[Windows MSVC CFG] Turning on linker CFG for all. Removing CFG compile.

[Windows MSVC CFG] Turning on linker CFG for all.  Disabling CFG compile.

This enables support for CFG that is already compiled into Microsoft
system DLLs - in any of our processes.

Also disabling compilation of CFG with MSVC compiler.  Preparing for
clang CFI instead.

TEST=CFGSupportTests.MsIndirectFailure in sbox_integration_tests suite.
BUG=584575
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.win:win10_chromium_x64_rel_ng
TBR=thestig or jochen (chrome\BUILD.gn)

Review-Url: https://codereview.chromium.org/2693193005
Cr-Commit-Position: refs/heads/master@{#451116}
Committed: https://chromium.googlesource.com/chromium/src/+/f160102d25fb949a3ed24478268b8835cc086477

[penny, 2017-02-15 00:36:04]

Description was changed from

==========
[Windows MSVC CFG] Turning on linker CFG for all.  Disabling CFG compile.

This enables support for CFG that is already compiled into Microsoft
system DLLs - in any of our processes.

Also disabling compilation of CFG with MSVC compiler.  Preparing for
clang CFI instead.

BUG=584575
==========

to

==========
[Windows MSVC CFG] Turning on linker CFG for all.  Disabling CFG compile.

This enables support for CFG that is already compiled into Microsoft
system DLLs - in any of our processes.

Also disabling compilation of CFG with MSVC compiler.  Preparing for
clang CFI instead.

BUG=584575
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.win:win10_chromium_x64_rel_ng
==========

[penny, 2017-02-15 00:37:17]

Description was changed from

==========
[Windows MSVC CFG] Turning on linker CFG for all.  Disabling CFG compile.

This enables support for CFG that is already compiled into Microsoft
system DLLs - in any of our processes.

Also disabling compilation of CFG with MSVC compiler.  Preparing for
clang CFI instead.

BUG=584575
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.win:win10_chromium_x64_rel_ng
==========

to

==========
[Windows MSVC CFG] Turning on linker CFG for all.  Disabling CFG compile.

This enables support for CFG that is already compiled into Microsoft
system DLLs - in any of our processes.

Also disabling compilation of CFG with MSVC compiler.  Preparing for
clang CFI instead.

BUG=584575
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.win:win10_chromium_x64_rel_ng
==========

[penny, 2017-02-15 00:37:17]

pennymac@chromium.org changed reviewers:
+ wfh@chromium.org

[penny, 2017-02-15 00:43:51]

pennymac@chromium.org changed reviewers:
+ scottmg@chromium.org

[penny, 2017-02-15 00:45:33]

pennymac@chromium.org changed reviewers:
+ thestig@chromium.org

[penny, 2017-02-15 00:53:02]

Hello folks,
Thanks for your help doing a small review.  This CL moves the linker CFG
argument into common_linker_setup so that all our processes make use of CFG
build into Microsoft system binaries.  I've also removed the compilation
support, which currently only impacts chrome_elf.  This is making way for clang
roll out. 

scottmg: build\config\win\BUILD.gn
thestig: chrome\BUILD.gn
wfh: everything else sanity check!

[penny, 2017-02-15 00:53:12]

The CQ bit was checked by pennymac@chromium.org to run a CQ dry run

[penny, 2017-02-15 00:53:53]

Description was changed from

==========
[Windows MSVC CFG] Turning on linker CFG for all.  Disabling CFG compile.

This enables support for CFG that is already compiled into Microsoft
system DLLs - in any of our processes.

Also disabling compilation of CFG with MSVC compiler.  Preparing for
clang CFI instead.

BUG=584575
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.win:win10_chromium_x64_rel_ng
==========

to

==========
[Windows MSVC CFG] Turning on linker CFG for all.  Disabling CFG compile.

This enables support for CFG that is already compiled into Microsoft
system DLLs - in any of our processes.

Also disabling compilation of CFG with MSVC compiler.  Preparing for
clang CFI instead.

TEST=CFGSupportTests.MsIndirectFailure in sbox_integration_tests suite.
BUG=584575
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.win:win10_chromium_x64_rel_ng
==========

[commit-bot: I haz the power, 2017-02-15 00:54:04]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-02-15 03:47:00]

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

[commit-bot: I haz the power, 2017-02-15 03:47:01]

Dry run: This issue passed the CQ dry run.

