Message ID | 20191031211857.18989-1-pshilov@microsoft.com (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
Series | CIFS: Fix SMB2 oplock break processing | expand |
tentatively merged into cifs-2.6.git for-next pending review and buildbot regression test runs On Thu, Oct 31, 2019 at 5:50 PM Pavel Shilovsky <piastryyy@gmail.com> wrote: > > Even when mounting modern protocol version the server may be > configured without supporting SMB2.1 leases and the client > uses SMB2 oplock to optimize IO performance through local caching. > > However there is a problem in oplock break handling that leads > to missing a break notification on the client who has a file > opened. It latter causes big latencies to other clients that > are trying to open the same file. > > The problem reproduces when there are multiple shares from the > same server mounted on the client. The processing code tries to > match persistent and volatile file ids from the break notification > with an open file but it skips all share besides the first one. > Fix this by looking up in all shares belonging to the server that > issued the oplock break. > > Cc: Stable <stable@vger.kernel.org> > Signed-off-by: Pavel Shilovsky <pshilov@microsoft.com> > --- > fs/cifs/smb2misc.c | 7 +++---- > 1 file changed, 3 insertions(+), 4 deletions(-) > > diff --git a/fs/cifs/smb2misc.c b/fs/cifs/smb2misc.c > index 8db6201b18ba..527c9efd3de0 100644 > --- a/fs/cifs/smb2misc.c > +++ b/fs/cifs/smb2misc.c > @@ -664,10 +664,10 @@ smb2_is_valid_oplock_break(char *buffer, struct TCP_Server_Info *server) > spin_lock(&cifs_tcp_ses_lock); > list_for_each(tmp, &server->smb_ses_list) { > ses = list_entry(tmp, struct cifs_ses, smb_ses_list); > + > list_for_each(tmp1, &ses->tcon_list) { > tcon = list_entry(tmp1, struct cifs_tcon, tcon_list); > > - cifs_stats_inc(&tcon->stats.cifs_stats.num_oplock_brks); > spin_lock(&tcon->open_file_lock); > list_for_each(tmp2, &tcon->openFileList) { > cfile = list_entry(tmp2, struct cifsFileInfo, > @@ -679,6 +679,8 @@ smb2_is_valid_oplock_break(char *buffer, struct TCP_Server_Info *server) > continue; > > cifs_dbg(FYI, "file id match, oplock break\n"); > + cifs_stats_inc( > + &tcon->stats.cifs_stats.num_oplock_brks); > cinode = CIFS_I(d_inode(cfile->dentry)); > spin_lock(&cfile->file_info_lock); > if (!CIFS_CACHE_WRITE(cinode) && > @@ -702,9 +704,6 @@ smb2_is_valid_oplock_break(char *buffer, struct TCP_Server_Info *server) > return true; > } > spin_unlock(&tcon->open_file_lock); > - spin_unlock(&cifs_tcp_ses_lock); > - cifs_dbg(FYI, "No matching file for oplock break\n"); > - return true; > } > } > spin_unlock(&cifs_tcp_ses_lock); > -- > 2.17.1 >
diff --git a/fs/cifs/smb2misc.c b/fs/cifs/smb2misc.c index 8db6201b18ba..527c9efd3de0 100644 --- a/fs/cifs/smb2misc.c +++ b/fs/cifs/smb2misc.c @@ -664,10 +664,10 @@ smb2_is_valid_oplock_break(char *buffer, struct TCP_Server_Info *server) spin_lock(&cifs_tcp_ses_lock); list_for_each(tmp, &server->smb_ses_list) { ses = list_entry(tmp, struct cifs_ses, smb_ses_list); + list_for_each(tmp1, &ses->tcon_list) { tcon = list_entry(tmp1, struct cifs_tcon, tcon_list); - cifs_stats_inc(&tcon->stats.cifs_stats.num_oplock_brks); spin_lock(&tcon->open_file_lock); list_for_each(tmp2, &tcon->openFileList) { cfile = list_entry(tmp2, struct cifsFileInfo, @@ -679,6 +679,8 @@ smb2_is_valid_oplock_break(char *buffer, struct TCP_Server_Info *server) continue; cifs_dbg(FYI, "file id match, oplock break\n"); + cifs_stats_inc( + &tcon->stats.cifs_stats.num_oplock_brks); cinode = CIFS_I(d_inode(cfile->dentry)); spin_lock(&cfile->file_info_lock); if (!CIFS_CACHE_WRITE(cinode) && @@ -702,9 +704,6 @@ smb2_is_valid_oplock_break(char *buffer, struct TCP_Server_Info *server) return true; } spin_unlock(&tcon->open_file_lock); - spin_unlock(&cifs_tcp_ses_lock); - cifs_dbg(FYI, "No matching file for oplock break\n"); - return true; } } spin_unlock(&cifs_tcp_ses_lock);
Even when mounting modern protocol version the server may be configured without supporting SMB2.1 leases and the client uses SMB2 oplock to optimize IO performance through local caching. However there is a problem in oplock break handling that leads to missing a break notification on the client who has a file opened. It latter causes big latencies to other clients that are trying to open the same file. The problem reproduces when there are multiple shares from the same server mounted on the client. The processing code tries to match persistent and volatile file ids from the break notification with an open file but it skips all share besides the first one. Fix this by looking up in all shares belonging to the server that issued the oplock break. Cc: Stable <stable@vger.kernel.org> Signed-off-by: Pavel Shilovsky <pshilov@microsoft.com> --- fs/cifs/smb2misc.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-)