[PATCH v3 0/7] target/riscv: Add XVentanaCondOps and supporting infrastructure changes

[PATCH v3 0/7] target/riscv: Add XVentanaCondOps and supporting infrastructure changes

[Philipp Tomsich]

In adding our first X-extension (i.e., vendor-defined) on RISC-V with
XVentanaCondOps, we need to add a few instructure improvements to make
it easier to add similar vendor-defined extensions in the future:
- refactor access to the cfg->ext_* fields by making a pointer to the
  cfg structure (as cfg_ptr) available via DisasContext
- add a table-based list of decoders to invoke, each being guarded by
  a guard/predicate-function, that can be used to either add vendor
  extensions, large extensions or override (by listing the decoder
  before the one for standard extensions) patterns to handle errata


Changes in v3:
- (new patch) refactor 'struct RISCVCPUConfig'
- (new patch) copy pointer to element cfg into DisasContext
- (new patch) test extension-availability through cfg_ptr in
  DisasContext, removing the fields that have been copied into
  DisasContext directly
- (new patch) change Zb[abcs] implementation to use cfg_ptr (copied
  into DisasContext) instead of going throuhg RISCV_CPU
- expose only the DisasContext* to predicate functions
- mark the table of decoder functions as static
- drop the inline from always_true_p, until the need arises (i.e.,
  someone finds a use for it and calls it directly)
- rewrite to drop the 'handled' temporary in iterating over the
  decoder table, removing the assignment in the condition of the if
- rename to trans_xventanacondops.c.inc (i.e. with the '.c')
- (in MATERIALISE_EXT_PREDICATE) don't annotate the predicate function
  for testing the availability of individual extensions as 'inline'
  and don't make CPURISCVState* visible to these predicate functions
- add a MAINTAINERS entry for XVentanaCondOps

Changes in v2:
- (new patch) iterate over a table of guarded decoder functions
- Split off decode table into XVentanaCondOps.decode
- Wire up XVentanaCondOps in the decoder-table

Philipp Tomsich (7):
  target/riscv: refactor (anonymous struct) RISCVCPU.cfg into 'struct
    RISCVCPUConfig'
  target/riscv: riscv_tr_init_disas_context: copy pointer-to-cfg into
    cfg_ptr
  target/riscv: access configuration through cfg_ptr in DisasContext
  target/riscv: access cfg structure through DisasContext
  target/riscv: iterate over a table of decoders
  target/riscv: Add XVentanaCondOps custom extension
  target/riscv: add a MAINTAINERS entry for XVentanaCondOps

 MAINTAINERS                                   |   7 ++
 target/riscv/XVentanaCondOps.decode           |  25 +++++
 target/riscv/cpu.c                            |   3 +
 target/riscv/cpu.h                            |  79 ++++++-------
 target/riscv/insn_trans/trans_rvb.c.inc       |   8 +-
 target/riscv/insn_trans/trans_rvi.c.inc       |   2 +-
 target/riscv/insn_trans/trans_rvv.c.inc       | 104 +++++++++---------
 target/riscv/insn_trans/trans_rvzfh.c.inc     |   4 +-
 .../insn_trans/trans_xventanacondops.c.inc    |  39 +++++++
 target/riscv/meson.build                      |   1 +
 target/riscv/translate.c                      |  62 +++++++----
 11 files changed, 218 insertions(+), 116 deletions(-)
 create mode 100644 target/riscv/XVentanaCondOps.decode
 create mode 100644 target/riscv/insn_trans/trans_xventanacondops.c.inc

-- 
2.33.1

[PATCH v3 1/7] target/riscv: refactor (anonymous struct) RISCVCPU.cfg into 'struct RISCVCPUConfig'

[Philipp Tomsich]

Signed-off-by: Philipp Tomsich <philipp.tomsich@vrull.eu>
Suggested-by: Richard Henderson <richard.henderson@linaro.org>

---

Changes in v3:
- (new patch) refactor 'struct RISCVCPUConfig'

 target/riscv/cpu.h | 76 ++++++++++++++++++++++++----------------------
 1 file changed, 39 insertions(+), 37 deletions(-)

diff --git a/target/riscv/cpu.h b/target/riscv/cpu.h
index 55635d68d5..9c795089b6 100644
--- a/target/riscv/cpu.h
+++ b/target/riscv/cpu.h
@@ -303,6 +303,44 @@ struct RISCVCPUClass {
     DeviceReset parent_reset;
 };
 
+struct RISCVCPUConfig {
+    bool ext_i;
+    bool ext_e;
+    bool ext_g;
+    bool ext_m;
+    bool ext_a;
+    bool ext_f;
+    bool ext_d;
+    bool ext_c;
+    bool ext_s;
+    bool ext_u;
+    bool ext_h;
+    bool ext_j;
+    bool ext_v;
+    bool ext_zba;
+    bool ext_zbb;
+    bool ext_zbc;
+    bool ext_zbs;
+    bool ext_counters;
+    bool ext_ifencei;
+    bool ext_icsr;
+    bool ext_zfh;
+    bool ext_zfhmin;
+    bool ext_zve32f;
+    bool ext_zve64f;
+
+    char *priv_spec;
+    char *user_spec;
+    char *bext_spec;
+    char *vext_spec;
+    uint16_t vlen;
+    uint16_t elen;
+    bool mmu;
+    bool pmp;
+    bool epmp;
+    uint64_t resetvec;
+};
+
 /**
  * RISCVCPU:
  * @env: #CPURISCVState
@@ -320,43 +358,7 @@ struct RISCVCPU {
     char *dyn_vreg_xml;
 
     /* Configuration Settings */
-    struct {
-        bool ext_i;
-        bool ext_e;
-        bool ext_g;
-        bool ext_m;
-        bool ext_a;
-        bool ext_f;
-        bool ext_d;
-        bool ext_c;
-        bool ext_s;
-        bool ext_u;
-        bool ext_h;
-        bool ext_j;
-        bool ext_v;
-        bool ext_zba;
-        bool ext_zbb;
-        bool ext_zbc;
-        bool ext_zbs;
-        bool ext_counters;
-        bool ext_ifencei;
-        bool ext_icsr;
-        bool ext_zfh;
-        bool ext_zfhmin;
-        bool ext_zve32f;
-        bool ext_zve64f;
-
-        char *priv_spec;
-        char *user_spec;
-        char *bext_spec;
-        char *vext_spec;
-        uint16_t vlen;
-        uint16_t elen;
-        bool mmu;
-        bool pmp;
-        bool epmp;
-        uint64_t resetvec;
-    } cfg;
+    struct RISCVCPUConfig cfg;
 };
 
 static inline int riscv_has_ext(CPURISCVState *env, target_ulong ext)
-- 
2.33.1

Re: [PATCH v3 1/7] target/riscv: refactor (anonymous struct) RISCVCPU.cfg into 'struct RISCVCPUConfig'

[Richard Henderson]

On 1/29/22 01:56, Philipp Tomsich wrote:
> +    struct RISCVCPUConfig cfg;

Coding style says use a typedef for the struct.

Otherwise,
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>


r~

[PATCH v3 2/7] target/riscv: riscv_tr_init_disas_context: copy pointer-to-cfg into cfg_ptr

[Philipp Tomsich]

As the number of extensions is growing, copying them individiually
into the DisasContext will scale less and less... instead we populate
a pointer to the RISCVCPUConfig structure in the DisasContext.

This adds an extra indirection when checking for the availability of
an extension (compared to copying the fields into DisasContext).
While not a performance problem today, we can always (shallow) copy
the entire structure into the DisasContext (instead of putting a
pointer to it) if this is ever deemed necessary.

Signed-off-by: Philipp Tomsich <philipp.tomsich@vrull.eu>
Suggested-by: Richard Henderson <richard.henderson@linaro.org>

---

Changes in v3:
- (new patch) copy pointer to element cfg into DisasContext

 target/riscv/translate.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index f0bbe80875..22d09af2df 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -76,6 +76,7 @@ typedef struct DisasContext {
     int frm;
     RISCVMXL ol;
     bool virt_enabled;
+    const struct RISCVCPUConfig *cfg_ptr;
     bool ext_ifencei;
     bool ext_zfh;
     bool ext_zfhmin;
@@ -908,6 +909,7 @@ static void riscv_tr_init_disas_context(DisasContextBase *dcbase, CPUState *cs)
 #endif
     ctx->misa_ext = env->misa_ext;
     ctx->frm = -1;  /* unknown rounding mode */
+    ctx->cfg_ptr = &(cpu->cfg);
     ctx->ext_ifencei = cpu->cfg.ext_ifencei;
     ctx->ext_zfh = cpu->cfg.ext_zfh;
     ctx->ext_zfhmin = cpu->cfg.ext_zfhmin;
-- 
2.33.1

Re: [PATCH v3 2/7] target/riscv: riscv_tr_init_disas_context: copy pointer-to-cfg into cfg_ptr

[Richard Henderson]

On 1/29/22 01:56, Philipp Tomsich wrote:
> As the number of extensions is growing, copying them individiually
> into the DisasContext will scale less and less... instead we populate
> a pointer to the RISCVCPUConfig structure in the DisasContext.
> 
> This adds an extra indirection when checking for the availability of
> an extension (compared to copying the fields into DisasContext).
> While not a performance problem today, we can always (shallow) copy
> the entire structure into the DisasContext (instead of putting a
> pointer to it) if this is ever deemed necessary.
> 
> Signed-off-by: Philipp Tomsich<philipp.tomsich@vrull.eu>
> Suggested-by: Richard Henderson<richard.henderson@linaro.org>
> 
> ---
> 
> Changes in v3:
> - (new patch) copy pointer to element cfg into DisasContext
> 
>   target/riscv/translate.c | 2 ++
>   1 file changed, 2 insertions(+)

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

r~

[PATCH v3 3/7] target/riscv: access configuration through cfg_ptr in DisasContext

[Philipp Tomsich]

The implementation in trans_{rvi,rvv,rvzfh}.c.inc accesses the shallow
copies (in DisasContext) of some of the elements available in the
RISCVCPUConfig structure.  This commit redirects accesses to use the
cfg_ptr copied into DisasContext and removes the shallow copies.

Signed-off-by: Philipp Tomsich <philipp.tomsich@vrull.eu>
Suggested-by: Richard Henderson <richard.henderson@linaro.org>

---

Changes in v3:
- (new patch) test extension-availability through cfg_ptr in
  DisasContext, removing the fields that have been copied into
  DisasContext directly

 target/riscv/insn_trans/trans_rvi.c.inc   |   2 +-
 target/riscv/insn_trans/trans_rvv.c.inc   | 104 +++++++++++-----------
 target/riscv/insn_trans/trans_rvzfh.c.inc |   4 +-
 target/riscv/translate.c                  |  14 ---
 4 files changed, 55 insertions(+), 69 deletions(-)

diff --git a/target/riscv/insn_trans/trans_rvi.c.inc b/target/riscv/insn_trans/trans_rvi.c.inc
index 3cd1b3f877..f1342f30f8 100644
--- a/target/riscv/insn_trans/trans_rvi.c.inc
+++ b/target/riscv/insn_trans/trans_rvi.c.inc
@@ -806,7 +806,7 @@ static bool trans_fence(DisasContext *ctx, arg_fence *a)
 
 static bool trans_fence_i(DisasContext *ctx, arg_fence_i *a)
 {
-    if (!ctx->ext_ifencei) {
+    if (!ctx->cfg_ptr->ext_ifencei) {
         return false;
     }
 
diff --git a/target/riscv/insn_trans/trans_rvv.c.inc b/target/riscv/insn_trans/trans_rvv.c.inc
index f85a9e83b4..ff09e345ad 100644
--- a/target/riscv/insn_trans/trans_rvv.c.inc
+++ b/target/riscv/insn_trans/trans_rvv.c.inc
@@ -74,7 +74,7 @@ static bool require_zve32f(DisasContext *s)
     }
 
     /* Zve32f doesn't support FP64. (Section 18.2) */
-    return s->ext_zve32f ? s->sew <= MO_32 : true;
+    return s->cfg_ptr->ext_zve32f ? s->sew <= MO_32 : true;
 }
 
 static bool require_scale_zve32f(DisasContext *s)
@@ -85,7 +85,7 @@ static bool require_scale_zve32f(DisasContext *s)
     }
 
     /* Zve32f doesn't support FP64. (Section 18.2) */
-    return s->ext_zve64f ? s->sew <= MO_16 : true;
+    return s->cfg_ptr->ext_zve64f ? s->sew <= MO_16 : true;
 }
 
 static bool require_zve64f(DisasContext *s)
@@ -96,7 +96,7 @@ static bool require_zve64f(DisasContext *s)
     }
 
     /* Zve64f doesn't support FP64. (Section 18.2) */
-    return s->ext_zve64f ? s->sew <= MO_32 : true;
+    return s->cfg_ptr->ext_zve64f ? s->sew <= MO_32 : true;
 }
 
 static bool require_scale_zve64f(DisasContext *s)
@@ -107,7 +107,7 @@ static bool require_scale_zve64f(DisasContext *s)
     }
 
     /* Zve64f doesn't support FP64. (Section 18.2) */
-    return s->ext_zve64f ? s->sew <= MO_16 : true;
+    return s->cfg_ptr->ext_zve64f ? s->sew <= MO_16 : true;
 }
 
 /* Destination vector register group cannot overlap source mask register. */
@@ -174,7 +174,7 @@ static bool do_vsetvl(DisasContext *s, int rd, int rs1, TCGv s2)
     TCGv s1, dst;
 
     if (!require_rvv(s) ||
-        !(has_ext(s, RVV) || s->ext_zve32f || s->ext_zve64f)) {
+        !(has_ext(s, RVV) || s->cfg_ptr->ext_zve32f || s->cfg_ptr->ext_zve64f)) {
         return false;
     }
 
@@ -210,7 +210,7 @@ static bool do_vsetivli(DisasContext *s, int rd, TCGv s1, TCGv s2)
     TCGv dst;
 
     if (!require_rvv(s) ||
-        !(has_ext(s, RVV) || s->ext_zve32f || s->ext_zve64f)) {
+        !(has_ext(s, RVV) || s->cfg_ptr->ext_zve32f || s->cfg_ptr->ext_zve64f)) {
         return false;
     }
 
