apolkosnik[m] | 1.12 | 00:35 |
---|---|---|
apolkosnik[m] | Let me switch to CS MK2 | 00:36 |
apolkosnik[m] | Same as above | 00:39 |
minute | apolkosnik[m]: i mean which version of the fw... default z3? | 00:40 |
apolkosnik[m] | I hope it's not due to MMUlibrary settings | 00:40 |
apolkosnik[m] | Yes Z3 version on A4k | 00:40 |
- S0rin (QUIT: Ping timeout: 268 seconds) (~S0rin@user/s0rin) | 00:44 | |
minute | hmm maybe i messed something up, will test tomorrow | 00:44 |
minute | i tested with cs mk2 in a3000 | 00:44 |
+ S0rin (~S0rin@user/s0rin) | 01:02 | |
- WoC (QUIT: Read error: Connection reset by peer) (WoC@2603-8080-cd40-00c6-ccd7-6745-059a-e424.res6.spectrum.com) | 03:04 | |
+ gdonner (gdonner@c-98-253-93-116.hsd1.in.comcast.net) | 04:18 | |
gdonner | Am I correct in assuming that AXMP3 in FW/drivers 1.12 is supposed to revert to v1.11 instead of the v1.12 it was in the 1.12 beta? | 04:19 |
gdonner | Also, ZZ9000Net.device is 6,936 bytes over 1.11's 4,252 bytes; but they're both still v1.2 (02/07/20)? | 04:21 |
gdonner | Not to be nit picky, but it could be helpful to embed a VER$ string in ZZTop :) | 04:23 |
gdonner | OK; I'm done. :) | 04:23 |
gdonner | Other than that, after updating to 1.12, I got a lockup when trying to play an MP3 in AmigaAMP... played the MP3 for a few seconds, then stopped, and sent Workbench into a deep lockup | 04:25 |
gdonner | i.e. using Ctrl-LAmiga-RAmiga couldn't reboot the system | 04:25 |
gdonner | I'm still trying to figure out what's causing this weird gfx issue (and occasional lockup) when running IBrowse: | 04:28 |
gdonner | https://www.gregdonner.org/transfer/IBrowse_gfx_issue_crash.mp4 | 04:28 |
gdonner | Might be IBrowse itself of course, but could also be ZZ9000 FW or maybe P96? | 04:29 |
gdonner | Happens on both my A4000 and my A3000 | 04:30 |
- gdonner (QUIT: ) (gdonner@c-98-253-93-116.hsd1.in.comcast.net) | 04:31 | |
- S0rin (QUIT: Ping timeout: 268 seconds) (~S0rin@user/s0rin) | 04:32 | |
+ S0rin (~S0rin@user/s0rin) | 04:34 | |
WoC` | Remind me in about two weeks and I'll check on my 4000 | 05:27 |
- lastebil (QUIT: *.net *.split) (truck@shell.suomiscene.fi) | 05:48 | |
+ lastebil (truck@shell.suomiscene.fi) | 05:48 | |
+ xet7 (~xet7@user/xet7) | 11:59 | |
- WoC` (QUIT: Remote host closed the connection) (woc@2603-8080-cd40-00c6-beef-beef-beef-beef.res6.spectrum.com) | 12:27 | |
+ WoC` (woc@2603-8080-cd40-00c6-beef-beef-beef-beef.res6.spectrum.com) | 12:28 | |
minute | running some tests now | 13:31 |
minute | i can't see any problem with the zorro 3 fw and driver | 13:39 |
minute | for me it is even faster than 1.11 in p96speed | 13:40 |
minute | unfortunate that noone is online when i'm actually at work | 14:24 |
Claude | . | 15:05 |
minute | ! | 15:11 |
minute | bought new p96, rerunning tests | 15:12 |
minute | works fine still | 15:50 |
- WoC` (QUIT: Remote host closed the connection) (woc@2603-8080-cd40-00c6-beef-beef-beef-beef.res6.spectrum.com) | 16:10 | |
minute | i think the problem manifests only on a4000 | 16:13 |
+ WoC (WoC@2603-8080-cd40-00c6-48c5-c981-4062-9f10.res6.spectrum.com) | 16:25 | |
minute | trying to set up my a4000, but i only get a short flash every ~5 seconds on the RGB out | 16:44 |
minute | with random color, only a few scanlines | 16:44 |
minute | guess i'm gonna burn some diagroms | 16:45 |
WoC | heh, making it sound like some Voodoo ritual ;) | 17:02 |
Jope | a big box amiga is a harsh mistress | 17:05 |
