RFC: How should logging verbosity work

[cratelyn]

Today our logging uses default values of INFO for viceroy-cli and viceroy-lib. For one v flag (-v) it's set at DEBUG, and for two v's (-vv) it's set to TRACE. This is fine as a beginning step for being able to have better instrumentation in Viceroy that's not just println and eprintln, but we should consider what other kinds of information would we want to expose.

Maybe hyper or tokio tracing events? Something else?

The implementation does allow users to set RUST_LOG themselves to bypass our default settings, but it would be nice to provide an experience where adding -v provides useful information to users without overwhelming them as well as making it easy to use. The existing implementation is subject to change after this RFC, and is meant to be a bare minimum.

This RFC exists so we can ask what other folks' opinions are on logging.

Originally reported by @mgattozzi.

[cratelyn]

Restating some previous discussion below, for posterity:

---

My instinct here is that end users shouldn't need to trace hyper/tokio events unless there's a bug in Viceroy, and that RUST_LOG is enough of a knob for that.

I think what's likely to be useful for most customers is the option to trace hostcall interactions in a lot more detail. That will need some design to do well, but especially given that wasm is a new platform for most folks, being able to log interactions with the host will probably help a lot.

One consideration there is that just about everyone will be using an SDK, so the actual hostcalls being made are at a lower level of abstraction than the code they are writing. That makes me wonder whether we could provide any kind of trace data from the SDKs (i.e. from within guest code). That could potentially be useful in production as well, though the considerations are very different there...

• @aturon

---

here are some assorted thoughts on the question of logging...

stadium Log Levels

I'll second the opinion that hyper/tokio events are probably a bit overkill, or a step in a different direction than customers would benefit from.

One notion I'd propose is that rather than increasing verbosity directly corresponding to more logs, for Viceroy this might mean displaying more information about things like runtime limits, expected billing numbers, or other related external resources like Edge Dictionaries.

That said, it could be pleasant to leave some sort of "show me everything" button, so that people reporting bugs in Viceroy could present us a tarball containing especially detailed logs. IMO, this is a related but distinct concern from the question of logging ergonomics for developers.

triangular_ruler SDK Diagnostics

That makes me wonder whether we could provide any kind of trace data from the SDKs (i.e. from within guest code). That could potentially be useful in production as well, though the considerations are very different there...

That's actually a really great point! I do remember that other folks had a similar idea of patching the fastly crate to emit more diagnostics, in order to support Fiddle iirc.

I suppose one vacuous point I'd ask is whether logging is prone to be affected by different host languages' respective idioms -- i.e. in Rust it's common practice to use tracing and emit structured spans/events, but I'm less confident OTTOMH whether other host languages would have an idiomatic language like that.

ballot_box Log Format

One option that could especially help people integrating Viceroy with other systems (hi @triblondon! hi @kailan!) would be a --message-format=json flag a la rustc, so that Viceroy could be run in a mode wherein all logs were presented in a machine readable format.

• @cratelyn

[GeeWee]

I have two comments.

• Currently running RUST_LOG=debug fastly compute serve on my machine just ends up outputting an endless stream of debug logs of some sort, and the server never starts

An example of this debug output.

Initializing...
Verifying package manifest...
Verifying local rust toolchain...
Initializing...
Building package using rust toolchain...
Creating package archive...

SUCCESS: Built package 'edge-api' (pkg/edge-api.tar.gz)

Initializing...
Running local server...

May 17 08:04:03.228 DEBUG RUST_LOG set to 'debug'
May 17 08:04:03.255 DEBUG timing: Starting Translate WASM function, (during <no pass>)    
May 17 08:04:03.255 DEBUG timing: Starting Translate WASM function, (during <no pass>)    
May 17 08:04:03.255 DEBUG timing: Starting Translate WASM function, (during <no pass>)    
May 17 08:04:03.255 DEBUG timing: Ending Translate WASM function    
May 17 08:04:03.255 DEBUG timing: Starting Compilation passes, (during <no pass>)    
May 17 08:04:03.255 DEBUG timing: Starting Translate WASM function, (during <no pass>)    
May 17 08:04:03.255 DEBUG timing: Starting Translate WASM function, (during <no pass>)    
May 17 08:04:03.255 DEBUG timing: Starting Translate WASM function, (during <no pass>)    
May 17 08:04:03.255 DEBUG timing: Ending Translate WASM function    
May 17 08:04:03.255 DEBUG Compiling (opt level Speed):
function u0:46(i64 vmctx, i64) fast {
    gv0 = vmctx
    gv1 = load.i64 notrap aligned readonly gv0
    gv2 = load.i64 notrap aligned gv1
    gv3 = vmctx
    sig0 = (i64 vmctx) wasmtime_system_v
    sig1 = (i64 vmctx, i64) fast
    fn0 = colocated u0:4739 sig1
    stack_limit = gv2

                                block0(v0: i64, v1: i64):
@2836                               v2 = global_value.i64 gv3
@2836                               v3 = load.i64 notrap aligned v2
                                    v14 -> v3
@2836                               v4 = load.i64 notrap aligned v3+8
@2836                               v5 = iadd_imm v4, 1
@2836                               v6 = iconst.i64 0
@2836                               v7 = ifcmp v5, v6
@2836                               brif sge v7, block2
@2836                               jump block3(v5)

                                block2:
@2836                               store.i64 notrap aligned v5, v3+8
@2836                               v8 = global_value.i64 gv3
@2836                               v9 = load.i64 notrap aligned readonly v8+0x0001_db38
@2836                               v10 = load.i64 notrap aligned readonly v9+152
@2836                               call_indirect sig0, v10(v8)
@2836                               v11 = load.i64 notrap aligned v3+8
@2836                               jump block3(v11)

                                block3(v12: i64):
@2837                               v13 = iadd_imm v12, 1
@2837                               store notrap aligned v13, v3+8
@2837                               call fn0(v0, v0)
@2837                               v15 = load.i64 notrap aligned v14+8
@283a                               jump block1

                                block1:
@283a                               store.i64 notrap aligned v15, v14+8
@283a                               return
}

May 17 08:04:03.255 DEBUG timing: Starting Compilation passes, (during <no pass>)    
May 17 08:04:03.255 DEBUG timing: Starting Control flow graph, (during Compilation passes)    
May 17 08:04:03.255 DEBUG timing: Ending Control flow graph    
May 17 08:04:03.255 DEBUG timing: Starting Pre-legalization rewriting, (during Compilation passes)    
May 17 08:04:03.255 DEBUG timing: Ending Pre-legalization rewriting    
May 17 08:04:03.255 DEBUG timing: Starting Dominator tree, (during Compilation passes)    
May 17 08:04:03.255 DEBUG timing: Ending Dominator tree    
May 17 08:04:03.255 DEBUG timing: Starting Loop analysis, (during Compilation passes)    
May 17 08:04:03.255 DEBUG timing: Ending Loop analysis    
May 17 08:04:03.255 DEBUG timing: Starting Loop invariant code motion, (during Compilation passes)    
May 17 08:04:03.255 DEBUG timing: Starting Control flow graph, (during Loop invariant code motion)    
May 17 08:04:03.255 DEBUG timing: Ending Control flow graph    
May 17 08:04:03.255 DEBUG timing: Starting Dominator tree, (during Loop invariant code motion)    
May 17 08:04:03.255 DEBUG timing: Ending Dominator tree    
May 17 08:04:03.255 DEBUG timing: Ending Loop invariant code motion    
May 17 08:04:03.255 DEBUG timing: Starting Global value numbering, (during Compilation passes)    
May 17 08:04:03.255 DEBUG timing: Ending Global value numbering    
May 17 08:04:03.255 DEBUG timing: Starting Dominator tree, (during Compilation passes)    
May 17 08:04:03.255 DEBUG timing: Ending Dominator tree    
May 17 08:04:03.255 DEBUG timing: Starting Remove unreachable blocks, (during Compilation passes)    
May 17 08:04:03.255 DEBUG timing: Ending Remove unreachable blocks    
May 17 08:04:03.255 DEBUG timing: Starting Dead code elimination, (during Compilation passes)    
May 17 08:04:03.255 DEBUG timing: Ending Dead code elimination    
May 17 08:04:03.255 DEBUG timing: Starting Remove constant phi-nodes, (during Compilation passes)    
May 17 08:04:03.255 DEBUG do_remove_constant_phis: done, 2 iters.   1 formals, of which 0 const.    
May 17 08:04:03.255 DEBUG timing: Ending Remove constant phi-nodes    
May 17 08:04:03.255 DEBUG Compiling (opt level Speed):
function u0:3742(i64 vmctx, i64, i32, i32) -> i32 wasmtime_system_v {
    gv0 = vmctx
    gv1 = load.i64 notrap aligned readonly gv0
    gv2 = load.i64 notrap aligned gv1
    gv3 = vmctx
    gv4 = load.i64 notrap aligned readonly gv3+1076
    heap0 = static gv4, min 0, bound 0x0001_0000_0000, offset_guard 0x8000_0000, index_type i32
    sig0 = (i64 vmctx) wasmtime_system_v
    sig1 = (i64 vmctx, i64, i32, i32, i32, i32) fast
    sig2 = (i64 vmctx, i64, i32, i32, i32) -> i32 fast
    sig3 = (i64 vmctx, i64, i32) -> i32 fast
    sig4 = (i64 vmctx, i64, i32, i32, i32) -> i64 fast
    sig5 = (i64 vmctx, i64, i32, i32, i32, i32, i32) -> i32 fast
    sig6 = (i64 vmctx, i64, i32) -> i32 fast
    sig7 = (i64 vmctx, i64, i32, i32, i32) -> i32 wasmtime_system_v
    fn0 = colocated u0:4906 sig1
    fn1 = colocated u0:4871 sig2
    fn2 = colocated u0:4872 sig3
    fn3 = colocated u0:4905 sig4
    fn4 = colocated u0:4847 sig5
    fn5 = colocated u0:4870 sig6
    fn6 = colocated u0:4898 sig7
    jt0 = jump_table [block11, block9, block8, block7, block6, block5]
    stack_limit = gv2

