Non-SFI Mode: Build base/ library by PNaCL toolchain for nacl_helper_nonsfi.

base/base_nacl.gyp

 # Copyright (c) 2012 The Chromium Authors. All rights reserved.
 # Use of this source code is governed by a BSD-style license that can be
 # found in the LICENSE file.
 
 {
   'variables': {
     'chromium_code': 1,
   },
   'includes': [
     '../build/common_untrusted.gypi',
@@ skipping 49 matching lines @@
           },
           'dependencies': [
             '<(DEPTH)/third_party/icu/icu_nacl.gyp:icudata_nacl',
             '<(DEPTH)/third_party/icu/icu_nacl.gyp:icui18n_nacl',
             '<(DEPTH)/third_party/icu/icu_nacl.gyp:icuuc_nacl',
             '<(DEPTH)/native_client/tools.gyp:prep_toolchain',
           ],
         },
       ],
     }],
+    ['disable_nacl==0', {
+      'targets': [
+        {
+          'target_name': 'base_nacl_nonsfi',
+          'type': 'none',
+          'include_dirs': [
+            '<(DEPTH)/native_client/src/public/linux_syscalls',

[Lei Zhang, 2014/10/14 21:32:23]

It's a bit weird for base to depend on things, but I see base_nacl.gyp already
does this.

[hidehiko, 2014/10/15 02:20:55]

The PNaCl toolchain does not provide some header files for syscalls by default,
such as <sys/socket.h>. This is the include path to use it.

[Mark Seaborn, 2014/10/16 23:03:09]

As an aside:  I suppose this could be added to include_dirs automatically by
NaCl's untrusted.gypi (for "build_nonsfi_helper" builds).  Do you think that
would be worthwhile doing?

[hidehiko, 2014/10/17 04:44:57]

I think, yes. Please let me send out a separate CL for it.

+          ],
+          'variables': {
+            'base_target': 1,
+            'nacl_untrusted_build': 1,
+            'nlib_target': 'libbase_nacl_nonsfi.a',
+            'build_glibc': 0,
+            'build_newlib': 0,
+            'build_irt': 0,
+            'build_pnacl_newlib': 0,
+            'build_nonsfi_helper': 1,
+
+            'sources': [
+              'base_switches.cc',
+              'base_switches.h',
+
+              # For PathExists and ReadFromFD.
+              'files/file_util_posix.cc',
+
+              # For MessageLoopForIO based on libevent.
+              'message_loop/message_pump_libevent.cc',
+              'message_loop/message_pump_libevent.h',
+
+              # For UnixDomainSocket::SendMsg and RecvMsg.
+              'posix/unix_domain_socket_linux.cc',
+
+              # For GetKnownDeadTerminationStatus and GetTerminationStatus.
+              'process/kill_posix.cc',
+
+              # Unlike libbase_nacl, for Non-SFI build, we need to use
+              # rand_util_posix for random implementation, intead of

[Lei Zhang, 2014/10/14 21:32:23]

typo: "intead"

[hidehiko, 2014/10/15 02:20:55]

Done.

+              # rand_util_nacl.cc, which is based on IRT. rand_util_nacl.cc is
+              # excluded below.
+              'rand_util_posix.cc',
+
+              # For CancelableSyncSocket.
+              'sync_socket_nacl.cc',
+            ],
+          },
+          'sources!': [
+            'rand_util_nacl.cc',

[Lei Zhang, 2014/10/14 21:32:23]

I don't think you need this because rand_util_nacl.cc is not in sources above.

[hidehiko, 2014/10/15 02:20:55]

rand_util_nacl.cc is actually included in base.gypi, and is not excluded by
suffix rule (described in build/filename_rules.gypi). So here, we need to
exclude it manually.

[Lei Zhang, 2014/10/15 06:04:02]

Ah, I see the "sources" above is a variable. That's rather confusing. I applied
your patch locally and I looked at the native_client/ gyp files. Now I better
understand what's going on and this makes sense.

+          ],
+          'dependencies': [
+            '<(DEPTH)/native_client/tools.gyp:prep_toolchain',
+            '<(DEPTH)/third_party/libevent/libevent_nacl_nonsfi.gyp:event_nacl_nonsfi',
+          ],
+        },
+      ],
+    }],
   ],
 }