system/linux
system/linux
1
0
Fork 0
mirror of https://github.com/torvalds/linux synced 2024-09-06 18:01:05 +00:00
Code Issues Packages Projects Releases Wiki Activity

generic_perform_write()/iomap_write_actor(): saner logics for short copy

Browse source 
if we run into a short copy and ->write_end() refuses to advance at all,
use the amount we'd managed to copy for the next iteration to handle.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
This commit is contained in:
Al Viro 2021-05-31 00:32:44 -04:00
parent 9067931236
commit bc1bb416bb
2 changed files with 19 additions and 30 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

25
fs/iomap/buffered-io.c
View file

@ -771,10 +771,6 @@ iomap_write_actor(struct inode *inode, loff_t pos, loff_t length, void *data,
* Otherwise there's a nasty deadlock on copying from the
* same page as we're writing to, without it being marked
* up-to-date.
*
* Not only is this an optimisation, but it is also required
* to check that the address is actually valid, when atomic
* usercopies are used, below.
*/
if (unlikely(iov_iter_fault_in_readable(i, bytes))) {
status = -EFAULT;
@ -791,25 +787,24 @@ iomap_write_actor(struct inode *inode, loff_t pos, loff_t length, void *data,
copied = iov_iter_copy_from_user_atomic(page, i, offset, bytes);
copied = iomap_write_end(inode, pos, bytes, copied, page, iomap,
status = iomap_write_end(inode, pos, bytes, copied, page, iomap,
srcmap);
cond_resched();
iov_iter_advance(i, copied);
if (unlikely(copied == 0)) {
if (unlikely(status == 0)) {
/*
* If we were unable to copy any data at all, we must
* fall back to a single segment length write.
*
* If we didn't fallback here, we could livelock
* because not all segments in the iov can be copied at
* once without a pagefault.
* A short copy made iomap_write_end() reject the
* thing entirely. Might be memory poisoning
* halfway through, might be a race with munmap,
* might be severe memory pressure.
*/
bytes = min_t(unsigned long, PAGE_SIZE - offset,
iov_iter_single_seg_count(i));
if (copied)
bytes = copied;
goto again;
}
copied = status;
iov_iter_advance(i, copied);
pos += copied;
written += copied;
length -= copied;

24
mm/filemap.c
View file

@ -3642,10 +3642,6 @@ ssize_t generic_perform_write(struct file *file,
* Otherwise there's a nasty deadlock on copying from the
* same page as we're writing to, without it being marked
* up-to-date.
*
* Not only is this an optimisation, but it is also required
* to check that the address is actually valid, when atomic
* usercopies are used, below.
*/
if (unlikely(iov_iter_fault_in_readable(i, bytes))) {
status = -EFAULT;
@ -3672,24 +3668,22 @@ ssize_t generic_perform_write(struct file *file,
page, fsdata);
if (unlikely(status < 0))
break;
copied = status;
cond_resched();
iov_iter_advance(i, copied);
if (unlikely(copied == 0)) {
if (unlikely(status == 0)) {
/*
* If we were unable to copy any data at all, we must
* fall back to a single segment length write.
*
* If we didn't fallback here, we could livelock
* because not all segments in the iov can be copied at
* once without a pagefault.
* A short copy made ->write_end() reject the
* thing entirely. Might be memory poisoning
* halfway through, might be a race with munmap,
* might be severe memory pressure.
*/
bytes = min_t(unsigned long, PAGE_SIZE - offset,
iov_iter_single_seg_count(i));
if (copied)
bytes = copied;
goto again;
}
copied = status;
iov_iter_advance(i, copied);
pos += copied;
written += copied;