similar to: [LLVMdev] Nested functions implementation

Does anyone know if LLVM will soon have a dynamic optimization infrastructure, namely support to generate trampolines back to the JIT/execution manager that get run when a piece of code's frequency exceeds a threshold? I see that there already is a relinkAndRecompileFunction in the JIT, which will be useful. Rather than roll my own mechanisms and have to worry about jumping back to code that

Hi! Attached set of patches splits llvm.init.trampoline into an "init" phase and an "adjust" phase, as discussed on the "Go on dragonegg" thread. Thanks! -- Sanjoy Das http://playingwithpointers.com -------------- next part -------------- A non-text attachment was scrubbed... Name: 0001-Split-intrinsics-and-DAG-nodes.patch Type: text/x-diff Size: 8808 bytes Desc:

Fellow developers, I'm parallelizing loops to be called by pthread. The thread body that I pass to pthread_create looks like define i8* @loop1({ i32*, i32* }* nest %parent_frame, i8* %arg) parent_frame is pointer to shared variables in original function 0x00007f0de11c41f0: mov (%r10),%rax 0x00007f0de11c41f3: cmpl $0x63,(%rax) 0x00007f0de11c41f6: jg 0x7f0de11c420c

Thanks much sir Ease I think mainly adding a macvtap is pretty quick, performant and works. And last but definitely not least, ignorance of other quick easy solutions. Well, also macvtap works on older hardware where I don't have physical functions to passthrough via sr-iov, that is what you are pointing to with "macvtaps in the most outer one VM and pass them thru to inner layer

Hi, > I'm parallelizing loops to be called by pthread. The thread body that I pass > to pthread_create looks like > > define i8* @loop1({ i32*, i32* }* nest %parent_frame, i8* %arg) > parent_frame is pointer to shared variables in original function > > 0x00007f0de11c41f0: mov (%r10),%rax > 0x00007f0de11c41f3: cmpl $0x63,(%rax) > 0x00007f0de11c41f6:

Hello all, hope all is well this maybe outside of libvirt-users.... Can you nest macvtap devices to ultimately receive a real routable ip on the nested vm? I have a nested vm up and running. Both vm and nested vm are centos 7 on arch linux host. The first vm uses a macvtap in bridge mode receives dhcp from an external dhcp server. I start the second vm and dhclient hangs and never receives an

Thanks again very much sir Will definitely try this path and report back On Mon, Jul 11, 2016 at 11:59 AM, Michal Privoznik <mprivozn@redhat.com> wrote: > On 11.07.2016 17:46, jsl6uy js16uy wrote: > > Thanks much sir > > Ease I think mainly > > adding a macvtap is pretty quick, performant and works. And last but > > definitely not least, ignorance of other quick

On Wed, Oct 13, 2010 at 2:41 AM, Duncan Sands <baldrick at free.fr> wrote: > Hi Nick, > >> I have no idea how to make use of llvm.invariant to help >> devirtualization. If no use or other use case for them can be found, >> maybe they should be removed. > > the dragonegg plugin uses llvm.invariant to help the code generators > optimize the code it produces for

Taking the address of a function inhibits optimisations for that function. Essentially any ABI changes are unavailable if we can't adjust the call site to match. The case of interest here is when a given function is called directly and indirectly, and we don't want the latter to impose a cost on the former. One approach to avoid the ABI constraint cost is to extract/outline the body of an

Was there already some reflection about how to lower the concept of nested functions (and the corresponding static links) into llvm? Dirk Muysers.

Signed-off-by: Christoph Egger <Christoph.Egger@amd.com> -- ---to satisfy European Law for business letters: Advanced Micro Devices GmbH Einsteinring 24, 85609 Dornach b. Muenchen Geschaeftsfuehrer: Alberto Bozzo, Andrew Bowd Sitz: Dornach, Gemeinde Aschheim, Landkreis Muenchen Registergericht Muenchen, HRB Nr. 43632 _______________________________________________ Xen-devel mailing list

On Feb 15, 2008, at 2:29 PM, Andrew Lenharth wrote: > On 2/15/08, Andrew Lenharth <andrewl at lenharth.org> wrote: >> I'll take a hack at the front end support for >> __sync_synchronize after this goes in. > > This is the gcc side of the patch. Thanks for tackling this Andrew. Please prepare a patch for LangRef.html that explains what this thing does :). What

I think we’re talking cross-purposes at this point. Let me try to step back and concisely restate what I’m trying to do: - I am writing a frontend for a high-level language and I emit LLVM IR. - I control the TBAA tree. I don’t use it to express the high-level language’s types, since that would make absolutely no sense. JavaScript doesn’t have types. But by the time my compiler is generating

> As an optimization for shallow nested functions (e.g., 3 levels or less), > it seems to me you could just avoid the stack walking entirely and add > $k-1$ arguments to each function at level $k$, i.e., at most 2 arguments > in all. This may even be an easy first implementation. I use this method with filtering unused in nested function args and local vars in my YAFL frontend

On Nov 22, 2004, at 12:34 PM, Vladimir Merzliakov wrote: >> As an optimization for shallow nested functions (e.g., 3 levels or >> less), it seems to me you could just avoid the stack walking entirely >> and add $k-1$ arguments to each function at level $k$, i.e., at most >> 2 arguments in all. This may even be an easy first implementation. > > I use this method

>> I use this method with filtering unused in nested function args and local >> vars >> in my YAFL frontend (not finished :(( - at this moment generate LLVM >> bytecode but doesn't have all runtime suport code writed) > > So do you mean you pass static links for parent function stack frames as > arguments? How do you generate them? > Sorry, I must more

On 08.07.2016 22:23, jsl6uy js16uy wrote: > Hello all, hope all is well > > this maybe outside of libvirt-users.... > Can you nest macvtap devices to ultimately receive a real routable ip on > the nested vm? > I have never tested it, but why do you want to nest macvtaps? Why not have macvtaps in the most outer one VM and pass them thru to inner layer VMs? Michal

TBAA MDNodes should describe the type hierarchy (with struct-path aware TBAA, it is a type DAG). It sounds okay to me to add !tbaa.call to function calls to describe which fields of the type system the call is touching. One example can be !tbaa.call {read a list of tag nodes, write a list of tag nodes}. Thanks, Manman On Mon, Oct 7, 2013 at 11:49 PM, Daniel Berlin <dberlin at dberlin.org>

On 11.07.2016 17:46, jsl6uy js16uy wrote: > Thanks much sir > Ease I think mainly > adding a macvtap is pretty quick, performant and works. And last but > definitely not least, ignorance of other quick easy solutions. > Well, also macvtap works on older hardware where I don't have physical > functions to passthrough via sr-iov, that is what you are pointing to with >

Hi! Attached patches split init.trampoline into adjust.trampoline and init.trampoline, like in gcc. As mentioned in the previous mail, I've not made a documentation patch, since I'm not sure about what the documented semantics of llvm.adjust.trampoline should be. Thanks! -- Sanjoy Das http://playingwithpointers.com -------------- next part -------------- A non-text attachment was