qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [PULL 25/30] Fix wrong behavior of cpu_memory_rw_debug() function in


From: Laszlo Ersek
Subject: Re: [PULL 25/30] Fix wrong behavior of cpu_memory_rw_debug() function in SMM
Date: Mon, 7 Oct 2019 10:29:56 +0200
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101 Thunderbird/52.9.1

Hi Paolo,

(+Peter)

On 10/02/19 18:51, Paolo Bonzini wrote:
> From: Dmitry Poletaev <address@hidden>
> 
> There is a problem, that you don't have access to the data using 
> cpu_memory_rw_debug() function when in SMM. You can't remotely debug SMM mode 
> program because of that for example.
> Likely attrs version of get_phys_page_debug should be used to get correct 
> asidx at the end to handle access properly.
> Here the patch to fix it.
> 
> Signed-off-by: Dmitry Poletaev <address@hidden>
> Signed-off-by: Paolo Bonzini <address@hidden>
> ---
>  target/i386/cpu.c    | 2 +-
>  target/i386/cpu.h    | 3 ++-
>  target/i386/helper.c | 5 ++++-
>  3 files changed, 7 insertions(+), 3 deletions(-)

If it's not too late yet -- I've just pulled the master branch and this patch 
doesn't appear to be on it --, can you please edit the commit message a little?

It would be nice if the commit message included the following two links, to the 
original issue report from Dmitry (both links point to the same message, just 
in different archives):

  "Can not read SMI handler code with cpu_memory_rw_debug while in SMM"
  https://lists.gnu.org/archive/html/qemu-devel/2019-08/msg06039.html
  http://mid.mail-archive.com/address@hidden

(Thank you Dmitry for the patch BTW, I'll probably rely quite a bit on it in 
the future.)

Thank you,
Laszlo

> diff --git a/target/i386/cpu.c b/target/i386/cpu.c
> index 313a2ef..8fcb571 100644
> --- a/target/i386/cpu.c
> +++ b/target/i386/cpu.c
> @@ -6245,7 +6245,7 @@ static void x86_cpu_common_class_init(ObjectClass *oc, 
> void *data)
>  #ifndef CONFIG_USER_ONLY
>      cc->asidx_from_attrs = x86_asidx_from_attrs;
>      cc->get_memory_mapping = x86_cpu_get_memory_mapping;
> -    cc->get_phys_page_debug = x86_cpu_get_phys_page_debug;
> +    cc->get_phys_page_attrs_debug = x86_cpu_get_phys_page_attrs_debug;
>      cc->write_elf64_note = x86_cpu_write_elf64_note;
>      cc->write_elf64_qemunote = x86_cpu_write_elf64_qemunote;
>      cc->write_elf32_note = x86_cpu_write_elf32_note;
> diff --git a/target/i386/cpu.h b/target/i386/cpu.h
> index 033991c..eaa5395 100644
> --- a/target/i386/cpu.h
> +++ b/target/i386/cpu.h
> @@ -1690,7 +1690,8 @@ void x86_cpu_get_memory_mapping(CPUState *cpu, 
> MemoryMappingList *list,
>  
>  void x86_cpu_dump_state(CPUState *cs, FILE *f, int flags);
>  
> -hwaddr x86_cpu_get_phys_page_debug(CPUState *cpu, vaddr addr);
> +hwaddr x86_cpu_get_phys_page_attrs_debug(CPUState *cpu, vaddr addr,
> +                                         MemTxAttrs *attrs);
>  
>  int x86_cpu_gdb_read_register(CPUState *cpu, uint8_t *buf, int reg);
>  int x86_cpu_gdb_write_register(CPUState *cpu, uint8_t *buf, int reg);
> diff --git a/target/i386/helper.c b/target/i386/helper.c
> index 0fa51be..c3a6e4f 100644
> --- a/target/i386/helper.c
> +++ b/target/i386/helper.c
> @@ -715,7 +715,8 @@ void cpu_x86_update_cr4(CPUX86State *env, uint32_t 
> new_cr4)
>  }
>  
>  #if !defined(CONFIG_USER_ONLY)
> -hwaddr x86_cpu_get_phys_page_debug(CPUState *cs, vaddr addr)
> +hwaddr x86_cpu_get_phys_page_attrs_debug(CPUState *cs, vaddr addr,
> +                                         MemTxAttrs *attrs)
>  {
>      X86CPU *cpu = X86_CPU(cs);
>      CPUX86State *env = &cpu->env;
> @@ -725,6 +726,8 @@ hwaddr x86_cpu_get_phys_page_debug(CPUState *cs, vaddr 
> addr)
>      uint32_t page_offset;
>      int page_size;
>  
> +    *attrs = cpu_get_mem_attrs(env);
> +
>      a20_mask = x86_get_a20_mask(env);
>      if (!(env->cr[0] & CR0_PG_MASK)) {
>          pte = addr & a20_mask;
> 




reply via email to

[Prev in Thread] Current Thread [Next in Thread]