The power compatibility is different to bus compatibility.

Max power on PCI bus is 25 W max, AGP bus is 48 W max and PCIe is 75 W max without additional power plug.

It means, if you use PCI-AGP/or PCI-PCIe adapter with AGP/PCIe cards, you are still limited to 25 W of your PCI bus in Sam.

I.e. PCIe card think, that is in the 75W bus, but you in real connected it to 25W bus. So not works without additional power.


And if the card R7 240 has 30W, is out range and you need additinal power.
I really recommend you to read again my article, all this is inside. Use google translate.

To check gfx cards parameter use this link: https://www.techpowerup.com/gpu-specs/

And PCI card HD 5450 is PCIe card with 19W power + PCI-PCIe bridge incorporated on single card.

@nbache
of course, it is copy-paste syndrome
fortunatelly, it is only text, &LX &LY values are OK.

@smarkusgQuote:
good tip, thank you

Quote:
exactly.
I have physical hw, so I can test something. It is far from ideal, but better than telephone consultation

Probably best is, if I learn SPE SIMD and try to rework critical calculations suitable for vectors.

@GebrochenQuote:

I know, I mentioned Sam460 only becouse you were curious about Sam460 speed vs. Sam440, and this I can answer - after testing

Quote:

You have several options with Sam440ep ( check the article above )

1. PCI card
Standart 9200, 9100 cards need no Enhancer or Radeon HD driver.
Disadvantage is, that Warp3D Nova and OGLES2 not works here.

( P.S: PCI versions of newer PCIe cards (like your radeon HD5450 probably ) is the same like PCIe cards below )

2. AGP cards
- needs PCI-AGP adapter ( powered). Best is Radeon 9000 PRO.
Much faster than PCI 9200 or 9250, no RadeonHD needed, Warp3D Nova and OGLES2 not works here.

3. PCIe cards
- needs PCI - PCIe adapter + powered extension. For howto do powered extension, look into article.
All these cards needs to purchase Enhancer or RadeonHD driver. There are RadeonHD version with limited functions on installation CD, but it is only for installation, it has no advanced functions.

Versions: look here.

RadeonHD.chip v3.x ( part of Enhacer ) - is must have if using PCIe cards ( or PCI HD cards )
compatible with X1300-X1950 and HD 4xxx - HD 7xxx, R7, R9
This version has no video decoder.

There are also RadeonHD v5, but only for X1000, X5000 and Sam460ex! It is no for Sam440!
This version is limited to HD 77++, R7, R9, it has GART ( except Sam460) and video decoding feature.
I will test it, if it can be used on Sam440 - CPU is by design similar to Sam460ex, and let know.

@GebrochenQuote:
This question is directly after m3x. He made HyperClock (thx!).
As I know, Sam CPU has several buses with several timings. Combinations in Hyperclock selected from many theoretical combinations to several working combinations.
And no, first Sam460 has no 133 PCI bus for sure.

Generally, all Sam440 and Sam460 are very nice and very stable piece of hardware, but with lower power. So we need some careful configuration ( especially with emulators etc ).

Enjoy it.
I have now also Sam460LE, so will I test it against Sam440ep-fxex and made some article.

@amigakit
thanks, it is great news!

@smarkusg

I have in plan to test SPE SIMD units.
I looked into mplayer source and here is a tons of altivec code.
Pls, can you recommend me the simpliest example of use of altivec code in mplayer package?

I want learn how to make a SPE SIMD version for this.

INTRO - this post will be long, if you want, jump down..
At first, I have to thanks all guys involved in A1222+ production, and all coders who tried to compile something for SPE. And especially to Hans, flash, HunoPPC and MickJT ( they kindly answered my stupid questions and requests )
The main reason for my thanks is that I now forced to return to programming in c. After twenty-five years of bash, terraform and ansible .

I read manual of P1022 CPU and found, that SPE is interesting feature - it not only Embeddded FPU, but also two SIMD engines ( int + float ). Not so powerful like AltiVec, but still nice.

1. Flash's example
Now back to flash ( and mine ) problem above:
Corrupt output from "printf" function ( and all external functions with float value parameters ) is, that float parameters are passed by registers. And P1022+SPE has no FPRs, uses 64-bit GPSs instead.

So in example above code is compiled for SPE, it puts "result" value in 64-bit GPR. And printf() is compilled for powerpc standart, so it reads parameter from FPR !!!.
And if P1022 has no FPR, exception is generated, and AmigaOS run LTE FPU emulator and retuns value from Emulated FPR - where is stored result from previous emulated operations. This value is of course totally irrelevant to our function call...

2. How to fix it
Answer is inside this nice Hans's Taborizing guide - see paragraph 2. But unfortunatelly, this guide is too short, is not written for blondes, who did c-coding 25 years before.
So I have to install Codebench ( thx @walkero for example ) and SDK and start to experiment on live A1222. And try some examples.
As Hans wrote, we cannot call standard powerpc-compilled functions with float arguments from SPE-code. But we can call call functions with pointer to float ( float* ), because value is address and is stored in GPRs both in SPE and powerpc code.
For every standard function call we can do replacement of this function called by reference.

