zdoom/qzdoom
0
0
Fork 0
mirror of https://github.com/ZDoom/qzdoom.git synced 2024-11-11 15:22:16 +00:00
Code Issues Projects Releases Packages Wiki Activity

Make wall drawers thread aware

Browse source
This commit is contained in:
Magnus Norddahl 2016-12-05 13:05:05 +01:00
parent d8a3174330
commit 246d1625e5
2 changed files with 163 additions and 16 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

176
src/r_draw_pal.cpp
View file

@ -102,6 +102,15 @@ namespace swrenderer
int bits = _vlinebits;
int pitch = _pitch;
count = thread->count_for_thread(_dest_y, count);
if (count <= 0)
return;
dest = thread->dest_for_thread(_dest_y, pitch, dest);
frac += fracstep * thread->skipped_by_thread(_dest_y);
fracstep *= thread->num_cores;
pitch *= thread->num_cores;
do
{
*dest = colormap[source[frac >> bits]];
@ -124,16 +133,32 @@ namespace swrenderer
auto buf1 = _bufplce[1];
auto buf2 = _bufplce[2];
auto buf3 = _bufplce[3];
const auto vince0 = _vince[0];
const auto vince1 = _vince[1];
const auto vince2 = _vince[2];
const auto vince3 = _vince[3];
auto vince0 = _vince[0];
auto vince1 = _vince[1];
auto vince2 = _vince[2];
auto vince3 = _vince[3];
auto vplce0 = _vplce[0];
auto vplce1 = _vplce[1];
auto vplce2 = _vplce[2];
auto vplce3 = _vplce[3];
auto pitch = _pitch;
count = thread->count_for_thread(_dest_y, count);
if (count <= 0)
return;
int skipped = thread->skipped_by_thread(_dest_y);
dest = thread->dest_for_thread(_dest_y, pitch, dest);
vplce0 += vince0 * skipped;
vplce1 += vince1 * skipped;
vplce2 += vince2 * skipped;
vplce3 += vince3 * skipped;
vince0 *= thread->num_cores;
vince1 *= thread->num_cores;
vince2 *= thread->num_cores;
vince3 *= thread->num_cores;
pitch *= thread->num_cores;
do
{
dest[0] = pal0[buf0[(place = vplce0) >> bits]]; vplce0 = place + vince0;
@ -155,6 +180,15 @@ namespace swrenderer
int bits = _mvlinebits;
int pitch = _pitch;
count = thread->count_for_thread(_dest_y, count);
if (count <= 0)
return;
dest = thread->dest_for_thread(_dest_y, pitch, dest);
frac += fracstep * thread->skipped_by_thread(_dest_y);
fracstep *= thread->num_cores;
pitch *= thread->num_cores;
do
{
uint8_t pix = source[frac >> bits];
@ -181,16 +215,32 @@ namespace swrenderer
auto buf1 = _bufplce[1];
auto buf2 = _bufplce[2];
auto buf3 = _bufplce[3];
const auto vince0 = _vince[0];
const auto vince1 = _vince[1];
const auto vince2 = _vince[2];
const auto vince3 = _vince[3];
auto vince0 = _vince[0];
auto vince1 = _vince[1];
auto vince2 = _vince[2];
auto vince3 = _vince[3];
auto vplce0 = _vplce[0];
auto vplce1 = _vplce[1];
auto vplce2 = _vplce[2];
auto vplce3 = _vplce[3];
auto pitch = _pitch;
count = thread->count_for_thread(_dest_y, count);
if (count <= 0)
return;
int skipped = thread->skipped_by_thread(_dest_y);
dest = thread->dest_for_thread(_dest_y, pitch, dest);
vplce0 += vince0 * skipped;
vplce1 += vince1 * skipped;
vplce2 += vince2 * skipped;
vplce3 += vince3 * skipped;
vince0 *= thread->num_cores;
vince1 *= thread->num_cores;
vince2 *= thread->num_cores;
vince3 *= thread->num_cores;
pitch *= thread->num_cores;
do
{
uint8_t pix;
@ -217,6 +267,15 @@ namespace swrenderer
uint32_t *fg2rgb = _srcblend;
uint32_t *bg2rgb = _destblend;
count = thread->count_for_thread(_dest_y, count);
if (count <= 0)
return;
dest = thread->dest_for_thread(_dest_y, pitch, dest);
frac += fracstep * thread->skipped_by_thread(_dest_y);
fracstep *= thread->num_cores;
pitch *= thread->num_cores;
do
{
uint8_t pix = source[frac >> bits];
@ -242,6 +301,21 @@ namespace swrenderer
uint32_t *bg2rgb = _destblend;
uint32_t vplce[4] = { _vplce[0], _vplce[1], _vplce[2], _vplce[3] };
uint32_t vince[4] = { _vince[0], _vince[1], _vince[2], _vince[3] };
count = thread->count_for_thread(_dest_y, count);
if (count <= 0)
return;
int pitch = _pitch;
int skipped = thread->skipped_by_thread(_dest_y);
dest = thread->dest_for_thread(_dest_y, pitch, dest);
for (int i = 0; i < 4; i++)
{
vplce[i] += vince[i] * skipped;
