diff options
author | 2023-02-21 18:24:12 -0800 | |
---|---|---|
committer | 2023-02-21 18:24:12 -0800 | |
commit | 5b7c4cabbb65f5c469464da6c5f614cbd7f730f2 (patch) | |
tree | cc5c2d0a898769fd59549594fedb3ee6f84e59a0 /drivers/net/ethernet/qlogic/qed/qed_spq.c | |
download | linux-5b7c4cabbb65f5c469464da6c5f614cbd7f730f2.tar.gz linux-5b7c4cabbb65f5c469464da6c5f614cbd7f730f2.zip |
Merge tag 'net-next-6.3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-nextgrafted
Pull networking updates from Jakub Kicinski:
"Core:
- Add dedicated kmem_cache for typical/small skb->head, avoid having
to access struct page at kfree time, and improve memory use.
- Introduce sysctl to set default RPS configuration for new netdevs.
- Define Netlink protocol specification format which can be used to
describe messages used by each family and auto-generate parsers.
Add tools for generating kernel data structures and uAPI headers.
- Expose all net/core sysctls inside netns.
- Remove 4s sleep in netpoll if carrier is instantly detected on
boot.
- Add configurable limit of MDB entries per port, and port-vlan.
- Continue populating drop reasons throughout the stack.
- Retire a handful of legacy Qdiscs and classifiers.
Protocols:
- Support IPv4 big TCP (TSO frames larger than 64kB).
- Add IP_LOCAL_PORT_RANGE socket option, to control local port range
on socket by socket basis.
- Track and report in procfs number of MPTCP sockets used.
- Support mixing IPv4 and IPv6 flows in the in-kernel MPTCP path
manager.
- IPv6: don't check net.ipv6.route.max_size and rely on garbage
collection to free memory (similarly to IPv4).
- Support Penultimate Segment Pop (PSP) flavor in SRv6 (RFC8986).
- ICMP: add per-rate limit counters.
- Add support for user scanning requests in ieee802154.
- Remove static WEP support.
- Support minimal Wi-Fi 7 Extremely High Throughput (EHT) rate
reporting.
- WiFi 7 EHT channel puncturing support (client & AP).
BPF:
- Add a rbtree data structure following the "next-gen data structure"
precedent set by recently added linked list, that is, by using
kfunc + kptr instead of adding a new BPF map type.
- Expose XDP hints via kfuncs with initial support for RX hash and
timestamp metadata.
- Add BPF_F_NO_TUNNEL_KEY extension to bpf_skb_set_tunnel_key to
better support decap on GRE tunnel devices not operating in collect
metadata.
- Improve x86 JIT's codegen for PROBE_MEM runtime error checks.
- Remove the need for trace_printk_lock for bpf_trace_printk and
bpf_trace_vprintk helpers.
- Extend libbpf's bpf_tracing.h support for tracing arguments of
kprobes/uprobes and syscall as a special case.
- Significantly reduce the search time for module symbols by
livepatch and BPF.
- Enable cpumasks to be used as kptrs, which is useful for tracing
programs tracking which tasks end up running on which CPUs in
different time intervals.
- Add support for BPF trampoline on s390x and riscv64.
- Add capability to export the XDP features supported by the NIC.
- Add __bpf_kfunc tag for marking kernel functions as kfuncs.
- Add cgroup.memory=nobpf kernel parameter option to disable BPF
memory accounting for container environments.
Netfilter:
- Remove the CLUSTERIP target. It has been marked as obsolete for
years, and we still have WARN splats wrt races of the out-of-band
/proc interface installed by this target.
- Add 'destroy' commands to nf_tables. They are identical to the
existing 'delete' commands, but do not return an error if the
referenced object (set, chain, rule...) did not exist.
Driver API:
- Improve cpumask_local_spread() locality to help NICs set the right
IRQ affinity on AMD platforms.
- Separate C22 and C45 MDIO bus transactions more clearly.
- Introduce new DCB table to control DSCP rewrite on egress.
- Support configuration of Physical Layer Collision Avoidance (PLCA)
Reconciliation Sublayer (RS) (802.3cg-2019). Modern version of
shared medium Ethernet.
- Support for MAC Merge layer (IEEE 802.3-2018 clause 99). Allowing
preemption of low priority frames by high priority frames.
- Add support for controlling MACSec offload using netlink SET.
- Rework devlink instance refcounts to allow registration and
de-registration under the instance lock. Split the code into
multiple files, drop some of the unnecessarily granular locks and
factor out common parts of netlink operation handling.
- Add TX frame aggregation parameters (for USB drivers).
- Add a new attr TCA_EXT_WARN_MSG to report TC (offload) warning
messages with notifications for debug.
- Allow offloading of UDP NEW connections via act_ct.
- Add support for per action HW stats in TC.
- Support hardware miss to TC action (continue processing in SW from
a specific point in the action chain).
- Warn if old Wireless Extension user space interface is used with
modern cfg80211/mac80211 drivers. Do not support Wireless
Extensions for Wi-Fi 7 devices at all. Everyone should switch to
using nl80211 interface instead.
- Improve the CAN bit timing configuration. Use extack to return
error messages directly to user space, update the SJW handling,
including the definition of a new default value that will benefit
CAN-FD controllers, by increasing their oscillator tolerance.
New hardware / drivers:
- Ethernet:
- nVidia BlueField-3 support (control traffic driver)
- Ethernet support for imx93 SoCs
- Motorcomm yt8531 gigabit Ethernet PHY
- onsemi NCN26000 10BASE-T1S PHY (with support for PLCA)
- Microchip LAN8841 PHY (incl. cable diagnostics and PTP)
- Amlogic gxl MDIO mux
- WiFi:
- RealTek RTL8188EU (rtl8xxxu)
- Qualcomm Wi-Fi 7 devices (ath12k)
- CAN:
- Renesas R-Car V4H
Drivers:
- Bluetooth:
- Set Per Platform Antenna Gain (PPAG) for Intel controllers.
- Ethernet NICs:
- Intel (1G, igc):
- support TSN / Qbv / packet scheduling features of i226 model
- Intel (100G, ice):
- use GNSS subsystem instead of TTY
- multi-buffer XDP support
- extend support for GPIO pins to E823 devices
- nVidia/Mellanox:
- update the shared buffer configuration on PFC commands
- implement PTP adjphase function for HW offset control
- TC support for Geneve and GRE with VF tunnel offload
- more efficient crypto key management method
- multi-port eswitch support
- Netronome/Corigine:
- add DCB IEEE support
- support IPsec offloading for NFP3800
- Freescale/NXP (enetc):
- support XDP_REDIRECT for XDP non-linear buffers
- improve reconfig, avoid link flap and waiting for idle
- support MAC Merge layer
- Other NICs:
- sfc/ef100: add basic devlink support for ef100
- ionic: rx_push mode operation (writing descriptors via MMIO)
- bnxt: use the auxiliary bus abstraction for RDMA
- r8169: disable ASPM and reset bus in case of tx timeout
- cpsw: support QSGMII mode for J721e CPSW9G
- cpts: support pulse-per-second output
- ngbe: add an mdio bus driver
- usbnet: optimize usbnet_bh() by avoiding unnecessary queuing
- r8152: handle devices with FW with NCM support
- amd-xgbe: support 10Mbps, 2.5GbE speeds and rx-adaptation
- virtio-net: support multi buffer XDP
- virtio/vsock: replace virtio_vsock_pkt with sk_buff
- tsnep: XDP support
- Ethernet high-speed switches:
- nVidia/Mellanox (mlxsw):
- add support for latency TLV (in FW control messages)
- Microchip (sparx5):
- separate explicit and implicit traffic forwarding rules, make
the implicit rules always active
- add support for egress DSCP rewrite
- IS0 VCAP support (Ingress Classification)
- IS2 VCAP filters (protos, L3 addrs, L4 ports, flags, ToS
etc.)
- ES2 VCAP support (Egress Access Control)
- support for Per-Stream Filtering and Policing (802.1Q,
8.6.5.1)
- Ethernet embedded switches:
- Marvell (mv88e6xxx):
- add MAB (port auth) offload support
- enable PTP receive for mv88e6390
- NXP (ocelot):
- support MAC Merge layer
- support for the the vsc7512 internal copper phys
- Microchip:
- lan9303: convert to PHYLINK
- lan966x: support TC flower filter statistics
- lan937x: PTP support for KSZ9563/KSZ8563 and LAN937x
- lan937x: support Credit Based Shaper configuration
- ksz9477: support Energy Efficient Ethernet
- other:
- qca8k: convert to regmap read/write API, use bulk operations
- rswitch: Improve TX timestamp accuracy
- Intel WiFi (iwlwifi):
- EHT (Wi-Fi 7) rate reporting
- STEP equalizer support: transfer some STEP (connection to radio
on platforms with integrated wifi) related parameters from the
BIOS to the firmware.
- Qualcomm 802.11ax WiFi (ath11k):
- IPQ5018 support
- Fine Timing Measurement (FTM) responder role support
- channel 177 support
- MediaTek WiFi (mt76):
- per-PHY LED support
- mt7996: EHT (Wi-Fi 7) support
- Wireless Ethernet Dispatch (WED) reset support
- switch to using page pool allocator
- RealTek WiFi (rtw89):
- support new version of Bluetooth co-existance
- Mobile:
- rmnet: support TX aggregation"
* tag 'net-next-6.3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next: (1872 commits)
page_pool: add a comment explaining the fragment counter usage
net: ethtool: fix __ethtool_dev_mm_supported() implementation
ethtool: pse-pd: Fix double word in comments
xsk: add linux/vmalloc.h to xsk.c
sefltests: netdevsim: wait for devlink instance after netns removal
selftest: fib_tests: Always cleanup before exit
net/mlx5e: Align IPsec ASO result memory to be as required by hardware
net/mlx5e: TC, Set CT miss to the specific ct action instance
net/mlx5e: Rename CHAIN_TO_REG to MAPPED_OBJ_TO_REG
net/mlx5: Refactor tc miss handling to a single function
net/mlx5: Kconfig: Make tc offload depend on tc skb extension
net/sched: flower: Support hardware miss to tc action
net/sched: flower: Move filter handle initialization earlier
net/sched: cls_api: Support hardware miss to tc action
net/sched: Rename user cookie and act cookie
sfc: fix builds without CONFIG_RTC_LIB
sfc: clean up some inconsistent indentings
net/mlx4_en: Introduce flexible array to silence overflow warning
net: lan966x: Fix possible deadlock inside PTP
net/ulp: Remove redundant ->clone() test in inet_clone_ulp().
...
Diffstat (limited to 'drivers/net/ethernet/qlogic/qed/qed_spq.c')
-rw-r--r-- | drivers/net/ethernet/qlogic/qed/qed_spq.c | 1051 |
1 files changed, 1051 insertions, 0 deletions
diff --git a/drivers/net/ethernet/qlogic/qed/qed_spq.c b/drivers/net/ethernet/qlogic/qed/qed_spq.c new file mode 100644 index 000000000..d01b9245f --- /dev/null +++ b/drivers/net/ethernet/qlogic/qed/qed_spq.c @@ -0,0 +1,1051 @@ +// SPDX-License-Identifier: (GPL-2.0-only OR BSD-3-Clause) +/* QLogic qed NIC Driver + * Copyright (c) 2015-2017 QLogic Corporation + * Copyright (c) 2019-2020 Marvell International Ltd. + */ + +#include <linux/types.h> +#include <asm/byteorder.h> +#include <linux/io.h> +#include <linux/delay.h> +#include <linux/dma-mapping.h> +#include <linux/errno.h> +#include <linux/kernel.h> +#include <linux/list.h> +#include <linux/pci.h> +#include <linux/slab.h> +#include <linux/spinlock.h> +#include <linux/string.h> +#include "qed.h" +#include "qed_cxt.h" +#include "qed_dev_api.h" +#include "qed_hsi.h" +#include "qed_iro_hsi.h" +#include "qed_hw.h" +#include "qed_int.h" +#include "qed_iscsi.h" +#include "qed_mcp.h" +#include "qed_ooo.h" +#include "qed_reg_addr.h" +#include "qed_sp.h" +#include "qed_sriov.h" +#include "qed_rdma.h" + +/*************************************************************************** + * Structures & Definitions + ***************************************************************************/ + +#define SPQ_HIGH_PRI_RESERVE_DEFAULT (1) + +#define SPQ_BLOCK_DELAY_MAX_ITER (10) +#define SPQ_BLOCK_DELAY_US (10) +#define SPQ_BLOCK_SLEEP_MAX_ITER (1000) +#define SPQ_BLOCK_SLEEP_MS (5) + +/*************************************************************************** + * Blocking Imp. (BLOCK/EBLOCK mode) + ***************************************************************************/ +static void qed_spq_blocking_cb(struct qed_hwfn *p_hwfn, + void *cookie, + union event_ring_data *data, u8 fw_return_code) +{ + struct qed_spq_comp_done *comp_done; + + comp_done = (struct qed_spq_comp_done *)cookie; + + comp_done->fw_return_code = fw_return_code; + + /* Make sure completion done is visible on waiting thread */ + smp_store_release(&comp_done->done, 0x1); +} + +static int __qed_spq_block(struct qed_hwfn *p_hwfn, + struct qed_spq_entry *p_ent, + u8 *p_fw_ret, bool sleep_between_iter) +{ + struct qed_spq_comp_done *comp_done; + u32 iter_cnt; + + comp_done = (struct qed_spq_comp_done *)p_ent->comp_cb.cookie; + iter_cnt = sleep_between_iter ? SPQ_BLOCK_SLEEP_MAX_ITER + : SPQ_BLOCK_DELAY_MAX_ITER; + + while (iter_cnt--) { + /* Validate we receive completion update */ + if (smp_load_acquire(&comp_done->done) == 1) { /* ^^^ */ + if (p_fw_ret) + *p_fw_ret = comp_done->fw_return_code; + return 0; + } + + if (sleep_between_iter) + msleep(SPQ_BLOCK_SLEEP_MS); + else + udelay(SPQ_BLOCK_DELAY_US); + } + + return -EBUSY; +} + +static int qed_spq_block(struct qed_hwfn *p_hwfn, + struct qed_spq_entry *p_ent, + u8 *p_fw_ret, bool skip_quick_poll) +{ + struct qed_spq_comp_done *comp_done; + struct qed_ptt *p_ptt; + int rc; + + /* A relatively short polling period w/o sleeping, to allow the FW to + * complete the ramrod and thus possibly to avoid the following sleeps. + */ + if (!skip_quick_poll) { + rc = __qed_spq_block(p_hwfn, p_ent, p_fw_ret, false); + if (!rc) + return 0; + } + + /* Move to polling with a sleeping period between iterations */ + rc = __qed_spq_block(p_hwfn, p_ent, p_fw_ret, true); + if (!rc) + return 0; + + p_ptt = qed_ptt_acquire(p_hwfn); + if (!p_ptt) { + DP_NOTICE(p_hwfn, "ptt, failed to acquire\n"); + return -EAGAIN; + } + + DP_INFO(p_hwfn, "Ramrod is stuck, requesting MCP drain\n"); + rc = qed_mcp_drain(p_hwfn, p_ptt); + qed_ptt_release(p_hwfn, p_ptt); + if (rc) { + DP_NOTICE(p_hwfn, "MCP drain failed\n"); + goto err; + } + + /* Retry after drain */ + rc = __qed_spq_block(p_hwfn, p_ent, p_fw_ret, true); + if (!rc) + return 0; + + comp_done = (struct qed_spq_comp_done *)p_ent->comp_cb.cookie; + if (comp_done->done == 1) { + if (p_fw_ret) + *p_fw_ret = comp_done->fw_return_code; + return 0; + } +err: + p_ptt = qed_ptt_acquire(p_hwfn); + if (!p_ptt) + return -EBUSY; + qed_hw_err_notify(p_hwfn, p_ptt, QED_HW_ERR_RAMROD_FAIL, + "Ramrod is stuck [CID %08x %s:%02x %s:%02x echo %04x]\n", + le32_to_cpu(p_ent->elem.hdr.cid), + qed_get_ramrod_cmd_id_str(p_ent->elem.hdr.protocol_id, + p_ent->elem.hdr.cmd_id), + p_ent->elem.hdr.cmd_id, + qed_get_protocol_type_str(p_ent->elem.hdr.protocol_id), + p_ent->elem.hdr.protocol_id, + le16_to_cpu(p_ent->elem.hdr.echo)); + qed_ptt_release(p_hwfn, p_ptt); + + return -EBUSY; +} + +/*************************************************************************** + * SPQ entries inner API + ***************************************************************************/ +static int qed_spq_fill_entry(struct qed_hwfn *p_hwfn, + struct qed_spq_entry *p_ent) +{ + p_ent->flags = 0; + + switch (p_ent->comp_mode) { + case QED_SPQ_MODE_EBLOCK: + case QED_SPQ_MODE_BLOCK: + p_ent->comp_cb.function = qed_spq_blocking_cb; + break; + case QED_SPQ_MODE_CB: + break; + default: + DP_NOTICE(p_hwfn, "Unknown SPQE completion mode %d\n", + p_ent->comp_mode); + return -EINVAL; + } + + DP_VERBOSE(p_hwfn, + QED_MSG_SPQ, + "Ramrod hdr: [CID 0x%08x %s:0x%02x %s:0x%02x] Data ptr: [%08x:%08x] Cmpltion Mode: %s\n", + p_ent->elem.hdr.cid, + qed_get_ramrod_cmd_id_str(p_ent->elem.hdr.protocol_id, + p_ent->elem.hdr.cmd_id), + p_ent->elem.hdr.cmd_id, + qed_get_protocol_type_str(p_ent->elem.hdr.protocol_id), + p_ent->elem.hdr.protocol_id, + p_ent->elem.data_ptr.hi, p_ent->elem.data_ptr.lo, + D_TRINE(p_ent->comp_mode, QED_SPQ_MODE_EBLOCK, + QED_SPQ_MODE_BLOCK, "MODE_EBLOCK", "MODE_BLOCK", + "MODE_CB")); + + return 0; +} + +/*************************************************************************** + * HSI access + ***************************************************************************/ +static void qed_spq_hw_initialize(struct qed_hwfn *p_hwfn, + struct qed_spq *p_spq) +{ + struct core_conn_context *p_cxt; + struct qed_cxt_info cxt_info; + u16 physical_q; + int rc; + + cxt_info.iid = p_spq->cid; + + rc = qed_cxt_get_cid_info(p_hwfn, &cxt_info); + + if (rc < 0) { + DP_NOTICE(p_hwfn, "Cannot find context info for cid=%d\n", + p_spq->cid); + return; + } + + p_cxt = cxt_info.p_cxt; + + SET_FIELD(p_cxt->xstorm_ag_context.flags10, + XSTORM_CORE_CONN_AG_CTX_DQ_CF_EN, 1); + SET_FIELD(p_cxt->xstorm_ag_context.flags1, + XSTORM_CORE_CONN_AG_CTX_DQ_CF_ACTIVE, 1); + SET_FIELD(p_cxt->xstorm_ag_context.flags9, + XSTORM_CORE_CONN_AG_CTX_CONSOLID_PROD_CF_EN, 1); + + /* QM physical queue */ + physical_q = qed_get_cm_pq_idx(p_hwfn, PQ_FLAGS_LB); + p_cxt->xstorm_ag_context.physical_q0 = cpu_to_le16(physical_q); + + p_cxt->xstorm_st_context.spq_base_addr.lo = + DMA_LO_LE(p_spq->chain.p_phys_addr); + p_cxt->xstorm_st_context.spq_base_addr.hi = + DMA_HI_LE(p_spq->chain.p_phys_addr); +} + +static int qed_spq_hw_post(struct qed_hwfn *p_hwfn, + struct qed_spq *p_spq, struct qed_spq_entry *p_ent) +{ + struct qed_chain *p_chain = &p_hwfn->p_spq->chain; + struct core_db_data *p_db_data = &p_spq->db_data; + u16 echo = qed_chain_get_prod_idx(p_chain); + struct slow_path_element *elem; + + p_ent->elem.hdr.echo = cpu_to_le16(echo); + elem = qed_chain_produce(p_chain); + if (!elem) { + DP_NOTICE(p_hwfn, "Failed to produce from SPQ chain\n"); + return -EINVAL; + } + + *elem = p_ent->elem; /* struct assignment */ + + /* send a doorbell on the slow hwfn session */ + p_db_data->spq_prod = cpu_to_le16(qed_chain_get_prod_idx(p_chain)); + + /* make sure the SPQE is updated before the doorbell */ + wmb(); + + DOORBELL(p_hwfn, p_spq->db_addr_offset, *(u32 *)p_db_data); + + /* make sure doorbell is rang */ + wmb(); + + DP_VERBOSE(p_hwfn, QED_MSG_SPQ, + "Doorbelled [0x%08x, CID 0x%08x] with Flags: %02x agg_params: %02x, prod: %04x\n", + p_spq->db_addr_offset, + p_spq->cid, + p_db_data->params, + p_db_data->agg_flags, qed_chain_get_prod_idx(p_chain)); + + return 0; +} + +/*************************************************************************** + * Asynchronous events + ***************************************************************************/ +static int +qed_async_event_completion(struct qed_hwfn *p_hwfn, + struct event_ring_entry *p_eqe) +{ + qed_spq_async_comp_cb cb; + + if (!p_hwfn->p_spq) + return -EINVAL; + + if (p_eqe->protocol_id >= MAX_PROTOCOL_TYPE) { + DP_ERR(p_hwfn, "Wrong protocol: %s:%d\n", + qed_get_protocol_type_str(p_eqe->protocol_id), + p_eqe->protocol_id); + + return -EINVAL; + } + + cb = p_hwfn->p_spq->async_comp_cb[p_eqe->protocol_id]; + if (cb) { + return cb(p_hwfn, p_eqe->opcode, p_eqe->echo, + &p_eqe->data, p_eqe->fw_return_code); + } else { + DP_NOTICE(p_hwfn, + "Unknown Async completion for %s:%d\n", + qed_get_protocol_type_str(p_eqe->protocol_id), + p_eqe->protocol_id); + + return -EINVAL; + } +} + +int +qed_spq_register_async_cb(struct qed_hwfn *p_hwfn, + enum protocol_type protocol_id, + qed_spq_async_comp_cb cb) +{ + if (!p_hwfn->p_spq || (protocol_id >= MAX_PROTOCOL_TYPE)) + return -EINVAL; + + p_hwfn->p_spq->async_comp_cb[protocol_id] = cb; + return 0; +} + +void +qed_spq_unregister_async_cb(struct qed_hwfn *p_hwfn, + enum protocol_type protocol_id) +{ + if (!p_hwfn->p_spq || (protocol_id >= MAX_PROTOCOL_TYPE)) + return; + + p_hwfn->p_spq->async_comp_cb[protocol_id] = NULL; +} + +/*************************************************************************** + * EQ API + ***************************************************************************/ +void qed_eq_prod_update(struct qed_hwfn *p_hwfn, u16 prod) +{ + u32 addr = GET_GTT_REG_ADDR(GTT_BAR0_MAP_REG_USDM_RAM, + USTORM_EQE_CONS, p_hwfn->rel_pf_id); + + REG_WR16(p_hwfn, addr, prod); +} + +int qed_eq_completion(struct qed_hwfn *p_hwfn, void *cookie) +{ + struct qed_eq *p_eq = cookie; + struct qed_chain *p_chain = &p_eq->chain; + int rc = 0; + + /* take a snapshot of the FW consumer */ + u16 fw_cons_idx = le16_to_cpu(*p_eq->p_fw_cons); + + DP_VERBOSE(p_hwfn, QED_MSG_SPQ, "fw_cons_idx %x\n", fw_cons_idx); + + /* Need to guarantee the fw_cons index we use points to a usuable + * element (to comply with our chain), so our macros would comply + */ + if ((fw_cons_idx & qed_chain_get_usable_per_page(p_chain)) == + qed_chain_get_usable_per_page(p_chain)) + fw_cons_idx += qed_chain_get_unusable_per_page(p_chain); + + /* Complete current segment of eq entries */ + while (fw_cons_idx != qed_chain_get_cons_idx(p_chain)) { + struct event_ring_entry *p_eqe = qed_chain_consume(p_chain); + + if (!p_eqe) { + rc = -EINVAL; + break; + } + + DP_VERBOSE(p_hwfn, QED_MSG_SPQ, + "op %x prot %x res0 %x echo %x fwret %x flags %x\n", + p_eqe->opcode, + p_eqe->protocol_id, + p_eqe->reserved0, + le16_to_cpu(p_eqe->echo), + p_eqe->fw_return_code, + p_eqe->flags); + + if (GET_FIELD(p_eqe->flags, EVENT_RING_ENTRY_ASYNC)) { + if (qed_async_event_completion(p_hwfn, p_eqe)) + rc = -EINVAL; + } else if (qed_spq_completion(p_hwfn, + p_eqe->echo, + p_eqe->fw_return_code, + &p_eqe->data)) { + rc = -EINVAL; + } + + qed_chain_recycle_consumed(p_chain); + } + + qed_eq_prod_update(p_hwfn, qed_chain_get_prod_idx(p_chain)); + + /* Attempt to post pending requests */ + spin_lock_bh(&p_hwfn->p_spq->lock); + rc = qed_spq_pend_post(p_hwfn); + spin_unlock_bh(&p_hwfn->p_spq->lock); + + return rc; +} + +int qed_eq_alloc(struct qed_hwfn *p_hwfn, u16 num_elem) +{ + struct qed_chain_init_params params = { + .mode = QED_CHAIN_MODE_PBL, + .intended_use = QED_CHAIN_USE_TO_PRODUCE, + .cnt_type = QED_CHAIN_CNT_TYPE_U16, + .num_elems = num_elem, + .elem_size = sizeof(union event_ring_element), + }; + struct qed_eq *p_eq; + int ret; + + /* Allocate EQ struct */ + p_eq = kzalloc(sizeof(*p_eq), GFP_KERNEL); + if (!p_eq) + return -ENOMEM; + + ret = qed_chain_alloc(p_hwfn->cdev, &p_eq->chain, ¶ms); + if (ret) { + DP_NOTICE(p_hwfn, "Failed to allocate EQ chain\n"); + goto eq_allocate_fail; + } + + /* register EQ completion on the SP SB */ + qed_int_register_cb(p_hwfn, qed_eq_completion, + p_eq, &p_eq->eq_sb_index, &p_eq->p_fw_cons); + + p_hwfn->p_eq = p_eq; + return 0; + +eq_allocate_fail: + kfree(p_eq); + + return ret; +} + +void qed_eq_setup(struct qed_hwfn *p_hwfn) +{ + qed_chain_reset(&p_hwfn->p_eq->chain); +} + +void qed_eq_free(struct qed_hwfn *p_hwfn) +{ + if (!p_hwfn->p_eq) + return; + + qed_chain_free(p_hwfn->cdev, &p_hwfn->p_eq->chain); + + kfree(p_hwfn->p_eq); + p_hwfn->p_eq = NULL; +} + +/*************************************************************************** + * CQE API - manipulate EQ functionality + ***************************************************************************/ +static int qed_cqe_completion(struct qed_hwfn *p_hwfn, + struct eth_slow_path_rx_cqe *cqe, + enum protocol_type protocol) +{ + if (IS_VF(p_hwfn->cdev)) + return 0; + + /* @@@tmp - it's possible we'll eventually want to handle some + * actual commands that can arrive here, but for now this is only + * used to complete the ramrod using the echo value on the cqe + */ + return qed_spq_completion(p_hwfn, cqe->echo, 0, NULL); +} + +int qed_eth_cqe_completion(struct qed_hwfn *p_hwfn, + struct eth_slow_path_rx_cqe *cqe) +{ + int rc; + + rc = qed_cqe_completion(p_hwfn, cqe, PROTOCOLID_ETH); + if (rc) + DP_NOTICE(p_hwfn, + "Failed to handle RXQ CQE [cmd 0x%02x]\n", + cqe->ramrod_cmd_id); + + return rc; +} + +/*************************************************************************** + * Slow hwfn Queue (spq) + ***************************************************************************/ +void qed_spq_setup(struct qed_hwfn *p_hwfn) +{ + struct qed_spq *p_spq = p_hwfn->p_spq; + struct qed_spq_entry *p_virt = NULL; + struct core_db_data *p_db_data; + void __iomem *db_addr; + dma_addr_t p_phys = 0; + u32 i, capacity; + int rc; + + INIT_LIST_HEAD(&p_spq->pending); + INIT_LIST_HEAD(&p_spq->completion_pending); + INIT_LIST_HEAD(&p_spq->free_pool); + INIT_LIST_HEAD(&p_spq->unlimited_pending); + spin_lock_init(&p_spq->lock); + + /* SPQ empty pool */ + p_phys = p_spq->p_phys + offsetof(struct qed_spq_entry, ramrod); + p_virt = p_spq->p_virt; + + capacity = qed_chain_get_capacity(&p_spq->chain); + for (i = 0; i < capacity; i++) { + DMA_REGPAIR_LE(p_virt->elem.data_ptr, p_phys); + + list_add_tail(&p_virt->list, &p_spq->free_pool); + + p_virt++; + p_phys += sizeof(struct qed_spq_entry); + } + + /* Statistics */ + p_spq->normal_count = 0; + p_spq->comp_count = 0; + p_spq->comp_sent_count = 0; + p_spq->unlimited_pending_count = 0; + + bitmap_zero(p_spq->p_comp_bitmap, SPQ_RING_SIZE); + p_spq->comp_bitmap_idx = 0; + + /* SPQ cid, cannot fail */ + qed_cxt_acquire_cid(p_hwfn, PROTOCOLID_CORE, &p_spq->cid); + qed_spq_hw_initialize(p_hwfn, p_spq); + + /* reset the chain itself */ + qed_chain_reset(&p_spq->chain); + + /* Initialize the address/data of the SPQ doorbell */ + p_spq->db_addr_offset = qed_db_addr(p_spq->cid, DQ_DEMS_LEGACY); + p_db_data = &p_spq->db_data; + memset(p_db_data, 0, sizeof(*p_db_data)); + SET_FIELD(p_db_data->params, CORE_DB_DATA_DEST, DB_DEST_XCM); + SET_FIELD(p_db_data->params, CORE_DB_DATA_AGG_CMD, DB_AGG_CMD_MAX); + SET_FIELD(p_db_data->params, CORE_DB_DATA_AGG_VAL_SEL, + DQ_XCM_CORE_SPQ_PROD_CMD); + p_db_data->agg_flags = DQ_XCM_CORE_DQ_CF_CMD; + + /* Register the SPQ doorbell with the doorbell recovery mechanism */ + db_addr = (void __iomem *)((u8 __iomem *)p_hwfn->doorbells + + p_spq->db_addr_offset); + rc = qed_db_recovery_add(p_hwfn->cdev, db_addr, &p_spq->db_data, + DB_REC_WIDTH_32B, DB_REC_KERNEL); + if (rc) + DP_INFO(p_hwfn, + "Failed to register the SPQ doorbell with the doorbell recovery mechanism\n"); +} + +int qed_spq_alloc(struct qed_hwfn *p_hwfn) +{ + struct qed_chain_init_params params = { + .mode = QED_CHAIN_MODE_SINGLE, + .intended_use = QED_CHAIN_USE_TO_PRODUCE, + .cnt_type = QED_CHAIN_CNT_TYPE_U16, + .elem_size = sizeof(struct slow_path_element), + }; + struct qed_dev *cdev = p_hwfn->cdev; + struct qed_spq_entry *p_virt = NULL; + struct qed_spq *p_spq = NULL; + dma_addr_t p_phys = 0; + u32 capacity; + int ret; + + /* SPQ struct */ + p_spq = kzalloc(sizeof(*p_spq), GFP_KERNEL); + if (!p_spq) + return -ENOMEM; + + /* SPQ ring */ + ret = qed_chain_alloc(cdev, &p_spq->chain, ¶ms); + if (ret) { + DP_NOTICE(p_hwfn, "Failed to allocate SPQ chain\n"); + goto spq_chain_alloc_fail; + } + + /* allocate and fill the SPQ elements (incl. ramrod data list) */ + capacity = qed_chain_get_capacity(&p_spq->chain); + ret = -ENOMEM; + + p_virt = dma_alloc_coherent(&cdev->pdev->dev, + capacity * sizeof(struct qed_spq_entry), + &p_phys, GFP_KERNEL); + if (!p_virt) + goto spq_alloc_fail; + + p_spq->p_virt = p_virt; + p_spq->p_phys = p_phys; + p_hwfn->p_spq = p_spq; + + return 0; + +spq_alloc_fail: + qed_chain_free(cdev, &p_spq->chain); +spq_chain_alloc_fail: + kfree(p_spq); + + return ret; +} + +void qed_spq_free(struct qed_hwfn *p_hwfn) +{ + struct qed_spq *p_spq = p_hwfn->p_spq; + void __iomem *db_addr; + u32 capacity; + + if (!p_spq) + return; + + /* Delete the SPQ doorbell from the doorbell recovery mechanism */ + db_addr = (void __iomem *)((u8 __iomem *)p_hwfn->doorbells + + p_spq->db_addr_offset); + qed_db_recovery_del(p_hwfn->cdev, db_addr, &p_spq->db_data); + + if (p_spq->p_virt) { + capacity = qed_chain_get_capacity(&p_spq->chain); + dma_free_coherent(&p_hwfn->cdev->pdev->dev, + capacity * + sizeof(struct qed_spq_entry), + p_spq->p_virt, p_spq->p_phys); + } + + qed_chain_free(p_hwfn->cdev, &p_spq->chain); + kfree(p_spq); + p_hwfn->p_spq = NULL; +} + +int qed_spq_get_entry(struct qed_hwfn *p_hwfn, struct qed_spq_entry **pp_ent) +{ + struct qed_spq *p_spq = p_hwfn->p_spq; + struct qed_spq_entry *p_ent = NULL; + int rc = 0; + + spin_lock_bh(&p_spq->lock); + + if (list_empty(&p_spq->free_pool)) { + p_ent = kzalloc(sizeof(*p_ent), GFP_ATOMIC); + if (!p_ent) { + DP_NOTICE(p_hwfn, + "Failed to allocate an SPQ entry for a pending ramrod\n"); + rc = -ENOMEM; + goto out_unlock; + } + p_ent->queue = &p_spq->unlimited_pending; + } else { + p_ent = list_first_entry(&p_spq->free_pool, + struct qed_spq_entry, list); + list_del(&p_ent->list); + p_ent->queue = &p_spq->pending; + } + + *pp_ent = p_ent; + +out_unlock: + spin_unlock_bh(&p_spq->lock); + return rc; +} + +/* Locked variant; Should be called while the SPQ lock is taken */ +static void __qed_spq_return_entry(struct qed_hwfn *p_hwfn, + struct qed_spq_entry *p_ent) +{ + list_add_tail(&p_ent->list, &p_hwfn->p_spq->free_pool); +} + +void qed_spq_return_entry(struct qed_hwfn *p_hwfn, struct qed_spq_entry *p_ent) +{ + spin_lock_bh(&p_hwfn->p_spq->lock); + __qed_spq_return_entry(p_hwfn, p_ent); + spin_unlock_bh(&p_hwfn->p_spq->lock); +} + +/** + * qed_spq_add_entry() - Add a new entry to the pending list. + * Should be used while lock is being held. + * + * @p_hwfn: HW device data. + * @p_ent: An entry to add. + * @priority: Desired priority. + * + * Adds an entry