| The Android Open Source Project | 89c1feb | 2008-12-17 18:03:55 -0800 | [diff] [blame^] | 1 | <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> |
| 2 | |
| 3 | <html> |
| 4 | |
| 5 | <head> |
| 6 | <title>binop</title> |
| 7 | <link rel=stylesheet href="opcode.css"> |
| 8 | </head> |
| 9 | |
| 10 | <body> |
| 11 | |
| 12 | <h1>binop</h1> |
| 13 | |
| 14 | <h2>Purpose</h2> |
| 15 | |
| 16 | <p> |
| 17 | Perform the identified binary operation on the two source registers, storing |
| 18 | the result in the first source register. |
| 19 | </p> |
| 20 | |
| 21 | <h2>Details</h2> |
| 22 | |
| 23 | <table class="instruc"> |
| 24 | <thead> |
| 25 | <tr> |
| 26 | <th>Op & Format</th> |
| 27 | <th>Mnemonic / Syntax</th> |
| 28 | <th>Arguments</th> |
| 29 | </tr> |
| 30 | </thead> |
| 31 | <tbody> |
| 32 | <tr> |
| 33 | <td>90..af 23x</td> |
| 34 | <td><i>binop</i> vAA, vBB, vCC<br/> |
| 35 | 90: add-int<br/> |
| 36 | 91: sub-int<br/> |
| 37 | 92: mul-int<br/> |
| 38 | 93: div-int<br/> |
| 39 | 94: rem-int<br/> |
| 40 | 95: and-int<br/> |
| 41 | 96: or-int<br/> |
| 42 | 97: xor-int<br/> |
| 43 | 98: shl-int<br/> |
| 44 | 99: shr-int<br/> |
| 45 | 9a: ushr-int<br/> |
| 46 | 9b: add-long<br/> |
| 47 | 9c: sub-long<br/> |
| 48 | 9d: mul-long<br/> |
| 49 | 9e: div-long<br/> |
| 50 | 9f: rem-long<br/> |
| 51 | a0: and-long<br/> |
| 52 | a1: or-long<br/> |
| 53 | a2: xor-long<br/> |
| 54 | a3: shl-long<br/> |
| 55 | a4: shr-long<br/> |
| 56 | a5: ushr-long<br/> |
| 57 | a6: add-float<br/> |
| 58 | a7: sub-float<br/> |
| 59 | a8: mul-float<br/> |
| 60 | a9: div-float<br/> |
| 61 | aa: rem-float<br/> |
| 62 | ab: add-double<br/> |
| 63 | ac: sub-double<br/> |
| 64 | ad: mul-double<br/> |
| 65 | ae: div-double<br/> |
| 66 | af: rem-double |
| 67 | </td> |
| 68 | <td><code>A:</code> destination register or pair (8 bits)<br/> |
| 69 | <code>B:</code> first source register or pair (8 bits)<br/> |
| 70 | <code>C:</code> second source register or pair (8 bits)</td> |
| 71 | </tr> |
| 72 | </tbody> |
| 73 | </table> |
| 74 | |
| 75 | <h2>Constraints</h2> |
| 76 | |
| 77 | <ul> |
| 78 | <li> |
| 79 | All A, B and C must be valid register indices in the current stackframe. |
| 80 | </li> |
| 81 | <li> |
| 82 | For the -long and -double variants, also A+1, B+1 and C+1 must be valid |
| 83 | register indices. |
| 84 | </li> |
| 85 | <li> |
| 86 | Registers vB and vC must be defined. They must both contain values that |
| 87 | match the variant of the instruction (it's probably better to split this up |
| 88 | into multiple pages again). |
| 89 | </li> |
| 90 | </ul> |
| 91 | |
| 92 | <h2>Behavior</h2> |
| 93 | |
| 94 | <ul> |
| 95 | <li> |
| 96 | The given operation <binop> is performed according to the semantics |
| 97 | specified in table XXX. |
| 98 | </li> |
| 99 | <li> |
| 100 | The result is stored in register vA, that is, vA'=<biop> vB. |
| 101 | </li> |
| 102 | <li> |
| 103 | For the -double and -long variants, (vA+1) is also affected. |
| 104 | </li> |
| 105 | <li> |
| 106 | As usual, neighboring registers might get undefined, if vA (and vA+1) were |
| 107 | part of a register pair originally. |
| 108 | </li> |
| 109 | </ul> |
| 110 | |
| 111 | <h2>Exceptions</h2> |
| 112 | |
| 113 | <ul> |
| 114 | <li> |
| 115 | ArithmeticException if an error occurs during the instruction. |
| 116 | </li> |
| 117 | </ul> |
| 118 | |
| 119 | </body> |
| 120 | </html> |