cmd/vet: panics with inconsistent import

[gyuho]

I think I found a strange edge case for go vet.

What version of Go are you using (go version)?

go version devel +691c5c1 Sun Jun 19 05:18:09 2016 +0000 linux/amd64

What operating system and processor architecture are you using (go env)?

go env
GOARCH="amd64"
GOBIN=""
GOEXE=""
GOHOSTARCH="amd64"
GOHOSTOS="linux"
GOOS="linux"
GOPATH="/home/gyuho/go"
GORACE=""
GOROOT="/home/gyuho/go-master"
GOTOOLDIR="/home/gyuho/go-master/pkg/tool/linux_amd64"
CC="gcc"
GOGCCFLAGS="-fPIC -m64 -pthread -fmessage-length=0 -fdebug-prefix-map=/tmp/go-build259744847=/tmp/go-build -gno-record-gcc-switches"
CXX="g++"
CGO_ENABLED="1"

What did you do?

Here are my file structure:

~/go/src/github.com/gyuho/aaa $ find .
.
./1.go
./2.go
./fileutil
./fileutil/fileutil.go

// 1.go
package aaa

import "github.com/gyuho/aaa/fileutil"

type WAL struct {
    lockedFiles []*fileutil.LockedFile
}

// 2.go
package aaa

import "time"

const (
    warnSyncDuration = time.Second
)

// fileutil/fileutil.go
package fileutil

import "os"

type LockedFile struct {
    *os.File
}

And I just ran go vet *.go in ~/go/src/github.com/gyuho/aaa

What did you expect to see?

It shouldn't panic.

What did you see instead?

go vet *.go

panic: inconsistent import:
    var time.UTC *time.Location
previously imported as:
    var UTC *time.Location
 [recovered]
    panic: inconsistent import:
    var time.UTC *time.Location
previously imported as:
    var UTC *time.Location


goroutine 1 [running]:
panic(0x662f00, 0xc4201e8420)
    /home/gyuho/go-master/src/runtime/panic.go:500 +0x1a1
go/types.(*Checker).handleBailout(0xc4200a61c0, 0xc4200b98e8)
    /home/gyuho/go-master/src/go/types/check.go:213 +0xae
panic(0x662f00, 0xc4201e8420)
    /home/gyuho/go-master/src/runtime/panic.go:458 +0x243
go/internal/gcimporter.(*importer).declare(0xc42007e480, 0x7de1e0, 0xc4201e7bd0)
    /home/gyuho/go-master/src/go/internal/gcimporter/bimport.go:173 +0x17c
go/internal/gcimporter.(*importer).obj(0xc42007e480, 0xfffffffffffffffc)
    /home/gyuho/go-master/src/go/internal/gcimporter/bimport.go:193 +0x1af
go/internal/gcimporter.BImportData(0xc42000b110, 0xc4201a0000, 0x39d34, 0x3fe00, 0xc42001199b, 0x4, 0x0, 0x4, 0x10, 0x10)
    /home/gyuho/go-master/src/go/internal/gcimporter/bimport.go:86 +0x34c
go/internal/gcimporter.Import(0xc42000b110, 0xc42001199b, 0x4, 0x6bba44, 0x1, 0xc42014c3c0, 0x0, 0x0)
    /home/gyuho/go-master/src/go/internal/gcimporter/gcimporter.go:166 +0x551
go/importer.gcimports.ImportFrom(0xc42000b110, 0xc42001199b, 0x4, 0x6bba44, 0x1, 0x0, 0x1, 0x2, 0x0)
    /home/gyuho/go-master/src/go/importer/importer.go:70 +0x67
go/types.(*Checker).collectObjects(0xc4200a61c0)
    /home/gyuho/go-master/src/go/types/resolver.go:191 +0x81d
go/types.(*Checker).checkFiles(0xc4200a61c0, 0xc420011a30, 0x2, 0x2, 0x0, 0x0)
    /home/gyuho/go-master/src/go/types/check.go:225 +0xaa
go/types.(*Checker).Files(0xc4200a61c0, 0xc420011a30, 0x2, 0x2, 0xc42013ede0, 0xc4200b9960)
    /home/gyuho/go-master/src/go/types/check.go:218 +0x49
go/types.(*Config).Check(0xc420015440, 0xc4200118b0, 0x3, 0xc420015240, 0xc420011a30, 0x2, 0x2, 0xc420141860, 0x20, 0xc4200001a0, ...)
    /home/gyuho/go-master/src/go/types/api.go:344 +0x19f
main.(*Package).check(0xc4200608a0, 0xc420015240, 0xc420011a30, 0x2, 0x2, 0xc420011a40, 0x1)
    /home/gyuho/go-master/src/cmd/vet/types.go:76 +0x36c
main.doPackage(0x6bba44, 0x1, 0xc42000a3d0, 0x2, 0x2, 0x0)
    /home/gyuho/go-master/src/cmd/vet/main.go:333 +0x992
main.main()
    /home/gyuho/go-master/src/cmd/vet/main.go:241 +0x335
exit status 2

Could anyone take a look? Or am I doing something wrong?

It's weird. If I put all the code in one file (without 2.go) in top aaa package, I don't see this panic.

Thanks!

[ALTree]

Bisected to CL 22123, so cc @griesemer I guess.

[ianlancetaylor]

I can't recreate the problem. I don't see any warnings or panics from vet.

Does it get fixed if you run go install?

[gyuho]

It's weird. It's not happening in my local machine either. I still keep the same go version. Guess my machine had cached something in tmp directory that might have caused this.

Please feel free to close it.

When I can reliably reproduce, I will revisit this issue.

Thanks!

[josharian]

OK. Thanks!