openpower/sv/po9_encoding/discussion.mdwn

   1 [[!toc]]
   2
   3 # introduction
   4
   5 the purpose of this page is to create and evaluate alternative
   6 encoding schemes that allow for a hybrid mixture of
   7 (ultra-precious) 32-bit and 64-bit (actually `x86 REP`-like
   8 prefixed) instructions.
   9
  10 a specific view to attaining high-speed massive-wide multi-issue decode
  11 is a high priority.  therefore especially length-detection must be kept
  12 brutally simple.
  13
  14 # **superseded** alternative 32-64 encoding (1)
  15
  16 conflict to resolve: EXT90x and EXT232. they are indistinguishable.
  17
  18 ```
  19 |0-5| 6-27 28 29|30-31|32|33-35|36-37| 38-59 | 60-63 | Description      |
  20 |---|-----|-----|-----|--|-----|-----|-------|-------|------------------|
  21 |PO9| rm0 | 0 0 | 0 0 | 0  000   rm1 |  xxxx | 0000  | SVP64:EXT900     |
  22 |PO9| rm0 | 0 0 | 0 0 | 1  000   rm1 |  xxxx | 0000  | SSingle:EXT900   |
  23 |PO9| xxx | x 0 | 0 0 | x !zero0     |  xxxx | !zero1| 55-bit RESERVED  |
  24 |PO9| xxx | 0 1 | 0 0 |              32-bit EXT900 (Vectorizable)       |
  25 |PO9| xxx | 1 1 | 0 0 |              32-bit EXT901 (Unvectorizable)     |
  26 |PO9|   !ZERO   | 0 0 |  1 | DWi                   | SSingle:EXT232-263 |
  27 |PO9|   0000    | 0 0 |  1 | DWi                   | Scalar EXT232-263  |
  28 |PO9|   RM      | 1 0 |  1 | DWi                   | SVP64:EXT232-263   |
  29 |PO9|   0000    | 0 1 | Defined Word-instruction   | 32-bit Unvec in 64b|
  30 |PO9|   !ZERO   | 0 1 | Defined Word-instruction   | SSingle:EXT000-063 |
  31 |PO9|   RM      | 1 1 | Defined Word-instruction   | SVP64:EXT000-063   |
  32 ```
  33
  34 Fields:
  35
  36 * `RM`: 24-bit SVP64 prefix
  37 * `RM <- rm0 || rm1` split field
  38 * `!zero0 || !zero1` a split field that must not be zero
  39
  40 Length detection:
  41
  42 ```
  43     if PO1                   return 64
  44     elif not PO9             return 32
  45     elif Word[31] = 1        return 64
  46     elif Word[29:30] = 0b10  return 32
  47     elif Word[29:30] = 0b00  return 64
  48     else                     return 32
  49 ```
  50
  51 Instruction allocation restrictions:
  52
  53 * setvl is Unvectorizable but needs more space later: use 55-bit
  54 * likewise svshape svindex and svshape2, all these need 32-bit
  55 * svstep *is* Vectorizable but needs to be encoded 32-bit in order
  56   to reduce loop-size
  57 * space for 32-bit instructions is needed: some instructions
  58   just make no sense if done as 64-bit because they **replace**
  59   two 32-bit instructions.
  60 * but those types of space-saving instructions
  61   **also need to be Vectorizable*
  62
  63 # **superseded** alternative 32-64 encoding (2)
  64
  65 requires reducing SVP64Single to 23 bits. luckily there are 2 spare
  66
  67 the complexity of attempting to fit 32-bit instructions into
  68 PO9 is very high. encoding (1) attempts to fit `setvl` etc. into
  69 32-bit but it is very tight.  the entire EXT900 area would be
  70 taken up by the SVP64 Management instructions, below, defeating
  71 the purpose of having it.
  72
  73 * `setvl` - Unvectorizable, EXT0xx, `RT RA CTR SVSTATE`, 6-bit XO (including Rc=1)
  74 * `psetvl` - Unvectorizable, EXT1xx, ditto
  75 * `svstep` - Vectorizable, either EXT900 (very precious) or EXT1xx (better), `RT, SVSTATE`, 6-bit XO (including Rc=1)
  76 * `svindex` - Unvectorizable, EXT0xx, `SVSTATE,SVSHAPE0-3`, 5-bit XO
  77 * `psvindex` - Unvectorizable, EXT1xx, ditto
  78 * `svstep(2)` - Unvectorizable, EXT0xx, `SVSTATE,SVSHAPE0-3`, 5-bit XO
  79 * `svstep(2)` - Unvectorizable, EXT1xx, ditto
  80
  81 ```
  82 |0-5| 6-28|29 30 31|32|33-36|37 | 38-60 | 61-63 | Description      |
  83 |---|-----|--------|--|-----|---|-------|-------|------------------|
  84 |PO9| rm0 | 1  0 0 | 0 0000  rm1|  xxxx | 000   | SVP64:EXT900     |
  85 |PO9|!ZERO| 1  0 0 | 0 1000  1  |  xxxx | 000   | SSingle:EXT900   |
  86 |PO9| xxx | 1  0 0 | 0 !zero0   |  xxxx | !zero1| 55-bit RESERVED  |
  87 |PO9| xxx | 0  0 0 |            32-bit EXT900                      |
  88 |PO9|!ZERO| 1  0 0 |  1 | DWi                 | SSingle:EXT232-263 |
  89 |PO9| 000 | 1  0 0 |  1 | DWi                 | Scalar EXT232-263  |
  90 |PO9|  nnnn  | 1 0 |  1 | DWi                 | SVP64:EXT232-263   |
  91 |PO9|  0000  | 0 1 | Defined Word-instruction | 32-bit Unvec in 64b|
  92 |PO9|!ZERO   | 0 1 | Defined Word-instruction | SSingle:EXT000-063 |
  93 |PO9|  nnnn  | 1 1 | Defined Word-instruction | SVP64:EXT000-063   |
  94 ```
  95
  96 Length detection:
  97
  98 ```
  99     if PO1                   return 64
 100     elif not PO9             return 32
 101     elif Word[29:31] = 0b000 return 32
 102     else                     return 64
 103 ```
 104
 105 # alternative **no 32 encoding** (3) **current**
 106
 107 aim of this idea is to attempt simplification of area identification
 108 and length.  the 55-bit area is eliminated and may only be reintroduced
 109 by sacrificing parts of EXT200-231, bear in mind that EXT240 is already
 110 unavailable (used by EXT900).
 111
 112 ```
 113 |0-5| 6-29   |30 31|32 |33-37   | 38-61 | 62-63 | Description      |
 114 |---|--------|-----|---|--------|-------|-------|------------------|
 115 |PO9| xxxx   | 0 0 | (next new instruction)     |  32-bit EXT900   |
 116 |PO9| 0000   | 1 0 | 0 | 10001  |  //// |  00   | RESERVED         |
 117 |PO9| !ZERO  | 1 0 | 0 | 10001  |  xxxx |  00   | SSingle:EXT900   |
 118 |PO9| SVRM   | 1 0 | 0 | 10001  |  xxxx |  01   | SVP64:EXT900     |
 119 |PO9| ////   | 1 0 | 0 | 10001  |  //// |  10   | RESERVED         |
 120 |PO9| ////   | 1 0 | 0 | 10001  |  //// |  11   | RESERVED         |
 121 |PO9| !ZERO  | 1 0 | 0 | !PO9   |     xxxx      | SSingle:EXT232-263 |
 122 |PO9|  0000  | 1 0 | 0 | !PO9   |     xxxx      | Scalar EXT232-263  |
 123 |PO9|  SVRM  | 1 0 | 1 | !PO9   |     xxxx      | SVP64:EXT232-263   |
 124 |PO9|  0000  | 0 1 | Defined Word-instruction   | 32-bit Unvec in 64b|
 125 |PO9| !ZERO  | 0 1 | Defined Word-instruction   | SSingle:EXT000-063 |
 126 |PO9|  SVRM  | 1 1 | Defined Word-instruction   | SVP64:EXT000-063   |
 127 ```
 128
 129 **Notes**:
 130
 131 * with `setvl` `svindex` `svshape` and `svshape2` being Unvectorizable Management
 132 instructions requiring 6-bit `XO` and needing EXT1xx variants later, they are
 133 best placed in EXT0xx.  EXT900 is Vectorizable and extremely precious.  Further
 134 complication of Decode to create Vectorized EXT900 and Unvectorized EXT901
 135 is inadviseable (rejected in alternative 1)
 136 * EXT200-231 is intended to be partially-filled with LD/ST-Immediate-Postincrement
 137   and it is strongly recommended that the exact same Primary Opcode be
 138   used, matching the lower 5 bits. `lzbu` is EXT033, therefore `lzbup` is
 139   recommended to be EXT233. This will greatly simplify Decode.
 140 * With no "overlapping" Encodings permitted on top of EXT232-263 (except
 141   EXT240) Speculative Decode may begin early in Multi-Issue systems safe
 142   that for the most part the 32-bit instruction is either going to be
 143   EXT0xx or EXT2xx, regardless of whether it is SVP64-Prefixed.
 144   LD/ST-Immediate-Postincrement discernment from
 145   LD/ST-Immediate is delayed, but Decode of other EXT232-263 operations
 146   are *also delayed* pending identification.
 147
 148
 149 **Identification**:
 150
 151 The following identification-criteria may be easily verified by taking
 152 a copy of the table above and striking out each line that fails each
 153 test.  Only EXT000-063 is not by a "process of elimination" (if ... return True)
 154
 155 **Length detection**: (critically-important to be short)
 156
 157 ```
 158     if PO1                   return 64
 159     elif not PO9             return 32
 160     elif Word[30:31] = 0b00  return 32
 161     else                     return 64
 162 ```
 163
 164 **EXT232-263:**
 165
 166 ```
 167     if not PO9                return False
 168     # eliminate EXT900
 169     if Word[30:31] == 0b00    return False
 170     # eliminate EXT0xx
 171     if Word[31] == 0b1        return False
 172     # eliminate EXT900 and reserved areas
 173     if Word[33:37] != 0b10001 return False
 174     #
 175     return True
 176 ```
 177
 178 **EXT900:**
 179
 180 ```
 181     # easy first step
 182     if not PO9                 return False
 183     # eliminate EXT0xx, 32-Unvec-in-64b
 184     if Word[31] = 0b1          return False
 185     # eliminate EXT2xx
 186     if Word[30:31] =  0b10 and
 187        Word[33:37] != 0b10001   return False
 188     # eliminate currently-RESERVED areas
 189     if Word[30:31] =  0b10 and
 190        Word[62]    =  0b1      return False
 191     # eliminate final RESERVED area
 192     if Word[30:31] =  0b10 and
 193        Word[62:63] =  0b00 and
 194        Word[6:29 ] =  0x000000 return False
 195     # last-remaining is EXT900
 196     return True
 197 ```
 198
 199 **EXT000-063**: (includes Prefixed)
 200
 201 ```
 202     if PO1                     return False # or other reserved
 203     if not PO9                 return True  # prefixed needs more
 204     # eliminate EXT900
 205     if Word[30:31] =  0b00     return False
 206     # eliminate Prefixed-EXT900, RESERVED and EXT200
 207     if Word[31]    =  0b0      return False
 208     # eliminate 32-bit Unvec in 64b area
 209     if Word[30:31] =  0b01 and
 210        Word[6:29 ] =  0x000000 return False
 211     return True
 212 ```
 213
 214 **SVP64**:
 215
 216 ```
 217     # easy-elimination, first
 218     if not PO9                 return False
 219     if Word[30]    =  0b0      return False
 220     # eliminate anything not SVP64:EXT900
 221     if Word[30:31] =  0b10 and
 222        Word[33:37] = 0b10001 and
 223        Word[62:63] != 0b01    return False
 224     # eliminate anything not SVP64:EXT200-231
 225     if Word[31:32] =  0b00    return False
 226     # all remaining combinations are SVP64:EXTnnn
 227     return True
 228 ```
 229
 230 **SVP64Single**:
 231
 232 ```
 233     # easy-elimination, first
 234     if not PO9                 return False
 235     # eliminate 32-bit EXT900
 236     if Word[30:31] =  0b00     return False
 237     # eliminate SVP64:EXT000-063
 238     if Word[30:31] =  0b11     return False
 239     # eliminate anything "Identity"
 240     if  Word[6:29 ] =  0x000000 return False
 241     # eliminate SVP64:EXT200-231
 242     if Word[30:32] =  0b101     return False
 243     # eliminate anything not SSingle:EXT900
 244     if Word[30:31] =  0b10 and
 245        Word[33:37] = 0b10001 and
 246        Word[62:63] != 0b00      return False
 247     # everything left is SVP64Single
 248     return True
 249 ```
 250
 251 **RESERVED**: (no need to pass for further detailed decode)
 252
 253 ```
 254     # easy-elimination, first
 255     if not PO9                 return False
 256     # eliminate EXT900
 257     if Word[30:31] =  0b00     return False
 258     # eliminate SVP64:EXT000-031
 259     if Word[30:31] =  0b11     return False
 260     # eliminate SSingle:EXT000-063
 261     if Word[6:29 ] != 0x000000 and
 262        Word[30:31]  =  0b01    return False
 263     # eliminate EXT200-231
 264     if Word[30:31] =  0b10 and
 265        Word[33:37] = 0b10001   return False
 266     # eliminate SSingle:EXT900
 267     if Word[30:31] =  0b10 and
 268        Word[6:29 ] != 0x000000 and
 269        Word[33:37] = 0b10001 and
 270        Word[62:63] = 0b00      return False
 271     # eliminate SVP64:EXT900
 272     if Word[30:31] =  0b10 and
 273        Word[33:37] = 0b10001 and
 274        Word[62:63] = 0b01      return False
 275     # all else needs further detailed decode
 276     return True
 277 ```
 278
 279 # alternative 32-64 encoding (4) **current**
 280
 281 this idea gets even simpler. the 55-bit area is eliminated and may only be
 282 reintroduced by sacrificing parts of EXT200-263. however doing so **WILL** damage
 283 high-performance decode.
