From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755929Ab3EXNIT (ORCPT ); Fri, 24 May 2013 09:08:19 -0400 Received: from e06smtp16.uk.ibm.com ([195.75.94.112]:58071 "EHLO e06smtp16.uk.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754780Ab3EXNIO (ORCPT ); Fri, 24 May 2013 09:08:14 -0400 From: Michael Holzheu To: Vivek Goyal , HATAYAMA Daisuke Cc: Jan Willeke , Martin Schwidefsky , Heiko Carstens , linux-kernel@vger.kernel.org, kexec@lists.infradead.org Subject: [PATCH 1/2] kdump/mmap: Introduce arch_oldmem_remap_pfn_range() Date: Fri, 24 May 2013 15:08:08 +0200 Message-Id: <1369400889-11064-2-git-send-email-holzheu@linux.vnet.ibm.com> X-Mailer: git-send-email 1.8.1.6 In-Reply-To: <1369400889-11064-1-git-send-email-holzheu@linux.vnet.ibm.com> References: <1369400889-11064-1-git-send-email-holzheu@linux.vnet.ibm.com> X-TM-AS-MML: No X-Content-Scanned: Fidelis XPS MAILER x-cbid: 13052413-3548-0000-0000-0000057CA388 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org From: Jan Willeke Currently the /proc/vmcore mmap patches are not working on s390. The problem is that on s390 the kernel in not relocatable and therefore always runs in the lower memory area. Therefore for kdump on s390 we swap the lower memory area with the crashkernel area before starting the kdump kernel: [0 - OLDMEM_SIZE] is mapped to [OLDMEM_BASE - OLDMEM_BASE + OLDMEM_SIZE] To fix /proc/vmcore mmap memory below OLDMEMSIZE needs to be mapped with OLDMEM_BASE as offset. To achieve that, a new weak function arch_oldmem_remap_pfn_range() is introduced. Signed-off-by: Jan Willeke Signed-off-by: Michael Holzheu --- fs/proc/vmcore.c | 15 ++++++++++++++- include/linux/crash_dump.h | 5 +++++ 2 files changed, 19 insertions(+), 1 deletion(-) diff --git a/fs/proc/vmcore.c b/fs/proc/vmcore.c index 80221d7..3eda0ac 100644 --- a/fs/proc/vmcore.c +++ b/fs/proc/vmcore.c @@ -123,6 +123,19 @@ static ssize_t read_from_oldmem(char *buf, size_t count, return read; } +/* + * Architetures may override this function to map oldmen + */ +int __weak arch_oldmem_remap_pfn_range(struct vm_area_struct *vma, + unsigned long from, + unsigned long pfn, + unsigned long size, + pgprot_t prot) +{ + return remap_pfn_range(vma, from, pfn, size, prot); +} + + /* Read from the ELF header and then the crash dump. On error, negative value is * returned otherwise number of bytes read are returned. */ @@ -267,7 +280,7 @@ static int mmap_vmcore(struct file *file, struct vm_area_struct *vma) if (size < tsz) tsz = size; paddr = m->paddr + start - m->offset; - if (remap_pfn_range(vma, vma->vm_start + len, + if (arch_oldmem_remap_pfn_range(vma, vma->vm_start + len, paddr >> PAGE_SHIFT, tsz, vma->vm_page_prot)) { do_munmap(vma->vm_mm, vma->vm_start, len); diff --git a/include/linux/crash_dump.h b/include/linux/crash_dump.h index 37e4f8d..da300a7 100644 --- a/include/linux/crash_dump.h +++ b/include/linux/crash_dump.h @@ -14,6 +14,11 @@ extern unsigned long long elfcorehdr_size; extern ssize_t copy_oldmem_page(unsigned long, char *, size_t, unsigned long, int); +extern int __weak arch_oldmem_remap_pfn_range(struct vm_area_struct *vma, + unsigned long from, + unsigned long pfn, + unsigned long size, + pgprot_t prot); /* Architecture code defines this if there are other possible ELF * machine types, e.g. on bi-arch capable hardware. */ -- 1.8.1.6 From mboxrd@z Thu Jan 1 00:00:00 1970 Return-path: Received: from e06smtp14.uk.ibm.com ([195.75.94.110]) by merlin.infradead.org with esmtps (Exim 4.80.1 #2 (Red Hat Linux)) id 1Ufrjc-00074N-IZ for kexec@lists.infradead.org; Fri, 24 May 2013 13:08:41 +0000 Received: from /spool/local by e06smtp14.uk.