xeirwn
commented
7 years ago What write_stream is can be found in the manual under the name write_out.
This callback receives as input the pointer to an element (const void *buffer), the size of that element (size_t size) and some context (void *app_key).
/* Dump the buffer out to the specified FILE */
static int write_out(const void *buffer, size_t size, void *key) {
FILE *fp = (FILE *)key;
return (fwrite(buffer, 1, size, fp) == size) ? 0 : -1;
}In this example, we can see that the user is using der_encode() which accepts a FILE * as the last parameter, which later is passed to write_out() as the context.
mouse07410
vlm
Expected behavior
Similar to DER encoder, I'd expect the call to
provide me with the exact size of the buffer required to UPER-encode the given PDU, which I would later convert to bytes like
size_t bytes_needed = (ret.encoded + 7)/8;Actually happens
causes SEGV (application crashes with segmentation fault, trying to encode octet string).
If PER encoder is not supposed to behave like DER encoder in this aspect - what's the correct way of determining the size of the buffer to hold the UPER-encoded PDU?
I'm aware of
uper_encode_to_new_buffer(), and it works fine. But it is documented that this function uses "complete" encoding, and therefore may take more space than necessary.What's the correct way of using
uper_encode_to_buffer()? How do I know how big a buffer to pre-allocate?And as we're at it, asn1c-usage.pdf mentions _writestream() callback, but never actually shows how to use/write it. What's the example? What are the parameters
(const void *buffer, size_t size, app_key)? Just the number of bytes in the given buffer that represent the encoded value?