fix: write memory for resident allocations in simulation mode

- refactor and call proceesFlushResdiency() on memoryOperationsHandler
- call free() to remove allocation from resident allocations when
graphics allocation is released

Related-To: NEO-11719

Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>

shared/test/common/mocks/mock_aub_csr.h

@@ -1,5 +1,5 @@
 /*
- * Copyright (C) 2018-2023 Intel Corporation
+ * Copyright (C) 2018-2024 Intel Corporation
  *
  * SPDX-License-Identifier: MIT
  *
@@ -93,6 +93,10 @@ struct MockAubCsr : public AUBCommandStreamReceiverHw<GfxFamily> {
         AUBCommandStreamReceiverHw<GfxFamily>::writeMemory(gpuAddress, cpuAddress, size, memoryBank, entryBits);
         writeMemoryCalled = true;
     }
+    bool writeMemory(GraphicsAllocation &gfxAllocation) override {
+        writeMemoryGfxAllocCalled = true;
+        return AUBCommandStreamReceiverHw<GfxFamily>::writeMemory(gfxAllocation);
+    }
     void writeMMIO(uint32_t offset, uint32_t value) override {
         AUBCommandStreamReceiverHw<GfxFamily>::writeMMIO(offset, value);
         writeMMIOCalled = true;
@@ -146,6 +150,7 @@ struct MockAubCsr : public AUBCommandStreamReceiverHw<GfxFamily> {
     bool initProgrammingFlagsCalled = false;
     bool initializeEngineCalled = false;
     bool writeMemoryCalled = false;
+    bool writeMemoryGfxAllocCalled = false;
     bool writeMemoryWithAubManagerCalled = false;
     bool writeMMIOCalled = false;
     bool submitBatchBufferCalled = false;

shared/test/common/mocks/mock_aub_memory_operations_handler.h

@@ -1,5 +1,5 @@
 /*
- * Copyright (C) 2023 Intel Corporation
+ * Copyright (C) 2023-2024 Intel Corporation
  *
  * SPDX-License-Identifier: MIT
  *
@@ -46,11 +46,17 @@ struct MockAubMemoryOperationsHandler : public AubMemoryOperationsHandler {
         return AubMemoryOperationsHandler::evictWithinOsContext(osContext, gfxAllocation);
     }
 
+    MemoryOperationsStatus free(Device *device, GraphicsAllocation &gfxAllocation) override {
+        freeCalled = true;
+        return AubMemoryOperationsHandler::free(device, gfxAllocation);
+    }
+
     bool makeResidentCalled = false;
     bool evictCalled = false;
     bool isResidentCalled = false;
     bool makeResidentWithinOsContextCalled = false;
     bool evictWithinOsContextCalled = false;
+    bool freeCalled = false;
 };
 
 } // namespace NEO

shared/test/common/mocks/mock_memory_operations_handler.h

@@ -102,9 +102,23 @@ class MockMemoryOperations : public MemoryOperationsHandler {
         return MemoryOperationsStatus::success;
     }
 
+    MemoryOperationsStatus free(Device *device, GraphicsAllocation &gfxAllocation) override {
+        freeCalledCount++;
+
+        if (captureGfxAllocationsForMakeResident) {
+            auto itor = std::find(gfxAllocationsForMakeResident.begin(), gfxAllocationsForMakeResident.end(), &gfxAllocation);
+            if (itor != gfxAllocationsForMakeResident.end()) {
+                gfxAllocationsForMakeResident.erase(itor, itor + 1);
+            }
+        }
+
+        return MemoryOperationsStatus::success;
+    }
+
     std::vector<GraphicsAllocation *> gfxAllocationsForMakeResident{};
     int makeResidentCalledCount = 0;
     int evictCalledCount = 0;
+    int freeCalledCount = 0;
     uint32_t isResidentCalledCount = 0;
     uint32_t lockCalledCount = 0;
     uint32_t makeResidentContextId = std::numeric_limits<uint32_t>::max();

shared/test/common/mocks/mock_tbx_csr.h

@@ -1,5 +1,5 @@
 /*
- * Copyright (C) 2018-2023 Intel Corporation
+ * Copyright (C) 2018-2024 Intel Corporation
  *
  * SPDX-License-Identifier: MIT
  *
@@ -42,6 +42,11 @@ class MockTbxCsr : public TbxCommandStreamReceiverHw<GfxFamily> {
         TbxCommandStreamReceiverHw<GfxFamily>::writeMemory(gpuAddress, cpuAddress, size, memoryBank, entryBits);
         writeMemoryCalled = true;
     }
+    bool writeMemory(GraphicsAllocation &graphicsAllocation) override {
+        writeMemoryGfxAllocCalled = true;
+        return TbxCommandStreamReceiverHw<GfxFamily>::writeMemory(graphicsAllocation);
+    }
+
     void submitBatchBufferTbx(uint64_t batchBufferGpuAddress, const void *batchBuffer, size_t batchBufferSize, uint32_t memoryBank, uint64_t entryBits, bool overrideRingHead) override {
         TbxCommandStreamReceiverHw<GfxFamily>::submitBatchBufferTbx(batchBufferGpuAddress, batchBuffer, batchBufferSize, memoryBank, entryBits, overrideRingHead);
         overrideRingHeadPassed = overrideRingHead;
@@ -62,6 +67,7 @@ class MockTbxCsr : public TbxCommandStreamReceiverHw<GfxFamily> {
     bool initializeEngineCalled = false;
     bool writeMemoryWithAubManagerCalled = false;
     bool writeMemoryCalled = false;
+    bool writeMemoryGfxAllocCalled = false;
     bool submitBatchBufferCalled = false;
     bool overrideRingHeadPassed = false;
     bool pollForCompletionCalled = false;