On 30/05/2023 14.53, Paolo Bonzini wrote:
> On Tue, May 30, 2023 at 2:04 PM Thomas Huth <thuth@redhat.com> wrote:
>> On 27/05/2023 11.28, Paolo Bonzini wrote:
>>> SLOF is required for building the s390-ccw firmware on s390x,
>>> since it is using the libnet code from SLOF for network booting.
>>>
>>> If SLOF is absent and submodules are not updated, pc-bios/s390-ccw
>>> cannot be built.
>>>
>>> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
>>> ---
>>> configure | 3 ++-
>>> 1 file changed, 2 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/configure b/configure
>>> index 1d1b8736c0eb..c92a3b30b9a4 100755
>>> --- a/configure
>>> +++ b/configure
>>> @@ -1662,7 +1662,8 @@ fi
>>>
>>> # Only build s390-ccw bios if the compiler has -march=z900 or -march=z10
>>> # (which is the lowest architecture level that Clang supports)
>>> -if have_target s390x-softmmu && probe_target_compiler s390x-softmmu; then
>>> +if have_target s390x-softmmu && probe_target_compiler s390x-softmmu && \
>>> + ( test "$git_submodules_action" != ignore || test -f roms/SLOF/VERSION ); then
>>> write_c_skeleton
>>> do_compiler "$target_cc" $target_cc_cflags -march=z900 -o $TMPO -c $TMPC
>>> has_z900=$?
>>
>> Not sure if we really need this. Only the networking part of the s390-ccw
>> bios cannot be build without SLOF, but the main binary still builds fine
>> also without it.
>
> Thinking more about it---considering we have prebuilt s390-ccw
> binaries, why would someone want to rebuild only one of the ROMs and
> risk breaking the other? Should we keep this patch, and also include
> $(SRC_PATH)/netboot.mak unconditionally from
> pc-bios/s390-ccw/Makefile?
Yes, I think it makes sense nowadays. In the beginning, the network
bootloader was considered experimental and thus optional, but I think it is
well established nowadays, so it makes sense to always build it together
with the other part of the s390-ccw bios.
Thomas