Jope | minute, what to look out for with the latest fw + a4000? | 17:05 |
minute | Jope: is it slow? i.e. low rectfill numbers in p96speed | 17:06 |
minute | Jope: normal is >10000 op/s, abnormal is <1000 | 17:06 |
WoC | Anyone ported glmark yet ? | 17:08 |
minute | ? | 17:13 |
WoC | would be nice to see some glmark2 benchmarks on the zz9000 | 17:14 |
Jope | minute, ok moment.. I'll upgrade | 17:14 |
WoC | (OpenGL benchmarks) | 17:14 |
WoC | Anyhow, I will try to build it... | 17:15 |
minute | WoC: why? zz doesn't support opengl | 17:16 |
WoC | oh | 17:17 |
WoC | didn´t know that | 17:18 |
minute | it is 2D only | 17:19 |
minute | damn EPROMs, half an hour in UV light was not enough :/// | 17:19 |
js | time for EEPROMs? 😄 | 17:20 |
WoC | ok, that´ll be interesting... maybe i can port mesa (and use fgpa for the 3D) | 17:20 |
minute | yeah but i don't have a flash adapter thing (yet), perhaps i should get one | 17:20 |
js | but yeah it's funny how everyone is "It's all gone as soon as you remove the label" | 17:20 |
minute | recommendations? | 17:20 |
js | but then when yoiu want to erase it it takes forever | 17:20 |
minute | WoC: good luck! | 17:20 |
js | minute: I think MiniPro is the standard, pretty much | 17:20 |
WoC | ty minute, worth a try | 17:20 |
js | it's pretty much the only thing I used for programming so far | 17:21 |
minute | js: i meant flash rom solution for amiga | 17:21 |
js | ah, I think the problem is that there are no pin-compatible EEPROMs ☹️ | 17:21 |
js | still EPROM for me, too | 17:21 |
minute | https://www.onyxsoft.se/flashrom.html | 17:22 |
js | but it sure is annoying | 17:22 |
js | oh nice | 17:22 |
js | next step: Replace the EPROM with an RPi that can swap ROM on the fly? 😄 | 17:22 |
Jope | noo | 17:23 |
js | I mean, how many RPis can you get into one Amiga? 🙂 | 17:23 |
WoC | Raspberry Jam ? | 17:23 |
minute | you don't need a full computer just to swap in some data | 17:24 |
Jope | minute, which screenmode are you benchmarking with | 17:25 |
minute | 640x480x8 | 17:25 |
js | minute: but then you can cross-compile and "flash" the new ROM via SSH 😉 | 17:25 |
js | extremely nice for debugging | 17:25 |
minute | ok, got the HI rom of diagrom burned | 17:25 |
Jope | rectal fill at 13590 | 17:26 |
Jope | (höhö) | 17:26 |
minute | ? | 17:27 |
Jope | a4000, csmk2 060/50, os3.2.1, 1.12 zorro3 nofast firmware | 17:27 |
minute | ah | 17:27 |
minute | well that's good! | 17:27 |
Jope | might as well run all the tests while I'm at it | 17:27 |
minute | i wonder what the hell is different with your a4000 compared to all the others who are seeing that super low number | 17:27 |
Jope | I don't have that many expansions left, plucked most out after the zz9000 came along | 17:27 |
Jope | now it's just the zz, ax and gvp ioextender | 17:28 |
minute | midwan also removed all other cards | 17:28 |
Jope | csmk2 + cybscsimk2 in the fast slot | 17:28 |
minute | maybe a software interaction after all? mmulib or sth? | 17:28 |
minute | 68060.library version? something like that | 17:28 |
Jope | I'll check which 060 lib I have | 17:28 |
minute | thanks | 17:29 |
Jope | probably mmulib in this os install, since I decided to try to be as thor approved as possible for a chance | 17:29 |
minute | ok | 17:29 |
Jope | + I was getting a bit weary of random crashes due to my 20 year old patched-to-eternity 3.9 setup | 17:29 |
minute | yeah | 17:29 |
Jope | held out over 3.1.4 but when 3.2 came along I began from scratch | 17:29 |
