i915/gem_ppgtt: Estimate resource usage and bail if it means swapping!

fi-kbl-guc's swap ran dry while running blt-vs-render-ctxN, which is
mildly concerning but conceivable as we never checked there was enough
memory to run the test to begin with.

Each child needs to keep its own surface and possible a pair of logical
contexts (one for rcs and one for bcs) so check that there is enough
memory to allow all children to co-exist. During execution, we require
another surface and batch, but these are temporary and so should fit
fine with a small amount of thrashing on the boundary.

References: https://bugs.freedesktop.org/show_bug.cgi?id=109801
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
14 jobs for master in 10 minutes and 45 seconds (queued for 48 minutes)
Status Job ID Name Coverage
  Build
passed #164649
build:tests-debian-autotools

00:01:09

passed #164646
build:tests-debian-meson

00:00:49

passed #164648
build:tests-debian-meson-arm64

00:00:39

passed #164647
build:tests-debian-meson-armhf

00:00:46

passed #164643
build:tests-fedora

00:01:34

passed #164644
build:tests-fedora-no-libunwind

00:01:03

passed #164645
build:tests-fedora-oldest-meson

00:01:25

 
  Test
passed #164650
test:ninja-test

00:00:39

passed #164651
test:ninja-test-arm64

00:00:45

passed #164652
test:ninja-test-armhf

00:00:32

passed #164653
test:test-list-diff

00:00:26

 
  Deploy
passed #164654
pages

00:00:52

passed pages:deploy

00:00:00

failed pages:deploy

00:00:00