commit 67d15c7aa0 upstream.
These are u64 variables that come from the user via
qaic_attach_slice_bo_ioctl(). Use check_add_overflow() to ensure that
the math doesn't have an integer wrapping bug.
Cc: stable@vger.kernel.org
Fixes: ff13be8303 ("accel/qaic: Add datapath")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Reviewed-by: Jeff Hugo <jeff.hugo@oss.qualcomm.com>
Signed-off-by: Jeff Hugo <jeff.hugo@oss.qualcomm.com>
Link: https://patchwork.freedesktop.org/patch/msgid/176388fa-40fe-4cb4-9aeb-2c91c22130bd@stanley.mountain
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 84a833d906 ]
When slicing a BO, we need to iterate through the BO's sgt to find the
right pieces to construct the slice. Some of the data types chosen for
this process are incorrectly too small, and can overflow. This can
result in the incorrect slice construction, which can lead to data
corruption in workload execution.
The device can only handle 32-bit sized transfers, and the scatterlist
struct only supports 32-bit buffer sizes, so our upper limit for an
individual transfer is an unsigned int. Using an int is incorrect due to
the reservation of the sign bit. Upgrade the length of a scatterlist
entry and the offsets into a scatterlist entry to unsigned int for a
correct representation.
While each transfer may be limited to 32-bits, the overall BO may exceed
that size. For counting the total length of the BO, we need a type that
can represent the largest allocation possible on the system. That is the
definition of size_t, so use it.
Fixes: ff13be8303 ("accel/qaic: Add datapath")
Signed-off-by: Jeffrey Hugo <quic_jhugo@quicinc.com>
Signed-off-by: Jeff Hugo <jeff.hugo@oss.qualcomm.com>
Reviewed-by: Lizhi Hou <lizhi.hou@amd.com>
Reviewed-by: Troy Hanson <quic_thanson@quicinc.com>
Reviewed-by: Youssef Samir <quic_yabdulra@quicinc.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250306171959.853466-1-jeff.hugo@oss.qualcomm.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 41a2d8286c ]
Disable runtime PM for the duration of reset/recovery so it is possible
to set the correct runtime PM state depending on the outcome of the
`ivpu_resume()`. Don’t suspend or reset the HW if the NPU is suspended
when the reset/recovery is requested. Also, move common reset/recovery
code to separate functions for better code readability.
Fixes: 27d19268cf ("accel/ivpu: Improve recovery and reset support")
Cc: stable@vger.kernel.org # v6.8+
Reviewed-by: Maciej Falkowski <maciej.falkowski@linux.intel.com>
Reviewed-by: Jeffrey Hugo <quic_jhugo@quicinc.com>
Signed-off-by: Jacek Lawrynowicz <jacek.lawrynowicz@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250129124009.1039982-4-jacek.lawrynowicz@linux.intel.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 5e162f872d ]
Send JSM state dump message at the beginning of TDR handler. This allows
FW to collect debug info in the FW log before the state of the NPU is
lost allowing to analyze the cause of a TDR.
Wait a predefined timeout (10 ms) so the FW has a chance to write debug
logs. We cannot wait for JSM response at this point because IRQs are
already disabled before TDR handler is invoked.
Signed-off-by: Tomasz Rusinowicz <tomasz.rusinowicz@intel.com>
Reviewed-by: Jacek Lawrynowicz <jacek.lawrynowicz@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240930195322.461209-9-jacek.lawrynowicz@linux.intel.com
Signed-off-by: Jacek Lawrynowicz <jacek.lawrynowicz@linux.intel.com>
Stable-dep-of: 41a2d8286c ("accel/ivpu: Fix error handling in recovery/reset")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit bade034052 ]
Use coredump (if available) to collect FW logs in case of a FW crash.
This makes dmesg more readable and allows to collect more log data.
Signed-off-by: Karol Wachowski <karol.wachowski@intel.com>
Reviewed-by: Jacek Lawrynowicz <jacek.lawrynowicz@linux.intel.com>
Reviewed-by: Jeffrey Hugo <quic_jhugo@quicinc.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240930195322.461209-8-jacek.lawrynowicz@linux.intel.com
Signed-off-by: Jacek Lawrynowicz <jacek.lawrynowicz@linux.intel.com>
Stable-dep-of: 41a2d8286c ("accel/ivpu: Fix error handling in recovery/reset")
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit f2bc2afe34 upstream.
pm_runtime_resume_and_get() sets dev->power.runtime_error that causes
all subsequent pm_runtime_get_sync() calls to fail.
