LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: "Sébastien Dugué" <sebastien.dugue@bull.net>
To: linux-kernel <linux-kernel@vger.kernel.org>
Cc: Andrew Morton <akpm@osdl.org>, linux-aio <linux-aio@kvack.org>,
	Bharata B Rao <bharata@in.ibm.com>,
	Christoph Hellwig <hch@infradead.org>,
	Suparna Bhattacharya <suparna@in.ibm.com>,
	Ulrich Drepper <drepper@redhat.com>,
	Zach Brown <zach.brown@oracle.com>,
	Oleg Nesterov <oleg@tv-sign.ru>,
	Badari Pulavarty <pbadari@us.ibm.com>,
	Benjamin LaHaise <bcrl@linux.intel.com>,
	Jean Pierre Dion <jean-pierre.dion@bull.net>
Subject: [PATCH -mm 1/7][AIO] - Rework compat_sys_io_submit
Date: Thu, 1 Feb 2007 10:26:40 +0100	[thread overview]
Message-ID: <20070201102640.2bdd6f37@frecb000686> (raw)
In-Reply-To: <20070201102252.240130c6@frecb000686>


From: Sébastien Dugué <sebastien.dugue@bull.net>

                 compat_sys_io_submit() cleanup


  Cleanup compat_sys_io_submit by duplicating some of the native syscall
logic in the compat layer and directly calling io_submit_one() instead
of fooling the syscall into thinking it is called from a native 64-bit
caller.

  This eliminates:

  - the overhead of copying the nr iocb pointers on the userspace stack

  - the PAGE_SIZE/(sizeof(void *)) limit on the number of iocbs that
    can be submitted.

  This is also needed for the completion notification patch to avoid having
to rewrite each iocb on the caller stack for io_submit_one() to find the
sigevents.


 compat.c |   61 ++++++++++++++++++++++++++++++++++---------------------------
 1 file changed, 34 insertions(+), 27 deletions(-)

Signed-off-by: Sébastien Dugué <sebastien.dugue@bull.net>

Index: linux-2.6.20-rc6-mm3/fs/compat.c
===================================================================
--- linux-2.6.20-rc6-mm3.orig/fs/compat.c	2007-01-30 10:05:08.000000000 +0100
+++ linux-2.6.20-rc6-mm3/fs/compat.c	2007-01-30 11:30:29.000000000 +0100
@@ -644,40 +644,47 @@ out:
 	return ret;
 }
 
-static inline long
-copy_iocb(long nr, u32 __user *ptr32, struct iocb __user * __user *ptr64)
-{
-	compat_uptr_t uptr;
-	int i;
-
-	for (i = 0; i < nr; ++i) {
-		if (get_user(uptr, ptr32 + i))
-			return -EFAULT;
-		if (put_user(compat_ptr(uptr), ptr64 + i))
-			return -EFAULT;
-	}
-	return 0;
-}
-
-#define MAX_AIO_SUBMITS 	(PAGE_SIZE/sizeof(struct iocb *))
-
 asmlinkage long
 compat_sys_io_submit(aio_context_t ctx_id, int nr, u32 __user *iocb)
 {
-	struct iocb __user * __user *iocb64; 
-	long ret;
+	struct kioctx *ctx;
+	long ret = 0;
+	int i;
 
 	if (unlikely(nr < 0))
 		return -EINVAL;
 
-	if (nr > MAX_AIO_SUBMITS)
-		nr = MAX_AIO_SUBMITS;
-	
-	iocb64 = compat_alloc_user_space(nr * sizeof(*iocb64));
-	ret = copy_iocb(nr, iocb, iocb64);
-	if (!ret)
-		ret = sys_io_submit(ctx_id, nr, iocb64);
-	return ret;
+	if (unlikely(!access_ok(VERIFY_READ, iocb, (nr * sizeof(u32)))))
+		return -EFAULT;
+
+	ctx = lookup_ioctx(ctx_id);
+	if (unlikely(!ctx))
+		return -EINVAL;
+
+	for (i=0; i<nr; i++) {
+		compat_uptr_t uptr;
+		struct iocb __user *user_iocb;
+		struct iocb tmp;
+
+		if (unlikely(get_user(uptr, iocb + i))) {
+			ret = -EFAULT;
+			break;
+		}
+
+		user_iocb = compat_ptr(uptr);
+
+		if (unlikely(copy_from_user(&tmp, user_iocb, sizeof(tmp)))) {
+			ret = -EFAULT;
+			break;
+		}
+
+		ret = io_submit_one(ctx, user_iocb, &tmp);
+		if (ret)
+			break;
+	}
+
+	put_ioctx(ctx);
+	return i ? i: ret;
 }
 
 struct compat_ncp_mount_data {

  reply	other threads:[~2007-02-01  9:39 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-02-01  9:22 [PATCH -mm 0/7][AIO] - AIO completion signal notification v6 Sébastien Dugué
2007-02-01  9:26 ` Sébastien Dugué [this message]
2007-02-01  9:29 ` [PATCH -mm 2/7][AIO] - fix aio.h includes Sébastien Dugué
2007-02-01  9:30 ` [PATCH -mm 3/7][AIO] - Fix access_ok() checks Sébastien Dugué
2007-02-01  9:30 ` [PATCH -mm 4/7][AIO] - Make good_sigevent() non-static Sébastien Dugué
2007-02-02 18:00   ` Oleg Nesterov
2007-02-05 11:09     ` Sébastien Dugué
2007-02-05 12:18     ` [PATCH -mm 4/7][AIO] Resend " Sébastien Dugué
2007-02-05 13:43       ` Oleg Nesterov
2007-02-05 16:00         ` [PATCH -mm][AIO] Fix AIO completion signal notification possible ref leak Sébastien Dugué
2007-02-05 17:13           ` Oleg Nesterov
2007-02-06  8:31             ` Sébastien Dugué
2007-02-06  9:22             ` [PATCH -mm][AIO] AIO completion signal notification fixes and cleanups Sébastien Dugué
2007-02-06 11:05               ` Oleg Nesterov
2007-02-06 11:39                 ` Sébastien Dugué
2007-02-06 11:57                 ` [PATCH -mm][AIO] AIO completion signal notification small cleanup Sébastien Dugué
2007-02-01  9:31 ` [PATCH -mm 5/7][AIO] - Make __sigqueue_free() and __sigqueue_alloc() non static Sébastien Dugué
2007-02-01  9:31 ` [PATCH -mm 6/7][AIO] - AIO completion signal notification Sébastien Dugué
2007-02-01  9:32 ` [PATCH -mm 7/7][AIO] - Add listio syscall support Sébastien Dugué

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=20070201102640.2bdd6f37@frecb000686 \
    --to=sebastien.dugue@bull.net \
    --cc=akpm@osdl.org \
    --cc=bcrl@linux.intel.com \
    --cc=bharata@in.ibm.com \
    --cc=drepper@redhat.com \
    --cc=hch@infradead.org \
    --cc=jean-pierre.dion@bull.net \
    --cc=linux-aio@kvack.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=oleg@tv-sign.ru \
    --cc=pbadari@us.ibm.com \
    --cc=suparna@in.ibm.com \
    --cc=zach.brown@oracle.com \
    --subject='Re: [PATCH -mm 1/7][AIO] - Rework compat_sys_io_submit' \
    /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

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).