Re: [RFC 1/3] x86/signal: add SA_{X32,IA32}_ABI sa_flags

[Andy Lutomirski]

On Mon, Apr 25, 2016 at 1:34 PM, Dmitry Safonov <0x7f454c46@xxxxxxxxx> wrote:
> 2016-04-25 22:20 GMT+03:00 Andy Lutomirski <luto@xxxxxxxxxxxxxx>:
>> On Mon, Apr 25, 2016 at 9:12 AM, Dmitry Safonov <dsafonov@xxxxxxxxxxxxx> wrote:
>>> Introduce new flags that defines which ABI to use on creating sigframe.
>>> Those flags one may set from the userspace, or kernel will set them
>>> according to syscall, which sets handler for a signal.
>>> So that will drop the dependency on TIF_IA32/TIF_X32 flags on syscall deliver.
>>> Those flags will be used only under CONFIG_COMPAT.
>>>
>>> The same way ARM uses sa_flags to differ in which mode deliver signal
>>> for 26-bit applications (look at SA_THIRYTWO).
>>
>> Hmm.  Do we want to make these user-visible at all, or should it be
>> purely an in-kernel thing?
>
> Yes, I'll rework it to not expose to userspace.
> I thought about it as a bonus when did it, but yeah, it's better
> not reveal a new interfaces until they really needed.
> But anyway, I did it for RFC, and I don't know what's better
> for hidden flag: reuse sa_flags or invent in ksig a new hidden
> member only for the kernel?

Either is fine with me.  If you hide it in sa_flags, make sure to mask
it off in the syscalls.

-- 
Andy Lutomirski
AMA Capital Management, LLC