Re: [PATCH v2 1/2] http: add option to enable 100 Continue responses

[Jonathan Nieder <jrnieder@gmail.com>]

brian m. carlson wrote:

> When using GSS-Negotiate authentication with libcurl, the authentication
> provided will change every time, and so the probe that git uses to determine if
> authentication is needed is not sufficient to guarantee that data can be sent.
> If the data fits entirely in http.postBuffer bytes, the data can be rewound and
> resent if authentication fails; otherwise, a 100 Continue must be requested in
> this case.
>
> By default, curl will send an Expect: 100-continue if a certain amount of data
> is to be uploaded, but when using chunked data this is never triggered.  Add an
> option http.continue, which defaults to disabled, to control whether this header
> is sent.  The addition of an option is necessary because some proxies break
> badly if sent this header.

"By default" means "when allowed to make its own choice", right?  (i.e.,
the behavior git never gave libcurl a chance to try :))

Makes sense.

[...]
> --- a/http.c
> +++ b/http.c
> @@ -11,6 +11,7 @@
>  int active_requests;
>  int http_is_verbose;
>  size_t http_post_buffer = 16 * LARGE_PACKET_MAX;
> +int http_use_100_continue = 0;

Style: git tends to omit the '= 0' implicit for globals, since they are
already 0 by default.

[...]
> --- a/remote-curl.c
> +++ b/remote-curl.c
> @@ -470,7 +470,12 @@ static int post_rpc(struct rpc_state *rpc)
>  
>  	headers = curl_slist_append(headers, rpc->hdr_content_type);
>  	headers = curl_slist_append(headers, rpc->hdr_accept);
> -	headers = curl_slist_append(headers, "Expect:");
> +
> +	/* Force it either on or off, since curl will try to decide based on how
> +	 * much data is to be uploaded and we want consistency.
> +	 */

Style:

	/*
	 * Multi-line comments in git have the starting and ending comment
	 * delimiters on their own lines, like this.
	 */

Will make the fixups mentioned above, squash with documentation, and
apply.  Thanks.