| /linux/tools/testing/selftests/drivers/net/ |
| H A D | shaper.py | 2 # SPDX-License-Identifier: GPL-2.0
3 # pylint: disable=too-many-lines
15 def _delete_shaper(cfg, nl_shaper, handle) -> None:
16 """ Delete the shaper identified by handle, ignoring a missing-shaper error. """
26 qcnt = len(glob.glob(f"/sys/class/net/{cfg.ifname}/queues/tx-*"))
41 def _require_caps(cfg, nl_shaper, scope, caps, msg) -> None:
53 def get_shapers(cfg, nl_shaper) -> None:
64 def get_caps(cfg, nl_shaper) -> None:
72 # Each device implementing shaper support must support some
76 def set_qshapers(cfg, nl_shaper) -> None:
[all …]
|
| /linux/Documentation/userspace-api/ |
| H A D | iommufd.rst | 1 .. SPDX-License-Identifier: GPL-2.0+
19 At minimum iommufd provides universal support of managing I/O address spaces and
20 I/O page tables for all IOMMUs, with room in the design to add non-generic
31 --------------------
35 - IOMMUFD_OBJ_IOAS, representing an I/O address space (IOAS), allowing map/unmap
41 - IOMMUFD_OBJ_DEVICE, representing a device that is bound to iommufd by an
44 - IOMMUFD_OBJ_HWPT_PAGING, representing an actual hardware I/O page table
48 feature flag. This can be either an UNMANAGED stage-1 domain for a device
49 running in the user space, or a nesting parent stage-2 domain for mappings
50 from guest-level physical addresses to host-level physical addresses.
[all …]
|
| /linux/Documentation/netlink/specs/ |
| H A D | net_shaper.yaml | 1 # SPDX-License-Identifier: ((GPL-2.0 WITH Linux-syscall-note) OR BSD-3-Clause)
2 ---
3 name: net-shaper
31 support for complex scheduling hierarchy, or for some shaping
33 @cap-get operation.
36 -
38 name: max-handle-id
41 -
45 render-max: true
47 - name: unspec
[all …]
|
| /linux/include/linux/ |
| H A D | kcsan.h | 1 /* SPDX-License-Identifier: GPL-2.0 */
4 * data structures to set up runtime. See kcsan-checks.h for explicit checks and
5 * modifiers. For more info please see Documentation/dev-tools/kcsan.rst.
13 #include <linux/kcsan-checks.h>
20 * task_struct, and interrupts access internal per-CPU storage.
30 * track of nesting. Both (a) and (b) are entirely independent of each
32 * vice-versa.
41 * To support these cases, we independently track the depth of nesting
48 * Access mask for all accesses if non-zero.
58 * barriers; only keep 1 to keep fast-path complexity manageable.
[all …]
|
| H A D | highmem.h | 1 /* SPDX-License-Identifier: GPL-2.0 */
14 #include "highmem-internal.h"
17 * kmap - Map a page for long term usage
40 * kunmap - Unmap the virtual address mapped by kmap()
49 * kmap_to_page - Get the page for a kmap'ed address
57 * kmap_flush_unused - Flush all unused kmap mappings in order to
63 * kmap_local_page - Map a page for temporary usage
70 * Requires careful handling when nesting multiple mappings because the map
99 * kmap_local_folio - Map a page in this folio for temporary usage
103 * Requires careful handling when nesting multiple mappings because the map
[all …]
|
| /linux/Documentation/virt/kvm/x86/ |
| H A D | running-nested-guests.rst | 1 .. SPDX-License-Identifier: GPL-2.0
8 can be KVM-based or a different hypervisor). The straightforward
12 .----------------. .----------------.
17 |----------------'--'----------------|
22 .------------------------------------------------------.
25 |------------------------------------------------------|
27 '------------------------------------------------------'
31 - L0 – level-0; the bare metal host, running KVM
33 - L1 – level-1 guest; a VM running on L0; also called the "guest
36 - L2 – level-2 guest; a VM running on L1, this is the "nested guest"
[all …]
|
| /linux/Documentation/bpf/ |
| H A D | map_of_maps.rst | 1 .. SPDX-License-Identifier: GPL-2.0-only
9 - ``BPF_MAP_TYPE_ARRAY_OF_MAPS`` and ``BPF_MAP_TYPE_HASH_OF_MAPS`` were
13 purpose support for map in map storage. One level of nesting is supported, where
15 ``array_of_maps->sock_map``.
27 - Multi-level nesting is not supported.
28 - Any BPF map type can be used as an inner map, except for
30 - A BPF program cannot update or delete outer map entries.
