* linux-next: manual merge of the net tree with the unicore32 tree
From: Stephen Rothwell @ 2011-08-23 3:02 UTC (permalink / raw)
To: David Miller, netdev
Cc: linux-next, linux-kernel, Guan Xuetao, Zhang Shu, Chu Xiaowei,
Su Yonggang, Jeff Kirsher
Hi all,
Today's linux-next merge of the net tree got conflicts in
drivers/net/Kconfig and drivers/net/Makefile between commit e8787de6fa83
("unicore32: add pkunity-v3 mac/net driver (umal)") from the unicore32
tree and the network driver rearrangement from the net tree.
I just added the new driver from the unicore32 tree commit into each file
(see below).
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig
index ef6b6be..df5e990 100644
--- a/drivers/net/Kconfig
+++ b/drivers/net/Kconfig
@@ -215,6 +215,12 @@ source "drivers/s390/net/Kconfig"
source "drivers/net/caif/Kconfig"
+config MAC_PUV3
+ tristate "PKUnity v3 UMAL Gigabit Network Adapter support"
+ depends on UNICORE32 && ARCH_PUV3
+ select MII
+ select PHYLIB
+
config XEN_NETDEV_FRONTEND
tristate "Xen network device frontend driver"
depends on XEN
diff --git a/drivers/net/Makefile b/drivers/net/Makefile
index c33009b..9896ad1 100644
--- a/drivers/net/Makefile
+++ b/drivers/net/Makefile
@@ -23,6 +23,7 @@ obj-$(CONFIG_RIONET) += rionet.o
obj-$(CONFIG_NET) += Space.o loopback.o
obj-$(CONFIG_NET_SB1000) += sb1000.o
+obj-$(CONFIG_MAC_PUV3) += mac-puv3.o
obj-$(CONFIG_PPP) += ppp_generic.o
obj-$(CONFIG_PPP_ASYNC) += ppp_async.o
obj-$(CONFIG_PPP_SYNC_TTY) += ppp_synctty.o
^ permalink raw reply related
* (3.1.0-rc2-git7) include/linux/inetdevice.h:209 invoked rcu_dereference_check() without protection!
From: Dave Jones @ 2011-08-23 2:44 UTC (permalink / raw)
To: netdev
Just hit this..
Is the fix as straight forward as changing that __in_dev_get_rcu to
an in_dev_get() ?
Dave
===================================================
[ INFO: suspicious rcu_dereference_check() usage. ]
---------------------------------------------------
include/linux/inetdevice.h:209 invoked rcu_dereference_check() without protection!
other info that might help us debug this:
rcu_scheduler_active = 1, debug_locks = 0
4 locks held by setfiles/2123:
#0: (&sb->s_type->i_mutex_key#13){+.+.+.}, at: [<ffffffff8114cbc4>] walk_component+0x1ef/0x3e8
#1: (&isec->lock){+.+.+.}, at: [<ffffffff81204bca>] inode_doinit_with_dentry+0x3f/0x41f
#2: (&tbl->proxy_timer){+.-...}, at: [<ffffffff8106a803>] run_timer_softirq+0x157/0x372
#3: (class){+.-...}, at: [<ffffffff8141f256>] neigh_proxy_process+0x36/0x103
stack backtrace:
Pid: 2123, comm: setfiles Tainted: G W 3.1.0-0.rc2.git7.2.fc16.x86_64 #1
Call Trace:
<IRQ> [<ffffffff8108ca23>] lockdep_rcu_dereference+0xa7/0xaf
[<ffffffff8146a0b7>] __in_dev_get_rcu+0x55/0x5d
[<ffffffff8146a751>] arp_process+0x25/0x4d7
[<ffffffff8146ac11>] parp_redo+0xe/0x10
[<ffffffff8141f2ba>] neigh_proxy_process+0x9a/0x103
[<ffffffff8106a8c4>] run_timer_softirq+0x218/0x372
[<ffffffff8106a803>] ? run_timer_softirq+0x157/0x372
[<ffffffff8141f220>] ? neigh_stat_seq_open+0x41/0x41
[<ffffffff8108f2f0>] ? mark_held_locks+0x6d/0x95
[<ffffffff81062bb6>] __do_softirq+0x112/0x25a
[<ffffffff8150d27c>] call_softirq+0x1c/0x30
[<ffffffff81010bf5>] do_softirq+0x4b/0xa2
[<ffffffff81062f65>] irq_exit+0x5d/0xcf
[<ffffffff8150dc11>] smp_apic_timer_interrupt+0x7c/0x8a
[<ffffffff8150baf3>] apic_timer_interrupt+0x73/0x80
<EOI> [<ffffffff8108f439>] ? trace_hardirqs_on_caller+0x121/0x158
[<ffffffff814fc285>] ? __slab_free+0x30/0x24c
[<ffffffff814fc283>] ? __slab_free+0x2e/0x24c
[<ffffffff81204e74>] ? inode_doinit_with_dentry+0x2e9/0x41f
[<ffffffff81204e74>] ? inode_doinit_with_dentry+0x2e9/0x41f
[<ffffffff81204e74>] ? inode_doinit_with_dentry+0x2e9/0x41f
[<ffffffff81130cb0>] kfree+0x108/0x131
[<ffffffff81204e74>] inode_doinit_with_dentry+0x2e9/0x41f
[<ffffffff81204fc6>] selinux_d_instantiate+0x1c/0x1e
[<ffffffff81200f4f>] security_d_instantiate+0x21/0x23
[<ffffffff81154625>] d_instantiate+0x5c/0x61
[<ffffffff811563ca>] d_splice_alias+0xbc/0xd2
[<ffffffff811b17ff>] ext4_lookup+0xba/0xeb
[<ffffffff8114bf1e>] d_alloc_and_lookup+0x45/0x6b
[<ffffffff8114cbea>] walk_component+0x215/0x3e8
[<ffffffff8114cdf8>] lookup_last+0x3b/0x3d
[<ffffffff8114daf3>] path_lookupat+0x82/0x2af
[<ffffffff8110fc53>] ? might_fault+0xa5/0xac
[<ffffffff8110fc0a>] ? might_fault+0x5c/0xac
[<ffffffff8114c564>] ? getname_flags+0x31/0x1ca
[<ffffffff8114dd48>] do_path_lookup+0x28/0x97
[<ffffffff8114df2c>] user_path_at+0x59/0x96
[<ffffffff811467ad>] ? cp_new_stat+0xf7/0x10d
[<ffffffff811469a6>] vfs_fstatat+0x44/0x6e
[<ffffffff811469ee>] vfs_lstat+0x1e/0x20
[<ffffffff81146b3d>] sys_newlstat+0x1a/0x33
[<ffffffff8108f439>] ? trace_hardirqs_on_caller+0x121/0x158
[<ffffffff812535fe>] ? trace_hardirqs_on_thunk+0x3a/0x3f
[<ffffffff8150af82>] system_call_fastpath+0x16/0x1b
^ permalink raw reply
* Re: linux-next: boot test failure (net tree)
From: Jeff Kirsher @ 2011-08-23 2:26 UTC (permalink / raw)
To: David Miller
Cc: sfr@canb.auug.org.au, netdev@vger.kernel.org,
linux-next@vger.kernel.org, linux-kernel@vger.kernel.org,
mikey@neuling.org, torvalds@linux-foundation.org,
akpm@linux-foundation.org, linuxppc-dev@lists.ozlabs.org,
benh@kernel.crashing.org, paulus@samba.org
In-Reply-To: <20110822.191348.2099822249437201579.davem@davemloft.net>
[-- Attachment #1: Type: text/plain, Size: 1586 bytes --]
On Mon, 2011-08-22 at 19:13 -0700, David Miller wrote:
> From: Stephen Rothwell <sfr@canb.auug.org.au>
> Date: Tue, 23 Aug 2011 11:41:29 +1000
>
> > On Tue, 23 Aug 2011 11:40:11 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> >>
> >> On Mon, 22 Aug 2011 11:30:32 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> >> >
> >> > Here's what I am applying as a merge fixup to the net tree today so that
> >> > my ppc64_defconfig builds actually build more or less the same set of
> >> > drivers as before this rearrangement.
> >>
> >> And this today:
> >
> > And this:
>
> I'm starting to get uncomfortable with this whole situation, and I
> feel more and more that these new kconfig guards are not tenable.
>
> Changing defconfig files might fix the "automated test boot with
> defconfig" case but it won't fix the case of someone trying to
> automate a build and boot using a different, existing, config file.
> It ought to work too, and I do know people really do this.
>
> And just the fact that we would have to merge all of these defconfig changes
> through the networking tree is evidence of how it's really not reasonable
> to be doing things this way.
>
> Jeff, I think we need to revert the dependencies back to what they were
> before the drivers/net moves. Could you prepare a patch which does that?
>
I was just finishing up those patches (not including any defconfig
changes) and started looking at a patch to fix/resolve the issues that
Stephen is seeing.
Let me see what I can come up with tonight to resolve this.
[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 490 bytes --]
^ permalink raw reply
* Re: linux-next: boot test failure (net tree)
From: David Miller @ 2011-08-23 2:13 UTC (permalink / raw)
To: sfr
Cc: netdev, linux-next, linux-kernel, jeffrey.t.kirsher, mikey,
torvalds, akpm, linuxppc-dev, benh, paulus
In-Reply-To: <20110823114129.ceb18da164bf7df3c145941b@canb.auug.org.au>
From: Stephen Rothwell <sfr@canb.auug.org.au>
Date: Tue, 23 Aug 2011 11:41:29 +1000
> On Tue, 23 Aug 2011 11:40:11 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>>
>> On Mon, 22 Aug 2011 11:30:32 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>> >
>> > Here's what I am applying as a merge fixup to the net tree today so that
>> > my ppc64_defconfig builds actually build more or less the same set of
>> > drivers as before this rearrangement.
>>
>> And this today:
>
> And this:
I'm starting to get uncomfortable with this whole situation, and I
feel more and more that these new kconfig guards are not tenable.
Changing defconfig files might fix the "automated test boot with
defconfig" case but it won't fix the case of someone trying to
automate a build and boot using a different, existing, config file.
It ought to work too, and I do know people really do this.
And just the fact that we would have to merge all of these defconfig changes
through the networking tree is evidence of how it's really not reasonable
to be doing things this way.
Jeff, I think we need to revert the dependencies back to what they were
before the drivers/net moves. Could you prepare a patch which does that?
^ permalink raw reply
* Re: linux-next: boot test failure (net tree)
From: Stephen Rothwell @ 2011-08-23 1:41 UTC (permalink / raw)
To: David Miller
Cc: netdev, linux-next, linux-kernel, jeffrey.t.kirsher, mikey,
torvalds, akpm, ppc-dev, Benjamin Herrenschmidt, Paul Mackerras
In-Reply-To: <20110823114011.a059aea0138b75bfa7eed1ce@canb.auug.org.au>
Hi Dave,
On Tue, 23 Aug 2011 11:40:11 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>
> On Mon, 22 Aug 2011 11:30:32 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> >
> > Here's what I am applying as a merge fixup to the net tree today so that
> > my ppc64_defconfig builds actually build more or less the same set of
> > drivers as before this rearrangement.
>
> And this today:
And this:
From: Stephen Rothwell <sfr@canb.auug.org.au>
Date: Tue, 23 Aug 2011 11:35:18 +1000
Subject: [PATCH] sparc: update sparc64_defconfig for net device movement
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
arch/sparc/configs/sparc64_defconfig | 20 +++++++++-----------
1 files changed, 9 insertions(+), 11 deletions(-)
diff --git a/arch/sparc/configs/sparc64_defconfig b/arch/sparc/configs/sparc64_defconfig
index 3c1e858..5732728 100644
--- a/arch/sparc/configs/sparc64_defconfig
+++ b/arch/sparc/configs/sparc64_defconfig
@@ -37,8 +37,6 @@ CONFIG_NET_KEY_MIGRATE=y
CONFIG_INET=y
CONFIG_IP_MULTICAST=y
CONFIG_NET_IPIP=m
-CONFIG_NET_IPGRE=m
-CONFIG_NET_IPGRE_BROADCAST=y
CONFIG_IP_MROUTE=y
CONFIG_IP_PIMSM_V1=y
CONFIG_IP_PIMSM_V2=y
@@ -95,17 +93,19 @@ CONFIG_DM_SNAPSHOT=m
CONFIG_DM_MIRROR=m
CONFIG_DM_ZERO=m
CONFIG_NETDEVICES=y
-CONFIG_NET_ETHERNET=y
CONFIG_MII=m
+CONFIG_NET_VENDOR_AMD=y
CONFIG_SUNLANCE=m
+CONFIG_NET_VENDOR_BROADCOM=y
+CONFIG_BNX2=m
+CONFIG_TIGON3=m
+CONFIG_NET_VENDOR_INTEL=y
+CONFIG_E1000=m
+CONFIG_E1000E=m
+CONFIG_NET_VENDOR_SUN=y
CONFIG_HAPPYMEAL=m
CONFIG_SUNGEM=m
CONFIG_SUNVNET=m
-CONFIG_NET_PCI=y
-CONFIG_E1000=m
-CONFIG_E1000E=m
-CONFIG_TIGON3=m
-CONFIG_BNX2=m
CONFIG_NIU=m
# CONFIG_WLAN is not set
CONFIG_PPP=m
@@ -126,13 +126,13 @@ CONFIG_INPUT_SPARCSPKR=y
# CONFIG_SERIO_SERPORT is not set
CONFIG_SERIO_PCIPS2=m
CONFIG_SERIO_RAW=m
+# CONFIG_LEGACY_PTYS is not set
# CONFIG_DEVKMEM is not set
CONFIG_SERIAL_SUNSU=y
CONFIG_SERIAL_SUNSU_CONSOLE=y
CONFIG_SERIAL_SUNSAB=y
CONFIG_SERIAL_SUNSAB_CONSOLE=y
CONFIG_SERIAL_SUNHV=y
-# CONFIG_LEGACY_PTYS is not set
CONFIG_FB=y
CONFIG_FB_TILEBLITTING=y
CONFIG_FB_SBUS=y
@@ -206,10 +206,8 @@ CONFIG_PRINTK_TIME=y
CONFIG_MAGIC_SYSRQ=y
CONFIG_DEBUG_KERNEL=y
CONFIG_LOCKUP_DETECTOR=y
-CONFIG_DETECT_HUNG_TASK=y
# CONFIG_SCHED_DEBUG is not set
CONFIG_SCHEDSTATS=y
-# CONFIG_RCU_CPU_STALL_DETECTOR is not set
CONFIG_SYSCTL_SYSCALL_CHECK=y
CONFIG_BLK_DEV_IO_TRACE=y
CONFIG_KEYS=y
--
1.7.5.4
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/
^ permalink raw reply related
* Re: linux-next: boot test failure (net tree)
From: Stephen Rothwell @ 2011-08-23 1:40 UTC (permalink / raw)
To: David Miller
Cc: netdev, linux-next, linux-kernel, jeffrey.t.kirsher, mikey,
torvalds, akpm, ppc-dev, Benjamin Herrenschmidt, Paul Mackerras
In-Reply-To: <20110822113032.15087c2e190e2b0c3ee7dfb8@canb.auug.org.au>
Hi Dave,
On Mon, 22 Aug 2011 11:30:32 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>
> Here's what I am applying as a merge fixup to the net tree today so that
> my ppc64_defconfig builds actually build more or less the same set of
> drivers as before this rearrangement.
And this today:
From: Stephen Rothwell <sfr@canb.auug.org.au>
Date: Tue, 23 Aug 2011 11:23:40 +1000
Subject: [PATCH] sparc: update sparc32_defconfig for net device movement
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
arch/sparc/configs/sparc32_defconfig | 9 ++-------
1 files changed, 2 insertions(+), 7 deletions(-)
diff --git a/arch/sparc/configs/sparc32_defconfig b/arch/sparc/configs/sparc32_defconfig
index fb23fd6..9bc241a 100644
--- a/arch/sparc/configs/sparc32_defconfig
+++ b/arch/sparc/configs/sparc32_defconfig
@@ -2,9 +2,7 @@ CONFIG_EXPERIMENTAL=y
CONFIG_SYSVIPC=y
CONFIG_POSIX_MQUEUE=y
CONFIG_LOG_BUF_SHIFT=14
-CONFIG_SYSFS_DEPRECATED_V2=y
CONFIG_BLK_DEV_INITRD=y
-# CONFIG_CC_OPTIMIZE_FOR_SIZE is not set
CONFIG_SLAB=y
CONFIG_MODULES=y
CONFIG_MODULE_UNLOAD=y
@@ -42,9 +40,10 @@ CONFIG_SCSI_QLOGICPTI=m
CONFIG_SCSI_SUNESP=y
CONFIG_NETDEVICES=y
CONFIG_DUMMY=m
-CONFIG_NET_ETHERNET=y
CONFIG_MII=m
+CONFIG_NET_VENDOR_AMD=y
CONFIG_SUNLANCE=y
+CONFIG_NET_VENDOR_SUN=y
CONFIG_HAPPYMEAL=m
CONFIG_SUNBMAC=m
CONFIG_SUNQE=m
@@ -64,26 +63,22 @@ CONFIG_SERIAL_SUNSU=y
CONFIG_SERIAL_SUNSU_CONSOLE=y
CONFIG_SPI=y
CONFIG_SPI_XILINX=m
-CONFIG_SPI_XILINX_PLTFM=m
CONFIG_SUN_OPENPROMIO=m
CONFIG_EXT2_FS=y
CONFIG_EXT2_FS_XATTR=y
CONFIG_EXT2_FS_POSIX_ACL=y
CONFIG_EXT2_FS_SECURITY=y
-CONFIG_AUTOFS_FS=m
CONFIG_AUTOFS4_FS=m
CONFIG_ISO9660_FS=m
CONFIG_PROC_KCORE=y
CONFIG_ROMFS_FS=m
CONFIG_NFS_FS=y
CONFIG_ROOT_NFS=y
-CONFIG_RPCSEC_GSS_KRB5=m
CONFIG_NLS=y
# CONFIG_ENABLE_WARN_DEPRECATED is not set
CONFIG_DEBUG_KERNEL=y
CONFIG_DETECT_HUNG_TASK=y
# CONFIG_SCHED_DEBUG is not set
-# CONFIG_RCU_CPU_STALL_DETECTOR is not set
CONFIG_KGDB=y
CONFIG_KGDB_TESTS=y
CONFIG_CRYPTO_NULL=m
--
1.7.5.4
--
Cheers,
Stephen Rothwell sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/
^ permalink raw reply related
* Re: [PATCH 00/11] various fixes v3
From: Greg KH @ 2011-08-23 1:32 UTC (permalink / raw)
To: Jason Baron
Cc: gregkh, joe, jim.cromie, bvanassche, linux-kernel, davem,
aloisio.almeida, netdev
In-Reply-To: <cover.1313085588.git.jbaron@redhat.com>
On Thu, Aug 11, 2011 at 02:36:17PM -0400, Jason Baron wrote:
> Hi,
>
> Dynamic debug fixes and cleanups. Only changes from v2 are a re-base against
> Linus's latest tree and the inclusion of the re-posted version for patch #11.
I applied the first 8, as it seemed like there was lots of discussion
about #9.
Care to send the rest when you all have those worked out?
thanks,
greg k-h
^ permalink raw reply
* Re: [Patch] Scm: Remove unnecessary pid & credential references in Unix socket's send and receive path
From: Tim Chen @ 2011-08-23 0:57 UTC (permalink / raw)
To: David Miller
Cc: eric.dumazet, viro, ebiederm, ak, matt.fleming, linux-kernel,
netdev
In-Reply-To: <20110822.174013.353730308624819543.davem@davemloft.net>
On Mon, 2011-08-22 at 17:40 -0700, David Miller wrote:
> From: Tim Chen <tim.c.chen@linux.intel.com>
> Date: Fri, 19 Aug 2011 09:44:58 -0700
>
> > - /* Only send the fds in the first buffer */
> > - err = unix_scm_to_skb(siocb->scm, skb, !fds_sent);
> > + /* Only send the fds and no ref to pid in the first buffer */
> > + if (fds_sent)
> > + err = unix_scm_to_skb(siocb->scm, skb, !fds_sent, true);
> > + else
> > + err = unix_scm_to_skb(siocb->scm, skb, !fds_sent, false);
>
> Just set this final boolean the way the third argument is, there is no
> reason to replicate the entire function call twice just to set the
> final argument to what "fds_sent" evaluates to as a boolean.
>
> err = unix_scm_to_skb(siocb->scm, skb, !fds_sent, fds_sent);
>
> ought to suffice.
Good point. Patch updated as suggested.
Tim
-----------
Patch series 109f6e39..7361c36c back in 2.6.36 added functionality to
allow credentials to work across pid namespaces for packets sent via
UNIX sockets. However, the atomic reference counts on pid and
credentials caused plenty of cache bouncing when there are numerous
threads of the same pid sharing a UNIX socket. This patch mitigates the
problem by eliminating extraneous reference counts on pid and
credentials on both send and receive path of UNIX sockets. I found a 2x
improvement in hackbench's threaded case.
On the receive path in unix_dgram_recvmsg, currently there is an
increment of reference count on pid and credentials in scm_set_cred.
Then there are two decrement of the reference counts. Once in scm_recv
and once when skb_free_datagram call skb->destructor function
unix_destruct_scm. One pair of increment and decrement of ref count on
pid and credentials can be eliminated from the receive path. Until we
destroy the skb, we already set a reference when we created the skb on
the send side.
On the send path, there are two increments of ref count on pid and
credentials, once in scm_send and once in unix_scm_to_skb. Then there
is a decrement of the reference counts in scm_destroy's call to
scm_destroy_cred at the end of unix_dgram_sendmsg functions. One pair
of increment and decrement of the reference counts can be removed so we
only need to increment the ref counts once.
By incorporating these changes, for hackbench running on a 4 socket
NHM-EX machine with 40 cores, the execution of hackbench on
50 groups of 20 threads sped up by factor of 2.
Hackbench command used for testing:
./hackbench 50 thread 2000
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
---
diff --git a/include/net/scm.h b/include/net/scm.h
index 745460f..68e1e48 100644
--- a/include/net/scm.h
+++ b/include/net/scm.h
@@ -53,6 +53,14 @@ static __inline__ void scm_set_cred(struct scm_cookie *scm,
cred_to_ucred(pid, cred, &scm->creds);
}
+static __inline__ void scm_set_cred_noref(struct scm_cookie *scm,
+ struct pid *pid, const struct cred *cred)
+{
+ scm->pid = pid;
+ scm->cred = cred;
+ cred_to_ucred(pid, cred, &scm->creds);
+}
+
static __inline__ void scm_destroy_cred(struct scm_cookie *scm)
{
put_pid(scm->pid);
@@ -70,6 +78,15 @@ static __inline__ void scm_destroy(struct scm_cookie *scm)
__scm_destroy(scm);
}
+static __inline__ void scm_release(struct scm_cookie *scm)
+{
+ /* keep ref on pid and cred */
+ scm->pid = NULL;
+ scm->cred = NULL;
+ if (scm->fp)
+ __scm_destroy(scm);
+}
+
static __inline__ int scm_send(struct socket *sock, struct msghdr *msg,
struct scm_cookie *scm)
{
@@ -108,15 +125,14 @@ static __inline__ void scm_recv(struct socket *sock, struct msghdr *msg,
if (!msg->msg_control) {
if (test_bit(SOCK_PASSCRED, &sock->flags) || scm->fp)
msg->msg_flags |= MSG_CTRUNC;
- scm_destroy(scm);
+ if (scm && scm->fp)
+ __scm_destroy(scm);
return;
}
if (test_bit(SOCK_PASSCRED, &sock->flags))
put_cmsg(msg, SOL_SOCKET, SCM_CREDENTIALS, sizeof(scm->creds), &scm->creds);
- scm_destroy_cred(scm);
-
scm_passec(sock, msg, scm);
if (!scm->fp)
diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c
index 0722a25..136298c 100644
--- a/net/unix/af_unix.c
+++ b/net/unix/af_unix.c
@@ -1382,11 +1382,17 @@ static int unix_attach_fds(struct scm_cookie *scm, struct sk_buff *skb)
return max_level;
}
-static int unix_scm_to_skb(struct scm_cookie *scm, struct sk_buff *skb, bool send_fds)
+static int unix_scm_to_skb(struct scm_cookie *scm, struct sk_buff *skb,
+ bool send_fds, bool ref)
{
int err = 0;
- UNIXCB(skb).pid = get_pid(scm->pid);
- UNIXCB(skb).cred = get_cred(scm->cred);
+ if (ref) {
+ UNIXCB(skb).pid = get_pid(scm->pid);
+ UNIXCB(skb).cred = get_cred(scm->cred);
+ } else {
+ UNIXCB(skb).pid = scm->pid;
+ UNIXCB(skb).cred = scm->cred;
+ }
UNIXCB(skb).fp = NULL;
if (scm->fp && send_fds)
err = unix_attach_fds(scm, skb);
@@ -1411,7 +1417,7 @@ static int unix_dgram_sendmsg(struct kiocb *kiocb, struct socket *sock,
int namelen = 0; /* fake GCC */
int err;
unsigned hash;
- struct sk_buff *skb;
+ struct sk_buff *skb = NULL;
long timeo;
struct scm_cookie tmp_scm;
int max_level;
@@ -1452,7 +1458,7 @@ static int unix_dgram_sendmsg(struct kiocb *kiocb, struct socket *sock,
if (skb == NULL)
goto out;
- err = unix_scm_to_skb(siocb->scm, skb, true);
+ err = unix_scm_to_skb(siocb->scm, skb, true, false);
if (err < 0)
goto out_free;
max_level = err + 1;
@@ -1548,7 +1554,7 @@ restart:
unix_state_unlock(other);
other->sk_data_ready(other, len);
sock_put(other);
- scm_destroy(siocb->scm);
+ scm_release(siocb->scm);
return len;
out_unlock:
@@ -1558,7 +1564,8 @@ out_free:
out:
if (other)
sock_put(other);
- scm_destroy(siocb->scm);
+ if (skb == NULL)
+ scm_destroy(siocb->scm);
return err;
}
@@ -1570,7 +1577,7 @@ static int unix_stream_sendmsg(struct kiocb *kiocb, struct socket *sock,
struct sock *sk = sock->sk;
struct sock *other = NULL;
int err, size;
- struct sk_buff *skb;
+ struct sk_buff *skb = NULL;
int sent = 0;
struct scm_cookie tmp_scm;
bool fds_sent = false;
@@ -1635,11 +1642,11 @@ static int unix_stream_sendmsg(struct kiocb *kiocb, struct socket *sock,
size = min_t(int, size, skb_tailroom(skb));
- /* Only send the fds in the first buffer */
- err = unix_scm_to_skb(siocb->scm, skb, !fds_sent);
+ /* Only send the fds and no ref to pid in the first buffer */
+ err = unix_scm_to_skb(siocb->scm, skb, !fds_sent, fds_sent);
if (err < 0) {
kfree_skb(skb);
- goto out_err;
+ goto out;
}
max_level = err + 1;
fds_sent = true;
@@ -1647,7 +1654,7 @@ static int unix_stream_sendmsg(struct kiocb *kiocb, struct socket *sock,
err = memcpy_fromiovec(skb_put(skb, size), msg->msg_iov, size);
if (err) {
kfree_skb(skb);
- goto out_err;
+ goto out;
}
unix_state_lock(other);
@@ -1664,7 +1671,10 @@ static int unix_stream_sendmsg(struct kiocb *kiocb, struct socket *sock,
sent += size;
}
- scm_destroy(siocb->scm);
+ if (skb)
+ scm_release(siocb->scm);
+ else
+ scm_destroy(siocb->scm);
siocb->scm = NULL;
return sent;
@@ -1677,7 +1687,9 @@ pipe_err:
send_sig(SIGPIPE, current, 0);
err = -EPIPE;
out_err:
- scm_destroy(siocb->scm);
+ if (skb == NULL)
+ scm_destroy(siocb->scm);
+out:
siocb->scm = NULL;
return sent ? : err;
}
@@ -1781,7 +1793,7 @@ static int unix_dgram_recvmsg(struct kiocb *iocb, struct socket *sock,
siocb->scm = &tmp_scm;
memset(&tmp_scm, 0, sizeof(tmp_scm));
}
- scm_set_cred(siocb->scm, UNIXCB(skb).pid, UNIXCB(skb).cred);
+ scm_set_cred_noref(siocb->scm, UNIXCB(skb).pid, UNIXCB(skb).cred);
unix_set_secdata(siocb->scm, skb);
if (!(flags & MSG_PEEK)) {
@@ -1943,7 +1955,8 @@ static int unix_stream_recvmsg(struct kiocb *iocb, struct socket *sock,
}
} else {
/* Copy credentials */
- scm_set_cred(siocb->scm, UNIXCB(skb).pid, UNIXCB(skb).cred);
+ scm_set_cred_noref(siocb->scm, UNIXCB(skb).pid,
+ UNIXCB(skb).cred);
check_creds = 1;
}
^ permalink raw reply related
* Re: Miscalculated TCP ACK ?
From: Gabriel Beddingfield @ 2011-08-23 0:41 UTC (permalink / raw)
To: netdev
In-Reply-To: <1314038973.3777.0.camel@edumazet-laptop>
On 08/22/2011 01:49 PM, Eric Dumazet wrote:
> Le lundi 22 août 2011 à 13:05 -0500, Gabriel Beddingfield a écrit :
>> On Mon, Aug 22, 2011 at 12:19 PM, Daniel Baluta<daniel.baluta@gmail.com> wrote:
>>>>> It appears to me that the ACK is off-by-one (should have been 27676).
>>>>> The server replies again with retransmissions. This is with several
>>>>> kernel versions on local machines, ranging from 2.6.31 to 2.6.38.
>>>
>>>> Not a bug if frame carried a FIN
>>
>> The server packet had the FIN bit set.
