[Piglit] [PATCH] arb_gpu_shader_fp64: expand get uniform even further.

[Dave Airlie]

On 12 October 2015 at 13:05, Ilia Mirkin <imirkin at alum.mit.edu> wrote:
> On Sun, Oct 11, 2015 at 11:02 PM, Dave Airlie <airlied at gmail.com> wrote:
>> From: Dave Airlie <airlied at redhat.com>
>>
>> Ilia pointed out the fix patch didn't go far enough, so add
>> some tests to cover more cases, specifically:
>>
>> a) reading from a dvec4 into a float
>> b) reading from a vec4 into doubles.
>>
>> Signed-off-by: Dave Airlie <airlied at redhat.com>
>> ---
>>  .../arb_gpu_shader_fp64/execution/getuniformdv.c   | 31 +++++++++++++++++++---
>>  1 file changed, 27 insertions(+), 4 deletions(-)
>>
>> diff --git a/tests/spec/arb_gpu_shader_fp64/execution/getuniformdv.c b/tests/spec/arb_gpu_shader_fp64/execution/getuniformdv.c
>> index 8cda7f4..5f7d85e 100644
>> --- a/tests/spec/arb_gpu_shader_fp64/execution/getuniformdv.c
>> +++ b/tests/spec/arb_gpu_shader_fp64/execution/getuniformdv.c
>> @@ -50,6 +50,7 @@ static const char vs_text[] =
>>     "uniform s1 s;\n"
>>     "uniform double d2; \n"
>>     "uniform float f1; \n"
>> +   "uniform vec4 fv1; \n"
>>     "uniform uint ui1; \n"
>>     "uniform int i1; \n"
>>     "uniform bool b1; \n"
>> @@ -59,6 +60,7 @@ static const char vs_text[] =
>>     "{\n"
>>     "  gl_Position = vec4(0.0, 0.0, 0.0, 1.0);\n"
>>     "  dvec4 t = dvec4(s.a, s.b, s.c, s.d) * d1 + d2 + double(f1) + double(ui1) + double(i1) + double(b1);\n"
>> +   "  t += fv1.x + fv1.y + fv1.z + fv1.w;\n"
>>     "  t += v[0] + v[1] + v[2]; \n"
>>     "  color = vec4(t); \n"
>>     "}\n";
>> @@ -81,8 +83,8 @@ piglit_init(int argc, char **argv)
>>  {
>>     GLuint vs, fs, prog;
>>     GLint numUniforms, i;
>> -   GLint expectedNum = 11;
>> -   GLint loc_d1, loc_d2, loc_sa, loc_sd, loc_v1, loc_f1, loc_ui1, loc_i1, loc_b1;
>> +   GLint expectedNum = 12;
>> +   GLint loc_d1, loc_d2, loc_sa, loc_sd, loc_v1, loc_f1, loc_fv1, loc_ui1, loc_i1, loc_b1;
>>     GLdouble v[4];
>>     GLfloat f[4];
>>     GLuint ui[4];
>> @@ -136,8 +138,11 @@ piglit_init(int argc, char **argv)
>>           expectedType = GL_DOUBLE;
>>           expectedSize = 1;
>>        } else if (name[0] == 'f') {
>> -         expectedType = GL_FLOAT;
>> -         expectedSize = 1;
>> +       if (name[1] == 'v')
>> +         expectedType = GL_FLOAT_VEC4;
>> +       else
>> +         expectedType = GL_FLOAT;
>> +       expectedSize = 1;
>>        } else if (name[0] == 'i') {
>>           expectedType = GL_INT;
>>           expectedSize = 1;
>> @@ -170,6 +175,7 @@ piglit_init(int argc, char **argv)
>>     loc_sd = glGetUniformLocation(prog, "s.d");
>>     loc_v1 = glGetUniformLocation(prog, "v[1]");
>>     loc_f1 = glGetUniformLocation(prog, "f1");
>> +   loc_fv1 = glGetUniformLocation(prog, "fv1");
>>     loc_ui1 = glGetUniformLocation(prog, "ui1");
>>     loc_i1 = glGetUniformLocation(prog, "i1");
>>     loc_b1 = glGetUniformLocation(prog, "b1");
>> @@ -179,6 +185,7 @@ piglit_init(int argc, char **argv)
>>     glUniform1d(loc_sd, 20.0);
>>     glUniform4dv(loc_v1, 1, vVals);
>>     glUniform1f(loc_f1, 40.0);
>> +   glUniform4f(loc_fv1, 40.0, 30.0, 20.0, 10.0);
>>     glUniform1ui(loc_ui1, 100);
>>     glUniform1i(loc_i1, -100);
>>
>> @@ -249,6 +256,15 @@ piglit_init(int argc, char **argv)
>>        piglit_report_result(PIGLIT_FAIL);
>>     }
>>
>> +   glGetUniformfv(prog, loc_v1, f);
>
> loc_fv1 right?

No we want to read the dvec4 at loc_v1 into floats here.

So loc_v1 is correct.
Dave.