@@ -248,7 +248,7 @@ static bool trans_vsetivli(DisasContext *s, arg_vsetivli *a)
 /* vector register offset from env */
 static uint32_t vreg_ofs(DisasContext *s, int reg)
 {
-    return offsetof(CPURISCVState, vreg) + reg * s->vlen / 8;
+    return offsetof(CPURISCVState, vreg) + reg * s->cfg_ptr->vlen / 8;
 }
 
 /* check functions */
@@ -318,7 +318,7 @@ static bool vext_check_st_index(DisasContext *s, int vd, int vs2, int nf,
      * when XLEN=32. (Section 18.2)
      */
     if (get_xl(s) == MXL_RV32) {
-        ret &= (!has_ext(s, RVV) && s->ext_zve64f ? eew != MO_64 : true);
+        ret &= (!has_ext(s, RVV) && s->cfg_ptr->ext_zve64f ? eew != MO_64 : true);
     }
 
     return ret;
@@ -454,7 +454,7 @@ static bool vext_wide_check_common(DisasContext *s, int vd, int vm)
 {
     return (s->lmul <= 2) &&
            (s->sew < MO_64) &&
-           ((s->sew + 1) <= (s->elen >> 4)) &&
+           ((s->sew + 1) <= (s->cfg_ptr->elen >> 4)) &&
            require_align(vd, s->lmul + 1) &&
            require_vm(vm, vd);
 }
@@ -482,7 +482,7 @@ static bool vext_narrow_check_common(DisasContext *s, int vd, int vs2,
 {
     return (s->lmul <= 2) &&
            (s->sew < MO_64) &&
-           ((s->sew + 1) <= (s->elen >> 4)) &&
+           ((s->sew + 1) <= (s->cfg_ptr->elen >> 4)) &&
            require_align(vs2, s->lmul + 1) &&
            require_align(vd, s->lmul) &&
            require_vm(vm, vd);
@@ -661,7 +661,7 @@ static bool ldst_us_trans(uint32_t vd, uint32_t rs1, uint32_t data,
      * The first part is vlen in bytes, encoded in maxsz of simd_desc.
      * The second part is lmul, encoded in data of simd_desc.
      */
-    desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data));
+    desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data));
 
     tcg_gen_addi_ptr(dest, cpu_env, vreg_ofs(s, vd));
     tcg_gen_addi_ptr(mask, cpu_env, vreg_ofs(s, 0));
@@ -819,7 +819,7 @@ static bool ldst_stride_trans(uint32_t vd, uint32_t rs1, uint32_t rs2,
     mask = tcg_temp_new_ptr();
     base = get_gpr(s, rs1, EXT_NONE);
     stride = get_gpr(s, rs2, EXT_NONE);
-    desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data));
+    desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data));
 
     tcg_gen_addi_ptr(dest, cpu_env, vreg_ofs(s, vd));
     tcg_gen_addi_ptr(mask, cpu_env, vreg_ofs(s, 0));
@@ -925,7 +925,7 @@ static bool ldst_index_trans(uint32_t vd, uint32_t rs1, uint32_t vs2,
     mask = tcg_temp_new_ptr();
     index = tcg_temp_new_ptr();
     base = get_gpr(s, rs1, EXT_NONE);
-    desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data));
+    desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data));
 
     tcg_gen_addi_ptr(dest, cpu_env, vreg_ofs(s, vd));
     tcg_gen_addi_ptr(index, cpu_env, vreg_ofs(s, vs2));
@@ -1065,7 +1065,7 @@ static bool ldff_trans(uint32_t vd, uint32_t rs1, uint32_t data,
     dest = tcg_temp_new_ptr();
     mask = tcg_temp_new_ptr();
     base = get_gpr(s, rs1, EXT_NONE);
-    desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data));
+    desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data));
 
     tcg_gen_addi_ptr(dest, cpu_env, vreg_ofs(s, vd));
     tcg_gen_addi_ptr(mask, cpu_env, vreg_ofs(s, 0));
@@ -1120,7 +1120,7 @@ static bool ldst_whole_trans(uint32_t vd, uint32_t rs1, uint32_t nf,
 
     uint32_t data = FIELD_DP32(0, VDATA, NF, nf);
     dest = tcg_temp_new_ptr();
-    desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data));
+    desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data));
 
     base = get_gpr(s, rs1, EXT_NONE);
     tcg_gen_addi_ptr(dest, cpu_env, vreg_ofs(s, vd));
@@ -1185,7 +1185,7 @@ GEN_LDST_WHOLE_TRANS(vs8r_v, 8, true)
 static inline uint32_t MAXSZ(DisasContext *s)
 {
     int scale = s->lmul - 3;
-    return scale < 0 ? s->vlen >> -scale : s->vlen << scale;
+    return scale < 0 ? s->cfg_ptr->vlen >> -scale : s->cfg_ptr->vlen << scale;
 }
 
 static bool opivv_check(DisasContext *s, arg_rmrr *a)
@@ -1220,7 +1220,7 @@ do_opivv_gvec(DisasContext *s, arg_rmrr *a, GVecGen3Fn *gvec_fn,
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
                            vreg_ofs(s, a->rs1), vreg_ofs(s, a->rs2),
-                           cpu_env, s->vlen / 8, s->vlen / 8, data, fn);
+                           cpu_env, s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data, fn);
     }
     mark_vs_dirty(s);
     gen_set_label(over);
@@ -1262,7 +1262,7 @@ static bool opivx_trans(uint32_t vd, uint32_t rs1, uint32_t vs2, uint32_t vm,
 
     data = FIELD_DP32(data, VDATA, VM, vm);
     data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
-    desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data));
+    desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data));
 
     tcg_gen_addi_ptr(dest, cpu_env, vreg_ofs(s, vd));
     tcg_gen_addi_ptr(src2, cpu_env, vreg_ofs(s, vs2));
@@ -1425,7 +1425,7 @@ static bool opivi_trans(uint32_t vd, uint32_t imm, uint32_t vs2, uint32_t vm,
 
     data = FIELD_DP32(data, VDATA, VM, vm);
     data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
-    desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data));
+    desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data));
 
     tcg_gen_addi_ptr(dest, cpu_env, vreg_ofs(s, vd));
     tcg_gen_addi_ptr(src2, cpu_env, vreg_ofs(s, vs2));
@@ -1508,7 +1508,7 @@ static bool do_opivv_widen(DisasContext *s, arg_rmrr *a,
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
                            vreg_ofs(s, a->rs1),
                            vreg_ofs(s, a->rs2),
-                           cpu_env, s->vlen / 8, s->vlen / 8,
+                           cpu_env, s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8,
                            data, fn);
         mark_vs_dirty(s);
         gen_set_label(over);
@@ -1587,7 +1587,7 @@ static bool do_opiwv_widen(DisasContext *s, arg_rmrr *a,
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
                            vreg_ofs(s, a->rs1),
                            vreg_ofs(s, a->rs2),
-                           cpu_env, s->vlen / 8, s->vlen / 8, data, fn);
+                           cpu_env, s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data, fn);
         mark_vs_dirty(s);
         gen_set_label(over);
         return true;
@@ -1663,7 +1663,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs1),                    \
                            vreg_ofs(s, a->rs2), cpu_env,           \
-                           s->vlen / 8, s->vlen / 8, data,         \
+                           s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data,         \
                            fns[s->sew]);                           \
         mark_vs_dirty(s);                                          \
         gen_set_label(over);                                       \
@@ -1843,7 +1843,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs1),                    \
                            vreg_ofs(s, a->rs2), cpu_env,           \
-                           s->vlen / 8, s->vlen / 8, data,         \
+                           s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data,         \
                            fns[s->sew]);                           \
         mark_vs_dirty(s);                                          \
         gen_set_label(over);                                       \
@@ -1963,7 +1963,7 @@ static bool vmulh_vv_check(DisasContext *s, arg_rmrr *a)
      * are not included for EEW=64 in Zve64*. (Section 18.2)
      */
     return opivv_check(s, a) &&
-           (!has_ext(s, RVV) && s->ext_zve64f ? s->sew != MO_64 : true);
+           (!has_ext(s, RVV) && s->cfg_ptr->ext_zve64f ? s->sew != MO_64 : true);
 }
 
 static bool vmulh_vx_check(DisasContext *s, arg_rmrr *a)
@@ -1976,7 +1976,7 @@ static bool vmulh_vx_check(DisasContext *s, arg_rmrr *a)
      * are not included for EEW=64 in Zve64*. (Section 18.2)
      */
     return opivx_check(s, a) &&
-           (!has_ext(s, RVV) && s->ext_zve64f ? s->sew != MO_64 : true);
+           (!has_ext(s, RVV) && s->cfg_ptr->ext_zve64f ? s->sew != MO_64 : true);
 }
 
 GEN_OPIVV_GVEC_TRANS(vmul_vv,  mul)
@@ -2046,7 +2046,7 @@ static bool trans_vmv_v_v(DisasContext *s, arg_vmv_v_v *a)
             tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
 
             tcg_gen_gvec_2_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, a->rs1),
-                               cpu_env, s->vlen / 8, s->vlen / 8, data,
+                               cpu_env, s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data,
                                fns[s->sew]);
             gen_set_label(over);
         }
@@ -2083,7 +2083,7 @@ static bool trans_vmv_v_x(DisasContext *s, arg_vmv_v_x *a)
             };
 
             tcg_gen_ext_tl_i64(s1_i64, s1);
-            desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data));
+            desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data));
             tcg_gen_addi_ptr(dest, cpu_env, vreg_ofs(s, a->rd));
             fns[s->sew](dest, s1_i64, cpu_env, desc);
 
@@ -2123,7 +2123,7 @@ static bool trans_vmv_v_i(DisasContext *s, arg_vmv_v_i *a)
 
             s1 = tcg_constant_i64(simm);
             dest = tcg_temp_new_ptr();
-            desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data));
+            desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data));
             tcg_gen_addi_ptr(dest, cpu_env, vreg_ofs(s, a->rd));
             fns[s->sew](dest, s1, cpu_env, desc);
 
@@ -2176,7 +2176,7 @@ static bool vsmul_vv_check(DisasContext *s, arg_rmrr *a)
      * for EEW=64 in Zve64*. (Section 18.2)
      */
     return opivv_check(s, a) &&
-           (!has_ext(s, RVV) && s->ext_zve64f ? s->sew != MO_64 : true);
+           (!has_ext(s, RVV) && s->cfg_ptr->ext_zve64f ? s->sew != MO_64 : true);
 }
 
 static bool vsmul_vx_check(DisasContext *s, arg_rmrr *a)
@@ -2187,7 +2187,7 @@ static bool vsmul_vx_check(DisasContext *s, arg_rmrr *a)
      * for EEW=64 in Zve64*. (Section 18.2)
      */
     return opivx_check(s, a) &&
-           (!has_ext(s, RVV) && s->ext_zve64f ? s->sew != MO_64 : true);
+           (!has_ext(s, RVV) && s->cfg_ptr->ext_zve64f ? s->sew != MO_64 : true);
 }
 
 GEN_OPIVV_TRANS(vsmul_vv, vsmul_vv_check)
@@ -2275,7 +2275,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs1),                    \
                            vreg_ofs(s, a->rs2), cpu_env,           \
-                           s->vlen / 8, s->vlen / 8, data,         \
+                           s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data,         \
                            fns[s->sew - 1]);                       \
         mark_vs_dirty(s);                                          \
         gen_set_label(over);                                       \
@@ -2302,7 +2302,7 @@ static bool opfvf_trans(uint32_t vd, uint32_t rs1, uint32_t vs2,
     dest = tcg_temp_new_ptr();
     mask = tcg_temp_new_ptr();
     src2 = tcg_temp_new_ptr();
-    desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data));
+    desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data));
 
     tcg_gen_addi_ptr(dest, cpu_env, vreg_ofs(s, vd));
     tcg_gen_addi_ptr(src2, cpu_env, vreg_ofs(s, vs2));
@@ -2391,7 +2391,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)           \
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),   \
                            vreg_ofs(s, a->rs1),                  \
                            vreg_ofs(s, a->rs2), cpu_env,         \
-                           s->vlen / 8, s->vlen / 8, data,       \
+                           s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data,       \
                            fns[s->sew - 1]);                     \
         mark_vs_dirty(s);                                        \
         gen_set_label(over);                                     \
@@ -2464,7 +2464,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs1),                    \
                            vreg_ofs(s, a->rs2), cpu_env,           \
-                           s->vlen / 8, s->vlen / 8, data,         \
+                           s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data,         \
                            fns[s->sew - 1]);                       \
         mark_vs_dirty(s);                                          \
         gen_set_label(over);                                       \
@@ -2583,7 +2583,7 @@ static bool do_opfv(DisasContext *s, arg_rmr *a,
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
         tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
                            vreg_ofs(s, a->rs2), cpu_env,
-                           s->vlen / 8, s->vlen / 8, data, fn);
+                           s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data, fn);
         mark_vs_dirty(s);
         gen_set_label(over);
         return true;
@@ -2696,7 +2696,7 @@ static bool trans_vfmv_v_f(DisasContext *s, arg_vfmv_v_f *a)
             do_nanbox(s, t1, cpu_fpr[a->rs1]);
 
             dest = tcg_temp_new_ptr();
-            desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data));
+            desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data));
             tcg_gen_addi_ptr(dest, cpu_env, vreg_ofs(s, a->rd));
 
             fns[s->sew - 1](dest, t1, cpu_env, desc);
@@ -2782,7 +2782,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
         tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs2), cpu_env,           \
-                           s->vlen / 8, s->vlen / 8, data,         \
+                           s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data,         \
                            fns[s->sew - 1]);                       \
         mark_vs_dirty(s);                                          \
         gen_set_label(over);                                       \
@@ -2831,7 +2831,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs2), cpu_env,           \
-                           s->vlen / 8, s->vlen / 8, data,         \
+                           s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data,         \
                            fns[s->sew]);                           \
         mark_vs_dirty(s);                                          \
         gen_set_label(over);                                       \
@@ -2896,7 +2896,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
         tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs2), cpu_env,           \
-                           s->vlen / 8, s->vlen / 8, data,         \
+                           s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data,         \
                            fns[s->sew - 1]);                       \
         mark_vs_dirty(s);                                          \
         gen_set_label(over);                                       \
@@ -2947,7 +2947,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
         data = FIELD_DP32(data, VDATA, VM, a->vm);                 \
         tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs2), cpu_env,           \
-                           s->vlen / 8, s->vlen / 8, data,         \
+                           s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data,         \
                            fns[s->sew]);                           \
         mark_vs_dirty(s);                                          \
         gen_set_label(over);                                       \
@@ -2986,7 +2986,7 @@ GEN_OPIVV_TRANS(vredxor_vs, reduction_check)
 static bool reduction_widen_check(DisasContext *s, arg_rmrr *a)
 {
     return reduction_check(s, a) && (s->sew < MO_64) &&
-           ((s->sew + 1) <= (s->elen >> 4));
+           ((s->sew + 1) <= (s->cfg_ptr->elen >> 4));
 }
 
