IntenseWebs/tuxclocker
3
0
Fork 0
mirror of https://github.com/Lurkki14/tuxclocker.git synced 2024-11-25 01:30:18 -06:00
Code Issues Packages Projects Releases Wiki Activity

nvidia: fix invalid function call by capturing by value instead of reference

Browse Source
This commit is contained in:
Jussi Kuokkanen 2020-04-06 01:17:56 +03:00
parent 9410102c89
commit 0a2bf68ac0
1 changed files with 13 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
src/plugins/Nvidia.cpp
View File

@ -172,7 +172,6 @@ NvidiaPlugin::NvidiaPlugin() : m_dpy() {
}
};
struct NvidiaGPUDataOpt {
std::string uuid;
std::string name;
@ -247,7 +246,7 @@ NvidiaPlugin::NvidiaPlugin() : m_dpy() {
gi++;
};
// [GPUOpt] -> [TreeNode DeviceNode]
// Map GPU data to readables
@ -257,20 +256,24 @@ NvidiaPlugin::NvidiaPlugin() : m_dpy() {
// Get nvml nodes if there is a device
if_let(pattern(some(arg)) = nvOpt.devHandle) = [&](auto dev) {
auto availableNVMLNodes = fp::keep_if([dev](auto rawNode) {
// These might be similar enough to make sense to be templates
auto availableNVMLNodes = fp::keep_if([&](auto rawNode) {
auto result = rawNode.func(dev);
match(result)(pattern(as<ReadError>(_)) = []() {return false;},
pattern(_) = []() {return true;});
return true;
}, rawNVMLNodes);
auto specializedNVMLNodes = fp::transform([&](auto rawNode) {
// After a day of debugging, I found out that this needs to be captured by value instead of reference or we get a bad function call when this goes out of scope
auto specializedNVMLNodes = fp::transform([=](auto rawNode) {
auto readable = DynamicReadable(
[&]() {return rawNode.func(dev);});
[=]() {return rawNode.func(dev);});
auto devNode = DeviceNode{.name = rawNode.nodeName,
.interface = readable
};
return devNode;
}, availableNVMLNodes);
for (auto &specNode : specializedNVMLNodes) {
gpuRoot.appendChild(specNode);
}
@ -284,9 +287,9 @@ NvidiaPlugin::NvidiaPlugin() : m_dpy() {
pattern(_) = []() {return true;});
return true;
}, rawNVCTRLNodes);
auto specializedNodes = fp::transform([index](auto rawNode) {
auto specializedNodes = fp::transform([=](auto rawNode) {
auto readable = DynamicReadable(
[&]() {return rawNode.func(index);});
[=]() {return rawNode.func(index);});
auto devNode = DeviceNode{.name = rawNode.nodeName,
.interface = readable
};
@ -297,8 +300,10 @@ NvidiaPlugin::NvidiaPlugin() : m_dpy() {
return gpuRoot;
}, optDataVec);
for (auto &gpuNode : gpuNodes) rootNode.appendChild(gpuNode);
//auto res = std::get<DynamicReadable>(gpuNodes[0].children()[0].value().interface.value()).read();
for (auto &gpuNode : gpuNodes) rootNode.appendChild(gpuNode);
m_rootDeviceNode = rootNode;
/*std::cout << nvctrlGPUCount << "GPUs\n";
for (auto nvmlGPU : nvmlGPUVec) {