AgeCommit message (Collapse)AuthorFilesLines
2010-09-14r600c: fix setting negative values to bitfields7.9-branchpointAndre Maasikas1-1/+1
when setting negative integers to bitfields we could overwrite other parts of it. So mask the value to be written correctly. This is used quite often in the driver - hope it doesnt affect performace or uncover behaviour relied before... fixes strange effects when setting negative lodbias on evergreen
2010-09-14mesa: upgrade wglext.h to version 22Brian Paul2-106/+139
2010-09-14mesa: upgrade glxext.h to version 32Brian Paul2-2/+48
2010-09-14mesa: update to version 64 of GL/glext.hBrian Paul14-102/+1050
A number of other files had to be updated as well because const qualifiers were added to the glMultiDrawArrays() function. Also, GL_FIXED is now defined in glext.h.
2010-09-14mesa: move, redefine MESA_GEOMETRY_PROGRAMBrian Paul2-9/+15
2010-09-14glsl2: add case for ir_unop_noiseBrian Paul1-0/+4
Silences a compiler warning. Still need to add some assertions for this case.
2010-09-14glsl2: fix commentsBrian Paul3-3/+3
2010-09-14mesa: Remove unnecessary FEATURE tests.Chia-I Wu2-18/+0
Remove all FEATURE tests in mesa/drivers/common/. They are not needed and the code looks better without them.
2010-09-14mesa: Less FEATURE_ARB_sync tests.Chia-I Wu5-25/+60
Add dummy static inline definitions to syncobj.h when FEATURE_ARB_sync is 0, and remove most FEATURE_ARB_sync tests.
2010-09-14mesa: Remove unused _MESA_INIT_*_FUNCTIONS.Chia-I Wu11-90/+0
They were intended to be used to build OpenGL ES only DRI drivers, but that never happened.
2010-09-14gallivm: Remove unnecessary header.Vinson Lee1-1/+0
2010-09-14llvmpipe: Initialize variable for potentially unhandled switch case.Vinson Lee1-0/+3
2010-09-14i965g: Fix 'control reaches end of non-void function' warning.Vinson Lee1-1/+2
Fixes the following GCC warning. brw_screen.c: In function 'brw_get_shader_param': brw_screen.c:241: warning: control reaches end of non-void function
2010-09-13i915g: Fix 'control reaches end of non-void function' warning.Vinson Lee1-1/+2
Fixes the following GCC warning. i915_screen.c: In function 'i915_get_shader_param': i915_screen.c:184: warning: control reaches end of non-void function
2010-09-13i915: Fix "implicit declaration of function 'draw_get_shader_param'" warning.Vinson Lee1-0/+1
Fixes the following GCC warning. i915_screen.c: In function 'i915_get_shader_param': i915_screen.c:147: warning: implicit declaration of function 'draw_get_shader_param'
2010-09-13identity: Fix 'assignment from incompatible pointer type' warning.Vinson Lee1-1/+1
This is a follow-up to commit a508d2dddcc67d0f92cc36b9ed6f36a9bbfc579d. Fixes the following GCC warning. id_screen.c: In function 'identity_screen_create': id_screen.c:317: warning: assignment from incompatible pointer type
2010-09-13rbug: Fix 'assignment from incompatible pointer type' warning.Vinson Lee1-1/+1
This is a follow-up to commit a508d2dddcc67d0f92cc36b9ed6f36a9bbfc579d. Fixes the following GCC warning. rbug_screen.c: In function 'rbug_screen_create': rbug_screen.c:331: warning: assignment from incompatible pointer type
2010-09-14mesa/st: ask GLSL to not emit noise since we have a dummy implementationLuca Barbieri1-0/+2
Note, BTW, that the Gallium implementation returns 0.5, which seems to violate the GLSL spec, where it should return 0.0 instead. Not sure whether changing it to 0 is correct or not.
