IORequest: Refactor IOOperation transferred-bytes and status accounting.

Until the introduction of the nvme_disk driver, these classes were
mostly only used directly by the IO scheduler, and then a few direct
usages of IOOperation itself in the individual disk drivers; so
API confusions were easily missed.

When writing the nvme_disk driver's IORequest support, however, it
became readily apparent that there were some pretty bad confusions
around transferred-bytes accounting in IOOperation. This commit
attempts to resolve all of those.

There are two basic changes here:

1. Move transferred-bytes accounting into IOOperation::SetStatus.

The "TransferredBytes" field of IOOperation is against the *original*
range, not the actual operation's range (which will be wider, due to
bouncing, etc.), and furthermore only applies to the actual content
of the request (and not e.g. to a read half of a bounced write.)

These two facts meant that determining what value to pass to
SetTransferredBytes was not trivial, and was easy to get wrong.
I recall messing that up when working on nvme_disk multiple times
before reading the API carefully.

2. Do not pass redundant values to IORequest::OperationFinished.

All of the values here can be derived (albeit indirectly) from the
IOOperation, and all consumers of this API basically did just that.
Rather than make them do it, make the IORequest take care of
computing all of those values itself.

Change-Id: Ic9ae29e1100319e5b7647647c4db7e5aad4d125e

src/add-ons/kernel/drivers/disk/nvme/nvme_disk.cpp

@@ -570,7 +570,6 @@ nvme_disk_bounced_io(nvme_disk_handle* handle, io_request* request)
 		if (status != B_OK)
 			break;
 
-		size_t transferredBytes = 0;
 		do {
 			TRACE("%p: IOO offset: %" B_PRIdOFF ", length: %" B_PRIuGENADDR
 				", write: %s\n", request, operation.Offset(),
@@ -583,10 +582,9 @@ nvme_disk_bounced_io(nvme_disk_handle* handle, io_request* request)
 			nvme_request.iovec_count = operation.VecCount();
 
 			status = do_nvme_io_request(handle->info, &nvme_request);
-			if (status == B_OK && nvme_request.write == request->IsWrite())
-				transferredBytes += operation.OriginalLength();
 
-			operation.SetStatus(status);
+			operation.SetStatus(status,
+				status == B_OK ? operation.Length() : 0);
 		} while (status == B_OK && !operation.Finish());
 
 		if (status == B_OK && operation.Status() != B_OK) {
@@ -594,9 +592,7 @@ nvme_disk_bounced_io(nvme_disk_handle* handle, io_request* request)
 			status = operation.Status();
 		}
 
-		operation.SetTransferredBytes(transferredBytes);
+		request->OperationFinished(&operation);
-		request->OperationFinished(&operation, status, status != B_OK,
-			operation.OriginalOffset() + transferredBytes);
 
 		handle->info->dma_resource.RecycleBuffer(operation.Buffer());
 

src/system/kernel/device_manager/IOCache.cpp

@@ -203,6 +203,8 @@ void
 IOCache::OperationCompleted(IOOperation* operation, status_t status,
 	generic_size_t transferredBytes)
 {
+	operation->SetStatus(status, transferredBytes);
+
 	if (status == B_OK) {
 		// always fail in case of partial transfers
 		((Operation*)operation)->finishedCondition.NotifyAll(
@@ -466,8 +468,7 @@ IOCache::_TransferRequestLineUncached(IORequest* request, off_t lineOffset,
 
 		error = _DoOperation(operation);
 
-		request->OperationFinished(&operation, error, false,
+		request->OperationFinished(&operation);
-			error == B_OK ? operation.OriginalLength() : 0);
 		request->SetUnfinished();
 			// Keep the request in unfinished state. ScheduleRequest() will set
 			// the final status and notify.

src/system/kernel/device_manager/IORequest.cpp

@@ -291,10 +291,33 @@ IOBuffer::Dump() const
 // #pragma mark -
 
 
+void
+IOOperation::SetStatus(status_t status, generic_size_t completedLength)
+{
+	IORequestChunk::SetStatus(status);
+	if (IsWrite() == fParent->IsWrite()) {
+		// Determine how many bytes we actually read or wrote,
+		// relative to the original range, not the translated range.
+		const generic_size_t partialBegin = (fOriginalOffset - fOffset);
+		generic_size_t originalTransferredBytes = completedLength;
+		if (originalTransferredBytes < partialBegin)
+			originalTransferredBytes = 0;
+		else
+			originalTransferredBytes -= partialBegin;
+
+		if (originalTransferredBytes > fOriginalLength)
+			originalTransferredBytes = fOriginalLength;
+
+		fTransferredBytes += originalTransferredBytes;
+	}
+}
+
+
 bool
 IOOperation::Finish()
 {
 	TRACE("IOOperation::Finish()\n");
+
 	if (fStatus == B_OK) {
 		if (fParent->IsWrite()) {
 			TRACE("  is write\n");
@@ -318,7 +341,7 @@ IOOperation::Finish()
 					return false;
 				}
 
-				SetStatus(error);
+				IORequestChunk::SetStatus(error);
 			} else if (fPhase == PHASE_READ_END) {
 				TRACE("  phase read end\n");
 				// repair phase adjusted vec
@@ -338,7 +361,7 @@ IOOperation::Finish()
 					return false;
 				}
 
-				SetStatus(error);
+				IORequestChunk::SetStatus(error);
 			}
 		}
 	}
@@ -402,7 +425,7 @@ IOOperation::Finish()
 		}
 
 		if (error != B_OK)
-			SetStatus(error);
+			IORequestChunk::SetStatus(error);
 	}
 
 	return true;
@@ -1012,8 +1035,7 @@ IORequest::SetStatusAndNotify(status_t status)
 
 
 void
-IORequest::OperationFinished(IOOperation* operation, status_t status,
+IORequest::OperationFinished(IOOperation* operation)
-	bool partialTransfer, generic_size_t transferEndOffset)
 {
 	TRACE("IORequest::OperationFinished(%p, %#" B_PRIx32 "): request: %p\n",
 		operation, status, this);
@@ -1023,6 +1045,12 @@ IORequest::OperationFinished(IOOperation* operation, status_t status,
 	fChildren.Remove(operation);
 	operation->SetParent(NULL);
 
+	const status_t status = operation->Status();
+	const bool partialTransfer =
+		(operation->TransferredBytes() < operation->OriginalLength());
+	const generic_size_t transferEndOffset =
+		(operation->OriginalOffset() + operation->TransferredBytes());
+
 	if (status != B_OK || partialTransfer) {
 		if (fTransferSize > transferEndOffset)
 			fTransferSize = transferEndOffset;

src/system/kernel/device_manager/IORequest.h

@@ -127,6 +127,9 @@ struct IOOperation : IORequestChunk, DoublyLinkedListLinkImpl<IOOperation> {
 public:
 			bool				Finish();
 									// returns true, if it can be recycled
+									// otherwise, there is more to be done
+
+			void				SetStatus(status_t status, generic_size_t completedLength);
 
 			status_t			Prepare(IORequest* request);
 			void				SetOriginalRange(off_t offset,
@@ -134,11 +137,9 @@ public:
 									// also sets range
 			void				SetRange(off_t offset, generic_size_t length);
 
-			void				SetStatus(status_t status)
-									{ IORequestChunk::SetStatus(status); }
-
 			off_t				Offset() const;
 			generic_size_t		Length() const;
+
 			off_t				OriginalOffset() const
 									{ return fOriginalOffset; }
 			generic_size_t		OriginalLength() const
@@ -146,8 +147,6 @@ public:
 
 			generic_size_t		TransferredBytes() const
 									{ return fTransferredBytes; }
-			void				SetTransferredBytes(generic_size_t bytes)
-									{ fTransferredBytes = bytes; }
 
 			generic_io_vec*		Vecs() const;
 			uint32				VecCount() const;
@@ -256,9 +255,7 @@ struct IORequest : IORequestChunk, DoublyLinkedListLinkImpl<IORequest> {
 			bool				HasCallbacks() const;
 			void				SetStatusAndNotify(status_t status);
 
-			void				OperationFinished(IOOperation* operation,
+			void				OperationFinished(IOOperation* operation);
-									status_t status, bool partialTransfer,
-									generic_size_t transferEndOffset);
 			void				SubRequestFinished(IORequest* request,
 									status_t status, bool partialTransfer,
 									generic_size_t transferEndOffset);

src/system/kernel/device_manager/IOSchedulerSimple.cpp

@@ -291,13 +291,7 @@ IOSchedulerSimple::OperationCompleted(IOOperation* operation, status_t status,
 	if (operation->Status() <= 0)
 		return;
 
-	operation->SetStatus(status);
+	operation->SetStatus(status, transferredBytes);
-
-	// set the bytes transferred (of the net data)
-	generic_size_t partialBegin
-		= operation->OriginalOffset() - operation->Offset();
-	operation->SetTransferredBytes(
-		transferredBytes > partialBegin ? transferredBytes - partialBegin : 0);
 
 	fCompletedOperations.Add(operation);
 	fFinishedOperationCondition.NotifyAll();
@@ -344,7 +338,6 @@ IOSchedulerSimple::_Finisher()
 		if (!operationFinished) {
 			TRACE("  operation: %p not finished yet\n", operation);
 			MutexLocker _(fLock);
-			operation->SetTransferredBytes(0);
 			operation->Parent()->Owner()->operations.Add(operation);
 			fPendingOperations--;
 			continue;
@@ -353,13 +346,7 @@ IOSchedulerSimple::_Finisher()
 		// notify request and remove operation
 		IORequest* request = operation->Parent();
 
-		generic_size_t operationOffset
+		request->OperationFinished(operation);
-			= operation->OriginalOffset() - request->Offset();
-		request->OperationFinished(operation, operation->Status(),
-			operation->TransferredBytes() < operation->OriginalLength(),
-			operation->Status() == B_OK
-				? operationOffset + operation->OriginalLength()
-				: operationOffset);
 
 		// recycle the operation
 		MutexLocker _(fLock);