summaryrefslogtreecommitdiffstats
path: root/js/src/jit-test/tests/wasm/binop-x64-ion-codegen.js
blob: 44031416586f71ccd5b14dd99adf99177823b7d3 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
// |jit-test| skip-if: !hasDisassembler() || wasmCompileMode() != "ion" || !getBuildConfiguration("x64") || getBuildConfiguration("simulator"); include:codegen-x64-test.js

// There may be some redundant moves to set some operations up on x64 (that's
// bug 1701164) so we avoid those with no_prefix/no_suffix flags when the issue
// comes up.

// Test that multiplication by -1 yields negation.

let neg32 =
    `(module
       (func (export "f") (param i32) (result i32)
         (i32.mul (local.get 0) (i32.const -1))))`;
codegenTestX64_adhoc(
    neg32,
    'f',
    'f7 d8  neg %eax', {no_prefix:true});
assertEq(wasmEvalText(neg32).exports.f(-37), 37)
assertEq(wasmEvalText(neg32).exports.f(42), -42)

let neg64 =
    `(module
       (func (export "f") (param i64) (result i64)
         (i64.mul (local.get 0) (i64.const -1))))`;
codegenTestX64_adhoc(
    neg64,
    'f',
    '48 f7 d8  neg %rax', {no_prefix:true});
assertEq(wasmEvalText(neg64).exports.f(-37000000000n), 37000000000n)
assertEq(wasmEvalText(neg64).exports.f(42000000000n), -42000000000n)

// Test that multiplication by zero yields zero

let zero32 =
    `(module
       (func (export "f") (param i32) (result i32)
         (i32.mul (local.get 0) (i32.const 0))))`;
codegenTestX64_adhoc(
    zero32,
    'f',
    '33 c0 xor %eax, %eax', {no_prefix:true});
assertEq(wasmEvalText(zero32).exports.f(-37), 0)
assertEq(wasmEvalText(zero32).exports.f(42), 0)

let zero64 = `(module
       (func (export "f") (param i64) (result i64)
         (i64.mul (local.get 0) (i64.const 0))))`
codegenTestX64_adhoc(
    zero64,
    'f',
    '48 33 c0 xor %rax, %rax', {no_prefix:true});
assertEq(wasmEvalText(zero64).exports.f(-37000000000n), 0n)
assertEq(wasmEvalText(zero64).exports.f(42000000000n), 0n)

// Test that multiplication by one yields no code

let one32 =
    `(module
       (func (export "f") (param i32) (result i32)
         (i32.mul (local.get 0) (i32.const 1))))`;
codegenTestX64_adhoc(
    one32,
    'f',
    '', {no_prefix:true});
assertEq(wasmEvalText(one32).exports.f(-37), -37)
assertEq(wasmEvalText(one32).exports.f(42), 42)

let one64 = `(module
       (func (export "f") (param i64) (result i64)
         (i64.mul (local.get 0) (i64.const 1))))`
codegenTestX64_adhoc(
    one64,
    'f',
    '', {no_prefix:true});
assertEq(wasmEvalText(one64).exports.f(-37000000000n), -37000000000n)
assertEq(wasmEvalText(one64).exports.f(42000000000n), 42000000000n)

// Test that multiplication by two yields an add

let double32 =
    `(module
       (func (export "f") (param i32) (result i32)
         (i32.mul (local.get 0) (i32.const 2))))`;
codegenTestX64_adhoc(
    double32,
    'f',
    '03 c0 add %eax, %eax', {no_prefix:true});
assertEq(wasmEvalText(double32).exports.f(-37), -74)
assertEq(wasmEvalText(double32).exports.f(42), 84)

let double64 = `(module
       (func (export "f") (param i64) (result i64)
         (i64.mul (local.get 0) (i64.const 2))))`
codegenTestX64_adhoc(
    double64,
    'f',
    '48 03 c0 add %rax, %rax', {no_prefix:true});
assertEq(wasmEvalText(double64).exports.f(-37000000000n), -74000000000n)
assertEq(wasmEvalText(double64).exports.f(42000000000n), 84000000000n)

// Test that multiplication by four yields a shift

let quad32 =
    `(module
       (func (export "f") (param i32) (result i32)
         (i32.mul (local.get 0) (i32.const 4))))`;
codegenTestX64_adhoc(
    quad32,
    'f',
    'c1 e0 02 shl \\$0x02, %eax', {no_prefix:true});
assertEq(wasmEvalText(quad32).exports.f(-37), -148)
assertEq(wasmEvalText(quad32).exports.f(42), 168)

let quad64 = `(module
       (func (export "f") (param i64) (result i64)
         (i64.mul (local.get 0) (i64.const 4))))`
codegenTestX64_adhoc(
    quad64,
    'f',
    '48 c1 e0 02 shl \\$0x02, %rax', {no_prefix:true});
assertEq(wasmEvalText(quad64).exports.f(-37000000000n), -148000000000n)
assertEq(wasmEvalText(quad64).exports.f(42000000000n), 168000000000n)

