* Added FS interface hooks io() and cancel_io(). The former is supposed

to provide asynchrounous (or only synchronous, if asynchronous is not
  supported) I/O request support. It will eventually replace
  {read,write}_pages(). None of the FS implementations implement them
  yet.
* Implemented some support functions for request-based I/O. File system
  implementations can use do_fd_io() which passes an I/O request to the
  layer responsible for a given FD, and do_iterative_fd_io(), which
  translates a request for a file to subrequests for the underlying
  device and passes them on. Both fall back to synchrounous processing
  when the io() hook is not supported.
  Furthermore added vfs_synchronous_io() which should be handy for the
  devfs to perform io_requests synchronously for devices that don't
  support the io() hook.


git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26655 a95241bf-73f2-0310-859d-f6bbb57e9c96

headers/os/drivers/fs_interface.h

@@ -21,6 +21,9 @@ struct stat;
 struct fs_info;
 struct select_sync;
 
+typedef struct IORequest io_request;
+
+
 /* additional flags passed to write_stat() (see NodeMonitor.h for the others) */
 // NOTE: Changing the constants here or in NodeMonitor.h will break
 // src/kits/storage/LibBeAdapter.cpp:_kern_write_stat().
@@ -124,6 +127,12 @@ struct fs_vnode_ops {
 				void *cookie, off_t pos, const iovec *vecs, size_t count,
 				size_t *_numBytes);
 
+	/* asynchronous I/O */
+	status_t (*io)(fs_volume *volume, fs_vnode *vnode, void *cookie,
+				io_request *request);
+	status_t (*cancel_io)(fs_volume *volume, fs_vnode *vnode, void *cookie,
+				io_request *request);
+
 	/* cache file access */
 	status_t (*get_file_map)(fs_volume *volume, fs_vnode *vnode, off_t offset,
 				size_t size, struct file_io_vec *vecs, size_t *_count);
@@ -284,6 +293,14 @@ typedef struct file_system_module_info {
 
 
 /* file system add-ons only prototypes */
+
+// callbacks for do_iterative_fd_io()
+typedef status_t (*iterative_io_get_vecs)(void *cookie, io_request* request,
+				off_t offset, size_t size, struct file_io_vec *vecs,
+				size_t *_count);
+typedef status_t (*iterative_io_finished)(void* cookie, io_request* request,
+				status_t status);
+
 extern status_t new_vnode(fs_volume *volume, ino_t vnodeID, void *privateNode,
 					fs_vnode_ops *ops);
 extern status_t publish_vnode(fs_volume *volume, ino_t vnodeID,
@@ -309,6 +326,10 @@ extern status_t write_file_io_vec_pages(int fd,
 					const struct file_io_vec *fileVecs, size_t fileVecCount,
 					const struct iovec *vecs, size_t vecCount,
 					uint32 *_vecIndex, size_t *_vecOffset, size_t *_bytes);
+extern status_t do_fd_io(int fd, io_request *request);
+extern status_t do_iterative_fd_io(int fd, io_request *request,
+					iterative_io_get_vecs getVecs,
+					iterative_io_finished finished, void *cookie);
 
 extern status_t notify_entry_created(dev_t device, ino_t directory,
 					const char *name, ino_t node);

headers/private/fs_shell/fssh_fs_interface.h

@@ -26,6 +26,8 @@ typedef fssh_ino_t fssh_vnode_id;
 /* the file system's private data structures */
 typedef void *fssh_fs_cookie;
 
+typedef struct FSSHIORequest fssh_io_request;
+
 /* additional flags passed to write_stat() */
 #define FSSH_B_STAT_SIZE_INSECURE	0x2000
 
@@ -142,6 +144,12 @@ struct fssh_fs_vnode_ops {
 				fssh_fs_cookie cookie, fssh_off_t pos, const fssh_iovec *vecs,
 				fssh_size_t count, fssh_size_t *_numBytes);
 
+	/* asynchronous I/O */
+	fssh_status_t (*io)(fssh_fs_volume *volume, fssh_fs_vnode *vnode,
+				void *cookie, fssh_io_request *request);
+	fssh_status_t (*cancel_io)(fssh_fs_volume *volume, fssh_fs_vnode *vnode,
+				void *cookie, fssh_io_request *request);
+
 	/* cache file access */
 	fssh_status_t (*get_file_map)(fssh_fs_volume *volume, fssh_fs_vnode *vnode,
 				fssh_off_t offset, fssh_size_t size,

headers/private/kernel/vfs.h

@@ -97,6 +97,10 @@ status_t vfs_read_pages(struct vnode *vnode, void *cookie, off_t pos,
 			const iovec *vecs, size_t count, size_t *_numBytes);
 status_t vfs_write_pages(struct vnode *vnode, void *cookie, off_t pos,
 			const iovec *vecs, size_t count, size_t *_numBytes);
+status_t vfs_synchronous_io(io_request* request,
+			status_t (*doIO)(void* cookie, off_t offset, void* buffer,
+				size_t length),
+			void* cookie);
 status_t vfs_get_vnode_cache(struct vnode *vnode, struct VMCache **_cache,
 			bool allocate);
 status_t vfs_get_file_map(struct vnode *vnode, off_t offset, size_t size,

src/add-ons/kernel/file_systems/bfs/kernel_interface.cpp

@@ -2165,6 +2165,9 @@ fs_vnode_ops gBFSVnodeOps = {
 	&bfs_read_pages,
 	&bfs_write_pages,
 
+	NULL,	// io()
+	NULL,	// cancel_io()
+
 	&bfs_get_file_map,
 
 	&bfs_ioctl,

src/add-ons/kernel/file_systems/cdda/kernel_interface.cpp

@@ -2014,6 +2014,9 @@ fs_vnode_ops gCDDAVnodeOps = {
 	cdda_read_pages,
 	cdda_write_pages,
 
+	NULL,	// io()
+	NULL,	// cancel_io()
+
 	NULL,	// get_file_map()
 
 	// common

src/add-ons/kernel/file_systems/ext2/kernel_interface.cpp

@@ -509,6 +509,9 @@ fs_vnode_ops gExt2VnodeOps = {
 	&ext2_read_pages,
 	NULL,
 
+	NULL,	// io()
+	NULL,	// cancel_io()
+
 	&ext2_get_file_map,
 
 	NULL,

src/add-ons/kernel/file_systems/fat/dosfs.c

@@ -1268,6 +1268,9 @@ fs_vnode_ops gFATVnodeOps = {
 	&dosfs_read_pages,
 	&dosfs_write_pages,
 
+	NULL,	// io()
+	NULL,	// cancel_io()
+
 	&dosfs_get_file_map,
 
 	&dosfs_ioctl,

src/add-ons/kernel/file_systems/iso9660/kernel_interface.cpp

@@ -679,6 +679,9 @@ fs_vnode_ops gISO9660VnodeOps = {
 	&fs_read_pages,
 	NULL,
 
+	NULL,	// io()
+	NULL,	// cancel_io()
+
 	/* cache file access */
 	NULL,
 

src/add-ons/kernel/file_systems/reiserfs/kernel_interface.cpp

@@ -6,12 +6,12 @@
 // it under the terms of the GNU General Public License as published by
 // the Free Software Foundation; either version 2 of the License, or
 // (at your option) any later version.
-// 
+//
 // This program is distributed in the hope that it will be useful,
 // but WITHOUT ANY WARRANTY; without even the implied warranty of
 // MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 // GNU General Public License for more details.
-// 
+//
 // You should have received a copy of the GNU General Public License
 // along with this program; if not, write to the Free Software
 // Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
@@ -444,7 +444,7 @@ reiserfs_read(fs_volume *fs, fs_vnode *_node, void *cookie, off_t pos,
 		else
 			error = B_BAD_VALUE;
 	}
-	
+
 	// read
 	StreamReader *reader = (StreamReader*)cookie;
 	if (error == B_OK) {
@@ -569,7 +569,7 @@ FUNCTION(("node: (%Ld: %lu, %lu)\n", node->GetID(), node->GetDirID(),
 	delete iterator;
 	return B_OK;
 }
-			
+
 // reiserfs_read_dir
 static status_t
 reiserfs_read_dir(fs_volume *fs, fs_vnode *_node, void *cookie,
@@ -754,6 +754,9 @@ fs_vnode_ops gReiserFSVnodeOps = {
 	NULL,	// &reiserfs_read_pages,
 	NULL,	// &reiserfs_write_pages,
 
