linux/drivers/net, branch v2.6.37

atl1: fix oops when changing tx/rx ring params

2011-01-03T19:04:49Z

Commit 3f5a2a713aad28480d86b0add00c68484b54febc zeroes out the statistics message block (SMB) and coalescing message block (CMB) when adapter ring resources are freed. This is desirable behavior, but, as a side effect, the commit leads to an oops when atl1_set_ringparam() attempts to alter the number of rx or tx elements in the ring buffer (by using ethtool -G, for example). We don't want SMB or CMB to change during this operation. Modify atl1_set_ringparam() to preserve SMB and CMB when changing ring parameters. Cc: stable@kernel.org Signed-off-by: Jay Cliburn Reported-by: Tõnu Raitviir Signed-off-by: David S. Miller

starfire: Fix dma_addr_t size test for MIPS

2010-12-31T20:33:17Z

Commit 56543af "starfire: use BUILD_BUG_ON for netdrv_addr_t" revealed that the preprocessor condition used to find the size of dma_addr_t yielded the wrong result for some architectures and configurations. This was kluged for 64-bit PowerPC in commit 3e502e6 by adding yet another case to the condition. However, 64-bit MIPS configurations are not detected reliably either. This should be fixed by using CONFIG_ARCH_DMA_ADDR_T_64BIT, but that isn't yet defined everywhere it should be. Signed-off-by: Ben Hutchings Signed-off-by: David S. Miller

tg3: fix return value check in tg3_read_vpd()

2010-12-31T20:31:01Z

Besides -ETIMEDOUT and -EINTR, pci_read_vpd may return other error values like -ENODEV or -EINVAL which are ignored due to the buggy check, but the data are not read from VPD anyway and this is checked subsequently with at most 3 needless loop iterations. This does not show up as a runtime bug. CC: Matt Carlson CC: Michael Chan Signed-off-by: David Sterba Signed-off-by: David S. Miller

Broadcom CNIC core network driver: fix mem leak on allocation failures in cnic_alloc_uio_rings()

2010-12-31T19:18:48Z

We are leaking memory in drivers/net/cnic.c::cnic_alloc_uio_rings() if either of the calls to dma_alloc_coherent() fail. This patch fixes it by freeing both the memory allocated with kzalloc() and memory allocated with previous calls to dma_alloc_coherent() when there's a failure. Thanks to Joe Perches for suggesting a better implementation than my initial version. Signed-off-by: Jesper Juhl Acked-by: Michael Chan Signed-off-by: David S. Miller

skfp: testing the wrong variable in skfp_driver_init()

2010-12-28T21:54:57Z

The intent here was to test if the allocation failed but we tested "SharedMemSize" instead of "SharedMemAddr" by mistake. Signed-off-by: Dan Carpenter Reviewed-by: Jiri Pirko Signed-off-by: David S. Miller

ppp: allow disabling multilink protocol ID compression

2010-12-28T21:53:39Z

Linux would not connect to other router running old version Cisco IOS (12.0). This is most likely a bug in that version of IOS, since it is fixed in later versions. As a workaround this patch allows a module parameter to be set to disable compressing the protocol ID. See: https://bugzilla.vyatta.com/show_bug.cgi?id=3979 RFC 1990 allows an implementation to formulate MP fragments as if protocol compression had been negotiated. This allows us to always send compressed protocol IDs. But some implementations don't accept MP fragments with compressed protocol IDs. This parameter allows us to interoperate with them. The default value of the configurable parameter is the same as the current behavior: protocol compression is enabled. If protocol compression is disabled we will not send compressed protocol IDs. This is based on an earlier patch by Bob Gilligan (using a sysctl). Module parameter is writable to allow for enabling even if ppp is already loaded for other uses. Signed-off-by: Stephen Hemminger Acked-by: Eric Dumazet Acked-by: Paul Mackerras Signed-off-by: David S. Miller

ehea: Avoid changing vlan flags

2010-12-28T21:51:22Z

This patch avoids disabling the vlan flags using ethtool. Signed-off-by: Breno Leitao Signed-off-by: David S. Miller

Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6

2010-12-26T20:06:56Z

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6: (42 commits) ipv4: dont create routes on down devices epic100: hamachi: yellowfin: Fix skb allocation size sundance: Fix oopses with corrupted skb_shared_info Revert "ipv4: Allow configuring subnets as local addresses" USB: mcs7830: return negative if auto negotiate fails irda: prevent integer underflow in IRLMP_ENUMDEVICES tcp: fix listening_get_next() atl1c: Do not use legacy PCI power management mac80211: fix mesh forwarding MAINTAINERS: email address change net: Fix range checks in tcf_valid_offset(). net_sched: sch_sfq: fix allot handling hostap: remove netif_stop_queue from init mac80211/rt2x00: add ieee80211_tx_status_ni() typhoon: memory corruption in typhoon_get_drvinfo() net: Add USB PID for new MOSCHIP USB ethernet controller MCS7832 variant net_sched: always clone skbs ipv6: Fragment locally generated tunnel-mode IPSec6 packets as needed. netlink: fix gcc -Wconversion compilation warning asix: add USB ID for Logitec LAN-GTJ U2A ...

epic100: hamachi: yellowfin: Fix skb allocation size

2010-12-26T03:37:49Z

Joel Soete reported oopses during pppoe over sundance NIC, caused by a bug in skb allocation and dma mapping code, where skb_reserve() bytes weren't taken into account. As a followup to the patch: "sundance: Fix oopses with corrupted skb_shared_info" very similar code is fixed here for three other drivers. Signed-off-by: Jarek Poplawski Cc: Joel Soete Cc: Eric Dumazet Signed-off-by: David S. Miller

sundance: Fix oopses with corrupted skb_shared_info

2010-12-26T03:37:49Z

Joel Soete reported oopses at the beginning of pppoe connections since v2.6.35. After debugging the bug was found in sundance skb allocation and dma mapping code, where skb_reserve() bytes aren't taken into account. This is an old bug, only uncovered by some change in 2.6.35. Initial debugging patch by: Eric Dumazet Reported-by: Joel Soete Tested-by: Joel Soete Signed-off-by: Jarek Poplawski Cc: Eric Dumazet Signed-off-by: David S. Miller