v3d-rpi4-gl: reduce the parallelism from 10 to 8

We are slightly over-subscribed right now, with 21 merge jobs (10 vk + 10 gl + 1 traces) for 20 RPi4, so let's split the tests between slightly fewer RPi4 and make each split job run for slightly longer, because it also means that all the jobs start immediately, reducing the overall delay in merging any MR that triggers this job.

The typical run time for this job was around 8 min with a 10-split; with an 8-split it is now 9 min, which is still within the 10 min target and well below the 15min limit.

---

Stress-tested over 8 * 100 runs overnight, no new flakes, but the board failed to boot (read timeout waiting for serial during bootloader) 3 times, and weston crashed (weston: ../xwayland/window-manager.c:1190: weston_wm_handle_map_request: Assertion '!window->shsurf' failed.) 2 times.
Not really sure what we could do about either of these, but at least it's a very low rate 2 and 3 out of 800 :)

>>> {j.name for j in jobs}
{'v3d-rpi4-gl:arm64 6/8', 'v3d-rpi4-gl:arm64 2/8', 'v3d-rpi4-gl:arm64 8/8', 'v3d-rpi4-gl:arm64 3/8', 'v3d-rpi4-gl:arm64 5/8', 'v3d-rpi4-gl:arm64 7/8', 'v3d-rpi4-gl:arm64 4/8', 'v3d-rpi4-gl:arm64 1/8'}
>>> len(jobs)
808
>>> # Not quite sure why there's more than 800, but we've verified the job names above so we know we're looking at the right data
>>> 
>>> durations = [j.duration for j in jobs if j.status == 'success']
>>> # skipping the ones that didn't boot or weston crashed as they finished sooner than normal
>>> 
>>> len(durations)
803
>>> mean(durations)
535.8316864097136
>>> median(durations)
531.765511
>>> stdev(durations)
28.866478407572778

8:55 ± 28sec < 10:00 👌