Linux 4.1 compat: loop device on ZFS

Starting from Linux 4.1 allows iov_iter with bio_vec to be passed into
iter_read/iter_write. Notably, the loop device will pass bio_vec to backend
filesystem. However, current ZFS code assumes iovec without any check, so it
will always crash when using loop device.

With the restructured uio_t, we can safely pass bio_vec in uio_t with UIO_BVEC
set. The uio* functions are modified to handle bio_vec case separately.

The const uio_iov causes some warning in xuio related stuff, so explicit
convert them to non const.

Signed-off-by: Chunwei Chen <[email protected]>
Signed-off-by: Richard Yao <[email protected]>
Signed-off-by: Brian Behlendorf <[email protected]>
Closes #3511
Closes #3640

1 files changed, 45 insertions, 36 deletions

diff --git a/module/zfs/zpl_file.c b/module/zfs/zpl_file.c
index 547114012..a23bc7d8d 100644
--- a/module/zfs/zpl_file.c
+++ b/module/zfs/zpl_file.c
@@ -20,6 +20,7 @@
  */
 /*
  * Copyright (c) 2011, Lawrence Livermore National Security, LLC.
+ * Copyright (c) 2015 by Chunwei Chen. All rights reserved.
  */
 
 
@@ -202,17 +203,18 @@ zpl_aio_fsync(struct kiocb *kiocb, int datasync)
 #error "Unsupported fops->fsync() implementation"
 #endif
 
-static inline ssize_t
+static ssize_t
 zpl_read_common_iovec(struct inode *ip, const struct iovec *iovp, size_t count,
-    unsigned long nr_segs, loff_t *ppos, uio_seg_t segment,
-    int flags, cred_t *cr)
+    unsigned long nr_segs, loff_t *ppos, uio_seg_t segment, int flags,
+    cred_t *cr, size_t skip)
 {
 	ssize_t read;
 	uio_t uio;
 	int error;
 	fstrans_cookie_t cookie;
 
-	uio.uio_iov = (struct iovec *)iovp;
+	uio.uio_iov = iovp;
+	uio.uio_skip = skip;
 	uio.uio_resid = count;
 	uio.uio_iovcnt = nr_segs;
 	uio.uio_loffset = *ppos;
@@ -242,7 +244,7 @@ zpl_read_common(struct inode *ip, const char *buf, size_t len, loff_t *ppos,
 	iov.iov_len = len;
 
 	return (zpl_read_common_iovec(ip, &iov, len, 1, ppos, segment,
-	    flags, cr));
+	    flags, cr, 0));
 }
 
 static ssize_t
@@ -261,24 +263,17 @@ zpl_read(struct file *filp, char __user *buf, size_t len, loff_t *ppos)
 
 static ssize_t
 zpl_iter_read_common(struct kiocb *kiocb, const struct iovec *iovp,
-    unsigned long nr_segs, size_t count)
+    unsigned long nr_segs, size_t count, uio_seg_t seg, size_t skip)
 {
 	cred_t *cr = CRED();
 	struct file *filp = kiocb->ki_filp;
 	ssize_t read;
-	size_t alloc_size = sizeof (struct iovec) * nr_segs;
-	struct iovec *iov_tmp = kmem_alloc(alloc_size, KM_SLEEP);
-	bcopy(iovp, iov_tmp, alloc_size);
-
-	ASSERT(iovp);
 
 	crhold(cr);
-	read = zpl_read_common_iovec(filp->f_mapping->host, iov_tmp, count,
-	    nr_segs, &kiocb->ki_pos, UIO_USERSPACE, filp->f_flags, cr);
+	read = zpl_read_common_iovec(filp->f_mapping->host, iovp, count,
+	    nr_segs, &kiocb->ki_pos, seg, filp->f_flags, cr, skip);
 	crfree(cr);
 
-	kmem_free(iov_tmp, alloc_size);
-
 	return (read);
 }
 
@@ -286,22 +281,32 @@ zpl_iter_read_common(struct kiocb *kiocb, const struct iovec *iovp,
 static ssize_t
 zpl_iter_read(struct kiocb *kiocb, struct iov_iter *to)
 {
-	return (zpl_iter_read_common(kiocb, to->iov, to->nr_segs,
-	    iov_iter_count(to)));
+	ssize_t ret;
+	uio_seg_t seg = UIO_USERSPACE;
+	if (to->type & ITER_KVEC)
+		seg = UIO_SYSSPACE;
+	if (to->type & ITER_BVEC)
+		seg = UIO_BVEC;
+	ret = zpl_iter_read_common(kiocb, to->iov, to->nr_segs,
+	    iov_iter_count(to), seg, to->iov_offset);
+	if (ret > 0)
+		iov_iter_advance(to, ret);
+	return (ret);
 }
 #else
 static ssize_t
 zpl_aio_read(struct kiocb *kiocb, const struct iovec *iovp,
     unsigned long nr_segs, loff_t pos)
 {
-	return (zpl_iter_read_common(kiocb, iovp, nr_segs, kiocb->ki_nbytes));
+	return (zpl_iter_read_common(kiocb, iovp, nr_segs, kiocb->ki_nbytes,
+	    UIO_USERSPACE, 0));
 }
 #endif /* HAVE_VFS_RW_ITERATE */
 
-static inline ssize_t
+static ssize_t
 zpl_write_common_iovec(struct inode *ip, const struct iovec *iovp, size_t count,
-    unsigned long nr_segs, loff_t *ppos, uio_seg_t segment,
-    int flags, cred_t *cr)
+    unsigned long nr_segs, loff_t *ppos, uio_seg_t segment, int flags,
+    cred_t *cr, size_t skip)
 {
 	ssize_t wrote;
 	uio_t uio;
@@ -311,7 +316,8 @@ zpl_write_common_iovec(struct inode *ip, const struct iovec *iovp, size_t count,
 	if (flags & O_APPEND)
 		*ppos = i_size_read(ip);
 
-	uio.uio_iov = (struct iovec *)iovp;
+	uio.uio_iov = iovp;
+	uio.uio_skip = skip;
 	uio.uio_resid = count;
 	uio.uio_iovcnt = nr_segs;
 	uio.uio_loffset = *ppos;
@@ -340,7 +346,7 @@ zpl_write_common(struct inode *ip, const char *buf, size_t len, loff_t *ppos,
 	iov.iov_len = len;
 
 	return (zpl_write_common_iovec(ip, &iov, len, 1, ppos, segment,
-	    flags, cr));
+	    flags, cr, 0));
 }
 
 static ssize_t
@@ -359,24 +365,17 @@ zpl_write(struct file *filp, const char __user *buf, size_t len, loff_t *ppos)
 
 static ssize_t
 zpl_iter_write_common(struct kiocb *kiocb, const struct iovec *iovp,
-    unsigned long nr_segs, size_t count)
+    unsigned long nr_segs, size_t count, uio_seg_t seg, size_t skip)
 {
 	cred_t *cr = CRED();
 	struct file *filp = kiocb->ki_filp;
 	ssize_t wrote;
-	size_t alloc_size = sizeof (struct iovec) * nr_segs;
-	struct iovec *iov_tmp = kmem_alloc(alloc_size, KM_SLEEP);
-	bcopy(iovp, iov_tmp, alloc_size);
-
-	ASSERT(iovp);
 
 	crhold(cr);
-	wrote = zpl_write_common_iovec(filp->f_mapping->host, iov_tmp, count,
-	    nr_segs, &kiocb->ki_pos, UIO_USERSPACE, filp->f_flags, cr);
+	wrote = zpl_write_common_iovec(filp->f_mapping->host, iovp, count,
+	    nr_segs, &kiocb->ki_pos, seg, filp->f_flags, cr, skip);
 	crfree(cr);
 
-	kmem_free(iov_tmp, alloc_size);
-
 	return (wrote);
 }
 
@@ -384,15 +383,25 @@ zpl_iter_write_common(struct kiocb *kiocb, const struct iovec *iovp,
 static ssize_t
 zpl_iter_write(struct kiocb *kiocb, struct iov_iter *from)
 {
-	return (zpl_iter_write_common(kiocb, from->iov, from->nr_segs,
-	    iov_iter_count(from)));
+	ssize_t ret;
+	uio_seg_t seg = UIO_USERSPACE;
+	if (from->type & ITER_KVEC)
+		seg = UIO_SYSSPACE;
+	if (from->type & ITER_BVEC)
+		seg = UIO_BVEC;
+	ret = zpl_iter_write_common(kiocb, from->iov, from->nr_segs,
+	    iov_iter_count(from), seg, from->iov_offset);
+	if (ret > 0)
+		iov_iter_advance(from, ret);
+	return (ret);
 }
 #else
 static ssize_t
 zpl_aio_write(struct kiocb *kiocb, const struct iovec *iovp,
     unsigned long nr_segs, loff_t pos)
 {
-	return (zpl_iter_write_common(kiocb, iovp, nr_segs, kiocb->ki_nbytes));
+	return (zpl_iter_write_common(kiocb, iovp, nr_segs, kiocb->ki_nbytes,
+	    UIO_USERSPACE, 0));
 }
 #endif /* HAVE_VFS_RW_ITERATE */