public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Andrew Morton <akpm@zip.com.au>
To: Linus Torvalds <torvalds@transmeta.com>
Cc: lkml <linux-kernel@vger.kernel.org>
Subject: Re: [patch 2/13] remove pages from the LRU in __free_pages_ok()
Date: Sun, 28 Jul 2002 17:32:56 -0700	[thread overview]
Message-ID: <3D448D38.D156C249@zip.com.au> (raw)
In-Reply-To: Pine.LNX.4.44.0207281649560.9427-100000@home.transmeta.com

Linus Torvalds wrote:
> 
> On Sun, 28 Jul 2002, Andrew Morton wrote:
> >
> > There are some situations where a page's final release is performed by
> > put_page().  Such as in access_process_vm().  This tends to go BUG()
> > because the page is on the LRU.
> 
> This is wrong.
> 
> If that happens, then you should just make access_process_vm() use
> "page_cache_release()". That's what you basically make "__free_pages_ok()"
> do, but since you still have to add the BUG_ON() check to make clear that
> it is illegal to do this from an interrupt context, it's much better to
> just do this check statically.
> 

OK.  This means that put_page() against userspace-mapped pages
is to be avoided.

Did an audit.  What on earth is drivers/scsi/sg.c:sg_rb_correct4mmap()
doing?

Also skb_release_data(), ___pskb_trim() and __pskb_pull_tail().  Can these
ever perform the final release against a page which is on the LRU? In
interrupt context?

tcp_sendmsg() is doing put_page() too.  I _think_ it's OK because
the page is mapped into the calling process and because of the
way in which shrink_cache() looks at page->count.  Not sure.



 fs/binfmt_elf.c |    3 ++-
 fs/smbfs/file.c |    8 ++++----
 kernel/ptrace.c |    3 ++-
 3 files changed, 8 insertions(+), 6 deletions(-)

--- 2.5.29/fs/binfmt_elf.c~put_page	Sun Jul 28 17:17:23 2002
+++ 2.5.29-akpm/fs/binfmt_elf.c	Sun Jul 28 17:31:33 2002
@@ -33,6 +33,7 @@
 #include <linux/smp_lock.h>
 #include <linux/compiler.h>
 #include <linux/highmem.h>
+#include <linux/pagemap.h>
 
 #include <asm/uaccess.h>
 #include <asm/param.h>
@@ -1249,7 +1250,7 @@ static int elf_core_dump(long signr, str
 					flush_page_to_ram(page);
 					kunmap(page);
 				}
-				put_page(page);
+				page_cache_release(page);
 			}
 		}
 	}
--- 2.5.29/fs/smbfs/file.c~put_page	Sun Jul 28 17:19:08 2002
+++ 2.5.29-akpm/fs/smbfs/file.c	Sun Jul 28 17:19:40 2002
@@ -105,9 +105,9 @@ smb_readpage(struct file *file, struct p
 	int		error;
 	struct dentry  *dentry = file->f_dentry;
 
-	get_page(page);
+	page_cache_get(page);
 	error = smb_readpage_sync(dentry, page);
-	put_page(page);
+	page_cache_release(page);
 	return error;
 }
 
@@ -194,11 +194,11 @@ smb_writepage(struct page *page)
 	if (page->index >= end_index+1 || !offset)
 		return -EIO;
 do_it:
-	get_page(page);
+	page_cache_get(page);
 	err = smb_writepage_sync(inode, page, 0, offset);
 	SetPageUptodate(page);
 	unlock_page(page);
-	put_page(page);
+	page_cache_release(page);
 	return err;
 }
 
--- 2.5.29/kernel/ptrace.c~put_page	Sun Jul 28 17:20:04 2002
+++ 2.5.29-akpm/kernel/ptrace.c	Sun Jul 28 17:26:13 2002
@@ -11,6 +11,7 @@
 #include <linux/errno.h>
 #include <linux/mm.h>
 #include <linux/highmem.h>
+#include <linux/pagemap.h>
 #include <linux/smp_lock.h>
 
 #include <asm/pgtable.h>
@@ -156,7 +157,7 @@ int access_process_vm(struct task_struct
 			flush_page_to_ram(page);
 		}
 		kunmap(page);
-		put_page(page);
+		page_cache_release(page);
 		len -= bytes;
 		buf += bytes;
 		addr += bytes;

.

  reply	other threads:[~2002-07-29  0:21 UTC|newest]

Thread overview: 29+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2002-07-28  7:32 [patch 2/13] remove pages from the LRU in __free_pages_ok() Andrew Morton
2002-07-28 23:54 ` Linus Torvalds
2002-07-29  0:32   ` Andrew Morton [this message]
2002-07-29  0:37     ` Linus Torvalds
2002-07-29  0:59       ` Andrew Morton
2002-07-29  0:59         ` Rik van Riel
2002-07-29  2:50     ` David S. Miller
2002-07-29  3:51       ` Linus Torvalds
2002-07-29  3:43         ` David S. Miller
2002-07-29  4:21           ` Linus Torvalds
2002-07-29  5:43             ` David S. Miller
2002-07-29  6:16               ` Linus Torvalds
2002-07-29  6:10                 ` David S. Miller
2002-07-29  6:27                   ` Linus Torvalds
2002-07-29  6:59                     ` Andrew Morton
2002-07-30 11:30                     ` Ed Tomlinson
     [not found]                     ` <200208011942.49342.tomlins@cam.org>
     [not found]                       ` <3D49C951.AB7C527E@zip.com.au>
2002-08-03 19:27                         ` [PATCH] slablru for linux-2.5 bk tree Ed Tomlinson
2002-08-03 20:43                           ` Rik van Riel
2002-08-04  3:17                           ` Andrew Morton
2002-07-29  8:35                 ` [patch 2/13] remove pages from the LRU in __free_pages_ok() Rik van Riel
2002-07-29  4:25           ` Andrew Morton
2002-07-29  4:28             ` Linus Torvalds
2002-07-29  4:52               ` Andrew Morton
2002-07-29  4:50                 ` Linus Torvalds
2002-07-29  5:15                   ` Andrew Morton
2002-07-29  4:17         ` Andrew Morton
2002-07-29  4:23           ` Linus Torvalds
2002-07-29  5:43           ` David S. Miller
2002-07-29  6:24           ` Paul Mackerras

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=3D448D38.D156C249@zip.com.au \
    --to=akpm@zip.com.au \
    --cc=linux-kernel@vger.kernel.org \
    --cc=torvalds@transmeta.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox