@afxgroup If that is the case, that means that the X5000 as sold by A-Eon, are broken. But why this happens only with some cards and not every card? Did Luigi Burdo shared that information anywhere, where I could have a look at?
Thank you. Two very lengthy discussions. Is it the chip, the motherboard design, UBoot, or some combination....?
This is the sort of issue I hope we can avoid with the Mirari boards. And if it ends up being a design issue with NXP chips, I at least expect Harald and Dave to be honest with us.
We might need to give such a card a try in the Mirari. I have never tested this. On the other hand..we have SATA and NVME.
Even if those are useless on the Mirari: Testing PCI(e) hardware known to have issues on other Amiga PPC systems, like the Sii3152 on classic Amigas (with FireStorm, Prometheus or Mediator) and on the Sam440 as well as the Sii3132 on the X5000, could help to find hardware and/or software (U-Boot) problems on the Mirari.
I agree, and i might order one of these cards to give it a try. As we use a way newer Uboot version compared to the X5K machines it would still be unclear where the problem lies regarding the X5K`s .. Anwyay i would agree that testing as much hardware as possible will always benefit!
AmigaOne X5000 -> 2GHz / 16GB RAM / Radeon RX 550 / ATI X1950 / M-Audio 5.1 -> AmigaOS 4.1 FE / Linux / MorphOS Amiga 1200 -> Recapped / PiStorm CM4 / SD HDD / WifiPi connected to the NET Vampire V4SE TrioBoot RPI4 AmiKit XE
Since you chimed-in, I'll ask this: if Sil3132 problems do carry over to the Mirari, as NXP developers, are you entitled to their tech support? Do you have the ability to open tickets, or is that a different support tier?
Well i do think we can we can create tickets at NXP if needed. On the other hand, asking support for a 21 year old SIL3132 chipset based budget controller might not be something they see as a high priority issue.. but again if needed we can always try.
AmigaOne X5000 -> 2GHz / 16GB RAM / Radeon RX 550 / ATI X1950 / M-Audio 5.1 -> AmigaOS 4.1 FE / Linux / MorphOS Amiga 1200 -> Recapped / PiStorm CM4 / SD HDD / WifiPi connected to the NET Vampire V4SE TrioBoot RPI4 AmiKit XE
Good to know, thanks. Yes, it makes sense that they won't get excited over one old, obscure card, but if this problem plagues one card, it probably plagues other, too. Somewhere.
Getting OT, but I wonder how many Amiga users have access to a wide range of PCI/PCie cards (maybe a friend owns a computer repair shop?) and checked them for recognition in UBoot/CFE? When you know there are drivers for only cards X, Y, and Z, I think the natural tendency is to steer away from anything else. Maybe some of the X5000 beta testers had a go?
The linux nerds have used more cards on the X5000 than anyone else and the Sil 3132 is the only one i ever seen beeing reported as invisible to the system, that one was reported about already in 2016.
Skateman wrote:Well i do think we can we can create tickets at NXP if needed. On the other hand, asking support for a 21 year old SIL3132 chipset based budget controller might not be something they see as a high priority issue.. but again if needed we can always try.
You're probably right, and it's not worth it for the X5000 or its Mirari board.
The internal controllers on both boards are fast enough, and very few people are likely to use more than two hard drives with this hardware.
For example, I use a 1 terabyte SSD and a 512 GB NVME, which is more storage than I use on my MacStudio
MacStudio ARM M1 Max Qemu//Pegasos2 AmigaOs4.1 FE / AmigaOne x5000/40 AmigaOs4.1 FE
Yes isn't that typical that the only driver is for hardware that is not working on X5000
But it isn't really the only option.
We used to have a driver for NVMe devices. iirc the author removed his driver from os4depot because of for me unknown reasons but maybe he will change his mind later on.
The author of the NVME and HDA Audio driver Harald ΅geennaam" Kanning removed his drivers due to a perceived violation of his non-commercial usage terms. As the driver was free to use for all, it just should not be used or distributed in a commercial way.
So every system sold with a soundblaster Audigy FX for example should not be able to produce sound..since there is not driver available for it. If it came with a driver... we are back at the root cause of why it got removed in the first place.
Anyway.. the NVME and HDA audio driver will be part of the new Enhancer software i was told.
Edited by Skateman on 2025/11/18 16:20:39
AmigaOne X5000 -> 2GHz / 16GB RAM / Radeon RX 550 / ATI X1950 / M-Audio 5.1 -> AmigaOS 4.1 FE / Linux / MorphOS Amiga 1200 -> Recapped / PiStorm CM4 / SD HDD / WifiPi connected to the NET Vampire V4SE TrioBoot RPI4 AmiKit XE
The AmigaOne (SE/XE/µA1) U-Boot versions don't initialize any IT, SII, LSI/NCR, etc. PCI(e) PATA, SATA and SCSI hardware either, but sg2's AmigaOS 4.x drives for it are still working on any supported hardware. At least his SII 3112 and 3114 drivers are working on the Sam440EP the same way well, using completely different versions of U-Boot.
I have a Si0680 IDE controller in my XE. Uboot is able to boot off it. From any device unit IIRC.
My X1000, having SATA built in, can only boot from one port and then AmigaBoot can only boot from one drive. Funny that the X1000, which has the best SATA support of all machines, providing four ports, can only boot from one drive and due to software limitations. Quote:
BTW: The X1000 doesn't use U-Boot, it's firmware is CFE (Common Firmware Environment), AFAIK OpenFirmware based and more similar to the Pegasos2 SmartFirmware, not comparable to the U-Boot firmware on the AmigaOne, Sam4x0, X5000 and A1222+
In the X1000 CFE is OpenFirmware based because it's some kind of hack. Varisys are likely responsible. CFE itself purely is not OpenFirmware based at all. It's independent. I don't know what happened to the X1000 CFE but the guts was ripped out and replaced with OpenFirmware. CFE is only the user interface on the terminal. Mind you, embedding OpenFirmware inside is no problem, but the way it was done means it's totally incompatible with Broadcoms CFE and any SDK that can compile binaries won't work. It's similar to SmartFirmware as both load OpenFirmware compatible binaries. So any binary running executes like any OpenFirmware binary using the OpenFirmware binary interface. The difference between the Pegasos and X1000 being the memory map is different. So any generic OpenFirmware binary cannot run.
The Sil3132 controller is not recognized at all on the PCIe bus, so this is not the same situation you reported with the AmigaOne example. It could be an incompatibility between the PCIe bridge and the controller, or an incorrect configuration of the PCIe bridge.
That's a bit more extreme than it showing up but not working with a known driver.
I once plugged in a USB2 card into my X1000. It showed up on the PCI bus but Ranger reported it as something different than expected. Maybe as storage and not USB. I recall posting something about the details and seemed to get an angry response back. Which I thought was rather strange as I did recall the same card plugged into my XE came up in Ranger as being a standard USB card. The same software Ranger reported the same card as being different in my X1000.
Same card. Same software. Different machine. Gave a different hardware info. No explanation as to how this would happen if the hardware should be hardwired but different firmware used in this case.
News feed
OS4 Feedback forum
OS4Depot Feedback forum
AmiCygnix forum
ABC shell forum
AmiKit forum
Cinnamon Writer forum
CodeBench forum
Digital Universe forum
Dopus 5 forum
E-UAE forum
Gnash forum
Ibrowse forum
JAmiga forum
Odyssey forum
OWB forum
Qt forum
SmartFileSystem forum
Timberwolf forum
TouchDevice forum
TuneNet forum
Unsatisfactory Software forum
Donate
