We can perform less moves by checking for scenarios where aliasing
occurs. Since addition is commutative (usually, general-case anyway),
order of inputs doesn't strictly matter here.
In the event no source vectors alias the destination,
we can just move the first source vector into it and
then perform the divide without needing to move afterword.
We can avoid needing to use movprfx here by moving
directly into the destination when possible and just
doing the UMAX directly.
Also expands the unsigned max tests to test values with
the sign bit set to ensure all behavior is caught.
Since SMAX performs a comparison and returns the max value regardless
of how the operands are provided, we can check for when the second
input aliases the destination.
Removes the truncating move that we perform inside the StoreResult
function and instead delegates the responsibility to the instruction
implementations themselves.
This removes a lot of redundant moves that occur on 128-bit variants
of AVX instructions.
Also fixes a weird case where we were handling 128-bit SVE
in VBroadcastFromMem when we already have AdvSIMD instructions
that will perfom the zero-extension behavior for us.
Allows for easier expansion without needing to expand the function definitons.
Also makes a few usages significantly less verbose and makes specifying
options a little more declarative, rather than having to memorize what
each argument is specifying.
We don't need zero in the upper bits for a push.
Makes a couple variants optimal.
Adds missing tests to the 32-bit file, since only 32-bit can push a
32-bit register.
These IR operations are required to support AFP's NEP mode which does
vector insert in to the destination register. Additionally it gives us
tracking information to allow optimizing out redundant inserts on
devices that don't support AFP natively.
In order to match x86 semantics we need to support binary and unary
scalar operations that do a final insert in to a vector. With optional
zeroing of the top 128-bits for AVX variants.
A tricky thing is that in binary operations this means that the
destination and first source have an intrinsically linked property
depending on if it is SSE or AVX.
SSE example:
- addss xmm0, xmm1
- xmm0 is both the destination and the first source.
- This means xmm0[31:0] = xmm0[31:0] + xmm1[31:0]
- Bits [127:32] are UNMODIFIED.
FEX's JIT jumps through some hoops so that if the destination register
equals the first source register, then it hits the optimal path the
AFP.NEP will insert in to the result. AVX throws a small wrench in to
this due to changed behaviour
AVX example:
- vaddss xmm0, xmm1, xmm2
- xmm0 is ONLY the destination, xmm1 and xmm2 are the sources
- This operation copies the bits above the scalar result from the
first source (xmm1).
- Additionally this will zero bits above the original 128-bit xmm
register.
- xmm0[31:0] = xmm1[31:0] + xmm2[31:0]
- xmm0[127:32] = xmm1[127:32]
- ymm0[255:127] = 0
This causes these instructions to support a fairly large table depending
on if the instruction is an SSE or AVX instruction, plus if the host CPU
supports AFP or not.
So while fairly complex, it's handling all the edge cases and gives us
optimization opportunities as we move forward. Currently on non-AFP
supporting devices this has a minor benefit that these IR operations
remove one temporary register, lowering the Register Allocation
overhead.
In the coming weeks I am likely to introduce an optimization pass that
removes redundant inserts because FEX currently does /really/ badly with
scalar code loops.
Needs #3184 merged first.
When FEX is in the JIT we need to make sure to enable NEP and AH and
then disable when leaving.
Explicitly disabled when the vixl simulator is used since even
attempting to set the bits will cause it to fault out. Ensures
InstCountCI keeps working.
There are some cases where we want to test multiple instructions where
we can do optimizations that would overwise be hard to see.
eg:
```asm
; Can be optimized to a single stp
push eax
push ebx
; Can remove half of the copy since we know the direction
cld
rep movsb
; Can remove a redundant insert
addss xmm0, xmm1
addss xmm0, xmm2
```
This lets us have arbitrary sized code in instruction count CI, with the
original json key becoming only a label if the instruction array is
provided.
There are still some major limitations to this, instructions that
generate side-effects might have "garbage" after the end of the block
that isn't correctly accounted for. So care must be taken.
Example in the json
```json
"push ax, bx": {
"ExpectedInstructionCount": 4,
"Optimal": "No",
"Comment": "0x50",
"x86Insts": [
"push ax",
"push bx"
],
"ExpectedArm64ASM": [
"uxth w20, w4",
"strh w20, [x8, #-2]!",
"uxth w20, w7",
"strh w20, [x8, #-2]!"
