Merge remote-tracking branch 'remotes/amarkovic/tags/mips-queue-oct-2018-part-2-v2' into staging

MIPS queue for October 2018 - part 2 - v2

# gpg: Signature made Wed 24 Oct 2018 14:22:54 BST
# gpg:                using RSA key D4972A8967F75A65
# gpg: Good signature from "Aleksandar Markovic <amarkovic@wavecomp.com>"
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg:          There is no indication that the signature belongs to the owner.
# Primary key fingerprint: 8526 FBF1 5DA3 811F 4A01  DD75 D497 2A89 67F7 5A65

* remotes/amarkovic/tags/mips-queue-oct-2018-part-2-v2: (33 commits)
  target/mips: Fix decoding of ALIGN and DALIGN instructions
  target/mips: Fix the title of translate.c
  linux-user/mips: Recognize the R5900 CPU model
  target/mips: Define the R5900 CPU
  tests/tcg/mips: Add tests for R5900 DIVU1
  tests/tcg/mips: Add tests for R5900 DIV1
  tests/tcg/mips: Add tests for R5900 MTLO1 and MTHI1
  tests/tcg/mips: Add tests for R5900 MFLO1 and MFHI1
  tests/tcg/mips: Add tests for R5900 three-operand MULTU1
  tests/tcg/mips: Add tests for R5900 three-operand MULT1
  tests/tcg/mips: Add tests for R5900 three-operand MULTU
  tests/tcg/mips: Add tests for R5900 three-operand MULT
  target/mips: Make R5900 DMULT[U], DDIV[U], LL[D] and SC[D] user only
  target/mips: Support R5900 MOVN, MOVZ and PREF instructions from MIPS IV
  target/mips: Support R5900 DIV1 and DIVU1 instructions
  target/mips: Support R5900 MFLO1, MTLO1, MFHI1 and MTHI1 instructions
  target/mips: Support R5900 three-operand MULT1 and MULTU1 instructions
  target/mips: Support R5900 three-operand MULT and MULTU instructions
  target/mips: Add a placeholder for R5900 MMI3 instruction subclass
  target/mips: Add a placeholder for R5900 MMI2 instruction subclass
  ...
Signed-off-by: NPeter Maydell <peter.maydell@linaro.org>

linux-user/mips/target_elf.h

@@ -12,6 +12,9 @@ static inline const char *cpu_get_model(uint32_t eflags)
     if ((eflags & EF_MIPS_ARCH) == EF_MIPS_ARCH_32R6) {
         return "mips32r6-generic";
     }
+    if ((eflags & EF_MIPS_MACH) == EF_MIPS_MACH_5900) {
+        return "R5900";
+    }
     return "24Kf";
 }
 #endif

target/mips/mips-defs.h

@@ -64,9 +64,11 @@
 #define INSN_LOONGSON2E   0x0001000000000000ULL
 #define INSN_LOONGSON2F   0x0002000000000000ULL
 #define INSN_VR54XX       0x0004000000000000ULL
+#define INSN_R5900        0x0008000000000000ULL
 /*
  *   bits 56-63: vendor-specific ASEs
  */
+#define ASE_MMI           0x0100000000000000ULL
 
 /* MIPS CPU defines. */
 #define		CPU_MIPS1	(ISA_MIPS1)
@@ -74,6 +76,7 @@
 #define		CPU_MIPS3	(CPU_MIPS2 | ISA_MIPS3)
 #define		CPU_MIPS4	(CPU_MIPS3 | ISA_MIPS4)
 #define		CPU_VR54XX	(CPU_MIPS4 | INSN_VR54XX)
+#define         CPU_R5900       (CPU_MIPS3 | INSN_R5900)
 #define		CPU_LOONGSON2E  (CPU_MIPS3 | INSN_LOONGSON2E)
 #define		CPU_LOONGSON2F  (CPU_MIPS3 | INSN_LOONGSON2F)
 

target/mips/translate_init.inc.c

@@ -410,6 +410,65 @@ const mips_def_t mips_defs[] =
         .insn_flags = CPU_MIPS32R5 | ASE_MSA,
         .mmu_type = MMU_TYPE_R4000,
     },
+    {
+        /*
+         * The Toshiba TX System RISC TX79 Core Architecture manual
+         *
+         * https://wiki.qemu.org/File:C790.pdf
+         *
+         * describes the C790 processor that is a follow-up to the R5900.
+         * There are a few notable differences in that the R5900 FPU
+         *
+         * - is not IEEE 754-1985 compliant,
+         * - does not implement double format, and
+         * - its machine code is nonstandard.
+         */
+        .name = "R5900",
+        .CP0_PRid = 0x00002E00,
+        /* No L2 cache, icache size 32k, dcache size 32k, uncached coherency. */
+        .CP0_Config0 = (0x3 << 9) | (0x3 << 6) | (0x2 << CP0C0_K0),
+        .CP0_Status_rw_bitmask = 0xF4C79C1F,
+#ifdef CONFIG_USER_ONLY
+        /*
+         * R5900 hardware traps to the Linux kernel for IEEE 754-1985 and LL/SC
+         * emulation. For user only, QEMU is the kernel, so we emulate the traps
+         * by simply emulating the instructions directly.
+         *
+         * Note: Config1 is only used internally, the R5900 has only Config0.
+         */
+        .CP0_Config1 = (1 << CP0C1_FP) | (47 << CP0C1_MMU),
+        .CP0_LLAddr_rw_bitmask = 0xFFFFFFFF,
+        .CP0_LLAddr_shift = 4,
+        .CP1_fcr0 = (0x38 << FCR0_PRID) | (0x0 << FCR0_REV),
+        .CP1_fcr31 = 0,
+        .CP1_fcr31_rw_bitmask = 0x0183FFFF,
+#else
+        /*
+         * The R5900 COP1 FPU implements single-precision floating-point
+         * operations but is not entirely IEEE 754-1985 compatible. In
+         * particular,
+         *
+         * - NaN (not a number) and +/- infinities are not supported;
+         * - exception mechanisms are not fully supported;
+         * - denormalized numbers are not supported;
+         * - rounding towards nearest and +/- infinities are not supported;
+         * - computed results usually differs in the least significant bit;
+         * - saturations can differ more than the least significant bit.
+         *
+         * Since only rounding towards zero is supported, the two least
+         * significant bits of FCR31 are hardwired to 01.
+         *
+         * FPU emulation is disabled here until it is implemented.
+         *
+         * Note: Config1 is only used internally, the R5900 has only Config0.
+         */
+        .CP0_Config1 = (47 << CP0C1_MMU),
+#endif /* !CONFIG_USER_ONLY */
+        .SEGBITS = 32,
+        .PABITS = 32,
+        .insn_flags = CPU_R5900 | ASE_MMI,
+        .mmu_type = MMU_TYPE_R4000,
+    },
     {
         /* A generic CPU supporting MIPS32 Release 6 ISA.
            FIXME: Support IEEE 754-2008 FP.

tests/tcg/mips/mipsr5900/Makefile

+-include ../../config-host.mak
+
+CROSS=mipsr5900el-unknown-linux-gnu-
+
+SIM=qemu-mipsel
+SIM_FLAGS=-cpu R5900
+
+CC      = $(CROSS)gcc
+CFLAGS  = -Wall -mabi=32 -march=r5900 -static
+
+TESTCASES = div1.tst
+TESTCASES += divu1.tst
+TESTCASES += mflohi1.tst
+TESTCASES += mtlohi1.tst
+TESTCASES += mult.tst
+TESTCASES += multu.tst
+
+all: $(TESTCASES)
+
+%.tst: %.c
+	$(CC) $(CFLAGS) $< -o $@
+
+check: $(TESTCASES)
+	@for case in $(TESTCASES); do \
+        echo $(SIM) $(SIM_FLAGS) ./$$case;\
+        $(SIM) $(SIM_FLAGS) ./$$case; \
+	done
+
+clean:
+	$(RM) -rf $(TESTCASES)

