5 * ternlogi <https://bugs.libre-soc.org/show_bug.cgi?id=745>
6 * grev <https://bugs.libre-soc.org/show_bug.cgi?id=755>
7 * remove Rc=1 from ternlog due to conflicts in encoding as well
8 as saving space <https://bugs.libre-soc.org/show_bug.cgi?id=753#c5>
14 this extension amalgamates bitmanipulation primitives from many sources, including RISC-V bitmanip, Packed SIMD, AVX-512 and OpenPOWER VSX. Vectorisation and SIMD are removed: these are straight scalar (element) operations making them suitable for embedded applications.
15 Vectorisation Context is provided by [[openpower/sv]].
17 When combined with SV, scalar variants of bitmanip operations found in VSX are added so that VSX may be retired as "legacy" in the far future (10 to 20 years). Also, VSX is hundreds of opcodes, requires 128 bit pathways, and is wholly unsuited to low power or embedded scenarios.
19 ternlogv is experimental and is the only operation that may be considered a "Packed SIMD". It is added as a variant of the already well-justified ternlog operation (done in AVX512 as an immediate only) "because it looks fun". As it is based on the LUT4 concept it will allow accelerated emulation of FPGAs. Other vendors of ISAs are buying FPGA companies to achieve similar objectives.
21 general-purpose Galois Field operations are added so as to avoid huge custom opcode proliferation across many areas of Computer Science. however for convenience and also to avoid setup costs, some of the more common operations (clmul, crc32) are also added. The expectation is that these operations would all be covered by the same pipeline.
23 note that there are brownfield spaces below that could incorporate some of the set-before-first and other scalar operations listed in [[sv/vector_ops]], and
24 the [[sv/av_opcodes]] as well as [[sv/setvl]]
28 * <https://en.wikiversity.org/wiki/Reed%E2%80%93Solomon_codes_for_coders>
29 * <https://maths-people.anu.edu.au/~brent/pd/rpb232tr.pdf>
33 minor opcode allocation
36 | ------ |--| --------- |
43 | 101 |0 | ternlogcr |
49 | dest | src1 | subop | op |
50 | ---- | ---- | ----- | -------- |
51 | RT | RA | .. | bmatflip |
55 | dest | src1 | src2 | subop | op |
56 | ---- | ---- | ---- | ----- | -------- |
57 | RT | RA | RB | or | bmatflip |
58 | RT | RA | RB | xor | bmatflip |
59 | RT | RA | RB | | grev |
60 | RT | RA | RB | | clmul* |
61 | RT | RA | RB | | gorc |
62 | RT | RA | RB | shuf | shuffle |
63 | RT | RA | RB | unshuf| shuffle |
64 | RT | RA | RB | width | xperm |
65 | RT | RA | RB | type | minmax |
66 | RT | RA | RB | | av abs avgadd |
67 | RT | RA | RB | type | vmask ops |
76 | 0.5|6.10|11.15|16.20|21..25 | 26....30 |31| name |
77 | -- | -- | --- | --- | ----- | -------- |--| ------ |
78 | NN | RT | RA | RB | RC | mode 000 |1 | ternlog |
79 | NN | RT | RA | RB | im0-4 | im5-7 00 |0 | ternlogi |
80 | NN | RS | RA | RB | RC | 00 011 |Rc| gfmul |
81 | NN | RS | RA | RB | RC | 01 011 |Rc| gfadd |
82 | NN | RT | RA | RB | deg | 10 011 |Rc| gfinv |
83 | NN | RS | RA | RB | deg | 11 011 |Rc| gfmuli |
