Re: [gdal-dev] The pthread_atfork() issue

2023-12-04 Thread Carsten Schultz via gdal-dev
Even, Thanks for the quick reply. If the getpid() has actually been measured to be an overhead, then it’s even better that this will now be resolved so that the faster code can be re-enabled for the Macs. Since you have already created a PR I will answer there. Best Carsten > On 4. Dec 2023,

Re: [gdal-dev] The pthread_atfork() issue

2023-12-04 Thread Even Rouault via gdal-dev
Carsten, thanks for the detailed analysis. There is definitely a cost to getpid() in some circumstances. The worst case I can imagine is on code like the one below that looks over OGRCoordinateTransformation::Transform() many times. For each call to that function, there's a OSRGetProjTLSConte

[gdal-dev] The pthread_atfork() issue

2023-12-04 Thread Carsten Schultz via gdal-dev
Hello everyone! Please let me know if you would have preferred to discuss this on Github. I have run into a problem with the use of pthread_atfork() in ogr_proj_p.cpp. I have since seen that the code has been disabled for macOS in 3.8 in response to https://github.com/OSGeo/gdal/issues/8497, an