 GEN_OPIVV_WIDEN_TRANS(vwredsum_vs, reduction_widen_check)
@@ -3034,7 +3034,7 @@ static bool trans_##NAME(DisasContext *s, arg_r *a)                \
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
                            vreg_ofs(s, a->rs1),                    \
                            vreg_ofs(s, a->rs2), cpu_env,           \
-                           s->vlen / 8, s->vlen / 8, data, fn);    \
+                           s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data, fn);    \
         mark_vs_dirty(s);                                          \
         gen_set_label(over);                                       \
         return true;                                               \
@@ -3067,7 +3067,7 @@ static bool trans_vcpop_m(DisasContext *s, arg_rmr *a)
         mask = tcg_temp_new_ptr();
         src2 = tcg_temp_new_ptr();
         dst = dest_gpr(s, a->rd);
-        desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data));
+        desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data));
 
         tcg_gen_addi_ptr(src2, cpu_env, vreg_ofs(s, a->rs2));
         tcg_gen_addi_ptr(mask, cpu_env, vreg_ofs(s, 0));
@@ -3099,7 +3099,7 @@ static bool trans_vfirst_m(DisasContext *s, arg_rmr *a)
         mask = tcg_temp_new_ptr();
         src2 = tcg_temp_new_ptr();
         dst = dest_gpr(s, a->rd);
-        desc = tcg_constant_i32(simd_desc(s->vlen / 8, s->vlen / 8, data));
+        desc = tcg_constant_i32(simd_desc(s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data));
 
         tcg_gen_addi_ptr(src2, cpu_env, vreg_ofs(s, a->rs2));
         tcg_gen_addi_ptr(mask, cpu_env, vreg_ofs(s, 0));
@@ -3134,7 +3134,7 @@ static bool trans_##NAME(DisasContext *s, arg_rmr *a)              \
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);             \
         tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd),                     \
                            vreg_ofs(s, 0), vreg_ofs(s, a->rs2),    \
-                           cpu_env, s->vlen / 8, s->vlen / 8,      \
+                           cpu_env, s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8,      \
                            data, fn);                              \
         mark_vs_dirty(s);                                          \
         gen_set_label(over);                                       \
@@ -3174,7 +3174,7 @@ static bool trans_viota_m(DisasContext *s, arg_viota_m *a)
         };
         tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
                            vreg_ofs(s, a->rs2), cpu_env,
-                           s->vlen / 8, s->vlen / 8, data, fns[s->sew]);
+                           s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data, fns[s->sew]);
         mark_vs_dirty(s);
         gen_set_label(over);
         return true;
@@ -3200,7 +3200,7 @@ static bool trans_vid_v(DisasContext *s, arg_vid_v *a)
             gen_helper_vid_v_w, gen_helper_vid_v_d,
         };
         tcg_gen_gvec_2_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
-                           cpu_env, s->vlen / 8, s->vlen / 8,
+                           cpu_env, s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8,
                            data, fns[s->sew]);
         mark_vs_dirty(s);
         gen_set_label(over);
@@ -3554,7 +3554,7 @@ static bool trans_vrgather_vx(DisasContext *s, arg_rmrr *a)
 
     if (a->vm && s->vl_eq_vlmax) {
         int scale = s->lmul - (s->sew + 3);
-        int vlmax = scale < 0 ? s->vlen >> -scale : s->vlen << scale;
+        int vlmax = scale < 0 ? s->cfg_ptr->vlen >> -scale : s->cfg_ptr->vlen << scale;
         TCGv_i64 dest = tcg_temp_new_i64();
 
         if (a->rs1 == 0) {
@@ -3586,7 +3586,7 @@ static bool trans_vrgather_vi(DisasContext *s, arg_rmrr *a)
 
     if (a->vm && s->vl_eq_vlmax) {
         int scale = s->lmul - (s->sew + 3);
-        int vlmax = scale < 0 ? s->vlen >> -scale : s->vlen << scale;
+        int vlmax = scale < 0 ? s->cfg_ptr->vlen >> -scale : s->cfg_ptr->vlen << scale;
         if (a->rs1 >= vlmax) {
             tcg_gen_gvec_dup_imm(MO_64, vreg_ofs(s, a->rd),
                                  MAXSZ(s), MAXSZ(s), 0);
@@ -3638,7 +3638,7 @@ static bool trans_vcompress_vm(DisasContext *s, arg_r *a)
         data = FIELD_DP32(data, VDATA, LMUL, s->lmul);
         tcg_gen_gvec_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
                            vreg_ofs(s, a->rs1), vreg_ofs(s, a->rs2),
-                           cpu_env, s->vlen / 8, s->vlen / 8, data,
+                           cpu_env, s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data,
                            fns[s->sew]);
         mark_vs_dirty(s);
         gen_set_label(over);
@@ -3657,7 +3657,7 @@ static bool trans_##NAME(DisasContext *s, arg_##NAME * a)               \
     if (require_rvv(s) &&                                               \
         QEMU_IS_ALIGNED(a->rd, LEN) &&                                  \
         QEMU_IS_ALIGNED(a->rs2, LEN)) {                                 \
-        uint32_t maxsz = (s->vlen >> 3) * LEN;                          \
+        uint32_t maxsz = (s->cfg_ptr->vlen >> 3) * LEN;                          \
         if (s->vstart == 0) {                                           \
             /* EEW = 8 */                                               \
             tcg_gen_gvec_mov(MO_8, vreg_ofs(s, a->rd),                  \
@@ -3742,7 +3742,7 @@ static bool int_ext_op(DisasContext *s, arg_rmr *a, uint8_t seq)
 
     tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),
                        vreg_ofs(s, a->rs2), cpu_env,
-                       s->vlen / 8, s->vlen / 8, data, fn);
+                       s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, data, fn);
 
     mark_vs_dirty(s);
     gen_set_label(over);
diff --git a/target/riscv/insn_trans/trans_rvzfh.c.inc b/target/riscv/insn_trans/trans_rvzfh.c.inc
index 5a7cac8958..608c51da2c 100644
--- a/target/riscv/insn_trans/trans_rvzfh.c.inc
+++ b/target/riscv/insn_trans/trans_rvzfh.c.inc
@@ -17,13 +17,13 @@
  */
 
 #define REQUIRE_ZFH(ctx) do { \
-    if (!ctx->ext_zfh) {      \
+    if (!ctx->cfg_ptr->ext_zfh) {      \
         return false;         \
     }                         \
 } while (0)
 
 #define REQUIRE_ZFH_OR_ZFHMIN(ctx) do {       \
-    if (!(ctx->ext_zfh || ctx->ext_zfhmin)) { \
+    if (!(ctx->cfg_ptr->ext_zfh || ctx->cfg_ptr->ext_zfhmin)) { \
         return false;                         \
     }                                         \
 } while (0)
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index 22d09af2df..441690846c 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -77,11 +77,6 @@ typedef struct DisasContext {
     RISCVMXL ol;
     bool virt_enabled;
     const struct RISCVCPUConfig *cfg_ptr;
-    bool ext_ifencei;
-    bool ext_zfh;
-    bool ext_zfhmin;
-    bool ext_zve32f;
-    bool ext_zve64f;
     bool hlsx;
     /* vector extension */
     bool vill;
@@ -99,8 +94,6 @@ typedef struct DisasContext {
      */
     int8_t lmul;
     uint8_t sew;
-    uint16_t vlen;
-    uint16_t elen;
     target_ulong vstart;
     bool vl_eq_vlmax;
     uint8_t ntemp;
@@ -910,13 +903,6 @@ static void riscv_tr_init_disas_context(DisasContextBase *dcbase, CPUState *cs)
     ctx->misa_ext = env->misa_ext;
     ctx->frm = -1;  /* unknown rounding mode */
     ctx->cfg_ptr = &(cpu->cfg);
-    ctx->ext_ifencei = cpu->cfg.ext_ifencei;
-    ctx->ext_zfh = cpu->cfg.ext_zfh;
-    ctx->ext_zfhmin = cpu->cfg.ext_zfhmin;
-    ctx->ext_zve32f = cpu->cfg.ext_zve32f;
-    ctx->ext_zve64f = cpu->cfg.ext_zve64f;
-    ctx->vlen = cpu->cfg.vlen;
-    ctx->elen = cpu->cfg.elen;
     ctx->mstatus_hs_fs = FIELD_EX32(tb_flags, TB_FLAGS, MSTATUS_HS_FS);
     ctx->mstatus_hs_vs = FIELD_EX32(tb_flags, TB_FLAGS, MSTATUS_HS_VS);
     ctx->hlsx = FIELD_EX32(tb_flags, TB_FLAGS, HLSX);
-- 
2.33.1

Re: [PATCH v3 3/7] target/riscv: access configuration through cfg_ptr in DisasContext

[Richard Henderson]

On 1/29/22 01:56, Philipp Tomsich wrote:
> The implementation in trans_{rvi,rvv,rvzfh}.c.inc accesses the shallow
> copies (in DisasContext) of some of the elements available in the
> RISCVCPUConfig structure.  This commit redirects accesses to use the
> cfg_ptr copied into DisasContext and removes the shallow copies.
> 
> Signed-off-by: Philipp Tomsich<philipp.tomsich@vrull.eu>
> Suggested-by: Richard Henderson<richard.henderson@linaro.org>
> 
> ---
> 
> Changes in v3:
> - (new patch) test extension-availability through cfg_ptr in
>    DisasContext, removing the fields that have been copied into
>    DisasContext directly
> 
>   target/riscv/insn_trans/trans_rvi.c.inc   |   2 +-
>   target/riscv/insn_trans/trans_rvv.c.inc   | 104 +++++++++++-----------
>   target/riscv/insn_trans/trans_rvzfh.c.inc |   4 +-
>   target/riscv/translate.c                  |  14 ---
>   4 files changed, 55 insertions(+), 69 deletions(-)

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

r~

[PATCH v3 4/7] target/riscv: access cfg structure through DisasContext

[Philipp Tomsich]

The Zb[abcs] support code still uses the RISCV_CPU macros to access
the configuration information (i.e., check whether an extension is
available/enabled).  Now that we provide this information directly
from DisasContext, we can access this directly via the cfg_ptr field.

Signed-off-by: Philipp Tomsich <philipp.tomsich@vrull.eu>
Suggested-by: Richard Henderson <richard.henderson@linaro.org>

---

Changes in v3:
- (new patch) change Zb[abcs] implementation to use cfg_ptr (copied
  into DisasContext) instead of going throuhg RISCV_CPU

 target/riscv/insn_trans/trans_rvb.c.inc | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/target/riscv/insn_trans/trans_rvb.c.inc b/target/riscv/insn_trans/trans_rvb.c.inc
index 810431a1d6..f9bd3b7ec4 100644
--- a/target/riscv/insn_trans/trans_rvb.c.inc
+++ b/target/riscv/insn_trans/trans_rvb.c.inc
@@ -19,25 +19,25 @@
  */
 
 #define REQUIRE_ZBA(ctx) do {                    \
-    if (!RISCV_CPU(ctx->cs)->cfg.ext_zba) {      \
+    if (ctx->cfg_ptr->ext_zba) {                 \
         return false;                            \
     }                                            \
 } while (0)
 
 #define REQUIRE_ZBB(ctx) do {                    \
-    if (!RISCV_CPU(ctx->cs)->cfg.ext_zbb) {      \
+    if (ctx->cfg_ptr->ext_zbb) {                 \
         return false;                            \
     }                                            \
 } while (0)
 
 #define REQUIRE_ZBC(ctx) do {                    \
-    if (!RISCV_CPU(ctx->cs)->cfg.ext_zbc) {      \
+    if (ctx->cfg_ptr->ext_zbc) {                 \
         return false;                            \
     }                                            \
 } while (0)
 
 #define REQUIRE_ZBS(ctx) do {                    \
-    if (!RISCV_CPU(ctx->cs)->cfg.ext_zbs) {      \
+    if (ctx->cfg_ptr->ext_zbs) {                 \
         return false;                            \
     }                                            \
 } while (0)
-- 
2.33.1

Re: [PATCH v3 4/7] target/riscv: access cfg structure through DisasContext

[Richard Henderson]

On 1/29/22 01:56, Philipp Tomsich wrote:
> The Zb[abcs] support code still uses the RISCV_CPU macros to access
> the configuration information (i.e., check whether an extension is
> available/enabled).  Now that we provide this information directly
> from DisasContext, we can access this directly via the cfg_ptr field.
> 
> Signed-off-by: Philipp Tomsich<philipp.tomsich@vrull.eu>
> Suggested-by: Richard Henderson<richard.henderson@linaro.org>
> 
> ---
> 
> Changes in v3:
> - (new patch) change Zb[abcs] implementation to use cfg_ptr (copied
>    into DisasContext) instead of going throuhg RISCV_CPU
> 
>   target/riscv/insn_trans/trans_rvb.c.inc | 8 ++++----
>   1 file changed, 4 insertions(+), 4 deletions(-)

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

r~

[PATCH v3 5/7] target/riscv: iterate over a table of decoders

[Philipp Tomsich]

To split up the decoder into multiple functions (both to support
vendor-specific opcodes in separate files and to simplify maintenance
of orthogonal extensions), this changes decode_op to iterate over a
table of decoders predicated on guard functions.

This commit only adds the new structure and the table, allowing for
the easy addition of additional decoders in the future.

