diff mbox series

[v1] RISC-V: Bugfix ICE for the vector return arg in mode switch

Message ID 20240411021628.3470772-1-pan2.li@intel.com
State New
Headers show
Series [v1] RISC-V: Bugfix ICE for the vector return arg in mode switch | expand

Commit Message

Li, Pan2 April 11, 2024, 2:16 a.m. UTC 
From: Pan Li <pan2.li@intel.com>

This patch would like to fix a ICE in mode sw for below example code.

during RTL pass: mode_sw
test.c: In function ‘vbool16_t j(vuint64m4_t)’:
test.c:15:1: internal compiler error: in create_pre_exit, at
mode-switching.cc:451
   15 | }
      | ^
0x3978f12 create_pre_exit
        __RISCV_BUILD__/../gcc/mode-switching.cc:451
0x3979e9e optimize_mode_switching
        __RISCV_BUILD__/../gcc/mode-switching.cc:849
0x397b9bc execute
        __RISCV_BUILD__/../gcc/mode-switching.cc:1324

extern size_t get_vl ();

vbool16_t
test (vuint64m4_t a)
{
  unsigned long b;
  return __riscv_vmsne_vx_u64m4_b16 (a, b, get_vl ());
}

The create_pre_exit would like to find a return value copy.  If
not, there will be a reason in assert but not available for above
sample code when vector calling convension is enabled by default.
This patch would like to override the TARGET_FUNCTION_VALUE_REGNO_P
for vector register and then we will have hard_regno_nregs for copy_num,
aka there is a return value copy.

As a side-effect of allow vector in TARGET_FUNCTION_VALUE_REGNO_P, the
TARGET_GET_RAW_RESULT_MODE will have vector mode and which is sizeless
cannot be converted to fixed_size_mode.  Thus override the hook
TARGET_GET_RAW_RESULT_MODE and return VOIDmode when the regno is-not-a
fixed_size_mode.

The below tests are passed for this patch.
* The fully riscv regression tests.
* The reproducing test in bugzilla PR114639.

	PR target/114639

gcc/ChangeLog:

	* config/riscv/riscv.cc (riscv_function_value_regno_p): New func
	impl for hook TARGET_FUNCTION_VALUE_REGNO_P.
	(riscv_get_raw_result_mode): New func imple for hook
	TARGET_GET_RAW_RESULT_MODE.
	(TARGET_FUNCTION_VALUE_REGNO_P): Impl the hook.
	(TARGET_GET_RAW_RESULT_MODE): Ditto.
	* config/riscv/riscv.h (V_RETURN): New macro for vector return.
	(GP_RETURN_FIRST): New macro for the first GPR in return.
	(GP_RETURN_LAST): New macro for the last GPR in return.
	(FP_RETURN_FIRST): Diito but for FPR.
	(FP_RETURN_LAST): Ditto.
	(FUNCTION_VALUE_REGNO_P): Remove as deprecated and replace by
	TARGET_FUNCTION_VALUE_REGNO_P.

gcc/testsuite/ChangeLog:

	* g++.target/riscv/rvv/base/pr114639-1.C: New test.
	* gcc.target/riscv/rvv/base/pr114639-1.c: New test.

Signed-off-by: Pan Li <pan2.li@intel.com>
---
 gcc/config/riscv/riscv.cc                     | 34 +++++++++++++++++++
 gcc/config/riscv/riscv.h                      |  8 +++--
 .../g++.target/riscv/rvv/base/pr114639-1.C    | 25 ++++++++++++++
 .../gcc.target/riscv/rvv/base/pr114639-1.c    | 14 ++++++++
 4 files changed, 79 insertions(+), 2 deletions(-)
 create mode 100644 gcc/testsuite/g++.target/riscv/rvv/base/pr114639-1.C
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/base/pr114639-1.c

Comments

juzhe.zhong@rivai.ai April 11, 2024, 2:22 a.m. UTC | #1 
Thanks for fixing it. LGTM from my side.

I prefer wait kito for another ACK.



juzhe.zhong@rivai.ai
 
From: pan2.li
Date: 2024-04-11 10:16
To: gcc-patches
CC: juzhe.zhong; kito.cheng; Pan Li
Subject: [PATCH v1] RISC-V: Bugfix ICE for the vector return arg in mode switch
From: Pan Li <pan2.li@intel.com>
 
This patch would like to fix a ICE in mode sw for below example code.
 
during RTL pass: mode_sw
test.c: In function ‘vbool16_t j(vuint64m4_t)’:
test.c:15:1: internal compiler error: in create_pre_exit, at
mode-switching.cc:451
   15 | }
      | ^
0x3978f12 create_pre_exit
        __RISCV_BUILD__/../gcc/mode-switching.cc:451
0x3979e9e optimize_mode_switching
        __RISCV_BUILD__/../gcc/mode-switching.cc:849
0x397b9bc execute
        __RISCV_BUILD__/../gcc/mode-switching.cc:1324
 
extern size_t get_vl ();
 
vbool16_t
test (vuint64m4_t a)
{
  unsigned long b;
  return __riscv_vmsne_vx_u64m4_b16 (a, b, get_vl ());
}
 
The create_pre_exit would like to find a return value copy.  If
not, there will be a reason in assert but not available for above
sample code when vector calling convension is enabled by default.
This patch would like to override the TARGET_FUNCTION_VALUE_REGNO_P
for vector register and then we will have hard_regno_nregs for copy_num,
aka there is a return value copy.
 
As a side-effect of allow vector in TARGET_FUNCTION_VALUE_REGNO_P, the
TARGET_GET_RAW_RESULT_MODE will have vector mode and which is sizeless
cannot be converted to fixed_size_mode.  Thus override the hook
TARGET_GET_RAW_RESULT_MODE and return VOIDmode when the regno is-not-a
fixed_size_mode.
 
The below tests are passed for this patch.
* The fully riscv regression tests.
* The reproducing test in bugzilla PR114639.
 
PR target/114639
 
gcc/ChangeLog:
 
* config/riscv/riscv.cc (riscv_function_value_regno_p): New func
impl for hook TARGET_FUNCTION_VALUE_REGNO_P.
(riscv_get_raw_result_mode): New func imple for hook
TARGET_GET_RAW_RESULT_MODE.
(TARGET_FUNCTION_VALUE_REGNO_P): Impl the hook.
(TARGET_GET_RAW_RESULT_MODE): Ditto.
* config/riscv/riscv.h (V_RETURN): New macro for vector return.
(GP_RETURN_FIRST): New macro for the first GPR in return.
(GP_RETURN_LAST): New macro for the last GPR in return.
(FP_RETURN_FIRST): Diito but for FPR.
(FP_RETURN_LAST): Ditto.
(FUNCTION_VALUE_REGNO_P): Remove as deprecated and replace by
TARGET_FUNCTION_VALUE_REGNO_P.
 
gcc/testsuite/ChangeLog:
 
* g++.target/riscv/rvv/base/pr114639-1.C: New test.
* gcc.target/riscv/rvv/base/pr114639-1.c: New test.
 
Signed-off-by: Pan Li <pan2.li@intel.com>
---
gcc/config/riscv/riscv.cc                     | 34 +++++++++++++++++++
gcc/config/riscv/riscv.h                      |  8 +++--
.../g++.target/riscv/rvv/base/pr114639-1.C    | 25 ++++++++++++++
.../gcc.target/riscv/rvv/base/pr114639-1.c    | 14 ++++++++
4 files changed, 79 insertions(+), 2 deletions(-)
create mode 100644 gcc/testsuite/g++.target/riscv/rvv/base/pr114639-1.C
create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/base/pr114639-1.c
 
diff --git a/gcc/config/riscv/riscv.cc b/gcc/config/riscv/riscv.cc
index 00defa69fd8..91f017dd52a 100644
--- a/gcc/config/riscv/riscv.cc
+++ b/gcc/config/riscv/riscv.cc
@@ -10997,6 +10997,34 @@ riscv_vector_mode_supported_any_target_p (machine_mode)
   return true;
}
+/* Implements hook TARGET_FUNCTION_VALUE_REGNO_P.  */
+
+static bool
+riscv_function_value_regno_p (const unsigned regno)
+{
+  if (GP_RETURN_FIRST <= regno && regno <= GP_RETURN_LAST)
+    return true;
+
+  if (FP_RETURN_FIRST <= regno && regno <= FP_RETURN_LAST)
+    return true;
+
+  if (regno == V_RETURN)
+    return true;
+
+  return false;
+}
+
+/* Implements hook TARGET_GET_RAW_RESULT_MODE.  */
+
+static fixed_size_mode
+riscv_get_raw_result_mode (int regno)
+{
+  if (!is_a <fixed_size_mode> (reg_raw_mode[regno]))
+    return as_a <fixed_size_mode> (VOIDmode);
+
+  return default_get_reg_raw_mode (regno);
+}
+
/* Initialize the GCC target structure.  */
#undef TARGET_ASM_ALIGNED_HI_OP
#define TARGET_ASM_ALIGNED_HI_OP "\t.half\t"
@@ -11343,6 +11371,12 @@ riscv_vector_mode_supported_any_target_p (machine_mode)
#undef TARGET_VECTOR_MODE_SUPPORTED_ANY_TARGET_P
#define TARGET_VECTOR_MODE_SUPPORTED_ANY_TARGET_P riscv_vector_mode_supported_any_target_p
+#undef TARGET_FUNCTION_VALUE_REGNO_P
+#define TARGET_FUNCTION_VALUE_REGNO_P riscv_function_value_regno_p
+
+#undef TARGET_GET_RAW_RESULT_MODE
+#define TARGET_GET_RAW_RESULT_MODE riscv_get_raw_result_mode
+
struct gcc_target targetm = TARGET_INITIALIZER;
#include "gt-riscv.h"
diff --git a/gcc/config/riscv/riscv.h b/gcc/config/riscv/riscv.h
index 269b8c1f076..7797e67317a 100644
--- a/gcc/config/riscv/riscv.h
+++ b/gcc/config/riscv/riscv.h
@@ -683,6 +683,12 @@ enum reg_class
#define GP_RETURN GP_ARG_FIRST
#define FP_RETURN (UNITS_PER_FP_ARG == 0 ? GP_RETURN : FP_ARG_FIRST)
+#define V_RETURN  V_REG_FIRST
+
+#define GP_RETURN_FIRST GP_ARG_FIRST
+#define GP_RETURN_LAST  GP_ARG_FIRST + 1
+#define FP_RETURN_FIRST FP_RETURN
+#define FP_RETURN_LAST  FP_RETURN + 1
#define MAX_ARGS_IN_REGISTERS \
   (riscv_abi == ABI_ILP32E || riscv_abi == ABI_LP64E \
@@ -714,8 +720,6 @@ enum reg_class
#define FUNCTION_VALUE(VALTYPE, FUNC) \
   riscv_function_value (VALTYPE, FUNC, VOIDmode)
-#define FUNCTION_VALUE_REGNO_P(N) ((N) == GP_RETURN || (N) == FP_RETURN)
-
/* 1 if N is a possible register number for function argument passing.
    We have no FP argument registers when soft-float.  */
diff --git a/gcc/testsuite/g++.target/riscv/rvv/base/pr114639-1.C b/gcc/testsuite/g++.target/riscv/rvv/base/pr114639-1.C
new file mode 100644
index 00000000000..9450b108ae5
--- /dev/null
+++ b/gcc/testsuite/g++.target/riscv/rvv/base/pr114639-1.C
@@ -0,0 +1,25 @@
+/* Test that we do not have ice when compile */
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv -mabi=lp64d -O3" } */
+
+typedef long c;
+
+#pragma riscv intrinsic "vector"
+
+template <unsigned long> struct d {};
+
+struct e {
+  using f = d<0>;
+};
+
+struct g {
+  using f = e::f;
+};
+
+template <typename, int> using h = g::f;
+template <unsigned long i> long get_vl (d<i>);
+
+vbool16_t test (vuint64m4_t a) {
+  c b;
+  return __riscv_vmsne_vx_u64m4_b16(a, b, get_vl (h<c, 2>()));
+}
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/base/pr114639-1.c b/gcc/testsuite/gcc.target/riscv/rvv/base/pr114639-1.c
new file mode 100644
index 00000000000..3ad91dbf6bb
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/base/pr114639-1.c
@@ -0,0 +1,14 @@
+/* Test that we do not have ice when compile */
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv -mabi=lp64d -O3" } */
+
+#include <riscv_vector.h>
+
+extern size_t get_vl ();
+
+vbool16_t
+test (vuint64m4_t a)
+{
+  unsigned long b;
+  return __riscv_vmsne_vx_u64m4_b16 (a, b, get_vl ());
+}
Kito Cheng April 11, 2024, 2:49 a.m. UTC | #2 
I was thinking we may guarded with TARGET_VECTOR and TARGET_HARD_FLOAT
or checking with ABI in riscv_function_value_regno_p, however I think
it's fine with current implementation (no checking) after checking all
use site of `targetm.calls.function_value_regno_p`, so LGTM :)

Thanks Pan for fixing this issue!

On Thu, Apr 11, 2024 at 10:23 AM juzhe.zhong@rivai.ai
<juzhe.zhong@rivai.ai> wrote:
>
> Thanks for fixing it. LGTM from my side.
>
> I prefer wait kito for another ACK.
>
> ________________________________
> juzhe.zhong@rivai.ai
>
>
> From: pan2.li
> Date: 2024-04-11 10:16
> To: gcc-patches
> CC: juzhe.zhong; kito.cheng; Pan Li
> Subject: [PATCH v1] RISC-V: Bugfix ICE for the vector return arg in mode switch
> From: Pan Li <pan2.li@intel.com>
>
> This patch would like to fix a ICE in mode sw for below example code.
>
> during RTL pass: mode_sw
> test.c: In function ‘vbool16_t j(vuint64m4_t)’:
> test.c:15:1: internal compiler error: in create_pre_exit, at
> mode-switching.cc:451
>    15 | }
>       | ^
> 0x3978f12 create_pre_exit
>         __RISCV_BUILD__/../gcc/mode-switching.cc:451
> 0x3979e9e optimize_mode_switching
>         __RISCV_BUILD__/../gcc/mode-switching.cc:849
> 0x397b9bc execute
>         __RISCV_BUILD__/../gcc/mode-switching.cc:1324
>
> extern size_t get_vl ();
>
> vbool16_t
> test (vuint64m4_t a)
> {
>   unsigned long b;
>   return __riscv_vmsne_vx_u64m4_b16 (a, b, get_vl ());
> }
>
> The create_pre_exit would like to find a return value copy.  If
> not, there will be a reason in assert but not available for above
> sample code when vector calling convension is enabled by default.
> This patch would like to override the TARGET_FUNCTION_VALUE_REGNO_P
> for vector register and then we will have hard_regno_nregs for copy_num,
> aka there is a return value copy.
>
> As a side-effect of allow vector in TARGET_FUNCTION_VALUE_REGNO_P, the
> TARGET_GET_RAW_RESULT_MODE will have vector mode and which is sizeless
> cannot be converted to fixed_size_mode.  Thus override the hook
> TARGET_GET_RAW_RESULT_MODE and return VOIDmode when the regno is-not-a
> fixed_size_mode.
>
> The below tests are passed for this patch.
> * The fully riscv regression tests.
> * The reproducing test in bugzilla PR114639.
>
> PR target/114639
>
> gcc/ChangeLog:
>
> * config/riscv/riscv.cc (riscv_function_value_regno_p): New func
> impl for hook TARGET_FUNCTION_VALUE_REGNO_P.
> (riscv_get_raw_result_mode): New func imple for hook
> TARGET_GET_RAW_RESULT_MODE.
> (TARGET_FUNCTION_VALUE_REGNO_P): Impl the hook.
> (TARGET_GET_RAW_RESULT_MODE): Ditto.
> * config/riscv/riscv.h (V_RETURN): New macro for vector return.
> (GP_RETURN_FIRST): New macro for the first GPR in return.
> (GP_RETURN_LAST): New macro for the last GPR in return.
> (FP_RETURN_FIRST): Diito but for FPR.
> (FP_RETURN_LAST): Ditto.
> (FUNCTION_VALUE_REGNO_P): Remove as deprecated and replace by
> TARGET_FUNCTION_VALUE_REGNO_P.
>
> gcc/testsuite/ChangeLog:
>
> * g++.target/riscv/rvv/base/pr114639-1.C: New test.
> * gcc.target/riscv/rvv/base/pr114639-1.c: New test.
>
> Signed-off-by: Pan Li <pan2.li@intel.com>
> ---
> gcc/config/riscv/riscv.cc                     | 34 +++++++++++++++++++
> gcc/config/riscv/riscv.h                      |  8 +++--
> .../g++.target/riscv/rvv/base/pr114639-1.C    | 25 ++++++++++++++
> .../gcc.target/riscv/rvv/base/pr114639-1.c    | 14 ++++++++
> 4 files changed, 79 insertions(+), 2 deletions(-)
> create mode 100644 gcc/testsuite/g++.target/riscv/rvv/base/pr114639-1.C
> create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/base/pr114639-1.c
>
> diff --git a/gcc/config/riscv/riscv.cc b/gcc/config/riscv/riscv.cc
> index 00defa69fd8..91f017dd52a 100644
> --- a/gcc/config/riscv/riscv.cc
> +++ b/gcc/config/riscv/riscv.cc
> @@ -10997,6 +10997,34 @@ riscv_vector_mode_supported_any_target_p (machine_mode)
>    return true;
> }
> +/* Implements hook TARGET_FUNCTION_VALUE_REGNO_P.  */
> +
> +static bool
> +riscv_function_value_regno_p (const unsigned regno)
> +{
> +  if (GP_RETURN_FIRST <= regno && regno <= GP_RETURN_LAST)
> +    return true;
> +
> +  if (FP_RETURN_FIRST <= regno && regno <= FP_RETURN_LAST)
> +    return true;
> +
> +  if (regno == V_RETURN)
> +    return true;
> +
> +  return false;
> +}
> +
> +/* Implements hook TARGET_GET_RAW_RESULT_MODE.  */
> +
> +static fixed_size_mode
> +riscv_get_raw_result_mode (int regno)
> +{
> +  if (!is_a <fixed_size_mode> (reg_raw_mode[regno]))
> +    return as_a <fixed_size_mode> (VOIDmode);
> +
> +  return default_get_reg_raw_mode (regno);
> +}
> +
> /* Initialize the GCC target structure.  */
> #undef TARGET_ASM_ALIGNED_HI_OP
> #define TARGET_ASM_ALIGNED_HI_OP "\t.half\t"
> @@ -11343,6 +11371,12 @@ riscv_vector_mode_supported_any_target_p (machine_mode)
> #undef TARGET_VECTOR_MODE_SUPPORTED_ANY_TARGET_P
> #define TARGET_VECTOR_MODE_SUPPORTED_ANY_TARGET_P riscv_vector_mode_supported_any_target_p
> +#undef TARGET_FUNCTION_VALUE_REGNO_P
> +#define TARGET_FUNCTION_VALUE_REGNO_P riscv_function_value_regno_p
> +
> +#undef TARGET_GET_RAW_RESULT_MODE
> +#define TARGET_GET_RAW_RESULT_MODE riscv_get_raw_result_mode
> +
> struct gcc_target targetm = TARGET_INITIALIZER;
> #include "gt-riscv.h"
> diff --git a/gcc/config/riscv/riscv.h b/gcc/config/riscv/riscv.h
> index 269b8c1f076..7797e67317a 100644
> --- a/gcc/config/riscv/riscv.h
> +++ b/gcc/config/riscv/riscv.h
> @@ -683,6 +683,12 @@ enum reg_class
> #define GP_RETURN GP_ARG_FIRST
> #define FP_RETURN (UNITS_PER_FP_ARG == 0 ? GP_RETURN : FP_ARG_FIRST)
> +#define V_RETURN  V_REG_FIRST
> +
> +#define GP_RETURN_FIRST GP_ARG_FIRST
> +#define GP_RETURN_LAST  GP_ARG_FIRST + 1
> +#define FP_RETURN_FIRST FP_RETURN
> +#define FP_RETURN_LAST  FP_RETURN + 1
> #define MAX_ARGS_IN_REGISTERS \
>    (riscv_abi == ABI_ILP32E || riscv_abi == ABI_LP64E \
> @@ -714,8 +720,6 @@ enum reg_class
> #define FUNCTION_VALUE(VALTYPE, FUNC) \
>    riscv_function_value (VALTYPE, FUNC, VOIDmode)
> -#define FUNCTION_VALUE_REGNO_P(N) ((N) == GP_RETURN || (N) == FP_RETURN)
> -
> /* 1 if N is a possible register number for function argument passing.
>     We have no FP argument registers when soft-float.  */
> diff --git a/gcc/testsuite/g++.target/riscv/rvv/base/pr114639-1.C b/gcc/testsuite/g++.target/riscv/rvv/base/pr114639-1.C
> new file mode 100644
> index 00000000000..9450b108ae5
> --- /dev/null
> +++ b/gcc/testsuite/g++.target/riscv/rvv/base/pr114639-1.C
> @@ -0,0 +1,25 @@
> +/* Test that we do not have ice when compile */
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv64gcv -mabi=lp64d -O3" } */
> +
> +typedef long c;
> +
> +#pragma riscv intrinsic "vector"
> +
> +template <unsigned long> struct d {};
> +
> +struct e {
> +  using f = d<0>;
> +};
> +
> +struct g {
> +  using f = e::f;
> +};
> +
> +template <typename, int> using h = g::f;
> +template <unsigned long i> long get_vl (d<i>);
> +
> +vbool16_t test (vuint64m4_t a) {
> +  c b;
> +  return __riscv_vmsne_vx_u64m4_b16(a, b, get_vl (h<c, 2>()));
> +}
> diff --git a/gcc/testsuite/gcc.target/riscv/rvv/base/pr114639-1.c b/gcc/testsuite/gcc.target/riscv/rvv/base/pr114639-1.c
> new file mode 100644
> index 00000000000..3ad91dbf6bb
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/base/pr114639-1.c
> @@ -0,0 +1,14 @@
> +/* Test that we do not have ice when compile */
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv64gcv -mabi=lp64d -O3" } */
> +
> +#include <riscv_vector.h>
> +
> +extern size_t get_vl ();
> +
> +vbool16_t
> +test (vuint64m4_t a)
> +{
> +  unsigned long b;
> +  return __riscv_vmsne_vx_u64m4_b16 (a, b, get_vl ());
> +}
> --
> 2.34.1
>
>
Li, Pan2 April 11, 2024, 3:03 a.m. UTC | #3 
Committed, thanks Juzhe and Kito.

