Fix handling of relocation names, so that prefix mangling works.

Fix handling of relocation names, so that prefix mangling works.

Fixes bug in the representation of relocation names (in either global
initializers or as constant expressions in code) so that they
understand when the name is externally defined. This allows us to test
this property using command line arguments, and fixes relocation tests
in cross compilations (where externnally referenced names shouldn't be
name mangled).

BUG=
R=jvoung@chromium.org, stichnot@chromium.org

Committed: https://gerrit.chromium.org/gerrit/gitweb?p=native_client/pnacl-subzero.git;a=commit;h=df6f9d18fce6750872dba5686ed6032949ca01c6

[Karl, 2014-10-20 17:08:42]

kschimpf@google.com changed reviewers:
+ jvoung@chromium.org, stichnot@chromium.org

[Jim Stichnoth, 2014-10-20 17:45:59]

LGTM.

Fix spelling of "externally" in the commit message.

https://codereview.chromium.org/667763002/diff/20001/src/IceGlobalInits.h
File src/IceGlobalInits.h (right):

https://codereview.chromium.org/667763002/diff/20001/src/IceGlobalInits.h#new...
src/IceGlobalInits.h:64: dump(nullptr, Stream);
GlobalContext *const Ctx = nullptr;
dump(Ctx, Stream);

makes it clearer what the nullptr is.

[jvoung (off chromium), 2014-10-20 18:10:05]

LGTM too

https://codereview.chromium.org/667763002/diff/20001/tests_lit/reader_tests/e...
File tests_lit/reader_tests/extern_globals.ll (right):

https://codereview.chromium.org/667763002/diff/20001/tests_lit/reader_tests/e...
tests_lit/reader_tests/extern_globals.ll:6: ; RUN: %lc2i -i %s --insts --args
--allow-uninitialized-globals | FileCheck %s
Add a comment on why lc2i and not p2i?

https://codereview.chromium.org/667763002/diff/20001/tests_lit/reader_tests/e...
tests_lit/reader_tests/extern_globals.ll:10: ; ModuleID =
'/tmp/tmpmLtol6/test_global.ll'
Could probably remove the ModuleID? I guess it depends on how close you want
this to be to the original source.

[Karl, 2014-10-20 21:09:17]

Committed patchset #3 (id:40001) manually as df6f9d1 (presubmit successful).

[Karl, 2014-10-20 21:09:33]

https://codereview.chromium.org/667763002/diff/20001/src/IceGlobalInits.h
File src/IceGlobalInits.h (right):

https://codereview.chromium.org/667763002/diff/20001/src/IceGlobalInits.h#new...
src/IceGlobalInits.h:64: dump(nullptr, Stream);
On 2014/10/20 17:45:59, stichnot wrote:
> GlobalContext *const Ctx = nullptr;
> dump(Ctx, Stream);
> 
> makes it clearer what the nullptr is.

Done.

https://codereview.chromium.org/667763002/diff/20001/tests_lit/reader_tests/e...
File tests_lit/reader_tests/extern_globals.ll (right):

https://codereview.chromium.org/667763002/diff/20001/tests_lit/reader_tests/e...
tests_lit/reader_tests/extern_globals.ll:6: ; RUN: %lc2i -i %s --insts --args
--allow-uninitialized-globals | FileCheck %s
On 2014/10/20 18:10:05, jvoung wrote:
> Add a comment on why lc2i and not p2i?

Done.

https://codereview.chromium.org/667763002/diff/20001/tests_lit/reader_tests/e...
tests_lit/reader_tests/extern_globals.ll:10: ; ModuleID =
'/tmp/tmpmLtol6/test_global.ll'
On 2014/10/20 18:10:04, jvoung wrote:
> Could probably remove the ModuleID? I guess it depends on how close you want
> this to be to the original source.

Done.