/linux/mm/ |
H A D | workingset.c | 97 * at the time of a page's eviction, and comparing it to another 175 * On eviction, a snapshot of this counter (along with some bits to 199 static void *pack_shadow(int memcgid, pg_data_t *pgdat, unsigned long eviction, in pack_shadow() argument 202 eviction &= EVICTION_MASK; in pack_shadow() 203 eviction = (eviction << MEM_CGROUP_ID_SHIFT) | memcgid; in pack_shadow() 204 eviction = (eviction << NODES_SHIFT) | pgdat->node_id; in pack_shadow() 205 eviction = (eviction << WORKINGSET_SHIF in pack_shadow() 384 unsigned long eviction; workingset_eviction() local 428 unsigned long eviction; workingset_test_recent() local [all...] |
/linux/Documentation/gpu/ |
H A D | drm-compute.rst | 40 This creates a third option, using cgroups to prevent eviction. Both GPU and 42 eviction would be made cgroup aware. This allows the GPU to be partitioned 47 interface, with similar semantics for min/low/high/max, if eviction can
|
H A D | drm-vm-bind-locking.rst | 150 Revalidation and eviction of local objects 219 The gpu_vm_bos marked for eviction are put on the gpu_vm's evict list, 220 which is protected by ``gpu_vm->resv``. During eviction all local 224 With VM_BIND, gpu_vmas don't need to be unbound before eviction, 225 since the driver must ensure that the eviction blit or copy will wait 229 section which will make sure all gpu_vmas are rebound. The eviction 231 new exec function may not race with the eviction. 252 At eviction time, the gpu_vm_bos of *all* the gpu_vms an external 257 is a ww_acquire context at hand at eviction time we could grab those 301 And the corresponding shared-object aware eviction woul [all...] |
/linux/Documentation/mm/ |
H A D | multigen_lru.rst | 110 eviction. They form a closed-loop system, i.e., the page reclaim. 123 the latter, when the eviction walks the rmap and finds a young PTE, 130 The eviction consumes old generations. Given an ``lruvec``, it 136 unmapped clean pages, which are the best bet. The eviction sorts a 148 set, an ``lruvec`` is protected from the eviction when its oldest 193 forms a feedback loop between the eviction and the aging. 201 In the eviction path, specifically, in ``lru_gen_look_around()``, if a 264 The aging and the eviction form a producer-consumer model; 268 Within the eviction, the PID controller uses refaults as the feedback
|
H A D | unevictable-lru.rst | 127 may be evicted. To prevent eviction of any such pages, the AS_UNEVICTABLE
|
/linux/Documentation/bpf/ |
H A D | map_hash.rst | 31 is used to select elements for eviction. This internal LRU list is 230 Updating elements in LRU maps may trigger eviction behaviour when the capacity 251 :alt: Diagram outlining the LRU eviction steps taken during map update. 253 LRU hash eviction during map update for ``BPF_MAP_TYPE_LRU_HASH`` and
|
/linux/drivers/gpu/drm/omapdrm/ |
H A D | TODO | 2 . Where should we do eviction (detatch_pages())? We aren't necessarily
|
/linux/drivers/gpu/drm/ttm/tests/ |
H A D | TODO | 5 - Update eviction tests so they use parametrized "from" memory type
|
/linux/Documentation/gpu/rfc/ |
H A D | gpusvm.rst | 40 * Races (especially against concurrent eviction or migrate_to_ram)
|
H A D | i915_vm_bind.rst | 66 the i915 driver. Hence all VA assignment, eviction are not applicable in 122 invalidated vmas (due to eviction and userptr invalidation) etc.
|
/linux/arch/arm/mach-omap2/ |
H A D | Kconfig | 263 bool "Errata 801819: An eviction from L1 data cache might stall indefinitely"
|
/linux/Documentation/filesystems/ |
H A D | vfat.rst | 201 supported but with cache eviction at NFS server, this could 230 last close if there is memory pressure resulting in eviction of
|
H A D | coda.rst | 262 (usually a cache eviction or replacement) and when it finishes
|
/linux/drivers/perf/ |
H A D | xgene_pmu.c | 381 XGENE_PMU_EVENT_ATTR(index-flush-eviction, 0x03), 384 XGENE_PMU_EVENT_ATTR(clean-eviction, 0x06), 385 XGENE_PMU_EVENT_ATTR(dirty-eviction, 0x07),
|
/linux/drivers/fpga/ |
H A D | dfl-fme-perf.c | 588 FME_EVENT_CACHE(eviction, CACHE_EVNT_EVICTIONS);
|
/linux/Documentation/block/ |
H A D | inline-encryption.rst | 486 - The programming and eviction of hardware-wrapped keys happens via
|
/linux/Documentation/admin-guide/device-mapper/ |
H A D | vdo-design.rst | 502 problems. Circumstances causing an eviction include an application
|
/linux/arch/arm64/ |
H A D | Kconfig | 520 Under certain conditions this erratum can cause a clean line eviction
|