----- Original Message ----- > From: "Pavel Shilovsky" <piastryyy@xxxxxxxxx> > To: "Xiong Zhou" <xzhou@xxxxxxxxxx> > Cc: "Steve French" <smfrench@xxxxxxxxx>, "CIFS" <linux-cifs@xxxxxxxxxxxxxxx> > Sent: Monday, 26 March, 2018 3:58:00 AM > Subject: Re: the second open(2) call to file in samba 2.0 cifs mountpoint hang > > 2018-03-24 17:12 GMT-07:00 Xiong Zhou <xzhou@xxxxxxxxxx>: > > Hi, > > > > Sorry for the top posting. > > > > test.sh > > https://paste.fedoraproject.org/paste/L-PsaUEQF~wagiIAabsg8Q > > > > dmesg-v2.0 > > https://paste.fedoraproject.org/paste/zliVSma3mwod-JEzIe0bUQ > > tcpdump -r output v2.0 > > https://paste.fedoraproject.org/paste/aePcPqkl3Rm8XwF3-4u49A > > > > dmesg-v3.0 > > https://paste.fedoraproject.org/paste/E-BBmpr3bbJD4y6fPagt6g > > tcpdump -r output v3.0 > > https://paste.fedoraproject.org/paste/qa-7ziaF~yOhU1u~7kxJqg > > > > Thanks, > > Xiong > > Thanks. The kernel logs prove the theory that the client doesn't > process the oplock break request from the server: > > [58173.188033] fs/cifs/smb2misc.c: Checking for oplock break > [58173.188035] fs/cifs/smb2misc.c: oplock level 0x1 > [58173.188036] fs/cifs/smb2misc.c: No matching file for oplock break > [58173.188037] fs/cifs/connect.c: Received oplock break > > that's why the 2nd program succeeds to open the file only after the > 1st finishes execution (and closes its file descriptor). I can not reproduce this on Steve's for-next branch. The oplock is broken correctly on for-next. Verified with wireshark. > > Xiong, do you know if this scenario works on some older kernel version? > > -- > Best regards, > Pavel Shilovsky > -- > To unsubscribe from this list: send the line "unsubscribe linux-cifs" in > the body of a message to majordomo@xxxxxxxxxxxxxxx > More majordomo info at http://vger.kernel.org/majordomo-info.html > -- To unsubscribe from this list: send the line "unsubscribe linux-cifs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html