Pan

-----Original Message-----
From: Kito Cheng <kito.cheng@gmail.com> 
Sent: Thursday, April 11, 2024 10:50 AM
To: juzhe.zhong@rivai.ai
Cc: Li, Pan2 <pan2.li@intel.com>; gcc-patches <gcc-patches@gcc.gnu.org>
Subject: Re: [PATCH v1] RISC-V: Bugfix ICE for the vector return arg in mode switch

I was thinking we may guarded with TARGET_VECTOR and TARGET_HARD_FLOAT
or checking with ABI in riscv_function_value_regno_p, however I think
it's fine with current implementation (no checking) after checking all
use site of `targetm.calls.function_value_regno_p`, so LGTM :)

Thanks Pan for fixing this issue!

On Thu, Apr 11, 2024 at 10:23 AM juzhe.zhong@rivai.ai
<juzhe.zhong@rivai.ai> wrote:
>
> Thanks for fixing it. LGTM from my side.
>
> I prefer wait kito for another ACK.
>
> ________________________________
> juzhe.zhong@rivai.ai
>
>
> From: pan2.li
> Date: 2024-04-11 10:16
> To: gcc-patches
> CC: juzhe.zhong; kito.cheng; Pan Li
> Subject: [PATCH v1] RISC-V: Bugfix ICE for the vector return arg in mode switch
> From: Pan Li <pan2.li@intel.com>
>
> This patch would like to fix a ICE in mode sw for below example code.
>
> during RTL pass: mode_sw
> test.c: In function ‘vbool16_t j(vuint64m4_t)’:
> test.c:15:1: internal compiler error: in create_pre_exit, at
> mode-switching.cc:451
>    15 | }
>       | ^
> 0x3978f12 create_pre_exit
>         __RISCV_BUILD__/../gcc/mode-switching.cc:451
> 0x3979e9e optimize_mode_switching
>         __RISCV_BUILD__/../gcc/mode-switching.cc:849
> 0x397b9bc execute
>         __RISCV_BUILD__/../gcc/mode-switching.cc:1324
>
> extern size_t get_vl ();
>
> vbool16_t
> test (vuint64m4_t a)
> {
>   unsigned long b;
>   return __riscv_vmsne_vx_u64m4_b16 (a, b, get_vl ());
> }
>
> The create_pre_exit would like to find a return value copy.  If
> not, there will be a reason in assert but not available for above
> sample code when vector calling convension is enabled by default.
> This patch would like to override the TARGET_FUNCTION_VALUE_REGNO_P
> for vector register and then we will have hard_regno_nregs for copy_num,
> aka there is a return value copy.
>
> As a side-effect of allow vector in TARGET_FUNCTION_VALUE_REGNO_P, the
> TARGET_GET_RAW_RESULT_MODE will have vector mode and which is sizeless
> cannot be converted to fixed_size_mode.  Thus override the hook
> TARGET_GET_RAW_RESULT_MODE and return VOIDmode when the regno is-not-a
> fixed_size_mode.
>
> The below tests are passed for this patch.
> * The fully riscv regression tests.
> * The reproducing test in bugzilla PR114639.
>
> PR target/114639
>
> gcc/ChangeLog:
>
> * config/riscv/riscv.cc (riscv_function_value_regno_p): New func
> impl for hook TARGET_FUNCTION_VALUE_REGNO_P.
> (riscv_get_raw_result_mode): New func imple for hook
> TARGET_GET_RAW_RESULT_MODE.
> (TARGET_FUNCTION_VALUE_REGNO_P): Impl the hook.
> (TARGET_GET_RAW_RESULT_MODE): Ditto.
> * config/riscv/riscv.h (V_RETURN): New macro for vector return.
> (GP_RETURN_FIRST): New macro for the first GPR in return.
> (GP_RETURN_LAST): New macro for the last GPR in return.
> (FP_RETURN_FIRST): Diito but for FPR.
> (FP_RETURN_LAST): Ditto.
> (FUNCTION_VALUE_REGNO_P): Remove as deprecated and replace by
> TARGET_FUNCTION_VALUE_REGNO_P.
>
> gcc/testsuite/ChangeLog:
>
> * g++.target/riscv/rvv/base/pr114639-1.C: New test.
> * gcc.target/riscv/rvv/base/pr114639-1.c: New test.
>
> Signed-off-by: Pan Li <pan2.li@intel.com>
> ---
> gcc/config/riscv/riscv.cc                     | 34 +++++++++++++++++++
> gcc/config/riscv/riscv.h                      |  8 +++--
> .../g++.target/riscv/rvv/base/pr114639-1.C    | 25 ++++++++++++++
> .../gcc.target/riscv/rvv/base/pr114639-1.c    | 14 ++++++++
> 4 files changed, 79 insertions(+), 2 deletions(-)
> create mode 100644 gcc/testsuite/g++.target/riscv/rvv/base/pr114639-1.C
> create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/base/pr114639-1.c
>
> diff --git a/gcc/config/riscv/riscv.cc b/gcc/config/riscv/riscv.cc
> index 00defa69fd8..91f017dd52a 100644
> --- a/gcc/config/riscv/riscv.cc
