Return-Path: <linux-kernel-owner@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
	id S1753978Ab3FJPhr (ORCPT <rfc822;w@1wt.eu>);
	Mon, 10 Jun 2013 11:37:47 -0400
Received: from e06smtp11.uk.ibm.com ([195.75.94.107]:59430 "EHLO
	e06smtp11.uk.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
	with ESMTP id S1752486Ab3FJPhq (ORCPT
	<rfc822;linux-kernel@vger.kernel.org>);
	Mon, 10 Jun 2013 11:37:46 -0400
Date: Mon, 10 Jun 2013 17:37:39 +0200
From: Michael Holzheu <holzheu@linux.vnet.ibm.com>
To: HATAYAMA Daisuke <d.hatayama@gmail.com>
Cc: Vivek Goyal <vgoyal@redhat.com>, Jan Willeke <willeke@de.ibm.com>,
        Martin Schwidefsky <schwidefsky@de.ibm.com>,
        Heiko Carstens <heiko.carstens@de.ibm.com>,
        linux-kernel@vger.kernel.org, kexec@lists.infradead.org
Subject: Re: [PATCH v5 3/5] vmcore: Introduce remap_oldmem_pfn_range()
Message-ID: <20130610173739.4d88d4ec@holzheu>
In-Reply-To: <CABOkKT0yk6tytN=aTSdpECQWw_30m5qrZYH-W1msovkeiSeZCA@mail.gmail.com>
References: <1370624161-2298-1-git-send-email-holzheu@linux.vnet.ibm.com>
	<1370624161-2298-4-git-send-email-holzheu@linux.vnet.ibm.com>
	<CABOkKT0yk6tytN=aTSdpECQWw_30m5qrZYH-W1msovkeiSeZCA@mail.gmail.com>
Organization: IBM
X-Mailer: Claws Mail 3.8.0 (GTK+ 2.24.10; i686-pc-linux-gnu)
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
X-TM-AS-MML: No
X-Content-Scanned: Fidelis XPS MAILER
x-cbid: 13061015-5024-0000-0000-0000064927AE
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org
Content-Length: 3260
Lines: 94

On Mon, 10 Jun 2013 22:40:24 +0900
HATAYAMA Daisuke <d.hatayama@gmail.com> wrote:

> 2013/6/8 Michael Holzheu <holzheu@linux.vnet.ibm.com>:
> <cut>
> > @@ -225,6 +251,56 @@ static ssize_t read_vmcore(struct file *file, char __user *buffer,
> >         return acc;
> >  }
> >
> > +static ssize_t read_vmcore(struct file *file, char __user *buffer,
> > +                          size_t buflen, loff_t *fpos)
> > +{
> > +       return __read_vmcore(buffer, buflen, fpos, 1);
> > +}
> > +
> > +/*
> > + * The vmcore fault handler uses the page cache and fills data using the
> > + * standard __vmcore_read() function.
> > + */
> > +static int mmap_vmcore_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
> > +{
> > +       struct address_space *mapping = vma->vm_private_data;
> > +       pgoff_t index = vmf->pgoff;
> > +       struct page *page;
> > +       loff_t src;
> > +       char *buf;
> > +       int rc;
> > +
> > +find_page:
> > +       page = find_lock_page(mapping, index);
> > +       if (page) {
> > +               unlock_page(page);
> > +               rc = VM_FAULT_MINOR;
> > +       } else {
> > +               page = page_cache_alloc_cold(mapping);
> > +               if (!page)
> > +                       return VM_FAULT_OOM;
> > +               rc = add_to_page_cache_lru(page, mapping, index, GFP_KERNEL);
> > +               if (rc) {
> > +                       page_cache_release(page);
> > +                       if (rc == -EEXIST)
> > +                               goto find_page;
> > +                       /* Probably ENOMEM for radix tree node */
> > +                       return VM_FAULT_OOM;
> > +               }
> > +               buf = (void *) (page_to_pfn(page) << PAGE_SHIFT);
> > +               src = index << PAGE_CACHE_SHIFT;
> > +               __read_vmcore(buf, PAGE_SIZE, &src, 0);
> > +               unlock_page(page);
> > +               rc = VM_FAULT_MAJOR;
> > +       }
> > +       vmf->page = page;
> > +       return rc;
> > +}
> 
> How about reusing find_or_create_page()?

The function would then look like the following:

static int mmap_vmcore_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
{
        struct address_space *mapping = vma->vm_private_data;
        pgoff_t index = vmf->pgoff;
        struct page *page;
        loff_t src;
        char *buf;

        page = find_or_create_page(mapping, index, GFP_KERNEL);
        if (!page)
                return VM_FAULT_OOM;
        src = index << PAGE_CACHE_SHIFT;
        buf = (void *) (page_to_pfn(page) << PAGE_SHIFT);
        __read_vmcore(buf, PAGE_SIZE, &src, 0);
        unlock_page(page);
        vmf->page = page;
        return 0;
}

I agree that this makes the function simpler but we have to copy
the page also if it has already been filled, correct?

But since normally only one process uses /proc/vmcore this might be
acceptable.

BTW: I also removed the VM_FAULT_MAJOR/MINOR because I think the fault
handler should return 0 if a page has been found.

Michael

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/