[PATCH v2 09/29] drm/xe: Add SVM range invalidation
Matthew Brost
matthew.brost at intel.com
Mon Dec 16 16:09:38 UTC 2024
On Mon, Dec 16, 2024 at 11:01:23AM +0100, Thomas Hellström wrote:
> On Wed, 2024-12-11 at 11:01 -0800, Matthew Brost wrote:
> > On Tue, Nov 19, 2024 at 02:56:12PM +0100, Thomas Hellström wrote:
> > > On Tue, 2024-10-15 at 20:24 -0700, Matthew Brost wrote:
> > > > Add SVM range invalidation vfunc.
> > > >
> > > > v2:
> > > > - Don't run invalidation if VM is closed
> > > > - Cycle notifier lock in xe_svm_close
> > > > - Drop xe_gt_tlb_invalidation_fence_fini
> > > >
> > > > Signed-off-by: Matthew Brost <matthew.brost at intel.com>
> > > > ---
> > > > drivers/gpu/drm/xe/xe_gt_pagefault.c | 17 ++-
> > > > drivers/gpu/drm/xe/xe_pt.c | 24 ++++
> > > > drivers/gpu/drm/xe/xe_pt.h | 3 +
> > > > drivers/gpu/drm/xe/xe_svm.c | 205
> > > > ++++++++++++++++++++++++++-
> > > > drivers/gpu/drm/xe/xe_svm.h | 13 ++
> > > > 5 files changed, 256 insertions(+), 6 deletions(-)
> > > >
> > > > diff --git a/drivers/gpu/drm/xe/xe_gt_pagefault.c
> > > > b/drivers/gpu/drm/xe/xe_gt_pagefault.c
> > > > index 79c426dc2505..92923947a12c 100644
> > > > --- a/drivers/gpu/drm/xe/xe_gt_pagefault.c
> > > > +++ b/drivers/gpu/drm/xe/xe_gt_pagefault.c
> > > > @@ -19,6 +19,7 @@
> > > > #include "xe_guc.h"
> > > > #include "xe_guc_ct.h"
> > > > #include "xe_migrate.h"
> > > > +#include "xe_svm.h"
> > > > #include "xe_trace_bo.h"
> > > > #include "xe_vm.h"
> > > >
> > > > @@ -125,18 +126,17 @@ static int xe_pf_begin(struct drm_exec
> > > > *exec,
> > > > struct xe_vma *vma,
> > > > return 0;
> > > > }
> > > >
> > > > -static int handle_vma_pagefault(struct xe_tile *tile, struct
> > > > pagefault *pf,
> > > > - struct xe_vma *vma)
> > > > +static int handle_vma_pagefault(struct xe_tile *tile, struct
> > > > xe_vma
> > > > *vma,
> > > > + bool atomic)
> > > > {
> > > > struct xe_vm *vm = xe_vma_vm(vma);
> > > > struct drm_exec exec;
> > > > struct dma_fence *fence;
> > > > ktime_t end = 0;
> > > > int err;
> > > > - bool atomic;
> > > >
> > > > + lockdep_assert_held_write(&vm->lock);
> > > > trace_xe_vma_pagefault(vma);
> > > > - atomic = access_is_atomic(pf->access_type);
> > > >
> > > > /* Check if VMA is valid */
> > > > if (vma_is_valid(tile, vma) && !atomic)
> > > > @@ -207,6 +207,7 @@ static int handle_pagefault(struct xe_gt *gt,
> > > > struct pagefault *pf)
> > > > struct xe_vm *vm;
> > > > struct xe_vma *vma = NULL;
> > > > int err;
> > > > + bool atomic;
> > > >
> > > > /* SW isn't expected to handle TRTT faults */
> > > > if (pf->trva_fault)
> > > > @@ -232,7 +233,13 @@ static int handle_pagefault(struct xe_gt
> > > > *gt,
> > > > struct pagefault *pf)
> > > > goto unlock_vm;
> > > > }
> > > >
> > > > - err = handle_vma_pagefault(tile, pf, vma);
> > > > + atomic = access_is_atomic(pf->access_type);
> > > > +
