CI: Fix r.texture test on macOS for v8.4

[neteler]

Attempt to fix the following CI error (log) on releasebranch_8_4 by syncing to main:

 Running ./raster/r.texture/testsuite/test_texture.py...
========================================================================
...........F.
======================================================================
FAIL: test_sv (__main__.TestRasterreport.test_sv)
Testing method sv
----------------------------------------------------------------------
Traceback (most recent call last):
  File "raster/r.texture/testsuite/test_texture.py", line 144, in test_sv
    self.assertRasterFitsUnivar(output, reference=values, precision=1e-2)
  File "etc/python/grass/gunittest/case.py", line 289, in assertRasterFitsUnivar
    self.assertModuleKeyValue(
  File "etc/python/grass/gunittest/case.py", line 271, in assertModuleKeyValue
    self.fail(self._formatMessage(msg, stdMsg))
AssertionError: r.univar map=sv_SV percentile=90.0 nprocs=1 separator== -g difference:
mismatch values (key, reference, actual): [('max', 45368496, 45368492), ('mean', 2248724.38215788, 2248724.35922656), ('variance', 2332049495199.41, 2332049429651.92)]
command: r.univar map=sv_SV percentile=90.0 nprocs=1 separator== -g {'map': 'sv_SV', 'separator': '=', 'flags': 'g'}

----------------------------------------------------------------------
Ran 13 tests in 17.655s
FAILED (failures=1)

Looks like an ugly workaround here but in G85 (#5750) this part has been rewritten and backporting is too much work (for me).

[wenzeslaus]

With your latest changes, it only adjusts the already inconsistent numbers, so I would say this is fine. You can link the actual fixes in the PR to make it clear that the ugliness is already resolved elsewhere.

[neteler]

Yes, I have edited the description above to refer to #5750

[echoix]

No opinion for me, but the justification by Vashek is fine: the numbers are probably wrong, so one wrong number for another is equivalent.