The EnumPrinters Win32 function takes and argument _Out_ LPBYTE pPrinterEnum, a pointer to an allocated buffer. In C, it works like this:
DWORD cbNeeded, nPrinters;
EnumPrinters(PRINTER_ENUM_LOCAL, NULL, 5, NULL, 0, &cbNeeded, &nPrinters);
BYTE *pPrnInfo = malloc(cbNeeded);
EnumPrinters(PRINTER_ENUM_LOCAL, NULL, 5, pPrnInfo, cbNeeded, &cbNeeded, &nPrinters);
PRINTER_INFO_5 *pPrinterInfo = (PRINTER_INFO_5 *) pPrnInfo;
for (int i=0; i < nPrinters; i++) {
printf("pPrinterName: %s\n", pPrinterInfo[i].pPrinterName);
}
How is the same accomplished in Go, using syscall instead of cgo? So far, this much compiles, but I don't know how to cast the resulting byte slice to an array of structs (without using cgo).
type PrinterInfo5 struct {
pPrinterName *uint16
pPortName *uint16
attributes uint32
deviceNotSelectedTimeout uint32
transmissionRetryTimeout uint32
}
...
dll := syscall.MustLoadDLL("winspool.drv")
f := dll.MustFindProc("EnumPrintersW")
var cbNeeded, nPrinters uint32
fmt.Println(cbNeeded, nPrinters)
f.Call(PRINTER_ENUM_LOCAL, 0, 5, 0, 0, uintptr(unsafe.Pointer(&cbNeeded)), uintptr(unsafe.Pointer(&nPrinters)))
fmt.Println(cbNeeded, nPrinters)
var pPrnInfo []byte = make([]byte, cbNeeded)
f.Call(PRINTER_ENUM_LOCAL, 0, 5, uintptr(unsafe.Pointer(&pPrnInfo)), uintptr(cbNeeded), uintptr(unsafe.Pointer(&cbNeeded)), uintptr(unsafe.Pointer(&nPrinters)))
I have tried this, which prints one iteration successfully, then fails with fatal error: heapBitsBulkBarrier: unaligned arguments:
hdr := reflect.SliceHeader{
Data: uintptr(unsafe.Pointer(&pPrnInfo)),
Len: int(nPrinters),
Cap: int(nPrinters),
}
s := *(*[]PrinterInfo5)(unsafe.Pointer(&hdr))
for _, t := range s {
fmt.Println(t)
}
uintptr(unsafe.Pointer(&pPrnInfo))
in both places in the code above is wrong; it gives you a pointer to the slice header, not to the actual backing array. You want this instead:
uintptr(unsafe.Pointer(&pPrnInfo[0]))
(Since the backing array is contiguous, a pointer to the first element of the backing array is the same as a pointer to the backing array itself.)
Related
Given a type such as:
type LicenseCards struct {
cards *[]int
}
I won't show the code that creates the slice. But this removes the top item, ignoring the zero-length case.
func (licenseCards *LicenseCards) PopLicenseCard() int {
l := len(*licenseCards.cards)
ret := (*licenseCards.cards)[l-1]
*licenseCards.cards = (*licenseCards.cards)[:l-1]
return ret
}
If I remove the last item from the slice and return a pointer to the removed item, is it guaranteed to still be available?
As #Volker said the memory will not be released by the GC if something is using it.
Another point with your code is that you do not need to dereference a pointer (using * operator) before using the . operator eg: just do this: l := len(licenseCards.cards).
Also you don't need cards and the receiver to both be pointers. If you don't mind I would like to suggest this:
type LicenseCards struct {
cards []int
}
func (lc *LicenseCards) PopLicenseCard() int {
l := len(lc.cards)
ret := lc.cards[l-1]
lc.cards = lc.cards[:l-1]
return ret
}
Given two nested types
type Inner struct {
InnerVal int
}
type Outer struct {
InnerStruct *Inner
OuterVal int
}
I need to duplicate a slice of pointers to Outer
originalSlice := []*Outer{<plenty_of_items>}
with itself, but having updated field values in the duplicates, including the Outer.InnerStruct.InnerVal.
