Tuesday, 2019-08-20

*** tpb has joined #tomu00:00
*** awe00 has quit IRC00:01
*** emeb has quit IRC00:01
*** rohitksingh has quit IRC00:10
*** rohitksingh has joined #tomu00:38
*** whatnick has joined #tomu00:42
*** tom_ has joined #tomu00:58
*** tom_ has quit IRC01:03
*** whatnick has quit IRC01:04
*** rohitksingh has quit IRC01:13
*** rohitksingh has joined #tomu01:44
*** futarisIRCcloud has joined #tomu03:35
TomKeddiemithro: I'm back in town, hoping to close that RGB PR.  The gateware I build unmodified from your repo doesn't give me a bios prompt on icebreaker.  Is this something new?  I'm following the micropython wiki (https://github.com/timvideos/litex-buildenv/wiki/HowTo-FuPy-on-iCE40-Boards)  Any debug tips?  I guess I could look at recent PRs and roll back to see where it broke?03:52
tpbTitle: HowTo FuPy on iCE40 Boards · timvideos/litex-buildenv Wiki · GitHub (at github.com)03:52
*** rohitksingh has quit IRC05:13
*** rohitksingh has joined #tomu05:36
*** rohitksingh has quit IRC05:50
mithroTomKeddie: I had something working on the icebreaker recently but will check again when I get to the camp06:54
mithroJust landed in Berlin06:54
*** xkapastel has joined #tomu07:48
*** cedric has quit IRC07:49
*** cedric has joined #tomu07:49
*** cedric has joined #tomu07:49
xobsWell, I think I have the SET_ADDRESS issue solved. It works with both the dummyusb and epfifo cpu interfaces.07:55
xobsThe only thing I have to do now is get the updater working again.07:55
xobsI recently rewrote the updater to make it harder to brick a Fomu by pulling power during an update. But it seems not well tested. At least it's not bricking my test unit!07:57
*** olasd has quit IRC09:41
*** olasd has joined #tomu09:44
*** awe00 has joined #tomu09:51
*** futarisIRCcloud has quit IRC10:34
*** xkapastel has quit IRC11:07
TomKeddiemithro: thanks, no rush.  Enjoy the camp.11:43
MadHackerxobs: "Not bricking" is a good starting point. :)12:17
*** awe00 has quit IRC12:47
tmbincxobs: is an EVT2 still good? (the one you gave me in december)12:54
tmbincI could test the SET_ADDRESS fix then if it resolves my issue12:54
tmbincbut actually i think for whatever reason windows is asking for descriptor 0x0600, which FOMU STALLs as it doesn't exist12:55
tmbinc... i have no idea why windows would ask for that descriptor12:55
*** awe00 has joined #tomu13:07
xobstmbinc: they're about the same, yeah. I just haven't built a bitstream for it yet.13:11
tmbincok just flashed top-of-tree foboot and worked fine (i.e. on linux; still not on windows)13:12
tmbincso yeah EVT2 seems to work fine for me; if you have a fix, I can test it13:13
MadHackerI've a fomu here for testing too, and I don't mind if it bricks it and I need to JTAG it back to health later.13:25
MadHacker(Well, SPI)13:25
*** awe00 has quit IRC13:37
*** emeb has joined #tomu14:01
tntpsa, I'm at the camp, I have free fomu hackerboards and also test jig to flash them.14:12
xobsI'm in an airport. I'll be at camp in... Some number of hours. Probably in about 18 hours.14:13
xobsMadHacker: thanks for the help. I'm glad you're able to test it, and I'm sorry you have to!14:14
xobsIt's requesting descriptor 0600?14:14
tmbincthis maaay be related with me having UsbDk (that spice USB redirection thingie) installed14:15
tmbincBut all other USB devices still work14:15
tmbincI've checked the Windows USB ETW but meh i fail at windows14:15
tmbincBest I could figure out it fails at "reading the device descriptor with 0xC0000004" (USBD_STATUS_STALL_PID )14:16
xobstmbinc: oh. Hmm... Yes, I didn't check for USB resets. It was a Stage 1 Hack.14:16
tmbincbut the Urb indicates that's while reading the device descriptor, not while reading a random other descriptor14:16
xobsThough if it's stalling, then that means it's actually responding.14:16
