Fix some compiler errors in exception_handler_unittest by casting like crazy

The unit test added in r1067 fails to compile for me with Ubuntu's GCC 4.6.3. I
get template errors in gtest:

In file included from
../google-breakpad/src/testing/gtest/include/gtest/gtest.h:55:0,
                 from ../google-breakpad/src/breakpad_googletest_includes.h:33,
                 from
../google-breakpad/src/client/linux/handler/exception_handler_unittest.cc:41:
../google-breakpad/src/testing/gtest/include/gtest/internal/gtest-internal.h: In
function ‘testing::internal::String
testing::internal::FormatForFailureMessage(T*) [with T = void(int)]’:
../google-breakpad/src/testing/gtest/include/gtest/gtest.h:1246:39:  
instantiated from ‘testing::internal::String
testing::internal::FormatForComparisonFailureMessage(const T1&, const T2&) [with
T1 = void (*)(int), T2 = void (*)(int)]’
../google-breakpad/src/testing/gtest/include/gtest/gtest.h:1377:1:  
instantiated from ‘testing::AssertionResult testing::internal::CmpHelperNE(const
char*, const char*, const T1&, const T2&) [with T1 = void (*)(int), T2 = void
(*)(int)]’
../google-breakpad/src/client/linux/handler/exception_handler_unittest.cc:341:5:
  instantiated from here
../google-breakpad/src/testing/gtest/include/gtest/internal/gtest-internal.h:238:62:
error: invalid static_cast from type ‘void (*)(int)’ to type ‘const void*’

Adding copious casts fixes the issue for me.

[jimb_red-bean.com]

Yeah, those SIG_ constants have stunk from the day they were introduced... :(

[digit]

lgtm (not owner)

http://breakpad.appspot.com/483002/diff/1/src/client/linux/handler/exception_...
File src/client/linux/handler/exception_handler_unittest.cc (right):

http://breakpad.appspot.com/483002/diff/1/src/client/linux/handler/exception_...
src/client/linux/handler/exception_handler_unittest.cc:344:
reinterpret_cast<void*>(SIG_ERR));
Interesting, I don't get these compiler errors with the GCC 4.6.3 compiler on
Ubuntu Precise, but I agree these casts are a good thing.

[Ted Mielczarek]

On Mon, Oct 15, 2012 at 7:47 AM,  <digit@chromium.org> wrote:
>
http://breakpad.appspot.com/483002/diff/1/src/client/linux/handler/exception_...
> src/client/linux/handler/exception_handler_unittest.cc:344:
> reinterpret_cast<void*>(SIG_ERR));
> Interesting, I don't get these compiler errors with the GCC 4.6.3
> compiler on Ubuntu Precise, but I agree these casts are a good thing.

Might be something specific to my build environment. Are you building
a 64-bit or 32-bit build? Either way, these casts shouldn't hurt
anything.

-Ted

[digit]

It's a 64-bit Ubuntu 12.04 install, gcc -v gives:

  ...
  gcc version 4.6.3 (Ubuntu/Linaro 4.6.3-1ubuntu5)

Also, the Android NDK toolchain that I use is based on GCC 4.6 too:

  gcc version 4.6.x-google 20120106 (prerelease) (GCC)

So, apparently an older version.
 
I also tried to run android/run-checks.sh --enable-w32 --all-tests --verbose and
the host test suite builds and runs appropriately, as well as the Android ARM
one.

I suspect this may be due to different definitions of SIG_ERR in the GLibc
headers. Mine are:

/usr/include/x86_64-linux-gnu/bits/signum.h:
  #define SIG_ERR	((__sighandler_t) -1)

/usr/include/i386-linux-gnu/bits/signum.h:
  #define SIG_ERR	((__sighandler_t) -1

[Ted Mielczarek]

On Mon, Oct 15, 2012 at 11:31 AM,  <digit@chromium.org> wrote:
> It's a 64-bit Ubuntu 12.04 install, gcc -v gives:
>
>   ...
>   gcc version 4.6.3 (Ubuntu/Linaro 4.6.3-1ubuntu5)

I have the exact same version (unsurprisingly).

> I suspect this may be due to different definitions of SIG_ERR in the
> GLibc headers. Mine are:
>
> /usr/include/x86_64-linux-gnu/bits/signum.h:
>   #define SIG_ERR       ((__sighandler_t) -1)

That matches mine exactly as well. I have no idea why I was hitting
this and you weren't.

-Ted