Re: BigRAMPlus seems to be detected now
Hi Stefan,
On Tue, Mar 31, 2020 at 7:07 AM Stefan Reinauer
<stefan.k.reinauer@gmail.com> wrote:
> On Tue, Mar 10, 2020 at 1:15 AM Geert Uytterhoeven <geert@linux-m68k.org> wrote:
> > On Tue, Mar 10, 2020 at 4:06 AM Stefan Reinauer
> > <stefan.k.reinauer@gmail.com> wrote:
> > > On Sun, Mar 8, 2020 at 3:43 AM Geert Uytterhoeven <geert@linux-m68k.org> wrote:
> > > > As well as trying to boot it with the BRP as primary memory (BRP first
> > > > in memfile) or sole memory (motherboard RAM disabled in memfile).
> > >
> > > With BigramPlus as sole memory the system can't seem to deal with the
> > > initramfs (see log)
> >
> > However, according to this log you have both enabled:
> >
> > > On node 0 totalpages: 4096
> > > DMA zone: 40 pages used for memmap
> > > DMA zone: 0 pages reserved
> > > DMA zone: 4096 pages, LIFO batch:0
> >
> > 16 MiB
> >
> > > On node 1 totalpages: 65536
> > > DMA zone: 640 pages used for memmap
> > > DMA zone: 0 pages reserved
> > > DMA zone: 65536 pages, LIFO batch:15
> >
> > 256 MiB
> >
> > > SLUB: Unable to allocate memory from node 0
> > > SLUB: Allocating a useless per node structure in order to be able to continue
> > > SLUB: HWalign=16, Order=0-3, MinObjects=0, CPUs=1, Nodes=8
> >
> > Probably the 16 MiB in the first chunk is too small to allocate
> > page tables from, especially if it contains the initramfs.
> >
> > What happens if you put the BRP first in the memfile?
>
> Before chipram?
> ---------------------------------
> 0x50000000 268435456
> 2097152
No, as the first FastRAM chunk.
ChipRAM must always be first ;-)
> This one works:
> ---------------------------------
> 2097152
> 0x50000000 268435456
> 0x07000000 16777216
> ---------------------------------
OK, that's what I meant.
> Same as only using the bigram (and I got into the OS just fine, but I
> had to do the installation through qemu)
Yeah, that won't make much of a difference, as it may take a while
before it actually starts using the on-board RAM.
Does the initramfs start working if you restrict the amount of RAM used,
by adding e.g. mem=16M or mem=64M to the kernel command line?
If there is a limitation on memory size somewhere in the code, that may
help identifying it.
Thanks!
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
Reply to: