2021-08-10.log

mntmnthe sn65dsi86 driver can set 1 dsi flag00:00
mntmnif (!(val & DPPLL_CLK_SRC_DSICLK))00:00
mntmndsi->mode_flags |= MIPI_DSI_CLOCK_NON_CONTINUOUS;00:00
mntmnwe do not use a refclk so DPPLL_CLK_SRC = DACP/N00:02
mntmnDPPLL_CLK_SRC_DSICLK should thus be true00:02
mntmnare you setting any refclk bits in u-boot let me see?00:03
chartreuseust want to give a slight update on the trackball, after a couple days use it has become much smoother and doesn't have any extra friction in certain directions00:04
chartreuseStill will probably make a model with the 1.5mm ball bearings when they get here as a test, but it has become much smoother just from use00:05
mntmnchartreuse: ok, that's very good to know, thanks00:05
mntmnbluerise: how do we know that 486 mhz is the correct dsi refclk?00:08
mntmnbluerise: but the PLL locking stuff in your bridge driver looks sane, it should lock on to the DSI clock and that also means that the DSI driver is working00:15
bluerisemntmn: because...00:22
bluerise(162 * 24) / (4 * 2)00:23
bluerise48600:23
bluerise(pixelclock * bits-per-pixel-or-so) / (lanes * 2)00:23
blueriseforgot what the 2 was00:23
mntmnah ok thanks00:26
mntmnbluerise: linux driver has a 10ms delay after setting the timings in the sn68 and before starting its video stream00:26
bluerisehmhm00:27
blueriseI can try that00:27
bluerisestarting video stream is the dcss init?00:27
bluerise _00:00:05        mntmn |  if (!(val & DPPLL_CLK_SRC_DSICLK))00:27
blueriseI grepped through the code and thought I don't need to set NON_CONTINUOUS00:28
blueriseDo I need to set it?00:28
mntmndon't set 00:29
mntmnwe need continuous clock00:29
mntmnfound one more thing maybe. we don't do any lane assignment or polarity setting00:29
mntmnregmap_write(pdata->regmap, SN_LN_ASSIGN_REG, pdata->ln_assign);00:30
mntmnregmap_update_bits(pdata->regmap, SN_ENH_FRAME_REG, LN_POLRS_MASK,00:30
mntmn   pdata->ln_polrs << LN_POLRS_OFFSET);00:30
mntmnlets see 00:30
bluerisetrue00:31
bluerisewell SN_ENH_FRAME_REG00:32
bluerisewe set to 0x500:32
mntmninteresting, these are commented out in my dts in linux00:32
mntmn  //data-lanes = <2 1 3 0>;00:32
mntmn  //lane-polarities = <0 1 0 1>;00:33
blueriseheh00:33
mntmnthe plot thickens...00:33
mntmnlane assign reg is 0x5900:34
blueriseinteresting00:34
bluerisewhy is it commented out? :P00:34
mntmn * Read config from the device tree about lane remapping and lane00:36
mntmn * polarities.  These are optional and we assume identity map and00:36
mntmn * normal polarity if nothing is specified.  It's OK to specify just00:36
mntmn * data-lanes but not lane-polarities but not vice versa.00:36
mntmnso linux' default would be 0123 for the lane mapping and 0000 for the polarity00:38
bluerisethat's the default00:39
bluerisein the datasheet00:39
blueriseu-boot=> i2c md 2c 59 100:40
bluerise0059: e4    .00:40
mntmnyep, these are fine.00:41
bluerise11 10 01 0000:41
bluerisesounds about right00:41
mntmnthat's all correct00:41
mntmnthanks00:41
mntmnmipi is also outputting on 4 lanes yes?00:42
mntmni see in your code:00:42
mntmn/* FIXME: somewhere else */00:42
mntmndevice->lanes = 4;00:42
mntmndevice->format = MIPI_DSI_FMT_RGB888;00:42
blueriseheh. I hope it is! ;)00:51
bluerisenwl_enable: min_hs_clock_multiple=0 lanes=400:51
bluerisenwl_dsi_mode_probe: phyref 0 2400000000:51
blueriselanes=4, data_rate=97200000000:51
bluerisenwl_dsi_config_dpi: hbp=80 hfp=40 hsa=4000:51
bluerisenwl_dsi_config_dpi: hbp=226 hfp=113 hsa=110 e00:51
mntmndata rate 972000000?00:53
mntmnsounds a lotr00:53
mntmnlike a lot00:53
- wagga (QUIT: Quit: Client closed) (~wagga@node-1w7jra22ildhwu44kvuq39uhc.ipv6.telus.net)00:53
bluerisehehe, that's bit rate00:53
mntmnmhm00:54
mntmn>>> (972000000/8.0)*400:54
mntmn486000000.000:54
mntmnok there's our 486mhz i guess?00:54
blueriseyup00:55
bluerise                hs_clk_rate *= bpp;00:55
bluerise                do_div(hs_clk_rate, lanes);00:55
mntmnok, i should go home like 2 hours ago :D00:55
blueriseI'm home...00:55
mntmngood for you!!00:55
blueriseall day long :/00:55
mntmnok depends!00:55
bluerisewell, I went to the golf range for half an hour00:55
blueriseah it's 2400:56
bluerise(972 * 4)/2400:56
bluerise16200:56
blueriseRGB888 is apparently 24 bpp00:57
mntmnmakes sense00:57
mntmnIIRC there needs to be some overhead though00:57
mntmnhmm maybe that was on the dp side00:58
chartreuse24bpp though you may have the H and V sync encoded into an extra few bits or byte in there00:59
mntmn /* fudge factor required to account for 8b/10b encoding */01:00
mntmnyeah that seems to be on the DP side01:00
mntmnbit_rate_khz = mode->clock * ti_sn_bridge_get_bpp(pdata);01:02
mntmndp_rate_mhz = DIV_ROUND_UP(bit_rate_khz * DP_CLK_FUDGE_NUM,01:02
mntmn   1000 * pdata->dp_lanes * DP_CLK_FUDGE_DEN);01:02
mntmnhmmm this gives me 3.24ghz01:03
mntmnnot 2.701:04
mntmn162000*8*1000*2*(10/8.0)01:04
mntmnwait, my calculation is off01:05
mntmn>>> 162000*24*10/(1000*2*8.0)01:06
mntmn2430.001:06
mntmnall good01:06
mntmnso 2.7ghz should be fine, but 2.43ghz should work as well01:06
blueriseyou should go home01:08
bluerisetomorrow is another day01:09
mntmnwe could try 0x60 instead of 0x80 for reg 0x94, but i will go home, yes01:09
mntmnvery true01:09
bluerise0x94? let's see01:09
blueriseone last test, sure01:09
mntmnthat's the dp bitrate01:09
mntmnas you have colorbars i don't think that'll fix it01:09
+ alex4nde1 (~alexander@c-76-126-12-166.hsd1.ca.comcast.net)01:09
mntmnbut who knows01:09
blueriseack01:10
blueriselet's try it01:10
bluerisenow it hangs01:11
bluerisedoesn't work01:12
mntmnhaha ok!01:12
mntmnthanks for trying01:12
blueriseno link01:12
blueriseso training or so fails01:12
blueriseok, zzz!01:13
mntmnn8n8n801:48
+ chartreu1e (~chartreus@S0106f0f249dfd9c3.cg.shawcable.net)03:24
- chartreuse (QUIT: Remote host closed the connection) (~chartreus@S0106f0f249dfd9c3.cg.shawcable.net)03:25
- Asmadeus (QUIT: Ping timeout: 272 seconds) (~asmadeus@240b:13:8c80:d300:e:98c:8000:d300)03:25
+ Asmadeus (~asmadeus@240b:13:8c80:d300:e:98c:8000:d300)03:25
- alex4nde1 (QUIT: Ping timeout: 248 seconds) (~alexander@c-76-126-12-166.hsd1.ca.comcast.net)04:04
+ alex4nde1 (~alexander@c-76-126-12-166.hsd1.ca.comcast.net)04:19
- alex4nde1 (QUIT: Ping timeout: 248 seconds) (~alexander@c-76-126-12-166.hsd1.ca.comcast.net)04:26
- chartreu1e (QUIT: Quit: leaving) (~chartreus@S0106f0f249dfd9c3.cg.shawcable.net)04:46
+ chartreuse (~chartreus@S0106f0f249dfd9c3.cg.shawcable.net)04:47
+ alex4nder (~alexander@50-254-129-180-static.hfc.comcastbusiness.net)06:59
- alex4nder (QUIT: Ping timeout: 268 seconds) (~alexander@50-254-129-180-static.hfc.comcastbusiness.net)08:32
- kklimonda (QUIT: Ping timeout: 258 seconds) (sid72883@user/kklimonda)09:44
+ kklimonda (sid72883@user/kklimonda)09:44
- kremlin (QUIT: Ping timeout: 252 seconds) (~kremlin@uglyman.kremlin.cc)10:12
+ kremlin (~kremlin@uglyman.kremlin.cc)10:12
+ wagga (~wagga@node-1w7jra22ildhwu44kvuq39uhc.ipv6.telus.net)12:00
- erlehmann (QUIT: Ping timeout: 248 seconds) (~erle@dynamic-046-114-033-055.46.114.pool.telefonica.de)13:24
+ erlehmann (~erle@dynamic-046-114-036-024.46.114.pool.telefonica.de)13:26
+ arminweigl_ (~arminweig@sourcehut/user/arminweigl)13:44
+ adjtm_ (~adjtm@13.red-88-1-141.dynamicip.rima-tde.net)13:44
+ ggoes_ (~gregf@pool-72-69-95-105.nycmny.fios.verizon.net)13:45
+ blast007_ (~blast@user/blast007)13:45
+ signop_ (~signop@li699-102.members.linode.com)13:47
+ jackhill_ (~jackhill@kalessin.dragonsnail.net)13:47
+ jnerula_ (~jnerula@li1009-93.members.linode.com)13:47
+ wagga79 (~wagga@node-1w7jra22ildhvg0963fkd9csj.ipv6.telus.net)13:49
- wagga (QUIT: Ping timeout: 246 seconds) (~wagga@node-1w7jra22ildhwu44kvuq39uhc.ipv6.telus.net)13:49
* wagga79 -> wagga13:50
+ tarxvf (~tarxvf@mail.tarxvf.tech)13:50
- scops (QUIT: *.net *.split) (~scopstchn@2001:470:69fc:105::8da)13:51
- marex (QUIT: *.net *.split) (~marex@195.140.253.37)13:51
- tarxvf_ (QUIT: *.net *.split) (~tarxvf@mail.tarxvf.tech)13:51
- jackhill (QUIT: *.net *.split) (~jackhill@kalessin.dragonsnail.net)13:51
- Kooda (QUIT: *.net *.split) (~kooda@natsu.upyum.com)13:51
- adjtm (QUIT: *.net *.split) (~adjtm@13.red-88-1-141.dynamicip.rima-tde.net)13:51
- nio (QUIT: *.net *.split) (~nio@2001:470:69fc:105::172d)13:51
- blast007 (QUIT: *.net *.split) (~blast@user/blast007)13:51
- arminweigl (QUIT: *.net *.split) (~arminweig@sourcehut/user/arminweigl)13:51
- jnerula (QUIT: *.net *.split) (~jnerula@li1009-93.members.linode.com)13:51
- frank2 (QUIT: *.net *.split) (~frank@91.229.143.153)13:51
- jomo (QUIT: *.net *.split) (~jomo@user/jomo)13:51
- jfred (QUIT: *.net *.split) (quassel@libera/sponsor/jfred)13:51
- signop (QUIT: *.net *.split) (~signop@li699-102.members.linode.com)13:51
- fsx (QUIT: *.net *.split) (~fsx@durian.61924.nl)13:51
- mtm (QUIT: *.net *.split) (~mtm@c-174-58-99-93.hsd1.fl.comcast.net)13:51
- ggoes (QUIT: *.net *.split) (~gregf@fsf/staff/ggoes)13:51
- mntmn (QUIT: *.net *.split) (~mntirc@softboy.mntmn.com)13:52
- BluRaf (QUIT: *.net *.split) (bluraf@junkcc.net)13:52
- schneider (QUIT: *.net *.split) (~schneider@irc2.xtort.eu)13:52
- dodo (QUIT: *.net *.split) (~dodo@user/dodo)13:52
