stable.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	stable@vger.kernel.org, Dwight Engen <dwight.engen@oracle.com>,
	"David S. Miller" <davem@davemloft.net>
Subject: [PATCH 3.10 08/70] vio: fix reuse of vio_dring slot
Date: Wed, 19 Nov 2014 12:51:59 -0800	[thread overview]
Message-ID: <20141119205211.194978535@linuxfoundation.org> (raw)
In-Reply-To: <20141119205210.913169042@linuxfoundation.org>

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

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

From: Dwight Engen <dwight.engen@oracle.com>

[ Upstream commit d0aedcd4f14a22e23b313f42b7e6e6ebfc0fbc31 ]

vio_dring_avail() will allow use of every dring entry, but when the last
entry is allocated then dr->prod == dr->cons which is indistinguishable from
the ring empty condition. This causes the next allocation to reuse an entry.
When this happens in sunvdc, the server side vds driver begins nack'ing the
messages and ends up resetting the ldc channel. This problem does not effect
sunvnet since it checks for < 2.

The fix here is to just never allocate the very last dring slot so that full
and empty are not the same condition. The request start path was changed to
check for the ring being full a bit earlier, and to stop the blk_queue if
there is no space left. The blk_queue will be restarted once the ring is
only half full again. The number of ring entries was increased to 512 which
matches the sunvnet and Solaris vdc drivers, and greatly reduces the
frequency of hitting the ring full condition and the associated blk_queue
stop/starting. The checks in sunvent were adjusted to account for
vio_dring_avail() returning 1 less.

Orabug: 19441666
OraBZ: 14983

Signed-off-by: Dwight Engen <dwight.engen@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 arch/sparc/include/asm/vio.h       |    2 -
 drivers/block/sunvdc.c             |   39 +++++++++++++++++++++----------------
 drivers/net/ethernet/sun/sunvnet.c |    4 +--
 3 files changed, 26 insertions(+), 19 deletions(-)

--- a/arch/sparc/include/asm/vio.h
+++ b/arch/sparc/include/asm/vio.h
@@ -265,7 +265,7 @@ static inline u32 vio_dring_avail(struct
 				  unsigned int ring_size)
 {
 	return (dr->pending -
-		((dr->prod - dr->cons) & (ring_size - 1)));
+		((dr->prod - dr->cons) & (ring_size - 1)) - 1);
 }
 
 #define VIO_MAX_TYPE_LEN	32
--- a/drivers/block/sunvdc.c
+++ b/drivers/block/sunvdc.c
@@ -33,7 +33,7 @@ MODULE_DESCRIPTION("Sun LDOM virtual dis
 MODULE_LICENSE("GPL");
 MODULE_VERSION(DRV_MODULE_VERSION);
 
-#define VDC_TX_RING_SIZE	256
+#define VDC_TX_RING_SIZE	512
 
 #define WAITING_FOR_LINK_UP	0x01
 #define WAITING_FOR_TX_SPACE	0x02
@@ -283,7 +283,9 @@ static void vdc_end_one(struct vdc_port
 
 	__blk_end_request(req, (desc->status ? -EIO : 0), desc->size);
 
-	if (blk_queue_stopped(port->disk->queue))
+	/* restart blk queue when ring is half emptied */
+	if (blk_queue_stopped(port->disk->queue) &&
+	    vdc_tx_dring_avail(dr) * 100 / VDC_TX_RING_SIZE >= 50)
 		blk_start_queue(port->disk->queue);
 }
 
@@ -435,12 +437,6 @@ static int __send_request(struct request
 	for (i = 0; i < nsg; i++)
 		len += sg[i].length;
 
-	if (unlikely(vdc_tx_dring_avail(dr) < 1)) {
-		blk_stop_queue(port->disk->queue);
-		err = -ENOMEM;
-		goto out;
-	}
-
 	desc = vio_dring_cur(dr);
 
 	err = ldc_map_sg(port->vio.lp, sg, nsg,
@@ -480,21 +476,32 @@ static int __send_request(struct request
 		port->req_id++;
 		dr->prod = (dr->prod + 1) & (VDC_TX_RING_SIZE - 1);
 	}
-out:
 
 	return err;
 }
 
-static void do_vdc_request(struct request_queue *q)
+static void do_vdc_request(struct request_queue *rq)
 {
-	while (1) {
-		struct request *req = blk_fetch_request(q);
+	struct request *req;
 