to the pending list is there is room (an empty + * element is available in the free_pool), or else places the + * entry in the unlimited_pending pool. + * + * Return: zero on success, -EINVAL on invalid @priority. + */ +static int qed_spq_add_entry(struct qed_hwfn *p_hwfn, + struct qed_spq_entry *p_ent, + enum spq_priority priority) +{ + struct qed_spq *p_spq = p_hwfn->p_spq; + + if (p_ent->queue == &p_spq->unlimited_pending) { + if (list_empty(&p_spq->free_pool)) { + list_add_tail(&p_ent->list, &p_spq->unlimited_pending); + p_spq->unlimited_pending_count++; + + return 0; + } else { + struct qed_spq_entry *p_en2; + + p_en2 = list_first_entry(&p_spq->free_pool, + struct qed_spq_entry, list); + list_del(&p_en2->list); + + /* Copy the ring element physical pointer to the new + * entry, since we are about to override the entire ring + * entry and don't want to lose the pointer. + */ + p_ent->elem.data_ptr = p_en2->elem.data_ptr; + + *p_en2 = *p_ent; + + /* EBLOCK responsible to free the allocated p_ent */ + if (p_ent->comp_mode != QED_SPQ_MODE_EBLOCK) + kfree(p_ent); + else + p_ent->post_ent = p_en2; + + p_ent = p_en2; + } + } + + /* entry is to be placed in 'pending' queue */ + switch (priority) { + case QED_SPQ_PRIORITY_NORMAL: + list_add_tail(&p_ent->list, &p_spq->pending); + p_spq->normal_count++; + break; + case QED_SPQ_PRIORITY_HIGH: + list_add(&p_ent->list, &p_spq->pending); + p_spq->high_count++; + break; + default: + return -EINVAL; + } + + return 0; +} + +/*************************************************************************** + * Accessor + ***************************************************************************/ +u32 qed_spq_get_cid(struct qed_hwfn *p_hwfn) +{ + if (!p_hwfn->p_spq) + return 0xffffffff; /* illegal */ + return p_hwfn->p_spq->cid; +} + +/*************************************************************************** + * Posting new Ramrods + ***************************************************************************/ +static int qed_spq_post_list(struct qed_hwfn *p_hwfn, + struct list_head *head, u32 keep_reserve) +{ + struct qed_spq *p_spq = p_hwfn->p_spq; + int rc; + + while (qed_chain_get_elem_left(&p_spq->chain) > keep_reserve && + !list_empty(head)) { + struct qed_spq_entry *p_ent = + list_first_entry(head, struct qed_spq_entry, list); + list_move_tail(&p_ent->list, &p_spq->completion_pending); + p_spq->comp_sent_count++; + + rc = qed_spq_hw_post(p_hwfn, p_spq, p_ent); + if (rc) { + list_del(&p_ent->list); + __qed_spq_return_entry(p_hwfn, p_ent); + return rc; + } + } + + return 0; +} + +int qed_spq_pend_post(struct qed_hwfn *p_hwfn) +{ + struct qed_spq *p_spq = p_hwfn->p_spq; + struct qed_spq_entry *p_ent = NULL; + + while (!list_empty(&p_spq->free_pool)) { + if (list_empty(&p_spq->unlimited_pending)) + break; + + p_ent = list_first_entry(&p_spq->unlimited_pending, + struct qed_spq_entry, list); + if (!p_ent) + return -EINVAL; + + list_del(&p_ent->list); + + qed_spq_add_entry(p_hwfn, p_ent, p_ent->priority); + } + + return qed_spq_post_list(p_hwfn, &p_spq->pending, + SPQ_HIGH_PRI_RESERVE_DEFAULT); +} + +static void qed_spq_recov_set_ret_code(struct qed_spq_entry *p_ent, + u8 *fw_return_code) +{ + if (!fw_return_code) + return; + + if (p_ent->elem.hdr.protocol_id == PROTOCOLID_ROCE || + p_ent->elem.hdr.protocol_id == PROTOCOLID_IWARP) + *fw_return_code = RDMA_RETURN_OK; +} + +/* Avoid overriding of SPQ entries when getting out-of-order completions, by + * marking the completions in a bitmap and increasing the chain consumer only + * for the first successive completed entries. + */ +static void qed_spq_comp_bmap_update(struct qed_hwfn *p_hwfn, __le16 echo) +{ + u16 pos = le16_to_cpu(echo) % SPQ_RING_SIZE; + struct qed_spq *p_spq = p_hwfn->p_spq; + + __set_bit(pos, p_spq->p_comp_bitmap); + while (test_bit(p_spq->comp_bitmap_idx, + p_spq->p_comp_bitmap)) { + __clear_bit(p_spq->comp_bitmap_idx, + p_spq->p_comp_bitmap); + p_spq->comp_bitmap_idx++; + qed_chain_return_produced(&p_spq->chain); + } +} + +int qed_spq_post(struct qed_hwfn *p_hwfn, + struct qed_spq_entry *p_ent, u8 *fw_return_code) +{ + int rc = 0; + struct qed_spq *p_spq = p_hwfn ? p_hwfn->p_spq : NULL; + bool b_ret_ent = true; + bool eblock; + + if (!p_hwfn) + return -EINVAL; + + if (!p_ent) { + DP_NOTICE(p_hwfn, "Got a NULL pointer\n"); + return -EINVAL; + } + + if (p_hwfn->cdev->recov_in_prog) { + DP_VERBOSE(p_hwfn, + QED_MSG_SPQ, + "Recovery is in progress. Skip spq post [%s:%02x %s:%02x]\n", + qed_get_ramrod_cmd_id_str(p_ent->elem.hdr.protocol_id, + p_ent->elem.hdr.cmd_id), + p_ent->elem.hdr.cmd_id, + qed_get_protocol_type_str(p_ent->elem.hdr.protocol_id), + p_ent->elem.hdr.protocol_id); + + /* Let the flow complete w/o any error