[PATCH 1/9] habanalabs: rename non_hard_reset to compute_reset

[PATCH 1/9] habanalabs: rename non_hard_reset to compute_reset

[Oded Gabbay]

From: Ofir Bitton <obitton@habana.ai>

In order to be more explicit we should use the term compute_reset
for describing the reset in which only the compute engines gets
reset.

Signed-off-by: Ofir Bitton <obitton@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
---
 drivers/misc/habanalabs/common/device.c     | 2 +-
 drivers/misc/habanalabs/common/habanalabs.h | 4 ++--
 drivers/misc/habanalabs/gaudi/gaudi.c       | 4 ++--
 drivers/misc/habanalabs/gaudi2/gaudi2.c     | 4 ++--
 drivers/misc/habanalabs/goya/goya.c         | 4 ++--
 5 files changed, 9 insertions(+), 9 deletions(-)

diff --git a/drivers/misc/habanalabs/common/device.c b/drivers/misc/habanalabs/common/device.c
index b30aeb1c657f..90e346727a7c 100644
--- a/drivers/misc/habanalabs/common/device.c
+++ b/drivers/misc/habanalabs/common/device.c
@@ -1556,7 +1556,7 @@ int hl_device_reset(struct hl_device *hdev, u32 flags)
 		if (!hdev->asic_prop.fw_security_enabled)
 			hl_fw_set_max_power(hdev);
 	} else {
-		rc = hdev->asic_funcs->non_hard_reset_late_init(hdev);
+		rc = hdev->asic_funcs->compute_reset_late_init(hdev);
 		if (rc) {
 			if (reset_upon_device_release)
 				dev_err(hdev->dev,
diff --git a/drivers/misc/habanalabs/common/habanalabs.h b/drivers/misc/habanalabs/common/habanalabs.h
index 4d6b31ea1efb..1e1f2c494423 100644
--- a/drivers/misc/habanalabs/common/habanalabs.h
+++ b/drivers/misc/habanalabs/common/habanalabs.h
@@ -1446,7 +1446,7 @@ struct engines_data {
  * @send_heartbeat: send is-alive packet to CPU-CP and verify response.
  * @debug_coresight: perform certain actions on Coresight for debugging.
  * @is_device_idle: return true if device is idle, false otherwise.
- * @non_hard_reset_late_init: perform certain actions needed after a reset which is not hard-reset
+ * @compute_reset_late_init: perform certain actions needed after a compute reset
  * @hw_queues_lock: acquire H/W queues lock.
  * @hw_queues_unlock: release H/W queues lock.
  * @kdma_lock: acquire H/W queues lock. Relevant from GRECO ASIC
@@ -1584,7 +1584,7 @@ struct hl_asic_funcs {
 	int (*debug_coresight)(struct hl_device *hdev, struct hl_ctx *ctx, void *data);
 	bool (*is_device_idle)(struct hl_device *hdev, u64 *mask_arr, u8 mask_len,
 				struct engines_data *e);
-	int (*non_hard_reset_late_init)(struct hl_device *hdev);
+	int (*compute_reset_late_init)(struct hl_device *hdev);
 	void (*hw_queues_lock)(struct hl_device *hdev);
 	void (*hw_queues_unlock)(struct hl_device *hdev);
 	void (*kdma_lock)(struct hl_device *hdev, int dcore_id);
diff --git a/drivers/misc/habanalabs/gaudi/gaudi.c b/drivers/misc/habanalabs/gaudi/gaudi.c
index a7923960fce1..20f62730be02 100644
--- a/drivers/misc/habanalabs/gaudi/gaudi.c
+++ b/drivers/misc/habanalabs/gaudi/gaudi.c
@@ -7427,7 +7427,7 @@ static void gaudi_print_nic_axi_irq_info(struct hl_device *hdev, u16 event_type,
 		event_type, desc);
 }
 
-static int gaudi_non_hard_reset_late_init(struct hl_device *hdev)
+static int gaudi_compute_reset_late_init(struct hl_device *hdev)
 {
 	/* GAUDI doesn't support any reset except hard-reset */
 	return -EPERM;
@@ -9193,7 +9193,7 @@ static const struct hl_asic_funcs gaudi_funcs = {
 	.send_heartbeat = gaudi_send_heartbeat,
 	.debug_coresight = gaudi_debug_coresight,
 	.is_device_idle = gaudi_is_device_idle,
-	.non_hard_reset_late_init = gaudi_non_hard_reset_late_init,
+	.compute_reset_late_init = gaudi_compute_reset_late_init,
 	.hw_queues_lock = gaudi_hw_queues_lock,
 	.hw_queues_unlock = gaudi_hw_queues_unlock,
 	.kdma_lock = NULL,
diff --git a/drivers/misc/habanalabs/gaudi2/gaudi2.c b/drivers/misc/habanalabs/gaudi2/gaudi2.c
index fd917e837075..ab6ad06cec03 100644
--- a/drivers/misc/habanalabs/gaudi2/gaudi2.c
+++ b/drivers/misc/habanalabs/gaudi2/gaudi2.c
@@ -6129,7 +6129,7 @@ static int gaudi2_test_queues(struct hl_device *hdev)
 	return ret_val;
 }
 
-static int gaudi2_non_hard_reset_late_init(struct hl_device *hdev)
+static int gaudi2_compute_reset_late_init(struct hl_device *hdev)
 {
 	struct gaudi2_device *gaudi2 = hdev->asic_specific;
 	size_t irq_arr_size;
@@ -9930,7 +9930,7 @@ static const struct hl_asic_funcs gaudi2_funcs = {
 	.send_heartbeat = gaudi2_send_heartbeat,
 	.debug_coresight = gaudi2_debug_coresight,
 	.is_device_idle = gaudi2_is_device_idle,
-	.non_hard_reset_late_init = gaudi2_non_hard_reset_late_init,
+	.compute_reset_late_init = gaudi2_compute_reset_late_init,
 	.hw_queues_lock = gaudi2_hw_queues_lock,
 	.hw_queues_unlock = gaudi2_hw_queues_unlock,
 	.kdma_lock = gaudi2_kdma_lock,
diff --git a/drivers/misc/habanalabs/goya/goya.c b/drivers/misc/habanalabs/goya/goya.c
index 7b9f7f8b51f4..d4459c290ea8 100644
--- a/drivers/misc/habanalabs/goya/goya.c
+++ b/drivers/misc/habanalabs/goya/goya.c
@@ -4559,7 +4559,7 @@ static int goya_unmask_irq_arr(struct hl_device *hdev, u32 *irq_arr,
 	return rc;
 }
 
-static int goya_non_hard_reset_late_init(struct hl_device *hdev)
+static int goya_compute_reset_late_init(struct hl_device *hdev)
 {
 	/*
 	 * Unmask all IRQs since some could have been received
@@ -5478,7 +5478,7 @@ static const struct hl_asic_funcs goya_funcs = {
 	.send_heartbeat = goya_send_heartbeat,
 	.debug_coresight = goya_debug_coresight,
 	.is_device_idle = goya_is_device_idle,
-	.non_hard_reset_late_init = goya_non_hard_reset_late_init,
+	.compute_reset_late_init = goya_compute_reset_late_init,
 	.hw_queues_lock = goya_hw_queues_lock,
 	.hw_queues_unlock = goya_hw_queues_unlock,
 	.kdma_lock = NULL,
-- 
2.25.1

[PATCH 2/9] habanalabs/gaudi2: enable all MMU SPI/SEI interrupts

[Oded Gabbay]

From: Tomer Tayar <ttayar@habana.ai>

Currently only part of the MMU SPI/SEI interrupts are enabled, although
there is no real reason to not enable all.
The only exception is "burst_fifo_full" which is expected for PMMU
because it has a 2 entries FIFO, and thus is it not enabled for it.

