If your IDE bus needs DMA off then hopefully this covers your CF card. As long as it detects it CF and shuts it off. But check this line out:kernel parameter libata.dma=1 causes that harddisks ( i.e. SSd) use DMA, and CD + CFcards not used DMA. What is exactly I need.
There if also possibility for detailed DMA/PIO settings, to use kernel parameter libata.force=1:pio4,2:pio4,3:udma6 .... but in my case I am not using it.
Code: Select all
# Uncomment to disable DMA for devices on IDE channels for old IDE drivers.
#SLB_CMDLINE_LINUX="${SLB_CMDLINE_LINUX} ide_core.nodma=0.0 ide_core.nodma=0.1 ide_core.nodma=1.0 ide_core.nodma=1.1"
The post install calls update-amigaos4-slb to create the boot entries. Once installed they are saved to the a1boot volume. In folders named after UUID. The old config is backed up but in case it's lost update-amigaos4-slb can recreate menu from a1boot folders. Using custom is also a good way to keep it saved. And the official way.I leave original rows untouched, as is generated with postinstall.sh script, and I add another row(s) with my testing configuration. This way my configuration appears on booting menu, like last line(s).
Yes. It's because the XE has lacks CPU cache coherency. So it creates a conflict for PCI devices. The end result is a system freeze. I've enabled hardware acceleration and even on a 9250SE (that I have) it works nice with windows. But it only lasts a few moments and then the system freezes. There's no real solution to this. Well apart from what OS4 driver is doing. And if there were it would take too much work modifying code to do it.Debian Jessie on XE has disabled hardware acceleration of graphics cards. Do somebody know the reason?
However, if you wish to see it in action, for a short while at least, disable the /etc/X11/xorg.conf.d/amigaone.conf file or settings inside and have a small moment of joy.
Check this as well:R100 + R200 graphics chip have no shaders, thus linux drivers cannot support kernel modesetting, only fbdev. So as drivers are very probably the same, this issue is on AmigaOne XE cards too.
And if we are using in our AmigaOne XE cards like 9250 or 9000 PRO with R200 chip, acceleration is disabled.
Code: Select all
# Uncomment to use radeon KMS driver as framebuffer driver.
SLB_CMDLINE_LINUX="${SLB_CMDLINE_LINUX} video=radeonfb:off"
Sorry about that. I'd need to find my notes but I recall there was a rare issue when this happened. I could not figure out what caused it nor why it stopped responding. The menu code, which is based on ub2lb for Sam, calls UBoot to check keyboard. I ran through the routine and could not see where it would be stuck. Bit when it did happen it was annoying and I never tracked it down.1) in early menu with boot options selections, if i touched any key (D/DMA, U/USB/... or arrow up/down to select boot option it looks normally and change the option, but after that if I press Enter / Return, nothing happens
It's possible UBoot could be doing it. It is a beta. Does the keyboard go dead when plugged in directly?Has anybody similar problem? Or it is on my Micro only? And can be the cause UBoot different version?
Also, do you have a serial cable? That may show more closely what UBoot was doing when it stopped. It's possible it could be stuck on RAM. Pulling out the RAM and/or cleaning the contacts may help. I find these PPC machines need this as the RAM goes stale sitting there.
On that note you may try my old ram tester. But it's limited to UBoot. By the looks of it I should update it with a full USB image. But check the comments with myself. It looks like this was the key problem I found. However it was found on a Sam 440.
http://os4depot.net/?function=showfile& ... tester.lha