#
8a5be2b6 |
|
28-Dec-2023 |
Malkoot Khan <engr.mkhan1990@gmail.com> |
accel/habanalabs: Remove unnecessary braces from if statement The coding style in the Linux kernel prefers not to use braces for single-statement if conditions. This patch removes the unnecessary braces from an if statement in the file drivers/accel/habanalabs/common/command_submission.c, which also resolves a coding style warning. Signed-off-by: Malkoot Khan <engr.mkhan1990@gmail.com> Reviewed-by: Oded Gabbay <ogabbay@kernel.org> Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
|
#
84190b92 |
|
19-Sep-2023 |
farah kassabri <fkassabri@habana.ai> |
accel/habanalabs: fix bug in decoder wait for cs completion The decoder interrupts are handled in the interrupt context same as all user interrupts. In such case, the wait list should be protected by spin_lock_irqsave in order to avoid deadlock that might happen with the user submission flow. Signed-off-by: farah kassabri <fkassabri@habana.ai> Reviewed-by: Oded Gabbay <ogabbay@kernel.org> Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
|
#
87c60e23 |
|
21-Sep-2023 |
Oded Gabbay <ogabbay@kernel.org> |
accel/habanalabs: remove leftover code This code was added as part of a bigger feature which was never upstreamed, so remove this code. Signed-off-by: Oded Gabbay <ogabbay@kernel.org> Reviewed-by: Ofir Bitton <obitton@habana.ai>
|
#
e5873f6b |
|
21-Sep-2023 |
Oded Gabbay <ogabbay@kernel.org> |
accel/habanalabs: remove unused field flags in struct wait_interrupt_data is not used anywhere so remove it. Signed-off-by: Oded Gabbay <ogabbay@kernel.org> Reviewed-by: Ofir Bitton <obitton@habana.ai>
|
#
ba24b5ec |
|
19-Jul-2023 |
farah kassabri <fkassabri@habana.ai> |
accel/habanalabs: split user interrupts pending list 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>
|
#
1157b5d6 |
|
02-Mar-2023 |
farah kassabri <fkassabri@habana.ai> |
accel/habanalabs: optimize timestamp registration handler 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>
|
#
57963ff8 |
|
17-Apr-2023 |
Tomer Tayar <ttayar@habana.ai> |
accel/habanalabs: Move ioctls to the device specific ioctls range 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>
|
#
2da9f8d8 |
|
12-Jun-2023 |
farah kassabri <fkassabri@habana.ai> |
accel/habanalabs: fix wait_for_interrupt abortion flow When the driver needs to abort waiters for interrupts, for cases such as critical events that occur and driver need to do hard reset, in such scenario the driver will complete the fence to wake up the waiting thread, and will set the fence error indication. The return value of the completion API will be greater than 0 since it will return the timeout, but as this indicates successful completion, the driver should mark it as aborted. Signed-off-by: farah kassabri <fkassabri@habana.ai> Reviewed-by: Oded Gabbay <ogabbay@kernel.org> Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
|
#
eaa43a06 |
|
12-Jun-2023 |
farah kassabri <fkassabri@habana.ai> |
accel/habanalabs: Allow single timestamp registration request at a time 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>
|
#
583f12a8 |
|
09-May-2023 |
Ofir Bitton <obitton@habana.ai> |
accel/habanalabs: remove support for mmu disable 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>
|
#
adda800c |
|
03-May-2023 |
Oded Gabbay <ogabbay@kernel.org> |
accel/habanalabs: print max timeout value on CS stuck If a workload got stuck, we print an error to the kernel log about it. Add to that print the configured max timeout value, as that value is not fixed between ASICs and in addition it can be configured using a kernel module parameter. Signed-off-by: Oded Gabbay <ogabbay@kernel.org> Reviewed-by: Ofir Bitton <obitton@habana.ai>
|
#
9a4e44a4 |
|
04-Dec-2022 |
Koby Elbaz <kelbaz@habana.ai> |
accel/habanalabs: refactor abort of completions and waits 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>
|
#
ad8bfd36 |
|
17-Apr-2023 |
Koby Elbaz <kelbaz@habana.ai> |
accel/habanalabs: minimize encapsulation signal mutex lock time Sync Stream Encapsulated Signal Handlers can be managed from different contexts, and as such they are protected via a spin_lock. However, spin_lock was unnecessarily protecting a larger code section than really needed, covering a sleepable code section as well. Since spin_lock disables preemption, it could lead to sleeping in atomic context. Signed-off-by: Koby Elbaz <kelbaz@habana.ai> Reviewed-by: Oded Gabbay <ogabbay@kernel.org> Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
|
#
f7f0085e |
|
15-Feb-2023 |
Koby Elbaz <kelbaz@habana.ai> |
accel/habanalabs: add uapi to stall/resume engine The user might want to stall/resume engines to perform power testing for various scenarios. Because our current HL_CS_FLAGS_ENGINE_CORE_COMMAND command only handles the engines' cores, we need to add another opcode for handling entire engine and not just its core. The user supplies an array, where each entry holds the engine's ID and the command to send to the engine. The size of the array is limited by the number of engines in the ASIC (only Gaudi2 is currently supported). Signed-off-by: Koby Elbaz <kelbaz@habana.ai> Reviewed-by: Oded Gabbay <ogabbay@kernel.org> Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
|
#
25ebbc57 |
|
23-Feb-2023 |
farah kassabri <fkassabri@habana.ai> |
