Displaying 2 results from an estimated 2 matches for "computedeadsymbols".
2017 May 01
3
RFC: Moving the module summary into the irsymtab
...e summaries (and store them in the
bitcode files) that look like this:
ThinLTO: foo -> {bar}
FullLTO: bar -> {}
The FullLTO summaries would have a flag set to prevent importing (at least
to start with).
At LTO time those summaries would be loaded into a single summary index and
we would run computeDeadSymbols over the index. In this case, bar would be
added to the set of dead symbols because it is not reachable from a GC
root. We would move the code that handles loading regular LTO modules to
LTO::runRegularLTO. In that function we would check whether bar is dead.
Because it is, we would not add it to t...
2017 Apr 25
2
RFC: Moving the module summary into the irsymtab
Hi all,
I've been making a number of changes to the summary representation
recently, and I wanted to lay out some of my plans so that folks are aware
of my ultimate direction with this.
Basically I want to move the summary into the irsymtab that we will be
storing to disk after D32061 lands. This would help solve a number of
problems:
- To read a summary, you need to read all summaries in a