84 | NN | RS | RA | RB | deg | 11 111 |Rc| gfaddi |
86 | 0.5|6.10|11.15| 16.23 |24.27 | 28.30 |31| name |
87 | -- | -- | --- | ----- | ---- | ----- |--| ------ |
88 | NN | RT | RA | imm | mask | 101 |1 | ternlogv |
90 | 0.5|6.8 | 9.11|12.14|15|16.23|24.27 | 28.30|31| name |
91 | -- | -- | --- | --- |- |-----|----- | -----|--| -------|
92 | NN | BA | BB | BC |0 |imm | mask | 101 |0 | ternlogcr |
94 ops (note that av avg and abs as well as vec scalar mask
97 | 0.5|6.10|11.15|16.20| 21.22 | 23 | 24....30 |31| name |
98 | -- | -- | --- | --- | ----- | -- | -------- |--| ---- |
99 | NN | RA | RB | | | 0 | 0000 110 |Rc| rsvd |
100 | NN | RA | RB | RC | itype | 1 | 0000 110 |Rc| xperm |
101 | NN | RA | RB | RC | itype | 0 | 0100 110 |Rc| minmax |
102 | NN | RA | RB | RC | 00 | 1 | 0100 110 |Rc| av avgadd |
103 | NN | RA | RB | RC | 01 | 1 | 0100 110 |Rc| av abs |
104 | NN | RA | RB | | 10 | 1 | 0100 110 |Rc| rsvd |
105 | NN | RA | RB | | 11 | 1 | 0100 110 |Rc| rsvd |
106 | NN | RA | RB | sh | itype | SH | 1000 110 |Rc| bmopsi |
107 | NN | RA | RB | | | | 1100 110 |Rc| rsvd |
108 | NN | RA | RB | | | 1 | 0001 110 |Rc| rsvd |
109 | NN | RA | RB | RC | 00 | 0 | 0001 110 |Rc| vec sbfm |
110 | NN | RA | RB | RC | 01 | 0 | 0001 110 |Rc| vec sofm |
111 | NN | RA | RB | RC | 10 | 0 | 0001 110 |Rc| vec sifm |
112 | NN | RA | RB | RC | 11 | 0 | 0001 110 |Rc| vec cprop |
113 | NN | RA | RB | | | 0 | 0101 110 |Rc| rsvd |
114 | NN | RA | RB | RC | 00 | 0 | 0010 110 |Rc| gorc |
115 | NN | RA | RB | sh | 00 | SH | 1010 110 |Rc| gorci |
116 | NN | RA | RB | RC | 00 | 0 | 0110 110 |Rc| gorcw |
117 | NN | RA | RB | sh | 00 | 0 | 1110 110 |Rc| gorcwi |
118 | NN | RA | RB | RC | 00 | 1 | 1110 110 |Rc| bmator |
119 | NN | RA | RB | RC | 01 | 0 | 0010 110 |Rc| grev |
120 | NN | RA | RB | RC | 01 | 1 | 0010 110 |Rc| clmul |
121 | NN | RA | RB | sh | 01 | SH | 1010 110 |Rc| grevi |
122 | NN | RA | RB | RC | 01 | 0 | 0110 110 |Rc| grevw |
123 | NN | RA | RB | sh | 01 | 0 | 1110 110 |Rc| grevwi |
124 | NN | RA | RB | RC | 01 | 1 | 1110 110 |Rc| bmatxor |
125 | NN | RA | RB | RC | 10 | 0 | 0010 110 |Rc| shfl |
126 | NN | RA | RB | sh | 10 | SH | 1010 110 |Rc| shfli |
127 | NN | RA | RB | RC | 10 | 0 | 0110 110 |Rc| shflw |
128 | NN | RA | RB | RC | 10 | | 1110 110 |Rc| rsvd |
129 | NN | RA | RB | RC | 11 | 0 | 1110 110 |Rc| clmulr |
130 | NN | RA | RB | RC | 11 | 1 | 1110 110 |Rc| clmulh |
131 | NN | | | | | | --11 110 |Rc| setvl |
133 # bit to byte permute
135 similar to matrix permute in RV bitmanip, which has XOR and OR variants
139 b = VSR[VRB+32].dword[i].byte[k].bit[j]
140 VSR[VRT+32].dword[i].byte[j].bit[k] = b
144 signed and unsigned min/max for integer. this is sort-of partly synthesiseable in [[sv/svp64]] with pred-result as long as the dest reg is one of the sources, but not both signed and unsigned. when the dest is also one of the srces and the mv fails due to the CR bittest failing this will only overwrite the dest where the src is greater (or less).
146 signed/unsigned min/max gives more flexibility.
