linux/drivers/net, branch v3.8.12

ixgbe: fix EICR write in ixgbe_msix_other

2013-05-08T03:08:24Z

commit d87d830720a1446403ed38bfc2da268be0d356d1 upstream. Previously, the ixgbe_msix_other was writing the full 32bits of the set interrupts, instead of only the ones which the ixgbe_msix_other is handling. This resulted in a loss of performance when the X540's PPS feature is enabled due to sometimes clearing queue interrupts which resulted in the driver not getting the interrupt for cleaning the q_vector rings often enough. The fix is to simply mask the lower 16bits off so that this handler does not write them in the EICR, which causes them to remain high and be properly handled by the clean_rings interrupt routine as normal. Signed-off-by: Jacob Keller Tested-by: Phil Schmitt Signed-off-by: Jeff Kirsher Signed-off-by: Greg Kroah-Hartman

rt2x00: Fix transmit power troubles on some Ralink RT30xx cards

2013-05-08T03:08:20Z

commit 7e9dafd873034dd64ababcb858be424c4780ae13 upstream. Some cards on Ralink RT30xx chipset not have correctly TX_MIXER_GAIN value in them EEPROM/EFUSE. In this case, we must use default value, but always used EEPROM/EFUSE value. As result we have tranmitt power range from -10dBm to +6dBm instead 0dBm to +16dBm. Correctly value in EEPROM/EFUSE is one or more for RT3070 and two or more for other RT30xx chips. Tested on Canyon CNP-WF518N1 usb Wi-Fi dongle and Jorjin WN8020 usb embedded Wi-Fi module. Signed-off-by: Alex A. Mihaylov Acked-by: Gertjan van Wingerde Acked-by: Stanislaw Gruszka Signed-off-by: John W. Linville Signed-off-by: Greg Kroah-Hartman

gianfar: do not advertise any alarm capability.

2013-05-08T03:08:19Z

commit cd4baaaa04b4aaa3b0ec4d13a6f3d203b92eadbd upstream. An early draft of the PHC patch series included an alarm in the gianfar driver. During the review process, the alarm code was dropped, but the capability removal was overlooked. This patch fixes the issue by advertising zero alarms. This patch should be applied to every 3.x stable kernel. Signed-off-by: Richard Cochran Reported-by: Chris LaRocque Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman

mwifiex: Call pci_release_region after calling pci_disable_device

2013-05-08T03:08:17Z

commit 5b0d9b218b74042ff72bf4bfda6eeb2e4bf98397 upstream. "drivers should call pci_release_region() AFTER calling pci_disable_device()" Please refer section 3.2 Request MMIO/IOP resources in Documentation/PCI/pci.txt Signed-off-by: Avinash Patil Signed-off-by: Amitkumar Karwar Signed-off-by: Yogesh Ashok Powar Signed-off-by: Bing Zhao Signed-off-by: John W. Linville Signed-off-by: Greg Kroah-Hartman

mwifiex: Use pci_release_region() instead of a pci_release_regions()

2013-05-08T03:08:17Z

commit c380aafb77b7435d010698fe3ca6d3e1cd745fde upstream. PCI regions are associated with the device using pci_request_region() call. Hence use pci_release_region() instead of pci_release_regions(). Signed-off-by: Yogesh Ashok Powar Signed-off-by: Amitkumar Karwar Signed-off-by: Avinash Patil Signed-off-by: Bing Zhao Signed-off-by: John W. Linville Signed-off-by: Greg Kroah-Hartman

iwlwifi: dvm: don't send zeroed LQ cmd

2013-05-08T03:08:17Z

commit 63b77bf489881747c5118476918cc8c29378ee63 upstream. When the stations are being restored because of unassoc RXON, the LQ cmd may not have been initialized because it is initialized only after association. Sending zeroed LQ_CMD makes the fw unhappy: it raises SYSASSERT_2078. Signed-off-by: Emmanuel Grumbach Reviewed-by: Johannes Berg [move zero_lq and make static const] Signed-off-by: Johannes Berg Signed-off-by: Greg Kroah-Hartman

iwlwifi: fix freeing uninitialized pointer

2013-05-08T03:08:17Z

commit 3309ccf7fcebceef540ebe90c65d2f94d745a45b upstream. If on iwl_dump_nic_event_log() error occurs before that function initialize buf, we process uninitiated pointer in iwl_dbgfs_log_event_read() and can hit "BUG at mm/slub.c:3409" Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=951241 Reported-by: ian.odette@eprize.com Signed-off-by: Stanislaw Gruszka Reviewed-by: Emmanuel Grumbach Signed-off-by: Johannes Berg Signed-off-by: Greg Kroah-Hartman

net: cdc_mbim: remove bogus sizeof()

2013-05-01T16:46:19Z

[ Upstream commit 32b161aa88aa40a83888a995c6e2ef81140219b1 ] The intention was to test against the constant, not the size of the constant. Signed-off-by: Bjørn Mork Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman

net: mvneta: fix improper tx queue usage in mvneta_tx()

2013-05-01T16:46:19Z

[ Upstream commit ee40a116ebf139f900c3d2e6febb8388738e96d0 ] mvneta_tx() was using a static tx queue number causing crashes as soon as a little bit of traffic was sent via the interface, because it is normally expected that the same queue should be used as in dev_queue_xmit(). As suggested by Ben Hutchings, let's use skb_get_queue_mapping() to get the proper Tx queue number, and use alloc_etherdev_mqs() instead of alloc_etherdev_mq() to create the queues. Both my Mirabox and my OpenBlocks AX3 used to crash without this patch and don't anymore with it. The issue appeared in 3.8 but became more visible after the fix allowing GSO to be enabled. Original work was done by Dmitri Epshtein and Thomas Petazzoni. I just adapted it to take care of Ben's comments. Signed-off-by: Willy Tarreau Cc: Dmitri Epshtein Cc: Thomas Petazzoni Cc: Gregory CLEMENT Cc: Ben Hutchings Tested-by: Gregory CLEMENT Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman

net: mvmdio: add select PHYLIB

2013-05-01T16:46:19Z

[ Upstream commit 2e0cbf2cc2c9371f0aa198857d799175ffe231a6 ] The mvmdio driver uses the phylib API, so it should select the PHYLIB symbol, otherwise, a build with mvmdio (but without mvneta) fails to build with undefined symbols such as mdiobus_unregister, mdiobus_free, etc. The mvneta driver does not use the phylib API directly, so it does not need to select PHYLIB. It already selects the mvmdio driver anyway. Historically, this problem is due to the fact that the PHY handling was originally part of mvneta, and was later moved to a separate driver, without updating the Kconfig select statements accordingly. And since there was no functional reason to use mvmdio without mvneta, this case was not tested. Signed-off-by: Thomas Petazzoni Reported-by: Fengguang Wu Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman