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 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>
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>
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>
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>
Currently, the code contains 2 types of FW version comparison functions:
- hl_is_fw_sw_ver_[below/equal_or_greater]()
- gaudi2 specific function of the type
gaudi2_is_fw_ver_[below/above]x_y_z()
Moreover, some functions use the inner FW version which shuold be only
stage during development but not version dependencies.
Finally, some tests are done to deprecated FW version to which LKD
should hold no compatibility.
This commit aligns all APIs to a single function that just compares the
version and return an integers indicator (similar in some way to
strcmp()).
In addition, this generic function now considers also the sub-minor FW
version and also remove dead code resulting in deprecated FW versions
compatibility.
Signed-off-by: Ohad Sharabi <osharabi@habana.ai>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
Signed-off-by: Ofir Bitton <obitton@habana.ai>
The reserved memory for FW is currently saved in an ASIC property in
units of MB, just like the value that comes from FW.
Except the fact that it is not clear from the property's name, it means
also that a calculation to actual size is required everywhere that it is
used.
Modify the property to hold the size in bytes.
Signed-off-by: Tomer Tayar <ttayar@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Reviewed-by: Carl Vanderlip <quic_carlv@quicinc.com>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
The glbl error cause handling has a wrong assumption that all error
bits are consecutive.
Fix the handling to check all relevant error bits per ASIC.
Signed-off-by: Tomer Tayar <ttayar@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Reviewed-by: Carl Vanderlip <quic_carlv@quicinc.com>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
The hop size related properties is a MMU properties and not
asic properties.
As for PMMU and HMMU we could have different sizes.
Signed-off-by: Farah Kassabri <fkassabri@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Reviewed-by: Carl Vanderlip <quic_carlv@quicinc.com>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
Currently the HMMU page tables reside in the host memory,
which will cause host access from the device for every page walk.
This can affect PCIe bandwidth in certain scenarios.
To prevent that problem, HMMU page tables will be moved to the device
memory so the miss transaction will read the hops from there instead of
going to the host.
Signed-off-by: Farah Kassabri <fkassabri@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
Fix a warning of a buffer overflow:
‘snprintf’ output between 38 and 47 bytes into a destination of size 32
Signed-off-by: Koby Elbaz <kelbaz@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
User interrupts are MSIx interrupts coming from Gaudi2, that have
specific range of IDs and are assigned to the sole use of the user
process that opened the Gaudi2 device (reminder: there can be only
a single user process running on Gaudi2 at any given time).
The interrupts are allocated and managed by the driver and therefore,
the user expects the driver to initialize them properly, which also
includes setting the affinity to the related CPU cores of the
device's NUMA node to get maximum performance.
Signed-off-by: Dani Liberman <dliberman@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
The device debugfs directory was modified to be named as the
device-name.
This name is the parent device name, i.e. either the PCI address in case
of an ASIC, or the simulator device name in case of a simulator.
This change makes it more difficult for a user to access the debugfs
directory for a specific accel device, because he can't just use the
accel minor id, but he needs to do more device-dependent operations to
get the device name.
To make it easier to get this name, add a 'parent_device' sysfs
attribute that the user can read using the minor id before accessing
debugfs.
Signed-off-by: Tomer Tayar <ttayar@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
User will provide a nonce via the INFO ioctl, and will retrieve
the signed device info generated using given nonce.
Signed-off-by: Moti Haimovski <mhaimovski@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
Gaudi2 with PCI revision ID with the value of '3' represents Gaudi2C
device and should be detected and initialized as Gaudi2.
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
When a PCIe AXI drain event happens, it is possible that the driver
cannot access the device through PCIe, and therefore cannot send a
hard-reset request to FW.
Starting from FW version 1.13, FW will initiate a hard-reset in such
a case without waiting for a reset request from the driver.
Signed-off-by: Tomer Tayar <ttayar@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
Two function stubs were removed in an earlier commit but are now needed
again:
drivers/accel/habanalabs/common/device.c: In function 'hl_device_init':
drivers/accel/habanalabs/common/device.c:2231:14: error: implicit declaration of function 'hl_debugfs_device_init'; did you mean 'drm_debugfs_dev_init'? [-Werror=implicit-function-declaration]
2231 | rc = hl_debugfs_device_init(hdev);
drivers/accel/habanalabs/common/device.c:2367:9: error: implicit declaration of function 'hl_debugfs_device_fini'; did you mean 'hl_debugfs_remove_file'? [-Werror=implicit-function-declaration]
2367 | hl_debugfs_device_fini(hdev);
| ^~~~~~~~~~~~~~~~~~~~~~
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
Traces the DMA [un]map_sgtable using the new traces we added.
Signed-off-by: Ohad Sharabi <osharabi@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
asic_dma_{un}map_single() asic-specific functions are no longer called
from the common code, so delete these functions.
In addition, delete the gaudi2 implementation as they are also not
called.
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
Reviewed-by: Ofir Bitton <obitton@habana.ai>
There are cases such when FW runs MBIST, that preboot is expected to take
longer than the usual. In such cases the firmware reports status
SECURITY_READY/IN_PREBOOT and we extend the timeout waiting for it.
This is currently implemented for Gaudi2 only.
Signed-off-by: Dafna Hirschfeld <dhirschfeld@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
On devices with virtual device memory (Gaudi2 onwards), user can provide
an offset within an allocated device memory from which he wants to
export a dma-buf object.
The offset value is verified by driver, but it is not taken into
consideration when the importer driver maps the dma-buf and the SG table
it prepared.
Add the missing offset handling.
Signed-off-by: Tomer Tayar <ttayar@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
The device_address member of 'struct hl_dmabuf_priv' is used only when
virtual device memory is not supported and dma-buf is exported from
address.
Set the value of this field only when it is relevant, and add "phys" to
its name so it would be clearer that it can't be a device virtual
address.
Signed-off-by: Tomer Tayar <ttayar@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
Currently driver maintain one list for both pending user interrupts
which seeks to wait till CQ reaches it's target value and also the ones
that seeks to get timestamp records when the CQ reaches it's target
value.
This causes delay in handling the waiters which gets higher priority
than the timestamp records.
In order to solve this, let's split the list into two,
one for each case and each one is protected by it's own spinlock.
Waiters will be handled within the interrupt context first,
then the timestamp records will be set.
Freeing the timestamp related memory will be handled in a workqueue.
Signed-off-by: farah kassabri <fkassabri@habana.ai>
Reviewed-by: Tomer Tayar <ttayar@habana.ai>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
Currently we use dynamic allocation inside the irq handler
in order to allocate free node to be used for the free jobs.
This operation is expensive, especially when we deal with large
burst of events records that get released at the same time.
The alternative is to have pre allocated pool of free nodes
and just fetch nodes from this pool at irq handling time instead
of allocating them.
In case the pool becomes full, then the driver will fallback to
dynamic allocations.
As part of the optimization also update the unregister flow
upon re-using a timestamp record, by making the operation much
simpler and quicker. We already have the record in the registration
flow and now we just seek to re-use with different interrupt.
Therefore, no need to look for buffer according to the user handle.
Signed-off-by: farah kassabri <fkassabri@habana.ai>
Reviewed-by: Tomer Tayar <ttayar@habana.ai>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
The 'exported_size' member in 'struct hl_vm_phys_pg_pack' is used to
keep the exported dma-buf size, to be later used when the buffer is
mapped.
However it is possible that the same phys_pg_pack will be exported more
than once, and independently of when the mapping takes place.
Remove this member from the phys_pg_pack structure, and simply use the
size in the dma-buf object as the exported size when mapping.
Signed-off-by: Tomer Tayar <ttayar@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
This is the second patch for applying the eq health check mechanism
which will add support for the interrupt flow for gaudi2 asic.
More info about the interrupt mechanism:
set a dedicated msix for the eq error interrupt, and add
interrupt handler for it.
when FW detects some issue with EQ like EQ_FULL, it'll
raise that interrupt and driver should reset the device.
Driver will inform the FW which msix index to use through
the already existing handshake mechanism which will
send msix info message to fw.
Signed-off-by: farah kassabri <fkassabri@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
Add mechanism for fw eq health check. this will be done using two flows:
using the heartbeat mechanism and raising a dedicated interrupt to
indicate an eq failure like EQ full.
This patch will add implementation for the eq heartbeat for gaudi2 asic.
More info about the heartbeat mechanism:
Expand the heartbeat mechanism to monitor a new event that
will be sent from FW upon receiving heartbeat message.
that way driver can know that the eq is working or not.
Signed-off-by: farah kassabri <fkassabri@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
The CPUCP interface is moved to a shared folder outside of accel as
a pre-requisite to upstream the NIC drivers that will also include
this file.