-		if (!req)
+	while ((req = blk_peek_request(rq)) != NULL) {
+		struct vdc_port *port;
+		struct vio_dring_state *dr;
+
+		port = req->rq_disk->private_data;
+		dr = &port->vio.drings[VIO_DRIVER_TX_RING];
+		if (unlikely(vdc_tx_dring_avail(dr) < 1))
+			goto wait;
+
+		blk_start_request(req);
+
+		if (__send_request(req) < 0) {
+			blk_requeue_request(rq, req);
+wait:
+			/* Avoid pointless unplugs. */
+			blk_stop_queue(rq);
 			break;
-
-		if (__send_request(req) < 0)
-			__blk_end_request_all(req, -EIO);
+		}
 	}
 }
 
--- a/drivers/net/ethernet/sun/sunvnet.c
+++ b/drivers/net/ethernet/sun/sunvnet.c
@@ -656,7 +656,7 @@ static int vnet_start_xmit(struct sk_buf
 	spin_lock_irqsave(&port->vio.lock, flags);
 
 	dr = &port->vio.drings[VIO_DRIVER_TX_RING];
-	if (unlikely(vnet_tx_dring_avail(dr) < 2)) {
+	if (unlikely(vnet_tx_dring_avail(dr) < 1)) {
 		if (!netif_queue_stopped(dev)) {
 			netif_stop_queue(dev);
 
@@ -704,7 +704,7 @@ static int vnet_start_xmit(struct sk_buf
 	dev->stats.tx_bytes += skb->len;
 
 	dr->prod = (dr->prod + 1) & (VNET_TX_RING_SIZE - 1);
-	if (unlikely(vnet_tx_dring_avail(dr) < 2)) {
+	if (unlikely(vnet_tx_dring_avail(dr) < 1)) {
 		netif_stop_queue(dev);
 		if (vnet_tx_dring_avail(dr) > VNET_TX_WAKEUP_THRESH(dr))
 			netif_wake_queue(dev);



  parent reply	other threads:[~2014-11-19 20:51 UTC|newest]

Thread overview: 68+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-11-19 20:51 [PATCH 3.10 00/70] 3.10.61-stable review Greg Kroah-Hartman
2014-11-19 20:51 ` [PATCH 3.10 01/70] ip6_tunnel: Use ip6_tnl_dev_init as the ndo_init function Greg Kroah-Hartman
2014-11-19 20:51 ` [PATCH 3.10 02/70] gre6: Move the setting of dev->iflink into the ndo_init functions Greg Kroah-Hartman
2014-11-19 20:51 ` [PATCH 3.10 03/70] net: sctp: fix NULL pointer dereference in af->from_addr_param on malformed packet Greg Kroah-Hartman
2014-11-19 20:51 ` [PATCH 3.10 04/70] net: sctp: fix memory leak in auth key management Greg Kroah-Hartman
2014-11-19 20:51 ` [PATCH 3.10 05/70] sunvdc: add cdrom and v1.1 protocol support Greg Kroah-Hartman
2014-11-19 20:51 ` [PATCH 3.10 06/70] sunvdc: compute vdisk geometry from capacity Greg Kroah-Hartman
2014-11-19 20:51 ` [PATCH 3.10 07/70] sunvdc: limit each sg segment to a page Greg Kroah-Hartman
2014-11-19 20:51 ` Greg Kroah-Hartman [this message]
2014-11-19 20:52 ` [PATCH 3.10 09/70] sunvdc: dont call VD_OP_GET_VTOC Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 10/70] sparc64: Fix crashes in schizo_pcierr_intr_other() Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 11/70] sparc64: Do irq_{enter,exit}() around generic_smp_call_function*() Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 12/70] sparc32: Implement xchg and atomic_xchg using ATOMIC_HASH locks Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 13/70] x86, x32, audit: Fix x32s AUDIT_ARCH wrt audit Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 14/70] audit: keep inode pinned Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 15/70] ahci: Add Device IDs for Intel Sunrise Point PCH Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 16/70] ahci: disable MSI instead of NCQ on Samsung pci-e SSDs on macbooks Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 17/70] ALSA: usb-audio: Fix memory leak in FTU quirk Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 18/70] xtensa: re-wire umount syscall to sys_oldumount Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 19/70] libceph: do not crash on large auth tickets Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 20/70] iwlwifi: configure the LTR Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 21/70] macvtap: Fix csum_start when VLAN tags are present Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 22/70] mac80211: fix use-after-free in defragmentation Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 23/70] drm/radeon: add missing crtc unlock when setting up the MC Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 26/70] firewire: cdev: prevent kernel stack leaking into ioctl arguments Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 27/70] nfs: fix pnfs direct write memory leak Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 28/70] scsi: only re-lock door after EH on devices that were reset Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 29/70] parisc: Use compat layer for msgctl, shmat, shmctl and semtimedop syscalls Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 30/70] block: Fix computation of merged request priority Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 31/70] dm btree: fix a recursion depth bug in btree walking code Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 32/70] dm raid: ensure superblocks size matches devices logical block size Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 35/70] NFSv4: Ensure that we remove NFSv4.0 delegations when state has expired Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 36/70] NFS: Dont try to reclaim delegation open state if recovery failed Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 37/70] nfs: Fix use of uninitialized variable in nfs_getattr() Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 38/70] NFSv4: Fix races between nfs_remove_bad_delegation() and delegation return Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 39/70] media: ttusb-dec: buffer overflow in ioctl Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 40/70] kgdb: Remove "weak" from kgdb_arch_pc() declaration Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 41/70] clocksource: Remove "weak" from clocksource_default_clock() declaration Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 42/70] ipc: always handle a new value of auto_msgmni Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 43/70] netfilter: nf_log: account for size of NLMSG_DONE attribute Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 44/70] netfilter: nfnetlink_log: fix maximum packet length logged to userspace Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 45/70] netfilter: nf_log: release skbuff on nlmsg put failure Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 46/70] netfilter: xt_bpf: add mising opaque struct sk_filter definition Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 47/70] netfilter: nf_nat: fix oops on netns removal Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 48/70] br: fix use of ->rx_handler_data in code executed on non-rx_handler path Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 49/70] ARM: probes: fix instruction fetch order with <asm/opcodes.h> Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 51/70] MIPS: Fix forgotten preempt_enable() when CPU has inclusive pcaches Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 52/70] perf: Handle compat ioctl Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 53/70] mei: bus: fix possible boundaries violation Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 54/70] perf/x86/intel: Use proper dTLB-load-misses event on IvyBridge Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 55/70] ARM: Correct BUG() assembly to ensure it is endian-agnostic Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 56/70] net/mlx4_en: Fix BlueFlame race Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 57/70] SCSI: hpsa: fix a race in cmd_free/scsi_done Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 58/70] KVM: x86: Dont report guest userspace emulation error to userspace Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 59/70] net: sctp: fix remote memory pressure from excessive queueing Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 60/70] net: sctp: fix panic on duplicate ASCONF chunks Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 61/70] net: sctp: fix skb_over_panic when receiving malformed " Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 62/70] mm: invoke oom-killer from remaining unconverted page fault handlers Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 63/70] arch: mm: remove obsolete init OOM protection Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 64/70] arch: mm: do not invoke OOM killer on kernel fault OOM Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 65/70] arch: mm: pass userspace fault flag to generic fault handler Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 66/70] x86: finish user fault error path with fatal signal Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 67/70] mm: memcg: enable memcg OOM killer only for user faults Greg Kroah-Hartman
2014-11-19 20:52 ` [PATCH 3.10 68/70] mm: memcg: rework and document OOM waiting and wakeup Greg Kroah-Hartman
2014-11-19 20:53 ` [PATCH 3.10 69/70] mm: memcg: do not trap chargers with full callstack on OOM Greg Kroah-Hartman
2014-11-19 20:53 ` [PATCH 3.10 70/70] mm: memcg: handle non-error OOM situations more gracefully Greg Kroah-Hartman
2014-11-20  5:30 ` [PATCH 3.10 00/70] 3.10.61-stable review Guenter Roeck
2014-11-21  1:38 ` Shuah Khan

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=20141119205211.194978535@linuxfoundation.org \
    --to=gregkh@linuxfoundation.org \
    --cc=davem@davemloft.net \
    --cc=dwight.engen@oracle.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).