IEEE P802.3ap D2.2 BackPlane Comments

Transcription

1 I P802.3ap D2.2 BackPlan ommnts l 45 S P 47 L 1 # 1 l 45 S P 25 L 1 # 4 Marris, Arthur adnc Marris, Arthur adnc orrct ditors commnt at th top of th pag rong claus numbr - should b rmov rfrnc to AP. Status l 45 S 45.5 P 49 L 8 Marris, Arthur orrct PIS by corrcting subclaus to for *BP, rmov shalls from PMA/PMD managmnt functions and add tabls for AN options and AN managmnt functions. As abov RJ. adnc ommnt Status R Status h rfrnc itm was dltd from th documnt and should hav bn shown in th compar documnt with a crossout. # 2 orrct th claus numbr for th bit dfinitions for rgistr 1.11 th xtndd ability rgistr. hy should b and Status AP IN PRINIPL. hy should b and l 45 S P 26 L 24 Marris, Arthur Missing word 'to' Rword to 'his bit maps to th stat variabl' also do this for , , , AP. adnc Status # 5 hs changs ar bing don in P802.3an. l 45 S P 28 L 12 # 6 l 45 S P 24 L 32 Marris, Arthur adnc Nd to add modifications for th 10G PMD transmit disabl rgistr modify 802.3an to add th following txt: G PMA/PMD transmit disabl rgistr # 3 Marris, Arthur adnc hang 'h updat gain dfins' to 'h updat gain bits dfin' as abov Status AP. h transmit disabl function for srial PMDs is dscribd in and h transmit disabl function for 10GBAS-KX4 is dscribd in AP. Status add transmit disabl rgistr dfinition as a modification to 802.3aq YP: R/tchnical rquird R/ditorial rquird GR/gnral rquird /tchnical /ditorial G/gnral OMMN SAUS: D/dispatchd A/accptd R/rjctd RSPONS SAUS: O/opn /writtn /closd U/unsatisfid Z/withdrawn SOR ORDR: ommnt ID ommnt ID # 6 Pag 1 of 54 2/15/2006 1:32:28 PM

2 I P802.3ap D2.2 BackPlan ommnts l 45 S P 28 L 27 # 7 l 72 S P 112 L 6 # 10 Marris, Arthur adnc Andr, Szczpank xas Instrumnts IN tabls 45-56, 45-57, 45-58, rmov th undrlining of bits 1:0, 3:2 5:4 and 15. as abov AP. Status l 45 S P L Bradshaw, Ptr R his rfrs to th following unsatisfid commnt: #21152, Both th suggstd rmdy and th rspons to this commnt us th wording "ports typs". h corrct form is "port typs". A sarch of svral sctions of th confirms that "ports typs" dos not appar (xcpt as part of "Supports typs", but all such possibl occurrncs ar givn as "port typs", th grammatically corrct form. h proposd "hang txt to" would introduc th rror. In th rspons "AP IN PRINIPL" to commnt # 152, undr "hang txt to", rplac "ports typs" by "port typs". AP. Status Intrsil orpn. # 8 MIN_LIMI and MAX_LIMI ar not constants. h limits ar subjct to th rstrictions on tap cofficint valus statd in Mov MIN_LIMI and MAX_LIMI to th variabls sction. hang th dscriptiv txt to somthing on lins of... MIN_LIMI Intgr variabl containing th minimum tap cofficint valu, subjct to th constraints dtaild in MAX_LIMI Intgr variabl containing th maximum tap cofficint valu, subjct to th constraints dtaild in AP. Status Although th commnt databas rfrs to "ports typs" whn th chang was mad to th documnt th corrct form "port typs" was usd. hrfor no furthr action is rquird. l 69 S 69B.4 P 208 L 10 Powll, Scott Broadcom # 9 R ** Pil onto Ali Ghiasi's unrsolvd commnt numbr ** S commnt Status AP IN PRINIPL. S rspons to commnt #40. YP: R/tchnical rquird R/ditorial rquird GR/gnral rquird /tchnical /ditorial G/gnral OMMN SAUS: D/dispatchd A/accptd R/rjctd RSPONS SAUS: O/opn /writtn /closd U/unsatisfid Z/withdrawn SOR ORDR: ommnt ID ommnt ID # 10 Pag 2 of 54

3 I P802.3ap D2.2 BackPlan ommnts l 72 S P 114 L 32 # 11 l 74 S P 187 L 5 # 12 Andr, Szczpank xas Instrumnts Andr, Szczpank xas Instrumnts R OFF_UPDA function lacks ""prst"" & ""initializ"" inputs urrntly... OFF_UPDA(cofficint,inc,dc,gain) Adds or subtracts th rqustd gain valu to th cofficint valu. If inc is RU thn th function rturns (cofficint + gain). If dc is RU thn th function rturns (cofficint û gain). Othrwis th function rturns cofficint. Rplac with.. OFF_UPDA(cofficint,prst, initializ,inc,dc,gain) Sts a fixd cofficint valu, or Adds or subtracts th rqustd gain valu to th cofficint valu. If initializ is RU thn th function rturns th cofficint val such that th transmit output mts th conditions dfind in If prst is RU thn th function rturns th cofficint valu quivalnt to no qualization ((-1) & (1) cofficints ar st to zro, (0) st to maximum). If inc is RU thn th function rturns (cofficint + gain). If dc is RU thn th function rturns (cofficint û gain). Othrwis th function rturns cofficint. AP. Status h chang to th initial PN-2112 stat improvs th D balanc of th transmittd transcod bit, but dos not rally scrambl it. A bttr solution would b to scrambl th transcod bit bfor Parity gnration/unscrambl bfor sync r-cration. A simpl way to do this (suggstd by Pat halr) would b to XOR th transcod bit with on of th associatd (scrambld) data bits. his avoids th nd to dfin anothr scramblr, and is vry asy to implmnt - it rquirs just on XOR gat. h adjacnt data bit (0) should not b usd, I hav pickd bit 8 in my suggstd rsolution but I am willing to lav this to th ditors discrtion. I think th chang can b accomplishd with ths two changs hang sction to: h F ncodr conncts to th PS Garbox function using th 16-bit tx_data-group. h F ncodr taks 32 x 64b/66b blocks from PS and ncods it into a singl F block of 2112 bits. h F ncodr comprsss th two sync bits to on transcod bits as xplaind in h transcod bit is thn XOR'd with data bit 8. h rsulting 32 x 65b = 2080 bits with th fram format as shown in abl 74û2 ar fd to th (2112,2080) ncodr, which producs 32 parity-chck bits. h parity chck bits ar appndd to th nd of th F fram. h F fram is scrambld using th PN-2112 psudo-nois squnc as dscribd in and snt to th PMA intrfac. hang sction paragraph 2 (pag 187 lin 34) to: h F dcodr rcovrs and xtracts th information bits using th parity-chck data. In cas of succssful dcoding th dcodr rstors th sync bits in ach of th 64b/66b blocks snt to th PS function, by first XOR'ing th rcivd transcod bit with th associatd data bit 8 and thn gnrating th two sync bits. hn th dcodr is configurd to indicat dcoding rror, th dcodr indicats rror to th PS by mans of stting both sync bits in ach of th 64B/66B block to th sam valu (11), thus forcing th PS sublayr to considr this block as invalid. YP: R/tchnical rquird R/ditorial rquird GR/gnral rquird /tchnical /ditorial G/gnral OMMN SAUS: D/dispatchd A/accptd R/rjctd RSPONS SAUS: O/opn /writtn /closd U/unsatisfid Z/withdrawn SOR ORDR: ommnt ID ommnt ID # 12 Pag 3 of 54

