Message ID | 20230815025525.3437008-1-pan2.li@intel.com |
---|---|
State | New |
Headers | show |
Series | [v1] RISC-V: Support RVV VFCVT.X.F.V rounding mode intrinsic API | expand |
For conversion API, I would prefer kito review since I am not sure about this stuff. juzhe.zhong@rivai.ai From: pan2.li Date: 2023-08-15 10:55 To: gcc-patches CC: juzhe.zhong; pan2.li; yanzhang.wang; kito.cheng Subject: [PATCH v1] RISC-V: Support RVV VFCVT.X.F.V rounding mode intrinsic API From: Pan Li <pan2.li@intel.com> This patch would like to support the rounding mode API for the VFCVT.X.F.V as the below samples. * __riscv_vfcvt_x_f_v_i32m1_rm * __riscv_vfcvt_x_f_v_i32m1_rm_m Signed-off-by: Pan Li <pan2.li@intel.com> gcc/ChangeLog: * config/riscv/riscv-vector-builtins-bases.cc (class vfcvt_x_frm): New class for frm. (vfcvt_x_frm_obj): New declaration. (BASE): Ditto. * config/riscv/riscv-vector-builtins-bases.h: Ditto. * config/riscv/riscv-vector-builtins-functions.def (vfcvt_x_frm): New intrinsic function definition. gcc/testsuite/ChangeLog: * gcc.target/riscv/rvv/base/float-point-cvt-x.c: New test. --- .../riscv/riscv-vector-builtins-bases.cc | 17 +++++++++++ .../riscv/riscv-vector-builtins-bases.h | 1 + .../riscv/riscv-vector-builtins-functions.def | 2 ++ .../riscv/rvv/base/float-point-cvt-x.c | 29 +++++++++++++++++++ 4 files changed, 49 insertions(+) create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/base/float-point-cvt-x.c diff --git a/gcc/config/riscv/riscv-vector-builtins-bases.cc b/gcc/config/riscv/riscv-vector-builtins-bases.cc index f2124080ef9..754a53efd3d 100644 --- a/gcc/config/riscv/riscv-vector-builtins-bases.cc +++ b/gcc/config/riscv/riscv-vector-builtins-bases.cc @@ -660,6 +660,21 @@ public: } }; +/* Implements below instructions for frm + - vfcvt_x +*/ +template<int UNSPEC> +class vfcvt_x_frm : public function_base +{ +public: + bool has_rounding_mode_operand_p () const override { return true; } + + rtx expand (function_expander &e) const override + { + return e.use_exact_insn (code_for_pred_fcvt_x_f (UNSPEC, e.arg_mode (0))); + } +}; + /* Implements vrsub. */ class vrsub : public function_base { @@ -2465,6 +2480,7 @@ static CONSTEXPR const vfclass vfclass_obj; static CONSTEXPR const vmerge vfmerge_obj; static CONSTEXPR const vmv_v vfmv_v_obj; static CONSTEXPR const vfcvt_x<UNSPEC_VFCVT> vfcvt_x_obj; +static CONSTEXPR const vfcvt_x_frm<UNSPEC_VFCVT> vfcvt_x_frm_obj; static CONSTEXPR const vfcvt_x<UNSPEC_UNSIGNED_VFCVT> vfcvt_xu_obj; static CONSTEXPR const vfcvt_rtz_x<FIX> vfcvt_rtz_x_obj; static CONSTEXPR const vfcvt_rtz_x<UNSIGNED_FIX> vfcvt_rtz_xu_obj; @@ -2714,6 +2730,7 @@ BASE (vfclass) BASE (vfmerge) BASE (vfmv_v) BASE (vfcvt_x) +BASE (vfcvt_x_frm) BASE (vfcvt_xu) BASE (vfcvt_rtz_x) BASE (vfcvt_rtz_xu) diff --git a/gcc/config/riscv/riscv-vector-builtins-bases.h b/gcc/config/riscv/riscv-vector-builtins-bases.h index 2a9381eec5e..50a7d7ffb6f 100644 --- a/gcc/config/riscv/riscv-vector-builtins-bases.h +++ b/gcc/config/riscv/riscv-vector-builtins-bases.h @@ -205,6 +205,7 @@ extern const function_base *const vfclass; extern const function_base *const vfmerge; extern const function_base *const vfmv_v; extern const function_base *const vfcvt_x; +extern const function_base *const vfcvt_x_frm; extern const function_base *const vfcvt_xu; extern const function_base *const vfcvt_rtz_x; extern const function_base *const vfcvt_rtz_xu; diff --git a/gcc/config/riscv/riscv-vector-builtins-functions.def b/gcc/config/riscv/riscv-vector-builtins-functions.def index 34def6bb82f..8b6a7cc49f3 100644 --- a/gcc/config/riscv/riscv-vector-builtins-functions.def +++ b/gcc/config/riscv/riscv-vector-builtins-functions.def @@ -445,6 +445,8 @@ DEF_RVV_FUNCTION (vfcvt_rtz_xu, alu, full_preds, f_to_u_f_v_ops) DEF_RVV_FUNCTION (vfcvt_f, alu, full_preds, i_to_f_x_v_ops) DEF_RVV_FUNCTION (vfcvt_f, alu, full_preds, u_to_f_xu_v_ops) +DEF_RVV_FUNCTION (vfcvt_x_frm, alu_frm, full_preds, f_to_i_f_v_ops) + // 13.18. Widening Floating-Point/Integer Type-Convert Instructions DEF_RVV_FUNCTION (vfwcvt_x, alu, full_preds, f_to_wi_f_v_ops) DEF_RVV_FUNCTION (vfwcvt_xu, alu, full_preds, f_to_wu_f_v_ops) diff --git a/gcc/testsuite/gcc.target/riscv/rvv/base/float-point-cvt-x.c b/gcc/testsuite/gcc.target/riscv/rvv/base/float-point-cvt-x.c new file mode 100644 index 00000000000..e090f0f97e9 --- /dev/null +++ b/gcc/testsuite/gcc.target/riscv/rvv/base/float-point-cvt-x.c @@ -0,0 +1,29 @@ +/* { dg-do compile } */ +/* { dg-options "-march=rv64gcv -mabi=lp64 -O3 -Wno-psabi" } */ + +#include "riscv_vector.h" + +vint32m1_t +test_riscv_vfcvt_x_f_vv_i32m1_rm (vfloat32m1_t op1, size_t vl) { + return __riscv_vfcvt_x_f_v_i32m1_rm (op1, 0, vl); +} + +vint32m1_t +test_vfcvt_x_f_vv_i32m1_rm_m (vbool32_t mask, vfloat32m1_t op1, size_t vl) { + return __riscv_vfcvt_x_f_v_i32m1_rm_m (mask, op1, 1, vl); +} + +vint32m1_t +test_riscv_vfcvt_x_f_vv_i32m1 (vfloat32m1_t op1, size_t vl) { + return __riscv_vfcvt_x_f_v_i32m1 (op1, vl); +} + +vint32m1_t +test_vfcvt_x_f_vv_i32m1_m (vbool32_t mask, vfloat32m1_t op1, size_t vl) { + return __riscv_vfcvt_x_f_v_i32m1_m (mask, op1, vl); +} + +/* { dg-final { scan-assembler-times {vfcvt\.x\.f\.v\s+v[0-9]+,\s*v[0-9]+} 4 } } */ +/* { dg-final { scan-assembler-times {frrm\s+[axs][0-9]+} 2 } } */ +/* { dg-final { scan-assembler-times {fsrm\s+[axs][0-9]+} 2 } } */ +/* { dg-final { scan-assembler-times {fsrmi\s+[01234]} 2 } } */
Just a random idea came to my mind, maybe we could introduce one more template argument to reduce those codes for rounding mode intrinsic stuff? example: diff --git a/gcc/config/riscv/riscv-vector-builtins-bases.cc b/gcc/config/riscv/riscv-vector-builtins-bases.cc index 2074dac0f16..9cc60842a5b 100644 --- a/gcc/config/riscv/riscv-vector-builtins-bases.cc +++ b/gcc/config/riscv/riscv-vector-builtins-bases.cc @@ -1648,10 +1648,11 @@ public: }; /* Implements vfcvt.x. */ -template<int UNSPEC> +template<int UNSPEC, bool HAS_FRM = false> class vfcvt_x : public function_base { public: + bool has_rounding_mode_operand_p () const override { return HAS_FRM; } rtx expand (function_expander &e) const override { return e.use_exact_insn (code_for_pred_fcvt_x_f (UNSPEC, e.arg_mode (0))); @@ -2451,6 +2452,7 @@ static CONSTEXPR const vmerge vfmerge_obj; static CONSTEXPR const vmv_v vfmv_v_obj; static CONSTEXPR const vfcvt_x<UNSPEC_VFCVT> vfcvt_x_obj; static CONSTEXPR const vfcvt_x<UNSPEC_UNSIGNED_VFCVT> vfcvt_xu_obj; +static CONSTEXPR const vfcvt_x<UNSPEC_VFCVT, /*HAS_FRM*/true> vfcvt_x_frm_obj; static CONSTEXPR const vfcvt_rtz_x<FIX> vfcvt_rtz_x_obj; static CONSTEXPR const vfcvt_rtz_x<UNSIGNED_FIX> vfcvt_rtz_xu_obj; static CONSTEXPR const vfcvt_f vfcvt_f_obj;
Thanks Kito for comments. How about leverage inherit instead of template? AFAIK, the bool argument isn't recommended up to a point.
