eugenis updated this revision to Diff 37795.
eugenis marked 3 inline comments as done.
Repository:
rL LLVM
http://reviews.llvm.org/D10677
Files:
include/__config
include/deque
test/libcxx/containers/sequences/deque/incomplete.pass.cpp
Index: test/libcxx/containers/sequences/deque/incomp
eugenis added a comment.
In http://reviews.llvm.org/D10677#266595, @EricWF wrote:
> For the most part this looks good. I'm a touch concerned though about the
> changes to the static initialization. The initializer is moved from within
> the function body to outside it. Could you have somebody c
eugenis created this revision.
eugenis added a reviewer: rsmith.
eugenis added subscribers: cfe-commits, EricWF, rnk.
eugenis set the repository for this revision to rL LLVM.
The attrubite is applicable to functions and variables and changes the linkage
of the subject to internal.
Following the
eugenis added inline comments.
Comment at: lib/Sema/SemaDeclAttr.cpp:1580-1585
@@ -1577,3 +1579,8 @@
+
+ if (InternalLinkageAttr *Internal = D->getAttr()) {
+S.Diag(Attr.getRange().getBegin(), diag::warn_attribute_ignored)
+<< Attr.getName();
+S.Diag(Internal->get
Author: eugenis
Date: Wed Oct 21 15:20:03 2015
New Revision: 250941
URL: http://llvm.org/viewvc/llvm-project?rev=250941&view=rev
Log:
MemorySanitizer does not require PIE.
Since r249754 MemorySanitizer should work equally well for PIE and
non-PIE executables.
Beware, with this change -fsanitize=
Author: eugenis
Date: Wed Oct 21 15:47:00 2015
New Revision: 250944
URL: http://llvm.org/viewvc/llvm-project?rev=250944&view=rev
Log:
Revert "MemorySanitizer does not require PIE."
It actually does require PIE on some targets.
Modified:
cfe/trunk/lib/Driver/SanitizerArgs.cpp
cfe/trunk/te
Author: eugenis
Date: Wed Oct 21 16:28:49 2015
New Revision: 250949
URL: http://llvm.org/viewvc/llvm-project?rev=250949&view=rev
Log:
MemorySanitizer does not require PIE.
Since r249754 MemorySanitizer should work equally well for PIE and
non-PIE executables on Linux/x86_64.
Beware, with this ch
eugenis added a comment.
What kind of confirmation are you looking for?
I've compiled the following code with 2 versions of : one as in this
review, another the same but with __block_size initializers moved back into
respective classes. Resulting object files are identical.
#include
int main
eugenis updated this revision to Diff 38071.
eugenis marked 2 inline comments as done.
eugenis added a comment.
This new version supports __attribute__((internal_linkage)) on classes and even
namespaces!
Repository:
rL LLVM
http://reviews.llvm.org/D13925
Files:
include/clang/Basic/Attr.td
eugenis added inline comments.
Comment at: lib/Sema/SemaDeclAttr.cpp:1580-1585
@@ -1577,3 +1579,8 @@
+
+ if (InternalLinkageAttr *Internal = D->getAttr()) {
+S.Diag(Attr.getRange().getBegin(), diag::warn_attribute_ignored)
+<< Attr.getName();
+S.Diag(Internal->get
eugenis added a comment.
ping
We will need this when libc++ is bumped to v2. There are no plans to do this
AFAIK but the ABI version macros are in so it is possible and clang should be
ready.
http://reviews.llvm.org/D12382
___
cfe-commits mailing
eugenis updated this revision to Diff 38679.
Repository:
rL LLVM
http://reviews.llvm.org/D13925
Files:
include/clang/Basic/Attr.td
include/clang/Basic/AttrDocs.td
include/clang/Sema/Sema.h
lib/AST/Decl.cpp
lib/Sema/Sema.cpp
lib/Sema/SemaDecl.cpp
lib/Sema/SemaDeclAttr.cpp
test/C
eugenis added a comment.
In http://reviews.llvm.org/D13925#276626, @majnemer wrote:
> No diagnostic is issued for the following C test case:
>
> int x __attribute__((internal_linkage));
> int x __attribute__((common));
> int *f() { return &x; }
Thanks for noticing! Added missing attribute
eugenis updated this revision to Diff 38979.
eugenis added a comment.