// Test that multiplication by five yields a multiply

let quint32 =
    `(module
       (func (export "f") (param i32) (result i32)
         (i32.mul (local.get 0) (i32.const 5))))`;
codegenTestX64_adhoc(
    quint32,
    'f',
    '6b c0 05 imul \\$0x05, %eax, %eax', {no_prefix:true});
assertEq(wasmEvalText(quint32).exports.f(-37), -37*5)
assertEq(wasmEvalText(quint32).exports.f(42), 42*5)

let quint64 = `(module
       (func (export "f") (param i64) (result i64)
         (i64.mul (local.get 0) (i64.const 5))))`
codegenTestX64_adhoc(
    quint64,
    'f',
    `48 6b c0 05               imul \\$0x05, %rax, %rax`, {no_prefix:true})
assertEq(wasmEvalText(quint64).exports.f(-37000000000n), -37000000000n*5n)
assertEq(wasmEvalText(quint64).exports.f(42000000000n), 42000000000n*5n)

// Test that 0-n yields negation.

let subneg32 =
    `(module
       (func (export "f") (param i32) (result i32)
         (i32.sub (i32.const 0) (local.get 0))))`
codegenTestX64_adhoc(
    subneg32,
    'f',
    'f7 d8  neg %eax', {no_prefix:true});
assertEq(wasmEvalText(subneg32).exports.f(-37), 37)
assertEq(wasmEvalText(subneg32).exports.f(42), -42)

let subneg64 =
    `(module
       (func (export "f") (param i64) (result i64)
         (i64.sub (i64.const 0) (local.get 0))))`
codegenTestX64_adhoc(
    subneg64,
    'f',
    '48 f7 d8  neg %rax', {no_prefix:true});
assertEq(wasmEvalText(subneg64).exports.f(-37000000000n), 37000000000n)
assertEq(wasmEvalText(subneg64).exports.f(42000000000n), -42000000000n)

// AND{32,64} followed by `== 0`: check the two operations are merged into a
// single 'test' insn, and no 'and' insn.  The merging isn't done for
// {OR,XOR}{32,64}.  This is for both arguments being non-constant.

for ( [ty, expect_test] of
      [['i32',   '85 ..     test %e.., %e..'],
       ['i64',   '48 85 ..  test %r.., %r..']] ) {
   codegenTestX64_adhoc(
    `(module
       (func (export "f") (param $p1 ${ty}) (param $p2 ${ty}) (result i32)
         (local $x i32)
         (local.set $x (i32.const 0x4D2))
         (if (${ty}.eq (${ty}.and (local.get $p1) (local.get $p2))
                       (${ty}.const 0))
           (then (local.set $x (i32.const 0x11D7)))
         )
         (local.get $x)
       )
    )`,
    'f',
    `${expect_test}
     0f 85 .. 00 00 00   jnz 0x00000000000000..
     b8 d7 11 00 00      mov \\$0x11D7, %eax
     e9 .. 00 00 00      jmp 0x00000000000000..
     b8 d2 04 00 00      mov \\$0x4D2, %eax`
   );
}

// AND64 followed by `== 0`, with one of the args being a constant.  Depending
// on the constant we can get one of three forms.

for ( [imm, expect1, expect2] of
      [ // in signed-32 range => imm in insn
        ['0x17654321',
         'f7 c. 21 43 65 17   test \\$0x17654321, %e..', // edi or ecx
         ''],
        // in unsigned-32 range => imm in reg via movl
        ['0x87654321',
         '41 bb 21 43 65 87   mov \\$-0x789ABCDF, %r11d',
         '4c 85 d.            test %r11, %r..'], // rdi or rcx
        // not in either range => imm in reg via mov(absq)
        ['0x187654321',
         '49 bb 21 43 65 87 01 00 00 00   mov \\$0x187654321, %r11',
         '4c 85 d.   test %r11, %r..']] // rdi or rcx
      ) {
   codegenTestX64_adhoc(
    `(module
       (func (export "f") (param $p1 i64) (result i32)
         (local $x i32)
         (local.set $x (i32.const 0x4D2))
         (if (i64.eq (i64.and (i64.const ${imm}) (local.get $p1))
                     (i64.const 0))
           (then (local.set $x (i32.const 0x11D7)))
         )
         (local.get $x)
       )
    )`,
    'f',
    `${expect1}
     ${expect2}
     0f 85 .. 00 00 00   jnz 0x00000000000000..
     b8 d7 11 00 00      mov \\$0x11D7, %eax
     e9 .. 00 00 00      jmp 0x00000000000000..
     b8 d2 04 00 00      mov \\$0x4D2, %eax`
   );
}

// For integer comparison followed by select, check that the comparison result
// isn't materialised into a register, for specific types.