Signed-off-by: Tomer Tayar <ttayar@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
---
 drivers/misc/habanalabs/gaudi2/gaudi2.c | 14 +++++++++++---
 1 file changed, 11 insertions(+), 3 deletions(-)

diff --git a/drivers/misc/habanalabs/gaudi2/gaudi2.c b/drivers/misc/habanalabs/gaudi2/gaudi2.c
index ab6ad06cec03..465d9c319c3c 100644
--- a/drivers/misc/habanalabs/gaudi2/gaudi2.c
+++ b/drivers/misc/habanalabs/gaudi2/gaudi2.c
@@ -117,6 +117,12 @@
 #define MMU_RANGE_INV_ASID_EN_SHIFT		1
 #define MMU_RANGE_INV_ASID_SHIFT		2
 
+/* The last SPI_SEI cause bit, "burst_fifo_full", is expected to be triggered in PMMU because it has
+ * a 2 entries FIFO, and hence it is not enabled for it.
+ */
+#define GAUDI2_PMMU_SPI_SEI_ENABLE_MASK		GENMASK(GAUDI2_NUM_OF_MMU_SPI_SEI_CAUSE - 2, 0)
+#define GAUDI2_HMMU_SPI_SEI_ENABLE_MASK		GENMASK(GAUDI2_NUM_OF_MMU_SPI_SEI_CAUSE - 1, 0)
+
 #define GAUDI2_MAX_STRING_LEN			64
 
 #define GAUDI2_VDEC_MSIX_ENTRIES		(GAUDI2_IRQ_NUM_SHARED_DEC1_ABNRM - \
@@ -4956,8 +4962,7 @@ static int gaudi2_mmu_update_hop0_addr(struct hl_device *hdev, u32 stlb_base)
 	return 0;
 }
 
-static int gaudi2_mmu_init_common(struct hl_device *hdev, u32 mmu_base,
-					u32 stlb_base)
+static int gaudi2_mmu_init_common(struct hl_device *hdev, u32 mmu_base, u32 stlb_base)
 {
 	u32 status, timeout_usec;
 	int rc;
@@ -4985,7 +4990,6 @@ static int gaudi2_mmu_init_common(struct hl_device *hdev, u32 mmu_base,
 		return rc;
 
 	WREG32(mmu_base + MMU_BYPASS_OFFSET, 0);
-	WREG32(mmu_base + MMU_SPI_SEI_MASK_OFFSET, 0xF);
 
 	rc = hl_poll_timeout(
 		hdev,
@@ -5042,6 +5046,8 @@ static int gaudi2_pci_mmu_init(struct hl_device *hdev)
 			DCORE0_HMMU0_MMU_STATIC_MULTI_PAGE_SIZE_CFG_8_BITS_HOP_MODE_EN_MASK);
 	}
 
+	WREG32(mmu_base + MMU_SPI_SEI_MASK_OFFSET, GAUDI2_PMMU_SPI_SEI_ENABLE_MASK);
+
 	rc = gaudi2_mmu_init_common(hdev, mmu_base, stlb_base);
 	if (rc)
 		return rc;
@@ -5092,6 +5098,8 @@ static int gaudi2_dcore_hmmu_init(struct hl_device *hdev, int dcore_id,
 	RMWREG32(stlb_base + STLB_HOP_CONFIGURATION_OFFSET, 1,
 			STLB_HOP_CONFIGURATION_ONLY_LARGE_PAGE_MASK);
 
+	WREG32(mmu_base + MMU_SPI_SEI_MASK_OFFSET, GAUDI2_HMMU_SPI_SEI_ENABLE_MASK);
+
 	rc = gaudi2_mmu_init_common(hdev, mmu_base, stlb_base);
 	if (rc)
 		return rc;
-- 
2.25.1

[PATCH 3/9] habanalabs: set gnu_printf attribute for hl_engine_data_sprintf

[Oded Gabbay]

From: Dani Liberman <dliberman@habana.ai>

Since hl_engine_data_sprintf receives formatted strings, it should be
compiled as gnu_printf format, so compiler could check the arguments
of the formatted string.

Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Dani Liberman <dliberman@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
---
 drivers/misc/habanalabs/common/habanalabs.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/misc/habanalabs/common/habanalabs.h b/drivers/misc/habanalabs/common/habanalabs.h
index 1e1f2c494423..6d7b66cd50f1 100644
--- a/drivers/misc/habanalabs/common/habanalabs.h
+++ b/drivers/misc/habanalabs/common/habanalabs.h
@@ -3755,7 +3755,7 @@ struct hl_mmap_mem_buf *
 hl_mmap_mem_buf_alloc(struct hl_mem_mgr *mmg,
 		      struct hl_mmap_mem_buf_behavior *behavior, gfp_t gfp,
 		      void *args);
-void hl_engine_data_sprintf(struct engines_data *e, const char *fmt, ...);
+__printf(2, 3) void hl_engine_data_sprintf(struct engines_data *e, const char *fmt, ...);
 
 #ifdef CONFIG_DEBUG_FS
 
-- 
2.25.1

[PATCH 4/9] habanalabs: add return code field to module iterator

[Oded Gabbay]

From: Ohad Sharabi <osharabi@habana.ai>

Up until now the module iterator called void callback functions
and so caller activating callback that may fail suffered from 2 issues:
1. The need to "plant" return called in the private data. This is a
   drawback since the iterator itself should not be aware of the private
   data of the caller.
2. Due to 1 even in a failure the iterator would keep iterating instead
   of break upon error.

To overcome this an optional rc field added to the iterator context.

Signed-off-by: Ohad Sharabi <osharabi@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
---
 drivers/misc/habanalabs/common/habanalabs.h   |  7 +++--
 drivers/misc/habanalabs/gaudi2/gaudi2.c       | 26 +++++++++++++------
 .../misc/habanalabs/gaudi2/gaudi2_security.c  | 19 +++++++-------
 3 files changed, 32 insertions(+), 20 deletions(-)

diff --git a/drivers/misc/habanalabs/common/habanalabs.h b/drivers/misc/habanalabs/common/habanalabs.h
index 6d7b66cd50f1..8c2c94fb1322 100644
--- a/drivers/misc/habanalabs/common/habanalabs.h
+++ b/drivers/misc/habanalabs/common/habanalabs.h
@@ -208,6 +208,7 @@ enum hl_protection_levels {
  * struct iterate_module_ctx - HW module iterator
  * @fn: function to apply to each HW module instance
  * @data: optional internal data to the function iterator
+ * @rc: return code for optional use of iterator/iterator-caller
  */
 struct iterate_module_ctx {
 	/*
@@ -217,10 +218,12 @@ struct iterate_module_ctx {
 	 * @inst: HW module instance within the block
 	 * @offset: current HW module instance offset from the 1-st HW module instance
 	 *          in the 1-st block
-	 * @data: function specific data
+	 * @ctx: the iterator context.
 	 */
-	void (*fn)(struct hl_device *hdev, int block, int inst, u32 offset, void *data);
+	void (*fn)(struct hl_device *hdev, int block, int inst, u32 offset,
+			struct iterate_module_ctx *ctx);
 	void *data;
+	int rc;
 };
 
 struct hl_block_glbl_sec {
diff --git a/drivers/misc/habanalabs/gaudi2/gaudi2.c b/drivers/misc/habanalabs/gaudi2/gaudi2.c
index 465d9c319c3c..3531a339e742 100644
--- a/drivers/misc/habanalabs/gaudi2/gaudi2.c
+++ b/drivers/misc/habanalabs/gaudi2/gaudi2.c
@@ -1712,6 +1712,9 @@ void gaudi2_iterate_tpcs(struct hl_device *hdev, struct iterate_module_ctx *ctx)
 	int dcore, inst, tpc_seq;
 	u32 offset;
 
+	/* init the return code */
+	ctx->rc = 0;
+
 	for (dcore = 0; dcore < NUM_OF_DCORES; dcore++) {
 		for (inst = 0; inst < NUM_OF_TPC_PER_DCORE; inst++) {
 			tpc_seq = dcore * NUM_OF_TPC_PER_DCORE + inst;
@@ -1721,7 +1724,12 @@ void gaudi2_iterate_tpcs(struct hl_device *hdev, struct iterate_module_ctx *ctx)
 
 			offset = (DCORE_OFFSET * dcore) + (DCORE_TPC_OFFSET * inst);
 
-			ctx->fn(hdev, dcore, inst, offset, ctx->data);
+			ctx->fn(hdev, dcore, inst, offset, ctx);
+			if (ctx->rc) {
+				dev_err(hdev->dev, "TPC iterator failed for DCORE%d TPC%d\n",
+							dcore, inst);
+				return;
+			}
 		}
 	}
 
@@ -1730,7 +1738,9 @@ void gaudi2_iterate_tpcs(struct hl_device *hdev, struct iterate_module_ctx *ctx)
 
 	/* special check for PCI TPC (DCORE0_TPC6) */
 	offset = DCORE_TPC_OFFSET * (NUM_DCORE0_TPC - 1);
-	ctx->fn(hdev, 0, NUM_DCORE0_TPC - 1, offset, ctx->data);
+	ctx->fn(hdev, 0, NUM_DCORE0_TPC - 1, offset, ctx);
+	if (ctx->rc)
+		dev_err(hdev->dev, "TPC iterator failed for DCORE0 TPC6\n");
 }
 
 static bool gaudi2_host_phys_addr_valid(u64 addr)
@@ -4507,10 +4517,10 @@ struct gaudi2_tpc_init_cfg_data {
 };
 
 static void gaudi2_init_tpc_config(struct hl_device *hdev, int dcore, int inst,
-							u32 offset, void *data)
+					u32 offset, struct iterate_module_ctx *ctx)
 {
 	struct gaudi2_device *gaudi2 = hdev->asic_specific;
-	struct gaudi2_tpc_init_cfg_data *cfg_data = data;
+	struct gaudi2_tpc_init_cfg_data *cfg_data = ctx->data;
 	u32 queue_id_base;
 	u8 seq;
 
@@ -6155,9 +6165,9 @@ static int gaudi2_compute_reset_late_init(struct hl_device *hdev)
 }
 
 static void gaudi2_is_tpc_engine_idle(struct hl_device *hdev, int dcore, int inst, u32 offset,
-					void *data)
+					struct iterate_module_ctx *ctx)
 {
-	struct gaudi2_tpc_idle_data *idle_data = (struct gaudi2_tpc_idle_data *)data;
+	struct gaudi2_tpc_idle_data *idle_data = ctx->data;
 	u32 tpc_cfg_sts, qm_glbl_sts0, qm_glbl_sts1, qm_cgm_sts;
 	bool is_eng_idle;
 	int engine_idx;
@@ -6736,9 +6746,9 @@ static int gaudi2_mmu_shared_prepare(struct hl_device *hdev, u32 asid)
 }
 
 static void gaudi2_tpc_mmu_prepare(struct hl_device *hdev, int dcore, int inst,	u32 offset,
-					void *data)
+					struct iterate_module_ctx *ctx)
 {
-	struct gaudi2_tpc_mmu_data *mmu_data = (struct gaudi2_tpc_mmu_data *)data;
+	struct gaudi2_tpc_mmu_data *mmu_data = ctx->data;
 
 	WREG32(mmDCORE0_TPC0_CFG_AXUSER_HB_MMU_BP + offset, 0);
 	WREG32(mmDCORE0_TPC0_CFG_AXUSER_HB_ASID + offset, mmu_data->rw_asid);
diff --git a/drivers/misc/habanalabs/gaudi2/gaudi2_security.c b/drivers/misc/habanalabs/gaudi2/gaudi2_security.c
index 89a06ff5ba34..c4165db06db2 100644
--- a/drivers/misc/habanalabs/gaudi2/gaudi2_security.c
+++ b/drivers/misc/habanalabs/gaudi2/gaudi2_security.c
@@ -2583,9 +2583,9 @@ struct gaudi2_tpc_pb_data {
 };
 
 static void gaudi2_config_tpcs_glbl_sec(struct hl_device *hdev, int dcore, int inst, u32 offset,
-					void *data)
+						struct iterate_module_ctx *ctx)
 {
-	struct gaudi2_tpc_pb_data *pb_data = (struct gaudi2_tpc_pb_data *)data;
+	struct gaudi2_tpc_pb_data *pb_data = ctx->data;
 
 	hl_config_glbl_sec(hdev, gaudi2_pb_dcr0_tpc0, pb_data->glbl_sec,
 					offset, pb_data->block_array_size);
@@ -2660,15 +2660,14 @@ static int gaudi2_init_pb_tpc(struct hl_device *hdev)
 struct gaudi2_tpc_arc_pb_data {
 	u32 unsecured_regs_arr_size;
 	u32 arc_regs_arr_size;
-	int rc;
 };
 
 static void gaudi2_config_tpcs_pb_ranges(struct hl_device *hdev, int dcore, int inst, u32 offset,
-					void *data)
+						struct iterate_module_ctx *ctx)
 {
-	struct gaudi2_tpc_arc_pb_data *pb_data = (struct gaudi2_tpc_arc_pb_data *)data;
+	struct gaudi2_tpc_arc_pb_data *pb_data = ctx->data;
 
-	pb_data->rc |= hl_init_pb_ranges(hdev, HL_PB_SHARED, HL_PB_NA, 1,
+	ctx->rc = hl_init_pb_ranges(hdev, HL_PB_SHARED, HL_PB_NA, 1,
 					offset, gaudi2_pb_dcr0_tpc0_arc,
 					pb_data->arc_regs_arr_size,
 					gaudi2_pb_dcr0_tpc0_arc_unsecured_regs,
@@ -2683,12 +2682,12 @@ static int gaudi2_init_pb_tpc_arc(struct hl_device *hdev)
 	tpc_arc_pb_data.arc_regs_arr_size = ARRAY_SIZE(gaudi2_pb_dcr0_tpc0_arc);
 	tpc_arc_pb_data.unsecured_regs_arr_size =
 			ARRAY_SIZE(gaudi2_pb_dcr0_tpc0_arc_unsecured_regs);
-	tpc_arc_pb_data.rc = 0;
+
 	tpc_iter.fn = &gaudi2_config_tpcs_pb_ranges;
 	tpc_iter.data = &tpc_arc_pb_data;
 	gaudi2_iterate_tpcs(hdev, &tpc_iter);
 