> +++ b/gcc/config/riscv/riscv.cc
> @@ -10997,6 +10997,34 @@ riscv_vector_mode_supported_any_target_p (machine_mode)
>    return true;
> }
> +/* Implements hook TARGET_FUNCTION_VALUE_REGNO_P.  */
> +
> +static bool
> +riscv_function_value_regno_p (const unsigned regno)
> +{
> +  if (GP_RETURN_FIRST <= regno && regno <= GP_RETURN_LAST)
> +    return true;
> +
> +  if (FP_RETURN_FIRST <= regno && regno <= FP_RETURN_LAST)
> +    return true;
> +
> +  if (regno == V_RETURN)
> +    return true;
> +
> +  return false;
> +}
> +
> +/* Implements hook TARGET_GET_RAW_RESULT_MODE.  */
> +
> +static fixed_size_mode
> +riscv_get_raw_result_mode (int regno)
> +{
> +  if (!is_a <fixed_size_mode> (reg_raw_mode[regno]))
> +    return as_a <fixed_size_mode> (VOIDmode);
> +
> +  return default_get_reg_raw_mode (regno);
> +}
> +
> /* Initialize the GCC target structure.  */
> #undef TARGET_ASM_ALIGNED_HI_OP
> #define TARGET_ASM_ALIGNED_HI_OP "\t.half\t"
> @@ -11343,6 +11371,12 @@ riscv_vector_mode_supported_any_target_p (machine_mode)
> #undef TARGET_VECTOR_MODE_SUPPORTED_ANY_TARGET_P
> #define TARGET_VECTOR_MODE_SUPPORTED_ANY_TARGET_P riscv_vector_mode_supported_any_target_p
> +#undef TARGET_FUNCTION_VALUE_REGNO_P
> +#define TARGET_FUNCTION_VALUE_REGNO_P riscv_function_value_regno_p
> +
> +#undef TARGET_GET_RAW_RESULT_MODE
> +#define TARGET_GET_RAW_RESULT_MODE riscv_get_raw_result_mode
> +
> struct gcc_target targetm = TARGET_INITIALIZER;
> #include "gt-riscv.h"
> diff --git a/gcc/config/riscv/riscv.h b/gcc/config/riscv/riscv.h
> index 269b8c1f076..7797e67317a 100644
> --- a/gcc/config/riscv/riscv.h
> +++ b/gcc/config/riscv/riscv.h
> @@ -683,6 +683,12 @@ enum reg_class
> #define GP_RETURN GP_ARG_FIRST
> #define FP_RETURN (UNITS_PER_FP_ARG == 0 ? GP_RETURN : FP_ARG_FIRST)
> +#define V_RETURN  V_REG_FIRST
> +
> +#define GP_RETURN_FIRST GP_ARG_FIRST
> +#define GP_RETURN_LAST  GP_ARG_FIRST + 1
> +#define FP_RETURN_FIRST FP_RETURN
> +#define FP_RETURN_LAST  FP_RETURN + 1
> #define MAX_ARGS_IN_REGISTERS \
>    (riscv_abi == ABI_ILP32E || riscv_abi == ABI_LP64E \
> @@ -714,8 +720,6 @@ enum reg_class
> #define FUNCTION_VALUE(VALTYPE, FUNC) \
>    riscv_function_value (VALTYPE, FUNC, VOIDmode)
> -#define FUNCTION_VALUE_REGNO_P(N) ((N) == GP_RETURN || (N) == FP_RETURN)
> -
> /* 1 if N is a possible register number for function argument passing.
>     We have no FP argument registers when soft-float.  */
> diff --git a/gcc/testsuite/g++.target/riscv/rvv/base/pr114639-1.C b/gcc/testsuite/g++.target/riscv/rvv/base/pr114639-1.C
> new file mode 100644
> index 00000000000..9450b108ae5
> --- /dev/null
> +++ b/gcc/testsuite/g++.target/riscv/rvv/base/pr114639-1.C
> @@ -0,0 +1,25 @@
> +/* Test that we do not have ice when compile */
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv64gcv -mabi=lp64d -O3" } */
> +
> +typedef long c;
> +
> +#pragma riscv intrinsic "vector"
> +
> +template <unsigned long> struct d {};
> +
> +struct e {
> +  using f = d<0>;
> +};
> +
> +struct g {
> +  using f = e::f;
> +};
> +
> +template <typename, int> using h = g::f;
> +template <unsigned long i> long get_vl (d<i>);
> +
> +vbool16_t test (vuint64m4_t a) {
> +  c b;
> +  return __riscv_vmsne_vx_u64m4_b16(a, b, get_vl (h<c, 2>()));
> +}
> diff --git a/gcc/testsuite/gcc.target/riscv/rvv/base/pr114639-1.c b/gcc/testsuite/gcc.target/riscv/rvv/base/pr114639-1.c
> new file mode 100644
> index 00000000000..3ad91dbf6bb
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/base/pr114639-1.c
> @@ -0,0 +1,14 @@
> +/* Test that we do not have ice when compile */
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv64gcv -mabi=lp64d -O3" } */
> +
> +#include <riscv_vector.h>
> +
> +extern size_t get_vl ();
> +
> +vbool16_t
> +test (vuint64m4_t a)
> +{
> +  unsigned long b;
> +  return __riscv_vmsne_vx_u64m4_b16 (a, b, get_vl ());
> +}
> --
> 2.34.1
>
>
Bernd Edlinger April 11, 2024, 11:52 a.m. UTC | #4 
On 4/11/24 05:03, Li, Pan2 wrote:
> Committed, thanks Juzhe and Kito.
> 
> Pan


Hi Pan,

this commit caused a regression:

FAIL: gcc.c-torture/compile/930623-1.c   -O0  (test for excess errors)
FAIL: gcc.c-torture/compile/930623-1.c   -O1  (internal compiler error: in emit_vec_extract, at config/riscv/riscv-v.cc:5059)
FAIL: gcc.c-torture/compile/930623-1.c   -O1  (test for excess errors)
FAIL: gcc.c-torture/compile/930623-1.c   -O2  (internal compiler error: in emit_vec_extract, at config/riscv/riscv-v.cc:5059)
FAIL: gcc.c-torture/compile/930623-1.c   -O2  (test for excess errors)
FAIL: gcc.c-torture/compile/930623-1.c   -O3 -g  (internal compiler error: in emit_vec_extract, at config/riscv/riscv-v.cc:5059)
FAIL: gcc.c-torture/compile/930623-1.c   -O3 -g  (test for excess errors)
FAIL: gcc.c-torture/compile/930623-1.c   -Os  (internal compiler error: in emit_vec_extract, at config/riscv/riscv-v.cc:5059)
FAIL: gcc.c-torture/compile/930623-1.c   -Os  (test for excess errors)
FAIL: gcc.c-torture/compile/930623-1.c   -O2 -flto -fno-use-linker-plugin -flto-partition=none  (internal compiler error: in emit_vec_extract, at config/riscv/riscv-v.cc:5059)
FAIL: gcc.c-torture/compile/930623-1.c   -O2 -flto -fno-use-linker-plugin -flto-partition=none  (test for excess errors)

gcc/testsuite/gcc.c-torture/compile/930623-1.c:10:3: internal compiler error: in emit_vec_extract, at config/riscv/riscv-v.cc:5059^M
0xbba2de riscv_vector::emit_vec_extract(rtx_def*, rtx_def*, rtx_def*)^M
        ../../gcc-trunk/gcc/config/riscv/riscv-v.cc:5059^M
0x186945f riscv_legitimize_move(machine_mode, rtx_def*, rtx_def*)^M
        ../../gcc-trunk/gcc/config/riscv/riscv.cc:2895^M
0x1ef50b2 gen_movsi(rtx_def*, rtx_def*)^M
        ../../gcc-trunk/gcc/config/riscv/riscv.md:2225^M
0xffc91c rtx_insn* insn_gen_fn::operator()<rtx_def*, rtx_def*>(rtx_def*, rtx_def*) const^M
        ../../gcc-trunk/gcc/recog.h:441^M
0xffc91c emit_move_insn_1(rtx_def*, rtx_def*)^M
        ../../gcc-trunk/gcc/expr.cc:4551^M
0xffcdf4 emit_move_insn(rtx_def*, rtx_def*)^M
        ../../gcc-trunk/gcc/expr.cc:4721^M
0x1002f17 emit_move_multi_word^M
        ../../gcc-trunk/gcc/expr.cc:4517^M
0xffcdf4 emit_move_insn(rtx_def*, rtx_def*)^M
        ../../gcc-trunk/gcc/expr.cc:4721^M
0x1efc6b7 gen_untyped_call(rtx_def*, rtx_def*, rtx_def*)^M
        ../../gcc-trunk/gcc/config/riscv/riscv.md:3478^M
0x185fc7c target_gen_untyped_call^M
        ../../gcc-trunk/gcc/config/riscv/riscv.md:3453^M
0xe8e81f expand_builtin_apply^M
        ../../gcc-trunk/gcc/builtins.cc:1761^M
0xea053c expand_builtin(tree_node*, rtx_def*, rtx_def*, machine_mode, int)^M
        ../../gcc-trunk/gcc/builtins.cc:8001^M
0xff9e27 expand_expr_real_1(tree_node*, rtx_def*, machine_mode, expand_modifier, rtx_def**, bool)^M
        ../../gcc-trunk/gcc/expr.cc:12353^M
0xec4c3d expand_expr(tree_node*, rtx_def*, machine_mode, expand_modifier)^M
        ../../gcc-trunk/gcc/expr.h:316^M
0xec4c3d expand_call_stmt^M
        ../../gcc-trunk/gcc/cfgexpand.cc:2865^M
0xec4c3d expand_gimple_stmt_1^M
        ../../gcc-trunk/gcc/cfgexpand.cc:3932^M
0xec4c3d expand_gimple_stmt^M
        ../../gcc-trunk/gcc/cfgexpand.cc:4077^M
0xeca206 expand_gimple_basic_block^M
        ../../gcc-trunk/gcc/cfgexpand.cc:6133^M
0xecc287 execute^M
        ../../gcc-trunk/gcc/cfgexpand.cc:6872^M
Please submit a full bug report, with preprocessed source (by using -freport-bug).^M
Please include the complete backtrace with any bug report.^M
See <https://gcc.gnu.org/bugs/> for instructions.^M
compiler exited with status 1

I've built the git revision f3fdcf4a37a with 
../gcc-trunk/configure --target=riscv-unknown-elf --prefix=/home/ed/gnu/riscv-unknown-elf --enable-languages=c,c++ --disable-multilib --with-arch=rv32imac --with-abi=ilp32

I am a bit surprised since the target is not supposed to support floating point
or vector instructions AFAIK.

The issue does not happen with gcc-trunk from yesterday.

Regards
Bernd.
Li, Pan2 April 11, 2024, 12:45 p.m. UTC | #5 
Thanks for reporting this. Just take a look from my test log that 930623-1.c is all pass.

Thus I bet this difference comes from the build option --with-arch=rv32imac but my test script take rv64gcv.

> I've built the git revision f3fdcf4a37a with 
> ../gcc-trunk/configure --target=riscv-unknown-elf --prefix=/home/ed/gnu/riscv-unknown-elf --enable-languages=c,c++ --disable-multilib --with-arch=rv32imac --with-abi=ilp32

> I am a bit surprised since the target is not supposed to support floating point
> or vector instructions AFAIK.

Because you specify rv32imac, with doesn't include f/d/v extension, aka single/double floating point and vector extension. Thus, related functionality are disabled.

> The issue does not happen with gcc-trunk from yesterday.

Ack, will look into it.

Pan

-----Original Message-----
From: Bernd Edlinger <bernd.edlinger@hotmail.de> 
Sent: Thursday, April 11, 2024 7:52 PM
To: Li, Pan2 <pan2.li@intel.com>; Kito Cheng <kito.cheng@gmail.com>; juzhe.zhong@rivai.ai
Cc: gcc-patches <gcc-patches@gcc.gnu.org>
Subject: Re: [PATCH v1] RISC-V: Bugfix ICE for the vector return arg in mode switch

On 4/11/24 05:03, Li, Pan2 wrote:
> Committed, thanks Juzhe and Kito.
> 
> Pan


Hi Pan,

this commit caused a regression:

FAIL: gcc.c-torture/compile/930623-1.c   -O0  (test for excess errors)
FAIL: gcc.c-torture/compile/930623-1.c   -O1  (internal compiler error: in emit_vec_extract, at config/riscv/riscv-v.cc:5059)
FAIL: gcc.c-torture/compile/930623-1.c   -O1  (test for excess errors)
FAIL: gcc.c-torture/compile/930623-1.c   -O2  (internal compiler error: in emit_vec_extract, at config/riscv/riscv-v.cc:5059)
FAIL: gcc.c-torture/compile/930623-1.c   -O2  (test for excess errors)
FAIL: gcc.c-torture/compile/930623-1.c   -O3 -g  (internal compiler error: in emit_vec_extract, at config/riscv/riscv-v.cc:5059)
FAIL: gcc.c-torture/compile/930623-1.c   -O3 -g  (test for excess errors)
FAIL: gcc.c-torture/compile/930623-1.c   -Os  (internal compiler error: in emit_vec_extract, at config/riscv/riscv-v.cc:5059)
FAIL: gcc.c-torture/compile/930623-1.c   -Os  (test for excess errors)
FAIL: gcc.c-torture/compile/930623-1.c   -O2 -flto -fno-use-linker-plugin -flto-partition=none  (internal compiler error: in emit_vec_extract, at config/riscv/riscv-v.cc:5059)
FAIL: gcc.c-torture/compile/930623-1.c   -O2 -flto -fno-use-linker-plugin -flto-partition=none  (test for excess errors)

gcc/testsuite/gcc.c-torture/compile/930623-1.c:10:3: internal compiler error: in emit_vec_extract, at config/riscv/riscv-v.cc:5059^M
0xbba2de riscv_vector::emit_vec_extract(rtx_def*, rtx_def*, rtx_def*)^M
        ../../gcc-trunk/gcc/config/riscv/riscv-v.cc:5059^M
0x186945f riscv_legitimize_move(machine_mode, rtx_def*, rtx_def*)^M
        ../../gcc-trunk/gcc/config/riscv/riscv.cc:2895^M
0x1ef50b2 gen_movsi(rtx_def*, rtx_def*)^M
        ../../gcc-trunk/gcc/config/riscv/riscv.md:2225^M
0xffc91c rtx_insn* insn_gen_fn::operator()<rtx_def*, rtx_def*>(rtx_def*, rtx_def*) const^M
        ../../gcc-trunk/gcc/recog.h:441^M
0xffc91c emit_move_insn_1(rtx_def*, rtx_def*)^M
        ../../gcc-trunk/gcc/expr.cc:4551^M
0xffcdf4 emit_move_insn(rtx_def*, rtx_def*)^M
        ../../gcc-trunk/gcc/expr.cc:4721^M
0x1002f17 emit_move_multi_word^M
        ../../gcc-trunk/gcc/expr.cc:4517^M
0xffcdf4 emit_move_insn(rtx_def*, rtx_def*)^M
        ../../gcc-trunk/gcc/expr.cc:4721^M
0x1efc6b7 gen_untyped_call(rtx_def*, rtx_def*, rtx_def*)^M
        ../../gcc-trunk/gcc/config/riscv/riscv.md:3478^M
0x185fc7c target_gen_untyped_call^M
        ../../gcc-trunk/gcc/config/riscv/riscv.md:3453^M
0xe8e81f expand_builtin_apply^M
        ../../gcc-trunk/gcc/builtins.cc:1761^M
0xea053c expand_builtin(tree_node*, rtx_def*, rtx_def*, machine_mode, int)^M
        ../../gcc-trunk/gcc/builtins.cc:8001^M
0xff9e27 expand_expr_real_1(tree_node*, rtx_def*, machine_mode, expand_modifier, rtx_def**, bool)^M
        ../../gcc-trunk/gcc/expr.cc:12353^M
0xec4c3d expand_expr(tree_node*, rtx_def*, machine_mode, expand_modifier)^M
        ../../gcc-trunk/gcc/expr.h:316^M
0xec4c3d expand_call_stmt^M
        ../../gcc-trunk/gcc/cfgexpand.cc:2865^M
0xec4c3d expand_gimple_stmt_1^M
        ../../gcc-trunk/gcc/cfgexpand.cc:3932^M
0xec4c3d expand_gimple_stmt^M
        ../../gcc-trunk/gcc/cfgexpand.cc:4077^M
0xeca206 expand_gimple_basic_block^M
        ../../gcc-trunk/gcc/cfgexpand.cc:6133^M
0xecc287 execute^M
        ../../gcc-trunk/gcc/cfgexpand.cc:6872^M
Please submit a full bug report, with preprocessed source (by using -freport-bug).^M
Please include the complete backtrace with any bug report.^M
See <https://gcc.gnu.org/bugs/> for instructions.^M
compiler exited with status 1

I've built the git revision f3fdcf4a37a with 
../gcc-trunk/configure --target=riscv-unknown-elf --prefix=/home/ed/gnu/riscv-unknown-elf --enable-languages=c,c++ --disable-multilib --with-arch=rv32imac --with-abi=ilp32

I am a bit surprised since the target is not supposed to support floating point
or vector instructions AFAIK.

The issue does not happen with gcc-trunk from yesterday.

Regards
Bernd.
Edwin Lu April 11, 2024, 9:19 p.m. UTC | #6 
On 4/11/2024 5:45 AM, Li, Pan2 wrote:
> Thanks for reporting this. Just take a look from my test log that 930623-1.c is all pass.
> 
> Thus I bet this difference comes from the build option --with-arch=rv32imac but my test script take rv64gcv.
> 
>> I've built the git revision f3fdcf4a37a with
>> ../gcc-trunk/configure --target=riscv-unknown-elf --prefix=/home/ed/gnu/riscv-unknown-elf --enable-languages=c,c++ --disable-multilib --with-arch=rv32imac --with-abi=ilp32
> 
>> I am a bit surprised since the target is not supposed to support floating point
>> or vector instructions AFAIK.
> 
> Because you specify rv32imac, with doesn't include f/d/v extension, aka single/double floating point and vector extension. Thus, related functionality are disabled.
> 
>> The issue does not happen with gcc-trunk from yesterday.
> 
> Ack, will look into it.
> 
> Pan
>  
Hi Pan,

Our postcommit-ci found that it breaks for non-vector targets on rv32/64 
newlib/linux https://github.com/patrick-rivos/gcc-postcommit-ci/issues/757.

The patchwork precommit-ci also appeared to have flagged it 
https://github.com/ewlu/gcc-precommit-ci/issues/1417#issuecomment-2048846532

Edwin
Li, Pan2 April 12, 2024, 1:38 a.m. UTC | #7 
Thanks Edwin, should be one silly mistake, will fix it ASAP.

