2] hw/intc/riscv_aplic: Factor out source_active() and remove duplicate checks Nikita Novikov
  2025-10-30 13:19   ` Daniel Henrique Barboza
@ 2025-11-12  0:57   ` Alistair Francis
  1 sibling, 0 replies; 8+ messages in thread
From: Alistair Francis @ 2025-11-12  0:57 UTC (permalink / raw)
  To: Nikita Novikov
  Cc: qemu-devel, qemu-riscv, Palmer Dabbelt, Alistair Francis,
	Weiwei Li, Daniel Henrique Barboza, Liu Zhiwei

On Wed, Oct 29, 2025 at 5:19 PM Nikita Novikov <n.novikov@syntacore.com> wrote:
>
> Refactor the APLIC code to consolidate repeated conditions checking
> whether an interrupt source is valid, delegated, or inactive.
>
> Signed-off-by: Nikita Novikov <n.novikov@syntacore.com>

Reviewed-by: Alistair Francis <alistair.francis@wdc.com>

Alistair

> ---
>  hw/intc/riscv_aplic.c | 44 +++++++-------------------------------------
>  1 file changed, 7 insertions(+), 37 deletions(-)
>
> diff --git a/hw/intc/riscv_aplic.c b/hw/intc/riscv_aplic.c
> index 8c3b16074cd3ca1bc3004cfaaa13f34b8860bd48..ccfbc9b4656f3e2a69eb5bcd1cee9e5762020351 100644
> --- a/hw/intc/riscv_aplic.c
> +++ b/hw/intc/riscv_aplic.c
> @@ -216,22 +216,13 @@ static inline bool riscv_aplic_source_active(RISCVAPLICState *aplic,
>  static bool riscv_aplic_irq_rectified_val(RISCVAPLICState *aplic,
>                                            uint32_t irq)
>  {
> -    uint32_t sourcecfg, sm, raw_input, irq_inverted;
> +    uint32_t sm, raw_input, irq_inverted;
>
> -    if (!irq || aplic->num_irqs <= irq) {
> -        return false;
> -    }
> -
> -    sourcecfg = aplic->sourcecfg[irq];
> -    if (sourcecfg & APLIC_SOURCECFG_D) {
> -        return false;
> -    }
> -
> -    sm = sourcecfg & APLIC_SOURCECFG_SM_MASK;
> -    if (sm == APLIC_SOURCECFG_SM_INACTIVE) {
> +    if (!riscv_aplic_source_active(aplic, irq)) {
>          return false;
>      }
>
> +    sm = aplic->sourcecfg[irq] & APLIC_SOURCECFG_SM_MASK;
>      raw_input = (aplic->state[irq] & APLIC_ISTATE_INPUT) ? 1 : 0;
>      irq_inverted = (sm == APLIC_SOURCECFG_SM_LEVEL_LOW ||
>                      sm == APLIC_SOURCECFG_SM_EDGE_FALL) ? 1 : 0;
> @@ -284,22 +275,13 @@ static void riscv_aplic_set_pending_raw(RISCVAPLICState *aplic,
>  static void riscv_aplic_set_pending(RISCVAPLICState *aplic,
>                                      uint32_t irq, bool pending)
>  {
> -    uint32_t sourcecfg, sm;
> +    uint32_t sm;
>
> -    if ((irq <= 0) || (aplic->num_irqs <= irq)) {
> -        return;
> -    }
> -
> -    sourcecfg = aplic->sourcecfg[irq];
> -    if (sourcecfg & APLIC_SOURCECFG_D) {
> -        return;
> -    }
> -
> -    sm = sourcecfg & APLIC_SOURCECFG_SM_MASK;
> -    if (sm == APLIC_SOURCECFG_SM_INACTIVE) {
> +    if (!riscv_aplic_source_active(aplic, irq)) {
>          return;
>      }
>
> +    sm = aplic->sourcecfg[irq] & APLIC_SOURCECFG_SM_MASK;
>      if ((sm == APLIC_SOURCECFG_SM_LEVEL_HIGH) ||
>          (sm == APLIC_SOURCECFG_SM_LEVEL_LOW)) {
>          if (!aplic->msimode) {
> @@ -370,19 +352,7 @@ static void riscv_aplic_set_enabled_raw(RISCVAPLICState *aplic,
>  static void riscv_aplic_set_enabled(RISCVAPLICState *aplic,
>                                      uint32_t irq, bool enabled)
>  {
> -    uint32_t sourcecfg, sm;
> -
> -    if ((irq <= 0) || (aplic->num_irqs <= irq)) {
> -        return;
> -    }
> -
> -    sourcecfg = aplic->sourcecfg[irq];
> -    if (sourcecfg & APLIC_SOURCECFG_D) {
> -        return;
> -    }
> -
> -    sm = sourcecfg & APLIC_SOURCECFG_SM_MASK;
> -    if (sm == APLIC_SOURCECFG_SM_INACTIVE) {
> +    if (!riscv_aplic_source_active(aplic, irq)) {
>          return;
>      }
>
>
> --
> 2.51.0
>
>


^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [PATCH 0/2] hw/intc/riscv_aplic: Expand AIA target[i] source handling and refactor related code
  2025-10-29  7:17 [PATCH 0/2] hw/intc/riscv_aplic: Expand AIA target[i] source handling and refactor related code Nikita Novikov
  2025-10-29  7:17 ` [PATCH 1/2] hw/intc/riscv_aplic: Expand inactive source handling for AIA target[i] Nikita Novikov
  2025-10-29  7:17 ` [PATCH 2/2] hw/intc/riscv_aplic: Factor out source_active() and remove duplicate checks Nikita Novikov
@ 2025-11-12  1:06 ` Alistair Francis
  2 siblings, 0 replies; 8+ messages in thread
From: Alistair Francis @ 2025-11-12  1:06 UTC (permalink / raw)
  To: Nikita Novikov
  Cc: qemu-devel, qemu-riscv, Palmer Dabbelt, Alistair Francis,
	Weiwei Li, Daniel Henrique Barboza, Liu Zhiwei

On Wed, Oct 29, 2025 at 5:20 PM Nikita Novikov <n.novikov@syntacore.com> wrote:
>
> During the debug of APLIC I faced with the problem with target[i]
> registers of source i. According to RISC-V AIA spec v1.0 ratificated on
> the 12-th of March, 2025, the section 4.5.2 ("Source configurations")
> says, than register target[i] shall be read-only zero when interrupt source i
> is inactive in this domain. A source is inactive if it is delegated to a child
> domain or its source mode is INACTIVE. Currently we checks the source status
> only using sm == INACTIVE, but we do not check whether the source is
> delegated.
>
> This series of patches expands current active source checking and reduces
> code the associated code dublication.
>
> Signed-off-by: Nikita Novikov <n.novikov@syntacore.com>
> ---
> Nikita Novikov (2):
>       hw/intc/riscv_aplic: Expand inactive source handling for AIA target[i]
>       hw/intc/riscv_aplic: Factor out source_active() and remove duplicate checks

Thanks!

Applied to riscv-to-apply.next

Alistair

>
>  hw/intc/riscv_aplic.c | 66 ++++++++++++++++++++++-----------------------------
>  1 file changed, 29 insertions(+), 37 deletions(-)
> ---
> base-commit: d1bf06e8cc8197a2b18dbde820d957a51899d374
> change-id: 20251028-n-novikov-aplic_aia_ro-1baa353cd672
>
> Best regards,
> --
> Nikita Novikov <n.novikov@syntacore.com>
>
>


^ permalink raw reply	[flat|nested] 8+ messages in thread

end of thread, other threads:[~2025-11-12  1:07 UTC | newest]

Thread overview: 8+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-10-29  7:17 [PATCH 0/2] hw/intc/riscv_aplic: Expand AIA target[i] source handling and refactor related code Nikita Novikov
2025-10-29  7:17 ` [PATCH 1/2] hw/intc/riscv_aplic: Expand inactive source handling for AIA target[i] Nikita Novikov
2025-10-30 13:19   ` Daniel Henrique Barboza
2025-11-12  0:55   ` Alistair Francis
2025-10-29  7:17 ` [PATCH 2/2] hw/intc/riscv_aplic: Factor out source_active() and remove duplicate checks Nikita Novikov
2025-10-30 13:19   ` Daniel Henrique Barboza
2025-11-12  0:57   ` Alistair Francis
2025-11-12  1:06 ` [PATCH 0/2] hw/intc/riscv_aplic: Expand AIA target[i] source handling and refactor related code Alistair Francis

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).