-	return tpc_arc_pb_data.rc;
+	return tpc_iter.rc;
 }
 
 static int gaudi2_init_pb_sm_objs(struct hl_device *hdev)
@@ -3547,9 +3546,9 @@ struct gaudi2_ack_pb_tpc_data {
 };
 
 static void gaudi2_ack_pb_tpc_config(struct hl_device *hdev, int dcore, int inst, u32 offset,
-					void *data)
+					struct iterate_module_ctx *ctx)
 {
-	struct gaudi2_ack_pb_tpc_data *pb_data = (struct gaudi2_ack_pb_tpc_data *)data;
+	struct gaudi2_ack_pb_tpc_data *pb_data = ctx->data;
 
 	hl_ack_pb_single_dcore(hdev, offset, HL_PB_SINGLE_INSTANCE, HL_PB_NA,
 				gaudi2_pb_dcr0_tpc0, pb_data->tpc_regs_array_size);
-- 
2.25.1

[PATCH 5/9] habanalabs/gaudi: increase default cs timeout to 10 minutes

[Oded Gabbay]

In order to improve scalability and reduce host overhead, it is better
to increase the default TDR timeout of Gaudi1 from 30 seconds to
10 minutes.

This will allow the DL Framework (e.g. PyTorch, TensorFlow) to remove
the host sync they are using now and improve overall performance on
scaleout training.

Note that one can always set the timeout to a custom value via
a kernel module parameter given during driver load.

Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
---
 .../misc/habanalabs/common/habanalabs_drv.c   | 23 +++++++++++++++----
 1 file changed, 18 insertions(+), 5 deletions(-)

diff --git a/drivers/misc/habanalabs/common/habanalabs_drv.c b/drivers/misc/habanalabs/common/habanalabs_drv.c
index f733ead605e7..d59d8cdf33e6 100644
--- a/drivers/misc/habanalabs/common/habanalabs_drv.c
+++ b/drivers/misc/habanalabs/common/habanalabs_drv.c
@@ -27,7 +27,10 @@ static struct class *hl_class;
 static DEFINE_IDR(hl_devs_idr);
 static DEFINE_MUTEX(hl_devs_idr_lock);
 
-static int timeout_locked = 30;
+#define HL_DEFAULT_TIMEOUT_LOCKED	30	/* 30 seconds */
+#define GAUDI_DEFAULT_TIMEOUT_LOCKED	600	/* 10 minutes */
+
+static int timeout_locked = HL_DEFAULT_TIMEOUT_LOCKED;
 static int reset_on_lockup = 1;
 static int memory_scrub;
 static ulong boot_error_status_mask = ULONG_MAX;
@@ -314,12 +317,22 @@ static void copy_kernel_module_params_to_device(struct hl_device *hdev)
 	hdev->boot_error_status_mask = boot_error_status_mask;
 }
 
-static void fixup_device_params_per_asic(struct hl_device *hdev)
+static void fixup_device_params_per_asic(struct hl_device *hdev, int timeout)
 {
 	switch (hdev->asic_type) {
-	case ASIC_GOYA:
 	case ASIC_GAUDI:
 	case ASIC_GAUDI_SEC:
+		/* If user didn't request a different timeout than the default one, we have
+		 * a different default timeout for Gaudi
+		 */
+		if (timeout == HL_DEFAULT_TIMEOUT_LOCKED)
+			hdev->timeout_jiffies = msecs_to_jiffies(GAUDI_DEFAULT_TIMEOUT_LOCKED *
+										MSEC_PER_SEC);
+
+		hdev->reset_upon_device_release = 0;
+		break;
+
+	case ASIC_GOYA:
 		hdev->reset_upon_device_release = 0;
 		break;
 
@@ -339,7 +352,7 @@ static int fixup_device_params(struct hl_device *hdev)
 	hdev->fw_comms_poll_interval_usec = HL_FW_STATUS_POLL_INTERVAL_USEC;
 
 	if (tmp_timeout)
-		hdev->timeout_jiffies = msecs_to_jiffies(tmp_timeout * 1000);
+		hdev->timeout_jiffies = msecs_to_jiffies(tmp_timeout * MSEC_PER_SEC);
 	else
 		hdev->timeout_jiffies = MAX_SCHEDULE_TIMEOUT;
 
@@ -360,7 +373,7 @@ static int fixup_device_params(struct hl_device *hdev)
 	if (!hdev->cpu_queues_enable)
 		hdev->heartbeat = 0;
 
-	fixup_device_params_per_asic(hdev);
+	fixup_device_params_per_asic(hdev, tmp_timeout);
 
 	return 0;
 }
-- 
2.25.1

[PATCH 6/9] habanalabs/gaudi2: remove old interrupt mappings

[Oded Gabbay]

From: Ofir Bitton <obitton@habana.ai>

Interrupt enumration has changed some time ago but the old mapping
was accidentally left in the driver.

Signed-off-by: Ofir Bitton <obitton@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
---
 drivers/misc/habanalabs/gaudi2/gaudi2P.h      |  3 -
 .../include/gaudi2/gaudi2_async_virt_events.h | 57 -------------------
 2 files changed, 60 deletions(-)
 delete mode 100644 drivers/misc/habanalabs/include/gaudi2/gaudi2_async_virt_events.h

diff --git a/drivers/misc/habanalabs/gaudi2/gaudi2P.h b/drivers/misc/habanalabs/gaudi2/gaudi2P.h
index e4bc4009f05b..5110574a650e 100644
--- a/drivers/misc/habanalabs/gaudi2/gaudi2P.h
+++ b/drivers/misc/habanalabs/gaudi2/gaudi2P.h
@@ -15,7 +15,6 @@
 #include "../include/gaudi2/gaudi2_packets.h"
 #include "../include/gaudi2/gaudi2_fw_if.h"
 #include "../include/gaudi2/gaudi2_async_events.h"
-#include "../include/gaudi2/gaudi2_async_virt_events.h"
 
