Linux 7.2
There was a lot of conflicts this round between fixes and next,
and I'd like to get the merge resolutions that we have in drm-tip.
Signed-off-by: Dave Airlie <[email protected]>
Add ttm_bo_evict_cgroup() to evict buffer objects charged to a specific
dmem cgroup pool from a resource manager's LRU until a byte target is
met. Add ttm_resource_manager_set_dmem_region() to associate a dmem
cgroup region with a resource manager; drivers supply their own
dmem_cgroup_ops with ttm_resource_manager_dmem_reclaim as the reclaim
function and the manager pointer as reclaim_priv in the dmem_cgroup_init
to wire up TTM eviction as the reclaim callback.
The eviction context is interruptible; signals abort the operation and
propagate back through the write() syscall.
Introduce a new mode for the bo LRU walker so that sleeping locks
can be taken. This can be used when the caller doesn't hold any
previous dma_resv locks, and where it intends to hold at most
one lock at a time.
Like the rest of the TTM eviction this should sooner than later
be converted to full WW transactions.
v3:
- Fix ttm_resource_manager_set_dmem_region() storing an error pointer
in man->cg unconditionally. (Sashiko-bot)
- Fix kernel-doc function name format for ttm_bo_evict_cgroup() and
ttm_resource_manager_set_dmem_region().
v5:
- Rebased on the introduction of struct dmem_cgroup_init.
- Handle NULL region in ttm_resource_manager_set_dmem_region() to clear
the reclaim callback, preventing use-after-free when the manager is
torn down while the dmem region outlives it. (Sashiko-bot)
- Return 0 on any progress (even partial eviction), -ENOSPC only when
nothing was freed; fixes callers that expected 0 on partial success.
- Document that the reclaim callback should return 0 if some progress
was made, -ENOSPC if no progress at all, or another error for fatal
failures.
v8:
- Fix ttm_resource_manager_set_dmem_region() using IS_ERR_OR_NULL(),
which skipped the assignment for a NULL region and thus never
cleared man->cg. Use IS_ERR() so that a NULL region detaches the
region as the kernel-doc and the v5 changelog intended. (Sashiko-bot)
v9:
- Don't leak cgroup charges for bos that may have survived dmemcg
region fini.
- Drop the misleading "Capture size before eviction in case res is
cleared" comment in ttm_bo_evict_cb(). (Maarten Lankhorst)
Assisted-by: GitHub_Copilot:claude-sonnet-4.6
Signed-off-by: Thomas Hellström <[email protected]>
Reviewed-by: Maarten Lankhorst <[email protected]> #v7
Tested-by: Thadeu Lima de Souza Cascardo <[email protected]>
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Maarten Lankhorst <[email protected]>
Coupling resource allocation and cgroup charging is racy when charging
succeeds, but subsequent resource allocation fails. Certain eviction
decisions are made on the basis of whether the allocating cgroup is
protected, i.e. within its min/low limits, but with the charge being
tied to resource allocation (and uncharged when the resource allocation
fails), this check is done at a point where the allocation is not actually
charged to the cgroup.
This is subtly wrong if the allocation were to cause the cgroup to exceed
the min/low protection, but it's even more wrong if the same cgroup tries
allocating multiple buffers concurrently: In this case, the min/low
protection may pass for all allocation attempts when the real min/low
protection covers only some, or potentially none of the allocated
buffers.
Instead, charge the allocation to the cgroup once and keep the charge
for as long as we try to allocate a ttm_resource, and only undo the charge
if allocating the resource is ultimately unsuccessful and we move on to
a different ttm_place.
Reviewed-by: Maarten Lankhorst <[email protected]>
Reviewed-by: Timur Kristóf <[email protected]>
Signed-off-by: Natalie Vock <[email protected]>
Link: https://patch.msgid.link/[email protected]
ttm_pool_split_for_swap() unconditionally splits high-order pool pages
into order-0 pages before backup, so every compound the shrinker
touches is shattered even when the rest of the system would prefer it
stay intact. Under sustained kswapd pressure this fragments memory
enough to drive other parts of MM into recovery loops.
Back up each compound at its native order instead. In
ttm_pool_backup(), hand the full compound to the new
ttm_backup_backup_folio(), which backs up subpages to a contiguous
range of shmem indices and returns the base handle plus the number of
subpages actually backed up (@nr_backed). On full success, free the
compound once at its native order -- no split_page(), no per-4K
refcount juggling.
A per-folio backup can't be made fully atomic under memory pressure:
ttm_backup_backup_folio() must allocate shmem folios before source
subpages can be released, so under true OOM any subpage may fail
while the rest of the compound is still live. Two mechanisms handle
this without regressing reclaim behaviour:
- alloc_gfp gets __GFP_NOMEMALLOC whenever order > 0 (cleared again
for order-0), so a high-order backup fails fast with -ENOMEM
instead of draining kernel reserves, leaving them for other
allocations under the same pressure.
- If ttm_backup_backup_folio() still returns a short @nr_backed with
a valid handle for the successfully-backed prefix, split the
source compound with ttm_pool_split_for_swap(), free the prefix as
order-0 pages (already safely in shmem), and retry the remaining
subpages at order 0, where __GFP_NOMEMALLOC is cleared and
reserves may be used as a last resort.
This preserves the original split-on-OOM fallback while keeping the
common case fragmentation-free, and preserves the "partial backup is
allowed" contract (shrunken is incremented per subpage backed up).
The restore-side leftover-page split in ttm_pool_restore_commit() is
left as-is: it's unreachable in practice and not worth complicating
the restore state machine to avoid.
Testing: the existing backup_fault_inject point only truncated
tt->num_pages, which never exercised the reactive split path above
since it never left a compound partially backed up. Wire fault
injection into ttm_backup_backup_folio() itself: past the first
subpage of a compound, synthesize a -ENOMEM in place of
shmem_read_folio_gfp() when should_fail() trips, producing the same
short @nr_pages_backed a real failure would and forcing
ttm_pool_backup() through the split-and-retry path. The fault_attr
stays private to ttm_pool.c; ttm_backup.c reaches it through
ttm_backup_fault_inject_folio(), declared in ttm_pool_internal.h.
While converting the writeback branch to operate on the whole folio,
the unlock condition after shmem_writeout() also changed from `if
(ret)` to `if (ret == AOP_WRITEPAGE_ACTIVATE)`, matching the actual
contract: shmem_writeout()/swap_writeout() only leave the folio locked
when returning AOP_WRITEPAGE_ACTIVATE; any other return (including a
hard error from arch_prepare_to_swap()) means the folio was already
unlocked internally. The old `if (ret)` check would have double-
unlocked in that hard-error case.
Cc: Christian Koenig <[email protected]>
Cc: Huang Rui <[email protected]>
Cc: Matthew Auld <[email protected]>
Cc: Maarten Lankhorst <[email protected]>
Cc: Maxime Ripard <[email protected]>
Cc: Thomas Zimmermann <[email protected]>
Cc: David Airlie <[email protected]>
Cc: Simona Vetter <[email protected]>
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
Fixes: b63d715b80 ("drm/ttm/pool, drm/ttm/tt: Provide a helper to shrink pages")
Suggested-by: Thomas Hellström <[email protected]>
Assisted-by: Claude:claude-opus-4.6
Signed-off-by: Matthew Brost <[email protected]>
Reviewed-by: Thomas Hellström <[email protected]>
Link: https://patch.msgid.link/[email protected]
Boris Brezillion needs the gem lru fixes 379e8f1ca5 ("drm/gem: Make
the GEM LRU lock part of drm_device") backmerged for drm-misc-next.
That also means we need to sort out the rename conflict in panthor with
the fixup patch from Boris from drm-tip.
Signed-off-by: Simona Vetter <[email protected]>
fpfn and lpfn in struct ttm_place are 32-bit page numbers. With 4KB page
size this can support up to 44-bit physical addressing. Grow these to
64-bit (uint64_t) to support larger physical addresses.
Signed-off-by: Felix Kuehling <[email protected]>
Reviewed-by: Christian König <[email protected]>
Signed-off-by: Arunpravin Paneer Selvam <[email protected]>
Link: https://patch.msgid.link/[email protected]
When ttm_tt_swapout() fails, the current code calls
ttm_resource_add_bulk_move() followed by ttm_resource_move_to_lru_tail()
to restore the resource's bulk_move membership.
