From mboxrd@z Thu Jan 1 00:00:00 1970 From: Borislav Petkov Subject: Re: [PATCH v2 1/2] mce: acpi/apei: Honour Firmware First for MCA banks listed in APEI HEST CMC Date: Thu, 20 Jun 2013 21:29:28 +0200 Message-ID: <20130620192928.GD19877@pd.tnic> References: <20130619175438.2852.93449.stgit@localhost.localdomain> <20130620073943.GE32694@pd.tnic> <51C3531D.3000600@linux.vnet.ibm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Return-path: Content-Disposition: inline In-Reply-To: <51C3531D.3000600@linux.vnet.ibm.com> Sender: linux-kernel-owner@vger.kernel.org To: "Naveen N. Rao" Cc: tony.luck@intel.com, ananth@in.ibm.com, masbock@linux.vnet.ibm.com, lcm@linux.vnet.ibm.com, linux-kernel@vger.kernel.org, linux-acpi@vger.kernel.org, ying.huang@intel.com List-Id: linux-acpi@vger.kernel.org On Fri, Jun 21, 2013 at 12:38:13AM +0530, Naveen N. Rao wrote: > We need this bitfield to prevent enabling CMCI in future > cmci_discover() invocations. See usage in cmci_discover() further > below. So?! /* Skip banks in firmware first mode */ if (!test_bit(i, __get_cpu_var(mce_poll_banks)) continue; ... > >Yeah, let's call it ...disable_poll_bank because we're disabling polling > >for those banks. And yes, we poll for errors for which no MCE exception > >is generated and those happen to be corrected but still... > > We actually also disable CMCI here. So, in essence, we are disabling > these banks for all sorts of direct corrected error reporting. I > thought of naming this disable_ce_on_bank() or disable_ce_bank(), > but felt that the mce_ prefix would be good to have. If that isn't > necessary, I can rename this to disable_ce_on_bank() which sounds > more accurate to me. Is that ok? No, mce_disable_bank() removes the respective bank from the polling bitfield and cmci_disable_bank() actually disables CMCI which is Intel-only. So leave it at mce_disable_bank and that should be fine. Thanks. -- Regards/Gruss, Boris. Sent from a fat crate under my desk. Formatting is fine. --