 284
 285 advantages:
 286
 287 * brutally simple length detection (multi-issue 1 clock cycle)
 288 * extremely easy SVP64 identification
 289 * extremely easy SVP64Single identification
 290
 291
 292 ```
 293 |0-5| 6-29   |30 31|32-37   | 38-663 | Description      |
 294 |---|--------|-----|--------|--------|------------------|
 295 |PO9| xxxx   | x x | 010001 |  ////  | RESERVED         |
 296 |PO9| xxxx   | x x | 000001 |  ////  | RESERVED         |
 297 |PO9| !ZERO  | 1 1 |  !PO9  |     xxxx        | SSingle:EXT232-263 |
 298 |PO9|  0000  | 1 1 |  !PO9  |     xxxx        | Scalar EXT232-263  |
 299 |PO9|  SVRM  | 1 0 |  !PO9  |     xxxx        | SVP64:EXT232-263   |
 300 |PO9|  0000  | 0 1 | Defined Word-instruction | 32-bit EXT300-363  |
 301 |PO9| !ZERO  | 0 1 | Defined Word-instruction | SSingle:EXT000-063 |
 302 |PO9|  SVRM  | 0 0 | Defined Word-instruction | SVP64:EXT000-063   |
 303 ```
 304 * bit 31 is "SSingle"
 305 * bit 30 is EXT200-263 if set otherwise EXT000-063
 306
 307 This was one of the original early PO9 Encodings.
 308
 309 **Identification**:
 310
 311 The following identification-criteria may be easily verified by taking
 312 a copy of the table above and striking out each line that fails each
 313 test.  Only EXT000-063 is not by a "process of elimination" (if ... return True)
 314
 315 **Length detection**: (critically-important to be short - 1 clock cycle)
 316
 317 ```
 318     if PO1 or PO9            return 64
 319     else                     return 32
 320 ```
 321
 322 **EXT232-263:**
 323
 324 ```
 325     if not PO9                return False
 326     # eliminate EXT0xx
 327     if Word[31] == 0b1        return False
 328     # remaining is EXT2xx
 329     return True
 330 ```
 331
 332
 333
 334 **EXT000-063**: (includes Prefixed)
 335
 336 ```
 337     if PO1                     return False # or other reserved
 338     if not PO9                 return True  # prefixed needs more
 339     # eliminate EXT2xx
 340     if Word[31]    =  0b0      return False
 341     # remaining is EXT0xx
 342     return True
 343 ```
 344
 345 **SVP64**:
 346
 347 ```
 348     # easy-elimination, first
 349     if not PO9                 return False
 350     # eliminate anything not SVP64:
 351     if Word[31]    =  0b1      return False
 352     # all remaining are SVP64:EXT0xx/EXT2xx
 353     return True
 354 ```
 355
 356 **SVP64Single**:
 357
 358 ```
 359     # easy-elimination, first
 360     if not PO9                 return False
 361     # eliminate SVP64:*
 362     if Word[31] =  0b0         return False
 363     # eliminate anything "Identity"
 364     if  Word[6:29 ] =  0x000000 return False
 365     # everything left is SVP64Single
 366     return True
 367 ```
 368
 369 **RESERVED**: (no need to pass for further detailed decode)
 370
 371 ```
 372     # easy-elimination, first
 373     if not PO9                 return False
 374     # eliminate double-PO9 and PO9-PO1
 375     if Word[32:37] = 0b010001  return False
 376     if Word[32:37] = 0b000001  return False
 377     # all else requires detailed decode
 378     return True
 379 ```