>>
>
> then its fine.
Thank you!
>> I clipped the conversation to the vicinity of the problem.
>
> This sounds as a TSO bug, maybe driver/NIC related
>
> What is the NIC/driver used on your server ?
New news: If I remove my router (Linksys WRTU54G-TM), the problem goes
away.
Thank you guys for your helpful insights. :-)
-gabriel
^ permalink raw reply
* Re: [Patch] Scm: Remove unnecessary pid & credential references in Unix socket's send and receive path
From: David Miller @ 2011-08-23 0:40 UTC (permalink / raw)
To: tim.c.chen
Cc: eric.dumazet, viro, ebiederm, ak, matt.fleming, linux-kernel,
netdev
In-Reply-To: <1313772298.2576.2916.camel@schen9-DESK>
From: Tim Chen <tim.c.chen@linux.intel.com>
Date: Fri, 19 Aug 2011 09:44:58 -0700
> - /* Only send the fds in the first buffer */
> - err = unix_scm_to_skb(siocb->scm, skb, !fds_sent);
> + /* Only send the fds and no ref to pid in the first buffer */
> + if (fds_sent)
> + err = unix_scm_to_skb(siocb->scm, skb, !fds_sent, true);
> + else
> + err = unix_scm_to_skb(siocb->scm, skb, !fds_sent, false);
Just set this final boolean the way the third argument is, there is no
reason to replicate the entire function call twice just to set the
final argument to what "fds_sent" evaluates to as a boolean.
err = unix_scm_to_skb(siocb->scm, skb, !fds_sent, fds_sent);
ought to suffice.
^ permalink raw reply
* Re: pull request: wireless-next 2011-08-22
From: David Miller @ 2011-08-23 0:14 UTC (permalink / raw)
To: linville; +Cc: linux-wireless, netdev
In-Reply-To: <20110822195857.GL2534@tuxdriver.com>
From: "John W. Linville" <linville@tuxdriver.com>
Date: Mon, 22 Aug 2011 15:58:57 -0400
> This is a batch of updates intended for 3.2.
>
> Highlights include a wireless extensions dependency cleanup amongst
> a number of drivers, and the inclusion of the ath6kl driver
> in drivers/net/wireless (along with its removal from staging).
> Also included are a number of updates to mac80211, bcma, mwifiex,
> libertas, ath9k, iwlagn, and a number of other drivers.
>
> Please let me know if there are problems!
Pulled, thanks!
^ permalink raw reply
* Re: [PATCH] net: add APIs for manipulating skb page fragments.
From: David Miller @ 2011-08-23 0:00 UTC (permalink / raw)
To: Ian.Campbell; +Cc: netdev, linux-kernel, eric.dumazet, mirq-linux
In-Reply-To: <1313915037.17978.686.camel@dagon.hellion.org.uk>
From: Ian Campbell <Ian.Campbell@eu.citrix.com>
Date: Sun, 21 Aug 2011 09:23:57 +0100
> On Sun, 2011-08-21 at 01:31 +0100, David Miller wrote:
>> From: Ian Campbell <ian.campbell@citrix.com>
>> Date: Fri, 19 Aug 2011 17:25:00 +0100
>>
>> > The primary aim is to add skb_frag_(ref|unref) in order to remove the use of
>> > bare get/put_page on SKB pages fragments and to isolate users from subsequent
>> > changes to the skb_frag_t data structure.
>> >
>> > Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
>>
>> You're going to have to protect all of the things using the interfaces
>> from linux/dma-mapping.h with CONFIG_HAS_DMA otherwise it won't build
>> on platforms like S390.
>
> s390 is one of the arches which I build tested and I initially saw this
> issue too. I did add CONFIG_HAS_DMA but it turns out that
> linux/dma-mapping.h takes care of this by including
> asm-generic/dma-mapping-broken.h for you so I removed the #ifdef again.
> The header defines the prototypes to allow building but causes a link
> time failure if anything actually uses the interfaces.
>
> I just tested a s390x defconfig build again and it appears to be fine.
Thanks for explaining this.
Patch applied, thanks!
^ permalink raw reply
* Re: net_device leak after "bridge: allow creating bridge devices with netlink"?
From: David Miller @ 2011-08-22 23:48 UTC (permalink / raw)
To: eric.dumazet; +Cc: shemminger, JBeulich, netdev
In-Reply-To: <1314029159.2307.99.camel@edumazet-HP-Compaq-6005-Pro-SFF-PC>
From: Eric Dumazet <eric.dumazet@gmail.com>
Date: Mon, 22 Aug 2011 18:05:59 +0200
> [PATCH] bridge: fix a possible net_device leak
>
> Jan Beulich reported a possible net_device leak in bridge code after
> commit bb900b27a2f4 (bridge: allow creating bridge devices with netlink)
>
> Reported-by: Jan Beulich <JBeulich@novell.com>
> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
> Acked-by: Stephen Hemminger <shemminger@vyatta.com>
Applied, thanks Eric.
^ permalink raw reply
* Re: [net-next 04/12] bna: TX Path and RX Path Changes
From: David Miller @ 2011-08-22 23:45 UTC (permalink / raw)
To: rmody; +Cc: netdev, adapter_linux_open_src_team, gkaraje
In-Reply-To: <1314052869-7407-5-git-send-email-rmody@brocade.com>
From: Rasesh Mody <rmody@brocade.com>
Date: Mon, 22 Aug 2011 15:41:01 -0700
> + if (unlikely(skb_is_gso(skb) > netdev->mtu)) {
skb_is_gso() is mean to be a boolean test, it happens to return
the value of ->gso_size as it's implementation but that's not
a good reason for you to abuse this interface in this way.
If we every fix skb_is_gso() to return a true 'bool' your driver
will break.
Access ->gso_size directly if that's what you're interested in,
that way it's future proof and won't break in the future.
^ permalink raw reply
* Re: [net-next 03/10] ixgbe: Drop the TX work limit and instead just leave it to budget
From: David Miller @ 2011-08-22 23:40 UTC (permalink / raw)
To: alexander.h.duyck; +Cc: bhutchings, jeffrey.t.kirsher, netdev, gospo
In-Reply-To: <4E52DEEF.40504@intel.com>
From: Alexander Duyck <alexander.h.duyck@intel.com>
Date: Mon, 22 Aug 2011 15:57:51 -0700
> The problem was occurring even without large rings.
> I was seeing issues with rings just 256 descriptors in size.
And the default in the ixgbe driver is 512 entries which I think
itself is quite excessive. Something like 128 is more in line with
what I'd call a sane default.
So the only side effect of your change is to decrease the TX quota to
64 (the default NAPI quota) from it's current value of 512
(IXGBE_DEFAULT_TXD).
Talking about the existing code, I can't even see how the current
driver private TX quota can trigger except in the most extreme cases.
This is because the quota is set the same as the size you're setting
the TX ring to.
> The problem seemed to be that the TX cleanup being a multiple of
> budget was allowing one CPU to overwhelm the other and the fact that
> the TX was essentially unbounded was just allowing the issue to
> feedback on itself.
I still don't understand what issue you could even be running into.
On each CPU we round robin against all NAPI requestors for that CPU.
In your routing test setup, we should have one cpu doing the RX and
another different cpu doing TX.
Therefore if the TX cpu simply spins in a loop doing nothing but TX
reclaim work it should not really matter.
And if you hit the TX budget on the TX cpu, it's just going to come
right back into the ixgbe NAPI handler and thus the TX reclaim
processing not even a dozen cycles later.
The only effect is to have us go through the whole function call
sequence and data structure setup into local variables more than you
would be doing so before.
> In addition since the RX and TX workload was balanced it kept both
> locked into polling while the CPU was saturated instead of allowing
> the TX to become interrupt driven. In addition since the TX was
> working on the same budget as the RX the number of SKBs freed up in
> the TX path would match the number consumed when being reallocated
> on the RX path.
So the only conclusion I can come to is that what happens is we're now
executing what are essentially wasted cpu cycles and this takes us
over the threshold such that we poll more and take interrupts less.
And this improves performance.
That's pretty unwise if you ask me, we should do something useful with
cpu cycles instead of wasting them merely to make us poll more.
> The problem seemed to be present as long as I allowed the TX budget to
> be a multiple of the RX budget. The easiest way to keep things
> balanced and avoid allowing the TX from one CPU to overwhelm the RX on
> another was just to keep the budgets equal.
You're executing 10 or 20 cpu cycles after every 64 TX reclaims,
that's the only effect of these changes. That's not even long enough
for a cache line to transfer between two cpus.
^ permalink raw reply
* Re: [net-next 03/10] ixgbe: Drop the TX work limit and instead just leave it to budget
From: Alexander Duyck @ 2011-08-22 22:57 UTC (permalink / raw)
To: David Miller; +Cc: bhutchings, jeffrey.t.kirsher, netdev, gospo
In-Reply-To: <20110822.135644.683110224886588181.davem@davemloft.net>
On 08/22/2011 01:56 PM, David Miller wrote:
> From: Alexander Duyck<alexander.h.duyck@intel.com>
> Date: Mon, 22 Aug 2011 10:29:51 -0700
>
>> The only problem I was seeing with that was that in certain cases it
>> seemed like the TX cleanup could consume enough CPU time to cause
>> pretty significant delays in processing the RX cleanup. This in turn
>> was causing single queue bi-directional routing tests to come out
>> pretty unbalanced since what seemed to happen is that one CPUs RX work
>> would overwhelm the other CPU with the TX processing resulting in an
>> unbalanced flow that was something like a 60/40 split between the
>> upstream and downstream throughput.
> But the problem is that now you're applying the budget to two operations
> that have much differing costs. Freeing up a TX ring packet is probably
> on the order of 1/10th the cost of processing an incoming RX ring frame.
>
> I've advocated to not apply the budget at all to TX ring processing.
I fully understand that the TX path is much cheaper than the RX path.
One step I have taken in all of this code is that the TX path only
counts SKBs cleaned, it doesn't count descriptors. So a single
descriptor 60byte transmit will cost the same as a 64K 18 descriptor
TSO. All I am really counting is the number of times I have called
dev_kfree_skb_any();
> I can see your delimma with respect to RX ring processing being delayed,
> but if that's really happening you can consider whether the TX ring is
> simply too large.
The problem was occurring even without large rings. I was seeing issues
with rings just 256 descriptors in size. The problem seemed to be that
the TX cleanup being a multiple of budget was allowing one CPU to
overwhelm the other and the fact that the TX was essentially unbounded
was just allowing the issue to feedback on itself.
In the routing test case I was actually seeing significant advantages to
this approach as we were essentially cleaning just the right number of
buffers to make room for the next set of transmits when the RX cleanup
came though. In addition since the RX and TX workload was balanced it
kept both locked into polling while the CPU was saturated instead of
allowing the TX to become interrupt driven. In addition since the TX
was working on the same budget as the RX the number of SKBs freed up in
the TX path would match the number consumed when being reallocated on
the RX path.
> In any event can you try something like dampening the cost applied to
> budget for TX work (1/2, 1/4, etc.)? Because as far as I can tell, if
> you are really hitting the budget limit on TX then you won't be doing
> any RX work on that device until a future NAPI round that depletes the
> TX ring work without going over the budget.
The problem seemed to be present as long as I allowed the TX budget to
be a multiple of the RX budget. The easiest way to keep things balanced
and avoid allowing the TX from one CPU to overwhelm the RX on another
was just to keep the budgets equal.
I'm a bit confused by this last comment. The full budget is used for TX
and RX, it isn't divided. I do a budget worth of TX cleanup and a
budget worth of RX cleanup within the ixgbe_poll routine, and if either
of them consume their full budget then I return the budget value as the
work done.
If you are referring to the case where two devices are sharing the CPU
then I would suspect this might lead to faster consumption of the
netdev_budget, but other than that I don't see any starvation issues for
RX or TX.
Thanks,
Alex
^ permalink raw reply
* [net-next 12/12] bna: Driver Version changed to 3.0.2.1
From: Rasesh Mody @ 2011-08-22 22:41 UTC (permalink / raw)
To: davem, netdev; +Cc: adapter_linux_open_src_team, Rasesh Mody, Gurunatha Karaje
In-Reply-To: <1314052869-7407-1-git-send-email-rmody@brocade.com>
Signed-off-by: Gurunatha Karaje <gkaraje@brocade.com>
Signed-off-by: Rasesh Mody <rmody@brocade.com>
---
drivers/net/ethernet/brocade/bna/bnad.h | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/net/ethernet/brocade/bna/bnad.h b/drivers/net/ethernet/brocade/bna/bnad.h
index 41c984c..7bdde74 100644
--- a/drivers/net/ethernet/brocade/bna/bnad.h
+++ b/drivers/net/ethernet/brocade/bna/bnad.h
@@ -71,7 +71,7 @@ struct bnad_rx_ctrl {
#define BNAD_NAME "bna"
#define BNAD_NAME_LEN 64
-#define BNAD_VERSION "3.0.2.0"
+#define BNAD_VERSION "3.0.2.1"
#define BNAD_MAILBOX_MSIX_INDEX 0
#define BNAD_MAILBOX_MSIX_VECTORS 1
--
1.7.1
^ permalink raw reply related
* [net-next 11/12] bna: SKB PCI UNMAP Fix
From: Rasesh Mody @ 2011-08-22 22:41 UTC (permalink / raw)
To: davem, netdev; +Cc: adapter_linux_open_src_team, Rasesh Mody, Gurunatha Karaje
In-Reply-To: <1314052869-7407-1-git-send-email-rmody@brocade.com>
Change details:
- Found a leak in sk_buff unmapping of PCI dma addresses where boundary
conditions are not properly handled in freeing all Tx buffers. Freeing
of all Tx buffers is done considering sk_buffs data and fragments can
be mapped at the boundary.
Signed-off-by: Gurunatha Karaje <gkaraje@brocade.com>
Signed-off-by: Rasesh Mody <rmody@brocade.com>
---
drivers/net/ethernet/brocade/bna/bnad.c | 39 ++++++++-----------------------
1 files changed, 10 insertions(+), 29 deletions(-)
diff --git a/drivers/net/ethernet/brocade/bna/bnad.c b/drivers/net/ethernet/brocade/bna/bnad.c
index d94497d..6ae531e 100644
--- a/drivers/net/ethernet/brocade/bna/bnad.c
+++ b/drivers/net/ethernet/brocade/bna/bnad.c
@@ -133,39 +133,20 @@ bnad_free_all_txbufs(struct bnad *bnad,
struct bnad_unmap_q *unmap_q = tcb->unmap_q;
struct bnad_skb_unmap *unmap_array;
struct sk_buff *skb = NULL;
- int i;
+ int q;
unmap_array = unmap_q->unmap_array;
- unmap_cons = 0;
- while (unmap_cons < unmap_q->q_depth) {
- skb = unmap_array[unmap_cons].skb;
- if (!skb) {
- unmap_cons++;
+ for (q = 0; q < unmap_q->q_depth; q++) {
+ skb = unmap_array[q].skb;
+ if (!skb)
continue;
- }
- unmap_array[unmap_cons].skb = NULL;
-
- dma_unmap_single(&bnad->pcidev->dev,
- dma_unmap_addr(&unmap_array[unmap_cons],
- dma_addr), skb_headlen(skb),
- DMA_TO_DEVICE);
- dma_unmap_addr_set(&unmap_array[unmap_cons], dma_addr, 0);
- if (++unmap_cons >= unmap_q->q_depth)
- break;
+ unmap_cons = q;
+ BNAD_PCI_UNMAP_SKB(&bnad->pcidev->dev, unmap_array, unmap_cons,
+ unmap_q->q_depth, skb,
+ skb_shinfo(skb)->nr_frags);
- for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
- dma_unmap_page(&bnad->pcidev->dev,
- dma_unmap_addr(&unmap_array[unmap_cons],
- dma_addr),
- skb_shinfo(skb)->frags[i].size,
- DMA_TO_DEVICE);
- dma_unmap_addr_set(&unmap_array[unmap_cons], dma_addr,
- 0);
- if (++unmap_cons >= unmap_q->q_depth)
- break;
- }
dev_kfree_skb_any(skb);
}
}
--
1.7.1
^ permalink raw reply related
* [net-next 10/12] bna: Queue Depth and SKB Unmap Array Fix
From: Rasesh Mody @ 2011-08-22 22:41 UTC (permalink / raw)
To: davem, netdev; +Cc: adapter_linux_open_src_team, Rasesh Mody, Gurunatha Karaje
In-Reply-To: <1314052869-7407-1-git-send-email-rmody@brocade.com>
Change details:
- sk_buff unmap_array grows greater than 65536 (x2) with Tx ring of 65536.
The index used for accessing it is incorrectly declared as u16. It quickly
wraps around and accesses null sk_buff ptr. So using u32 to handle
unmap_array.
- Reducing TXQ depth and safe(max) acking of Tx events to 32768 (same as Rx)
Signed-off-by: Gurunatha Karaje <gkaraje@brocade.com>
Signed-off-by: Rasesh Mody <rmody@brocade.com>
---
drivers/net/ethernet/brocade/bna/bnad.c | 4 ++--
drivers/net/ethernet/brocade/bna/bnad.h | 4 ++++
drivers/net/ethernet/brocade/bna/bnad_ethtool.c | 8 ++++----
3 files changed, 10 insertions(+), 6 deletions(-)
diff --git a/drivers/net/ethernet/brocade/bna/bnad.c b/drivers/net/ethernet/brocade/bna/bnad.c
index 5df21d0..d94497d 100644
--- a/drivers/net/ethernet/brocade/bna/bnad.c
+++ b/drivers/net/ethernet/brocade/bna/bnad.c
@@ -182,8 +182,8 @@ static u32
bnad_free_txbufs(struct bnad *bnad,
struct bna_tcb *tcb)
{
- u32 sent_packets = 0, sent_bytes = 0;
- u16 wis, unmap_cons, updated_hw_cons;
+ u32 unmap_cons, sent_packets = 0, sent_bytes = 0;
+ u16 wis, updated_hw_cons;
struct bnad_unmap_q *unmap_q = tcb->unmap_q;
struct bnad_skb_unmap *unmap_array;
struct sk_buff *skb;
diff --git a/drivers/net/ethernet/brocade/bna/bnad.h b/drivers/net/ethernet/brocade/bna/bnad.h
index b31b893..41c984c 100644
--- a/drivers/net/ethernet/brocade/bna/bnad.h
+++ b/drivers/net/ethernet/brocade/bna/bnad.h
@@ -86,6 +86,10 @@ struct bnad_rx_ctrl {
#define BNAD_MAX_Q_DEPTH 0x10000
#define BNAD_MIN_Q_DEPTH 0x200
+#define BNAD_MAX_RXQ_DEPTH (BNAD_MAX_Q_DEPTH / bnad_rxqs_per_cq)
+/* keeping MAX TX and RX Q depth equal */
+#define BNAD_MAX_TXQ_DEPTH BNAD_MAX_RXQ_DEPTH
+
#define BNAD_JUMBO_MTU 9000
#define BNAD_NETIF_WAKE_THRESHOLD 8
diff --git a/drivers/net/ethernet/brocade/bna/bnad_ethtool.c b/drivers/net/ethernet/brocade/bna/bnad_ethtool.c
index 1199f01..e85fb2b 100644
--- a/drivers/net/ethernet/brocade/bna/bnad_ethtool.c
+++ b/drivers/net/ethernet/brocade/bna/bnad_ethtool.c
@@ -407,10 +407,10 @@ bnad_get_ringparam(struct net_device *netdev,
{
struct bnad *bnad = netdev_priv(netdev);
- ringparam->rx_max_pending = BNAD_MAX_Q_DEPTH / bnad_rxqs_per_cq;
+ ringparam->rx_max_pending = BNAD_MAX_RXQ_DEPTH;
ringparam->rx_mini_max_pending = 0;
ringparam->rx_jumbo_max_pending = 0;
- ringparam->tx_max_pending = BNAD_MAX_Q_DEPTH;
+ ringparam->tx_max_pending = BNAD_MAX_TXQ_DEPTH;
ringparam->rx_pending = bnad->rxq_depth;
ringparam->rx_mini_max_pending = 0;
@@ -434,13 +434,13 @@ bnad_set_ringparam(struct net_device *netdev,
}
if (ringparam->rx_pending < BNAD_MIN_Q_DEPTH ||
- ringparam->rx_pending > BNAD_MAX_Q_DEPTH / bnad_rxqs_per_cq ||
+ ringparam->rx_pending > BNAD_MAX_RXQ_DEPTH ||
!BNA_POWER_OF_2(ringparam->rx_pending)) {
mutex_unlock(&bnad->conf_mutex);
return -EINVAL;
}
if (ringparam->tx_pending < BNAD_MIN_Q_DEPTH ||
- ringparam->tx_pending > BNAD_MAX_Q_DEPTH ||
+ ringparam->tx_pending > BNAD_MAX_TXQ_DEPTH ||
!BNA_POWER_OF_2(ringparam->tx_pending)) {
mutex_unlock(&bnad->conf_mutex);
return -EINVAL;
--
1.7.1
^ permalink raw reply related
* [net-next 09/12] bna: MBOX IRQ Flag Check after Locking
From: Rasesh Mody @ 2011-08-22 22:41 UTC (permalink / raw)
To: davem, netdev; +Cc: adapter_linux_open_src_team, Rasesh Mody, Gurunatha Karaje
In-Reply-To: <1314052869-7407-1-git-send-email-rmody@brocade.com>
Change details:
- Check the BNAD_RF_MBOX_IRQ_DISABLED flag after acquiring the bna_lock.
Signed-off-by: Gurunatha Karaje <gkaraje@brocade.com>
Signed-off-by: Rasesh Mody <rmody@brocade.com>
---
drivers/net/ethernet/brocade/bna/bnad.c | 18 +++++++++++-------
1 files changed, 11 insertions(+), 7 deletions(-)
diff --git a/drivers/net/ethernet/brocade/bna/bnad.c b/drivers/net/ethernet/brocade/bna/bnad.c
index 2f808c7..5df21d0 100644
--- a/drivers/net/ethernet/brocade/bna/bnad.c
+++ b/drivers/net/ethernet/brocade/bna/bnad.c
@@ -582,10 +582,11 @@ bnad_msix_mbox_handler(int irq, void *data)
unsigned long flags;
struct bnad *bnad = (struct bnad *)data;
- if (unlikely(test_bit(BNAD_RF_MBOX_IRQ_DISABLED, &bnad->run_flags)))
- return IRQ_HANDLED;
-
spin_lock_irqsave(&bnad->bna_lock, flags);
+ if (unlikely(test_bit(BNAD_RF_MBOX_IRQ_DISABLED, &bnad->run_flags))) {
+ spin_unlock_irqrestore(&bnad->bna_lock, flags);
+ return IRQ_HANDLED;
+ }
bna_intr_status_get(&bnad->bna, intr_status);
@@ -608,15 +609,18 @@ bnad_isr(int irq, void *data)
struct bnad_rx_ctrl *rx_ctrl;
struct bna_tcb *tcb = NULL;
- if (unlikely(test_bit(BNAD_RF_MBOX_IRQ_DISABLED, &bnad->run_flags)))
+ spin_lock_irqsave(&bnad->bna_lock, flags);
+ if (unlikely(test_bit(BNAD_RF_MBOX_IRQ_DISABLED, &bnad->run_flags))) {
+ spin_unlock_irqrestore(&bnad->bna_lock, flags);
return IRQ_NONE;
+ }
bna_intr_status_get(&bnad->bna, intr_status);
- if (unlikely(!intr_status))
+ if (unlikely(!intr_status)) {
+ spin_unlock_irqrestore(&bnad->bna_lock, flags);
return IRQ_NONE;
-
- spin_lock_irqsave(&bnad->bna_lock, flags);
+ }
if (BNA_IS_MBOX_ERR_INTR(&bnad->bna, intr_status))
bna_mbox_handler(&bnad->bna, intr_status);
--
1.7.1
^ permalink raw reply related
* [net-next 08/12] bna: Async Mode Tx Rx Init Fix
From: Rasesh Mody @ 2011-08-22 22:41 UTC (permalink / raw)
To: davem, netdev; +Cc: adapter_linux_open_src_team, Rasesh Mody, Gurunatha Karaje
In-Reply-To: <1314052869-7407-1-git-send-email-rmody@brocade.com>
Change details:
- Async mode of Tx/Rx queue initialization in BNAD from a task queue context
runs into non-unique taskq allocation issues. Get rid of Tx/Rx
initialization from task q context
- In the attach function, wait for IOC enable, then do Tx/Rx queue
initialization. Default BNA attributes are used when IOC enable from attach
fails and values are set to:
1 TxQ, 1 RxQ, 1 Unicast MAC, 1 RIT entry
Signed-off-by: Gurunatha Karaje <gkaraje@brocade.com>
Signed-off-by: Rasesh Mody <rmody@brocade.com>
---
drivers/net/ethernet/brocade/bna/bna_enet.c | 29 ++++++++++++++++++-----
drivers/net/ethernet/brocade/bna/bna_hw_defs.h | 4 +++
drivers/net/ethernet/brocade/bna/bna_types.h | 1 +
3 files changed, 27 insertions(+), 7 deletions(-)
diff --git a/drivers/net/ethernet/brocade/bna/bna_enet.c b/drivers/net/ethernet/brocade/bna/bna_enet.c
index 68a275d..26f5c5a 100644
--- a/drivers/net/ethernet/brocade/bna/bna_enet.c
+++ b/drivers/net/ethernet/brocade/bna/bna_enet.c
@@ -167,13 +167,14 @@ bna_bfi_attr_get_rsp(struct bna_ioceth *ioceth,
* Store only if not set earlier, since BNAD can override the HW
* attributes
*/
- if (!ioceth->attr.num_txq)
+ if (!ioceth->attr.fw_query_complete) {
ioceth->attr.num_txq = ntohl(rsp->max_cfg);
- if (!ioceth->attr.num_rxp)
ioceth->attr.num_rxp = ntohl(rsp->max_cfg);
- ioceth->attr.num_ucmac = ntohl(rsp->max_ucmac);
- ioceth->attr.num_mcmac = BFI_ENET_MAX_MCAM;
- ioceth->attr.max_rit_size = ntohl(rsp->rit_size);
+ ioceth->attr.num_ucmac = ntohl(rsp->max_ucmac);
+ ioceth->attr.num_mcmac = BFI_ENET_MAX_MCAM;
+ ioceth->attr.max_rit_size = ntohl(rsp->rit_size);
+ ioceth->attr.fw_query_complete = true;
+ }
bfa_fsm_send_event(ioceth, IOCETH_E_ENET_ATTR_RESP);
}
@@ -1693,6 +1694,16 @@ static struct bfa_ioc_cbfn bna_ioceth_cbfn = {
bna_cb_ioceth_reset
};
+static void bna_attr_init(struct bna_ioceth *ioceth)
+{
+ ioceth->attr.num_txq = BFI_ENET_DEF_TXQ;
+ ioceth->attr.num_rxp = BFI_ENET_DEF_RXP;
+ ioceth->attr.num_ucmac = BFI_ENET_DEF_UCAM;
+ ioceth->attr.num_mcmac = BFI_ENET_MAX_MCAM;
+ ioceth->attr.max_rit_size = BFI_ENET_DEF_RITSZ;
+ ioceth->attr.fw_query_complete = false;
+}
+
static void
bna_ioceth_init(struct bna_ioceth *ioceth, struct bna *bna,
struct bna_res_info *res_info)
@@ -1738,6 +1749,8 @@ bna_ioceth_init(struct bna_ioceth *ioceth, struct bna *bna,
ioceth->stop_cbfn = NULL;
ioceth->stop_cbarg = NULL;
+ bna_attr_init(ioceth);
+
bfa_fsm_set_state(ioceth, bna_ioceth_sm_stopped);
}
@@ -2036,7 +2049,8 @@ bna_uninit(struct bna *bna)
int
bna_num_txq_set(struct bna *bna, int num_txq)
{
- if (num_txq > 0 && (num_txq <= bna->ioceth.attr.num_txq)) {
+ if (bna->ioceth.attr.fw_query_complete &&
+ (num_txq <= bna->ioceth.attr.num_txq)) {
bna->ioceth.attr.num_txq = num_txq;
return BNA_CB_SUCCESS;
}
@@ -2047,7 +2061,8 @@ bna_num_txq_set(struct bna *bna, int num_txq)
int
bna_num_rxp_set(struct bna *bna, int num_rxp)
{
- if (num_rxp > 0 && (num_rxp <= bna->ioceth.attr.num_rxp)) {
+ if (bna->ioceth.attr.fw_query_complete &&
+ (num_rxp <= bna->ioceth.attr.num_rxp)) {
bna->ioceth.attr.num_rxp = num_rxp;
return BNA_CB_SUCCESS;
}
diff --git a/drivers/net/ethernet/brocade/bna/bna_hw_defs.h b/drivers/net/ethernet/brocade/bna/bna_hw_defs.h
index 7ecdca5..dde8a46 100644
--- a/drivers/net/ethernet/brocade/bna/bna_hw_defs.h
+++ b/drivers/net/ethernet/brocade/bna/bna_hw_defs.h
@@ -30,6 +30,10 @@
* SW imposed limits
*
*/
+#define BFI_ENET_DEF_TXQ 1
+#define BFI_ENET_DEF_RXP 1
+#define BFI_ENET_DEF_UCAM 1
+#define BFI_ENET_DEF_RITSZ 1
#define BFI_ENET_MAX_MCAM 256
diff --git a/drivers/net/ethernet/brocade/bna/bna_types.h b/drivers/net/ethernet/brocade/bna/bna_types.h
index 59417b1..242d799 100644
--- a/drivers/net/ethernet/brocade/bna/bna_types.h
+++ b/drivers/net/ethernet/brocade/bna/bna_types.h
@@ -323,6 +323,7 @@ struct bna_qpt {
};
struct bna_attr {
+ bool fw_query_complete;
int num_txq;
int num_rxp;
int num_ucmac;
--
1.7.1
^ permalink raw reply related
* [net-next 07/12] bna: Ethtool Enhancements and Fix
From: Rasesh Mody @ 2011-08-22 22:41 UTC (permalink / raw)
To: davem, netdev; +Cc: adapter_linux_open_src_team, Rasesh Mody, Gurunatha Karaje
In-Reply-To: <1314052869-7407-1-git-send-email-rmody@brocade.com>
Change details:
- Use available bnad_dim_timer_stop macro in bnad_set_coalesce.
- Add tx_skb counters and NAPI debug counters to ethtool stats.
- Add rlb stats strings to bnad_net_stats_strings{} array. rlb_stats field
was added to struct bfi_enet_stats {} but the corresponding name structure
array for ethtool was not initialized with right strings, even though the
actual name structure array got expanded. This caused a NULL pointer
violation and a crash when doing ehtool -S <if_name>.
- While setting the ring parameter restore the rx, vlan configuration and
set rx mode
- Indentation fix
Signed-off-by: Gurunatha Karaje <gkaraje@brocade.com>
Signed-off-by: Rasesh Mody <rmody@brocade.com>
---
drivers/net/ethernet/brocade/bna/bnad.c | 8 +-
drivers/net/ethernet/brocade/bna/bnad.h | 10 ++-
drivers/net/ethernet/brocade/bna/bnad_ethtool.c | 88 +++++++++++++++++++----
3 files changed, 84 insertions(+), 22 deletions(-)
diff --git a/drivers/net/ethernet/brocade/bna/bnad.c b/drivers/net/ethernet/brocade/bna/bnad.c
index 4171c6f..2f808c7 100644
--- a/drivers/net/ethernet/brocade/bna/bnad.c
+++ b/drivers/net/ethernet/brocade/bna/bnad.c
@@ -2015,7 +2015,7 @@ bnad_rx_coalescing_timeo_set(struct bnad *bnad)
/*
* Called with bnad->bna_lock held
*/
-static int
+int
bnad_mac_addr_set_locked(struct bnad *bnad, u8 *mac_addr)
{
int ret;
@@ -2035,7 +2035,7 @@ bnad_mac_addr_set_locked(struct bnad *bnad, u8 *mac_addr)
}
/* Should be called with conf_lock held */
-static int
+int
bnad_enable_default_bcast(struct bnad *bnad)
{
struct bnad_rx_info *rx_info = &bnad->rx_info[0];
@@ -2061,7 +2061,7 @@ bnad_enable_default_bcast(struct bnad *bnad)
}
/* Called with mutex_lock(&bnad->conf_mutex) held */
-static void
+void
bnad_restore_vlans(struct bnad *bnad, u32 rx_id)
{
u16 vid;
@@ -2812,7 +2812,7 @@ bnad_get_stats64(struct net_device *netdev, struct rtnl_link_stats64 *stats)
return stats;
}
-static void
+void
bnad_set_rx_mode(struct net_device *netdev)
{
struct bnad *bnad = netdev_priv(netdev);
diff --git a/drivers/net/ethernet/brocade/bna/bnad.h b/drivers/net/ethernet/brocade/bna/bnad.h
index b03e3a9..b31b893 100644
--- a/drivers/net/ethernet/brocade/bna/bnad.h
+++ b/drivers/net/ethernet/brocade/bna/bnad.h
@@ -328,6 +328,12 @@ extern u32 bnad_rxqs_per_cq;
*/
extern u32 *cna_get_firmware_buf(struct pci_dev *pdev);
/* Netdev entry point prototypes */
+extern void bnad_set_rx_mode(struct net_device *netdev);
+extern struct net_device_stats *bnad_get_netdev_stats(
+ struct net_device *netdev);
+extern int bnad_mac_addr_set_locked(struct bnad *bnad, u8 *mac_addr);
+extern int bnad_enable_default_bcast(struct bnad *bnad);
+extern void bnad_restore_vlans(struct bnad *bnad, u32 rx_id);
extern void bnad_set_ethtool_ops(struct net_device *netdev);
/* Configuration & setup */
@@ -366,10 +372,6 @@ extern void bnad_netdev_hwstats_fill(struct bnad *bnad,
} \
}
-#define bnad_dim_timer_running(_bnad) \
- (((_bnad)->cfg_flags & BNAD_CF_DIM_ENABLED) && \
- (test_bit(BNAD_RF_DIM_TIMER_RUNNING, &((_bnad)->run_flags))))
-
/*
* Stops the DIM timer
* Called with bnad->bna_lock held
diff --git a/drivers/net/ethernet/brocade/bna/bnad_ethtool.c b/drivers/net/ethernet/brocade/bna/bnad_ethtool.c
index 1c19dce..1199f01 100644
--- a/drivers/net/ethernet/brocade/bna/bnad_ethtool.c
+++ b/drivers/net/ethernet/brocade/bna/bnad_ethtool.c
@@ -75,14 +75,25 @@ static char *bnad_net_stats_strings[BNAD_ETHTOOL_STATS_NUM] = {
"tcpcsum_offload",
"udpcsum_offload",
"csum_help",
- "csum_help_err",
+ "tx_skb_too_short",
+ "tx_skb_stopping",
+ "tx_skb_max_vectors",
+ "tx_skb_mss_too_long",
+ "tx_skb_tso_too_short",
+ "tx_skb_tso_prepare",
+ "tx_skb_non_tso_too_long",
+ "tx_skb_tcp_hdr",
+ "tx_skb_udp_hdr",
+ "tx_skb_csum_err",
+ "tx_skb_headlen_too_long",
+ "tx_skb_headlen_zero",
+ "tx_skb_frag_zero",
+ "tx_skb_len_mismatch",
"hw_stats_updates",
- "netif_rx_schedule",
- "netif_rx_complete",
"netif_rx_dropped",
"link_toggle",
- "cee_up",
+ "cee_toggle",
"rxp_info_alloc_failed",
"mbox_intr_disabled",
@@ -201,6 +212,20 @@ static char *bnad_net_stats_strings[BNAD_ETHTOOL_STATS_NUM] = {
"rad_rx_bcast_vlan",
"rad_rx_drops",
+ "rlb_rad_rx_frames",
+ "rlb_rad_rx_octets",
+ "rlb_rad_rx_vlan_frames",
+ "rlb_rad_rx_ucast",
+ "rlb_rad_rx_ucast_octets",
+ "rlb_rad_rx_ucast_vlan",
+ "rlb_rad_rx_mcast",
+ "rlb_rad_rx_mcast_octets",
+ "rlb_rad_rx_mcast_vlan",
+ "rlb_rad_rx_bcast",
+ "rlb_rad_rx_bcast_octets",
+ "rlb_rad_rx_bcast_vlan",
+ "rlb_rad_rx_drops",
+
"fc_rx_ucast_octets",
"fc_rx_ucast",
"fc_rx_ucast_vlan",
@@ -321,7 +346,6 @@ bnad_set_coalesce(struct net_device *netdev, struct ethtool_coalesce *coalesce)
{
struct bnad *bnad = netdev_priv(netdev);
unsigned long flags;
- int dim_timer_del = 0;
if (coalesce->rx_coalesce_usecs == 0 ||
coalesce->rx_coalesce_usecs >
@@ -348,14 +372,7 @@ bnad_set_coalesce(struct net_device *netdev, struct ethtool_coalesce *coalesce)
} else {
if (bnad->cfg_flags & BNAD_CF_DIM_ENABLED) {
bnad->cfg_flags &= ~BNAD_CF_DIM_ENABLED;
- dim_timer_del = bnad_dim_timer_running(bnad);
- if (dim_timer_del) {
- clear_bit(BNAD_RF_DIM_TIMER_RUNNING,
- &bnad->run_flags);
- spin_unlock_irqrestore(&bnad->bna_lock, flags);
- del_timer_sync(&bnad->dim_timer);
- spin_lock_irqsave(&bnad->bna_lock, flags);
- }
+ bnad_dim_timer_stop(bnad, flags);
bnad_rx_coalescing_timeo_set(bnad);
}
}
@@ -407,6 +424,7 @@ bnad_set_ringparam(struct net_device *netdev,
{
int i, current_err, err = 0;
struct bnad *bnad = netdev_priv(netdev);
+ unsigned long flags;
mutex_lock(&bnad->conf_mutex);
if (ringparam->rx_pending == bnad->rxq_depth &&
@@ -430,6 +448,11 @@ bnad_set_ringparam(struct net_device *netdev,
if (ringparam->rx_pending != bnad->rxq_depth) {
bnad->rxq_depth = ringparam->rx_pending;
+ if (!netif_running(netdev)) {
+ mutex_unlock(&bnad->conf_mutex);
+ return 0;
+ }
+
for (i = 0; i < bnad->num_rx; i++) {
if (!bnad->rx_info[i].rx)
continue;
@@ -437,10 +460,26 @@ bnad_set_ringparam(struct net_device *netdev,
current_err = bnad_setup_rx(bnad, i);
if (current_err && !err)
err = current_err;
+ if (!err)
+ bnad_restore_vlans(bnad, i);
+ }
+
+ if (!err && bnad->rx_info[0].rx) {
+ /* restore rx configuration */
+ bnad_enable_default_bcast(bnad);
+ spin_lock_irqsave(&bnad->bna_lock, flags);
+ bnad_mac_addr_set_locked(bnad, netdev->dev_addr);
+ spin_unlock_irqrestore(&bnad->bna_lock, flags);
+ bnad_set_rx_mode(netdev);
}
}
if (ringparam->tx_pending != bnad->txq_depth) {
bnad->txq_depth = ringparam->tx_pending;
+ if (!netif_running(netdev)) {
+ mutex_unlock(&bnad->conf_mutex);
+ return 0;
+ }
+
for (i = 0; i < bnad->num_tx; i++) {
if (!bnad->tx_info[i].tx)
continue;
@@ -578,6 +617,16 @@ bnad_get_strings(struct net_device *netdev, u32 stringset, u8 * string)
sprintf(string, "cq%d_hw_producer_index",
q_num);
string += ETH_GSTRING_LEN;
+ sprintf(string, "cq%d_intr", q_num);
+ string += ETH_GSTRING_LEN;
+ sprintf(string, "cq%d_poll", q_num);
+ string += ETH_GSTRING_LEN;
+ sprintf(string, "cq%d_schedule", q_num);
+ string += ETH_GSTRING_LEN;
+ sprintf(string, "cq%d_keep_poll", q_num);
+ string += ETH_GSTRING_LEN;
+ sprintf(string, "cq%d_complete", q_num);
+ string += ETH_GSTRING_LEN;
q_num++;
}
}
@@ -660,7 +709,7 @@ static int
bnad_get_stats_count_locked(struct net_device *netdev)
{
struct bnad *bnad = netdev_priv(netdev);
- int i, j, count, rxf_active_num = 0, txf_active_num = 0;
+ int i, j, count = 0, rxf_active_num = 0, txf_active_num = 0;
u32 bmap;
bmap = bna_tx_rid_mask(&bnad->bna);
@@ -718,6 +767,17 @@ bnad_per_q_stats_fill(struct bnad *bnad, u64 *buf, int bi)
buf[bi++] = 0; /* ccb->consumer_index */
buf[bi++] = *(bnad->rx_info[i].rx_ctrl[j].
ccb->hw_producer_index);
+
+ buf[bi++] = bnad->rx_info[i].
+ rx_ctrl[j].rx_intr_ctr;
+ buf[bi++] = bnad->rx_info[i].
+ rx_ctrl[j].rx_poll_ctr;
+ buf[bi++] = bnad->rx_info[i].
+ rx_ctrl[j].rx_schedule;
+ buf[bi++] = bnad->rx_info[i].
+ rx_ctrl[j].rx_keep_poll;
+ buf[bi++] = bnad->rx_info[i].
+ rx_ctrl[j].rx_complete;
}
}
for (i = 0; i < bnad->num_rx; i++) {
--
1.7.1
^ permalink raw reply related
* [net-next 06/12] bna: Initialization and Locking Fix
From: Rasesh Mody @ 2011-08-22 22:41 UTC (permalink / raw)
To: davem, netdev; +Cc: adapter_linux_open_src_team, Rasesh Mody, Gurunatha Karaje
In-Reply-To: <1314052869-7407-1-git-send-email-rmody@brocade.com>
Change details:
- Initialize rx_id to 0 for bnad_cleanup_rx
- Return -ENOMEM in case if bna_rx_create fails
- Count the Rx buffer allocation failures in bnad_alloc_n_post_rxbufs()
- Remove unnecessary initialization of using_dac to false in bnad_pci_probe
- Release lock if error while doing bna_num_txq_set in bnad_pci_probe
- Release all the locks while doing free_netdev
Signed-off-by: Gurunatha Karaje <gkaraje@brocade.com>
Signed-off-by: Rasesh Mody <rmody@brocade.com>
---
drivers/net/ethernet/brocade/bna/bna_hw_defs.h | 1 +
drivers/net/ethernet/brocade/bna/bnad.c | 15 ++++++++++++---
2 files changed, 13 insertions(+), 3 deletions(-)
diff --git a/drivers/net/ethernet/brocade/bna/bna_hw_defs.h b/drivers/net/ethernet/brocade/bna/bna_hw_defs.h
index 07bb792..7ecdca5 100644
--- a/drivers/net/ethernet/brocade/bna/bna_hw_defs.h
+++ b/drivers/net/ethernet/brocade/bna/bna_hw_defs.h
@@ -99,6 +99,7 @@
(_bna)->bits.error_status_bits = (__HFN_INT_ERR_MASK); \
(_bna)->bits.error_mask_bits = (__HFN_INT_ERR_MASK); \
(_bna)->bits.halt_status_bits = __HFN_INT_LL_HALT; \
+ (_bna)->bits.halt_mask_bits = __HFN_INT_LL_HALT; \
}
#define ct2_reg_addr_init(_bna, _pcidev) \
diff --git a/drivers/net/ethernet/brocade/bna/bnad.c b/drivers/net/ethernet/brocade/bna/bnad.c
index b5fb265..4171c6f 100644
--- a/drivers/net/ethernet/brocade/bna/bnad.c
+++ b/drivers/net/ethernet/brocade/bna/bnad.c
@@ -389,6 +389,7 @@ bnad_alloc_n_post_rxbufs(struct bnad *bnad, struct bna_rcb *rcb)
rcb->rxq->buffer_size);
if (unlikely(!skb)) {
BNAD_UPDATE_CTR(bnad, rxbuf_alloc_failed);
+ rcb->rxq->rxbuf_alloc_failed++;
goto finishing;
}
unmap_array[unmap_prod].skb = skb;
@@ -1880,6 +1881,7 @@ bnad_cleanup_rx(struct bnad *bnad, u32 rx_id)
spin_unlock_irqrestore(&bnad->bna_lock, flags);
rx_info->rx = NULL;
+ rx_info->rx_id = 0;
bnad_rx_res_free(bnad, res_info);
}
@@ -1935,8 +1937,10 @@ bnad_setup_rx(struct bnad *bnad, u32 rx_id)
rx = bna_rx_create(&bnad->bna, bnad, rx_config, &rx_cbfn, res_info,
rx_info);
spin_unlock_irqrestore(&bnad->bna_lock, flags);
- if (!rx)
+ if (!rx) {
+ err = -ENOMEM;
goto err_return;
+ }
rx_info->rx = rx;
/*
@@ -3192,7 +3196,7 @@ static int __devinit
bnad_pci_probe(struct pci_dev *pdev,
const struct pci_device_id *pcidev_id)
{
- bool using_dac = false;
+ bool using_dac;
int err;
struct bnad *bnad;
struct bna *bna;
@@ -3315,6 +3319,11 @@ bnad_pci_probe(struct pci_dev *pdev,
bna_num_rxp_set(bna, BNAD_NUM_RXP + 1))
err = -EIO;
}
+ spin_unlock_irqrestore(&bnad->bna_lock, flags);
+ if (err)
+ goto disable_ioceth;
+
+ spin_lock_irqsave(&bnad->bna_lock, flags);
bna_mod_res_req(&bnad->bna, &bnad->mod_res_info[0]);
spin_unlock_irqrestore(&bnad->bna_lock, flags);
@@ -3368,9 +3377,9 @@ drv_uninit:
bnad_uninit(bnad);
pci_uninit:
bnad_pci_uninit(pdev);
+free_netdev:
mutex_unlock(&bnad->conf_mutex);
bnad_lock_uninit(bnad);
-free_netdev:
free_netdev(netdev);
return err;
}
--
1.7.1
^ permalink raw reply related
* [net-next 05/12] bna: Formatting and Code Cleanup
From: Rasesh Mody @ 2011-08-22 22:41 UTC (permalink / raw)
To: davem, netdev; +Cc: adapter_linux_open_src_team, Rasesh Mody, Gurunatha Karaje
In-Reply-To: <1314052869-7407-1-git-send-email-rmody@brocade.com>
Change details:
- Print log messages when running with reduced number of MSI-X vectors
and when defaulting to INTx mode.