                                block0(v0: i64, v1: i64, v2: i32, v3: i32):
                                    v21 -> v2
                                    v36 -> v2
                                    v28 -> v3
@1ebf32                             v5 = iconst.i32 0
@1ebf32                             v6 = global_value.i64 gv3
@1ebf32                             v7 = load.i64 notrap aligned v6
                                    v34 -> v7
@1ebf32                             v8 = load.i64 notrap aligned v7+8
@1ebf32                             v9 = iadd_imm v8, 1
@1ebf32                             v10 = iconst.i64 0
@1ebf32                             v11 = ifcmp v9, v10
@1ebf32                             brif sge v11, block2
@1ebf32                             jump block3(v9)

                                block2:
@1ebf32                             store.i64 notrap aligned v9, v7+8
@1ebf32                             v12 = global_value.i64 gv3
@1ebf32                             v13 = load.i64 notrap aligned readonly v12+0x0001_db38
@1ebf32                             v14 = load.i64 notrap aligned readonly v13+152
@1ebf32                             call_indirect sig0, v14(v12)
@1ebf32                             v15 = load.i64 notrap aligned v7+8
@1ebf32                             jump block3(v15)

                                block3(v24: i64):
@1ebf34                             v16 = global_value.i64 gv3
@1ebf34                             v17 = load.i32 notrap aligned v16+1104
@1ebf3a                             v18 = iconst.i32 16
@1ebf3c                             v19 = isub v17, v18
                                    v26 -> v19
@1ebf3f                             v20 = global_value.i64 gv3
@1ebf3f                             store notrap aligned v19, v20+1104
@1ebf55                             v22 = heap_addr.i64 heap0, v2, 1
@1ebf55                             v23 = uload8.i32 little v22
@1ebf58                             v25 = iadd_imm v24, 8
                                    v31 -> v25
@1ebf58                             br_table v23, block11, jt0

                                block11:
                                    jump block10(v28, v7)

                                block12:
                                    jump block10(v28, v34)

                                block10(v27: i32, v33: i64):
@1ebf66                             v29 = iconst.i32 0x0050_e268
@1ebf6c                             v30 = iconst.i32 4
@1ebf6e                             v32 = iadd_imm.i64 v25, 5
@1ebf6e                             store notrap aligned v32, v33+8
@1ebf6e                             call fn0(v0, v0, v19, v27, v29, v30)
@1ebf6e                             v35 = load.i64 notrap aligned v33+8
@1ebf78                             v37 = iconst.i32 4
@1ebf7a                             v38 = iadd.i32 v2, v37
@1ebf7b                             v39 = heap_addr.i64 heap0, v26, 1
@1ebf7b                             store little v38, v39+12
@1ebf82                             v40 = iconst.i32 12
@1ebf84                             v41 = iadd.i32 v26, v40
@1ebf85                             v42 = iconst.i32 0x0050_e26c
@1ebf8b                             v43 = iadd_imm v35, 11
@1ebf8b                             store notrap aligned v43, v33+8
@1ebf8b                             v44 = call fn1(v0, v0, v26, v41, v42)
@1ebf8b                             v45 = load.i64 notrap aligned v33+8
@1ebf94                             v46 = iadd_imm v45, 2
@1ebf94                             store notrap aligned v46, v33+8
@1ebf94                             v47 = call fn2(v0, v0, v26)
@1ebf94                             v48 = load.i64 notrap aligned v33+8
@1ebf9c                             v49 = iadd_imm v48, 2
@1ebf9c                             jump block4(v26, v47, v49, v33)

                                block9:
@1ebfa3                             v50 = iconst.i32 0x0050_e24c
@1ebfa9                             v51 = iconst.i32 12
@1ebfab                             v52 = iadd_imm.i64 v25, 5
@1ebfab                             store notrap aligned v52, v34+8
@1ebfab                             call fn0(v0, v0, v19, v28, v50, v51)
@1ebfab                             v53 = load.i64 notrap aligned v34+8
@1ebfb5                             v54 = iconst.i32 4
@1ebfb7                             v55 = iadd.i32 v21, v54
@1ebfb8                             v56 = heap_addr.i64 heap0, v19, 1
@1ebfb8                             store little v55, v56+12
@1ebfbf                             v57 = iconst.i32 12
@1ebfc1                             v58 = iadd.i32 v19, v57
@1ebfc2                             v59 = iconst.i32 0x0050_e258
@1ebfc8                             v60 = iadd_imm v53, 11
@1ebfc8                             store notrap aligned v60, v34+8
@1ebfc8                             v61 = call fn1(v0, v0, v19, v58, v59)
@1ebfc8                             v62 = load.i64 notrap aligned v34+8
@1ebfd1                             v63 = iadd_imm v62, 2
@1ebfd1                             store notrap aligned v63, v34+8
@1ebfd1                             v64 = call fn2(v0, v0, v19)
@1ebfd1                             v65 = load.i64 notrap aligned v34+8
@1ebfd9                             v66 = iadd_imm v65, 2
@1ebfd9                             jump block4(v19, v64, v66, v34)

                                block8:
@1ebfe0                             v67 = iconst.i32 0x0050_e22f
@1ebfe6                             v68 = iconst.i32 10
@1ebfe8                             v69 = iadd_imm.i64 v25, 5
@1ebfe8                             store notrap aligned v69, v34+8
@1ebfe8                             call fn0(v0, v0, v19, v28, v67, v68)
@1ebfe8                             v70 = load.i64 notrap aligned v34+8
@1ebff2                             v71 = iconst.i32 4
@1ebff4                             v72 = iadd.i32 v21, v71
@1ebff5                             v73 = heap_addr.i64 heap0, v19, 1
@1ebff5                             store little v72, v73+12
@1ebffc                             v74 = iconst.i32 12
@1ebffe                             v75 = iadd.i32 v19, v74
@1ebfff                             v76 = iconst.i32 0x0050_e23c
@1ec005                             v77 = iadd_imm v70, 11
@1ec005                             store notrap aligned v77, v34+8
@1ec005                             v78 = call fn1(v0, v0, v19, v75, v76)
@1ec005                             v79 = load.i64 notrap aligned v34+8
@1ec00e                             v80 = iadd_imm v79, 2
@1ec00e                             store notrap aligned v80, v34+8
@1ec00e                             v81 = call fn2(v0, v0, v19)
@1ec00e                             v82 = load.i64 notrap aligned v34+8
@1ec016                             v83 = iadd_imm v82, 2
@1ec016                             jump block4(v19, v81, v83, v34)

                                block7:
@1ec01d                             v84 = iconst.i32 0x0050_e221
@1ec023                             v85 = iconst.i32 5
@1ec025                             v86 = iadd_imm.i64 v25, 5
@1ec025                             store notrap aligned v86, v34+8
@1ec025                             v87 = call fn3(v0, v0, v28, v84, v85)
@1ec025                             v88 = load.i64 notrap aligned v34+8
@1ec02b                             v89 = heap_addr.i64 heap0, v19, 1
@1ec02b                             store little v87, v89
@1ec032                             v90 = iconst.i32 1
@1ec034                             v91 = iadd.i32 v21, v90
@1ec035                             v92 = heap_addr.i64 heap0, v19, 1
@1ec035                             store little v91, v92+12
@1ec03a                             v93 = iconst.i32 0x0050_e226
@1ec040                             v94 = iconst.i32 9
@1ec044                             v95 = iconst.i32 12
@1ec046                             v96 = iadd.i32 v19, v95
@1ec047                             v97 = iconst.i32 0x0050_e200
@1ec04d                             v98 = iadd_imm v88, 14
@1ec04d                             store notrap aligned v98, v34+8
@1ec04d                             v99 = call fn4(v0, v0, v19, v93, v94, v96, v97)
@1ec04d                             v100 = load.i64 notrap aligned v34+8
@1ec056                             v101 = iadd_imm v100, 2
@1ec056                             store notrap aligned v101, v34+8
@1ec056                             v102 = call fn5(v0, v0, v19)
@1ec056                             v103 = load.i64 notrap aligned v34+8
@1ec05e                             v104 = iadd_imm v103, 2
@1ec05e                             jump block4(v19, v102, v104, v34)

                                block6:
@1ec063                             v105 = iconst.i32 0x0050_e21b
@1ec069                             v106 = iconst.i32 6
@1ec06b                             v107 = iadd_imm.i64 v25, 4
@1ec06b                             store notrap aligned v107, v34+8
@1ec06b                             v108 = call fn6(v0, v0, v28, v105, v106)
@1ec06b                             v109 = load.i64 notrap aligned v34+8
@1ec073                             v110 = iadd_imm v109, 2
@1ec073                             jump block4(v19, v108, v110, v34)

                                block5:
@1ec078                             v111 = iconst.i32 0x0050_e210
@1ec07e                             v112 = iconst.i32 11
@1ec080                             v113 = iadd_imm.i64 v25, 4
@1ec080                             store notrap aligned v113, v34+8
@1ec080                             v114 = call fn6(v0, v0, v28, v111, v112)
@1ec080                             v115 = load.i64 notrap aligned v34+8
@1ec088                             v116 = iadd_imm v115, 1
@1ec088                             jump block4(v19, v114, v116, v34)

                                block4(v117: i32, v121: i32, v122: i64, v124: i64):
@1ec08b                             v118 = iconst.i32 16
@1ec08d                             v119 = iadd v117, v118
@1ec08e                             v120 = global_value.i64 gv3
@1ec08e                             store notrap aligned v119, v120+1104
@1ec096                             v123 = iadd_imm v122, 5
@1ec096                             jump block1(v121)

                                block1(v4: i32):
@1ec096                             store.i64 notrap aligned v123, v124+8
@1ec096                             return v4
}

May 17 08:04:03.255 DEBUG timing: Starting Control flow graph, (during Compilation passes)    
May 17 08:04:03.255 DEBUG timing: Starting VCode lowering, (during Compilation passes)    
May 17 08:04:03.255 DEBUG timing: Ending Control flow graph    
May 17 08:04:03.255 DEBUG timing: Starting Pre-legalization rewriting, (during Compilation passes)    
May 17 08:04:03.255 DEBUG timing: Ending Pre-legalization rewriting    
May 17 08:04:03.255 DEBUG timing: Starting Dominator tree, (during Compilation passes)    
May 17 08:04:03.255 DEBUG timing: Ending Dominator tree    
May 17 08:04:03.255 DEBUG timing: Starting Loop analysis, (during Compilation passes)    
May 17 08:04:03.255 DEBUG timing: Ending Loop analysis    
May 17 08:04:03.255 DEBUG timing: Starting Loop invariant code motion, (during Compilation passes)    
May 17 08:04:03.255 DEBUG timing: Starting Control flow graph, (during Loop invariant code motion)    
May 17 08:04:03.255 DEBUG timing: Ending Control flow graph    
May 17 08:04:03.255 DEBUG timing: Starting Dominator tree, (during Loop invariant code motion)    
May 17 08:04:03.255 DEBUG timing: Ending Dominator tree    
May 17 08:04:03.255 DEBUG timing: Ending VCode lowering    
May 17 08:04:03.255 DEBUG timing: Ending Loop invariant code motion    
May 17 08:04:03.255 DEBUG timing: Starting Global value numbering, (during Compilation passes)    
May 17 08:04:03.255 DEBUG timing: Ending Translate WASM function    
May 17 08:04:03.255 DEBUG timing: Ending Translate WASM function    
May 17 08:04:03.255 DEBUG timing: Starting Compilation passes, (during <no pass>)    
May 17 08:04:03.255 DEBUG timing: Ending Global value numbering    
May 17 08:04:03.255 DEBUG timing: Starting Compilation passes, (during <no pass>)    
May 17 08:04:03.255 DEBUG timing: Starting Dominator tree, (during Compilation passes)    
May 17 08:04:03.255 DEBUG timing: Starting Register allocation, (during Compilation passes)    
May 17 08:04:03.255 DEBUG timing: Ending Dominator tree    
May 17 08:04:03.255 DEBUG     
May 17 08:04:03.255 DEBUG ================ regalloc.rs: BEGIN function ================    
May 17 08:04:03.255 DEBUG timing: Starting Remove unreachable blocks, (during Compilation passes)    
May 17 08:04:03.255 DEBUG with options: checker: false, algorithm: backtracking (block annotations: false)    
May 17 08:04:03.255 DEBUG timing: Ending Remove unreachable blocks    
May 17 08:04:03.255 DEBUG timing: Starting Dead code elimination, (during Compilation passes)    
May 17 08:04:03.255 DEBUG using RealRegUniverse:    
May 17 08:04:03.255 DEBUG   class I32(J) at [16 .. 29]: %rsi %rdi %rax %rcx %rdx %r8 %r9 %r10 %r11 %r12* %r13 %r14 %rbx %r15     
May 17 08:04:03.255 DEBUG   class V128(V) at [0 .. 15]: %xmm0 %xmm1 %xmm2 %xmm3 %xmm4 %xmm5 %xmm6 %xmm7 %xmm8 %xmm9 %xmm10 %xmm11 %xmm12 %xmm13 %xmm14 %xmm15*     
May 17 08:04:03.255 DEBUG timing: Ending Dead code elimination    
May 17 08:04:03.255 DEBUG timing: Starting Remove constant phi-nodes, (during Compilation passes)    
May 17 08:04:03.255 DEBUG   not allocable at [30 .. 31]: %rsp %rbp     
May 17 08:04:03.255 DEBUG do_remove_constant_phis: done, 2 iters.   8 formals, of which 4 const.    
May 17 08:04:03.255 DEBUG timing: Ending Remove constant phi-nodes    
May 17 08:04:03.255 DEBUG alloc_main: begin    
May 17 08:04:03.255 DEBUG alloc_main:   in: 30 insns in 6 blocks    
May 17 08:04:03.255 DEBUG alloc_main:   in: 12 VLRs, 55 RLRs    
May 17 08:04:03.255 DEBUG timing: Ending Translate WASM function    
May 17 08:04:03.255 DEBUG timing: Ending Translate WASM function    
May 17 08:04:03.255 DEBUG     
May 17 08:04:03.255 DEBUG timing: Starting Compilation passes, (during <no pass>)    
May 17 08:04:03.255 DEBUG <<<<====---- RA state at 'Initial' ----====    
May 17 08:04:03.255 DEBUG TODO  vr2 = (VR: v2J, sz=27, tc=7, sc=0.259, [(RF: i2.d-i28.u)])    
May 17 08:04:03.255 DEBUG TODO  vr3 = (VR: v5J, sz=7, tc=4, sc=0.571, [(RF: i4.d-i9.u), (RF: i18.u-i18.u)])    
May 17 08:04:03.255 DEBUG TODO  vr0 = (VR: v0J, sz=25, tc=6, sc=0.240, [(RF: i0.d-i24.u)])    
May 17 08:04:03.255 DEBUG Compiling (opt level Speed):
function u0:1278(i64 vmctx, i64, i32, i32, i32) wasmtime_system_v {
    gv0 = vmctx
    gv1 = load.i64 notrap aligned readonly gv0
    gv2 = load.i64 notrap aligned gv1
    gv3 = vmctx
    sig0 = (i64 vmctx) wasmtime_system_v
    sig1 = (i64 vmctx, i64, i32, i32, i32) fast
    sig2 = (i64 vmctx, i64) -> i64 fast
    fn0 = colocated u0:1605 sig1
    fn1 = colocated u0:1576 sig2
    stack_limit = gv2

                                block0(v0: i64, v1: i64, v2: i32, v3: i32, v4: i32):
                                    v15 -> v2
                                    v16 -> v3
                                    v17 -> v4
@d1e51                              v5 = global_value.i64 gv3
@d1e51                              v6 = load.i64 notrap aligned v5
                                    v20 -> v6
@d1e51                              v7 = load.i64 notrap aligned v6+8
@d1e51                              v8 = iadd_imm v7, 1
@d1e51                              v9 = iconst.i64 0
@d1e51                              v10 = ifcmp v8, v9
@d1e51                              brif sge v10, block2
@d1e51                              jump block3(v8)

                                block2:
@d1e51                              store.i64 notrap aligned v8, v6+8
@d1e51                              v11 = global_value.i64 gv3
@d1e51                              v12 = load.i64 notrap aligned readonly v11+0x0001_db38
@d1e51                              v13 = load.i64 notrap aligned readonly v12+152
@d1e51                              call_indirect sig0, v13(v11)
@d1e51                              v14 = load.i64 notrap aligned v6+8
@d1e51                              jump block3(v14)

                                block3(v18: i64):
@d1e58                              v19 = iadd_imm v18, 4
@d1e58                              store notrap aligned v19, v6+8
@d1e58                              call fn0(v0, v0, v2, v3, v4)
@d1e58                              v21 = load.i64 notrap aligned v20+8
@d1e5e                              v22 = iadd_imm v21, 1
@d1e5e                              store notrap aligned v22, v20+8
@d1e5e                              v23 = call fn1(v0, v0)
@d1e5e                              v24 = load.i64 notrap aligned v20+8
@d1e65                              v25 = iadd_imm v24, 1
@d1e65                              store notrap aligned v25, v20+8
@d1e65                              v26 = call fn1(v0, v0)
@d1e65                              v27 = load.i64 notrap aligned v20+8
@d1e6c                              jump block1

                                block1:
@d1e6c                              store.i64 notrap aligned v27, v20+8
@d1e6c                              return
}

