When running igt@gem_exec_balancer@individual for multiple iterations,
it is seen that the delta busyness returned by PMU is 0. The issue stems
from a combination of 2 implementation specific details:
1) gt_park is throttling __update_guc_busyness_stats() so that it does
not hog PCI bandwidth for some use cases. (Ref: 59bcdb564b)
2) busyness implementation always returns monotonically increasing
counters. (Ref: cf907f6d29)
If an application queried an engine while it was active,
engine->stats.guc.running is set to true. Following that, if all PM
wakeref's are released, then gt is parked. At this time the throttling
of __update_guc_busyness_stats() may result in a missed update to the
running state of the engine (due to (1) above). This means subsequent
calls to guc_engine_busyness() will think that the engine is still
running and they will keep updating the cached counter (stats->total).
This results in an inflated cached counter.
Later when the application runs a workload and queries for busyness, we
return the cached value since it is larger than the actual value (due to
(2) above)
All subsequent queries will return the same large (inflated) value, so
the application sees a delta busyness of zero.
Fix the issue by resetting the running state of engines each time
intel_guc_busyness_park() is called.
v2: (Rodrigo)
- Use the correct tag in commit message
- Drop the redundant wakeref check in guc_engine_busyness() and update
commit message
Closes: https://gitlab.freedesktop.org/drm/i915/kernel/-/issues/13366
Fixes: cf907f6d29 ("i915/guc: Ensure busyness counter increases motonically")
Signed-off-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250123193839.2394694-1-umesh.nerlige.ramappa@intel.com
(cherry picked from commit 431b742e2b)
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Pull kthread updates from Frederic Weisbecker:
"Kthreads affinity follow either of 4 existing different patterns:
1) Per-CPU kthreads must stay affine to a single CPU and never
execute relevant code on any other CPU. This is currently handled
by smpboot code which takes care of CPU-hotplug operations.
Affinity here is a correctness constraint.
2) Some kthreads _have_ to be affine to a specific set of CPUs and
can't run anywhere else. The affinity is set through
kthread_bind_mask() and the subsystem takes care by itself to
handle CPU-hotplug operations. Affinity here is assumed to be a
correctness constraint.
3) Per-node kthreads _prefer_ to be affine to a specific NUMA node.
This is not a correctness constraint but merely a preference in
terms of memory locality. kswapd and kcompactd both fall into this
category. The affinity is set manually like for any other task and
CPU-hotplug is supposed to be handled by the relevant subsystem so
that the task is properly reaffined whenever a given CPU from the
node comes up. Also care should be taken so that the node affinity
doesn't cross isolated (nohz_full) cpumask boundaries.
4) Similar to the previous point except kthreads have a _preferred_
affinity different than a node. Both RCU boost kthreads and RCU
exp kworkers fall into this category as they refer to "RCU nodes"
from a distinctly distributed tree.
Currently the preferred affinity patterns (3 and 4) have at least 4
identified users, with more or less success when it comes to handle
CPU-hotplug operations and CPU isolation. Each of which do it in its
own ad-hoc way.
This is an infrastructure proposal to handle this with the following
API changes:
- kthread_create_on_node() automatically affines the created kthread
to its target node unless it has been set as per-cpu or bound with
kthread_bind[_mask]() before the first wake-up.
- kthread_affine_preferred() is a new function that can be called
right after kthread_create_on_node() to specify a preferred
affinity different than the specified node.
When the preferred affinity can't be applied because the possible
targets are offline or isolated (nohz_full), the kthread is affine to
the housekeeping CPUs (which means to all online CPUs most of the time
or only the non-nohz_full CPUs when nohz_full= is set).
kswapd, kcompactd, RCU boost kthreads and RCU exp kworkers have been
converted, along with a few old drivers.
Summary of the changes:
- Consolidate a bunch of ad-hoc implementations of
kthread_run_on_cpu()
- Introduce task_cpu_fallback_mask() that defines the default last
resort affinity of a task to become nohz_full aware
- Add some correctness check to ensure kthread_bind() is always
called before the first kthread wake up.
