Hexagon (target/hexagon) circular addressing

The following instructions are added
    L2_loadrub_pci          Rd32 = memub(Rx32++#s4:0:circ(Mu2))
    L2_loadrb_pci           Rd32 = memb(Rx32++#s4:0:circ(Mu2))
    L2_loadruh_pci          Rd32 = memuh(Rx32++#s4:1:circ(Mu2))
    L2_loadrh_pci           Rd32 = memh(Rx32++#s4:1:circ(Mu2))
    L2_loadri_pci           Rd32 = memw(Rx32++#s4:2:circ(Mu2))
    L2_loadrd_pci           Rdd32 = memd(Rx32++#s4:3:circ(Mu2))
    S2_storerb_pci          memb(Rx32++#s4:0:circ(Mu2)) = Rt32
    S2_storerh_pci          memh(Rx32++#s4:1:circ(Mu2)) = Rt32
    S2_storerf_pci          memh(Rx32++#s4:1:circ(Mu2)) = Rt.H32
    S2_storeri_pci          memw(Rx32++#s4:2:circ(Mu2)) = Rt32
    S2_storerd_pci          memd(Rx32++#s4:3:circ(Mu2)) = Rtt32
    S2_storerbnew_pci       memb(Rx32++#s4:0:circ(Mu2)) = Nt8.new
    S2_storerhnew_pci       memw(Rx32++#s4:1:circ(Mu2)) = Nt8.new
    S2_storerinew_pci       memw(Rx32++#s4:2:circ(Mu2)) = Nt8.new
    L2_loadrub_pcr          Rd32 = memub(Rx32++I:circ(Mu2))
    L2_loadrb_pcr           Rd32 = memb(Rx32++I:circ(Mu2))
    L2_loadruh_pcr          Rd32 = memuh(Rx32++I:circ(Mu2))
    L2_loadrh_pcr           Rd32 = memh(Rx32++I:circ(Mu2))
    L2_loadri_pcr           Rd32 = memw(Rx32++I:circ(Mu2))
    L2_loadrd_pcr           Rdd32 = memd(Rx32++I:circ(Mu2))
    S2_storerb_pcr          memb(Rx32++I:circ(Mu2)) = Rt32
    S2_storerh_pcr          memh(Rx32++I:circ(Mu2)) = Rt32
    S2_storerf_pcr          memh(Rx32++I:circ(Mu2)) = Rt32.H32
    S2_storeri_pcr          memw(Rx32++I:circ(Mu2)) = Rt32
    S2_storerd_pcr          memd(Rx32++I:circ(Mu2)) = Rtt32
    S2_storerbnew_pcr       memb(Rx32++I:circ(Mu2)) = Nt8.new
    S2_storerhnew_pcr       memh(Rx32++I:circ(Mu2)) = Nt8.new
    S2_storerinew_pcr       memw(Rx32++I:circ(Mu2)) = Nt8.new

Test cases in tests/tcg/hexagon/circ.c

Signed-off-by: Taylor Simpson <tsimpson@quicinc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <1617930474-31979-23-git-send-email-tsimpson@quicinc.com>
[rth: Squash <1619667142-29636-1-git-send-email-tsimpson@quicinc.com>
removing gen_read_reg and gen_set_byte to avoid clang Werror.]
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

7 files changed, 346 insertions, 23 deletions

diff --git a/target/hexagon/gen_tcg.h b/target/hexagon/gen_tcg.h
index 6bc578dfda..25c228c112 100644
--- a/target/hexagon/gen_tcg.h
+++ b/target/hexagon/gen_tcg.h
@@ -38,6 +38,8 @@
  *     _ap       absolute set                      r0 = memw(r1=##variable)
  *     _pr       post increment register           r0 = memw(r1++m1)
  *     _pi       post increment immediate          r0 = memb(r1++#1)
+ *     _pci      post increment circular immediate r0 = memw(r1++#4:circ(m0))
+ *     _pcr      post increment circular register  r0 = memw(r1++I:circ(m0))
  */
 
 /* Macros for complex addressing modes */
@@ -56,7 +58,22 @@
         fEA_REG(RxV); \
         fPM_I(RxV, siV); \
     } while (0)
-
+#define GET_EA_pci \
+    do { \
+        TCGv tcgv_siV = tcg_const_tl(siV); \
+        tcg_gen_mov_tl(EA, RxV); \
+        gen_helper_fcircadd(RxV, RxV, tcgv_siV, MuV, \
+                            hex_gpr[HEX_REG_CS0 + MuN]); \
+        tcg_temp_free(tcgv_siV); \
+    } while (0)
+#define GET_EA_pcr(SHIFT) \
+    do { \
+        TCGv ireg = tcg_temp_new(); \
+        tcg_gen_mov_tl(EA, RxV); \
+        gen_read_ireg(ireg, MuV, (SHIFT)); \
+        gen_helper_fcircadd(RxV, RxV, ireg, MuV, hex_gpr[HEX_REG_CS0 + MuN]); \
+        tcg_temp_free(ireg); \
+    } while (0)
 