May 17 08:04:03.255 DEBUG timing: Starting Control flow graph, (during Compilation passes)    
May 17 08:04:03.255 DEBUG TODO  vr7 = (VR: v11J, sz=5, tc=0, sc=0.000, [(RF: i16.d-i17.d), (RF: i18.d-i20.u)])    
May 17 08:04:03.255 DEBUG timing: Ending Control flow graph    
May 17 08:04:03.255 DEBUG TODO  vr9 = (VR: v13J, sz=3, tc=2, sc=0.667, [(RF: i26.d-i28.u)])    
May 17 08:04:03.255 DEBUG timing: Starting Pre-legalization rewriting, (during Compilation passes)    
May 17 08:04:03.255 DEBUG TODO  vr5 = (VR: v9J, sz=3, tc=2, sc=0.667, [(RF: i11.d-i13.u)])    
May 17 08:04:03.255 DEBUG TODO  vr6 = (VR: v10J, sz=3, tc=1, sc=0.333, [(RF: i14.d-i16.u)])    
May 17 08:04:03.255 DEBUG TODO  vr1 = (VR: v1J, sz=1, tc=1, sc=1.000, [(RF: i1.d-i1.d)])    
May 17 08:04:03.255 DEBUG TODO  vr8 = (VR: v12J, sz=3, tc=3, sc=1.000, [(RF: i20.d-i22.u)])    
May 17 08:04:03.255 DEBUG timing: Ending Pre-legalization rewriting    
May 17 08:04:03.255 DEBUG timing: Starting VCode lowering, (during Compilation passes)    
May 17 08:04:03.255 DEBUG TODO  vr4 = (VR: v8J, sz=2, tc=2, sc=1.000, [(RF: i10.d-i11.u)])    
May 17 08:04:03.255 DEBUG TODO  vr10 = (VR: v17J, sz=2, tc=2, sc=1.000, [(RF: i6.d-i7.u)])    
May 17 08:04:03.255 DEBUG TODO  vr11 = (VR: v18J, sz=2, tc=1, sc=0.500, [(RF: i3.d-i4.u)])    
May 17 08:04:03.255 DEBUG >>>>    
May 17 08:04:03.255 DEBUG     
May 17 08:04:03.255 DEBUG -- MAIN ALLOCATION LOOP (DI means 'direct', CO means 'coalesced'):    
May 17 08:04:03.255 DEBUG alloc_main:   main allocation loop: begin    
May 17 08:04:03.255 DEBUG -- still TODO          12    
May 17 08:04:03.255 DEBUG --   considering       vr2:  (VR: v2J, sz=27, tc=7, sc=0.259, [(RF: i2.d-i28.u)])    
May 17 08:04:03.255 DEBUG --   DI alloc to       %r12    
May 17 08:04:03.255 DEBUG -- still TODO          11    
May 17 08:04:03.255 DEBUG --   considering       vr0:  (VR: v0J, sz=25, tc=6, sc=0.240, [(RF: i0.d-i24.u)])    
May 17 08:04:03.255 DEBUG --   CO candidates     pri { %rdi %rsi }    
May 17 08:04:03.255 DEBUG --   DI alloc to       %r13    
May 17 08:04:03.255 DEBUG -- still TODO          10    
May 17 08:04:03.255 DEBUG --   considering       vr3:  (VR: v5J, sz=7, tc=4, sc=0.571, [(RF: i4.d-i9.u), (RF: i18.u-i18.u)])    
May 17 08:04:03.255 DEBUG --   DI alloc to       %rsi    
May 17 08:04:03.255 DEBUG -- still TODO          9    
May 17 08:04:03.255 DEBUG --   considering       vr7:  (VR: v11J, sz=5, tc=0, sc=0.000, [(RF: i16.d-i17.d), (RF: i18.d-i20.u)])    
May 17 08:04:03.255 DEBUG --   CO candidates     pri { %rsi }    
May 17 08:04:03.255 DEBUG --   CO alloc to       %rsi    
May 17 08:04:03.255 DEBUG -- still TODO          8    
May 17 08:04:03.255 DEBUG --   considering       vr5:  (VR: v9J, sz=3, tc=2, sc=0.667, [(RF: i11.d-i13.u)])    
May 17 08:04:03.255 DEBUG --   DI alloc to       %rsi    
May 17 08:04:03.255 DEBUG -- still TODO          7    
May 17 08:04:03.255 DEBUG --   considering       vr6:  (VR: v10J, sz=3, tc=1, sc=0.333, [(RF: i14.d-i16.u)])    
May 17 08:04:03.255 DEBUG --   CO candidates     pri { %rsi }    
May 17 08:04:03.255 DEBUG --   CO alloc to       %rsi    
May 17 08:04:03.255 DEBUG -- still TODO          6    
May 17 08:04:03.255 DEBUG --   considering       vr8:  (VR: v12J, sz=3, tc=3, sc=1.000, [(RF: i20.d-i22.u)])    
May 17 08:04:03.255 DEBUG --   CO candidates     pri { %rsi }    
May 17 08:04:03.255 DEBUG --   CO alloc to       %rsi    
May 17 08:04:03.255 DEBUG -- still TODO          5    
May 17 08:04:03.255 DEBUG --   considering       vr9:  (VR: v13J, sz=3, tc=2, sc=0.667, [(RF: i26.d-i28.u)])    
May 17 08:04:03.255 DEBUG --   DI alloc to       %rsi    
May 17 08:04:03.255 DEBUG -- still TODO          4    
May 17 08:04:03.255 DEBUG --   considering       vr4:  (VR: v8J, sz=2, tc=2, sc=1.000, [(RF: i10.d-i11.u)])    
May 17 08:04:03.255 DEBUG --   DI alloc to       %rsi    
May 17 08:04:03.255 DEBUG -- still TODO          3    
May 17 08:04:03.255 DEBUG --   considering       vr10:  (VR: v17J, sz=2, tc=2, sc=1.000, [(RF: i6.d-i7.u)])    
May 17 08:04:03.255 DEBUG --   DI alloc to       %rdi    
May 17 08:04:03.255 DEBUG -- still TODO          2    
May 17 08:04:03.255 DEBUG --   considering       vr11:  (VR: v18J, sz=2, tc=1, sc=0.500, [(RF: i3.d-i4.u)])    
May 17 08:04:03.255 DEBUG --   CO candidates     pri { %rsi }    
May 17 08:04:03.255 DEBUG --   CO alloc to       %rsi    
May 17 08:04:03.255 DEBUG -- still TODO          1    
May 17 08:04:03.255 DEBUG --   considering       vr1:  (VR: v1J, sz=1, tc=1, sc=1.000, [(RF: i1.d-i1.d)])    
May 17 08:04:03.255 DEBUG --   CO candidates     pri { %rsi }    
May 17 08:04:03.255 DEBUG --   CO alloc to       %rsi    
May 17 08:04:03.255 DEBUG -- still TODO          0    
May 17 08:04:03.255 DEBUG alloc_main:   main allocation loop: end    
May 17 08:04:03.255 DEBUG     
May 17 08:04:03.255 DEBUG <<<<====---- RA state at 'Final' ----====    
May 17 08:04:03.255 DEBUG >>>>    
May 17 08:04:03.255 DEBUG     
May 17 08:04:03.255 DEBUG alloc_main:   create spills_n_reloads for MOVE insns    
May 17 08:04:03.255 DEBUG     
May 17 08:04:03.255 DEBUG alloc_main:   create spills_n_reloads for other insns    
May 17 08:04:03.255 DEBUG alloc_main:   create frag_map    
May 17 08:04:03.255 DEBUG timing: Ending VCode lowering    
May 17 08:04:03.255 DEBUG alloc_main:   edit_inst_stream    
May 17 08:04:03.255 DEBUG timing: Starting Register allocation, (during Compilation passes)    
May 17 08:04:03.255 DEBUG     
May 17 08:04:03.255 DEBUG ================ regalloc.rs: BEGIN function ================    
May 17 08:04:03.255 DEBUG Compiling (opt level Speed):
function u0:662(i64 vmctx, i64, i32, i32, i32) wasmtime_system_v {
    gv0 = vmctx
    gv1 = load.i64 notrap aligned readonly gv0
    gv2 = load.i64 notrap aligned gv1
    gv3 = vmctx
    gv4 = load.i64 notrap aligned readonly gv3+1076
    heap0 = static gv4, min 0, bound 0x0001_0000_0000, offset_guard 0x8000_0000, index_type i32
    sig0 = (i64 vmctx) wasmtime_system_v
    sig1 = (i64 vmctx, i64, i32) -> i32 fast
    sig2 = (i64 vmctx, i64, i32, i32, i32) -> i32 fast
    sig3 = (i64 vmctx, i64, i32, i32) fast
    sig4 = (i64 vmctx, i64, i32, i32, i32) fast
    sig5 = (i64 vmctx, i64, i32, i32) fast
    sig6 = (i64 vmctx, i64, i32, i32) fast
    sig7 = (i64 vmctx, i64, i32) fast
    sig8 = (i64 vmctx, i64, i32, i32) fast
    sig9 = (i64 vmctx, i64, i32, i32, i32) fast
    sig10 = (i64 vmctx, i64, i32, i32) fast
    sig11 = (i64 vmctx, i64, i32) fast
    sig12 = (i64 vmctx, i64, i32, i32, i32) fast
    sig13 = (i64 vmctx, i64, i32, i32, i32) fast
    sig14 = (i64 vmctx, i64, i32) fast
    sig15 = (i64 vmctx, i64, i32, i32) fast
    sig16 = (i64 vmctx, i64, i32) -> i32 fast
    sig17 = (i64 vmctx, i64, i32) fast
    sig18 = (i64 vmctx, i64, i32) fast
    sig19 = (i64 vmctx, i64, i32, i32, i32) fast
    fn0 = colocated u0:4029 sig1
    fn1 = colocated u0:4752 sig2
    fn2 = colocated u0:4325 sig3
    fn3 = colocated u0:4171 sig4
    fn4 = colocated u0:1070 sig5
    fn5 = colocated u0:564 sig6
    fn6 = colocated u0:3211 sig7
    fn7 = colocated u0:3212 sig8
    fn8 = colocated u0:3210 sig9
    fn9 = colocated u0:3065 sig10
    fn10 = colocated u0:3950 sig11
    fn11 = colocated u0:1047 sig12
    fn12 = colocated u0:112 sig13
    fn13 = colocated u0:655 sig14
    fn14 = colocated u0:3208 sig15
    fn15 = colocated u0:3937 sig16
    fn16 = colocated u0:657 sig17
    fn17 = colocated u0:656 sig18
    fn18 = colocated u0:3882 sig19
    stack_limit = gv2