+	NULL,	// io()
+	NULL,	// cancel_io()
+
 	NULL,	// &reiserfs_get_file_map,
 
 	NULL,	// &reiserfs_ioctl,

src/system/kernel/device_manager/devfs.cpp

@@ -1836,6 +1836,9 @@ fs_vnode_ops kVnodeOps = {
 	&devfs_read_pages,
 	&devfs_write_pages,
 
+	NULL,	// io()
+	NULL,	// cancel_io()
+
 	NULL,	// get_file_map
 
 	/* common */

src/system/kernel/fs/Jamfile

@@ -5,6 +5,8 @@ UsePrivateHeaders [ FDirName kernel fs ] ;
 UsePrivateHeaders [ FDirName kernel util ] ;
 UsePrivateHeaders net shared storage ;
 
+UseHeaders [ FDirName $(SUBDIR) $(DOTDOT) device_manager ] ;
+
 KernelMergeObject kernel_fs.o :
 	fd.cpp
 	fifo.cpp

src/system/kernel/fs/fifo.cpp

@@ -980,6 +980,9 @@ static fs_vnode_ops sFIFOVnodeOps = {
 	&fifo_read_pages,
 	&fifo_write_pages,
 
+	NULL,	// io()
+	NULL,	// cancel_io()
+
 	NULL,	// get_file_map
 
 	/* common */

src/system/kernel/fs/rootfs.cpp

@@ -1146,6 +1146,9 @@ fs_vnode_ops sVnodeOps = {
 	&rootfs_read_pages,
 	&rootfs_write_pages,
 
+	NULL,	// io()
+	NULL,	// cancel_io()
+
 	NULL,	// get_file_map()
 
 	/* common */

src/system/kernel/fs/vfs.cpp

@@ -8705,3 +8705,6 @@ _user_open_query(dev_t device, const char *userQuery, size_t queryLength,
 	free(query);
 	return fd;
 }
+
+
+#include "vfs_request_io.cpp"

