<aeris> lekernel, ben, si je prends le Pegasus chez moi, le cable servira, mais c'est comme tu veux...
<aeris> (j'ai emnag, j'ai internet et une voiture dans quelque jours)
<aeris> Sinon je lui envoie et il me le renvoie...
<adamw_> lekernel, I compared the original jtag/serial pod's sch and bom, yes I found some values are not the same.
<adamw_> 1), C14 was 10uF, now is 2.2uF  2), 480pf --> 470pF 3), C1/C3 value was 27pF, now is 33pF
<adamw_> all of those three differences I think it should not be the reasons.
<adamw_> the only one thing is the history of removing 'C2/C4/R3/R4' components in D-/D+ line loop.
<adamw_> So could you help me check the older pod, did the R3 and R4 (22 ohms) were still on board?
<adamw_> or just 'short' already?
<lekernel> on the other pod I shorted R3+R4
<lekernel> so really it has almost the same schematics
<lekernel> maybe some parts were swapped during production?
<lekernel> SMT?
<adamw_> i measured all resistors, their values are all ok.
<adamw_> but for capacitors, I quite not easy to measure them on board all correctly.
<adamw_> so i'm now checking both old and newest gerber now.
<adamw_> the schematic I compared both old and newest, they must be correct excepts I mentioned the values above.
<wolfspraul> adamw_: you have to be careful to consider anything correct
<wolfspraul> there have been too many changes and too many manual interventions
<wolfspraul> if we cannot track it down now, I see 3 options
<wolfspraul> 1) sebastien who has both old and new boards finds the difference
<wolfspraul> 2) adam sends a new board to yanjun luo who finds the difference
<wolfspraul> 3) sebastien sends an old board to adam who finds the differenc
<wolfspraul> :-)
<wolfspraul> I am a little worried that #2 is too slow, but of course Adam can just send a board and we see when/what yanjun luo finds.
<adamw_> yes, schematic I checked ...checked *.lst( bom ) compared..and manually measured resistors...so I need to compare gerbers now
<wolfspraul> for #1 the problem is whether sebastien has enough time or equipment to track it down
<adamw_> yeah...too slow
<wolfspraul> #3 will also be slow, because first sebastien has to send the board to Taipei
<wolfspraul> I can't do anything :-) oh well...
<lekernel> i'm traveling with little equipment atm so everything will be slow...
<wolfspraul> maybe you should drop an old one into an airmail letter to Taipei...
<lekernel> i'd ideally need the 2 JTAG pods for a workshop on feb 4th...
<lekernel> i'll try to get you another old board
<adamw_> wolfspraul, can we send more new pods to lekernel ? meanwhile lekernel send me an old pod?
<lekernel> i'll be moving all around until feb 4, so it's going to be complicated. i'll try to get you aeris's old pod
<mwalle> hi
<larsc> mwalle: i send out most of the patches i have in my tree. the scheduling could maybe help with the problem you were seeing yesterday
<mwalle> larsc: thx
<mwalle> larsc: i think i found the problem
<mwalle> what was the problem with the serial uart?
<mwalle> you had
<larsc> none
<mwalle> the problem where the output suddenly wasnt working anymore
<larsc> that was because of the messed up lm32 irq handling
<lekernel> I got invited to this: http://renpar.irisa.fr/satellite_eng.htm ... mwalle / larsc: do you want to go (I don't think I will)?
<aeris> I can mail the cable only this weekend ='(
<aeris> wait a car
<mwalle> larsc:
<mwalle> /home/mw/repo/milkymist-linux/arch/lm32/kernel/time.c:96:16: Fehler: IRQ_SYSTMR undeclared (first use in this function)
<mwalle> /home/mw/repo/milkymist-linux/arch/lm32/kernel/time.c:96:16: Anmerkung: each undeclared identifier is reported only once for each function it appears in
<mwalle> /home/mw/repo/milkymist-linux/arch/lm32/kernel/time.c:99:2: Fehler: Implizite Deklaration der Funktion lm32_irq_unmask
<mwalle> larsc: did you forget to add the this file to the patch?
<larsc> it replaced in another patch
<mwalle> lekernel: no thanks, its the weekend before my birthday
<larsc> mwalle: "lm32: milkymist: Add clocksource and clockevents support"
<mwalle> larsc: ok, btw the first line within a git commit should not exceed 50 chars
<mwalle> yey! :)
<mwalle> executed code without theobromas 'fdpic' hack
<mwalle> compiled with own toolchain and uclibc
<larsc> party!
<mwalle> gn8
<lekernel> larsc: great work with the patches! :)
<kristianpaul> indeed
<larsc> 3thanks
<larsc> but threre is still lots to do. my estimation is that we'll get it merged upstream around august