summaryrefslogtreecommitdiff
path: root/drivers/infiniband/hw
AgeCommit message (Collapse)Author
2009-06-23Merge branches 'ehca', 'misc', 'mlx4', 'mthca' and 'nes' into for-linusRoland Dreier
2009-06-23IB/ehca: Bump version numberAlexander Schmidt
Increment version number for DMEM toleration. Signed-off-by: Alexander Schmidt <alexs@linux.vnet.ibm.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-06-22IB/mthca: Replace dma_sync_single() use with proper functionsRoland Dreier
dma_sync_single() is deprecated now, and the use in mthca is wrong: there should be a dma_sync_single_for_cpu() before touching the memory from the CPU, and a dma_sync_single_for_device() afterwards. Fix this, prompted by a kick in the pants from a patch from FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>. Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-06-22RDMA/nes: Fix FIN state handling under error conditionsFaisal Latif
During cluster testing, one QP was not closed, as FIN is not handled properly when its rexmit count expires or in some cases when RST is is received after sending FIN. The reason is that the cm_id does not get decremented under these conditions. Signed-off-by: Faisal Latif <faisal.latif@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-06-22RDMA/nes: Fix max_qp_init_rd_atom returned from query deviceFaisal Latif
In nes_query_device(), max_qp_init_rd_atom is incorrectly set to max_qp_wr. This was found when a test application had a dapl async event error. Signed-off-by: Faisal Latif <faisal.latif@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-06-22IB/ehca: Ensure that guid_entry index is not negativeRoel Kluin
This prevents the memcpy() of a guid_entries element using a negative index. Signed-off-by: Roel Kluin <roel.kluin@gmail.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-06-22IB/ehca: Tolerate dynamic memory operations before driver loadHannes Hering
Implement toleration of dynamic memory operations and 16 GB gigantic pages, where "toleration" means that the driver can cope with dynamic memory operations that happen before the driver is loaded. While the ehca driver is loaded, dynamic memory operations are still prohibited by returning NOTIFY_BAD from the memory notifier. On module load the driver walks through available system memory, checks for available memory ranges and then registers the kernel internal memory region accordingly. The translation of address ranges is implemented via a 3-level busmap. Signed-off-by: Hannes Hering <hering2@de.ibm.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-06-15infiniband: ehca: remove driver_data direct access of struct deviceGreg Kroah-Hartman
In the near future, the driver core is going to not allow direct access to the driver_data pointer in struct device. Instead, the functions dev_get_drvdata() and dev_set_drvdata() should be used. These functions have been around since the beginning, so are backwards compatible with all older kernel versions. Cc: Sean Hefty <sean.hefty@intel.com> Cc: Roland Dreier <rolandd@cisco.com> Cc: Hal Rosenstock <hal.rosenstock@gmail.com> Cc: general@lists.openfabrics.org Cc: Christoph Raisch <raisch@de.ibm.com> Acked-by: Hoang-Nam Nguyen <hnguyen@de.ibm.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2009-06-14Merge branches 'cxgb3', 'ehca', 'misc', 'mlx4', 'mthca' and 'nes' into for-linusRoland Dreier
2009-06-13IB/mthca: Don't double-free IRQs when falling back from MSI-X to INTxRoland Dreier
When both MSI-X and legacy INTx fail to generate an interrupt, the driver frees the MSI-X interrupts twice. Fix this by clearing the have_irq flag for the MSI-X interrupts when they are freed the first time. Reported-by: Yinghai Lu <yhlu.kernel@gmail.com> Tested-by: Yinghai Lu <yhlu.kernel@gmail.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-06-05IB/mlx4: Add strong ordering to local inval and fast reg work requestsJack Morgenstein
The ConnectX Programmer's Reference Manual states that the "SO" bit must be set when posting Fast Register and Local Invalidate send work requests. When this bit is set, the work request will be executed only after all previous work requests on the send queue have been executed. (If the bit is not set, Fast Register and Local Invalidate WQEs may begin execution too early, which violates the defined semantics for these operations) This fixes the issue with NFS/RDMA reported in <http://lists.openfabrics.org/pipermail/general/2009-April/059253.html> Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il> Cc: <stable@kernel.org> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-06-03IB/ehca: Remove superfluous bitmasks from QP control blockJoachim Fenkes