32 For ``BPF_MAP_TYPE_ARRAY_OF_MAPS`` the key is an unsigned 32-bit integer index
38 the max_entries limit that you specify. Hash maps use pre-allocation of hash
40 pre-allocation when it is too memory expensive.
[all …]
|
| /linux/Documentation/networking/ |
| H A D | openvswitch.rst | 1 .. SPDX-License-Identifier: GPL-2.0
8 flow-level packet processing on selected network devices. It can be
10 VLAN processing, network access control, flow-based network control,
26 packets of the same type entirely in-kernel).
30 ----------------------
36 might even be desirable, someday, to drop support for parsing
41 To support this forward and backward compatibility, whenever the
45 kernel-provided version:
47 - If userspace's notion of the flow key for the packet matches the
50 - If the kernel's flow key includes more fields than the userspace
[all …]
|
| /linux/Documentation/userspace-api/netlink/ |
| H A D | genetlink-legacy.rst | 1 .. SPDX-License-Identifier: BSD-3-Clause
4 Netlink specification support for legacy Generic Netlink families
9 the ``genetlink-legacy`` protocol level.
15 -------
29 --------------------
31 New Netlink families should use ``multi-attr`` to define arrays.
35 For reference the ``multi-attr`` array may look like this::
37 [ARRAY-ATTR]
41 [SOME-OTHER-ATTR]
42 [ARRAY-ATTR]
[all …]
|
| /linux/drivers/acpi/acpica/ |
| H A D | utownerid.c | 1 // SPDX-License-Identifier: BSD-3-Clause OR GPL-2.0
4 * Module Name: utownerid - Support for Table/Method Owner IDs
19 * PARAMETERS: owner_id - Where the new owner ID is returned
114 * If this error happens, there may be very deep nesting of invoked in acpi_ut_allocate_owner_id()
130 * PARAMETERS: owner_id_ptr - Pointer to a previously allocated owner_ID
136 * DESCRIPTION: Release a table or method owner ID. Valid IDs are 1 - 255
169 owner_id--; in acpi_ut_release_owner_id()
182 "Attempted release of non-allocated OwnerId: 0x%3.3X", in acpi_ut_release_owner_id()
|
| /linux/Documentation/RCU/Design/Data-Structures/ |
| H A D | Data-Structures.rst | 15 Data-Structure Relationships
25 .. kernel-figure:: BigTreeClassicRCU.svg
34 which results in a three-level ``rcu_node`` tree.
38 The purpose of this combining tree is to allow per-CPU events
39 such as quiescent states, dyntick-idle transitions,
42 Quiescent states are recorded by the per-CPU ``rcu_data`` structures,
43 and other events are recorded by the leaf-level ``rcu_node``
54 As can be seen from the diagram, on a 64-bit system
55 a two-level tree with 64 leaves can accommodate 1,024 CPUs, with a fanout
58 +-----------------------------------------------------------------------+
[all …]
|
| /linux/scripts/ |
| H A D | unifdef.c | 2 * Copyright (c) 2002 - 2011 Tony Finch <dot@dotat.at>
27 * unifdef - remove ifdef'ed lines
30 * It was rewritten to support ANSI C by Tony Finch. The original version
31 * of unifdef carried the 4-clause BSD copyright licence. None of its code
60 "@(#) $Version: unifdef-2.5 $\n"
99 IS_TRUE_PREFIX, /* first non-false #(el)if is true */
100 IS_PASS_MIDDLE, /* first non-false #(el)if is unknown */
122 STARTING_COMMENT, /* just after slash-backslash-newline */
123 FINISHING_COMMENT, /* star-backslash-newline in a C comment */
146 #define MAXDEPTH 64 /* maximum #if nesting */
[all …]
|
| /linux/kernel/bpf/ |
| H A D | rqspinlock.c | 1 // SPDX-License-Identifier: GPL-2.0-or-later
5 * (C) Copyright 2013-2015 Hewlett-Packard Development Company, L.P.
6 * (C) Copyright 2013-2014,2018 Red Hat, Inc.
8 * (C) Copyright 2015 Hewlett-Packard Enterprise Development LP
9 * (C) Copyright 2024-2025 Meta Platforms, Inc. and affiliates.
44 * The basic principle of a queue-based spinlock can best be understood
45 * by studying a classic queue-based spinlock implementation called the
47 * Synchronization on Shared-Memory Multiprocessors by Mellor-Crummey and
58 * unlock the next pending (next->locked), we compress both these: {tail,
59 * next->locked} into a single u32 value.