function cmpSel32vs64(cmpTy, cmpOp, selTy) {
    return `(module
              (func (export "f")
                    (param $p1 ${cmpTy}) (param $p2 ${cmpTy})
                    (param $p3 ${selTy}) (param $p4 ${selTy})
                    (result ${selTy})
                (select (local.get $p3)
                        (local.get $p4)
                        (${cmpTy}.${cmpOp} (local.get $p1) (local.get $p2)))
              )
            )`;
}
if (getBuildConfiguration("windows")) {
    for ( [cmpTy, cmpOp, selTy, insn1, insn2, insn3] of
          [ ['i32', 'le_s', 'i32',  '8b c3        mov %ebx, %eax',
                                    '3b ca        cmp %edx, %ecx',
                                    '41 0f 4f c1  cmovnle %r9d, %eax'],
            ['i32', 'lt_u', 'i64',  '48 89 d8     mov %rbx, %rax',
                                    '3b ca        cmp %edx, %ecx',
                                    '49 0f 43 c1  cmovnb %r9, %rax'],
            ['i64', 'le_s', 'i32',  '8b c3        mov %ebx, %eax',
                                    '48 3b ca     cmp %rdx, %rcx',
                                    '41 0f 4f c1  cmovnle %r9d, %eax'],
            ['i64', 'lt_u', 'i64',  '48 89 d8     mov %rbx, %rax',
                                    '48 3b ca     cmp %rdx, %rcx',
                                    '49 0f 43 c1  cmovnb %r9, %rax']
          ] ) {
        codegenTestX64_adhoc(cmpSel32vs64(cmpTy, cmpOp, selTy), 'f',
          `4. (89 c3|8b d8)   mov %r8.*, %.bx
           ${insn1}
           ${insn2}
           ${insn3}`
        );
    }
} else {
    for ( [cmpTy, cmpOp, selTy, insn1, insn2, insn3] of
          [ ['i32', 'le_s', 'i32',  '8b c2        mov %edx, %eax',
                                    '3b fe        cmp %esi, %edi',
                                    '0f 4f c1     cmovnle %ecx, %eax'],
            ['i32', 'lt_u', 'i64',  '48 89 d0     mov %rdx, %rax',
                                    '3b fe        cmp %esi, %edi',
                                    '48 0f 43 c1  cmovnb %rcx, %rax'],
            ['i64', 'le_s', 'i32',  '8b c2        mov %edx, %eax',
                                    '48 3b fe     cmp %rsi, %rdi',
                                    '0f 4f c1     cmovnle %ecx, %eax'],
            ['i64', 'lt_u', 'i64',  '48 89 d0     mov %rdx, %rax',
                                    '48 3b fe     cmp %rsi, %rdi',
                                    '48 0f 43 c1  cmovnb %rcx, %rax']
          ] ) {
        codegenTestX64_adhoc(cmpSel32vs64(cmpTy, cmpOp, selTy), 'f',
          `${insn1}
           ${insn2}
           ${insn3}`
        );
    }
}

// For integer comparison followed by select, check correct use of operands in
// registers vs memory.  At least for the 64-bit-cmp/64-bit-sel case.

for ( [pAnyCmp, pAnySel, cmpBytes, cmpArgL, cmovBytes, cmovArgL ] of
      [ // r, r
        ['$pReg1', '$pReg2',
         '4. .. ..', '%r.+',  '4. .. .. ..', '%r.+'],
        // r, m
        ['$pReg1', '$pMem2',
         '4. .. ..', '%r.+',  '4. .. .. .. ..', '0x..\\(%rbp\\)'],
        // m, r
        ['$pMem1', '$pReg2',
         '4. .. .. ..', '0x..\\(%rbp\\)', '4. .. .. ..', '%r.+'],
        // m, m
        ['$pMem1', '$pMem2',
         '4. .. .. ..', '0x..\\(%rbp\\)', '4. .. .. .. ..', '0x..\\(%rbp\\)']
      ] ) {
   codegenTestX64_adhoc(
    `(module
       (func (export "f")
             (param $p1 i64)    (param $p2 i64)     ;; in regs for both ABIs
             (param $pReg1 i64) (param $pReg2 i64)  ;; in regs for both ABIs
             (param i64)        (param i64)         ;; ignored
             (param $pMem1 i64) (param $pMem2 i64)  ;; in mem for both ABIs
             (result i64)
         (select (local.get $p1)
                 (local.get ${pAnySel})
                 (i64.eq (local.get $p2) (local.get ${pAnyCmp})))
       )
    )`,
    'f',
    // On Linux we have an extra move
    (getBuildConfiguration("windows") ? '' : '48 89 ..       mov %r.+, %r.+\n') +
    // 'q*' because the disassembler shows 'q' only for the memory cases
    `48 89 ..       mov %r.+, %r.+
     ${cmpBytes}    cmpq*    ${cmpArgL}, %r.+
     ${cmovBytes}   cmovnzq* ${cmovArgL}, %r.+`
   );
}