DRM GPUVA Manager GPU-VM features
Subsystems (4)
v3 cover DRM GPUVA Manager GPU-VM features
So far the DRM GPUVA manager offers common infrastructure to track GPU VA allocations and mappings, generically connect GPU VA mappings to their backing buffers and perform more complex mapping operations on the GPU VA space.
However, there are more design patterns commonly used by drivers, which can potentially be generalized in order to make the DRM GPUVA manager represent a basic GPU-VM implementation. In this context, this patch series aims at generalizing the following elements.
1) Provide a common dma-resv for GEM objects not being used outside of this GPU-VM.
2) Provide tracking of external GEM objects (GEM objects which are shared with other GPU-VMs).
3) Provide functions to efficiently lock all GEM objects dma-resv the GPU-VM contains mappings of.
4) Provide tracking of evicted GEM objects the GPU-VM contains mappings of, such that validation of evicted GEM objects is accelerated.
5) Provide some convinience functions for common patterns.
The implementation introduces struct drm_gpuvm_bo, which serves as abstraction combining a struct drm_gpuvm and struct drm_gem_object, similar to what amdgpu does with struct amdgpu_bo_vm. While this adds a bit of complexity it improves the efficiency of tracking external and evicted GEM objects.
This patch series also renames struct drm_gpuva_manager to struct drm_gpuvm including corresponding functions. This way the GPUVA manager's structures align better with the documentation of VM_BIND [1] and VM_BIND locking [2]. It also provides a better foundation for the naming of data structures and functions introduced for implementing the features of this patch series.
This patch series is also available at [3].
[1] Documentation/gpu/drm-vm-bind-async.rst
[2] Documentation/gpu/drm-vm-bind-locking.rst
[3] https://gitlab.freedesktop.org/nouvelles/kernel/-/commits/gpuvm-next
Changes in V2:
==============
- rename 'drm_gpuva_manager' -> 'drm_gpuvm' which generally leads to more consistent naming
- properly separate commits (introduce common dma-resv, drm_gpuvm_bo abstraction, etc.)
- remove maple tree for tracking external objects, use a list drm_gpuvm_bos per drm_gpuvm instead
- rework dma-resv locking helpers (Thomas)
- add a locking helper for a given range of the VA space (Christian)
- make the GPUVA manager buildable as module, rather than drm_exec builtin (Christian)
Changes in V3:
==============
- rename missing function and files (Boris)
- warn if vm_obj->obj != obj in drm_gpuva_link() (Boris)
- don't expose drm_gpuvm_bo_destroy() (Boris)
- unlink VM_BO from GEM in drm_gpuvm_bo_destroy() rather than drm_gpuva_unlink() and link within drm_gpuvm_bo_obtain() to keep drm_gpuvm_bo instances unique
- add internal locking to external and evicted object lists to support drivers updating the VA space from within the fence signalling critical path (Boris)
- unlink external objects and evicted objects from the GPUVM's list in drm_gpuvm_bo_destroy()
- add more documentation and fix some kernel doc issues
Danilo Krummrich (7):
drm/gpuvm: rename struct drm_gpuva_manager to struct drm_gpuvm drm/gpuvm: allow building as module
drm/nouveau: uvmm: rename 'umgr' to 'base'
drm/gpuvm: common dma-resv per struct drm_gpuvm drm/gpuvm: add an abstraction for a VM / BO combination drm/gpuvm: generalize dma_resv/extobj handling and GEM validation drm/nouveau: GPUVM dma-resv/extobj handling, GEM validation
drivers/gpu/drm/Kconfig | 7 +
drivers/gpu/drm/Makefile | 2 +-
drivers/gpu/drm/drm_debugfs.c | 16 +-
.../gpu/drm/{drm_gpuva_mgr.c => drm_gpuvm.c} | 1209 ++++++++++++++---
drivers/gpu/drm/nouveau/Kconfig | 1 +
drivers/gpu/drm/nouveau/nouveau_bo.c | 4 +-
drivers/gpu/drm/nouveau/nouveau_debugfs.c | 2 +-
drivers/gpu/drm/nouveau/nouveau_exec.c | 52 +-
drivers/gpu/drm/nouveau/nouveau_exec.h | 4 -
drivers/gpu/drm/nouveau/nouveau_gem.c | 4 +-
drivers/gpu/drm/nouveau/nouveau_sched.h | 4 +-
drivers/gpu/drm/nouveau/nouveau_uvmm.c | 207 +--
drivers/gpu/drm/nouveau/nouveau_uvmm.h | 8 +-
include/drm/drm_debugfs.h | 6 +-
include/drm/drm_gem.h | 32 +-
include/drm/{drm_gpuva_mgr.h => drm_gpuvm.h} | 510 +++++--
16 files changed, 1605 insertions(+), 463 deletions(-)
rename drivers/gpu/drm/{drm_gpuva_mgr.c => drm_gpuvm.c} (51%)
rename include/drm/{drm_gpuva_mgr.h => drm_gpuvm.h} (53%)base-commit: 6bd3d8da51ca1ec97c724016466606aec7739b9f
Signature · 1 line
2.41.0
PATCH v3 1/7 drm/gpuvm: rename struct drm_gpuva_manager to struct drm_gpuvm
Rename struct drm_gpuva_manager to struct drm_gpuvm including corresponding functions. This way the GPUVA manager's structures align very well with the documentation of VM_BIND [1] and VM_BIND locking [2].
It also provides a better foundation for the naming of data structures and functions introduced for implementing a common dma-resv per GPU-VM including tracking of external and evicted objects in subsequent patches.
[1] Documentation/gpu/drm-vm-bind-async.rst
[2] Documentation/gpu/drm-vm-bind-locking.rst
---
drivers/gpu/drm/Makefile | 2 +-
drivers/gpu/drm/drm_debugfs.c | 16 +-
.../gpu/drm/{drm_gpuva_mgr.c => drm_gpuvm.c} | 392 +++++++++---------
drivers/gpu/drm/nouveau/nouveau_exec.c | 2 +-
drivers/gpu/drm/nouveau/nouveau_uvmm.c | 24 +-
drivers/gpu/drm/nouveau/nouveau_uvmm.h | 6 +-
include/drm/drm_debugfs.h | 6 +-
include/drm/{drm_gpuva_mgr.h => drm_gpuvm.h} | 153 ++++---
8 files changed, 300 insertions(+), 301 deletions(-)
rename drivers/gpu/drm/{drm_gpuva_mgr.c => drm_gpuvm.c} (78%)
rename include/drm/{drm_gpuva_mgr.h => drm_gpuvm.h} (78%)drivers/gpu/drm/Makefile +1 -1
@@ -45,7 +45,7 @@ drm-y := \ drm_vblank.o \ drm_vblank_work.o \ drm_vma_manager.o \ - drm_gpuva_mgr.o \ + drm_gpuvm.o \ drm_writeback.o drm-$(CONFIG_DRM_LEGACY) += \ drm_agpsupport.o \
drivers/gpu/drm/drm_debugfs.c +8 -8
@@ -40,7 +40,7 @@#include <drm/drm_file.h> #include <drm/drm_gem.h> #include <drm/drm_managed.h> -#include <drm/drm_gpuva_mgr.h> +#include <drm/drm_gpuvm.h> #include "drm_crtc_internal.h" #include "drm_internal.h"
@@ -187,31 +187,31 @@ static const struct file_operations drm_debugfs_fops = {/** * drm_debugfs_gpuva_info - dump the given DRM GPU VA space * @m: pointer to the &seq_file to write -* @mgr: the &drm_gpuva_manager representing the GPU VA space +* @gpuvm: the &drm_gpuvm representing the GPU VA space * * Dumps the GPU VA mappings of a given DRM GPU VA manager. * * For each DRM GPU VA space drivers should call this function from their * &drm_info_list's show callback. * -* Returns: 0 on success, -ENODEV if the &mgr is not initialized +* Returns: 0 on success, -ENODEV if the &gpuvm is not initialized */ int drm_debugfs_gpuva_info (struct seq_file *m, -struct drm_gpuva_manager *mgr) +struct drm_gpuvm *gpuvm) { -struct drm_gpuva *va, *kva = &mgr->kernel_alloc_node; +struct drm_gpuva *va, *kva = &gpuvm->kernel_alloc_node; -if (!mgr->name) +if (!gpuvm->name)return -ENODEV ;seq_printf (m,"DRM GPU VA space (%s) [0x%016llx;0x%016llx]\n" , - mgr->name, mgr->mm_start, mgr->mm_start + mgr->mm_range); + gpuvm->name, gpuvm->mm_start, gpuvm->mm_start + gpuvm->mm_range);seq_printf (m,"Kernel reserved node [0x%016llx;0x%016llx]\n" , kva->va.addr, kva->va.addr + kva->va.range);seq_puts (m,"\n" );seq_puts (m," VAs | start | range | end | object | object offset\n" );seq_puts (m,"-------------------------------------------------------------------------------------------------------------\n" ); -drm_gpuva_for_each_va (va, mgr) { +drm_gpuvm_for_each_va (va, gpuvm) {if (unlikely (va == kva))continue ;
drivers/gpu/drm/drm_gpuvm.c renamed from drivers/gpu/drm/drm_gpuva_mgr.c +196 -196
1,020 lines not on this page · raw message
drivers/gpu/drm/nouveau/nouveau_exec.c +1 -1
@@ -106,7 +106,7 @@ nouveau_exec_job_submit(struct nouveau_job *job)drm_exec_until_all_locked (exec) {struct drm_gpuva *va; -drm_gpuva_for_each_va (va, &uvmm->umgr) { +drm_gpuvm_for_each_va (va, &uvmm->umgr) {if (unlikely (va == &uvmm->umgr.kernel_alloc_node))continue ;
drivers/gpu/drm/nouveau/nouveau_uvmm.c +12 -12
@@ -444,7 +444,7 @@ op_map_prepare_unwind(struct nouveau_uvma *uvma)static void op_unmap_prepare_unwind (struct drm_gpuva *va) { -drm_gpuva_insert (va->mgr, va); +drm_gpuva_insert (va->vm, va); } static void
@@ -1194,7 +1194,7 @@ nouveau_uvmm_bind_job_submit(struct nouveau_job *job) goto unwind_continue; } - op->ops =drm_gpuva_sm_unmap_ops_create (&uvmm->umgr, + op->ops =drm_gpuvm_sm_unmap_ops_create (&uvmm->umgr, op->va.addr, op->va.range);if (IS_ERR (op->ops)) {
@@ -1240,7 +1240,7 @@ nouveau_uvmm_bind_job_submit(struct nouveau_job *job) } } - op->ops = drm_gpuva_sm_map_ops_create(&uvmm->umgr, + op->ops = drm_gpuvm_sm_map_ops_create(&uvmm->umgr, op->va.addr, op->va.range, op->gem.obj,
@@ -1264,7 +1264,7 @@ nouveau_uvmm_bind_job_submit(struct nouveau_job *job)break ; }case OP_UNMAP : - op->ops =drm_gpuva_sm_unmap_ops_create (&uvmm->umgr, + op->ops =drm_gpuvm_sm_unmap_ops_create (&uvmm->umgr, op->va.addr, op->va.range);if (IS_ERR (op->ops)) {
@@ -1836,11 +1836,11 @@ nouveau_uvmm_init(struct nouveau_uvmm *uvmm, struct nouveau_cli *cli, uvmm->kernel_managed_addr = kernel_managed_addr; uvmm->kernel_managed_size = kernel_managed_size; -drm_gpuva_manager_init (&uvmm->umgr, cli->name, -NOUVEAU_VA_SPACE_START , -NOUVEAU_VA_SPACE_END , - kernel_managed_addr, kernel_managed_size, -NULL ); +drm_gpuvm_init (&uvmm->umgr, cli->name, +NOUVEAU_VA_SPACE_START , +NOUVEAU_VA_SPACE_END , + kernel_managed_addr, kernel_managed_size, +NULL ); ret = nvif_vmm_ctor(&cli->mmu, "uvmm", cli->vmm.vmm.object.oclass, RAW,
@@ -1855,7 +1855,7 @@ nouveau_uvmm_init(struct nouveau_uvmm *uvmm, struct nouveau_cli *cli,return 0 ; out_free_gpuva_mgr: -drm_gpuva_manager_destroy (&uvmm->umgr); +drm_gpuvm_destroy (&uvmm->umgr); out_unlock:mutex_unlock (&cli->mutex);return ret;
@@ -1877,7 +1877,7 @@ nouveau_uvmm_fini(struct nouveau_uvmm *uvmm)wait_event (entity->job.wq,list_empty (&entity->job.list.head));nouveau_uvmm_lock (uvmm); -drm_gpuva_for_each_va_safe (va, next, &uvmm->umgr) { +drm_gpuvm_for_each_va_safe (va, next, &uvmm->umgr) {struct nouveau_uvma *uvma =uvma_from_va (va);struct drm_gem_object *obj = va->gem.obj;
@@ -1910,7 +1910,7 @@ nouveau_uvmm_fini(struct nouveau_uvmm *uvmm)mutex_lock (&cli->mutex);nouveau_vmm_fini (&uvmm->vmm); -drm_gpuva_manager_destroy (&uvmm->umgr); +drm_gpuvm_destroy (&uvmm->umgr);mutex_unlock (&cli->mutex);dma_resv_fini (&uvmm->resv);
drivers/gpu/drm/nouveau/nouveau_uvmm.h +3 -3
@@ -3,13 +3,13 @@ #ifndef __NOUVEAU_UVMM_H__ #define __NOUVEAU_UVMM_H__ -#include <drm/drm_gpuva_mgr.h> +#include <drm/drm_gpuvm.h> #include "nouveau_drv.h" struct nouveau_uvmm { struct nouveau_vmm vmm; - struct drm_gpuva_manager umgr; + struct drm_gpuvm umgr; struct maple_tree region_mt; struct mutex mutex; struct dma_resv resv;
@@ -44,7 +44,7 @@ struct nouveau_uvma {#define uvmm_from_mgr (x) container_of((x), struct nouveau_uvmm, umgr)#define uvma_from_va (x) container_of((x), struct nouveau_uvma, va) -#define to_uvmm (x) uvmm_from_mgr((x)->va.mgr) +#define to_uvmm (x) uvmm_from_mgr((x)->va.vm)struct nouveau_uvmm_bind_job {struct nouveau_job base;
include/drm/drm_debugfs.h +3 -3
@@ -35,7 +35,7 @@#include <linux/types.h> #include <linux/seq_file.h> -#include <drm/drm_gpuva_mgr.h> +#include <drm/drm_gpuvm.h> /** * DRM_DEBUGFS_GPUVA_INFO - &drm_info_list entry to dump a GPU VA space
@@ -152,7 +152,7 @@ void drm_debugfs_add_files(struct drm_device *dev, const struct drm_debugfs_info *files, int count);int drm_debugfs_gpuva_info (struct seq_file *m, -struct drm_gpuva_manager *mgr); +struct drm_gpuvm *gpuvm);#else static inline void drm_debugfs_create_files(const struct drm_info_list *files, int count, struct dentry *root,
@@ -176,7 +176,7 @@ static inline void drm_debugfs_add_files(struct drm_device *dev, {}static inline int drm_debugfs_gpuva_info (struct seq_file *m, -struct drm_gpuva_manager *mgr) +struct drm_gpuvm *gpuvm) {return 0 ; }
include/drm/drm_gpuvm.h renamed from include/drm/drm_gpuva_mgr.h +76 -77
352 lines not on this page · raw message
Signature · 1 line
2.41.0
kernel test robot
build botRe: [PATCH drm-misc-next v3 1/7] drm/gpuvm: rename struct drm_gpuva_manager to struct drm_gpuvm
kernel test robot noticed the following build warnings:
Hi Danilo,
kernel test robot noticed the following build warnings:
[auto build test WARNING on 6bd3d8da51ca1ec97c724016466606aec7739b9f]
url: https://github.com/intel-lab-lkp/linux/commits/Danilo-Krummrich/drm-gpuvm-rename-struct-drm_gpuva_manager-to-struct-drm_gpuvm/20230909-233346 base: 6bd3d8da51ca1ec97c724016466606aec7739b9f patch link: https://lore.kernel.org/r/20230909153125.30032-2-dakr%40redhat.com patch subject: [PATCH drm-misc-next v3 1/7] drm/gpuvm: rename struct drm_gpuva_manager to struct drm_gpuvm config: riscv-defconfig (https://download.01.org/0day-ci/archive/20230910/[email protected]/config) compiler: riscv64-linux-gcc (GCC) 13.2.0 reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20230910/[email protected]/reproduce)
If you fix the issue in a separate patch/commit (i.e. not just a new version of the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <[email protected]>
| Closes: https://lore.kernel.org/oe-kbuild-all/[email protected]/
All warnings (new ones prefixed by >>):
drivers/gpu/drm/drm_gpuvm.c: In function '__drm_gpuvm_sm_map':
>> drivers/gpu/drm/drm_gpuvm.c:1079:39: warning: variable 'prev' set but not used [-Wunused-but-set-variable]
1079 | struct drm_gpuva *va, *next, *prev = NULL;
| ^~~~vim +/prev +1079 drivers/gpu/drm/drm_gpuvm.c
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1072
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1073 static int
5d2000a9816e19 drivers/gpu/drm/drm_gpuvm.c Danilo Krummrich 2023-09-09 1074 __drm_gpuvm_sm_map(struct drm_gpuvm *gpuvm,
5d2000a9816e19 drivers/gpu/drm/drm_gpuvm.c Danilo Krummrich 2023-09-09 1075 const struct drm_gpuvm_ops *ops, void *priv,
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1076 u64 req_addr, u64 req_range,
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1077 struct drm_gem_object *req_obj, u64 req_offset)
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1078 {
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 @1079 struct drm_gpuva *va, *next, *prev = NULL;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1080 u64 req_end = req_addr + req_range;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1081 int ret;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1082
5d2000a9816e19 drivers/gpu/drm/drm_gpuvm.c Danilo Krummrich 2023-09-09 1083 if (unlikely(!drm_gpuva_range_valid(gpuvm, req_addr, req_range)))
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1084 return -EINVAL;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1085
5d2000a9816e19 drivers/gpu/drm/drm_gpuvm.c Danilo Krummrich 2023-09-09 1086 drm_gpuvm_for_each_va_range_safe(va, next, gpuvm, req_addr, req_end) {
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1087 struct drm_gem_object *obj = va->gem.obj;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1088 u64 offset = va->gem.offset;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1089 u64 addr = va->va.addr;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1090 u64 range = va->va.range;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1091 u64 end = addr + range;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1092 bool merge = !!va->gem.obj;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1093
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1094 if (addr == req_addr) {
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1095 merge &= obj == req_obj &&
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1096 offset == req_offset;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1097
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1098 if (end == req_end) {
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1099 ret = op_unmap_cb(ops, priv, va, merge);
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1100 if (ret)
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1101 return ret;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1102 break;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1103 }
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1104
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1105 if (end < req_end) {
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1106 ret = op_unmap_cb(ops, priv, va, merge);
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1107 if (ret)
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1108 return ret;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1109 goto next;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1110 }
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1111
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1112 if (end > req_end) {
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1113 struct drm_gpuva_op_map n = {
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1114 .va.addr = req_end,
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1115 .va.range = range - req_range,
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1116 .gem.obj = obj,
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1117 .gem.offset = offset + req_range,
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1118 };
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1119 struct drm_gpuva_op_unmap u = {
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1120 .va = va,
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1121 .keep = merge,
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1122 };
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1123
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1124 ret = op_remap_cb(ops, priv, NULL, &n, &u);
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1125 if (ret)
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1126 return ret;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1127 break;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1128 }
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1129 } else if (addr < req_addr) {
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1130 u64 ls_range = req_addr - addr;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1131 struct drm_gpuva_op_map p = {
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1132 .va.addr = addr,
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1133 .va.range = ls_range,
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1134 .gem.obj = obj,
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1135 .gem.offset = offset,
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1136 };
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1137 struct drm_gpuva_op_unmap u = { .va = va };
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1138
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1139 merge &= obj == req_obj &&
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1140 offset + ls_range == req_offset;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1141 u.keep = merge;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1142
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1143 if (end == req_end) {
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1144 ret = op_remap_cb(ops, priv, &p, NULL, &u);
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1145 if (ret)
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1146 return ret;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1147 break;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1148 }
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1149
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1150 if (end < req_end) {
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1151 ret = op_remap_cb(ops, priv, &p, NULL, &u);
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1152 if (ret)
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1153 return ret;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1154 goto next;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1155 }
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1156
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1157 if (end > req_end) {
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1158 struct drm_gpuva_op_map n = {
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1159 .va.addr = req_end,
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1160 .va.range = end - req_end,
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1161 .gem.obj = obj,
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1162 .gem.offset = offset + ls_range +
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1163 req_range,
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1164 };
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1165
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1166 ret = op_remap_cb(ops, priv, &p, &n, &u);
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1167 if (ret)
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1168 return ret;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1169 break;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1170 }
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1171 } else if (addr > req_addr) {
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1172 merge &= obj == req_obj &&
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1173 offset == req_offset +
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1174 (addr - req_addr);
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1175
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1176 if (end == req_end) {
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1177 ret = op_unmap_cb(ops, priv, va, merge);
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1178 if (ret)
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1179 return ret;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1180 break;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1181 }
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1182
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1183 if (end < req_end) {
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1184 ret = op_unmap_cb(ops, priv, va, merge);
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1185 if (ret)
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1186 return ret;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1187 goto next;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1188 }
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1189
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1190 if (end > req_end) {
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1191 struct drm_gpuva_op_map n = {
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1192 .va.addr = req_end,
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1193 .va.range = end - req_end,
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1194 .gem.obj = obj,
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1195 .gem.offset = offset + req_end - addr,
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1196 };
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1197 struct drm_gpuva_op_unmap u = {
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1198 .va = va,
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1199 .keep = merge,
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1200 };
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1201
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1202 ret = op_remap_cb(ops, priv, NULL, &n, &u);
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1203 if (ret)
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1204 return ret;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1205 break;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1206 }
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1207 }
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1208 next:
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1209 prev = va;
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1210 }
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1211
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1212 return op_map_cb(ops, priv,
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1213 req_addr, req_range,
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1214 req_obj, req_offset);
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1215 }
e6303f323b1ad9 drivers/gpu/drm/drm_gpuva_mgr.c Danilo Krummrich 2023-07-20 1216 Signature · 2 lines
0-DAY CI Kernel Test Service https://github.com/intel/lkp-tests/wiki
PATCH v3 2/7 drm/gpuvm: allow building as module
In reply to Danilo KrummrichCurrently, the DRM GPUVM does not have any core dependencies preventing a module build.
Also, new features from subsequent patches require helpers (namely drm_exec) which can be built as module.
--- drivers/gpu/drm/Kconfig | 7 +++++++ drivers/gpu/drm/Makefile | 2 +- drivers/gpu/drm/drm_gpuvm.c | 3 +++ drivers/gpu/drm/nouveau/Kconfig | 1 + 4 files changed, 12 insertions(+), 1 deletion(-)
drivers/gpu/drm/Kconfig +7 -0
@@ -216,6 +216,13 @@ config DRM_EXEC help Execution context for command submissions +config DRM_GPUVM + tristate + depends on DRM && DRM_EXEC + help + GPU-VM representation providing helpers to manage a GPUs virtual + address space + config DRM_BUDDY tristate depends on DRM
drivers/gpu/drm/Makefile +1 -1
@@ -45,7 +45,6 @@ drm-y := \ drm_vblank.o \ drm_vblank_work.o \ drm_vma_manager.o \ - drm_gpuvm.o \ drm_writeback.o drm-$(CONFIG_DRM_LEGACY) += \ drm_agpsupport.o \
@@ -81,6 +80,7 @@ obj-$(CONFIG_DRM_PANEL_ORIENTATION_QUIRKS) += drm_panel_orientation_quirks.o # # obj-$(CONFIG_DRM_EXEC) += drm_exec.o +obj-$(CONFIG_DRM_GPUVM) += drm_gpuvm.o obj-$(CONFIG_DRM_BUDDY) += drm_buddy.o
drivers/gpu/drm/drm_gpuvm.c +3 -0
@@ -1723,3 +1723,6 @@ drm_gpuva_ops_free(struct drm_gpuvm *gpuvm,kfree (ops); }EXPORT_SYMBOL_GPL (drm_gpuva_ops_free); + +MODULE_DESCRIPTION ("DRM GPUVM" ); +MODULE_LICENSE ("GPL" );
drivers/gpu/drm/nouveau/Kconfig +1 -0
@@ -11,6 +11,7 @@ config DRM_NOUVEAU select DRM_TTM select DRM_TTM_HELPER select DRM_EXEC + select DRM_GPUVM select DRM_SCHED select I2C select I2C_ALGOBIT
Signature · 1 line
2.41.0
Re: [PATCH drm-misc-next v3 2/7] drm/gpuvm: allow building as module
Am 09.09.23 um 17:31 schrieb Danilo Krummrich:
Quoted 7 lines
Currently, the DRM GPUVM does not have any core dependencies preventing a module build.
Also, new features from subsequent patches require helpers (namely
drm_exec) which can be built as module.Signed-off-by Danilo Krummrich <[email protected]>
I hope that I can get somebody to work on the remaining patches with the end goal of using this in amdgpu as well.
Regards,
Christian.
Quoted 68 lines
--- drivers/gpu/drm/Kconfig | 7 +++++++ drivers/gpu/drm/Makefile | 2 +- drivers/gpu/drm/drm_gpuvm.c | 3 +++ drivers/gpu/drm/nouveau/Kconfig | 1 + 4 files changed, 12 insertions(+), 1 deletion(-)drivers/gpu/drm/Kconfig +7
-0@@ -216,6 +216,13 @@ config DRM_EXEC help Execution context for command submissions +config DRM_GPUVM + tristate + depends on DRM && DRM_EXEC + help + GPU-VM representation providing helpers to manage a GPUs virtual + address space + config DRM_BUDDY tristate depends on DRMdrivers/gpu/drm/Makefile +1
-1@@ -45,7 +45,6 @@ drm-y := \ drm_vblank.o \ drm_vblank_work.o \ drm_vma_manager.o \ - drm_gpuvm.o \ drm_writeback.o drm-$(CONFIG_DRM_LEGACY) += \ drm_agpsupport.o \@@ -81,6 +80,7 @@ obj-$(CONFIG_DRM_PANEL_ORIENTATION_QUIRKS) += drm_panel_orientation_quirks.o # # obj-$(CONFIG_DRM_EXEC) += drm_exec.o +obj-$(CONFIG_DRM_GPUVM) += drm_gpuvm.o obj-$(CONFIG_DRM_BUDDY) += drm_buddy.odrivers/gpu/drm/drm_gpuvm.c +3
-0@@ -1723,3 +1723,6 @@ drm_gpuva_ops_free(struct drm_gpuvm *gpuvm,kfree (ops); }EXPORT_SYMBOL_GPL (drm_gpuva_ops_free); + +MODULE_DESCRIPTION ("DRM GPUVM" ); +MODULE_LICENSE ("GPL" );drivers/gpu/drm/nouveau/Kconfig +1
-0@@ -11,6 +11,7 @@ config DRM_NOUVEAU select DRM_TTM select DRM_TTM_HELPER select DRM_EXEC + select DRM_GPUVM select DRM_SCHED select I2C select I2C_ALGOBIT
PATCH v3 3/7 drm/nouveau: uvmm: rename 'umgr' to 'base'
In reply to Danilo KrummrichRename struct drm_gpuvm within struct nouveau_uvmm from 'umgr' to base.
--- drivers/gpu/drm/nouveau/nouveau_debugfs.c | 2 +- drivers/gpu/drm/nouveau/nouveau_exec.c | 4 +-- drivers/gpu/drm/nouveau/nouveau_uvmm.c | 32 +++++++++++------------ drivers/gpu/drm/nouveau/nouveau_uvmm.h | 6 ++--- 4 files changed, 22 insertions(+), 22 deletions(-)
drivers/gpu/drm/nouveau/nouveau_debugfs.c +1 -1
@@ -231,7 +231,7 @@ nouveau_debugfs_gpuva(struct seq_file *m, void *data)continue ;nouveau_uvmm_lock (uvmm); -drm_debugfs_gpuva_info (m, &uvmm->umgr); +drm_debugfs_gpuva_info (m, &uvmm->base);seq_puts (m,"\n" );nouveau_debugfs_gpuva_regions (m, uvmm);nouveau_uvmm_unlock (uvmm);
drivers/gpu/drm/nouveau/nouveau_exec.c +2 -2
@@ -106,8 +106,8 @@ nouveau_exec_job_submit(struct nouveau_job *job)drm_exec_until_all_locked (exec) {struct drm_gpuva *va; -drm_gpuvm_for_each_va (va, &uvmm->umgr) { -if (unlikely (va == &uvmm->umgr.kernel_alloc_node)) +drm_gpuvm_for_each_va (va, &uvmm->base) { +if (unlikely (va == &uvmm->base.kernel_alloc_node))continue ; ret =drm_exec_prepare_obj (exec, va->gem.obj,1 );
drivers/gpu/drm/nouveau/nouveau_uvmm.c +16 -16
@@ -329,7 +329,7 @@ nouveau_uvma_region_create(struct nouveau_uvmm *uvmm,struct nouveau_uvma_region *reg;int ret; -if (!drm_gpuva_interval_empty (&uvmm->umgr, addr, range)) +if (!drm_gpuva_interval_empty (&uvmm->base, addr, range))return -ENOSPC ; ret =nouveau_uvma_region_alloc (®);
@@ -384,7 +384,7 @@ nouveau_uvma_region_empty(struct nouveau_uvma_region *reg) {struct nouveau_uvmm *uvmm = reg->uvmm; -return drm_gpuva_interval_empty (&uvmm->umgr, +return drm_gpuva_interval_empty (&uvmm->base, reg->va.addr, reg->va.range); }
@@ -589,7 +589,7 @@ op_map_prepare(struct nouveau_uvmm *uvmm, uvma->region = args->region; uvma->kind = args->kind; -drm_gpuva_map (&uvmm->umgr, &uvma->va, op); +drm_gpuva_map (&uvmm->base, &uvma->va, op);/* Keep a reference until this uvma is destroyed. */ nouveau_uvma_gem_get (uvma);
@@ -1194,7 +1194,7 @@ nouveau_uvmm_bind_job_submit(struct nouveau_job *job) goto unwind_continue; } - op->ops =drm_gpuvm_sm_unmap_ops_create (&uvmm->umgr, + op->ops =drm_gpuvm_sm_unmap_ops_create (&uvmm->base, op->va.addr, op->va.range);if (IS_ERR (op->ops)) {
@@ -1205,7 +1205,7 @@ nouveau_uvmm_bind_job_submit(struct nouveau_job *job) ret =nouveau_uvmm_sm_unmap_prepare (uvmm, &op->new, op->ops);if (ret) { -drm_gpuva_ops_free (&uvmm->umgr, op->ops); +drm_gpuva_ops_free (&uvmm->base, op->ops); op->ops =NULL ; op->reg =NULL ; goto unwind_continue;
@@ -1240,7 +1240,7 @@ nouveau_uvmm_bind_job_submit(struct nouveau_job *job) } } - op->ops = drm_gpuvm_sm_map_ops_create(&uvmm->umgr, + op->ops = drm_gpuvm_sm_map_ops_create(&uvmm->base, op->va.addr, op->va.range, op->gem.obj,
@@ -1256,7 +1256,7 @@ nouveau_uvmm_bind_job_submit(struct nouveau_job *job) op->va.range, op->flags & 0xff);if (ret) { -drm_gpuva_ops_free (&uvmm->umgr, op->ops); +drm_gpuva_ops_free (&uvmm->base, op->ops); op->ops =NULL ; goto unwind_continue; }
@@ -1264,7 +1264,7 @@ nouveau_uvmm_bind_job_submit(struct nouveau_job *job)break ; }case OP_UNMAP : - op->ops =drm_gpuvm_sm_unmap_ops_create (&uvmm->umgr, + op->ops =drm_gpuvm_sm_unmap_ops_create (&uvmm->base, op->va.addr, op->va.range);if (IS_ERR (op->ops)) {
@@ -1275,7 +1275,7 @@ nouveau_uvmm_bind_job_submit(struct nouveau_job *job) ret =nouveau_uvmm_sm_unmap_prepare (uvmm, &op->new, op->ops);if (ret) { -drm_gpuva_ops_free (&uvmm->umgr, op->ops); +drm_gpuva_ops_free (&uvmm->base, op->ops); op->ops =NULL ; goto unwind_continue; }
@@ -1404,7 +1404,7 @@ nouveau_uvmm_bind_job_submit(struct nouveau_job *job)break ; } -drm_gpuva_ops_free (&uvmm->umgr, op->ops); +drm_gpuva_ops_free (&uvmm->base, op->ops); op->ops =NULL ; op->reg =NULL ; }
@@ -1509,7 +1509,7 @@ nouveau_uvmm_bind_job_free_work_fn(struct work_struct *work) }if (!IS_ERR_OR_NULL (op->ops)) -drm_gpuva_ops_free (&uvmm->umgr, op->ops); +drm_gpuva_ops_free (&uvmm->base, op->ops);if (obj)drm_gem_object_put (obj);
@@ -1836,7 +1836,7 @@ nouveau_uvmm_init(struct nouveau_uvmm *uvmm, struct nouveau_cli *cli, uvmm->kernel_managed_addr = kernel_managed_addr; uvmm->kernel_managed_size = kernel_managed_size; - drm_gpuvm_init(&uvmm->umgr, cli->name, + drm_gpuvm_init(&uvmm->base, cli->name, NOUVEAU_VA_SPACE_START, NOUVEAU_VA_SPACE_END, kernel_managed_addr, kernel_managed_size,
@@ -1855,7 +1855,7 @@ nouveau_uvmm_init(struct nouveau_uvmm *uvmm, struct nouveau_cli *cli,return 0 ; out_free_gpuva_mgr: -drm_gpuvm_destroy (&uvmm->umgr); +drm_gpuvm_destroy (&uvmm->base); out_unlock:mutex_unlock (&cli->mutex);return ret;
@@ -1877,11 +1877,11 @@ nouveau_uvmm_fini(struct nouveau_uvmm *uvmm)wait_event (entity->job.wq,list_empty (&entity->job.list.head));nouveau_uvmm_lock (uvmm); -drm_gpuvm_for_each_va_safe (va, next, &uvmm->umgr) { +drm_gpuvm_for_each_va_safe (va, next, &uvmm->base) {struct nouveau_uvma *uvma =uvma_from_va (va);struct drm_gem_object *obj = va->gem.obj; -if (unlikely (va == &uvmm->umgr.kernel_alloc_node)) +if (unlikely (va == &uvmm->base.kernel_alloc_node))continue ;drm_gpuva_remove (va);
@@ -1910,7 +1910,7 @@ nouveau_uvmm_fini(struct nouveau_uvmm *uvmm)mutex_lock (&cli->mutex);nouveau_vmm_fini (&uvmm->vmm); -drm_gpuvm_destroy (&uvmm->umgr); +drm_gpuvm_destroy (&uvmm->base);mutex_unlock (&cli->mutex);dma_resv_fini (&uvmm->resv);
drivers/gpu/drm/nouveau/nouveau_uvmm.h +3 -3
@@ -8,8 +8,8 @@#include "nouveau_drv.h" struct nouveau_uvmm { +struct drm_gpuvm base;struct nouveau_vmm vmm; -struct drm_gpuvm umgr;struct maple_tree region_mt;struct mutex mutex;struct dma_resv resv;
@@ -41,10 +41,10 @@ struct nouveau_uvma {u8 kind; }; -#define uvmm_from_mgr (x) container_of((x), struct nouveau_uvmm, umgr) +#define uvmm_from_gpuvm (x) container_of((x), struct nouveau_uvmm, base)#define uvma_from_va (x) container_of((x), struct nouveau_uvma, va) -#define to_uvmm (x) uvmm_from_mgr((x)->va.vm) +#define to_uvmm (x) uvmm_from_gpuvm((x)->va.vm)struct nouveau_uvmm_bind_job {struct nouveau_job base;
Signature · 1 line
2.41.0
PATCH v3 4/7 drm/gpuvm: common dma-resv per struct drm_gpuvm
In reply to Danilo KrummrichProvide a common dma-resv for GEM objects not being used outside of this GPU-VM. This is used in a subsequent patch to generalize dma-resv, external and evicted object handling and GEM validation.
--- drivers/gpu/drm/drm_gpuvm.c | 10 ++++++++-- drivers/gpu/drm/nouveau/nouveau_uvmm.c | 2 +- include/drm/drm_gpuvm.h | 15 ++++++++++++++- 3 files changed, 23 insertions(+), 4 deletions(-)
drivers/gpu/drm/drm_gpuvm.c +8 -2
@@ -655,6 +655,7 @@ drm_gpuva_range_valid(struct drm_gpuvm *gpuvm,/** * drm_gpuvm_init() - initialize a &drm_gpuvm * @gpuvm: pointer to the &drm_gpuvm to initialize +* @drm: the drivers &drm_device * @name: the name of the GPU VA space * @start_offset: the start offset of the GPU VA space * @range: the size of the GPU VA space
@@ -668,7 +669,7 @@ drm_gpuva_range_valid(struct drm_gpuvm *gpuvm,* &name is expected to be managed by the surrounding driver structures. */ void -drm_gpuvm_init(struct drm_gpuvm *gpuvm, +drm_gpuvm_init(struct drm_gpuvm *gpuvm, struct drm_device *drm, const char *name, u64 start_offset, u64 range, u64 reserve_offset, u64 reserve_range,
@@ -694,6 +695,9 @@ drm_gpuvm_init(struct drm_gpuvm *gpuvm, reserve_range))) __drm_gpuva_insert(gpuvm, &gpuvm->kernel_alloc_node); } + +drm_gem_private_object_init (drm, &gpuvm->d_obj,0 ); + gpuvm->resv = gpuvm->d_obj.resv; }EXPORT_SYMBOL_GPL (drm_gpuvm_init);
@@ -713,7 +717,9 @@ drm_gpuvm_destroy(struct drm_gpuvm *gpuvm)__drm_gpuva_remove (&gpuvm->kernel_alloc_node);WARN (!RB_EMPTY_ROOT (&gpuvm->rb.tree.rb_root), -"GPUVA tree is not empty, potentially leaking memory." ); +"GPUVA tree is not empty, potentially leaking memory.\n" ); + +drm_gem_private_object_fini (&gpuvm->d_obj); }EXPORT_SYMBOL_GPL (drm_gpuvm_destroy);
drivers/gpu/drm/nouveau/nouveau_uvmm.c +1 -1
@@ -1836,7 +1836,7 @@ nouveau_uvmm_init(struct nouveau_uvmm *uvmm, struct nouveau_cli *cli, uvmm->kernel_managed_addr = kernel_managed_addr; uvmm->kernel_managed_size = kernel_managed_size; - drm_gpuvm_init(&uvmm->base, cli->name, + drm_gpuvm_init(&uvmm->base, cli->drm->dev, cli->name, NOUVEAU_VA_SPACE_START, NOUVEAU_VA_SPACE_END, kernel_managed_addr, kernel_managed_size,
include/drm/drm_gpuvm.h +14 -1
@@ -240,9 +240,22 @@ struct drm_gpuvm {* @ops: &drm_gpuvm_ops providing the split/merge steps to drivers */ const struct drm_gpuvm_ops *ops; + +/** +* @d_obj: Dummy GEM object; used internally to pass the GPU VMs +* dma-resv to &drm_exec. +*/ +struct drm_gem_object d_obj; + +/** +* @resv: the &dma_resv for &drm_gem_objects mapped in this GPU VA +* space +*/ +struct dma_resv *resv; }; -void drm_gpuvm_init (struct drm_gpuvm *gpuvm,const char *name, +void drm_gpuvm_init (struct drm_gpuvm *gpuvm,struct drm_device *drm, +const char *name,u64 start_offset,u64 range,u64 reserve_offset,u64 reserve_range,const struct drm_gpuvm_ops *ops);
Signature · 1 line
2.41.0
Re: [PATCH drm-misc-next v3 4/7] drm/gpuvm: common dma-resv per struct drm_gpuvm
Quoted 6 lines · On Sat, 9 Sep 2023 17:31:11 +0200 Danilo Krummrich <[email protected]> wrote:
include/drm/drm_gpuvm.h +2
-0@@ -240,9 +240,22 @@ struct drm_gpuvm {* @ops: &drm_gpuvm_ops providing the split/merge steps to drivers */ const struct drm_gpuvm_ops *ops; + +/**
include/drm/drm_gpuvm.h struct drm_gpuvm { · line 254
Quoted 10 lines
include/drm/drm_gpuvm.h +10
-0+* @d_obj: Dummy GEM object; used internally to pass the GPU VMs +* dma-resv to &drm_exec. +*/ +struct drm_gem_object d_obj; + +/** +* @resv: the &dma_resv for &drm_gem_objects mapped in this GPU VA +* space +*/ +struct dma_resv *resv;
Hm, I'd be tempted to drop this field and add a drm_gpuvm_resv() helper returning vm->d_obj.resv;
Quoted 1 line
(quoted diff) +0
-0};
Re: [PATCH drm-misc-next v3 4/7] drm/gpuvm: common dma-resv per struct drm_gpuvm
Quoted 14 lines · Boris Brezillon
Quoted 11 lines · On Sat, 9 Sep 2023 17:31:11 +0200 Danilo Krummrich <[email protected]> wrote:
include/drm/drm_gpuvm.h +7
-0@@ -240,9 +240,22 @@ struct drm_gpuvm {* @ops: &drm_gpuvm_ops providing the split/merge steps to drivers */ const struct drm_gpuvm_ops *ops; + +/** +* @d_obj: Dummy GEM object; used internally to pass the GPU VMs +* dma-resv to &drm_exec. +*/ +struct drm_gem_object d_obj; +
Quoted 8 lines
include/drm/drm_gpuvm.h struct drm_gpuvm { · line 254
Quoted 5 lines
include/drm/drm_gpuvm.h +5
-0+/** +* @resv: the &dma_resv for &drm_gem_objects mapped in this GPU VA +* space +*/ +struct dma_resv *resv;Hm, I'd be tempted to drop this field and add a
drm_gpuvm_resv()helper returningvm->d_obj.resv;
Makes sense, will do that for V4.
Quoted 3 lines
Quoted 2 lines
(quoted diff) +0
-0};
PATCH v3 5/7 drm/gpuvm: add an abstraction for a VM / BO combination
In reply to Danilo KrummrichThis patch adds an abstraction layer between the drm_gpuva mappings of a particular drm_gem_object and this GEM object itself. The abstraction represents a combination of a drm_gem_object and drm_gpuvm. The drm_gem_object holds a list of drm_gpuvm_bo structures (the structure representing this abstraction), while each drm_gpuvm_bo contains list of mappings of this GEM object.
This has multiple advantages:
1) We can use the drm_gpuvm_bo structure to attach it to various lists of the drm_gpuvm. This is useful for tracking external and evicted objects per VM, which is introduced in subsequent patches.
2) Finding mappings of a certain drm_gem_object mapped in a certain drm_gpuvm becomes much cheaper.
3) Drivers can derive and extend the structure to easily represent driver specific states of a BO for a certain GPUVM.
The idea of this abstraction was taken from amdgpu, hence the credit for this idea goes to the developers of amdgpu.
--- drivers/gpu/drm/drm_gpuvm.c | 304 ++++++++++++++++++++++--- drivers/gpu/drm/nouveau/nouveau_uvmm.c | 68 ++++-- include/drm/drm_gem.h | 32 +-- include/drm/drm_gpuvm.h | 149 +++++++++++- 4 files changed, 478 insertions(+), 75 deletions(-)
drivers/gpu/drm/drm_gpuvm.c +269 -35
@@ -61,6 +61,18 @@* contained within struct drm_gpuva already. Hence, for inserting &drm_gpuva * entries from within dma-fence signalling critical sections it is enough to * pre-allocate the &drm_gpuva structures. +* +* In order to connect a struct drm_gpuva its backing &drm_gem_object each +* &drm_gem_object maintains a list of &drm_gpuvm_bo structures, and each +* &drm_gpuvm_bo contains a list of &&drm_gpuva structures. +* +* A &drm_gpuvm_bo is an abstraction that represents a combination of a +* &drm_gpuvm and a &drm_gem_object. Every such combination should be unique. +* This is ensured by the API through drm_gpuvm_bo_obtain() and +* drm_gpuvm_bo_obtain_prealloc() which first look into the corresponding +* &drm_gem_object list of &drm_gpuvm_bos for an existing instance of this +* particular combination. If not existent a new instance is created and linked +* to the &drm_gem_object. */ /**
@@ -393,14 +405,21 @@* split / merge or prefetch. * * The GPU VA manager also does not take care of the locking of the backing -* &drm_gem_object buffers GPU VA lists by itself; drivers are responsible to -* enforce mutual exclusion using either the GEMs dma_resv lock or alternatively -* a driver specific external lock. For the latter see also -* drm_gem_gpuva_set_lock(). +* &drm_gem_object buffers GPU VA lists and &drm_gpuvm_bo abstractions by +* itself; drivers are responsible to enforce mutual exclusion using either the +* GEMs dma_resv lock or alternatively a driver specific external lock. For the +* latter see also drm_gem_gpuva_set_lock(). * * However, the GPU VA manager contains lockdep checks to ensure callers of its * API hold the corresponding lock whenever the &drm_gem_objects GPU VA list is -* accessed by functions such as drm_gpuva_link() or drm_gpuva_unlink(). +* accessed by functions such as drm_gpuva_link() or drm_gpuva_unlink(), but +* also drm_gpuvm_bo_obtain() and drm_gpuvm_bo_put(). +* +* The latter is required since on creation and destruction of a &drm_gpuvm_bo +* the &drm_gpuvm_bo is attached / removed from the &drm_gem_objects gpuva list. +* Subsequent calls to drm_gpuvm_bo_obtain() for the same &drm_gpuvm and +* &drm_gem_object must be able to observe previous creations and destructions +* of &drm_gpuvm_bos in order to keep instances unique. */ /**
@@ -430,6 +449,7 @@* { * struct drm_gpuva_ops *ops; * struct drm_gpuva_op *op +* struct drm_gpuvm_bo *vm_bo; * * driver_lock_va_space(); * ops = drm_gpuvm_sm_map_ops_create(gpuvm, addr, range,
@@ -437,6 +457,10 @@* if (IS_ERR(ops)) * return PTR_ERR(ops); * +* vm_bo = drm_gpuvm_bo_obtain(gpuvm, obj); +* if (IS_ERR(vm_bo)) +* return PTR_ERR(vm_bo); +* * drm_gpuva_for_each_op(op, ops) { * struct drm_gpuva *va; *
@@ -449,7 +473,7 @@* * driver_vm_map(); * drm_gpuva_map(gpuvm, va, &op->map); -* drm_gpuva_link(va); +* drm_gpuva_link(va, vm_bo); * * break; * case DRM_GPUVA_OP_REMAP: {
@@ -476,11 +500,11 @@* driver_vm_remap(); * drm_gpuva_remap(prev, next, &op->remap); * -* drm_gpuva_unlink(va); * if (prev) -* drm_gpuva_link(prev); +* drm_gpuva_link(prev, va->vm_bo); * if (next) -* drm_gpuva_link(next); +* drm_gpuva_link(next, va->vm_bo); +* drm_gpuva_unlink(va); * * break; * }
@@ -496,6 +520,7 @@* break; * } * } +* drm_gpuvm_bo_put(vm_bo); * driver_unlock_va_space(); * * return 0;
@@ -505,6 +530,7 @@* * struct driver_context { * struct drm_gpuvm *gpuvm; +* struct drm_gpuvm_bo *vm_bo; * struct drm_gpuva *new_va; * struct drm_gpuva *prev_va; * struct drm_gpuva *next_va;
@@ -525,6 +551,7 @@* struct drm_gem_object *obj, u64 offset) * { * struct driver_context ctx; +* struct drm_gpuvm_bo *vm_bo; * struct drm_gpuva_ops *ops; * struct drm_gpuva_op *op; * int ret = 0;
@@ -534,16 +561,23 @@* ctx.new_va = kzalloc(sizeof(*ctx.new_va), GFP_KERNEL); * ctx.prev_va = kzalloc(sizeof(*ctx.prev_va), GFP_KERNEL); * ctx.next_va = kzalloc(sizeof(*ctx.next_va), GFP_KERNEL); -* if (!ctx.new_va || !ctx.prev_va || !ctx.next_va) { +* ctx.vm_bo = drm_gpuvm_bo_create(gpuvm, obj); +* if (!ctx.new_va || !ctx.prev_va || !ctx.next_va || !vm_bo) { * ret = -ENOMEM; * goto out; * } * +* // Typically protected with a driver specific GEM gpuva lock +* // used in the fence signaling path for drm_gpuva_link() and +* // drm_gpuva_unlink(), hence pre-allocate. +* ctx.vm_bo = drm_gpuvm_bo_obtain_prealloc(ctx.vm_bo); +* * driver_lock_va_space(); * ret = drm_gpuvm_sm_map(gpuvm, &ctx, addr, range, obj, offset); * driver_unlock_va_space(); * * out: +* drm_gpuvm_bo_put(ctx.vm_bo); * kfree(ctx.new_va); * kfree(ctx.prev_va); * kfree(ctx.next_va);
@@ -556,7 +590,7 @@* * drm_gpuva_map(ctx->vm, ctx->new_va, &op->map); * -* drm_gpuva_link(ctx->new_va); +* drm_gpuva_link(ctx->new_va, ctx->vm_bo); * * // prevent the new GPUVA from being freed in * // driver_mapping_create()
@@ -568,22 +602,23 @@* int driver_gpuva_remap(struct drm_gpuva_op *op, void *__ctx) * { * struct driver_context *ctx = __ctx; +* struct drm_gpuva *va = op->remap.unmap->va; * * drm_gpuva_remap(ctx->prev_va, ctx->next_va, &op->remap); * -* drm_gpuva_unlink(op->remap.unmap->va); -* kfree(op->remap.unmap->va); -* * if (op->remap.prev) { -* drm_gpuva_link(ctx->prev_va); +* drm_gpuva_link(ctx->prev_va, va->vm_bo); * ctx->prev_va = NULL; * } * * if (op->remap.next) { -* drm_gpuva_link(ctx->next_va); +* drm_gpuva_link(ctx->next_va, va->vm_bo); * ctx->next_va = NULL; * } * +* drm_gpuva_unlink(va); +* kfree(va); +* * return 0; * } *
@@ -723,6 +758,186 @@ drm_gpuvm_destroy(struct drm_gpuvm *gpuvm) }EXPORT_SYMBOL_GPL (drm_gpuvm_destroy); +/** +* drm_gpuvm_bo_create() - create a new instance of struct drm_gpuvm_bo +* @gpuvm: The &drm_gpuvm the @obj is mapped in. +* @obj: The &drm_gem_object being mapped in the @gpuvm. +* +* If provided by the driver, this function uses the &drm_gpuvm_ops +* vm_bo_alloc() callback to allocate. +* +* Returns: a pointer to the &drm_gpuvm_bo on success, NULL on failure +*/ +struct drm_gpuvm_bo * +drm_gpuvm_bo_create (struct drm_gpuvm *gpuvm, +struct drm_gem_object *obj) +{ +const struct drm_gpuvm_ops *ops = gpuvm->ops; +struct drm_gpuvm_bo *vm_bo; + +if (ops && ops->vm_bo_alloc) + vm_bo = ops->vm_bo_alloc (); +else + vm_bo =kzalloc (sizeof (*vm_bo),GFP_KERNEL ); + +if (unlikely (!vm_bo)) +return NULL ; + + vm_bo->vm = gpuvm; + vm_bo->obj = obj; + +kref_init (&vm_bo->kref); +INIT_LIST_HEAD (&vm_bo->list.gpuva); +INIT_LIST_HEAD (&vm_bo->list.entry.gem); + +drm_gem_object_get (obj); + +return vm_bo; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_create); + +static void +drm_gpuvm_bo_destroy (struct kref *kref) +{ + struct drm_gpuvm_bo *vm_bo = container_of(kref, struct drm_gpuvm_bo, + kref); +struct drm_gpuvm *gpuvm = vm_bo->vm; +const struct drm_gpuvm_ops *ops = gpuvm->ops; +struct drm_gem_object *obj = vm_bo->obj; + +drm_gem_gpuva_assert_lock_held (vm_bo->obj); + +list_del (&vm_bo->list.entry.gem); + +drm_gem_object_put (obj); + +if (ops && ops->vm_bo_free) + ops->vm_bo_free (vm_bo); +else +kfree (vm_bo); +} + +/** +* drm_gpuvm_bo_put() - drop a struct drm_gpuvm_bo reference +* @vm_bo: the &drm_gpuvm_bo to release the reference of +* +* This releases a reference to @vm_bo. +*/ +void +drm_gpuvm_bo_put (struct drm_gpuvm_bo *vm_bo) +{ +if (vm_bo) +kref_put (&vm_bo->kref, drm_gpuvm_bo_destroy); +} +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_put); + +static struct drm_gpuvm_bo * +__drm_gpuvm_bo_find (struct drm_gpuvm *gpuvm, +struct drm_gem_object *obj) +{ +struct drm_gpuvm_bo *vm_bo; + +drm_gem_gpuva_assert_lock_held (obj); + +drm_gem_for_each_gpuvm_bo (vm_bo, obj) +if (vm_bo->vm == gpuvm) +return vm_bo; + +return NULL ; +} + +/** +* drm_gpuvm_bo_find() - find the &drm_gpuvm_bo for the given +* &drm_gpuvm and &drm_gem_object +* @gpuvm: The &drm_gpuvm the @obj is mapped in. +* @obj: The &drm_gem_object being mapped in the @gpuvm. +* +* Find the &drm_gpuvm_bo representing the combination of the given +* &drm_gpuvm and &drm_gem_object. If found, increases the reference +* count of the &drm_gpuvm_bo accordingly. +* +* Returns: a pointer to the &drm_gpuvm_bo on success, NULL on failure +*/ +struct drm_gpuvm_bo * +drm_gpuvm_bo_find (struct drm_gpuvm *gpuvm, +struct drm_gem_object *obj) +{ +struct drm_gpuvm_bo *vm_bo =__drm_gpuvm_bo_find (gpuvm, obj); + +return vm_bo ?drm_gpuvm_bo_get (vm_bo) :NULL ; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_find); + +/** +* drm_gpuvm_bo_obtain() - obtains and instance of the &drm_gpuvm_bo for the +* given &drm_gpuvm and &drm_gem_object +* @gpuvm: The &drm_gpuvm the @obj is mapped in. +* @obj: The &drm_gem_object being mapped in the @gpuvm. +* +* Find the &drm_gpuvm_bo representing the combination of the given +* &drm_gpuvm and &drm_gem_object. If found, increases the reference +* count of the &drm_gpuvm_bo accordingly. If not found, allocates a new +* &drm_gpuvm_bo. +* +* A new &drm_gpuvm_bo is added to the GEMs gpuva list. +* +* Returns: a pointer to the &drm_gpuvm_bo on success, an ERR_PTR on failure +*/ +struct drm_gpuvm_bo * +drm_gpuvm_bo_obtain (struct drm_gpuvm *gpuvm, +struct drm_gem_object *obj) +{ +struct drm_gpuvm_bo *vm_bo; + + vm_bo =drm_gpuvm_bo_find (gpuvm, obj); +if (vm_bo) +return vm_bo; + + vm_bo =drm_gpuvm_bo_create (gpuvm, obj); +if (!vm_bo) +return ERR_PTR (-ENOMEM ); + +list_add_tail (&vm_bo->list.entry.gem, &obj->gpuva.list); + +return vm_bo; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_obtain); + +/** +* drm_gpuvm_bo_obtain_prealloc() - obtains and instance of the &drm_gpuvm_bo +* for the given &drm_gpuvm and &drm_gem_object +* @__vm_bo: A pre-allocated struct drm_gpuvm_bo. +* +* Find the &drm_gpuvm_bo representing the combination of the given +* &drm_gpuvm and &drm_gem_object. If found, increases the reference +* count of the found &drm_gpuvm_bo accordingly, while the @__vm_bo reference +* count is decreased. If not found @__vm_bo is returned without further +* increase of the reference count. +* +* A new &drm_gpuvm_bo is added to the GEMs gpuva list. +* +* Returns: a pointer to the found &drm_gpuvm_bo or @__vm_bo if no existing +* &drm_gpuvm_bo was found +*/ +struct drm_gpuvm_bo * +drm_gpuvm_bo_obtain_prealloc (struct drm_gpuvm_bo *__vm_bo) +{ +struct drm_gpuvm *gpuvm = __vm_bo->vm; +struct drm_gem_object *obj = __vm_bo->obj; +struct drm_gpuvm_bo *vm_bo; + + vm_bo =drm_gpuvm_bo_find (gpuvm, obj); +if (vm_bo) { +drm_gpuvm_bo_put (__vm_bo); +return vm_bo; + } + +list_add_tail (&__vm_bo->list.entry.gem, &obj->gpuva.list); + +return __vm_bo; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_obtain_prealloc); +static int __drm_gpuva_insert (struct drm_gpuvm *gpuvm,struct drm_gpuva *va)
@@ -812,24 +1027,33 @@ EXPORT_SYMBOL_GPL(drm_gpuva_remove);/** * drm_gpuva_link() - link a &drm_gpuva * @va: the &drm_gpuva to link +* @vm_bo: the &drm_gpuvm_bo to add the &drm_gpuva to * -* This adds the given &va to the GPU VA list of the &drm_gem_object it is -* associated with. +* This adds the given &va to the GPU VA list of the &drm_gpuvm_bo and the +* &drm_gpuvm_bo to the &drm_gem_object it is associated with. +* +* For every &drm_gpuva entry added to the &drm_gpuvm_bo an additional +* reference of the latter is taken. * * This function expects the caller to protect the GEM's GPUVA list against -* concurrent access using the GEMs dma_resv lock. +* concurrent access using either the GEMs dma_resv lock or a driver specific +* lock set through drm_gem_gpuva_set_lock(). */ void -drm_gpuva_link (struct drm_gpuva *va) +drm_gpuva_link (struct drm_gpuva *va,struct drm_gpuvm_bo *vm_bo) {struct drm_gem_object *obj = va->gem.obj;if (unlikely (!obj))return ; +WARN_ON (obj != vm_bo->obj);drm_gem_gpuva_assert_lock_held (obj); -list_add_tail (&va->gem.entry, &obj->gpuva.list); +drm_gpuvm_bo_get (vm_bo); + va->vm_bo = vm_bo; + +list_add_tail (&va->gem.entry, &vm_bo->list.gpuva); }EXPORT_SYMBOL_GPL (drm_gpuva_link);
@@ -840,13 +1064,22 @@ EXPORT_SYMBOL_GPL(drm_gpuva_link);* This removes the given &va from the GPU VA list of the &drm_gem_object it is * associated with. * +* This removes the given &va from the GPU VA list of the &drm_gpuvm_bo and +* the &drm_gpuvm_bo from the &drm_gem_object it is associated with in case +* this call unlinks the last &drm_gpuva from the &drm_gpuvm_bo. +* +* For every &drm_gpuva entry removed from the &drm_gpuvm_bo a reference of +* the latter is dropped. +* * This function expects the caller to protect the GEM's GPUVA list against -* concurrent access using the GEMs dma_resv lock. +* concurrent access using either the GEMs dma_resv lock or a driver specific +* lock set through drm_gem_gpuva_set_lock(). */ void drm_gpuva_unlink (struct drm_gpuva *va) {struct drm_gem_object *obj = va->gem.obj; +struct drm_gpuvm_bo *vm_bo = va->vm_bo;if (unlikely (!obj))return ;
@@ -854,6 +1087,9 @@ drm_gpuva_unlink(struct drm_gpuva *va)drm_gem_gpuva_assert_lock_held (obj);list_del_init (&va->gem.entry); + + va->vm_bo =NULL ; +drm_gpuvm_bo_put (vm_bo); }EXPORT_SYMBOL_GPL (drm_gpuva_unlink);
@@ -998,10 +1234,10 @@ drm_gpuva_remap(struct drm_gpuva *prev,struct drm_gpuva *next, struct drm_gpuva_op_remap *op) { - struct drm_gpuva *curr = op->unmap->va; -struct drm_gpuvm *gpuvm = curr->vm; + struct drm_gpuva *va = op->unmap->va; +struct drm_gpuvm *gpuvm = va->vm; -drm_gpuva_remove (curr); +drm_gpuva_remove (va);if (op->prev) {drm_gpuva_init_from_op (prev, op->prev);
@@ -1645,9 +1881,8 @@ drm_gpuvm_prefetch_ops_create(struct drm_gpuvm *gpuvm,EXPORT_SYMBOL_GPL (drm_gpuvm_prefetch_ops_create);/** -* drm_gpuvm_gem_unmap_ops_create() - creates the &drm_gpuva_ops to unmap a GEM -* @gpuvm: the &drm_gpuvm representing the GPU VA space -* @obj: the &drm_gem_object to unmap +* drm_gpuvm_bo_unmap_ops_create() - creates the &drm_gpuva_ops to unmap a GEM +* @vm_bo: the &drm_gpuvm_bo abstraction * * This function creates a list of operations to perform unmapping for every * GPUVA attached to a GEM.
@@ -1664,15 +1899,14 @@ EXPORT_SYMBOL_GPL(drm_gpuvm_prefetch_ops_create);* Returns: a pointer to the &drm_gpuva_ops on success, an ERR_PTR on failure */ struct drm_gpuva_ops * -drm_gpuvm_gem_unmap_ops_create (struct drm_gpuvm *gpuvm, -struct drm_gem_object *obj) +drm_gpuvm_bo_unmap_ops_create (struct drm_gpuvm_bo *vm_bo) {struct drm_gpuva_ops *ops;struct drm_gpuva_op *op;struct drm_gpuva *va;int ret; -drm_gem_gpuva_assert_lock_held (obj); +drm_gem_gpuva_assert_lock_held (vm_bo->obj); ops =kzalloc (sizeof (*ops),GFP_KERNEL );if (!ops)
@@ -1680,8 +1914,8 @@ drm_gpuvm_gem_unmap_ops_create(struct drm_gpuvm *gpuvm,INIT_LIST_HEAD (&ops->list); -drm_gem_for_each_gpuva (va, obj) { - op =gpuva_op_alloc (gpuvm); +drm_gpuvm_bo_for_each_va (va, vm_bo) { + op =gpuva_op_alloc (vm_bo->vm);if (!op) { ret = -ENOMEM ; goto err_free_ops;
@@ -1695,10 +1929,10 @@ drm_gpuvm_gem_unmap_ops_create(struct drm_gpuvm *gpuvm,return ops; err_free_ops: -drm_gpuva_ops_free (gpuvm, ops); +drm_gpuva_ops_free (vm_bo->vm, ops);return ERR_PTR (ret); } -EXPORT_SYMBOL_GPL (drm_gpuvm_gem_unmap_ops_create); +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_unmap_ops_create);/** * drm_gpuva_ops_free() - free the given &drm_gpuva_ops
drivers/gpu/drm/nouveau/nouveau_uvmm.c +48 -20
143 lines not on this page · raw message
include/drm/drm_gem.h +16 -16
@@ -571,7 +571,7 @@ int drm_gem_evict(struct drm_gem_object *obj);* drm_gem_gpuva_init() - initialize the gpuva list of a GEM object * @obj: the &drm_gem_object * -* This initializes the &drm_gem_object's &drm_gpuva list. +* This initializes the &drm_gem_object's &drm_gpuvm_bo list. * * Calling this function is only necessary for drivers intending to support the * &drm_driver_feature DRIVER_GEM_GPUVA.
@@ -584,28 +584,28 @@ static inline void drm_gem_gpuva_init(struct drm_gem_object *obj) }/** -* drm_gem_for_each_gpuva() - iternator to walk over a list of gpuvas -* @entry__: &drm_gpuva structure to assign to in each iteration step -* @obj__: the &drm_gem_object the &drm_gpuvas to walk are associated with +* drm_gem_for_each_gpuvm_bo() - iterator to walk over a list of &drm_gpuvm_bo +* @entry__: &drm_gpuvm_bo structure to assign to in each iteration step +* @obj__: the &drm_gem_object the &drm_gpuvm_bo to walk are associated with * -* This iterator walks over all &drm_gpuva structures associated with the -* &drm_gpuva_manager. +* This iterator walks over all &drm_gpuvm_bo structures associated with the +* &drm_gem_object. */ -#define drm_gem_for_each_gpuva (entry__, obj__) \ - list_for_each_entry(entry__, &(obj__)->gpuva.list, gem.entry) +#define drm_gem_for_each_gpuvm_bo (entry__, obj__) \ + list_for_each_entry(entry__, &(obj__)->gpuva.list, list.entry.gem)/** -* drm_gem_for_each_gpuva_safe() - iternator to safely walk over a list of -* gpuvas -* @entry__: &drm_gpuva structure to assign to in each iteration step -* @next__: &next &drm_gpuva to store the next step -* @obj__: the &drm_gem_object the &drm_gpuvas to walk are associated with +* drm_gem_for_each_gpuvm_bo_safe() - iterator to safely walk over a list of +* &drm_gpuvm_bo +* @entry__: &drm_gpuvm_bostructure to assign to in each iteration step +* @next__: &next &drm_gpuvm_bo to store the next step +* @obj__: the &drm_gem_object the &drm_gpuvm_bo to walk are associated with * -* This iterator walks over all &drm_gpuva structures associated with the +* This iterator walks over all &drm_gpuvm_bo structures associated with the * &drm_gem_object. It is implemented with list_for_each_entry_safe(), hence * it is save against removal of elements. */ -#define drm_gem_for_each_gpuva_safe (entry__, next__, obj__) \ - list_for_each_entry_safe(entry__, next__, &(obj__)->gpuva.list, gem.entry) +#define drm_gem_for_each_gpuvm_bo_safe (entry__, next__, obj__) \ + list_for_each_entry_safe(entry__, next__, &(obj__)->gpuva.list, list.entry.gem)#endif /* __DRM_GEM_H__ */
include/drm/drm_gpuvm.h +145 -4
191 lines not on this page · raw message
Signature · 1 line
2.41.0
Thomas Hellström
reviewerRe: [PATCH drm-misc-next v3 5/7] drm/gpuvm: add an abstraction for a VM / BO combination
Hi, Danilo
Quoted 16 lines · Danilo Krummrich
This patch adds an abstraction layer between the
drm_gpuvamappings of a particulardrm_gem_objectand this GEM object itself. The abstraction represents a combination of adrm_gem_objectanddrm_gpuvm. Thedrm_gem_objectholds a list ofdrm_gpuvm_bostructures (the structure representing this abstraction), while eachdrm_gpuvm_bocontains list of mappings of this GEM object.This has multiple advantages:
1) We can use the
drm_gpuvm_bostructure to attach it to various lists of thedrm_gpuvm. This is useful for tracking external and evicted objects per VM, which is introduced in subsequent patches.2) Finding mappings of a certain
drm_gem_objectmapped in a certaindrm_gpuvmbecomes much cheaper.
Quoted 8 lines
3) Drivers can derive and extend the structure to easily represent driver specific states of a BO for a certain GPUVM.
The idea of this abstraction was taken from amdgpu, hence the credit for this idea goes to the developers of amdgpu.
Cc Christian König <[email protected]>Signed-off-by Danilo Krummrich <[email protected]>
Did you consider having the drivers embed the struct drm_gpuvm_bo in their own bo definition? I figure that would mean using the gem bo's refcounting and providing a helper to call from the driver's bo release. Looks like that could potentially save a lot of code? Or is there something that won't work with that approach?
Thanks,
Thomas
Re: [PATCH drm-misc-next v3 5/7] drm/gpuvm: add an abstraction for a VM / BO combination
Hi Thomas,
Quoted 22 lines · Thomas Hellström
Hi, Danilo
Quoted 19 lines · Danilo Krummrich
This patch adds an abstraction layer between the
drm_gpuvamappings of a particulardrm_gem_objectand this GEM object itself. The abstraction represents a combination of adrm_gem_objectanddrm_gpuvm. Thedrm_gem_objectholds a list ofdrm_gpuvm_bostructures (the structure representing this abstraction), while eachdrm_gpuvm_bocontains list of mappings of this GEM object.This has multiple advantages:
1) We can use the
drm_gpuvm_bostructure to attach it to various lists of thedrm_gpuvm. This is useful for tracking external and evicted objects per VM, which is introduced in subsequent patches.2) Finding mappings of a certain
drm_gem_objectmapped in a certaindrm_gpuvmbecomes much cheaper.3) Drivers can derive and extend the structure to easily represent driver specific states of a BO for a certain GPUVM.
Quoted 7 lines
Quoted 5 lines
The idea of this abstraction was taken from amdgpu, hence the credit for this idea goes to the developers of amdgpu.
Cc Christian König <[email protected]>Signed-off-by Danilo Krummrich <[email protected]>Did you consider having the drivers embed the struct
drm_gpuvm_boin their own bo definition? I figure that would mean using the gem bo's refcounting and providing a helper to call from the driver's bo release. Looks like that could potentially save a lot of code? Or is there something that won't work with that approach?
There are drm_gpuvm_ops::vm_bo_alloc and drm_gpuvm_ops::vm_bo_free callback for drivers to register for that purpose.
- Danilo
Quoted 6 lines
Thanks,
Thomas
Thomas Hellström
reviewerRe: [PATCH drm-misc-next v3 5/7] drm/gpuvm: add an abstraction for a VM / BO combination
Quoted 28 lines · Danilo Krummrich
Hi Thomas,
Quoted 25 lines · Thomas Hellström
Hi, Danilo
Quoted 22 lines · Danilo Krummrich
This patch adds an abstraction layer between the
drm_gpuvamappings of a particulardrm_gem_objectand this GEM object itself. The abstraction represents a combination of adrm_gem_objectanddrm_gpuvm. Thedrm_gem_objectholds a list ofdrm_gpuvm_bostructures (the structure representing this abstraction), while eachdrm_gpuvm_bocontains list of mappings of this GEM object.This has multiple advantages:
1) We can use the
drm_gpuvm_bostructure to attach it to various lists of thedrm_gpuvm. This is useful for tracking external and evicted objects per VM, which is introduced in subsequent patches.2) Finding mappings of a certain
drm_gem_objectmapped in a certaindrm_gpuvmbecomes much cheaper.3) Drivers can derive and extend the structure to easily represent driver specific states of a BO for a certain GPUVM.
The idea of this abstraction was taken from amdgpu, hence the credit for this idea goes to the developers of amdgpu.
Quoted 6 lines
Quoted 4 lines
Quoted 2 lines
Cc Christian König <[email protected]>Signed-off-by Danilo Krummrich <[email protected]>Did you consider having the drivers embed the struct
drm_gpuvm_boin their own bo definition? I figure that would mean using the gem bo's refcounting and providing a helper to call from the driver's bo release. Looks like that could potentially save a lot of code? Or is there something that won't work with that approach?There are
drm_gpuvm_ops::vm_bo_allocanddrm_gpuvm_ops::vm_bo_freecallback for drivers to register for that purpose.
Ah OK. Thanks, I'll take a deeper look.
/Thomas
Quoted 10 lines
- Danilo
Quoted 6 lines
Thanks,
Thomas
Thomas Hellström
reviewerRe: [PATCH drm-misc-next v3 5/7] drm/gpuvm: add an abstraction for a VM / BO combination
In reply to Danilo KrummrichHi, Danilo
Quoted 28 lines · Danilo Krummrich
Hi Thomas,
Quoted 25 lines · Thomas Hellström
Hi, Danilo
Quoted 22 lines · Danilo Krummrich
This patch adds an abstraction layer between the
drm_gpuvamappings of a particulardrm_gem_objectand this GEM object itself. The abstraction represents a combination of adrm_gem_objectanddrm_gpuvm. Thedrm_gem_objectholds a list ofdrm_gpuvm_bostructures (the structure representing this abstraction), while eachdrm_gpuvm_bocontains list of mappings of this GEM object.This has multiple advantages:
1) We can use the
drm_gpuvm_bostructure to attach it to various lists of thedrm_gpuvm. This is useful for tracking external and evicted objects per VM, which is introduced in subsequent patches.2) Finding mappings of a certain
drm_gem_objectmapped in a certaindrm_gpuvmbecomes much cheaper.3) Drivers can derive and extend the structure to easily represent driver specific states of a BO for a certain GPUVM.
The idea of this abstraction was taken from amdgpu, hence the credit for this idea goes to the developers of amdgpu.
Quoted 8 lines
Quoted 4 lines
Quoted 2 lines
Cc Christian König <[email protected]>Signed-off-by Danilo Krummrich <[email protected]>Did you consider having the drivers embed the struct
drm_gpuvm_boin their own bo definition? I figure that would mean using the gem bo's refcounting and providing a helper to call from the driver's bo release. Looks like that could potentially save a lot of code? Or is there something that won't work with that approach?There are
drm_gpuvm_ops::vm_bo_allocanddrm_gpuvm_ops::vm_bo_freecallback for drivers to register for that purpose.- Danilo
Now after looking a bit deeper, I think actually the question could be rephrased as, why don't we just use the
struct drm_gem_object::gpuva struct as the drm_gpuvm_bo in the spirit of keeping things simple? Drivers would then just embed it in their bo subclass and we'd avoid unnecessary fields in the struct drm_gem_object for drivers that don't do VM_BIND yet.
Sure, this won't be per bo and per vm, but it'd really only make a slight difference where we have multiple VMAs per bo, where per-vm per-bo state either needs to be duplicated or attached to a single vma (as in the case of the external bo list).
To me that looks like a substantial amount of less code / complexity?
/Thomas
Quoted 8 lines
Quoted 7 lines
Thanks,
Thomas
Re: [PATCH drm-misc-next v3 5/7] drm/gpuvm: add an abstraction for a VM / BO combination
Quoted 45 lines · Thomas Hellström
Hi, Danilo
Quoted 42 lines · Danilo Krummrich
Hi Thomas,
Quoted 27 lines · Thomas Hellström
Hi, Danilo
Quoted 24 lines · Danilo Krummrich
This patch adds an abstraction layer between the
drm_gpuvamappings of a particulardrm_gem_objectand this GEM object itself. The abstraction represents a combination of adrm_gem_objectanddrm_gpuvm. Thedrm_gem_objectholds a list ofdrm_gpuvm_bostructures (the structure representing this abstraction), while eachdrm_gpuvm_bocontains list of
mappings of this GEM object.This has multiple advantages:
1) We can use the
drm_gpuvm_bostructure to attach it to various lists of thedrm_gpuvm. This is useful for tracking external and evicted objects per VM, which is introduced in subsequent patches.2) Finding mappings of a certain
drm_gem_objectmapped in a certaindrm_gpuvmbecomes much cheaper.3) Drivers can derive and extend the structure to easily represent driver specific states of a BO for a certain GPUVM.
The idea of this abstraction was taken from amdgpu, hence the credit for
this idea goes to the developers of amdgpu.Quoted 8 lines
Quoted 2 lines
Cc Christian König <[email protected]>Signed-off-by Danilo Krummrich <[email protected]>Did you consider having the drivers embed the struct
drm_gpuvm_boin their own bo definition? I figure that would mean using the gem bo's refcounting and providing a helper to call from the driver's bo release. Looks like that could potentially save a lot of code? Or is there something that won't work with that approach?There are
drm_gpuvm_ops::vm_bo_allocanddrm_gpuvm_ops::vm_bo_freecallback for drivers to register for that purpose.
Quoted 8 lines
Quoted 1 line
- Danilo
Now after looking a bit deeper, I think actually the question could be rephrased as, why don't we just use the
structdrm_gem_object::gpuvastruct as thedrm_gpuvm_boin the spirit of keeping things simple? Drivers would then just embed it in their bo subclass and we'd avoid unnecessary fields in the structdrm_gem_objectfor drivers that don't doVM_BINDyet.
struct drm_gem_object::gpuva is just a container containing a list in order to (currently) attach drm_gpuva structs to it and with this patch attach drm_gpuvm_bo structs (combination of BO + VM) to it. Doing the above basically means "leave everything as it is, but move the list_head of drm_gpuvs per GEM to the driver specific BO structure". Having a common connection between GEM objects and drm_gpuva structs was one of the goals of the initial GPUVA manager patch series however.
Quoted 5 lines
Sure, this won't be per bo and per vm, but it'd really only make a slight difference where we have multiple VMAs per bo, where per-vm per-bo state either needs to be duplicated or attached to a single vma (as in the case of the external bo list).
Correct, one implication is that we don't get a per VM and BO abstraction, and hence are left with a list of all drm_gpuva structs having the same backing BO, regardless of the VM.
For amdgpu this was always a concern. Now that we want to keep track of external and evicted objects it's going to be a concern for most drivers I guess. Because the only structure we could use for tracking external and evicted objects we are left with (without having a VM_BO abstraction) is struct drm_gpuva. But this structure isn't unique and we need to consider cases where userspace just allocates rather huge BOs and creates tons of mappings from it. Running the full list of drm_gpuva structs (with even the ones from other VMs included) for adding an external or evicted object isn't very efficient. Not to mention that the maintenance when the mapping we've (randomly) picked as an entry for the external/evicted object list is unmapped, but there are still mappings left in the VM with the same backing BO.
Now, a way to get rid of the VM_BO abstraction would be to use maple trees instead, since then we can store drm_gem_object structs directly for each VM. However, Xe had concerns about using maple trees and preferred lists, plus having maple trees wouldn't get rid of the concerns of amdgpu not having a VM_BO abstraction for cases with tons of VMs and tons of mappings per BO. Hence, having a VM_BO abstraction enabling us to track external/evicted objects with lists seems to satisfy everyone's needs.
- Danilo
Quoted 15 lines
To me that looks like a substantial amount of less code / complexity?
/Thomas
Quoted 8 lines
Quoted 7 lines
Thanks,
Thomas
Thomas Hellström
reviewerRe: [PATCH drm-misc-next v3 5/7] drm/gpuvm: add an abstraction for a VM / BO combination
Quoted 68 lines · Danilo Krummrich
Quoted 43 lines · Thomas Hellström
Hi, Danilo
Quoted 40 lines · Danilo Krummrich
Hi Thomas,
Quoted 27 lines · Thomas Hellström
Hi, Danilo
Quoted 24 lines · Danilo Krummrich
This patch adds an abstraction layer between the
drm_gpuvamappings of a particulardrm_gem_objectand this GEM object itself. The abstraction represents a combination of adrm_gem_objectanddrm_gpuvm. Thedrm_gem_objectholds a list ofdrm_gpuvm_bostructures (the structure representing this abstraction), while eachdrm_gpuvm_bocontains list of
mappings of this GEM object.This has multiple advantages:
1) We can use the
drm_gpuvm_bostructure to attach it to various lists of thedrm_gpuvm. This is useful for tracking external and evicted objects per VM, which is introduced in subsequent patches.2) Finding mappings of a certain
drm_gem_objectmapped in a certaindrm_gpuvmbecomes much cheaper.3) Drivers can derive and extend the structure to easily represent driver specific states of a BO for a certain GPUVM.
The idea of this abstraction was taken from amdgpu, hence the credit for
this idea goes to the developers of amdgpu.Quoted 7 lines
Quoted 2 lines
Cc Christian König <[email protected]>Signed-off-by Danilo Krummrich <[email protected]>Did you consider having the drivers embed the struct
drm_gpuvm_boin their own bo definition? I figure that would mean using the gem bo's refcounting and providing a helper to call from the driver's bo release. Looks like that could potentially save a lot of code? Or is there something that won't work with that approach?There are
drm_gpuvm_ops::vm_bo_allocanddrm_gpuvm_ops::vm_bo_freecallback for drivers to register for that purpose.Quoted 7 lines
Quoted 1 line
- Danilo
Now after looking a bit deeper, I think actually the question could be rephrased as, why don't we just use the
structdrm_gem_object::gpuvastruct as thedrm_gpuvm_boin the spirit of keeping things simple? Drivers would then just embed it in their bo subclass and we'd avoid unnecessary fields in the structdrm_gem_objectfor drivers that don't doVM_BINDyet.struct
drm_gem_object::gpuvais just a container containing a list in order to (currently) attachdrm_gpuvastructs to it and with this patch attachdrm_gpuvm_bostructs (combination of BO + VM) to it. Doing the above basically means "leave everything as it is, but move thelist_headofdrm_gpuvsper GEM to the driver specific BO structure". Having a common connection between GEM objects anddrm_gpuvastructs was one of the goals of the initial GPUVA manager patch series however.Quoted 4 lines
Sure, this won't be per bo and per vm, but it'd really only make a slight difference where we have multiple VMAs per bo, where per-vm per-bo state either needs to be duplicated or attached to a single vma (as in the case of the external bo list).
Correct, one implication is that we don't get a per VM and BO abstraction, and hence are left with a list of all
drm_gpuvastructs having the same backing BO, regardless of the VM.
Quoted 11 lines
For amdgpu this was always a concern. Now that we want to keep track of external and evicted objects it's going to be a concern for most drivers I guess. Because the only structure we could use for tracking external and evicted objects we are left with (without having a
VM_BOabstraction) is structdrm_gpuva. But this structure isn't unique and we need to consider cases where userspace just allocates rather huge BOs and creates tons of mappings from it. Running the full list ofdrm_gpuvastructs (with even the ones from other VMs included) for adding an external or evicted object isn't very efficient. Not to mention that the maintenance when the mapping we've (randomly) picked as an entry for the external/evicted object list is unmapped, but there are still mappings left in the VM with the same backing BO.
For the evicted object it's not much of an issue; we maintain a list of vmas needing rebinding for each VM rather than objects evicted, so there is no or very little additional overhead there. The extobj list is indeed a problem if many VMAs are bound to the same bo. Not that the code snippets are complicated, but the list traversals would be excessive.
Quoted 8 lines
Now, a way to get rid of the
VM_BOabstraction would be to use maple trees instead, since then we can storedrm_gem_objectstructs directly for each VM. However, Xe had concerns about using maple trees and preferred lists, plus having maple trees wouldn't get rid of the concerns of amdgpu not having aVM_BOabstraction for cases with tons of VMs and tons of mappings per BO. Hence, having aVM_BOabstraction enabling us to track external/evicted objects with lists seems to satisfy everyone's needs.
Indeed this is a tradeoff between a simple implementation that is OK for situations with not many VMs nor VMAs per bo vs a more complex implementation that optimizes for the opposite case.
So if this latter is a case we need to optimize for at this point then I guess it's the way to go.
(I'm in the process of adapting the xe driver to this, so I just wanted to bring up areas where the implementations differ quite a lot and make sure options are discussed).
Thanks,
Thomas
Quoted 13 lines
- Danilo
Quoted 10 lines
To me that looks like a substantial amount of less code / complexity?
/Thomas
Quoted 5 lines
Quoted 5 lines
Thanks,
Thomas
Re: [PATCH drm-misc-next v3 5/7] drm/gpuvm: add an abstraction for a VM / BO combination
Quoted 99 lines · Thomas Hellström
Quoted 68 lines · Danilo Krummrich
Quoted 43 lines · Thomas Hellström
Hi, Danilo
Quoted 40 lines · Danilo Krummrich
Hi Thomas,
Quoted 27 lines · Thomas Hellström
Hi, Danilo
Quoted 24 lines · Danilo Krummrich
This patch adds an abstraction layer between the
drm_gpuvamappings of a particulardrm_gem_objectand this GEM object itself. The abstraction represents a combination of adrm_gem_objectanddrm_gpuvm. Thedrm_gem_objectholds a list ofdrm_gpuvm_bostructures (the structure representing this abstraction), while eachdrm_gpuvm_bocontains list of
mappings of this GEM object.This has multiple advantages:
1) We can use the
drm_gpuvm_bostructure to attach it to various lists of thedrm_gpuvm. This is useful for tracking external and evicted objects per VM, which is introduced in subsequent patches.2) Finding mappings of a certain
drm_gem_objectmapped in a certaindrm_gpuvmbecomes much cheaper.3) Drivers can derive and extend the structure to easily represent driver specific states of a BO for a certain GPUVM.
The idea of this abstraction was taken from amdgpu, hence the credit for
this idea goes to the developers of amdgpu.Quoted 7 lines
Quoted 2 lines
Cc Christian König <[email protected]>Signed-off-by Danilo Krummrich <[email protected]>Did you consider having the drivers embed the struct
drm_gpuvm_boin their own bo definition? I figure that would mean using the gem bo's refcounting and providing a helper to call from the driver's bo release. Looks like that could potentially save a lot of code? Or is there something that won't work with that approach?There are
drm_gpuvm_ops::vm_bo_allocanddrm_gpuvm_ops::vm_bo_freecallback for drivers to register for that purpose.Quoted 7 lines
Quoted 1 line
- Danilo
Now after looking a bit deeper, I think actually the question could be rephrased as, why don't we just use the
structdrm_gem_object::gpuvastruct as thedrm_gpuvm_boin the spirit of keeping things simple? Drivers would then just embed it in their bo subclass and we'd avoid unnecessary fields in the structdrm_gem_objectfor drivers that don't doVM_BINDyet.struct
drm_gem_object::gpuvais just a container containing a list in order to (currently) attachdrm_gpuvastructs to it and with this patch attachdrm_gpuvm_bostructs (combination of BO + VM) to it. Doing the above basically means "leave everything as it is, but move thelist_headofdrm_gpuvsper GEM to the driver specific BO structure". Having a common connection between GEM objects anddrm_gpuvastructs was one of the goals of the initial GPUVA manager patch series however.Quoted 4 lines
Sure, this won't be per bo and per vm, but it'd really only make a slight difference where we have multiple VMAs per bo, where per-vm per-bo state either needs to be duplicated or attached to a single vma (as in the case of the external bo list).
Correct, one implication is that we don't get a per VM and BO abstraction, and hence are left with a list of all
drm_gpuvastructs having the same backing BO, regardless of the VM.Quoted 11 lines
For amdgpu this was always a concern. Now that we want to keep track of external and evicted objects it's going to be a concern for most drivers I guess. Because the only structure we could use for tracking external and evicted objects we are left with (without having a
VM_BOabstraction) is structdrm_gpuva. But this structure isn't unique and we need to consider cases where userspace just allocates rather huge BOs and creates tons of mappings from it. Running the full list ofdrm_gpuvastructs (with even the ones from other VMs included) for adding an external or evicted object isn't very efficient. Not to mention that the maintenance when the mapping we've (randomly) picked as an entry for the external/evicted object list is unmapped, but there are still mappings left in the VM with the same backing BO.For the evicted object it's not much of an issue; we maintain a list of vmas needing rebinding for each VM rather than objects evicted, so there is no or very little additional overhead there. The extobj list is indeed a problem if many VMAs are bound to the same bo. Not that the code snippets are complicated, but the list traversals would be excessive.
Quoted 8 lines
Now, a way to get rid of the
VM_BOabstraction would be to use maple trees instead, since then we can storedrm_gem_objectstructs directly for each VM. However, Xe had concerns about using maple trees and preferred lists, plus having maple trees wouldn't get rid of the concerns of amdgpu not having aVM_BOabstraction for cases with tons of VMs and tons of mappings per BO. Hence, having aVM_BOabstraction enabling us to track external/evicted objects with lists seems to satisfy everyone's needs.Indeed this is a tradeoff between a simple implementation that is OK for situations with not many VMs nor VMAs per bo vs a more complex implementation that optimizes for the opposite case.
Quoted 5 lines
So if this latter is a case we need to optimize for at this point then I guess it's the way to go.
(I'm in the process of adapting the xe driver to this, so I just wanted to bring up areas where the implementations differ quite a lot and make sure options are discussed).
Thanks, I appreciate that. Just be aware of the locking issue in V3 that Boris has pointed out. I don't know if I will get to sending out a V4 today to fix that, but I'll try to do it by tomorrow.
- Danilo
Quoted 20 lines
Thanks,
Thomas
Quoted 13 lines
- Danilo
Quoted 10 lines
To me that looks like a substantial amount of less code / complexity?
/Thomas
Quoted 5 lines
Quoted 5 lines
Thanks,
Thomas
PATCH v3 6/7 drm/gpuvm: generalize dma_resv/extobj handling and GEM validation
In reply to Danilo KrummrichSo far the DRM GPUVA manager offers common infrastructure to track GPU VA allocations and mappings, generically connect GPU VA mappings to their backing buffers and perform more complex mapping operations on the GPU VA space.
However, there are more design patterns commonly used by drivers, which can potentially be generalized in order to make the DRM GPUVA manager represent a basic GPU-VM implementation. In this context, this patch aims at generalizing the following elements.
1) Provide a common dma-resv for GEM objects not being used outside of this GPU-VM.
2) Provide tracking of external GEM objects (GEM objects which are shared with other GPU-VMs).
3) Provide functions to efficiently lock all GEM objects dma-resv the GPU-VM contains mappings of.
4) Provide tracking of evicted GEM objects the GPU-VM contains mappings of, such that validation of evicted GEM objects is accelerated.
5) Provide some convinience functions for common patterns.
Rather than being designed as a "framework", the target is to make all features appear as a collection of optional helper functions, such that drivers are free to make use of the DRM GPUVA managers basic functionality and opt-in for other features without setting any feature flags, just by making use of the corresponding functions.
Big kudos to Boris Brezillon for his help to figure out locking for drivers updating the GPU VA space within the fence signalling path.
--- drivers/gpu/drm/drm_gpuvm.c | 516 ++++++++++++++++++++++++++++++++++++ include/drm/drm_gpuvm.h | 197 ++++++++++++++ 2 files changed, 713 insertions(+)
drivers/gpu/drm/drm_gpuvm.c +516 -0
@@ -73,6 +73,21 @@* &drm_gem_object list of &drm_gpuvm_bos for an existing instance of this * particular combination. If not existent a new instance is created and linked * to the &drm_gem_object. +* +* &drm_gpuvm_bo structures, since unique for a given &drm_gpuvm, are also used +* as entry for the &drm_gpuvm's lists of external and evicted objects. Those +* list are maintained in order to accelerate locking of dma-resv locks and +* validation of evicted objects bound in a &drm_gpuvm. For instance the all +* &drm_gem_object's &dma_resv of a given &drm_gpuvm can be locked by calling +* drm_gpuvm_exec_lock(). Once locked drivers can call drm_gpuvm_validate() in +* order to validate all evicted &drm_gem_objects. It is also possible to lock +* additional &drm_gem_objects by providing the corresponding parameters to +* drm_gpuvm_exec_lock() as well as open code the &drm_exec loop while making +* use of helper functions such as drm_gpuvm_prepare_range() or +* drm_gpuvm_prepare_objects(). +* +* Every bound &drm_gem_object is treated as external object when its &dma_resv +* structure is different than the &drm_gpuvm's common &dma_resv structure. */ /**
@@ -420,6 +435,20 @@* Subsequent calls to drm_gpuvm_bo_obtain() for the same &drm_gpuvm and * &drm_gem_object must be able to observe previous creations and destructions * of &drm_gpuvm_bos in order to keep instances unique. +* +* The &drm_gpuvm's lists for keeping track of external and evicted objects are +* protected against concurrent insertion / removal and iteration internally. +* +* However, drivers still need ensure to protect concurrent calls to functions +* iterating those lists, such as drm_gpuvm_validate() and +* drm_gpuvm_prepare_objects(). Every such function contains a particular +* comment and lockdep checks if possible. +* +* Functions adding or removing entries from those lists, such as +* drm_gpuvm_bo_evict() or drm_gpuvm_bo_extobj_add() may be called with external +* locks being held, e.g. in order to avoid the corresponding list to be +* (safely) modified while potentially being iternated by other API functions. +* However, this is entirely optional. */ /**
@@ -632,6 +661,131 @@* } */ +/** +* get_next_vm_bo_from_list() - get the next vm_bo element +* @__gpuvm: The GPU VM +* @__list_name: The name of the list we're iterating on +* @__local_list: A pointer to the local list used to store already iterated items +* @__prev_vm_bo: The previous element we got from drm_gpuvm_get_next_cached_vm_bo() +* +* This helper is here to provide lockless list iteration. Lockless as in, the +* iterator releases the lock immediately after picking the first element from +* the list, so list insertion deletion can happen concurrently. +* +* Elements popped from the original list are kept in a local list, so removal +* and is_empty checks can still happen while we're iterating the list. +*/ +#define get_next_vm_bo_from_list (__gpuvm, __list_name, __local_list, __prev_vm_bo) \ + ({ \ + struct drm_gpuvm_bo *__vm_bo; \ + \ + drm_gpuvm_bo_put(__prev_vm_bo); \ + \ + spin_lock(&(__gpuvm)->__list_name.lock); \ + while (!list_empty(&(__gpuvm)->__list_name.list)) { \ + __vm_bo = list_first_entry(&(__gpuvm)->__list_name.list, \ + struct drm_gpuvm_bo, \ + list.entry.__list_name); \ + if (drm_gpuvm_bo_get_unless_zero(__vm_bo)) { \ + list_move_tail(&(__vm_bo)->list.entry.__list_name, \ + __local_list); \ + break; \ + } else { \ + list_del_init(&(__vm_bo)->list.entry.__list_name); \ + __vm_bo = NULL; \ + } \ + } \ + spin_unlock(&(__gpuvm)->__list_name.lock); \ + \ + __vm_bo; \ + }) + +/** +* for_each_vm_bo_in_list() - internal vm_bo list iterator +* +* This helper is here to provide lockless list iteration. Lockless as in, the +* iterator releases the lock immediately after picking the first element from the +* list, so list insertion and deletion can happen concurrently. +* +* Typical use: +* +* struct drm_gpuvm_bo *vm_bo; +* LIST_HEAD(my_local_list); +* +* ret = 0; +* drm_gpuvm_for_each_vm_bo(gpuvm, <list_name>, &my_local_list, vm_bo) { +* ret = do_something_with_vm_bo(..., vm_bo); +* if (ret) +* break; +* } +* drm_gpuvm_bo_put(vm_bo); +* drm_gpuvm_restore_vm_bo_list(gpuvm, <list_name>, &my_local_list); +* +* +* Only used for internal list iterations, not meant to be exposed to the outside +* world. +*/ +#define for_each_vm_bo_in_list (__gpuvm, __list_name, __local_list, __vm_bo) \ + for (__vm_bo = get_next_vm_bo_from_list(__gpuvm, __list_name, \ + __local_list, NULL); \ + __vm_bo; \ + __vm_bo = get_next_vm_bo_from_list(__gpuvm, __list_name, \ + __local_list, __vm_bo)) \ + +/** +* restore_vm_bo_list() - move vm_bo elements back to their original list +* @__gpuvm: The GPU VM +* @__list_name: The name of the list we're iterating on +* @__local_list: A pointer to the local list used to store already iterated items +* +* When we're done iterating a vm_bo list, we should call restore_vm_bo_list() +* to restore the original state and let new iterations take place. +*/ +#define restore_vm_bo_list (__gpuvm, __list_name, __local_list) \ + do { \ +/* Merge back the two lists, moving local list elements to the \ +* head to preserve previous ordering, in case it matters. \ +*/ \ +spin_lock (&(__gpuvm)->__list_name.lock); \ +list_splice (__local_list, &(__gpuvm)->__list_name.list); \ +spin_unlock (&(__gpuvm)->__list_name.lock); \ + }while (0 ) +/** +* drm_gpuvm_bo_list_add() - insert a vm_bo into the given list +* @__vm_bo: the &drm_gpuvm_bo +* @__list_name: the name of the list to insert into +* +* Inserts the given @__vm_bo into the list specified by @__list_name and +* increases the vm_bo's reference count. + */ +#define drm_gpuvm_bo_list_add (__vm_bo, __list_name) \ + do { \ + spin_lock(&(__vm_bo)->vm->__list_name.lock); \ + if (list_empty(&(__vm_bo)->list.entry.__list_name)) \ + list_add_tail(&(__vm_bo)->list.entry.__list_name, \ + &(__vm_bo)->vm->__list_name.list); \ + spin_unlock(&(__vm_bo)->vm->__list_name.lock); \ + } while (0) + +/** +* drm_gpuvm_bo_list_del() - remove a vm_bo from the given list +* @__vm_bo: the &drm_gpuvm_bo +* @__list_name: the name of the list to insert into +* +* Removes the given @__vm_bo from the list specified by @__list_name and +* decreases the vm_bo's reference count. +*/ +#define drm_gpuvm_bo_list_del (__vm_bo, __list_name) \ + do { \ + spin_lock(&(__vm_bo)->vm->__list_name.lock); \ + if (!list_empty(&(__vm_bo)->list.entry.__list_name)) \ + list_del_init(&(__vm_bo)->list.entry.__list_name); \ + spin_unlock(&(__vm_bo)->vm->__list_name.lock); \ + } while (0) + +static int __must_check +drm_gpuvm_bo_get_unless_zero (struct drm_gpuvm_bo *vm_bo); +#define to_drm_gpuva (__node) container_of((__node), struct drm_gpuva, rb.node)#define GPUVA_START (node) ((node)->va.addr)
@@ -713,6 +867,12 @@ drm_gpuvm_init(struct drm_gpuvm *gpuvm, struct drm_device *drm, gpuvm->rb.tree =RB_ROOT_CACHED ;INIT_LIST_HEAD (&gpuvm->rb.list); +INIT_LIST_HEAD (&gpuvm->extobj.list); +spin_lock_init (&gpuvm->extobj.lock); + +INIT_LIST_HEAD (&gpuvm->evict.list); +spin_lock_init (&gpuvm->evict.lock); +drm_gpuva_check_overflow (start_offset, range); gpuvm->mm_start = start_offset; gpuvm->mm_range = range;
@@ -754,10 +914,302 @@ drm_gpuvm_destroy(struct drm_gpuvm *gpuvm)WARN (!RB_EMPTY_ROOT (&gpuvm->rb.tree.rb_root),"GPUVA tree is not empty, potentially leaking memory.\n" ); +WARN (!list_empty (&gpuvm->extobj.list),"Extobj list should be empty.\n" ); +WARN (!list_empty (&gpuvm->evict.list),"Evict list should be empty.\n" ); +drm_gem_private_object_fini (&gpuvm->d_obj); }EXPORT_SYMBOL_GPL (drm_gpuvm_destroy); +/** +* drm_gpuvm_prepare_objects() - prepare all assoiciated BOs +* @gpuvm: the &drm_gpuvm +* @exec: the &drm_exec locking context +* @num_fences: the amount of &dma_fences to reserve +* +* Calls drm_exec_prepare_obj() for all &drm_gem_objects the given +* &drm_gpuvm contains mappings of. +* +* Using this function directly, it is the drivers responsibility to call +* drm_exec_init() and drm_exec_fini() accordingly. +* +* Note: This function is safe against concurrent insertion and removal of +* external objects, however it is not safe against concurrent usage itself. +* +* Drivers need to make sure to protect this case with either an outer VM lock +* or by calling drm_gpuvm_prepare_vm() before this function within the +* drm_exec_until_all_locked() loop, such that the GPUVM's dma-resv lock ensures +* mutual exclusion. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_prepare_objects (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +unsigned int num_fences) +{ +struct drm_gpuvm_bo *vm_bo; +LIST_HEAD (extobjs); +int ret =0 ; + +for_each_vm_bo_in_list (gpuvm, extobj, &extobjs, vm_bo) { + ret =drm_exec_prepare_obj (exec, vm_bo->obj, num_fences); +if (ret) +break ; + } +/* Drop ref in case we break out of the loop. */ +drm_gpuvm_bo_put (vm_bo); +restore_vm_bo_list (gpuvm, extobj, &extobjs); + +return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_prepare_objects); + +/** +* drm_gpuvm_prepare_range() - prepare all BOs mapped within a given range +* @gpuvm: the &drm_gpuvm +* @exec: the &drm_exec locking context +* @addr: the start address within the VA space +* @range: the range to iterate within the VA space +* @num_fences: the amount of &dma_fences to reserve +* +* Calls drm_exec_prepare_obj() for all &drm_gem_objects mapped between @addr +* and @addr + @range. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_prepare_range (struct drm_gpuvm *gpuvm,struct drm_exec *exec, +u64 addr,u64 range,unsigned int num_fences) +{ +struct drm_gpuva *va; +u64 end = addr + range; +int ret; + +drm_gpuvm_for_each_va_range (va, gpuvm, addr, end) { +struct drm_gem_object *obj = va->gem.obj; + + ret =drm_exec_prepare_obj (exec, obj, num_fences); +if (ret) +return ret; + } + +return 0 ; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_prepare_range); + +/** +* drm_gpuvm_exec_lock() - lock all dma-resv of all assoiciated BOs +* @vm_exec: the &drm_gpuvm_exec abstraction +* @num_fences: the amount of &dma_fences to reserve +* @interruptible: sleep interruptible if waiting +* +* Acquires all dma-resv locks of all &drm_gem_objects the given +* &drm_gpuvm contains mappings of. +* +* Addionally, when calling this function with struct drm_gpuvm_exec::extra +* being set the driver receives the given @fn callback to lock additional +* dma-resv in the context of the &drm_gpuvm_exec instance. Typically, drivers +* would call drm_exec_prepare_obj() from within this callback. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_exec_lock (struct drm_gpuvm_exec *vm_exec, +unsigned int num_fences, +bool interruptible) +{ +struct drm_gpuvm *gpuvm = vm_exec->vm; +struct drm_exec *exec = &vm_exec->exec; +uint32_t flags; +int ret; + + flags = interruptible ?DRM_EXEC_INTERRUPTIBLE_WAIT :0 | +DRM_EXEC_IGNORE_DUPLICATES ; + +drm_exec_init (exec, flags); + +drm_exec_until_all_locked (exec) { + ret =drm_gpuvm_prepare_vm (gpuvm, exec, num_fences); +drm_exec_retry_on_contention (exec); +if (ret) + goto err; + + ret =drm_gpuvm_prepare_objects (gpuvm, exec, num_fences); +drm_exec_retry_on_contention (exec); +if (ret) + goto err; + +if (vm_exec->extra.fn) { + ret = vm_exec->extra.fn (vm_exec, num_fences); +drm_exec_retry_on_contention (exec); +if (ret) + goto err; + } + } + +return 0 ; + +err: +drm_exec_fini (exec); +return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock); + +static int +fn_lock_array (struct drm_gpuvm_exec *vm_exec,unsigned int num_fences) +{ +struct { +struct drm_gem_object **objs; +unsigned int num_objs; + } *args = vm_exec->extra.priv; + +return drm_exec_prepare_array (&vm_exec->exec, args->objs, + args->num_objs, num_fences); +} + +/** +* drm_gpuvm_exec_lock_array() - lock all dma-resv of all assoiciated BOs +* @vm_exec: the &drm_gpuvm_exec abstraction +* @objs: additional &drm_gem_objects to lock +* @num_objs: the number of additional &drm_gem_objects to lock +* @num_fences: the amount of &dma_fences to reserve +* @interruptible: sleep interruptible if waiting +* +* Acquires all dma-resv locks of all &drm_gem_objects the given &drm_gpuvm +* contains mappings of, plus the ones given through @objs. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_exec_lock_array (struct drm_gpuvm_exec *vm_exec, +struct drm_gem_object **objs, +unsigned int num_objs, +unsigned int num_fences, +bool interruptible) +{ +struct { +struct drm_gem_object **objs; +unsigned int num_objs; + } args; + + args.objs = objs; + args.num_objs = num_objs; + + vm_exec->extra.fn = fn_lock_array; + vm_exec->extra.priv = &args; + +return drm_gpuvm_exec_lock (vm_exec, num_fences, interruptible); +} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock_array); + +/** +* drm_gpuvm_exec_lock_range() - prepare all BOs mapped within a given range +* @vm_exec: the &drm_gpuvm_exec abstraction +* @addr: the start address within the VA space +* @range: the range to iterate within the VA space +* @num_fences: the amount of &dma_fences to reserve +* @interruptible: sleep interruptible if waiting +* +* Acquires all dma-resv locks of all &drm_gem_objects mapped between @addr and +* @addr + @range. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_exec_lock_range (struct drm_gpuvm_exec *vm_exec, +u64 addr,u64 range, +unsigned int num_fences, +bool interruptible) +{ +struct drm_gpuvm *gpuvm = vm_exec->vm; +struct drm_exec *exec = &vm_exec->exec; +uint32_t flags; +int ret; + + flags = interruptible ?DRM_EXEC_INTERRUPTIBLE_WAIT :0 | +DRM_EXEC_IGNORE_DUPLICATES ; + +drm_exec_init (exec, flags); + +drm_exec_until_all_locked (exec) { + ret =drm_gpuvm_prepare_range (gpuvm, exec, addr, range, + num_fences); +drm_exec_retry_on_contention (exec); +if (ret) + goto err; + } + +return ret; + +err: +drm_exec_fini (exec); +return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock_range); + +/** +* drm_gpuvm_validate() - validate all BOs marked as evicted +* @gpuvm: the &drm_gpuvm to validate evicted BOs +* +* Calls the &drm_gpuvm_ops.bo_validate callback for all evicted buffer +* objects being mapped in the given &drm_gpuvm. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_validate (struct drm_gpuvm *gpuvm) +{ +const struct drm_gpuvm_ops *ops = gpuvm->ops; +struct drm_gpuvm_bo *vm_bo; +LIST_HEAD (evict); +int ret =0 ; + +if (unlikely (!ops || !ops->bo_validate)) +return -ENOTSUPP ; + +for_each_vm_bo_in_list (gpuvm, evict, &evict, vm_bo) { +dma_resv_assert_held (vm_bo->obj->resv); + ret = ops->bo_validate (vm_bo->obj); +if (ret) +break ; + } +/* Drop ref in case we break out of the loop. */ +drm_gpuvm_bo_put (vm_bo); +restore_vm_bo_list (gpuvm, evict, &evict); + +return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_validate); + +/** +* drm_gpuvm_resv_add_fence - add fence to private and all extobj +* dma-resv +* @gpuvm: the &drm_gpuvm to add a fence to +* @exec: the &drm_exec locking context +* @fence: fence to add +* @private_usage: private dma-resv usage +* @extobj_usage: extobj dma-resv usage +*/ +void +drm_gpuvm_resv_add_fence (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +struct dma_fence *fence, +enum dma_resv_usage private_usage, +enum dma_resv_usage extobj_usage) +{ +struct drm_gem_object *obj; +unsigned long index; + +drm_exec_for_each_locked_object (exec, index, obj) { +dma_resv_assert_held (obj->resv); +dma_resv_add_fence (obj->resv, fence, +drm_gpuvm_is_extobj (gpuvm, obj) ? + private_usage : extobj_usage); + } +} +EXPORT_SYMBOL_GPL (drm_gpuvm_resv_add_fence); +/** * drm_gpuvm_bo_create() - create a new instance of struct drm_gpuvm_bo * @gpuvm: The &drm_gpuvm the @obj is mapped in.
@@ -790,6 +1242,9 @@ drm_gpuvm_bo_create(struct drm_gpuvm *gpuvm,INIT_LIST_HEAD (&vm_bo->list.gpuva);INIT_LIST_HEAD (&vm_bo->list.entry.gem); +INIT_LIST_HEAD (&vm_bo->list.entry.extobj); +INIT_LIST_HEAD (&vm_bo->list.entry.evict); +drm_gem_object_get (obj);return vm_bo;
@@ -807,6 +1262,14 @@ drm_gpuvm_bo_destroy(struct kref *kref)drm_gem_gpuva_assert_lock_held (vm_bo->obj); +spin_lock (&gpuvm->extobj.lock); +list_del (&vm_bo->list.entry.extobj); +spin_unlock (&gpuvm->extobj.lock); + +spin_lock (&gpuvm->evict.lock); +list_del (&vm_bo->list.entry.evict); +spin_unlock (&gpuvm->evict.lock); +list_del (&vm_bo->list.entry.gem);drm_gem_object_put (obj);
@@ -822,6 +1285,11 @@ drm_gpuvm_bo_destroy(struct kref *kref)* @vm_bo: the &drm_gpuvm_bo to release the reference of * * This releases a reference to @vm_bo. +* +* If the reference count drops to zero, the &gpuvm_bo is destroyed, which +* includes removing it from the GEMs gpuva list. Hence, if a call to this +* function can potentially let the reference count to zero the caller must +* hold the dma-resv or driver specific GEM gpuva lock. */ void drm_gpuvm_bo_put (struct drm_gpuvm_bo *vm_bo)
@@ -831,6 +1299,12 @@ drm_gpuvm_bo_put(struct drm_gpuvm_bo *vm_bo) }EXPORT_SYMBOL_GPL (drm_gpuvm_bo_put); +static int __must_check +drm_gpuvm_bo_get_unless_zero (struct drm_gpuvm_bo *vm_bo) +{ +return kref_get_unless_zero (&vm_bo->kref); +} +static struct drm_gpuvm_bo *__drm_gpuvm_bo_find (struct drm_gpuvm *gpuvm,struct drm_gem_object *obj)
@@ -938,6 +1412,48 @@ drm_gpuvm_bo_obtain_prealloc(struct drm_gpuvm_bo *__vm_bo) }EXPORT_SYMBOL_GPL (drm_gpuvm_bo_obtain_prealloc); +/** +* drm_gpuvm_bo_extobj_add() - adds the &drm_gpuvm_bo to its &drm_gpuvm's +* extobj list +* @vm_bo: The &drm_gpuvm_bo to add to its &drm_gpuvm's the extobj list. +* +* Adds the given @vm_bo to its &drm_gpuvm's extobj list if not on the list +* already and if the corresponding &drm_gem_object is an external object, +* actually. +*/ +void +drm_gpuvm_bo_extobj_add (struct drm_gpuvm_bo *vm_bo) +{ +struct drm_gpuvm *gpuvm = vm_bo->vm; + +if (drm_gpuvm_is_extobj (gpuvm, vm_bo->obj)) +drm_gpuvm_bo_list_add (vm_bo, extobj); +} +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_extobj_add); + +/** +* drm_gpuvm_bo_evict() - add / remove a &drm_gem_object to / from a +* &drm_gpuvms evicted list +* @obj: the &drm_gem_object to add or remove +* @evict: indicates whether the object is evicted +* +* Adds a &drm_gem_object to or removes it from all &drm_gpuvms evicted +* list containing a mapping of this &drm_gem_object. +*/ +void +drm_gpuvm_bo_evict (struct drm_gem_object *obj,bool evict) +{ +struct drm_gpuvm_bo *vm_bo; + +drm_gem_for_each_gpuvm_bo (vm_bo, obj) { +if (evict) +drm_gpuvm_bo_list_add (vm_bo, evict); +else +drm_gpuvm_bo_list_del (vm_bo, evict); + } +} +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_evict); +static int __drm_gpuva_insert (struct drm_gpuvm *gpuvm,struct drm_gpuva *va)
include/drm/drm_gpuvm.h +197 -0
243 lines not on this page · raw message
Signature · 1 line
2.41.0
kernel test robot
build botRe: [PATCH drm-misc-next v3 6/7] drm/gpuvm: generalize dma_resv/extobj handling and GEM validation
kernel test robot noticed the following build warnings:
Hi Danilo,
kernel test robot noticed the following build warnings:
[auto build test WARNING on 6bd3d8da51ca1ec97c724016466606aec7739b9f]
url: https://github.com/intel-lab-lkp/linux/commits/Danilo-Krummrich/drm-gpuvm-rename-struct-drm_gpuva_manager-to-struct-drm_gpuvm/20230909-233346 base: 6bd3d8da51ca1ec97c724016466606aec7739b9f patch link: https://lore.kernel.org/r/20230909153125.30032-7-dakr%40redhat.com patch subject: [PATCH drm-misc-next v3 6/7] drm/gpuvm: generalize dma_resv/extobj handling and GEM validation config: riscv-defconfig (https://download.01.org/0day-ci/archive/20230910/[email protected]/config) compiler: riscv64-linux-gcc (GCC) 13.2.0 reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20230910/[email protected]/reproduce)
If you fix the issue in a separate patch/commit (i.e. not just a new version of the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <[email protected]>
| Closes: https://lore.kernel.org/oe-kbuild-all/[email protected]/
All warnings (new ones prefixed by >>):
>> drivers/gpu/drm/drm_gpuvm.c:734: warning: Function parameter or member '__gpuvm' not described in 'for_each_vm_bo_in_list' >> drivers/gpu/drm/drm_gpuvm.c:734: warning: Function parameter or member '__list_name' not described in 'for_each_vm_bo_in_list' >> drivers/gpu/drm/drm_gpuvm.c:734: warning: Function parameter or member '__local_list' not described in 'for_each_vm_bo_in_list' >> drivers/gpu/drm/drm_gpuvm.c:734: warning: Function parameter or member '__vm_bo' not described in 'for_each_vm_bo_in_list'
vim +734 drivers/gpu/drm/drm_gpuvm.c
32
33 /**
34 * DOC: Overview
35 *
36 * The DRM GPU VA Manager, represented by struct drm_gpuvm keeps track of a
37 * GPU's virtual address (VA) space and manages the corresponding virtual
38 * mappings represented by &drm_gpuva objects. It also keeps track of the
39 * mapping's backing &drm_gem_object buffers.
40 *
41 * &drm_gem_object buffers maintain a list of &drm_gpuva objects representing
42 * all existent GPU VA mappings using this &drm_gem_object as backing buffer.
43 *
44 * GPU VAs can be flagged as sparse, such that drivers may use GPU VAs to also
45 * keep track of sparse PTEs in order to support Vulkan 'Sparse Resources'.
46 *
47 * The GPU VA manager internally uses a rb-tree to manage the
48 * &drm_gpuva mappings within a GPU's virtual address space.
49 *
50 * The &drm_gpuvm structure contains a special &drm_gpuva representing the
51 * portion of VA space reserved by the kernel. This node is initialized together
52 * with the GPU VA manager instance and removed when the GPU VA manager is
53 * destroyed.
54 *
55 * In a typical application drivers would embed struct drm_gpuvm and
56 * struct drm_gpuva within their own driver specific structures, there won't be
57 * any memory allocations of its own nor memory allocations of &drm_gpuva
58 * entries.
59 *
60 * The data structures needed to store &drm_gpuvas within the &drm_gpuvm are
61 * contained within struct drm_gpuva already. Hence, for inserting &drm_gpuva
62 * entries from within dma-fence signalling critical sections it is enough to
63 * pre-allocate the &drm_gpuva structures.
64 *
65 * In order to connect a struct drm_gpuva its backing &drm_gem_object each
66 * &drm_gem_object maintains a list of &drm_gpuvm_bo structures, and each
67 * &drm_gpuvm_bo contains a list of &&drm_gpuva structures.
68 *
69 * A &drm_gpuvm_bo is an abstraction that represents a combination of a
70 * &drm_gpuvm and a &drm_gem_object. Every such combination should be unique.
71 * This is ensured by the API through drm_gpuvm_bo_obtain() and
72 * drm_gpuvm_bo_obtain_prealloc() which first look into the corresponding
73 * &drm_gem_object list of &drm_gpuvm_bos for an existing instance of this
74 * particular combination. If not existent a new instance is created and linked
75 * to the &drm_gem_object.
76 *
77 * &drm_gpuvm_bo structures, since unique for a given &drm_gpuvm, are also used
78 * as entry for the &drm_gpuvm's lists of external and evicted objects. Those
79 * list are maintained in order to accelerate locking of dma-resv locks and
80 * validation of evicted objects bound in a &drm_gpuvm. For instance the all
81 * &drm_gem_object's &dma_resv of a given &drm_gpuvm can be locked by calling
82 * drm_gpuvm_exec_lock(). Once locked drivers can call drm_gpuvm_validate() in
83 * order to validate all evicted &drm_gem_objects. It is also possible to lock
84 * additional &drm_gem_objects by providing the corresponding parameters to
85 * drm_gpuvm_exec_lock() as well as open code the &drm_exec loop while making
86 * use of helper functions such as drm_gpuvm_prepare_range() or
87 * drm_gpuvm_prepare_objects().
88 *
89 * Every bound &drm_gem_object is treated as external object when its &dma_resv
90 * structure is different than the &drm_gpuvm's common &dma_resv structure.
91 */
92
93 /**
94 * DOC: Split and Merge
95 *
96 * Besides its capability to manage and represent a GPU VA space, the
97 * GPU VA manager also provides functions to let the &drm_gpuvm calculate a
98 * sequence of operations to satisfy a given map or unmap request.
99 *
100 * Therefore the DRM GPU VA manager provides an algorithm implementing splitting
101 * and merging of existent GPU VA mappings with the ones that are requested to
102 * be mapped or unmapped. This feature is required by the Vulkan API to
103 * implement Vulkan 'Sparse Memory Bindings' - drivers UAPIs often refer to this
104 * as VM BIND.
105 *
106 * Drivers can call drm_gpuvm_sm_map() to receive a sequence of callbacks
107 * containing map, unmap and remap operations for a given newly requested
108 * mapping. The sequence of callbacks represents the set of operations to
109 * execute in order to integrate the new mapping cleanly into the current state
110 * of the GPU VA space.
111 *
112 * Depending on how the new GPU VA mapping intersects with the existent mappings
113 * of the GPU VA space the &drm_gpuvm_ops callbacks contain an arbitrary amount
114 * of unmap operations, a maximum of two remap operations and a single map
115 * operation. The caller might receive no callback at all if no operation is
116 * required, e.g. if the requested mapping already exists in the exact same way.
117 *
118 * The single map operation represents the original map operation requested by
119 * the caller.
120 *
121 * &drm_gpuva_op_unmap contains a 'keep' field, which indicates whether the
122 * &drm_gpuva to unmap is physically contiguous with the original mapping
123 * request. Optionally, if 'keep' is set, drivers may keep the actual page table
124 * entries for this &drm_gpuva, adding the missing page table entries only and
125 * update the &drm_gpuvm's view of things accordingly.
126 *
127 * Drivers may do the same optimization, namely delta page table updates, also
128 * for remap operations. This is possible since &drm_gpuva_op_remap consists of
129 * one unmap operation and one or two map operations, such that drivers can
130 * derive the page table update delta accordingly.
131 *
132 * Note that there can't be more than two existent mappings to split up, one at
133 * the beginning and one at the end of the new mapping, hence there is a
134 * maximum of two remap operations.
135 *
136 * Analogous to drm_gpuvm_sm_map() drm_gpuvm_sm_unmap() uses &drm_gpuvm_ops to
137 * call back into the driver in order to unmap a range of GPU VA space. The
138 * logic behind this function is way simpler though: For all existent mappings
139 * enclosed by the given range unmap operations are created. For mappings which
140 * are only partically located within the given range, remap operations are
141 * created such that those mappings are split up and re-mapped partically.
142 *
143 * As an alternative to drm_gpuvm_sm_map() and drm_gpuvm_sm_unmap(),
144 * drm_gpuvm_sm_map_ops_create() and drm_gpuvm_sm_unmap_ops_create() can be used
145 * to directly obtain an instance of struct drm_gpuva_ops containing a list of
146 * &drm_gpuva_op, which can be iterated with drm_gpuva_for_each_op(). This list
147 * contains the &drm_gpuva_ops analogous to the callbacks one would receive when
148 * calling drm_gpuvm_sm_map() or drm_gpuvm_sm_unmap(). While this way requires
149 * more memory (to allocate the &drm_gpuva_ops), it provides drivers a way to
150 * iterate the &drm_gpuva_op multiple times, e.g. once in a context where memory
151 * allocations are possible (e.g. to allocate GPU page tables) and once in the
152 * dma-fence signalling critical path.
153 *
154 * To update the &drm_gpuvm's view of the GPU VA space drm_gpuva_insert() and
155 * drm_gpuva_remove() may be used. These functions can safely be used from
156 * &drm_gpuvm_ops callbacks originating from drm_gpuvm_sm_map() or
157 * drm_gpuvm_sm_unmap(). However, it might be more convenient to use the
158 * provided helper functions drm_gpuva_map(), drm_gpuva_remap() and
159 * drm_gpuva_unmap() instead.
160 *
161 * The following diagram depicts the basic relationships of existent GPU VA
162 * mappings, a newly requested mapping and the resulting mappings as implemented
163 * by drm_gpuvm_sm_map() - it doesn't cover any arbitrary combinations of these.
164 *
165 * 1) Requested mapping is identical. Replace it, but indicate the backing PTEs
166 * could be kept.
167 *
168 * ::
169 *
170 * 0 a 1
171 * old: |-----------| (bo_offset=n)
172 *
173 * 0 a 1
174 * req: |-----------| (bo_offset=n)
175 *
176 * 0 a 1
177 * new: |-----------| (bo_offset=n)
178 *
179 *
180 * 2) Requested mapping is identical, except for the BO offset, hence replace
181 * the mapping.
182 *
183 * ::
184 *
185 * 0 a 1
186 * old: |-----------| (bo_offset=n)
187 *
188 * 0 a 1
189 * req: |-----------| (bo_offset=m)
190 *
191 * 0 a 1
192 * new: |-----------| (bo_offset=m)
193 *
194 *
195 * 3) Requested mapping is identical, except for the backing BO, hence replace
196 * the mapping.
197 *
198 * ::
199 *
200 * 0 a 1
201 * old: |-----------| (bo_offset=n)
202 *
203 * 0 b 1
204 * req: |-----------| (bo_offset=n)
205 *
206 * 0 b 1
207 * new: |-----------| (bo_offset=n)
208 *
209 *
210 * 4) Existent mapping is a left aligned subset of the requested one, hence
211 * replace the existent one.
212 *
213 * ::
214 *
215 * 0 a 1
216 * old: |-----| (bo_offset=n)
217 *
218 * 0 a 2
219 * req: |-----------| (bo_offset=n)
220 *
221 * 0 a 2
222 * new: |-----------| (bo_offset=n)
223 *
224 * .. note::
225 * We expect to see the same result for a request with a different BO
226 * and/or non-contiguous BO offset.
227 *
228 *
229 * 5) Requested mapping's range is a left aligned subset of the existent one,
230 * but backed by a different BO. Hence, map the requested mapping and split
231 * the existent one adjusting its BO offset.
232 *
233 * ::
234 *
235 * 0 a 2
236 * old: |-----------| (bo_offset=n)
237 *
238 * 0 b 1
239 * req: |-----| (bo_offset=n)
240 *
241 * 0 b 1 a' 2
242 * new: |-----|-----| (b.bo_offset=n, a.bo_offset=n+1)
243 *
244 * .. note::
245 * We expect to see the same result for a request with a different BO
246 * and/or non-contiguous BO offset.
247 *
248 *
249 * 6) Existent mapping is a superset of the requested mapping. Split it up, but
250 * indicate that the backing PTEs could be kept.
251 *
252 * ::
253 *
254 * 0 a 2
255 * old: |-----------| (bo_offset=n)
256 *
257 * 0 a 1
258 * req: |-----| (bo_offset=n)
259 *
260 * 0 a 1 a' 2
261 * new: |-----|-----| (a.bo_offset=n, a'.bo_offset=n+1)
262 *
263 *
264 * 7) Requested mapping's range is a right aligned subset of the existent one,
265 * but backed by a different BO. Hence, map the requested mapping and split
266 * the existent one, without adjusting the BO offset.
267 *
268 * ::
269 *
270 * 0 a 2
271 * old: |-----------| (bo_offset=n)
272 *
273 * 1 b 2
274 * req: |-----| (bo_offset=m)
275 *
276 * 0 a 1 b 2
277 * new: |-----|-----| (a.bo_offset=n,b.bo_offset=m)
278 *
279 *
280 * 8) Existent mapping is a superset of the requested mapping. Split it up, but
281 * indicate that the backing PTEs could be kept.
282 *
283 * ::
284 *
285 * 0 a 2
286 * old: |-----------| (bo_offset=n)
287 *
288 * 1 a 2
289 * req: |-----| (bo_offset=n+1)
290 *
291 * 0 a' 1 a 2
292 * new: |-----|-----| (a'.bo_offset=n, a.bo_offset=n+1)
293 *
294 *
295 * 9) Existent mapping is overlapped at the end by the requested mapping backed
296 * by a different BO. Hence, map the requested mapping and split up the
297 * existent one, without adjusting the BO offset.
298 *
299 * ::
300 *
301 * 0 a 2
302 * old: |-----------| (bo_offset=n)
303 *
304 * 1 b 3
305 * req: |-----------| (bo_offset=m)
306 *
307 * 0 a 1 b 3
308 * new: |-----|-----------| (a.bo_offset=n,b.bo_offset=m)
309 *
310 *
311 * 10) Existent mapping is overlapped by the requested mapping, both having the
312 * same backing BO with a contiguous offset. Indicate the backing PTEs of
313 * the old mapping could be kept.
314 *
315 * ::
316 *
317 * 0 a 2
318 * old: |-----------| (bo_offset=n)
319 *
320 * 1 a 3
321 * req: |-----------| (bo_offset=n+1)
322 *
323 * 0 a' 1 a 3
324 * new: |-----|-----------| (a'.bo_offset=n, a.bo_offset=n+1)
325 *
326 *
327 * 11) Requested mapping's range is a centered subset of the existent one
328 * having a different backing BO. Hence, map the requested mapping and split
329 * up the existent one in two mappings, adjusting the BO offset of the right
330 * one accordingly.
331 *
332 * ::
333 *
334 * 0 a 3
335 * old: |-----------------| (bo_offset=n)
336 *
337 * 1 b 2
338 * req: |-----| (bo_offset=m)
339 *
340 * 0 a 1 b 2 a' 3
341 * new: |-----|-----|-----| (a.bo_offset=n,b.bo_offset=m,a'.bo_offset=n+2)
342 *
343 *
344 * 12) Requested mapping is a contiguous subset of the existent one. Split it
345 * up, but indicate that the backing PTEs could be kept.
346 *
347 * ::
348 *
349 * 0 a 3
350 * old: |-----------------| (bo_offset=n)
351 *
352 * 1 a 2
353 * req: |-----| (bo_offset=n+1)
354 *
355 * 0 a' 1 a 2 a'' 3
356 * old: |-----|-----|-----| (a'.bo_offset=n, a.bo_offset=n+1, a''.bo_offset=n+2)
357 *
358 *
359 * 13) Existent mapping is a right aligned subset of the requested one, hence
360 * replace the existent one.
361 *
362 * ::
363 *
364 * 1 a 2
365 * old: |-----| (bo_offset=n+1)
366 *
367 * 0 a 2
368 * req: |-----------| (bo_offset=n)
369 *
370 * 0 a 2
371 * new: |-----------| (bo_offset=n)
372 *
373 * .. note::
374 * We expect to see the same result for a request with a different bo
375 * and/or non-contiguous bo_offset.
376 *
377 *
378 * 14) Existent mapping is a centered subset of the requested one, hence
379 * replace the existent one.
380 *
381 * ::
382 *
383 * 1 a 2
384 * old: |-----| (bo_offset=n+1)
385 *
386 * 0 a 3
387 * req: |----------------| (bo_offset=n)
388 *
389 * 0 a 3
390 * new: |----------------| (bo_offset=n)
391 *
392 * .. note::
393 * We expect to see the same result for a request with a different bo
394 * and/or non-contiguous bo_offset.
395 *
396 *
397 * 15) Existent mappings is overlapped at the beginning by the requested mapping
398 * backed by a different BO. Hence, map the requested mapping and split up
399 * the existent one, adjusting its BO offset accordingly.
400 *
401 * ::
402 *
403 * 1 a 3
404 * old: |-----------| (bo_offset=n)
405 *
406 * 0 b 2
407 * req: |-----------| (bo_offset=m)
408 *
409 * 0 b 2 a' 3
410 * new: |-----------|-----| (b.bo_offset=m,a.bo_offset=n+2)
411 */
412
413 /**
414 * DOC: Locking
415 *
416 * Generally, the GPU VA manager does not take care of locking itself, it is
417 * the drivers responsibility to take care about locking. Drivers might want to
418 * protect the following operations: inserting, removing and iterating
419 * &drm_gpuva objects as well as generating all kinds of operations, such as
420 * split / merge or prefetch.
421 *
422 * The GPU VA manager also does not take care of the locking of the backing
423 * &drm_gem_object buffers GPU VA lists and &drm_gpuvm_bo abstractions by
424 * itself; drivers are responsible to enforce mutual exclusion using either the
425 * GEMs dma_resv lock or alternatively a driver specific external lock. For the
426 * latter see also drm_gem_gpuva_set_lock().
427 *
428 * However, the GPU VA manager contains lockdep checks to ensure callers of its
429 * API hold the corresponding lock whenever the &drm_gem_objects GPU VA list is
430 * accessed by functions such as drm_gpuva_link() or drm_gpuva_unlink(), but
431 * also drm_gpuvm_bo_obtain() and drm_gpuvm_bo_put().
432 *
433 * The latter is required since on creation and destruction of a &drm_gpuvm_bo
434 * the &drm_gpuvm_bo is attached / removed from the &drm_gem_objects gpuva list.
435 * Subsequent calls to drm_gpuvm_bo_obtain() for the same &drm_gpuvm and
436 * &drm_gem_object must be able to observe previous creations and destructions
437 * of &drm_gpuvm_bos in order to keep instances unique.
438 *
439 * The &drm_gpuvm's lists for keeping track of external and evicted objects are
440 * protected against concurrent insertion / removal and iteration internally.
441 *
442 * However, drivers still need ensure to protect concurrent calls to functions
443 * iterating those lists, such as drm_gpuvm_validate() and
444 * drm_gpuvm_prepare_objects(). Every such function contains a particular
445 * comment and lockdep checks if possible.
446 *
447 * Functions adding or removing entries from those lists, such as
448 * drm_gpuvm_bo_evict() or drm_gpuvm_bo_extobj_add() may be called with external
449 * locks being held, e.g. in order to avoid the corresponding list to be
450 * (safely) modified while potentially being iternated by other API functions.
451 * However, this is entirely optional.
452 */
453
454 /**
455 * DOC: Examples
456 *
457 * This section gives two examples on how to let the DRM GPUVA Manager generate
458 * &drm_gpuva_op in order to satisfy a given map or unmap request and how to
459 * make use of them.
460 *
461 * The below code is strictly limited to illustrate the generic usage pattern.
462 * To maintain simplicitly, it doesn't make use of any abstractions for common
463 * code, different (asyncronous) stages with fence signalling critical paths,
464 * any other helpers or error handling in terms of freeing memory and dropping
465 * previously taken locks.
466 *
467 * 1) Obtain a list of &drm_gpuva_op to create a new mapping::
468 *
469 * // Allocates a new &drm_gpuva.
470 * struct drm_gpuva * driver_gpuva_alloc(void);
471 *
472 * // Typically drivers would embedd the &drm_gpuvm and &drm_gpuva
473 * // structure in individual driver structures and lock the dma-resv with
474 * // drm_exec or similar helpers.
475 * int driver_mapping_create(struct drm_gpuvm *gpuvm,
476 * u64 addr, u64 range,
477 * struct drm_gem_object *obj, u64 offset)
478 * {
479 * struct drm_gpuva_ops *ops;
480 * struct drm_gpuva_op *op
481 * struct drm_gpuvm_bo *vm_bo;
482 *
483 * driver_lock_va_space();
484 * ops = drm_gpuvm_sm_map_ops_create(gpuvm, addr, range,
485 * obj, offset);
486 * if (IS_ERR(ops))
487 * return PTR_ERR(ops);
488 *
489 * vm_bo = drm_gpuvm_bo_obtain(gpuvm, obj);
490 * if (IS_ERR(vm_bo))
491 * return PTR_ERR(vm_bo);
492 *
493 * drm_gpuva_for_each_op(op, ops) {
494 * struct drm_gpuva *va;
495 *
496 * switch (op->op) {
497 * case DRM_GPUVA_OP_MAP:
498 * va = driver_gpuva_alloc();
499 * if (!va)
500 * ; // unwind previous VA space updates,
501 * // free memory and unlock
502 *
503 * driver_vm_map();
504 * drm_gpuva_map(gpuvm, va, &op->map);
505 * drm_gpuva_link(va, vm_bo);
506 *
507 * break;
508 * case DRM_GPUVA_OP_REMAP: {
509 * struct drm_gpuva *prev = NULL, *next = NULL;
510 *
511 * va = op->remap.unmap->va;
512 *
513 * if (op->remap.prev) {
514 * prev = driver_gpuva_alloc();
515 * if (!prev)
516 * ; // unwind previous VA space
517 * // updates, free memory and
518 * // unlock
519 * }
520 *
521 * if (op->remap.next) {
522 * next = driver_gpuva_alloc();
523 * if (!next)
524 * ; // unwind previous VA space
525 * // updates, free memory and
526 * // unlock
527 * }
528 *
529 * driver_vm_remap();
530 * drm_gpuva_remap(prev, next, &op->remap);
531 *
532 * if (prev)
533 * drm_gpuva_link(prev, va->vm_bo);
534 * if (next)
535 * drm_gpuva_link(next, va->vm_bo);
536 * drm_gpuva_unlink(va);
537 *
538 * break;
539 * }
540 * case DRM_GPUVA_OP_UNMAP:
541 * va = op->unmap->va;
542 *
543 * driver_vm_unmap();
544 * drm_gpuva_unlink(va);
545 * drm_gpuva_unmap(&op->unmap);
546 *
547 * break;
548 * default:
549 * break;
550 * }
551 * }
552 * drm_gpuvm_bo_put(vm_bo);
553 * driver_unlock_va_space();
554 *
555 * return 0;
556 * }
557 *
558 * 2) Receive a callback for each &drm_gpuva_op to create a new mapping::
559 *
560 * struct driver_context {
561 * struct drm_gpuvm *gpuvm;
562 * struct drm_gpuvm_bo *vm_bo;
563 * struct drm_gpuva *new_va;
564 * struct drm_gpuva *prev_va;
565 * struct drm_gpuva *next_va;
566 * };
567 *
568 * // ops to pass to drm_gpuvm_init()
569 * static const struct drm_gpuvm_ops driver_gpuvm_ops = {
570 * .sm_step_map = driver_gpuva_map,
571 * .sm_step_remap = driver_gpuva_remap,
572 * .sm_step_unmap = driver_gpuva_unmap,
573 * };
574 *
575 * // Typically drivers would embedd the &drm_gpuvm and &drm_gpuva
576 * // structure in individual driver structures and lock the dma-resv with
577 * // drm_exec or similar helpers.
578 * int driver_mapping_create(struct drm_gpuvm *gpuvm,
579 * u64 addr, u64 range,
580 * struct drm_gem_object *obj, u64 offset)
581 * {
582 * struct driver_context ctx;
583 * struct drm_gpuvm_bo *vm_bo;
584 * struct drm_gpuva_ops *ops;
585 * struct drm_gpuva_op *op;
586 * int ret = 0;
587 *
588 * ctx.gpuvm = gpuvm;
589 *
590 * ctx.new_va = kzalloc(sizeof(*ctx.new_va), GFP_KERNEL);
591 * ctx.prev_va = kzalloc(sizeof(*ctx.prev_va), GFP_KERNEL);
592 * ctx.next_va = kzalloc(sizeof(*ctx.next_va), GFP_KERNEL);
593 * ctx.vm_bo = drm_gpuvm_bo_create(gpuvm, obj);
594 * if (!ctx.new_va || !ctx.prev_va || !ctx.next_va || !vm_bo) {
595 * ret = -ENOMEM;
596 * goto out;
597 * }
598 *
599 * // Typically protected with a driver specific GEM gpuva lock
600 * // used in the fence signaling path for drm_gpuva_link() and
601 * // drm_gpuva_unlink(), hence pre-allocate.
602 * ctx.vm_bo = drm_gpuvm_bo_obtain_prealloc(ctx.vm_bo);
603 *
604 * driver_lock_va_space();
605 * ret = drm_gpuvm_sm_map(gpuvm, &ctx, addr, range, obj, offset);
606 * driver_unlock_va_space();
607 *
608 * out:
609 * drm_gpuvm_bo_put(ctx.vm_bo);
610 * kfree(ctx.new_va);
611 * kfree(ctx.prev_va);
612 * kfree(ctx.next_va);
613 * return ret;
614 * }
615 *
616 * int driver_gpuva_map(struct drm_gpuva_op *op, void *__ctx)
617 * {
618 * struct driver_context *ctx = __ctx;
619 *
620 * drm_gpuva_map(ctx->vm, ctx->new_va, &op->map);
621 *
622 * drm_gpuva_link(ctx->new_va, ctx->vm_bo);
623 *
624 * // prevent the new GPUVA from being freed in
625 * // driver_mapping_create()
626 * ctx->new_va = NULL;
627 *
628 * return 0;
629 * }
630 *
631 * int driver_gpuva_remap(struct drm_gpuva_op *op, void *__ctx)
632 * {
633 * struct driver_context *ctx = __ctx;
634 * struct drm_gpuva *va = op->remap.unmap->va;
635 *
636 * drm_gpuva_remap(ctx->prev_va, ctx->next_va, &op->remap);
637 *
638 * if (op->remap.prev) {
639 * drm_gpuva_link(ctx->prev_va, va->vm_bo);
640 * ctx->prev_va = NULL;
641 * }
642 *
643 * if (op->remap.next) {
644 * drm_gpuva_link(ctx->next_va, va->vm_bo);
645 * ctx->next_va = NULL;
646 * }
647 *
648 * drm_gpuva_unlink(va);
649 * kfree(va);
650 *
651 * return 0;
652 * }
653 *
654 * int driver_gpuva_unmap(struct drm_gpuva_op *op, void *__ctx)
655 * {
656 * drm_gpuva_unlink(op->unmap.va);
657 * drm_gpuva_unmap(&op->unmap);
658 * kfree(op->unmap.va);
659 *
660 * return 0;
661 * }
662 */
663
664 /**
665 * get_next_vm_bo_from_list() - get the next vm_bo element
666 * @__gpuvm: The GPU VM
667 * @__list_name: The name of the list we're iterating on
668 * @__local_list: A pointer to the local list used to store already iterated items
669 * @__prev_vm_bo: The previous element we got from drm_gpuvm_get_next_cached_vm_bo()
670 *
671 * This helper is here to provide lockless list iteration. Lockless as in, the
672 * iterator releases the lock immediately after picking the first element from
673 * the list, so list insertion deletion can happen concurrently.
674 *
675 * Elements popped from the original list are kept in a local list, so removal
676 * and is_empty checks can still happen while we're iterating the list.
677 */
678 #define get_next_vm_bo_from_list(__gpuvm, __list_name, __local_list, __prev_vm_bo) \
679 ({ \
680 struct drm_gpuvm_bo *__vm_bo; \
681 \
682 drm_gpuvm_bo_put(__prev_vm_bo); \
683 \
684 spin_lock(&(__gpuvm)->__list_name.lock); \
685 while (!list_empty(&(__gpuvm)->__list_name.list)) { \
686 __vm_bo = list_first_entry(&(__gpuvm)->__list_name.list, \
687 struct drm_gpuvm_bo, \
688 list.entry.__list_name); \
689 if (drm_gpuvm_bo_get_unless_zero(__vm_bo)) { \
690 list_move_tail(&(__vm_bo)->list.entry.__list_name, \
691 __local_list); \
692 break; \
693 } else { \
694 list_del_init(&(__vm_bo)->list.entry.__list_name); \
695 __vm_bo = NULL; \
696 } \
697 } \
698 spin_unlock(&(__gpuvm)->__list_name.lock); \
699 \
700 __vm_bo; \
701 })
702
703 /**
704 * for_each_vm_bo_in_list() - internal vm_bo list iterator
705 *
706 * This helper is here to provide lockless list iteration. Lockless as in, the
707 * iterator releases the lock immediately after picking the first element from the
708 * list, so list insertion and deletion can happen concurrently.
709 *
710 * Typical use:
711 *
712 * struct drm_gpuvm_bo *vm_bo;
713 * LIST_HEAD(my_local_list);
714 *
715 * ret = 0;
716 * drm_gpuvm_for_each_vm_bo(gpuvm, <list_name>, &my_local_list, vm_bo) {
717 * ret = do_something_with_vm_bo(..., vm_bo);
718 * if (ret)
719 * break;
720 * }
721 * drm_gpuvm_bo_put(vm_bo);
722 * drm_gpuvm_restore_vm_bo_list(gpuvm, <list_name>, &my_local_list);
723 *
724 *
725 * Only used for internal list iterations, not meant to be exposed to the outside
726 * world.
727 */
728 #define for_each_vm_bo_in_list(__gpuvm, __list_name, __local_list, __vm_bo) \
729 for (__vm_bo = get_next_vm_bo_from_list(__gpuvm, __list_name, \
730 __local_list, NULL); \
731 __vm_bo; \
732 __vm_bo = get_next_vm_bo_from_list(__gpuvm, __list_name, \
733 __local_list, __vm_bo)) \
> 734 Signature · 2 lines
0-DAY CI Kernel Test Service https://github.com/intel/lkp-tests/wiki
Re: [PATCH drm-misc-next v3 6/7] drm/gpuvm: generalize dma_resv/extobj handling and GEM validation
In reply to Danilo KrummrichHello Danilo,
Quoted 15 lines · On Sat, 9 Sep 2023 17:31:13 +0200 Danilo Krummrich <[email protected]> wrote:
drivers/gpu/drm/drm_gpuvm.c +11
-0@@ -632,6 +661,131 @@* } */ +/** +* get_next_vm_bo_from_list() - get the next vm_bo element +* @__gpuvm: The GPU VM +* @__list_name: The name of the list we're iterating on +* @__local_list: A pointer to the local list used to store already iterated items +* @__prev_vm_bo: The previous element we got from drm_gpuvm_get_next_cached_vm_bo() +* +* This helper is here to provide lockless list iteration. Lockless as in, the +* iterator releases the lock immediately after picking the first element from +* the list, so list insertion deletion can happen concurrently. +*
drivers/gpu/drm/drm_gpuvm.c line 684
Quoted 10 lines
drivers/gpu/drm/drm_gpuvm.c +10
-0+* Elements popped from the original list are kept in a local list, so removal +* and is_empty checks can still happen while we're iterating the list. +*/ +#define get_next_vm_bo_from_list (__gpuvm, __list_name, __local_list, __prev_vm_bo) \ + ({ \ + struct drm_gpuvm_bo *__vm_bo; \ + \ + drm_gpuvm_bo_put(__prev_vm_bo); \ + \ + spin_lock(&(__gpuvm)->__list_name.lock); \
I'm tempted to add a drm_gpuvm::<list_name>::local_list field, so we can catch concurrent iterations with something like:
if (!(__gpuvm)->__list_name.local_list) (__gpuvm)->__list_name.local_list = __local_list; else WARN_ON((__gpuvm)->__list_name.local_list != __local_list);
with (__gpuvm)->__list_name.local_list being restored to NULL in restore_vm_bo_list().
Quoted 28 lines
drivers/gpu/drm/drm_gpuvm.c +28
-0+while (!list_empty (&(__gpuvm)->__list_name.list)) { \ + __vm_bo =list_first_entry (&(__gpuvm)->__list_name.list, \ + struct drm_gpuvm_bo, \ + list.entry.__list_name); \ +if (drm_gpuvm_bo_get_unless_zero (__vm_bo)) { \ +list_move_tail (&(__vm_bo)->list.entry.__list_name, \ + __local_list); \ +break ; \ + }else { \ +list_del_init (&(__vm_bo)->list.entry.__list_name); \ + __vm_bo =NULL ; \ + } \ + } \ +spin_unlock (&(__gpuvm)->__list_name.lock); \ + \ + __vm_bo; \ + }) + +/** +* for_each_vm_bo_in_list() - internal vm_bo list iterator +* +* This helper is here to provide lockless list iteration. Lockless as in, the +* iterator releases the lock immediately after picking the first element from the +* list, so list insertion and deletion can happen concurrently. +* +* Typical use: +* +* struct drm_gpuvm_bo *vm_bo;
drivers/gpu/drm/drm_gpuvm.c line 722
Quoted 10 lines
drivers/gpu/drm/drm_gpuvm.c +10
-0+* LIST_HEAD(my_local_list); +* +* ret = 0; +* drm_gpuvm_for_each_vm_bo(gpuvm, <list_name>, &my_local_list, vm_bo) { +* ret = do_something_with_vm_bo(..., vm_bo); +* if (ret) +* break; +* } +* drm_gpuvm_bo_put(vm_bo); +* drm_gpuvm_restore_vm_bo_list(gpuvm, <list_name>, &my_local_list);
The names in this example and the helper names don't match.
Quoted 55 lines
drivers/gpu/drm/drm_gpuvm.c +55
-0+* +* +* Only used for internal list iterations, not meant to be exposed to the outside +* world. +*/ +#define for_each_vm_bo_in_list (__gpuvm, __list_name, __local_list, __vm_bo) \ + for (__vm_bo = get_next_vm_bo_from_list(__gpuvm, __list_name, \ + __local_list, NULL); \ + __vm_bo; \ + __vm_bo = get_next_vm_bo_from_list(__gpuvm, __list_name, \ + __local_list, __vm_bo)) \ + +/** +* restore_vm_bo_list() - move vm_bo elements back to their original list +* @__gpuvm: The GPU VM +* @__list_name: The name of the list we're iterating on +* @__local_list: A pointer to the local list used to store already iterated items +* +* When we're done iterating a vm_bo list, we should call restore_vm_bo_list() +* to restore the original state and let new iterations take place. +*/ +#define restore_vm_bo_list (__gpuvm, __list_name, __local_list) \ + do { \ +/* Merge back the two lists, moving local list elements to the \ +* head to preserve previous ordering, in case it matters. \ +*/ \ +spin_lock (&(__gpuvm)->__list_name.lock); \ +list_splice (__local_list, &(__gpuvm)->__list_name.list); \ +spin_unlock (&(__gpuvm)->__list_name.lock); \ + }while (0 ) +/** +* drm_gpuvm_bo_list_add() - insert a vm_bo into the given list +* @__vm_bo: the &drm_gpuvm_bo +* @__list_name: the name of the list to insert into +* +* Inserts the given @__vm_bo into the list specified by @__list_name and +* increases the vm_bo's reference count. + */ +#define drm_gpuvm_bo_list_add (__vm_bo, __list_name) \ + do { \ + spin_lock(&(__vm_bo)->vm->__list_name.lock); \ + if (list_empty(&(__vm_bo)->list.entry.__list_name)) \ + list_add_tail(&(__vm_bo)->list.entry.__list_name, \ + &(__vm_bo)->vm->__list_name.list); \ + spin_unlock(&(__vm_bo)->vm->__list_name.lock); \ + } while (0) + +/** +* drm_gpuvm_bo_list_del() - remove a vm_bo from the given list +* @__vm_bo: the &drm_gpuvm_bo +* @__list_name: the name of the list to insert into +* +* Removes the given @__vm_bo from the list specified by @__list_name and +* decreases the vm_bo's reference count. +*/
drivers/gpu/drm/drm_gpuvm.c line 787
Quoted 10 lines
drivers/gpu/drm/drm_gpuvm.c +10
-0+#define drm_gpuvm_bo_list_del (__vm_bo, __list_name) \ + do { \ + spin_lock(&(__vm_bo)->vm->__list_name.lock); \ + if (!list_empty(&(__vm_bo)->list.entry.__list_name)) \ + list_del_init(&(__vm_bo)->list.entry.__list_name); \ + spin_unlock(&(__vm_bo)->vm->__list_name.lock); \ + } while (0) + +static int __must_check +drm_gpuvm_bo_get_unless_zero (struct drm_gpuvm_bo *vm_bo);
I see no obvious reason to have a forward declaration for this helper, if we decide to keep it, let's at least move the declaration here.
Quoted 16 lines
drivers/gpu/drm/drm_gpuvm.c +8
-0@@ -807,6 +1262,14 @@ drm_gpuvm_bo_destroy(struct kref *kref)drm_gem_gpuva_assert_lock_held (vm_bo->obj); +spin_lock (&gpuvm->extobj.lock); +list_del (&vm_bo->list.entry.extobj); +spin_unlock (&gpuvm->extobj.lock); + +spin_lock (&gpuvm->evict.lock); +list_del (&vm_bo->list.entry.evict); +spin_unlock (&gpuvm->evict.lock); +list_del (&vm_bo->list.entry.gem);drm_gem_object_put (obj);
drivers/gpu/drm/drm_gpuvm.c drm_gpuvm_bo_destroy(struct kref *kref) · line 1292
Quoted 8 lines
drivers/gpu/drm/drm_gpuvm.c +5
-0@@ -822,6 +1285,11 @@ drm_gpuvm_bo_destroy(struct kref *kref)* @vm_bo: the &drm_gpuvm_bo to release the reference of * * This releases a reference to @vm_bo. +* +* If the reference count drops to zero, the &gpuvm_bo is destroyed, which +* includes removing it from the GEMs gpuva list. Hence, if a call to this +* function can potentially let the reference count to zero the caller must +* hold the dma-resv or driver specific GEM gpuva lock.
Looks like this should have been part of the previous patch. I hate the fact we have to worry about GEM gpuva lock being held when we call _put() only if the ref drops to zero though. I think I'd feel more comfortable if the function was named differently. Maybe _return() or _release() to match the _obtain() function, where the object is inserted in the GEM vm_bo list. I would also do the lock_is_held() check unconditionally, move the list removal in this function with a del_init(), and have a WARN_ON(!list_empty) in vm_bo_destroy().
drivers/gpu/drm/drm_gpuvm.c drm_gpuvm_bo_put(struct drm_gpuvm_bo *vm_bo) · line 1305
Quoted 11 lines
drivers/gpu/drm/drm_gpuvm.c +4
-0*/ void drm_gpuvm_bo_put (struct drm_gpuvm_bo *vm_bo)@@ -831,6 +1299,12 @@ drm_gpuvm_bo_put(struct drm_gpuvm_bo *vm_bo) }EXPORT_SYMBOL_GPL (drm_gpuvm_bo_put); +static int __must_check +drm_gpuvm_bo_get_unless_zero (struct drm_gpuvm_bo *vm_bo) +{ +return kref_get_unless_zero (&vm_bo->kref);
Not convinced this helper is needed. It's only used once, and I don't think we'll need it elsewhere.
drivers/gpu/drm/drm_gpuvm.c drm_gpuvm_bo_put(struct drm_gpuvm_bo *vm_bo) · line 1310
Quoted 5 lines
drivers/gpu/drm/drm_gpuvm.c +2
-0+} +static struct drm_gpuvm_bo *__drm_gpuvm_bo_find (struct drm_gpuvm *gpuvm,struct drm_gem_object *obj)
Regards,
Boris
Re: [PATCH drm-misc-next v3 6/7] drm/gpuvm: generalize dma_resv/extobj handling and GEM validation
Quoted 179 lines · Boris Brezillon
Hello Danilo,
Quoted 15 lines · On Sat, 9 Sep 2023 17:31:13 +0200 Danilo Krummrich <[email protected]> wrote:
drivers/gpu/drm/drm_gpuvm.c +11
-0@@ -632,6 +661,131 @@* } */ +/** +* get_next_vm_bo_from_list() - get the next vm_bo element +* @__gpuvm: The GPU VM +* @__list_name: The name of the list we're iterating on +* @__local_list: A pointer to the local list used to store already iterated items +* @__prev_vm_bo: The previous element we got from drm_gpuvm_get_next_cached_vm_bo() +* +* This helper is here to provide lockless list iteration. Lockless as in, the +* iterator releases the lock immediately after picking the first element from +* the list, so list insertion deletion can happen concurrently. +* drivers/gpu/drm/drm_gpuvm.c line 684
Quoted 10 lines
drivers/gpu/drm/drm_gpuvm.c +10
-0+* Elements popped from the original list are kept in a local list, so removal +* and is_empty checks can still happen while we're iterating the list. +*/ +#define get_next_vm_bo_from_list (__gpuvm, __list_name, __local_list, __prev_vm_bo) \ + ({ \ + struct drm_gpuvm_bo *__vm_bo; \ + \ + drm_gpuvm_bo_put(__prev_vm_bo); \ + \ + spin_lock(&(__gpuvm)->__list_name.lock); \I'm tempted to add a drm_gpuvm::<list_name>::local_list field, so we can catch concurrent iterations with something like:
if (!(__gpuvm)->__list_name.local_list) (__gpuvm)->__list_name.local_list = __local_list; else WARN_ON((__gpuvm)->__list_name.local_list != __local_list);with (__gpuvm)->__list_name.local_list being restored to NULL in
restore_vm_bo_list().Quoted 28 lines
drivers/gpu/drm/drm_gpuvm.c +28
-0+while (!list_empty (&(__gpuvm)->__list_name.list)) { \ + __vm_bo =list_first_entry (&(__gpuvm)->__list_name.list, \ + struct drm_gpuvm_bo, \ + list.entry.__list_name); \ +if (drm_gpuvm_bo_get_unless_zero (__vm_bo)) { \ +list_move_tail (&(__vm_bo)->list.entry.__list_name, \ + __local_list); \ +break ; \ + }else { \ +list_del_init (&(__vm_bo)->list.entry.__list_name); \ + __vm_bo =NULL ; \ + } \ + } \ +spin_unlock (&(__gpuvm)->__list_name.lock); \ + \ + __vm_bo; \ + }) + +/** +* for_each_vm_bo_in_list() - internal vm_bo list iterator +* +* This helper is here to provide lockless list iteration. Lockless as in, the +* iterator releases the lock immediately after picking the first element from the +* list, so list insertion and deletion can happen concurrently. +* +* Typical use: +* +* struct drm_gpuvm_bo *vm_bo; drivers/gpu/drm/drm_gpuvm.c line 722
Quoted 10 lines
drivers/gpu/drm/drm_gpuvm.c +10
-0+* LIST_HEAD(my_local_list); +* +* ret = 0; +* drm_gpuvm_for_each_vm_bo(gpuvm, <list_name>, &my_local_list, vm_bo) { +* ret = do_something_with_vm_bo(..., vm_bo); +* if (ret) +* break; +* } +* drm_gpuvm_bo_put(vm_bo); +* drm_gpuvm_restore_vm_bo_list(gpuvm, <list_name>, &my_local_list); The names in this example and the helper names don't match.
Quoted 55 lines
drivers/gpu/drm/drm_gpuvm.c +55
-0+* +* +* Only used for internal list iterations, not meant to be exposed to the outside +* world. +*/ +#define for_each_vm_bo_in_list (__gpuvm, __list_name, __local_list, __vm_bo) \ + for (__vm_bo = get_next_vm_bo_from_list(__gpuvm, __list_name, \ + __local_list, NULL); \ + __vm_bo; \ + __vm_bo = get_next_vm_bo_from_list(__gpuvm, __list_name, \ + __local_list, __vm_bo)) \ + +/** +* restore_vm_bo_list() - move vm_bo elements back to their original list +* @__gpuvm: The GPU VM +* @__list_name: The name of the list we're iterating on +* @__local_list: A pointer to the local list used to store already iterated items +* +* When we're done iterating a vm_bo list, we should call restore_vm_bo_list() +* to restore the original state and let new iterations take place. +*/ +#define restore_vm_bo_list (__gpuvm, __list_name, __local_list) \ + do { \ +/* Merge back the two lists, moving local list elements to the \ +* head to preserve previous ordering, in case it matters. \ +*/ \ +spin_lock (&(__gpuvm)->__list_name.lock); \ +list_splice (__local_list, &(__gpuvm)->__list_name.list); \ +spin_unlock (&(__gpuvm)->__list_name.lock); \ + }while (0 ) +/** +* drm_gpuvm_bo_list_add() - insert a vm_bo into the given list +* @__vm_bo: the &drm_gpuvm_bo +* @__list_name: the name of the list to insert into +* +* Inserts the given @__vm_bo into the list specified by @__list_name and +* increases the vm_bo's reference count. + */ +#define drm_gpuvm_bo_list_add (__vm_bo, __list_name) \ + do { \ + spin_lock(&(__vm_bo)->vm->__list_name.lock); \ + if (list_empty(&(__vm_bo)->list.entry.__list_name)) \ + list_add_tail(&(__vm_bo)->list.entry.__list_name, \ + &(__vm_bo)->vm->__list_name.list); \ + spin_unlock(&(__vm_bo)->vm->__list_name.lock); \ + } while (0) + +/** +* drm_gpuvm_bo_list_del() - remove a vm_bo from the given list +* @__vm_bo: the &drm_gpuvm_bo +* @__list_name: the name of the list to insert into +* +* Removes the given @__vm_bo from the list specified by @__list_name and +* decreases the vm_bo's reference count. +*/ drivers/gpu/drm/drm_gpuvm.c line 787
Quoted 10 lines
drivers/gpu/drm/drm_gpuvm.c +10
-0+#define drm_gpuvm_bo_list_del (__vm_bo, __list_name) \ + do { \ + spin_lock(&(__vm_bo)->vm->__list_name.lock); \ + if (!list_empty(&(__vm_bo)->list.entry.__list_name)) \ + list_del_init(&(__vm_bo)->list.entry.__list_name); \ + spin_unlock(&(__vm_bo)->vm->__list_name.lock); \ + } while (0) + +static int __must_check +drm_gpuvm_bo_get_unless_zero (struct drm_gpuvm_bo *vm_bo);I see no obvious reason to have a forward declaration for this helper, if we decide to keep it, let's at least move the declaration here.
Quoted 16 lines
drivers/gpu/drm/drm_gpuvm.c +8
-0@@ -807,6 +1262,14 @@ drm_gpuvm_bo_destroy(struct kref *kref)drm_gem_gpuva_assert_lock_held (vm_bo->obj); +spin_lock (&gpuvm->extobj.lock); +list_del (&vm_bo->list.entry.extobj); +spin_unlock (&gpuvm->extobj.lock); + +spin_lock (&gpuvm->evict.lock); +list_del (&vm_bo->list.entry.evict); +spin_unlock (&gpuvm->evict.lock); +list_del (&vm_bo->list.entry.gem);drm_gem_object_put (obj);drivers/gpu/drm/drm_gpuvm.c drm_gpuvm_bo_destroy(struct kref *kref) · line 1292
Quoted 8 lines
drivers/gpu/drm/drm_gpuvm.c +5
-0@@ -822,6 +1285,11 @@ drm_gpuvm_bo_destroy(struct kref *kref)* @vm_bo: the &drm_gpuvm_bo to release the reference of * * This releases a reference to @vm_bo. +* +* If the reference count drops to zero, the &gpuvm_bo is destroyed, which +* includes removing it from the GEMs gpuva list. Hence, if a call to this +* function can potentially let the reference count to zero the caller must +* hold the dma-resv or driver specific GEM gpuva lock.
Quoted 9 lines
Looks like this should have been part of the previous patch. I hate the fact we have to worry about GEM gpuva lock being held when we call
_put()only if the ref drops to zero though. I think I'd feel more comfortable if the function was named differently. Maybe_return()or_release()to match the_obtain()function, where the object is inserted in the GEMvm_bolist. I would also do thelock_is_held()check unconditionally, move the list removal in this function with adel_init(), and have aWARN_ON(!list_empty)invm_bo_destroy().
We can't move the list removal to drm_gpuvm_bo_put(), we need to make sure we can't create duplicate drm_gpuvm_bo structures. Everything else pretty much goes away with a dedicated GEM gpuva list lock, as I had in my first patch series when I introduced the GPUVA manager. At that time it wasn't always needed, hence the optional driver specific lock, however with the VM_BO abstraction it really makes sense to have a dedicated one.
I agree with the other feedback from this reply and will address it in a V4.
Quoted 26 lines
drivers/gpu/drm/drm_gpuvm.c drm_gpuvm_bo_put(struct drm_gpuvm_bo *vm_bo) · line 1305
Quoted 11 lines
drivers/gpu/drm/drm_gpuvm.c +4
-0*/ void drm_gpuvm_bo_put (struct drm_gpuvm_bo *vm_bo)@@ -831,6 +1299,12 @@ drm_gpuvm_bo_put(struct drm_gpuvm_bo *vm_bo) }EXPORT_SYMBOL_GPL (drm_gpuvm_bo_put); +static int __must_check +drm_gpuvm_bo_get_unless_zero (struct drm_gpuvm_bo *vm_bo) +{ +return kref_get_unless_zero (&vm_bo->kref);Not convinced this helper is needed. It's only used once, and I don't think we'll need it elsewhere.
drivers/gpu/drm/drm_gpuvm.c drm_gpuvm_bo_put(struct drm_gpuvm_bo *vm_bo) · line 1310
Quoted 5 lines
drivers/gpu/drm/drm_gpuvm.c +2
-0+} +static struct drm_gpuvm_bo *__drm_gpuvm_bo_find (struct drm_gpuvm *gpuvm,struct drm_gem_object *obj)Regards,
Boris
Re: [PATCH drm-misc-next v3 6/7] drm/gpuvm: generalize dma_resv/extobj handling and GEM validation
In reply to Danilo KrummrichQuoted 7 lines · On Sat, 9 Sep 2023 17:31:13 +0200 Danilo Krummrich <[email protected]> wrote:
drivers/gpu/drm/drm_gpuvm.c +7
-0+/** +* get_next_vm_bo_from_list() - get the next vm_bo element +* @__gpuvm: The GPU VM +* @__list_name: The name of the list we're iterating on +* @__local_list: A pointer to the local list used to store already iterated items +* @__prev_vm_bo: The previous element we got from drm_gpuvm_get_next_cached_vm_bo() +*
drivers/gpu/drm/drm_gpuvm.c line 680
Quoted 10 lines
drivers/gpu/drm/drm_gpuvm.c +10
-0+* This helper is here to provide lockless list iteration. Lockless as in, the +* iterator releases the lock immediately after picking the first element from +* the list, so list insertion deletion can happen concurrently. +* +* Elements popped from the original list are kept in a local list, so removal +* and is_empty checks can still happen while we're iterating the list. +*/ +#define get_next_vm_bo_from_list (__gpuvm, __list_name, __local_list, __prev_vm_bo) \ + ({ \ + struct drm_gpuvm_bo *__vm_bo; \
Missing NULL assignment here.
Quoted 32 lines
drivers/gpu/drm/drm_gpuvm.c +32
-0+ \ +drm_gpuvm_bo_put (__prev_vm_bo); \ + \ +spin_lock (&(__gpuvm)->__list_name.lock); \ +while (!list_empty (&(__gpuvm)->__list_name.list)) { \ + __vm_bo =list_first_entry (&(__gpuvm)->__list_name.list, \ + struct drm_gpuvm_bo, \ + list.entry.__list_name); \ +if (drm_gpuvm_bo_get_unless_zero (__vm_bo)) { \ +list_move_tail (&(__vm_bo)->list.entry.__list_name, \ + __local_list); \ +break ; \ + }else { \ +list_del_init (&(__vm_bo)->list.entry.__list_name); \ + __vm_bo =NULL ; \ + } \ + } \ +spin_unlock (&(__gpuvm)->__list_name.lock); \ + \ + __vm_bo; \ + }) + +/** +* for_each_vm_bo_in_list() - internal vm_bo list iterator +* +* This helper is here to provide lockless list iteration. Lockless as in, the +* iterator releases the lock immediately after picking the first element from the +* list, so list insertion and deletion can happen concurrently. +* +* Typical use: +* +* struct drm_gpuvm_bo *vm_bo;
drivers/gpu/drm/drm_gpuvm.c line 722
Quoted 10 lines
drivers/gpu/drm/drm_gpuvm.c +10
-0+* LIST_HEAD(my_local_list); +* +* ret = 0; +* drm_gpuvm_for_each_vm_bo(gpuvm, <list_name>, &my_local_list, vm_bo) { +* ret = do_something_with_vm_bo(..., vm_bo); +* if (ret) +* break; +* } +* drm_gpuvm_bo_put(vm_bo); +* drm_gpuvm_restore_vm_bo_list(gpuvm, <list_name>, &my_local_list);
Might be worth mentioning that the vm_bo pointer shouldn't be re-assigned from inside for loop, otherwise
the next get_next_vm_bo_from_list() will be passed a wrong prev_vm_bo.
Quoted 11 lines
+ *
+ *
+ * Only used for internal list iterations, not meant to be exposed to the outside
+ * world.
+ */drivers/gpu/drm/drm_gpuvm.c +6
-0+#define for_each_vm_bo_in_list (__gpuvm, __list_name, __local_list, __vm_bo) \ + for (__vm_bo = get_next_vm_bo_from_list(__gpuvm, __list_name, \ + __local_list, NULL); \ + __vm_bo; \ + __vm_bo = get_next_vm_bo_from_list(__gpuvm, __list_name, \ + __local_list, __vm_bo)) \
Re: [PATCH drm-misc-next v3 6/7] drm/gpuvm: generalize dma_resv/extobj handling and GEM validation
In reply to Danilo KrummrichQuoted 5 lines · On Sat, 9 Sep 2023 17:31:13 +0200 Danilo Krummrich <[email protected]> wrote:
drivers/gpu/drm/drm_gpuvm.c +1
-0@@ -807,6 +1262,14 @@ drm_gpuvm_bo_destroy(struct kref *kref)drm_gem_gpuva_assert_lock_held (vm_bo->obj); +spin_lock (&gpuvm->extobj.lock);
drivers/gpu/drm/drm_gpuvm.c drm_gpuvm_bo_destroy(struct kref *kref) · line 1275
Quoted 10 lines
drivers/gpu/drm/drm_gpuvm.c +7
-0+list_del (&vm_bo->list.entry.extobj); +spin_unlock (&gpuvm->extobj.lock); + +spin_lock (&gpuvm->evict.lock); +list_del (&vm_bo->list.entry.evict); +spin_unlock (&gpuvm->evict.lock); +list_del (&vm_bo->list.entry.gem);drm_gem_object_put (obj);
I ran into a UAF situation when the drm_gpuvm_bo object is the last owner of obj, because the lock that's supposed to be held when calling this function (drm_gem_gpuva_assert_lock_held() call above), belongs to obj (either obj->resv, or a driver specific lock that's attached to the driver-specific GEM object). I worked around it by taking a ref to obj before calling lock()+drm_gpuvm_bo_put()+unlock(), and releasing it after I'm node with the lock, but that just feels wrong.
Re: [PATCH drm-misc-next v3 6/7] drm/gpuvm: generalize dma_resv/extobj handling and GEM validation
Quoted 17 lines · Boris Brezillon
Quoted 14 lines · On Sat, 9 Sep 2023 17:31:13 +0200 Danilo Krummrich <[email protected]> wrote:
drivers/gpu/drm/drm_gpuvm.c +8
-0@@ -807,6 +1262,14 @@ drm_gpuvm_bo_destroy(struct kref *kref)drm_gem_gpuva_assert_lock_held (vm_bo->obj); +spin_lock (&gpuvm->extobj.lock); +list_del (&vm_bo->list.entry.extobj); +spin_unlock (&gpuvm->extobj.lock); + +spin_lock (&gpuvm->evict.lock); +list_del (&vm_bo->list.entry.evict); +spin_unlock (&gpuvm->evict.lock); +list_del (&vm_bo->list.entry.gem);
Quoted 10 lines
drivers/gpu/drm/drm_gpuvm.c drm_gpuvm_bo_destroy(struct kref *kref) · line 1275
Quoted 1 line
drivers/gpu/drm/drm_gpuvm.c +0
-0drm_gem_object_put (obj);I ran into a UAF situation when the
drm_gpuvm_boobject is the last owner of obj, because the lock that's supposed to be held when calling this function (drm_gem_gpuva_assert_lock_held()call above), belongs to obj (eitherobj->resv, or a driver specific lock that's attached to the driver-specific GEM object). I worked around it by taking a ref to obj before calling lock()+drm_gpuvm_bo_put()+unlock(), and releasing it after I'm node with the lock, but that just feels wrong.
As mentioned in a previous reply, I think we want to bring the dedicated GEM gpuva list lock back instead of abusing the dma-resv lock. This way we can handle locking internally and don't run into such issues.
There is also no reason for a driver to already hold the GEM gpuva list lock when when calling drm_gpuvm_bo_put(). Drivers would only acquire the lock to iterate the GEMs list of drm_gpuvm_bos or the drm_gpuvm_bos list of drm_gpuvas. And dropping the drm_gpuvm_bo from within such a loop is forbidden anyways.
Thomas Hellström
reviewerRe: [PATCH drm-misc-next v3 6/7] drm/gpuvm: generalize dma_resv/extobj handling and GEM validation
In reply to Danilo KrummrichHi, Danilo,
Quoted 91 lines · Danilo Krummrich
So far the DRM GPUVA manager offers common infrastructure to track GPU VA allocations and mappings, generically connect GPU VA mappings to their backing buffers and perform more complex mapping operations on the GPU VA space.
However, there are more design patterns commonly used by drivers, which can potentially be generalized in order to make the DRM GPUVA manager represent a basic GPU-VM implementation. In this context, this patch aims at generalizing the following elements.
1) Provide a common dma-resv for GEM objects not being used outside of this GPU-VM.
2) Provide tracking of external GEM objects (GEM objects which are shared with other GPU-VMs).
3) Provide functions to efficiently lock all GEM objects dma-resv the GPU-VM contains mappings of.
4) Provide tracking of evicted GEM objects the GPU-VM contains mappings of, such that validation of evicted GEM objects is accelerated.
5) Provide some convinience functions for common patterns.
Rather than being designed as a "framework", the target is to make all features appear as a collection of optional helper functions, such that drivers are free to make use of the DRM GPUVA managers basic functionality and opt-in for other features without setting any feature flags, just by making use of the corresponding functions.
Big kudos to Boris Brezillon for his help to figure out locking for drivers updating the GPU VA space within the fence signalling path.
Suggested-by Matthew Brost <[email protected]>Signed-off-by Danilo Krummrich <[email protected]>--- drivers/gpu/drm/drm_gpuvm.c | 516 ++++++++++++++++++++++++++++++++++++ include/drm/drm_gpuvm.h | 197 ++++++++++++++ 2 files changed, 713 insertions(+)drivers/gpu/drm/drm_gpuvm.c +29
-0@@ -73,6 +73,21 @@* &drm_gem_object list of &drm_gpuvm_bos for an existing instance of this * particular combination. If not existent a new instance is created and linked * to the &drm_gem_object. +* +* &drm_gpuvm_bo structures, since unique for a given &drm_gpuvm, are also used +* as entry for the &drm_gpuvm's lists of external and evicted objects. Those +* list are maintained in order to accelerate locking of dma-resv locks and +* validation of evicted objects bound in a &drm_gpuvm. For instance the all +* &drm_gem_object's &dma_resv of a given &drm_gpuvm can be locked by calling +* drm_gpuvm_exec_lock(). Once locked drivers can call drm_gpuvm_validate() in +* order to validate all evicted &drm_gem_objects. It is also possible to lock +* additional &drm_gem_objects by providing the corresponding parameters to +* drm_gpuvm_exec_lock() as well as open code the &drm_exec loop while making +* use of helper functions such as drm_gpuvm_prepare_range() or +* drm_gpuvm_prepare_objects(). +* +* Every bound &drm_gem_object is treated as external object when its &dma_resv +* structure is different than the &drm_gpuvm's common &dma_resv structure. */ /** @@ -420,6 +435,20 @@* Subsequent calls to drm_gpuvm_bo_obtain() for the same &drm_gpuvm and * &drm_gem_object must be able to observe previous creations and destructions * of &drm_gpuvm_bos in order to keep instances unique. +* +* The &drm_gpuvm's lists for keeping track of external and evicted objects are +* protected against concurrent insertion / removal and iteration internally. +* +* However, drivers still need ensure to protect concurrent calls to functions +* iterating those lists, such as drm_gpuvm_validate() and +* drm_gpuvm_prepare_objects(). Every such function contains a particular +* comment and lockdep checks if possible. +* +* Functions adding or removing entries from those lists, such as +* drm_gpuvm_bo_evict() or drm_gpuvm_bo_extobj_add() may be called with external +* locks being held, e.g. in order to avoid the corresponding list to be +* (safely) modified while potentially being iternated by other API functions. +* However, this is entirely optional. */ /** @@ -632,6 +661,131 @@* } */
drivers/gpu/drm/drm_gpuvm.c line 673
Quoted 10 lines
drivers/gpu/drm/drm_gpuvm.c +10
-0+/** +* get_next_vm_bo_from_list() - get the next vm_bo element +* @__gpuvm: The GPU VM +* @__list_name: The name of the list we're iterating on +* @__local_list: A pointer to the local list used to store already iterated items +* @__prev_vm_bo: The previous element we got from drm_gpuvm_get_next_cached_vm_bo() +* +* This helper is here to provide lockless list iteration. Lockless as in, the +* iterator releases the lock immediately after picking the first element from +* the list, so list insertion deletion can happen concurrently.
Are the list spinlocks needed for that async state update from within the dma-fence critical section we've discussed previously?
Otherwise it should be sufficient to protect the lists with the gpuvm's resv (or for the extobj list with an outer lock).
If those spinlocks are still needed in some situations, perhaps could we have an option to set them to NULL (Like IIRC the maple tree allows for)?
For such drivers, that would require anybody calling unlink to hold the vm's resv, though.
It seems that with that also the refcount could be make non-atomic.
All in the spirit of the drm locking guidelines "use big locks when possible". Lower level locks only when necessary for performance or locking inversion?
/Thomas
Quoted 787 lines
drivers/gpu/drm/drm_gpuvm.c +477
-0+* +* Elements popped from the original list are kept in a local list, so removal +* and is_empty checks can still happen while we're iterating the list. +*/ +#define get_next_vm_bo_from_list (__gpuvm, __list_name, __local_list, __prev_vm_bo) \ + ({ \ + struct drm_gpuvm_bo *__vm_bo; \ + \ + drm_gpuvm_bo_put(__prev_vm_bo); \ + \ + spin_lock(&(__gpuvm)->__list_name.lock); \ + while (!list_empty(&(__gpuvm)->__list_name.list)) { \ + __vm_bo = list_first_entry(&(__gpuvm)->__list_name.list, \ + struct drm_gpuvm_bo, \ + list.entry.__list_name); \ + if (drm_gpuvm_bo_get_unless_zero(__vm_bo)) { \ + list_move_tail(&(__vm_bo)->list.entry.__list_name, \ + __local_list); \ + break; \ + } else { \ + list_del_init(&(__vm_bo)->list.entry.__list_name); \ + __vm_bo = NULL; \ + } \ + } \ + spin_unlock(&(__gpuvm)->__list_name.lock); \ + \ + __vm_bo; \ + }) + +/** +* for_each_vm_bo_in_list() - internal vm_bo list iterator +* +* This helper is here to provide lockless list iteration. Lockless as in, the +* iterator releases the lock immediately after picking the first element from the +* list, so list insertion and deletion can happen concurrently. +* +* Typical use: +* +* struct drm_gpuvm_bo *vm_bo; +* LIST_HEAD(my_local_list); +* +* ret = 0; +* drm_gpuvm_for_each_vm_bo(gpuvm, <list_name>, &my_local_list, vm_bo) { +* ret = do_something_with_vm_bo(..., vm_bo); +* if (ret) +* break; +* } +* drm_gpuvm_bo_put(vm_bo); +* drm_gpuvm_restore_vm_bo_list(gpuvm, <list_name>, &my_local_list); +* +* +* Only used for internal list iterations, not meant to be exposed to the outside +* world. +*/ +#define for_each_vm_bo_in_list (__gpuvm, __list_name, __local_list, __vm_bo) \ + for (__vm_bo = get_next_vm_bo_from_list(__gpuvm, __list_name, \ + __local_list, NULL); \ + __vm_bo; \ + __vm_bo = get_next_vm_bo_from_list(__gpuvm, __list_name, \ + __local_list, __vm_bo)) \ + +/** +* restore_vm_bo_list() - move vm_bo elements back to their original list +* @__gpuvm: The GPU VM +* @__list_name: The name of the list we're iterating on +* @__local_list: A pointer to the local list used to store already iterated items +* +* When we're done iterating a vm_bo list, we should call restore_vm_bo_list() +* to restore the original state and let new iterations take place. +*/ +#define restore_vm_bo_list (__gpuvm, __list_name, __local_list) \ + do { \ +/* Merge back the two lists, moving local list elements to the \ +* head to preserve previous ordering, in case it matters. \ +*/ \ +spin_lock (&(__gpuvm)->__list_name.lock); \ +list_splice (__local_list, &(__gpuvm)->__list_name.list); \ +spin_unlock (&(__gpuvm)->__list_name.lock); \ + }while (0 ) +/** +* drm_gpuvm_bo_list_add() - insert a vm_bo into the given list +* @__vm_bo: the &drm_gpuvm_bo +* @__list_name: the name of the list to insert into +* +* Inserts the given @__vm_bo into the list specified by @__list_name and +* increases the vm_bo's reference count. + */ +#define drm_gpuvm_bo_list_add (__vm_bo, __list_name) \ + do { \ + spin_lock(&(__vm_bo)->vm->__list_name.lock); \ + if (list_empty(&(__vm_bo)->list.entry.__list_name)) \ + list_add_tail(&(__vm_bo)->list.entry.__list_name, \ + &(__vm_bo)->vm->__list_name.list); \ + spin_unlock(&(__vm_bo)->vm->__list_name.lock); \ + } while (0) + +/** +* drm_gpuvm_bo_list_del() - remove a vm_bo from the given list +* @__vm_bo: the &drm_gpuvm_bo +* @__list_name: the name of the list to insert into +* +* Removes the given @__vm_bo from the list specified by @__list_name and +* decreases the vm_bo's reference count. +*/ +#define drm_gpuvm_bo_list_del (__vm_bo, __list_name) \ + do { \ + spin_lock(&(__vm_bo)->vm->__list_name.lock); \ + if (!list_empty(&(__vm_bo)->list.entry.__list_name)) \ + list_del_init(&(__vm_bo)->list.entry.__list_name); \ + spin_unlock(&(__vm_bo)->vm->__list_name.lock); \ + } while (0) + +static int __must_check +drm_gpuvm_bo_get_unless_zero (struct drm_gpuvm_bo *vm_bo); +#define to_drm_gpuva (__node) container_of((__node), struct drm_gpuva, rb.node)#define GPUVA_START (node) ((node)->va.addr)@@ -713,6 +867,12 @@ drm_gpuvm_init(struct drm_gpuvm *gpuvm, struct drm_device *drm, gpuvm->rb.tree =RB_ROOT_CACHED ;INIT_LIST_HEAD (&gpuvm->rb.list); +INIT_LIST_HEAD (&gpuvm->extobj.list); +spin_lock_init (&gpuvm->extobj.lock); + +INIT_LIST_HEAD (&gpuvm->evict.list); +spin_lock_init (&gpuvm->evict.lock); +drm_gpuva_check_overflow (start_offset, range); gpuvm->mm_start = start_offset; gpuvm->mm_range = range;@@ -754,10 +914,302 @@ drm_gpuvm_destroy(struct drm_gpuvm *gpuvm)WARN (!RB_EMPTY_ROOT (&gpuvm->rb.tree.rb_root),"GPUVA tree is not empty, potentially leaking memory.\n" ); +WARN (!list_empty (&gpuvm->extobj.list),"Extobj list should be empty.\n" ); +WARN (!list_empty (&gpuvm->evict.list),"Evict list should be empty.\n" ); +drm_gem_private_object_fini (&gpuvm->d_obj); }EXPORT_SYMBOL_GPL (drm_gpuvm_destroy); +/** +* drm_gpuvm_prepare_objects() - prepare all assoiciated BOs +* @gpuvm: the &drm_gpuvm +* @exec: the &drm_exec locking context +* @num_fences: the amount of &dma_fences to reserve +* +* Calls drm_exec_prepare_obj() for all &drm_gem_objects the given +* &drm_gpuvm contains mappings of. +* +* Using this function directly, it is the drivers responsibility to call +* drm_exec_init() and drm_exec_fini() accordingly. +* +* Note: This function is safe against concurrent insertion and removal of +* external objects, however it is not safe against concurrent usage itself. +* +* Drivers need to make sure to protect this case with either an outer VM lock +* or by calling drm_gpuvm_prepare_vm() before this function within the +* drm_exec_until_all_locked() loop, such that the GPUVM's dma-resv lock ensures +* mutual exclusion. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_prepare_objects (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +unsigned int num_fences) +{ +struct drm_gpuvm_bo *vm_bo; +LIST_HEAD (extobjs); +int ret =0 ; + +for_each_vm_bo_in_list (gpuvm, extobj, &extobjs, vm_bo) { + ret =drm_exec_prepare_obj (exec, vm_bo->obj, num_fences); +if (ret) +break ; + } +/* Drop ref in case we break out of the loop. */ +drm_gpuvm_bo_put (vm_bo); +restore_vm_bo_list (gpuvm, extobj, &extobjs); + +return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_prepare_objects); + +/** +* drm_gpuvm_prepare_range() - prepare all BOs mapped within a given range +* @gpuvm: the &drm_gpuvm +* @exec: the &drm_exec locking context +* @addr: the start address within the VA space +* @range: the range to iterate within the VA space +* @num_fences: the amount of &dma_fences to reserve +* +* Calls drm_exec_prepare_obj() for all &drm_gem_objects mapped between @addr +* and @addr + @range. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_prepare_range (struct drm_gpuvm *gpuvm,struct drm_exec *exec, +u64 addr,u64 range,unsigned int num_fences) +{ +struct drm_gpuva *va; +u64 end = addr + range; +int ret; + +drm_gpuvm_for_each_va_range (va, gpuvm, addr, end) { +struct drm_gem_object *obj = va->gem.obj; + + ret =drm_exec_prepare_obj (exec, obj, num_fences); +if (ret) +return ret; + } + +return 0 ; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_prepare_range); + +/** +* drm_gpuvm_exec_lock() - lock all dma-resv of all assoiciated BOs +* @vm_exec: the &drm_gpuvm_exec abstraction +* @num_fences: the amount of &dma_fences to reserve +* @interruptible: sleep interruptible if waiting +* +* Acquires all dma-resv locks of all &drm_gem_objects the given +* &drm_gpuvm contains mappings of. +* +* Addionally, when calling this function with struct drm_gpuvm_exec::extra +* being set the driver receives the given @fn callback to lock additional +* dma-resv in the context of the &drm_gpuvm_exec instance. Typically, drivers +* would call drm_exec_prepare_obj() from within this callback. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_exec_lock (struct drm_gpuvm_exec *vm_exec, +unsigned int num_fences, +bool interruptible) +{ +struct drm_gpuvm *gpuvm = vm_exec->vm; +struct drm_exec *exec = &vm_exec->exec; +uint32_t flags; +int ret; + + flags = interruptible ?DRM_EXEC_INTERRUPTIBLE_WAIT :0 | +DRM_EXEC_IGNORE_DUPLICATES ; + +drm_exec_init (exec, flags); + +drm_exec_until_all_locked (exec) { + ret =drm_gpuvm_prepare_vm (gpuvm, exec, num_fences); +drm_exec_retry_on_contention (exec); +if (ret) + goto err; + + ret =drm_gpuvm_prepare_objects (gpuvm, exec, num_fences); +drm_exec_retry_on_contention (exec); +if (ret) + goto err; + +if (vm_exec->extra.fn) { + ret = vm_exec->extra.fn (vm_exec, num_fences); +drm_exec_retry_on_contention (exec); +if (ret) + goto err; + } + } + +return 0 ; + +err: +drm_exec_fini (exec); +return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock); + +static int +fn_lock_array (struct drm_gpuvm_exec *vm_exec,unsigned int num_fences) +{ +struct { +struct drm_gem_object **objs; +unsigned int num_objs; + } *args = vm_exec->extra.priv; + +return drm_exec_prepare_array (&vm_exec->exec, args->objs, + args->num_objs, num_fences); +} + +/** +* drm_gpuvm_exec_lock_array() - lock all dma-resv of all assoiciated BOs +* @vm_exec: the &drm_gpuvm_exec abstraction +* @objs: additional &drm_gem_objects to lock +* @num_objs: the number of additional &drm_gem_objects to lock +* @num_fences: the amount of &dma_fences to reserve +* @interruptible: sleep interruptible if waiting +* +* Acquires all dma-resv locks of all &drm_gem_objects the given &drm_gpuvm +* contains mappings of, plus the ones given through @objs. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_exec_lock_array (struct drm_gpuvm_exec *vm_exec, +struct drm_gem_object **objs, +unsigned int num_objs, +unsigned int num_fences, +bool interruptible) +{ +struct { +struct drm_gem_object **objs; +unsigned int num_objs; + } args; + + args.objs = objs; + args.num_objs = num_objs; + + vm_exec->extra.fn = fn_lock_array; + vm_exec->extra.priv = &args; + +return drm_gpuvm_exec_lock (vm_exec, num_fences, interruptible); +} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock_array); + +/** +* drm_gpuvm_exec_lock_range() - prepare all BOs mapped within a given range +* @vm_exec: the &drm_gpuvm_exec abstraction +* @addr: the start address within the VA space +* @range: the range to iterate within the VA space +* @num_fences: the amount of &dma_fences to reserve +* @interruptible: sleep interruptible if waiting +* +* Acquires all dma-resv locks of all &drm_gem_objects mapped between @addr and +* @addr + @range. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_exec_lock_range (struct drm_gpuvm_exec *vm_exec, +u64 addr,u64 range, +unsigned int num_fences, +bool interruptible) +{ +struct drm_gpuvm *gpuvm = vm_exec->vm; +struct drm_exec *exec = &vm_exec->exec; +uint32_t flags; +int ret; + + flags = interruptible ?DRM_EXEC_INTERRUPTIBLE_WAIT :0 | +DRM_EXEC_IGNORE_DUPLICATES ; + +drm_exec_init (exec, flags); + +drm_exec_until_all_locked (exec) { + ret =drm_gpuvm_prepare_range (gpuvm, exec, addr, range, + num_fences); +drm_exec_retry_on_contention (exec); +if (ret) + goto err; + } + +return ret; + +err: +drm_exec_fini (exec); +return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock_range); + +/** +* drm_gpuvm_validate() - validate all BOs marked as evicted +* @gpuvm: the &drm_gpuvm to validate evicted BOs +* +* Calls the &drm_gpuvm_ops.bo_validate callback for all evicted buffer +* objects being mapped in the given &drm_gpuvm. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_validate (struct drm_gpuvm *gpuvm) +{ +const struct drm_gpuvm_ops *ops = gpuvm->ops; +struct drm_gpuvm_bo *vm_bo; +LIST_HEAD (evict); +int ret =0 ; + +if (unlikely (!ops || !ops->bo_validate)) +return -ENOTSUPP ; + +for_each_vm_bo_in_list (gpuvm, evict, &evict, vm_bo) { +dma_resv_assert_held (vm_bo->obj->resv); + ret = ops->bo_validate (vm_bo->obj); +if (ret) +break ; + } +/* Drop ref in case we break out of the loop. */ +drm_gpuvm_bo_put (vm_bo); +restore_vm_bo_list (gpuvm, evict, &evict); + +return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_validate); + +/** +* drm_gpuvm_resv_add_fence - add fence to private and all extobj +* dma-resv +* @gpuvm: the &drm_gpuvm to add a fence to +* @exec: the &drm_exec locking context +* @fence: fence to add +* @private_usage: private dma-resv usage +* @extobj_usage: extobj dma-resv usage +*/ +void +drm_gpuvm_resv_add_fence (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +struct dma_fence *fence, +enum dma_resv_usage private_usage, +enum dma_resv_usage extobj_usage) +{ +struct drm_gem_object *obj; +unsigned long index; + +drm_exec_for_each_locked_object (exec, index, obj) { +dma_resv_assert_held (obj->resv); +dma_resv_add_fence (obj->resv, fence, +drm_gpuvm_is_extobj (gpuvm, obj) ? + private_usage : extobj_usage); + } +} +EXPORT_SYMBOL_GPL (drm_gpuvm_resv_add_fence); +/** * drm_gpuvm_bo_create() - create a new instance of struct drm_gpuvm_bo * @gpuvm: The &drm_gpuvm the @obj is mapped in. @@ -790,6 +1242,9 @@ drm_gpuvm_bo_create(struct drm_gpuvm *gpuvm,INIT_LIST_HEAD (&vm_bo->list.gpuva);INIT_LIST_HEAD (&vm_bo->list.entry.gem); +INIT_LIST_HEAD (&vm_bo->list.entry.extobj); +INIT_LIST_HEAD (&vm_bo->list.entry.evict); +drm_gem_object_get (obj);return vm_bo;@@ -807,6 +1262,14 @@ drm_gpuvm_bo_destroy(struct kref *kref)drm_gem_gpuva_assert_lock_held (vm_bo->obj); +spin_lock (&gpuvm->extobj.lock); +list_del (&vm_bo->list.entry.extobj); +spin_unlock (&gpuvm->extobj.lock); + +spin_lock (&gpuvm->evict.lock); +list_del (&vm_bo->list.entry.evict); +spin_unlock (&gpuvm->evict.lock); +list_del (&vm_bo->list.entry.gem);drm_gem_object_put (obj);@@ -822,6 +1285,11 @@ drm_gpuvm_bo_destroy(struct kref *kref)* @vm_bo: the &drm_gpuvm_bo to release the reference of * * This releases a reference to @vm_bo. +* +* If the reference count drops to zero, the &gpuvm_bo is destroyed, which +* includes removing it from the GEMs gpuva list. Hence, if a call to this +* function can potentially let the reference count to zero the caller must +* hold the dma-resv or driver specific GEM gpuva lock. */ void drm_gpuvm_bo_put (struct drm_gpuvm_bo *vm_bo)@@ -831,6 +1299,12 @@ drm_gpuvm_bo_put(struct drm_gpuvm_bo *vm_bo) }EXPORT_SYMBOL_GPL (drm_gpuvm_bo_put); +static int __must_check +drm_gpuvm_bo_get_unless_zero (struct drm_gpuvm_bo *vm_bo) +{ +return kref_get_unless_zero (&vm_bo->kref); +} +static struct drm_gpuvm_bo *__drm_gpuvm_bo_find (struct drm_gpuvm *gpuvm,struct drm_gem_object *obj)@@ -938,6 +1412,48 @@ drm_gpuvm_bo_obtain_prealloc(struct drm_gpuvm_bo *__vm_bo) }EXPORT_SYMBOL_GPL (drm_gpuvm_bo_obtain_prealloc); +/** +* drm_gpuvm_bo_extobj_add() - adds the &drm_gpuvm_bo to its &drm_gpuvm's +* extobj list +* @vm_bo: The &drm_gpuvm_bo to add to its &drm_gpuvm's the extobj list. +* +* Adds the given @vm_bo to its &drm_gpuvm's extobj list if not on the list +* already and if the corresponding &drm_gem_object is an external object, +* actually. +*/ +void +drm_gpuvm_bo_extobj_add (struct drm_gpuvm_bo *vm_bo) +{ +struct drm_gpuvm *gpuvm = vm_bo->vm; + +if (drm_gpuvm_is_extobj (gpuvm, vm_bo->obj)) +drm_gpuvm_bo_list_add (vm_bo, extobj); +} +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_extobj_add); + +/** +* drm_gpuvm_bo_evict() - add / remove a &drm_gem_object to / from a +* &drm_gpuvms evicted list +* @obj: the &drm_gem_object to add or remove +* @evict: indicates whether the object is evicted +* +* Adds a &drm_gem_object to or removes it from all &drm_gpuvms evicted +* list containing a mapping of this &drm_gem_object. +*/ +void +drm_gpuvm_bo_evict (struct drm_gem_object *obj,bool evict) +{ +struct drm_gpuvm_bo *vm_bo; + +drm_gem_for_each_gpuvm_bo (vm_bo, obj) { +if (evict) +drm_gpuvm_bo_list_add (vm_bo, evict); +else +drm_gpuvm_bo_list_del (vm_bo, evict); + } +} +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_evict); +static int __drm_gpuva_insert (struct drm_gpuvm *gpuvm,struct drm_gpuva *va)include/drm/drm_gpuvm.h +197
-0@@ -26,10 +26,12 @@*/ #include <linux/list.h> +#include <linux/dma-resv.h> #include <linux/rbtree.h> #include <linux/types.h> #include <drm/drm_gem.h> +#include <drm/drm_exec.h> struct drm_gpuvm ;struct drm_gpuvm_bo ;@@ -259,6 +261,38 @@ struct drm_gpuvm {* space */ struct dma_resv *resv; + +/** +* @extobj: structure holding the extobj list +*/ +struct { +/** +* @list: &list_head storing &drm_gpuvm_bos serving as +* external object +*/ +struct list_head list; + +/** +* @lock: spinlock to protect the extobj list +*/ +spinlock_t lock; + } extobj; + +/** +* @evict: structure holding the evict list and evict list lock +*/ +struct { +/** +* @list: &list_head storing &drm_gpuvm_bos currently being +* evicted +*/ +struct list_head list; + +/** +* @lock: spinlock to protect the evict list +*/ +spinlock_t lock; + } evict; }; void drm_gpuvm_init(struct drm_gpuvm *gpuvm, struct drm_device *drm,@@ -268,6 +302,21 @@ void drm_gpuvm_init(struct drm_gpuvm *gpuvm, struct drm_device *drm, const struct drm_gpuvm_ops *ops);void drm_gpuvm_destroy (struct drm_gpuvm *gpuvm); +/** +* drm_gpuvm_is_extobj() - indicates whether the given &drm_gem_object is an +* external object +* @gpuvm: the &drm_gpuvm to check +* @obj: the &drm_gem_object to check +* +* Returns: true if the &drm_gem_object &dma_resv differs from the +* &drm_gpuvms &dma_resv, false otherwise +*/ +static inline bool drm_gpuvm_is_extobj (struct drm_gpuvm *gpuvm, +struct drm_gem_object *obj) +{ +return obj && obj->resv != gpuvm->resv; +} +static inline struct drm_gpuva *__drm_gpuva_next (struct drm_gpuva *va) {@@ -346,6 +395,128 @@ __drm_gpuva_next(struct drm_gpuva *va)#define drm_gpuvm_for_each_va_safe (va__, next__, gpuvm__) \ list_for_each_entry_safe(va__, next__, &(gpuvm__)->rb.list, rb.entry) +/** +* struct drm_gpuvm_exec - &drm_gpuvm abstraction of &drm_exec +* +* This structure should be created on the stack as &drm_exec should be. +* +* Optionally, @extra can be set in order to lock additional &drm_gem_objects. +*/ +struct drm_gpuvm_exec { +/** +* @exec: the &drm_exec structure +*/ +struct drm_exec exec; + +/** +* @vm: the &drm_gpuvm to lock its DMA reservations +*/ +struct drm_gpuvm *vm; + +/** +* @extra: Callback and corresponding private data for the driver to +* lock arbitrary additional &drm_gem_objects. +*/ +struct { +/** +* @fn: The driver callback to lock additional &drm_gem_objects. +*/ +int (*fn)(struct drm_gpuvm_exec *vm_exec, +unsigned int num_fences); + +/** +* @priv: driver private data for the @fn callback +*/ +void *priv; + } extra; +}; + +/** +* drm_gpuvm_prepare_vm() - prepare the GPUVMs common dma-resv +* @gpuvm: the &drm_gpuvm +* @exec: the &drm_exec context +* @num_fences: the amount of &dma_fences to reserve +* +* Calls drm_exec_prepare_obj() for the GPUVMs dummy &drm_gem_object. +* +* Using this function directly, it is the drivers responsibility to call +* drm_exec_init() and drm_exec_fini() accordingly. +* +* Returns: 0 on success, negative error code on failure. +*/ +static inline int +drm_gpuvm_prepare_vm (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +unsigned int num_fences) +{ +return drm_exec_prepare_obj (exec, &gpuvm->d_obj, num_fences); +} + +int drm_gpuvm_prepare_objects (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +unsigned int num_fences); + +int drm_gpuvm_prepare_range (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +u64 addr,u64 range, +unsigned int num_fences); + +int drm_gpuvm_exec_lock (struct drm_gpuvm_exec *vm_exec, +unsigned int num_fences, +bool interruptible); + +int drm_gpuvm_exec_lock_array (struct drm_gpuvm_exec *vm_exec, +struct drm_gem_object **objs, +unsigned int num_objs, +unsigned int num_fences, +bool interruptible); + +int drm_gpuvm_exec_lock_range (struct drm_gpuvm_exec *vm_exec, +u64 addr,u64 range, +unsigned int num_fences, +bool interruptible); + +/** +* drm_gpuvm_lock() - lock all dma-resv of all assoiciated BOs +* @gpuvm: the &drm_gpuvm +* +* Releases all dma-resv locks of all &drm_gem_objects previously acquired +* through drm_gpuvm_lock() or its variants. +* +* Returns: 0 on success, negative error code on failure. +*/ +static inline void +drm_gpuvm_exec_unlock (struct drm_gpuvm_exec *vm_exec) +{ +drm_exec_fini (&vm_exec->exec); +} + +int drm_gpuvm_validate (struct drm_gpuvm *gpuvm); +void drm_gpuvm_resv_add_fence (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +struct dma_fence *fence, +enum dma_resv_usage private_usage, +enum dma_resv_usage extobj_usage); + +/** +* drm_gpuvm_exec_resv_add_fence() +* @vm_exec: the &drm_gpuvm_exec abstraction +* @fence: fence to add +* @private_usage: private dma-resv usage +* @extobj_usage: extobj dma-resv usage +* +* See drm_gpuvm_resv_add_fence(). +*/ +static inline void +drm_gpuvm_exec_resv_add_fence (struct drm_gpuvm_exec *vm_exec, +struct dma_fence *fence, +enum dma_resv_usage private_usage, +enum dma_resv_usage extobj_usage) +{ +drm_gpuvm_resv_add_fence (vm_exec->vm, &vm_exec->exec, fence, + private_usage, extobj_usage); +} +/** * struct drm_gpuvm_bo - structure representing a &drm_gpuvm and * &drm_gem_object combination @@ -398,6 +569,18 @@ struct drm_gpuvm_bo {* gpuva list. */ struct list_head gem; + +/** +* @evict: List entry to attach to the &drm_gpuvms +* extobj list. +*/ +struct list_head extobj; + +/** +* @evict: List entry to attach to the &drm_gpuvms evict +* list. +*/ +struct list_head evict; } entry; } list; };@@ -432,6 +615,9 @@ struct drm_gpuvm_bo * drm_gpuvm_bo_find(struct drm_gpuvm *gpuvm, struct drm_gem_object *obj); +void drm_gpuvm_bo_evict (struct drm_gem_object *obj,bool evict); +void drm_gpuvm_bo_extobj_add (struct drm_gpuvm_bo *vm_bo); +/** * drm_gpuvm_bo_for_each_va() - iterator to walk over a list of &drm_gpuva * @va__: &drm_gpuva structure to assign to in each iteration step @@ -837,6 +1023,17 @@ struct drm_gpuvm_ops {* used. */ int (*sm_step_unmap)(struct drm_gpuva_op *op,void *priv); + +/** +* @bo_validate: called from drm_gpuvm_validate() +* +* Drivers receive this callback for every evicted &drm_gem_object being +* mapped in the corresponding &drm_gpuvm. +* +* Typically, drivers would call their driver specific variant of +* ttm_bo_validate() from within this callback. +*/ +int (*bo_validate)(struct drm_gem_object *obj); }; int drm_gpuvm_sm_map(struct drm_gpuvm *gpuvm, void *priv,
Re: [PATCH drm-misc-next v3 6/7] drm/gpuvm: generalize dma_resv/extobj handling and GEM validation
Quoted 96 lines · Thomas Hellström
Hi, Danilo,
Quoted 93 lines · Danilo Krummrich
So far the DRM GPUVA manager offers common infrastructure to track GPU VA allocations and mappings, generically connect GPU VA mappings to their backing buffers and perform more complex mapping operations on the GPU VA space.
However, there are more design patterns commonly used by drivers, which can potentially be generalized in order to make the DRM GPUVA manager represent a basic GPU-VM implementation. In this context, this patch aims at generalizing the following elements.
1) Provide a common dma-resv for GEM objects not being used outside of this GPU-VM.
2) Provide tracking of external GEM objects (GEM objects which are shared with other GPU-VMs).
3) Provide functions to efficiently lock all GEM objects dma-resv the GPU-VM contains mappings of.
4) Provide tracking of evicted GEM objects the GPU-VM contains mappings of, such that validation of evicted GEM objects is accelerated.
5) Provide some convinience functions for common patterns.
Rather than being designed as a "framework", the target is to make all features appear as a collection of optional helper functions, such that drivers are free to make use of the DRM GPUVA managers basic functionality and opt-in for other features without setting any feature flags, just by making use of the corresponding functions.
Big kudos to Boris Brezillon for his help to figure out locking for drivers updating the GPU VA space within the fence signalling path.
Suggested-by Matthew Brost <[email protected]>Signed-off-by Danilo Krummrich <[email protected]>--- drivers/gpu/drm/drm_gpuvm.c | 516 ++++++++++++++++++++++++++++++++++++ include/drm/drm_gpuvm.h | 197 ++++++++++++++ 2 files changed, 713 insertions(+)drivers/gpu/drm/drm_gpuvm.c +34
-0@@ -73,6 +73,21 @@* &drm_gem_object list of &drm_gpuvm_bos for an existing instance of this * particular combination. If not existent a new instance is created and linked * to the &drm_gem_object. +* +* &drm_gpuvm_bo structures, since unique for a given &drm_gpuvm, are also used +* as entry for the &drm_gpuvm's lists of external and evicted objects. Those +* list are maintained in order to accelerate locking of dma-resv locks and +* validation of evicted objects bound in a &drm_gpuvm. For instance the all +* &drm_gem_object's &dma_resv of a given &drm_gpuvm can be locked by calling +* drm_gpuvm_exec_lock(). Once locked drivers can call drm_gpuvm_validate() in +* order to validate all evicted &drm_gem_objects. It is also possible to lock +* additional &drm_gem_objects by providing the corresponding parameters to +* drm_gpuvm_exec_lock() as well as open code the &drm_exec loop while making +* use of helper functions such as drm_gpuvm_prepare_range() or +* drm_gpuvm_prepare_objects(). +* +* Every bound &drm_gem_object is treated as external object when its &dma_resv +* structure is different than the &drm_gpuvm's common &dma_resv structure. */ /** @@ -420,6 +435,20 @@* Subsequent calls to drm_gpuvm_bo_obtain() for the same &drm_gpuvm and * &drm_gem_object must be able to observe previous creations and destructions * of &drm_gpuvm_bos in order to keep instances unique. +* +* The &drm_gpuvm's lists for keeping track of external and evicted objects are +* protected against concurrent insertion / removal and iteration internally. +* +* However, drivers still need ensure to protect concurrent calls to functions +* iterating those lists, such as drm_gpuvm_validate() and +* drm_gpuvm_prepare_objects(). Every such function contains a particular +* comment and lockdep checks if possible. +* +* Functions adding or removing entries from those lists, such as +* drm_gpuvm_bo_evict() or drm_gpuvm_bo_extobj_add() may be called with external +* locks being held, e.g. in order to avoid the corresponding list to be +* (safely) modified while potentially being iternated by other API functions. +* However, this is entirely optional. */ /** @@ -632,6 +661,131 @@* } */ +/** +* get_next_vm_bo_from_list() - get the next vm_bo element +* @__gpuvm: The GPU VM +* @__list_name: The name of the list we're iterating on +* @__local_list: A pointer to the local list used to store already iterated items
Quoted 8 lines
drivers/gpu/drm/drm_gpuvm.c line 673
Quoted 5 lines
drivers/gpu/drm/drm_gpuvm.c +5
-0+* @__prev_vm_bo: The previous element we got from drm_gpuvm_get_next_cached_vm_bo() +* +* This helper is here to provide lockless list iteration. Lockless as in, the +* iterator releases the lock immediately after picking the first element from +* the list, so list insertion deletion can happen concurrently. Are the list spinlocks needed for that async state update from within the dma-fence critical section we've discussed previously?
Yes, but also for other reasons, see below.
Quoted 6 lines
Otherwise it should be sufficient to protect the lists with the gpuvm's resv (or for the extobj list with an outer lock).
If those spinlocks are still needed in some situations, perhaps could we have an option to set them to NULL (Like IIRC the maple tree allows for)?
The evict spinlock is needed in any case, since in drm_gpuvm_bo_evict() we're holding only the dma-resv lock from the BO this function gets called for. Hence, the spinlock protects concurrent drm_gpuvm_bo_evict() calls with different BOs.
For extobjs an outer lock would be enough in case of Xe, but I really would not like to add even more complexity just to get the spinlock out of the way in case the driver already has an outer lock protecting this path.
Quoted 3 lines
For such drivers, that would require anybody calling unlink to hold the vm's resv, though.
In V4 I want to go back to having a dedicated lock for the GEMs gpuva list (or VM_BO list to be more precise). We can't just use the dma-resv lock for that with VM_BO abstractions, because on destruction of a VM_BO we otherwise wouldn't be allowed to already hold the dma-resv lock. That's the fix I was referring to earlier.
Quoted 779 lines
It seems that with that also the refcount could be make non-atomic.
All in the spirit of the drm locking guidelines "use big locks when possible".
Lower level locks only when necessary for performance or locking inversion?/Thomas
Quoted 768 lines
drivers/gpu/drm/drm_gpuvm.c +477
-0+* +* Elements popped from the original list are kept in a local list, so removal +* and is_empty checks can still happen while we're iterating the list. +*/ +#define get_next_vm_bo_from_list (__gpuvm, __list_name, __local_list, __prev_vm_bo) \ + ({ \ + struct drm_gpuvm_bo *__vm_bo; \ + \ + drm_gpuvm_bo_put(__prev_vm_bo); \ + \ + spin_lock(&(__gpuvm)->__list_name.lock); \ + while (!list_empty(&(__gpuvm)->__list_name.list)) { \ + __vm_bo = list_first_entry(&(__gpuvm)->__list_name.list, \ + struct drm_gpuvm_bo, \ + list.entry.__list_name); \ + if (drm_gpuvm_bo_get_unless_zero(__vm_bo)) { \ + list_move_tail(&(__vm_bo)->list.entry.__list_name, \ + __local_list); \ + break; \ + } else { \ + list_del_init(&(__vm_bo)->list.entry.__list_name); \ + __vm_bo = NULL; \ + } \ + } \ + spin_unlock(&(__gpuvm)->__list_name.lock); \ + \ + __vm_bo; \ + }) + +/** +* for_each_vm_bo_in_list() - internal vm_bo list iterator +* +* This helper is here to provide lockless list iteration. Lockless as in, the +* iterator releases the lock immediately after picking the first element from the +* list, so list insertion and deletion can happen concurrently. +* +* Typical use: +* +* struct drm_gpuvm_bo *vm_bo; +* LIST_HEAD(my_local_list); +* +* ret = 0; +* drm_gpuvm_for_each_vm_bo(gpuvm, <list_name>, &my_local_list, vm_bo) { +* ret = do_something_with_vm_bo(..., vm_bo); +* if (ret) +* break; +* } +* drm_gpuvm_bo_put(vm_bo); +* drm_gpuvm_restore_vm_bo_list(gpuvm, <list_name>, &my_local_list); +* +* +* Only used for internal list iterations, not meant to be exposed to the outside +* world. +*/ +#define for_each_vm_bo_in_list (__gpuvm, __list_name, __local_list, __vm_bo) \ + for (__vm_bo = get_next_vm_bo_from_list(__gpuvm, __list_name, \ + __local_list, NULL); \ + __vm_bo; \ + __vm_bo = get_next_vm_bo_from_list(__gpuvm, __list_name, \ + __local_list, __vm_bo)) \ + +/** +* restore_vm_bo_list() - move vm_bo elements back to their original list +* @__gpuvm: The GPU VM +* @__list_name: The name of the list we're iterating on +* @__local_list: A pointer to the local list used to store already iterated items +* +* When we're done iterating a vm_bo list, we should call restore_vm_bo_list() +* to restore the original state and let new iterations take place. +*/ +#define restore_vm_bo_list (__gpuvm, __list_name, __local_list) \ + do { \ +/* Merge back the two lists, moving local list elements to the \ +* head to preserve previous ordering, in case it matters. \ +*/ \ +spin_lock (&(__gpuvm)->__list_name.lock); \ +list_splice (__local_list, &(__gpuvm)->__list_name.list); \ +spin_unlock (&(__gpuvm)->__list_name.lock); \ + }while (0 ) +/** +* drm_gpuvm_bo_list_add() - insert a vm_bo into the given list +* @__vm_bo: the &drm_gpuvm_bo +* @__list_name: the name of the list to insert into +* +* Inserts the given @__vm_bo into the list specified by @__list_name and +* increases the vm_bo's reference count. + */ +#define drm_gpuvm_bo_list_add (__vm_bo, __list_name) \ + do { \ + spin_lock(&(__vm_bo)->vm->__list_name.lock); \ + if (list_empty(&(__vm_bo)->list.entry.__list_name)) \ + list_add_tail(&(__vm_bo)->list.entry.__list_name, \ + &(__vm_bo)->vm->__list_name.list); \ + spin_unlock(&(__vm_bo)->vm->__list_name.lock); \ + } while (0) + +/** +* drm_gpuvm_bo_list_del() - remove a vm_bo from the given list +* @__vm_bo: the &drm_gpuvm_bo +* @__list_name: the name of the list to insert into +* +* Removes the given @__vm_bo from the list specified by @__list_name and +* decreases the vm_bo's reference count. +*/ +#define drm_gpuvm_bo_list_del (__vm_bo, __list_name) \ + do { \ + spin_lock(&(__vm_bo)->vm->__list_name.lock); \ + if (!list_empty(&(__vm_bo)->list.entry.__list_name)) \ + list_del_init(&(__vm_bo)->list.entry.__list_name); \ + spin_unlock(&(__vm_bo)->vm->__list_name.lock); \ + } while (0) + +static int __must_check +drm_gpuvm_bo_get_unless_zero (struct drm_gpuvm_bo *vm_bo); +#define to_drm_gpuva (__node) container_of((__node), struct drm_gpuva, rb.node)#define GPUVA_START (node) ((node)->va.addr)@@ -713,6 +867,12 @@ drm_gpuvm_init(struct drm_gpuvm *gpuvm, struct drm_device *drm, gpuvm->rb.tree =RB_ROOT_CACHED ;INIT_LIST_HEAD (&gpuvm->rb.list); +INIT_LIST_HEAD (&gpuvm->extobj.list); +spin_lock_init (&gpuvm->extobj.lock); + +INIT_LIST_HEAD (&gpuvm->evict.list); +spin_lock_init (&gpuvm->evict.lock); +drm_gpuva_check_overflow (start_offset, range); gpuvm->mm_start = start_offset; gpuvm->mm_range = range;@@ -754,10 +914,302 @@ drm_gpuvm_destroy(struct drm_gpuvm *gpuvm)WARN (!RB_EMPTY_ROOT (&gpuvm->rb.tree.rb_root),"GPUVA tree is not empty, potentially leaking memory.\n" ); +WARN (!list_empty (&gpuvm->extobj.list),"Extobj list should be empty.\n" ); +WARN (!list_empty (&gpuvm->evict.list),"Evict list should be empty.\n" ); +drm_gem_private_object_fini (&gpuvm->d_obj); }EXPORT_SYMBOL_GPL (drm_gpuvm_destroy); +/** +* drm_gpuvm_prepare_objects() - prepare all assoiciated BOs +* @gpuvm: the &drm_gpuvm +* @exec: the &drm_exec locking context +* @num_fences: the amount of &dma_fences to reserve +* +* Calls drm_exec_prepare_obj() for all &drm_gem_objects the given +* &drm_gpuvm contains mappings of. +* +* Using this function directly, it is the drivers responsibility to call +* drm_exec_init() and drm_exec_fini() accordingly. +* +* Note: This function is safe against concurrent insertion and removal of +* external objects, however it is not safe against concurrent usage itself. +* +* Drivers need to make sure to protect this case with either an outer VM lock +* or by calling drm_gpuvm_prepare_vm() before this function within the +* drm_exec_until_all_locked() loop, such that the GPUVM's dma-resv lock ensures +* mutual exclusion. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_prepare_objects (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +unsigned int num_fences) +{ +struct drm_gpuvm_bo *vm_bo; +LIST_HEAD (extobjs); +int ret =0 ; + +for_each_vm_bo_in_list (gpuvm, extobj, &extobjs, vm_bo) { + ret =drm_exec_prepare_obj (exec, vm_bo->obj, num_fences); +if (ret) +break ; + } +/* Drop ref in case we break out of the loop. */ +drm_gpuvm_bo_put (vm_bo); +restore_vm_bo_list (gpuvm, extobj, &extobjs); + +return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_prepare_objects); + +/** +* drm_gpuvm_prepare_range() - prepare all BOs mapped within a given range +* @gpuvm: the &drm_gpuvm +* @exec: the &drm_exec locking context +* @addr: the start address within the VA space +* @range: the range to iterate within the VA space +* @num_fences: the amount of &dma_fences to reserve +* +* Calls drm_exec_prepare_obj() for all &drm_gem_objects mapped between @addr +* and @addr + @range. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_prepare_range (struct drm_gpuvm *gpuvm,struct drm_exec *exec, +u64 addr,u64 range,unsigned int num_fences) +{ +struct drm_gpuva *va; +u64 end = addr + range; +int ret; + +drm_gpuvm_for_each_va_range (va, gpuvm, addr, end) { +struct drm_gem_object *obj = va->gem.obj; + + ret =drm_exec_prepare_obj (exec, obj, num_fences); +if (ret) +return ret; + } + +return 0 ; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_prepare_range); + +/** +* drm_gpuvm_exec_lock() - lock all dma-resv of all assoiciated BOs +* @vm_exec: the &drm_gpuvm_exec abstraction +* @num_fences: the amount of &dma_fences to reserve +* @interruptible: sleep interruptible if waiting +* +* Acquires all dma-resv locks of all &drm_gem_objects the given +* &drm_gpuvm contains mappings of. +* +* Addionally, when calling this function with struct drm_gpuvm_exec::extra +* being set the driver receives the given @fn callback to lock additional +* dma-resv in the context of the &drm_gpuvm_exec instance. Typically, drivers +* would call drm_exec_prepare_obj() from within this callback. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_exec_lock (struct drm_gpuvm_exec *vm_exec, +unsigned int num_fences, +bool interruptible) +{ +struct drm_gpuvm *gpuvm = vm_exec->vm; +struct drm_exec *exec = &vm_exec->exec; +uint32_t flags; +int ret; + + flags = interruptible ?DRM_EXEC_INTERRUPTIBLE_WAIT :0 | +DRM_EXEC_IGNORE_DUPLICATES ; + +drm_exec_init (exec, flags); + +drm_exec_until_all_locked (exec) { + ret =drm_gpuvm_prepare_vm (gpuvm, exec, num_fences); +drm_exec_retry_on_contention (exec); +if (ret) + goto err; + + ret =drm_gpuvm_prepare_objects (gpuvm, exec, num_fences); +drm_exec_retry_on_contention (exec); +if (ret) + goto err; + +if (vm_exec->extra.fn) { + ret = vm_exec->extra.fn (vm_exec, num_fences); +drm_exec_retry_on_contention (exec); +if (ret) + goto err; + } + } + +return 0 ; + +err: +drm_exec_fini (exec); +return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock); + +static int +fn_lock_array (struct drm_gpuvm_exec *vm_exec,unsigned int num_fences) +{ +struct { +struct drm_gem_object **objs; +unsigned int num_objs; + } *args = vm_exec->extra.priv; + +return drm_exec_prepare_array (&vm_exec->exec, args->objs, + args->num_objs, num_fences); +} + +/** +* drm_gpuvm_exec_lock_array() - lock all dma-resv of all assoiciated BOs +* @vm_exec: the &drm_gpuvm_exec abstraction +* @objs: additional &drm_gem_objects to lock +* @num_objs: the number of additional &drm_gem_objects to lock +* @num_fences: the amount of &dma_fences to reserve +* @interruptible: sleep interruptible if waiting +* +* Acquires all dma-resv locks of all &drm_gem_objects the given &drm_gpuvm +* contains mappings of, plus the ones given through @objs. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_exec_lock_array (struct drm_gpuvm_exec *vm_exec, +struct drm_gem_object **objs, +unsigned int num_objs, +unsigned int num_fences, +bool interruptible) +{ +struct { +struct drm_gem_object **objs; +unsigned int num_objs; + } args; + + args.objs = objs; + args.num_objs = num_objs; + + vm_exec->extra.fn = fn_lock_array; + vm_exec->extra.priv = &args; + +return drm_gpuvm_exec_lock (vm_exec, num_fences, interruptible); +} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock_array); + +/** +* drm_gpuvm_exec_lock_range() - prepare all BOs mapped within a given range +* @vm_exec: the &drm_gpuvm_exec abstraction +* @addr: the start address within the VA space +* @range: the range to iterate within the VA space +* @num_fences: the amount of &dma_fences to reserve +* @interruptible: sleep interruptible if waiting +* +* Acquires all dma-resv locks of all &drm_gem_objects mapped between @addr and +* @addr + @range. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_exec_lock_range (struct drm_gpuvm_exec *vm_exec, +u64 addr,u64 range, +unsigned int num_fences, +bool interruptible) +{ +struct drm_gpuvm *gpuvm = vm_exec->vm; +struct drm_exec *exec = &vm_exec->exec; +uint32_t flags; +int ret; + + flags = interruptible ?DRM_EXEC_INTERRUPTIBLE_WAIT :0 | +DRM_EXEC_IGNORE_DUPLICATES ; + +drm_exec_init (exec, flags); + +drm_exec_until_all_locked (exec) { + ret =drm_gpuvm_prepare_range (gpuvm, exec, addr, range, + num_fences); +drm_exec_retry_on_contention (exec); +if (ret) + goto err; + } + +return ret; + +err: +drm_exec_fini (exec); +return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock_range); + +/** +* drm_gpuvm_validate() - validate all BOs marked as evicted +* @gpuvm: the &drm_gpuvm to validate evicted BOs +* +* Calls the &drm_gpuvm_ops.bo_validate callback for all evicted buffer +* objects being mapped in the given &drm_gpuvm. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_validate (struct drm_gpuvm *gpuvm) +{ +const struct drm_gpuvm_ops *ops = gpuvm->ops; +struct drm_gpuvm_bo *vm_bo; +LIST_HEAD (evict); +int ret =0 ; + +if (unlikely (!ops || !ops->bo_validate)) +return -ENOTSUPP ; + +for_each_vm_bo_in_list (gpuvm, evict, &evict, vm_bo) { +dma_resv_assert_held (vm_bo->obj->resv); + ret = ops->bo_validate (vm_bo->obj); +if (ret) +break ; + } +/* Drop ref in case we break out of the loop. */ +drm_gpuvm_bo_put (vm_bo); +restore_vm_bo_list (gpuvm, evict, &evict); + +return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_validate); + +/** +* drm_gpuvm_resv_add_fence - add fence to private and all extobj +* dma-resv +* @gpuvm: the &drm_gpuvm to add a fence to +* @exec: the &drm_exec locking context +* @fence: fence to add +* @private_usage: private dma-resv usage +* @extobj_usage: extobj dma-resv usage +*/ +void +drm_gpuvm_resv_add_fence (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +struct dma_fence *fence, +enum dma_resv_usage private_usage, +enum dma_resv_usage extobj_usage) +{ +struct drm_gem_object *obj; +unsigned long index; + +drm_exec_for_each_locked_object (exec, index, obj) { +dma_resv_assert_held (obj->resv); +dma_resv_add_fence (obj->resv, fence, +drm_gpuvm_is_extobj (gpuvm, obj) ? + private_usage : extobj_usage); + } +} +EXPORT_SYMBOL_GPL (drm_gpuvm_resv_add_fence); +/** * drm_gpuvm_bo_create() - create a new instance of struct drm_gpuvm_bo * @gpuvm: The &drm_gpuvm the @obj is mapped in. @@ -790,6 +1242,9 @@ drm_gpuvm_bo_create(struct drm_gpuvm *gpuvm,INIT_LIST_HEAD (&vm_bo->list.gpuva);INIT_LIST_HEAD (&vm_bo->list.entry.gem); +INIT_LIST_HEAD (&vm_bo->list.entry.extobj); +INIT_LIST_HEAD (&vm_bo->list.entry.evict); +drm_gem_object_get (obj);return vm_bo;@@ -807,6 +1262,14 @@ drm_gpuvm_bo_destroy(struct kref *kref)drm_gem_gpuva_assert_lock_held (vm_bo->obj); +spin_lock (&gpuvm->extobj.lock); +list_del (&vm_bo->list.entry.extobj); +spin_unlock (&gpuvm->extobj.lock); + +spin_lock (&gpuvm->evict.lock); +list_del (&vm_bo->list.entry.evict); +spin_unlock (&gpuvm->evict.lock); +list_del (&vm_bo->list.entry.gem);drm_gem_object_put (obj);@@ -822,6 +1285,11 @@ drm_gpuvm_bo_destroy(struct kref *kref)* @vm_bo: the &drm_gpuvm_bo to release the reference of * * This releases a reference to @vm_bo. +* +* If the reference count drops to zero, the &gpuvm_bo is destroyed, which +* includes removing it from the GEMs gpuva list. Hence, if a call to this +* function can potentially let the reference count to zero the caller must +* hold the dma-resv or driver specific GEM gpuva lock. */ void drm_gpuvm_bo_put (struct drm_gpuvm_bo *vm_bo)@@ -831,6 +1299,12 @@ drm_gpuvm_bo_put(struct drm_gpuvm_bo *vm_bo) }EXPORT_SYMBOL_GPL (drm_gpuvm_bo_put); +static int __must_check +drm_gpuvm_bo_get_unless_zero (struct drm_gpuvm_bo *vm_bo) +{ +return kref_get_unless_zero (&vm_bo->kref); +} +static struct drm_gpuvm_bo *__drm_gpuvm_bo_find (struct drm_gpuvm *gpuvm,struct drm_gem_object *obj)@@ -938,6 +1412,48 @@ drm_gpuvm_bo_obtain_prealloc(struct drm_gpuvm_bo *__vm_bo) }EXPORT_SYMBOL_GPL (drm_gpuvm_bo_obtain_prealloc); +/** +* drm_gpuvm_bo_extobj_add() - adds the &drm_gpuvm_bo to its &drm_gpuvm's +* extobj list +* @vm_bo: The &drm_gpuvm_bo to add to its &drm_gpuvm's the extobj list. +* +* Adds the given @vm_bo to its &drm_gpuvm's extobj list if not on the list +* already and if the corresponding &drm_gem_object is an external object, +* actually. +*/ +void +drm_gpuvm_bo_extobj_add (struct drm_gpuvm_bo *vm_bo) +{ +struct drm_gpuvm *gpuvm = vm_bo->vm; + +if (drm_gpuvm_is_extobj (gpuvm, vm_bo->obj)) +drm_gpuvm_bo_list_add (vm_bo, extobj); +} +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_extobj_add); + +/** +* drm_gpuvm_bo_evict() - add / remove a &drm_gem_object to / from a +* &drm_gpuvms evicted list +* @obj: the &drm_gem_object to add or remove +* @evict: indicates whether the object is evicted +* +* Adds a &drm_gem_object to or removes it from all &drm_gpuvms evicted +* list containing a mapping of this &drm_gem_object. +*/ +void +drm_gpuvm_bo_evict (struct drm_gem_object *obj,bool evict) +{ +struct drm_gpuvm_bo *vm_bo; + +drm_gem_for_each_gpuvm_bo (vm_bo, obj) { +if (evict) +drm_gpuvm_bo_list_add (vm_bo, evict); +else +drm_gpuvm_bo_list_del (vm_bo, evict); + } +} +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_evict); +static int __drm_gpuva_insert (struct drm_gpuvm *gpuvm,struct drm_gpuva *va)include/drm/drm_gpuvm.h +197
-0@@ -26,10 +26,12 @@*/ #include <linux/list.h> +#include <linux/dma-resv.h> #include <linux/rbtree.h> #include <linux/types.h> #include <drm/drm_gem.h> +#include <drm/drm_exec.h> struct drm_gpuvm ;struct drm_gpuvm_bo ;@@ -259,6 +261,38 @@ struct drm_gpuvm {* space */ struct dma_resv *resv; + +/** +* @extobj: structure holding the extobj list +*/ +struct { +/** +* @list: &list_head storing &drm_gpuvm_bos serving as +* external object +*/ +struct list_head list; + +/** +* @lock: spinlock to protect the extobj list +*/ +spinlock_t lock; + } extobj; + +/** +* @evict: structure holding the evict list and evict list lock +*/ +struct { +/** +* @list: &list_head storing &drm_gpuvm_bos currently being +* evicted +*/ +struct list_head list; + +/** +* @lock: spinlock to protect the evict list +*/ +spinlock_t lock; + } evict; }; void drm_gpuvm_init(struct drm_gpuvm *gpuvm, struct drm_device *drm,@@ -268,6 +302,21 @@ void drm_gpuvm_init(struct drm_gpuvm *gpuvm, struct drm_device *drm, const struct drm_gpuvm_ops *ops);void drm_gpuvm_destroy (struct drm_gpuvm *gpuvm); +/** +* drm_gpuvm_is_extobj() - indicates whether the given &drm_gem_object is an +* external object +* @gpuvm: the &drm_gpuvm to check +* @obj: the &drm_gem_object to check +* +* Returns: true if the &drm_gem_object &dma_resv differs from the +* &drm_gpuvms &dma_resv, false otherwise +*/ +static inline bool drm_gpuvm_is_extobj (struct drm_gpuvm *gpuvm, +struct drm_gem_object *obj) +{ +return obj && obj->resv != gpuvm->resv; +} +static inline struct drm_gpuva *__drm_gpuva_next (struct drm_gpuva *va) {@@ -346,6 +395,128 @@ __drm_gpuva_next(struct drm_gpuva *va)#define drm_gpuvm_for_each_va_safe (va__, next__, gpuvm__) \ list_for_each_entry_safe(va__, next__, &(gpuvm__)->rb.list, rb.entry) +/** +* struct drm_gpuvm_exec - &drm_gpuvm abstraction of &drm_exec +* +* This structure should be created on the stack as &drm_exec should be. +* +* Optionally, @extra can be set in order to lock additional &drm_gem_objects. +*/ +struct drm_gpuvm_exec { +/** +* @exec: the &drm_exec structure +*/ +struct drm_exec exec; + +/** +* @vm: the &drm_gpuvm to lock its DMA reservations +*/ +struct drm_gpuvm *vm; + +/** +* @extra: Callback and corresponding private data for the driver to +* lock arbitrary additional &drm_gem_objects. +*/ +struct { +/** +* @fn: The driver callback to lock additional &drm_gem_objects. +*/ +int (*fn)(struct drm_gpuvm_exec *vm_exec, +unsigned int num_fences); + +/** +* @priv: driver private data for the @fn callback +*/ +void *priv; + } extra; +}; + +/** +* drm_gpuvm_prepare_vm() - prepare the GPUVMs common dma-resv +* @gpuvm: the &drm_gpuvm +* @exec: the &drm_exec context +* @num_fences: the amount of &dma_fences to reserve +* +* Calls drm_exec_prepare_obj() for the GPUVMs dummy &drm_gem_object. +* +* Using this function directly, it is the drivers responsibility to call +* drm_exec_init() and drm_exec_fini() accordingly. +* +* Returns: 0 on success, negative error code on failure. +*/ +static inline int +drm_gpuvm_prepare_vm (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +unsigned int num_fences) +{ +return drm_exec_prepare_obj (exec, &gpuvm->d_obj, num_fences); +} + +int drm_gpuvm_prepare_objects (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +unsigned int num_fences); + +int drm_gpuvm_prepare_range (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +u64 addr,u64 range, +unsigned int num_fences); + +int drm_gpuvm_exec_lock (struct drm_gpuvm_exec *vm_exec, +unsigned int num_fences, +bool interruptible); + +int drm_gpuvm_exec_lock_array (struct drm_gpuvm_exec *vm_exec, +struct drm_gem_object **objs, +unsigned int num_objs, +unsigned int num_fences, +bool interruptible); + +int drm_gpuvm_exec_lock_range (struct drm_gpuvm_exec *vm_exec, +u64 addr,u64 range, +unsigned int num_fences, +bool interruptible); + +/** +* drm_gpuvm_lock() - lock all dma-resv of all assoiciated BOs +* @gpuvm: the &drm_gpuvm +* +* Releases all dma-resv locks of all &drm_gem_objects previously acquired +* through drm_gpuvm_lock() or its variants. +* +* Returns: 0 on success, negative error code on failure. +*/ +static inline void +drm_gpuvm_exec_unlock (struct drm_gpuvm_exec *vm_exec) +{ +drm_exec_fini (&vm_exec->exec); +} + +int drm_gpuvm_validate (struct drm_gpuvm *gpuvm); +void drm_gpuvm_resv_add_fence (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +struct dma_fence *fence, +enum dma_resv_usage private_usage, +enum dma_resv_usage extobj_usage); + +/** +* drm_gpuvm_exec_resv_add_fence() +* @vm_exec: the &drm_gpuvm_exec abstraction +* @fence: fence to add +* @private_usage: private dma-resv usage +* @extobj_usage: extobj dma-resv usage +* +* See drm_gpuvm_resv_add_fence(). +*/ +static inline void +drm_gpuvm_exec_resv_add_fence (struct drm_gpuvm_exec *vm_exec, +struct dma_fence *fence, +enum dma_resv_usage private_usage, +enum dma_resv_usage extobj_usage) +{ +drm_gpuvm_resv_add_fence (vm_exec->vm, &vm_exec->exec, fence, + private_usage, extobj_usage); +} +/** * struct drm_gpuvm_bo - structure representing a &drm_gpuvm and * &drm_gem_object combination @@ -398,6 +569,18 @@ struct drm_gpuvm_bo {* gpuva list. */ struct list_head gem; + +/** +* @evict: List entry to attach to the &drm_gpuvms +* extobj list. +*/ +struct list_head extobj; + +/** +* @evict: List entry to attach to the &drm_gpuvms evict +* list. +*/ +struct list_head evict; } entry; } list; };@@ -432,6 +615,9 @@ struct drm_gpuvm_bo * drm_gpuvm_bo_find(struct drm_gpuvm *gpuvm, struct drm_gem_object *obj); +void drm_gpuvm_bo_evict (struct drm_gem_object *obj,bool evict); +void drm_gpuvm_bo_extobj_add (struct drm_gpuvm_bo *vm_bo); +/** * drm_gpuvm_bo_for_each_va() - iterator to walk over a list of &drm_gpuva * @va__: &drm_gpuva structure to assign to in each iteration step @@ -837,6 +1023,17 @@ struct drm_gpuvm_ops {* used. */ int (*sm_step_unmap)(struct drm_gpuva_op *op,void *priv); + +/** +* @bo_validate: called from drm_gpuvm_validate() +* +* Drivers receive this callback for every evicted &drm_gem_object being +* mapped in the corresponding &drm_gpuvm. +* +* Typically, drivers would call their driver specific variant of +* ttm_bo_validate() from within this callback. +*/ +int (*bo_validate)(struct drm_gem_object *obj); }; int drm_gpuvm_sm_map(struct drm_gpuvm *gpuvm, void *priv,
Thomas Hellström
reviewerRe: [PATCH drm-misc-next v3 6/7] drm/gpuvm: generalize dma_resv/extobj handling and GEM validation
Quoted 106 lines · Danilo Krummrich
Quoted 95 lines · Thomas Hellström
Hi, Danilo,
Quoted 92 lines · Danilo Krummrich
So far the DRM GPUVA manager offers common infrastructure to track GPU VA allocations and mappings, generically connect GPU VA mappings to their backing buffers and perform more complex mapping operations on the GPU VA space.
However, there are more design patterns commonly used by drivers, which can potentially be generalized in order to make the DRM GPUVA manager represent a basic GPU-VM implementation. In this context, this patch aims at generalizing the following elements.
1) Provide a common dma-resv for GEM objects not being used outside of this GPU-VM.
2) Provide tracking of external GEM objects (GEM objects which are shared with other GPU-VMs).
3) Provide functions to efficiently lock all GEM objects dma-resv the GPU-VM contains mappings of.
4) Provide tracking of evicted GEM objects the GPU-VM contains mappings of, such that validation of evicted GEM objects is accelerated.
5) Provide some convinience functions for common patterns.
Rather than being designed as a "framework", the target is to make all features appear as a collection of optional helper functions, such that drivers are free to make use of the DRM GPUVA managers basic functionality and opt-in for other features without setting any feature flags, just by making use of the corresponding functions.
Big kudos to Boris Brezillon for his help to figure out locking for drivers updating the GPU VA space within the fence signalling path.
Suggested-by Matthew Brost <[email protected]>Signed-off-by Danilo Krummrich <[email protected]>--- drivers/gpu/drm/drm_gpuvm.c | 516 ++++++++++++++++++++++++++++++++++++ include/drm/drm_gpuvm.h | 197 ++++++++++++++ 2 files changed, 713 insertions(+)drivers/gpu/drm/drm_gpuvm.c +33
-0@@ -73,6 +73,21 @@* &drm_gem_object list of &drm_gpuvm_bos for an existing instance of this * particular combination. If not existent a new instance is created and linked * to the &drm_gem_object. +* +* &drm_gpuvm_bo structures, since unique for a given &drm_gpuvm, are also used +* as entry for the &drm_gpuvm's lists of external and evicted objects. Those +* list are maintained in order to accelerate locking of dma-resv locks and +* validation of evicted objects bound in a &drm_gpuvm. For instance the all +* &drm_gem_object's &dma_resv of a given &drm_gpuvm can be locked by calling +* drm_gpuvm_exec_lock(). Once locked drivers can call drm_gpuvm_validate() in +* order to validate all evicted &drm_gem_objects. It is also possible to lock +* additional &drm_gem_objects by providing the corresponding parameters to +* drm_gpuvm_exec_lock() as well as open code the &drm_exec loop while making +* use of helper functions such as drm_gpuvm_prepare_range() or +* drm_gpuvm_prepare_objects(). +* +* Every bound &drm_gem_object is treated as external object when its &dma_resv +* structure is different than the &drm_gpuvm's common &dma_resv structure. */ /** @@ -420,6 +435,20 @@* Subsequent calls to drm_gpuvm_bo_obtain() for the same &drm_gpuvm and * &drm_gem_object must be able to observe previous creations and destructions * of &drm_gpuvm_bos in order to keep instances unique. +* +* The &drm_gpuvm's lists for keeping track of external and evicted objects are +* protected against concurrent insertion / removal and iteration internally. +* +* However, drivers still need ensure to protect concurrent calls to functions +* iterating those lists, such as drm_gpuvm_validate() and +* drm_gpuvm_prepare_objects(). Every such function contains a particular +* comment and lockdep checks if possible. +* +* Functions adding or removing entries from those lists, such as +* drm_gpuvm_bo_evict() or drm_gpuvm_bo_extobj_add() may be called with external +* locks being held, e.g. in order to avoid the corresponding list to be +* (safely) modified while potentially being iternated by other API functions. +* However, this is entirely optional. */ /** @@ -632,6 +661,131 @@* } */ +/** +* get_next_vm_bo_from_list() - get the next vm_bo element +* @__gpuvm: The GPU VM +* @__list_name: The name of the list we're iterating on Quoted 8 lines
drivers/gpu/drm/drm_gpuvm.c line 673
Quoted 6 lines
drivers/gpu/drm/drm_gpuvm.c +6
-0+* @__local_list: A pointer to the local list used to store already iterated items +* @__prev_vm_bo: The previous element we got from drm_gpuvm_get_next_cached_vm_bo() +* +* This helper is here to provide lockless list iteration. Lockless as in, the +* iterator releases the lock immediately after picking the first element from +* the list, so list insertion deletion can happen concurrently. Are the list spinlocks needed for that async state update from within the dma-fence critical section we've discussed previously?
Yes, but also for other reasons, see below.
Quoted 8 lines
Quoted 5 lines
Otherwise it should be sufficient to protect the lists with the gpuvm's resv (or for the extobj list with an outer lock).
If those spinlocks are still needed in some situations, perhaps could we have an option to set them to NULL (Like IIRC the maple tree allows for)?
The evict spinlock is needed in any case, since in
drm_gpuvm_bo_evict()we're holding only the dma-resv lock from the BO this function gets called for. Hence, the spinlock protects concurrentdrm_gpuvm_bo_evict()calls with different BOs.
No. Only if you try to add external objects to the vm's evict list from within the evict code. That's not necessary since you loop through all external objects anyway when locking them so an "evicted" bool in the vm_bo, protected by the bo resv would be sufficient. The extobj locking loop can then add the bo to the evicted list.
Quoted 4 lines
For extobjs an outer lock would be enough in case of Xe, but I really would not like to add even more complexity just to get the spinlock out of the way in case the driver already has an outer lock protecting this path.
I must disagree here. These spinlocks and atomic operations are pretty costly and as discussed earlier this type of locking was the reason (at least according to the commit message) that made Christian drop the XArray use in drm_exec for the same set of objects: "The locking overhead is unecessary and measurable". IMHO the spinlock is the added complexity and a single wide lock following the drm locking guidelines set out by Daniel and David should really be the default choice with an opt-in for a spinlock if needed for async and pushing out to a wq is not an option.
A pretty simple way that would not add much code would be
static void gpuvm_cond_spin_lock(const struct drm_gpuvm *gpuvm, spinlock_t *lock)
{
if (!gpuvm->resv_protected_lists)
spin_lock(lock);
}
Quoted 7 lines
Quoted 2 lines
For such drivers, that would require anybody calling unlink to hold the vm's resv, though.
In V4 I want to go back to having a dedicated lock for the GEMs gpuva list (or
VM_BOlist to be more precise). We can't just use the dma-resv lock for that withVM_BOabstractions, because on destruction of aVM_BOwe otherwise wouldn't be allowed to already hold the dma-resv lock. That's the fix I was referring to earlier.
Yeah, I can see the need for a dedicated lock for the GEM's gpuva list, but holding the vm's dma-resv lock across the unlink shouldn't be a problem. We may free the object and a pointer to the vm's resv during unlink but we don't free the vm's resv. It'd be a matter of ensuring that any calls to unlink from *within* drm_gpuvm allows it to be held.
/Thomas
Quoted 777 lines
Quoted 777 lines
It seems that with that also the refcount could be make non-atomic.
All in the spirit of the drm locking guidelines "use big locks when possible".
Lower level locks only when necessary for performance or locking inversion?/Thomas
Quoted 768 lines
drivers/gpu/drm/drm_gpuvm.c +477
-0+* +* Elements popped from the original list are kept in a local list, so removal +* and is_empty checks can still happen while we're iterating the list. +*/ +#define get_next_vm_bo_from_list (__gpuvm, __list_name, __local_list, __prev_vm_bo) \ + ({ \ + struct drm_gpuvm_bo *__vm_bo; \ + \ + drm_gpuvm_bo_put(__prev_vm_bo); \ + \ + spin_lock(&(__gpuvm)->__list_name.lock); \ + while (!list_empty(&(__gpuvm)->__list_name.list)) { \ + __vm_bo = list_first_entry(&(__gpuvm)->__list_name.list, \ + struct drm_gpuvm_bo, \ + list.entry.__list_name); \ + if (drm_gpuvm_bo_get_unless_zero(__vm_bo)) { \ + list_move_tail(&(__vm_bo)->list.entry.__list_name, \ + __local_list); \ + break; \ + } else { \ + list_del_init(&(__vm_bo)->list.entry.__list_name); \ + __vm_bo = NULL; \ + } \ + } \ + spin_unlock(&(__gpuvm)->__list_name.lock); \ + \ + __vm_bo; \ + }) + +/** +* for_each_vm_bo_in_list() - internal vm_bo list iterator +* +* This helper is here to provide lockless list iteration. Lockless as in, the +* iterator releases the lock immediately after picking the first element from the +* list, so list insertion and deletion can happen concurrently. +* +* Typical use: +* +* struct drm_gpuvm_bo *vm_bo; +* LIST_HEAD(my_local_list); +* +* ret = 0; +* drm_gpuvm_for_each_vm_bo(gpuvm, <list_name>, &my_local_list, vm_bo) { +* ret = do_something_with_vm_bo(..., vm_bo); +* if (ret) +* break; +* } +* drm_gpuvm_bo_put(vm_bo); +* drm_gpuvm_restore_vm_bo_list(gpuvm, <list_name>, &my_local_list); +* +* +* Only used for internal list iterations, not meant to be exposed to the outside +* world. +*/ +#define for_each_vm_bo_in_list (__gpuvm, __list_name, __local_list, __vm_bo) \ + for (__vm_bo = get_next_vm_bo_from_list(__gpuvm, __list_name, \ + __local_list, NULL); \ + __vm_bo; \ + __vm_bo = get_next_vm_bo_from_list(__gpuvm, __list_name, \ + __local_list, __vm_bo)) \ + +/** +* restore_vm_bo_list() - move vm_bo elements back to their original list +* @__gpuvm: The GPU VM +* @__list_name: The name of the list we're iterating on +* @__local_list: A pointer to the local list used to store already iterated items +* +* When we're done iterating a vm_bo list, we should call restore_vm_bo_list() +* to restore the original state and let new iterations take place. +*/ +#define restore_vm_bo_list (__gpuvm, __list_name, __local_list) \ + do { \ +/* Merge back the two lists, moving local list elements to the \ +* head to preserve previous ordering, in case it matters. \ +*/ \ +spin_lock (&(__gpuvm)->__list_name.lock); \ +list_splice (__local_list, &(__gpuvm)->__list_name.list); \ +spin_unlock (&(__gpuvm)->__list_name.lock); \ + }while (0 ) +/** +* drm_gpuvm_bo_list_add() - insert a vm_bo into the given list +* @__vm_bo: the &drm_gpuvm_bo +* @__list_name: the name of the list to insert into +* +* Inserts the given @__vm_bo into the list specified by @__list_name and +* increases the vm_bo's reference count. + */ +#define drm_gpuvm_bo_list_add (__vm_bo, __list_name) \ + do { \ + spin_lock(&(__vm_bo)->vm->__list_name.lock); \ + if (list_empty(&(__vm_bo)->list.entry.__list_name)) \ + list_add_tail(&(__vm_bo)->list.entry.__list_name, \ + &(__vm_bo)->vm->__list_name.list); \ + spin_unlock(&(__vm_bo)->vm->__list_name.lock); \ + } while (0) + +/** +* drm_gpuvm_bo_list_del() - remove a vm_bo from the given list +* @__vm_bo: the &drm_gpuvm_bo +* @__list_name: the name of the list to insert into +* +* Removes the given @__vm_bo from the list specified by @__list_name and +* decreases the vm_bo's reference count. +*/ +#define drm_gpuvm_bo_list_del (__vm_bo, __list_name) \ + do { \ + spin_lock(&(__vm_bo)->vm->__list_name.lock); \ + if (!list_empty(&(__vm_bo)->list.entry.__list_name)) \ + list_del_init(&(__vm_bo)->list.entry.__list_name); \ + spin_unlock(&(__vm_bo)->vm->__list_name.lock); \ + } while (0) + +static int __must_check +drm_gpuvm_bo_get_unless_zero (struct drm_gpuvm_bo *vm_bo); +#define to_drm_gpuva (__node) container_of((__node), struct drm_gpuva, rb.node)#define GPUVA_START (node) ((node)->va.addr)@@ -713,6 +867,12 @@ drm_gpuvm_init(struct drm_gpuvm *gpuvm, struct drm_device *drm, gpuvm->rb.tree =RB_ROOT_CACHED ;INIT_LIST_HEAD (&gpuvm->rb.list); +INIT_LIST_HEAD (&gpuvm->extobj.list); +spin_lock_init (&gpuvm->extobj.lock); + +INIT_LIST_HEAD (&gpuvm->evict.list); +spin_lock_init (&gpuvm->evict.lock); +drm_gpuva_check_overflow (start_offset, range); gpuvm->mm_start = start_offset; gpuvm->mm_range = range;@@ -754,10 +914,302 @@ drm_gpuvm_destroy(struct drm_gpuvm *gpuvm)WARN (!RB_EMPTY_ROOT (&gpuvm->rb.tree.rb_root),"GPUVA tree is not empty, potentially leaking memory.\n" ); +WARN (!list_empty (&gpuvm->extobj.list),"Extobj list should be empty.\n" ); +WARN (!list_empty (&gpuvm->evict.list),"Evict list should be empty.\n" ); +drm_gem_private_object_fini (&gpuvm->d_obj); }EXPORT_SYMBOL_GPL (drm_gpuvm_destroy); +/** +* drm_gpuvm_prepare_objects() - prepare all assoiciated BOs +* @gpuvm: the &drm_gpuvm +* @exec: the &drm_exec locking context +* @num_fences: the amount of &dma_fences to reserve +* +* Calls drm_exec_prepare_obj() for all &drm_gem_objects the given +* &drm_gpuvm contains mappings of. +* +* Using this function directly, it is the drivers responsibility to call +* drm_exec_init() and drm_exec_fini() accordingly. +* +* Note: This function is safe against concurrent insertion and removal of +* external objects, however it is not safe against concurrent usage itself. +* +* Drivers need to make sure to protect this case with either an outer VM lock +* or by calling drm_gpuvm_prepare_vm() before this function within the +* drm_exec_until_all_locked() loop, such that the GPUVM's dma-resv lock ensures +* mutual exclusion. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_prepare_objects (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +unsigned int num_fences) +{ +struct drm_gpuvm_bo *vm_bo; +LIST_HEAD (extobjs); +int ret =0 ; + +for_each_vm_bo_in_list (gpuvm, extobj, &extobjs, vm_bo) { + ret =drm_exec_prepare_obj (exec, vm_bo->obj, num_fences); +if (ret) +break ; + } +/* Drop ref in case we break out of the loop. */ +drm_gpuvm_bo_put (vm_bo); +restore_vm_bo_list (gpuvm, extobj, &extobjs); + +return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_prepare_objects); + +/** +* drm_gpuvm_prepare_range() - prepare all BOs mapped within a given range +* @gpuvm: the &drm_gpuvm +* @exec: the &drm_exec locking context +* @addr: the start address within the VA space +* @range: the range to iterate within the VA space +* @num_fences: the amount of &dma_fences to reserve +* +* Calls drm_exec_prepare_obj() for all &drm_gem_objects mapped between @addr +* and @addr + @range. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_prepare_range (struct drm_gpuvm *gpuvm,struct drm_exec *exec, +u64 addr,u64 range,unsigned int num_fences) +{ +struct drm_gpuva *va; +u64 end = addr + range; +int ret; + +drm_gpuvm_for_each_va_range (va, gpuvm, addr, end) { +struct drm_gem_object *obj = va->gem.obj; + + ret =drm_exec_prepare_obj (exec, obj, num_fences); +if (ret) +return ret; + } + +return 0 ; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_prepare_range); + +/** +* drm_gpuvm_exec_lock() - lock all dma-resv of all assoiciated BOs +* @vm_exec: the &drm_gpuvm_exec abstraction +* @num_fences: the amount of &dma_fences to reserve +* @interruptible: sleep interruptible if waiting +* +* Acquires all dma-resv locks of all &drm_gem_objects the given +* &drm_gpuvm contains mappings of. +* +* Addionally, when calling this function with struct drm_gpuvm_exec::extra +* being set the driver receives the given @fn callback to lock additional +* dma-resv in the context of the &drm_gpuvm_exec instance. Typically, drivers +* would call drm_exec_prepare_obj() from within this callback. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_exec_lock (struct drm_gpuvm_exec *vm_exec, +unsigned int num_fences, +bool interruptible) +{ +struct drm_gpuvm *gpuvm = vm_exec->vm; +struct drm_exec *exec = &vm_exec->exec; +uint32_t flags; +int ret; + + flags = interruptible ?DRM_EXEC_INTERRUPTIBLE_WAIT :0 | +DRM_EXEC_IGNORE_DUPLICATES ; + +drm_exec_init (exec, flags); + +drm_exec_until_all_locked (exec) { + ret =drm_gpuvm_prepare_vm (gpuvm, exec, num_fences); +drm_exec_retry_on_contention (exec); +if (ret) + goto err; + + ret =drm_gpuvm_prepare_objects (gpuvm, exec, num_fences); +drm_exec_retry_on_contention (exec); +if (ret) + goto err; + +if (vm_exec->extra.fn) { + ret = vm_exec->extra.fn (vm_exec, num_fences); +drm_exec_retry_on_contention (exec); +if (ret) + goto err; + } + } + +return 0 ; + +err: +drm_exec_fini (exec); +return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock); + +static int +fn_lock_array (struct drm_gpuvm_exec *vm_exec,unsigned int num_fences) +{ +struct { +struct drm_gem_object **objs; +unsigned int num_objs; + } *args = vm_exec->extra.priv; + +return drm_exec_prepare_array (&vm_exec->exec, args->objs, + args->num_objs, num_fences); +} + +/** +* drm_gpuvm_exec_lock_array() - lock all dma-resv of all assoiciated BOs +* @vm_exec: the &drm_gpuvm_exec abstraction +* @objs: additional &drm_gem_objects to lock +* @num_objs: the number of additional &drm_gem_objects to lock +* @num_fences: the amount of &dma_fences to reserve +* @interruptible: sleep interruptible if waiting +* +* Acquires all dma-resv locks of all &drm_gem_objects the given &drm_gpuvm +* contains mappings of, plus the ones given through @objs. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_exec_lock_array (struct drm_gpuvm_exec *vm_exec, +struct drm_gem_object **objs, +unsigned int num_objs, +unsigned int num_fences, +bool interruptible) +{ +struct { +struct drm_gem_object **objs; +unsigned int num_objs; + } args; + + args.objs = objs; + args.num_objs = num_objs; + + vm_exec->extra.fn = fn_lock_array; + vm_exec->extra.priv = &args; + +return drm_gpuvm_exec_lock (vm_exec, num_fences, interruptible); +} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock_array); + +/** +* drm_gpuvm_exec_lock_range() - prepare all BOs mapped within a given range +* @vm_exec: the &drm_gpuvm_exec abstraction +* @addr: the start address within the VA space +* @range: the range to iterate within the VA space +* @num_fences: the amount of &dma_fences to reserve +* @interruptible: sleep interruptible if waiting +* +* Acquires all dma-resv locks of all &drm_gem_objects mapped between @addr and +* @addr + @range. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_exec_lock_range (struct drm_gpuvm_exec *vm_exec, +u64 addr,u64 range, +unsigned int num_fences, +bool interruptible) +{ +struct drm_gpuvm *gpuvm = vm_exec->vm; +struct drm_exec *exec = &vm_exec->exec; +uint32_t flags; +int ret; + + flags = interruptible ?DRM_EXEC_INTERRUPTIBLE_WAIT :0 | +DRM_EXEC_IGNORE_DUPLICATES ; + +drm_exec_init (exec, flags); + +drm_exec_until_all_locked (exec) { + ret =drm_gpuvm_prepare_range (gpuvm, exec, addr, range, + num_fences); +drm_exec_retry_on_contention (exec); +if (ret) + goto err; + } + +return ret; + +err: +drm_exec_fini (exec); +return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock_range); + +/** +* drm_gpuvm_validate() - validate all BOs marked as evicted +* @gpuvm: the &drm_gpuvm to validate evicted BOs +* +* Calls the &drm_gpuvm_ops.bo_validate callback for all evicted buffer +* objects being mapped in the given &drm_gpuvm. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_validate (struct drm_gpuvm *gpuvm) +{ +const struct drm_gpuvm_ops *ops = gpuvm->ops; +struct drm_gpuvm_bo *vm_bo; +LIST_HEAD (evict); +int ret =0 ; + +if (unlikely (!ops || !ops->bo_validate)) +return -ENOTSUPP ; + +for_each_vm_bo_in_list (gpuvm, evict, &evict, vm_bo) { +dma_resv_assert_held (vm_bo->obj->resv); + ret = ops->bo_validate (vm_bo->obj); +if (ret) +break ; + } +/* Drop ref in case we break out of the loop. */ +drm_gpuvm_bo_put (vm_bo); +restore_vm_bo_list (gpuvm, evict, &evict); + +return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_validate); + +/** +* drm_gpuvm_resv_add_fence - add fence to private and all extobj +* dma-resv +* @gpuvm: the &drm_gpuvm to add a fence to +* @exec: the &drm_exec locking context +* @fence: fence to add +* @private_usage: private dma-resv usage +* @extobj_usage: extobj dma-resv usage +*/ +void +drm_gpuvm_resv_add_fence (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +struct dma_fence *fence, +enum dma_resv_usage private_usage, +enum dma_resv_usage extobj_usage) +{ +struct drm_gem_object *obj; +unsigned long index; + +drm_exec_for_each_locked_object (exec, index, obj) { +dma_resv_assert_held (obj->resv); +dma_resv_add_fence (obj->resv, fence, +drm_gpuvm_is_extobj (gpuvm, obj) ? + private_usage : extobj_usage); + } +} +EXPORT_SYMBOL_GPL (drm_gpuvm_resv_add_fence); +/** * drm_gpuvm_bo_create() - create a new instance of struct drm_gpuvm_bo * @gpuvm: The &drm_gpuvm the @obj is mapped in. @@ -790,6 +1242,9 @@ drm_gpuvm_bo_create(struct drm_gpuvm *gpuvm,INIT_LIST_HEAD (&vm_bo->list.gpuva);INIT_LIST_HEAD (&vm_bo->list.entry.gem); +INIT_LIST_HEAD (&vm_bo->list.entry.extobj); +INIT_LIST_HEAD (&vm_bo->list.entry.evict); +drm_gem_object_get (obj);return vm_bo;@@ -807,6 +1262,14 @@ drm_gpuvm_bo_destroy(struct kref *kref)drm_gem_gpuva_assert_lock_held (vm_bo->obj); +spin_lock (&gpuvm->extobj.lock); +list_del (&vm_bo->list.entry.extobj); +spin_unlock (&gpuvm->extobj.lock); + +spin_lock (&gpuvm->evict.lock); +list_del (&vm_bo->list.entry.evict); +spin_unlock (&gpuvm->evict.lock); +list_del (&vm_bo->list.entry.gem);drm_gem_object_put (obj);@@ -822,6 +1285,11 @@ drm_gpuvm_bo_destroy(struct kref *kref)* @vm_bo: the &drm_gpuvm_bo to release the reference of * * This releases a reference to @vm_bo. +* +* If the reference count drops to zero, the &gpuvm_bo is destroyed, which +* includes removing it from the GEMs gpuva list. Hence, if a call to this +* function can potentially let the reference count to zero the caller must +* hold the dma-resv or driver specific GEM gpuva lock. */ void drm_gpuvm_bo_put (struct drm_gpuvm_bo *vm_bo)@@ -831,6 +1299,12 @@ drm_gpuvm_bo_put(struct drm_gpuvm_bo *vm_bo) }EXPORT_SYMBOL_GPL (drm_gpuvm_bo_put); +static int __must_check +drm_gpuvm_bo_get_unless_zero (struct drm_gpuvm_bo *vm_bo) +{ +return kref_get_unless_zero (&vm_bo->kref); +} +static struct drm_gpuvm_bo *__drm_gpuvm_bo_find (struct drm_gpuvm *gpuvm,struct drm_gem_object *obj)@@ -938,6 +1412,48 @@ drm_gpuvm_bo_obtain_prealloc(struct drm_gpuvm_bo *__vm_bo) }EXPORT_SYMBOL_GPL (drm_gpuvm_bo_obtain_prealloc); +/** +* drm_gpuvm_bo_extobj_add() - adds the &drm_gpuvm_bo to its &drm_gpuvm's +* extobj list +* @vm_bo: The &drm_gpuvm_bo to add to its &drm_gpuvm's the extobj list. +* +* Adds the given @vm_bo to its &drm_gpuvm's extobj list if not on the list +* already and if the corresponding &drm_gem_object is an external object, +* actually. +*/ +void +drm_gpuvm_bo_extobj_add (struct drm_gpuvm_bo *vm_bo) +{ +struct drm_gpuvm *gpuvm = vm_bo->vm; + +if (drm_gpuvm_is_extobj (gpuvm, vm_bo->obj)) +drm_gpuvm_bo_list_add (vm_bo, extobj); +} +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_extobj_add); + +/** +* drm_gpuvm_bo_evict() - add / remove a &drm_gem_object to / from a +* &drm_gpuvms evicted list +* @obj: the &drm_gem_object to add or remove +* @evict: indicates whether the object is evicted +* +* Adds a &drm_gem_object to or removes it from all &drm_gpuvms evicted +* list containing a mapping of this &drm_gem_object. +*/ +void +drm_gpuvm_bo_evict (struct drm_gem_object *obj,bool evict) +{ +struct drm_gpuvm_bo *vm_bo; + +drm_gem_for_each_gpuvm_bo (vm_bo, obj) { +if (evict) +drm_gpuvm_bo_list_add (vm_bo, evict); +else +drm_gpuvm_bo_list_del (vm_bo, evict); + } +} +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_evict); +static int __drm_gpuva_insert (struct drm_gpuvm *gpuvm,struct drm_gpuva *va)include/drm/drm_gpuvm.h +197
-0@@ -26,10 +26,12 @@*/ #include <linux/list.h> +#include <linux/dma-resv.h> #include <linux/rbtree.h> #include <linux/types.h> #include <drm/drm_gem.h> +#include <drm/drm_exec.h> struct drm_gpuvm ;struct drm_gpuvm_bo ;@@ -259,6 +261,38 @@ struct drm_gpuvm {* space */ struct dma_resv *resv; + +/** +* @extobj: structure holding the extobj list +*/ +struct { +/** +* @list: &list_head storing &drm_gpuvm_bos serving as +* external object +*/ +struct list_head list; + +/** +* @lock: spinlock to protect the extobj list +*/ +spinlock_t lock; + } extobj; + +/** +* @evict: structure holding the evict list and evict list lock +*/ +struct { +/** +* @list: &list_head storing &drm_gpuvm_bos currently being +* evicted +*/ +struct list_head list; + +/** +* @lock: spinlock to protect the evict list +*/ +spinlock_t lock; + } evict; }; void drm_gpuvm_init(struct drm_gpuvm *gpuvm, struct drm_device *drm,@@ -268,6 +302,21 @@ void drm_gpuvm_init(struct drm_gpuvm *gpuvm, struct drm_device *drm, const struct drm_gpuvm_ops *ops);void drm_gpuvm_destroy (struct drm_gpuvm *gpuvm); +/** +* drm_gpuvm_is_extobj() - indicates whether the given &drm_gem_object is an +* external object +* @gpuvm: the &drm_gpuvm to check +* @obj: the &drm_gem_object to check +* +* Returns: true if the &drm_gem_object &dma_resv differs from the +* &drm_gpuvms &dma_resv, false otherwise +*/ +static inline bool drm_gpuvm_is_extobj (struct drm_gpuvm *gpuvm, +struct drm_gem_object *obj) +{ +return obj && obj->resv != gpuvm->resv; +} +static inline struct drm_gpuva *__drm_gpuva_next (struct drm_gpuva *va) {@@ -346,6 +395,128 @@ __drm_gpuva_next(struct drm_gpuva *va)#define drm_gpuvm_for_each_va_safe (va__, next__, gpuvm__) \ list_for_each_entry_safe(va__, next__, &(gpuvm__)->rb.list, rb.entry) +/** +* struct drm_gpuvm_exec - &drm_gpuvm abstraction of &drm_exec +* +* This structure should be created on the stack as &drm_exec should be. +* +* Optionally, @extra can be set in order to lock additional &drm_gem_objects. +*/ +struct drm_gpuvm_exec { +/** +* @exec: the &drm_exec structure +*/ +struct drm_exec exec; + +/** +* @vm: the &drm_gpuvm to lock its DMA reservations +*/ +struct drm_gpuvm *vm; + +/** +* @extra: Callback and corresponding private data for the driver to +* lock arbitrary additional &drm_gem_objects. +*/ +struct { +/** +* @fn: The driver callback to lock additional &drm_gem_objects. +*/ +int (*fn)(struct drm_gpuvm_exec *vm_exec, +unsigned int num_fences); + +/** +* @priv: driver private data for the @fn callback +*/ +void *priv; + } extra; +}; + +/** +* drm_gpuvm_prepare_vm() - prepare the GPUVMs common dma-resv +* @gpuvm: the &drm_gpuvm +* @exec: the &drm_exec context +* @num_fences: the amount of &dma_fences to reserve +* +* Calls drm_exec_prepare_obj() for the GPUVMs dummy &drm_gem_object. +* +* Using this function directly, it is the drivers responsibility to call +* drm_exec_init() and drm_exec_fini() accordingly. +* +* Returns: 0 on success, negative error code on failure. +*/ +static inline int +drm_gpuvm_prepare_vm (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +unsigned int num_fences) +{ +return drm_exec_prepare_obj (exec, &gpuvm->d_obj, num_fences); +} + +int drm_gpuvm_prepare_objects (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +unsigned int num_fences); + +int drm_gpuvm_prepare_range (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +u64 addr,u64 range, +unsigned int num_fences); + +int drm_gpuvm_exec_lock (struct drm_gpuvm_exec *vm_exec, +unsigned int num_fences, +bool interruptible); + +int drm_gpuvm_exec_lock_array (struct drm_gpuvm_exec *vm_exec, +struct drm_gem_object **objs, +unsigned int num_objs, +unsigned int num_fences, +bool interruptible); + +int drm_gpuvm_exec_lock_range (struct drm_gpuvm_exec *vm_exec, +u64 addr,u64 range, +unsigned int num_fences, +bool interruptible); + +/** +* drm_gpuvm_lock() - lock all dma-resv of all assoiciated BOs +* @gpuvm: the &drm_gpuvm +* +* Releases all dma-resv locks of all &drm_gem_objects previously acquired +* through drm_gpuvm_lock() or its variants. +* +* Returns: 0 on success, negative error code on failure. +*/ +static inline void +drm_gpuvm_exec_unlock (struct drm_gpuvm_exec *vm_exec) +{ +drm_exec_fini (&vm_exec->exec); +} + +int drm_gpuvm_validate (struct drm_gpuvm *gpuvm); +void drm_gpuvm_resv_add_fence (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +struct dma_fence *fence, +enum dma_resv_usage private_usage, +enum dma_resv_usage extobj_usage); + +/** +* drm_gpuvm_exec_resv_add_fence() +* @vm_exec: the &drm_gpuvm_exec abstraction +* @fence: fence to add +* @private_usage: private dma-resv usage +* @extobj_usage: extobj dma-resv usage +* +* See drm_gpuvm_resv_add_fence(). +*/ +static inline void +drm_gpuvm_exec_resv_add_fence (struct drm_gpuvm_exec *vm_exec, +struct dma_fence *fence, +enum dma_resv_usage private_usage, +enum dma_resv_usage extobj_usage) +{ +drm_gpuvm_resv_add_fence (vm_exec->vm, &vm_exec->exec, fence, + private_usage, extobj_usage); +} +/** * struct drm_gpuvm_bo - structure representing a &drm_gpuvm and * &drm_gem_object combination @@ -398,6 +569,18 @@ struct drm_gpuvm_bo {* gpuva list. */ struct list_head gem; + +/** +* @evict: List entry to attach to the &drm_gpuvms +* extobj list. +*/ +struct list_head extobj; + +/** +* @evict: List entry to attach to the &drm_gpuvms evict +* list. +*/ +struct list_head evict; } entry; } list; };@@ -432,6 +615,9 @@ struct drm_gpuvm_bo * drm_gpuvm_bo_find(struct drm_gpuvm *gpuvm, struct drm_gem_object *obj); +void drm_gpuvm_bo_evict (struct drm_gem_object *obj,bool evict); +void drm_gpuvm_bo_extobj_add (struct drm_gpuvm_bo *vm_bo); +/** * drm_gpuvm_bo_for_each_va() - iterator to walk over a list of &drm_gpuva * @va__: &drm_gpuva structure to assign to in each iteration step @@ -837,6 +1023,17 @@ struct drm_gpuvm_ops {* used. */ int (*sm_step_unmap)(struct drm_gpuva_op *op,void *priv); + +/** +* @bo_validate: called from drm_gpuvm_validate() +* +* Drivers receive this callback for every evicted &drm_gem_object being +* mapped in the corresponding &drm_gpuvm. +* +* Typically, drivers would call their driver specific variant of +* ttm_bo_validate() from within this callback. +*/ +int (*bo_validate)(struct drm_gem_object *obj); }; int drm_gpuvm_sm_map(struct drm_gpuvm *gpuvm, void *priv,
Re: [PATCH drm-misc-next v3 6/7] drm/gpuvm: generalize dma_resv/extobj handling and GEM validation
Quoted 113 lines · Thomas Hellström
Quoted 111 lines · Danilo Krummrich
Quoted 95 lines · Thomas Hellström
Hi, Danilo,
Quoted 92 lines · Danilo Krummrich
So far the DRM GPUVA manager offers common infrastructure to track GPU VA allocations and mappings, generically connect GPU VA mappings to their backing buffers and perform more complex mapping operations on the GPU VA space.
However, there are more design patterns commonly used by drivers, which can potentially be generalized in order to make the DRM GPUVA manager represent a basic GPU-VM implementation. In this context, this patch aims at generalizing the following elements.
1) Provide a common dma-resv for GEM objects not being used outside of this GPU-VM.
2) Provide tracking of external GEM objects (GEM objects which are shared with other GPU-VMs).
3) Provide functions to efficiently lock all GEM objects dma-resv the GPU-VM contains mappings of.
4) Provide tracking of evicted GEM objects the GPU-VM contains mappings of, such that validation of evicted GEM objects is accelerated.
5) Provide some convinience functions for common patterns.
Rather than being designed as a "framework", the target is to make all features appear as a collection of optional helper functions, such that drivers are free to make use of the DRM GPUVA managers basic functionality and opt-in for other features without setting any feature flags, just by making use of the corresponding functions.
Big kudos to Boris Brezillon for his help to figure out locking for drivers updating the GPU VA space within the fence signalling path.
Suggested-by Matthew Brost <[email protected]>Signed-off-by Danilo Krummrich <[email protected]>--- drivers/gpu/drm/drm_gpuvm.c | 516 ++++++++++++++++++++++++++++++++++++ include/drm/drm_gpuvm.h | 197 ++++++++++++++ 2 files changed, 713 insertions(+)drivers/gpu/drm/drm_gpuvm.c +33
-0@@ -73,6 +73,21 @@* &drm_gem_object list of &drm_gpuvm_bos for an existing instance of this * particular combination. If not existent a new instance is created and linked * to the &drm_gem_object. +* +* &drm_gpuvm_bo structures, since unique for a given &drm_gpuvm, are also used +* as entry for the &drm_gpuvm's lists of external and evicted objects. Those +* list are maintained in order to accelerate locking of dma-resv locks and +* validation of evicted objects bound in a &drm_gpuvm. For instance the all +* &drm_gem_object's &dma_resv of a given &drm_gpuvm can be locked by calling +* drm_gpuvm_exec_lock(). Once locked drivers can call drm_gpuvm_validate() in +* order to validate all evicted &drm_gem_objects. It is also possible to lock +* additional &drm_gem_objects by providing the corresponding parameters to +* drm_gpuvm_exec_lock() as well as open code the &drm_exec loop while making +* use of helper functions such as drm_gpuvm_prepare_range() or +* drm_gpuvm_prepare_objects(). +* +* Every bound &drm_gem_object is treated as external object when its &dma_resv +* structure is different than the &drm_gpuvm's common &dma_resv structure. */ /** @@ -420,6 +435,20 @@* Subsequent calls to drm_gpuvm_bo_obtain() for the same &drm_gpuvm and * &drm_gem_object must be able to observe previous creations and destructions * of &drm_gpuvm_bos in order to keep instances unique. +* +* The &drm_gpuvm's lists for keeping track of external and evicted objects are +* protected against concurrent insertion / removal and iteration internally. +* +* However, drivers still need ensure to protect concurrent calls to functions +* iterating those lists, such as drm_gpuvm_validate() and +* drm_gpuvm_prepare_objects(). Every such function contains a particular +* comment and lockdep checks if possible. +* +* Functions adding or removing entries from those lists, such as +* drm_gpuvm_bo_evict() or drm_gpuvm_bo_extobj_add() may be called with external +* locks being held, e.g. in order to avoid the corresponding list to be +* (safely) modified while potentially being iternated by other API functions. +* However, this is entirely optional. */ /** @@ -632,6 +661,131 @@* } */ +/** +* get_next_vm_bo_from_list() - get the next vm_bo element +* @__gpuvm: The GPU VM +* @__list_name: The name of the list we're iterating on Quoted 8 lines
drivers/gpu/drm/drm_gpuvm.c line 673
Quoted 6 lines
drivers/gpu/drm/drm_gpuvm.c +6
-0+* @__local_list: A pointer to the local list used to store already iterated items +* @__prev_vm_bo: The previous element we got from drm_gpuvm_get_next_cached_vm_bo() +* +* This helper is here to provide lockless list iteration. Lockless as in, the +* iterator releases the lock immediately after picking the first element from +* the list, so list insertion deletion can happen concurrently. Are the list spinlocks needed for that async state update from within the dma-fence critical section we've discussed previously?
Yes, but also for other reasons, see below.
Quoted 5 lines
Otherwise it should be sufficient to protect the lists with the gpuvm's resv (or for the extobj list with an outer lock).
If those spinlocks are still needed in some situations, perhaps could we have an option to set them to NULL (Like IIRC the maple tree allows for)?
Quoted 8 lines
Quoted 3 lines
The evict spinlock is needed in any case, since in
drm_gpuvm_bo_evict()we're holding only the dma-resv lock from the BO this function gets called for. Hence, the spinlock protects concurrentdrm_gpuvm_bo_evict()calls with different BOs.No. Only if you try to add external objects to the vm's evict list from within the evict code. That's not necessary since you loop through all external objects anyway when locking them so an "evicted" bool in the
vm_bo, protected by the bo resv would be sufficient. The extobj locking loop can then add the bo to the evicted list.
And validate() can remove it while still holding all dma-resv locks, neat! However, what if two tasks are trying to lock the VA space concurrently? What do we do when the drm_gpuvm_bo's refcount drops to zero in drm_gpuva_unlink()? Are we guaranteed that at this point of time the drm_gpuvm_bo is not on the evicted list? Because otherwise we would call drm_gpuvm_bo_destroy() with the dma-resv lock held, which wouldn't be allowed, since drm_gpuvm_bo_destroy() might drop the last reference to the drm_gem_object and hence we'd potentially free the dma-resv lock while holding it, at least if it's an external object.
Quoted 5 lines
Quoted 4 lines
For extobjs an outer lock would be enough in case of Xe, but I really would not like to add even more complexity just to get the spinlock out of the way in case the driver already has an outer lock protecting this path.
Quoted 8 lines
I must disagree here. These spinlocks and atomic operations are pretty costly and as discussed earlier this type of locking was the reason (at least according to the commit message) that made Christian drop the XArray use in
drm_execfor the same set of objects: "The locking overhead is unecessary and measurable". IMHO the spinlock is the added complexity and a single wide lock following the drm locking guidelines set out by Daniel and David should really be the default choice with an opt-in for a spinlock if needed for async and pushing out to a wq is not an option.
For the external object list an outer lock would work as long as it's not the dma-resv lock of the corresponding GEM object, since here we actually need to remove the list entry from the external object list on drm_gpuvm_bo_destroy(). It's just a bit weird design wise that drivers would need to take this outer lock on:
- drm_gpuvm_bo_extobj_add()
- drm_gpuvm_bo_destroy() (and hence also drm_gpuvm_bo_put())
- drm_gpuva_unlink() (because it needs to call drm_gpuvm_bo_put())
- drm_gpuvm_exec_lock()
- drm_gpuvm_exec_lock_array()
- drm_gpuvm_prepare_range()
Given that it seems reasonable to do all the required locking internally.
In order to at least place lockdep checks, the driver would need to supply the corresponding lock's lockdep_map, because the GPUVM otherwise doesn't know about the lock.
Out of curiosity, what is the overhead of a spin_lock() that doesn't need to spin?
Quoted 21 lines
A pretty simple way that would not add much code would be
static void
gpuvm_cond_spin_lock(const struct drm_gpuvm *gpuvm, spinlock_t *lock){
if (!gpuvm->resv_protected_lists)
spin_lock(lock);}
Quoted 7 lines
Quoted 2 lines
For such drivers, that would require anybody calling unlink to hold the vm's resv, though.
In V4 I want to go back to having a dedicated lock for the GEMs gpuva list (or
VM_BOlist to be more precise). We can't just use the dma-resv lock for that withVM_BOabstractions, because on destruction of aVM_BOwe otherwise wouldn't be allowed to already hold the dma-resv lock. That's the fix I was referring to earlier.
Quoted 5 lines
Yeah, I can see the need for a dedicated lock for the GEM's gpuva list, but holding the vm's dma-resv lock across the unlink shouldn't be a problem. We may free the object and a pointer to the vm's resv during unlink but we don't free the vm's resv. It'd be a matter of ensuring that any calls to unlink from *within*
drm_gpuvmallows it to be held.
Drivers calling unlink() from the fence signaling path can't use the VM's dma-resv lock.
Also, what if the object is an external object? We can't use the VM's dma-resv lock here. And we can't have the GEM objs dma-resv lock held when calling unlink(), since unlink() calls drm_gpuvm_bo_put(), which if the refcount drops to zero calls drm_gpuvm_bo_destroy() and drm_gpuvm_bo_destroy() might drop the last reference of the GEM object. All those problems go away with a dedicated GEM gpuva list lock.
Quoted 782 lines
/Thomas
Quoted 777 lines
Quoted 777 lines
It seems that with that also the refcount could be make non-atomic.
All in the spirit of the drm locking guidelines "use big locks when possible".
Lower level locks only when necessary for performance or locking inversion?/Thomas
Quoted 768 lines
drivers/gpu/drm/drm_gpuvm.c +477
-0+* +* Elements popped from the original list are kept in a local list, so removal +* and is_empty checks can still happen while we're iterating the list. +*/ +#define get_next_vm_bo_from_list (__gpuvm, __list_name, __local_list, __prev_vm_bo) \ + ({ \ + struct drm_gpuvm_bo *__vm_bo; \ + \ + drm_gpuvm_bo_put(__prev_vm_bo); \ + \ + spin_lock(&(__gpuvm)->__list_name.lock); \ + while (!list_empty(&(__gpuvm)->__list_name.list)) { \ + __vm_bo = list_first_entry(&(__gpuvm)->__list_name.list, \ + struct drm_gpuvm_bo, \ + list.entry.__list_name); \ + if (drm_gpuvm_bo_get_unless_zero(__vm_bo)) { \ + list_move_tail(&(__vm_bo)->list.entry.__list_name, \ + __local_list); \ + break; \ + } else { \ + list_del_init(&(__vm_bo)->list.entry.__list_name); \ + __vm_bo = NULL; \ + } \ + } \ + spin_unlock(&(__gpuvm)->__list_name.lock); \ + \ + __vm_bo; \ + }) + +/** +* for_each_vm_bo_in_list() - internal vm_bo list iterator +* +* This helper is here to provide lockless list iteration. Lockless as in, the +* iterator releases the lock immediately after picking the first element from the +* list, so list insertion and deletion can happen concurrently. +* +* Typical use: +* +* struct drm_gpuvm_bo *vm_bo; +* LIST_HEAD(my_local_list); +* +* ret = 0; +* drm_gpuvm_for_each_vm_bo(gpuvm, <list_name>, &my_local_list, vm_bo) { +* ret = do_something_with_vm_bo(..., vm_bo); +* if (ret) +* break; +* } +* drm_gpuvm_bo_put(vm_bo); +* drm_gpuvm_restore_vm_bo_list(gpuvm, <list_name>, &my_local_list); +* +* +* Only used for internal list iterations, not meant to be exposed to the outside +* world. +*/ +#define for_each_vm_bo_in_list (__gpuvm, __list_name, __local_list, __vm_bo) \ + for (__vm_bo = get_next_vm_bo_from_list(__gpuvm, __list_name, \ + __local_list, NULL); \ + __vm_bo; \ + __vm_bo = get_next_vm_bo_from_list(__gpuvm, __list_name, \ + __local_list, __vm_bo)) \ + +/** +* restore_vm_bo_list() - move vm_bo elements back to their original list +* @__gpuvm: The GPU VM +* @__list_name: The name of the list we're iterating on +* @__local_list: A pointer to the local list used to store already iterated items +* +* When we're done iterating a vm_bo list, we should call restore_vm_bo_list() +* to restore the original state and let new iterations take place. +*/ +#define restore_vm_bo_list (__gpuvm, __list_name, __local_list) \ + do { \ +/* Merge back the two lists, moving local list elements to the \ +* head to preserve previous ordering, in case it matters. \ +*/ \ +spin_lock (&(__gpuvm)->__list_name.lock); \ +list_splice (__local_list, &(__gpuvm)->__list_name.list); \ +spin_unlock (&(__gpuvm)->__list_name.lock); \ + }while (0 ) +/** +* drm_gpuvm_bo_list_add() - insert a vm_bo into the given list +* @__vm_bo: the &drm_gpuvm_bo +* @__list_name: the name of the list to insert into +* +* Inserts the given @__vm_bo into the list specified by @__list_name and +* increases the vm_bo's reference count. + */ +#define drm_gpuvm_bo_list_add (__vm_bo, __list_name) \ + do { \ + spin_lock(&(__vm_bo)->vm->__list_name.lock); \ + if (list_empty(&(__vm_bo)->list.entry.__list_name)) \ + list_add_tail(&(__vm_bo)->list.entry.__list_name, \ + &(__vm_bo)->vm->__list_name.list); \ + spin_unlock(&(__vm_bo)->vm->__list_name.lock); \ + } while (0) + +/** +* drm_gpuvm_bo_list_del() - remove a vm_bo from the given list +* @__vm_bo: the &drm_gpuvm_bo +* @__list_name: the name of the list to insert into +* +* Removes the given @__vm_bo from the list specified by @__list_name and +* decreases the vm_bo's reference count. +*/ +#define drm_gpuvm_bo_list_del (__vm_bo, __list_name) \ + do { \ + spin_lock(&(__vm_bo)->vm->__list_name.lock); \ + if (!list_empty(&(__vm_bo)->list.entry.__list_name)) \ + list_del_init(&(__vm_bo)->list.entry.__list_name); \ + spin_unlock(&(__vm_bo)->vm->__list_name.lock); \ + } while (0) + +static int __must_check +drm_gpuvm_bo_get_unless_zero (struct drm_gpuvm_bo *vm_bo); +#define to_drm_gpuva (__node) container_of((__node), struct drm_gpuva, rb.node)#define GPUVA_START (node) ((node)->va.addr)@@ -713,6 +867,12 @@ drm_gpuvm_init(struct drm_gpuvm *gpuvm, struct drm_device *drm, gpuvm->rb.tree =RB_ROOT_CACHED ;INIT_LIST_HEAD (&gpuvm->rb.list); +INIT_LIST_HEAD (&gpuvm->extobj.list); +spin_lock_init (&gpuvm->extobj.lock); + +INIT_LIST_HEAD (&gpuvm->evict.list); +spin_lock_init (&gpuvm->evict.lock); +drm_gpuva_check_overflow (start_offset, range); gpuvm->mm_start = start_offset; gpuvm->mm_range = range;@@ -754,10 +914,302 @@ drm_gpuvm_destroy(struct drm_gpuvm *gpuvm)WARN (!RB_EMPTY_ROOT (&gpuvm->rb.tree.rb_root),"GPUVA tree is not empty, potentially leaking memory.\n" ); +WARN (!list_empty (&gpuvm->extobj.list),"Extobj list should be empty.\n" ); +WARN (!list_empty (&gpuvm->evict.list),"Evict list should be empty.\n" ); +drm_gem_private_object_fini (&gpuvm->d_obj); }EXPORT_SYMBOL_GPL (drm_gpuvm_destroy); +/** +* drm_gpuvm_prepare_objects() - prepare all assoiciated BOs +* @gpuvm: the &drm_gpuvm +* @exec: the &drm_exec locking context +* @num_fences: the amount of &dma_fences to reserve +* +* Calls drm_exec_prepare_obj() for all &drm_gem_objects the given +* &drm_gpuvm contains mappings of. +* +* Using this function directly, it is the drivers responsibility to call +* drm_exec_init() and drm_exec_fini() accordingly. +* +* Note: This function is safe against concurrent insertion and removal of +* external objects, however it is not safe against concurrent usage itself. +* +* Drivers need to make sure to protect this case with either an outer VM lock +* or by calling drm_gpuvm_prepare_vm() before this function within the +* drm_exec_until_all_locked() loop, such that the GPUVM's dma-resv lock ensures +* mutual exclusion. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_prepare_objects (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +unsigned int num_fences) +{ +struct drm_gpuvm_bo *vm_bo; +LIST_HEAD (extobjs); +int ret =0 ; + +for_each_vm_bo_in_list (gpuvm, extobj, &extobjs, vm_bo) { + ret =drm_exec_prepare_obj (exec, vm_bo->obj, num_fences); +if (ret) +break ; + } +/* Drop ref in case we break out of the loop. */ +drm_gpuvm_bo_put (vm_bo); +restore_vm_bo_list (gpuvm, extobj, &extobjs); + +return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_prepare_objects); + +/** +* drm_gpuvm_prepare_range() - prepare all BOs mapped within a given range +* @gpuvm: the &drm_gpuvm +* @exec: the &drm_exec locking context +* @addr: the start address within the VA space +* @range: the range to iterate within the VA space +* @num_fences: the amount of &dma_fences to reserve +* +* Calls drm_exec_prepare_obj() for all &drm_gem_objects mapped between @addr +* and @addr + @range. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_prepare_range (struct drm_gpuvm *gpuvm,struct drm_exec *exec, +u64 addr,u64 range,unsigned int num_fences) +{ +struct drm_gpuva *va; +u64 end = addr + range; +int ret; + +drm_gpuvm_for_each_va_range (va, gpuvm, addr, end) { +struct drm_gem_object *obj = va->gem.obj; + + ret =drm_exec_prepare_obj (exec, obj, num_fences); +if (ret) +return ret; + } + +return 0 ; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_prepare_range); + +/** +* drm_gpuvm_exec_lock() - lock all dma-resv of all assoiciated BOs +* @vm_exec: the &drm_gpuvm_exec abstraction +* @num_fences: the amount of &dma_fences to reserve +* @interruptible: sleep interruptible if waiting +* +* Acquires all dma-resv locks of all &drm_gem_objects the given +* &drm_gpuvm contains mappings of. +* +* Addionally, when calling this function with struct drm_gpuvm_exec::extra +* being set the driver receives the given @fn callback to lock additional +* dma-resv in the context of the &drm_gpuvm_exec instance. Typically, drivers +* would call drm_exec_prepare_obj() from within this callback. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_exec_lock (struct drm_gpuvm_exec *vm_exec, +unsigned int num_fences, +bool interruptible) +{ +struct drm_gpuvm *gpuvm = vm_exec->vm; +struct drm_exec *exec = &vm_exec->exec; +uint32_t flags; +int ret; + + flags = interruptible ?DRM_EXEC_INTERRUPTIBLE_WAIT :0 | +DRM_EXEC_IGNORE_DUPLICATES ; + +drm_exec_init (exec, flags); + +drm_exec_until_all_locked (exec) { + ret =drm_gpuvm_prepare_vm (gpuvm, exec, num_fences); +drm_exec_retry_on_contention (exec); +if (ret) + goto err; + + ret =drm_gpuvm_prepare_objects (gpuvm, exec, num_fences); +drm_exec_retry_on_contention (exec); +if (ret) + goto err; + +if (vm_exec->extra.fn) { + ret = vm_exec->extra.fn (vm_exec, num_fences); +drm_exec_retry_on_contention (exec); +if (ret) + goto err; + } + } + +return 0 ; + +err: +drm_exec_fini (exec); +return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock); + +static int +fn_lock_array (struct drm_gpuvm_exec *vm_exec,unsigned int num_fences) +{ +struct { +struct drm_gem_object **objs; +unsigned int num_objs; + } *args = vm_exec->extra.priv; + +return drm_exec_prepare_array (&vm_exec->exec, args->objs, + args->num_objs, num_fences); +} + +/** +* drm_gpuvm_exec_lock_array() - lock all dma-resv of all assoiciated BOs +* @vm_exec: the &drm_gpuvm_exec abstraction +* @objs: additional &drm_gem_objects to lock +* @num_objs: the number of additional &drm_gem_objects to lock +* @num_fences: the amount of &dma_fences to reserve +* @interruptible: sleep interruptible if waiting +* +* Acquires all dma-resv locks of all &drm_gem_objects the given &drm_gpuvm +* contains mappings of, plus the ones given through @objs. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_exec_lock_array (struct drm_gpuvm_exec *vm_exec, +struct drm_gem_object **objs, +unsigned int num_objs, +unsigned int num_fences, +bool interruptible) +{ +struct { +struct drm_gem_object **objs; +unsigned int num_objs; + } args; + + args.objs = objs; + args.num_objs = num_objs; + + vm_exec->extra.fn = fn_lock_array; + vm_exec->extra.priv = &args; + +return drm_gpuvm_exec_lock (vm_exec, num_fences, interruptible); +} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock_array); + +/** +* drm_gpuvm_exec_lock_range() - prepare all BOs mapped within a given range +* @vm_exec: the &drm_gpuvm_exec abstraction +* @addr: the start address within the VA space +* @range: the range to iterate within the VA space +* @num_fences: the amount of &dma_fences to reserve +* @interruptible: sleep interruptible if waiting +* +* Acquires all dma-resv locks of all &drm_gem_objects mapped between @addr and +* @addr + @range. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_exec_lock_range (struct drm_gpuvm_exec *vm_exec, +u64 addr,u64 range, +unsigned int num_fences, +bool interruptible) +{ +struct drm_gpuvm *gpuvm = vm_exec->vm; +struct drm_exec *exec = &vm_exec->exec; +uint32_t flags; +int ret; + + flags = interruptible ?DRM_EXEC_INTERRUPTIBLE_WAIT :0 | +DRM_EXEC_IGNORE_DUPLICATES ; + +drm_exec_init (exec, flags); + +drm_exec_until_all_locked (exec) { + ret =drm_gpuvm_prepare_range (gpuvm, exec, addr, range, + num_fences); +drm_exec_retry_on_contention (exec); +if (ret) + goto err; + } + +return ret; + +err: +drm_exec_fini (exec); +return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock_range); + +/** +* drm_gpuvm_validate() - validate all BOs marked as evicted +* @gpuvm: the &drm_gpuvm to validate evicted BOs +* +* Calls the &drm_gpuvm_ops.bo_validate callback for all evicted buffer +* objects being mapped in the given &drm_gpuvm. +* +* Returns: 0 on success, negative error code on failure. +*/ +int +drm_gpuvm_validate (struct drm_gpuvm *gpuvm) +{ +const struct drm_gpuvm_ops *ops = gpuvm->ops; +struct drm_gpuvm_bo *vm_bo; +LIST_HEAD (evict); +int ret =0 ; + +if (unlikely (!ops || !ops->bo_validate)) +return -ENOTSUPP ; + +for_each_vm_bo_in_list (gpuvm, evict, &evict, vm_bo) { +dma_resv_assert_held (vm_bo->obj->resv); + ret = ops->bo_validate (vm_bo->obj); +if (ret) +break ; + } +/* Drop ref in case we break out of the loop. */ +drm_gpuvm_bo_put (vm_bo); +restore_vm_bo_list (gpuvm, evict, &evict); + +return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_validate); + +/** +* drm_gpuvm_resv_add_fence - add fence to private and all extobj +* dma-resv +* @gpuvm: the &drm_gpuvm to add a fence to +* @exec: the &drm_exec locking context +* @fence: fence to add +* @private_usage: private dma-resv usage +* @extobj_usage: extobj dma-resv usage +*/ +void +drm_gpuvm_resv_add_fence (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +struct dma_fence *fence, +enum dma_resv_usage private_usage, +enum dma_resv_usage extobj_usage) +{ +struct drm_gem_object *obj; +unsigned long index; + +drm_exec_for_each_locked_object (exec, index, obj) { +dma_resv_assert_held (obj->resv); +dma_resv_add_fence (obj->resv, fence, +drm_gpuvm_is_extobj (gpuvm, obj) ? + private_usage : extobj_usage); + } +} +EXPORT_SYMBOL_GPL (drm_gpuvm_resv_add_fence); +/** * drm_gpuvm_bo_create() - create a new instance of struct drm_gpuvm_bo * @gpuvm: The &drm_gpuvm the @obj is mapped in. @@ -790,6 +1242,9 @@ drm_gpuvm_bo_create(struct drm_gpuvm *gpuvm,INIT_LIST_HEAD (&vm_bo->list.gpuva);INIT_LIST_HEAD (&vm_bo->list.entry.gem); +INIT_LIST_HEAD (&vm_bo->list.entry.extobj); +INIT_LIST_HEAD (&vm_bo->list.entry.evict); +drm_gem_object_get (obj);return vm_bo;@@ -807,6 +1262,14 @@ drm_gpuvm_bo_destroy(struct kref *kref)drm_gem_gpuva_assert_lock_held (vm_bo->obj); +spin_lock (&gpuvm->extobj.lock); +list_del (&vm_bo->list.entry.extobj); +spin_unlock (&gpuvm->extobj.lock); + +spin_lock (&gpuvm->evict.lock); +list_del (&vm_bo->list.entry.evict); +spin_unlock (&gpuvm->evict.lock); +list_del (&vm_bo->list.entry.gem);drm_gem_object_put (obj);@@ -822,6 +1285,11 @@ drm_gpuvm_bo_destroy(struct kref *kref)* @vm_bo: the &drm_gpuvm_bo to release the reference of * * This releases a reference to @vm_bo. +* +* If the reference count drops to zero, the &gpuvm_bo is destroyed, which +* includes removing it from the GEMs gpuva list. Hence, if a call to this +* function can potentially let the reference count to zero the caller must +* hold the dma-resv or driver specific GEM gpuva lock. */ void drm_gpuvm_bo_put (struct drm_gpuvm_bo *vm_bo)@@ -831,6 +1299,12 @@ drm_gpuvm_bo_put(struct drm_gpuvm_bo *vm_bo) }EXPORT_SYMBOL_GPL (drm_gpuvm_bo_put); +static int __must_check +drm_gpuvm_bo_get_unless_zero (struct drm_gpuvm_bo *vm_bo) +{ +return kref_get_unless_zero (&vm_bo->kref); +} +static struct drm_gpuvm_bo *__drm_gpuvm_bo_find (struct drm_gpuvm *gpuvm,struct drm_gem_object *obj)@@ -938,6 +1412,48 @@ drm_gpuvm_bo_obtain_prealloc(struct drm_gpuvm_bo *__vm_bo) }EXPORT_SYMBOL_GPL (drm_gpuvm_bo_obtain_prealloc); +/** +* drm_gpuvm_bo_extobj_add() - adds the &drm_gpuvm_bo to its &drm_gpuvm's +* extobj list +* @vm_bo: The &drm_gpuvm_bo to add to its &drm_gpuvm's the extobj list. +* +* Adds the given @vm_bo to its &drm_gpuvm's extobj list if not on the list +* already and if the corresponding &drm_gem_object is an external object, +* actually. +*/ +void +drm_gpuvm_bo_extobj_add (struct drm_gpuvm_bo *vm_bo) +{ +struct drm_gpuvm *gpuvm = vm_bo->vm; + +if (drm_gpuvm_is_extobj (gpuvm, vm_bo->obj)) +drm_gpuvm_bo_list_add (vm_bo, extobj); +} +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_extobj_add); + +/** +* drm_gpuvm_bo_evict() - add / remove a &drm_gem_object to / from a +* &drm_gpuvms evicted list +* @obj: the &drm_gem_object to add or remove +* @evict: indicates whether the object is evicted +* +* Adds a &drm_gem_object to or removes it from all &drm_gpuvms evicted +* list containing a mapping of this &drm_gem_object. +*/ +void +drm_gpuvm_bo_evict (struct drm_gem_object *obj,bool evict) +{ +struct drm_gpuvm_bo *vm_bo; + +drm_gem_for_each_gpuvm_bo (vm_bo, obj) { +if (evict) +drm_gpuvm_bo_list_add (vm_bo, evict); +else +drm_gpuvm_bo_list_del (vm_bo, evict); + } +} +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_evict); +static int __drm_gpuva_insert (struct drm_gpuvm *gpuvm,struct drm_gpuva *va)include/drm/drm_gpuvm.h +197
-0@@ -26,10 +26,12 @@*/ #include <linux/list.h> +#include <linux/dma-resv.h> #include <linux/rbtree.h> #include <linux/types.h> #include <drm/drm_gem.h> +#include <drm/drm_exec.h> struct drm_gpuvm ;struct drm_gpuvm_bo ;@@ -259,6 +261,38 @@ struct drm_gpuvm {* space */ struct dma_resv *resv; + +/** +* @extobj: structure holding the extobj list +*/ +struct { +/** +* @list: &list_head storing &drm_gpuvm_bos serving as +* external object +*/ +struct list_head list; + +/** +* @lock: spinlock to protect the extobj list +*/ +spinlock_t lock; + } extobj; + +/** +* @evict: structure holding the evict list and evict list lock +*/ +struct { +/** +* @list: &list_head storing &drm_gpuvm_bos currently being +* evicted +*/ +struct list_head list; + +/** +* @lock: spinlock to protect the evict list +*/ +spinlock_t lock; + } evict; }; void drm_gpuvm_init(struct drm_gpuvm *gpuvm, struct drm_device *drm,@@ -268,6 +302,21 @@ void drm_gpuvm_init(struct drm_gpuvm *gpuvm, struct drm_device *drm, const struct drm_gpuvm_ops *ops);void drm_gpuvm_destroy (struct drm_gpuvm *gpuvm); +/** +* drm_gpuvm_is_extobj() - indicates whether the given &drm_gem_object is an +* external object +* @gpuvm: the &drm_gpuvm to check +* @obj: the &drm_gem_object to check +* +* Returns: true if the &drm_gem_object &dma_resv differs from the +* &drm_gpuvms &dma_resv, false otherwise +*/ +static inline bool drm_gpuvm_is_extobj (struct drm_gpuvm *gpuvm, +struct drm_gem_object *obj) +{ +return obj && obj->resv != gpuvm->resv; +} +static inline struct drm_gpuva *__drm_gpuva_next (struct drm_gpuva *va) {@@ -346,6 +395,128 @@ __drm_gpuva_next(struct drm_gpuva *va)#define drm_gpuvm_for_each_va_safe (va__, next__, gpuvm__) \ list_for_each_entry_safe(va__, next__, &(gpuvm__)->rb.list, rb.entry) +/** +* struct drm_gpuvm_exec - &drm_gpuvm abstraction of &drm_exec +* +* This structure should be created on the stack as &drm_exec should be. +* +* Optionally, @extra can be set in order to lock additional &drm_gem_objects. +*/ +struct drm_gpuvm_exec { +/** +* @exec: the &drm_exec structure +*/ +struct drm_exec exec; + +/** +* @vm: the &drm_gpuvm to lock its DMA reservations +*/ +struct drm_gpuvm *vm; + +/** +* @extra: Callback and corresponding private data for the driver to +* lock arbitrary additional &drm_gem_objects. +*/ +struct { +/** +* @fn: The driver callback to lock additional &drm_gem_objects. +*/ +int (*fn)(struct drm_gpuvm_exec *vm_exec, +unsigned int num_fences); + +/** +* @priv: driver private data for the @fn callback +*/ +void *priv; + } extra; +}; + +/** +* drm_gpuvm_prepare_vm() - prepare the GPUVMs common dma-resv +* @gpuvm: the &drm_gpuvm +* @exec: the &drm_exec context +* @num_fences: the amount of &dma_fences to reserve +* +* Calls drm_exec_prepare_obj() for the GPUVMs dummy &drm_gem_object. +* +* Using this function directly, it is the drivers responsibility to call +* drm_exec_init() and drm_exec_fini() accordingly. +* +* Returns: 0 on success, negative error code on failure. +*/ +static inline int +drm_gpuvm_prepare_vm (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +unsigned int num_fences) +{ +return drm_exec_prepare_obj (exec, &gpuvm->d_obj, num_fences); +} + +int drm_gpuvm_prepare_objects (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +unsigned int num_fences); + +int drm_gpuvm_prepare_range (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +u64 addr,u64 range, +unsigned int num_fences); + +int drm_gpuvm_exec_lock (struct drm_gpuvm_exec *vm_exec, +unsigned int num_fences, +bool interruptible); + +int drm_gpuvm_exec_lock_array (struct drm_gpuvm_exec *vm_exec, +struct drm_gem_object **objs, +unsigned int num_objs, +unsigned int num_fences, +bool interruptible); + +int drm_gpuvm_exec_lock_range (struct drm_gpuvm_exec *vm_exec, +u64 addr,u64 range, +unsigned int num_fences, +bool interruptible); + +/** +* drm_gpuvm_lock() - lock all dma-resv of all assoiciated BOs +* @gpuvm: the &drm_gpuvm +* +* Releases all dma-resv locks of all &drm_gem_objects previously acquired +* through drm_gpuvm_lock() or its variants. +* +* Returns: 0 on success, negative error code on failure. +*/ +static inline void +drm_gpuvm_exec_unlock (struct drm_gpuvm_exec *vm_exec) +{ +drm_exec_fini (&vm_exec->exec); +} + +int drm_gpuvm_validate (struct drm_gpuvm *gpuvm); +void drm_gpuvm_resv_add_fence (struct drm_gpuvm *gpuvm, +struct drm_exec *exec, +struct dma_fence *fence, +enum dma_resv_usage private_usage, +enum dma_resv_usage extobj_usage); + +/** +* drm_gpuvm_exec_resv_add_fence() +* @vm_exec: the &drm_gpuvm_exec abstraction +* @fence: fence to add +* @private_usage: private dma-resv usage +* @extobj_usage: extobj dma-resv usage +* +* See drm_gpuvm_resv_add_fence(). +*/ +static inline void +drm_gpuvm_exec_resv_add_fence (struct drm_gpuvm_exec *vm_exec, +struct dma_fence *fence, +enum dma_resv_usage private_usage, +enum dma_resv_usage extobj_usage) +{ +drm_gpuvm_resv_add_fence (vm_exec->vm, &vm_exec->exec, fence, + private_usage, extobj_usage); +} +/** * struct drm_gpuvm_bo - structure representing a &drm_gpuvm and * &drm_gem_object combination @@ -398,6 +569,18 @@ struct drm_gpuvm_bo {* gpuva list. */ struct list_head gem; + +/** +* @evict: List entry to attach to the &drm_gpuvms +* extobj list. +*/ +struct list_head extobj; + +/** +* @evict: List entry to attach to the &drm_gpuvms evict +* list. +*/ +struct list_head evict; } entry; } list; };@@ -432,6 +615,9 @@ struct drm_gpuvm_bo * drm_gpuvm_bo_find(struct drm_gpuvm *gpuvm, struct drm_gem_object *obj); +void drm_gpuvm_bo_evict (struct drm_gem_object *obj,bool evict); +void drm_gpuvm_bo_extobj_add (struct drm_gpuvm_bo *vm_bo); +/** * drm_gpuvm_bo_for_each_va() - iterator to walk over a list of &drm_gpuva * @va__: &drm_gpuva structure to assign to in each iteration step @@ -837,6 +1023,17 @@ struct drm_gpuvm_ops {* used. */ int (*sm_step_unmap)(struct drm_gpuva_op *op,void *priv); + +/** +* @bo_validate: called from drm_gpuvm_validate() +* +* Drivers receive this callback for every evicted &drm_gem_object being +* mapped in the corresponding &drm_gpuvm. +* +* Typically, drivers would call their driver specific variant of +* ttm_bo_validate() from within this callback. +*/ +int (*bo_validate)(struct drm_gem_object *obj); }; int drm_gpuvm_sm_map(struct drm_gpuvm *gpuvm, void *priv,
Thomas Hellström
reviewerRe: [PATCH drm-misc-next v3 6/7] drm/gpuvm: generalize dma_resv/extobj handling and GEM validation
Hi!
Quoted 180 lines · Danilo Krummrich
Quoted 169 lines · Thomas Hellström
Quoted 161 lines · Danilo Krummrich
Quoted 142 lines · Thomas Hellström
Hi, Danilo,
Quoted 139 lines · Danilo Krummrich
So far the DRM GPUVA manager offers common infrastructure to track GPU VA
allocations and mappings, generically connect GPU VA mappings to their
backing buffers and perform more complex mapping operations on the GPU VA
space.However, there are more design patterns commonly used by drivers, which
can potentially be generalized in order to make the DRM GPUVA manager
represent a basic GPU-VM implementation. In this context, this patch aims
at generalizing the following elements.1) Provide a common dma-resv for GEM objects not being used outside of
this GPU-VM.2) Provide tracking of external GEM objects (GEM objects which are
shared with other GPU-VMs).3) Provide functions to efficiently lock all GEM objects dma- resv the
GPU-VM contains mappings of.4) Provide tracking of evicted GEM objects the GPU-VM contains mappings
of, such that validation of evicted GEM objects is accelerated.5) Provide some convinience functions for common patterns.
Rather than being designed as a "framework", the target is to make all
features appear as a collection of optional helper functions, such that
drivers are free to make use of the DRM GPUVA managers basic functionality and opt-in for other features without setting any feature
flags, just by making use of the corresponding functions.Big kudos to Boris Brezillon for his help to figure out locking for drivers
updating the GPU VA space within the fence signalling path.Suggested-by Matthew Brost <[email protected]>Signed-off-by Danilo Krummrich <[email protected]>---
drivers/gpu/drm/drm_gpuvm.c| 516
++++++++++++++++++++++++++++++++++++
include/drm/drm_gpuvm.h | 197 ++++++++++++++2 files changed, 713 insertions(+)drivers/gpu/drm/drm_gpuvm.c +0
-0@@ -73,6 +73,21 @@ * &drm_gem_object list of &drm_gpuvm_bos for an existinginstance of this
* particular combination. If not existent a new instance is created and linked
* to the&drm_gem_object.
+ *
+ *&drm_gpuvm_bostructures, since unique for a given&drm_gpuvm, are also used
+ * as entry for the &drm_gpuvm's lists of external and evicted objects. Those
+ * list are maintained in order to accelerate locking of dma-resv locks and
+ * validation of evicted objects bound in a&drm_gpuvm. For instance the all
+ * &drm_gem_object's&dma_resvof a given&drm_gpuvmcan be locked by calling
+ *drm_gpuvm_exec_lock(). Once locked drivers can calldrm_gpuvm_validate()in
+ * order to validate all evicted&drm_gem_objects. It is also possible to lock
+ * additional&drm_gem_objectsby providing the corresponding parameters to
+ *drm_gpuvm_exec_lock()as well as open code the&drm_execloop while making
+ * use of helper functions such asdrm_gpuvm_prepare_range()or
+ *drm_gpuvm_prepare_objects().
+ *
+ * Every bound&drm_gem_objectis treated as external object when its&dma_resv
+ * structure is different than the &drm_gpuvm's common&dma_resvstructure.
*/
/**drivers/gpu/drm/drm_gpuvm.c +0
-0@@ -420,6 +435,20 @@ * Subsequent calls to drm_gpuvm_bo_obtain() for the same
&drm_gpuvmand
*&drm_gem_objectmust be able to observe previous creations and destructions
* of&drm_gpuvm_bosin order to keep instances unique.
+ *
+ * The &drm_gpuvm's lists for keeping track of external and evicted objects are
+ * protected against concurrent insertion / removal and iteration internally.
+ *
+ * However, drivers still need ensure to protect concurrent calls to functions
+ * iterating those lists, such asdrm_gpuvm_validate()and
+ *drm_gpuvm_prepare_objects(). Every such function contains a particular
+ * comment and lockdep checks if possible.
+ *
+ * Functions adding or removing entries from those lists, such as
+ *drm_gpuvm_bo_evict()ordrm_gpuvm_bo_extobj_add()may be called with external
+ * locks being held, e.g. in order to avoid the corresponding list to be
+ * (safely) modified while potentially being iternated by other API functions.
+ * However, this is entirely optional.
*/
/**drivers/gpu/drm/drm_gpuvm.c +5
-0@@ -632,6 +661,131 @@* } */ +/** +* get_next_vm_bo_from_list() - get the next vm_bo element +* @__gpuvm: The GPU VM +* @__list_name: The name of the list we're iterating on +* @__local_list: A pointer to the local list used to store already iterated items
+ * @__prev_vm_bo: The previous element we got fromdrm_gpuvm_get_next_cached_vm_bo()
+ *
+ * This helper is here to provide lockless list iteration.Quoted 8 lines
Quoted 5 lines
Lockless as in, the
+ * iterator releases the lock immediately after picking the first element from
+ * the list, so list insertion deletion can happen concurrently.Are the list spinlocks needed for that async state update from within the
dma-fence critical section we've discussed previously?Yes, but also for other reasons, see below.
Quoted 8 lines
Otherwise it should be sufficient to protect the lists with the gpuvm's resv
(or for the extobj list with an outer lock).If those spinlocks are still needed in some situations, perhaps could we
have an option to set them to NULL (Like IIRC the maple tree allows for)?Quoted 6 lines
The evict spinlock is needed in any case, since in
drm_gpuvm_bo_evict()we're
holding only the dma-resv lock from the BO this function gets called for. Hence,
the spinlock protects concurrentdrm_gpuvm_bo_evict()calls with different BOs.Quoted 9 lines
No. Only if you try to add external objects to the vm's evict list from
within the evict code. That's not necessary since you loop through all
external objects anyway when locking them so an "evicted" bool in thevm_bo,
protected by the bo resv would be sufficient. The extobj locking loop can
then add the bo to the evicted list.
Quoted 16 lines
And
validate()can remove it while still holding all dma-resv locks, neat!
However, what if two tasks are trying to lock the VA space concurrently? What
do we do when the drm_gpuvm_bo's refcount drops to zero indrm_gpuva_unlink()?
Are we guaranteed that at this point of time thedrm_gpuvm_bois not on the
evicted list? Because otherwise we would calldrm_gpuvm_bo_destroy()with the
dma-resv lock held, which wouldn't be allowed, sincedrm_gpuvm_bo_destroy()
might drop the last reference to thedrm_gem_objectand hence we'd potentially
free the dma-resv lock while holding it, at least if it's an external object.
Easiest way in this scheme is to think of the lists as being protected by the vm's resv lock. That means anybody calling unlink() must also hold the vm's resv lock. (Which is OK from an UAF point of view, but perhaps not from a locking inversion POW from an async list update).
Quoted 42 lines
Quoted 8 lines
Quoted 7 lines
For extobjs an outer lock would be enough in case of Xe, but I really would not
like to add even more complexity just to get the spinlock out of the way in case
the driver already has an outer lock protecting this path.Quoted 15 lines
I must disagree here. These spinlocks and atomic operations are pretty
costly and as discussed earlier this type of locking was the reason (at
least according to the commit message) that made Christian drop the XArray
use indrm_execfor the same set of objects: "The locking overhead is
unecessary and measurable". IMHO the spinlock is the added complexity and a
single wide lock following the drm locking guidelines set out by Daniel and
David should really be the default choice with an opt-in for a spinlock if
needed for async and pushing out to a wq is not an option.For the external object list an outer lock would work as long as it's not the
dma-resv lock of the corresponding GEM object, since here we actually need to
remove the list entry from the external object list ondrm_gpuvm_bo_destroy().
It's just a bit weird design wise that drivers would need to take this outer
lock on:-
drm_gpuvm_bo_extobj_add()
- drm_gpuvm_bo_destroy() (and hence alsodrm_gpuvm_bo_put())
- drm_gpuva_unlink() (because it needs to calldrm_gpuvm_bo_put())
-drm_gpuvm_exec_lock()
-drm_gpuvm_exec_lock_array()
-drm_gpuvm_prepare_range()
Quoted 2 lines
Given that it seems reasonable to do all the required locking internally.
From a design POW, there has been a clear direction in XE to make things similar to mmap() / munmap(), so this outer lock, which in Xe is an rwsem, is used in a similar way as the mmap_lock. It's protecting the page-table structures and vma rb tree, the userptr structures and the extobj list. Basically it's taken early in the exec IOCTL, the VM_BIND ioctl, the compute rebind worker and the pagefault handler, so all of the above are just asserting that it is taken in the correct mode.
But strictly with this scheme one could also use the vm's dma_resv for the extobj list since with drm_exec, it's locked before traversing the list.
The whole point of this scheme is to rely on locks that you already are supposed to be holding for various reasons and is simple to comprehend.
Quoted 6 lines
In order to at least place lockdep checks, the driver would need to supply the
corresponding lock'slockdep_map, because the GPUVM otherwise doesn't know about
the lock.
Yes, that sounds reasonable. One lockdep map per list.
Quoted 4 lines
Out of curiosity, what is the overhead of a
spin_lock()that doesn't need to
spin?
I guess it's hard to tell exactly, but it is much lower on modern x86 than what it used to be. Not sure about ARM, which is the other architecture important to us. I figure if there is little cache-line bouncing the main overhead comes from the implied barriers.
Quoted 38 lines
Quoted 28 lines
A pretty simple way that would not add much code would be
static void gpuvm_cond_spin_lock(const struct
drm_gpuvm*gpuvm,spinlock_t
*lock){
if (!gpuvm->resv_protected_lists)
spin_lock(lock);}
Quoted 12 lines
Quoted 3 lines
For such drivers, that would require anybody calling unlink to hold the vm's
resv, though.In V4 I want to go back to having a dedicated lock for the GEMs gpuva list (or
VM_BOlist to be more precise). We can't just use the dma-resv lock for that
withVM_BOabstractions, because on destruction of aVM_BOwe otherwise wouldn't
be allowed to already hold the dma-resv lock. That's the fix I was referring to
earlier.Quoted 9 lines
Yeah, I can see the need for a dedicated lock for the GEM's gpuva list, but
holding the vm's dma-resv lock across the unlink shouldn't be a problem. We
may free the object and a pointer to the vm's resv during unlink but we
don't free the vm's resv. It'd be a matter of ensuring that any calls to
unlink from *within*drm_gpuvmallows it to be held.
Quoted 3 lines
Drivers calling
unlink()from the fence signaling path can't use the VM's
dma-resv lock.
Yes, that made me a bit curious because in the current version the code required the object's dma_resv for unlink() which can't be grabbed either from the fence signaling path. So are there any drivers actually wanting to do that? If so, they will either need to resort to the current spinlock solution or they will need to call unlink from a workqueue item.
Quoted 4 lines
Also, what if the object is an external object? We can't use the VM's dma-resv
lock here.
Why? Typically (sync) unlink is only ever called from an unbind-like operation where it should be trivial to grab the vm's resv. Or, for that matter any outer lock protecting the extobj list. Rule would be the drm_gpuvm_bo::entry::extobj and drm_gpuvm_bo::entry::evict would be protected by either the vm's dma_resv (or possibly an outer lock in the case of the extobj list).
Quoted 6 lines
And we can't have the GEM objs dma-resv lock held when calling
unlink(), sinceunlink()callsdrm_gpuvm_bo_put(), which if the refcount drops
to zero callsdrm_gpuvm_bo_destroy()anddrm_gpuvm_bo_destroy()might drop the
last reference of the GEM object.
Yes, but this is a different problem as to what exactly protects drm_gpuvm_bo::entry::gem. Either as you suggest an internal per bo list lock, or if we want to keep the bo's dma_resv we need to ensure that the caller of dma_resv_unlock(obj->resv) actually refcounts its obj pointer, and doesn't implicitly rely on the gpuvm_bo's refcount (I know Boris didn't like that, but requiring an explicit refcount for a pointer you dereference unless you're under a lock that ensures keeping the object alive is pretty much required?) But anyway for the drm_gpuvm_bo::entry::gem list protection (bo resv or internal spinlock) I don't have a strong preference.
Quoted 2 lines
All those problems go away with a dedicated
GEM gpuva list lock.
I don't think these are real problems.
With the excepton of the eviction list "trick" where we currently have slightly different approach to collect external bos needing rebinding, we have this working fine.
TBH I think pretty much the only situation where the spinlock is needed is for async updates of these lists, unless a wq item can be used for that, but it doesn't really seem like the current code allows for such updates anyway? It complicates the code a lot, adds overhead and also adds the requirement for refcounting during list traversal.
/Thomas
Quoted 959 lines
Quoted 958 lines
/Thomas
Quoted 952 lines
Quoted 952 lines
It seems that with that also the refcount could be make non- atomic.
All in the spirit of the drm locking guidelines "use big locks when
possible".
Lower level locks only when necessary for performance or locking inversion?/Thomas
Quoted 940 lines
+ *
+ * Elements popped from the original list are kept in a local list, so removal
+ * andis_emptychecks can still happen while we're iterating the list.
+ */
+#defineget_next_vm_bo_from_list(__gpuvm, __list_name, __local_list, __prev_vm_bo)\
+ ({
\
+ structdrm_gpuvm_bo
*__vm_bo; \
+
\
+ drm_gpuvm_bo_put(__prev_vm_bo); \
+
\
+ spin_lock(&(__gpuvm)-Quoted 1 line
__list_name.lock); \
+ while (!list_empty(&(__gpuvm)-
Quoted 1 line
__list_name.list)) { \
+ __vm_bo = list_first_entry(&(__gpuvm)->__list_name.list, \
+ struct drm_gpuvm_bo, \
+
list.entry.__list_name); \
+ if (drm_gpuvm_bo_get_unless_zero(__vm_bo))
{ \
+ list_move_tail(&(__vm_bo)-Quoted 1 line
list.entry.__list_name, \
+
__local_list); \
+ break;
\
+ } else
{ \
+ list_del_init(&(__vm_bo)-Quoted 1 line
list.entry.__list_name); \
+ __vm_bo =
NULL; \
+ }
\
+ }
\
+ spin_unlock(&(__gpuvm)-Quoted 1 line
__list_name.lock); \
+
\
+ __vm_bo;
\
+ })drivers/gpu/drm/drm_gpuvm.c +5
-0+ +/** +* for_each_vm_bo_in_list() - internal vm_bo list iterator +* +* This helper is here to provide lockless list iteration. Lockless as in, the
+ * iterator releases the lock immediately after picking the first element from the
+ * list, so list insertion and deletion can happen concurrently.
+ *
+ * Typical use:
+ *
+ * structdrm_gpuvm_bo*vm_bo;
+ * LIST_HEAD(my_local_list);
+ *
+ * ret = 0;
+ * drm_gpuvm_for_each_vm_bo(gpuvm, <list_name>,&my_local_list,vm_bo) {
+ * ret =do_something_with_vm_bo(..., vm_bo);
+ * if (ret)
+ * break;
+ * }
+ * drm_gpuvm_bo_put(vm_bo);
+ * drm_gpuvm_restore_vm_bo_list(gpuvm, <list_name>,&my_local_list);
+ *
+ *
+ * Only used for internal list iterations, not meant to be exposed to the outside
+ * world.
+ */
+#definefor_each_vm_bo_in_list(__gpuvm, __list_name, __local_list, __vm_bo)\
+ for (__vm_bo= get_next_vm_bo_from_list(__gpuvm, __list_name, \
+ __local_list, NULL); \
+__vm_bo;
\
+__vm_bo= get_next_vm_bo_from_list(__gpuvm, __list_name, \
+ __local_list, __vm_bo)) \drivers/gpu/drm/drm_gpuvm.c +3
-0+ +/** +* restore_vm_bo_list() - move vm_bo elements back to their original list
+ * @__gpuvm: The GPU VM
+ * @__list_name: The name of the list we're iterating on
+ * @__local_list: A pointer to the local list used to store already iterated items
+ *
+ * When we're done iterating avm_bolist, we should callrestore_vm_bo_list()
+ * to restore the original state and let new iterations take place.
+ */
+#definerestore_vm_bo_list(__gpuvm, __list_name, __local_list)\
+ do
{
\
+ /* Merge back the two lists, moving local list elements to the \
+ * head to preserve previous ordering, in case it matters. \
+
*/
\
+ spin_lock(&(__gpuvm)-Quoted 1 line
__list_name.lock); \
+ list_splice(__local_list, &(__gpuvm)-
Quoted 1 line
__list_name.list); \
+ spin_unlock(&(__gpuvm)-
Quoted 1 line
__list_name.lock); \
+ } while (0)
+/**
+ *drm_gpuvm_bo_list_add()- insert avm_bointo the given list
+ * @__vm_bo: the&drm_gpuvm_bo
+ * @__list_name: the name of the list to insert into
+ *
+ * Inserts the given @__vm_bo into the list specified by @__list_name and
+ * increases the vm_bo's reference count.
+ */
+#definedrm_gpuvm_bo_list_add(__vm_bo, __list_name)\
+ do
{
\
+ spin_lock(&(__vm_bo)->vm-Quoted 1 line
__list_name.lock); \
+ if (list_empty(&(__vm_bo)-
Quoted 1 line
list.entry.__list_name)) \
+ list_add_tail(&(__vm_bo)-
Quoted 1 line
list.entry.__list_name, \
+ &(__vm_bo)->vm-
Quoted 1 line
__list_name.list); \
+ spin_unlock(&(__vm_bo)->vm-
Quoted 1 line
__list_name.lock); \
+ } while (0)
drivers/gpu/drm/drm_gpuvm.c +3
-0+ +/** +* drm_gpuvm_bo_list_del() - remove a vm_bo from the given list
+ * @__vm_bo: the&drm_gpuvm_bo
+ * @__list_name: the name of the list to insert into
+ *
+ * Removes the given @__vm_bo from the list specified by @__list_name and
+ * decreases the vm_bo's reference count.
+ */
+#definedrm_gpuvm_bo_list_del(__vm_bo, __list_name)\
+ do
{
\
+ spin_lock(&(__vm_bo)->vm-Quoted 1 line
__list_name.lock); \
+ if (!list_empty(&(__vm_bo)-
Quoted 1 line
list.entry.__list_name)) \
+ list_del_init(&(__vm_bo)-
Quoted 1 line
list.entry.__list_name); \
+ spin_unlock(&(__vm_bo)->vm-
Quoted 1 line
__list_name.lock); \
+ } while (0)
drivers/gpu/drm/drm_gpuvm.c +4
-0+ +static int __must_check +drm_gpuvm_bo_get_unless_zero (struct drm_gpuvm_bo *vm_bo); +#define to_drm_gpuva(__node) container_of((__node), struct
drm_gpuva, rb.node)
#defineGPUVA_START(node)((node)->va.addr)drivers/gpu/drm/drm_gpuvm.c +6
-0@@ -713,6 +867,12 @@ drm_gpuvm_init(struct drm_gpuvm *gpuvm,struct drm_device *drm, gpuvm->rb.tree = RB_ROOT_CACHED; INIT_LIST_HEAD(&gpuvm->rb.list); + INIT_LIST_HEAD(&gpuvm->extobj.list); + spin_lock_init(&gpuvm->extobj.lock); + + INIT_LIST_HEAD(&gpuvm->evict.list); + spin_lock_init(&gpuvm->evict.lock); + drm_gpuva_check_overflow(start_offset, range); gpuvm->mm_start = start_offset; gpuvm->mm_range = range;@@ -754,10 +914,302 @@ drm_gpuvm_destroy(struct drm_gpuvm *gpuvm) WARN(!RB_EMPTY_ROOT(&gpuvm->rb.tree.rb_root),"GPUVA tree is not empty, potentially leaking memory.\n");
+ WARN(!list_empty(&gpuvm->extobj.list), "Extobj list should be empty.\n");
+ WARN(!list_empty(&gpuvm->evict.list), "Evict list should be empty.\n");
+
drm_gem_private_object_fini(&gpuvm->d_obj); }
EXPORT_SYMBOL_GPL(drm_gpuvm_destroy);
+/**
+ *drm_gpuvm_prepare_objects()- prepare all assoiciated BOs
+ * @gpuvm: the&drm_gpuvm
+ * @exec: the&drm_execlocking context
+ * @num_fences: the amount of&dma_fencesto reserve
+ *
+ * Callsdrm_exec_prepare_obj()for all&drm_gem_objectsthe given
+ *&drm_gpuvmcontains mappings of.
+ *
+ * Using this function directly, it is the drivers responsibility to call
+ *drm_exec_init()anddrm_exec_fini()accordingly.
+ *
+ * Note: This function is safe against concurrent insertion and removal of
+ * external objects, however it is not safe against concurrent usage itself.
+ *
+ * Drivers need to make sure to protect this case with either an outer VM lock
+ * or by callingdrm_gpuvm_prepare_vm()before this function within the
+ *drm_exec_until_all_locked()loop, such that the GPUVM's dma-resv lock ensures
+ * mutual exclusion.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +10
-0+int +drm_gpuvm_prepare_objects (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + unsigned int num_fences) +{ + struct drm_gpuvm_bo *vm_bo; + LIST_HEAD(extobjs); + int ret = 0; + + for_each_vm_bo_in_list(gpuvm, extobj, &extobjs,
vm_bo) {
+ ret =drm_exec_prepare_obj(exec, vm_bo->obj, num_fences);
+ if (ret)
+ break;
+ }
+ /* Drop ref in case we break out of the loop. */ + drm_gpuvm_bo_put(vm_bo);
+ restore_vm_bo_list(gpuvm, extobj, &extobjs);drivers/gpu/drm/drm_gpuvm.c +7
-0+ + return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_prepare_objects); + +/** +* drm_gpuvm_prepare_range() - prepare all BOs mapped within a given range
+ * @gpuvm: the&drm_gpuvm
+ * @exec: the&drm_execlocking context
+ * @addr: the start address within the VA space
+ * @range: the range to iterate within the VA space
+ * @num_fences: the amount of&dma_fencesto reserve
+ *
+ * Callsdrm_exec_prepare_obj()for all&drm_gem_objectsmapped between @addr
+ * and @addr + @range.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +2
-0+int +drm_gpuvm_prepare_range(struct drm_gpuvm *gpuvm, struct
drm_exec*exec,
+ u64 addr, u64 range, unsigned intnum_fences)drivers/gpu/drm/drm_gpuvm.c +9
-0+{ + struct drm_gpuva *va; + u64 end = addr + range; + int ret; + + drm_gpuvm_for_each_va_range(va, gpuvm, addr, end) { + struct drm_gem_object *obj = va->gem.obj; + + ret = drm_exec_prepare_obj(exec, obj,
num_fences);
+ if (ret)
+ return ret;
+ }drivers/gpu/drm/drm_gpuvm.c +7
-0+ + return 0; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_prepare_range); + +/** +* drm_gpuvm_exec_lock() - lock all dma-resv of all assoiciated BOs
+ * @vm_exec: the&drm_gpuvm_execabstraction
+ * @num_fences: the amount of&dma_fencesto reserve
+ * @interruptible: sleep interruptible if waiting
+ *
+ * Acquires all dma-resv locks of all&drm_gem_objectsthe given
+ *&drm_gpuvmcontains mappings of.
+ *
+ * Addionally, when calling this function with structdrm_gpuvm_exec::extra
+ * being set the driver receives the given @fn callback to lock additional
+ * dma-resv in the context of the&drm_gpuvm_execinstance. Typically, drivers
+ * would calldrm_exec_prepare_obj()from within this callback.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +11
-0+int +drm_gpuvm_exec_lock (struct drm_gpuvm_exec *vm_exec, + unsigned int num_fences, + bool interruptible) +{ + struct drm_gpuvm *gpuvm = vm_exec->vm; + struct drm_exec *exec = &vm_exec->exec; + uint32_t flags; + int ret; + + flags = interruptible ? DRM_EXEC_INTERRUPTIBLE_WAIT :0 |
+ DRM_EXEC_IGNORE_DUPLICATES;drivers/gpu/drm/drm_gpuvm.c +5
-0+ + drm_exec_init(exec, flags); + + drm_exec_until_all_locked(exec) { + ret = drm_gpuvm_prepare_vm(gpuvm, exec,
num_fences);
+ drm_exec_retry_on_contention(exec); + if (ret)
+ goto err;
+
+ ret =drm_gpuvm_prepare_objects(gpuvm, exec, num_fences);
+ drm_exec_retry_on_contention(exec); + if (ret)
+ goto err;
+
+ if (vm_exec->extra.fn) {
+ ret =vm_exec->extra.fn(vm_exec, num_fences);
+ drm_exec_retry_on_contention(exec); + if (ret)
+ goto err;
+ }
+ }drivers/gpu/drm/drm_gpuvm.c +11
-0+ + return 0; + +err: + drm_exec_fini(exec); + return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock); + +static int +fn_lock_array(struct drm_gpuvm_exec *vm_exec, unsigned int
num_fences)drivers/gpu/drm/drm_gpuvm.c +7
-0+{ + struct { + struct drm_gem_object **objs; + unsigned int num_objs; + } *args = vm_exec->extra.priv; + + return drm_exec_prepare_array(&vm_exec->exec, args-Quoted 1 line
objs,
+
args->num_objs,num_fences);drivers/gpu/drm/drm_gpuvm.c +4
-0+} + +/** +* drm_gpuvm_exec_lock_array() - lock all dma-resv of all assoiciated BOs
+ * @vm_exec: the&drm_gpuvm_execabstraction
+ * @objs: additional&drm_gem_objectsto lock
+ * @num_objs: the number of additional&drm_gem_objectsto lock
+ * @num_fences: the amount of&dma_fencesto reserve
+ * @interruptible: sleep interruptible if waiting
+ *
+ * Acquires all dma-resv locks of all&drm_gem_objectsthe given&drm_gpuvm
+ * contains mappings of, plus the ones given through @objs.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +19
-0+int +drm_gpuvm_exec_lock_array (struct drm_gpuvm_exec *vm_exec, + struct drm_gem_object **objs, + unsigned int num_objs, + unsigned int num_fences, + bool interruptible) +{ + struct { + struct drm_gem_object **objs; + unsigned int num_objs; + } args; + + args.objs = objs; + args.num_objs = num_objs; + + vm_exec->extra.fn = fn_lock_array; + vm_exec->extra.priv = &args; + + return drm_gpuvm_exec_lock(vm_exec, num_fences,interruptible);
drivers/gpu/drm/drm_gpuvm.c +5
-0+} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock_array); + +/** +* drm_gpuvm_exec_lock_range() - prepare all BOs mapped within a given range
+ * @vm_exec: the&drm_gpuvm_execabstraction
+ * @addr: the start address within the VA space
+ * @range: the range to iterate within the VA space
+ * @num_fences: the amount of&dma_fencesto reserve
+ * @interruptible: sleep interruptible if waiting
+ *
+ * Acquires all dma-resv locks of all&drm_gem_objectsmapped between @addr and
+ * @addr + @range.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +12
-0+int +drm_gpuvm_exec_lock_range (struct drm_gpuvm_exec *vm_exec, + u64 addr, u64 range, + unsigned int num_fences, + bool interruptible) +{ + struct drm_gpuvm *gpuvm = vm_exec->vm; + struct drm_exec *exec = &vm_exec->exec; + uint32_t flags; + int ret; + + flags = interruptible ? DRM_EXEC_INTERRUPTIBLE_WAIT :0 |
+ DRM_EXEC_IGNORE_DUPLICATES;drivers/gpu/drm/drm_gpuvm.c +5
-0+ + drm_exec_init(exec, flags); + + drm_exec_until_all_locked(exec) { + ret = drm_gpuvm_prepare_range(gpuvm, exec,addr, range,
+num_fences); + drm_exec_retry_on_contention(exec); + if (ret)
+ goto err;
+ }drivers/gpu/drm/drm_gpuvm.c +14
-0+ + return ret; + +err: + drm_exec_fini(exec); + return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock_range); + +/** +* drm_gpuvm_validate() - validate all BOs marked as evicted +* @gpuvm: the &drm_gpuvm to validate evicted BOs +* +* Calls the &drm_gpuvm_ops.bo_validate callback for all evicted buffer
+ * objects being mapped in the given&drm_gpuvm.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +27
-0+int +drm_gpuvm_validate (struct drm_gpuvm *gpuvm) +{ + const struct drm_gpuvm_ops *ops = gpuvm->ops; + struct drm_gpuvm_bo *vm_bo; + LIST_HEAD(evict); + int ret = 0; + + if (unlikely(!ops || !ops->bo_validate)) + return -ENOTSUPP; + + for_each_vm_bo_in_list(gpuvm, evict, &evict, vm_bo) { + dma_resv_assert_held(vm_bo->obj->resv); + ret = ops->bo_validate(vm_bo->obj); + if (ret) + break; + } + /* Drop ref in case we break out of the loop. */ + drm_gpuvm_bo_put(vm_bo); + restore_vm_bo_list(gpuvm, evict, &evict); + + return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_validate); + +/** +* drm_gpuvm_resv_add_fence - add fence to private and all extobj
+ * dma-resv
+ * @gpuvm: the&drm_gpuvmto add a fence to
+ * @exec: the&drm_execlocking context
+ * @fence: fence to add
+ * @private_usage: private dma-resv usage
+ * @extobj_usage: extobj dma-resv usage
+ */drivers/gpu/drm/drm_gpuvm.c +14
-0+void +drm_gpuvm_resv_add_fence (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + struct dma_fence *fence, + enum dma_resv_usage private_usage, + enum dma_resv_usage extobj_usage) +{ + struct drm_gem_object *obj; + unsigned long index; + + drm_exec_for_each_locked_object(exec, index, obj) { + dma_resv_assert_held(obj->resv); + dma_resv_add_fence(obj->resv, fence, + drm_gpuvm_is_extobj(gpuvm,obj) ?
+private_usage:extobj_usage);
+ }drivers/gpu/drm/drm_gpuvm.c +3
-0+} +EXPORT_SYMBOL_GPL (drm_gpuvm_resv_add_fence); +/**
*drm_gpuvm_bo_create()- create a new instance of structdrm_gpuvm_bo
* @gpuvm: The&drm_gpuvmthe @obj is mapped in.drivers/gpu/drm/drm_gpuvm.c +13
-0@@ -790,6 +1242,9 @@ drm_gpuvm_bo_create(struct drm_gpuvm *gpuvm, INIT_LIST_HEAD(&vm_bo->list.gpuva); INIT_LIST_HEAD(&vm_bo->list.entry.gem); + INIT_LIST_HEAD(&vm_bo->list.entry.extobj); + INIT_LIST_HEAD(&vm_bo->list.entry.evict); + drm_gem_object_get(obj); return vm_bo;@@ -807,6 +1262,14 @@ drm_gpuvm_bo_destroy(struct kref *kref) drm_gem_gpuva_assert_lock_held(vm_bo->obj); + spin_lock(&gpuvm->extobj.lock); + list_del(&vm_bo->list.entry.extobj); + spin_unlock(&gpuvm->extobj.lock); + + spin_lock(&gpuvm->evict.lock); + list_del(&vm_bo->list.entry.evict); + spin_unlock(&gpuvm->evict.lock); + list_del(&vm_bo->list.entry.gem); drm_gem_object_put(obj);@@ -822,6 +1285,11 @@ drm_gpuvm_bo_destroy(struct kref *kref) * @vm_bo: the &drm_gpuvm_bo to release the reference of * * This releases a reference to @vm_bo. + * + * If the reference count drops to zero, the &gpuvm_bo isdestroyed, which
+ * includes removing it from the GEMs gpuva list. Hence, if a call to this
+ * function can potentially let the reference count to zero the caller must
+ * hold the dma-resv or driver specific GEM gpuva lock. */
void
drm_gpuvm_bo_put(struct drm_gpuvm_bo *vm_bo)drivers/gpu/drm/drm_gpuvm.c +6
-0@@ -831,6 +1299,12 @@ drm_gpuvm_bo_put(struct drm_gpuvm_bo *vm_bo) } EXPORT_SYMBOL_GPL(drm_gpuvm_bo_put); +static int __must_check +drm_gpuvm_bo_get_unless_zero (struct drm_gpuvm_bo *vm_bo) +{ + return kref_get_unless_zero(&vm_bo->kref); +} +static struct
drm_gpuvm_bo*
__drm_gpuvm_bo_find(structdrm_gpuvm*gpuvm,
structdrm_gem_object*obj)drivers/gpu/drm/drm_gpuvm.c +2
-0@@ -938,6 +1412,48 @@ drm_gpuvm_bo_obtain_prealloc(struct drm_gpuvm_bo *__vm_bo) } EXPORT_SYMBOL_GPL(drm_gpuvm_bo_obtain_prealloc); +/** +* drm_gpuvm_bo_extobj_add() - adds the &drm_gpuvm_bo to its &drm_gpuvm's
+ * extobj list
+ * @vm_bo: The&drm_gpuvm_boto add to its &drm_gpuvm's the extobj list.
+ *
+ * Adds the given @vm_bo to its &drm_gpuvm's extobj list if not on the list
+ * already and if the corresponding&drm_gem_objectis an external object,
+ * actually.
+ */drivers/gpu/drm/drm_gpuvm.c +12
-0+void +drm_gpuvm_bo_extobj_add (struct drm_gpuvm_bo *vm_bo) +{ + struct drm_gpuvm *gpuvm = vm_bo->vm; + + if (drm_gpuvm_is_extobj(gpuvm, vm_bo->obj)) + drm_gpuvm_bo_list_add(vm_bo, extobj); +} +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_extobj_add); + +/** +* drm_gpuvm_bo_evict() - add / remove a &drm_gem_object to / from a
+ *&drm_gpuvmsevicted list
+ * @obj: the&drm_gem_objectto add or remove
+ * @evict: indicates whether the object is evicted
+ *
+ * Adds a&drm_gem_objectto or removes it from all&drm_gpuvmsevicted
+ * list containing a mapping of this&drm_gem_object.
+ */drivers/gpu/drm/drm_gpuvm.c +14
-0+void +drm_gpuvm_bo_evict (struct drm_gem_object *obj,bool evict) +{ + struct drm_gpuvm_bo *vm_bo; + + drm_gem_for_each_gpuvm_bo(vm_bo, obj) { + if (evict) + drm_gpuvm_bo_list_add(vm_bo, evict); + else + drm_gpuvm_bo_list_del(vm_bo, evict); + } +} +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_evict); +static int
__drm_gpuva_insert(structdrm_gpuvm*gpuvm,
structdrm_gpuva*va)include/drm/drm_gpuvm.h +9
-0@@ -26,10 +26,12 @@*/ #include <linux/list.h> +#include <linux/dma-resv.h> #include <linux/rbtree.h> #include <linux/types.h> #include <drm/drm_gem.h> +#include <drm/drm_exec.h> struct drm_gpuvm; struct drm_gpuvm_bo;@@ -259,6 +261,38 @@ struct drm_gpuvm {* space */ struct dma_resv *resv; + + /** + * @extobj: structure holding the extobj list + */ + struct { + /** +* @list: &list_head storing &drm_gpuvm_bos serving as
+ * external object
+ */
+ structlist_headlist;include/drm/drm_gpuvm.h +9
-0+ + /** + * @lock: spinlock to protect the extobj list + */ + spinlock_t lock; + } extobj; + + /** +* @evict: structure holding the evict list and evict list lock
+ */
+ struct {
+ /**
+ * @list:&list_headstoring&drm_gpuvm_boscurrently being
+ * evicted
+ */
+ structlist_headlist;
+
+ /**
+ * @lock: spinlock to protect the evict list + */
+ spinlock_t lock;
+ } evict;
};
void drm_gpuvm_init(structdrm_gpuvm*gpuvm, structdrm_device*drm,include/drm/drm_gpuvm.h +2
-0@@ -268,6 +302,21 @@ void drm_gpuvm_init(struct drm_gpuvm *gpuvm, struct drm_device *drm, const struct drm_gpuvm_ops *ops); void drm_gpuvm_destroy(struct drm_gpuvm *gpuvm); +/** +* drm_gpuvm_is_extobj() - indicates whether the given
&drm_gem_objectis an
+ * external object
+ * @gpuvm: the&drm_gpuvmto check
+ * @obj: the&drm_gem_objectto check
+ *
+ * Returns: true if the&drm_gem_object&dma_resvdiffers from the
+ *&drm_gpuvms&dma_resv, false otherwise
+ */
+static inline bool drm_gpuvm_is_extobj(structdrm_gpuvm*gpuvm,
+ structdrm_gem_object*obj)include/drm/drm_gpuvm.h +4
-0+{ + return obj && obj->resv != gpuvm->resv; +} +static inline struct
drm_gpuva*
__drm_gpuva_next(struct drm_gpuva *va)
{include/drm/drm_gpuvm.h +0
-0@@ -346,6 +395,128 @@ __drm_gpuva_next(struct drm_gpuva *va) #define drm_gpuvm_for_each_va_safe(va__, next__, gpuvm__) \list_for_each_entry_safe(va__,
next__, &(gpuvm__)-Quoted 1 line
rb.list, rb.entry)
+/**
+ * structdrm_gpuvm_exec-&drm_gpuvmabstraction of&drm_exec
+ *
+ * This structure should be created on the stack as&drm_execshould be.
+ *
+ * Optionally, @extra can be set in order to lock additional&drm_gem_objects.
+ */include/drm/drm_gpuvm.h +13
-0+struct drm_gpuvm_exec { + /** + * @exec: the &drm_exec structure + */ + struct drm_exec exec; + + /** + * @vm: the &drm_gpuvm to lock its DMA reservations + */ + struct drm_gpuvm *vm; + + /** + * @extra: Callback and corresponding private datafor the driver to
+ * lock arbitrary additional&drm_gem_objects. + */
+ struct {
+ /**
+ * @fn: The driver callback to lock additional&drm_gem_objects.
+ */
+ int (*fn)(structdrm_gpuvm_exec*vm_exec, + unsigned intnum_fences); +
+ /**
+ * @priv: driver private data for the @fn callback
+ */
+ void *priv;
+ } extra;include/drm/drm_gpuvm.h +4
-0+}; + +/** +* drm_gpuvm_prepare_vm() - prepare the GPUVMs common dma- resv
+ * @gpuvm: the&drm_gpuvm
+ * @exec: the&drm_execcontext
+ * @num_fences: the amount of&dma_fencesto reserve
+ *
+ * Callsdrm_exec_prepare_obj()for the GPUVMs dummy&drm_gem_object.
+ *
+ * Using this function directly, it is the drivers responsibility to call
+ *drm_exec_init()anddrm_exec_fini()accordingly.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */include/drm/drm_gpuvm.h +6
-0+static inline int +drm_gpuvm_prepare_vm (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + unsigned int num_fences) +{ + return drm_exec_prepare_obj(exec, &gpuvm->d_obj,
num_fences);include/drm/drm_gpuvm.h +16
-0+} + +int drm_gpuvm_prepare_objects (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + unsigned int num_fences); + +int drm_gpuvm_prepare_range (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + u64 addr, u64 range, + unsigned int num_fences); + +int drm_gpuvm_exec_lock (struct drm_gpuvm_exec *vm_exec, + unsigned int num_fences, + bool interruptible); + +int drm_gpuvm_exec_lock_array(struct drm_gpuvm_exec
*vm_exec,
+ structdrm_gem_object**objs, + unsigned intnum_objs, + unsigned intnum_fences, + bool interruptible);include/drm/drm_gpuvm.h +2
-0+ +int drm_gpuvm_exec_lock_range(struct drm_gpuvm_exec
*vm_exec,
+ u64 addr, u64 range, + unsigned intnum_fences, + bool interruptible);include/drm/drm_gpuvm.h +3
-0+ +/** +* drm_gpuvm_lock() - lock all dma-resv of all assoiciated BOs
+ * @gpuvm: the&drm_gpuvm
+ *
+ * Releases all dma-resv locks of all&drm_gem_objectspreviously acquired
+ * throughdrm_gpuvm_lock()or its variants.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */include/drm/drm_gpuvm.h +11
-0+static inline void +drm_gpuvm_exec_unlock (struct drm_gpuvm_exec *vm_exec) +{ + drm_exec_fini(&vm_exec->exec); +} + +int drm_gpuvm_validate (struct drm_gpuvm *gpuvm); +void drm_gpuvm_resv_add_fence (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + struct dma_fence *fence, + enum dma_resv_usage
private_usage,
+ enumdma_resv_usageextobj_usage);include/drm/drm_gpuvm.h +12
-0+ +/** +* drm_gpuvm_exec_resv_add_fence() +* @vm_exec: the &drm_gpuvm_exec abstraction +* @fence: fence to add +* @private_usage: private dma-resv usage +* @extobj_usage: extobj dma-resv usage +* +* See drm_gpuvm_resv_add_fence(). +*/ +static inline void +drm_gpuvm_exec_resv_add_fence(struct drm_gpuvm_exec
*vm_exec,
+ structdma_fence*fence, + enumdma_resv_usageprivate_usage,
+ enumdma_resv_usageextobj_usage)
+{
+ drm_gpuvm_resv_add_fence(vm_exec->vm,&vm_exec->exec, fence,
+private_usage,extobj_usage);include/drm/drm_gpuvm.h +2
-0+} +/**
* structdrm_gpuvm_bo- structure representing a&drm_gpuvmand
*&drm_gem_objectcombinationinclude/drm/drm_gpuvm.h +3
-0@@ -398,6 +569,18 @@ struct drm_gpuvm_bo {* gpuva list. */ struct list_head gem; + + /** +* @evict: List entry to attach to the
&drm_gpuvms
+ * extobj list.
+ */
+ structlist_headextobj; +
+ /**
+ * @evict: List entry to attach to the&drm_gpuvmsevict
+ * list.
+ */
+ structlist_headevict;
} entry;
} list;
};include/drm/drm_gpuvm.h +3
-0@@ -432,6 +615,9 @@ struct drm_gpuvm_bo * drm_gpuvm_bo_find(struct drm_gpuvm *gpuvm, struct drm_gem_object *obj); +void drm_gpuvm_bo_evict (struct drm_gem_object *obj,bool evict); +void drm_gpuvm_bo_extobj_add (struct drm_gpuvm_bo *vm_bo); + /*** drm_gpuvm_bo_for_each_va() - iterator to walk over a list of
&drm_gpuva
* @va__:&drm_gpuvastructure to assign to in each iteration stepinclude/drm/drm_gpuvm.h +0
-0@@ -837,6 +1023,17 @@ struct drm_gpuvm_ops {* used. */ int (*sm_step_unmap)(struct
drm_gpuva_op*op, void *priv);
+
+ /**
+ * @bo_validate: called fromdrm_gpuvm_validate()+ *
+ * Drivers receive this callback for every evicted&drm_gem_objectbeing
+ * mapped in the corresponding&drm_gpuvm. + *
+ * Typically, drivers would call their driver specific variant of
+ *ttm_bo_validate()from within this callback. + */
+ int (*bo_validate)(structdrm_gem_object*obj); };
int drm_gpuvm_sm_map(structdrm_gpuvm*gpuvm, void *priv,
Re: [PATCH drm-misc-next v3 6/7] drm/gpuvm: generalize dma_resv/extobj handling and GEM validation
As mentioned in a different mail thread, the reply is based on the assumption that we don't support anything else than GPUVM updates from the IOCTL.
Quoted 200 lines · Thomas Hellström
Hi!
Quoted 180 lines · Danilo Krummrich
Quoted 169 lines · Thomas Hellström
Quoted 161 lines · Danilo Krummrich
Quoted 142 lines · Thomas Hellström
Hi, Danilo,
Quoted 139 lines · Danilo Krummrich
So far the DRM GPUVA manager offers common infrastructure to track GPU VA
allocations and mappings, generically connect GPU VA mappings to their
backing buffers and perform more complex mapping operations on the GPU VA
space.However, there are more design patterns commonly used by drivers, which
can potentially be generalized in order to make the DRM GPUVA manager
represent a basic GPU-VM implementation. In this context, this patch aims
at generalizing the following elements.1) Provide a common dma-resv for GEM objects not being used outside of
this GPU-VM.2) Provide tracking of external GEM objects (GEM objects which are
shared with other GPU-VMs).3) Provide functions to efficiently lock all GEM objects dma- resv the
GPU-VM contains mappings of.4) Provide tracking of evicted GEM objects the GPU-VM contains mappings
of, such that validation of evicted GEM objects is accelerated.5) Provide some convinience functions for common patterns.
Rather than being designed as a "framework", the target is to make all
features appear as a collection of optional helper functions, such that
drivers are free to make use of the DRM GPUVA managers basic functionality and opt-in for other features without setting any feature
flags, just by making use of the corresponding functions.Big kudos to Boris Brezillon for his help to figure out locking for drivers
updating the GPU VA space within the fence signalling path.Suggested-by Matthew Brost <[email protected]>Signed-off-by Danilo Krummrich <[email protected]>---
drivers/gpu/drm/drm_gpuvm.c| 516
++++++++++++++++++++++++++++++++++++
include/drm/drm_gpuvm.h | 197 ++++++++++++++2 files changed, 713 insertions(+)drivers/gpu/drm/drm_gpuvm.c +0
-0@@ -73,6 +73,21 @@ * &drm_gem_object list of &drm_gpuvm_bos for an existinginstance of this
* particular combination. If not existent a new instance is created and linked
* to the&drm_gem_object.
+ *
+ *&drm_gpuvm_bostructures, since unique for a given&drm_gpuvm, are also used
+ * as entry for the &drm_gpuvm's lists of external and evicted objects. Those
+ * list are maintained in order to accelerate locking of dma-resv locks and
+ * validation of evicted objects bound in a&drm_gpuvm. For instance the all
+ * &drm_gem_object's&dma_resvof a given&drm_gpuvmcan be locked by calling
+ *drm_gpuvm_exec_lock(). Once locked drivers can calldrm_gpuvm_validate()in
+ * order to validate all evicted&drm_gem_objects. It is also possible to lock
+ * additional&drm_gem_objectsby providing the corresponding parameters to
+ *drm_gpuvm_exec_lock()as well as open code the&drm_execloop while making
+ * use of helper functions such asdrm_gpuvm_prepare_range()or
+ *drm_gpuvm_prepare_objects().
+ *
+ * Every bound&drm_gem_objectis treated as external object when its&dma_resv
+ * structure is different than the &drm_gpuvm's common&dma_resvstructure.
*/
/**drivers/gpu/drm/drm_gpuvm.c +0
-0@@ -420,6 +435,20 @@ * Subsequent calls to drm_gpuvm_bo_obtain() for the same
&drm_gpuvmand
*&drm_gem_objectmust be able to observe previous creations and destructions
* of&drm_gpuvm_bosin order to keep instances unique.
+ *
+ * The &drm_gpuvm's lists for keeping track of external and evicted objects are
+ * protected against concurrent insertion / removal and iteration internally.
+ *
+ * However, drivers still need ensure to protect concurrent calls to functions
+ * iterating those lists, such asdrm_gpuvm_validate()and
+ *drm_gpuvm_prepare_objects(). Every such function contains a particular
+ * comment and lockdep checks if possible.
+ *
+ * Functions adding or removing entries from those lists, such as
+ *drm_gpuvm_bo_evict()ordrm_gpuvm_bo_extobj_add()may be called with external
+ * locks being held, e.g. in order to avoid the corresponding list to be
+ * (safely) modified while potentially being iternated by other API functions.
+ * However, this is entirely optional.
*/
/**drivers/gpu/drm/drm_gpuvm.c +5
-0@@ -632,6 +661,131 @@* } */ +/** +* get_next_vm_bo_from_list() - get the next vm_bo element +* @__gpuvm: The GPU VM +* @__list_name: The name of the list we're iterating on +* @__local_list: A pointer to the local list used to store already iterated items
+ * @__prev_vm_bo: The previous element we got fromdrm_gpuvm_get_next_cached_vm_bo()
+ *
+ * This helper is here to provide lockless list iteration.Quoted 8 lines
Quoted 5 lines
Lockless as in, the
+ * iterator releases the lock immediately after picking the first element from
+ * the list, so list insertion deletion can happen concurrently.Are the list spinlocks needed for that async state update from within the
dma-fence critical section we've discussed previously?Yes, but also for other reasons, see below.
Quoted 8 lines
Otherwise it should be sufficient to protect the lists with the gpuvm's resv
(or for the extobj list with an outer lock).If those spinlocks are still needed in some situations, perhaps could we
have an option to set them to NULL (Like IIRC the maple tree allows for)?Quoted 6 lines
The evict spinlock is needed in any case, since in
drm_gpuvm_bo_evict()we're
holding only the dma-resv lock from the BO this function gets called for. Hence,
the spinlock protects concurrentdrm_gpuvm_bo_evict()calls with different BOs.Quoted 9 lines
No. Only if you try to add external objects to the vm's evict list from
within the evict code. That's not necessary since you loop through all
external objects anyway when locking them so an "evicted" bool in thevm_bo,
protected by the bo resv would be sufficient. The extobj locking loop can
then add the bo to the evicted list.Quoted 16 lines
And
validate()can remove it while still holding all dma-resv locks, neat!
However, what if two tasks are trying to lock the VA space concurrently? What
do we do when the drm_gpuvm_bo's refcount drops to zero indrm_gpuva_unlink()?
Are we guaranteed that at this point of time thedrm_gpuvm_bois not on the
evicted list? Because otherwise we would calldrm_gpuvm_bo_destroy()with the
dma-resv lock held, which wouldn't be allowed, sincedrm_gpuvm_bo_destroy()
might drop the last reference to thedrm_gem_objectand hence we'd potentially
free the dma-resv lock while holding it, at least if it's an external object.
Quoted 4 lines
Easiest way in this scheme is to think of the lists as being protected by the vm's resv lock. That means anybody calling
unlink()must also hold the vm's resv lock. (Which is OK from an UAF point of view, but perhaps not from a locking inversion POW from an async list update).
This would mean that on unlink() we'd need to hold the VM's resv lock and the corresponding GEM's resv lock (in case they're not the same anyways) because the VM's resv lock would protect the external / evicted object lists and the GEM objects resv lock protects the GEM's list of drm_gpuvm_bos and the drm_gpuvm_bo's list of drm_gpuvas.
Quoted 55 lines
Quoted 43 lines
Quoted 9 lines
Quoted 8 lines
For extobjs an outer lock would be enough in case of Xe, but I really would not
like to add even more complexity just to get the spinlock out of the way in case
the driver already has an outer lock protecting this path.Quoted 15 lines
I must disagree here. These spinlocks and atomic operations are pretty
costly and as discussed earlier this type of locking was the reason (at
least according to the commit message) that made Christian drop the XArray
use indrm_execfor the same set of objects: "The locking overhead is
unecessary and measurable". IMHO the spinlock is the added complexity and a
single wide lock following the drm locking guidelines set out by Daniel and
David should really be the default choice with an opt-in for a spinlock if
needed for async and pushing out to a wq is not an option.For the external object list an outer lock would work as long as it's not the
dma-resv lock of the corresponding GEM object, since here we actually need to
remove the list entry from the external object list ondrm_gpuvm_bo_destroy().
It's just a bit weird design wise that drivers would need to take this outer
lock on:-
drm_gpuvm_bo_extobj_add()
- drm_gpuvm_bo_destroy() (and hence alsodrm_gpuvm_bo_put())
- drm_gpuva_unlink() (because it needs to calldrm_gpuvm_bo_put())
-drm_gpuvm_exec_lock()
-drm_gpuvm_exec_lock_array()
-drm_gpuvm_prepare_range()Quoted 2 lines
Given that it seems reasonable to do all the required locking internally.
From a design POW, there has been a clear direction in XE to make things similar to
mmap()/munmap(), so this outer lock, which in Xe is an rwsem, is used in a similar way as themmap_lock. It's protecting the page-table structures and vma rb tree, the userptr structures and the extobj list. Basically it's taken early in the exec IOCTL, theVM_BINDioctl, the compute rebind worker and the pagefault handler, so all of the above are just asserting that it is taken in the correct mode.
Quoted 6 lines
But strictly with this scheme one could also use the vm's
dma_resvfor the extobj list since withdrm_exec, it's locked before traversing the list.The whole point of this scheme is to rely on locks that you already are supposed to be holding for various reasons and is simple to comprehend.
I don't agree that we're supposed to hold the VM's resv lock anyways for functions like drm_gpuvm_bo_put() or drm_gpuva_unlink(), but I'm fine using it for that purpose nevertheless.
Quoted 9 lines
Quoted 7 lines
In order to at least place lockdep checks, the driver would need to supply the
corresponding lock'slockdep_map, because the GPUVM otherwise doesn't know about
the lock.Yes, that sounds reasonable. One lockdep map per list.
I'd really like to avoid that, especially now that everything got simpler. We should define the actual locks to take instead.
Quoted 53 lines
Quoted 5 lines
Out of curiosity, what is the overhead of a
spin_lock()that doesn't need to
spin?I guess it's hard to tell exactly, but it is much lower on modern x86 than what it used to be. Not sure about ARM, which is the other architecture important to us. I figure if there is little cache-line bouncing the main overhead comes from the implied barriers.
Quoted 38 lines
Quoted 28 lines
A pretty simple way that would not add much code would be
static void gpuvm_cond_spin_lock(const struct
drm_gpuvm*gpuvm,spinlock_t
*lock){
if (!gpuvm->resv_protected_lists)
spin_lock(lock);}
Quoted 12 lines
Quoted 3 lines
For such drivers, that would require anybody calling unlink to hold the vm's
resv, though.In V4 I want to go back to having a dedicated lock for the GEMs gpuva list (or
VM_BOlist to be more precise). We can't just use the dma-resv lock for that
withVM_BOabstractions, because on destruction of aVM_BOwe otherwise wouldn't
be allowed to already hold the dma-resv lock. That's the fix I was referring to
earlier.Quoted 9 lines
Yeah, I can see the need for a dedicated lock for the GEM's gpuva list, but
holding the vm's dma-resv lock across the unlink shouldn't be a problem. We
may free the object and a pointer to the vm's resv during unlink but we
don't free the vm's resv. It'd be a matter of ensuring that any calls to
unlink from *within*drm_gpuvmallows it to be held.Quoted 3 lines
Drivers calling
unlink()from the fence signaling path can't use the VM's
dma-resv lock.
Quoted 6 lines
Yes, that made me a bit curious because in the current version the code required the object's
dma_resvforunlink()which can't be grabbed either from the fence signaling path. So are there any drivers actually wanting to do that? If so, they will either need to resort to the current spinlock solution or they will need to call unlink from a workqueue item.
As Boris already mentioned we have the dma-resv lock by default or a driver specific GEM gpuva lock as opt-in. Now, we can get rid of the latter.
Quoted 11 lines
Quoted 4 lines
Also, what if the object is an external object? We can't use the VM's dma-resv
lock here.Why? Typically (sync) unlink is only ever called from an unbind-like operation where it should be trivial to grab the vm's resv. Or, for that matter any outer lock protecting the extobj list. Rule would be the
drm_gpuvm_bo::entry::extobjanddrm_gpuvm_bo::entry::evictwould be protected by either the vm'sdma_resv(or possibly an outer lock in the case of the extobj list).
Outer lock wouldn't have been working for updates in the async path, but shouldn't be relevant anymore. We could use the VM's resv for that.
Quoted 8 lines
Quoted 7 lines
And we can't have the GEM objs dma-resv lock held when calling
unlink(), sinceunlink()callsdrm_gpuvm_bo_put(), which if the refcount drops
to zero callsdrm_gpuvm_bo_destroy()anddrm_gpuvm_bo_destroy()might drop the
last reference of the GEM object.
Quoted 10 lines
Yes, but this is a different problem as to what exactly protects
drm_gpuvm_bo::entry::gem. Either as you suggest an internal per bo list lock, or if we want to keep the bo'sdma_resvwe need to ensure that the caller ofdma_resv_unlock(obj->resv)actually refcounts its obj pointer, and doesn't implicitly rely on the gpuvm_bo's refcount (I know Boris didn't like that, but requiring an explicit refcount for a pointer you dereference unless you're under a lock that ensures keeping the object alive is pretty much required?) But anyway for thedrm_gpuvm_bo::entry::gemlist protection (bo resv or internal spinlock) I don't have a strong preference.
We can keep the GEM objects dma-resv lock, however as mentioned above drm_gpuva_unlink() and drm_gpuvm_bo_put() then requires both the VM's resv lock and the GEM's resv lock in case they differ.
Quoted 977 lines
Quoted 3 lines
All those problems go away with a dedicated
GEM gpuva list lock.I don't think these are real problems.
With the excepton of the eviction list "trick" where we currently have slightly different approach to collect external bos needing rebinding, we have this working fine.TBH I think pretty much the only situation where the spinlock is needed is for async updates of these lists, unless a wq item can be used for that, but it doesn't really seem like the current code allows for such updates anyway? It complicates the code a lot, adds overhead and also adds the requirement for refcounting during list traversal.
/Thomas
Quoted 959 lines
Quoted 958 lines
/Thomas
Quoted 952 lines
Quoted 952 lines
It seems that with that also the refcount could be make non- atomic.
All in the spirit of the drm locking guidelines "use big locks when
possible".
Lower level locks only when necessary for performance or locking inversion?/Thomas
Quoted 940 lines
+ *
+ * Elements popped from the original list are kept in a local list, so removal
+ * andis_emptychecks can still happen while we're iterating the list.
+ */
+#defineget_next_vm_bo_from_list(__gpuvm, __list_name, __local_list, __prev_vm_bo)\
+ ({
\
+ structdrm_gpuvm_bo
*__vm_bo; \
+
\
+ drm_gpuvm_bo_put(__prev_vm_bo); \
+
\
+ spin_lock(&(__gpuvm)-Quoted 1 line
__list_name.lock); \
+ while (!list_empty(&(__gpuvm)-
Quoted 1 line
__list_name.list)) { \
+ __vm_bo = list_first_entry(&(__gpuvm)->__list_name.list, \
+ struct drm_gpuvm_bo, \
+
list.entry.__list_name); \
+ if (drm_gpuvm_bo_get_unless_zero(__vm_bo))
{ \
+ list_move_tail(&(__vm_bo)-Quoted 1 line
list.entry.__list_name, \
+
__local_list); \
+ break;
\
+ } else
{ \
+ list_del_init(&(__vm_bo)-Quoted 1 line
list.entry.__list_name); \
+ __vm_bo =
NULL; \
+ }
\
+ }
\
+ spin_unlock(&(__gpuvm)-Quoted 1 line
__list_name.lock); \
+
\
+ __vm_bo;
\
+ })drivers/gpu/drm/drm_gpuvm.c +5
-0+ +/** +* for_each_vm_bo_in_list() - internal vm_bo list iterator +* +* This helper is here to provide lockless list iteration. Lockless as in, the
+ * iterator releases the lock immediately after picking the first element from the
+ * list, so list insertion and deletion can happen concurrently.
+ *
+ * Typical use:
+ *
+ * structdrm_gpuvm_bo*vm_bo;
+ * LIST_HEAD(my_local_list);
+ *
+ * ret = 0;
+ * drm_gpuvm_for_each_vm_bo(gpuvm, <list_name>,&my_local_list,vm_bo) {
+ * ret =do_something_with_vm_bo(..., vm_bo);
+ * if (ret)
+ * break;
+ * }
+ * drm_gpuvm_bo_put(vm_bo);
+ * drm_gpuvm_restore_vm_bo_list(gpuvm, <list_name>,&my_local_list);
+ *
+ *
+ * Only used for internal list iterations, not meant to be exposed to the outside
+ * world.
+ */
+#definefor_each_vm_bo_in_list(__gpuvm, __list_name, __local_list, __vm_bo)\
+ for (__vm_bo= get_next_vm_bo_from_list(__gpuvm, __list_name, \
+ __local_list, NULL); \
+__vm_bo;
\
+__vm_bo= get_next_vm_bo_from_list(__gpuvm, __list_name, \
+ __local_list, __vm_bo)) \drivers/gpu/drm/drm_gpuvm.c +3
-0+ +/** +* restore_vm_bo_list() - move vm_bo elements back to their original list
+ * @__gpuvm: The GPU VM
+ * @__list_name: The name of the list we're iterating on
+ * @__local_list: A pointer to the local list used to store already iterated items
+ *
+ * When we're done iterating avm_bolist, we should callrestore_vm_bo_list()
+ * to restore the original state and let new iterations take place.
+ */
+#definerestore_vm_bo_list(__gpuvm, __list_name, __local_list)\
+ do
{
\
+ /* Merge back the two lists, moving local list elements to the \
+ * head to preserve previous ordering, in case it matters. \
+
*/
\
+ spin_lock(&(__gpuvm)-Quoted 1 line
__list_name.lock); \
+ list_splice(__local_list, &(__gpuvm)-
Quoted 1 line
__list_name.list); \
+ spin_unlock(&(__gpuvm)-
Quoted 1 line
__list_name.lock); \
+ } while (0)
+/**
+ *drm_gpuvm_bo_list_add()- insert avm_bointo the given list
+ * @__vm_bo: the&drm_gpuvm_bo
+ * @__list_name: the name of the list to insert into
+ *
+ * Inserts the given @__vm_bo into the list specified by @__list_name and
+ * increases the vm_bo's reference count.
+ */
+#definedrm_gpuvm_bo_list_add(__vm_bo, __list_name)\
+ do
{
\
+ spin_lock(&(__vm_bo)->vm-Quoted 1 line
__list_name.lock); \
+ if (list_empty(&(__vm_bo)-
Quoted 1 line
list.entry.__list_name)) \
+ list_add_tail(&(__vm_bo)-
Quoted 1 line
list.entry.__list_name, \
+ &(__vm_bo)->vm-
Quoted 1 line
__list_name.list); \
+ spin_unlock(&(__vm_bo)->vm-
Quoted 1 line
__list_name.lock); \
+ } while (0)
drivers/gpu/drm/drm_gpuvm.c +3
-0+ +/** +* drm_gpuvm_bo_list_del() - remove a vm_bo from the given list
+ * @__vm_bo: the&drm_gpuvm_bo
+ * @__list_name: the name of the list to insert into
+ *
+ * Removes the given @__vm_bo from the list specified by @__list_name and
+ * decreases the vm_bo's reference count.
+ */
+#definedrm_gpuvm_bo_list_del(__vm_bo, __list_name)\
+ do
{
\
+ spin_lock(&(__vm_bo)->vm-Quoted 1 line
__list_name.lock); \
+ if (!list_empty(&(__vm_bo)-
Quoted 1 line
list.entry.__list_name)) \
+ list_del_init(&(__vm_bo)-
Quoted 1 line
list.entry.__list_name); \
+ spin_unlock(&(__vm_bo)->vm-
Quoted 1 line
__list_name.lock); \
+ } while (0)
drivers/gpu/drm/drm_gpuvm.c +4
-0+ +static int __must_check +drm_gpuvm_bo_get_unless_zero (struct drm_gpuvm_bo *vm_bo); +#define to_drm_gpuva(__node) container_of((__node), struct
drm_gpuva, rb.node)
#defineGPUVA_START(node)((node)->va.addr)drivers/gpu/drm/drm_gpuvm.c +6
-0@@ -713,6 +867,12 @@ drm_gpuvm_init(struct drm_gpuvm *gpuvm,struct drm_device *drm, gpuvm->rb.tree = RB_ROOT_CACHED; INIT_LIST_HEAD(&gpuvm->rb.list); + INIT_LIST_HEAD(&gpuvm->extobj.list); + spin_lock_init(&gpuvm->extobj.lock); + + INIT_LIST_HEAD(&gpuvm->evict.list); + spin_lock_init(&gpuvm->evict.lock); + drm_gpuva_check_overflow(start_offset, range); gpuvm->mm_start = start_offset; gpuvm->mm_range = range;@@ -754,10 +914,302 @@ drm_gpuvm_destroy(struct drm_gpuvm *gpuvm) WARN(!RB_EMPTY_ROOT(&gpuvm->rb.tree.rb_root),"GPUVA tree is not empty, potentially leaking memory.\n");
+ WARN(!list_empty(&gpuvm->extobj.list), "Extobj list should be empty.\n");
+ WARN(!list_empty(&gpuvm->evict.list), "Evict list should be empty.\n");
+
drm_gem_private_object_fini(&gpuvm->d_obj); }
EXPORT_SYMBOL_GPL(drm_gpuvm_destroy);
+/**
+ *drm_gpuvm_prepare_objects()- prepare all assoiciated BOs
+ * @gpuvm: the&drm_gpuvm
+ * @exec: the&drm_execlocking context
+ * @num_fences: the amount of&dma_fencesto reserve
+ *
+ * Callsdrm_exec_prepare_obj()for all&drm_gem_objectsthe given
+ *&drm_gpuvmcontains mappings of.
+ *
+ * Using this function directly, it is the drivers responsibility to call
+ *drm_exec_init()anddrm_exec_fini()accordingly.
+ *
+ * Note: This function is safe against concurrent insertion and removal of
+ * external objects, however it is not safe against concurrent usage itself.
+ *
+ * Drivers need to make sure to protect this case with either an outer VM lock
+ * or by callingdrm_gpuvm_prepare_vm()before this function within the
+ *drm_exec_until_all_locked()loop, such that the GPUVM's dma-resv lock ensures
+ * mutual exclusion.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +10
-0+int +drm_gpuvm_prepare_objects (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + unsigned int num_fences) +{ + struct drm_gpuvm_bo *vm_bo; + LIST_HEAD(extobjs); + int ret = 0; + + for_each_vm_bo_in_list(gpuvm, extobj, &extobjs,
vm_bo) {
+ ret =drm_exec_prepare_obj(exec, vm_bo->obj, num_fences);
+ if (ret)
+ break;
+ }
+ /* Drop ref in case we break out of the loop. */ + drm_gpuvm_bo_put(vm_bo);
+ restore_vm_bo_list(gpuvm, extobj, &extobjs);drivers/gpu/drm/drm_gpuvm.c +7
-0+ + return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_prepare_objects); + +/** +* drm_gpuvm_prepare_range() - prepare all BOs mapped within a given range
+ * @gpuvm: the&drm_gpuvm
+ * @exec: the&drm_execlocking context
+ * @addr: the start address within the VA space
+ * @range: the range to iterate within the VA space
+ * @num_fences: the amount of&dma_fencesto reserve
+ *
+ * Callsdrm_exec_prepare_obj()for all&drm_gem_objectsmapped between @addr
+ * and @addr + @range.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +2
-0+int +drm_gpuvm_prepare_range(struct drm_gpuvm *gpuvm, struct
drm_exec*exec,
+ u64 addr, u64 range, unsigned intnum_fences)drivers/gpu/drm/drm_gpuvm.c +9
-0+{ + struct drm_gpuva *va; + u64 end = addr + range; + int ret; + + drm_gpuvm_for_each_va_range(va, gpuvm, addr, end) { + struct drm_gem_object *obj = va->gem.obj; + + ret = drm_exec_prepare_obj(exec, obj,
num_fences);
+ if (ret)
+ return ret;
+ }drivers/gpu/drm/drm_gpuvm.c +7
-0+ + return 0; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_prepare_range); + +/** +* drm_gpuvm_exec_lock() - lock all dma-resv of all assoiciated BOs
+ * @vm_exec: the&drm_gpuvm_execabstraction
+ * @num_fences: the amount of&dma_fencesto reserve
+ * @interruptible: sleep interruptible if waiting
+ *
+ * Acquires all dma-resv locks of all&drm_gem_objectsthe given
+ *&drm_gpuvmcontains mappings of.
+ *
+ * Addionally, when calling this function with structdrm_gpuvm_exec::extra
+ * being set the driver receives the given @fn callback to lock additional
+ * dma-resv in the context of the&drm_gpuvm_execinstance. Typically, drivers
+ * would calldrm_exec_prepare_obj()from within this callback.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +11
-0+int +drm_gpuvm_exec_lock (struct drm_gpuvm_exec *vm_exec, + unsigned int num_fences, + bool interruptible) +{ + struct drm_gpuvm *gpuvm = vm_exec->vm; + struct drm_exec *exec = &vm_exec->exec; + uint32_t flags; + int ret; + + flags = interruptible ? DRM_EXEC_INTERRUPTIBLE_WAIT :0 |
+ DRM_EXEC_IGNORE_DUPLICATES;drivers/gpu/drm/drm_gpuvm.c +5
-0+ + drm_exec_init(exec, flags); + + drm_exec_until_all_locked(exec) { + ret = drm_gpuvm_prepare_vm(gpuvm, exec,
num_fences);
+ drm_exec_retry_on_contention(exec); + if (ret)
+ goto err;
+
+ ret =drm_gpuvm_prepare_objects(gpuvm, exec, num_fences);
+ drm_exec_retry_on_contention(exec); + if (ret)
+ goto err;
+
+ if (vm_exec->extra.fn) {
+ ret =vm_exec->extra.fn(vm_exec, num_fences);
+ drm_exec_retry_on_contention(exec); + if (ret)
+ goto err;
+ }
+ }drivers/gpu/drm/drm_gpuvm.c +11
-0+ + return 0; + +err: + drm_exec_fini(exec); + return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock); + +static int +fn_lock_array(struct drm_gpuvm_exec *vm_exec, unsigned int
num_fences)drivers/gpu/drm/drm_gpuvm.c +7
-0+{ + struct { + struct drm_gem_object **objs; + unsigned int num_objs; + } *args = vm_exec->extra.priv; + + return drm_exec_prepare_array(&vm_exec->exec, args-Quoted 1 line
objs,
+
args->num_objs,num_fences);drivers/gpu/drm/drm_gpuvm.c +4
-0+} + +/** +* drm_gpuvm_exec_lock_array() - lock all dma-resv of all assoiciated BOs
+ * @vm_exec: the&drm_gpuvm_execabstraction
+ * @objs: additional&drm_gem_objectsto lock
+ * @num_objs: the number of additional&drm_gem_objectsto lock
+ * @num_fences: the amount of&dma_fencesto reserve
+ * @interruptible: sleep interruptible if waiting
+ *
+ * Acquires all dma-resv locks of all&drm_gem_objectsthe given&drm_gpuvm
+ * contains mappings of, plus the ones given through @objs.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +19
-0+int +drm_gpuvm_exec_lock_array (struct drm_gpuvm_exec *vm_exec, + struct drm_gem_object **objs, + unsigned int num_objs, + unsigned int num_fences, + bool interruptible) +{ + struct { + struct drm_gem_object **objs; + unsigned int num_objs; + } args; + + args.objs = objs; + args.num_objs = num_objs; + + vm_exec->extra.fn = fn_lock_array; + vm_exec->extra.priv = &args; + + return drm_gpuvm_exec_lock(vm_exec, num_fences,interruptible);
drivers/gpu/drm/drm_gpuvm.c +5
-0+} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock_array); + +/** +* drm_gpuvm_exec_lock_range() - prepare all BOs mapped within a given range
+ * @vm_exec: the&drm_gpuvm_execabstraction
+ * @addr: the start address within the VA space
+ * @range: the range to iterate within the VA space
+ * @num_fences: the amount of&dma_fencesto reserve
+ * @interruptible: sleep interruptible if waiting
+ *
+ * Acquires all dma-resv locks of all&drm_gem_objectsmapped between @addr and
+ * @addr + @range.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +12
-0+int +drm_gpuvm_exec_lock_range (struct drm_gpuvm_exec *vm_exec, + u64 addr, u64 range, + unsigned int num_fences, + bool interruptible) +{ + struct drm_gpuvm *gpuvm = vm_exec->vm; + struct drm_exec *exec = &vm_exec->exec; + uint32_t flags; + int ret; + + flags = interruptible ? DRM_EXEC_INTERRUPTIBLE_WAIT :0 |
+ DRM_EXEC_IGNORE_DUPLICATES;drivers/gpu/drm/drm_gpuvm.c +5
-0+ + drm_exec_init(exec, flags); + + drm_exec_until_all_locked(exec) { + ret = drm_gpuvm_prepare_range(gpuvm, exec,addr, range,
+num_fences); + drm_exec_retry_on_contention(exec); + if (ret)
+ goto err;
+ }drivers/gpu/drm/drm_gpuvm.c +14
-0+ + return ret; + +err: + drm_exec_fini(exec); + return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock_range); + +/** +* drm_gpuvm_validate() - validate all BOs marked as evicted +* @gpuvm: the &drm_gpuvm to validate evicted BOs +* +* Calls the &drm_gpuvm_ops.bo_validate callback for all evicted buffer
+ * objects being mapped in the given&drm_gpuvm.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +27
-0+int +drm_gpuvm_validate (struct drm_gpuvm *gpuvm) +{ + const struct drm_gpuvm_ops *ops = gpuvm->ops; + struct drm_gpuvm_bo *vm_bo; + LIST_HEAD(evict); + int ret = 0; + + if (unlikely(!ops || !ops->bo_validate)) + return -ENOTSUPP; + + for_each_vm_bo_in_list(gpuvm, evict, &evict, vm_bo) { + dma_resv_assert_held(vm_bo->obj->resv); + ret = ops->bo_validate(vm_bo->obj); + if (ret) + break; + } + /* Drop ref in case we break out of the loop. */ + drm_gpuvm_bo_put(vm_bo); + restore_vm_bo_list(gpuvm, evict, &evict); + + return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_validate); + +/** +* drm_gpuvm_resv_add_fence - add fence to private and all extobj
+ * dma-resv
+ * @gpuvm: the&drm_gpuvmto add a fence to
+ * @exec: the&drm_execlocking context
+ * @fence: fence to add
+ * @private_usage: private dma-resv usage
+ * @extobj_usage: extobj dma-resv usage
+ */drivers/gpu/drm/drm_gpuvm.c +14
-0+void +drm_gpuvm_resv_add_fence (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + struct dma_fence *fence, + enum dma_resv_usage private_usage, + enum dma_resv_usage extobj_usage) +{ + struct drm_gem_object *obj; + unsigned long index; + + drm_exec_for_each_locked_object(exec, index, obj) { + dma_resv_assert_held(obj->resv); + dma_resv_add_fence(obj->resv, fence, + drm_gpuvm_is_extobj(gpuvm,obj) ?
+private_usage:extobj_usage);
+ }drivers/gpu/drm/drm_gpuvm.c +3
-0+} +EXPORT_SYMBOL_GPL (drm_gpuvm_resv_add_fence); +/**
*drm_gpuvm_bo_create()- create a new instance of structdrm_gpuvm_bo
* @gpuvm: The&drm_gpuvmthe @obj is mapped in.drivers/gpu/drm/drm_gpuvm.c +13
-0@@ -790,6 +1242,9 @@ drm_gpuvm_bo_create(struct drm_gpuvm *gpuvm, INIT_LIST_HEAD(&vm_bo->list.gpuva); INIT_LIST_HEAD(&vm_bo->list.entry.gem); + INIT_LIST_HEAD(&vm_bo->list.entry.extobj); + INIT_LIST_HEAD(&vm_bo->list.entry.evict); + drm_gem_object_get(obj); return vm_bo;@@ -807,6 +1262,14 @@ drm_gpuvm_bo_destroy(struct kref *kref) drm_gem_gpuva_assert_lock_held(vm_bo->obj); + spin_lock(&gpuvm->extobj.lock); + list_del(&vm_bo->list.entry.extobj); + spin_unlock(&gpuvm->extobj.lock); + + spin_lock(&gpuvm->evict.lock); + list_del(&vm_bo->list.entry.evict); + spin_unlock(&gpuvm->evict.lock); + list_del(&vm_bo->list.entry.gem); drm_gem_object_put(obj);@@ -822,6 +1285,11 @@ drm_gpuvm_bo_destroy(struct kref *kref) * @vm_bo: the &drm_gpuvm_bo to release the reference of * * This releases a reference to @vm_bo. + * + * If the reference count drops to zero, the &gpuvm_bo isdestroyed, which
+ * includes removing it from the GEMs gpuva list. Hence, if a call to this
+ * function can potentially let the reference count to zero the caller must
+ * hold the dma-resv or driver specific GEM gpuva lock. */
void
drm_gpuvm_bo_put(struct drm_gpuvm_bo *vm_bo)drivers/gpu/drm/drm_gpuvm.c +6
-0@@ -831,6 +1299,12 @@ drm_gpuvm_bo_put(struct drm_gpuvm_bo *vm_bo) } EXPORT_SYMBOL_GPL(drm_gpuvm_bo_put); +static int __must_check +drm_gpuvm_bo_get_unless_zero (struct drm_gpuvm_bo *vm_bo) +{ + return kref_get_unless_zero(&vm_bo->kref); +} +static struct
drm_gpuvm_bo*
__drm_gpuvm_bo_find(structdrm_gpuvm*gpuvm,
structdrm_gem_object*obj)drivers/gpu/drm/drm_gpuvm.c +2
-0@@ -938,6 +1412,48 @@ drm_gpuvm_bo_obtain_prealloc(struct drm_gpuvm_bo *__vm_bo) } EXPORT_SYMBOL_GPL(drm_gpuvm_bo_obtain_prealloc); +/** +* drm_gpuvm_bo_extobj_add() - adds the &drm_gpuvm_bo to its &drm_gpuvm's
+ * extobj list
+ * @vm_bo: The&drm_gpuvm_boto add to its &drm_gpuvm's the extobj list.
+ *
+ * Adds the given @vm_bo to its &drm_gpuvm's extobj list if not on the list
+ * already and if the corresponding&drm_gem_objectis an external object,
+ * actually.
+ */drivers/gpu/drm/drm_gpuvm.c +12
-0+void +drm_gpuvm_bo_extobj_add (struct drm_gpuvm_bo *vm_bo) +{ + struct drm_gpuvm *gpuvm = vm_bo->vm; + + if (drm_gpuvm_is_extobj(gpuvm, vm_bo->obj)) + drm_gpuvm_bo_list_add(vm_bo, extobj); +} +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_extobj_add); + +/** +* drm_gpuvm_bo_evict() - add / remove a &drm_gem_object to / from a
+ *&drm_gpuvmsevicted list
+ * @obj: the&drm_gem_objectto add or remove
+ * @evict: indicates whether the object is evicted
+ *
+ * Adds a&drm_gem_objectto or removes it from all&drm_gpuvmsevicted
+ * list containing a mapping of this&drm_gem_object.
+ */drivers/gpu/drm/drm_gpuvm.c +14
-0+void +drm_gpuvm_bo_evict (struct drm_gem_object *obj,bool evict) +{ + struct drm_gpuvm_bo *vm_bo; + + drm_gem_for_each_gpuvm_bo(vm_bo, obj) { + if (evict) + drm_gpuvm_bo_list_add(vm_bo, evict); + else + drm_gpuvm_bo_list_del(vm_bo, evict); + } +} +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_evict); +static int
__drm_gpuva_insert(structdrm_gpuvm*gpuvm,
structdrm_gpuva*va)include/drm/drm_gpuvm.h +9
-0@@ -26,10 +26,12 @@*/ #include <linux/list.h> +#include <linux/dma-resv.h> #include <linux/rbtree.h> #include <linux/types.h> #include <drm/drm_gem.h> +#include <drm/drm_exec.h> struct drm_gpuvm; struct drm_gpuvm_bo;@@ -259,6 +261,38 @@ struct drm_gpuvm {* space */ struct dma_resv *resv; + + /** + * @extobj: structure holding the extobj list + */ + struct { + /** +* @list: &list_head storing &drm_gpuvm_bos serving as
+ * external object
+ */
+ structlist_headlist;include/drm/drm_gpuvm.h +9
-0+ + /** + * @lock: spinlock to protect the extobj list + */ + spinlock_t lock; + } extobj; + + /** +* @evict: structure holding the evict list and evict list lock
+ */
+ struct {
+ /**
+ * @list:&list_headstoring&drm_gpuvm_boscurrently being
+ * evicted
+ */
+ structlist_headlist;
+
+ /**
+ * @lock: spinlock to protect the evict list + */
+ spinlock_t lock;
+ } evict;
};
void drm_gpuvm_init(structdrm_gpuvm*gpuvm, structdrm_device*drm,include/drm/drm_gpuvm.h +2
-0@@ -268,6 +302,21 @@ void drm_gpuvm_init(struct drm_gpuvm *gpuvm, struct drm_device *drm, const struct drm_gpuvm_ops *ops); void drm_gpuvm_destroy(struct drm_gpuvm *gpuvm); +/** +* drm_gpuvm_is_extobj() - indicates whether the given
&drm_gem_objectis an
+ * external object
+ * @gpuvm: the&drm_gpuvmto check
+ * @obj: the&drm_gem_objectto check
+ *
+ * Returns: true if the&drm_gem_object&dma_resvdiffers from the
+ *&drm_gpuvms&dma_resv, false otherwise
+ */
+static inline bool drm_gpuvm_is_extobj(structdrm_gpuvm*gpuvm,
+ structdrm_gem_object*obj)include/drm/drm_gpuvm.h +4
-0+{ + return obj && obj->resv != gpuvm->resv; +} +static inline struct
drm_gpuva*
__drm_gpuva_next(struct drm_gpuva *va)
{include/drm/drm_gpuvm.h +0
-0@@ -346,6 +395,128 @@ __drm_gpuva_next(struct drm_gpuva *va) #define drm_gpuvm_for_each_va_safe(va__, next__, gpuvm__) \list_for_each_entry_safe(va__,
next__, &(gpuvm__)-Quoted 1 line
rb.list, rb.entry)
+/**
+ * structdrm_gpuvm_exec-&drm_gpuvmabstraction of&drm_exec
+ *
+ * This structure should be created on the stack as&drm_execshould be.
+ *
+ * Optionally, @extra can be set in order to lock additional&drm_gem_objects.
+ */include/drm/drm_gpuvm.h +13
-0+struct drm_gpuvm_exec { + /** + * @exec: the &drm_exec structure + */ + struct drm_exec exec; + + /** + * @vm: the &drm_gpuvm to lock its DMA reservations + */ + struct drm_gpuvm *vm; + + /** + * @extra: Callback and corresponding private datafor the driver to
+ * lock arbitrary additional&drm_gem_objects. + */
+ struct {
+ /**
+ * @fn: The driver callback to lock additional&drm_gem_objects.
+ */
+ int (*fn)(structdrm_gpuvm_exec*vm_exec, + unsigned intnum_fences); +
+ /**
+ * @priv: driver private data for the @fn callback
+ */
+ void *priv;
+ } extra;include/drm/drm_gpuvm.h +4
-0+}; + +/** +* drm_gpuvm_prepare_vm() - prepare the GPUVMs common dma- resv
+ * @gpuvm: the&drm_gpuvm
+ * @exec: the&drm_execcontext
+ * @num_fences: the amount of&dma_fencesto reserve
+ *
+ * Callsdrm_exec_prepare_obj()for the GPUVMs dummy&drm_gem_object.
+ *
+ * Using this function directly, it is the drivers responsibility to call
+ *drm_exec_init()anddrm_exec_fini()accordingly.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */include/drm/drm_gpuvm.h +6
-0+static inline int +drm_gpuvm_prepare_vm (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + unsigned int num_fences) +{ + return drm_exec_prepare_obj(exec, &gpuvm->d_obj,
num_fences);include/drm/drm_gpuvm.h +16
-0+} + +int drm_gpuvm_prepare_objects (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + unsigned int num_fences); + +int drm_gpuvm_prepare_range (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + u64 addr, u64 range, + unsigned int num_fences); + +int drm_gpuvm_exec_lock (struct drm_gpuvm_exec *vm_exec, + unsigned int num_fences, + bool interruptible); + +int drm_gpuvm_exec_lock_array(struct drm_gpuvm_exec
*vm_exec,
+ structdrm_gem_object**objs, + unsigned intnum_objs, + unsigned intnum_fences, + bool interruptible);include/drm/drm_gpuvm.h +2
-0+ +int drm_gpuvm_exec_lock_range(struct drm_gpuvm_exec
*vm_exec,
+ u64 addr, u64 range, + unsigned intnum_fences, + bool interruptible);include/drm/drm_gpuvm.h +3
-0+ +/** +* drm_gpuvm_lock() - lock all dma-resv of all assoiciated BOs
+ * @gpuvm: the&drm_gpuvm
+ *
+ * Releases all dma-resv locks of all&drm_gem_objectspreviously acquired
+ * throughdrm_gpuvm_lock()or its variants.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */include/drm/drm_gpuvm.h +11
-0+static inline void +drm_gpuvm_exec_unlock (struct drm_gpuvm_exec *vm_exec) +{ + drm_exec_fini(&vm_exec->exec); +} + +int drm_gpuvm_validate (struct drm_gpuvm *gpuvm); +void drm_gpuvm_resv_add_fence (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + struct dma_fence *fence, + enum dma_resv_usage
private_usage,
+ enumdma_resv_usageextobj_usage);include/drm/drm_gpuvm.h +12
-0+ +/** +* drm_gpuvm_exec_resv_add_fence() +* @vm_exec: the &drm_gpuvm_exec abstraction +* @fence: fence to add +* @private_usage: private dma-resv usage +* @extobj_usage: extobj dma-resv usage +* +* See drm_gpuvm_resv_add_fence(). +*/ +static inline void +drm_gpuvm_exec_resv_add_fence(struct drm_gpuvm_exec
*vm_exec,
+ structdma_fence*fence, + enumdma_resv_usageprivate_usage,
+ enumdma_resv_usageextobj_usage)
+{
+ drm_gpuvm_resv_add_fence(vm_exec->vm,&vm_exec->exec, fence,
+private_usage,extobj_usage);include/drm/drm_gpuvm.h +2
-0+} +/**
* structdrm_gpuvm_bo- structure representing a&drm_gpuvmand
*&drm_gem_objectcombinationinclude/drm/drm_gpuvm.h +3
-0@@ -398,6 +569,18 @@ struct drm_gpuvm_bo {* gpuva list. */ struct list_head gem; + + /** +* @evict: List entry to attach to the
&drm_gpuvms
+ * extobj list.
+ */
+ structlist_headextobj; +
+ /**
+ * @evict: List entry to attach to the&drm_gpuvmsevict
+ * list.
+ */
+ structlist_headevict;
} entry;
} list;
};include/drm/drm_gpuvm.h +3
-0@@ -432,6 +615,9 @@ struct drm_gpuvm_bo * drm_gpuvm_bo_find(struct drm_gpuvm *gpuvm, struct drm_gem_object *obj); +void drm_gpuvm_bo_evict (struct drm_gem_object *obj,bool evict); +void drm_gpuvm_bo_extobj_add (struct drm_gpuvm_bo *vm_bo); + /*** drm_gpuvm_bo_for_each_va() - iterator to walk over a list of
&drm_gpuva
* @va__:&drm_gpuvastructure to assign to in each iteration stepinclude/drm/drm_gpuvm.h +0
-0@@ -837,6 +1023,17 @@ struct drm_gpuvm_ops {* used. */ int (*sm_step_unmap)(struct
drm_gpuva_op*op, void *priv);
+
+ /**
+ * @bo_validate: called fromdrm_gpuvm_validate()+ *
+ * Drivers receive this callback for every evicted&drm_gem_objectbeing
+ * mapped in the corresponding&drm_gpuvm. + *
+ * Typically, drivers would call their driver specific variant of
+ *ttm_bo_validate()from within this callback. + */
+ int (*bo_validate)(structdrm_gem_object*obj); };
int drm_gpuvm_sm_map(structdrm_gpuvm*gpuvm, void *priv,
Re: [PATCH drm-misc-next v3 6/7] drm/gpuvm: generalize dma_resv/extobj handling and GEM validation
Am 13.09.23 um 14:16 schrieb Danilo Krummrich:
Quoted 2 lines
As mentioned in a different mail thread, the reply is based on the assumption that we don't support anything else than GPUVM updates from the IOCTL.
I think that this assumption is incorrect.
Vulkan is just once specific use case, but this here should probably be able to handle other use cases as well.
Especially with HMM you get the requirement that you need to be able to invalidate GPUVM mappings without grabbing a reservation lock.
See what the eviction lock in amdgpu is doing for example.
Regards,
Christian.
Quoted 1,333 lines · Thomas Hellström
Quoted 197 lines
Hi!
Quoted 178 lines · Danilo Krummrich
Quoted 168 lines · Thomas Hellström
Quoted 161 lines · Danilo Krummrich
Quoted 142 lines · Thomas Hellström
Hi, Danilo,
Quoted 139 lines · Danilo Krummrich
So far the DRM GPUVA manager offers common infrastructure to track GPU VA
allocations and mappings, generically connect GPU VA mappings to their
backing buffers and perform more complex mapping operations on the GPU VA
space.However, there are more design patterns commonly used by drivers, which
can potentially be generalized in order to make the DRM GPUVA manager
represent a basic GPU-VM implementation. In this context, this patch aims
at generalizing the following elements.1) Provide a common dma-resv for GEM objects not being used outside of
this GPU-VM.2) Provide tracking of external GEM objects (GEM objects which are
shared with other GPU-VMs).3) Provide functions to efficiently lock all GEM objects dma- resv the
GPU-VM contains mappings of.4) Provide tracking of evicted GEM objects the GPU-VM contains mappings
of, such that validation of evicted GEM objects is
accelerated.5) Provide some convinience functions for common patterns.
Rather than being designed as a "framework", the target is to make all
features appear as a collection of optional helper functions, such that
drivers are free to make use of the DRM GPUVA managers basic functionality and opt-in for other features without setting any feature
flags, just by making use of the corresponding functions.Big kudos to Boris Brezillon for his help to figure out locking for drivers
updating the GPU VA space within the fence signalling path.Suggested-by Matthew Brost <[email protected]>Signed-off-by Danilo Krummrich <[email protected]>--- drivers/gpu/drm/drm_gpuvm.c | 516 ++++++++++++++++++++++++++++++++++++ include/drm/drm_gpuvm.h | 197 ++++++++++++++ 2 files changed, 713 insertions(+)drivers/gpu/drm/drm_gpuvm.c +0
-0@@ -73,6 +73,21 @@ * &drm_gem_object list of &drm_gpuvm_bos for an existinginstance of this
* particular combination. If not existent a new instance
is created and linked
* to the&drm_gem_object.
+ *
+ *&drm_gpuvm_bostructures, since unique for a given&drm_gpuvm, are also used
+ * as entry for the &drm_gpuvm's lists of external and evicted objects. Those
+ * list are maintained in order to accelerate locking of dma-resv locks and
+ * validation of evicted objects bound in a&drm_gpuvm. For instance the all
+ * &drm_gem_object's&dma_resvof a given&drm_gpuvmcan be locked by calling
+ *drm_gpuvm_exec_lock(). Once locked drivers can calldrm_gpuvm_validate()in
+ * order to validate all evicted&drm_gem_objects. It is also possible to lock
+ * additional&drm_gem_objectsby providing the corresponding parameters to
+ *drm_gpuvm_exec_lock()as well as open code the&drm_execloop while making
+ * use of helper functions such asdrm_gpuvm_prepare_range()or
+ *drm_gpuvm_prepare_objects().
+ *
+ * Every bound&drm_gem_objectis treated as external object when its&dma_resv
+ * structure is different than the &drm_gpuvm's common&dma_resvstructure.
*/
/**drivers/gpu/drm/drm_gpuvm.c +0
-0@@ -420,6 +435,20 @@ * Subsequent calls to drm_gpuvm_bo_obtain() for the same
&drm_gpuvmand
*&drm_gem_objectmust be able to observe previous
creations and destructions
* of&drm_gpuvm_bosin order to keep instances unique.
+ *
+ * The &drm_gpuvm's lists for keeping track of external and evicted objects are
+ * protected against concurrent insertion / removal and iteration internally.
+ *
+ * However, drivers still need ensure to protect concurrent calls to functions
+ * iterating those lists, such asdrm_gpuvm_validate()and
+ *drm_gpuvm_prepare_objects(). Every such function contains a particular
+ * comment and lockdep checks if possible.
+ *
+ * Functions adding or removing entries from those lists, such as
+ *drm_gpuvm_bo_evict()ordrm_gpuvm_bo_extobj_add()may be called with external
+ * locks being held, e.g. in order to avoid the corresponding list to be
+ * (safely) modified while potentially being iternated by other API functions.
+ * However, this is entirely optional.
*/
/**drivers/gpu/drm/drm_gpuvm.c +5
-0@@ -632,6 +661,131 @@* } */ +/** +* get_next_vm_bo_from_list() - get the next vm_bo element +* @__gpuvm: The GPU VM +* @__list_name: The name of the list we're iterating on +* @__local_list: A pointer to the local list used to store already iterated items
+ * @__prev_vm_bo: The previous element we got fromdrm_gpuvm_get_next_cached_vm_bo()
+ *
+ * This helper is here to provide lockless list iteration.Quoted 8 lines
Quoted 5 lines
Lockless as in, the
+ * iterator releases the lock immediately after picking the first element from
+ * the list, so list insertion deletion can happen concurrently.Are the list spinlocks needed for that async state update from within the
dma-fence critical section we've discussed previously?Yes, but also for other reasons, see below.
Quoted 8 lines
Otherwise it should be sufficient to protect the lists with the gpuvm's resv
(or for the extobj list with an outer lock).If those spinlocks are still needed in some situations, perhaps could we
have an option to set them to NULL (Like IIRC the maple tree allows for)?Quoted 6 lines
The evict spinlock is needed in any case, since in
drm_gpuvm_bo_evict()we're
holding only the dma-resv lock from the BO this function gets called for. Hence,
the spinlock protects concurrentdrm_gpuvm_bo_evict()calls with different BOs.Quoted 9 lines
No. Only if you try to add external objects to the vm's evict list from
within the evict code. That's not necessary since you loop through all
external objects anyway when locking them so an "evicted" bool in thevm_bo,
protected by the bo resv would be sufficient. The extobj locking loop can
then add the bo to the evicted list.Quoted 16 lines
And
validate()can remove it while still holding all dma-resv locks, neat!
However, what if two tasks are trying to lock the VA space concurrently? What
do we do when the drm_gpuvm_bo's refcount drops to zero indrm_gpuva_unlink()?
Are we guaranteed that at this point of time thedrm_gpuvm_bois not on the
evicted list? Because otherwise we would calldrm_gpuvm_bo_destroy()with the
dma-resv lock held, which wouldn't be allowed, sincedrm_gpuvm_bo_destroy()
might drop the last reference to thedrm_gem_objectand hence we'd potentially
free the dma-resv lock while holding it, at least if it's an external object.Quoted 4 lines
Easiest way in this scheme is to think of the lists as being protected by the vm's resv lock. That means anybody calling
unlink()must also hold the vm's resv lock. (Which is OK from an UAF point of view, but perhaps not from a locking inversion POW from an async list update).This would mean that on
unlink()we'd need to hold the VM's resv lock and the corresponding GEM's resv lock (in case they're not the same anyways) because the VM's resv lock would protect the external / evicted object lists and the GEM objects resv lock protects the GEM's list ofdrm_gpuvm_bosand the drm_gpuvm_bo's list ofdrm_gpuvas.Quoted 49 lines
Quoted 38 lines
Quoted 5 lines
Quoted 5 lines
For extobjs an outer lock would be enough in case of Xe, but I really would not
like to add even more complexity just to get the spinlock out of the way in case
the driver already has an outer lock protecting this path.Quoted 15 lines
I must disagree here. These spinlocks and atomic operations are pretty
costly and as discussed earlier this type of locking was the reason (at
least according to the commit message) that made Christian drop the XArray
use indrm_execfor the same set of objects: "The locking overhead is
unecessary and measurable". IMHO the spinlock is the added complexity and a
single wide lock following the drm locking guidelines set out by Daniel and
David should really be the default choice with an opt-in for a spinlock if
needed for async and pushing out to a wq is not an option.For the external object list an outer lock would work as long as it's not the
dma-resv lock of the corresponding GEM object, since here we actually need to
remove the list entry from the external object list ondrm_gpuvm_bo_destroy().
It's just a bit weird design wise that drivers would need to take this outer
lock on:-
drm_gpuvm_bo_extobj_add()
- drm_gpuvm_bo_destroy() (and hence alsodrm_gpuvm_bo_put())
- drm_gpuva_unlink() (because it needs to calldrm_gpuvm_bo_put())
-drm_gpuvm_exec_lock()
-drm_gpuvm_exec_lock_array()
-drm_gpuvm_prepare_range()Quoted 2 lines
Given that it seems reasonable to do all the required locking internally.
From a design POW, there has been a clear direction in XE to make
things similar tommap()/munmap(), so this outer lock, which in Xe is an rwsem, is used in a similar way as themmap_lock. It's protecting the page-table structures and vma rb tree, the userptr structures and the extobj list. Basically it's taken early in the exec IOCTL, theVM_BINDioctl, the compute rebind worker and the pagefault handler, so all of the above are just asserting that it is taken in the correct mode.Quoted 6 lines
But strictly with this scheme one could also use the vm's
dma_resvfor the extobj list since withdrm_exec, it's locked before traversing the list.The whole point of this scheme is to rely on locks that you already are supposed to be holding for various reasons and is simple to comprehend.
I don't agree that we're supposed to hold the VM's resv lock anyways for functions like
drm_gpuvm_bo_put()ordrm_gpuva_unlink(), but I'm fine using it for that purpose nevertheless.Quoted 6 lines
Quoted 5 lines
In order to at least place lockdep checks, the driver would need to supply the
corresponding lock'slockdep_map, because the GPUVM otherwise doesn't know about
the lock.Yes, that sounds reasonable. One lockdep map per list.
I'd really like to avoid that, especially now that everything got simpler. We should define the actual locks to take instead.
Quoted 45 lines
Quoted 3 lines
Out of curiosity, what is the overhead of a
spin_lock()that doesn't need to
spin?I guess it's hard to tell exactly, but it is much lower on modern x86 than what it used to be. Not sure about ARM, which is the other architecture important to us. I figure if there is little cache-line bouncing the main overhead comes from the implied barriers.
Quoted 34 lines
Quoted 25 lines
A pretty simple way that would not add much code would be
static void gpuvm_cond_spin_lock(const struct
drm_gpuvm*gpuvm,spinlock_t
*lock){
if (!gpuvm->resv_protected_lists)
spin_lock(lock);}
Quoted 12 lines
Quoted 3 lines
For such drivers, that would require anybody calling unlink to hold the vm's
resv, though.In V4 I want to go back to having a dedicated lock for the GEMs gpuva list (or
VM_BOlist to be more precise). We can't just use the dma-resv lock for that
withVM_BOabstractions, because on destruction of aVM_BOwe otherwise wouldn't
be allowed to already hold the dma-resv lock. That's the fix I was referring to
earlier.Quoted 9 lines
Yeah, I can see the need for a dedicated lock for the GEM's gpuva list, but
holding the vm's dma-resv lock across the unlink shouldn't be a problem. We
may free the object and a pointer to the vm's resv during unlink but we
don't free the vm's resv. It'd be a matter of ensuring that any calls to
unlink from *within*drm_gpuvmallows it to be held.Quoted 3 lines
Drivers calling
unlink()from the fence signaling path can't use the VM's
dma-resv lock.Quoted 6 lines
Yes, that made me a bit curious because in the current version the code required the object's
dma_resvforunlink()which can't be grabbed either from the fence signaling path. So are there any drivers actually wanting to do that? If so, they will either need to resort to the current spinlock solution or they will need to call unlink from a workqueue item.As Boris already mentioned we have the dma-resv lock by default or a driver specific GEM gpuva lock as opt-in. Now, we can get rid of the latter.
Quoted 9 lines
Quoted 3 lines
Also, what if the object is an external object? We can't use the VM's dma-resv
lock here.Why? Typically (sync) unlink is only ever called from an unbind-like operation where it should be trivial to grab the vm's resv. Or, for that matter any outer lock protecting the extobj list. Rule would be the
drm_gpuvm_bo::entry::extobjanddrm_gpuvm_bo::entry::evictwould be protected by either the vm'sdma_resv(or possibly an outer lock in the case of the extobj list).Outer lock wouldn't have been working for updates in the async path, but shouldn't be relevant anymore. We could use the VM's resv for that.
Quoted 6 lines
Quoted 6 lines
And we can't have the GEM objs dma-resv lock held when calling
unlink(), sinceunlink()callsdrm_gpuvm_bo_put(), which if the refcount drops
to zero callsdrm_gpuvm_bo_destroy()anddrm_gpuvm_bo_destroy()might drop the
last reference of the GEM object.Quoted 10 lines
Yes, but this is a different problem as to what exactly protects
drm_gpuvm_bo::entry::gem. Either as you suggest an internal per bo list lock, or if we want to keep the bo'sdma_resvwe need to ensure that the caller ofdma_resv_unlock(obj->resv)actually refcounts its obj pointer, and doesn't implicitly rely on the gpuvm_bo's refcount (I know Boris didn't like that, but requiring an explicit refcount for a pointer you dereference unless you're under a lock that ensures keeping the object alive is pretty much required?) But anyway for thedrm_gpuvm_bo::entry::gemlist protection (bo resv or internal spinlock) I don't have a strong preference.We can keep the GEM objects dma-resv lock, however as mentioned above
drm_gpuva_unlink()anddrm_gpuvm_bo_put()then requires both the VM's resv lock and the GEM's resv lock in case they differ.Quoted 970 lines
Quoted 2 lines
All those problems go away with a dedicated
GEM gpuva list lock.I don't think these are real problems.
With the excepton of the eviction list "trick" where we currently have slightly different approach to collect external bos needing rebinding, we have this working fine.TBH I think pretty much the only situation where the spinlock is needed is for async updates of these lists, unless a wq item can be used for that, but it doesn't really seem like the current code allows for such updates anyway? It complicates the code a lot, adds overhead and also adds the requirement for refcounting during list traversal.
/Thomas
Quoted 955 lines
Quoted 955 lines
/Thomas
Quoted 952 lines
Quoted 952 lines
It seems that with that also the refcount could be make non- atomic.
All in the spirit of the drm locking guidelines "use big locks when
possible".
Lower level locks only when necessary for performance or locking inversion?/Thomas
Quoted 940 lines
+ *
+ * Elements popped from the original list are kept in a local list, so removal
+ * andis_emptychecks can still happen while we're iterating the list.
+ */
+#defineget_next_vm_bo_from_list(__gpuvm, __list_name, __local_list, __prev_vm_bo)\
+ ({
\
+ structdrm_gpuvm_bo
*__vm_bo; \drivers/gpu/drm/drm_gpuvm.c +4
-0+ \ + drm_gpuvm_bo_put(__prev_vm_bo); \ + \ + spin_lock(&(__gpuvm)-Quoted 1 line
__list_name.lock); \
+ while (!list_empty(&(__gpuvm)-
Quoted 1 line
__list_name.list)) { \
+ __vm_bo = list_first_entry(&(__gpuvm)->__list_name.list, \
+ struct drm_gpuvm_bo, \
+
list.entry.__list_name); \
+ if (drm_gpuvm_bo_get_unless_zero(__vm_bo))
{ \
+ list_move_tail(&(__vm_bo)-Quoted 1 line
list.entry.__list_name, \
+
__local_list); \
+ break;
\
+ } else
{ \
+ list_del_init(&(__vm_bo)-Quoted 1 line
list.entry.__list_name); \
+ __vm_bo =
NULL; \
+ }
\
+ }
\
+ spin_unlock(&(__gpuvm)-Quoted 1 line
__list_name.lock); \
drivers/gpu/drm/drm_gpuvm.c +8
-0+ \ + __vm_bo; \ + }) + +/** +* for_each_vm_bo_in_list() - internal vm_bo list iterator +* +* This helper is here to provide lockless list iteration. Lockless as in, the
+ * iterator releases the lock immediately after picking the first element from the
+ * list, so list insertion and deletion can happen concurrently.
+ *
+ * Typical use:
+ *
+ * structdrm_gpuvm_bo*vm_bo;
+ * LIST_HEAD(my_local_list);
+ *
+ * ret = 0;
+ * drm_gpuvm_for_each_vm_bo(gpuvm, <list_name>,&my_local_list,vm_bo) {
+ * ret =do_something_with_vm_bo(..., vm_bo);
+ * if (ret)
+ * break;
+ * }
+ * drm_gpuvm_bo_put(vm_bo);
+ * drm_gpuvm_restore_vm_bo_list(gpuvm, <list_name>,&my_local_list);
+ *
+ *
+ * Only used for internal list iterations, not meant to be exposed to the outside
+ * world.
+ */
+#definefor_each_vm_bo_in_list(__gpuvm, __list_name, __local_list, __vm_bo)\
+ for (__vm_bo= get_next_vm_bo_from_list(__gpuvm, __list_name, \
+ __local_list, NULL); \
+__vm_bo;
\
+__vm_bo= get_next_vm_bo_from_list(__gpuvm, __list_name, \
+ __local_list, __vm_bo)) \drivers/gpu/drm/drm_gpuvm.c +3
-0+ +/** +* restore_vm_bo_list() - move vm_bo elements back to their original list
+ * @__gpuvm: The GPU VM
+ * @__list_name: The name of the list we're iterating on
+ * @__local_list: A pointer to the local list used to store already iterated items
+ *
+ * When we're done iterating avm_bolist, we should callrestore_vm_bo_list()
+ * to restore the original state and let new iterations take place.
+ */
+#definerestore_vm_bo_list(__gpuvm, __list_name, __local_list)\
+ do
{
\
+ /* Merge back the two lists, moving local list elements to the \
+ * head to preserve previous ordering, in case it matters. \
+
*/
\
+ spin_lock(&(__gpuvm)-Quoted 1 line
__list_name.lock); \
+ list_splice(__local_list, &(__gpuvm)-
Quoted 1 line
__list_name.list); \
+ spin_unlock(&(__gpuvm)-
Quoted 1 line
__list_name.lock); \
+ } while (0)
+/**
+ *drm_gpuvm_bo_list_add()- insert avm_bointo the given list
+ * @__vm_bo: the&drm_gpuvm_bo
+ * @__list_name: the name of the list to insert into
+ *
+ * Inserts the given @__vm_bo into the list specified by @__list_name and
+ * increases the vm_bo's reference count.
+ */
+#definedrm_gpuvm_bo_list_add(__vm_bo, __list_name)\
+ do
{
\
+ spin_lock(&(__vm_bo)->vm-Quoted 1 line
__list_name.lock); \
+ if (list_empty(&(__vm_bo)-
Quoted 1 line
list.entry.__list_name)) \
+ list_add_tail(&(__vm_bo)-
Quoted 1 line
list.entry.__list_name, \
+ &(__vm_bo)->vm-
Quoted 1 line
__list_name.list); \
+ spin_unlock(&(__vm_bo)->vm-
Quoted 1 line
__list_name.lock); \
+ } while (0)
drivers/gpu/drm/drm_gpuvm.c +3
-0+ +/** +* drm_gpuvm_bo_list_del() - remove a vm_bo from the given list
+ * @__vm_bo: the&drm_gpuvm_bo
+ * @__list_name: the name of the list to insert into
+ *
+ * Removes the given @__vm_bo from the list specified by @__list_name and
+ * decreases the vm_bo's reference count.
+ */
+#definedrm_gpuvm_bo_list_del(__vm_bo, __list_name)\
+ do
{
\
+ spin_lock(&(__vm_bo)->vm-Quoted 1 line
__list_name.lock); \
+ if (!list_empty(&(__vm_bo)-
Quoted 1 line
list.entry.__list_name)) \
+ list_del_init(&(__vm_bo)-
Quoted 1 line
list.entry.__list_name); \
+ spin_unlock(&(__vm_bo)->vm-
Quoted 1 line
__list_name.lock); \
+ } while (0)
drivers/gpu/drm/drm_gpuvm.c +4
-0+ +static int __must_check +drm_gpuvm_bo_get_unless_zero (struct drm_gpuvm_bo *vm_bo); + #define to_drm_gpuva(__node) container_of((__node), struct
drm_gpuva, rb.node)
#defineGPUVA_START(node)((node)->va.addr)drivers/gpu/drm/drm_gpuvm.c +16
-0@@ -713,6 +867,12 @@ drm_gpuvm_init(struct drm_gpuvm *gpuvm,struct drm_device *drm, gpuvm->rb.tree = RB_ROOT_CACHED; INIT_LIST_HEAD(&gpuvm->rb.list); + INIT_LIST_HEAD(&gpuvm->extobj.list); + spin_lock_init(&gpuvm->extobj.lock); + + INIT_LIST_HEAD(&gpuvm->evict.list); + spin_lock_init(&gpuvm->evict.lock); + drm_gpuva_check_overflow(start_offset, range); gpuvm->mm_start = start_offset; gpuvm->mm_range = range;@@ -754,10 +914,302 @@ drm_gpuvm_destroy(struct drm_gpuvm *gpuvm) WARN(!RB_EMPTY_ROOT(&gpuvm->rb.tree.rb_root), "GPUVA tree is not empty, potentially leaking memory.\n"); + WARN(!list_empty(&gpuvm->extobj.list), "Extobj list should be empty.\n"); + WARN(!list_empty(&gpuvm->evict.list), "Evict list should be empty.\n"); + drm_gem_private_object_fini(&gpuvm->d_obj); } EXPORT_SYMBOL_GPL(drm_gpuvm_destroy); +/** +* drm_gpuvm_prepare_objects() - prepare all assoiciated BOs +* @gpuvm: the &drm_gpuvm +* @exec: the &drm_exec locking context +* @num_fences: the amount of &dma_fences to reserve +* +* Calls drm_exec_prepare_obj() for all &drm_gem_objects the given
+ *&drm_gpuvmcontains mappings of.
+ *
+ * Using this function directly, it is the drivers responsibility to call
+ *drm_exec_init()anddrm_exec_fini()accordingly.
+ *
+ * Note: This function is safe against concurrent insertion and removal of
+ * external objects, however it is not safe against concurrent usage itself.
+ *
+ * Drivers need to make sure to protect this case with either an outer VM lock
+ * or by callingdrm_gpuvm_prepare_vm()before this function within the
+ *drm_exec_until_all_locked()loop, such that the GPUVM's dma-resv lock ensures
+ * mutual exclusion.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +10
-0+int +drm_gpuvm_prepare_objects (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + unsigned int num_fences) +{ + struct drm_gpuvm_bo *vm_bo; + LIST_HEAD(extobjs); + int ret = 0; + + for_each_vm_bo_in_list(gpuvm, extobj, &extobjs,
vm_bo) {
+ ret =drm_exec_prepare_obj(exec, vm_bo->obj, num_fences);
+ if (ret)
+ break;
+ }
+ /* Drop ref in case we break out of the loop. */ + drm_gpuvm_bo_put(vm_bo);
+ restore_vm_bo_list(gpuvm, extobj, &extobjs);drivers/gpu/drm/drm_gpuvm.c +7
-0+ + return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_prepare_objects); + +/** +* drm_gpuvm_prepare_range() - prepare all BOs mapped within a given range
+ * @gpuvm: the&drm_gpuvm
+ * @exec: the&drm_execlocking context
+ * @addr: the start address within the VA space
+ * @range: the range to iterate within the VA space
+ * @num_fences: the amount of&dma_fencesto reserve
+ *
+ * Callsdrm_exec_prepare_obj()for all&drm_gem_objectsmapped between @addr
+ * and @addr + @range.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +2
-0+int +drm_gpuvm_prepare_range(struct drm_gpuvm *gpuvm, struct
drm_exec*exec,
+ u64 addr, u64 range, unsigned intnum_fences)drivers/gpu/drm/drm_gpuvm.c +9
-0+{ + struct drm_gpuva *va; + u64 end = addr + range; + int ret; + + drm_gpuvm_for_each_va_range(va, gpuvm, addr, end) { + struct drm_gem_object *obj = va->gem.obj; + + ret = drm_exec_prepare_obj(exec, obj,
num_fences);
+ if (ret)
+ return ret;
+ }drivers/gpu/drm/drm_gpuvm.c +7
-0+ + return 0; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_prepare_range); + +/** +* drm_gpuvm_exec_lock() - lock all dma-resv of all assoiciated BOs
+ * @vm_exec: the&drm_gpuvm_execabstraction
+ * @num_fences: the amount of&dma_fencesto reserve
+ * @interruptible: sleep interruptible if waiting
+ *
+ * Acquires all dma-resv locks of all&drm_gem_objectsthe given
+ *&drm_gpuvmcontains mappings of.
+ *
+ * Addionally, when calling this function with structdrm_gpuvm_exec::extra
+ * being set the driver receives the given @fn callback to lock additional
+ * dma-resv in the context of the&drm_gpuvm_execinstance. Typically, drivers
+ * would calldrm_exec_prepare_obj()from within this callback.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +11
-0+int +drm_gpuvm_exec_lock (struct drm_gpuvm_exec *vm_exec, + unsigned int num_fences, + bool interruptible) +{ + struct drm_gpuvm *gpuvm = vm_exec->vm; + struct drm_exec *exec = &vm_exec->exec; + uint32_t flags; + int ret; + + flags = interruptible ? DRM_EXEC_INTERRUPTIBLE_WAIT :0 |
+ DRM_EXEC_IGNORE_DUPLICATES;drivers/gpu/drm/drm_gpuvm.c +5
-0+ + drm_exec_init(exec, flags); + + drm_exec_until_all_locked(exec) { + ret = drm_gpuvm_prepare_vm(gpuvm, exec,
num_fences);
+ drm_exec_retry_on_contention(exec); + if (ret)
+ goto err;
+
+ ret =drm_gpuvm_prepare_objects(gpuvm, exec, num_fences);
+ drm_exec_retry_on_contention(exec); + if (ret)
+ goto err;
+
+ if (vm_exec->extra.fn) {
+ ret =vm_exec->extra.fn(vm_exec, num_fences);
+ drm_exec_retry_on_contention(exec); + if (ret)
+ goto err;
+ }
+ }drivers/gpu/drm/drm_gpuvm.c +11
-0+ + return 0; + +err: + drm_exec_fini(exec); + return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock); + +static int +fn_lock_array(struct drm_gpuvm_exec *vm_exec, unsigned int
num_fences)drivers/gpu/drm/drm_gpuvm.c +7
-0+{ + struct { + struct drm_gem_object **objs; + unsigned int num_objs; + } *args = vm_exec->extra.priv; + + return drm_exec_prepare_array(&vm_exec->exec, args-Quoted 1 line
objs,
+
args->num_objs,num_fences);drivers/gpu/drm/drm_gpuvm.c +4
-0+} + +/** +* drm_gpuvm_exec_lock_array() - lock all dma-resv of all assoiciated BOs
+ * @vm_exec: the&drm_gpuvm_execabstraction
+ * @objs: additional&drm_gem_objectsto lock
+ * @num_objs: the number of additional&drm_gem_objectsto lock
+ * @num_fences: the amount of&dma_fencesto reserve
+ * @interruptible: sleep interruptible if waiting
+ *
+ * Acquires all dma-resv locks of all&drm_gem_objectsthe given&drm_gpuvm
+ * contains mappings of, plus the ones given through @objs.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +19
-0+int +drm_gpuvm_exec_lock_array (struct drm_gpuvm_exec *vm_exec, + struct drm_gem_object **objs, + unsigned int num_objs, + unsigned int num_fences, + bool interruptible) +{ + struct { + struct drm_gem_object **objs; + unsigned int num_objs; + } args; + + args.objs = objs; + args.num_objs = num_objs; + + vm_exec->extra.fn = fn_lock_array; + vm_exec->extra.priv = &args; + + return drm_gpuvm_exec_lock(vm_exec, num_fences,interruptible);
drivers/gpu/drm/drm_gpuvm.c +5
-0+} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock_array); + +/** +* drm_gpuvm_exec_lock_range() - prepare all BOs mapped within a given range
+ * @vm_exec: the&drm_gpuvm_execabstraction
+ * @addr: the start address within the VA space
+ * @range: the range to iterate within the VA space
+ * @num_fences: the amount of&dma_fencesto reserve
+ * @interruptible: sleep interruptible if waiting
+ *
+ * Acquires all dma-resv locks of all&drm_gem_objectsmapped between @addr and
+ * @addr + @range.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +12
-0+int +drm_gpuvm_exec_lock_range (struct drm_gpuvm_exec *vm_exec, + u64 addr, u64 range, + unsigned int num_fences, + bool interruptible) +{ + struct drm_gpuvm *gpuvm = vm_exec->vm; + struct drm_exec *exec = &vm_exec->exec; + uint32_t flags; + int ret; + + flags = interruptible ? DRM_EXEC_INTERRUPTIBLE_WAIT :0 |
+ DRM_EXEC_IGNORE_DUPLICATES;drivers/gpu/drm/drm_gpuvm.c +5
-0+ + drm_exec_init(exec, flags); + + drm_exec_until_all_locked(exec) { + ret = drm_gpuvm_prepare_range(gpuvm, exec,addr, range,
+num_fences); + drm_exec_retry_on_contention(exec); + if (ret)
+ goto err;
+ }drivers/gpu/drm/drm_gpuvm.c +14
-0+ + return ret; + +err: + drm_exec_fini(exec); + return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock_range); + +/** +* drm_gpuvm_validate() - validate all BOs marked as evicted +* @gpuvm: the &drm_gpuvm to validate evicted BOs +* +* Calls the &drm_gpuvm_ops.bo_validate callback for all evicted buffer
+ * objects being mapped in the given&drm_gpuvm.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +27
-0+int +drm_gpuvm_validate (struct drm_gpuvm *gpuvm) +{ + const struct drm_gpuvm_ops *ops = gpuvm->ops; + struct drm_gpuvm_bo *vm_bo; + LIST_HEAD(evict); + int ret = 0; + + if (unlikely(!ops || !ops->bo_validate)) + return -ENOTSUPP; + + for_each_vm_bo_in_list(gpuvm, evict, &evict, vm_bo) { + dma_resv_assert_held(vm_bo->obj->resv); + ret = ops->bo_validate(vm_bo->obj); + if (ret) + break; + } + /* Drop ref in case we break out of the loop. */ + drm_gpuvm_bo_put(vm_bo); + restore_vm_bo_list(gpuvm, evict, &evict); + + return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_validate); + +/** +* drm_gpuvm_resv_add_fence - add fence to private and all extobj
+ * dma-resv
+ * @gpuvm: the&drm_gpuvmto add a fence to
+ * @exec: the&drm_execlocking context
+ * @fence: fence to add
+ * @private_usage: private dma-resv usage
+ * @extobj_usage: extobj dma-resv usage
+ */drivers/gpu/drm/drm_gpuvm.c +14
-0+void +drm_gpuvm_resv_add_fence (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + struct dma_fence *fence, + enum dma_resv_usage private_usage, + enum dma_resv_usage extobj_usage) +{ + struct drm_gem_object *obj; + unsigned long index; + + drm_exec_for_each_locked_object(exec, index, obj) { + dma_resv_assert_held(obj->resv); + dma_resv_add_fence(obj->resv, fence, + drm_gpuvm_is_extobj(gpuvm,obj) ?
+private_usage:extobj_usage);
+ }drivers/gpu/drm/drm_gpuvm.c +3
-0+} +EXPORT_SYMBOL_GPL (drm_gpuvm_resv_add_fence); + /*** drm_gpuvm_bo_create() - create a new instance of struct
drm_gpuvm_bo
* @gpuvm: The&drm_gpuvmthe @obj is mapped in.drivers/gpu/drm/drm_gpuvm.c +13
-0@@ -790,6 +1242,9 @@ drm_gpuvm_bo_create(struct drm_gpuvm *gpuvm, INIT_LIST_HEAD(&vm_bo->list.gpuva); INIT_LIST_HEAD(&vm_bo->list.entry.gem); + INIT_LIST_HEAD(&vm_bo->list.entry.extobj); + INIT_LIST_HEAD(&vm_bo->list.entry.evict); + drm_gem_object_get(obj); return vm_bo;@@ -807,6 +1262,14 @@ drm_gpuvm_bo_destroy(struct kref *kref) drm_gem_gpuva_assert_lock_held(vm_bo->obj); + spin_lock(&gpuvm->extobj.lock); + list_del(&vm_bo->list.entry.extobj); + spin_unlock(&gpuvm->extobj.lock); + + spin_lock(&gpuvm->evict.lock); + list_del(&vm_bo->list.entry.evict); + spin_unlock(&gpuvm->evict.lock); + list_del(&vm_bo->list.entry.gem); drm_gem_object_put(obj);@@ -822,6 +1285,11 @@ drm_gpuvm_bo_destroy(struct kref *kref) * @vm_bo: the &drm_gpuvm_bo to release the reference of * * This releases a reference to @vm_bo. + * + * If the reference count drops to zero, the &gpuvm_bo isdestroyed, which
+ * includes removing it from the GEMs gpuva list. Hence, if a call to this
+ * function can potentially let the reference count to zero the caller must
+ * hold the dma-resv or driver specific GEM gpuva lock. */
void
drm_gpuvm_bo_put(struct drm_gpuvm_bo *vm_bo)drivers/gpu/drm/drm_gpuvm.c +8
-0@@ -831,6 +1299,12 @@ drm_gpuvm_bo_put(struct drm_gpuvm_bo *vm_bo) } EXPORT_SYMBOL_GPL(drm_gpuvm_bo_put); +static int __must_check +drm_gpuvm_bo_get_unless_zero (struct drm_gpuvm_bo *vm_bo) +{ + return kref_get_unless_zero(&vm_bo->kref); +} + static struct drm_gpuvm_bo * __drm_gpuvm_bo_find(struct drm_gpuvm *gpuvm, struct drm_gem_object *obj)@@ -938,6 +1412,48 @@ drm_gpuvm_bo_obtain_prealloc(struct drm_gpuvm_bo *__vm_bo) } EXPORT_SYMBOL_GPL(drm_gpuvm_bo_obtain_prealloc); +/** +* drm_gpuvm_bo_extobj_add() - adds the &drm_gpuvm_bo to its &drm_gpuvm's
+ * extobj list
+ * @vm_bo: The&drm_gpuvm_boto add to its &drm_gpuvm's the extobj list.
+ *
+ * Adds the given @vm_bo to its &drm_gpuvm's extobj list if not on the list
+ * already and if the corresponding&drm_gem_objectis an external object,
+ * actually.
+ */drivers/gpu/drm/drm_gpuvm.c +12
-0+void +drm_gpuvm_bo_extobj_add (struct drm_gpuvm_bo *vm_bo) +{ + struct drm_gpuvm *gpuvm = vm_bo->vm; + + if (drm_gpuvm_is_extobj(gpuvm, vm_bo->obj)) + drm_gpuvm_bo_list_add(vm_bo, extobj); +} +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_extobj_add); + +/** +* drm_gpuvm_bo_evict() - add / remove a &drm_gem_object to / from a
+ *&drm_gpuvmsevicted list
+ * @obj: the&drm_gem_objectto add or remove
+ * @evict: indicates whether the object is evicted
+ *
+ * Adds a&drm_gem_objectto or removes it from all&drm_gpuvmsevicted
+ * list containing a mapping of this&drm_gem_object.
+ */drivers/gpu/drm/drm_gpuvm.c +14
-0+void +drm_gpuvm_bo_evict (struct drm_gem_object *obj,bool evict) +{ + struct drm_gpuvm_bo *vm_bo; + + drm_gem_for_each_gpuvm_bo(vm_bo, obj) { + if (evict) + drm_gpuvm_bo_list_add(vm_bo, evict); + else + drm_gpuvm_bo_list_del(vm_bo, evict); + } +} +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_evict); + static int __drm_gpuva_insert(struct drm_gpuvm *gpuvm, struct drm_gpuva *va)include/drm/drm_gpuvm.h +9
-0@@ -26,10 +26,12 @@*/ #include <linux/list.h> +#include <linux/dma-resv.h> #include <linux/rbtree.h> #include <linux/types.h> #include <drm/drm_gem.h> +#include <drm/drm_exec.h> struct drm_gpuvm; struct drm_gpuvm_bo;@@ -259,6 +261,38 @@ struct drm_gpuvm {* space */ struct dma_resv *resv; + + /** + * @extobj: structure holding the extobj list + */ + struct { + /** +* @list: &list_head storing &drm_gpuvm_bos serving as
+ * external object
+ */
+ structlist_headlist;include/drm/drm_gpuvm.h +9
-0+ + /** + * @lock: spinlock to protect the extobj list + */ + spinlock_t lock; + } extobj; + + /** +* @evict: structure holding the evict list and evict list lock
+ */
+ struct {
+ /**
+ * @list:&list_headstoring&drm_gpuvm_boscurrently being
+ * evicted
+ */
+ structlist_headlist;
+
+ /**
+ * @lock: spinlock to protect the evict list + */
+ spinlock_t lock;
+ } evict;
};
void drm_gpuvm_init(structdrm_gpuvm*gpuvm, structdrm_device*drm,include/drm/drm_gpuvm.h +2
-0@@ -268,6 +302,21 @@ void drm_gpuvm_init(struct drm_gpuvm *gpuvm, struct drm_device *drm, const struct drm_gpuvm_ops *ops); void drm_gpuvm_destroy(struct drm_gpuvm *gpuvm); +/** +* drm_gpuvm_is_extobj() - indicates whether the given
&drm_gem_objectis an
+ * external object
+ * @gpuvm: the&drm_gpuvmto check
+ * @obj: the&drm_gem_objectto check
+ *
+ * Returns: true if the&drm_gem_object&dma_resvdiffers from the
+ *&drm_gpuvms&dma_resv, false otherwise
+ */
+static inline bool drm_gpuvm_is_extobj(structdrm_gpuvm*gpuvm,
+ structdrm_gem_object*obj)include/drm/drm_gpuvm.h +6
-0+{ + return obj && obj->resv != gpuvm->resv; +} + static inline struct drm_gpuva * __drm_gpuva_next(struct drm_gpuva *va) {@@ -346,6 +395,128 @@ __drm_gpuva_next(struct drm_gpuva *va) #define drm_gpuvm_for_each_va_safe(va__, next__, gpuvm__) \ list_for_each_entry_safe(va__, next__, &(gpuvm__)- > rb.list, rb.entry) +/** +* struct drm_gpuvm_exec - &drm_gpuvm abstraction of
&drm_exec
+ *
+ * This structure should be created on the stack as&drm_execshould be.
+ *
+ * Optionally, @extra can be set in order to lock additional&drm_gem_objects.
+ */include/drm/drm_gpuvm.h +13
-0+struct drm_gpuvm_exec { + /** + * @exec: the &drm_exec structure + */ + struct drm_exec exec; + + /** + * @vm: the &drm_gpuvm to lock its DMA reservations + */ + struct drm_gpuvm *vm; + + /** + * @extra: Callback and corresponding private datafor the driver to
+ * lock arbitrary additional&drm_gem_objects. + */
+ struct {
+ /**
+ * @fn: The driver callback to lock additional&drm_gem_objects.
+ */
+ int (*fn)(structdrm_gpuvm_exec*vm_exec, + unsigned intnum_fences); +
+ /**
+ * @priv: driver private data for the @fn callback
+ */
+ void *priv;
+ } extra;include/drm/drm_gpuvm.h +4
-0+}; + +/** +* drm_gpuvm_prepare_vm() - prepare the GPUVMs common dma- resv
+ * @gpuvm: the&drm_gpuvm
+ * @exec: the&drm_execcontext
+ * @num_fences: the amount of&dma_fencesto reserve
+ *
+ * Callsdrm_exec_prepare_obj()for the GPUVMs dummy&drm_gem_object.
+ *
+ * Using this function directly, it is the drivers responsibility to call
+ *drm_exec_init()anddrm_exec_fini()accordingly.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */include/drm/drm_gpuvm.h +6
-0+static inline int +drm_gpuvm_prepare_vm (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + unsigned int num_fences) +{ + return drm_exec_prepare_obj(exec, &gpuvm->d_obj,
num_fences);include/drm/drm_gpuvm.h +16
-0+} + +int drm_gpuvm_prepare_objects (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + unsigned int num_fences); + +int drm_gpuvm_prepare_range (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + u64 addr, u64 range, + unsigned int num_fences); + +int drm_gpuvm_exec_lock (struct drm_gpuvm_exec *vm_exec, + unsigned int num_fences, + bool interruptible); + +int drm_gpuvm_exec_lock_array(struct drm_gpuvm_exec
*vm_exec,
+ structdrm_gem_object**objs, + unsigned intnum_objs, + unsigned intnum_fences, + bool interruptible);include/drm/drm_gpuvm.h +2
-0+ +int drm_gpuvm_exec_lock_range(struct drm_gpuvm_exec
*vm_exec,
+ u64 addr, u64 range, + unsigned intnum_fences, + bool interruptible);include/drm/drm_gpuvm.h +3
-0+ +/** +* drm_gpuvm_lock() - lock all dma-resv of all assoiciated BOs
+ * @gpuvm: the&drm_gpuvm
+ *
+ * Releases all dma-resv locks of all&drm_gem_objectspreviously acquired
+ * throughdrm_gpuvm_lock()or its variants.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */include/drm/drm_gpuvm.h +11
-0+static inline void +drm_gpuvm_exec_unlock (struct drm_gpuvm_exec *vm_exec) +{ + drm_exec_fini(&vm_exec->exec); +} + +int drm_gpuvm_validate (struct drm_gpuvm *gpuvm); +void drm_gpuvm_resv_add_fence (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + struct dma_fence *fence, + enum dma_resv_usage
private_usage,
+ enumdma_resv_usageextobj_usage);include/drm/drm_gpuvm.h +12
-0+ +/** +* drm_gpuvm_exec_resv_add_fence() +* @vm_exec: the &drm_gpuvm_exec abstraction +* @fence: fence to add +* @private_usage: private dma-resv usage +* @extobj_usage: extobj dma-resv usage +* +* See drm_gpuvm_resv_add_fence(). +*/ +static inline void +drm_gpuvm_exec_resv_add_fence(struct drm_gpuvm_exec
*vm_exec,
+ structdma_fence*fence, + enumdma_resv_usageprivate_usage,
+ enumdma_resv_usageextobj_usage)
+{
+ drm_gpuvm_resv_add_fence(vm_exec->vm,&vm_exec->exec, fence,
+private_usage,extobj_usage);include/drm/drm_gpuvm.h +2
-0+} + /*** struct drm_gpuvm_bo - structure representing a
&drm_gpuvmand
*&drm_gem_objectcombinationinclude/drm/drm_gpuvm.h +3
-0@@ -398,6 +569,18 @@ struct drm_gpuvm_bo {* gpuva list. */ struct list_head gem; + + /** +* @evict: List entry to attach to the
&drm_gpuvms
+ * extobj list.
+ */
+ structlist_headextobj; +
+ /**
+ * @evict: List entry to attach to the&drm_gpuvmsevict
+ * list.
+ */
+ structlist_headevict;
} entry;
} list;
};include/drm/drm_gpuvm.h +3
-0@@ -432,6 +615,9 @@ struct drm_gpuvm_bo * drm_gpuvm_bo_find(struct drm_gpuvm *gpuvm, struct drm_gem_object *obj); +void drm_gpuvm_bo_evict (struct drm_gem_object *obj,bool evict); +void drm_gpuvm_bo_extobj_add (struct drm_gpuvm_bo *vm_bo); + /*** drm_gpuvm_bo_for_each_va() - iterator to walk over a list of
&drm_gpuva
* @va__:&drm_gpuvastructure to assign to in each
iteration stepinclude/drm/drm_gpuvm.h +5
-0@@ -837,6 +1023,17 @@ struct drm_gpuvm_ops {* used. */ int (*sm_step_unmap)(struct drm_gpuva_op *op, void *priv); + + /** +* @bo_validate: called from drm_gpuvm_validate() +* +* Drivers receive this callback for every evicted
&drm_gem_objectbeing
+ * mapped in the corresponding&drm_gpuvm. + *
+ * Typically, drivers would call their driver specific variant of
+ *ttm_bo_validate()from within this callback. + */
+ int (*bo_validate)(structdrm_gem_object*obj);
};
int drm_gpuvm_sm_map(structdrm_gpuvm*gpuvm, void *priv,
Re: [PATCH drm-misc-next v3 6/7] drm/gpuvm: generalize dma_resv/extobj handling and GEM validation
Quoted 5 lines · Christian König
Am 13.09.23 um 14:16 schrieb Danilo Krummrich:
Quoted 2 lines
As mentioned in a different mail thread, the reply is based on the assumption that we don't support anything else than GPUVM updates from the IOCTL.
I think that this assumption is incorrect.
Well, more precisely I should have said "don't support GPUVM updated from within fence signaling critical sections". And looking at the code, that doesn't seem what you're doing there.
Quoted 4 lines
Vulkan is just once specific use case, but this here should probably be able to handle other use cases as well.
Especially with HMM you get the requirement that you need to be able to invalidate GPUVM mappings without grabbing a reservation lock.
What do you mean with "invalidate GPUVM mappings" in this context? drm_gpuvm_bo_evict() should only be called from a ttm_device_funcs::move callback, we should hold the dma-resv lock there.
Quoted 2 lines
See what the eviction lock in amdgpu is doing for example.
The eviction_lock seems to protect a VM state "evicting" of whether any BO that is associated with the VM is currently evicting. At the same time amdgpu protects the eviceted list of the VM with a different lock. So this seems to be entirely unrelated. Tracking a "currently evicting" state is not part of the GPUVM implementation currently and hence nothing would change for amdgpu there.
Quoted 1,338 lines
Regards,
Christian.Quoted 1,333 lines · Thomas Hellström
Quoted 197 lines
Hi!
Quoted 178 lines · Danilo Krummrich
Quoted 168 lines · Thomas Hellström
Quoted 161 lines · Danilo Krummrich
Quoted 142 lines · Thomas Hellström
Hi, Danilo,
Quoted 139 lines · Danilo Krummrich
So far the DRM GPUVA manager offers common infrastructure to track GPU VA
allocations and mappings, generically connect GPU VA mappings to their
backing buffers and perform more complex mapping operations on the GPU VA
space.However, there are more design patterns commonly used by drivers, which
can potentially be generalized in order to make the DRM GPUVA manager
represent a basic GPU-VM implementation. In this context, this patch aims
at generalizing the following elements.1) Provide a common dma-resv for GEM objects not being used outside of
this GPU-VM.2) Provide tracking of external GEM objects (GEM objects which are
shared with other GPU-VMs).3) Provide functions to efficiently lock all GEM objects dma- resv the
GPU-VM contains mappings of.4) Provide tracking of evicted GEM objects the GPU-VM contains mappings
of, such that validation of evicted GEM objects is accelerated.5) Provide some convinience functions for common patterns.
Rather than being designed as a "framework", the target is to make all
features appear as a collection of optional helper functions, such that
drivers are free to make use of the DRM GPUVA managers basic functionality and opt-in for other features without setting any feature
flags, just by making use of the corresponding functions.Big kudos to Boris Brezillon for his help to figure out locking for drivers
updating the GPU VA space within the fence signalling path.Suggested-by Matthew Brost <[email protected]>Signed-off-by Danilo Krummrich <[email protected]>---
drivers/gpu/drm/drm_gpuvm.c| 516
++++++++++++++++++++++++++++++++++++
include/drm/drm_gpuvm.h | 197 ++++++++++++++2 files changed, 713 insertions(+)drivers/gpu/drm/drm_gpuvm.c +0
-0@@ -73,6 +73,21 @@ * &drm_gem_object list of &drm_gpuvm_bos for an existinginstance of this
* particular combination. If not existent a new instance is created and linked
* to the&drm_gem_object.
+ *
+ *&drm_gpuvm_bostructures, since unique for a given&drm_gpuvm, are also used
+ * as entry for the &drm_gpuvm's lists of external and evicted objects. Those
+ * list are maintained in order to accelerate locking of dma-resv locks and
+ * validation of evicted objects bound in a&drm_gpuvm. For instance the all
+ * &drm_gem_object's&dma_resvof a given&drm_gpuvmcan be locked by calling
+ *drm_gpuvm_exec_lock(). Once locked drivers can calldrm_gpuvm_validate()in
+ * order to validate all evicted&drm_gem_objects. It is also possible to lock
+ * additional&drm_gem_objectsby providing the corresponding parameters to
+ *drm_gpuvm_exec_lock()as well as open code the&drm_execloop while making
+ * use of helper functions such asdrm_gpuvm_prepare_range()or
+ *drm_gpuvm_prepare_objects().
+ *
+ * Every bound&drm_gem_objectis treated as external object when its&dma_resv
+ * structure is different than the &drm_gpuvm's common&dma_resvstructure.
*/
/**drivers/gpu/drm/drm_gpuvm.c +0
-0@@ -420,6 +435,20 @@ * Subsequent calls to drm_gpuvm_bo_obtain() for the same
&drm_gpuvmand
*&drm_gem_objectmust be able to observe previous creations and destructions
* of&drm_gpuvm_bosin order to keep instances unique.
+ *
+ * The &drm_gpuvm's lists for keeping track of external and evicted objects are
+ * protected against concurrent insertion / removal and iteration internally.
+ *
+ * However, drivers still need ensure to protect concurrent calls to functions
+ * iterating those lists, such asdrm_gpuvm_validate()and
+ *drm_gpuvm_prepare_objects(). Every such function contains a particular
+ * comment and lockdep checks if possible.
+ *
+ * Functions adding or removing entries from those lists, such as
+ *drm_gpuvm_bo_evict()ordrm_gpuvm_bo_extobj_add()may be called with external
+ * locks being held, e.g. in order to avoid the corresponding list to be
+ * (safely) modified while potentially being iternated by other API functions.
+ * However, this is entirely optional.
*/
/**drivers/gpu/drm/drm_gpuvm.c +5
-0@@ -632,6 +661,131 @@* } */ +/** +* get_next_vm_bo_from_list() - get the next vm_bo element +* @__gpuvm: The GPU VM +* @__list_name: The name of the list we're iterating on +* @__local_list: A pointer to the local list used to store already iterated items
+ * @__prev_vm_bo: The previous element we got fromdrm_gpuvm_get_next_cached_vm_bo()
+ *
+ * This helper is here to provide lockless list iteration.Quoted 8 lines
Quoted 5 lines
Lockless as in, the
+ * iterator releases the lock immediately after picking the first element from
+ * the list, so list insertion deletion can happen concurrently.Are the list spinlocks needed for that async state update from within the
dma-fence critical section we've discussed previously?Yes, but also for other reasons, see below.
Quoted 8 lines
Otherwise it should be sufficient to protect the lists with the gpuvm's resv
(or for the extobj list with an outer lock).If those spinlocks are still needed in some situations, perhaps could we
have an option to set them to NULL (Like IIRC the maple tree allows for)?Quoted 6 lines
The evict spinlock is needed in any case, since in
drm_gpuvm_bo_evict()we're
holding only the dma-resv lock from the BO this function gets called for. Hence,
the spinlock protects concurrentdrm_gpuvm_bo_evict()calls with different BOs.Quoted 9 lines
No. Only if you try to add external objects to the vm's evict list from
within the evict code. That's not necessary since you loop through all
external objects anyway when locking them so an "evicted" bool in thevm_bo,
protected by the bo resv would be sufficient. The extobj locking loop can
then add the bo to the evicted list.Quoted 16 lines
And
validate()can remove it while still holding all dma-resv locks, neat!
However, what if two tasks are trying to lock the VA space concurrently? What
do we do when the drm_gpuvm_bo's refcount drops to zero indrm_gpuva_unlink()?
Are we guaranteed that at this point of time thedrm_gpuvm_bois not on the
evicted list? Because otherwise we would calldrm_gpuvm_bo_destroy()with the
dma-resv lock held, which wouldn't be allowed, sincedrm_gpuvm_bo_destroy()
might drop the last reference to thedrm_gem_objectand hence we'd potentially
free the dma-resv lock while holding it, at least if it's an external object.Quoted 4 lines
Easiest way in this scheme is to think of the lists as being protected by the vm's resv lock. That means anybody calling
unlink()must also hold the vm's resv lock. (Which is OK from an UAF point of view, but perhaps not from a locking inversion POW from an async list update).This would mean that on
unlink()we'd need to hold the VM's resv lock and the corresponding GEM's resv lock (in case they're not the same anyways) because the VM's resv lock would protect the external / evicted object lists and the GEM objects resv lock protects the GEM's list ofdrm_gpuvm_bosand the drm_gpuvm_bo's list ofdrm_gpuvas.Quoted 49 lines
Quoted 38 lines
Quoted 5 lines
Quoted 5 lines
For extobjs an outer lock would be enough in case of Xe, but I really would not
like to add even more complexity just to get the spinlock out of the way in case
the driver already has an outer lock protecting this path.Quoted 15 lines
I must disagree here. These spinlocks and atomic operations are pretty
costly and as discussed earlier this type of locking was the reason (at
least according to the commit message) that made Christian drop the XArray
use indrm_execfor the same set of objects: "The locking overhead is
unecessary and measurable". IMHO the spinlock is the added complexity and a
single wide lock following the drm locking guidelines set out by Daniel and
David should really be the default choice with an opt-in for a spinlock if
needed for async and pushing out to a wq is not an option.For the external object list an outer lock would work as long as it's not the
dma-resv lock of the corresponding GEM object, since here we actually need to
remove the list entry from the external object list ondrm_gpuvm_bo_destroy().
It's just a bit weird design wise that drivers would need to take this outer
lock on:-
drm_gpuvm_bo_extobj_add()
- drm_gpuvm_bo_destroy() (and hence alsodrm_gpuvm_bo_put())
- drm_gpuva_unlink() (because it needs to calldrm_gpuvm_bo_put())
-drm_gpuvm_exec_lock()
-drm_gpuvm_exec_lock_array()
-drm_gpuvm_prepare_range()Quoted 2 lines
Given that it seems reasonable to do all the required locking internally.
From a design POW, there has been a clear direction in XE to make things similar to
mmap()/munmap(), so this outer lock, which in Xe is an rwsem, is used in a similar way as themmap_lock. It's protecting the page-table structures and vma rb tree, the userptr structures and the extobj list. Basically it's taken early in the exec IOCTL, theVM_BINDioctl, the compute rebind worker and the pagefault handler, so all of the above are just asserting that it is taken in the correct mode.Quoted 6 lines
But strictly with this scheme one could also use the vm's
dma_resvfor the extobj list since withdrm_exec, it's locked before traversing the list.The whole point of this scheme is to rely on locks that you already are supposed to be holding for various reasons and is simple to comprehend.
I don't agree that we're supposed to hold the VM's resv lock anyways for functions like
drm_gpuvm_bo_put()ordrm_gpuva_unlink(), but I'm fine using it for that purpose nevertheless.Quoted 6 lines
Quoted 5 lines
In order to at least place lockdep checks, the driver would need to supply the
corresponding lock'slockdep_map, because the GPUVM otherwise doesn't know about
the lock.Yes, that sounds reasonable. One lockdep map per list.
I'd really like to avoid that, especially now that everything got simpler. We should define the actual locks to take instead.
Quoted 45 lines
Quoted 3 lines
Out of curiosity, what is the overhead of a
spin_lock()that doesn't need to
spin?I guess it's hard to tell exactly, but it is much lower on modern x86 than what it used to be. Not sure about ARM, which is the other architecture important to us. I figure if there is little cache-line bouncing the main overhead comes from the implied barriers.
Quoted 34 lines
Quoted 25 lines
A pretty simple way that would not add much code would be
static void gpuvm_cond_spin_lock(const struct
drm_gpuvm*gpuvm,spinlock_t
*lock){
if (!gpuvm->resv_protected_lists)
spin_lock(lock);}
Quoted 12 lines
Quoted 3 lines
For such drivers, that would require anybody calling unlink to hold the vm's
resv, though.In V4 I want to go back to having a dedicated lock for the GEMs gpuva list (or
VM_BOlist to be more precise). We can't just use the dma-resv lock for that
withVM_BOabstractions, because on destruction of aVM_BOwe otherwise wouldn't
be allowed to already hold the dma-resv lock. That's the fix I was referring to
earlier.Quoted 9 lines
Yeah, I can see the need for a dedicated lock for the GEM's gpuva list, but
holding the vm's dma-resv lock across the unlink shouldn't be a problem. We
may free the object and a pointer to the vm's resv during unlink but we
don't free the vm's resv. It'd be a matter of ensuring that any calls to
unlink from *within*drm_gpuvmallows it to be held.Quoted 3 lines
Drivers calling
unlink()from the fence signaling path can't use the VM's
dma-resv lock.Quoted 6 lines
Yes, that made me a bit curious because in the current version the code required the object's
dma_resvforunlink()which can't be grabbed either from the fence signaling path. So are there any drivers actually wanting to do that? If so, they will either need to resort to the current spinlock solution or they will need to call unlink from a workqueue item.As Boris already mentioned we have the dma-resv lock by default or a driver specific GEM gpuva lock as opt-in. Now, we can get rid of the latter.
Quoted 9 lines
Quoted 3 lines
Also, what if the object is an external object? We can't use the VM's dma-resv
lock here.Why? Typically (sync) unlink is only ever called from an unbind-like operation where it should be trivial to grab the vm's resv. Or, for that matter any outer lock protecting the extobj list. Rule would be the drm_gpuvm_bo::entry::extobj and
drm_gpuvm_bo::entry::evictwould be protected by either the vm'sdma_resv(or possibly an outer lock in the case of the extobj list).Outer lock wouldn't have been working for updates in the async path, but shouldn't be relevant anymore. We could use the VM's resv for that.
Quoted 6 lines
Quoted 6 lines
And we can't have the GEM objs dma-resv lock held when calling
unlink(), sinceunlink()callsdrm_gpuvm_bo_put(), which if the refcount drops
to zero callsdrm_gpuvm_bo_destroy()anddrm_gpuvm_bo_destroy()might drop the
last reference of the GEM object.Quoted 10 lines
Yes, but this is a different problem as to what exactly protects
drm_gpuvm_bo::entry::gem. Either as you suggest an internal per bo list lock, or if we want to keep the bo'sdma_resvwe need to ensure that the caller ofdma_resv_unlock(obj->resv)actually refcounts its obj pointer, and doesn't implicitly rely on the gpuvm_bo's refcount (I know Boris didn't like that, but requiring an explicit refcount for a pointer you dereference unless you're under a lock that ensures keeping the object alive is pretty much required?) But anyway for thedrm_gpuvm_bo::entry::gemlist protection (bo resv or internal spinlock) I don't have a strong preference.We can keep the GEM objects dma-resv lock, however as mentioned above
drm_gpuva_unlink()anddrm_gpuvm_bo_put()then requires both the VM's resv lock and the GEM's resv lock in case they differ.Quoted 970 lines
Quoted 2 lines
All those problems go away with a dedicated
GEM gpuva list lock.I don't think these are real problems.
With the excepton of the eviction list "trick" where we currently have slightly different approach to collect external bos needing rebinding, we have this working fine.TBH I think pretty much the only situation where the spinlock is needed is for async updates of these lists, unless a wq item can be used for that, but it doesn't really seem like the current code allows for such updates anyway? It complicates the code a lot, adds overhead and also adds the requirement for refcounting during list traversal.
/Thomas
Quoted 955 lines
Quoted 955 lines
/Thomas
Quoted 952 lines
Quoted 952 lines
It seems that with that also the refcount could be make non- atomic.
All in the spirit of the drm locking guidelines "use big locks when
possible".
Lower level locks only when necessary for performance or locking inversion?/Thomas
Quoted 940 lines
+ *
+ * Elements popped from the original list are kept in a local list, so removal
+ * andis_emptychecks can still happen while we're iterating the list.
+ */
+#defineget_next_vm_bo_from_list(__gpuvm, __list_name, __local_list, __prev_vm_bo)\
+ ({
\
+ structdrm_gpuvm_bo
*__vm_bo; \
+
\
+ drm_gpuvm_bo_put(__prev_vm_bo); \
+
\
+ spin_lock(&(__gpuvm)-Quoted 1 line
__list_name.lock); \
+ while (!list_empty(&(__gpuvm)-
Quoted 1 line
__list_name.list)) { \
+ __vm_bo = list_first_entry(&(__gpuvm)->__list_name.list, \
+ struct drm_gpuvm_bo, \
+
list.entry.__list_name); \
+ if (drm_gpuvm_bo_get_unless_zero(__vm_bo))
{ \
+ list_move_tail(&(__vm_bo)-Quoted 1 line
list.entry.__list_name, \
+
__local_list); \
+ break;
\
+ } else
{ \
+ list_del_init(&(__vm_bo)-Quoted 1 line
list.entry.__list_name); \
+ __vm_bo =
NULL; \
+ }
\
+ }
\
+ spin_unlock(&(__gpuvm)-Quoted 1 line
__list_name.lock); \
+
\
+ __vm_bo;
\
+ })drivers/gpu/drm/drm_gpuvm.c +5
-0+ +/** +* for_each_vm_bo_in_list() - internal vm_bo list iterator +* +* This helper is here to provide lockless list iteration. Lockless as in, the
+ * iterator releases the lock immediately after picking the first element from the
+ * list, so list insertion and deletion can happen concurrently.
+ *
+ * Typical use:
+ *
+ * structdrm_gpuvm_bo*vm_bo;
+ * LIST_HEAD(my_local_list);
+ *
+ * ret = 0;
+ * drm_gpuvm_for_each_vm_bo(gpuvm, <list_name>,&my_local_list,vm_bo) {
+ * ret =do_something_with_vm_bo(..., vm_bo);
+ * if (ret)
+ * break;
+ * }
+ * drm_gpuvm_bo_put(vm_bo);
+ * drm_gpuvm_restore_vm_bo_list(gpuvm, <list_name>,&my_local_list);
+ *
+ *
+ * Only used for internal list iterations, not meant to be exposed to the outside
+ * world.
+ */
+#definefor_each_vm_bo_in_list(__gpuvm, __list_name, __local_list, __vm_bo)\
+ for (__vm_bo= get_next_vm_bo_from_list(__gpuvm, __list_name, \
+ __local_list, NULL); \
+__vm_bo;
\
+__vm_bo= get_next_vm_bo_from_list(__gpuvm, __list_name, \
+ __local_list, __vm_bo)) \drivers/gpu/drm/drm_gpuvm.c +3
-0+ +/** +* restore_vm_bo_list() - move vm_bo elements back to their original list
+ * @__gpuvm: The GPU VM
+ * @__list_name: The name of the list we're iterating on
+ * @__local_list: A pointer to the local list used to store already iterated items
+ *
+ * When we're done iterating avm_bolist, we should callrestore_vm_bo_list()
+ * to restore the original state and let new iterations take place.
+ */
+#definerestore_vm_bo_list(__gpuvm, __list_name, __local_list)\
+ do
{
\
+ /* Merge back the two lists, moving local list elements to the \
+ * head to preserve previous ordering, in case it matters. \
+
*/
\
+ spin_lock(&(__gpuvm)-Quoted 1 line
__list_name.lock); \
+ list_splice(__local_list, &(__gpuvm)-
Quoted 1 line
__list_name.list); \
+ spin_unlock(&(__gpuvm)-
Quoted 1 line
__list_name.lock); \
+ } while (0)
+/**
+ *drm_gpuvm_bo_list_add()- insert avm_bointo the given list
+ * @__vm_bo: the&drm_gpuvm_bo
+ * @__list_name: the name of the list to insert into
+ *
+ * Inserts the given @__vm_bo into the list specified by @__list_name and
+ * increases the vm_bo's reference count.
+ */
+#definedrm_gpuvm_bo_list_add(__vm_bo, __list_name)\
+ do
{
\
+ spin_lock(&(__vm_bo)->vm-Quoted 1 line
__list_name.lock); \
+ if (list_empty(&(__vm_bo)-
Quoted 1 line
list.entry.__list_name)) \
+ list_add_tail(&(__vm_bo)-
Quoted 1 line
list.entry.__list_name, \
+ &(__vm_bo)->vm-
Quoted 1 line
__list_name.list); \
+ spin_unlock(&(__vm_bo)->vm-
Quoted 1 line
__list_name.lock); \
+ } while (0)
drivers/gpu/drm/drm_gpuvm.c +3
-0+ +/** +* drm_gpuvm_bo_list_del() - remove a vm_bo from the given list
+ * @__vm_bo: the&drm_gpuvm_bo
+ * @__list_name: the name of the list to insert into
+ *
+ * Removes the given @__vm_bo from the list specified by @__list_name and
+ * decreases the vm_bo's reference count.
+ */
+#definedrm_gpuvm_bo_list_del(__vm_bo, __list_name)\
+ do
{
\
+ spin_lock(&(__vm_bo)->vm-Quoted 1 line
__list_name.lock); \
+ if (!list_empty(&(__vm_bo)-
Quoted 1 line
list.entry.__list_name)) \
+ list_del_init(&(__vm_bo)-
Quoted 1 line
list.entry.__list_name); \
+ spin_unlock(&(__vm_bo)->vm-
Quoted 1 line
__list_name.lock); \
+ } while (0)
drivers/gpu/drm/drm_gpuvm.c +4
-0+ +static int __must_check +drm_gpuvm_bo_get_unless_zero (struct drm_gpuvm_bo *vm_bo); +#define to_drm_gpuva(__node) container_of((__node), struct
drm_gpuva, rb.node)
#defineGPUVA_START(node)((node)->va.addr)drivers/gpu/drm/drm_gpuvm.c +6
-0@@ -713,6 +867,12 @@ drm_gpuvm_init(struct drm_gpuvm *gpuvm,struct drm_device *drm, gpuvm->rb.tree = RB_ROOT_CACHED; INIT_LIST_HEAD(&gpuvm->rb.list); + INIT_LIST_HEAD(&gpuvm->extobj.list); + spin_lock_init(&gpuvm->extobj.lock); + + INIT_LIST_HEAD(&gpuvm->evict.list); + spin_lock_init(&gpuvm->evict.lock); + drm_gpuva_check_overflow(start_offset, range); gpuvm->mm_start = start_offset; gpuvm->mm_range = range;@@ -754,10 +914,302 @@ drm_gpuvm_destroy(struct drm_gpuvm *gpuvm) WARN(!RB_EMPTY_ROOT(&gpuvm->rb.tree.rb_root),"GPUVA tree is not empty, potentially leaking memory.\n");
+ WARN(!list_empty(&gpuvm->extobj.list), "Extobj list should be empty.\n");
+ WARN(!list_empty(&gpuvm->evict.list), "Evict list should be empty.\n");
+
drm_gem_private_object_fini(&gpuvm->d_obj); }
EXPORT_SYMBOL_GPL(drm_gpuvm_destroy);
+/**
+ *drm_gpuvm_prepare_objects()- prepare all assoiciated BOs
+ * @gpuvm: the&drm_gpuvm
+ * @exec: the&drm_execlocking context
+ * @num_fences: the amount of&dma_fencesto reserve
+ *
+ * Callsdrm_exec_prepare_obj()for all&drm_gem_objectsthe given
+ *&drm_gpuvmcontains mappings of.
+ *
+ * Using this function directly, it is the drivers responsibility to call
+ *drm_exec_init()anddrm_exec_fini()accordingly.
+ *
+ * Note: This function is safe against concurrent insertion and removal of
+ * external objects, however it is not safe against concurrent usage itself.
+ *
+ * Drivers need to make sure to protect this case with either an outer VM lock
+ * or by callingdrm_gpuvm_prepare_vm()before this function within the
+ *drm_exec_until_all_locked()loop, such that the GPUVM's dma-resv lock ensures
+ * mutual exclusion.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +10
-0+int +drm_gpuvm_prepare_objects (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + unsigned int num_fences) +{ + struct drm_gpuvm_bo *vm_bo; + LIST_HEAD(extobjs); + int ret = 0; + + for_each_vm_bo_in_list(gpuvm, extobj, &extobjs,
vm_bo) {
+ ret =drm_exec_prepare_obj(exec, vm_bo->obj, num_fences);
+ if (ret)
+ break;
+ }
+ /* Drop ref in case we break out of the loop. */ + drm_gpuvm_bo_put(vm_bo);
+ restore_vm_bo_list(gpuvm, extobj, &extobjs);drivers/gpu/drm/drm_gpuvm.c +7
-0+ + return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_prepare_objects); + +/** +* drm_gpuvm_prepare_range() - prepare all BOs mapped within a given range
+ * @gpuvm: the&drm_gpuvm
+ * @exec: the&drm_execlocking context
+ * @addr: the start address within the VA space
+ * @range: the range to iterate within the VA space
+ * @num_fences: the amount of&dma_fencesto reserve
+ *
+ * Callsdrm_exec_prepare_obj()for all&drm_gem_objectsmapped between @addr
+ * and @addr + @range.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +2
-0+int +drm_gpuvm_prepare_range(struct drm_gpuvm *gpuvm, struct
drm_exec*exec,
+ u64 addr, u64 range, unsigned intnum_fences)drivers/gpu/drm/drm_gpuvm.c +9
-0+{ + struct drm_gpuva *va; + u64 end = addr + range; + int ret; + + drm_gpuvm_for_each_va_range(va, gpuvm, addr, end) { + struct drm_gem_object *obj = va->gem.obj; + + ret = drm_exec_prepare_obj(exec, obj,
num_fences);
+ if (ret)
+ return ret;
+ }drivers/gpu/drm/drm_gpuvm.c +7
-0+ + return 0; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_prepare_range); + +/** +* drm_gpuvm_exec_lock() - lock all dma-resv of all assoiciated BOs
+ * @vm_exec: the&drm_gpuvm_execabstraction
+ * @num_fences: the amount of&dma_fencesto reserve
+ * @interruptible: sleep interruptible if waiting
+ *
+ * Acquires all dma-resv locks of all&drm_gem_objectsthe given
+ *&drm_gpuvmcontains mappings of.
+ *
+ * Addionally, when calling this function with structdrm_gpuvm_exec::extra
+ * being set the driver receives the given @fn callback to lock additional
+ * dma-resv in the context of the&drm_gpuvm_execinstance. Typically, drivers
+ * would calldrm_exec_prepare_obj()from within this callback.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +11
-0+int +drm_gpuvm_exec_lock (struct drm_gpuvm_exec *vm_exec, + unsigned int num_fences, + bool interruptible) +{ + struct drm_gpuvm *gpuvm = vm_exec->vm; + struct drm_exec *exec = &vm_exec->exec; + uint32_t flags; + int ret; + + flags = interruptible ? DRM_EXEC_INTERRUPTIBLE_WAIT :0 |
+ DRM_EXEC_IGNORE_DUPLICATES;drivers/gpu/drm/drm_gpuvm.c +5
-0+ + drm_exec_init(exec, flags); + + drm_exec_until_all_locked(exec) { + ret = drm_gpuvm_prepare_vm(gpuvm, exec,
num_fences);
+ drm_exec_retry_on_contention(exec); + if (ret)
+ goto err;
+
+ ret =drm_gpuvm_prepare_objects(gpuvm, exec, num_fences);
+ drm_exec_retry_on_contention(exec); + if (ret)
+ goto err;
+
+ if (vm_exec->extra.fn) {
+ ret =vm_exec->extra.fn(vm_exec, num_fences);
+ drm_exec_retry_on_contention(exec); + if (ret)
+ goto err;
+ }
+ }drivers/gpu/drm/drm_gpuvm.c +11
-0+ + return 0; + +err: + drm_exec_fini(exec); + return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock); + +static int +fn_lock_array(struct drm_gpuvm_exec *vm_exec, unsigned int
num_fences)drivers/gpu/drm/drm_gpuvm.c +7
-0+{ + struct { + struct drm_gem_object **objs; + unsigned int num_objs; + } *args = vm_exec->extra.priv; + + return drm_exec_prepare_array(&vm_exec->exec, args-Quoted 1 line
objs,
+
args->num_objs,num_fences);drivers/gpu/drm/drm_gpuvm.c +4
-0+} + +/** +* drm_gpuvm_exec_lock_array() - lock all dma-resv of all assoiciated BOs
+ * @vm_exec: the&drm_gpuvm_execabstraction
+ * @objs: additional&drm_gem_objectsto lock
+ * @num_objs: the number of additional&drm_gem_objectsto lock
+ * @num_fences: the amount of&dma_fencesto reserve
+ * @interruptible: sleep interruptible if waiting
+ *
+ * Acquires all dma-resv locks of all&drm_gem_objectsthe given&drm_gpuvm
+ * contains mappings of, plus the ones given through @objs.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +19
-0+int +drm_gpuvm_exec_lock_array (struct drm_gpuvm_exec *vm_exec, + struct drm_gem_object **objs, + unsigned int num_objs, + unsigned int num_fences, + bool interruptible) +{ + struct { + struct drm_gem_object **objs; + unsigned int num_objs; + } args; + + args.objs = objs; + args.num_objs = num_objs; + + vm_exec->extra.fn = fn_lock_array; + vm_exec->extra.priv = &args; + + return drm_gpuvm_exec_lock(vm_exec, num_fences,interruptible);
drivers/gpu/drm/drm_gpuvm.c +5
-0+} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock_array); + +/** +* drm_gpuvm_exec_lock_range() - prepare all BOs mapped within a given range
+ * @vm_exec: the&drm_gpuvm_execabstraction
+ * @addr: the start address within the VA space
+ * @range: the range to iterate within the VA space
+ * @num_fences: the amount of&dma_fencesto reserve
+ * @interruptible: sleep interruptible if waiting
+ *
+ * Acquires all dma-resv locks of all&drm_gem_objectsmapped between @addr and
+ * @addr + @range.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +12
-0+int +drm_gpuvm_exec_lock_range (struct drm_gpuvm_exec *vm_exec, + u64 addr, u64 range, + unsigned int num_fences, + bool interruptible) +{ + struct drm_gpuvm *gpuvm = vm_exec->vm; + struct drm_exec *exec = &vm_exec->exec; + uint32_t flags; + int ret; + + flags = interruptible ? DRM_EXEC_INTERRUPTIBLE_WAIT :0 |
+ DRM_EXEC_IGNORE_DUPLICATES;drivers/gpu/drm/drm_gpuvm.c +5
-0+ + drm_exec_init(exec, flags); + + drm_exec_until_all_locked(exec) { + ret = drm_gpuvm_prepare_range(gpuvm, exec,addr, range,
+num_fences); + drm_exec_retry_on_contention(exec); + if (ret)
+ goto err;
+ }drivers/gpu/drm/drm_gpuvm.c +14
-0+ + return ret; + +err: + drm_exec_fini(exec); + return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock_range); + +/** +* drm_gpuvm_validate() - validate all BOs marked as evicted +* @gpuvm: the &drm_gpuvm to validate evicted BOs +* +* Calls the &drm_gpuvm_ops.bo_validate callback for all evicted buffer
+ * objects being mapped in the given&drm_gpuvm.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +27
-0+int +drm_gpuvm_validate (struct drm_gpuvm *gpuvm) +{ + const struct drm_gpuvm_ops *ops = gpuvm->ops; + struct drm_gpuvm_bo *vm_bo; + LIST_HEAD(evict); + int ret = 0; + + if (unlikely(!ops || !ops->bo_validate)) + return -ENOTSUPP; + + for_each_vm_bo_in_list(gpuvm, evict, &evict, vm_bo) { + dma_resv_assert_held(vm_bo->obj->resv); + ret = ops->bo_validate(vm_bo->obj); + if (ret) + break; + } + /* Drop ref in case we break out of the loop. */ + drm_gpuvm_bo_put(vm_bo); + restore_vm_bo_list(gpuvm, evict, &evict); + + return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_validate); + +/** +* drm_gpuvm_resv_add_fence - add fence to private and all extobj
+ * dma-resv
+ * @gpuvm: the&drm_gpuvmto add a fence to
+ * @exec: the&drm_execlocking context
+ * @fence: fence to add
+ * @private_usage: private dma-resv usage
+ * @extobj_usage: extobj dma-resv usage
+ */drivers/gpu/drm/drm_gpuvm.c +14
-0+void +drm_gpuvm_resv_add_fence (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + struct dma_fence *fence, + enum dma_resv_usage private_usage, + enum dma_resv_usage extobj_usage) +{ + struct drm_gem_object *obj; + unsigned long index; + + drm_exec_for_each_locked_object(exec, index, obj) { + dma_resv_assert_held(obj->resv); + dma_resv_add_fence(obj->resv, fence, + drm_gpuvm_is_extobj(gpuvm,obj) ?
+private_usage:extobj_usage);
+ }drivers/gpu/drm/drm_gpuvm.c +3
-0+} +EXPORT_SYMBOL_GPL (drm_gpuvm_resv_add_fence); +/**
*drm_gpuvm_bo_create()- create a new instance of structdrm_gpuvm_bo
* @gpuvm: The&drm_gpuvmthe @obj is mapped in.drivers/gpu/drm/drm_gpuvm.c +13
-0@@ -790,6 +1242,9 @@ drm_gpuvm_bo_create(struct drm_gpuvm *gpuvm, INIT_LIST_HEAD(&vm_bo->list.gpuva); INIT_LIST_HEAD(&vm_bo->list.entry.gem); + INIT_LIST_HEAD(&vm_bo->list.entry.extobj); + INIT_LIST_HEAD(&vm_bo->list.entry.evict); + drm_gem_object_get(obj); return vm_bo;@@ -807,6 +1262,14 @@ drm_gpuvm_bo_destroy(struct kref *kref) drm_gem_gpuva_assert_lock_held(vm_bo->obj); + spin_lock(&gpuvm->extobj.lock); + list_del(&vm_bo->list.entry.extobj); + spin_unlock(&gpuvm->extobj.lock); + + spin_lock(&gpuvm->evict.lock); + list_del(&vm_bo->list.entry.evict); + spin_unlock(&gpuvm->evict.lock); + list_del(&vm_bo->list.entry.gem); drm_gem_object_put(obj);@@ -822,6 +1285,11 @@ drm_gpuvm_bo_destroy(struct kref *kref) * @vm_bo: the &drm_gpuvm_bo to release the reference of * * This releases a reference to @vm_bo. + * + * If the reference count drops to zero, the &gpuvm_bo isdestroyed, which
+ * includes removing it from the GEMs gpuva list. Hence, if a call to this
+ * function can potentially let the reference count to zero the caller must
+ * hold the dma-resv or driver specific GEM gpuva lock. */
void
drm_gpuvm_bo_put(struct drm_gpuvm_bo *vm_bo)drivers/gpu/drm/drm_gpuvm.c +6
-0@@ -831,6 +1299,12 @@ drm_gpuvm_bo_put(struct drm_gpuvm_bo *vm_bo) } EXPORT_SYMBOL_GPL(drm_gpuvm_bo_put); +static int __must_check +drm_gpuvm_bo_get_unless_zero (struct drm_gpuvm_bo *vm_bo) +{ + return kref_get_unless_zero(&vm_bo->kref); +} +static struct
drm_gpuvm_bo*__drm_gpuvm_bo_find(struct drm_gpuvm *gpuvm, struct drm_gem_object *obj)drivers/gpu/drm/drm_gpuvm.c +2
-0@@ -938,6 +1412,48 @@ drm_gpuvm_bo_obtain_prealloc(struct drm_gpuvm_bo *__vm_bo) } EXPORT_SYMBOL_GPL(drm_gpuvm_bo_obtain_prealloc); +/** +* drm_gpuvm_bo_extobj_add() - adds the &drm_gpuvm_bo to its &drm_gpuvm's
+ * extobj list
+ * @vm_bo: The&drm_gpuvm_boto add to its &drm_gpuvm's the extobj list.
+ *
+ * Adds the given @vm_bo to its &drm_gpuvm's extobj list if not on the list
+ * already and if the corresponding&drm_gem_objectis an external object,
+ * actually.
+ */drivers/gpu/drm/drm_gpuvm.c +12
-0+void +drm_gpuvm_bo_extobj_add (struct drm_gpuvm_bo *vm_bo) +{ + struct drm_gpuvm *gpuvm = vm_bo->vm; + + if (drm_gpuvm_is_extobj(gpuvm, vm_bo->obj)) + drm_gpuvm_bo_list_add(vm_bo, extobj); +} +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_extobj_add); + +/** +* drm_gpuvm_bo_evict() - add / remove a &drm_gem_object to / from a
+ *&drm_gpuvmsevicted list
+ * @obj: the&drm_gem_objectto add or remove
+ * @evict: indicates whether the object is evicted
+ *
+ * Adds a&drm_gem_objectto or removes it from all&drm_gpuvmsevicted
+ * list containing a mapping of this&drm_gem_object.
+ */drivers/gpu/drm/drm_gpuvm.c +14
-0+void +drm_gpuvm_bo_evict (struct drm_gem_object *obj,bool evict) +{ + struct drm_gpuvm_bo *vm_bo; + + drm_gem_for_each_gpuvm_bo(vm_bo, obj) { + if (evict) + drm_gpuvm_bo_list_add(vm_bo, evict); + else + drm_gpuvm_bo_list_del(vm_bo, evict); + } +} +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_evict); +static int
__drm_gpuva_insert(structdrm_gpuvm*gpuvm,
structdrm_gpuva*va)include/drm/drm_gpuvm.h +9
-0@@ -26,10 +26,12 @@*/ #include <linux/list.h> +#include <linux/dma-resv.h> #include <linux/rbtree.h> #include <linux/types.h> #include <drm/drm_gem.h> +#include <drm/drm_exec.h> struct drm_gpuvm; struct drm_gpuvm_bo;@@ -259,6 +261,38 @@ struct drm_gpuvm {* space */ struct dma_resv *resv; + + /** + * @extobj: structure holding the extobj list + */ + struct { + /** +* @list: &list_head storing &drm_gpuvm_bos serving as
+ * external object
+ */
+ structlist_headlist;include/drm/drm_gpuvm.h +9
-0+ + /** + * @lock: spinlock to protect the extobj list + */ + spinlock_t lock; + } extobj; + + /** +* @evict: structure holding the evict list and evict list lock
+ */
+ struct {
+ /**
+ * @list:&list_headstoring&drm_gpuvm_boscurrently being
+ * evicted
+ */
+ structlist_headlist;
+
+ /**
+ * @lock: spinlock to protect the evict list + */
+ spinlock_t lock;
+ } evict;
};
void drm_gpuvm_init(structdrm_gpuvm*gpuvm, structdrm_device*drm,include/drm/drm_gpuvm.h +2
-0@@ -268,6 +302,21 @@ void drm_gpuvm_init(struct drm_gpuvm *gpuvm, struct drm_device *drm, const struct drm_gpuvm_ops *ops); void drm_gpuvm_destroy(struct drm_gpuvm *gpuvm); +/** +* drm_gpuvm_is_extobj() - indicates whether the given
&drm_gem_objectis an
+ * external object
+ * @gpuvm: the&drm_gpuvmto check
+ * @obj: the&drm_gem_objectto check
+ *
+ * Returns: true if the&drm_gem_object&dma_resvdiffers from the
+ *&drm_gpuvms&dma_resv, false otherwise
+ */
+static inline bool drm_gpuvm_is_extobj(structdrm_gpuvm*gpuvm,
+ structdrm_gem_object*obj)include/drm/drm_gpuvm.h +4
-0+{ + return obj && obj->resv != gpuvm->resv; +} +static inline struct
drm_gpuva*
__drm_gpuva_next(struct drm_gpuva *va)
{include/drm/drm_gpuvm.h +0
-0@@ -346,6 +395,128 @@ __drm_gpuva_next(struct drm_gpuva *va) #define drm_gpuvm_for_each_va_safe(va__, next__, gpuvm__) \list_for_each_entry_safe(va__,
next__, &(gpuvm__)-Quoted 1 line
rb.list, rb.entry)
+/**
+ * structdrm_gpuvm_exec-&drm_gpuvmabstraction of&drm_exec
+ *
+ * This structure should be created on the stack as&drm_execshould be.
+ *
+ * Optionally, @extra can be set in order to lock additional&drm_gem_objects.
+ */include/drm/drm_gpuvm.h +13
-0+struct drm_gpuvm_exec { + /** + * @exec: the &drm_exec structure + */ + struct drm_exec exec; + + /** + * @vm: the &drm_gpuvm to lock its DMA reservations + */ + struct drm_gpuvm *vm; + + /** + * @extra: Callback and corresponding private datafor the driver to
+ * lock arbitrary additional&drm_gem_objects. + */
+ struct {
+ /**
+ * @fn: The driver callback to lock additional&drm_gem_objects.
+ */
+ int (*fn)(structdrm_gpuvm_exec*vm_exec, + unsigned intnum_fences); +
+ /**
+ * @priv: driver private data for the @fn callback
+ */
+ void *priv;
+ } extra;include/drm/drm_gpuvm.h +4
-0+}; + +/** +* drm_gpuvm_prepare_vm() - prepare the GPUVMs common dma- resv
+ * @gpuvm: the&drm_gpuvm
+ * @exec: the&drm_execcontext
+ * @num_fences: the amount of&dma_fencesto reserve
+ *
+ * Callsdrm_exec_prepare_obj()for the GPUVMs dummy&drm_gem_object.
+ *
+ * Using this function directly, it is the drivers responsibility to call
+ *drm_exec_init()anddrm_exec_fini()accordingly.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */include/drm/drm_gpuvm.h +6
-0+static inline int +drm_gpuvm_prepare_vm (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + unsigned int num_fences) +{ + return drm_exec_prepare_obj(exec, &gpuvm->d_obj,
num_fences);include/drm/drm_gpuvm.h +16
-0+} + +int drm_gpuvm_prepare_objects (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + unsigned int num_fences); + +int drm_gpuvm_prepare_range (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + u64 addr, u64 range, + unsigned int num_fences); + +int drm_gpuvm_exec_lock (struct drm_gpuvm_exec *vm_exec, + unsigned int num_fences, + bool interruptible); + +int drm_gpuvm_exec_lock_array(struct drm_gpuvm_exec
*vm_exec,
+ structdrm_gem_object**objs, + unsigned intnum_objs, + unsigned intnum_fences, + bool interruptible);include/drm/drm_gpuvm.h +2
-0+ +int drm_gpuvm_exec_lock_range(struct drm_gpuvm_exec
*vm_exec,
+ u64 addr, u64 range, + unsigned intnum_fences, + bool interruptible);include/drm/drm_gpuvm.h +3
-0+ +/** +* drm_gpuvm_lock() - lock all dma-resv of all assoiciated BOs
+ * @gpuvm: the&drm_gpuvm
+ *
+ * Releases all dma-resv locks of all&drm_gem_objectspreviously acquired
+ * throughdrm_gpuvm_lock()or its variants.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */include/drm/drm_gpuvm.h +11
-0+static inline void +drm_gpuvm_exec_unlock (struct drm_gpuvm_exec *vm_exec) +{ + drm_exec_fini(&vm_exec->exec); +} + +int drm_gpuvm_validate (struct drm_gpuvm *gpuvm); +void drm_gpuvm_resv_add_fence (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + struct dma_fence *fence, + enum dma_resv_usage
private_usage,
+ enumdma_resv_usageextobj_usage);include/drm/drm_gpuvm.h +12
-0+ +/** +* drm_gpuvm_exec_resv_add_fence() +* @vm_exec: the &drm_gpuvm_exec abstraction +* @fence: fence to add +* @private_usage: private dma-resv usage +* @extobj_usage: extobj dma-resv usage +* +* See drm_gpuvm_resv_add_fence(). +*/ +static inline void +drm_gpuvm_exec_resv_add_fence(struct drm_gpuvm_exec
*vm_exec,
+ structdma_fence*fence, + enumdma_resv_usageprivate_usage,
+ enumdma_resv_usageextobj_usage)
+{
+ drm_gpuvm_resv_add_fence(vm_exec->vm,&vm_exec->exec, fence,
+private_usage,extobj_usage);include/drm/drm_gpuvm.h +2
-0+} +/**
* structdrm_gpuvm_bo- structure representing a&drm_gpuvmand
*&drm_gem_objectcombinationinclude/drm/drm_gpuvm.h +3
-0@@ -398,6 +569,18 @@ struct drm_gpuvm_bo {* gpuva list. */ struct list_head gem; + + /** +* @evict: List entry to attach to the
&drm_gpuvms
+ * extobj list.
+ */
+ structlist_headextobj; +
+ /**
+ * @evict: List entry to attach to the&drm_gpuvmsevict
+ * list.
+ */
+ structlist_headevict;
} entry;
} list;
};include/drm/drm_gpuvm.h +3
-0@@ -432,6 +615,9 @@ struct drm_gpuvm_bo * drm_gpuvm_bo_find(struct drm_gpuvm *gpuvm, struct drm_gem_object *obj); +void drm_gpuvm_bo_evict (struct drm_gem_object *obj,bool evict); +void drm_gpuvm_bo_extobj_add (struct drm_gpuvm_bo *vm_bo); + /*** drm_gpuvm_bo_for_each_va() - iterator to walk over a list of
&drm_gpuva
* @va__:&drm_gpuvastructure to assign to in each iteration stepinclude/drm/drm_gpuvm.h +0
-0@@ -837,6 +1023,17 @@ struct drm_gpuvm_ops {* used. */ int (*sm_step_unmap)(struct
drm_gpuva_op*op, void *priv);
+
+ /**
+ * @bo_validate: called fromdrm_gpuvm_validate()+ *
+ * Drivers receive this callback for every evicted&drm_gem_objectbeing
+ * mapped in the corresponding&drm_gpuvm. + *
+ * Typically, drivers would call their driver specific variant of
+ *ttm_bo_validate()from within this callback. + */
+ int (*bo_validate)(structdrm_gem_object*obj); };
int drm_gpuvm_sm_map(structdrm_gpuvm*gpuvm, void *priv,
Re: [PATCH drm-misc-next v3 6/7] drm/gpuvm: generalize dma_resv/extobj handling and GEM validation
Am 13.09.23 um 17:15 schrieb Danilo Krummrich:
Quoted 12 lines · Christian König
Quoted 5 lines
Am 13.09.23 um 14:16 schrieb Danilo Krummrich:
Quoted 2 lines
As mentioned in a different mail thread, the reply is based on the assumption that we don't support anything else than GPUVM updates from the IOCTL.
I think that this assumption is incorrect.
Well, more precisely I should have said "don't support GPUVM updated from within fence signaling critical sections". And looking at the code, that doesn't seem what you're doing there.
Quoted 1 line
Quoted 7 lines
Quoted 3 lines
Vulkan is just once specific use case, but this here should probably be able to handle other use cases as well.
Especially with HMM you get the requirement that you need to be able to invalidate GPUVM mappings without grabbing a reservation lock.
What do you mean with "invalidate GPUVM mappings" in this context?
drm_gpuvm_bo_evict()should only be called from a ttm_device_funcs::move callback, we should hold the dma-resv lock there.
Well the question is which dma-resv lock do we hold?
In the move callback we only hold the dma-resv lock of the BO which is moved, but when that is a shared BO then that's not the same as the one for the VM.
Quoted 9 lines
Quoted 3 lines
See what the eviction lock in amdgpu is doing for example.
The
eviction_lockseems to protect a VM state "evicting" of whether any BO that is associated with the VM is currently evicting. At the same time amdgpu protects the eviceted list of the VM with a different lock. So this seems to be entirely unrelated. Tracking a "currently evicting" state is not part of the GPUVM implementation currently and hence nothing would change for amdgpu there.
Sorry for the confusion we use different terminology in amdgpu.
The eviction lock and evicted state is for the VM page tables, e.g. if the whole VM is currently not used and swapped out or even de-allocated.
This is necessary because we have cases where we need to access the VM data without holding the dma-resv lock of this VM. Especially figuring out which parts of an address space contain mappings and which doesn't.
This is a requirement which comes with HMM handling, you won't see this with Vulkan (or OpenGL, VAAPI etc..).
The invalidation lock on the other hand is what in this discussion is called eviction lock. This one is needed because what I wrote above, during the move callback only the dma-resv of the BO which is moved is locked, but not necessarily the dma-resv of the VM.
Regards,
Christian.
Quoted 1,340 lines
Quoted 1,339 lines
Regards,
Christian.Quoted 1,333 lines · Thomas Hellström
Quoted 197 lines
Hi!
Quoted 178 lines · Danilo Krummrich
Quoted 168 lines · Thomas Hellström
Quoted 161 lines · Danilo Krummrich
Quoted 142 lines · Thomas Hellström
Hi, Danilo,
Quoted 139 lines · Danilo Krummrich
So far the DRM GPUVA manager offers common infrastructure to track GPU VA
allocations and mappings, generically connect GPU VA mappings to their
backing buffers and perform more complex mapping operations on the GPU VA
space.However, there are more design patterns commonly used by drivers, which
can potentially be generalized in order to make the DRM GPUVA manager
represent a basic GPU-VM implementation. In this context, this patch aims
at generalizing the following elements.1) Provide a common dma-resv for GEM objects not being used outside of
this GPU-VM.2) Provide tracking of external GEM objects (GEM objects which are
shared with other GPU-VMs).3) Provide functions to efficiently lock all GEM objects dma- resv the
GPU-VM contains mappings of.4) Provide tracking of evicted GEM objects the GPU-VM contains mappings
of, such that validation of evicted GEM objects is accelerated.5) Provide some convinience functions for common patterns.
Rather than being designed as a "framework", the target is to make all
features appear as a collection of optional helper functions, such that
drivers are free to make use of the DRM GPUVA managers basic functionality and opt-in for other features without setting any feature
flags, just by making use of the corresponding functions.Big kudos to Boris Brezillon for his help to figure out locking for drivers
updating the GPU VA space within the fence signalling path.Suggested-by Matthew Brost <[email protected]>Signed-off-by Danilo Krummrich <[email protected]>---
drivers/gpu/drm/drm_gpuvm.c| 516
++++++++++++++++++++++++++++++++++++
include/drm/drm_gpuvm.h | 197 ++++++++++++++2 files changed, 713 insertions(+)drivers/gpu/drm/drm_gpuvm.c +0
-0@@ -73,6 +73,21 @@ * &drm_gem_object list of &drm_gpuvm_bos for an existinginstance of this
* particular combination. If not existent a new instance is created and linked
* to the&drm_gem_object.
+ *
+ *&drm_gpuvm_bostructures, since unique for a given&drm_gpuvm, are also used
+ * as entry for the &drm_gpuvm's lists of external and evicted objects. Those
+ * list are maintained in order to accelerate locking of dma-resv locks and
+ * validation of evicted objects bound in a&drm_gpuvm. For instance the all
+ * &drm_gem_object's&dma_resvof a given&drm_gpuvmcan be locked by calling
+ *drm_gpuvm_exec_lock(). Once locked drivers can calldrm_gpuvm_validate()in
+ * order to validate all evicted&drm_gem_objects. It is also possible to lock
+ * additional&drm_gem_objectsby providing the corresponding parameters to
+ *drm_gpuvm_exec_lock()as well as open code the&drm_execloop while making
+ * use of helper functions such asdrm_gpuvm_prepare_range()or
+ *drm_gpuvm_prepare_objects().
+ *
+ * Every bound&drm_gem_objectis treated as external object when its&dma_resv
+ * structure is different than the &drm_gpuvm's common&dma_resvstructure.
*/
/**drivers/gpu/drm/drm_gpuvm.c +0
-0@@ -420,6 +435,20 @@ * Subsequent calls to drm_gpuvm_bo_obtain() for the same
&drm_gpuvmand
*&drm_gem_objectmust be able to observe previous creations and destructions
* of&drm_gpuvm_bosin order to keep instances unique.
+ *
+ * The &drm_gpuvm's lists for keeping track of external and evicted objects are
+ * protected against concurrent insertion / removal and iteration internally.
+ *
+ * However, drivers still need ensure to protect concurrent calls to functions
+ * iterating those lists, such asdrm_gpuvm_validate()and
+ *drm_gpuvm_prepare_objects(). Every such function contains a particular
+ * comment and lockdep checks if possible.
+ *
+ * Functions adding or removing entries from those lists, such as
+ *drm_gpuvm_bo_evict()ordrm_gpuvm_bo_extobj_add()may be called with external
+ * locks being held, e.g. in order to avoid the corresponding list to be
+ * (safely) modified while potentially being iternated by other API functions.
+ * However, this is entirely optional.
*/
/**drivers/gpu/drm/drm_gpuvm.c +5
-0@@ -632,6 +661,131 @@* } */ +/** +* get_next_vm_bo_from_list() - get the next vm_bo element +* @__gpuvm: The GPU VM +* @__list_name: The name of the list we're iterating on +* @__local_list: A pointer to the local list used to store already iterated items
+ * @__prev_vm_bo: The previous element we got fromdrm_gpuvm_get_next_cached_vm_bo()
+ *
+ * This helper is here to provide lockless list iteration.Quoted 8 lines
Quoted 5 lines
Lockless as in, the
+ * iterator releases the lock immediately after picking the first element from
+ * the list, so list insertion deletion can happen concurrently.Are the list spinlocks needed for that async state update from within the
dma-fence critical section we've discussed previously?Yes, but also for other reasons, see below.
Quoted 8 lines
Otherwise it should be sufficient to protect the lists with the gpuvm's resv
(or for the extobj list with an outer lock).If those spinlocks are still needed in some situations, perhaps could we
have an option to set them to NULL (Like IIRC the maple tree allows for)?Quoted 6 lines
The evict spinlock is needed in any case, since in
drm_gpuvm_bo_evict()we're
holding only the dma-resv lock from the BO this function gets called for. Hence,
the spinlock protects concurrentdrm_gpuvm_bo_evict()calls with different BOs.Quoted 9 lines
No. Only if you try to add external objects to the vm's evict list from
within the evict code. That's not necessary since you loop through all
external objects anyway when locking them so an "evicted" bool in thevm_bo,
protected by the bo resv would be sufficient. The extobj locking loop can
then add the bo to the evicted list.Quoted 16 lines
And
validate()can remove it while still holding all dma-resv locks, neat!
However, what if two tasks are trying to lock the VA space concurrently? What
do we do when the drm_gpuvm_bo's refcount drops to zero indrm_gpuva_unlink()?
Are we guaranteed that at this point of time thedrm_gpuvm_bois not on the
evicted list? Because otherwise we would calldrm_gpuvm_bo_destroy()with the
dma-resv lock held, which wouldn't be allowed, sincedrm_gpuvm_bo_destroy()
might drop the last reference to thedrm_gem_objectand hence we'd potentially
free the dma-resv lock while holding it, at least if it's an external object.Quoted 4 lines
Easiest way in this scheme is to think of the lists as being protected by the vm's resv lock. That means anybody calling
unlink()must also hold the vm's resv lock. (Which is OK from an UAF point of view, but perhaps not from a locking inversion POW from an async list update).This would mean that on
unlink()we'd need to hold the VM's resv lock and the corresponding GEM's resv lock (in case they're not the same anyways) because the VM's resv lock would protect the external / evicted object lists and the GEM objects resv lock protects the GEM's list ofdrm_gpuvm_bosand the drm_gpuvm_bo's list ofdrm_gpuvas.Quoted 49 lines
Quoted 38 lines
Quoted 5 lines
Quoted 5 lines
For extobjs an outer lock would be enough in case of Xe, but I really would not
like to add even more complexity just to get the spinlock out of the way in case
the driver already has an outer lock protecting this path.Quoted 15 lines
I must disagree here. These spinlocks and atomic operations are pretty
costly and as discussed earlier this type of locking was the reason (at
least according to the commit message) that made Christian drop the XArray
use indrm_execfor the same set of objects: "The locking overhead is
unecessary and measurable". IMHO the spinlock is the added complexity and a
single wide lock following the drm locking guidelines set out by Daniel and
David should really be the default choice with an opt-in for a spinlock if
needed for async and pushing out to a wq is not an option.For the external object list an outer lock would work as long as it's not the
dma-resv lock of the corresponding GEM object, since here we actually need to
remove the list entry from the external object list ondrm_gpuvm_bo_destroy().
It's just a bit weird design wise that drivers would need to take this outer
lock on:-
drm_gpuvm_bo_extobj_add()
- drm_gpuvm_bo_destroy() (and hence alsodrm_gpuvm_bo_put())
- drm_gpuva_unlink() (because it needs to calldrm_gpuvm_bo_put())
-drm_gpuvm_exec_lock()
-drm_gpuvm_exec_lock_array()
-drm_gpuvm_prepare_range()Quoted 2 lines
Given that it seems reasonable to do all the required locking internally.
From a design POW, there has been a clear direction in XE to make things similar to
mmap()/munmap(), so this outer lock, which in Xe is an rwsem, is used in a similar way as themmap_lock. It's protecting the page-table structures and vma rb tree, the userptr structures and the extobj list. Basically it's taken early in the exec IOCTL, theVM_BINDioctl, the compute rebind worker and the pagefault handler, so all of the above are just asserting that it is taken in the correct mode.Quoted 6 lines
But strictly with this scheme one could also use the vm's
dma_resvfor the extobj list since withdrm_exec, it's locked before traversing the list.The whole point of this scheme is to rely on locks that you already are supposed to be holding for various reasons and is simple to comprehend.
I don't agree that we're supposed to hold the VM's resv lock anyways for functions like
drm_gpuvm_bo_put()ordrm_gpuva_unlink(), but I'm fine using it for that purpose nevertheless.Quoted 6 lines
Quoted 5 lines
In order to at least place lockdep checks, the driver would need to supply the
corresponding lock'slockdep_map, because the GPUVM otherwise doesn't know about
the lock.Yes, that sounds reasonable. One lockdep map per list.
I'd really like to avoid that, especially now that everything got simpler. We should define the actual locks to take instead.
Quoted 45 lines
Quoted 3 lines
Out of curiosity, what is the overhead of a
spin_lock()that doesn't need to
spin?I guess it's hard to tell exactly, but it is much lower on modern x86 than what it used to be. Not sure about ARM, which is the other architecture important to us. I figure if there is little cache-line bouncing the main overhead comes from the implied barriers.
Quoted 34 lines
Quoted 25 lines
A pretty simple way that would not add much code would be
static void gpuvm_cond_spin_lock(const struct
drm_gpuvm*gpuvm,spinlock_t
*lock){
if (!gpuvm->resv_protected_lists)
spin_lock(lock);}
Quoted 12 lines
Quoted 3 lines
For such drivers, that would require anybody calling unlink to hold the vm's
resv, though.In V4 I want to go back to having a dedicated lock for the GEMs gpuva list (or
VM_BOlist to be more precise). We can't just use the dma-resv lock for that
withVM_BOabstractions, because on destruction of aVM_BOwe otherwise wouldn't
be allowed to already hold the dma-resv lock. That's the fix I was referring to
earlier.Quoted 9 lines
Yeah, I can see the need for a dedicated lock for the GEM's gpuva list, but
holding the vm's dma-resv lock across the unlink shouldn't be a problem. We
may free the object and a pointer to the vm's resv during unlink but we
don't free the vm's resv. It'd be a matter of ensuring that any calls to
unlink from *within*drm_gpuvmallows it to be held.Quoted 3 lines
Drivers calling
unlink()from the fence signaling path can't use the VM's
dma-resv lock.Quoted 6 lines
Yes, that made me a bit curious because in the current version the code required the object's
dma_resvforunlink()which can't be grabbed either from the fence signaling path. So are there any drivers actually wanting to do that? If so, they will either need to resort to the current spinlock solution or they will need to call unlink from a workqueue item.As Boris already mentioned we have the dma-resv lock by default or a driver specific GEM gpuva lock as opt-in. Now, we can get rid of the latter.
Quoted 9 lines
Quoted 3 lines
Also, what if the object is an external object? We can't use the VM's dma-resv
lock here.Why? Typically (sync) unlink is only ever called from an unbind-like operation where it should be trivial to grab the vm's resv. Or, for that matter any outer lock protecting the extobj list. Rule would be the drm_gpuvm_bo::entry::extobj and
drm_gpuvm_bo::entry::evictwould be protected by either the vm'sdma_resv(or possibly an outer lock in the case of the extobj list).Outer lock wouldn't have been working for updates in the async path, but shouldn't be relevant anymore. We could use the VM's resv for that.
Quoted 6 lines
Quoted 6 lines
And we can't have the GEM objs dma-resv lock held when calling
unlink(), sinceunlink()callsdrm_gpuvm_bo_put(), which if the refcount drops
to zero callsdrm_gpuvm_bo_destroy()anddrm_gpuvm_bo_destroy()might drop the
last reference of the GEM object.Quoted 10 lines
Yes, but this is a different problem as to what exactly protects
drm_gpuvm_bo::entry::gem. Either as you suggest an internal per bo list lock, or if we want to keep the bo'sdma_resvwe need to ensure that the caller ofdma_resv_unlock(obj->resv)actually refcounts its obj pointer, and doesn't implicitly rely on the gpuvm_bo's refcount (I know Boris didn't like that, but requiring an explicit refcount for a pointer you dereference unless you're under a lock that ensures keeping the object alive is pretty much required?) But anyway for thedrm_gpuvm_bo::entry::gemlist protection (bo resv or internal spinlock) I don't have a strong preference.We can keep the GEM objects dma-resv lock, however as mentioned above
drm_gpuva_unlink()anddrm_gpuvm_bo_put()then requires both the VM's resv lock and the GEM's resv lock in case they differ.Quoted 970 lines
Quoted 2 lines
All those problems go away with a dedicated
GEM gpuva list lock.I don't think these are real problems.
With the excepton of the eviction list "trick" where we currently have slightly different approach to collect external bos needing rebinding, we have this working fine.TBH I think pretty much the only situation where the spinlock is needed is for async updates of these lists, unless a wq item can be used for that, but it doesn't really seem like the current code allows for such updates anyway? It complicates the code a lot, adds overhead and also adds the requirement for refcounting during list traversal.
/Thomas
Quoted 955 lines
Quoted 955 lines
/Thomas
Quoted 952 lines
Quoted 952 lines
It seems that with that also the refcount could be make non- atomic.
All in the spirit of the drm locking guidelines "use big locks when
possible".
Lower level locks only when necessary for performance or locking inversion?/Thomas
Quoted 940 lines
+ *
+ * Elements popped from the original list are kept in a local list, so removal
+ * andis_emptychecks can still happen while we're iterating the list.
+ */
+#defineget_next_vm_bo_from_list(__gpuvm, __list_name, __local_list, __prev_vm_bo)\
+ ({
\
+ structdrm_gpuvm_bo
*__vm_bo; \
+
\
+ drm_gpuvm_bo_put(__prev_vm_bo); \
+
\
+ spin_lock(&(__gpuvm)-Quoted 1 line
__list_name.lock); \
+ while (!list_empty(&(__gpuvm)-
Quoted 1 line
__list_name.list)) { \
+ __vm_bo = list_first_entry(&(__gpuvm)->__list_name.list, \
+ struct
drm_gpuvm_bo, \
+
list.entry.__list_name); \
+ if (drm_gpuvm_bo_get_unless_zero(__vm_bo))
{ \
+ list_move_tail(&(__vm_bo)-Quoted 1 line
list.entry.__list_name, \
+
__local_list); \
+ break;
\
+ } else
{ \
+ list_del_init(&(__vm_bo)-Quoted 1 line
list.entry.__list_name); \
+ __vm_bo =
NULL; \
+ }
\
+ }
\
+ spin_unlock(&(__gpuvm)-Quoted 1 line
__list_name.lock); \
+
\
+ __vm_bo;
\
+ })drivers/gpu/drm/drm_gpuvm.c +5
-0+ +/** +* for_each_vm_bo_in_list() - internal vm_bo list iterator +* +* This helper is here to provide lockless list iteration. Lockless as in, the
+ * iterator releases the lock immediately after picking the first element from the
+ * list, so list insertion and deletion can happen concurrently.
+ *
+ * Typical use:
+ *
+ * structdrm_gpuvm_bo*vm_bo;
+ * LIST_HEAD(my_local_list);
+ *
+ * ret = 0;
+ * drm_gpuvm_for_each_vm_bo(gpuvm, <list_name>,&my_local_list,vm_bo) {
+ * ret =do_something_with_vm_bo(..., vm_bo);
+ * if (ret)
+ * break;
+ * }
+ * drm_gpuvm_bo_put(vm_bo);
+ * drm_gpuvm_restore_vm_bo_list(gpuvm, <list_name>,&my_local_list);
+ *
+ *
+ * Only used for internal list iterations, not meant to be exposed to the outside
+ * world.
+ */
+#definefor_each_vm_bo_in_list(__gpuvm, __list_name, __local_list, __vm_bo)\
+ for (__vm_bo= get_next_vm_bo_from_list(__gpuvm, __list_name, \
+ __local_list, NULL); \
+__vm_bo;
\
+__vm_bo= get_next_vm_bo_from_list(__gpuvm, __list_name, \
+ __local_list, __vm_bo)) \drivers/gpu/drm/drm_gpuvm.c +3
-0+ +/** +* restore_vm_bo_list() - move vm_bo elements back to their original list
+ * @__gpuvm: The GPU VM
+ * @__list_name: The name of the list we're iterating on
+ * @__local_list: A pointer to the local list used to store already iterated items
+ *
+ * When we're done iterating avm_bolist, we should callrestore_vm_bo_list()
+ * to restore the original state and let new iterations take place.
+ */
+#definerestore_vm_bo_list(__gpuvm, __list_name, __local_list)\
+ do
{
\
+ /* Merge back the two lists, moving local list elements to the \
+ * head to preserve previous ordering, in case it matters. \
+
*/
\
+ spin_lock(&(__gpuvm)-Quoted 1 line
__list_name.lock); \
+ list_splice(__local_list, &(__gpuvm)-
Quoted 1 line
__list_name.list); \
+ spin_unlock(&(__gpuvm)-
Quoted 1 line
__list_name.lock); \
+ } while (0)
+/**
+ *drm_gpuvm_bo_list_add()- insert avm_bointo the given list
+ * @__vm_bo: the&drm_gpuvm_bo
+ * @__list_name: the name of the list to insert into
+ *
+ * Inserts the given @__vm_bo into the list specified by @__list_name and
+ * increases the vm_bo's reference count.
+ */
+#definedrm_gpuvm_bo_list_add(__vm_bo, __list_name)\
+ do
{
\
+ spin_lock(&(__vm_bo)->vm-Quoted 1 line
__list_name.lock); \
+ if (list_empty(&(__vm_bo)-
Quoted 1 line
list.entry.__list_name)) \
+ list_add_tail(&(__vm_bo)-
Quoted 1 line
list.entry.__list_name, \
+ &(__vm_bo)->vm-
Quoted 1 line
__list_name.list); \
+ spin_unlock(&(__vm_bo)->vm-
Quoted 1 line
__list_name.lock); \
+ } while (0)
drivers/gpu/drm/drm_gpuvm.c +3
-0+ +/** +* drm_gpuvm_bo_list_del() - remove a vm_bo from the given list
+ * @__vm_bo: the&drm_gpuvm_bo
+ * @__list_name: the name of the list to insert into
+ *
+ * Removes the given @__vm_bo from the list specified by @__list_name and
+ * decreases the vm_bo's reference count.
+ */
+#definedrm_gpuvm_bo_list_del(__vm_bo, __list_name)\
+ do
{
\
+ spin_lock(&(__vm_bo)->vm-Quoted 1 line
__list_name.lock); \
+ if (!list_empty(&(__vm_bo)-
Quoted 1 line
list.entry.__list_name)) \
+ list_del_init(&(__vm_bo)-
Quoted 1 line
list.entry.__list_name); \
+ spin_unlock(&(__vm_bo)->vm-
Quoted 1 line
__list_name.lock); \
+ } while (0)
drivers/gpu/drm/drm_gpuvm.c +4
-0+ +static int __must_check +drm_gpuvm_bo_get_unless_zero (struct drm_gpuvm_bo *vm_bo); +#define to_drm_gpuva(__node) container_of((__node), struct
drm_gpuva, rb.node)
#defineGPUVA_START(node)((node)->va.addr)drivers/gpu/drm/drm_gpuvm.c +6
-0@@ -713,6 +867,12 @@ drm_gpuvm_init(struct drm_gpuvm *gpuvm,struct drm_device *drm, gpuvm->rb.tree = RB_ROOT_CACHED; INIT_LIST_HEAD(&gpuvm->rb.list); + INIT_LIST_HEAD(&gpuvm->extobj.list); + spin_lock_init(&gpuvm->extobj.lock); + + INIT_LIST_HEAD(&gpuvm->evict.list); + spin_lock_init(&gpuvm->evict.lock); + drm_gpuva_check_overflow(start_offset, range); gpuvm->mm_start = start_offset; gpuvm->mm_range = range;@@ -754,10 +914,302 @@ drm_gpuvm_destroy(struct drm_gpuvm *gpuvm) WARN(!RB_EMPTY_ROOT(&gpuvm->rb.tree.rb_root),"GPUVA tree is not empty, potentially leaking memory.\n");
+ WARN(!list_empty(&gpuvm->extobj.list), "Extobj list should be empty.\n");
+ WARN(!list_empty(&gpuvm->evict.list), "Evict list should be empty.\n");
+
drm_gem_private_object_fini(&gpuvm->d_obj); }
EXPORT_SYMBOL_GPL(drm_gpuvm_destroy);
+/**
+ *drm_gpuvm_prepare_objects()- prepare all assoiciated BOs
+ * @gpuvm: the&drm_gpuvm
+ * @exec: the&drm_execlocking context
+ * @num_fences: the amount of&dma_fencesto reserve
+ *
+ * Callsdrm_exec_prepare_obj()for all&drm_gem_objectsthe given
+ *&drm_gpuvmcontains mappings of.
+ *
+ * Using this function directly, it is the drivers responsibility to call
+ *drm_exec_init()anddrm_exec_fini()accordingly.
+ *
+ * Note: This function is safe against concurrent insertion and removal of
+ * external objects, however it is not safe against concurrent usage itself.
+ *
+ * Drivers need to make sure to protect this case with either an outer VM lock
+ * or by callingdrm_gpuvm_prepare_vm()before this function within the
+ *drm_exec_until_all_locked()loop, such that the GPUVM's dma-resv lock ensures
+ * mutual exclusion.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +10
-0+int +drm_gpuvm_prepare_objects (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + unsigned int num_fences) +{ + struct drm_gpuvm_bo *vm_bo; + LIST_HEAD(extobjs); + int ret = 0; + + for_each_vm_bo_in_list(gpuvm, extobj, &extobjs,
vm_bo) {
+ ret =drm_exec_prepare_obj(exec, vm_bo->obj, num_fences);
+ if (ret)
+ break;
+ }
+ /* Drop ref in case we break out of the loop. */ + drm_gpuvm_bo_put(vm_bo);
+ restore_vm_bo_list(gpuvm, extobj, &extobjs);drivers/gpu/drm/drm_gpuvm.c +7
-0+ + return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_prepare_objects); + +/** +* drm_gpuvm_prepare_range() - prepare all BOs mapped within a given range
+ * @gpuvm: the&drm_gpuvm
+ * @exec: the&drm_execlocking context
+ * @addr: the start address within the VA space
+ * @range: the range to iterate within the VA space
+ * @num_fences: the amount of&dma_fencesto reserve
+ *
+ * Callsdrm_exec_prepare_obj()for all&drm_gem_objectsmapped between @addr
+ * and @addr + @range.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +2
-0+int +drm_gpuvm_prepare_range(struct drm_gpuvm *gpuvm, struct
drm_exec*exec,
+ u64 addr, u64 range, unsigned intnum_fences)drivers/gpu/drm/drm_gpuvm.c +9
-0+{ + struct drm_gpuva *va; + u64 end = addr + range; + int ret; + + drm_gpuvm_for_each_va_range(va, gpuvm, addr, end) { + struct drm_gem_object *obj = va->gem.obj; + + ret = drm_exec_prepare_obj(exec, obj,
num_fences);
+ if (ret)
+ return ret;
+ }drivers/gpu/drm/drm_gpuvm.c +7
-0+ + return 0; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_prepare_range); + +/** +* drm_gpuvm_exec_lock() - lock all dma-resv of all assoiciated BOs
+ * @vm_exec: the&drm_gpuvm_execabstraction
+ * @num_fences: the amount of&dma_fencesto reserve
+ * @interruptible: sleep interruptible if waiting
+ *
+ * Acquires all dma-resv locks of all&drm_gem_objectsthe given
+ *&drm_gpuvmcontains mappings of.
+ *
+ * Addionally, when calling this function with structdrm_gpuvm_exec::extra
+ * being set the driver receives the given @fn callback to lock additional
+ * dma-resv in the context of the&drm_gpuvm_execinstance. Typically, drivers
+ * would calldrm_exec_prepare_obj()from within this callback.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +11
-0+int +drm_gpuvm_exec_lock (struct drm_gpuvm_exec *vm_exec, + unsigned int num_fences, + bool interruptible) +{ + struct drm_gpuvm *gpuvm = vm_exec->vm; + struct drm_exec *exec = &vm_exec->exec; + uint32_t flags; + int ret; + + flags = interruptible ? DRM_EXEC_INTERRUPTIBLE_WAIT :0 |
+ DRM_EXEC_IGNORE_DUPLICATES;drivers/gpu/drm/drm_gpuvm.c +5
-0+ + drm_exec_init(exec, flags); + + drm_exec_until_all_locked(exec) { + ret = drm_gpuvm_prepare_vm(gpuvm, exec,
num_fences);
+ drm_exec_retry_on_contention(exec); + if (ret)
+ goto err;
+
+ ret =drm_gpuvm_prepare_objects(gpuvm, exec, num_fences);
+ drm_exec_retry_on_contention(exec); + if (ret)
+ goto err;
+
+ if (vm_exec->extra.fn) {
+ ret =vm_exec->extra.fn(vm_exec, num_fences);
+ drm_exec_retry_on_contention(exec); + if (ret)
+ goto err;
+ }
+ }drivers/gpu/drm/drm_gpuvm.c +11
-0+ + return 0; + +err: + drm_exec_fini(exec); + return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock); + +static int +fn_lock_array(struct drm_gpuvm_exec *vm_exec, unsigned int
num_fences)drivers/gpu/drm/drm_gpuvm.c +7
-0+{ + struct { + struct drm_gem_object **objs; + unsigned int num_objs; + } *args = vm_exec->extra.priv; + + return drm_exec_prepare_array(&vm_exec->exec, args-Quoted 1 line
objs,
+
args->num_objs,num_fences);drivers/gpu/drm/drm_gpuvm.c +4
-0+} + +/** +* drm_gpuvm_exec_lock_array() - lock all dma-resv of all assoiciated BOs
+ * @vm_exec: the&drm_gpuvm_execabstraction
+ * @objs: additional&drm_gem_objectsto lock
+ * @num_objs: the number of additional&drm_gem_objectsto lock
+ * @num_fences: the amount of&dma_fencesto reserve
+ * @interruptible: sleep interruptible if waiting
+ *
+ * Acquires all dma-resv locks of all&drm_gem_objectsthe given&drm_gpuvm
+ * contains mappings of, plus the ones given through @objs.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +19
-0+int +drm_gpuvm_exec_lock_array (struct drm_gpuvm_exec *vm_exec, + struct drm_gem_object **objs, + unsigned int num_objs, + unsigned int num_fences, + bool interruptible) +{ + struct { + struct drm_gem_object **objs; + unsigned int num_objs; + } args; + + args.objs = objs; + args.num_objs = num_objs; + + vm_exec->extra.fn = fn_lock_array; + vm_exec->extra.priv = &args; + + return drm_gpuvm_exec_lock(vm_exec, num_fences,interruptible);
drivers/gpu/drm/drm_gpuvm.c +5
-0+} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock_array); + +/** +* drm_gpuvm_exec_lock_range() - prepare all BOs mapped within a given range
+ * @vm_exec: the&drm_gpuvm_execabstraction
+ * @addr: the start address within the VA space
+ * @range: the range to iterate within the VA space
+ * @num_fences: the amount of&dma_fencesto reserve
+ * @interruptible: sleep interruptible if waiting
+ *
+ * Acquires all dma-resv locks of all&drm_gem_objectsmapped between @addr and
+ * @addr + @range.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +12
-0+int +drm_gpuvm_exec_lock_range (struct drm_gpuvm_exec *vm_exec, + u64 addr, u64 range, + unsigned int num_fences, + bool interruptible) +{ + struct drm_gpuvm *gpuvm = vm_exec->vm; + struct drm_exec *exec = &vm_exec->exec; + uint32_t flags; + int ret; + + flags = interruptible ? DRM_EXEC_INTERRUPTIBLE_WAIT :0 |
+ DRM_EXEC_IGNORE_DUPLICATES;drivers/gpu/drm/drm_gpuvm.c +5
-0+ + drm_exec_init(exec, flags); + + drm_exec_until_all_locked(exec) { + ret = drm_gpuvm_prepare_range(gpuvm, exec,addr, range,
+num_fences);
+ drm_exec_retry_on_contention(exec); + if (ret)
+ goto err;
+ }drivers/gpu/drm/drm_gpuvm.c +14
-0+ + return ret; + +err: + drm_exec_fini(exec); + return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_exec_lock_range); + +/** +* drm_gpuvm_validate() - validate all BOs marked as evicted +* @gpuvm: the &drm_gpuvm to validate evicted BOs +* +* Calls the &drm_gpuvm_ops.bo_validate callback for all evicted buffer
+ * objects being mapped in the given&drm_gpuvm.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */drivers/gpu/drm/drm_gpuvm.c +27
-0+int +drm_gpuvm_validate (struct drm_gpuvm *gpuvm) +{ + const struct drm_gpuvm_ops *ops = gpuvm->ops; + struct drm_gpuvm_bo *vm_bo; + LIST_HEAD(evict); + int ret = 0; + + if (unlikely(!ops || !ops->bo_validate)) + return -ENOTSUPP; + + for_each_vm_bo_in_list(gpuvm, evict, &evict, vm_bo) { + dma_resv_assert_held(vm_bo->obj->resv); + ret = ops->bo_validate(vm_bo->obj); + if (ret) + break; + } + /* Drop ref in case we break out of the loop. */ + drm_gpuvm_bo_put(vm_bo); + restore_vm_bo_list(gpuvm, evict, &evict); + + return ret; +} +EXPORT_SYMBOL_GPL (drm_gpuvm_validate); + +/** +* drm_gpuvm_resv_add_fence - add fence to private and all extobj
+ * dma-resv
+ * @gpuvm: the&drm_gpuvmto add a fence to
+ * @exec: the&drm_execlocking context
+ * @fence: fence to add
+ * @private_usage: private dma-resv usage
+ * @extobj_usage: extobj dma-resv usage
+ */drivers/gpu/drm/drm_gpuvm.c +14
-0+void +drm_gpuvm_resv_add_fence (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + struct dma_fence *fence, + enum dma_resv_usage private_usage, + enum dma_resv_usage extobj_usage) +{ + struct drm_gem_object *obj; + unsigned long index; + + drm_exec_for_each_locked_object(exec, index, obj) { + dma_resv_assert_held(obj->resv); + dma_resv_add_fence(obj->resv, fence, + drm_gpuvm_is_extobj(gpuvm,obj) ?
+private_usage:extobj_usage);
+ }drivers/gpu/drm/drm_gpuvm.c +3
-0+} +EXPORT_SYMBOL_GPL (drm_gpuvm_resv_add_fence); +/**
*drm_gpuvm_bo_create()- create a new instance of structdrm_gpuvm_bo
* @gpuvm: The&drm_gpuvmthe @obj is mapped in.drivers/gpu/drm/drm_gpuvm.c +13
-0@@ -790,6 +1242,9 @@ drm_gpuvm_bo_create(struct drm_gpuvm *gpuvm, INIT_LIST_HEAD(&vm_bo->list.gpuva); INIT_LIST_HEAD(&vm_bo->list.entry.gem); + INIT_LIST_HEAD(&vm_bo->list.entry.extobj); + INIT_LIST_HEAD(&vm_bo->list.entry.evict); + drm_gem_object_get(obj); return vm_bo;@@ -807,6 +1262,14 @@ drm_gpuvm_bo_destroy(struct kref *kref) drm_gem_gpuva_assert_lock_held(vm_bo->obj); + spin_lock(&gpuvm->extobj.lock); + list_del(&vm_bo->list.entry.extobj); + spin_unlock(&gpuvm->extobj.lock); + + spin_lock(&gpuvm->evict.lock); + list_del(&vm_bo->list.entry.evict); + spin_unlock(&gpuvm->evict.lock); + list_del(&vm_bo->list.entry.gem); drm_gem_object_put(obj);@@ -822,6 +1285,11 @@ drm_gpuvm_bo_destroy(struct kref *kref) * @vm_bo: the &drm_gpuvm_bo to release the reference of * * This releases a reference to @vm_bo. + * + * If the reference count drops to zero, the &gpuvm_bo isdestroyed, which
+ * includes removing it from the GEMs gpuva list. Hence, if a call to this
+ * function can potentially let the reference count to zero the caller must
+ * hold the dma-resv or driver specific GEM gpuva lock. */
void
drm_gpuvm_bo_put(struct drm_gpuvm_bo *vm_bo)drivers/gpu/drm/drm_gpuvm.c +6
-0@@ -831,6 +1299,12 @@ drm_gpuvm_bo_put(struct drm_gpuvm_bo *vm_bo) } EXPORT_SYMBOL_GPL(drm_gpuvm_bo_put); +static int __must_check +drm_gpuvm_bo_get_unless_zero (struct drm_gpuvm_bo *vm_bo) +{ + return kref_get_unless_zero(&vm_bo->kref); +} +static struct
drm_gpuvm_bo*__drm_gpuvm_bo_find(struct drm_gpuvm *gpuvm, struct drm_gem_object *obj)drivers/gpu/drm/drm_gpuvm.c +2
-0@@ -938,6 +1412,48 @@ drm_gpuvm_bo_obtain_prealloc(struct drm_gpuvm_bo *__vm_bo) }EXPORT_SYMBOL_GPL (drm_gpuvm_bo_obtain_prealloc); +/** +* drm_gpuvm_bo_extobj_add() - adds the &drm_gpuvm_bo to its &drm_gpuvm's
+ * extobj list
+ * @vm_bo: The&drm_gpuvm_boto add to its &drm_gpuvm's the extobj list.
+ *
+ * Adds the given @vm_bo to its &drm_gpuvm's extobj list if not on the list
+ * already and if the corresponding&drm_gem_objectis an external object,
+ * actually.
+ */drivers/gpu/drm/drm_gpuvm.c +12
-0+void +drm_gpuvm_bo_extobj_add (struct drm_gpuvm_bo *vm_bo) +{ + struct drm_gpuvm *gpuvm = vm_bo->vm; + + if (drm_gpuvm_is_extobj(gpuvm, vm_bo->obj)) + drm_gpuvm_bo_list_add(vm_bo, extobj); +} +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_extobj_add); + +/** +* drm_gpuvm_bo_evict() - add / remove a &drm_gem_object to / from a
+ *&drm_gpuvmsevicted list
+ * @obj: the&drm_gem_objectto add or remove
+ * @evict: indicates whether the object is evicted
+ *
+ * Adds a&drm_gem_objectto or removes it from all&drm_gpuvmsevicted
+ * list containing a mapping of this&drm_gem_object.
+ */drivers/gpu/drm/drm_gpuvm.c +14
-0+void +drm_gpuvm_bo_evict (struct drm_gem_object *obj,bool evict) +{ + struct drm_gpuvm_bo *vm_bo; + + drm_gem_for_each_gpuvm_bo(vm_bo, obj) { + if (evict) + drm_gpuvm_bo_list_add(vm_bo, evict); + else + drm_gpuvm_bo_list_del(vm_bo, evict); + } +} +EXPORT_SYMBOL_GPL (drm_gpuvm_bo_evict); +static int
__drm_gpuva_insert(structdrm_gpuvm*gpuvm,
structdrm_gpuva*va)include/drm/drm_gpuvm.h +9
-0@@ -26,10 +26,12 @@*/ #include <linux/list.h> +#include <linux/dma-resv.h> #include <linux/rbtree.h> #include <linux/types.h> #include <drm/drm_gem.h> +#include <drm/drm_exec.h> struct drm_gpuvm; struct drm_gpuvm_bo;@@ -259,6 +261,38 @@ struct drm_gpuvm {* space */ struct dma_resv *resv; + + /** + * @extobj: structure holding the extobj list + */ + struct { + /** +* @list: &list_head storing &drm_gpuvm_bos serving as
+ * external object
+ */
+ structlist_headlist;include/drm/drm_gpuvm.h +9
-0+ + /** + * @lock: spinlock to protect the extobj list + */ + spinlock_t lock; + } extobj; + + /** +* @evict: structure holding the evict list and evict list lock
+ */
+ struct {
+ /**
+ * @list:&list_headstoring&drm_gpuvm_boscurrently being
+ * evicted
+ */
+ structlist_headlist;
+
+ /**
+ * @lock: spinlock to protect the evict list + */
+ spinlock_t lock;
+ } evict;
};
void drm_gpuvm_init(structdrm_gpuvm*gpuvm, structdrm_device*drm,include/drm/drm_gpuvm.h +2
-0@@ -268,6 +302,21 @@ void drm_gpuvm_init(struct drm_gpuvm *gpuvm, struct drm_device *drm, const struct drm_gpuvm_ops *ops); void drm_gpuvm_destroy(struct drm_gpuvm *gpuvm); +/** +* drm_gpuvm_is_extobj() - indicates whether the given
&drm_gem_objectis an
+ * external object
+ * @gpuvm: the&drm_gpuvmto check
+ * @obj: the&drm_gem_objectto check
+ *
+ * Returns: true if the&drm_gem_object&dma_resvdiffers from the
+ *&drm_gpuvms&dma_resv, false otherwise
+ */
+static inline bool drm_gpuvm_is_extobj(structdrm_gpuvm*gpuvm,
+ structdrm_gem_object*obj)include/drm/drm_gpuvm.h +4
-0+{ + return obj && obj->resv != gpuvm->resv; +} +static inline struct
drm_gpuva*
__drm_gpuva_next(struct drm_gpuva *va)
{include/drm/drm_gpuvm.h +0
-0@@ -346,6 +395,128 @@ __drm_gpuva_next(struct drm_gpuva *va) #define drm_gpuvm_for_each_va_safe(va__, next__, gpuvm__) \list_for_each_entry_safe(va__,
next__, &(gpuvm__)-Quoted 1 line
rb.list, rb.entry)
+/**
+ * structdrm_gpuvm_exec-&drm_gpuvmabstraction of&drm_exec
+ *
+ * This structure should be created on the stack as&drm_execshould be.
+ *
+ * Optionally, @extra can be set in order to lock additional&drm_gem_objects.
+ */include/drm/drm_gpuvm.h +13
-0+struct drm_gpuvm_exec { + /** + * @exec: the &drm_exec structure + */ + struct drm_exec exec; + + /** + * @vm: the &drm_gpuvm to lock its DMA reservations + */ + struct drm_gpuvm *vm; + + /** + * @extra: Callback and corresponding private datafor the driver to
+ * lock arbitrary additional&drm_gem_objects. + */
+ struct {
+ /**
+ * @fn: The driver callback to lock additional&drm_gem_objects.
+ */
+ int (*fn)(structdrm_gpuvm_exec*vm_exec, + unsigned intnum_fences); +
+ /**
+ * @priv: driver private data for the @fn callback
+ */
+ void *priv;
+ } extra;include/drm/drm_gpuvm.h +4
-0+}; + +/** +* drm_gpuvm_prepare_vm() - prepare the GPUVMs common dma- resv
+ * @gpuvm: the&drm_gpuvm
+ * @exec: the&drm_execcontext
+ * @num_fences: the amount of&dma_fencesto reserve
+ *
+ * Callsdrm_exec_prepare_obj()for the GPUVMs dummy&drm_gem_object.
+ *
+ * Using this function directly, it is the drivers responsibility to call
+ *drm_exec_init()anddrm_exec_fini()accordingly.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */include/drm/drm_gpuvm.h +6
-0+static inline int +drm_gpuvm_prepare_vm (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + unsigned int num_fences) +{ + return drm_exec_prepare_obj(exec, &gpuvm->d_obj,
num_fences);include/drm/drm_gpuvm.h +16
-0+} + +int drm_gpuvm_prepare_objects (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + unsigned int num_fences); + +int drm_gpuvm_prepare_range (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + u64 addr, u64 range, + unsigned int num_fences); + +int drm_gpuvm_exec_lock (struct drm_gpuvm_exec *vm_exec, + unsigned int num_fences, + bool interruptible); + +int drm_gpuvm_exec_lock_array(struct drm_gpuvm_exec
*vm_exec,
+ structdrm_gem_object**objs, + unsigned intnum_objs, + unsigned intnum_fences, + bool interruptible);include/drm/drm_gpuvm.h +2
-0+ +int drm_gpuvm_exec_lock_range(struct drm_gpuvm_exec
*vm_exec,
+ u64 addr, u64 range, + unsigned intnum_fences, + bool interruptible);include/drm/drm_gpuvm.h +3
-0+ +/** +* drm_gpuvm_lock() - lock all dma-resv of all assoiciated BOs
+ * @gpuvm: the&drm_gpuvm
+ *
+ * Releases all dma-resv locks of all&drm_gem_objectspreviously acquired
+ * throughdrm_gpuvm_lock()or its variants.
+ *
+ * Returns: 0 on success, negative error code on failure.
+ */include/drm/drm_gpuvm.h +11
-0+static inline void +drm_gpuvm_exec_unlock (struct drm_gpuvm_exec *vm_exec) +{ + drm_exec_fini(&vm_exec->exec); +} + +int drm_gpuvm_validate (struct drm_gpuvm *gpuvm); +void drm_gpuvm_resv_add_fence (struct drm_gpuvm *gpuvm, + struct drm_exec *exec, + struct dma_fence *fence, + enum dma_resv_usage
private_usage,
+ enumdma_resv_usageextobj_usage);include/drm/drm_gpuvm.h +12
-0+ +/** +* drm_gpuvm_exec_resv_add_fence() +* @vm_exec: the &drm_gpuvm_exec abstraction +* @fence: fence to add +* @private_usage: private dma-resv usage +* @extobj_usage: extobj dma-resv usage +* +* See drm_gpuvm_resv_add_fence(). +*/ +static inline void +drm_gpuvm_exec_resv_add_fence(struct drm_gpuvm_exec
*vm_exec,
+ structdma_fence*fence, + enumdma_resv_usageprivate_usage,
+ enumdma_resv_usageextobj_usage)
+{
+ drm_gpuvm_resv_add_fence(vm_exec->vm,&vm_exec->exec, fence,
+private_usage,extobj_usage);include/drm/drm_gpuvm.h +2
-0+} +/**
* structdrm_gpuvm_bo- structure representing a&drm_gpuvmand
*&drm_gem_objectcombinationinclude/drm/drm_gpuvm.h +3
-0@@ -398,6 +569,18 @@ struct drm_gpuvm_bo {* gpuva list. */ struct list_head gem; + + /** +* @evict: List entry to attach to the
&drm_gpuvms
+ * extobj list.
+ */
+ structlist_headextobj; +
+ /**
+ * @evict: List entry to attach to the&drm_gpuvmsevict
+ * list.
+ */
+ structlist_headevict;
} entry;
} list;
};include/drm/drm_gpuvm.h +3
-0@@ -432,6 +615,9 @@ struct drm_gpuvm_bo * drm_gpuvm_bo_find(struct drm_gpuvm *gpuvm, struct drm_gem_object *obj); +void drm_gpuvm_bo_evict (struct drm_gem_object *obj,bool evict); +void drm_gpuvm_bo_extobj_add (struct drm_gpuvm_bo *vm_bo); + /*** drm_gpuvm_bo_for_each_va() - iterator to walk over a list of
&drm_gpuva
* @va__:&drm_gpuvastructure to assign to in each iteration stepinclude/drm/drm_gpuvm.h +0
-0@@ -837,6 +1023,17 @@ struct drm_gpuvm_ops {* used. */ int (*sm_step_unmap)(struct
drm_gpuva_op*op, void *priv);
+
+ /**
+ * @bo_validate: called fromdrm_gpuvm_validate()+ *
+ * Drivers receive this callback for every evicted&drm_gem_objectbeing
+ * mapped in the corresponding&drm_gpuvm. + *
+ * Typically, drivers would call their driver specific variant of
+ *ttm_bo_validate()from within this callback. + */
+ int (*bo_validate)(structdrm_gem_object*obj); };
int drm_gpuvm_sm_map(structdrm_gpuvm*gpuvm, void *priv,