 /* Instructions with multiple definitions */
 #define fGEN_TCG_LOAD_AP(RES, SIZE, SIGN) \
@@ -80,6 +97,36 @@
 #define fGEN_TCG_L4_loadrd_ap(SHORTCODE) \
     fGEN_TCG_LOAD_AP(RddV, 8, u)
 
+#define fGEN_TCG_L2_loadrub_pci(SHORTCODE)    SHORTCODE
+#define fGEN_TCG_L2_loadrb_pci(SHORTCODE)     SHORTCODE
+#define fGEN_TCG_L2_loadruh_pci(SHORTCODE)    SHORTCODE
+#define fGEN_TCG_L2_loadrh_pci(SHORTCODE)     SHORTCODE
+#define fGEN_TCG_L2_loadri_pci(SHORTCODE)     SHORTCODE
+#define fGEN_TCG_L2_loadrd_pci(SHORTCODE)     SHORTCODE
+
+#define fGEN_TCG_LOAD_pcr(SHIFT, LOAD) \
+    do { \
+        TCGv ireg = tcg_temp_new(); \
+        tcg_gen_mov_tl(EA, RxV); \
+        gen_read_ireg(ireg, MuV, SHIFT); \
+        gen_helper_fcircadd(RxV, RxV, ireg, MuV, hex_gpr[HEX_REG_CS0 + MuN]); \
+        LOAD; \
+        tcg_temp_free(ireg); \
+    } while (0)
+
+#define fGEN_TCG_L2_loadrub_pcr(SHORTCODE) \
+      fGEN_TCG_LOAD_pcr(0, fLOAD(1, 1, u, EA, RdV))
+#define fGEN_TCG_L2_loadrb_pcr(SHORTCODE) \
+      fGEN_TCG_LOAD_pcr(0, fLOAD(1, 1, s, EA, RdV))
+#define fGEN_TCG_L2_loadruh_pcr(SHORTCODE) \
+      fGEN_TCG_LOAD_pcr(1, fLOAD(1, 2, u, EA, RdV))
+#define fGEN_TCG_L2_loadrh_pcr(SHORTCODE) \
+      fGEN_TCG_LOAD_pcr(1, fLOAD(1, 2, s, EA, RdV))
+#define fGEN_TCG_L2_loadri_pcr(SHORTCODE) \
+      fGEN_TCG_LOAD_pcr(2, fLOAD(1, 4, u, EA, RdV))
+#define fGEN_TCG_L2_loadrd_pcr(SHORTCODE) \
+      fGEN_TCG_LOAD_pcr(3, fLOAD(1, 8, u, EA, RddV))
+
 #define fGEN_TCG_L2_loadrub_pr(SHORTCODE)      SHORTCODE
 #define fGEN_TCG_L2_loadrub_pi(SHORTCODE)      SHORTCODE
 #define fGEN_TCG_L2_loadrb_pr(SHORTCODE)       SHORTCODE
@@ -195,6 +242,69 @@
 #define fGEN_TCG_S4_stored_locked(SHORTCODE) \
     do { SHORTCODE; READ_PREG(PdV, PdN); } while (0)
 
+#define fGEN_TCG_STORE(SHORTCODE) \
+    do { \
+        TCGv HALF = tcg_temp_new(); \
+        TCGv BYTE = tcg_temp_new(); \
+        SHORTCODE; \
+        tcg_temp_free(HALF); \
+        tcg_temp_free(BYTE); \
+    } while (0)
+
+#define fGEN_TCG_STORE_pcr(SHIFT, STORE) \
+    do { \
+        TCGv ireg = tcg_temp_new(); \
+        TCGv HALF = tcg_temp_new(); \
+        TCGv BYTE = tcg_temp_new(); \
+        tcg_gen_mov_tl(EA, RxV); \
+        gen_read_ireg(ireg, MuV, SHIFT); \
+        gen_helper_fcircadd(RxV, RxV, ireg, MuV, hex_gpr[HEX_REG_CS0 + MuN]); \
+        STORE; \
+        tcg_temp_free(ireg); \
+        tcg_temp_free(HALF); \
+        tcg_temp_free(BYTE); \
+    } while (0)
+
+#define fGEN_TCG_S2_storerb_pci(SHORTCODE) \
+    fGEN_TCG_STORE(SHORTCODE)
+#define fGEN_TCG_S2_storerb_pcr(SHORTCODE) \
+    fGEN_TCG_STORE_pcr(0, fSTORE(1, 1, EA, fGETBYTE(0, RtV)))
+
+#define fGEN_TCG_S2_storerh_pci(SHORTCODE) \
+    fGEN_TCG_STORE(SHORTCODE)
+#define fGEN_TCG_S2_storerh_pcr(SHORTCODE) \
+    fGEN_TCG_STORE_pcr(1, fSTORE(1, 2, EA, fGETHALF(0, RtV)))
+
+#define fGEN_TCG_S2_storerf_pci(SHORTCODE) \
+    fGEN_TCG_STORE(SHORTCODE)
+#define fGEN_TCG_S2_storerf_pcr(SHORTCODE) \
+    fGEN_TCG_STORE_pcr(1, fSTORE(1, 2, EA, fGETHALF(1, RtV)))
+
+#define fGEN_TCG_S2_storeri_pci(SHORTCODE) \
+    fGEN_TCG_STORE(SHORTCODE)
+#define fGEN_TCG_S2_storeri_pcr(SHORTCODE) \
+    fGEN_TCG_STORE_pcr(2, fSTORE(1, 4, EA, RtV))
+
+#define fGEN_TCG_S2_storerd_pci(SHORTCODE) \
+    fGEN_TCG_STORE(SHORTCODE)
+#define fGEN_TCG_S2_storerd_pcr(SHORTCODE) \
+    fGEN_TCG_STORE_pcr(3, fSTORE(1, 8, EA, RttV))
+
+#define fGEN_TCG_S2_storerbnew_pci(SHORTCODE) \
+    fGEN_TCG_STORE(SHORTCODE)
+#define fGEN_TCG_S2_storerbnew_pcr(SHORTCODE) \
+    fGEN_TCG_STORE_pcr(0, fSTORE(1, 1, EA, fGETBYTE(0, NtN)))
+
+#define fGEN_TCG_S2_storerhnew_pci(SHORTCODE) \
+    fGEN_TCG_STORE(SHORTCODE)
+#define fGEN_TCG_S2_storerhnew_pcr(SHORTCODE) \
+    fGEN_TCG_STORE_pcr(1, fSTORE(1, 2, EA, fGETHALF(0, NtN)))
+
+#define fGEN_TCG_S2_storerinew_pci(SHORTCODE) \
+    fGEN_TCG_STORE(SHORTCODE)
+#define fGEN_TCG_S2_storerinew_pcr(SHORTCODE) \
+    fGEN_TCG_STORE_pcr(2, fSTORE(1, 4, EA, NtN))
+
 /*
  * Mathematical operations with more than one definition require
  * special handling
diff --git a/target/hexagon/genptr.c b/target/hexagon/genptr.c
index 333f7d74bf..c6928d6184 100644
--- a/target/hexagon/genptr.c
+++ b/target/hexagon/genptr.c
@@ -266,6 +266,16 @@ static inline void gen_write_ctrl_reg_pair(DisasContext *ctx, int reg_num,
     }
 }
 
+static TCGv gen_get_byte(TCGv result, int N, TCGv src, bool sign)
+{
+    if (sign) {
+        tcg_gen_sextract_tl(result, src, N * 8, 8);
+    } else {
+        tcg_gen_extract_tl(result, src, N * 8, 8);
+    }
+    return result;
+}
+
 static TCGv gen_get_byte_i64(TCGv result, int N, TCGv_i64 src, bool sign)
 {
     TCGv_i64 res64 = tcg_temp_new_i64();
@@ -280,6 +290,16 @@ static TCGv gen_get_byte_i64(TCGv result, int N, TCGv_i64 src, bool sign)
     return result;
 }
 
+static inline TCGv gen_get_half(TCGv result, int N, TCGv src, bool sign)
+{
+    if (sign) {
+        tcg_gen_sextract_tl(result, src, N * 16, 16);
+    } else {
+        tcg_gen_extract_tl(result, src, N * 16, 16);
+    }
+    return result;
+}
+
 static void gen_set_byte_i64(int N, TCGv_i64 result, TCGv src)
 {
     TCGv_i64 src64 = tcg_temp_new_i64();
@@ -361,6 +381,75 @@ static inline void gen_store_conditional8(CPUHexagonState *env,
     tcg_gen_movi_tl(hex_llsc_addr, ~0);
 }
 
+static inline void gen_store32(TCGv vaddr, TCGv src, int width, int slot)
+{
+    tcg_gen_mov_tl(hex_store_addr[slot], vaddr);
+    tcg_gen_movi_tl(hex_store_width[slot], width);
+    tcg_gen_mov_tl(hex_store_val32[slot], src);
+}
+
+static inline void gen_store1(TCGv_env cpu_env, TCGv vaddr, TCGv src,
+                              DisasContext *ctx, int slot)
+{
+    gen_store32(vaddr, src, 1, slot);
+    ctx->store_width[slot] = 1;
+}
+
+static inline void gen_store1i(TCGv_env cpu_env, TCGv vaddr, int32_t src,
+                               DisasContext *ctx, int slot)
+{
+    TCGv tmp = tcg_const_tl(src);
+    gen_store1(cpu_env, vaddr, tmp, ctx, slot);
+    tcg_temp_free(tmp);
+}
+
+static inline void gen_store2(TCGv_env cpu_env, TCGv vaddr, TCGv src,
+                              DisasContext *ctx, int slot)
+{
+    gen_store32(vaddr, src, 2, slot);
+    ctx->store_width[slot] = 2;
+}
+
+static inline void gen_store2i(TCGv_env cpu_env, TCGv vaddr, int32_t src,
+                               DisasContext *ctx, int slot)
+{
+    TCGv tmp = tcg_const_tl(src);
+    gen_store2(cpu_env, vaddr, tmp, ctx, slot);
+    tcg_temp_free(tmp);
+}
+
+static inline void gen_store4(TCGv_env cpu_env, TCGv vaddr, TCGv src,
+                              DisasContext *ctx, int slot)
+{
+    gen_store32(vaddr, src, 4, slot);
+    ctx->store_width[slot] = 4;
+}
+
+static inline void gen_store4i(TCGv_env cpu_env, TCGv vaddr, int32_t src,
+                               DisasContext *ctx, int slot)
+{
+    TCGv tmp = tcg_const_tl(src);
+    gen_store4(cpu_env, vaddr, tmp, ctx, slot);
+    tcg_temp_free(tmp);
+}
+
+static inline void gen_store8(TCGv_env cpu_env, TCGv vaddr, TCGv_i64 src,
+                              DisasContext *ctx, int slot)
+{
+    tcg_gen_mov_tl(hex_store_addr[slot], vaddr);
+    tcg_gen_movi_tl(hex_store_width[slot], 8);
+    tcg_gen_mov_i64(hex_store_val64[slot], src);
+    ctx->store_width[slot] = 8;
+}
+
+static inline void gen_store8i(TCGv_env cpu_env, TCGv vaddr, int64_t src,
+                               DisasContext *ctx, int slot)
+{
+    TCGv_i64 tmp = tcg_const_i64(src);
+    gen_store8(cpu_env, vaddr, tmp, ctx, slot);
+    tcg_temp_free_i64(tmp);
+}
+
 static TCGv gen_8bitsof(TCGv result, TCGv value)
 {
     TCGv zero = tcg_const_tl(0);
diff --git a/target/hexagon/imported/encode_pp.def b/target/hexagon/imported/encode_pp.def
index 514c2404ce..68b435ebe7 100644
--- a/target/hexagon/imported/encode_pp.def
+++ b/target/hexagon/imported/encode_pp.def
@@ -294,6 +294,7 @@ DEF_CLASS32(ICLASS_LD" ---- -------- PP------ --------",LD)
 
 
 DEF_CLASS32(ICLASS_LD" 0--- -------- PP------ --------",LD_ADDR_ROFFSET)
+DEF_CLASS32(ICLASS_LD" 100- -------- PP----0- --------",LD_ADDR_POST_CIRC_IMMED)
 DEF_CLASS32(ICLASS_LD" 101- -------- PP00---- --------",LD_ADDR_POST_IMMED)
 DEF_CLASS32(ICLASS_LD" 101- -------- PP01---- --------",LD_ADDR_ABS_UPDATE_V4)
 DEF_CLASS32(ICLASS_LD" 101- -------- PP1----- --------",LD_ADDR_POST_IMMED_PRED_V2)
@@ -308,18 +309,23 @@ DEF_FIELD32(ICLASS_LD" ---- --!----- PP------ --------",LD_UN,"Unsigned")
 
 #define STD_LD_ENC(TAG,OPC) \
 DEF_ENC32(L2_load##TAG##_io,   ICLASS_LD" 0 ii "OPC"  sssss  PPiiiiii  iiiddddd")\
+DEF_ENC32(L2_load##TAG##_pci,  ICLASS_LD" 1 00 "OPC"  xxxxx  PPu0--0i  iiiddddd")\
 DEF_ENC32(L2_load##TAG##_pi,   ICLASS_LD" 1 01 "OPC"  xxxxx  PP00---i  iiiddddd")\
 DEF_ENC32(L4_load##TAG##_ap,   ICLASS_LD" 1 01 "OPC"  eeeee  PP01IIII  -IIddddd")\
 DEF_ENC32(L2_load##TAG##_pr,   ICLASS_LD" 1 10 "OPC"  xxxxx  PPu0----  0--ddddd")\
 DEF_ENC32(L4_load##TAG##_ur,   ICLASS_LD" 1 10 "OPC"  ttttt  PPi1IIII  iIIddddd")\
+DEF_ENC32(L2_load##TAG##_pcr,  ICLASS_LD" 1 00 "OPC"  xxxxx  PPu0--1-  0--ddddd")\
 
