Update on "[2/2] Intel GPU Runtime Upstreaming for Generator"

# Motivation
According to [[1/2] Intel GPU Runtime Upstreaming for Generator](#118528), as mentioned in [[RFC] Intel GPU Runtime Upstreaming](#114842), the second PR covers the changes under `python frontend`.

# Design
Currently, it primarily offers geneartor-related APIs, including

- `torch.xpu.default_generators`
- `torch.xpu.get_rng_state`
- `torch.xpu.get_rng_state_all`
- `torch.xpu.initial_seed`
- `torch.xpu.manual_seed`
- `torch.xpu.manual_seed_all`
- `torch.xpu.seed`
- `torch.xpu.seed_all`
- `torch.xpu.set_rng_state`
- `torch.xpu.set_rng_state_all`

# Additional Context
The differences with CUDA:
The generator-related frontend python APIs are 1:1 mapping with CUDA.

cc jgong5 mingfeima XiaobingSuper sanchitintel ashokei jingxu10

[ghstack-poisoned]

aten/src/ATen/detail/XPUHooksInterface.h

@@ -43,6 +43,10 @@ struct TORCH_API XPUHooksInterface {
     TORCH_CHECK(false, "Cannot get XPU global device index without ATen_xpu library.");
   }
 
+  virtual Generator getXPUGenerator(C10_UNUSED DeviceIndex device_index = -1) const {
+    TORCH_CHECK(false, "Cannot get XPU generator without Intel Extension for Pytorch. ", XPU_HELP);
+  }
+
   virtual const Generator& getDefaultXPUGenerator(C10_UNUSED DeviceIndex device_index = -1) const {
     TORCH_CHECK(false, "Cannot get default XPU generator without Intel Extension for Pytorch. ", XPU_HELP);
   }
@@ -52,7 +56,7 @@ struct TORCH_API XPUHooksInterface {
   }
 
   virtual DeviceIndex current_device() const {
-    return -1;
+    TORCH_CHECK(false, "Cannot get current device on XPU without ATen_xpu library.");
   }
 
   virtual Device getDeviceFromPtr(void* /*data*/) const {

aten/src/ATen/xpu/detail/XPUHooks.cpp

@@ -27,6 +27,10 @@ int XPUHooks::getGlobalIdxFromDevice(const at::Device& device) const {
   return at::xpu::getGlobalIdxFromDevice(device.index());
 }
 
+Generator XPUHooks::getXPUGenerator(DeviceIndex device_index) const {
+  return make_generator<at::XPUGeneratorImpl>(device_index);
+}
+
 const Generator& XPUHooks::getDefaultXPUGenerator(
     DeviceIndex device_index) const {
   return at::xpu::detail::getDefaultXPUGenerator(device_index);

aten/src/ATen/xpu/detail/XPUHooks.h

@@ -14,6 +14,7 @@ struct XPUHooks : public at::XPUHooksInterface {
   bool hasXPU() const override;
   std::string showConfig() const override;
   int getGlobalIdxFromDevice(const at::Device& device) const override;
+  Generator getXPUGenerator(DeviceIndex device_index = -1) const override;
   const Generator& getDefaultXPUGenerator(
       DeviceIndex device_index = -1) const override;
   Device getDeviceFromPtr(void* data) const override;

torch/csrc/Generator.cpp

@@ -25,10 +25,6 @@
 #include <ATen/mps/MPSGeneratorImpl.h>
 #endif
 
-#ifdef USE_XPU
-#include <ATen/xpu/XPUGeneratorImpl.h>
-#endif
-
 using namespace at;
 using namespace torch;
 
@@ -76,12 +72,9 @@ static PyObject* THPGenerator_pynew(
     self->cdata = make_generator<MPSGeneratorImpl>();
   }
 #endif
-#ifdef USE_XPU
   else if (device.type() == at::kXPU) {
-    self->cdata = make_generator<XPUGeneratorImpl>(device.index());
-  }
-#endif
-  else if (device.type() == at::kIPU) {
+    self->cdata = at::detail::getXPUHooks().getXPUGenerator(device.index());
+  } else if (device.type() == at::kIPU) {
     self->cdata = at::detail::getIPUHooks().newIPUGenerator(device.index());
   } else if (device.type() == at::kPrivateUse1) {
     self->cdata = at::GetGeneratorForPrivateuse1(device.index());