Skip to content
Open
Show file tree
Hide file tree
Changes from 2 commits
Commits
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
14 changes: 5 additions & 9 deletions test/Feature/WaveOps/WaveActiveAnyTrue.test
Original file line number Diff line number Diff line change
Expand Up @@ -5,15 +5,11 @@ RWStructuredBuffer<uint> Out : register(u1);
[numthreads(4, 1, 1)]
void main(uint3 threadID : SV_DispatchThreadID) {
bool B1 = false;
switch (value[threadID.x]) {
case 0:
case 2: // threads 0, 1, 2; result for each false
Out[threadID.x] = WaveActiveAnyTrue(B1);
B1 = true;
break;
default: // thread 3; result is false
Out[threadID.x] = WaveActiveAnyTrue(B1);
break;
if (value[threadID.x] == 0 || value[threadID.x] == 2) {
Copy link

Choose a reason for hiding this comment

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

While this matches the behavior of the replaced switch, the comment seemed to imply that thread 1 was meant to also go through this branch. The later comment also seems to assume this ("result for all threads is true because B1 is true for threads 0-2"). However thread numbers are translated through the LUT value, which is [0, 0, 1, 2].

Even accounting for this looked up value, the removed and remaining comments don't seem to align with the test's design and expected behavior. The test seems to be written in a way to make it a bit confusing on purpose.

I don't understand the intention behind these odd details of the test. Was this a mistake in the code or in the comments?

It would be nice to have an explanatory comment for the test design.

The second test has a similar issue.

Out[threadID.x] = WaveActiveAnyTrue(B1);
B1 = true;
} else {
Out[threadID.x] = WaveActiveAnyTrue(B1);
}
// result for all threads is true because B1 is true for threads 0-2
Out[threadID.x + 4] = WaveActiveAnyTrue(B1);
Expand Down
25 changes: 12 additions & 13 deletions test/Feature/WaveOps/WaveActiveCountBits.test
Original file line number Diff line number Diff line change
Expand Up @@ -6,17 +6,16 @@ RWStructuredBuffer<uint> Out : register(u1);
void main(uint3 threadID : SV_DispatchThreadID) {
bool B1 = false;

switch (value[threadID.x]) {
case 0: // threads 0 and 1; result is number of active lanes (2)
Out[threadID.x + 4] = WaveActiveCountBits(true); // threads 0 and 1
case 2:
B1 = true; // set b1 to true for thread 3
break;
default:
Out[threadID.x + 4] = WaveActiveCountBits(false); // thread 2; expect 0
break;
}
// should be 3 because B1 set to true for threads 0,1, and 3.
if (value[threadID.x] == 0) // thread 0 and 1.
Out[threadID.x + 4] = WaveActiveCountBits(true);
// thread 0-3 set b1 to true.
if (value[threadID.x] == 0 || value[threadID.x] == 1)
Copy link

Choose a reason for hiding this comment

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

Note: This isn't equivalent to the original behavior. == 1 should be == 2 for equivalent behavior.

B1 = true;
else // thread 4
Out[threadID.x + 4] = WaveActiveCountBits(false);

// Out[threadID.x + 4] on thread 3 is never written to and should remain 0.
// All threads count 3 because B1 set to true for threads 0,1, and 3.
uint Count = WaveActiveCountBits(B1);
Out[threadID.x] = Count;
}
Expand All @@ -35,11 +34,11 @@ Buffers:
- Name: Out
Format: UInt32
Stride: 4
ZeroInitSize: 28
ZeroInitSize: 32
- Name: ExpectedOut
Format: UInt32
Stride: 4
Data: [3, 3, 3, 3, 2, 2, 0]
Data: [3, 3, 3, 3, 2, 2, 0, 0]
Results:
- Result: Test
Rule: BufferExact
Expand Down
Loading