Added a [[clang::internal_linkage]] spelling to the attribute.
Added tests for namespace re-declarations with and without the attribute.
Repository:
rL LLVM
http://reviews.llvm.org/D13925
Files:
include/clang/Basic/Attr.
eugenis added inline comments.
Comment at: include/clang/Basic/Attr.td:2125
@@ +2124,3 @@
+
+def InternalLinkage : InheritableAttr {
+ let Spellings = [GNU<"internal_linkage">, CXX11<"clang",
"internal_linkage">];
aaron.ballman wrote:
> rsmith wrote:
> > `Inheri
eugenis updated this revision to Diff 39124.
eugenis marked 5 inline comments as done.
eugenis added a comment.
Disabled the new attribute on namespaces.
Repository:
rL LLVM
http://reviews.llvm.org/D13925
Files:
include/clang/Basic/Attr.td
include/clang/Basic/AttrDocs.td
include/clang/
eugenis updated this revision to Diff 39128.
eugenis added a comment.
One more test for forward declarations.
Repository:
rL LLVM
http://reviews.llvm.org/D13925
Files:
include/clang/Basic/Attr.td
include/clang/Basic/AttrDocs.td
include/clang/Sema/Sema.h
lib/AST/Decl.cpp
lib/Sema/Se
eugenis updated this revision to Diff 39135.
eugenis marked 2 inline comments as done.
Repository:
rL LLVM
http://reviews.llvm.org/D13925
Files:
include/clang/Basic/Attr.td
include/clang/Basic/AttrDocs.td
include/clang/Sema/Sema.h
lib/AST/Decl.cpp
lib/Sema/SemaDecl.cpp
lib/Sema/Sem
eugenis added inline comments.
Comment at: lib/AST/Decl.cpp:635-641
@@ -634,2 +634,9 @@
assert(!isa(D) && "Didn't expect a FieldDecl!");
+ for (const DeclContext *DC = D->getDeclContext();
+ !isa(DC); DC = DC->getParent()) {
+const NamespaceDecl *ND = dyn_cast(DC);
eugenis added a comment.
Hi Eric,
could you please clarify what exactly you are looking for here?
Repository:
rL LLVM
http://reviews.llvm.org/D10677
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/li
eugenis added a comment.
Hm, the current implementation allows all of the following:
void f(int a [[clang::internal_linkage]]) { // 1
int b [[clang::internal_linkage]]; // 2
static int c [[clang::internal_linkage]]; // 3
}
I'll fix (1). Is it OK to allow (2) and (3)? The attribute has
eugenis added a comment.
How do I check if a Var is a local variable?
Comment at: test/CodeGenCXX/attribute_internal_linkage.cpp:35-36
@@ +34,4 @@
+__attribute__((internal_linkage)) void A::f3() {
+}
+
+// Forward declaration w/o an attribute.
OK, done. Still al
eugenis updated this revision to Diff 39245.
Repository:
rL LLVM
http://reviews.llvm.org/D13925
Files:
include/clang/Basic/Attr.td
include/clang/Basic/AttrDocs.td
include/clang/Basic/DiagnosticSemaKinds.td
include/clang/Sema/Sema.h
lib/AST/Decl.cpp
lib/Sema/SemaDecl.cpp
lib/Sema/
eugenis updated this revision to Diff 39249.
Repository:
rL LLVM
http://reviews.llvm.org/D13925
Files:
include/clang/Basic/Attr.td
include/clang/Basic/AttrDocs.td
include/clang/Basic/DiagnosticSemaKinds.td
include/clang/Sema/Sema.h
lib/AST/Decl.cpp
lib/Sema/SemaDecl.cpp
lib/Sema/
eugenis added a comment.
In http://reviews.llvm.org/D13925#281349, @eugenis wrote:
> Hm, the current implementation allows all of the following:
>
> void f(int a [[clang::internal_linkage]]) { // 1
>
> int b [[clang::internal_linkage]]; // 2
> static int c [[clang::internal_linkage]]; //
eugenis added inline comments.
Comment at: test/SemaCXX/internal_linkage.cpp:23
@@ +22,3 @@
+
+__attribute__((internal_linkage)) void A::f4() {} //
expected-error{{'internal_linkage' attribute does not appear on the first
declaration of 'f4'}}
+
Btw, this trigge
eugenis updated this revision to Diff 39389.
eugenis marked an inline comment as done.
eugenis added a comment.
Added the new warning to a -W group.
Repository:
rL LLVM
http://reviews.llvm.org/D13925
Files:
include/clang/Basic/Attr.td
include/clang/Basic/AttrDocs.td
include/clang/Basic
eugenis added inline comments.
Comment at: lib/Sema/SemaDeclAttr.cpp:3407
@@ +3406,3 @@
+ : ExpectedVariableOrFunction);
+D->dropAttr();
+ }
aaron.ballman wrote:
> Why is this dropping AlwaysInlineAttr instead of re
eugenis created this revision.
eugenis added reviewers: mclow.lists, EricWF.
eugenis added a subscriber: cfe-commits.
eugenis set the repository for this revision to rL LLVM.
No point in pretending that these methods are hidden - they are
actually exported from libc++.so. Extern template declarati
eugenis created this revision.
eugenis added reviewers: EricWF, mclow.lists.
eugenis added a subscriber: cfe-commits.
eugenis set the repository for this revision to rL LLVM.
This change moves visibility attributes from out-of-class method definitions to
in-class declaration.
This is needed for a
eugenis created this revision.
eugenis added reviewers: EricWF, mclow.lists.
eugenis added a subscriber: cfe-commits.
eugenis set the repository for this revision to rL LLVM.
Use __attribute__((internal_linkage)) instead of always_inline and
visibility("hidden") when it is available.
Repository:
eugenis marked an inline comment as done.
eugenis added a comment.
Repository:
rL LLVM
http://reviews.llvm.org/D12382
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
eugenis set the repository for this revision to rL LLVM.
eugenis updated this revision to Diff 39587.
Repository:
rL LLVM
http://reviews.llvm.org/D12382
Files:
lib/Driver/ToolChains.cpp
test/Driver/Inputs/basic_linux_libcxxv2_tree/usr/bin/.keep
test/Driver/Inputs/basic_linux_libcxxv2_tre
eugenis added a comment.
This is a mechanical change, and AFAIR I checked that no bits of libc++.so are
affected.
The same mechanical change in streambuf/istream/ostream (the stuff that's parts
of extern templates) breaks libc++ because of
https://llvm.org/bugs/show_bug.cgi?id=25427, that's wh
eugenis closed this revision.
eugenis added a comment.
Thanks!
Landed as r252350.
Repository:
rL LLVM
http://reviews.llvm.org/D10677
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commit
Author: eugenis
Date: Fri Nov 6 16:02:29 2015
New Revision: 252350
URL: http://llvm.org/viewvc/llvm-project?rev=252350&view=rev
Log:
Allow deque to handle incomplete types.
Allow deque and deque::iterator instantiation with incomplete element
type. This is an ABI breaking change, and it is only
eugenis added a comment.
Yes. I'm just using libc++.so as a (incomplete) test that this is actually NFC.
It did catch the extern template issue.
Repository:
rL LLVM
http://reviews.llvm.org/D14410
___
cfe-commits mailing list
cfe-commits@lists.ll
Author: eugenis
Date: Fri Nov 6 19:22:13 2015
New Revision: 252385
URL: http://llvm.org/viewvc/llvm-project?rev=252385&view=rev
Log:
Cleanup: move visibility/linkage attributes to the first declaration.
This change moves visibility attributes from out-of-class method
definitions to in-class decl
eugenis closed this revision.
eugenis added a comment.
Confirmed and landed as r252385.
Thanks for the quick response!
Repository:
rL LLVM
http://reviews.llvm.org/D14410
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.or
eugenis added a comment.
Richard, are you OK with this?
Repository:
rL LLVM
http://reviews.llvm.org/D13925
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Author: eugenis
Date: Mon Nov 9 15:10:54 2015
New Revision: 252514
URL: http://llvm.org/viewvc/llvm-project?rev=252514&view=rev
Log:
Extend linux header search to find libc++ headers in c++/vN for any N.
Added:
cfe/trunk/test/Driver/Inputs/basic_linux_libcxxv2_tree/
cfe/trunk/test/Driver
eugenis closed this revision.
eugenis added a comment.
r252514
Repository:
rL LLVM
http://reviews.llvm.org/D12382
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
eugenis updated this revision to Diff 39750.
eugenis added a comment.
Applied the same change to . No idea how I missed it.
Repository:
rL LLVM
http://reviews.llvm.org/D14409
Files:
include/istream
include/ostream
include/sstream
include/streambuf
Index: include/streambuf
==
eugenis updated this revision to Diff 39751.
eugenis added a comment.
Rebase.
Repository:
rL LLVM
http://reviews.llvm.org/D13925
Files:
include/clang/Basic/Attr.td
include/clang/Basic/AttrDocs.td
include/clang/Basic/DiagnosticSemaKinds.td
include/clang/Sema/Sema.h
lib/AST/Decl.cpp
eugenis added a comment.
is an interesting case.
There are no cases of visibility attribute present on an out-of-class
definition but missing on an in-class declaration, so nothing needs to be done
for a switch to internal_linkage.
Setting hidden visibility on an "extern template" method is pr
eugenis closed this revision.
eugenis added a comment.
r252648.
Thanks everyone for the very detailed review!
Repository:
rL LLVM
http://reviews.llvm.org/D13925
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin
Author: eugenis
Date: Tue Nov 10 15:28:44 2015
New Revision: 252648
URL: http://llvm.org/viewvc/llvm-project?rev=252648&view=rev
Log:
Implement __attribute__((internal_linkage)).
The attrubite is applicable to functions and variables and changes
the linkage of the subject to internal.
This is th
Author: eugenis
Date: Thu Jan 3 14:41:10 2019
New Revision: 350361
URL: http://llvm.org/viewvc/llvm-project?rev=350361&view=rev
Log:
[cmake] Fix monorepo + LLVM_BUILD_EXTERNAL_COMPILER_RT=ON.
In cmake 3.10.2, if (${VARIABLE_NAME}) seems to always be false no
matter what documentation says (or ma
Author: eugenis
Date: Thu Jan 3 14:50:45 2019
New Revision: 350363
URL: http://llvm.org/viewvc/llvm-project?rev=350363&view=rev
Log:
Fix check-hwasan with LLVM_BUILD_EXTERNAL_COMPILER_RT=ON
Add a forwarding target for check-hwasan in clang.
Modified:
cfe/trunk/runtime/CMakeLists.txt
Modifi
Author: eugenis
Date: Thu Jul 6 17:48:12 2017
New Revision: 307341
URL: http://llvm.org/viewvc/llvm-project?rev=307341&view=rev
Log:
Update Cross-DSO CFI documentation.
Reviewers: pcc
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D35094
Modified:
cfe/trunk/docs
Author: eugenis
Date: Tue May 9 16:57:39 2017
New Revision: 302590
URL: http://llvm.org/viewvc/llvm-project?rev=302590&view=rev
Log:
Remove unnecessary calls to MakeArgString.
Modified:
cfe/trunk/lib/Driver/SanitizerArgs.cpp
cfe/trunk/lib/Driver/ToolChains/Arch/Mips.cpp
Modified: cfe/tr
Author: eugenis
Date: Tue May 9 16:57:43 2017
New Revision: 302591
URL: http://llvm.org/viewvc/llvm-project?rev=302591&view=rev
Log:
[asan] A clang flag to enable ELF globals-gc.
This feature is subtly broken when the linker is gold 2.26 or
earlier. See the following bug for details:
https://s
Author: eugenis
Date: Tue May 9 17:28:57 2017
New Revision: 302594
URL: http://llvm.org/viewvc/llvm-project?rev=302594&view=rev
Log:
Fix clang_cl argument in fsanitize.c driver test.
Modified:
cfe/trunk/test/Driver/fsanitize.c
Modified: cfe/trunk/test/Driver/fsanitize.c
URL:
http://llvm.or
Author: eugenis
Date: Mon May 15 15:43:48 2017
New Revision: 303114
URL: http://llvm.org/viewvc/llvm-project?rev=303114&view=rev
Log:
[asan] One more test for -fsanitize-address-globals-dead-stripping.
Added:
cfe/trunk/test/CodeGen/asan-no-globals-no-comdat.cpp
Added: cfe/trunk/test/CodeGen/
Author: eugenis
Date: Fri Apr 7 18:00:38 2017
New Revision: 299806
URL: http://llvm.org/viewvc/llvm-project?rev=299806&view=rev
Log:
[cfi] Emit __cfi_check stub in the frontend.
Previously __cfi_check was created in LTO optimization pipeline, which
means LLD has no way of knowing about the exist
Author: eugenis
Date: Mon Apr 24 14:34:12 2017
New Revision: 301225
URL: http://llvm.org/viewvc/llvm-project?rev=301225&view=rev
Log:
[asan] Disable ASan global-GC depending on the target and compiler flags.
Added:
cfe/trunk/test/CodeGen/asan-globals-gc.cpp
Modified:
cfe/trunk/lib/CodeGen
Author: eugenis
Date: Tue Apr 25 19:51:06 2017
New Revision: 301374
URL: http://llvm.org/viewvc/llvm-project?rev=301374&view=rev
Log:
[asan] Unconditionally enable GC of globals on COFF.
This change restores pre-r301225 behavior, where linker GC compatible global
instrumentation was used on COFF
Author: eugenis
Date: Fri Nov 11 12:49:49 2016
New Revision: 286613
URL: http://llvm.org/viewvc/llvm-project?rev=286613&view=rev
Log:
[cfi] Enable cfi-icall on ARM and AArch64.
Modified:
cfe/trunk/lib/Driver/ToolChain.cpp
cfe/trunk/test/Driver/fsanitize.c
Modified: cfe/trunk/lib/Driver/T
eugenis added inline comments.
Comment at: lib/Frontend/CompilerInvocation.cpp:732
Args.hasArg(OPT_fsanitize_coverage_trace_pc_guard);
+ Opts.SanitizeThreadMemoryAccess =
+ Args.hasFlag(OPT_fsanitize_thread_memory_access,
It looks like lib/Frontend c
eugenis accepted this revision.
eugenis added a comment.
LGTM
https://reviews.llvm.org/D26461
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Author: eugenis
Date: Fri Nov 11 17:17:36 2016
New Revision: 286669
URL: http://llvm.org/viewvc/llvm-project?rev=286669&view=rev
Log:
Tread TSan LLVM flags to driver: add TSan controlling flags to clang.
Summary:
New clang flags, all default to true:
-f[no-]sanitize-thread-data-races
-f[no-]sanit
This revision was automatically updated to reflect the committed changes.
Closed by commit rL286669: Tread TSan LLVM flags to driver: add TSan
controlling flags to clang. (authored by eugenis).
Changed prior to commit:
https://reviews.llvm.org/D26461?vs=77667&id=77683#toc
Repository:
rL LLVM
Author: eugenis
Date: Wed Feb 15 19:35:23 2017
New Revision: 295267
URL: http://llvm.org/viewvc/llvm-project?rev=295267&view=rev
Log:
Add missing regexp quantifiers in a test.
Modified:
cfe/trunk/test/CodeGen/sanitize-init-order.cpp
Modified: cfe/trunk/test/CodeGen/sanitize-init-order.cpp
UR
eugenis created this revision.
eugenis added a reviewer: pcc.
eugenis added a subscriber: cfe-commits.
eugenis set the repository for this revision to rL LLVM.
Herald added subscribers: srhines, danalbert, tberghammer, rengolin, aemerson.
Backend support added in https://reviews.llvm.org/D25927
eugenis created this revision.
eugenis added a reviewer: pcc.
eugenis added a subscriber: cfe-commits.
eugenis set the repository for this revision to rL LLVM.
CFI (only in the cross-dso mode) fails to set !type annotations when
a function is used before it is defined.
Repository:
rL LLVM
htt
eugenis updated this revision to Diff 76489.
Repository:
rL LLVM
https://reviews.llvm.org/D26164
Files:
lib/CodeGen/CodeGenModule.cpp
test/CodeGen/cfi-icall-cross-dso2.c
Index: test/CodeGen/cfi-icall-cross-dso2.c
===
--- /de
Author: eugenis
Date: Mon Oct 31 17:28:10 2016
New Revision: 285650
URL: http://llvm.org/viewvc/llvm-project?rev=285650&view=rev
Log:
[cfi] Fix missing !type annotation.
CFI (only in the cross-dso mode) fails to set !type annotations when
a function is used before it is defined.
Added:
cfe/t
eugenis closed this revision.
eugenis added a comment.
thanks!
r285650
Repository:
rL LLVM
https://reviews.llvm.org/D26164
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
eugenis accepted this revision.
eugenis added a comment.
This revision is now accepted and ready to land.
LGTM
It appears that this code is not testable in clang. There is a compiler-rt test
in https://reviews.llvm.org/D26355.
https://reviews.llvm.org/D26354
Author: eugenis
Date: Mon Nov 7 15:02:11 2016
New Revision: 286148
URL: http://llvm.org/viewvc/llvm-project?rev=286148&view=rev
Log:
Use -fsanitize-recover instead of -mllvm -msan-keep-going: clang.
Summary: Use -fsanitize-recover instead of -mllvm -msan-keep-going: pass
-fsanitize-recover valu
eugenis added a comment.
This is a good change, but I don't think it is the right fix for PR30940.
Instead of handling this in the NDK, we should change *::getIRStackGuard to
fallback to __stack_chk_guard when targeting an old version.
https://reviews.llvm.org/D26385
___
eugenis accepted this revision.
eugenis added a reviewer: eugenis.
eugenis added a comment.
LGTM
https://reviews.llvm.org/D26385
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
eugenis added inline comments.
Comment at: lib/Frontend/CompilerInvocation.cpp:735
+ OPT_fno_sanitize_thread_data_races,
+ Opts.SanitizeThreadDataRaces);
+ Opts.SanitizeThreadStackTraces =
It seems common to hardcode the defau
eugenis added a comment.
Oh, and this needs a test. See test/Driver/fsanitize.c (search for
-fsanitize-address-use-after-scope as an example).
https://reviews.llvm.org/D26461
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm
eugenis added inline comments.
Comment at: lib/Frontend/CompilerInvocation.cpp:735
+ OPT_fno_sanitize_thread_data_races,
+ Opts.SanitizeThreadDataRaces);
+ Opts.SanitizeThreadStackTraces =
alekseyshl wrote:
> eugenis wrote:
>
eugenis added inline comments.
Comment at: test/Driver/fsanitize.c:288
+// RUN: %clang -target x86_64-linux-gnu -fsanitize=thread
-fsanitize-thread-data-races -fno-sanitize-thread-data-races %s -### 2>&1 |
FileCheck %s --check-prefix=CHECK-TSAN-DATA-RACES-OFF
+// CHECK-TSAN-DAT
eugenis accepted this revision.
eugenis added a reviewer: eugenis.
eugenis added a comment.
Looks great!
https://reviews.llvm.org/D26491
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commi
eugenis added a comment.
In http://reviews.llvm.org/D14858#293651, @danalbert wrote:
> I assume we're going to actually want to go the other direction on this and
> build a shared library for Android's TSAN (see eugenis' comment on
> https://android-review.googlesource.com/#/c/120507/1/core/con
eugenis added a comment.
I vaguely recall that android runtime loader had some symbol lookup
differencies with glibc and that prevented interceptors (when statically linked
into the main executable) from working. Maybe it is not the case now.
http://reviews.llvm.org/D14858
_
eugenis added a comment.
OK, it sounds like static runtime would work fine.
We would still need a way to switch to the shared runtime for the apps (the
workflow when we LD_PRELOAD the runtime into the Zygote to run instrumented
apps on a non-instrumented device). Something like -shared-libasan f
eugenis added a comment.
ping
Repository:
rL LLVM
http://reviews.llvm.org/D14409
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
eugenis accepted this revision.
This revision is now accepted and ready to land.
Comment at: docs/MemorySanitizer.rst:109
@@ -107,3 +108,3 @@
MemorySanitizer can track origins of unitialized values, similar to
Valgrind's --track-origins option. This feature is enabled by
-
eugenis added a comment.
Better -fsanitize-merge-checks, and it should apply to non-trap checks as well
(i.e. SIGILL address should uniquely correspond to the failure source location).
Repository:
rL LLVM
http://reviews.llvm.org/D15208
___
cfe-c
eugenis added a comment.
I misunderstood the meaning of -fsanitize-trap, and now I prefer
-fsanitize-merge-traps for the flag name.
Repository:
rL LLVM
http://reviews.llvm.org/D15208
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http:/
eugenis created this revision.
eugenis added reviewers: kcc, pcc.
eugenis added a subscriber: cfe-commits.
eugenis set the repository for this revision to rL LLVM.
Clang-side cross-DSO CFI.
* Enabled with -fsanitize-cfi-cross-dso
* uses a runtime library, unlike "plain" CFI
* does not yet support
eugenis added a comment.
Hi,
Have you had a chance to look at this?
Repository:
rL LLVM
http://reviews.llvm.org/D14409
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
eugenis added a comment.
In http://reviews.llvm.org/D14409#306272, @EricWF wrote:
> Does the `inline` keyword have any effect when it's on function definitions
> that are externally instantiated?
I could not detect any difference in behavior with or without inline keyword.
Remove it?
Reposit
eugenis created this revision.
eugenis added reviewers: EricWF, mclow.lists.
eugenis added a subscriber: cfe-commits.
eugenis set the repository for this revision to rL LLVM.
These are the cases when an out-of-class definition of a method is marked
_LIBCPP_INLINE_VISIBILITY, but the in-class decl
eugenis added a comment.
Should be fine.
Repository:
rL LLVM
http://reviews.llvm.org/D15395
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
Author: eugenis
Date: Wed Dec 9 16:32:36 2015
New Revision: 255166
URL: http://llvm.org/viewvc/llvm-project?rev=255166&view=rev
Log:
Add 3 more missing inline/visibility attributes.
These are the cases when an out-of-class definition of a method is
marked _LIBCPP_INLINE_VISIBILITY, but the in-cl
eugenis closed this revision.
eugenis added a comment.
r255166
Repository:
rL LLVM
http://reviews.llvm.org/D15395
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
eugenis added a comment.
In http://reviews.llvm.org/D14409#306379, @eugenis wrote:
> In http://reviews.llvm.org/D14409#306272, @EricWF wrote:
>
> > Does the `inline` keyword have any effect when it's on function definitions
> > that are externally instantiated?
>
>
> I could not detect any diffe
eugenis added a comment.
In http://reviews.llvm.org/D14409#306531, @eugenis wrote:
> In http://reviews.llvm.org/D14409#306379, @eugenis wrote:
>
> > In http://reviews.llvm.org/D14409#306272, @EricWF wrote:
> >
> > > Does the `inline` keyword have any effect when it's on function
> > > definition
Author: eugenis
Date: Wed Dec 9 17:42:30 2015
New Revision: 255177
URL: http://llvm.org/viewvc/llvm-project?rev=255177&view=rev
Log:
Remove visibility attributes from out-of-class method definitions in iostreams.
No point in pretending that these methods are hidden - they are
actually exported f
eugenis closed this revision.
eugenis added a comment.
r255177
Thanks for the review!
Repository:
rL LLVM
http://reviews.llvm.org/D14409
___
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-co
Author: eugenis
Date: Wed Dec 9 18:44:42 2015
New Revision: 255186
URL: http://llvm.org/viewvc/llvm-project?rev=255186&view=rev
Log:
Replace cmake check for printf with a check for fopen.
Printf is a builtin, and the check fails with -Werror because of a clang
warning about an incompatible redec
Author: eugenis
Date: Wed Dec 9 18:45:18 2015
New Revision: 255187
URL: http://llvm.org/viewvc/llvm-project?rev=255187&view=rev
Log:
Replace cmake check for printf with a check for fopen.
Printf is a builtin, and the check fails with -Werror because of a clang
warning about an incompatible redec
Author: eugenis
Date: Wed Dec 9 18:47:08 2015
New Revision: 255188
URL: http://llvm.org/viewvc/llvm-project?rev=255188&view=rev
Log:
Replace cmake check for printf with a check for fopen.
Printf is a builtin, and the check fails with -Werror because of a clang
warning about an incompatible redec
eugenis added a dependency: D12502: [libcxx] Better constain tuples
constructors -- Fix PR23256 and PR22806.
eugenis added a comment.
Note, this breaks tuple_cat.pass.cpp test.
With -O0, replacing always_inline with internal_linkage results in less
optimization being done (namely, no inlining h
eugenis created this revision.
eugenis added reviewers: EricWF, mclow.lists.
eugenis added a subscriber: cfe-commits.
eugenis set the repository for this revision to rL LLVM.
This is a follow-up to r252385.
For some reason, I missed a lot of cases when the visibility attribute was
applied to the
101 - 200 of 304 matches
Mail list logo