> > Why not just pass the C pointer from one package to another? > >From docs:
> Cgo translates C types into equivalent unexported Go types. Because the > translations are unexported, a Go package should not expose C types in its > exported API: a C type used in one Go package is different from the same C > type used in another. There is a issue to allow that tho https://github.com/golang/go/issues/13467 Or an unsafe.Pointer? Same reason syscall doesn't use unsafe.Pointer, so you have to import unsafe to convert. Make procAddr a *byte I don't want to risk the GC (or even the user) thinking it's a Go pointer and doing stuff it shouldn't. CGO uses uintptr for certain pointer types, so thought it's better to be safe, the docs says: These types are uintptr on the Go side because they would otherwise confuse > the Go garbage collector; they are sometimes not really pointers but data > structures encoded in a pointer type. All operations on these types must > happen in C. The proper constant to initialize an empty such reference is > 0, not nil. On Friday, 14 February 2020 07:40:46 UTC+8, Keith Randall wrote: > > Why all the conversions? Why not just pass the C pointer from one package > to another? Or an unsafe.Pointer? You can use a placeholder type, even. > Make procAddr a *byte, for example. > > The pointer conversion rules are intended to ensure safety when playing > with Go pointers (by that, I mean pointers to objects in the Go heap or > stack). They're definitely overkill for pointers that you know point into > the C heap. Unfortunately, the vet rules don't know where the pointers came > from. > > On Thursday, February 13, 2020 at 9:37:47 AM UTC-8 Alex wrote: > >> I have to use a massive third party API so it's not really a choice to >> use cgo. >> >> Do you have any safer options to pass C pointers between packages? >> The C pointers are all C allocated and package B needs to call C >> functions directly and pass those pointers. >> >> On Friday, 14 February 2020 01:02:24 UTC+8, Jake Montgomery wrote: >>> >>> You need to read https://golang.org/pkg/unsafe/#Pointer *very, very, >>> very* carefully before using unsafe.Pointer in any way. It spells out 6 >>> conversions that are considered "valid". It says: "Code not using these >>> patterns is likely to be invalid today or to become invalid in the future." >>> AFAICT, your code does not fit any of those 6 allowable patterns. >>> >>> Go is simple and easy, CGO is tricky, difficult and full of dragons. >>> >>> >>> On Thursday, February 13, 2020 at 11:42:30 AM UTC-5, Alex wrote: >>>> >>>> I have to pass C pointers between packages so I used uintptr like how >>>> syscall does things. >>>> However go vet gives the message "possible misuse of unsafe.Pointer". >>>> >>>> Is there something I could do to avoid vet complaining? >>>> >>>> Package A: >>>> type Foo struct { >>>> procAddr uintptr >>>> } >>>> >>>> func (f Foo) ProcAddr() uintptr { >>>> return f.procAddr >>>> } >>>> >>>> func Bar() Foo { >>>> return Foo{C.SomeFunc()} >>>> } >>>> >>>> Package B: >>>> p := C.ASDF(unsafe.Pointer(A.Bar().ProcAddr())) // possible misuse of >>>> unsafe.Pointer >>>> >>>> -- You received this message because you are subscribed to the Google Groups "golang-nuts" group. To unsubscribe from this group and stop receiving emails from it, send an email to golang-nuts+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/golang-nuts/d7f4adab-0459-43fe-a299-0f27b6f1be5c%40googlegroups.com.
