kernel/linux.git/drivers/infiniband, branch linux-2.6.35.y

IB/cm: Bump reference count on cm_id before invoking callback

2011-03-31T18:58:38+00:00

commit 29963437a48475036353b95ab142bf199adb909e upstream. When processing a SIDR REQ, the ib_cm allocates a new cm_id. The refcount of the cm_id is initialized to 1. However, cm_process_work will decrement the refcount after invoking all callbacks. The result is that the cm_id will end up with refcount set to 0 by the end of the sidr req handler. If a user tries to destroy the cm_id, the destruction will proceed, under the incorrect assumption that no other threads are referencing the cm_id. This can lead to a crash when the cm callback thread tries to access the cm_id. This problem was noticed as part of a larger investigation with kernel crashes in the rdma_cm when running on a real time OS. Signed-off-by: Sean Hefty Acked-by: Doug Ledford Signed-off-by: Roland Dreier Signed-off-by: Greg Kroah-Hartman Signed-off-by: Andi Kleen

RDMA/cma: Fix crash in request handlers

2011-03-31T18:58:38+00:00

commit 25ae21a10112875763c18b385624df713a288a05 upstream. Doug Ledford and Red Hat reported a crash when running the rdma_cm on a real-time OS. The crash has the following call trace: cm_process_work cma_req_handler cma_disable_callback rdma_create_id kzalloc init_completion cma_get_net_info cma_save_net_info cma_any_addr cma_zero_addr rdma_translate_ip rdma_copy_addr cma_acquire_dev rdma_addr_get_sgid ib_find_cached_gid cma_attach_to_dev ucma_event_handler kzalloc ib_copy_ah_attr_to_user cma_comp [ preempted ] cma_write copy_from_user ucma_destroy_id copy_from_user _ucma_find_context ucma_put_ctx ucma_free_ctx rdma_destroy_id cma_exch cma_cancel_operation rdma_node_get_transport rt_mutex_slowunlock bad_area_nosemaphore oops_enter They were able to reproduce the crash multiple times with the following details: Crash seems to always happen on the: mutex_unlock(&conn_id->handler_mutex); as conn_id looks to have been freed during this code path. An examination of the code shows that a race exists in the request handlers. When a new connection request is received, the rdma_cm allocates a new connection identifier. This identifier has a single reference count on it. If a user calls rdma_destroy_id() from another thread after receiving a callback, rdma_destroy_id will proceed to destroy the id and free the associated memory. However, the request handlers may still be in the process of running. When control returns to the request handlers, they can attempt to access the newly created identifiers. Fix this by holding a reference on the newly created rdma_cm_id until the request handler is through accessing it. Signed-off-by: Sean Hefty Acked-by: Doug Ledford Signed-off-by: Roland Dreier Signed-off-by: Greg Kroah-Hartman Signed-off-by: Andi Kleen

IB/uverbs: Handle large number of entries in poll CQ

2011-02-06T19:03:29+00:00

commit 7182afea8d1afd432a17c18162cc3fd441d0da93 upstream. In ib_uverbs_poll_cq() code there is a potential integer overflow if userspace passes in a large cmd.ne. The calls to kmalloc() would allocate smaller buffers than intended, leading to memory corruption. There iss also an information leak if resp wasn't all used. Unprivileged userspace may call this function, although only if an RDMA device that uses this function is present. Fix this by copying CQ entries one at a time, which avoids the allocation entirely, and also by moving this copying into a function that makes sure to initialize all memory copied to userspace. Special thanks to Jason Gunthorpe for his help and advice. Signed-off-by: Dan Carpenter Signed-off-by: Andi Kleen [ Monkey around with things a bit to avoid bad code generation by gcc when designated initializers are used. - Roland ] Signed-off-by: Roland Dreier Signed-off-by: Greg Kroah-Hartman

RDMA/cxgb3: Turn off RX coalescing for iWARP connections

2010-10-29T04:51:11+00:00

commit bec658ff31453a5726b1c188674d587a5d40c482 upstream. The HW by default has RX coalescing on. For iWARP connections, this causes a 100ms delay in connection establishement due to the ingress MPA Start message being stalled in HW. So explicitly turn RX coalescing off when setting up iWARP connections. This was causing very bad performance for NP64 gather operations using Open MPI, due to the way it sets up connections on larger jobs. Signed-off-by: Steve Wise Signed-off-by: Roland Dreier Signed-off-by: Greg Kroah-Hartman

RDMA/cxgb3: Don't exceed the max HW CQ depth

2010-09-20T20:36:43+00:00

commit dc4e96ce2dceb649224ee84f83592aac8c54c9b7 upstream. The max depth supported by T3 is 64K entries. This fixes a bug introduced in commit 9918b28d ("RDMA/cxgb3: Increase the max CQ depth") that causes stalls and possibly crashes in large MPI clusters. Signed-off-by: Steve Wise Signed-off-by: Roland Dreier Signed-off-by: Greg Kroah-Hartman

IB/qib: Use request_firmware() to load SD7220 firmware

2010-07-08T20:27:05+00:00

Extract the microcode for the QLogic QLE7220 series IB HCA and use the kernel microcode request facility to load the microcode. This supports Debian Linux's requirements to separate microcode which doesn't have open source code available from the device driver. Signed-off-by: Ben Hutchings Signed-off-by: Roland Dreier

Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband

2010-07-08T19:20:54+00:00

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband: IPoIB: Fix world-writable child interface control sysfs attributes IB/qib: Clean up properly if qib_init() fails IB/qib: Completion queue callback needs to be single threaded IB/qib: Update 7322 serdes tables IB/qib: Clear 6120 hardware error register IB/qib: Clear eager buffer memory for each new process IB/qib: Mask hardware error during link reset IB/qib: Don't mark VL15 bufs as WC to avoid a rare 7322 chip problem RDMA/cxgb4: Derive smac_idx from port viid RDMA/cxgb4: Avoid false GTS CIDX_INC overflows RDMA/cxgb4: Don't call abort_connection() for active connect failures RDMA/cxgb4: Use the DMA state API instead of the pci equivalents

Merge branches 'cxgb4', 'ipoib' and 'qib' into for-next

2010-07-08T16:10:24+00:00

IPoIB: Fix world-writable child interface control sysfs attributes

2010-07-06T21:23:22+00:00

Sumeet Lahorani reported that the IPoIB child entries are world-writable; however we don't want ordinary users to be able to create and destroy child interfaces, so fix them to be writable only by root. Signed-off-by: Or Gerlitz Cc: Signed-off-by: Roland Dreier

IB/qib: Clean up properly if qib_init() fails

2010-07-06T21:14:04+00:00

If qib_init() fails, the driver fails to free memory, unregister device files, and unregister with the PCIe framework. The driver will unload without error but a subsequent driver load will cause the system to panic. This was found by changing the 7220 code to load the serdes microcode separately and not installing the microcode file. Signed-off-by: Ralph Campbell Signed-off-by: Roland Dreier