Even Rouault

Are you sure it is really a regression of GDAL 3.8.4 and not something else that changed since last time you did a 3.8.3 build? Because there's no obvious iconv...

> I'm going to propose closing this in favor of a separate PR that includes this batch of test changes but does not enable random ordering or parallel tests for...

> As a suggestion, it may be good to implement a `VSIFilesystemHandler::PathFileExists` function, which on Windows calls [PathFileExistsW](https://docs.microsoft.com/en-us/windows/win32/api/shlwapi/nf-shlwapi-pathfileexistsw), and other platforms keeps the existing stat behaviour I don't think we'd...

@VirtualTim Can you check https://github.com/OSGeo/gdal/pull/9256 works as expected?

Relevant code: - IMapInfoFile::CreateTABFeature() in the case wkbGeometryCollection - TABCollection::SetRegionDirectly() / SetPolylineDirectly() / SetMultiPointDirectly() that could probably be used

> For the `SORT_BY_BBOX` option, one thought: currently you made it into a YES/NO option, but I can imagine that there could be multiple sort _methods_ For the sake of...

@schwehr Could you confirm https://github.com/OSGeo/gdal/pull/9253 fixes this issue ? I couldn't reproduce a TSAN warning with gcc 9 -sanitize=thread

> Allowing zoom level 23 based on extent to avoid the overflows That would certainly be the prefered solution

- How many CPUs/virtual CPUs does your system has ? - Please include the output of gdalinfo on the geotiff file

ok, so you are definitely in one use case where multithreading is going to be the less efficient, because you use nearest neighbouring which isn't computing intensive at all, and...