The src_owner field in struct migrate_vma is being used for two purposes,
it acts as a selection filter for which types of pages are to be migrated
and it identifies device private pages owned by the caller. Split this
into separate parameters so the src_owner field can be used just to
identify
The goal for this series is to avoid device private memory TLB
invalidations when migrating a range of addresses from system
memory to device private memory and some of those pages have already
been migrated. The approach taken is to introduce a new mmu notifier
invalidation event type and use
Use the new MMU_NOTIFY_MIGRATE event to skip GPU MMU invalidations of
device private memory and handle the invalidation in the driver as part
of migrating device private memory.
Signed-off-by: Ralph Campbell
---
drivers/gpu/drm/nouveau/nouveau_dmem.c | 15 ---
When migrating a range of system memory to device private memory, some
of the pages in the address range may not be migrating. In this case,
the non migrating pages won't have a new GPU MMU entry to store but
the nvif_object_ioctl() NVIF_VMM_V0_PFNMAP method doesn't check the input
and stores a
Currently migrate_vma_setup() calls mmu_notifier_invalidate_range_start()
which flushes all device private page mappings whether or not a page
is being migrated to/from device private memory. In order to not disrupt
device mappings that are not being migrated, shift the responsibility
for clearing
Use the new MMU_NOTIFY_MIGRATE event to skip MMU invalidations of device
private memory and handle the invalidation in the driver as part of
migrating device private memory.
Signed-off-by: Ralph Campbell
---
lib/test_hmm.c | 30 +++---
On Tue, 2020-07-21 at 12:00 -0400, Lyude Paul wrote:
> On Tue, 2020-07-21 at 18:27 +0300, Mika Westerberg wrote:
> > On Tue, Jul 21, 2020 at 11:01:55AM -0400, Lyude Paul wrote:
> > > Sure thing. Also, feel free to let me know if you'd like access to one
> > > of
> > > the
> > > systems we saw
On Tue, 2020-07-21 at 18:27 +0300, Mika Westerberg wrote:
> On Tue, Jul 21, 2020 at 11:01:55AM -0400, Lyude Paul wrote:
> > Sure thing. Also, feel free to let me know if you'd like access to one of
> > the
> > systems we saw breaking with this patch - I'm fairly sure I've got one of
> > them
> >
Reviewed-by: Lyude Paul
Thanks!
On Tue, 2020-07-21 at 15:17 +, Wei Yongjun wrote:
> When using single_open() for opening, single_release() should be
> used instead of seq_release(), otherwise there is a memory leak.
>
> Fixes: 12885ecbfe62 ("drm/nouveau/kms/nvd9-: Add CRC support")
>
On Tue, Jul 21, 2020 at 11:01:55AM -0400, Lyude Paul wrote:
> Sure thing. Also, feel free to let me know if you'd like access to one of the
> systems we saw breaking with this patch - I'm fairly sure I've got one of them
> locally at my apartment and don't mind setting up AMT/KVM/SSH
Probably no
Sure thing. Also, feel free to let me know if you'd like access to one of the
systems we saw breaking with this patch - I'm fairly sure I've got one of them
locally at my apartment and don't mind setting up AMT/KVM/SSH
On Tue, 2020-07-21 at 15:22 +0300, Mika Westerberg wrote:
> Hi,
>
> [Sorry
On Fri, Jul 17, 2020 at 09:52:09PM +0200, Lukas Wunner wrote:
> On Fri, Jul 17, 2020 at 03:04:10PM -0400, Lyude Paul wrote:
> > Isn't it possible to tell whether a PCI device is connected through
> > thunderbolt or not? We could probably get away with just defaulting
> > to 100ms for thunderbolt
Hi,
[Sorry for the delay, I was on vacation]
On Fri, Jul 17, 2020 at 01:32:10PM +0200, Karol Herbst wrote:
> Filed at https://bugzilla.kernel.org/show_bug.cgi?id=208597
Thanks for reporting.
I'll check your logs and try to figure if there is something we can do
to make both nouveau and TBT
13 matches
Mail list logo