- 27 Aug, 2008 10 commits
-
-
Alan Cox authored
Signed-off-by:
Alan 'pass the paper bags' Cox <alan@redhat.com> Signed-off-by:
Jeff Garzik <jgarzik@redhat.com>
-
Heiko Schocher authored
Signed-off-by:
Heiko Schocher <hs@denx.de> Signed-off-by:
Vitaly Bordug <vitb@kernel.crashing.org> Signed-off-by:
Kumar Gala <galak@kernel.crashing.org> Signed-off-by:
-
Alexander Duyck authored
The real_num_tx_queues was not being set when in MSI-X only mode. This patch corrects that path so all interrupt types are correctly configured. Signed-off-by:
Alexander Duyck <alexander.h.duyck@intel.com> Signed-off-by:
Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by:
-
Alexander Duyck authored
Ethtool -d is reading the EICR and ICR registers which is currently clearing these registers and masking off interrupts. To prevent this we read the EICS and ICS equivilents as they can be read without clearing or masking. Signed-off-by:
-
Alexander Duyck authored
Set the EICS bit for each of the RX queues at least once every 2 seconds to prevent the rx queues from stalling. Signed-off-by:
-
Francois Romieu authored
The leak hurts with swiotlb and jumbo frames. Fix http://bugzilla.kernel.org/show_bug.cgi?id=9468. Heavily hinted by Ilpo Järvinen <ilpo.jarvinen@helsinki.fi>. Signed-off-by:
Francois Romieu <romieu@fr.zoreil.com> Tested-by:
Alistair John Strachan <alistair@devzero.co.uk> Tested-by:
Timothy J Fontaine <tjfontaine@atxconsulting.com> Cc: Edward Hsu <edward_hsu@realtek.com.tw> Signed-off-by:
-
Brice Goglin authored
Update myri10ge version string to 1.4.3-1.358. Signed-off-by:
Brice Goglin <brice@myri.com> Signed-off-by:
-
Alexander Duyck authored
VLAN filtering is broken, due to reading the incorrect register for the VLAN filtering settings. Fixed by reading/writing the correct register. Signed-off-by:
Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by:
-
Mike Frysinger authored
so update things accordingly Signed-off-by:
Mike Frysinger <vapier.adi@gmail.com> Signed-off-by:
Bryan Wu <cooloney@kernel.org> Signed-off-by:
-
-
- 25 Aug, 2008 12 commits
-
-
Eilon Greenstein authored
Version update Signed-off-by:
Eilon Greenstein <eilong@broadcom.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Yitchak Gertner authored
The multi queue support is still disabled by default for the bnx2x (needs some more testing and validation), but there are 2 obvious bug in it which are fixed in this patch Signed-off-by:
Yitchak Gertner <gertner@broadcom.com> Signed-off-by:
-
Yitchak Gertner authored
Fixing the order of enabling and disabling NAPI and the interrupts Signed-off-by:
-
Yitchak Gertner authored
Load failures were not handled correctly Signed-off-by:
-
Yitchak Gertner authored
The TPA initialization is part of the FW internal memory initialization and so it is moved to the appropriate function Signed-off-by:
-
Eilon Greenstein authored
Increasing the lock timeout to 5 seconds instead of 1 second to minimize the chance of failures due to timeout Signed-off-by:
-
Eilon Greenstein authored
After iSCSI boot, the HW lock should only protect the flag so only the first function will reset the chip and not then entire chip reset process Signed-off-by:
-
Eilon Greenstein authored
The A1021G board is also using the fan failure mechanism in the same way the A1022G board does Signed-off-by:
-
Eilon Greenstein authored
The has Rx work check was wrong: when the FW was at the end of the page, the driver was already at the beginning of the next page. Since the check only validated that both driver and FW are pointing to the same place, it concluded that there is still work to be done. This caused some serious issues including long latency results on ping-pong test and lockups while unloading the driver in that condition. Signed-off-by:
-
Al Viro authored
Braino: net.ipv6 in ipv6 skeleton has no business in rotable class Signed-off-by:
Al Viro <viro@zeniv.linux.org.uk> Signed-off-by:
-
Al Viro authored
net.ipv4.neigh should be a part of skeleton to avoid ordering problems Signed-off-by:
-
Vlad Yasevich authored
The structure used for SCTP_AUTH_KEY option contains a length that needs to be verfied to prevent buffer overflow conditions. Spoted by Eugene Teo <eteo@redhat.com>. Signed-off-by:
Vlad Yasevich <vladislav.yasevich@hp.com> Signed-off-by:
-
- 24 Aug, 2008 6 commits
-
-
Lennert Buytenhek authored
Signed-off-by:Lennert Buytenhek <buytenh@marvell.com>
-
Lennert Buytenhek authored
The mv643xx_eth hardware ignores the lower three bits of the buffer size field in receive descriptors, causing the reception of full-sized packets to fail at some MTUs. Fix this by rounding the size of allocated receive buffers up to a multiple of eight bytes. While we are at it, add a bit of extra space to each receive buffer so that we can handle multiple vlan tags on ingress. Signed-off-by: -
Lennert Buytenhek authored
When we are low on memory, the assumption that every descriptor in the receive ring will have an skbuff associated with it does not hold. rxq_process() was assuming that if the receive descriptor it is working on is not owned by the hardware, it can safely be processed and handed to the networking stack. But a descriptor in the receive ring not being owned by the hardware can also happen when we are low on memory and did not manage to refill the receive ring fully. This patch changes rxq_process()'s bailout condition from "the first receive descriptor to be processed is owned by the hardware" to "the first receive descriptor to be processed is owned by the hardware OR the number of valid receive descriptors in the ring is zero". Signed-off-by: -
Lennert Buytenhek authored
Nicolas Pitre noted that mv643xx_eth_poll was incorrectly using non-IRQ-safe locks while checking whether to wake up the netdevice's transmit queue. Convert the locking to *_irq() variants, since we are running from softirq context where interrupts are enabled. Signed-off-by: -
Lennert Buytenhek authored
Commit 12e4ab79 ("mv643xx_eth: be more agressive about RX refill") changed the condition for the receive out-of-memory timer to be scheduled from "the receive ring is empty" to "the receive ring is not full". This can lead to a situation where the receive out-of-memory timer is pending because a previous rxq_refill() didn't manage to refill the receive ring entirely as a result of being out of memory, and rxq_refill() is then called again as a side effect of a packet receive interrupt, and that rxq_refill() call then again does not succeed to refill the entire receive ring with fresh empty skbuffs because we are still out of memory, and then tries to call add_timer() on the already scheduled out-of-memory timer. This patch fixes this issue by changing the add_timer() call in rxq_refill() to a mod_timer() call. If the OOM timer was not already scheduled, this will behave as before, whereas if it was already scheduled, this patch will push back its firing time a bit, which is safe because we've (unsuccessfully) attempted to refill the receive ring just before we do this. Signed-off-by:
-
Lennert Buytenhek authored
When a receive interrupt occurs, mv643xx_eth would first process the receive descriptors and then ACK the receive interrupt, instead of the other way round. This would leave a small race window between processing the last receive descriptor and clearing the receive interrupt status in which a new packet could come in, which would then 'rot' in the receive ring until the next receive interrupt would come in. Fix this by ACKing (clearing) the receive interrupt condition before processing the receive descriptors. Signed-off-by:
-
- 23 Aug, 2008 2 commits
-
-
Stephen Hemminger authored
This fixes a problem spotted with zebra, but not sure if it is necessary a kernel problem. With IPV6 when an address is added to an interface, Zebra creates a duplicate RIB entry, one as a connected route, and other as a kernel route. When an address is added to an interface the RTN_NEWADDR message causes Zebra to create a connected route. In IPV4 when an address is added to an interface a RTN_NEWROUTE message is set to user space with the protocol RTPROT_KERNEL. Zebra ignores these messages, because it already has the connected route. The problem is that route created in IPV6 has route protocol == RTPROT_BOOT. Was this a design decision or a bug? This fixes it. Same patch applies to both net-2.6 and stable. Signed-off-by:
