From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Laight Subject: RE: [RESEND PATCH v2] devres: Really align data field to unsigned long long Date: Mon, 9 Jul 2018 09:16:38 +0000 Message-ID: References: <20180709044444.6397-1-abrodkin@synopsys.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 8BIT Return-path: In-Reply-To: <20180709044444.6397-1-abrodkin@synopsys.com> Content-Language: en-US Sender: linux-kernel-owner@vger.kernel.org To: 'Alexey Brodkin' , "linux-kernel@vger.kernel.org" Cc: "linux-snps-arc@lists.infradead.org" , "linux-arch@vger.kernel.org" , Greg KH , Greg Kroah-Hartman , Thomas Gleixner , "stable@vger.kernel.org" List-Id: linux-arch.vger.kernel.org From: Alexey Brodkin > Sent: 09 July 2018 05:45 > Depending on ABI "long long" type of a particular 32-bit CPU > might be aligned by either word (32-bits) or double word (64-bits). > Make sure "data" is really 64-bit aligned for any 32-bit CPU. > > At least for 32-bit ARC cores ABI requires "long long" types > to be aligned by normal 32-bit word. This makes "data" field aligned to > 12 bytes. Which is still OK as long as we use 32-bit data only. > > But once we want to use native atomic64_t type (i.e. when we use special > instructions LLOCKD/SCONDD for accessing 64-bit data) we easily hit > misaligned access exception. Shouldn't there be a typedef for the actual type. Perhaps it is even atomic64_t ? And have the __aligned(8) applied to that typedef ?? > That's because even on CPUs capable of non-aligned data access LL/SC > instructions require strict alignment. ... > --- a/drivers/base/devres.c > +++ b/drivers/base/devres.c > @@ -24,8 +24,12 @@ struct devres_node { > > struct devres { > struct devres_node node; > - /* -- 3 pointers */ > - unsigned long long data[]; /* guarantee ull alignment */ > + /* > + * Depending on ABI "long long" type of a particular 32-bit CPU > + * might be aligned by either word (32-bits) or double word (64-bits). > + * Make sure "data" is really 64-bit aligned for any 32-bit CPU. Just: /* data[] must be 64 bit aligned even on 32 bit architectures * because it might be accessed by instructions that require * aligned memory arguments. > + */ > + unsigned long long data[] __aligned(sizeof(unsigned long long)); One day assuming that 'unsigned long long' is exactly 64 bits will bite. This probably ought to be u64 (or similar). (If not atomic64_t) David From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from smtp-out4.electric.net ([192.162.216.195]:59097 "EHLO smtp-out4.electric.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754142AbeGIJPG (ORCPT ); Mon, 9 Jul 2018 05:15:06 -0400 From: David Laight Subject: RE: [RESEND PATCH v2] devres: Really align data field to unsigned long long Date: Mon, 9 Jul 2018 09:16:38 +0000 Message-ID: References: <20180709044444.6397-1-abrodkin@synopsys.com> In-Reply-To: <20180709044444.6397-1-abrodkin@synopsys.com> Content-Language: en-US Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 8BIT MIME-Version: 1.0 Sender: linux-arch-owner@vger.kernel.org List-ID: To: 'Alexey Brodkin' , "linux-kernel@vger.kernel.org" Cc: "linux-snps-arc@lists.infradead.org" , "linux-arch@vger.kernel.org" , Greg KH , Greg Kroah-Hartman , Thomas Gleixner , "stable@vger.kernel.org" Message-ID: <20180709091638.PNcoErjwFw3sXKDCXV-SzISL82wiaPDAQmqismJebeg@z> From: Alexey Brodkin > Sent: 09 July 2018 05:45 > Depending on ABI "long long" type of a particular 32-bit CPU > might be aligned by either word (32-bits) or double word (64-bits). > Make sure "data" is really 64-bit aligned for any 32-bit CPU. > > At least for 32-bit ARC cores ABI requires "long long" types > to be aligned by normal 32-bit word. This makes "data" field aligned to > 12 bytes. Which is still OK as long as we use 32-bit data only. > > But once we want to use native atomic64_t type (i.e. when we use special > instructions LLOCKD/SCONDD for accessing 64-bit data) we easily hit > misaligned access exception. Shouldn't there be a typedef for the actual type. Perhaps it is even atomic64_t ? And have the __aligned(8) applied to that typedef ?? > That's because even on CPUs capable of non-aligned data access LL/SC > instructions require strict alignment. ... > --- a/drivers/base/devres.c > +++ b/drivers/base/devres.c > @@ -24,8 +24,12 @@ struct devres_node { > > struct devres { > struct devres_node node; > - /* -- 3 pointers */ > - unsigned long long data[]; /* guarantee ull alignment */ > + /* > + * Depending on ABI "long long" type of a particular 32-bit CPU > + * might be aligned by either word (32-bits) or double word (64-bits). > + * Make sure "data" is really 64-bit aligned for any 32-bit CPU. Just: /* data[] must be 64 bit aligned even on 32 bit architectures * because it might be accessed by instructions that require * aligned memory arguments. > + */ > + unsigned long long data[] __aligned(sizeof(unsigned long long)); One day assuming that 'unsigned long long' is exactly 64 bits will bite. This probably ought to be u64 (or similar). (If not atomic64_t) David