From: Duy Nguyen Date: Thu, 02 Jun 2016 01:02:48 GMT Subject: Re: Segfault in the attr stack Message-ID: In-Reply-To: On Thu, Jun 2, 2016 at 5:17 AM, Junio C Hamano wrote: > Gaah, of course. > > This is coming from the cache preload codepath, where multiple threads > try to run ce_path_match(). > It used to be OK because pathspec magic never looked at attributes, > but now it does, and attribute system is not thread-safe. Off topic. I wonder if we can annotate which functions are thread-safe (minority currently in git code base, I believe) and make gcc, clang or sparse spot them in threaded code? -- Duy