Jope | rebooting, didn't want to run through all of p96speed .. | 17:30 |
Jope | oh! I actually have the phase 5 060 lib installed | 17:30 |
Jope | 46.15 by ralph schmidt | 17:30 |
Jope | mulibs are installed however. this might have been a conscious decision after doing a bit of research on the subject | 17:31 |
Jope | ok, let's try thor's 060 lib then | 17:33 |
Jope | from 2018 | 17:33 |
Jope | 14137 ops/s! | 17:34 |
Jope | I gained 600 rectangles just like that | 17:34 |
minute | huh. | 17:48 |
Jope | axmp3 doesn't seem to do much more than write dots to my screen | 18:04 |
Jope | didn't try it before | 18:04 |
minute | forget about axmp3 for now | 18:05 |
minute | the included version is probably wrong/old | 18:05 |
Jope | aight | 18:06 |
Jope | not a very exciting use case for me in any case :-D | 18:06 |
Jope | ahi works fine otherwise.. | 18:06 |
Jope | been messign around with sined + hippo player and they seem to play happily | 18:06 |
+ WoC` (woc@2603-8080-cd40-00c6-beef-beef-beef-beef.res6.spectrum.com) | 18:14 | |
minute | Jope: did you set up MHI already? | 18:18 |
minute | ah, you said not exciting | 18:18 |
Jope | I didn't. that might be the next thing then | 18:19 |
apolkosnik[m] | Delitracker wants a real-time mixer or something like that, it won't play ahi without user intervention in any case | 18:19 |
Jope | mhiplay works | 18:23 |
Jope | a little crackle in the beginning but playback seems fine, albeit a bit filtered | 18:24 |
Jope | now there's no crackle. I was playing some xm via ahi previously | 18:25 |
WoC` | Is there any Documentations that I can read up on for the FPGA portion of the ZZ9000 ? My idea is to define the OpenGL functions in FPGA and map them from mesa... | 18:27 |
Jope | amigaamp3-m68k via mhi also seems happy | 18:28 |
minute | the crackle is a bug, it's playing the initial buffer which might not be filled yet, i think | 18:57 |
Jope | aight | 18:58 |
+ gdonner (~gdonner@c-98-253-93-116.hsd1.in.comcast.net) | 19:03 | |
gdonner | Ah; I was wondering about that crackle--good to know! | 19:03 |
gdonner | But yeah, I suspect some A4000s may just not play well with other Amiga hardware. | 19:04 |
gdonner | A GVP T-Rex I bought some time ago, refused to boot on my system (power light always on, black screen), so I sent it to TheBajaGuy, and it booted up straight away without any adjustments. Go figure. | 19:05 |
gdonner | I'm holding out for the TF4060 (until the Warp 3060/4060 is available in the distant future). | 19:06 |
gdonner | Just as well on the T-Rex, as it's a fairly rare card, and I can't think of a better person than TheBajaGuy (ex-GVP tech support) to own it and use it. | 19:07 |
gdonner | There was a rare and little-known issue with some early A4000s that may have a wrong speed crystal installed in them | 19:08 |
gdonner | Namely, U104--it's supposed to be 50 MHz; some early A4000s had a 28 MHz crystal erroneously installed | 19:10 |
gdonner | (the crystal near the ROMs) | 19:14 |
gdonner | Thankfully, mine is 50 MHz :) | 19:15 |
minute | gdonner: interesting | 19:16 |
minute | i finally have diagroms burned. | 19:16 |
minute | 1084s can't sync to the rgb output of the machine | 19:19 |
minute | so a4000 is not generating a proper sync | 19:19 |
WoC | is that using the DB23 out from the Amiga ? | 19:39 |
minute | WoC: yes | 19:41 |
WoC | oh, never had that issue with my 1084s back in the day, but then I also made the cable | 19:42 |
WoC | iirc I traded that 1084s for some sony screen with much sharper picture | 19:45 |
