Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

RE: Question 3.5.11 Transmit local fault bit?




Justin,

No, you do not set the transmit local fault bit based upon receiving a Local
Fault sequence. 

If you are inserting Local Fault into the transmit path, as you do when the
transmit state machine is in reset, then you would set the bit.  That means
one will not ordinarily see this bit set as while in reset you can't read
the register. However, I think the bit should allow for an implementation
dependent detection of an internal problem and should allow being set due to
that. Right now there isn't text that makes that specifically allows or
disallows such usage.

Your question has caused me to notice that there is a bit of a race
condition in the definition of the bit. The definition says the bit is to be
cleared by a reset though it is vague as to whether this is at the
beginning, the end or all through reset. During a reset we will be sending
local fault so we will try to be setting the bit. During reset we will be
simultaneously setting and clearing the bit so at the end of reset the value
is ambiguous.

The receive bit has a similar problem but since we know that at the end of
reset it will be set because the receive machines won't have achieved lock
because the machines are held in a reset state during reset. Thus we know
that at the end of reset the local fault bit will be set and we don't care
what its value is during reset when it is unreadable. The requirement to
clear the receive local fault bit during reset is useless but doesn't create
the same ambiguity.

Pat

-----Original Message-----
From: Justin Gaither [mailto:jgaither@xxxxxxxxxxxxxxx]
Sent: Monday, May 14, 2001 1:30 PM
To: 802.3ae
Subject: Question 3.5.11 Transmit local fault bit?


Everyone,
	In the Clause 45.2.3.5.2 Transmit Local Fault it says "When read as
one
bit 3.5.11 indicates that the PCS has detected a local fault condition
on the transmit path"  It does not say what a local fault condition is,
and I can assume it means that the Transmit state machine is in reset
because that is the only time the PCS will generate the LBLOCK_T. 
However, I am wondering if I receive a Local Fault Sequence on the XGMII
inputs to the transmit block, am I suppose to set this bit?

justin
-- 
Justin Gaither                       Phone: 512-306-7292  x529
RocketChips a Division of Xilinx     Fax:   512-306-7293
500 N. Capital of TX Hwy.
Bldg 3                         email: jgaither@xxxxxxxxxxxxxxx
Austin, TX 78746               WWW:   www.rocketchips.com