accel/habanalabs: fix few misspelled words in the code Run spell checker on the code and fix accordingly. Signed-off-by: farah kassabri <fkassabri@habana.ai> Reviewed-by: Oded Gabbay <ogabbay@kernel.org> Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
|
#
60122358 |
|
25-Jan-2023 |
Tal Cohen <talcohen@habana.ai> |
accel/habanalabs: change user interrupt to threaded IRQ We prefer not to handle the user interrupt job inside the interrupt context. Instead, use threaded IRQ to handle the user interrupts. This will allow to avoid disabling interrupts when the user process registers for a new event and to avoid long handling inside an interrupt. Signed-off-by: Tal Cohen <talcohen@habana.ai> Reviewed-by: Oded Gabbay <ogabbay@kernel.org> Signed-off-by: Oded Gabbay <ogabbay@kernel.org> Reviewed-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
|
#
d43bce6e |
|
18-Jan-2023 |
Tomer Tayar <ttayar@habana.ai> |
accel/habanalabs: add info when FD released while device still in use When user closes the device file descriptor, it is checked whether the device is still in use, and a message is printed if it is. To make this message more informative, add to this print also the reason due to which the device is considered as in use. The possible reasons which are checked for now are active CS and exported dma-buf. Signed-off-by: Tomer Tayar <ttayar@habana.ai> Reviewed-by: Oded Gabbay <ogabbay@kernel.org> Signed-off-by: Oded Gabbay <ogabbay@kernel.org> Reviewed-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
|
#
6168a837 |
|
18-Jan-2023 |
Ofir Bitton <obitton@habana.ai> |
accel/habanalabs: increase user interrupt grace time Currently we support scenarios where a timestamp registration request of a certain offset is received during the interrupt handling of the same offset. In this case we give a grace period of up to 100us for the interrupt handler to finish. It seems that sometimes the interrupt handling takes more than expected, and therefore this path should be optimized. Until that happens, let's increase the grace period in order not to reach timeout which will cause user call to be rejected. Signed-off-by: Ofir Bitton <obitton@habana.ai> Reviewed-by: Oded Gabbay <ogabbay@kernel.org> Signed-off-by: Oded Gabbay <ogabbay@kernel.org> Reviewed-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
|
#
75b6984e |
|
10-Jan-2023 |
Ofir Bitton <obitton@habana.ai> |
habanalabs: optimize command submission completion timestamp Completion timestamp is taken during the actual command submission release. As the release happens in a work queue, the timestamp taken is not accurate. Hence, we will take the timestamp in the interrupt handler itself while propagating it to the release function. Signed-off-by: Ofir Bitton <obitton@habana.ai> Reviewed-by: Oded Gabbay <ogabbay@kernel.org> Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
|
#
ac5af990 |
|
09-Jan-2023 |
farah kassabri <fkassabri@habana.ai> |
habanalabs: fix bug in timestamps registration code Protect re-using the same timestamp buffer record before actually adding it to the to interrupt wait list. Mark ts buff offset as in use in the spinlock protection area of the interrupt wait list to avoid getting in the re-use section in ts_buff_get_kernel_ts_record before adding the node to the list. this scenario might happen when multiple threads are racing on same offset and one thread could set data in the ts buff in ts_buff_get_kernel_ts_record then the other thread takes over and get to ts_buff_get_kernel_ts_record and we will try to re-use the same ts buff offset then we will try to delete a non existing node from the list. Signed-off-by: farah kassabri <fkassabri@habana.ai> Reviewed-by: Oded Gabbay <ogabbay@kernel.org> Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
|
#
72848de0 |
|
03-Jan-2023 |
farah kassabri <fkassabri@habana.ai> |
habanalabs: check pad and reserved fields in ioctls Make sure all reserved/pad fields in uapi input structures are set to 0. Signed-off-by: farah kassabri <fkassabri@habana.ai> Reviewed-by: Oded Gabbay <ogabbay@kernel.org> Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
|
#
20faaeec |
|
18-Dec-2022 |
Ohad Sharabi <osharabi@habana.ai> |
habanalabs: add uapi to flush inbound HBM transactions When doing p2p with a NIC device, the NIC needs to make sure all the writes to the HBM (through the PCI bar of the Gaudi device) were flushed. It can be done by either the NIC or the host reading through the PCI bar. To support the host side, we supply a simple uapi to perform this flush through the driver, because the user can't create such a transaction by itself (the PCI bar isn't exposed to normal users). Signed-off-by: Ohad Sharabi <osharabi@habana.ai> Reviewed-by: Oded Gabbay <ogabbay@kernel.org> Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
|
#
e65e175b |
|
26-Dec-2022 |
Oded Gabbay <ogabbay@kernel.org> |
habanalabs: move driver to accel subsystem Now that we have a subsystem for compute accelerators, move the habanalabs driver to it. This patch only moves the files and fixes the Makefiles. Future patches will change the existing code to register to the accel subsystem and expose the accel device char files instead of the habanalabs device char files. Update the MAINTAINERS file to reflect this change. Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
|