 
 #define STD_LDX_ENC(TAG,OPC) \
 DEF_ENC32(L2_load##TAG##_io,   ICLASS_LD" 0 ii "OPC"  sssss  PPiiiiii  iiiyyyyy")\
+DEF_ENC32(L2_load##TAG##_pci,  ICLASS_LD" 1 00 "OPC"  xxxxx  PPu0--0i  iiiyyyyy")\
 DEF_ENC32(L2_load##TAG##_pi,   ICLASS_LD" 1 01 "OPC"  xxxxx  PP00---i  iiiyyyyy")\
 DEF_ENC32(L4_load##TAG##_ap,   ICLASS_LD" 1 01 "OPC"  eeeee  PP01IIII  -IIyyyyy")\
 DEF_ENC32(L2_load##TAG##_pr,   ICLASS_LD" 1 10 "OPC"  xxxxx  PPu0----  0--yyyyy")\
 DEF_ENC32(L4_load##TAG##_ur,   ICLASS_LD" 1 10 "OPC"  ttttt  PPi1IIII  iIIyyyyy")\
+DEF_ENC32(L2_load##TAG##_pcr,  ICLASS_LD" 1 00 "OPC"  xxxxx  PPu0--1-  0--yyyyy")\
+DEF_ENC32(L2_load##TAG##_pbr,  ICLASS_LD" 1 11 "OPC"  xxxxx  PPu0----  0--yyyyy")
 
 
 #define STD_PLD_ENC(TAG,OPC) \
@@ -351,6 +357,7 @@ STD_PLD_ENC(rd,  "1 110") /* note dest reg field LSB=0, 1 is reserved */
 
 DEF_CLASS32(    ICLASS_LD" 0--0 000----- PP------ --------",LD_MISC)
 DEF_ANTICLASS32(ICLASS_LD" 0--0 000----- PP------ --------",LD_ADDR_ROFFSET)
+DEF_ANTICLASS32(ICLASS_LD" 1000 000----- PP------ --------",LD_ADDR_POST_CIRC_IMMED)
 DEF_ANTICLASS32(ICLASS_LD" 1010 000----- PP------ --------",LD_ADDR_POST_IMMED)
 DEF_ANTICLASS32(ICLASS_LD" 1100 000----- PP------ --------",LD_ADDR_POST_REG)
 DEF_ANTICLASS32(ICLASS_LD" 1110 000----- PP------ --------",LD_ADDR_POST_REG)
@@ -397,6 +404,7 @@ DEF_FIELD32(ICLASS_ST" ---! !!------ PP------ --------",ST_Type,"Type")
 DEF_FIELD32(ICLASS_ST" ---- --!----- PP------ --------",ST_UN,"Unsigned")
 
 DEF_CLASS32(ICLASS_ST" 0--1 -------- PP------ --------",ST_ADDR_ROFFSET)
+DEF_CLASS32(ICLASS_ST" 1001 -------- PP------ ------0-",ST_ADDR_POST_CIRC_IMMED)
 DEF_CLASS32(ICLASS_ST" 1011 -------- PP0----- 0-----0-",ST_ADDR_POST_IMMED)
 DEF_CLASS32(ICLASS_ST" 1011 -------- PP0----- 1-------",ST_ADDR_ABS_UPDATE_V4)
 DEF_CLASS32(ICLASS_ST" 1011 -------- PP1----- --------",ST_ADDR_POST_IMMED_PRED_V2)
@@ -411,10 +419,12 @@ DEF_CLASS32(ICLASS_ST" 0--0 0------- PP------ --------",ST_MISC_CACHEOP)
 
 #define STD_ST_ENC(TAG,OPC,SRC) \
 DEF_ENC32(S2_store##TAG##_io,   ICLASS_ST" 0 ii "OPC"  sssss  PPi"SRC"  iiiiiiii")\
+DEF_ENC32(S2_store##TAG##_pci,  ICLASS_ST" 1 00 "OPC"  xxxxx  PPu"SRC"  0iiii-0-")\
 DEF_ENC32(S2_store##TAG##_pi,   ICLASS_ST" 1 01 "OPC"  xxxxx  PP0"SRC"  0iiii-0-")\
 DEF_ENC32(S4_store##TAG##_ap,   ICLASS_ST" 1 01 "OPC"  eeeee  PP0"SRC"  1-IIIIII")\
 DEF_ENC32(S2_store##TAG##_pr,   ICLASS_ST" 1 10 "OPC"  xxxxx  PPu"SRC"  0-------")\
 DEF_ENC32(S4_store##TAG##_ur,   ICLASS_ST" 1 10 "OPC"  uuuuu  PPi"SRC"  1iIIIIII")\
+DEF_ENC32(S2_store##TAG##_pcr,  ICLASS_ST" 1 00 "OPC"  xxxxx  PPu"SRC"  0-----1-")\
 