- Default affine kthread to its preferred node.
- Convert kswapd / kcompactd and remove their halfway working ad-hoc
affinity implementation
- Implement kthreads preferred affinity
- Unify kthread worker and kthread API's style
- Convert RCU kthreads to the new API and remove the ad-hoc affinity
implementation"
* tag 'kthread-for-6.14-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/frederic/linux-dynticks:
kthread: modify kernel-doc function name to match code
rcu: Use kthread preferred affinity for RCU exp kworkers
treewide: Introduce kthread_run_worker[_on_cpu]()
kthread: Unify kthread_create_on_cpu() and kthread_create_worker_on_cpu() automatic format
rcu: Use kthread preferred affinity for RCU boost
kthread: Implement preferred affinity
mm: Create/affine kswapd to its preferred node
mm: Create/affine kcompactd to its preferred node
kthread: Default affine kthread to its preferred NUMA node
kthread: Make sure kthread hasn't started while binding it
sched,arm64: Handle CPU isolation on last resort fallback rq selection
arm64: Exclude nohz_full CPUs from 32bits el0 support
lib: test_objpool: Use kthread_run_on_cpu()
kallsyms: Use kthread_run_on_cpu()
soc/qman: test: Use kthread_run_on_cpu()
arm/bL_switcher: Use kthread_run_on_cpu()
kthread_create() creates a kthread without running it yet. kthread_run()
creates a kthread and runs it.
On the other hand, kthread_create_worker() creates a kthread worker and
runs it.
This difference in behaviours is confusing. Also there is no way to
create a kthread worker and affine it using kthread_bind_mask() or
kthread_affine_preferred() before starting it.
Consolidate the behaviours and introduce kthread_run_worker[_on_cpu]()
that behaves just like kthread_run(). kthread_create_worker[_on_cpu]()
will now only create a kthread worker without starting it.
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Add RC6 & RC0 frequency printing to ensure accurate energy
readings aimed at addressing GPU energy leaks and power
measurement failures.
Also update sleep time for RC6 mode to match RC0.
v2:
- Improved commit message.
v3:
- Used pr_err log to display frequency (Anshuman)
- Sorted headers alphabetically (Sai Teja)
v4:
- Improved commit message.
- Fix pr_err log (Sai Teja)
v5:
- Add error & debug logging for RC0 power and frequency checks (Anshuman)
v6:
- Modify debug logging for RC0 power and frequency checks (Sai Teja)
v7:
- Use pr_debug if RC0 power isn't measured but frequency is (Anshuman)
- Improved commit message (Badal)
- Change API to read actual frequency without applying forcewake (Badal)
- Update sleep time for RC6 mode (Anshuman)
Signed-off-by: Sk Anirban <sk.anirban@intel.com>
Reviewed-by: Sai Teja Pottumuttu <sai.teja.pottumuttu@intel.com>
Reviewed-by: Badal Nilawar <badal.nilawar@intel.com>
Reviewed-by: Andi Shyti <andi.shyti@linux.intel.com>
Signed-off-by: Andi Shyti <andi.shyti@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241129154716.2764974-1-sk.anirban@intel.com
Active busyness of an engine is calculated using gt timestamp and the
context switch in time. While capturing the gt timestamp, it's possible
that the context switches out. This race could result in an active
busyness value that is greater than the actual context runtime value by a
small amount. This leads to a negative delta and throws off busyness
calculations for the user.
If a subsequent count is smaller than the previous one, just return the
previous one, since we expect the busyness to catch up.
Fixes: 77cdd054dd ("drm/i915/pmu: Connect engine busyness stats from GuC to pmu")
Signed-off-by: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
Reviewed-by: John Harrison <John.C.Harrison@Intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241127174006.190128-3-umesh.nerlige.ramappa@intel.com
Currently intel_dpt_resume() tries to blindly rewrite all the
PTEs for currently bound DPT VMAs. That is problematic because
the CPU mapping for the DPT is only really guaranteed to exist
while the DPT object has been pinned. In the past we worked
around this issue by making DPT objects unshrinkable, but that
is undesirable as it'll waste physical RAM.
