Reindexing from scratch way too often

[borbmizzet]

On my VisualGDB, it reindexes way too often, taking tens of seconds almost every time I try to use any of the clang-based features. My PC has 20GB RAM with a quad-core i7 @ 3.6GHz. My Indexing settings are below.
Clang cache settings

[borbmizzet]

The Error List seems to only work for the current file even when it should be showing solution-wide errors, and it re-indexes every time I use try to view the Use Tree, Call Hierarchy, Go To Definition, or Find All References. Peek Definition just plain doesn’t work. I’m using VS 2015 Community with VisualGDB 5.2(build 1357)

[support]

Hi,

Most likely either you modify some global header file used by all other headers, or VisualGDB believes it is modified and does a full rebuild.

Please try opening the Clang IntelliSense Diagnostics Console and check for messages like thi:

[+0:00:33.084] Found an outdated PSF 0 due to <file> during initial check. It will be re-built

It should explain why VisualGDB reparses the files.

The Error List behavior is by design – showing all errors from the entire project would require constantly rebuilding all the files in the background. Currently VisualGDB only does that when you use the ‘Go to definition’ or ‘Find all references’ commands.

[borbmizzet]

Can you make an option for visualgdb to [check if a rebuild is needed and run the rebuild if needed] as a background process every so often? I have been modifying a lot of headers used by a lot of other headers (whoever built my company’s internal library clearly did not set encapsulation as a priority), and it gets really annoying to have to wait 15-20 seconds every time I want to view the call or use tree.

[support]

Hi,

We could add an option like that, but it could be confusing as the information will obviously be outdated (e.g. the offsets of function definitions will be old). Let us know which commands in your opinion would still be useful in such mode and we will consider adding this option.

[Author]

Posts