Back
[02:33:39] <OndraSterver> !seen rikusw
[02:33:39] <tobbor> RikusW was last seen in #avr on Sep 01 10:43 2012
[02:33:43] <OndraSterver> alright
[02:57:37] <Richard_Cavell> How's everyone doing today in AVR land?
[03:11:07] <tomatto> how can i do USART in attiny45 and only RX?
[04:07:58] <tomatto> in pwm mode timer1's OCR1B doesn't work with OCR1B ?
[04:17:30] * Richard_Cavell wishes everyone a good evening
[04:21:21] <mitsakos> anybode used ST7540?
[06:48:06] <karlp> has anyone gotten avarice + dragon to work with newer versions of gdb?
[06:48:20] <karlp> I always get a thread assertion failed, and it's game over from there
[06:49:30] <karlp> clearly there's no thread info to be had, and I'm presuming that gdb isn't handling the lack of thread info replies from avarice's gdb server?
[06:56:26] <megal0maniac> karlp: which version of gdb?
[06:56:34] <karlp> 7.3 and 7.5
[06:56:41] <karlp> with avarice 2.12
[06:56:48] <karlp> 2.10 works fine with gdb 6.4
[06:58:10] <megal0maniac> Hmmm... Mine's pretty outdated. Avarice 2.9 with gdb 6.8
[06:59:30] <karlp> newer gdb than my working one.
[06:59:53] <karlp> I was really just trying ot use a newer avr-gcc, to see if the problems I'm seeing are just an old compiler bug.
[07:00:32] <megal0maniac> Yeah, but I haven't tried it yet :)
[07:00:40] <megal0maniac> Haven't taken the time to figure out how
[07:00:57] <karlp> might go back to the machine with the newer avr-gcc/binutils, and try and rebuild an older gdb for it. see how that goes.
[07:01:02] <megal0maniac> Thought I might give it a go if I had a newer version
[07:01:28] <megal0maniac> Although gut feeling says gdb is to blame
[07:01:34] <karlp> I'm sure there should be a way to tell gdb to not try and find out thread info, or to make avarice understand it, but I find the gdb remote protocol rather hard to follow
[07:02:11] <karlp> my gut is that gdb is making an assumption about lack of reply to various status commands that is appropriate for desktop systems, with OS,s but is not appropriate for avr
[07:12:31] <megal0maniac> What's weird is that it worked in older versions, so there must have been a reason that is (whatever "it" is) was changed
[07:12:42] <megal0maniac> *it
[07:16:19] <karlp> yeah, new "features" in gdb and bad assumptions about the support/lack thereof in older gdbservers
[07:16:28] <karlp> seen that with arm gdb too :|
[07:16:55] <karlp> "hat? your gdbserver doesn't reply to this command we just made up? that's it, no fallback, no use the old wya, we're just not going to run anymore"
[07:24:42] <megal0maniac> I suppose there doesn't really have to be a reason for changes. Thing is, gdb is so widely used, that it seems almost impossible to change one thing without breaking another
[07:36:40] <megal0maniac> Where do I start if I want to use a jtagice mkI clone?
[07:37:23] <megal0maniac> Specifically, with a device which isn't officially supported by AS4.18
[07:37:31] <megal0maniac> (ATmega1284P)
[07:37:35] <megal0maniac> @karlp
[07:38:17] <karlp> for the dragon, on the setup that works, "[sudo] avarice -g -w :4242" works,
[07:38:31] <karlp> then avr-gdb blah.elf; target extended-remote :4242; run....
[07:39:10] <megal0maniac> I'm using RikusW's board
[07:39:17] <megal0maniac> https://sites.google.com/site/megau2s/home/jtag
[07:39:50] <karlp> bleh, trying to build gdb 6.8 fails with newer gcc :|
[07:39:52] <megal0maniac> Connected JTAG lines, VCC & GND. Fuses are set to use internal osc.
[07:40:11] <karlp> if rikus doesn't have documentation on how to actually _use_ his board, perhaps you should get him to add some
[07:40:29] <megal0maniac> He does, but it seems to be aimed at the informed. Or Linux users :)
[07:41:03] <megal0maniac> "will need custom PC side software"
[07:41:15] <megal0maniac> However, avarice should theoretically work
[07:55:09] <karlp> ok, yay, rebuilt avr-gdb 6.8, it still works with avarice 2.12.
[07:55:22] <karlp> now, back to 4 days ago, and the actual problem at hand :|
[08:00:41] <megal0maniac> JTAG ICE communication failed: Inappropriate ioctl for device
[08:00:45] <megal0maniac> Meh
[08:01:42] <OndraSterver> !seen rikusw
[08:01:43] <tobbor> RikusW was last seen in #avr on Sep 01 10:43 2012
[08:02:20] <megal0maniac> avarice --program --file BlinkLED2.elf --part atmega1284p --jtag COM3 :4242
[08:02:24] <megal0maniac> Gives me that error
[08:03:48] <OndraSterver> challenge accepted: write USB stack for xmega in asm!
[08:04:05] <karlp> why would i tbe COM3? isn't it a usb device?
[08:04:12] <OndraSterver> CDC?
[08:04:54] <megal0maniac> Yeah, cdc
[08:05:14] <OndraSterver> CapnKernel, eya, you around?
[08:05:15] <megal0maniac> JTAG ICE MKI clone
[08:05:22] <OndraSterver> that would use virtual serial port
[08:05:39] <megal0maniac> Correct. But why won't jtag work?
[08:05:47] <OndraSterver> hard to say :)
[08:05:49] <megal0maniac> What does that error even mean?
[08:05:56] <OndraSterver> what chip are you trying to debug with it?
[08:06:11] <megal0maniac> mega1284P
[08:06:33] <OndraSterver> hmm are you sure it will work? Original jtagice didn't support it... who knows how much it is (un)supported on clones :)
[08:07:31] <megal0maniac> Hmmm... Good point. I know the firmware will work with it, but perhaps that's why there's a problem client-side
[08:08:20] <megal0maniac> Tried substituting "atmega8" in the command string just to see, but same error
[08:08:38] <OndraSterver> hmm no idea
[08:11:31] <megal0maniac> Is there any real advantage in using jtag to program?
[08:12:22] <megal0maniac> Pin count is 7, which is the same is ISP if you're providing vtg
[08:13:56] <OndraSterver> programming? not really
[08:13:59] <OndraSterver> but you can quickly switch to debug :)
[08:14:51] <megal0maniac> Only if you have software which supports it...
[08:14:57] <OndraSterver> yep
[08:15:00] <OndraSterver> software and hardware :P
[08:16:55] <megal0maniac> Well, yeah. But there are only about 8 devices which officially support jtag debugging with the MKI. Further than that, you need to jump through hoops
[08:17:30] * megal0maniac watches mega1284p blink an LED while waiting for g++ to download
[08:17:42] <OndraSterver> yes, the original jtagice is now pretty much useless
[08:19:43] <OndraSterver> afk
[08:51:05] <Tom_itx> megal0maniac, no advantage for programming
[08:52:24] <megal0maniac> Tom_itx: Didn't think so. debugwire is the way, it seems
[08:58:52] <_raven> hi
[08:59:44] <OndraSterver> hi
[09:03:56] <megal0maniac> hi
[09:09:09] <megal0maniac> SO MANY WIRES!
[09:09:19] <megal0maniac> (testing HVPP)
[09:14:32] <specing> megal0maniac: go with HVSP then
[09:16:11] <megal0maniac> specing: It's only really for testing. But do larger devices support HVSP? I thought it was either one or the other
[09:19:37] <OndraSterver> only small ones know HVSP
[09:19:57] <_raven> i am trying to use an avrispmkII programmer with avrdude 5.11.1 but i am unable to use it. after a firmware-upgrade i am at least able to use it via windows avrstudio 4. i did everything what is important with udev rules but it doesnt work
[09:20:11] <CapnKernel> OndraSterver: Yes.
[09:20:22] <CapnKernel> But going to sleep soon.
[09:20:29] <OndraSterver> CapnKernel, you said that 5x5cm board would be the same as 2.5x5cm board
[09:20:34] <CapnKernel> Yes
[09:20:37] <OndraSterver> what about 5x5 board with huge slot in the middle? :P
[09:20:44] <OndraSterver> not splitting the boards totally
[09:20:44] <CapnKernel> And 10 for the price of 5
[09:21:18] <CapnKernel> It depends on whether it appears you're trying to get two boards for the price of one.
[09:21:36] <OndraSterver> I could connect one wire across it...
[09:21:36] <CapnKernel> If you have a slot that appears to have a mechanical function, that's fine. You can have whatever slots and curves you want.
[09:22:01] <megal0maniac> OndraSterver: That's what I thought. Pity
[09:22:05] <CapnKernel> If it appears you're trying to get two boards out of the space, whether they be the same or different, then there may be a problem with that.
[09:22:26] <OndraSterver> well, it would ease up also pre-programming and building them - I can do two at once... That is mechanical function, isn't it? :D
[09:22:30] <megal0maniac> _raven: What command line are you using with avrdude?
[09:22:56] <CapnKernel> If you had a board with a break-away programming header, that would be fine
[09:23:09] <OndraSterver> hmm
[09:23:23] <Tom_itx> cheater
[09:23:44] <_raven_> megal0maniac avrdude -c avrispmkII -p m8 -P usb
[09:23:54] <CapnKernel> If you had two boards that were clearly designed to be used with each other, for example, as a stack, that would be fine too.
[09:24:02] <OndraSterver> hmm
[09:24:14] <OndraSterver> two same boards? or two completely different boards?
[09:24:24] <Tom_itx> CapnKernel i don't see the difference if they're selling board space.
[09:24:31] <OndraSterver> neither do I
[09:24:33] <Tom_itx> i never understoood that
[09:24:33] <CapnKernel> I don't either, but that's what they say.
[09:24:39] <karlp> anyone had any problems with avr-gdb (and avr-addr2line) coming up with bogus line numbers?
[09:24:56] <OndraSterver> I remember somebody somewhere saying that he wrote to iTead and included gerbers and they gave him 20 2.5x5cm boards for the same price as 5x5 boards
[09:25:01] <karlp> it thinks my own code is in things like SPI.h:846, when SPI.h has only 100lines or so
[09:25:02] <CapnKernel> If you take the space and attempt to do your own panelisation, they either reject the order, or charge an extra "panelisation fee"
[09:25:55] <CapnKernel> While it's interesting to know what the other suppliers do, I can't comment on it.
[09:26:02] <CapnKernel> I can only tell you what I can offer.
[09:26:14] <OndraSterver> sure
[09:26:32] <OndraSterver> because all my boards are <25mm wide and thus I am wasting quite a lot of money on the boards actually :/
[09:26:58] <OndraSterver> $0.8 vs $0.4 on $3 device makes quite a difference
[09:27:18] <_raven_> megal0maniac and this gives me something like that in verbose mode
http://pastebin.com/e69e2EL7
[09:28:07] <CapnKernel> I'm off soon...