minute | well, something is broken on the A4000 motherboard and i'm trying to figure out the issue, | 19:54 |
minute | gonna scope some signals | 20:09 |
apolkosnik[m] | That's what the diagroms were for? | 20:12 |
apolkosnik[m] | My last stupid mistake was spraying DeOxit on the A1200 expansion slot | 20:14 |
minute | well, the diagrom doesn't work | 20:15 |
minute | the a4000 is too broken to run it | 20:15 |
minute | ok so SCLK behaves randomly | 20:15 |
minute | it goes up and down every few seconds with seemingly random intervals | 20:16 |
apolkosnik[m] | Voltages are ok? | 20:17 |
minute | 5V is ok | 20:19 |
minute | sounds a bit like > Certainly replace U975/U976 if there has been any visible corrosion around these shift registers. I've seen faults where the computer fails to boot because the MDAT line from U975 to Lisa/U450 is floating which causes U450 not to generate certain clocks and the system gets stuck in this weird half-reset state at power on | 20:19 |
apolkosnik[m] | I got two A3000T PSUs for recap the other day, while one is shorting out, the other one has voltages going up and down | 20:20 |
apolkosnik[m] | Seems like the usual problem on Varta plagued A4Ks | 20:24 |
apolkosnik[m] | The MDAT floating, I meant | 20:24 |
minute | aha? | 20:29 |
minute | i measured it though and it doesn't look floating, it's more like 0V | 20:29 |
minute | what's MDAT carrying? sounds like mouse data | 20:29 |
gdonner | I've always thought that "Varta" should have had an "F" and not a "V" :) | 20:30 |
gdonner | For all the damage and havoc they've caused to our beloved Amigas | 20:30 |
gdonner | It recently sunk in that in the tower versions, the leakage damage is much worse, because the mobo is mounted vertically... dripping down all kinds of traces and components | 20:32 |
gdonner | nasty stuff; thanks so much, C= | 20:32 |
minute | ah, > These addresses each read a 16 bit register. These in turn are loaded | 20:32 |
minute | from the MDAT serial stream and are clocked in on the rising edge of | 20:32 |
minute | SCLK | 20:33 |
minute | http://amiga-dev.wikidot.com/hardware:joy0dat | 20:33 |
minute | C28M is high all the time | 20:34 |
minute | actually, there are some traces of activity on C28M a few times after cold start | 20:39 |
minute | it's like the clock wants to start sometimes, but can run only for a moment | 20:41 |
minute | aha! | 20:43 |
minute | the cpu connector is warped and was not connecting properly on one side | 20:44 |
minute | now i have a C28M | 20:44 |
minute | okay and i have output on the screen, blinking colors and stuff | 20:45 |
minute | probably should put in the chipram now | 20:45 |
minute | ok DiagROM works!! | 20:46 |
Jope | yay | 20:51 |
minute | ok so according to a1k users, the slowness happens when you have no network cable connected to zz9000! | 20:51 |
Jope | intriguing | 20:51 |
minute | connecting the network cable makes it fast | 20:51 |
minute | can someone here confirm this? | 20:51 |
Jope | moment | 20:51 |
Jope | first boot: start up with network cable connected, have the system load the network driver as part of the boot up, disconnect cable. rectfill 14406 | 20:53 |
Jope | now reboot with network cable disconnected, but with network loading on bootup: rectfill 15361 | 20:55 |
Jope | now reboot with network cable disconnected, network startup commented out: 15093 | 20:56 |