Clear the runtime_error using pm_runtime_set_suspended(), so the driver
doesn't have to be reloaded to recover when the NPU fails to boot during
runtime resume.
Fixes: 7d4b4c7443 ("accel/ivpu: Remove suspend_reschedule_counter")
Cc: stable@vger.kernel.org # v6.11+
Reviewed-by: Maciej Falkowski <maciej.falkowski@linux.intel.com>
Reviewed-by: Jeffrey Hugo <quic_jhugo@quicinc.com>
Signed-off-by: Jacek Lawrynowicz <jacek.lawrynowicz@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20250129124009.1039982-3-jacek.lawrynowicz@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 901dd2617c upstream.
Restore PCI state after putting the NPU in D0.
Restoring state before powering up the device caused a Qemu crash
if NPU was running in passthrough mode and recovery was performed.
Fixes: 3534eacbf1 ("accel/ivpu: Fix PCI D0 state entry in resume")
Cc: stable@vger.kernel.org # v6.8+
Reviewed-by: Karol Wachowski <karol.wachowski@linux.intel.com>
Signed-off-by: Jacek Lawrynowicz <jacek.lawrynowicz@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241106105549.2757115-1-jacek.lawrynowicz@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 0f6482caa6 upstream.
Move pm_runtime_set_active() to ivpu_pm_init() so when
ivpu_ipc_send_receive_internal() is executed before ivpu_pm_enable()
it already has correct runtime state, even if last resume was
not successful.
Fixes: 8ed520ff46 ("accel/ivpu: Move set autosuspend delay to HW specific code")
Cc: stable@vger.kernel.org # v6.7+
Reviewed-by: Karol Wachowski <karol.wachowski@intel.com>
Reviewed-by: Jeffrey Hugo <quic_jhugo@quicinc.com>
Signed-off-by: Jacek Lawrynowicz <jacek.lawrynowicz@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241210130939.1575610-4-jacek.lawrynowicz@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit b8128f7815 ]
Add basic support for the new AIC080 product. The PCIe Device ID is
0xa080. AIC080 is a lower cost, lower performance SKU variant of AIC100.
From the qaic perspective, it is the same as AIC100.
Reviewed-by: Troy Hanson <quic_thanson@quicinc.com>
Signed-off-by: Jeffrey Hugo <quic_jhugo@quicinc.com>
Reviewed-by: Jacek Lawrynowicz <jacek.lawrynowicz@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241004195209.3910996-1-quic_jhugo@quicinc.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 5eaa497411 ]
Refactor IPC send and receive functions to allow correct
handling of operations that should not trigger a recovery process.
Expose ivpu_send_receive_internal(), which is now utilized by the D0i3
entry, DCT initialization, and HWS initialization functions.
These functions have been modified to return error codes gracefully,
rather than initiating recovery.
The updated functions are invoked within ivpu_probe() and ivpu_resume(),
ensuring that any errors encountered during these stages result in a proper
teardown or shutdown sequence. The previous approach of triggering recovery
within these functions could lead to a race condition, potentially causing
undefined behavior and kernel crashes due to null pointer dereferences.
Fixes: 45e45362e0 ("accel/ivpu: Introduce ivpu_ipc_send_receive_active()")
Signed-off-by: Karol Wachowski <karol.wachowski@intel.com>
Reviewed-by: Jacek Lawrynowicz <jacek.lawrynowicz@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240930195322.461209-23-jacek.lawrynowicz@linux.intel.com
Signed-off-by: Jacek Lawrynowicz <jacek.lawrynowicz@linux.intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
The NOC firewall interrupt means that the HW prevented
unauthorized access to a protected resource, so there
is no need to trigger device reset in such case.
To facilitate security testing add firewall_irq_counter
debugfs file that tracks firewall interrupts.
Fixes: 8a27ad81f7 ("accel/ivpu: Split IP and buttress code")
Cc: stable@vger.kernel.org # v6.11+
Signed-off-by: Andrzej Kacprowski <Andrzej.Kacprowski@intel.com>
Reviewed-by: Jacek Lawrynowicz <jacek.lawrynowicz@linux.intel.com>
Reviewed-by: Jeffrey Hugo <quic_jhugo@quicinc.com>
Signed-off-by: Jacek Lawrynowicz <jacek.lawrynowicz@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241017144958.79327-1-jacek.lawrynowicz@linux.intel.com
Only for_each_sgtable_dma_sg() should be used to walk through a SG table
to grab correct bus address and length pair after calling DMA MAP API on
a SG table as DMA MAP APIs updates the SG table and for_each_sgtable_sg()
walks through the original SG table.
Fixes: ff13be8303 ("accel/qaic: Add datapath")
Fixes: 129776ac2e ("accel/qaic: Add control path")
Signed-off-by: Pranjal Ramajor Asha Kanojiya <quic_pkanojiy@quicinc.com>
Reviewed-by: Jeffrey Hugo <quic_jhugo@quicinc.com>
Signed-off-by: Jeffrey Hugo <quic_jhugo@quicinc.com>
Reviewed-by: Jacek Lawrynowicz <jacek.lawrynowicz@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241004193252.3888544-1-quic_jhugo@quicinc.com
- support DMA zones for arm64 systems where memory starts at > 4GB
(Baruch Siach, Catalin Marinas)
- support direct calls into dma-iommu and thus obsolete dma_map_ops for
many common configurations (Leon Romanovsky)
- add DMA-API tracing (Sean Anderson)
- remove the not very useful return value from various dma_set_* APIs
(Christoph Hellwig)
- misc cleanups and minor optimizations (Chen Y, Yosry Ahmed,
Christoph Hellwig)
-----BEGIN PGP SIGNATURE-----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=DZVc
-----END PGP SIGNATURE-----
Merge tag 'dma-mapping-6.12-2024-09-19' of git://git.infradead.org/users/hch/dma-mapping
Pull dma-mapping updates from Christoph Hellwig:
- support DMA zones for arm64 systems where memory starts at > 4GB
(Baruch Siach, Catalin Marinas)
- support direct calls into dma-iommu and thus obsolete dma_map_ops for
many common configurations (Leon Romanovsky)
- add DMA-API tracing (Sean Anderson)
- remove the not very useful return value from various dma_set_* APIs
(Christoph Hellwig)
- misc cleanups and minor optimizations (Chen Y, Yosry Ahmed, Christoph
Hellwig)
* tag 'dma-mapping-6.12-2024-09-19' of git://git.infradead.org/users/hch/dma-mapping:
dma-mapping: reflow dma_supported
dma-mapping: reliably inform about DMA support for IOMMU
dma-mapping: add tracing for dma-mapping API calls
dma-mapping: use IOMMU DMA calls for common alloc/free page calls
dma-direct: optimize page freeing when it is not addressable
dma-mapping: clearly mark DMA ops as an architecture feature
vdpa_sim: don't select DMA_OPS
arm64: mm: keep low RAM dma zone
dma-mapping: don't return errors from dma_set_max_seg_size
dma-mapping: don't return errors from dma_set_seg_boundary
dma-mapping: don't return errors from dma_set_min_align_mask
scsi: check that busses support the DMA API before setting dma parameters
arm64: mm: fix DMA zone when dma-ranges is missing
dma-mapping: direct calls for dma-iommu
dma-mapping: call ->unmap_page and ->unmap_sg unconditionally
arm64: support DMA zone above 4GB
dma-mapping: replace zone_dma_bits by zone_dma_limit
dma-mapping: use bit masking to check VM_DMA_COHERENT
A NULL dev->dma_parms indicates either a bus that is not DMA capable or
grave bug in the implementation of the bus code.
There isn't much the driver can do in terms of error handling for either
case, so just warn and continue as DMA operations will fail anyway.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Robin Murphy <robin.murphy@arm.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Acked-by: Ulf Hansson <ulf.hansson@linaro.org> # For MMC
Accel minor management is based on DRM (and is also using struct
drm_minor internally), since DRM is using XArray for minors, it makes
sense to also convert accel.
As the two implementations are identical (only difference being the
underlying xarray), move the accel_minor_* functionality to DRM.
Signed-off-by: Michał Winiarski <michal.winiarski@intel.com>
Acked-by: James Zhu <James.Zhu@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240823163048.2676257-3-michal.winiarski@intel.com
Signed-off-by: Christian König <christian.koenig@amd.com>
Modules that load firmware from various paths at runtime must declare
those paths at compile time, via the MODULE_FIRMWARE macro, so that the
firmware paths are included in the module's metadata.
The accel/ivpu driver loads firmware but lacks this metadata,
preventing dracut from correctly locating firmware files. Fix it.
Fixes: 9ab43e95f9 ("accel/ivpu: Switch to generation based FW names")
Fixes: 02d5b0aacd ("accel/ivpu: Implement firmware parsing and booting")
Signed-off-by: Alexander F. Lent <lx@xanderlent.com>
Reviewed-by: Jacek Lawrynowicz <jacek.lawrynowicz@linux.intel.com>
Signed-off-by: Jacek Lawrynowicz <jacek.lawrynowicz@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240709-fix-ivpu-firmware-metadata-v3-1-55f70bba055b@xanderlent.com
This tag contains habanalabs driver changes for v6.11.
The notable changes are:
- uAPI changes:
- Use device-name directory in debugfs-driver-habanalabs.
- Expose server type in debugfs.
- New features and improvements:
- Gradual sleep in polling memory macro.
- Reduce Gaudi2 MSI-X interrupt count to 128.
- Add Gaudi2-D revision support.
- Firmware related changes:
- Add timestamp to CPLD info.
- Gaudi2: Assume hard-reset by firmware upon MC SEI severe error.
- Align Gaudi2 interrupt names.
- Check for errors after preboot is ready.
- Bug fixes and code cleanups:
- Move heartbeat work initialization to early init.
- Fix a race when receiving events during reset.
- Change the heartbeat scheduling point.
- Maintainers:
- Change habanalabs maintainer and git repo path.
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Ofir Bitton <obitton@habana.ai>
Link: https://patchwork.freedesktop.org/patch/msgid/ZnfIjTH5AYQvPe7n@obitton-vm-u22.habana-labs.com
It’s better to avoid long sleeps right from the beginning of the polling
since the data may be available much sooner than the sleep period.
Because polling host memory is inexpensive, this change gradually
increases the sleep time up to the user-requested period.
Signed-off-by: Didi Freiman <dfreiman@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
The device heartbeat work is currently initialized at
device_heartbeat_schedule() which is called at the end of
hl_device_init().
However hl_device_init() can fail at a previous step, and in such a
case, a subsequent call to hl_device_fini() will lead to calling
cleanup_resources() and accessing this work uninitialized.
As there is no real need to re-initialize this work every time it is
rescheduled, move this initialization to device_early_init() to be done
once and early enough.
Signed-off-by: Tomer Tayar <ttayar@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
The test packet which is sent to FW for the PQ heartbeat is used also as
the trigger in FW to send the EQ heartbeat event.
Add the time of the last sent packet to the debug info which is printed
upon a EQ heartbeat failure.
Signed-off-by: Tomer Tayar <ttayar@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
Add a dump of the EQ entries headers upon a EQ heartbeat failure.
Signed-off-by: Tomer Tayar <ttayar@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
Don't print the "previous EQ index" value in case of a EQ heartbeat
failure, because it is incremented along with the EQ CI and therefore
redundant.
In addition, as the CPU-CP PI is zeroed when it reaches a value that is
twice the queue size, add a value of the CI with a similar wrap around,
to make it easier to compare the values.
Signed-off-by: Tomer Tayar <ttayar@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
In order to have better debuggability upon encountering FW issues,
We are adding additional info once CPU packet timeout expires.
Signed-off-by: Farah Kassabri <fkassabri@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
When device release triggers a hard reset, there is a printout of
the cause. Currently listed causes (that increment context refcount)
are active command submissions and exported DMA buffer objects. In
any other case, the printout emits "unknown reason". We identify and
print another reason - allocated command buffers.
Signed-off-by: Ilia Levi <illevi@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
Some systems allow a maximum number of 128 MSI-X interrupts.
Hence we reduce the interrupt count to 128 instead of 512.
Reviewed-by: Tomer Tayar <ttayar@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
When sending disable pci msg towards firmware, there is a
possibility that an EQ packet is already pending,
disabling EQ interrupt will prevent this from happening.
The interrupt will be re-enabled after reset.
Signed-off-by: Tal Cohen <talcohen@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
Currently we schedule the heartbeat thread at late init, only then
we set the INTS_REGISTER packet which enables events to be received
from firmware.
Init may take some time and we want to give firmware 2 full cycles of
heartbeat thread after it received INTS_REGISTER.
The patch will move the heartbeat thread scheduling to be after driver
is done with all initializations.
Signed-off-by: Farah Kassabri <fkassabri@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
Netowrk EDMAs uses more outstanding transfers so this needs to be
programmed by EDMA firmware.
Signed-off-by: Rakesh Ughreja <rughreja@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
There are several timestamp registration debug prints which
spams the kernel log whenever dyn debug is enabled.
Remove those prints.
Reviewed-by: Tomer Tayar <ttayar@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
Add cpld_timestamp field to cpucp_info structure and return cpld
timestamp as part of cpld version
Signed-off-by: Vitaly Margolin <vmargolin@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
As the new dynamic EQ includes clock change events which are common and
not ASIC-specific, add a common handler for these events.
Signed-off-by: Tomer Tayar <ttayar@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
Gaudi2 with PCI revision ID with the value of '4' represents Gaudi2D
device and should be detected and initialized as Gaudi2.
Signed-off-by: Farah Kassabri <fkassabri@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
hl_eq_heartbeat_event_handle() doesn't have ASIC specific code, and
therefore can be moved from Gaudi2-only code to common code, and
possibly used for other ASICs.
Signed-off-by: Tomer Tayar <ttayar@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
Future supported ASICs might use the dynamic EQ mechanism with the
firmware, and in that case the EQ size won't be equal to the default
HL_EQ_SIZE_IN_BYTES value.
Add an ASIC property to enable overriding this value.
Signed-off-by: Tomer Tayar <ttayar@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
FW initiates a hard reset upon an MC SEI severe error.
Align the driver to expect this reset and avoid accessing the device
until the reset is done.
Signed-off-by: Tomer Tayar <ttayar@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
The return value in gaudi2_hbm_sei_handle_read_err() is boolean and not
a bitmask, so there is need for "|= true".
In addition, rename the 'rc' variable, as no "return code" is returned
here but an indication if a hard reset is required.
Signed-off-by: Tomer Tayar <ttayar@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
when reporting tpc events, the dcore and tpc in dcore should
be reported and propagated, and not the generatl tpc number
Signed-off-by: Ariel Suller <asuller@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
Driver should check and report any fatal errors detected by preboot,
before it attempts to load the boot fit.
Some errors may cause the driver to stop the boot process and mark
the device as unusable.
This check will allow the driver to fail and print the error reported
by preboot and skip the time wasting attempt of trying to load the
boot fit, which will fail due to the error.
Signed-off-by: Farah Kassabri <fkassabri@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
Struct comms_desc_header is deprecated and replaced by struct
comms_msg_header. As a preparation for removing comms_desc_header
from FW, all it's usage in code is replaced by comms_msg_header.
Signed-off-by: Igal Zeltser <izeltser@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
It is hard to debug the reason for heartbeat check failures.
As an attempt to ease this task, this patch will provide more
information when this failure happens.
Heartbeat checks the communication with FW, so printing
the CPU queue pi/ci and the counter of how many times that event
was received would help in debugging the issue.
Signed-off-by: Farah Kassabri <fkassabri@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
This addition helps log parsers better define the error without the need
to go back and search the device name on former log lines.
Signed-off-by: Ohad Sharabi <osharabi@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
Exposing server type through debugfs to enable easier access via
scripts.
Signed-off-by: Tal Risin <trisin@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
Trace events might still be recorded after the accel device is released,
while the device name is no longer available.
Modify the trace functions to use the parent device instead, which is
available at that point and still informative as the device name.
Signed-off-by: Tomer Tayar <ttayar@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
If we detected heartbet event while some daemon in the background send
(via driver interface) CPUCP messages the dmesg will be flooded.
Instead, a slight refactor in hl_fw_send_cpu_message() returns -EAGAIN
when CPU is disabled (i.e. heartbeat failure) and only then.
Later, all calling functions that may be invoked by user space can issue
prints only if the error code is not -EAGAIN.
Signed-off-by: Ohad Sharabi <osharabi@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
The function returned an error code which isn't propagated up the stack
(nor is it printed).
The return value is only checked for =0 or !=0 which implies bool return
value.
The function signature is updated accordingly, renamed, and slightly
refactored.
Signed-off-by: Ohad Sharabi <osharabi@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
