On 20/02/2017 13:07, Markus Armbruster wrote: > Paolo Bonzini <[email protected]> writes: > >> On 15/02/2017 14:10, Markus Armbruster wrote: >>> Paolo Bonzini <[email protected]> writes: >>> >>>> On 15/02/2017 13:18, Markus Armbruster wrote: >>>>> Drives defined with if=scsi get connected to buses created with >>>>> -device, unlike other interface types. Deprecate this usage. >>>>> >>>>> There is no good default SCSI HBA for PC machines. Deprecate if=scsi >>>>> there entirely. >>>>> >>>>> Before this series, frontends for -drive if=scsi get created by SCSI >>>>> HBAs. Frontends for other interface types get created by machine >>>>> initialization. >>>>> >>>>> After this series, they get created by machine initialization, except >>>>> for deprecated cases, which get created by new function >>>>> scsi_legacy_handle_cmdline(). >>>>> >>>>> Based on "[PATCH v3 0/8] More sensible default for -drive interface >>>>> type". >>>>> >>>>> v2: >>>>> * Trivially rebased >>>>> * PATCH 2: Suppress warning for default drives >>>>> * PATCH 3: Update user manual and qemu-iotests/051.pc.out >>>>> >>>>> Markus Armbruster (3): >>>>> hw/scsi: Concentrate -drive if=scsi auto-create in one place >>>>> hw: Deprecate -drive if=scsi with non-onboard HBAs >>>>> hw/i386: Deprecate -drive if=scsi with PC machine types >>>>> >>>>> hw/arm/realview.c | 2 +- >>>>> hw/arm/versatilepb.c | 2 +- >>>>> hw/i386/pc.c | 6 +++++ >>>>> hw/scsi/esp-pci.c | 3 --- >>>>> hw/scsi/esp.c | 6 ----- >>>>> hw/scsi/lsi53c895a.c | 10 +++++--- >>>>> hw/scsi/megasas.c | 4 ---- >>>>> hw/scsi/mptsas.c | 4 ---- >>>>> hw/scsi/scsi-bus.c | 56 >>>>> +++++++++++++++++++++++++++++++++++++------ >>>>> hw/scsi/spapr_vscsi.c | 5 ++-- >>>>> hw/scsi/virtio-scsi.c | 8 ------- >>>>> include/hw/pci/pci.h | 2 ++ >>>>> include/hw/scsi/scsi.h | 3 ++- >>>>> qemu-options.hx | 5 ---- >>>>> tests/qemu-iotests/051.pc.out | 12 ++++++---- >>>>> vl.c | 10 ++++++++ >>>>> 16 files changed, 88 insertions(+), 50 deletions(-) >>>>> >>>> >>>> Because of the ESC characters in 051.pc.out, please send the pull >>>> request for this yourself. >>> >>> Is that an Acked-by? >> >> Not a sufficient one because the patch touches files outside my area of >> maintenance, but for hw/scsi you can treat it as one. > > Okay, that covers PATCH 1/3. > > What about PATCH 3/3? It's actually your idea...
Yep, hw/i386 too I guess. Paolo
