ZaneYork
/
zoneminder
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Added ARM Neon (AArch32) Delta functions for 32bit color and 8bit grayscale

This commit is contained in:
Kfir Itzhak 2017-03-19 13:02:42 +02:00
parent e02e033cc8
commit c5c9b52d18
1 changed files with 90 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

90
src/zm_image.cpp
View File

@ -249,6 +249,14 @@ void Image::Initialise()
// fptr_delta8_abgr = &std_delta8_abgr;
fptr_delta8_gray8 = &sse2_delta8_gray8;
Debug(4,"Delta: Using SSE2 delta functions");
} else if(neonversion >= 1) {
/* ARM Neon available */
fptr_delta8_rgba = &neon32_armv7_delta8_rgba;
fptr_delta8_bgra = &neon32_armv7_delta8_bgra;
fptr_delta8_argb = &neon32_armv7_delta8_argb;
fptr_delta8_abgr = &neon32_armv7_delta8_abgr;
fptr_delta8_gray8 = &neon32_armv7_delta8_gray8;
Debug(4,"Delta: Using ARM Neon delta functions");
} else {
/* No suitable SSE version available */
fptr_delta8_rgba = &std_delta8_rgba;
@ -3501,6 +3509,88 @@ __attribute__((noinline)) void std_delta8_abgr(const uint8_t* col1, const uint8_
}
}
/* Grayscale Neon for AArch32 */
#if defined(__arm__)
__attribute__((noinline,__target__("fpu=neon")))
#endif
void neon32_armv7_delta8_gray8(uint8_t* col1, uint8_t* col2, uint8_t* result, unsigned long count) {
#if defined(__arm__)
/* Q0(D0,D1) = col1 */
/* Q1(D2,D3) = col2 */
__asm__ __volatile__ (
"neon32_armv7_delta8_gray8_iter:\n\t"
"vldm %0!, {q0}\n\t"
"vldm %1!, {q1}\n\t"
"vabd.u8 q0, q0, q1\n\t"
"vstm %2!, {q0}\n\t"
"subs %3, %3, #16\n\t"
"bne neon32_armv7_delta8_gray8_iter\n\t"
:
: "r" (col1), "r" (col2), "r" (result), "r" (count)
: "%q0", "%q1", "cc", "memory"
);
#else
Panic("Neon function called on a non ARM platform");
#endif
}
/* RGB32 Neon for AArch32 */
#if defined(__arm__)
__attribute__((noinline,__target__("fpu=neon")))
#endif
void neon32_armv7_delta8_rgb32(uint8_t* col1, uint8_t* col2, uint8_t* result, unsigned long count, uint32_t multiplier) {
#if defined(__arm__)
/* Q0(D0,D1) = col1 */
/* Q1(D2,D3) = col2 */
/* Q2(D4,D5) = multiplier */
__asm__ __volatile__ (
"mov r12, %4\n\t"
"vdup.32 q2, r12\n\t"
"neon32_armv7_delta8_rgb32_iter:\n\t"
"vldm %0!, {q0}\n\t"
"vldm %1!, {q1}\n\t"
"vabd.u8 q0, q0, q1\n\t"
"vrshr.u8 q0, q0, #3\n\t"
"vmul.i8 q0, q0, q2\n\t"
"vpadd.i8 d0, d0, d1\n\t"
"vpadd.i8 d2, d2, d3\n\t"
"vpadd.i8 d0, d0, d2\n\t"
"vst1.32 {d0[0]}, [%2]!\n\t"
"subs %3, %3, #4\n\t"
"bne neon32_armv7_delta8_rgb32_iter\n\t"
:
: "r" (col1), "r" (col2), "r" (result), "r" (count), "r" (multiplier)
: "%r12", "%q0", "%q1", "%q2", "cc", "memory"
);
#else
Panic("Neon function called on a non ARM platform");
#endif
}
/* RGB32: RGBA Neon for AArch32 */
void neon32_armv7_delta8_rgba(uint8_t* col1, uint8_t* col2, uint8_t* result, unsigned long count) {
neon32_armv7_delta8_rgb32(col1, col2, result, count, 0x00010502);
}
/* RGB32: BGRA Neon for AArch32 */
void neon32_armv7_delta8_bgra(uint8_t* col1, uint8_t* col2, uint8_t* result, unsigned long count) {
neon32_armv7_delta8_rgb32(col1, col2, result, count, 0x00020501);
}
/* RGB32: ARGB Neon for AArch32 */
void neon32_armv7_delta8_argb(uint8_t* col1, uint8_t* col2, uint8_t* result, unsigned long count) {
neon32_armv7_delta8_rgb32(col1, col2, result, count, 0x01050200);
}
/* RGB32: ABGR Neon for AArch32 */
void neon32_armv7_delta8_abgr(uint8_t* col1, uint8_t* col2, uint8_t* result, unsigned long count) {
neon32_armv7_delta8_rgb32(col1, col2, result, count, 0x02050100);
}
/* Grayscale SSE2 */
#if defined(__i386__) || defined(__x86_64__)
__attribute__((noinline,__target__("sse2")))