Skip to contentSkip to conversation

s390x/tcg: Don't model FP registers as globals

PATCH Superseded David Hildenbrand 2 participants 3 messages Jan 31, 2019 – Feb 4, 2019

v1 of 2 Superseded Latest is v2 →

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 s390x/tcg: Don't model FP registers as globals

As floating point registers overlay some vector registers and we want to make use of the general tcg_gvec infrastructure that assumes vectors are not stored in globals but in memory, don't model flaoting point registers as globals anymore. This is than similar to how arm handles it.

Reading/writing a floating point register means reading/writing memory now.

Break up ugly in2_x2() handling that modifies both, in1 and in2 into in1_x2l and in2_x2h. This makes things more readable. Also, in1_x1() is ugly as it touches out/out2, get rid of that and use prep_x1() instead.

As we are no longer able to use the original global variables for out/out2, we have to use new temporary variables and write from them to the target registers using wout_ helpers.

E.g. an instruction that reads and writes x1 will use
- prep_x1 to get the values into out/out2
- wout_x1 to write the values from out/out2
This special handling is needed for x1 as it is often used along with other inputs, so in1/in2 is already used.

---
 target/s390x/insn-data.def | 150 ++++++++++++++++++-------------------
 target/s390x/translate.c   | 133 +++++++++++++++++++-------------
 2 files changed, 158 insertions(+), 125 deletions(-)
target/s390x/insn-data.def +75 -75
index 54e39df831..e275910dfa 100644
@@ -33,10 +33,10 @@
     C(0xe308, AG,      RXY_a, Z,   r1, m2_64, r1, 0, add, adds64)
     C(0xe318, AGF,     RXY_a, Z,   r1, m2_32s, r1, 0, add, adds64)
     F(0xb30a, AEBR,    RRE,   Z,   e1, e2, new, e1, aeb, f32, IF_BFP)
-    F(0xb31a, ADBR,    RRE,   Z,   f1_o, f2_o, f1, 0, adb, f64, IF_BFP)
-    F(0xb34a, AXBR,    RRE,   Z,   0, x2_o, x1, 0, axb, f128, IF_BFP)
+    F(0xb31a, ADBR,    RRE,   Z,   f1, f2, new, f1, adb, f64, IF_BFP)
+    F(0xb34a, AXBR,    RRE,   Z,   x2l, x2h, x1, x1, axb, f128, IF_BFP)
     F(0xed0a, AEB,     RXE,   Z,   e1, m2_32u, new, e1, aeb, f32, IF_BFP)
-    F(0xed1a, ADB,     RXE,   Z,   f1_o, m2_64, f1, 0, adb, f64, IF_BFP)
+    F(0xed1a, ADB,     RXE,   Z,   f1, m2_64, new, f1, adb, f64, IF_BFP)
 /* ADD HIGH */
     C(0xb9c8, AHHHR,   RRF_a, HW,  r2_sr32, r3_sr32, new, r1_32h, add, adds32)
     C(0xb9d8, AHHLR,   RRF_a, HW,  r2_sr32, r3, new, r1_32h, add, adds32)
@@ -154,7 +154,7 @@
     C(0xb241, CKSM,    RRE,   Z,   r1_o, ra2, new, r1_32, cksm, 0)
 
 /* COPY SIGN */
-    F(0xb372, CPSDR,   RRF_b, FPSSH, f3_o, f2_o, f1, 0, cps, 0, IF_AFP1 | IF_AFP2 | IF_AFP3)
+    F(0xb372, CPSDR,   RRF_b, FPSSH, f3, f2, new, f1, cps, 0, IF_AFP1 | IF_AFP2 | IF_AFP3)
 
 /* COMPARE */
     C(0x1900, CR,      RR_a,  Z,   r1_o, r2_o, 0, 0, 0, cmps32)
@@ -165,16 +165,16 @@
     C(0xe320, CG,      RXY_a, Z,   r1_o, m2_64, 0, 0, 0, cmps64)
     C(0xe330, CGF,     RXY_a, Z,   r1_o, m2_32s, 0, 0, 0, cmps64)
     F(0xb309, CEBR,    RRE,   Z,   e1, e2, 0, 0, ceb, 0, IF_BFP)
-    F(0xb319, CDBR,    RRE,   Z,   f1_o, f2_o, 0, 0, cdb, 0, IF_BFP)
-    F(0xb349, CXBR,    RRE,   Z,   x1_o, x2_o, 0, 0, cxb, 0, IF_BFP)
+    F(0xb319, CDBR,    RRE,   Z,   f1, f2, 0, 0, cdb, 0, IF_BFP)
+    F(0xb349, CXBR,    RRE,   Z,   x2l, x2h, x1, 0, cxb, 0, IF_BFP)
     F(0xed09, CEB,     RXE,   Z,   e1, m2_32u, 0, 0, ceb, 0, IF_BFP)
-    F(0xed19, CDB,     RXE,   Z,   f1_o, m2_64, 0, 0, cdb, 0, IF_BFP)
+    F(0xed19, CDB,     RXE,   Z,   f1, m2_64, 0, 0, cdb, 0, IF_BFP)
 /* COMPARE AND SIGNAL */
     F(0xb308, KEBR,    RRE,   Z,   e1, e2, 0, 0, keb, 0, IF_BFP)
-    F(0xb318, KDBR,    RRE,   Z,   f1_o, f2_o, 0, 0, kdb, 0, IF_BFP)
-    F(0xb348, KXBR,    RRE,   Z,   x1_o, x2_o, 0, 0, kxb, 0, IF_BFP)
+    F(0xb318, KDBR,    RRE,   Z,   f1, f2, 0, 0, kdb, 0, IF_BFP)
+    F(0xb348, KXBR,    RRE,   Z,   x2l, x2h, x1, 0, kxb, 0, IF_BFP)
     F(0xed08, KEB,     RXE,   Z,   e1, m2_32u, 0, 0, keb, 0, IF_BFP)
-    F(0xed18, KDB,     RXE,   Z,   f1_o, m2_64, 0, 0, kdb, 0, IF_BFP)
+    F(0xed18, KDB,     RXE,   Z,   f1, m2_64, 0, 0, kdb, 0, IF_BFP)
 /* COMPARE IMMEDIATE */
     C(0xc20d, CFI,     RIL_a, EI,  r1, i2, 0, 0, 0, cmps32)
     C(0xc20c, CGFI,    RIL_a, EI,  r1, i2, 0, 0, 0, cmps64)
@@ -292,32 +292,32 @@
     C(0xe326, CVDY,    RXY_a, LD,  r1_o, a2, 0, 0, cvd, 0)
 /* CONVERT TO FIXED */
     F(0xb398, CFEBR,   RRF_e, Z,   0, e2, new, r1_32, cfeb, 0, IF_BFP)
-    F(0xb399, CFDBR,   RRF_e, Z,   0, f2_o, new, r1_32, cfdb, 0, IF_BFP)
-    F(0xb39a, CFXBR,   RRF_e, Z,   0, x2_o, new, r1_32, cfxb, 0, IF_BFP)
+    F(0xb399, CFDBR,   RRF_e, Z,   0, f2, new, r1_32, cfdb, 0, IF_BFP)
+    F(0xb39a, CFXBR,   RRF_e, Z,   x2l, x2h, new, r1_32, cfxb, 0, IF_BFP)
     F(0xb3a8, CGEBR,   RRF_e, Z,   0, e2, r1, 0, cgeb, 0, IF_BFP)
-    F(0xb3a9, CGDBR,   RRF_e, Z,   0, f2_o, r1, 0, cgdb, 0, IF_BFP)
-    F(0xb3aa, CGXBR,   RRF_e, Z,   0, x2_o, r1, 0, cgxb, 0, IF_BFP)
+    F(0xb3a9, CGDBR,   RRF_e, Z,   0, f2, r1, 0, cgdb, 0, IF_BFP)
+    F(0xb3aa, CGXBR,   RRF_e, Z,   x2l, x2h, r1, 0, cgxb, 0, IF_BFP)
 /* CONVERT FROM FIXED */
     F(0xb394, CEFBR,   RRF_e, Z,   0, r2_32s, new, e1, cegb, 0, IF_BFP)
-    F(0xb395, CDFBR,   RRF_e, Z,   0, r2_32s, f1, 0, cdgb, 0, IF_BFP)
-    F(0xb396, CXFBR,   RRF_e, Z,   0, r2_32s, x1, 0, cxgb, 0, IF_BFP)
+    F(0xb395, CDFBR,   RRF_e, Z,   0, r2_32s, new, f1, cdgb, 0, IF_BFP)
+    F(0xb396, CXFBR,   RRF_e, Z,   0, r2_32s, new_P, x1, cxgb, 0, IF_BFP)
     F(0xb3a4, CEGBR,   RRF_e, Z,   0, r2_o, new, e1, cegb, 0, IF_BFP)
-    F(0xb3a5, CDGBR,   RRF_e, Z,   0, r2_o, f1, 0, cdgb, 0, IF_BFP)
-    F(0xb3a6, CXGBR,   RRF_e, Z,   0, r2_o, x1, 0, cxgb, 0, IF_BFP)
+    F(0xb3a5, CDGBR,   RRF_e, Z,   0, r2_o, new, f1, cdgb, 0, IF_BFP)
+    F(0xb3a6, CXGBR,   RRF_e, Z,   0, r2_o, new_P, x1, cxgb, 0, IF_BFP)
 /* CONVERT TO LOGICAL */
     F(0xb39c, CLFEBR,  RRF_e, FPE, 0, e2, new, r1_32, clfeb, 0, IF_BFP)
-    F(0xb39d, CLFDBR,  RRF_e, FPE, 0, f2_o, new, r1_32, clfdb, 0, IF_BFP)
-    F(0xb39e, CLFXBR,  RRF_e, FPE, 0, x2_o, new, r1_32, clfxb, 0, IF_BFP)
+    F(0xb39d, CLFDBR,  RRF_e, FPE, 0, f2, new, r1_32, clfdb, 0, IF_BFP)
+    F(0xb39e, CLFXBR,  RRF_e, FPE, x2l, x2h, new, r1_32, clfxb, 0, IF_BFP)
     F(0xb3ac, CLGEBR,  RRF_e, FPE, 0, e2, r1, 0, clgeb, 0, IF_BFP)
-    F(0xb3ad, CLGDBR,  RRF_e, FPE, 0, f2_o, r1, 0, clgdb, 0, IF_BFP)
-    F(0xb3ae, CLGXBR,  RRF_e, FPE, 0, x2_o, r1, 0, clgxb, 0, IF_BFP)
+    F(0xb3ad, CLGDBR,  RRF_e, FPE, 0, f2, r1, 0, clgdb, 0, IF_BFP)
+    F(0xb3ae, CLGXBR,  RRF_e, FPE, x2l, x2h, r1, 0, clgxb, 0, IF_BFP)
 /* CONVERT FROM LOGICAL */
     F(0xb390, CELFBR,  RRF_e, FPE, 0, r2_32u, new, e1, celgb, 0, IF_BFP)
-    F(0xb391, CDLFBR,  RRF_e, FPE, 0, r2_32u, f1, 0, cdlgb, 0, IF_BFP)
-    F(0xb392, CXLFBR,  RRF_e, FPE, 0, r2_32u, x1, 0, cxlgb, 0, IF_BFP)
+    F(0xb391, CDLFBR,  RRF_e, FPE, 0, r2_32u, new, f1, cdlgb, 0, IF_BFP)
+    F(0xb392, CXLFBR,  RRF_e, FPE, 0, r2_32u, new_P, x1, cxlgb, 0, IF_BFP)
     F(0xb3a0, CELGBR,  RRF_e, FPE, 0, r2_o, new, e1, celgb, 0, IF_BFP)
-    F(0xb3a1, CDLGBR,  RRF_e, FPE, 0, r2_o, f1, 0, cdlgb, 0, IF_BFP)
-    F(0xb3a2, CXLGBR,  RRF_e, FPE, 0, r2_o, x1, 0, cxlgb, 0, IF_BFP)
+    F(0xb3a1, CDLGBR,  RRF_e, FPE, 0, r2_o, new, f1, cdlgb, 0, IF_BFP)
+    F(0xb3a2, CXLGBR,  RRF_e, FPE, 0, r2_o, new_P, x1, cxlgb, 0, IF_BFP)
 
 /* CONVERT UTF-8 TO UTF-16 */
     D(0xb2a7, CU12,    RRF_c, Z,   0, 0, 0, 0, cuXX, 0, 12)
@@ -336,10 +336,10 @@
     C(0x1d00, DR,      RR_a,  Z,   r1_D32, r2_32s, new_P, r1_P32, divs32, 0)
     C(0x5d00, D,       RX_a,  Z,   r1_D32, m2_32s, new_P, r1_P32, divs32, 0)
     F(0xb30d, DEBR,    RRE,   Z,   e1, e2, new, e1, deb, 0, IF_BFP)
-    F(0xb31d, DDBR,    RRE,   Z,   f1_o, f2_o, f1, 0, ddb, 0, IF_BFP)
-    F(0xb34d, DXBR,    RRE,   Z,   0, x2_o, x1, 0, dxb, 0, IF_BFP)
+    F(0xb31d, DDBR,    RRE,   Z,   f1, f2, new, f1, ddb, 0, IF_BFP)
+    F(0xb34d, DXBR,    RRE,   Z,   x2l, x2h, x1, x1, dxb, 0, IF_BFP)
     F(0xed0d, DEB,     RXE,   Z,   e1, m2_32u, new, e1, deb, 0, IF_BFP)
-    F(0xed1d, DDB,     RXE,   Z,   f1_o, m2_64, f1, 0, ddb, 0, IF_BFP)
+    F(0xed1d, DDB,     RXE,   Z,   f1, m2_64, new, f1, ddb, 0, IF_BFP)
 /* DIVIDE LOGICAL */
     C(0xb997, DLR,     RRE,   Z,   r1_D32, r2_32u, new_P, r1_P32, divu32, 0)
     C(0xe397, DL,      RXY_a, Z,   r1_D32, m2_32u, new_P, r1_P32, divu32, 0)
@@ -410,13 +410,13 @@
     C(0xb914, LGFR,    RRE,   Z,   0, r2_32s, 0, r1, mov2, 0)
     C(0xe304, LG,      RXY_a, Z,   0, a2, r1, 0, ld64, 0)
     C(0xe314, LGF,     RXY_a, Z,   0, a2, r1, 0, ld32s, 0)
-    F(0x2800, LDR,     RR_a,  Z,   0, f2_o, 0, f1, mov2, 0, IF_AFP1 | IF_AFP2)
+    F(0x2800, LDR,     RR_a,  Z,   0, f2, 0, f1, mov2, 0, IF_AFP1 | IF_AFP2)
     F(0x6800, LD,      RX_a,  Z,   0, m2_64, 0, f1, mov2, 0, IF_AFP1)
     F(0xed65, LDY,     RXY_a, LD,  0, m2_64, 0, f1, mov2, 0, IF_AFP1)
     F(0x3800, LER,     RR_a,  Z,   0, e2, 0, cond_e1e2, mov2, 0, IF_AFP1 | IF_AFP2)
     F(0x7800, LE,      RX_a,  Z,   0, m2_32u, 0, e1, mov2, 0, IF_AFP1)
     F(0xed64, LEY,     RXY_a, LD,  0, m2_32u, 0, e1, mov2, 0, IF_AFP1)
-    F(0xb365, LXR,     RRE,   Z,   0, x2_o, 0, x1, movx, 0, IF_AFP1)
+    F(0xb365, LXR,     RRE,   Z,   x2l, x2h, 0, x1, movx, 0, IF_AFP1)
 /* LOAD IMMEDIATE */
     C(0xc001, LGFI,    RIL_a, EI,  0, i2, 0, r1, mov2, 0)
 /* LOAD RELATIVE LONG */
@@ -454,8 +454,8 @@
     C(0xe302, LTG,     RXY_a, EI,  0, a2, r1, 0, ld64, s64)
     C(0xe332, LTGF,    RXY_a, GIE, 0, a2, r1, 0, ld32s, s64)
     F(0xb302, LTEBR,   RRE,   Z,   0, e2, 0, cond_e1e2, mov2, f32, IF_BFP)
-    F(0xb312, LTDBR,   RRE,   Z,   0, f2_o, 0, f1, mov2, f64, IF_BFP)
-    F(0xb342, LTXBR,   RRE,   Z,   0, x2_o, 0, x1, movx, f128, IF_BFP)
+    F(0xb312, LTDBR,   RRE,   Z,   0, f2, 0, f1, mov2, f64, IF_BFP)
+    F(0xb342, LTXBR,   RRE,   Z,   x2l, x2h, 0, x1, movx, f128, IF_BFP)
 /* LOAD AND TRAP */
     C(0xe39f, LAT,     RXY_a, LAT, 0, m2_32u, r1, 0, lat, 0)
     C(0xe385, LGAT,    RXY_a, LAT, 0, a2, r1, 0, lgat, 0)
@@ -476,9 +476,9 @@
     C(0xb903, LCGR,    RRE,   Z,   0, r2, r1, 0, neg, neg64)
     C(0xb913, LCGFR,   RRE,   Z,   0, r2_32s, r1, 0, neg, neg64)
     F(0xb303, LCEBR,   RRE,   Z,   0, e2, new, e1, negf32, f32, IF_BFP)
-    F(0xb313, LCDBR,   RRE,   Z,   0, f2_o, f1, 0, negf64, f64, IF_BFP)
-    F(0xb343, LCXBR,   RRE,   Z,   0, x2_o, x1, 0, negf128, f128, IF_BFP)
-    F(0xb373, LCDFR,   RRE,   FPSSH, 0, f2_o, f1, 0, negf64, 0, IF_AFP1 | IF_AFP2)
+    F(0xb313, LCDBR,   RRE,   Z,   0, f2, new, f1, negf64, f64, IF_BFP)
+    F(0xb343, LCXBR,   RRE,   Z,   x2l, x2h, new_P, x1, negf128, f128, IF_BFP)
+    F(0xb373, LCDFR,   RRE,   FPSSH, 0, f2, new, f1, negf64, 0, IF_AFP1 | IF_AFP2)
 /* LOAD HALFWORD */
     C(0xb927, LHR,     RRE,   EI,  0, r2_16s, 0, r1_32, mov2, 0)
     C(0xb907, LGHR,    RRE,   EI,  0, r2_16s, 0, r1, mov2, 0)
@@ -537,15 +537,15 @@
 /* LOAD FPR FROM GR */
     F(0xb3c1, LDGR,    RRE,   FPRGR, 0, r2_o, 0, f1, mov2, 0, IF_AFP1)
 /* LOAD GR FROM FPR */
-    F(0xb3cd, LGDR,    RRE,   FPRGR, 0, f2_o, 0, r1, mov2, 0, IF_AFP2)
+    F(0xb3cd, LGDR,    RRE,   FPRGR, 0, f2, 0, r1, mov2, 0, IF_AFP2)
 /* LOAD NEGATIVE */
     C(0x1100, LNR,     RR_a,  Z,   0, r2_32s, new, r1_32, nabs, nabs32)
     C(0xb901, LNGR,    RRE,   Z,   0, r2, r1, 0, nabs, nabs64)
     C(0xb911, LNGFR,   RRE,   Z,   0, r2_32s, r1, 0, nabs, nabs64)
     F(0xb301, LNEBR,   RRE,   Z,   0, e2, new, e1, nabsf32, f32, IF_BFP)
-    F(0xb311, LNDBR,   RRE,   Z,   0, f2_o, f1, 0, nabsf64, f64, IF_BFP)
-    F(0xb341, LNXBR,   RRE,   Z,   0, x2_o, x1, 0, nabsf128, f128, IF_BFP)
-    F(0xb371, LNDFR,   RRE,   FPSSH, 0, f2_o, f1, 0, nabsf64, 0, IF_AFP1 | IF_AFP2)
+    F(0xb311, LNDBR,   RRE,   Z,   0, f2, new, f1, nabsf64, f64, IF_BFP)
+    F(0xb341, LNXBR,   RRE,   Z,   x2l, x2h, new_P, x1, nabsf128, f128, IF_BFP)
+    F(0xb371, LNDFR,   RRE,   FPSSH, 0, f2, new, f1, nabsf64, 0, IF_AFP1 | IF_AFP2)
 /* LOAD ON CONDITION */
     C(0xb9f2, LOCR,    RRF_c, LOC, r1, r2, new, r1_32, loc, 0)
     C(0xb9e2, LOCGR,   RRF_c, LOC, r1, r2, r1, 0, loc, 0)
@@ -568,9 +568,9 @@
     C(0xb900, LPGR,    RRE,   Z,   0, r2, r1, 0, abs, abs64)
     C(0xb910, LPGFR,   RRE,   Z,   0, r2_32s, r1, 0, abs, abs64)
     F(0xb300, LPEBR,   RRE,   Z,   0, e2, new, e1, absf32, f32, IF_BFP)
-    F(0xb310, LPDBR,   RRE,   Z,   0, f2_o, f1, 0, absf64, f64, IF_BFP)
-    F(0xb340, LPXBR,   RRE,   Z,   0, x2_o, x1, 0, absf128, f128, IF_BFP)
-    F(0xb370, LPDFR,   RRE,   FPSSH, 0, f2_o, f1, 0, absf64, 0, IF_AFP1 | IF_AFP2)
+    F(0xb310, LPDBR,   RRE,   Z,   0, f2, new, f1, absf64, f64, IF_BFP)
+    F(0xb340, LPXBR,   RRE,   Z,   x2l, x2h, new_P, x1, absf128, f128, IF_BFP)
+    F(0xb370, LPDFR,   RRE,   FPSSH, 0, f2, new, f1, absf64, 0, IF_AFP1 | IF_AFP2)
 /* LOAD REVERSED */
     C(0xb91f, LRVR,    RRE,   Z,   0, r2_32u, new, r1_32, rev32, 0)
     C(0xb90f, LRVGR,   RRE,   Z,   0, r2_o, r1, 0, rev64, 0)
@@ -588,20 +588,20 @@
     F(0xb2bd, LFAS,    S,     IEEEE_SIM, 0, m2_32u, 0, 0, sfas, 0, IF_DFP)
 /* LOAD FP INTEGER */
     F(0xb357, FIEBR,   RRF_e, Z,   0, e2, new, e1, fieb, 0, IF_BFP)
-    F(0xb35f, FIDBR,   RRF_e, Z,   0, f2_o, f1, 0, fidb, 0, IF_BFP)
-    F(0xb347, FIXBR,   RRF_e, Z,   0, x2_o, x1, 0, fixb, 0, IF_BFP)
+    F(0xb35f, FIDBR,   RRF_e, Z,   0, f2, new, f1, fidb, 0, IF_BFP)
+    F(0xb347, FIXBR,   RRF_e, Z,   x2l, x2h, new_P, x1, fixb, 0, IF_BFP)
 
 /* LOAD LENGTHENED */
-    F(0xb304, LDEBR,   RRE,   Z,   0, e2, f1, 0, ldeb, 0, IF_BFP)
-    F(0xb305, LXDBR,   RRE,   Z,   0, f2_o, x1, 0, lxdb, 0, IF_BFP)
-    F(0xb306, LXEBR,   RRE,   Z,   0, e2, x1, 0, lxeb, 0, IF_BFP)
-    F(0xed04, LDEB,    RXE,   Z,   0, m2_32u, f1, 0, ldeb, 0, IF_BFP)
-    F(0xed05, LXDB,    RXE,   Z,   0, m2_64, x1, 0, lxdb, 0, IF_BFP)
-    F(0xed06, LXEB,    RXE,   Z,   0, m2_32u, x1, 0, lxeb, 0, IF_BFP)
+    F(0xb304, LDEBR,   RRE,   Z,   0, e2, new, f1, ldeb, 0, IF_BFP)
+    F(0xb305, LXDBR,   RRE,   Z,   0, f2, new_P, x1, lxdb, 0, IF_BFP)
+    F(0xb306, LXEBR,   RRE,   Z,   0, e2, new_P, x1, lxeb, 0, IF_BFP)
+    F(0xed04, LDEB,    RXE,   Z,   0, m2_32u, new, f1, ldeb, 0, IF_BFP)
+    F(0xed05, LXDB,    RXE,   Z,   0, m2_64, new_P, x1, lxdb, 0, IF_BFP)
+    F(0xed06, LXEB,    RXE,   Z,   0, m2_32u, new_P, x1, lxeb, 0, IF_BFP)
 /* LOAD ROUNDED */
-    F(0xb344, LEDBR,   RRE,   Z,   0, f2_o, new, e1, ledb, 0, IF_BFP)
-    F(0xb345, LDXBR,   RRE,   Z,   0, x2_o, f1, 0, ldxb, 0, IF_BFP)
-    F(0xb346, LEXBR,   RRE,   Z,   0, x2_o, new, e1, lexb, 0, IF_BFP)
+    F(0xb344, LEDBR,   RRE,   Z,   0, f2, new, e1, ledb, 0, IF_BFP)
+    F(0xb345, LDXBR,   RRE,   Z,   x2l, x2h, new, f1, ldxb, 0, IF_BFP)
+    F(0xb346, LEXBR,   RRE,   Z,   x2l, x2h, new, e1, lexb, 0, IF_BFP)
 
 /* LOAD MULTIPLE */
     C(0x9800, LM,      RS_a,  Z,   0, a2, 0, 0, lm32, 0)
@@ -648,14 +648,14 @@
     C(0x5c00, M,       RX_a,  Z,   r1p1_32s, m2_32s, new, r1_D32, mul, 0)
     C(0xe35c, MFY,     RXY_a, GIE, r1p1_32s, m2_32s, new, r1_D32, mul, 0)
     F(0xb317, MEEBR,   RRE,   Z,   e1, e2, new, e1, meeb, 0, IF_BFP)
-    F(0xb31c, MDBR,    RRE,   Z,   f1_o, f2_o, f1, 0, mdb, 0, IF_BFP)
-    F(0xb34c, MXBR,    RRE,   Z,   0, x2_o, x1, 0, mxb, 0, IF_BFP)
-    F(0xb30c, MDEBR,   RRE,   Z,   f1_o, e2, f1, 0, mdeb, 0, IF_BFP)
-    F(0xb307, MXDBR,   RRE,   Z,   0, f2_o, x1, 0, mxdb, 0, IF_BFP)
+    F(0xb31c, MDBR,    RRE,   Z,   f1, f2, new, f1, mdb, 0, IF_BFP)
+    F(0xb34c, MXBR,    RRE,   Z,   x2l, x2h, x1, x1, mxb, 0, IF_BFP)
+    F(0xb30c, MDEBR,   RRE,   Z,   f1, e2, new, f1, mdeb, 0, IF_BFP)
+    F(0xb307, MXDBR,   RRE,   Z,   0, f2, x1, x1, mxdb, 0, IF_BFP)
     F(0xed17, MEEB,    RXE,   Z,   e1, m2_32u, new, e1, meeb, 0, IF_BFP)
-    F(0xed1c, MDB,     RXE,   Z,   f1_o, m2_64, f1, 0, mdb, 0, IF_BFP)
-    F(0xed0c, MDEB,    RXE,   Z,   f1_o, m2_32u, f1, 0, mdeb, 0, IF_BFP)
-    F(0xed07, MXDB,    RXE,   Z,   0, m2_64, x1, 0, mxdb, 0, IF_BFP)
+    F(0xed1c, MDB,     RXE,   Z,   f1, m2_64, new, f1, mdb, 0, IF_BFP)
+    F(0xed0c, MDEB,    RXE,   Z,   f1, m2_32u, new, f1, mdeb, 0, IF_BFP)
+    F(0xed07, MXDB,    RXE,   Z,   0, m2_64, x1, x1, mxdb, 0, IF_BFP)
 /* MULTIPLY HALFWORD */
     C(0x4c00, MH,      RX_a,  Z,   r1_o, m2_16s, new, r1_32, mul, 0)
     C(0xe37c, MHY,     RXY_a, GIE, r1_o, m2_16s, new, r1_32, mul, 0)
@@ -681,14 +681,14 @@
 