Signed-off-by: David Meriin <dmeriin@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
To use drm_ioctl(), move the ioctls to the device specific ioctls
range at [DRM_COMMAND_BASE, DRM_COMMAND_END).
Signed-off-by: Tomer Tayar <ttayar@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
Register the compute device as an accel device, and remove the creation
of the habanalabs compute char device.
The IOCTLs in this patch are still handled by the current driver
handler. Moving to DRM IOCTL handling requires moving the IOCTLs
numbers to a specific range, so it will be handled in subsequent
patches.
Signed-off-by: Tomer Tayar <ttayar@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
User gets notification for every engine error report, but he still
lacks the exact engine information. Hence, we allow user to query
for the exact engine reported an error.
Signed-off-by: Ofir Bitton <obitton@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
It is possible for FW to request reserved space in dram.
If the device supports this option, it will retrieve the size from the
f/w and will reserve it.
Currently we add the common code infrastructure to support it.
Signed-off-by: Dani Liberman <dliberman@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
Protect against concurrency of user requesting to register a timestamp
offset (where the driver fills the timestamp when the command submission
has finished executing) to a specific user interrupt ID. The
protection is basically to allow only one timestamp registration
request to be handled at a time.
This is needed because the user can decide to re-use a timestamp
offset (register an already registered offset, to a different
interrupt ID). This means the request will cause the timestamp node to
move from one interrupt list to another interrupt list. In such
scenario, without proper protection, we could end up adding the same
node twice to the interrupts wait lists.
Signed-off-by: farah kassabri <fkassabri@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
Two functions got added with normal prototypes for debugfs, but not
alternative when building without it:
drivers/accel/habanalabs/common/device.c: In function 'hl_device_init':
drivers/accel/habanalabs/common/device.c:2177:14: error: implicit declaration of function 'hl_debugfs_device_init'; did you mean 'hl_debugfs_init'? [-Werror=implicit-function-declaration]
drivers/accel/habanalabs/common/device.c:2305:9: error: implicit declaration of function 'hl_debugfs_device_fini'; did you mean 'hl_debugfs_remove_file'? [-Werror=implicit-function-declaration]
Add stubs for these as well.
Fixes: 3b9abb4fa6 ("accel/habanalabs: expose debugfs files later")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: Tomer Tayar <ttayar@habana.ai>
Acked-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20230609120636.3969045-1-arnd@kernel.org
Moved error info reset code to single function for future use from
other places in the driver.
Signed-off-by: Dani Liberman <dliberman@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
There were a few places where simulator only code got into the upstream.
Remove those places that can confuse other developers.
Fixes: 2a0a839b6a ("habanalabs: extend fatal messages to contain PCI info")
Cc: Moti Haimovski <mhaimovski@habana.ai>
Cc: Dan Carpenter <dan.carpenter@linaro.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
As mmu disable mode is only used for bring-up stages, let's remove this
option and all code related to it.
Signed-off-by: Ofir Bitton <obitton@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
Currently the debugfs root folder and files for a device are created at
an early step, before the device initialization and before the char
device and sysfs files are exposed to user.
As there is no real reason not to do it together with the device
creation, postpone it to be done right afterwards.
The initialization of the debugfs entry structure is left in its
current position because it is used before creating the files.
Signed-off-by: Tomer Tayar <ttayar@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
Currently upon a heartbeat failure, we don't know if the failure
is due to firmware hang or due to a bad PCI link. Hence, we
are reading a PCI config space register with a known value (vendor ID)
so we will know which of the two possibilities caused the heartbeat
failure.
Signed-off-by: Ofir Bitton <obitton@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
Aborting CS completions should be in command_submission.c but aborting
waiting for user interrupts should be in device.c.
This separation is also for adding more abort operations in the future.
Signed-off-by: Koby Elbaz <kelbaz@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
This is done depending on the FW version. The cpucp method is
preferable and saves scratchpads resource.
Signed-off-by: Dafna Hirschfeld <dhirschfeld@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
The fw inner version is less trustable, instead use the fw general
sw release version.
Signed-off-by: Dafna Hirschfeld <dhirschfeld@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
It is not always possible to know the FW's SW version from the inner FW
version. Therefore we should extract the general SW version in addition
to the FW version and use it in functions like
'hl_is_fw_ver_below_1_9' etc.
Signed-off-by: Dafna Hirschfeld <dhirschfeld@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