]
}
```
Six of the EFLAGS can't be used directly in a bitmask because they are
either contained in a different flags location or has multiple bits
stored in it.
SF, ZF, CF, OF are stored in ARM's NZCV format in offset 24.
PF calculation is deferred but stored in the regular offset.
AF is also deferred in relation to the PF but stored in the regular
offset.
These /need/ to be reconstructed using the `ReconstructCompactedEFLAGS`
function when wanting to read the EFLAGS.
When setting these flags they /need/ to be set using
`SetFlagsFromCompactedEFLAGS`.
If either of these functions are not used when managing EFLAGs then the
internal representation will get mangled and the state will be
corrupted.
Having a little `_RAW` on these to signify that these aren't just
regular single bit representations like the other flags in EFLAGS should
make us puzzle about this issue before writing more broken code that
tries accessing it directly.
This allows us to use reciprocal instructions which matches precision of
what x86 expects rather than converting everything to float divides.
Currently no hardware supports this, and even the upcoming X4/A720/A520
won't support it, but it was trivial to implement so wire it up.
Suggested by Alyssa. Adding an IR operation can be a little tedious
since you need to add the definition to JIT.cpp for the dispatch switch,
JITClass.h for the function declared, and then actually defining the
implementation in the correct file.
Instead support the common case where an IR operation just gets
dispatched through to the regular handler. This lets the developer just
put the function definition in to the json and the relevent cpp file and
it just gets picked up.
Some minor things:
- Needs to support dynamic dispatch for {Load,Store}Register and
{Load,Store}Mem
- This is just a bool in the json
- It needs to not output JIT dispatch for some IR operations
- SSE4.2 string instructions and x87 operations
- These go down the "Unhandled" path
- Needs to support a Dispatcher function override
- This is just for handling NoOp IR operations that get used for
other reasons.
- Finally removes VSMul and VUMul, consolidating to VMul
- Unlike V{U,S}Mull, signed or unsigned doesn't change behaviour here
- Fixed a couple random handler names not matching the IR operation
name.
Enables in InstCountCI so Pi users can run InstCountCI can run the tests
without breaking on crypto operations.
When crypto is enabled or disabled just wholesale change AES, CRC32, and
PMULL 128-bit in one step. We don't really care about partial support
here.
The motivation towards just having a pointer array in CpuState was that
initialization was fairly cheap and that we have limited space inside
the encoding depending on what we want to do.
Initialization cost is still a concern but doing a memcpy of 128-bytes
isn't that big of a deal.
Limited space in CpuState, while a concern isn't a significant one.
- Needs to currently be less than 1 page in size
- Needs to be under the architectural offset limitations of loadstore
scaled offsets. Which is 65KB for 128-bit vectors
Still keeps the pointer array around for cases when we would need
synthesize an address offset and it's just easier to load the
process-wide table.
The performance improvement here is removing the dependency in the
ldr+ldr chain. In microbenchmarks this has shown to have an improvement
of ~4% by removing this dependency chain on Cortex-X1C.
Using the cached zero value is less efficient than loading it in to the
register for all these cases.
Lets us use rename hardware more efficiently and removes a dependency
chain on a single register.
Original:
```
movi v2.2d, #0x0
mov z16.d, p7/m, z2.d
<... 16 more times>
mov z31.d, p7/m, z2.d
```
Result:
```
movi v16.2d, #0x0
<... 16 more times>
movi v31.2d, #0x0
```
With the removal of the x86 JIT, there is no need to have these be
independent classes.
Merges the Arm64Dispatcher in to the base Dispatcher class.
No functional change, just moving code.
If we ever get around to fusing ops with shifts in the ConstProp optimizer (may
or may not be worthwhile), this will delete an instruction from things like "or
al, bh".
Even though lsr is the same speed as bfe on Firestorm, I feel if you ask for
garbage you should get garbage C:
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Pointless, upper bits ignored anyway. Deletes piles of uxt and even some 32-bit
instruction moves.
Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Lioncache