[09:28:19] <Tom_itx> when are you back to China?
[09:28:20] <megal0maniac> _raven: Firstly, you can leave out "-P usb"
[09:28:35] <OndraSterver> also CapnKernel what about selling reels of LEDs and resistors and capacitors? :)
[09:29:11] <CapnKernel> I arrived back in Melbourne on Friday, and I'll return to Shenzhen in mid-October.
[09:29:22] <_raven_> megal0maniac but there is no ttyS0: avrdude: ser_open(): can't set attributes for device "/dev/ttyS0": Inappropriate ioctl for device
[09:29:23] <CapnKernel> I'm still doing PCBs
[09:29:28] <OndraSterver> only?
[09:29:37] <CapnKernel> While I'm not in China, yes.
[09:29:39] <OndraSterver> hackvana.org not operating so far?
[09:29:40] <OndraSterver> ok
[09:29:51] <CapnKernel> Some people are testing it now.
[09:29:59] <CapnKernel> When I'm in China, I do reels of SMD components.
[09:30:06] <OndraSterver> oh ok
[09:30:07] <OndraSterver> thanks
[09:30:10] <OndraSterver> will ask later then :)
[09:30:10] <megal0maniac> _raven: Unplug and re-plug the programmer, then run "dmesg" and pastebin the last 10 lines or so
[09:30:26] <CapnKernel> :-)
[09:30:28] <CapnKernel> Good night all.
[09:30:31] <OndraSterver> gn
[09:31:02] <megal0maniac> _raven: Also, what exactly are you trying to do with that command string? You haven't specified a .hex files
[09:31:04] <megal0maniac> *file
[09:31:56] <_raven_> megal0maniac
http://pastebin.com/9pw5MDMF
[09:32:12] <_raven_> megal0maniac i am trying to read the state of the device
[09:33:05] <megal0maniac> _raven: Odd. So if you leave out -p, then it defaults to /dev/ttyS0 ?
[09:33:14] <megal0maniac> Sorry, -P
[09:33:16] <_raven_> megal0maniac yes
[09:37:24] <megal0maniac> _raven: One sec, want to copy your setup
[09:37:24] <_raven_> ok
[09:53:22] <_raven_> megal0maniac ?
[09:53:43] <megal0maniac> _raven_: Well, your pc is seeing the programmer as it should. The connection to the target won't be the issue because avrdude isn't getting that far. Maybe try downloading Arduino and trying with that. It has its own copy of avrdude with isolated config
[09:55:44] <_raven_> megal0maniac you mean "arduino ide" for example?
[09:55:59] <megal0maniac> Yes
[10:02:58] <_raven_> megal0maniac how to test connection with it i cannot find any option to change the target from com1 to anything else
[10:03:11] <megal0maniac> Also needs to be run as root
[10:03:22] <megal0maniac> Otherwise you won't get access to ports
[10:04:31] <megal0maniac> Although, you can also cd into the arduino directory and run avrdude from terminal. Makes more sense, since Arduino core doesn't directly support the m8
[10:05:16] <megal0maniac> _raven_ The point being, you now have another copy of avrdude with seperate configs
[10:22:00] <_raven_> megal0maniac where is the right arduino dir? i cannot find it
[10:22:23] <megal0maniac> Did you apt-get or download the tar.gz
[10:22:47] <_raven_> apt-get
[10:23:15] <megal0maniac> Then it might use system-wide config... Sorry, should've been more specific
[10:23:39] <_raven_> what to do now?
[10:25:50] <megal0maniac> Uninstall, then download the tar.gz from
http://arduino.cc/en/Main/Software
[12:01:41] <_raven_> megal0maniac sry i am back
[12:22:04] <OndraSterver> why is Dragon giving me on PDI in log that it doesn't work?
[12:22:11] <OndraSterver> 19:05:48: [ERROR] Got error setting up PDI mode: Device is not supported in this emulator mode. Debugger command setParameter failed., ModuleName: TCF (TCF command: Device:startSession failed.)
[12:22:21] <OndraSterver> afk
[12:22:24] <OndraSterver> then I will search the interwebz
[12:26:55] <_raven_> megal0maniac i compiled avrdude 5.11 myself now but its exactly the same error
[12:37:04] <OndraSterver> http://8515.avrfreaks.net/index.php?name=PNphpBB2&file=viewtopic&t=118076
[12:37:04] <OndraSterver> hmm
[12:59:40] <OndraSterver> yep
[12:59:43] <OndraSterver> clock line is sitting low
[12:59:44] <OndraSterver> like a duck
[12:59:48] <OndraSterver> tried VMeter on it
[12:59:58] <OndraSterver> data line jumps for fraction of a second high
[13:00:01] <OndraSterver> but clock sits low
[13:00:02] <OndraSterver> :(
[13:00:47] <OndraSterver> I can JTAG just fine with my 40cm cable
[13:00:58] <OndraSterver> but PDI CLK is low even with no cable on it :(
[13:02:30] <OndraSterver> guys, USB's ground is earthed, right?
[13:02:37] <OndraSterver> so I can hook up oscope to it
[13:03:05] <OndraSterver> I might just as well grab some laptop instead :D
[13:03:09] <OndraSterver> that one is not earthed and is floating
[13:05:59] <Kevin`> OndraSterver: usb ground would be earthed yes
[13:06:53] <OndraSterver> ko
[13:06:53] <OndraSterver> ok
[13:07:00] <OndraSterver> I don't have to be scared of blowing something up
[13:07:02] <OndraSterver> (like last time lol)
[13:13:20] <OndraSterver> HOLY COW
[13:13:25] <OndraSterver> I connected scope to SCK pin
[13:13:30] <OndraSterver> and it read signature just fine ?!
[13:13:35] <OndraSterver> I have GOT to make video of that
[13:14:47] <OndraSterver> that is 1M/25pF
[13:14:48] <OndraSterver> input
[13:14:49] <OndraSterver> on the scope
[13:14:53] <OndraSterver> let me find 1M resistor lol
[13:15:01] <OndraSterver> http://clip2net.com/s/2gxWK
[13:26:23] <jacekowski> OndraSterver: do you know that i2c requires pullups
[13:26:31] <OndraSterver> yes
[13:26:35] <OndraSterver> who said anything about I2C?
[13:26:44] <jacekowski> SCK
[13:26:48] <OndraSterver> okay
[13:26:50] <OndraSterver> CLK
[13:26:50] <OndraSterver> on PDI
[13:26:54] <jacekowski> ahm
[13:26:57] <OndraSterver> CLK ... SCK... it all is the same
[13:27:00] <OndraSterver> serial clock
[13:27:22] <OndraSterver> damnit, where is my box with resistors
[13:30:21] <OndraSterver> nope, 1M pulldown does nothing
[13:31:21] <OndraSterver> hah
[13:31:25] <OndraSterver> 22pF added to SCK
[13:31:26] <OndraSterver> BOOM WORKS
[14:16:57] <DanFrederiksen> can you start a function from an interrupt function in such a way that it doesn't block the next interrupt?
[14:17:15] <DanFrederiksen> atmega8
[14:17:26] <OndraSterver> ISR(..., NO_BLOCK)
[14:18:02] <DanFrederiksen> ok thanks. is that normally the way you mix timer and non timer code?
[14:18:42] <OndraSterver> no, I just design it the way that the timer doesn't run nonblocking code :)
[14:20:46] <DanFrederiksen> my pwm timer fluctuates so if I want to add display functionality that way I'll have to do a lot of code to make sure I use the free part of the pwm cycle. it's ugly
[14:21:25] <DanFrederiksen> an interrupted function resumes afterwards right?
[14:21:40] <OndraSterver> yes
[14:22:06] <OndraSterver> you do know that in order to use PWM you don't need any pin handling, right?
[14:22:09] <OndraSterver> in software
[14:22:10] <DanFrederiksen> ok. that has to be the nicest way assuming the protocol can tolerate pauses in the data stream
[14:22:46] <DanFrederiksen> OndraSterver, yeah. I active an ADC current sensor right before the pwm ends to measure current and such
[14:22:57] <OndraSterver> ah ok
[14:23:14] <DanFrederiksen> I hate realtime code. it's so fragile :)
[14:24:13] <DanFrederiksen> and I'd really like multiple ADCs so I don't have to swap around all the time and time everything just right
[14:24:50] <DanFrederiksen> I want all channels to sample all the time so I can just read a register
[14:25:15] <DanFrederiksen> and no 26 clockcycle bullshit. it has one cycle. no more :)
[14:27:52] <Kevin`> DanFrederiksen: xmega is a bit better about that, worth reading the datasheet
[14:27:58] <DanFrederiksen> 1MHz samplers. 26us is a life time in high voltage electronics
[14:38:40] <OndraSterver> guys, any idea under which name I could find these connectors on mouser?
[14:38:41] <OndraSterver> http://www.gme.cz/dokumentace/821/821-257/pctdetail.821-257.1.jpg
[14:39:22] <OndraSterver> there is few hundreds of thousands connectors on mouser
[14:39:27] <OndraSterver> with a hundred categories
[15:02:34] <OndraSterver> Dear GCC
[15:02:37] <OndraSterver> can you explain
[15:02:38] <OndraSterver> THIS
[15:02:38] <OndraSterver> http://pastebin.com/hJUFTNPF
[15:02:52] <OndraSterver> ldd r24, 0x0050
[15:02:56] <OndraSterver> sbrs r24, 2
[15:02:58] <OndraSterver> rjmp .-6
[15:03:00] <OndraSterver> THANKS
[15:03:10] <OndraSterver> err
[15:03:10] <OndraSterver> lds
[15:03:54] <OndraSterver> 0x0051*
[15:12:22] <megal0maniac> !seen RikusW
[15:12:40] <OndraSterver> RikusW was last seen on Sep 1 at 00:10
[15:12:43] <OndraSterver> something like that
[15:12:49] <OndraSterver> !seen RikusW
[15:12:50] <tobbor> RikusW was last seen in #avr on Sep 01 10:43 2012
[15:12:52] <OndraSterver> ha
[15:12:55] <OndraSterver> I wasn't so far
[15:13:29] <megal0maniac> Close enough :) I thought I'd done it wrong. Never spoken to tobbor before
[15:13:45] <OndraSterver> he is random
[15:13:55] <OndraSterver> randomly refuses to reply
[15:14:12] <megal0maniac> Or takes his time
[15:14:27] <OndraSterver> no
[15:14:29] <OndraSterver> refuses to reply
[15:14:38] <OndraSterver> I think he is sometimes even racist
[15:16:14] <megal0maniac> Well you managed to type out and send a reply before a bot could. That sounds like sleeping on the job
[15:16:31] <megal0maniac> tobbor: hi
[15:20:39] <OndraSterver> http://pastebin.com/ntWUKzHq
[15:20:44] <OndraSterver> this is optimalized version
[15:20:54] <OndraSterver> of that gcc's crap
[15:21:01] * megal0maniac is on his phone
[15:21:08] <OndraSterver> 4 instructions
[15:21:10] <OndraSterver> instead of 20
[15:21:29] <megal0maniac> Which is a Sony Ericsson W810i
[15:21:35] <OndraSterver> ugh
[15:21:45] <OndraSterver> are you using terminal to connect to IRC? :D
[15:22:00] <megal0maniac> virca
[15:22:20] <megal0maniac> The 80mhz cpu is barely breaking a sweat :P
[15:22:25] <OndraSterver> :D
[15:24:42] <megal0maniac> I actually have the schematic. But it doesn't specify ICs
[15:25:30] <OndraSterver> heh
[15:25:36] <OndraSterver> I have got schematics for few phones
[15:25:51] <OndraSterver> but it is not like there is much really
[15:25:53] <OndraSterver> one SoC
[15:25:54] <OndraSterver> one DRAM
[15:25:56] <OndraSterver> one flash
[15:26:03] <OndraSterver> few capacitors
[15:26:05] <OndraSterver> few resistors
[15:26:06] <OndraSterver> JTAG header
[15:26:08] <OndraSterver> done
[15:26:13] <OndraSterver> charging circuitry maybe
[15:26:53] <megal0maniac> You'd be amazed
[15:27:16] <megal0maniac> Remind me to send it to you tomorrow
[15:28:44] <OndraSterver> I have got plenty
[15:28:48] <megal0maniac> 15 A4 pages
[15:28:49] <OndraSterver> omnia7, lumia800/710
[15:28:54] <OndraSterver> motorola f3
[15:29:28] <megal0maniac> Motorola f3 components: Motorola f3
[15:29:34] <megal0maniac> Done
[15:29:55] <megal0maniac> Component placement: box
[15:30:13] <megal0maniac> Still want one :)
[15:30:32] <OndraSterver> heh
[15:30:34] <OndraSterver> I have one :P
[15:30:48] <OndraSterver> it is creepy - I don't have battery in it and it still shows NO SIM error... E-INK FTW
[15:31:34] <donryanocero> neato. homebrew phone with e-ink?
[15:32:31] <OndraSterver> hehe
[15:32:37] <OndraSterver> I don't have use for it :/
[15:32:42] <OndraSterver> it was $4 incl. shipping
[15:32:44] <OndraSterver> so I bought it :D
[15:32:50] <OndraSterver> it has got SPI interface
[15:32:54] <megal0maniac> You can ship it with my xboard :)
[15:32:57] <OndraSterver> :D
[15:35:00] <donryanocero> where did you get it for $4?
[15:35:03] <megal0maniac> What can you do with the SPI interface?
[15:35:17] <OndraSterver> donryanocero, local ebay-like website
[15:35:19] <megal0maniac> And is shipping international?
[15:35:19] <OndraSterver> somebody was selling it
[15:35:24] <OndraSterver> no battery, no back cover
[15:35:24] <donryanocero> me want.
[15:35:28] <megal0maniac> And did you buy the new launchpad?
[15:35:36] <OndraSterver> no, I haven't bought new cortex launchpad
[15:35:37] <donryanocero> i has! 4.30 from TI! woo woo
[15:35:42] <OndraSterver> heh
[15:35:47] <OndraSterver> I have got SO MANY dev platforms here
[15:35:52] <megal0maniac> 4.30?
[15:35:54] <OndraSterver> I still haven't even opened the launchpad :(
[15:35:55] <donryanocero> $
[15:35:59] <OndraSterver> 430
[15:36:02] <megal0maniac> You got the wrong one, my friend :)
[15:36:03] <OndraSterver> MSP430*
[15:36:04] <OndraSterver> heh
[15:36:12] <donryanocero> i haven't done anything with it either.
[15:36:13] <OndraSterver> this one was made available like a week after 430 arrived
[15:36:58] <megal0maniac> It's 5 bucks but far more powerful. ARM based
[15:37:27] <donryanocero> Ahh. never messed with ARM stuff. but neato!
[15:37:29] <OndraSterver> I know
[15:37:34] <OndraSterver> ARM is powerful
[15:37:39] <OndraSterver> but can it beat my 80MHz xmega?
[15:37:40] <OndraSterver> :P
[15:38:04] <specing> 80 MHz xmega?
[15:38:10] <OndraSterver> overclocked
[15:38:15] <specing> Does it work?
[15:38:17] <OndraSterver> LED was blinking on 80MHz clock just fine
[15:38:21] <OndraSterver> 84MHz not anymore
[15:38:22] <OndraSterver> :D
[15:38:29] <specing> USART?
[15:38:30] <OndraSterver> I suppose that 64MHz could be long term stable just fine
[15:38:38] <OndraSterver> peripherals have their own prescalers
[15:38:41] <OndraSterver> global ones
[15:39:17] <specing> oh
[15:39:21] <specing> and the core works?
[15:39:26] <specing> do some math to be sure
[15:39:33] <specing> use all the instructions
[15:39:42] <specing> send the result to a port
[15:39:54] <OndraSterver> heh
[15:39:55] <OndraSterver> maybe laters
[15:40:13] <megal0maniac> What is standard clock?
[15:40:27] <donryanocero> Erm. so it works so-far-so-good? :-P
[15:40:29] <OndraSterver> 32
[15:40:40] <OndraSterver> 48MHz is 100% no problem since it is USB specced
[15:40:47] <OndraSterver> and good half of the chip has to be able to run at such speeds
[15:40:56] <OndraSterver> 64MHz shouldn't be problem either IMHO :)
[15:41:25] <megal0maniac> You're nearing the power of my phone :P
[15:41:34] <OndraSterver> I think I am past it
[15:41:39] <OndraSterver> does it do 1 instruction per cycle? :P
[15:41:57] <megal0maniac> It does many things
[15:42:06] <Kevin`> phones can't do anything useful with their processing power, so you are far past it already =p
[15:42:07] <megal0maniac> Probably not that
[15:42:25] <OndraSterver> can it blink two LEDs?
