On Wed, Mar 10, 2021 at 08:12:54PM +0100, Philippe Mathieu-Daudé wrote: > No need to create a local ISA I/O MemoryRegion, use get_system_io(). > > This partly reverts commit 5c63bcf7501527b844f61624957bdba254d75bfc.
I think it's not a clean revert of that, see below. > > Signed-off-by: Philippe Mathieu-Daudé <f4...@amsat.org> > --- > hw/mips/jazz.c | 6 ++---- > 1 file changed, 2 insertions(+), 4 deletions(-) > > diff --git a/hw/mips/jazz.c b/hw/mips/jazz.c > index 1a0888a0fd5..9ac9361b7eb 100644 > --- a/hw/mips/jazz.c > +++ b/hw/mips/jazz.c > @@ -158,7 +158,6 @@ static void mips_jazz_init(MachineState *machine, > rc4030_dma *dmas; > IOMMUMemoryRegion *rc4030_dma_mr; > MemoryRegion *isa_mem = g_new(MemoryRegion, 1); > - MemoryRegion *isa_io = g_new(MemoryRegion, 1); > MemoryRegion *rtc = g_new(MemoryRegion, 1); > MemoryRegion *i8042 = g_new(MemoryRegion, 1); > MemoryRegion *dma_dummy = g_new(MemoryRegion, 1); > @@ -259,11 +258,10 @@ static void mips_jazz_init(MachineState *machine, > memory_region_add_subregion(address_space, 0x8000d000, dma_dummy); > > /* ISA bus: IO space at 0x90000000, mem space at 0x91000000 */ > - memory_region_init(isa_io, NULL, "isa-io", 0x00010000); > memory_region_init(isa_mem, NULL, "isa-mem", 0x01000000); > - memory_region_add_subregion(address_space, 0x90000000, isa_io); > + memory_region_add_subregion(address_space, 0x90000000, get_system_io()); The old code has an alias created just for adding subregion into address_space: - /* ISA IO space at 0x90000000 */ - memory_region_init_alias(isa, NULL, "isa_mmio", - get_system_io(), 0, 0x01000000); - memory_region_add_subregion(address_space, 0x90000000, isa); - isa_mem_base = 0x11000000; While you didn't revert that part. Maybe that's the issue? -- Peter Xu
