- 05 2月, 2008 1 次提交
-
-
由 Hoang-Nam Nguyen 提交于
This patch enables ehca to redirect any PMA queries to the actual PMA QP. Signed-off-by: NHoang-Nam Nguyen <hnguyen@de.ibm.com> Reviewed-by: NJoachim Fenkes <fenkes@de.ibm.com> Reviewed-by: NChristoph Raisch <raisch@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
- 26 1月, 2008 3 次提交
-
-
由 Joachim Fenkes 提交于
Some HW revisions of eHCA2 may cause an RC connection to break if they received RDMA Reads over that connection before. This can be prevented by assuring that, after the first RDMA Read, the QP receives a new RDMA Read every few million link packets. Include code into the driver that inserts an empty (size 0) RDMA Read into the message stream every now and then if the consumer doesn't post them frequently enough. Signed-off-by: NJoachim Fenkes <fenkes@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Hoang-Nam Nguyen 提交于
This patch enhances ehca with a capability to "autodetect" the ports being connected physically. In order to utilize that function the module option nr_ports must be set to -1 (default is 2 - two ports). This feature is experimental and will made the default later. More detail: If the user connects only one port to the switch, current code requires 1) port one to be connected and 2) module option nr_ports=1 to be given. If autodetect is enabled, ehca will not wait at creation of the GSI QP for the respective port to become active. Since firmware does not accept modify_qp() while the port is down at initialization, we need to cache all calls to modify_qp() for the SMI/GSI QP and just return a good return code. When a port is activated and we get a PORT_ACTIVE event, we replay the cached modify-qp() parms and re-trigger any posted recv WRs. Only then do we forward the PORT_ACTIVE event to registered clients. The result of this autodetect patch is that all ports will be accessible by the users. Depending on their respective cabling only those ports that are connected properly will become operable. If a user tries to modify a regular QP of a non-connected port, modify_qp() will fail. Furthermore, ibv_devinfo should show the port state accordingly. Note that this patch primarily improves the loading behaviour of ehca. If the cable is removed while the driver is operating and plugged in again, firmware will handle that properly by sending an appropriate async event. Signed-off-by: NHoang-Nam Nguyen <hnguyen@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Hoang-Nam Nguyen 提交于
Signed-off-by: NHoang-Nam Nguyen <hnguyen@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
- 14 12月, 2007 1 次提交
-
-
由 Joachim Fenkes 提交于
Signed-off-by: NJoachim Fenkes <fenkes@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
- 14 11月, 2007 1 次提交
-
-
由 Joachim Fenkes 提交于
The IPD (inter-packet delay) formula was a little off and assumed a fixed physical link rate; fix the formula and query the actual physical link rate, now that we can get it. Also, refactor the calculation into a common function ehca_calc_ipd() and use that instead of duplicating code. Signed-off-by: NJoachim Fenkes <fenkes@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
- 18 10月, 2007 1 次提交
-
-
由 Joachim Fenkes 提交于
ehca_shca.hca_cap_mr_pgsize now contains all supported page sizes ORed together. This makes some checks easier to code and understand, plus we can return this value verbatim in query_hca(), fixing a problem with SRP (reported by Anton Blanchard -- thanks!). Signed-off-by: NJoachim Fenkes <fenkes@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
- 17 10月, 2007 1 次提交
-
-
由 Joachim Fenkes 提交于
Replace struct ibmebus_dev and struct ibmebus_driver with struct of_device and struct of_platform_driver, respectively. Match the external ibmebus interface and drivers using it. Signed-off-by: NJoachim Fenkes <fenkes@de.ibm.com> Acked-by: NArnd Bergmann <arnd@arndb.de> Acked-by: NRoland Dreier <rolandd@cisco.com> Acked-by: NJeff Garzik <jeff@garzik.org> Signed-off-by: NPaul Mackerras <paulus@samba.org>
-
- 10 10月, 2007 4 次提交
-
-
由 Hoang-Nam Nguyen 提交于
Signed-off-by: NHoang-Nam Nguyen <hnguyen@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Hoang-Nam Nguyen 提交于
Signed-off-by: NJoachim Fenkes <fenkes@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Hoang-Nam Nguyen 提交于
