Re: [PATCH-for-8.0 2/4] gdbstub: Use vaddr type for generic insert/remove_breakpoint() API

[Peter Maydell]

On Wed, 7 Dec 2022 at 17:42, Philippe Mathieu-Daudé <philmd@linaro.org> wrote:
>
> Both insert/remove_breakpoint() handlers are used in system and
> user emulation. We can not use the 'hwaddr' type on user emulation,
> we have to use 'vaddr' which is defined as "wide enough to contain
> any #target_ulong virtual address".
>
> Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
> ---
>  gdbstub/internals.h        | 6 ++++--
>  include/sysemu/accel-ops.h | 6 +++---
>  2 files changed, 7 insertions(+), 5 deletions(-)
>
> diff --git a/gdbstub/internals.h b/gdbstub/internals.h
> index eabb0341d1..b23999f951 100644
> --- a/gdbstub/internals.h
> +++ b/gdbstub/internals.h
> @@ -9,9 +9,11 @@
>  #ifndef _INTERNALS_H_
>  #define _INTERNALS_H_
>
> +#include "exec/cpu-common.h"
> +
>  bool gdb_supports_guest_debug(void);
> -int gdb_breakpoint_insert(CPUState *cs, int type, hwaddr addr, hwaddr len);
> -int gdb_breakpoint_remove(CPUState *cs, int type, hwaddr addr, hwaddr len);
> +int gdb_breakpoint_insert(CPUState *cs, int type, vaddr addr, vaddr len);
> +int gdb_breakpoint_remove(CPUState *cs, int type, vaddr addr, vaddr len);
>  void gdb_breakpoint_remove_all(CPUState *cs);
>
>  #endif /* _INTERNALS_H_ */
> diff --git a/include/sysemu/accel-ops.h b/include/sysemu/accel-ops.h
> index 8cc7996def..30690c71bd 100644
> --- a/include/sysemu/accel-ops.h
> +++ b/include/sysemu/accel-ops.h
> @@ -10,7 +10,7 @@
>  #ifndef ACCEL_OPS_H
>  #define ACCEL_OPS_H
>
> -#include "exec/hwaddr.h"
> +#include "exec/cpu-common.h"
>  #include "qom/object.h"
>
>  #define ACCEL_OPS_SUFFIX "-ops"
> @@ -48,8 +48,8 @@ struct AccelOpsClass {
>
>      /* gdbstub hooks */
>      bool (*supports_guest_debug)(void);
> -    int (*insert_breakpoint)(CPUState *cpu, int type, hwaddr addr, hwaddr 
> len);
> -    int (*remove_breakpoint)(CPUState *cpu, int type, hwaddr addr, hwaddr 
> len);
> +    int (*insert_breakpoint)(CPUState *cpu, int type, vaddr addr, vaddr len);
> +    int (*remove_breakpoint)(CPUState *cpu, int type, vaddr addr, vaddr len);
>      void (*remove_all_breakpoints)(CPUState *cpu);
>  };

If you're changing the prototype of these methods on AccelOpsClass
don't you also want to change the implementations, eg tcg_breakpoint_insert()?
Interestingly that function calls cpu_breakpoint_insert() which
already takes a 'vaddr' rather than a 'hwaddr'.

In looking at this I discovered some rather confusing gdbstub behaviour:
if you use the qemu.PhyMemMode custom gdb flag to put the stub into
"physical memory mode", data reads and writes are done on physical
addresses, but breakpoints and watchpoints continue to take virtual
addresses.

But at any rate given that currently breakpoints are always on virtual
addresses, vaddr is definitely the right type here and probably all
the way down through the callstack.

thanks
-- PMM