On Thu, Jul 11, 2019 at 2:53 AM Alistair Francis <address@hidden
On Wed, Jul 10, 2019 at 10:51 AM Chih-Min Chao <address@hidden> wrote:
> Similar to the mips + malta test, it boots a Linux kernel on a virt
> board and verify the serial is working. Also, it relies on the serial
> device set by the machine itself.
> If riscv64 is a target being built, "make check-acceptance" will
> automatically include this test by the use of the "arch:riscv64" tags.
> Alternatively, this test can be run using:
> $ avocado run -t arch:riscv64 tests/acceptance
> Signed-off-by: Chih-Min Chao <address@hidden>
Awesome! Thanks for the test case. This will help a lot with RISC-V
regressions in QEMU.
> tests/acceptance/boot_linux_console.py | 40 ++++++++++++++++++++++++++++++++++
> 1 file changed, 40 insertions(+)
> diff --git a/tests/acceptance/boot_linux_console.py b/tests/acceptance/boot_linux_console.py
> index 3215950..bbc6b06 100644
> --- a/tests/acceptance/boot_linux_console.py
> +++ b/tests/acceptance/boot_linux_console.py
> @@ -354,3 +354,43 @@ class BootLinuxConsole(Test):
> console_pattern = 'Kernel command line: %s' % kernel_command_line
> + def test_riscv64_virt(self):
> + """
> + :avocado: tags=arch:riscv64
> + :avocado: tags=machine:virt
> + """
> + kernel_url = ('https://github.com/chihminchao/test-binary/raw/'
These images need to be built with a standard build flow. Having them
built from SiFive's custom scripts will make debugging problems in the
future impossible. I'm also a little worried here about GPL
violations, I'm not sure if it's enough to just point to a script SHA
to meet GPL source disclosure. I know companies have huge headaches
meeting GPL requirements so this seems too easy.
I am not very familiar with this kind of binary and source license conflict.
Is it ok if I write a simple script with BSD license to build kernel and image from sifive's linux/buildroot repo and commit the script to my test-binary repo ?
> + '0b7787305d9e40815c05a805266cc74ff356239e/qemu/riscv64/'
> + 'bbl_w_kernel.gz')
Don't use BBL, most people use OpenSBI now which is what we should be
I will try to move to OpenSBI in next version. My environment is based freedom-u-sdk and It still relays on BBL. So ..:P
> + kernel_hash = 'c7f6cc7967975ad42dc61ee0535db01c9cbd0968'
> + kernel_path_gz = self.fetch_asset(kernel_url, asset_hash=kernel_hash)
> + kernel_path = self.workdir + "bbl_w_kernel"
> + with gzip.open(kernel_path_gz, 'rb') as f_in:
> + with open(kernel_path, 'wb') as f_out:
> + shutil.copyfileobj(f_in, f_out)
> + initrd_url = ('https://github.com/groeck/linux-build-test/raw/'
> + '8584a59ed9e5eb5ee7ca91f6d74bbb06619205b8/rootfs/'
> + 'riscv64/rootfs.cpio.gz')
Same comment about build tools.
> + initrd_hash = 'f4867d263754961b6f626cdcdc0cb334c47e3b49'
> + initrd_path = self.fetch_asset(initrd_url, asset_hash=initrd_hash)
> + self.vm.set_machine('virt')
> + self.vm.set_console()
> + kernel_command_line = (self.KERNEL_COMMON_COMMAND_LINE
> + + 'console=ttyS0 noreboot')
> + self.vm.add_args('-kernel', kernel_path,
> + '-initrd', initrd_path,
> + '-append', kernel_command_line)
> + self.vm.launch()
> + self.wait_for_console_pattern('Boot successful.')
> + self.exec_command_and_wait_for_pattern('cat /proc/cpuinfo',
> + 'isa')
> + self.exec_command_and_wait_for_pattern('uname -a',
> + 'sifive')
> + self.exec_command_and_wait_for_pattern('reboot',
> + 'reboot: Restarting system')