> > > > + if (xe_vma_is_system_allocator(vma))
> > > > + err = xe_svm_handle_pagefault(vm, vma, tile,
> > > > + pf->page_addr,
> > > > atomic);
> > > > + else
> > > > + err = handle_vma_pagefault(tile, vma, atomic);
> > > >
> > > > unlock_vm:
> > > > if (!err)
> > > > diff --git a/drivers/gpu/drm/xe/xe_pt.c
> > > > b/drivers/gpu/drm/xe/xe_pt.c
> > > > index 39357e829b6d..282476c4edbd 100644
> > > > --- a/drivers/gpu/drm/xe/xe_pt.c
> > > > +++ b/drivers/gpu/drm/xe/xe_pt.c
> > > > @@ -20,6 +20,7 @@
> > > > #include "xe_res_cursor.h"
> > > > #include "xe_sched_job.h"
> > > > #include "xe_sync.h"
> > > > +#include "xe_svm.h"
> > > > #include "xe_trace.h"
> > > > #include "xe_ttm_stolen_mgr.h"
> > > > #include "xe_vm.h"
> > > > @@ -829,6 +830,29 @@ bool xe_pt_zap_ptes(struct xe_tile *tile,
> > > > struct
> > > > xe_vma *vma)
> > > > return xe_walk.needs_invalidate;
> > > > }
> > > >
> > > > +bool xe_pt_zap_ptes_range(struct xe_tile *tile, struct xe_vm
> > > > *vm,
> > > > + struct xe_svm_range *range)
> > >
> > > Kerneldoc.
> > >
> >
> > Will add.
> >
> > > Here, (and I saw Oak also commented around this some time ago)
> > > ideally
> > > we should make xe_pt.c unaware of vmas and svm ranges, and in this
> > > case, use the same xe_pt function for both.
> > >
> >
> > See some of other comments, agree we should do in a follow up.
> >
> > >
> > >
> > > > +{
> > > > + struct xe_pt_zap_ptes_walk xe_walk = {
> > > > + .base = {
> > > > + .ops = &xe_pt_zap_ptes_ops,
> > > > + .shifts = xe_normal_pt_shifts,
> > > > + .max_level = XE_PT_HIGHEST_LEVEL,
> > > > + },
> > > > + .tile = tile,
> > > > + };
> > > > + struct xe_pt *pt = vm->pt_root[tile->id];
> > > > + u8 pt_mask = (range->tile_present & ~range-
> > > > > tile_invalidated);
> > > > +
> > > > + if (!(pt_mask & BIT(tile->id)))
> > > > + return false;
> > > > +
> > > > + (void)xe_pt_walk_shared(&pt->base, pt->level, range-
> > > > > base.va.start,
> > > > + range->base.va.end,
> > > > &xe_walk.base);
> > > > +
> > > > + return xe_walk.needs_invalidate;
> > > > +}
> > > > +
> > > > static void
> > > > xe_vm_populate_pgtable(struct xe_migrate_pt_update *pt_update,
> > > > struct xe_tile *tile,
> > > > struct iosys_map *map, void *data,
> > > > diff --git a/drivers/gpu/drm/xe/xe_pt.h
> > > > b/drivers/gpu/drm/xe/xe_pt.h
> > > > index 9ab386431cad..5f333eeedf5c 100644
> > > > --- a/drivers/gpu/drm/xe/xe_pt.h
> > > > +++ b/drivers/gpu/drm/xe/xe_pt.h
> > > > @@ -13,6 +13,7 @@ struct dma_fence;
> > > > struct xe_bo;
> > > > struct xe_device;
> > > > struct xe_exec_queue;
> > > > +struct xe_svm_range;
> > > > struct xe_sync_entry;
> > > > struct xe_tile;
> > > > struct xe_vm;
> > > > @@ -42,5 +43,7 @@ void xe_pt_update_ops_fini(struct xe_tile
> > > > *tile,
> > > > struct xe_vma_ops *vops);
> > > > void xe_pt_update_ops_abort(struct xe_tile *tile, struct
> > > > xe_vma_ops
> > > > *vops);
