[PATCH 1/9] accel/habanalabs: minor cosmetics update to cpucp_if.h

[PATCH 1/9] accel/habanalabs: minor cosmetics update to cpucp_if.h

[Oded Gabbay]

- Update copyright years
- Align comments

Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
---
 include/linux/habanalabs/cpucp_if.h | 17 +++++++----------
 1 file changed, 7 insertions(+), 10 deletions(-)

diff --git a/include/linux/habanalabs/cpucp_if.h b/include/linux/habanalabs/cpucp_if.h
index 84d74c4ee4d3..86ea7c63a0d2 100644
--- a/include/linux/habanalabs/cpucp_if.h
+++ b/include/linux/habanalabs/cpucp_if.h
@@ -1,6 +1,6 @@
 /* SPDX-License-Identifier: GPL-2.0
  *
- * Copyright 2020-2022 HabanaLabs, Ltd.
+ * Copyright 2020-2023 HabanaLabs, Ltd.
  * All Rights Reserved.
  *
  */
@@ -668,18 +668,15 @@ enum pq_init_status {
  *       Obsolete.
  *
  * CPUCP_PACKET_GENERIC_PASSTHROUGH -
- *      Generic opcode for all firmware info that is only passed to host
- *      through the LKD, without getting parsed there.
+ *       Generic opcode for all firmware info that is only passed to host
+ *       through the LKD, without getting parsed there.
  *
  * CPUCP_PACKET_ACTIVE_STATUS_SET -
  *       LKD sends FW indication whether device is free or in use, this indication is reported
  *       also to the BMC.
  *
- * CPUCP_PACKET_REGISTER_INTERRUPTS -
- *       Packet to register interrupts indicating LKD is ready to receive events from FW.
- *
  * CPUCP_PACKET_SOFT_RESET -
- *      Packet to perform soft-reset.
+ *       Packet to perform soft-reset.
  *
  * CPUCP_PACKET_INTS_REGISTER -
  *       Packet to inform FW that queues have been established and LKD is ready to receive
@@ -750,9 +747,9 @@ enum cpucp_packet_id {
 	CPUCP_PACKET_RESERVED11,		/* not used */
 	CPUCP_PACKET_RESERVED12,		/* internal */
 	CPUCP_PACKET_RESERVED13,                /* internal */
-	CPUCP_PACKET_SOFT_RESET,                /* internal */
-	CPUCP_PACKET_INTS_REGISTER,             /* internal */
-	CPUCP_PACKET_ID_MAX                     /* must be last */
+	CPUCP_PACKET_SOFT_RESET,		/* internal */
+	CPUCP_PACKET_INTS_REGISTER,		/* internal */
+	CPUCP_PACKET_ID_MAX			/* must be last */
 };
 
 #define CPUCP_PACKET_FENCE_VAL	0xFE8CE7A5
-- 
2.34.1

[PATCH 2/9] accel/habanalabs: minor cosmetics update to trace file

[Oded Gabbay]

- Update copyright years
- Add missing newline at end of file

Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
---
 include/trace/events/habanalabs.h | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/include/trace/events/habanalabs.h b/include/trace/events/habanalabs.h
index 0d3667399d01..cf85753a33b1 100644
--- a/include/trace/events/habanalabs.h
+++ b/include/trace/events/habanalabs.h
@@ -1,6 +1,6 @@
 /* SPDX-License-Identifier: GPL-2.0
  *
- * Copyright 2016-2021 HabanaLabs, Ltd.
+ * Copyright 2022-2023 HabanaLabs, Ltd.
  * All Rights Reserved.
  *
  */
@@ -209,3 +209,4 @@ DEFINE_EVENT(habanalabs_reg_access_template, habanalabs_elbi_write,
 
 /* This part must be outside protection */
 #include <trace/define_trace.h>
+
-- 
2.34.1

[PATCH 3/9] accel/habanalabs: change Greco to Gaudi2

[Oded Gabbay]

Greco was not upstreamed so no point of mentioning it here.

Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
---
 include/uapi/drm/habanalabs_accel.h | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/include/uapi/drm/habanalabs_accel.h b/include/uapi/drm/habanalabs_accel.h
index dfe47db24ae6..347c7b62e60e 100644
--- a/include/uapi/drm/habanalabs_accel.h
+++ b/include/uapi/drm/habanalabs_accel.h
@@ -1,6 +1,6 @@
 /* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note
  *
- * Copyright 2016-2022 HabanaLabs, Ltd.
+ * Copyright 2016-2023 HabanaLabs, Ltd.
  * All Rights Reserved.
  *
  */
@@ -885,11 +885,11 @@ enum hl_server_type {
  * @dram_enabled: Whether the DRAM is enabled.
  * @security_enabled: Whether security is enabled on device.
  * @mme_master_slave_mode: Indicate whether the MME is working in master/slave
- *                         configuration. Relevant for Greco and later.
+ *                         configuration. Relevant for Gaudi2 and later.
  * @cpucp_version: The CPUCP f/w version.
  * @card_name: The card name as passed by the f/w.
  * @tpc_enabled_mask_ext: Bit-mask that represents which TPCs are enabled.
- *                        Relevant for Greco and later.
+ *                        Relevant for Gaudi2 and later.
  * @dram_page_size: The DRAM physical page size.
  * @edma_enabled_mask: Bit-mask that represents which EDMAs are enabled.
  *                     Relevant for Gaudi2 and later.
@@ -1425,7 +1425,7 @@ union hl_cb_args {
  *
  * HL_CS_CHUNK_FLAGS_USER_ALLOC_CB:
  *      Indicates if the CB was allocated and mapped by userspace
- *      (relevant to greco and above). User allocated CB is a command buffer,
+ *      (relevant to Gaudi2 and later). User allocated CB is a command buffer,
  *      allocated by the user, via malloc (or similar). After allocating the
  *      CB, the user invokes - “memory ioctl” to map the user memory into a
  *      device virtual address. The user provides this address via the
@@ -1450,7 +1450,7 @@ struct hl_cs_chunk {
 		 * a DRAM address of the internal CB. In Gaudi, this might also
 		 * represent a mapped host address of the CB.
 		 *
-		 * Greco onwards:
+		 * Gaudi2 onwards:
 		 * For H/W queue, this represents either a Handle of CB on the
 		 * Host, or an SRAM, a DRAM, or a mapped host address of the CB.
 		 *
@@ -2227,7 +2227,7 @@ struct hl_debug_args {
  * internal. The driver will get completion notifications from the device only
  * on JOBS which are enqueued in the external queues.
  *
- * Greco onwards:
+ * Gaudi2 onwards:
  * There is a single type of queue for all types of engines, either DMA engines
  * for transfers from/to the host or inside the device, or compute engines.
  * The driver will get completion notifications from the device for all queues.
-- 
2.34.1

[PATCH 4/9] accel/habanalabs/gaudi: remove unused structure definition

[Oded Gabbay]

struct gaudi_nic_status is not used anywhere in the code.

Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
---
 .../habanalabs/include/gaudi/gaudi_fw_if.h    | 32 -------------------
 1 file changed, 32 deletions(-)

diff --git a/drivers/accel/habanalabs/include/gaudi/gaudi_fw_if.h b/drivers/accel/habanalabs/include/gaudi/gaudi_fw_if.h
index 2dba02757d37..a2547f306750 100644
--- a/drivers/accel/habanalabs/include/gaudi/gaudi_fw_if.h
+++ b/drivers/accel/habanalabs/include/gaudi/gaudi_fw_if.h
@@ -44,38 +44,6 @@ struct eq_nic_sei_event {
 	__u8 pad[6];
 };
 
-/*
- * struct gaudi_nic_status - describes the status of a NIC port.
- * @port: NIC port index.
- * @bad_format_cnt: e.g. CRC.
- * @responder_out_of_sequence_psn_cnt: e.g NAK.
- * @high_ber_reinit_cnt: link reinit due to high BER.
- * @correctable_err_cnt: e.g. bit-flip.
- * @uncorrectable_err_cnt: e.g. MAC errors.
- * @retraining_cnt: re-training counter.
- * @up: is port up.
- * @pcs_link: has PCS link.
- * @phy_ready: is PHY ready.
- * @auto_neg: is Autoneg enabled.
- * @timeout_retransmission_cnt: timeout retransmission events
- * @high_ber_cnt: high ber events
- */
-struct gaudi_nic_status {
-	__u32 port;
-	__u32 bad_format_cnt;
-	__u32 responder_out_of_sequence_psn_cnt;
-	__u32 high_ber_reinit;
-	__u32 correctable_err_cnt;
-	__u32 uncorrectable_err_cnt;
-	__u32 retraining_cnt;
-	__u8 up;
-	__u8 pcs_link;
-	__u8 phy_ready;
-	__u8 auto_neg;
-	__u32 timeout_retransmission_cnt;
-	__u32 high_ber_cnt;
-};
-
 struct gaudi_cold_rst_data {
 	union {
 		struct {
-- 
2.34.1

[PATCH 5/9] accel/habanalabs: remove unused field

[Oded Gabbay]

flags in struct wait_interrupt_data is not used anywhere so remove it.

Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
---
 drivers/accel/habanalabs/common/command_submission.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/drivers/accel/habanalabs/common/command_submission.c b/drivers/accel/habanalabs/common/command_submission.c
index 751d2c7d3fb8..e5ca490b7fca 100644
--- a/drivers/accel/habanalabs/common/command_submission.c
+++ b/drivers/accel/habanalabs/common/command_submission.c
@@ -47,7 +47,6 @@ struct wait_interrupt_data {
 	u64 cq_offset;
 	u64 target_value;
 	u64 intr_timeout_us;
-	unsigned long flags;
 };
 
 static void job_wq_completion(struct work_struct *work);
-- 
2.34.1

[PATCH 6/9] accel/habanalabs: print device name when it is removed

[Oded Gabbay]

Notifies the user which device was removed. It is important in
a server with multiple devices.

Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
---
 drivers/accel/habanalabs/common/device.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/accel/habanalabs/common/device.c b/drivers/accel/habanalabs/common/device.c
index b1d330cd2537..9711e8fc979d 100644
--- a/drivers/accel/habanalabs/common/device.c
+++ b/drivers/accel/habanalabs/common/device.c
@@ -2420,7 +2420,7 @@ void hl_device_fini(struct hl_device *hdev)
 	u64 reset_sec;
 	int i, rc;
 
-	dev_info(hdev->dev, "Removing device\n");
+	dev_info(hdev->dev, "Removing device %s\n", dev_name(&(hdev)->pdev->dev));
 
 	hdev->device_fini_pending = 1;
 	flush_delayed_work(&hdev->device_reset_work.reset_work);
-- 
2.34.1

[PATCH 7/9] accel/habanalabs: remove leftover code

[Oded Gabbay]

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>
---
 drivers/accel/habanalabs/common/command_submission.c | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)

diff --git a/drivers/accel/habanalabs/common/command_submission.c b/drivers/accel/habanalabs/common/command_submission.c
index e5ca490b7fca..4f7b70d9754c 100644
--- a/drivers/accel/habanalabs/common/command_submission.c
+++ b/drivers/accel/habanalabs/common/command_submission.c
@@ -1751,16 +1751,11 @@ static int hl_cs_ctx_switch(struct hl_fpriv *hpriv, union hl_cs_args *args,
 		/* Need to wait for restore completion before execution phase */
 		if (num_chunks) {
 			enum hl_cs_wait_status status;
-wait_again:
+
 			ret = _hl_cs_wait_ioctl(hdev, ctx,
 					jiffies_to_usecs(hdev->timeout_jiffies),
 					*cs_seq, &status, NULL);
 			if (ret) {
-				if (ret == -ERESTARTSYS) {
-					usleep_range(100, 200);
-					goto wait_again;
-				}
-
 				dev_err(hdev->dev,
 					"Restore CS for context %d failed to complete %d\n",
 					ctx->asid, ret);
-- 
2.34.1

[PATCH 8/9] accel/habanalabs/gaudi: remove define used for simulator

[Oded Gabbay]

We don't support simulator in upstream.

Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
---
 drivers/accel/habanalabs/common/habanalabs.h | 2 --
 drivers/accel/habanalabs/gaudi/gaudi.c       | 3 +--
 2 files changed, 1 insertion(+), 4 deletions(-)

diff --git a/drivers/accel/habanalabs/common/habanalabs.h b/drivers/accel/habanalabs/common/habanalabs.h
index b1b1c51aa5b1..a8ccc04e7f92 100644
--- a/drivers/accel/habanalabs/common/habanalabs.h
+++ b/drivers/accel/habanalabs/common/habanalabs.h
@@ -85,8 +85,6 @@ struct hl_fpriv;
 
 #define HL_PCI_ELBI_TIMEOUT_MSEC	10 /* 10ms */
 
-#define HL_SIM_MAX_TIMEOUT_US		100000000 /* 100s */
-
 #define HL_INVALID_QUEUE		UINT_MAX
 
 #define HL_COMMON_USER_CQ_INTERRUPT_ID	0xFFF
diff --git a/drivers/accel/habanalabs/gaudi/gaudi.c b/drivers/accel/habanalabs/gaudi/gaudi.c
index 31e04af46876..53292d4c15c8 100644
--- a/drivers/accel/habanalabs/gaudi/gaudi.c
+++ b/drivers/accel/habanalabs/gaudi/gaudi.c
@@ -4623,8 +4623,7 @@ static int gaudi_scrub_device_dram(struct hl_device *hdev, u64 val)
 static int gaudi_scrub_device_mem(struct hl_device *hdev)
 {
 	struct asic_fixed_properties *prop = &hdev->asic_prop;
-	u64 wait_to_idle_time = hdev->pdev ? HBM_SCRUBBING_TIMEOUT_US :
-			min_t(u64, HBM_SCRUBBING_TIMEOUT_US * 10, HL_SIM_MAX_TIMEOUT_US);
+	u64 wait_to_idle_time = HBM_SCRUBBING_TIMEOUT_US;
 	u64 addr, size, val = hdev->memory_scrub_val;
 	ktime_t timeout;
 	int rc = 0;
-- 
2.34.1

[PATCH 9/9] accel/habanalabs: minor cosmetic update to habanalabs.h

[Oded Gabbay]

- Update copyright years
- Align fields in struct hl_userptr
- Fix comments

Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
---
 drivers/accel/habanalabs/common/habanalabs.h | 31 ++++++++++----------
 1 file changed, 15 insertions(+), 16 deletions(-)

diff --git a/drivers/accel/habanalabs/common/habanalabs.h b/drivers/accel/habanalabs/common/habanalabs.h
index a8ccc04e7f92..2a3acdbf9171 100644
--- a/drivers/accel/habanalabs/common/habanalabs.h
+++ b/drivers/accel/habanalabs/common/habanalabs.h
@@ -1,6 +1,6 @@
 /* SPDX-License-Identifier: GPL-2.0
  *
- * Copyright 2016-2022 HabanaLabs, Ltd.
+ * Copyright 2016-2023 HabanaLabs, Ltd.
  * All Rights Reserved.
  *
  */
@@ -553,8 +553,7 @@ struct hl_hints_range {
  *              allocated with huge pages.
  * @hints_dram_reserved_va_range: dram hint addresses reserved range.
  * @hints_host_reserved_va_range: host hint addresses reserved range.
- * @hints_host_hpage_reserved_va_range: host huge page hint addresses reserved
- *                                      range.
+ * @hints_host_hpage_reserved_va_range: host huge page hint addresses reserved range.
  * @sram_base_address: SRAM physical start address.
  * @sram_end_address: SRAM physical end address.
  * @sram_user_base_address - SRAM physical start address for user access.
@@ -593,7 +592,7 @@ struct hl_hints_range {
  * @mmu_pte_size: PTE size in MMU page tables.
  * @mmu_hop_table_size: MMU hop table size.
  * @mmu_hop0_tables_total_size: total size of MMU hop0 tables.
- * @dram_page_size: page size for MMU DRAM allocation.
+ * @dram_page_size: The DRAM physical page size.
  * @cfg_size: configuration space size on SRAM.
  * @sram_size: total size of SRAM.
  * @max_asid: maximum number of open contexts (ASIDs).
@@ -695,7 +694,7 @@ struct hl_hints_range {
  * @configurable_stop_on_err: is stop-on-error option configurable via debugfs.
  * @set_max_power_on_device_init: true if need to set max power in F/W on device init.
  * @supports_user_set_page_size: true if user can set the allocation page size.
- * @dma_mask: the dma mask to be set for this device
+ * @dma_mask: the dma mask to be set for this device.
  * @supports_advanced_cpucp_rc: true if new cpucp opcodes are supported.
  * @supports_engine_modes: true if changing engines/engine_cores modes is supported.
  * @support_dynamic_resereved_fw_size: true if we support dynamic reserved size for fw.
@@ -1959,17 +1958,17 @@ struct hl_ctx_mgr {
  * @dma_mapped: true if the SG was mapped to DMA addresses, false otherwise.
  */
 struct hl_userptr {
-	enum vm_type		vm_type; /* must be first */
-	struct list_head	job_node;
-	struct page		**pages;
-	unsigned int		npages;
-	struct sg_table		*sgt;
-	enum dma_data_direction dir;
-	struct list_head	debugfs_list;
-	pid_t			pid;
-	u64			addr;
-	u64			size;
-	u8			dma_mapped;
+	enum vm_type			vm_type; /* must be first */
+	struct list_head		job_node;
+	struct page			**pages;
+	unsigned int			npages;
+	struct sg_table			*sgt;
+	enum dma_data_direction		dir;
+	struct list_head		debugfs_list;
+	pid_t				pid;
+	u64				addr;
+	u64				size;
+	u8				dma_mapped;
 };
 
