Re: [tip:x86/x32] x32: Fix alignment fail in struct compat_siginfo

[H.J. Lu]

On Wed, Apr 18, 2012 at 11:21 AM, H. Peter Anvin <hpa@xxxxxxxxx> wrote:
> On 04/18/2012 11:15 AM, H.J. Lu wrote:
>>
>> In x32 glibc, I have
>>
>> # if compile for X32
>> /* si_utime and si_stime must be 4 byte aligned for x32 to match the
>>    kernel.  We align siginfo_t to 8 bytes so that si_utime and si_stime
>>    are actually aligned to 8 bytes since their offsets are multiple of
>>    8 bytes.  */
>> typedef __clock_t __attribute__ ((__aligned__ (4))) __sigchld_clock_t;
>> #  define __SI_ALIGNMENT __attribute__ ((__aligned__ (8)))
>> # else
>> typedef __clock_t __sigchld_clock_t;
>> #  define __SI_ALIGNMENT
>> # endif
>>
>> typedef struct siginfo
>>   {
>> ...
>>       /* SIGCHLD.  */
>>       struct
>>         {
>>           __pid_t si_pid;     /* Which child.  */
>>           __uid_t si_uid;     /* Real user ID of sending process.  */
>>           int si_status;      /* Exit value or signal.  */
>>           __sigchld_clock_t si_utime;
>>           __sigchld_clock_t si_stime;
>>         } _sigchld;
>> ...
>>   } siginfo_t __SI_ALIGNMENT;
>>
>
> So in other words, a separate type just for this.  However, the question

Yes.

> still holds: could we simply change __kernel_clock_t for x32 and be done
> with it?  It would affect the alignment of struct tms, I guess; it
> wouldn't really affect the kernel directly since the kernel gets a
> pointer straight from userspace for times(2), but it would kind of mess
> up a userspace which uses __kernel_clock_t directly.

That is true.

> Urk, okay...
>

This is the best approach I can think of.

-- 
H.J.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/