For example, as below to reuse the expand part.
class vfcvt_x : public function_base
{
public:
+ virtual bool has_rounding_mode_operand_p () const { return false; }
+
rtx expand (function_expander &e) const override
{
return e.use_exact_insn (code_for_pred_fcvt_x_f (UNSPEC, e.arg_mode (0)));
}
};
+/* Implements below instructions for frm
+ - vfcvt_x
+*/
+template<int UNSPEC>
+class vfcvt_x_frm : public vfcvt_x<UNSPEC>
+{
+public:
+ bool has_rounding_mode_operand_p () const override { return true; }
+};
Pan
-----Original Message-----
From: Kito Cheng <kito.cheng@gmail.com>
Sent: Tuesday, August 15, 2023 11:34 PM
To: juzhe.zhong@rivai.ai
Cc: Li, Pan2 <pan2.li@intel.com>; gcc-patches <gcc-patches@gcc.gnu.org>; Wang, Yanzhang <yanzhang.wang@intel.com>
Subject: Re: [PATCH v1] RISC-V: Support RVV VFCVT.X.F.V rounding mode intrinsic API
Just a random idea came to my mind, maybe we could introduce one more
template argument to reduce those codes for rounding mode intrinsic
stuff?
example:
diff --git a/gcc/config/riscv/riscv-vector-builtins-bases.cc
b/gcc/config/riscv/riscv-vector-builtins-bases.cc
index 2074dac0f16..9cc60842a5b 100644
--- a/gcc/config/riscv/riscv-vector-builtins-bases.cc
+++ b/gcc/config/riscv/riscv-vector-builtins-bases.cc
@@ -1648,10 +1648,11 @@ public:
};
/* Implements vfcvt.x. */
-template<int UNSPEC>
+template<int UNSPEC, bool HAS_FRM = false>
class vfcvt_x : public function_base
{
public:
+ bool has_rounding_mode_operand_p () const override { return HAS_FRM; }
rtx expand (function_expander &e) const override
{
return e.use_exact_insn (code_for_pred_fcvt_x_f (UNSPEC, e.arg_mode (0)));
@@ -2451,6 +2452,7 @@ static CONSTEXPR const vmerge vfmerge_obj;
static CONSTEXPR const vmv_v vfmv_v_obj;
static CONSTEXPR const vfcvt_x<UNSPEC_VFCVT> vfcvt_x_obj;
static CONSTEXPR const vfcvt_x<UNSPEC_UNSIGNED_VFCVT> vfcvt_xu_obj;
+static CONSTEXPR const vfcvt_x<UNSPEC_VFCVT, /*HAS_FRM*/true> vfcvt_x_frm_obj;
static CONSTEXPR const vfcvt_rtz_x<FIX> vfcvt_rtz_x_obj;
static CONSTEXPR const vfcvt_rtz_x<UNSIGNED_FIX> vfcvt_rtz_xu_obj;
static CONSTEXPR const vfcvt_f vfcvt_f_obj;
Or using an enum value rather than bool? I am thinking we could also simplify/remove most other frm classes, some practical example: diff --git a/gcc/config/riscv/riscv-vector-builtins-bases.cc b/gcc/config/riscv/riscv-vector-builtins-bases.cc index 2074dac0f16..ace63e963a5 100644 --- a/gcc/config/riscv/riscv-vector-builtins-bases.cc +++ b/gcc/config/riscv/riscv-vector-builtins-bases.cc @@ -58,6 +58,11 @@ enum lst_type LST_INDEXED, }; +enum frm_op_type +{ + NO_FRM, + HAS_FRM +}; /* Helper function to fold vleff and vlsegff. */ static gimple * fold_fault_load (gimple_folder &f) @@ -256,41 +261,22 @@ public: vremu/vsadd/vsaddu/vssub/vssubu vfadd/vfsub/ */ -template<rtx_code CODE> +template<rtx_code CODE, enum frm_op_type FRM_OP = NO_FRM> class binop : public function_base { public: - rtx expand (function_expander &e) const override + bool has_rounding_mode_operand_p () const override { - switch (e.op_info->op) - { - case OP_TYPE_vx: - case OP_TYPE_vf: - return e.use_exact_insn (code_for_pred_scalar (CODE, e.vector_mode ())); - case OP_TYPE_vv: - return e.use_exact_insn (code_for_pred (CODE, e.vector_mode ())); - default: - gcc_unreachable (); - } + return FRM_OP == HAS_FRM; } -}; - -/* Implements below instructions for now. - - vfadd - - vfsub - - vfmul - - vfdiv -*/ -template<rtx_code CODE> -class binop_frm : public function_base -{ -public: - bool has_rounding_mode_operand_p () const override { return true; } rtx expand (function_expander &e) const override { switch (e.op_info->op) { + case OP_TYPE_vx: + gcc_assert (FRM_OP == NO_FRM); + gcc_fallthrough (); case OP_TYPE_vf: return e.use_exact_insn (code_for_pred_scalar (CODE, e.vector_mode ())); case OP_TYPE_vv: @@ -1648,10 +1634,15 @@ public: }; /* Implements vfcvt.x. */ -template<int UNSPEC> +template<int UNSPEC, enum frm_op_type FRM_OP = NO_FRM> class vfcvt_x : public function_base { public: + bool has_rounding_mode_operand_p () const override + { + return FRM_OP == HAS_FRM; + } + rtx expand (function_expander &e) const override { return e.use_exact_insn (code_for_pred_fcvt_x_f (UNSPEC, e.arg_mode (0))); @@ -2389,8 +2380,8 @@ static CONSTEXPR const viota viota_obj; static CONSTEXPR const vid vid_obj; static CONSTEXPR const binop<PLUS> vfadd_obj; static CONSTEXPR const binop<MINUS> vfsub_obj; -static CONSTEXPR const binop_frm<PLUS> vfadd_frm_obj; -static CONSTEXPR const binop_frm<MINUS> vfsub_frm_obj; +static CONSTEXPR const binop<PLUS, HAS_FRM> vfadd_frm_obj; +static CONSTEXPR const binop<MINUS, HAS_FRM> vfsub_frm_obj; static CONSTEXPR const reverse_binop<MINUS> vfrsub_obj; static CONSTEXPR const reverse_binop_frm<MINUS> vfrsub_frm_obj; static CONSTEXPR const widen_binop<PLUS> vfwadd_obj; @@ -2398,9 +2389,9 @@ static CONSTEXPR const widen_binop_frm<PLUS> vfwadd_frm_obj; static CONSTEXPR const widen_binop<MINUS> vfwsub_obj; static CONSTEXPR const widen_binop_frm<MINUS> vfwsub_frm_obj; static CONSTEXPR const binop<MULT> vfmul_obj; -static CONSTEXPR const binop_frm<MULT> vfmul_frm_obj; +static CONSTEXPR const binop<MULT, HAS_FRM> vfmul_frm_obj; static CONSTEXPR const binop<DIV> vfdiv_obj; -static CONSTEXPR const binop_frm<DIV> vfdiv_frm_obj; +static CONSTEXPR const binop<DIV, HAS_FRM> vfdiv_frm_obj; static CONSTEXPR const reverse_binop<DIV> vfrdiv_obj; static CONSTEXPR const reverse_binop_frm<DIV> vfrdiv_frm_obj; static CONSTEXPR const widen_binop<MULT> vfwmul_obj;
That should work as well, but may require some changes to existing codes like declaration, etc.
I am OK for both the enum or inherit, and will start with the CVT parts, then refactor the existing frm class.
Do you have any suggestion for the decision making?
Pan
-----Original Message-----
From: Kito Cheng <kito.cheng@gmail.com>
Sent: Wednesday, August 16, 2023 11:30 AM
To: Li, Pan2 <pan2.li@intel.com>
Cc: juzhe.zhong@rivai.ai; gcc-patches <gcc-patches@gcc.gnu.org>; Wang, Yanzhang <yanzhang.wang@intel.com>
Subject: Re: [PATCH v1] RISC-V: Support RVV VFCVT.X.F.V rounding mode intrinsic API
Or using an enum value rather than bool?
I am thinking we could also simplify/remove most other frm classes,
some practical example:
diff --git a/gcc/config/riscv/riscv-vector-builtins-bases.cc
b/gcc/config/riscv/riscv-vector-builtins-bases.cc
index 2074dac0f16..ace63e963a5 100644
--- a/gcc/config/riscv/riscv-vector-builtins-bases.cc
+++ b/gcc/config/riscv/riscv-vector-builtins-bases.cc
@@ -58,6 +58,11 @@ enum lst_type
LST_INDEXED,
};
+enum frm_op_type
+{
+ NO_FRM,
+ HAS_FRM
+};
/* Helper function to fold vleff and vlsegff. */
static gimple *
fold_fault_load (gimple_folder &f)
@@ -256,41 +261,22 @@ public:
vremu/vsadd/vsaddu/vssub/vssubu
vfadd/vfsub/
*/
-template<rtx_code CODE>
+template<rtx_code CODE, enum frm_op_type FRM_OP = NO_FRM>
class binop : public function_base
{
public:
- rtx expand (function_expander &e) const override
+ bool has_rounding_mode_operand_p () const override
{
- switch (e.op_info->op)
- {
- case OP_TYPE_vx:
- case OP_TYPE_vf:
- return e.use_exact_insn (code_for_pred_scalar (CODE, e.vector_mode ()));
- case OP_TYPE_vv:
- return e.use_exact_insn (code_for_pred (CODE, e.vector_mode ()));
- default:
- gcc_unreachable ();
- }
+ return FRM_OP == HAS_FRM;
}
-};
-
-/* Implements below instructions for now.
