Hi James,<div><br></div><div>Joshua is right, what you're trying to accomplish is quite known in the Java VM world (<a href="http://domino.research.ibm.com/comm/research_people.nsf/pages/dgrove.oopsla01.html">http://domino.research.ibm.com/comm/research_people.nsf/pages/dgrove.oopsla01.html</a>).</div>
<div><br></div><div>In order to express the "thunk" code in LLVM you need a full control of how registers are used (because otherwise they would mess up with the arguments). I haven't investigated enough to know if that's possible today in LLVM (I think it wasn't a few years ago when I added the optimization). So what I ended up with was a generic IR that walks the interface table of an object and detects collisions. The IR was inline in the caller to make sure that arguments in registers are not overriden.</div>
<div><br></div><div>This may not be the best approach today, but I believe it was the easiest way to have something efficient at that point.</div><div><br></div><div>Cheers,</div><div>Nicolas<br><br><div class="gmail_quote">
On Mon, Jan 31, 2011 at 11:01 PM, James Williams <span dir="ltr"><<a href="mailto:junk@giantblob.com">junk@giantblob.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
Thanks, that's a good idea - I'll have a look through the VMKit source.<br><font color="#888888"><br>-- James</font><div><div></div><div class="h5"><br><br><div class="gmail_quote">On 31 January 2011 21:39, Joshua Warner <span dir="ltr"><<a href="mailto:joshuawarner32@gmail.com" target="_blank">joshuawarner32@gmail.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0pt 0pt 0pt 0.8ex;border-left:1px solid rgb(204, 204, 204);padding-left:1ex">Hi James,<div><br><div><div>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).</div>
<div><br></div><div>I've CC'd the main dev behind VMKit - he might be able to lend some insight.</div><div><br></div><font color="#888888"><div>--Joshua</div></font><div><div></div><div><br><div class="gmail_quote">
On Mon, Jan 31, 2011 at 2:24 PM, James Williams <span dir="ltr"><<a href="mailto:junk@giantblob.com" target="_blank">junk@giantblob.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0pt 0pt 0pt 0.8ex;border-left:1px solid rgb(204, 204, 204);padding-left:1ex">Hi Joshua,<br><br>Thanks - I was hoping that would be the case.<br><br>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.<br><br>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.<br><font color="#888888"><br>-- James</font><div><div></div><div><br> <br><div class="gmail_quote">On 27 January 2011 17:37, Joshua Warner <span dir="ltr"><<a href="mailto:joshuawarner32@gmail.com" target="_blank">joshuawarner32@gmail.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0pt 0pt 0pt 0.8ex;border-left:1px solid rgb(204, 204, 204);padding-left:1ex">Hi James,<div><br></div><div>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.</div>
<div><br></div><div>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.</div><div><br></div><div>-Joshua<br><br><div class="gmail_quote">
<div><div></div><div>
On Tue, Jan 25, 2011 at 2:04 AM, James Williams <span dir="ltr"><<a href="mailto:junk@giantblob.com" target="_blank">junk@giantblob.com</a>></span> wrote:<br></div></div><blockquote class="gmail_quote" style="margin:0pt 0pt 0pt 0.8ex;border-left:1px solid rgb(204, 204, 204);padding-left:1ex">
<div><div></div><div>
Hi,<br><br>Can anyone tell me, is it possible to express in LLVM IR:<br> - that, for a specific function, register allocator can use only limited set of registers? (specifically, cannot touch any registers that might contain parameters)<br>
- that stack can't be touched? (or at least must balance on exit from thunk)<br> - jump, not call, to another function leaving any received parameters unchanged in registers and on stack?<br><br>Thanks,<br>-- James Williams<br>
<br>Background:<br><br>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.<br>
<br>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.<br>
<br>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.<br>
<br>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:<br><br>interface X:<br>
method A(a, b)<br><br>interface Y:<br> method B(c, d, e)<br><br>class Z implements X, y:<br> method A(a, b) ...<br> method B(c, d, e) ...<br><br>X.A + Y.B happen to hash to same vtable index, say -3<br><br>This would require a thunk something like:<br>
<br>vtable[-3] = <br> thunk_Z_AorB(selector_id, ...)<br> // binary search for matching selector id:<br> if selector_id <= selector_Z_A then<br> Z.A(selector_id, ...)<br> else<br> Z.B(selector_id, ...)<br>
fi<br><br>which would ideally would compile on x64 to something like:<br><br>thunk_Z_AorB:<br> cmp $selector_Z_A, %rdi<br> jle Z.A<br> jmp Z.B <br><br><br>
<br></div></div>_______________________________________________<br>
LLVM Developers mailing list<br>
<a href="mailto:LLVMdev@cs.uiuc.edu" target="_blank">LLVMdev@cs.uiuc.edu</a> <a href="http://llvm.cs.uiuc.edu" target="_blank">http://llvm.cs.uiuc.edu</a><br>
<a href="http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev" target="_blank">http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev</a><br>
<br></blockquote></div><br></div>
</blockquote></div><br>
</div></div></blockquote></div><br></div></div></div></div>
</blockquote></div><br>
</div></div></blockquote></div><br></div>