 #define GAUDI2_LINUX_FW_FILE	"habanalabs/gaudi2/gaudi2-fit.itb"
 #define GAUDI2_BOOT_FIT_FILE	"habanalabs/gaudi2/gaudi2-boot-fit.itb"
@@ -511,8 +510,6 @@ struct dup_block_ctx {
  * @hbm_cfg: HBM subsystem settings
  * @hw_queues_lock_mutex: used by simulator instead of hw_queues_lock.
  * @kdma_lock_mutex: used by simulator instead of kdma_lock.
- * @use_deprecated_event_mappings: use old event mappings which are about to be
- *                                 deprecated
  */
 struct gaudi2_device {
 	int (*cpucp_info_get)(struct hl_device *hdev);
diff --git a/drivers/misc/habanalabs/include/gaudi2/gaudi2_async_virt_events.h b/drivers/misc/habanalabs/include/gaudi2/gaudi2_async_virt_events.h
deleted file mode 100644
index 6d6ed7838a64..000000000000
--- a/drivers/misc/habanalabs/include/gaudi2/gaudi2_async_virt_events.h
+++ /dev/null
@@ -1,57 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0
- *
- * Copyright 2022 HabanaLabs, Ltd.
- * All Rights Reserved.
- *
- */
-
-#ifndef __GAUDI2_ASYNC_VIRT_EVENTS_H_
-#define __GAUDI2_ASYNC_VIRT_EVENTS_H_
-
-enum gaudi2_async_virt_event_id {
-	GAUDI2_EVENT_NIC3_QM1_OLD = 1206,
-	GAUDI2_EVENT_NIC4_QM0_OLD = 1207,
-	GAUDI2_EVENT_NIC4_QM1_OLD = 1208,
-	GAUDI2_EVENT_NIC5_QM0_OLD = 1209,
-	GAUDI2_EVENT_NIC5_QM1_OLD = 1210,
-	GAUDI2_EVENT_NIC6_QM0_OLD = 1211,
-	GAUDI2_EVENT_NIC6_QM1_OLD = 1212,
-	GAUDI2_EVENT_NIC7_QM0_OLD = 1213,
-	GAUDI2_EVENT_NIC7_QM1_OLD = 1214,
-	GAUDI2_EVENT_NIC8_QM0_OLD = 1215,
-	GAUDI2_EVENT_NIC8_QM1_OLD = 1216,
-	GAUDI2_EVENT_NIC9_QM0_OLD = 1217,
-	GAUDI2_EVENT_NIC9_QM1_OLD = 1218,
-	GAUDI2_EVENT_NIC10_QM0_OLD = 1219,
-	GAUDI2_EVENT_NIC10_QM1_OLD = 1220,
-	GAUDI2_EVENT_NIC11_QM0_OLD = 1221,
-	GAUDI2_EVENT_NIC11_QM1_OLD = 1222,
-	GAUDI2_EVENT_CPU_PKT_SANITY_FAILED_OLD = 1223,
-	GAUDI2_EVENT_CPU0_STATUS_NIC0_ENG0_OLD = 1224,
-	GAUDI2_EVENT_CPU0_STATUS_NIC0_ENG1_OLD = 1225,
-	GAUDI2_EVENT_CPU1_STATUS_NIC1_ENG0_OLD = 1226,
-	GAUDI2_EVENT_CPU1_STATUS_NIC1_ENG1_OLD = 1227,
-	GAUDI2_EVENT_CPU2_STATUS_NIC2_ENG0_OLD = 1228,
-	GAUDI2_EVENT_CPU2_STATUS_NIC2_ENG1_OLD = 1229,
-	GAUDI2_EVENT_CPU3_STATUS_NIC3_ENG0_OLD = 1230,
-	GAUDI2_EVENT_CPU3_STATUS_NIC3_ENG1_OLD = 1231,
-	GAUDI2_EVENT_CPU4_STATUS_NIC4_ENG0_OLD = 1232,
-	GAUDI2_EVENT_CPU4_STATUS_NIC4_ENG1_OLD = 1233,
-	GAUDI2_EVENT_CPU5_STATUS_NIC5_ENG0_OLD = 1234,
-	GAUDI2_EVENT_CPU5_STATUS_NIC5_ENG1_OLD = 1235,
-	GAUDI2_EVENT_CPU6_STATUS_NIC6_ENG0_OLD = 1236,
-	GAUDI2_EVENT_CPU6_STATUS_NIC6_ENG1_OLD = 1237,
-	GAUDI2_EVENT_CPU7_STATUS_NIC7_ENG0_OLD = 1238,
-	GAUDI2_EVENT_CPU7_STATUS_NIC7_ENG1_OLD = 1239,
-	GAUDI2_EVENT_CPU8_STATUS_NIC8_ENG0_OLD = 1240,
-	GAUDI2_EVENT_CPU8_STATUS_NIC8_ENG1_OLD = 1241,
-	GAUDI2_EVENT_CPU9_STATUS_NIC9_ENG0_OLD = 1242,
-	GAUDI2_EVENT_CPU9_STATUS_NIC9_ENG1_OLD = 1243,
-	GAUDI2_EVENT_CPU10_STATUS_NIC10_ENG0_OLD = 1244,
-	GAUDI2_EVENT_CPU10_STATUS_NIC10_ENG1_OLD = 1245,
-	GAUDI2_EVENT_CPU11_STATUS_NIC11_ENG0_OLD = 1246,
-	GAUDI2_EVENT_CPU11_STATUS_NIC11_ENG1_OLD = 1247,
-	GAUDI2_EVENT_ARC_DCCM_FULL_OLD = 1248,
-};
-
-#endif /* __GAUDI2_ASYNC_VIRT_EVENTS_H_ */
-- 
2.25.1

[PATCH 7/9] habanalabs: fix spelling mistakes

[Oded Gabbay]

From: Bharat Jauhari <bjauhari@habana.ai>

Cosmetic commit, no logical changes. It just fixes the spelling
mistakes.

Signed-off-by: Bharat Jauhari <bjauhari@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
---
 drivers/misc/habanalabs/common/firmware_if.c |  8 +++---
 drivers/misc/habanalabs/common/habanalabs.h  | 27 ++++++++++----------
 drivers/misc/habanalabs/common/memory_mgr.c  | 10 ++++----
 include/uapi/misc/habanalabs.h               | 10 ++++----
 4 files changed, 27 insertions(+), 28 deletions(-)

diff --git a/drivers/misc/habanalabs/common/firmware_if.c b/drivers/misc/habanalabs/common/firmware_if.c
index 608ca67527a5..58c1eff16df6 100644
--- a/drivers/misc/habanalabs/common/firmware_if.c
+++ b/drivers/misc/habanalabs/common/firmware_if.c
@@ -1514,7 +1514,7 @@ int hl_fw_read_preboot_status(struct hl_device *hdev)
 	hdev->asic_funcs->init_firmware_preload_params(hdev);
 
 	/*
-	 * In order to determine boot method (static VS dymanic) we need to
+	 * In order to determine boot method (static VS dynamic) we need to
 	 * read the boot caps register
 	 */
 	rc = hl_fw_read_preboot_caps(hdev);
@@ -1781,7 +1781,7 @@ int hl_fw_dynamic_send_protocol_cmd(struct hl_device *hdev,
  *
  * @return the CRC32 result
  *
- * NOTE: kernel's CRC32 differ's from standard CRC32 calculation.
+ * NOTE: kernel's CRC32 differs from standard CRC32 calculation.
  *       in order to be aligned we need to flip the bits of both the input
  *       initial CRC and kernel's CRC32 result.
  *       in addition both sides use initial CRC of 0,
@@ -1798,7 +1798,7 @@ static u32 hl_fw_compat_crc32(u8 *data, size_t size)
  *
  * @hdev: pointer to the habanalabs device structure
  * @addr: device address of memory transfer
- * @size: memory transter size
+ * @size: memory transfer size
  * @region: PCI memory region
  *
  * @return 0 on success, otherwise non-zero error code
@@ -2547,7 +2547,7 @@ static int hl_fw_dynamic_init_cpu(struct hl_device *hdev,
 	/*
 	 * when testing FW load (without Linux) on PLDM we don't want to
 	 * wait until boot fit is active as it may take several hours.
-	 * instead, we load the bootfit and let it do all initializations in
+	 * instead, we load the bootfit and let it do all initialization in
 	 * the background.
 	 */
 	if (hdev->pldm && !(hdev->fw_components & FW_TYPE_LINUX))
diff --git a/drivers/misc/habanalabs/common/habanalabs.h b/drivers/misc/habanalabs/common/habanalabs.h
index 8c2c94fb1322..350bc5b9f174 100644
--- a/drivers/misc/habanalabs/common/habanalabs.h
+++ b/drivers/misc/habanalabs/common/habanalabs.h
@@ -94,7 +94,7 @@ struct hl_fpriv;
 #define MMU_HASH_TABLE_BITS		7 /* 1 << 7 buckets */
 
 /**
- * enum hl_mmu_page_table_locaion - mmu page table location
+ * enum hl_mmu_page_table_location - mmu page table location
  * @MMU_DR_PGT: page-table is located on device DRAM.
  * @MMU_HR_PGT: page-table is located on host memory.
  * @MMU_NUM_PGT_LOCATIONS: number of page-table locations currently supported.
@@ -800,7 +800,7 @@ struct hl_fence {
  * @lock: spinlock to protect fence.
  * @hdev: habanalabs device structure.
  * @hw_sob: the H/W SOB used in this signal/wait CS.
- * @encaps_sig_hdl: encaps signals hanlder.
+ * @encaps_sig_hdl: encaps signals handler.
  * @cs_seq: command submission sequence number.
  * @type: type of the CS - signal/wait.
  * @sob_val: the SOB value that is used in this signal/wait CS.
@@ -908,7 +908,7 @@ struct hl_mmap_mem_buf {
  * @size: holds the CB's size.
  * @cs_cnt: holds number of CS that this CB participates in.
  * @is_pool: true if CB was acquired from the pool, false otherwise.
- * @is_internal: internaly allocated
+ * @is_internal: internally allocated
  * @is_mmu_mapped: true if the CB is mapped to the device's MMU.
  */
 struct hl_cb {
@@ -1116,7 +1116,7 @@ struct timestamp_reg_info {
  * @fence: hl fence object for interrupt completion
  * @cq_target_value: CQ target value
  * @cq_kernel_addr: CQ kernel address, to be used in the cq interrupt
- *                  handler for taget value comparison
+ *                  handler for target value comparison
  */
 struct hl_user_pending_interrupt {
 	struct timestamp_reg_info	ts_reg_info;
@@ -1742,10 +1742,10 @@ struct hl_cs_outcome {
 
 /**
  * struct hl_cs_outcome_store - represents a limited store of completed CS outcomes
- * @outcome_map: index of completed CS searcheable by sequence number
+ * @outcome_map: index of completed CS searchable by sequence number
  * @used_list: list of outcome objects currently in use
  * @free_list: list of outcome objects currently not in use
- * @nodes_pool: a static pool of preallocated outcome objects
+ * @nodes_pool: a static pool of pre-allocated outcome objects
  * @db_lock: any operation on the store must take this lock
  */
 struct hl_cs_outcome_store {
@@ -1769,7 +1769,7 @@ struct hl_cs_outcome_store {
  * @refcount: reference counter for the context. Context is released only when
  *		this hits 0l. It is incremented on CS and CS_WAIT.
  * @cs_pending: array of hl fence objects representing pending CS.
- * @outcome_store: storage data structure used to remember ouitcomes of completed
+ * @outcome_store: storage data structure used to remember outcomes of completed
  *                 command submissions for a long time after CS id wraparound.
  * @va_range: holds available virtual addresses for host and dram mappings.
  * @mem_hash_lock: protects the mem_hash.
@@ -1838,7 +1838,6 @@ struct hl_ctx_mgr {
 };
 
 
-
 /*
  * COMMAND SUBMISSIONS
  */
@@ -1904,7 +1903,7 @@ struct hl_userptr {
  * @tdr_active: true if TDR was activated for this CS (to prevent
  *		double TDR activation).
  * @aborted: true if CS was aborted due to some device error.
- * @timestamp: true if a timestmap must be captured upon completion.
+ * @timestamp: true if a timestamp must be captured upon completion.
  * @staged_last: true if this is the last staged CS and needs completion.
  * @staged_first: true if this is the first staged CS and we need to receive
  *                timeout for this CS.
@@ -2229,7 +2228,7 @@ struct hl_info_list {
 
 /**
  * struct hl_debugfs_entry - debugfs dentry wrapper.
- * @info_ent: dentry realted ops.
+ * @info_ent: dentry related ops.
  * @dev_entry: ASIC specific debugfs manager.
  */
 struct hl_debugfs_entry {
@@ -2934,7 +2933,7 @@ struct razwi_info {
  * struct undefined_opcode_info - info about last undefined opcode error
  * @timestamp: timestamp of the undefined opcode error
  * @cb_addr_streams: CB addresses (per stream) that are currently exists in the PQ
- *                   entiers. In case all streams array entries are
+ *                   entries. In case all streams array entries are
  *                   filled with values, it means the execution was in Lower-CP.
  * @cq_addr: the address of the current handled command buffer
  * @cq_size: the size of the current handled command buffer
@@ -2975,7 +2974,7 @@ struct last_error_session_info {
 /**
  * struct hl_reset_info - holds current device reset information.
  * @lock: lock to protect critical reset flows.
- * @compute_reset_cnt: number of compte resets since the driver was loaded.
+ * @compute_reset_cnt: number of compute resets since the driver was loaded.
  * @hard_reset_cnt: number of hard resets since the driver was loaded.
  * @hard_reset_schedule_flags: hard reset is scheduled to after current compute reset,
  *                             here we hold the hard reset flags.
@@ -2986,7 +2985,7 @@ struct last_error_session_info {
  * @hard_reset_pending: is there a hard reset work pending.
  * @curr_reset_cause: saves an enumerated reset cause when a hard reset is
  *                    triggered, and cleared after it is shared with preboot.
- * @prev_reset_trigger: saves the previous trigger which caused a reset, overidden
+ * @prev_reset_trigger: saves the previous trigger which caused a reset, overridden
  *                      with a new value on next reset
  * @reset_trigger_repeated: set if device reset is triggered more than once with
  *                          same cause.
@@ -3064,7 +3063,7 @@ struct hl_reset_info {
  * @hl_chip_info: ASIC's sensors information.
  * @device_status_description: device status description.
  * @hl_debugfs: device's debugfs manager.
- * @cb_pool: list of preallocated CBs.
+ * @cb_pool: list of pre allocated CBs.
  * @cb_pool_lock: protects the CB pool.
  * @internal_cb_pool_virt_addr: internal command buffer pool virtual address.
  * @internal_cb_pool_dma_addr: internal command buffer pool dma address.
diff --git a/drivers/misc/habanalabs/common/memory_mgr.c b/drivers/misc/habanalabs/common/memory_mgr.c
index 56df962d2f3c..1936d653699e 100644
--- a/drivers/misc/habanalabs/common/memory_mgr.c
+++ b/drivers/misc/habanalabs/common/memory_mgr.c
@@ -11,7 +11,7 @@
  * hl_mmap_mem_buf_get - increase the buffer refcount and return a pointer to
  *                        the buffer descriptor.
  *
- * @mmg: parent unifed memory manager
+ * @mmg: parent unified memory manager
  * @handle: requested buffer handle
  *
  * Find the buffer in the store and return a pointer to its descriptor.
@@ -104,7 +104,7 @@ int hl_mmap_mem_buf_put(struct hl_mmap_mem_buf *buf)
  * hl_mmap_mem_buf_put_handle - decrease the reference to the buffer with the
  *                              given handle.
  *
- * @mmg: parent unifed memory manager
+ * @mmg: parent unified memory manager
  * @handle: requested buffer handle
  *
  * Decrease the reference to the buffer, and release it if it was the last one.
@@ -137,7 +137,7 @@ int hl_mmap_mem_buf_put_handle(struct hl_mem_mgr *mmg, u64 handle)
 /**
  * hl_mmap_mem_buf_alloc - allocate a new mappable buffer
  *
- * @mmg: parent unifed memory manager
+ * @mmg: parent unified memory manager
  * @behavior: behavior object describing this buffer polymorphic behavior
  * @gfp: gfp flags to use for the memory allocations
  * @args: additional args passed to behavior->alloc
@@ -222,7 +222,7 @@ static const struct vm_operations_struct hl_mmap_mem_buf_vm_ops = {
 /**
  * hl_mem_mgr_mmap - map the given buffer to the user
  *
- * @mmg: unifed memory manager
+ * @mmg: unified memory manager
  * @vma: the vma object for which mmap was closed.
  * @args: additional args passed to behavior->mmap
  *
@@ -322,7 +322,7 @@ void hl_mem_mgr_init(struct device *dev, struct hl_mem_mgr *mmg)
 /**
  * hl_mem_mgr_fini - release unified memory manager
  *
- * @mmg: parent unifed memory manager
+ * @mmg: parent unified memory manager
  *
  * Release the unified memory manager. Shall be called from an interrupt context.
  */
diff --git a/include/uapi/misc/habanalabs.h b/include/uapi/misc/habanalabs.h
index 5d06d5c74dd1..be06b1307c44 100644
--- a/include/uapi/misc/habanalabs.h
+++ b/include/uapi/misc/habanalabs.h
@@ -876,13 +876,13 @@ struct hl_info_hw_idle {
 	__u32 is_idle;
 	/*
 	 * Bitmask of busy engines.
-	 * Bits definition is according to `enum <chip>_enging_id'.
+	 * Bits definition is according to `enum <chip>_engine_id'.
 	 */
 	__u32 busy_engines_mask;
 
