qcms 0.1.0

lightweight color management
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267

use crate::transform::{qcms_transform, Format, BGRA, CLAMPMAXVAL, FLOATSCALE, RGB, RGBA};
use ::libc;
#[cfg(target_arch = "aarch64")]
use core::arch::aarch64::{float32x4_t, int32x4_t, vaddq_f32};
#[cfg(target_arch = "arm")]
use core::arch::arm::{float32x4_t, int32x4_t, vaddq_f32};
use std::mem::zeroed;

static mut floatScale: f32 = FLOATSCALE;
static mut clampMaxValue: f32 = CLAMPMAXVAL;

#[target_feature(enable = "neon")]
#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
unsafe extern "C" fn qcms_transform_data_template_lut_neon<F: Format>(
    mut transform: *const qcms_transform,
    mut src: *const libc::c_uchar,
    mut dest: *mut libc::c_uchar,
    mut length: usize,
) {
    let mut mat: *const [f32; 4] = (*transform).matrix.as_ptr();
    let mut input_back: [libc::c_char; 32] = [0; 32];
    /* Ensure we have a buffer that's 16 byte aligned regardless of the original
     * stack alignment. We can't use __attribute__((aligned(16))) or __declspec(align(32))
     * because they don't work on stack variables. gcc 4.4 does do the right thing
     * on x86 but that's too new for us right now. For more info: gcc bug #16660 */
    let mut input: *const f32 = (&mut *input_back.as_mut_ptr().offset(16isize) as *mut libc::c_char
        as usize
        & !(0xf) as usize) as *mut f32;
    /* share input and output locations to save having to keep the
     * locations in separate registers */
    let mut output: *const u32 = input as *mut u32;
    /* deref *transform now to avoid it in loop */
    let mut igtbl_r: *const f32 = (*transform).input_gamma_table_r.as_ref().unwrap().as_ptr();
    let mut igtbl_g: *const f32 = (*transform).input_gamma_table_g.as_ref().unwrap().as_ptr();
    let mut igtbl_b: *const f32 = (*transform).input_gamma_table_b.as_ref().unwrap().as_ptr();
    /* deref *transform now to avoid it in loop */
    let mut otdata_r: *const u8 = (*transform)
        .output_table_r
        .as_deref()
        .unwrap()
        .data
        .as_ptr();
    let mut otdata_g: *const u8 = (*transform)
        .output_table_g
        .as_deref()
        .unwrap()
        .data
        .as_ptr();
    let mut otdata_b: *const u8 = (*transform)
        .output_table_b
        .as_deref()
        .unwrap()
        .data
        .as_ptr();
    /* input matrix values never change */
    let mat0: float32x4_t = vld1q_f32((*mat.offset(0isize)).as_ptr());
    let mat1: float32x4_t = vld1q_f32((*mat.offset(1isize)).as_ptr());
    let mat2: float32x4_t = vld1q_f32((*mat.offset(2isize)).as_ptr());
    /* these values don't change, either */
    let max: float32x4_t = vld1q_dup_f32(&clampMaxValue);
    let min: float32x4_t = zeroed();
    let scale: float32x4_t = vld1q_dup_f32(&floatScale);
    let components: libc::c_uint = if F::kAIndex == 0xff { 3 } else { 4 } as libc::c_uint;
    /* working variables */
    let mut vec_r: float32x4_t;
    let mut vec_g: float32x4_t;
    let mut vec_b: float32x4_t;
    let mut result: int32x4_t;
    let mut alpha: libc::c_uchar = 0;
    /* CYA */
    if length == 0 {
        return;
    }
    /* one pixel is handled outside of the loop */
    length = length.wrapping_sub(1);
    /* setup for transforming 1st pixel */
    vec_r = vld1q_dup_f32(&*igtbl_r.offset(*src.offset(F::kRIndex as isize) as isize));
    vec_g = vld1q_dup_f32(&*igtbl_g.offset(*src.offset(F::kGIndex as isize) as isize));
    vec_b = vld1q_dup_f32(&*igtbl_b.offset(*src.offset(F::kBIndex as isize) as isize));
    if F::kAIndex != 0xff {
        alpha = *src.offset(F::kAIndex as isize)
    }
    src = src.offset(components as isize);
    let mut i: libc::c_uint = 0;
    while (i as usize) < length {
        /* gamma * matrix */
        vec_r = vmulq_f32(vec_r, mat0);
        vec_g = vmulq_f32(vec_g, mat1);
        vec_b = vmulq_f32(vec_b, mat2);
        /* store alpha for this pixel; load alpha for next */
        if F::kAIndex != 0xff {
            *dest.offset(F::kAIndex as isize) = alpha;
            alpha = *src.offset(F::kAIndex as isize)
        }
        /* crunch, crunch, crunch */
        vec_r = vaddq_f32(vec_r, vaddq_f32(vec_g, vec_b));
        vec_r = vmaxq_f32(min, vec_r);
        vec_r = vminq_f32(max, vec_r);
        result = vcvtq_s32_f32(vmulq_f32(vec_r, scale));

        /* use calc'd indices to output RGB values */
        *dest.offset(F::kRIndex as isize) = *otdata_r.offset(vgetq_lane_s32(result, 0) as isize);
        *dest.offset(F::kGIndex as isize) = *otdata_g.offset(vgetq_lane_s32(result, 1) as isize);
        *dest.offset(F::kBIndex as isize) = *otdata_b.offset(vgetq_lane_s32(result, 2) as isize);

        /* load gamma values for next loop while store completes */
        vec_r = vld1q_dup_f32(&*igtbl_r.offset(*src.offset(F::kRIndex as isize) as isize));
        vec_g = vld1q_dup_f32(&*igtbl_g.offset(*src.offset(F::kGIndex as isize) as isize));
        vec_b = vld1q_dup_f32(&*igtbl_b.offset(*src.offset(F::kBIndex as isize) as isize));

        dest = dest.offset(components as isize);
        src = src.offset(components as isize);
        i = i.wrapping_add(1)
    }
    /* handle final (maybe only) pixel */
    vec_r = vmulq_f32(vec_r, mat0);
    vec_g = vmulq_f32(vec_g, mat1);
    vec_b = vmulq_f32(vec_b, mat2);
    if F::kAIndex != 0xff {
        *dest.offset(F::kAIndex as isize) = alpha
    }
    vec_r = vaddq_f32(vec_r, vaddq_f32(vec_g, vec_b));
    vec_r = vmaxq_f32(min, vec_r);
    vec_r = vminq_f32(max, vec_r);
    result = vcvtq_s32_f32(vmulq_f32(vec_r, scale));

    *dest.offset(F::kRIndex as isize) = *otdata_r.offset(vgetq_lane_s32(result, 0) as isize);
    *dest.offset(F::kGIndex as isize) = *otdata_g.offset(vgetq_lane_s32(result, 1) as isize);
    *dest.offset(F::kBIndex as isize) = *otdata_b.offset(vgetq_lane_s32(result, 2) as isize);
}
#[no_mangle]
#[target_feature(enable = "neon")]
#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
pub unsafe extern "C" fn qcms_transform_data_rgb_out_lut_neon(
    mut transform: *const qcms_transform,
    mut src: *const libc::c_uchar,
    mut dest: *mut libc::c_uchar,
    mut length: usize,
) {
    qcms_transform_data_template_lut_neon::<RGB>(transform, src, dest, length);
}
#[no_mangle]
#[target_feature(enable = "neon")]
#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
pub unsafe extern "C" fn qcms_transform_data_rgba_out_lut_neon(
    mut transform: *const qcms_transform,
    mut src: *const libc::c_uchar,
    mut dest: *mut libc::c_uchar,
    mut length: usize,
) {
    qcms_transform_data_template_lut_neon::<RGBA>(transform, src, dest, length);
}

#[no_mangle]
#[target_feature(enable = "neon")]
#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
pub unsafe extern "C" fn qcms_transform_data_bgra_out_lut_neon(
    mut transform: *const qcms_transform,
    mut src: *const libc::c_uchar,
    mut dest: *mut libc::c_uchar,
    mut length: usize,
) {
    qcms_transform_data_template_lut_neon::<BGRA>(transform, src, dest, length);
}

use std::mem::transmute;

