| + gdonner (gdonner@c-98-253-93-116.hsd1.in.comcast.net) | 01:00 | |
| gdonner | Re: Super Buster revs: Yeah, my A3000 is one of the early 16 MHz '030s, and it had a rev. 7 Super Buster in it. One of the first things I did when upgrading it was replace it with a rev. 11 chip. | 01:01 |
|---|---|---|
| gdonner | My A4000 had a rev. 9; also got swiftly replaced with a rev. 11 (among other upgrades). | 01:03 |
| gdonner | If socketed, very easy to upgrade with the right chip puller :) | 01:04 |
| - gdonner (QUIT: ) (gdonner@c-98-253-93-116.hsd1.in.comcast.net) | 01:11 | |
| + mnemotron (mnemotron@194.135.47.239) | 08:07 | |
| mnemotron | good morning | 08:07 |
| + CottonUndies (~CottonUnd@2405:6580:33c0:b400:3884:240d:eb9a:34e6) | 08:13 | |
| CottonUndies | mnemotron good day to you | 08:13 |
| mnemotron | hey there | 08:13 |
| mnemotron | did you get your problems sorted out? | 08:14 |
| CottonUndies | minute I think I've found the culprit for my freezes. The happy outcome of non-stop testing...... | 08:14 |
| CottonUndies | I'm pretty sure I know whats causing ATK to freeze between 1.10FW and 1.11FW. | 08:14 |
| CottonUndies | I'm not sure if its documented in the release notes, but the closest I can find that associates with my fix for the issue is....... none other than... the first and foremost on the list!!!! | 08:16 |
| CottonUndies | "Improves/Fixes ZZ9000AX playback on many Amiga models through improved IRQ behavior (also improves Ethernet)" | 08:16 |
| CottonUndies | I say this, becuase if I comment out of my s:network-startup "AddNetInterface DEVS:NetInterfaces/~(#?.info) QUIET , then ATK no longer freezes. The only device in that location is ZZ9000Net | 08:17 |
| CottonUndies | This also might explain while Ibrowse freezes occasionally while scrolling. | 08:18 |
| mnemotron | Interesting! | 08:23 |
| mnemotron | Did you also try to disable the cpu data cache like on my 3000? | 08:23 |
| mnemotron | I only start networking with MiamiDX when I need it | 08:23 |
| CottonUndies | No, as I now know this is the problem for me I need to see how the boss views this, as the difference between 1.10 and 1.11 was to also "improve Ethernet" "through improved IRQ behavior". So I'm sure this isn't "my problem" per say. Also, in a properly working system, I would never want to turn off any CPU caching, the performance hit is | 08:26 |
| CottonUndies | significant. | 08:26 |
| mnemotron | Yeah, of course. I meant it only for double-checking | 08:27 |
| mnemotron | or cross-checking | 08:27 |
| CottonUndies | BTW, you say you only start networking when you need it, but in your startup when/where do you eventually do the ""AddNetInterface DEVS:NetInterfaces/~(#?.info) QUIET " as I assume you need to to that as well at some point? | 08:28 |
| mnemotron | i don't think so. it's all in MiamiDX | 08:31 |
| mnemotron | I start the MiamiDX gui | 08:31 |
| CottonUndies | Without starting MiamiDX does Sysinfo show under DEVICES ZZ9000Net V1.2? | 08:37 |
| mnemotron | wait | 08:37 |
| mnemotron | when i look in ShowConfig it only lists zzusb.device under Devices | 08:38 |
| CottonUndies | Do you have Sysinfo? What argurments do you use for SHowConfig? | 08:39 |
| mnemotron | no arguments with ShowConfig. it's in Tools | 08:40 |
| mnemotron | hm, i should have sysinfo somewhere | 08:40 |
| mnemotron | ah. Work: Download/SysInfo :) | 08:41 |
| CottonUndies | My Showconfig doesn't show that level of detail.... my version is 45.2 | 08:41 |
| mnemotron | i have OS 3.2.1 | 08:41 |
| CottonUndies | that might explain it, I think I'm 3.2 | 08:41 |
| mnemotron | sysinfo shows zzusb V43.0 in Devices | 08:41 |
| CottonUndies | ok, well, the bad news is I don't have a solution for you. The good news is you don't need to be concerned about my issue potentially being your issue, as I assume your testing/crashing happens without you loading MiamiDX, and therefore at that time the device. | 08:43 |
| mnemotron | sysinfo: https://i.imgur.com/f9EZ6Nk.jpg | 08:45 |
| mnemotron | showconfig https://i.imgur.com/tT1Wrqh.jpg | 08:45 |
| mnemotron | CottonUndies: yep | 08:45 |
| CottonUndies | Ohhh, very nice looking showconfig! | 08:46 |
| mnemotron | indeed | 08:54 |
| minute | CottonUndies: i see, good investigation | 08:56 |
| _Bnu | Also yes, IRQ problems are the main reason for Amiga Test Kit locking up or crashing to the debug trace screen. | 08:56 |
| minute | this is why i did the "weird" old IRQ behavior where the interrupt was just a pulse, because sometimes, things would hang | 08:57 |
| _Bnu | But the AX IRQ shouldn't be active unless the driver has enabled it. | 08:57 |
| minute | it's not AX, it's ethernet | 08:57 |
| _Bnu | Ah, I see. | 08:57 |
| minute | i have rewritten how both of the interrupts work | 08:58 |
| minute | (in 1.11) | 08:58 |
| minute | for example, if something like ATK prevents interrupts from being handled, but it doesn't shut down ethernet before, this makes sense | 09:00 |
| _Bnu | I don't know much about what Amiga Test Kit does if you launch it from Workbench, never tried. | 09:02 |
| _Bnu | I always boot it from floppy to avoid problems with random libraries that may be loaded. | 09:02 |
| mnemotron | any quick idea why my a3000 is 0.88 as fast as an a3000? https://i.imgur.com/w4A3gLJ.jpg | 09:06 |
| CottonUndies | Could this also be why I get freezes when scrolling in IBrowse | 09:06 |
| CottonUndies | ? | 09:06 |
| mnemotron | i enabled the datacache for this. without it was 0.77 | 09:06 |
| minute | CottonUndies: well, you can test that by opening a website, then calling c:netshutdown in a shell while keeping the browser open, and then scroll the website a lot | 09:08 |
| _Bnu | mnemotron: Probably because you're using the physical Kickstart ROM rather than having it mapped to Fast RAM. | 09:32 |
| _Bnu | It slows things down quite a bit. | 09:33 |
| mnemotron | oh? let me enable fastrom then... sec | 09:33 |
| CottonUndies | minute It seemed better for a moment, then froze. One thing I did check before scrolling the 2nd time was check the device in Sysinfo, and it was still showing up, even though I had issued c:netshutdown | 09:35 |
| mnemotron | haha, now it's slower... only 0.80 instead of 0.88 | 09:35 |
| CottonUndies | even the locally loaded iBrowse splash screen feezes. Let me try not loading the device the first time. | 09:36 |
| CottonUndies | freeze happens even if I dont add the net device. hmmm. | 09:38 |
| mnemotron | aha! | 09:45 |
| - CottonUndies (QUIT: Quit: Client closed) (~CottonUnd@2405:6580:33c0:b400:3884:240d:eb9a:34e6) | 10:12 | |
| _Bnu | ('A`) | 10:13 |
| _Bnu | Is the Fast RAM on the A3000 motherboard that slow? | 10:13 |
| _Bnu | I'm not familiar with the 3000, really. | 10:13 |
| _Bnu | But typically you get really good speeds in SysInfo with a 68030... | 10:14 |
| + CottonUndies (~CottonUnd@2405:6580:33c0:b400:3884:240d:eb9a:34e6) | 10:15 | |
| mnemotron | strange | 11:50 |
| Jope | amiga test kit will definitely hang your system if you start it from the wb with a tcp stack running | 12:22 |
| Jope | same problem as with whdload if you leave your tcp stack up when you start a game | 12:22 |
| Jope | the hang comes from the network hw interrupt that is no longr served when the os is thrown out | 12:22 |
| CottonUndies | Not under 1.10FW | 12:41 |
| CottonUndies | at least that is my own repeatable testing result! | 12:41 |
| CottonUndies | Interesting whdload comment, as I also use it without problems (with roadshow running) up to FW1.10, haven't tried 1.11FW and whdload yet. Something to try! | 12:42 |
| CottonUndies | minute iBrowse scrolling also prolematic under 1.10FW. Just did a quick test. | 12:52 |
| CottonUndies | Jope Whdload seems to be working for me with my stack loaded on FW1.11. | 13:06 |
| - CottonUndies (QUIT: Quit: Client closed) (~CottonUnd@2405:6580:33c0:b400:3884:240d:eb9a:34e6) | 13:20 | |
| + CottonUndies (~CottonUnd@2405:6580:33c0:b400:3884:240d:eb9a:34e6) | 14:50 | |
| _Bnu | It's very hit and miss in my experience. | 16:04 |
| - CottonUndies (QUIT: Quit: Client closed) (~CottonUnd@2405:6580:33c0:b400:3884:240d:eb9a:34e6) | 16:50 | |
| - mnemotron (QUIT: Quit: L8r) (mnemotron@194.135.47.239) | 17:07 | |
| + xet7 (~xet7@user/xet7) | 20:38 | |
| - xet7 (QUIT: Remote host closed the connection) (~xet7@user/xet7) | 20:40 | |
| + xet7 (~xet7@user/xet7) | 20:41 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
