Return-Path: MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Message-ID: <17696.48842.41851.905802@cargo.ozlabs.ibm.com> Date: Mon, 2 Oct 2006 17:24:58 +1000 From: Paul Mackerras To: Kim Phillips Subject: Re: [PATCH 1/15] Update for ucc_geth driver In-Reply-To: <20060930183311.1d50638c.kim.phillips@freescale.com> References: <20060930183311.1d50638c.kim.phillips@freescale.com> Cc: linuxppc-dev@ozlabs.org List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Kim Phillips writes: > if (bcsr_regs == NULL) > - bcsr_regs = (u8 *) ioremap(BCSR_PHYS_ADDR, BCSR_SIZE); > + bcsr_regs = get_bcsr(); > bcsr_regs[14] |= 0x40; Umm, if the BSCR is really so unique to each board, how come this driver is frobbing a specific bit? How does the ucc_geth driver know about the layout of a given board's BCSR? Also, this patch means that the kernel won't build (assuming you have the ucc_geth driver in your config) until we get to patch 9/15, where get_bcsr gets defined. Lastly, shouldn't this patch be going through Jeff Garzik? Paul.