What are the commands you used to generate each intermediate step?
In your .wast file, there should be a segments section which, among other
things, initializes the stack pointer. Since that is not there, it's being
initialized to 0, and then the translation of alloca yields a negative
address. I suspect this is what's causing the out of bounds.
As a workaround, what happens if you compile from C to LLVM with
optimization enabled? It seems like it'd likely be able to get rid of the
alloca and avoid any memory access at all, at least for this tin
On Wednesday, October 5, 2016 at 4:26:23 AM UTC-7, David Klassen wrote:
>
> Hi,
>
> I started playing with the current wasm implementation and there are some
> things that are not clear for me.
>
> If I compile this wast i get the expected result:
>
> (module
> (func $add (param $x f32) (param $y f32) (result f32) (f32.add
> (get_local $x) (get_local $y)))
> (export "add" (func $add))
> )
>
> module = Wasm.instantiateModule(new Uint8Array(buffer), {});
> module.exports.add(1,2) // => 3
>
> But when I try to compile from C, I get an exception "Error: memory access
> out of bounds"
>
> This is the C source:
>
> double add(double a, double b) { return a + b; }
>
> Fist I compile it to llvm ir:
>
> ; ModuleID = 'sample.c'
> source_filename = "sample.c"
> target datalayout = "e-m:e-p:32:32-i64:64-n32:64-S128"
> target triple = "wasm32"
>
> ; Function Attrs: nounwind
> define hidden double @add(double %a, double %b) #0 {
> entry:
> %a.addr = alloca double, align 8
> %b.addr = alloca double, align 8
> store double %a, double* %a.addr, align 8
> store double %b, double* %b.addr, align 8
> %0 = load double, double* %a.addr, align 8
> %1 = load double, double* %b.addr, align 8
> %add = fadd double %0, %1
> ret double %add
> }
>
> attributes #0 = { nounwind "correctly-rounded-divide-sqrt-fp-math"="false"
> "disable-tail-calls"="false" "less-precise-fpmad"="false"
> "no-frame-pointer-elim"="false" "no-infs-fp-math"="false"
> "no-jump-tables"="false" "no-nans-fp-math"="false"
> "no-signed-zeros-fp-math"="false" "no-trapping-math"="false"
> "stack-protector-buffer-size"="8" "target-cpu"="generic"
> "unsafe-fp-math"="false" "use-soft-float"="false" }
>
> !llvm.ident = !{!0}
>
> !0 = !{!"clang version 4.0.0 (http://llvm.org/git/clang.git
> b1b3141a30d56f4feb3cc68f57d1ebaf6fbad246) (http://llvm.org/git/llvm.git
> cf04e5f0f84bd44c375bf866d777708df39935df)"}
>
>
> Then to .s:
>
> .text
> .file "sample.ll"
> .hidden add
> .globl add
> .type add,@function
> add: # @add
> .param f64, f64
> .result f64
> .local i32
> # BB#0: # %entry
> i32.const $push2=, 0
> i32.load $push3=, __stack_pointer($pop2)
> i32.const $push4=, 16
> i32.sub $push6=, $pop3, $pop4
> tee_local $push5=, $2=, $pop6
> f64.store $drop=, 8($pop5), $0
> f64.store $drop=, 0($2), $1
> f64.load $push0=, 8($2)
> f64.add $push1=, $pop0, $1
> # fallthrough-return: $pop1
> .endfunc
> .Lfunc_end0:
> .size add, .Lfunc_end0-add
>
>
> .ident "clang version 4.0.0 (http://llvm.org/git/clang.git
> b1b3141a30d56f4feb3cc68f57d1ebaf6fbad246) (http://llvm.org/git/llvm.git
> cf04e5f0f84bd44c375bf866d777708df39935df)"
>
>
> Then i use binaryen s2wast:
>
> (module
> (memory $0 1)
> (export "memory" (memory $0))
> (export "add" (func $add))
> (table 0 anyfunc)
>
> (func $add (param $0 f64) (param $1 f64) (result f64)
> (local $2 i32)
> (f64.store offset=8
> (tee_local $2
> (i32.sub
> (i32.load offset=4
> (i32.const 0)
> )
> (i32.const 16)
> )
> )
> (get_local $0)
> )
> (f64.store
> (get_local $2)
> (get_local $1)
> )
> (f64.add
> (f64.load offset=8
> (get_local $2)
> )
> (get_local $1)
> )
> )
> )
>
> And finally use wabt to get wasm file.
>
> If I try to run the result in Chrome Canary i get "Error: memory access
> out of bounds"
>
> My goal is to be able to pass ArrayBuffers and other args from JS to wasm
> module and to be able to write the code in C. Can someone help me to solve
> this issue and maybe point to some useful resources. At this moment it is
> hard to find any working examples.
>
>
--
--
v8-users mailing list
[email protected]
http://groups.google.com/group/v8-users
---
You received this message because you are subscribed to the Google Groups
"v8-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/d/optout.
