Skip to content
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

Jump to bottom

feature: add support for querying number of l3 banks from xe kmd #722

Closed
wants to merge 1 commit into from
Closed

feature: add support for querying number of l3 banks from xe kmd #722

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
17 changes: 15 additions & 2 deletions shared/source/os_interface/linux/xe/ioctl_helper_xe.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -347,9 +347,10 @@ bool IoctlHelperXe::setGpuCpuTimes(TimeStampData *pGpuCpuTime, OSTime *osTime) {
}

void IoctlHelperXe::getTopologyData(size_t nTiles, std::vector<std::bitset<8>> *geomDss, std::vector<std::bitset<8>> *computeDss,
std::vector<std::bitset<8>> *euDss, DrmQueryTopologyData &topologyData, bool &isComputeDssEmpty) {
std::vector<std::bitset<8>> *euDss, std::vector<std::bitset<8>> *l3BanksMask, DrmQueryTopologyData &topologyData, bool &isComputeDssEmpty) {
int subSliceCount = 0;
int euPerDss = 0;
int l3Banks = 0;

for (auto tileId = 0u; tileId < nTiles; tileId++) {

Expand All @@ -371,9 +372,15 @@ void IoctlHelperXe::getTopologyData(size_t nTiles, std::vector<std::bitset<8>> *
euPerDssPerTile += euDss[tileId][byte].count();
}

int l3BanksPerTile = 0;
for (auto byte = 0u; byte < l3BanksMask[tileId].size(); byte++) {
l3BanksPerTile += l3BanksMask[tileId][byte].count();
}

// pick smallest config
subSliceCount = (subSliceCount == 0) ? subSliceCountPerTile : std::min(subSliceCount, subSliceCountPerTile);
euPerDss = (euPerDss == 0) ? euPerDssPerTile : std::min(euPerDss, euPerDssPerTile);
l3Banks = (l3Banks == 0) ? l3BanksPerTile : std::min(l3Banks, l3BanksPerTile);

// pick max config
topologyData.maxSubSliceCount = std::max(topologyData.maxSubSliceCount, subSliceCountPerTile);
Expand All @@ -384,6 +391,7 @@ void IoctlHelperXe::getTopologyData(size_t nTiles, std::vector<std::bitset<8>> *
topologyData.subSliceCount = subSliceCount;
topologyData.euCount = subSliceCount * euPerDss;
topologyData.maxSliceCount = 1;
topologyData.numL3Banks = l3Banks;
}

void IoctlHelperXe::getTopologyMap(size_t nTiles, std::vector<std::bitset<8>> *dssInfo, TopologyMap &topologyMap) {
Expand Down Expand Up @@ -420,6 +428,7 @@ bool IoctlHelperXe::getTopologyDataAndMap(const HardwareInfo &hwInfo, DrmQueryTo
StackVec<std::vector<std::bitset<8>>, 2> geomDss;
StackVec<std::vector<std::bitset<8>>, 2> computeDss;
StackVec<std::vector<std::bitset<8>>, 2> euDss;
StackVec<std::vector<std::bitset<8>>, 2> l3Banks;

auto topologySize = queryGtTopology.size();
auto dataPtr = queryGtTopology.data();
Expand All @@ -428,6 +437,7 @@ bool IoctlHelperXe::getTopologyDataAndMap(const HardwareInfo &hwInfo, DrmQueryTo
geomDss.resize(numTiles);
computeDss.resize(numTiles);
euDss.resize(numTiles);
l3Banks.resize(numTiles);
bool receivedDssInfo = false;
bool receivedEuPerDssInfo = false;
while (topologySize >= sizeof(drm_xe_query_topology_mask)) {
Expand All @@ -451,6 +461,9 @@ bool IoctlHelperXe::getTopologyDataAndMap(const HardwareInfo &hwInfo, DrmQueryTo
fillMask(euDss[tileId], topo);
receivedEuPerDssInfo = true;
break;
case DRM_XE_TOPO_L3_BANK:
fillMask(l3Banks[tileId], topo);
break;
default:
xeLog("Unhandle GT Topo type: %d\n", topo->type);
}
Expand All @@ -462,7 +475,7 @@ bool IoctlHelperXe::getTopologyDataAndMap(const HardwareInfo &hwInfo, DrmQueryTo
}

bool isComputeDssEmpty = false;
getTopologyData(numTiles, geomDss.begin(), computeDss.begin(), euDss.begin(), topologyData, isComputeDssEmpty);
getTopologyData(numTiles, geomDss.begin(), computeDss.begin(), euDss.begin(), l3Banks.begin(), topologyData, isComputeDssEmpty);

