3 Jun
2011
3 Jun
'11
3:55 p.m.
While the paper on Linux's spinlock semantics was very interesting, it remains the
fact that this is not the case in Windows in this particular instance.
A lot of ReactOS code *is* missing calls such as KeMemoryBarrier() and (volatile), and
only works by chance, so the argument that "otherwise our code wouldn't
work" is a bit of a fallacy.
You also need to think outside the strict-ordering x86 box. Most of ReactOS' code is
totally borked on IA64, PPC or ARM (and semi-broken on x64 too, which has looser
ordering).
Of course, feel free to ignore the suggestion.
--
Best regards,
Alex Ionescu
On 2011-06-03, at 8:08 AM, Timo Kreuzer wrote:
Am 03.06.2011 13:24, schrieb Alex Ionescu:
Ah, I didn't see the caller.
There's still the issue of the missing volatile. It's required to make sure there
is strict ordering between the spinlock acquisition and the increment/decrement.
Ordering is guaranteed, since the spinlock functions act as memory barriers,
otherwise a lot of our code wouldn't work.
Suggested reading: http://kernel.org/doc/Documentation/volatile-considered-harmful.txt
Regards,
Timo
_______________________________________________
Ros-dev mailing list
Ros-dev(a)reactos.org
http://www.reactos.org/mailman/listinfo/ros-dev