- Remove BUG_ONs and header file inclusion that are not needed
- Comments addition/cleanup
- Unused code cleanup
- Add New Line to Print msg in bfa_sm_fault
- Formatting fix
Signed-off-by: Gurunatha Karaje <gkaraje@brocade.com>
Signed-off-by: Rasesh Mody <rmody@brocade.com>
---
drivers/net/ethernet/brocade/bna/bfa_cee.c | 2 -
.../net/ethernet/brocade/bna/bfa_defs_mfg_comm.h | 1 -
drivers/net/ethernet/brocade/bna/bfi.h | 46 --------------------
drivers/net/ethernet/brocade/bna/bna.h | 18 +++-----
drivers/net/ethernet/brocade/bna/bna_types.h | 1 -
drivers/net/ethernet/brocade/bna/bnad.c | 46 ++++++--------------
drivers/net/ethernet/brocade/bna/bnad.h | 13 +++---
drivers/net/ethernet/brocade/bna/cna.h | 11 ++---
8 files changed, 31 insertions(+), 107 deletions(-)
diff --git a/drivers/net/ethernet/brocade/bna/bfa_cee.c b/drivers/net/ethernet/brocade/bna/bfa_cee.c
index b45b8eb..8e62718 100644
--- a/drivers/net/ethernet/brocade/bna/bfa_cee.c
+++ b/drivers/net/ethernet/brocade/bna/bfa_cee.c
@@ -16,8 +16,6 @@
* www.brocade.com
*/
-#include "bfa_defs_cna.h"
-#include "cna.h"
#include "bfa_cee.h"
#include "bfi_cna.h"
#include "bfa_ioc.h"
diff --git a/drivers/net/ethernet/brocade/bna/bfa_defs_mfg_comm.h b/drivers/net/ethernet/brocade/bna/bfa_defs_mfg_comm.h
index 7ddd16f..7e5df90 100644
--- a/drivers/net/ethernet/brocade/bna/bfa_defs_mfg_comm.h
+++ b/drivers/net/ethernet/brocade/bna/bfa_defs_mfg_comm.h
@@ -18,7 +18,6 @@
#ifndef __BFA_DEFS_MFG_COMM_H__
#define __BFA_DEFS_MFG_COMM_H__
-#include "cna.h"
#include "bfa_defs.h"
/**
diff --git a/drivers/net/ethernet/brocade/bna/bfi.h b/drivers/net/ethernet/brocade/bna/bfi.h
index 19654cc..4e04c14 100644
--- a/drivers/net/ethernet/brocade/bna/bfi.h
+++ b/drivers/net/ethernet/brocade/bna/bfi.h
@@ -73,20 +73,6 @@ struct bfi_mhdr {
****************************************************************************
*/
-#define BFI_SGE_INLINE 1
-#define BFI_SGE_INLINE_MAX (BFI_SGE_INLINE + 1)
-
-/**
- * SG Flags
- */
-enum {
- BFI_SGE_DATA = 0, /*!< data address, not last */
- BFI_SGE_DATA_CPL = 1, /*!< data addr, last in current page */
- BFI_SGE_DATA_LAST = 3, /*!< data address, last */
- BFI_SGE_LINK = 2, /*!< link address */
- BFI_SGE_PGDLEN = 2, /*!< cumulative data length for page */
-};
-
/**
* DMA addresses
*/
@@ -97,33 +83,6 @@ union bfi_addr_u {
} a32;
};
-/**
- * Scatter Gather Element
- */
-struct bfi_sge {
-#ifdef __BIGENDIAN
- u32 flags:2,
- rsvd:2,
- sg_len:28;
-#else
- u32 sg_len:28,
- rsvd:2,
- flags:2;
-#endif
- union bfi_addr_u sga;
-};
-
-/**
- * Scatter Gather Page
- */
-#define BFI_SGPG_DATA_SGES 7
-#define BFI_SGPG_SGES_MAX (BFI_SGPG_DATA_SGES + 1)
-#define BFI_SGPG_RSVD_WD_LEN 8
-struct bfi_sgpg {
- struct bfi_sge sges[BFI_SGPG_SGES_MAX];
- u32 rsvd[BFI_SGPG_RSVD_WD_LEN];
-};
-
/*
* Large Message structure - 128 Bytes size Msgs
*/
@@ -131,11 +90,6 @@ struct bfi_sgpg {
#define BFI_LMSG_PL_WSZ \
((BFI_LMSG_SZ - sizeof(struct bfi_mhdr)) / 4)
-struct bfi_msg {
- struct bfi_mhdr mhdr;
- u32 pl[BFI_LMSG_PL_WSZ];
-};
-
/**
* Mailbox message structure
*/
diff --git a/drivers/net/ethernet/brocade/bna/bna.h b/drivers/net/ethernet/brocade/bna/bna.h
index 2a587c5..3a6e790 100644
--- a/drivers/net/ethernet/brocade/bna/bna.h
+++ b/drivers/net/ethernet/brocade/bna/bna.h
@@ -10,12 +10,17 @@
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* General Public License for more details.
*/
+/*
+ * Copyright (c) 2005-2011 Brocade Communications Systems, Inc.
+ * All rights reserved
+ * www.brocade.com
+ */
#ifndef __BNA_H__
#define __BNA_H__
-#include "bfa_cs.h"
+#include "bfa_defs.h"
#include "bfa_ioc.h"
-#include "cna.h"
+#include "bfi_enet.h"
#include "bna_types.h"
extern const u32 bna_napi_dim_vector[][BNA_BIAS_T_MAX];
@@ -395,12 +400,8 @@ void bna_mod_init(struct bna *bna, struct bna_res_info *res_info);
void bna_uninit(struct bna *bna);
int bna_num_txq_set(struct bna *bna, int num_txq);
int bna_num_rxp_set(struct bna *bna, int num_rxp);
-void bna_stats_get(struct bna *bna);
-void bna_get_perm_mac(struct bna *bna, u8 *mac);
void bna_hw_stats_get(struct bna *bna);
-/* APIs for Rx */
-
/* APIs for RxF */
struct bna_mac *bna_ucam_mod_mac_get(struct bna_ucam_mod *ucam_mod);
void bna_ucam_mod_mac_put(struct bna_ucam_mod *ucam_mod,
@@ -521,11 +522,6 @@ bna_rx_mode_set(struct bna_rx *rx, enum bna_rxmode rxmode,
void bna_rx_vlan_add(struct bna_rx *rx, int vlan_id);
void bna_rx_vlan_del(struct bna_rx *rx, int vlan_id);
void bna_rx_vlanfilter_enable(struct bna_rx *rx);
-void bna_rx_hds_enable(struct bna_rx *rx, struct bna_hds_config *hds_config,
- void (*cbfn)(struct bnad *, struct bna_rx *));
-void bna_rx_hds_disable(struct bna_rx *rx,
- void (*cbfn)(struct bnad *, struct bna_rx *));
-
/**
* ENET
*/
diff --git a/drivers/net/ethernet/brocade/bna/bna_types.h b/drivers/net/ethernet/brocade/bna/bna_types.h
index 8a6da0c..59417b1 100644
--- a/drivers/net/ethernet/brocade/bna/bna_types.h
+++ b/drivers/net/ethernet/brocade/bna/bna_types.h
@@ -21,7 +21,6 @@
#include "cna.h"
#include "bna_hw_defs.h"
#include "bfa_cee.h"
-#include "bfi_enet.h"
#include "bfa_msgq.h"
/**
diff --git a/drivers/net/ethernet/brocade/bna/bnad.c b/drivers/net/ethernet/brocade/bna/bnad.c
index 802d3f1..b5fb265 100644
--- a/drivers/net/ethernet/brocade/bna/bnad.c
+++ b/drivers/net/ethernet/brocade/bna/bnad.c
@@ -382,10 +382,9 @@ bnad_alloc_n_post_rxbufs(struct bnad *bnad, struct bna_rcb *rcb)
BNA_RXQ_QPGE_PTR_GET(unmap_prod, rcb->sw_qpt, rxent, wi_range);
while (to_alloc--) {
- if (!wi_range) {
+ if (!wi_range)
BNA_RXQ_QPGE_PTR_GET(unmap_prod, rcb->sw_qpt, rxent,
wi_range);
- }
skb = netdev_alloc_skb_ip_align(bnad->netdev,
rcb->rxq->buffer_size);
if (unlikely(!skb)) {
@@ -547,27 +546,6 @@ next:
}
static void
-bnad_disable_rx_irq(struct bnad *bnad, struct bna_ccb *ccb)
-{
- if (unlikely(!test_bit(BNAD_RXQ_STARTED, &ccb->rcb[0]->flags)))
- return;
-
- bna_ib_coalescing_timer_set(ccb->i_dbell, 0);
- bna_ib_ack(ccb->i_dbell, 0);
-}
-
-static void
-bnad_enable_rx_irq(struct bnad *bnad, struct bna_ccb *ccb)
-{
- unsigned long flags;
-
- /* Because of polling context */
- spin_lock_irqsave(&bnad->bna_lock, flags);
- bnad_enable_rx_irq_unsafe(ccb);
- spin_unlock_irqrestore(&bnad->bna_lock, flags);
-}
-
-static void
bnad_netif_rx_schedule_poll(struct bnad *bnad, struct bna_ccb *ccb)
{
struct bnad_rx_ctrl *rx_ctrl = (struct bnad_rx_ctrl *)(ccb->ctrl);
@@ -1667,7 +1645,7 @@ bnad_napi_poll_rx(struct napi_struct *napi, int budget)
return rcvd;
poll_exit:
- napi_complete((napi));
+ napi_complete(napi);
rx_ctrl->rx_complete++;
@@ -2078,15 +2056,13 @@ bnad_enable_default_bcast(struct bnad *bnad)
return 0;
}
-/* Called with bnad_conf_lock() held */
+/* Called with mutex_lock(&bnad->conf_mutex) held */
static void
bnad_restore_vlans(struct bnad *bnad, u32 rx_id)
{
u16 vid;
unsigned long flags;
- BUG_ON(!(VLAN_N_VID == BFI_ENET_VLAN_ID_MAX));
-
for_each_set_bit(vid, bnad->active_vlans, VLAN_N_VID) {
spin_lock_irqsave(&bnad->bna_lock, flags);
bna_rx_vlan_add(bnad->rx_info[rx_id].rx, vid);
@@ -2195,9 +2171,6 @@ bnad_tso_prepare(struct bnad *bnad, struct sk_buff *skb)
{
int err;
- /* SKB_GSO_TCPV4 and SKB_GSO_TCPV6 is defined since 2.6.18. */
- BUG_ON(!(skb_shinfo(skb)->gso_type == SKB_GSO_TCPV4 ||
- skb_shinfo(skb)->gso_type == SKB_GSO_TCPV6));
if (skb_header_cloned(skb)) {
err = pskb_expand_head(skb, 0, 0, GFP_ATOMIC);
if (err) {
@@ -2224,7 +2197,6 @@ bnad_tso_prepare(struct bnad *bnad, struct sk_buff *skb)
} else {
struct ipv6hdr *ipv6h = ipv6_hdr(skb);
- BUG_ON(!(skb->protocol == htons(ETH_P_IPV6)));
ipv6h->payload_len = 0;
tcp_hdr(skb)->check =
~csum_ipv6_magic(&ipv6h->saddr, &ipv6h->daddr, 0,
@@ -2375,6 +2347,8 @@ bnad_enable_msix(struct bnad *bnad)
ret = pci_enable_msix(bnad->pcidev, bnad->msix_table, bnad->msix_num);
if (ret > 0) {
/* Not enough MSI-X vectors. */
+ pr_warn("BNA: %d MSI-X vectors allocated < %d requested\n",
+ ret, bnad->msix_num);
spin_lock_irqsave(&bnad->bna_lock, flags);
/* ret = #of vectors that we got */
@@ -2403,6 +2377,7 @@ bnad_enable_msix(struct bnad *bnad)
return;
intx_mode:
+ pr_warn("BNA: MSI-X enable failed - operating in INTx mode\n");
kfree(bnad->msix_table);
bnad->msix_table = NULL;
@@ -2564,7 +2539,7 @@ bnad_start_xmit(struct sk_buff *skb, struct net_device *netdev)
/*
* Takes care of the Tx that is scheduled between clearing the flag
- * and the netif_stop_all_queue() call.
+ * and the netif_tx_stop_all_queues() call.