2010-09-14mesa/st: set compiler options based on Gallium shader capsLuca Barbieri1-0/+9
This turns on if conversion and unlimited loop unrolling if control flow is not supported. NOTE: this will change the behavior of r300g and any other driver that doesn't advertise control flow
2010-09-14gallium: introduce get_shader_param (ALL DRIVERS CHANGED) (v3)Luca Barbieri25-345/+612
Changes in v3: - Also change trace, which I forgot about Changes in v2: - No longer adds tessellation shaders Currently each shader cap has FS and VS versions. However, we want a version of them for geometry, tessellation control, and tessellation evaluation shaders, and want to be able to easily query a given cap type for a given shader stage. Since having 5 duplicates of each shader cap is unmanageable, add a new get_shader_param function that takes both a shader cap from a new enum and a shader stage. Drivers with non-unified shaders will first switch on the shader and, within each case, switch on the cap. Drivers with unified shaders instead first check whether the shader is supported, and then switch on the cap. MAX_CONST_BUFFERS is now per-stage. The geometry shader cap is removed in favor of checking whether the limit of geometry shader instructions is greater than 0, which is also used for tessellation shaders. WARNING: all drivers changed and compiled but only nvfx tested
2010-09-13glsl2: Port equal() and notEqual() to ir_unop_all_equal and ir_unop_any_nequalIan Romanick3-240/+48
2010-09-13glsl: introduce ir_binop_all_equal and ir_binop_any_equal, allow vector cmpsLuca Barbieri8-23/+74
Currently GLSL IR forbids any vector comparisons, and defines "ir_binop_equal" and "ir_binop_nequal" to compare all elements and give a single bool. This is highly unintuitive and prevents generation of optimal Mesa IR. Hence, first rename "ir_binop_equal" to "ir_binop_all_equal" and "ir_binop_nequal" to "ir_binop_any_nequal". Second, readd "ir_binop_equal" and "ir_binop_nequal" with the same semantics as less, lequal, etc. Third, allow all comparisons to acts on vectors. Signed-off-by: Ian Romanick <>
2010-09-13loop_unroll: unroll loops with (lowered) breaksLuca Barbieri1-4/+89
If the loop ends with an if with one break or in a single break unroll it. Loops that end with a continue will have that continue removed by the redundant jump optimizer. Likewise loops that end with an if-statement with a break at the end of both branches will have the break pulled out after the if-statement. Loops of the form for (...) { do_something1(); if (cond) { do_something2(); break; } else { do_something3(); } } will be unrolled as do_something1(); if (cond) { do_something2(); } else { do_something3(); do_something1(); if (cond) { do_something2(); } else { do_something3(); /* Repeat inserting iterations here.*/ } } ir_lower_jumps can guarantee that all loops are put in this form and thus all loops are now potentially unrollable if an upper bound on the number of iterations can be found. Signed-off-by: Ian Romanick <>
2010-09-13glsl2: Add pass to remove redundant jumpsIan Romanick6-1/+118
2010-09-13glsl: Explain file naming conventionIan Romanick1-0/+12
2010-09-13loop_controls: fix analysis of already analyzed loopsLuca Barbieri1-1/+8
The loop_controls pass didn't look at the counter values it put in ir_loop on previous iterations, so while the first iteration worked, subsequent ones couldn't determine max_iterations.
2010-09-13i965: Request that returns be lowered in shader mainIan Romanick1-0/+1
Fixes piglit tests glsl-vs-main-return and glsl-fs-main-return.
2010-09-13glsl: call ir_lower_jumps according to compiler optionsLuca Barbieri1-0/+2
2010-09-13glsl: add continue/break/return unification/elimination pass (v2)Luca Barbieri6-250/+548
Changes in v2: - Base class renamed to ir_control_flow_visitor - Tried to comply with coding style This is a new pass that supersedes ir_if_return and "lowers" jumps to if/else structures. Currently it causes no regressions on softpipe and nv40, but I'm not sure whether the piglit glsl tests are thorough enough, so consider this experimental. It can be asked to: 1. Pull jumps out of ifs where possible 2. Remove all "continue"s, replacing them with an "execute flag" 3. Replace all "break" with a single conditional one at the end of the loop 4. Replace all "return"s with a single return at the end of the function, for the main function and/or other functions This gives several great benefits: 1. All functions can be inlined after this pass 2. nv40 and other pre-DX10 chips without "continue" can be supported 3. nv30 and other pre-DX10 chips with no control flow at all are better supported Note that for full effect we should also teach the unroller to unroll loops with a fixed maximum number of iterations but with the canonical conditional "break" that this pass will insert if asked to. Continues are lowered by adding a per-loop "execute flag", initialized to TRUE, that when cleared inhibits all execution until the end of the loop. Breaks are lowered to continues, plus setting a "break flag" that is checked at the end of the loop, and trigger the unique "break". Returns are lowered to breaks/continues, plus adding a "return flag" that causes loops to break again out of their enclosing loops until all the loops are exited: then the "execute flag" logic will ignore everything until the end of the function. Note that "continue" and "return" can also be implemented by adding a dummy loop and using break. However, this is bad for hardware with limited nesting depth, and prevents further optimization, and thus is not currently performed.
2010-09-13glsl: add ir_control_flow_visitorLuca Barbieri1-0/+17
This is just a subclass of ir_visitor with empty implementations of all the visit methods for non-control flow nodes. Used to avoid duplicating that in ir_visitor subclasses. ir_hierarchical_visitor is another way to solve this, but is less natural for some applications.
2010-09-13llvmpipe: Fix non SSE2 builds.José Fonseca1-2/+2
Should fix fdo 30168.
2010-09-13r300g/swtcl: unlock VBO after draw_flushMarek Olšák1-4/+1
2010-09-13llvmpipe: Change asm to __asm__.Witold Baryluk1-3/+3
According to gcc documentation both are equivalent, second are prefered as first can make conflict with existing symbols. Signed-off-by: José Fonseca <>
2010-09-13EGL DRI2: 0xa011 is Pineview not IronlakeJesse Barnes1-1/+1
Point about needing a better way to do this validated.
2010-09-13r600c: const buffer sizes must be a multiple of 16 constsAlex Deucher3-29/+21
This applies to r6xx/r7xx/evergreen
2010-09-13EGL DRI2: add PCI ID for Ironlake mobileJesse Barnes1-0/+1
Allows KMS EGL driver to load. We need a better way of doing this.
2010-09-13r600c/eg: remove obselete commentAlex Deucher1-2/+0
2010-09-13r600c/eg: remove unused emit timestamp functionAlex Deucher1-8/+0
2010-09-13r600c/eg: emit CB_BLEND_ALPHA with the other blend valuesAlex Deucher1-5/+5
saves a few dwords
2010-09-13r600c: remove redundant state emit on evergreenAlex Deucher1-17/+0
r700start3d already emits the context control packets
2010-09-13mesa: Revert accidentally committed vertex code chunkKristian Høgsberg1-2/+0
2010-09-13r600c: eg: fix typoAndre Maasikas1-1/+1
probably copy/paste error
2010-09-13r600c: eg: 256 float4 constants may need more than 256 bytesAndre Maasikas2-2/+2
2010-09-13r600c: eg - fix uninitialized variableAndre Maasikas1-0/+2
2010-09-13glx: Don't destroy DRI2 drawables for legacy glx drawablesKristian Høgsberg2-1/+13
For GLX 1.3 drawables, we can destroy the DRI2 drawable when the GLX drawable is destroyed. However, for legacy drawables, there os no good way of knowing when the application is done with it, so we just let the DRI2 drawable linger on the server. The server will destroy the DRI2 drawable when it destroys the X drawable or the client exits anyway.
2010-09-13r300g: fix SWTCLMarek Olšák4-41/+99
2010-09-13llvmpipe: Unbreak rasterization on 64bit.José Fonseca1-24/+22
2010-09-13gallium: Change the resource_copy_region semantics to allow copies between ↵José Fonseca1-3/+5
different yet compatible formats
2010-09-13r600g: evergreen fixup dsa state for running query.Dave Airlie2-3/+2
evergreen is always the same as r700 here.
2010-09-13r600c: remove stray unmap callAndre Maasikas1-1/+0
no idea how/why it got there