- 23 Feb, 2009 2 commits
-
-
Vasanthakumar Thiagarajan authored
[246916.338046] [246916.338048] Pid: 29265, comm: insmod Not tainted (2.6.29-rc4-wl #64) 9461DUU [246916.338051] EIP: 0060:[<c02ca274>] EFLAGS: 00010202 CPU: 0 [246916.338055] EIP is at rollback_registered+0x24/0x220 [246916.338057] EAX: 00000001 EBX: 00000000 ECX: 00000000 EDX: f122e8fc [246916.338059] ESI: 00000000 EDI: 00000000 EBP: f6595d30 ESP: f6595d1c [246916.338062] DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068 [246916.338064] Process insmod (pid: 29265, ti=f6594000 task=f7343fe0 task.ti=f6594000) [246916.338067] Stack: [246916.338068] c04a2920 22222222 f6595d48 00000000 f122f080 f6595d48 c02ca489 f122e8fc [246916.338076] f122e220 f122f080 f122e220 f6595d5c f8a03156 f122e220 f122f080 f122e220 [246916.338085] f6595d80 f87359af f122f080 00002000 f874e129 f122f150 f122f080 f6290000 [246916.338094] Call Trace: [246916.338096] [<c02ca489>] ? unregister_netdevice+0x19/0x70 [246916.338100] [<f8a03156>] ? ieee80211_unregister_hw+0x36/0xd0 [mac80211] [246916.338112] [<f87359af>] ? ath_detach+0xcf/0x250 [ath9k] [246916.338127] [<f8735d9c>] ? ath_attach+0x26c/0x740 [ath9k] [246916.338139] [<f873c33a>] ? ath_pci_probe+0x13a/0x310 [ath9k] [246916.338151] [<c0233e28>] ? _raw_spin_unlock+0x68/0x80 [246916.338158] [<c023ab8e>] ? local_pci_probe+0xe/0x10 [246916.338162] [<c023b8e0>] ? pci_device_probe+0x60/0x80 [246916.338169] [<c029e042>] ? driver_probe_device+0x82/0x1b0 [246916.338174] [<c029e1f9>] ? __driver_attach+0x89/0x90 [246916.338180] [<c029d97b>] ? bus_for_each_dev+0x4b/0x70 [246916.338184] [<c023b820>] ? pci_device_remove+0x0/0x40 [246916.338190] [<c029ded9>] ? driver_attach+0x19/0x20 [246916.338193] [<c029e170>] ? __driver_attach+0x0/0x90 [246916.338197] [<c029d317>] ? bus_add_driver+0x1b7/0x230 [246916.338203] [<c023b820>] ? pci_device_remove+0x0/0x40 [246916.338206] [<c029e399>] ? driver_register+0x69/0x140 [246916.338212] [<f859d000>] ? ath9k_init+0x0/0x54 [ath9k] [246916.338221] [<c023bb4e>] ? __pci_register_driver+0x4e/0x90 [246916.338225] [<f859d000>] ? ath9k_init+0x0/0x54 [ath9k] [246916.338232] [<f859d06b>] ? ath_pci_init+0x17/0x19 [ath9k] [246916.338238] [<f859d017>] ? ath9k_init+0x17/0x54 [ath9k] [246916.338245] [<c017148e>] ? tracepoint_update_probe_range+0x7e/0xb0 [246916.338249] [<c010111a>] ? do_one_initcall+0x2a/0x170 [246916.338252] [<c0149f26>] ? up_read+0x16/0x30 [246916.338256] [<c014aa9d>] ? __blocking_notifier_call_chain+0x4d/0x60 [246916.338265] [<c0162b1a>] ? sys_init_module+0x8a/0x1c0 [246916.338269] [<c022f888>] ? trace_hardirqs_on_thunk+0xc/0x10 [246916.338272] [<c0103ebf>] ? sysenter_do_call+0x12/0x43 [246916.338276] Code: 8d bc 27 00 00 00 00 55 89 e5 56 89 c6 53 83 ec 0c a1 74 27 4a c0 85 c0 0f 85 4b 01 00 00 e8 04 7d 00 00 85 c0 0f 84 c9 01 00 00 <8b> 86 18 03 00 00 85 c0 0f 84 86 01 00 00 83 e8 01 0f 85 71 01 [246916.338328] EIP: [<c02ca274>] rollback_registered+0x24/0x220 SS:ESP 0068:f6595d1c [246916.338335] ---[ end trace 76357c56a75ea34e ]--- Signed-off-by:
Vasanthakumar Thiagarajan <vasanth@atheros.com> Signed-off-by:
John W. Linville <linville@tuxdriver.com>
-
Andrey Borzenkov authored
With DEBUG_NOTIFIERS it results in [11330.890966] WARNING: at /home/bor/src/linux-git/kernel/notifier.c:88 notifier_call_chain+0x91/0xa0() [11330.890977] Hardware name: PORTEGE 4000 [11330.890983] Invalid notifier called! ... Without DEBUG_NOTIFIERS it most likely crashes on NULL pointer. Signed-off-by:
Andrey Borzenkov <arvidjaar@mail.ru> Acked-by:
David Kilroy <kilroyd@googlemail.com> Signed-off-by:
-
- 22 Feb, 2009 3 commits
-
-
Daniel Lezcano authored
This patch fix a double free when a network namespace fails. The previous code does a kfree of the net_generic structure when one of the init subsystem initialization fails. The 'setup_net' function does kfree(ng) and returns an error. The caller, 'copy_net_ns', call net_free on error, and this one calls kfree(net->gen), making this pointer freed twice. This patch make the code symetric, the net_alloc does the net_generic allocation and the net_free frees the net_generic. Signed-off-by:
Daniel Lezcano <daniel.lezcano@free.fr> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Daniel Lezcano authored
Fix lost set_mac_address capability. Signed-off-by:
-
Roel Kluin authored
Do not go beyond ARRAY_SIZE of ib->btx_ring Signed-off-by:
Roel Kluin <roel.kluin@gmail.com> Signed-off-by:
-
- 20 Feb, 2009 7 commits
-
-
roel kluin authored
Another error was printed one too early. Signed-off-by:
-
Randy Dunlap authored
Fix isdn/sc/shmem.c printk format warning: drivers/isdn/sc/shmem.c:57: warning: format '%d' expects type 'int', but argument 3 has type 'size_t' Signed-off-by:
Randy Dunlap <randy.dunlap@oracle.com> Signed-off-by:
-
Roel Kluin authored
With a postfix decrement timeouts will reach -1 rather than 0, so the error path does not appear. Signed-off-by:
-
Steve Glendinning authored
ethtool.h says the driver should set the magic field in get_eeprom and verify it in set_eeprom. This patch adds this functionality using an arbitary driver-specific magic value constant (0x9420). Signed-off-by:
Steve Glendinning <steve.glendinning@smsc.com> Signed-off-by:
-
Roel Kluin authored
Signed-off-by:
-
Steve Glendinning authored
Roel Kluin recently fixed several instances where variables reach -1, but 0 is tested afterwards. This patch fixes another, so the timeout will be correctly detected and a warning printed. Signed-off-by:
-
Inaky Perez-Gonzalez authored
This is a one liner change to have the driver use by default the v1.4 of the i2400m firmware instead of v1.3. The v1.4 version of the firmware has been submitted to David Woodhouse for inclusion in the linux-firmware tree and it is already available at http://linuxwimax.org/Download. The reason for this change is that the 1.3 release of the user space software and firmware has a few issues that will make it difficult to use with currently deployed commercial networks such as Xohm and Clearwire. As well, the new 1.4 release of the user space software (which matches the 1.4 firmware) has intermitent issues with the 1.3 firmware. The 1.4 release in http://linuxwimax.org/Download has been widely deployed and tested with the codebase in 2.6.29-rc, the 1.4 firmware and the 1.4 user space components. We understand it is quite late in the rc process for such a change, but would like to ask for the change to be taken into consideration. Alternatively, a user could always force feed a 1.4 firmware into a driver that doesn't have this modification by: $ cd /lib/firmware $ mv i2400m-fw-usb-1.3.sbcf i2400m-fw-usb-1.3.real.sbcf $ ln -sf i2400m-fw-usb-1.4.sbc i2400m-fw-usb-1.3.sbcf Signed-off-by:
Inaky Perez-Gonzalez <inaky@linux.intel.com> Signed-off-by:
-
- 19 Feb, 2009 8 commits
-
-
Jarek Poplawski authored
After moving mac addresses in __vlan_put_tag() skb->mac_header needs to be updated. Reported-by:
Karl Hiramoto <karl@hiramoto.org> Signed-off-by:
Jarek Poplawski <jarkao2@gmail.com> Signed-off-by:
-
Divy Le Ray authored
Add support for adapters with a PCI id equal to 0x35. Signed-off-by:
Divy Le Ray <divy@chelsio.com> Signed-off-by:
-
Ilpo Järvinen authored
This is obsolete since the passes got combined. Signed-off-by:
Ilpo Järvinen <ilpo.jarvinen@helsinki.fi> Signed-off-by:
-
Roel Kluin authored
phyid Can't be both TG3_PHY_OUI_1 and TG3_PHY_OUI_2 and TG3_PHY_OUI_3. Signed-off-by:
Matt Carlson <mcarlson@broadcom.com> Signed-off-by:
-
Roel Kluin authored
Add missing parentheses Signed-off-by:
-
Sebastian Siewior authored
mib_counters_update() also restarts the timer. So the timer is dequeued, the stats are read and then the timer is enqueued again. This is "okay" unless someone unloads the module. The locking here is also broken: mib_counters_update() grabs just a simple spinlock. The only thing the lock is good for is to protect the timer func against other callers namely mv643xx_eth_stop() && mv643xx_eth_get_ethtool_stats(). That means if the spinlock is taken via the ethtool path and than the timer kicks in then the box will lock up. Signed-off-by:
Sebastian Andrzej Siewior <sebastian@breakpoint.cc> Acked-by:
Lennert Buytenhek <buytenh@marvell.com> Signed-off-by:
-
Sebastian Siewior authored
dev_set_rx_mode() grabs netif_addr_lock_bh(): |BUG: sleeping function called from invalid context at /home/bigeasy/git/cryptodev-2.6/mm/slub.c:1599 |in_atomic(): 1, irqs_disabled(): 0, pid: 859, name: ifconfig |2 locks held by ifconfig/859: | #0: (rtnl_mutex){--..}, at: [<c0239ccc>] rtnl_lock+0x18/0x20 | #1: (_xmit_ETHER){-...}, at: [<c022d094>] dev_set_rx_mode+0x1c/0x30 |[<c029f118>] (dump_stack+0x0/0x14) from [<c003df28>] (__might_sleep+0x11c/0x13c) |[<c003de0c>] (__might_sleep+0x0/0x13c) from [<c00a8854>] (kmem_cache_alloc+0x30/0xd4) | r5:c78093a0 r4:c034a47c |[<c00a8824>] (kmem_cache_alloc+0x0/0xd4) from [<c01a5fd0>] (mv643xx_eth_set_rx_mode+0x70/0x188) |[<c01a5f60>] (mv643xx_eth_set_rx_mode+0x0/0x188) from [<c022ced0>] (__dev_set_rx_mode+0x40/0xac) |[<c022ce90>] (__dev_set_rx_mode+0x0/0xac) from [<c022d09c>] (dev_set_rx_mode+0x24/0x30) | r6:00001043 r5:c78090f8 r4:c7809000 |[<c022d078>] (dev_set_rx_mode+0x0/0x30) from [<c02304c4>] (dev_open+0xe4/0x114) | r5:c7809350 r4:c7809000 |[<c02303e0>] (dev_open+0x0/0x114) from [<c022fd18>] (dev_change_flags+0xb0/0x190) | r5:00000041 r4:c7809000 |[<c022fc68>] (dev_change_flags+0x0/0x190) from [<c0270250>] (devinet_ioctl+0x2f0/0x710) | r7:c7221e70 r6:c7aadb00 r5:00000000 r4:00000001 |[<c026ff60>] (devinet_ioctl+0x0/0x710) from [<c02717c8>] (inet_ioctl+0xd4/0x110) |[<c02716f4>] (inet_ioctl+0x0/0x110) from [<c021fb74>] (sock_ioctl+0x1f4/0x254) | r4:c7242b40 |[<c021f980>] (sock_ioctl+0x0/0x254) from [<c00b8160>] (vfs_ioctl+0x38/0x98) | r6:beec9bb8 r5:00008914 r4:c7242b40 |[<c00b8128>] (vfs_ioctl+0x0/0x98) from [<c00b873c>] (do_vfs_ioctl+0x484/0x4d4) | r6:00008914 r5:c7242b40 r4:c74db1c0 |[<c00b82b8>] (do_vfs_ioctl+0x0/0x4d4) from [<c00b87cc>] (sys_ioctl+0x40/0x64) |[<c00b878c>] (sys_ioctl+0x0/0x64) from [<c00269a0>] (ret_fast_syscall+0x0/0x2c) |[42949399.520000] r7:00000036 r6:beec9c80 r5:00000041 r4:beec9bb8 Signed-off-by: -
Jie Yang authored
Supporting AR8131, and AR8132. Signed-off-by:
Jie Yang <jie.yang@atheros.com> Signed-off-by:
-
- 18 Feb, 2009 1 commit
-
-
David S. Miller authored
