Missing timeout deadlines on `Modify/DeleteResourceGroup` calls

## Bug Report

In the function `onAlterResourceGroup` within `pkg/ddl/resource_group.go`, `ModifyResourceGroup` is called without a deadline:

```go
err = infosync.ModifyResourceGroup(context.TODO(), protoGroup)
```

Similarly, `DeleteResourceGroup` in `onDropResourceGroup` has the same issue:

```go
err = infosync.DeleteResourceGroup(context.TODO(), groupInfo.Name.L)
```

Both `ModifyResourceGroup` and `DeleteResourceGroup` should be called with a deadline. Otherwise, they may stuck if the internal RPC requests encounter timeout errors.

These functions should follow the same pattern as `AddResourceGroup` in `onCreateResourceGroup`, which properly implements a timeout context:

```go
ctx, cancel := context.WithTimeout(jobCtx.stepCtx, defaultInfosyncTimeout)
defer cancel()
err = infosync.AddResourceGroup(ctx, protoGroup)
```

### 1. Minimal reproduce step (Required)

The following statements call all three functions.

```sql
create resource group rg1 ru_per_sec = 100;
create user user1;
alter user `user1` resource group `rg1`;
alter resource group rg1 ru_per_sec = 200;
drop user user1;
drop resource group if exists rg1;
```

### 2. What did you expect to see? (Required)

Internal RPCs of `ModifyResourceGroup` and `DeleteResourceGroup` return `DeadlineExceeded`

### 3. What did you see instead (Required)

Internal RPCs of `ModifyResourceGroup` and `DeleteResourceGroup` stuck, further block themselves, and further block `onAlterResourceGroup` and `onDropResourceGroup`.

### 4. What is your TiDB version? (Required)

v8.5.0