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Fri, 24 May 2013 14:02:01 +0100 Received: from b06cxnps3075.portsmouth.uk.ibm.com (d06relay10.portsmouth.uk.ibm.com [9.149.109.195]) by d06dlp01.portsmouth.uk.ibm.com (Postfix) with ESMTP id AB73A17D8025 for ; Fri, 24 May 2013 14:09:22 +0100 (BST) Received: from d06av03.portsmouth.uk.ibm.com (d06av03.portsmouth.uk.ibm.com [9.149.37.213]) by b06cxnps3075.portsmouth.uk.ibm.com (8.13.8/8.13.8/NCO v10.0) with ESMTP id r4OD80pi54853816 for ; Fri, 24 May 2013 13:08:00 GMT Received: from d06av03.portsmouth.uk.ibm.com (localhost [127.0.0.1]) by d06av03.portsmouth.uk.ibm.com (8.14.4/8.14.4/NCO v10.0 AVout) with ESMTP id r4OD8A1g022872 for ; Fri, 24 May 2013 07:08:10 -0600 From: Michael Holzheu Subject: [PATCH 1/2] kdump/mmap: Introduce arch_oldmem_remap_pfn_range() Date: Fri, 24 May 2013 15:08:08 +0200 Message-Id: <1369400889-11064-2-git-send-email-holzheu@linux.vnet.ibm.com> In-Reply-To: <1369400889-11064-1-git-send-email-holzheu@linux.vnet.ibm.com> References: <1369400889-11064-1-git-send-email-holzheu@linux.vnet.ibm.com> List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Sender: "kexec" Errors-To: kexec-bounces+dwmw2=twosheds.infradead.org@lists.infradead.org To: Vivek Goyal , HATAYAMA Daisuke Cc: Martin Schwidefsky , kexec@lists.infradead.org, Heiko Carstens , Jan Willeke , linux-kernel@vger.kernel.org From: Jan Willeke Currently the /proc/vmcore mmap patches are not working on s390. The problem is that on s390 the kernel in not relocatable and therefore always runs in the lower memory area. Therefore for kdump on s390 we swap the lower memory area with the crashkernel area before starting the kdump kernel: [0 - OLDMEM_SIZE] is mapped to [OLDMEM_BASE - OLDMEM_BASE + OLDMEM_SIZE] To fix /proc/vmcore mmap memory below OLDMEMSIZE needs to be mapped with OLDMEM_BASE as offset. To achieve that, a new weak function arch_oldmem_remap_pfn_range() is introduced. Signed-off-by: Jan Willeke Signed-off-by: Michael Holzheu --- fs/proc/vmcore.c | 15 ++++++++++++++- include/linux/crash_dump.h | 5 +++++ 2 files changed, 19 insertions(+), 1 deletion(-) diff --git a/fs/proc/vmcore.c b/fs/proc/vmcore.c index 80221d7..3eda0ac 100644 --- a/fs/proc/vmcore.c +++ b/fs/proc/vmcore.c @@ -123,6 +123,19 @@ static ssize_t read_from_oldmem(char *buf, size_t count, return read; } +/* + * Architetures may override this function to map oldmen + */ +int __weak arch_oldmem_remap_pfn_range(struct vm_area_struct *vma, + unsigned long from, + unsigned long pfn, + unsigned long size, + pgprot_t prot) +{ + return remap_pfn_range(vma, from, pfn, size, prot); +} + + /* Read from the ELF header and then the crash dump. On error, negative value is * returned otherwise number of bytes read are returned. */ @@ -267,7 +280,7 @@ static int mmap_vmcore(struct file *file, struct vm_area_struct *vma) if (size < tsz) tsz = size; paddr = m->paddr + start - m->offset; - if (remap_pfn_range(vma, vma->vm_start + len, + if (arch_oldmem_remap_pfn_range(vma, vma->vm_start + len, paddr >> PAGE_SHIFT, tsz, vma->vm_page_prot)) { do_munmap(vma->vm_mm, vma->vm_start, len); diff --git a/include/linux/crash_dump.h b/include/linux/crash_dump.h index 37e4f8d..da300a7 100644 --- a/include/linux/crash_dump.h +++ b/include/linux/crash_dump.h @@ -14,6 +14,11 @@ extern unsigned long long elfcorehdr_size; extern ssize_t copy_oldmem_page(unsigned long, char *, size_t, unsigned long, int); +extern int __weak arch_oldmem_remap_pfn_range(struct vm_area_struct *vma, + unsigned long from, + unsigned long pfn, + unsigned long size, + pgprot_t prot); /* Architecture code defines this if there are other possible ELF * machine types, e.g. on bi-arch capable hardware. */ -- 1.8.1.6 _______________________________________________ kexec mailing list kexec@lists.infradead.org http://lists.infradead.org/mailman/listinfo/kexec