src/system/kernel/fs/vfs_request_io.cpp

@@ -0,0 +1,403 @@
+/*
+ * Copyright 2008, Ingo Weinhold, ingo_weinhold@gmx.de.
+ * Distributed under the terms of the MIT License.
+ */
+
+// included by vfs.cpp
+
+#include "io_requests.h"
+
+
+struct iterative_io_cookie {
+	struct vnode*					vnode;
+	file_descriptor*				descriptor;
+	iterative_io_get_vecs			get_vecs;
+	iterative_io_finished			finished;
+	void*							cookie;
+	off_t							request_offset;
+	io_request_finished_callback	next_finished_callback;
+	void*							next_finished_cookie;
+};
+
+
+class DoIO {
+public:
+	DoIO(bool write, bool isPhysical)
+		:
+		fWrite(write),
+		fIsPhysical(isPhysical)
+	{
+	}
+
+	virtual	~DoIO()
+	{
+	}
+
+	status_t IO(off_t offset, void* _buffer, size_t length)
+	{
+		if (!fIsPhysical)
+			return InternalIO(offset, _buffer, length);
+
+		// buffer points to physical address -- map pages
+		addr_t buffer = (addr_t)_buffer;
+
+		while (length > 0) {
+			addr_t pageOffset = buffer % B_PAGE_SIZE;
+			addr_t virtualAddress;
+			status_t error = vm_get_physical_page(buffer - pageOffset,
+				&virtualAddress, 0);
+			if (error != B_OK)
+				return error;
+
+			size_t toTransfer = min_c(length, B_PAGE_SIZE - pageOffset);
+			error = InternalIO(offset, (void*)(virtualAddress + pageOffset),
+				toTransfer);
+
+			vm_put_physical_page(virtualAddress);
+
+			if (error != B_OK)
+				return error;
+
+			buffer += toTransfer;
+			length -= toTransfer;
+		}
+
+		return B_OK;
+	}
+
+protected:
+	virtual status_t InternalIO(off_t offset, void* buffer, size_t length) = 0;
+
+protected:
+	bool	fWrite;
+	bool	fIsPhysical;
+};
+
+
+class CallbackIO : public DoIO {
+public:
+	CallbackIO(bool write, bool isPhysical,
+			status_t (*doIO)(void* cookie, off_t offset, void* buffer,
+				size_t length),
+			void* cookie)
+		:
+		DoIO(write, isPhysical),
+		fDoIO(doIO),
+		fCookie(cookie)
+	{
+	}
+
+protected:
+	virtual status_t InternalIO(off_t offset, void* buffer, size_t length)
+	{
+		return fDoIO(fCookie, offset, buffer, length);
+	}
+
+private:
+	status_t (*fDoIO)(void*, off_t, void*, size_t);
+	void*		fCookie;
+};
+
+
+class VnodeIO : public DoIO {
+public:
+	VnodeIO(bool write, bool isPhysical, struct vnode* vnode, void* cookie)
+		:
+		DoIO(write, isPhysical),
+		fVnode(vnode),
+		fCookie(cookie)
+	{
+	}
+
+protected:
+	virtual status_t InternalIO(off_t offset, void* buffer, size_t length)
+	{
+		size_t bytesTransferred = length;
+		status_t error;
+		if (fWrite) {
+			error = FS_CALL(fVnode, write, fCookie, offset, buffer,
+				&bytesTransferred);
+		} else {
+			error = FS_CALL(fVnode, read, fCookie, offset, buffer,
+				&bytesTransferred);
+		}
+
+		return error == B_OK && bytesTransferred != length
+			? B_FILE_ERROR : error;
+	}
+
+private:
+	struct vnode*	fVnode;
+	void*			fCookie;
+};
+
+
+static status_t
+do_iterative_fd_io_iterate(void* _cookie, io_request* request)
+{
+	static const int32 kMaxSubRequests = 8;
+
+	iterative_io_cookie* cookie = (iterative_io_cookie*)_cookie;
+
+	request->DeleteSubRequests();
+
+	off_t requestOffset = cookie->request_offset;
+	size_t requestLength = request->Length()
+		- (requestOffset - request->Offset());
+
+	// get the next file vecs
+	file_io_vec vecs[kMaxSubRequests];
+	uint32 vecCount = kMaxSubRequests;
+	status_t error = cookie->get_vecs(cookie->cookie, request, requestOffset,
+		requestLength, vecs, &vecCount);
+	if (error != B_OK)
+		return error;
