aboutsummaryrefslogtreecommitdiff
path: root/drivers/infiniband
AgeCommit message (Collapse)AuthorFilesLines
2011-01-28RDMA/amso1100: Fix compile warningsRalf Thielow1-3/+3
Fix compile warnings on 32-bit by using "0" instead of "(u64) NULL" to assign to "c2_vq_req->reply_msg". Signed-off-by: Ralf Thielow <[email protected]> [ Change from "(unsigned long) NULL" to plain old "0" as suggested by Bart Van Assche <[email protected]>. - Roland ] Signed-off-by: Roland Dreier <[email protected]>
2011-01-28RDMA/cxgb4: Set the correct device physical function for iWARP connectionsSteve Wise1-1/+1
The PF passed to FW was 0, causing PCI failures in an SR-IOV environment. Signed-off-by: Steve Wise <[email protected]> Cc: <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-28RDMA/cxgb4: Limit MAXBURST EQ context field to 256BSteve Wise1-2/+2
MAXBURST cannot exceed 256B for on-chip queues. With a 512B MAXBURST, we can lock up the chip. Signed-off-by: Steve Wise <[email protected]> Cc: <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-28IB/qib: Hold link for TX SERDES settingsMitko Haralanov1-19/+11
Hold the IB link at DISABLED until we get the correct TX settings on mezz boards. Signed-off-by: Mike Marciniszyn <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-20kconfig: rename CONFIG_EMBEDDED to CONFIG_EXPERTDavid Rientjes2-2/+2
The meaning of CONFIG_EMBEDDED has long since been obsoleted; the option is used to configure any non-standard kernel with a much larger scope than only small devices. This patch renames the option to CONFIG_EXPERT in init/Kconfig and fixes references to the option throughout the kernel. A new CONFIG_EMBEDDED option is added that automatically selects CONFIG_EXPERT when enabled and can be used in the future to isolate options that should only be considered for embedded systems (RISC architectures, SLOB, etc). Calling the option "EXPERT" more accurately represents its intention: only expert users who understand the impact of the configuration changes they are making should enable it. Reviewed-by: Ingo Molnar <[email protected]> Acked-by: David Woodhouse <[email protected]> Signed-off-by: David Rientjes <[email protected]> Cc: Greg KH <[email protected]> Cc: "David S. Miller" <[email protected]> Cc: Jens Axboe <[email protected]> Cc: Arnd Bergmann <[email protected]> Cc: Robin Holt <[email protected]> Cc: <[email protected]> Signed-off-by: Andrew Morton <[email protected]> Signed-off-by: Linus Torvalds <[email protected]>
2011-01-17Merge branch 'for-linus' of ↵Linus Torvalds31-122/+320
git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband * 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband: RDMA: Update workqueue usage RDMA/nes: Fix incorrect SFP+ link status detection on driver init RDMA/nes: Fix SFP+ link down detection issue with switch port disable RDMA/nes: Generate IB_EVENT_PORT_ERR/PORT_ACTIVE events RDMA/nes: Fix bonding on iw_nes IB/srp: Test only once whether iu allocation succeeded IB/mlx4: Handle protocol field in multicast table RDMA: Use vzalloc() to replace vmalloc()+memset(0) mlx4_{core, ib, en}: Fix driver when sizeof (phys_addr_t) > sizeof (long) IB/mthca: Fix driver when sizeof (phys_addr_t) > sizeof (long)
2011-01-16Merge branches 'misc', 'mlx4', 'mthca', 'nes' and 'srp' into for-nextRoland Dreier14-38/+269
2011-01-16RDMA: Update workqueue usageTejun Heo12-50/+36
* ib_wq is added, which is used as the common workqueue for infiniband instead of the system workqueue. All system workqueue usages including flush_scheduled_work() callers are converted to use and flush ib_wq. * cancel_delayed_work() + flush_scheduled_work() converted to cancel_delayed_work_sync(). * qib_wq is removed and ib_wq is used instead. This is to prepare for deprecation of flush_scheduled_work(). Signed-off-by: Tejun Heo <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-16RDMA/nes: Fix incorrect SFP+ link status detection on driver initMaciej Sosnowski1-4/+39
