stable.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Ben Hutchings <ben@decadent.org.uk>
To: linux-kernel@vger.kernel.org, stable@vger.kernel.org
Cc: akpm@linux-foundation.org, Ben Hutchings <bhutchings@solarflare.com>
Subject: [ 045/104] sfc: Properly sync RX DMA buffer when it is not the  last in the page
Date: Mon, 25 Mar 2013 01:06:09 +0000	[thread overview]
Message-ID: <20130325010528.719100602@decadent.org.uk> (raw)
In-Reply-To: <20130325010524.240972766@decadent.org.uk>

3.2-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Ben Hutchings <bhutchings@solarflare.com>

[ Upstream commit 3a68f19d7afb80f548d016effbc6ed52643a8085 ]

We may currently allocate two RX DMA buffers to a page, and only unmap
the page when the second is completed.  We do not sync the first RX
buffer to be completed; this can result in packet loss or corruption
if the last RX buffer completed in a NAPI poll is the first in a page
and is not DMA-coherent.  (In the middle of a NAPI poll, we will
handle the following RX completion and unmap the page *before* looking
at the content of the first buffer.)

Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
[bwh: Backported to 3.2: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
---
 drivers/net/ethernet/sfc/rx.c | 15 ++++++++++-----
 1 file changed, 10 insertions(+), 5 deletions(-)

diff --git a/drivers/net/ethernet/sfc/rx.c b/drivers/net/ethernet/sfc/rx.c
index 5ef4cc0..b4d3cd5 100644
--- a/drivers/net/ethernet/sfc/rx.c
+++ b/drivers/net/ethernet/sfc/rx.c
@@ -246,7 +246,8 @@ static int efx_init_rx_buffers_page(struct efx_rx_queue *rx_queue)
 }
 
 static void efx_unmap_rx_buffer(struct efx_nic *efx,
-				struct efx_rx_buffer *rx_buf)
+				struct efx_rx_buffer *rx_buf,
+				unsigned int used_len)
 {
 	if (rx_buf->is_page && rx_buf->u.page) {
 		struct efx_rx_page_state *state;
@@ -257,6 +258,10 @@ static void efx_unmap_rx_buffer(struct efx_nic *efx,
 				       state->dma_addr,
 				       efx_rx_buf_size(efx),
 				       PCI_DMA_FROMDEVICE);
+		} else if (used_len) {
+			dma_sync_single_for_cpu(&efx->pci_dev->dev,
+						rx_buf->dma_addr, used_len,
+						DMA_FROM_DEVICE);
 		}
 	} else if (!rx_buf->is_page && rx_buf->u.skb) {
 		pci_unmap_single(efx->pci_dev, rx_buf->dma_addr,
@@ -279,7 +284,7 @@ static void efx_free_rx_buffer(struct efx_nic *efx,
 static void efx_fini_rx_buffer(struct efx_rx_queue *rx_queue,
 			       struct efx_rx_buffer *rx_buf)
 {
-	efx_unmap_rx_buffer(rx_queue->efx, rx_buf);
+	efx_unmap_rx_buffer(rx_queue->efx, rx_buf, 0);
 	efx_free_rx_buffer(rx_queue->efx, rx_buf);
 }
 
@@ -550,10 +555,10 @@ void efx_rx_packet(struct efx_rx_queue *rx_queue, unsigned int index,
 		goto out;
 	}
 
-	/* Release card resources - assumes all RX buffers consumed in-order
-	 * per RX queue
+	/* Release and/or sync DMA mapping - assumes all RX buffers
+	 * consumed in-order per RX queue
 	 */
-	efx_unmap_rx_buffer(efx, rx_buf);
+	efx_unmap_rx_buffer(efx, rx_buf, len);
 
 	/* Prefetch nice and early so data will (hopefully) be in cache by
 	 * the time we look at it.



  parent reply	other threads:[~2013-03-25  1:06 UTC|newest]

