RE: [802.3ae] Clause 48: Check_end Question
pat_thaler@xxxxxxxxxxx said:
>The normative text for check_end has the same meaning as the text you
>quote. The relevant sentence from the check_end definition is: The
>XGMII Error control character is returned in all lanes less than n in
>||T||, where n identifies the specific Terminate ordered-set ||T n ||,
>for which a running disparity error or any code-groups other than /A/
>or /K/ are recognized in the column following ||T||.
>
>Error is returned in lanes less than n in ||T|| for which an error is
>recognized in the column following the ||T||.
>
>Doug's interpretation of the sentence is not correct. It does not say
>that an error is returned in lanes less than n when an error occurs in
>the column after the ||T||. The subtle thing in the sentence is "lanes
>... for which".
Fair enough. I'm happy to hear that I was wrong, since I'd designed
my core the other way . . . :-)
>The sentence is perhaps a little hard to read and the next sentence
>(the one about returning an error in the column before the T) is
>written a bit more clearly. Adding "in the corresponding lane" would
>hammer the point home better (especially since when the next sentence
>uses it the reader thinks that this sentence not having it means
>something), but it shouldn't be essential.
I'd agree -- it's what threw off my verification expert (who initially
pointed out the problem to me) and confused me more the more I thought
about it. I would certainly agree that adding "in the corresponding
lane" would make it much more understandable.
Thanks to everyone who responded!
Doug
-- -------------------------------------------------------------------
___, Doug Massey, ASIC Digital Logic Designer
\o IBM Microelectronics Division, Burlington, Vermont |>
| Phone: (802)769-7095 t/l: 446-7095 fax: x6752 |
/ \ |
. My homepage: http://doug.obscurestuff.com (|)