During iw_nes initialization the link status for SFP+ PHY is always detected as "up" regardless of real state (cable either connected or disconnected). Add SFP+ PHY specific link status detection to the iw_nes initialization procedure. Use link status recheck for netdev_open to detect delayed state updates. Signed-off-by: Maciej Sosnowski <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-16RDMA/nes: Fix SFP+ link down detection issue with switch port disableMaciej Sosnowski4-0/+100
In case of SFP+ PHY, link status check at interrupt processing can give false results. For proper link status change detection a delayed recheck is needed to give nes registers time to settle. Add a periodic link status recheck scheduled at interrupt to detect potential delayed registers state changes. Addresses: http://bugs.openfabrics.org/bugzilla/show_bug.cgi?id=2117 Signed-off-by: Maciej Sosnowski <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-16RDMA/nes: Generate IB_EVENT_PORT_ERR/PORT_ACTIVE eventsMaciej Sosnowski4-9/+83
Depending on link state change, IB_EVENT_PORT_ERR or IB_EVENT_PORT_ACTIVE should be generated when handling MAC interrupts. Plugging in a cable happens to result in series of interrupts changing driver's link state a number of times before finally staying at link up (e.g. link up, link down, link up, link down, ..., link up). To prevent sending series of redundant IB_EVENT_PORT_ACTIVE and IB_EVENT_PORT_ERR events, we use a timer to debounce them in nes_port_ibevent(). Signed-off-by: Maciej Sosnowski <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-16RDMA/nes: Fix bonding on iw_nesMaciej Sosnowski2-6/+26
Enable configuring bonds on nes devices by adding missing support for master net_device to the driver. Signed-off-by: Maciej Sosnowski <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-13IB/srp: Test only once whether iu allocation succeededBart Van Assche1-8/+7
Merge the two tests in srp_queuecommand() of whether information unit allocation succeeded into one. An intended side effect of this change is that we fix the warning: drivers/infiniband/ulp/srp/ib_srp.c: In function 'srp_queuecommand': drivers/infiniband/ulp/srp/ib_srp.c:1116: warning: 'req' may be used uninitialized in this function (seen with CONFIG_CC_OPTIMIZE_FOR_SIZE=y at least with gcc 4.4.4) Signed-off-by: Bart Van Assche <[email protected]> Acked-by: David Dillow <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-13Merge branch 'for-next' of ↵Linus Torvalds3-3/+2
git://git.kernel.org/pub/scm/linux/kernel/git/jikos/trivial * 'for-next' of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/trivial: (43 commits) Documentation/trace/events.txt: Remove obsolete sched_signal_send. writeback: fix global_dirty_limits comment runtime -> real-time ppc: fix comment typo singal -> signal drivers: fix comment typo diable -> disable. m68k: fix comment typo diable -> disable. wireless: comment typo fix diable -> disable. media: comment typo fix diable -> disable. remove doc for obsolete dynamic-printk kernel-parameter remove extraneous 'is' from Documentation/iostats.txt Fix spelling milisec -> ms in snd_ps3 module parameter description Fix spelling mistakes in comments Revert conflicting V4L changes i7core_edac: fix typos in comments mm/rmap.c: fix comment sound, ca0106: Fix assignment to 'channel'. hrtimer: fix a typo in comment init/Kconfig: fix typo anon_inodes: fix wrong function name in comment fix comment typos concerning "consistent" poll: fix a typo in comment ... Fix up trivial conflicts in: - drivers/net/wireless/iwlwifi/iwl-core.c (moved to iwl-legacy.c) - fs/ext4/ext4.h Also fix missed 'diabled' typo in drivers/net/bnx2x/bnx2x.h while at it.
2011-01-12IB/mlx4: Handle protocol field in multicast tableAleksey Senin1-4/+5
The newest device firmware stores IB vs. Ethernet protocol in two bits in members_count field of multicast group table (0: Infiniband, 1: Ethernet). When changing the QP members count for a multicast group, it important not to reset this information. When calling multicast attach first time, the protocol type should be specified. In this patch we always set it IB, but in the future we will handle Ethernet too. When looking for a QP, the protocol type shoud be checked too. Signed-off-by: Aleksey Senin <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-12RDMA: Use vzalloc() to replace vmalloc()+memset(0)Joe Perches8-34/+15
Signed-off-by: Joe Perches <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-12mlx4_{core, ib, en}: Fix driver when sizeof (phys_addr_t) > sizeof (long)Roland Dreier1-1/+2
Some systems have PCI addresses that don't fit in unsigned long (eg some 32-bit PowerPC 440 systems have 36-bit bus addresses). Fix up mlx4 drivers by using phys_addr_t where appropriate, so we don't truncate any PCI resource addresses before ioremapping them. Signed-off-by: Roland Dreier <[email protected]>
2011-01-11IB/mthca: Fix driver when sizeof (phys_addr_t) > sizeof (long)John L. Burr5-6/+7
Some systems have PCI addresses that don't fit in unsigned long (eg some 32-bit PowerPC 440 systems have 36-bit bus addresses). Fix up the driver by using phys_addr_t where appropriate, so we don't truncate any PCI resource addresses before ioremapping them. Signed-off-by: John L. Burr <[email protected]> [ Update to apply to current driver source. - Roland ] Signed-off-by: Roland Dreier <[email protected]>
2011-01-11Merge branch 'for-linus' of ↵Linus Torvalds35-618/+886
git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband * 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband: (42 commits) IB/qib: Fix refcount leak in lkey/rkey validation IB/qib: Improve SERDES tunning on QMH boards IB/qib: Unnecessary delayed completions on RC connection IB/qib: Issue pre-emptive NAKs on eager buffer overflow IB/qib: RDMA lkey/rkey validation is inefficient for large MRs IB/qib: Change QPN increment IB/qib: Add fix missing from earlier patch IB/qib: Change receive queue/QPN selection IB/qib: Fix interrupt mitigation IB/qib: Avoid duplicate writes to the rcv head register IB/qib: Add a few new SERDES tunings IB/qib: Reset packet list after freeing IB/qib: New SERDES init routine and improvements to SI quality IB/qib: Clear WAIT_SEND flags when setting QP to error state IB/qib: Fix context allocation with multiple HCAs IB/qib: Fix multi-Florida HCA host panic on reboot IB/qib: Handle transitions from ACTIVE_DEFERRED to ACTIVE better IB/qib: UD send with immediate receive completion has wrong size IB/qib: Set port physical state even if other fields are invalid IB/qib: Generate completion callback on errors ...
2011-01-10Merge branches 'cxgb4', 'ipath', 'ipoib', 'mlx4', 'mthca', 'nes', 'qib' and ↵Roland Dreier30-527/+884
'srp' into for-next
2011-01-10IB/qib: Fix refcount leak in lkey/rkey validationMike Marciniszyn1-16/+14
The mr optimization introduced a reference count leak on an exception test. The lock/refcount manipulation is moved down and the problematic exception test now calls bail to insure that the lock is released. Additional fixes as suggested by Ralph Campbell <[email protected]>: - reduce lock scope of dma regions - use explicit values on returns vs. automatic ret value Signed-off-by: Mike Marciniszyn <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/qib: Improve SERDES tunning on QMH boardsMike Marciniszyn1-3/+23
Improve the QMH SERDES tunning on initial driver load by having the driver go through a link state change. Signed-off-by: Mike Marciniszyn <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/qib: Unnecessary delayed completions on RC connectionMike Marciniszyn1-0/+24
Currently on receipt of a response message (ACKs, RDMA Response, Atomic Responses etc.) if the SDMA completion counter is not advanced the driver delays the completion of the WQE. In most cases this is overly pessimistic as the response (ACK) to a previously transmitted send implies that the send is complete. Ensure that SDMA queue is progressed appropriately before determining if a send has delayed completions. Signed-off-by: Mike Marciniszyn <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/qib: Issue pre-emptive NAKs on eager buffer overflowMike Marciniszyn1-4/+137
Under congestion resulting in eager buffer overflow attempt to send pre-emptive NAKs if header queue entries with TID errors are generated and a valid header is present. This prevents long timeouts and flow restarts if a trailing set of packets are dropped due to eager overflows. Pre-emptive NAKs are currently only supported for RDMA writes. Signed-off-by: Mike Marciniszyn <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/qib: RDMA lkey/rkey validation is inefficient for large MRsMike Marciniszyn3-23/+60
The current code loops during rkey/lkey validiation to isolate the MR for the RDMA, which is expensive when the current operation is inside a very large memory region. This fix optimizes rkey/lkey validation routines for user memory regions and fast memory regions. The MR entry can be isolated by shifts/mods instead of looping. The existing loop is preserved for phys memory regions for now. Signed-off-by: Mike Marciniszyn <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/qib: Change QPN incrementMike Marciniszyn1-1/+1
Changing from +1 to +2 allows for better QP distribution across receive contexts. Signed-off-by: Mike Marciniszyn <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/qib: Add fix missing from earlier patchMike Marciniszyn1-31/+24
The upstream code was missing part of a receive/error race fix from the internal tree. Add the missing part, which makes future merges possible. Signed-off-by: Mike Marciniszyn <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/qib: Change receive queue/QPN selectionMike Marciniszyn4-31/+15
The basic idea is that on SusieQ, the difficult part of mapping QPN to context is handled by the mapping registers so the generic QPN allocation doesn't need to worry about chip specifics. For Monty and Linda, there is no mapping table so the qpt->mask (same as dd->qpn_mask), is used to see if the QPN to context falls within [zero..dd->n_krcv_queues). Signed-off-by: Mike Marciniszyn <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/qib: Fix interrupt mitigationMike Marciniszyn5-8/+12
For SusieQ we need to write to the interrupt timer register before updating the header queue head with interrupt count. This is to ensure that the timer is enabled properly and a receive available interrupt is delivered. Otherwise this interrupt can be lost if the receiver header/eager queues are full before the timer is enabled. Signed-off-by: Mike Marciniszyn <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/qib: Avoid duplicate writes to the rcv head registerMike Marciniszyn1-1/+9
Avoid duplicate writes to the head register as this can lead to lost interrupts if the context goes full before the second write is done. Signed-off-by: Mike Marciniszyn <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/qib: Add a few new SERDES tuningsMike Marciniszyn1-3/+26
Add new SERDES tuning to aid manufacturing. Signed-off-by: Mike Marciniszyn <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/qib: Reset packet list after freeingMike Marciniszyn1-0/+1
Reset the list pointers after freeing the SDMA packet list. This is done to any potential double-free cases. Signed-off-by: Mike Marciniszyn <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/qib: New SERDES init routine and improvements to SI qualityMike Marciniszyn1-15/+255
Implement new SERDES initialization routine and improvements to signal integrity -- disable LE1 adaptation, disable LOS after link-up, set better SERDES parameters. Signed-off-by: Mike Marciniszyn <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/qib: Clear WAIT_SEND flags when setting QP to error stateMike Marciniszyn1-0/+4
If these flags are set when the QP is transitioned to the error state, it will wait until the flags are cleared, which may never happen if the error transition is due to a link going down. Signed-off-by: Mike Marciniszyn <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/qib: Fix context allocation with multiple HCAsMike Marciniszyn1-5/+5
The driver was incorrectly choosing HCAs on which to allocate new user contexts based on overall count of usable ports regardless whether the usable port was on the currently selected HCA. Signed-off-by: Mike Marciniszyn <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/qib: Fix multi-Florida HCA host panic on rebootMike Marciniszyn1-2/+4