4 I P802.3ap D2.2 BackPlan ommnts AP IN PRINIPL. Status In addition to th proposd rspons, nsur that it is clar which bit is bit 8. Rvrt PN-2112 initialization back to th txt in draft 2.1. (binary squnc ) l 69B S 69B.4.1 P 209 L 5 Moor, harls R # 13 hannl attnuation is too larg. For KR port typ, signal for worst cas channl with worst cas attnuation and rippl will not not hav nough margin to work in th prsnc of worst cas PSX. rduc channl attnuation by 3dB at GHz. Suggstd valus for b1-b4 will b providd in a sparat prsntation. AP IN PRINIPL. Status Avago chnologis Adopt valus on P. 8 in gorgn_01_0206.pdf, basd on I rsults prsntd to dat. l 69B S 69B P 214 L 42 Moor, harls R # 14 rosstalk lvls ar too high. hy rprsnt an intrfrnc tolranc which rasonabl rcivrs will not b abl to handl vn with rducd channl loss. ithr incras IR by 12dB or chang mthod of spcifying crosstalk. Altrnat mthod will b dscribd in a sparat prsntation. AP IN PRINIPL. ithr chang quation 69B-21 from Status Avago chnologis IRfit (f ) > = IRmin (f) = log(f/5ghz) l 72 S P 129 L 6 Moor, harls R # 15 orst cas combination of ffcts of crosstalk, slf intrfrnc, and DD add up to an quivalnt of an Ibaslin of 21mV, 15. hang Ibaslin valu in spcification to 21mV. Justification for this numbr was snt to th rflctor and will b rpatd in a sparat prsntation. RJ. ommnt Status R Status Avago chnologis In light of changs to Amax and IRmin (S ommnts #13 and #14 ), th study of th propr I valu nds to b rvisitd. l 72 S P 120 L 26 Mllitz, Richard R abl UI DD is too high Rfr to krooswyk_c1_0106.pdf Suggst changing DD to 0.03 UI RJ. h task forc can not rach consnsus to mak th chang. Straw Poll #3 a. Rduc DD to 0.03UI b. Kp DD at 0.05UI A - 6 B - 9 abstain - 3 U ommnt Status R Status U # 16 to: IRfit (f ) > = IRmin (f) = log(f/5GHz) YP: R/tchnical rquird R/ditorial rquird GR/gnral rquird /tchnical /ditorial G/gnral OMMN SAUS: D/dispatchd A/accptd R/rjctd RSPONS SAUS: O/opn /writtn /closd U/unsatisfid Z/withdrawn SOR ORDR: ommnt ID ommnt ID # 16 Pag 4 of 54

5 I P802.3ap D2.2 BackPlan ommnts l 69A S 69a.2.2 P 203 L 23 # 17 l 71 S P 95 L 43 # 19 Mllitz, Richard Dudk, Mik Picolight R h tst channl nd to hav a rturn loss spc so that it is not doubl counting in I voltag. h tst channl hav a rturn loss gratr than 20 db for a rang spanning f1 and f2. AP IN PRINIPL. S commnt 136. Status l 71 S P 85 L 36 Dudk, Mik R # 18 Sub claus says that transmit disabl shall driv a constant lvl that dos not xcd th maximum diffrntial pak-to-pak output voltag in abl h only maximum diffrntial pak-to-pak output voltag in that tabl is th normal valu (not disabld). Insrt an xtra row in tabl 71.4 with Maximum diffrntial pak-to-pak output voltag whn disabld. RJ. xt in is saying that th diffrntial voltag during disabl is constraind by th sam spcification as th maximum diffrntial pak-to-pak output voltag during normal opration, as dfind in abl Motion #3 chnical (>=75%) Mov to accpt rspons. Mov - John D'Ambrosia Scond - om Palkrt All Y- 13 N- 0 A- 2 Picolight ommnt Status R Status R hr is tchnical dtail in th PIS that contradicts th main documnt in FS10, FS11, FS12, and 5 Fix it. Status AP. Mak FS10, FS11, FS12, and 5 consistnt with th txt. l 72 S P 102 L 53 Dudk, Mik h rfrnc to figur 72-2 dosn't mak sns. hang Figur 72-2 to Figur 72-1 AP. Picolight Status l 71 S P 90 L 19 Dudk, Mik # 20 # 21 h rfrncs to th figurs appar to b wrong. (Fig 71-5 is diffrntial rturn loss which dosn't mak sns. hang fig 71-5 to fig 71-6 (or mayb it should b fig 71-1) RJ. xt in lan vrsion was corrct Fig.rf is to fig71-2 and fig71-5 his is a Fram difrntial Issu Picolight ommnt Status R Status Motion passs YP: R/tchnical rquird R/ditorial rquird GR/gnral rquird /tchnical /ditorial G/gnral OMMN SAUS: D/dispatchd A/accptd R/rjctd RSPONS SAUS: O/opn /writtn /closd U/unsatisfid Z/withdrawn SOR ORDR: ommnt ID ommnt ID # 21 Pag 5 of 54

6 I P802.3ap D2.2 BackPlan ommnts l 72 S P 120 L 16 # 22 l 71 S P 91 L 32 # 24 Dudk, Mik Picolight Spagna, Fulvio INL R Subclaus stats that th transmit disabl shall driv a constant lvl not to xcd th maximum diffrntial voltag pak-to-pak voltag in abl h only valu of maximum diffrntial voltag in that tabl is th normal valu which dosn't mak sns. Add row to abl 72-6 with maximum diffrntial voltag with x disabl. RJ. xt in is saying that th diffrntial voltag during disabl is constraind by th sam spcification as th maximum diffrntial pak-to-pak output voltag during normal opration, as dfind in abl S also commnt #18. ommnt Status R Status l 69 S B.4 P 208 L 1 Agarwal, Punt R # 23 ommnt #127 on draft 2.1 (from Ali Ghiasi) was rjctd saying that ""no srvicabl rturn loss mask had bn proposd"", implying that th suggstd rmdy was not tchnically complt. I think th draft itslf is not tchnically complt unlss it provids spcification for th channl. Spcifically, th draft nds to at last spcify crosstalk, rturn loss, insrtion loss and insrtion loss to crosstalk ratio. Dfin channl spcifications in th spc as statd abov. AP IN PRINIPL. S rspons to commnt #40 Status Broadcom R I bliv that th intnt of Not 1 to abl 71-8 should b to dfin th total RMS jittr, J, as a function th sinusoidal jittr, SJ, and th random jittr RJ and not as a function of DJ and RJ. hang not 1 to rad: RMS jittr is dfind to b: Jrms = Sqrt[ ((SJrms)^2)/8 +((RJpp/14.06)^2) ] whr SJrms is th sinusoidal jittr rfrncd to in 69A.2.1 and RJpp is th pak-to-pak random jittr dfind in abl 71-4 AP IN PRINIPL. h not is intndd to dfin th origin of th rms jittr valu in abl h quantity SJrms in th not should b rnamd RMS jittr to b consistnt with th tabl. S commnt 32, 35. Status l 72 S P 109 L 7 Spagna, Fulvio R # 25 Mak th Initializ updat protocol consistnt with th officint updat protocol. Insrt following sntnc at thn of lin 7 : ""At that point th outgoing rqust shall b st to zro"" AP IN PRINIPL. INL Status Insrt th following sntnc at thn of lin 7: At that poin th outgoing initializ fild shall b st to zro. YP: R/tchnical rquird R/ditorial rquird GR/gnral rquird /tchnical /ditorial G/gnral OMMN SAUS: D/dispatchd A/accptd R/rjctd RSPONS SAUS: O/opn /writtn /closd U/unsatisfid Z/withdrawn SOR ORDR: ommnt ID ommnt ID # 25 Pag 6 of 54

