| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
| |
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
| |
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
In the P9 hello script we look for $P9MAMBO_BINARY, but then if we don't
find it we print a message using $MAMBO_BINARY, which is confusing:
$ set -x P9MAMBO_BINARY run/p9/run_cmdline
$ ./test/hello_world/run_mambo_p9_hello_world.sh
Could not find executable P9MAMBO_BINARY (/opt/ibm/systemsim-p9//).
Fix it so the right thing is printed:
Could not find executable P9MAMBO_BINARY (/opt/ibm/systemsim-p9//run/p9/run_cmdline).
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
|
|
| |
This allows us to fail a lot faster if we checkstop
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
|
|
| |
A regression test for the mambo implementation of OPAL_SIGNAL_SYSTEM_RESET.
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
|
|
|
|
| |
Helps if you're building mambo from source (or haven't used the
packages in exactly the way they install) to ensure you run both
the p8 and p9 mambo simulators for testing.
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
|
|
|
|
|
|
|
| |
convert test cases to diff DTS rather than DTB.
This means we also have to build dtc on CentOS 7 to be able to run
the test suite.
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
[oohall@gmail.com moved the test cases into seperate patches]
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
| |
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
| |
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
|
|
| |
Now based on Cédric's branch.
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
|
|
|
|
|
|
|
| |
The hello world kernel fails to correctly set r3 before making the
shutdown opal call. On FSP machines only shutdown types 0 and 1 are
recognised as valid shutdown types. If any other type is specified
(in r3) the call is rejected with an OPAL_PARAMETER error and the
machine will continue running.
Signed-off-by: Oliver O'Halloran <oohall@gmail.com>
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
|
| |
Fixes: 2ceb8b8c0ac53a9014e83d12e1c758d5f9e07fe6
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Commit 2ceb8b8c (make check: make test runs less noisy) used tempfile
command to create temporary files. Looks like distros like Fedora
doesn't have tempfile command.
Looks like mktemp is supported on all distros (at least Fedora and
Ubuntu supports this command). Hence replace tempfile with mktemp.
Output without patch:
[ CLEANUP ] gard-test-clean
./test/run.sh: line 3: tempfile: command not found
skiboot/external/gard/test/Makefile.check:18: recipe for target 'gard-test-clean' failed
make: *** [gard-test-clean] Error 1
Signed-off-by: Vasant Hegde <hegdevasant@linux.vnet.ibm.com>
Reported-by: Balbir Singh <bsingharora@gmail.com>
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
| |
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
| |
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
|
|
|
|
|
|
|
| |
Run a small wrapper around some unit tests with the QTEST makefile macro
(QTEST=Quiet TEST). Also, wrap boot tests in mambo and qemu to be quiet
by default.
Both ./test/run.sh and the modified mambo/qemu test runner scripts output
full stdout and stderr in the event of error.
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Call OPAL_CEC_POWER_DOWN at end rather than using a raw attn
instruction. We are doing this since attn should be disabled in the
host kernel otherwise userspace may crash the machine by calling it.
We are now doing two opal calls so need to save some registers which
are volatile over opal calls.
This bloats the code from 10 instructions up to 16. May god have mercy
on my soul.
Signed-off-by: Michael Neuling <mikey@neuling.org>
[stewart@linux.vnet.ibm.com: change hello world test run to look for power
down rather than ATTN]
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
|
|
|
|
|
| |
Adds make targets to perform boot tests using multithreaded mambo. There
are seperate targets for the kernel and hello_world payloads.
Signed-off-by: Oliver O'Halloran <oohall@gmail.com>
[stewart@linux.vnet.ibm.com: Use SMT rather than threaded in test description]
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|\
| |
| |
| | |
Merge device tree sorting
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Moved the dt_dump() into test/dt_common.c so that it can be shared
between hdata/test/hdata_to_dt.c and core/test/run-device.c
run-device.c contains two tests, one basic sorting test and a
generate-and-sort test.
Signed-off-by: Oliver O'Halloran <oohall@gmail.com>
[stewart@linux.vnet.ibm.com: remove trailing whitespace]
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
| |
| |
| |
| | |
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
| |
| |
| |
| | |
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
| |
| |
| |
| | |
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|/
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
We now run tests under both Mambo and QEMU PowerNV model.
Also added are scripts in opal-ci/ for building powernv model qemu.
Currently, this is not yet in upstream qemu, so we build from Ben's branch.
You can start using qemu along with/instead of Mambo by:
1) (cd ./opal-ci; sudo ./install-deps-qemu-powernv.sh; ./build-qemu-powernv.sh)
2) Pointing QEMU_PATH and QEMU_BINARY environment variables to appropriate
qemu binary with powernv model
When building qemu ourselves, we build a *specific* known good tag from
the open-power tree. This should ensure that into the future existing test
scripts should continue to function.
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
| |
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
|
| |
Signed-off-by: Michael Neuling <mikey@neuling.org>
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
|
|
|
|
|
|
|
| |
Document it a lot more. Remove magic numbers. Remove include file. Make it more
flexible.
... and most importantly remove the horrendous bloat. From 17 instructions
down to 10!
Signed-off-by: Michael Neuling <mikey@neuling.org>
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
|
| |
Signed-off-by: Michael Neuling <mikey@neuling.org>
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
|
| |
Signed-off-by: Michael Neuling <mikey@neuling.org>
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
| |
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
| |
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
|
|
|
| |
We can boot FSP machines and extract GCOV coverage report from them
combining with Mambo reports
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
|
|
| |
We also add a makefile to help people re-create the images.
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
|
|
|
|
|
|
|
| |
This target will run the (two) current boot tests and produce lcov
coverage of skiboot from running them in Mambo.
Everything is pretty hard coded at this stage and should most certainly
be improved upon, especially if we want input from real hardware or to
have more boot tests.
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
|
|
|
|
| |
we now properly control the simulation rather than just killing off
Mambo. For boot test, we wait for petitboot and actually shut down
the simulation properly with 'halt'.
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
|
|
|
|
| |
If you've built a zImage.epapr using op-build, you can now drop it
in the top level skiboot source tree, run 'make check' and these
test files will pick it up and run a boot test using it.
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
powerpc-linux-gnu-gcc -m64 -c -MMD -Itest/hello_world/hello_kernel/ -o
test/hello_world/hello_kernel/hello_kernel.o
test/hello_world/hello_kernel/hello_kernel.S
test/hello_world/hello_kernel/asm-utils.h
powerpc-linux-gnu-gcc: fatal error: cannot specify -o with -c, -S or -E
with multiple files
GCC barfed as we were passing the header and the c file while compiling
with -c. Fix the rule to just build source file.
Signed-off-by: Joel Stanley <joel@jms.id.au>
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The test for expect would always return true. This change makes it do
the correct thing in the presence and absence of expect on my Ubuntu
machine.
Also, skip the test if the user has KERNEL set. With this set skiboot
contains an embedded kernel, and will load that before falling back to
our hello_world test kernel, causing all kinds of confusion.
Signed-off-by: Joel Stanley <joel@jms.id.au>
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
|
|
|
|
|
|
|
| |
Instead of having individual rules to generate .d, add -MMD to
HOSTCC parameters, and just include the generated .d files.
This fixes a few weird dependency issues.
Also, make the mambo hello_kernel test depend on skiboot.lid
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
|
|
| |
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|
|
Just calls OPAL_CONSOLE_WRITE with "Hello World!" and with mambo
we can execute this tiny boot test in not much time at all.
Good little sanity check.
Signed-off-by: Stewart Smith <stewart@linux.vnet.ibm.com>
|