Wednesday, 2018-12-12

*** tpb has joined #timvideos00:00
*** TheAssassin has quit IRC00:16
*** TheAssassin has joined #timvideos00:21
*** futarisIRCcloud has joined #timvideos01:42
*** CarlFK has quit IRC01:48
xobsmithro: I remember an ancient article about being able to write FORTRAN in any language.  I'm sure it applies somehow.02:21
*** Kripton has quit IRC03:11
*** Kripton has joined #timvideos03:19
*** froztbyte has quit IRC03:37
*** froztbyte has joined #timvideos03:45
*** olasd has quit IRC03:58
*** CarlFK has joined #timvideos04:02
*** ChanServ sets mode: +v CarlFK04:02
*** olasd has joined #timvideos04:05
mithrotannewt: ping?05:19
mithrotannewt: Is "set_config #:1, process_set_config: 463: failed" expected?05:19
tannewtNo it’s not. I’m at choir now though. Probably an endian issue05:21
tannewtI think that was due to the stall not being cleared by the setup token05:22
tannewtThe setup packet should always be acked05:22
tpbTitle: Snippet | IRCCloud (at
mithrotannewt: that is happening now05:22
tannewtI’m not sure what that output means. I may have time tomorrow to look. I’m on my phone now05:23
mithrotannewt: Okay05:24
xobsmithro: your string descriptors look like you're feeding them utf8/ascii data when they should be utf16 (or is it ucs2?)05:36
xobsIt also sounds like you might be missing the SET_CONFIGURATION command?  For that you just need to respond with a zero-byte DATA packet.05:44
*** CarlFK has quit IRC05:44
xobsHere's what it looks like in Grainuum: (note that response_len remains 0 for both SET_ADDRESS and SET_CONFIGURATION)05:44
tpbTitle: grainuum/grainuum-state.c at master · xobs/grainuum · GitHub (at
mithrotannewt: I'm not seeing any of the endpoint opens....05:47
mithro[278609.785183] cdc_acm 2-13.1.3:1.0: acm_port_activate - usb_submit_urb(ctrl irq) failed05:47
mithroalso this suspect line...05:48
mithro[278494.791232] xhci_hcd 0000:00:14.0: xHCI xhci_add_endpoint called with enabled ep 00000000150719c905:48
mithrothe 463 == line 46305:52
mithroThe issue seems to be05:56
mithro      TU_ASSERT( 0xff == _usbd_dev.itf2drv[desc_itf->bInterfaceNumber] );05:56
xobsIs there an off-by-one error with bInterfaceNumber?  I can't remember if it's 0-indexed or 1-indexed.05:57
*** rohitksingh_work has joined #timvideos06:14
mithroI removed that assert and now I get endpt_open06:30
mithroIt seems to have an issue with06:52
mithro[282354.299357] cdc_acm 2-13.1.3:1.0: acm_ctrl_msg - rq 0x20, val 0x0, len 0x7, result -11006:52
mithroxobs / tannewt: Pushed the latest version with a lot of updates, reworked / rewrote a lot of the endpoint interface stuff06:59
xobsmithro: sounds like a lot of really good progress!06:59
mithroIt'll now correctly always ACK a setup packet07:00
mithroand you can stall endpoints07:00
mithroas far as I can tell the hardware seems to be working pretty well07:01
mithroEverything should be pushed now07:06
tannewtNice job mithro! I may have time to test tonight07:15
mithrotannewt: I also put everything into a nice structure ->
tpbTitle: tinyusb/dcd_valentyusb.c at develop · mithro/tinyusb · GitHub (at
*** thaytan has quit IRC07:39
*** thaytan has joined #timvideos07:40
*** ChanServ sets mode: +v thaytan07:40
tannewtmithro, updating everything now!07:42
*** swalladge_ has joined #timvideos07:47
mithrotannewt: I think I'm going to head home in a second07:47
*** swalladge has quit IRC07:47
tannewtk np. I'm surprised you stay at work to work on it07:47
tpbTitle: Snippet | IRCCloud (at
mithrotannewt: My work setup is much nicer than my home setup, 3 * 4k monitors07:48
tannewtnice! I have one ultrawide and its my work and home setup :-)07:48
tannewtthat looks like its working07:48
tannewtwhat are you working on now?07:49
tpbTitle: valentyusb/ at master · mithro/valentyusb · GitHub (at
mithrotannewt: I'm pretty happy with how the tests are starting to turn out - but they are very slow....07:50
mithrohrm, I should finish my dinner before going home...07:50
tannewtya, definitely07:51
tannewtare you updating litex-buildenv or just valentyusb?07:52
mithrotannewt: you'll want to update everything07:52
mithrotannewt: there are some changes in litex, valentyusb and tinyusb packages07:52