vince[i] *= thread->num_cores;
}
pitch *= thread->num_cores;
do
{
@ -255,9 +329,9 @@ namespace swrenderer
fg = (fg + bg) | 0x1f07c1f;
dest[i] = RGB32k.All[fg & (fg >> 15)];
}
vplce[i] += _vince[i];
vplce[i] += vince[i];
}
dest += _pitch;
dest += pitch;
} while (--count);
}
@ -275,6 +349,15 @@ namespace swrenderer
uint32_t *fg2rgb = _srcblend;
uint32_t *bg2rgb = _destblend;
count = thread->count_for_thread(_dest_y, count);
if (count <= 0)
return;
dest = thread->dest_for_thread(_dest_y, pitch, dest);
frac += fracstep * thread->skipped_by_thread(_dest_y);
fracstep *= thread->num_cores;
pitch *= thread->num_cores;
do
{
uint8_t pix = source[frac >> bits];
@ -305,6 +388,21 @@ namespace swrenderer
uint32_t *bg2rgb = _destblend;
uint32_t vplce[4] = { _vplce[0], _vplce[1], _vplce[2], _vplce[3] };
uint32_t vince[4] = { _vince[0], _vince[1], _vince[2], _vince[3] };
count = thread->count_for_thread(_dest_y, count);
if (count <= 0)
return;
int pitch = _pitch;
int skipped = thread->skipped_by_thread(_dest_y);
dest = thread->dest_for_thread(_dest_y, pitch, dest);
for (int i = 0; i < 4; i++)
{
vplce[i] += vince[i] * skipped;
vince[i] *= thread->num_cores;
}
pitch *= thread->num_cores;
do
{
@ -323,9 +421,9 @@ namespace swrenderer
a |= b;
dest[i] = RGB32k.All[a & (a >> 15)];
}
vplce[i] += _vince[i];
vplce[i] += vince[i];
}
dest += _pitch;
dest += pitch;
} while (--count);
}
@ -343,6 +441,15 @@ namespace swrenderer
uint32_t *fg2rgb = _srcblend;
uint32_t *bg2rgb = _destblend;
count = thread->count_for_thread(_dest_y, count);
if (count <= 0)
return;
dest = thread->dest_for_thread(_dest_y, pitch, dest);
frac += fracstep * thread->skipped_by_thread(_dest_y);
fracstep *= thread->num_cores;
pitch *= thread->num_cores;
do
{
uint8_t pix = source[frac >> bits];
@ -372,6 +479,21 @@ namespace swrenderer
uint32_t *bg2rgb = _destblend;
uint32_t vplce[4] = { _vplce[0], _vplce[1], _vplce[2], _vplce[3] };
uint32_t vince[4] = { _vince[0], _vince[1], _vince[2], _vince[3] };
count = thread->count_for_thread(_dest_y, count);
if (count <= 0)
return;
int pitch = _pitch;
int skipped = thread->skipped_by_thread(_dest_y);
dest = thread->dest_for_thread(_dest_y, pitch, dest);
for (int i = 0; i < 4; i++)
{
vplce[i] += vince[i] * skipped;
vince[i] *= thread->num_cores;
}
pitch *= thread->num_cores;
do
{
@ -389,9 +511,9 @@ namespace swrenderer
a |= 0x01f07c1f;
dest[i] = RGB32k.All[a & (a >> 15)];
}
vplce[i] += _vince[i];
vplce[i] += vince[i];
}
dest += _pitch;
dest += pitch;
} while (--count);
}
@ -409,6 +531,15 @@ namespace swrenderer
uint32_t *fg2rgb = _srcblend;
uint32_t *bg2rgb = _destblend;
count = thread->count_for_thread(_dest_y, count);
if (count <= 0)
return;
dest = thread->dest_for_thread(_dest_y, pitch, dest);
frac += fracstep * thread->skipped_by_thread(_dest_y);
fracstep *= thread->num_cores;
pitch *= thread->num_cores;
do
{
uint8_t pix = source[frac >> bits];
@ -438,6 +569,21 @@ namespace swrenderer
uint32_t *bg2rgb = _destblend;
uint32_t vplce[4] = { _vplce[0], _vplce[1], _vplce[2], _vplce[3] };
uint32_t vince[4] = { _vince[0], _vince[1], _vince[2], _vince[3] };
count = thread->count_for_thread(_dest_y, count);
if (count <= 0)
return;
int pitch = _pitch;
int skipped = thread->skipped_by_thread(_dest_y);
dest = thread->dest_for_thread(_dest_y, pitch, dest);
for (int i = 0; i < 4; i++)
{
vplce[i] += vince[i] * skipped;
vince[i] *= thread->num_cores;
}
pitch *= thread->num_cores;
do
{
@ -455,7 +601,7 @@ namespace swrenderer
a |= 0x01f07c1f;
dest[i] = RGB32k.All[a & (a >> 15)];
}
vplce[i] += _vince[i];
vplce[i] += vince[i];
}
dest += _pitch;
} while (--count);

3
src/r_thread.h
View file

@ -84,7 +84,8 @@ public:
}
// Calculate the dest address for the first line to be rendered by this thread
uint32_t *dest_for_thread(int first_line, int pitch, uint32_t *dest)
template<typename T>
T *dest_for_thread(int first_line, int pitch, T *dest)
{
return dest + skipped_by_thread(first_line) * pitch;
}