*/
if (unlikely(!test_bit(BNAD_TXQ_TX_STARTED, &tcb->flags))) {
dev_kfree_skb(skb);
@@ -2617,7 +2592,6 @@ bnad_start_xmit(struct sk_buff *skb, struct net_device *netdev)
txq_prod = tcb->producer_index;
BNA_TXQ_QPGE_PTR_GET(txq_prod, tcb->sw_qpt, txqent, wi_range);
- BUG_ON(!(wi_range <= tcb->q_depth));
txqent->hdr.wi.reserved = 0;
txqent->hdr.wi.num_vectors = vectors;
@@ -3022,6 +2996,12 @@ bnad_netpoll(struct net_device *netdev)
bnad_isr(bnad->pcidev->irq, netdev);
bna_intx_enable(&bnad->bna, curr_mask);
} else {
+ /*
+ * Tx processing may happen in sending context, so no need
+ * to explicitly process completions here
+ */
+
+ /* Rx processing */
for (i = 0; i < bnad->num_rx; i++) {
rx_info = &bnad->rx_info[i];
if (!rx_info->rx)
diff --git a/drivers/net/ethernet/brocade/bna/bnad.h b/drivers/net/ethernet/brocade/bna/bnad.h
index 8a31882..b03e3a9 100644
--- a/drivers/net/ethernet/brocade/bna/bnad.h
+++ b/drivers/net/ethernet/brocade/bna/bnad.h
@@ -65,8 +65,6 @@ struct bnad_rx_ctrl {
#define BNAD_RXMODE_PROMISC_DEFAULT BNA_RXMODE_PROMISC
-#define BNAD_GET_TX_ID(_skb) (0)
-
/*
* GLOBAL #defines (CONSTANTS)
*/
@@ -152,7 +150,6 @@ struct bnad_drv_stats {
u64 tcpcsum_offload;
u64 udpcsum_offload;
u64 csum_help;
- u64 csum_help_err;
u64 tx_skb_too_short;
u64 tx_skb_stopping;
u64 tx_skb_max_vectors;
@@ -169,13 +166,10 @@ struct bnad_drv_stats {
u64 tx_skb_len_mismatch;
u64 hw_stats_updates;
- u64 netif_rx_schedule;
- u64 netif_rx_complete;
u64 netif_rx_dropped;
u64 link_toggle;
u64 cee_toggle;
- u64 cee_up;
u64 rxp_info_alloc_failed;
u64 mbox_intr_disabled;
@@ -375,6 +369,13 @@ extern void bnad_netdev_hwstats_fill(struct bnad *bnad,
#define bnad_dim_timer_running(_bnad) \
(((_bnad)->cfg_flags & BNAD_CF_DIM_ENABLED) && \
(test_bit(BNAD_RF_DIM_TIMER_RUNNING, &((_bnad)->run_flags))))
+
+/*
+ * Stops the DIM timer
+ * Called with bnad->bna_lock held
+ * Implemented as macro, since we want to use
+ * the correct flags(on stack) while unlocking.
+ */
#define bnad_dim_timer_stop(_bnad, _flags) \
do { \
int to_del = 0; \
diff --git a/drivers/net/ethernet/brocade/bna/cna.h b/drivers/net/ethernet/brocade/bna/cna.h
index 50fce15..cb48742 100644
--- a/drivers/net/ethernet/brocade/bna/cna.h
+++ b/drivers/net/ethernet/brocade/bna/cna.h
@@ -21,21 +21,18 @@
#include <linux/kernel.h>
#include <linux/types.h>
+#include <linux/mutex.h>
#include <linux/pci.h>
#include <linux/delay.h>
#include <linux/bitops.h>
#include <linux/timer.h>
#include <linux/interrupt.h>
+#include <linux/if_vlan.h>
#include <linux/if_ether.h>
-#include <asm/page.h>
-#include <asm/io.h>
-#include <asm/string.h>
-
-#include <linux/list.h>
#define bfa_sm_fault(__event) do { \
- pr_err("SM Assertion failure: %s: %d: event = %d", __FILE__, __LINE__, \
- __event); \
+ pr_err("SM Assertion failure: %s: %d: event = %d\n", \
+ __FILE__, __LINE__, __event); \
} while (0)
extern char bfa_version[];
--
1.7.1
^ permalink raw reply related
* [net-next 04/12] bna: TX Path and RX Path Changes
From: Rasesh Mody @ 2011-08-22 22:41 UTC (permalink / raw)
To: davem, netdev; +Cc: adapter_linux_open_src_team, Rasesh Mody, Gurunatha Karaje
In-Reply-To: <1314052869-7407-1-git-send-email-rmody@brocade.com>
Change details:
- Add BNAD_PCI_UNMAP_SKB macro to unmap skb from transmit path. Add more
checks for illegal skbs in transmit path. Add tx_skb counters for dropped
skbs.
- Disable and enable interrupts from the same polling context to prevent
reordering in Rx path.
- Add Rx NAPI debug counters.
- Make NAPI budget check more generic.
- Add a macro bnad_dim_timer_stop for DIM(Dynamic Interrupt Moderation)
timer stop.
- Handle reduced MSI-X vectors case in bnad_enable_msix.
- Check for single frame TSO skbs and send them out as non-TSO.
- Put memory barrier after bna_txq_prod_indx_doorbell().
Signed-off-by: Gurunatha Karaje <gkaraje@brocade.com>
Signed-off-by: Rasesh Mody <rmody@brocade.com>
---
drivers/net/ethernet/brocade/bna/bnad.c | 259 +++++++++++++++++++------------
drivers/net/ethernet/brocade/bna/bnad.h | 35 ++++-
2 files changed, 195 insertions(+), 99 deletions(-)
diff --git a/drivers/net/ethernet/brocade/bna/bnad.c b/drivers/net/ethernet/brocade/bna/bnad.c
index 095eac9..802d3f1 100644
--- a/drivers/net/ethernet/brocade/bna/bnad.c
+++ b/drivers/net/ethernet/brocade/bna/bnad.c
@@ -74,6 +74,24 @@ do { \
#define BNAD_TXRX_SYNC_MDELAY 250 /* 250 msecs */
+#define BNAD_PCI_UNMAP_SKB(_pdev, _array, _index, _depth, _skb, _frag) \
+{ \
+ int j; \
+ (_array)[_index].skb = NULL; \
+ dma_unmap_single(_pdev, dma_unmap_addr(&(_array)[_index], dma_addr), \
+ skb_headlen(_skb), DMA_TO_DEVICE); \
+ dma_unmap_addr_set(&(_array)[_index], dma_addr, 0); \
+ BNA_QE_INDX_ADD(_index, 1, _depth); \
+ for (j = 0; j < (_frag); j++) { \
+ prefetch(&(_array)[(_index) + 1]); \
+ dma_unmap_page(_pdev, dma_unmap_addr(&(_array)[_index], \
+ dma_addr), \
+ skb_shinfo(_skb)->frags[j].size, DMA_TO_DEVICE); \
+ dma_unmap_addr_set(&(_array)[_index], dma_addr, 0); \
+ BNA_QE_INDX_ADD(_index, 1, _depth); \
+ } \
+}
+
/*
* Reinitialize completions in CQ, once Rx is taken down
*/
@@ -169,7 +187,6 @@ bnad_free_txbufs(struct bnad *bnad,
struct bnad_unmap_q *unmap_q = tcb->unmap_q;
struct bnad_skb_unmap *unmap_array;
struct sk_buff *skb;
- int i;
/*
* Just return if TX is stopped. This check is useful
@@ -195,32 +212,14 @@ bnad_free_txbufs(struct bnad *bnad,
while (wis) {
skb = unmap_array[unmap_cons].skb;
- unmap_array[unmap_cons].skb = NULL;
-
sent_packets++;
sent_bytes += skb->len;
wis -= BNA_TXQ_WI_NEEDED(1 + skb_shinfo(skb)->nr_frags);
- dma_unmap_single(&bnad->pcidev->dev,
- dma_unmap_addr(&unmap_array[unmap_cons],
- dma_addr), skb_headlen(skb),
- DMA_TO_DEVICE);
- dma_unmap_addr_set(&unmap_array[unmap_cons], dma_addr, 0);
- BNA_QE_INDX_ADD(unmap_cons, 1, unmap_q->q_depth);
+ BNAD_PCI_UNMAP_SKB(&bnad->pcidev->dev, unmap_array, unmap_cons,
+ unmap_q->q_depth, skb,
+ skb_shinfo(skb)->nr_frags);
- prefetch(&unmap_array[unmap_cons + 1]);
- for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
- prefetch(&unmap_array[unmap_cons + 1]);
-
- dma_unmap_page(&bnad->pcidev->dev,
- dma_unmap_addr(&unmap_array[unmap_cons],
- dma_addr),
- skb_shinfo(skb)->frags[i].size,
- DMA_TO_DEVICE);
- dma_unmap_addr_set(&unmap_array[unmap_cons], dma_addr,
- 0);
- BNA_QE_INDX_ADD(unmap_cons, 1, unmap_q->q_depth);
- }
dev_kfree_skb_any(skb);
}
@@ -536,7 +535,8 @@ next:
BNA_QE_INDX_ADD(ccb->producer_index, wis, ccb->q_depth);
if (likely(test_bit(BNAD_RXQ_STARTED, &ccb->rcb[0]->flags)))
- bna_ib_ack(ccb->i_dbell, packets);
+ bna_ib_ack_disable_irq(ccb->i_dbell, packets);
+
bnad_refill_rxq(bnad, ccb->rcb[0]);
if (ccb->rcb[1])
bnad_refill_rxq(bnad, ccb->rcb[1]);
@@ -574,10 +574,9 @@ bnad_netif_rx_schedule_poll(struct bnad *bnad, struct bna_ccb *ccb)
struct napi_struct *napi = &rx_ctrl->napi;
if (likely(napi_schedule_prep(napi))) {
- bnad_disable_rx_irq(bnad, ccb);
__napi_schedule(napi);
+ rx_ctrl->rx_schedule++;
}
- BNAD_UPDATE_CTR(bnad, netif_rx_schedule);
}
/* MSIX Rx Path Handler */
@@ -586,8 +585,10 @@ bnad_msix_rx(int irq, void *data)
{
struct bna_ccb *ccb = (struct bna_ccb *)data;
- if (ccb)
+ if (ccb) {
+ ((struct bnad_rx_ctrl *)(ccb->ctrl))->rx_intr_ctr++;
bnad_netif_rx_schedule_poll(ccb->bnad, ccb);
+ }
return IRQ_HANDLED;
}
@@ -1656,22 +1657,23 @@ bnad_napi_poll_rx(struct napi_struct *napi, int budget)
struct bnad *bnad = rx_ctrl->bnad;
int rcvd = 0;
+ rx_ctrl->rx_poll_ctr++;
if (!netif_carrier_ok(bnad->netdev))
goto poll_exit;
rcvd = bnad_poll_cq(bnad, rx_ctrl->ccb, budget);
- if (rcvd == budget)
+ if (rcvd >= budget)
return rcvd;
poll_exit:
napi_complete((napi));
- BNAD_UPDATE_CTR(bnad, netif_rx_complete);
-
+ rx_ctrl->rx_complete++;
if (rx_ctrl->ccb)
- bnad_enable_rx_irq(bnad, rx_ctrl->ccb);
+ bnad_enable_rx_irq_unsafe(rx_ctrl->ccb);
+
return rcvd;
}
@@ -1875,20 +1877,14 @@ bnad_cleanup_rx(struct bnad *bnad, u32 rx_id)
struct bna_rx_config *rx_config = &bnad->rx_config[rx_id];
struct bna_res_info *res_info = &bnad->rx_res_info[rx_id].res_info[0];
unsigned long flags;
- int dim_timer_del = 0;
if (!rx_info->rx)
return;
- if (0 == rx_id) {
- spin_lock_irqsave(&bnad->bna_lock, flags);
- dim_timer_del = bnad_dim_timer_running(bnad);
- if (dim_timer_del)
- clear_bit(BNAD_RF_DIM_TIMER_RUNNING, &bnad->run_flags);
- spin_unlock_irqrestore(&bnad->bna_lock, flags);
- if (dim_timer_del)
- del_timer_sync(&bnad->dim_timer);
- }
+ spin_lock_irqsave(&bnad->bna_lock, flags);
+ if (0 == rx_id)
+ bnad_dim_timer_stop(bnad, flags);
+ spin_unlock_irqrestore(&bnad->bna_lock, flags);
init_completion(&bnad->bnad_completions.rx_comp);
spin_lock_irqsave(&bnad->bna_lock, flags);
@@ -2382,12 +2378,11 @@ bnad_enable_msix(struct bnad *bnad)
spin_lock_irqsave(&bnad->bna_lock, flags);
/* ret = #of vectors that we got */
- bnad_q_num_adjust(bnad, ret, 0);
+ bnad_q_num_adjust(bnad, (ret - BNAD_MAILBOX_MSIX_VECTORS) / 2,
+ (ret - BNAD_MAILBOX_MSIX_VECTORS) / 2);
spin_unlock_irqrestore(&bnad->bna_lock, flags);
- bnad->msix_num = (bnad->num_tx * bnad->num_txq_per_tx)
- + (bnad->num_rx
- * bnad->num_rxp_per_rx) +
+ bnad->msix_num = BNAD_NUM_TXQ + BNAD_NUM_RXP +
BNAD_MAILBOX_MSIX_VECTORS;
if (bnad->msix_num > ret)
@@ -2544,15 +2539,26 @@ bnad_start_xmit(struct sk_buff *skb, struct net_device *netdev)
u32 unmap_prod, wis, wis_used, wi_range;
u32 vectors, vect_id, i, acked;
int err;
+ unsigned int len;
struct bnad_unmap_q *unmap_q = tcb->unmap_q;
dma_addr_t dma_addr;
struct bna_txq_entry *txqent;
u16 flags;
- if (unlikely
- (skb->len <= ETH_HLEN || skb->len > BFI_TX_MAX_DATA_PER_PKT)) {
+ if (unlikely(skb->len <= ETH_HLEN)) {
dev_kfree_skb(skb);
+ BNAD_UPDATE_CTR(bnad, tx_skb_too_short);
+ return NETDEV_TX_OK;
+ }
+ if (unlikely(skb_headlen(skb) > BFI_TX_MAX_DATA_PER_VECTOR)) {
+ dev_kfree_skb(skb);
+ BNAD_UPDATE_CTR(bnad, tx_skb_headlen_too_long);
+ return NETDEV_TX_OK;
+ }
+ if (unlikely(skb_headlen(skb) == 0)) {
+ dev_kfree_skb(skb);
+ BNAD_UPDATE_CTR(bnad, tx_skb_headlen_zero);
return NETDEV_TX_OK;
}