Let's instead forcefully evict all the DPT VMAs on suspend,
thus guaranteeing that intel_dpt_resume() has nothing to do.
To guarantee that all the DPT VMAs are evictable by
intel_dpt_suspend() we need to flush the cleanup workqueue
after the display output has been shut down.
And for good measure throw in a few extra WARNs to catch
any mistakes.
Cc: Brian Geffon <bgeffon@google.com>
Cc: Vidya Srinivas <vidya.srinivas@intel.com>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241127061117.25622-4-ville.syrjala@linux.intel.com
Reviewed-by: Vidya Srinivas <vidya.srinivas@intel.com>
Tested-by: Vidya Srinivas <vidya.srinivas@intel.com>
Pull drm updates from Dave Airlie:
"There's a lot of rework, the panic helper support is being added to
more drivers, v3d gets support for HW superpages, scheduler
documentation, drm client and video aperture reworks, some new
MAINTAINERS added, amdgpu has the usual lots of IP refactors, Intel
has some Pantherlake enablement and xe is getting some SRIOV bits, but
just lots of stuff everywhere.
core:
- split DSC helpers from DP helpers
- clang build fixes for drm/mm test
- drop simple pipeline support for gem vram
- document submission error signaling
- move drm_rect to drm core module from kms helper
- add default client setup to most drivers
- move to video aperture helpers instead of drm ones
tests:
- new framebuffer tests
ttm:
- remove swapped and pinned BOs from TTM lru
panic:
- fix uninit spinlock
- add ABGR2101010 support
bridge:
- add TI TDP158 support
- use standard PM OPS
dma-fence:
- use read_trylock instead of read_lock to help lockdep
scheduler:
- add errno to sched start to report different errors
- add locking to drm_sched_entity_modify_sched
- improve documentation
xe:
- add drm_line_printer
- lots of refactoring
- Enable Xe2 + PES disaggregation
- add new ARL PCI ID
- SRIOV development work
- fix exec unnecessary implicit fence
- define and parse OA sync props
- forcewake refactoring
i915:
- Enable BMG/LNL ultra joiner
- Enable 10bpx + CCS scanout on ICL+, fp16/CCS on TGL+
- use DSB for plane/color mgmt
- Arrow lake PCI IDs
- lots of i915/xe display refactoring
- enable PXP GuC autoteardown
- Pantherlake (PTL) Xe3 LPD display enablement
- Allow fastset HDR infoframe changes
- write DP source OUI for non-eDP sinks
- share PCI IDs between i915 and xe
amdgpu:
- SDMA queue reset support
- SMU 13.0.6, JPEG 4.0.3 updates
- Initial runtime repartitioning support
- rework IP structs for multiple IP instances
- Fetch EDID from _DDC if available
- SMU13 zero rpm user control
- lots of fixes/cleanups
amdkfd:
- Increase event FIFO size
- add topology cap flag for per queue reset
msm:
- DPU:
- SA8775P support
- (disabled by default) MSM8917, MSM8937, MSM8953 and MSM8996 support
- Enable large framebuffer support
- Drop MSM8998 and SDM845
- DP:
- SA8775P support
- GPU:
- a7xx preemption support
- Adreno A663 support
ast:
- warn about unsupported TX chips
ivpu:
- add coredump
- add pantherlake support
rockchip:
- 4K@60Hz display enablement
- generate pll programming tables
panthor:
- add timestamp query API
- add realtime group priority
- add fdinfo support
etnaviv:
- improve handling of DMA address limits
- improve GPU hangcheck
exynos:
- Decon Exynos7870 support
mediatek:
- add OF graph support
omap:
- locking fixes
bochs:
- convert to gem/shmem from simpledrm
v3d:
- support big/super pages
- add gemfs
vc4:
- BCM2712 support refactoring
- add YUV444 format support
udmabuf:
- folio related fixes
nouveau:
- add panic support on nv50+"