+	if (vecCount == 0)
+		return B_FILE_ERROR;
+
+	// create subrequests for the file vecs we've got
+	int32 subRequestCount = 0;
+	for (uint32 i = 0; i < vecCount && subRequestCount < kMaxSubRequests; i++) {
+		off_t vecOffset = vecs[i].offset;
+		off_t vecLength = vecs[i].length;
+
+		while (vecLength > 0 && subRequestCount < kMaxSubRequests) {
+			IORequest* subRequest;
+			error = request->CreateSubRequest(requestOffset, vecOffset,
+				vecLength, subRequest);
+			if (error != B_OK)
+				break;
+
+			subRequestCount++;
+
+			size_t lengthProcessed = subRequest->Length();
+			vecOffset += lengthProcessed;
+			vecLength -= lengthProcessed;
+			requestOffset += lengthProcessed;
+			requestLength -= lengthProcessed;
+		}
+	}
+
+	// Only if we couldn't create any subrequests, we fail.
+	if (error != B_OK && subRequestCount == 0)
+		return error;
+
+	request->Advance(requestOffset - cookie->request_offset);
+	cookie->request_offset = requestOffset;
+
+	// Schedule the subrequests.
+	for (IORequest* subRequest = request->FirstSubRequest();
+			(subRequest = request->NextSubRequest(subRequest)) != NULL;) {
+		if (error == B_OK) {
+			error = FS_CALL(cookie->vnode, io, cookie->descriptor->cookie,
+				subRequest);
+		} else {
+			// Once scheduling a subrequest failed, we cancel all subsequent
+			// subrequests.
+			subRequest->SetStatusAndNotify(B_CANCELED);
+		}
+	}
+
+	// TODO: Cancel the subrequests that were scheduled successfully.
+
+	return B_OK;
+}
+
+
+static status_t
+do_iterative_fd_io_finish(void* _cookie, io_request* request, status_t status)
+{
+	iterative_io_cookie* cookie = (iterative_io_cookie*)_cookie;
+
+	if (cookie->finished != NULL)
+		cookie->finished(cookie->cookie, request, status);
+
+	put_fd(cookie->descriptor);
+
+	if (cookie->next_finished_callback != NULL) {
+		cookie->next_finished_callback(cookie->next_finished_cookie, request,
+			status);
+	}
+
+	delete cookie;
+
+	return B_OK;
+}
+
+
+static status_t
+do_synchronous_iterative_vnode_io(struct vnode* vnode, void* openCookie,
+	io_request* request, iterative_io_get_vecs getVecs,
+	iterative_io_finished finished, void* cookie)
+{
+	IOBuffer* buffer = request->Buffer();
+	VnodeIO io(request->IsWrite(), buffer->IsPhysical(), vnode, openCookie);
+
+	iovec* vecs = buffer->Vecs();
+	int32 vecCount = buffer->VecCount();
+	off_t offset = request->Offset();
+	size_t length = request->Length();
+
+	status_t error = B_OK;
+
+	for (int32 i = 0; error == B_OK && length > 0 && i < vecCount; i++) {
+		uint8* vecBase = (uint8*)vecs[i].iov_base;
+		size_t vecLength = vecs[i].iov_len;
+		if (vecLength > length)
+			vecLength = length;
+
+		while (error == B_OK && vecLength > 0) {
+			file_io_vec fileVecs[8];
+			uint32 fileVecCount = 8;
+			error = getVecs(cookie, request, offset, vecLength, fileVecs,
+				&fileVecCount);
+			if (error == B_OK && fileVecCount == 0)
+				error = B_FILE_ERROR;
+			if (error != B_OK)
+				break;
+
+			for (uint32 k = 0; k < fileVecCount; k++) {
+				const file_io_vec& fileVec = fileVecs[i];
+				error = io.IO(fileVec.offset, vecBase, fileVec.length);
+				if (error != B_OK)
+					break;
+
+				offset += fileVec.length;
+				length -= fileVec.length;
+				vecBase += fileVec.length;
+				vecLength -= fileVec.length;
+			}
+		}
+	}
+
+	finished(cookie, request, error);
+	request->SetStatusAndNotify(error);
+	return error;
+}
+
+
+static status_t
+synchronous_io(io_request* request, DoIO& io)
