intel/compute-runtime
1
0
Fork 0
mirror of https://github.com/intel/compute-runtime.git synced 2025-12-21 09:14:47 +08:00
Code Issues Packages Projects Releases Wiki Activity

Refactor TagAllocator

Browse Source 
Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
This commit is contained in:
Bartosz Dunajski
2021-03-24 18:21:13 +00:00
committed by Compute-Runtime-Automation
parent cb4db7767e
commit 5a50ad098c
49 changed files with 868 additions and 430 deletions
Download Patch File Download Diff File Expand all files Collapse all files

50
shared/source/helpers/timestamp_packet.h
View File

@@ -31,7 +31,7 @@ constexpr uint32_t preferredPacketCount = 16u;
#pragma pack(1)
template <typename TSize>
class TimestampPackets {
class TimestampPackets : public TagTypeBase {
public:
struct Packet {
TSize contextStart = 1u;
@@ -40,10 +40,14 @@ class TimestampPackets {
TSize globalEnd = 1u;
};
static GraphicsAllocation::AllocationType getAllocationType() {
static constexpr GraphicsAllocation::AllocationType getAllocationType() {
return GraphicsAllocation::AllocationType::TIMESTAMP_PACKET_TAG_BUFFER;
}
static constexpr TagNodeType getTagNodeType() { return TagNodeType::TimestampPacket; }
size_t getSinglePacketSize() const { return sizeof(Packet); }
bool isCompleted() const {
if (DebugManager.flags.DisableAtomicForPostSyncs.get()) {
return false;
@@ -96,29 +100,25 @@ class TimestampPackets {
};
#pragma pack()
using TimestampPacketStorage = TimestampPackets<uint32_t>;
static_assert(((4 * TimestampPacketSizeControl::preferredPacketCount + 2) * sizeof(uint32_t)) == sizeof(TimestampPacketStorage),
static_assert(((4 * TimestampPacketSizeControl::preferredPacketCount + 2) * sizeof(uint32_t)) == sizeof(TimestampPackets<uint32_t>),
"This structure is consumed by GPU and has to follow specific restrictions for padding and size");
class TimestampPacketContainer : public NonCopyableClass {
public:
using Node = TagNode<TimestampPacketStorage>;
TimestampPacketContainer() = default;
TimestampPacketContainer(TimestampPacketContainer &&) = default;
TimestampPacketContainer &operator=(TimestampPacketContainer &&) = default;
MOCKABLE_VIRTUAL ~TimestampPacketContainer();
const std::vector<Node *> &peekNodes() const { return timestampPacketNodes; }
void add(Node *timestampPacketNode);
const std::vector<TagNodeBase *> &peekNodes() const { return timestampPacketNodes; }
void add(TagNodeBase *timestampPacketNode);
void swapNodes(TimestampPacketContainer &timestampPacketContainer);
void assignAndIncrementNodesRefCounts(const TimestampPacketContainer &inputTimestampPacketContainer);
void resolveDependencies(bool clearAllDependencies);
void makeResident(CommandStreamReceiver &commandStreamReceiver);
protected:
std::vector<Node *> timestampPacketNodes;
std::vector<TagNodeBase *> timestampPacketNodes;
};
struct TimestampPacketDependencies : public NonCopyableClass {
@@ -130,27 +130,27 @@ struct TimestampPacketDependencies : public NonCopyableClass {
};
struct TimestampPacketHelper {
static uint64_t getContextEndGpuAddress(const TagNode<TimestampPacketStorage> &timestampPacketNode) {
return timestampPacketNode.getGpuAddress() + timestampPacketNode.tagForCpuAccess->getContextEndOffset();
static uint64_t getContextEndGpuAddress(const TagNodeBase &timestampPacketNode) {
return timestampPacketNode.getGpuAddress() + timestampPacketNode.getContextEndOffset();
}
static uint64_t getContextStartGpuAddress(const TagNode<TimestampPacketStorage> &timestampPacketNode) {
