Skip to content

Cuda interop vk13 #90

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

Open
wants to merge 21 commits into
base: master
Choose a base branch
from
Open

Cuda interop vk13 #90

Changes from 1 commit
Commits
Show all changes
21 commits
Select commit Hold shift + click to select a range
3a02134
working cuda vulkan interop example
atkurtul Jul 9, 2023
c90032d
proper cleanup
atkurtul Jul 9, 2023
6d0a577
add cuda to vulkan example
atkurtul Jul 9, 2023
867b435
refactor
atkurtul Jul 9, 2023
831e61e
semaphore interop
atkurtul Jul 15, 2023
1396829
update
atkurtul Jul 16, 2023
d7fb50e
texture sharing
atkurtul Jul 27, 2023
682a8c1
update to latest changes
atkurtul Aug 6, 2023
faddda4
Merge branch 'cuda-interop' into HEAD
atkurtul Jan 4, 2024
6ce21d5
cuda-interop-vk13
atkurtul Jan 4, 2024
4159025
fix sync
atkurtul Jan 4, 2024
c1b0ddc
Merge remote-tracking branch 'origin/vulkan_1_3' into HEAD
atkurtul Jan 13, 2024
386608b
rename 63 example folder
atkurtul Jan 13, 2024
f9a2d1b
derive from monodevice
atkurtul Jan 14, 2024
138356a
rewrite stuff
atkurtul Jan 14, 2024
e96a4bf
Merge commit '558f9724' into HEAD
atkurtul Jan 14, 2024
db2ac71
reflect nabla changes
atkurtul Jan 18, 2024
744dd44
address pr comments
atkurtul Jan 18, 2024
59ebd29
filter cuda compatible devices
atkurtul Jan 18, 2024
9897e11
Merge branch 'vulkan_1_3' into cuda-interop-vk13
atkurtul Jan 18, 2024
73f1479
fix some comments & add test code
atkurtul Jan 19, 2024
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
30 changes: 14 additions & 16 deletions 63.CUDAInterop/main.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -251,8 +251,6 @@ class CUDA2VKApp : public examples::MonoSystemMonoLoggerApplication
CUDA_EXTERNAL_SEMAPHORE_SIGNAL_PARAMS signalParams = { .params = {.fence = {.value = 1 } } };
auto semaphore = cusema->getInternalObject();
ASSERT_SUCCESS(cu.pcuSignalExternalSemaphoresAsync(&semaphore, &signalParams, 1, stream)); // Signal the imported semaphore

std::string abc = "123";
}

// After the cuda kernel has signalled our exported vk semaphore, we will download the results through the buffer imported from CUDA
Expand Down Expand Up @@ -311,13 +309,16 @@ class CUDA2VKApp : public examples::MonoSystemMonoLoggerApplication

re &= cmd->copyImageToBuffer(importedimg.get(), imgBarrier.newLayout, stagingbuf2.get(), 1, &imgRegion);
re &= cmd->end();

IQueue::SSubmitInfo submitInfo = {
.waitSemaphores = std::array{IQueue::SSubmitInfo::SSemaphoreInfo{ .semaphore = sema.get(), .value = 0, .stageMask = PIPELINE_STAGE_FLAGS::ALL_COMMANDS_BITS, }},
.commandBuffers = std::array{IQueue::SSubmitInfo::SCommandBufferInfo{cmd.get()}},
};

auto submitRe = queue->submit(std::array{submitInfo});;

auto waitSemaphores = std::array{IQueue::SSubmitInfo::SSemaphoreInfo{.semaphore = sema.get(), .value = 1, .stageMask = PIPELINE_STAGE_FLAGS::ALL_COMMANDS_BITS, }};
auto signalSemaphores = std::array{IQueue::SSubmitInfo::SSemaphoreInfo{.semaphore = sema.get(), .value = 2, .stageMask = PIPELINE_STAGE_FLAGS::ALL_COMMANDS_BITS, }};
auto commandBuffers = std::array{IQueue::SSubmitInfo::SCommandBufferInfo{cmd.get()}};
auto submitInfo = std::array{IQueue::SSubmitInfo {
.waitSemaphores = waitSemaphores,
.commandBuffers = commandBuffers,
.signalSemaphores = signalSemaphores,
}};
auto submitRe = queue->submit(submitInfo);
re &= IQueue::RESULT::SUCCESS == submitRe;
assert(re);
}
Expand All @@ -328,11 +329,8 @@ class CUDA2VKApp : public examples::MonoSystemMonoLoggerApplication
void kernelCallback()
{
// Make sure we are also done with the readback
{
//IGPUFence* fences[] = { fence.get() };
//auto status = logicalDevice->waitForFences(1, fences, true, -1);
//assert(IGPUFence::ES_SUCCESS == status);
}
auto wait = std::array{ILogicalDevice::SSemaphoreWaitInfo{.semaphore = sema.get(), .value = 2}};
logicalDevice->waitForSemaphores(wait, true, -1);

float* A = reinterpret_cast<float*>(cpubuffers[0]->getPointer());
float* B = reinterpret_cast<float*>(cpubuffers[1]->getPointer());
Expand All @@ -343,8 +341,8 @@ class CUDA2VKApp : public examples::MonoSystemMonoLoggerApplication

for (auto i = 0; i < numElements; i++)
{
assert(abs(CBuf[i] - A[i] - B[i]) < 0.01f);
assert(abs(CImg[i] - A[i] - B[i]) < 0.01f);
bool re = (abs(CBuf[i] - A[i] - B[i]) < 0.01f) && (abs(CImg[i] - A[i] - B[i]) < 0.01f);
assert(re);
}

std::cout << "Success\n";
Expand Down