mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Luis Henriques <luis.henriques@canonical.com>
To: linux-kernel@vger.kernel.org, stable@vger.kernel.org,
	kernel-team@lists.ubuntu.com
Cc: "Eric W. Biederman" <ebiederm@xmission.com>,
	Stephen Hemminger <stephen@networkplumber.org>,
	Julian Anastasov <ja@ssi.bg>,
	"David S. Miller" <davem@davemloft.net>,
	Luis Henriques <luis.henriques@canonical.com>
Subject: [PATCH 3.16.y-ckt 062/118] net: do not process device backlog during unregistration
Date: Wed, 12 Aug 2015 09:56:04 +0100	[thread overview]
Message-ID: <1439369820-27005-63-git-send-email-luis.henriques@canonical.com> (raw)
In-Reply-To: <1439369820-27005-1-git-send-email-luis.henriques@canonical.com>

3.16.7-ckt16 -stable review patch.  If anyone has any objections, please let me know.

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

From: Julian Anastasov <ja@ssi.bg>

commit e9e4dd3267d0c5234c5c0f47440456b10875dec9 upstream.

commit 381c759d9916 ("ipv4: Avoid crashing in ip_error")
fixes a problem where processed packet comes from device
with destroyed inetdev (dev->ip_ptr). This is not expected
because inetdev_destroy is called in NETDEV_UNREGISTER
phase and packets should not be processed after
dev_close_many() and synchronize_net(). Above fix is still
required because inetdev_destroy can be called for other
reasons. But it shows the real problem: backlog can keep
packets for long time and they do not hold reference to
device. Such packets are then delivered to upper levels
at the same time when device is unregistered.
Calling flush_backlog after NETDEV_UNREGISTER_FINAL still
accounts all packets from backlog but before that some packets
continue to be delivered to upper levels long after the
synchronize_net call which is supposed to wait the last
ones. Also, as Eric pointed out, processed packets, mostly
from other devices, can continue to add new packets to backlog.

Fix the problem by moving flush_backlog early, after the
device driver is stopped and before the synchronize_net() call.
Then use netif_running check to make sure we do not add more
packets to backlog. We have to do it in enqueue_to_backlog
context when the local IRQ is disabled. As result, after the
flush_backlog and synchronize_net sequence all packets
should be accounted.

Thanks to Eric W. Biederman for the test script and his
valuable feedback!

Reported-by: Vittorio Gambaletta <linuxbugs@vittgam.net>
Fixes: 6e583ce5242f ("net: eliminate refcounting in backlog queue")
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Stephen Hemminger <stephen@networkplumber.org>
Signed-off-by: Julian Anastasov <ja@ssi.bg>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Luis Henriques <luis.henriques@canonical.com>
---
 net/core/dev.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/net/core/dev.c b/net/core/dev.c
index 2e616d68a7aa..395830206d73 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -3262,6 +3262,8 @@ static int enqueue_to_backlog(struct sk_buff *skb, int cpu,
 	local_irq_save(flags);
 
 	rps_lock(sd);
+	if (!netif_running(skb->dev))
+		goto drop;
 	qlen = skb_queue_len(&sd->input_pkt_queue);
 	if (qlen <= netdev_max_backlog && !skb_flow_limit(skb, qlen)) {
 		if (skb_queue_len(&sd->input_pkt_queue)) {
@@ -3283,6 +3285,7 @@ enqueue:
 		goto enqueue;
 	}
 
+drop:
 	sd->dropped++;
 	rps_unlock(sd);
 
@@ -5788,6 +5791,7 @@ static void rollback_registered_many(struct list_head *head)
 		unlist_netdevice(dev);
 
 		dev->reg_state = NETREG_UNREGISTERING;
+		on_each_cpu(flush_backlog, dev, 1);
 	}
 
 	synchronize_net();
@@ -6408,8 +6412,6 @@ void netdev_run_todo(void)
 
 		dev->reg_state = NETREG_UNREGISTERED;
 
-		on_each_cpu(flush_backlog, dev, 1);
-
 		netdev_wait_allrefs(dev);
 
 		/* paranoia */

  parent reply	other threads:[~2015-08-12  8:58 UTC|newest]

Thread overview: 119+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-08-12  8:55 [3.16.y-ckt stable] Linux 3.16.7-ckt16 stable review Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 001/118] netfilter: nfnetlink_cthelper: Remove 'const' and '&' to avoid warnings Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 002/118] Bluetooth: ath3k: Add a new ID 0cf3:e006 to ath3k list Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 003/118] Btrfs: use kmem_cache_free when freeing entry in inode cache Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 004/118] Btrfs: fix race between caching kthread and returning inode to " Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 005/118] Btrfs: fix fsync data loss after append write Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 006/118] ext4: fix reservation release on invalidatepage for delalloc fs Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 007/118] ext4: be more strict when migrating to non-extent based file Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 008/118] ext4: correctly migrate a file with a hole at the beginning Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 009/118] ext4: replace open coded nofail allocation in ext4_free_blocks() Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 010/118] drm/radeon: Handle irqs only based on irq ring, not irq status regs Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 011/118] drm/radeon: unpin cursor BOs on suspend and pin them again on resume (v2) Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 012/118] hpfs: kstrdup() out of memory handling Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 013/118] hpfs: hpfs_error: Remove static buffer, use vsprintf extension %pV instead Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 014/118] 9p: don't leave a half-initialized inode sitting around Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 015/118] MIPS: kernel: traps: Fix broken indentation Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 016/118] thermal: step_wise: fix: Prevent from binary overflow when trend is dropping Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 017/118] spi: pl022: Specify 'num-cs' property as required in devicetree binding Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 018/118] iio: twl4030-madc: Pass the IRQF_ONESHOT flag Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 019/118] iio: inv-mpu: Specify the expected format/precision for write channels Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 020/118] iio: DAC: ad5624r_spi: fix bit shift of output data value Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 021/118] iio: adc: at91_adc: allow to use full range of startup time Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 022/118] ALSA: usb-audio: Add MIDI support for Steinberg MI2/MI4 Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 023/118] iio: tmp006: Check channel info on write Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 024/118] dm btree remove: fix bug in redistribute3 Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 025/118] kbuild: Allow arch Makefiles to override {cpp,ld,c}flags Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 026/118] ARC: Override toplevel default -O2 with -O3 Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 027/118] crypto: omap-des - Fix unmapping of dma channels Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 028/118] USB: option: add 2020:4000 ID Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 029/118] USB: cp210x: add ID for Aruba Networks controllers Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 030/118] dm btree: silence lockdep lock inversion in dm_btree_del() Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 031/118] usb: musb: host: rely on port_mode to call musb_start() Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 032/118] usb: f_mass_storage: limit number of reported LUNs Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 033/118] drm: add a check for x/y in drm_mode_setcrtc Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 034/118] bio integrity: do not assume bio_integrity_pool exists if bioset exists Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 035/118] ARM: dts: mx23: fix iio-hwmon support Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 036/118] tracing: Have branch tracer use recursive field of task struct Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 037/118] drivers: net: cpsw: fix crash while accessing second slave ethernet interface Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 038/118] USB: serial: Destroy serial_minors IDR on module exit Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 039/118] Btrfs: fix memory leak in the extent_same ioctl Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 040/118] Btrfs: fix list transaction->pending_ordered corruption Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 041/118] can: rcar_can: fix IRQ check Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 042/118] ARC: make sure instruction_pointer() returns unsigned value Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 043/118] Btrfs: fix file corruption after cloning inline extents Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 044/118] st: null pointer dereference panic caused by use after kref_put by st_open Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 045/118] drm/radeon: add a dpm quirk for Sapphire Radeon R9 270X 2GB GDDR5 Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 046/118] drm/radeon: Don't flush the GART TLB if rdev->gart.ptr == NULL Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 047/118] genirq: Prevent resend to interrupts marked IRQ_NESTED_THREAD Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 048/118] x86/asm/entry/64: Fold the 'test_in_nmi' macro into its only user Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 049/118] x86/asm/entry/64: Remove a redundant jump Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 050/118] x86/nmi: Enable nested do_nmi() handling for 64-bit kernels Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 051/118] x86/nmi/64: Remove asm code that saves CR2 Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 052/118] x86/nmi/64: Switch stacks on userspace NMI entry Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 053/118] x86/nmi/64: Improve nested NMI comments Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 054/118] x86/nmi/64: Reorder nested NMI checks Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 055/118] x86/nmi/64: Use DF to avoid userspace RSP confusing nested NMI detection Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 056/118] ARM: 8404/1: dma-mapping: fix off-by-one error in bitmap size check Luis Henriques
2015-08-12  8:55 ` [PATCH 3.16.y-ckt 057/118] ipv6: Make MLD packets to only be processed locally Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 058/118] bridge: mdb: start delete timer for temp static entries Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 059/118] net: graceful exit from netif_alloc_netdev_queues() Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 060/118] ip_tunnel: fix ipv4 pmtu check to honor inner ip header df Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 061/118] bridge: mdb: zero out the local br_ip variable before use Luis Henriques
