Home | History | Annotate | Download | only in libpixelflinger
      1 /* libs/pixelflinger/col32cb16blend.S
      2 **
      3 ** (C) COPYRIGHT 2009 ARM Limited.
      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 
     19     .text
     20     .align
     21 
     22     .global scanline_col32cb16blend_arm
     23 
     24 //
     25 // This function alpha blends a fixed color into a destination scanline, using
     26 // the formula:
     27 //
     28 //     d = s + (((a + (a >> 7)) * d) >> 8)
     29 //
     30 // where d is the destination pixel,
     31 //       s is the source color,
     32 //       a is the alpha channel of the source color.
     33 //
     34 
     35 // r0 = destination buffer pointer
     36 // r1 = color value
     37 // r2 = count
     38 
     39 
     40 scanline_col32cb16blend_arm:
     41     push        {r4-r10, lr}                    // stack ARM regs
     42 
     43     mov         r5, r1, lsr #24                 // shift down alpha
     44     mov         r9, #0xff                       // create mask
     45     add         r5, r5, r5, lsr #7              // add in top bit
     46     rsb         r5, r5, #256                    // invert alpha
     47     and         r10, r1, #0xff                  // extract red
     48     and         r12, r9, r1, lsr #8             // extract green
     49     and         r4, r9, r1, lsr #16             // extract blue
     50     mov         r10, r10, lsl #5                // prescale red
     51     mov         r12, r12, lsl #6                // prescale green
     52     mov         r4, r4, lsl #5                  // prescale blue
     53     mov         r9, r9, lsr #2                  // create dest green mask
     54 
     55 1:
     56     ldrh        r8, [r0]                        // load dest pixel
     57     subs        r2, r2, #1                      // decrement loop counter
     58     mov         r6, r8, lsr #11                 // extract dest red
     59     and         r7, r9, r8, lsr #5              // extract dest green
     60     and         r8, r8, #0x1f                   // extract dest blue
     61 
     62     smlabb      r6, r6, r5, r10                 // dest red * alpha + src red
     63     smlabb      r7, r7, r5, r12                 // dest green * alpha + src green
     64     smlabb      r8, r8, r5, r4                  // dest blue * alpha + src blue
     65 
     66     mov         r6, r6, lsr #8                  // shift down red
     67     mov         r7, r7, lsr #8                  // shift down green
     68     mov         r6, r6, lsl #11                 // shift red into 565
     69     orr         r6, r7, lsl #5                  // shift green into 565
     70     orr         r6, r8, lsr #8                  // shift blue into 565
     71 
     72     strh        r6, [r0], #2                    // store pixel to dest, update ptr
     73     bne         1b                              // if count != 0, loop
     74 
     75     pop         {r4-r10, pc}                    // return
     76 
     77 
     78 
     79