tmbincI think the STALL response is actually fine14:17
tmbincafter that it's doing 80 00 00 00 00 00 02 (GET_STATUS)14:17
tmbincwhich the device responds "00 00" which should be fine14:17
tmbincthen windows is doing 00 09 01 00 00 00 00 00 (set configuration 0), which should be fine as well14:18
xobsYou sure it's not Windows asking for the USB high speed descriptor? You're supposed to stall those.14:18
tmbincthen it doesn't do anything [visible; i think OV isn't showing me the port reset] and then it's reading the device descriptor from address 014:18
xobsThat seems normal.14:18
tmbincbut why is it not happy  / why is it port-resetting?14:19
xobsAnd dfu-util -l or wishbone-tool can't see it?14:19
tmbinc(aside from the fact that it doesn't handle the reset)14:19
tmbincwell it keeps attempting to read the device descriptor after that, and it appears as "code 43" in the device manager14:19
xobsThat does seem ungood.14:20
xobsWell. Hopefully we can coax the Beagle into working again.14:20
tnttmbinc: can you post the ovftdi log ?14:21
tmbincxobs: I have a $$$ commercial USB3 analyzer (Teledyne/LeCroy) but I'm not ready to admit defeat yet and stop using OV to debug it :)14:21
tmbinctnt: http://tmb.elitedvb.net/tmp/M5N7G6xXmH/fomu2.pcap14:21
tmbincignore the crap at beginning and end14:22
tmbinc(needs nightly wireshark)14:22
tnttmbinc: arf yeah, I don't use wireshark, I just use the text log.14:23
tmbinci have that, too, just a second14:24
tmbincxobs: you're right, 0x0600 is the "device qualifer" and should be STALLed14:30
tmbincso that's not the issue14:30
tntYeah, there is really nothing wrong in that log.14:32
tmbincooh, maybe this is the missing address filter overall?14:32
tmbincif i unplug my FS keyboard, it'll work14:32
tmbincso likely the hub doesn't have a per-port TT14:32
tmbinctnt: I don't understand the final "READ DEVICE DESCRIPTOR" from address 0 though14:33
tmbincwhy is it doing that?14:33
tntWell , but bus was reset between that and the set configuration before.14:37
*** awe00 has joined #tomu14:38
tmbincok indeed the set_address fix (taken from master) fixes it14:40
tntinteresting the OV log shows really nothing wrong :/14:41
tmbincmaybe the issue happens in the beginning - we see a couple of "IN" tokens to address 2114:46
tmbincand the OV will respond to those14:46
tmbincso it caused a reset of my FS keyboard on the same TT14:47
tmbincso the READ DEVICE DESCRIPTOR we see at the end is really meant for the keyboard, but since both devices respond, it never works14:47
tmbincthis is also why - in the other log - we see a couple of NAKs but no furhter retries14:47
tmbinc(because the OV saw the NAK from FOMU, but the host likely saw something else / nothing valid)14:48
xobstmbinc: it shouldn't be naking with the fix. Are you still seeing that?14:48
tmbincnote how it keeps polling 30.114:49
tmbinc(which appears to be the keyboard)14:49
tmbincand no NAKs anymore14:49
tmbinc(from fomu)14:49
tmbincI think it's fixed14:50
xobsThe beginning is kinda ugly, though.14:50
tmbincthe noise?14:51
xobsYeah. Is that just something floating?14:51
tmbincI think that's just the hub being high impedance or maybe high speed?14:51
tmbincit's an openvizsla problem14:51
xobsOh, okay!14:52
xobsSo I'll work on the updater and then we can push the update.14:53
tmbinccan the bootloader be updated?14:53
tmbinci forgot, is it just in flash at 0?14:53
xobsIt's possible to load 1.9 onto a production unit to test it. I've done that, so I concur it works.14:53
xobsI have an updater that's half written. Two, actually.14:53
tmbincHappy to test that on an EVT :)14:54
xobsOne is a naive proof of concept that'll Brick your device if you unplug it during the 17 seconds it takes to update.14:54
xobsThe new one cuts the critical window down to just a few ms.14:54
xobstmbinc: the first thing it does is check itself to make sure it's valid. The next thing it does is erase block 0 and make it point to itself.14:56
xobsSp it'll resume the update if it gets reset.14:56
tmbincah using this "boot applet"?15:00
xobstmbinc: https://github.com/im-tomu/foboot/blob/master/booster/README.md15:04
tpbTitle: foboot/README.md at master · im-tomu/foboot · GitHub (at github.com)15:04
xobsWow. I stopped writing mid sentence.15:10
tmbincnit: 20097151 has an extra 0 :)15:12
xobsThat's around the time I started working on the new version of the USB core. The idea was to try avoiding needing to update at all. Guess that's not happening now.15:12
tmbincwell it was close i guess?15:15
xobsI'm glad I had it ready.15:15
xobsWhen I tried it earlier today it just gave an error. I don't have the ability to debug it at the moment.15:18
xobsI'll try hooking up https://github.com/im-tomu/valentyusb/blob/master/src/usb-epfifo.c (and adding SET_ADDRESS support to that file), which should be enough to placate the host to let me hook up gdb.15:20
tpbTitle: valentyusb/usb-epfifo.c at master · im-tomu/valentyusb · GitHub (at github.com)15:20
xobsBrb, taking off now.15:20
tmbincsee you tomorrow15:23
*** xkapastel has joined #tomu15:34
*** awe00 has quit IRC15:35
mithroI'm sitting in a hammock in the hardware hacking village16:03
mithrotnt: I'm at camp too!16:05
*** awe00 has joined #tomu16:16
*** im-tomu_ has left #tomu16:28
*** kristianpaul has joined #tomu16:28
*** im-tomu_ has joined #tomu16:28
*** awe00 has quit IRC16:32
xobsmithro: sounds pleasant.16:47
*** awe00 has joined #tomu16:59
tntmithro: \o/17:01
tntI'm at the Singularity city hacker tent. I also have a bunch of icebreaker boards for sale at a discount in addition to the free fomus.17:02
xobsI feel like I should have brought more evt boards.17:07
xobsBut I also have a bunch of busted dvt boards that people can turn into jewellery.17:08
*** awe00 has quit IRC17:12
*** im-tomu_ has left #tomu17:24
*** im-tomu_ has joined #tomu17:24
mithrotnt: I have lots of Fomu boards + chubby + pano logic devices17:30
*** xkapastel has quit IRC17:44
*** NoGodDamnIdea has joined #tomu18:00
*** emeb has left #tomu18:10
*** awe00 has joined #tomu18:24
*** r has joined #tomu19:16
*** tmarble has joined #tomu20:04
*** renatosrabelo has joined #tomu20:19
renatosrabelohi guys. Is there a fomu channel ?20:20
tntsort of. You're in it :)20:21
renatosrabelomy fomu just arrived. Is there a quick start guide like tomu ?20:21
renatosrabeloi guess the this one20:24
tpbTitle: GitHub - im-tomu/fomu-toolchain: A collection of tools for developing for Fomu (at github.com)20:24
tmarbleI just got my fomu too.. it's adorable! micropython just worked^TM :)20:41
shaprI've been working through https://p.xobs.io/td19/ and https://github.com/im-tomu/fomu-workshop with my fomu20:47
tpbTitle: Fomu: An FPGA in your USB Port (at p.xobs.io)20:47
*** W1lkins has quit IRC20:52
*** W1lkins has joined #tomu20:53
shaprI'm cautiously trying to upgrade my crowdsupply fomu from 1.8.7, is it correct that I should run "python3 foboot-bitstream.py --revision evt"  ?21:55
shaprthat is, inside foboot/hw from https://github.com/im-tomu/foboot ?21:55
tpbTitle: GitHub - im-tomu/foboot: Bootloader for Fomu (at github.com)21:55
*** emeb has joined #tomu21:58
*** xkapastel has joined #tomu22:00
shapronce I've modified the source and produced riscv-blink.elf how do I load that file onto the fomu?22:12
shaprwell, dfu-util -D riscv-blink.elf didn't seem to kill anything?22:14
shaprI want to build verilog-blink for my crowdsupply-shipped fomu, would I set FOMU_REV=evt1 ?22:18
shaprneeds to be FOMU_REV=pvt1 the other options don't work!22:23
*** futarisIRCcloud has joined #tomu23:17

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