From mboxrd@z Thu Jan 1 00:00:00 1970 From: Matt Fleming Subject: Re: [PATCH 1/2] efi: Provide a non-blocking SetVariable() operation Date: Wed, 1 Oct 2014 17:11:51 +0100 Message-ID: <20141001161151.GD14343@console-pimps.org> References: <1412165200-32141-1-git-send-email-matt@console-pimps.org> <1412165200-32141-2-git-send-email-matt@console-pimps.org> <20141001152624.GB2843@worktop.programming.kicks-ass.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <20141001152624.GB2843-IIpfhp3q70z/8w/KjCw3T+5/BudmfyzbbVWyRVo5IupeoWH0uzbU5w@public.gmane.org> Sender: linux-efi-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org To: Peter Zijlstra Cc: Ingo Molnar , "H. Peter Anvin" , Thomas Gleixner , linux-efi-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, Ard Biesheuvel , Leif Lindholm , Matthew Garrett , Matt Fleming List-Id: linux-efi@vger.kernel.org On Wed, 01 Oct, at 05:26:24PM, Peter Zijlstra wrote: > > If you want to have this usable from NMI context, you need to convert > efi_runtime_lock to a raw_spinlock_t. Hmm.. I note that none of the spinlocks in the pstore code path that we execute to get here are raw_spinlock_t. And that's the only code path that calls this function. They need to be raw_spinlock_t for -rt? > Also, it would probably be a good idea to have some selftest like thing > that actually calls this from NMI context, right? Yeah probably, although that'd need to be a core pstore test since there's nothing really interesting going on in virt_efi_set_variable_nonblocking(), and it would be good to test it out in the larger pstore context. -- Matt Fleming, Intel Open Source Technology Center