Changes between Initial Version and Version 1 of Ticket #24986, comment 35


Ignore:
Timestamp:
03/16/18 16:45:01 (4 years ago)
Author:
embray
Comment:

Legend:

Unmodified
Added
Removed
Modified
  • Ticket #24986, comment 35

    initial v1  
    1 I suppose one other possibility, is we replace the GMP memory allocation functions with ones that trace memory allocations relevant to the integer pool, and use that to perform a quick consistency check before returning objects to the pool.  Would have to think about how to do this in a way that does not have too awful overhead...
     1~~I suppose one other possibility, is we replace the GMP memory allocation functions with ones that trace memory allocations relevant to the integer pool, and use that to perform a quick consistency check before returning objects to the pool.  Would have to think about how to do this in a way that does not have too awful overhead...~~
     2
     3No, this would suck, because there would be no obvious way to enable that behavior only for the relevant contexts.