On Mobile chipsets, we often enable PipeB instead of PipeA, but the test
in here was insufficient, falling back to pipe A if the area
intersection returned zero. Therefore, in the case where a window went
off to the top of the left of the screen, it would freeze, waiting on
the wrong vblank.
Fix this mess by checking the sarea for a crtc being zero sized, and in
that case always default to the other one.
(cherry picked from commit 0b5266ff64)
intelMakeCurrent is called before intelWindowMoved (in fact, it calls
it), so calculation of the correct vblank crtc has not happened yet.
Fix this by making a function that fixes up a set of vblank flags and
call if from both functions.
(cherry picked from commit 39b4061bb9)
When doing line stipple, the stipple count resets on each line segment,
unless the primitive is a GL_LINE_LOOP or a GL_LINE_STRIP.
The existing code correctly identifies the need for a software fallback
to handle conformant line stipple on GL_LINE_LOOP primitives, but
neglects to make the same assessment on GL_LINE_STRIP primitives.
This fixes it so they match.
(cherry picked from commit 73658ff04f)
Back-face stencil operations didn't work correctly because this value was
zero. It needs to be 1 or 2. The only place it's set otherwise is in
glEnable/Disable(GL_STENCIL_TEST_TWO_SIDE_EXT).
(cherry picked from commit 2a968113a9)
Track separate back-face stencil state for OpenGL 2.0 /
GL_ATI_separate_stencil and GL_EXT_stencil_two_side. This allows all
three to be enabled in a driver. One set of state is set via the 2.0
or ATI functions and is used when STENCIL_TEST_TWO_SIDE_EXT is
disabled. The other is set by StencilFunc and StencilOp when the
active stencil face is set to BACK. The GL_EXT_stencil_two_side spec has
more details.
http://opengl.org/registry/specs/EXT/stencil_two_side.txt
(cherry picked from commit dde7cb9628)
If we're using anything but GL_NEAREST sampling of a cube map, we need to
use the BRW_TEXCOORDMODE_CUBE texcoord wrap mode. Before this, the GPU
would either lock up or subsequent texture filtering would be corrupted.
(cherry picked from master, commit 6f915b10d5)
This catches the linux-uclibc case and any others that were being set
prior to 98fcdf3f. Fixes bug 20345.
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
(cherry picked from commit ac55db1d7d)
When a hw driver fell back to swrast, swrast wasn't always getting informed
of program changes. When fixed function is translated into shaders, flags
like _NEW_LIGHT, _NEW_TEXTURE, etc. should really signal _NEW_PROGRAM.
In this case, swrast wasn't seeing _NEW_PROGRAM when new fragment shaders
were generated.
(cherry picked from master, commit de1caa5507)
If glBufferDataARB() is called while a buffer object is currently mapped
we're supposed to unmap the current buffer, then replace it. Don't generate
an error.
(cherry picked from master, commit 75e3ccf6a5)
This fixes the case of "infinitely" nested swizzles such as EXPR.wzyx.yxwz.xxyz
This doesn't appear in typical shaders but with function inlining and the
compiler's internal use of swizzles it can happen.
New glean glsl1 test case added for this.
(cherry picked from master, commit d9881356a6)
Only 15 actually worked before since we always reserved generic[0] as an
alias for vertex position.
The case of vertex attribute 0 is tricky. The spec says that there is no
aliasing between generic vertex attributes 0..MAX_VERTEX_ATTRIBS-1 and the
conventional attributes. But it also says that calls to glVertexAttrib(0, v)
are equivalent to glVertex(v). The distinction seems to be in glVertex-mode
versus vertex array mode.
So update the VBO code so that if the shader uses generic[0] but not gl_Vertex,
route the attribute data set with glVertex() to go to shader input generic[0].
No change needed for the glDrawArrays/Elements() path.
This is a potentially risky change so regressions are possible. All the usual
tests seem OK though.
(cherry picked from mesa, commit dea0d4d563)
Follow-up: This patch doesn't seem to be as risky as noted above.
There have been no known regressions on Mesa/master because of it.
Move the is_boolean/integer_type() calls out of the loops.
Move the is_sampler_type() function near the bool/int functions.
Add a bunch of comments.
(cherry picked from master, commit c4ffbf009e)
If the vertex shader writes to a varying array with a variable index,
mark all the elements of that array as being written.
For example, if the vertex shader does:
for (i = 0; i < 4; i++)
gl_TexCoord[i] = expr;
Mark all texcoord outputs as being written, not just the first.
Linking will fail if a fragment shader tries to read an input that's not
written by the vertex shader. Before this fix, this linker test could fail.
(cherry picked from master, commit dac19f17f3)
Old limit was 256. Note that no arrays are declared to this size.
The only place we have to be careful about raising this limit is the
prog_src/dst_register Index bitfields. These have been bumped up too.
Added assertions to check we don't exceed the bitfield in the future too.
(cherry picked from master, commit 5b2f8dc013)
The new array features, precision/invariant/centroid qualifiers, etc. were
done a while back. The glGetString(GL_SHADING_LANGUAGE_VERSION) query returns
"1.20" now (for drivers that support it anyway).
(cherry picked from master, commit f59719c6c7)
This fixes a bug found with swizzled array indexes such as in "array[index.z]"
where "index" is an ivec4.
(cherry picked from master, commit 6eabfc27f1)
Arrays of sampler vars haven't been tested much and might actually be broken.
Will need to be revisited someday.
Another fix for bug 20056.
(cherry picked from master, commit 2b4f0216bf)
location = -1 is silently ignored, but other negative values should raise
an error.
Another fix for bug 20056.
(cherry picked from master, commit 234f03e90a)