However, ttm_resource_move_to_lru_tail() places the resource at the tail
of the LRU list which, relative to the walk cursor's hitch node (placed
immediately after the resource when it was yielded), puts the resource
*in front of the* the hitch. The next list_for_each_entry_continue() from
the hitch finds the same resource again, causing an infinite loop.
Fix by deferring del_bulk_move to the success path only.
On the success path, TTM_TT_FLAG_SWAPPED has just been set by
ttm_tt_swapout() but the resource is still tracked in the bulk_move range,
so ttm_resource_del_bulk_move()'s !ttm_resource_unevictable() guard would
incorrectly skip the removal. Introduce
ttm_resource_del_bulk_move_unevictable() which bypasses that guard.
Reported-by: Jatin Kataria <[email protected]>
Fixes: fc5d96670e ("drm/ttm: Move swapped objects off the manager's LRU list")
Cc: Christian König <[email protected]>
Cc: Matthew Brost <[email protected]>
Cc: <[email protected]>
Cc: <[email protected]> # v6.13+
Assisted-by: GitHub_Copilot:claude-sonnet-4.6
Signed-off-by: Thomas Hellström <[email protected]>
Reviewed-by: Christian König <[email protected]>
Tested-by: Boqun Feng <[email protected]>
Link: https://patch.msgid.link/[email protected]
This is an initial port of the TTM pools for
write combined and uncached pages to use the list_lru.
This makes the pool's more NUMA aware and avoids
needing separate NUMA pools (later commit enables this).
Cc: Christian Koenig <[email protected]>
Cc: Johannes Weiner <[email protected]>
Cc: Dave Chinner <[email protected]>
Reviewed-by: Christian König <[email protected]>
Signed-off-by: Dave Airlie <[email protected]>
In situations where the system is very short on RAM, the shmem
readback from swap-space may invoke the OOM killer.
However, since this might be a recoverable situation where the caller
is indicating this by setting
struct ttm_operation_ctx::gfp_retry_mayfail to true, adjust the gfp
value used by the allocation accordingly.
Signed-off-by: Thomas Hellström <[email protected]>
Reviewed-by: Maarten Lankhorst <[email protected]>
Acked-by: Christian König <[email protected]>
Link: https://patch.msgid.link/[email protected]
Until now ttm stored a single pipelined eviction fence which means
drivers had to use a single entity for these evictions.
To lift this requirement, this commit allows up to 8 entities to
be used.
Ideally a dma_resv object would have been used as a container of
the eviction fences, but the locking rules makes it complex.
dma_resv all have the same ww_class, which means "Attempting to
lock more mutexes after ww_acquire_done." is an error.
One alternative considered was to introduced a 2nd ww_class for
specific resv to hold a single "transient" lock (= the resv lock
would only be held for a short period, without taking any other
locks).
The other option, is to statically reserve a fence array, and
extend the existing code to deal with N fences, instead of 1.
The driver is still responsible to reserve the correct number
of fence slots.
Signed-off-by: Pierre-Eric Pelloux-Prayer <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Reviewed-by: Christian König <[email protected]>
Signed-off-by: Christian König <[email protected]>
Some graphics APIs differentiate between out-of-graphics-memory and
out-of-host-memory (system memory). Add a device init flag to have -ENOSPC
propagated from the resource managers instead of being converted to
-ENOMEM, to aid driver stacks in determining what error code to return or
whether corrective action can be taken at the driver level.
Co-developed-by: Thomas Hellström <[email protected]>
Cc: Christian König <[email protected]>
Cc: Matthew Brost <[email protected]>
Signed-off-by: Tvrtko Ursulin <[email protected]>
Reviewed-by: Thomas Hellström <[email protected]>
Reviewed-by: Christian König <[email protected]>
Signed-off-by: Tvrtko Ursulin <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
GPUs typically benefit from contiguous memory via reduced TLB pressure and
improved caching performance, where the maximum size of contiguous block
which adds a performance benefit is related to hardware design.
TTM pool allocator by default tries (hard) to allocate up to the system
MAX_PAGE_ORDER blocks. This varies by the CPU platform and can also be
configured via Kconfig.
If that limit was set to be higher than the GPU can make an extra use of,
lets allow the individual drivers to let TTM know over which allocation
order can the pool allocator afford to make a little bit less effort with.