+{
+	IOBuffer* buffer = request->Buffer();
+
+	iovec* vecs = buffer->Vecs();
+	int32 vecCount = buffer->VecCount();
+	off_t offset = request->Offset();
+	size_t length = request->Length();
+
+	for (int32 i = 0; length > 0 && i < vecCount; i++) {
+		void* vecBase = vecs[i].iov_base;
+		size_t vecLength = vecs[i].iov_len;
+		if (vecLength > length)
+			vecLength = length;
+
+		status_t error = io.IO(offset, vecBase, vecLength);
+		if (error != B_OK) {
+			request->SetStatusAndNotify(error);
+			return error;
+		}
+
+		offset += vecLength;
+		length -= vecLength;
+	}
+
+	request->SetStatusAndNotify(B_OK);
+	return B_OK;
+}
+
+
+// #pragma mark - kernel private API
+
+
+extern "C" status_t
+vfs_synchronous_io(io_request* request,
+	status_t (*doIO)(void* cookie, off_t offset, void* buffer, size_t length),
+	void* cookie)
+{
+	IOBuffer* buffer = request->Buffer();
+	CallbackIO io(request->IsWrite(), buffer->IsPhysical(), doIO, cookie);
+
+	return synchronous_io(request, io);
+}
+
+
+// #pragma mark - public API
+
+
+extern "C" status_t
+do_fd_io(int fd, io_request* request)
+{
+	struct vnode* vnode;
+	file_descriptor* descriptor = get_fd_and_vnode(fd, &vnode, true);
+	if (descriptor == NULL) {
+		request->SetStatusAndNotify(B_FILE_ERROR);
+		return B_FILE_ERROR;
+	}
+
+	CObjectDeleter<file_descriptor> descriptorPutter(descriptor, put_fd);
+
+	if (!HAS_FS_CALL(vnode, io)) {
+		// no io() call -- fall back to synchronous I/O
+		IOBuffer* buffer = request->Buffer();
+		VnodeIO io(request->IsWrite(), buffer->IsPhysical(), vnode,
+			descriptor->cookie);
+		return synchronous_io(request, io);
+	}
+
+	return FS_CALL(vnode, io, descriptor->cookie, request);
+}
+
+
+extern "C" status_t
+do_iterative_fd_io(int fd, io_request* request, iterative_io_get_vecs getVecs,
+	iterative_io_finished finished, void* cookie)
+{
+	struct vnode* vnode;
+	file_descriptor* descriptor = get_fd_and_vnode(fd, &vnode, true);
+	if (descriptor == NULL) {
+		finished(cookie, request, B_FILE_ERROR);
+		request->SetStatusAndNotify(B_FILE_ERROR);
+		return B_FILE_ERROR;
+	}
+
+	CObjectDeleter<file_descriptor> descriptorPutter(descriptor, put_fd);
+
+	if (!HAS_FS_CALL(vnode, io)) {
+		// no io() call -- fall back to synchronous I/O
+		return do_synchronous_iterative_vnode_io(vnode, descriptor->cookie,
+			request, getVecs, finished, cookie);
+	}
+
+	iterative_io_cookie* iterationCookie
+		= new(std::nothrow) iterative_io_cookie;
+		// TODO: Heed B_VIP_IO_REQUEST!
+	if (iterationCookie == NULL) {
+		// no memory -- fall back to synchronous I/O
+		return do_synchronous_iterative_vnode_io(vnode, descriptor->cookie,
+			request, getVecs, finished, cookie);
+	}
+
+	iterationCookie->vnode = vnode;
+	iterationCookie->descriptor = descriptor;
+	iterationCookie->get_vecs = getVecs;
+	iterationCookie->finished = finished;
+	iterationCookie->cookie = cookie;
+	iterationCookie->request_offset = request->Offset();
+	iterationCookie->next_finished_callback = request->FinishedCallback(
+		&iterationCookie->next_finished_cookie);
+
+	request->SetFinishedCallback(&do_iterative_fd_io_finish, iterationCookie);
+	request->SetIterationCallback(&do_iterative_fd_io_iterate, iterationCookie);
+
+	status_t error = do_iterative_fd_io_iterate(iterationCookie, request);
+	if (error != B_OK) {
+		finished(cookie, request, error);
+		request->SetStatusAndNotify(error);
+		return error;
+	}
+
+	descriptorPutter.Detach();
+
+	return B_OK;
+}