bruno abandoned this revision.
bruno added a comment.
Hi Ben,
Comment at: lib/Basic/FileManager.cpp:221-223
@@ -220,2 +220,5 @@
// See if there is already an entry in the map.
+ // FIXME: Note that when first requested, the returned file name equals to
the
+ // requested p
benlangmuir added inline comments.
Comment at: lib/Basic/FileManager.cpp:221-223
@@ -220,2 +220,5 @@
// See if there is already an entry in the map.
+ // FIXME: Note that when first requested, the returned file name equals to
the
+ // requested path. This is not true when re
bruno added a comment.
Hi Ben,
Comment at: lib/Basic/FileManager.cpp:221-223
@@ -220,2 +220,5 @@
// See if there is already an entry in the map.
+ // FIXME: Note that when first requested, the returned file name equals to
the
+ // requested path. This is not true when retu
benlangmuir added a comment.
> From what I understand from
> http://lists.llvm.org/pipermail/cfe-dev/2014-July/038273.html, changing the
> behavior of the cached result has side-effects and demands a non trivial
> change, am I assuming the right thing?
Yes.
Threading through the Filename see
bruno updated this revision to Diff 53363.
bruno added a comment.
Hi Richard & Ben,
Thanks for the review!
I've attached a new patch and changed the approach: made
findUsableModuleForHeader receive a FileName
instead of FileEntry. Answering the questions below:
FileManager::getFile returns th
rsmith added a comment.
Can you find who's poisoning the `FileManager`'s idea of the name of the file
with a full path to the VFS overlay and fix that?
Comment at: lib/Lex/HeaderSearch.cpp:1163
@@ +1162,3 @@
+SmallString<256> RootName = StringRef(Root->getName());
+path
benlangmuir added a comment.
This doesn't appear to be a safe change. We can't assume that `RootName`
wouldn't ever be legitimately found somewhere in a path (even without a VFS)
and changing the path prefix could give a completely different location. I
would also be concerned about the perfo
bruno added a comment.
Ping!
http://reviews.llvm.org/D18849
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
bruno created this revision.
bruno added reviewers: rsmith, benlangmuir, klimek.
bruno added subscribers: aprantl, dexonsmith, cfe-commits.
The reproducer script (generated by the crash reproducer) invokes clang
and tries to rebuild modules using the headers in the .cache/vfs
directory. Depending