We implement this by disabling direct reclaim for those allocations, which
reduces the allocation latency and lowers the demands on the page
allocator, in cases where expending this effort is not critical for the
GPU in question.
Signed-off-by: Tvrtko Ursulin <[email protected]>
Cc: Christian König <[email protected]>
Cc: Thadeu Lima de Souza Cascardo <[email protected]>
Reviewed-by: Christian König <[email protected]>
Signed-off-by: Tvrtko Ursulin <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Give TTM BOs a separate cleanup function.
No funktional change, but the next step in removing the TTM BO reference
counting and replacing it with the GEM object reference counting.
v2: move the code around a bit to make it clearer what's happening
v3: fix nouveau_bo_fini as well
Signed-off-by: Christian König <[email protected]>
Acked-by: Thomas Hellström <[email protected]>
Acked-by: Joonas Lahtinen <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
While discussing cgroups we noticed a problem where you could export
a BO to a dma-buf without having it ever being backed or accounted for.
This meant in low memory situations or eventually with cgroups, a
lower privledged process might cause the compositor to try and allocate
a lot of memory on it's behalf and this could fail. At least make
sure the exporter has managed to allocate the RAM at least once
before exporting the object.
This only applies currently to TTM_PL_SYSTEM objects, because
GTT objects get populated on first validate, and VRAM doesn't
use TT.
Reviewed-by: Christian Koenig <[email protected]>
Cc: Thomas Hellström <[email protected]>
Cc: Simona Vetter <[email protected]>
Signed-off-by: Dave Airlie <[email protected]>
Reviewed-by: Thomas Hellström <[email protected]>
Signed-off-by: Dave Airlie <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
drm/i915 feature pull #2 for v6.17:
Features and functionality:
- Add drm_panic support for both i915 and xe drivers (Jocelyn Falempe)
- Add initial flip queue implementation, disabled by default, for LNL and PTL
(Ville)
- Add support for Wildcat Lake (WCL) display, version 30.02 (Matt Roper, Matt
Atwood, Dnyaneshwar)
- Extend drm_panel and follower support to DDI eDP (Arun)
Refactoring and cleanups:
- Make all global state objects opaque (Jani)
- Move display works to display specific unordered workqueue (Luca)
- Add and use struct drm_device based pcode interface (Jani, Lucas)
- Use clamp() instead of max()+min() combo (Ankit)
- Simplify wait for power well disable (Jani)
- Various stylistics cleanups and renames (Jani)
Fixes:
- Deal with loss of pipe DMC state (Ville)
- Fix PTL HDCP2 stream status check (Suraj)
- Add workaround for ADL-P DKL PHY DP and HDMI (Nemesa)
- Fix skl_print_wm_changes() stack usage with KMSAN (Arnd Bergmann)
- Fix PCON capability reads on non-branch devices (Chaitanya)
- Fix which platforms have ultra joiner (Ankit)
DRM core changes:
- Add ttm_bo_kmap_try_from_panic() for xe drm_panic support (Jocelyn Falempe)
- Add private pointer to struct drm_scanout buffer for xe/i915 drm_panic support
(Jocelyn Falempe)
Merges:
- Backmerge drm-next for drm_panel and xe changes (Jani)
Signed-off-by: Simona Vetter <[email protected]>
From: Jani Nikula <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
If the ttm bo is backed by pages, then it's possible to safely kmap
one page at a time, using kmap_try_from_panic().
Unfortunately there is no way to do the same with ioremap, so it
only supports the kmap case.
This is needed for proper drm_panic support with xe driver.
Signed-off-by: Jocelyn Falempe <[email protected]>
Reviewed-by: Christian König <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Maarten Lankhorst <[email protected]>
To avoid duplicating the tricky bo locking implementation,
Implement ttm_lru_walk_for_evict() using the guarded bo LRU iteration.
To facilitate this, support ticketlocking from the guarded bo LRU
iteration.
v2:
- Clean up some static function interfaces (Christian König)
- Fix Handling -EALREADY from ticketlocking in the loop by
skipping to the next item. (Intel CI)
Signed-off-by: Thomas Hellström <[email protected]>
Reviewed-by: Christian König <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Instead of the struct ttm_operation_ctx, Pass a struct ttm_lru_walk_arg
to enable us to easily extend the walk functionality, and to
implement ttm_lru_walk_for_evict() using the guarded LRU iteration.
