Home | History | Annotate | Download | only in libpixelflinger
      1 /* libs/pixelflinger/col32cb16blend.S
      2  *
      3  * Copyright (C) 2009 The Android Open Source Project
      4  *
      5  * Licensed under the Apache License, Version 2.0 (the "License");
      6  * you may not use this file except in compliance with the License.
      7  * You may obtain a copy of the License at
      8  *
      9  *      http://www.apache.org/licenses/LICENSE-2.0
     10  *
     11  * Unless required by applicable law or agreed to in writing, software
     12  * distributed under the License is distributed on an "AS IS" BASIS,
     13  * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
     14  * See the License for the specific language governing permissions and
     15  * limitations under the License.
     16  */
     17 
     18     .text
     19     .balign 4
     20 
     21     .global scanline_col32cb16blend_arm
     22 
     23 //
     24 // This function alpha blends a fixed color into a destination scanline, using
     25 // the formula:
     26 //
     27 //     d = s + (((a + (a >> 7)) * d) >> 8)
     28 //
     29 // where d is the destination pixel,
     30 //       s is the source color,
     31 //       a is the alpha channel of the source color.
     32 //
     33 
     34 // r0 = destination buffer pointer
     35 // r1 = color value
     36 // r2 = count
     37 
     38 
     39 scanline_col32cb16blend_arm:
     40     push        {r4-r10, lr}                    // stack ARM regs
     41 
     42     mov         r5, r1, lsr #24                 // shift down alpha
     43     mov         r9, #0xff                       // create mask
     44     add         r5, r5, r5, lsr #7              // add in top bit
     45     rsb         r5, r5, #256                    // invert alpha
     46     and         r10, r1, #0xff                  // extract red
     47     and         r12, r9, r1, lsr #8             // extract green
     48     and         r4, r9, r1, lsr #16             // extract blue
     49     mov         r10, r10, lsl #5                // prescale red
     50     mov         r12, r12, lsl #6                // prescale green
     51     mov         r4, r4, lsl #5                  // prescale blue
     52     mov         r9, r9, lsr #2                  // create dest green mask
     53 
     54 1:
     55     ldrh        r8, [r0]                        // load dest pixel
     56     subs        r2, r2, #1                      // decrement loop counter
     57     mov         r6, r8, lsr #11                 // extract dest red
     58     and         r7, r9, r8, lsr #5              // extract dest green
     59     and         r8, r8, #0x1f                   // extract dest blue
     60 
     61     smlabb      r6, r6, r5, r10                 // dest red * alpha + src red
     62     smlabb      r7, r7, r5, r12                 // dest green * alpha + src green
     63     smlabb      r8, r8, r5, r4                  // dest blue * alpha + src blue
     64 
     65     mov         r6, r6, lsr #8                  // shift down red
     66     mov         r7, r7, lsr #8                  // shift down green
     67     mov         r6, r6, lsl #11                 // shift red into 565
     68     orr         r6, r7, lsl #5                  // shift green into 565
     69     orr         r6, r8, lsr #8                  // shift blue into 565
     70 
     71     strh        r6, [r0], #2                    // store pixel to dest, update ptr
     72     bne         1b                              // if count != 0, loop
     73 
     74     pop         {r4-r10, pc}                    // return
     75 
     76 
     77 
     78