szepet added inline comments.
Comment at: lib/StaticAnalyzer/Core/LoopUnrolling.cpp:28-46
struct LoopState {
private:
enum Kind { Normal, Unrolled } K;
- const Stmt *LoopStmt;
- const LocationContext *LCtx;
- unsigned maxStep;
- LoopState(Kind InK, const Stmt *S, const L
szepet updated this revision to Diff 130516.
szepet marked 2 inline comments as done.
szepet added a comment.
First, sorry for this delayed update, however, I was working on this and
running this on real projects. I wanted to make sure that this update will be
complete enough that this patch wou
NoQ added a comment.
//*asked stuff in https://reviews.llvm.org/D39398 regarding how indirect gotos
are supported*//
https://reviews.llvm.org/D41151
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listi
NoQ added a comment.
Or not. Loop counter has its own whole-loop scope.
I guess `LoopContext` can be treated as a sub-class of `ScopeContext`. And i
don't mind having `ScopeContext` be split into small distinct sub-classes.
Because we're stuck in cornercases for covering all possible scopes, wh
NoQ added a comment.
So, essentially, `LoopContext` is per-loop, while `ScopeContext` is
per-iteration?
Comment at: lib/StaticAnalyzer/Core/LoopUnrolling.cpp:28-46
struct LoopState {
private:
enum Kind { Normal, Unrolled } K;
- const Stmt *LoopStmt;
- const LocationCont
a.sidorin added a comment.
This thing is very similar to https://reviews.llvm.org/D19979. Do we really
need to create a separate LoopContext or we can reuse ScopeContext instead?
https://reviews.llvm.org/D41151
___
cfe-commits mailing list
cfe-comm
xazax.hun added a comment.
Just to be sure, this is just a refactoring to make this cleaner or you expect
this to have other effects as well, like better performance?
https://reviews.llvm.org/D41151
___
cfe-commits mailing list
cfe-commits@lists.ll
MTC added inline comments.
Comment at: include/clang/Analysis/ProgramPoint.h:619
/// CallExitBegin and CallExitEnd. The following operations occur between the
-/// two program points:
+/// two preogram points:
/// - CallExitBegin
A minor typo, it should be 'pro
szepet created this revision.
szepet added reviewers: dcoughlin, NoQ, zaks.anna, xazax.hun, a.sidorin.
Herald added subscribers: dkrupp, baloghadamsoftware, whisperity.
Based on the CFGLoopEntrance element, it is possible to have a CFG driven
LocationContext update which contains loop information
