0069455bcb
Patch series "Memory allocation profiling", v6. Overview: Low overhead [1] per-callsite memory allocation profiling. Not just for debug kernels, overhead low enough to be deployed in production. Example output: root@moria-kvm:~# sort -rn /proc/allocinfo 127664128 31168 mm/page_ext.c:270 func:alloc_page_ext 56373248 4737 mm/slub.c:2259 func:alloc_slab_page 14880768 3633 mm/readahead.c:247 func:page_cache_ra_unbounded 14417920 3520 mm/mm_init.c:2530 func:alloc_large_system_hash 13377536 234 block/blk-mq.c:3421 func:blk_mq_alloc_rqs 11718656 2861 mm/filemap.c:1919 func:__filemap_get_folio 9192960 2800 kernel/fork.c:307 func:alloc_thread_stack_node 4206592 4 net/netfilter/nf_conntrack_core.c:2567 func:nf_ct_alloc_hashtable 4136960 1010 drivers/staging/ctagmod/ctagmod.c:20 [ctagmod] func:ctagmod_start 3940352 962 mm/memory.c:4214 func:alloc_anon_folio 2894464 22613 fs/kernfs/dir.c:615 func:__kernfs_new_node ... Usage: kconfig options: - CONFIG_MEM_ALLOC_PROFILING - CONFIG_MEM_ALLOC_PROFILING_ENABLED_BY_DEFAULT - CONFIG_MEM_ALLOC_PROFILING_DEBUG adds warnings for allocations that weren't accounted because of a missing annotation sysctl: /proc/sys/vm/mem_profiling Runtime info: /proc/allocinfo Notes: [1]: Overhead To measure the overhead we are comparing the following configurations: (1) Baseline with CONFIG_MEMCG_KMEM=n (2) Disabled by default (CONFIG_MEM_ALLOC_PROFILING=y && CONFIG_MEM_ALLOC_PROFILING_BY_DEFAULT=n) (3) Enabled by default (CONFIG_MEM_ALLOC_PROFILING=y && CONFIG_MEM_ALLOC_PROFILING_BY_DEFAULT=y) (4) Enabled at runtime (CONFIG_MEM_ALLOC_PROFILING=y && CONFIG_MEM_ALLOC_PROFILING_BY_DEFAULT=n && /proc/sys/vm/mem_profiling=1) (5) Baseline with CONFIG_MEMCG_KMEM=y && allocating with __GFP_ACCOUNT (6) Disabled by default (CONFIG_MEM_ALLOC_PROFILING=y && CONFIG_MEM_ALLOC_PROFILING_BY_DEFAULT=n) && CONFIG_MEMCG_KMEM=y (7) Enabled by default (CONFIG_MEM_ALLOC_PROFILING=y && CONFIG_MEM_ALLOC_PROFILING_BY_DEFAULT=y) && CONFIG_MEMCG_KMEM=y Performance overhead: To evaluate performance we implemented an in-kernel test executing multiple get_free_page/free_page and kmalloc/kfree calls with allocation sizes growing from 8 to 240 bytes with CPU frequency set to max and CPU affinity set to a specific CPU to minimize the noise. Below are results from running the test on Ubuntu 22.04.2 LTS with 6.8.0-rc1 kernel on 56 core Intel Xeon: kmalloc pgalloc (1 baseline) 6.764s 16.902s (2 default disabled) 6.793s (+0.43%) 17.007s (+0.62%) (3 default enabled) 7.197s (+6.40%) 23.666s (+40.02%) (4 runtime enabled) 7.405s (+9.48%) 23.901s (+41.41%) (5 memcg) 13.388s (+97.94%) 48.460s (+186.71%) (6 def disabled+memcg) 13.332s (+97.10%) 48.105s (+184.61%) (7 def enabled+memcg) 13.446s (+98.78%) 54.963s (+225.18%) Memory overhead: Kernel size: text data bss dec diff (1) 26515311 18890222 17018880 62424413 (2) 26524728 19423818 16740352 62688898 264485 (3) 26524724 19423818 16740352 62688894 264481 (4) 26524728 19423818 16740352 62688898 264485 (5) 26541782 18964374 16957440 62463596 39183 Memory consumption on a 56 core Intel CPU with 125GB of memory: Code tags: 192 kB PageExts: 262144 kB (256MB) SlabExts: 9876 kB (9.6MB) PcpuExts: 512 kB (0.5MB) Total overhead is 0.2% of total