- marlun (QUIT: *.net *.split) (~marlun@46.101.246.245)13:52
- XgF (QUIT: *.net *.split) (~quassel@2001:19f0:5001:1174:dbf:47a1:9a3f:787d)13:52
- q66 (QUIT: *.net *.split) (~q66@q66.moe)13:52
* arminweigl_ -> arminweigl13:52
+ scops (~scopstchn@2001:470:69fc:105::8da)13:53
+ marex (~marex@195.140.253.37)13:53
+ Kooda (~kooda@natsu.upyum.com)13:53
- erlehmann (QUIT: Quit: Just say no, then the virus can not enter your body without your consent.) (~erle@dynamic-046-114-036-024.46.114.pool.telefonica.de)13:55
- scops (QUIT: Ping timeout: 245 seconds) (~scopstchn@2001:470:69fc:105::8da)13:56
- indefini[m] (QUIT: Ping timeout: 252 seconds) (~indefinim@2001:470:69fc:105::1e2a)13:56
- cryptix (QUIT: Ping timeout: 252 seconds) (~cryptxxma@2001:470:69fc:105::94a)13:56
+ erlehmann (~erle@dynamic-046-114-036-024.46.114.pool.telefonica.de)13:56
+ XgF (~quassel@2001:19f0:5001:1174:dbf:47a1:9a3f:787d)13:57
+ q66 (~q66@q66.moe)13:57
+ nio (~nio@2001:470:69fc:105::172d)13:57
+ mtm (~mtm@c-174-58-99-93.hsd1.fl.comcast.net)13:57
+ mntmn (~mntirc@softboy.mntmn.com)13:57
+ BluRaf (bluraf@junkcc.net)13:57
+ schneider (~schneider@irc2.xtort.eu)13:57
+ dodo (~dodo@user/dodo)13:57
+ marlun (~marlun@46.101.246.245)13:57
+ frank2 (~frank@91.229.143.153)13:58
+ jomo (~jomo@user/jomo)13:58
+ jfred (quassel@libera/sponsor/jfred)13:58
+ fsx (~fsx@durian.61924.nl)13:58
- nio (QUIT: Ping timeout: 256 seconds) (~nio@2001:470:69fc:105::172d)14:01
- jryans (QUIT: Ping timeout: 268 seconds) (~jryans@2001:470:69fc:105::1d)14:01
+ cryptix (~cryptxxma@2001:470:69fc:105::94a)14:13
+ indefini[m] (~indefinim@2001:470:69fc:105::1e2a)14:16
+ scops (~scopstchn@2001:470:69fc:105::8da)14:20
- wiedi (QUIT: Read error: Connection reset by peer) (~wiedi@37.228.191.159)14:22
* ggoes_ -> ggoes14:35
- ggoes (QUIT: Changing host) (~gregf@pool-72-69-95-105.nycmny.fios.verizon.net)14:35
+ ggoes (~gregf@fsf/staff/ggoes)14:35
+ nio (~nio@2001:470:69fc:105::172d)14:39
* blast007_ -> blast00714:49
- S0rin (QUIT: Ping timeout: 240 seconds) (~S0rin@user/s0rin)15:07
+ odnes (~odnes@ppp141237116116.access.hol.gr)15:08
+ S0rin (~S0rin@user/s0rin)15:08
- odnes (QUIT: Read error: Connection reset by peer) (~odnes@ppp141237116116.access.hol.gr)15:09
+ jryans (~jryans@2001:470:69fc:105::1d)15:36
* jackhill_ -> jackhill16:13
- khm_ (QUIT: Ping timeout: 248 seconds) (~kfx@wopr.sciops.net)16:21
bluerisemntmn: everyday I appreciate that nice laptop design16:22
blueriseit's a bit bulky, but it's still awesome16:22
mntmnhappy to hear!16:23
blueriseI wish the side panels were a bit more sturdy, like metal instead of 3d printed16:23
blueriseon the other hand, having it plastic shows that it's easily replaceable, which is nice16:23
bluerisenow if only we got that MIPI running, right16:24
mntmnyes, and yes16:27
mntmnit should be easy to get them made from lasercut steel or something16:27
mntmnthe drawings are in the repo16:27
bluerise:)16:32
qbitmm17:10
qbitmntmn: the boards with more ram, are they going to be the defacto board at some point? or is it an 'upgrade' kinda situation?17:10
qbittrying to budget for 2022 :D17:11
+ cwebber (~user@user/cwebber)17:18
cwebberoh hello17:18
cwebbermntmn: got our reform and holy shit that boot-up menu17:18
cwebberit's so cool17:18
cwebberhonestly the whole thing is beautiful17:19
cwebbertbh I probably won't start using it seriously until I have GuixSD running on here and I might be waiting for someone else (jackhill? :)) to do that first17:20
technomancycwebber: yessss you got yours =D17:22
cwebbertechnomancy: yes it's so beautiful17:22
cwebberalso hi technomancy :)17:22
cwebbertechnomancy: we haven't chatted as much lately and I've kind of missed it tbh!17:23
jackhillACTION waves17:23
technomancyyeah same! hoping to get a chance to try out some of the spritely stuff you've been working on soon.17:23
cwebbertechnomancy: :D17:24
cwebbertechnomancy: I am currently porting Goblins to Guile in "phases"17:24
cwebbersaving each milestone17:24
cwebberwhich means porting to eg fennel may be very possible17:24
technomancycwebber: is this about switching your platform or more about having multiple working implementations that will both be kept active?17:24
technomancymwahaha love to hear it17:25
mntmncwebber: thank you!17:27
mntmnqbit: sorry, no idea17:27
mntmn(yet)17:27
qbitkew17:27
mntmnwe need to test them first for a while to be able to gauge how it works in practice17:27
qbitmakes sense 17:28
mntmnthe ls1028a is currently in the ddr4 calibration phase17:28
cwebbertechnomancy: multiple implementations, at least Guile and Racket will both be active17:28
mntmnapparently they made it boot from sd card but now it's all about getting the ram parameters right and tested. then i will get the module17:28
cwebberit's partly that I want to demonstrate distributed programming with CapTP across different language runtimes17:29
cwebberyou could say two schemes is kind of cheating17:29
jackhillmntmn: have you had any more thoughts on a "desktop" case? 17:29
technomancycwebber: awesome. too many social protocols out there with only one implementation; always feels like a huge red flag (*cough*SSB*cough*)17:29
cwebbertechnomancy: yes I want to avoid that17:29
mntmnjackhill: yeah. would be interesting i guess. could replace my old NAS with something passive cooled...17:29
cwebbertechnomancy: also we're working on interoperability with JS with Agoric's stuff and also I should stop spamming this channel about spritely things17:29
technomancyhaha, unless you want to talk about how your reform is allowing you to make spritely so much better17:30
mntmncwebber: it's ok, sounds interesting ^^17:30
cwebberwell I do want to hack on spritely from the reform :)17:30
cwebberthat's a serious goal there17:30
mntmncool17:30
technomancycode that is written on free hardware is 15 to 30% more freedom-effective, don't you know?17:31
cwebbertechnomancy: obviously ;)17:31
cwebbertechnomancy: I'm going to buy "artisnal organic produce" stickers in bulk, transmute them into pure data, and transmute them right onto every release shipped to every user17:32