All the fields in the control block are nicely right-aligned, so no masking is necessary. Signed-off-by: Joachim Fenkes <fenkes@de.ibm.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-05-27RDMA/cxgb3: Limit fast register size based on T3 limitationsSteve Wise
T3 firmware only supports one WRs worth of page list for fast register work requests. The driver currently allows 2 WRs worth, which doesn't work for T3, so reduce the limit in the driver. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-05-27RDMA/cxgb3: Report correct port state and MTUSteve Wise
Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-05-27IB/mthca: Add module parameter for number of MTTs per segmentEli Cohen
The current MTT allocator uses kmalloc() to allocate a buffer for its buddy allocator, and thus is limited in the amount of MTT segments that it can control. As a result, the size of memory that can be registered is limited too. This patch uses a module parameter to control the number of MTT entries that each segment represents, allowing more memory to be registered with the same number of segments. Signed-off-by: Eli Cohen <eli@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-05-15RDMA/nes: Fix off-by-one bugs in reset_adapter_ne020() and init_serdes()Roel Kluin
With a postfix increment, i is incremented one past 10K/5K before the loop ends, so the error messages will be displayed too soon if the test succeeds on the last iteration. Fix the comparisons to be > instead of >=. Signed-off-by: Roel Kluin <roel.kluin@gmail.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-05-13infiniband: Remove void castsJack Stone
Remove uneeded casts of void *. Signed-off-by: Jack Stone <jwjstone@fastmail.fm> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-05-13IB/ehca: Increment version numberStefan Roscher
Signed-off-by: Stefan Roscher <stefan.roscher@de.ibm.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-05-13IB/ehca: Remove unnecessary memory operations for userspace queue pairsStefan Roscher
The queue map for flush completion circumvention is only used for kernel space queue pairs. This patch skips the allocation of the queue maps in case the QP is created for userspace. In addition, this patch does not iomap the galpas for kernel usage if the queue pair is only used in userspace. These changes will improve the performance of creation of userspace queue pairs. Signed-off-by: Stefan Roscher <stefan.roscher@de.ibm.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-05-13IB/ehca: Fall back to vmalloc() for big allocationsStefan Roscher
In case of large queue pairs there is the possibillity of allocation failures due to memory fragmentation when using kmalloc(). To ensure the memory is allocated even if kmalloc() can not find chunks which are big enough, we fall back to allocating the memory with vmalloc(). Signed-off-by: Stefan Roscher <stefan.roscher@de.ibm.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-05-13IB/ehca: Replace vmalloc() with kmalloc() for queue allocationAnton Blanchard
To improve performance of driver resource allocation, replace vmalloc() calls with kmalloc(). Signed-off-by: Stefan Roscher <stefan.roscher@de.ibm.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-05-13Merge branch 'for-linus' of ↵Linus Torvalds
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: IB/mlx4: Don't overwrite fast registration page list when posting work request RDMA/cxgb3: Don't complete flushed send work requests twice
2009-05-13Merge branches 'cxgb3' and 'mlx4' into for-linusRoland Dreier
2009-05-09Fix deadlock in ipathfs ->get_sb()Al Viro
forgot to unlock superblock before calling deactivate_super()... Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2009-05-07IB/mlx4: Don't overwrite fast registration page list when posting work requestJack Morgenstein
The low-level mlx4 driver modified the page-list addresses for fast register work requests post send to big-endian, and set a "present" bit. This caused problems later when the consumer attempted to unmap the pages using the page-list (using the list addresses which were assumed to be still in CPU-endian order). Fix the mlx4 driver to allocate two buffers and use a private buffer for the hardware-format bus addresses. This patch fixes <https://bugs.openfabrics.org/show_bug.cgi?id=1571>, an NFS/RDMA server crash. The cause of the crash was found by Vu Pham of Mellanox. The fix is along the lines suggested by Steve Wise in comment #21 in bug 1571. Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-04-29RDMA/cxgb3: Don't complete flushed send work requests twiceSteve Wise
When the SQ is flushed, mark the flushed entries as not signaled so the poll logic doesn't re-insert the CQ entry thinking its an out of order completion. The bug can cause the NFS/RDMA server to crash due to processing the same completed work request twice. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-04-28Merge branches 'cxgb3', 'ipoib', 'mthca', 'mlx4' and 'nes' into for-linusRoland Dreier
2009-04-27RDMA/nes: Update iw_nes versionChien Tung
Update version number to 1.5.0.0 Signed-off-by: Chien Tung <chien.tin.tung@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-04-27RDMA/nes: Fix error path in nes_accept()Faisal Latif
If reg_phys_mem() fails, we need to free memory allocated for MPA frame with private data before returning the error. Also move nes_add_ref() after the reg_phys_mem() is successful. Signed-off-by: Faisal Latif <faisal.latif@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-04-27RDMA/nes: Fix hang issues for large cluster dynamic connectionsFaisal Latif
Running large cluster setup, we are hanging after many hours of testing. Fixing this required going over the code and making sure the rexmit entry was properly removed based on the cm_node's state and packet received. Also when receiving a FIN packet, check seq# and make sure there were no errors before calling handle_fin(). Following are the changes done in nes_cm.c: * handle_ack_pkt() needs to return error value, so in case of error, handle_fin() is not called. Some cleanup done while going over the code. * handle_rst_pkt(), handling of cm_node's NES_CM_STATE_LAST_ACK is missing. * process_packet(), in case of FIN only packet is received, call check_seq() before processing. * in handle_fin_pkt(), we are calling cleanup_retrans_entry() for all conditions, even if the packets need to be dropped. Signed-off-by: Faisal Latif <faisal.latif@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-04-27RDMA/nes: Increase rexmit timeout intervalFaisal Latif
Under heavy load with large cluster testing, it may take longer to receive a response to MPA requests. Change the driver to wait longer after each rexmit to max time value. Signed-off-by: Faisal Latif <faisal.latif@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-04-27RDMA/nes: Check for sequence number wrap-aroundFaisal Latif
check_seq() was not checking if the seq#s have wrapped. Fix it. Signed-off-by: Faisal Latif <faisal.latif@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-04-27RDMA/nes: Do not set apbvt entry for loopbackFaisal Latif
When a connect request comes, apbvt should only be set for non-loopback connections. Signed-off-by: Faisal Latif <faisal.latif@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-04-27RDMA/nes: Fix unused variable compile warning when INFINIBAND_NES_DEBUG=nChien Tung
Remove the NES_DEBUG that is causing the compile warning about an unused variable when INFINIBAND_NES_DEBUG is not enabled. Signed-off-by: Chien Tung <chien.tin.tung@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-04-27RDMA/nes: Fix fw_ver in /sysChien Tung
/sys/class/infiniband/nes?/fw_ver is not displaying firmware version properly (it shows 0.0.0 with the current code). Fill in the correct firmware version number. Signed-off-by: Chien Tung <chien.tin.tung@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-04-27RDMA/nes: Set trace length to 1 inch for SFP_DChien Tung
With updated PHY firmware for SFP_D, setting the trace length to 1 inch for SFP_D provides a more stable link. Signed-off-by: Chien Tung <chien.tin.tung@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-04-27RDMA/nes: Enable repause timer for port 1Chien Tung
Enable repause timer for port 1. Without this setting, under stress, the chip may misbehave. Signed-off-by: Chien Tung <chien.tin.tung@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-04-27RDMA/nes: Correct CDR loop filter setting for port 1Chien Tung
In commit 1b949324 ("RDMA/nes: Fix SFP+ PHY initialization") there is a mistake in the clean up code that removed port 1 CDR loop filter settings for 10G cards other than CX4. Put the correct setting back for appropriate PHY types. Signed-off-by: Chien Tung <chien.tin.tung@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-04-27RDMA/nes: Modify thermo mitigation to flip SerDes1 ref clk to internalChien Tung
Change thermo mitigation code to flip the SerDes1 reference clock to internal, to match the change in commit a4849fc1 ("RDMA/nes: Add wide_ppm_offset parm for switch compatibility"). Signed-off-by: Chien Tung <chien.tin.tung@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-04-21RDMA/nes: Fix resource issues in nes_create_cq() and nes_destroy_cq()Miroslaw Walukiewicz
In error paths where a CQ is not created, pbl is not freeed properly. In nes_destroy_cq(), add the corresponding check for nescq->mcrqf to not call nes_free_resource() when it is already done in nes_create_cq(). Signed-off-by: Miroslaw Walukiewicz <miroslaw.walukiewicz@intel.com> Signed-off-by: Chien Tung <chien.tin.tung@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-04-21RDMA/nes: Remove root_256()'s unused pbl_count_256 parameterMatt Kraai
Signed-off-by: Matt Kraai <kraai@ftbfs.org> Acked-by: Chien Tung <chien.tin.tung@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-04-20IB/mthca: Fix timeout for INIT_HCA and a few other commandsJack Morgenstein
Commands INIT_HCA, CLOSE_HCA, SYS_EN, SYS_DIS, and CLOSE_IB all have 1 second timeouts. For INIT_HCA this causes problems when had more than 2^18 are QPs configured, since the command takes more than 1 second to complete. All other commands have 60-second timeouts. This patch makes the above commands consistent with the rest of the commands (and with the chip documentation). This patch is an expansion of a patch from Arthur Kepner <akepner@sgi.com> fixing just the INIT_HCA timeout. Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-04-20RDMA/cxgb3: Don't zero QP attrs when moving to IDLESteve Wise
QP attributes must stay initialized when moving back to IDLE. Zeroing them will crash the system in _flush_qp() if the QP is subsequently moved to ERROR and back to IDLE. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-04-20RDMA/nes: Fix bugs in nes_reg_phys_mr()Don Wood
The code incorrectly failed memory registration if the buffer was not page aligned. Also, the length field is mangled causing the hardware to think the registration is much larger than it really is. The fix is to remove the page alignment restriction as well the incorrect length adjustment. Also make sure that all buffers after the first start at a page boundary, and all buffers except the last end on a page boundary. Signed-off-by: Don Wood <donald.e.wood@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-04-20RDMA/nes: Fix compiler warning at nes_verbs.c:1955Chien Tung
Initialize pbl_count_256 to 0 to get rid of the warning: drivers/infiniband/hw/nes/nes_verbs.c: In function 'nes_reg_mr': drivers/infiniband/hw/nes/nes_verbs.c:1955: warning: 'pbl_count_256' may be used uninitialized in this function Reported-by: Roland Dreier <rdreier@cisco.com> Signed-off-by: Chien Tung <chien.tin.tung@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-04-20RDMA/cxgb3: Adjust ORD/IRD (if needed) for peer2peer connectionsSteve Wise
NFS/RDMA currently fails to set up connections if peer2peer is on. This is due to the fact that the NFS/RDMA client sets its ORD to 0. If peer2peer is set, make sure the active side ORD is >= 1 and the passive side IRD is >=1. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-04-09Merge branch 'for-linus' of ↵Linus Torvalds
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/nes: Add support for new SFP+ PHY RDMA/nes: Add wide_ppm_offset parm for switch compatibility RDMA/nes: Fix SFP+ PHY initialization RDMA/nes: Fix nes_nic_cm_xmit() error handling RDMA/nes: Fix error handling issues RDMA/nes: Fix incorrect casts on 32-bit architectures IPoIB: Document newish features RDMA/cma: Create cm id even when IB port is down RDMA/cma: Use rate from IPoIB broadcast when joining IPoIB multicast groups IPoIB: Avoid free_netdev() BUG when destroying a child interface mlx4_core: Don't leak mailbox for SET_PORT on Ethernet ports RDMA/cxgb3: Release dependent resources only when endpoint memory is freed. RDMA/cxgb3: Handle EEH events IB/mlx4: Use pgprot_writecombine() for BlueFlame pages
2009-04-08Merge branches 'cma', 'cxgb3', 'ipoib', 'mlx4' and 'nes' into for-nextRoland Dreier
2009-04-08RDMA/nes: Add support for new SFP+ PHYChien Tung
Add new register settings for new SFP+ PHY/firmware. Add new PHY to to nes_netdev_get/set_settings. Signed-off-by: Chien Tung <chien.tin.tung@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2009-04-08RDMA/nes: Add wide_ppm_offset parm for switch compatibilityChien Tung
We have observed unstable link with a new BNT switch. Add wide_ppm_offset parameter to allow the user to control the clock ppm offset on the CX4 interface for better compatibility. Default is 100ppm, setting it to 1 will increase it to 300ppm. Change default SerDes1 reference clock to external source. Signed-off-by: Chien Tung <chien.tin.tung@intel.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>