                                block0(v0: i64, v1: i64, v2: i32, v3: i32, v4: i32):
                                    v343 -> v2
                                    v22 -> v3
                                    v474 -> v4
@76093                              v5 = iconst.i32 0
@76095                              v6 = iconst.i64 0
@76095                              v7 = global_value.i64 gv3
@76095                              v8 = load.i64 notrap aligned v7
                                    v25 -> v8
@76095                              v9 = load.i64 notrap aligned v8+8
@76095                              v10 = iadd_imm v9, 1
@76095                              v11 = iconst.i64 0
@76095                              v12 = ifcmp v10, v11
@76095                              brif sge v12, block2
@76095                              jump block3(v10)

                                block2:
@76095                              store.i64 notrap aligned v10, v8+8
@76095                              v13 = global_value.i64 gv3
@76095                              v14 = load.i64 notrap aligned readonly v13+0x0001_db38
@76095                              v15 = load.i64 notrap aligned readonly v14+152
@76095                              call_indirect sig0, v15(v13)
@76095                              v16 = load.i64 notrap aligned v8+8
@76095                              jump block3(v16)

                                block3(v23: i64):
@76097                              v17 = global_value.i64 gv3
@76097                              v18 = load.i32 notrap aligned v17+1104
@7609d                              v19 = iconst.i32 1328
@760a0                              v20 = isub v18, v19
@760a3                              v21 = global_value.i64 gv3
@760a3                              store notrap aligned v20, v21+1104
@760b5                              v24 = iadd_imm v23, 7
@760b5                              store notrap aligned v24, v8+8
@760b5                              v26 = call fn0(v0, v0, v3)
@760b5                              v27 = load.i64 notrap aligned v25+8
@760bb                              v28 = heap_addr.i64 heap0, v26, 1
@760bb                              v29 = uload8.i32 little v28
@760be                              v30 = iconst.i32 1
@760c0                              v31 = icmp ne v29, v30
@760c0                              v32 = bint.i32 v31
@760c1                              v33 = iadd_imm v27, 4
@760c1                              brnz v32, block8
@760c1                              jump block9

                                block9:
@760c5                              v34 = iconst.i32 24
@760c7                              v35 = iadd.i32 v20, v34
@760ca                              v36 = iconst.i32 288
@760cd                              v37 = iadd_imm.i64 v33, 6
@760cd                              store notrap aligned v37, v25+8
@760cd                              v38 = call fn1(v0, v0, v35, v22, v36)
@760cd                              v39 = load.i64 notrap aligned v25+8
@760d6                              v40 = iconst.i32 16
@760d8                              v41 = iadd.i32 v20, v40
@760db                              v42 = iconst.i32 108
@760de                              v43 = iadd.i32 v20, v42
@760df                              v44 = iadd_imm v39, 7
@760df                              store notrap aligned v44, v25+8
@760df                              call fn2(v0, v0, v41, v43)
@760df                              v45 = load.i64 notrap aligned v25+8
@760e7                              v46 = iconst.i32 496
@760ea                              v47 = iadd.i32 v20, v46
@760ed                              v48 = heap_addr.i64 heap0, v20, 1
@760ed                              v49 = load.i32 little v48+16
@760f2                              v50 = iconst.i32 0x0010_7690
@760fa                              v51 = select v49, v49, v50
@760fd                              v52 = heap_addr.i64 heap0, v20, 1
@760fd                              v53 = load.i32 little v52+20
@76100                              v54 = iconst.i32 0
@76104                              v55 = select v49, v53, v54
@76105                              v56 = iadd_imm v45, 15
@76105                              store notrap aligned v56, v25+8
@76105                              call fn3(v0, v0, v47, v51, v55)
@76105                              v57 = load.i64 notrap aligned v25+8
@7610d                              v58 = iconst.i32 608
@76110                              v59 = iadd.i32 v20, v58
@76111                              v60 = iconst.i32 24
@76113                              v61 = iadd v59, v60
@76116                              v62 = iconst.i32 496
@76119                              v63 = iadd.i32 v20, v62
@7611a                              v64 = iconst.i32 24
@7611c                              v65 = iadd v63, v64
@7611d                              v66 = heap_addr.i64 heap0, v65, 1
@7611d                              v67 = load.i32 little v66
@76120                              v68 = heap_addr.i64 heap0, v61, 1
@76120                              store little v67, v68
@76125                              v69 = iconst.i32 608
@76128                              v70 = iadd.i32 v20, v69
@76129                              v71 = iconst.i32 16
@7612b                              v72 = iadd v70, v71
@7612e                              v73 = iconst.i32 496
@76131                              v74 = iadd.i32 v20, v73
@76132                              v75 = iconst.i32 16
@76134                              v76 = iadd v74, v75
@76135                              v77 = heap_addr.i64 heap0, v76, 1
@76135                              v78 = load.i64 little v77
@76138                              v79 = heap_addr.i64 heap0, v72, 1
@76138                              store little v78, v79
@7613d                              v80 = iconst.i32 608
@76140                              v81 = iadd.i32 v20, v80
@76141                              v82 = iconst.i32 8
@76143                              v83 = iadd v81, v82
@76146                              v84 = iconst.i32 496
@76149                              v85 = iadd.i32 v20, v84
@7614a                              v86 = iconst.i32 8
@7614c                              v87 = iadd v85, v86
@7614d                              v88 = heap_addr.i64 heap0, v87, 1
@7614d                              v89 = load.i64 little v88
@76150                              v90 = heap_addr.i64 heap0, v83, 1
@76150                              store little v89, v90
@76157                              v91 = heap_addr.i64 heap0, v20, 1
@76157                              v92 = load.i64 little v91+496
@7615b                              v93 = heap_addr.i64 heap0, v20, 1
@7615b                              store little v92, v93+608
@76161                              v94 = iconst.i32 904
@76164                              v95 = iadd.i32 v20, v94
@76167                              v96 = iconst.i32 608
@7616a                              v97 = iadd.i32 v20, v96
@7616b                              v98 = iadd_imm v57, 47
@7616b                              store notrap aligned v98, v25+8
@7616b                              call fn4(v0, v0, v95, v97)
@7616b                              v99 = load.i64 notrap aligned v25+8
@76173                              v100 = iconst.i32 904
@76176                              v101 = iadd.i32 v20, v100
@76177                              v102 = iconst.i32 8
@76179                              v103 = iadd v101, v102
@7617c                              v104 = heap_addr.i64 heap0, v103, 1
@7617c                              v105 = load.i32 little v104
@76183                              v106 = heap_addr.i64 heap0, v20, 1
@76183                              v107 = load.i32 little v106+908
@7618b                              v108 = heap_addr.i64 heap0, v20, 1
@7618b                              v109 = load.i32 little v108+904
@7618f                              v110 = iadd_imm v99, 14
@7618f                              brnz v109, block7
@7618f                              jump block10

                                block10:
@76193                              v111 = iconst.i32 904
@76196                              v112 = iadd.i32 v20, v111
@76197                              v113 = iconst.i32 12
@76199                              v114 = iadd v112, v113
@7619a                              v115 = heap_addr.i64 heap0, v114, 1
@7619a                              v116 = load.i32 little v115
@761a1                              v117 = iconst.i32 312
@761a4                              v118 = iadd.i32 v20, v117
@761a7                              v119 = iconst.i32 904
@761aa                              v120 = iadd.i32 v20, v119
@761ab                              v121 = iconst.i32 16
@761ad                              v122 = iadd v120, v121
@761ae                              v123 = iconst.i32 96
@761b1                              v124 = iadd_imm.i64 v110, 17
@761b1                              store notrap aligned v124, v25+8
@761b1                              v125 = call fn1(v0, v0, v118, v122, v123)
@761b1                              v126 = load.i64 notrap aligned v25+8
@761ba                              v127 = iconst.i32 608
@761bd                              v128 = iadd.i32 v20, v127
@761c0                              v129 = iconst.i32 312
@761c3                              v130 = iadd.i32 v20, v129
@761c4                              v131 = iconst.i32 96
@761c7                              v132 = iadd_imm v126, 8
@761c7                              store notrap aligned v132, v25+8
@761c7                              v133 = call fn1(v0, v0, v128, v130, v131)
@761c7                              v134 = load.i64 notrap aligned v25+8
@761d2                              v135 = heap_addr.i64 heap0, v103, 1
@761d2                              store little v116, v135
@761d7                              v136 = iconst.i32 904
@761da                              v137 = iadd.i32 v20, v136
@761db                              v138 = iconst.i32 20
@761dd                              v139 = iadd v137, v138
@761e0                              v140 = iconst.i32 608
@761e3                              v141 = iadd.i32 v20, v140
@761e4                              v142 = iconst.i32 68
@761e7                              v143 = iadd v141, v142
@761e8                              v144 = heap_addr.i64 heap0, v143, 1
@761e8                              v145 = load.i32 little v144
@761eb                              v146 = heap_addr.i64 heap0, v139, 1
@761eb                              store little v145, v146
@761f0                              v147 = iconst.i32 904
@761f3                              v148 = iadd.i32 v20, v147
@761f4                              v149 = iconst.i32 32
@761f6                              v150 = iadd v148, v149
@761f9                              v151 = iconst.i32 312
@761fc                              v152 = iadd.i32 v20, v151
@761fd                              v153 = iconst.i32 8
@761ff                              v154 = iadd v152, v153
@76200                              v155 = heap_addr.i64 heap0, v154, 1
@76200                              v156 = load.i32 little v155
@76203                              v157 = heap_addr.i64 heap0, v150, 1
@76203                              store little v156, v157
@76208                              v158 = iconst.i32 904
@7620b                              v159 = iadd.i32 v20, v158
@7620c                              v160 = iconst.i32 44
@7620e                              v161 = iadd v159, v160
@76211                              v162 = iconst.i32 608
@76214                              v163 = iadd.i32 v20, v162
@76215                              v164 = iconst.i32 20
@76217                              v165 = iadd v163, v164
@76218                              v166 = heap_addr.i64 heap0, v165, 1
@76218                              v167 = load.i32 little v166
@7621b                              v168 = heap_addr.i64 heap0, v161, 1
@7621b                              store little v167, v168
@76220                              v169 = iconst.i32 904
@76223                              v170 = iadd.i32 v20, v169
@76224                              v171 = iconst.i32 56
@76226                              v172 = iadd v170, v171
@76229                              v173 = iconst.i32 608
@7622c                              v174 = iadd.i32 v20, v173
@7622d                              v175 = iconst.i32 32
@7622f                              v176 = iadd v174, v175
@76230                              v177 = heap_addr.i64 heap0, v176, 1
@76230                              v178 = load.i32 little v177
@76233                              v179 = heap_addr.i64 heap0, v172, 1
@76233                              store little v178, v179
@7623a                              v180 = heap_addr.i64 heap0, v20, 1
@7623a                              store.i32 little v105, v180+908
@76242                              v181 = heap_addr.i64 heap0, v20, 1
@76242                              store.i32 little v107, v181+904
@7624a                              v182 = heap_addr.i64 heap0, v20, 1
@7624a                              v183 = load.i64 little v182+668
@7624e                              v184 = heap_addr.i64 heap0, v20, 1
@7624e                              store little v183, v184+916
@76256                              v185 = heap_addr.i64 heap0, v20, 1
@76256                              v186 = load.i64 little v185+312
@7625a                              v187 = heap_addr.i64 heap0, v20, 1
@7625a                              store little v186, v187+928
@76262                              v188 = heap_addr.i64 heap0, v20, 1
@76262                              v189 = load.i64 little v188+620
@76266                              v190 = heap_addr.i64 heap0, v20, 1
@76266                              store little v189, v190+940
@7626e                              v191 = heap_addr.i64 heap0, v20, 1
@7626e                              v192 = load.i64 little v191+632
@76272                              v193 = heap_addr.i64 heap0, v20, 1
@76272                              store little v192, v193+952
@76278                              v194 = iconst.i32 904
@7627b                              v195 = iadd.i32 v20, v194
@7627c                              v196 = iconst.i32 68
@7627f                              v197 = iadd v195, v196
@76282                              v198 = iconst.i32 608
@76285                              v199 = iadd.i32 v20, v198
@76286                              v200 = iconst.i32 44
@76288                              v201 = iadd v199, v200
@76289                              v202 = heap_addr.i64 heap0, v201, 1
@76289                              v203 = load.i32 little v202
@7628c                              v204 = heap_addr.i64 heap0, v197, 1
@7628c                              store little v203, v204
@76291                              v205 = iconst.i32 904
@76294                              v206 = iadd.i32 v20, v205
@76295                              v207 = iconst.i32 80
@76298                              v208 = iadd v206, v207
@7629b                              v209 = iconst.i32 608
@7629e                              v210 = iadd.i32 v20, v209
@7629f                              v211 = iconst.i32 56
@762a1                              v212 = iadd v210, v211
@762a2                              v213 = heap_addr.i64 heap0, v212, 1
@762a2                              v214 = load.i32 little v213
@762a5                              v215 = heap_addr.i64 heap0, v208, 1
@762a5                              store little v214, v215
@762aa                              v216 = iconst.i32 904
@762ad                              v217 = iadd.i32 v20, v216
@762ae                              v218 = iconst.i32 92
@762b1                              v219 = iadd v217, v218
@762b4                              v220 = iconst.i32 608
@762b7                              v221 = iadd.i32 v20, v220
@762b8                              v222 = iconst.i32 80
@762bb                              v223 = iadd v221, v222
@762bc                              v224 = heap_addr.i64 heap0, v223, 1
@762bc                              v225 = load.i32 little v224
@762bf                              v226 = heap_addr.i64 heap0, v219, 1
@762bf                              store little v225, v226
@762c4                              v227 = iconst.i32 1008
@762c7                              v228 = iadd.i32 v20, v227
@762ca                              v229 = iconst.i32 608
@762cd                              v230 = iadd.i32 v20, v229
@762ce                              v231 = iconst.i32 92
@762d1                              v232 = iadd v230, v231
@762d2                              v233 = heap_addr.i64 heap0, v232, 1
@762d2                              v234 = load.i32 little v233
@762d5                              v235 = heap_addr.i64 heap0, v228, 1
@762d5                              store little v234, v235
@762dc                              v236 = heap_addr.i64 heap0, v20, 1
@762dc                              v237 = load.i64 little v236+644
@762e0                              v238 = heap_addr.i64 heap0, v20, 1
@762e0                              store little v237, v238+964
@762e8                              v239 = heap_addr.i64 heap0, v20, 1
@762e8                              v240 = load.i64 little v239+656
@762ec                              v241 = heap_addr.i64 heap0, v20, 1
@762ec                              store little v240, v241+976
@762f4                              v242 = heap_addr.i64 heap0, v20, 1
@762f4                              v243 = load.i64 little v242+680
@762f8                              v244 = heap_addr.i64 heap0, v20, 1
@762f8                              store little v243, v244+988
@76300                              v245 = heap_addr.i64 heap0, v20, 1
@76300                              v246 = load.i64 little v245+692
@76304                              v247 = heap_addr.i64 heap0, v20, 1
@76304                              store little v246, v247+1000
@7630a                              v248 = iconst.i32 496
@7630d                              v249 = iadd.i32 v20, v248
@76310                              v250 = iconst.i32 904
@76313                              v251 = iadd.i32 v20, v250
@76314                              v252 = iadd_imm v134, 142
@76314                              store notrap aligned v252, v25+8
@76314                              call fn5(v0, v0, v249, v251)
@76314                              v253 = load.i64 notrap aligned v25+8
@7631c                              v254 = iconst.i32 496
@7631f                              v255 = iadd.i32 v20, v254
@76320                              v256 = iconst.i32 4
@76322                              v257 = bor v255, v256
@76327                              v258 = heap_addr.i64 heap0, v20, 1
@76327                              v259 = load.i32 little v258+496
@7632b                              v260 = iadd_imm v253, 9
@7632b                              brnz v259, block6
@7632b                              jump block11

                                block11:
@7632f                              v261 = iconst.i32 720
@76332                              v262 = iadd.i32 v20, v261
@76335                              v263 = iconst.i32 76
@76338                              v264 = iadd_imm.i64 v260, 6
@76338                              store notrap aligned v264, v25+8
@76338                              v265 = call fn1(v0, v0, v262, v257, v263)
@76338                              v266 = load.i64 notrap aligned v25+8
@76341                              v267 = iconst.i32 408
@76344                              v268 = iadd.i32 v20, v267
@76347                              v269 = iconst.i32 720
@7634a                              v270 = iadd.i32 v20, v269
@7634b                              v271 = iconst.i32 76
@7634e                              v272 = iadd_imm v266, 8
@7634e                              store notrap aligned v272, v25+8
@7634e                              v273 = call fn1(v0, v0, v268, v270, v271)
@7634e                              v274 = load.i64 notrap aligned v25+8
@76357                              v275 = iconst.i32 8
@76359                              v276 = iadd.i32 v20, v275
@7635a                              v277 = iadd_imm v274, 4
@7635a                              store notrap aligned v277, v25+8
@7635a                              call fn6(v0, v0, v276)
@7635a                              v278 = load.i64 notrap aligned v25+8
@76364                              v279 = heap_addr.i64 heap0, v20, 1
@76364                              v280 = load.i64 little v279+8
@76367                              v281 = heap_addr.i64 heap0, v20, 1
@76367                              store little v280, v281+488
@7636d                              v282 = iconst.i32 824
@76370                              v283 = iadd.i32 v20, v282
@76373                              v284 = iconst.i32 408
@76376                              v285 = iadd.i32 v20, v284
@76377                              v286 = iconst.i32 76
@7637a                              v287 = iadd_imm v278, 12
@7637a                              store notrap aligned v287, v25+8
@7637a                              v288 = call fn1(v0, v0, v283, v285, v286)
@7637a                              v289 = load.i64 notrap aligned v25+8
@76383                              v290 = iconst.i32 720
@76386                              v291 = iadd.i32 v20, v290
@76389                              v292 = iconst.i32 824
@7638c                              v293 = iadd.i32 v20, v292
@7638d                              v294 = iadd_imm v289, 7
@7638d                              store notrap aligned v294, v25+8
@7638d                              call fn7(v0, v0, v291, v293)
@7638d                              v295 = load.i64 notrap aligned v25+8
@76395                              v296 = iconst.i32 904
@76398                              v297 = iadd.i32 v20, v296
@7639b                              v298 = iconst.i32 488
@7639e                              v299 = iadd.i32 v20, v298
@763a1                              v300 = iconst.i32 720
@763a4                              v301 = iadd.i32 v20, v300
@763a5                              v302 = iadd_imm v295, 10
@763a5                              store notrap aligned v302, v25+8
@763a5                              call fn8(v0, v0, v297, v299, v301)
@763a5                              v303 = load.i64 notrap aligned v25+8
@763ad                              v304 = iconst.i32 904
@763b0                              v305 = iadd.i32 v20, v304
@763b1                              v306 = iconst.i32 4
@763b3                              v307 = bor v305, v306
@763ba                              v308 = heap_addr.i64 heap0, v20, 1
@763ba                              v309 = load.i32 little v308+904
@763be                              v310 = iadd_imm v303, 9
@763be                              brnz v309, block12
@763be                              jump block13

                                block13:
@763c2                              v311 = iconst.i32 608
@763c5                              v312 = iadd.i32 v20, v311
@763c8                              v313 = iconst.i32 112
@763cb                              v314 = iadd_imm.i64 v310, 6
@763cb                              store notrap aligned v314, v25+8
@763cb                              v315 = call fn1(v0, v0, v312, v307, v313)
@763cb                              v316 = load.i64 notrap aligned v25+8
@763d4                              v317 = iconst.i32 496
@763d7                              v318 = iadd.i32 v20, v317
@763da                              v319 = iconst.i32 608
@763dd                              v320 = iadd.i32 v20, v319
@763de                              v321 = iconst.i32 112
@763e1                              v322 = iadd_imm v316, 8
@763e1                              store notrap aligned v322, v25+8
@763e1                              v323 = call fn1(v0, v0, v318, v320, v321)
@763e1                              v324 = load.i64 notrap aligned v25+8
@763ea                              v325 = iconst.i32 608
@763ed                              v326 = iadd.i32 v20, v325
@763f0                              v327 = iconst.i32 496
@763f3                              v328 = iadd.i32 v20, v327
@763f4                              v329 = iadd_imm v324, 7
@763f4                              store notrap aligned v329, v25+8
@763f4                              call fn9(v0, v0, v326, v328)
@763f4                              v330 = load.i64 notrap aligned v25+8
@763fc                              v331 = iconst.i32 904
@763ff                              v332 = iadd.i32 v20, v331
@76400                              v333 = iadd_imm v330, 4
@76400                              store notrap aligned v333, v25+8
@76400                              call fn10(v0, v0, v332)
@76400                              v334 = load.i64 notrap aligned v25+8
@76408                              v335 = iconst.i32 904
@7640b                              v336 = iadd.i32 v20, v335
@7640e                              v337 = iconst.i32 24
@76410                              v338 = iadd.i32 v20, v337
@76413                              v339 = iconst.i32 608
@76416                              v340 = iadd.i32 v20, v339
@76417                              v341 = iadd_imm v334, 10
@76417                              store notrap aligned v341, v25+8
@76417                              call fn11(v0, v0, v336, v338, v340)
@76417                              v342 = load.i64 notrap aligned v25+8
@7641f                              v344 = iconst.i32 8
@76421                              v345 = iadd.i32 v2, v344
@76424                              v346 = iconst.i32 904
@76427                              v347 = iadd.i32 v20, v346
@76428                              v348 = iconst.i32 424
@7642b                              v349 = iadd_imm v342, 8
@7642b                              store notrap aligned v349, v25+8
@7642b                              v350 = call fn1(v0, v0, v345, v347, v348)
@7642b                              v351 = load.i64 notrap aligned v25+8
@76434                              v352 = iconst.i32 0
@76436                              v353 = heap_addr.i64 heap0, v2, 1
@76436                              store little v352, v353
@7643d                              v354 = heap_addr.i64 heap0, v20, 1
@7643d                              v355 = load.i32 little v354+616
@76443                              v356 = icmp_imm eq v355, 0
@76443                              v357 = bint.i32 v356
@76444                              v358 = iadd_imm v351, 8
@76444                              brnz v357, block14(v20, v358, v25, v474)
@76444                              jump block15

                                block15:
@76448                              v359 = heap_addr.i64 heap0, v20, 1
@76448                              v360 = load.i32 little v359+608
@76450                              v361 = iconst.i32 12
@76452                              v362 = imul.i32 v355, v361
@76455                              v363 = iadd_imm.i64 v358, 7
@76455                              jump block16(v363, v25, v360, v362, v20, v474)

                                block16(v365: i64, v367: i64, v373: i32, v393: i32, v402: i32, v542: i32):
                                    v386 -> v367
                                    v372 -> v373
                                    v388 -> v373
                                    v392 -> v393
                                    v401 -> v402
                                    v541 -> v542
@76455                              v364 = iconst.i64 0
@76455                              v366 = ifcmp v365, v364
@76455                              brif sge v366, block18
@76455                              jump block19(v365)

                                block18:
@76455                              store.i64 notrap aligned v365, v367+8
@76455                              v368 = global_value.i64 gv3
@76455                              v369 = load.i64 notrap aligned readonly v368+0x0001_db38
@76455                              v370 = load.i64 notrap aligned readonly v369+152
@76455                              call_indirect sig0, v370(v368)
@76455                              v371 = load.i64 notrap aligned v367+8
@76455                              jump block19(v371)

                                block19(v380: i64):
@7645b                              v374 = iconst.i32 4
@7645d                              v375 = iadd.i32 v373, v374
@7645e                              v376 = heap_addr.i64 heap0, v375, 1
@7645e                              v377 = load.i32 little v376
@76463                              v378 = icmp_imm eq v377, 0
@76463                              v379 = bint.i32 v378
@76464                              v381 = iadd_imm v380, 7
@76464                              brnz v379, block20(v392, v381, v386, v401, v541)
@76464                              jump block21

                                block21:
@76468                              v382 = heap_addr.i64 heap0, v372, 1
@76468                              v383 = load.i32 little v382
@7646d                              v384 = iconst.i32 1
@7646f                              v385 = iadd_imm.i64 v381, 5
@7646f                              store notrap aligned v385, v367+8
@7646f                              call fn12(v0, v0, v383, v377, v384)
@7646f                              v387 = load.i64 notrap aligned v367+8
@76475                              jump block20(v392, v387, v367, v401, v541)

                                block20(v391: i32, v396: i64, v398: i64, v400: i32, v540: i32):
@76478                              v389 = iconst.i32 12
@7647a                              v390 = iadd.i32 v373, v389
@7647f                              v394 = iconst.i32 -12
@76481                              v395 = iadd v391, v394
@76484                              v397 = iadd_imm v396, 9
@76484                              brnz v395, block16(v397, v398, v390, v395, v400, v540)
@76484                              jump block22

                                block22:
@76486                              jump block17

                                block17:
@76487                              jump block14(v400, v397, v398, v540)

                                block14(v399: i32, v407: i64, v418: i64, v539: i32):
                                    v420 -> v399
                                    v425 -> v418
                                    v538 -> v539
@7648c                              v403 = heap_addr.i64 heap0, v399, 1
@7648c                              v404 = load.i32 little v403+612
@76492                              v405 = icmp_imm eq v404, 0
@76492                              v406 = bint.i32 v405
@76493                              v408 = iadd_imm v407, 5
@76493                              brnz v406, block23(v408)
@76493                              jump block24

                                block24:
@76497                              v409 = iconst.i32 12
@76499                              v410 = imul.i32 v404, v409
@7649c                              v411 = icmp_imm eq v410, 0
@7649c                              v412 = bint.i32 v411
@7649d                              v413 = iadd_imm.i64 v408, 6
@7649d                              brnz v412, block23(v413)
@7649d                              jump block25

                                block25:
@764a1                              v414 = heap_addr.i64 heap0, v399, 1
@764a1                              v415 = load.i32 little v414+608
@764a7                              v416 = iconst.i32 4
@764a9                              v417 = iadd_imm.i64 v413, 5
@764a9                              store notrap aligned v417, v418+8
@764a9                              call fn12(v0, v0, v415, v410, v416)
@764a9                              v419 = load.i64 notrap aligned v418+8
@764af                              jump block23(v419)

                                block23(v423: i64):
@764b2                              v421 = iconst.i32 496
@764b5                              v422 = iadd.i32 v399, v421
@764b6                              v424 = iadd_imm v423, 4
@764b6                              store notrap aligned v424, v418+8
@764b6                              call fn13(v0, v0, v422)
@764b6                              v426 = load.i64 notrap aligned v425+8
@764bc                              v427 = iadd_imm v426, 1
@764bc                              jump block5(v420, v427, v425, v538)

                                block12:
@764c1                              v428 = iconst.i32 608
@764c4                              v429 = iadd.i32 v20, v428
@764c5                              v430 = iconst.i32 8
@764c7                              v431 = iadd v429, v430
@764ca                              v432 = iconst.i32 8
@764cc                              v433 = iadd.i32 v307, v432
@764cd                              v434 = heap_addr.i64 heap0, v433, 1
@764cd                              v435 = load.i64 little v434
@764d2                              v436 = heap_addr.i64 heap0, v431, 1
@764d2                              store little v435, v436
@764d9                              v437 = heap_addr.i64 heap0, v307, 1
@764d9                              v438 = load.i64 little v437
@764de                              v439 = heap_addr.i64 heap0, v20, 1
@764de                              store little v438, v439+608
@764e4                              v440 = iconst.i32 904
@764e7                              v441 = iadd.i32 v20, v440
@764e8                              v442 = iconst.i32 8
@764ea                              v443 = iadd v441, v442
@764ed                              v444 = heap_addr.i64 heap0, v443, 1
@764ed                              store little v435, v444
@764f4                              v445 = heap_addr.i64 heap0, v20, 1
@764f4                              store little v438, v445+904
@764fa                              v446 = iconst.i32 4
@764fc                              v447 = iadd.i32 v2, v446
@764ff                              v448 = iconst.i32 904
@76502                              v449 = iadd.i32 v20, v448
@76503                              v450 = iadd_imm.i64 v310, 33
@76503                              store notrap aligned v450, v25+8
@76503                              call fn14(v0, v0, v447, v449)
@76503                              v451 = load.i64 notrap aligned v25+8
@7650b                              v452 = iconst.i32 1
@7650d                              v453 = heap_addr.i64 heap0, v2, 1
@7650d                              store little v452, v453
@76510                              v454 = iadd_imm v451, 4
@76510                              jump block5(v20, v454, v25, v474)

                                block8:
@76515                              v455 = iconst.i32 904
@76518                              v456 = iadd.i32 v20, v455
@76519                              v457 = iadd_imm.i64 v33, 4
@76519                              store notrap aligned v457, v25+8
@76519                              call fn10(v0, v0, v456)
@76519                              v458 = load.i64 notrap aligned v25+8
@76521                              v459 = iconst.i32 405
@76524                              v460 = iadd_imm v458, 3
@76524                              store notrap aligned v460, v25+8
@76524                              v461 = call fn15(v0, v0, v459)
@76524                              v462 = load.i64 notrap aligned v25+8
@7652a                              v463 = heap_addr.i64 heap0, v20, 1
@7652a                              istore16 little v461, v463+1320
@76530                              v464 = iconst.i32 8
@76532                              v465 = iadd.i32 v343, v464
@76535                              v466 = iconst.i32 904
@76538                              v467 = iadd.i32 v20, v466
@76539                              v468 = iconst.i32 424
@7653c                              v469 = iadd_imm v462, 9
@7653c                              store notrap aligned v469, v25+8
@7653c                              v470 = call fn1(v0, v0, v465, v467, v468)
@7653c                              v471 = load.i64 notrap aligned v25+8
@76545                              v472 = iconst.i32 0
@76547                              v473 = heap_addr.i64 heap0, v343, 1
@76547                              store little v472, v473
@7654c                              v475 = iadd_imm v471, 5
@7654c                              store notrap aligned v475, v25+8
@7654c                              call fn16(v0, v0, v4)
@7654c                              v476 = load.i64 notrap aligned v25+8
@76554                              v477 = iadd_imm v476, 2
@76554                              store notrap aligned v477, v25+8
@76554                              call fn17(v0, v0, v22)
@76554                              v478 = load.i64 notrap aligned v25+8
@7655a                              v479 = iadd_imm v478, 1
@7655a                              jump block4(v20, v479, v25)

                                block7:
@7655f                              v480 = iconst.i32 4
@76561                              v481 = iadd.i32 v2, v480
@76566                              v482 = iadd_imm.i64 v110, 6
@76566                              store notrap aligned v482, v25+8
@76566                              call fn18(v0, v0, v481, v107, v105)
@76566                              v483 = load.i64 notrap aligned v25+8
@7656e                              v484 = iconst.i32 1
@76570                              v485 = heap_addr.i64 heap0, v2, 1
@76570                              store little v484, v485
@76573                              v486 = iadd_imm v483, 4
@76573                              jump block5(v20, v486, v25, v474)

                                block6:
@76578                              v487 = iconst.i32 720
@7657b                              v488 = iadd.i32 v20, v487
@7657c                              v489 = iconst.i32 24
@7657e                              v490 = iadd v488, v489
@76581                              v491 = iconst.i32 24
@76583                              v492 = iadd.i32 v257, v491
@76584                              v493 = heap_addr.i64 heap0, v492, 1
@76584                              v494 = load.i32 little v493
@76589                              v495 = heap_addr.i64 heap0, v490, 1
@76589                              store little v494, v495
@7658e                              v496 = iconst.i32 720
@76591                              v497 = iadd.i32 v20, v496
@76592                              v498 = iconst.i32 16
@76594                              v499 = iadd v497, v498
@76597                              v500 = iconst.i32 16
@76599                              v501 = iadd.i32 v257, v500
@7659a                              v502 = heap_addr.i64 heap0, v501, 1
@7659a                              v503 = load.i64 little v502
@7659f                              v504 = heap_addr.i64 heap0, v499, 1
@7659f                              store little v503, v504
@765a4                              v505 = iconst.i32 720
@765a7                              v506 = iadd.i32 v20, v505
@765a8                              v507 = iconst.i32 8
@765aa                              v508 = iadd v506, v507
@765ad                              v509 = iconst.i32 8
@765af                              v510 = iadd.i32 v257, v509
@765b0                              v511 = heap_addr.i64 heap0, v510, 1
@765b0                              v512 = load.i64 little v511
@765b5                              v513 = heap_addr.i64 heap0, v508, 1
@765b5                              store little v512, v513
@765bc                              v514 = heap_addr.i64 heap0, v257, 1
@765bc                              v515 = load.i64 little v514
@765c1                              v516 = heap_addr.i64 heap0, v20, 1
@765c1                              store little v515, v516+720
@765c7                              v517 = iconst.i32 28
@765c9                              v518 = iadd.i32 v2, v517
@765cc                              v519 = heap_addr.i64 heap0, v518, 1
@765cc                              store little v494, v519
@765d1                              v520 = iconst.i32 20
@765d3                              v521 = iadd.i32 v2, v520
@765d6                              v522 = heap_addr.i64 heap0, v521, 1
@765d6                              store little v503, v522
@765db                              v523 = iconst.i32 12
@765dd                              v524 = iadd.i32 v2, v523
@765e0                              v525 = heap_addr.i64 heap0, v524, 1
@765e0                              store little v512, v525
@765e7                              v526 = heap_addr.i64 heap0, v2, 1
@765e7                              store little v515, v526+4
@765ec                              v527 = iconst.i32 1
@765ee                              v528 = heap_addr.i64 heap0, v2, 1
@765ee                              store little v527, v528
@765f1                              v529 = iadd_imm.i64 v260, 59
@765f1                              jump block5(v20, v529, v25, v474)

                                block5(v530: i32, v533: i64, v535: i64, v537: i32):
@765f4                              v531 = iconst.i32 24
@765f6                              v532 = iadd v530, v531
@765f7                              v534 = iadd_imm v533, 4
@765f7                              store notrap aligned v534, v535+8
@765f7                              call fn17(v0, v0, v532)
@765f7                              v536 = load.i64 notrap aligned v535+8
@765ff                              v543 = iadd_imm v536, 2
@765ff                              store notrap aligned v543, v535+8
@765ff                              call fn16(v0, v0, v537)
@765ff                              v544 = load.i64 notrap aligned v535+8
@76605                              jump block4(v530, v544, v535)

                                block4(v545: i32, v549: i64, v551: i64):
@76608                              v546 = iconst.i32 1328
@7660b                              v547 = iadd v545, v546
@7660c                              v548 = global_value.i64 gv3
@7660c                              store notrap aligned v547, v548+1104
@76612                              v550 = iadd_imm v549, 4
@76612                              jump block1

                                block1:
@76612                              store.i64 notrap aligned v550, v551+8
@76612                              return
}

May 17 08:04:03.255 DEBUG with options: checker: false, algorithm: backtracking (block annotations: false)    
May 17 08:04:03.255 DEBUG Compiling (opt level Speed):
function u0:2510(i64 vmctx, i64, i32, i32) fast {
    gv0 = vmctx
    gv1 = load.i64 notrap aligned readonly gv0
    gv2 = load.i64 notrap aligned gv1
    gv3 = vmctx
    gv4 = load.i64 notrap aligned readonly gv3+1076
    heap0 = static gv4, min 0, bound 0x0001_0000_0000, offset_guard 0x8000_0000, index_type i32
    sig0 = (i64 vmctx) wasmtime_system_v
    sig1 = (i64 vmctx, i64, i32, i32, i32, i32) fast
    sig2 = (i64 vmctx, i64, i32, i32, i32) fast
    sig3 = (i64 vmctx, i64, i32, i32) -> i32 fast
    sig4 = (i64 vmctx, i64, i32, i32, i32) fast
    sig5 = (i64 vmctx, i64, i32, i32) fast
    fn0 = colocated u0:2503 sig1
    fn1 = colocated u0:112 sig2
    fn2 = colocated u0:111 sig3
    fn3 = colocated u0:2505 sig4
    fn4 = colocated u0:4772 sig5
    stack_limit = gv2

• We've also had this discussion a little bit inside #146, but off-the-cuff I would expect that RUST_LOG would have an effect on "application code", and not on the server. I know you're vary about exporting environment variables to the wasm environment, but not being able to change log levels locally without changing the code is a bit of a pain. I know this isn't the issue for that, but consider ensuring that you're not locking yourself into a solution where this becomes impossible in the future.