Date: Tue, 6 Feb 2018 10:20:48 +0200 From: Konstantin Belousov <kostikbel@gmail.com> To: Mori Hiroki <yamori813@yahoo.co.jp> Cc: "freebsd-arm@freebsd.org" <freebsd-arm@freebsd.org> Subject: Re: not close critical section at uma_core.c Message-ID: <20180206082047.GA33564@kib.kiev.ua> In-Reply-To: <458964.25252.qm@web101712.mail.ssk.yahoo.co.jp>
index | next in thread | previous in thread | raw e-mail
On Tue, Feb 06, 2018 at 03:00:23PM +0900, Mori Hiroki wrote: > Hi > > I'm still debugging armv5t. > > I found not good code in vm/uma_core.c. Thisšcritical_enter šhavešnošcritical_exit. > > https://svnweb.freebsd.org/base/head/sys/vm/uma_core.c?revision=328916&view=markup#l2344 > The code path is convoluted, but each critical_enter() in the uma_zalloc_arg() function is accompanied by the goto zalloc_start;, and there the conditions are prepared by an origin site to enter the if(), which does the critical_exit(). Would critical_enter() leaked in so often used function, locking operations started panicing a lot.home | help
Want to link to this message? Use this
URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20180206082047.GA33564>