Pull NTB updates from Jon Mason: "New feature to add support for NTB virtual MSI interrupts, the ability to test and use this feature in the NTB transport layer. Also, bug fixes for the AMD and Switchtec drivers, as well as some general patches" * tag 'ntb-5.3' of git://github.com/jonmason/ntb: (22 commits) NTB: Describe the ntb_msi_test client in the documentation. NTB: Add MSI interrupt support to ntb_transport NTB: Add ntb_msi_test support to ntb_test NTB: Introduce NTB MSI Test Client NTB: Introduce MSI library NTB: Rename ntb.c to support multiple source files in the module NTB: Introduce functions to calculate multi-port resource index NTB: Introduce helper functions to calculate logical port number PCI/switchtec: Add module parameter to request more interrupts PCI/MSI: Support allocating virtual MSI interrupts ntb_hw_switchtec: Fix setup MW with failure bug ntb_hw_switchtec: Skip unnecessary re-setup of shared memory window for crosslink case ntb_hw_switchtec: Remove redundant steps of switchtec_ntb_reinit_peer() function NTB: correct ntb_dev_ops and ntb_dev comment typos NTB: amd: Silence shift wrapping warning in amd_ntb_db_vector_mask() ntb_hw_switchtec: potential shift wrapping bug in switchtec_ntb_init_sndev() NTB: ntb_transport: Ensure qp->tx_mw_dma_addr is initaliazed NTB: ntb_hw_amd: set peer limit register NTB: ntb_perf: Clear stale values in doorbell and command SPAD register NTB: ntb_perf: Disable NTB link after clearing peer XLAT registers ...
Introduce the module parameter 'use_msi' which, when set, uses MSI interrupts instead of doorbells for each queue pair (QP). The parameter is only available if NTB MSI support is configured into the kernel. We also require there to be more than one memory window (MW) so that an extra one is available to forward the APIC region. To use MSIs, we request one interrupt per QP and forward the MSI address and data to the peer using scratch pad registers (SPADS) above the MW SPADS. (If there are not enough SPADS the MSI interrupt will not be used.) Once registered, we simply use ntb_msi_peer_trigger and the receiving ISR simply queues up the rxc_db_work for the queue. This addition can significantly improve performance of ntb_transport. In a simple, untuned, apples-to-apples comparision using ntb_netdev and iperf with switchtec hardware, I see 3.88Gb/s without MSI interrupts and 14.1Gb/s wit MSI, which is a more than 3x improvement. Signed-off-by: Logan Gunthorpe <logang@deltatee.com> Cc: Dave Jiang <dave.jiang@intel.com> Cc: Allen Hubbe <allenbh@gmail.com> Signed-off-by: Jon Mason <jdmason@kudzu.us>
Introduce a tool to test NTB MSI interrupts similar to the other NTB test tools. This tool creates a debugfs directory for each NTB device with the following files: port irqX_occurrences peerX/port peerX/count peerX/trigger The 'port' file tells the user the local port number and the 'occurrences' files tell the number of local interrupts that have been received for each interrupt. For each peer, the 'port' file and the 'count' file tell you the peer's port number and number of interrupts respectively. Writing the interrupt number to the 'trigger' file triggers the interrupt handler for the peer which should increment their corresponding 'occurrences' file. The 'ready' file indicates if a peer is ready, writing to this file blocks until it is ready. The module parameter num_irqs can be used to set the number of local interrupts. By default this is 4. This is only limited by the number of unused MSI interrupts registered by the hardware (this will require support of the hardware driver) and there must be at least 2*num_irqs + 1 spads registers available. Signed-off-by: Logan Gunthorpe <logang@deltatee.com> Cc: Dave Jiang <dave.jiang@intel.com> Cc: Allen Hubbe <allenbh@gmail.com> Signed-off-by: Jon Mason <jdmason@kudzu.us>