tannewtmithro, trace looks wayyyyyy cleaner!07:58
tannewtvery nice job!07:59
mithrotannewt: so, need to figure out what is going that means cdc-acm isn't working08:02
tannewtya, looking at that now08:03
tannewtits trying to get or set the linecoding I think08:03
tannewtthe strings are the incorrect endian as well08:03
tannewtmy bx isn't staying in the bootloader after reset08:12
mithrotannewt: I had that - needed to reset the usb bus08:12
tannewthow do you do that?08:13
tpbTitle: Snippet | IRCCloud (at
mithrotannewt: sudo rmmod xhci; sudo modprobe xhci08:13
mithrotannewt: Actually seems to be "sudo rmmod xhci_pci; sudo modprobe xhci_pci"08:14
tannewtkk thanks08:14
tannewtstill not happy08:16
tannewtmay need to restart08:16
tannewtya, not enumerating08:17
mithrowell, I'm *actually* heading home now08:17
tannewtk thanks! I'll see if I can figure this out in the next 20 minutes or so08:17
tannewtthen bedtime :-)08:17
*** CarlFK has joined #timvideos08:25
*** ChanServ sets mode: +v CarlFK08:25
*** swalladge_ is now known as swalladge09:02
*** cr1901_modern has quit IRC09:06
tannewtmithro: I got the strings looking ok and the endpoint size correct but its still crashing the xhci every time. (changes here: The beagle doesn't show any token coming from the host after the zero length IN after the set line coding. Is it possible its something lower level? Anyway I'm off to bed. Great work on this! Sooo close. :-)09:06
tpbTitle: Comparing mithro:develop...tannewt:endian_cleanup · mithro/tinyusb · GitHub (at
*** futarisIRCcloud has quit IRC09:12
thaytananyone else get email from Ark Jadoon <[email protected]> about FPGA video compression?11:01
*** tumbleweed_ has quit IRC11:05
*** tumbleweed_ has joined #timvideos11:05
*** tumbleweed_ is now known as tumbleweed11:05
tumbleweedyeah, assuming because of timvideos11:06
tumbleweedsounds like someone who wants people to do his schoolwork for him11:07
*** puck_ has quit IRC11:23
*** puck_ has joined #timvideos11:25
*** cr1901_modern has joined #timvideos12:02
*** rohitksingh_work has quit IRC13:33
*** rohitksingh has joined #timvideos14:30
*** rohitksingh has quit IRC15:36
*** thaytan has quit IRC16:18
*** thaytan has joined #timvideos16:22
*** ChanServ sets mode: +v thaytan16:22
mithrotmmmcci: morning?16:45
*** rohitksingh has joined #timvideos18:17
*** rohitksingh has quit IRC21:02
*** rpifan has joined #timvideos21:02
*** rohitksingh has joined #timvideos21:05
*** rpifan has quit IRC21:51
mithrotannewt: morning?21:58
tannewthi hi mithro21:58
mithroSo I'm getting to21:59
tpbTitle: Snippet | IRCCloud (at
mithrotannewt: It looks like it was expecting a status packet and got a setup packet?22:00
mithrobut it was a tok == out....22:00
tannewtya, that looks like a set config setup packet22:01
mithrotannewt: I'm going to change to tok being per endpoint22:02
tannewtgood call22:02
tannewtonly endpoint 0 should get setup22:02
tannewtI think its the only endpoint that will get a mix of tokens22:03
mithrotannewt: Yeah - it's confusing like that22:03
mithroI was pondering making the endpoint "address" == tok_type + epnum22:04
tannewtaddress is usually num + direction22:04
mithrotannewt: Yeah - but setup is kind of special....22:05
tannewtya, it is22:06
mithroThe only thing which stopped me is the stall state being kind of shared...22:07
mithroHrm, it didn't seem to change anything...22:08
mithro[337536.763623] usb 2-13.1.3: can't set config #1, error -11022:10
tannewtI wasn't getting that last night22:11
tannewtdid you take a peek at the changes I posted last night?22:11
mithroHrm, I was getting /dev/ttyACM0 last night....22:13
mithroSo, the "ep0: tok out w:0 0 9 1 0 0 0 0 0 g:8 -- ERR long!!" is the "SET CONFIGURATION" request...22:17
mithrotannewt: It kind of looks like the status token and setup token are getting mundged together or something?22:18
tannewtmaybe. I don't know. I can't look now. adding midi to tinyusb22:19
*** rohitksingh has quit IRC22:22
*** Toba__ has joined #timvideos22:32
*** nancy[m]1 has joined #timvideos22:35
*** pzieba[m]1 has joined #timvideos22:35
*** Toba has quit IRC22:35
*** pzieba[m] has quit IRC22:35
*** nancy[m] has quit IRC22:35

Generated by 2.13.1 by Marius Gedminas - find it at!