Add check when setting configured contexts that the value does not exceed the number of contexts allocated for the card. If the value exceeds the already allocated count, set it to what is already allocated. Signed-off-by: Mike Marciniszyn <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/qib: Handle transitions from ACTIVE_DEFERRED to ACTIVE betterMike Marciniszyn1-1/+2
When the link transitions from ACTIVE_DEFERRED to ACTIVE, the driver only sees the ACTIVE state. With this change, it will check whether the state was already ACTIVE and if so, it will not generated IB events and will not clear symbol error counts. Signed-off-by: Mike Marciniszyn <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/qib: UD send with immediate receive completion has wrong sizeMike Marciniszyn1-1/+1
The code to generate receive completion entries for UD send with immediate contains the wrong payload length. This is because when the code to compute the payload size was moved, the value of hdrsize didn't get moved too. The fix is to update tlen directly. Signed-off-by: Mike Marciniszyn <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/qib: Set port physical state even if other fields are invalidMike Marciniszyn1-20/+25
The IBTA vol. 1 release 1.2.1 spec. says: C14-24.2.1: If PortInfo:Portstate=Down, then a SubnSet(PortInfo) shall make any changes it specifies to PortInfo:PortPhysicalState; any other result is vendor-dependent. The patch changes the error handling so that the reply says there are invalid fields but still attempts to set fields that are in range including PortInfo:PortPhysicalState. Signed-off-by: Mike Marciniszyn <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/qib: Generate completion callback on errorsMike Marciniszyn1-1/+2
According to IBTA vol. 1, C11-30.1.1, a notification callback is invoked if the CQ is armed for the next solicited completion event or an error completion. The error case wasn't being generated correctly. Signed-off-by: Mike Marciniszyn <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/qib: Add support for the new QME7362 cardMike Marciniszyn1-0/+4
Add support to recognize another board variation named QME7362. Signed-off-by: Mike Marciniszyn <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/qib: Add receive header queue size module parametersMike Marciniszyn1-4/+24
The receive header queue sizes need to modified for performance tuning. Three module parameters are added to support this. Signed-off-by: Mike Marciniszyn <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/qib: Remove IB latency turnoffMike Marciniszyn1-2/+0
This is required for hardware testing. Signed-off-by: Mike Marciniszyn <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10RDMA/nes: Fix string continuation lineJoe Perches1-2/+2
Signed-off-by: Joe Perches <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/mthca: Handle -ENOMEM in forward_trap()Dan Carpenter1-0/+2
ib_create_send_mad() can return ERR_PTR(-ENOMEM) here. Signed-off-by: Dan Carpenter <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/mlx4: Handle -ENOMEM in forward_trap()Dan Carpenter1-0/+2
ib_create_send_mad() can return ERR_PTR(-ENOMEM) here. Signed-off-by: Dan Carpenter <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/mlx4: Don't call dma_free_coherent() with irqs disabledVladimir Sokolovsky1-1/+8
mlx4_ib_free_cq_buf() should not be called under spin_lock_irq() since it calls dma_free_coherent(), which needs irqs enabled. Fix this by deferring the free to outside the locked region. This was found due to the WARN_ON(irqs_disabled()); in swiotlb_free_coherent(). Signed-off-by: Vladimir Sokolovsky <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IPoIB: Add GRO supportOr Gerlitz3-1/+4
Signed-off-by: Or Gerlitz <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IPoIB: Remove LRO supportOr Gerlitz5-132/+2
As a first step in moving from LRO to GRO, revert commit af40da894e9 ("IPoIB: add LRO support"). Also eliminate the ethtool set_flags callback which isn't needed anymore. Finally, we need to include <linux/sched.h> directly to get the declaration of restart_syscall() (which used to be included implicitly through <linux/inet_lro.h>). Cc: Ben Hutchings <[email protected]> Cc: Eric W. Biederman <[email protected]> Cc: Vladimir Sokolovsky <[email protected]> Signed-off-by: Or Gerlitz <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
2011-01-10IB/ipath: Use printf extension %pR for struct resourceJoe Perches1-3/+2
Using %pR standardizes the struct resource output. Signed-off-by: Joe Perches <[email protected]> Signed-off-by: Roland Dreier <[email protected]>
Powered by cgit, Gruvboxed by Blaster4385.