From cda8cc22bcd2cea397ee833f39bd1bad0b4b4ea0 Mon Sep 17 00:00:00 2001
From: Patrick Schmidt <patrick.schmidt@innogames.com>
Date: Tue, 23 Aug 2022 10:47:18 +0200
Subject: [PATCH] Implement lseek syscall in FUSE (#3491)

See the man page of lseek:
https://man7.org/linux/man-pages/man2/lseek.2.html
---
 weed/mount/weedfs_file_lseek.go  | 110 +++++++++++++++++++++++++++++++
 weed/mount/weedfs_unsupported.go |   7 --
 2 files changed, 110 insertions(+), 7 deletions(-)
 create mode 100644 weed/mount/weedfs_file_lseek.go

diff --git a/weed/mount/weedfs_file_lseek.go b/weed/mount/weedfs_file_lseek.go
new file mode 100644
index 000000000..ed495f5b5
--- /dev/null
+++ b/weed/mount/weedfs_file_lseek.go
@@ -0,0 +1,110 @@
+package mount
+
+import (
+	"syscall"
+
+	"github.com/hanwen/go-fuse/v2/fuse"
+
+	"github.com/seaweedfs/seaweedfs/weed/filer"
+	"github.com/seaweedfs/seaweedfs/weed/glog"
+)
+
+// These are non-POSIX extensions
+const (
+	SEEK_DATA uint32 = 3 // seek to next data after the offset
+	SEEK_HOLE uint32 = 4 // seek to next hole after the offset
+	ENXIO            = fuse.Status(syscall.ENXIO)
+)
+
+// Lseek finds next data or hole segments after the specified offset
+// See https://man7.org/linux/man-pages/man2/lseek.2.html
+func (wfs *WFS) Lseek(cancel <-chan struct{}, in *fuse.LseekIn, out *fuse.LseekOut) fuse.Status {
+	// not a documented feature
+	if in.Padding != 0 {
+		return fuse.EINVAL
+	}
+
+	if in.Whence != SEEK_DATA && in.Whence != SEEK_HOLE {
+		return fuse.EINVAL
+	}
+
+	// file must exist
+	fh := wfs.GetHandle(FileHandleId(in.Fh))
+	if fh == nil {
+		return fuse.EBADF
+	}
+
+	// lock the file until the proper offset was calculated
+	fh.Lock()
+	defer fh.Unlock()
+	fh.entryLock.Lock()
+	defer fh.entryLock.Unlock()
+
+	fileSize := int64(filer.FileSize(fh.entry))
+	offset := max(int64(in.Offset), 0)
+
+	glog.V(4).Infof(
+		"Lseek %s fh %d in [%d,%d], whence %d",
+		fh.FullPath(), fh.fh, offset, fileSize, in.Whence,
+	)
+
+	// can neither seek beyond file nor seek to a hole at the end of the file with SEEK_DATA
+	if offset > fileSize {
+		return ENXIO
+	} else if in.Whence == SEEK_DATA && offset == fileSize {
+		return ENXIO
+	}
+
+	// refresh view cache if necessary
+	if fh.entryViewCache == nil {
+		var err error
+		fh.entryViewCache, err = filer.NonOverlappingVisibleIntervals(fh.wfs.LookupFn(), fh.entry.Chunks, 0, fileSize)
+		if err != nil {
+			return fuse.EIO
+		}
+	}
+
+	// search chunks for the offset
+	found, offset := searchChunks(fh, offset, fileSize, in.Whence)
+	if found {
+		out.Offset = uint64(offset)
+		return fuse.OK
+	}
+
+	// in case we found no exact matches, we return the recommended fallbacks, that is:
+	// original offset for SEEK_DATA or end of file for an implicit hole
+	if in.Whence == SEEK_DATA {
+		out.Offset = in.Offset
+	} else {
+		out.Offset = uint64(fileSize)
+	}
+
+	return fuse.OK
+}
+
+// searchChunks goes through all chunks to find the correct offset
+func searchChunks(fh *FileHandle, offset, fileSize int64, whence uint32) (found bool, out int64) {
+	chunkViews := filer.ViewFromVisibleIntervals(fh.entryViewCache, offset, fileSize)
+
+	for _, chunkView := range chunkViews {
+		if offset < chunkView.LogicOffset {
+			if whence == SEEK_HOLE {
+				out = offset
+			} else {
+				out = chunkView.LogicOffset
+			}
+
+			return true, out
+		}
+
+		if offset >= chunkView.LogicOffset && offset < chunkView.Offset+int64(chunkView.Size) && whence == SEEK_DATA {
+			out = offset
+
+			return true, out
+		}
+
+		offset += int64(chunkView.Size)
+	}
+
+	return
+}
diff --git a/weed/mount/weedfs_unsupported.go b/weed/mount/weedfs_unsupported.go
index 08347aec1..a8342a2fc 100644
--- a/weed/mount/weedfs_unsupported.go
+++ b/weed/mount/weedfs_unsupported.go
@@ -16,13 +16,6 @@ func (wfs *WFS) Fallocate(cancel <-chan struct{}, in *fuse.FallocateIn) (code fu
 	return fuse.ENOSYS
 }
 
-/**
- * Find next data or hole after the specified offset
- */
-func (wfs *WFS) Lseek(cancel <-chan struct{}, in *fuse.LseekIn, out *fuse.LseekOut) fuse.Status {
-	return fuse.ENOSYS
-}
-
 func (wfs *WFS) GetLk(cancel <-chan struct{}, in *fuse.LkIn, out *fuse.LkOut) (code fuse.Status) {
 	return fuse.ENOSYS
 }