 
 #define STD_PST_ENC(TAG,OPC,SRC) \
diff --git a/target/hexagon/imported/ldst.idef b/target/hexagon/imported/ldst.idef
index 78a2ea441c..6ce0635e32 100644
--- a/target/hexagon/imported/ldst.idef
+++ b/target/hexagon/imported/ldst.idef
@@ -26,6 +26,8 @@ Q6INSN(L4_##TAG##_ur,  OPER"(Rt32<<#u2+#U6)",             ATTRIB,DESCR,{fMUST_IM
 Q6INSN(L4_##TAG##_ap,  OPER"(Re32=#U6)",                  ATTRIB,DESCR,{fMUST_IMMEXT(UiV); fEA_IMM(UiV); SEMANTICS; ReV=UiV; })\
 Q6INSN(L2_##TAG##_pr,  OPER"(Rx32++Mu2)",                 ATTRIB,DESCR,{fEA_REG(RxV); fPM_M(RxV,MuV); SEMANTICS;})\
 Q6INSN(L2_##TAG##_pi,  OPER"(Rx32++#s4:"SHFT")",          ATTRIB,DESCR,{fEA_REG(RxV); fPM_I(RxV,siV); SEMANTICS;})\
+Q6INSN(L2_##TAG##_pci, OPER"(Rx32++#s4:"SHFT":circ(Mu2))",ATTRIB,DESCR,{fEA_REG(RxV); fPM_CIRI(RxV,siV,MuV); SEMANTICS;})\
+Q6INSN(L2_##TAG##_pcr, OPER"(Rx32++I:circ(Mu2))",  ATTRIB,DESCR,{fEA_REG(RxV); fPM_CIRR(RxV,fREAD_IREG(MuV)<<SCALE,MuV); SEMANTICS;})
 
 /* The set of 32-bit load instructions */
 STD_LD_AMODES(loadrub,"Rd32=memub","Load Unsigned Byte",ATTRIBS(A_LOAD),"0",fLOAD(1,1,u,EA,RdV),0)
@@ -42,6 +44,8 @@ Q6INSN(S2_##TAG##_pi,  OPER"(Rx32++#s4:"SHFT")="DEST,     ATTRIB,DESCR,{fEA_REG(
 Q6INSN(S4_##TAG##_ap,  OPER"(Re32=#U6)="DEST,             ATTRIB,DESCR,{fMUST_IMMEXT(UiV); fEA_IMM(UiV); SEMANTICS; ReV=UiV; })\
 Q6INSN(S2_##TAG##_pr,  OPER"(Rx32++Mu2)="DEST,            ATTRIB,DESCR,{fEA_REG(RxV); fPM_M(RxV,MuV); SEMANTICS; })\
 Q6INSN(S4_##TAG##_ur,  OPER"(Ru32<<#u2+#U6)="DEST,            ATTRIB,DESCR,{fMUST_IMMEXT(UiV); fEA_IRs(UiV,RuV,uiV); SEMANTICS;})\
+Q6INSN(S2_##TAG##_pci, OPER"(Rx32++#s4:"SHFT":circ(Mu2))="DEST,  ATTRIB,DESCR,{fEA_REG(RxV); fPM_CIRI(RxV,siV,MuV); SEMANTICS;})\
+Q6INSN(S2_##TAG##_pcr, OPER"(Rx32++I:circ(Mu2))="DEST,  ATTRIB,DESCR,{fEA_REG(RxV); fPM_CIRR(RxV,fREAD_IREG(MuV)<<SCALE,MuV); SEMANTICS;})
 
 
 /* The set of 32-bit store instructions */
diff --git a/target/hexagon/imported/macros.def b/target/hexagon/imported/macros.def
index 65292c7afa..25f57b6f19 100755
--- a/target/hexagon/imported/macros.def
+++ b/target/hexagon/imported/macros.def
@@ -277,6 +277,12 @@ DEF_MACRO(
 /*************************************/
 
 DEF_MACRO(
+    fREAD_IREG, /* read modifier register */
+    (fSXTN(11,64,(((VAL) & 0xf0000000)>>21) | ((VAL>>17)&0x7f) )),          /* behavior */
+    ()
+)
+
+DEF_MACRO(
     fREAD_LR, /* read link register */
     (READ_RREG(REG_LR)),          /* behavior */
     ()
@@ -307,6 +313,12 @@ DEF_MACRO(
 )
 
 DEF_MACRO(
+    fREAD_CSREG, /* read  CS register */
+    (READ_RREG(REG_CSA+N)),          /* behavior */
+    ()
+)
+
+DEF_MACRO(
     fREAD_LC0, /* read loop count */
     (READ_RREG(REG_LC0)),          /* behavior */
     ()
@@ -825,6 +837,20 @@ DEF_MACRO(
 )
 
 DEF_MACRO(
+    fPM_CIRI, /* Post Modify Register using Circular arithmetic by Immediate */
+    do { fcirc_add(REG,siV,MuV); } while (0),
+    ()
+)
+
+DEF_MACRO(
+    fPM_CIRR, /* Post Modify Register using Circular arithmetic by register */
+    do { fcirc_add(REG,VAL,MuV); } while (0),
+    ()
+)
+
+
+
+DEF_MACRO(
     fSCALE, /* scale by N */
     (((size8s_t)(A))<<N),
     /* optional attributes */
diff --git a/target/hexagon/macros.h b/target/hexagon/macros.h
index 8cb211dfde..494ea8dfd8 100644
--- a/target/hexagon/macros.h
+++ b/target/hexagon/macros.h
@@ -133,6 +133,38 @@
         CHECK_NOSHUF; \
         tcg_gen_qemu_ld64(DST, VA, ctx->mem_idx); \
     } while (0)
+
+#define MEM_STORE1_FUNC(X) \
+    __builtin_choose_expr(TYPE_INT(X), \
+        gen_store1i, \
+        __builtin_choose_expr(TYPE_TCGV(X), \
+            gen_store1, (void)0))
+#define MEM_STORE1(VA, DATA, SLOT) \
+    MEM_STORE1_FUNC(DATA)(cpu_env, VA, DATA, ctx, SLOT)
+
+#define MEM_STORE2_FUNC(X) \
+    __builtin_choose_expr(TYPE_INT(X), \
+        gen_store2i, \
+        __builtin_choose_expr(TYPE_TCGV(X), \
+            gen_store2, (void)0))
+#define MEM_STORE2(VA, DATA, SLOT) \
+    MEM_STORE2_FUNC(DATA)(cpu_env, VA, DATA, ctx, SLOT)
+
+#define MEM_STORE4_FUNC(X) \
+    __builtin_choose_expr(TYPE_INT(X), \
+        gen_store4i, \
+        __builtin_choose_expr(TYPE_TCGV(X), \
+            gen_store4, (void)0))
+#define MEM_STORE4(VA, DATA, SLOT) \
+    MEM_STORE4_FUNC(DATA)(cpu_env, VA, DATA, ctx, SLOT)
+
+#define MEM_STORE8_FUNC(X) \
+    __builtin_choose_expr(TYPE_INT(X), \
+        gen_store8i, \
+        __builtin_choose_expr(TYPE_TCGV_I64(X), \
+            gen_store8, (void)0))
+#define MEM_STORE8(VA, DATA, SLOT) \
+    MEM_STORE8_FUNC(DATA)(cpu_env, VA, DATA, ctx, SLOT)
 #else
 #define MEM_LOAD1s(VA) ((int8_t)mem_load1(env, slot, VA))
 #define MEM_LOAD1u(VA) ((uint8_t)mem_load1(env, slot, VA))
@@ -285,6 +317,39 @@ static inline void gen_logical_not(TCGv dest, TCGv src)
 
 #define fPCALIGN(IMM) IMM = (IMM & ~PCALIGN_MASK)
 
+#ifdef QEMU_GENERATE
+static inline TCGv gen_read_ireg(TCGv result, TCGv val, int shift)
+{
+    /*
+     * Section 2.2.4 of the Hexagon V67 Programmer's Reference Manual
+     *
+     *  The "I" value from a modifier register is divided into two pieces
+     *      LSB         bits 23:17
+     *      MSB         bits 31:28
+     * The value is signed
+     *
+     * At the end we shift the result according to the shift argument
+     */
+    TCGv msb = tcg_temp_new();
+    TCGv lsb = tcg_temp_new();
+
+    tcg_gen_extract_tl(lsb, val, 17, 7);
+    tcg_gen_sari_tl(msb, val, 21);
+    tcg_gen_deposit_tl(result, msb, lsb, 0, 7);
+
+    tcg_gen_shli_tl(result, result, shift);
+
+    tcg_temp_free(msb);
+    tcg_temp_free(lsb);
+
+    return result;
+}
+#define fREAD_IREG(VAL, SHIFT) gen_read_ireg(ireg, (VAL), (SHIFT))
+#else
+#define fREAD_IREG(VAL) \
+    (fSXTN(11, 64, (((VAL) & 0xf0000000) >> 21) | ((VAL >> 17) & 0x7f)))
+#endif
+
 #define fREAD_LR() (READ_REG(HEX_REG_LR))
 
 #define fWRITE_LR(A) WRITE_RREG(HEX_REG_LR, A)
@@ -418,6 +483,13 @@ static inline void gen_logical_not(TCGv dest, TCGv src)
 #define fEA_REG(REG) tcg_gen_mov_tl(EA, REG)
 #define fPM_I(REG, IMM)     tcg_gen_addi_tl(REG, REG, IMM)
 #define fPM_M(REG, MVAL)    tcg_gen_add_tl(REG, REG, MVAL)
+#define fPM_CIRI(REG, IMM, MVAL) \
+    do { \
+        TCGv tcgv_siV = tcg_const_tl(siV); \
+        gen_helper_fcircadd(REG, REG, tcgv_siV, MuV, \
+                            hex_gpr[HEX_REG_CS0 + MuN]); \
+        tcg_temp_free(tcgv_siV); \
+    } while (0)
 #else
 #define fEA_IMM(IMM)        do { EA = (IMM); } while (0)
 #define fEA_REG(REG)        do { EA = (REG); } while (0)
@@ -494,23 +566,43 @@ static inline void gen_logical_not(TCGv dest, TCGv src)
     gen_load_locked##SIZE##SIGN(DST, EA, ctx->mem_idx);
 #endif
 