* tag 'drm-next-2024-11-21' of https://gitlab.freedesktop.org/drm/kernel: (1583 commits)
drm/xe/guc: Fix dereference before NULL check
drm/amd: Fix initialization mistake for NBIO 7.7.0
Revert "drm/amd/display: parse umc_info or vram_info based on ASIC"
drm/amd/display: Fix failure to read vram info due to static BP_RESULT
drm/amdgpu: enable GTT fallback handling for dGPUs only
drm/amd/amdgpu: limit single process inside MES
drm/fourcc: add AMD_FMT_MOD_TILE_GFX9_4K_D_X
drm/amdgpu/mes12: correct kiq unmap latency
drm/amdgpu: Support vcn and jpeg error info parsing
drm/amd : Update MES API header file for v11 & v12
drm/amd/amdkfd: add/remove kfd queues on start/stop KFD scheduling
drm/amdkfd: change kfd process kref count at creation
drm/amdgpu: Cleanup shift coding style
drm/amd/amdgpu: Increase MES log buffer to dump mes scratch data
drm/amdgpu: Implement virt req_ras_err_count
drm/amdgpu: VF Query RAS Caps from Host if supported
drm/amdgpu: Add msg handlers for SRIOV RAS Telemetry
drm/amdgpu: Update SRIOV Exchange Headers for RAS Telemetry Support
drm/amd/display: 3.2.309
drm/amd/display: Adjust VSDB parser for replay feature
...
Pull vdso data page handling updates from Thomas Gleixner:
"First steps of consolidating the VDSO data page handling.
The VDSO data page handling is architecture specific for historical
reasons, but there is no real technical reason to do so.
Aside of that VDSO data has become a dump ground for various
mechanisms and fail to provide a clear separation of the
functionalities.
Clean this up by:
- consolidating the VDSO page data by getting rid of architecture
specific warts especially in x86 and PowerPC.
- removing the last includes of header files which are pulling in
other headers outside of the VDSO namespace.
- seperating timekeeping and other VDSO data accordingly.
Further consolidation of the VDSO page handling is done in subsequent
changes scheduled for the next merge window.
This also lays the ground for expanding the VDSO time getters for
independent PTP clocks in a generic way without making every
architecture add support seperately"
* tag 'timers-vdso-2024-11-18' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (42 commits)
x86/vdso: Add missing brackets in switch case
vdso: Rename struct arch_vdso_data to arch_vdso_time_data
powerpc: Split systemcfg struct definitions out from vdso
powerpc: Split systemcfg data out of vdso data page
powerpc: Add kconfig option for the systemcfg page
powerpc/pseries/lparcfg: Use num_possible_cpus() for potential processors
powerpc/pseries/lparcfg: Fix printing of system_active_processors
powerpc/procfs: Propagate error of remap_pfn_range()
powerpc/vdso: Remove offset comment from 32bit vdso_arch_data
x86/vdso: Split virtual clock pages into dedicated mapping
x86/vdso: Delete vvar.h
x86/vdso: Access vdso data without vvar.h
x86/vdso: Move the rng offset to vsyscall.h
x86/vdso: Access rng vdso data without vvar.h
x86/vdso: Access timens vdso data without vvar.h
x86/vdso: Allocate vvar page from C code
x86/vdso: Access rng data from kernel without vvar
x86/vdso: Place vdso_data at beginning of vvar page
x86/vdso: Use __arch_get_vdso_data() to access vdso data
x86/mm/mmap: Remove arch_vma_name()
...
Pull vfs file updates from Christian Brauner:
"This contains changes the changes for files for this cycle:
- Introduce a new reference counting mechanism for files.
As atomic_inc_not_zero() is implemented with a try_cmpxchg() loop
it has O(N^2) behaviour under contention with N concurrent
operations and it is in a hot path in __fget_files_rcu().
