@@ -156,7 +156,8 @@ static int rdma_read_chunk_lcl(struct svcxprt_rdma *xprt,
ctxt->read_hdr = head;
pages_needed =
min_t(int, pages_needed, rdma_read_max_sge(xprt, pages_needed));
- read = min_t(int, pages_needed << PAGE_SHIFT, rs_length);
+ read = min_t(int, (pages_needed << PAGE_SHIFT) - *page_offset,
+ rs_length);
for (pno = 0; pno < pages_needed; pno++) {
int len = min_t(int, rs_length, PAGE_SIZE - pg_off);
@@ -255,7 +256,8 @@ static int rdma_read_chunk_frmr(struct svcxprt_rdma *xprt,
ctxt->direction = DMA_FROM_DEVICE;
ctxt->frmr = frmr;
pages_needed = min_t(int, pages_needed, xprt->sc_frmr_pg_list_len);
- read = min_t(int, pages_needed << PAGE_SHIFT, rs_length);
+ read = min_t(int, (pages_needed << PAGE_SHIFT) - *page_offset,
+ rs_length);
frmr->kva = page_address(rqstp->rq_arg.pages[pg_no]);
frmr->direction = DMA_FROM_DEVICE;
This is a note to let you know that I have just added a patch titled svcrdma: handle rdma read with a non-zero initial page offset to the linux-3.16.y-queue branch of the 3.16.y-ckt extended stable tree which can be found at: http://kernel.ubuntu.com/git/ubuntu/linux.git/log/?h=linux-3.16.y-queue This patch is scheduled to be released in version 3.16.7-ckt20. If you, or anyone else, feels it should not be added to this tree, please reply to this email. For more information about the 3.16.y-ckt tree, see https://wiki.ubuntu.com/Kernel/Dev/ExtendedStable Thanks. -Luis ------ From 24d9380db4bde66e7dcf32b42608a2ae3f104e3e Mon Sep 17 00:00:00 2001 From: Steve Wise <swise@opengridcomputing.com> Date: Mon, 28 Sep 2015 16:46:06 -0500 Subject: svcrdma: handle rdma read with a non-zero initial page offset commit c91aed9896946721bb30705ea2904edb3725dd61 upstream. The server rdma_read_chunk_lcl() and rdma_read_chunk_frmr() functions were not taking into account the initial page_offset when determining the rdma read length. This resulted in a read who's starting address and length exceeded the base/bounds of the frmr. The server gets an async error from the rdma device and kills the connection, and the client then reconnects and resends. This repeats indefinitely, and the application hangs. Most work loads don't tickle this bug apparently, but one test hit it every time: building the linux kernel on a 16 core node with 'make -j 16 O=/mnt/0' where /mnt/0 is a ramdisk mounted via NFSRDMA. This bug seems to only be tripped with devices having small fastreg page list depths. I didn't see it with mlx4, for instance. Fixes: 0bf4828983df ('svcrdma: refactor marshalling logic') Signed-off-by: Steve Wise <swise@opengridcomputing.com> Tested-by: Chuck Lever <chuck.lever@oracle.com> Signed-off-by: J. Bruce Fields <bfields@redhat.com> [ luis: backported to 3.16: adjusted context ] Signed-off-by: Luis Henriques <luis.henriques@canonical.com> --- net/sunrpc/xprtrdma/svc_rdma_recvfrom.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)