[scottmg, 2017-02-15 20:42:57]

lgtm

[Will Harris, 2017-02-15 20:49:04]

I think this means we'll only use CFG on LINK.EXE from now on, and not CL?

Before this CL, what was being compiled with CFG on CL? Do we lose any CFG
coverage with this CL?

Is there a roadmap/plan for the future of CFG vs clang-cfi?

[penny, 2017-02-15 21:44:39]

On 2017/02/15 20:49:04, Will Harris wrote:
> I think this means we'll only use CFG on LINK.EXE from now on, and not CL?
> 
> Before this CL, what was being compiled with CFG on CL? Do we lose any CFG
> coverage with this CL?
> 
> Is there a roadmap/plan for the future of CFG vs clang-cfi?

As seen in my first message:
"I've also removed the compilation support, which currently only impacts
chrome_elf."

Yes, this support as seen in build\config\win\BUILD.gn is not programmatically
forced on only link.exe - but of course only link.exe would work with this
cmdline linker argument for now.  Which is fine for the up-coming rollout of
clang as a *compiler*.

Once lexan starts to work on the linker switch out, they'll have to update clang
support for this. (And I'll be supporting them on this.)  I'm happy to report
that if they "forget" to do this, the new cfg test case added in
https://codereview.chromium.org/2679793002/ would break.  :)

Let me know if you have any other questions.
Thanks!

[Will Harris, 2017-02-15 21:46:20]

Is removing compiler-time CFG from chrome_elf an acceptable security regression?

[penny, 2017-02-15 21:51:34]

On 2017/02/15 21:46:20, Will Harris wrote:
> Is removing compiler-time CFG from chrome_elf an acceptable security
regression?

My assessment of this was "yes".  It was in fact only a very small portion of
"chrome_elf".  Only the (I believe only one) file compiled under the gn
"chrome_elf" project that had cfg compiled in.  Any dependency projects were NOT
previously opted in for cfg compilation.

[penny, 2017-02-15 21:52:50]

On 2017/02/15 21:51:34, penny wrote:
> On 2017/02/15 21:46:20, Will Harris wrote:
> > Is removing compiler-time CFG from chrome_elf an acceptable security
> regression?
> 
> My assessment of this was "yes".  It was in fact only a very small portion of
> "chrome_elf".  Only the (I believe only one) file compiled under the gn
> "chrome_elf" project that had cfg compiled in.  Any dependency projects were
NOT
> previously opted in for cfg compilation.

Yup, chrome_elf_main.cc.

[Will Harris, 2017-02-15 21:53:28]

lgtm

[penny, 2017-02-15 21:54:48]

On 2017/02/15 21:52:50, penny wrote:
> On 2017/02/15 21:51:34, penny wrote:
> > On 2017/02/15 21:46:20, Will Harris wrote:
> > > Is removing compiler-time CFG from chrome_elf an acceptable security
> > regression?
> > 
> > My assessment of this was "yes".  It was in fact only a very small portion
of
> > "chrome_elf".  Only the (I believe only one) file compiled under the gn
> > "chrome_elf" project that had cfg compiled in.  Any dependency projects were
> NOT
> > previously opted in for cfg compilation.
> 
> Yup, chrome_elf_main.cc.

Sadly, I am halting a fair bit of work I had already done (but not landed) to
compile cfg into most of our chrome code (with updates to our sandbox hooking
infrastructure).  At this point, it's worth holding off until CFI can be tackled
completely differently.

[penny, 2017-02-15 22:00:31]

pennymac@chromium.org changed reviewers:
+ jochen@chromium.org
- thestig@chromium.org

[penny, 2017-02-15 22:02:27]

jochen, could I get you to review a tiny change to src\chrome\BUILD.gn please? 
thestig appears to be ooo.

It's tiny.  Ping me if you have any questions.  Thanks!

[penny, 2017-02-16 20:04:27]

Description was changed from

==========
[Windows MSVC CFG] Turning on linker CFG for all.  Disabling CFG compile.

This enables support for CFG that is already compiled into Microsoft
system DLLs - in any of our processes.

Also disabling compilation of CFG with MSVC compiler.  Preparing for
clang CFI instead.

TEST=CFGSupportTests.MsIndirectFailure in sbox_integration_tests suite.
BUG=584575
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.win:win10_chromium_x64_rel_ng
==========

to

==========
[Windows MSVC CFG] Turning on linker CFG for all.  Disabling CFG compile.

This enables support for CFG that is already compiled into Microsoft
system DLLs - in any of our processes.

Also disabling compilation of CFG with MSVC compiler.  Preparing for
clang CFI instead.

TEST=CFGSupportTests.MsIndirectFailure in sbox_integration_tests suite.
BUG=584575
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.win:win10_chromium_x64_rel_ng
TBR=thestig@chromium.org,jochen@chromium.org (chrome\BUILD.gn)
==========

[penny, 2017-02-16 20:05:24]

Description was changed from

==========
[Windows MSVC CFG] Turning on linker CFG for all.  Disabling CFG compile.

This enables support for CFG that is already compiled into Microsoft
system DLLs - in any of our processes.

Also disabling compilation of CFG with MSVC compiler.  Preparing for
clang CFI instead.

TEST=CFGSupportTests.MsIndirectFailure in sbox_integration_tests suite.
BUG=584575
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.win:win10_chromium_x64_rel_ng
TBR=thestig@chromium.org,jochen@chromium.org (chrome\BUILD.gn)
==========

to

==========
[Windows MSVC CFG] Turning on linker CFG for all.  Disabling CFG compile.

This enables support for CFG that is already compiled into Microsoft
system DLLs - in any of our processes.

Also disabling compilation of CFG with MSVC compiler.  Preparing for
clang CFI instead.

TEST=CFGSupportTests.MsIndirectFailure in sbox_integration_tests suite.
BUG=584575
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.win:win10_chromium_x64_rel_ng
TBR=thestig or jochen (chrome\BUILD.gn)
==========

[penny, 2017-02-16 20:05:44]

The CQ bit was checked by pennymac@chromium.org

[commit-bot: I haz the power, 2017-02-16 20:07:06]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[jochen (gone - plz use gerrit), 2017-02-16 20:08:34]

lgtm

[commit-bot: I haz the power, 2017-02-16 22:21:20]

CQ is committing da patch.

Bot data: {"patchset_id": 1, "attempt_start_ts": 1487275544087020, "parent_rev":
"f1716821e93c7d19a81a0c22f9cbd03fdb3e037d", "commit_rev":
"f160102d25fb949a3ed24478268b8835cc086477"}

[commit-bot: I haz the power, 2017-02-16 22:22:21]

Description was changed from

==========
[Windows MSVC CFG] Turning on linker CFG for all.  Disabling CFG compile.

This enables support for CFG that is already compiled into Microsoft
system DLLs - in any of our processes.

Also disabling compilation of CFG with MSVC compiler.  Preparing for
clang CFI instead.

TEST=CFGSupportTests.MsIndirectFailure in sbox_integration_tests suite.
BUG=584575
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.win:win10_chromium_x64_rel_ng
TBR=thestig or jochen (chrome\BUILD.gn)
==========

to

==========
[Windows MSVC CFG] Turning on linker CFG for all.  Disabling CFG compile.

This enables support for CFG that is already compiled into Microsoft
system DLLs - in any of our processes.

Also disabling compilation of CFG with MSVC compiler.  Preparing for
clang CFI instead.

TEST=CFGSupportTests.MsIndirectFailure in sbox_integration_tests suite.
BUG=584575
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.win:win10_chromium_x64_rel_ng
TBR=thestig or jochen (chrome\BUILD.gn)

Review-Url: https://codereview.chromium.org/2693193005
Cr-Commit-Position: refs/heads/master@{#451116}
Committed:
https://chromium.googlesource.com/chromium/src/+/f160102d25fb949a3ed24478268b...
==========

[commit-bot: I haz the power, 2017-02-16 22:22:22]

Committed patchset #1 (id:1) as
https://chromium.googlesource.com/chromium/src/+/f160102d25fb949a3ed24478268b...