[PATCH] compat: Fix endian issue in union sigval

[Chris Metcalf]

On 2/13/2015 5:44 AM, Catalin Marinas wrote:
> On Fri, Feb 13, 2015 at 04:00:43PM +0800, Bamvor Jian Zhang wrote:
>> On 2015/2/11 23:40, Catalin Marinas wrote:
>>> On Wed, Feb 11, 2015 at 07:22:08PM +0800, Bamvor Jian Zhang wrote:
>>>> On 2015/2/10 20:27, Catalin Marinas wrote:
>>>>> On Tue, Feb 10, 2015 at 10:10:11AM +0000, Zhang Jian(Bamvor) wrote:
>> ...
>>> The native sigval_t is also a union but on 64-bit big endian, the
>>> sival_int overlaps with the most significant 32-bit of the sival_ptr.
>>> So reading sival_int would always be 0. When the compat siginfo is
>>> copied to user, arm64 reads the native sival_ptr (si_ptr) and converts
>>> it to the compat one, getting the correct 32-bit value. However, other
>>> architectures access sival_int (si_int) instead which breaks with your
>>> get_compat_sigevent() changes.
>
>> tile, s390:          arch/xxx/kernel/compat_signal.c
>
> tile seems to be bi-endian, though I couldn't see a Kconfig option, nor
> something defining __BIG_ENDIAN__ in headers or Makefile. I guess it's
> coming from the compiler directly.

Yes, we just pick up the compiler's __BIG_ENDIAN__ if specified.

> Anyway, on big endian tile, I we have
> the same issue as on big endian arm64.
>
> I think it's only tile that needs fixing for big endian, something like
> the arm64 patch below:
>
>> diff --git a/arch/arm64/kernel/signal32.c b/arch/arm64/kernel/signal32.c
>> index e299de396e9b..32601939a3c8 100644
>> --- a/arch/arm64/kernel/signal32.c
>> +++ b/arch/arm64/kernel/signal32.c
>> @@ -154,8 +154,7 @@ int copy_siginfo_to_user32(compat_siginfo_t __user *to, const siginfo_t *from)
>>   	case __SI_TIMER:
>>   		 err |= __put_user(from->si_tid, &to->si_tid);
>>   		 err |= __put_user(from->si_overrun, &to->si_overrun);
>> -		 err |= __put_user((compat_uptr_t)(unsigned long)from->si_ptr,
>> -				   &to->si_ptr);
>> +		 err |= __put_user(from->si_int, &to->si_int);
>>   		break;
>>   	case __SI_POLL:
>>   		err |= __put_user(from->si_band, &to->si_band);
>> @@ -184,7 +183,7 @@ int copy_siginfo_to_user32(compat_siginfo_t __user *to, const siginfo_t *from)
>>   	case __SI_MESGQ: /* But this is */
>>   		err |= __put_user(from->si_pid, &to->si_pid);
>>   		err |= __put_user(from->si_uid, &to->si_uid);
>> -		err |= __put_user((compat_uptr_t)(unsigned long)from->si_ptr, &to->si_ptr);
>> +		err |= __put_user(from->si_int, &to->si_int);
>>   		break;
>>   	case __SI_SYS:
>>   		err |= __put_user((compat_uptr_t)(unsigned long)

I must be confused here, but I don't see that these do anything different.

If we are writing 32 bits to to->si_ptr or to->si_int, either way the high 32 bits
are irrelevant.  So whether we read it from from->si_ptr and massage the high bits,
or just read it from from->si_int as a straight-up 32-bit quantity, either way it
seems we should end up writing the same bits to userspace.

I would understand the argument if we were overlaying the si_ptr/si_int union
from a kernel-side siginfo_t where si_ptr and si_int are different sizes
onto userspace, but it doesn't seem we ever do that.

All that said, it certainly seems like the si_int version is simpler, so I don't have
a problem with switching to it, but I don't see how it fixes a problem.

-- 
Chris Metcalf, EZChip Semiconductor
http://www.ezchip.com