[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference 7.286::fddi

Title:	FDDI - The Next Generation

Moderator:	NETCAD::STEFANI

Created:	Thu Apr 27 1989
Last Modified:	Thu Jun 05 1997
Last Successful Update:	Fri Jun 06 1997
Number of topics:	2259
Total number of notes:	8590

1859.0. "XMI BYTE COUNT ERRORS. FXDRIVER QUESTION." by KERNEL::SMITH_P () Mon Nov 06 1995 14:07

Hello,
      I have a customer who recently received a patched driver for VMS 5.5-2.
The patch was to cure his XMI BYTE COUNT ERRORS. The driver causing the problem
was version "X-22A1" and is the latest (The patch was a special for this
particular customer)

The customer now wishes to upgrade to VMS 6.1. and the most recent FXDRIVER for
this version appears to be "X-24" and has a similar creation date/time of
March94. The customer is worried since nodes dropped from his cluster whenever
the XMI errors occurred. He requires some reassurance that the X-24 driver in
VMS6.1 wont cause him a similar problem.

Has anyone experienced XMI BYTE COUNT ERRORS with VMS6.1?

If anyone in engineering is reading, is this a known problem with the X-24
driver?  

I am also posting this note in the FDDI conference. Any replies would be 
greatly appreciated.

Paul.

T.R Title User Personal
Name Date Lines

1859.1 STAR::STOCKDALE Mon Nov 06 1995 14:10 18

T.R	Title	User	Personal Name	Date	Lines
1859.1		STAR::STOCKDALE		`Mon Nov 06 1995 14:10`	18
	There were two sources of XMI Byte Count Errors, first, which is fixed in X-24 is when a chained transmit request is received by the driver for which the 2nd chain segment is zero bytes long. The driver fix was to check for a zero byte length 2nd segment and ignore it. The second is fixed in a TIMA kit, and is fixed in V6.2, X-26, which is the case where a workaround for a hardware problem induces a zero byte count segment. The DEMFA cannot transmit packets for which the packet length MOD 200 (hex) is either 1E0 or 0. The driver has to detect these transmit cases and ensure that the last byte of the packet is transmitted in the first hexword of a page aligned buffer, so what it does is decrement the byte count of the last segment by one byte and create a new segment which consists of one byte located in the special hexword buffer. What X-26 does is to verify that the decrement of the byte count of the last segment does not result in a zero length segment, but instead, in that case, replaces the last segment by the new segment instead of adding a new segment. - Dick

There were two sources of XMI Byte Count Errors, first, which is fixed
in X-24 is when a chained transmit request is received by the driver for
which the 2nd chain segment is zero bytes long.  The driver fix was to
check for a zero byte length 2nd segment and ignore it.

The second is fixed in a TIMA kit, and is fixed in V6.2, X-26, which is
the case where a workaround for a hardware problem induces a zero byte
count segment.  The DEMFA cannot transmit packets for which the packet
length MOD 200 (hex) is either 1E0 or 0.  The driver has to detect these
transmit cases and ensure that the last byte of the packet is transmitted
in the first hexword of a page aligned buffer, so what it does is decrement
the byte count of the last segment by one byte and create a new segment
which consists of one byte located in the special hexword buffer.  What X-26
does is to verify that the decrement of the byte count of the last segment
does not result in a zero length segment, but instead, in that case, replaces
the last segment by the new segment instead of adding a new segment.

- Dick