diff --git a/Docs/Whats_New.md b/Docs/Whats_New.md
index 31b0bbde..babb789f 100644
--- a/Docs/Whats_New.md
+++ b/Docs/Whats_New.md
@@ -18,7 +18,11 @@ MoltenVK 1.0.34
 
 Released TBD
 
+- Allow zero offset and stride combo in VkVertexInputBindingDescription.
+- Fix potential memory leak on synchronous command buffer submission.
 - Increase shader float constant accuracy beyond 6 digits of precision.
+- MoltenVKShaderConverterTool support cs & csh for compute shader file extensions.
+- `fetchDependencies`: Stop on first error.
 
 
 
@@ -46,6 +50,7 @@ Released 2019/02/28
 - Fix crash from use of MTLDevice registryID on early OS versions.
 - `fetchDependencies`: Fix issue loading from `Vulkan-Portability_repo_revision`.
 - `fetchDependencies`: Clean MoltenVK build to ensure using latest dependency libs.
+- Update `VK_MVK_MOLTENVK_SPEC_VERSION` to 18.
 - Update to latest dependency libraries to support SDK 1.1.101.
 - Update to latest SPIRV-Cross version:
 	- MSL: Implement 8-bit part of `VK_KHR_shader_float16_int8`.
diff --git a/MoltenVK/MoltenVK/GPUObjects/MVKPipeline.mm b/MoltenVK/MoltenVK/GPUObjects/MVKPipeline.mm
index f2900ab3..84063271 100644
--- a/MoltenVK/MoltenVK/GPUObjects/MVKPipeline.mm
+++ b/MoltenVK/MoltenVK/GPUObjects/MVKPipeline.mm
@@ -839,16 +839,23 @@ bool MVKGraphicsPipeline::addVertexInputToPipeline(MTLRenderPipelineDescriptor*
     for (uint32_t i = 0; i < vaCnt; i++) {
         const VkVertexInputAttributeDescription* pVKVA = &pVI->pVertexAttributeDescriptions[i];
         if (shaderContext.isVertexAttributeLocationUsed(pVKVA->location)) {
-            // Vulkan allows offsets to exceed the buffer stride, but Metal doesn't.
-            const VkVertexInputBindingDescription* pVKVB = pVI->pVertexBindingDescriptions;
-            for (uint32_t j = 0; j < vbCnt; j++, pVKVB++) {
-                if (pVKVB->binding == pVKVA->binding) { break; }
-            }
-            if (pVKVA->offset >= pVKVB->stride) {
-                setConfigurationResult(mvkNotifyErrorWithText(VK_ERROR_INITIALIZATION_FAILED, "Under Metal, vertex attribute offsets must not exceed the vertex buffer stride."));
-                return false;
-            }
-            MTLVertexAttributeDescriptor* vaDesc = plDesc.vertexDescriptor.attributes[pVKVA->location];
+
+      // Vulkan allows offsets to exceed the buffer stride, but Metal doesn't.
+			// Only check non-zero offsets, as it's common for both to be zero when step rate is instance.
+			if (pVKVA->offset > 0) {
+				const VkVertexInputBindingDescription* pVKVB = pVI->pVertexBindingDescriptions;
+				for (uint32_t j = 0; j < vbCnt; j++, pVKVB++) {
+					if (pVKVB->binding == pVKVA->binding) {
+						if (pVKVA->offset >= pVKVB->stride) {
+							setConfigurationResult(mvkNotifyErrorWithText(VK_ERROR_INITIALIZATION_FAILED, "Under Metal, vertex attribute offsets must not exceed the vertex buffer stride."));
+							return false;
+						}
+						break;
+					}
+				}
+			}
+
+			MTLVertexAttributeDescriptor* vaDesc = plDesc.vertexDescriptor.attributes[pVKVA->location];
             vaDesc.format = mvkMTLVertexFormatFromVkFormat(pVKVA->format);
             vaDesc.bufferIndex = _device->getMetalBufferIndexForVertexAttributeBinding(pVKVA->binding);
             vaDesc.offset = pVKVA->offset;
@@ -860,15 +867,16 @@ bool MVKGraphicsPipeline::addVertexInputToPipeline(MTLRenderPipelineDescriptor*
         const VkVertexInputBindingDescription* pVKVB = &pVI->pVertexBindingDescriptions[i];
         uint32_t vbIdx = _device->getMetalBufferIndexForVertexAttributeBinding(pVKVB->binding);
         if (shaderContext.isVertexBufferUsed(vbIdx)) {
-            MTLVertexBufferLayoutDescriptor* vbDesc = plDesc.vertexDescriptor.layouts[vbIdx];
-            // Vulkan allows any stride, but Metal only allows multiples of 4.
-            // TODO: We should try to expand the buffer to the required alignment
-            // in that case.
+
+			// Vulkan allows any stride, but Metal only allows multiples of 4.
+            // TODO: We should try to expand the buffer to the required alignment in that case.
             if ((pVKVB->stride % 4) != 0) {
                 setConfigurationResult(mvkNotifyErrorWithText(VK_ERROR_INITIALIZATION_FAILED, "Under Metal, vertex buffer strides must be aligned to four bytes."));
                 return false;
             }
-            vbDesc.stride = (pVKVB->stride == 0) ? sizeof(simd::float4) : pVKVB->stride;      // Vulkan allows zero stride but Metal doesn't. Default to float4
+
+			MTLVertexBufferLayoutDescriptor* vbDesc = plDesc.vertexDescriptor.layouts[vbIdx];
+			vbDesc.stride = (pVKVB->stride == 0) ? sizeof(simd::float4) : pVKVB->stride;      // Vulkan allows zero stride but Metal doesn't. Default to float4
             vbDesc.stepFunction = mvkMTLVertexStepFunctionFromVkVertexInputRate(pVKVB->inputRate);
             vbDesc.stepRate = 1;
         }
diff --git a/MoltenVK/MoltenVK/GPUObjects/MVKQueue.mm b/MoltenVK/MoltenVK/GPUObjects/MVKQueue.mm
index 2b9f79d2..102d93a4 100644
--- a/MoltenVK/MoltenVK/GPUObjects/MVKQueue.mm
+++ b/MoltenVK/MoltenVK/GPUObjects/MVKQueue.mm
@@ -63,6 +63,10 @@ MVKQueueFamily::~MVKQueueFamily() {
 
 #pragma mark Queue submissions
 
+// Execute the queue submission under an autorelease pool to ensure transient Metal objects are autoreleased.
+// This is critical for apps that don't use standard OS autoreleasing runloop threading.
+static inline void execute(MVKQueueSubmission* qSubmit) { @autoreleasepool { qSubmit->execute(); } }
+
 // Executes the submmission, either immediately, or by dispatching to an execution queue.
 // Submissions to the execution queue are wrapped in a dedicated autorelease pool.
 // Relying on the dispatch queue to find time to drain the autorelease pool can
@@ -72,9 +76,9 @@ VkResult MVKQueue::submit(MVKQueueSubmission* qSubmit) {
 
 	VkResult rslt = qSubmit->_submissionResult;     // Extract result before submission to avoid race condition with early destruction
 	if (_execQueue) {
-		dispatch_async(_execQueue, ^{ @autoreleasepool { qSubmit->execute(); } } );
+		dispatch_async(_execQueue, ^{ execute(qSubmit); } );
 	} else {
-		qSubmit->execute();
+		execute(qSubmit);
 	}
 	return rslt;
 }
diff --git a/MoltenVKShaderConverter/MoltenVKShaderConverterTool/MoltenVKShaderConverterTool.cpp b/MoltenVKShaderConverter/MoltenVKShaderConverterTool/MoltenVKShaderConverterTool.cpp
index a8980ad8..fd757083 100644
--- a/MoltenVKShaderConverter/MoltenVKShaderConverterTool/MoltenVKShaderConverterTool.cpp
+++ b/MoltenVKShaderConverter/MoltenVKShaderConverterTool/MoltenVKShaderConverterTool.cpp
@@ -34,7 +34,7 @@ static const char* _defaultVertexShaderExtns = "vs vsh vert vertex";
 static const char* _defaultFragShaderExtns = "fs fsh frag fragment";
 
 // The default list of compute file extensions.
-static const char* _defaultCompShaderExtns = "cp cmp comp compute kn kl krn kern kernel";
+static const char* _defaultCompShaderExtns = "cs csh cp cmp comp compute kn kl krn kern kernel";
 
 // The default list of SPIR-V file extensions.
 static const char* _defaultSPIRVShaderExtns = "spv spirv";