https://github.com/torvalds/linux
Revision 914ee295af418e936ec20a08c1663eaabe4cd07a authored by Xin Zhong on 09 December 2010, 09:30:14 UTC, committed by Chris Mason on 10 December 2010, 21:29:10 UTC
This problem is found in meego testing:
http://bugs.meego.com/show_bug.cgi?id=6672
A file in btrfs is mmaped and the mmaped buffer is passed to pwrite to write to the same page
of the same file. In btrfs_file_aio_write(), the pages is locked by prepare_pages(). So when
btrfs_copy_from_user() is called, page fault happens and the same page needs to be locked again
in filemap_fault(). The fix is to move iov_iter_fault_in_readable() before prepage_pages() to make page
fault happen before pages are locked. And also disable page fault in critical region in
btrfs_copy_from_user().

Reviewed-by: Yan, Zheng<zheng.z.yan@intel.com>
Signed-off-by: Zhong, Xin <xin.zhong@intel.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
1 parent f106e82
History
Tip revision: 914ee295af418e936ec20a08c1663eaabe4cd07a authored by Xin Zhong on 09 December 2010, 09:30:14 UTC
Btrfs: pwrite blocked when writing from the mmaped buffer of the same page
Tip revision: 914ee29
File Mode Size
Kconfig -rw-r--r-- 686 bytes
Makefile -rw-r--r-- 182 bytes
vxfs.h -rw-r--r-- 9.1 KB
vxfs_bmap.c -rw-r--r-- 7.2 KB
vxfs_dir.h -rw-r--r-- 3.2 KB
vxfs_extern.h -rw-r--r-- 3.0 KB
vxfs_fshead.c -rw-r--r-- 5.4 KB
vxfs_fshead.h -rw-r--r-- 2.5 KB
vxfs_immed.c -rw-r--r-- 3.5 KB
vxfs_inode.c -rw-r--r-- 8.8 KB
vxfs_inode.h -rw-r--r-- 5.4 KB
vxfs_lookup.c -rw-r--r-- 8.1 KB
vxfs_olt.c -rw-r--r-- 3.6 KB
vxfs_olt.h -rw-r--r-- 4.4 KB
vxfs_subr.c -rw-r--r-- 4.6 KB
vxfs_super.c -rw-r--r-- 7.3 KB

back to top