Hi, thanks. Those tools might be good if it were a situation where a developer 
needs to do a one-time investigation with a special build, perhaps to trim down 
the memory consumption generally.

But actually, the use case I had in mind is for incorporating into detailed 
statistics available in the customer-delivered package. I can already give 
fairly accurate per-file counts for tiles read, time spent on I/O, memory held 
by my own code, etc., for each of the many thousands of image files accessed in 
a run. But the missing component would be knowing how much memory is being held 
in the TIFF library for each open file. The existence of the  SetMax*MemAlloc 
functions mean it must already be tracked internally, it's just a matter of 
whether I can easily get to it.


> On Sep 19, 2024, at 12:59 AM, Milian Wolff <[email protected]> wrote:
> 
> While I agree that forwarding more information might be useful, I would like 
> to point out that your question might also be answered in a more library-
> agnostic fashion via generic memory profiling. 

--
Larry Gritz
[email protected]





_______________________________________________
Tiff mailing list
[email protected]
https://lists.osgeo.org/mailman/listinfo/tiff

Reply via email to