Hi,
On Thu, Jan 07, 2016 at 01:49:08AM +0000, Steve McIntyre wrote:
> Hey folks,
>
> I'm seeing this both in the 4.11 tarball (which I was just hoping to
> push into Debian) and wuth current git HEAD. Everything builds fine
> for me on amd64 and tests run for 64-bit fine, but in tests-32 I get a
> single failure for "times":
>
> FAIL: times
> ===========
>
> 1c1
> < times({tms_utime=24, tms_stime=14, tms_cutime=36, tms_cstime=22}) =
> 2246491019
> ---
> > times({tms_utime=24, tms_stime=14, tms_cutime=36, tms_cstime=22}) =
> > 18446744071661075339
> times.test: failed test: ../strace -etimes -esignal=none ./times
> output mismatch
> FAIL times.test (exit status: 1)
>
> The numbers are suspicious to me: 2246491019 == 0x85E6BB8B
> 18446744071661075339 == 0xFFFFFFFF85E6BB8B
>
> Maybe a sign extension problem in the test or something? This is a sign extension bug in strace, thanks for reporting. I've pushed commit v4.11-124-gbe1cb92 that hopefully fixes it without breaking anything else. -- ldv
pgpWPGlNMovfw.pgp
Description: PGP signature
------------------------------------------------------------------------------
_______________________________________________ Strace-devel mailing list Strace-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/strace-devel