[all …]
|
| /linux/arch/x86/include/asm/ |
| H A D | nospec-branch.h | 1 /* SPDX-License-Identifier: GPL-2.0 */
13 #include <asm/msr-index.h>
16 #include <asm/ptrace-abi.h>
79 #include <asm/asm-offsets.h>
82 movq $-1, PER_CPU_VAR(__x86_call_depth);
112 * This is required in various cases for retpoline and IBRS-based
120 * from C via asm(".include <asm/nospec-branch.h>") but let's not go there.
138 * Google experimented with loop-unrolling and this turned out to be
139 * the optimal version - two calls, each with their own speculation
170 * To mitigate Post-Barrier RSB speculation, one CALL instruction must be
[all …]
|
| /linux/arch/arm64/kernel/probes/ |
| H A D | kprobes.c | 1 // SPDX-License-Identifier: GPL-2.0-only
5 * Kprobes support for ARM64
28 #include <asm/debug-monitors.h>
31 #include <asm/text-patching.h>
37 #include "decode-insn.h"
61 kprobe_opcode_t *addr = p->ainsn.xol_insn; in arch_prepare_ss_slot()
67 * - That the I-cache maintenance for these instructions is complete in arch_prepare_ss_slot()
69 * ensures this, but just omits causing a Context-Synchronization-Event in arch_prepare_ss_slot()
72 * - That the kprobe BRK results in an exception (and consequently a in arch_prepare_ss_slot()
73 * Context-Synchronoization-Event), which ensures that the CPU will in arch_prepare_ss_slot()
[all …]
|
| /linux/include/linux/mmc/ |
| H A D | host.h | 1 /* SPDX-License-Identifier: GPL-2.0-only */
12 #include <linux/fault-inject.h>
18 #include <linux/dma-direction.h>
19 #include <linux/blk-crypto-profile.h>
164 * order to support double buffering of requests (prepare one
187 * ios->clock might be 0. For some controllers, setting 0Hz
197 * 1 for a read-only card
198 * -ENOSYS when not supported (equal to NULL callback)
207 * -ENOSYS when not supported (equal to NULL callback)
233 /* Optional callback to prepare for SD high-speed tuning */
[all …]
|
| /linux/kernel/events/ |
| H A D | ring_buffer.c | 1 // SPDX-License-Identifier: GPL-2.0
3 * Performance events ring-buffer code:
6 * Copyright (C) 2008-2011 Red Hat, Inc., Ingo Molnar
7 * Copyright (C) 2008-2011 Red Hat, Inc., Peter Zijlstra
22 atomic_set(&handle->rb->poll, EPOLLIN | EPOLLRDNORM); in perf_output_wakeup()
24 handle->event->pending_wakeup = 1; in perf_output_wakeup()
26 if (*perf_event_fasync(handle->event) && !handle->event->pending_kill) in perf_output_wakeup()
27 handle->event->pending_kill = POLL_IN; in perf_output_wakeup()
29 irq_work_queue(&handle->event->pending_irq); in perf_output_wakeup()
37 * We only publish the head (and generate a wakeup) when the outer-most
[all …]
|
| /linux/kernel/rcu/ |
| H A D | tree_stall.h | 1 // SPDX-License-Identifier: GPL-2.0+
110 /* Limit-check stall timeouts specified at boottime and runtime. */
161 /* If so specified via sysctl, panic, yielding cleaner stall-warning output. */
181 * rcu_cpu_stall_reset - restart stall-warning timeout for current grace period
208 smp_mb(); // ->gp_start before ->jiffies_stall and caller's ->gp_seq. in record_gp_stall_check_time()
215 /* Zero ->ticks_this_gp and snapshot the number of RCU softirq handlers. */
218 rdp->ticks_this_gp = 0; in zero_cpu_stall_ticks()
219 rdp->softirq_snap = kstat_softirqs_cpu(RCU_SOFTIRQ, smp_processor_id()); in zero_cpu_stall_ticks()
220 WRITE_ONCE(rdp->last_fqs_resched, jiffies); in zero_cpu_stall_ticks()
236 WARN_ONCE(1, "Kicking %s grace-period kthread\n", in rcu_stall_kick_kthreads()
[all …]
|
| /linux/drivers/iommu/arm/arm-smmu-v3/ |
| H A D | arm-smmu-v3-iommufd.c | 1 // SPDX-License-Identifier: GPL-2.0
8 #include "arm-smmu-v3.h"
14 const struct arm_smmu_impl_ops *impl_ops = master->smmu->impl_ops; in arm_smmu_hw_info()
21 if (!impl_ops || !impl_ops->hw_info) in arm_smmu_hw_info()
22 return ERR_PTR(-EOPNOTSUPP); in arm_smmu_hw_info()
23 return impl_ops->hw_info(master->smmu, length, type); in arm_smmu_hw_info()