Thread overview: 110+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-03-25  0:34 [ 000/104] 3.2.42-stable review Ben Hutchings
2013-03-25  1:05 ` [ 001/104] TTY: do not reset masters packet mode Ben Hutchings
2013-03-25  1:05 ` [ 002/104] perf,x86: fix kernel crash with PEBS/BTS after suspend/resume Ben Hutchings
2013-03-25  1:05 ` [ 003/104] perf,x86: fix wrmsr_on_cpu() warning on suspend/resume Ben Hutchings
2013-03-25  1:05 ` [ 004/104] perf,x86: fix link failure for non-Intel configs Ben Hutchings
2013-03-25  1:05 ` [ 005/104] l2tp: Restore socket refcount when sendmsg succeeds Ben Hutchings
2013-03-25  1:05 ` [ 006/104] rds: limit the size allocated by rds_message_alloc() Ben Hutchings
2013-03-25  1:05 ` [ 007/104] net: ipv6: Dont purge default router if accept_ra=2 Ben Hutchings
2013-03-25  1:05 ` [ 008/104] tcp: fix double-counted receiver RTT when leaving receiver fast path Ben Hutchings
2013-03-25  1:05 ` [ 009/104] tun: add a missing nf_reset() in tun_net_xmit() Ben Hutchings
2013-03-25  1:05 ` [ 010/104] macvlan: Set IFF_UNICAST_FLT flag to prevent unnecessary promisc mode Ben Hutchings
2013-03-25  1:05 ` [ 011/104] netlabel: correctly list all the static label mappings Ben Hutchings
2013-03-25  1:05 ` [ 012/104] bridging: fix rx_handlers return code Ben Hutchings
2013-03-25  1:05 ` [ 013/104] ipv6: stop multicast forwarding to process interface scoped addresses Ben Hutchings
2013-03-25  1:05 ` [ 014/104] rtnl: fix info leak on RTM_GETLINK request for VF devices Ben Hutchings
2013-03-25  1:05 ` [ 015/104] dcbnl: fix various netlink info leaks Ben Hutchings
2013-03-25  1:05 ` [ 016/104] 6lowpan: Fix endianness issue in is_addr_link_local() Ben Hutchings
2013-03-25  1:05 ` [ 017/104] drm/i915: Increase the RC6p threshold Ben Hutchings
2013-03-25  1:05 ` [ 018/104] perf: Revert duplicated commit Ben Hutchings
2013-03-25  1:05 ` [ 019/104] i915: initialize CADL in opregion Ben Hutchings
2013-03-25  1:05 ` [ 020/104] s390/mm: fix flush_tlb_kernel_range() Ben Hutchings
2013-03-25  1:05 ` [ 021/104] mwifiex: fix potential out-of-boundary access to ibss rate table Ben Hutchings
2013-03-25  1:05 ` [ 022/104] rtlwifi: rtl8192cu: Fix schedule while atomic bug splat Ben Hutchings
2013-03-25  1:05 ` [ 023/104] powerpc: Fix cputable entry for 970MP rev 1.0 Ben Hutchings
2013-03-25  1:05 ` [ 024/104] rtlwifi: rtl8192cu: Fix problem that prevents reassociation Ben Hutchings
2013-03-25  1:05 ` [ 025/104] vhost/net: fix heads usage of ubuf_info Ben Hutchings
2013-03-25  1:05 ` [ 026/104] selinux: use GFP_ATOMIC under spin_lock Ben Hutchings
2013-03-25  1:05 ` [ 027/104] tools: hv: Netlink source address validation allows DoS Ben Hutchings
2013-03-25  1:05 ` [ 028/104] udf: avoid info leak on export Ben Hutchings
2013-03-25  1:05 ` [ 029/104] isofs: " Ben Hutchings
2013-03-25  1:05 ` [ 030/104] sunsu: Fix panic in case of nonexistent port at "console=ttySY" cmdline option Ben Hutchings
2013-03-25  1:05 ` [ 031/104] net/ipv4: Ensure that location of timestamp option is stored Ben Hutchings
2013-03-25  1:05 ` [ 032/104] netconsole: dont call __netpoll_cleanup() while atomic Ben Hutchings
2013-03-25  1:05 ` [ 033/104] bonding: dont call update_speed_duplex() under spinlocks Ben Hutchings
2013-03-25  1:05 ` [ 034/104] sctp: Use correct sideffect command in duplicate cookie handling Ben Hutchings
2013-03-25  1:05 ` [ 035/104] sctp: dont break the loop while meeting the active_path so as to find the matched transport Ben Hutchings
2013-03-25  1:06 ` [ 036/104] ipv4: fix definition of FIB_TABLE_HASHSZ Ben Hutchings
2013-03-25  1:06 ` [ 037/104] tcp: fix skb_availroom() Ben Hutchings
2013-03-25  1:06 ` [ 038/104] rtnetlink: Mask the rta_type when range checking Ben Hutchings
2013-03-25  1:06 ` [ 039/104] inet: limit length of fragment queue hash table bucket lists Ben Hutchings
2013-03-25  1:06 ` [ 040/104] sfc: Do not attempt to flush queues if DMA is disabled Ben Hutchings
2013-03-25  1:06 ` [ 041/104] sfc: Convert firmware subtypes to native byte order in efx_mcdi_get_board_cfg() Ben Hutchings
2013-03-25  1:06 ` [ 042/104] sfc: Fix two causes of flush failure Ben Hutchings
2013-03-25  1:06 ` [ 043/104] sfc: lock TX queues when calling netif_device_detach() Ben Hutchings
2013-03-25  1:06 ` [ 044/104] sfc: Fix timekeeping in efx_mcdi_poll() Ben Hutchings
2013-03-25  1:06 ` Ben Hutchings [this message]
2013-03-25  1:06 ` [ 046/104] sfc: Fix efx_rx_buf_offset() in the presence of swiotlb Ben Hutchings
2013-03-25  1:06 ` [ 047/104] sfc: Detach net device when stopping queues for reconfiguration Ben Hutchings
2013-03-25  1:06 ` [ 048/104] sfc: Disable soft interrupt handling during efx_device_detach_sync() Ben Hutchings
2013-03-25  1:06 ` [ 049/104] sfc: Only use TX push if a single descriptor is to be written Ben Hutchings
2013-03-25  1:06 ` [ 050/104] ext4: fix the wrong number of the allocated blocks in Ben Hutchings
2013-03-25  1:06 ` [ 051/104] jbd2: fix use after free in jbd2_journal_dirty_metadata() Ben Hutchings
2013-03-25  1:06 ` [ 052/104] ext4: convert number of blocks to clusters properly Ben Hutchings
2013-03-25  1:06 ` [ 053/104] ext4: use atomic64_t for the per-flexbg free_clusters count Ben Hutchings
2013-03-25  1:06 ` [ 054/104] tracing: Fix race in snapshot swapping Ben Hutchings
2013-03-25  1:06 ` [ 055/104] cifs: delay super block destruction until all cifsFileInfo objects Ben Hutchings
2013-03-25  1:06 ` [ 056/104] drm/i915: restrict kernel address leak in debugfs Ben Hutchings
2013-03-25  1:06 ` [ 057/104] drm/i915: bounds check execbuffer relocation count Ben Hutchings
2013-03-25  1:06 ` [ 058/104] tracing: Fix free of probe entry by calling call_rcu_sched() Ben Hutchings
2013-03-25  1:06 ` [ 059/104] tracing: Protect tracer flags with trace_types_lock Ben Hutchings
2013-03-25  1:06 ` [ 060/104] tracing: Keep overwrite in sync between regular and snapshot buffers Ben Hutchings
2013-03-25  1:06 ` [ 061/104] tracing: Prevent buffer overwrite disabled for latency tracers Ben Hutchings
2013-03-25  1:06 ` [ 062/104] USB: xhci: correctly enable interrupts Ben Hutchings
2013-03-25  1:06 ` [ 063/104] usb-storage: add unusual_devs entry for Samsung YP-Z3 mp3 player Ben Hutchings
2013-03-25  1:06 ` [ 064/104] drm/radeon/benchmark: make sure bo blit copy exists before using it Ben Hutchings
2013-03-25  1:06 ` [ 065/104] ALSA: hda/cirrus - Fix the digital beep registration Ben Hutchings
2013-03-25  1:06 ` [ 066/104] USB: xhci - fix bit definitions for IMAN register Ben Hutchings
2013-03-25  1:06 ` [ 067/104] x86-64: Fix the failure case in copy_user_handle_tail() Ben Hutchings
2013-03-25  1:06 ` [ 068/104] ALSA: snd-usb: mixer: propagate errors up the call chain Ben Hutchings
2013-03-25  1:06 ` [ 069/104] ALSA: snd-usb: mixer: ignore -EINVAL in snd_usb_mixer_controls() Ben Hutchings
2013-03-25  1:06 ` [ 070/104] ext4: fix data=journal fast mount/umount hang Ben Hutchings
2013-03-25  1:06 ` [ 071/104] ALSA: hda - Fix typo in checking IEC958 emphasis bit Ben Hutchings
2013-03-25  1:06 ` [ 072/104] usb: gadget: udc-core: fix a regression during gadget driver Ben Hutchings
2013-03-25  1:06 ` [ 073/104] dm thin: fix discard corruption Ben Hutchings
2013-03-25  1:06 ` [ 074/104] efivars: Allow disabling use as a pstore backend Ben Hutchings
2013-03-25  1:06 ` [ 075/104] efivars: Add module parameter to disable " Ben Hutchings
2013-03-25  1:06 ` [ 076/104] efivars: Fix check for CONFIG_EFI_VARS_PSTORE_DEFAULT_DISABLE Ben Hutchings
2013-03-25  1:06 ` [ 077/104] efi_pstore: Introducing workqueue updating sysfs Ben Hutchings
2013-03-25  1:06 ` [ 078/104] efivars: explicitly calculate length of VariableName Ben Hutchings
2013-03-25  1:06 ` [ 079/104] efivars: Handle duplicate names from get_next_variable() Ben Hutchings
2013-03-25  1:06 ` [ 080/104] cifs: ignore everything in SPNEGO blob after mechTypes Ben Hutchings
2013-03-25  1:06 ` [ 081/104] USB: garmin_gps: fix memory leak on disconnect Ben Hutchings
2013-03-25  1:06 ` [ 082/104] USB: io_ti: fix get_icount for two port adapters Ben Hutchings
2013-03-25  1:06 ` [ 083/104] USB: serial: fix interface refcounting Ben Hutchings
2013-03-25  1:06 ` [ 084/104] USB: serial: add modem-status-change wait queue Ben Hutchings
2013-03-26 15:36   ` Johan Hovold
2013-03-27  2:35     ` Ben Hutchings
2013-03-25  1:06 ` [ 085/104] USB: ark3116: fix use-after-free in TIOCMIWAIT Ben Hutchings
2013-03-25  1:06 ` [ 086/104] USB: ch341: " Ben Hutchings
2013-03-25  1:06 ` [ 087/104] USB: cypress_m8: " Ben Hutchings
2013-03-25  1:06 ` [ 088/104] USB: ftdi_sio: " Ben Hutchings
2013-03-25  1:06 ` [ 089/104] USB: io_edgeport: " Ben Hutchings
2013-03-25  1:06 ` [ 090/104] USB: io_ti: " Ben Hutchings
2013-03-25  1:06 ` [ 091/104] USB: mct_u232: " Ben Hutchings
2013-03-25  1:06 ` [ 092/104] USB: mos7840: fix broken TIOCMIWAIT Ben Hutchings
2013-03-25  1:06 ` [ 093/104] USB: mos7840: fix use-after-free in TIOCMIWAIT Ben Hutchings
2013-03-25  1:06 ` [ 094/104] USB: oti6858: " Ben Hutchings
2013-03-25  1:06 ` [ 095/104] USB: pl2303: " Ben Hutchings
2013-03-25  1:07 ` [ 096/104] USB: spcp8x5: " Ben Hutchings
2013-03-25  1:07 ` [ 097/104] USB: ssu100: " Ben Hutchings
2013-03-25  1:07 ` [ 098/104] USB: ti_usb_3410_5052: " Ben Hutchings
2013-03-25  1:07 ` [ 099/104] i2c: tegra: check the clk_prepare_enable() return value Ben Hutchings
2013-03-25  1:07 ` [ 100/104] vfs,proc: guarantee unique inodes in /proc Ben Hutchings
2013-03-25  1:07 ` [ 101/104] mm/hugetlb: fix total hugetlbfs pages count when using memory Ben Hutchings
2013-03-25  1:07 ` [ 102/104] KMS: fix EDID detailed timing vsync parsing Ben Hutchings
2013-03-25  1:07 ` [ 103/104] KMS: fix EDID detailed timing frame rate Ben Hutchings
2013-03-25  1:07 ` [ 104/104] efivars: pstore: Do not check size when erasing variable Ben Hutchings
2013-03-25  2:06 ` [ 000/104] 3.2.42-stable review Ben Hutchings
2013-03-26 13:01 ` Satoru Takeuchi
2013-03-26 19:03   ` Ben Hutchings

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=20130325010528.719100602@decadent.org.uk \
    --to=ben@decadent.org.uk \
    --cc=akpm@linux-foundation.org \
    --cc=bhutchings@solarflare.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=stable@vger.kernel.org \
    /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;
as well as URLs for NNTP newsgroup(s).