7 I P802.3ap D2.2 BackPlan ommnts l 45 S P 28 L 25 # 26 l 73 S 5.3 P 141 L 13 # 29 Spagna, Fulvio INL Spagna, Fulvio INL R R In abl 45-57, bits 12 and 13 ar not corrctly dfind. DM pag timing summary abl numbr is duplicatd. Updat tabl to match bits dfinition in abl 72-7 on pag 108 Status AP. l 45 S P 30 L 6 Spagna, Fulvio INL R In abl 45-59, bits 12 and 13 ar not corrctly dfind. Updat tabl to match bits dfinition in abl 72-7 on pag 108 Status AP. # 27 hang tabl hading to abl On lin 13, pag 141 chang rfrnc to abl 73-1 inot abl 73-2 AP. Status l 45 S P 30 L 6 Spagna, Fulvio R INL # 30 h bits in this rgistr ar dfind as R/. I intrprt this to man that th rgistr's contnt can b writtn by th dvic, to rport th valu of th vairabls associatd to ths bits, and by th usr through MDIO managmnt. his bing th cas, thr is nothing that prvnts a conflict from happning, that is an instanc in which both th dvic and th usr try to accss th rgistr. l 70 S 7.1 P 68 L 45 Spagna, Fulvio R # 28 For notational consistncy with claus 71 and 72 chang th jittr units from UIpp to UI. h paramtr listd in th tabl is alrady dfind as pak-to-pak. hang units from UIpp to UI. AP. INL Status Mor in gnral, if th intnt is to provid managmnt a way to ovrrid th LD cofficint updat, thr dos not sm a mchanism to insur that a valu writtn to th rgistr by managmnt can b actd upon bfor bing ovrwrittn by th dvic. Mak th rgistr rad only (RO). AP IN PRINIPL. Status Mak th LD status rport rgistr rad only. YP: R/tchnical rquird R/ditorial rquird GR/gnral rquird /tchnical /ditorial G/gnral OMMN SAUS: D/dispatchd A/accptd R/rjctd RSPONS SAUS: O/opn /writtn /closd U/unsatisfid Z/withdrawn SOR ORDR: ommnt ID ommnt ID # 30 Pag 7 of 54

8 I P802.3ap D2.2 BackPlan ommnts l 45 S P L # 31 l 70 S P 73 L 32 # 32 Spagna, Fulvio INL Spagna, Fulvio INL R h bits in this rgistr ar dfind as R/. I intrprt this to man that th rgistr's contnt can b writtn by th dvic, to rport th valu of th vairabls associatd to ths bits, and by th usr through MDIO managmnt. his bing th cas, thr is nothing that prvnts a conflict from happning, that is an instanc in which both th dvic and th usr try to accss th rgistr. Mor in gnral, if th intnt is to provid managmnt a way to ovrrid th LD cofficint updat, thr dos not sm a mchanism to insur that a valu writtn to th rgistr by managmnt can b actd upon bfor bing ovrwrittn by th dvic. Mak th rgistr rad only (RO). AP IN PRINIPL. Status h rgistr will b mad rad only (RO). As a corollary to this discussion, rgistr 152 will b mad R with th writ function accssibl only whn training_nabl is dassrtd. R I bliv that th intnt of Not 1 to abl 70-8 should b to dfin th total RMS jittr, J, as a function th sinusoidal jittr, SJ, and th random jittr RJ and not as a function of DJ and RJ. hang not 1 to rad: RMS jittr is dfind to b: Jrms = Sqrt[ ((SJrms)^2)/8 +((RJpp/14.06)^2) ] whr SJrms is th sinusoidal jittr rfrncd to in 69A.2.1 and RJpp is th pak-to-pak random jittr dfind in abl 70-4 AP IN PRINIPL. h not is intndd to dfin th origin of th rms jittr valu in abl h quantity SJrms in th not should b rnamd RMS jittr to b consistnt with th tabl. ditorial - corrct splling of "jittr" in tst pattrn row. S commnt #35, 24. Status l 72 S P 108 L 46 Spagna, Fulvio INL # 33 R Mak th Prst updat protocol consistnt with th officint updat protocol. Insrt following sntnc in lin 46 : ""At that point th outgoing rqust shall b st to zro"" prior to ""Maximum status... "" Status AP IN PRINIPL. Insrt th following sntnc at thn of lin 7: At that point th outgoing initializ fild shall b st to zro. YP: R/tchnical rquird R/ditorial rquird GR/gnral rquird /tchnical /ditorial G/gnral OMMN SAUS: D/dispatchd A/accptd R/rjctd RSPONS SAUS: O/opn /writtn /closd U/unsatisfid Z/withdrawn SOR ORDR: ommnt ID ommnt ID # 33 Pag 8 of 54

9 I P802.3ap D2.2 BackPlan ommnts l 72 S 7.1 P 120 L 3 # 34 l 72 S P 128 L 14 # 35 Spagna, Fulvio INL Spagna, Fulvio INL R Add dfinition for Duty ycl Distortion Rplac lin 31 with th following txt: ""h Duty ycl Distortion is dfind as th ratio of th diffrnc in th man puls width of a æ1æ puls compard to th man puls width of a æ0æ puls (as masurd in a clock-lik rpating 0,1,0,1,à bit squnc) and th nominal puls width. Duty ycl Distortion is considrd part of th dtrministic jittr distribution."" AP. ""h Duty ycl Distortion is dfind as th unsignd ratio of th diffrnc in th man puls width of a "1" puls compard to th man puls width of a "0" puls (as masurd at th A coupld "0" crossing in a clock-lik rpating 0,1,0,1 bit squnc) and th nominal puls width. Duty ycl Distortion is considrd part of th dtrministic jittr distribution."" S commnt #44 Motion#4 Moiton to rconsidr. Movd by Fulvio Spagna Scond - Howard Baumr Passd without objction Status Rconsidrd rsolution - ""h pak-to-pak duty cycl distortion (DD) is th unsignd ratio of th diffrnc in th man puls width of a "1" puls compard to th man puls width of a "0" puls (as masurd at th A coupld "0" crossing in a clock-lik rpating 0,1,0,1 bit squnc) and th nominal puls width. DD is considrd part of th dtrministic jittr distribution."" R I bliv that th intnt of Not 1 to abl should b to dfin th total RMS jittr, J, as a function th sinusoidal jittr, SJ, and th random jittr RJ and not as a function of DJ and RJ. hang not 1 to rad: RMS jittr is dfind to b: Jrms = Sqrt[ ((SJrms)^2)/8 +((RJpp/14.06)^2) ] whr SJrms is th sinusoidal jittr rfrncd to in 69A.2.1 and RJpp is th pak-to-pak random jittr dfind in abl 72.9 AP IN PRINIPL. h not is intndd to dfin th origin of th rms jittr valu in abl h quantity SJrms in th not should b rnamd RMS jittr to b consistnt with th tabl. S commnt 24, 32. Status l 69B S 4.3 P 210 L 13 Spagna, Fulvio R # 36 quations 69B-7 and 69B-8 ar missing. Also, th ""clan"" vrsion of D2p2, which shows th two quations, but rfrncs A(f) instad of Amax(f) quations 69B-7 and 69B-8 in D2p2 should rfrnc Amax(f) and not A(f) AP. INL Status YP: R/tchnical rquird R/ditorial rquird GR/gnral rquird /tchnical /ditorial G/gnral OMMN SAUS: D/dispatchd A/accptd R/rjctd RSPONS SAUS: O/opn /writtn /closd U/unsatisfid Z/withdrawn SOR ORDR: ommnt ID ommnt ID # 36 Pag 9 of 54

