Cross Memory Attach

The basic idea behind cross memory attach is to allow MPI programs doing
intra-node communication to do a single copy of the message rather than a
double copy of the message via shared memory.

The following patch attempts to achieve this by allowing a destination
process, given an address and size from a source process, to copy memory
directly from the source process into its own address space via a system
call.  There is also a symmetrical ability to copy from the current
process's address space into a destination process's address space.

- Use of /proc/pid/mem has been considered, but there are issues with
  using it:
  - Does not allow for specifying iovecs for both src and dest, assuming
    preadv or pwritev was implemented either the area read from or
  written to would need to be contiguous.
  - Currently mem_read allows only processes who are currently
  ptrace'ing the target and are still able to ptrace the target to read
  from the target. This check could possibly be moved to the open call,
  but its not clear exactly what race this restriction is stopping
  (reason  appears to have been lost)
  - Having to send the fd of /proc/self/mem via SCM_RIGHTS on unix
  domain socket is a bit ugly from a userspace point of view,
  especially when you may have hundreds if not (eventually) thousands
  of processes  that all need to do this with each other
  - Doesn't allow for some future use of the interface we would like to
  consider adding in the future (see below)
  - Interestingly reading from /proc/pid/mem currently actually
  involves two copies! (But this could be fixed pretty easily)

As mentioned previously use of vmsplice instead was considered, but has
problems.  Since you need the reader and writer working co-operatively if
the pipe is not drained then you block.  Which requires some wrapping to
do non blocking on the send side or polling on the receive.  In all to all
communication it requires ordering otherwise you can deadlock.  And in the
example of many MPI tasks writing to one MPI task vmsplice serialises the
copying.

There are some cases of MPI collectives where even a single copy interface
does not get us the performance gain we could.  For example in an
MPI_Reduce rather than copy the data from the source we would like to
instead use it directly in a mathops (say the reduce is doing a sum) as
this would save us doing a copy.  We don't need to keep a copy of the data
from the source.  I haven't implemented this, but I think this interface
could in the future do all this through the use of the flags - eg could
specify the math operation and type and the kernel rather than just
copying the data would apply the specified operation between the source
and destination and store it in the destination.

Although we don't have a "second user" of the interface (though I've had
some nibbles from people who may be interested in using it for intra
process messaging which is not MPI).  This interface is something which
hardware vendors are already doing for their custom drivers to implement
fast local communication.  And so in addition to this being useful for
OpenMPI it would mean the driver maintainers don't have to fix things up
when the mm changes.

There was some discussion about how much faster a true zero copy would
go. Here's a link back to the email with some testing I did on that:

http://marc.info/?l=linux-mm&m=130105930902915&w=2

There is a basic man page for the proposed interface here:

http://ozlabs.org/~cyeoh/cma/process_vm_readv.txt

This has been implemented for x86 and powerpc, other architecture should
mainly (I think) just need to add syscall numbers for the process_vm_readv
and process_vm_writev. There are 32 bit compatibility versions for
64-bit kernels.

For arch maintainers there are some simple tests to be able to quickly
verify that the syscalls are working correctly here:

http://ozlabs.org/~cyeoh/cma/cma-test-20110718.tgz

Signed-off-by: Chris Yeoh <yeohc@au1.ibm.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: David Howells <dhowells@redhat.com>
Cc: James Morris <jmorris@namei.org>
Cc: <linux-man@vger.kernel.org>
Cc: <linux-arch@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

3 files changed, 11 insertions, 8 deletions

diff --git a/fs/aio.c b/fs/aio.c
index e29ec485af25..632b235f4fbe 100644
--- a/fs/aio.c
+++ b/fs/aio.c
@@ -1387,13 +1387,13 @@ static ssize_t aio_setup_vectored_rw(int type, struct kiocb *kiocb, bool compat)
 		ret = compat_rw_copy_check_uvector(type,
 				(struct compat_iovec __user *)kiocb->ki_buf,
 				kiocb->ki_nbytes, 1, &kiocb->ki_inline_vec,
-				&kiocb->ki_iovec);
+				&kiocb->ki_iovec, 1);
 	else
 #endif
 		ret = rw_copy_check_uvector(type,
 				(struct iovec __user *)kiocb->ki_buf,
 				kiocb->ki_nbytes, 1, &kiocb->ki_inline_vec,
-				&kiocb->ki_iovec);
+				&kiocb->ki_iovec, 1);
 	if (ret < 0)
 		goto out;
 
diff --git a/fs/compat.c b/fs/compat.c
index 302e761bd0aa..c98787536bb8 100644
--- a/fs/compat.c
+++ b/fs/compat.c
@@ -546,7 +546,7 @@ out:
 ssize_t compat_rw_copy_check_uvector(int type,
 		const struct compat_iovec __user *uvector, unsigned long nr_segs,
 		unsigned long fast_segs, struct iovec *fast_pointer,
-		struct iovec **ret_pointer)
+		struct iovec **ret_pointer, int check_access)
 {
 	compat_ssize_t tot_len;
 	struct iovec *iov = *ret_pointer = fast_pointer;
@@ -593,7 +593,8 @@ ssize_t compat_rw_copy_check_uvector(int type,
 		}
 		if (len < 0)	/* size_t not fitting in compat_ssize_t .. */
 			goto out;
-		if (!access_ok(vrfy_dir(type), compat_ptr(buf), len)) {
+		if (check_access &&
+		    !access_ok(vrfy_dir(type), compat_ptr(buf), len)) {
 			ret = -EFAULT;
 			goto out;
 		}
@@ -1107,7 +1108,7 @@ static ssize_t compat_do_readv_writev(int type, struct file *file,
 		goto out;
 
 	tot_len = compat_rw_copy_check_uvector(type, uvector, nr_segs,
-					       UIO_FASTIOV, iovstack, &iov);
+					       UIO_FASTIOV, iovstack, &iov, 1);
 	if (tot_len == 0) {
 		ret = 0;
 		goto out;
diff --git a/fs/read_write.c b/fs/read_write.c
index dfd125798791..5ad4248b0cd8 100644
--- a/fs/read_write.c
+++ b/fs/read_write.c
@@ -633,7 +633,8 @@ ssize_t do_loop_readv_writev(struct file *filp, struct iovec *iov,
 ssize_t rw_copy_check_uvector(int type, const struct iovec __user * uvector,
 			      unsigned long nr_segs, unsigned long fast_segs,
 			      struct iovec *fast_pointer,
-			      struct iovec **ret_pointer)
+			      struct iovec **ret_pointer,
+			      int check_access)
 {
 	unsigned long seg;
 	ssize_t ret;
@@ -689,7 +690,8 @@ ssize_t rw_copy_check_uvector(int type, const struct iovec __user * uvector,
 			ret = -EINVAL;
 			goto out;
 		}
-		if (unlikely(!access_ok(vrfy_dir(type), buf, len))) {
+		if (check_access
+		    && unlikely(!access_ok(vrfy_dir(type), buf, len))) {
 			ret = -EFAULT;
 			goto out;
 		}
@@ -721,7 +723,7 @@ static ssize_t do_readv_writev(int type, struct file *file,
 	}
 
 	ret = rw_copy_check_uvector(type, uvector, nr_segs,
-			ARRAY_SIZE(iovstack), iovstack, &iov);
+				    ARRAY_SIZE(iovstack), iovstack, &iov, 1);
 	if (ret <= 0)
 		goto out;