GPU/Shader Instruction Set - Revision history

Oreo639: Clarify LITP instruction

2023-10-02T07:38:34Z

Clarify LITP instruction

← Older revision		Revision as of 11:38, 2 October 2023
Line 349:		Line 349:
	\| 1u		\| 1u
	\| LITP		\| LITP
	\| ~~Appears to~~ be ~~related~~ to ~~Microsoft's~~ [https://msdn.microsoft.com/en-us/library/windows/desktop/bb174703.aspx lit~~] instruction~~; DST = ~~clamp~~(~~SRC1~~, min~~={0~~, -127.9961, 0~~, 0}~~, max~~={inf, 127~~.~~9961~~, 0~~, inf~~}~~); n~~.b.~~: 127~~.~~9961~~ = ~~0x7FFF / 0x100~~		\| Partial lighting computation, may be used in conjunction with EX2, LG2, etc to compute the vertex lighting coefficients. See the [https://msdn.microsoft.com/en-us/library/windows/desktop/bb174703.aspx Microsoft] and [https://registry.khronos.org/OpenGL/extensions/ARB/ARB_vertex_program.txt ARB] docs for more information on how to implement the full lit function; DST = {max(src.x, 0), max(min(src.y, 127.9961), -127.9961), 0, max(src.w, 0)} and it sets the cmp.x and cmp.y flags based on if the respective src.x and src.w components are >= 0.
	\|-		\|-
	\| 0x08		\| 0x08

Steveice10: More thorough behavior testing

2023-08-24T15:54:39Z

More thorough behavior testing

← Older revision		Revision as of 19:54, 24 August 2023
Line 664:		Line 664:

	a0.x and a0.y are set manually through the MOVA instruction by rounding a float value to integer precision. Hence, they may take negative values. The way out-of-bounds values behave when reading uniforms is as follows:		a0.x and a0.y are set manually through the MOVA instruction by rounding a float value to integer precision. Hence, they may take negative values. The way out-of-bounds values behave when reading uniforms is as follows:
	* If the ~~value~~ is out of byte bounds (less than -128 or greater than 127), the ~~result~~ is ~~the vector~~ (~~0, 0, 0,~~ 0).		* If the offset is out of byte bounds (less than -128 or greater than 127), the offset is not applied (treated as 0).
	* If the ~~value without its sign bit (value & 0x7F)~~ is greater than 95, the result is ~~the vector~~ (1, 1, 1, 1).		* The offset is added to the constant register index and masked by 0x7F.
	* Otherwise, the ~~index of~~ the ~~resulting uniform~~ value is the ~~value without its sign bit (value & 0x7F)~~.		* If the resulting index is greater than 95, the result is (1, 1, 1, 1).
			* Otherwise, the result is the value at the indexed constant register.

	aL can only be set indirectly by the LOOP instruction. It is still accessible and valid after exiting a LOOP block, though.		aL can only be set indirectly by the LOOP instruction. It is still accessible and valid after exiting a LOOP block, though.

Steveice10: /* Relative addressing */

2023-08-24T08:53:23Z

Relative addressing

← Older revision		Revision as of 12:53, 24 August 2023
Line 663:		Line 663:
	There are 3 address registers: a0.x, a0.y and aL (loop counter). For format 1 instructions, when IDX != 0, the value of the corresponding address register is added to SRC1's value. For example, if IDX = 2, a0.y = 3 and SRC1 = c8, then instead SRC1+a0.y = c11 will be used for the instruction. It is only possible to use address registers on constant registers, attempting to use them on input attribute or temporary registers results in the address register being ignored (i.e. read as zero).		There are 3 address registers: a0.x, a0.y and aL (loop counter). For format 1 instructions, when IDX != 0, the value of the corresponding address register is added to SRC1's value. For example, if IDX = 2, a0.y = 3 and SRC1 = c8, then instead SRC1+a0.y = c11 will be used for the instruction. It is only possible to use address registers on constant registers, attempting to use them on input attribute or temporary registers results in the address register being ignored (i.e. read as zero).

	a0.x and a0.y are set manually through the MOVA instruction by rounding a float value to integer precision. Hence, they may take negative values.		a0.x and a0.y are set manually through the MOVA instruction by rounding a float value to integer precision. Hence, they may take negative values. The way out-of-bounds values behave when reading uniforms is as follows:
			* If the value is out of byte bounds (less than -128 or greater than 127), the result is the vector (0, 0, 0, 0).
			* If the value without its sign bit (value & 0x7F) is greater than 95, the result is the vector (1, 1, 1, 1).
			* Otherwise, the index of the resulting uniform value is the value without its sign bit (value & 0x7F).

	aL can only be set indirectly by the LOOP instruction. It is still accessible and valid after exiting a LOOP block, though.		aL can only be set indirectly by the LOOP instruction. It is still accessible and valid after exiting a LOOP block, though.

Oreo639 at 19:41, 29 October 2022

2022-10-29T17:41:02Z

← Older revision		Revision as of 21:41, 29 October 2022
Line 775:		Line 775:
	\| float		\| float
	\| Read only		\| Read only
	\| Application		\| Application/Vertex-stream
	\| Floating-point Constant registers.		\| Floating-point Constant registers.
	\|-		\|-

Oreo639: Forgot to fix the register names too

2021-08-13T21:08:41Z

Forgot to fix the register names too

← Older revision		Revision as of 01:08, 14 August 2021
Line 828:		Line 828:
	\|-		\|-
	\| 0x0-0xF		\| 0x0-0xF
	\| o0-o6		\| o0-o15
	\| Output registers.		\| Output registers.
	\|-		\|-
Line 845:		Line 845:
	\|-		\|-
	\| 0x0-0xF		\| 0x0-0xF
	\| v0-v7		\| v0-v15
	\| Input attribute registers.		\| Input attribute registers.
	\|-		\|-

Oreo639: Fix SRC/DST mapping

2021-08-13T21:05:52Z

Fix SRC/DST mapping

← Older revision		Revision as of 01:05, 14 August 2021
Line 827:		Line 827:
	! Description		! Description
	\|-		\|-
	\| 0x0-~~0x6~~		\| 0x0-0xF
	\| o0-o6		\| o0-o6
	\| Output registers.		\| Output registers.
Line 840:		Line 840:
	{\| class="wikitable" border="1"		{\| class="wikitable" border="1"
	\|-		\|-
	! ~~SRC1~~ raw value		! SRC raw value
	! Register name		! Register name
	! Description		! Description
	\|-		\|-
	\| 0x0-~~0x7~~		\| 0x0-0xF
	\| v0-v7		\| v0-v7
	\| Input attribute registers.		\| Input attribute registers.

Oreo639 at 00:55, 13 August 2021

2021-08-12T22:55:30Z

← Older revision		Revision as of 02:55, 13 August 2021
Line 810:		Line 810:

	Output registers hold the data to be passed to the later GPU stages and are write-only. Each of the output register is assigned a semantic by setting the corresponding [[GPU_Internal_Registers]]. Output registers o7-o15 are only available in vertex shaders.		Output registers hold the data to be passed to the later GPU stages and are write-only. Each of the output register is assigned a semantic by setting the corresponding [[GPU_Internal_Registers]]. Output registers o7-o15 are only available in vertex shaders.
	Keep in mind that writing to an output register's component ~~twice~~ appears ~~that writing twice~~ appears to cause problems (e.g. GPU hangs).		Keep in mind that writing to the same output register/component more than once appears appears to cause problems (e.g. GPU hangs).

	Temporary registers can be used for intermediate calculations and can be both read and written.		Temporary registers can be used for intermediate calculations and can be both read and written.

Oreo639 at 00:52, 13 August 2021

2021-08-12T22:52:32Z

← Older revision		Revision as of 02:52, 13 August 2021
Line 810:		Line 810:

	Output registers hold the data to be passed to the later GPU stages and are write-only. Each of the output register is assigned a semantic by setting the corresponding [[GPU_Internal_Registers]]. Output registers o7-o15 are only available in vertex shaders.		Output registers hold the data to be passed to the later GPU stages and are write-only. Each of the output register is assigned a semantic by setting the corresponding [[GPU_Internal_Registers]]. Output registers o7-o15 are only available in vertex shaders.
	Keep in mind that writing to an output register that ~~has been previously written to within 8 cycles of the first write~~ appears to cause ~~the GPU to hang~~. ~~After 8 cycles, all further writes to an output register appear to be ignored and do not cause a~~ GPU ~~hang~~.		Keep in mind that writing to an output register's component twice appears that writing twice appears to cause problems (e.g. GPU hangs).

	Temporary registers can be used for intermediate calculations and can be both read and written.		Temporary registers can be used for intermediate calculations and can be both read and written.

Oreo639: Update the output attribute section

2021-08-12T22:38:00Z

Update the output attribute section

← Older revision		Revision as of 02:38, 13 August 2021
Line 809:		Line 809:
	Input attribute registers store the per-vertex data given by the CPU and hence are read-only.		Input attribute registers store the per-vertex data given by the CPU and hence are read-only.

	Output ~~attribute~~ registers hold the data to be passed to the later GPU stages and are write-only. Each of the output ~~attribute~~ register ~~components~~ is assigned a semantic by setting the corresponding [[GPU_Internal_Registers]]. Output registers o7-o15 are only available in vertex shaders.		Output registers hold the data to be passed to the later GPU stages and are write-only. Each of the output register is assigned a semantic by setting the corresponding [[GPU_Internal_Registers]]. Output registers o7-o15 are only available in vertex shaders.
	~~It appears~~ that writing ~~twice~~ to ~~a component of~~ an output register that ~~was~~ written to ~~before can~~ cause ~~problems (e.g~~. GPU ~~hangs)~~.		Keep in mind that writing to an output register that has been previously written to within 8 cycles of the first write appears to cause the GPU to hang. After 8 cycles, all further writes to an output register appear to be ignored and do not cause a GPU hang.

	Temporary registers can be used for intermediate calculations and can be both read and written.		Temporary registers can be used for intermediate calculations and can be both read and written.

Oreo639: Fix typo

2021-07-25T02:31:24Z

Fix typo

← Older revision		Revision as of 06:31, 25 July 2021
Line 803:		Line 803:
	\| integer		\| integer
	\| Use		\| Use
	\| ~~Application~~		\| Vertex shader
	\| Loop count register.		\| Loop count register.
	\|}		\|}