 /* MULTIPLY AND ADD */
     F(0xb30e, MAEBR,   RRD,   Z,   e1, e2, new, e1, maeb, 0, IF_BFP)
-    F(0xb31e, MADBR,   RRD,   Z,   f1_o, f2_o, f1, 0, madb, 0, IF_BFP)
+    F(0xb31e, MADBR,   RRD,   Z,   f1, f2, new, f1, madb, 0, IF_BFP)
     F(0xed0e, MAEB,    RXF,   Z,   e1, m2_32u, new, e1, maeb, 0, IF_BFP)
-    F(0xed1e, MADB,    RXF,   Z,   f1_o, m2_64, f1, 0, madb, 0, IF_BFP)
+    F(0xed1e, MADB,    RXF,   Z,   f1, m2_64, new, f1, madb, 0, IF_BFP)
 /* MULTIPLY AND SUBTRACT */
     F(0xb30f, MSEBR,   RRD,   Z,   e1, e2, new, e1, mseb, 0, IF_BFP)
-    F(0xb31f, MSDBR,   RRD,   Z,   f1_o, f2_o, f1, 0, msdb, 0, IF_BFP)
+    F(0xb31f, MSDBR,   RRD,   Z,   f1, f2, new, f1, msdb, 0, IF_BFP)
     F(0xed0f, MSEB,    RXF,   Z,   e1, m2_32u, new, e1, mseb, 0, IF_BFP)
-    F(0xed1f, MSDB,    RXF,   Z,   f1_o, m2_64, f1, 0, msdb, 0, IF_BFP)
+    F(0xed1f, MSDB,    RXF,   Z,   f1, m2_64, new, f1, msdb, 0, IF_BFP)
 
