static int lfs_copy_from_user( loff_t pos, int npages, int nbytes,
struct page **segpages, const char __user *buf)
{
long page_fault=0;
int i, offset;
for ( i = 0, offset = (pos & (PAGE_CACHE_SIZE-1));
i < npages; i++,offset=0) {
size_t count = min_t(size_t,PAGE_CACHE_SIZE-offset, nbytes);
struct page *page = segpages[i];
fault_in_pages_readable( buf, count);
kmap(page);
page_fault = __copy_from_user(page_address(page)+offset,
buf, count);
flush_dcache_page(page);
kunmap(page);
buf += count;
nbytes -=count;
if (page_fault)
break;
}
return page_fault?-EFAULT:0;
}
/* simple helper to fault in pages and copy. This should go away
* and be replaced with calls into generic code.
*/
static noinline int btrfs_copy_from_user(loff_t pos, int num_pages,
int write_bytes,
struct page **prepared_pages,
const char __user *buf)
{
long page_fault = 0;
int i;
int offset = pos & (PAGE_CACHE_SIZE - 1);
for (i = 0; i < num_pages && write_bytes > 0; i++, offset = 0) {
size_t count = min_t(size_t,
PAGE_CACHE_SIZE - offset, write_bytes);
struct page *page = prepared_pages[i];
fault_in_pages_readable(buf, count);
/* Copy data from userspace to the current page */
kmap(page);
page_fault = __copy_from_user(page_address(page) + offset,
buf, count);
/* Flush processor's dcache for this page */
flush_dcache_page(page);
kunmap(page);
buf += count;
write_bytes -= count;
if (page_fault)
break;
}
return page_fault ? -EFAULT : 0;
}
/*
* Fault in the first iovec of the given iov_iter, to a maximum length
* of bytes. Returns 0 on success, or non-zero if the memory could not be
* accessed (ie. because it is an invalid address).
*
* writev-intensive code may want this to prefault several iovecs -- that
* would be possible (callers must not rely on the fact that _only_ the
* first iovec will be faulted with the current implementation).
*/
static int ii_iovec_fault_in_readable(struct iov_iter *i, size_t bytes)
{
struct iovec *iov = (struct iovec *)i->data;
char __user *buf = iov->iov_base + i->iov_offset;
bytes = min(bytes, iov->iov_len - i->iov_offset);
return fault_in_pages_readable(buf, bytes);
}
/*
* Fault in the first iovec of the given iov_iter, to a maximum length
* of bytes. Returns 0 on success, or non-zero if the memory could not be
* accessed (ie. because it is an invalid address).
*
* writev-intensive code may want this to prefault several iovecs -- that
* would be possible (callers must not rely on the fact that _only_ the
* first iovec will be faulted with the current implementation).
*/
int iov_iter_fault_in_readable(struct iov_iter *i, size_t bytes)
{
if (!(i->type & (ITER_BVEC|ITER_KVEC))) {
char __user *buf = i->iov->iov_base + i->iov_offset;
bytes = min(bytes, i->iov->iov_len - i->iov_offset);
return fault_in_pages_readable(buf, bytes);
}
return 0;
}
/*
* Pre-fault in the user memory, so we can use atomic copies.
*/
static void iov_fault_in_pages_read(struct iovec *iov, unsigned long len)
{
while (!iov->iov_len)
iov++;
while (len > 0) {
unsigned long this_len;
this_len = min_t(unsigned long, len, iov->iov_len);
fault_in_pages_readable(iov->iov_base, this_len);
len -= this_len;
iov++;
}
}
/*
* Fault in one or more iovecs of the given iov_iter, to a maximum length of
* bytes. For each iovec, fault in each page that constitutes the iovec.
*
* Return 0 on success, or non-zero if the memory could not be accessed (i.e.
* because it is an invalid address).
*/
int iov_iter_fault_in_readable(struct iov_iter *i, size_t bytes)
{
size_t skip = i->iov_offset;
const struct iovec *iov;
int err;
struct iovec v;
if (!(i->type & (ITER_BVEC|ITER_KVEC))) {
iterate_iovec(i, bytes, v, iov, skip, ({
err = fault_in_pages_readable(v.iov_base, v.iov_len);
if (unlikely(err))
return err;
0;}))
}
return 0;
static void kvm_use_magic_page(void)
{
u32 *p;
u32 *start, *end;
u32 features;
/* Tell the host to map the magic page to -4096 on all CPUs */
on_each_cpu(kvm_map_magic_page, &features, 1);
/* Quick self-test to see if the mapping works */
if (!fault_in_pages_readable((const char *)KVM_MAGIC_PAGE, sizeof(u32))) {
kvm_patching_worked = false;
return;
}
/* Now loop through all code and find instructions */
start = (void*)_stext;
end = (void*)_etext;
/*
* Being interrupted in the middle of patching would
* be bad for SPRG4-7, which KVM can't keep in sync
* with emulated accesses because reads don't trap.
*/
local_irq_disable();
for (p = start; p < end; p++) {
/* Avoid patching the template code */
if (p >= kvm_template_start && p < kvm_template_end) {
p = kvm_template_end - 1;
continue;
}
kvm_check_ins(p, features);
}
local_irq_enable();
printk(KERN_INFO "KVM: Live patching for a fast VM %s\n",
kvm_patching_worked ? "worked" : "failed");
}
static size_t copy_page_from_iter_iovec(struct page *page, size_t offset, size_t bytes,
struct iov_iter *i)
{
size_t skip, copy, left, wanted;
const struct iovec *iov;
char __user *buf;
void *kaddr, *to;
if (unlikely(bytes > i->count))
bytes = i->count;
if (unlikely(!bytes))
return 0;
might_fault();
wanted = bytes;
iov = i->iov;
skip = i->iov_offset;
buf = iov->iov_base + skip;
copy = min(bytes, iov->iov_len - skip);
if (IS_ENABLED(CONFIG_HIGHMEM) && !fault_in_pages_readable(buf, copy)) {
kaddr = kmap_atomic(page);
to = kaddr + offset;
/* first chunk, usually the only one */
left = copyin(to, buf, copy);
copy -= left;
skip += copy;
to += copy;
bytes -= copy;
while (unlikely(!left && bytes)) {
iov++;
buf = iov->iov_base;
copy = min(bytes, iov->iov_len);
left = copyin(to, buf, copy);
copy -= left;
skip = copy;
to += copy;
bytes -= copy;
}
if (likely(!bytes)) {
kunmap_atomic(kaddr);
goto done;
}
offset = to - kaddr;
buf += copy;
kunmap_atomic(kaddr);
copy = min(bytes, iov->iov_len - skip);
}
/* Too bad - revert to non-atomic kmap */
kaddr = kmap(page);
to = kaddr + offset;
left = copyin(to, buf, copy);
copy -= left;
skip += copy;
to += copy;
bytes -= copy;
while (unlikely(!left && bytes)) {
iov++;
buf = iov->iov_base;
copy = min(bytes, iov->iov_len);
left = copyin(to, buf, copy);
copy -= left;
skip = copy;
to += copy;
bytes -= copy;
}
kunmap(page);
done:
if (skip == iov->iov_len) {
iov++;
skip = 0;
}
i->count -= wanted - bytes;
i->nr_segs -= iov - i->iov;
i->iov = iov;
i->iov_offset = skip;
return wanted - bytes;
}
/* FIXME: Ugliest function of all in LFS, need I say more? */
static ssize_t lfs_file_write( struct file *file, const char __user *buf,
size_t count, loff_t *ppos)
{
loff_t pos;
struct page *page;
ssize_t res, written, bytes;
struct inode *inode = file->f_dentry->d_inode;
struct super_block *sb = inode->i_sb;
struct segment *segp = LFS_SBI(sb)->s_curr;
//dprintk("lfs_file_write called for %lu at pos %Lu\n", inode->i_ino, *ppos);
if(file->f_flags & O_DIRECT) {
dprintk("The file is requesting direct IO\n");
return -EINVAL;
}
if (unlikely(count < 0 ))
return -EINVAL;
if (unlikely(!access_ok(VERIFY_READ, buf, count)))
return -EFAULT;
//down(&inode->i_sem); /* lock the file */
mutex_lock(&inode->i_mutex); //BrechREiZ: We need this for Kernel 2.6.17
lfs_lock(sb);
pos = *ppos;
res = generic_write_checks(file, &pos, &count, 0);
if (res)
goto out;
if(count == 0)
goto out;
res = remove_suid(file->f_dentry);
if(res)
goto out;
//inode_update_time(inode, 1); /* update mtime and ctime */
file_update_time(inode); //BrechREiZ: We need this for Kernel 2.6.17
written = 0;
do {
long offset;
size_t copied;
int i, siblock, eiblock, boffset;
sector_t block;
offset = (segp->offset % BUF_IN_PAGE) * LFS_BSIZE;
offset += pos & (LFS_BSIZE - 1); /* within block */
bytes = PAGE_CACHE_SIZE - offset; /* number of bytes written
in this iteration */
invalidate_old_page(inode, pos);
if (bytes > count)
bytes = count;
//dprintk("1:segp->start=%Lu,segp->offset=%d,segp->end=%Lu,offset=%lu,bytes=%d\n", segp->start, segp->offset, segp->end,offset,bytes);
siblock = pos >> LFS_BSIZE_BITS;
eiblock = (pos + bytes - 1) >> LFS_BSIZE_BITS;
//dprintk("writing %d bytes at offset %ld (pos = %Lu)\n", bytes, offset, pos);
//dprintk("siblock = %d, eiblock = %d\n", siblock, eiblock);
/*
* Bring in the user page that we will copy from _first_.
* Otherwise there's a nasty deadlock on copying from the
* same page as we're writing to, without it being marked
* up-to-date.
*/
fault_in_pages_readable(buf, bytes);
page = get_seg_page(segp);
if (!page) {
res = -ENOMEM;
break;
}
/* fill the page with current inode blocks if any */
boffset = offset / LFS_BSIZE;;
for(i = siblock; i <= eiblock; ++i, ++boffset) {
struct buffer_head *bh;
//dprintk("Asking for block %d\n", i);
bh = lfs_read_block(inode, i);
if(!bh) /* new block */
break;
//dprintk("boffset = %d\n", boffset);
memcpy(page_address(page) + LFS_BSIZE * boffset, bh->b_data, LFS_BSIZE);
brelse(bh);
}
copied = __copy_from_user(page_address(page) + offset, buf, bytes);
flush_dcache_page(page);
block = segp->start + segp->offset;
for(i = siblock;i <= eiblock; ++i, ++block)
segsum_update_finfo(segp, inode->i_ino, i, block);
block = segp->start + segp->offset;
segp->offset += (bytes - 1)/LFS_BSIZE + 1;
//dprintk("2:segp->start=%Lu,segp->offset=%d,segp->end=%Lu,offset=%lu,bytes=%d\n",
//segp->start, segp->offset, segp->end,offset,bytes);
BUG_ON(segp->start + segp->offset > segp->end);
if(segp->start + segp->offset == segp->end) {
dprintk("allocating new segment\n");
/* This also is going to write the previous segment */
segment_allocate_new(inode->i_sb, segp, segp->start + segp->offset);
segp = LFS_SBI(sb)->s_curr;
}
/* update the inode */
for(i = siblock;i <= eiblock; ++i, ++block)
update_inode(inode, i, block);
//dprintk("start=%Lu,offset=%d,end=%Lu\n", segp->start, segp->offset, segp->end);
segusetbl_add_livebytes(sb, segp->segnum, bytes);
written += bytes;
buf += bytes;
pos += bytes;
count -= bytes;
} while(count);
*ppos = pos;
if(pos > inode->i_size)
i_size_write(inode, pos);
if(written)
mark_inode_dirty(inode);
lfs_unlock(sb);
//up(&inode->i_sem);
mutex_unlock(&inode->i_mutex); //BrechREiZ: and unlocking...
return written ? written : res;
out:
lfs_unlock(sb);
//up(&inode->i_sem);
mutex_unlock(&inode->i_mutex); //BrechREiZ: and unlocking...
return res;
}