mac: Pull a prebuilt libc++-static.a.

mac: Pull a prebuilt libc++-static.a.

Not used for anything yet.

BUG=400091
R=mark@chromium.org
TBR=torne

Committed: https://chromium.googlesource.com/chromium/src/+/7ef5508bd7b471d7b4a67886e23feb89a3a13713

[Nico, 2015-10-22 19:13:25]

thakis@chromium.org changed reviewers:
+ mark@chromium.org

[Nico, 2015-10-22 19:13:25]

This is pass 1 at a prebuilt libc++-static. It doesn't strip the call to
cxa_demangle() in libcxxabi/src/cxa_default_handlers.cpp yet, so binaries linked
against this will be 500k larger than necessary. If you think this is basically
ok, I'll tweak the stuff in the README to do that too, and I'll upload a new
prebuilt with that change.

[Mark Mentovai, 2015-10-22 21:54:27]

Yeah, it’s basically OK.

https://codereview.chromium.org/1415213004/diff/1/third_party/libc++-static/R...
File third_party/libc++-static/README.chromium (right):

https://codereview.chromium.org/1415213004/diff/1/third_party/libc++-static/R...
third_party/libc++-static/README.chromium:28: cd /tmp/
This is basically a script. Can you make it a real build.sh or something?

https://codereview.chromium.org/1415213004/diff/1/third_party/libc++-static/R...
third_party/libc++-static/README.chromium:42: c++ -c -nostdinc++
-I../../libcxx/include/ -O3 -std=c++11 -fstrict-aliasing ../../libcxx/src/*.cpp
-fvisibility=hidden -fvisibility-inlines-hidden -mmacosx-version-min=10.6 -arch
i386 -arch x86_64
Whose c++, Chrome’s or the system’s? Preferably Chrome’s, right?

https://codereview.chromium.org/1415213004/diff/1/third_party/libc++-static/R...
third_party/libc++-static/README.chromium:48: c++ -c -nostdinc++
-I../../libcxx/include/ -I../../libcxxabi/include -O3 -std=c++11
-fstrict-aliasing ../../libcxxabi/src/*.cpp -fvisibility=hidden
-fvisibility-inlines-hidden -mmacosx-version-min=10.6 -arch i386 -arch x86_64
Where do these object files go?

https://codereview.chromium.org/1415213004/diff/1/third_party/libc++-static/R...
third_party/libc++-static/README.chromium:57: python
~/depot_tools/upload_to_google_storage.py -b chromium-libcpp
third_party/libc++-static.a
depot_tools should be in the PATH for most people, and
upload_to_google_storage.py is +x, so this can just be
“upload_to_google_storage.py -b …”

[Nico, 2015-10-22 22:05:41]

Thanks! New CL coming later today.

https://codereview.chromium.org/1415213004/diff/1/third_party/libc++-static/R...
File third_party/libc++-static/README.chromium (right):

https://codereview.chromium.org/1415213004/diff/1/third_party/libc++-static/R...
third_party/libc++-static/README.chromium:28: cd /tmp/
On 2015/10/22 21:54:27, Mark Mentovai wrote:
> This is basically a script. Can you make it a real build.sh or something?

Yeah, I really should. Will do.

https://codereview.chromium.org/1415213004/diff/1/third_party/libc++-static/R...
third_party/libc++-static/README.chromium:42: c++ -c -nostdinc++
-I../../libcxx/include/ -O3 -std=c++11 -fstrict-aliasing ../../libcxx/src/*.cpp
-fvisibility=hidden -fvisibility-inlines-hidden -mmacosx-version-min=10.6 -arch
i386 -arch x86_64
On 2015/10/22 21:54:27, Mark Mentovai wrote:
> Whose c++, Chrome’s or the system’s? Preferably Chrome’s, right?

That'd be better, yes. Will change.

https://codereview.chromium.org/1415213004/diff/1/third_party/libc++-static/R...
third_party/libc++-static/README.chromium:48: c++ -c -nostdinc++
-I../../libcxx/include/ -I../../libcxxabi/include -O3 -std=c++11
-fstrict-aliasing ../../libcxxabi/src/*.cpp -fvisibility=hidden
-fvisibility-inlines-hidden -mmacosx-version-min=10.6 -arch i386 -arch x86_64
On 2015/10/22 21:54:27, Mark Mentovai wrote:
> Where do these object files go?

Note the subtle first * in `libcxx*/*.o` in the libtool line below :-)

[Mark Mentovai, 2015-10-22 22:57:45]

https://codereview.chromium.org/1415213004/diff/1/third_party/libc++-static/R...
File third_party/libc++-static/README.chromium (right):

https://codereview.chromium.org/1415213004/diff/1/third_party/libc++-static/R...
third_party/libc++-static/README.chromium:48: c++ -c -nostdinc++
-I../../libcxx/include/ -I../../libcxxabi/include -O3 -std=c++11
-fstrict-aliasing ../../libcxxabi/src/*.cpp -fvisibility=hidden
-fvisibility-inlines-hidden -mmacosx-version-min=10.6 -arch i386 -arch x86_64
Nico wrote:
> On 2015/10/22 21:54:27, Mark Mentovai wrote:
> > Where do these object files go?
> 
> Note the subtle first * in `libcxx*/*.o` in the libtool line below :-)

DOH!

[Mark Mentovai, 2015-10-22 22:59:43]

Also, do things get too hefty if you build with -g? Source file and line
information might be handy.