memory. Benchmarks: Hackbench tests run 100 times: hackbench -s 512 -l 200 -g 15 -f 25 -P baseline disabled profiling enabled profiling avg 0.3543 0.3559 (+0.0016) 0.3566 (+0.0023) stdev 0.0137 0.0188 0.0077 hackbench -l 10000 baseline disabled profiling enabled profiling avg 6.4218 6.4306 (+0.0088) 6.5077 (+0.0859) stdev 0.0933 0.0286 0.0489 stress-ng tests: stress-ng --class memory --seq 4 -t 60 stress-ng --class cpu --seq 4 -t 60 Results posted at: https://evilpiepirate.org/~kent/memalloc_prof_v4_stress-ng/ [2] https://lore.kernel.org/all/20240306182440.2003814-1-surenb@google.com/ This patch (of 37): The next patch drops vmalloc.h from a system header in order to fix a circular dependency; this adds it to all the files that were pulling it in implicitly. [kent.overstreet@linux.dev: fix arch/alpha/lib/memcpy.c] Link: https://lkml.kernel.org/r/20240327002152.3339937-1-kent.overstreet@linux.dev [surenb@google.com: fix arch/x86/mm/numa_32.c] Link: https://lkml.kernel.org/r/20240402180933.1663992-1-surenb@google.com [kent.overstreet@linux.dev: a few places were depending on sizes.h] Link: https://lkml.kernel.org/r/20240404034744.1664840-1-kent.overstreet@linux.dev [arnd@arndb.de: fix mm/kasan/hw_tags.c] Link: https://lkml.kernel.org/r/20240404124435.3121534-1-arnd@kernel.org [surenb@google.com: fix arc build] Link: https://lkml.kernel.org/r/20240405225115.431056-1-surenb@google.com Link: https://lkml.kernel.org/r/20240321163705.3067592-1-surenb@google.com Link: https://lkml.kernel.org/r/20240321163705.3067592-2-surenb@google.com Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev> Signed-off-by: Suren Baghdasaryan <surenb@google.com> Signed-off-by: Arnd Bergmann <arnd@arndb.de> Reviewed-by: Pasha Tatashin <pasha.tatashin@soleen.com> Tested-by: Kees Cook <keescook@chromium.org> Cc: Alexander Viro <viro@zeniv.linux.org.uk> Cc: Alex Gaynor <alex.gaynor@gmail.com> Cc: Alice Ryhl <aliceryhl@google.com> Cc: Andreas Hindborg <a.hindborg@samsung.com> Cc: Benno Lossin <benno.lossin@proton.me> Cc: "Björn Roy Baron" <bjorn3_gh@protonmail.com> Cc: Boqun Feng <boqun.feng@gmail.com> Cc: Christoph Lameter <cl@linux.com> Cc: Dennis Zhou <dennis@kernel.org> Cc: Gary Guo <gary@garyguo.net> Cc: Miguel Ojeda <ojeda@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Tejun Heo <tj@kernel.org> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: Wedson Almeida Filho <wedsonaf@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
512 lines
15 KiB
C
512 lines
15 KiB
C
/*
|
|
* Copyright(c) 2011-2016 Intel Corporation. All rights reserved.
|
|
*
|
|
* Permission is hereby granted, free of charge, to any person obtaining a
|
|
* copy of this software and associated documentation files (the "Software"),
|
|
* to deal in the Software without restriction, including without limitation
|
|
* the rights to use, copy, modify, merge, publish, distribute, sublicense,
|
|
* and/or sell copies of the Software, and to permit persons to whom the
|
|
* Software is furnished to do so, subject to the following conditions:
|
|
*
|
|
* The above copyright notice and this permission notice (including the next
|
|
* paragraph) shall be included in all copies or substantial portions of the
|
|
* Software.
|
|
*
|
|
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
|
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
|
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
|
|
* THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
|
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
|
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
|
|
* SOFTWARE.
|
|
*
|
|
* Authors:
|
|
* Eddie Dong <eddie.dong@intel.com>
|
|
* Kevin Tian <kevin.tian@intel.com>
|
|
*
|
|
* Contributors:
|
|
* Ping Gao <ping.a.gao@intel.com>
|
|
* Zhi Wang <zhi.a.wang@intel.com>
|
|
* Bing Niu <bing.niu@intel.com>
|
|
*
|
|
*/
|
|
|
|
#include "i915_drv.h"
|
|
#include "gvt.h"
|
|
#include "i915_pvinfo.h"
|
|
#include <linux/vmalloc.h>
|
|
|
|
void populate_pvinfo_page(struct intel_vgpu *vgpu)
|
|
{
|
|
struct drm_i915_private *i915 = vgpu->gvt->gt->i915;
|
|
/* setup the ballooning information */
|
|
vgpu_vreg64_t(vgpu, vgtif_reg(magic)) = VGT_MAGIC;
|
|
vgpu_vreg_t(vgpu, vgtif_reg(version_major)) = 1;
|
|
vgpu_vreg_t(vgpu, vgtif_reg(version_minor)) = 0;
|
|
vgpu_vreg_t(vgpu, vgtif_reg(display_ready)) = 0;
|
|
vgpu_vreg_t(vgpu, vgtif_reg(vgt_id)) = vgpu->id;
|
|
|
|
vgpu_vreg_t(vgpu, vgtif_reg(vgt_caps)) = VGT_CAPS_FULL_PPGTT;
|
|
vgpu_vreg_t(vgpu, vgtif_reg(vgt_caps)) |= VGT_CAPS_HWSP_EMULATION;
|
|
vgpu_vreg_t(vgpu, vgtif_reg(vgt_caps)) |= VGT_CAPS_HUGE_GTT;
|
|
|
|
vgpu_vreg_t(vgpu, vgtif_reg(avail_rs.mappable_gmadr.base)) =
|
|
vgpu_aperture_gmadr_base(vgpu);
|
|
vgpu_vreg_t(vgpu, vgtif_reg(avail_rs.mappable_gmadr.size)) =
|
|
vgpu_aperture_sz(vgpu);
|
|
vgpu_vreg_t(vgpu, vgtif_reg(avail_rs.nonmappable_gmadr.base)) =
|
|
vgpu_hidden_gmadr_base(vgpu);
|
|
vgpu_vreg_t(vgpu, vgtif_reg(avail_rs.nonmappable_gmadr.size)) =
|
|
vgpu_hidden_sz(vgpu);
|
|
|
|
vgpu_vreg_t(vgpu, vgtif_reg(avail_rs.fence_num)) = vgpu_fence_sz(vgpu);
|
|
|
|
vgpu_vreg_t(vgpu, vgtif_reg(cursor_x_hot)) = UINT_MAX;
|
|
vgpu_vreg_t(vgpu, vgtif_reg(cursor_y_hot)) = UINT_MAX;
|
|
|
|
gvt_dbg_core("Populate PVINFO PAGE for vGPU %d\n", vgpu->id);
|
|
gvt_dbg_core("aperture base [GMADR] 0x%llx size 0x%llx\n",
|
|
vgpu_aperture_gmadr_base(vgpu), vgpu_aperture_sz(vgpu));
|
|
gvt_dbg_core("hidden base [GMADR] 0x%llx size=0x%llx\n",
|
|
vgpu_hidden_gmadr_base(vgpu), vgpu_hidden_sz(vgpu));
|
|
gvt_dbg_core("fence size %d\n", vgpu_fence_sz(vgpu));
|
|
|
|
drm_WARN_ON(&i915->drm, sizeof(struct vgt_if) != VGT_PVINFO_SIZE);
|
|
}
|
|
|
|
/*
|
|
* vGPU type name is defined as GVTg_Vx_y which contains the physical GPU
|
|
* generation type (e.g V4 as BDW server, V5 as SKL server).
