DevKit7/libctru
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

SHDR : various fixes

Browse Source
This commit is contained in:
smea 2014-03-11 19:00:44 +01:00
parent 1cab4c80a9
commit 656cb1a923
2 changed files with 49 additions and 71 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

43
gpu/source/test.vsh
View File

@ -1,43 +0,0 @@
; setup constants
.const 5, 0.0, 1.0, 2.0, 3.0
; setup outmap
.out o0, result.position
.out o1, result.color
.out o2, result.texcoord0
.out o3, result.texcoord1
.out o4, result.texcoord2
; setup uniform map (not required)
.uniform 0x10, 0x13, mdlvMtx
.uniform 0x14, 0x17, projMtx
;code
main:
; result.pos = mdlvMtx * in.pos
dp4 d40, d40, d00 (0x0)
dp4 d40, d41, d00 (0x1)
dp4 d40, d42, d00 (0x2)
mov d40, d25 (0x4)
; result.pos = projMtx * in.pos
dp4 d00, d44, d40 (0x0)
dp4 d00, d45, d40 (0x1)
dp4 d00, d46, d40 (0x2)
dp4 d00, d47, d40 (0x3)
; result.color = in.pos
mov d04, d25 (0x5)
; result.texcoord = const
mov d08, d25 (0x5)
mov d0C, d25 (0x5)
mov d10, d25 (0x5)
flush
end
endmain:
;operand descriptors
.opdesc x___, xyzw, xyzw ; 0x0
.opdesc _y__, xyzw, xyzw ; 0x1
.opdesc __z_, xyzw, xyzw ; 0x2
.opdesc ___w, xyzw, xyzw ; 0x3
.opdesc ___w, yyyy, xyzw ; 0x4
.opdesc xyzw, xyzw, xyzw ; 0x5

