external_mesa3d.git - Unnamed repository; edit this file 'description' to name the repository.

	Commit message (Collapse)	Author	Age	Files	Lines
*	gallivm: Help for combined extraction and broadcasting.	José Fonseca	2010-10-08	1	-0/+77
\| \| \| \|	Doesn't change generated code quality, but saves some typing.
*	gallivm: move i32_vec_type inside the #ifdef	Brian Paul	2010-09-16	1	-1/+2
\|
*	gallivm: fix incorrect vector shuffle datatype	Brian Paul	2010-09-16	1	-1/+5
\| \| \| \|	The permutation vector must always be a vector of int32 values.
*	gallivm: s/lp_build_broadcast_aos/lp_build_swizzle_scalar_aos/	José Fonseca	2010-09-11	1	-4/+7
\| \| \| \|	More accurate description of this function purpose.
*	gallivm: Pass condition masks as an unsigned bitmask.	José Fonseca	2010-09-05	1	-8/+7
\| \| \| \|	Much more convenient than boolean arrays.
*	gallivm: Support extended swizzles in lp_build_swizzle1_aos().	José Fonseca	2010-07-01	1	-15/+161
\| \| \| \|	And rename to lp_build_swizzle_aos().
*	llvmpipe: Remove lp_build_swizzle2_aos().	José Fonseca	2010-07-01	1	-49/+0
\| \| \| \|	Unnecessary special case.
*	Conditionally revert "gallivm: Use a more compact approach for ↵	José Fonseca	2010-05-13	1	-0/+10
\| \| \| \| \| \|	lp_build_broadcast_scalar()." This reverts commit a09e46c72461183c879d8472b44fe740ecc79b9f.
*	gallivm: Use a more compact approach for lp_build_broadcast_scalar().	José Fonseca	2010-05-12	1	-8/+10
\| \| \| \| \| \| \|	It produces exactly the same machine code, but it cuts 5% of the number of instructions generated for a typical shader. Also, preserve the scalar when length is 1.
*	gallivm: Actually do floor/ceil/trunc for scalars.	José Fonseca	2010-05-08	1	-0/+2
\| \| \| \| \| \|	Also start axing the code duplication for scalar case. The olution is to treat the scalar case specially in a few innermost functions, and leave outer functions untouched.
*	gallivm: Centralize SoA swizzling into a single place.	José Fonseca	2010-05-08	1	-0/+78
\|
*	gallivm/llvmpipe: rename some constant building functions	Brian Paul	2010-03-15	1	-2/+2
\|
*	llvmpipe: export the tgsi translation code to a common layer	Zack Rusin	2010-02-08	1	-0/+239
	the llvmpipe tgsi translation is a lot more complete than what was in gallivm so replacing the latter with the former. this is needed since the draw llvm paths will use the same code. effectively the proven llvmpipe code becomes gallivm.