Pan

-----Original Message-----
From: Edwin Lu <ewlu@rivosinc.com> 
Sent: Friday, April 12, 2024 5:20 AM
To: Li, Pan2 <pan2.li@intel.com>; Bernd Edlinger <bernd.edlinger@hotmail.de>; Kito Cheng <kito.cheng@gmail.com>; juzhe.zhong@rivai.ai
Cc: gcc-patches <gcc-patches@gcc.gnu.org>
Subject: Re: [PATCH v1] RISC-V: Bugfix ICE for the vector return arg in mode switch

On 4/11/2024 5:45 AM, Li, Pan2 wrote:
> Thanks for reporting this. Just take a look from my test log that 930623-1.c is all pass.
> 
> Thus I bet this difference comes from the build option --with-arch=rv32imac but my test script take rv64gcv.
> 
>> I've built the git revision f3fdcf4a37a with
>> ../gcc-trunk/configure --target=riscv-unknown-elf --prefix=/home/ed/gnu/riscv-unknown-elf --enable-languages=c,c++ --disable-multilib --with-arch=rv32imac --with-abi=ilp32
> 
>> I am a bit surprised since the target is not supposed to support floating point
>> or vector instructions AFAIK.
> 
> Because you specify rv32imac, with doesn't include f/d/v extension, aka single/double floating point and vector extension. Thus, related functionality are disabled.
> 
>> The issue does not happen with gcc-trunk from yesterday.
> 
> Ack, will look into it.
> 
> Pan
>  
Hi Pan,

Our postcommit-ci found that it breaks for non-vector targets on rv32/64 
newlib/linux https://github.com/patrick-rivos/gcc-postcommit-ci/issues/757.

The patchwork precommit-ci also appeared to have flagged it 
https://github.com/ewlu/gcc-precommit-ci/issues/1417#issuecomment-2048846532

Edwin
diff mbox series

Patch

diff --git a/gcc/config/riscv/riscv.cc b/gcc/config/riscv/riscv.cc
index 00defa69fd8..91f017dd52a 100644
--- a/gcc/config/riscv/riscv.cc
+++ b/gcc/config/riscv/riscv.cc
@@ -10997,6 +10997,34 @@  riscv_vector_mode_supported_any_target_p (machine_mode)
   return true;
 }
 
+/* Implements hook TARGET_FUNCTION_VALUE_REGNO_P.  */
+
+static bool
+riscv_function_value_regno_p (const unsigned regno)
+{
+  if (GP_RETURN_FIRST <= regno && regno <= GP_RETURN_LAST)
+    return true;
+
+  if (FP_RETURN_FIRST <= regno && regno <= FP_RETURN_LAST)
+    return true;
+
+  if (regno == V_RETURN)
+    return true;
+
+  return false;
+}
+
+/* Implements hook TARGET_GET_RAW_RESULT_MODE.  */
+
+static fixed_size_mode
+riscv_get_raw_result_mode (int regno)
+{
+  if (!is_a <fixed_size_mode> (reg_raw_mode[regno]))
+    return as_a <fixed_size_mode> (VOIDmode);
+
+  return default_get_reg_raw_mode (regno);
+}
+
 /* Initialize the GCC target structure.  */
 #undef TARGET_ASM_ALIGNED_HI_OP
 #define TARGET_ASM_ALIGNED_HI_OP "\t.half\t"
@@ -11343,6 +11371,12 @@  riscv_vector_mode_supported_any_target_p (machine_mode)
 #undef TARGET_VECTOR_MODE_SUPPORTED_ANY_TARGET_P
 #define TARGET_VECTOR_MODE_SUPPORTED_ANY_TARGET_P riscv_vector_mode_supported_any_target_p
 
+#undef TARGET_FUNCTION_VALUE_REGNO_P
+#define TARGET_FUNCTION_VALUE_REGNO_P riscv_function_value_regno_p
+
+#undef TARGET_GET_RAW_RESULT_MODE
+#define TARGET_GET_RAW_RESULT_MODE riscv_get_raw_result_mode
+
 struct gcc_target targetm = TARGET_INITIALIZER;
 
 #include "gt-riscv.h"
diff --git a/gcc/config/riscv/riscv.h b/gcc/config/riscv/riscv.h
index 269b8c1f076..7797e67317a 100644
--- a/gcc/config/riscv/riscv.h
+++ b/gcc/config/riscv/riscv.h
@@ -683,6 +683,12 @@  enum reg_class
 
 #define GP_RETURN GP_ARG_FIRST
 #define FP_RETURN (UNITS_PER_FP_ARG == 0 ? GP_RETURN : FP_ARG_FIRST)
+#define V_RETURN  V_REG_FIRST
+
+#define GP_RETURN_FIRST GP_ARG_FIRST
+#define GP_RETURN_LAST  GP_ARG_FIRST + 1
+#define FP_RETURN_FIRST FP_RETURN
+#define FP_RETURN_LAST  FP_RETURN + 1
 
 #define MAX_ARGS_IN_REGISTERS \
   (riscv_abi == ABI_ILP32E || riscv_abi == ABI_LP64E \
@@ -714,8 +720,6 @@  enum reg_class
 #define FUNCTION_VALUE(VALTYPE, FUNC) \
   riscv_function_value (VALTYPE, FUNC, VOIDmode)
 
-#define FUNCTION_VALUE_REGNO_P(N) ((N) == GP_RETURN || (N) == FP_RETURN)
-
 /* 1 if N is a possible register number for function argument passing.
    We have no FP argument registers when soft-float.  */
 
diff --git a/gcc/testsuite/g++.target/riscv/rvv/base/pr114639-1.C b/gcc/testsuite/g++.target/riscv/rvv/base/pr114639-1.C
new file mode 100644
index 00000000000..9450b108ae5
--- /dev/null
+++ b/gcc/testsuite/g++.target/riscv/rvv/base/pr114639-1.C
@@ -0,0 +1,25 @@ 
+/* Test that we do not have ice when compile */
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv -mabi=lp64d -O3" } */
+
+typedef long c;
+
+#pragma riscv intrinsic "vector"
+
+template <unsigned long> struct d {};
+
+struct e {
+  using f = d<0>;
+};
+
+struct g {
+  using f = e::f;
+};
+
+template <typename, int> using h = g::f;
+template <unsigned long i> long get_vl (d<i>);
+
+vbool16_t test (vuint64m4_t a) {
+  c b;
+  return __riscv_vmsne_vx_u64m4_b16(a, b, get_vl (h<c, 2>()));
+}
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/base/pr114639-1.c b/gcc/testsuite/gcc.target/riscv/rvv/base/pr114639-1.c
new file mode 100644
index 00000000000..3ad91dbf6bb
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/base/pr114639-1.c
@@ -0,0 +1,14 @@ 
+/* Test that we do not have ice when compile */
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv -mabi=lp64d -O3" } */
+
+#include <riscv_vector.h>
+
+extern size_t get_vl ();
+
+vbool16_t
+test (vuint64m4_t a)
+{
+  unsigned long b;
+  return __riscv_vmsne_vx_u64m4_b16 (a, b, get_vl ());
+}