The rcuref infrastructures remedies this problem by using an
unconditional increment relying on safe- and dead zones to make
this work and requiring rcu protection for the data structure in
question. This not just scales better it also introduces overflow
protection.
However, in contrast to generic rcuref, files require a memory
barrier and thus cannot rely on *_relaxed() atomic operations and
also require to be built on atomic_long_t as having massive amounts
of reference isn't unheard of even if it is just an attack.
This adds a file specific variant instead of making this a generic
library.
This has been tested by various people and it gives consistent
improvement up to 3-5% on workloads with loads of threads.
- Add a fastpath for find_next_zero_bit(). Skip 2-levels searching
via find_next_zero_bit() when there is a free slot in the word that
contains the next fd. This improves pts/blogbench-1.1.0 read by 8%
and write by 4% on Intel ICX 160.
- Conditionally clear full_fds_bits since it's very likely that a bit
in full_fds_bits has been cleared during __clear_open_fds(). This
improves pts/blogbench-1.1.0 read up to 13%, and write up to 5% on
Intel ICX 160.
- Get rid of all lookup_*_fdget_rcu() variants. They were used to
lookup files without taking a reference count. That became invalid
once files were switched to SLAB_TYPESAFE_BY_RCU and now we're
always taking a reference count. Switch to an already existing
helper and remove the legacy variants.
- Remove pointless includes of <linux/fdtable.h>.
- Avoid cmpxchg() in close_files() as nobody else has a reference to
the files_struct at that point.
- Move close_range() into fs/file.c and fold __close_range() into it.
- Cleanup calling conventions of alloc_fdtable() and expand_files().
- Merge __{set,clear}_close_on_exec() into one.
- Make __set_open_fd() set cloexec as well instead of doing it in two
separate steps"
* tag 'vfs-6.13.file' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
selftests: add file SLAB_TYPESAFE_BY_RCU recycling stressor
fs: port files to file_ref
fs: add file_ref
expand_files(): simplify calling conventions
make __set_open_fd() set cloexec state as well
fs: protect backing files with rcu
file.c: merge __{set,clear}_close_on_exec()
alloc_fdtable(): change calling conventions.
fs/file.c: add fast path in find_next_fd()
fs/file.c: conditionally clear full_fds
fs/file.c: remove sanity_check and add likely/unlikely in alloc_fd()
move close_range(2) into fs/file.c, fold __close_range() into it
close_files(): don't bother with xchg()
remove pointless includes of <linux/fdtable.h>
get rid of ...lookup...fdget_rcu() family
Port files to rely on file_ref reference to improve scaling and gain
overflow protection.
- We continue to WARN during get_file() in case a file that is already
marked dead is revived as get_file() is only valid if the caller
already holds a reference to the file. This hasn't changed just the
check changes.
- The semantics for epoll and ttm's dmabuf usage have changed. Both
epoll and ttm synchronize with __fput() to prevent the underlying file
from beeing freed.
(1) epoll
Explaining epoll is straightforward using a simple diagram.
Essentially, the mutex of the epoll instance needs to be taken in both
__fput() and around epi_fget() preventing the file from being freed
while it is polled or preventing the file from being resurrected.
