- emit_load4f_1(p, dest, arg0);
+ /* Loading from memory erases the upper bits. */
+ sse_movss(&p->func, dest, arg0);
- emit_load4f_1(p, dest, arg0);
+ /* Loading from memory erases the upper bits. */
+ sse_movss(&p->func, dest, arg0);
struct x86_reg temp = x86_make_reg(file_XMM, 0);
struct x86_reg vp0 = x86_make_reg(file_XMM, 1);
struct x86_reg vp1 = x86_make_reg(file_XMM, 2);
struct x86_reg temp = x86_make_reg(file_XMM, 0);
struct x86_reg vp0 = x86_make_reg(file_XMM, 1);
struct x86_reg vp1 = x86_make_reg(file_XMM, 2);
sse_shufps(&p->func, temp, temp, SHUF(W,X,Y,Z));
get_src_ptr(p, srcECX, vtxESI, &a[1]);
sse_shufps(&p->func, temp, temp, SHUF(W,X,Y,Z));
get_src_ptr(p, srcECX, vtxESI, &a[1]);
- emit_load(p, temp, 1, x86_deref(srcECX), a[1].inputsize);
+ emit_load(p, temp2, 1, x86_deref(srcECX), a[1].inputsize);
+ sse_movss(&p->func, temp, temp2);
update_src_ptr(p, srcECX, vtxESI, &a[1]);
/* Rearrange and possibly do BGR conversion:
update_src_ptr(p, srcECX, vtxESI, &a[1]);
/* Rearrange and possibly do BGR conversion: