chandlerc added a comment.
Other thoughts:
Can we get 'bootstrap-check'? 'bootstrap-ccheck-foo' for all the 'check-foo'
variants?
I think 'bootstrap' shouldn't do 'bootstrap-install', it should be equivalent
to 'bootstrap-check-all' IMO.
http://reviews.llvm.org/D11743
chandlerc added a comment.
This is pretty awesome.
A question -- how can we document better what this does? In particular, the
fact that we have both 'bootstrap-install' and 'install' seems... confusing. Is
it possible to make 'install' == 'bootstrap-install' in the case where the
CMake
Author: chandlerc
Date: Mon Aug 3 22:52:52 2015
New Revision: 243945
URL: http://llvm.org/viewvc/llvm-project?rev=243945view=rev
Log:
[UB] Fix two cases of UB in copy/pasted code from SmallVector.
We should really stop copying and pasting code around. =/
Found by UBSan.
Modified:
Author: chandlerc
Date: Mon Aug 3 22:52:56 2015
New Revision: 243946
URL: http://llvm.org/viewvc/llvm-project?rev=243946view=rev
Log:
[UB] Fix the two ways that we would try to memcpy from a null buffer in
the nested name specifier code.
First, skip the entire thing when the input is empty.
Author: chandlerc
Date: Mon Aug 3 22:52:58 2015
New Revision: 243947
URL: http://llvm.org/viewvc/llvm-project?rev=243947view=rev
Log:
[UB] When attaching empty strings to the AST, use an empty StringRef
rather than forcing the bump pointer allocator to produce a viable
pointer. This also fixes
Author: chandlerc
Date: Mon Aug 3 22:53:04 2015
New Revision: 243950
URL: http://llvm.org/viewvc/llvm-project?rev=243950view=rev
Log:
[UB] Avoid a really broken call to realloc that would later result in
a bad call to memcpy.
When we only have a buffer from one of the two reparse calls, we can
Author: chandlerc
Date: Mon Aug 3 22:53:01 2015
New Revision: 243949
URL: http://llvm.org/viewvc/llvm-project?rev=243949view=rev
Log:
[UB] Guard two calls to memcpy in generated attribute code to handle
null StringRef objects as inputs.
Found by UBSan.
Modified:
Author: chandlerc
Date: Mon Aug 3 22:53:00 2015
New Revision: 243948
URL: http://llvm.org/viewvc/llvm-project?rev=243948view=rev
Log:
[UB] Another place where we were trying to put string data into
a BumpPtrAllocator. This at least now handles the case where there is no
concatentation without
chandlerc accepted this revision.
chandlerc added a comment.
This revision is now accepted and ready to land.
Very much prefer preserving the original order.
We were already working around the reorderings by putting the D and U groups
into a single call. =/ This approach seems *much* cleaner.
chandlerc added a subscriber: chandlerc.
chandlerc added a comment.
Comments from the peanut gallery...
Comment at: lib/Driver/Tools.cpp:9482-9488
@@ -9481,13 +9481,9 @@
- // Any -O option passes through without translation. What about -Ofast ?
- if (Arg *A =
FWIW, I agree, this seems really important.
On Sun, Jul 26, 2015 at 8:54 AM Logan Chien tzuhsiang.ch...@gmail.com
wrote:
Hi Hans and Nick,
I would like to propose to merge following patches to libunwind 3.7 branch:
r242642: libunwind: Introduce __libunwind_config.h.
r243073: unwind: Fix
chandlerc accepted this revision.
chandlerc added a comment.
This revision is now accepted and ready to land.
Feel free to address the comments below in a follow-up patch. This seems a fine
next step, LGTM
Comment at: clang-tidy/misc/UnusedParametersCheck.cpp:76
@@ +75,3 @@
+
chandlerc accepted this revision.
chandlerc added a comment.
This revision is now accepted and ready to land.
LGTM
http://reviews.llvm.org/D11322
___
cfe-commits mailing list
cfe-commits@cs.uiuc.edu
On Mon, Jul 13, 2015 at 7:27 PM Richard Smith rich...@metafoo.co.uk wrote:
On Mon, Jul 13, 2015 at 6:02 PM, Adrian Prantl apra...@apple.com wrote:
On Jul 13, 2015, at 5:47 PM, Richard Smith rich...@metafoo.co.uk wrote:
On Mon, Jul 13, 2015 at 3:06 PM, Adrian Prantl apra...@apple.com wrote:
On Tue, Jul 7, 2015 at 10:27 PM Xinliang David Li davi...@google.com
wrote:
On Tue, Jul 7, 2015 at 6:06 PM, Chandler Carruth chandl...@gmail.com
wrote:
On Tue, Jul 7, 2015 at 4:11 PM Easwaran Raman era...@google.com wrote:
I'm reviving this thread after a while and CCing cfe-commits
On Wed, Jul 8, 2015 at 1:46 PM Hal Finkel hfin...@anl.gov wrote:
- Original Message -
From: Xinliang David Li davi...@google.com
To: Chandler Carruth chandl...@gmail.com
Cc: cfe-commits@cs.uiuc.edu, llvm...@cs.uiuc.edu List
llvm...@cs.uiuc.edu
Sent: Wednesday, July 8, 2015 12
On Tue, Jul 7, 2015 at 4:11 PM Easwaran Raman era...@google.com wrote:
I'm reviving this thread after a while and CCing cfe-commits as
suggested by David Blaikie. I've also collected numbers building
chrome (from chromium, on Linux) with and without this patch as
suggested by David. I've
LGTM with a trivial comment tweak below.
Comment at: lib/Driver/Driver.cpp:1255-1256
@@ -1257,4 +1254,4 @@
std::unique_ptrAction Current(new InputAction(*InputArg, InputType));
-for (SmallVectorImplphases::ID::iterator
- i = PL.begin(), e = PL.end(); i != e;
I think you should just fix the code. This seems clearly a bug.
With that fix, this LGTM.
Comment at: lib/Driver/ToolChains.cpp:3096
@@ -3093,2 +3095,3 @@
return powerpc64-linux-gnu;
+// Fallthrough
case llvm::Triple::ppc64le:
This is certainly
OMG I hate the commandeer thing in Phab. I always think it will make me a
reviewr. :: sigh ::
I go to try and fix this.
http://reviews.llvm.org/D10609
EMAIL PREFERENCES
http://reviews.llvm.org/settings/panel/emailpreferences/
___
cfe-commits
Oh great, I can't. Doug, sorry, just commandeer the revision back. Sorry i
broke the review tool...
http://reviews.llvm.org/D10609
EMAIL PREFERENCES
http://reviews.llvm.org/settings/panel/emailpreferences/
___
cfe-commits mailing list
Please update once this is rebased on top of the formatted variant. I'd like to
glance at it then before I stamp it. Thanks for the cleanup!!
http://reviews.llvm.org/D10609
EMAIL PREFERENCES
http://reviews.llvm.org/settings/panel/emailpreferences/
Author: chandlerc
Date: Wed Jun 24 05:24:30 2015
New Revision: 240530
URL: http://llvm.org/viewvc/llvm-project?rev=240530view=rev
Log:
Remove a limited and somewhat questionable DenseMapInfo specialization
for StringRef now that the core DenseMap library provides this facility.
Modified:
LGTM, thanks for the cleanup!
Comment at: lib/Driver/Tools.h:196-197
@@ -192,7 +195,4 @@
namespace hexagon {
- // For Hexagon, we do not need to instantiate tools for PreProcess,
PreCompile and Compile.
- // We simply use clang -cc1 for those actions.
- class
The content here seems fine, but clang-format has changed a lot of unrelated
lines of code.
How about sending me just a fresh clang-format of lib/Driver? You'll want to
look at the result and revert some hunks that don't make sense I suspect, but
that would give you a much more consistent and
Comment at: lib/Driver/Tools.h:170
@@ -171,1 +169,3 @@
+Compiler(const ToolChain TC)
+: Common(gcc::Compile, gcc frontend, TC) {}
Above you renamed gcc::Preprocess to gcc::Preprocessor, but here (and
almost everywhere else in my spot check) the text
Just to be clear, this commit seems good to me, and I'm fine with it being
in, but please don't submit patches without review once you've asked for it.
On Tue, Jun 16, 2015 at 2:22 PM Chandler Carruth chandl...@google.com
wrote:
Doug, why did you commit this without review? That's not really
Doug, why did you commit this without review? That's not really cool after
you've asked folks to code review it and without any update.
On Tue, Jun 16, 2015 at 12:43 PM Douglas Katzman do...@google.com wrote:
REPOSITORY
rL LLVM
http://reviews.llvm.org/D10444
Files:
I agree with Joerg about submitting the separate cleanup change separately
(feel free to just pull it out and commit it).
More detailed comments below.
Comment at: lib/Driver/ToolChains.h:875-877
@@ -874,1 +874,5 @@
+/// MoviClang - A tool chain using the compiler that the
I think we should do this, and I'll try to give a plausibly principled
rationale.
We have two uses of {}s that I care about: braced init lists and blocks. For
braced init lists, we routinely omit padding spaces around the {}s:
foo({x, y z});
return {{1, 2, 3}};
On the other hand, with
In all of these cases, the cause of the spacing is not the {. We put spaces
on both sides of a multiply, or after a comma.
The debug thing is (IMO) a bad example in every way. It's a macro, the
contents are a block, and yuck.
On Fri, Jun 12, 2015 at 1:00 AM Daniel Jasper djas...@google.com
int *f(); // no braces, irrelevant
int *f() { // space before the brace
auto f() { // space before the brace
auto f() - int { // space before the brace
auto f() - int * { // space before the brace
On Fri, Jun 12, 2015 at 1:16 AM Daniel Jasper djas...@google.com wrote:
Ok. To be clear, I
I think the logic here should be bundled up in a macro in the LLVM cmake tree
so that we can use it both for Clang's man page generation but also generating
other man pages. Does that make sense?
http://reviews.llvm.org/D10304
EMAIL PREFERENCES
Folks on cfe-dev seem happy with this direction, so go ahead and submit with a
tweaked comment as below...
Comment at: include/clang/Driver/ToolChain.h:166-171
@@ -165,3 +165,8 @@
/// Choose a tool to use to handle the action \p JA.
- Tool *SelectTool(const JobAction JA)
It's possible. I don't think my script and my clang format binary are
version locked. Should they be? Is that important?
On Thu, Jun 4, 2015, 21:45 Manuel Klimek kli...@google.com wrote:
Were you on an outdated version of clang-format?
On Thu, Jun 4, 2015, 11:27 PM Chandler Carruth chandl
Suggested some improvements to the comments below. Generally, this makes sense,
but I think you should also mention your end-goal in the commit log. I happen
to know because we've chatted, and so this change makes sense to me, but others
won't have that context.
For those reading the review
After this, all my code produces this error... The message isn't very
helpful, how should I debug this?
On Tue, Jun 2, 2015 at 5:08 AM Manuel Klimek kli...@google.com wrote:
Author: klimek
Date: Tue Jun 2 07:01:50 2015
New Revision: 238823
URL:
It's because the dict doesn't always have the key.
Fixed in r239098.
On Thu, Jun 4, 2015 at 2:21 PM Chandler Carruth chandl...@google.com
wrote:
After this, all my code produces this error... The message isn't very
helpful, how should I debug this?
On Tue, Jun 2, 2015 at 5:08 AM Manuel
Author: chandlerc
Date: Thu Jun 4 16:23:07 2015
New Revision: 239098
URL: http://llvm.org/viewvc/llvm-project?rev=239098view=rev
Log:
Fix terrible python goof in clang-format.py which broke my vim
integration.
Nothing is more important in life than clang-format integration with
vim. ;]
On Fri, May 29, 2015 at 11:40 AM İsmail Dönmez ism...@donmez.ws wrote:
On Fri, May 29, 2015 at 9:32 PM, Richard Smith rich...@metafoo.co.uk
wrote:
On Fri, May 29, 2015 at 5:56 AM, İsmail Dönmez ism...@donmez.ws wrote:
On Thu, May 28, 2015 at 11:48 PM, Chandler Carruth chandl...@gmail.com
On Thu, May 28, 2015 at 5:09 AM İsmail Dönmez ism...@donmez.ws wrote:
Hi,
On Thu, May 28, 2015 at 4:52 AM, Chandler Carruth chandl...@gmail.com
wrote:
Author: chandlerc
Date: Wed May 27 20:52:38 2015
New Revision: 238389
URL: http://llvm.org/viewvc/llvm-project?rev=238389view=rev
Author: chandlerc
Date: Thu May 28 16:10:31 2015
New Revision: 238498
URL: http://llvm.org/viewvc/llvm-project?rev=238498view=rev
Log:
[omp] Fix a typo in a comment and a line I forgot to clang-format that
Justin pointed out in post-commit review.
Modified:
cfe/trunk/lib/Driver/Tools.cpp
Author: chandlerc
Date: Thu May 28 16:20:14 2015
New Revision: 238500
URL: http://llvm.org/viewvc/llvm-project?rev=238500view=rev
Log:
[omp] Loosen the driver test enough so that overriding the defaults
works well for folks.
This isn't terribly clean (sadly) but after chatting with both Eric and
On Thu, May 28, 2015 at 2:46 PM Andrey Bokhanko andreybokha...@gmail.com
wrote:
Chandler,
Issue #1 -- not an issue at all. Personally, I *welcome* your fixes,
and it's logical they are coming from you -- as apart of users you
mentioned, there are zero documented users who *rely* on clang
Thanks for the review!
I'll make these changes shortly, just responding below:
On Thu, May 28, 2015 at 12:39 AM Justin Bogner m...@justinbogner.com
wrote:
Chandler Carruth chandl...@gmail.com writes:
+ /// The legacy name for the LLVM OpenMP runtim from when it was the
Intel
On Thu, May 28, 2015 at 5:48 AM Andrey Bokhanko andreybokha...@gmail.com
wrote:
Chandler,
+set(CLANG_DEFAULT_OPENMP_RUNTIME libgomp CACHE STRING
+ Default OpenMP runtime used by -fopenmp.)
This effectively invalidates instructions recently published on LLVM
blog
Looks good (assuming with the LLVM change).
REPOSITORY
rL LLVM
http://reviews.llvm.org/D9153
EMAIL PREFERENCES
http://reviews.llvm.org/settings/panel/emailpreferences/
___
cfe-commits mailing list
cfe-commits@cs.uiuc.edu
On Wed, May 27, 2015 at 3:51 PM Chandler Carruth chandl...@google.com
wrote:
On Wed, May 20, 2015 at 12:02 AM Chandler Carruth chandl...@google.com
wrote:
I don't think this should have been committed yet.
It's been a week without response.
1) The CMake build for the OpenMP stuff
Author: chandlerc
Date: Wed May 27 20:52:38 2015
New Revision: 238389
URL: http://llvm.org/viewvc/llvm-project?rev=238389view=rev
Log:
[omp] Re-work Clang's handling of -fopenmp and undo r237769.
This isn't an actual revert of r237769, it just restores the behavior of
the Clang driver prior to
On Wed, May 20, 2015 at 12:02 AM Chandler Carruth chandl...@google.com
wrote:
I don't think this should have been committed yet.
It's been a week without response.
1) The CMake build for the OpenMP stuff is still really broken and not
well integrated with the rest of the LLVM project
2
On Tue, May 26, 2015 at 7:54 PM David Majnemer david.majne...@gmail.com
wrote:
On Tue, May 26, 2015 at 7:22 PM, Chaoren Lin chaor...@google.com wrote:
Hi David,
Thanks for your concern.
The commit was intentional. I was hoping this small commit would fall
under this part:
smaller
Yea, i think Daniel's patch was a closer start. Anyways, thanks for taking
a stab at this.
http://reviews.llvm.org/D9875
EMAIL PREFERENCES
http://reviews.llvm.org/settings/panel/emailpreferences/
___
cfe-commits mailing list
Yea, i think Daniel's patch was a closer start. Anyways, thanks for taking
a stab at this.
On Wed, May 20, 2015 at 8:04 PM Alexey Bataev a.bat...@hotmail.com wrote:
Already has been fixed by Richard Smith.
http://reviews.llvm.org/D9875
EMAIL PREFERENCES
LGTM
http://reviews.llvm.org/D8784
EMAIL PREFERENCES
http://reviews.llvm.org/settings/panel/emailpreferences/
___
cfe-commits mailing list
cfe-commits@cs.uiuc.edu
http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
Can you add some basic testing in test/Driver for the flag handling, and in
test/Preprocessor for the macros that should be predefined?
http://reviews.llvm.org/D8784
EMAIL PREFERENCES
http://reviews.llvm.org/settings/panel/emailpreferences/
___
Before we go here, I think we should seriously ask whether we want to undo that
restriction on shift left.
I can see arguments both for and against it.
Does GCC already do this optimization?
http://reviews.llvm.org/D8899
EMAIL PREFERENCES
This doesn't seem correct for a negative LHS where we only shift out 1s -- I
think that should be valid.
Essentially, we seem to have a strange circumstance here where NUW as specified
is not a subset of the restrictions of NSW.
http://reviews.llvm.org/D8899
EMAIL PREFERENCES
Looks good, go for it.
http://reviews.llvm.org/D8694
EMAIL PREFERENCES
http://reviews.llvm.org/settings/panel/emailpreferences/
___
cfe-commits mailing list
cfe-commits@cs.uiuc.edu
http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
Unsure if you were actually looking for review of this Chris, but it looks fine.
http://reviews.llvm.org/D7797
EMAIL PREFERENCES
http://reviews.llvm.org/settings/panel/emailpreferences/
___
cfe-commits mailing list
cfe-commits@cs.uiuc.edu
This seems stalled. Update and add me back if you want me to take a look at it.
http://reviews.llvm.org/D2898
EMAIL PREFERENCES
http://reviews.llvm.org/settings/panel/emailpreferences/
___
cfe-commits mailing list
cfe-commits@cs.uiuc.edu
Resigning as this patch has gotten quite stale. If someone is still interested,
update or send a new patch.
http://reviews.llvm.org/D3542
EMAIL PREFERENCES
http://reviews.llvm.org/settings/panel/emailpreferences/
___
cfe-commits mailing list
I'm not a good reviewer here (I don't work on Windows) and Reid seems like an
excellent reviewer.
http://reviews.llvm.org/D5268
EMAIL PREFERENCES
http://reviews.llvm.org/settings/panel/emailpreferences/
___
cfe-commits mailing list
I'm definitely not the right reviewer for this. Richard or Doug make much more
sense.
http://reviews.llvm.org/D5611
EMAIL PREFERENCES
http://reviews.llvm.org/settings/panel/emailpreferences/
___
cfe-commits mailing list
cfe-commits@cs.uiuc.edu
Dave, not sure there is anything else you need from me here, but if so, add me
back and let me know. Until then, i'm moving this off my review queue.
http://reviews.llvm.org/D4956
EMAIL PREFERENCES
http://reviews.llvm.org/settings/panel/emailpreferences/
FYI, as I commented on the original discussion, I'd personally like to see this
have some versioning scheme before it goes in rather than a binary unstable.
Or some plan for getting there.
Otherwise this looks good as a first step.
http://reviews.llvm.org/D6712
EMAIL PREFERENCES
Er, I believe you already submitted this with Manuel's review. In the future,
please have him actually mark the review on the mailing list so its clear to
others that it was in fact reviewed.
http://reviews.llvm.org/D7289
EMAIL PREFERENCES
Author: chandlerc
Date: Fri Mar 27 16:40:58 2015
New Revision: 233426
URL: http://llvm.org/viewvc/llvm-project?rev=233426view=rev
Log:
[Modules] Work around PR23030 again, in a different code path, where
I again added the reasonable assertions and they again fired during
a modules self-host.
Author: chandlerc
Date: Fri Mar 27 20:10:44 2015
New Revision: 233462
URL: http://llvm.org/viewvc/llvm-project?rev=233462view=rev
Log:
[Modules] Don't compute a modules cache path if we're not using modules!
Notably, this prevents us from doing *tons* of work to compute the
modules hash,
Author: chandlerc
Date: Thu Mar 26 03:49:55 2015
New Revision: 233263
URL: http://llvm.org/viewvc/llvm-project?rev=233263view=rev
Log:
[Modules] Add some more fun code to my modules stress test, this time
templates. Turns out all of this works correctly (so far). But it should
cover more code
Author: chandlerc
Date: Thu Mar 26 03:32:49 2015
New Revision: 233261
URL: http://llvm.org/viewvc/llvm-project?rev=233261view=rev
Log:
[Modules] Make #pragma weak undeclared identifiers be tracked
deterministically.
This fixes a latent issue where even Clang's Sema (and diagnostics) were
Author: chandlerc
Date: Thu Mar 26 04:08:15 2015
New Revision: 233264
URL: http://llvm.org/viewvc/llvm-project?rev=233264view=rev
Log:
[Modules] Preserve source order for the map of late parsed templates.
Clang was inserting these into a dense map. While it never iterated the
dense map during
Just FYI in case others see this:
On Wed, Mar 25, 2015 at 8:11 PM, Chandler Carruth chandl...@gmail.com
wrote:
+assert(ConstructorNameSet.empty() Failed to find all of the
visible
+ constructors by walking all
On Thu, Mar 26, 2015 at 1:12 PM, Richard Smith rich...@metafoo.co.uk
wrote:
So, the problem seems to be that the class has implicit special members
that are not in the list of lexical declarations of the class but are in
the visible lookup results. In order for this to happen, you might need
Author: chandlerc
Date: Thu Mar 26 19:31:20 2015
New Revision: 29
URL: http://llvm.org/viewvc/llvm-project?rev=29view=rev
Log:
[Modules] Sort the file IDs prior to building the flattened array of
DeclIDs so that in addition to be grouped by file, the order of these
groups is stable.
Author: chandlerc
Date: Thu Mar 26 19:47:43 2015
New Revision: 233342
URL: http://llvm.org/viewvc/llvm-project?rev=233342view=rev
Log:
[Modules] Make our on-disk hash table of selector IDs be built in
a deterministic order.
This uses a MapVector to track the insertion order of selectors.
Found
Author: chandlerc
Date: Thu Mar 26 19:55:05 2015
New Revision: 233343
URL: http://llvm.org/viewvc/llvm-project?rev=233343view=rev
Log:
[Modules] Make Sema's map of referenced selectors have a deterministic
order based on order of insertion.
This should cause both our warnings about these and the
Author: chandlerc
Date: Thu Mar 26 18:45:40 2015
New Revision: 21
URL: http://llvm.org/viewvc/llvm-project?rev=21view=rev
Log:
[Modules] Delete stale, pointless code. All tests still pass with this
logic removed.
This logic was both inserting all builtins into the identifier table and
Author: chandlerc
Date: Thu Mar 26 18:54:15 2015
New Revision: 22
URL: http://llvm.org/viewvc/llvm-project?rev=22view=rev
Log:
[Modules] Make the AST serialization always use lexicographic order when
traversing the identifier table.
No easy test case as this table is somewhere between
Author: chandlerc
Date: Thu Mar 26 18:58:11 2015
New Revision: 23
URL: http://llvm.org/viewvc/llvm-project?rev=23view=rev
Log:
[Modules] Fix an obvious lack of deterministic ordering when processing
rewritten decls for Objective-C modules.
Found by inspection and completely obvious, so
Author: chandlerc
Date: Thu Mar 26 18:59:47 2015
New Revision: 24
URL: http://llvm.org/viewvc/llvm-project?rev=24view=rev
Log:
[Modules] Clean up some code that was manually replicating what
SmallSetVector provides directly.
Modified:
cfe/trunk/include/clang/Serialization/ASTWriter.h
Author: chandlerc
Date: Thu Mar 26 19:01:44 2015
New Revision: 25
URL: http://llvm.org/viewvc/llvm-project?rev=25view=rev
Log:
[Modules] Fix another pointer keyed set that we iterate over while
writing a module to be a set-vector to preserve insertion order.
No test case, found by
On Thu, Mar 26, 2015 at 1:39 PM, Chandler Carruth chandl...@gmail.com
wrote:
On Thu, Mar 26, 2015 at 1:12 PM, Richard Smith rich...@metafoo.co.uk
wrote:
So, the problem seems to be that the class has implicit special members
that are not in the list of lexical declarations of the class
Author: chandlerc
Date: Thu Mar 26 17:22:22 2015
New Revision: 233325
URL: http://llvm.org/viewvc/llvm-project?rev=233325view=rev
Log:
[Modules] Fix tiny bug where we failed to get the canonical decl when
deserializing an inherited constructor.
This is the exact same logic we use when
Author: chandlerc
Date: Thu Mar 26 17:27:09 2015
New Revision: 233327
URL: http://llvm.org/viewvc/llvm-project?rev=233327view=rev
Log:
[Modules] Fix a sneaky bug in r233249 where we would look for implicit
constructors in the current lexical context even though name lookup
found them via some
Author: chandlerc
Date: Thu Mar 26 20:48:11 2015
New Revision: 233348
URL: http://llvm.org/viewvc/llvm-project?rev=233348view=rev
Log:
[Modules] When walking the lookup results in a namespace, sort them by
declaration name so that we mark declarations for emission in
a deterministic order (and in
, and would serve the exact same purpose.
On Tue, Mar 24, 2015 at 2:21 PM, Chandler Carruth chandl...@google.com
wrote:
On Mon, Mar 23, 2015 at 10:37 PM, Ben Langmuir blangm...@apple.com
wrote:
On Mar 23, 2015, at 10:21 PM, Chandler Carruth chandl...@google.com
wrote:
On Mon, Mar 23, 2015 at 9:54
Sorry, yes, the correct fix was to disable the tests.
I tried to do that originally, but apparently didn't disable enough
(despite running the tests many times).
On Wed, Mar 25, 2015 at 12:02 AM, Daniel Jasper djas...@google.com wrote:
The problem is that the test is for testing
On Tue, Mar 24, 2015 at 7:07 PM, David Blaikie dblai...@gmail.com wrote:
On Mar 24, 2015 5:42 PM, Chandler Carruth chandl...@gmail.com wrote:
Author: chandlerc
Date: Tue Mar 24 19:34:51 2015
New Revision: 233156
URL: http://llvm.org/viewvc/llvm-project?rev=233156view=rev
Log
Author: chandlerc
Date: Wed Mar 25 22:11:40 2015
New Revision: 233249
URL: http://llvm.org/viewvc/llvm-project?rev=233249view=rev
Log:
[Modules] A second attempt at writing out on-disk hash tables for the
decl context lookup tables.
The first attepmt at this caused problems. We had significantly
Author: chandlerc
Date: Wed Mar 25 23:27:10 2015
New Revision: 233251
URL: http://llvm.org/viewvc/llvm-project?rev=233251view=rev
Log:
[Modules] Delete a bunch of complex code for ensuring visible decls in
updated decl contexts get emitted.
Since this code was added, we have newer vastly simpler
Author: chandlerc
Date: Tue Mar 24 16:44:25 2015
New Revision: 233117
URL: http://llvm.org/viewvc/llvm-project?rev=233117view=rev
Log:
[Modules] Stop creating timestamps for the modules cache and trying to
prune it when we have disabled implicit module generation and thus are
not using any cached
Author: chandlerc
Date: Tue Mar 24 20:30:02 2015
New Revision: 233163
URL: http://llvm.org/viewvc/llvm-project?rev=233163view=rev
Log:
[Modules] Disable the diff of the merged module, there is still some
non-determinism here, I just got lucky a bunch of times on my system.
Modified:
Author: chandlerc
Date: Tue Mar 24 19:34:51 2015
New Revision: 233156
URL: http://llvm.org/viewvc/llvm-project?rev=233156view=rev
Log:
[Modules] When writing out the on-disk hash table for the decl context
lookup tables, we need to establish a stable ordering for constructing
the hash table. This
Author: chandlerc
Date: Tue Mar 24 20:02:12 2015
New Revision: 233162
URL: http://llvm.org/viewvc/llvm-project?rev=233162view=rev
Log:
[Modules] Make the DeclUpdates map be processed in insertion order.
This fixes my stress tests non-determinism so far. However, I've not
started playing with
Author: chandlerc
Date: Tue Mar 24 16:18:10 2015
New Revision: 233115
URL: http://llvm.org/viewvc/llvm-project?rev=233115view=rev
Log:
[Modules] Start making explicit modules produce deterministic output.
There are two aspects of non-determinism fixed here, which was the
minimum required to
On Mon, Mar 23, 2015 at 10:37 PM, Ben Langmuir blangm...@apple.com wrote:
On Mar 23, 2015, at 10:21 PM, Chandler Carruth chandl...@google.com
wrote:
On Mon, Mar 23, 2015 at 9:54 PM, Ben Langmuir blangm...@apple.com wrote:
On Mar 23, 2015, at 5:38 PM, Chandler Carruth chandl...@google.com
On Mon, Mar 23, 2015 at 9:54 PM, Ben Langmuir blangm...@apple.com wrote:
On Mar 23, 2015, at 5:38 PM, Chandler Carruth chandl...@google.com
wrote:
Digging this thread back up, I'm starting to work on deterministic output
of modules. This is *really* important. Unfortunately, it's almost
On Mon, Mar 23, 2015 at 10:37 PM, Ben Langmuir blangm...@apple.com wrote:
On Mar 23, 2015, at 10:21 PM, Chandler Carruth chandl...@google.com
wrote:
On Mon, Mar 23, 2015 at 9:54 PM, Ben Langmuir blangm...@apple.com wrote:
On Mar 23, 2015, at 5:38 PM, Chandler Carruth chandl...@google.com
Digging this thread back up, I'm starting to work on deterministic output
of modules. This is *really* important. Unfortunately, it's almost
impossible to make progress as long as this code is actually firing because
it kind of makes things definitively non-deterministic. ;]
Ben, do you have any
Author: chandlerc
Date: Thu Mar 19 17:39:51 2015
New Revision: 232778
URL: http://llvm.org/viewvc/llvm-project?rev=232778view=rev
Log:
[Modules] Implement __builtin_isinf_sign in Clang.
Somehow, we never managed to implement this fully. We could constant
fold it like crazy, including constant
1 - 100 of 2467 matches
Mail list logo