 	/*
 	 * Extended Bitmask of busy engines.
-	 * Bits definition is according to `enum <chip>_enging_id'.
+	 * Bits definition is according to `enum <chip>_engine_id'.
 	 */
 	__u64 busy_engines_mask_ext[HL_BUSY_ENGINES_MASK_EXT_SIZE];
 };
@@ -1078,12 +1078,12 @@ struct hl_info_razwi_event {
  * struct hl_info_undefined_opcode_event - info about last undefined opcode error
  * @timestamp: timestamp of the undefined opcode error
  * @cb_addr_streams: CB addresses (per stream) that are currently exists in the PQ
- *                   entiers. In case all streams array entries are
+ *                   entries. In case all streams array entries are
  *                   filled with values, it means the execution was in Lower-CP.
  * @cq_addr: the address of the current handled command buffer
  * @cq_size: the size of the current handled command buffer
  * @cb_addr_streams_len: num of streams - actual len of cb_addr_streams array.
- *                       should be equal to 1 incase of undefined opcode
+ *                       should be equal to 1 in case of undefined opcode
  *                       in Upper-CP (specific stream) and equal to 4 incase
  *                       of undefined opcode in Lower-CP.
  * @engine_id: engine-id that the error occurred on
@@ -1412,7 +1412,7 @@ struct hl_cs_out {
 
 		/* Valid only when HL_CS_FLAGS_RESERVE_SIGNALS_ONLY is set */
 		struct {
-			/* This is the resereved signal handle id */
+			/* This is the reserved signal handle id */
 			__u32 handle_id;
 
 			/* This is the signals count */
-- 
2.25.1

[PATCH 8/9] habanalabs: wrap macro arg with parentheses

[Oded Gabbay]

From: Ohad Sharabi <osharabi@habana.ai>

The macro argument <val> is cast-ed to u32 in some of the places.
Because this arg can be some arithmetic computation (e.g. address +
offset) the cast should be on the whole expression.

Signed-off-by: Ohad Sharabi <osharabi@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
---
 drivers/misc/habanalabs/common/habanalabs.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/misc/habanalabs/common/habanalabs.h b/drivers/misc/habanalabs/common/habanalabs.h
index 350bc5b9f174..d3efec600458 100644
--- a/drivers/misc/habanalabs/common/habanalabs.h
+++ b/drivers/misc/habanalabs/common/habanalabs.h
@@ -2506,7 +2506,7 @@ void hl_wreg(struct hl_device *hdev, u32 reg, u32 val);
 				break; \
 			(val) = __elbi_read; \
 		} else {\
-			(val) = RREG32((u32)addr); \
+			(val) = RREG32((u32)(addr)); \
 		} \
 		if (cond) \
 			break; \
@@ -2517,7 +2517,7 @@ void hl_wreg(struct hl_device *hdev, u32 reg, u32 val);
 					break; \
 				(val) = __elbi_read; \
 			} else {\
-				(val) = RREG32((u32)addr); \
+				(val) = RREG32((u32)(addr)); \
 			} \
 			break; \
 		} \
-- 
2.25.1

[PATCH 9/9] habanalabs: remove all kdma locks

[Oded Gabbay]

We don't use KDMA concurrently in the driver. The only use is through
debugfs and we don't protect concurrent access through it.

Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>
---
 drivers/misc/habanalabs/common/habanalabs.h |  4 ----
 drivers/misc/habanalabs/gaudi/gaudi.c       |  2 --
 drivers/misc/habanalabs/gaudi2/gaudi2.c     | 23 ---------------------
 drivers/misc/habanalabs/gaudi2/gaudi2P.h    |  3 ---
 drivers/misc/habanalabs/goya/goya.c         |  2 --
 5 files changed, 34 deletions(-)

