Marek Chalupa
Marek Chalupa
Can you show the log of the errors?
Or you can do something like this: https://github.com/mchalupa/wayland-tracer/commit/a1307610bc1b4f03a589ddaeb393d6bbd0e193f5 Haven't tested it extensively, so maybe there's an error - but seems working.
Thanks a lot, such patches are always very welcome! I created the branch `llvm-latest` as you wanted. If you're OK with that, I would later apply this PR with some...
Thanks for reporting it. This is indeed a bug. The behavior of `llvm-pta-dump` was changed when a generic API for pointer analyses was introduced. `llvm-pta-dump` now uses this API to...
DG can do only data dependency computation, check the `llvm-dda-dump` tool. But I reckon that the analysis that takes too long for PostgreSQL is exactly the data dependency analysis.
DG can use SVF for points-to analysis and this then propagates to data dependence analysis. If you want to use the value-flow analysis from SVF, you must do that directly...
To answer the question from the title: both, depends on the options you pass on the command line (and if you compile DG with SVF support).
Yes, that should do the trick, check this code: https://github.com/mchalupa/dg/blob/master/tools/llvm-pta-dump.cpp#L816
This part is important: ```C LLVMDataDependenceAnalysis DDA(M.get(), &PTA, options.dgOptions.DDAOptions); DDA.run(); ``` Then you can query directly the `DDA` object for the definitions.
SSA is run either way, but this code makes it to compute all the results immediately (otherwise they are computed on-demand upon querying DDA object) so that `llvm-dda-dump` can show...