@@ -2562,12 +2568,14 @@ bnad_start_xmit(struct sk_buff *skb, struct net_device *netdev)
*/
if (unlikely(!test_bit(BNAD_TXQ_TX_STARTED, &tcb->flags))) {
dev_kfree_skb(skb);
+ BNAD_UPDATE_CTR(bnad, tx_skb_stopping);
return NETDEV_TX_OK;
}
vectors = 1 + skb_shinfo(skb)->nr_frags;
- if (vectors > BFI_TX_MAX_VECTORS_PER_PKT) {
+ if (unlikely(vectors > BFI_TX_MAX_VECTORS_PER_PKT)) {
dev_kfree_skb(skb);
+ BNAD_UPDATE_CTR(bnad, tx_skb_max_vectors);
return NETDEV_TX_OK;
}
wis = BNA_TXQ_WI_NEEDED(vectors); /* 4 vectors per work item */
@@ -2605,8 +2613,6 @@ bnad_start_xmit(struct sk_buff *skb, struct net_device *netdev)
}
unmap_prod = unmap_q->producer_index;
- wis_used = 1;
- vect_id = 0;
flags = 0;
txq_prod = tcb->producer_index;
@@ -2614,9 +2620,6 @@ bnad_start_xmit(struct sk_buff *skb, struct net_device *netdev)
BUG_ON(!(wi_range <= tcb->q_depth));
txqent->hdr.wi.reserved = 0;
txqent->hdr.wi.num_vectors = vectors;
- txqent->hdr.wi.opcode =
- htons((skb_is_gso(skb) ? BNA_TXQ_WI_SEND_LSO :
- BNA_TXQ_WI_SEND));
if (vlan_tx_tag_present(skb)) {
vlan_tag = (u16) vlan_tx_tag_get(skb);
@@ -2631,62 +2634,91 @@ bnad_start_xmit(struct sk_buff *skb, struct net_device *netdev)
txqent->hdr.wi.vlan_tag = htons(vlan_tag);
if (skb_is_gso(skb)) {
+ if (unlikely(skb_is_gso(skb) > netdev->mtu)) {
+ dev_kfree_skb(skb);
+ BNAD_UPDATE_CTR(bnad, tx_skb_mss_too_long);
+ return NETDEV_TX_OK;
+ }
+ if (unlikely((skb_is_gso(skb) + skb_transport_offset(skb) +
+ tcp_hdrlen(skb)) >= skb->len)) {
+ txqent->hdr.wi.opcode =
+ __constant_htons(BNA_TXQ_WI_SEND);
+ txqent->hdr.wi.lso_mss = 0;
+ BNAD_UPDATE_CTR(bnad, tx_skb_tso_too_short);
+ } else {
+ txqent->hdr.wi.opcode =
+ __constant_htons(BNA_TXQ_WI_SEND_LSO);
+ txqent->hdr.wi.lso_mss = htons(skb_is_gso(skb));
+ }
+
err = bnad_tso_prepare(bnad, skb);
- if (err) {
+ if (unlikely(err)) {
dev_kfree_skb(skb);
+ BNAD_UPDATE_CTR(bnad, tx_skb_tso_prepare);
return NETDEV_TX_OK;
}
- txqent->hdr.wi.lso_mss = htons(skb_is_gso(skb));
flags |= (BNA_TXQ_WI_CF_IP_CKSUM | BNA_TXQ_WI_CF_TCP_CKSUM);
txqent->hdr.wi.l4_hdr_size_n_offset =
htons(BNA_TXQ_WI_L4_HDR_N_OFFSET
(tcp_hdrlen(skb) >> 2,
skb_transport_offset(skb)));
- } else if (skb->ip_summed == CHECKSUM_PARTIAL) {
- u8 proto = 0;
-
+ } else {
+ txqent->hdr.wi.opcode = __constant_htons(BNA_TXQ_WI_SEND);
txqent->hdr.wi.lso_mss = 0;
- if (skb->protocol == htons(ETH_P_IP))
- proto = ip_hdr(skb)->protocol;
- else if (skb->protocol == htons(ETH_P_IPV6)) {
- /* nexthdr may not be TCP immediately. */
- proto = ipv6_hdr(skb)->nexthdr;
+ if (unlikely(skb->len > (netdev->mtu + ETH_HLEN))) {
+ dev_kfree_skb(skb);
+ BNAD_UPDATE_CTR(bnad, tx_skb_non_tso_too_long);
+ return NETDEV_TX_OK;
}
- if (proto == IPPROTO_TCP) {
- flags |= BNA_TXQ_WI_CF_TCP_CKSUM;
- txqent->hdr.wi.l4_hdr_size_n_offset =
- htons(BNA_TXQ_WI_L4_HDR_N_OFFSET
- (0, skb_transport_offset(skb)));
- BNAD_UPDATE_CTR(bnad, tcpcsum_offload);
+ if (skb->ip_summed == CHECKSUM_PARTIAL) {
+ u8 proto = 0;
- BUG_ON(!(skb_headlen(skb) >=
- skb_transport_offset(skb) + tcp_hdrlen(skb)));
-
- } else if (proto == IPPROTO_UDP) {
- flags |= BNA_TXQ_WI_CF_UDP_CKSUM;
- txqent->hdr.wi.l4_hdr_size_n_offset =
- htons(BNA_TXQ_WI_L4_HDR_N_OFFSET
- (0, skb_transport_offset(skb)));
-
- BNAD_UPDATE_CTR(bnad, udpcsum_offload);
+ if (skb->protocol == __constant_htons(ETH_P_IP))
+ proto = ip_hdr(skb)->protocol;
+ else if (skb->protocol ==
+ __constant_htons(ETH_P_IPV6)) {
+ /* nexthdr may not be TCP immediately. */
+ proto = ipv6_hdr(skb)->nexthdr;
+ }
+ if (proto == IPPROTO_TCP) {
+ flags |= BNA_TXQ_WI_CF_TCP_CKSUM;
+ txqent->hdr.wi.l4_hdr_size_n_offset =
+ htons(BNA_TXQ_WI_L4_HDR_N_OFFSET
+ (0, skb_transport_offset(skb)));
+
+ BNAD_UPDATE_CTR(bnad, tcpcsum_offload);
+
+ if (unlikely(skb_headlen(skb) <
+ skb_transport_offset(skb) + tcp_hdrlen(skb))) {
+ dev_kfree_skb(skb);
+ BNAD_UPDATE_CTR(bnad, tx_skb_tcp_hdr);
+ return NETDEV_TX_OK;
+ }
- BUG_ON(!(skb_headlen(skb) >=
- skb_transport_offset(skb) +
- sizeof(struct udphdr)));
- } else {
- err = skb_checksum_help(skb);
- BNAD_UPDATE_CTR(bnad, csum_help);
- if (err) {
+ } else if (proto == IPPROTO_UDP) {
+ flags |= BNA_TXQ_WI_CF_UDP_CKSUM;
+ txqent->hdr.wi.l4_hdr_size_n_offset =
+ htons(BNA_TXQ_WI_L4_HDR_N_OFFSET
+ (0, skb_transport_offset(skb)));
+
+ BNAD_UPDATE_CTR(bnad, udpcsum_offload);
+ if (unlikely(skb_headlen(skb) <
+ skb_transport_offset(skb) +
+ sizeof(struct udphdr))) {
+ dev_kfree_skb(skb);
+ BNAD_UPDATE_CTR(bnad, tx_skb_udp_hdr);
+ return NETDEV_TX_OK;
+ }
+ } else {
dev_kfree_skb(skb);
- BNAD_UPDATE_CTR(bnad, csum_help_err);
+ BNAD_UPDATE_CTR(bnad, tx_skb_csum_err);
return NETDEV_TX_OK;
}
+ } else {
+ txqent->hdr.wi.l4_hdr_size_n_offset = 0;
}
- } else {
- txqent->hdr.wi.lso_mss = 0;
- txqent->hdr.wi.l4_hdr_size_n_offset = 0;
}
txqent->hdr.wi.flags = htons(flags);
@@ -2694,20 +2726,38 @@ bnad_start_xmit(struct sk_buff *skb, struct net_device *netdev)
txqent->hdr.wi.frame_length = htonl(skb->len);
unmap_q->unmap_array[unmap_prod].skb = skb;
- BUG_ON(!(skb_headlen(skb) <= BFI_TX_MAX_DATA_PER_VECTOR));
- txqent->vector[vect_id].length = htons(skb_headlen(skb));
+ len = skb_headlen(skb);
+ txqent->vector[0].length = htons(len);
dma_addr = dma_map_single(&bnad->pcidev->dev, skb->data,
skb_headlen(skb), DMA_TO_DEVICE);
dma_unmap_addr_set(&unmap_q->unmap_array[unmap_prod], dma_addr,
dma_addr);
- BNA_SET_DMA_ADDR(dma_addr, &txqent->vector[vect_id].host_addr);
+ BNA_SET_DMA_ADDR(dma_addr, &txqent->vector[0].host_addr);
BNA_QE_INDX_ADD(unmap_prod, 1, unmap_q->q_depth);
+ vect_id = 0;
+ wis_used = 1;
+
for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
struct skb_frag_struct *frag = &skb_shinfo(skb)->frags[i];
u16 size = frag->size;
+ if (unlikely(size == 0)) {
+ unmap_prod = unmap_q->producer_index;
+ prefetch(&unmap_q->unmap_array[unmap_prod + 1]);
+
+ BNAD_PCI_UNMAP_SKB(&bnad->pcidev->dev,
+ unmap_q->unmap_array,
+ unmap_prod, unmap_q->q_depth, skb,
+ i);
+ dev_kfree_skb(skb);
+ BNAD_UPDATE_CTR(bnad, tx_skb_frag_zero);
+ return NETDEV_TX_OK;
+ }
+
+ len += size;
+
if (++vect_id == BFI_TX_MAX_VECTORS_PER_WI) {
vect_id = 0;
if (--wi_range)
@@ -2718,10 +2768,10 @@ bnad_start_xmit(struct sk_buff *skb, struct net_device *netdev)
wis_used = 0;
BNA_TXQ_QPGE_PTR_GET(txq_prod, tcb->sw_qpt,
txqent, wi_range);
- BUG_ON(!(wi_range <= tcb->q_depth));
}
wis_used++;
- txqent->hdr.wi_ext.opcode = htons(BNA_TXQ_WI_EXTENSION);
+ txqent->hdr.wi_ext.opcode =
+ __constant_htons(BNA_TXQ_WI_EXTENSION);
}
BUG_ON(!(size <= BFI_TX_MAX_DATA_PER_VECTOR));
@@ -2734,6 +2784,18 @@ bnad_start_xmit(struct sk_buff *skb, struct net_device *netdev)
BNA_QE_INDX_ADD(unmap_prod, 1, unmap_q->q_depth);
}
+ if (unlikely(len != skb->len)) {
+ unmap_prod = unmap_q->producer_index;
+ prefetch(&unmap_q->unmap_array[unmap_prod + 1]);
+
+ BNAD_PCI_UNMAP_SKB(&bnad->pcidev->dev, unmap_q->unmap_array,
+ unmap_prod, unmap_q->q_depth, skb,
+ skb_shinfo(skb)->nr_frags);
+ dev_kfree_skb(skb);
+ BNAD_UPDATE_CTR(bnad, tx_skb_len_mismatch);
+ return NETDEV_TX_OK;
+ }
+
unmap_q->producer_index = unmap_prod;
BNA_QE_INDX_ADD(txq_prod, wis_used, tcb->q_depth);
tcb->producer_index = txq_prod;
@@ -2744,6 +2806,7 @@ bnad_start_xmit(struct sk_buff *skb, struct net_device *netdev)
return NETDEV_TX_OK;
bna_txq_prod_indx_doorbell(tcb);
+ smp_mb();
if ((u16) (*tcb->hw_consumer_index) != tcb->consumer_index)
tasklet_schedule(&bnad->tx_free_tasklet);
@@ -2810,6 +2873,9 @@ bnad_set_rx_mode(struct net_device *netdev)
}
}
+ if (bnad->rx_info[0].rx == NULL)
+ goto unlock;
+
bna_rx_mode_set(bnad->rx_info[0].rx, new_mask, valid_mask, NULL);
if (!netdev_mc_empty(netdev)) {
@@ -2962,12 +3028,9 @@ bnad_netpoll(struct net_device *netdev)
continue;
for (j = 0; j < bnad->num_rxp_per_rx; j++) {
rx_ctrl = &rx_info->rx_ctrl[j];
- if (rx_ctrl->ccb) {
- bnad_disable_rx_irq(bnad,
- rx_ctrl->ccb);
+ if (rx_ctrl->ccb)
bnad_netif_rx_schedule_poll(bnad,
rx_ctrl->ccb);
- }
}
}
}
diff --git a/drivers/net/ethernet/brocade/bna/bnad.h b/drivers/net/ethernet/brocade/bna/bnad.h
index 60c2e9d..8a31882 100644
--- a/drivers/net/ethernet/brocade/bna/bnad.h
+++ b/drivers/net/ethernet/brocade/bna/bnad.h
@@ -56,6 +56,11 @@ struct bnad_rx_ctrl {
struct bnad *bnad;
unsigned long flags;
struct napi_struct napi;
+ u64 rx_intr_ctr;
+ u64 rx_poll_ctr;
+ u64 rx_schedule;
+ u64 rx_keep_poll;
+ u64 rx_complete;
};
#define BNAD_RXMODE_PROMISC_DEFAULT BNA_RXMODE_PROMISC
@@ -148,6 +153,20 @@ struct bnad_drv_stats {
u64 udpcsum_offload;
u64 csum_help;
u64 csum_help_err;
+ u64 tx_skb_too_short;
+ u64 tx_skb_stopping;
+ u64 tx_skb_max_vectors;
+ u64 tx_skb_mss_too_long;
+ u64 tx_skb_tso_too_short;
+ u64 tx_skb_tso_prepare;
+ u64 tx_skb_non_tso_too_long;
+ u64 tx_skb_tcp_hdr;
+ u64 tx_skb_udp_hdr;
+ u64 tx_skb_csum_err;
+ u64 tx_skb_headlen_too_long;
+ u64 tx_skb_headlen_zero;
+ u64 tx_skb_frag_zero;
+ u64 tx_skb_len_mismatch;
u64 hw_stats_updates;
u64 netif_rx_schedule;
@@ -346,7 +365,7 @@ extern void bnad_netdev_hwstats_fill(struct bnad *bnad,
#define bnad_enable_rx_irq_unsafe(_ccb) \
{ \
- if (likely(test_bit(BNAD_RXQ_STARTED, &ccb->rcb[0]->flags))) {\
+ if (likely(test_bit(BNAD_RXQ_STARTED, &(_ccb)->rcb[0]->flags))) {\
bna_ib_coalescing_timer_set((_ccb)->i_dbell, \
(_ccb)->rx_coalescing_timeo); \
bna_ib_ack((_ccb)->i_dbell, 0); \
@@ -356,5 +375,19 @@ extern void bnad_netdev_hwstats_fill(struct bnad *bnad,
#define bnad_dim_timer_running(_bnad) \
(((_bnad)->cfg_flags & BNAD_CF_DIM_ENABLED) && \
(test_bit(BNAD_RF_DIM_TIMER_RUNNING, &((_bnad)->run_flags))))
+#define bnad_dim_timer_stop(_bnad, _flags) \
+do { \
+ int to_del = 0; \
+ \
+ if ((_bnad)->cfg_flags & BNAD_CF_DIM_ENABLED && \
+ test_bit(BNAD_RF_DIM_TIMER_RUNNING, &(_bnad)->run_flags)) {\
+ clear_bit(BNAD_RF_DIM_TIMER_RUNNING, &(_bnad)->run_flags);\
+ to_del = 1; \
+ } \
+ spin_unlock_irqrestore(&(_bnad)->bna_lock, (_flags)); \
+ if (to_del) \
+ del_timer_sync(&(_bnad)->dim_timer); \
+ spin_lock_irqsave(&(_bnad)->bna_lock, (_flags));\
+} while (0)
#endif /* __BNAD_H__ */
--
1.7.1
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox