Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 8 Aug 1995 14:27:03 +0200 (SAT)
From:      Bertus Pretorius <bertus@mikom.csir.co.za>
To:        current@freebsd.org
Subject:   Re: rlogin bug -> OOB bug?
Message-ID:  <199508081227.OAA01386@dolphin.mikom.csir.co.za>

next in thread | raw e-mail | index | archive | help
Hi,
    I'm a bit late on this discussion because i do not read current.  Jhay pointed
it out to me.  My experience is as follows:

It wrote a little tcp relay that is suppose to take care of OOB but it failed.
After detailed and painfull searching I found the following:

  Telnet and ftp (rlogin not checked) send more than one OOB byte - ftp 2 and telnet
  often a long string.

  When OOB is read inline or with the OOB flag set in the read you recieve only
  one byte.  This byte being the last byte of the OOB bytes in the packet.  The
  ATMARK ioctl marks also only the last byte, that is when you perform a read, OOB
  set to inline, the read returns before the last byte and the mark is set to
  indicate to following byte is OOB.

  When the line or read is slow that byte is moved forward as the specs say with
  the following result:
    packet:  OOB bytes <ABCDEF>, normal bytes <abcdef>
    data read: OOB bytes <F> normal bytes <ABCDEabcdef>

  With a fast line it is as follows:
    packet:  OOB bytes <ABCDEF>, normal bytes <abcdef>
    data read: normal bytes <ABCDE>, OOB bytes <F>, normal bytes <abcdef>

I tried to read the tcp code in the kernel and it seems to me that FreeBSD only
handles the last OOB byte in a packet as OOB data - not that I know what the
code really does :).

-- 
+-Bertus Pretorius------------ (O) (O) ---------------bertus@mikom.csir.co.za-+
| mikomtek                        ^                   +27 12 841-3001 (Voice) |
| CSIR                          \___/                 +27 12 841-4720 (FAX)   |
+-----------------A smile is the same in all languages------------------------+



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199508081227.OAA01386>