Signed-off-by: Philipp Tomsich <philipp.tomsich@vrull.eu>
---

Changes in v3:
- expose only the DisasContext* to predicate functions
- mark the table of decoder functions as static
- drop the inline from always_true_p, until the need arises (i.e.,
  someone finds a use for it and calls it directly)
- rewrite to drop the 'handled' temporary in iterating over the
  decoder table, removing the assignment in the condition of the if

Changes in v2:
- (new patch) iterate over a table of guarded decoder functions

 target/riscv/translate.c | 34 ++++++++++++++++++++++++++++------
 1 file changed, 28 insertions(+), 6 deletions(-)

diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index 441690846c..38fdec2fd9 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -111,6 +111,11 @@ static inline bool has_ext(DisasContext *ctx, uint32_t ext)
     return ctx->misa_ext & ext;
 }
 
+static bool always_true_p(DisasContext *ctx  __attribute__((__unused__)))
+{
+    return true;
+}
+
 #ifdef TARGET_RISCV32
 #define get_xl(ctx)    MXL_RV32
 #elif defined(CONFIG_USER_ONLY)
@@ -855,16 +860,26 @@ static uint32_t opcode_at(DisasContextBase *dcbase, target_ulong pc)
 
 static void decode_opc(CPURISCVState *env, DisasContext *ctx, uint16_t opcode)
 {
-    /* check for compressed insn */
+    /*
+     * A table with predicate (i.e., guard) functions and decoder functions
+     * that are tested in-order until a decoder matches onto the opcode.
+     */
+    static const struct {
+        bool (*guard_func)(DisasContext *);
+        bool (*decode_func)(DisasContext *, uint32_t);
+    } decoders[] = {
+        { always_true_p,  decode_insn32 },
+    };
+
+    /* Check for compressed insn */
     if (extract16(opcode, 0, 2) != 3) {
         if (!has_ext(ctx, RVC)) {
             gen_exception_illegal(ctx);
         } else {
             ctx->opcode = opcode;
             ctx->pc_succ_insn = ctx->base.pc_next + 2;
-            if (!decode_insn16(ctx, opcode)) {
-                gen_exception_illegal(ctx);
-            }
+            if (decode_insn16(ctx, opcode))
+                return;
         }
     } else {
         uint32_t opcode32 = opcode;
@@ -873,10 +888,17 @@ static void decode_opc(CPURISCVState *env, DisasContext *ctx, uint16_t opcode)
                                              ctx->base.pc_next + 2));
         ctx->opcode = opcode32;
         ctx->pc_succ_insn = ctx->base.pc_next + 4;
-        if (!decode_insn32(ctx, opcode32)) {
-            gen_exception_illegal(ctx);
+
+        for (size_t i = 0; i < ARRAY_SIZE(decoders); ++i) {
+            if (!decoders[i].guard_func(ctx))
+                continue;
+
+            if (decoders[i].decode_func(ctx, opcode32))
+                return;
         }
     }
+
+    gen_exception_illegal(ctx);
 }
 
 static void riscv_tr_init_disas_context(DisasContextBase *dcbase, CPUState *cs)
-- 
2.33.1

Re: [PATCH v3 5/7] target/riscv: iterate over a table of decoders

[Richard Henderson]

On 1/29/22 01:56, Philipp Tomsich wrote:
> -            if (!decode_insn16(ctx, opcode)) {
> -                gen_exception_illegal(ctx);
> -            }
> +            if (decode_insn16(ctx, opcode))
> +                return;
...
> -        if (!decode_insn32(ctx, opcode32)) {
> -            gen_exception_illegal(ctx);
> +
> +        for (size_t i = 0; i < ARRAY_SIZE(decoders); ++i) {
> +            if (!decoders[i].guard_func(ctx))
> +                continue;
> +
> +            if (decoders[i].decode_func(ctx, opcode32))
> +                return;

Missing braces, per style.  Otherwise,

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

r~

[PATCH v3 6/7] target/riscv: Add XVentanaCondOps custom extension

[Philipp Tomsich]

This adds the decoder and translation for the XVentanaCondOps custom
extension (vendor-defined by Ventana Micro Systems), which is
documented at https://github.com/ventanamicro/ventana-custom-extensions/releases/download/v1.0.0/ventana-custom-extensions-v1.0.0.pdf

This commit then also adds a guard-function (has_XVentanaCondOps_p)
and the decoder function to the table of decoders, enabling the
support for the XVentanaCondOps extension.

Signed-off-by: Philipp Tomsich <philipp.tomsich@vrull.eu>

---

Changes in v3:
- rename to trans_xventanacondops.c.inc (i.e. with the '.c')
- (in MATERIALISE_EXT_PREDICATE) don't annotate the predicate function
  for testing the availability of individual extensions as 'inline'
  and don't make CPURISCVState* visible to these predicate functions

Changes in v2:
- Split off decode table into XVentanaCondOps.decode
- Wire up XVentanaCondOps in the decoder-table

 target/riscv/XVentanaCondOps.decode           | 25 ++++++++++++
 target/riscv/cpu.c                            |  3 ++
 target/riscv/cpu.h                            |  3 ++
 .../insn_trans/trans_xventanacondops.c.inc    | 39 +++++++++++++++++++
 target/riscv/meson.build                      |  1 +
 target/riscv/translate.c                      | 12 ++++++
 6 files changed, 83 insertions(+)
 create mode 100644 target/riscv/XVentanaCondOps.decode
 create mode 100644 target/riscv/insn_trans/trans_xventanacondops.c.inc

diff --git a/target/riscv/XVentanaCondOps.decode b/target/riscv/XVentanaCondOps.decode
new file mode 100644
index 0000000000..5aef7c3d72
--- /dev/null
+++ b/target/riscv/XVentanaCondOps.decode
@@ -0,0 +1,25 @@
+#
+# RISC-V translation routines for the XVentanaCondOps extension
+#
+# Copyright (c) 2022 Dr. Philipp Tomsich, philipp.tomsich@vrull.eu
+#
+# SPDX-License-Identifier: LGPL-2.1-or-later
+#
+# Reference: VTx-family custom instructions
+#            Custom ISA extensions for Ventana Micro Systems RISC-V cores
+#            (https://github.com/ventanamicro/ventana-custom-extensions/releases/download/v1.0.0/ventana-custom-extensions-v1.0.0.pdf)
+
+# Fields
+%rs2  20:5
+%rs1  15:5
+%rd    7:5
+
+# Argument sets
+&r    rd rs1 rs2  !extern
+
+# Formats
+@r         .......  ..... ..... ... ..... ....... &r                %rs2 %rs1 %rd
+
+# *** RV64 Custom-3 Extension ***
+vt_maskc   0000000  ..... ..... 110 ..... 1111011 @r
+vt_maskcn  0000000  ..... ..... 111 ..... 1111011 @r
diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c
index 1cb0436187..6df07b8289 100644
--- a/target/riscv/cpu.c
+++ b/target/riscv/cpu.c
@@ -734,6 +734,9 @@ static Property riscv_cpu_properties[] = {
     DEFINE_PROP_BOOL("zbc", RISCVCPU, cfg.ext_zbc, true),
     DEFINE_PROP_BOOL("zbs", RISCVCPU, cfg.ext_zbs, true),
 