10 I P802.3ap D2.2 BackPlan ommnts l 69B S 69B.4 P 208 L 17 # 37 l 72 S P 129 L 5 # 39 lang, Vivk Broadcom Baumr, Howard Broadcom R his is a pil-on to Ali Ghiasi's UNSAISFID commnt #21127 For introprability to b guarantd, th standard should hav a normativ spcification for th channl rturn loss. S quation in Howard Baumr's commnt AP IN PRINIPL. S rspons to commnt #40 Status l 72 S P 120 L 22 Baumr, Howard # 38 ransition tims in abl 72-9 do not match th txt in abl has 24-40ps and txt has 24-47ps. Pick on rang and mak txt and tabl match. AP IN PRINIPL. Status Broadcom Fix tabl 72.6 to indicat max transition tim of 47ps R Pil onto commt #56 D2.1 ith backplan systms having upwards of 100+ links a targt BR spcifid to 1-15 should at last b spcifid along with th standard BR of hr sould at last b a st of ""shalls"" that go with a BR of 1-15 vn if it maks th inclusion of F mandatory. ommnt Status R Add a nw tabl ntry of: BR with F on, 1-15 or add a scond st of tst with BR at 1-15 Status RJ. h following rspons was accptd by th ask Forc without objction. ommnt #57 from D2.1 was a pil-on to #613 from D2.0. A pil-on to a pil-on is not considrd a valid nw commnt. Rcirculation rquirmnts hav bn mt. h rspons to commnt #57 is rpatd blow. h ask Forc objctiv is to support a BR of 1-12 or bttr, and thrfor th prformanc targts ar within th objctivs. U Howvr, th ask Forc rcognizs that som systms may rquir backplan links that prform bttr than th statd 1-12 targt. It is suggstd that th Forward rror orrction sublayr dfind in laus 74 b utilizd to supply this additional prformanc. It has bn shown in: that links xhibiting 1-9 pformanc improv to bttr than hrfor, links oprating at 1-12 can b xpctd to improv to 1-15 or bttr via us of th laus 74 F. ith rgards to tstability, th intrfrnc tolranc tst procdur vrifis rcivr prformanc, without F, to a BR targt of 1-12 or bttr. Mathmatic tchniqus may thn b applid to driv th rcivr prformanc with th bnfit of F. YP: R/tchnical rquird R/ditorial rquird GR/gnral rquird /tchnical /ditorial G/gnral OMMN SAUS: D/dispatchd A/accptd R/rjctd RSPONS SAUS: O/opn /writtn /closd U/unsatisfid Z/withdrawn SOR ORDR: ommnt ID ommnt ID # 39 Pag 10 of 54

11 I P802.3ap D2.2 BackPlan ommnts l 69B S 69B.4 P 242 L 20 Baumr, Howard Broadcom R his is a pil on to D2.1 commnt #21127 # 40 1.ircuits for producing a rturn loss wors than som lin ovr a broad rang of frquncis without gratly distorting th transfr charactristics hav not bn shown. 2.Having rturn loss abov -10 to -4dB but not othrwis spcifid will rsult in vry unrpatabl tst rsults. Rcommnd chang 69B hr ar thr impairmts to th systm: 1) loss in th channl 2) rflctions in th channl 3) nois injctd into th channl from xtrnal sourcs All thr of ths impairmnts nd to b compltly spcifid as conditions of th tst. h loss of th ""tst channl"" (condition #1) is only partially spcifid, th slop of th loss is spcifid (m) whras th D.. componnt is not (b). h rturn loss of th ""tst channl"" and of th tst signal is not spcifid at all rsulting in partially spcifying th tst condtions for th rflctions in th channl (th rcivr's rturn loss is spcifid as it is what is undr tst). Add sub claus 69B4.5 and chang numbr of xisting 69B4.5 to 69B4.5 tc. 69B4.5 Rturn loss h rcommndd rturn loss of th intrconnct, masurd at ithr nd, b such that: Rturn Loss >= *log10(f/350MHz) for 500MHz <= f <= 3GHz Rturn Loss > 4 for 3GHz < f <= 8GHz ithout th complt spcification for #1 & #2, th signal going into th rcivr undr tst will b unknown thrfor invalidating any tst rsults. vn using th I tst to mulat th affcts of rturn loss (slf intrfrnc) is invalidatd for without knowing what th tru signal into th rcivr undr tst is you cannot know what lvl of I is appropriat. It cannot b assumd that all tst stups will b built with th bst possibl conditions (i.. no rturn loss). Spcify that th rturn loss of th channl must b Rturn Loss >= *log10(f/350MHz) for 500MHz <= f <= 3GHz Rturn Loss > -4 for 3GHz < f <= 8GHz Not: rturn loss is spcifid as ngativ db hr and th limits ar such that th compianc (tst) channl should hav a rturn loss that is at ths lvls or wors. h rason for making th tst channl wors is that ths ar complianc conditions on th rcivr and thrfor th rcivr undr tst nds to b abl to tollrat ths conditions or wors. Not #2: h tst complianc channl can b built by crating a vry clan PB with a passiv circuit ntwork that crats th appropriat insrtion loss and rturn loss. AP IN PRINIPL. Status Howard has indicatd that this is rally 2 commnts, rqusting a intrconnct rturn loss limit to go into annx 69B and an idntical spc but with th opposit sign for th intrfrnc tolranc tst channl rturn loss in annx 69A. A diffrnt I hannl Rturn Loss modl was adoptd (s commnt #136) for th following rasons : YP: R/tchnical rquird R/ditorial rquird GR/gnral rquird /tchnical /ditorial G/gnral OMMN SAUS: D/dispatchd A/accptd R/rjctd RSPONS SAUS: O/opn /writtn /closd U/unsatisfid Z/withdrawn SOR ORDR: ommnt ID ommnt ID # 40 Pag 11 of 54

12 I P802.3ap D2.2 BackPlan ommnts l 73A S 73A.2 P 187 L 36 # 41 l 73A S 73A.3 P 187 L 51 # 42 Baumr, Howard Broadcom Baumr, Howard Broadcom R Subclaus, pag and lin numbr ar for fil p802.3ap-d2.2.pdf Mssag cod #5 is improprly ncodd. h whol point of th xtndd nxt pags was to hav to xchang lss nxt pags. h currnt ncoding of MP5 rquirs 1 mssag nxt pag and 4 unformattd nxt pags, howvr, if it took full advantag of th xtdd nxt pags it would only nd 1 mssag nxt pag and 1 unformattd nxt pag. hang th ntir 73A.2 to rad: h OUI aggd Mssag shall consist of a singl mssag cod of followd by on usr cod dfind as follows. h unformattd cod fild of Mssag Nxt Pag 5 shall contain th most significant 11 bits of th OUI (bits 23:13) with th most significant OUI bit in bit 26 (bit U11) of th unformattd cod fild, th nxt 11 most significant bits of th OUI (bits 12:2) in bits 42:32 (bit U26 to U16) of th unformattd cod fild and th rmaing unformattd cod fild bits, bits 31:27 (bits U15 to U10) and bits 47:43 (bits U31 to U27) as a usr-dfid usr cod valu that is spcific to th OUI transmittd. h unformattd nxt pag usr cod fild shall contain th rmaining last significant 2 bits of th OUI (bits 1:0) with OUI bit 1 in bit 1 (bit U1) with th rmaining bits 10:2, 47:16 (bits U10 to U2, U47 to U16) as a usr-dfind usr cod valu that is spcific to th OUI transmittd. For xampl, assum that a manufacturr's I-assignd OUI valu is A-D-48 and th manufacturr-slctd usr-dfind usr cod associatd with th OUI is 3--1F-A-35- D9-B2. h mssag cod valus gnratd from ths two numbrs is ncodd into th mssag nxt pag and unformattd nxt pag cods, as spcifid in Figur 73Aû7. For clarity, th position of th global broadcast g is illustratd. hang Figur 73A-7 to match abov paragraph. Status AP. Match to th way th xtndd nxt pags ar usd to snd ths mssags in 28. ditor''s not: in adding this dit, I found that th figur showing th bit placmnt didn't match th txt. hat also applis to th sourc txt in laus 28. Sinc th figur is nwr than th txt (it was addd in Rv AM ballot), th laus 73 figur has bn modifid to match th txt. R Subclaus, pag and lin numbr ar for fil p802.3ap-d2.2.pdf Mssag cod #6 is improprly ncodd. h whol point of th xtndd nxt pags was to hav to xchang lss nxt pags. h currnt ncoding of MP6 rquirs 1 mssag nxt pag and 4 unformattd nxt pags, howvr, if it took full advantag of th xtdd nxt pags it would only nd 1 mssag nxt pag and 1 unformattd nxt pag. h PHY ID tag cod mssag shall consist of a singl mssag cod of followd by on usr cods dfind as follows. h unformattd cod fild of Mssag Nxt Pag 6 shall contain th most significant 11 bits of th PHY ID (2.15:5) with th most significant PHY ID bit in bit 26 (bit U11) of th unformattd cod fild, th nxt 11 most significant bits of th PHY ID (bits 2.4:0 to 3.15:10) in bits 42:32 (bit U26 to U16) of th unformattd cod fild and th rmaing unformattd cod fild bits, bits 31:27 (bits U15 to U10) and bits 47:43 (bits U31 to U27) as a usr-dfid usr cod valu that is spcific to th PHY ID transmittd. h unformattd nxt pag usr cod fild shall contain th rmaining last significant 10 bits of th PHY ID (bits 3.9:0) in bits 9:0 (bit U9 to U10) with th rmaining bits 10, 47:16 (bits U10, U47 to U16) as a usr-dfind usr cod valu that is spcific to th OUI transmittd. Status AP IN PRINIPL. Match to th way th xtndd nxt pags ar usd to snd ths mssags in 28. l 72 S P 125 L 5 Baumr, Howard R # 43 hr is nothing constraiing th ngativ dg of th transmittr output wavform. Add txt that constrains th nngativ dg th sam way th positiv dg is constraind. RJ. S commnt #45. ommnt Status R Status Broadcom U YP: R/tchnical rquird R/ditorial rquird GR/gnral rquird /tchnical /ditorial G/gnral OMMN SAUS: D/dispatchd A/accptd R/rjctd RSPONS SAUS: O/opn /writtn /closd U/unsatisfid Z/withdrawn SOR ORDR: ommnt ID ommnt ID # 43 Pag 12 of 54

13 I P802.3ap D2.2 BackPlan ommnts l 72 S P 102 L 47 # 44 l 99 S 99 P 1 L 1 # 46 Baumr, Howard Broadcom Haly, Adam Agr Systms R Duty ycl Distortion is not dfind Dfin Duty ycl Distortion Status AP. S commnt #34 R Draft covr pag should contain a dscription of th status of this documnt as it rlats to th standards procss as wll as an xpiration dat. In addition, th hadr of ach pag should not th publication dat of th draft. Updat covr pag and documnt tmplat accordingly. AP. Status l 72 S P 106 L 49 Baumr, Howard Broadcom # 45 l 72 S abl 72-9 P 120 L 22 Haly, Adam Agr Systms # 47 R ransmittr output wavform is not constraind tightly nough allowing many diffrnt output wavform shaps. S attchd xampl, Vout_KR.pdf, showing two diffrnt transmittr output wavforms that would rsult in diffrnt pr-qualization. his crats an unknown for th rcivr dsign. Although I do not hav a mor prcis transmittr output wavform constraints th task forc nds to mor prcisly spcify th transmittr. RJ. Rfrnc baumr_01_0206. Suggstd rmdy is incomplt and prsntd matrial dos not hav consnsus. ommnts on this issu in D2.3 rcirculation will not b ruld out of scop. Motion #8 chnical (>=75%) Mov to accpt rspons. Mov John D'Ambrosia Scond Schlto van Doorn ommnt Status R Status U abl is not consistnt with rfrncd txt ( ). hang transition tim rang to ps AP. S 38 Status l 72 S P 124 L 41 Haly, Adam # 48 h draft now contains a prcis dscription of ""no qualization,"" and it should b utilizd. Add statmnt to th ffct that: ""ransmittr qualization may b disabld via th prst control dfind in "" AP. Status Agr Systms Motion passs by acclamation. YP: R/tchnical rquird R/ditorial rquird GR/gnral rquird /tchnical /ditorial G/gnral OMMN SAUS: D/dispatchd A/accptd R/rjctd RSPONS SAUS: O/opn /writtn /closd U/unsatisfid Z/withdrawn SOR ORDR: ommnt ID ommnt ID # 48 Pag 13 of 54

14 I P802.3ap D2.2 BackPlan ommnts l 72 S P 125 L 3 # 49 l 72 S P 127 L 6 # 51 Haly, Adam Agr Systms Haly, Adam Agr Systms h draft now contains a prcis dscription of ""qualization off,"" and it should b utilizd. Add statmnt to th ffct that: ""ransmittr qualization may b disabld via th prst control dfind in "" AP. Status l 72 S abl P 126 L 40 Haly, Adam # 50 h draft now contains a prcis dscription of ""qualization off,"" and it should b utilizd. o simply th tabl, chang rfrncs from ""qualization off"" to ""disabld"". Add a footnot to abl stating that: ""A cofficint may b disabld by first assrting th prst control dfind in , and thn manipulating th othr cofficints as rquird by th tst."" Status Agr Systms AP IN PRINIPL. Us proposd txt and updat th cross rfrnc to point to Rgistr 152 in claus 45. h conditions citd in c) ar only valid whn th transition tims ar in th lowr nd of th prmittd rang. At th highr nd of th rang, this formula will tnd to undrstimat th pak voltag. It is mor prcis to vrify th pak voltag using a pattrn. Dfin a scond tst pattrn ( ) to b usd for pak output voltag masurmnt. Not that whn qualization is off, it is possibl that th stady-stat voltag masurd pr will xcd th pak voltag of this nw tst pattrn. hrfor, th rquirmnt must b r-phrasd such that thlargr of th two valus, Vpk and Vss, must b lss than th 600 mv limit. Appropriat txt will b supplid to ditor on rqust. AP IN PRINIPL. Dlt Bullt in Status Add to th rquirmnt that for a 1010 pattrn, th pak to pak diffrntial output voltag shall b lss than 1200mV, rgardlss of qualization stting. l 74 S P 187 L 38 Haly, Adam # 52 h txt stats that, whn th dcodr is st to indicat at dcoding rror, th sync hadr for th block is st to 11. It is not clar whthr this is intndd to vry block in th F fram, or a spcific block within th fram. If it is intndd to b vry block in th fram, this would translat to 32 conscutiv sync hadr rrors, which would forc th claus 49 lock stat machin out of block lock. Is this what was intndd? larify whthr all of th blocks in th F fram ar populatd with th invalid sync hadr, or spcifically which blocks is to b corruptd. AP IN PRINIPL. Status Agr Systms Provid which blocks ar to b corruptd with Snding '11'. Nd to mark "11" in atlast on in vry 8th block to guarant a minimum siz thrnt fram is droppd. his is blow th thrshold that caus th stat machin to go out of sync. Mark th 1st, 9th, 17th, 25th, 32nd, 66 bit block of th dcodd F blocks. YP: R/tchnical rquird R/ditorial rquird GR/gnral rquird /tchnical /ditorial G/gnral OMMN SAUS: D/dispatchd A/accptd R/rjctd RSPONS SAUS: O/opn /writtn /closd U/unsatisfid Z/withdrawn SOR ORDR: ommnt ID ommnt ID # 52 Pag 14 of 54

