#milkymist IRC log for Monday, 2013-12-09

--- Mon Dec 9 201300:00
larschm, does somebody know if there is a possibility to tell par to be a bit more conservative with placement regarding to timing09:29
larscI often have the problem that the post routing timing check fails with a couple of picoseconds off. And it is more or less random when this happens and where this happens09:29
ysionneauanyone has any idea about my issue with PSW.dtlbe being cleared somehow?10:18
ysionneauit seems that running a bit of code leads to PSW.dtlbe being cleared10:20
ysionneauI thought it was caused by some exception code, but I don't see where since dtlb miss only reads PSW and never writes to it10:20
ysionneauso exception followed by eret should do the usual edtlbe <= dtlbe; dtlbe <= 0; and upon eret : dtlbe <= edtlbe;10:21
larscyeay, just got my mixxeo board12:26
larscit's smaller than it looks in the brochures though ;)12:26
ysionneaularsc: ask for a refund! :p13:21
ysionneauyou've been lied to13:21
ysionneauI think since lekernel started to play with MEMS things he likes to do small things ;)13:21
larscbut it is right on time, friday is the last day of work before a month vacation13:31
ysionneauah :) enjoy vacation !13:32
ysionneaugoing somewhere? or just enjoying sitting idly in front of some screen @home?13:32
larscfirst week will be my week off with nothing to do at all13:33
larscand then the usual christmas roundtrip13:33
larscvisiting family, friends, etc13:34
larscbought my four day 30c3 ticket yesterday, so I also have no excuse no to go there13:35
ysionneauhéhé :)13:35
ysionneauit's cool that now you can buy tickets normally without destroying your mouse by overclicking13:36
wpwrakbecause of larger space ?13:40
ysionneausoon it's going to be the same for EHSM :)13:44
Action: ysionneau is ordering a new mouse13:44
ysionneauwpwrak: you should come by !13:45
wpwrakEHSM ? yeah, would be tempting13:52
ysionneaucome on, to visit the particule accelerator13:54
ysionneauor show us in real life some anelok prototypes!13:54
ysionneauyou could probably even do a presentation at ehsm about it13:55
wpwrakphew. if i'm still at the prototype level in june, then i'm already well past declaring defeat :)13:56
ysionneaueheh sure13:58
ysionneauanyway, you could present your work, (investigation, how you chosed components, the criteria etc)13:58
ysionneaupcb process, case process13:58
ysionneauI'm sure it would be as interesting as those TLDR (as Too Long Do Read) emails of yours :)13:59
larschm, just wondered what WSH3 is...14:05
ysionneaudunno, what is it?14:06
larscbut turns out I was just holding the card upside down14:06
ysionneauahah :)14:10
wpwrakvery easy to get wrong ;-))14:13
Action: ysionneau has a big EHSM poster in his appartment14:13
ysionneauvery nice on a white wall14:14
ysionneauwith a bullet passing through a graphic card's PCB :)14:14
wpwrakysionneau: (tl;dr) bah. and i try to lighten them up with pictures. maybe i should get a kitten. then the internet will be all over the project :)14:14
ysionneauyour mails are very good!14:14
wpwrak"can i haz anelok ?"14:14
ysionneauthe pictures are a very good idea14:14
larsconly if the kitten solders the board14:15
ysionneauI wouldn't be against 1 or 2 lolcat though :D14:15
ysionneauif you incist ;)14:15
lekernelysionneau, you should get a 2014 poster, with a custom-made X-ray photograph of a hand soldering a board :)14:15
ysionneauoh, I just noticed it's a 2014 edition picture14:16
ysionneaupicture looks really good :)14:16
wpwraklarsc: hmm, maybe a baby dragon then. they should be pretty good at soldering :)14:20
larscdo I understand this correctly? when I have a OSERDES should fCLK = DATA_WIDTH x fCLKDIV15:18
lekerneland you have to use PLL + BUFPLL to generate those, and deal with a bunch of bugs, restrictions and various annoyances15:21
wpwrakthe curse of success: eventually, people will want to run your stuff under windows ...15:22
lekernelthe behaviour of those is very confusing at times. for example, the BUFPLL will never lock unless its *output* clock is properly connected to a ISERDES/OSERDES ...15:23
larschm, ok thanks15:29
lekernelhttps://events.ccc.de/congress/2013/Fahrplan/events/5417.html "Extracting keys from FPGAs, OTP Tokens and Door Locks"17:54
larschm, I get the feeling that my laptop is not able to provide enough power on the usbport for the mixxeo board19:46
rjolarsc: some usb hubs do in fact enforce the negotiated current limits. apple keyboard for example. others don't20:23
davidc__Some enforce at 500ma; and some just have anti-short protection (usually a PTC)20:24
davidc__.... and then the really cheap ones are limited by the maximum output of their power supply20:25
larscthe board just keeps reseting itself during the boot sequence20:36
lekernellarsc, try reflashing the ftdi-chip with the maximum current limit20:42
lekerneland good luck with the libftdi shitware... I'd recommend trying the windows tool20:42
lekernelboards were flashed with "MaxPower 0mA". doesn't cause problems on any of my computers.20:43
davidc__yeah, don't touch libftdi with a million foot pole20:43
davidc__Last time I had to do MFG-time serialization and programming of FTDI parts, I ended up rewriting libftdi in python since it wouldn't reliably program the EEPROM20:44
davidc__(nor did it even use the right format!)20:44
davidc__I might be able to dredge up + open-source that code if there's demand for it20:44
lekernelyeah, the quantity of bugs per line of code is mindboggling. I'd even bet it's higher than ISE's.20:45
lekernelif all else fails, this sort of thing can also help: http://www.streakwave.com/Itemdesc.asp?ic=5VUSB20:51
larscI think I'll try a hub first20:53
larscit worked when I booted the board for the first time, I got a bios promt and everything20:53
larscbut then at somepoint it started resetting itself20:53
larscnow after some cooldown time it is working again20:54
mumptaii can only recommand those 7-port hubs with 4A PSUs21:24
rjoanyone: any pointers to "the best USB hub"? 8-ports, supporting both self-powered and upstream-powered operation, stable, cascadable to the USB spec max, etc23:18
rjoor 1-to-6 ports.23:18
--- Tue Dec 10 201300:00

Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!