Re: Re: [PATCH v1] RISC-V: Allow rounding mode control for RVV floating-point add

["juzhe.zhong@rivai.ai" <juzhe.zhong@rivai.ai>]

[-- Attachment #1: Type: text/plain, Size: 18844 bytes --]

Hi, Pan.

I think the last step is to support dynamic mode switching which may need to change the mode-switching PASS.

After this done, I suggest you go over all rounding mode API (including fixed-point and floating-point.)

Check SPIKE implementation, make sure which API needs rounding mode, which API doesn't need rounding mode.
Do not trust the rvv-intrinsic-doc since it's often wrong. 
You should check doc too, if doc is wrong,  you should not only correct GCC implementation but also make a fix PR to the doc.

Thanks.


juzhe.zhong@rivai.ai
 
From: Kito Cheng
Date: 2023-06-29 10:35
To: Li, Pan2
CC: juzhe.zhong@rivai.ai; gcc-patches; Wang, Yanzhang; jeffreyalaw
Subject: Re: [PATCH v1] RISC-V: Allow rounding mode control for RVV floating-point add
LGTM, thanks!
 
On Tue, Jun 27, 2023 at 3:02 PM Li, Pan2 <pan2.li@intel.com> wrote:
>
> Ack, thanks Juzhe.
>
>
>
> Pan
>
>
>
> From: juzhe.zhong@rivai.ai <juzhe.zhong@rivai.ai>
> Sent: Tuesday, June 27, 2023 3:00 PM
> To: Li, Pan2 <pan2.li@intel.com>; gcc-patches <gcc-patches@gcc.gnu.org>
> Cc: Kito.cheng <kito.cheng@sifive.com>; Li, Pan2 <pan2.li@intel.com>; Wang, Yanzhang <yanzhang.wang@intel.com>; jeffreyalaw <jeffreyalaw@gmail.com>
> Subject: Re: [PATCH v1] RISC-V: Allow rounding mode control for RVV floating-point add
>
>
>
> LGTM.
>
> You can go ahead to implement rounding mode of floating-point by mode-switching:
>
>
>
> Suggest you implement rounding mode for floating-poing as follows:
>
>
>
> 1st step: Implement mode-switching for floating-point rounding mode except DYNAMIC which should be totally same as fixed-point.
>
> 2nd step: Support DYNAMIC rounding mode on mode-switching which may need to modify the mode-switching PASS.
>
>
>
> Thanks.
>
> ________________________________
>
> juzhe.zhong@rivai.ai
>
>
>
> From: pan2.li
>
> Date: 2023-06-27 14:06
>
> To: gcc-patches
>
> CC: juzhe.zhong; kito.cheng; pan2.li; yanzhang.wang; jeffreyalaw
>
> Subject: [PATCH v1] RISC-V: Allow rounding mode control for RVV floating-point add
>
> From: Pan Li <pan2.li@intel.com>
>
>
>
> According to the doc as below, we need to support the rounding mode of
>
> the RVV floating-point, both the static and dynamice frm.
>
>
>
> https://github.com/riscv-non-isa/rvv-intrinsic-doc/pull/226
>
>
>
> For tracking and development friendly, We will take some steps to support
>
> all rounding modes for the RVV floating-point rounding modes.
>
>
>
> 1. Allow rounding mode control by one intrinsic (aka this patch), vfadd.
>
> 2. Support static rounding mode control by mode switch, like fixed-point.
>
> 3. Support dynamice round mode control by mode switch.
>
> 4. Support the rest floating-point instructions for frm.
>
>
>
> Please *NOTE* this patch only allow the rounding mode control for the
>
> vfadd intrinsic API, and the related frm will be coverred by step 2.
>
>
>
> Signed-off-by: Pan Li <pan2.li@intel.com>
>
> Co-Authored by: Juzhe-Zhong <juzhe.zhong@rivai.ai>
>
>
>
> gcc/ChangeLog:
>
>
>
> * config/riscv/riscv-protos.h (enum floating_point_rounding_mode):
>
> Add macro for static frm min and max.
