Re: What's cooking in git.git (Sep 2016, #08; Tue, 27)

From: Junio C Hamano <gitster@pobox.com>
Date: Oct 3, 2016, 20:49 UTC
Message-ID: <xmqqwphp5f9q.fsf@gitster.mtv.corp.google.com>
In-Reply-To: <CAGZ79ka8dO1AHJftKAqD6LvxJSP+8yGGa7Citcdxxrnc5DMeYg@mail.gmail.com>

Stefan Beller <sbeller@google.com> writes:

> I am looking at the tip of jc/attr-more and for a minimum
> thread safety we'd want to change the call sites to be aware of the
> threads, i.e. instead of doing
>

      static struct git_attr_check *check;

Show 12 quoted lines

>     if (!check)
>         check = git_attr_check_initl("crlf", "ident",
>                     "filter", "eol", "text",
>                     NULL);
>
> We'd rather call
>
>         struct git_attr_check *check;
>         check = git_attr_check_lookup_or_initl_threadsafe(
>                 "crlf", "ident", "filter", "eol", "text", NULL);
>          if (!git_check_attr(path, check)) {
>              ...

I actually am hoping that the "static" can be kept so that we can minimize the cost of interning these symbols into struct git_attr.

The initialization would thus become something like:

	static struct git_attr_check *check;
	git_attr_check_initl(&check, "crlf", "ident", ..., NULL);

The structure will have an array of git_attr[], once interned will be shared and used by everybody. _initl() will need to make sure that the "check" pointer is updated atomically so that multiple people racing to initialize this variable will not step on each others' toes.

Then the use site would do something like

	const char *result[... some size ...];
	git_check_attr(path, check, result[]);

to receive result regardless of anybody else who is using the same "check" structure to ask the same question (for different paths).