From mboxrd@z Thu Jan 1 00:00:00 1970 From: eric.y.miao@gmail.com (Eric Miao) Date: Sat, 2 Apr 2011 14:48:54 +0800 Subject: [PATCH 2/9] ARM: pxa: enable sparsemem in saarb In-Reply-To: <20110401183325.GB8482@n2100.arm.linux.org.uk> References: <1301625568-16583-1-git-send-email-haojian.zhuang@marvell.com> <1301625568-16583-2-git-send-email-haojian.zhuang@marvell.com> <25B60CDC2F704E4E9D88FFD52780CB4C05CEEB84A9@SC-VEXCH1.marvell.com> <20110401183325.GB8482@n2100.arm.linux.org.uk> Message-ID: To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On Sat, Apr 2, 2011 at 2:33 AM, Russell King - ARM Linux wrote: > On Fri, Apr 01, 2011 at 01:10:52PM +0800, Eric Miao wrote: >> On Fri, Apr 1, 2011 at 12:35 PM, Haojian Zhuang wrote: >> > If I use HIGHMEM now, our some drivers and apps should be changed also. We don't want to change them. >> >> I don't understand this will cause side effects to other drivers and apps? > > If they're not using kmap_atomic() but are dealing directly with > page addresses, they'll fail if the pages are coming from highmem. > > Given Linus' complaints in the OMAP pull request thread, I think we > should refuse, and have the drivers fixed to work with highmem. > That's what I thought as well. Haojian, Please work on enabling highmem, anything that's going to prevent us from having a single kernel binary, will just make the situation worse. Thanks