Jan Kiszka [Mon, 8 Feb 2010 10:12:36 +0000 (10:12 +0000)]
CAPI: Rework capiminor RX handler
Avoid re-queuing skbs unless the error detected in handle_recv_skb is
expected to be recoverable such as lacking memory, a full CAPI queue, a
full TTY input buffer, or a not yet existing TTY.
Signed-off-by: Jan Kiszka <jan.kiszka@web.de> Signed-off-by: David S. Miller <davem@davemloft.net>
Jan Kiszka [Mon, 8 Feb 2010 10:12:35 +0000 (10:12 +0000)]
CAPI: Drop handle_minor_recv from capinc_tty_write
Sending a message down the CAPI stack may trigger the reception of an
answer, but this will go through capi_recv_message and call
handle_minor_recv from there. There is no need to walk the receive queue
on capinc_tty_write.
Signed-off-by: Jan Kiszka <jan.kiszka@web.de> Signed-off-by: David S. Miller <davem@davemloft.net>
Jan Kiszka [Mon, 8 Feb 2010 10:12:32 +0000 (10:12 +0000)]
CAPI: Drop obsolete nccip from capiminor struct
The nccip in capiminor used to serve as an indicator that the NCCI was
close. But we don't need this, we issue a hangup on capincci_free_minor.
So drop this legacy.
Signed-off-by: Jan Kiszka <jan.kiszka@web.de> Signed-off-by: David S. Miller <davem@davemloft.net>
Jan Kiszka [Mon, 8 Feb 2010 10:12:27 +0000 (10:12 +0000)]
CAPI: Use kref on capiminor
Install a reference counter for capiminor objects. Acquire it when
obtaining a capiminor from the array during capinc_tty_open, drop it
when closing the tty again. Another reference is held for the hook-up
with capincci.
Signed-off-by: Jan Kiszka <jan.kiszka@web.de> Signed-off-by: David S. Miller <davem@davemloft.net>
Jan Kiszka [Mon, 8 Feb 2010 10:12:25 +0000 (10:12 +0000)]
CAPI: Dynamically register minor devices
Register capiminors dynamically with the TTY core so that udev can make
them show up as the NCCIs appear or disappear. This removes the need to
check if the capiminor requested in capinc_tty_open actually exists.
And this completely obsoletes capifs which will be scheduled for removal
in a later patch.
Signed-off-by: Jan Kiszka <jan.kiszka@web.de> Signed-off-by: David S. Miller <davem@davemloft.net>
Jan Kiszka [Mon, 8 Feb 2010 10:12:24 +0000 (10:12 +0000)]
CAPI: Clean up capinc_tty_init/exit
Return proper error code if tty_register_driver fails. In contrast,
tty_unregister_driver cannot practically fail, so drop that error
handling. Finally, mark capinc_tty_init/exit with __init/__exit.
Signed-off-by: Jan Kiszka <jan.kiszka@web.de> Signed-off-by: David S. Miller <davem@davemloft.net>
Jan Kiszka [Mon, 8 Feb 2010 10:12:19 +0000 (10:12 +0000)]
CAPI: Rework locking of capidev members
Rename 'ncci_list_mtx' to 'lock', expressing that it now protects a
larger set of capidev members: the NCCI list, ap.applid (ie. the
registration of the application), and modifications of userflags.
We do not need to protect each and every check for ap.applid because,
once an application is registered, it will stay for the whole lifetime
of the device.
Also, there is no need to apply the capidev mutex during release (if
there could be concurrent users, we would crash them anyway by freeing
the device at the end of capi_release).
Signed-off-by: Jan Kiszka <jan.kiszka@web.de> Signed-off-by: David S. Miller <davem@davemloft.net>
Jan Kiszka [Mon, 8 Feb 2010 10:12:18 +0000 (10:12 +0000)]
CAPI: Clean up capi_open/release
Fold capidev_alloc and capidev_free into capi_open and capi_release -
there are no other users. Someone pushed a lock_kernel into capi_open.
Drop it, we don't need it. Also remove the useless test from open that
checks for private_data == NULL.