Signed-off-by: Thomas Hellström <[email protected]>
Reviewed-by: Christian König <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Pull drm updates from Dave Airlie:
"As part of building up nova-core/nova-drm pieces we've brought in some
rust abstractions through this tree, aux bus being the main one, with
devres changes also in the driver-core tree. Along with the drm core
abstractions and enough nova-core/nova-drm to use them. This is still
all stub work under construction, to build the nova driver upstream.
The other big NVIDIA related one is nouveau adds support for
Hopper/Blackwell GPUs, this required a new GSP firmware update to
570.144, and a bunch of rework in order to support multiple fw
interfaces.
There is also the introduction of an asahi uapi header file as a
precursor to getting the real driver in later, but to unblock
userspace mesa packages while the driver is trapped behind rust
enablement.
Otherwise it's the usual mixture of stuff all over, amdgpu, i915/xe,
and msm being the main ones, and some changes to vsprintf.
new drivers:
- bring in the asahi uapi header standalone
- nova-drm: stub driver
rust dependencies (for nova-core):
- auxiliary
- bus abstractions
- driver registration
- sample driver
- devres changes from driver-core
- revocable changes
core:
- add Apple fourcc modifiers
- add virtio capset definitions
- extend EXPORT_SYNC_FILE for timeline syncobjs
- convert to devm_platform_ioremap_resource
- refactor shmem helper page pinning
- DP powerup/down link helpers
- extended %p4cc in vsprintf.c to support fourcc prints
- change vsprintf %p4cn to %p4chR, remove %p4cn
- Add drm_file_err function
- IN_FORMATS_ASYNC property
- move sitronix from tiny to their own subdir
rust:
- add drm core infrastructure rust abstractions
(device/driver, ioctl, file, gem)
dma-buf:
- adjust sg handling to not cache map on attach
- allow setting dma-device for import
- Add a helper to sort and deduplicate dma_fence arrays
docs:
- updated drm scheduler docs
- fbdev todo update
- fb rendering
- actual brightness
ttm:
- fix delayed destroy resv object
bridge:
- add kunit tests
- convert tc358775 to atomic
- convert drivers to devm_drm_bridge_alloc
- convert rk3066_hdmi to bridge driver
scheduler:
- add kunit tests
panel:
- refcount panels to improve lifetime handling
- Powertip PH128800T004-ZZA01
- NLT NL13676BC25-03F, Tianma TM070JDHG34-00
- Himax HX8279/HX8279-D DDIC
- Visionox G2647FB105
- Sitronix ST7571
- ZOTAC rotation quirk
vkms:
- allow attaching more displays
i915:
- xe3lpd display updates
- vrr refactor
- intel_display struct conversions
- xe2hpd memory type identification
- add link rate/count to i915_display_info
- cleanup VGA plane handling
- refactor HDCP GSC
- fix SLPC wait boosting reference counting
- add 20ms delay to engine reset
- fix fence release on early probe errors
xe:
- SRIOV updates
- BMG PCI ID update
- support separate firmware for each GT
- SVM fix, prelim SVM multi-device work
- export fan speed
- temp disable d3cold on BMG
- backup VRAM in PM notifier instead of suspend/freeze
- update xe_ttm_access_memory to use GPU for non-visible access
- fix guc_info debugfs for VFs
- use copy_from_user instead of __copy_from_user
- append PCIe gen5 limitations to xe_firmware document
amdgpu:
- DSC cleanup
- DC Scaling updates
- Fused I2C-over-AUX updates
- DMUB updates
- Use drm_file_err in amdgpu
- Enforce isolation updates
- Use new dma_fence helpers
- USERQ fixes
- Documentation updates
- SR-IOV updates
- RAS updates
- PSP 12 cleanups
- GC 9.5 updates
- SMU 13.x updates
- VCN / JPEG SR-IOV updates
amdkfd:
- Update error messages for SDMA
- Userptr updates
- XNACK fixes
radeon:
- CIK doorbell cleanup
nouveau:
- add support for NVIDIA r570 GSP firmware
- enable Hopper/Blackwell support
nova-core:
- fix task list
- register definition infrastructure
- move firmware into own rust module
- register auxiliary device for nova-drm
nova-drm:
- initial driver skeleton
msm:
- GPU:
- ACD (adaptive clock distribution) for X1-85
- drop fictional address_space_size
- improve GMU HFI response time out robustness
- fix crash when throttling during boot
- DPU:
- use single CTL path for flushing on DPU 5.x+
- improve SSPP allocation code for better sharing
- Enabled SmartDMA on SM8150, SC8180X, SC8280XP, SM8550
- Added SAR2130P support
- Disabled DSC support on MSM8937, MSM8917, MSM8953, SDM660
- DP:
- switch to new audio helpers
- better LTTPR handling
- DSI:
- Added support for SA8775P
- Added SAR2130P support
- HDMI:
- Switched to use new helpers for ACR data
- Fixed old standing issue of HPD not working in some cases
amdxdna:
- add dma-buf support
- allow empty command submits
renesas:
- add dma-buf support
- add zpos, alpha, blend support
panthor:
- fail properly for NO_MMAP bos
- add SET_LABEL ioctl
- debugfs BO dumping support
imagination:
- update DT bindings
- support TI AM68 GPU
hibmc:
- improve interrupt handling and HPD support
virtio:
- add panic handler support
rockchip:
- add RK3588 support
- add DP AUX bus panel support
ivpu:
- add heartbeat based hangcheck
mediatek:
- prepares support for MT8195/99 HDMIv2/DDCv2
anx7625:
- improve HPD
tegra:
- speed up firmware loading
* tag 'drm-next-2025-05-28' of https://gitlab.freedesktop.org/drm/kernel: (1627 commits)
drm/nouveau/tegra: Fix error pointer vs NULL return in nvkm_device_tegra_resource_addr()
drm/xe: Default auto_link_downgrade status to false
drm/xe/guc: Make creation of SLPC debugfs files conditional
drm/i915/display: Add check for alloc_ordered_workqueue() and alloc_workqueue()
drm/i915/dp_mst: Work around Thunderbolt sink disconnect after SINK_COUNT_ESI read
drm/i915/ptl: Use everywhere the correct DDI port clock select mask
drm/nouveau/kms: add support for GB20x
drm/dp: add option to disable zero sized address only transactions.
drm/nouveau: add support for GB20x
drm/nouveau/gsp: add hal for fifo.chan.doorbell_handle
drm/nouveau: add support for GB10x
drm/nouveau/gf100-: track chan progress with non-WFI semaphore release
drm/nouveau/nv50-: separate CHANNEL_GPFIFO handling out from CHANNEL_DMA
drm/nouveau: add helper functions for allocating pinned/cpu-mapped bos
drm/nouveau: add support for GH100
drm/nouveau: improve handling of 64-bit BARs
drm/nouveau/gv100-: switch to volta semaphore methods
drm/nouveau/gsp: support deeper page tables in COPY_SERVER_RESERVED_PDES
drm/nouveau/gsp: init client VMMs with NV0080_CTRL_DMA_SET_PAGE_DIRECTORY
drm/nouveau/gsp: fetch level shift and PDE from BAR2 VMM
...
This flag used to be used in the old memory tracking code, that
code got migrated into the vmwgfx driver[1], and then got removed
from the tree[2], but this piece got left behind.
[1] f07069da6b ("drm/ttm: move memory accounting into vmwgfx v4")
[2] 8aadeb8ad8 ("drm/vmwgfx: Remove the dedicated memory accounting")
Cleanup the dead code.
Reviewed-by: Matthew Brost <[email protected]>
Signed-off-by: Dave Airlie <[email protected]>
Add a number of helpers for shrinking that access core TTM and
core MM functionality in a way that make them unsuitable for
driver open-coding.
v11:
- New patch (split off from previous) and additional helpers.
v13:
- Adapt to ttm_backup interface change.
- Take resource off LRU when backed up.
Signed-off-by: Thomas Hellström <[email protected]>
Reviewed-by: Matthew Brost <[email protected]>
Acked-by: Dave Airlie <[email protected]>
Acked-by: Christian König <[email protected]>
Link: https://lore.kernel.org/intel-xe/[email protected]
Following the design direction communicated here:
https://lore.kernel.org/linux-mm/[email protected]/T/#ma918844aa8a6efe8768fdcda0c6590d5c93850c9
Export a LRU walker for driver shrinker use. The walker
initially supports only trylocking, since that's the
method used by shrinkes. The walker makes use of
scoped_guard() to allow exiting from the LRU walk loop
without performing any explicit unlocking or
cleanup.
v8:
- Split out from another patch.
- Use a struct for bool arguments to increase readability (Matt Brost).
- Unmap user-space cpu-mappings before shrinking pages.
- Explain non-fatal error codes (Matt Brost)
v10:
- Instead of using the existing helper, Wrap the interface inside out and
provide a loop to de-midlayer things the LRU iteration (Christian König).
- Removing the R-B by Matt Brost since the patch was significantly changed.
v11:
- Split the patch up to include just the LRU walk helper.
v12:
- Indent after scoped_guard() (Matt Brost)
v15:
- Adapt to new definition of scoped_guard()
Signed-off-by: Thomas Hellström <[email protected]>
Reviewed-by: Matthew Brost <[email protected]>
Acked-by: Dave Airlie <[email protected]>
Acked-by: Christian König <[email protected]>
Link: https://lore.kernel.org/intel-xe/[email protected]
Provide a helper to shrink ttm_tt page-vectors on a per-page
basis. A ttm_backup backend could then in theory get away with
allocating a single temporary page for each struct ttm_tt.
This is accomplished by splitting larger pages before trying to
back them up.
In the future we could allow ttm_backup to handle backing up
large pages as well, but currently there's no benefit in
doing that, since the shmem backup backend would have to
split those anyway to avoid allocating too much temporary
memory, and if the backend instead inserts pages into the
swap-cache, those are split on reclaim by the core.
Due to potential backup- and recover errors, allow partially swapped
out struct ttm_tt's, although mark them as swapped out stopping them
from being swapped out a second time. More details in the ttm_pool.c
DOC section.
v2:
- A couple of cleanups and error fixes in ttm_pool_back_up_tt.
- s/back_up/backup/
- Add a writeback parameter to the exported interface.
v8:
- Use a struct for flags for readability (Matt Brost)
- Address misc other review comments (Matt Brost)
v9:
- Update the kerneldoc for the ttm_tt::backup field.
v10:
- Rebase.
v13:
- Rebase on ttm_backup interface change. Update kerneldoc.
- Rebase and adjust ttm_tt_is_swapped().
v15:
- Rebase on ttm_backup return value change.
- Rebase on previous restructuring of ttm_pool_alloc()
- Rework the ttm_pool backup interface (Christian König)
- Remove cond_resched() (Christian König)
- Get rid of the need to allocate an intermediate page array
when restoring a multi-order page (Christian König)
- Update documentation.
Cc: Christian König <[email protected]>
Cc: Somalapuram Amaranath <[email protected]>
Cc: Matthew Brost <[email protected]>
Cc: <[email protected]>
Signed-off-by: Thomas Hellström <[email protected]>
Reviewed-by: Matthew Brost <[email protected]>
Acked-by: Christian Koenig <[email protected]>
Link: https://lore.kernel.org/intel-xe/[email protected]
Provide a standalone shmem backup implementation.
Given the ttm_backup interface, this could
later on be extended to providing other backup
implementation than shmem, with one use-case being
GPU swapout to a user-provided fd.
v5:
- Fix a UAF. (kernel test robot, Dan Carptenter)
v6:
- Rename ttm_backup_shmem_copy_page() function argument
(Matthew Brost)
- Add some missing documentation
v8:
- Use folio_file_page to get to the page we want to writeback
instead of using the first page of the folio.
v13:
- Remove the base class abstraction (Christian König)
- Include ttm_backup_bytes_avail().
v14:
- Fix kerneldoc for ttm_backup_bytes_avail() (0-day)
- Work around casting of __randomize_layout struct pointer (0-day)
v15:
- Return negative error code from ttm_backup_backup_page()
(Christian König)
- Doc fixes. (Christian König).
Cc: Christian König <[email protected]>
Cc: Somalapuram Amaranath <[email protected]>
Cc: Matthew Brost <[email protected]>
Cc: <[email protected]>
Signed-off-by: Thomas Hellström <[email protected]>
Reviewed-by: Matthew Brost <[email protected]>
Reviewed-by: Christian König <[email protected]>
Link: https://lore.kernel.org/intel-xe/[email protected]
cgroup resource allocation has to be handled in TTM, so -EAGAIN from
cgroups can be converted into -ENOSPC, and the limitcg can be properly
evicted in ttm code.
When hitting a resource limit through -EAGAIN, the cgroup for which the
limit is hit is also returned. This allows eviction to delete only from
cgroups which are a subgroup of the current cgroup.
The returned CSS is used to determine if eviction is valuable for a
given resource, and allows TTM to only target specific resources to
lower memory usage.
Co-developed-by: Friedrich Vock <[email protected]>
Signed-off-by: Friedrich Vock <[email protected]>
Co-developed-by: Maxime Ripard <[email protected]>
Signed-off-by: Maarten Lankhorst <[email protected]>
Acked-by: Tejun Heo <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Maxime Ripard <[email protected]>
Non-contiguous VRAM cannot easily be mapped in TTM nor can non-visible
VRAM easily be accessed. Add ttm_bo_access, which is similar to
ttm_bo_vm_access, to access such memory.
v4:
- Fix checkpatch warnings (CI)
v5:
- Fix checkpatch warnings (CI)
v6:
- Fix kernel doc (Auld)
v7:
- Move ttm_bo_access to ttm_bo_vm.c (Christian)
Cc: Christian König <[email protected]>
Reported-by: Christoph Manszewski <[email protected]>
Suggested-by: Thomas Hellström <[email protected]>
Signed-off-by: Matthew Brost <[email protected]>
Tested-by: Mika Kuoppala <[email protected]>
Reviewed-by: Matthew Auld <[email protected]>
Reviewed-by: Christian König <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
Resources of swapped objects remains on the TTM_PL_SYSTEM manager's
LRU list, which is bad for the LRU walk efficiency.
Rename the device-wide "pinned" list to "unevictable" and move
also resources of swapped-out objects to that list.
An alternative would be to create an "UNEVICTABLE" priority to
be able to keep the pinned- and swapped objects on their
respective manager's LRU without affecting the LRU walk efficiency.
v2:
- Remove a bogus WARN_ON (Christian König)
- Update ttm_resource_[add|del] bulk move (Christian König)
- Fix TTM KUNIT tests (Intel CI)
v3:
- Check for non-NULL bo->resource in ttm_bo_populate().
v4:
- Don't move to LRU tail during swapout until the resource
is properly swapped or there was a swapout failure.
(Intel Ci)
- Add a newline after checkpatch check.
v5:
- Introduce ttm_resource_is_swapped() to avoid a corner-case where
a newly created resource was considered swapped. (Intel CI)
v6:
- Move an assert.
Cc: Christian König <[email protected]>
Cc: Matthew Brost <[email protected]>
Cc: <[email protected]>
Signed-off-by: Thomas Hellström <[email protected]>
Reviewed-by: Christian König <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
Thomas needs 5a498d4d06 ("drm/fbdev-dma: Only install deferred I/O
if necessary") in drm-misc, so start the backmerge cascade.
Signed-off-by: Simona Vetter <[email protected]>
Use the LRU walker for eviction. This helps
removing a lot of code with weird locking
semantics.
The functionality is slightly changed so that
when trylocked buffer objects are exhausted, we
continue to interleave walks with ticket-locks while
there is still progress made. The list walks are
not restarted in-between evictions.
Also provide a separate ttm_bo_evict_first()
function for its single user. The context of that
user allows sleeping dma_resv locks.
v6:
- Various cleanups suggested by Matthew Brost.
- Fix error return code of ttm_bo_evict_first(). (Matthew Brost)
- Fix an error check that was inverted. (Matthew Brost)
v7:
- Use s64 rather than long (Christian König)
- Early ttm_resource_cursor_fini() in ttm_bo_evict_first().
- Simplify check for bo_moved in ttm_bo_evict_first().
(Christian König)
- Don't evict pinned bos.
Cc: Christian König <[email protected]>
Cc: Somalapuram Amaranath <[email protected]>
Cc: Matthew Brost <[email protected]>
Cc: <[email protected]>
Signed-off-by: Thomas Hellström <[email protected]>
Reviewed-by: Matthew Brost <[email protected]> #v6
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
Signed-off-by: Christian König <[email protected]>
Dave Airlie