habanalabs: set max power according to card type

In Gaudi, the default max power setting is different between PCI and PMC
cards. Therefore, the driver need to set the default after knowing what is
the card type.

The current code has a bug where it limits the maximum power of the PMC
card to 200W after a reset occurs.

Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>

diff --git a/drivers/misc/habanalabs/common/device.c b/drivers/misc/habanalabs/common/device.c
index be16b75..8e34c39 100644 (file)
--- a/drivers/misc/habanalabs/common/device.c
+++ b/drivers/misc/habanalabs/common/device.c
@@ -1069,7 +1069,7 @@ again:
                        goto out_err;
                }
 
-               hl_set_max_power(hdev, hdev->max_power);
+               hl_set_max_power(hdev);
        } else {
                rc = hdev->asic_funcs->soft_reset_late_init(hdev);
                if (rc) {
@@ -1318,6 +1318,11 @@ int hl_device_init(struct hl_device *hdev, struct class *hclass)
                goto out_disabled;
        }
 
+       /* Need to call this again because the max power might change,
+        * depending on card type for certain ASICs
+        */
+       hl_set_max_power(hdev);
+
        /*
         * hl_hwmon_init() must be called after device_late_init(), because only
         * there we get the information from the device about which

diff --git a/drivers/misc/habanalabs/common/habanalabs.h b/drivers/misc/habanalabs/common/habanalabs.h
index 13c18f3..edbd627 100644 (file)
--- a/drivers/misc/habanalabs/common/habanalabs.h
+++ b/drivers/misc/habanalabs/common/habanalabs.h
@@ -1462,6 +1462,8 @@ struct hl_device_idle_busy_ts {
  *                     details.
  * @in_reset: is device in reset flow.
  * @curr_pll_profile: current PLL profile.
+ * @card_type: Various ASICs have several card types. This indicates the card
+ *             type of the current device.
  * @cs_active_cnt: number of active command submissions on this device (active
  *                 means already in H/W queues)
  * @major: habanalabs kernel driver major.
@@ -1566,6 +1568,7 @@ struct hl_device {
        u64                             clock_gating_mask;
        atomic_t                        in_reset;
        enum hl_pll_frequency           curr_pll_profile;
+       enum armcp_card_types           card_type;
        int                             cs_active_cnt;
        u32                             major;
        u32                             high_pll;
@@ -1858,7 +1861,7 @@ int hl_get_pwm_info(struct hl_device *hdev,
 void hl_set_pwm_info(struct hl_device *hdev, int sensor_index, u32 attr,
                        long value);
 u64 hl_get_max_power(struct hl_device *hdev);
-void hl_set_max_power(struct hl_device *hdev, u64 value);
+void hl_set_max_power(struct hl_device *hdev);
 int hl_set_voltage(struct hl_device *hdev,
                        int sensor_index, u32 attr, long value);
 int hl_set_current(struct hl_device *hdev,

diff --git a/drivers/misc/habanalabs/common/sysfs.c b/drivers/misc/habanalabs/common/sysfs.c
index b3cb0ac..5ae484c 100644 (file)
--- a/drivers/misc/habanalabs/common/sysfs.c
+++ b/drivers/misc/habanalabs/common/sysfs.c
@@ -81,7 +81,7 @@ u64 hl_get_max_power(struct hl_device *hdev)
        return result;
 }
 
-void hl_set_max_power(struct hl_device *hdev, u64 value)
+void hl_set_max_power(struct hl_device *hdev)
 {
        struct armcp_packet pkt;
        int rc;
@@ -90,7 +90,7 @@ void hl_set_max_power(struct hl_device *hdev, u64 value)
 
        pkt.ctl = cpu_to_le32(ARMCP_PACKET_MAX_POWER_SET <<
                                ARMCP_PKT_CTL_OPCODE_SHIFT);
-       pkt.value = cpu_to_le64(value);
+       pkt.value = cpu_to_le64(hdev->max_power);
 
        rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt),
                                                0, NULL);
@@ -316,7 +316,7 @@ static ssize_t max_power_store(struct device *dev,
        }
 
        hdev->max_power = value;
-       hl_set_max_power(hdev, value);
+       hl_set_max_power(hdev);
 
 out:
        return count;
@@ -422,6 +422,7 @@ int hl_sysfs_init(struct hl_device *hdev)
                hdev->pm_mng_profile = PM_AUTO;
        else
                hdev->pm_mng_profile = PM_MANUAL;
+
        hdev->max_power = hdev->asic_prop.max_power_default;
 
        hdev->asic_funcs->add_device_attr(hdev, &hl_dev_clks_attr_group);

diff --git a/drivers/misc/habanalabs/gaudi/gaudi.c b/drivers/misc/habanalabs/gaudi/gaudi.c
index 7e0f9f6..4009b7d 100644 (file)
--- a/drivers/misc/habanalabs/gaudi/gaudi.c
+++ b/drivers/misc/habanalabs/gaudi/gaudi.c
@@ -456,7 +456,7 @@ static int gaudi_get_fixed_properties(struct hl_device *hdev)
        prop->num_of_events = GAUDI_EVENT_SIZE;
        prop->tpc_enabled_mask = TPC_ENABLED_MASK;
 
-       prop->max_power_default = MAX_POWER_DEFAULT;
+       prop->max_power_default = MAX_POWER_DEFAULT_PCI;
 
        prop->cb_pool_cb_cnt = GAUDI_CB_POOL_CB_CNT;
        prop->cb_pool_cb_size = GAUDI_CB_POOL_CB_SIZE;
@@ -6055,6 +6055,15 @@ static int gaudi_armcp_info_get(struct hl_device *hdev)
                strncpy(prop->armcp_info.card_name, GAUDI_DEFAULT_CARD_NAME,
                                CARD_NAME_MAX_LEN);
 
+       hdev->card_type = le32_to_cpu(hdev->asic_prop.armcp_info.card_type);
+
+       if (hdev->card_type == armcp_card_type_pci)
+               prop->max_power_default = MAX_POWER_DEFAULT_PCI;
+       else if (hdev->card_type == armcp_card_type_pmc)
+               prop->max_power_default = MAX_POWER_DEFAULT_PMC;
+
+       hdev->max_power = prop->max_power_default;
+
        return 0;
 }
 

diff --git a/drivers/misc/habanalabs/gaudi/gaudiP.h b/drivers/misc/habanalabs/gaudi/gaudiP.h
index 5dc99f6..82137c3 100644 (file)
--- a/drivers/misc/habanalabs/gaudi/gaudiP.h
+++ b/drivers/misc/habanalabs/gaudi/gaudiP.h
@@ -41,7 +41,8 @@
 
 #define GAUDI_MAX_CLK_FREQ             2200000000ull   /* 2200 MHz */
 
-#define MAX_POWER_DEFAULT              200000          /* 200W */
+#define MAX_POWER_DEFAULT_PCI          200000          /* 200W */
+#define MAX_POWER_DEFAULT_PMC          350000          /* 350W */
 
 #define GAUDI_CPU_TIMEOUT_USEC         15000000        /* 15s */