 /* OR */
     C(0x1600, OR,      RR_a,  Z,   r1, r2, new, r1_32, or, nz32)
@@ -793,17 +793,17 @@
 
 /* SQUARE ROOT */
     F(0xb314, SQEBR,   RRE,   Z,   0, e2, new, e1, sqeb, 0, IF_BFP)
-    F(0xb315, SQDBR,   RRE,   Z,   0, f2_o, f1, 0, sqdb, 0, IF_BFP)
-    F(0xb316, SQXBR,   RRE,   Z,   0, x2_o, x1, 0, sqxb, 0, IF_BFP)
+    F(0xb315, SQDBR,   RRE,   Z,   0, f2, new, f1, sqdb, 0, IF_BFP)
+    F(0xb316, SQXBR,   RRE,   Z,   x2l, x2h, new, x1, sqxb, 0, IF_BFP)
     F(0xed14, SQEB,    RXE,   Z,   0, m2_32u, new, e1, sqeb, 0, IF_BFP)
-    F(0xed15, SQDB,    RXE,   Z,   0, m2_64, f1, 0, sqdb, 0, IF_BFP)
+    F(0xed15, SQDB,    RXE,   Z,   0, m2_64, new, f1, sqdb, 0, IF_BFP)
 
 /* STORE */
     C(0x5000, ST,      RX_a,  Z,   r1_o, a2, 0, 0, st32, 0)
     C(0xe350, STY,     RXY_a, LD,  r1_o, a2, 0, 0, st32, 0)
     C(0xe324, STG,     RXY_a, Z,   r1_o, a2, 0, 0, st64, 0)
-    F(0x6000, STD,     RX_a,  Z,   f1_o, a2, 0, 0, st64, 0, IF_AFP1)
-    F(0xed67, STDY,    RXY_a, LD,  f1_o, a2, 0, 0, st64, 0, IF_AFP1)
+    F(0x6000, STD,     RX_a,  Z,   f1, a2, 0, 0, st64, 0, IF_AFP1)
+    F(0xed67, STDY,    RXY_a, LD,  f1, a2, 0, 0, st64, 0, IF_AFP1)
     F(0x7000, STE,     RX_a,  Z,   e1, a2, 0, 0, st32, 0, IF_AFP1)
     F(0xed66, STEY,    RXY_a, LD,  e1, a2, 0, 0, st32, 0, IF_AFP1)
 /* STORE RELATIVE LONG */
@@ -865,10 +865,10 @@
     C(0xe309, SG,      RXY_a, Z,   r1, m2_64, r1, 0, sub, subs64)
     C(0xe319, SGF,     RXY_a, Z,   r1, m2_32s, r1, 0, sub, subs64)
     F(0xb30b, SEBR,    RRE,   Z,   e1, e2, new, e1, seb, f32, IF_BFP)
-    F(0xb31b, SDBR,    RRE,   Z,   f1_o, f2_o, f1, 0, sdb, f64, IF_BFP)
-    F(0xb34b, SXBR,    RRE,   Z,   0, x2_o, x1, 0, sxb, f128, IF_BFP)
+    F(0xb31b, SDBR,    RRE,   Z,   f1, f2, new, f1, sdb, f64, IF_BFP)
+    F(0xb34b, SXBR,    RRE,   Z,   x2l, x2h, x1, x1, sxb, f128, IF_BFP)
     F(0xed0b, SEB,     RXE,   Z,   e1, m2_32u, new, e1, seb, f32, IF_BFP)
-    F(0xed1b, SDB,     RXE,   Z,   f1_o, m2_64, f1, 0, sdb, f64, IF_BFP)
+    F(0xed1b, SDB,     RXE,   Z,   f1, m2_64, new, f1, sdb, f64, IF_BFP)
 /* SUBTRACT HALFWORD */
     C(0x4b00, SH,      RX_a,  Z,   r1, m2_16s, new, r1_32, sub, subs32)
     C(0xe37b, SHY,     RXY_a, LD,  r1, m2_16s, new, r1_32, sub, subs32)
@@ -908,8 +908,8 @@
 
 /* TEST DATA CLASS */
     F(0xed10, TCEB,    RXE,   Z,   e1, a2, 0, 0, tceb, 0, IF_BFP)
-    F(0xed11, TCDB,    RXE,   Z,   f1_o, a2, 0, 0, tcdb, 0, IF_BFP)
-    F(0xed12, TCXB,    RXE,   Z,   x1_o, a2, 0, 0, tcxb, 0, IF_BFP)
+    F(0xed11, TCDB,    RXE,   Z,   f1, a2, 0, 0, tcdb, 0, IF_BFP)
+    F(0xed12, TCXB,    RXE,   Z,   0, a2, x1, 0, tcxb, 0, IF_BFP)
 
 /* TEST DECIMAL */
     C(0xebc0, TP,      RSL,   E2,  la1, 0, 0, 0, tp, 0)
target/s390x/translate.c +83 -50

213 lines not on this page · raw message

Signature · 1 line
2.17.2
Reviewed-by on v1 1/1
#1 lore raw
Quoted 22 lines · David Hildenbrand

As floating point registers overlay some vector registers and we want to make use of the general tcg_gvec infrastructure that assumes vectors are not stored in globals but in memory, don't model flaoting point registers as globals anymore. This is than similar to how arm handles it.

Reading/writing a floating point register means reading/writing memory now.

Break up ugly in2_x2() handling that modifies both, in1 and in2 into in1_x2l and in2_x2h. This makes things more readable. Also, in1_x1() is ugly as it touches out/out2, get rid of that and use prep_x1() instead.

As we are no longer able to use the original global variables for out/out2, we have to use new temporary variables and write from them to the target registers using wout_ helpers.

E.g. an instruction that reads and writes x1 will use
- prep_x1 to get the values into out/out2
- wout_x1 to write the values from out/out2
This special handling is needed for x1 as it is often used along with other inputs, so in1/in2 is already used.

Quoted 5 lines
---
 target/s390x/insn-data.def | 150 ++++++++++++++++++-------------------
 target/s390x/translate.c   | 133 +++++++++++++++++++-------------
 2 files changed, 158 insertions(+), 125 deletions(-)

r~

#2 lore raw
Quoted 23 lines · Richard Henderson
Quoted 22 lines · David Hildenbrand

As floating point registers overlay some vector registers and we want to make use of the general tcg_gvec infrastructure that assumes vectors are not stored in globals but in memory, don't model flaoting point registers as globals anymore. This is than similar to how arm handles it.

Reading/writing a floating point register means reading/writing memory now.

Break up ugly in2_x2() handling that modifies both, in1 and in2 into in1_x2l and in2_x2h. This makes things more readable. Also, in1_x1() is ugly as it touches out/out2, get rid of that and use prep_x1() instead.

As we are no longer able to use the original global variables for out/out2, we have to use new temporary variables and write from them to the target registers using wout_ helpers.

E.g. an instruction that reads and writes x1 will use
- prep_x1 to get the values into out/out2
- wout_x1 to write the values from out/out2
This special handling is needed for x1 as it is often used along with other inputs, so in1/in2 is already used.

Quoted 7 lines
Quoted 5 lines
---
 target/s390x/insn-data.def | 150 ++++++++++++++++++-------------------
 target/s390x/translate.c   | 133 +++++++++++++++++++-------------
 2 files changed, 158 insertions(+), 125 deletions(-)

Thanks!

I'll probably switch naming of in1_x2l and in2_x2h to

in1_x2h and in2_x2l (will double check, but I guess this is the right thing to do - high double word always comes first, e.g. in memory).

Quoted 4 lines

r~

Signature · 3 lines
Thanks,

David / dhildenb