similar to: [LLVMdev] Proposal: "load linked" and "store conditional" atomic instructions

Hi Philip, On 29 May 2014 17:03, Philip Reames <listmail at philipreames.com> wrote: > I have some reservations about this proposal. I don't have anything > particularly concrete, but the idea of supporting both LL/SC and atomicrwm > in the IR concerns me from a complexity perspective. Well, I'll start by saying my particular optimisation use case looks like it's not

Quoting Tim Northover <t.p.northover at gmail.com>: > On 24 April 2015 at 13:53, Tyler Denniston <tyler at csail.mit.edu> wrote: >> I'm wondering how I can create an atomic load and add instruction for >> floating point values. If I use IRBuilder::CreateAtomicRMW() I get the >> error message: "atomicrmw operand must have integer type". > >

Hi all, The C++11 (& C11) compare_exchange functions with explicit memory order allow you to specify two sets of semantics, one for when the exchange actually happens and one for when it fails. Unfortunately, at the moment the LLVM IR "cmpxchg" instruction only has one ordering, which means we get sub-optimal codegen. This probably affects all architectures which use

> } while (__c11_atomic_compare_exchange_weak( > addr, &oldval, newval, memory_order_seq_cst, memory_order_relaxed)); Actually, I think this condition is inverted. Should be "while (!_c11...". Sorry about that. Tim.

On Fri, Mar 25, 2016 at 02:22:11PM -0400, Jan Vesely wrote: > Hi Tom, Matt, > > I'm working on a project that needs few coherent atomic operations (HSA > mode: load, store, compare-and-swap) for std::atomic_uint in HCC. > > the attached patch implements atomic compare and swap for SI+ > (untested). I tried to stay within what was available, but there are > few issues

# RFC: Atomic LL/SC loops in LLVM revisited ## Summary This proposal gives a brief overview of the challenges of lowering to LL/SC loops and details the approach I am taking for RISC-V. Beyond getting feedback on that work, my intention is to find consensus on moving other backends towards a similar approach and sharing common code where feasible. Scroll down to 'Questions' for a summary

Currently the AtomicExpandPass will lower the following IR: define i1 @foo(i32* %obj, i32 %old, i32 %new) { entry: %v0 = cmpxchg weak volatile i32* %obj, i32 %old, i32 %new _*release acquire*_ %v1 = extractvalue { i32, i1 } %v0, 1 ret i1 %v1 } to the equivalent of the following on AArch64: _*ldxr w8, [x0]*_ cmp w8, w1 b.ne .LBB0_3 // BB#1:

On 12/11/2015 01:29 PM, James Y Knight wrote: > > On Fri, Dec 11, 2015 at 3:05 PM, Philip Reames via llvm-dev > <llvm-dev at lists.llvm.org <mailto:llvm-dev at lists.llvm.org>> wrote: > >> One open question I don't know the answer to: Are there any >> special semantics required from floating point stores which >> aren't met

Hi, I have a simple swift code from which I generate IR code with ‘swiftc test.swift -emit-ir -o test.ll' When I try to run the .ll file or apply optimization with opt, I get errors like this one: lli: test.ll:548:110: error: expected instruction opcode %9 = cmpxchg i64* bitcast (%swift.type*** @field_type_vector_TipCalculator to i64*), i64 0, i64 %8 seq_cst seq_cst

On 10 May 2014, at 13:53, Tim Northover <t.p.northover at gmail.com> wrote: > It doesn't make sense for everything though, particularly if you want > target-specific IR to simply not exist. What would you map ARM's > "ldrex" to on x86? This isn't a great example. Having load-linked / store-conditional in the IR would make a number of transforms related to

Looking further during down time at the dev meeting today, it actually seems that PTX atom.* and red.* intrinsics map extremely naturally onto the LLVM atomicrmw and cmpxchg instructions. The biggest issue is that a subset of things expressible with these LLVM instructions do not trivially map to PTX, and the range of things naturally supported depends on the features of a given target. With

I'm working on getting the LLVM/projects/Test-Suite/UnitTest to compile for a target that I am developing. There is an example: AtomicOps, that uses calls to: __sync_fetch_and_add __sync_val_compare_and_swap __sync_lock_test_and_set These get converted into llvm IR operations like: atomicrmw cmpxchg Is there any way to keep these as function calls, as they are easier to map to the

Greetings everyone, I am working on Aarch64 LSE out-of-line atomics support in LLVM, porting this GCC series: https://gcc.gnu.org/legacy-ml/gcc-patches/2019-09/msg01034.html After local design experiments I've got some questions about upstream-suitable ways of implementation. More specifically: 1. Pass to expand atomics to library helper functions calls. These helpers test for the presence

Chris Lattner wrote: > On Jun 27, 2008, at 8:27 AM, Gary Benson wrote: > > def CMP_UNRESw : Pseudo<(outs), (ins GPRC:$rA, GPRC:$rB, i32imm: > > $label), > > "cmpw $rA, $rB\n\tbne- La${label}_exit", > > [(PPCcmp_unres GPRC:$rA, GPRC:$rB, imm: > > $label)]>; > > } > > > > ...and

This provides a mechanism to attach arbitrary configure options into the ssh_config file and use them from the LocalCommand and ProxyCommand. Examples: # set FOO to foo LocalEnvMod FOO = foo # append bar to FOO with default separator "," LocalEnvMod FOO += bar # unset FOO LocalEnvMod FOO = # append foo to BAR with separator ":", if BAR is empty

We've relaxed `atomicrmw xchg` to support floating point types but not cmpxchg -- the cmpxchg comparison behavior is not a floating point comparison, so that would be potentially misleading. I'd say adding the assertion is a good idea. Cheers, Nicolai On Thu, Aug 13, 2020 at 10:59 PM Chris Lattner via llvm-dev <llvm-dev at lists.llvm.org> wrote: > > Does the code generator

You need to insert new basic blocks and update CFG to accomplish this. There is a hackish way to do this right now. Add a pseudo instruction to represent this operation and mark it usesCustomDAGSchedInserter. This means the intrinsic is mapped to a single (pseudo) node. But it is then expanded into instructions that can span multiple basic blocks. See

Cool, that worked. New patch attached... Cheers, Gary Evan Cheng wrote: > Just cast both values to const TargetRegisterClass*. > > Evan > > On Jul 10, 2008, at 7:36 AM, Gary Benson wrote: > > Evan Cheng wrote: > > > How about? > > > > > > const TargetRegisterClass *RC = is64Bit ? &PPC:GPRCRegClass : > > > &PPC:G8RCRegClass; >

On 12/11/2015 12:05 AM, JF Bastien wrote: > On Fri, Dec 11, 2015 at 3:22 AM, Philip Reames via llvm-dev > <llvm-dev at lists.llvm.org <mailto:llvm-dev at lists.llvm.org>> wrote: > > Currently, we limit atomic loads and stores to either pointer or > integer types. I would like to propose that we extend this to > allow both floating point and vector types

> I am planning in doing in IR, but with target specific-passes (such as X86ExpandAtomicPass) > that just share some of the code This would more normally be done via target hooks in LLVM, though the principle is sound. > But it must be target-dependent as for example on Power a > seq_cst store has a fence before it, while on ARM it has a fence > both before and after it (per