diff --git a/drivers/misc/habanalabs/common/habanalabs.h b/drivers/misc/habanalabs/common/habanalabs.h
index d3efec600458..8d9e96c6092a 100644
--- a/drivers/misc/habanalabs/common/habanalabs.h
+++ b/drivers/misc/habanalabs/common/habanalabs.h
@@ -1452,8 +1452,6 @@ struct engines_data {
  * @compute_reset_late_init: perform certain actions needed after a compute reset
  * @hw_queues_lock: acquire H/W queues lock.
  * @hw_queues_unlock: release H/W queues lock.
- * @kdma_lock: acquire H/W queues lock. Relevant from GRECO ASIC
- * @kdma_unlock: release H/W queues lock. Relevant from GRECO ASIC
  * @get_pci_id: retrieve PCI ID.
  * @get_eeprom_data: retrieve EEPROM data from F/W.
  * @get_monitor_dump: retrieve monitor registers dump from F/W.
@@ -1590,8 +1588,6 @@ struct hl_asic_funcs {
 	int (*compute_reset_late_init)(struct hl_device *hdev);
 	void (*hw_queues_lock)(struct hl_device *hdev);
 	void (*hw_queues_unlock)(struct hl_device *hdev);
-	void (*kdma_lock)(struct hl_device *hdev, int dcore_id);
-	void (*kdma_unlock)(struct hl_device *hdev, int dcore_id);
 	u32 (*get_pci_id)(struct hl_device *hdev);
 	int (*get_eeprom_data)(struct hl_device *hdev, void *data, size_t max_size);
 	int (*get_monitor_dump)(struct hl_device *hdev, void *data);
diff --git a/drivers/misc/habanalabs/gaudi/gaudi.c b/drivers/misc/habanalabs/gaudi/gaudi.c
index 20f62730be02..4d11efed3e64 100644
--- a/drivers/misc/habanalabs/gaudi/gaudi.c
+++ b/drivers/misc/habanalabs/gaudi/gaudi.c
@@ -9196,8 +9196,6 @@ static const struct hl_asic_funcs gaudi_funcs = {
 	.compute_reset_late_init = gaudi_compute_reset_late_init,
 	.hw_queues_lock = gaudi_hw_queues_lock,
 	.hw_queues_unlock = gaudi_hw_queues_unlock,
-	.kdma_lock = NULL,
-	.kdma_unlock = NULL,
 	.get_pci_id = gaudi_get_pci_id,
 	.get_eeprom_data = gaudi_get_eeprom_data,
 	.get_monitor_dump = gaudi_get_monitor_dump,
diff --git a/drivers/misc/habanalabs/gaudi2/gaudi2.c b/drivers/misc/habanalabs/gaudi2/gaudi2.c
index 3531a339e742..2c43ed403509 100644
--- a/drivers/misc/habanalabs/gaudi2/gaudi2.c
+++ b/drivers/misc/habanalabs/gaudi2/gaudi2.c
@@ -3010,7 +3010,6 @@ static int gaudi2_sw_init(struct hl_device *hdev)
 	}
 
 	spin_lock_init(&gaudi2->hw_queues_lock);
-	spin_lock_init(&gaudi2->kdma_lock);
 
 	gaudi2->scratchpad_kernel_address = hl_asic_dma_alloc_coherent(hdev, PAGE_SIZE,
 							&gaudi2->scratchpad_bus_address,
@@ -6464,22 +6463,6 @@ static void gaudi2_hw_queues_unlock(struct hl_device *hdev)
 	spin_unlock(&gaudi2->hw_queues_lock);
 }
 
-static void gaudi2_kdma_lock(struct hl_device *hdev, int dcore_id)
-	__acquires(&gaudi2->kdma_lock)
-{
-	struct gaudi2_device *gaudi2 = hdev->asic_specific;
-
-	spin_lock(&gaudi2->kdma_lock);
-}
-
-static void gaudi2_kdma_unlock(struct hl_device *hdev, int dcore_id)
-	__releases(&gaudi2->kdma_lock)
-{
-	struct gaudi2_device *gaudi2 = hdev->asic_specific;
-
-	spin_unlock(&gaudi2->kdma_lock);
-}
-
 static u32 gaudi2_get_pci_id(struct hl_device *hdev)
 {
 	return hdev->pdev->device;
@@ -9122,8 +9105,6 @@ static int gaudi2_debugfs_read_dma(struct hl_device *hdev, u64 addr, u32 size, v
 		goto unreserve_va;
 	}
 
-	hdev->asic_funcs->kdma_lock(hdev, 0);
-
 	/* Enable MMU on KDMA */
 	gaudi2_kdma_set_mmbp_asid(hdev, false, ctx->asid);
 
@@ -9151,8 +9132,6 @@ static int gaudi2_debugfs_read_dma(struct hl_device *hdev, u64 addr, u32 size, v
 
 	gaudi2_kdma_set_mmbp_asid(hdev, true, HL_KERNEL_ASID_ID);
 
-	hdev->asic_funcs->kdma_unlock(hdev, 0);
-
 	mutex_lock(&ctx->mmu_lock);
 	hl_mmu_unmap_contiguous(ctx, reserved_va_base, SZ_2M);
 	hl_mmu_invalidate_cache_range(hdev, false, MMU_OP_USERPTR,
@@ -9951,8 +9930,6 @@ static const struct hl_asic_funcs gaudi2_funcs = {
 	.compute_reset_late_init = gaudi2_compute_reset_late_init,
 	.hw_queues_lock = gaudi2_hw_queues_lock,
 	.hw_queues_unlock = gaudi2_hw_queues_unlock,
-	.kdma_lock = gaudi2_kdma_lock,
-	.kdma_unlock = gaudi2_kdma_unlock,
 	.get_pci_id = gaudi2_get_pci_id,
 	.get_eeprom_data = gaudi2_get_eeprom_data,
 	.get_monitor_dump = gaudi2_get_monitor_dump,
diff --git a/drivers/misc/habanalabs/gaudi2/gaudi2P.h b/drivers/misc/habanalabs/gaudi2/gaudi2P.h
index 5110574a650e..347ea1dd78e2 100644
--- a/drivers/misc/habanalabs/gaudi2/gaudi2P.h
+++ b/drivers/misc/habanalabs/gaudi2/gaudi2P.h
@@ -457,7 +457,6 @@ struct dup_block_ctx {
  *                 the user can map.
  * @lfsr_rand_seeds: array of MME ACC random seeds to set.
  * @hw_queues_lock: protects the H/W queues from concurrent access.
- * @kdma_lock: protects the KDMA engine from concurrent access.
  * @scratchpad_kernel_address: general purpose PAGE_SIZE contiguous memory,
  *                             this memory region should be write-only.
  *                             currently used for HBW QMAN writes which is
@@ -509,7 +508,6 @@ struct dup_block_ctx {
  * @flush_db_fifo: flag to force flush DB FIFO after a write.
  * @hbm_cfg: HBM subsystem settings
  * @hw_queues_lock_mutex: used by simulator instead of hw_queues_lock.
- * @kdma_lock_mutex: used by simulator instead of kdma_lock.
  */
 struct gaudi2_device {
 	int (*cpucp_info_get)(struct hl_device *hdev);
@@ -518,7 +516,6 @@ struct gaudi2_device {
 	int				lfsr_rand_seeds[MME_NUM_OF_LFSR_SEEDS];
 
 	spinlock_t			hw_queues_lock;
-	spinlock_t			kdma_lock;
 
 	void				*scratchpad_kernel_address;
 	dma_addr_t			scratchpad_bus_address;
diff --git a/drivers/misc/habanalabs/goya/goya.c b/drivers/misc/habanalabs/goya/goya.c
index d4459c290ea8..91429d6ea037 100644
--- a/drivers/misc/habanalabs/goya/goya.c
+++ b/drivers/misc/habanalabs/goya/goya.c
@@ -5481,8 +5481,6 @@ static const struct hl_asic_funcs goya_funcs = {
 	.compute_reset_late_init = goya_compute_reset_late_init,
 	.hw_queues_lock = goya_hw_queues_lock,
 	.hw_queues_unlock = goya_hw_queues_unlock,
-	.kdma_lock = NULL,
-	.kdma_unlock = NULL,
 	.get_pci_id = goya_get_pci_id,
 	.get_eeprom_data = goya_get_eeprom_data,
 	.get_monitor_dump = goya_get_monitor_dump,
-- 
2.25.1