To do so I create a new slice of the same type and length as originalSlice, append pointers to newly created structs with altered values to it, and finally append these items to the originalSlice
duplicateSlice := make([]*Outer, len(originalSlice))
for _, originalItem := range originalSlice {
duplicateSlice = append(duplicateSlice, &Outer{
InnerStruct: &Inner{
InnerVal: originalItem.InnerStruct.InnerVal + 1
},
OuterVal: originalItem.OuterVal + 1,
})
}
originalSlice = append(originalSlice, duplicateSlice...)
While this is verbose enough to follow the pointers around, or so I thought, when passed to a function right after as nowDoubledSlice, and accessed via loop
someOtherSlice := make([]*types.Inner, len(nowDoubledSlice))
for i, doubledItem := range nowDoubledSlice {
someOtherSlice[i] = doubledItem.InnerStruct
}
I get a
runtime error: invalid memory address or nil pointer dereference
Why is that? And is there a more elegant or idiomatic way to duplicate a slice of pointers to structs, while altering the duplicates' fields?
It's nothing to do with your pointer creation, it's your slice allocation. This line:
duplicateSlice := make([]*Outer, len(originalSlice))
Creates a new slice of length len(originalSlice), filled with zero-value elements. What you likely want instead is:
duplicateSlice := make([]*Outer, 0, len(originalSlice))
to create a slice of length 0 but capacity of len(originalSlice). This works fine, as you can see here.
Alternatively, you could keep make([]*Outer, len(originalSlice)) and use indexing instead of append in your loop:
for i, originalItem := range originalSlice {
duplicateSlice[i] =&Outer{
InnerStruct: &Inner{
InnerVal: originalItem.InnerStruct.InnerVal + 1,
},
OuterVal: originalItem.OuterVal + 1,
}
}
Which works just as well, as you can see here.
I'm using syscall.Syscall(...) to call a C method in a dll.
This is the C method signature:
SENSEI_API HSENSEI SENSEI_open(const char* sensigrafo, const char* options, SENSEI_ERR* se);
This is the SENSEI_ERR struct:
typedef struct
{
int code;
char* error_string;
} SENSEI_ERR;
In my GO program I declared a struct:
type senseiErr struct {
code int
error_string *byte
}
And tried to call the method:
var nargs uintptr = 3
var err senseiErr
ret, _, callErr := syscall.Syscall(uintptr(senseiOpen),
nargs,
uintptr(unsafe.Pointer(syscall.StringToUTF16Ptr("en"))),
uintptr(unsafe.Pointer(syscall.StringToUTF16Ptr(""))),
uintptr(unsafe.Pointer(&err)),
)
As you may have guessed, the SENSEI_open method fill the SENSEI_ERR argument with the code and the text of the error.
Now I need to read the content of that error.
err.code actually has the correct value.
About err.error_string I don't know. I'm new to GO and i have some questions:
Since the C struct has the field char* error_string, is error_string *byte in my GO struct correct?
Should I use []byte or something else?
How do I read the content of the error_string field?
fmt.Println(err.error_string) prints the memory address
fmt.Println(*err.error_string) prints always "101"
1) I doubt that cost char* meant to be UTF16 encoded. So all what you need is just getting raw data:
sensigrafo := "en\000" // \000 = 0 = null termination, \0 does not valid
options := "\000"
...
uintptr(*(*unsafe.Pointer)(unsafe.Pointer(&sensigrafo))
uintptr(*(*unsafe.Pointer)(unsafe.Pointer(&options))
// *(*unsafe.Pointer) are accessing the first field of string header:
type string struct {
data *byte
len int
}
// same with slices
// but for them there's less ugly way:
sensigrafo := []byte("en\000")
options := []byte("\000")
uintptr(unsafe.Pointer(&sensigrafo[0]))
uintptr(unsafe.Pointer(&options[0]))
2) C's int and Golang's int might have different sizeof, so this requires cgo declaration (C.int) or manual matching with random selection (try also int32, int64 if you don't want to use cgo)
type senseiErr struct {
code C.int /* Golang's int32/int64 */
error_string *byte // pointer types are same as C's void* or Golang's unsafe.Pointer
}
Wrong offset might cause error_string be empty or point to random addr.