[Nico, 2015-10-22 23:12:43]

On 2015/10/22 22:59:43, Mark Mentovai wrote:
> Also, do things get too hefty if you build with -g? Source file and line
> information might be handy.

It makes the library 19MB instead of 3.1MB. I haven't measured how much it slows
down linking. Not sure if that's worth it? 19MB isn't that huge, but line
numbers in libc++ also don't seem that useful maybe.

[Mark Mentovai, 2015-10-22 23:16:53]

Maybe not worthwhile, then. We wouldn’t have a .dSYM for the system libc++,
either.

[Nico, 2015-10-22 23:20:33]

Now with script.

Still uses c++, it doesn't build with our clang (!) I'll investigate what's up
there.

[Mark Mentovai, 2015-10-22 23:26:46]

LGTM otherwise. I’d prefer it if we could build with our own clang because
c’mon, really, but I’m sure you’ll figure out what’s wrong.

https://codereview.chromium.org/1415213004/diff/40001/third_party/libc++-stat...
File third_party/libc++-static/build.sh (right):

https://codereview.chromium.org/1415213004/diff/40001/third_party/libc++-stat...
third_party/libc++-static/build.sh:39: cat << 'EOF' |
Instead of an inline here-doc, put a .patch in the same directory. Easier to
maintain.

https://codereview.chromium.org/1415213004/diff/40001/third_party/libc++-stat...
third_party/libc++-static/build.sh:65: cp libc++-static.a "${THIS_DIR}"
Is this gitignored?

[Nico, 2015-10-22 23:40:00]

https://codereview.chromium.org/1415213004/diff/40001/third_party/libc++-stat...
File third_party/libc++-static/build.sh (right):

https://codereview.chromium.org/1415213004/diff/40001/third_party/libc++-stat...
third_party/libc++-static/build.sh:39: cat << 'EOF' |
On 2015/10/22 23:26:46, Mark Mentovai wrote:
> Instead of an inline here-doc, put a .patch in the same directory. Easier to
> maintain.

Done.

https://codereview.chromium.org/1415213004/diff/40001/third_party/libc++-stat...
third_party/libc++-static/build.sh:65: cp libc++-static.a "${THIS_DIR}"
On 2015/10/22 23:26:46, Mark Mentovai wrote:
> Is this gitignored?

Done.

[Nico, 2015-10-22 23:43:59]

Description was changed from

==========
mac: Pull a prebuilt libc++-static.a.

Not used for anything yet.

BUG=400091
==========

to

==========
mac: Pull a prebuilt libc++-static.a.

Not used for anything yet.

BUG=400091
TBR=torne
==========

[Nico, 2015-10-22 23:44:21]

thakis@chromium.org changed reviewers:
+ torne@chromium.org

[Nico, 2015-10-22 23:44:21]

tbr torne for the webview_licenses.py spelling fix

[Nico, 2015-10-22 23:44:41]

The CQ bit was checked by thakis@chromium.org

[Nico, 2015-10-22 23:44:42]

The patchset sent to the CQ was uploaded after l-g-t-m from mark@chromium.org
Link to the patchset: https://codereview.chromium.org/1415213004/#ps80001
(title: "sha")

[commit-bot: I haz the power, 2015-10-22 23:45:17]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1415213004/80001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1415213004/80001

[commit-bot: I haz the power, 2015-10-23 00:12:39]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-10-23 00:12:40]

Try jobs failed on following builders:
  cast_shell_linux on tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/cast_shell_linu...)
  linux_chromium_chromeos_rel_ng on tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_...)

[Nico, 2015-10-23 00:14:01]

The CQ bit was checked by thakis@chromium.org

[commit-bot: I haz the power, 2015-10-23 00:15:01]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1415213004/80001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1415213004/80001

[commit-bot: I haz the power, 2015-10-23 01:01:20]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-10-23 01:01:21]

Try jobs failed on following builders:
  linux_chromium_rel_ng on tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_...)

[Nico, 2015-10-23 02:25:37]

The CQ bit was checked by thakis@chromium.org

[Nico, 2015-10-23 02:25:37]

The patchset sent to the CQ was uploaded after l-g-t-m from mark@chromium.org
Link to the patchset: https://codereview.chromium.org/1415213004/#ps100001
(title: "licenseeeee")

[commit-bot: I haz the power, 2015-10-23 02:25:51]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1415213004/100001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1415213004/100001

[Nico, 2015-10-23 03:23:17]

The CQ bit was checked by thakis@chromium.org

[Nico, 2015-10-23 03:23:17]

The patchset sent to the CQ was uploaded after l-g-t-m from mark@chromium.org
Link to the patchset: https://codereview.chromium.org/1415213004/#ps120001
(title: "licenseeeeeeeeeee")

[commit-bot: I haz the power, 2015-10-23 03:23:43]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1415213004/120001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1415213004/120001

[commit-bot: I haz the power, 2015-10-23 04:33:58]

Patchset 7 (id:??) landed as
https://crrev.com/7ef5508bd7b471d7b4a67886e23feb89a3a13713
Cr-Commit-Position: refs/heads/master@{#355729}

[Nico, 2015-10-23 04:34:21]

Committed patchset #7 (id:120001) manually as
7ef5508bd7b471d7b4a67886e23feb89a3a13713 (presubmit successful).

[Torne, 2015-10-23 10:34:25]

lgtm