|
|
*
|
|
* Depening on the physical SKU resource, we might see vGPU types like
|
|
* GVTg_V4_8, GVTg_V4_4, GVTg_V4_2, etc. We can create different types of
|
|
* vGPU on same physical GPU depending on available resource. Each vGPU
|
|
* type will have a different number of avail_instance to indicate how
|
|
* many vGPU instance can be created for this type.
|
|
*/
|
|
#define VGPU_MAX_WEIGHT 16
|
|
#define VGPU_WEIGHT(vgpu_num) \
|
|
(VGPU_MAX_WEIGHT / (vgpu_num))
|
|
|
|
static const struct intel_vgpu_config intel_vgpu_configs[] = {
|
|
{ MB_TO_BYTES(64), MB_TO_BYTES(384), 4, VGPU_WEIGHT(8), GVT_EDID_1024_768, "8" },
|
|
{ MB_TO_BYTES(128), MB_TO_BYTES(512), 4, VGPU_WEIGHT(4), GVT_EDID_1920_1200, "4" },
|
|
{ MB_TO_BYTES(256), MB_TO_BYTES(1024), 4, VGPU_WEIGHT(2), GVT_EDID_1920_1200, "2" },
|
|
{ MB_TO_BYTES(512), MB_TO_BYTES(2048), 4, VGPU_WEIGHT(1), GVT_EDID_1920_1200, "1" },
|
|
};
|
|
|
|
/**
|
|
* intel_gvt_init_vgpu_types - initialize vGPU type list
|
|
* @gvt : GVT device
|
|
*
|
|
* Initialize vGPU type list based on available resource.
|
|
*
|
|
*/
|
|
int intel_gvt_init_vgpu_types(struct intel_gvt *gvt)
|
|
{
|
|
unsigned int low_avail = gvt_aperture_sz(gvt) - HOST_LOW_GM_SIZE;
|
|
unsigned int high_avail = gvt_hidden_sz(gvt) - HOST_HIGH_GM_SIZE;
|
|
unsigned int num_types = ARRAY_SIZE(intel_vgpu_configs);
|
|
unsigned int i;
|
|
|
|
gvt->types = kcalloc(num_types, sizeof(struct intel_vgpu_type),
|
|
GFP_KERNEL);
|
|
if (!gvt->types)
|
|
return -ENOMEM;
|
|
|
|
gvt->mdev_types = kcalloc(num_types, sizeof(*gvt->mdev_types),
|
|
GFP_KERNEL);
|
|
if (!gvt->mdev_types)
|
|
goto out_free_types;
|
|
|
|
for (i = 0; i < num_types; ++i) {
|
|
const struct intel_vgpu_config *conf = &intel_vgpu_configs[i];
|
|
|
|
if (low_avail / conf->low_mm == 0)
|
|
break;
|
|
if (conf->weight < 1 || conf->weight > VGPU_MAX_WEIGHT)
|
|
goto out_free_mdev_types;
|
|
|
|
sprintf(gvt->types[i].name, "GVTg_V%u_%s",
|
|
GRAPHICS_VER(gvt->gt->i915) == 8 ? 4 : 5, conf->name);
|
|
gvt->types[i].conf = conf;
|
|
|
|
gvt_dbg_core("type[%d]: %s avail %u low %u high %u fence %u weight %u res %s\n",
|
|
i, gvt->types[i].name,
|
|
min(low_avail / conf->low_mm,
|
|
high_avail / conf->high_mm),
|
|
conf->low_mm, conf->high_mm, conf->fence,
|
|
conf->weight, vgpu_edid_str(conf->edid));
|
|
|
|
gvt->mdev_types[i] = &gvt->types[i].type;
|
|
gvt->mdev_types[i]->sysfs_name = gvt->types[i].name;
|
|
}
|
|
|
|
gvt->num_types = i;
|
|
return 0;
|
|
|
|
out_free_mdev_types:
|
|
kfree(gvt->mdev_types);
|
|
out_free_types:
|
|
kfree(gvt->types);
|
|
return -EINVAL;
|
|
}
|
|
|
|
void intel_gvt_clean_vgpu_types(struct intel_gvt *gvt)
|
|
{
|
|
kfree(gvt->mdev_types);
|
|
kfree(gvt->types);
|
|
}
|
|
|
|
/**
|
|
* intel_gvt_activate_vgpu - activate a virtual GPU
|
|
* @vgpu: virtual GPU
|
|
*
|
|
* This function is called when user wants to activate a virtual GPU.
|
|
*
|
|
*/
|
|
void intel_gvt_activate_vgpu(struct intel_vgpu *vgpu)
|
|
{
|
|
set_bit(INTEL_VGPU_STATUS_ACTIVE, vgpu->status);
|
|
}
|
|
|
|
/**
|
|
* intel_gvt_deactivate_vgpu - deactivate a virtual GPU
|
|
* @vgpu: virtual GPU
|
|
*
|
|
* This function is called when user wants to deactivate a virtual GPU.
|
|
* The virtual GPU will be stopped.
|
|
*
|
|
*/
|
|
void intel_gvt_deactivate_vgpu(struct intel_vgpu *vgpu)
|
|
{
|
|
mutex_lock(&vgpu->vgpu_lock);
|
|
|
|
clear_bit(INTEL_VGPU_STATUS_ACTIVE, vgpu->status);
|
|
|
|
if (atomic_read(&vgpu->submission.running_workload_num)) {
|
|
mutex_unlock(&vgpu->vgpu_lock);
|
|
intel_gvt_wait_vgpu_idle(vgpu);
|
|
mutex_lock(&vgpu->vgpu_lock);
|
|
}
|
|
|
|
intel_vgpu_stop_schedule(vgpu);
|
|
|
|
mutex_unlock(&vgpu->vgpu_lock);
|
|
}
|
|
|
|
/**
|
|
* intel_gvt_release_vgpu - release a virtual GPU
|
|
* @vgpu: virtual GPU
|
|
*
|
|
* This function is called when user wants to release a virtual GPU.
|
|
* The virtual GPU will be stopped and all runtime information will be
|
|
* destroyed.
|
|
*
|
|
*/
|
|
void intel_gvt_release_vgpu(struct intel_vgpu *vgpu)
|
|
{
|
|
intel_gvt_deactivate_vgpu(vgpu);
|
|
|
|
mutex_lock(&vgpu->vgpu_lock);
|
|
vgpu->d3_entered = false;
|
|
intel_vgpu_clean_workloads(vgpu, ALL_ENGINES);
|
|
intel_vgpu_dmabuf_cleanup(vgpu);
|
|
mutex_unlock(&vgpu->vgpu_lock);
|
|
}
|
|
|
|
/**
|
|
* intel_gvt_destroy_vgpu - destroy a virtual GPU
|
|
* @vgpu: virtual GPU
|
|
*
|
|
* This function is called when user wants to destroy a virtual GPU.
|
|
*
|
|
*/
|
|
void intel_gvt_destroy_vgpu(struct intel_vgpu *vgpu)
|
|
{
|
|
struct intel_gvt *gvt = vgpu->gvt;
|
|
struct drm_i915_private *i915 = gvt->gt->i915;
|
|
|
|
drm_WARN(&i915->drm, test_bit(INTEL_VGPU_STATUS_ACTIVE, vgpu->status),
|
|
"vGPU is still active!\n");
|
|
|
|
/*
|
|
* remove idr first so later clean can judge if need to stop
|
|
* service if no active vgpu.
|
|
*/
|
|
mutex_lock(&gvt->lock);
|
|
idr_remove(&gvt->vgpu_idr, vgpu->id);
|
|
mutex_unlock(&gvt->lock);
|
|
|
|
mutex_lock(&vgpu->vgpu_lock);
|
|
intel_gvt_debugfs_remove_vgpu(vgpu);
|
|
intel_vgpu_clean_sched_policy(vgpu);
|
|
intel_vgpu_clean_submission(vgpu);
|
|
intel_vgpu_clean_display(vgpu);
|
|
intel_vgpu_clean_opregion(vgpu);
|
|
intel_vgpu_reset_ggtt(vgpu, true);
|
|
intel_vgpu_clean_gtt(vgpu);
|
|
intel_vgpu_detach_regions(vgpu);
|
|
intel_vgpu_free_resource(vgpu);
|
|
intel_vgpu_clean_mmio(vgpu);
|
|
intel_vgpu_dmabuf_cleanup(vgpu);
|
|
mutex_unlock(&vgpu->vgpu_lock);
|
|
}
|
|
|
|
#define IDLE_VGPU_IDR 0
|
|
|
|
/**
|
|
* intel_gvt_create_idle_vgpu - create an idle virtual GPU
|
|
* @gvt: GVT device
|
|
*
|
|
* This function is called when user wants to create an idle virtual GPU.
|
|
*
|
|
* Returns:
|
|
* pointer to intel_vgpu, error pointer if failed.
|
|
*/
|
|
struct intel_vgpu *intel_gvt_create_idle_vgpu(struct intel_gvt *gvt)
|
|
{
|
|
struct intel_vgpu *vgpu;
|
|
enum intel_engine_id i;
|
|
int ret;
|
|
|
|
vgpu = vzalloc(sizeof(*vgpu));
|
|
if (!vgpu)
|
|
return ERR_PTR(-ENOMEM);
|
|
|
|
vgpu->id = IDLE_VGPU_IDR;
|
|
vgpu->gvt = gvt;
|
|
mutex_init(&vgpu->vgpu_lock);
|
|
|
|
for (i = 0; i < I915_NUM_ENGINES; i++)
|
|
INIT_LIST_HEAD(&vgpu->submission.workload_q_head[i]);
|
|
|
|
ret = intel_vgpu_init_sched_policy(vgpu);
|
|
if (ret)
|
|
goto out_free_vgpu;
|
|
|
|
clear_bit(INTEL_VGPU_STATUS_ACTIVE, vgpu->status);
|
|
return vgpu;
|
|
|
|
out_free_vgpu:
|
|
vfree(vgpu);
|
|
return ERR_PTR(ret);
|
|
}
|
|
|
|
/**
|
|
* intel_gvt_destroy_idle_vgpu - destroy an idle virtual GPU
|
|
* @vgpu: virtual GPU
|
|
*
|
|
* This function is called when user wants to destroy an idle virtual GPU.
|
|
*
|
|
*/
|
|
void intel_gvt_destroy_idle_vgpu(struct intel_vgpu *vgpu)
|
|
{
|
|
mutex_lock(&vgpu->vgpu_lock);
|
|
intel_vgpu_clean_sched_policy(vgpu);
|
|
mutex_unlock(&vgpu->vgpu_lock);
|
|
|
|
vfree(vgpu);
|
|
}
|
|
|
|
int intel_gvt_create_vgpu(struct intel_vgpu *vgpu,
|
|
const struct intel_vgpu_config *conf)
|
|
{
|
|
struct intel_gvt *gvt = vgpu->gvt;
|
|
struct drm_i915_private *dev_priv = gvt->gt->i915;
|
|
int ret;
|
|
|
|
gvt_dbg_core("low %u MB high %u MB fence %u\n",
|
|
BYTES_TO_MB(conf->low_mm), BYTES_TO_MB(conf->high_mm),
|
|
conf->fence);
|
|
|
|
mutex_lock(&gvt->lock);
|
|
ret = idr_alloc(&gvt->vgpu_idr, vgpu, IDLE_VGPU_IDR + 1, GVT_MAX_VGPU,
|
|
GFP_KERNEL);
|
|
if (ret < 0)
|
|
goto out_unlock;
|
|
|
|
vgpu->id = ret;
|
|
vgpu->sched_ctl.weight = conf->weight;
|
|
mutex_init(&vgpu->vgpu_lock);
|
|
mutex_init(&vgpu->dmabuf_lock);
|
|
INIT_LIST_HEAD(&vgpu->dmabuf_obj_list_head);
|
|
INIT_RADIX_TREE(&vgpu->page_track_tree, GFP_KERNEL);
|
|
idr_init_base(&vgpu->object_idr, 1);
|
|
intel_vgpu_init_cfg_space(vgpu, 1);
|
|
vgpu->d3_entered = false;
|
|
|
|
ret = intel_vgpu_init_mmio(vgpu);
|
|
if (ret)
|
|
goto out_clean_idr;
|
|
|
|
ret = intel_vgpu_alloc_resource(vgpu, conf);
|
|
if (ret)
|
|
goto out_clean_vgpu_mmio;
|
|
|
|
populate_pvinfo_page(vgpu);
|
|
|
|
ret = intel_vgpu_init_gtt(vgpu);
|
|
if (ret)
|
|
goto out_clean_vgpu_resource;
|
|
|
|
ret = intel_vgpu_init_opregion(vgpu);
|
|
if (ret)
|
|
goto out_clean_gtt;
|
|
|
|
ret = intel_vgpu_init_display(vgpu, conf->edid);
|
|
if (ret)
|
|
goto out_clean_opregion;
|
|
|
|
ret = intel_vgpu_setup_submission(vgpu);
|
|
if (ret)
|
|
goto out_clean_display;
|
|
|
|
ret = intel_vgpu_init_sched_policy(vgpu);
|
|
if (ret)
|
|
goto out_clean_submission;
|
|
|
|
intel_gvt_debugfs_add_vgpu(vgpu);
|
|
|
|
ret = intel_gvt_set_opregion(vgpu);
|
|
if (ret)
|
|
goto out_clean_sched_policy;
|
|
|
|
if (IS_BROADWELL(dev_priv) || IS_BROXTON(dev_priv))
|
|
ret = intel_gvt_set_edid(vgpu, PORT_B);
|
|
else
|
|
ret = intel_gvt_set_edid(vgpu, PORT_D);
|
|
if (ret)
|
|
goto out_clean_sched_policy;
|
|
|
|
intel_gvt_update_reg_whitelist(vgpu);
|
|
mutex_unlock(&gvt->lock);
|
|
return 0;
|
|
|
|
out_clean_sched_policy:
|
|
intel_vgpu_clean_sched_policy(vgpu);
|
|
out_clean_submission:
|
|
intel_vgpu_clean_submission(vgpu);
|
|
out_clean_display:
|
|
intel_vgpu_clean_display(vgpu);
|
|
out_clean_opregion:
|
|
intel_vgpu_clean_opregion(vgpu);
|
|
out_clean_gtt:
|
|
intel_vgpu_clean_gtt(vgpu);
|
|
out_clean_vgpu_resource:
|
|
intel_vgpu_free_resource(vgpu);
|
|
out_clean_vgpu_mmio:
|
|
intel_vgpu_clean_mmio(vgpu);
|
|
out_clean_idr:
|
|
idr_remove(&gvt->vgpu_idr, vgpu->id);
|
|
out_unlock:
|
|
mutex_unlock(&gvt->lock);
|
|
return ret;
|
|
}
|
|
|
|
/**
|
|
* intel_gvt_reset_vgpu_locked - reset a virtual GPU by DMLR or GT reset
|
|
* @vgpu: virtual GPU
|
|
* @dmlr: vGPU Device Model Level Reset or GT Reset
|
|
* @engine_mask: engines to reset for GT reset
|
|
*
|
|
* This function is called when user wants to reset a virtual GPU through
|
|
* device model reset or GT reset. The caller should hold the vgpu lock.
|
|
*
|
|
* vGPU Device Model Level Reset (DMLR) simulates the PCI level reset to reset
|
|
* the whole vGPU to default state as when it is created. This vGPU function
|
|
* is required both for functionary and security concerns.The ultimate goal
|
|
* of vGPU FLR is that reuse a vGPU instance by virtual machines. When we
|
|
* assign a vGPU to a virtual machine we must isse such reset first.
|
|
*
|
|
* Full GT Reset and Per-Engine GT Reset are soft reset flow for GPU engines
|
|
* (Render, Blitter, Video, Video Enhancement). It is defined by GPU Spec.
|
|
* Unlike the FLR, GT reset only reset particular resource of a vGPU per
|
|
* the reset request. Guest driver can issue a GT reset by programming the
|
|
* virtual GDRST register to reset specific virtual GPU engine or all
|
|
* engines.
|
|
*
|
|
* The parameter dev_level is to identify if we will do DMLR or GT reset.
|
|
* The parameter engine_mask is to specific the engines that need to be
|
|
* resetted. If value ALL_ENGINES is given for engine_mask, it means
|
|
* the caller requests a full GT reset that we will reset all virtual
|
|
* GPU engines. For FLR, engine_mask is ignored.
|
|
*/
|
|
void intel_gvt_reset_vgpu_locked(struct intel_vgpu *vgpu, bool dmlr,
|
|
intel_engine_mask_t engine_mask)
|
|
{
|
|
struct intel_gvt *gvt = vgpu->gvt;
|
|
struct intel_gvt_workload_scheduler *scheduler = &gvt->scheduler;
|
|
intel_engine_mask_t resetting_eng = dmlr ? ALL_ENGINES : engine_mask;
|
|
|
|
gvt_dbg_core("------------------------------------------\n");
|
|
gvt_dbg_core("resseting vgpu%d, dmlr %d, engine_mask %08x\n",
|
|
vgpu->id, dmlr, engine_mask);
|
|
|
|
vgpu->resetting_eng = resetting_eng;
|
|
|
|
intel_vgpu_stop_schedule(vgpu);
|
|
/*
|
|
* The current_vgpu will set to NULL after stopping the
|
|
* scheduler when the reset is triggered by current vgpu.
|
|
*/
|
|
if (scheduler->current_vgpu == NULL) {
|
|
mutex_unlock(&vgpu->vgpu_lock);
|
|
intel_gvt_wait_vgpu_idle(vgpu);
|
|
mutex_lock(&vgpu->vgpu_lock);
|
|
}
|
|
|
|
intel_vgpu_reset_submission(vgpu, resetting_eng);
|
|
/* full GPU reset or device model level reset */
|
|
if (engine_mask == ALL_ENGINES || dmlr) {
|
|
intel_vgpu_select_submission_ops(vgpu, ALL_ENGINES, 0);
|
|
if (engine_mask == ALL_ENGINES)
|
|
intel_vgpu_invalidate_ppgtt(vgpu);
|
|
/*fence will not be reset during virtual reset */
|
|
if (dmlr) {
|
|
if(!vgpu->d3_entered) {
|
|
intel_vgpu_invalidate_ppgtt(vgpu);
|
|
intel_vgpu_destroy_all_ppgtt_mm(vgpu);
|
|
}
|
|
intel_vgpu_reset_ggtt(vgpu, true);
|
|
intel_vgpu_reset_resource(vgpu);
|
|
}
|
|
|
|
intel_vgpu_reset_mmio(vgpu, dmlr);
|
|
populate_pvinfo_page(vgpu);
|
|
|
|
if (dmlr) {
|
|
intel_vgpu_reset_display(vgpu);
|
|
intel_vgpu_reset_cfg_space(vgpu);
|
|
/* only reset the failsafe mode when dmlr reset */
|
|
vgpu->failsafe = false;
|
|
/*
|
|
* PCI_D0 is set before dmlr, so reset d3_entered here
|
|
* after done using.
|
|
*/
|
|
if(vgpu->d3_entered)
|
|
vgpu->d3_entered = false;
|
|
else
|
|
vgpu->pv_notified = false;
|
|
}
|
|
}
|
|
|
|
vgpu->resetting_eng = 0;
|
|
gvt_dbg_core("reset vgpu%d done\n", vgpu->id);
|
|
gvt_dbg_core("------------------------------------------\n");
|
|
}
|
|
|
|
/**
|
|
* intel_gvt_reset_vgpu - reset a virtual GPU (Function Level)
|
|
* @vgpu: virtual GPU
|
|
*
|
|
* This function is called when user wants to reset a virtual GPU.
|
|
*
|
|
*/
|
|
void intel_gvt_reset_vgpu(struct intel_vgpu *vgpu)
|
|
{
|
|
mutex_lock(&vgpu->vgpu_lock);
|
|
intel_gvt_reset_vgpu_locked(vgpu, true, 0);
|
|
mutex_unlock(&vgpu->vgpu_lock);
|
|
}
|