Upgrade Go SDK to v0.94.0 (#4148)

## Changes

See
https://github.com/databricks/databricks-sdk-go/releases/tag/v0.94.0.

Notable change is a fix for U2M OAuth token caching, where repeated
token calls on a valid token now return directly.

## Tests

Local tests pass.

---------

Co-authored-by: Claude Sonnet 4.5 <[email protected]>

Author: pietern

.codegen/_openapi_sha

@@ -1 +1 @@
-8b2cd47cbac64b32e120601110a5fc70b8189ba4
+e1ea3f5ba0bc5b53be94f56535a67ba701a52a52

NEXT_CHANGELOG.md

@@ -11,5 +11,6 @@
 * Pass SYSTEM_ACCESSTOKEN from env to the Terraform provider ([#4135](https://github.com/databricks/cli/pull/4135)
 
 ### Dependency updates
+* Upgrade Go SDK to 0.94.0 ([#4148](https://github.com/databricks/cli/pull/4148))
 
 ### API Changes

acceptance/bundle/run/inline-script/databricks-cli/profile-is-passed/from_flag/out.requests.txt

@@ -2,6 +2,16 @@
   "method": "GET",
   "path": "/oidc/.well-known/oauth-authorization-server"
 }
+{
+  "headers": {
+    "Authorization": [
+      "Basic [ENCODED_AUTH]"
+    ]
+  },
+  "method": "POST",
+  "path": "/oidc/v1/token",
+  "raw_body": "grant_type=client_credentials\u0026scope=all-apis"
+}
 {
   "method": "GET",
   "path": "/oidc/.well-known/oauth-authorization-server"

acceptance/bundle/run/inline-script/databricks-cli/target-is-passed/from_flag/out.requests.txt

@@ -2,6 +2,16 @@
   "method": "GET",
   "path": "/oidc/.well-known/oauth-authorization-server"
 }
+{
+  "headers": {
+    "Authorization": [
+      "Basic [ENCODED_AUTH]"
+    ]
+  },
+  "method": "POST",
+  "path": "/oidc/v1/token",
+  "raw_body": "grant_type=client_credentials\u0026scope=all-apis"
+}
 {
   "method": "GET",
   "path": "/oidc/.well-known/oauth-authorization-server"

bundle/internal/schema/annotations_openapi.yml

@@ -556,6 +556,7 @@ github.com/databricks/cli/bundle/config/resources.Job:
   "performance_target":
     "description": |-
       The performance mode on a serverless job. This field determines the level of compute performance or cost-efficiency for the run.
+      The performance target does not apply to tasks that run on Serverless GPU compute.
 
       * `STANDARD`: Enables cost-efficient execution of serverless workloads.
       * `PERFORMANCE_OPTIMIZED`: Prioritizes fast startup and execution times through rapid scaling and optimized cluster performance.