>
> * config/riscv/riscv-vector-builtins-bases.cc
>
> (class binop_frm): New class for floating-point with frm.
>
> (BASE): Add vfadd for frm.
>
> * config/riscv/riscv-vector-builtins-bases.h: Likewise.
>
> * config/riscv/riscv-vector-builtins-functions.def
>
> (vfadd_frm): Likewise.
>
> * config/riscv/riscv-vector-builtins-shapes.cc
>
> (struct alu_frm_def): New struct for alu with frm.
>
> (SHAPE): Add alu with frm.
>
> * config/riscv/riscv-vector-builtins-shapes.h: Likewise.
>
> * config/riscv/riscv-vector-builtins.cc
>
> (function_checker::report_out_of_range_and_not): New function
>
> for report out of range and not val.
>
> (function_checker::require_immediate_range_or): New function
>
> for checking in range or one val.
>
> * config/riscv/riscv-vector-builtins.h: Add function decl.
>
>
>
> gcc/testsuite/ChangeLog:
>
>
>
> * gcc.target/riscv/rvv/base/float-point-frm-error.c: New test.
>
> * gcc.target/riscv/rvv/base/float-point-frm.c: New test.
>
> ---
>
> gcc/config/riscv/riscv-protos.h               |  2 +
>
> .../riscv/riscv-vector-builtins-bases.cc      | 25 +++++++
>
> .../riscv/riscv-vector-builtins-bases.h       |  1 +
>
> .../riscv/riscv-vector-builtins-functions.def |  2 +
>
> .../riscv/riscv-vector-builtins-shapes.cc     | 68 +++++++++++++++++++
>
> .../riscv/riscv-vector-builtins-shapes.h      |  1 +
>
> gcc/config/riscv/riscv-vector-builtins.cc     | 41 +++++++++++
>
> gcc/config/riscv/riscv-vector-builtins.h      |  4 ++
>
> .../riscv/rvv/base/float-point-frm-error.c    | 15 ++++
>
> .../riscv/rvv/base/float-point-frm.c          | 30 ++++++++
>
> 10 files changed, 189 insertions(+)
>
> create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/base/float-point-frm-error.c
>
> create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/base/float-point-frm.c
>
>
>
> diff --git a/gcc/config/riscv/riscv-protos.h b/gcc/config/riscv/riscv-protos.h
>
> index f686edab3d1..bee64eee504 100644
>
> --- a/gcc/config/riscv/riscv-protos.h
>
> +++ b/gcc/config/riscv/riscv-protos.h
>
> @@ -278,6 +278,8 @@ enum floating_point_rounding_mode
>
>    FRM_RUP = 3, /* Aka 0b011.  */
>
>    FRM_RMM = 4, /* Aka 0b100.  */
>
>    FRM_DYN = 7, /* Aka 0b111.  */
>
> +  FRM_STATIC_MIN = FRM_RNE,
>
> +  FRM_STATIC_MAX = FRM_RMM,
>
> };
>
> opt_machine_mode vectorize_related_mode (machine_mode, scalar_mode,
>
> diff --git a/gcc/config/riscv/riscv-vector-builtins-bases.cc b/gcc/config/riscv/riscv-vector-builtins-bases.cc
>
> index 5c8deda900d..1b4c2c6ad66 100644
>
> --- a/gcc/config/riscv/riscv-vector-builtins-bases.cc
>
> +++ b/gcc/config/riscv/riscv-vector-builtins-bases.cc
>
> @@ -281,6 +281,29 @@ public:
>
>    }
>
> };
>
> +/* Implements below instructions for now.
>
> +   - vfadd
>
> +*/
>
> +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_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 ();
>
> +      }
>
> +  }
>
> +};
>
> +
>
> /* Implements vrsub.  */
>
> class vrsub : public function_base
>
> {
>
> @@ -2006,6 +2029,7 @@ 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 reverse_binop<MINUS> vfrsub_obj;
>
> static CONSTEXPR const widen_binop<PLUS> vfwadd_obj;
>
> static CONSTEXPR const widen_binop<MINUS> vfwsub_obj;
>
> @@ -2231,6 +2255,7 @@ BASE (vmsof)
>
> BASE (viota)
>
> BASE (vid)
>
> BASE (vfadd)
>
> +BASE (vfadd_frm)
>
> BASE (vfsub)
>
> BASE (vfrsub)
>
> BASE (vfwadd)
>
> diff --git a/gcc/config/riscv/riscv-vector-builtins-bases.h b/gcc/config/riscv/riscv-vector-builtins-bases.h
>
> index 62ff38a2811..54a81eab269 100644
>
> --- a/gcc/config/riscv/riscv-vector-builtins-bases.h
>
> +++ b/gcc/config/riscv/riscv-vector-builtins-bases.h
>
> @@ -145,6 +145,7 @@ extern const function_base *const viota;
>
> extern const function_base *const vid;
>
> extern const function_base *const vfadd;
>
> extern const function_base *const vfadd;
>
> +extern const function_base *const vfadd_frm;
>
> extern const function_base *const vfsub;
>
> extern const function_base *const vfsub;
>
> extern const function_base *const vfrsub;
>
> diff --git a/gcc/config/riscv/riscv-vector-builtins-functions.def b/gcc/config/riscv/riscv-vector-builtins-functions.def
>
> index 89aff27bf26..035c9e4252f 100644
>
> --- a/gcc/config/riscv/riscv-vector-builtins-functions.def
>
> +++ b/gcc/config/riscv/riscv-vector-builtins-functions.def
>
> @@ -289,6 +289,8 @@ DEF_RVV_FUNCTION (vfadd, alu, full_preds, f_vvf_ops)
>
> DEF_RVV_FUNCTION (vfsub, alu, full_preds, f_vvv_ops)
>
> DEF_RVV_FUNCTION (vfsub, alu, full_preds, f_vvf_ops)
>
> DEF_RVV_FUNCTION (vfrsub, alu, full_preds, f_vvf_ops)
>
> +DEF_RVV_FUNCTION (vfadd_frm, alu_frm, full_preds, f_vvv_ops)
>
> +DEF_RVV_FUNCTION (vfadd_frm, alu_frm, full_preds, f_vvf_ops)
>
> // 13.3. Vector Widening Floating-Point Add/Subtract Instructions
>
> DEF_RVV_FUNCTION (vfwadd, widen_alu, full_preds, f_wvv_ops)
>
> diff --git a/gcc/config/riscv/riscv-vector-builtins-shapes.cc b/gcc/config/riscv/riscv-vector-builtins-shapes.cc
>
> index c8daae01f91..69a67106418 100644
>
> --- a/gcc/config/riscv/riscv-vector-builtins-shapes.cc
>
> +++ b/gcc/config/riscv/riscv-vector-builtins-shapes.cc
>
> @@ -226,6 +226,73 @@ struct alu_def : public build_base
>
>    }
>
> };
>
> +/* alu_frm_def class.  */
>
> +struct alu_frm_def : public build_base
>
> +{
>
> +  /* Normalize vf<op>_frm to vf<op>.  */
>
> +  static void normalize_base_name (char *to, const char *from, int limit)
>
> +  {
>
> +    strncpy (to, from, limit - 1);
>
> +    char *suffix = strstr (to, "_frm");
>
> +
>
> +    if (suffix)
>
> +      *suffix = '\0';
>
> +
>
> +    to[limit - 1] = '\0';
>
> +  }
>
> +
>
> +  char *get_name (function_builder &b, const function_instance &instance,
>
> +   bool overloaded_p) const override
>
> +  {
>
> +    char base_name[16] = {};
>
> +
>
> +    /* Return nullptr if it can not be overloaded.  */
>
> +    if (overloaded_p && !instance.base->can_be_overloaded_p (instance.pred))
>
> +      return nullptr;
>
> +
>
> +    normalize_base_name (base_name, instance.base_name, sizeof (base_name));
>
> +
>
> +    b.append_base_name (base_name);
>
> +
>
> +    /* vop<sew>_<op> --> vop<sew>_<op>_<type>.  */
>
> +    if (!overloaded_p)
>
> +      {
>
> + b.append_name (operand_suffixes[instance.op_info->op]);
>
> + b.append_name (type_suffixes[instance.type.index].vector);
>
> +      }
>
> +
>
> +    /* According to rvv-intrinsic-doc, it does not add "_m" suffix
>
> +       for vop_m C++ overloaded API.  */
>
> +    if (overloaded_p && instance.pred == PRED_TYPE_m)
>
> +      return b.finish_name ();
>
> +
>
> +    b.append_name (predication_suffixes[instance.pred]);
>
> +
>
> +    /* According to rvv-intrinsic-doc, it does not add "_rm" suffix
>
> +       for vop_rm C++ overloaded API.  */
>
> +    if (!overloaded_p)
>
> +      b.append_name ("_rm");
>
> +
>
> +    return b.finish_name ();
>
> +  }
>
> +
>
> +  bool check (function_checker &c) const override
>
> +  {
>
> +    gcc_assert (c.any_type_float_p ());
>
> +
>
> +    /* Check whether rounding mode argument is a valid immediate.  */
>
> +    if (c.base->has_rounding_mode_operand_p ())
>
> +      {
>
> + unsigned int frm_num = c.arg_num () - 2;
>
> +
>
> + return c.require_immediate_range_or (frm_num, FRM_STATIC_MIN,
>
> +      FRM_STATIC_MAX, FRM_DYN);
>
> +      }
>
> +
>
> +    return true;
>
> +  }
>
> +};
>
> +
>
> /* widen_alu_def class. Handle vwadd/vwsub. Unlike
>
>     vadd.vx/vadd.vv/vwmul.vv/vwmul.vx, vwadd.vv/vwadd.vx/vwadd.wv/vwadd.wx has
>
>     'OP' suffix in overloaded API.  */
>
> @@ -743,6 +810,7 @@ SHAPE(vsetvl, vsetvlmax)
>
> SHAPE(loadstore, loadstore)
>
> SHAPE(indexed_loadstore, indexed_loadstore)
>
> SHAPE(alu, alu)
>
> +SHAPE(alu_frm, alu_frm)
>
> SHAPE(widen_alu, widen_alu)
>
> SHAPE(no_mask_policy, no_mask_policy)
>
> SHAPE(return_mask, return_mask)
>
> diff --git a/gcc/config/riscv/riscv-vector-builtins-shapes.h b/gcc/config/riscv/riscv-vector-builtins-shapes.h
>
> index 6a51713c12c..15fef8342ec 100644
>
> --- a/gcc/config/riscv/riscv-vector-builtins-shapes.h
>
> +++ b/gcc/config/riscv/riscv-vector-builtins-shapes.h
>
> @@ -29,6 +29,7 @@ extern const function_shape *const vsetvlmax;
>
> extern const function_shape *const loadstore;
>
> extern const function_shape *const indexed_loadstore;
>
> extern const function_shape *const alu;
>
> +extern const function_shape *const alu_frm;
>
> extern const function_shape *const widen_alu;
>
> extern const function_shape *const no_mask_policy;
>
> extern const function_shape *const return_mask;
>
> diff --git a/gcc/config/riscv/riscv-vector-builtins.cc b/gcc/config/riscv/riscv-vector-builtins.cc
>
> index 466e36d50b7..648c765a5d1 100644
>
> --- a/gcc/config/riscv/riscv-vector-builtins.cc
>
> +++ b/gcc/config/riscv/riscv-vector-builtins.cc
>
> @@ -3852,6 +3852,23 @@ function_checker::report_out_of_range (unsigned int argno, HOST_WIDE_INT actual,
>
>     actual, argno + 1, fndecl, min, max);
>
> }
>
> +/* Report that LOCATION has a call to FNDECL in which argument ARGNO has
>
> +   the value ACTUAL, whereas the function requires a value in the range
>
> +   [MIN, MAX] or OR_VAL.  ARGNO counts from zero.  */
>
> +void
>
> +function_checker::report_out_of_range_and_not (unsigned int argno,
>
> +        HOST_WIDE_INT actual,
>
> +        HOST_WIDE_INT min,
>
> +        HOST_WIDE_INT max,
>
> +        HOST_WIDE_INT or_val) const
>
> +{
>
> +  error_at (location,
>
> +     "passing %wd to argument %d of %qE, which expects"
>
> +     " a value in the range [%wd, %wd] or %wd",
>
> +     actual, argno + 1, fndecl, min, max, or_val);
>
> +}
>
> +
>
> +
>
> /* Check that argument ARGNO is an integer constant expression and
>
>     store its value in VALUE_OUT if so.  The caller should first
>
>     check that argument ARGNO exists.  */
>
> @@ -3893,6 +3910,30 @@ function_checker::require_immediate_range (unsigned int argno,
>
>    return true;
>
> }
>
> +/* Check that argument REL_ARGNO is an integer constant expression in the
>
> +   range [MIN, MAX] or OR_VAL.  REL_ARGNO counts from the end of the
>
> +   predication arguments.  */
>
> +bool
>
> +function_checker::require_immediate_range_or (unsigned int argno,
>
> +       HOST_WIDE_INT min,
>
> +       HOST_WIDE_INT max,
>
> +       HOST_WIDE_INT or_val) const
>
> +{
>
> +  gcc_assert (min >= 0 && min <= max);
>
> +  gcc_assert (argno < m_nargs);
>
> +
>
> +  tree arg = m_args[argno];
>
> +  HOST_WIDE_INT actual = tree_to_uhwi (arg);
>
> +
>
> +  if (!IN_RANGE (actual, min, max) && actual != or_val)
>
> +    {
>
> +      report_out_of_range_and_not (argno, actual, min, max, or_val);
>
> +      return false;
>
> +    }
>
> +
>
> +  return true;
>
> +}
>
> +
>
> /* Perform semantic checks on the call.  Return true if the call is valid,
>
>     otherwise report a suitable error.  */
>
> bool
>
> diff --git a/gcc/config/riscv/riscv-vector-builtins.h b/gcc/config/riscv/riscv-vector-builtins.h
>
> index b0c3a42d820..e358a8e4d91 100644
>
> --- a/gcc/config/riscv/riscv-vector-builtins.h
>
> +++ b/gcc/config/riscv/riscv-vector-builtins.h
>
> @@ -442,6 +442,8 @@ public:
>
>    bool check (void);
>
>    bool require_immediate (unsigned int, HOST_WIDE_INT, HOST_WIDE_INT) const;
>
> +  bool require_immediate_range_or (unsigned int, HOST_WIDE_INT,
>
> +    HOST_WIDE_INT, HOST_WIDE_INT) const;
>
> private:
>
>    bool require_immediate_range (unsigned int, HOST_WIDE_INT,
>
> @@ -449,6 +451,8 @@ private:
>
>    void report_non_ice (unsigned int) const;
>
>    void report_out_of_range (unsigned int, HOST_WIDE_INT, HOST_WIDE_INT,
>
>     HOST_WIDE_INT) const;
>
> +  void report_out_of_range_and_not (unsigned int, HOST_WIDE_INT, HOST_WIDE_INT,
>
> +     HOST_WIDE_INT, HOST_WIDE_INT) const;
>
>    /* The type of the resolved function.  */
>
>    tree m_fntype;
>
> diff --git a/gcc/testsuite/gcc.target/riscv/rvv/base/float-point-frm-error.c b/gcc/testsuite/gcc.target/riscv/rvv/base/float-point-frm-error.c
>
> new file mode 100644
>
> index 00000000000..4ebaa15ab0b
>
> --- /dev/null
>
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/base/float-point-frm-error.c
>
> @@ -0,0 +1,15 @@
>
> +/* { dg-do compile } */
>
> +/* { dg-options "-march=rv64gcv -mabi=lp64 -O3 -Wno-psabi" } */
>
> +
>
> +#include "riscv_vector.h"
>
> +
>
> +typedef float float32_t;
>
> +
>
> +void test_float_point_frm_error (float32_t *out, vfloat32m1_t op1, vfloat32m1_t op2, size_t vl)
>
> +{
>
> +  vfloat32m1_t v1 = __riscv_vfadd_vv_f32m1_rm (op1, op2, 5, vl); /* { dg-error {passing 5 to argument 3 of '__riscv_vfadd_vv_f32m1_rm', which expects a value in the range \[0, 4\] or 7} } */
>
> +  vfloat32m1_t v2 = __riscv_vfadd_vv_f32m1_rm (v1, v1, 6, vl);   /* { dg-error {passing 6 to argument 3 of '__riscv_vfadd_vv_f32m1_rm', which expects a value in the range \[0, 4\] or 7} } */
>
> +  vfloat32m1_t v3 = __riscv_vfadd_vv_f32m1_rm (v2, v2, 8, vl);   /* { dg-error {passing 8 to argument 3 of '__riscv_vfadd_vv_f32m1_rm', which expects a value in the range \[0, 4\] or 7} } */
>
> +
>
> +  __riscv_vse32_v_f32m1 (out, v3, vl);
>
> +}
>
> diff --git a/gcc/testsuite/gcc.target/riscv/rvv/base/float-point-frm.c b/gcc/testsuite/gcc.target/riscv/rvv/base/float-point-frm.c
>
> new file mode 100644
>
> index 00000000000..95271b2c822
>
> --- /dev/null
>
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/base/float-point-frm.c
>
> @@ -0,0 +1,30 @@
>
> +/* { dg-do compile } */
>
> +/* { dg-options "-march=rv64gcv -mabi=lp64 -O3 -Wno-psabi" } */
>
> +
>
> +#include "riscv_vector.h"
>
> +
>
> +typedef float float32_t;
>
> +
>
> +vfloat32m1_t
>
> +test_riscv_vfadd_vv_f32m1_rm (vfloat32m1_t op1, vfloat32m1_t op2, size_t vl) {
>
> +  return __riscv_vfadd_vv_f32m1_rm (op1, op2, 0, vl);
>
> +}
>
> +
>
> +vfloat32m1_t
>
> +test_vfadd_vv_f32m1_m_rm(vbool32_t mask, vfloat32m1_t op1, vfloat32m1_t op2,
>
> + size_t vl) {
>
> +  return __riscv_vfadd_vv_f32m1_m_rm(mask, op1, op2, 0, vl);
>
> +}
>
> +
>
> +vfloat32m1_t
>
> +test_vfadd_vf_f32m1_rm(vfloat32m1_t op1, float32_t op2, size_t vl) {
>
> +  return __riscv_vfadd_vf_f32m1_rm(op1, op2, 0, vl);
>
> +}
>
> +
>
> +vfloat32m1_t
>
> +test_vfadd_vf_f32m1_m_rm(vbool32_t mask, vfloat32m1_t op1, float32_t op2,
>
> + size_t vl) {
>
> +  return __riscv_vfadd_vf_f32m1_m_rm(mask, op1, op2, 0, vl);
>
> +}
>
> +
>
> +/* { dg-final { scan-assembler-times {vfadd\.v[vf]\s+v[0-9]+,\s*v[0-9]+,\s*[fav]+[0-9]+} 4 } } */
>
> --
>
> 2.34.1
>
>
>
>