habanalabs: fix resetting args in wait for CS IOCTL

In wait for CS IOCTL code, the driver resets the incoming args structure
before returning to the user, regardless of the return value of the
IOCTL.

In case the IOCTL returns EINTR, resetting the args will result in error
in case the userspace will repeat the ioctl call immediately (which is
the behavior in the hl-thunk userspace library).

The solution is to reset the args only if the driver returns success (0)
as a return value for the IOCTL.

Signed-off-by: Rajaravi Krishna Katta <rkatta@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>

diff --git a/drivers/misc/habanalabs/common/command_submission.c b/drivers/misc/habanalabs/common/command_submission.c
index 91b5754..6dafff3 100644 (file)
--- a/drivers/misc/habanalabs/common/command_submission.c
+++ b/drivers/misc/habanalabs/common/command_submission.c
@@ -2649,11 +2649,18 @@ put_ctx:
 free_seq_arr:
        kfree(cs_seq_arr);
 
-       /* update output args */
-       memset(args, 0, sizeof(*args));
        if (rc)
                return rc;
 
+       if (mcs_data.wait_status == -ERESTARTSYS) {
+               dev_err_ratelimited(hdev->dev,
+                               "user process got signal while waiting for Multi-CS\n");
+               return -EINTR;
+       }
+
+       /* update output args */
+       memset(args, 0, sizeof(*args));
+
        if (mcs_data.completion_bitmap) {
                args->out.status = HL_WAIT_CS_STATUS_COMPLETED;
                args->out.cs_completion_map = mcs_data.completion_bitmap;
@@ -2667,8 +2674,6 @@ free_seq_arr:
                /* update if some CS was gone */
                if (mcs_data.timestamp)
                        args->out.flags |= HL_WAIT_CS_STATUS_FLAG_GONE;
-       } else if (mcs_data.wait_status == -ERESTARTSYS) {
-               args->out.status = HL_WAIT_CS_STATUS_INTERRUPTED;
        } else {
                args->out.status = HL_WAIT_CS_STATUS_BUSY;
        }
@@ -2688,16 +2693,17 @@ static int hl_cs_wait_ioctl(struct hl_fpriv *hpriv, void *data)
        rc = _hl_cs_wait_ioctl(hdev, hpriv->ctx, args->in.timeout_us, seq,
                                &status, &timestamp);
 
+       if (rc == -ERESTARTSYS) {
+               dev_err_ratelimited(hdev->dev,
+                       "user process got signal while waiting for CS handle %llu\n",
+                       seq);
+               return -EINTR;
+       }
+
        memset(args, 0, sizeof(*args));
 
        if (rc) {
-               if (rc == -ERESTARTSYS) {
-                       dev_err_ratelimited(hdev->dev,
-                               "user process got signal while waiting for CS handle %llu\n",
-                               seq);
-                       args->out.status = HL_WAIT_CS_STATUS_INTERRUPTED;
-                       rc = -EINTR;
-               } else if (rc == -ETIMEDOUT) {
+               if (rc == -ETIMEDOUT) {
                        dev_err_ratelimited(hdev->dev,
                                "CS %llu has timed-out while user process is waiting for it\n",
                                seq);
@@ -2823,7 +2829,6 @@ wait_again:
                dev_err_ratelimited(hdev->dev,
                        "user process got signal while waiting for interrupt ID %d\n",
                        interrupt->interrupt_id);
-               *status = HL_WAIT_CS_STATUS_INTERRUPTED;
                rc = -EINTR;
        } else {
                *status = CS_WAIT_STATUS_BUSY;
@@ -2878,8 +2883,6 @@ static int hl_interrupt_wait_ioctl(struct hl_fpriv *hpriv, void *data)
                                args->in.interrupt_timeout_us, args->in.addr,
                                args->in.target, interrupt_offset, &status);
 
-       memset(args, 0, sizeof(*args));
-
        if (rc) {
                if (rc != -EINTR)
                        dev_err_ratelimited(hdev->dev,
@@ -2888,6 +2891,8 @@ static int hl_interrupt_wait_ioctl(struct hl_fpriv *hpriv, void *data)
                return rc;
        }
 
+       memset(args, 0, sizeof(*args));
+
        switch (status) {
        case CS_WAIT_STATUS_COMPLETED:
                args->out.status = HL_WAIT_CS_STATUS_COMPLETED;

diff --git a/include/uapi/misc/habanalabs.h b/include/uapi/misc/habanalabs.h
index 7cc2a0f..d13bb8c 100644 (file)
--- a/include/uapi/misc/habanalabs.h
+++ b/include/uapi/misc/habanalabs.h
@@ -917,7 +917,6 @@ struct hl_wait_cs_in {
 #define HL_WAIT_CS_STATUS_BUSY         1
 #define HL_WAIT_CS_STATUS_TIMEDOUT     2
 #define HL_WAIT_CS_STATUS_ABORTED      3
-#define HL_WAIT_CS_STATUS_INTERRUPTED  4
 
 #define HL_WAIT_CS_STATUS_FLAG_GONE            0x1
 #define HL_WAIT_CS_STATUS_FLAG_TIMESTAMP_VLD   0x2
@@ -1286,7 +1285,8 @@ struct hl_debug_args {
  * EIO       - The CS was aborted (usually because the device was reset)
  * ENODEV    - The device wants to do hard-reset (so user need to close FD)
  *
- * The driver also returns a custom define inside the IOCTL which can be:
+ * The driver also returns a custom define in case the IOCTL call returned 0.
+ * The define can be one of the following:
  *
  * HL_WAIT_CS_STATUS_COMPLETED   - The CS has been completed successfully (0)
  * HL_WAIT_CS_STATUS_BUSY        - The CS is still executing (0)
@@ -1294,8 +1294,6 @@ struct hl_debug_args {
  *                                 (ETIMEDOUT)
  * HL_WAIT_CS_STATUS_ABORTED     - The CS was aborted, usually because the
  *                                 device was reset (EIO)
- * HL_WAIT_CS_STATUS_INTERRUPTED - Waiting for the CS was interrupted (EINTR)
- *
  */
 
 #define HL_IOCTL_WAIT_CS                       \