15 I P802.3ap D2.2 BackPlan ommnts l 45 S P 28 L 5 # 53 l 70 S P 69 L 51 # 56 Haly, Adam Agr Systms Haly, Adam Agr Systms Prst and Initializ controls addd in laus 72 nd to b rflctd in th laus 45 rgistrs. Add Prst and Initializ controls and cross-rfrnc to claus 72 (s also ). AP. Status l 70 S P 67 L 30 Haly, Adam h txt rally dosn't rflct th optional natur of th signal dtct function. # 54 Rplac txt of with: ""PMD signal dtct is optional for 1000BAS-KX and its dfinition is byond th scop of this spcification. hn PMD signal dtct is not implmntd, th valu of SIGNAL_D shall b st to OK for purposs of managmnt and signaling of th primitiv."" AP. Status Agr Systms l 71 S P 83 L 44 Haly, Adam # 55 h txt rally dosn't rflct th optional natur of th signal dtct function. It could also b r-organizd to add clarity. Agr Systms Rplac th txt of with: ""Globabl PMD signal dtct is optional for 10GBAS-KX4 and its dfinition is byond th scop of this spcification. hn Global PMD signal dtct is not implmntd, th valu of SIGNAL_D shall b st to OK for purposs of managmnt and signaling of th primitiv. h txt in this subclaus was basd on similar txt in laus 54, which has bn clarifid via Maintnanc Rqust his is usful clarification that should b includd in P802.3ap. hang: ""h nominal diffrntial impdanc of th transmit tst fixtur dpictd in..."" to: ""h diffrntial load impdanc applid to th transmittr output by th tst fixtur dpictd in..."" AP. Status l 71 S P 86 L 51 Haly, Adam # 57 h txt in this subclaus was basd on similar txt in laus 54, which has bn clarifid via Maintnanc Rqust his is usful clarification that should b includd in P802.3ap. hang: ""h nominal diffrntial impdanc of th transmit tst fixtur dpictd in..."" to: ""h diffrntial load impdanc applid to th transmittr output by th tst fixtur dpictd in..."" AP. Status Agr Systms hn th MDIO is implmntd, ach PMD_signal_dtct_n valu, whr n rprsnts th lan numbr in th rang 0:3, shall rport th valu of signal_dtct for th corrsponding lan whn signal dtct is implmntd, or OK othrwis."" AP. Status YP: R/tchnical rquird R/ditorial rquird GR/gnral rquird /tchnical /ditorial G/gnral OMMN SAUS: D/dispatchd A/accptd R/rjctd RSPONS SAUS: O/opn /writtn /closd U/unsatisfid Z/withdrawn SOR ORDR: ommnt ID ommnt ID # 57 Pag 15 of 54

16 I P802.3ap D2.2 BackPlan ommnts l 72 S P 121 L 30 # 58 l 72 S P 121 L 49 # 59 Haly, Adam Agr Systms Haly, Adam Agr Systms h txt in this subclaus was basd on similar txt in laus 54, which has bn clarifid via Maintnanc Rqust his is usful clarification that should b includd in P802.3ap. hang: ""h nominal diffrntial impdanc of th transmit tst fixtur dpictd in..."" to: ""h diffrntial load impdanc applid to th transmittr output by th tst fixtur dpictd in..."" AP. Status Insrtion loss to crosstalk ratio limits assum that th transmittr and aggrssors ar of th sam PHY typ, hav similar output amplitud, and similar qualization sttings. Howvr, if it this is not th cas, hypothtical scnarios xist whr th transmittr is configurd for a low pak output voltag and high transmit qualizr gain (Vpk = 400 mv, Vss = 40 mv) whil th aggrssors hav high output voltag and low qualization gain (Vpk = Vss = 600 mv). his crats an unfavorabl signal-to-nois ratio. ommnt Status R If it cannot b assumd that th transmittr and aggrssors ar idntical, thr ar multipl approachs to mitigating this issu. 1. Sinc th pak output amplitud is programmabl, incras th minimum output amplitud rquirmnt to accomodat longr channls and highr nois (lowr valus can b configurd for lowr loss/nois channls) 2. Limit th maximum output voltag to only b availabl for transmittr configurations with a crtain lvl of qualization gain 3. Or, som combination throf. A supporting prsntation will b supplid with additional dtail. Status RJ. his commnt was IHDRAN by th commntr. l 72 S P 114 L 6 Nosworthy, Bob UNH-IOL # 60 hangbar vrsion, updat_status variabl rfrs to th officint Status fild ""as dfind in abl 72-7"", but this tabl is abl 72-8 Status rport fild. non-changbar vrsion updat_status pag 98 lin 43, rfrs to abl 72-4, should b abl 72-5 Status Rport Fild Fix rfrnc to Status rport fild tabl in dfinition of updat_status AP. Status YP: R/tchnical rquird R/ditorial rquird GR/gnral rquird /tchnical /ditorial G/gnral OMMN SAUS: D/dispatchd A/accptd R/rjctd RSPONS SAUS: O/opn /writtn /closd U/unsatisfid Z/withdrawn SOR ORDR: ommnt ID ommnt ID # 60 Pag 16 of 54

17 I P802.3ap D2.2 BackPlan ommnts l 69A S 69A.1.1 P 201 L 32 # 61 l 73 S P 153 L 13 # 64 Nosworthy, Bob UNH-IOL Nosworthy, Bob UNH-IOL rplac 'ar' with 'and': ""Othr trms ar abbrviations ar"" to ""Othr trms and abbrviations ar"" rplac 'ar' with 'and': ""Othr trms ar abbrviations ar"" to ""Othr trms and abbrviations ar"" AP. Status l 74 S Figur 74-9 P 188 L 25 Nosworthy, Bob # 62 figur 74-9 and 74-8 (both ""F (2112,2080) dcoding"") would sm to b duplicatd, strik on and rnumbr appropriatly. Dlt duplicat figur AP IN PRINIPL. Status Strik 74-8 and rnumbr as appropriat UNH-IOL l 73 S Figur 73-2 P 140 L 24 Nosworthy, Bob # 63 Figur 73-2 is calld ""PS idl randomizr"" but has nothing to do with a PS or idl Rnam Figur 73-2 to ""DM pag bit 49 randomizr"" AP. Status UNH-IOL ditorial: in dfinition of ack_nonc_match ""acknowldg_match"" should b ""acknowldg_match"" ""acknowldg_match"" should b ""acknowldg_match"" AP. Status l 73 S P 152 L 49 Nosworthy, Bob ability_match_word[48:1] and dfinition ar improprly indntd. orrct indntation of ability_match_word variabl and dfinition AP. Status UNH-IOL l 72 S P 115 L 21 Nosworthy, Bob # 65 # 66 changbar vrsion, rfrs to ""officint Updat stat diagram as dpictd in Figur 72-7"" but is Figur non-changbar vrsion lin 53 pag 99, rfrs to Figur 72-5, should b 72-6 Fix rfrncd figur to rfr to actual officint Updat stat diagram. AP IN PRINIPL. Status UNH-IOL YP: R/tchnical rquird R/ditorial rquird GR/gnral rquird /tchnical /ditorial G/gnral OMMN SAUS: D/dispatchd A/accptd R/rjctd RSPONS SAUS: O/opn /writtn /closd U/unsatisfid Z/withdrawn SOR ORDR: ommnt ID ommnt ID # 66 Pag 17 of 54