- - vfadd
- - vfsub
- - vfmul
- - vfdiv
-*/
-template<rtx_code CODE>
-class binop_frm : public function_base
-{
-public:
- bool has_rounding_mode_operand_p () const override { return true; }
rtx expand (function_expander &e) const override
{
switch (e.op_info->op)
{
+ case OP_TYPE_vx:
+ gcc_assert (FRM_OP == NO_FRM);
+ gcc_fallthrough ();
case OP_TYPE_vf:
return e.use_exact_insn (code_for_pred_scalar (CODE, e.vector_mode ()));
case OP_TYPE_vv:
@@ -1648,10 +1634,15 @@ public:
};
/* Implements vfcvt.x. */
-template<int UNSPEC>
+template<int UNSPEC, enum frm_op_type FRM_OP = NO_FRM>
class vfcvt_x : public function_base
{
public:
+ bool has_rounding_mode_operand_p () const override
+ {
+ return FRM_OP == HAS_FRM;
+ }
+
rtx expand (function_expander &e) const override
{
return e.use_exact_insn (code_for_pred_fcvt_x_f (UNSPEC, e.arg_mode (0)));
@@ -2389,8 +2380,8 @@ static CONSTEXPR const viota viota_obj;
static CONSTEXPR const vid vid_obj;
static CONSTEXPR const binop<PLUS> vfadd_obj;
static CONSTEXPR const binop<MINUS> vfsub_obj;
-static CONSTEXPR const binop_frm<PLUS> vfadd_frm_obj;
-static CONSTEXPR const binop_frm<MINUS> vfsub_frm_obj;
+static CONSTEXPR const binop<PLUS, HAS_FRM> vfadd_frm_obj;
+static CONSTEXPR const binop<MINUS, HAS_FRM> vfsub_frm_obj;
static CONSTEXPR const reverse_binop<MINUS> vfrsub_obj;
static CONSTEXPR const reverse_binop_frm<MINUS> vfrsub_frm_obj;
static CONSTEXPR const widen_binop<PLUS> vfwadd_obj;
@@ -2398,9 +2389,9 @@ static CONSTEXPR const widen_binop_frm<PLUS>
vfwadd_frm_obj;
static CONSTEXPR const widen_binop<MINUS> vfwsub_obj;
static CONSTEXPR const widen_binop_frm<MINUS> vfwsub_frm_obj;
static CONSTEXPR const binop<MULT> vfmul_obj;
-static CONSTEXPR const binop_frm<MULT> vfmul_frm_obj;
+static CONSTEXPR const binop<MULT, HAS_FRM> vfmul_frm_obj;
static CONSTEXPR const binop<DIV> vfdiv_obj;
-static CONSTEXPR const binop_frm<DIV> vfdiv_frm_obj;
+static CONSTEXPR const binop<DIV, HAS_FRM> vfdiv_frm_obj;
static CONSTEXPR const reverse_binop<DIV> vfrdiv_obj;
static CONSTEXPR const reverse_binop_frm<DIV> vfrdiv_frm_obj;
static CONSTEXPR const widen_binop<MULT> vfwmul_obj;
I would prefer to introduce an enum template argument and refactor existing code later :) On Wed, Aug 16, 2023 at 11:40 AM Li, Pan2 via Gcc-patches <gcc-patches@gcc.gnu.org> wrote: > > That should work as well, but may require some changes to existing codes like declaration, etc. > I am OK for both the enum or inherit, and will start with the CVT parts, then refactor the existing frm class. > > Do you have any suggestion for the decision making? > > Pan > > -----Original Message----- > From: Kito Cheng <kito.cheng@gmail.com> > Sent: Wednesday, August 16, 2023 11:30 AM > To: Li, Pan2 <pan2.li@intel.com> > Cc: juzhe.zhong@rivai.ai; gcc-patches <gcc-patches@gcc.gnu.org>; Wang, Yanzhang <yanzhang.wang@intel.com> > Subject: Re: [PATCH v1] RISC-V: Support RVV VFCVT.X.F.V rounding mode intrinsic API > > Or using an enum value rather than bool? > > I am thinking we could also simplify/remove most other frm classes, > some practical example: > > > diff --git a/gcc/config/riscv/riscv-vector-builtins-bases.cc > b/gcc/config/riscv/riscv-vector-builtins-bases.cc > index 2074dac0f16..ace63e963a5 100644 > --- a/gcc/config/riscv/riscv-vector-builtins-bases.cc > +++ b/gcc/config/riscv/riscv-vector-builtins-bases.cc > @@ -58,6 +58,11 @@ enum lst_type > LST_INDEXED, > }; > > +enum frm_op_type > +{ > + NO_FRM, > + HAS_FRM > +}; > /* Helper function to fold vleff and vlsegff. */ > static gimple * > fold_fault_load (gimple_folder &f) > @@ -256,41 +261,22 @@ public: > vremu/vsadd/vsaddu/vssub/vssubu > vfadd/vfsub/ > */ > -template<rtx_code CODE> > +template<rtx_code CODE, enum frm_op_type FRM_OP = NO_FRM> > class binop : public function_base > { > public: > - rtx expand (function_expander &e) const override > + bool has_rounding_mode_operand_p () const override > { > - switch (e.op_info->op) > - { > - case OP_TYPE_vx: > - case OP_TYPE_vf: > - return e.use_exact_insn (code_for_pred_scalar (CODE, e.vector_mode ())); > - case OP_TYPE_vv: > - return e.use_exact_insn (code_for_pred (CODE, e.vector_mode ())); > - default: > - gcc_unreachable (); > - } > + return FRM_OP == HAS_FRM; > } > -}; > - > -/* Implements below instructions for now. > - - vfadd > - - vfsub > - - vfmul > - - vfdiv > -*/ > -template<rtx_code CODE> > -class binop_frm : public function_base > -{ > -public: > - bool has_rounding_mode_operand_p () const override { return true; } > > rtx expand (function_expander &e) const override > { > switch (e.op_info->op) > { > + case OP_TYPE_vx: > + gcc_assert (FRM_OP == NO_FRM); > + gcc_fallthrough (); > case OP_TYPE_vf: > return e.use_exact_insn (code_for_pred_scalar (CODE, e.vector_mode ())); > case OP_TYPE_vv: > @@ -1648,10 +1634,15 @@ public: > }; > > /* Implements vfcvt.x. */ > -template<int UNSPEC> > +template<int UNSPEC, enum frm_op_type FRM_OP = NO_FRM> > class vfcvt_x : public function_base > { > public: > + bool has_rounding_mode_operand_p () const override > + { > + return FRM_OP == HAS_FRM; > + } > + > rtx expand (function_expander &e) const override > { > return e.use_exact_insn (code_for_pred_fcvt_x_f (UNSPEC, e.arg_mode (0))); > @@ -2389,8 +2380,8 @@ static CONSTEXPR const viota viota_obj; > static CONSTEXPR const vid vid_obj; > static CONSTEXPR const binop<PLUS> vfadd_obj; > static CONSTEXPR const binop<MINUS> vfsub_obj; > -static CONSTEXPR const binop_frm<PLUS> vfadd_frm_obj; > -static CONSTEXPR const binop_frm<MINUS> vfsub_frm_obj; > +static CONSTEXPR const binop<PLUS, HAS_FRM> vfadd_frm_obj; > +static CONSTEXPR const binop<MINUS, HAS_FRM> vfsub_frm_obj; > static CONSTEXPR const reverse_binop<MINUS> vfrsub_obj; > static CONSTEXPR const reverse_binop_frm<MINUS> vfrsub_frm_obj; > static CONSTEXPR const widen_binop<PLUS> vfwadd_obj; > @@ -2398,9 +2389,9 @@ static CONSTEXPR const widen_binop_frm<PLUS> > vfwadd_frm_obj; > static CONSTEXPR const widen_binop<MINUS> vfwsub_obj; > static CONSTEXPR const widen_binop_frm<MINUS> vfwsub_frm_obj; > static CONSTEXPR const binop<MULT> vfmul_obj; > -static CONSTEXPR const binop_frm<MULT> vfmul_frm_obj; > +static CONSTEXPR const binop<MULT, HAS_FRM> vfmul_frm_obj; > static CONSTEXPR const binop<DIV> vfdiv_obj; > -static CONSTEXPR const binop_frm<DIV> vfdiv_frm_obj; > +static CONSTEXPR const binop<DIV, HAS_FRM> vfdiv_frm_obj; > static CONSTEXPR const reverse_binop<DIV> vfrdiv_obj; > static CONSTEXPR const reverse_binop_frm<DIV> vfrdiv_frm_obj; > static CONSTEXPR const widen_binop<MULT> vfwmul_obj;
Got it, thanks! Will start with CVT and rest frm instructions first, and then refactor. Pan -----Original Message----- From: Kito Cheng <kito.cheng@gmail.com> Sent: Wednesday, August 16, 2023 11:44 AM To: Li, Pan2 <pan2.li@intel.com> Cc: juzhe.zhong@rivai.ai; gcc-patches <gcc-patches@gcc.gnu.org>; Wang, Yanzhang <yanzhang.wang@intel.com> Subject: Re: [PATCH v1] RISC-V: Support RVV VFCVT.X.F.V rounding mode intrinsic API I would prefer to introduce an enum template argument and refactor existing code later :) On Wed, Aug 16, 2023 at 11:40 AM Li, Pan2 via Gcc-patches <gcc-patches@gcc.gnu.org> wrote: > > That should work as well, but may require some changes to existing codes like declaration, etc. > I am OK for both the enum or inherit, and will start with the CVT parts, then refactor the existing frm class. > > Do you have any suggestion for the decision making? > > Pan > > -----Original Message----- > From: Kito Cheng <kito.cheng@gmail.com> > Sent: Wednesday, August 16, 2023 11:30 AM > To: Li, Pan2 <pan2.li@intel.com> > Cc: juzhe.zhong@rivai.ai; gcc-patches <gcc-patches@gcc.gnu.org>; Wang, Yanzhang <yanzhang.wang@intel.com> > Subject: Re: [PATCH v1] RISC-V: Support RVV VFCVT.X.F.V rounding mode intrinsic API > > Or using an enum value rather than bool? > > I am thinking we could also simplify/remove most other frm classes, > some practical example: > > > diff --git a/gcc/config/riscv/riscv-vector-builtins-bases.cc > b/gcc/config/riscv/riscv-vector-builtins-bases.cc > index 2074dac0f16..ace63e963a5 100644 > --- a/gcc/config/riscv/riscv-vector-builtins-bases.cc > +++ b/gcc/config/riscv/riscv-vector-builtins-bases.cc > @@ -58,6 +58,11 @@ enum lst_type > LST_INDEXED, > }; > > +enum frm_op_type > +{ > + NO_FRM, > + HAS_FRM > +}; > /* Helper function to fold vleff and vlsegff. */ > static gimple * > fold_fault_load (gimple_folder &f) > @@ -256,41 +261,22 @@ public: > vremu/vsadd/vsaddu/vssub/vssubu > vfadd/vfsub/ > */ > -template<rtx_code CODE> > +template<rtx_code CODE, enum frm_op_type FRM_OP = NO_FRM> > class binop : public function_base > { > public: > - rtx expand (function_expander &e) const override > + bool has_rounding_mode_operand_p () const override > { > - switch (e.op_info->op) > - { > - case OP_TYPE_vx: > - case OP_TYPE_vf: > - return e.use_exact_insn (code_for_pred_scalar (CODE, e.vector_mode ())); > - case OP_TYPE_vv: > - return e.use_exact_insn (code_for_pred (CODE, e.vector_mode ())); > - default: > - gcc_unreachable (); > - } > + return FRM_OP == HAS_FRM; > } > -}; > - > -/* Implements below instructions for now. > - - vfadd > - - vfsub > - - vfmul > - - vfdiv > -*/ > -template<rtx_code CODE> > -class binop_frm : public function_base > -{ > -public: > - bool has_rounding_mode_operand_p () const override { return true; } > > rtx expand (function_expander &e) const override > { > switch (e.op_info->op) > { > + case OP_TYPE_vx: > + gcc_assert (FRM_OP == NO_FRM); > + gcc_fallthrough (); > case OP_TYPE_vf: > return e.use_exact_insn (code_for_pred_scalar (CODE, e.vector_mode ())); > case OP_TYPE_vv: > @@ -1648,10 +1634,15 @@ public: > }; > > /* Implements vfcvt.x. */ > -template<int UNSPEC> > +template<int UNSPEC, enum frm_op_type FRM_OP = NO_FRM> > class vfcvt_x : public function_base > { > public: > + bool has_rounding_mode_operand_p () const override > + { > + return FRM_OP == HAS_FRM; > + } > + > rtx expand (function_expander &e) const override > { > return e.use_exact_insn (code_for_pred_fcvt_x_f (UNSPEC, e.arg_mode (0))); > @@ -2389,8 +2380,8 @@ static CONSTEXPR const viota viota_obj; > static CONSTEXPR const vid vid_obj; > static CONSTEXPR const binop<PLUS> vfadd_obj; > static CONSTEXPR const binop<MINUS> vfsub_obj; > -static CONSTEXPR const binop_frm<PLUS> vfadd_frm_obj; > -static CONSTEXPR const binop_frm<MINUS> vfsub_frm_obj; > +static CONSTEXPR const binop<PLUS, HAS_FRM> vfadd_frm_obj; > +static CONSTEXPR const binop<MINUS, HAS_FRM> vfsub_frm_obj; > static CONSTEXPR const reverse_binop<MINUS> vfrsub_obj; > static CONSTEXPR const reverse_binop_frm<MINUS> vfrsub_frm_obj; > static CONSTEXPR const widen_binop<PLUS> vfwadd_obj; > @@ -2398,9 +2389,9 @@ static CONSTEXPR const widen_binop_frm<PLUS> > vfwadd_frm_obj; > static CONSTEXPR const widen_binop<MINUS> vfwsub_obj; > static CONSTEXPR const widen_binop_frm<MINUS> vfwsub_frm_obj; > static CONSTEXPR const binop<MULT> vfmul_obj; > -static CONSTEXPR const binop_frm<MULT> vfmul_frm_obj; > +static CONSTEXPR const binop<MULT, HAS_FRM> vfmul_frm_obj; > static CONSTEXPR const binop<DIV> vfdiv_obj; > -static CONSTEXPR const binop_frm<DIV> vfdiv_frm_obj; > +static CONSTEXPR const binop<DIV, HAS_FRM> vfdiv_frm_obj; > static CONSTEXPR const reverse_binop<DIV> vfrdiv_obj; > static CONSTEXPR const reverse_binop_frm<DIV> vfrdiv_frm_obj; > static CONSTEXPR const widen_binop<MULT> vfwmul_obj;
diff --git a/gcc/config/riscv/riscv-vector-builtins-bases.cc b/gcc/config/riscv/riscv-vector-builtins-bases.cc index f2124080ef9..754a53efd3d 100644 --- a/gcc/config/riscv/riscv-vector-builtins-bases.cc +++ b/gcc/config/riscv/riscv-vector-builtins-bases.cc @@ -660,6 +660,21 @@ public: } }; +/* Implements below instructions for frm + - vfcvt_x +*/ +template<int UNSPEC> +class vfcvt_x_frm : public function_base +{ +public: + bool has_rounding_mode_operand_p () const override { return true; } + + rtx expand (function_expander &e) const override + { + return e.use_exact_insn (code_for_pred_fcvt_x_f (UNSPEC, e.arg_mode (0))); + } +}; + /* Implements vrsub. */ class vrsub : public function_base { @@ -2465,6 +2480,7 @@ static CONSTEXPR const vfclass vfclass_obj; static CONSTEXPR const vmerge vfmerge_obj; static CONSTEXPR const vmv_v vfmv_v_obj; static CONSTEXPR const vfcvt_x<UNSPEC_VFCVT> vfcvt_x_obj; +static CONSTEXPR const vfcvt_x_frm<UNSPEC_VFCVT> vfcvt_x_frm_obj; static CONSTEXPR const vfcvt_x<UNSPEC_UNSIGNED_VFCVT> vfcvt_xu_obj; static CONSTEXPR const vfcvt_rtz_x<FIX> vfcvt_rtz_x_obj; static CONSTEXPR const vfcvt_rtz_x<UNSIGNED_FIX> vfcvt_rtz_xu_obj; @@ -2714,6 +2730,7 @@ BASE (vfclass) BASE (vfmerge) BASE (vfmv_v) BASE (vfcvt_x) +BASE (vfcvt_x_frm) BASE (vfcvt_xu) BASE (vfcvt_rtz_x) BASE (vfcvt_rtz_xu) diff --git a/gcc/config/riscv/riscv-vector-builtins-bases.h b/gcc/config/riscv/riscv-vector-builtins-bases.h index 2a9381eec5e..50a7d7ffb6f 100644 --- a/gcc/config/riscv/riscv-vector-builtins-bases.h +++ b/gcc/config/riscv/riscv-vector-builtins-bases.h @@ -205,6 +205,7 @@ extern const function_base *const vfclass; extern const function_base *const vfmerge; extern const function_base *const vfmv_v; extern const function_base *const vfcvt_x; +extern const function_base *const vfcvt_x_frm; extern const function_base *const vfcvt_xu; extern const function_base *const vfcvt_rtz_x; extern const function_base *const vfcvt_rtz_xu; diff --git a/gcc/config/riscv/riscv-vector-builtins-functions.def b/gcc/config/riscv/riscv-vector-builtins-functions.def index 34def6bb82f..8b6a7cc49f3 100644 --- a/gcc/config/riscv/riscv-vector-builtins-functions.def +++ b/gcc/config/riscv/riscv-vector-builtins-functions.def @@ -445,6 +445,8 @@ DEF_RVV_FUNCTION (vfcvt_rtz_xu, alu, full_preds, f_to_u_f_v_ops) DEF_RVV_FUNCTION (vfcvt_f, alu, full_preds, i_to_f_x_v_ops) DEF_RVV_FUNCTION (vfcvt_f, alu, full_preds, u_to_f_xu_v_ops) +DEF_RVV_FUNCTION (vfcvt_x_frm, alu_frm, full_preds, f_to_i_f_v_ops) + // 13.18. Widening Floating-Point/Integer Type-Convert Instructions DEF_RVV_FUNCTION (vfwcvt_x, alu, full_preds, f_to_wi_f_v_ops) DEF_RVV_FUNCTION (vfwcvt_xu, alu, full_preds, f_to_wu_f_v_ops) diff --git a/gcc/testsuite/gcc.target/riscv/rvv/base/float-point-cvt-x.c b/gcc/testsuite/gcc.target/riscv/rvv/base/float-point-cvt-x.c new file mode 100644 index 00000000000..e090f0f97e9 --- /dev/null +++ b/gcc/testsuite/gcc.target/riscv/rvv/base/float-point-cvt-x.c @@ -0,0 +1,29 @@ +/* { dg-do compile } */ +/* { dg-options "-march=rv64gcv -mabi=lp64 -O3 -Wno-psabi" } */ + +#include "riscv_vector.h" + +vint32m1_t +test_riscv_vfcvt_x_f_vv_i32m1_rm (vfloat32m1_t op1, size_t vl) { + return __riscv_vfcvt_x_f_v_i32m1_rm (op1, 0, vl); +} + +vint32m1_t +test_vfcvt_x_f_vv_i32m1_rm_m (vbool32_t mask, vfloat32m1_t op1, size_t vl) { + return __riscv_vfcvt_x_f_v_i32m1_rm_m (mask, op1, 1, vl); +} + +vint32m1_t +test_riscv_vfcvt_x_f_vv_i32m1 (vfloat32m1_t op1, size_t vl) { + return __riscv_vfcvt_x_f_v_i32m1 (op1, vl); +} + +vint32m1_t +test_vfcvt_x_f_vv_i32m1_m (vbool32_t mask, vfloat32m1_t op1, size_t vl) { + return __riscv_vfcvt_x_f_v_i32m1_m (mask, op1, vl); +} + +/* { dg-final { scan-assembler-times {vfcvt\.x\.f\.v\s+v[0-9]+,\s*v[0-9]+} 4 } } */ +/* { dg-final { scan-assembler-times {frrm\s+[axs][0-9]+} 2 } } */ +/* { dg-final { scan-assembler-times {fsrm\s+[axs][0-9]+} 2 } } */ +/* { dg-final { scan-assembler-times {fsrmi\s+[01234]} 2 } } */