handling */ + qed_spq_recov_set_ret_code(p_ent, fw_return_code); + return 0; + } + + /* Complete the entry */ + rc = qed_spq_fill_entry(p_hwfn, p_ent); + + spin_lock_bh(&p_spq->lock); + + /* Check return value after LOCK is taken for cleaner error flow */ + if (rc) + goto spq_post_fail; + + /* Check if entry is in block mode before qed_spq_add_entry, + * which might kfree p_ent. + */ + eblock = (p_ent->comp_mode == QED_SPQ_MODE_EBLOCK); + + /* Add the request to the pending queue */ + rc = qed_spq_add_entry(p_hwfn, p_ent, p_ent->priority); + if (rc) + goto spq_post_fail; + + rc = qed_spq_pend_post(p_hwfn); + if (rc) { + /* Since it's possible that pending failed for a different + * entry [although unlikely], the failed entry was already + * dealt with; No need to return it here. + */ + b_ret_ent = false; + goto spq_post_fail; + } + + spin_unlock_bh(&p_spq->lock); + + if (eblock) { + /* For entries in QED BLOCK mode, the completion code cannot + * perform the necessary cleanup - if it did, we couldn't + * access p_ent here to see whether it's successful or not. + * Thus, after gaining the answer perform the cleanup here. + */ + rc = qed_spq_block(p_hwfn, p_ent, fw_return_code, + p_ent->queue == &p_spq->unlimited_pending); + + if (p_ent->queue == &p_spq->unlimited_pending) { + struct qed_spq_entry *p_post_ent = p_ent->post_ent; + + kfree(p_ent); + + /* Return the entry which was actually posted */ + p_ent = p_post_ent; + } + + if (rc) + goto spq_post_fail2; + + /* return to pool */ + qed_spq_return_entry(p_hwfn, p_ent); + } + return rc; + +spq_post_fail2: + spin_lock_bh(&p_spq->lock); + list_del(&p_ent->list); + qed_spq_comp_bmap_update(p_hwfn, p_ent->elem.hdr.echo); + +spq_post_fail: + /* return to the free pool */ + if (b_ret_ent) + __qed_spq_return_entry(p_hwfn, p_ent); + spin_unlock_bh(&p_spq->lock); + + return rc; +} + +int qed_spq_completion(struct qed_hwfn *p_hwfn, + __le16 echo, + u8 fw_return_code, + union event_ring_data *p_data) +{ + struct qed_spq *p_spq; + struct qed_spq_entry *p_ent = NULL; + struct qed_spq_entry *tmp; + struct qed_spq_entry *found = NULL; + + if (!p_hwfn) + return -EINVAL; + + p_spq = p_hwfn->p_spq; + if (!p_spq) + return -EINVAL; + + spin_lock_bh(&p_spq->lock); + list_for_each_entry_safe(p_ent, tmp, &p_spq->completion_pending, list) { + if (p_ent->elem.hdr.echo == echo) { + list_del(&p_ent->list); + qed_spq_comp_bmap_update(p_hwfn, echo); + p_spq->comp_count++; + found = p_ent; + break; + } + + /* This is relatively uncommon - depends on scenarios + * which have mutliple per-PF sent ramrods. + */ + DP_VERBOSE(p_hwfn, QED_MSG_SPQ, + "Got completion for echo %04x - doesn't match echo %04x in completion pending list\n", + le16_to_cpu(echo), + le16_to_cpu(p_ent->elem.hdr.echo)); + } + + /* Release lock before callback, as callback may post + * an additional ramrod. + */ + spin_unlock_bh(&p_spq->lock); + + if (!found) { + DP_NOTICE(p_hwfn, + "Failed to find an entry this EQE [echo %04x] completes\n", + le16_to_cpu(echo)); + return -EEXIST; + } + + DP_VERBOSE(p_hwfn, QED_MSG_SPQ, + "Complete EQE [echo %04x]: func %p cookie %p)\n", + le16_to_cpu(echo), + p_ent->comp_cb.function, p_ent->comp_cb.cookie); + if (found->comp_cb.function) + found->comp_cb.function(p_hwfn, found->comp_cb.cookie, p_data, + fw_return_code); + else + DP_VERBOSE(p_hwfn, + QED_MSG_SPQ, + "Got a completion without a callback function\n"); + + if (found->comp_mode != QED_SPQ_MODE_EBLOCK) + /* EBLOCK is responsible for returning its own entry into the + * free list. + */ + qed_spq_return_entry(p_hwfn, found); + + return 0; +} + +#define QED_SPQ_CONSQ_ELEM_SIZE 0x80 + +int qed_consq_alloc(struct qed_hwfn *p_hwfn) +{ + struct qed_chain_init_params params = { + .mode = QED_CHAIN_MODE_PBL, + .intended_use = QED_CHAIN_USE_TO_PRODUCE, + .cnt_type = QED_CHAIN_CNT_TYPE_U16, + .num_elems = QED_CHAIN_PAGE_SIZE / QED_SPQ_CONSQ_ELEM_SIZE, + .elem_size = QED_SPQ_CONSQ_ELEM_SIZE, + }; + struct qed_consq *p_consq; + int ret; + + /* Allocate ConsQ struct */ + p_consq = kzalloc(sizeof(*p_consq), GFP_KERNEL); + if (!p_consq) + return -ENOMEM; + + /* Allocate and initialize ConsQ chain */ + ret = qed_chain_alloc(p_hwfn->cdev, &p_consq->chain, ¶ms); + if (ret) { + DP_NOTICE(p_hwfn, "Failed to allocate ConsQ chain"); + goto consq_alloc_fail; + } + + p_hwfn->p_consq = p_consq; + + return 0; + +consq_alloc_fail: + kfree(p_consq); + + return ret; +} + +void qed_consq_setup(struct qed_hwfn *p_hwfn) +{ + qed_chain_reset(&p_hwfn->p_consq->chain); +} + +void qed_consq_free(struct qed_hwfn *p_hwfn) +{ + if (!p_hwfn->p_consq) + return; + + qed_chain_free(p_hwfn->cdev, &p_hwfn->p_consq->chain); + + kfree(p_hwfn->p_consq); + p_hwfn->p_consq = NULL; +} |