CPU1 CPU2
fput(file)
-> __fput(file)
-> eventpoll_release(file)
-> eventpoll_release_file(file)
mutex_lock(&ep->mtx)
epi_item_poll()
-> epi_fget()
-> file_ref_get(file)
mutex_unlock(&ep->mtx)
mutex_lock(&ep->mtx);
__ep_remove()
mutex_unlock(&ep->mtx);
-> kmem_cache_free(file)
(2) ttm dmabuf
This explanation is a bit more involved. A regular dmabuf file stashed
the dmabuf in file->private_data and the file in dmabuf->file:
file->private_data = dmabuf;
dmabuf->file = file;
The generic release method of a dmabuf file handles file specific
things:
f_op->release::dma_buf_file_release()
while the generic dentry release method of a dmabuf handles dmabuf
freeing including driver specific things:
dentry->d_release::dma_buf_release()
During ttm dmabuf initialization in ttm_object_device_init() the ttm
driver copies the provided struct dma_buf_ops into a private location:
struct ttm_object_device {
spinlock_t object_lock;
struct dma_buf_ops ops;
void (*dmabuf_release)(struct dma_buf *dma_buf);
struct idr idr;
};
ttm_object_device_init(const struct dma_buf_ops *ops)
{
// copy original dma_buf_ops in private location
tdev->ops = *ops;
// stash the release method of the original struct dma_buf_ops
tdev->dmabuf_release = tdev->ops.release;
// override the release method in the copy of the struct dma_buf_ops
// with ttm's own dmabuf release method
tdev->ops.release = ttm_prime_dmabuf_release;
}
When a new dmabuf is created the struct dma_buf_ops with the overriden
release method set to ttm_prime_dmabuf_release is passed in exp_info.ops:
DEFINE_DMA_BUF_EXPORT_INFO(exp_info);
exp_info.ops = &tdev->ops;
exp_info.size = prime->size;
exp_info.flags = flags;
exp_info.priv = prime;
The call to dma_buf_export() then sets
mutex_lock_interruptible(&prime->mutex);
dma_buf = dma_buf_export(&exp_info)
{
dmabuf->ops = exp_info->ops;
}
mutex_unlock(&prime->mutex);
which creates a new dmabuf file and then install a file descriptor to
it in the callers file descriptor table:
ret = dma_buf_fd(dma_buf, flags);
When that dmabuf file is closed we now get:
fput(file)
-> __fput(file)
-> f_op->release::dma_buf_file_release()
-> dput()
-> d_op->d_release::dma_buf_release()
-> dmabuf->ops->release::ttm_prime_dmabuf_release()
mutex_lock(&prime->mutex);
if (prime->dma_buf == dma_buf)
prime->dma_buf = NULL;
mutex_unlock(&prime->mutex);
Where we can see that prime->dma_buf is set to NULL. So when we have
the following diagram:
CPU1 CPU2
fput(file)
-> __fput(file)
-> f_op->release::dma_buf_file_release()
-> dput()
-> d_op->d_release::dma_buf_release()
-> dmabuf->ops->release::ttm_prime_dmabuf_release()
ttm_prime_handle_to_fd()
mutex_lock_interruptible(&prime->mutex)
dma_buf = prime->dma_buf
dma_buf && get_dma_buf_unless_doomed(dma_buf)
-> file_ref_get(dma_buf->file)
mutex_unlock(&prime->mutex);
mutex_lock(&prime->mutex);
if (prime->dma_buf == dma_buf)
prime->dma_buf = NULL;
mutex_unlock(&prime->mutex);
-> kmem_cache_free(file)
The logic of the mechanism is the same as for epoll: sync with
__fput() preventing the file from being freed. Here the
synchronization happens through the ttm instance's prime->mutex.
Basically, the lifetime of the dma_buf and the file are tighly
coupled.
Both (1) and (2) used to call atomic_inc_not_zero() to check whether
the file has already been marked dead and then refuse to revive it.
This is only safe because both (1) and (2) sync with __fput() and thus
prevent kmem_cache_free() on the file being called and thus prevent
the file from being immediately recycled due to SLAB_TYPESAFE_BY_RCU.
Both (1) and (2) have been ported from atomic_inc_not_zero() to
file_ref_get(). That means a file that is already in the process of
being marked as FILE_REF_DEAD:
file_ref_put()
cnt = atomic_long_dec_return()
-> __file_ref_put(cnt)
if (cnt == FIlE_REF_NOREF)
atomic_long_try_cmpxchg_release(cnt, FILE_REF_DEAD)
can be revived again:
CPU1 CPU2
file_ref_put()
cnt = atomic_long_dec_return()
-> __file_ref_put(cnt)
if (cnt == FIlE_REF_NOREF)
file_ref_get()
// Brings reference back to FILE_REF_ONEREF
atomic_long_add_negative()
atomic_long_try_cmpxchg_release(cnt, FILE_REF_DEAD)
This is fine and inherent to the file_ref_get()/file_ref_put()
semantics. For both (1) and (2) this is safe because __fput() is
prevented from making progress if file_ref_get() fails due to the
aforementioned synchronization mechanisms.
Two cases need to be considered that affect both (1) epoll and (2) ttm
dmabuf:
(i) fput()'s file_ref_put() and marks the file as FILE_REF_NOREF but
before that fput() can mark the file as FILE_REF_DEAD someone
manages to sneak in a file_ref_get() and brings the refcount back
from FILE_REF_NOREF to FILE_REF_ONEREF. In that case the original
fput() doesn't call __fput(). For epoll the poll will finish and
for ttm dmabuf the file can be used again. For ttm dambuf this is
actually an advantage because it avoids immediately allocating
a new dmabuf object.
CPU1 CPU2
file_ref_put()
cnt = atomic_long_dec_return()
-> __file_ref_put(cnt)
if (cnt == FIlE_REF_NOREF)
file_ref_get()
// Brings reference back to FILE_REF_ONEREF
atomic_long_add_negative()
atomic_long_try_cmpxchg_release(cnt, FILE_REF_DEAD)
(ii) fput()'s file_ref_put() marks the file FILE_REF_NOREF and
also suceeds in actually marking it FILE_REF_DEAD and then calls
into __fput() to free the file.
When either (1) or (2) call file_ref_get() they fail as
atomic_long_add_negative() will return true.
At the same time, both (1) and (2) all file_ref_get() under
mutexes that __fput() must also acquire preventing
kmem_cache_free() from freeing the file.
So while this might be treated as a change in semantics for (1) and
(2) it really isn't. It if should end up causing issues this can be
fixed by adding a helper that does something like:
long cnt = atomic_long_read(&ref->refcnt);
do {
if (cnt < 0)
return false;
} while (!atomic_long_try_cmpxchg(&ref->refcnt, &cnt, cnt + 1));
return true;
which would block FILE_REF_NOREF to FILE_REF_ONEREF transitions.
- Jann correctly pointed out that kmem_cache_zalloc() cannot be used
anymore once files have been ported to file_ref_t.
The kmem_cache_zalloc() call will memset() the whole struct file to
zero when it is reallocated. This will also set file->f_ref to zero
which mens that a concurrent file_ref_get() can return true:
CPU1 CPU2
__get_file_rcu()
rcu_dereference_raw()
close()
[frees file]
alloc_empty_file()
kmem_cache_zalloc()
[reallocates same file]
memset(..., 0, ...)
file_ref_get()
[increments 0->1, returns true]
init_file()
file_ref_init(..., 1)
[sets to 0]
rcu_dereference_raw()
fput()
file_ref_put()
[decrements 0->FILE_REF_NOREF, frees file]
[UAF]
causing a concurrent __get_file_rcu() call to acquire a reference to
the file that is about to be reallocated and immediately freeing it
on realizing that it has been recycled. This causes a UAF for the
task that reallocated/recycled the file.
This is prevented by switching from kmem_cache_zalloc() to
kmem_cache_alloc() and initializing the fields manually. With
file->f_ref initialized last.
Note that a memset() also isn't guaranteed to atomically update an
unsigned long so it's theoretically possible to see torn and
therefore bogus counter values.
Link: https://lore.kernel.org/r/20241007-brauner-file-rcuref-v2-3-387e24dc9163@kernel.org
Signed-off-by: Christian Brauner <brauner@kernel.org>
Add missing tag for "Wa_14019159160 - Case 2" (for existing
PXP code that ensures run alone mode bit is set to allow
PxP-decryption.
v5: - remove the max IP_VER check since new platforms that
i915 supports needs this fix and tag the caller too
(John Harrison).
v4: - Include IP_VER 12.71. (Matt Roper)
v3: - Check targeted platforms using IP_VAL. (John Harrison)
v2: - Fix WA id number (John Harrison).
- Improve comments and code to be specific
for the targeted platforms (John Harrison)
Signed-off-by: Alan Previn <alan.previn.teres.alexis@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241016001658.2671225-1-alan.previn.teres.alexis@intel.com
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Fault is currently of type u32 and with the introduction of the
generalized vdso/page.h we trigger the error below:
drivers/gpu/drm/i915/gt/intel_gt_print.h:29:36: error: format ‘%lx’ expects
argument of type ‘long unsigned int’, but argument 6 has type ‘u32’ {aka
‘unsigned int’} [-Werror=format=]
29 | drm_dbg(&(_gt)->i915->drm, "GT%u: " _fmt, (_gt)->info.id,
| ^~~~~~~~
include/drm/drm_print.h:424:39: note: in definition of macro ‘drm_dev_dbg’
424 | __drm_dev_dbg(NULL, dev, cat, fmt, ##__VA_ARGS__)
| ^~~
include/drm/drm_print.h:524:33: note: in expansion of macro ‘drm_dbg_driver’
524 | #define drm_dbg(drm, fmt, ...) drm_dbg_driver(drm, fmt, ##__VA_ARGS__)
| ^~~~~~~~~~~~~~
linux/drivers/gpu/drm/i915/gt/intel_gt_print.h:29:9: note: in expansion of macro
‘drm_dbg’
29 | drm_dbg(&(_gt)->i915->drm, "GT%u: " _fmt, (_gt)->info.id,
| ^~~~~~~
drivers/gpu/drm/i915/gt/intel_gt.c:310:25: note: in expansion of macro ‘gt_dbg’
310 | gt_dbg(gt, "Unexpected fault\n"
| ^~~~~~
This happens because the type of PAGE_MASK depends on the architecture.
Prevent the compilation error changing the 'fault' type to unsigned
long.
Signed-off-by: Vincenzo Frascino <vincenzo.frascino@arm.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Arnd Bergmann <arnd@arndb.de>
Link: https://lore.kernel.org/all/20241014151340.1639555-2-vincenzo.frascino@arm.com
Implement pmu support for gen2 so that one can use intel_gpu_top
on it once again.
Gen2 lacks MI_MODE/MODE_IDLE so we'll have to do a bit more work
to determine the state of the engine:
- to determine if the ring contains unconsumed data we can simply
compare RING_TAIL vs. RING_HEAD
- also check RING_HEAD vs. ACTHD to catch cases where the hardware
is still executing a batch buffer but the ring head has already
caught up with the tail. Not entirely sure if that's actually
possible or not, but maybe it can happen if the batch buffer is
initiated from the very end of the ring? But even if not strictly
necessary there's no real harm in checking anyway.
- MI_WAIT_FOR_EVENT can be detected via a dedicated bit in RING_HEAD
v2: Use genX_ prefix rarther than suffix (Jani)
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241008214349.23331-5-ville.syrjala@linux.intel.com
Acked-by: Jani Nikula <jani.nikula@intel.com>
For intel_wakeref_t, opaque is reasonable, but disguising the underlying
struct ref_tracker * as an unsigned long is not so great. Update the
typedef to remove one level of disguise.
Although the kernel coding style strongly discourages pointer typedefs,
it's a better alternative, and an incremental improvement on the status
quo. It provides much better type safety than an unsigned long could,
and prevents passing magic -1 instead of INTEL_WAKEREF_DEF. Moreover, it
provides a gradual path for replacing intel_wakeref_t with struct
ref_tracker * if desired.
As an extra safety measure, check for error pointers in
intel_ref_tracker_free() before passing them on to ref_tracker_free(),
to catch any mistakes with mock gt special wakeref value.
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/cca2b0631f816ad90461aa1bf4fe3f80c0e13464.1726680898.git.jani.nikula@intel.com
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
