Fix the llm module e2e test

[phaiting]

The generate API for llm has two issues.

• The input image could be null, but prefillImages assume non-null, this cause prefillImages return Error::EndOfMethod, this can be fixed by adding a null check on the java layer
• The generate didn't pass prompt which caused stack overflow

Test plan
Run all e2e test on the local emulator and run via CI

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/15275

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

❗ 1 Active SEVs

There are 1 currently active SEVs. If your PR is affected, please view them below:

• AWS was down, GHA infrastructure effected / recovering

❌ 2 New Failures, 3 Unrelated Failures

As of commit 4638277 with merge base 73f7286 ():

NEW FAILURES - The following jobs have failed:

• pull / unittest-arm-backend-with-no-fvp (test_pytest_models) / linux-job (gh)
  RuntimeError: Command docker exec -t e92acf08a0ee086814a5836ebf9a45fb0be6a124f12c4738ff66bd4a867f9f87 /exec failed with exit code 1
• pull / unittest-nxp-neutron / linux-job (gh)
  RuntimeError: Command docker exec -t 9d907a226177e6e1245bd8f8e5f81b4a38ad66cd056392e1db100416fb89a730 /exec failed with exit code 1

BROKEN TRUNK - The following jobs failed but were present on the merge base:

👉 Rebase onto the `viable/strict` branch to avoid these failures

• Build documentation / build (buck2) / Build doc (gh) (trunk failure)
• pull / unittest / linux / linux-job (gh) (trunk failure)
• pull / unittest-editable / linux / linux-job (gh) (trunk failure)

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[phaiting]

@pytorchbot label "release notes: none"