Re: [PATCH 5/6] x86/ptrace: down with test_thread_flag(TIF_IA32)

[Dmitry Safonov <dsafonov@virtuozzo.com>]

On 06/07/2016 12:19 AM, Oleg Nesterov wrote:
> On 06/01, Dmitry Safonov wrote:
>>
>> Note, that this will work only if application has changed it's CS.
>
> So, suppose it changes it's CS and crashes,
>
>> If the application does 32-bit syscall with __USER_CS, ptrace
>> @@ -1355,7 +1355,7 @@ void update_regset_xstate_info(unsigned int size, u64 xstate_mask)
>>  const struct user_regset_view *task_user_regset_view(struct task_struct *task)
>>  {
>>  #ifdef CONFIG_IA32_EMULATION
>> -	if (test_tsk_thread_flag(task, TIF_IA32))
>> +	if (!user_64bit_mode(task_pt_regs(task)))
>>  #endif
>
> then coredump will do fill_elf_header(view->e_machine) and use EM_X86_64
> instead of EM_386, or vice versa...
>
> I simply can't understand is this better or worse, I guess gdb or any
> other tool which looks at this coredump will be confused anyway.

Oleg, thanks for your reply!

At least, gdb will have all registers at that moment.
Firstly, I did it for returning with GET_REGSET corresponding
register set. So ptrace-attaching to a task with 64-bit code
selector would return 64-bit register set and contrariwise.
(as task may use all 64-bit registers, that seems logical
for me even if it has been loaded as 32-bit ELF)

And for coredump -- I guess it will show full register set
on the crash moment and show proper disasm around that place.
And for my purpose -- that's really what I need, to generate
32-bit ELF core file on crash after changing CS.

I did it for C/R of 32-bit application with changing code
selector from 64-bit. So last thing restorer does - unmap
itself (it's pie blob) and sigreturn to application.
So if the application crashes after it - it will be good
to have 32-bit coredump as for the original application.

Thanks,
Dmitry Safonov