77
libctru/source/SHDR.c
View File

@ -15,42 +15,55 @@ DVLB_s* SHDR_ParseSHBIN(u32* shbinData, u32 shbinSize)
DVLB_s* ret=malloc(sizeof(DVLB_s)); DVLB_s* ret=malloc(sizeof(DVLB_s));
if(!ret)goto exit; if(!ret)goto exit;
// //parse DVLB //parse DVLB
// ret->numDVLE=shbinData[1]; ret->numDVLE=shbinData[1];
// ret->DVLE=malloc(sizeof(DVLE_s)*ret->numDVLE); ret->DVLE=malloc(sizeof(DVLE_s)*ret->numDVLE);
// if(!ret->DVLE)goto clean1; if(!ret->DVLE)goto clean1;
// //parse DVLP //parse DVLP
// u32* dvlpData=&shbinData[2+ret->numDVLE]; u32* dvlpData=&shbinData[2+ret->numDVLE];
// ret->DVLP.codeSize=dvlpData[3]; ret->DVLP.codeSize=dvlpData[3];
// ret->DVLP.codeData=&dvlpData[dvlpData[2]/4]; ret->DVLP.codeData=&dvlpData[dvlpData[2]/4];
// ret->DVLP.opdescSize=dvlpData[5]; ret->DVLP.opdescSize=dvlpData[5];
// ret->DVLP.opcdescData=&dvlpData[dvlpData[4]/4]; ret->DVLP.opcdescData=&dvlpData[dvlpData[4]/4];
// //parse DVLE //parse DVLE
// int i; int i;
// for(i=0;i<ret->numDVLE;i++) for(i=0;i<ret->numDVLE;i++)
// { {
// DVLE_s* dvle=&ret->DVLE[i]; DVLE_s* dvle=&ret->DVLE[i];
// u32* dvleData=&shbinData[shbinData[2+i]/4]; u32* dvleData=&shbinData[shbinData[2+i]/4];
// dvle->type=(dvleData[1]>>16)&0xFF; dvle->type=(dvleData[1]>>16)&0xFF;
// dvle->mainOffset=dvleData[2]; dvle->mainOffset=dvleData[2];
// dvle->endmainOffset=dvleData[3]; dvle->endmainOffset=dvleData[3];
// dvle->constTableSize=dvleData[7]; dvle->constTableSize=dvleData[7];
// dvle->constTableData=(DVLE_constEntry_s*)&dvleData[dvleData[6]/4]; dvle->constTableData=(DVLE_constEntry_s*)&dvleData[dvleData[6]/4];
// dvle->outTableSize=dvleData[11]; dvle->outTableSize=dvleData[11];
// dvle->outTableData=(DVLE_outEntry_s*)&dvleData[dvleData[10]/4]; dvle->outTableData=(DVLE_outEntry_s*)&dvleData[dvleData[10]/4];
// } }
goto exit;
clean1: clean1:
free(ret); free(ret);
exit: exit:
return ret; return ret;
} }
//hm
static inline minu8(u8 a, u8 b)
{
if(a<b)return a;
return b;
}
static inline maxu8(u8 a, u8 b)
{
if(a<b)return b;
return a;
}
void DVLP_SendCode(DVLP_s* dvlp) void DVLP_SendCode(DVLP_s* dvlp)
{ {
if(!dvlp)return; if(!dvlp)return;
@ -58,7 +71,8 @@ void DVLP_SendCode(DVLP_s* dvlp)
GPUCMD_AddSingleParam(0x000F02CB, 0x00000000); GPUCMD_AddSingleParam(0x000F02CB, 0x00000000);
int i; int i;
for(i=0;i<dvlp->codeSize;i+=0x80)GPUCMD_Add(0x000F02CC, &dvlp->codeData[i], ((dvlp->codeSize-i)<0x80)?(dvlp->codeSize-i):0x80); // for(i=0;i<dvlp->codeSize;i+=0x80)GPUCMD_Add(0x000F02CC, &dvlp->codeData[i], ((dvlp->codeSize-i)<0x80)?(dvlp->codeSize-i):0x80);
for(i=0;i<dvlp->codeSize;i+=0x80)GPUCMD_Add(0x000F02CC, &dvlp->codeData[i], maxu8(minu8(dvlp->codeSize-i,0x80),0x37)); //not sure why, but anything smaller than 0x37 seems to break stuff atm...
GPUCMD_AddSingleParam(0x000F02BF, 0x00000001); GPUCMD_AddSingleParam(0x000F02BF, 0x00000001);
} }
@ -69,17 +83,24 @@ void DVLP_SendOpDesc(DVLP_s* dvlp)
GPUCMD_AddSingleParam(0x000F02D5, 0x00000000); GPUCMD_AddSingleParam(0x000F02D5, 0x00000000);
GPUCMD_Add(0x000F02D6, dvlp->opcdescData, dvlp->opdescSize); u32 param[0x20];
int i;
//TODO : should probably preprocess this
for(i=0;i<dvlp->opdescSize;i++)param[i]=dvlp->opcdescData[i*2];
GPUCMD_Add(0x000F02D6, param, dvlp->opdescSize);
} }
void DVLE_SendOutmap(DVLE_s* dvle) void DVLE_SendOutmap(DVLE_s* dvle)
{ {
if(!dvle)return; if(!dvle)return;
u32 param[0x8]={0x1F1F1F1F,0x1F1F1F1F,0x1F1F1F1F,0x1F1F1F1F, u32 param[0x7]={0x1F1F1F1F,0x1F1F1F1F,0x1F1F1F1F,0x1F1F1F1F,
0x1F1F1F1F,0x1F1F1F1F,0x1F1F1F1F,0x1F1F1F1F}; 0x1F1F1F1F,0x1F1F1F1F,0x1F1F1F1F};
int i; int i;
//TODO : should probably preprocess this
for(i=0;i<dvle->outTableSize;i++) for(i=0;i<dvle->outTableSize;i++)
{ {
u32* out=&param[dvle->outTableData[i].regID]; u32* out=&param[dvle->outTableData[i].regID];