From mboxrd@z Thu Jan 1 00:00:00 1970 From: Arnd Bergmann Subject: Re: [PATCH] arc: use little endian accesses Date: Thu, 10 Mar 2016 20:23:59 +0100 Message-ID: <201603102023.59425.arnd@arndb.de> References: <1457544064-16167-1-git-send-email-ltrimas@synopsys.com> <1457636268.457.34.camel@synopsys.com> Mime-Version: 1.0 Content-Type: Text/Plain; charset="utf-8" Content-Transfer-Encoding: 7bit Return-path: Received: from mout.kundenserver.de ([212.227.126.134]:56323 "EHLO mout.kundenserver.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752217AbcCJTYe (ORCPT ); Thu, 10 Mar 2016 14:24:34 -0500 In-Reply-To: <1457636268.457.34.camel@synopsys.com> Sender: linux-arch-owner@vger.kernel.org List-ID: To: Lada Trimasova Cc: Vineet Gupta , "linux-kernel@vger.kernel.org" , "linux-arch@vger.kernel.org" , "noamc@ezchip.com" , Alexey Brodkin , "linux-snps-arc@lists.infradead.org" Archived-At: List-Archive: List-Post: On Thursday 10 March 2016, Lada Trimasova wrote: > Driver is 8250, kernel is built for BE arc, nsim option in model "nsim_isa_big_endian = 1". > > With current "readl" and "writel" implementation for ARC we read word from memory without any endianess manipulation. So in case of little endian architecture we get what we want: the first memory byte is the low byte in the word. But in case of big endian architecture the word endianess is swapped: the first memory byte is the high word byte. > > And for example, let's see what happens when we use "readl" in function "serial8250_early_in" in driver/tty/serial/8250. > > Take a look to one line from memory dump: > 0xf0000010: 0x0b 0x00 0x00 0x00 0x60 0x00 0x00 0x00 > > When kernel is built for little endian architecture, we read this data in status register in function "serial_putc" using "readl" function in driver/tty/serial/8250 as: > r0: 0x0000006 > The low byte is 0x0b, so the condition "if ((status & BOTH_EMPTY) == BOTH_EMPTY)" is true, as BOTH_EMPTY is some mask with low bytes set. > > When kernel is built with "CPU_BIG_ENDIAN" and model nsim option is "nsim_isa_big_endian=1", we read this data as: > r0: 0x6000000 > So as you can see the low byte is 0x00 and above mentioned condition never becomes true, we can't continue initialization. > Ok, this sounds like a completely normal architecture implementation then, and your patch looks correct. If some other driver breaks because of this change, you should investigate what went wrong there, and treat it as a driver specific problem. Arnd