Re: [HACKERS] MinGW versus _strtoui64() ?

2016-03-12 Thread Tom Lane
Petr Jelinek writes: > On 13/03/16 04:24, Tom Lane wrote: >> So you're thinking "#ifdef _MSC_VER"? Or something else? > Sorry for brevity, yes, that should work, afaics mingw has strtoul so > the code should compile fine with that ifdef. Also just checked to make > sure,

Re: [HACKERS] MinGW versus _strtoui64() ?

2016-03-12 Thread Petr Jelinek
On 13/03/16 04:24, Tom Lane wrote: Petr Jelinek writes: On 13/03/16 03:30, Tom Lane wrote: Per a comment from Petr Jelinek, I added this in commit 23a27b039d94ba35: #ifdef WIN32 return _strtoui64(str, endptr, base); #else ... Several of the Windows buildfarm members

Re: [HACKERS] MinGW versus _strtoui64() ?

2016-03-12 Thread Tom Lane
Petr Jelinek writes: > On 13/03/16 03:30, Tom Lane wrote: >> Per a comment from Petr Jelinek, I added this in commit 23a27b039d94ba35: >> >> #ifdef WIN32 >> return _strtoui64(str, endptr, base); >> #else ... >> >> Several of the Windows buildfarm members are good with

Re: [HACKERS] MinGW versus _strtoui64() ?

2016-03-12 Thread Petr Jelinek
On 13/03/16 03:30, Tom Lane wrote: Per a comment from Petr Jelinek, I added this in commit 23a27b039d94ba35: #ifdef WIN32 return _strtoui64(str, endptr, base); #else ... Several of the Windows buildfarm members are good with that, but narwhal is not: numutils.c: In function