Skip to content

8342103: C2 compiler support for Float16 type and associated scalar operations #22754

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
wants to merge 20 commits into from
+2,855 −67
Closed

8342103: C2 compiler support for Float16 type and associated scalar operations #22754

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
20 commits
Select commit Hold shift + click to select a range
c215eac
C2 compiler support for float16 scalar operations.
Dec 15, 2024
7cb694f
Adding missed check in container type detection.
Dec 16, 2024
3a6697e
Adding more test points
Dec 16, 2024
246cb27
Fixing obfuscation due to intrinsic entries
Dec 17, 2024
ec0834a
Addressing review comments
Dec 17, 2024
dd444c4
Review comments resolutions
Dec 23, 2024
c3f3226
Review suggestions incorporated.
Jan 3, 2025
e40c66a
Merge branch 'master' of http://github.com/openjdk/jdk into JDK-8342103
Jan 3, 2025
175f4ed
Updating copyright year of modified files.
Jan 3, 2025
43aa3eb
Review comments resolutions
Jan 13, 2025
692de9c
Review suggestions incorporated.
Jan 17, 2025
b64655c
Refining IR match rule
Jan 24, 2025
c0cdbb1
Merge branch 'master' of http://github.com/openjdk/jdk into JDK-8342103
Jan 24, 2025
e0602c1
Rebasing to jdk mainline
Jan 24, 2025
4f22ed8
Copyright header fix
Jan 27, 2025
854fc73
Updating typos in comments
Jan 28, 2025
19fc6c2
Fixing a typo error
Jan 29, 2025
8207c9f
Update test/micro/org/openjdk/bench/jdk/incubator/vector/Float16Opera…
Jan 30, 2025
82a4221
Fixing typos
Feb 4, 2025
111c808
Review comments resolutions
Feb 11, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 1 addition & 2 deletions src/hotspot/share/opto/convertnode.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -987,8 +987,7 @@ int Float16NodeFactory::get_float16_binary_oper(int opc) {
return Op_MaxHF;
case Op_MinF:
return Op_MinHF;
default:
return false;
default: ShouldNotReachHere();
}
}

Expand Down
19 changes: 11 additions & 8 deletions src/hotspot/share/opto/library_call.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8654,14 +8654,17 @@ bool LibraryCallKit::inline_fp16_operations(vmIntrinsics::ID id, int num_args) {

Node* result = nullptr;
switch (id) {
// Unary operations
case vmIntrinsics::_sqrt_float16: result = _gvn.transform(new SqrtHFNode(C, control(), fld1)); break;

// Ternary operations
case vmIntrinsics::_fma_float16: result = _gvn.transform(new FmaHFNode(control(), fld1, fld2, fld3)); break;
default:
fatal_unexpected_iid(id);
break;
// Unary operations
case vmIntrinsics::_sqrt_float16:
result = _gvn.transform(new SqrtHFNode(C, control(), fld1));
break;
// Ternary operations
case vmIntrinsics::_fma_float16:
result = _gvn.transform(new FmaHFNode(control(), fld1, fld2, fld3));
break;
default:
fatal_unexpected_iid(id);
break;
}
set_result(_gvn.transform(new ReinterpretHF2SNode(result)));
return true;
Expand Down
4 changes: 3 additions & 1 deletion src/hotspot/share/opto/mulnode.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -557,7 +557,9 @@ Node* MulHFNode::Ideal(PhaseGVN* phase, bool can_reshape) {

// Compute the product type of two half float ranges into this node.
const Type* MulHFNode::mul_ring(const Type* t0, const Type* t1) const {
if(t0 == Type::HALF_FLOAT || t1 == Type::HALF_FLOAT) return Type::HALF_FLOAT;
if (t0 == Type::HALF_FLOAT || t1 == Type::HALF_FLOAT) {
return Type::HALF_FLOAT;
}
return TypeH::make(t0->getf() * t1->getf());
}

Expand Down
10 changes: 5 additions & 5 deletions src/hotspot/share/opto/type.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -457,7 +457,7 @@ void Type::Initialize_shared(Compile* current) {
ABIO = make(Abio); // State-of-machine only
RETURN_ADDRESS=make(Return_Address);
FLOAT = make(FloatBot); // All floats
HALF_FLOAT = make(HalfFloatBot); // All half floats
HALF_FLOAT = make(HalfFloatBot); // All half floats
DOUBLE = make(DoubleBot); // All doubles
BOTTOM = make(Bottom); // Everything
HALF = make(Half); // Placeholder half of doublewide type
Expand Down Expand Up @@ -1083,11 +1083,11 @@ const Type *Type::xmeet( const Type *t ) const {
return t;

case HalfFloatTop:
if (_base == HalfFloatTop) return this;
if (_base == HalfFloatTop) { return this; }
case HalfFloatBot: // Half Float
if (_base == HalfFloatBot || _base == HalfFloatTop) return HALF_FLOAT;
if (_base == FloatBot || _base == FloatTop) return Type::BOTTOM;
if (_base == DoubleTop || _base == DoubleBot) return Type::BOTTOM;
if (_base == HalfFloatBot || _base == HalfFloatTop) { return HALF_FLOAT; }
if (_base == FloatBot || _base == FloatTop) { return Type::BOTTOM; }
if (_base == DoubleTop || _base == DoubleBot) { return Type::BOTTOM; }
typerr(t);
return Type::BOTTOM;
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Please use curly-braces even for single-line ifs


Expand Down
1 change: 0 additions & 1 deletion src/java.base/share/classes/jdk/internal/vm/vector/Float16Math.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -32,7 +32,6 @@
/**
* The class {@code Float16Math} constains intrinsic entry points corresponding
* to scalar numeric operations defined in Float16 class.
* @author
* @since 24
*/
public final class Float16Math {
Expand Down