I have attached two bug reports, the first without any extra kernel parameters, and the second with pci=nocrs. When I use pci=nocrs, the Nvidia modules load because my GeForce RT 710 is found, but starting X results either in a blank screen or a green screen, depending on the resolution settings set or the monitor connected (green screen occurs over HDMI when connected to an LG 24UD58-B 4K monitor).
My motherboard is an Asus PRIME X399-A with the latest BIOS (Oct. 2018, version 0808).
nvidia-bug-report.log.gz (54.8 KB)
nvidia-bug-report.log.pci_nocrs.gz (1.01 MB)
[ 0.432802] pci 0000:41:00.0: BAR 1: no space for [mem size 0x08000000 64bit pref]
[ 0.432958] pci 0000:41:00.0: BAR 1: trying firmware assignment [mem 0x98000000-0x9fffffff 64bit pref]
[ 0.433118] pci 0000:41:00.0: BAR 1: [mem 0x98000000-0x9fffffff 64bit pref] conflicts with PCI Bus 0000:40 [mem 0x9c000000-0xa95fffff window]
[ 0.433287] pci 0000:41:00.0: BAR 1: failed to assign [mem size 0x08000000 64bit pref]
Looks like a kernel/bios bug. Try a different pcie slot.
I did, and I still got the same address collision issue. Why doesn’t pci=nocrs work, though?
Would adding pci=realloc, too, help?
BAR1 is the memory window needed for graphics communication with the gpu. nocrs skips the allocation so the module loads but no graphics output work. realloc might work.
realloc led to a kernel panic upon boot, so it does seem to be a kernel issue.
Please check if you have an option “Above 4G Decoding” in bios and enable that, if possible.
I couldn’t find that BIOS option.
Just to make sure the card isn’t broken, you should check it in another system. Otherwise you can only up-/downgrade the kernel and maybe issue a bug report at the kernel bugzilla.
It works with the VESA driver just fine.
[/quote]
filed