149 uint_xlen_t min(uint_xlen_t rs1, uint_xlen_t rs2)
150 { return (int_xlen_t)rs1 < (int_xlen_t)rs2 ? rs1 : rs2;
152 uint_xlen_t max(uint_xlen_t rs1, uint_xlen_t rs2)
153 { return (int_xlen_t)rs1 > (int_xlen_t)rs2 ? rs1 : rs2;
155 uint_xlen_t minu(uint_xlen_t rs1, uint_xlen_t rs2)
156 { return rs1 < rs2 ? rs1 : rs2;
158 uint_xlen_t maxu(uint_xlen_t rs1, uint_xlen_t rs2)
159 { return rs1 > rs2 ? rs1 : rs2;
166 Similar to FPGA LUTs: for every bit perform a lookup into a table using an 8bit immediate, or in another register.
168 Like the x86 AVX512F [vpternlogd/vpternlogq](https://www.felixcloutier.com/x86/vpternlogd:vpternlogq) instructions.
172 | 0.5|6.10|11.15|16.20| 21..25| 26..30 |31|
173 | -- | -- | --- | --- | ----- | -------- |--|
174 | NN | RT | RA | RB | im0-4 | im5-7 00 |0 |
177 idx = RT[i] << 2 | RA[i] << 1 | RB[i]
178 RT[i] = (imm & (1<<idx)) != 0
180 bits 21..22 may be used to specify a mode, such as treating the whole integer zero/nonzero and putting 1/0 in the result, rather than bitwise test.
184 a 4 operand variant which becomes more along the lines of an FPGA:
186 | 0.5|6.10|11.15|16.20|21.25| 26...30 |31|
187 | -- | -- | --- | --- | --- | -------- |--|
188 | NN | RT | RA | RB | RC | mode 100 |1 |
191 idx = RT[i] << 2 | RA[i] << 1 | RB[i]
192 RT[i] = (RC & (1<<idx)) != 0
194 mode (2 bit) may be used to do inversion of ordering, similar to carryless mul,
199 also, another possible variant involving swizzle and vec4:
201 | 0.5|6.10|11.15| 16.23 |24.27 | 28.30 |31|
202 | -- | -- | --- | ----- | ---- | ----- |--|
203 | NN | RT | RA | imm | mask | 101 |1 |
206 idx = RA.x[i] << 2 | RA.y[i] << 1 | RA.z[i]
207 res = (imm & (1<<idx)) != 0
209 if mask[j]: RT[i+j*8] = res
213 another mode selection would be CRs not Ints.
215 | 0.5|6.8 | 9.11|12.14|15|16.23|24.27 | 28.30|31|
216 | -- | -- | --- | --- |- |-----|----- | -----|--|
217 | NN | BA | BB | BC |0 |imm | mask | 101 |0 |
220 if not mask[i] continue
221 idx = crregs[BA][i] << 2 |
224 crregs[BA][i] = (imm & (1<<idx)) != 0
228 based on RV bitmanip, covered by ternlog bitops
231 uint_xlen_t cmix(uint_xlen_t RA, uint_xlen_t RB, uint_xlen_t RC) {
232 return (RA & RB) | (RC & ~RB);
239 based on RV bitmanip singlebit set, instruction format similar to shift
240 [[isa/fixedshift]]. bmext is actually covered already (shift-with-mask rldicl but only immediate version).
241 however bitmask-invert is not, and set/clr are not covered, although they can use the same Shift ALU.
243 bmext (RB) version is not the same as rldicl because bmext is a right shift by RC, where rldicl is a left rotate. for the immediate version this does not matter, so a bmexti is not required.
244 bmrev however there is no direct equivalent and consequently a bmrevi is required.
246 bmset (register for mask amount) is particularly useful for creating
247 predicate masks where the length is a dynamic runtime quantity.
248 bmset(RA=0, RB=0, RC=mask) will produce a run of ones of length "mask" in a single instruction without needing to initialise or depend on any other registers.
250 | 0.5|6.10|11.15|16.20|21.25| 26..30 |31| name |
251 | -- | -- | --- | --- | --- | ------- |--| ----- |
252 | NN | RT | RA | RB | RC | mode 010 |Rc| bm* |
253 | NN | RT | RA | RB | RC | 0 1 111 |Rc| bmrev |
257 uint_xlen_t bmset(RA, RB, sh)
259 int shamt = RB & (XLEN - 1);
261 return RA | (mask << shamt);
264 uint_xlen_t bmclr(RA, RB, sh)
266 int shamt = RB & (XLEN - 1);
268 return RA & ~(mask << shamt);
271 uint_xlen_t bminv(RA, RB, sh)
273 int shamt = RB & (XLEN - 1);
275 return RA ^ (mask << shamt);
278 uint_xlen_t bmext(RA, RB, sh)
280 int shamt = RB & (XLEN - 1);
282 return mask & (RA >> shamt);
286 bitmask extract with reverse. can be done by bitinverting all of RA and getting bits of RA from the opposite end.
290 rev[0:msb] = ra[msb:0];
293 uint_xlen_t bmextrev(RA, RB, sh)
295 int shamt = (RB & (XLEN - 1));
296 shamt = (XLEN-1)-shamt; # shift other end
297 bra = bitreverse(RA) # swap LSB-MSB
299 return mask & (bra >> shamt);
303 | 0.5|6.10|11.15|16.20|21.26| 27..30 |31| name |
304 | -- | -- | --- | --- | --- | ------- |--| ------ |
305 | NN | RT | RA | RB | sh | 0 111 |Rc| bmrevi |
313 <img src="https://upload.wikimedia.org/wikipedia/commons/thumb/8/8c/Butterfly_Network.jpg/474px-Butterfly_Network.jpg" />
316 uint64_t grev64(uint64_t RA, uint64_t RB)
320 if (shamt & 1) x = ((x & 0x5555555555555555LL) << 1) |
321 ((x & 0xAAAAAAAAAAAAAAAALL) >> 1);
322 if (shamt & 2) x = ((x & 0x3333333333333333LL) << 2) |
323 ((x & 0xCCCCCCCCCCCCCCCCLL) >> 2);
324 if (shamt & 4) x = ((x & 0x0F0F0F0F0F0F0F0FLL) << 4) |
325 ((x & 0xF0F0F0F0F0F0F0F0LL) >> 4);
326 if (shamt & 8) x = ((x & 0x00FF00FF00FF00FFLL) << 8) |
327 ((x & 0xFF00FF00FF00FF00LL) >> 8);
328 if (shamt & 16) x = ((x & 0x0000FFFF0000FFFFLL) << 16) |
329 ((x & 0xFFFF0000FFFF0000LL) >> 16);
330 if (shamt & 32) x = ((x & 0x00000000FFFFFFFFLL) << 32) |
331 ((x & 0xFFFFFFFF00000000LL) >> 32);
337 # shuffle / unshuffle
342 uint32_t shfl32(uint32_t RA, uint32_t RB)
346 if (shamt & 8) x = shuffle32_stage(x, 0x00ff0000, 0x0000ff00, 8);
347 if (shamt & 4) x = shuffle32_stage(x, 0x0f000f00, 0x00f000f0, 4);
348 if (shamt & 2) x = shuffle32_stage(x, 0x30303030, 0x0c0c0c0c, 2);
349 if (shamt & 1) x = shuffle32_stage(x, 0x44444444, 0x22222222, 1);
352 uint32_t unshfl32(uint32_t RA, uint32_t RB)
356 if (shamt & 1) x = shuffle32_stage(x, 0x44444444, 0x22222222, 1);
357 if (shamt & 2) x = shuffle32_stage(x, 0x30303030, 0x0c0c0c0c, 2);
358 if (shamt & 4) x = shuffle32_stage(x, 0x0f000f00, 0x00f000f0, 4);
359 if (shamt & 8) x = shuffle32_stage(x, 0x00ff0000, 0x0000ff00, 8);
363 uint64_t shuffle64_stage(uint64_t src, uint64_t maskL, uint64_t maskR, int N)
365 uint64_t x = src & ~(maskL | maskR);
366 x |= ((src << N) & maskL) | ((src >> N) & maskR);
369 uint64_t shfl64(uint64_t RA, uint64_t RB)
373 if (shamt & 16) x = shuffle64_stage(x, 0x0000ffff00000000LL,
374 0x00000000ffff0000LL, 16);
375 if (shamt & 8) x = shuffle64_stage(x, 0x00ff000000ff0000LL,
376 0x0000ff000000ff00LL, 8);
377 if (shamt & 4) x = shuffle64_stage(x, 0x0f000f000f000f00LL,
378 0x00f000f000f000f0LL, 4);
379 if (shamt & 2) x = shuffle64_stage(x, 0x3030303030303030LL,
380 0x0c0c0c0c0c0c0c0cLL, 2);
381 if (shamt & 1) x = shuffle64_stage(x, 0x4444444444444444LL,
382 0x2222222222222222LL, 1);
385 uint64_t unshfl64(uint64_t RA, uint64_t RB)
389 if (shamt & 1) x = shuffle64_stage(x, 0x4444444444444444LL,
390 0x2222222222222222LL, 1);
391 if (shamt & 2) x = shuffle64_stage(x, 0x3030303030303030LL,
392 0x0c0c0c0c0c0c0c0cLL, 2);
393 if (shamt & 4) x = shuffle64_stage(x, 0x0f000f000f000f00LL,
394 0x00f000f000f000f0LL, 4);
395 if (shamt & 8) x = shuffle64_stage(x, 0x00ff000000ff0000LL,
396 0x0000ff000000ff00LL, 8);
397 if (shamt & 16) x = shuffle64_stage(x, 0x0000ffff00000000LL,
398 0x00000000ffff0000LL, 16);
408 uint_xlen_t xperm(uint_xlen_t RA, uint_xlen_t RB, int sz_log2)
411 uint_xlen_t sz = 1LL << sz_log2;
412 uint_xlen_t mask = (1LL << sz) - 1;
413 for (int i = 0; i < XLEN; i += sz) {
414 uint_xlen_t pos = ((RB >> i) & mask) << sz_log2;
416 r |= ((RA >> pos) & mask) << i;
420 uint_xlen_t xperm_n (uint_xlen_t RA, uint_xlen_t RB)
421 { return xperm(RA, RB, 2); }
422 uint_xlen_t xperm_b (uint_xlen_t RA, uint_xlen_t RB)
423 { return xperm(RA, RB, 3); }
424 uint_xlen_t xperm_h (uint_xlen_t RA, uint_xlen_t RB)
425 { return xperm(RA, RB, 4); }
426 uint_xlen_t xperm_w (uint_xlen_t RA, uint_xlen_t RB)
427 { return xperm(RA, RB, 5); }
435 uint32_t gorc32(uint32_t RA, uint32_t RB)
439 if (shamt & 1) x |= ((x & 0x55555555) << 1) | ((x & 0xAAAAAAAA) >> 1);
440 if (shamt & 2) x |= ((x & 0x33333333) << 2) | ((x & 0xCCCCCCCC) >> 2);
441 if (shamt & 4) x |= ((x & 0x0F0F0F0F) << 4) | ((x & 0xF0F0F0F0) >> 4);
442 if (shamt & 8) x |= ((x & 0x00FF00FF) << 8) | ((x & 0xFF00FF00) >> 8);
443 if (shamt & 16) x |= ((x & 0x0000FFFF) << 16) | ((x & 0xFFFF0000) >> 16);
446 uint64_t gorc64(uint64_t RA, uint64_t RB)
450 if (shamt & 1) x |= ((x & 0x5555555555555555LL) << 1) |
451 ((x & 0xAAAAAAAAAAAAAAAALL) >> 1);
452 if (shamt & 2) x |= ((x & 0x3333333333333333LL) << 2) |
453 ((x & 0xCCCCCCCCCCCCCCCCLL) >> 2);
454 if (shamt & 4) x |= ((x & 0x0F0F0F0F0F0F0F0FLL) << 4) |
455 ((x & 0xF0F0F0F0F0F0F0F0LL) >> 4);
456 if (shamt & 8) x |= ((x & 0x00FF00FF00FF00FFLL) << 8) |
457 ((x & 0xFF00FF00FF00FF00LL) >> 8);
458 if (shamt & 16) x |= ((x & 0x0000FFFF0000FFFFLL) << 16) |
459 ((x & 0xFFFF0000FFFF0000LL) >> 16);
460 if (shamt & 32) x |= ((x & 0x00000000FFFFFFFFLL) << 32) |
461 ((x & 0xFFFFFFFF00000000LL) >> 32);
469 see <https://courses.csail.mit.edu/6.857/2016/files/ffield.py>
473 this requires 3 parameters and a "degree"
475 RT = GFMUL(RA, RB, gfdegree, modulo=RC)
477 realistically with the degree also needing to be an immediate it should be brought down to an overwrite version:
479 RS = GFMUL(RS, RA, gfdegree, modulo=RC)
480 RS = GFMUL(RS, RA, gfdegree=RB, modulo=RC)
482 | 0.5|6.10|11.15|16.20|21.25| 26..30 |31|
483 | -- | -- | --- | --- | --- | ------- |--|
484 | NN | RS | RA | deg | RC | 00 011 |Rc|
485 | NN | RS | RA | RB | RC | 11 011 |Rc|
487 where the SimpleV variant may override RS-as-src differently from RS-as-dest
492 from functools import reduce
494 # constants used in the multGF2 function
495 mask1 = mask2 = polyred = None
497 def setGF2(degree, irPoly):
498 """Define parameters of binary finite field GF(2^m)/g(x)
499 - degree: extension degree of binary field
500 - irPoly: coefficients of irreducible polynomial g(x)
503 """Convert an integer into a polynomial"""
504 return [(sInt >> i) & 1
505 for i in reversed(range(sInt.bit_length()))]
507 global mask1, mask2, polyred
508 mask1 = mask2 = 1 << degree
510 polyred = reduce(lambda x, y: (x << 1) + y, i2P(irPoly)[1:])
513 """Multiply two polynomials in GF(2^m)/g(x)"""
524 if __name__ == "__main__":
526 # Define binary field GF(2^3)/x^3 + x + 1
529 # Evaluate the product (x^2 + x + 1)(x^2 + 1)
530 print("{:02x}".format(multGF2(0b111, 0b101)))
532 # Define binary field GF(2^8)/x^8 + x^4 + x^3 + x + 1
533 # (used in the Advanced Encryption Standard-AES)
534 setGF2(8, 0b100011011)
536 # Evaluate the product (x^7)(x^7 + x + 1)
537 print("{:02x}".format(multGF2(0b10000000, 0b10000011)))
541 RS = GFADDI(RS, RA|0, gfdegree, modulo=RC)
542 RS = GFADD(RS, RA|0, gfdegree=RB, modulo=RC)
544 | 0.5|6.10|11.15|16.20|21.25| 26..30 |31| name |
545 | -- | -- | --- | --- | --- | ------- |--| ----- |
546 | NN | RS | RA | deg | RC | 0 1 011 |Rc| gfaddi |
547 | NN | RS | RA | RB | RC | 1 1 111 |Rc| gfadd |
549 GFMOD is a pseudo-op where RA=0
562 def gf_invert(a, mod=0x1B) :
577 a %= 256 # Emulating 8-bit overflow
578 g1 %= 256 # Emulating 8-bit overflow
580 j = gf_degree(a) - gf_degree(v)
588 see https://en.wikipedia.org/wiki/CLMUL_instruction_set
590 these are GF2 operations with the modulo set to 2^degree.
591 they are worth adding as their own non-overwrite operations
592 (in the same pipeline).
595 uint_xlen_t clmul(uint_xlen_t RA, uint_xlen_t RB)
598 for (int i = 0; i < XLEN; i++)
603 uint_xlen_t clmulh(uint_xlen_t RA, uint_xlen_t RB)
606 for (int i = 1; i < XLEN; i++)
611 uint_xlen_t clmulr(uint_xlen_t RA, uint_xlen_t RB)
614 for (int i = 0; i < XLEN; i++)
616 x ^= RA >> (XLEN-i-1);
624 uint64_t bmatflip(uint64_t RA)
632 uint64_t bmatxor(uint64_t RA, uint64_t RB)
635 uint64_t RBt = bmatflip(RB);
636 uint8_t u[8]; // rows of RA
637 uint8_t v[8]; // cols of RB
638 for (int i = 0; i < 8; i++) {
643 for (int i = 0; i < 64; i++) {
644 if (pcnt(u[i / 8] & v[i % 8]) & 1)
649 uint64_t bmator(uint64_t RA, uint64_t RB)
652 uint64_t RBt = bmatflip(RB);
653 uint8_t u[8]; // rows of RA
654 uint8_t v[8]; // cols of RB
655 for (int i = 0; i < 8; i++) {
660 for (int i = 0; i < 64; i++) {
661 if ((u[i / 8] & v[i % 8]) != 0)
669 # Already in POWER ISA
671 ## count leading/trailing zeros with mask
677 do i = 0 to 63 if((RB)i=1) then do
678 if((RS)i=1) then break end end count ← count + 1
684 vpdepd VRT,VRA,VRB, identical to RV bitmamip bdep, found already in v3.1 p106
687 if VSR[VRB+32].dword[i].bit[63-m]=1 then do
688 result = VSR[VRA+32].dword[i].bit[63-k]
689 VSR[VRT+32].dword[i].bit[63-m] = result
695 uint_xlen_t bdep(uint_xlen_t RA, uint_xlen_t RB)
698 for (int i = 0, j = 0; i < XLEN; i++)
701 r |= uint_xlen_t(1) << i;
711 other way round: identical to RV bext, found in v3.1 p196
714 uint_xlen_t bext(uint_xlen_t RA, uint_xlen_t RB)
717 for (int i = 0, j = 0; i < XLEN; i++)
720 r |= uint_xlen_t(1) << j;
729 found in v3.1 p106 so not to be added here
739 if((RB)63-i==1) then do
740 result63-ptr1 = (RS)63-i