Signed-off-by: Jan Kiszka <jan.kiszka@web.de> Signed-off-by: David S. Miller <davem@davemloft.net>
Jan Kiszka [Mon, 8 Feb 2010 10:12:17 +0000 (10:12 +0000)]
CAPI: Convert capidev_list_lock into a mutex
No need for anything "harder" here (specifically no need for
irqsave...). Also, make the list removal the first operation of
capidev_free to avoid dumping half-released devices via /proc.
Signed-off-by: Jan Kiszka <jan.kiszka@web.de> Signed-off-by: David S. Miller <davem@davemloft.net>
Jan Kiszka [Mon, 8 Feb 2010 10:12:16 +0000 (10:12 +0000)]
CAPI: Reduce #ifdef mess around CONFIG_ISDN_CAPI_MIDDLEWARE
Make the code a bit more readable be providing stub functions for the
!CONFIG_ISDN_CAPI_MIDDLEWARE case. Though a few lines are moved around,
this comes with no functional changes.
Signed-off-by: Jan Kiszka <jan.kiszka@web.de> Signed-off-by: David S. Miller <davem@davemloft.net>
Jan Kiszka [Mon, 8 Feb 2010 10:12:15 +0000 (10:12 +0000)]
CAPI: Rework application locking
Drop the application rw-lock in favour of RCU. This synchronizes
capi20_release against capi_ctr_handle_message which may dereference an
application from (soft-)IRQ context. Any other access to the application
list is now protected by the capi_controller_lock as well. This also
allows to safely inspect applications for /proc dumping by holding
capi_controller_lock.
At this chance, drop some useless release_in_progress checks where we
obtained the application pointer from the list (which becomes NULL on
release_in_progress).
Signed-off-by: Jan Kiszka <jan.kiszka@web.de> Signed-off-by: David S. Miller <davem@davemloft.net>
Jan Kiszka [Mon, 8 Feb 2010 10:12:14 +0000 (10:12 +0000)]
CAPI: Rework locking of controller data structures
This patch applies the mutex so far only protecting the controller list
to (almost) all accesses of controller data structures. It also reworks
waiting on state changes in old_capi_manufacturer so that it no longer
poll and holds a module reference to the controller owner while waiting
(the latter was partly done already). Modification and checking of the
blocked state remains racy by design, the caller is responsible for
dealing with this.
Signed-off-by: Jan Kiszka <jan.kiszka@web.de> Signed-off-by: David S. Miller <davem@davemloft.net>
Jan Kiszka [Mon, 8 Feb 2010 10:12:13 +0000 (10:12 +0000)]
CAPI: Rework controller state notifier
Another step towards proper locking: Rework the callback provided to
capidrv for controller state changes. This is so far attached to an
application, which would require us to hold the corresponding lock
across notification calls.
But there is no direct relation between a controller up/down event and
an application, so let's decouple them and provide a notifier call chain
for those events instead. This notifier chain is first of all used
internally. Here we request the highest priority to unsure that
housekeeping work is done before any other notifications. The chain is
exported via [un]register_capictr_notifier to our only user, capidrv, to
replace the racy and unfixable capi20_set_callback.
Signed-off-by: Jan Kiszka <jan.kiszka@web.de> Signed-off-by: David S. Miller <davem@davemloft.net>
Jan Kiszka [Mon, 8 Feb 2010 10:12:12 +0000 (10:12 +0000)]
CAPI: Rework capi_ctr_ready/down
This step prepares the application of proper controller locking: Push
all state changing work into the notify handler that are called by
capi_ctr_ready and capi_ctr_down, switch detach_capi_ctr to issue a
synchronous ctr_down. Also ensure that we do not go through any action
if the state did not change.
Signed-off-by: Jan Kiszka <jan.kiszka@web.de> Signed-off-by: David S. Miller <davem@davemloft.net>
Jan Kiszka [Mon, 8 Feb 2010 10:12:09 +0000 (10:12 +0000)]
CAPI: Reduce chattiness during module loading/removal
The CVS revisions dumped by all CAPI modules are meaningless today. And
that some CAPI module is loaded or removed does not necessarily deserve
a message. Just keep the message of the central module, capi.ko, drop
the rest.
Signed-off-by: Jan Kiszka <jan.kiszka@web.de> Signed-off-by: David S. Miller <davem@davemloft.net>
Jan Kiszka [Mon, 8 Feb 2010 10:12:05 +0000 (10:12 +0000)]
CAPI: Fix leaks in capifs_new_ncci
When something went wrong during capifs_new_ncci, the looked up dentry
was not properly released. Neither was the allocated inode. Refactor the
function to avoid leaks.
Signed-off-by: Jan Kiszka <jan.kiszka@web.de> Signed-off-by: David S. Miller <davem@davemloft.net>
net neigh: Decouple per interface neighbour table controls from binary sysctls
Stop computing the number of neighbour table settings we have by
counting the number of binary sysctls. This behaviour was silly
and meant that we could not add another neighbour table setting
without also adding another binary sysctl.
Don't pass the binary sysctl path for neighour table entries
into neigh_sysctl_register. These parameters are no longer
used and so are just dead code.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net ipv4: Decouple ipv4 interface parameters from binary sysctl numbers
Stop using the binary sysctl enumeartion in sysctl.h as an index into
a per interface array. This leads to unnecessary binary sysctl number
allocation, and a fragility in data structure and implementation
because of unnecessary coupling.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Tue, 16 Feb 2010 23:23:23 +0000 (15:23 -0800)]
sparc64: Kill bogus ip_tables.h include.
Fixes the following build failure:
CC arch/sparc/kernel/sys_sparc32.o
In file included from include/linux/netfilter_ipv4/ip_tables.h:28,
from arch/sparc/kernel/sys_sparc32.c:46:
include/linux/netfilter/x_tables.h:525: error: expected declaration specifiers or ‘...’ before ‘nf_hookfn’
Signed-off-by: David S. Miller <davem@davemloft.net>
Remove unnecessary code that works around older versions of ethtool
that can pass down invalid advertisement speed values. This old
code prevents the user from specifying multiple advertisement values.
The new code uses simple masking to mask out invalid advertisment bits.
Reported-by: Vlad Yasevich <vladislav.yasevich@hp.com> Signed-off-by: Michael Chan <mchan@broadcom.com> Signed-off-by: Benjamin Li <benli@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Michael Chan [Mon, 15 Feb 2010 19:42:07 +0000 (19:42 +0000)]
bnx2: Check BNX2_FLAG_USING_MSIX flag when setting up MSIX.
Checking the flag is more correct than checking bp->irq_nvecs. By
accident it is not a problem because we always have more than 1
vectors when using MSIX mode.
Signed-off-by: Michael Chan <mchan@broadcom.com> Signed-off-by: Benjamin Li <benli@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Alexey Dobriyan [Tue, 16 Feb 2010 07:57:44 +0000 (07:57 +0000)]
gre: fix netns vs proto registration ordering
GRE protocol receive hook can be called right after protocol addition is done.
If netns stuff is not yet initialized, we're going to oops in
net_generic().
This is remotely oopsable if ip_gre is compiled as module and packet
comes at unfortunate moment of module loading.
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sadanand Mutyala <Sadanand.Mutyala@xilinx.com> Signed-off-by: John Linn <john.linn@xilinx.com> Acked-by: Grant Likely <grant.likely@secretlab.ca> Signed-off-by: David S. Miller <davem@davemloft.net>
To see the effect make sure you have an empty SPD.
On window1 "ip xfrm mon" and on window2 issue "ip xfrm policy flush"
You get prompt back in window1 and you see the flush event on window2.
With this fix, you still get prompt on window1 but no event on window2.
Signed-off-by: Jamal Hadi Salim <hadi@cyberus.ca> Signed-off-by: David S. Miller <davem@davemloft.net>
jamal [Wed, 10 Feb 2010 23:51:27 +0000 (23:51 +0000)]
xfrm: Flushing empty SAD generates false events
To see the effect make sure you have an empty SAD.
-On window1 "ip xfrm mon"
-on window2 issue "ip xfrm state flush"
You get prompt back in window1
and you see the flush event on window2.
With this fix, you still get prompt on window1 but no
event on window2.
I was tempted to return -ESRCH on window1 (which would
show "RTNETLINK answers: No such process") but didnt want
to change current behavior.
To see the effect make sure you have an empty SAD.
On window1 "ip xfrm mon" and on window2 issue "ip xfrm state flush"
You get prompt back in window1 and you see the flush event on window2.
With this fix, you still get prompt on window1 but no event on window2.
Signed-off-by: Jamal Hadi Salim <hadi@cyberus.ca> Signed-off-by: David S. Miller <davem@davemloft.net>
Shan Wei [Thu, 11 Feb 2010 00:12:45 +0000 (00:12 +0000)]
IPv6: Delete redundant counter of IPSTATS_MIB_REASMFAILS
When no more memory can be allocated, fq_find() will return NULL and
increase the value of IPSTATS_MIB_REASMFAILS. In this case,
ipv6_frag_rcv() also increase the value of IPSTATS_MIB_REASMFAILS.
So, the patch deletes redundant counter of IPSTATS_MIB_REASMFAILS in fq_find().
and deletes the unused parameter of idev.
Signed-off-by: Shan Wei <shanwei@cn.fujitsu.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Arnd Bergmann [Thu, 11 Feb 2010 05:55:39 +0000 (05:55 +0000)]
net/macvtap: fix reference counting
The RCU usage in the original code was broken because
there are cases where we possibly sleep with rcu_read_lock
held. As a fix, change the macvtap_file_get_queue to
get a reference on the socket and the netdev instead of
taking the full rcu_read_lock.
Also, change macvtap_file_get_queue failure case to
not require a subsequent macvtap_file_put_queue, as
pointed out by Ed Swierk.
Signed-off-by: Arnd Bergmann <arnd@arndb.de> Cc: Ed Swierk <eswierk@aristanetworks.com> Cc: Sridhar Samudrala <sri@us.ibm.com> Acked-by: Sridhar Samudrala <sri@us.ibm.com> Acked-by: Ed Swierk <eswierk@aristanetworks.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Divy Le Ray [Sat, 13 Feb 2010 09:44:35 +0000 (09:44 +0000)]
cxgb3: fix link flap
The driver is expected to report that the link is up
when the phy Rx signal is established and the mac
has not detected a link fault.
The code is however broken, the driver does not check the link fault
status when the phy link status changes.
The link fault status being checked within a short period of time,
it leads to link up/link down events.
Signed-off-by: Divy Le Ray <divy@chelsio.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Divy Le Ray [Sat, 13 Feb 2010 09:44:30 +0000 (09:44 +0000)]
cxgb3: FIx VLAN over Jumbo frames
The mac is expected to auto-inflate the Maximum Frame size for VLAN
tagged frames. It however does not work with jumbo frames.
Work around the bug adding 4 to the Maximum Frame for MTUs
greater than 1536.
Signed-off-by: Divy Le Ray <divy@chelsio.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Peter Waskiewicz [Fri, 12 Feb 2010 13:48:44 +0000 (13:48 +0000)]
ixgbe: Cleanup incorrect header comments
The recent n-tuple patches added some comments to the headers
of the Flow Director functions that aren't accurate. This
cleans them up, and is a purely cosmetic patch.
Signed-off-by: Peter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Peter Waskiewicz [Fri, 12 Feb 2010 13:48:25 +0000 (13:48 +0000)]
ethtool: Move n-tuple capability check into set_flags
set_flags should check if the underlying device supports
n-tuple filter programming before setting the device flags
on the netdevice.
Signed-off-by: Peter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Peter Waskiewicz [Fri, 12 Feb 2010 13:48:05 +0000 (13:48 +0000)]
ethtool: Fix filter addition when caching n-tuple filters
We can allow a filter to be added successfully to the underlying
hardware, but still return an error if the cached list memory
allocation fails. This patch fixes that condition.
Signed-off-by: Peter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Florian Westphal [Mon, 15 Feb 2010 17:17:10 +0000 (18:17 +0100)]
netfilter: CONFIG_COMPAT: allow delta to exceed 32767
with 32 bit userland and 64 bit kernels, it is unlikely but possible
that insertion of new rules fails even tough there are only about 2000
iptables rules.
This happens because the compat delta is using a short int.
Easily reproducible via "iptables -m limit" ; after about 2050
rules inserting new ones fails with -ELOOP.
Note that compat_delta included 2 bytes of padding on x86_64, so
structure size remains the same.
Signed-off-by: Florian Westphal <fw@strlen.de> Signed-off-by: Patrick McHardy <kaber@trash.net>
Florian Westphal [Mon, 15 Feb 2010 17:16:26 +0000 (18:16 +0100)]
netfilter: ebtables: avoid explicit XT_ALIGN() in match/targets
This will cause trouble once CONFIG_COMPAT support is added to ebtables.
xt_compat_*_offset() calculate the kernel/userland structure size delta
using:
XT_ALIGN(size) - COMPAT_XT_ALIGN(size)
If the match/target sizes are aligned at registration time,
delta is always zero.
Should have zero effect for existing systems: xtables uses
XT_ALIGN() whenever it deals with match/target sizes.
Signed-off-by: Florian Westphal <fwestphal@astaro.com> Signed-off-by: Patrick McHardy <kaber@trash.net>
Florian Westphal [Mon, 15 Feb 2010 17:15:55 +0000 (18:15 +0100)]
netfilter: ebtables: abort if next_offset is too small
next_offset must be > 0, otherwise this loops forever.
The offset also contains the size of the ebt_entry structure
itself, so anything smaller is invalid.
Signed-off-by: Florian Westphal <fwestphal@astaro.com> Signed-off-by: Patrick McHardy <kaber@trash.net>
Patrick McHardy [Mon, 15 Feb 2010 17:13:33 +0000 (18:13 +0100)]
netfilter: nf_conntrack: add support for "conntrack zones"
Normally, each connection needs a unique identity. Conntrack zones allow
to specify a numerical zone using the CT target, connections in different
zones can use the same identity.
Example:
iptables -t raw -A PREROUTING -i veth0 -j CT --zone 1
iptables -t raw -A OUTPUT -o veth1 -j CT --zone 1
Jan Engelhardt [Mon, 15 Feb 2010 15:31:35 +0000 (16:31 +0100)]
netfilter: xt_recent: inform user when hitcount is too large
It is one of these things that iptables cannot catch and which can
cause "Invalid argument" to be printed. Without a hint in dmesg, it is
not going to be helpful.
Signed-off-by: Jan Engelhardt <jengelh@medozas.de> Signed-off-by: Patrick McHardy <kaber@trash.net>
which is definitely not correct, introduced a warning (reported
by Stephen Rothwell):
net/mac80211/iface.c: In function 'ieee80211_stop':
net/mac80211/iface.c:416: warning: passing argument 4 of '__dev_addr_unsync' makes pointer from integer without a cast
include/linux/netdevice.h:1967: note: expected 'int *' but argument is of type 'int'
and is thus reverted here.
Signed-off-by: David S. Miller <davem@davemloft.net>
vhost-net only uses memory barriers to control SMP effects
(communication with userspace potentially running on a different CPU),
so it should use SMP barriers and not mandatory barriers for memory
access ordering, as suggested by Documentation/memory-barriers.txt
Signed-off-by: Michael S. Tsirkin <mst@redhat.com> Acked-by: Rusty Russell <rusty@rustcorp.com.au> Signed-off-by: David S. Miller <davem@davemloft.net>
Joe Perches [Mon, 15 Feb 2010 06:40:39 +0000 (22:40 -0800)]
niu: Use pr_<level>, netdev_<level> and netif_<level> macros
Remove #define PFX
Add pr_fmt(fmt) KBUILD_MODNAME ": " fmt
Convert printks to pr_<level>
Convert printks without levels to pr_cont
Convert pr_<level> with np->dev to netdev_<level>
Convert dev_<level> to netdev_<level>
Convert niudbg to netif_printk
Convert niuinfo, niuwarn macros to netif_<level>(priv, type, dev...
Coalesce long formats
Convert embedded function names to "%s", __func__
Always use "%s()..." when __func__ is printed
Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Matt Carlson [Fri, 12 Feb 2010 14:47:12 +0000 (14:47 +0000)]
tg3: Fix AC131 loopback test errors for 5785
The AC131 does not enable the forced transmit clock settings
immediately. The workaround is to read the register again to get the
setting to take effect.
Signed-off-by: Matt Carlson <mcarlson@broadcom.com> Reviewed-by: Michael Chan <mchan@broadcom.com> Reviewed-by: Benjamin Li <benli@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Matt Carlson [Fri, 12 Feb 2010 14:47:11 +0000 (14:47 +0000)]
tg3: Fix napi assignments in loopback test
The 57765 lacks TSS support. This renders the napi assignments
incorrect in the loopback test function. This patch fixes the problem.
Signed-off-by: Matt Carlson <mcarlson@broadcom.com> Reviewed-by: Michael Chan <mchan@broadcom.com> Reviewed-by: Benjamin Li <benli@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Matt Carlson [Fri, 12 Feb 2010 14:47:10 +0000 (14:47 +0000)]
tg3: Reset phy during bringup when using phylib
The driver puts the phy into low-power mode when it releases the device.
If the device were to be reacquired, the phy needs a reset to bring it
back to full powered operation. This patch allows phylib-enabled
devices to reset the phy.
Signed-off-by: Matt Carlson <mcarlson@broadcom.com> Reviewed-by: Michael Chan <mchan@broadcom.com> Reviewed-by: Benjamin Li <benli@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Matt Carlson [Fri, 12 Feb 2010 14:47:09 +0000 (14:47 +0000)]
tg3: Turn off multiple DMA reads for 5717
The 5717's DMA read engine has a bug when initiating multiple DMA reads
across the PCIe bus. This patch disables the feature.
Signed-off-by: Matt Carlson <mcarlson@broadcom.com> Reviewed-by: Michael Chan <mchan@broadcom.com> Reviewed-by: Benjamin Li <benli@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Matt Carlson [Fri, 12 Feb 2010 14:47:08 +0000 (14:47 +0000)]
tg3: Fix 57765 A0 bootcode race condition
On A0 revision of 57765 asic rev devices, the bootcode will perform some
hardware operations, after the magic signature is presented, that will
collide with setup operations performed by the driver. The best way to
avoid the contention is to have the driver delay an additional 10
milliseconds. B0 revisions of the chip will make this workaround
unnecessary.
Signed-off-by: Matt Carlson <mcarlson@broadcom.com> Signed-off-by: Benjamin Li <benli@broadcom.com> Reviewed-by: Michael Chan <mchan@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Matt Carlson [Fri, 12 Feb 2010 14:47:07 +0000 (14:47 +0000)]
tg3: Unwedge stuck MSI-X vectors
The previous patch changed the code so that new rx buffer submissions to
the hardware stall if a new submission would overwrite data needed by an
unserviced rx packet. On very busy 5717 and 57765 asic rev devices,
there is a corner case where the hardware will fail to assert an MSI-X
interrupt for rx traffic. If that vector's interrupt never has another
reason to assert, any rx buffers held will never be serviced. If the
buffers are never serviced and the hardware consumes all the available
rx packets for other rx rings, deadlock will result.
The most reliable and least intrusive way to work around the problem is
to detect the case where new submissions would overwrite existing data
and force all rx interrupt vectors to fire.
Signed-off-by: Matt Carlson <mcarlson@broadcom.com> Reviewed-by: Michael Chan <mchan@broadcom.com> Reviewed-by: Benjamin Li <benli@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Matt Carlson [Fri, 12 Feb 2010 14:47:06 +0000 (14:47 +0000)]
tg3: Prevent rx producer ring overruns
When operating in RSS mode, it is possible for one rx return ring to
submit enough rx buffers back to the hardware such that it inadvertently
overwrites data needed by another rx return ring. This patch addresses
the problem by looking for non-NULL skb pointers in the
rx_[std|jmb]_buffers rings that parallel the rx producer rings.
Signed-off-by: Matt Carlson <mcarlson@broadcom.com> Signed-off-by: Michael Chan <mchan@broadcom.com> Reviewed-by: Benjamin Li <benli@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Matt Carlson [Fri, 12 Feb 2010 14:47:05 +0000 (14:47 +0000)]
tg3: Give MSI-X vec 1 rx backlog space
RSS ring 1 is responsible for submitting new rx buffers to the
hardware on behalf of all the other RSS rx return rings. Up until now
this ring submitted its new rx buffers to the producer ring directly.
The following patch will require that this ring have a place to put
backlogged rx packets. As a consequence, it can no longer submit new
buffers to the producer ring.
This patch adds code to allocate an extra shadow producer ring for this
RSS ring and adds RSS ring 1 to the list of rings needing buffer
transfers.
Signed-off-by: Matt Carlson <mcarlson@broadcom.com> Signed-off-by: Michael Chan <mchan@broadcom.com> Reviewed-by: Benjamin Li <benli@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
projects / linux-beck.git / log