| Message ID | 20180822095421.11765-9-quintela@redhat.com |
|---|---|
| State | New |
| Headers | show |
| Series | Don't check disabled drivers | expand |
On 2018-08-22 11:54, Juan Quintela wrote: > We protect it with CONFIG_VMXNET3_PCI now, so no need to also put it > on i386. > > Signed-off-by: Juan Quintela <quintela@redhat.com> > --- > tests/Makefile.include | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > > diff --git a/tests/Makefile.include b/tests/Makefile.include > index d524e1bdeb..91ffde3103 100644 > --- a/tests/Makefile.include > +++ b/tests/Makefile.include > @@ -247,6 +247,8 @@ check-qtest-pci-$(CONFIG_IVSHMEM_DEVICE) += tests/ivshmem-test$(EXESUF) > gcov-files-pci-$(CONFIG_IVSHMEM_DEVICE) += hw/misc/ivshmem.c > check-qtest-pci-y += tests/megasas-test$(EXESUF) > gcov-files-pci-y += hw/scsi/megasas.c > +check-qtest-$(CONFIG_VMXNET3_PCI) += tests/vmxnet3-test$(EXESUF) > +gcov-files-$(CONFIG_VMXNET3_PCI) += hw/net/vmxnet3.c > > check-qtest-i386-$(CONFIG_ISA_TESTDEV) = tests/endianness-test$(EXESUF) > check-qtest-i386-y += tests/fdc-test$(EXESUF) > @@ -270,8 +272,6 @@ gcov-files-i386-$(CONFIG_WDT_IB700) += hw/watchdog/watchdog.c hw/watchdog/wdt_ib > check-qtest-i386-y += tests/tco-test$(EXESUF) > check-qtest-i386-y += $(check-qtest-pci-y) > gcov-files-i386-y += $(gcov-files-pci-y) > -check-qtest-i386-$(CONFIG_VMXNET3_PCI) += tests/vmxnet3-test$(EXESUF) > -gcov-files-i386-$(CONFIG_VMXNET3_PCI) += hw/net/vmxnet3.c > gcov-files-i386-y += hw/net/net_rx_pkt.c > gcov-files-i386-y += hw/net/net_tx_pkt.c > check-qtest-i386-$(CONFIG_PVPANIC) += tests/pvpanic-test$(EXESUF) It shouldn't matter much ... it's a x86-only device, so we could also keep it there? Or is there an urgent reason to move it? Thomas
Thomas Huth <thuth@redhat.com> wrote: > On 2018-08-22 11:54, Juan Quintela wrote: >> We protect it with CONFIG_VMXNET3_PCI now, so no need to also put it >> on i386. >> >> Signed-off-by: Juan Quintela <quintela@redhat.com> >> --- >> tests/Makefile.include | 4 ++-- >> 1 file changed, 2 insertions(+), 2 deletions(-) >> >> diff --git a/tests/Makefile.include b/tests/Makefile.include >> index d524e1bdeb..91ffde3103 100644 >> --- a/tests/Makefile.include >> +++ b/tests/Makefile.include >> @@ -247,6 +247,8 @@ check-qtest-pci-$(CONFIG_IVSHMEM_DEVICE) += >> tests/ivshmem-test$(EXESUF) >> gcov-files-pci-$(CONFIG_IVSHMEM_DEVICE) += hw/misc/ivshmem.c >> check-qtest-pci-y += tests/megasas-test$(EXESUF) >> gcov-files-pci-y += hw/scsi/megasas.c >> +check-qtest-$(CONFIG_VMXNET3_PCI) += tests/vmxnet3-test$(EXESUF) >> +gcov-files-$(CONFIG_VMXNET3_PCI) += hw/net/vmxnet3.c >> >> check-qtest-i386-$(CONFIG_ISA_TESTDEV) = tests/endianness-test$(EXESUF) >> check-qtest-i386-y += tests/fdc-test$(EXESUF) >> @@ -270,8 +272,6 @@ gcov-files-i386-$(CONFIG_WDT_IB700) += >> hw/watchdog/watchdog.c hw/watchdog/wdt_ib >> check-qtest-i386-y += tests/tco-test$(EXESUF) >> check-qtest-i386-y += $(check-qtest-pci-y) >> gcov-files-i386-y += $(gcov-files-pci-y) >> -check-qtest-i386-$(CONFIG_VMXNET3_PCI) += tests/vmxnet3-test$(EXESUF) >> -gcov-files-i386-$(CONFIG_VMXNET3_PCI) += hw/net/vmxnet3.c >> gcov-files-i386-y += hw/net/net_rx_pkt.c >> gcov-files-i386-y += hw/net/net_tx_pkt.c >> check-qtest-i386-$(CONFIG_PVPANIC) += tests/pvpanic-test$(EXESUF) > > It shouldn't matter much ... it's a x86-only device, so we could also > keep it there? Or is there an urgent reason to move it? I was trying to minimize the amount that are architecture specific. In this particular case it is defined already on i386-softmmu.mak. So, why should we maintain that info in two places? Anyways, I had to stop soon this "cleanup" because there are things like boot-serial-test that don't work on all the boards that define CONFIG_SERIAL :-( Later, Juan.
On 2018-08-22 12:41, Juan Quintela wrote: > Thomas Huth <thuth@redhat.com> wrote: >> On 2018-08-22 11:54, Juan Quintela wrote: >>> We protect it with CONFIG_VMXNET3_PCI now, so no need to also put it >>> on i386. >>> >>> Signed-off-by: Juan Quintela <quintela@redhat.com> >>> --- >>> tests/Makefile.include | 4 ++-- >>> 1 file changed, 2 insertions(+), 2 deletions(-) >>> >>> diff --git a/tests/Makefile.include b/tests/Makefile.include >>> index d524e1bdeb..91ffde3103 100644 >>> --- a/tests/Makefile.include >>> +++ b/tests/Makefile.include >>> @@ -247,6 +247,8 @@ check-qtest-pci-$(CONFIG_IVSHMEM_DEVICE) += >>> tests/ivshmem-test$(EXESUF) >>> gcov-files-pci-$(CONFIG_IVSHMEM_DEVICE) += hw/misc/ivshmem.c >>> check-qtest-pci-y += tests/megasas-test$(EXESUF) >>> gcov-files-pci-y += hw/scsi/megasas.c >>> +check-qtest-$(CONFIG_VMXNET3_PCI) += tests/vmxnet3-test$(EXESUF) >>> +gcov-files-$(CONFIG_VMXNET3_PCI) += hw/net/vmxnet3.c >>> >>> check-qtest-i386-$(CONFIG_ISA_TESTDEV) = tests/endianness-test$(EXESUF) >>> check-qtest-i386-y += tests/fdc-test$(EXESUF) >>> @@ -270,8 +272,6 @@ gcov-files-i386-$(CONFIG_WDT_IB700) += >>> hw/watchdog/watchdog.c hw/watchdog/wdt_ib >>> check-qtest-i386-y += tests/tco-test$(EXESUF) >>> check-qtest-i386-y += $(check-qtest-pci-y) >>> gcov-files-i386-y += $(gcov-files-pci-y) >>> -check-qtest-i386-$(CONFIG_VMXNET3_PCI) += tests/vmxnet3-test$(EXESUF) >>> -gcov-files-i386-$(CONFIG_VMXNET3_PCI) += hw/net/vmxnet3.c >>> gcov-files-i386-y += hw/net/net_rx_pkt.c >>> gcov-files-i386-y += hw/net/net_tx_pkt.c >>> check-qtest-i386-$(CONFIG_PVPANIC) += tests/pvpanic-test$(EXESUF) >> >> It shouldn't matter much ... it's a x86-only device, so we could also >> keep it there? Or is there an urgent reason to move it? > > I was trying to minimize the amount that are architecture specific. In > this particular case it is defined already on i386-softmmu.mak. So, why > should we maintain that info in two places? Ok, fair. Reviewed-by: Thomas Huth <thuth@redhat.com> > Anyways, I had to stop soon this "cleanup" because there are things like > boot-serial-test that don't work on all the boards that define > CONFIG_SERIAL :-( Which ones are missing? Feel free to contribute small assembler programs there to increase the test coverage ;-) Thomas
Thomas Huth <thuth@redhat.com> wrote: > On 2018-08-22 12:41, Juan Quintela wrote: >> Thomas Huth <thuth@redhat.com> wrote: >>> On 2018-08-22 11:54, Juan Quintela wrote: >>>> We protect it with CONFIG_VMXNET3_PCI now, so no need to also put it >>>> on i386. >>>> >>>> Signed-off-by: Juan Quintela <quintela@redhat.com> >>>> --- >>>> tests/Makefile.include | 4 ++-- >>>> 1 file changed, 2 insertions(+), 2 deletions(-) >>>> >>>> diff --git a/tests/Makefile.include b/tests/Makefile.include >>>> index d524e1bdeb..91ffde3103 100644 >>>> --- a/tests/Makefile.include >>>> +++ b/tests/Makefile.include >>>> @@ -247,6 +247,8 @@ check-qtest-pci-$(CONFIG_IVSHMEM_DEVICE) += >>>> tests/ivshmem-test$(EXESUF) >>>> gcov-files-pci-$(CONFIG_IVSHMEM_DEVICE) += hw/misc/ivshmem.c >>>> check-qtest-pci-y += tests/megasas-test$(EXESUF) >>>> gcov-files-pci-y += hw/scsi/megasas.c >>>> +check-qtest-$(CONFIG_VMXNET3_PCI) += tests/vmxnet3-test$(EXESUF) >>>> +gcov-files-$(CONFIG_VMXNET3_PCI) += hw/net/vmxnet3.c >>>> >>>> check-qtest-i386-$(CONFIG_ISA_TESTDEV) = tests/endianness-test$(EXESUF) >>>> check-qtest-i386-y += tests/fdc-test$(EXESUF) >>>> @@ -270,8 +272,6 @@ gcov-files-i386-$(CONFIG_WDT_IB700) += >>>> hw/watchdog/watchdog.c hw/watchdog/wdt_ib >>>> check-qtest-i386-y += tests/tco-test$(EXESUF) >>>> check-qtest-i386-y += $(check-qtest-pci-y) >>>> gcov-files-i386-y += $(gcov-files-pci-y) >>>> -check-qtest-i386-$(CONFIG_VMXNET3_PCI) += tests/vmxnet3-test$(EXESUF) >>>> -gcov-files-i386-$(CONFIG_VMXNET3_PCI) += hw/net/vmxnet3.c >>>> gcov-files-i386-y += hw/net/net_rx_pkt.c >>>> gcov-files-i386-y += hw/net/net_tx_pkt.c >>>> check-qtest-i386-$(CONFIG_PVPANIC) += tests/pvpanic-test$(EXESUF) >>> >>> It shouldn't matter much ... it's a x86-only device, so we could also >>> keep it there? Or is there an urgent reason to move it? >> >> I was trying to minimize the amount that are architecture specific. In >> this particular case it is defined already on i386-softmmu.mak. So, why >> should we maintain that info in two places? > > Ok, fair. > > Reviewed-by: Thomas Huth <thuth@redhat.com> > >> Anyways, I had to stop soon this "cleanup" because there are things like >> boot-serial-test that don't work on all the boards that define >> CONFIG_SERIAL :-( > > Which ones are missing? Feel free to contribute small assembler programs > there to increase the test coverage ;-) moxie, ti and another two or so. I will try to take another round of serial "cleanups" next week and will come with a better answer. Later, Juan.
diff --git a/tests/Makefile.include b/tests/Makefile.include index d524e1bdeb..91ffde3103 100644 --- a/tests/Makefile.include +++ b/tests/Makefile.include @@ -247,6 +247,8 @@ check-qtest-pci-$(CONFIG_IVSHMEM_DEVICE) += tests/ivshmem-test$(EXESUF) gcov-files-pci-$(CONFIG_IVSHMEM_DEVICE) += hw/misc/ivshmem.c check-qtest-pci-y += tests/megasas-test$(EXESUF) gcov-files-pci-y += hw/scsi/megasas.c +check-qtest-$(CONFIG_VMXNET3_PCI) += tests/vmxnet3-test$(EXESUF) +gcov-files-$(CONFIG_VMXNET3_PCI) += hw/net/vmxnet3.c check-qtest-i386-$(CONFIG_ISA_TESTDEV) = tests/endianness-test$(EXESUF) check-qtest-i386-y += tests/fdc-test$(EXESUF) @@ -270,8 +272,6 @@ gcov-files-i386-$(CONFIG_WDT_IB700) += hw/watchdog/watchdog.c hw/watchdog/wdt_ib check-qtest-i386-y += tests/tco-test$(EXESUF) check-qtest-i386-y += $(check-qtest-pci-y) gcov-files-i386-y += $(gcov-files-pci-y) -check-qtest-i386-$(CONFIG_VMXNET3_PCI) += tests/vmxnet3-test$(EXESUF) -gcov-files-i386-$(CONFIG_VMXNET3_PCI) += hw/net/vmxnet3.c gcov-files-i386-y += hw/net/net_rx_pkt.c gcov-files-i386-y += hw/net/net_tx_pkt.c check-qtest-i386-$(CONFIG_PVPANIC) += tests/pvpanic-test$(EXESUF)
We protect it with CONFIG_VMXNET3_PCI now, so no need to also put it on i386. Signed-off-by: Juan Quintela <quintela@redhat.com> --- tests/Makefile.include | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)