[llvm-commits] [llvm] r150197 - in /llvm/trunk: include/llvm/Support/system_error.h lib/Support/DataStream.cpp lib/Support/FileUtilities.cpp lib/Support/MemoryBuffer.cpp lib/Support/PathV2.cpp lib/Support/Unix/PathV2.inc lib/Support/Windows/PathV

[David Blaikie]

On Thu, Feb 9, 2012 at 3:36 PM, Michael Spencer <bigcheesegs at gmail.com> wrote:
> On Thu, Feb 9, 2012 at 11:24 AM, David Blaikie <dblaikie at gmail.com> wrote:
>> Author: dblaikie
>> Date: Thu Feb  9 13:24:12 2012
>> New Revision: 150197
>>
>> URL: http://llvm.org/viewvc/llvm-project?rev=150197&view=rev
>> Log:
>> Change default error_code ctor to a 'named ctor' so it's more self-documenting.
>>
>> Unify default construction of error_code uses on this idiom so that users don't
>> feel compelled to make static globals for naming convenience. (unfortunately I
>> couldn't make the original ctor private as some APIs don't return their result,
>> instead using an out parameter (that makes sense to default construct) - which
>> is a bit of a pity. I did, however, find/fix some cases of unnecessary default
>> construction of error_code before I hit the unfixable cases)
>
> The out parameters are needed for constructors since we can't use
> exceptions :P.

Actually I didn't get far enough to find ctors using this idiom. What
I ran into was the error-providing iterator "increment(errorcode&)"
API.

> I like the rest of it. It's sad that the standard
> doesn't define an automatic conversion between errc::success and
> error_code(0, generic_category()).

Yeah - that is a bit awkward if there's not a nicer way to create a
success system_error in the standard API