 /**
-- 
2.34.1

Re: [PATCH 1/9] accel/habanalabs: minor cosmetics update to cpucp_if.h

[Ofir Bitton]

On 26/09/2023 11:13, Oded Gabbay wrote:
> - Update copyright years
> - Align comments
> 
> Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
> ---
>   include/linux/habanalabs/cpucp_if.h | 17 +++++++----------
>   1 file changed, 7 insertions(+), 10 deletions(-)
> 
> diff --git a/include/linux/habanalabs/cpucp_if.h b/include/linux/habanalabs/cpucp_if.h
> index 84d74c4ee4d3..86ea7c63a0d2 100644
> --- a/include/linux/habanalabs/cpucp_if.h
> +++ b/include/linux/habanalabs/cpucp_if.h
> @@ -1,6 +1,6 @@
>   /* SPDX-License-Identifier: GPL-2.0
>    *
> - * Copyright 2020-2022 HabanaLabs, Ltd.
> + * Copyright 2020-2023 HabanaLabs, Ltd.
>    * All Rights Reserved.
>    *
>    */
> @@ -668,18 +668,15 @@ enum pq_init_status {
>    *       Obsolete.
>    *
>    * CPUCP_PACKET_GENERIC_PASSTHROUGH -
> - *      Generic opcode for all firmware info that is only passed to host
> - *      through the LKD, without getting parsed there.
> + *       Generic opcode for all firmware info that is only passed to host
> + *       through the LKD, without getting parsed there.
>    *
>    * CPUCP_PACKET_ACTIVE_STATUS_SET -
>    *       LKD sends FW indication whether device is free or in use, this indication is reported
>    *       also to the BMC.
>    *
> - * CPUCP_PACKET_REGISTER_INTERRUPTS -
> - *       Packet to register interrupts indicating LKD is ready to receive events from FW.
> - *
>    * CPUCP_PACKET_SOFT_RESET -
> - *      Packet to perform soft-reset.
> + *       Packet to perform soft-reset.
>    *
>    * CPUCP_PACKET_INTS_REGISTER -
>    *       Packet to inform FW that queues have been established and LKD is ready to receive
> @@ -750,9 +747,9 @@ enum cpucp_packet_id {
>   	CPUCP_PACKET_RESERVED11,		/* not used */
>   	CPUCP_PACKET_RESERVED12,		/* internal */
>   	CPUCP_PACKET_RESERVED13,                /* internal */
> -	CPUCP_PACKET_SOFT_RESET,                /* internal */
> -	CPUCP_PACKET_INTS_REGISTER,             /* internal */
> -	CPUCP_PACKET_ID_MAX                     /* must be last */
> +	CPUCP_PACKET_SOFT_RESET,		/* internal */
> +	CPUCP_PACKET_INTS_REGISTER,		/* internal */
> +	CPUCP_PACKET_ID_MAX			/* must be last */
>   };
>   
>   #define CPUCP_PACKET_FENCE_VAL	0xFE8CE7A5

Ack for the whole series.

Reviewed-by: Ofir Bitton <obitton@habana.ai>