Convert COMPILE_ASSERT to static_assert

Convert COMPILE_ASSERT to static_assert

(Perhaps I should have just left it in mini_chromium, but anyway.)

R=mark@chromium.org
BUG=crashpad:1
Committed: https://chromium.googlesource.com/crashpad/crashpad/+/d198c50abe9bf2132d2e65f814cf9d0d48e49cc0

[Mark Mentovai, 2014-10-01 03:55:13]

util/stdlib/strlcpy_test.cc also needs #include "base/basictypes.h" for its use
of arraysize. It was previously picking it up by chance.

You should have a Mac to test these sorts of changes on (and slash or we should
get try and buildbots set up).

https://codereview.chromium.org/615923004/diff/1/client/simple_string_diction...
File client/simple_string_dictionary.cc (right):

https://codereview.chromium.org/615923004/diff/1/client/simple_string_diction...
client/simple_string_dictionary.cc:17: #include "base/basictypes.h"
No longer needed.

https://codereview.chromium.org/615923004/diff/1/util/file/file_writer.cc
File util/file/file_writer.cc (right):

https://codereview.chromium.org/615923004/diff/1/util/file/file_writer.cc#new...
util/file/file_writer.cc:29: static_assert(offsetof(WritableIoVec, iov_base) ==
offsetof(iovec, "iov base"),
You changed the wrong parameter.

../../util/file/file_writer.cc:29:68: error: expected identifier
static_assert(offsetof(WritableIoVec, iov_base) == offsetof(iovec, "iov base"),
                                                                   ^
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/../lib/clang/6.0/include/stddef.h:87:46:
note: expanded from macro 'offsetof'
#define offsetof(t, d) __builtin_offsetof(t, d)

Same on line 31.

https://codereview.chromium.org/615923004/diff/1/util/file/string_file_writer...
File util/file/string_file_writer_test.cc (right):

https://codereview.chromium.org/615923004/diff/1/util/file/string_file_writer...
util/file/string_file_writer_test.cc:358: "three must be invalid for whence");
This can be a numeral 3 now, since there’s no restriction on starting a string
literal with a number like there is on starting an identifier with a number.

https://codereview.chromium.org/615923004/diff/1/util/mac/mach_o_image_reader.cc
File util/mac/mach_o_image_reader.cc (right):

https://codereview.chromium.org/615923004/diff/1/util/mac/mach_o_image_reader...
util/mac/mach_o_image_reader.cc:331: static_assert(NO_SECT == 0, "no sect must
be zero");
"NO_SECT must be 0"

https://codereview.chromium.org/615923004/diff/1/util/mach/exc_server_variant...
File util/mach/exc_server_variants_test.cc (right):

https://codereview.chromium.org/615923004/diff/1/util/mach/exc_server_variant...
util/mach/exc_server_variants_test.cc:1117: static assert(arraysize(kTestData)
>= 1, "must have something to test");
Missing underscore: static_assert. Same as in uuid.cc.

https://codereview.chromium.org/615923004/diff/1/util/misc/uuid.cc
File util/misc/uuid.cc (right):

https://codereview.chromium.org/615923004/diff/1/util/misc/uuid.cc#newcode30
util/misc/uuid.cc:30: static assert(sizeof(UUID) == 16, "UUID must be 16
bytes");
Missing underscore.

../../util/misc/uuid.cc:30:35: error: too many arguments provided to
function-like macro invocation
static assert(sizeof(UUID) == 16, "UUID must be 16 bytes");
                                  ^
/usr/include/assert.h:56:9: note: macro 'assert' defined here
#define assert(e)       ((void)0)
        ^
../../util/misc/uuid.cc:30:8: error: C++ requires a type specifier for all
declarations
static assert(sizeof(UUID) == 16, "UUID must be 16 bytes");
~~~~~~ ^

https://codereview.chromium.org/615923004/diff/1/util/numeric/int128.h
File util/numeric/int128.h (right):

https://codereview.chromium.org/615923004/diff/1/util/numeric/int128.h#newcode20
util/numeric/int128.h:20: #include "base/basictypes.h"
No longer necessary.

https://codereview.chromium.org/615923004/diff/1/util/stdlib/string_number_co...
File util/stdlib/string_number_conversion.cc (right):

https://codereview.chromium.org/615923004/diff/1/util/stdlib/string_number_co...
util/stdlib/string_number_conversion.cc:24: #include "base/basictypes.h"
Not needed.

https://codereview.chromium.org/615923004/diff/1/util/stdlib/string_number_co...
util/stdlib/string_number_conversion.cc:34: #define
CONSTEXPR_COMPILE_ASSERT(condition, message) \
Probably you should rename this to CONSTEXPR_STATIC_ASSERT.

https://codereview.chromium.org/615923004/diff/1/util/stdlib/string_number_co...
util/stdlib/string_number_conversion.cc:37: #define
CONSTEXPR_COMPILE_ASSERT(condition, message) CHECK(condition)
You can make this CHECK(condition) << message, nuke the # from line 35, and pass
strings to this macro so it’s more of a drop-in for static_assert.

[scottmg, 2014-10-01 04:03:40]

Yeesh, sorry. I'll set up a Mac before I send any more.

Thanks for the careful review.

[scottmg, 2014-10-01 16:37:19]

https://codereview.chromium.org/615923004/diff/1/client/simple_string_diction...
File client/simple_string_dictionary.cc (right):

https://codereview.chromium.org/615923004/diff/1/client/simple_string_diction...
client/simple_string_dictionary.cc:17: #include "base/basictypes.h"
On 2014/10/01 03:55:12, Mark Mentovai wrote:
> No longer needed.

Done.

https://codereview.chromium.org/615923004/diff/1/util/file/file_writer.cc
File util/file/file_writer.cc (right):

https://codereview.chromium.org/615923004/diff/1/util/file/file_writer.cc#new...
util/file/file_writer.cc:29: static_assert(offsetof(WritableIoVec, iov_base) ==
offsetof(iovec, "iov base"),
On 2014/10/01 03:55:12, Mark Mentovai wrote:
> You changed the wrong parameter.
> 
> ../../util/file/file_writer.cc:29:68: error: expected identifier
> static_assert(offsetof(WritableIoVec, iov_base) == offsetof(iovec, "iov
base"),
>                                                                    ^
>
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/../lib/clang/6.0/include/stddef.h:87:46:
> note: expanded from macro 'offsetof'
> #define offsetof(t, d) __builtin_offsetof(t, d)
> 
> Same on line 31.

Done.

https://codereview.chromium.org/615923004/diff/1/util/file/string_file_writer...
File util/file/string_file_writer_test.cc (right):

https://codereview.chromium.org/615923004/diff/1/util/file/string_file_writer...
util/file/string_file_writer_test.cc:358: "three must be invalid for whence");
On 2014/10/01 03:55:12, Mark Mentovai wrote:
> This can be a numeral 3 now, since there’s no restriction on starting a string
> literal with a number like there is on starting an identifier with a number.

Done.

https://codereview.chromium.org/615923004/diff/1/util/mac/mach_o_image_reader.cc
File util/mac/mach_o_image_reader.cc (right):

https://codereview.chromium.org/615923004/diff/1/util/mac/mach_o_image_reader...
util/mac/mach_o_image_reader.cc:331: static_assert(NO_SECT == 0, "no sect must
be zero");
On 2014/10/01 03:55:12, Mark Mentovai wrote:
> "NO_SECT must be 0"

Done.

https://codereview.chromium.org/615923004/diff/1/util/mach/exc_server_variant...
File util/mach/exc_server_variants_test.cc (right):

https://codereview.chromium.org/615923004/diff/1/util/mach/exc_server_variant...
util/mach/exc_server_variants_test.cc:1117: static assert(arraysize(kTestData)
>= 1, "must have something to test");
On 2014/10/01 03:55:12, Mark Mentovai wrote:
> Missing underscore: static_assert. Same as in uuid.cc.

Done.

https://codereview.chromium.org/615923004/diff/1/util/misc/uuid.cc
File util/misc/uuid.cc (right):

https://codereview.chromium.org/615923004/diff/1/util/misc/uuid.cc#newcode30
util/misc/uuid.cc:30: static assert(sizeof(UUID) == 16, "UUID must be 16
bytes");
On 2014/10/01 03:55:12, Mark Mentovai wrote:
> Missing underscore.
> 
> ../../util/misc/uuid.cc:30:35: error: too many arguments provided to
> function-like macro invocation
> static assert(sizeof(UUID) == 16, "UUID must be 16 bytes");
>                                   ^
> /usr/include/assert.h:56:9: note: macro 'assert' defined here
> #define assert(e)       ((void)0)
>         ^
> ../../util/misc/uuid.cc:30:8: error: C++ requires a type specifier for all
> declarations
> static assert(sizeof(UUID) == 16, "UUID must be 16 bytes");
> ~~~~~~ ^

Done.

https://codereview.chromium.org/615923004/diff/1/util/numeric/int128.h
File util/numeric/int128.h (right):

https://codereview.chromium.org/615923004/diff/1/util/numeric/int128.h#newcode20
util/numeric/int128.h:20: #include "base/basictypes.h"
On 2014/10/01 03:55:12, Mark Mentovai wrote:
> No longer necessary.

Done.

https://codereview.chromium.org/615923004/diff/1/util/stdlib/string_number_co...
File util/stdlib/string_number_conversion.cc (right):

https://codereview.chromium.org/615923004/diff/1/util/stdlib/string_number_co...
util/stdlib/string_number_conversion.cc:24: #include "base/basictypes.h"
On 2014/10/01 03:55:12, Mark Mentovai wrote:
> Not needed.

Done.

https://codereview.chromium.org/615923004/diff/1/util/stdlib/string_number_co...
util/stdlib/string_number_conversion.cc:34: #define
CONSTEXPR_COMPILE_ASSERT(condition, message) \
On 2014/10/01 03:55:12, Mark Mentovai wrote:
> Probably you should rename this to CONSTEXPR_STATIC_ASSERT.

Done.

https://codereview.chromium.org/615923004/diff/1/util/stdlib/string_number_co...
util/stdlib/string_number_conversion.cc:37: #define
CONSTEXPR_COMPILE_ASSERT(condition, message) CHECK(condition)
On 2014/10/01 03:55:12, Mark Mentovai wrote:
> You can make this CHECK(condition) << message, nuke the # from line 35, and
pass
> strings to this macro so it’s more of a drop-in for static_assert.

Done.

[Mark Mentovai, 2014-10-01 18:57:56]

LGTM. Thanks!

https://codereview.chromium.org/615923004/diff/1/util/numeric/int128.h
File util/numeric/int128.h (right):

https://codereview.chromium.org/615923004/diff/1/util/numeric/int128.h#newcode20
util/numeric/int128.h:20: #include "base/basictypes.h"
scottmg wrote:
> On 2014/10/01 03:55:12, Mark Mentovai wrote:
> > No longer necessary.
> 
> Done.

Can you add base/basictypes.h to int128_test.cc now? The test uses bit_cast<>.

https://codereview.chromium.org/615923004/diff/40001/minidump/minidump_system...
File minidump/minidump_system_info_writer.cc (right):

https://codereview.chromium.org/615923004/diff/40001/minidump/minidump_system...
minidump/minidump_system_info_writer.cc:51: "vendor id must have 3 elements");
I’d say to call this one VendorId, and the same on line 66. And on line 109,
call it ProcessorFeatures. These ought to match the things they’re testing.

https://codereview.chromium.org/615923004/diff/40001/util/stdlib/string_numbe...
File util/stdlib/string_number_conversion.cc (right):

https://codereview.chromium.org/615923004/diff/40001/util/stdlib/string_numbe...
util/stdlib/string_number_conversion.cc:27: // CONSTEXPR_STATIC_ASSERT will be a
normal static_assert if the C++ library
Nit: this whole paragraph can reflow.

[scottmg, 2014-10-01 19:27:40]

Thanks

https://codereview.chromium.org/615923004/diff/40001/minidump/minidump_system...
File minidump/minidump_system_info_writer.cc (right):

https://codereview.chromium.org/615923004/diff/40001/minidump/minidump_system...
minidump/minidump_system_info_writer.cc:51: "vendor id must have 3 elements");
On 2014/10/01 18:57:55, Mark Mentovai wrote:
> I’d say to call this one VendorId, and the same on line 66. And on line 109,
> call it ProcessorFeatures. These ought to match the things they’re testing.

Done.

https://codereview.chromium.org/615923004/diff/40001/util/stdlib/string_numbe...
File util/stdlib/string_number_conversion.cc (right):

https://codereview.chromium.org/615923004/diff/40001/util/stdlib/string_numbe...
util/stdlib/string_number_conversion.cc:27: // CONSTEXPR_STATIC_ASSERT will be a
normal static_assert if the C++ library
On 2014/10/01 18:57:55, Mark Mentovai wrote:
> Nit: this whole paragraph can reflow.

Done.

[scottmg, 2014-10-01 19:29:05]

Committed patchset #4 (id:60001) manually as
d198c50abe9bf2132d2e65f814cf9d0d48e49cc0 (presubmit successful).