Tuesday, 2015-12-29

*** tpb has joined #timvideos00:00
*** Bertl_zZ is now known as Bertl00:30
mithroCarlFK: https://docs.google.com/a/mithis.com/forms/d/18JPnOeavLZDBsA_WOXru34bbcpXRuF7qpfwdQnQdFOM/viewform00:36
tpbTitle: Sign in - Google Accounts (at docs.google.com)00:36
CarlFKmithro: Your response has been recorded.  thanks for poking me01:15
*** rohitksingh has quit IRC02:25
*** rohitksingh has joined #timvideos02:40
*** rohitksingh has quit IRC02:46
*** miselin2 is now known as miselin03:33
*** miselin has joined #timvideos03:33
*** mithro-work has joined #timvideos03:55
*** mithro-work has quit IRC03:55
*** Bertl is now known as Bertl_zZ06:10
mithroysionneau: you wouldn't happen to be around?07:36
ysionneauI'm around!09:41
mithroysionneau: so it turns out that EEPROM part we ended up getting ignores the address pins10:25
mithroysionneau: And responds to all addresses in the EEPROM range :(10:25
ysionneauouch :(10:26
mithrothe FX2 also detects people fiddlying with it's master (IE a master collision) and then stops loading10:26
mithrothe best solution I've come up with so far is to create a very minimal "second stage bootloader" and put that in the EEPROM10:27
ysionneauyep I guess it's your best move :/10:27
ysionneauit always boots from i2c/real-eeprom, but then downloads another firmware via another protocol talking to the fpga10:28
mithroysionneau: so it seems like the smallest bootloader I've been able to make still uses the I2C interface10:28
mithroI can fit one in ~50 bytes10:29
ysionneaubut if the bootloader uses i2c to retrieve the fw, it will still get answers from the real eeprom, right? isn't that an issue?10:30
mithroysionneau: I send a non-EEPROM addres10:31
ysionneauah so the eeprom ignores the address pins but still only answers to some eeprom related slave addresses10:32
mithroBasically the last 3 bits of the address are in "don't care mode"10:37
mithroysionneau: what I wanted to ask you is, why is the fx2 firmware segment in firmware.h 0x3FB8 big?10:39
*** se6astian|away is now known as se6astian11:20
*** rohitksingh has joined #timvideos11:40
*** Bertl_zZ is now known as Bertl11:46
*** se6astian is now known as se6astian|away11:51
*** Bertl is now known as Bertl_oO12:57
MaZderMindfyi: the first releases produced with voctomix on the 32C3 are online: https://www.youtube.com/playlist?list=PLIoqMTM7qDWouZPzI-JZP5uOyZgWfIzDK14:49
MaZderMindLivestream is running here: https://streaming.media.ccc.de/32c3/sendezentrum/14:49
tpbTitle: Sendezentrum FullHD Video – 32C3 Streaming (at streaming.media.ccc.de)14:49
*** Bertl_oO is now known as Bertl15:21
*** se6astian|away is now known as se6astian15:58
*** Bertl is now known as Bertl_oO19:40
*** CarlFK has quit IRC20:15
*** rohitksingh has quit IRC20:55
*** se6astian is now known as se6astian|away22:04
*** CarlFK has joined #timvideos22:07
*** ChanServ sets mode: +v CarlFK22:07
*** nueces has joined #timvideos22:12
nueceshi all, I see that a few pages that are included in a iframe like this https://code.timvideos.us/ideas/ don't work correctly22:19
mithronueces: yeah - I tried to enable https:// and it broke a bunch of things22:19
nuecesI get a 301 (redirect) but the iframe is not loaded, I test that on chrome and ffox22:19
mithronueces: and I haven't gotten around to fixing it yet22:20
nuecesmithro, that could be because the iframe is loading a few imgs from http instead off https like images in the issue 20722:31
nuecesthen the browser refuse to load the whole content22:31
*** nueces has quit IRC22:41
ysionneaumithro: what do you mean about 0x3fb8 ? what's this number about?23:14
ysionneauif the question is "why is thhe firmware so big?" it's because I'm using a very naive approach to convert the fx2 fw into the eeprom content23:16
ysionneauin the eeprom you can store "regions" with a start address and a size, allowing you to leave areas untouched23:16
ysionneauinstead of that, I'm dumping the entire firmware in .bin and then I'm only using one big fat region which start at the beginning of the fw and has the size of the fw23:17
ysionneauone way to have a smaller eeprom content would be to parse the intel hex format and use several smaller regions, to skip big unused spaces23:17
ysionneaudid you get the i2c stuff to work then? with your own bootloader protocole ?23:19
ysionneauwere you able to load an fx2 fw?23:19
mithroI ran into the problem that some of my firmware has data in the scratch region - 0xE000 which made the EEPROM 64k big23:19
mithroBut I have now created C+Python code which is able to understand the fx2 config format23:20
ysionneauin my tests, I think the fw was almost 64k big also IIRC23:25
ysionneauah now I see the 3Fb8 indeed in the fx2fw array, yes it's big and could be optimized23:31

Generated by irclog2html.py 2.13.1 by Marius Gedminas - find it at mg.pov.lt!