Re: Return value of 'times' on MS-Windows is constant and overflows

2014-06-10 Thread Eli Zaretskii
> Date: Tue, 10 Jun 2014 23:54:41 +0100 > From: Pádraig Brady > CC: bug-gnulib@gnu.org > > Pushed the attached, thanks. Great, thanks.

Re: Return value of 'times' on MS-Windows is constant and overflows

2014-06-10 Thread Pádraig Brady
On 06/10/2014 03:43 AM, Eli Zaretskii wrote: >> Date: Tue, 10 Jun 2014 00:42:22 +0100 >> From: Pádraig Brady >> CC: bug-gnulib@gnu.org >> >>> . The value is constant: every call to 'times' within the same >>> process returns the same value. Callers generally expect the >>> value to chan

Re: Return value of 'times' on MS-Windows is constant and overflows

2014-06-09 Thread Eli Zaretskii
> Date: Tue, 10 Jun 2014 00:42:22 +0100 > From: Pádraig Brady > CC: bug-gnulib@gnu.org > > > . The value is constant: every call to 'times' within the same > > process returns the same value. Callers generally expect the > > value to change, since Posix says the value is the elapsed ti

Re: Return value of 'times' on MS-Windows is constant and overflows

2014-06-09 Thread Pádraig Brady
On 06/08/2014 04:23 PM, Eli Zaretskii wrote: > The gnulib implementation of 'times' for MS-Windows uses the process > creation time returned by the GetProcessTimes API to construct the > value that the function should return. > > This has 2 problems: > > . The value is constant: every call to '

Return value of 'times' on MS-Windows is constant and overflows

2014-06-08 Thread Eli Zaretskii
The gnulib implementation of 'times' for MS-Windows uses the process creation time returned by the GetProcessTimes API to construct the value that the function should return. This has 2 problems: . The value is constant: every call to 'times' within the same process returns the same value.