18 I P802.3ap D2.2 BackPlan ommnts l 74 S Figur P 190 L 25 # 67 l 74 S P 191 L 50 # 68 Nosworthy, Bob UNH-IOL Nosworthy, Bob UNH-IOL h bhavior whn F rror Indication is not st is not clar to th radr. (pag 36) dfins bit ""nabl F rror Indication"" indicats that if zro ""th rror indication function is disabld""(lin 30) dos not clarly stat what occurs if = allows forcd rror indication to th PS to b nabld or disabld by managmnt. As ach F block rcivd with forcd rrors would caus th laus 49 PS to loos block_lock, it may b dsirabl to st =0 for som implmntations, howvr th procss followd in this condition is not clarly dfind. Furthr confusion ariss from lin 37 pag 187, which indicats sh=11 is usd to indicat rrors. 1. ithr: A) Updat th flow diagram of Figur to show bhavior whn =0 by adding a dcision block "" =0?"" btwn F Dcodr and Dcoding succssful to show configuration chck. Add 'ys' transition to th lft-hand sid ""x1=0?"" dcision block, and add 'no' transition to continu to th ""Dcoding succssful?"" block lin 50 pag 191 indicats ""h dfault valu for variabls m=8 and n=4."" h trm 'dfault valu' implis managmnt control, but thr is no dfind managmnt mthod for altring th valu of m or n, or rading thir currnt valus. Dlt th word 'dfault' from th final sntanc, and rwrit as follows: ""h valu for variabls ""m"" and ""n"" ar as follows, m=8 and n=4."" Altrnativly, dfin a managmnt mthod for rading th currnt valus of ths variabls and writing nw valus. AP IN PRINIPL. Modify txt as follow: Status Dlt th word 'dfault' from th final sntnc, and rwrit as follows: ""h valus for ""m"" and ""n"" ar as follows, m=8 and n=4."" Or B) Add clarifing languag to to dfin bhavior whn =0. 2) orrct txt in lin 37 pag 187 from ""to th sam valu (11),"" to th corrctd txt ""to th sam valu (11 or 00),"" AP IN PRINIPL. Status Add clarifing languag to to dfin bhavior whn =0. YP: R/tchnical rquird R/ditorial rquird GR/gnral rquird /tchnical /ditorial G/gnral OMMN SAUS: D/dispatchd A/accptd R/rjctd RSPONS SAUS: O/opn /writtn /closd U/unsatisfid Z/withdrawn SOR ORDR: ommnt ID ommnt ID # 68 Pag 18 of 54

19 I P802.3ap D2.2 BackPlan ommnts l 74 S P 191 L 40 # 69 l 73 S P 146 L 13 # 70 Nosworthy, Bob UNH-IOL Nosworthy, Bob UNH-IOL lin 5 pag 187 indicats that ""Bfor ach F block procssing (ncoding or dcoding) th PN-2112 gnrator is initializd"" hil for propr opration this rquirmnt should b clar, figurs 74-9(74-8) indicat that th PN-2112 Gnrator is xor'd prior to F Block Sync. "" F block synchronization"" and ""Figur F(2112,2080) block sync and dcoding"" mak no mntion of th rquirmnts of At a minimum, modify th procdur of as follows, insrting a nw itm #1 blow: ""a) st a candidat block start position 1) Dscrambl block using PN-2112 Gnrator pr ) valuat parity for th potntial block i) If it fails shift candidat by on bit position and try again"" Also, optionally strik th PN-2112 Gnrator block and XOR junction from Figur 74-9(74-8). AP IN PRINIPL. Add txt to as proposd. Status Do not rmov th XOR junction and PN-2112 gnrator as suggstd, instad rfr rspons to commnts #109,148 rgarding modifications to Figur 74-6, 74-9 that illustrat th initialization at ach F block start How dos paralll dtction work in th prsnc of 10GBAS-R F? As F rsids blow th XSBI of a 10GBAS-KR intrfac, thn nothing in claus 73 xplicitly forbids paralll dtction of a 10GBAS-KR intrfac with F, as a systm capabl of paralll dtcting an F ncodd KR link would ultimatly rport sync_status_kr=rady. Howvr, it is not clar that th dtctd link would b indicatd to managmnt as F nabld. Modify th last sntanc of th scond to last paragraph in which currntly nds: ""th paralll dtction function shall st th bit in th link partnr ability rgistrs (S ) corrsponding to th tchnology dtctd by th paralll dtction function."" his statmnt rfrs to a singl bit, if F paralll dtction is to b supportd, modify th statmnt to rad: ""th paralll dtction function shall st th bit in th link partnr ability rgistrs (S ) corrsponding to th tchnology dtctd by th paralll dtction function, including th prsnc of F via th F0 bit"" AP IN PRINIPL. Status Add not to indicat that if you paralll dtct for 10GBAS-KR, thn it is up to th implmntation to dtrmin whthr to turn on F. YP: R/tchnical rquird R/ditorial rquird GR/gnral rquird /tchnical /ditorial G/gnral OMMN SAUS: D/dispatchd A/accptd R/rjctd RSPONS SAUS: O/opn /writtn /closd U/unsatisfid Z/withdrawn SOR ORDR: ommnt ID ommnt ID # 70 Pag 19 of 54

20 I P802.3ap D2.2 BackPlan ommnts l 69B S 69B.2 P 207 L 25 # 71 l 72 S P 120 L 22 # 74 Ablr, Jo IBM Ablr, Jo IBM radr should b advisd that th rcommndations for channl attributs hav bn individually dfind and that concurrntly taking all attributs to thir absolut limits may b byond th bounds of th standard. rcommndd txt will b providd at Fb mting. RJ. h group was unabl to rach consnsus on making th chang. Straw Poll # 4 Option #1 - Mak th chang Option #2 - Don't mak th chang ommnt Status R Status transition tim in abl 72-9 nds to b updatd chang to 24-47ps Status AP. S 38 l 72 S P 124 L 46 Ablr, Jo IBM nd to updat tx jittr dscription to includ a DD componnt. # 75 Option 1-4 Option 2-8 updat txt to spcify a max DD of 0.05UIpp, which is a componnt of total DJ. l 72 S P 110 L 49 Ablr, Jo IBM # 72 AP. Status main tap is no longr rfrrd to as gain tap. dlt "",or gain,"" Status l 44 S P 19 L 49 Avago chnologis # 76 In this draft th optional F sublayr is not tid to Backplan thrnt any mor: it's an indpndnt option. AP. l 45 S P 28 L 25 Ablr, Jo IBM # 73 Add anothr paragraph: 'An optional F sublayr is dfind in laus 74.' AP. Status Nd to add nw bit filds, prst and initializ, to abl updat tabl to includ bits 13 & 12. add corrsponding subclaus dfining prst and initializ bits. carry through to claus AP. Status YP: R/tchnical rquird R/ditorial rquird GR/gnral rquird /tchnical /ditorial G/gnral OMMN SAUS: D/dispatchd A/accptd R/rjctd RSPONS SAUS: O/opn /writtn /closd U/unsatisfid Z/withdrawn SOR ORDR: ommnt ID ommnt ID # 76 Pag 20 of 54

