If resultant length of a frame exceeds the maximum value (which is
2**24 - 1 for SPDY/2 and 3), SPDYLAY_ERR_FRAME_TOO_LARGE is used to
indicate this error. This error will be notified by
on_ctrl_not_send_callback.
This option sets maximum receive buffer size for incoming control
frame. Basically the library checks the length field of the incoming
control frame. For frames with name/value header block, the library
also checks the length of inflated block is also under the limit. This
is done while incrementally inflating block. If the length of frames
with name/value header block exceeds the limit, the library will issue
RST_STREAM with FRAME_TOO_LARGE. For other frames, it will issue
GOAWAY.
We say the header block is invalid if at least one of the following
condition is true:
There are duplicate header names; or the header names are not
encoded in US-ASCII character set and not lower cased; or the
header name is zero-length string; or the header value contains
multiple in-sequence NUL bytes.
spdylay_frame_unpack_nv() returns SPDYLAY_ERR_INVALID_HEADER_BLOCK
if the unpacking suceeded but it found the header block is invalid.
This means that caller treats it as success, but do additional
processing for invalid header block if it wants.
The functions calling spdylay_frame_unpack_nv() also return
SPDYLAY_ERR_INVALID_HEADER_BLOCK.
spdylay_net.h is compatibility layer for network related header files.
In the nature of spdylay library it should not depend on the actual
networking implementations, but we need some system headers for
optimization. Currently, arpha/inet.h and netinet/in.h are needed for
ntoh*/hton* functions.
If we return negative error code in that function, it means frame will not
unpacked and ends up to lose any information the frame contains.
The spec says it should send RST_STREAM with PROTOCOL_ERROR, so we need
at least stream ID.
Therefore, the check should be performed in
spdylay_session_on_syn_stream_received().
We use just single buffer to store name/value headers fields, instead of
allocating memory for each name/value strings.
It is now more than 2 times faster than old one.
