Mail Archives: cygwin-developers/2001/11/18/09:54:58
Hi!
Sunday, 18 November, 2001 Robert Collins robert DOT collins AT itdomain DOT com DOT au wrote:
>> >> the problem remains that the patched code still won't work on w95,
>> >> but the current code doesn't work at all!
>>
>> RC> The current code should run fine on w95 and 2k - it did for me
>> RC> before committing it 0 I tested on both 9x and 2k.
>>
>> RC> However the bug is likely the cause for Jasons continued python
>> RC> failures.
>>
>> really? have you checked if exactly one thread wakes up when condition
>> variable is signaled? i _can't_ see how current code can work at all.
>> please run winsup.api/pthread/condvar3_1.c test on w2k and tell what
>> the result is. btw, are you sure you've been testing on stock w95? or
>> it was w98? or osr2? msdn says that InterlockedIncrement() behavior
>> changed from w95 to w98 and from nt3.5 to nt4.0.
>>
>> again, i can't see how current code can work at all either on w95 or
>> on w98 or on nt or on w2k. what exactly have you tested?
RC> 95 OSR 2 and W2k SP2. Why can't it work? I haven't checked that _only_
RC> one thread wakes up no,
because, according to msdn, on w95 the value InterlockedIncrement
returns is:
==================================================================
Windows 95, Windows NT 3.51 and earlier: If the result of the
operation is zero, the return value is zero. If the result of the
operation is less than zero, the return value is negative, but it is
not necessarily equal to the result. If the result of the operation is
greater than zero, the return value is positive, but it is not
necessarily equal to the result.
==================================================================
osr2 is possibly different from w95 in this regard.
RC> but the woken threads are serialised regardless.
i'm not sure i understand what you mean "serialized" in this
context.
Egor. mailto:deo AT logos-m DOT ru ICQ 5165414 FidoNet 2:5020/496.19
- Raw text -