[LLVMdev] Fw: Compile function with limited set of registers? Jump to another function?

James Williams junk at giantblob.com
Fri Feb 4 03:37:19 PST 2011 

Hi Sam,

Thanks for this + sorry for not replying earlier.

I'm afraid I'm not sure I've understood your advice. I have a working
implementation for interfaces with inheritance - I use constant negative
indexes into class vtables for class methods that implement methods in
interfaces. These indexes are allocated at compile time either via graph
coloring (if whole program is compiled as a unit) or by hashing method
signatures (if parts of program are compiled separately). Any collisions
arising from the hashing method are resolved on method call by thunks in any
vtable slots hashed to by more than one selector.

Advantages of this mechanism are it's as fast as regular virtual method
dispatch (except in the method signature hash collision case) and that there
is no difference between calls through objects and calls through interfaces
(in fact there are no interfaces as such at runtime, only references to
regular objects)

What I'm looking to do is speed up the thunk that's called when two or more
interface method signatures hash to the same slot in the vtable. Currently
it just does a linear search over a compile time generated table of selector
ids->method address. I think common improvements are using a self ordering
list or using a binary search.

My conclusion is that I could probably use LLVM to generate a thunk to
resolve each vtable collision to do a binary search over the expected method
signatures, providing I use varargs calling convention. I'll need to do some
profiling to determine if this is worthwhile, given the overhead of adding
an additional parameter to both all calls to methods interfaces and methods
in classes that implement methods in interfaces.

-- James.

On 31 January 2011 22:28, Samuel Crow <samuraileumas at yahoo.com> wrote:

> Forgot to cc the list.
>
>
> ----- Forwarded Message ----
> *From:* Samuel Crow <samuraileumas at yahoo.com>
> *To:* James Williams <junk at giantblob.com>
> *Sent:* Mon, January 31, 2011 4:27:45 PM
> *Subject:* Re: [LLVMdev] Compile function with limited set of registers?
> Jump to another function?
>
> Hi James,
>
> If you're looking for a way to implement interface inheritance, I think the
> easiest way is to use double-indirection.  It's slow but it does the job.
>  You just use an array to hold the interface and have each entry point to
> the vtable entry for the appropriate functions in the class.  Then you can
> treat the interface as a read-only class once the constructor is run and the
> interface will assign its values when the class is allocated before the
> invocation of the actual constructor.
>
> The faster way to implement interfaces is with generics but that generates
> bigger code.
>
> I hope this answers your question,
>
> --Sam
>
>
> *From:* James Williams <junk at giantblob.com>
> *To:* joshuawarner32 at gmail.com
> *Cc:* llvmdev at cs.uiuc.edu
> *Sent:* Mon, January 31, 2011 4:01:24 PM
> *Subject:* Re: [LLVMdev] Compile function with limited set of registers?
> Jump to another function?
>
> Thanks, that's a good idea - I'll have a look through the VMKit source.
>
> -- James
>
> On 31 January 2011 21:39, Joshua Warner <joshuawarner32 at gmail.com> wrote:
>
>> Hi James,
>>
>> I see the problem now.  You might look at VMKit (a Java VM build with the
>> LLVM JIT compiler)  - I would expect it uses a similar method for resolving
>> interface calls (the method, if I understand it correctly, is well-known in
>> the Java world).
>>
>> I've CC'd the main dev behind VMKit - he might be able to lend some
>> insight.
>>
>> --Joshua
>>
>> On Mon, Jan 31, 2011 at 2:24 PM, James Williams <junk at giantblob.com>wrote:
>>
>>> Hi Joshua,
>>>
>>> Thanks - I was hoping that would be the case.
>>>
>>> However, I've had a think about this since I posted to the list and I
>>> believe the only way to handle these issues safely in LLVM IR would be to
>>> define the thunk as varargs. I'm not sure how well LLVM handles varargs but
>>> ideally it would all compile down to nothing since the parameters to the the
>>> thunk would be in the same registers/stack locations as required by the
>>> target method.
>>>
>>> Unfortunately, varargs has some downsides: there's the additional
>>> overhead for the extra hidden parameter to every interface method call for
>>> the parameter count plus it doesn't (I don't think) support tail calls.
>>>
>>> -- James
>>>
>>>
>>> On 27 January 2011 17:37, Joshua Warner <joshuawarner32 at gmail.com>wrote:
>>>
>>>> Hi James,
>>>>
>>>> I'm no expert in LLVM IR, but I think that just encoding each *actual*
>>>> method invocation in the thunk as a tail call would work.  This would
>>>> require trusting that LLVM passes / code generators will translate down to a
>>>> jump, as is normal.  If the passes / code generators are smart, I see no
>>>> reason that LLVM wouldn't emit code that fits your requirements.  Either
>>>> way, you know that your thunk will be correct - it just might not be as
>>>> efficient as you want.
>>>>
>>>> I would suggest experimenting with generating a thunk this way, and look
>>>> at the resultant target assembly to make sure it's doing what you want.
>>>>
>>>> -Joshua
>>>>
>>>>  On Tue, Jan 25, 2011 at 2:04 AM, James Williams <junk at giantblob.com>wrote:
>>>>
>>>>>  Hi,
>>>>>
>>>>> Can anyone tell me, is it possible to express in LLVM IR:
>>>>>   - that, for a specific function, register allocator can use only
>>>>> limited set of registers? (specifically, cannot touch any registers that
>>>>> might contain parameters)
>>>>>   - that stack can't be touched? (or at least must balance on exit from
>>>>> thunk)
>>>>>   - jump, not call, to another function leaving any received parameters
>>>>> unchanged in registers and on stack?
>>>>>
>>>>> Thanks,
>>>>> -- James Williams
>>>>>
>>>>> Background:
>>>>>
>>>>> I'm looking for some advice on implementing thunks required by my
>>>>> language's interface call mechanism. This is a fairly conventional
>>>>> arrangement where method selectors in interfaces are hashed to determine
>>>>> their index within a vtable and hash collisions are disambiguated at runtime
>>>>> by a thunk, which determines which method to call from a selector id passed
>>>>> as the first method parameter.
>>>>>
>>>>> I'm currently using a single thunk (written in assembly) for all
>>>>> collisions that walks a table to determine what method to call. This works
>>>>> but it's inefficient and requires the a hand written thunk for each
>>>>> supported target.
>>>>>
>>>>> I'd like to instead generate IR for a specific thunk for each vtable
>>>>> collisoin that does a binary search of possible selectors because this will
>>>>> avoid some pointer dereferences and an additional indirect call.
>>>>>
>>>>> The problem is that a thunk may need to decide between methods with
>>>>> different signatures without disturbing parameters in registers and on the
>>>>> stack and then jump to, rather than call, another function:
>>>>>
>>>>> interface X:
>>>>>   method A(a, b)
>>>>>
>>>>> interface Y:
>>>>>   method B(c, d, e)
>>>>>
>>>>> class Z implements X, y:
>>>>>   method A(a, b) ...
>>>>>   method B(c, d, e) ...
>>>>>
>>>>> X.A + Y.B happen to hash to same vtable index, say -3
>>>>>
>>>>> This would require a thunk something like:
>>>>>
>>>>> vtable[-3] =
>>>>>   thunk_Z_AorB(selector_id, ...)
>>>>>     // binary search for matching selector id:
>>>>>     if selector_id <= selector_Z_A then
>>>>>       Z.A(selector_id, ...)
>>>>>     else
>>>>>       Z.B(selector_id, ...)
>>>>>     fi
>>>>>
>>>>> which would ideally would compile on x64 to something like:
>>>>>
>>>>> thunk_Z_AorB:
>>>>>   cmp $selector_Z_A, %rdi
>>>>>   jle Z.A
>>>>>   jmp Z.B
>>>>>
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> LLVM Developers mailing list
>>>>> LLVMdev at cs.uiuc.edu         http://llvm.cs.uiuc.edu
>>>>> http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev
>>>>>
>>>>>
>>>>
>>>
>>
>
>
>
> _______________________________________________
> LLVM Developers mailing list
> LLVMdev at cs.uiuc.edu         http://llvm.cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20110204/255f0f46/attachment.html>

More information about the llvm-dev mailing list