return timestampPacketNode.getGpuAddress() + timestampPacketNode.tagForCpuAccess->getContextStartOffset();
static uint64_t getContextStartGpuAddress(const TagNodeBase &timestampPacketNode) {
return timestampPacketNode.getGpuAddress() + timestampPacketNode.getContextStartOffset();
}
static uint64_t getGlobalEndGpuAddress(const TagNode<TimestampPacketStorage> &timestampPacketNode) {
return timestampPacketNode.getGpuAddress() + timestampPacketNode.tagForCpuAccess->getGlobalEndOffset();
static uint64_t getGlobalEndGpuAddress(const TagNodeBase &timestampPacketNode) {
return timestampPacketNode.getGpuAddress() + timestampPacketNode.getGlobalEndOffset();
}
static uint64_t getGlobalStartGpuAddress(const TagNode<TimestampPacketStorage> &timestampPacketNode) {
return timestampPacketNode.getGpuAddress() + timestampPacketNode.tagForCpuAccess->getGlobalStartOffset();
static uint64_t getGlobalStartGpuAddress(const TagNodeBase &timestampPacketNode) {
return timestampPacketNode.getGpuAddress() + timestampPacketNode.getGlobalStartOffset();
}
static uint64_t getGpuDependenciesCountGpuAddress(const TagNode<TimestampPacketStorage> &timestampPacketNode) {
return timestampPacketNode.getGpuAddress() + timestampPacketNode.tagForCpuAccess->getImplicitGpuDependenciesCountOffset();
static uint64_t getGpuDependenciesCountGpuAddress(const TagNodeBase &timestampPacketNode) {
return timestampPacketNode.getGpuAddress() + timestampPacketNode.getImplicitGpuDependenciesCountOffset();
}
static void overrideSupportedDevicesCount(uint32_t &numSupportedDevices);
template <typename GfxFamily>
static void programSemaphoreWithImplicitDependency(LinearStream &cmdStream, TagNode<TimestampPacketStorage> &timestampPacketNode, uint32_t numSupportedDevices) {
static void programSemaphoreWithImplicitDependency(LinearStream &cmdStream, TagNodeBase &timestampPacketNode, uint32_t numSupportedDevices) {
using MI_ATOMIC = typename GfxFamily::MI_ATOMIC;
using COMPARE_OPERATION = typename GfxFamily::MI_SEMAPHORE_WAIT::COMPARE_OPERATION;
using MI_SEMAPHORE_WAIT = typename GfxFamily::MI_SEMAPHORE_WAIT;
@@ -158,8 +158,8 @@ struct TimestampPacketHelper {
auto compareAddress = getContextEndGpuAddress(timestampPacketNode);
auto dependenciesCountAddress = getGpuDependenciesCountGpuAddress(timestampPacketNode);
for (uint32_t packetId = 0; packetId < timestampPacketNode.tagForCpuAccess->getPacketsUsed(); packetId++) {
uint64_t compareOffset = packetId * sizeof(TimestampPacketStorage::Packet);
for (uint32_t packetId = 0; packetId < timestampPacketNode.getPacketsUsed(); packetId++) {
uint64_t compareOffset = packetId * timestampPacketNode.getSinglePacketSize();
EncodeSempahore<GfxFamily>::addMiSemaphoreWaitCommand(cmdStream, compareAddress + compareOffset, 1, COMPARE_OPERATION::COMPARE_OPERATION_SAD_NOT_EQUAL_SDD);
}
@@ -231,8 +231,8 @@ struct TimestampPacketHelper {
}
template <typename GfxFamily>
static size_t getRequiredCmdStreamSizeForNodeDependency(TagNode<TimestampPacketStorage> &timestampPacketNode) {
size_t totalMiSemaphoreWaitSize = timestampPacketNode.tagForCpuAccess->getPacketsUsed() * sizeof(typename GfxFamily::MI_SEMAPHORE_WAIT);
static size_t getRequiredCmdStreamSizeForNodeDependency(TagNodeBase &timestampPacketNode) {
size_t totalMiSemaphoreWaitSize = timestampPacketNode.getPacketsUsed() * sizeof(typename GfxFamily::MI_SEMAPHORE_WAIT);
return totalMiSemaphoreWaitSize + sizeof(typename GfxFamily::MI_ATOMIC);
}