tests/tcg/mips/mipsr5900/div1.c

+/*
+ * Test R5900-specific DIV1.
+ */
+
+#include <stdio.h>
+#include <inttypes.h>
+#include <assert.h>
+
+struct quotient_remainder { int32_t quotient, remainder; };
+
+static struct quotient_remainder div1(int32_t rs, int32_t rt)
+{
+    int32_t lo, hi;
+
+    __asm__ __volatile__ (
+            "    div1 $0, %2, %3\n"
+            "    mflo1 %0\n"
+            "    mfhi1 %1\n"
+            : "=r" (lo), "=r" (hi)
+            : "r" (rs), "r" (rt));
+
+    assert(rs / rt == lo);
+    assert(rs % rt == hi);
+
+    return (struct quotient_remainder) { .quotient = lo, .remainder = hi };
+}
+
+static void verify_div1(int32_t rs, int32_t rt,
+                        int32_t expected_quotient,
+                        int32_t expected_remainder)
+{
+    struct quotient_remainder qr = div1(rs, rt);
+
+    assert(qr.quotient == expected_quotient);
+    assert(qr.remainder == expected_remainder);
+}
+
+static void verify_div1_negations(int32_t rs, int32_t rt,
+                                  int32_t expected_quotient,
+                                  int32_t expected_remainder)
+{
+    verify_div1(rs, rt, expected_quotient, expected_remainder);
+    verify_div1(rs, -rt, -expected_quotient, expected_remainder);
+    verify_div1(-rs, rt, -expected_quotient, -expected_remainder);
+    verify_div1(-rs, -rt, expected_quotient, -expected_remainder);
+}
+
+int main()
+{
+    verify_div1_negations(0, 1, 0, 0);
+    verify_div1_negations(1, 1, 1, 0);
+    verify_div1_negations(1, 2, 0, 1);
+    verify_div1_negations(17, 19, 0, 17);
+    verify_div1_negations(19, 17, 1, 2);
+    verify_div1_negations(77773, 101, 770, 3);
+
+    verify_div1(-0x80000000,  1, -0x80000000, 0);
+
+    /*
+     * Supplementary explanation from the Toshiba TX System RISC TX79 Core
+     * Architecture manual, A-38 and B-7, https://wiki.qemu.org/File:C790.pdf
+     *
+     * Normally, when 0x80000000 (-2147483648) the signed minimum value is
+     * divided by 0xFFFFFFFF (-1), the operation will result in an overflow.
+     * However, in this instruction an overflow exception doesn't occur and
+     * the result will be as follows:
+     *
+     * Quotient is 0x80000000 (-2147483648), and remainder is 0x00000000 (0).
+     */
+    verify_div1(-0x80000000, -1, -0x80000000, 0);
+
+    return 0;
+}

tests/tcg/mips/mipsr5900/divu1.c

+/*
+ * Test R5900-specific DIVU1.
+ */
+
+#include <stdio.h>
+#include <inttypes.h>
+#include <assert.h>
+
+struct quotient_remainder { uint32_t quotient, remainder; };
+
+static struct quotient_remainder divu1(uint32_t rs, uint32_t rt)
+{
+    uint32_t lo, hi;
+
+    __asm__ __volatile__ (
+            "    divu1 $0, %2, %3\n"
+            "    mflo1 %0\n"
+            "    mfhi1 %1\n"
+            : "=r" (lo), "=r" (hi)
+            : "r" (rs), "r" (rt));
+
+    assert(rs / rt == lo);
+    assert(rs % rt == hi);
+
+    return (struct quotient_remainder) { .quotient = lo, .remainder = hi };
+}
+
+static void verify_divu1(uint32_t rs, uint32_t rt,
+                         uint32_t expected_quotient,
+                         uint32_t expected_remainder)
+{
+    struct quotient_remainder qr = divu1(rs, rt);
+
+    assert(qr.quotient == expected_quotient);
+    assert(qr.remainder == expected_remainder);
+}
+
+int main()
+{
+    verify_divu1(0, 1, 0, 0);
+    verify_divu1(1, 1, 1, 0);
+    verify_divu1(1, 2, 0, 1);
+    verify_divu1(17, 19, 0, 17);
+    verify_divu1(19, 17, 1, 2);
+    verify_divu1(77773, 101, 770, 3);
+
+    return 0;
+}

tests/tcg/mips/mipsr5900/mflohi1.c

+/*
+ * Test R5900-specific MFLO1 and MFHI1.
+ */
+
+#include <stdio.h>
+#include <inttypes.h>
+#include <assert.h>
+
+int main()
+{
+    int32_t rs  = 12207031, rt  = 305175781;
+    int32_t rs1 = 32452867, rt1 = 49979687;
+    int64_t lo, hi, lo1, hi1;
+    int64_t r, r1;
+
+    /* Test both LO/HI and LO1/HI1 to verify separation. */
+    __asm__ __volatile__ (
+            "    mult $0, %4, %5\n"
+            "    mult1 $0, %6, %7\n"
+            "    mflo %0\n"
+            "    mfhi %1\n"
+            "    mflo1 %2\n"
+            "    mfhi1 %3\n"
+            : "=r" (lo),  "=r" (hi),
+              "=r" (lo1), "=r" (hi1)
+            : "r" (rs),  "r" (rt),
+              "r" (rs1), "r" (rt1));
+    r  = ((int64_t)hi  << 32) | (uint32_t)lo;
+    r1 = ((int64_t)hi1 << 32) | (uint32_t)lo1;
+
+    assert(r  == 3725290219116211);
+    assert(r1 == 1621984134912629);
+
+    return 0;
+}

tests/tcg/mips/mipsr5900/mtlohi1.c

+/*
+ * Test R5900-specific MTLO1 and MTHI1.
+ */
+
+#include <stdio.h>
+#include <inttypes.h>
+#include <assert.h>
+
+int main()
+{
+    int32_t tlo  = 12207031, thi  = 305175781;
+    int32_t tlo1 = 32452867, thi1 = 49979687;
+    int32_t flo, fhi, flo1, fhi1;
+
+    /* Test both LO/HI and LO1/HI1 to verify separation. */
+    __asm__ __volatile__ (
+            "    mtlo  %4\n"
+            "    mthi  %5\n"
+            "    mtlo1 %6\n"
+            "    mthi1 %7\n"
+            "    move  %0, $0\n"
+            "    move  %1, $0\n"
+            "    move  %2, $0\n"
+            "    move  %3, $0\n"
+            "    mflo  %0\n"
+            "    mfhi  %1\n"
+            "    mflo1 %2\n"
+            "    mfhi1 %3\n"
+            : "=r" (flo),  "=r" (fhi),
+              "=r" (flo1), "=r" (fhi1)
+            : "r" (tlo),  "r" (thi),
+              "r" (tlo1), "r" (thi1));
+
+    assert(flo  == 12207031);
+    assert(fhi  == 305175781);
+    assert(flo1 == 32452867);
+    assert(fhi1 == 49979687);
+
+    return 0;
+}

tests/tcg/mips/mipsr5900/mult.c

+/*
+ * Test R5900-specific three-operand MULT and MULT1.
+ */
+
+#include <stdio.h>
+#include <inttypes.h>
+#include <assert.h>
+
+static int64_t mult(int32_t rs, int32_t rt)
+{
+    int32_t rd, lo, hi;
+    int64_t r;
+
+    __asm__ __volatile__ (
+            "    mult %0, %3, %4\n"
+            "    mflo %1\n"
+            "    mfhi %2\n"
+            : "=r" (rd), "=r" (lo), "=r" (hi)
+            : "r" (rs), "r" (rt));
+    r = ((int64_t)hi << 32) | (uint32_t)lo;
+
+    assert((int64_t)rs * rt == r);
+    assert(rd == lo);
+
+    return r;
+}
+
+static int64_t mult1(int32_t rs, int32_t rt)
+{
+    int32_t rd, lo, hi;
+    int64_t r;
+
+    __asm__ __volatile__ (
+            "    mult1 %0, %3, %4\n"
+            "    mflo1 %1\n"
+            "    mfhi1 %2\n"
+            : "=r" (rd), "=r" (lo), "=r" (hi)
+            : "r" (rs), "r" (rt));
+    r = ((int64_t)hi << 32) | (uint32_t)lo;
+
+    assert((int64_t)rs * rt == r);
+    assert(rd == lo);
+
+    return r;
+}
+
+static int64_t mult_variants(int32_t rs, int32_t rt)
+{
+    int64_t rd  = mult(rs, rt);
+    int64_t rd1 = mult1(rs, rt);
+
+    assert(rd == rd1);
+
+    return rd;
+}
+
+static void verify_mult_negations(int32_t rs, int32_t rt, int64_t expected)
+{
+    assert(mult_variants(rs, rt) == expected);
+    assert(mult_variants(-rs, rt) == -expected);
+    assert(mult_variants(rs, -rt) == -expected);
+    assert(mult_variants(-rs, -rt) == expected);
+}
+
+int main()
+{
+    verify_mult_negations(17, 19, 323);
+    verify_mult_negations(77773, 99991, 7776600043);
+    verify_mult_negations(12207031, 305175781, 3725290219116211);
+
+    assert(mult_variants(-0x80000000,  0x7FFFFFFF) == -0x3FFFFFFF80000000);
+    assert(mult_variants(-0x80000000, -0x7FFFFFFF) ==  0x3FFFFFFF80000000);
+    assert(mult_variants(-0x80000000, -0x80000000) ==  0x4000000000000000);
+
+    return 0;
+}

tests/tcg/mips/mipsr5900/multu.c

+/*
+ * Test R5900-specific three-operand MULTU and MULTU1.
+ */
+
+#include <stdio.h>
+#include <inttypes.h>
+#include <assert.h>
+
+static uint64_t multu(uint32_t rs, uint32_t rt)
+{
+    uint32_t rd, lo, hi;
+    uint64_t r;
+
+    __asm__ __volatile__ (
+            "    multu %0, %3, %4\n"
+            "    mflo %1\n"
+            "    mfhi %2\n"
+            : "=r" (rd), "=r" (lo), "=r" (hi)
+            : "r" (rs), "r" (rt));
+    r = ((uint64_t)hi << 32) | (uint32_t)lo;
+
+    assert((uint64_t)rs * rt == r);
+    assert(rd == lo);
+
+    return r;
+}
+
+static uint64_t multu1(uint32_t rs, uint32_t rt)
+{
+    uint32_t rd, lo, hi;
+    uint64_t r;
+
+    __asm__ __volatile__ (
+            "    multu1 %0, %3, %4\n"
+            "    mflo1 %1\n"
+            "    mfhi1 %2\n"
+            : "=r" (rd), "=r" (lo), "=r" (hi)
+            : "r" (rs), "r" (rt));
+    r = ((uint64_t)hi << 32) | (uint32_t)lo;
+
+    assert((uint64_t)rs * rt == r);
+    assert(rd == lo);
+
+    return r;
+}
+
+static uint64_t multu_variants(uint32_t rs, uint32_t rt)
+{
+    uint64_t rd  = multu(rs, rt);
+    uint64_t rd1 = multu1(rs, rt);
+
+    assert(rd == rd1);
+
+    return rd;
+}
+
+int main()
+{
+    assert(multu_variants(17, 19) == 323);
+    assert(multu_variants(77773, 99991) == 7776600043);
+    assert(multu_variants(12207031, 305175781) == 3725290219116211);
+
+    assert(multu_variants(0x80000000U, 0x7FFFFFFF) == 0x3FFFFFFF80000000);
+    assert(multu_variants(0x80000000U, 0x80000000U) ==  0x4000000000000000);
+    assert(multu_variants(0xFFFFFFFFU, 0xFFFFFFFFU) ==  0xFFFFFFFE00000001U);
+
+    return 0;
+}