cwebberand yet somehow have the same hash every time17:32
cwebberartisnal AND reproducible obviously17:32
jackhillmntmn: just thinking if I end up with multiple compute modules, I'll of course want to run them all at once :)17:33
+ darth-cheney (~darth-che@2603-7000-8d00-1f72-20a1-b00f-18b2-57b6.res6.spectrum.com)18:02
darth-cheneyYo, I'm having some trouble following the instructions in this community thread: wget http://mntre.com/reform_md/flash-rescue-reform-init.bin18:02
darth-cheneywhoops, I mean: https://community.mnt.re/t/operating-system-on-nvme-without-sd-card/11018:03
darth-cheneyNo matter what I do I am unable to boot from either the nvme or the sd card. I'm always in emmc18:03
mntmnjackhill: yes, same18:06
mntmndarth-cheney: what are you trying to do? boot nvme via emmc?18:07
darth-cheneymntmn: Yes, I was originally but I couldn't get that to work 18:07
darth-cheneySo now I'm trying to just get back to the sd card and that's a no go too18:07
mntmndarth-cheney: you have to switch the switch back18:08
mntmnto go back to SD18:08
mntmnthe one on the module18:08
darth-cheneyoh I see18:08
darth-cheneyso emmc can only boot into itself, nvme, and I guess a usb device, right?18:09
darth-cheney(I'm thinking I may not have migrated correctly and that's why nvme won't boot)18:09
+ odnes (~odnes@ppp141237116116.access.hol.gr)18:11
- odnes (QUIT: Read error: Connection reset by peer) (~odnes@ppp141237116116.access.hol.gr)18:11
+ wiedi (~wiedi@37.228.191.159)18:21
blueriseonce I have my u-boot working with LCD, I'll probably replace the eMMC content so I don't need the SD card anymore18:21
+ khm (~kfx@wopr.sciops.net)18:23
darth-cheneyok back in action on the sd card, thanks. I'll try the full migration with boot some other time18:28
mntmndarth-cheney: the answer to your question is yes18:30
darth-cheneymntmn: thanks18:31
- qbit (QUIT: Quit: WeeChat 3.2) (~qbit@foof.suah.dev)19:28
+ qbit (~qbit@foof.suah.dev)19:36
- darth-cheney (QUIT: Ping timeout: 272 seconds) (~darth-che@2603-7000-8d00-1f72-20a1-b00f-18b2-57b6.res6.spectrum.com)19:52
mtmACTION zooms in and tries to read the notes, hoping to see his name: https://mastodon.social/@mntmn/10673139705210613821:40
mntmnhaha, no names on that21:46
technomancyI was kind of surprised that booting off the SSD instead of the SD card was considered an advanced topic in the manual22:11
mntmnwell, it was untested at the time23:15
mntmnor, not really untested but it was still kind of new23:16
mntmnthere was also a point where i thought dual display + nvme cannot work together, was a real headache23:16
technomancyHDMI needs to die in a fire23:17
mntmnand the fix looks like this... https://source.mnt.re/reform/reform-system-image/-/blob/main/reform2-imx8mq/template-kernel/patches/mnt4001-lcdif-fix-pcie-interference.patch#L1823:17
- adjtm_ (QUIT: Ping timeout: 268 seconds) (~adjtm@13.red-88-1-141.dynamicip.rima-tde.net)23:24
+ adjtm (~adjtm@229.red-81-36-31.dynamicip.rima-tde.net)23:28

Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!