A long time ago we had bugs, primarily in TCP, where we would modify skb->truesize (for TSO queue collapsing) in ways which would corrupt the socket memory accounting. skb_truesize_check() was added in order to try and catch this error more systematically. However this debugging check has morphed into a Frankenstein of sorts and these days it does nothing other than catch false-positives. Signed-off-by:
-
- 16 Feb, 2009 1 commit
-
-
Tobias Diedrich authored
Commit f55c21fd ("forcedeth: call restore mac addr in nv_shutdown path"), which was introduced to fix the regression tracked at http://bugzilla.kernel.org/show_bug.cgi?id=11358 causes the wake-on-lan mac to be reversed in the shutdown path. Apparently the forcedeth situation is rather messy in that the mac we need to writeback for a subsequent modprobe to work is exactly the reverse of what is needed for proper wake-on-lan. The following patch explains the situation in the comments and makes the call to nv_restore_mac_addr() conditional (only called if we are not really going for poweroff). Tobias Diedrich wrote: > Hmm, I had not tried WOL for some time. > With 2.6.29-rc3 is see the following behaviour: > > State WOL Behaviour > ------------------------------ > shutdown reversed MAC > disk/shutdown reversed MAC > disk/platform OK > > Apparently nv_restore_mac_addr() restores the MAC in the wrong order > for WOL (at least for my PCI_DEVICE_ID_NVIDIA_NVENET_15). platform > works, because the MAC is not touched in the nv_suspend() path. > > A possible fix might be to only call nv_restore_mac_addr() if > system_state != SYSTEM_POWER_OFF. With the following patch: shutdown OK disk/shutdown OK disk/platform OK kexec OK Signed-off-by:
Tobias Diedrich <ranma+kernel@tdiedrich.de> Tested-by:
Philipp Matthias Hahn <pmhahn@titan.lahn.de> Signed-off-by:
-
- 13 Feb, 2009 18 commits
-
-
Inaky Perez-Gonzalez authored
When a non-wimax interface is looked up by the stack, a bad pointer is returned when the looked-up interface is not found in the list (of registered WiMAX interfaces). This causes an oops in the caller when trying to use the pointer. Fix by properly setting the pointer to NULL if we don't exit from the list_for_each() with a found entry. Signed-off-by:
-
Clément Lecigne authored
In function sock_getsockopt() located in net/core/sock.c, optval v.val is not correctly initialized and directly returned in userland in case we have SO_BSDCOMPAT option set. This dummy code should trigger the bug: int main(void) { unsigned char buf[4] = { 0, 0, 0, 0 }; int len; int sock; sock = socket(33, 2, 2); getsockopt(sock, 1, SO_BSDCOMPAT, &buf, &len); printf("%x%x%x%x\n", buf[0], buf[1], buf[2], buf[3]); close(sock); } Here is a patch that fix this bug by initalizing v.val just after its declaration. Signed-off-by:Clément Lecigne <clement.lecigne@netasq.com> Signed-off-by:
-
Yang Hongyang authored
When compile the latest kernel on IA64 platform,I got a warning: drivers/net/netxen/netxen_nic_main.c:203: warning: label ‘set_32_bit_mask’ defined but not used We do not need label ‘set_32_bit_mask’ on IA64 platform,So move it to #else. Signed-off-by:
Yang Hongyang <yanghy@cn.fujitsu.com> Signed-off-by:
-
Michael Chan authored
Signed-off-by:
Michael Chan <mchan@broadcom.com> Signed-off-by:
-
Michael Chan authored
If errors are reported on a frame descriptor, we need to account for the buffer pages that may have been used for this error packet and recycle them. Otherwise, we may get the wrong pages for the next packet. Signed-off-by:
Benjamin Li <benli@broadcom.com> Signed-off-by:
-
Michael Chan authored
New firmware fixes a data corruption issue when receiving and placing jumbo frames into host buffers. In some cases, the buffer descriptor is not updated correctly and this will lead to the driver linking the wrong number of pages into the SKB. Signed-off-by:
-
Michael Chan authored
New firmware fixes a data corruption issue when receiving and placing jumbo frames into host buffers. In some cases, the buffer descriptor is not updated correctly and this will lead to the driver linking the wrong number of pages into the SKB. Signed-off-by:
-
Roel Kluin authored
Ensure that we do not set pcb->data.raw beyond its size, print an error message and return false if we attempt to. A timout message was printed one too early. Signed-off-by:
-
Andrew Morton authored
cn_test_timer_func() is a timer handler and can never use GFP_KERNEL - there's no point in using gfp_any() here. Also, use setup_timer(). Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
-
Andrew Morton authored
The problem is that in_atomic() will return false inside spinlocks if CONFIG_PREEMPT=n. This will lead to deadlockable GFP_KERNEL allocations from spinlocked regions. Secondly, if CONFIG_PREEMPT=y, this bug solves itself because networking will instead use GFP_ATOMIC from this callsite. Hence we won't get the might_sleep() debugging warnings which would have informed us of the buggy callsites. Solve both these problems by switching to in_interrupt(). Now, if someone runs a gfp_any() allocation from inside spinlock we will get the warning if CONFIG_PREEMPT=y. I reviewed all callsites and most of them were too complex for my little brain and none of them documented their interface requirements. I have no idea what this patch will do. Signed-off-by:
-
Roel Kluin authored
If no prior break occurs, cnt reaches 101 after the loop, so we are still able to change speed when cnt has become 100. Signed-off-by:
-
Dhananjay Phadke authored
Remove workaround for pcie bug in early revisions of NX3031 (rev 41 or earlier). This is taken care of during firmware init. The workaround required writing pcie config reg of every pcie function on a card, not all of which are enabled. Signed-off-by:
Dhananjay Phadke <dhananjay@netxen.com> Signed-off-by:
-
Roel Kluin authored
With while (--i > 0) { ... } i reaches 0; print when lance_open() fails Signed-off-by: -
Ron Mercer authored
The large receive buffer queue is not properly tracking the current index in the case where an early exit occurs. This can happen when a page alloc or dma mapping fails. If this occurs the queue will get out of sync and invalid indexes can be written to the hardware. Signed-off-by:
Ron Mercer <ron.mercer@qlogic.com> Signed-off-by:
-
Ron Mercer authored
Receive packets were only scaling across 2 of the receive queues. The value was hardcoded to 2 instead of being based on how many rx queues were running. Signed-off-by:
-
Ron Mercer authored
Moved the buffer mapping to a point after TSO logic has modified the iph->check field. We were seeing stale data on the PCIe bus. Signed-off-by:
-
Ron Mercer authored
We put the skb back if we can't get mapping for it. We don't want unmapped buffers on our receive buffer queue. Signed-off-by:
-
Ron Mercer authored
We put the page back if we can't get mapping for it. We don't want unmapped buffers on our receive buffer queue. Signed-off-by:
-