21 I P802.3ap D2.2 BackPlan ommnts l 45 S P 22 L 54 # 77 l 45 S P 24 L 27 # 79 Avago chnologis Avago chnologis If adding this sntnc, th lack of an quivalnt for 10G bcoms apparnt. Insrt anothr sntnc: 'hn st to 0000, bits 5:2 slct th us of a 10G PMA/PMD. Mor spcific slction is prformd using th 10G PMA/PMD control 2 rgistr (Rgistr 1.7) (s ) and th laus 74 F control rgistr (Rgistr 1.171) (s ). AP. Status l 45 S P 23 L 46 # 78 Bcaus 10GBAS-KR is a srial PMD, th nw sntnc 'h dscription of th transmit fault function for th 10GBAS-KR PMD is givn in ' contradicts 802.3's 'h dscription of th transmit fault function for srial PMDs is givn in '. But not th lattr is bing modifid by P802.3aq. Anothr rmdy would b to kp th nw sntnc for 10GBAS-KR and chang 'othr srial' to 'othr optical srial'. hang: h dscription of th transmit fault function for 10GBAS-LRM srial PMDs is givn in , and for othr srial PMDs in ' to: h dscription of th transmit fault function for th 10GBAS-KR PMD is givn in , for 10GBAS-LRM srial PMDs in , and for othr srial PMDs in and dlt th sntnc: h dscription of th transmit fault function for th 10GBAS-KR PMD is givn in Status Avago chnologis AP IN PRINIPL. Adopt th proposd rmdy and add ditors not to say it is an 802.3aq chang Bcaus 10GBAS-KR is a srial PMD, th nw sntnc 'h dscription of th rciv fault function for th 10GBAS-KR PMD is givn in ' contradicts 802.3's 'h dscription of th rciv fault function for srial PMDs is givn in '. But not th lattr is bing modifid by P802.3aq. Anothr rmdy would b to kp th nw sntnc for 10GBAS-KR and chang 'othr srial' to 'othr optical srial'. hang: h dscription of th rciv fault function for 10GBAS-LRM srial PMDs is givn in , and for othr srial PMDs in ' to: h dscription of th rciv fault function for th 10GBAS-KR PMD is givn in , for 10GBAS-LRM srial PMDs in , and for othr srial PMDs in and dlt th sntnc: h dscription of th rciv fault function for th 10GBAS-KR PMD is givn in Status AP IN PRINIPL. Adopt th proposd rmdy and add ditors not to say it is an 802.3aq chang. l 45 S P 37 L 5 # 80 As for laus 74 F corrctd blocks countr, ths ar non-rollovr rgistrs. Add sntnc as in : 'hs bits shall b hld at all ons in th cas of ovrflow.' Add appropriat PIS. AP. Status Avago chnologis YP: R/tchnical rquird R/ditorial rquird GR/gnral rquird /tchnical /ditorial G/gnral OMMN SAUS: D/dispatchd A/accptd R/rjctd RSPONS SAUS: O/opn /writtn /closd U/unsatisfid Z/withdrawn SOR ORDR: ommnt ID ommnt ID # 80 Pag 21 of 54

22 I P802.3ap D2.2 BackPlan ommnts l 45 S P 40 L 34 # 81 l 45 S P 47 L 39 # 84 Avago chnologis Avago chnologis Nam of bit not consistnt. Somtims 'abl', somtims 'ability'. Also, somtims 'LP' coms bfor 'Auto-Ngotiation', somtims aftr. Using 'ability' brings it in lin with th majority of laus 45, lin 36 and 'Auto-Ngotiation ability'. hang 'LP AN abl' to 'AN LP ability' hr, chang 'LP Auto-Ngotiation abl' to 'AN LP ability' in abl AP. Status 'th PHY typ': which PHY typ? hang to 'If at last on Backplan thrnt PHY typ (1000BAS-KX, 10GBAS-KX4 or 10GBAS-KR)' is implmntd...'. AP IN PRINIPL. S rspons to commnt 105 Status l 45 S P 43 L 39 Avago chnologis # 82 l 74 S 74.1 P 174 L 13 Avago chnologis # 85 nst ach subclaus usually has a layr diagram. nxt Status AP. l 45 S P 47 L 35 Avago chnologis on of thr... is st (not ar st) Status AP. # 83 Insrt diagram 'Figur GBAS-R F rlationship to th ISO/I Opn Systms Intrconnction (OSI) rfrnc modl and th I SMA/D LAN modl' opy and modify fig opy and modify sntnc to introduc figur from on of claus 46 upwards (wording varis slightly). AP IN PRINIPL. Status Insrt diagram as proposd and provid th following introductory txt: 74.3 Rlationship to othr sublayrs Figur 74-1 dpicts th rlationships among th 10GBAS-R F (shown shadd), th 10 Gb/s MA and Rconciliation Sublayrs, th 10GBAS-R PS and PMA, th ISO/I LL, and th ISO/I Opn Systm Intrconnction (OSI) rfrnc modl. YP: R/tchnical rquird R/ditorial rquird GR/gnral rquird /tchnical /ditorial G/gnral OMMN SAUS: D/dispatchd A/accptd R/rjctd RSPONS SAUS: O/opn /writtn /closd U/unsatisfid Z/withdrawn SOR ORDR: ommnt ID ommnt ID # 85 Pag 22 of 54

23 I P802.3ap D2.2 BackPlan ommnts l 74 S P 177 L 27 # 86 l 74 S 74 P 174 L 1 # 89 Avago chnologis Avago chnologis Nam of on functional block dos not match hang DOD to DODR or dcodr. Status AP IN PRINIPL. o b consistnt also chang F NOD to F NODR Also rfr to commnt #107 l 74 S P 177 L 8 Avago chnologis Do not s why som labls ar all-capitals. hang to uppr/lowr cas as in body txt or tabl. Status AP IN PRINIPL. # 87 waking th claus titl to b mor lik othrs.g. 51, 52, and to gt th sublayr hav 'F' in its nam. Rmoving stray capitals. Do not know xactly what 'Physical Layr signaling systms' is mant to man. Shortning. Prfrably, chang to: 74. Forward rror orrction (F) sublayr for 10GBAS-R or, to: 74. Forward rror orrction (F) sublayr for 10GBAS-R PHYs or, to: 74. Forward rror orrction (F) sublayr for 10GBAS-R Physical Layrs or, to: 74. Forward rror orrction (F) sublayr for 10GBAS-R Physical Layr signaling systms AP IN PRINIPL. hang laus 74 titl to: Status Forward rror orrction (F) sublayr for 10GBAS-R PHYs l 74 S P 187 L 5 Avago chnologis # 90 o b consistnt, mak similar changs to Figur 74-8 which has labls with all-capital lttrs. 'v gon 180 pags without using this unncssary notation; it's a pity to giv up now. l 74 S P 177 L 32 Avago chnologis Somthing whit on top of PMA_SIGNAL.indication? St background of rx_data-group<15:0> to transparnt? Status AP IN PRINIPL. Mov th whit on top of PMA_Signal.indication # 88 Plas chang '0x2AAAAAAAAAAA2F8' to 'hxadcimal 2AAAAAAAAAAA2F8' (as in ). AP IN PRINIPL. Status Ovr takn by vnts. Rfr rspons to commnt #127. YP: R/tchnical rquird R/ditorial rquird GR/gnral rquird /tchnical /ditorial G/gnral OMMN SAUS: D/dispatchd A/accptd R/rjctd RSPONS SAUS: O/opn /writtn /closd U/unsatisfid Z/withdrawn SOR ORDR: ommnt ID ommnt ID # 90 Pag 23 of 54