Skip to content

Optimize color conversions in non-JIT vertex decoder #19241

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
hrydgard merged 4 commits into from
Jun 4, 2024
Merged

Optimize color conversions in non-JIT vertex decoder #19241

Changes from 1 commit
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
f2b4ad0
Write new 16-bit to 32-bit color conversion utility functions
hrydgard Jun 4, 2024
5b14cf9
Switch to the new functions.
hrydgard Jun 4, 2024
9ac7054
Vertex decoder (non-JIT): Optimize 16-bit color decoders.
hrydgard Jun 4, 2024
2174a89
Clarify operator precedence
hrydgard Jun 4, 2024
File filter

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Prev Previous commit
Next Next commit
Vertex decoder (non-JIT): Optimize 16-bit color decoders.
  • Loading branch information
@hrydgard
hrydgard committed Jun 4, 2024
commit 9ac7054b01a5db99ec70295cfb827d6557411991
33 changes: 16 additions & 17 deletions GPU/Common/VertexDecoderCommon.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -506,40 +506,39 @@ void VertexDecoder::Step_ColorInvalid() const

void VertexDecoder::Step_Color565() const
{
u8 *c = decoded_ + decFmt.c0off;
u16 cdata = *(const u16_le *)(ptr_ + coloff);
c[0] = Convert5To8(cdata & 0x1f);
c[1] = Convert6To8((cdata >> 5) & 0x3f);
c[2] = Convert5To8((cdata >> 11) & 0x1f);
c[3] = 255;
// Always full alpha.
u32 *c = (u32 *)(decoded_ + decFmt.c0off);
*c = RGB565ToRGBA8888(cdata);
}

void VertexDecoder::Step_Color5551() const
{
u8 *c = decoded_ + decFmt.c0off;
u16 cdata = *(const u16_le *)(ptr_ + coloff);
gstate_c.vertexFullAlpha = gstate_c.vertexFullAlpha && (cdata >> 15) != 0;
c[0] = Convert5To8(cdata & 0x1f);
c[1] = Convert5To8((cdata >> 5) & 0x1f);
c[2] = Convert5To8((cdata >> 10) & 0x1f);
c[3] = (cdata >> 15) ? 255 : 0;
u32 *c = (u32 *)(decoded_ + decFmt.c0off);
int alpha = (cdata >> 15);
if (!alpha) {
gstate_c.vertexFullAlpha = false;
}
*c = RGBA5551ToRGBA8888(cdata);
}

void VertexDecoder::Step_Color4444() const
{
u8 *c = decoded_ + decFmt.c0off;
u16 cdata = *(const u16_le *)(ptr_ + coloff);
gstate_c.vertexFullAlpha = gstate_c.vertexFullAlpha && (cdata >> 12) == 0xF;
for (int j = 0; j < 4; j++)
c[j] = Convert4To8((cdata >> (j * 4)) & 0xF);
u32 *c = (u32 *)(decoded_ + decFmt.c0off);
if ((cdata >> 12) != 0xF) {
gstate_c.vertexFullAlpha = false;
}
*c = RGBA4444ToRGBA8888(cdata);
}

void VertexDecoder::Step_Color8888() const
{
u8 *c = decoded_ + decFmt.c0off;
const u8 *cdata = (const u8*)(ptr_ + coloff);
gstate_c.vertexFullAlpha = gstate_c.vertexFullAlpha && cdata[3] == 255;
if (cdata[3] != 255) {
gstate_c.vertexFullAlpha = false;
}
memcpy(c, cdata, sizeof(u8) * 4);
}

Expand Down