From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from psmtp.com (na3sys010amx185.postini.com [74.125.245.185]) by kanga.kvack.org (Postfix) with SMTP id 8308E6B0070 for ; Wed, 24 Oct 2012 08:18:38 -0400 (EDT) From: Mario Smarduch Subject: questions about free_area_init_core(), SPARSEMEM, DMA Date: Wed, 24 Oct 2012 12:18:34 +0000 Message-ID: <2DDB038789B01B4B80B0D3F1FF7CBDC20645F788@lhreml509-mbb.china.huawei.com> Content-Language: en-US Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Sender: owner-linux-mm@kvack.org List-ID: To: "linux-mm@kvack.org" In frea_area_init_core() for each zone calculates=20 memmap_pages=3DPAGE_ALIGN(size * sizeof(struct page)) >> PAGE_SHIFT; Which is total memory range zone spans and then it subtracts that value fro= m realsize (memory present in zone). I'm bit confused for SPARSEMEM case where sections not occupied don't have memmap arrays (mem_section[] indexed= ) allocated. Should not the calculation of memmap_pages above take that into consideration? Also related to same function I've notice 'dma_reserve' is hardly defined anywhere. The boards I've looked at (ARM PBX, PXA) have their kernel text, data, bss allocated in DMA zone, any reasons why 'dma_reserve' is not defined to correctly determine zone watermarks and other things later on? - Mario -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: email@kvack.org