[15:42:28] <OndraSterver> SEPARATELY?
[15:42:36] <megal0maniac> The peripherals make it cool
[15:42:49] <megal0maniac> YES :D
[15:43:06] <specing> This makes me wonder how far I could push my AT91SAM9260...
[15:43:08] <megal0maniac> This firmware is hackable
[15:44:15] <OndraSterver> gcc: 0x6A bytes
[15:44:17] <OndraSterver> me: 0x31
[15:44:20] <OndraSterver> SUCK ON THAT
[15:44:29] <OndraSterver> and I could even optimalize it even more!
[15:45:12] <megal0maniac> Different platforms handle it differently. For example, the mega1284 can't be overclocked, the mega644 can go up to around 28mhz and the atxmega doesn't care much for manufacturer spec
[15:45:49] <specing> Thats right
[15:45:49] <specing> fucking gcc.
[15:45:49] <specing> :P
[15:45:49] <OndraSterver> haha specing
[15:45:49] <OndraSterver> did you finally gave up on it
[15:45:51] <OndraSterver> and started writing ASM only?
[15:45:51] <specing> yup
[15:46:04] <specing> lets code a high-level assembler, shall we?
[15:46:10] <OndraSterver> hehe
[15:46:13] <OndraSterver> I had such an idea!
[15:46:20] <OndraSterver> simplifying all branching etc
[15:46:25] <specing> register allocator and type checking ;P
[15:46:26] <OndraSterver> I have got even draft somewhere on my HDD
[15:46:29] <OndraSterver> oh
[15:46:33] <OndraSterver> not THAT high level
[15:46:37] <specing> automatic intXXX_t handling
[15:46:45] <specing> I've wrote an assembler once
[15:46:55] <specing> for a motorola mc6803
[15:46:56] <OndraSterver> I've simplified it by far
[15:46:59] <OndraSterver> stuff like
[15:47:04] <OndraSterver> R16 <- 0x5A
[15:47:13] <specing> :)
[15:47:17] <OndraSterver> R16 < R15 ? Go somewhere
[15:47:28] <specing> I'd do it like this:
[15:47:33] <OndraSterver> R16(5) <- 1
[15:47:37] <OndraSterver> aka bit5 in R16 set to 1
[15:47:44] <OndraSterver> works on any IO register or memory
[15:47:51] <OndraSterver> would work*
[15:47:54] <specing> @alloc (myreg, anothereg)
[15:48:11] <specing> ... code using $myreg and $anothereg
[15:48:32] <specing> and the assembler would automatically determine what register to use
[15:49:26] <OndraSterver> gcc is register waster
[15:49:35] * megal0maniac goes back to Arduino
[15:49:46] <OndraSterver> lol
[15:49:51] <OndraSterver> that is even bigger waster
[15:50:24] <megal0maniac> Well it's a wrapper (waster) around C
[15:50:36] <megal0maniac> gcc to be exact
[15:50:36] <OndraSterver> why do they call it "wiring"
[15:50:39] <OndraSterver> when it is c++?
[15:51:20] <megal0maniac> Derived from Processing. And the code in Arduino IDE isn't real C++
[15:51:32] <specing> it is real C++
[15:51:41] <specing> they've only supplied their own classes
[15:51:53] <megal0maniac> Isn't a lot of the syntax different?
[15:51:55] <specing> fuckTARDs :)
[15:51:56] <megal0maniac> Oh
[15:51:58] <megal0maniac> True
[15:52:01] <specing> megal0maniac: no
[15:53:13] <megal0maniac> I had a cool idea for the Arduino uno
[15:53:48] <megal0maniac> It's based on a mega328p with a mega16u2 acting as a usb to serial converter
[15:53:48] <specing> Undoino.
[15:54:16] <megal0maniac> But there's a PDI header for the 16u2
[15:55:38] <megal0maniac> So why not program it to be a STK500 programmer? have a jumper to determine the reset pin direction and you've got yourself a 2 in 1 board. And you no longer need a bootloader in the 328
[15:56:05] <specing> Who would want to pay $50 for a 'duino programmer
[15:56:10] <megal0maniac> Although Arduino is designed to be as simple as possible... But it would be a cool option
[15:56:15] <specing> you can get a dragon for that
[15:56:21] <specing> or rikusW's stuff
[15:56:26] <specing> for even less
[15:56:30] <megal0maniac> Well I paid about half that
[15:56:56] <megal0maniac> I've got rikus' board. It's amazing and gave me the idea :)
[15:57:15] <megal0maniac> Want to buy more, but can't justify it
[15:57:57] <OndraSterver> one can do it all!
[15:59:34] <specing> we need an all-in-one xmega firmware
[15:59:43] <specing> kinda like the bus pirate
[15:59:49] <specing> only 50 times faster
[16:00:48] <megal0maniac> Well the bus pirate is open source... That should make it easy to get started
[16:01:09] <specing> SPI, PDI, TPI, USB-{1-8}*serial, USB-{1-4}*-SPI, USB-{1-4}*-I2C, USB-ADC, USB-DAC, USB-bitbang, USB-PWM,...
