Re: xgetbv nondeterminism

From: Andy Lutomirski
Date: Fri Jun 16 2017 - 13:56:54 EST

On Fri, Jun 16, 2017 at 10:44 AM, H.J. Lu <> wrote:
> On Fri, Jun 16, 2017 at 9:38 AM, Andy Lutomirski <luto@xxxxxxxxxx> wrote:
>> On Fri, Jun 16, 2017 at 9:17 AM, H.J. Lu <> wrote:
>>> On Fri, Jun 16, 2017 at 9:01 AM, Andy Lutomirski <luto@xxxxxxxxxx> wrote:
>>>> On Thu, Jun 15, 2017 at 9:34 PM, H.J. Lu <> wrote:
>>>>> On Thu, Jun 15, 2017 at 8:05 PM, Andy Lutomirski <luto@xxxxxxxxxx> wrote:
>>>>>> On Thu, Jun 15, 2017 at 7:17 PM, H.J. Lu <> wrote:
>>>>>>> On Thu, Jun 15, 2017 at 4:28 PM, Andy Lutomirski <luto@xxxxxxxxxx> wrote:
>>>>>>>> On Thu, Jun 15, 2017 at 4:11 PM, H.J. Lu <> wrote:
>>>>>>>>> It is used for lazy binding the first time when an external function is called.
>>>>>>>> Maybe I'm just being dense, but why? What does need to do to
>>>>>>>> resolve a symbol and update the GOT that requires using extended
>>>>>>>> state?
>>>>>>> Since the first 8 vector registers are used to pass function parameters
>>>>>>> and uses vector registers, _dl_runtime_resolve needs to preserve
>>>>>>> the first 8 vector registers when transferring control to
>>>>>> Wouldn't it be faster and more future-proof to recompile the relevant
>>>>>> parts of to avoid using extended state?
>>>>> Are you suggesting not to use vector in
>>>> Yes, exactly.
>>>>> We used to do that
>>>>> several years ago, which leads to some subtle bugs, like
>>>> I don't think x86_64 has the issue that ARM has there. The Linux
>>>> kernel, for example, has always been compiled to not use vector or
>>>> floating point registers on x86 (32 and 64), and it works fine. Linux
>>>> doesn't save extended regs on kernel entry and it doesn't restore them
>>>> on exit.
>>>> I would suggest that be compiled without use of vector
>>>> registers, that the normal lazy binding path not try to save any extra
>>>> regs, and that ifuncs be called through a thunk that saves whatever
>>>> registers need saving, possibly just using XSAVEOPT. After all, ifunc
>>>> is used for only a tiny fraction of symbols.
>>> x86-64 was the only target which used FOREIGN_CALL macros
>>> in, FOREIGN_CALL macros were the cause of race condition
>>> in
>>> Not to save and restore the first 8 vector registers means that
>>> FOREIGN_CALL macros have to be used. We don't want to
>>> do that on x86-64.
>> You're talking about this, right:
>> commit f3dcae82d54e5097e18e1d6ef4ff55c2ea4e621e
>> Author: H.J. Lu <>
>> Date: Tue Aug 25 04:33:54 2015 -0700
>> Save and restore vector registers in x86-64
>> It seems to me that the problem wasn't that the save/restore happened
>> on some of the time -- it was that the save and restore code used a
>> TLS variable to track its own state. Shouldn't it have been a stack
>> variable or even just implicit in the control flow?
> No, it can't use stack variable since _dl_runtime_resolve never
> returns.

I haven't dug all the way through the source, but surely ifuncs are
CALLed, not JMPed to. That means you have a stack somewhere. This
stuff is mostly written in C, and local variables should work just

>> In any case, glibc is effectively doing a foreign call anyway, right?
> No.
>> It's doing the foreign call to itself on every lazy binding
>> resolution, though, which seems quite expensive. I'm saying that it
>> seems like it would be more sensible to do the complicated foreign
>> call logic only when doing the dangerous case, which is when lazy
>> binding calls an ifunc.
>> If I were to rewrite this, I would do it like this:
>> void *call_runtime_ifunc(void (*ifunc)()); // or whatever the
>> signature needs to be
> It is unrelated to IFUNC. This is how external function call works.

External function call to what external function? Are there any calls
to any non-IFUNC external functions that are triggered by runtime

In any event, I still don't understand the issue. The code does this,

GOT points to a stub that transfers control to resolves the symbol (_dl_fixup, I think) patches the GOT jumps to the resolved function

As far as I can tell, the only part of the whole process that might
touch vector registers at all is elf_ifunc_invoke(). Couldn't all the
register saving and restoring be moved to elf_ifunc_invoke()?

>> call_runtime_ifunc would be implemented in asm (or maybe even C!) and
>> would use XSAVEOPT or similar to save the state to a buffer on the
>> stack. Then it would call the ifunc and restore the state. No TLS
>> needed, so there wouldn't be any races. In fact, it would work very
>> much like your current save/restore code, except that it wouldn't need
>> to be as highly optimized because it would be called much less
>> frequently. This should improve performance and could be quite a bit
>> simpler.
>> As an aside, why is saving the first eight registers enough? I don't
>> think there's any particular guarantee that a call through the GOT
>> uses the psABI, is there? Compilers can and do produce custom calling
>> conventions, and ISTM that some day a compiler might do that between
>> DSOs. Or those DSOs might not be written in C in the first place.
> The result is undefined if psABI isn't followed.

That's unfortunate. Does that mean that, if you use a custom ABI
across DSO boundaries, you have to use -z now?

> --
> H.J.