2015-08-12  8:56 ` Luis Henriques [this message]
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 063/118] net: dsa: Test array index " Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 064/118] net: dsa: Fix off-by-one in switch address parsing Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 065/118] can: rcar_can: print signed IRQ # Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 066/118] perf symbols: Store if there is a filter in place Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 067/118] perf hists browser: Take the --comm, --dsos, etc filters into account Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 068/118] rds: rds_ib_device.refcount overflow Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 069/118] KEYS: ensure we free the assoc array edit if edit is valid Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 070/118] mm: avoid setting up anonymous pages into file mapping Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 071/118] evm: labeling pseudo filesystems exception Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 072/118] USB: usbfs: allow URBs to be reaped after disconnection Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 073/118] sg_start_req(): make sure that there's not too many elements in iovec Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 074/118] HID: cp2112: fix to force single data-report reply Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 075/118] ata: pmp: add quirk for Marvell 4140 SATA PMP Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 076/118] libata: add ATA_HORKAGE_BROKEN_FPDMA_AA quirk for HP 250GB SATA disk VB0250EAVER Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 077/118] libata: add ATA_HORKAGE_NOTRIM Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 078/118] libata: force disable trim for SuperSSpeed S238 Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 079/118] libata: increase the timeout when setting transfer mode Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 080/118] can: mcp251x: fix resume when device is down Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 081/118] libata: Do not blacklist M510DC Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 082/118] mac80211: clear subdir_stations when removing debugfs Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 083/118] iio: adc: vf610: fix the adc register read fail issue Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 084/118] net: mvneta: fix refilling for Rx DMA buffers Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 085/118] ALSA: hda - Add new GPU codec ID 0x10de007d to snd-hda Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 086/118] xdrm/i915: Use two 32bit reads for select 64bit REG_READ ioctls Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 087/118] usb: dwc3: gadget: return error if command sent to DEPCMD register fails Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 088/118] usb: dwc3: Reset the transfer resource index on SET_INTERFACE Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 089/118] usb: xhci: Bugfix for NULL pointer deference in xhci_endpoint_init() function Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 090/118] xhci: Calculate old endpoints correctly on device reset Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 091/118] xhci: report U3 when link is in resume state Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 092/118] xhci: prevent bus_suspend if SS port resuming in phase 1 Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 093/118] xhci: do not report PLC when link is in internal resume state Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 094/118] usb: core: lpm: set lpm_capable for root hub device Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 095/118] USB: OHCI: Fix race between ED unlink and URB submission Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 096/118] usb-storage: ignore ZTE MF 823 card reader in mode 0x1225 Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 097/118] blkcg: fix gendisk reference leak in blkg_conf_prep() Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 098/118] tile: use free_bootmem_late() for initrd Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 099/118] Input: usbtouchscreen - avoid unresponsive TSC-30 touch screen Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 100/118] block: Do a full clone when splitting discard bios Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 101/118] md/raid1: fix test for 'was read error from last working device' Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 102/118] mmc: omap_hsmmc: Fix DTO and DCRC handling Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 103/118] mtd: nand: Fix NAND_USE_BOUNCE_BUFFER flag conflict Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 104/118] net/xen-netback: off by one in BUG_ON() condition Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 105/118] bridge: mdb: fix double add notification Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 106/118] isdn/gigaset: reset tty->receive_room when attaching ser_gigaset Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 107/118] usb: gadget: mv_udc_core: fix phy_regs I/O memory leak Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 108/118] bonding: fix destruction of bond with devices different from arphrd_ether Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 109/118] bonding: correctly handle bonding type change on enslave failure Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 110/118] inet: frags: fix defragmented packet's IP header for af_packet Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 111/118] mmc: block: Add missing mmc_blk_put() in power_ro_lock_show() Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 112/118] mmc: sdhci-esdhc: Make 8BIT bus work Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 113/118] mmc: sdhci-pxav3: fix platform_data is not initialized Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 114/118] freeing unlinked file indefinitely delayed Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 115/118] s390/sclp: clear upper register halves in _sclp_print_early Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 116/118] s390/process: fix sfpc inline assembly Luis Henriques
2015-08-12  8:56 ` [PATCH 3.16.y-ckt 117/118] mmc: sdhci: Fix FSL ESDHC reset handling quirk Luis Henriques
2015-08-12  8:57 ` [PATCH 3.16.y-ckt 118/118] md: fix a build warning Luis Henriques

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=1439369820-27005-63-git-send-email-luis.henriques@canonical.com \
    --to=luis.henriques@canonical.com \
    --cc=davem@davemloft.net \
    --cc=ebiederm@xmission.com \
    --cc=ja@ssi.bg \
    --cc=kernel-team@lists.ubuntu.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=stable@vger.kernel.org \
    --cc=stephen@networkplumber.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

all inboxes | Powered by JetHome®