#[inline]
#[target_feature(enable = "neon")]
#[cfg(target_arch = "aarch64")]
pub unsafe fn vld1q_f32(addr: *const f32) -> float32x4_t {
    transmute([*addr, *addr.offset(1), *addr.offset(2), *addr.offset(3)])
}

#[inline]
#[cfg(target_arch = "arm")]
#[target_feature(enable = "neon")]
#[target_feature(enable = "v7")]
pub unsafe fn vld1q_f32(addr: *const f32) -> float32x4_t {
    vld1q_v4f32(addr as *const u8, 4)
}

#[cfg(target_arch = "arm")]
#[allow(improper_ctypes)]
extern "C" {
    #[cfg_attr(target_arch = "arm", link_name = "llvm.arm.neon.vld1.v4f32.p0i8")]
    fn vld1q_v4f32(addr: *const u8, align: u32) -> float32x4_t;
}

#[cfg(target_arch = "aarch64")]
#[allow(improper_ctypes)]
extern "C" {
    #[link_name = "llvm.aarch64.neon.fcvtzs.v4.v4f32"]
    fn vcvtq_s32_f32_(a: float32x4_t) -> int32x4_t;
}

#[allow(improper_ctypes)]
extern "C" {
    #[cfg_attr(target_arch = "arm", link_name = "llvm.arm.neon.vmaxs.v4f32")]
    #[cfg_attr(target_arch = "aarch64", link_name = "llvm.aarch64.neon.fmax.v4f32")]
    fn vmaxq_f32_(a: float32x4_t, b: float32x4_t) -> float32x4_t;
    #[cfg_attr(target_arch = "arm", link_name = "llvm.arm.neon.vmins.v4f32")]
    #[cfg_attr(target_arch = "aarch64", link_name = "llvm.aarch64.neon.fmin.v4f32")]
    fn vminq_f32_(a: float32x4_t, b: float32x4_t) -> float32x4_t;
}

/// Move vector element to general-purpose register
#[inline]
#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
pub unsafe fn vgetq_lane_s32(v: int32x4_t, imm5: i32) -> i32 {
    assert!(imm5 >= 0 && imm5 <= 3);
    simd_extract(v, imm5 as u32)
}

/// Multiply
#[inline]
#[target_feature(enable = "neon")]
#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
pub unsafe fn vmulq_f32(a: float32x4_t, b: float32x4_t) -> float32x4_t {
    simd_mul(a, b)
}

/// Floating-point minimum (vector).
#[inline]
#[target_feature(enable = "neon")]
#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
pub unsafe fn vminq_f32(a: float32x4_t, b: float32x4_t) -> float32x4_t {
    vminq_f32_(a, b)
}

/// Floating-point maxmimum (vector).
#[inline]
#[target_feature(enable = "neon")]
#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
pub unsafe fn vmaxq_f32(a: float32x4_t, b: float32x4_t) -> float32x4_t {
    vmaxq_f32_(a, b)
}

#[inline]
#[cfg(target_arch = "aarch64")]
#[target_feature(enable = "neon")]
pub unsafe fn vcvtq_s32_f32(a: float32x4_t) -> int32x4_t {
    vcvtq_s32_f32_(a)
}
/// Floating-point Convert to Signed fixed-point, rounding toward Zero (vector)
#[inline]
#[cfg(target_arch = "arm")]
#[target_feature(enable = "neon")]
#[target_feature(enable = "v7")]
pub unsafe fn vcvtq_s32_f32(a: float32x4_t) -> int32x4_t {
    simd_cast::<_, int32x4_t>(a)
}

/// Load one single-element structure and Replicate to all lanes (of one register).
#[inline]
#[target_feature(enable = "neon")]
#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
pub unsafe fn vld1q_dup_f32(addr: *const f32) -> float32x4_t {
    let v = *addr;
    transmute([v, v, v, v])
}

extern "platform-intrinsic" {
    pub fn simd_mul<T>(x: T, y: T) -> T;
    pub fn simd_extract<T, U>(x: T, idx: u32) -> U;
    pub fn simd_cast<T, U>(x: T) -> U;
}