Stephen Hemminger <shemminger@vyatta.com> Signed-off-by:
-
Denis V. Lunev authored
Pass namespace into icmp_xmit_lock, obtain socket inside and return it as a result for caller. Thanks Alexey Dobryan for this report: Steps to reproduce: CONFIG_PREEMPT=y CONFIG_DEBUG_PREEMPT=y tracepath <something> BUG: using smp_processor_id() in preemptible [00000000] code: tracepath/3205 caller is icmp_sk+0x15/0x30 Pid: 3205, comm: tracepath Not tainted 2.6.27-rc4 #1 Call Trace: [<ffffffff8031af14>] debug_smp_processor_id+0xe4/0xf0 [<ffffffff80409405>] icmp_sk+0x15/0x30 [<ffffffff8040a17b>] icmp_send+0x4b/0x3f0 [<ffffffff8025a415>] ? trace_hardirqs_on_caller+0xd5/0x160 [<ffffffff8025a4ad>] ? trace_hardirqs_on+0xd/0x10 [<ffffffff8023a475>] ? local_bh_enable_ip+0x95/0x110 [<ffffffff804285b9>] ? _spin_unlock_bh+0x39/0x40 [<ffffffff8025a26c>] ? mark_held_locks+0x4c/0x90 [<ffffffff8025a4ad>] ? trace_hardirqs_on+0xd/0x10 [<ffffffff8025a415>] ? trace_hardirqs_on_caller+0xd5/0x160 [<ffffffff803e91b4>] ip_fragment+0x8d4/0x900 [<ffffffff803e7030>] ? ip_finish_output2+0x0/0x290 [<ffffffff803e91e0>] ? ip_finish_output+0x0/0x60 [<ffffffff803e6650>] ? dst_output+0x0/0x10 [<ffffffff803e922c>] ip_finish_output+0x4c/0x60 [<ffffffff803e92e3>] ip_output+0xa3/0xf0 [<ffffffff803e68d0>] ip_local_out+0x20/0x30 [<ffffffff803e753f>] ip_push_pending_frames+0x27f/0x400 [<ffffffff80406313>] udp_push_pending_frames+0x233/0x3d0 [<ffffffff804067d1>] udp_sendmsg+0x321/0x6f0 [<ffffffff8040d155>] inet_sendmsg+0x45/0x80 [<ffffffff803b967f>] sock_sendmsg+0xdf/0x110 [<ffffffff8024a100>] ? autoremove_wake_function+0x0/0x40 [<ffffffff80257ce5>] ? validate_chain+0x415/0x1010 [<ffffffff8027dc10>] ? __do_fault+0x140/0x450 [<ffffffff802597d0>] ? __lock_acquire+0x260/0x590 [<ffffffff803b9e55>] ? sockfd_lookup_light+0x45/0x80 [<ffffffff803ba50a>] sys_sendto+0xea/0x120 [<ffffffff80428e42>] ? _spin_unlock_irqrestore+0x42/0x80 [<ffffffff803134bc>] ? __up_read+0x4c/0xb0 [<ffffffff8024e0c6>] ? up_read+0x26/0x30 [<ffffffff8020b8bb>] system_call_fastpath+0x16/0x1b icmp6_sk() is similar. Signed-off-by:
Denis V. Lunev <den@openvz.org> Signed-off-by:
-
- 22 Aug, 2008 1 commit
-
-
Jarek Poplawski authored
Since some qdiscs call qdisc_tree_decrease_qlen() (so qdisc_lookup()) without rtnl_lock(), adding and deleting from a qdisc list needs additional locking. This patch adds global spinlock qdisc_list_lock and wrapper functions for modifying the list. It is considered as a temporary solution until hfsc_dequeue(), netem_dequeue() and tbf_dequeue() (or qdisc_tree_decrease_qlen()) are redone. With feedback from Herbert Xu and David S. Miller. Signed-off-by:
Jarek Poplawski <jarkao2@gmail.com> Acked-by:
Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by:
-
- 21 Aug, 2008 3 commits
-
-
Jarek Poplawski authored
dev_deactivate() can skip rescheduling of a qdisc by qdisc_watchdog() or other timer calling netif_schedule() after dev_queue_deactivate(). We prevent this checking aliveness before scheduling the timer. Since during deactivation the root qdisc is available only as qdisc_sleeping additional accessor qdisc_root_sleeping() is created. With feedback from Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by:
-
Vlad Yasevich authored
All of the SCTP-AUTH socket options could cause a panic if the extension is disabled and the API is envoked. Additionally, there were some additional assumptions that certain pointers would always be valid which may not always be the case. This patch hardens the API and address all of the crash scenarios. Signed-off-by:
-
Linus Torvalds authored
-
- 20 Aug, 2008 6 commits
-
-
Al Viro authored
After commit a97c9bf3 (fix cramfs making duplicate entries in inode cache) in kernel 2.6.14, named-pipe on cramfs does not work properly. It seems the commit make all named-pipe on cramfs share their inode (and named-pipe buffer). Make ..._test() refuse to merge inodes with ->i_ino == 1, take inode setup back to get_cramfs_inode() and make ->drop_inode() evict ones with ->i_ino == 1 immediately. Reported-by:
Atsushi Nemoto <anemo@mba.ocn.ne.jp> Cc: Al Viro <viro@zeniv.linux.org.uk> Cc: <stable@kernel.org> [2.6.14 and later] Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Ian Campbell authored
Fixes kernel BUG at lib/radix-tree.c:473. Previously the handler was incidentally provided by tmpfs but this was removed with: commit 14fcc23f Author: Hugh Dickins <hugh@veritas.com> Date: Mon Jul 28 15:46:19 2008 -0700 tmpfs: fix kernel BUG in shmem_delete_inode relying on this behaviour was incorrect in any case and the BUG also appeared when the device node was on an ext3 filesystem. v2: override a_ops at open() time rather than mmap() time to minimise races per AKPM's concerns. Signed-off-by:
Ian Campbell <ijc@hellion.org.uk> Cc: Jaya Kumar <jayakumar.lkml@gmail.com> Cc: Nick Piggin <npiggin@suse.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Hugh Dickins <hugh@veritas.com> Cc: Johannes Weiner <hannes@saeurebad.de> Cc: Jeremy Fitzhardinge <jeremy@goop.org> Cc: Kel Modderman <kel@otaku42.de> Cc: Markus Armbruster <armbru@redhat.com> Cc: Krzysztof Helt <krzysztof.h1@poczta.fm> Cc: <stable@kernel.org> [14fcc23f is in 2.6.25.14 and 2.6.26.1] Signed-off-by:
-
Anton Vorontsov authored
On a PowerPC board with ds1374 RTC I'm getting this error while RTC tries to probe: rtc-ds1374 0-0068: unable to request IRQ This happens because I2C probing code (drivers/of/of_i2c.c) is specifying IRQ0 for 'no irq' case, which is correct. The driver handles this incorrectly, though. This patch fixes it. Signed-off-by:
Anton Vorontsov <avorontsov@ru.mvista.com> Cc: David Brownell <david-b@pacbell.net> Cc: Alessandro Zummo <a.zummo@towertech.it> Acked-by:
Peter Korsgaard <jacmet@sunsite.dk> Signed-off-by:
-
Nick Piggin authored
XIP can call into get_xip_mem concurrently with the same file,offset with create=1. This usually maps down to get_block, which expects the page lock to prevent such a situation. This causes ext2 to explode for one reason or another. Serialise those calls for the moment. For common usages today, I suspect get_xip_mem rarely is called to create new blocks. In future as XIP technologies evolve we might need to look at which operations require scalability, and rework the locking to suit. Signed-off-by:
Nick Piggin <npiggin@suse.de> Cc: Jared Hulbert <jaredeh@gmail.com> Acked-by:
Carsten Otte <cotte@freenet.de> Cc: Hugh Dickins <hugh@veritas.com> Signed-off-by:
-
Nick Piggin authored
XIP has a race between sparse pages being inserted into page tables, and sparse pages being zapped when its time to put a non-sparse page in. What can happen is that a process can be left with a dangling sparse page in a MAP_SHARED mapping, while the rest of the world sees the non-sparse version. Ie. data corruption. Guard these operations with a seqlock, making fault-in-sparse-pages the slowpath, and try-to-unmap-sparse-pages the fastpath. Signed-off-by:
-
Nick Piggin authored
There is a race with dirty page accounting where a page may not properly be accounted for. clear_page_dirty_for_io() calls page_mkclean; then TestClearPageDirty. page_mkclean walks the rmaps for that page, and for each one it cleans and write protects the pte if it was dirty. It uses page_check_address to find the pte. That function has a shortcut to avoid the ptl if the pte is not present. Unfortunately, the pte can be switched to not-present then back to present by other code while holding the page table lock -- this should not be a signal for page_mkclean to ignore that pte, because it may be dirty. For example, powerpc64's set_pte_at will clear a previously present pte before setting it to the desired value. There may also be other code in core mm or in arch which do similar things. The consequence of the bug is loss of data integrity due to msync, and loss of dirty page accounting accuracy. XIP's __xip_unmap could easily also be unreliable (depending on the exact XIP locking scheme), which can lead to data corruption. Fix this by having an option to always take ptl to check the pte in page_check_address. It's possible to retain this optimization for page_referenced and try_to_unmap. Signed-off-by:
Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by:
-
