From mboxrd@z Thu Jan 1 00:00:00 1970 From: =?iso-8859-1?q?Markus_Klotzb=FCcher?= Date: Wed, 31 May 2006 14:11:30 +0200 Subject: [U-Boot-Users] USB OHCI drivers unification In-Reply-To: <20060531103446.GA4390@enneenne.com> (Rodolfo Giometti's message of "Wed, 31 May 2006 12:34:47 +0200") References: <20060530140405.GB21734@enneenne.com> <20060530141700.8F726353450@atlas.denx.de> <20060530145421.GU21995@enneenne.com> <8764jmfuu6.fsf@denx.de> <20060531092135.GB3630@enneenne.com> <87slmqcwrl.fsf@denx.de> <20060531103446.GA4390@enneenne.com> Message-ID: <878xoics25.fsf@denx.de> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: u-boot@lists.denx.de Rodolfo Giometti writes: > On Wed, May 31, 2006 at 12:29:50PM +0200, Markus Klotzb?cher wrote: >> I understand that the virt_to_phys() are required, but why do you need >> the ohci_cpu_to_le16 macros? They seem to be the same as the m16_swap, >> m32_swap macros? > > They are useful only for better reading the code since if I see > m16_swap() I may think that the variable _must_ be swapped in any > case, but if I read ohci_cpu_to_le16() I well understand that the > variable _may_ be swapped according to CPU endianess. Well, then we should probably use the existing macros (include/asm/byteorder.h), but the downside is that this will grow the diffs between the remaining ohci drivers and the generic one which will make merging them more work. I'd prefer to leave cosmetic stuff for now, until the other drivers are merged. Regards Markus Klotzbuecher