[16:01:24] <specing> I mean an all in one firmware that can switch its mode on the fly
[16:01:33] <specing> not by reflashing
[16:01:56] <megal0maniac> I know. Isn't the firmware open source?
[16:01:59] <specing> since there is only a limited number of usb endpoints, the device should just reenumerate itself
[16:02:17] <specing> or embed its own protocol inside the usb stream
[16:04:16] <megal0maniac> I like it. ASM everything
[16:04:31] <specing> Yup
[16:05:33] <megal0maniac> Generic HID should suffice, no?
[16:07:18] <megal0maniac> Well I'm off. Goodnight all
[16:43:54] <jadew> any idea how to .def SOMETHING, Rxx with avr-as?
[16:44:05] <jadew> .equ doesn't seem to cut it
[16:49:12] <jadew> gonna try it under nix
[17:57:18] <OndraSterver> OH NOES
[17:57:21] <OndraSterver> THE HOUSE IS GOING TO FALL APART
[17:57:25] <OndraSterver> I AM RUNNING OUT OF DUCKTAPE!
[18:07:54] <donryanocero> nooo
[18:08:42] <Essobi> heh
[18:12:04] <OndraSterver> you need only 2 things to fix anything
[18:12:06] <OndraSterver> ducktape
[18:12:08] <OndraSterver> and WD40
[18:12:15] * Tom_itx sets OndraSterver on fire
[18:12:19] <OndraSterver> ducktape
[18:12:30] <OndraSterver> make blanket out of ducktape
[18:12:30] <Tom_itx> WD40?
[18:12:31] <OndraSterver> lol
[18:12:33] <OndraSterver> yes
[18:12:42] <OndraSterver> it is that lubricant for motors and what not
[18:15:08] <Essobi> WD40 sucks on locks.
[18:15:17] <Kevin`> wd40 evaporates, don't use it as a lubricant
[18:15:26] <Kevin`> it's for cleaning
[18:15:29] <OndraSterver> or cleaning
[18:15:30] <OndraSterver> I don't use it
[18:15:38] <Kevin`> it's great for cleaning
[18:15:44] <OndraSterver> I just take it apart
[18:15:47] <Essobi> I've herd that
[18:16:25] <Kevin`> anything metal and mechanical, soak it in wd40
[18:16:38] <OndraSterver> anything metal and mechanical, TAKE IT APART
[18:16:43] <OndraSterver> I love taking stuff apart
[18:16:46] <OndraSterver> and learn from it
[18:16:52] <OndraSterver> I hate learning at school
[18:16:54] <OndraSterver> it sucks
[18:16:56] <OndraSterver> only theory
[18:16:59] <OndraSterver> no real cases etc
[18:17:07] <OndraSterver> all I've ever used was because I learnt it at home
[18:17:40] <Thetawaves> 'equipment autopsies' are really educational
[18:17:47] <Thetawaves> see how/why they failed
[18:18:04] <OndraSterver> I take apart even new stuff :( :D
[18:18:13] <OndraSterver> I got new laptop for $1k... 2 months I hold it!
[18:18:16] <OndraSterver> till I had to take it apart
[18:18:20] <OndraSterver> I just couldn't hold it anymore
[18:19:07] <OndraSterver> Tom_itx,
http://www.cyclelicio.us/2011/simple-bicycle-repair-guide/flowchart/
[18:19:09] <OndraSterver> here
[18:19:10] <OndraSterver> lol
[18:19:38] <OndraSterver> also - PDI works on my XBoards
[18:19:42] <OndraSterver> tested on dragon :P
[18:30:58] <OndraSterver> !seen Corwin
[18:30:59] <tobbor> Corwin was last seen in #avr on Sep 01 09:35 2012
[18:31:03] <OndraSterver> !seen RikusW
[18:31:04] <tobbor> RikusW was last seen in #avr on Sep 01 10:43 2012
[18:31:07] <OndraSterver> where have they gone?
[20:06:56] <jadew> in order to add some value to Z, do I have to go like:
[20:06:58] <jadew> add ZL, Rx
[20:07:04] <jadew> adc ZH, zero
[20:07:44] <jadew> that's a question btw
[20:11:00] <Tom_itx> without an answer
[20:12:18] <Tom_itx> i haven't done enough asm to tell you
[20:14:02] <OndraSterver> yes
[20:14:20] <OndraSterver> for 16bit adding you need add low1, low2 adc hi1, hi2
[20:14:26] <OndraSterver> if you want immediate you can use adiw
[20:14:31] <OndraSterver> but that goes only upto 64
[20:14:35] <OndraSterver> or 63
[20:18:22] <jadew> no, I just want to add a few bytes to an address
[20:18:25] <jadew> thanks for the answer
[20:18:49] <OndraSterver> np
[20:18:53] <OndraSterver> when it comes to asm.. :)
[20:19:07] <jadew> I'm trying to rewrite that dht11 to 1-wire slave thingie, I think it's gonna fit just fine in 1k
[20:21:04] <OndraSterver> :)
[20:21:05] <OndraSterver> gl
[20:21:07] <OndraSterver> tiny13a?
[20:21:11] <OndraSterver> with software SPI?