3. solution
Instead of use printf() directly in SPE code I used _pritnf__SPEfp_1 () custom replacement. _pritnf__SPEfp_1 has no body here, only declaration. All this code is compilled with SPE:
> gcc -c -mcpu=8540 -mfloat-gprs=double -mabi=spe -mspe -c SPEfloat-printf-output.c -o SPEfloat-printf-output.o



Now we call function with &LX, which is address stored in GPR.
Now we made _pritnf__SPEfp_1() body with simple code ( it calls printf() ), compilled for powerpc ABI.
> gcc -c -mcpu=powerpc -c printf__SPEfp_1.c -o printf__SPEfp_1.o


This code is powerpc ABI, so put float value into standard FPR and simply calls printf(). Of course, all this FPR call is emulated in LTE, but it works with consistent data.

Last steps are simple:
> gcc -c -mcpu=powerpc -c printf__SPEfp_1.c -o printf__SPEfp_1.o
> gcc SPEfloat-printf-output.o printf__SPEfp_1.o -o SPEfloat-printf-output
> SPEfloat-printf-output
Hodnota SPE LX=1.500000
Hodnota SPE LX=15.500000

I did it!
So it is the same like Hans advice, but for us who are slower...

4. What next ?
It is only workaround to made compatible powerpc ABI to SPE ABI and utilize hardwired embedded FPU.
Btw, did you know, that today's modern ARM CPUs has exactly the same problem with different FPUs and FPRs? But they have soulution - gcc parameter "mfloat-abi=softfp" which made code which stores f-p values in integer registers. The same what we did above.

We don't do anything yet with main SPE feature - SIMD units.
It is for next time - it is very time consuming to do it only with Basics of SIMD Programming for AltiVec + spe.h + SPEProgrammingEnvironmentsManual

If you find some SPE SIMD examples, it will be great....

5. The end
P1022 is not crippled CPU - it is more powerful CPU than G3!!!
Only it takes some work and time to do it properly...

And my last question: Where is the best place for such discussion and examples? Here or os4coding, or elsewhere?

@Hypex
Thanks!

@Gebrochen
limpid clock runnig, CPU usage increase to 8%
If I moving with clock+calendar fast around the workbench, usage goes to 50-70%. But I think it is normal.
Moving around a notepad full of text utilizes CPU to 35%..

Quote:
factory default for 800 MHz version is FSB 133, PCI 66
my 733 MHz version have FSB 147, PCI 73 - i.e. slower CPU, faster buses

800 CPU, FSB 133, PCI 133 in Hyperclock is one of many oveclocking options.
I recommend try modes where all three frequencies are same or a little higher than yours. PCI 133 MHz is too much = 2x standard value, your PCI-PCIe bridge or other device probably freeze.
Try:
800-160-80
833-167-83
867-144-72
probably none of them will not work for you, it is too high. But try it.

For example, for me worked 778-156-78 for some application and games, not for all. You should also try my frequentions, CPU a little bit slower, but faster busses - it depend of task, if it will be ok or not.

@Gebrochen
I cannot second this. I have Sam460ep-flex, 733MHz version with R9 270X.
Windows move is smooth, when I typing fast in shell, CPU is utilized 4-7%, in standby 3-5% ( CPUInfo.docky )

First check, if you have right RadonHD.chip ( v3.7 from last Enhancer works for me, and according to wiki v1.20, v2.21 or v3.6 should work too )

I have compositing desktop, so try it with Prefs/Gui/Effects/compositing enable + optimizing for performance.
If not works, check wbstartup and user-startup if there are no some power-consuming apps.

Then, I can only recommend to wait for new R7 240 card, and if it not helps, reinstall AmigaOS.

Here is my old article about testing graphics cards on Sam440ep-flex ( Czech language ). Look inside and compare results with your cards.

@AmigaOldskooler

thanks, it is very nice! It's already becoming a magazine...

Great! Thanks @geennaam!

Interesting reading, thanks for sharing.
I will try for sure.

@Kamelito

I will make some tests, but it takes some time, I will play with everything.
As eliyahu wrote on his pages, all SPE compilled, and most of Warp3D Nova are very fast.

For example just tested A1222+ + RX 580: Tower 57 SPE = 71 FPS.
My AmigaOne X1000 + HD7970 has 99 FPS and Powermac G5 2.7 GHz + Radeon 9800 Pro = 65-74 FPS ( according to Trevor's blog )

So if you will wait long enough, I will wrote some article. But first I made some technical comparison of P1022 and AMCC 460ex CPUs.

Also, regarding the card question, very probably is best RX 580. But it is a little overkill for A1222, it has 185 W max . Maybe is more wise some compromise like RX 550 - 50 W or RX 560 - 75 W.

@geennaam
thanks for support!

This line missing in A1222+ log:



P.S: this log is without any fsl_ddr: parameter. But with it logs not differs.

@geennaam
thank you, it will ne nice!

Just now I tested both fsl_ddr:bank_intlv=cs2_cs3 and fsl_ddr:bank_intlv=auto, with the same result - speed is the same like eliyahu`s 1R memory.

Also can be, that option in A1222 U-Boot missing.

@balaton

I didn't noticed interrupts failing.

I never tested rtl8139, only rtl8169 and this not works for me, even if worked in AmigaOne XE.
I was some years ago, so I am not remember if it freezes or just not working.

Just tested interleaving with:
hwconfig=esdhc;fsl_ddr:bank_intlv=cs0_cs1;audclk:12;usb1:dr_mode=host,phy_type=ulpi

the good thing: A1222+ not bricked
the bad thing: interleaving not working - both output from UBoot and ragemem speed are the same

@balaton
please, do you have any idea what u-boot repo is closest to our A1222+? There are hundreds of them on git.

@all
please, are there any way how ask A1222+ U-boot developers?