Addresses regressions from the reverted merge
of the tbx fault manager for host memory.
Recursive locking of mutex caused deadlock.
To fix, separate tbx fault data from base
cpu fault data, allowing separate mutexes
for each, eliminating recursive locks on
the same mutex.
By separating, we also help ensure that tbx-related
changes don't affect the original cpu fault manager code
paths.
As an added safe guard preventing critical regressions
and avoiding another auto-revert, the tbx fault manager
is hidden behind a new debug flag which is disabled by default.
Related-To: NEO-12268
Signed-off-by: Jack Myers <jack.myers@intel.com>
Addresses regressions from the reverted merge
of the tbx fault manager for host memory.
This fixes attempts by the tbx fault manager
to protect/unprotect host buffer memory, even
if the host ptr was not driver-allocated.
In the case of the smoke test that triggered
the critical regression, clCreateBuffer was
called with the CL_MEM_USE_HOST_PTR flag.
The subsequent `mprotect` calls on the
provided host ptr then failed.
Related-To: NEO-12268
Signed-off-by: Jack Myers <jack.myers@intel.com>
In TBX mode, the host could not write to host buffers after access from device
code due to the lack of a migration mechanism post-initial TBX upload.
Migration is unnecessary with real hardware, but required for TBX.
This patch introduces a new page fault manager type that extends the original
CPU fault manager, enabling automatic migration of host buffers in TBX mode.
Refactoring was necessary to avoid diamond inheritance, achieved by using a
template parameter as the base class for OS-specific fault managers.
Related-To: NEO-12268
Signed-off-by: Jack Myers <jack.myers@intel.com>
Related-To: NEO-11755
Removing trim candidate list reduces overhead
caused by residency handling. Allocations required
for eviction are placed in eviction container managed
by CSR.
Signed-off-by: Szymon Morek <szymon.morek@intel.com>
- refactor and call proceesFlushResdiency() on memoryOperationsHandler
- call free() to remove allocation from resident allocations when
graphics allocation is released
Related-To: NEO-11719
Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
In this mode AUB csr will be created, however, no aub file will be created
Related-To: NEO-11097
Signed-off-by: Mateusz Jablonski <mateusz.jablonski@intel.com>
- not only gpuAddress is offset but also cpu address with data needs
to be offset while writing memory.
Related-To: GSD-6604
Signed-off-by: Mateusz Hoppe <mateusz.hoppe@intel.com>
We want to return error code to the application instead of aborting when
we are not able to make more memory resident.
Related-To: NEO-7289
Signed-off-by: Fabian Zwolinski <fabian.zwolinski@intel.com>
When running multiple threads, one thread could clear
allocationsForDownload while another was iterating over it.
Signed-off-by: Dominik Dabek <dominik.dabek@intel.com>
This change uses DRM_IOCTL_I915_GET_RESET_STATS to detect
GPU hangs. When such situation is encountered, then
zeCommandQueueSynchronize returns ZE_RESULT_ERROR_DEVICE_LOST.
Related-To: NEO-5313
Signed-off-by: Patryk Wrobel <patryk.wrobel@intel.com>