Jope | now cold boot with network cable disconnected, network startup commented out: ui seems sluggish, I can see the redraws. result: a paltry 667 | 20:58 |
Jope | plugged the network cable in, no reboot: 14782 | 20:59 |
Jope | can't be bothered to test any more permutations | 20:59 |
minute | perfect | 21:02 |
minute | thanks jope | 21:02 |
Jope | np | 21:02 |
minute | 100% confirmed then. | 21:02 |
- buckket (QUIT: Remote host closed the connection) (~buckket@pdp8.buckket.org) | 21:10 | |
- xet7 (QUIT: Remote host closed the connection) (~xet7@user/xet7) | 21:11 | |
+ buckket (~buckket@pdp8.buckket.org) | 21:15 | |
- buckket (QUIT: Quit: buckket) (~buckket@pdp8.buckket.org) | 21:30 | |
+ buckket (~buckket@pdp8.buckket.org) | 21:41 | |
+ mntamiga (~amiga3000@p54a435d6.dip0.t-ipconnect.de) | 22:02 | |
mntamiga | ok ok | 22:02 |
mntamiga | fw12.1 seems snappy | 22:02 |
mntamiga | can chat and listen to mp3 with 10-20% cpu usage | 22:04 |
mntamiga | openSSL accel on ZZ would be really nice tbh | 22:05 |
mntamiga | (or any TLS accel) | 22:05 |
apolkosnik[m] | I'm surprised that the upload speed is 3x the download speed when tested with AmigaSpeedTest | 22:06 |
minute | this is expected | 22:07 |
minute | receiving packets is more processing/queueing things up for the amiga then just sending packets | 22:07 |
apolkosnik[m] | And still, the 3c589 pcmcia card is way faster | 22:07 |
minute | define "way" | 22:08 |
apolkosnik[m] | Maybe it has some special sauce | 22:08 |
apolkosnik[m] | Let me find the comparisons I did | 22:09 |
mntamiga | i'm testing with amispeedtest now, avg download speed 3298 Kb/s | 22:09 |
mntamiga | upload 6845 Kb/s | 22:10 |
mntamiga | i guess these are kilobits? | 22:10 |
mntamiga | for me downloading from aminet in ibrowse gets faster after a while. it's not super great but i'm not sure what amigaos/roadshow/060 can handle | 22:11 |
apolkosnik[m] | ACTION uploaded an image: (138KiB) < https://libera.ems.host/_matrix/media/r0/download/matrix.org/IVPwUXKsymJgQijqmxNdauIX/Screenshot_20221019-161114.png > | 22:12 |
minute | apolkosnik[m]: that link gives me 502 bad gateway | 22:12 |
apolkosnik[m] | Ugh | 22:13 |
apolkosnik[m] | Discord is so much easier | 22:13 |
mntamiga | just use pastebin | 22:13 |
minute | https://mister-muffin.de/post.htm | 22:13 |
minute | browse for the file and click post | 22:13 |
apolkosnik[m] | https://photos.app.goo.gl/7N78DFboZg5Wdyru7 | 22:14 |
minute | apolkosnik[m]: upload speed looks similar, same kind of variations, download speed is almost double yeah? | 22:15 |
minute | wait, those are both pcmcia cards? | 22:15 |
apolkosnik[m] | Yes | 22:15 |
minute | i thought you were comparing to zz | 22:15 |
apolkosnik[m] | Both have code on aminet | 22:15 |
apolkosnik[m] | Cnet is asm | 22:16 |
minute | with zz i get basically the same speed like cnet | 22:16 |
minute | except faster upload | 22:16 |
apolkosnik[m] | 3c is C | 22:16 |
minute | but yeah, might be worth to take a look at the code, if there is some special trick | 22:17 |
apolkosnik[m] | The upload speed got better with 1.12, I think | 22:17 |
minute | FW 1.12.1 https://mntre.com/media/ZZ9000_info_md/zz9000-firmware-1_12_1-zorro3.zip | 22:20 |
minute | this should fix the speed problems | 22:20 |
minute | with network unplugged etc | 22:20 |
- mntamiga (QUIT: Read error: Connection reset by peer) (~amiga3000@p54a435d6.dip0.t-ipconnect.de) | 23:20 | |
- gdonner (QUIT: ) (~gdonner@c-98-253-93-116.hsd1.in.comcast.net) | 23:46 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!