3) To read content you have to use same methods as C does (read data until null terminated byte, considering that *byte points to first element of string), but I propose to use already implemented runtime functions:
//go:linkname gostringn runtime.gostringn
func gostringn(p *byte, l int) string
//go:linkname findnull runtime.findnull
//go:nosplit
func findnull(s *byte) int
...
error_string := gostringn(err.error_string, findnull(err.error_string))
// or cgo one:
type senseiErr struct {
code C.int
error_string *C.char
}
...
error_string := C.GoString(err.error_string)
I'v been trying to play with netapi32.dll, but I'm having mixed results.
The following works as expected
type SERVER_INFO_101 struct {
PlatformID uint32
Name *uint16
VersionMajor uint32
VersionMinor uint32
Type uint32
Comment *uint16
}
func NetServerGetInfo() {
info := &SERVER_INFO_101{}
ret, _, err := procNetServerGetInfo.Call(0, 101, uintptr(unsafe.Pointer(&info)))
if ret != 0 {
log.Fatal(err)
}
spew.Dump(info)
}
However, I'm not sure why info has to have & inside the unsafe.Pointer also.
The following does not work, and I can't seem to find out why. No error codes get thrown. Neither the struct or variables gets filled out.
type SESSION_INFO_10 struct {
Cname *uint16
Username *uint16
Time uint32
IdleTime uint32
}
func NetSessionEnum() {
info := &SESSION_INFO_10{}
var prefmaxlen int32 = -1
var entriesread uint32
var totalentries uint32
var resumehandle uint32
x, y, z := procNetSessionEnum.Call(0, 0, 0, 10, uintptr(unsafe.Pointer(info)), uintptr(prefmaxlen), uintptr(unsafe.Pointer(&entriesread)), uintptr(unsafe.Pointer(&totalentries)), uintptr(unsafe.Pointer(&resumehandle)))
fmt.Println(x, y, z)
fmt.Println(entriesread, totalentries)
spew.Dump(info)
}
…because you're not supposed to pass a pointer to your memory block there—to cite the manual:
This buffer is allocated by the system and must be freed using the NetApiBufferFree function.
The type of that pointer is misleading but you're supposed to pass a pointer to a pointer there, something like this:
func NetSessionEnum() {
var pinfo *SESSION_INFO_10
var prefmaxlen int32 = -1
var entriesread uint32
var totalentries uint32
var resumehandle uint32
x, y, z := procNetSessionEnum.Call(0, 0, 0, 10,
uintptr(unsafe.Pointer(&pinfo)), uintptr(prefmaxlen),
uintptr(unsafe.Pointer(&entriesread)),
uintptr(unsafe.Pointer(&totalentries)),
uintptr(unsafe.Pointer(&resumehandle)))
fmt.Println(x, y, z)
fmt.Println(entriesread, totalentries)
spew.Dump(info)
}
// Now use `*pinfo.Cname` etc
// Don't forget to later call `NetApiBufferFree()` on that pointer.
What happens here:
The variable pinfo is a pointer to a value of type SESSION_INFO_10.
You take the address of the memory block occupied by the value kept in that variable (which is a pointer) and pass it to NetSessionEnum().
That function allocates the buffer by itself and writes its address to the memory block pointed to by the address you have passed to the function.
Since you've passed an address of the pinfo variable, the address of the buffer ends up being written into the variable pinfo.
You then use that address stored in pinfo to access the memory allocated by NetSessionEnum().
That's called "double indirection" and is used in quite many places of Win32 API. Please read the manual page and study the code example it includes.
Update: as it turned out, there were more problems with the original code so I've took time to provide full solution—here is the gist (tested with Go 1.6 amd64 and i386 on Windows XP 32-bit, Windows 2003 R2 64-bit and Windows 8.1 64-bit).
I'm working in Go 1.6 on Windows and trying to export a certificate container to a PFX (the ultimate goal here is to access an exportable private key from the certificate store).
I have opened a memory store and inserted a certificate into the store:
var storedCertCtx *syscall.CertContext
storeHandle, err := syscall.CertOpenStore(syscall.CERT_STORE_PROV_MEMORY, 0, 0, syscall.CERT_STORE_DEFER_CLOSE_UNTIL_LAST_FREE_FLAG, 0)
err = syscall.CertAddCertificateContextToStore(storeHandle, certenum, syscall.CERT_STORE_ADD_ALWAYS, &storedCertCtx)
Now I want to generate a PFX of that store. I have defined a struct for containing the data blob and want to use PFXExportCertStoreEx to get a PFX of the store:
var (
crypt32 = syscall.NewLazyDLL("crypt32.dll")
procPFXExportCertStoreEx = crypt32.NewProc("PFXExportCertStoreEx")
)
type CRYPTOAPI_BLOB struct {
DataSize uint32
Data *byte
}
var pfxBlob CRYPTOAPI_BLOB
err = PfxExportCertStore(storeHandle, &pfxBlob, syscall.StringToUTF16Ptr("MyPassword"), 0, 0)
syscall.Syscall6(procPFXExportCertStoreEx.Addr(), 5,
uintptr(storeHandle), //hStore
uintptr(unsafe.Pointer(&pfxBlob)), //*pPFX
uintptr(unsafe.Pointer(syscall.StringToUTF16Ptr("password"))), //szPassword
0, //*pvPara
0, //dwFlags
0)
And this half works.
DataSize is populated with what looks like an appropriate value (i.e. if I add more certificates to the store, it grows bigger), however Data is always <nil>.
Seeing as it's meant to be populated with a pointer, I have tried declaring it as *uintptr and uint32 (just to see if anything gets populated), but nothing. The value is always untouched (if I manually put junk data in there, the junk data stays after the syscall is executed).
Have I defined the struct incorrectly? There is precious few examples to go for getting this done in Go, but from what I can see from the numerous C examples, this should be working.
This is the expected behavior.
According to this: https://msdn.microsoft.com/en-us/library/windows/desktop/aa387313(v=vs.85).aspx, the pPFX struct requires a pre-allocated buffer, with the size in the cbData field, which will be updated with the size of the data copied in.
If the call is made with pbData equal to NULL, only the cbData field is updated to reflect the size needed for the output buffer.
JimB's answer is most certainly correct, but I want to add this for followup in case anyone else is going down this path. The actual code that I had to use to get the PFX file into CRYPTOAPI_BLOB was:
var (
crypt32 = syscall.NewLazyDLL("crypt32.dll")
procPFXExportCertStoreEx = crypt32.NewProc("PFXExportCertStoreEx")
procCryptMemAlloc = crypt32.NewProc("CryptMemAlloc")
procCryptMemFree = crypt32.NewProc("CryptMemFree")
)
type CRYPTOAPI_BLOB struct {
cbData uint32
pbData *byte
}
func (b *CRYPTOAPI_BLOB) ToByteArray() []byte {
d := make([]byte, b.cbData)
copy(d, (*[1 << 30]byte)(unsafe.Pointer(b.pbData))[:])
return d
}
func PfxExportCertStore(storeHandle syscall.Handle, password string, flags uint32) (returnData []byte, err error) {
var pfxBlob CRYPTOAPI_BLOB
r1, _, _ := syscall.Syscall6(procPFXExportCertStoreEx.Addr(), 5,
uintptr(storeHandle), //hStore
uintptr(unsafe.Pointer(&pfxBlob)), //*pPFX
uintptr(unsafe.Pointer(syscall.StringToUTF16Ptr(password))), //szPassword
0, //*pvPara
uintptr(flags), //dwFlags
0)
r2, _, _ := syscall.Syscall(procCryptMemAlloc.Addr(), 1, uintptr(unsafe.Pointer(&pfxBlob.cbData)), 0, 0)
p := unsafe.Pointer(&r2)
q := (*byte)(p)
pfxBlob.pbData = q
defer syscall.Syscall(procCryptMemFree.Addr(), 1, uintptr(unsafe.Pointer(pfxBlob.pbData)), 0, 0)
r3, _, _ := syscall.Syscall6(procPFXExportCertStoreEx.Addr(), 5,
uintptr(storeHandle), //hStore
uintptr(unsafe.Pointer(&pfxBlob)), //*pPFX
uintptr(unsafe.Pointer(syscall.StringToUTF16Ptr(password))), //szPassword
0, //*pvPara
uintptr(flags), //dwFlags
0)
returnData = pfxBlob.ToByteArray()
return
}
(I have stripped the error handling to make it easier to read). The first call to PFXExportCertStoreEx just returns the size, and once we have the size we can do a call to PFXExportCertStoreEx to allocate a buffer, and then we pass the same pointer to PFXExportCertStoreEx, but this time it has the allocated buffer, and we get the full PFX file returned.