2015-08-14 1:55 GMT-03:00 Ilia Mirkin <imir...@alum.mit.edu>:
> On Fri, Aug 14, 2015 at 12:52 AM, Marcos Paulo de souza
> <marcos.souza....@gmail.com> wrote:
> > Now I'll take a look about the last problem of LRP and MOV.
>
> That should ideally have solved itself too... if not, do you have the
> full shader that demonstrates the problem?
>
Yes, there it is:
TESS_CTRL
PROPERTY TCS_VERTICES_OUT 9
DCL IN[][0], POSITION
DCL SV[0], INVOCATIONID
DCL SV[1], VERTICESIN
DCL OUT[0], TESSOUTER
DCL OUT[1], TESSINNER
DCL OUT[][2], GENERIC[0]
DCL OUT[][3], GENERIC[1]
DCL TEMP[0..3], LOCAL
DCL ADDR[0..1]
IMM[0] FLT32 { 21.0000, 0.5000, 0.0000, 1.0000}
IMM[1] INT32 {3, 4, 0, 0}
0: MOV OUT[1].x, IMM[0].xxxx
1: MOV OUT[1].y, IMM[0].xxxx
2: MOV OUT[0].x, IMM[0].xxxx
3: MOV OUT[0].y, IMM[0].xxxx
4: MOV OUT[0].z, IMM[0].xxxx
5: MOV OUT[0].w, IMM[0].xxxx
6: MOD TEMP[0].x, SV[0].xxxx, IMM[1].xxxx
7: I2F TEMP[0].x, TEMP[0].xxxx
8: MUL TEMP[0].x, TEMP[0].xxxx, IMM[0].yyyy
9: IDIV TEMP[1].x, SV[0].xxxx, IMM[1].xxxx
10: I2F TEMP[1].x, TEMP[1].xxxx
11: MUL TEMP[1].x, TEMP[1].xxxx, IMM[0].yyyy
12: LRP TEMP[2], TEMP[0].xxxx, IN[1][0], IN[0][0]
13: LRP TEMP[3], TEMP[0].xxxx, IN[3][0], IN[2][0]
14: UARL ADDR[1].x, SV[0].xxxx
15: LRP OUT[ADDR[1].x][3], TEMP[1].xxxx, TEMP[3], TEMP[2]
16: USEQ TEMP[2].x, SV[1].xxxx, IMM[1].yyyy
17: UIF TEMP[2].xxxx :0
18: MOV TEMP[2].zw, IMM[0].wwzw
19: MOV TEMP[2].x, TEMP[0].xxxx
20: MOV TEMP[2].y, TEMP[1].xxxx
21: MOV TEMP[0], TEMP[2]
22: ELSE :0
23: MOV TEMP[0], IMM[0].zzzz
24: ENDIF
25: UARL ADDR[1].x, SV[0].xxxx
26: MOV OUT[ADDR[1].x][2], TEMP[0]
27: END
Can you give me some tip to fix it, or do you think this can be sent in
another patch rather than the tgsi and the nouai part?
Thanks
--
Att,
Marcos Paulo de Souza
Github: https://github.com/marcosps/
_______________________________________________
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