> > > >
> > > > bool xe_pt_zap_ptes(struct xe_tile *tile, struct xe_vma *vma);
> > > > +bool xe_pt_zap_ptes_range(struct xe_tile *tile, struct xe_vm
> > > > *vm,
> > > > + struct xe_svm_range *range);
> > > >
> > > > #endif
> > > > diff --git a/drivers/gpu/drm/xe/xe_svm.c
> > > > b/drivers/gpu/drm/xe/xe_svm.c
> > > > index 57b740367843..b2bc259978c4 100644
> > > > --- a/drivers/gpu/drm/xe/xe_svm.c
> > > > +++ b/drivers/gpu/drm/xe/xe_svm.c
> > > > @@ -5,18 +5,188 @@
> > > >
> > > > #include "drm_gpusvm.h"
> > > >
> > > > +#include "xe_gt_tlb_invalidation.h"
> > > > +#include "xe_pt.h"
> > > > #include "xe_svm.h"
> > > > #include "xe_vm.h"
> > > > #include "xe_vm_types.h"
> > > >
> > > > +static struct xe_vm *gpusvm_to_vm(struct drm_gpusvm *gpusvm)
> > > > +{
> > > > + return container_of(gpusvm, struct xe_vm, svm.gpusvm);
> > > > +}
> > > > +
> > > > +static struct xe_vm *range_to_vm(struct drm_gpusvm_range *r)
> > > > +{
> > > > + return gpusvm_to_vm(r->gpusvm);
> > > > +}
> > > > +
> > > > +static struct drm_gpusvm_range *
> > > > +xe_svm_range_alloc(struct drm_gpusvm *gpusvm)
> > > > +{
> > > > + struct xe_svm_range *range;
> > > > +
> > > > + range = kzalloc(sizeof(*range), GFP_KERNEL);
> > > > + if (!range)
> > > > + return ERR_PTR(-ENOMEM);
> > > > +
> > > > + xe_vm_get(gpusvm_to_vm(gpusvm));
> > > > +
> > > > + return &range->base;
> > > > +}
> > > > +
> > > > +static void xe_svm_range_free(struct drm_gpusvm_range *range)
> > > > +{
> > > > + xe_vm_put(range_to_vm(range));
> > > > + kfree(range);
> > > > +}
> > > > +
> > > > +static struct xe_svm_range *to_xe_range(struct drm_gpusvm_range
> > > > *r)
> > > > +{
> > > > + return container_of(r, struct xe_svm_range, base);
> > > > +}
> > > > +
> > > > +static u8
> > > > +xe_svm_range_notifier_event_begin(struct xe_vm *vm, struct
> > > > drm_gpusvm_range *r,
> > > > + const struct
> > > > mmu_notifier_range
> > > > *mmu_range,
> > > > + u64 *adj_start, u64 *adj_end)
> > > > +{
> > > > + struct xe_svm_range *range = to_xe_range(r);
> > > > + struct xe_device *xe = vm->xe;
> > > > + struct xe_tile *tile;
> > > > + u8 tile_mask = 0;
> > > > + u8 id;
> > > > +
> > >
> > > lockdep assert?
> > >
> >
> > Sure.
> >
> > > > + /* Skip if already unmapped or if no binding exist */
> > > > + if (range->base.flags.unmapped || !range->tile_present)
> > > > + return 0;
> > > > +
> > > > + /* Adjust invalidation to range boundaries */
> > > > + if (range->base.va.start < mmu_range->start)
> > > > + *adj_start = range->base.va.start;
> > > > + if (range->base.va.end > mmu_range->end)
> > > > + *adj_end = range->base.va.end;
> > > > +
> > > > + /*
> > > > + * XXX: Ideally would zap PTEs in one shot in
> > > > xe_svm_invalidate but the
> > > > + * invalidation code can't correctly cope with sparse
> > > > ranges
> > > > or
> > > > + * invalidations spanning multiple ranges.
> > > > + */
> > > > + for_each_tile(tile, xe, id)
> > > > + if (xe_pt_zap_ptes_range(tile, vm, range)) {
> > > > + tile_mask |= BIT(id);
> > > > + range->tile_invalidated |= BIT(id);
> > > > + }
> > > > +
> > > > + return tile_mask;
> > > > +}
> > > > +
> > > > +static void
> > > > +xe_svm_range_notifier_event_end(struct xe_vm *vm, struct
> > > > drm_gpusvm_range *r,
> > > > + const struct mmu_notifier_range
> > > > *mmu_range)
> > > > +{
> > > > + struct drm_gpusvm_ctx ctx = { .in_notifier = true, };
> > > > +
> > > > + drm_gpusvm_range_unmap_pages(&vm->svm.gpusvm, r, &ctx);
> > > > + /* TODO: Add range to garbage collector */
> > > > +}
> > > > +
> > > > static void xe_svm_invalidate(struct drm_gpusvm *gpusvm,
> > > > struct drm_gpusvm_notifier
> > > > *notifier,
> > > > const struct mmu_notifier_range
> > > > *mmu_range)
> > > > {
> > > > - /* TODO: Implement */
> > > > + struct xe_vm *vm = gpusvm_to_vm(gpusvm);
> > > > + struct xe_device *xe = vm->xe;
> > > > + struct xe_tile *tile;
> > > > + struct drm_gpusvm_range *r, *first;
> > > > + struct xe_gt_tlb_invalidation_fence
> > > > + fence[XE_MAX_TILES_PER_DEVICE *
> > > > XE_MAX_GT_PER_TILE];
> > > > + u64 adj_start = mmu_range->start, adj_end = mmu_range-
> > > > >end;
> > > > + u8 tile_mask = 0;
> > > > + u8 id;
> > > > + u32 fence_id = 0;
> > > > + long err;
> > > > +
> > > > + if (xe_vm_is_closed(vm))
> > > > + return;
> > >
> > > How do we ensure we don't race here? Are we sure that all dma
> > > mappings
> > > and all PTEs pointing to the range is gone at this point? Becase
> > > "They
> > > will soon be gone anyway" isn't enough.
> > >
> >
> > I think this is to prevent touching PTs which are being destroyed in
> > parallel which resulted in kernel explosion, so I think we need this.
>
> IIRC, the pt structure change is committed under the notifier lock when
> unbinding, which means that a racing pt zap shouldn't do any harm and
> just have the commit phase re-run?
>
It is destroying the PTs in xe_vm_close_and_put which can race with the
notifier.
> So if we follow the
> 1.) take vm->lock : waits for existing binding to complete
> 2.) mark vm closed : inhibits future binding
> 3.) unbind page-table
> 4.) Remove notifiers
>
> We should be ok?
>
Yea, I think you missing a few locks but yea probably something like you
describe works. I have already coded like I describe below in my
previous reply though - next rev is likely to posted in the next day or
so. I think either works but what you suggest is likely a bit of
larger rework.
I'm thinking really we should completely rework xe_vm_close_and_put as
it is a bit of a mess and we have just continually bolted on things over
time without a ton of deep thought on what we want the flow to look
rather just get stuff working and move on. Would it be ok for me to open
a Jira for this rework, I can take ownership, and do this rework in a
follow on series?
Matt
> /Thomas
>
>
> >
> > How to prevent a race? How about on VM close we invalidate the PT
> > root?
> > I had patch at one point which did this. We'd still have dma mappings
> > too but I think if need to we can safely dma-unmap the pages if the
> > VM
> > is closed too. Thoughts?
> >
> > > > +
> > > > + /* Adjust invalidation to notifier boundaries */
> > > > + if (adj_start < notifier->interval.start)
> > > > + adj_start = notifier->interval.start;
> > > > + if (adj_end > notifier->interval.end)
> > > > + adj_end = notifier->interval.end;
> > > > +
> > > > + first = drm_gpusvm_range_find(notifier, adj_start,
> > > > adj_end);
> > > > + if (!first)
> > > > + return;
> > > > +
> > > > + /*
> > > > + * XXX: Less than ideal to always wait on VM's resv
> > > > slots if
> > > > an
> > > > + * invalidation is not required. Could walk range list
> > > > twice
> > > > to figure
> > > > + * out if an invalidations is need, but also not ideal.
> > > > Maybe a counter
> > > > + * within the notifier, seems like that could work.
> > > > + */
> > > > + err = dma_resv_wait_timeout(xe_vm_resv(vm),
> > > > + DMA_RESV_USAGE_BOOKKEEP,
> > > > + false,
> > > > MAX_SCHEDULE_TIMEOUT);
> > > > + XE_WARN_ON(err <= 0);
> > > > +
> > > > + r = first;
> > > > + drm_gpusvm_for_each_range(r, notifier, adj_start,
> > > > adj_end)
> > > > + tile_mask |=
> > > > xe_svm_range_notifier_event_begin(vm,
> > > > r, mmu_range,
> > > > +
> > > > &adj_start,
> > > > +
> > > > &adj_end);
> > > > + if (!tile_mask)
> > > > + goto range_notifier_event_end;
> > > > +
> > > > + xe_device_wmb(xe);
> > > > +
> > > > + for_each_tile(tile, xe, id) {
> > > > + if (tile_mask & BIT(id)) {
> > > > + int err;
> > > > +
> > > > + xe_gt_tlb_invalidation_fence_init(tile-
> > > > > primary_gt,
> > > > +
> > > > &fence[fence_id], true);
> > > > +
> > > > + err = xe_gt_tlb_invalidation_range(tile-
> > > > > primary_gt,
> > > > +
> > > > &fence[fence_id],
> > > > +
> > > > adj_start,
> > > > +
> > > > adj_end,
> > > > + vm-
> > > > > usm.asid);
> > > > + if (WARN_ON_ONCE(err < 0))
> > > > + goto wait;
> > > > + ++fence_id;
> > > > +
> > > > + if (!tile->media_gt)
> > > > + continue;
> > > > +
> > > > + xe_gt_tlb_invalidation_fence_init(tile-
> > > > > media_gt,
> > > > +
> > > > &fence[fence_id], true);
> > > > +
> > > > + err = xe_gt_tlb_invalidation_range(tile-
> > > > > media_gt,
> > > > +
> > > > &fence[fence_id],
> > > > +
> > > > adj_start,
> > > > +
> > > > adj_end,
> > > > + vm-
> > > > > usm.asid);
> > > > + if (WARN_ON_ONCE(err < 0))
> > > > + goto wait;
> > > > + ++fence_id;
> > > > + }
> > > > + }
> > > > +
> > > > +wait:
> > > > + for (id = 0; id < fence_id; ++id)
> > > > + xe_gt_tlb_invalidation_fence_wait(&fence[id]);
> > > > +
> > > > +range_notifier_event_end:
> > > > + r = first;
> > > > + drm_gpusvm_for_each_range(r, notifier, adj_start,
> > > > adj_end)
> > > > + xe_svm_range_notifier_event_end(vm, r,
> > > > mmu_range);
> > > > }
> > > >
> > > > static const struct drm_gpusvm_ops gpusvm_ops = {
> > > > + .range_alloc = xe_svm_range_alloc,
> > > > + .range_free = xe_svm_range_free,
> > > > .invalidate = xe_svm_invalidate,
> > > > };
> > > >
> > > > @@ -36,6 +206,11 @@ int xe_svm_init(struct xe_vm *vm)
> > > >
> > > > void xe_svm_close(struct xe_vm *vm)
> > > > {
> > > > + xe_assert(vm->xe, xe_vm_is_closed(vm));
> > > > +
> > > > + /* Flush running notifiers making xe_vm_close() visable
> > > > */
> > > > + drm_gpusvm_notifier_lock(&vm->svm.gpusvm);
> > > > + drm_gpusvm_notifier_unlock(&vm->svm.gpusvm);
> > >
> > > Calling mmu_notifier_read_begin() ensures that nothing is
> > > invalidating
> > > on the range. Probably a better choice.
> > >
> >
> > We'd have to call that on every notifier rather than just cycle the
> > lock, so with that I'd prefer to leave it as is.
> >
> > > > }
> > > >
> > > > void xe_svm_fini(struct xe_vm *vm)
> > > > @@ -44,3 +219,31 @@ void xe_svm_fini(struct xe_vm *vm)
> > > >
> > > > drm_gpusvm_fini(&vm->svm.gpusvm);
> > > > }
> > > > +
> > > > +int xe_svm_handle_pagefault(struct xe_vm *vm, struct xe_vma
> > > > *vma,
> > > > + struct xe_tile *tile, u64
> > > > fault_addr,
> > > > + bool atomic)
> > > > +{
> > > > + struct drm_gpusvm_ctx ctx = { .read_only =
> > > > xe_vma_read_only(vma), };
> > > > + struct drm_gpusvm_range *r;
> > > > + int err;
> > > > +
> > > > + lockdep_assert_held_write(&vm->lock);
> > > > +
> > > > +retry:
> > > > + /* TODO: Run garbage collector */
> > > > +
> > > > + r = drm_gpusvm_range_find_or_insert(&vm->svm.gpusvm,
> > > > fault_addr,
> > > > + xe_vma_start(vma),
> > > > xe_vma_end(vma),
> > > > + &ctx);
> > > > + if (IS_ERR(r))
> > > > + return PTR_ERR(r);
> > > > +
> > > > + err = drm_gpusvm_range_get_pages(&vm->svm.gpusvm, r,
> > > > false);
> > > > + if (err == -EFAULT || err == -EPERM) /* Corner where
> > > > CPU
> > > > mappings have change */
> > >
> > > s/change/changed/
> > >
> >
> > Yep.
> >
> > > > + goto retry;
> > > > +
> > > > + /* TODO: Issue bind */
> > > > +
> > > > + return err;
> > > > +}
> > > > diff --git a/drivers/gpu/drm/xe/xe_svm.h
> > > > b/drivers/gpu/drm/xe/xe_svm.h
> > > > index 376e86876a11..c91c5f538024 100644
> > > > --- a/drivers/gpu/drm/xe/xe_svm.h
> > > > +++ b/drivers/gpu/drm/xe/xe_svm.h
> > > > @@ -6,14 +6,27 @@
> > > > #ifndef _XE_SVM_H_
> > > > #define _XE_SVM_H_
> > > >
> > > > +#include "drm_gpusvm.h"
> > > > #include "drm_pagemap.h"
> > > >
> > > > #define XE_INTERCONNECT_VRAM DRM_INTERCONNECT_DRIVER
> > >
> > > Not used yet
> > >
> >
> > Will remove.
> >
> > > >
> > > > +struct xe_tile;
> > > > struct xe_vm;
> > > > +struct xe_vma;
> > > > +
> > > > +struct xe_svm_range {
> > > > + struct drm_gpusvm_range base;
> > > > + u8 tile_present;
> > > > + u8 tile_invalidated;
> > > > +};
> > >
> > > Kerneldoc
> > >
> >
> > Will add.
> >
> > >
> > > >
> > > > int xe_svm_init(struct xe_vm *vm);
> > > > void xe_svm_fini(struct xe_vm *vm);
> > > > void xe_svm_close(struct xe_vm *vm);
> > > >
> > > > +int xe_svm_handle_pagefault(struct xe_vm *vm, struct xe_vma
> > > > *vma,
> > > > + struct xe_tile *tile, u64
> > > > fault_addr,
> > > > + bool atomic);
> > > > +
> > > > #endif
> > >
> > > Thanks,
> >
> > Thanks,
> > Matt
> >
> > > Thomas
> > >
>
More information about the Intel-xe
mailing list