http://bugzilla.gdcproject.org/show_bug.cgi?id=224

--- Comment #1 from Iain Buclaw <ibuclaw@gdcproject.org> ---
This happens when compiling using -fPIC.

-- 
You are receiving this mail because:
You are watching all bug changes.

http://bugzilla.gdcproject.org/show_bug.cgi?id=224

Iain Buclaw <ibuclaw@gdcproject.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Hardware|All                         |x86
                 OS|All                         |Linux

-- 
You are receiving this mail because:
You are watching all bug changes.

http://bugzilla.gdcproject.org/show_bug.cgi?id=224

Johannes Pfau <johannespfau@gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |johannespfau@gmail.com

--- Comment #2 from Johannes Pfau <johannespfau@gmail.com> ---
Strange, I think I've tested 32bit when I initially implemented the automake/libtool stuff. So this is probably a regression in the last 2-3 weeks?

-- 
You are receiving this mail because:
You are watching all bug changes.

http://bugzilla.gdcproject.org/show_bug.cgi?id=224

--- Comment #3 from Iain Buclaw <ibuclaw@gdcproject.org> ---
This is the "reduced" test.
---
import core.time;
import core.sys.posix.sys.time;

struct Clock
{
    static currTime(immutable TimeZone = LocalTime())
    {
        return SysTime(currStdTime);
    }

    static currStdTime()
    {
        return convert!("seconds", "hnsecs")(0);
    }
}

struct SysTime
{
    import std.typecons;

    this(TimeZone) { }
    this(DateTime) { }

    this(Date date, immutable TimeZone tz = null)
    {
        _timezone = tz is null ? LocalTime() : tz;

        try
        {
            immutable adjustedTime = (date - Date()).total!"hnsecs";
            immutable standardTime = _timezone.tzToUTC(adjustedTime);
            this(standardTime);
        }
        catch
            assert(0);
    }

    this(long stdTime, immutable TimeZone tz = null)
    {
        _stdTime = stdTime;
        _timezone = tz is null ? LocalTime() : tz;
    }

    @property stdTime()
    {
        return _stdTime;
    }

    @property dayOfGregorianCal()
    {
        adjTime;

        long hnsecs;
        days = hnsecs;

        return 0;
    }

    Date opCast()
    {
        return Date(dayOfGregorianCal);
    }

    DateTime opCast()() if(is(Unqual))
    {
    }

    string toISOString()
    {
        import std.format;
        try
        {
            auto dateTime = DateTime();
            return format("%s",
                          dateTime.toISOString);
        }
        catch
            assert(0);
    }

    @property adjTime()
    {
        _timezone.utcToTZ(_stdTime);
    }


    long  _stdTime;
    Rebindable!(immutable TimeZone) _timezone;
}

struct Date
{
    short _year;
    ubyte _day;

    this(int, int) { }
    this(int) { }
    @property year() { }
    @property month() { }
    @property day() { }

    @property day()
    {
        return _day;
    }

    Duration opBinary(string op)(Date)
    {
        return dur!"days"(dayOfGregorianCal);
    }

    @property dayOfGregorianCal()
    {
        return 0;
    }
}

struct TimeOfDay
{
    @property second() { }
}

struct DateTime
{
    this(Date) { }
    this(int) { }
    @property date() { }
    @property timeOfDay() { }
    string toISOString() { return ""; }
}


class TimeZone
{
    long utcToTZ(long) const;
    long tzToUTC(long) const;
    this() immutable { }

    Duration utcOffsetAt(long stdTime)
    {
        return dur!"hnsecs"(stdTime);
    }
}

class LocalTime : TimeZone
{
    static immutable(LocalTime) opCall()
    {
        return (cast()&singleton)();
    }

    bool hasDST()
    {
        try
        {
            cast(Date)Clock.currTime;
            auto janOffset = SysTime(Date()).stdTime - SysTime(Date()).stdTime;
            auto julyOffset = SysTime(Date()).stdTime -
SysTime(Date()).stdTime;

            return janOffset != julyOffset;
        }
        catch
            assert(0);
    }

    this() immutable { }

    static localTime = new immutable(LocalTime);
    immutable(LocalTime) singleton()
    {
        synchronized
            return localTime;
    }
}

time_t stdTimeToUnixTime()
{
    return cast(time_t)convert!("hnsecs", "seconds")(0);
}

SysTime DosFileTimeToSysTime()
{
    try
        return SysTime(DateTime());
    catch
        assert(0);
}

-- 
You are receiving this mail because:
You are watching all bug changes.

http://bugzilla.gdcproject.org/show_bug.cgi?id=224

--- Comment #4 from Iain Buclaw <ibuclaw@gdcproject.org> ---
I'll try updating gcc, just incase it's a trunk bug.

-- 
You are receiving this mail because:
You are watching all bug changes.

http://bugzilla.gdcproject.org/show_bug.cgi?id=224

--- Comment #5 from Johannes Pfau <johannespfau@gmail.com> ---
I can reproduce this on 6.1.0 and 5.3.0, so it's probably not a trunk bug.
(unfortunately)

-- 
You are receiving this mail because:
You are watching all bug changes.

http://bugzilla.gdcproject.org/show_bug.cgi?id=224

--- Comment #6 from Iain Buclaw <ibuclaw@gdcproject.org> ---
(In reply to Johannes Pfau from comment #5)
> I can reproduce this on 6.1.0 and 5.3.0, so it's probably not a trunk bug.
> (unfortunately)

OK. It seems like it's related to #211.  There is something not right in the way we set method and lambdas as "inline".  We should probably get rid of this and make way for pragma(inline).

-- 
You are receiving this mail because:
You are watching all bug changes.

http://bugzilla.gdcproject.org/show_bug.cgi?id=224

--- Comment #7 from Iain Buclaw <ibuclaw@gdcproject.org> ---
Nope, it's just enough to not trigger the "reduced", but the larger test
(std.datetime) still fails.

-- 
You are receiving this mail because:
You are watching all bug changes.

http://bugzilla.gdcproject.org/show_bug.cgi?id=224

--- Comment #8 from Iain Buclaw <ibuclaw@gdcproject.org> ---
(In reply to Johannes Pfau from comment #2)
> Strange, I think I've tested 32bit when I initially implemented the automake/libtool stuff. So this is probably a regression in the last 2-3 weeks?

Yes, you are right.  Just running a git bisect between master and 05c618d.

-- 
You are receiving this mail because:
You are watching all bug changes.

http://bugzilla.gdcproject.org/show_bug.cgi?id=224

--- Comment #9 from Iain Buclaw <ibuclaw@gdcproject.org> ---
((86b6f18...)|BISECTING) $ git bisect good
aa70af04b8c407a169fc65b3fde8a6966b134842 is the first bad commit
commit aa70af04b8c407a169fc65b3fde8a6966b134842
Author: Johannes Pfau <johannespfau@gmail.com>
Date:   Mon May 16 10:41:45 2016 +0200

    Fix some more issue #223 cases

:040000 040000 d30e448ff4270f117f7860fdf8a72713503f0e3c 5a6b8f354fef6ad62c5311e4d96455215ecbc0b1 M      gcc



OK, so maybe I was wrong about putting save_expr inside build_boolop. >:-)

-- 
You are receiving this mail because:
You are watching all bug changes.