28 return ERR_PTR(-ENOMEM); in arm_smmu_hw_info()
30 base_idr = master->smmu->base + ARM_SMMU_IDR0; in arm_smmu_hw_info()
32 info->idr[i] = readl_relaxed(base_idr + i); in arm_smmu_hw_info()
33 info->iidr = readl_relaxed(master->smmu->base + ARM_SMMU_IIDR); in arm_smmu_hw_info()
[all …]
|
| /linux/net/rds/ |
| H A D | rds.h | 1 /* SPDX-License-Identifier: GPL-2.0 */
56 /* Used to limit both RDMA and non-RDMA RDS message to 1MB */
96 #define RDS_MPATH_HASH(rs, n) (jhash_1word(ntohs((rs)->rs_bound_port), \
97 (rs)->rs_hash_initval) & ((n) - 1))
180 return read_pnet(&conn->c_net); in rds_conn_net()
186 write_pnet(&conn->c_net, net); in rds_conn_net_set()
226 * Reserved - indicates end of extensions
352 * conn->c_lock. m_sock_item has additional meaning in that once it is empty
366 * nesting. As paths iterate over messages on a sock, or conn, they must
408 spin_lock_init(&q->lock); in rds_message_zcopy_queue_init()
[all …]
|
| /linux/drivers/mtd/chips/ |
| H A D | cfi_probe.c | 1 // SPDX-License-Identifier: GPL-2.0
83 for (f = fixups; f->fixup; f++) { in cfi_early_fixup()
84 if (((f->mfr == CFI_MFR_ANY) || (f->mfr == cfi->mfr)) && in cfi_early_fixup()
85 ((f->id == CFI_ID_ANY) || (f->id == cfi->id))) { in cfi_early_fixup()
86 f->fixup(cfi); in cfi_early_fixup()
101 if ((base + 0) >= map->size) { in cfi_probe_chip()
104 (unsigned long)base, map->size -1); in cfi_probe_chip()
107 if ((base + 0xff) >= map->size) { in cfi_probe_chip()
110 (unsigned long)base + 0x55, map->size -1); in cfi_probe_chip()
120 if (!cfi->numchips) { in cfi_probe_chip()
[all …]
|
| /linux/drivers/dma-buf/ |
| H A D | dma-fence.c | 1 // SPDX-License-Identifier: GPL-2.0-only
3 * Fence mechanism for dma-buf and to allow for asynchronous dma access
16 #include <linux/dma-fence.h>
49 * Since the purposes of fences is to facilitate cross-device and
50 * cross-application synchronization, there's multiple ways to use one:
52 * - Individual fences can be exposed as a &sync_file, accessed as a file
57 * - Som
[all...] |
| /linux/Documentation/locking/ |
| H A D | rt-mutex-design.rst | 2 RT-mutex implementation design
12 Documentation/locking/rt-mutex.rst. Although this document does explain problems
22 ----------------------------
49 A ---+
52 C +----+
54 B +-------->
59 -------------------------
74 -----------
80 - The PI chain is an ordered series of locks and processes that cause
86 - In this document, to differentiate from locks that implement
[all …]
|
| /linux/drivers/base/ |
| H A D | arch_topology.c | 1 // SPDX-License-Identifier: GPL-2.0
59 * Task scheduler behavior depends on frequency invariance support, in update_scale_freq_invariant()
60 * either cpufreq or counter driven. If the support status changes as in update_scale_freq_invariant()
89 if (!sfd || sfd->source != SCALE_FREQ_SOURCE_ARCH) { in topology_set_scale_freq_source()
112 if (sfd && sfd->source == source) { in topology_clear_scale_freq_source()
122 * use-after-free races. in topology_clear_scale_freq_source()
135 sfd->set_freq_scale(); in topology_scale_freq_tick()
167 * topology_update_hw_pressure() - Update HW pressure for CPUs
173 * operating on stale data when hot-plug is used for some CPUs. The
200 pressure = max_capacity - capacity; in topology_update_hw_pressure()
[all …]
|
| /linux/kernel/trace/ |
| H A D | Kconfig | 1 # SPDX-License-Identifier: GPL-2.0-only
21 API, which will be used by other function-entry hooking
27 See Documentation/trace/ftrace-design.rst
32 See Documentation/trace/ftrace-design.rst
45 See Documentation/trace/ftrace-design.rst
95 See Documentation/trace/ftrace-design.rst
100 Arch supports the gcc options -pg with -mfentry
105 Arch supports the gcc options -pg with -mrecord-mcount and -nop-mcount
110 Arch supports objtool --mcount
115 Arch supports the objtool options --mcount with --mnop.
[all …]
|