+    /* Vendor-specific custom extensions */
+    DEFINE_PROP_BOOL("xventanacondops", RISCVCPU, cfg.ext_XVentanaCondOps, false),
+
     /* These are experimental so mark with 'x-' */
     DEFINE_PROP_BOOL("x-j", RISCVCPU, cfg.ext_j, false),
     /* ePMP 0.9.3 */
diff --git a/target/riscv/cpu.h b/target/riscv/cpu.h
index 9c795089b6..ccf9ea9922 100644
--- a/target/riscv/cpu.h
+++ b/target/riscv/cpu.h
@@ -329,6 +329,9 @@ struct RISCVCPUConfig {
     bool ext_zve32f;
     bool ext_zve64f;
 
+    /* Vendor-specific custom extensions */
+    bool ext_XVentanaCondOps;
+
     char *priv_spec;
     char *user_spec;
     char *bext_spec;
diff --git a/target/riscv/insn_trans/trans_xventanacondops.c.inc b/target/riscv/insn_trans/trans_xventanacondops.c.inc
new file mode 100644
index 0000000000..b8a5d031b5
--- /dev/null
+++ b/target/riscv/insn_trans/trans_xventanacondops.c.inc
@@ -0,0 +1,39 @@
+/*
+ * RISC-V translation routines for the XVentanaCondOps extension.
+ *
+ * Copyright (c) 2021-2022 VRULL GmbH.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2 or later, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+static bool gen_condmask(DisasContext *ctx, arg_r *a, TCGCond cond)
+{
+    TCGv dest = dest_gpr(ctx, a->rd);
+    TCGv src1 = get_gpr(ctx, a->rs1, EXT_NONE);
+    TCGv src2 = get_gpr(ctx, a->rs2, EXT_NONE);
+
+    tcg_gen_movcond_tl(cond, dest, src2, ctx->zero, src1, ctx->zero);
+
+    gen_set_gpr(ctx, a->rd, dest);
+    return true;
+}
+
+static bool trans_vt_maskc(DisasContext *ctx, arg_r *a)
+{
+    return gen_condmask(ctx, a, TCG_COND_NE);
+}
+
+static bool trans_vt_maskcn(DisasContext *ctx, arg_r *a)
+{
+    return gen_condmask(ctx, a, TCG_COND_EQ);
+}
diff --git a/target/riscv/meson.build b/target/riscv/meson.build
index a3997ed580..91f0ac32ff 100644
--- a/target/riscv/meson.build
+++ b/target/riscv/meson.build
@@ -4,6 +4,7 @@ dir = meson.current_source_dir()
 gen = [
   decodetree.process('insn16.decode', extra_args: ['--static-decode=decode_insn16', '--insnwidth=16']),
   decodetree.process('insn32.decode', extra_args: '--static-decode=decode_insn32'),
+  decodetree.process('XVentanaCondOps.decode', extra_args: '--static-decode=decode_XVentanaCodeOps'),
 ]
 
 riscv_ss = ss.source_set()
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index 38fdec2fd9..89cf96b2e6 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -116,6 +116,14 @@ static bool always_true_p(DisasContext *ctx  __attribute__((__unused__)))
     return true;
 }
 
+#define MATERIALISE_EXT_PREDICATE(ext)  \
+    static bool has_ ## ext ## _p(DisasContext *ctx)    \
+    { \
+        return ctx->cfg_ptr->ext_ ## ext ; \
+    }
+
+MATERIALISE_EXT_PREDICATE(XVentanaCondOps);
+
 #ifdef TARGET_RISCV32
 #define get_xl(ctx)    MXL_RV32
 #elif defined(CONFIG_USER_ONLY)
@@ -854,9 +862,12 @@ static uint32_t opcode_at(DisasContextBase *dcbase, target_ulong pc)
 #include "insn_trans/trans_rvb.c.inc"
 #include "insn_trans/trans_rvzfh.c.inc"
 #include "insn_trans/trans_privileged.c.inc"
+#include "insn_trans/trans_xventanacondops.c.inc"
 
 /* Include the auto-generated decoder for 16 bit insn */
 #include "decode-insn16.c.inc"
+/* Include decoders for factored-out extensions */
+#include "decode-XVentanaCondOps.c.inc"
 
 static void decode_opc(CPURISCVState *env, DisasContext *ctx, uint16_t opcode)
 {
@@ -869,6 +880,7 @@ static void decode_opc(CPURISCVState *env, DisasContext *ctx, uint16_t opcode)
         bool (*decode_func)(DisasContext *, uint32_t);
     } decoders[] = {
         { always_true_p,  decode_insn32 },
+        { has_XVentanaCondOps_p,  decode_XVentanaCodeOps },
     };
 
     /* Check for compressed insn */
-- 
2.33.1

Re: [PATCH v3 6/7] target/riscv: Add XVentanaCondOps custom extension

[Richard Henderson]

On 1/29/22 01:56, Philipp Tomsich wrote:
> This adds the decoder and translation for the XVentanaCondOps custom
> extension (vendor-defined by Ventana Micro Systems), which is
> documented athttps://github.com/ventanamicro/ventana-custom-extensions/releases/download/v1.0.0/ventana-custom-extensions-v1.0.0.pdf
> 
> This commit then also adds a guard-function (has_XVentanaCondOps_p)
> and the decoder function to the table of decoders, enabling the
> support for the XVentanaCondOps extension.
> 
> Signed-off-by: Philipp Tomsich<philipp.tomsich@vrull.eu>
> 
> ---

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

r~

[PATCH v3 7/7] target/riscv: add a MAINTAINERS entry for XVentanaCondOps

[Philipp Tomsich]

The XVentanaCondOps extension is supported by VRULL on behalf of the
Ventana Micro.  Add myself as a point-of-contact.

Signed-off-by: Philipp Tomsich <philipp.tomsich@vrull.eu>

---

Changes in v3:
- add a MAINTAINERS entry for XVentanaCondOps

 MAINTAINERS | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/MAINTAINERS b/MAINTAINERS
index e4b3a4bcdf..2ffcc5a87b 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -286,6 +286,13 @@ F: include/hw/riscv/
 F: linux-user/host/riscv32/
 F: linux-user/host/riscv64/
 
+RISC-V XVentanaCondOps extension
+M: Philipp Tomsich <philipp.tomsich@vrull.eu>
+L: qemu-riscv@nongnu.org
+S: Supported
+F: target/riscv/XVentanaCondOps.decode
+F: target/riscv/insn_trans/trans_xventanacondops.c.inc
+
 RENESAS RX CPUs
 R: Yoshinori Sato <ysato@users.sourceforge.jp>
 S: Orphan
-- 
2.33.1

Re: [PATCH v3 7/7] target/riscv: add a MAINTAINERS entry for XVentanaCondOps

[Richard Henderson]

On 1/29/22 01:56, Philipp Tomsich wrote:
> The XVentanaCondOps extension is supported by VRULL on behalf of the
> Ventana Micro.  Add myself as a point-of-contact.
> 
> Signed-off-by: Philipp Tomsich<philipp.tomsich@vrull.eu>
> 
> ---

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

r~