On Wed, Feb 7, 2024 at 7:58 PM Philippe Mathieu-Daudé <phi...@linaro.org> wrote: > > @@ -6,8 +6,7 @@ config MALTA > > > > config MIPSSIM > > bool > > - select ISA_BUS > > - select SERIAL_ISA > > + select SERIAL > > Hmm there is an ISA bus which can be exposed with: > > -- >8 -- > diff --git a/hw/mips/mipssim.c b/hw/mips/mipssim.c > index 16af31648e..a1a4688861 100644 > --- a/hw/mips/mipssim.c > +++ b/hw/mips/mipssim.c > @@ -209,8 +209,9 @@ mips_mipssim_init(MachineState *machine) > /* Register 64 KB of ISA IO space at 0x1fd00000. */ > memory_region_init_alias(isa, NULL, "isa_mmio", > get_system_io(), 0, 0x00010000); > memory_region_add_subregion(get_system_memory(), 0x1fd00000, isa); > + isa_bus_new(NULL, get_system_memory(), get_system_io(), &error_abort);
Quoting you from https://lists.gnu.org/archive/html/qemu-devel/2020-09/msg08752.html, "there is an ISA MMIO space mapped at 0x1fd00000, however this is not a real ISA bus (no ISA IRQ)". If mipssim cannot support "-device isa-serial" as a replacement for "-serial", there's no reason for it to expose the bus. In the end, -device support is the main thing that an ISA bus provides over sysbus, and if it cannot work due to the missing interrupts, I think this patch is correct. I can add a comment: diff --git a/hw/mips/mipssim.c b/hw/mips/mipssim.c index 01e323904d9..47b37b454e7 100644 --- a/hw/mips/mipssim.c +++ b/hw/mips/mipssim.c @@ -204,7 +204,11 @@ mips_mipssim_init(MachineState *machine) cpu_mips_irq_init_cpu(cpu); cpu_mips_clock_init(cpu); - /* Register 64 KB of ISA IO space at 0x1fd00000. */ + /* + * Register 64 KB of ISA IO space at 0x1fd00000. But without interrupts + * (except for the hardcoded serial port interrupt) -device cannot work, + * so do not expose the ISA bus to the user. + */ memory_region_init_alias(isa, NULL, "isa_mmio", get_system_io(), 0, 0x00010000); memory_region_add_subregion(get_system_memory(), 0x1fd00000, isa); Paolo > /* > * A single 16450 sits at offset 0x3f8. It is attached to > * MIPS CPU INT2, which is interrupt 4. > --- >
