Psim and kickstart

--0015175cb150070615049bb4d8f8
Content-Type: text/plain; charset=ISO-8859-1

I have managed to compile the kernel and the user level applications for
l4ka:pistachio for powerpc now.
When I tried to run it on the p2020 machine I have here it didn't work,
which didn't really surprise me.
Then however I decided that I wanted to set up psim and check what is going
on inside the machine, but it turns out that kickstart crashes on the very
first instruction in the binary, which should be:

mfspr    %r24, 0x11e


Here is the log from my attempt to run kickstart in psim, I build psim as
suggested in ppc-build.txt and used the psim.tree that was specified there.
I did not apply the patches to psim though, since with the most recent gdb
distribution they seem to have been fixed.
Here is the log from what I ran and what the results were.

powerpc-unknown-eabi-run -e chirp -i2 -f psim.tree
user/util/kickstart/kickstart

cpu 1, cia 0x700: double interrupt - MSR[RI] bit clear when attempting to
deliver interrupt, cia=0x700, msr=0x1000; srr0=0x400000(cia),
srr1=0x89032(msr); trap-vector=0x700, trap-msr=0x1000
CPU #1 executed 1 Move from Special Purpose Register instruction.

CPU #1 executed 2 icache misses.
CPU #1 executed 1 instructions in total.

CPU #2 executed 1 Branch instruction.

CPU #2 executed 1 icache miss.
CPU #2 executed 1 instructions in total.

CPU #3 executed 1 Branch instruction.

CPU #3 executed 1 icache miss.
CPU #3 executed 1 instructions in total.

CPU #4 executed 1 Branch instruction.

CPU #4 executed 1 icache miss.
CPU #4 executed 1 instructions in total.

All CPUs executed 4 instructions in total.
Simulator speed was 111 instructions/second.
user/util/kickstart/kickstart: Caught signal -1 at address 0x700

I hope somebody can offer some advice on this issue.

Best regards,
Nils Ole Timm

--0015175cb150070615049bb4d8f8
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

I have managed to compile the kernel and the user level applications for l4=
ka:pistachio for powerpc now.<br>When I tried to run it on the p2020 machin=
e I have here it didn&#39;t work, which didn&#39;t really surprise me.<br>
Then however I decided that I wanted to set up psim and check what is going=
 on inside the machine, but it turns out that kickstart crashes on the very=
 first instruction in the binary, which should be:<br><br>mfspr=A0=A0=A0 %r=
24, 0x11e<br>
<br><br>Here is the log from my attempt to run kickstart in psim, I build p=
sim as suggested in ppc-build.txt and used the psim.tree that was specified=
 there.<br>I did not apply the patches to psim though, since with the most =
recent gdb distribution they seem to have been fixed.<br>
Here is the log from what I ran and what the results were.<br><br>powerpc-u=
nknown-eabi-run -e chirp -i2 -f psim.tree user/util/kickstart/kickstart <br=
><br>cpu 1, cia 0x700: double interrupt - MSR[RI] bit clear when attempting=
 to deliver interrupt, cia=3D0x700, msr=3D0x1000; srr0=3D0x400000(cia), srr=
1=3D0x89032(msr); trap-vector=3D0x700, trap-msr=3D0x1000<br>
CPU #1 executed 1 Move from Special Purpose Register instruction.<br><br>CP=
U #1 executed 2 icache misses.<br>CPU #1 executed 1 instructions in total.<=
br><br>CPU #2 executed 1 Branch instruction.<br><br>CPU #2 executed 1 icach=
e miss.<br>
CPU #2 executed 1 instructions in total.<br><br>CPU #3 executed 1 Branch in=
struction.<br><br>CPU #3 executed 1 icache miss.<br>CPU #3 executed 1 instr=
uctions in total.<br><br>CPU #4 executed 1 Branch instruction.<br><br>CPU #=
4 executed 1 icache miss.<br>
CPU #4 executed 1 instructions in total.<br><br>All CPUs executed 4 instruc=
tions in total.<br>Simulator speed was 111 instructions/second.<br>user/uti=
l/kickstart/kickstart: Caught signal -1 at address 0x700<br><br>I hope some=
body can offer some advice on this issue.<br>
<br>Best regards,<br>Nils Ole Timm<br><br>

--0015175cb150070615049bb4d8f8--