Re: [Ocfs2-devel] [PATCH 2/3] fs: ocfs2: Fix a possible null-pointer dereference in ocfs2_write_end_nolock()

Hi Jia-ju, Could you please point out how ->w_handle can be NULL if we are changing disk inode? I just checked the ocfs2 code but can't find any clue ... In my opinion, it's impossible to change disk inode without an existed journal transaction. If truly so, it's a another problem.

Re: [PATCH 2/3] fs: ocfs2: Fix a possible null-pointer dereference in ocfs2_write_end_nolock()

On 19/7/26 11:37, Jia-Ju Bai wrote: > In ocfs2_write_end_nolock(), there are an if statement on lines 1976, > 2047 and 2058, to check whether handle is NULL: > if (handle) > > When handle is NULL, it is used on line 2045: > ocfs2_update_inode_fsync_trans(handle, inode, 1); >

[PATCH 2/3] fs: ocfs2: Fix a possible null-pointer dereference in ocfs2_write_end_nolock()

In ocfs2_write_end_nolock(), there are an if statement on lines 1976, 2047 and 2058, to check whether handle is NULL: if (handle) When handle is NULL, it is used on line 2045: ocfs2_update_inode_fsync_trans(handle, inode, 1); oi->i_sync_tid = handle->h_transaction->t_tid;