Use Paul's new remap_4k_pfn() function to map our 4K firmware contexts into user space on 64K-page machines without exposing neighboring firmware contexts. Return the context's offset within a 64K page to user space so it can determine the proper virtual address. For details about remap_4k_pfn(), see commit 721151d0 or http://patchwork.ozlabs.org/linuxppc/patch?id=10281Signed-off-by: NJoachim Fenkes <fenkes@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Roland Dreier 提交于
ehca_classes.h uses struct mutex, so while <linux/mutex.h> seems to be pulled in indirectly by one of the headers it includes, the right thing is to include <linux/mutex.h> directly. Signed-off-by: NMichael S. Tsirkin <mst@dev.mellanox.co.il> Acked-by: NStefan Roscher <stefan.roscher@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
- 29 7月, 2007 1 次提交
-
-
由 Hoang-Nam Nguyen 提交于
Make sure declarations stay in sync with definitions by keeping all extern declarations in common .h files. Signed-off-by: NHoang-Nam Nguyen <hnguyen@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
- 21 7月, 2007 2 次提交
-
-
由 Stefan Roscher 提交于
eHCA2 supports QP queues that can be as small as 512 bytes. This greatly reduces memory overhead for consumers that use lots of QPs with small queues (e.g. RDMA-only QPs). Apart from dealing with firmware, this code needs to manage bite-sized chunks of kernel pages, making sure that no kernel page is shared between different protection domains. Signed-off-by: NHoang-Nam Nguyen <hnguyen@de.ibm.com>
-
由 Hoang-Nam Nguyen 提交于
Add support for MR pages larger than 4K on eHCA2. This reduces firmware memory consumption. If enabled via the mr_largepage module parameter, the MR page size will be determined based on the MR length and the hardware capabilities -- if the MR is >= 16M, 16M pages are used, for example. Signed-off-by: NJoachim Fenkes <fenkes@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
- 18 7月, 2007 2 次提交
-
-
由 Hoang-Nam Nguyen 提交于
Run the existing ehca code through checkpatch.pl and clean up the worst of the coding style violations. Signed-off-by: NJoachim Fenkes <fenkes@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Hoang-Nam Nguyen 提交于
- Rename struct ehca_mr fields to clearly distinguish between kernel and HW page size. - Sort struct ehca_mr_pginfo into a common part and a union containing specific fields for physical, user and fast MR Signed-off-by: NJoachim Fenkes <fenkes@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
- 10 7月, 2007 7 次提交
-
-
由 Joachim Fenkes 提交于
When firmware reports a nondisruptive port configuration change event, previous versions of the eHCA driver didn't forward the event to consumers like IPoIB. Add code that determines the type of configuration change by comparing old and new port attributes and reports it. Signed-off-by: NJoachim Fenkes <fenkes@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Joachim Fenkes 提交于
This eliminates lock contention among IRQs as well as the need to disable IRQs around idr_find, because there are no IRQ writers. Signed-off-by: NJoachim Fenkes <fenkes@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Joachim Fenkes 提交于
- ehca_cq.nr_events is made an atomic_t, eliminating a lot of locking. - The CQ is removed from the CQ idr first now to make sure no more completions are scheduled on that CQ. The "wait for all completions to end" code becomes much simpler this way. Signed-off-by: NJoachim Fenkes <fenkes@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Joachim Fenkes 提交于
- Rename all spinlock flags to "flags", matching the vast majority of kernel code. - Move hcall_lock into the only file it's used in. - Replaced spin_lock_init() and friends with static initializers for global variables. Signed-off-by: NJoachim Fenkes <fenkes@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Joachim Fenkes 提交于
Support SRQs on eHCA2. Since an SRQ is a QP for eHCA2, a lot of code (structures, create, destroy, post_recv) can be shared between QP and SRQ. Signed-off-by: NJoachim Fenkes <fenkes@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Joachim Fenkes 提交于
- Replace init_qp_queues() by a shorter init_qp_queue(), eliminating duplicate code. - hipz_h_alloc_resource_qp() doesn't need a pointer to struct ehca_qp any longer. All input and output data is transferred through the parms parameter. - Change the interface to also support SRQ. Signed-off-by: NJoachim Fenkes <fenkes@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Joachim Fenkes 提交于
In preparation for support of new eHCA2 features, change adapter probing: - Hardware level is changed to encode major and minor chip version - Hardware capabilities are queried from the firmware - The maximum MTU is queried from the firmware instead of assuming a fixed value Signed-off-by: NJoachim Fenkes <fenkes@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
- 15 5月, 2007 1 次提交
-
-
由 Stefan Roscher 提交于
Some pSeries hypervisor versions show a race condition in the allocate MR hCall. Serialize this call per adapter to circumvent this problem. Signed-off-by: NJoachim Fenkes <fenkes@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
- 09 5月, 2007 1 次提交
-
-
由 Roland Dreier 提交于
Export ib_umem_get()/ib_umem_release() and put low-level drivers in control of when to call ib_umem_get() to pin and DMA map userspace, rather than always calling it in ib_uverbs_reg_mr() before calling the low-level driver's reg_user_mr method. Also move these functions to be in the ib_core module instead of ib_uverbs, so that driver modules using them do not depend on ib_uverbs. This has a number of advantages: - It is better design from the standpoint of making generic code a library that can be used or overridden by device-specific code as the details of specific devices dictate. - Drivers that do not need to pin userspace memory regions do not need to take the performance hit of calling ib_mem_get(). For example, although I have not tried to implement it in this patch, the ipath driver should be able to avoid pinning memory and just use copy_{to,from}_user() to access userspace memory regions. - Buffers that need special mapping treatment can be identified by the low-level driver. For example, it may be possible to solve some Altix-specific memory ordering issues with mthca CQs in userspace by mapping CQ buffers with extra flags. - Drivers that need to pin and DMA map userspace memory for things other than memory regions can use ib_umem_get() directly, instead of hacks using extra parameters to their reg_phys_mr method. For example, the mlx4 driver that is pending being merged needs to pin and DMA map QP and CQ buffers, but it does not need to create a memory key for these buffers. So the cleanest solution is for mlx4 to call ib_umem_get() in the create_qp and create_cq methods. Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
- 25 4月, 2007 1 次提交
-
-
由 Joachim Fenkes 提交于
Add "Modify Port" verb support to eHCA driver. The IB communication manager needs this to set the IsCM port capability bit when initializing. Signed-off-by: NJoachim Fenkes <fenkes@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
- 02 3月, 2007 1 次提交
-
-
由 Hoang-Nam Nguyen 提交于
This patch fixes two issues reported by Roland Dreier and Christoph Hellwig: - Mismatched sync/locking between completion handler and destroy cq We introduced a counter nr_events per cq to track number of irq events seen. This counter is incremented when an event queue entry is seen and decremented after completion handler has been called regardless if scaling code is active or not. Note that nr_callbacks tracks number of events assigned to a cpu and both counters can potentially diverge. The sync between running completion handler and destroy cq is done by using the global spin lock ehca_cq_idr_lock. - Replace yield by wait_event on the counter above to become zero. Signed-off-by: NHoang-Nam Nguyen <hnguyen@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
- 17 2月, 2007 2 次提交
-
-
由 Hoang-Nam Nguyen 提交于
Allow users to en/disable scaling code when loading ib_ehca module, rather than requiring the module to be rebuilt to change the setting. Signed-off-by: NHoang-Nam Nguyen <hnguyen@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Hoang-Nam Nguyen 提交于
Rework ehca interrupt handling to avoid/reduce missed irq events. Signed-off-by: NHoang-Nam Nguyen <hnguyen@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
- 05 2月, 2007 2 次提交
-
-
由 Hoang-Nam Nguyen 提交于
Remove prototypes for functions that don't exist. Signed-off-by: NHoang-Nam Nguyen <hnguyen@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
由 Hoang-Nam Nguyen 提交于
This patch removes do_mmap() from ehca: - Call remap_pfn_range() for hardware register block - Use vm_insert_page() to register memory allocated for completion queues and queue pairs - The actual mmap() call/trigger is now controlled by user space, ie. libehca Signed-off-by: NHoang-Nam Nguyen <hnguyen@de.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-
- 23 9月, 2006 1 次提交
-
-
由 Heiko J Schick 提交于
Add a driver for IBM GX bus InfiniBand adapters, which are usable with some pSeries/System p systems. Signed-off-by: Heiko J Schick <schickhj.ibm.com> Signed-off-by: NRoland Dreier <rolandd@cisco.com>
-