The NTB MSI library allows passing MSI interrupts across a memory window. This offers similar functionality to doorbells or messages except will often have much better latency and the client can potentially use significantly more remote interrupts than typical hardware provides for doorbells. (Which can be important in high-multiport setups.) The library utilizes one memory window per peer and uses the highest index memory windows. Before any ntb_msi function may be used, the user must call ntb_msi_init(). It may then setup and tear down the memory windows when the link state changes using ntb_msi_setup_mws() and ntb_msi_clear_mws(). The peer which receives the interrupt must call ntb_msim_request_irq() to assign the interrupt handler (this function is functionally similar to devm_request_irq()) and the returned descriptor must be transferred to the peer which can use it to trigger the interrupt. The triggering peer, once having received the descriptor, can trigger the interrupt by calling ntb_msi_peer_trigger(). Signed-off-by: Logan Gunthorpe <logang@deltatee.com> Cc: Dave Jiang <dave.jiang@intel.com> Cc: Allen Hubbe <allenbh@gmail.com> Signed-off-by: Jon Mason <jdmason@kudzu.us>
The kbuild system does not support having multiple source files in a module if one of those source files has the same name as the module. Therefore, we must rename ntb.c to core.c, while the module remains ntb.ko. This is similar to the way the nvme modules are structured. Signed-off-by: Logan Gunthorpe <logang@deltatee.com> Cc: Dave Jiang <dave.jiang@intel.com> Cc: Allen Hubbe <allenbh@gmail.com> Signed-off-by: Jon Mason <jdmason@kudzu.us>
Switchtec does not support setting multiple MWs simultaneously. The driver takes a hardware lock to ensure that two peers are not doing this simultaneously and it fails if someone else takes the lock. In most cases, this is fine as clients only setup the MWs once on one side of the link. However, there's a race condition when a re-initialization is caused by a link event. The driver will re-setup the shared memory window asynchronously and this races with the client setting up it's memory windows on the link up event. To fix this we ensure do the entire initialization in a work queue and signal the client once it's done. Signed-off-by: Joey Zhang <joey.zhang@microchip.com> Signed-off-by: Wesley Sheng <wesley.sheng@microchip.com> Signed-off-by: Jon Mason <jdmason@kudzu.us>
In case of NTB crosslink topology, the setting of shared memory window in the virtual partition doesn't reset on peer's reboot. So skip the unnecessary re-setup of shared memory window for that case. Signed-off-by: Wesley Sheng <wesley.sheng@microchip.com> Signed-off-by: Jon Mason <jdmason@kudzu.us>
When a re-initialization is caused by a link event, the driver will re-setup the shared memory window. But at that time, the shared memory is still valid, and it's unnecessary to free, reallocate and then initialize it again. We only need to reconfigure the hardware registers. Remove the redundant steps from switchtec_ntb_reinit_peer() function. Signed-off-by: Joey Zhang <joey.zhang@microchip.com> Signed-off-by: Wesley Sheng <wesley.sheng@microchip.com> Reviewed-by: Logan Gunthorpe <logang@deltatee.com> Signed-off-by: Jon Mason <jdmason@kudzu.us>
This code triggers a Smatch warning:
drivers/ntb/hw/amd/ntb_hw_amd.c:336 amd_ntb_db_vector_mask()
warn: should '(1 << db_vector)' be a 64 bit type?
I don't think "db_vector" can be higher than 16 so this doesn't affect
runtime, but it's nice to silence the static checker warning and we
might increase "ndev->db_count" in the future.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Acked-by: Shyam Sundar S K <Shyam-sundar.S-k@amd.com>
Signed-off-by: Jon Mason <jdmason@kudzu.us>
This code triggers a Smatch warning:
drivers/ntb/hw/mscc/ntb_hw_switchtec.c:884 switchtec_ntb_init_sndev()
warn: should '(1 << sndev->peer_partition)' be a 64 bit type?
The "part_map" and "tpart_vec" variables are u64 type so this seems like
a valid warning.
Fixes: 3df54c870f52 ("ntb_hw_switchtec: Allow using Switchtec NTB in multi-partition setups")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Jon Mason <jdmason@kudzu.us>