Remove CommandQueue redirecting usages straight to HWCQ (#17219)

### Ticket
#17208

### Problem description
Over time `CommandQueue` stopped playing any particular function.
Today it is simply proxying calls to HWCQ via creating temporary command
structures and immediately routing their execution with a big switch
statement, calling into hwcq methods.
This adds unreasonable complexity. 

After the sync with @tt-asaigal @dmakoviichuk-tt @cfjchu @omilyutin-tt ,
we agreed to remove this class.

### What's changed
Remove CommandQueue class.
Redirect all usage to HWCQ.
HWCommandQueue is renamed to CommandQueue

Note:
* command_queue.hpp was not removed. It contains definitions of commands
used by hardware_command_queue, I would prefer us to clean this up in
the next PR.

### Checklist
- [x] [Post commit
CI](https://github.com/tenstorrent/tt-metal/actions/runs/13023052350)
- [x] [Blackhole Post
commit](https://github.com/tenstorrent/tt-metal/actions/runs/13023279261)
- [x] [T3k
Frequent](https://github.com/tenstorrent/tt-metal/actions/runs/13021799389)
-

docs/source/tt-metalium/tt_metal/apis/host_apis/command_queue/EnqueueWriteSubBuffer.rst

@@ -1,5 +1,5 @@
 EnqueueWriteSubBuffer
 =====================
 
-.. doxygenfunction:: tt::tt_metal::v0::EnqueueWriteSubBuffer(CommandQueue& cq, std::variant<std::reference_wrapper<Buffer>, std::shared_ptr<Buffer> > buffer, HostDataType src, const BufferRegion& region, bool blocking)
-.. doxygenfunction:: tt::tt_metal::v0::EnqueueWriteSubBuffer(CommandQueue& cq, std::variant<std::reference_wrapper<Buffer>, std::shared_ptr<Buffer> > buffer, std::vector<DType>& src, const BufferRegion& region, bool blocking)
+.. doxygenfunction:: tt::tt_metal::v0::EnqueueWriteSubBuffer(CommandQueue& cq, const std::variant<std::reference_wrapper<Buffer>, std::shared_ptr<Buffer> >& buffer, HostDataType src, const BufferRegion& region, bool blocking)
+.. doxygenfunction:: tt::tt_metal::v0::EnqueueWriteSubBuffer(CommandQueue& cq, const std::variant<std::reference_wrapper<Buffer>, std::shared_ptr<Buffer> >& buffer, std::vector<DType>& src, const BufferRegion& region, bool blocking)

tests/tt_metal/tt_metal/api/CMakeLists.txt

@@ -17,7 +17,6 @@ set(UNIT_TESTS_API_SRC
     ${CMAKE_CURRENT_SOURCE_DIR}/test_banked.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/test_bit_utils.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/test_buffer_region.cpp
-    ${CMAKE_CURRENT_SOURCE_DIR}/test_CommandQueue.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/test_direct.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/test_dram_to_l1_multicast.cpp
     ${CMAKE_CURRENT_SOURCE_DIR}/test_dram.cpp

tests/tt_metal/tt_metal/dispatch/dispatch_buffer/test_EnqueueWriteBuffer_and_EnqueueReadBuffer.cpp

@@ -462,52 +462,6 @@ bool test_EnqueueWriteBuffer_and_EnqueueReadBuffer_multi_queue(
 namespace basic_tests {
 namespace dram_tests {
 
-TEST_F(CommandQueueBufferFixture, DISABLED_TestAsyncBufferRW) {
-    // Test Async Enqueue Read and Write + Get Addr + Buffer Allocation and Deallocation
-    auto& command_queue = this->device_->command_queue();
-    auto current_mode = CommandQueue::default_mode();
-    command_queue.set_mode(CommandQueue::CommandQueueMode::ASYNC);
-    Program program;
-    for (int j = 0; j < 10; j++) {
-        // Asynchronously initialize a buffer on device
-        uint32_t first_buf_value = j + 1;
-        uint32_t second_buf_value = j + 2;
-        uint32_t first_buf_size = 4096;
-        uint32_t second_buf_size = 2048;
-        // Asynchronously allocate buffer on device
-        std::shared_ptr<Buffer> buffer =
-            Buffer::create(this->device_, first_buf_size, first_buf_size, BufferType::DRAM);
-        std::shared_ptr<uint32_t> allocated_buffer_address = std::make_shared<uint32_t>();
-        EnqueueGetBufferAddr(this->device_->command_queue(), allocated_buffer_address.get(), buffer.get(), true);
-        // Ensure returned addr is correct
-        EXPECT_EQ((*allocated_buffer_address), buffer->address());
-
-        std::shared_ptr<std::vector<uint32_t>> vec =
-            std::make_shared<std::vector<uint32_t>>(first_buf_size / 4, first_buf_value);
-        std::vector<uint32_t> readback_vec = {};
-        // Write first vector to existing on device buffer.
-        EnqueueWriteBuffer(this->device_->command_queue(), buffer, vec, false);
-        // Reallocate the vector in the main thread after asynchronously pushing it (ensure that worker still has access
-        // to this data)
-        vec = std::make_shared<std::vector<uint32_t>>(second_buf_size / 4, second_buf_value);
-        // Simulate what tt-eager does: Share buffer ownership with program
-        AssignGlobalBufferToProgram(buffer, program);
-        // Reallocate buffer (this is safe, since the program also owns the existing buffer, which will not be
-        // deallocated)
-        buffer = Buffer::create(this->device_, second_buf_size, second_buf_size, BufferType::DRAM);
-        // Write second vector to second buffer
-        EnqueueWriteBuffer(this->device_->command_queue(), buffer, vec, false);
-        // Have main thread give up ownership immediately after writing
-        vec.reset();
-        // Read both buffer and ensure data is correct
-        EnqueueReadBuffer(this->device_->command_queue(), buffer, readback_vec, true);
-        for (int i = 0; i < readback_vec.size(); i++) {
-            EXPECT_EQ(readback_vec[i], second_buf_value);
-        }
-    }
-    command_queue.set_mode(current_mode);
-}
-
 TEST_F(CommandQueueSingleCardBufferFixture, WriteOneTileToDramBank0) {
     TestBufferConfig config = {.num_pages = 1, .page_size = 2048, .buftype = BufferType::DRAM};
     for (IDevice* device : devices_) {

tests/tt_metal/tt_metal/integration/test_flatten.cpp

@@ -299,17 +299,3 @@ TEST_F(DispatchFixture, TensixFlatten) {
         ASSERT_TRUE(test_flatten::flatten(this, this->devices_.at(id), num_tiles_r, num_tiles_c));
     }
 }
-
-TEST_F(CommandQueueProgramFixture, DISABLED_TensixTestAsyncFlattenStress) {
-    auto& command_queue = this->device_->command_queue();
-    auto current_mode = CommandQueue::default_mode();
-    command_queue.set_mode(CommandQueue::CommandQueueMode::ASYNC);
-    uint32_t num_tiles_r = 2;
-    uint32_t num_tiles_c = 2;
-    if (!this->IsSlowDispatch()) {
-        num_tiles_r = 1;
-        num_tiles_c = 1;
-    }
-    ASSERT_TRUE(test_flatten::flatten_stress(this->device_, num_tiles_r, num_tiles_c));
-    command_queue.set_mode(current_mode);
-}

tests/tt_metal/tt_metal/perf_microbenchmark/dispatch/test_dispatcher.cpp

@@ -420,9 +420,6 @@ int main(int argc, char** argv) {
     try {
         int device_id = 0;
         tt_metal::IDevice* device = tt_metal::CreateDevice(device_id);
-
-        CommandQueue& cq = device->command_queue();
-
         tt_metal::Program program = tt_metal::CreateProgram();
 
         CoreCoord spoof_prefetch_core = {0, 0};