[20:21:14] <OndraSterver> oh
[20:21:16] <OndraSterver> into 1wire
[20:21:18] <jadew> tiny 13
[20:21:24] <OndraSterver> it is getting late for me to think :D
[20:22:42] <jadew> heh, I feel quite confortable with asm, however I need to get used to the instructions and how things work
[20:23:25] <OndraSterver> my biggest issue is to work with 8bit stuff while the stuff I often work with is 16 or 32bit :/
[20:23:30] <jadew> so I hope I won't need any luck :P
[20:23:31] <OndraSterver> so unless one has plenty of macros
[20:23:46] <OndraSterver> same for always LDI R16, something; STS somewhere, R16
[20:23:55] <OndraSterver> instead just somewhere = something
[20:24:04] <OndraSterver> plus no visual helper for ASM in AS6 :(
[20:24:29] <jadew> not really much to help with
[20:24:37] <OndraSterver> sure there is
[20:24:49] <OndraSterver> all those definitions from iom????.inc :)
[20:25:01] <OndraSterver> who is supposed to remember all those 500 names?
[20:25:07] <OndraSterver> bit masks, registers...
[20:25:19] <OndraSterver> it is easy on megas and tiny
[20:25:26] <OndraSterver> but when you go to xmegas there is really huge amount of registers
[20:25:30] <jadew> ah, I keep the spreadsheet opened
[20:25:45] <jadew> ah
[20:26:10] <jadew> I haven't used anything bigger than a mega
[20:27:05] <OndraSterver> I often fall back to asm
[20:27:27] <OndraSterver> because I get annoyed by C and its pointer-depointing that originally was pointing at pointer that was pointing to some value
[20:27:28] <OndraSterver> ARRGH
[20:27:40] <OndraSterver> then I fiddle with * and & to get it working
[20:27:45] <OndraSterver> but it still doesn't work
[20:28:13] <OndraSterver> some things are just simpler to be made in asm
[20:28:37] <OndraSterver> place on position FF80 in flash code that jumps into another function?
[20:28:40] <OndraSterver> in asm you just do JMP something
[20:28:41] <OndraSterver> in C?
[20:28:42] <jadew> readability has to suffer tho
[20:28:43] <OndraSterver> good luck.
[20:28:49] <OndraSterver> readability can be maintained
[20:29:05] <jadew> it's not the same
[20:29:59] <OndraSterver> sure it is not the same
[20:30:08] <OndraSterver> but with a bit of luck it can be still maintainable
[20:31:22] <jadew> one big problem with readability is brne
[20:31:37] <jadew> which is the usual jump after a comparation
[20:31:46] <OndraSterver> hehe
[20:31:58] <jadew> and it usually goes to some end_routing:
[20:32:01] <OndraSterver> that was "solved" by my low level human thing
[20:32:03] <jadew> *end_routine:
[20:32:11] <OndraSterver> which would support { } but would be on asm level
[20:32:14] <OndraSterver> or very near it
[20:32:31] <OndraSterver> R15 < R16 ? { ... }
[20:32:41] <jadew> like a preprocessor?
[20:32:45] <OndraSterver> SREG(c) 1 ? { .. }
[20:32:53] <OndraSterver> no, it would directly translate to binary
[20:32:56] <OndraSterver> probably
[20:33:07] <OndraSterver> that is easier maybe
[20:33:08] <jadew> preprocessor would be easier to implement
[20:33:14] <jadew> like... 2 hours
[20:33:21] <OndraSterver> there is not that much instructions on AVR
[20:33:22] <OndraSterver> they repeat
[20:33:23] <jadew> it's not a bad idea
[20:33:37] <jadew> well, you don't have to care about the instructions
[20:33:45] <jadew> you just take care about your syntax
[20:33:55] <jadew> then pass the output the assembler
[20:33:58] <OndraSterver> and translate it into what, if not raw instructions?
[20:34:05] <jadew> raw instructions
[20:34:15] <OndraSterver> so you do have to care about instructions :)
[20:34:21] <jadew> why?
[20:34:25] <OndraSterver> so you know what you can use
[20:34:28] <OndraSterver> and translate what to what
[20:34:32] <OndraSterver> there was more than just { }
[20:34:36] <jadew> ah, you mean based on different avrs?
[20:34:36] <OndraSterver> in the idea
[20:34:41] <OndraSterver> that too
[20:35:04] <jadew> shouldn't be that big of a drag, I think the condition thing is great
[20:35:19] <jadew> it actually sounds awesome
[20:35:46] <OndraSterver> like I said, there was more things I wanted to "implement" - bit operations on any register - both Rx and from register file
[20:35:48] <OndraSterver> or even memory
[20:36:12] <OndraSterver> aka [0x3254](3) <- 0
[20:36:13] <jadew> can't you do that with macros?
[20:36:19] <OndraSterver> aka bit 3 on mem location 0x3254 = 0
[20:36:21] <OndraSterver> no idea
[20:36:27] <jadew> you should be
[20:36:37] <jadew> macros can take arguments
[20:36:41] <OndraSterver> yes
[20:36:48] <OndraSterver> but I dare you to define such macro
[20:36:55] <OndraSterver> something <- something else
[20:37:10] <jadew> you don't have to use that exact syntax
[20:37:16] <OndraSterver> but I want to :)
[20:37:31] <jadew> you can go like mmbi 0x3254, 3, 0
[20:38:38] <OndraSterver> that looks awful though
[20:39:56] <OndraSterver> anyway
[20:40:00] <OndraSterver> 0328
[20:40:02] <OndraSterver> gn
[20:42:22] <jadew> night
[22:58:51] <Richard_Cavell> I just got my Tiny85s and mega328Ps in the mail
[22:59:02] <Richard_Cavell> About half a kilo of packaging for 8 tiny microchips
[23:14:14] <Casper> hehe
[23:14:20] <Casper> yeah they love to do over packing
[23:15:10] <Casper> seriously, sometime I wonder if those packing stuff making compagny don't corrupt all the compagny to use way too much of their products