auto &dssInfo = isComputeDssEmpty ? geomDss : computeDss;
getTopologyMap(numTiles, dssInfo.begin(), topologyMap);
Expand Down
2 changes: 1 addition & 1 deletion shared/source/os_interface/linux/xe/ioctl_helper_xe.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -105,7 +105,7 @@ class IoctlHelperXe : public IoctlHelper {
bool isWaitBeforeBindRequired(bool bind) const override;
std::unique_ptr<EngineInfo> createEngineInfo(bool isSysmanEnabled) override;
std::unique_ptr<MemoryInfo> createMemoryInfo() override;
void getTopologyData(size_t nTiles, std::vector<std::bitset<8>> *geomDss, std::vector<std::bitset<8>> *computeDss, std::vector<std::bitset<8>> *euDss, DrmQueryTopologyData &topologyData, bool &isComputeDssEmpty);
void getTopologyData(size_t nTiles, std::vector<std::bitset<8>> *geomDss, std::vector<std::bitset<8>> *computeDss, std::vector<std::bitset<8>> *euDss, std::vector<std::bitset<8>> *l3BanksMask, DrmQueryTopologyData &topologyData, bool &isComputeDssEmpty);
void getTopologyMap(size_t nTiles, std::vector<std::bitset<8>> *dssInfo, TopologyMap &topologyMap);

bool setGpuCpuTimes(TimeStampData *pGpuCpuTime, OSTime *osTime) override;
Expand Down
25 changes: 25 additions & 0 deletions shared/test/unit_test/os_interface/linux/xe/ioctl_helper_xe_tests.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -733,6 +733,31 @@ TEST(IoctlHelperXeTest, givenGeomDssWhenGetTopologyDataAndMapThenResultsAreCorre
}
}

TEST(IoctlHelperXeTest, givenL3BankMaskInQueryTopolgyWhenGetTopologyDataAndMapThenProperNumberOfL3BankIsSet) {

auto executionEnvironment = std::make_unique<MockExecutionEnvironment>();
DrmMockXe drm{*executionEnvironment->rootDeviceEnvironments[0]};
auto &hwInfo = *executionEnvironment->rootDeviceEnvironments[0]->getHardwareInfo();
auto xeIoctlHelper = std::make_unique<MockIoctlHelperXe>(drm);

xeIoctlHelper->initialize();

for (auto gtId = 0u; gtId < 3u; gtId++) {
drm.addMockedQueryTopologyData(gtId, DRM_XE_TOPO_DSS_GEOMETRY, 8, {0b11'1111, 0, 0, 0, 0, 0, 0, 0});
drm.addMockedQueryTopologyData(gtId, DRM_XE_TOPO_DSS_COMPUTE, 8, {0, 0, 0, 0, 0, 0, 0, 0});
drm.addMockedQueryTopologyData(gtId, DRM_XE_TOPO_EU_PER_DSS, 8, {0b1111'1111, 0b1111'1111, 0, 0, 0, 0, 0, 0});
drm.addMockedQueryTopologyData(gtId, DRM_XE_TOPO_L3_BANK, 4, {0b1110'1011, 0b1111'1100, 0, 0});
}
DrmQueryTopologyData topologyData{};
TopologyMap topologyMap{};

auto result = xeIoctlHelper->getTopologyDataAndMap(hwInfo, topologyData, topologyMap);
ASSERT_TRUE(result);

// verify topology data
EXPECT_EQ(12, topologyData.numL3Banks);
}

TEST(IoctlHelperXeTest, givenUnknownTopologyTypeWhenGetTopologyDataAndMapThenNotRecognizedTopologyIsIgnored) {

auto executionEnvironment = std::make_unique<MockExecutionEnvironment>();
Expand Down
8 changes: 5 additions & 3 deletions third_party/uapi/xe/xe_drm.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -500,6 +500,7 @@ struct drm_xe_query_gt_list {
* containing the following in mask:
* ``DSS_COMPUTE ff ff ff ff 00 00 00 00``
* means 32 DSS are available for compute.
* - %DRM_XE_TOPO_L3_BANK - To query the mask of enabled L3 banks
* - %DRM_XE_TOPO_EU_PER_DSS - To query the mask of Execution Units (EU)
* available per Dual Sub Slices (DSS). For example a query response
* containing the following in mask:
Expand All @@ -510,9 +511,10 @@ struct drm_xe_query_topology_mask {
/** @gt_id: GT ID the mask is associated with */
__u16 gt_id;

#define DRM_XE_TOPO_DSS_GEOMETRY (1 << 0)
#define DRM_XE_TOPO_DSS_COMPUTE (1 << 1)
#define DRM_XE_TOPO_EU_PER_DSS (1 << 2)
#define DRM_XE_TOPO_DSS_GEOMETRY 1
#define DRM_XE_TOPO_DSS_COMPUTE 2
#define DRM_XE_TOPO_L3_BANK 3
#define DRM_XE_TOPO_EU_PER_DSS 4
/** @type: type of mask */
__u16 type;

Expand Down
Loading