+#ifdef QEMU_GENERATE
+#define fSTORE(NUM, SIZE, EA, SRC) MEM_STORE##SIZE(EA, SRC, insn->slot)
+#else
 #define fSTORE(NUM, SIZE, EA, SRC) MEM_STORE##SIZE(EA, SRC, slot)
+#endif
 
 #ifdef QEMU_GENERATE
 #define fSTORE_LOCKED(NUM, SIZE, EA, SRC, PRED) \
     gen_store_conditional##SIZE(env, ctx, PdN, PRED, EA, SRC);
 #endif
 
+#ifdef QEMU_GENERATE
+#define GETBYTE_FUNC(X) \
+    __builtin_choose_expr(TYPE_TCGV(X), \
+        gen_get_byte, \
+        __builtin_choose_expr(TYPE_TCGV_I64(X), \
+            gen_get_byte_i64, (void)0))
+#define fGETBYTE(N, SRC) GETBYTE_FUNC(SRC)(BYTE, N, SRC, true)
+#define fGETUBYTE(N, SRC) GETBYTE_FUNC(SRC)(BYTE, N, SRC, false)
+#else
 #define fGETBYTE(N, SRC) ((int8_t)((SRC >> ((N) * 8)) & 0xff))
 #define fGETUBYTE(N, SRC) ((uint8_t)((SRC >> ((N) * 8)) & 0xff))
+#endif
 
 #define fSETBYTE(N, DST, VAL) \
     do { \
         DST = (DST & ~(0x0ffLL << ((N) * 8))) | \
         (((uint64_t)((VAL) & 0x0ffLL)) << ((N) * 8)); \
     } while (0)
+
+#ifdef QEMU_GENERATE
+#define fGETHALF(N, SRC)  gen_get_half(HALF, N, SRC, true)
+#define fGETUHALF(N, SRC) gen_get_half(HALF, N, SRC, false)
+#else
 #define fGETHALF(N, SRC) ((int16_t)((SRC >> ((N) * 16)) & 0xffff))
 #define fGETUHALF(N, SRC) ((uint16_t)((SRC >> ((N) * 16)) & 0xffff))
+#endif
 #define fSETHALF(N, DST, VAL) \
     do { \
         DST = (DST & ~(0x0ffffLL << ((N) * 16))) | \
diff --git a/target/hexagon/op_helper.c b/target/hexagon/op_helper.c
index f9fb65555b..2319b9313e 100644
--- a/target/hexagon/op_helper.c
+++ b/target/hexagon/op_helper.c
@@ -251,33 +251,25 @@ void HELPER(debug_commit_end)(CPUHexagonState *env, int has_st0, int has_st1)
 
 }
 
-static int32_t fcircadd_v4(int32_t RxV, int32_t offset, int32_t M, int32_t CS)
-{
-    int32_t length = M & 0x0001ffff;
-    uint32_t new_ptr = RxV + offset;
-    uint32_t start_addr = CS;
-    uint32_t end_addr = start_addr + length;
-
-    if (new_ptr >= end_addr) {
-        new_ptr -= length;
-    } else if (new_ptr < start_addr) {
-        new_ptr += length;
-    }
-
-    return new_ptr;
-}
-
 int32_t HELPER(fcircadd)(int32_t RxV, int32_t offset, int32_t M, int32_t CS)
 {
-    int32_t K_const = (M >> 24) & 0xf;
-    int32_t length = M & 0x1ffff;
-    int32_t mask = (1 << (K_const + 2)) - 1;
+    int32_t K_const = sextract32(M, 24, 4);
+    int32_t length = sextract32(M, 0, 17);
     uint32_t new_ptr = RxV + offset;
-    uint32_t start_addr = RxV & (~mask);
-    uint32_t end_addr = start_addr | length;
+    uint32_t start_addr;
+    uint32_t end_addr;
 
     if (K_const == 0 && length >= 4) {
-        return fcircadd_v4(RxV, offset, M, CS);
+        start_addr = CS;
+        end_addr = start_addr + length;
+    } else {
+        /*
+         * Versions v3 and earlier used the K value to specify a power-of-2 size
+         * 2^(K+2) that is greater than the buffer length
+         */
+        int32_t mask = (1 << (K_const + 2)) - 1;
+        start_addr = RxV & (~mask);
+        end_addr = start_addr | length;
     }
 
     if (new_ptr >= end_addr) {