This is mainly about splitting off monitor-related code. There's also a few UI fixes to HMP command pcie_aer_inject_error. One UI issue remains: when the second argument is symbolic (found in table pcie_aer_error_list[]), then any -c is silently ignored. Should it be rejected? Should it override the value from the table?
Markus Armbruster (12): pci: Clean up a few things checkpatch.pl would flag later on pci: Move QMP commands to new hw/pci/pci-qmp-cmds.c pci: Move HMP commands from monitor/ to new hw/pci/pci-hmp-cmds.c pci: Make query-pci stub consistent with the real one pci: Build hw/pci/pci-hmp-cmds.c only when CONFIG_PCI pci: Deduplicate get_class_desc() pci: Move pcibus_dev_print() to pci-hmp-cmds.c pci: Fix silent truncation of pcie_aer_inject_error argument pci: Move HMP command from hw/pci/pcie_aer.c to pci-hmp-cmds.c pci: Inline do_pcie_aer_inject_error() into its only caller pci: Rename hmp_pcie_aer_inject_error()'s local variable @err pci: Improve do_pcie_aer_inject_error()'s error messages hw/pci/pci-internal.h | 25 +++++ include/monitor/hmp.h | 1 + include/sysemu/sysemu.h | 3 - hw/pci/pci-hmp-cmds.c | 234 ++++++++++++++++++++++++++++++++++++++++ hw/pci/pci-qmp-cmds.c | 201 ++++++++++++++++++++++++++++++++++ hw/pci/pci-stub.c | 9 +- hw/pci/pci.c | 226 +------------------------------------- hw/pci/pcie_aer.c | 113 +------------------ monitor/hmp-cmds.c | 107 ------------------ hw/pci/meson.build | 2 + 10 files changed, 476 insertions(+), 445 deletions(-) create mode 100644 hw/pci/pci-internal.h create mode 100644 hw/pci/pci-hmp-cmds.c create mode 100644 hw/pci/pci-qmp-cmds.c -- 2.37.3
